diff --git "a/trainer_state.json" "b/trainer_state.json"
--- "a/trainer_state.json"
+++ "b/trainer_state.json"
@@ -1,172692 +1,196148 @@
 {
-  "best_metric": 0.2483615279197693,
-  "best_model_checkpoint": "./checkpoints/ckpt_8_6_r32_claude_all_fp16_lr5e-05_batch8/checkpoint-141500",
-  "epoch": 4.995953980930937,
+  "best_metric": 0.427830308675766,
+  "best_model_checkpoint": "../checkpoints/iccad-contest-results/checkpoint-110000",
+  "epoch": 4.991530615922442,
   "eval_steps": 500,
-  "global_step": 142000,
+  "global_step": 138500,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
     {
-      "epoch": 0.0,
-      "learning_rate": 5.863039399624766e-08,
-      "loss": 0.796,
+      "epoch": 0.0001801996612246369,
+      "grad_norm": 0.279486745595932,
+      "learning_rate": 6.005284650492435e-08,
+      "loss": 1.012,
       "step": 5
     },
     {
-      "epoch": 0.0,
-      "learning_rate": 1.1726078799249532e-07,
-      "loss": 0.7443,
+      "epoch": 0.0003603993224492738,
+      "grad_norm": 0.22080670297145844,
+      "learning_rate": 1.201056930098487e-07,
+      "loss": 0.9384,
       "step": 10
     },
     {
-      "epoch": 0.0,
-      "learning_rate": 1.7589118198874298e-07,
-      "loss": 0.8128,
+      "epoch": 0.0005405989836739107,
+      "grad_norm": 0.16033534705638885,
+      "learning_rate": 1.80158539514773e-07,
+      "loss": 0.993,
       "step": 15
     },
     {
-      "epoch": 0.0,
-      "learning_rate": 2.3452157598499064e-07,
-      "loss": 0.8111,
+      "epoch": 0.0007207986448985476,
+      "grad_norm": 0.21021625399589539,
+      "learning_rate": 2.402113860196974e-07,
+      "loss": 0.9073,
       "step": 20
     },
     {
-      "epoch": 0.0,
-      "learning_rate": 2.931519699812383e-07,
-      "loss": 0.6772,
+      "epoch": 0.0009009983061231845,
+      "grad_norm": 0.1960284262895584,
+      "learning_rate": 3.002642325246217e-07,
+      "loss": 0.9033,
       "step": 25
     },
     {
-      "epoch": 0.0,
-      "learning_rate": 3.5178236397748596e-07,
-      "loss": 0.8012,
+      "epoch": 0.0010811979673478213,
+      "grad_norm": 0.31484127044677734,
+      "learning_rate": 3.60317079029546e-07,
+      "loss": 1.0241,
       "step": 30
     },
     {
-      "epoch": 0.0,
-      "learning_rate": 4.1041275797373365e-07,
-      "loss": 0.7822,
+      "epoch": 0.0012613976285724582,
+      "grad_norm": 0.35444363951683044,
+      "learning_rate": 4.2036992553447037e-07,
+      "loss": 0.9982,
       "step": 35
     },
     {
-      "epoch": 0.0,
-      "learning_rate": 4.690431519699813e-07,
-      "loss": 0.7554,
+      "epoch": 0.0014415972897970951,
+      "grad_norm": 0.25069698691368103,
+      "learning_rate": 4.804227720393948e-07,
+      "loss": 0.9843,
       "step": 40
     },
     {
-      "epoch": 0.0,
-      "learning_rate": 5.276735459662289e-07,
-      "loss": 0.7366,
+      "epoch": 0.001621796951021732,
+      "grad_norm": 0.23986054956912994,
+      "learning_rate": 5.40475618544319e-07,
+      "loss": 0.9416,
       "step": 45
     },
     {
-      "epoch": 0.0,
-      "learning_rate": 5.863039399624765e-07,
-      "loss": 0.7378,
+      "epoch": 0.001801996612246369,
+      "grad_norm": 0.2893970012664795,
+      "learning_rate": 6.005284650492434e-07,
+      "loss": 0.9448,
       "step": 50
     },
     {
-      "epoch": 0.0,
-      "learning_rate": 6.449343339587242e-07,
-      "loss": 0.8267,
+      "epoch": 0.0019821962734710057,
+      "grad_norm": 0.19641688466072083,
+      "learning_rate": 6.605813115541676e-07,
+      "loss": 0.8997,
       "step": 55
     },
     {
-      "epoch": 0.0,
-      "learning_rate": 7.035647279549719e-07,
-      "loss": 0.8,
+      "epoch": 0.0021623959346956426,
+      "grad_norm": 0.2509602904319763,
+      "learning_rate": 7.20634158059092e-07,
+      "loss": 0.9826,
       "step": 60
     },
     {
-      "epoch": 0.0,
-      "learning_rate": 7.621951219512196e-07,
-      "loss": 0.8192,
+      "epoch": 0.0023425955959202795,
+      "grad_norm": 0.25828176736831665,
+      "learning_rate": 7.806870045640163e-07,
+      "loss": 0.9222,
       "step": 65
     },
     {
-      "epoch": 0.0,
-      "learning_rate": 8.208255159474673e-07,
-      "loss": 0.8117,
+      "epoch": 0.0025227952571449165,
+      "grad_norm": 0.19463542103767395,
+      "learning_rate": 8.407398510689407e-07,
+      "loss": 0.8906,
       "step": 70
     },
     {
-      "epoch": 0.0,
-      "learning_rate": 8.794559099437148e-07,
-      "loss": 0.7438,
+      "epoch": 0.0027029949183695534,
+      "grad_norm": 0.1942373514175415,
+      "learning_rate": 9.007926975738651e-07,
+      "loss": 0.9204,
       "step": 75
     },
     {
-      "epoch": 0.0,
-      "learning_rate": 9.380863039399626e-07,
-      "loss": 0.7744,
+      "epoch": 0.0028831945795941903,
+      "grad_norm": 0.2071942389011383,
+      "learning_rate": 9.608455440787895e-07,
+      "loss": 0.9334,
       "step": 80
     },
     {
-      "epoch": 0.0,
-      "learning_rate": 9.9671669793621e-07,
-      "loss": 0.7994,
+      "epoch": 0.0030633942408188272,
+      "grad_norm": 0.2774100601673126,
+      "learning_rate": 1.0208983905837138e-06,
+      "loss": 0.9543,
       "step": 85
     },
     {
-      "epoch": 0.0,
-      "learning_rate": 1.0553470919324578e-06,
-      "loss": 0.7964,
+      "epoch": 0.003243593902043464,
+      "grad_norm": 0.21016931533813477,
+      "learning_rate": 1.080951237088638e-06,
+      "loss": 0.9632,
       "step": 90
     },
     {
-      "epoch": 0.0,
-      "learning_rate": 1.1139774859287056e-06,
-      "loss": 0.7616,
+      "epoch": 0.003423793563268101,
+      "grad_norm": 0.24393318593502045,
+      "learning_rate": 1.1410040835935624e-06,
+      "loss": 0.8989,
       "step": 95
     },
     {
-      "epoch": 0.0,
-      "learning_rate": 1.172607879924953e-06,
-      "loss": 0.7486,
+      "epoch": 0.003603993224492738,
+      "grad_norm": 0.22489190101623535,
+      "learning_rate": 1.2010569300984867e-06,
+      "loss": 0.9678,
       "step": 100
     },
     {
-      "epoch": 0.0,
-      "learning_rate": 1.2312382739212008e-06,
-      "loss": 0.7867,
+      "epoch": 0.003784192885717375,
+      "grad_norm": 0.2544313669204712,
+      "learning_rate": 1.261109776603411e-06,
+      "loss": 0.9295,
       "step": 105
     },
     {
-      "epoch": 0.0,
-      "learning_rate": 1.2898686679174484e-06,
-      "loss": 0.7258,
+      "epoch": 0.003964392546942011,
+      "grad_norm": 0.21903349459171295,
+      "learning_rate": 1.3211626231083353e-06,
+      "loss": 0.9345,
       "step": 110
     },
     {
-      "epoch": 0.0,
-      "learning_rate": 1.348499061913696e-06,
-      "loss": 0.7761,
+      "epoch": 0.004144592208166649,
+      "grad_norm": 0.2654012143611908,
+      "learning_rate": 1.3812154696132596e-06,
+      "loss": 0.9683,
       "step": 115
     },
     {
-      "epoch": 0.0,
-      "learning_rate": 1.4071294559099438e-06,
-      "loss": 0.7119,
+      "epoch": 0.004324791869391285,
+      "grad_norm": 0.25590232014656067,
+      "learning_rate": 1.441268316118184e-06,
+      "loss": 0.9614,
       "step": 120
     },
     {
-      "epoch": 0.0,
-      "learning_rate": 1.4657598499061916e-06,
-      "loss": 0.7707,
+      "epoch": 0.004504991530615923,
+      "grad_norm": 0.25431954860687256,
+      "learning_rate": 1.5013211626231084e-06,
+      "loss": 1.0293,
       "step": 125
     },
     {
-      "epoch": 0.0,
-      "learning_rate": 1.5243902439024391e-06,
-      "loss": 0.7129,
+      "epoch": 0.004685191191840559,
+      "grad_norm": 0.18761098384857178,
+      "learning_rate": 1.5613740091280327e-06,
+      "loss": 0.8426,
       "step": 130
     },
     {
-      "epoch": 0.0,
-      "learning_rate": 1.5830206378986866e-06,
-      "loss": 0.7411,
+      "epoch": 0.004865390853065196,
+      "grad_norm": 0.21217921376228333,
+      "learning_rate": 1.621426855632957e-06,
+      "loss": 0.9276,
       "step": 135
     },
     {
-      "epoch": 0.0,
-      "learning_rate": 1.6416510318949346e-06,
-      "loss": 0.781,
+      "epoch": 0.005045590514289833,
+      "grad_norm": 0.23869900405406952,
+      "learning_rate": 1.6814797021378815e-06,
+      "loss": 0.8987,
       "step": 140
     },
     {
-      "epoch": 0.01,
-      "learning_rate": 1.7002814258911821e-06,
-      "loss": 0.7319,
+      "epoch": 0.00522579017551447,
+      "grad_norm": 0.18920210003852844,
+      "learning_rate": 1.7415325486428058e-06,
+      "loss": 0.9036,
       "step": 145
     },
     {
-      "epoch": 0.01,
-      "learning_rate": 1.7589118198874296e-06,
-      "loss": 0.7559,
+      "epoch": 0.005405989836739107,
+      "grad_norm": 0.2015441209077835,
+      "learning_rate": 1.8015853951477303e-06,
+      "loss": 0.8491,
       "step": 150
     },
     {
-      "epoch": 0.01,
-      "learning_rate": 1.8175422138836774e-06,
-      "loss": 0.8009,
+      "epoch": 0.005586189497963744,
+      "grad_norm": 0.18248215317726135,
+      "learning_rate": 1.8616382416526544e-06,
+      "loss": 0.9322,
       "step": 155
     },
     {
-      "epoch": 0.01,
-      "learning_rate": 1.8761726078799251e-06,
-      "loss": 0.7054,
+      "epoch": 0.005766389159188381,
+      "grad_norm": 0.23022598028182983,
+      "learning_rate": 1.921691088157579e-06,
+      "loss": 0.8488,
       "step": 160
     },
     {
-      "epoch": 0.01,
-      "learning_rate": 1.934803001876173e-06,
-      "loss": 0.6913,
+      "epoch": 0.005946588820413018,
+      "grad_norm": 0.18739941716194153,
+      "learning_rate": 1.981743934662503e-06,
+      "loss": 0.8359,
       "step": 165
     },
     {
-      "epoch": 0.01,
-      "learning_rate": 1.99343339587242e-06,
-      "loss": 0.6887,
+      "epoch": 0.0061267884816376544,
+      "grad_norm": 0.21552830934524536,
+      "learning_rate": 2.0417967811674277e-06,
+      "loss": 0.8829,
       "step": 170
     },
     {
-      "epoch": 0.01,
-      "learning_rate": 2.052063789868668e-06,
-      "loss": 0.6825,
+      "epoch": 0.006306988142862292,
+      "grad_norm": 0.15109090507030487,
+      "learning_rate": 2.101849627672352e-06,
+      "loss": 0.7759,
       "step": 175
     },
     {
-      "epoch": 0.01,
-      "learning_rate": 2.1106941838649157e-06,
-      "loss": 0.7041,
+      "epoch": 0.006487187804086928,
+      "grad_norm": 0.2417735457420349,
+      "learning_rate": 2.161902474177276e-06,
+      "loss": 0.9047,
       "step": 180
     },
     {
-      "epoch": 0.01,
-      "learning_rate": 2.1693245778611634e-06,
-      "loss": 0.686,
+      "epoch": 0.006667387465311566,
+      "grad_norm": 0.23814047873020172,
+      "learning_rate": 2.2219553206822005e-06,
+      "loss": 0.8425,
       "step": 185
     },
     {
-      "epoch": 0.01,
-      "learning_rate": 2.227954971857411e-06,
-      "loss": 0.7251,
+      "epoch": 0.006847587126536202,
+      "grad_norm": 0.23794889450073242,
+      "learning_rate": 2.282008167187125e-06,
+      "loss": 0.7744,
       "step": 190
     },
     {
-      "epoch": 0.01,
-      "learning_rate": 2.2865853658536584e-06,
-      "loss": 0.6662,
+      "epoch": 0.007027786787760839,
+      "grad_norm": 0.18542109429836273,
+      "learning_rate": 2.342061013692049e-06,
+      "loss": 0.7658,
       "step": 195
     },
     {
-      "epoch": 0.01,
-      "learning_rate": 2.345215759849906e-06,
-      "loss": 0.7285,
+      "epoch": 0.007207986448985476,
+      "grad_norm": 0.18955494463443756,
+      "learning_rate": 2.4021138601969734e-06,
+      "loss": 0.8884,
       "step": 200
     },
     {
-      "epoch": 0.01,
-      "learning_rate": 2.403846153846154e-06,
-      "loss": 0.6286,
+      "epoch": 0.0073881861102101124,
+      "grad_norm": 0.17132216691970825,
+      "learning_rate": 2.4621667067018977e-06,
+      "loss": 0.8408,
       "step": 205
     },
     {
-      "epoch": 0.01,
-      "learning_rate": 2.4624765478424017e-06,
-      "loss": 0.6775,
+      "epoch": 0.00756838577143475,
+      "grad_norm": 0.20857086777687073,
+      "learning_rate": 2.522219553206822e-06,
+      "loss": 0.826,
       "step": 210
     },
     {
-      "epoch": 0.01,
-      "learning_rate": 2.5211069418386494e-06,
-      "loss": 0.6458,
+      "epoch": 0.007748585432659386,
+      "grad_norm": 0.23217755556106567,
+      "learning_rate": 2.5822723997117463e-06,
+      "loss": 0.7692,
       "step": 215
     },
     {
-      "epoch": 0.01,
-      "learning_rate": 2.5797373358348967e-06,
-      "loss": 0.6852,
+      "epoch": 0.007928785093884023,
+      "grad_norm": 0.26247748732566833,
+      "learning_rate": 2.6423252462166706e-06,
+      "loss": 0.758,
       "step": 220
     },
     {
-      "epoch": 0.01,
-      "learning_rate": 2.638367729831145e-06,
-      "loss": 0.736,
+      "epoch": 0.008108984755108661,
+      "grad_norm": 0.1590353399515152,
+      "learning_rate": 2.7023780927215953e-06,
+      "loss": 0.7827,
       "step": 225
     },
     {
-      "epoch": 0.01,
-      "learning_rate": 2.696998123827392e-06,
-      "loss": 0.6177,
+      "epoch": 0.008289184416333297,
+      "grad_norm": 0.1754877269268036,
+      "learning_rate": 2.762430939226519e-06,
+      "loss": 0.8534,
       "step": 230
     },
     {
-      "epoch": 0.01,
-      "learning_rate": 2.75562851782364e-06,
-      "loss": 0.6665,
+      "epoch": 0.008469384077557934,
+      "grad_norm": 0.2057289332151413,
+      "learning_rate": 2.822483785731444e-06,
+      "loss": 0.7856,
       "step": 235
     },
     {
-      "epoch": 0.01,
-      "learning_rate": 2.8142589118198877e-06,
-      "loss": 0.6038,
+      "epoch": 0.00864958373878257,
+      "grad_norm": 0.15028229355812073,
+      "learning_rate": 2.882536632236368e-06,
+      "loss": 0.7429,
       "step": 240
     },
     {
-      "epoch": 0.01,
-      "learning_rate": 2.872889305816135e-06,
-      "loss": 0.6792,
+      "epoch": 0.008829783400007209,
+      "grad_norm": 0.20948627591133118,
+      "learning_rate": 2.9425894787412925e-06,
+      "loss": 0.7745,
       "step": 245
     },
     {
-      "epoch": 0.01,
-      "learning_rate": 2.931519699812383e-06,
-      "loss": 0.6346,
+      "epoch": 0.009009983061231845,
+      "grad_norm": 0.2172427922487259,
+      "learning_rate": 3.0026423252462168e-06,
+      "loss": 0.7531,
       "step": 250
     },
     {
-      "epoch": 0.01,
-      "learning_rate": 2.9901500938086305e-06,
-      "loss": 0.5861,
+      "epoch": 0.009190182722456482,
+      "grad_norm": 0.2265809327363968,
+      "learning_rate": 3.062695171751141e-06,
+      "loss": 0.8168,
       "step": 255
     },
     {
-      "epoch": 0.01,
-      "learning_rate": 3.0487804878048782e-06,
-      "loss": 0.6147,
+      "epoch": 0.009370382383681118,
+      "grad_norm": 0.216866135597229,
+      "learning_rate": 3.1227480182560654e-06,
+      "loss": 0.7149,
       "step": 260
     },
     {
-      "epoch": 0.01,
-      "learning_rate": 3.1074108818011255e-06,
-      "loss": 0.5976,
+      "epoch": 0.009550582044905756,
+      "grad_norm": 0.21732446551322937,
+      "learning_rate": 3.1828008647609896e-06,
+      "loss": 0.6542,
       "step": 265
     },
     {
-      "epoch": 0.01,
-      "learning_rate": 3.1660412757973733e-06,
-      "loss": 0.6033,
+      "epoch": 0.009730781706130393,
+      "grad_norm": 0.21542949974536896,
+      "learning_rate": 3.242853711265914e-06,
+      "loss": 0.7051,
       "step": 270
     },
     {
-      "epoch": 0.01,
-      "learning_rate": 3.224671669793621e-06,
-      "loss": 0.6113,
+      "epoch": 0.00991098136735503,
+      "grad_norm": 0.19147822260856628,
+      "learning_rate": 3.3029065577708387e-06,
+      "loss": 0.7055,
       "step": 275
     },
     {
-      "epoch": 0.01,
-      "learning_rate": 3.283302063789869e-06,
-      "loss": 0.5681,
+      "epoch": 0.010091181028579666,
+      "grad_norm": 0.24729080498218536,
+      "learning_rate": 3.362959404275763e-06,
+      "loss": 0.7251,
       "step": 280
     },
     {
-      "epoch": 0.01,
-      "learning_rate": 3.3419324577861165e-06,
-      "loss": 0.5717,
+      "epoch": 0.010271380689804302,
+      "grad_norm": 0.21902547776699066,
+      "learning_rate": 3.423012250780687e-06,
+      "loss": 0.7087,
       "step": 285
     },
     {
-      "epoch": 0.01,
-      "learning_rate": 3.4005628517823642e-06,
-      "loss": 0.5843,
+      "epoch": 0.01045158035102894,
+      "grad_norm": 0.1997164785861969,
+      "learning_rate": 3.4830650972856115e-06,
+      "loss": 0.7187,
       "step": 290
     },
     {
-      "epoch": 0.01,
-      "learning_rate": 3.4591932457786115e-06,
-      "loss": 0.5899,
+      "epoch": 0.010631780012253577,
+      "grad_norm": 0.16925355792045593,
+      "learning_rate": 3.543117943790536e-06,
+      "loss": 0.6483,
       "step": 295
     },
     {
-      "epoch": 0.01,
-      "learning_rate": 3.5178236397748593e-06,
-      "loss": 0.5319,
+      "epoch": 0.010811979673478214,
+      "grad_norm": 0.11117614805698395,
+      "learning_rate": 3.6031707902954605e-06,
+      "loss": 0.6396,
       "step": 300
     },
     {
-      "epoch": 0.01,
-      "learning_rate": 3.5764540337711074e-06,
-      "loss": 0.6013,
+      "epoch": 0.01099217933470285,
+      "grad_norm": 0.15845130383968353,
+      "learning_rate": 3.6632236368003844e-06,
+      "loss": 0.6586,
       "step": 305
     },
     {
-      "epoch": 0.01,
-      "learning_rate": 3.6350844277673548e-06,
-      "loss": 0.6063,
+      "epoch": 0.011172378995927488,
+      "grad_norm": 0.15920360386371613,
+      "learning_rate": 3.7232764833053087e-06,
+      "loss": 0.6414,
       "step": 310
     },
     {
-      "epoch": 0.01,
-      "learning_rate": 3.6937148217636025e-06,
-      "loss": 0.5724,
+      "epoch": 0.011352578657152125,
+      "grad_norm": 0.1967982053756714,
+      "learning_rate": 3.7833293298102334e-06,
+      "loss": 0.6027,
       "step": 315
     },
     {
-      "epoch": 0.01,
-      "learning_rate": 3.7523452157598502e-06,
-      "loss": 0.5849,
+      "epoch": 0.011532778318376761,
+      "grad_norm": 0.1459701806306839,
+      "learning_rate": 3.843382176315158e-06,
+      "loss": 0.6115,
       "step": 320
     },
     {
-      "epoch": 0.01,
-      "learning_rate": 3.8109756097560976e-06,
-      "loss": 0.5275,
+      "epoch": 0.011712977979601398,
+      "grad_norm": 0.1541188508272171,
+      "learning_rate": 3.903435022820082e-06,
+      "loss": 0.6935,
       "step": 325
     },
     {
-      "epoch": 0.01,
-      "learning_rate": 3.869606003752346e-06,
-      "loss": 0.5455,
+      "epoch": 0.011893177640826036,
+      "grad_norm": 0.15788233280181885,
+      "learning_rate": 3.963487869325006e-06,
+      "loss": 0.6226,
       "step": 330
     },
     {
-      "epoch": 0.01,
-      "learning_rate": 3.928236397748593e-06,
-      "loss": 0.518,
+      "epoch": 0.012073377302050672,
+      "grad_norm": 0.15022051334381104,
+      "learning_rate": 4.023540715829931e-06,
+      "loss": 0.6914,
       "step": 335
     },
     {
-      "epoch": 0.01,
-      "learning_rate": 3.98686679174484e-06,
-      "loss": 0.5309,
+      "epoch": 0.012253576963275309,
+      "grad_norm": 0.16778388619422913,
+      "learning_rate": 4.083593562334855e-06,
+      "loss": 0.6599,
       "step": 340
     },
     {
-      "epoch": 0.01,
-      "learning_rate": 4.0454971857410885e-06,
-      "loss": 0.5405,
+      "epoch": 0.012433776624499945,
+      "grad_norm": 0.154036745429039,
+      "learning_rate": 4.143646408839779e-06,
+      "loss": 0.6181,
       "step": 345
     },
     {
-      "epoch": 0.01,
-      "learning_rate": 4.104127579737336e-06,
-      "loss": 0.4894,
+      "epoch": 0.012613976285724584,
+      "grad_norm": 0.15167488157749176,
+      "learning_rate": 4.203699255344704e-06,
+      "loss": 0.6736,
       "step": 350
     },
     {
-      "epoch": 0.01,
-      "learning_rate": 4.162757973733584e-06,
-      "loss": 0.5709,
+      "epoch": 0.01279417594694922,
+      "grad_norm": 0.1559118628501892,
+      "learning_rate": 4.263752101849628e-06,
+      "loss": 0.6087,
       "step": 355
     },
     {
-      "epoch": 0.01,
-      "learning_rate": 4.221388367729831e-06,
-      "loss": 0.5289,
+      "epoch": 0.012974375608173857,
+      "grad_norm": 0.11976916342973709,
+      "learning_rate": 4.323804948354552e-06,
+      "loss": 0.6182,
       "step": 360
     },
     {
-      "epoch": 0.01,
-      "learning_rate": 4.280018761726079e-06,
-      "loss": 0.5217,
+      "epoch": 0.013154575269398493,
+      "grad_norm": 0.12271628528833389,
+      "learning_rate": 4.383857794859476e-06,
+      "loss": 0.594,
       "step": 365
     },
     {
-      "epoch": 0.01,
-      "learning_rate": 4.338649155722327e-06,
-      "loss": 0.5165,
+      "epoch": 0.013334774930623131,
+      "grad_norm": 0.16896136105060577,
+      "learning_rate": 4.443910641364401e-06,
+      "loss": 0.6635,
       "step": 370
     },
     {
-      "epoch": 0.01,
-      "learning_rate": 4.397279549718574e-06,
-      "loss": 0.4879,
+      "epoch": 0.013514974591847768,
+      "grad_norm": 0.1389530748128891,
+      "learning_rate": 4.503963487869326e-06,
+      "loss": 0.62,
       "step": 375
     },
     {
-      "epoch": 0.01,
-      "learning_rate": 4.455909943714822e-06,
-      "loss": 0.5202,
+      "epoch": 0.013695174253072404,
+      "grad_norm": 0.1768377125263214,
+      "learning_rate": 4.56401633437425e-06,
+      "loss": 0.6179,
       "step": 380
     },
     {
-      "epoch": 0.01,
-      "learning_rate": 4.51454033771107e-06,
-      "loss": 0.4872,
+      "epoch": 0.01387537391429704,
+      "grad_norm": 0.12114609777927399,
+      "learning_rate": 4.6240691808791735e-06,
+      "loss": 0.5923,
       "step": 385
     },
     {
-      "epoch": 0.01,
-      "learning_rate": 4.573170731707317e-06,
-      "loss": 0.493,
+      "epoch": 0.014055573575521677,
+      "grad_norm": 0.13337160646915436,
+      "learning_rate": 4.684122027384098e-06,
+      "loss": 0.6363,
       "step": 390
     },
     {
-      "epoch": 0.01,
-      "learning_rate": 4.631801125703565e-06,
-      "loss": 0.4825,
+      "epoch": 0.014235773236746315,
+      "grad_norm": 0.12390133738517761,
+      "learning_rate": 4.744174873889023e-06,
+      "loss": 0.5967,
       "step": 395
     },
     {
-      "epoch": 0.01,
-      "learning_rate": 4.690431519699812e-06,
-      "loss": 0.456,
+      "epoch": 0.014415972897970952,
+      "grad_norm": 0.13579072058200836,
+      "learning_rate": 4.804227720393947e-06,
+      "loss": 0.6168,
       "step": 400
     },
     {
-      "epoch": 0.01,
-      "learning_rate": 4.7490619136960605e-06,
-      "loss": 0.4817,
+      "epoch": 0.014596172559195588,
+      "grad_norm": 0.13865922391414642,
+      "learning_rate": 4.8642805668988715e-06,
+      "loss": 0.6619,
       "step": 405
     },
     {
-      "epoch": 0.01,
-      "learning_rate": 4.807692307692308e-06,
-      "loss": 0.4762,
+      "epoch": 0.014776372220420225,
+      "grad_norm": 0.13145041465759277,
+      "learning_rate": 4.924333413403795e-06,
+      "loss": 0.6132,
       "step": 410
     },
     {
-      "epoch": 0.01,
-      "learning_rate": 4.866322701688555e-06,
-      "loss": 0.4508,
+      "epoch": 0.014956571881644863,
+      "grad_norm": 0.13919544219970703,
+      "learning_rate": 4.98438625990872e-06,
+      "loss": 0.6158,
       "step": 415
     },
     {
-      "epoch": 0.01,
-      "learning_rate": 4.924953095684803e-06,
-      "loss": 0.4522,
+      "epoch": 0.0151367715428695,
+      "grad_norm": 0.14985574781894684,
+      "learning_rate": 5.044439106413644e-06,
+      "loss": 0.6162,
       "step": 420
     },
     {
-      "epoch": 0.01,
-      "learning_rate": 4.983583489681051e-06,
-      "loss": 0.4894,
+      "epoch": 0.015316971204094136,
+      "grad_norm": 0.16457071900367737,
+      "learning_rate": 5.104491952918569e-06,
+      "loss": 0.6272,
       "step": 425
     },
     {
-      "epoch": 0.02,
-      "learning_rate": 5.042213883677299e-06,
-      "loss": 0.4979,
+      "epoch": 0.015497170865318773,
+      "grad_norm": 0.09835664927959442,
+      "learning_rate": 5.164544799423493e-06,
+      "loss": 0.5721,
       "step": 430
     },
     {
-      "epoch": 0.02,
-      "learning_rate": 5.100844277673546e-06,
-      "loss": 0.4948,
+      "epoch": 0.01567737052654341,
+      "grad_norm": 0.12352655827999115,
+      "learning_rate": 5.224597645928417e-06,
+      "loss": 0.5868,
       "step": 435
     },
     {
-      "epoch": 0.02,
-      "learning_rate": 5.1594746716697934e-06,
-      "loss": 0.4799,
+      "epoch": 0.015857570187768046,
+      "grad_norm": 0.11715588718652725,
+      "learning_rate": 5.284650492433341e-06,
+      "loss": 0.5729,
       "step": 440
     },
     {
-      "epoch": 0.02,
-      "learning_rate": 5.218105065666042e-06,
-      "loss": 0.4551,
+      "epoch": 0.016037769848992686,
+      "grad_norm": 0.14185546338558197,
+      "learning_rate": 5.344703338938266e-06,
+      "loss": 0.6224,
       "step": 445
     },
     {
-      "epoch": 0.02,
-      "learning_rate": 5.27673545966229e-06,
-      "loss": 0.4232,
+      "epoch": 0.016217969510217322,
+      "grad_norm": 0.10686635971069336,
+      "learning_rate": 5.404756185443191e-06,
+      "loss": 0.5762,
       "step": 450
     },
     {
-      "epoch": 0.02,
-      "learning_rate": 5.335365853658537e-06,
-      "loss": 0.4446,
+      "epoch": 0.01639816917144196,
+      "grad_norm": 0.1367388218641281,
+      "learning_rate": 5.4648090319481145e-06,
+      "loss": 0.6108,
       "step": 455
     },
     {
-      "epoch": 0.02,
-      "learning_rate": 5.393996247654784e-06,
-      "loss": 0.4542,
+      "epoch": 0.016578368832666595,
+      "grad_norm": 0.1376735419034958,
+      "learning_rate": 5.524861878453038e-06,
+      "loss": 0.6249,
       "step": 460
     },
     {
-      "epoch": 0.02,
-      "learning_rate": 5.452626641651032e-06,
-      "loss": 0.4734,
+      "epoch": 0.01675856849389123,
+      "grad_norm": 0.1213606670498848,
+      "learning_rate": 5.584914724957963e-06,
+      "loss": 0.5733,
       "step": 465
     },
     {
-      "epoch": 0.02,
-      "learning_rate": 5.51125703564728e-06,
-      "loss": 0.4564,
+      "epoch": 0.016938768155115868,
+      "grad_norm": 0.14439453184604645,
+      "learning_rate": 5.644967571462888e-06,
+      "loss": 0.636,
       "step": 470
     },
     {
-      "epoch": 0.02,
-      "learning_rate": 5.569887429643528e-06,
-      "loss": 0.4991,
+      "epoch": 0.017118967816340504,
+      "grad_norm": 0.1540927290916443,
+      "learning_rate": 5.7050204179678125e-06,
+      "loss": 0.5924,
       "step": 475
     },
     {
-      "epoch": 0.02,
-      "learning_rate": 5.628517823639775e-06,
-      "loss": 0.4756,
+      "epoch": 0.01729916747756514,
+      "grad_norm": 0.13569463789463043,
+      "learning_rate": 5.765073264472736e-06,
+      "loss": 0.5908,
       "step": 480
     },
     {
-      "epoch": 0.02,
-      "learning_rate": 5.687148217636023e-06,
-      "loss": 0.474,
+      "epoch": 0.017479367138789777,
+      "grad_norm": 0.13307730853557587,
+      "learning_rate": 5.82512611097766e-06,
+      "loss": 0.5795,
       "step": 485
     },
     {
-      "epoch": 0.02,
-      "learning_rate": 5.74577861163227e-06,
-      "loss": 0.4501,
+      "epoch": 0.017659566800014417,
+      "grad_norm": 0.17746399343013763,
+      "learning_rate": 5.885178957482585e-06,
+      "loss": 0.5938,
       "step": 490
     },
     {
-      "epoch": 0.02,
-      "learning_rate": 5.804409005628518e-06,
-      "loss": 0.4726,
+      "epoch": 0.017839766461239054,
+      "grad_norm": 0.11576178669929504,
+      "learning_rate": 5.94523180398751e-06,
+      "loss": 0.5914,
       "step": 495
     },
     {
-      "epoch": 0.02,
-      "learning_rate": 5.863039399624766e-06,
-      "loss": 0.4845,
+      "epoch": 0.01801996612246369,
+      "grad_norm": 0.12623916566371918,
+      "learning_rate": 6.0052846504924335e-06,
+      "loss": 0.5679,
       "step": 500
     },
     {
-      "epoch": 0.02,
-      "eval_loss": 0.45026421546936035,
-      "eval_runtime": 10.5589,
-      "eval_samples_per_second": 9.471,
-      "eval_steps_per_second": 9.471,
+      "epoch": 0.01801996612246369,
+      "eval_loss": 0.6039146184921265,
+      "eval_runtime": 3.5675,
+      "eval_samples_per_second": 28.031,
+      "eval_steps_per_second": 7.008,
       "step": 500
     },
     {
-      "epoch": 0.02,
-      "learning_rate": 5.921669793621013e-06,
-      "loss": 0.4809,
+      "epoch": 0.018200165783688327,
+      "grad_norm": 0.13079528510570526,
+      "learning_rate": 6.065337496997358e-06,
+      "loss": 0.5946,
       "step": 505
     },
     {
-      "epoch": 0.02,
-      "learning_rate": 5.980300187617261e-06,
-      "loss": 0.4719,
+      "epoch": 0.018380365444912963,
+      "grad_norm": 0.1083323135972023,
+      "learning_rate": 6.125390343502282e-06,
+      "loss": 0.5892,
       "step": 510
     },
     {
-      "epoch": 0.02,
-      "learning_rate": 6.038930581613509e-06,
-      "loss": 0.4421,
+      "epoch": 0.0185605651061376,
+      "grad_norm": 0.16727234423160553,
+      "learning_rate": 6.185443190007207e-06,
+      "loss": 0.5907,
       "step": 515
     },
     {
-      "epoch": 0.02,
-      "learning_rate": 6.0975609756097564e-06,
-      "loss": 0.4295,
+      "epoch": 0.018740764767362236,
+      "grad_norm": 0.12499808520078659,
+      "learning_rate": 6.245496036512131e-06,
+      "loss": 0.6076,
       "step": 520
     },
     {
-      "epoch": 0.02,
-      "learning_rate": 6.156191369606005e-06,
-      "loss": 0.4448,
+      "epoch": 0.018920964428586873,
+      "grad_norm": 0.16318917274475098,
+      "learning_rate": 6.3055488830170554e-06,
+      "loss": 0.5995,
       "step": 525
     },
     {
-      "epoch": 0.02,
-      "learning_rate": 6.214821763602251e-06,
-      "loss": 0.4497,
+      "epoch": 0.019101164089811513,
+      "grad_norm": 0.12780463695526123,
+      "learning_rate": 6.365601729521979e-06,
+      "loss": 0.5618,
       "step": 530
     },
     {
-      "epoch": 0.02,
-      "learning_rate": 6.2734521575985e-06,
-      "loss": 0.4438,
+      "epoch": 0.01928136375103615,
+      "grad_norm": 0.11932645738124847,
+      "learning_rate": 6.425654576026904e-06,
+      "loss": 0.5856,
       "step": 535
     },
     {
-      "epoch": 0.02,
-      "learning_rate": 6.3320825515947465e-06,
-      "loss": 0.4549,
+      "epoch": 0.019461563412260786,
+      "grad_norm": 0.1256912350654602,
+      "learning_rate": 6.485707422531828e-06,
+      "loss": 0.5552,
       "step": 540
     },
     {
-      "epoch": 0.02,
-      "learning_rate": 6.390712945590995e-06,
-      "loss": 0.4767,
+      "epoch": 0.019641763073485422,
+      "grad_norm": 0.09648074209690094,
+      "learning_rate": 6.545760269036752e-06,
+      "loss": 0.5326,
       "step": 545
     },
     {
-      "epoch": 0.02,
-      "learning_rate": 6.449343339587242e-06,
-      "loss": 0.4291,
+      "epoch": 0.01982196273471006,
+      "grad_norm": 0.14793181419372559,
+      "learning_rate": 6.605813115541677e-06,
+      "loss": 0.5664,
       "step": 550
     },
     {
-      "epoch": 0.02,
-      "learning_rate": 6.50797373358349e-06,
-      "loss": 0.4421,
+      "epoch": 0.020002162395934695,
+      "grad_norm": 0.12733140587806702,
+      "learning_rate": 6.665865962046601e-06,
+      "loss": 0.571,
       "step": 555
     },
     {
-      "epoch": 0.02,
-      "learning_rate": 6.566604127579738e-06,
-      "loss": 0.4297,
+      "epoch": 0.02018236205715933,
+      "grad_norm": 0.12296765297651291,
+      "learning_rate": 6.725918808551526e-06,
+      "loss": 0.5242,
       "step": 560
     },
     {
-      "epoch": 0.02,
-      "learning_rate": 6.625234521575985e-06,
-      "loss": 0.4256,
+      "epoch": 0.020362561718383968,
+      "grad_norm": 0.13922452926635742,
+      "learning_rate": 6.78597165505645e-06,
+      "loss": 0.5367,
       "step": 565
     },
     {
-      "epoch": 0.02,
-      "learning_rate": 6.683864915572233e-06,
-      "loss": 0.4481,
+      "epoch": 0.020542761379608605,
+      "grad_norm": 0.14491942524909973,
+      "learning_rate": 6.846024501561374e-06,
+      "loss": 0.5708,
       "step": 570
     },
     {
-      "epoch": 0.02,
-      "learning_rate": 6.74249530956848e-06,
-      "loss": 0.4512,
+      "epoch": 0.020722961040833245,
+      "grad_norm": 0.1201525330543518,
+      "learning_rate": 6.906077348066299e-06,
+      "loss": 0.6137,
       "step": 575
     },
     {
-      "epoch": 0.02,
-      "learning_rate": 6.8011257035647285e-06,
-      "loss": 0.4377,
+      "epoch": 0.02090316070205788,
+      "grad_norm": 0.11606668680906296,
+      "learning_rate": 6.966130194571223e-06,
+      "loss": 0.5537,
       "step": 580
     },
     {
-      "epoch": 0.02,
-      "learning_rate": 6.859756097560977e-06,
-      "loss": 0.4654,
+      "epoch": 0.021083360363282518,
+      "grad_norm": 0.14227990806102753,
+      "learning_rate": 7.026183041076147e-06,
+      "loss": 0.5886,
       "step": 585
     },
     {
-      "epoch": 0.02,
-      "learning_rate": 6.918386491557223e-06,
-      "loss": 0.4026,
+      "epoch": 0.021263560024507154,
+      "grad_norm": 0.13029618561267853,
+      "learning_rate": 7.086235887581072e-06,
+      "loss": 0.5999,
       "step": 590
     },
     {
-      "epoch": 0.02,
-      "learning_rate": 6.977016885553471e-06,
-      "loss": 0.4547,
+      "epoch": 0.02144375968573179,
+      "grad_norm": 0.11412752419710159,
+      "learning_rate": 7.1462887340859955e-06,
+      "loss": 0.5678,
       "step": 595
     },
     {
-      "epoch": 0.02,
-      "learning_rate": 7.0356472795497186e-06,
-      "loss": 0.4349,
+      "epoch": 0.021623959346956427,
+      "grad_norm": 0.12800736725330353,
+      "learning_rate": 7.206341580590921e-06,
+      "loss": 0.6086,
       "step": 600
     },
     {
-      "epoch": 0.02,
-      "learning_rate": 7.094277673545967e-06,
-      "loss": 0.4776,
+      "epoch": 0.021804159008181064,
+      "grad_norm": 0.14154621958732605,
+      "learning_rate": 7.266394427095845e-06,
+      "loss": 0.539,
       "step": 605
     },
     {
-      "epoch": 0.02,
-      "learning_rate": 7.152908067542215e-06,
-      "loss": 0.4121,
+      "epoch": 0.0219843586694057,
+      "grad_norm": 0.10308407992124557,
+      "learning_rate": 7.326447273600769e-06,
+      "loss": 0.5246,
       "step": 610
     },
     {
-      "epoch": 0.02,
-      "learning_rate": 7.211538461538461e-06,
-      "loss": 0.4229,
+      "epoch": 0.02216455833063034,
+      "grad_norm": 0.15192610025405884,
+      "learning_rate": 7.3865001201056936e-06,
+      "loss": 0.5498,
       "step": 615
     },
     {
-      "epoch": 0.02,
-      "learning_rate": 7.2701688555347095e-06,
-      "loss": 0.445,
+      "epoch": 0.022344757991854976,
+      "grad_norm": 0.15246666967868805,
+      "learning_rate": 7.446552966610617e-06,
+      "loss": 0.5671,
       "step": 620
     },
     {
-      "epoch": 0.02,
-      "learning_rate": 7.328799249530957e-06,
-      "loss": 0.4528,
+      "epoch": 0.022524957653079613,
+      "grad_norm": 0.14070971310138702,
+      "learning_rate": 7.506605813115541e-06,
+      "loss": 0.5181,
       "step": 625
     },
     {
-      "epoch": 0.02,
-      "learning_rate": 7.387429643527205e-06,
-      "loss": 0.4382,
+      "epoch": 0.02270515731430425,
+      "grad_norm": 0.1048644632101059,
+      "learning_rate": 7.566658659620467e-06,
+      "loss": 0.5661,
       "step": 630
     },
     {
-      "epoch": 0.02,
-      "learning_rate": 7.446060037523453e-06,
-      "loss": 0.4237,
+      "epoch": 0.022885356975528886,
+      "grad_norm": 0.12176632136106491,
+      "learning_rate": 7.626711506125391e-06,
+      "loss": 0.6039,
       "step": 635
     },
     {
-      "epoch": 0.02,
-      "learning_rate": 7.5046904315197005e-06,
-      "loss": 0.4303,
+      "epoch": 0.023065556636753522,
+      "grad_norm": 0.11277682334184647,
+      "learning_rate": 7.686764352630316e-06,
+      "loss": 0.5394,
       "step": 640
     },
     {
-      "epoch": 0.02,
-      "learning_rate": 7.563320825515948e-06,
-      "loss": 0.4365,
+      "epoch": 0.02324575629797816,
+      "grad_norm": 0.12266416847705841,
+      "learning_rate": 7.74681719913524e-06,
+      "loss": 0.5814,
       "step": 645
     },
     {
-      "epoch": 0.02,
-      "learning_rate": 7.621951219512195e-06,
-      "loss": 0.4397,
+      "epoch": 0.023425955959202795,
+      "grad_norm": 0.1094772145152092,
+      "learning_rate": 7.806870045640164e-06,
+      "loss": 0.5854,
       "step": 650
     },
     {
-      "epoch": 0.02,
-      "learning_rate": 7.680581613508442e-06,
-      "loss": 0.4157,
+      "epoch": 0.023606155620427435,
+      "grad_norm": 0.16953280568122864,
+      "learning_rate": 7.866922892145088e-06,
+      "loss": 0.5464,
       "step": 655
     },
     {
-      "epoch": 0.02,
-      "learning_rate": 7.739212007504691e-06,
-      "loss": 0.4351,
+      "epoch": 0.023786355281652072,
+      "grad_norm": 0.16109640896320343,
+      "learning_rate": 7.926975738650012e-06,
+      "loss": 0.5889,
       "step": 660
     },
     {
-      "epoch": 0.02,
-      "learning_rate": 7.797842401500939e-06,
-      "loss": 0.4226,
+      "epoch": 0.02396655494287671,
+      "grad_norm": 0.1488220989704132,
+      "learning_rate": 7.987028585154936e-06,
+      "loss": 0.57,
       "step": 665
     },
     {
-      "epoch": 0.02,
-      "learning_rate": 7.856472795497186e-06,
-      "loss": 0.4086,
+      "epoch": 0.024146754604101345,
+      "grad_norm": 0.15228040516376495,
+      "learning_rate": 8.047081431659861e-06,
+      "loss": 0.5437,
       "step": 670
     },
     {
-      "epoch": 0.02,
-      "learning_rate": 7.915103189493433e-06,
-      "loss": 0.4167,
+      "epoch": 0.02432695426532598,
+      "grad_norm": 0.11605563014745712,
+      "learning_rate": 8.107134278164785e-06,
+      "loss": 0.5315,
       "step": 675
     },
     {
-      "epoch": 0.02,
-      "learning_rate": 7.97373358348968e-06,
-      "loss": 0.436,
+      "epoch": 0.024507153926550618,
+      "grad_norm": 0.1317128986120224,
+      "learning_rate": 8.16718712466971e-06,
+      "loss": 0.5731,
       "step": 680
     },
     {
-      "epoch": 0.02,
-      "learning_rate": 8.03236397748593e-06,
-      "loss": 0.4404,
+      "epoch": 0.024687353587775254,
+      "grad_norm": 0.14042159914970398,
+      "learning_rate": 8.227239971174635e-06,
+      "loss": 0.5542,
       "step": 685
     },
     {
-      "epoch": 0.02,
-      "learning_rate": 8.090994371482177e-06,
-      "loss": 0.4268,
+      "epoch": 0.02486755324899989,
+      "grad_norm": 0.16029533743858337,
+      "learning_rate": 8.287292817679558e-06,
+      "loss": 0.5782,
       "step": 690
     },
     {
-      "epoch": 0.02,
-      "learning_rate": 8.149624765478424e-06,
-      "loss": 0.4325,
+      "epoch": 0.025047752910224527,
+      "grad_norm": 0.1571711301803589,
+      "learning_rate": 8.347345664184484e-06,
+      "loss": 0.5533,
       "step": 695
     },
     {
-      "epoch": 0.02,
-      "learning_rate": 8.208255159474672e-06,
-      "loss": 0.41,
+      "epoch": 0.025227952571449167,
+      "grad_norm": 0.13758143782615662,
+      "learning_rate": 8.407398510689408e-06,
+      "loss": 0.5316,
       "step": 700
     },
     {
-      "epoch": 0.02,
-      "learning_rate": 8.266885553470919e-06,
-      "loss": 0.4027,
+      "epoch": 0.025408152232673804,
+      "grad_norm": 0.11138833314180374,
+      "learning_rate": 8.467451357194332e-06,
+      "loss": 0.5075,
       "step": 705
     },
     {
-      "epoch": 0.02,
-      "learning_rate": 8.325515947467168e-06,
-      "loss": 0.4532,
+      "epoch": 0.02558835189389844,
+      "grad_norm": 0.14913310110569,
+      "learning_rate": 8.527504203699256e-06,
+      "loss": 0.5363,
       "step": 710
     },
     {
-      "epoch": 0.03,
-      "learning_rate": 8.384146341463415e-06,
-      "loss": 0.4342,
+      "epoch": 0.025768551555123077,
+      "grad_norm": 0.11564284563064575,
+      "learning_rate": 8.58755705020418e-06,
+      "loss": 0.5616,
       "step": 715
     },
     {
-      "epoch": 0.03,
-      "learning_rate": 8.442776735459663e-06,
-      "loss": 0.4072,
+      "epoch": 0.025948751216347713,
+      "grad_norm": 0.15782320499420166,
+      "learning_rate": 8.647609896709103e-06,
+      "loss": 0.5368,
       "step": 720
     },
     {
-      "epoch": 0.03,
-      "learning_rate": 8.50140712945591e-06,
-      "loss": 0.3962,
+      "epoch": 0.02612895087757235,
+      "grad_norm": 0.16397272050380707,
+      "learning_rate": 8.707662743214029e-06,
+      "loss": 0.6325,
       "step": 725
     },
     {
-      "epoch": 0.03,
-      "learning_rate": 8.560037523452157e-06,
-      "loss": 0.4224,
+      "epoch": 0.026309150538796986,
+      "grad_norm": 0.16753605008125305,
+      "learning_rate": 8.767715589718953e-06,
+      "loss": 0.5532,
       "step": 730
     },
     {
-      "epoch": 0.03,
-      "learning_rate": 8.618667917448406e-06,
-      "loss": 0.443,
+      "epoch": 0.026489350200021623,
+      "grad_norm": 0.14978809654712677,
+      "learning_rate": 8.827768436223878e-06,
+      "loss": 0.5431,
       "step": 735
     },
     {
-      "epoch": 0.03,
-      "learning_rate": 8.677298311444654e-06,
-      "loss": 0.4262,
+      "epoch": 0.026669549861246263,
+      "grad_norm": 0.15160630643367767,
+      "learning_rate": 8.887821282728802e-06,
+      "loss": 0.576,
       "step": 740
     },
     {
-      "epoch": 0.03,
-      "learning_rate": 8.735928705440901e-06,
-      "loss": 0.4276,
+      "epoch": 0.0268497495224709,
+      "grad_norm": 0.1296982318162918,
+      "learning_rate": 8.947874129233726e-06,
+      "loss": 0.5488,
       "step": 745
     },
     {
-      "epoch": 0.03,
-      "learning_rate": 8.794559099437148e-06,
-      "loss": 0.4333,
+      "epoch": 0.027029949183695536,
+      "grad_norm": 0.19447606801986694,
+      "learning_rate": 9.007926975738652e-06,
+      "loss": 0.5878,
       "step": 750
     },
     {
-      "epoch": 0.03,
-      "learning_rate": 8.853189493433396e-06,
-      "loss": 0.4199,
+      "epoch": 0.027210148844920172,
+      "grad_norm": 0.13086961209774017,
+      "learning_rate": 9.067979822243575e-06,
+      "loss": 0.5713,
       "step": 755
     },
     {
-      "epoch": 0.03,
-      "learning_rate": 8.911819887429645e-06,
-      "loss": 0.4505,
+      "epoch": 0.02739034850614481,
+      "grad_norm": 0.1761922836303711,
+      "learning_rate": 9.1280326687485e-06,
+      "loss": 0.5605,
       "step": 760
     },
     {
-      "epoch": 0.03,
-      "learning_rate": 8.970450281425892e-06,
-      "loss": 0.4286,
+      "epoch": 0.027570548167369445,
+      "grad_norm": 0.13089734315872192,
+      "learning_rate": 9.188085515253423e-06,
+      "loss": 0.5187,
       "step": 765
     },
     {
-      "epoch": 0.03,
-      "learning_rate": 9.02908067542214e-06,
-      "loss": 0.4126,
+      "epoch": 0.02775074782859408,
+      "grad_norm": 0.14569701254367828,
+      "learning_rate": 9.248138361758347e-06,
+      "loss": 0.5301,
       "step": 770
     },
     {
-      "epoch": 0.03,
-      "learning_rate": 9.087711069418386e-06,
-      "loss": 0.3909,
+      "epoch": 0.027930947489818718,
+      "grad_norm": 0.16195152699947357,
+      "learning_rate": 9.308191208263273e-06,
+      "loss": 0.5554,
       "step": 775
     },
     {
-      "epoch": 0.03,
-      "learning_rate": 9.146341463414634e-06,
-      "loss": 0.4132,
+      "epoch": 0.028111147151043354,
+      "grad_norm": 0.13685005903244019,
+      "learning_rate": 9.368244054768196e-06,
+      "loss": 0.5719,
       "step": 780
     },
     {
-      "epoch": 0.03,
-      "learning_rate": 9.204971857410883e-06,
-      "loss": 0.3725,
+      "epoch": 0.028291346812267994,
+      "grad_norm": 0.1750328093767166,
+      "learning_rate": 9.42829690127312e-06,
+      "loss": 0.5635,
       "step": 785
     },
     {
-      "epoch": 0.03,
-      "learning_rate": 9.26360225140713e-06,
-      "loss": 0.3996,
+      "epoch": 0.02847154647349263,
+      "grad_norm": 0.16423344612121582,
+      "learning_rate": 9.488349747778046e-06,
+      "loss": 0.6248,
       "step": 790
     },
     {
-      "epoch": 0.03,
-      "learning_rate": 9.322232645403377e-06,
-      "loss": 0.434,
+      "epoch": 0.028651746134717267,
+      "grad_norm": 0.15460895001888275,
+      "learning_rate": 9.54840259428297e-06,
+      "loss": 0.5727,
       "step": 795
     },
     {
-      "epoch": 0.03,
-      "learning_rate": 9.380863039399625e-06,
-      "loss": 0.432,
+      "epoch": 0.028831945795941904,
+      "grad_norm": 0.15125848352909088,
+      "learning_rate": 9.608455440787894e-06,
+      "loss": 0.5535,
       "step": 800
     },
     {
-      "epoch": 0.03,
-      "learning_rate": 9.439493433395872e-06,
-      "loss": 0.407,
+      "epoch": 0.02901214545716654,
+      "grad_norm": 0.18755701184272766,
+      "learning_rate": 9.66850828729282e-06,
+      "loss": 0.5292,
       "step": 805
     },
     {
-      "epoch": 0.03,
-      "learning_rate": 9.498123827392121e-06,
-      "loss": 0.3913,
+      "epoch": 0.029192345118391177,
+      "grad_norm": 0.14782796800136566,
+      "learning_rate": 9.728561133797743e-06,
+      "loss": 0.5155,
       "step": 810
     },
     {
-      "epoch": 0.03,
-      "learning_rate": 9.556754221388368e-06,
-      "loss": 0.4433,
+      "epoch": 0.029372544779615813,
+      "grad_norm": 0.1731768250465393,
+      "learning_rate": 9.788613980302667e-06,
+      "loss": 0.5752,
       "step": 815
     },
     {
-      "epoch": 0.03,
-      "learning_rate": 9.615384615384616e-06,
-      "loss": 0.4153,
+      "epoch": 0.02955274444084045,
+      "grad_norm": 0.17426101863384247,
+      "learning_rate": 9.84866682680759e-06,
+      "loss": 0.5796,
       "step": 820
     },
     {
-      "epoch": 0.03,
-      "learning_rate": 9.674015009380863e-06,
-      "loss": 0.4182,
+      "epoch": 0.02973294410206509,
+      "grad_norm": 0.1710687130689621,
+      "learning_rate": 9.908719673312515e-06,
+      "loss": 0.5363,
       "step": 825
     },
     {
-      "epoch": 0.03,
-      "learning_rate": 9.73264540337711e-06,
-      "loss": 0.4045,
+      "epoch": 0.029913143763289726,
+      "grad_norm": 0.12665097415447235,
+      "learning_rate": 9.96877251981744e-06,
+      "loss": 0.5302,
       "step": 830
     },
     {
-      "epoch": 0.03,
-      "learning_rate": 9.79127579737336e-06,
-      "loss": 0.4041,
+      "epoch": 0.030093343424514363,
+      "grad_norm": 0.20195025205612183,
+      "learning_rate": 1.0028825366322364e-05,
+      "loss": 0.6092,
       "step": 835
     },
     {
-      "epoch": 0.03,
-      "learning_rate": 9.849906191369607e-06,
-      "loss": 0.3941,
+      "epoch": 0.030273543085739,
+      "grad_norm": 0.18237920105457306,
+      "learning_rate": 1.0088878212827288e-05,
+      "loss": 0.5281,
       "step": 840
     },
     {
-      "epoch": 0.03,
-      "learning_rate": 9.908536585365854e-06,
-      "loss": 0.3963,
+      "epoch": 0.030453742746963636,
+      "grad_norm": 0.17698000371456146,
+      "learning_rate": 1.0148931059332214e-05,
+      "loss": 0.5705,
       "step": 845
     },
     {
-      "epoch": 0.03,
-      "learning_rate": 9.967166979362101e-06,
-      "loss": 0.432,
+      "epoch": 0.030633942408188272,
+      "grad_norm": 0.16873596608638763,
+      "learning_rate": 1.0208983905837137e-05,
+      "loss": 0.5534,
       "step": 850
     },
     {
-      "epoch": 0.03,
-      "learning_rate": 1.0025797373358349e-05,
-      "loss": 0.4181,
+      "epoch": 0.03081414206941291,
+      "grad_norm": 0.18587124347686768,
+      "learning_rate": 1.0269036752342063e-05,
+      "loss": 0.5482,
       "step": 855
     },
     {
-      "epoch": 0.03,
-      "learning_rate": 1.0084427767354598e-05,
-      "loss": 0.4433,
+      "epoch": 0.030994341730637545,
+      "grad_norm": 0.15320129692554474,
+      "learning_rate": 1.0329089598846985e-05,
+      "loss": 0.584,
       "step": 860
     },
     {
-      "epoch": 0.03,
-      "learning_rate": 1.0143058161350845e-05,
-      "loss": 0.401,
+      "epoch": 0.03117454139186218,
+      "grad_norm": 0.15630275011062622,
+      "learning_rate": 1.0389142445351909e-05,
+      "loss": 0.5771,
       "step": 865
     },
     {
-      "epoch": 0.03,
-      "learning_rate": 1.0201688555347092e-05,
-      "loss": 0.3956,
+      "epoch": 0.03135474105308682,
+      "grad_norm": 0.20800472795963287,
+      "learning_rate": 1.0449195291856835e-05,
+      "loss": 0.5642,
       "step": 870
     },
     {
-      "epoch": 0.03,
-      "learning_rate": 1.0260318949343341e-05,
-      "loss": 0.4052,
+      "epoch": 0.031534940714311455,
+      "grad_norm": 0.12712302803993225,
+      "learning_rate": 1.0509248138361758e-05,
+      "loss": 0.5427,
       "step": 875
     },
     {
-      "epoch": 0.03,
-      "learning_rate": 1.0318949343339587e-05,
-      "loss": 0.4146,
+      "epoch": 0.03171514037553609,
+      "grad_norm": 0.1484389752149582,
+      "learning_rate": 1.0569300984866682e-05,
+      "loss": 0.5385,
       "step": 880
     },
     {
-      "epoch": 0.03,
-      "learning_rate": 1.0377579737335836e-05,
-      "loss": 0.4049,
+      "epoch": 0.03189534003676073,
+      "grad_norm": 0.12968121469020844,
+      "learning_rate": 1.0629353831371608e-05,
+      "loss": 0.5315,
       "step": 885
     },
     {
-      "epoch": 0.03,
-      "learning_rate": 1.0436210131332083e-05,
-      "loss": 0.4177,
+      "epoch": 0.03207553969798537,
+      "grad_norm": 0.14982539415359497,
+      "learning_rate": 1.0689406677876532e-05,
+      "loss": 0.5571,
       "step": 890
     },
     {
-      "epoch": 0.03,
-      "learning_rate": 1.049484052532833e-05,
-      "loss": 0.4537,
+      "epoch": 0.03225573935921001,
+      "grad_norm": 0.16306400299072266,
+      "learning_rate": 1.0749459524381456e-05,
+      "loss": 0.5379,
       "step": 895
     },
     {
-      "epoch": 0.03,
-      "learning_rate": 1.055347091932458e-05,
-      "loss": 0.4175,
+      "epoch": 0.032435939020434644,
+      "grad_norm": 0.19171880185604095,
+      "learning_rate": 1.0809512370886381e-05,
+      "loss": 0.5652,
       "step": 900
     },
     {
-      "epoch": 0.03,
-      "learning_rate": 1.0612101313320825e-05,
-      "loss": 0.3695,
+      "epoch": 0.03261613868165928,
+      "grad_norm": 0.1954376995563507,
+      "learning_rate": 1.0869565217391305e-05,
+      "loss": 0.5635,
       "step": 905
     },
     {
-      "epoch": 0.03,
-      "learning_rate": 1.0670731707317074e-05,
-      "loss": 0.4037,
+      "epoch": 0.03279633834288392,
+      "grad_norm": 0.1810484379529953,
+      "learning_rate": 1.0929618063896229e-05,
+      "loss": 0.5046,
       "step": 910
     },
     {
-      "epoch": 0.03,
-      "learning_rate": 1.0729362101313321e-05,
-      "loss": 0.4141,
+      "epoch": 0.03297653800410855,
+      "grad_norm": 0.18302740156650543,
+      "learning_rate": 1.0989670910401153e-05,
+      "loss": 0.5501,
       "step": 915
     },
     {
-      "epoch": 0.03,
-      "learning_rate": 1.0787992495309569e-05,
-      "loss": 0.4023,
+      "epoch": 0.03315673766533319,
+      "grad_norm": 0.1687111258506775,
+      "learning_rate": 1.1049723756906077e-05,
+      "loss": 0.5069,
       "step": 920
     },
     {
-      "epoch": 0.03,
-      "learning_rate": 1.0846622889305818e-05,
-      "loss": 0.3647,
+      "epoch": 0.033336937326557826,
+      "grad_norm": 0.16762186586856842,
+      "learning_rate": 1.1109776603411002e-05,
+      "loss": 0.5167,
       "step": 925
     },
     {
-      "epoch": 0.03,
-      "learning_rate": 1.0905253283302063e-05,
-      "loss": 0.4172,
+      "epoch": 0.03351713698778246,
+      "grad_norm": 0.22583533823490143,
+      "learning_rate": 1.1169829449915926e-05,
+      "loss": 0.5251,
       "step": 930
     },
     {
-      "epoch": 0.03,
-      "learning_rate": 1.0963883677298312e-05,
-      "loss": 0.3895,
+      "epoch": 0.0336973366490071,
+      "grad_norm": 0.19438394904136658,
+      "learning_rate": 1.122988229642085e-05,
+      "loss": 0.6164,
       "step": 935
     },
     {
-      "epoch": 0.03,
-      "learning_rate": 1.102251407129456e-05,
-      "loss": 0.3876,
+      "epoch": 0.033877536310231736,
+      "grad_norm": 0.1634751260280609,
+      "learning_rate": 1.1289935142925776e-05,
+      "loss": 0.5606,
       "step": 940
     },
     {
-      "epoch": 0.03,
-      "learning_rate": 1.1081144465290807e-05,
-      "loss": 0.3759,
+      "epoch": 0.03405773597145637,
+      "grad_norm": 0.1744074523448944,
+      "learning_rate": 1.13499879894307e-05,
+      "loss": 0.5853,
       "step": 945
     },
     {
-      "epoch": 0.03,
-      "learning_rate": 1.1139774859287056e-05,
-      "loss": 0.4043,
+      "epoch": 0.03423793563268101,
+      "grad_norm": 0.18406233191490173,
+      "learning_rate": 1.1410040835935625e-05,
+      "loss": 0.5132,
       "step": 950
     },
     {
-      "epoch": 0.03,
-      "learning_rate": 1.1198405253283302e-05,
-      "loss": 0.3948,
+      "epoch": 0.034418135293905645,
+      "grad_norm": 0.14555774629116058,
+      "learning_rate": 1.1470093682440549e-05,
+      "loss": 0.546,
       "step": 955
     },
     {
-      "epoch": 0.03,
-      "learning_rate": 1.125703564727955e-05,
-      "loss": 0.4055,
+      "epoch": 0.03459833495513028,
+      "grad_norm": 0.14365147054195404,
+      "learning_rate": 1.1530146528945473e-05,
+      "loss": 0.5515,
       "step": 960
     },
     {
-      "epoch": 0.03,
-      "learning_rate": 1.1315666041275798e-05,
-      "loss": 0.4142,
+      "epoch": 0.03477853461635492,
+      "grad_norm": 0.19130977988243103,
+      "learning_rate": 1.1590199375450397e-05,
+      "loss": 0.5576,
       "step": 965
     },
     {
-      "epoch": 0.03,
-      "learning_rate": 1.1374296435272045e-05,
-      "loss": 0.4328,
+      "epoch": 0.034958734277579555,
+      "grad_norm": 0.1700170785188675,
+      "learning_rate": 1.165025222195532e-05,
+      "loss": 0.5344,
       "step": 970
     },
     {
-      "epoch": 0.03,
-      "learning_rate": 1.1432926829268294e-05,
-      "loss": 0.4047,
+      "epoch": 0.0351389339388042,
+      "grad_norm": 0.14722155034542084,
+      "learning_rate": 1.1710305068460244e-05,
+      "loss": 0.5644,
       "step": 975
     },
     {
-      "epoch": 0.03,
-      "learning_rate": 1.149155722326454e-05,
-      "loss": 0.408,
+      "epoch": 0.035319133600028835,
+      "grad_norm": 0.1315588653087616,
+      "learning_rate": 1.177035791496517e-05,
+      "loss": 0.5136,
       "step": 980
     },
     {
-      "epoch": 0.03,
-      "learning_rate": 1.1550187617260787e-05,
-      "loss": 0.405,
+      "epoch": 0.03549933326125347,
+      "grad_norm": 0.22789514064788818,
+      "learning_rate": 1.1830410761470094e-05,
+      "loss": 0.5709,
       "step": 985
     },
     {
-      "epoch": 0.03,
-      "learning_rate": 1.1608818011257036e-05,
-      "loss": 0.4044,
+      "epoch": 0.03567953292247811,
+      "grad_norm": 0.19900259375572205,
+      "learning_rate": 1.189046360797502e-05,
+      "loss": 0.5815,
       "step": 990
     },
     {
-      "epoch": 0.04,
-      "learning_rate": 1.1667448405253284e-05,
-      "loss": 0.4023,
+      "epoch": 0.035859732583702744,
+      "grad_norm": 0.20573367178440094,
+      "learning_rate": 1.1950516454479943e-05,
+      "loss": 0.5188,
       "step": 995
     },
     {
-      "epoch": 0.04,
-      "learning_rate": 1.1726078799249533e-05,
-      "loss": 0.3951,
+      "epoch": 0.03603993224492738,
+      "grad_norm": 0.17396429181098938,
+      "learning_rate": 1.2010569300984867e-05,
+      "loss": 0.5032,
       "step": 1000
     },
     {
-      "epoch": 0.04,
-      "eval_loss": 0.3985450863838196,
-      "eval_runtime": 10.5677,
-      "eval_samples_per_second": 9.463,
-      "eval_steps_per_second": 9.463,
+      "epoch": 0.03603993224492738,
+      "eval_loss": 0.5565588474273682,
+      "eval_runtime": 3.5193,
+      "eval_samples_per_second": 28.415,
+      "eval_steps_per_second": 7.104,
       "step": 1000
     },
     {
-      "epoch": 0.04,
-      "learning_rate": 1.178470919324578e-05,
-      "loss": 0.3828,
+      "epoch": 0.03622013190615202,
+      "grad_norm": 0.1656351536512375,
+      "learning_rate": 1.2070622147489793e-05,
+      "loss": 0.5045,
       "step": 1005
     },
     {
-      "epoch": 0.04,
-      "learning_rate": 1.1843339587242026e-05,
-      "loss": 0.3952,
+      "epoch": 0.036400331567376654,
+      "grad_norm": 0.15851475298404694,
+      "learning_rate": 1.2130674993994717e-05,
+      "loss": 0.5142,
       "step": 1010
     },
     {
-      "epoch": 0.04,
-      "learning_rate": 1.1901969981238275e-05,
-      "loss": 0.4074,
+      "epoch": 0.03658053122860129,
+      "grad_norm": 0.18795153498649597,
+      "learning_rate": 1.219072784049964e-05,
+      "loss": 0.5408,
       "step": 1015
     },
     {
-      "epoch": 0.04,
-      "learning_rate": 1.1960600375234522e-05,
-      "loss": 0.3446,
+      "epoch": 0.03676073088982593,
+      "grad_norm": 0.18724803626537323,
+      "learning_rate": 1.2250780687004564e-05,
+      "loss": 0.5132,
       "step": 1020
     },
     {
-      "epoch": 0.04,
-      "learning_rate": 1.2019230769230771e-05,
-      "loss": 0.3972,
+      "epoch": 0.03694093055105056,
+      "grad_norm": 0.17133454978466034,
+      "learning_rate": 1.2310833533509488e-05,
+      "loss": 0.5231,
       "step": 1025
     },
     {
-      "epoch": 0.04,
-      "learning_rate": 1.2077861163227018e-05,
-      "loss": 0.3501,
+      "epoch": 0.0371211302122752,
+      "grad_norm": 0.19401347637176514,
+      "learning_rate": 1.2370886380014414e-05,
+      "loss": 0.5672,
       "step": 1030
     },
     {
-      "epoch": 0.04,
-      "learning_rate": 1.2136491557223264e-05,
-      "loss": 0.4003,
+      "epoch": 0.037301329873499836,
+      "grad_norm": 0.23305334150791168,
+      "learning_rate": 1.2430939226519338e-05,
+      "loss": 0.5496,
       "step": 1035
     },
     {
-      "epoch": 0.04,
-      "learning_rate": 1.2195121951219513e-05,
-      "loss": 0.3998,
+      "epoch": 0.03748152953472447,
+      "grad_norm": 0.1804538369178772,
+      "learning_rate": 1.2490992073024261e-05,
+      "loss": 0.5241,
       "step": 1040
     },
     {
-      "epoch": 0.04,
-      "learning_rate": 1.225375234521576e-05,
-      "loss": 0.3966,
+      "epoch": 0.03766172919594911,
+      "grad_norm": 0.16487692296504974,
+      "learning_rate": 1.2551044919529187e-05,
+      "loss": 0.537,
       "step": 1045
     },
     {
-      "epoch": 0.04,
-      "learning_rate": 1.231238273921201e-05,
-      "loss": 0.4103,
+      "epoch": 0.037841928857173746,
+      "grad_norm": 0.17555435001850128,
+      "learning_rate": 1.2611097766034111e-05,
+      "loss": 0.5381,
       "step": 1050
     },
     {
-      "epoch": 0.04,
-      "learning_rate": 1.2371013133208257e-05,
-      "loss": 0.4126,
+      "epoch": 0.03802212851839838,
+      "grad_norm": 0.19943825900554657,
+      "learning_rate": 1.2671150612539035e-05,
+      "loss": 0.5353,
       "step": 1055
     },
     {
-      "epoch": 0.04,
-      "learning_rate": 1.2429643527204502e-05,
-      "loss": 0.4029,
+      "epoch": 0.038202328179623025,
+      "grad_norm": 0.19467459619045258,
+      "learning_rate": 1.2731203459043959e-05,
+      "loss": 0.5634,
       "step": 1060
     },
     {
-      "epoch": 0.04,
-      "learning_rate": 1.2488273921200751e-05,
-      "loss": 0.4234,
+      "epoch": 0.03838252784084766,
+      "grad_norm": 0.1945466697216034,
+      "learning_rate": 1.2791256305548884e-05,
+      "loss": 0.5709,
       "step": 1065
     },
     {
-      "epoch": 0.04,
-      "learning_rate": 1.2546904315197e-05,
-      "loss": 0.4048,
+      "epoch": 0.0385627275020723,
+      "grad_norm": 0.17574776709079742,
+      "learning_rate": 1.2851309152053808e-05,
+      "loss": 0.496,
       "step": 1070
     },
     {
-      "epoch": 0.04,
-      "learning_rate": 1.2605534709193246e-05,
-      "loss": 0.4306,
+      "epoch": 0.038742927163296935,
+      "grad_norm": 0.19129513204097748,
+      "learning_rate": 1.2911361998558732e-05,
+      "loss": 0.4818,
       "step": 1075
     },
     {
-      "epoch": 0.04,
-      "learning_rate": 1.2664165103189493e-05,
-      "loss": 0.4036,
+      "epoch": 0.03892312682452157,
+      "grad_norm": 0.16488486528396606,
+      "learning_rate": 1.2971414845063656e-05,
+      "loss": 0.5433,
       "step": 1080
     },
     {
-      "epoch": 0.04,
-      "learning_rate": 1.272279549718574e-05,
-      "loss": 0.3782,
+      "epoch": 0.03910332648574621,
+      "grad_norm": 0.1613866090774536,
+      "learning_rate": 1.303146769156858e-05,
+      "loss": 0.5058,
       "step": 1085
     },
     {
-      "epoch": 0.04,
-      "learning_rate": 1.278142589118199e-05,
-      "loss": 0.408,
+      "epoch": 0.039283526146970844,
+      "grad_norm": 0.21093931794166565,
+      "learning_rate": 1.3091520538073504e-05,
+      "loss": 0.5199,
       "step": 1090
     },
     {
-      "epoch": 0.04,
-      "learning_rate": 1.2840056285178237e-05,
-      "loss": 0.4104,
+      "epoch": 0.03946372580819548,
+      "grad_norm": 0.1792885661125183,
+      "learning_rate": 1.315157338457843e-05,
+      "loss": 0.5298,
       "step": 1095
     },
     {
-      "epoch": 0.04,
-      "learning_rate": 1.2898686679174484e-05,
-      "loss": 0.3797,
+      "epoch": 0.03964392546942012,
+      "grad_norm": 0.20177467167377472,
+      "learning_rate": 1.3211626231083355e-05,
+      "loss": 0.5183,
       "step": 1100
     },
     {
-      "epoch": 0.04,
-      "learning_rate": 1.2957317073170733e-05,
-      "loss": 0.4013,
+      "epoch": 0.039824125130644754,
+      "grad_norm": 0.1867968887090683,
+      "learning_rate": 1.3271679077588279e-05,
+      "loss": 0.5268,
       "step": 1105
     },
     {
-      "epoch": 0.04,
-      "learning_rate": 1.301594746716698e-05,
-      "loss": 0.401,
+      "epoch": 0.04000432479186939,
+      "grad_norm": 0.17943917214870453,
+      "learning_rate": 1.3331731924093202e-05,
+      "loss": 0.555,
       "step": 1110
     },
     {
-      "epoch": 0.04,
-      "learning_rate": 1.307457786116323e-05,
-      "loss": 0.4132,
+      "epoch": 0.04018452445309403,
+      "grad_norm": 0.24387885630130768,
+      "learning_rate": 1.3391784770598126e-05,
+      "loss": 0.5367,
       "step": 1115
     },
     {
-      "epoch": 0.04,
-      "learning_rate": 1.3133208255159477e-05,
-      "loss": 0.3838,
+      "epoch": 0.04036472411431866,
+      "grad_norm": 0.27169400453567505,
+      "learning_rate": 1.3451837617103052e-05,
+      "loss": 0.5293,
       "step": 1120
     },
     {
-      "epoch": 0.04,
-      "learning_rate": 1.3191838649155722e-05,
-      "loss": 0.4303,
+      "epoch": 0.0405449237755433,
+      "grad_norm": 0.22766202688217163,
+      "learning_rate": 1.3511890463607976e-05,
+      "loss": 0.5438,
       "step": 1125
     },
     {
-      "epoch": 0.04,
-      "learning_rate": 1.325046904315197e-05,
-      "loss": 0.4177,
+      "epoch": 0.040725123436767936,
+      "grad_norm": 0.17502576112747192,
+      "learning_rate": 1.35719433101129e-05,
+      "loss": 0.482,
       "step": 1130
     },
     {
-      "epoch": 0.04,
-      "learning_rate": 1.3309099437148217e-05,
-      "loss": 0.3646,
+      "epoch": 0.04090532309799257,
+      "grad_norm": 0.1690770983695984,
+      "learning_rate": 1.3631996156617823e-05,
+      "loss": 0.5331,
       "step": 1135
     },
     {
-      "epoch": 0.04,
-      "learning_rate": 1.3367729831144466e-05,
-      "loss": 0.3996,
+      "epoch": 0.04108552275921721,
+      "grad_norm": 0.18362893164157867,
+      "learning_rate": 1.3692049003122747e-05,
+      "loss": 0.4705,
       "step": 1140
     },
     {
-      "epoch": 0.04,
-      "learning_rate": 1.3426360225140713e-05,
-      "loss": 0.3913,
+      "epoch": 0.04126572242044185,
+      "grad_norm": 0.2055690884590149,
+      "learning_rate": 1.3752101849627675e-05,
+      "loss": 0.4958,
       "step": 1145
     },
     {
-      "epoch": 0.04,
-      "learning_rate": 1.348499061913696e-05,
-      "loss": 0.3774,
+      "epoch": 0.04144592208166649,
+      "grad_norm": 0.20133014023303986,
+      "learning_rate": 1.3812154696132598e-05,
+      "loss": 0.5269,
       "step": 1150
     },
     {
-      "epoch": 0.04,
-      "learning_rate": 1.354362101313321e-05,
-      "loss": 0.4188,
+      "epoch": 0.041626121742891126,
+      "grad_norm": 0.18268916010856628,
+      "learning_rate": 1.3872207542637522e-05,
+      "loss": 0.5104,
       "step": 1155
     },
     {
-      "epoch": 0.04,
-      "learning_rate": 1.3602251407129457e-05,
-      "loss": 0.3658,
+      "epoch": 0.04180632140411576,
+      "grad_norm": 0.22504407167434692,
+      "learning_rate": 1.3932260389142446e-05,
+      "loss": 0.5622,
       "step": 1160
     },
     {
-      "epoch": 0.04,
-      "learning_rate": 1.3660881801125704e-05,
-      "loss": 0.387,
+      "epoch": 0.0419865210653404,
+      "grad_norm": 0.2050849348306656,
+      "learning_rate": 1.399231323564737e-05,
+      "loss": 0.5176,
       "step": 1165
     },
     {
-      "epoch": 0.04,
-      "learning_rate": 1.3719512195121953e-05,
-      "loss": 0.3991,
+      "epoch": 0.042166720726565035,
+      "grad_norm": 0.1802973598241806,
+      "learning_rate": 1.4052366082152294e-05,
+      "loss": 0.5322,
       "step": 1170
     },
     {
-      "epoch": 0.04,
-      "learning_rate": 1.37781425891182e-05,
-      "loss": 0.3824,
+      "epoch": 0.04234692038778967,
+      "grad_norm": 0.19005241990089417,
+      "learning_rate": 1.411241892865722e-05,
+      "loss": 0.5582,
       "step": 1175
     },
     {
-      "epoch": 0.04,
-      "learning_rate": 1.3836772983114446e-05,
-      "loss": 0.4034,
+      "epoch": 0.04252712004901431,
+      "grad_norm": 0.2026328444480896,
+      "learning_rate": 1.4172471775162143e-05,
+      "loss": 0.5726,
       "step": 1180
     },
     {
-      "epoch": 0.04,
-      "learning_rate": 1.3895403377110693e-05,
-      "loss": 0.3582,
+      "epoch": 0.042707319710238945,
+      "grad_norm": 0.22983752191066742,
+      "learning_rate": 1.4232524621667067e-05,
+      "loss": 0.5469,
       "step": 1185
     },
     {
-      "epoch": 0.04,
-      "learning_rate": 1.3954033771106943e-05,
-      "loss": 0.3816,
+      "epoch": 0.04288751937146358,
+      "grad_norm": 0.15863169729709625,
+      "learning_rate": 1.4292577468171991e-05,
+      "loss": 0.5368,
       "step": 1190
     },
     {
-      "epoch": 0.04,
-      "learning_rate": 1.401266416510319e-05,
-      "loss": 0.38,
+      "epoch": 0.04306771903268822,
+      "grad_norm": 0.19948942959308624,
+      "learning_rate": 1.4352630314676915e-05,
+      "loss": 0.518,
       "step": 1195
     },
     {
-      "epoch": 0.04,
-      "learning_rate": 1.4071294559099437e-05,
-      "loss": 0.3887,
+      "epoch": 0.043247918693912854,
+      "grad_norm": 0.20229573547840118,
+      "learning_rate": 1.4412683161181842e-05,
+      "loss": 0.5379,
       "step": 1200
     },
     {
-      "epoch": 0.04,
-      "learning_rate": 1.4129924953095686e-05,
-      "loss": 0.4033,
+      "epoch": 0.04342811835513749,
+      "grad_norm": 0.22024467587471008,
+      "learning_rate": 1.4472736007686766e-05,
+      "loss": 0.5648,
       "step": 1205
     },
     {
-      "epoch": 0.04,
-      "learning_rate": 1.4188555347091933e-05,
-      "loss": 0.3784,
+      "epoch": 0.04360831801636213,
+      "grad_norm": 0.22408129274845123,
+      "learning_rate": 1.453278885419169e-05,
+      "loss": 0.5228,
       "step": 1210
     },
     {
-      "epoch": 0.04,
-      "learning_rate": 1.424718574108818e-05,
-      "loss": 0.3889,
+      "epoch": 0.043788517677586763,
+      "grad_norm": 0.21704581379890442,
+      "learning_rate": 1.4592841700696614e-05,
+      "loss": 0.5289,
       "step": 1215
     },
     {
-      "epoch": 0.04,
-      "learning_rate": 1.430581613508443e-05,
-      "loss": 0.3671,
+      "epoch": 0.0439687173388114,
+      "grad_norm": 0.1930111199617386,
+      "learning_rate": 1.4652894547201538e-05,
+      "loss": 0.5176,
       "step": 1220
     },
     {
-      "epoch": 0.04,
-      "learning_rate": 1.4364446529080677e-05,
-      "loss": 0.3575,
+      "epoch": 0.044148917000036036,
+      "grad_norm": 0.23059917986392975,
+      "learning_rate": 1.4712947393706462e-05,
+      "loss": 0.5323,
       "step": 1225
     },
     {
-      "epoch": 0.04,
-      "learning_rate": 1.4423076923076923e-05,
-      "loss": 0.3871,
+      "epoch": 0.04432911666126068,
+      "grad_norm": 0.21172671020030975,
+      "learning_rate": 1.4773000240211387e-05,
+      "loss": 0.544,
       "step": 1230
     },
     {
-      "epoch": 0.04,
-      "learning_rate": 1.448170731707317e-05,
-      "loss": 0.4034,
+      "epoch": 0.044509316322485316,
+      "grad_norm": 0.20656728744506836,
+      "learning_rate": 1.4833053086716311e-05,
+      "loss": 0.5019,
       "step": 1235
     },
     {
-      "epoch": 0.04,
-      "learning_rate": 1.4540337711069419e-05,
-      "loss": 0.396,
+      "epoch": 0.04468951598370995,
+      "grad_norm": 0.19450949132442474,
+      "learning_rate": 1.4893105933221235e-05,
+      "loss": 0.5262,
       "step": 1240
     },
     {
-      "epoch": 0.04,
-      "learning_rate": 1.4598968105065666e-05,
-      "loss": 0.3741,
+      "epoch": 0.04486971564493459,
+      "grad_norm": 0.18495753407478333,
+      "learning_rate": 1.4953158779726159e-05,
+      "loss": 0.4986,
       "step": 1245
     },
     {
-      "epoch": 0.04,
-      "learning_rate": 1.4657598499061914e-05,
-      "loss": 0.3861,
+      "epoch": 0.045049915306159226,
+      "grad_norm": 0.20766963064670563,
+      "learning_rate": 1.5013211626231083e-05,
+      "loss": 0.5081,
       "step": 1250
     },
     {
-      "epoch": 0.04,
-      "learning_rate": 1.4716228893058163e-05,
-      "loss": 0.4007,
+      "epoch": 0.04523011496738386,
+      "grad_norm": 0.18297769129276276,
+      "learning_rate": 1.507326447273601e-05,
+      "loss": 0.5523,
       "step": 1255
     },
     {
-      "epoch": 0.04,
-      "learning_rate": 1.477485928705441e-05,
-      "loss": 0.367,
+      "epoch": 0.0454103146286085,
+      "grad_norm": 0.18521635234355927,
+      "learning_rate": 1.5133317319240934e-05,
+      "loss": 0.5279,
       "step": 1260
     },
     {
-      "epoch": 0.04,
-      "learning_rate": 1.4833489681050657e-05,
-      "loss": 0.355,
+      "epoch": 0.045590514289833135,
+      "grad_norm": 0.22829121351242065,
+      "learning_rate": 1.5193370165745858e-05,
+      "loss": 0.5055,
       "step": 1265
     },
     {
-      "epoch": 0.04,
-      "learning_rate": 1.4892120075046906e-05,
-      "loss": 0.4029,
+      "epoch": 0.04577071395105777,
+      "grad_norm": 0.21849414706230164,
+      "learning_rate": 1.5253423012250781e-05,
+      "loss": 0.539,
       "step": 1270
     },
     {
-      "epoch": 0.04,
-      "learning_rate": 1.4950750469043154e-05,
-      "loss": 0.3722,
+      "epoch": 0.04595091361228241,
+      "grad_norm": 0.21421493589878082,
+      "learning_rate": 1.5313475858755704e-05,
+      "loss": 0.5275,
       "step": 1275
     },
     {
-      "epoch": 0.05,
-      "learning_rate": 1.5009380863039401e-05,
-      "loss": 0.3901,
+      "epoch": 0.046131113273507045,
+      "grad_norm": 0.23160327970981598,
+      "learning_rate": 1.5373528705260633e-05,
+      "loss": 0.5409,
       "step": 1280
     },
     {
-      "epoch": 0.05,
-      "learning_rate": 1.5068011257035647e-05,
-      "loss": 0.4027,
+      "epoch": 0.04631131293473168,
+      "grad_norm": 0.16949445009231567,
+      "learning_rate": 1.5433581551765555e-05,
+      "loss": 0.5284,
       "step": 1285
     },
     {
-      "epoch": 0.05,
-      "learning_rate": 1.5126641651031896e-05,
-      "loss": 0.3618,
+      "epoch": 0.04649151259595632,
+      "grad_norm": 0.18024949729442596,
+      "learning_rate": 1.549363439827048e-05,
+      "loss": 0.5375,
       "step": 1290
     },
     {
-      "epoch": 0.05,
-      "learning_rate": 1.5185272045028143e-05,
-      "loss": 0.3801,
+      "epoch": 0.046671712257180954,
+      "grad_norm": 0.22767312824726105,
+      "learning_rate": 1.5553687244775402e-05,
+      "loss": 0.5214,
       "step": 1295
     },
     {
-      "epoch": 0.05,
-      "learning_rate": 1.524390243902439e-05,
-      "loss": 0.4104,
+      "epoch": 0.04685191191840559,
+      "grad_norm": 0.21435782313346863,
+      "learning_rate": 1.5613740091280328e-05,
+      "loss": 0.4854,
       "step": 1300
     },
     {
-      "epoch": 0.05,
-      "learning_rate": 1.530253283302064e-05,
-      "loss": 0.4028,
+      "epoch": 0.04703211157963023,
+      "grad_norm": 0.22171509265899658,
+      "learning_rate": 1.567379293778525e-05,
+      "loss": 0.5017,
       "step": 1305
     },
     {
-      "epoch": 0.05,
-      "learning_rate": 1.5361163227016885e-05,
-      "loss": 0.3972,
+      "epoch": 0.04721231124085487,
+      "grad_norm": 0.23257885873317719,
+      "learning_rate": 1.5733845784290176e-05,
+      "loss": 0.5319,
       "step": 1310
     },
     {
-      "epoch": 0.05,
-      "learning_rate": 1.5419793621013134e-05,
-      "loss": 0.4118,
+      "epoch": 0.04739251090207951,
+      "grad_norm": 0.24968785047531128,
+      "learning_rate": 1.57938986307951e-05,
+      "loss": 0.5511,
       "step": 1315
     },
     {
-      "epoch": 0.05,
-      "learning_rate": 1.5478424015009383e-05,
-      "loss": 0.3466,
+      "epoch": 0.047572710563304144,
+      "grad_norm": 0.19016103446483612,
+      "learning_rate": 1.5853951477300024e-05,
+      "loss": 0.5075,
       "step": 1320
     },
     {
-      "epoch": 0.05,
-      "learning_rate": 1.5537054409005632e-05,
-      "loss": 0.4084,
+      "epoch": 0.04775291022452878,
+      "grad_norm": 0.19356337189674377,
+      "learning_rate": 1.591400432380495e-05,
+      "loss": 0.5068,
       "step": 1325
     },
     {
-      "epoch": 0.05,
-      "learning_rate": 1.5595684803001878e-05,
-      "loss": 0.3734,
+      "epoch": 0.04793310988575342,
+      "grad_norm": 0.28713881969451904,
+      "learning_rate": 1.597405717030987e-05,
+      "loss": 0.5403,
       "step": 1330
     },
     {
-      "epoch": 0.05,
-      "learning_rate": 1.5654315196998123e-05,
-      "loss": 0.4115,
+      "epoch": 0.04811330954697805,
+      "grad_norm": 0.18049970269203186,
+      "learning_rate": 1.60341100168148e-05,
+      "loss": 0.5184,
       "step": 1335
     },
     {
-      "epoch": 0.05,
-      "learning_rate": 1.5712945590994372e-05,
-      "loss": 0.382,
+      "epoch": 0.04829350920820269,
+      "grad_norm": 0.22419008612632751,
+      "learning_rate": 1.6094162863319722e-05,
+      "loss": 0.5497,
       "step": 1340
     },
     {
-      "epoch": 0.05,
-      "learning_rate": 1.5771575984990618e-05,
-      "loss": 0.361,
+      "epoch": 0.048473708869427326,
+      "grad_norm": 0.24275891482830048,
+      "learning_rate": 1.6154215709824648e-05,
+      "loss": 0.5375,
       "step": 1345
     },
     {
-      "epoch": 0.05,
-      "learning_rate": 1.5830206378986867e-05,
-      "loss": 0.413,
+      "epoch": 0.04865390853065196,
+      "grad_norm": 0.21218125522136688,
+      "learning_rate": 1.621426855632957e-05,
+      "loss": 0.5069,
       "step": 1350
     },
     {
-      "epoch": 0.05,
-      "learning_rate": 1.5888836772983116e-05,
-      "loss": 0.3922,
+      "epoch": 0.0488341081918766,
+      "grad_norm": 0.17719610035419464,
+      "learning_rate": 1.6274321402834496e-05,
+      "loss": 0.5194,
       "step": 1355
     },
     {
-      "epoch": 0.05,
-      "learning_rate": 1.594746716697936e-05,
-      "loss": 0.3477,
+      "epoch": 0.049014307853101236,
+      "grad_norm": 0.27654048800468445,
+      "learning_rate": 1.633437424933942e-05,
+      "loss": 0.5144,
       "step": 1360
     },
     {
-      "epoch": 0.05,
-      "learning_rate": 1.600609756097561e-05,
-      "loss": 0.386,
+      "epoch": 0.04919450751432587,
+      "grad_norm": 0.20031145215034485,
+      "learning_rate": 1.6394427095844343e-05,
+      "loss": 0.5053,
       "step": 1365
     },
     {
-      "epoch": 0.05,
-      "learning_rate": 1.606472795497186e-05,
-      "loss": 0.3448,
+      "epoch": 0.04937470717555051,
+      "grad_norm": 0.21770700812339783,
+      "learning_rate": 1.645447994234927e-05,
+      "loss": 0.5087,
       "step": 1370
     },
     {
-      "epoch": 0.05,
-      "learning_rate": 1.6123358348968105e-05,
-      "loss": 0.3657,
+      "epoch": 0.049554906836775145,
+      "grad_norm": 0.26108381152153015,
+      "learning_rate": 1.651453278885419e-05,
+      "loss": 0.5614,
       "step": 1375
     },
     {
-      "epoch": 0.05,
-      "learning_rate": 1.6181988742964354e-05,
-      "loss": 0.3743,
+      "epoch": 0.04973510649799978,
+      "grad_norm": 0.21475346386432648,
+      "learning_rate": 1.6574585635359117e-05,
+      "loss": 0.533,
       "step": 1380
     },
     {
-      "epoch": 0.05,
-      "learning_rate": 1.62406191369606e-05,
-      "loss": 0.392,
+      "epoch": 0.04991530615922442,
+      "grad_norm": 0.2189481258392334,
+      "learning_rate": 1.663463848186404e-05,
+      "loss": 0.5106,
       "step": 1385
     },
     {
-      "epoch": 0.05,
-      "learning_rate": 1.629924953095685e-05,
-      "loss": 0.3744,
+      "epoch": 0.050095505820449054,
+      "grad_norm": 0.22523830831050873,
+      "learning_rate": 1.6694691328368968e-05,
+      "loss": 0.5047,
       "step": 1390
     },
     {
-      "epoch": 0.05,
-      "learning_rate": 1.6357879924953094e-05,
-      "loss": 0.383,
+      "epoch": 0.0502757054816737,
+      "grad_norm": 0.2538076639175415,
+      "learning_rate": 1.675474417487389e-05,
+      "loss": 0.4805,
       "step": 1395
     },
     {
-      "epoch": 0.05,
-      "learning_rate": 1.6416510318949343e-05,
-      "loss": 0.3739,
+      "epoch": 0.050455905142898334,
+      "grad_norm": 0.23015667498111725,
+      "learning_rate": 1.6814797021378816e-05,
+      "loss": 0.5236,
       "step": 1400
     },
     {
-      "epoch": 0.05,
-      "learning_rate": 1.6475140712945592e-05,
-      "loss": 0.3644,
+      "epoch": 0.05063610480412297,
+      "grad_norm": 0.26876676082611084,
+      "learning_rate": 1.6874849867883738e-05,
+      "loss": 0.541,
       "step": 1405
     },
     {
-      "epoch": 0.05,
-      "learning_rate": 1.6533771106941838e-05,
-      "loss": 0.3429,
+      "epoch": 0.05081630446534761,
+      "grad_norm": 0.25886017084121704,
+      "learning_rate": 1.6934902714388663e-05,
+      "loss": 0.5363,
       "step": 1410
     },
     {
-      "epoch": 0.05,
-      "learning_rate": 1.6592401500938087e-05,
-      "loss": 0.3876,
+      "epoch": 0.050996504126572244,
+      "grad_norm": 0.22038790583610535,
+      "learning_rate": 1.699495556089359e-05,
+      "loss": 0.4836,
       "step": 1415
     },
     {
-      "epoch": 0.05,
-      "learning_rate": 1.6651031894934336e-05,
-      "loss": 0.3877,
+      "epoch": 0.05117670378779688,
+      "grad_norm": 0.22495049238204956,
+      "learning_rate": 1.705500840739851e-05,
+      "loss": 0.5435,
       "step": 1420
     },
     {
-      "epoch": 0.05,
-      "learning_rate": 1.670966228893058e-05,
-      "loss": 0.4003,
+      "epoch": 0.05135690344902152,
+      "grad_norm": 0.22736889123916626,
+      "learning_rate": 1.7115061253903437e-05,
+      "loss": 0.4876,
       "step": 1425
     },
     {
-      "epoch": 0.05,
-      "learning_rate": 1.676829268292683e-05,
-      "loss": 0.3824,
+      "epoch": 0.05153710311024615,
+      "grad_norm": 0.21810127794742584,
+      "learning_rate": 1.717511410040836e-05,
+      "loss": 0.5353,
       "step": 1430
     },
     {
-      "epoch": 0.05,
-      "learning_rate": 1.682692307692308e-05,
-      "loss": 0.375,
+      "epoch": 0.05171730277147079,
+      "grad_norm": 0.19743861258029938,
+      "learning_rate": 1.7235166946913284e-05,
+      "loss": 0.5042,
       "step": 1435
     },
     {
-      "epoch": 0.05,
-      "learning_rate": 1.6885553470919325e-05,
-      "loss": 0.3732,
+      "epoch": 0.051897502432695426,
+      "grad_norm": 0.22972045838832855,
+      "learning_rate": 1.7295219793418207e-05,
+      "loss": 0.4953,
       "step": 1440
     },
     {
-      "epoch": 0.05,
-      "learning_rate": 1.694418386491557e-05,
-      "loss": 0.3874,
+      "epoch": 0.05207770209392006,
+      "grad_norm": 0.19471551477909088,
+      "learning_rate": 1.7355272639923136e-05,
+      "loss": 0.5055,
       "step": 1445
     },
     {
-      "epoch": 0.05,
-      "learning_rate": 1.700281425891182e-05,
-      "loss": 0.3687,
+      "epoch": 0.0522579017551447,
+      "grad_norm": 0.29399409890174866,
+      "learning_rate": 1.7415325486428058e-05,
+      "loss": 0.547,
       "step": 1450
     },
     {
-      "epoch": 0.05,
-      "learning_rate": 1.706144465290807e-05,
-      "loss": 0.3991,
+      "epoch": 0.052438101416369336,
+      "grad_norm": 0.24069805443286896,
+      "learning_rate": 1.7475378332932983e-05,
+      "loss": 0.5267,
       "step": 1455
     },
     {
-      "epoch": 0.05,
-      "learning_rate": 1.7120075046904315e-05,
-      "loss": 0.3493,
+      "epoch": 0.05261830107759397,
+      "grad_norm": 0.23223073780536652,
+      "learning_rate": 1.7535431179437905e-05,
+      "loss": 0.4945,
       "step": 1460
     },
     {
-      "epoch": 0.05,
-      "learning_rate": 1.7178705440900564e-05,
-      "loss": 0.3338,
+      "epoch": 0.05279850073881861,
+      "grad_norm": 0.19976119697093964,
+      "learning_rate": 1.759548402594283e-05,
+      "loss": 0.5135,
       "step": 1465
     },
     {
-      "epoch": 0.05,
-      "learning_rate": 1.7237335834896813e-05,
-      "loss": 0.3733,
+      "epoch": 0.052978700400043245,
+      "grad_norm": 0.2490244060754776,
+      "learning_rate": 1.7655536872447757e-05,
+      "loss": 0.5031,
       "step": 1470
     },
     {
-      "epoch": 0.05,
-      "learning_rate": 1.7295966228893058e-05,
-      "loss": 0.3407,
+      "epoch": 0.05315890006126788,
+      "grad_norm": 0.23295678198337555,
+      "learning_rate": 1.771558971895268e-05,
+      "loss": 0.5652,
       "step": 1475
     },
     {
-      "epoch": 0.05,
-      "learning_rate": 1.7354596622889307e-05,
-      "loss": 0.3743,
+      "epoch": 0.053339099722492525,
+      "grad_norm": 0.22554819285869598,
+      "learning_rate": 1.7775642565457604e-05,
+      "loss": 0.5279,
       "step": 1480
     },
     {
-      "epoch": 0.05,
-      "learning_rate": 1.7413227016885556e-05,
-      "loss": 0.3911,
+      "epoch": 0.05351929938371716,
+      "grad_norm": 0.2634689509868622,
+      "learning_rate": 1.7835695411962526e-05,
+      "loss": 0.5486,
       "step": 1485
     },
     {
-      "epoch": 0.05,
-      "learning_rate": 1.7471857410881802e-05,
-      "loss": 0.3591,
+      "epoch": 0.0536994990449418,
+      "grad_norm": 0.1896762251853943,
+      "learning_rate": 1.7895748258467452e-05,
+      "loss": 0.5419,
       "step": 1490
     },
     {
-      "epoch": 0.05,
-      "learning_rate": 1.7530487804878047e-05,
-      "loss": 0.3793,
+      "epoch": 0.053879698706166435,
+      "grad_norm": 0.21280480921268463,
+      "learning_rate": 1.7955801104972378e-05,
+      "loss": 0.5541,
       "step": 1495
     },
     {
-      "epoch": 0.05,
-      "learning_rate": 1.7589118198874296e-05,
-      "loss": 0.3851,
+      "epoch": 0.05405989836739107,
+      "grad_norm": 0.19312447309494019,
+      "learning_rate": 1.8015853951477303e-05,
+      "loss": 0.5315,
       "step": 1500
     },
     {
-      "epoch": 0.05,
-      "eval_loss": 0.3775494694709778,
-      "eval_runtime": 10.5367,
-      "eval_samples_per_second": 9.491,
-      "eval_steps_per_second": 9.491,
+      "epoch": 0.05405989836739107,
+      "eval_loss": 0.5359941720962524,
+      "eval_runtime": 3.5176,
+      "eval_samples_per_second": 28.429,
+      "eval_steps_per_second": 7.107,
       "step": 1500
     },
     {
-      "epoch": 0.05,
-      "learning_rate": 1.7647748592870545e-05,
-      "loss": 0.3634,
+      "epoch": 0.05424009802861571,
+      "grad_norm": 0.22356046736240387,
+      "learning_rate": 1.8075906797982225e-05,
+      "loss": 0.5439,
       "step": 1505
     },
     {
-      "epoch": 0.05,
-      "learning_rate": 1.770637898686679e-05,
-      "loss": 0.4114,
+      "epoch": 0.054420297689840344,
+      "grad_norm": 0.24172358214855194,
+      "learning_rate": 1.813595964448715e-05,
+      "loss": 0.5436,
       "step": 1510
     },
     {
-      "epoch": 0.05,
-      "learning_rate": 1.776500938086304e-05,
-      "loss": 0.378,
+      "epoch": 0.05460049735106498,
+      "grad_norm": 0.25496965646743774,
+      "learning_rate": 1.8196012490992073e-05,
+      "loss": 0.5427,
       "step": 1515
     },
     {
-      "epoch": 0.05,
-      "learning_rate": 1.782363977485929e-05,
-      "loss": 0.3722,
+      "epoch": 0.05478069701228962,
+      "grad_norm": 0.21974816918373108,
+      "learning_rate": 1.8256065337497e-05,
+      "loss": 0.5261,
       "step": 1520
     },
     {
-      "epoch": 0.05,
-      "learning_rate": 1.7882270168855535e-05,
-      "loss": 0.3827,
+      "epoch": 0.05496089667351425,
+      "grad_norm": 0.2528958320617676,
+      "learning_rate": 1.8316118184001924e-05,
+      "loss": 0.5123,
       "step": 1525
     },
     {
-      "epoch": 0.05,
-      "learning_rate": 1.7940900562851784e-05,
-      "loss": 0.383,
+      "epoch": 0.05514109633473889,
+      "grad_norm": 0.2141297310590744,
+      "learning_rate": 1.8376171030506846e-05,
+      "loss": 0.4936,
       "step": 1530
     },
     {
-      "epoch": 0.05,
-      "learning_rate": 1.7999530956848033e-05,
-      "loss": 0.3992,
+      "epoch": 0.055321295995963526,
+      "grad_norm": 0.2709108293056488,
+      "learning_rate": 1.8436223877011772e-05,
+      "loss": 0.5397,
       "step": 1535
     },
     {
-      "epoch": 0.05,
-      "learning_rate": 1.805816135084428e-05,
-      "loss": 0.3733,
+      "epoch": 0.05550149565718816,
+      "grad_norm": 0.2779485285282135,
+      "learning_rate": 1.8496276723516694e-05,
+      "loss": 0.5001,
       "step": 1540
     },
     {
-      "epoch": 0.05,
-      "learning_rate": 1.8116791744840524e-05,
-      "loss": 0.3744,
+      "epoch": 0.0556816953184128,
+      "grad_norm": 0.22084900736808777,
+      "learning_rate": 1.855632957002162e-05,
+      "loss": 0.4713,
       "step": 1545
     },
     {
-      "epoch": 0.05,
-      "learning_rate": 1.8175422138836773e-05,
-      "loss": 0.3748,
+      "epoch": 0.055861894979637436,
+      "grad_norm": 0.1823149174451828,
+      "learning_rate": 1.8616382416526545e-05,
+      "loss": 0.4974,
       "step": 1550
     },
     {
-      "epoch": 0.05,
-      "learning_rate": 1.8234052532833022e-05,
-      "loss": 0.3561,
+      "epoch": 0.05604209464086207,
+      "grad_norm": 0.20994675159454346,
+      "learning_rate": 1.867643526303147e-05,
+      "loss": 0.5241,
       "step": 1555
     },
     {
-      "epoch": 0.05,
-      "learning_rate": 1.8292682926829268e-05,
-      "loss": 0.3616,
+      "epoch": 0.05622229430208671,
+      "grad_norm": 0.24369290471076965,
+      "learning_rate": 1.8736488109536393e-05,
+      "loss": 0.5478,
       "step": 1560
     },
     {
-      "epoch": 0.06,
-      "learning_rate": 1.8351313320825517e-05,
-      "loss": 0.399,
+      "epoch": 0.05640249396331135,
+      "grad_norm": 0.259897381067276,
+      "learning_rate": 1.879654095604132e-05,
+      "loss": 0.4991,
       "step": 1565
     },
     {
-      "epoch": 0.06,
-      "learning_rate": 1.8409943714821766e-05,
-      "loss": 0.375,
+      "epoch": 0.05658269362453599,
+      "grad_norm": 0.25176161527633667,
+      "learning_rate": 1.885659380254624e-05,
+      "loss": 0.5066,
       "step": 1570
     },
     {
-      "epoch": 0.06,
-      "learning_rate": 1.846857410881801e-05,
-      "loss": 0.3917,
+      "epoch": 0.056762893285760625,
+      "grad_norm": 0.2606624364852905,
+      "learning_rate": 1.8916646649051163e-05,
+      "loss": 0.4857,
       "step": 1575
     },
     {
-      "epoch": 0.06,
-      "learning_rate": 1.852720450281426e-05,
-      "loss": 0.3619,
+      "epoch": 0.05694309294698526,
+      "grad_norm": 0.197222039103508,
+      "learning_rate": 1.8976699495556092e-05,
+      "loss": 0.5225,
       "step": 1580
     },
     {
-      "epoch": 0.06,
-      "learning_rate": 1.858583489681051e-05,
-      "loss": 0.3954,
+      "epoch": 0.0571232926082099,
+      "grad_norm": 0.24445274472236633,
+      "learning_rate": 1.9036752342061014e-05,
+      "loss": 0.4929,
       "step": 1585
     },
     {
-      "epoch": 0.06,
-      "learning_rate": 1.8644465290806755e-05,
-      "loss": 0.3946,
+      "epoch": 0.057303492269434535,
+      "grad_norm": 0.21949337422847748,
+      "learning_rate": 1.909680518856594e-05,
+      "loss": 0.5149,
       "step": 1590
     },
     {
-      "epoch": 0.06,
-      "learning_rate": 1.8703095684803e-05,
-      "loss": 0.3951,
+      "epoch": 0.05748369193065917,
+      "grad_norm": 0.19175410270690918,
+      "learning_rate": 1.9156858035070862e-05,
+      "loss": 0.4939,
       "step": 1595
     },
     {
-      "epoch": 0.06,
-      "learning_rate": 1.876172607879925e-05,
-      "loss": 0.3481,
+      "epoch": 0.05766389159188381,
+      "grad_norm": 0.2463282346725464,
+      "learning_rate": 1.9216910881575787e-05,
+      "loss": 0.5435,
       "step": 1600
     },
     {
-      "epoch": 0.06,
-      "learning_rate": 1.88203564727955e-05,
-      "loss": 0.3811,
+      "epoch": 0.057844091253108444,
+      "grad_norm": 0.2281898409128189,
+      "learning_rate": 1.9276963728080713e-05,
+      "loss": 0.5538,
       "step": 1605
     },
     {
-      "epoch": 0.06,
-      "learning_rate": 1.8878986866791744e-05,
-      "loss": 0.3951,
+      "epoch": 0.05802429091433308,
+      "grad_norm": 0.2389582097530365,
+      "learning_rate": 1.933701657458564e-05,
+      "loss": 0.5149,
       "step": 1610
     },
     {
-      "epoch": 0.06,
-      "learning_rate": 1.8937617260787993e-05,
-      "loss": 0.3794,
+      "epoch": 0.05820449057555772,
+      "grad_norm": 0.23086562752723694,
+      "learning_rate": 1.939706942109056e-05,
+      "loss": 0.5109,
       "step": 1615
     },
     {
-      "epoch": 0.06,
-      "learning_rate": 1.8996247654784242e-05,
-      "loss": 0.4044,
+      "epoch": 0.058384690236782354,
+      "grad_norm": 0.2583639621734619,
+      "learning_rate": 1.9457122267595486e-05,
+      "loss": 0.5174,
       "step": 1620
     },
     {
-      "epoch": 0.06,
-      "learning_rate": 1.9054878048780488e-05,
-      "loss": 0.3489,
+      "epoch": 0.05856488989800699,
+      "grad_norm": 0.20627906918525696,
+      "learning_rate": 1.951717511410041e-05,
+      "loss": 0.5224,
       "step": 1625
     },
     {
-      "epoch": 0.06,
-      "learning_rate": 1.9113508442776737e-05,
-      "loss": 0.3395,
+      "epoch": 0.05874508955923163,
+      "grad_norm": 0.30713996291160583,
+      "learning_rate": 1.9577227960605334e-05,
+      "loss": 0.5218,
       "step": 1630
     },
     {
-      "epoch": 0.06,
-      "learning_rate": 1.9172138836772986e-05,
-      "loss": 0.4209,
+      "epoch": 0.05892528922045626,
+      "grad_norm": 0.26830917596817017,
+      "learning_rate": 1.963728080711026e-05,
+      "loss": 0.5101,
       "step": 1635
     },
     {
-      "epoch": 0.06,
-      "learning_rate": 1.923076923076923e-05,
-      "loss": 0.3852,
+      "epoch": 0.0591054888816809,
+      "grad_norm": 0.21486251056194305,
+      "learning_rate": 1.969733365361518e-05,
+      "loss": 0.4967,
       "step": 1640
     },
     {
-      "epoch": 0.06,
-      "learning_rate": 1.928939962476548e-05,
-      "loss": 0.3539,
+      "epoch": 0.059285688542905536,
+      "grad_norm": 0.26688727736473083,
+      "learning_rate": 1.9757386500120107e-05,
+      "loss": 0.5322,
       "step": 1645
     },
     {
-      "epoch": 0.06,
-      "learning_rate": 1.9348030018761726e-05,
-      "loss": 0.4027,
+      "epoch": 0.05946588820413018,
+      "grad_norm": 0.23560921847820282,
+      "learning_rate": 1.981743934662503e-05,
+      "loss": 0.491,
       "step": 1650
     },
     {
-      "epoch": 0.06,
-      "learning_rate": 1.9406660412757975e-05,
-      "loss": 0.369,
+      "epoch": 0.059646087865354816,
+      "grad_norm": 0.27083316445350647,
+      "learning_rate": 1.9877492193129955e-05,
+      "loss": 0.5475,
       "step": 1655
     },
     {
-      "epoch": 0.06,
-      "learning_rate": 1.946529080675422e-05,
-      "loss": 0.3624,
+      "epoch": 0.05982628752657945,
+      "grad_norm": 0.24693989753723145,
+      "learning_rate": 1.993754503963488e-05,
+      "loss": 0.5349,
       "step": 1660
     },
     {
-      "epoch": 0.06,
-      "learning_rate": 1.952392120075047e-05,
-      "loss": 0.365,
+      "epoch": 0.06000648718780409,
+      "grad_norm": 0.2298416942358017,
+      "learning_rate": 1.9997597886139806e-05,
+      "loss": 0.5048,
       "step": 1665
     },
     {
-      "epoch": 0.06,
-      "learning_rate": 1.958255159474672e-05,
-      "loss": 0.3642,
+      "epoch": 0.060186686849028725,
+      "grad_norm": 0.2370372712612152,
+      "learning_rate": 2.0057650732644728e-05,
+      "loss": 0.4959,
       "step": 1670
     },
     {
-      "epoch": 0.06,
-      "learning_rate": 1.9641181988742964e-05,
-      "loss": 0.3889,
+      "epoch": 0.06036688651025336,
+      "grad_norm": 0.25343507528305054,
+      "learning_rate": 2.011770357914965e-05,
+      "loss": 0.5119,
       "step": 1675
     },
     {
-      "epoch": 0.06,
-      "learning_rate": 1.9699812382739213e-05,
-      "loss": 0.3828,
+      "epoch": 0.060547086171478,
+      "grad_norm": 0.28533029556274414,
+      "learning_rate": 2.0177756425654576e-05,
+      "loss": 0.5539,
       "step": 1680
     },
     {
-      "epoch": 0.06,
-      "learning_rate": 1.9758442776735462e-05,
-      "loss": 0.3699,
+      "epoch": 0.060727285832702635,
+      "grad_norm": 0.2781274914741516,
+      "learning_rate": 2.02378092721595e-05,
+      "loss": 0.4953,
       "step": 1685
     },
     {
-      "epoch": 0.06,
-      "learning_rate": 1.9817073170731708e-05,
-      "loss": 0.3726,
+      "epoch": 0.06090748549392727,
+      "grad_norm": 0.33222508430480957,
+      "learning_rate": 2.0297862118664427e-05,
+      "loss": 0.5152,
       "step": 1690
     },
     {
-      "epoch": 0.06,
-      "learning_rate": 1.9875703564727957e-05,
-      "loss": 0.3941,
+      "epoch": 0.06108768515515191,
+      "grad_norm": 0.2313944697380066,
+      "learning_rate": 2.035791496516935e-05,
+      "loss": 0.5293,
       "step": 1695
     },
     {
-      "epoch": 0.06,
-      "learning_rate": 1.9934333958724203e-05,
-      "loss": 0.3476,
+      "epoch": 0.061267884816376544,
+      "grad_norm": 0.28105735778808594,
+      "learning_rate": 2.0417967811674275e-05,
+      "loss": 0.5141,
       "step": 1700
     },
     {
-      "epoch": 0.06,
-      "learning_rate": 1.999296435272045e-05,
-      "loss": 0.3631,
+      "epoch": 0.06144808447760118,
+      "grad_norm": 0.2256317436695099,
+      "learning_rate": 2.0478020658179197e-05,
+      "loss": 0.4813,
       "step": 1705
     },
     {
-      "epoch": 0.06,
-      "learning_rate": 2.0051594746716697e-05,
-      "loss": 0.3619,
+      "epoch": 0.06162828413882582,
+      "grad_norm": 0.23726746439933777,
+      "learning_rate": 2.0538073504684126e-05,
+      "loss": 0.5469,
       "step": 1710
     },
     {
-      "epoch": 0.06,
-      "learning_rate": 2.0110225140712946e-05,
-      "loss": 0.3699,
+      "epoch": 0.061808483800050454,
+      "grad_norm": 0.27420201897621155,
+      "learning_rate": 2.0598126351189048e-05,
+      "loss": 0.4634,
       "step": 1715
     },
     {
-      "epoch": 0.06,
-      "learning_rate": 2.0168855534709195e-05,
-      "loss": 0.3832,
+      "epoch": 0.06198868346127509,
+      "grad_norm": 0.2140614092350006,
+      "learning_rate": 2.065817919769397e-05,
+      "loss": 0.5011,
       "step": 1720
     },
     {
-      "epoch": 0.06,
-      "learning_rate": 2.022748592870544e-05,
-      "loss": 0.3747,
+      "epoch": 0.06216888312249973,
+      "grad_norm": 0.21357890963554382,
+      "learning_rate": 2.0718232044198896e-05,
+      "loss": 0.4775,
       "step": 1725
     },
     {
-      "epoch": 0.06,
-      "learning_rate": 2.028611632270169e-05,
-      "loss": 0.3727,
+      "epoch": 0.06234908278372436,
+      "grad_norm": 0.2980591058731079,
+      "learning_rate": 2.0778284890703818e-05,
+      "loss": 0.5457,
       "step": 1730
     },
     {
-      "epoch": 0.06,
-      "learning_rate": 2.034474671669794e-05,
-      "loss": 0.3725,
+      "epoch": 0.062529282444949,
+      "grad_norm": 0.2486668825149536,
+      "learning_rate": 2.0838337737208744e-05,
+      "loss": 0.4759,
       "step": 1735
     },
     {
-      "epoch": 0.06,
-      "learning_rate": 2.0403377110694185e-05,
-      "loss": 0.3771,
+      "epoch": 0.06270948210617364,
+      "grad_norm": 0.24250219762325287,
+      "learning_rate": 2.089839058371367e-05,
+      "loss": 0.5026,
       "step": 1740
     },
     {
-      "epoch": 0.06,
-      "learning_rate": 2.0462007504690434e-05,
-      "loss": 0.3689,
+      "epoch": 0.06288968176739827,
+      "grad_norm": 0.2787621319293976,
+      "learning_rate": 2.0958443430218595e-05,
+      "loss": 0.5464,
       "step": 1745
     },
     {
-      "epoch": 0.06,
-      "learning_rate": 2.0520637898686683e-05,
-      "loss": 0.3746,
+      "epoch": 0.06306988142862291,
+      "grad_norm": 0.23946985602378845,
+      "learning_rate": 2.1018496276723517e-05,
+      "loss": 0.5052,
       "step": 1750
     },
     {
-      "epoch": 0.06,
-      "learning_rate": 2.0579268292682928e-05,
-      "loss": 0.3559,
+      "epoch": 0.06325008108984755,
+      "grad_norm": 0.26933711767196655,
+      "learning_rate": 2.1078549123228443e-05,
+      "loss": 0.5287,
       "step": 1755
     },
     {
-      "epoch": 0.06,
-      "learning_rate": 2.0637898686679174e-05,
-      "loss": 0.3715,
+      "epoch": 0.06343028075107218,
+      "grad_norm": 0.2348264902830124,
+      "learning_rate": 2.1138601969733365e-05,
+      "loss": 0.5193,
       "step": 1760
     },
     {
-      "epoch": 0.06,
-      "learning_rate": 2.0696529080675423e-05,
-      "loss": 0.3851,
+      "epoch": 0.06361048041229682,
+      "grad_norm": 0.26587164402008057,
+      "learning_rate": 2.1198654816238294e-05,
+      "loss": 0.5169,
       "step": 1765
     },
     {
-      "epoch": 0.06,
-      "learning_rate": 2.0755159474671672e-05,
-      "loss": 0.3851,
+      "epoch": 0.06379068007352146,
+      "grad_norm": 0.2152547687292099,
+      "learning_rate": 2.1258707662743216e-05,
+      "loss": 0.5034,
       "step": 1770
     },
     {
-      "epoch": 0.06,
-      "learning_rate": 2.0813789868667917e-05,
-      "loss": 0.3639,
+      "epoch": 0.0639708797347461,
+      "grad_norm": 0.27491626143455505,
+      "learning_rate": 2.1318760509248138e-05,
+      "loss": 0.5476,
       "step": 1775
     },
     {
-      "epoch": 0.06,
-      "learning_rate": 2.0872420262664166e-05,
-      "loss": 0.3669,
+      "epoch": 0.06415107939597074,
+      "grad_norm": 0.22118522226810455,
+      "learning_rate": 2.1378813355753064e-05,
+      "loss": 0.5291,
       "step": 1780
     },
     {
-      "epoch": 0.06,
-      "learning_rate": 2.0931050656660415e-05,
-      "loss": 0.3942,
+      "epoch": 0.06433127905719538,
+      "grad_norm": 0.2910890281200409,
+      "learning_rate": 2.1438866202257986e-05,
+      "loss": 0.5183,
       "step": 1785
     },
     {
-      "epoch": 0.06,
-      "learning_rate": 2.098968105065666e-05,
-      "loss": 0.3495,
+      "epoch": 0.06451147871842002,
+      "grad_norm": 0.2631491720676422,
+      "learning_rate": 2.149891904876291e-05,
+      "loss": 0.4692,
       "step": 1790
     },
     {
-      "epoch": 0.06,
-      "learning_rate": 2.104831144465291e-05,
-      "loss": 0.3516,
+      "epoch": 0.06469167837964465,
+      "grad_norm": 0.2912589907646179,
+      "learning_rate": 2.1558971895267837e-05,
+      "loss": 0.5036,
       "step": 1795
     },
     {
-      "epoch": 0.06,
-      "learning_rate": 2.110694183864916e-05,
-      "loss": 0.3966,
+      "epoch": 0.06487187804086929,
+      "grad_norm": 0.2575485408306122,
+      "learning_rate": 2.1619024741772762e-05,
+      "loss": 0.4924,
       "step": 1800
     },
     {
-      "epoch": 0.06,
-      "learning_rate": 2.1165572232645405e-05,
-      "loss": 0.3839,
+      "epoch": 0.06505207770209392,
+      "grad_norm": 0.24182820320129395,
+      "learning_rate": 2.1679077588277685e-05,
+      "loss": 0.4854,
       "step": 1805
     },
     {
-      "epoch": 0.06,
-      "learning_rate": 2.122420262664165e-05,
-      "loss": 0.3659,
+      "epoch": 0.06523227736331856,
+      "grad_norm": 0.23146407306194305,
+      "learning_rate": 2.173913043478261e-05,
+      "loss": 0.4882,
       "step": 1810
     },
     {
-      "epoch": 0.06,
-      "learning_rate": 2.12828330206379e-05,
-      "loss": 0.3885,
+      "epoch": 0.0654124770245432,
+      "grad_norm": 0.2757367789745331,
+      "learning_rate": 2.1799183281287532e-05,
+      "loss": 0.4886,
       "step": 1815
     },
     {
-      "epoch": 0.06,
-      "learning_rate": 2.134146341463415e-05,
-      "loss": 0.3648,
+      "epoch": 0.06559267668576783,
+      "grad_norm": 0.22752000391483307,
+      "learning_rate": 2.1859236127792458e-05,
+      "loss": 0.49,
       "step": 1820
     },
     {
-      "epoch": 0.06,
-      "learning_rate": 2.1400093808630394e-05,
-      "loss": 0.38,
+      "epoch": 0.06577287634699247,
+      "grad_norm": 0.2603028416633606,
+      "learning_rate": 2.1919288974297383e-05,
+      "loss": 0.4861,
       "step": 1825
     },
     {
-      "epoch": 0.06,
-      "learning_rate": 2.1458724202626643e-05,
-      "loss": 0.3699,
+      "epoch": 0.0659530760082171,
+      "grad_norm": 0.226707324385643,
+      "learning_rate": 2.1979341820802306e-05,
+      "loss": 0.4598,
       "step": 1830
     },
     {
-      "epoch": 0.06,
-      "learning_rate": 2.1517354596622892e-05,
-      "loss": 0.3835,
+      "epoch": 0.06613327566944174,
+      "grad_norm": 0.2270716428756714,
+      "learning_rate": 2.203939466730723e-05,
+      "loss": 0.5174,
       "step": 1835
     },
     {
-      "epoch": 0.06,
-      "learning_rate": 2.1575984990619138e-05,
-      "loss": 0.3604,
+      "epoch": 0.06631347533066638,
+      "grad_norm": 0.2932356595993042,
+      "learning_rate": 2.2099447513812153e-05,
+      "loss": 0.5836,
       "step": 1840
     },
     {
-      "epoch": 0.06,
-      "learning_rate": 2.1634615384615387e-05,
-      "loss": 0.3876,
+      "epoch": 0.06649367499189102,
+      "grad_norm": 0.24968333542346954,
+      "learning_rate": 2.2159500360317082e-05,
+      "loss": 0.5215,
       "step": 1845
     },
     {
-      "epoch": 0.07,
-      "learning_rate": 2.1693245778611636e-05,
-      "loss": 0.3233,
+      "epoch": 0.06667387465311565,
+      "grad_norm": 0.3082146942615509,
+      "learning_rate": 2.2219553206822005e-05,
+      "loss": 0.5236,
       "step": 1850
     },
     {
-      "epoch": 0.07,
-      "learning_rate": 2.175187617260788e-05,
-      "loss": 0.3619,
+      "epoch": 0.06685407431434029,
+      "grad_norm": 0.2535383999347687,
+      "learning_rate": 2.227960605332693e-05,
+      "loss": 0.4725,
       "step": 1855
     },
     {
-      "epoch": 0.07,
-      "learning_rate": 2.1810506566604127e-05,
-      "loss": 0.3788,
+      "epoch": 0.06703427397556493,
+      "grad_norm": 0.380136102437973,
+      "learning_rate": 2.2339658899831852e-05,
+      "loss": 0.5161,
       "step": 1860
     },
     {
-      "epoch": 0.07,
-      "learning_rate": 2.1869136960600376e-05,
-      "loss": 0.3812,
+      "epoch": 0.06721447363678956,
+      "grad_norm": 0.24711187183856964,
+      "learning_rate": 2.2399711746336778e-05,
+      "loss": 0.5302,
       "step": 1865
     },
     {
-      "epoch": 0.07,
-      "learning_rate": 2.1927767354596625e-05,
-      "loss": 0.3977,
+      "epoch": 0.0673946732980142,
+      "grad_norm": 0.285430908203125,
+      "learning_rate": 2.24597645928417e-05,
+      "loss": 0.5437,
       "step": 1870
     },
     {
-      "epoch": 0.07,
-      "learning_rate": 2.198639774859287e-05,
-      "loss": 0.3496,
+      "epoch": 0.06757487295923884,
+      "grad_norm": 0.2444837987422943,
+      "learning_rate": 2.2519817439346626e-05,
+      "loss": 0.4988,
       "step": 1875
     },
     {
-      "epoch": 0.07,
-      "learning_rate": 2.204502814258912e-05,
-      "loss": 0.3678,
+      "epoch": 0.06775507262046347,
+      "grad_norm": 0.3459201455116272,
+      "learning_rate": 2.257987028585155e-05,
+      "loss": 0.5126,
       "step": 1880
     },
     {
-      "epoch": 0.07,
-      "learning_rate": 2.210365853658537e-05,
-      "loss": 0.3521,
+      "epoch": 0.06793527228168811,
+      "grad_norm": 0.23254404962062836,
+      "learning_rate": 2.2639923132356473e-05,
+      "loss": 0.5229,
       "step": 1885
     },
     {
-      "epoch": 0.07,
-      "learning_rate": 2.2162288930581614e-05,
-      "loss": 0.3765,
+      "epoch": 0.06811547194291274,
+      "grad_norm": 0.3149000406265259,
+      "learning_rate": 2.26999759788614e-05,
+      "loss": 0.5091,
       "step": 1890
     },
     {
-      "epoch": 0.07,
-      "learning_rate": 2.2220919324577863e-05,
-      "loss": 0.3787,
+      "epoch": 0.06829567160413738,
+      "grad_norm": 0.31007108092308044,
+      "learning_rate": 2.276002882536632e-05,
+      "loss": 0.487,
       "step": 1895
     },
     {
-      "epoch": 0.07,
-      "learning_rate": 2.2279549718574112e-05,
-      "loss": 0.3749,
+      "epoch": 0.06847587126536202,
+      "grad_norm": 0.25598689913749695,
+      "learning_rate": 2.282008167187125e-05,
+      "loss": 0.5061,
       "step": 1900
     },
     {
-      "epoch": 0.07,
-      "learning_rate": 2.2338180112570358e-05,
-      "loss": 0.3536,
+      "epoch": 0.06865607092658665,
+      "grad_norm": 0.27555733919143677,
+      "learning_rate": 2.2880134518376172e-05,
+      "loss": 0.5177,
       "step": 1905
     },
     {
-      "epoch": 0.07,
-      "learning_rate": 2.2396810506566603e-05,
-      "loss": 0.367,
+      "epoch": 0.06883627058781129,
+      "grad_norm": 0.28343528509140015,
+      "learning_rate": 2.2940187364881098e-05,
+      "loss": 0.5267,
       "step": 1910
     },
     {
-      "epoch": 0.07,
-      "learning_rate": 2.2455440900562852e-05,
-      "loss": 0.3746,
+      "epoch": 0.06901647024903593,
+      "grad_norm": 0.1938653290271759,
+      "learning_rate": 2.300024021138602e-05,
+      "loss": 0.4926,
       "step": 1915
     },
     {
-      "epoch": 0.07,
-      "learning_rate": 2.25140712945591e-05,
-      "loss": 0.3574,
+      "epoch": 0.06919666991026056,
+      "grad_norm": 0.2610265910625458,
+      "learning_rate": 2.3060293057890945e-05,
+      "loss": 0.5187,
       "step": 1920
     },
     {
-      "epoch": 0.07,
-      "learning_rate": 2.2572701688555347e-05,
-      "loss": 0.3853,
+      "epoch": 0.0693768695714852,
+      "grad_norm": 0.20911858975887299,
+      "learning_rate": 2.3120345904395868e-05,
+      "loss": 0.4979,
       "step": 1925
     },
     {
-      "epoch": 0.07,
-      "learning_rate": 2.2631332082551596e-05,
-      "loss": 0.3839,
+      "epoch": 0.06955706923270984,
+      "grad_norm": 0.2716982662677765,
+      "learning_rate": 2.3180398750900793e-05,
+      "loss": 0.4976,
       "step": 1930
     },
     {
-      "epoch": 0.07,
-      "learning_rate": 2.2689962476547845e-05,
-      "loss": 0.3809,
+      "epoch": 0.06973726889393447,
+      "grad_norm": 0.3215219974517822,
+      "learning_rate": 2.324045159740572e-05,
+      "loss": 0.5193,
       "step": 1935
     },
     {
-      "epoch": 0.07,
-      "learning_rate": 2.274859287054409e-05,
-      "loss": 0.3989,
+      "epoch": 0.06991746855515911,
+      "grad_norm": 0.3229633867740631,
+      "learning_rate": 2.330050444391064e-05,
+      "loss": 0.5057,
       "step": 1940
     },
     {
-      "epoch": 0.07,
-      "learning_rate": 2.280722326454034e-05,
-      "loss": 0.3575,
+      "epoch": 0.07009766821638376,
+      "grad_norm": 0.20286689698696136,
+      "learning_rate": 2.3360557290415567e-05,
+      "loss": 0.4657,
       "step": 1945
     },
     {
-      "epoch": 0.07,
-      "learning_rate": 2.286585365853659e-05,
-      "loss": 0.3877,
+      "epoch": 0.0702778678776084,
+      "grad_norm": 0.30475685000419617,
+      "learning_rate": 2.342061013692049e-05,
+      "loss": 0.5316,
       "step": 1950
     },
     {
-      "epoch": 0.07,
-      "learning_rate": 2.2924484052532834e-05,
-      "loss": 0.3573,
+      "epoch": 0.07045806753883303,
+      "grad_norm": 0.273539662361145,
+      "learning_rate": 2.3480662983425418e-05,
+      "loss": 0.5268,
       "step": 1955
     },
     {
-      "epoch": 0.07,
-      "learning_rate": 2.298311444652908e-05,
-      "loss": 0.3377,
+      "epoch": 0.07063826720005767,
+      "grad_norm": 0.2909047305583954,
+      "learning_rate": 2.354071582993034e-05,
+      "loss": 0.4885,
       "step": 1960
     },
     {
-      "epoch": 0.07,
-      "learning_rate": 2.304174484052533e-05,
-      "loss": 0.3755,
+      "epoch": 0.0708184668612823,
+      "grad_norm": 0.23149296641349792,
+      "learning_rate": 2.3600768676435265e-05,
+      "loss": 0.534,
       "step": 1965
     },
     {
-      "epoch": 0.07,
-      "learning_rate": 2.3100375234521575e-05,
-      "loss": 0.373,
+      "epoch": 0.07099866652250694,
+      "grad_norm": 0.32449668645858765,
+      "learning_rate": 2.3660821522940188e-05,
+      "loss": 0.508,
       "step": 1970
     },
     {
-      "epoch": 0.07,
-      "learning_rate": 2.3159005628517824e-05,
-      "loss": 0.3544,
+      "epoch": 0.07117886618373158,
+      "grad_norm": 0.2767580449581146,
+      "learning_rate": 2.3720874369445113e-05,
+      "loss": 0.5106,
       "step": 1975
     },
     {
-      "epoch": 0.07,
-      "learning_rate": 2.3217636022514073e-05,
-      "loss": 0.3509,
+      "epoch": 0.07135906584495622,
+      "grad_norm": 0.3347299098968506,
+      "learning_rate": 2.378092721595004e-05,
+      "loss": 0.4898,
       "step": 1980
     },
     {
-      "epoch": 0.07,
-      "learning_rate": 2.327626641651032e-05,
-      "loss": 0.3458,
+      "epoch": 0.07153926550618085,
+      "grad_norm": 0.2920180559158325,
+      "learning_rate": 2.384098006245496e-05,
+      "loss": 0.5102,
       "step": 1985
     },
     {
-      "epoch": 0.07,
-      "learning_rate": 2.3334896810506567e-05,
-      "loss": 0.3339,
+      "epoch": 0.07171946516740549,
+      "grad_norm": 0.2103186845779419,
+      "learning_rate": 2.3901032908959886e-05,
+      "loss": 0.5,
       "step": 1990
     },
     {
-      "epoch": 0.07,
-      "learning_rate": 2.3393527204502816e-05,
-      "loss": 0.3466,
+      "epoch": 0.07189966482863012,
+      "grad_norm": 0.276729941368103,
+      "learning_rate": 2.396108575546481e-05,
+      "loss": 0.4737,
       "step": 1995
     },
     {
-      "epoch": 0.07,
-      "learning_rate": 2.3452157598499065e-05,
-      "loss": 0.3545,
+      "epoch": 0.07207986448985476,
+      "grad_norm": 0.25289759039878845,
+      "learning_rate": 2.4021138601969734e-05,
+      "loss": 0.4869,
       "step": 2000
     },
     {
-      "epoch": 0.07,
-      "eval_loss": 0.3645670711994171,
-      "eval_runtime": 10.5377,
-      "eval_samples_per_second": 9.49,
-      "eval_steps_per_second": 9.49,
+      "epoch": 0.07207986448985476,
+      "eval_loss": 0.5271518230438232,
+      "eval_runtime": 3.5218,
+      "eval_samples_per_second": 28.395,
+      "eval_steps_per_second": 7.099,
       "step": 2000
     },
     {
-      "epoch": 0.07,
-      "learning_rate": 2.351078799249531e-05,
-      "loss": 0.3561,
+      "epoch": 0.0722600641510794,
+      "grad_norm": 0.2662065923213959,
+      "learning_rate": 2.4081191448474656e-05,
+      "loss": 0.5005,
       "step": 2005
     },
     {
-      "epoch": 0.07,
-      "learning_rate": 2.356941838649156e-05,
-      "loss": 0.3915,
+      "epoch": 0.07244026381230403,
+      "grad_norm": 0.2379211187362671,
+      "learning_rate": 2.4141244294979585e-05,
+      "loss": 0.4959,
       "step": 2010
     },
     {
-      "epoch": 0.07,
-      "learning_rate": 2.3628048780487806e-05,
-      "loss": 0.3427,
+      "epoch": 0.07262046347352867,
+      "grad_norm": 0.25165653228759766,
+      "learning_rate": 2.4201297141484507e-05,
+      "loss": 0.4869,
       "step": 2015
     },
     {
-      "epoch": 0.07,
-      "learning_rate": 2.368667917448405e-05,
-      "loss": 0.3641,
+      "epoch": 0.07280066313475331,
+      "grad_norm": 0.21869604289531708,
+      "learning_rate": 2.4261349987989433e-05,
+      "loss": 0.4938,
       "step": 2020
     },
     {
-      "epoch": 0.07,
-      "learning_rate": 2.37453095684803e-05,
-      "loss": 0.3611,
+      "epoch": 0.07298086279597794,
+      "grad_norm": 0.21399500966072083,
+      "learning_rate": 2.4321402834494355e-05,
+      "loss": 0.5029,
       "step": 2025
     },
     {
-      "epoch": 0.07,
-      "learning_rate": 2.380393996247655e-05,
-      "loss": 0.3538,
+      "epoch": 0.07316106245720258,
+      "grad_norm": 0.24735891819000244,
+      "learning_rate": 2.438145568099928e-05,
+      "loss": 0.503,
       "step": 2030
     },
     {
-      "epoch": 0.07,
-      "learning_rate": 2.3862570356472795e-05,
-      "loss": 0.3584,
+      "epoch": 0.07334126211842722,
+      "grad_norm": 0.27130669355392456,
+      "learning_rate": 2.4441508527504206e-05,
+      "loss": 0.4726,
       "step": 2035
     },
     {
-      "epoch": 0.07,
-      "learning_rate": 2.3921200750469044e-05,
-      "loss": 0.3742,
+      "epoch": 0.07352146177965185,
+      "grad_norm": 0.33456870913505554,
+      "learning_rate": 2.450156137400913e-05,
+      "loss": 0.4906,
       "step": 2040
     },
     {
-      "epoch": 0.07,
-      "learning_rate": 2.3979831144465293e-05,
-      "loss": 0.3686,
+      "epoch": 0.07370166144087649,
+      "grad_norm": 0.3058684766292572,
+      "learning_rate": 2.4561614220514054e-05,
+      "loss": 0.5345,
       "step": 2045
     },
     {
-      "epoch": 0.07,
-      "learning_rate": 2.4038461538461542e-05,
-      "loss": 0.4034,
+      "epoch": 0.07388186110210113,
+      "grad_norm": 0.24727894365787506,
+      "learning_rate": 2.4621667067018976e-05,
+      "loss": 0.5108,
       "step": 2050
     },
     {
-      "epoch": 0.07,
-      "learning_rate": 2.4097091932457787e-05,
-      "loss": 0.3634,
+      "epoch": 0.07406206076332576,
+      "grad_norm": 0.32694023847579956,
+      "learning_rate": 2.4681719913523902e-05,
+      "loss": 0.4917,
       "step": 2055
     },
     {
-      "epoch": 0.07,
-      "learning_rate": 2.4155722326454036e-05,
-      "loss": 0.3832,
+      "epoch": 0.0742422604245504,
+      "grad_norm": 0.3927519917488098,
+      "learning_rate": 2.4741772760028827e-05,
+      "loss": 0.4762,
       "step": 2060
     },
     {
-      "epoch": 0.07,
-      "learning_rate": 2.4214352720450282e-05,
-      "loss": 0.364,
+      "epoch": 0.07442246008577504,
+      "grad_norm": 0.22874149680137634,
+      "learning_rate": 2.4801825606533753e-05,
+      "loss": 0.4849,
       "step": 2065
     },
     {
-      "epoch": 0.07,
-      "learning_rate": 2.4272983114446528e-05,
-      "loss": 0.3563,
+      "epoch": 0.07460265974699967,
+      "grad_norm": 0.24440160393714905,
+      "learning_rate": 2.4861878453038675e-05,
+      "loss": 0.4888,
       "step": 2070
     },
     {
-      "epoch": 0.07,
-      "learning_rate": 2.4331613508442777e-05,
-      "loss": 0.3542,
+      "epoch": 0.07478285940822431,
+      "grad_norm": 0.2987866997718811,
+      "learning_rate": 2.49219312995436e-05,
+      "loss": 0.4926,
       "step": 2075
     },
     {
-      "epoch": 0.07,
-      "learning_rate": 2.4390243902439026e-05,
-      "loss": 0.382,
+      "epoch": 0.07496305906944895,
+      "grad_norm": 0.2661680281162262,
+      "learning_rate": 2.4981984146048523e-05,
+      "loss": 0.4854,
       "step": 2080
     },
     {
-      "epoch": 0.07,
-      "learning_rate": 2.444887429643527e-05,
-      "loss": 0.3793,
+      "epoch": 0.07514325873067358,
+      "grad_norm": 0.23432804644107819,
+      "learning_rate": 2.504203699255345e-05,
+      "loss": 0.4452,
       "step": 2085
     },
     {
-      "epoch": 0.07,
-      "learning_rate": 2.450750469043152e-05,
-      "loss": 0.3624,
+      "epoch": 0.07532345839189822,
+      "grad_norm": 0.28125226497650146,
+      "learning_rate": 2.5102089839058374e-05,
+      "loss": 0.4941,
       "step": 2090
     },
     {
-      "epoch": 0.07,
-      "learning_rate": 2.456613508442777e-05,
-      "loss": 0.3771,
+      "epoch": 0.07550365805312285,
+      "grad_norm": 0.2805047631263733,
+      "learning_rate": 2.5162142685563296e-05,
+      "loss": 0.4885,
       "step": 2095
     },
     {
-      "epoch": 0.07,
-      "learning_rate": 2.462476547842402e-05,
-      "loss": 0.3546,
+      "epoch": 0.07568385771434749,
+      "grad_norm": 0.24063357710838318,
+      "learning_rate": 2.5222195532068222e-05,
+      "loss": 0.4973,
       "step": 2100
     },
     {
-      "epoch": 0.07,
-      "learning_rate": 2.4683395872420264e-05,
-      "loss": 0.3705,
+      "epoch": 0.07586405737557213,
+      "grad_norm": 0.27712422609329224,
+      "learning_rate": 2.5282248378573147e-05,
+      "loss": 0.5314,
       "step": 2105
     },
     {
-      "epoch": 0.07,
-      "learning_rate": 2.4742026266416513e-05,
-      "loss": 0.3835,
+      "epoch": 0.07604425703679676,
+      "grad_norm": 0.2542542517185211,
+      "learning_rate": 2.534230122507807e-05,
+      "loss": 0.5142,
       "step": 2110
     },
     {
-      "epoch": 0.07,
-      "learning_rate": 2.4800656660412762e-05,
-      "loss": 0.3706,
+      "epoch": 0.07622445669802141,
+      "grad_norm": 0.19768337905406952,
+      "learning_rate": 2.5402354071582995e-05,
+      "loss": 0.4794,
       "step": 2115
     },
     {
-      "epoch": 0.07,
-      "learning_rate": 2.4859287054409004e-05,
-      "loss": 0.3756,
+      "epoch": 0.07640465635924605,
+      "grad_norm": 0.2631590664386749,
+      "learning_rate": 2.5462406918087917e-05,
+      "loss": 0.5131,
       "step": 2120
     },
     {
-      "epoch": 0.07,
-      "learning_rate": 2.4917917448405253e-05,
-      "loss": 0.3632,
+      "epoch": 0.07658485602047069,
+      "grad_norm": 0.29528722167015076,
+      "learning_rate": 2.5522459764592843e-05,
+      "loss": 0.5186,
       "step": 2125
     },
     {
-      "epoch": 0.07,
-      "learning_rate": 2.4976547842401502e-05,
-      "loss": 0.3488,
+      "epoch": 0.07676505568169532,
+      "grad_norm": 0.22340206801891327,
+      "learning_rate": 2.558251261109777e-05,
+      "loss": 0.4905,
       "step": 2130
     },
     {
-      "epoch": 0.08,
-      "learning_rate": 2.5035178236397748e-05,
-      "loss": 0.3553,
+      "epoch": 0.07694525534291996,
+      "grad_norm": 0.21741612255573273,
+      "learning_rate": 2.564256545760269e-05,
+      "loss": 0.4528,
       "step": 2135
     },
     {
-      "epoch": 0.08,
-      "learning_rate": 2.5093808630394e-05,
-      "loss": 0.3656,
+      "epoch": 0.0771254550041446,
+      "grad_norm": 0.2538367807865143,
+      "learning_rate": 2.5702618304107616e-05,
+      "loss": 0.5053,
       "step": 2140
     },
     {
-      "epoch": 0.08,
-      "learning_rate": 2.5152439024390246e-05,
-      "loss": 0.3598,
+      "epoch": 0.07730565466536923,
+      "grad_norm": 0.21461638808250427,
+      "learning_rate": 2.5762671150612538e-05,
+      "loss": 0.5069,
       "step": 2145
     },
     {
-      "epoch": 0.08,
-      "learning_rate": 2.521106941838649e-05,
-      "loss": 0.3478,
+      "epoch": 0.07748585432659387,
+      "grad_norm": 0.2098444551229477,
+      "learning_rate": 2.5822723997117464e-05,
+      "loss": 0.4934,
       "step": 2150
     },
     {
-      "epoch": 0.08,
-      "learning_rate": 2.526969981238274e-05,
-      "loss": 0.3963,
+      "epoch": 0.0776660539878185,
+      "grad_norm": 0.26539066433906555,
+      "learning_rate": 2.5882776843622393e-05,
+      "loss": 0.525,
       "step": 2155
     },
     {
-      "epoch": 0.08,
-      "learning_rate": 2.5328330206378986e-05,
-      "loss": 0.3649,
+      "epoch": 0.07784625364904314,
+      "grad_norm": 0.23766759037971497,
+      "learning_rate": 2.594282969012731e-05,
+      "loss": 0.473,
       "step": 2160
     },
     {
-      "epoch": 0.08,
-      "learning_rate": 2.538696060037524e-05,
-      "loss": 0.3606,
+      "epoch": 0.07802645331026778,
+      "grad_norm": 0.35496985912323,
+      "learning_rate": 2.600288253663224e-05,
+      "loss": 0.5096,
       "step": 2165
     },
     {
-      "epoch": 0.08,
-      "learning_rate": 2.544559099437148e-05,
-      "loss": 0.383,
+      "epoch": 0.07820665297149242,
+      "grad_norm": 0.23349033296108246,
+      "learning_rate": 2.606293538313716e-05,
+      "loss": 0.5294,
       "step": 2170
     },
     {
-      "epoch": 0.08,
-      "learning_rate": 2.5504221388367733e-05,
-      "loss": 0.3918,
+      "epoch": 0.07838685263271705,
+      "grad_norm": 0.26091858744621277,
+      "learning_rate": 2.6122988229642088e-05,
+      "loss": 0.4993,
       "step": 2175
     },
     {
-      "epoch": 0.08,
-      "learning_rate": 2.556285178236398e-05,
-      "loss": 0.3658,
+      "epoch": 0.07856705229394169,
+      "grad_norm": 0.22804076969623566,
+      "learning_rate": 2.6183041076147007e-05,
+      "loss": 0.4726,
       "step": 2180
     },
     {
-      "epoch": 0.08,
-      "learning_rate": 2.5621482176360228e-05,
-      "loss": 0.3517,
+      "epoch": 0.07874725195516633,
+      "grad_norm": 0.24472366273403168,
+      "learning_rate": 2.6243093922651936e-05,
+      "loss": 0.4801,
       "step": 2185
     },
     {
-      "epoch": 0.08,
-      "learning_rate": 2.5680112570356473e-05,
-      "loss": 0.3743,
+      "epoch": 0.07892745161639096,
+      "grad_norm": 0.30233868956565857,
+      "learning_rate": 2.630314676915686e-05,
+      "loss": 0.5144,
       "step": 2190
     },
     {
-      "epoch": 0.08,
-      "learning_rate": 2.573874296435272e-05,
-      "loss": 0.3722,
+      "epoch": 0.0791076512776156,
+      "grad_norm": 0.2996428608894348,
+      "learning_rate": 2.6363199615661784e-05,
+      "loss": 0.518,
       "step": 2195
     },
     {
-      "epoch": 0.08,
-      "learning_rate": 2.5797373358348968e-05,
-      "loss": 0.4099,
+      "epoch": 0.07928785093884023,
+      "grad_norm": 0.25679364800453186,
+      "learning_rate": 2.642325246216671e-05,
+      "loss": 0.4516,
       "step": 2200
     },
     {
-      "epoch": 0.08,
-      "learning_rate": 2.5856003752345214e-05,
-      "loss": 0.3321,
+      "epoch": 0.07946805060006487,
+      "grad_norm": 0.311570942401886,
+      "learning_rate": 2.648330530867163e-05,
+      "loss": 0.5341,
       "step": 2205
     },
     {
-      "epoch": 0.08,
-      "learning_rate": 2.5914634146341466e-05,
-      "loss": 0.3664,
+      "epoch": 0.07964825026128951,
+      "grad_norm": 0.2966238260269165,
+      "learning_rate": 2.6543358155176557e-05,
+      "loss": 0.4725,
       "step": 2210
     },
     {
-      "epoch": 0.08,
-      "learning_rate": 2.5973264540337712e-05,
-      "loss": 0.4203,
+      "epoch": 0.07982844992251414,
+      "grad_norm": 0.254808247089386,
+      "learning_rate": 2.6603411001681483e-05,
+      "loss": 0.4943,
       "step": 2215
     },
     {
-      "epoch": 0.08,
-      "learning_rate": 2.603189493433396e-05,
-      "loss": 0.3258,
+      "epoch": 0.08000864958373878,
+      "grad_norm": 0.23695941269397736,
+      "learning_rate": 2.6663463848186405e-05,
+      "loss": 0.5048,
       "step": 2220
     },
     {
-      "epoch": 0.08,
-      "learning_rate": 2.6090525328330206e-05,
-      "loss": 0.368,
+      "epoch": 0.08018884924496342,
+      "grad_norm": 0.29264283180236816,
+      "learning_rate": 2.672351669469133e-05,
+      "loss": 0.4968,
       "step": 2225
     },
     {
-      "epoch": 0.08,
-      "learning_rate": 2.614915572232646e-05,
-      "loss": 0.3918,
+      "epoch": 0.08036904890618805,
+      "grad_norm": 0.24488191306591034,
+      "learning_rate": 2.6783569541196252e-05,
+      "loss": 0.4911,
       "step": 2230
     },
     {
-      "epoch": 0.08,
-      "learning_rate": 2.62077861163227e-05,
-      "loss": 0.348,
+      "epoch": 0.08054924856741269,
+      "grad_norm": 0.281082421541214,
+      "learning_rate": 2.6843622387701178e-05,
+      "loss": 0.4946,
       "step": 2235
     },
     {
-      "epoch": 0.08,
-      "learning_rate": 2.6266416510318953e-05,
-      "loss": 0.3674,
+      "epoch": 0.08072944822863733,
+      "grad_norm": 0.2860982418060303,
+      "learning_rate": 2.6903675234206104e-05,
+      "loss": 0.5168,
       "step": 2240
     },
     {
-      "epoch": 0.08,
-      "learning_rate": 2.63250469043152e-05,
-      "loss": 0.3862,
+      "epoch": 0.08090964788986196,
+      "grad_norm": 0.2372538298368454,
+      "learning_rate": 2.6963728080711026e-05,
+      "loss": 0.4603,
       "step": 2245
     },
     {
-      "epoch": 0.08,
-      "learning_rate": 2.6383677298311445e-05,
-      "loss": 0.3807,
+      "epoch": 0.0810898475510866,
+      "grad_norm": 0.3054511845111847,
+      "learning_rate": 2.702378092721595e-05,
+      "loss": 0.4975,
       "step": 2250
     },
     {
-      "epoch": 0.08,
-      "learning_rate": 2.6442307692307694e-05,
-      "loss": 0.3581,
+      "epoch": 0.08127004721231124,
+      "grad_norm": 0.3154446482658386,
+      "learning_rate": 2.7083833773720874e-05,
+      "loss": 0.5038,
       "step": 2255
     },
     {
-      "epoch": 0.08,
-      "learning_rate": 2.650093808630394e-05,
-      "loss": 0.3731,
+      "epoch": 0.08145024687353587,
+      "grad_norm": 0.2663957476615906,
+      "learning_rate": 2.71438866202258e-05,
+      "loss": 0.5158,
       "step": 2260
     },
     {
-      "epoch": 0.08,
-      "learning_rate": 2.6559568480300188e-05,
-      "loss": 0.399,
+      "epoch": 0.08163044653476051,
+      "grad_norm": 0.2612464427947998,
+      "learning_rate": 2.7203939466730728e-05,
+      "loss": 0.5184,
       "step": 2265
     },
     {
-      "epoch": 0.08,
-      "learning_rate": 2.6618198874296434e-05,
-      "loss": 0.3581,
+      "epoch": 0.08181064619598515,
+      "grad_norm": 0.22396814823150635,
+      "learning_rate": 2.7263992313235647e-05,
+      "loss": 0.5009,
       "step": 2270
     },
     {
-      "epoch": 0.08,
-      "learning_rate": 2.6676829268292686e-05,
-      "loss": 0.3314,
+      "epoch": 0.08199084585720978,
+      "grad_norm": 0.2750130295753479,
+      "learning_rate": 2.7324045159740576e-05,
+      "loss": 0.5154,
       "step": 2275
     },
     {
-      "epoch": 0.08,
-      "learning_rate": 2.6735459662288932e-05,
-      "loss": 0.3697,
+      "epoch": 0.08217104551843442,
+      "grad_norm": 0.29694536328315735,
+      "learning_rate": 2.7384098006245495e-05,
+      "loss": 0.4796,
       "step": 2280
     },
     {
-      "epoch": 0.08,
-      "learning_rate": 2.679409005628518e-05,
-      "loss": 0.3752,
+      "epoch": 0.08235124517965907,
+      "grad_norm": 0.20941223204135895,
+      "learning_rate": 2.7444150852750424e-05,
+      "loss": 0.4525,
       "step": 2285
     },
     {
-      "epoch": 0.08,
-      "learning_rate": 2.6852720450281427e-05,
-      "loss": 0.3518,
+      "epoch": 0.0825314448408837,
+      "grad_norm": 0.3018248677253723,
+      "learning_rate": 2.750420369925535e-05,
+      "loss": 0.5211,
       "step": 2290
     },
     {
-      "epoch": 0.08,
-      "learning_rate": 2.691135084427768e-05,
-      "loss": 0.3563,
+      "epoch": 0.08271164450210834,
+      "grad_norm": 0.22867879271507263,
+      "learning_rate": 2.756425654576027e-05,
+      "loss": 0.4702,
       "step": 2295
     },
     {
-      "epoch": 0.08,
-      "learning_rate": 2.696998123827392e-05,
-      "loss": 0.3996,
+      "epoch": 0.08289184416333298,
+      "grad_norm": 0.26447781920433044,
+      "learning_rate": 2.7624309392265197e-05,
+      "loss": 0.4893,
       "step": 2300
     },
     {
-      "epoch": 0.08,
-      "learning_rate": 2.7028611632270167e-05,
-      "loss": 0.3122,
+      "epoch": 0.08307204382455761,
+      "grad_norm": 0.22963860630989075,
+      "learning_rate": 2.768436223877012e-05,
+      "loss": 0.509,
       "step": 2305
     },
     {
-      "epoch": 0.08,
-      "learning_rate": 2.708724202626642e-05,
-      "loss": 0.353,
+      "epoch": 0.08325224348578225,
+      "grad_norm": 0.34469008445739746,
+      "learning_rate": 2.7744415085275045e-05,
+      "loss": 0.4985,
       "step": 2310
     },
     {
-      "epoch": 0.08,
-      "learning_rate": 2.7145872420262665e-05,
-      "loss": 0.3473,
+      "epoch": 0.08343244314700689,
+      "grad_norm": 0.23541411757469177,
+      "learning_rate": 2.7804467931779963e-05,
+      "loss": 0.4892,
       "step": 2315
     },
     {
-      "epoch": 0.08,
-      "learning_rate": 2.7204502814258914e-05,
-      "loss": 0.3672,
+      "epoch": 0.08361264280823152,
+      "grad_norm": 0.2870751619338989,
+      "learning_rate": 2.7864520778284892e-05,
+      "loss": 0.4806,
       "step": 2320
     },
     {
-      "epoch": 0.08,
-      "learning_rate": 2.726313320825516e-05,
-      "loss": 0.3672,
+      "epoch": 0.08379284246945616,
+      "grad_norm": 0.24497903883457184,
+      "learning_rate": 2.7924573624789818e-05,
+      "loss": 0.4938,
       "step": 2325
     },
     {
-      "epoch": 0.08,
-      "learning_rate": 2.732176360225141e-05,
-      "loss": 0.3638,
+      "epoch": 0.0839730421306808,
+      "grad_norm": 0.29667189717292786,
+      "learning_rate": 2.798462647129474e-05,
+      "loss": 0.5338,
       "step": 2330
     },
     {
-      "epoch": 0.08,
-      "learning_rate": 2.7380393996247654e-05,
-      "loss": 0.3466,
+      "epoch": 0.08415324179190543,
+      "grad_norm": 0.2696093022823334,
+      "learning_rate": 2.8044679317799666e-05,
+      "loss": 0.4821,
       "step": 2335
     },
     {
-      "epoch": 0.08,
-      "learning_rate": 2.7439024390243906e-05,
-      "loss": 0.358,
+      "epoch": 0.08433344145313007,
+      "grad_norm": 0.2720743715763092,
+      "learning_rate": 2.8104732164304588e-05,
+      "loss": 0.4867,
       "step": 2340
     },
     {
-      "epoch": 0.08,
-      "learning_rate": 2.7497654784240152e-05,
-      "loss": 0.3614,
+      "epoch": 0.0845136411143547,
+      "grad_norm": 0.3151923418045044,
+      "learning_rate": 2.8164785010809513e-05,
+      "loss": 0.4833,
       "step": 2345
     },
     {
-      "epoch": 0.08,
-      "learning_rate": 2.75562851782364e-05,
-      "loss": 0.3242,
+      "epoch": 0.08469384077557934,
+      "grad_norm": 0.2476382553577423,
+      "learning_rate": 2.822483785731444e-05,
+      "loss": 0.4882,
       "step": 2350
     },
     {
-      "epoch": 0.08,
-      "learning_rate": 2.7614915572232647e-05,
-      "loss": 0.3651,
+      "epoch": 0.08487404043680398,
+      "grad_norm": 0.2863009572029114,
+      "learning_rate": 2.828489070381936e-05,
+      "loss": 0.4852,
       "step": 2355
     },
     {
-      "epoch": 0.08,
-      "learning_rate": 2.7673545966228892e-05,
-      "loss": 0.3657,
+      "epoch": 0.08505424009802862,
+      "grad_norm": 0.26534658670425415,
+      "learning_rate": 2.8344943550324287e-05,
+      "loss": 0.5093,
       "step": 2360
     },
     {
-      "epoch": 0.08,
-      "learning_rate": 2.773217636022514e-05,
-      "loss": 0.3347,
+      "epoch": 0.08523443975925325,
+      "grad_norm": 0.2766381800174713,
+      "learning_rate": 2.840499639682921e-05,
+      "loss": 0.4981,
       "step": 2365
     },
     {
-      "epoch": 0.08,
-      "learning_rate": 2.7790806754221387e-05,
-      "loss": 0.3429,
+      "epoch": 0.08541463942047789,
+      "grad_norm": 0.3110561966896057,
+      "learning_rate": 2.8465049243334134e-05,
+      "loss": 0.5176,
       "step": 2370
     },
     {
-      "epoch": 0.08,
-      "learning_rate": 2.784943714821764e-05,
-      "loss": 0.3619,
+      "epoch": 0.08559483908170253,
+      "grad_norm": 0.2818463444709778,
+      "learning_rate": 2.8525102089839063e-05,
+      "loss": 0.4864,
       "step": 2375
     },
     {
-      "epoch": 0.08,
-      "learning_rate": 2.7908067542213885e-05,
-      "loss": 0.3741,
+      "epoch": 0.08577503874292716,
+      "grad_norm": 0.2455795407295227,
+      "learning_rate": 2.8585154936343982e-05,
+      "loss": 0.5048,
       "step": 2380
     },
     {
-      "epoch": 0.08,
-      "learning_rate": 2.7966697936210134e-05,
-      "loss": 0.321,
+      "epoch": 0.0859552384041518,
+      "grad_norm": 0.3062375783920288,
+      "learning_rate": 2.864520778284891e-05,
+      "loss": 0.5028,
       "step": 2385
     },
     {
-      "epoch": 0.08,
-      "learning_rate": 2.802532833020638e-05,
-      "loss": 0.3701,
+      "epoch": 0.08613543806537644,
+      "grad_norm": 0.29522883892059326,
+      "learning_rate": 2.870526062935383e-05,
+      "loss": 0.5034,
       "step": 2390
     },
     {
-      "epoch": 0.08,
-      "learning_rate": 2.808395872420263e-05,
-      "loss": 0.3606,
+      "epoch": 0.08631563772660107,
+      "grad_norm": 0.3509838283061981,
+      "learning_rate": 2.876531347585876e-05,
+      "loss": 0.461,
       "step": 2395
     },
     {
-      "epoch": 0.08,
-      "learning_rate": 2.8142589118198874e-05,
-      "loss": 0.3868,
+      "epoch": 0.08649583738782571,
+      "grad_norm": 0.2348550707101822,
+      "learning_rate": 2.8825366322363684e-05,
+      "loss": 0.5092,
       "step": 2400
     },
     {
-      "epoch": 0.08,
-      "learning_rate": 2.820121951219512e-05,
-      "loss": 0.324,
+      "epoch": 0.08667603704905034,
+      "grad_norm": 0.28098031878471375,
+      "learning_rate": 2.8885419168868607e-05,
+      "loss": 0.5046,
       "step": 2405
     },
     {
-      "epoch": 0.08,
-      "learning_rate": 2.8259849906191372e-05,
-      "loss": 0.3593,
+      "epoch": 0.08685623671027498,
+      "grad_norm": 0.29102039337158203,
+      "learning_rate": 2.8945472015373532e-05,
+      "loss": 0.4669,
       "step": 2410
     },
     {
-      "epoch": 0.08,
-      "learning_rate": 2.8318480300187618e-05,
-      "loss": 0.3776,
+      "epoch": 0.08703643637149962,
+      "grad_norm": 0.273709774017334,
+      "learning_rate": 2.900552486187845e-05,
+      "loss": 0.5121,
       "step": 2415
     },
     {
-      "epoch": 0.09,
-      "learning_rate": 2.8377110694183867e-05,
-      "loss": 0.3843,
+      "epoch": 0.08721663603272425,
+      "grad_norm": 0.3082577586174011,
+      "learning_rate": 2.906557770838338e-05,
+      "loss": 0.5158,
       "step": 2420
     },
     {
-      "epoch": 0.09,
-      "learning_rate": 2.8435741088180113e-05,
-      "loss": 0.3398,
+      "epoch": 0.08739683569394889,
+      "grad_norm": 0.26482829451560974,
+      "learning_rate": 2.9125630554888305e-05,
+      "loss": 0.4774,
       "step": 2425
     },
     {
-      "epoch": 0.09,
-      "learning_rate": 2.849437148217636e-05,
-      "loss": 0.3671,
+      "epoch": 0.08757703535517353,
+      "grad_norm": 0.26536524295806885,
+      "learning_rate": 2.9185683401393228e-05,
+      "loss": 0.4929,
       "step": 2430
     },
     {
-      "epoch": 0.09,
-      "learning_rate": 2.8553001876172607e-05,
-      "loss": 0.353,
+      "epoch": 0.08775723501639816,
+      "grad_norm": 0.26633599400520325,
+      "learning_rate": 2.9245736247898153e-05,
+      "loss": 0.4826,
       "step": 2435
     },
     {
-      "epoch": 0.09,
-      "learning_rate": 2.861163227016886e-05,
-      "loss": 0.3404,
+      "epoch": 0.0879374346776228,
+      "grad_norm": 0.2031547874212265,
+      "learning_rate": 2.9305789094403075e-05,
+      "loss": 0.4906,
       "step": 2440
     },
     {
-      "epoch": 0.09,
-      "learning_rate": 2.8670262664165105e-05,
-      "loss": 0.3265,
+      "epoch": 0.08811763433884744,
+      "grad_norm": 0.2550112307071686,
+      "learning_rate": 2.9365841940908e-05,
+      "loss": 0.4968,
       "step": 2445
     },
     {
-      "epoch": 0.09,
-      "learning_rate": 2.8728893058161354e-05,
-      "loss": 0.3741,
+      "epoch": 0.08829783400007207,
+      "grad_norm": 0.28357723355293274,
+      "learning_rate": 2.9425894787412923e-05,
+      "loss": 0.4956,
       "step": 2450
     },
     {
-      "epoch": 0.09,
-      "learning_rate": 2.87875234521576e-05,
-      "loss": 0.3315,
+      "epoch": 0.08847803366129672,
+      "grad_norm": 0.32640430331230164,
+      "learning_rate": 2.948594763391785e-05,
+      "loss": 0.5074,
       "step": 2455
     },
     {
-      "epoch": 0.09,
-      "learning_rate": 2.8846153846153845e-05,
-      "loss": 0.3615,
+      "epoch": 0.08865823332252136,
+      "grad_norm": 0.240611732006073,
+      "learning_rate": 2.9546000480422774e-05,
+      "loss": 0.4661,
       "step": 2460
     },
     {
-      "epoch": 0.09,
-      "learning_rate": 2.8904784240150094e-05,
-      "loss": 0.3375,
+      "epoch": 0.088838432983746,
+      "grad_norm": 0.27230045199394226,
+      "learning_rate": 2.9606053326927696e-05,
+      "loss": 0.4959,
       "step": 2465
     },
     {
-      "epoch": 0.09,
-      "learning_rate": 2.896341463414634e-05,
-      "loss": 0.3464,
+      "epoch": 0.08901863264497063,
+      "grad_norm": 0.20566551387310028,
+      "learning_rate": 2.9666106173432622e-05,
+      "loss": 0.518,
       "step": 2470
     },
     {
-      "epoch": 0.09,
-      "learning_rate": 2.9022045028142592e-05,
-      "loss": 0.405,
+      "epoch": 0.08919883230619527,
+      "grad_norm": 0.2713625729084015,
+      "learning_rate": 2.9726159019937544e-05,
+      "loss": 0.5178,
       "step": 2475
     },
     {
-      "epoch": 0.09,
-      "learning_rate": 2.9080675422138838e-05,
-      "loss": 0.3392,
+      "epoch": 0.0893790319674199,
+      "grad_norm": 0.2567339837551117,
+      "learning_rate": 2.978621186644247e-05,
+      "loss": 0.498,
       "step": 2480
     },
     {
-      "epoch": 0.09,
-      "learning_rate": 2.9139305816135087e-05,
-      "loss": 0.3461,
+      "epoch": 0.08955923162864454,
+      "grad_norm": 0.27121180295944214,
+      "learning_rate": 2.98462647129474e-05,
+      "loss": 0.5194,
       "step": 2485
     },
     {
-      "epoch": 0.09,
-      "learning_rate": 2.9197936210131333e-05,
-      "loss": 0.348,
+      "epoch": 0.08973943128986918,
+      "grad_norm": 0.35053110122680664,
+      "learning_rate": 2.9906317559452317e-05,
+      "loss": 0.533,
       "step": 2490
     },
     {
-      "epoch": 0.09,
-      "learning_rate": 2.9256566604127582e-05,
-      "loss": 0.368,
+      "epoch": 0.08991963095109382,
+      "grad_norm": 0.2476474940776825,
+      "learning_rate": 2.9966370405957246e-05,
+      "loss": 0.4895,
       "step": 2495
     },
     {
-      "epoch": 0.09,
-      "learning_rate": 2.9315196998123827e-05,
-      "loss": 0.367,
+      "epoch": 0.09009983061231845,
+      "grad_norm": 0.22306469082832336,
+      "learning_rate": 3.0026423252462165e-05,
+      "loss": 0.4847,
       "step": 2500
     },
     {
-      "epoch": 0.09,
-      "eval_loss": 0.3577040433883667,
-      "eval_runtime": 10.5307,
-      "eval_samples_per_second": 9.496,
-      "eval_steps_per_second": 9.496,
+      "epoch": 0.09009983061231845,
+      "eval_loss": 0.5190762877464294,
+      "eval_runtime": 3.513,
+      "eval_samples_per_second": 28.466,
+      "eval_steps_per_second": 7.116,
       "step": 2500
     },
     {
-      "epoch": 0.09,
-      "learning_rate": 2.937382739212008e-05,
-      "loss": 0.3809,
+      "epoch": 0.09028003027354309,
+      "grad_norm": 0.2834235429763794,
+      "learning_rate": 3.008647609896709e-05,
+      "loss": 0.5589,
       "step": 2505
     },
     {
-      "epoch": 0.09,
-      "learning_rate": 2.9432457786116325e-05,
-      "loss": 0.3426,
+      "epoch": 0.09046022993476772,
+      "grad_norm": 0.24811461567878723,
+      "learning_rate": 3.014652894547202e-05,
+      "loss": 0.4679,
       "step": 2510
     },
     {
-      "epoch": 0.09,
-      "learning_rate": 2.949108818011257e-05,
-      "loss": 0.3887,
+      "epoch": 0.09064042959599236,
+      "grad_norm": 0.25787869095802307,
+      "learning_rate": 3.020658179197694e-05,
+      "loss": 0.4961,
       "step": 2515
     },
     {
-      "epoch": 0.09,
-      "learning_rate": 2.954971857410882e-05,
-      "loss": 0.3503,
+      "epoch": 0.090820629257217,
+      "grad_norm": 0.1960582286119461,
+      "learning_rate": 3.0266634638481867e-05,
+      "loss": 0.5253,
       "step": 2520
     },
     {
-      "epoch": 0.09,
-      "learning_rate": 2.9608348968105066e-05,
-      "loss": 0.3561,
+      "epoch": 0.09100082891844163,
+      "grad_norm": 0.28364789485931396,
+      "learning_rate": 3.0326687484986786e-05,
+      "loss": 0.4797,
       "step": 2525
     },
     {
-      "epoch": 0.09,
-      "learning_rate": 2.9666979362101315e-05,
-      "loss": 0.3581,
+      "epoch": 0.09118102857966627,
+      "grad_norm": 0.30647027492523193,
+      "learning_rate": 3.0386740331491715e-05,
+      "loss": 0.5131,
       "step": 2530
     },
     {
-      "epoch": 0.09,
-      "learning_rate": 2.972560975609756e-05,
-      "loss": 0.3724,
+      "epoch": 0.09136122824089091,
+      "grad_norm": 0.2918342351913452,
+      "learning_rate": 3.044679317799664e-05,
+      "loss": 0.4649,
       "step": 2535
     },
     {
-      "epoch": 0.09,
-      "learning_rate": 2.9784240150093813e-05,
-      "loss": 0.378,
+      "epoch": 0.09154142790211554,
+      "grad_norm": 0.34838005900382996,
+      "learning_rate": 3.0506846024501563e-05,
+      "loss": 0.5218,
       "step": 2540
     },
     {
-      "epoch": 0.09,
-      "learning_rate": 2.9842870544090058e-05,
-      "loss": 0.3481,
+      "epoch": 0.09172162756334018,
+      "grad_norm": 0.22675694525241852,
+      "learning_rate": 3.056689887100649e-05,
+      "loss": 0.5072,
       "step": 2545
     },
     {
-      "epoch": 0.09,
-      "learning_rate": 2.9901500938086307e-05,
-      "loss": 0.3424,
+      "epoch": 0.09190182722456482,
+      "grad_norm": 0.3220086991786957,
+      "learning_rate": 3.062695171751141e-05,
+      "loss": 0.4664,
       "step": 2550
     },
     {
-      "epoch": 0.09,
-      "learning_rate": 2.9960131332082553e-05,
-      "loss": 0.3445,
+      "epoch": 0.09208202688578945,
+      "grad_norm": 0.2494743913412094,
+      "learning_rate": 3.068700456401633e-05,
+      "loss": 0.5067,
       "step": 2555
     },
     {
-      "epoch": 0.09,
-      "learning_rate": 3.0018761726078802e-05,
-      "loss": 0.3466,
+      "epoch": 0.09226222654701409,
+      "grad_norm": 0.35692298412323,
+      "learning_rate": 3.0747057410521265e-05,
+      "loss": 0.5213,
       "step": 2560
     },
     {
-      "epoch": 0.09,
-      "learning_rate": 3.0077392120075048e-05,
-      "loss": 0.3691,
+      "epoch": 0.09244242620823873,
+      "grad_norm": 0.352923721075058,
+      "learning_rate": 3.0807110257026184e-05,
+      "loss": 0.5052,
       "step": 2565
     },
     {
-      "epoch": 0.09,
-      "learning_rate": 3.0136022514071293e-05,
-      "loss": 0.3692,
+      "epoch": 0.09262262586946336,
+      "grad_norm": 0.2893941402435303,
+      "learning_rate": 3.086716310353111e-05,
+      "loss": 0.4696,
       "step": 2570
     },
     {
-      "epoch": 0.09,
-      "learning_rate": 3.0194652908067546e-05,
-      "loss": 0.3332,
+      "epoch": 0.092802825530688,
+      "grad_norm": 0.26177549362182617,
+      "learning_rate": 3.092721595003603e-05,
+      "loss": 0.5036,
       "step": 2575
     },
     {
-      "epoch": 0.09,
-      "learning_rate": 3.025328330206379e-05,
-      "loss": 0.3502,
+      "epoch": 0.09298302519191264,
+      "grad_norm": 0.22046709060668945,
+      "learning_rate": 3.098726879654096e-05,
+      "loss": 0.4604,
       "step": 2580
     },
     {
-      "epoch": 0.09,
-      "learning_rate": 3.031191369606004e-05,
-      "loss": 0.3345,
+      "epoch": 0.09316322485313727,
+      "grad_norm": 0.28608959913253784,
+      "learning_rate": 3.104732164304588e-05,
+      "loss": 0.5139,
       "step": 2585
     },
     {
-      "epoch": 0.09,
-      "learning_rate": 3.0370544090056286e-05,
-      "loss": 0.3595,
+      "epoch": 0.09334342451436191,
+      "grad_norm": 0.31415843963623047,
+      "learning_rate": 3.1107374489550805e-05,
+      "loss": 0.4915,
       "step": 2590
     },
     {
-      "epoch": 0.09,
-      "learning_rate": 3.0429174484052535e-05,
-      "loss": 0.3554,
+      "epoch": 0.09352362417558654,
+      "grad_norm": 0.24263805150985718,
+      "learning_rate": 3.116742733605573e-05,
+      "loss": 0.5423,
       "step": 2595
     },
     {
-      "epoch": 0.09,
-      "learning_rate": 3.048780487804878e-05,
-      "loss": 0.3632,
+      "epoch": 0.09370382383681118,
+      "grad_norm": 0.30432233214378357,
+      "learning_rate": 3.1227480182560656e-05,
+      "loss": 0.4955,
       "step": 2600
     },
     {
-      "epoch": 0.09,
-      "learning_rate": 3.054643527204503e-05,
-      "loss": 0.3838,
+      "epoch": 0.09388402349803582,
+      "grad_norm": 0.2743891775608063,
+      "learning_rate": 3.128753302906558e-05,
+      "loss": 0.5167,
       "step": 2605
     },
     {
-      "epoch": 0.09,
-      "learning_rate": 3.060506566604128e-05,
-      "loss": 0.3539,
+      "epoch": 0.09406422315926045,
+      "grad_norm": 0.2813490629196167,
+      "learning_rate": 3.13475858755705e-05,
+      "loss": 0.5374,
       "step": 2610
     },
     {
-      "epoch": 0.09,
-      "learning_rate": 3.0663696060037524e-05,
-      "loss": 0.3716,
+      "epoch": 0.09424442282048509,
+      "grad_norm": 0.3482913076877594,
+      "learning_rate": 3.1407638722075426e-05,
+      "loss": 0.4621,
       "step": 2615
     },
     {
-      "epoch": 0.09,
-      "learning_rate": 3.072232645403377e-05,
-      "loss": 0.3561,
+      "epoch": 0.09442462248170974,
+      "grad_norm": 0.3260897696018219,
+      "learning_rate": 3.146769156858035e-05,
+      "loss": 0.5003,
       "step": 2620
     },
     {
-      "epoch": 0.09,
-      "learning_rate": 3.0780956848030015e-05,
-      "loss": 0.3504,
+      "epoch": 0.09460482214293438,
+      "grad_norm": 0.27140331268310547,
+      "learning_rate": 3.152774441508528e-05,
+      "loss": 0.5161,
       "step": 2625
     },
     {
-      "epoch": 0.09,
-      "learning_rate": 3.083958724202627e-05,
-      "loss": 0.3515,
+      "epoch": 0.09478502180415901,
+      "grad_norm": 0.24953608214855194,
+      "learning_rate": 3.15877972615902e-05,
+      "loss": 0.5059,
       "step": 2630
     },
     {
-      "epoch": 0.09,
-      "learning_rate": 3.089821763602251e-05,
-      "loss": 0.3483,
+      "epoch": 0.09496522146538365,
+      "grad_norm": 0.2428867667913437,
+      "learning_rate": 3.164785010809512e-05,
+      "loss": 0.5285,
       "step": 2635
     },
     {
-      "epoch": 0.09,
-      "learning_rate": 3.0956848030018766e-05,
-      "loss": 0.3907,
+      "epoch": 0.09514542112660829,
+      "grad_norm": 0.2688390910625458,
+      "learning_rate": 3.170790295460005e-05,
+      "loss": 0.4971,
       "step": 2640
     },
     {
-      "epoch": 0.09,
-      "learning_rate": 3.101547842401501e-05,
-      "loss": 0.3511,
+      "epoch": 0.09532562078783292,
+      "grad_norm": 0.292272686958313,
+      "learning_rate": 3.176795580110498e-05,
+      "loss": 0.4781,
       "step": 2645
     },
     {
-      "epoch": 0.09,
-      "learning_rate": 3.1074108818011264e-05,
-      "loss": 0.3438,
+      "epoch": 0.09550582044905756,
+      "grad_norm": 0.22563514113426208,
+      "learning_rate": 3.18280086476099e-05,
+      "loss": 0.4319,
       "step": 2650
     },
     {
-      "epoch": 0.09,
-      "learning_rate": 3.11327392120075e-05,
-      "loss": 0.3856,
+      "epoch": 0.0956860201102822,
+      "grad_norm": 0.31794285774230957,
+      "learning_rate": 3.1888061494114824e-05,
+      "loss": 0.5238,
       "step": 2655
     },
     {
-      "epoch": 0.09,
-      "learning_rate": 3.1191369606003755e-05,
-      "loss": 0.33,
+      "epoch": 0.09586621977150683,
+      "grad_norm": 0.36745190620422363,
+      "learning_rate": 3.194811434061974e-05,
+      "loss": 0.4985,
       "step": 2660
     },
     {
-      "epoch": 0.09,
-      "learning_rate": 3.125e-05,
-      "loss": 0.373,
+      "epoch": 0.09604641943273147,
+      "grad_norm": 0.2950117588043213,
+      "learning_rate": 3.200816718712467e-05,
+      "loss": 0.5128,
       "step": 2665
     },
     {
-      "epoch": 0.09,
-      "learning_rate": 3.1308630393996246e-05,
-      "loss": 0.3941,
+      "epoch": 0.0962266190939561,
+      "grad_norm": 0.30142202973365784,
+      "learning_rate": 3.20682200336296e-05,
+      "loss": 0.4884,
       "step": 2670
     },
     {
-      "epoch": 0.09,
-      "learning_rate": 3.13672607879925e-05,
-      "loss": 0.3685,
+      "epoch": 0.09640681875518074,
+      "grad_norm": 0.34995657205581665,
+      "learning_rate": 3.212827288013452e-05,
+      "loss": 0.5074,
       "step": 2675
     },
     {
-      "epoch": 0.09,
-      "learning_rate": 3.1425891181988744e-05,
-      "loss": 0.3595,
+      "epoch": 0.09658701841640538,
+      "grad_norm": 0.2538526952266693,
+      "learning_rate": 3.2188325726639445e-05,
+      "loss": 0.4964,
       "step": 2680
     },
     {
-      "epoch": 0.09,
-      "learning_rate": 3.148452157598499e-05,
-      "loss": 0.3686,
+      "epoch": 0.09676721807763002,
+      "grad_norm": 0.3425311744213104,
+      "learning_rate": 3.2248378573144364e-05,
+      "loss": 0.4721,
       "step": 2685
     },
     {
-      "epoch": 0.09,
-      "learning_rate": 3.1543151969981236e-05,
-      "loss": 0.3551,
+      "epoch": 0.09694741773885465,
+      "grad_norm": 0.3075348138809204,
+      "learning_rate": 3.2308431419649296e-05,
+      "loss": 0.5145,
       "step": 2690
     },
     {
-      "epoch": 0.09,
-      "learning_rate": 3.160178236397749e-05,
-      "loss": 0.357,
+      "epoch": 0.09712761740007929,
+      "grad_norm": 0.24702075123786926,
+      "learning_rate": 3.236848426615422e-05,
+      "loss": 0.4793,
       "step": 2695
     },
     {
-      "epoch": 0.09,
-      "learning_rate": 3.1660412757973734e-05,
-      "loss": 0.3788,
+      "epoch": 0.09730781706130393,
+      "grad_norm": 0.2921466827392578,
+      "learning_rate": 3.242853711265914e-05,
+      "loss": 0.5097,
       "step": 2700
     },
     {
-      "epoch": 0.1,
-      "learning_rate": 3.1719043151969986e-05,
-      "loss": 0.346,
+      "epoch": 0.09748801672252856,
+      "grad_norm": 0.22841855883598328,
+      "learning_rate": 3.2488589959164066e-05,
+      "loss": 0.4905,
       "step": 2705
     },
     {
-      "epoch": 0.1,
-      "learning_rate": 3.177767354596623e-05,
-      "loss": 0.3612,
+      "epoch": 0.0976682163837532,
+      "grad_norm": 0.24366632103919983,
+      "learning_rate": 3.254864280566899e-05,
+      "loss": 0.4773,
       "step": 2710
     },
     {
-      "epoch": 0.1,
-      "learning_rate": 3.1836303939962484e-05,
-      "loss": 0.3425,
+      "epoch": 0.09784841604497783,
+      "grad_norm": 0.2745325565338135,
+      "learning_rate": 3.260869565217392e-05,
+      "loss": 0.492,
       "step": 2715
     },
     {
-      "epoch": 0.1,
-      "learning_rate": 3.189493433395872e-05,
-      "loss": 0.3502,
+      "epoch": 0.09802861570620247,
+      "grad_norm": 0.22801122069358826,
+      "learning_rate": 3.266874849867884e-05,
+      "loss": 0.4883,
       "step": 2720
     },
     {
-      "epoch": 0.1,
-      "learning_rate": 3.195356472795497e-05,
-      "loss": 0.3617,
+      "epoch": 0.09820881536742711,
+      "grad_norm": 0.316967636346817,
+      "learning_rate": 3.272880134518376e-05,
+      "loss": 0.4811,
       "step": 2725
     },
     {
-      "epoch": 0.1,
-      "learning_rate": 3.201219512195122e-05,
-      "loss": 0.3544,
+      "epoch": 0.09838901502865174,
+      "grad_norm": 0.28945192694664,
+      "learning_rate": 3.278885419168869e-05,
+      "loss": 0.4782,
       "step": 2730
     },
     {
-      "epoch": 0.1,
-      "learning_rate": 3.2070825515947466e-05,
-      "loss": 0.3392,
+      "epoch": 0.09856921468987638,
+      "grad_norm": 0.25695520639419556,
+      "learning_rate": 3.284890703819361e-05,
+      "loss": 0.5038,
       "step": 2735
     },
     {
-      "epoch": 0.1,
-      "learning_rate": 3.212945590994372e-05,
-      "loss": 0.3688,
+      "epoch": 0.09874941435110102,
+      "grad_norm": 0.22989587485790253,
+      "learning_rate": 3.290895988469854e-05,
+      "loss": 0.4745,
       "step": 2740
     },
     {
-      "epoch": 0.1,
-      "learning_rate": 3.2188086303939964e-05,
-      "loss": 0.3643,
+      "epoch": 0.09892961401232565,
+      "grad_norm": 0.28844088315963745,
+      "learning_rate": 3.296901273120346e-05,
+      "loss": 0.4886,
       "step": 2745
     },
     {
-      "epoch": 0.1,
-      "learning_rate": 3.224671669793621e-05,
-      "loss": 0.3766,
+      "epoch": 0.09910981367355029,
+      "grad_norm": 0.21202966570854187,
+      "learning_rate": 3.302906557770838e-05,
+      "loss": 0.4926,
       "step": 2750
     },
     {
-      "epoch": 0.1,
-      "learning_rate": 3.2305347091932456e-05,
-      "loss": 0.3322,
+      "epoch": 0.09929001333477493,
+      "grad_norm": 0.4138276278972626,
+      "learning_rate": 3.308911842421331e-05,
+      "loss": 0.5199,
       "step": 2755
     },
     {
-      "epoch": 0.1,
-      "learning_rate": 3.236397748592871e-05,
-      "loss": 0.3508,
+      "epoch": 0.09947021299599956,
+      "grad_norm": 0.22605939209461212,
+      "learning_rate": 3.3149171270718233e-05,
+      "loss": 0.4753,
       "step": 2760
     },
     {
-      "epoch": 0.1,
-      "learning_rate": 3.2422607879924954e-05,
-      "loss": 0.3747,
+      "epoch": 0.0996504126572242,
+      "grad_norm": 0.25468626618385315,
+      "learning_rate": 3.320922411722316e-05,
+      "loss": 0.483,
       "step": 2765
     },
     {
-      "epoch": 0.1,
-      "learning_rate": 3.24812382739212e-05,
-      "loss": 0.3758,
+      "epoch": 0.09983061231844884,
+      "grad_norm": 0.25252848863601685,
+      "learning_rate": 3.326927696372808e-05,
+      "loss": 0.5189,
       "step": 2770
     },
     {
-      "epoch": 0.1,
-      "learning_rate": 3.253986866791745e-05,
-      "loss": 0.3588,
+      "epoch": 0.10001081197967347,
+      "grad_norm": 0.29405850172042847,
+      "learning_rate": 3.3329329810233003e-05,
+      "loss": 0.4849,
       "step": 2775
     },
     {
-      "epoch": 0.1,
-      "learning_rate": 3.25984990619137e-05,
-      "loss": 0.3316,
+      "epoch": 0.10019101164089811,
+      "grad_norm": 0.37302690744400024,
+      "learning_rate": 3.3389382656737936e-05,
+      "loss": 0.4963,
       "step": 2780
     },
     {
-      "epoch": 0.1,
-      "learning_rate": 3.265712945590994e-05,
-      "loss": 0.3572,
+      "epoch": 0.10037121130212275,
+      "grad_norm": 0.23915189504623413,
+      "learning_rate": 3.3449435503242855e-05,
+      "loss": 0.5278,
       "step": 2785
     },
     {
-      "epoch": 0.1,
-      "learning_rate": 3.271575984990619e-05,
-      "loss": 0.3326,
+      "epoch": 0.1005514109633474,
+      "grad_norm": 0.3481108546257019,
+      "learning_rate": 3.350948834974778e-05,
+      "loss": 0.4938,
       "step": 2790
     },
     {
-      "epoch": 0.1,
-      "learning_rate": 3.277439024390244e-05,
-      "loss": 0.3435,
+      "epoch": 0.10073161062457203,
+      "grad_norm": 0.23707103729248047,
+      "learning_rate": 3.35695411962527e-05,
+      "loss": 0.4766,
       "step": 2795
     },
     {
-      "epoch": 0.1,
-      "learning_rate": 3.283302063789869e-05,
-      "loss": 0.3519,
+      "epoch": 0.10091181028579667,
+      "grad_norm": 0.25163257122039795,
+      "learning_rate": 3.362959404275763e-05,
+      "loss": 0.4613,
       "step": 2800
     },
     {
-      "epoch": 0.1,
-      "learning_rate": 3.289165103189494e-05,
-      "loss": 0.3572,
+      "epoch": 0.1010920099470213,
+      "grad_norm": 0.2651137113571167,
+      "learning_rate": 3.368964688926256e-05,
+      "loss": 0.4841,
       "step": 2805
     },
     {
-      "epoch": 0.1,
-      "learning_rate": 3.2950281425891185e-05,
-      "loss": 0.3398,
+      "epoch": 0.10127220960824594,
+      "grad_norm": 0.32314032316207886,
+      "learning_rate": 3.3749699735767476e-05,
+      "loss": 0.4957,
       "step": 2810
     },
     {
-      "epoch": 0.1,
-      "learning_rate": 3.300891181988743e-05,
-      "loss": 0.3444,
+      "epoch": 0.10145240926947058,
+      "grad_norm": 0.3319461941719055,
+      "learning_rate": 3.38097525822724e-05,
+      "loss": 0.492,
       "step": 2815
     },
     {
-      "epoch": 0.1,
-      "learning_rate": 3.3067542213883676e-05,
-      "loss": 0.3467,
+      "epoch": 0.10163260893069521,
+      "grad_norm": 0.24138695001602173,
+      "learning_rate": 3.386980542877733e-05,
+      "loss": 0.511,
       "step": 2820
     },
     {
-      "epoch": 0.1,
-      "learning_rate": 3.312617260787992e-05,
-      "loss": 0.3668,
+      "epoch": 0.10181280859191985,
+      "grad_norm": 0.3053799271583557,
+      "learning_rate": 3.392985827528225e-05,
+      "loss": 0.4911,
       "step": 2825
     },
     {
-      "epoch": 0.1,
-      "learning_rate": 3.3184803001876174e-05,
-      "loss": 0.3473,
+      "epoch": 0.10199300825314449,
+      "grad_norm": 0.26419997215270996,
+      "learning_rate": 3.398991112178718e-05,
+      "loss": 0.4836,
       "step": 2830
     },
     {
-      "epoch": 0.1,
-      "learning_rate": 3.324343339587242e-05,
-      "loss": 0.3297,
+      "epoch": 0.10217320791436912,
+      "grad_norm": 0.2777937352657318,
+      "learning_rate": 3.4049963968292097e-05,
+      "loss": 0.5124,
       "step": 2835
     },
     {
-      "epoch": 0.1,
-      "learning_rate": 3.330206378986867e-05,
-      "loss": 0.3245,
+      "epoch": 0.10235340757559376,
+      "grad_norm": 0.2962092459201813,
+      "learning_rate": 3.411001681479702e-05,
+      "loss": 0.4948,
       "step": 2840
     },
     {
-      "epoch": 0.1,
-      "learning_rate": 3.336069418386492e-05,
-      "loss": 0.3357,
+      "epoch": 0.1025336072368184,
+      "grad_norm": 0.3595232665538788,
+      "learning_rate": 3.417006966130195e-05,
+      "loss": 0.5122,
       "step": 2845
     },
     {
-      "epoch": 0.1,
-      "learning_rate": 3.341932457786116e-05,
-      "loss": 0.3226,
+      "epoch": 0.10271380689804303,
+      "grad_norm": 0.20213566720485687,
+      "learning_rate": 3.423012250780687e-05,
+      "loss": 0.4886,
       "step": 2850
     },
     {
-      "epoch": 0.1,
-      "learning_rate": 3.347795497185741e-05,
-      "loss": 0.3611,
+      "epoch": 0.10289400655926767,
+      "grad_norm": 0.3061508536338806,
+      "learning_rate": 3.42901753543118e-05,
+      "loss": 0.5068,
       "step": 2855
     },
     {
-      "epoch": 0.1,
-      "learning_rate": 3.353658536585366e-05,
-      "loss": 0.3491,
+      "epoch": 0.1030742062204923,
+      "grad_norm": 0.27155801653862,
+      "learning_rate": 3.435022820081672e-05,
+      "loss": 0.4583,
       "step": 2860
     },
     {
-      "epoch": 0.1,
-      "learning_rate": 3.359521575984991e-05,
-      "loss": 0.3388,
+      "epoch": 0.10325440588171694,
+      "grad_norm": 0.2824406623840332,
+      "learning_rate": 3.441028104732164e-05,
+      "loss": 0.4857,
       "step": 2865
     },
     {
-      "epoch": 0.1,
-      "learning_rate": 3.365384615384616e-05,
-      "loss": 0.3568,
+      "epoch": 0.10343460554294158,
+      "grad_norm": 0.2800888121128082,
+      "learning_rate": 3.447033389382657e-05,
+      "loss": 0.4928,
       "step": 2870
     },
     {
-      "epoch": 0.1,
-      "learning_rate": 3.3712476547842405e-05,
-      "loss": 0.3603,
+      "epoch": 0.10361480520416622,
+      "grad_norm": 0.2530428469181061,
+      "learning_rate": 3.4530386740331494e-05,
+      "loss": 0.4841,
       "step": 2875
     },
     {
-      "epoch": 0.1,
-      "learning_rate": 3.377110694183865e-05,
-      "loss": 0.3639,
+      "epoch": 0.10379500486539085,
+      "grad_norm": 0.2559506595134735,
+      "learning_rate": 3.459043958683641e-05,
+      "loss": 0.4919,
       "step": 2880
     },
     {
-      "epoch": 0.1,
-      "learning_rate": 3.3829737335834896e-05,
-      "loss": 0.3595,
+      "epoch": 0.10397520452661549,
+      "grad_norm": 0.35989275574684143,
+      "learning_rate": 3.465049243334134e-05,
+      "loss": 0.5059,
       "step": 2885
     },
     {
-      "epoch": 0.1,
-      "learning_rate": 3.388836772983114e-05,
-      "loss": 0.3682,
+      "epoch": 0.10415540418784013,
+      "grad_norm": 0.2567185163497925,
+      "learning_rate": 3.471054527984627e-05,
+      "loss": 0.4752,
       "step": 2890
     },
     {
-      "epoch": 0.1,
-      "learning_rate": 3.3946998123827394e-05,
-      "loss": 0.3693,
+      "epoch": 0.10433560384906476,
+      "grad_norm": 0.3426368832588196,
+      "learning_rate": 3.477059812635119e-05,
+      "loss": 0.5043,
       "step": 2895
     },
     {
-      "epoch": 0.1,
-      "learning_rate": 3.400562851782364e-05,
-      "loss": 0.3433,
+      "epoch": 0.1045158035102894,
+      "grad_norm": 0.26572245359420776,
+      "learning_rate": 3.4830650972856115e-05,
+      "loss": 0.5032,
       "step": 2900
     },
     {
-      "epoch": 0.1,
-      "learning_rate": 3.406425891181989e-05,
-      "loss": 0.3549,
+      "epoch": 0.10469600317151403,
+      "grad_norm": 0.2190892994403839,
+      "learning_rate": 3.4890703819361034e-05,
+      "loss": 0.4637,
       "step": 2905
     },
     {
-      "epoch": 0.1,
-      "learning_rate": 3.412288930581614e-05,
-      "loss": 0.3357,
+      "epoch": 0.10487620283273867,
+      "grad_norm": 0.3017270267009735,
+      "learning_rate": 3.4950756665865967e-05,
+      "loss": 0.4957,
       "step": 2910
     },
     {
-      "epoch": 0.1,
-      "learning_rate": 3.418151969981238e-05,
-      "loss": 0.3879,
+      "epoch": 0.10505640249396331,
+      "grad_norm": 0.2628999650478363,
+      "learning_rate": 3.501080951237089e-05,
+      "loss": 0.469,
       "step": 2915
     },
     {
-      "epoch": 0.1,
-      "learning_rate": 3.424015009380863e-05,
-      "loss": 0.355,
+      "epoch": 0.10523660215518794,
+      "grad_norm": 0.26654675602912903,
+      "learning_rate": 3.507086235887581e-05,
+      "loss": 0.4872,
       "step": 2920
     },
     {
-      "epoch": 0.1,
-      "learning_rate": 3.429878048780488e-05,
-      "loss": 0.3463,
+      "epoch": 0.10541680181641258,
+      "grad_norm": 0.33271774649620056,
+      "learning_rate": 3.5130915205380736e-05,
+      "loss": 0.465,
       "step": 2925
     },
     {
-      "epoch": 0.1,
-      "learning_rate": 3.435741088180113e-05,
-      "loss": 0.3445,
+      "epoch": 0.10559700147763722,
+      "grad_norm": 0.2927808165550232,
+      "learning_rate": 3.519096805188566e-05,
+      "loss": 0.5051,
       "step": 2930
     },
     {
-      "epoch": 0.1,
-      "learning_rate": 3.441604127579737e-05,
-      "loss": 0.3564,
+      "epoch": 0.10577720113886185,
+      "grad_norm": 0.2696301341056824,
+      "learning_rate": 3.525102089839059e-05,
+      "loss": 0.459,
       "step": 2935
     },
     {
-      "epoch": 0.1,
-      "learning_rate": 3.4474671669793625e-05,
-      "loss": 0.3453,
+      "epoch": 0.10595740080008649,
+      "grad_norm": 0.21873407065868378,
+      "learning_rate": 3.531107374489551e-05,
+      "loss": 0.4857,
       "step": 2940
     },
     {
-      "epoch": 0.1,
-      "learning_rate": 3.453330206378987e-05,
-      "loss": 0.3659,
+      "epoch": 0.10613760046131113,
+      "grad_norm": 0.22248321771621704,
+      "learning_rate": 3.537112659140043e-05,
+      "loss": 0.4911,
       "step": 2945
     },
     {
-      "epoch": 0.1,
-      "learning_rate": 3.4591932457786116e-05,
-      "loss": 0.3705,
+      "epoch": 0.10631780012253576,
+      "grad_norm": 0.2554193139076233,
+      "learning_rate": 3.543117943790536e-05,
+      "loss": 0.5309,
       "step": 2950
     },
     {
-      "epoch": 0.1,
-      "learning_rate": 3.465056285178236e-05,
-      "loss": 0.3214,
+      "epoch": 0.1064979997837604,
+      "grad_norm": 0.27886924147605896,
+      "learning_rate": 3.549123228441028e-05,
+      "loss": 0.4637,
       "step": 2955
     },
     {
-      "epoch": 0.1,
-      "learning_rate": 3.4709193245778614e-05,
-      "loss": 0.374,
+      "epoch": 0.10667819944498505,
+      "grad_norm": 0.23546633124351501,
+      "learning_rate": 3.555128513091521e-05,
+      "loss": 0.4657,
       "step": 2960
     },
     {
-      "epoch": 0.1,
-      "learning_rate": 3.476782363977486e-05,
-      "loss": 0.3363,
+      "epoch": 0.10685839910620969,
+      "grad_norm": 0.2205863744020462,
+      "learning_rate": 3.5611337977420134e-05,
+      "loss": 0.493,
       "step": 2965
     },
     {
-      "epoch": 0.1,
-      "learning_rate": 3.482645403377111e-05,
-      "loss": 0.3535,
+      "epoch": 0.10703859876743432,
+      "grad_norm": 0.27504095435142517,
+      "learning_rate": 3.567139082392505e-05,
+      "loss": 0.4543,
       "step": 2970
     },
     {
-      "epoch": 0.1,
-      "learning_rate": 3.488508442776736e-05,
-      "loss": 0.3567,
+      "epoch": 0.10721879842865896,
+      "grad_norm": 0.27117300033569336,
+      "learning_rate": 3.573144367042998e-05,
+      "loss": 0.4604,
       "step": 2975
     },
     {
-      "epoch": 0.1,
-      "learning_rate": 3.4943714821763604e-05,
-      "loss": 0.3518,
+      "epoch": 0.1073989980898836,
+      "grad_norm": 0.29087886214256287,
+      "learning_rate": 3.5791496516934904e-05,
+      "loss": 0.515,
       "step": 2980
     },
     {
-      "epoch": 0.11,
-      "learning_rate": 3.500234521575985e-05,
-      "loss": 0.3385,
+      "epoch": 0.10757919775110823,
+      "grad_norm": 0.2880057096481323,
+      "learning_rate": 3.585154936343983e-05,
+      "loss": 0.5116,
       "step": 2985
     },
     {
-      "epoch": 0.11,
-      "learning_rate": 3.5060975609756095e-05,
-      "loss": 0.3581,
+      "epoch": 0.10775939741233287,
+      "grad_norm": 0.26126590371131897,
+      "learning_rate": 3.5911602209944755e-05,
+      "loss": 0.5085,
       "step": 2990
     },
     {
-      "epoch": 0.11,
-      "learning_rate": 3.511960600375235e-05,
-      "loss": 0.3654,
+      "epoch": 0.1079395970735575,
+      "grad_norm": 0.34922003746032715,
+      "learning_rate": 3.5971655056449674e-05,
+      "loss": 0.5063,
       "step": 2995
     },
     {
-      "epoch": 0.11,
-      "learning_rate": 3.517823639774859e-05,
-      "loss": 0.353,
+      "epoch": 0.10811979673478214,
+      "grad_norm": 0.34584909677505493,
+      "learning_rate": 3.6031707902954606e-05,
+      "loss": 0.5323,
       "step": 3000
     },
     {
-      "epoch": 0.11,
-      "eval_loss": 0.3522051274776459,
-      "eval_runtime": 10.5471,
-      "eval_samples_per_second": 9.481,
-      "eval_steps_per_second": 9.481,
+      "epoch": 0.10811979673478214,
+      "eval_loss": 0.515049397945404,
+      "eval_runtime": 3.5163,
+      "eval_samples_per_second": 28.439,
+      "eval_steps_per_second": 7.11,
       "step": 3000
     },
     {
-      "epoch": 0.11,
-      "learning_rate": 3.5236866791744845e-05,
-      "loss": 0.3403,
+      "epoch": 0.10829999639600678,
+      "grad_norm": 0.22614799439907074,
+      "learning_rate": 3.6091760749459525e-05,
+      "loss": 0.4944,
       "step": 3005
     },
     {
-      "epoch": 0.11,
-      "learning_rate": 3.529549718574109e-05,
-      "loss": 0.3401,
+      "epoch": 0.10848019605723142,
+      "grad_norm": 0.2877223789691925,
+      "learning_rate": 3.615181359596445e-05,
+      "loss": 0.5352,
       "step": 3010
     },
     {
-      "epoch": 0.11,
-      "learning_rate": 3.5354127579737336e-05,
-      "loss": 0.3447,
+      "epoch": 0.10866039571845605,
+      "grad_norm": 0.2686571478843689,
+      "learning_rate": 3.621186644246937e-05,
+      "loss": 0.4605,
       "step": 3015
     },
     {
-      "epoch": 0.11,
-      "learning_rate": 3.541275797373358e-05,
-      "loss": 0.3644,
+      "epoch": 0.10884059537968069,
+      "grad_norm": 0.3369406759738922,
+      "learning_rate": 3.62719192889743e-05,
+      "loss": 0.477,
       "step": 3020
     },
     {
-      "epoch": 0.11,
-      "learning_rate": 3.5471388367729835e-05,
-      "loss": 0.3646,
+      "epoch": 0.10902079504090532,
+      "grad_norm": 0.27363601326942444,
+      "learning_rate": 3.633197213547923e-05,
+      "loss": 0.503,
       "step": 3025
     },
     {
-      "epoch": 0.11,
-      "learning_rate": 3.553001876172608e-05,
-      "loss": 0.3469,
+      "epoch": 0.10920099470212996,
+      "grad_norm": 0.21706698834896088,
+      "learning_rate": 3.6392024981984146e-05,
+      "loss": 0.4783,
       "step": 3030
     },
     {
-      "epoch": 0.11,
-      "learning_rate": 3.5588649155722326e-05,
-      "loss": 0.3766,
+      "epoch": 0.1093811943633546,
+      "grad_norm": 0.2553368806838989,
+      "learning_rate": 3.645207782848907e-05,
+      "loss": 0.514,
       "step": 3035
     },
     {
-      "epoch": 0.11,
-      "learning_rate": 3.564727954971858e-05,
-      "loss": 0.3477,
+      "epoch": 0.10956139402457923,
+      "grad_norm": 0.22971071302890778,
+      "learning_rate": 3.6512130674994e-05,
+      "loss": 0.4765,
       "step": 3040
     },
     {
-      "epoch": 0.11,
-      "learning_rate": 3.5705909943714824e-05,
-      "loss": 0.349,
+      "epoch": 0.10974159368580387,
+      "grad_norm": 0.2552284598350525,
+      "learning_rate": 3.657218352149892e-05,
+      "loss": 0.4497,
       "step": 3045
     },
     {
-      "epoch": 0.11,
-      "learning_rate": 3.576454033771107e-05,
-      "loss": 0.3357,
+      "epoch": 0.1099217933470285,
+      "grad_norm": 0.336775541305542,
+      "learning_rate": 3.663223636800385e-05,
+      "loss": 0.5162,
       "step": 3050
     },
     {
-      "epoch": 0.11,
-      "learning_rate": 3.5823170731707315e-05,
-      "loss": 0.3592,
+      "epoch": 0.11010199300825314,
+      "grad_norm": 0.27510499954223633,
+      "learning_rate": 3.669228921450877e-05,
+      "loss": 0.4996,
       "step": 3055
     },
     {
-      "epoch": 0.11,
-      "learning_rate": 3.588180112570357e-05,
-      "loss": 0.3172,
+      "epoch": 0.11028219266947778,
+      "grad_norm": 0.2974676787853241,
+      "learning_rate": 3.675234206101369e-05,
+      "loss": 0.4634,
       "step": 3060
     },
     {
-      "epoch": 0.11,
-      "learning_rate": 3.594043151969981e-05,
-      "loss": 0.3552,
+      "epoch": 0.11046239233070242,
+      "grad_norm": 0.2689826190471649,
+      "learning_rate": 3.681239490751862e-05,
+      "loss": 0.5105,
       "step": 3065
     },
     {
-      "epoch": 0.11,
-      "learning_rate": 3.5999061913696065e-05,
-      "loss": 0.356,
+      "epoch": 0.11064259199192705,
+      "grad_norm": 0.27652692794799805,
+      "learning_rate": 3.6872447754023544e-05,
+      "loss": 0.5259,
       "step": 3070
     },
     {
-      "epoch": 0.11,
-      "learning_rate": 3.605769230769231e-05,
-      "loss": 0.3345,
+      "epoch": 0.11082279165315169,
+      "grad_norm": 0.29832860827445984,
+      "learning_rate": 3.693250060052847e-05,
+      "loss": 0.5042,
       "step": 3075
     },
     {
-      "epoch": 0.11,
-      "learning_rate": 3.611632270168856e-05,
-      "loss": 0.3301,
+      "epoch": 0.11100299131437633,
+      "grad_norm": 0.28680893778800964,
+      "learning_rate": 3.699255344703339e-05,
+      "loss": 0.5209,
       "step": 3080
     },
     {
-      "epoch": 0.11,
-      "learning_rate": 3.61749530956848e-05,
-      "loss": 0.3499,
+      "epoch": 0.11118319097560096,
+      "grad_norm": 0.3082677125930786,
+      "learning_rate": 3.7052606293538314e-05,
+      "loss": 0.503,
       "step": 3085
     },
     {
-      "epoch": 0.11,
-      "learning_rate": 3.623358348968105e-05,
-      "loss": 0.3707,
+      "epoch": 0.1113633906368256,
+      "grad_norm": 0.2548987865447998,
+      "learning_rate": 3.711265914004324e-05,
+      "loss": 0.4714,
       "step": 3090
     },
     {
-      "epoch": 0.11,
-      "learning_rate": 3.62922138836773e-05,
-      "loss": 0.3458,
+      "epoch": 0.11154359029805024,
+      "grad_norm": 0.39584243297576904,
+      "learning_rate": 3.7172711986548165e-05,
+      "loss": 0.5219,
       "step": 3095
     },
     {
-      "epoch": 0.11,
-      "learning_rate": 3.6350844277673546e-05,
-      "loss": 0.3501,
+      "epoch": 0.11172378995927487,
+      "grad_norm": 0.36074209213256836,
+      "learning_rate": 3.723276483305309e-05,
+      "loss": 0.5255,
       "step": 3100
     },
     {
-      "epoch": 0.11,
-      "learning_rate": 3.64094746716698e-05,
-      "loss": 0.3591,
+      "epoch": 0.11190398962049951,
+      "grad_norm": 0.22818799316883087,
+      "learning_rate": 3.729281767955801e-05,
+      "loss": 0.4651,
       "step": 3105
     },
     {
-      "epoch": 0.11,
-      "learning_rate": 3.6468105065666044e-05,
-      "loss": 0.3478,
+      "epoch": 0.11208418928172414,
+      "grad_norm": 0.22764305770397186,
+      "learning_rate": 3.735287052606294e-05,
+      "loss": 0.4482,
       "step": 3110
     },
     {
-      "epoch": 0.11,
-      "learning_rate": 3.652673545966229e-05,
-      "loss": 0.3257,
+      "epoch": 0.11226438894294878,
+      "grad_norm": 0.2548796236515045,
+      "learning_rate": 3.741292337256786e-05,
+      "loss": 0.4823,
       "step": 3115
     },
     {
-      "epoch": 0.11,
-      "learning_rate": 3.6585365853658535e-05,
-      "loss": 0.3472,
+      "epoch": 0.11244458860417342,
+      "grad_norm": 0.2939002811908722,
+      "learning_rate": 3.7472976219072786e-05,
+      "loss": 0.4909,
       "step": 3120
     },
     {
-      "epoch": 0.11,
-      "learning_rate": 3.664399624765479e-05,
-      "loss": 0.3647,
+      "epoch": 0.11262478826539805,
+      "grad_norm": 0.3162885308265686,
+      "learning_rate": 3.753302906557771e-05,
+      "loss": 0.4936,
       "step": 3125
     },
     {
-      "epoch": 0.11,
-      "learning_rate": 3.670262664165103e-05,
-      "loss": 0.3712,
+      "epoch": 0.1128049879266227,
+      "grad_norm": 0.2588067352771759,
+      "learning_rate": 3.759308191208264e-05,
+      "loss": 0.4921,
       "step": 3130
     },
     {
-      "epoch": 0.11,
-      "learning_rate": 3.6761257035647286e-05,
-      "loss": 0.3607,
+      "epoch": 0.11298518758784734,
+      "grad_norm": 0.3024102747440338,
+      "learning_rate": 3.765313475858756e-05,
+      "loss": 0.4931,
       "step": 3135
     },
     {
-      "epoch": 0.11,
-      "learning_rate": 3.681988742964353e-05,
-      "loss": 0.3935,
+      "epoch": 0.11316538724907198,
+      "grad_norm": 0.40092089772224426,
+      "learning_rate": 3.771318760509248e-05,
+      "loss": 0.5415,
       "step": 3140
     },
     {
-      "epoch": 0.11,
-      "learning_rate": 3.687851782363978e-05,
-      "loss": 0.3593,
+      "epoch": 0.11334558691029661,
+      "grad_norm": 0.29862499237060547,
+      "learning_rate": 3.777324045159741e-05,
+      "loss": 0.4613,
       "step": 3145
     },
     {
-      "epoch": 0.11,
-      "learning_rate": 3.693714821763602e-05,
-      "loss": 0.3689,
+      "epoch": 0.11352578657152125,
+      "grad_norm": 0.3131415843963623,
+      "learning_rate": 3.7833293298102326e-05,
+      "loss": 0.5001,
       "step": 3150
     },
     {
-      "epoch": 0.11,
-      "learning_rate": 3.699577861163227e-05,
-      "loss": 0.3612,
+      "epoch": 0.11370598623274589,
+      "grad_norm": 0.2650774419307709,
+      "learning_rate": 3.789334614460726e-05,
+      "loss": 0.4916,
       "step": 3155
     },
     {
-      "epoch": 0.11,
-      "learning_rate": 3.705440900562852e-05,
-      "loss": 0.3765,
+      "epoch": 0.11388618589397052,
+      "grad_norm": 0.20322157442569733,
+      "learning_rate": 3.7953398991112184e-05,
+      "loss": 0.4615,
       "step": 3160
     },
     {
-      "epoch": 0.11,
-      "learning_rate": 3.7113039399624766e-05,
-      "loss": 0.3804,
+      "epoch": 0.11406638555519516,
+      "grad_norm": 0.2679648995399475,
+      "learning_rate": 3.80134518376171e-05,
+      "loss": 0.5039,
       "step": 3165
     },
     {
-      "epoch": 0.11,
-      "learning_rate": 3.717166979362102e-05,
-      "loss": 0.3414,
+      "epoch": 0.1142465852164198,
+      "grad_norm": 0.3185202181339264,
+      "learning_rate": 3.807350468412203e-05,
+      "loss": 0.5249,
       "step": 3170
     },
     {
-      "epoch": 0.11,
-      "learning_rate": 3.7230300187617264e-05,
-      "loss": 0.3391,
+      "epoch": 0.11442678487764443,
+      "grad_norm": 0.31188127398490906,
+      "learning_rate": 3.8133557530626954e-05,
+      "loss": 0.4923,
       "step": 3175
     },
     {
-      "epoch": 0.11,
-      "learning_rate": 3.728893058161351e-05,
-      "loss": 0.39,
+      "epoch": 0.11460698453886907,
+      "grad_norm": 0.2713908553123474,
+      "learning_rate": 3.819361037713188e-05,
+      "loss": 0.4823,
       "step": 3180
     },
     {
-      "epoch": 0.11,
-      "learning_rate": 3.7347560975609755e-05,
-      "loss": 0.3589,
+      "epoch": 0.1147871842000937,
+      "grad_norm": 0.2642289102077484,
+      "learning_rate": 3.8253663223636805e-05,
+      "loss": 0.4711,
       "step": 3185
     },
     {
-      "epoch": 0.11,
-      "learning_rate": 3.7406191369606e-05,
-      "loss": 0.3554,
+      "epoch": 0.11496738386131834,
+      "grad_norm": 0.30651018023490906,
+      "learning_rate": 3.8313716070141724e-05,
+      "loss": 0.5242,
       "step": 3190
     },
     {
-      "epoch": 0.11,
-      "learning_rate": 3.7464821763602253e-05,
-      "loss": 0.3238,
+      "epoch": 0.11514758352254298,
+      "grad_norm": 0.31043529510498047,
+      "learning_rate": 3.837376891664665e-05,
+      "loss": 0.4922,
       "step": 3195
     },
     {
-      "epoch": 0.11,
-      "learning_rate": 3.75234521575985e-05,
-      "loss": 0.3561,
+      "epoch": 0.11532778318376762,
+      "grad_norm": 0.29356715083122253,
+      "learning_rate": 3.8433821763151575e-05,
+      "loss": 0.46,
       "step": 3200
     },
     {
-      "epoch": 0.11,
-      "learning_rate": 3.758208255159475e-05,
-      "loss": 0.3526,
+      "epoch": 0.11550798284499225,
+      "grad_norm": 0.3174915611743927,
+      "learning_rate": 3.84938746096565e-05,
+      "loss": 0.4916,
       "step": 3205
     },
     {
-      "epoch": 0.11,
-      "learning_rate": 3.7640712945591e-05,
-      "loss": 0.3791,
+      "epoch": 0.11568818250621689,
+      "grad_norm": 0.2561566233634949,
+      "learning_rate": 3.8553927456161426e-05,
+      "loss": 0.4666,
       "step": 3210
     },
     {
-      "epoch": 0.11,
-      "learning_rate": 3.769934333958724e-05,
-      "loss": 0.3545,
+      "epoch": 0.11586838216744152,
+      "grad_norm": 0.24369752407073975,
+      "learning_rate": 3.8613980302666345e-05,
+      "loss": 0.483,
       "step": 3215
     },
     {
-      "epoch": 0.11,
-      "learning_rate": 3.775797373358349e-05,
-      "loss": 0.3523,
+      "epoch": 0.11604858182866616,
+      "grad_norm": 0.24509058892726898,
+      "learning_rate": 3.867403314917128e-05,
+      "loss": 0.4709,
       "step": 3220
     },
     {
-      "epoch": 0.11,
-      "learning_rate": 3.781660412757974e-05,
-      "loss": 0.3347,
+      "epoch": 0.1162287814898908,
+      "grad_norm": 0.23607365787029266,
+      "learning_rate": 3.8734085995676196e-05,
+      "loss": 0.5009,
       "step": 3225
     },
     {
-      "epoch": 0.11,
-      "learning_rate": 3.7875234521575986e-05,
-      "loss": 0.3463,
+      "epoch": 0.11640898115111543,
+      "grad_norm": 0.30435293912887573,
+      "learning_rate": 3.879413884218112e-05,
+      "loss": 0.4995,
       "step": 3230
     },
     {
-      "epoch": 0.11,
-      "learning_rate": 3.793386491557224e-05,
-      "loss": 0.3371,
+      "epoch": 0.11658918081234007,
+      "grad_norm": 0.20363330841064453,
+      "learning_rate": 3.885419168868605e-05,
+      "loss": 0.4809,
       "step": 3235
     },
     {
-      "epoch": 0.11,
-      "learning_rate": 3.7992495309568484e-05,
-      "loss": 0.3451,
+      "epoch": 0.11676938047356471,
+      "grad_norm": 0.25914672017097473,
+      "learning_rate": 3.891424453519097e-05,
+      "loss": 0.4811,
       "step": 3240
     },
     {
-      "epoch": 0.11,
-      "learning_rate": 3.805112570356473e-05,
-      "loss": 0.3648,
+      "epoch": 0.11694958013478934,
+      "grad_norm": 0.2945806682109833,
+      "learning_rate": 3.89742973816959e-05,
+      "loss": 0.4874,
       "step": 3245
     },
     {
-      "epoch": 0.11,
-      "learning_rate": 3.8109756097560976e-05,
-      "loss": 0.372,
+      "epoch": 0.11712977979601398,
+      "grad_norm": 0.25838515162467957,
+      "learning_rate": 3.903435022820082e-05,
+      "loss": 0.493,
       "step": 3250
     },
     {
-      "epoch": 0.11,
-      "learning_rate": 3.816838649155722e-05,
-      "loss": 0.3444,
+      "epoch": 0.11730997945723862,
+      "grad_norm": 0.27623122930526733,
+      "learning_rate": 3.909440307470574e-05,
+      "loss": 0.4873,
       "step": 3255
     },
     {
-      "epoch": 0.11,
-      "learning_rate": 3.8227016885553474e-05,
-      "loss": 0.355,
+      "epoch": 0.11749017911846325,
+      "grad_norm": 0.24453672766685486,
+      "learning_rate": 3.915445592121067e-05,
+      "loss": 0.4419,
       "step": 3260
     },
     {
-      "epoch": 0.11,
-      "learning_rate": 3.828564727954972e-05,
-      "loss": 0.3656,
+      "epoch": 0.11767037877968789,
+      "grad_norm": 0.26414886116981506,
+      "learning_rate": 3.9214508767715593e-05,
+      "loss": 0.4597,
       "step": 3265
     },
     {
-      "epoch": 0.12,
-      "learning_rate": 3.834427767354597e-05,
-      "loss": 0.3422,
+      "epoch": 0.11785057844091253,
+      "grad_norm": 0.246024951338768,
+      "learning_rate": 3.927456161422052e-05,
+      "loss": 0.4923,
       "step": 3270
     },
     {
-      "epoch": 0.12,
-      "learning_rate": 3.840290806754222e-05,
-      "loss": 0.361,
+      "epoch": 0.11803077810213716,
+      "grad_norm": 0.25583988428115845,
+      "learning_rate": 3.933461446072544e-05,
+      "loss": 0.4346,
       "step": 3275
     },
     {
-      "epoch": 0.12,
-      "learning_rate": 3.846153846153846e-05,
-      "loss": 0.3587,
+      "epoch": 0.1182109777633618,
+      "grad_norm": 0.2518952190876007,
+      "learning_rate": 3.939466730723036e-05,
+      "loss": 0.5041,
       "step": 3280
     },
     {
-      "epoch": 0.12,
-      "learning_rate": 3.852016885553471e-05,
-      "loss": 0.3556,
+      "epoch": 0.11839117742458644,
+      "grad_norm": 0.29746881127357483,
+      "learning_rate": 3.945472015373529e-05,
+      "loss": 0.5066,
       "step": 3285
     },
     {
-      "epoch": 0.12,
-      "learning_rate": 3.857879924953096e-05,
-      "loss": 0.3273,
+      "epoch": 0.11857137708581107,
+      "grad_norm": 0.2888292074203491,
+      "learning_rate": 3.9514773000240214e-05,
+      "loss": 0.4976,
       "step": 3290
     },
     {
-      "epoch": 0.12,
-      "learning_rate": 3.8637429643527207e-05,
-      "loss": 0.339,
+      "epoch": 0.11875157674703571,
+      "grad_norm": 0.2132648229598999,
+      "learning_rate": 3.957482584674514e-05,
+      "loss": 0.4901,
       "step": 3295
     },
     {
-      "epoch": 0.12,
-      "learning_rate": 3.869606003752345e-05,
-      "loss": 0.3469,
+      "epoch": 0.11893177640826036,
+      "grad_norm": 0.2614684998989105,
+      "learning_rate": 3.963487869325006e-05,
+      "loss": 0.5054,
       "step": 3300
     },
     {
-      "epoch": 0.12,
-      "learning_rate": 3.8754690431519705e-05,
-      "loss": 0.3426,
+      "epoch": 0.119111976069485,
+      "grad_norm": 0.2751060128211975,
+      "learning_rate": 3.9694931539754984e-05,
+      "loss": 0.5719,
       "step": 3305
     },
     {
-      "epoch": 0.12,
-      "learning_rate": 3.881332082551595e-05,
-      "loss": 0.3597,
+      "epoch": 0.11929217573070963,
+      "grad_norm": 0.2760029435157776,
+      "learning_rate": 3.975498438625991e-05,
+      "loss": 0.4718,
       "step": 3310
     },
     {
-      "epoch": 0.12,
-      "learning_rate": 3.8871951219512196e-05,
-      "loss": 0.3435,
+      "epoch": 0.11947237539193427,
+      "grad_norm": 0.3538672626018524,
+      "learning_rate": 3.9815037232764836e-05,
+      "loss": 0.5002,
       "step": 3315
     },
     {
-      "epoch": 0.12,
-      "learning_rate": 3.893058161350844e-05,
-      "loss": 0.3399,
+      "epoch": 0.1196525750531589,
+      "grad_norm": 0.34897580742836,
+      "learning_rate": 3.987509007926976e-05,
+      "loss": 0.4935,
       "step": 3320
     },
     {
-      "epoch": 0.12,
-      "learning_rate": 3.8989212007504694e-05,
-      "loss": 0.3429,
+      "epoch": 0.11983277471438354,
+      "grad_norm": 0.3150857388973236,
+      "learning_rate": 3.993514292577468e-05,
+      "loss": 0.5232,
       "step": 3325
     },
     {
-      "epoch": 0.12,
-      "learning_rate": 3.904784240150094e-05,
-      "loss": 0.3387,
+      "epoch": 0.12001297437560818,
+      "grad_norm": 0.22704921662807465,
+      "learning_rate": 3.999519577227961e-05,
+      "loss": 0.4848,
       "step": 3330
     },
     {
-      "epoch": 0.12,
-      "learning_rate": 3.910647279549719e-05,
-      "loss": 0.3409,
+      "epoch": 0.12019317403683281,
+      "grad_norm": 0.25814276933670044,
+      "learning_rate": 4.005524861878453e-05,
+      "loss": 0.4773,
       "step": 3335
     },
     {
-      "epoch": 0.12,
-      "learning_rate": 3.916510318949344e-05,
-      "loss": 0.3602,
+      "epoch": 0.12037337369805745,
+      "grad_norm": 0.2505935728549957,
+      "learning_rate": 4.0115301465289457e-05,
+      "loss": 0.5039,
       "step": 3340
     },
     {
-      "epoch": 0.12,
-      "learning_rate": 3.922373358348968e-05,
-      "loss": 0.3468,
+      "epoch": 0.12055357335928209,
+      "grad_norm": 0.26671484112739563,
+      "learning_rate": 4.017535431179438e-05,
+      "loss": 0.486,
       "step": 3345
     },
     {
-      "epoch": 0.12,
-      "learning_rate": 3.928236397748593e-05,
-      "loss": 0.3489,
+      "epoch": 0.12073377302050672,
+      "grad_norm": 0.289005309343338,
+      "learning_rate": 4.02354071582993e-05,
+      "loss": 0.5201,
       "step": 3350
     },
     {
-      "epoch": 0.12,
-      "learning_rate": 3.9340994371482174e-05,
-      "loss": 0.3501,
+      "epoch": 0.12091397268173136,
+      "grad_norm": 0.284462571144104,
+      "learning_rate": 4.029546000480423e-05,
+      "loss": 0.5013,
       "step": 3355
     },
     {
-      "epoch": 0.12,
-      "learning_rate": 3.939962476547843e-05,
-      "loss": 0.3368,
+      "epoch": 0.121094172342956,
+      "grad_norm": 0.29563307762145996,
+      "learning_rate": 4.035551285130915e-05,
+      "loss": 0.5042,
       "step": 3360
     },
     {
-      "epoch": 0.12,
-      "learning_rate": 3.945825515947467e-05,
-      "loss": 0.372,
+      "epoch": 0.12127437200418063,
+      "grad_norm": 0.2727985680103302,
+      "learning_rate": 4.041556569781408e-05,
+      "loss": 0.46,
       "step": 3365
     },
     {
-      "epoch": 0.12,
-      "learning_rate": 3.9516885553470925e-05,
-      "loss": 0.3321,
+      "epoch": 0.12145457166540527,
+      "grad_norm": 0.24392500519752502,
+      "learning_rate": 4.0475618544319e-05,
+      "loss": 0.4903,
       "step": 3370
     },
     {
-      "epoch": 0.12,
-      "learning_rate": 3.957551594746717e-05,
-      "loss": 0.3535,
+      "epoch": 0.1216347713266299,
+      "grad_norm": 0.261043518781662,
+      "learning_rate": 4.053567139082393e-05,
+      "loss": 0.4818,
       "step": 3375
     },
     {
-      "epoch": 0.12,
-      "learning_rate": 3.9634146341463416e-05,
-      "loss": 0.3559,
+      "epoch": 0.12181497098785454,
+      "grad_norm": 0.294289767742157,
+      "learning_rate": 4.0595724237328854e-05,
+      "loss": 0.48,
       "step": 3380
     },
     {
-      "epoch": 0.12,
-      "learning_rate": 3.969277673545966e-05,
-      "loss": 0.3842,
+      "epoch": 0.12199517064907918,
+      "grad_norm": 0.300879567861557,
+      "learning_rate": 4.065577708383377e-05,
+      "loss": 0.4833,
       "step": 3385
     },
     {
-      "epoch": 0.12,
-      "learning_rate": 3.9751407129455914e-05,
-      "loss": 0.3575,
+      "epoch": 0.12217537031030382,
+      "grad_norm": 0.23128901422023773,
+      "learning_rate": 4.07158299303387e-05,
+      "loss": 0.5047,
       "step": 3390
     },
     {
-      "epoch": 0.12,
-      "learning_rate": 3.981003752345216e-05,
-      "loss": 0.3682,
+      "epoch": 0.12235556997152845,
+      "grad_norm": 0.3030316233634949,
+      "learning_rate": 4.0775882776843624e-05,
+      "loss": 0.478,
       "step": 3395
     },
     {
-      "epoch": 0.12,
-      "learning_rate": 3.9868667917448405e-05,
-      "loss": 0.3297,
+      "epoch": 0.12253576963275309,
+      "grad_norm": 0.25957420468330383,
+      "learning_rate": 4.083593562334855e-05,
+      "loss": 0.5159,
       "step": 3400
     },
     {
-      "epoch": 0.12,
-      "learning_rate": 3.992729831144466e-05,
-      "loss": 0.3697,
+      "epoch": 0.12271596929397773,
+      "grad_norm": 0.2559145390987396,
+      "learning_rate": 4.0895988469853475e-05,
+      "loss": 0.4766,
       "step": 3405
     },
     {
-      "epoch": 0.12,
-      "learning_rate": 3.99859287054409e-05,
-      "loss": 0.3504,
+      "epoch": 0.12289616895520236,
+      "grad_norm": 0.3006746768951416,
+      "learning_rate": 4.0956041316358394e-05,
+      "loss": 0.4874,
       "step": 3410
     },
     {
-      "epoch": 0.12,
-      "learning_rate": 4.004455909943715e-05,
-      "loss": 0.3731,
+      "epoch": 0.123076368616427,
+      "grad_norm": 0.3356633484363556,
+      "learning_rate": 4.101609416286332e-05,
+      "loss": 0.5086,
       "step": 3415
     },
     {
-      "epoch": 0.12,
-      "learning_rate": 4.0103189493433394e-05,
-      "loss": 0.3451,
+      "epoch": 0.12325656827765163,
+      "grad_norm": 0.3298927843570709,
+      "learning_rate": 4.107614700936825e-05,
+      "loss": 0.5317,
       "step": 3420
     },
     {
-      "epoch": 0.12,
-      "learning_rate": 4.016181988742965e-05,
-      "loss": 0.3504,
+      "epoch": 0.12343676793887627,
+      "grad_norm": 0.30408477783203125,
+      "learning_rate": 4.113619985587317e-05,
+      "loss": 0.5221,
       "step": 3425
     },
     {
-      "epoch": 0.12,
-      "learning_rate": 4.022045028142589e-05,
-      "loss": 0.3507,
+      "epoch": 0.12361696760010091,
+      "grad_norm": 0.300484299659729,
+      "learning_rate": 4.1196252702378096e-05,
+      "loss": 0.4728,
       "step": 3430
     },
     {
-      "epoch": 0.12,
-      "learning_rate": 4.0279080675422145e-05,
-      "loss": 0.331,
+      "epoch": 0.12379716726132554,
+      "grad_norm": 0.3350473642349243,
+      "learning_rate": 4.1256305548883015e-05,
+      "loss": 0.4809,
       "step": 3435
     },
     {
-      "epoch": 0.12,
-      "learning_rate": 4.033771106941839e-05,
-      "loss": 0.3374,
+      "epoch": 0.12397736692255018,
+      "grad_norm": 0.23786677420139313,
+      "learning_rate": 4.131635839538794e-05,
+      "loss": 0.4957,
       "step": 3440
     },
     {
-      "epoch": 0.12,
-      "learning_rate": 4.0396341463414636e-05,
-      "loss": 0.3286,
+      "epoch": 0.12415756658377482,
+      "grad_norm": 0.29893916845321655,
+      "learning_rate": 4.1376411241892866e-05,
+      "loss": 0.5178,
       "step": 3445
     },
     {
-      "epoch": 0.12,
-      "learning_rate": 4.045497185741088e-05,
-      "loss": 0.3469,
+      "epoch": 0.12433776624499945,
+      "grad_norm": 0.29565131664276123,
+      "learning_rate": 4.143646408839779e-05,
+      "loss": 0.4833,
       "step": 3450
     },
     {
-      "epoch": 0.12,
-      "learning_rate": 4.051360225140713e-05,
-      "loss": 0.3475,
+      "epoch": 0.12451796590622409,
+      "grad_norm": 0.27151501178741455,
+      "learning_rate": 4.149651693490272e-05,
+      "loss": 0.4696,
       "step": 3455
     },
     {
-      "epoch": 0.12,
-      "learning_rate": 4.057223264540338e-05,
-      "loss": 0.3546,
+      "epoch": 0.12469816556744873,
+      "grad_norm": 0.23224970698356628,
+      "learning_rate": 4.1556569781407636e-05,
+      "loss": 0.4969,
       "step": 3460
     },
     {
-      "epoch": 0.12,
-      "learning_rate": 4.0630863039399625e-05,
-      "loss": 0.3629,
+      "epoch": 0.12487836522867338,
+      "grad_norm": 0.20471958816051483,
+      "learning_rate": 4.161662262791257e-05,
+      "loss": 0.4783,
       "step": 3465
     },
     {
-      "epoch": 0.12,
-      "learning_rate": 4.068949343339588e-05,
-      "loss": 0.3342,
+      "epoch": 0.125058564889898,
+      "grad_norm": 0.24922746419906616,
+      "learning_rate": 4.167667547441749e-05,
+      "loss": 0.475,
       "step": 3470
     },
     {
-      "epoch": 0.12,
-      "learning_rate": 4.0748123827392123e-05,
-      "loss": 0.3458,
+      "epoch": 0.12523876455112265,
+      "grad_norm": 0.29032933712005615,
+      "learning_rate": 4.173672832092241e-05,
+      "loss": 0.4964,
       "step": 3475
     },
     {
-      "epoch": 0.12,
-      "learning_rate": 4.080675422138837e-05,
-      "loss": 0.3519,
+      "epoch": 0.12541896421234727,
+      "grad_norm": 0.2781813144683838,
+      "learning_rate": 4.179678116742734e-05,
+      "loss": 0.5235,
       "step": 3480
     },
     {
-      "epoch": 0.12,
-      "learning_rate": 4.0865384615384615e-05,
-      "loss": 0.3356,
+      "epoch": 0.12559916387357192,
+      "grad_norm": 0.20956522226333618,
+      "learning_rate": 4.1856834013932264e-05,
+      "loss": 0.4646,
       "step": 3485
     },
     {
-      "epoch": 0.12,
-      "learning_rate": 4.092401500938087e-05,
-      "loss": 0.3324,
+      "epoch": 0.12577936353479655,
+      "grad_norm": 0.2584211528301239,
+      "learning_rate": 4.191688686043719e-05,
+      "loss": 0.5034,
       "step": 3490
     },
     {
-      "epoch": 0.12,
-      "learning_rate": 4.098264540337711e-05,
-      "loss": 0.3544,
+      "epoch": 0.1259595631960212,
+      "grad_norm": 0.2762541174888611,
+      "learning_rate": 4.197693970694211e-05,
+      "loss": 0.4318,
       "step": 3495
     },
     {
-      "epoch": 0.12,
-      "learning_rate": 4.1041275797373365e-05,
-      "loss": 0.3438,
+      "epoch": 0.12613976285724582,
+      "grad_norm": 0.27284300327301025,
+      "learning_rate": 4.2036992553447034e-05,
+      "loss": 0.5069,
       "step": 3500
     },
     {
-      "epoch": 0.12,
-      "eval_loss": 0.3493571877479553,
-      "eval_runtime": 10.5406,
-      "eval_samples_per_second": 9.487,
-      "eval_steps_per_second": 9.487,
+      "epoch": 0.12613976285724582,
+      "eval_loss": 0.5114214420318604,
+      "eval_runtime": 3.5175,
+      "eval_samples_per_second": 28.429,
+      "eval_steps_per_second": 7.107,
       "step": 3500
     },
     {
-      "epoch": 0.12,
-      "learning_rate": 4.109990619136961e-05,
-      "loss": 0.3457,
+      "epoch": 0.12631996251847047,
+      "grad_norm": 0.27490076422691345,
+      "learning_rate": 4.209704539995196e-05,
+      "loss": 0.5157,
       "step": 3505
     },
     {
-      "epoch": 0.12,
-      "learning_rate": 4.1158536585365856e-05,
-      "loss": 0.332,
+      "epoch": 0.1265001621796951,
+      "grad_norm": 0.27542826533317566,
+      "learning_rate": 4.2157098246456885e-05,
+      "loss": 0.4931,
       "step": 3510
     },
     {
-      "epoch": 0.12,
-      "learning_rate": 4.12171669793621e-05,
-      "loss": 0.3388,
+      "epoch": 0.12668036184091974,
+      "grad_norm": 0.25971895456314087,
+      "learning_rate": 4.221715109296181e-05,
+      "loss": 0.4658,
       "step": 3515
     },
     {
-      "epoch": 0.12,
-      "learning_rate": 4.127579737335835e-05,
-      "loss": 0.3509,
+      "epoch": 0.12686056150214436,
+      "grad_norm": 0.26690730452537537,
+      "learning_rate": 4.227720393946673e-05,
+      "loss": 0.4838,
       "step": 3520
     },
     {
-      "epoch": 0.12,
-      "learning_rate": 4.13344277673546e-05,
-      "loss": 0.3456,
+      "epoch": 0.12704076116336901,
+      "grad_norm": 0.3250514566898346,
+      "learning_rate": 4.2337256785971655e-05,
+      "loss": 0.5006,
       "step": 3525
     },
     {
-      "epoch": 0.12,
-      "learning_rate": 4.1393058161350846e-05,
-      "loss": 0.335,
+      "epoch": 0.12722096082459364,
+      "grad_norm": 0.23514829576015472,
+      "learning_rate": 4.239730963247659e-05,
+      "loss": 0.4743,
       "step": 3530
     },
     {
-      "epoch": 0.12,
-      "learning_rate": 4.14516885553471e-05,
-      "loss": 0.3253,
+      "epoch": 0.1274011604858183,
+      "grad_norm": 0.2410300076007843,
+      "learning_rate": 4.2457362478981506e-05,
+      "loss": 0.4643,
       "step": 3535
     },
     {
-      "epoch": 0.12,
-      "learning_rate": 4.1510318949343344e-05,
-      "loss": 0.3525,
+      "epoch": 0.1275813601470429,
+      "grad_norm": 0.3281730115413666,
+      "learning_rate": 4.251741532548643e-05,
+      "loss": 0.4714,
       "step": 3540
     },
     {
-      "epoch": 0.12,
-      "learning_rate": 4.156894934333959e-05,
-      "loss": 0.3493,
+      "epoch": 0.12776155980826756,
+      "grad_norm": 0.28623196482658386,
+      "learning_rate": 4.257746817199135e-05,
+      "loss": 0.5064,
       "step": 3545
     },
     {
-      "epoch": 0.12,
-      "learning_rate": 4.1627579737335835e-05,
-      "loss": 0.3619,
+      "epoch": 0.1279417594694922,
+      "grad_norm": 0.2539494037628174,
+      "learning_rate": 4.2637521018496276e-05,
+      "loss": 0.4777,
       "step": 3550
     },
     {
-      "epoch": 0.13,
-      "learning_rate": 4.168621013133208e-05,
-      "loss": 0.3241,
+      "epoch": 0.12812195913071683,
+      "grad_norm": 0.26674479246139526,
+      "learning_rate": 4.269757386500121e-05,
+      "loss": 0.4719,
       "step": 3555
     },
     {
-      "epoch": 0.13,
-      "learning_rate": 4.174484052532833e-05,
-      "loss": 0.3642,
+      "epoch": 0.12830215879194148,
+      "grad_norm": 0.23590758442878723,
+      "learning_rate": 4.275762671150613e-05,
+      "loss": 0.4984,
       "step": 3560
     },
     {
-      "epoch": 0.13,
-      "learning_rate": 4.180347091932458e-05,
-      "loss": 0.3621,
+      "epoch": 0.1284823584531661,
+      "grad_norm": 0.26926812529563904,
+      "learning_rate": 4.281767955801105e-05,
+      "loss": 0.5151,
       "step": 3565
     },
     {
-      "epoch": 0.13,
-      "learning_rate": 4.186210131332083e-05,
-      "loss": 0.3247,
+      "epoch": 0.12866255811439076,
+      "grad_norm": 0.2809188663959503,
+      "learning_rate": 4.287773240451597e-05,
+      "loss": 0.5129,
       "step": 3570
     },
     {
-      "epoch": 0.13,
-      "learning_rate": 4.1920731707317077e-05,
-      "loss": 0.3477,
+      "epoch": 0.12884275777561538,
+      "grad_norm": 0.2685711681842804,
+      "learning_rate": 4.2937785251020904e-05,
+      "loss": 0.4916,
       "step": 3575
     },
     {
-      "epoch": 0.13,
-      "learning_rate": 4.197936210131332e-05,
-      "loss": 0.3352,
+      "epoch": 0.12902295743684003,
+      "grad_norm": 0.2880760431289673,
+      "learning_rate": 4.299783809752582e-05,
+      "loss": 0.504,
       "step": 3580
     },
     {
-      "epoch": 0.13,
-      "learning_rate": 4.203799249530957e-05,
-      "loss": 0.3672,
+      "epoch": 0.12920315709806465,
+      "grad_norm": 0.28242960572242737,
+      "learning_rate": 4.305789094403075e-05,
+      "loss": 0.51,
       "step": 3585
     },
     {
-      "epoch": 0.13,
-      "learning_rate": 4.209662288930582e-05,
-      "loss": 0.3714,
+      "epoch": 0.1293833567592893,
+      "grad_norm": 0.22377252578735352,
+      "learning_rate": 4.3117943790535674e-05,
+      "loss": 0.471,
       "step": 3590
     },
     {
-      "epoch": 0.13,
-      "learning_rate": 4.2155253283302066e-05,
-      "loss": 0.3332,
+      "epoch": 0.12956355642051393,
+      "grad_norm": 0.23554405570030212,
+      "learning_rate": 4.31779966370406e-05,
+      "loss": 0.497,
       "step": 3595
     },
     {
-      "epoch": 0.13,
-      "learning_rate": 4.221388367729832e-05,
-      "loss": 0.3222,
+      "epoch": 0.12974375608173858,
+      "grad_norm": 0.3754778802394867,
+      "learning_rate": 4.3238049483545525e-05,
+      "loss": 0.4746,
       "step": 3600
     },
     {
-      "epoch": 0.13,
-      "learning_rate": 4.2272514071294564e-05,
-      "loss": 0.3322,
+      "epoch": 0.1299239557429632,
+      "grad_norm": 0.2620472013950348,
+      "learning_rate": 4.3298102330050444e-05,
+      "loss": 0.4682,
       "step": 3605
     },
     {
-      "epoch": 0.13,
-      "learning_rate": 4.233114446529081e-05,
-      "loss": 0.345,
+      "epoch": 0.13010415540418785,
+      "grad_norm": 0.25988104939460754,
+      "learning_rate": 4.335815517655537e-05,
+      "loss": 0.4681,
       "step": 3610
     },
     {
-      "epoch": 0.13,
-      "learning_rate": 4.2389774859287055e-05,
-      "loss": 0.3589,
+      "epoch": 0.13028435506541247,
+      "grad_norm": 0.2066163271665573,
+      "learning_rate": 4.3418208023060295e-05,
+      "loss": 0.4465,
       "step": 3615
     },
     {
-      "epoch": 0.13,
-      "learning_rate": 4.24484052532833e-05,
-      "loss": 0.3631,
+      "epoch": 0.13046455472663712,
+      "grad_norm": 0.28050124645233154,
+      "learning_rate": 4.347826086956522e-05,
+      "loss": 0.4737,
       "step": 3620
     },
     {
-      "epoch": 0.13,
-      "learning_rate": 4.250703564727955e-05,
-      "loss": 0.3355,
+      "epoch": 0.13064475438786174,
+      "grad_norm": 0.32841894030570984,
+      "learning_rate": 4.3538313716070146e-05,
+      "loss": 0.4578,
       "step": 3625
     },
     {
-      "epoch": 0.13,
-      "learning_rate": 4.25656660412758e-05,
-      "loss": 0.3479,
+      "epoch": 0.1308249540490864,
+      "grad_norm": 0.2937125861644745,
+      "learning_rate": 4.3598366562575065e-05,
+      "loss": 0.4655,
       "step": 3630
     },
     {
-      "epoch": 0.13,
-      "learning_rate": 4.262429643527205e-05,
-      "loss": 0.3446,
+      "epoch": 0.13100515371031102,
+      "grad_norm": 0.2880992889404297,
+      "learning_rate": 4.365841940907999e-05,
+      "loss": 0.48,
       "step": 3635
     },
     {
-      "epoch": 0.13,
-      "learning_rate": 4.26829268292683e-05,
-      "loss": 0.3448,
+      "epoch": 0.13118535337153567,
+      "grad_norm": 0.23047612607479095,
+      "learning_rate": 4.3718472255584916e-05,
+      "loss": 0.5327,
       "step": 3640
     },
     {
-      "epoch": 0.13,
-      "learning_rate": 4.274155722326454e-05,
-      "loss": 0.3515,
+      "epoch": 0.1313655530327603,
+      "grad_norm": 0.27166831493377686,
+      "learning_rate": 4.377852510208984e-05,
+      "loss": 0.4772,
       "step": 3645
     },
     {
-      "epoch": 0.13,
-      "learning_rate": 4.280018761726079e-05,
-      "loss": 0.3433,
+      "epoch": 0.13154575269398494,
+      "grad_norm": 0.3133472502231598,
+      "learning_rate": 4.383857794859477e-05,
+      "loss": 0.5161,
       "step": 3650
     },
     {
-      "epoch": 0.13,
-      "learning_rate": 4.285881801125704e-05,
-      "loss": 0.3223,
+      "epoch": 0.13172595235520956,
+      "grad_norm": 0.25573498010635376,
+      "learning_rate": 4.3898630795099686e-05,
+      "loss": 0.5252,
       "step": 3655
     },
     {
-      "epoch": 0.13,
-      "learning_rate": 4.2917448405253286e-05,
-      "loss": 0.3672,
+      "epoch": 0.1319061520164342,
+      "grad_norm": 0.23422662913799286,
+      "learning_rate": 4.395868364160461e-05,
+      "loss": 0.4795,
       "step": 3660
     },
     {
-      "epoch": 0.13,
-      "learning_rate": 4.297607879924953e-05,
-      "loss": 0.3445,
+      "epoch": 0.13208635167765884,
+      "grad_norm": 0.19992312788963318,
+      "learning_rate": 4.4018736488109544e-05,
+      "loss": 0.5103,
       "step": 3665
     },
     {
-      "epoch": 0.13,
-      "learning_rate": 4.3034709193245784e-05,
-      "loss": 0.3612,
+      "epoch": 0.1322665513388835,
+      "grad_norm": 0.22987040877342224,
+      "learning_rate": 4.407878933461446e-05,
+      "loss": 0.5037,
       "step": 3670
     },
     {
-      "epoch": 0.13,
-      "learning_rate": 4.309333958724203e-05,
-      "loss": 0.3635,
+      "epoch": 0.1324467510001081,
+      "grad_norm": 0.35055404901504517,
+      "learning_rate": 4.413884218111939e-05,
+      "loss": 0.5178,
       "step": 3675
     },
     {
-      "epoch": 0.13,
-      "learning_rate": 4.3151969981238275e-05,
-      "loss": 0.3712,
+      "epoch": 0.13262695066133276,
+      "grad_norm": 0.2555431127548218,
+      "learning_rate": 4.419889502762431e-05,
+      "loss": 0.4925,
       "step": 3680
     },
     {
-      "epoch": 0.13,
-      "learning_rate": 4.321060037523452e-05,
-      "loss": 0.3576,
+      "epoch": 0.13280715032255738,
+      "grad_norm": 0.2736440598964691,
+      "learning_rate": 4.425894787412924e-05,
+      "loss": 0.5068,
       "step": 3685
     },
     {
-      "epoch": 0.13,
-      "learning_rate": 4.326923076923077e-05,
-      "loss": 0.3451,
+      "epoch": 0.13298734998378203,
+      "grad_norm": 0.20614711940288544,
+      "learning_rate": 4.4319000720634165e-05,
+      "loss": 0.4919,
       "step": 3690
     },
     {
-      "epoch": 0.13,
-      "learning_rate": 4.332786116322702e-05,
-      "loss": 0.337,
+      "epoch": 0.13316754964500666,
+      "grad_norm": 0.29815688729286194,
+      "learning_rate": 4.4379053567139083e-05,
+      "loss": 0.5117,
       "step": 3695
     },
     {
-      "epoch": 0.13,
-      "learning_rate": 4.338649155722327e-05,
-      "loss": 0.3955,
+      "epoch": 0.1333477493062313,
+      "grad_norm": 0.23376838862895966,
+      "learning_rate": 4.443910641364401e-05,
+      "loss": 0.5105,
       "step": 3700
     },
     {
-      "epoch": 0.13,
-      "learning_rate": 4.344512195121952e-05,
-      "loss": 0.3417,
+      "epoch": 0.13352794896745593,
+      "grad_norm": 0.27070969343185425,
+      "learning_rate": 4.4499159260148935e-05,
+      "loss": 0.4692,
       "step": 3705
     },
     {
-      "epoch": 0.13,
-      "learning_rate": 4.350375234521576e-05,
-      "loss": 0.353,
+      "epoch": 0.13370814862868058,
+      "grad_norm": 0.2089190036058426,
+      "learning_rate": 4.455921210665386e-05,
+      "loss": 0.4596,
       "step": 3710
     },
     {
-      "epoch": 0.13,
-      "learning_rate": 4.356238273921201e-05,
-      "loss": 0.3516,
+      "epoch": 0.1338883482899052,
+      "grad_norm": 0.331286758184433,
+      "learning_rate": 4.461926495315878e-05,
+      "loss": 0.5144,
       "step": 3715
     },
     {
-      "epoch": 0.13,
-      "learning_rate": 4.3621013133208254e-05,
-      "loss": 0.3788,
+      "epoch": 0.13406854795112985,
+      "grad_norm": 0.2310916930437088,
+      "learning_rate": 4.4679317799663705e-05,
+      "loss": 0.5188,
       "step": 3720
     },
     {
-      "epoch": 0.13,
-      "learning_rate": 4.3679643527204506e-05,
-      "loss": 0.3611,
+      "epoch": 0.1342487476123545,
+      "grad_norm": 0.21956689655780792,
+      "learning_rate": 4.473937064616863e-05,
+      "loss": 0.4799,
       "step": 3725
     },
     {
-      "epoch": 0.13,
-      "learning_rate": 4.373827392120075e-05,
-      "loss": 0.3479,
+      "epoch": 0.13442894727357912,
+      "grad_norm": 0.2846265733242035,
+      "learning_rate": 4.4799423492673556e-05,
+      "loss": 0.525,
       "step": 3730
     },
     {
-      "epoch": 0.13,
-      "learning_rate": 4.3796904315197004e-05,
-      "loss": 0.3494,
+      "epoch": 0.13460914693480378,
+      "grad_norm": 0.2660646140575409,
+      "learning_rate": 4.485947633917848e-05,
+      "loss": 0.451,
       "step": 3735
     },
     {
-      "epoch": 0.13,
-      "learning_rate": 4.385553470919325e-05,
-      "loss": 0.3495,
+      "epoch": 0.1347893465960284,
+      "grad_norm": 0.30038926005363464,
+      "learning_rate": 4.49195291856834e-05,
+      "loss": 0.4577,
       "step": 3740
     },
     {
-      "epoch": 0.13,
-      "learning_rate": 4.3914165103189495e-05,
-      "loss": 0.3221,
+      "epoch": 0.13496954625725305,
+      "grad_norm": 0.23106199502944946,
+      "learning_rate": 4.4979582032188326e-05,
+      "loss": 0.5007,
       "step": 3745
     },
     {
-      "epoch": 0.13,
-      "learning_rate": 4.397279549718574e-05,
-      "loss": 0.3679,
+      "epoch": 0.13514974591847767,
+      "grad_norm": 0.24195370078086853,
+      "learning_rate": 4.503963487869325e-05,
+      "loss": 0.4571,
       "step": 3750
     },
     {
-      "epoch": 0.13,
-      "learning_rate": 4.4031425891181993e-05,
-      "loss": 0.3789,
+      "epoch": 0.13532994557970232,
+      "grad_norm": 0.3315066993236542,
+      "learning_rate": 4.509968772519818e-05,
+      "loss": 0.5199,
       "step": 3755
     },
     {
-      "epoch": 0.13,
-      "learning_rate": 4.409005628517824e-05,
-      "loss": 0.3436,
+      "epoch": 0.13551014524092694,
+      "grad_norm": 0.23671256005764008,
+      "learning_rate": 4.51597405717031e-05,
+      "loss": 0.486,
       "step": 3760
     },
     {
-      "epoch": 0.13,
-      "learning_rate": 4.4148686679174485e-05,
-      "loss": 0.3464,
+      "epoch": 0.1356903449021516,
+      "grad_norm": 0.2768378257751465,
+      "learning_rate": 4.521979341820802e-05,
+      "loss": 0.478,
       "step": 3765
     },
     {
-      "epoch": 0.13,
-      "learning_rate": 4.420731707317074e-05,
-      "loss": 0.3243,
+      "epoch": 0.13587054456337622,
+      "grad_norm": 0.20859333872795105,
+      "learning_rate": 4.5279846264712947e-05,
+      "loss": 0.4651,
       "step": 3770
     },
     {
-      "epoch": 0.13,
-      "learning_rate": 4.426594746716698e-05,
-      "loss": 0.3591,
+      "epoch": 0.13605074422460087,
+      "grad_norm": 0.28573447465896606,
+      "learning_rate": 4.533989911121788e-05,
+      "loss": 0.4645,
       "step": 3775
     },
     {
-      "epoch": 0.13,
-      "learning_rate": 4.432457786116323e-05,
-      "loss": 0.3671,
+      "epoch": 0.1362309438858255,
+      "grad_norm": 0.2684060335159302,
+      "learning_rate": 4.53999519577228e-05,
+      "loss": 0.4407,
       "step": 3780
     },
     {
-      "epoch": 0.13,
-      "learning_rate": 4.4383208255159474e-05,
-      "loss": 0.3528,
+      "epoch": 0.13641114354705014,
+      "grad_norm": 0.24884513020515442,
+      "learning_rate": 4.546000480422772e-05,
+      "loss": 0.4747,
       "step": 3785
     },
     {
-      "epoch": 0.13,
-      "learning_rate": 4.4441838649155726e-05,
-      "loss": 0.3637,
+      "epoch": 0.13659134320827476,
+      "grad_norm": 0.25004905462265015,
+      "learning_rate": 4.552005765073264e-05,
+      "loss": 0.4725,
       "step": 3790
     },
     {
-      "epoch": 0.13,
-      "learning_rate": 4.450046904315197e-05,
-      "loss": 0.3185,
+      "epoch": 0.1367715428694994,
+      "grad_norm": 0.28418663144111633,
+      "learning_rate": 4.5580110497237574e-05,
+      "loss": 0.5091,
       "step": 3795
     },
     {
-      "epoch": 0.13,
-      "learning_rate": 4.4559099437148224e-05,
-      "loss": 0.3577,
+      "epoch": 0.13695174253072404,
+      "grad_norm": 0.2636741101741791,
+      "learning_rate": 4.56401633437425e-05,
+      "loss": 0.4824,
       "step": 3800
     },
     {
-      "epoch": 0.13,
-      "learning_rate": 4.461772983114447e-05,
-      "loss": 0.347,
+      "epoch": 0.13713194219194869,
+      "grad_norm": 0.22315765917301178,
+      "learning_rate": 4.570021619024742e-05,
+      "loss": 0.4479,
       "step": 3805
     },
     {
-      "epoch": 0.13,
-      "learning_rate": 4.4676360225140716e-05,
-      "loss": 0.3868,
+      "epoch": 0.1373121418531733,
+      "grad_norm": 0.26647093892097473,
+      "learning_rate": 4.5760269036752344e-05,
+      "loss": 0.4664,
       "step": 3810
     },
     {
-      "epoch": 0.13,
-      "learning_rate": 4.473499061913696e-05,
-      "loss": 0.3359,
+      "epoch": 0.13749234151439796,
+      "grad_norm": 0.26308777928352356,
+      "learning_rate": 4.582032188325727e-05,
+      "loss": 0.5274,
       "step": 3815
     },
     {
-      "epoch": 0.13,
-      "learning_rate": 4.479362101313321e-05,
-      "loss": 0.3245,
+      "epoch": 0.13767254117562258,
+      "grad_norm": 0.2349170744419098,
+      "learning_rate": 4.5880374729762195e-05,
+      "loss": 0.4634,
       "step": 3820
     },
     {
-      "epoch": 0.13,
-      "learning_rate": 4.485225140712946e-05,
-      "loss": 0.3556,
+      "epoch": 0.13785274083684723,
+      "grad_norm": 0.21153903007507324,
+      "learning_rate": 4.594042757626712e-05,
+      "loss": 0.4549,
       "step": 3825
     },
     {
-      "epoch": 0.13,
-      "learning_rate": 4.4910881801125705e-05,
-      "loss": 0.3418,
+      "epoch": 0.13803294049807185,
+      "grad_norm": 0.282583624124527,
+      "learning_rate": 4.600048042277204e-05,
+      "loss": 0.5048,
       "step": 3830
     },
     {
-      "epoch": 0.13,
-      "learning_rate": 4.496951219512196e-05,
-      "loss": 0.3455,
+      "epoch": 0.1382131401592965,
+      "grad_norm": 0.24750690162181854,
+      "learning_rate": 4.6060533269276965e-05,
+      "loss": 0.5016,
       "step": 3835
     },
     {
-      "epoch": 0.14,
-      "learning_rate": 4.50281425891182e-05,
-      "loss": 0.3431,
+      "epoch": 0.13839333982052113,
+      "grad_norm": 0.28650569915771484,
+      "learning_rate": 4.612058611578189e-05,
+      "loss": 0.503,
       "step": 3840
     },
     {
-      "epoch": 0.14,
-      "learning_rate": 4.508677298311445e-05,
-      "loss": 0.3509,
+      "epoch": 0.13857353948174578,
+      "grad_norm": 0.28856590390205383,
+      "learning_rate": 4.6180638962286817e-05,
+      "loss": 0.4748,
       "step": 3845
     },
     {
-      "epoch": 0.14,
-      "learning_rate": 4.5145403377110694e-05,
-      "loss": 0.3297,
+      "epoch": 0.1387537391429704,
+      "grad_norm": 0.24171088635921478,
+      "learning_rate": 4.6240691808791735e-05,
+      "loss": 0.4835,
       "step": 3850
     },
     {
-      "epoch": 0.14,
-      "learning_rate": 4.5204033771106947e-05,
-      "loss": 0.3432,
+      "epoch": 0.13893393880419505,
+      "grad_norm": 0.3100346028804779,
+      "learning_rate": 4.630074465529666e-05,
+      "loss": 0.4756,
       "step": 3855
     },
     {
-      "epoch": 0.14,
-      "learning_rate": 4.526266416510319e-05,
-      "loss": 0.3348,
+      "epoch": 0.13911413846541967,
+      "grad_norm": 0.18996688723564148,
+      "learning_rate": 4.6360797501801586e-05,
+      "loss": 0.5116,
       "step": 3860
     },
     {
-      "epoch": 0.14,
-      "learning_rate": 4.5321294559099445e-05,
-      "loss": 0.3428,
+      "epoch": 0.13929433812664432,
+      "grad_norm": 0.2496083527803421,
+      "learning_rate": 4.642085034830651e-05,
+      "loss": 0.4598,
       "step": 3865
     },
     {
-      "epoch": 0.14,
-      "learning_rate": 4.537992495309569e-05,
-      "loss": 0.3588,
+      "epoch": 0.13947453778786895,
+      "grad_norm": 0.27395179867744446,
+      "learning_rate": 4.648090319481144e-05,
+      "loss": 0.4652,
       "step": 3870
     },
     {
-      "epoch": 0.14,
-      "learning_rate": 4.543855534709193e-05,
-      "loss": 0.3464,
+      "epoch": 0.1396547374490936,
+      "grad_norm": 0.24019189178943634,
+      "learning_rate": 4.6540956041316356e-05,
+      "loss": 0.4794,
       "step": 3875
     },
     {
-      "epoch": 0.14,
-      "learning_rate": 4.549718574108818e-05,
-      "loss": 0.3704,
+      "epoch": 0.13983493711031822,
+      "grad_norm": 0.24794286489486694,
+      "learning_rate": 4.660100888782128e-05,
+      "loss": 0.4893,
       "step": 3880
     },
     {
-      "epoch": 0.14,
-      "learning_rate": 4.555581613508443e-05,
-      "loss": 0.3672,
+      "epoch": 0.14001513677154287,
+      "grad_norm": 0.2861761450767517,
+      "learning_rate": 4.6661061734326214e-05,
+      "loss": 0.5465,
       "step": 3885
     },
     {
-      "epoch": 0.14,
-      "learning_rate": 4.561444652908068e-05,
-      "loss": 0.3704,
+      "epoch": 0.14019533643276752,
+      "grad_norm": 0.24302253127098083,
+      "learning_rate": 4.672111458083113e-05,
+      "loss": 0.5152,
       "step": 3890
     },
     {
-      "epoch": 0.14,
-      "learning_rate": 4.5673076923076925e-05,
-      "loss": 0.3336,
+      "epoch": 0.14037553609399214,
+      "grad_norm": 0.25176292657852173,
+      "learning_rate": 4.678116742733606e-05,
+      "loss": 0.5149,
       "step": 3895
     },
     {
-      "epoch": 0.14,
-      "learning_rate": 4.573170731707318e-05,
-      "loss": 0.3272,
+      "epoch": 0.1405557357552168,
+      "grad_norm": 0.3168286085128784,
+      "learning_rate": 4.684122027384098e-05,
+      "loss": 0.5241,
       "step": 3900
     },
     {
-      "epoch": 0.14,
-      "learning_rate": 4.579033771106942e-05,
-      "loss": 0.3223,
+      "epoch": 0.14073593541644142,
+      "grad_norm": 0.3177297115325928,
+      "learning_rate": 4.690127312034591e-05,
+      "loss": 0.5131,
       "step": 3905
     },
     {
-      "epoch": 0.14,
-      "learning_rate": 4.584896810506567e-05,
-      "loss": 0.361,
+      "epoch": 0.14091613507766607,
+      "grad_norm": 0.30368152260780334,
+      "learning_rate": 4.6961325966850835e-05,
+      "loss": 0.5248,
       "step": 3910
     },
     {
-      "epoch": 0.14,
-      "learning_rate": 4.5907598499061914e-05,
-      "loss": 0.3281,
+      "epoch": 0.1410963347388907,
+      "grad_norm": 0.2546997368335724,
+      "learning_rate": 4.7021378813355754e-05,
+      "loss": 0.4857,
       "step": 3915
     },
     {
-      "epoch": 0.14,
-      "learning_rate": 4.596622889305816e-05,
-      "loss": 0.3614,
+      "epoch": 0.14127653440011534,
+      "grad_norm": 0.23904703557491302,
+      "learning_rate": 4.708143165986068e-05,
+      "loss": 0.4906,
       "step": 3920
     },
     {
-      "epoch": 0.14,
-      "learning_rate": 4.602485928705441e-05,
-      "loss": 0.3411,
+      "epoch": 0.14145673406133996,
+      "grad_norm": 0.26533734798431396,
+      "learning_rate": 4.7141484506365605e-05,
+      "loss": 0.4906,
       "step": 3925
     },
     {
-      "epoch": 0.14,
-      "learning_rate": 4.608348968105066e-05,
-      "loss": 0.3313,
+      "epoch": 0.1416369337225646,
+      "grad_norm": 0.3533400297164917,
+      "learning_rate": 4.720153735287053e-05,
+      "loss": 0.4932,
       "step": 3930
     },
     {
-      "epoch": 0.14,
-      "learning_rate": 4.614212007504691e-05,
-      "loss": 0.3629,
+      "epoch": 0.14181713338378923,
+      "grad_norm": 0.31380826234817505,
+      "learning_rate": 4.7261590199375456e-05,
+      "loss": 0.4983,
       "step": 3935
     },
     {
-      "epoch": 0.14,
-      "learning_rate": 4.620075046904315e-05,
-      "loss": 0.3631,
+      "epoch": 0.14199733304501388,
+      "grad_norm": 0.25796714425086975,
+      "learning_rate": 4.7321643045880375e-05,
+      "loss": 0.4602,
       "step": 3940
     },
     {
-      "epoch": 0.14,
-      "learning_rate": 4.62593808630394e-05,
-      "loss": 0.3539,
+      "epoch": 0.1421775327062385,
+      "grad_norm": 0.2745930850505829,
+      "learning_rate": 4.73816958923853e-05,
+      "loss": 0.4592,
       "step": 3945
     },
     {
-      "epoch": 0.14,
-      "learning_rate": 4.631801125703565e-05,
-      "loss": 0.3254,
+      "epoch": 0.14235773236746316,
+      "grad_norm": 0.24087023735046387,
+      "learning_rate": 4.7441748738890226e-05,
+      "loss": 0.4606,
       "step": 3950
     },
     {
-      "epoch": 0.14,
-      "learning_rate": 4.63766416510319e-05,
-      "loss": 0.3704,
+      "epoch": 0.14253793202868778,
+      "grad_norm": 0.2588431239128113,
+      "learning_rate": 4.750180158539515e-05,
+      "loss": 0.4881,
       "step": 3955
     },
     {
-      "epoch": 0.14,
-      "learning_rate": 4.6435272045028145e-05,
-      "loss": 0.3478,
+      "epoch": 0.14271813168991243,
+      "grad_norm": 0.26731452345848083,
+      "learning_rate": 4.756185443190008e-05,
+      "loss": 0.4715,
       "step": 3960
     },
     {
-      "epoch": 0.14,
-      "learning_rate": 4.64939024390244e-05,
-      "loss": 0.3576,
+      "epoch": 0.14289833135113705,
+      "grad_norm": 0.2952512204647064,
+      "learning_rate": 4.7621907278404996e-05,
+      "loss": 0.4721,
       "step": 3965
     },
     {
-      "epoch": 0.14,
-      "learning_rate": 4.655253283302064e-05,
-      "loss": 0.3851,
+      "epoch": 0.1430785310123617,
+      "grad_norm": 0.19790342450141907,
+      "learning_rate": 4.768196012490992e-05,
+      "loss": 0.473,
       "step": 3970
     },
     {
-      "epoch": 0.14,
-      "learning_rate": 4.661116322701688e-05,
-      "loss": 0.3313,
+      "epoch": 0.14325873067358633,
+      "grad_norm": 0.23839952051639557,
+      "learning_rate": 4.774201297141485e-05,
+      "loss": 0.5295,
       "step": 3975
     },
     {
-      "epoch": 0.14,
-      "learning_rate": 4.6669793621013135e-05,
-      "loss": 0.3325,
+      "epoch": 0.14343893033481098,
+      "grad_norm": 0.2507927715778351,
+      "learning_rate": 4.780206581791977e-05,
+      "loss": 0.4815,
       "step": 3980
     },
     {
-      "epoch": 0.14,
-      "learning_rate": 4.672842401500938e-05,
-      "loss": 0.3399,
+      "epoch": 0.1436191299960356,
+      "grad_norm": 0.24117940664291382,
+      "learning_rate": 4.786211866442469e-05,
+      "loss": 0.4546,
       "step": 3985
     },
     {
-      "epoch": 0.14,
-      "learning_rate": 4.678705440900563e-05,
-      "loss": 0.3315,
+      "epoch": 0.14379932965726025,
+      "grad_norm": 0.2980533838272095,
+      "learning_rate": 4.792217151092962e-05,
+      "loss": 0.4767,
       "step": 3990
     },
     {
-      "epoch": 0.14,
-      "learning_rate": 4.684568480300188e-05,
-      "loss": 0.3511,
+      "epoch": 0.14397952931848487,
+      "grad_norm": 0.2951893210411072,
+      "learning_rate": 4.798222435743455e-05,
+      "loss": 0.4729,
       "step": 3995
     },
     {
-      "epoch": 0.14,
-      "learning_rate": 4.690431519699813e-05,
-      "loss": 0.3507,
+      "epoch": 0.14415972897970952,
+      "grad_norm": 0.19777187705039978,
+      "learning_rate": 4.804227720393947e-05,
+      "loss": 0.4814,
       "step": 4000
     },
     {
-      "epoch": 0.14,
-      "eval_loss": 0.34727856516838074,
-      "eval_runtime": 10.5437,
-      "eval_samples_per_second": 9.484,
-      "eval_steps_per_second": 9.484,
+      "epoch": 0.14415972897970952,
+      "eval_loss": 0.5076904296875,
+      "eval_runtime": 3.5171,
+      "eval_samples_per_second": 28.432,
+      "eval_steps_per_second": 7.108,
       "step": 4000
     },
     {
-      "epoch": 0.14,
-      "learning_rate": 4.696294559099437e-05,
-      "loss": 0.3252,
+      "epoch": 0.14433992864093415,
+      "grad_norm": 0.21895861625671387,
+      "learning_rate": 4.8102330050444394e-05,
+      "loss": 0.495,
       "step": 4005
     },
     {
-      "epoch": 0.14,
-      "learning_rate": 4.702157598499062e-05,
-      "loss": 0.3467,
+      "epoch": 0.1445201283021588,
+      "grad_norm": 0.2850351631641388,
+      "learning_rate": 4.816238289694931e-05,
+      "loss": 0.4767,
       "step": 4010
     },
     {
-      "epoch": 0.14,
-      "learning_rate": 4.708020637898687e-05,
-      "loss": 0.3564,
+      "epoch": 0.14470032796338342,
+      "grad_norm": 0.26362937688827515,
+      "learning_rate": 4.8222435743454245e-05,
+      "loss": 0.5319,
       "step": 4015
     },
     {
-      "epoch": 0.14,
-      "learning_rate": 4.713883677298312e-05,
-      "loss": 0.3565,
+      "epoch": 0.14488052762460807,
+      "grad_norm": 0.25553131103515625,
+      "learning_rate": 4.828248858995917e-05,
+      "loss": 0.4739,
       "step": 4020
     },
     {
-      "epoch": 0.14,
-      "learning_rate": 4.7197467166979365e-05,
-      "loss": 0.3467,
+      "epoch": 0.1450607272858327,
+      "grad_norm": 0.2785678505897522,
+      "learning_rate": 4.834254143646409e-05,
+      "loss": 0.5092,
       "step": 4025
     },
     {
-      "epoch": 0.14,
-      "learning_rate": 4.725609756097561e-05,
-      "loss": 0.3248,
+      "epoch": 0.14524092694705734,
+      "grad_norm": 0.25108572840690613,
+      "learning_rate": 4.8402594282969015e-05,
+      "loss": 0.496,
       "step": 4030
     },
     {
-      "epoch": 0.14,
-      "learning_rate": 4.7314727954971863e-05,
-      "loss": 0.3628,
+      "epoch": 0.14542112660828196,
+      "grad_norm": 0.21198877692222595,
+      "learning_rate": 4.8462647129473934e-05,
+      "loss": 0.4941,
       "step": 4035
     },
     {
-      "epoch": 0.14,
-      "learning_rate": 4.73733583489681e-05,
-      "loss": 0.3153,
+      "epoch": 0.14560132626950661,
+      "grad_norm": 0.23192477226257324,
+      "learning_rate": 4.8522699975978866e-05,
+      "loss": 0.486,
       "step": 4040
     },
     {
-      "epoch": 0.14,
-      "learning_rate": 4.7431988742964355e-05,
-      "loss": 0.3289,
+      "epoch": 0.14578152593073124,
+      "grad_norm": 0.25340479612350464,
+      "learning_rate": 4.858275282248379e-05,
+      "loss": 0.5056,
       "step": 4045
     },
     {
-      "epoch": 0.14,
-      "learning_rate": 4.74906191369606e-05,
-      "loss": 0.3446,
+      "epoch": 0.1459617255919559,
+      "grad_norm": 0.28752046823501587,
+      "learning_rate": 4.864280566898871e-05,
+      "loss": 0.4944,
       "step": 4050
     },
     {
-      "epoch": 0.14,
-      "learning_rate": 4.754924953095685e-05,
-      "loss": 0.3631,
+      "epoch": 0.1461419252531805,
+      "grad_norm": 0.2980968952178955,
+      "learning_rate": 4.8702858515493636e-05,
+      "loss": 0.4604,
       "step": 4055
     },
     {
-      "epoch": 0.14,
-      "learning_rate": 4.76078799249531e-05,
-      "loss": 0.3446,
+      "epoch": 0.14632212491440516,
+      "grad_norm": 0.2499280720949173,
+      "learning_rate": 4.876291136199856e-05,
+      "loss": 0.52,
       "step": 4060
     },
     {
-      "epoch": 0.14,
-      "learning_rate": 4.766651031894935e-05,
-      "loss": 0.3435,
+      "epoch": 0.1465023245756298,
+      "grad_norm": 0.2250458002090454,
+      "learning_rate": 4.882296420850349e-05,
+      "loss": 0.475,
       "step": 4065
     },
     {
-      "epoch": 0.14,
-      "learning_rate": 4.772514071294559e-05,
-      "loss": 0.3314,
+      "epoch": 0.14668252423685443,
+      "grad_norm": 0.20067259669303894,
+      "learning_rate": 4.888301705500841e-05,
+      "loss": 0.4465,
       "step": 4070
     },
     {
-      "epoch": 0.14,
-      "learning_rate": 4.778377110694184e-05,
-      "loss": 0.3587,
+      "epoch": 0.14686272389807908,
+      "grad_norm": 0.2999953329563141,
+      "learning_rate": 4.894306990151333e-05,
+      "loss": 0.5123,
       "step": 4075
     },
     {
-      "epoch": 0.14,
-      "learning_rate": 4.784240150093809e-05,
-      "loss": 0.3589,
+      "epoch": 0.1470429235593037,
+      "grad_norm": 0.26574891805648804,
+      "learning_rate": 4.900312274801826e-05,
+      "loss": 0.4827,
       "step": 4080
     },
     {
-      "epoch": 0.14,
-      "learning_rate": 4.790103189493433e-05,
-      "loss": 0.3404,
+      "epoch": 0.14722312322052836,
+      "grad_norm": 0.2696743309497833,
+      "learning_rate": 4.906317559452318e-05,
+      "loss": 0.4896,
       "step": 4085
     },
     {
-      "epoch": 0.14,
-      "learning_rate": 4.7959662288930586e-05,
-      "loss": 0.3554,
+      "epoch": 0.14740332288175298,
+      "grad_norm": 0.3257639706134796,
+      "learning_rate": 4.912322844102811e-05,
+      "loss": 0.5203,
       "step": 4090
     },
     {
-      "epoch": 0.14,
-      "learning_rate": 4.801829268292683e-05,
-      "loss": 0.3638,
+      "epoch": 0.14758352254297763,
+      "grad_norm": 0.2311761975288391,
+      "learning_rate": 4.9183281287533034e-05,
+      "loss": 0.502,
       "step": 4095
     },
     {
-      "epoch": 0.14,
-      "learning_rate": 4.8076923076923084e-05,
-      "loss": 0.3528,
+      "epoch": 0.14776372220420225,
+      "grad_norm": 0.26127567887306213,
+      "learning_rate": 4.924333413403795e-05,
+      "loss": 0.4544,
       "step": 4100
     },
     {
-      "epoch": 0.14,
-      "learning_rate": 4.813555347091932e-05,
-      "loss": 0.3536,
+      "epoch": 0.1479439218654269,
+      "grad_norm": 0.25755074620246887,
+      "learning_rate": 4.9303386980542885e-05,
+      "loss": 0.4996,
       "step": 4105
     },
     {
-      "epoch": 0.14,
-      "learning_rate": 4.8194183864915575e-05,
-      "loss": 0.3846,
+      "epoch": 0.14812412152665153,
+      "grad_norm": 0.23005364835262299,
+      "learning_rate": 4.9363439827047804e-05,
+      "loss": 0.5183,
       "step": 4110
     },
     {
-      "epoch": 0.14,
-      "learning_rate": 4.825281425891182e-05,
-      "loss": 0.3407,
+      "epoch": 0.14830432118787618,
+      "grad_norm": 0.20392531156539917,
+      "learning_rate": 4.942349267355273e-05,
+      "loss": 0.4814,
       "step": 4115
     },
     {
-      "epoch": 0.14,
-      "learning_rate": 4.831144465290807e-05,
-      "loss": 0.3505,
+      "epoch": 0.1484845208491008,
+      "grad_norm": 0.2658933401107788,
+      "learning_rate": 4.9483545520057655e-05,
+      "loss": 0.5149,
       "step": 4120
     },
     {
-      "epoch": 0.15,
-      "learning_rate": 4.837007504690432e-05,
-      "loss": 0.3477,
+      "epoch": 0.14866472051032545,
+      "grad_norm": 0.19606181979179382,
+      "learning_rate": 4.9543598366562574e-05,
+      "loss": 0.4621,
       "step": 4125
     },
     {
-      "epoch": 0.15,
-      "learning_rate": 4.8428705440900564e-05,
-      "loss": 0.3686,
+      "epoch": 0.14884492017155007,
+      "grad_norm": 0.25544169545173645,
+      "learning_rate": 4.9603651213067506e-05,
+      "loss": 0.5054,
       "step": 4130
     },
     {
-      "epoch": 0.15,
-      "learning_rate": 4.8487335834896817e-05,
-      "loss": 0.3406,
+      "epoch": 0.14902511983277472,
+      "grad_norm": 0.25996389985084534,
+      "learning_rate": 4.9663704059572425e-05,
+      "loss": 0.4607,
       "step": 4135
     },
     {
-      "epoch": 0.15,
-      "learning_rate": 4.8545966228893055e-05,
-      "loss": 0.3638,
+      "epoch": 0.14920531949399934,
+      "grad_norm": 0.22277650237083435,
+      "learning_rate": 4.972375690607735e-05,
+      "loss": 0.4594,
       "step": 4140
     },
     {
-      "epoch": 0.15,
-      "learning_rate": 4.860459662288931e-05,
-      "loss": 0.3467,
+      "epoch": 0.149385519155224,
+      "grad_norm": 0.24612875282764435,
+      "learning_rate": 4.978380975258227e-05,
+      "loss": 0.4942,
       "step": 4145
     },
     {
-      "epoch": 0.15,
-      "learning_rate": 4.8663227016885553e-05,
-      "loss": 0.3367,
+      "epoch": 0.14956571881644862,
+      "grad_norm": 0.3200278878211975,
+      "learning_rate": 4.98438625990872e-05,
+      "loss": 0.497,
       "step": 4150
     },
     {
-      "epoch": 0.15,
-      "learning_rate": 4.8721857410881806e-05,
-      "loss": 0.3366,
+      "epoch": 0.14974591847767327,
+      "grad_norm": 0.2536075711250305,
+      "learning_rate": 4.990391544559213e-05,
+      "loss": 0.4744,
       "step": 4155
     },
     {
-      "epoch": 0.15,
-      "learning_rate": 4.878048780487805e-05,
-      "loss": 0.3646,
+      "epoch": 0.1499261181388979,
+      "grad_norm": 0.22233828902244568,
+      "learning_rate": 4.9963968292097046e-05,
+      "loss": 0.4684,
       "step": 4160
     },
     {
-      "epoch": 0.15,
-      "learning_rate": 4.8839118198874304e-05,
-      "loss": 0.3653,
+      "epoch": 0.15010631780012254,
+      "grad_norm": 0.2499028444290161,
+      "learning_rate": 4.999999997275039e-05,
+      "loss": 0.4484,
       "step": 4165
     },
     {
-      "epoch": 0.15,
-      "learning_rate": 4.889774859287054e-05,
-      "loss": 0.3176,
+      "epoch": 0.15028651746134716,
+      "grad_norm": 0.2228631228208542,
+      "learning_rate": 4.9999999666192246e-05,
+      "loss": 0.4891,
       "step": 4170
     },
     {
-      "epoch": 0.15,
-      "learning_rate": 4.8956378986866795e-05,
-      "loss": 0.3331,
+      "epoch": 0.1504667171225718,
+      "grad_norm": 0.25544434785842896,
+      "learning_rate": 4.9999999019013944e-05,
+      "loss": 0.4893,
       "step": 4175
     },
     {
-      "epoch": 0.15,
-      "learning_rate": 4.901500938086304e-05,
-      "loss": 0.3376,
+      "epoch": 0.15064691678379644,
+      "grad_norm": 0.22523342072963715,
+      "learning_rate": 4.99999980312155e-05,
+      "loss": 0.4708,
       "step": 4180
     },
     {
-      "epoch": 0.15,
-      "learning_rate": 4.9073639774859286e-05,
-      "loss": 0.3516,
+      "epoch": 0.1508271164450211,
+      "grad_norm": 0.2670489549636841,
+      "learning_rate": 4.999999670279692e-05,
+      "loss": 0.4553,
       "step": 4185
     },
     {
-      "epoch": 0.15,
-      "learning_rate": 4.913227016885554e-05,
-      "loss": 0.3702,
+      "epoch": 0.1510073161062457,
+      "grad_norm": 0.19969885051250458,
+      "learning_rate": 4.999999503375823e-05,
+      "loss": 0.4742,
       "step": 4190
     },
     {
-      "epoch": 0.15,
-      "learning_rate": 4.9190900562851784e-05,
-      "loss": 0.3589,
+      "epoch": 0.15118751576747036,
+      "grad_norm": 0.26136794686317444,
+      "learning_rate": 4.9999993024099446e-05,
+      "loss": 0.5093,
       "step": 4195
     },
     {
-      "epoch": 0.15,
-      "learning_rate": 4.924953095684804e-05,
-      "loss": 0.3326,
+      "epoch": 0.15136771542869498,
+      "grad_norm": 0.265652596950531,
+      "learning_rate": 4.999999067382059e-05,
+      "loss": 0.4894,
       "step": 4200
     },
     {
-      "epoch": 0.15,
-      "learning_rate": 4.9308161350844276e-05,
-      "loss": 0.3562,
+      "epoch": 0.15154791508991963,
+      "grad_norm": 0.22276762127876282,
+      "learning_rate": 4.999998798292171e-05,
+      "loss": 0.4643,
       "step": 4205
     },
     {
-      "epoch": 0.15,
-      "learning_rate": 4.936679174484053e-05,
-      "loss": 0.3464,
+      "epoch": 0.15172811475114426,
+      "grad_norm": 0.18586856126785278,
+      "learning_rate": 4.9999984951402834e-05,
+      "loss": 0.4719,
       "step": 4210
     },
     {
-      "epoch": 0.15,
-      "learning_rate": 4.9425422138836774e-05,
-      "loss": 0.3674,
+      "epoch": 0.1519083144123689,
+      "grad_norm": 0.2172873467206955,
+      "learning_rate": 4.9999981579263997e-05,
+      "loss": 0.4829,
       "step": 4215
     },
     {
-      "epoch": 0.15,
-      "learning_rate": 4.9484052532833026e-05,
-      "loss": 0.3546,
+      "epoch": 0.15208851407359353,
+      "grad_norm": 0.24789679050445557,
+      "learning_rate": 4.999997786650525e-05,
+      "loss": 0.4808,
       "step": 4220
     },
     {
-      "epoch": 0.15,
-      "learning_rate": 4.954268292682927e-05,
-      "loss": 0.3918,
+      "epoch": 0.15226871373481818,
+      "grad_norm": 0.2855437397956848,
+      "learning_rate": 4.9999973813126654e-05,
+      "loss": 0.4822,
       "step": 4225
     },
     {
-      "epoch": 0.15,
-      "learning_rate": 4.9601313320825524e-05,
-      "loss": 0.3447,
+      "epoch": 0.15244891339604283,
+      "grad_norm": 0.29781678318977356,
+      "learning_rate": 4.999996941912825e-05,
+      "loss": 0.5047,
       "step": 4230
     },
     {
-      "epoch": 0.15,
-      "learning_rate": 4.965994371482176e-05,
-      "loss": 0.3669,
+      "epoch": 0.15262911305726745,
+      "grad_norm": 0.25594308972358704,
+      "learning_rate": 4.9999964684510104e-05,
+      "loss": 0.5076,
       "step": 4235
     },
     {
-      "epoch": 0.15,
-      "learning_rate": 4.971857410881801e-05,
-      "loss": 0.3438,
+      "epoch": 0.1528093127184921,
+      "grad_norm": 0.21112053096294403,
+      "learning_rate": 4.999995960927228e-05,
+      "loss": 0.4764,
       "step": 4240
     },
     {
-      "epoch": 0.15,
-      "learning_rate": 4.977720450281426e-05,
-      "loss": 0.3425,
+      "epoch": 0.15298951237971672,
+      "grad_norm": 0.29285240173339844,
+      "learning_rate": 4.999995419341485e-05,
+      "loss": 0.4841,
       "step": 4245
     },
     {
-      "epoch": 0.15,
-      "learning_rate": 4.9835834896810507e-05,
-      "loss": 0.3144,
+      "epoch": 0.15316971204094137,
+      "grad_norm": 0.25095129013061523,
+      "learning_rate": 4.9999948436937873e-05,
+      "loss": 0.4723,
       "step": 4250
     },
     {
-      "epoch": 0.15,
-      "learning_rate": 4.989446529080676e-05,
-      "loss": 0.3529,
+      "epoch": 0.153349911702166,
+      "grad_norm": 0.2660590410232544,
+      "learning_rate": 4.999994233984145e-05,
+      "loss": 0.4527,
       "step": 4255
     },
     {
-      "epoch": 0.15,
-      "learning_rate": 4.9953095684803005e-05,
-      "loss": 0.3431,
+      "epoch": 0.15353011136339065,
+      "grad_norm": 0.19300931692123413,
+      "learning_rate": 4.999993590212564e-05,
+      "loss": 0.4346,
       "step": 4260
     },
     {
-      "epoch": 0.15,
-      "learning_rate": 4.999999999350783e-05,
-      "loss": 0.3392,
+      "epoch": 0.15371031102461527,
+      "grad_norm": 0.2573135197162628,
+      "learning_rate": 4.999992912379055e-05,
+      "loss": 0.4827,
       "step": 4265
     },
     {
-      "epoch": 0.15,
-      "learning_rate": 4.999999976628187e-05,
-      "loss": 0.3406,
+      "epoch": 0.15389051068583992,
+      "grad_norm": 0.2267155647277832,
+      "learning_rate": 4.999992200483626e-05,
+      "loss": 0.4743,
       "step": 4270
     },
     {
-      "epoch": 0.15,
-      "learning_rate": 4.9999999214447376e-05,
-      "loss": 0.3477,
+      "epoch": 0.15407071034706454,
+      "grad_norm": 0.22252587974071503,
+      "learning_rate": 4.999991454526289e-05,
+      "loss": 0.4899,
       "step": 4275
     },
     {
-      "epoch": 0.15,
-      "learning_rate": 4.9999998338004374e-05,
-      "loss": 0.3506,
+      "epoch": 0.1542509100082892,
+      "grad_norm": 0.24364939332008362,
+      "learning_rate": 4.9999906745070515e-05,
+      "loss": 0.4584,
       "step": 4280
     },
     {
-      "epoch": 0.15,
-      "learning_rate": 4.999999713695287e-05,
-      "loss": 0.3389,
+      "epoch": 0.15443110966951382,
+      "grad_norm": 0.2882271409034729,
+      "learning_rate": 4.999989860425924e-05,
+      "loss": 0.4897,
       "step": 4285
     },
     {
-      "epoch": 0.15,
-      "learning_rate": 4.999999561129288e-05,
-      "loss": 0.3461,
+      "epoch": 0.15461130933073847,
+      "grad_norm": 0.22930704057216644,
+      "learning_rate": 4.9999890122829205e-05,
+      "loss": 0.482,
       "step": 4290
     },
     {
-      "epoch": 0.15,
-      "learning_rate": 4.999999376102442e-05,
-      "loss": 0.3139,
+      "epoch": 0.1547915089919631,
+      "grad_norm": 0.22965483367443085,
+      "learning_rate": 4.9999881300780495e-05,
+      "loss": 0.4587,
       "step": 4295
     },
     {
-      "epoch": 0.15,
-      "learning_rate": 4.999999158614751e-05,
-      "loss": 0.3506,
+      "epoch": 0.15497170865318774,
+      "grad_norm": 0.2293003350496292,
+      "learning_rate": 4.999987213811325e-05,
+      "loss": 0.532,
       "step": 4300
     },
     {
-      "epoch": 0.15,
-      "learning_rate": 4.999998908666219e-05,
-      "loss": 0.3673,
+      "epoch": 0.15515190831441236,
+      "grad_norm": 0.3403257727622986,
+      "learning_rate": 4.999986263482758e-05,
+      "loss": 0.4722,
       "step": 4305
     },
     {
-      "epoch": 0.15,
-      "learning_rate": 4.9999986262568495e-05,
-      "loss": 0.3583,
+      "epoch": 0.155332107975637,
+      "grad_norm": 0.27816566824913025,
+      "learning_rate": 4.9999852790923626e-05,
+      "loss": 0.4603,
       "step": 4310
     },
     {
-      "epoch": 0.15,
-      "learning_rate": 4.9999983113866445e-05,
-      "loss": 0.341,
+      "epoch": 0.15551230763686164,
+      "grad_norm": 0.24829629063606262,
+      "learning_rate": 4.9999842606401516e-05,
+      "loss": 0.5108,
       "step": 4315
     },
     {
-      "epoch": 0.15,
-      "learning_rate": 4.9999979640556095e-05,
-      "loss": 0.3345,
+      "epoch": 0.15569250729808629,
+      "grad_norm": 0.1773914247751236,
+      "learning_rate": 4.999983208126139e-05,
+      "loss": 0.4665,
       "step": 4320
     },
     {
-      "epoch": 0.15,
-      "learning_rate": 4.9999975842637484e-05,
-      "loss": 0.3396,
+      "epoch": 0.1558727069593109,
+      "grad_norm": 0.22621670365333557,
+      "learning_rate": 4.9999821215503396e-05,
+      "loss": 0.4857,
       "step": 4325
     },
     {
-      "epoch": 0.15,
-      "learning_rate": 4.999997172011066e-05,
-      "loss": 0.3524,
+      "epoch": 0.15605290662053556,
+      "grad_norm": 0.232326477766037,
+      "learning_rate": 4.999981000912767e-05,
+      "loss": 0.4696,
       "step": 4330
     },
     {
-      "epoch": 0.15,
-      "learning_rate": 4.999996727297568e-05,
-      "loss": 0.3158,
+      "epoch": 0.15623310628176018,
+      "grad_norm": 0.30691394209861755,
+      "learning_rate": 4.999979846213438e-05,
+      "loss": 0.5046,
       "step": 4335
     },
     {
-      "epoch": 0.15,
-      "learning_rate": 4.99999625012326e-05,
-      "loss": 0.37,
+      "epoch": 0.15641330594298483,
+      "grad_norm": 0.2561699151992798,
+      "learning_rate": 4.9999786574523675e-05,
+      "loss": 0.4833,
       "step": 4340
     },
     {
-      "epoch": 0.15,
-      "learning_rate": 4.999995740488148e-05,
-      "loss": 0.3554,
+      "epoch": 0.15659350560420945,
+      "grad_norm": 0.33284783363342285,
+      "learning_rate": 4.9999774346295716e-05,
+      "loss": 0.5305,
       "step": 4345
     },
     {
-      "epoch": 0.15,
-      "learning_rate": 4.999995198392239e-05,
-      "loss": 0.3597,
+      "epoch": 0.1567737052654341,
+      "grad_norm": 0.18784891068935394,
+      "learning_rate": 4.999976177745067e-05,
+      "loss": 0.4514,
       "step": 4350
     },
     {
-      "epoch": 0.15,
-      "learning_rate": 4.9999946238355405e-05,
-      "loss": 0.3107,
+      "epoch": 0.15695390492665873,
+      "grad_norm": 0.23666001856327057,
+      "learning_rate": 4.999974886798872e-05,
+      "loss": 0.5089,
       "step": 4355
     },
     {
-      "epoch": 0.15,
-      "learning_rate": 4.999994016818059e-05,
-      "loss": 0.3513,
+      "epoch": 0.15713410458788338,
+      "grad_norm": 0.2393905371427536,
+      "learning_rate": 4.999973561791002e-05,
+      "loss": 0.508,
       "step": 4360
     },
     {
-      "epoch": 0.15,
-      "learning_rate": 4.999993377339802e-05,
-      "loss": 0.3312,
+      "epoch": 0.157314304249108,
+      "grad_norm": 0.2510823905467987,
+      "learning_rate": 4.999972202721477e-05,
+      "loss": 0.4712,
       "step": 4365
     },
     {
-      "epoch": 0.15,
-      "learning_rate": 4.99999270540078e-05,
-      "loss": 0.3627,
+      "epoch": 0.15749450391033265,
+      "grad_norm": 0.21815423667430878,
+      "learning_rate": 4.999970809590314e-05,
+      "loss": 0.4702,
       "step": 4370
     },
     {
-      "epoch": 0.15,
-      "learning_rate": 4.9999920010009995e-05,
-      "loss": 0.3232,
+      "epoch": 0.15767470357155727,
+      "grad_norm": 0.26132404804229736,
+      "learning_rate": 4.999969382397534e-05,
+      "loss": 0.4778,
       "step": 4375
     },
     {
-      "epoch": 0.15,
-      "learning_rate": 4.99999126414047e-05,
-      "loss": 0.3548,
+      "epoch": 0.15785490323278192,
+      "grad_norm": 0.21386702358722687,
+      "learning_rate": 4.999967921143154e-05,
+      "loss": 0.4859,
       "step": 4380
     },
     {
-      "epoch": 0.15,
-      "learning_rate": 4.9999904948192025e-05,
-      "loss": 0.3475,
+      "epoch": 0.15803510289400655,
+      "grad_norm": 0.3148122727870941,
+      "learning_rate": 4.999966425827195e-05,
+      "loss": 0.5291,
       "step": 4385
     },
     {
-      "epoch": 0.15,
-      "learning_rate": 4.999989693037205e-05,
-      "loss": 0.3552,
+      "epoch": 0.1582153025552312,
+      "grad_norm": 0.26083052158355713,
+      "learning_rate": 4.999964896449678e-05,
+      "loss": 0.4753,
       "step": 4390
     },
     {
-      "epoch": 0.15,
-      "learning_rate": 4.99998885879449e-05,
-      "loss": 0.3655,
+      "epoch": 0.15839550221645585,
+      "grad_norm": 0.24049343168735504,
+      "learning_rate": 4.9999633330106234e-05,
+      "loss": 0.5006,
       "step": 4395
     },
     {
-      "epoch": 0.15,
-      "learning_rate": 4.9999879920910664e-05,
-      "loss": 0.3671,
+      "epoch": 0.15857570187768047,
+      "grad_norm": 0.31498703360557556,
+      "learning_rate": 4.999961735510052e-05,
+      "loss": 0.4786,
       "step": 4400
     },
     {
-      "epoch": 0.15,
-      "learning_rate": 4.999987092926946e-05,
-      "loss": 0.3337,
+      "epoch": 0.15875590153890512,
+      "grad_norm": 0.29424259066581726,
+      "learning_rate": 4.999960103947986e-05,
+      "loss": 0.5437,
       "step": 4405
     },
     {
-      "epoch": 0.16,
-      "learning_rate": 4.9999861613021416e-05,
-      "loss": 0.3321,
+      "epoch": 0.15893610120012974,
+      "grad_norm": 0.28579822182655334,
+      "learning_rate": 4.999958438324448e-05,
+      "loss": 0.4814,
       "step": 4410
     },
     {
-      "epoch": 0.16,
-      "learning_rate": 4.999985197216663e-05,
-      "loss": 0.3543,
+      "epoch": 0.1591163008613544,
+      "grad_norm": 0.2311839759349823,
+      "learning_rate": 4.9999567386394595e-05,
+      "loss": 0.4655,
       "step": 4415
     },
     {
-      "epoch": 0.16,
-      "learning_rate": 4.999984200670526e-05,
-      "loss": 0.3368,
+      "epoch": 0.15929650052257902,
+      "grad_norm": 0.3129260241985321,
+      "learning_rate": 4.9999550048930455e-05,
+      "loss": 0.4614,
       "step": 4420
     },
     {
-      "epoch": 0.16,
-      "learning_rate": 4.99998317166374e-05,
-      "loss": 0.351,
+      "epoch": 0.15947670018380367,
+      "grad_norm": 0.2093297243118286,
+      "learning_rate": 4.999953237085228e-05,
+      "loss": 0.4925,
       "step": 4425
     },
     {
-      "epoch": 0.16,
-      "learning_rate": 4.999982110196322e-05,
-      "loss": 0.3441,
+      "epoch": 0.1596568998450283,
+      "grad_norm": 0.2211044728755951,
+      "learning_rate": 4.999951435216032e-05,
+      "loss": 0.4632,
       "step": 4430
     },
     {
-      "epoch": 0.16,
-      "learning_rate": 4.9999810162682824e-05,
-      "loss": 0.3481,
+      "epoch": 0.15983709950625294,
+      "grad_norm": 0.2064119130373001,
+      "learning_rate": 4.9999495992854806e-05,
+      "loss": 0.4794,
       "step": 4435
     },
     {
-      "epoch": 0.16,
-      "learning_rate": 4.999979889879637e-05,
-      "loss": 0.3649,
+      "epoch": 0.16001729916747756,
+      "grad_norm": 0.19835302233695984,
+      "learning_rate": 4.999947729293601e-05,
+      "loss": 0.4831,
       "step": 4440
     },
     {
-      "epoch": 0.16,
-      "learning_rate": 4.9999787310304e-05,
-      "loss": 0.3493,
+      "epoch": 0.1601974988287022,
+      "grad_norm": 0.2594761550426483,
+      "learning_rate": 4.9999458252404176e-05,
+      "loss": 0.4982,
       "step": 4445
     },
     {
-      "epoch": 0.16,
-      "learning_rate": 4.9999775397205875e-05,
-      "loss": 0.3405,
+      "epoch": 0.16037769848992683,
+      "grad_norm": 0.22020448744297028,
+      "learning_rate": 4.999943887125955e-05,
+      "loss": 0.4638,
       "step": 4450
     },
     {
-      "epoch": 0.16,
-      "learning_rate": 4.9999763159502146e-05,
-      "loss": 0.3482,
+      "epoch": 0.16055789815115148,
+      "grad_norm": 0.16716046631336212,
+      "learning_rate": 4.9999419149502426e-05,
+      "loss": 0.4716,
       "step": 4455
     },
     {
-      "epoch": 0.16,
-      "learning_rate": 4.999975059719296e-05,
-      "loss": 0.3288,
+      "epoch": 0.1607380978123761,
+      "grad_norm": 0.2861575782299042,
+      "learning_rate": 4.9999399087133044e-05,
+      "loss": 0.4905,
       "step": 4460
     },
     {
-      "epoch": 0.16,
-      "learning_rate": 4.9999737710278485e-05,
-      "loss": 0.3454,
+      "epoch": 0.16091829747360076,
+      "grad_norm": 0.22994907200336456,
+      "learning_rate": 4.99993786841517e-05,
+      "loss": 0.486,
       "step": 4465
     },
     {
-      "epoch": 0.16,
-      "learning_rate": 4.99997244987589e-05,
-      "loss": 0.3426,
+      "epoch": 0.16109849713482538,
+      "grad_norm": 0.17052839696407318,
+      "learning_rate": 4.9999357940558656e-05,
+      "loss": 0.4374,
       "step": 4470
     },
     {
-      "epoch": 0.16,
-      "learning_rate": 4.999971096263436e-05,
-      "loss": 0.3203,
+      "epoch": 0.16127869679605003,
+      "grad_norm": 0.20029401779174805,
+      "learning_rate": 4.99993368563542e-05,
+      "loss": 0.4667,
       "step": 4475
     },
     {
-      "epoch": 0.16,
-      "learning_rate": 4.999969710190505e-05,
-      "loss": 0.3542,
+      "epoch": 0.16145889645727465,
+      "grad_norm": 0.26389387249946594,
+      "learning_rate": 4.9999315431538616e-05,
+      "loss": 0.4532,
       "step": 4480
     },
     {
-      "epoch": 0.16,
-      "learning_rate": 4.999968291657115e-05,
-      "loss": 0.3207,
+      "epoch": 0.1616390961184993,
+      "grad_norm": 0.27325350046157837,
+      "learning_rate": 4.999929366611221e-05,
+      "loss": 0.4707,
       "step": 4485
     },
     {
-      "epoch": 0.16,
-      "learning_rate": 4.999966840663285e-05,
-      "loss": 0.336,
+      "epoch": 0.16181929577972393,
+      "grad_norm": 0.28567302227020264,
+      "learning_rate": 4.9999271560075256e-05,
+      "loss": 0.4888,
       "step": 4490
     },
     {
-      "epoch": 0.16,
-      "learning_rate": 4.999965357209032e-05,
-      "loss": 0.3624,
+      "epoch": 0.16199949544094858,
+      "grad_norm": 0.16461628675460815,
+      "learning_rate": 4.999924911342807e-05,
+      "loss": 0.4651,
       "step": 4495
     },
     {
-      "epoch": 0.16,
-      "learning_rate": 4.999963841294377e-05,
-      "loss": 0.3367,
+      "epoch": 0.1621796951021732,
+      "grad_norm": 0.1844627559185028,
+      "learning_rate": 4.999922632617096e-05,
+      "loss": 0.4724,
       "step": 4500
     },
     {
-      "epoch": 0.16,
-      "eval_loss": 0.3431379199028015,
-      "eval_runtime": 10.5416,
-      "eval_samples_per_second": 9.486,
-      "eval_steps_per_second": 9.486,
+      "epoch": 0.1621796951021732,
+      "eval_loss": 0.5034152865409851,
+      "eval_runtime": 3.5158,
+      "eval_samples_per_second": 28.443,
+      "eval_steps_per_second": 7.111,
       "step": 4500
     },
     {
-      "epoch": 0.16,
-      "learning_rate": 4.999962292919339e-05,
-      "loss": 0.3374,
+      "epoch": 0.16235989476339785,
+      "grad_norm": 0.21949437260627747,
+      "learning_rate": 4.999920319830423e-05,
+      "loss": 0.4663,
       "step": 4505
     },
     {
-      "epoch": 0.16,
-      "learning_rate": 4.9999607120839374e-05,
-      "loss": 0.345,
+      "epoch": 0.16254009442462247,
+      "grad_norm": 0.23475132882595062,
+      "learning_rate": 4.9999179729828195e-05,
+      "loss": 0.4849,
       "step": 4510
     },
     {
-      "epoch": 0.16,
-      "learning_rate": 4.9999590987881944e-05,
-      "loss": 0.3261,
+      "epoch": 0.16272029408584712,
+      "grad_norm": 0.20308546721935272,
+      "learning_rate": 4.999915592074318e-05,
+      "loss": 0.4665,
       "step": 4515
     },
     {
-      "epoch": 0.16,
-      "learning_rate": 4.99995745303213e-05,
-      "loss": 0.303,
+      "epoch": 0.16290049374707175,
+      "grad_norm": 0.18522381782531738,
+      "learning_rate": 4.9999131771049496e-05,
+      "loss": 0.4843,
       "step": 4520
     },
     {
-      "epoch": 0.16,
-      "learning_rate": 4.999955774815766e-05,
-      "loss": 0.3219,
+      "epoch": 0.1630806934082964,
+      "grad_norm": 0.2702092230319977,
+      "learning_rate": 4.9999107280747484e-05,
+      "loss": 0.5005,
       "step": 4525
     },
     {
-      "epoch": 0.16,
-      "learning_rate": 4.999954064139122e-05,
-      "loss": 0.3604,
+      "epoch": 0.16326089306952102,
+      "grad_norm": 0.20005741715431213,
+      "learning_rate": 4.999908244983748e-05,
+      "loss": 0.4915,
       "step": 4530
     },
     {
-      "epoch": 0.16,
-      "learning_rate": 4.999952321002224e-05,
-      "loss": 0.3528,
+      "epoch": 0.16344109273074567,
+      "grad_norm": 0.2156675010919571,
+      "learning_rate": 4.9999057278319817e-05,
+      "loss": 0.4701,
       "step": 4535
     },
     {
-      "epoch": 0.16,
-      "learning_rate": 4.9999505454050915e-05,
-      "loss": 0.3385,
+      "epoch": 0.1636212923919703,
+      "grad_norm": 0.2057947814464569,
+      "learning_rate": 4.999903176619484e-05,
+      "loss": 0.4371,
       "step": 4540
     },
     {
-      "epoch": 0.16,
-      "learning_rate": 4.9999487373477486e-05,
-      "loss": 0.344,
+      "epoch": 0.16380149205319494,
+      "grad_norm": 0.20522654056549072,
+      "learning_rate": 4.9999005913462896e-05,
+      "loss": 0.474,
       "step": 4545
     },
     {
-      "epoch": 0.16,
-      "learning_rate": 4.99994689683022e-05,
-      "loss": 0.3397,
+      "epoch": 0.16398169171441956,
+      "grad_norm": 0.20008191466331482,
+      "learning_rate": 4.999897972012433e-05,
+      "loss": 0.4826,
       "step": 4550
     },
     {
-      "epoch": 0.16,
-      "learning_rate": 4.999945023852527e-05,
-      "loss": 0.3112,
+      "epoch": 0.16416189137564421,
+      "grad_norm": 0.2456914186477661,
+      "learning_rate": 4.999895318617951e-05,
+      "loss": 0.47,
       "step": 4555
     },
     {
-      "epoch": 0.16,
-      "learning_rate": 4.999943118414696e-05,
-      "loss": 0.361,
+      "epoch": 0.16434209103686884,
+      "grad_norm": 0.21113859117031097,
+      "learning_rate": 4.999892631162879e-05,
+      "loss": 0.4505,
       "step": 4560
     },
     {
-      "epoch": 0.16,
-      "learning_rate": 4.9999411805167515e-05,
-      "loss": 0.3396,
+      "epoch": 0.1645222906980935,
+      "grad_norm": 0.2364521026611328,
+      "learning_rate": 4.999889909647254e-05,
+      "loss": 0.4626,
       "step": 4565
     },
     {
-      "epoch": 0.16,
-      "learning_rate": 4.999939210158718e-05,
-      "loss": 0.3423,
+      "epoch": 0.16470249035931814,
+      "grad_norm": 0.29241201281547546,
+      "learning_rate": 4.9998871540711126e-05,
+      "loss": 0.5229,
       "step": 4570
     },
     {
-      "epoch": 0.16,
-      "learning_rate": 4.999937207340622e-05,
-      "loss": 0.3222,
+      "epoch": 0.16488269002054276,
+      "grad_norm": 0.22053126990795135,
+      "learning_rate": 4.9998843644344926e-05,
+      "loss": 0.5066,
       "step": 4575
     },
     {
-      "epoch": 0.16,
-      "learning_rate": 4.999935172062488e-05,
-      "loss": 0.2994,
+      "epoch": 0.1650628896817674,
+      "grad_norm": 0.2373570203781128,
+      "learning_rate": 4.999881540737433e-05,
+      "loss": 0.4642,
       "step": 4580
     },
     {
-      "epoch": 0.16,
-      "learning_rate": 4.9999331043243436e-05,
-      "loss": 0.3463,
+      "epoch": 0.16524308934299203,
+      "grad_norm": 0.24440720677375793,
+      "learning_rate": 4.999878682979972e-05,
+      "loss": 0.4755,
       "step": 4585
     },
     {
-      "epoch": 0.16,
-      "learning_rate": 4.9999310041262154e-05,
-      "loss": 0.3634,
+      "epoch": 0.16542328900421668,
+      "grad_norm": 0.29946649074554443,
+      "learning_rate": 4.999875791162146e-05,
+      "loss": 0.4903,
       "step": 4590
     },
     {
-      "epoch": 0.16,
-      "learning_rate": 4.9999288714681306e-05,
-      "loss": 0.3839,
+      "epoch": 0.1656034886654413,
+      "grad_norm": 0.2605338394641876,
+      "learning_rate": 4.9998728652839974e-05,
+      "loss": 0.478,
       "step": 4595
     },
     {
-      "epoch": 0.16,
-      "learning_rate": 4.999926706350117e-05,
-      "loss": 0.345,
+      "epoch": 0.16578368832666596,
+      "grad_norm": 0.20642036199569702,
+      "learning_rate": 4.999869905345565e-05,
+      "loss": 0.4739,
       "step": 4600
     },
     {
-      "epoch": 0.16,
-      "learning_rate": 4.999924508772202e-05,
-      "loss": 0.3378,
+      "epoch": 0.16596388798789058,
+      "grad_norm": 0.2667411267757416,
+      "learning_rate": 4.999866911346889e-05,
+      "loss": 0.4881,
       "step": 4605
     },
     {
-      "epoch": 0.16,
-      "learning_rate": 4.9999222787344156e-05,
-      "loss": 0.3724,
+      "epoch": 0.16614408764911523,
+      "grad_norm": 0.20082050561904907,
+      "learning_rate": 4.999863883288011e-05,
+      "loss": 0.4852,
       "step": 4610
     },
     {
-      "epoch": 0.16,
-      "learning_rate": 4.999920016236786e-05,
-      "loss": 0.3219,
+      "epoch": 0.16632428731033985,
+      "grad_norm": 0.27822422981262207,
+      "learning_rate": 4.99986082116897e-05,
+      "loss": 0.4941,
       "step": 4615
     },
     {
-      "epoch": 0.16,
-      "learning_rate": 4.9999177212793424e-05,
-      "loss": 0.353,
+      "epoch": 0.1665044869715645,
+      "grad_norm": 0.28651654720306396,
+      "learning_rate": 4.999857724989811e-05,
+      "loss": 0.5,
       "step": 4620
     },
     {
-      "epoch": 0.16,
-      "learning_rate": 4.999915393862115e-05,
-      "loss": 0.3162,
+      "epoch": 0.16668468663278913,
+      "grad_norm": 0.2600594162940979,
+      "learning_rate": 4.9998545947505734e-05,
+      "loss": 0.48,
       "step": 4625
     },
     {
-      "epoch": 0.16,
-      "learning_rate": 4.9999130339851325e-05,
-      "loss": 0.3491,
+      "epoch": 0.16686488629401378,
+      "grad_norm": 0.2715243697166443,
+      "learning_rate": 4.999851430451301e-05,
+      "loss": 0.472,
       "step": 4630
     },
     {
-      "epoch": 0.16,
-      "learning_rate": 4.9999106416484274e-05,
-      "loss": 0.3276,
+      "epoch": 0.1670450859552384,
+      "grad_norm": 0.259139746427536,
+      "learning_rate": 4.999848232092037e-05,
+      "loss": 0.4986,
       "step": 4635
     },
     {
-      "epoch": 0.16,
-      "learning_rate": 4.99990821685203e-05,
-      "loss": 0.3685,
+      "epoch": 0.16722528561646305,
+      "grad_norm": 0.24041247367858887,
+      "learning_rate": 4.999844999672825e-05,
+      "loss": 0.4546,
       "step": 4640
     },
     {
-      "epoch": 0.16,
-      "learning_rate": 4.999905759595972e-05,
-      "loss": 0.3249,
+      "epoch": 0.16740548527768767,
+      "grad_norm": 0.2729252278804779,
+      "learning_rate": 4.9998417331937086e-05,
+      "loss": 0.4713,
       "step": 4645
     },
     {
-      "epoch": 0.16,
-      "learning_rate": 4.9999032698802853e-05,
-      "loss": 0.3665,
+      "epoch": 0.16758568493891232,
+      "grad_norm": 0.25324806571006775,
+      "learning_rate": 4.999838432654733e-05,
+      "loss": 0.4919,
       "step": 4650
     },
     {
-      "epoch": 0.16,
-      "learning_rate": 4.999900747705002e-05,
-      "loss": 0.3448,
+      "epoch": 0.16776588460013694,
+      "grad_norm": 0.25361695885658264,
+      "learning_rate": 4.9998350980559427e-05,
+      "loss": 0.4811,
       "step": 4655
     },
     {
-      "epoch": 0.16,
-      "learning_rate": 4.9998981930701556e-05,
-      "loss": 0.3512,
+      "epoch": 0.1679460842613616,
+      "grad_norm": 0.24803782999515533,
+      "learning_rate": 4.999831729397383e-05,
+      "loss": 0.4998,
       "step": 4660
     },
     {
-      "epoch": 0.16,
-      "learning_rate": 4.999895605975778e-05,
-      "loss": 0.3621,
+      "epoch": 0.16812628392258622,
+      "grad_norm": 0.28923875093460083,
+      "learning_rate": 4.9998283266791e-05,
+      "loss": 0.4713,
       "step": 4665
     },
     {
-      "epoch": 0.16,
-      "learning_rate": 4.9998929864219035e-05,
-      "loss": 0.3259,
+      "epoch": 0.16830648358381087,
+      "grad_norm": 0.17523469030857086,
+      "learning_rate": 4.9998248899011405e-05,
+      "loss": 0.4995,
       "step": 4670
     },
     {
-      "epoch": 0.16,
-      "learning_rate": 4.999890334408566e-05,
-      "loss": 0.3641,
+      "epoch": 0.1684866832450355,
+      "grad_norm": 0.21582381427288055,
+      "learning_rate": 4.9998214190635495e-05,
+      "loss": 0.4943,
       "step": 4675
     },
     {
-      "epoch": 0.16,
-      "learning_rate": 4.9998876499358e-05,
-      "loss": 0.3532,
+      "epoch": 0.16866688290626014,
+      "grad_norm": 0.35510435700416565,
+      "learning_rate": 4.999817914166377e-05,
+      "loss": 0.5125,
       "step": 4680
     },
     {
-      "epoch": 0.16,
-      "learning_rate": 4.9998849330036414e-05,
-      "loss": 0.3338,
+      "epoch": 0.16884708256748476,
+      "grad_norm": 0.255281001329422,
+      "learning_rate": 4.9998143752096684e-05,
+      "loss": 0.5037,
       "step": 4685
     },
     {
-      "epoch": 0.17,
-      "learning_rate": 4.999882183612123e-05,
-      "loss": 0.3298,
+      "epoch": 0.1690272822287094,
+      "grad_norm": 0.26469287276268005,
+      "learning_rate": 4.999810802193473e-05,
+      "loss": 0.4958,
       "step": 4690
     },
     {
-      "epoch": 0.17,
-      "learning_rate": 4.999879401761283e-05,
-      "loss": 0.3281,
+      "epoch": 0.16920748188993404,
+      "grad_norm": 0.23432576656341553,
+      "learning_rate": 4.9998071951178405e-05,
+      "loss": 0.4646,
       "step": 4695
     },
     {
-      "epoch": 0.17,
-      "learning_rate": 4.9998765874511554e-05,
-      "loss": 0.3606,
+      "epoch": 0.1693876815511587,
+      "grad_norm": 0.19142989814281464,
+      "learning_rate": 4.999803553982818e-05,
+      "loss": 0.4931,
       "step": 4700
     },
     {
-      "epoch": 0.17,
-      "learning_rate": 4.999873740681779e-05,
-      "loss": 0.3785,
+      "epoch": 0.1695678812123833,
+      "grad_norm": 0.21128664910793304,
+      "learning_rate": 4.999799878788456e-05,
+      "loss": 0.4901,
       "step": 4705
     },
     {
-      "epoch": 0.17,
-      "learning_rate": 4.9998708614531885e-05,
-      "loss": 0.3164,
+      "epoch": 0.16974808087360796,
+      "grad_norm": 0.22904910147190094,
+      "learning_rate": 4.999796169534805e-05,
+      "loss": 0.4975,
       "step": 4710
     },
     {
-      "epoch": 0.17,
-      "learning_rate": 4.999867949765422e-05,
-      "loss": 0.334,
+      "epoch": 0.16992828053483258,
+      "grad_norm": 0.27078938484191895,
+      "learning_rate": 4.999792426221915e-05,
+      "loss": 0.4434,
       "step": 4715
     },
     {
-      "epoch": 0.17,
-      "learning_rate": 4.999865005618519e-05,
-      "loss": 0.3884,
+      "epoch": 0.17010848019605723,
+      "grad_norm": 0.2444290965795517,
+      "learning_rate": 4.9997886488498374e-05,
+      "loss": 0.4923,
       "step": 4720
     },
     {
-      "epoch": 0.17,
-      "learning_rate": 4.999862029012516e-05,
-      "loss": 0.3537,
+      "epoch": 0.17028867985728185,
+      "grad_norm": 0.2548987865447998,
+      "learning_rate": 4.999784837418623e-05,
+      "loss": 0.4474,
       "step": 4725
     },
     {
-      "epoch": 0.17,
-      "learning_rate": 4.9998590199474516e-05,
-      "loss": 0.2878,
+      "epoch": 0.1704688795185065,
+      "grad_norm": 0.2576419711112976,
+      "learning_rate": 4.999780991928325e-05,
+      "loss": 0.5217,
       "step": 4730
     },
     {
-      "epoch": 0.17,
-      "learning_rate": 4.999855978423365e-05,
-      "loss": 0.3349,
+      "epoch": 0.17064907917973116,
+      "grad_norm": 0.2640141546726227,
+      "learning_rate": 4.999777112378994e-05,
+      "loss": 0.4983,
       "step": 4735
     },
     {
-      "epoch": 0.17,
-      "learning_rate": 4.999852904440296e-05,
-      "loss": 0.3547,
+      "epoch": 0.17082927884095578,
+      "grad_norm": 0.2303144633769989,
+      "learning_rate": 4.999773198770684e-05,
+      "loss": 0.477,
       "step": 4740
     },
     {
-      "epoch": 0.17,
-      "learning_rate": 4.999849797998286e-05,
-      "loss": 0.3466,
+      "epoch": 0.17100947850218043,
+      "grad_norm": 0.22392189502716064,
+      "learning_rate": 4.999769251103449e-05,
+      "loss": 0.4617,
       "step": 4745
     },
     {
-      "epoch": 0.17,
-      "learning_rate": 4.999846659097373e-05,
-      "loss": 0.3519,
+      "epoch": 0.17118967816340505,
+      "grad_norm": 0.26157307624816895,
+      "learning_rate": 4.9997652693773414e-05,
+      "loss": 0.4668,
       "step": 4750
     },
     {
-      "epoch": 0.17,
-      "learning_rate": 4.999843487737598e-05,
-      "loss": 0.3389,
+      "epoch": 0.1713698778246297,
+      "grad_norm": 0.24970059096813202,
+      "learning_rate": 4.999761253592415e-05,
+      "loss": 0.4901,
       "step": 4755
     },
     {
-      "epoch": 0.17,
-      "learning_rate": 4.999840283919003e-05,
-      "loss": 0.3458,
+      "epoch": 0.17155007748585432,
+      "grad_norm": 0.25207483768463135,
+      "learning_rate": 4.999757203748727e-05,
+      "loss": 0.4811,
       "step": 4760
     },
     {
-      "epoch": 0.17,
-      "learning_rate": 4.999837047641629e-05,
-      "loss": 0.3598,
+      "epoch": 0.17173027714707897,
+      "grad_norm": 0.23375752568244934,
+      "learning_rate": 4.99975311984633e-05,
+      "loss": 0.4795,
       "step": 4765
     },
     {
-      "epoch": 0.17,
-      "learning_rate": 4.9998337789055195e-05,
-      "loss": 0.3349,
+      "epoch": 0.1719104768083036,
+      "grad_norm": 0.2202087789773941,
+      "learning_rate": 4.999749001885281e-05,
+      "loss": 0.5194,
       "step": 4770
     },
     {
-      "epoch": 0.17,
-      "learning_rate": 4.9998304777107155e-05,
-      "loss": 0.3621,
+      "epoch": 0.17209067646952825,
+      "grad_norm": 0.17962399125099182,
+      "learning_rate": 4.999744849865636e-05,
+      "loss": 0.4766,
       "step": 4775
     },
     {
-      "epoch": 0.17,
-      "learning_rate": 4.99982714405726e-05,
-      "loss": 0.3359,
+      "epoch": 0.17227087613075287,
+      "grad_norm": 0.18169254064559937,
+      "learning_rate": 4.999740663787452e-05,
+      "loss": 0.5046,
       "step": 4780
     },
     {
-      "epoch": 0.17,
-      "learning_rate": 4.999823777945196e-05,
-      "loss": 0.3329,
+      "epoch": 0.17245107579197752,
+      "grad_norm": 0.21856611967086792,
+      "learning_rate": 4.999736443650784e-05,
+      "loss": 0.4918,
       "step": 4785
     },
     {
-      "epoch": 0.17,
-      "learning_rate": 4.999820379374568e-05,
-      "loss": 0.3453,
+      "epoch": 0.17263127545320214,
+      "grad_norm": 0.24348682165145874,
+      "learning_rate": 4.999732189455692e-05,
+      "loss": 0.4751,
       "step": 4790
     },
     {
-      "epoch": 0.17,
-      "learning_rate": 4.99981694834542e-05,
-      "loss": 0.3564,
+      "epoch": 0.1728114751144268,
+      "grad_norm": 0.22899436950683594,
+      "learning_rate": 4.9997279012022324e-05,
+      "loss": 0.4198,
       "step": 4795
     },
     {
-      "epoch": 0.17,
-      "learning_rate": 4.999813484857796e-05,
-      "loss": 0.3436,
+      "epoch": 0.17299167477565142,
+      "grad_norm": 0.26833006739616394,
+      "learning_rate": 4.9997235788904646e-05,
+      "loss": 0.4782,
       "step": 4800
     },
     {
-      "epoch": 0.17,
-      "learning_rate": 4.9998099889117416e-05,
-      "loss": 0.35,
+      "epoch": 0.17317187443687607,
+      "grad_norm": 0.2631605565547943,
+      "learning_rate": 4.9997192225204466e-05,
+      "loss": 0.4366,
       "step": 4805
     },
     {
-      "epoch": 0.17,
-      "learning_rate": 4.9998064605073025e-05,
-      "loss": 0.336,
+      "epoch": 0.1733520740981007,
+      "grad_norm": 0.22278451919555664,
+      "learning_rate": 4.999714832092238e-05,
+      "loss": 0.4783,
       "step": 4810
     },
     {
-      "epoch": 0.17,
-      "learning_rate": 4.999802899644524e-05,
-      "loss": 0.3788,
+      "epoch": 0.17353227375932534,
+      "grad_norm": 0.22488220036029816,
+      "learning_rate": 4.999710407605899e-05,
+      "loss": 0.4497,
       "step": 4815
     },
     {
-      "epoch": 0.17,
-      "learning_rate": 4.9997993063234505e-05,
-      "loss": 0.3505,
+      "epoch": 0.17371247342054996,
+      "grad_norm": 0.19599469006061554,
+      "learning_rate": 4.99970594906149e-05,
+      "loss": 0.5147,
       "step": 4820
     },
     {
-      "epoch": 0.17,
-      "learning_rate": 4.9997956805441316e-05,
-      "loss": 0.36,
+      "epoch": 0.1738926730817746,
+      "grad_norm": 0.22859926521778107,
+      "learning_rate": 4.9997014564590706e-05,
+      "loss": 0.4891,
       "step": 4825
     },
     {
-      "epoch": 0.17,
-      "learning_rate": 4.999792022306613e-05,
-      "loss": 0.3545,
+      "epoch": 0.17407287274299924,
+      "grad_norm": 0.29516199231147766,
+      "learning_rate": 4.9996969297987036e-05,
+      "loss": 0.454,
       "step": 4830
     },
     {
-      "epoch": 0.17,
-      "learning_rate": 4.999788331610943e-05,
-      "loss": 0.3646,
+      "epoch": 0.17425307240422389,
+      "grad_norm": 0.2656870484352112,
+      "learning_rate": 4.9996923690804495e-05,
+      "loss": 0.4491,
       "step": 4835
     },
     {
-      "epoch": 0.17,
-      "learning_rate": 4.999784608457169e-05,
-      "loss": 0.3414,
+      "epoch": 0.1744332720654485,
+      "grad_norm": 0.2088516801595688,
+      "learning_rate": 4.999687774304371e-05,
+      "loss": 0.489,
       "step": 4840
     },
     {
-      "epoch": 0.17,
-      "learning_rate": 4.9997808528453386e-05,
-      "loss": 0.3378,
+      "epoch": 0.17461347172667316,
+      "grad_norm": 0.20423473417758942,
+      "learning_rate": 4.99968314547053e-05,
+      "loss": 0.5235,
       "step": 4845
     },
     {
-      "epoch": 0.17,
-      "learning_rate": 4.999777064775502e-05,
-      "loss": 0.3516,
+      "epoch": 0.17479367138789778,
+      "grad_norm": 0.2307443767786026,
+      "learning_rate": 4.999678482578991e-05,
+      "loss": 0.4527,
       "step": 4850
     },
     {
-      "epoch": 0.17,
-      "learning_rate": 4.999773244247706e-05,
-      "loss": 0.343,
+      "epoch": 0.17497387104912243,
+      "grad_norm": 0.16717872023582458,
+      "learning_rate": 4.9996737856298156e-05,
+      "loss": 0.4769,
       "step": 4855
     },
     {
-      "epoch": 0.17,
-      "learning_rate": 4.9997693912620035e-05,
-      "loss": 0.3429,
+      "epoch": 0.17515407071034705,
+      "grad_norm": 0.23562130331993103,
+      "learning_rate": 4.99966905462307e-05,
+      "loss": 0.4787,
       "step": 4860
     },
     {
-      "epoch": 0.17,
-      "learning_rate": 4.999765505818442e-05,
-      "loss": 0.3423,
+      "epoch": 0.1753342703715717,
+      "grad_norm": 0.2515668272972107,
+      "learning_rate": 4.9996642895588166e-05,
+      "loss": 0.4721,
       "step": 4865
     },
     {
-      "epoch": 0.17,
-      "learning_rate": 4.999761587917073e-05,
-      "loss": 0.321,
+      "epoch": 0.17551447003279633,
+      "grad_norm": 0.20563237369060516,
+      "learning_rate": 4.9996594904371215e-05,
+      "loss": 0.481,
       "step": 4870
     },
     {
-      "epoch": 0.17,
-      "learning_rate": 4.999757637557948e-05,
-      "loss": 0.3474,
+      "epoch": 0.17569466969402098,
+      "grad_norm": 0.24153757095336914,
+      "learning_rate": 4.99965465725805e-05,
+      "loss": 0.49,
       "step": 4875
     },
     {
-      "epoch": 0.17,
-      "learning_rate": 4.9997536547411165e-05,
-      "loss": 0.3708,
+      "epoch": 0.1758748693552456,
+      "grad_norm": 0.19717536866664886,
+      "learning_rate": 4.999649790021667e-05,
+      "loss": 0.4529,
       "step": 4880
     },
     {
-      "epoch": 0.17,
-      "learning_rate": 4.999749639466632e-05,
-      "loss": 0.3307,
+      "epoch": 0.17605506901647025,
+      "grad_norm": 0.26905906200408936,
+      "learning_rate": 4.99964488872804e-05,
+      "loss": 0.4775,
       "step": 4885
     },
     {
-      "epoch": 0.17,
-      "learning_rate": 4.999745591734545e-05,
-      "loss": 0.3124,
+      "epoch": 0.17623526867769487,
+      "grad_norm": 0.18942050635814667,
+      "learning_rate": 4.999639953377235e-05,
+      "loss": 0.4558,
       "step": 4890
     },
     {
-      "epoch": 0.17,
-      "learning_rate": 4.999741511544909e-05,
-      "loss": 0.3249,
+      "epoch": 0.17641546833891952,
+      "grad_norm": 0.25289055705070496,
+      "learning_rate": 4.99963498396932e-05,
+      "loss": 0.4752,
       "step": 4895
     },
     {
-      "epoch": 0.17,
-      "learning_rate": 4.9997373988977776e-05,
-      "loss": 0.321,
+      "epoch": 0.17659566800014415,
+      "grad_norm": 0.28434181213378906,
+      "learning_rate": 4.9996299805043625e-05,
+      "loss": 0.498,
       "step": 4900
     },
     {
-      "epoch": 0.17,
-      "learning_rate": 4.999733253793203e-05,
-      "loss": 0.3264,
+      "epoch": 0.1767758676613688,
+      "grad_norm": 0.2828446328639984,
+      "learning_rate": 4.99962494298243e-05,
+      "loss": 0.498,
       "step": 4905
     },
     {
-      "epoch": 0.17,
-      "learning_rate": 4.999729076231239e-05,
-      "loss": 0.3436,
+      "epoch": 0.17695606732259345,
+      "grad_norm": 0.24182403087615967,
+      "learning_rate": 4.999619871403592e-05,
+      "loss": 0.4959,
       "step": 4910
     },
     {
-      "epoch": 0.17,
-      "learning_rate": 4.999724866211941e-05,
-      "loss": 0.3249,
+      "epoch": 0.17713626698381807,
+      "grad_norm": 0.2429310828447342,
+      "learning_rate": 4.999614765767917e-05,
+      "loss": 0.4828,
       "step": 4915
     },
     {
-      "epoch": 0.17,
-      "learning_rate": 4.999720623735363e-05,
-      "loss": 0.3331,
+      "epoch": 0.17731646664504272,
+      "grad_norm": 0.2273981124162674,
+      "learning_rate": 4.999609626075475e-05,
+      "loss": 0.4961,
       "step": 4920
     },
     {
-      "epoch": 0.17,
-      "learning_rate": 4.9997163488015595e-05,
-      "loss": 0.3406,
+      "epoch": 0.17749666630626734,
+      "grad_norm": 0.24011553823947906,
+      "learning_rate": 4.999604452326335e-05,
+      "loss": 0.4782,
       "step": 4925
     },
     {
-      "epoch": 0.17,
-      "learning_rate": 4.999712041410587e-05,
-      "loss": 0.3397,
+      "epoch": 0.177676865967492,
+      "grad_norm": 0.22934234142303467,
+      "learning_rate": 4.999599244520569e-05,
+      "loss": 0.5014,
       "step": 4930
     },
     {
-      "epoch": 0.17,
-      "learning_rate": 4.9997077015625007e-05,
-      "loss": 0.3426,
+      "epoch": 0.17785706562871662,
+      "grad_norm": 0.19686634838581085,
+      "learning_rate": 4.999594002658247e-05,
+      "loss": 0.446,
       "step": 4935
     },
     {
-      "epoch": 0.17,
-      "learning_rate": 4.9997033292573575e-05,
-      "loss": 0.3314,
+      "epoch": 0.17803726528994127,
+      "grad_norm": 0.2651787996292114,
+      "learning_rate": 4.999588726739441e-05,
+      "loss": 0.5142,
       "step": 4940
     },
     {
-      "epoch": 0.17,
-      "learning_rate": 4.999698924495214e-05,
-      "loss": 0.3547,
+      "epoch": 0.1782174649511659,
+      "grad_norm": 0.1980464607477188,
+      "learning_rate": 4.999583416764222e-05,
+      "loss": 0.4975,
       "step": 4945
     },
     {
-      "epoch": 0.17,
-      "learning_rate": 4.999694487276128e-05,
-      "loss": 0.3362,
+      "epoch": 0.17839766461239054,
+      "grad_norm": 0.24422787129878998,
+      "learning_rate": 4.999578072732664e-05,
+      "loss": 0.455,
       "step": 4950
     },
     {
-      "epoch": 0.17,
-      "learning_rate": 4.999690017600155e-05,
-      "loss": 0.3154,
+      "epoch": 0.17857786427361516,
+      "grad_norm": 0.31826359033584595,
+      "learning_rate": 4.9995726946448375e-05,
+      "loss": 0.5055,
       "step": 4955
     },
     {
-      "epoch": 0.17,
-      "learning_rate": 4.999685515467355e-05,
-      "loss": 0.3549,
+      "epoch": 0.1787580639348398,
+      "grad_norm": 0.26964086294174194,
+      "learning_rate": 4.999567282500817e-05,
+      "loss": 0.4871,
       "step": 4960
     },
     {
-      "epoch": 0.17,
-      "learning_rate": 4.999680980877787e-05,
-      "loss": 0.3302,
+      "epoch": 0.17893826359606443,
+      "grad_norm": 0.18171383440494537,
+      "learning_rate": 4.999561836300676e-05,
+      "loss": 0.4757,
       "step": 4965
     },
     {
-      "epoch": 0.17,
-      "learning_rate": 4.999676413831507e-05,
-      "loss": 0.3415,
+      "epoch": 0.17911846325728908,
+      "grad_norm": 0.26201415061950684,
+      "learning_rate": 4.99955635604449e-05,
+      "loss": 0.4891,
       "step": 4970
     },
     {
-      "epoch": 0.18,
-      "learning_rate": 4.999671814328577e-05,
-      "loss": 0.3449,
+      "epoch": 0.1792986629185137,
+      "grad_norm": 0.24988138675689697,
+      "learning_rate": 4.999550841732331e-05,
+      "loss": 0.4845,
       "step": 4975
     },
     {
-      "epoch": 0.18,
-      "learning_rate": 4.9996671823690565e-05,
-      "loss": 0.3202,
+      "epoch": 0.17947886257973836,
+      "grad_norm": 0.21902771294116974,
+      "learning_rate": 4.999545293364277e-05,
+      "loss": 0.4645,
       "step": 4980
     },
     {
-      "epoch": 0.18,
-      "learning_rate": 4.9996625179530046e-05,
-      "loss": 0.3373,
+      "epoch": 0.17965906224096298,
+      "grad_norm": 0.1887025535106659,
+      "learning_rate": 4.999539710940402e-05,
+      "loss": 0.4777,
       "step": 4985
     },
     {
-      "epoch": 0.18,
-      "learning_rate": 4.999657821080481e-05,
-      "loss": 0.3344,
+      "epoch": 0.17983926190218763,
+      "grad_norm": 0.21480071544647217,
+      "learning_rate": 4.9995340944607824e-05,
+      "loss": 0.499,
       "step": 4990
     },
     {
-      "epoch": 0.18,
-      "learning_rate": 4.999653091751549e-05,
-      "loss": 0.3466,
+      "epoch": 0.18001946156341225,
+      "grad_norm": 0.2067321240901947,
+      "learning_rate": 4.9995284439254944e-05,
+      "loss": 0.4782,
       "step": 4995
     },
     {
-      "epoch": 0.18,
-      "learning_rate": 4.9996483299662694e-05,
-      "loss": 0.3277,
+      "epoch": 0.1801996612246369,
+      "grad_norm": 0.22930487990379333,
+      "learning_rate": 4.999522759334616e-05,
+      "loss": 0.5074,
       "step": 5000
     },
     {
-      "epoch": 0.18,
-      "eval_loss": 0.33953267335891724,
-      "eval_runtime": 10.5352,
-      "eval_samples_per_second": 9.492,
-      "eval_steps_per_second": 9.492,
+      "epoch": 0.1801996612246369,
+      "eval_loss": 0.5006341338157654,
+      "eval_runtime": 3.5194,
+      "eval_samples_per_second": 28.414,
+      "eval_steps_per_second": 7.104,
       "step": 5000
     },
     {
-      "epoch": 0.18,
-      "learning_rate": 4.999643535724703e-05,
-      "loss": 0.345,
+      "epoch": 0.18037986088586153,
+      "grad_norm": 0.23178088665008545,
+      "learning_rate": 4.999517040688223e-05,
+      "loss": 0.4822,
       "step": 5005
     },
     {
-      "epoch": 0.18,
-      "learning_rate": 4.999638709026913e-05,
-      "loss": 0.3589,
+      "epoch": 0.18056006054708618,
+      "grad_norm": 0.21988582611083984,
+      "learning_rate": 4.999511287986395e-05,
+      "loss": 0.4559,
       "step": 5010
     },
     {
-      "epoch": 0.18,
-      "learning_rate": 4.999633849872961e-05,
-      "loss": 0.3331,
+      "epoch": 0.1807402602083108,
+      "grad_norm": 0.25729215145111084,
+      "learning_rate": 4.999505501229209e-05,
+      "loss": 0.4693,
       "step": 5015
     },
     {
-      "epoch": 0.18,
-      "learning_rate": 4.999628958262911e-05,
-      "loss": 0.3244,
+      "epoch": 0.18092045986953545,
+      "grad_norm": 0.23540961742401123,
+      "learning_rate": 4.999499680416745e-05,
+      "loss": 0.5181,
       "step": 5020
     },
     {
-      "epoch": 0.18,
-      "learning_rate": 4.9996240341968265e-05,
-      "loss": 0.3285,
+      "epoch": 0.18110065953076007,
+      "grad_norm": 0.17820635437965393,
+      "learning_rate": 4.9994938255490814e-05,
+      "loss": 0.4853,
       "step": 5025
     },
     {
-      "epoch": 0.18,
-      "learning_rate": 4.999619077674771e-05,
-      "loss": 0.3572,
+      "epoch": 0.18128085919198472,
+      "grad_norm": 0.2419961541891098,
+      "learning_rate": 4.999487936626299e-05,
+      "loss": 0.4927,
       "step": 5030
     },
     {
-      "epoch": 0.18,
-      "learning_rate": 4.999614088696809e-05,
-      "loss": 0.3537,
+      "epoch": 0.18146105885320934,
+      "grad_norm": 0.2524636685848236,
+      "learning_rate": 4.999482013648476e-05,
+      "loss": 0.4986,
       "step": 5035
     },
     {
-      "epoch": 0.18,
-      "learning_rate": 4.9996090672630056e-05,
-      "loss": 0.3523,
+      "epoch": 0.181641258514434,
+      "grad_norm": 0.18317490816116333,
+      "learning_rate": 4.999476056615696e-05,
+      "loss": 0.4806,
       "step": 5040
     },
     {
-      "epoch": 0.18,
-      "learning_rate": 4.999604013373426e-05,
-      "loss": 0.3402,
+      "epoch": 0.18182145817565862,
+      "grad_norm": 0.2592169940471649,
+      "learning_rate": 4.999470065528038e-05,
+      "loss": 0.4741,
       "step": 5045
     },
     {
-      "epoch": 0.18,
-      "learning_rate": 4.999598927028135e-05,
-      "loss": 0.3613,
+      "epoch": 0.18200165783688327,
+      "grad_norm": 0.24432708323001862,
+      "learning_rate": 4.999464040385584e-05,
+      "loss": 0.4726,
       "step": 5050
     },
     {
-      "epoch": 0.18,
-      "learning_rate": 4.9995938082271996e-05,
-      "loss": 0.356,
+      "epoch": 0.1821818574981079,
+      "grad_norm": 0.23248982429504395,
+      "learning_rate": 4.999457981188417e-05,
+      "loss": 0.4759,
       "step": 5055
     },
     {
-      "epoch": 0.18,
-      "learning_rate": 4.999588656970686e-05,
-      "loss": 0.3402,
+      "epoch": 0.18236205715933254,
+      "grad_norm": 0.23487761616706848,
+      "learning_rate": 4.999451887936618e-05,
+      "loss": 0.4953,
       "step": 5060
     },
     {
-      "epoch": 0.18,
-      "learning_rate": 4.999583473258661e-05,
-      "loss": 0.3801,
+      "epoch": 0.18254225682055716,
+      "grad_norm": 0.28242379426956177,
+      "learning_rate": 4.9994457606302714e-05,
+      "loss": 0.4772,
       "step": 5065
     },
     {
-      "epoch": 0.18,
-      "learning_rate": 4.999578257091192e-05,
-      "loss": 0.341,
+      "epoch": 0.18272245648178181,
+      "grad_norm": 0.1956668496131897,
+      "learning_rate": 4.999439599269461e-05,
+      "loss": 0.5024,
       "step": 5070
     },
     {
-      "epoch": 0.18,
-      "learning_rate": 4.999573008468347e-05,
-      "loss": 0.3114,
+      "epoch": 0.18290265614300646,
+      "grad_norm": 0.25298696756362915,
+      "learning_rate": 4.999433403854269e-05,
+      "loss": 0.5046,
       "step": 5075
     },
     {
-      "epoch": 0.18,
-      "learning_rate": 4.9995677273901924e-05,
-      "loss": 0.356,
+      "epoch": 0.1830828558042311,
+      "grad_norm": 0.2178865671157837,
+      "learning_rate": 4.999427174384781e-05,
+      "loss": 0.5063,
       "step": 5080
     },
     {
-      "epoch": 0.18,
-      "learning_rate": 4.999562413856799e-05,
-      "loss": 0.3366,
+      "epoch": 0.18326305546545574,
+      "grad_norm": 0.2657603621482849,
+      "learning_rate": 4.9994209108610814e-05,
+      "loss": 0.4846,
       "step": 5085
     },
     {
-      "epoch": 0.18,
-      "learning_rate": 4.999557067868235e-05,
-      "loss": 0.3245,
+      "epoch": 0.18344325512668036,
+      "grad_norm": 0.23558183014392853,
+      "learning_rate": 4.999414613283256e-05,
+      "loss": 0.4673,
       "step": 5090
     },
     {
-      "epoch": 0.18,
-      "learning_rate": 4.999551689424569e-05,
-      "loss": 0.3187,
+      "epoch": 0.183623454787905,
+      "grad_norm": 0.23601466417312622,
+      "learning_rate": 4.999408281651391e-05,
+      "loss": 0.5041,
       "step": 5095
     },
     {
-      "epoch": 0.18,
-      "learning_rate": 4.999546278525872e-05,
-      "loss": 0.3288,
+      "epoch": 0.18380365444912963,
+      "grad_norm": 0.22025693953037262,
+      "learning_rate": 4.999401915965572e-05,
+      "loss": 0.4699,
       "step": 5100
     },
     {
-      "epoch": 0.18,
-      "learning_rate": 4.9995408351722145e-05,
-      "loss": 0.35,
+      "epoch": 0.18398385411035428,
+      "grad_norm": 0.23485559225082397,
+      "learning_rate": 4.9993955162258844e-05,
+      "loss": 0.4927,
       "step": 5105
     },
     {
-      "epoch": 0.18,
-      "learning_rate": 4.999535359363666e-05,
-      "loss": 0.3453,
+      "epoch": 0.1841640537715789,
+      "grad_norm": 0.2333470582962036,
+      "learning_rate": 4.999389082432417e-05,
+      "loss": 0.472,
       "step": 5110
     },
     {
-      "epoch": 0.18,
-      "learning_rate": 4.999529851100298e-05,
-      "loss": 0.3534,
+      "epoch": 0.18434425343280356,
+      "grad_norm": 0.20295509696006775,
+      "learning_rate": 4.999382614585258e-05,
+      "loss": 0.4978,
       "step": 5115
     },
     {
-      "epoch": 0.18,
-      "learning_rate": 4.999524310382182e-05,
-      "loss": 0.3667,
+      "epoch": 0.18452445309402818,
+      "grad_norm": 0.23132069408893585,
+      "learning_rate": 4.999376112684495e-05,
+      "loss": 0.4796,
       "step": 5120
     },
     {
-      "epoch": 0.18,
-      "learning_rate": 4.9995187372093906e-05,
-      "loss": 0.3592,
+      "epoch": 0.18470465275525283,
+      "grad_norm": 0.19237832725048065,
+      "learning_rate": 4.9993695767302165e-05,
+      "loss": 0.5032,
       "step": 5125
     },
     {
-      "epoch": 0.18,
-      "learning_rate": 4.999513131581995e-05,
-      "loss": 0.3338,
+      "epoch": 0.18488485241647745,
+      "grad_norm": 0.19899766147136688,
+      "learning_rate": 4.999363006722511e-05,
+      "loss": 0.4785,
       "step": 5130
     },
     {
-      "epoch": 0.18,
-      "learning_rate": 4.9995074935000694e-05,
-      "loss": 0.3393,
+      "epoch": 0.1850650520777021,
+      "grad_norm": 0.22194600105285645,
+      "learning_rate": 4.999356402661469e-05,
+      "loss": 0.5207,
       "step": 5135
     },
     {
-      "epoch": 0.18,
-      "learning_rate": 4.999501822963686e-05,
-      "loss": 0.3725,
+      "epoch": 0.18524525173892673,
+      "grad_norm": 0.22985856235027313,
+      "learning_rate": 4.999349764547179e-05,
+      "loss": 0.4558,
       "step": 5140
     },
     {
-      "epoch": 0.18,
-      "learning_rate": 4.999496119972918e-05,
-      "loss": 0.3396,
+      "epoch": 0.18542545140015138,
+      "grad_norm": 0.2950509488582611,
+      "learning_rate": 4.9993430923797324e-05,
+      "loss": 0.4956,
       "step": 5145
     },
     {
-      "epoch": 0.18,
-      "learning_rate": 4.999490384527841e-05,
-      "loss": 0.3536,
+      "epoch": 0.185605651061376,
+      "grad_norm": 0.2260800004005432,
+      "learning_rate": 4.9993363861592204e-05,
+      "loss": 0.435,
       "step": 5150
     },
     {
-      "epoch": 0.18,
-      "learning_rate": 4.9994846166285284e-05,
-      "loss": 0.3628,
+      "epoch": 0.18578585072260065,
+      "grad_norm": 0.18863672018051147,
+      "learning_rate": 4.999329645885734e-05,
+      "loss": 0.5241,
       "step": 5155
     },
     {
-      "epoch": 0.18,
-      "learning_rate": 4.999478816275054e-05,
-      "loss": 0.343,
+      "epoch": 0.18596605038382527,
+      "grad_norm": 0.24623405933380127,
+      "learning_rate": 4.999322871559365e-05,
+      "loss": 0.4892,
       "step": 5160
     },
     {
-      "epoch": 0.18,
-      "learning_rate": 4.999472983467496e-05,
-      "loss": 0.3212,
+      "epoch": 0.18614625004504992,
+      "grad_norm": 0.2146414965391159,
+      "learning_rate": 4.999316063180206e-05,
+      "loss": 0.4762,
       "step": 5165
     },
     {
-      "epoch": 0.18,
-      "learning_rate": 4.9994671182059274e-05,
-      "loss": 0.3293,
+      "epoch": 0.18632644970627454,
+      "grad_norm": 0.24258659780025482,
+      "learning_rate": 4.999309220748349e-05,
+      "loss": 0.4825,
       "step": 5170
     },
     {
-      "epoch": 0.18,
-      "learning_rate": 4.999461220490427e-05,
-      "loss": 0.3619,
+      "epoch": 0.1865066493674992,
+      "grad_norm": 0.24239155650138855,
+      "learning_rate": 4.999302344263889e-05,
+      "loss": 0.4796,
       "step": 5175
     },
     {
-      "epoch": 0.18,
-      "learning_rate": 4.999455290321068e-05,
-      "loss": 0.3588,
+      "epoch": 0.18668684902872382,
+      "grad_norm": 0.25734949111938477,
+      "learning_rate": 4.999295433726917e-05,
+      "loss": 0.5216,
       "step": 5180
     },
     {
-      "epoch": 0.18,
-      "learning_rate": 4.9994493276979304e-05,
-      "loss": 0.3385,
+      "epoch": 0.18686704868994847,
+      "grad_norm": 0.2938891649246216,
+      "learning_rate": 4.999288489137529e-05,
+      "loss": 0.4744,
       "step": 5185
     },
     {
-      "epoch": 0.18,
-      "learning_rate": 4.99944333262109e-05,
-      "loss": 0.3556,
+      "epoch": 0.1870472483511731,
+      "grad_norm": 0.20838968455791473,
+      "learning_rate": 4.999281510495819e-05,
+      "loss": 0.4786,
       "step": 5190
     },
     {
-      "epoch": 0.18,
-      "learning_rate": 4.999437305090625e-05,
-      "loss": 0.3072,
+      "epoch": 0.18722744801239774,
+      "grad_norm": 0.25164559483528137,
+      "learning_rate": 4.999274497801883e-05,
+      "loss": 0.523,
       "step": 5195
     },
     {
-      "epoch": 0.18,
-      "learning_rate": 4.999431245106615e-05,
-      "loss": 0.3087,
+      "epoch": 0.18740764767362236,
+      "grad_norm": 0.23008379340171814,
+      "learning_rate": 4.999267451055815e-05,
+      "loss": 0.4591,
       "step": 5200
     },
     {
-      "epoch": 0.18,
-      "learning_rate": 4.999425152669136e-05,
-      "loss": 0.3562,
+      "epoch": 0.187587847334847,
+      "grad_norm": 0.21487128734588623,
+      "learning_rate": 4.9992603702577124e-05,
+      "loss": 0.4903,
       "step": 5205
     },
     {
-      "epoch": 0.18,
-      "learning_rate": 4.9994190277782696e-05,
-      "loss": 0.3484,
+      "epoch": 0.18776804699607164,
+      "grad_norm": 0.2193828970193863,
+      "learning_rate": 4.999253255407671e-05,
+      "loss": 0.4357,
       "step": 5210
     },
     {
-      "epoch": 0.18,
-      "learning_rate": 4.9994128704340936e-05,
-      "loss": 0.3419,
+      "epoch": 0.1879482466572963,
+      "grad_norm": 0.20415742695331573,
+      "learning_rate": 4.999246106505788e-05,
+      "loss": 0.4428,
       "step": 5215
     },
     {
-      "epoch": 0.18,
-      "learning_rate": 4.999406680636689e-05,
-      "loss": 0.3368,
+      "epoch": 0.1881284463185209,
+      "grad_norm": 0.21222014725208282,
+      "learning_rate": 4.9992389235521606e-05,
+      "loss": 0.4969,
       "step": 5220
     },
     {
-      "epoch": 0.18,
-      "learning_rate": 4.9994004583861356e-05,
-      "loss": 0.3485,
+      "epoch": 0.18830864597974556,
+      "grad_norm": 0.2923332750797272,
+      "learning_rate": 4.999231706546886e-05,
+      "loss": 0.5121,
       "step": 5225
     },
     {
-      "epoch": 0.18,
-      "learning_rate": 4.999394203682515e-05,
-      "loss": 0.3424,
+      "epoch": 0.18848884564097018,
+      "grad_norm": 0.26699545979499817,
+      "learning_rate": 4.999224455490065e-05,
+      "loss": 0.4731,
       "step": 5230
     },
     {
-      "epoch": 0.18,
-      "learning_rate": 4.9993879165259074e-05,
-      "loss": 0.3392,
+      "epoch": 0.18866904530219483,
+      "grad_norm": 0.27180203795433044,
+      "learning_rate": 4.999217170381794e-05,
+      "loss": 0.4727,
       "step": 5235
     },
     {
-      "epoch": 0.18,
-      "learning_rate": 4.999381596916395e-05,
-      "loss": 0.3408,
+      "epoch": 0.18884924496341948,
+      "grad_norm": 0.23332376778125763,
+      "learning_rate": 4.999209851222172e-05,
+      "loss": 0.4673,
       "step": 5240
     },
     {
-      "epoch": 0.18,
-      "learning_rate": 4.9993752448540596e-05,
-      "loss": 0.3023,
+      "epoch": 0.1890294446246441,
+      "grad_norm": 0.19692248106002808,
+      "learning_rate": 4.9992024980113e-05,
+      "loss": 0.4728,
       "step": 5245
     },
     {
-      "epoch": 0.18,
-      "learning_rate": 4.999368860338983e-05,
-      "loss": 0.3318,
+      "epoch": 0.18920964428586876,
+      "grad_norm": 0.22370965778827667,
+      "learning_rate": 4.999195110749278e-05,
+      "loss": 0.4951,
       "step": 5250
     },
     {
-      "epoch": 0.18,
-      "learning_rate": 4.99936244337125e-05,
-      "loss": 0.3265,
+      "epoch": 0.18938984394709338,
+      "grad_norm": 0.20484954118728638,
+      "learning_rate": 4.9991876894362064e-05,
+      "loss": 0.4878,
       "step": 5255
     },
     {
-      "epoch": 0.19,
-      "learning_rate": 4.9993559939509426e-05,
-      "loss": 0.3451,
+      "epoch": 0.18957004360831803,
+      "grad_norm": 0.18314266204833984,
+      "learning_rate": 4.9991802340721866e-05,
+      "loss": 0.5089,
       "step": 5260
     },
     {
-      "epoch": 0.19,
-      "learning_rate": 4.9993495120781445e-05,
-      "loss": 0.3405,
+      "epoch": 0.18975024326954265,
+      "grad_norm": 0.1618998497724533,
+      "learning_rate": 4.99917274465732e-05,
+      "loss": 0.5125,
       "step": 5265
     },
     {
-      "epoch": 0.19,
-      "learning_rate": 4.9993429977529405e-05,
-      "loss": 0.3369,
+      "epoch": 0.1899304429307673,
+      "grad_norm": 0.18222755193710327,
+      "learning_rate": 4.9991652211917086e-05,
+      "loss": 0.5117,
       "step": 5270
     },
     {
-      "epoch": 0.19,
-      "learning_rate": 4.999336450975414e-05,
-      "loss": 0.3678,
+      "epoch": 0.19011064259199192,
+      "grad_norm": 0.25865301489830017,
+      "learning_rate": 4.999157663675454e-05,
+      "loss": 0.4643,
       "step": 5275
     },
     {
-      "epoch": 0.19,
-      "learning_rate": 4.999329871745652e-05,
-      "loss": 0.3559,
+      "epoch": 0.19029084225321657,
+      "grad_norm": 0.21296486258506775,
+      "learning_rate": 4.999150072108661e-05,
+      "loss": 0.4639,
       "step": 5280
     },
     {
-      "epoch": 0.19,
-      "learning_rate": 4.9993232600637374e-05,
-      "loss": 0.3529,
+      "epoch": 0.1904710419144412,
+      "grad_norm": 0.19290651381015778,
+      "learning_rate": 4.9991424464914324e-05,
+      "loss": 0.4825,
       "step": 5285
     },
     {
-      "epoch": 0.19,
-      "learning_rate": 4.999316615929759e-05,
-      "loss": 0.3264,
+      "epoch": 0.19065124157566585,
+      "grad_norm": 0.28650933504104614,
+      "learning_rate": 4.9991347868238714e-05,
+      "loss": 0.5179,
       "step": 5290
     },
     {
-      "epoch": 0.19,
-      "learning_rate": 4.9993099393438006e-05,
-      "loss": 0.3242,
+      "epoch": 0.19083144123689047,
+      "grad_norm": 0.1798352748155594,
+      "learning_rate": 4.9991270931060826e-05,
+      "loss": 0.4681,
       "step": 5295
     },
     {
-      "epoch": 0.19,
-      "learning_rate": 4.9993032303059486e-05,
-      "loss": 0.326,
+      "epoch": 0.19101164089811512,
+      "grad_norm": 0.238149955868721,
+      "learning_rate": 4.9991193653381704e-05,
+      "loss": 0.4844,
       "step": 5300
     },
     {
-      "epoch": 0.19,
-      "learning_rate": 4.999296488816293e-05,
-      "loss": 0.3212,
+      "epoch": 0.19119184055933974,
+      "grad_norm": 0.27041110396385193,
+      "learning_rate": 4.9991116035202425e-05,
+      "loss": 0.5273,
       "step": 5305
     },
     {
-      "epoch": 0.19,
-      "learning_rate": 4.999289714874918e-05,
-      "loss": 0.341,
+      "epoch": 0.1913720402205644,
+      "grad_norm": 0.2148786038160324,
+      "learning_rate": 4.999103807652401e-05,
+      "loss": 0.4877,
       "step": 5310
     },
     {
-      "epoch": 0.19,
-      "learning_rate": 4.999282908481915e-05,
-      "loss": 0.3389,
+      "epoch": 0.19155223988178902,
+      "grad_norm": 0.20174206793308258,
+      "learning_rate": 4.999095977734755e-05,
+      "loss": 0.4858,
       "step": 5315
     },
     {
-      "epoch": 0.19,
-      "learning_rate": 4.9992760696373695e-05,
-      "loss": 0.3637,
+      "epoch": 0.19173243954301367,
+      "grad_norm": 0.26444557309150696,
+      "learning_rate": 4.9990881137674103e-05,
+      "loss": 0.4992,
       "step": 5320
     },
     {
-      "epoch": 0.19,
-      "learning_rate": 4.999269198341372e-05,
-      "loss": 0.3309,
+      "epoch": 0.1919126392042383,
+      "grad_norm": 0.21838323771953583,
+      "learning_rate": 4.9990802157504734e-05,
+      "loss": 0.4098,
       "step": 5325
     },
     {
-      "epoch": 0.19,
-      "learning_rate": 4.99926229459401e-05,
-      "loss": 0.3434,
+      "epoch": 0.19209283886546294,
+      "grad_norm": 0.23885351419448853,
+      "learning_rate": 4.999072283684052e-05,
+      "loss": 0.4674,
       "step": 5330
     },
     {
-      "epoch": 0.19,
-      "learning_rate": 4.999255358395374e-05,
-      "loss": 0.3455,
+      "epoch": 0.19227303852668756,
+      "grad_norm": 0.24196289479732513,
+      "learning_rate": 4.9990643175682554e-05,
+      "loss": 0.5008,
       "step": 5335
     },
     {
-      "epoch": 0.19,
-      "learning_rate": 4.999248389745556e-05,
-      "loss": 0.342,
+      "epoch": 0.1924532381879122,
+      "grad_norm": 0.18864691257476807,
+      "learning_rate": 4.999056317403191e-05,
+      "loss": 0.4603,
       "step": 5340
     },
     {
-      "epoch": 0.19,
-      "learning_rate": 4.999241388644643e-05,
-      "loss": 0.337,
+      "epoch": 0.19263343784913683,
+      "grad_norm": 0.2206612378358841,
+      "learning_rate": 4.9990482831889685e-05,
+      "loss": 0.4875,
       "step": 5345
     },
     {
-      "epoch": 0.19,
-      "learning_rate": 4.999234355092728e-05,
-      "loss": 0.3357,
+      "epoch": 0.19281363751036149,
+      "grad_norm": 0.18795616924762726,
+      "learning_rate": 4.9990402149256964e-05,
+      "loss": 0.4718,
       "step": 5350
     },
     {
-      "epoch": 0.19,
-      "learning_rate": 4.9992272890899024e-05,
-      "loss": 0.3281,
+      "epoch": 0.1929938371715861,
+      "grad_norm": 0.28613436222076416,
+      "learning_rate": 4.999032112613485e-05,
+      "loss": 0.4476,
       "step": 5355
     },
     {
-      "epoch": 0.19,
-      "learning_rate": 4.999220190636257e-05,
-      "loss": 0.3208,
+      "epoch": 0.19317403683281076,
+      "grad_norm": 0.22328707575798035,
+      "learning_rate": 4.999023976252445e-05,
+      "loss": 0.4631,
       "step": 5360
     },
     {
-      "epoch": 0.19,
-      "learning_rate": 4.999213059731885e-05,
-      "loss": 0.3598,
+      "epoch": 0.19335423649403538,
+      "grad_norm": 0.20430655777454376,
+      "learning_rate": 4.9990158058426875e-05,
+      "loss": 0.523,
       "step": 5365
     },
     {
-      "epoch": 0.19,
-      "learning_rate": 4.999205896376878e-05,
-      "loss": 0.3522,
+      "epoch": 0.19353443615526003,
+      "grad_norm": 0.19645529985427856,
+      "learning_rate": 4.999007601384324e-05,
+      "loss": 0.4338,
       "step": 5370
     },
     {
-      "epoch": 0.19,
-      "learning_rate": 4.99919870057133e-05,
-      "loss": 0.3178,
+      "epoch": 0.19371463581648465,
+      "grad_norm": 0.25969836115837097,
+      "learning_rate": 4.998999362877464e-05,
+      "loss": 0.4712,
       "step": 5375
     },
     {
-      "epoch": 0.19,
-      "learning_rate": 4.999191472315333e-05,
-      "loss": 0.3168,
+      "epoch": 0.1938948354777093,
+      "grad_norm": 0.19297480583190918,
+      "learning_rate": 4.9989910903222234e-05,
+      "loss": 0.471,
       "step": 5380
     },
     {
-      "epoch": 0.19,
-      "learning_rate": 4.999184211608983e-05,
-      "loss": 0.3462,
+      "epoch": 0.19407503513893393,
+      "grad_norm": 0.24897582828998566,
+      "learning_rate": 4.998982783718712e-05,
+      "loss": 0.4915,
       "step": 5385
     },
     {
-      "epoch": 0.19,
-      "learning_rate": 4.999176918452372e-05,
-      "loss": 0.351,
+      "epoch": 0.19425523480015858,
+      "grad_norm": 0.2619524896144867,
+      "learning_rate": 4.998974443067044e-05,
+      "loss": 0.5119,
       "step": 5390
     },
     {
-      "epoch": 0.19,
-      "learning_rate": 4.999169592845596e-05,
-      "loss": 0.3342,
+      "epoch": 0.1944354344613832,
+      "grad_norm": 0.17749615013599396,
+      "learning_rate": 4.998966068367334e-05,
+      "loss": 0.4782,
       "step": 5395
     },
     {
-      "epoch": 0.19,
-      "learning_rate": 4.99916223478875e-05,
-      "loss": 0.3408,
+      "epoch": 0.19461563412260785,
+      "grad_norm": 0.20441953837871552,
+      "learning_rate": 4.998957659619694e-05,
+      "loss": 0.4346,
       "step": 5400
     },
     {
-      "epoch": 0.19,
-      "learning_rate": 4.99915484428193e-05,
-      "loss": 0.3531,
+      "epoch": 0.19479583378383247,
+      "grad_norm": 0.22351542115211487,
+      "learning_rate": 4.9989492168242405e-05,
+      "loss": 0.4833,
       "step": 5405
     },
     {
-      "epoch": 0.19,
-      "learning_rate": 4.999147421325231e-05,
-      "loss": 0.359,
+      "epoch": 0.19497603344505712,
+      "grad_norm": 0.19759854674339294,
+      "learning_rate": 4.998940739981087e-05,
+      "loss": 0.5104,
       "step": 5410
     },
     {
-      "epoch": 0.19,
-      "learning_rate": 4.999139965918749e-05,
-      "loss": 0.3325,
+      "epoch": 0.19515623310628177,
+      "grad_norm": 0.2504650354385376,
+      "learning_rate": 4.9989322290903504e-05,
+      "loss": 0.4585,
       "step": 5415
     },
     {
-      "epoch": 0.19,
-      "learning_rate": 4.999132478062582e-05,
-      "loss": 0.3474,
+      "epoch": 0.1953364327675064,
+      "grad_norm": 0.2255609631538391,
+      "learning_rate": 4.998923684152145e-05,
+      "loss": 0.4881,
       "step": 5420
     },
     {
-      "epoch": 0.19,
-      "learning_rate": 4.9991249577568264e-05,
-      "loss": 0.3529,
+      "epoch": 0.19551663242873105,
+      "grad_norm": 0.25729531049728394,
+      "learning_rate": 4.99891510516659e-05,
+      "loss": 0.4705,
       "step": 5425
     },
     {
-      "epoch": 0.19,
-      "learning_rate": 4.999117405001581e-05,
-      "loss": 0.3142,
+      "epoch": 0.19569683208995567,
+      "grad_norm": 0.24127353727817535,
+      "learning_rate": 4.998906492133798e-05,
+      "loss": 0.4939,
       "step": 5430
     },
     {
-      "epoch": 0.19,
-      "learning_rate": 4.9991098197969426e-05,
-      "loss": 0.3436,
+      "epoch": 0.19587703175118032,
+      "grad_norm": 0.2850863039493561,
+      "learning_rate": 4.99889784505389e-05,
+      "loss": 0.4461,
       "step": 5435
     },
     {
-      "epoch": 0.19,
-      "learning_rate": 4.99910220214301e-05,
-      "loss": 0.3514,
+      "epoch": 0.19605723141240494,
+      "grad_norm": 0.1722259372472763,
+      "learning_rate": 4.998889163926983e-05,
+      "loss": 0.4819,
       "step": 5440
     },
     {
-      "epoch": 0.19,
-      "learning_rate": 4.999094552039883e-05,
-      "loss": 0.331,
+      "epoch": 0.1962374310736296,
+      "grad_norm": 0.21394529938697815,
+      "learning_rate": 4.998880448753195e-05,
+      "loss": 0.4985,
       "step": 5445
     },
     {
-      "epoch": 0.19,
-      "learning_rate": 4.999086869487659e-05,
-      "loss": 0.3635,
+      "epoch": 0.19641763073485421,
+      "grad_norm": 0.18665440380573273,
+      "learning_rate": 4.998871699532644e-05,
+      "loss": 0.5014,
       "step": 5450
     },
     {
-      "epoch": 0.19,
-      "learning_rate": 4.9990791544864404e-05,
-      "loss": 0.3251,
+      "epoch": 0.19659783039607887,
+      "grad_norm": 0.226554274559021,
+      "learning_rate": 4.998862916265451e-05,
+      "loss": 0.4596,
       "step": 5455
     },
     {
-      "epoch": 0.19,
-      "learning_rate": 4.999071407036325e-05,
-      "loss": 0.3262,
+      "epoch": 0.1967780300573035,
+      "grad_norm": 0.2326064258813858,
+      "learning_rate": 4.998854098951734e-05,
+      "loss": 0.4634,
       "step": 5460
     },
     {
-      "epoch": 0.19,
-      "learning_rate": 4.9990636271374144e-05,
-      "loss": 0.3029,
+      "epoch": 0.19695822971852814,
+      "grad_norm": 0.20283064246177673,
+      "learning_rate": 4.9988452475916134e-05,
+      "loss": 0.4553,
       "step": 5465
     },
     {
-      "epoch": 0.19,
-      "learning_rate": 4.99905581478981e-05,
-      "loss": 0.3293,
+      "epoch": 0.19713842937975276,
+      "grad_norm": 0.19399242103099823,
+      "learning_rate": 4.998836362185211e-05,
+      "loss": 0.472,
       "step": 5470
     },
     {
-      "epoch": 0.19,
-      "learning_rate": 4.999047969993612e-05,
-      "loss": 0.3535,
+      "epoch": 0.1973186290409774,
+      "grad_norm": 0.24312461912631989,
+      "learning_rate": 4.998827442732646e-05,
+      "loss": 0.4887,
       "step": 5475
     },
     {
-      "epoch": 0.19,
-      "learning_rate": 4.9990400927489234e-05,
-      "loss": 0.3354,
+      "epoch": 0.19749882870220203,
+      "grad_norm": 0.23328492045402527,
+      "learning_rate": 4.9988184892340414e-05,
+      "loss": 0.4683,
       "step": 5480
     },
     {
-      "epoch": 0.19,
-      "learning_rate": 4.999032183055846e-05,
-      "loss": 0.3349,
+      "epoch": 0.19767902836342668,
+      "grad_norm": 0.1696726530790329,
+      "learning_rate": 4.998809501689518e-05,
+      "loss": 0.4872,
       "step": 5485
     },
     {
-      "epoch": 0.19,
-      "learning_rate": 4.9990242409144835e-05,
-      "loss": 0.3415,
+      "epoch": 0.1978592280246513,
+      "grad_norm": 0.20958632230758667,
+      "learning_rate": 4.998800480099199e-05,
+      "loss": 0.4747,
       "step": 5490
     },
     {
-      "epoch": 0.19,
-      "learning_rate": 4.9990162663249373e-05,
-      "loss": 0.3273,
+      "epoch": 0.19803942768587596,
+      "grad_norm": 0.23482240736484528,
+      "learning_rate": 4.998791424463208e-05,
+      "loss": 0.4519,
       "step": 5495
     },
     {
-      "epoch": 0.19,
-      "learning_rate": 4.999008259287312e-05,
-      "loss": 0.3508,
+      "epoch": 0.19821962734710058,
+      "grad_norm": 0.20664586126804352,
+      "learning_rate": 4.998782334781668e-05,
+      "loss": 0.4516,
       "step": 5500
     },
     {
-      "epoch": 0.19,
-      "eval_loss": 0.3391436040401459,
-      "eval_runtime": 10.541,
-      "eval_samples_per_second": 9.487,
-      "eval_steps_per_second": 9.487,
+      "epoch": 0.19821962734710058,
+      "eval_loss": 0.49830162525177,
+      "eval_runtime": 3.5194,
+      "eval_samples_per_second": 28.414,
+      "eval_steps_per_second": 7.103,
       "step": 5500
     },
     {
-      "epoch": 0.19,
-      "learning_rate": 4.999000219801712e-05,
-      "loss": 0.3461,
+      "epoch": 0.19839982700832523,
+      "grad_norm": 0.2016601264476776,
+      "learning_rate": 4.998773211054701e-05,
+      "loss": 0.5032,
       "step": 5505
     },
     {
-      "epoch": 0.19,
-      "learning_rate": 4.99899214786824e-05,
-      "loss": 0.3272,
+      "epoch": 0.19858002666954985,
+      "grad_norm": 0.18409425020217896,
+      "learning_rate": 4.998764053282433e-05,
+      "loss": 0.4853,
       "step": 5510
     },
     {
-      "epoch": 0.19,
-      "learning_rate": 4.9989840434870025e-05,
-      "loss": 0.3521,
+      "epoch": 0.1987602263307745,
+      "grad_norm": 0.18686065077781677,
+      "learning_rate": 4.998754861464989e-05,
+      "loss": 0.4274,
       "step": 5515
     },
     {
-      "epoch": 0.19,
-      "learning_rate": 4.998975906658103e-05,
-      "loss": 0.3151,
+      "epoch": 0.19894042599199913,
+      "grad_norm": 0.20691046118736267,
+      "learning_rate": 4.9987456356024944e-05,
+      "loss": 0.5276,
       "step": 5520
     },
     {
-      "epoch": 0.19,
-      "learning_rate": 4.9989677373816494e-05,
-      "loss": 0.3279,
+      "epoch": 0.19912062565322378,
+      "grad_norm": 0.1962127685546875,
+      "learning_rate": 4.9987363756950736e-05,
+      "loss": 0.4697,
       "step": 5525
     },
     {
-      "epoch": 0.19,
-      "learning_rate": 4.9989595356577465e-05,
-      "loss": 0.3475,
+      "epoch": 0.1993008253144484,
+      "grad_norm": 0.20173537731170654,
+      "learning_rate": 4.9987270817428535e-05,
+      "loss": 0.434,
       "step": 5530
     },
     {
-      "epoch": 0.19,
-      "learning_rate": 4.998951301486501e-05,
-      "loss": 0.3282,
+      "epoch": 0.19948102497567305,
+      "grad_norm": 0.21217285096645355,
+      "learning_rate": 4.998717753745961e-05,
+      "loss": 0.4603,
       "step": 5535
     },
     {
-      "epoch": 0.19,
-      "learning_rate": 4.9989430348680186e-05,
-      "loss": 0.3457,
+      "epoch": 0.19966122463689767,
+      "grad_norm": 0.2284974455833435,
+      "learning_rate": 4.998708391704522e-05,
+      "loss": 0.476,
       "step": 5540
     },
     {
-      "epoch": 0.2,
-      "learning_rate": 4.998934735802409e-05,
-      "loss": 0.3311,
+      "epoch": 0.19984142429812232,
+      "grad_norm": 0.22764502465724945,
+      "learning_rate": 4.998698995618666e-05,
+      "loss": 0.5092,
       "step": 5545
     },
     {
-      "epoch": 0.2,
-      "learning_rate": 4.998926404289779e-05,
-      "loss": 0.3496,
+      "epoch": 0.20002162395934694,
+      "grad_norm": 0.19202855229377747,
+      "learning_rate": 4.998689565488519e-05,
+      "loss": 0.4573,
       "step": 5550
     },
     {
-      "epoch": 0.2,
-      "learning_rate": 4.998918040330235e-05,
-      "loss": 0.3349,
+      "epoch": 0.2002018236205716,
+      "grad_norm": 0.20149993896484375,
+      "learning_rate": 4.998680101314211e-05,
+      "loss": 0.4563,
       "step": 5555
     },
     {
-      "epoch": 0.2,
-      "learning_rate": 4.9989096439238884e-05,
-      "loss": 0.3327,
+      "epoch": 0.20038202328179622,
+      "grad_norm": 0.1978190690279007,
+      "learning_rate": 4.9986706030958705e-05,
+      "loss": 0.4785,
       "step": 5560
     },
     {
-      "epoch": 0.2,
-      "learning_rate": 4.998901215070846e-05,
-      "loss": 0.3649,
+      "epoch": 0.20056222294302087,
+      "grad_norm": 0.186653271317482,
+      "learning_rate": 4.998661070833627e-05,
+      "loss": 0.4688,
       "step": 5565
     },
     {
-      "epoch": 0.2,
-      "learning_rate": 4.998892753771219e-05,
-      "loss": 0.3336,
+      "epoch": 0.2007424226042455,
+      "grad_norm": 0.20869889855384827,
+      "learning_rate": 4.9986515045276094e-05,
+      "loss": 0.4916,
       "step": 5570
     },
     {
-      "epoch": 0.2,
-      "learning_rate": 4.998884260025116e-05,
-      "loss": 0.3531,
+      "epoch": 0.20092262226547014,
+      "grad_norm": 0.2684619128704071,
+      "learning_rate": 4.9986419041779485e-05,
+      "loss": 0.4965,
       "step": 5575
     },
     {
-      "epoch": 0.2,
-      "learning_rate": 4.9988757338326474e-05,
-      "loss": 0.3581,
+      "epoch": 0.2011028219266948,
+      "grad_norm": 0.2171335518360138,
+      "learning_rate": 4.9986322697847765e-05,
+      "loss": 0.4418,
       "step": 5580
     },
     {
-      "epoch": 0.2,
-      "learning_rate": 4.998867175193924e-05,
-      "loss": 0.3178,
+      "epoch": 0.20128302158791941,
+      "grad_norm": 0.24311456084251404,
+      "learning_rate": 4.998622601348223e-05,
+      "loss": 0.4689,
       "step": 5585
     },
     {
-      "epoch": 0.2,
-      "learning_rate": 4.998858584109059e-05,
-      "loss": 0.3431,
+      "epoch": 0.20146322124914406,
+      "grad_norm": 0.19631347060203552,
+      "learning_rate": 4.998612898868421e-05,
+      "loss": 0.4882,
       "step": 5590
     },
     {
-      "epoch": 0.2,
-      "learning_rate": 4.99884996057816e-05,
-      "loss": 0.3362,
+      "epoch": 0.2016434209103687,
+      "grad_norm": 0.2597336173057556,
+      "learning_rate": 4.9986031623455006e-05,
+      "loss": 0.5101,
       "step": 5595
     },
     {
-      "epoch": 0.2,
-      "learning_rate": 4.998841304601342e-05,
-      "loss": 0.359,
+      "epoch": 0.20182362057159334,
+      "grad_norm": 0.2971448600292206,
+      "learning_rate": 4.9985933917795966e-05,
+      "loss": 0.4788,
       "step": 5600
     },
     {
-      "epoch": 0.2,
-      "learning_rate": 4.998832616178716e-05,
-      "loss": 0.331,
+      "epoch": 0.20200382023281796,
+      "grad_norm": 0.2276962250471115,
+      "learning_rate": 4.998583587170842e-05,
+      "loss": 0.4641,
       "step": 5605
     },
     {
-      "epoch": 0.2,
-      "learning_rate": 4.998823895310396e-05,
-      "loss": 0.3425,
+      "epoch": 0.2021840198940426,
+      "grad_norm": 0.16654905676841736,
+      "learning_rate": 4.99857374851937e-05,
+      "loss": 0.4612,
       "step": 5610
     },
     {
-      "epoch": 0.2,
-      "learning_rate": 4.9988151419964935e-05,
-      "loss": 0.3234,
+      "epoch": 0.20236421955526723,
+      "grad_norm": 0.17817789316177368,
+      "learning_rate": 4.998563875825313e-05,
+      "loss": 0.4524,
       "step": 5615
     },
     {
-      "epoch": 0.2,
-      "learning_rate": 4.998806356237124e-05,
-      "loss": 0.3417,
+      "epoch": 0.20254441921649188,
+      "grad_norm": 0.1932527720928192,
+      "learning_rate": 4.998553969088807e-05,
+      "loss": 0.4654,
       "step": 5620
     },
     {
-      "epoch": 0.2,
-      "learning_rate": 4.9987975380324e-05,
-      "loss": 0.351,
+      "epoch": 0.2027246188777165,
+      "grad_norm": 0.22446228563785553,
+      "learning_rate": 4.9985440283099885e-05,
+      "loss": 0.5021,
       "step": 5625
     },
     {
-      "epoch": 0.2,
-      "learning_rate": 4.9987886873824374e-05,
-      "loss": 0.352,
+      "epoch": 0.20290481853894116,
+      "grad_norm": 0.22577793896198273,
+      "learning_rate": 4.99853405348899e-05,
+      "loss": 0.5042,
       "step": 5630
     },
     {
-      "epoch": 0.2,
-      "learning_rate": 4.99877980428735e-05,
-      "loss": 0.364,
+      "epoch": 0.20308501820016578,
+      "grad_norm": 0.19390153884887695,
+      "learning_rate": 4.998524044625949e-05,
+      "loss": 0.4851,
       "step": 5635
     },
     {
-      "epoch": 0.2,
-      "learning_rate": 4.998770888747254e-05,
-      "loss": 0.3268,
+      "epoch": 0.20326521786139043,
+      "grad_norm": 0.22741837799549103,
+      "learning_rate": 4.998514001721002e-05,
+      "loss": 0.4954,
       "step": 5640
     },
     {
-      "epoch": 0.2,
-      "learning_rate": 4.998761940762265e-05,
-      "loss": 0.3856,
+      "epoch": 0.20344541752261505,
+      "grad_norm": 0.20795418322086334,
+      "learning_rate": 4.998503924774285e-05,
+      "loss": 0.4748,
       "step": 5645
     },
     {
-      "epoch": 0.2,
-      "learning_rate": 4.998752960332498e-05,
-      "loss": 0.3381,
+      "epoch": 0.2036256171838397,
+      "grad_norm": 0.20817185938358307,
+      "learning_rate": 4.998493813785936e-05,
+      "loss": 0.5023,
       "step": 5650
     },
     {
-      "epoch": 0.2,
-      "learning_rate": 4.998743947458072e-05,
-      "loss": 0.3492,
+      "epoch": 0.20380581684506432,
+      "grad_norm": 0.24967476725578308,
+      "learning_rate": 4.9984836687560924e-05,
+      "loss": 0.5128,
       "step": 5655
     },
     {
-      "epoch": 0.2,
-      "learning_rate": 4.998734902139102e-05,
-      "loss": 0.3318,
+      "epoch": 0.20398601650628898,
+      "grad_norm": 0.2160884588956833,
+      "learning_rate": 4.998473489684892e-05,
+      "loss": 0.4386,
       "step": 5660
     },
     {
-      "epoch": 0.2,
-      "learning_rate": 4.9987258243757055e-05,
-      "loss": 0.3393,
+      "epoch": 0.2041662161675136,
+      "grad_norm": 0.19086240231990814,
+      "learning_rate": 4.998463276572475e-05,
+      "loss": 0.4722,
       "step": 5665
     },
     {
-      "epoch": 0.2,
-      "learning_rate": 4.998716714168e-05,
-      "loss": 0.3467,
+      "epoch": 0.20434641582873825,
+      "grad_norm": 0.21120908856391907,
+      "learning_rate": 4.9984530294189794e-05,
+      "loss": 0.4662,
       "step": 5670
     },
     {
-      "epoch": 0.2,
-      "learning_rate": 4.9987075715161064e-05,
-      "loss": 0.3667,
+      "epoch": 0.20452661548996287,
+      "grad_norm": 0.20728902518749237,
+      "learning_rate": 4.9984427482245445e-05,
+      "loss": 0.4166,
       "step": 5675
     },
     {
-      "epoch": 0.2,
-      "learning_rate": 4.998698396420141e-05,
-      "loss": 0.3488,
+      "epoch": 0.20470681515118752,
+      "grad_norm": 0.24865606427192688,
+      "learning_rate": 4.998432432989311e-05,
+      "loss": 0.49,
       "step": 5680
     },
     {
-      "epoch": 0.2,
-      "learning_rate": 4.998689188880223e-05,
-      "loss": 0.3166,
+      "epoch": 0.20488701481241214,
+      "grad_norm": 0.3118506968021393,
+      "learning_rate": 4.99842208371342e-05,
+      "loss": 0.5066,
       "step": 5685
     },
     {
-      "epoch": 0.2,
-      "learning_rate": 4.998679948896474e-05,
-      "loss": 0.3351,
+      "epoch": 0.2050672144736368,
+      "grad_norm": 0.21482154726982117,
+      "learning_rate": 4.998411700397011e-05,
+      "loss": 0.5173,
       "step": 5690
     },
     {
-      "epoch": 0.2,
-      "learning_rate": 4.998670676469012e-05,
-      "loss": 0.3503,
+      "epoch": 0.20524741413486142,
+      "grad_norm": 0.2417951226234436,
+      "learning_rate": 4.998401283040226e-05,
+      "loss": 0.4688,
       "step": 5695
     },
     {
-      "epoch": 0.2,
-      "learning_rate": 4.998661371597958e-05,
-      "loss": 0.3226,
+      "epoch": 0.20542761379608607,
+      "grad_norm": 0.22649191319942474,
+      "learning_rate": 4.9983908316432084e-05,
+      "loss": 0.4748,
       "step": 5700
     },
     {
-      "epoch": 0.2,
-      "learning_rate": 4.998652034283432e-05,
-      "loss": 0.3403,
+      "epoch": 0.2056078134573107,
+      "grad_norm": 0.28141918778419495,
+      "learning_rate": 4.998380346206099e-05,
+      "loss": 0.4986,
       "step": 5705
     },
     {
-      "epoch": 0.2,
-      "learning_rate": 4.998642664525557e-05,
-      "loss": 0.3472,
+      "epoch": 0.20578801311853534,
+      "grad_norm": 0.25580498576164246,
+      "learning_rate": 4.99836982672904e-05,
+      "loss": 0.5207,
       "step": 5710
     },
     {
-      "epoch": 0.2,
-      "learning_rate": 4.998633262324453e-05,
-      "loss": 0.3504,
+      "epoch": 0.20596821277975996,
+      "grad_norm": 0.2193039357662201,
+      "learning_rate": 4.998359273212177e-05,
+      "loss": 0.5007,
       "step": 5715
     },
     {
-      "epoch": 0.2,
-      "learning_rate": 4.9986238276802435e-05,
-      "loss": 0.3345,
+      "epoch": 0.2061484124409846,
+      "grad_norm": 0.22306658327579498,
+      "learning_rate": 4.998348685655653e-05,
+      "loss": 0.4909,
       "step": 5720
     },
     {
-      "epoch": 0.2,
-      "learning_rate": 4.998614360593049e-05,
-      "loss": 0.3224,
+      "epoch": 0.20632861210220924,
+      "grad_norm": 0.20013074576854706,
+      "learning_rate": 4.998338064059611e-05,
+      "loss": 0.472,
       "step": 5725
     },
     {
-      "epoch": 0.2,
-      "learning_rate": 4.998604861062995e-05,
-      "loss": 0.3267,
+      "epoch": 0.20650881176343389,
+      "grad_norm": 0.2772786319255829,
+      "learning_rate": 4.998327408424196e-05,
+      "loss": 0.4913,
       "step": 5730
     },
     {
-      "epoch": 0.2,
-      "learning_rate": 4.9985953290902024e-05,
-      "loss": 0.3438,
+      "epoch": 0.2066890114246585,
+      "grad_norm": 0.23623016476631165,
+      "learning_rate": 4.998316718749555e-05,
+      "loss": 0.4423,
       "step": 5735
     },
     {
-      "epoch": 0.2,
-      "learning_rate": 4.998585764674797e-05,
-      "loss": 0.3184,
+      "epoch": 0.20686921108588316,
+      "grad_norm": 0.2390371859073639,
+      "learning_rate": 4.9983059950358316e-05,
+      "loss": 0.4747,
       "step": 5740
     },
     {
-      "epoch": 0.2,
-      "learning_rate": 4.998576167816902e-05,
-      "loss": 0.3406,
+      "epoch": 0.20704941074710778,
+      "grad_norm": 0.16959403455257416,
+      "learning_rate": 4.9982952372831724e-05,
+      "loss": 0.4875,
       "step": 5745
     },
     {
-      "epoch": 0.2,
-      "learning_rate": 4.9985665385166416e-05,
-      "loss": 0.3495,
+      "epoch": 0.20722961040833243,
+      "grad_norm": 0.1819695085287094,
+      "learning_rate": 4.998284445491726e-05,
+      "loss": 0.4731,
       "step": 5750
     },
     {
-      "epoch": 0.2,
-      "learning_rate": 4.998556876774141e-05,
-      "loss": 0.3361,
+      "epoch": 0.20740981006955708,
+      "grad_norm": 0.21502423286437988,
+      "learning_rate": 4.998273619661636e-05,
+      "loss": 0.4587,
       "step": 5755
     },
     {
-      "epoch": 0.2,
-      "learning_rate": 4.998547182589527e-05,
-      "loss": 0.3761,
+      "epoch": 0.2075900097307817,
+      "grad_norm": 0.1912410408258438,
+      "learning_rate": 4.998262759793052e-05,
+      "loss": 0.4497,
       "step": 5760
     },
     {
-      "epoch": 0.2,
-      "learning_rate": 4.998537455962924e-05,
-      "loss": 0.3232,
+      "epoch": 0.20777020939200636,
+      "grad_norm": 0.21954499185085297,
+      "learning_rate": 4.9982518658861224e-05,
+      "loss": 0.4768,
       "step": 5765
     },
     {
-      "epoch": 0.2,
-      "learning_rate": 4.998527696894458e-05,
-      "loss": 0.3331,
+      "epoch": 0.20795040905323098,
+      "grad_norm": 0.23011371493339539,
+      "learning_rate": 4.998240937940993e-05,
+      "loss": 0.4931,
       "step": 5770
     },
     {
-      "epoch": 0.2,
-      "learning_rate": 4.998517905384258e-05,
-      "loss": 0.3593,
+      "epoch": 0.20813060871445563,
+      "grad_norm": 0.18826115131378174,
+      "learning_rate": 4.998229975957816e-05,
+      "loss": 0.4547,
       "step": 5775
     },
     {
-      "epoch": 0.2,
-      "learning_rate": 4.998508081432448e-05,
-      "loss": 0.3332,
+      "epoch": 0.20831080837568025,
+      "grad_norm": 0.23759086430072784,
+      "learning_rate": 4.998218979936739e-05,
+      "loss": 0.4687,
       "step": 5780
     },
     {
-      "epoch": 0.2,
-      "learning_rate": 4.998498225039158e-05,
-      "loss": 0.3465,
+      "epoch": 0.2084910080369049,
+      "grad_norm": 0.2513539493083954,
+      "learning_rate": 4.9982079498779125e-05,
+      "loss": 0.4908,
       "step": 5785
     },
     {
-      "epoch": 0.2,
-      "learning_rate": 4.998488336204515e-05,
-      "loss": 0.3313,
+      "epoch": 0.20867120769812952,
+      "grad_norm": 0.23931877315044403,
+      "learning_rate": 4.998196885781485e-05,
+      "loss": 0.4848,
       "step": 5790
     },
     {
-      "epoch": 0.2,
-      "learning_rate": 4.998478414928647e-05,
-      "loss": 0.3309,
+      "epoch": 0.20885140735935417,
+      "grad_norm": 0.21470136940479279,
+      "learning_rate": 4.99818578764761e-05,
+      "loss": 0.4397,
       "step": 5795
     },
     {
-      "epoch": 0.2,
-      "learning_rate": 4.998468461211684e-05,
-      "loss": 0.3517,
+      "epoch": 0.2090316070205788,
+      "grad_norm": 0.181662455201149,
+      "learning_rate": 4.9981746554764366e-05,
+      "loss": 0.4663,
       "step": 5800
     },
     {
-      "epoch": 0.2,
-      "learning_rate": 4.998458475053754e-05,
-      "loss": 0.3301,
+      "epoch": 0.20921180668180345,
+      "grad_norm": 0.20043453574180603,
+      "learning_rate": 4.9981634892681175e-05,
+      "loss": 0.5032,
       "step": 5805
     },
     {
-      "epoch": 0.2,
-      "learning_rate": 4.9984484564549884e-05,
-      "loss": 0.3438,
+      "epoch": 0.20939200634302807,
+      "grad_norm": 0.21683253347873688,
+      "learning_rate": 4.998152289022804e-05,
+      "loss": 0.4492,
       "step": 5810
     },
     {
-      "epoch": 0.2,
-      "learning_rate": 4.998438405415515e-05,
-      "loss": 0.334,
+      "epoch": 0.20957220600425272,
+      "grad_norm": 0.24924011528491974,
+      "learning_rate": 4.998141054740649e-05,
+      "loss": 0.4736,
       "step": 5815
     },
     {
-      "epoch": 0.2,
-      "learning_rate": 4.998428321935466e-05,
-      "loss": 0.3086,
+      "epoch": 0.20975240566547734,
+      "grad_norm": 0.2400321364402771,
+      "learning_rate": 4.998129786421807e-05,
+      "loss": 0.4451,
       "step": 5820
     },
     {
-      "epoch": 0.2,
-      "learning_rate": 4.998418206014971e-05,
-      "loss": 0.3291,
+      "epoch": 0.209932605326702,
+      "grad_norm": 0.19398939609527588,
+      "learning_rate": 4.9981184840664294e-05,
+      "loss": 0.4288,
       "step": 5825
     },
     {
-      "epoch": 0.21,
-      "learning_rate": 4.9984080576541636e-05,
-      "loss": 0.3316,
+      "epoch": 0.21011280498792662,
+      "grad_norm": 0.20897026360034943,
+      "learning_rate": 4.9981071476746717e-05,
+      "loss": 0.4522,
       "step": 5830
     },
     {
-      "epoch": 0.21,
-      "learning_rate": 4.9983978768531724e-05,
-      "loss": 0.3494,
+      "epoch": 0.21029300464915127,
+      "grad_norm": 0.19112837314605713,
+      "learning_rate": 4.998095777246687e-05,
+      "loss": 0.4699,
       "step": 5835
     },
     {
-      "epoch": 0.21,
-      "learning_rate": 4.9983876636121325e-05,
-      "loss": 0.3372,
+      "epoch": 0.2104732043103759,
+      "grad_norm": 0.27279478311538696,
+      "learning_rate": 4.998084372782631e-05,
+      "loss": 0.4821,
       "step": 5840
     },
     {
-      "epoch": 0.21,
-      "learning_rate": 4.998377417931175e-05,
-      "loss": 0.3628,
+      "epoch": 0.21065340397160054,
+      "grad_norm": 0.17838047444820404,
+      "learning_rate": 4.99807293428266e-05,
+      "loss": 0.4548,
       "step": 5845
     },
     {
-      "epoch": 0.21,
-      "learning_rate": 4.998367139810433e-05,
-      "loss": 0.3209,
+      "epoch": 0.21083360363282516,
+      "grad_norm": 0.23038989305496216,
+      "learning_rate": 4.9980614617469284e-05,
+      "loss": 0.4893,
       "step": 5850
     },
     {
-      "epoch": 0.21,
-      "learning_rate": 4.99835682925004e-05,
-      "loss": 0.3382,
+      "epoch": 0.2110138032940498,
+      "grad_norm": 0.21518467366695404,
+      "learning_rate": 4.998049955175593e-05,
+      "loss": 0.4707,
       "step": 5855
     },
     {
-      "epoch": 0.21,
-      "learning_rate": 4.998346486250131e-05,
-      "loss": 0.3258,
+      "epoch": 0.21119400295527443,
+      "grad_norm": 0.2731831669807434,
+      "learning_rate": 4.998038414568811e-05,
+      "loss": 0.4649,
       "step": 5860
     },
     {
-      "epoch": 0.21,
-      "learning_rate": 4.998336110810838e-05,
-      "loss": 0.3392,
+      "epoch": 0.21137420261649909,
+      "grad_norm": 0.2282124161720276,
+      "learning_rate": 4.998026839926738e-05,
+      "loss": 0.4721,
       "step": 5865
     },
     {
-      "epoch": 0.21,
-      "learning_rate": 4.9983257029322975e-05,
-      "loss": 0.3324,
+      "epoch": 0.2115544022777237,
+      "grad_norm": 0.24522550404071808,
+      "learning_rate": 4.9980152312495345e-05,
+      "loss": 0.4881,
       "step": 5870
     },
     {
-      "epoch": 0.21,
-      "learning_rate": 4.998315262614645e-05,
-      "loss": 0.3524,
+      "epoch": 0.21173460193894836,
+      "grad_norm": 0.24952565133571625,
+      "learning_rate": 4.998003588537356e-05,
+      "loss": 0.4917,
       "step": 5875
     },
     {
-      "epoch": 0.21,
-      "learning_rate": 4.9983047898580134e-05,
-      "loss": 0.3644,
+      "epoch": 0.21191480160017298,
+      "grad_norm": 0.19938194751739502,
+      "learning_rate": 4.997991911790363e-05,
+      "loss": 0.4714,
       "step": 5880
     },
     {
-      "epoch": 0.21,
-      "learning_rate": 4.998294284662542e-05,
-      "loss": 0.3748,
+      "epoch": 0.21209500126139763,
+      "grad_norm": 0.25844866037368774,
+      "learning_rate": 4.997980201008713e-05,
+      "loss": 0.485,
       "step": 5885
     },
     {
-      "epoch": 0.21,
-      "learning_rate": 4.998283747028366e-05,
-      "loss": 0.326,
+      "epoch": 0.21227520092262225,
+      "grad_norm": 0.19596533477306366,
+      "learning_rate": 4.9979684561925663e-05,
+      "loss": 0.4782,
       "step": 5890
     },
     {
-      "epoch": 0.21,
-      "learning_rate": 4.9982731769556216e-05,
-      "loss": 0.3364,
+      "epoch": 0.2124554005838469,
+      "grad_norm": 0.17095403373241425,
+      "learning_rate": 4.9979566773420836e-05,
+      "loss": 0.4786,
       "step": 5895
     },
     {
-      "epoch": 0.21,
-      "learning_rate": 4.9982625744444464e-05,
-      "loss": 0.3307,
+      "epoch": 0.21263560024507153,
+      "grad_norm": 0.20486395061016083,
+      "learning_rate": 4.9979448644574254e-05,
+      "loss": 0.4701,
       "step": 5900
     },
     {
-      "epoch": 0.21,
-      "learning_rate": 4.9982519394949776e-05,
-      "loss": 0.3832,
+      "epoch": 0.21281579990629618,
+      "grad_norm": 0.19251783192157745,
+      "learning_rate": 4.997933017538751e-05,
+      "loss": 0.4164,
       "step": 5905
     },
     {
-      "epoch": 0.21,
-      "learning_rate": 4.998241272107355e-05,
-      "loss": 0.3344,
+      "epoch": 0.2129959995675208,
+      "grad_norm": 0.21283471584320068,
+      "learning_rate": 4.9979211365862235e-05,
+      "loss": 0.4842,
       "step": 5910
     },
     {
-      "epoch": 0.21,
-      "learning_rate": 4.998230572281715e-05,
-      "loss": 0.3497,
+      "epoch": 0.21317619922874545,
+      "grad_norm": 0.21387127041816711,
+      "learning_rate": 4.997909221600003e-05,
+      "loss": 0.4786,
       "step": 5915
     },
     {
-      "epoch": 0.21,
-      "learning_rate": 4.9982198400181975e-05,
-      "loss": 0.3361,
+      "epoch": 0.2133563988899701,
+      "grad_norm": 0.22376859188079834,
+      "learning_rate": 4.997897272580254e-05,
+      "loss": 0.4504,
       "step": 5920
     },
     {
-      "epoch": 0.21,
-      "learning_rate": 4.998209075316942e-05,
-      "loss": 0.3129,
+      "epoch": 0.21353659855119472,
+      "grad_norm": 0.15793459117412567,
+      "learning_rate": 4.997885289527139e-05,
+      "loss": 0.4533,
       "step": 5925
     },
     {
-      "epoch": 0.21,
-      "learning_rate": 4.998198278178088e-05,
-      "loss": 0.3266,
+      "epoch": 0.21371679821241937,
+      "grad_norm": 0.2030140906572342,
+      "learning_rate": 4.9978732724408195e-05,
+      "loss": 0.4771,
       "step": 5930
     },
     {
-      "epoch": 0.21,
-      "learning_rate": 4.9981874486017746e-05,
-      "loss": 0.3299,
+      "epoch": 0.213896997873644,
+      "grad_norm": 0.19730666279792786,
+      "learning_rate": 4.997861221321461e-05,
+      "loss": 0.4598,
       "step": 5935
     },
     {
-      "epoch": 0.21,
-      "learning_rate": 4.9981765865881444e-05,
-      "loss": 0.3378,
+      "epoch": 0.21407719753486865,
+      "grad_norm": 0.28604263067245483,
+      "learning_rate": 4.9978491361692255e-05,
+      "loss": 0.5014,
       "step": 5940
     },
     {
-      "epoch": 0.21,
-      "learning_rate": 4.998165692137338e-05,
-      "loss": 0.3558,
+      "epoch": 0.21425739719609327,
+      "grad_norm": 0.2171340137720108,
+      "learning_rate": 4.99783701698428e-05,
+      "loss": 0.4597,
       "step": 5945
     },
     {
-      "epoch": 0.21,
-      "learning_rate": 4.998154765249496e-05,
-      "loss": 0.3249,
+      "epoch": 0.21443759685731792,
+      "grad_norm": 0.24565933644771576,
+      "learning_rate": 4.9978248637667883e-05,
+      "loss": 0.4856,
       "step": 5950
     },
     {
-      "epoch": 0.21,
-      "learning_rate": 4.998143805924761e-05,
-      "loss": 0.3064,
+      "epoch": 0.21461779651854254,
+      "grad_norm": 0.20074883103370667,
+      "learning_rate": 4.997812676516917e-05,
+      "loss": 0.4681,
       "step": 5955
     },
     {
-      "epoch": 0.21,
-      "learning_rate": 4.998132814163275e-05,
-      "loss": 0.3358,
+      "epoch": 0.2147979961797672,
+      "grad_norm": 0.17820361256599426,
+      "learning_rate": 4.9978004552348314e-05,
+      "loss": 0.485,
       "step": 5960
     },
     {
-      "epoch": 0.21,
-      "learning_rate": 4.998121789965181e-05,
-      "loss": 0.3356,
+      "epoch": 0.21497819584099181,
+      "grad_norm": 0.19465869665145874,
+      "learning_rate": 4.997788199920698e-05,
+      "loss": 0.4693,
       "step": 5965
     },
     {
-      "epoch": 0.21,
-      "learning_rate": 4.998110733330621e-05,
-      "loss": 0.3289,
+      "epoch": 0.21515839550221647,
+      "grad_norm": 0.22670099139213562,
+      "learning_rate": 4.997775910574685e-05,
+      "loss": 0.5077,
       "step": 5970
     },
     {
-      "epoch": 0.21,
-      "learning_rate": 4.99809964425974e-05,
-      "loss": 0.3755,
+      "epoch": 0.2153385951634411,
+      "grad_norm": 0.14773029088974,
+      "learning_rate": 4.997763587196957e-05,
+      "loss": 0.4664,
       "step": 5975
     },
     {
-      "epoch": 0.21,
-      "learning_rate": 4.998088522752681e-05,
-      "loss": 0.3381,
+      "epoch": 0.21551879482466574,
+      "grad_norm": 0.26467835903167725,
+      "learning_rate": 4.997751229787685e-05,
+      "loss": 0.5078,
       "step": 5980
     },
     {
-      "epoch": 0.21,
-      "learning_rate": 4.9980773688095896e-05,
-      "loss": 0.3545,
+      "epoch": 0.21569899448589036,
+      "grad_norm": 0.20489193499088287,
+      "learning_rate": 4.9977388383470356e-05,
+      "loss": 0.4256,
       "step": 5985
     },
     {
-      "epoch": 0.21,
-      "learning_rate": 4.99806618243061e-05,
-      "loss": 0.3409,
+      "epoch": 0.215879194147115,
+      "grad_norm": 0.2557709217071533,
+      "learning_rate": 4.997726412875178e-05,
+      "loss": 0.4677,
       "step": 5990
     },
     {
-      "epoch": 0.21,
-      "learning_rate": 4.998054963615886e-05,
-      "loss": 0.336,
+      "epoch": 0.21605939380833963,
+      "grad_norm": 0.20980533957481384,
+      "learning_rate": 4.997713953372282e-05,
+      "loss": 0.4954,
       "step": 5995
     },
     {
-      "epoch": 0.21,
-      "learning_rate": 4.998043712365566e-05,
-      "loss": 0.3559,
+      "epoch": 0.21623959346956428,
+      "grad_norm": 0.20784328877925873,
+      "learning_rate": 4.997701459838517e-05,
+      "loss": 0.515,
       "step": 6000
     },
     {
-      "epoch": 0.21,
-      "eval_loss": 0.3358333706855774,
-      "eval_runtime": 10.5554,
-      "eval_samples_per_second": 9.474,
-      "eval_steps_per_second": 9.474,
+      "epoch": 0.21623959346956428,
+      "eval_loss": 0.4947231411933899,
+      "eval_runtime": 3.5181,
+      "eval_samples_per_second": 28.424,
+      "eval_steps_per_second": 7.106,
       "step": 6000
     },
     {
-      "epoch": 0.21,
-      "learning_rate": 4.9980324286797935e-05,
-      "loss": 0.3347,
+      "epoch": 0.2164197931307889,
+      "grad_norm": 0.2340945601463318,
+      "learning_rate": 4.997688932274053e-05,
+      "loss": 0.4411,
       "step": 6005
     },
     {
-      "epoch": 0.21,
-      "learning_rate": 4.9980211125587166e-05,
-      "loss": 0.3317,
+      "epoch": 0.21659999279201356,
+      "grad_norm": 0.1668865829706192,
+      "learning_rate": 4.997676370679061e-05,
+      "loss": 0.4891,
       "step": 6010
     },
     {
-      "epoch": 0.21,
-      "learning_rate": 4.998009764002482e-05,
-      "loss": 0.3474,
+      "epoch": 0.21678019245323818,
+      "grad_norm": 0.22756971418857574,
+      "learning_rate": 4.997663775053712e-05,
+      "loss": 0.4992,
       "step": 6015
     },
     {
-      "epoch": 0.21,
-      "learning_rate": 4.9979983830112366e-05,
-      "loss": 0.3507,
+      "epoch": 0.21696039211446283,
+      "grad_norm": 0.21588781476020813,
+      "learning_rate": 4.997651145398177e-05,
+      "loss": 0.4388,
       "step": 6020
     },
     {
-      "epoch": 0.21,
-      "learning_rate": 4.997986969585129e-05,
-      "loss": 0.3373,
+      "epoch": 0.21714059177568745,
+      "grad_norm": 0.1796243041753769,
+      "learning_rate": 4.9976384817126295e-05,
+      "loss": 0.4314,
       "step": 6025
     },
     {
-      "epoch": 0.21,
-      "learning_rate": 4.997975523724306e-05,
-      "loss": 0.3213,
+      "epoch": 0.2173207914369121,
+      "grad_norm": 0.2105339765548706,
+      "learning_rate": 4.9976257839972406e-05,
+      "loss": 0.4898,
       "step": 6030
     },
     {
-      "epoch": 0.21,
-      "learning_rate": 4.9979640454289175e-05,
-      "loss": 0.3132,
+      "epoch": 0.21750099109813673,
+      "grad_norm": 0.23153090476989746,
+      "learning_rate": 4.9976130522521845e-05,
+      "loss": 0.4549,
       "step": 6035
     },
     {
-      "epoch": 0.21,
-      "learning_rate": 4.9979525346991126e-05,
-      "loss": 0.3358,
+      "epoch": 0.21768119075936138,
+      "grad_norm": 0.240932434797287,
+      "learning_rate": 4.997600286477634e-05,
+      "loss": 0.4623,
       "step": 6040
     },
     {
-      "epoch": 0.21,
-      "learning_rate": 4.997940991535039e-05,
-      "loss": 0.3218,
+      "epoch": 0.217861390420586,
+      "grad_norm": 0.19893620908260345,
+      "learning_rate": 4.997587486673763e-05,
+      "loss": 0.4703,
       "step": 6045
     },
     {
-      "epoch": 0.21,
-      "learning_rate": 4.997929415936849e-05,
-      "loss": 0.3567,
+      "epoch": 0.21804159008181065,
+      "grad_norm": 0.17426280677318573,
+      "learning_rate": 4.9975746528407466e-05,
+      "loss": 0.4738,
       "step": 6050
     },
     {
-      "epoch": 0.21,
-      "learning_rate": 4.9979178079046914e-05,
-      "loss": 0.3256,
+      "epoch": 0.21822178974303527,
+      "grad_norm": 0.20304343104362488,
+      "learning_rate": 4.997561784978758e-05,
+      "loss": 0.4971,
       "step": 6055
     },
     {
-      "epoch": 0.21,
-      "learning_rate": 4.997906167438717e-05,
-      "loss": 0.3522,
+      "epoch": 0.21840198940425992,
+      "grad_norm": 0.20934487879276276,
+      "learning_rate": 4.997548883087974e-05,
+      "loss": 0.4552,
       "step": 6060
     },
     {
-      "epoch": 0.21,
-      "learning_rate": 4.997894494539077e-05,
-      "loss": 0.351,
+      "epoch": 0.21858218906548454,
+      "grad_norm": 0.17177551984786987,
+      "learning_rate": 4.99753594716857e-05,
+      "loss": 0.4991,
       "step": 6065
     },
     {
-      "epoch": 0.21,
-      "learning_rate": 4.997882789205924e-05,
-      "loss": 0.3127,
+      "epoch": 0.2187623887267092,
+      "grad_norm": 0.2459387630224228,
+      "learning_rate": 4.9975229772207224e-05,
+      "loss": 0.4845,
       "step": 6070
     },
     {
-      "epoch": 0.21,
-      "learning_rate": 4.997871051439409e-05,
-      "loss": 0.3374,
+      "epoch": 0.21894258838793382,
+      "grad_norm": 0.23840418457984924,
+      "learning_rate": 4.9975099732446085e-05,
+      "loss": 0.4877,
       "step": 6075
     },
     {
-      "epoch": 0.21,
-      "learning_rate": 4.997859281239684e-05,
-      "loss": 0.3133,
+      "epoch": 0.21912278804915847,
+      "grad_norm": 0.2193634808063507,
+      "learning_rate": 4.9974969352404036e-05,
+      "loss": 0.4465,
       "step": 6080
     },
     {
-      "epoch": 0.21,
-      "learning_rate": 4.997847478606902e-05,
-      "loss": 0.3224,
+      "epoch": 0.21930298771038312,
+      "grad_norm": 0.15216214954853058,
+      "learning_rate": 4.997483863208288e-05,
+      "loss": 0.4538,
       "step": 6085
     },
     {
-      "epoch": 0.21,
-      "learning_rate": 4.997835643541218e-05,
-      "loss": 0.3394,
+      "epoch": 0.21948318737160774,
+      "grad_norm": 0.1821216493844986,
+      "learning_rate": 4.997470757148437e-05,
+      "loss": 0.4328,
       "step": 6090
     },
     {
-      "epoch": 0.21,
-      "learning_rate": 4.997823776042782e-05,
-      "loss": 0.3583,
+      "epoch": 0.2196633870328324,
+      "grad_norm": 0.2372681200504303,
+      "learning_rate": 4.99745761706103e-05,
+      "loss": 0.4391,
       "step": 6095
     },
     {
-      "epoch": 0.21,
-      "learning_rate": 4.997811876111753e-05,
-      "loss": 0.3395,
+      "epoch": 0.219843586694057,
+      "grad_norm": 0.1972285658121109,
+      "learning_rate": 4.9974444429462476e-05,
+      "loss": 0.4586,
       "step": 6100
     },
     {
-      "epoch": 0.21,
-      "learning_rate": 4.997799943748281e-05,
-      "loss": 0.3493,
+      "epoch": 0.22002378635528166,
+      "grad_norm": 0.20691488683223724,
+      "learning_rate": 4.997431234804267e-05,
+      "loss": 0.5046,
       "step": 6105
     },
     {
-      "epoch": 0.21,
-      "learning_rate": 4.997787978952524e-05,
-      "loss": 0.3356,
+      "epoch": 0.2202039860165063,
+      "grad_norm": 0.20699431002140045,
+      "learning_rate": 4.9974179926352706e-05,
+      "loss": 0.4709,
       "step": 6110
     },
     {
-      "epoch": 0.22,
-      "learning_rate": 4.997775981724636e-05,
-      "loss": 0.3484,
+      "epoch": 0.22038418567773094,
+      "grad_norm": 0.24073725938796997,
+      "learning_rate": 4.997404716439438e-05,
+      "loss": 0.4774,
       "step": 6115
     },
     {
-      "epoch": 0.22,
-      "learning_rate": 4.9977639520647735e-05,
-      "loss": 0.3517,
+      "epoch": 0.22056438533895556,
+      "grad_norm": 0.19937781989574432,
+      "learning_rate": 4.997391406216948e-05,
+      "loss": 0.4387,
       "step": 6120
     },
     {
-      "epoch": 0.22,
-      "learning_rate": 4.9977518899730915e-05,
-      "loss": 0.3185,
+      "epoch": 0.2207445850001802,
+      "grad_norm": 0.203308567404747,
+      "learning_rate": 4.997378061967984e-05,
+      "loss": 0.4535,
       "step": 6125
     },
     {
-      "epoch": 0.22,
-      "learning_rate": 4.997739795449747e-05,
-      "loss": 0.3593,
+      "epoch": 0.22092478466140483,
+      "grad_norm": 0.1698613315820694,
+      "learning_rate": 4.997364683692728e-05,
+      "loss": 0.4852,
       "step": 6130
     },
     {
-      "epoch": 0.22,
-      "learning_rate": 4.997727668494898e-05,
-      "loss": 0.3254,
+      "epoch": 0.22110498432262948,
+      "grad_norm": 0.22055956721305847,
+      "learning_rate": 4.997351271391362e-05,
+      "loss": 0.5054,
       "step": 6135
     },
     {
-      "epoch": 0.22,
-      "learning_rate": 4.9977155091087e-05,
-      "loss": 0.3356,
+      "epoch": 0.2212851839838541,
+      "grad_norm": 0.25628286600112915,
+      "learning_rate": 4.997337825064068e-05,
+      "loss": 0.4846,
       "step": 6140
     },
     {
-      "epoch": 0.22,
-      "learning_rate": 4.997703317291313e-05,
-      "loss": 0.355,
+      "epoch": 0.22146538364507876,
+      "grad_norm": 0.17073556780815125,
+      "learning_rate": 4.9973243447110294e-05,
+      "loss": 0.4608,
       "step": 6145
     },
     {
-      "epoch": 0.22,
-      "learning_rate": 4.997691093042896e-05,
-      "loss": 0.3249,
+      "epoch": 0.22164558330630338,
+      "grad_norm": 0.13316601514816284,
+      "learning_rate": 4.99731083033243e-05,
+      "loss": 0.4658,
       "step": 6150
     },
     {
-      "epoch": 0.22,
-      "learning_rate": 4.997678836363604e-05,
-      "loss": 0.3273,
+      "epoch": 0.22182578296752803,
+      "grad_norm": 0.22222983837127686,
+      "learning_rate": 4.997297281928455e-05,
+      "loss": 0.4558,
       "step": 6155
     },
     {
-      "epoch": 0.22,
-      "learning_rate": 4.997666547253599e-05,
-      "loss": 0.3471,
+      "epoch": 0.22200598262875265,
+      "grad_norm": 0.21334873139858246,
+      "learning_rate": 4.997283699499287e-05,
+      "loss": 0.4884,
       "step": 6160
     },
     {
-      "epoch": 0.22,
-      "learning_rate": 4.99765422571304e-05,
-      "loss": 0.3326,
+      "epoch": 0.2221861822899773,
+      "grad_norm": 0.1857357621192932,
+      "learning_rate": 4.997270083045112e-05,
+      "loss": 0.4372,
       "step": 6165
     },
     {
-      "epoch": 0.22,
-      "learning_rate": 4.997641871742087e-05,
-      "loss": 0.3423,
+      "epoch": 0.22236638195120192,
+      "grad_norm": 0.18568672239780426,
+      "learning_rate": 4.997256432566116e-05,
+      "loss": 0.488,
       "step": 6170
     },
     {
-      "epoch": 0.22,
-      "learning_rate": 4.9976294853409e-05,
-      "loss": 0.3468,
+      "epoch": 0.22254658161242657,
+      "grad_norm": 0.19653025269508362,
+      "learning_rate": 4.997242748062485e-05,
+      "loss": 0.4907,
       "step": 6175
     },
     {
-      "epoch": 0.22,
-      "learning_rate": 4.9976170665096405e-05,
-      "loss": 0.3496,
+      "epoch": 0.2227267812736512,
+      "grad_norm": 0.19124481081962585,
+      "learning_rate": 4.9972290295344046e-05,
+      "loss": 0.5021,
       "step": 6180
     },
     {
-      "epoch": 0.22,
-      "learning_rate": 4.997604615248469e-05,
-      "loss": 0.3701,
+      "epoch": 0.22290698093487585,
+      "grad_norm": 0.20403480529785156,
+      "learning_rate": 4.997215276982062e-05,
+      "loss": 0.4855,
       "step": 6185
     },
     {
-      "epoch": 0.22,
-      "learning_rate": 4.997592131557548e-05,
-      "loss": 0.34,
+      "epoch": 0.22308718059610047,
+      "grad_norm": 0.18618778884410858,
+      "learning_rate": 4.9972014904056446e-05,
+      "loss": 0.4315,
       "step": 6190
     },
     {
-      "epoch": 0.22,
-      "learning_rate": 4.997579615437039e-05,
-      "loss": 0.3093,
+      "epoch": 0.22326738025732512,
+      "grad_norm": 0.24033483862876892,
+      "learning_rate": 4.997187669805341e-05,
+      "loss": 0.4607,
       "step": 6195
     },
     {
-      "epoch": 0.22,
-      "learning_rate": 4.997567066887105e-05,
-      "loss": 0.3141,
+      "epoch": 0.22344757991854974,
+      "grad_norm": 0.26656222343444824,
+      "learning_rate": 4.997173815181339e-05,
+      "loss": 0.5022,
       "step": 6200
     },
     {
-      "epoch": 0.22,
-      "learning_rate": 4.9975544859079074e-05,
-      "loss": 0.3394,
+      "epoch": 0.2236277795797744,
+      "grad_norm": 0.22280102968215942,
+      "learning_rate": 4.997159926533826e-05,
+      "loss": 0.4928,
       "step": 6205
     },
     {
-      "epoch": 0.22,
-      "learning_rate": 4.997541872499611e-05,
-      "loss": 0.3649,
+      "epoch": 0.22380797924099902,
+      "grad_norm": 0.2294902205467224,
+      "learning_rate": 4.997146003862994e-05,
+      "loss": 0.5015,
       "step": 6210
     },
     {
-      "epoch": 0.22,
-      "learning_rate": 4.997529226662381e-05,
-      "loss": 0.3132,
+      "epoch": 0.22398817890222367,
+      "grad_norm": 0.22047486901283264,
+      "learning_rate": 4.9971320471690295e-05,
+      "loss": 0.4545,
       "step": 6215
     },
     {
-      "epoch": 0.22,
-      "learning_rate": 4.997516548396378e-05,
-      "loss": 0.3583,
+      "epoch": 0.2241683785634483,
+      "grad_norm": 0.21225684881210327,
+      "learning_rate": 4.9971180564521254e-05,
+      "loss": 0.4758,
       "step": 6220
     },
     {
-      "epoch": 0.22,
-      "learning_rate": 4.9975038377017705e-05,
-      "loss": 0.3419,
+      "epoch": 0.22434857822467294,
+      "grad_norm": 0.18276849389076233,
+      "learning_rate": 4.9971040317124706e-05,
+      "loss": 0.4767,
       "step": 6225
     },
     {
-      "epoch": 0.22,
-      "learning_rate": 4.99749109457872e-05,
-      "loss": 0.3315,
+      "epoch": 0.22452877788589756,
+      "grad_norm": 0.19218437373638153,
+      "learning_rate": 4.9970899729502576e-05,
+      "loss": 0.466,
       "step": 6230
     },
     {
-      "epoch": 0.22,
-      "learning_rate": 4.997478319027394e-05,
-      "loss": 0.3451,
+      "epoch": 0.2247089775471222,
+      "grad_norm": 0.22989146411418915,
+      "learning_rate": 4.997075880165677e-05,
+      "loss": 0.4543,
       "step": 6235
     },
     {
-      "epoch": 0.22,
-      "learning_rate": 4.9974655110479576e-05,
-      "loss": 0.3336,
+      "epoch": 0.22488917720834684,
+      "grad_norm": 0.25233665108680725,
+      "learning_rate": 4.997061753358921e-05,
+      "loss": 0.501,
       "step": 6240
     },
     {
-      "epoch": 0.22,
-      "learning_rate": 4.997452670640578e-05,
-      "loss": 0.3494,
+      "epoch": 0.22506937686957149,
+      "grad_norm": 0.23017632961273193,
+      "learning_rate": 4.997047592530182e-05,
+      "loss": 0.463,
       "step": 6245
     },
     {
-      "epoch": 0.22,
-      "learning_rate": 4.997439797805421e-05,
-      "loss": 0.3194,
+      "epoch": 0.2252495765307961,
+      "grad_norm": 0.18937349319458008,
+      "learning_rate": 4.9970333976796526e-05,
+      "loss": 0.4387,
       "step": 6250
     },
     {
-      "epoch": 0.22,
-      "learning_rate": 4.997426892542654e-05,
-      "loss": 0.3357,
+      "epoch": 0.22542977619202076,
+      "grad_norm": 0.15060578286647797,
+      "learning_rate": 4.997019168807527e-05,
+      "loss": 0.4603,
       "step": 6255
     },
     {
-      "epoch": 0.22,
-      "learning_rate": 4.997413954852445e-05,
-      "loss": 0.3278,
+      "epoch": 0.2256099758532454,
+      "grad_norm": 0.2040947675704956,
+      "learning_rate": 4.997004905913998e-05,
+      "loss": 0.4982,
       "step": 6260
     },
     {
-      "epoch": 0.22,
-      "learning_rate": 4.997400984734961e-05,
-      "loss": 0.3397,
+      "epoch": 0.22579017551447003,
+      "grad_norm": 0.2734396159648895,
+      "learning_rate": 4.9969906089992616e-05,
+      "loss": 0.4854,
       "step": 6265
     },
     {
-      "epoch": 0.22,
-      "learning_rate": 4.997387982190372e-05,
-      "loss": 0.3387,
+      "epoch": 0.22597037517569468,
+      "grad_norm": 0.17175772786140442,
+      "learning_rate": 4.996976278063511e-05,
+      "loss": 0.4451,
       "step": 6270
     },
     {
-      "epoch": 0.22,
-      "learning_rate": 4.997374947218845e-05,
-      "loss": 0.3167,
+      "epoch": 0.2261505748369193,
+      "grad_norm": 0.21332839131355286,
+      "learning_rate": 4.996961913106942e-05,
+      "loss": 0.4869,
       "step": 6275
     },
     {
-      "epoch": 0.22,
-      "learning_rate": 4.997361879820551e-05,
-      "loss": 0.2976,
+      "epoch": 0.22633077449814396,
+      "grad_norm": 0.16605186462402344,
+      "learning_rate": 4.9969475141297504e-05,
+      "loss": 0.4448,
       "step": 6280
     },
     {
-      "epoch": 0.22,
-      "learning_rate": 4.997348779995659e-05,
-      "loss": 0.3262,
+      "epoch": 0.22651097415936858,
+      "grad_norm": 0.18335530161857605,
+      "learning_rate": 4.9969330811321325e-05,
+      "loss": 0.4778,
       "step": 6285
     },
     {
-      "epoch": 0.22,
-      "learning_rate": 4.997335647744338e-05,
-      "loss": 0.3545,
+      "epoch": 0.22669117382059323,
+      "grad_norm": 0.20451930165290833,
+      "learning_rate": 4.996918614114285e-05,
+      "loss": 0.4643,
       "step": 6290
     },
     {
-      "epoch": 0.22,
-      "learning_rate": 4.997322483066759e-05,
-      "loss": 0.3226,
+      "epoch": 0.22687137348181785,
+      "grad_norm": 0.1627047061920166,
+      "learning_rate": 4.9969041130764046e-05,
+      "loss": 0.4595,
       "step": 6295
     },
     {
-      "epoch": 0.22,
-      "learning_rate": 4.9973092859630945e-05,
-      "loss": 0.35,
+      "epoch": 0.2270515731430425,
+      "grad_norm": 0.18496522307395935,
+      "learning_rate": 4.9968895780186884e-05,
+      "loss": 0.4552,
       "step": 6300
     },
     {
-      "epoch": 0.22,
-      "learning_rate": 4.997296056433514e-05,
-      "loss": 0.3216,
+      "epoch": 0.22723177280426712,
+      "grad_norm": 0.209730327129364,
+      "learning_rate": 4.9968750089413365e-05,
+      "loss": 0.4515,
       "step": 6305
     },
     {
-      "epoch": 0.22,
-      "learning_rate": 4.997282794478191e-05,
-      "loss": 0.3178,
+      "epoch": 0.22741197246549177,
+      "grad_norm": 0.21571126580238342,
+      "learning_rate": 4.996860405844545e-05,
+      "loss": 0.4874,
       "step": 6310
     },
     {
-      "epoch": 0.22,
-      "learning_rate": 4.997269500097296e-05,
-      "loss": 0.3279,
+      "epoch": 0.2275921721267164,
+      "grad_norm": 0.20048457384109497,
+      "learning_rate": 4.996845768728514e-05,
+      "loss": 0.4921,
       "step": 6315
     },
     {
-      "epoch": 0.22,
-      "learning_rate": 4.997256173291002e-05,
-      "loss": 0.3349,
+      "epoch": 0.22777237178794105,
+      "grad_norm": 0.211773082613945,
+      "learning_rate": 4.996831097593443e-05,
+      "loss": 0.4907,
       "step": 6320
     },
     {
-      "epoch": 0.22,
-      "learning_rate": 4.9972428140594824e-05,
-      "loss": 0.3049,
+      "epoch": 0.22795257144916567,
+      "grad_norm": 0.22399649024009705,
+      "learning_rate": 4.996816392439532e-05,
+      "loss": 0.4853,
       "step": 6325
     },
     {
-      "epoch": 0.22,
-      "learning_rate": 4.99722942240291e-05,
-      "loss": 0.3463,
+      "epoch": 0.22813277111039032,
+      "grad_norm": 0.2886163294315338,
+      "learning_rate": 4.9968016532669805e-05,
+      "loss": 0.4592,
       "step": 6330
     },
     {
-      "epoch": 0.22,
-      "learning_rate": 4.9972159983214605e-05,
-      "loss": 0.3329,
+      "epoch": 0.22831297077161494,
+      "grad_norm": 0.21131619811058044,
+      "learning_rate": 4.9967868800759895e-05,
+      "loss": 0.4624,
       "step": 6335
     },
     {
-      "epoch": 0.22,
-      "learning_rate": 4.9972025418153065e-05,
-      "loss": 0.3367,
+      "epoch": 0.2284931704328396,
+      "grad_norm": 0.2376997023820877,
+      "learning_rate": 4.996772072866762e-05,
+      "loss": 0.4935,
       "step": 6340
     },
     {
-      "epoch": 0.22,
-      "learning_rate": 4.9971890528846235e-05,
-      "loss": 0.337,
+      "epoch": 0.22867337009406422,
+      "grad_norm": 0.24256786704063416,
+      "learning_rate": 4.996757231639497e-05,
+      "loss": 0.4595,
       "step": 6345
     },
     {
-      "epoch": 0.22,
-      "learning_rate": 4.997175531529586e-05,
-      "loss": 0.3383,
+      "epoch": 0.22885356975528887,
+      "grad_norm": 0.22123675048351288,
+      "learning_rate": 4.9967423563943994e-05,
+      "loss": 0.4759,
       "step": 6350
     },
     {
-      "epoch": 0.22,
-      "learning_rate": 4.9971619777503705e-05,
-      "loss": 0.3563,
+      "epoch": 0.2290337694165135,
+      "grad_norm": 0.22762177884578705,
+      "learning_rate": 4.996727447131669e-05,
+      "loss": 0.4417,
       "step": 6355
     },
     {
-      "epoch": 0.22,
-      "learning_rate": 4.9971483915471516e-05,
-      "loss": 0.3604,
+      "epoch": 0.22921396907773814,
+      "grad_norm": 0.1603131741285324,
+      "learning_rate": 4.9967125038515116e-05,
+      "loss": 0.4618,
       "step": 6360
     },
     {
-      "epoch": 0.22,
-      "learning_rate": 4.997134772920107e-05,
-      "loss": 0.3429,
+      "epoch": 0.22939416873896276,
+      "grad_norm": 0.2037096470594406,
+      "learning_rate": 4.99669752655413e-05,
+      "loss": 0.4785,
       "step": 6365
     },
     {
-      "epoch": 0.22,
-      "learning_rate": 4.997121121869414e-05,
-      "loss": 0.3422,
+      "epoch": 0.2295743684001874,
+      "grad_norm": 0.25867146253585815,
+      "learning_rate": 4.996682515239728e-05,
+      "loss": 0.451,
       "step": 6370
     },
     {
-      "epoch": 0.22,
-      "learning_rate": 4.997107438395247e-05,
-      "loss": 0.3272,
+      "epoch": 0.22975456806141203,
+      "grad_norm": 0.15876412391662598,
+      "learning_rate": 4.996667469908509e-05,
+      "loss": 0.4362,
       "step": 6375
     },
     {
-      "epoch": 0.22,
-      "learning_rate": 4.997093722497787e-05,
-      "loss": 0.3286,
+      "epoch": 0.22993476772263668,
+      "grad_norm": 0.22342482209205627,
+      "learning_rate": 4.99665239056068e-05,
+      "loss": 0.4629,
       "step": 6380
     },
     {
-      "epoch": 0.22,
-      "learning_rate": 4.99707997417721e-05,
-      "loss": 0.3103,
+      "epoch": 0.2301149673838613,
+      "grad_norm": 0.1261013001203537,
+      "learning_rate": 4.9966372771964456e-05,
+      "loss": 0.4501,
       "step": 6385
     },
     {
-      "epoch": 0.22,
-      "learning_rate": 4.997066193433695e-05,
-      "loss": 0.3118,
+      "epoch": 0.23029516704508596,
+      "grad_norm": 0.20957724750041962,
+      "learning_rate": 4.996622129816011e-05,
+      "loss": 0.4736,
       "step": 6390
     },
     {
-      "epoch": 0.22,
-      "learning_rate": 4.997052380267421e-05,
-      "loss": 0.3341,
+      "epoch": 0.23047536670631058,
+      "grad_norm": 0.23294906318187714,
+      "learning_rate": 4.996606948419583e-05,
+      "loss": 0.4565,
       "step": 6395
     },
     {
-      "epoch": 0.23,
-      "learning_rate": 4.997038534678568e-05,
-      "loss": 0.35,
+      "epoch": 0.23065556636753523,
+      "grad_norm": 0.18201923370361328,
+      "learning_rate": 4.99659173300737e-05,
+      "loss": 0.5131,
       "step": 6400
     },
     {
-      "epoch": 0.23,
-      "learning_rate": 4.997024656667315e-05,
-      "loss": 0.3533,
+      "epoch": 0.23083576602875985,
+      "grad_norm": 0.2046813815832138,
+      "learning_rate": 4.996576483579577e-05,
+      "loss": 0.4632,
       "step": 6405
     },
     {
-      "epoch": 0.23,
-      "learning_rate": 4.997010746233843e-05,
-      "loss": 0.3333,
+      "epoch": 0.2310159656899845,
+      "grad_norm": 0.17519819736480713,
+      "learning_rate": 4.9965612001364124e-05,
+      "loss": 0.5046,
       "step": 6410
     },
     {
-      "epoch": 0.23,
-      "learning_rate": 4.996996803378331e-05,
-      "loss": 0.336,
+      "epoch": 0.23119616535120913,
+      "grad_norm": 0.19435188174247742,
+      "learning_rate": 4.996545882678086e-05,
+      "loss": 0.4378,
       "step": 6415
     },
     {
-      "epoch": 0.23,
-      "learning_rate": 4.996982828100962e-05,
-      "loss": 0.3255,
+      "epoch": 0.23137636501243378,
+      "grad_norm": 0.2328794300556183,
+      "learning_rate": 4.9965305312048043e-05,
+      "loss": 0.4884,
       "step": 6420
     },
     {
-      "epoch": 0.23,
-      "learning_rate": 4.996968820401916e-05,
-      "loss": 0.3319,
+      "epoch": 0.23155656467365843,
+      "grad_norm": 0.1869356334209442,
+      "learning_rate": 4.996515145716778e-05,
+      "loss": 0.4929,
       "step": 6425
     },
     {
-      "epoch": 0.23,
-      "learning_rate": 4.996954780281375e-05,
-      "loss": 0.358,
+      "epoch": 0.23173676433488305,
+      "grad_norm": 0.18540406227111816,
+      "learning_rate": 4.996499726214216e-05,
+      "loss": 0.4421,
       "step": 6430
     },
     {
-      "epoch": 0.23,
-      "learning_rate": 4.996940707739523e-05,
-      "loss": 0.3364,
+      "epoch": 0.2319169639961077,
+      "grad_norm": 0.23471662402153015,
+      "learning_rate": 4.9964842726973286e-05,
+      "loss": 0.5024,
       "step": 6435
     },
     {
-      "epoch": 0.23,
-      "learning_rate": 4.9969266027765404e-05,
-      "loss": 0.3389,
+      "epoch": 0.23209716365733232,
+      "grad_norm": 0.1900070160627365,
+      "learning_rate": 4.996468785166326e-05,
+      "loss": 0.4652,
       "step": 6440
     },
     {
-      "epoch": 0.23,
-      "learning_rate": 4.9969124653926127e-05,
-      "loss": 0.3334,
+      "epoch": 0.23227736331855697,
+      "grad_norm": 0.23058246076107025,
+      "learning_rate": 4.99645326362142e-05,
+      "loss": 0.5072,
       "step": 6445
     },
     {
-      "epoch": 0.23,
-      "learning_rate": 4.996898295587921e-05,
-      "loss": 0.344,
+      "epoch": 0.2324575629797816,
+      "grad_norm": 0.20412348210811615,
+      "learning_rate": 4.9964377080628215e-05,
+      "loss": 0.4847,
       "step": 6450
     },
     {
-      "epoch": 0.23,
-      "learning_rate": 4.9968840933626506e-05,
-      "loss": 0.3552,
+      "epoch": 0.23263776264100625,
+      "grad_norm": 0.20064546167850494,
+      "learning_rate": 4.9964221184907424e-05,
+      "loss": 0.4501,
       "step": 6455
     },
     {
-      "epoch": 0.23,
-      "learning_rate": 4.996869858716986e-05,
-      "loss": 0.3372,
+      "epoch": 0.23281796230223087,
+      "grad_norm": 0.16838903725147247,
+      "learning_rate": 4.996406494905396e-05,
+      "loss": 0.4665,
       "step": 6460
     },
     {
-      "epoch": 0.23,
-      "learning_rate": 4.9968555916511114e-05,
-      "loss": 0.3244,
+      "epoch": 0.23299816196345552,
+      "grad_norm": 0.17399536073207855,
+      "learning_rate": 4.9963908373069935e-05,
+      "loss": 0.4616,
       "step": 6465
     },
     {
-      "epoch": 0.23,
-      "learning_rate": 4.996841292165213e-05,
-      "loss": 0.3458,
+      "epoch": 0.23317836162468014,
+      "grad_norm": 0.18575124442577362,
+      "learning_rate": 4.9963751456957494e-05,
+      "loss": 0.4336,
       "step": 6470
     },
     {
-      "epoch": 0.23,
-      "learning_rate": 4.996826960259475e-05,
-      "loss": 0.3342,
+      "epoch": 0.2333585612859048,
+      "grad_norm": 0.22949109971523285,
+      "learning_rate": 4.996359420071877e-05,
+      "loss": 0.5,
       "step": 6475
     },
     {
-      "epoch": 0.23,
-      "learning_rate": 4.996812595934085e-05,
-      "loss": 0.3492,
+      "epoch": 0.23353876094712941,
+      "grad_norm": 0.20516782999038696,
+      "learning_rate": 4.9963436604355916e-05,
+      "loss": 0.4611,
       "step": 6480
     },
     {
-      "epoch": 0.23,
-      "learning_rate": 4.996798199189229e-05,
-      "loss": 0.343,
+      "epoch": 0.23371896060835406,
+      "grad_norm": 0.2293572872877121,
+      "learning_rate": 4.996327866787106e-05,
+      "loss": 0.501,
       "step": 6485
     },
     {
-      "epoch": 0.23,
-      "learning_rate": 4.996783770025094e-05,
-      "loss": 0.3175,
+      "epoch": 0.2338991602695787,
+      "grad_norm": 0.18813565373420715,
+      "learning_rate": 4.9963120391266376e-05,
+      "loss": 0.4347,
       "step": 6490
     },
     {
-      "epoch": 0.23,
-      "learning_rate": 4.9967693084418665e-05,
-      "loss": 0.3195,
+      "epoch": 0.23407935993080334,
+      "grad_norm": 0.2553117275238037,
+      "learning_rate": 4.9962961774544006e-05,
+      "loss": 0.4794,
       "step": 6495
     },
     {
-      "epoch": 0.23,
-      "learning_rate": 4.9967548144397356e-05,
-      "loss": 0.3236,
+      "epoch": 0.23425955959202796,
+      "grad_norm": 0.20878198742866516,
+      "learning_rate": 4.996280281770611e-05,
+      "loss": 0.4326,
       "step": 6500
     },
     {
-      "epoch": 0.23,
-      "eval_loss": 0.3339012861251831,
-      "eval_runtime": 10.5429,
-      "eval_samples_per_second": 9.485,
-      "eval_steps_per_second": 9.485,
+      "epoch": 0.23425955959202796,
+      "eval_loss": 0.4936707615852356,
+      "eval_runtime": 3.5146,
+      "eval_samples_per_second": 28.453,
+      "eval_steps_per_second": 7.113,
       "step": 6500
     },
     {
-      "epoch": 0.23,
-      "learning_rate": 4.9967402880188884e-05,
-      "loss": 0.3493,
+      "epoch": 0.2344397592532526,
+      "grad_norm": 0.20102368295192719,
+      "learning_rate": 4.996264352075487e-05,
+      "loss": 0.4871,
       "step": 6505
     },
     {
-      "epoch": 0.23,
-      "learning_rate": 4.9967257291795147e-05,
-      "loss": 0.3457,
+      "epoch": 0.23461995891447723,
+      "grad_norm": 0.20165607333183289,
+      "learning_rate": 4.996248388369243e-05,
+      "loss": 0.4737,
       "step": 6510
     },
     {
-      "epoch": 0.23,
-      "learning_rate": 4.996711137921802e-05,
-      "loss": 0.3331,
+      "epoch": 0.23480015857570188,
+      "grad_norm": 0.1898958534002304,
+      "learning_rate": 4.996232390652099e-05,
+      "loss": 0.4562,
       "step": 6515
     },
     {
-      "epoch": 0.23,
-      "learning_rate": 4.996696514245941e-05,
-      "loss": 0.3379,
+      "epoch": 0.2349803582369265,
+      "grad_norm": 0.2387424260377884,
+      "learning_rate": 4.996216358924272e-05,
+      "loss": 0.4935,
       "step": 6520
     },
     {
-      "epoch": 0.23,
-      "learning_rate": 4.9966818581521214e-05,
-      "loss": 0.3319,
+      "epoch": 0.23516055789815116,
+      "grad_norm": 0.2199474722146988,
+      "learning_rate": 4.99620029318598e-05,
+      "loss": 0.4808,
       "step": 6525
     },
     {
-      "epoch": 0.23,
-      "learning_rate": 4.996667169640533e-05,
-      "loss": 0.3487,
+      "epoch": 0.23534075755937578,
+      "grad_norm": 0.20181065797805786,
+      "learning_rate": 4.996184193437442e-05,
+      "loss": 0.4586,
       "step": 6530
     },
     {
-      "epoch": 0.23,
-      "learning_rate": 4.996652448711366e-05,
-      "loss": 0.3356,
+      "epoch": 0.23552095722060043,
+      "grad_norm": 0.21171115338802338,
+      "learning_rate": 4.9961680596788784e-05,
+      "loss": 0.451,
       "step": 6535
     },
     {
-      "epoch": 0.23,
-      "learning_rate": 4.996637695364813e-05,
-      "loss": 0.3379,
+      "epoch": 0.23570115688182505,
+      "grad_norm": 0.22186800837516785,
+      "learning_rate": 4.996151891910508e-05,
+      "loss": 0.4912,
       "step": 6540
     },
     {
-      "epoch": 0.23,
-      "learning_rate": 4.996622909601065e-05,
-      "loss": 0.3411,
+      "epoch": 0.2358813565430497,
+      "grad_norm": 0.20023906230926514,
+      "learning_rate": 4.9961356901325515e-05,
+      "loss": 0.481,
       "step": 6545
     },
     {
-      "epoch": 0.23,
-      "learning_rate": 4.996608091420314e-05,
-      "loss": 0.323,
+      "epoch": 0.23606155620427433,
+      "grad_norm": 0.2267458438873291,
+      "learning_rate": 4.9961194543452296e-05,
+      "loss": 0.4896,
       "step": 6550
     },
     {
-      "epoch": 0.23,
-      "learning_rate": 4.996593240822752e-05,
-      "loss": 0.3342,
+      "epoch": 0.23624175586549898,
+      "grad_norm": 0.1733904927968979,
+      "learning_rate": 4.996103184548763e-05,
+      "loss": 0.4545,
       "step": 6555
     },
     {
-      "epoch": 0.23,
-      "learning_rate": 4.996578357808572e-05,
-      "loss": 0.336,
+      "epoch": 0.2364219555267236,
+      "grad_norm": 0.1978592723608017,
+      "learning_rate": 4.9960868807433734e-05,
+      "loss": 0.4491,
       "step": 6560
     },
     {
-      "epoch": 0.23,
-      "learning_rate": 4.996563442377967e-05,
-      "loss": 0.3418,
+      "epoch": 0.23660215518794825,
+      "grad_norm": 0.1597129851579666,
+      "learning_rate": 4.9960705429292836e-05,
+      "loss": 0.4196,
       "step": 6565
     },
     {
-      "epoch": 0.23,
-      "learning_rate": 4.996548494531131e-05,
-      "loss": 0.3527,
+      "epoch": 0.23678235484917287,
+      "grad_norm": 0.19397877156734467,
+      "learning_rate": 4.996054171106716e-05,
+      "loss": 0.4751,
       "step": 6570
     },
     {
-      "epoch": 0.23,
-      "learning_rate": 4.996533514268259e-05,
-      "loss": 0.3421,
+      "epoch": 0.23696255451039752,
+      "grad_norm": 0.20515479147434235,
+      "learning_rate": 4.996037765275894e-05,
+      "loss": 0.4897,
       "step": 6575
     },
     {
-      "epoch": 0.23,
-      "learning_rate": 4.996518501589545e-05,
-      "loss": 0.3535,
+      "epoch": 0.23714275417162214,
+      "grad_norm": 0.240402951836586,
+      "learning_rate": 4.99602132543704e-05,
+      "loss": 0.4842,
       "step": 6580
     },
     {
-      "epoch": 0.23,
-      "learning_rate": 4.996503456495182e-05,
-      "loss": 0.3448,
+      "epoch": 0.2373229538328468,
+      "grad_norm": 0.20333066582679749,
+      "learning_rate": 4.996004851590379e-05,
+      "loss": 0.4708,
       "step": 6585
     },
     {
-      "epoch": 0.23,
-      "learning_rate": 4.996488378985367e-05,
-      "loss": 0.3458,
+      "epoch": 0.23750315349407142,
+      "grad_norm": 0.2151239812374115,
+      "learning_rate": 4.995988343736135e-05,
+      "loss": 0.4748,
       "step": 6590
     },
     {
-      "epoch": 0.23,
-      "learning_rate": 4.9964732690602964e-05,
-      "loss": 0.3429,
+      "epoch": 0.23768335315529607,
+      "grad_norm": 0.1680813729763031,
+      "learning_rate": 4.995971801874533e-05,
+      "loss": 0.4726,
       "step": 6595
     },
     {
-      "epoch": 0.23,
-      "learning_rate": 4.9964581267201646e-05,
-      "loss": 0.3458,
+      "epoch": 0.23786355281652072,
+      "grad_norm": 0.18834345042705536,
+      "learning_rate": 4.995955226005799e-05,
+      "loss": 0.4443,
       "step": 6600
     },
     {
-      "epoch": 0.23,
-      "learning_rate": 4.99644295196517e-05,
-      "loss": 0.3492,
+      "epoch": 0.23804375247774534,
+      "grad_norm": 0.21085232496261597,
+      "learning_rate": 4.995938616130158e-05,
+      "loss": 0.4451,
       "step": 6605
     },
     {
-      "epoch": 0.23,
-      "learning_rate": 4.996427744795509e-05,
-      "loss": 0.3352,
+      "epoch": 0.23822395213897,
+      "grad_norm": 0.20523454248905182,
+      "learning_rate": 4.9959219722478365e-05,
+      "loss": 0.4476,
       "step": 6610
     },
     {
-      "epoch": 0.23,
-      "learning_rate": 4.996412505211378e-05,
-      "loss": 0.3193,
+      "epoch": 0.2384041518001946,
+      "grad_norm": 0.20655976235866547,
+      "learning_rate": 4.995908632657321e-05,
+      "loss": 0.445,
       "step": 6615
     },
     {
-      "epoch": 0.23,
-      "learning_rate": 4.996397233212978e-05,
-      "loss": 0.3528,
+      "epoch": 0.23858435146141926,
+      "grad_norm": 0.1544501781463623,
+      "learning_rate": 4.995891927563548e-05,
+      "loss": 0.454,
       "step": 6620
     },
     {
-      "epoch": 0.23,
-      "learning_rate": 4.996381928800503e-05,
-      "loss": 0.3292,
+      "epoch": 0.2387645511226439,
+      "grad_norm": 0.20439325273036957,
+      "learning_rate": 4.9958751884637286e-05,
+      "loss": 0.4784,
       "step": 6625
     },
     {
-      "epoch": 0.23,
-      "learning_rate": 4.996366591974155e-05,
-      "loss": 0.3336,
+      "epoch": 0.23894475078386854,
+      "grad_norm": 0.15891632437705994,
+      "learning_rate": 4.9958584153580933e-05,
+      "loss": 0.4433,
       "step": 6630
     },
     {
-      "epoch": 0.23,
-      "learning_rate": 4.9963512227341315e-05,
-      "loss": 0.3314,
+      "epoch": 0.23912495044509316,
+      "grad_norm": 0.20959921181201935,
+      "learning_rate": 4.995841608246871e-05,
+      "loss": 0.4862,
       "step": 6635
     },
     {
-      "epoch": 0.23,
-      "learning_rate": 4.996335821080632e-05,
-      "loss": 0.3274,
+      "epoch": 0.2393051501063178,
+      "grad_norm": 0.17294225096702576,
+      "learning_rate": 4.995824767130289e-05,
+      "loss": 0.436,
       "step": 6640
     },
     {
-      "epoch": 0.23,
-      "learning_rate": 4.996320387013858e-05,
-      "loss": 0.3435,
+      "epoch": 0.23948534976754243,
+      "grad_norm": 0.26996639370918274,
+      "learning_rate": 4.995807892008578e-05,
+      "loss": 0.4688,
       "step": 6645
     },
     {
-      "epoch": 0.23,
-      "learning_rate": 4.996304920534008e-05,
-      "loss": 0.3747,
+      "epoch": 0.23966554942876708,
+      "grad_norm": 0.20224200189113617,
+      "learning_rate": 4.995790982881968e-05,
+      "loss": 0.4688,
       "step": 6650
     },
     {
-      "epoch": 0.23,
-      "learning_rate": 4.9962894216412845e-05,
-      "loss": 0.3473,
+      "epoch": 0.2398457490899917,
+      "grad_norm": 0.17374387383460999,
+      "learning_rate": 4.995774039750689e-05,
+      "loss": 0.4692,
       "step": 6655
     },
     {
-      "epoch": 0.23,
-      "learning_rate": 4.996273890335888e-05,
-      "loss": 0.32,
+      "epoch": 0.24002594875121636,
+      "grad_norm": 0.14203490316867828,
+      "learning_rate": 4.995757062614972e-05,
+      "loss": 0.4516,
       "step": 6660
     },
     {
-      "epoch": 0.23,
-      "learning_rate": 4.9962583266180194e-05,
-      "loss": 0.3586,
+      "epoch": 0.24020614841244098,
+      "grad_norm": 0.19781635701656342,
+      "learning_rate": 4.9957400514750484e-05,
+      "loss": 0.5077,
       "step": 6665
     },
     {
-      "epoch": 0.23,
-      "learning_rate": 4.9962427304878825e-05,
-      "loss": 0.3452,
+      "epoch": 0.24038634807366563,
+      "grad_norm": 0.2605617344379425,
+      "learning_rate": 4.99572300633115e-05,
+      "loss": 0.4947,
       "step": 6670
     },
     {
-      "epoch": 0.23,
-      "learning_rate": 4.996227101945678e-05,
-      "loss": 0.3305,
+      "epoch": 0.24056654773489025,
+      "grad_norm": 0.22600892186164856,
+      "learning_rate": 4.995705927183508e-05,
+      "loss": 0.4973,
       "step": 6675
     },
     {
-      "epoch": 0.24,
-      "learning_rate": 4.9962114409916105e-05,
-      "loss": 0.2937,
+      "epoch": 0.2407467473961149,
+      "grad_norm": 0.1905760020017624,
+      "learning_rate": 4.995688814032357e-05,
+      "loss": 0.5074,
       "step": 6680
     },
     {
-      "epoch": 0.24,
-      "learning_rate": 4.996195747625881e-05,
-      "loss": 0.3321,
+      "epoch": 0.24092694705733952,
+      "grad_norm": 0.19111768901348114,
+      "learning_rate": 4.995671666877928e-05,
+      "loss": 0.4408,
       "step": 6685
     },
     {
-      "epoch": 0.24,
-      "learning_rate": 4.9961800218486965e-05,
-      "loss": 0.3408,
+      "epoch": 0.24110714671856417,
+      "grad_norm": 0.16061218082904816,
+      "learning_rate": 4.995654485720456e-05,
+      "loss": 0.4824,
       "step": 6690
     },
     {
-      "epoch": 0.24,
-      "learning_rate": 4.996164263660258e-05,
-      "loss": 0.3322,
+      "epoch": 0.2412873463797888,
+      "grad_norm": 0.21395018696784973,
+      "learning_rate": 4.9956372705601754e-05,
+      "loss": 0.4909,
       "step": 6695
     },
     {
-      "epoch": 0.24,
-      "learning_rate": 4.996148473060773e-05,
-      "loss": 0.3353,
+      "epoch": 0.24146754604101345,
+      "grad_norm": 0.18952535092830658,
+      "learning_rate": 4.9956200213973195e-05,
+      "loss": 0.476,
       "step": 6700
     },
     {
-      "epoch": 0.24,
-      "learning_rate": 4.996132650050444e-05,
-      "loss": 0.3284,
+      "epoch": 0.24164774570223807,
+      "grad_norm": 0.23430414497852325,
+      "learning_rate": 4.9956027382321244e-05,
+      "loss": 0.4684,
       "step": 6705
     },
     {
-      "epoch": 0.24,
-      "learning_rate": 4.996116794629478e-05,
-      "loss": 0.3554,
+      "epoch": 0.24182794536346272,
+      "grad_norm": 0.23189213871955872,
+      "learning_rate": 4.9955854210648246e-05,
+      "loss": 0.4627,
       "step": 6710
     },
     {
-      "epoch": 0.24,
-      "learning_rate": 4.99610090679808e-05,
-      "loss": 0.34,
+      "epoch": 0.24200814502468734,
+      "grad_norm": 0.19272887706756592,
+      "learning_rate": 4.995568069895657e-05,
+      "loss": 0.4993,
       "step": 6715
     },
     {
-      "epoch": 0.24,
-      "learning_rate": 4.9960849865564576e-05,
-      "loss": 0.3264,
+      "epoch": 0.242188344685912,
+      "grad_norm": 0.24815817177295685,
+      "learning_rate": 4.995550684724858e-05,
+      "loss": 0.4526,
       "step": 6720
     },
     {
-      "epoch": 0.24,
-      "learning_rate": 4.996069033904816e-05,
-      "loss": 0.3286,
+      "epoch": 0.24236854434713662,
+      "grad_norm": 0.2101772576570511,
+      "learning_rate": 4.995533265552663e-05,
+      "loss": 0.4711,
       "step": 6725
     },
     {
-      "epoch": 0.24,
-      "learning_rate": 4.9960530488433634e-05,
-      "loss": 0.3279,
+      "epoch": 0.24254874400836127,
+      "grad_norm": 0.18738162517547607,
+      "learning_rate": 4.995515812379311e-05,
+      "loss": 0.4603,
       "step": 6730
     },
     {
-      "epoch": 0.24,
-      "learning_rate": 4.996037031372306e-05,
-      "loss": 0.3283,
+      "epoch": 0.2427289436695859,
+      "grad_norm": 0.20007184147834778,
+      "learning_rate": 4.9954983252050393e-05,
+      "loss": 0.4957,
       "step": 6735
     },
     {
-      "epoch": 0.24,
-      "learning_rate": 4.9960209814918534e-05,
-      "loss": 0.3508,
+      "epoch": 0.24290914333081054,
+      "grad_norm": 0.22411781549453735,
+      "learning_rate": 4.995480804030086e-05,
+      "loss": 0.4503,
       "step": 6740
     },
     {
-      "epoch": 0.24,
-      "learning_rate": 4.996004899202213e-05,
-      "loss": 0.3734,
+      "epoch": 0.24308934299203516,
+      "grad_norm": 0.19585049152374268,
+      "learning_rate": 4.99546324885469e-05,
+      "loss": 0.4572,
       "step": 6745
     },
     {
-      "epoch": 0.24,
-      "learning_rate": 4.995988784503595e-05,
-      "loss": 0.3299,
+      "epoch": 0.2432695426532598,
+      "grad_norm": 0.1979917585849762,
+      "learning_rate": 4.99544565967909e-05,
+      "loss": 0.436,
       "step": 6750
     },
     {
-      "epoch": 0.24,
-      "learning_rate": 4.995972637396206e-05,
-      "loss": 0.3035,
+      "epoch": 0.24344974231448444,
+      "grad_norm": 0.16787096858024597,
+      "learning_rate": 4.995428036503527e-05,
+      "loss": 0.4745,
       "step": 6755
     },
     {
-      "epoch": 0.24,
-      "learning_rate": 4.995956457880257e-05,
-      "loss": 0.3312,
+      "epoch": 0.24362994197570909,
+      "grad_norm": 0.1978481262922287,
+      "learning_rate": 4.995410379328239e-05,
+      "loss": 0.4553,
       "step": 6760
     },
     {
-      "epoch": 0.24,
-      "learning_rate": 4.99594024595596e-05,
-      "loss": 0.319,
+      "epoch": 0.24381014163693374,
+      "grad_norm": 0.22523565590381622,
+      "learning_rate": 4.995392688153468e-05,
+      "loss": 0.4355,
       "step": 6765
     },
     {
-      "epoch": 0.24,
-      "learning_rate": 4.9959240016235233e-05,
-      "loss": 0.3293,
+      "epoch": 0.24399034129815836,
+      "grad_norm": 0.16378116607666016,
+      "learning_rate": 4.995374962979455e-05,
+      "loss": 0.4136,
       "step": 6770
     },
     {
-      "epoch": 0.24,
-      "learning_rate": 4.995907724883158e-05,
-      "loss": 0.3427,
+      "epoch": 0.244170540959383,
+      "grad_norm": 0.17417633533477783,
+      "learning_rate": 4.995357203806441e-05,
+      "loss": 0.4231,
       "step": 6775
     },
     {
-      "epoch": 0.24,
-      "learning_rate": 4.995891415735075e-05,
-      "loss": 0.3499,
+      "epoch": 0.24435074062060763,
+      "grad_norm": 0.20466962456703186,
+      "learning_rate": 4.9953394106346686e-05,
+      "loss": 0.5051,
       "step": 6780
     },
     {
-      "epoch": 0.24,
-      "learning_rate": 4.995875074179488e-05,
-      "loss": 0.355,
+      "epoch": 0.24453094028183228,
+      "grad_norm": 0.1599748432636261,
+      "learning_rate": 4.99532158346438e-05,
+      "loss": 0.4419,
       "step": 6785
     },
     {
-      "epoch": 0.24,
-      "learning_rate": 4.995858700216607e-05,
-      "loss": 0.3269,
+      "epoch": 0.2447111399430569,
+      "grad_norm": 0.18782804906368256,
+      "learning_rate": 4.995303722295816e-05,
+      "loss": 0.4459,
       "step": 6790
     },
     {
-      "epoch": 0.24,
-      "learning_rate": 4.995842293846647e-05,
-      "loss": 0.3251,
+      "epoch": 0.24489133960428155,
+      "grad_norm": 0.19041045010089874,
+      "learning_rate": 4.995285827129224e-05,
+      "loss": 0.458,
       "step": 6795
     },
     {
-      "epoch": 0.24,
-      "learning_rate": 4.995825855069818e-05,
-      "loss": 0.3315,
+      "epoch": 0.24507153926550618,
+      "grad_norm": 0.2037457972764969,
+      "learning_rate": 4.995267897964845e-05,
+      "loss": 0.4666,
       "step": 6800
     },
     {
-      "epoch": 0.24,
-      "learning_rate": 4.995809383886336e-05,
-      "loss": 0.3505,
+      "epoch": 0.24525173892673083,
+      "grad_norm": 0.2316879779100418,
+      "learning_rate": 4.995249934802925e-05,
+      "loss": 0.4435,
       "step": 6805
     },
     {
-      "epoch": 0.24,
-      "learning_rate": 4.9957928802964135e-05,
-      "loss": 0.3231,
+      "epoch": 0.24543193858795545,
+      "grad_norm": 0.19015929102897644,
+      "learning_rate": 4.995231937643706e-05,
+      "loss": 0.4936,
       "step": 6810
     },
     {
-      "epoch": 0.24,
-      "learning_rate": 4.995776344300266e-05,
-      "loss": 0.3082,
+      "epoch": 0.2456121382491801,
+      "grad_norm": 0.17666086554527283,
+      "learning_rate": 4.995213906487436e-05,
+      "loss": 0.443,
       "step": 6815
     },
     {
-      "epoch": 0.24,
-      "learning_rate": 4.995759775898107e-05,
-      "loss": 0.3222,
+      "epoch": 0.24579233791040472,
+      "grad_norm": 0.17423413693904877,
+      "learning_rate": 4.995195841334359e-05,
+      "loss": 0.4887,
       "step": 6820
     },
     {
-      "epoch": 0.24,
-      "learning_rate": 4.995743175090152e-05,
-      "loss": 0.327,
+      "epoch": 0.24597253757162937,
+      "grad_norm": 0.19055670499801636,
+      "learning_rate": 4.9951777421847225e-05,
+      "loss": 0.4609,
       "step": 6825
     },
     {
-      "epoch": 0.24,
-      "learning_rate": 4.9957265418766165e-05,
-      "loss": 0.3357,
+      "epoch": 0.246152737232854,
+      "grad_norm": 0.21230706572532654,
+      "learning_rate": 4.995159609038772e-05,
+      "loss": 0.4283,
       "step": 6830
     },
     {
-      "epoch": 0.24,
-      "learning_rate": 4.9957098762577174e-05,
-      "loss": 0.3527,
+      "epoch": 0.24633293689407865,
+      "grad_norm": 0.20157566666603088,
+      "learning_rate": 4.995141441896754e-05,
+      "loss": 0.4772,
       "step": 6835
     },
     {
-      "epoch": 0.24,
-      "learning_rate": 4.9956931782336694e-05,
-      "loss": 0.3178,
+      "epoch": 0.24651313655530327,
+      "grad_norm": 0.177265927195549,
+      "learning_rate": 4.995123240758919e-05,
+      "loss": 0.4816,
       "step": 6840
     },
     {
-      "epoch": 0.24,
-      "learning_rate": 4.99567644780469e-05,
-      "loss": 0.3072,
+      "epoch": 0.24669333621652792,
+      "grad_norm": 0.20768406987190247,
+      "learning_rate": 4.995105005625511e-05,
+      "loss": 0.482,
       "step": 6845
     },
     {
-      "epoch": 0.24,
-      "learning_rate": 4.995659684970997e-05,
-      "loss": 0.338,
+      "epoch": 0.24687353587775254,
+      "grad_norm": 0.18862581253051758,
+      "learning_rate": 4.9950867364967814e-05,
+      "loss": 0.4796,
       "step": 6850
     },
     {
-      "epoch": 0.24,
-      "learning_rate": 4.995642889732808e-05,
-      "loss": 0.3291,
+      "epoch": 0.2470537355389772,
+      "grad_norm": 0.20289580523967743,
+      "learning_rate": 4.995068433372978e-05,
+      "loss": 0.509,
       "step": 6855
     },
     {
-      "epoch": 0.24,
-      "learning_rate": 4.9956260620903406e-05,
-      "loss": 0.3177,
+      "epoch": 0.24723393520020182,
+      "grad_norm": 0.23247657716274261,
+      "learning_rate": 4.9950500962543503e-05,
+      "loss": 0.4609,
       "step": 6860
     },
     {
-      "epoch": 0.24,
-      "learning_rate": 4.995609202043814e-05,
-      "loss": 0.3261,
+      "epoch": 0.24741413486142647,
+      "grad_norm": 0.17218773066997528,
+      "learning_rate": 4.995031725141147e-05,
+      "loss": 0.4785,
       "step": 6865
     },
     {
-      "epoch": 0.24,
-      "learning_rate": 4.9955923095934454e-05,
-      "loss": 0.3423,
+      "epoch": 0.2475943345226511,
+      "grad_norm": 0.21331791579723358,
+      "learning_rate": 4.99501332003362e-05,
+      "loss": 0.4784,
       "step": 6870
     },
     {
-      "epoch": 0.24,
-      "learning_rate": 4.995575384739456e-05,
-      "loss": 0.3177,
+      "epoch": 0.24777453418387574,
+      "grad_norm": 0.25006669759750366,
+      "learning_rate": 4.99499488093202e-05,
+      "loss": 0.4931,
       "step": 6875
     },
     {
-      "epoch": 0.24,
-      "learning_rate": 4.995558427482065e-05,
-      "loss": 0.3263,
+      "epoch": 0.24795473384510036,
+      "grad_norm": 0.18483485281467438,
+      "learning_rate": 4.994976407836598e-05,
+      "loss": 0.4738,
       "step": 6880
     },
     {
-      "epoch": 0.24,
-      "learning_rate": 4.995541437821492e-05,
-      "loss": 0.3622,
+      "epoch": 0.248134933506325,
+      "grad_norm": 0.18278944492340088,
+      "learning_rate": 4.994957900747606e-05,
+      "loss": 0.4416,
       "step": 6885
     },
     {
-      "epoch": 0.24,
-      "learning_rate": 4.995524415757958e-05,
-      "loss": 0.3131,
+      "epoch": 0.24831513316754963,
+      "grad_norm": 0.17712728679180145,
+      "learning_rate": 4.9949393596652936e-05,
+      "loss": 0.5227,
       "step": 6890
     },
     {
-      "epoch": 0.24,
-      "learning_rate": 4.995507361291685e-05,
-      "loss": 0.3471,
+      "epoch": 0.24849533282877428,
+      "grad_norm": 0.18737031519412994,
+      "learning_rate": 4.994920784589917e-05,
+      "loss": 0.4913,
       "step": 6895
     },
     {
-      "epoch": 0.24,
-      "learning_rate": 4.995490274422893e-05,
-      "loss": 0.3,
+      "epoch": 0.2486755324899989,
+      "grad_norm": 0.2273300439119339,
+      "learning_rate": 4.9949021755217276e-05,
+      "loss": 0.4476,
       "step": 6900
     },
     {
-      "epoch": 0.24,
-      "learning_rate": 4.995473155151804e-05,
-      "loss": 0.339,
+      "epoch": 0.24885573215122356,
+      "grad_norm": 0.238374263048172,
+      "learning_rate": 4.9948835324609786e-05,
+      "loss": 0.4756,
       "step": 6905
     },
     {
-      "epoch": 0.24,
-      "learning_rate": 4.995456003478641e-05,
-      "loss": 0.3626,
+      "epoch": 0.24903593181244818,
+      "grad_norm": 0.21403105556964874,
+      "learning_rate": 4.9948648554079246e-05,
+      "loss": 0.448,
       "step": 6910
     },
     {
-      "epoch": 0.24,
-      "learning_rate": 4.995438819403626e-05,
-      "loss": 0.2978,
+      "epoch": 0.24921613147367283,
+      "grad_norm": 0.23275822401046753,
+      "learning_rate": 4.9948461443628205e-05,
+      "loss": 0.48,
       "step": 6915
     },
     {
-      "epoch": 0.24,
-      "learning_rate": 4.995421602926983e-05,
-      "loss": 0.3304,
+      "epoch": 0.24939633113489745,
+      "grad_norm": 0.13692793250083923,
+      "learning_rate": 4.99482739932592e-05,
+      "loss": 0.4495,
       "step": 6920
     },
     {
-      "epoch": 0.24,
-      "learning_rate": 4.9954043540489346e-05,
-      "loss": 0.2999,
+      "epoch": 0.2495765307961221,
+      "grad_norm": 0.2151869684457779,
+      "learning_rate": 4.9948086202974795e-05,
+      "loss": 0.4986,
       "step": 6925
     },
     {
-      "epoch": 0.24,
-      "learning_rate": 4.995387072769706e-05,
-      "loss": 0.3274,
+      "epoch": 0.24975673045734675,
+      "grad_norm": 0.19741404056549072,
+      "learning_rate": 4.9947898072777557e-05,
+      "loss": 0.4688,
       "step": 6930
     },
     {
-      "epoch": 0.24,
-      "learning_rate": 4.99536975908952e-05,
-      "loss": 0.3499,
+      "epoch": 0.24993693011857138,
+      "grad_norm": 0.1841607540845871,
+      "learning_rate": 4.994770960267002e-05,
+      "loss": 0.4252,
       "step": 6935
     },
     {
-      "epoch": 0.24,
-      "learning_rate": 4.995352413008603e-05,
-      "loss": 0.337,
+      "epoch": 0.250117129779796,
+      "grad_norm": 0.16864712536334991,
+      "learning_rate": 4.994752079265478e-05,
+      "loss": 0.4635,
       "step": 6940
     },
     {
-      "epoch": 0.24,
-      "learning_rate": 4.995335034527178e-05,
-      "loss": 0.3155,
+      "epoch": 0.2502973294410207,
+      "grad_norm": 0.1859603226184845,
+      "learning_rate": 4.9947331642734394e-05,
+      "loss": 0.4722,
       "step": 6945
     },
     {
-      "epoch": 0.24,
-      "learning_rate": 4.995317623645474e-05,
-      "loss": 0.3544,
+      "epoch": 0.2504775291022453,
+      "grad_norm": 0.19181598722934723,
+      "learning_rate": 4.994714215291144e-05,
+      "loss": 0.5074,
       "step": 6950
     },
     {
-      "epoch": 0.24,
-      "learning_rate": 4.9953001803637136e-05,
-      "loss": 0.3439,
+      "epoch": 0.2506577287634699,
+      "grad_norm": 0.17568619549274445,
+      "learning_rate": 4.9946952323188514e-05,
+      "loss": 0.4396,
       "step": 6955
     },
     {
-      "epoch": 0.24,
-      "learning_rate": 4.995282704682125e-05,
-      "loss": 0.3484,
+      "epoch": 0.25083792842469455,
+      "grad_norm": 0.18716076016426086,
+      "learning_rate": 4.9946762153568195e-05,
+      "loss": 0.4776,
       "step": 6960
     },
     {
-      "epoch": 0.25,
-      "learning_rate": 4.995265196600935e-05,
-      "loss": 0.3429,
+      "epoch": 0.2510181280859192,
+      "grad_norm": 0.18289898335933685,
+      "learning_rate": 4.994657164405306e-05,
+      "loss": 0.4569,
       "step": 6965
     },
     {
-      "epoch": 0.25,
-      "learning_rate": 4.9952476561203706e-05,
-      "loss": 0.3532,
+      "epoch": 0.25119832774714385,
+      "grad_norm": 0.1580553501844406,
+      "learning_rate": 4.994638079464572e-05,
+      "loss": 0.4865,
       "step": 6970
     },
     {
-      "epoch": 0.25,
-      "learning_rate": 4.9952300832406605e-05,
-      "loss": 0.3107,
+      "epoch": 0.25137852740836847,
+      "grad_norm": 0.200215145945549,
+      "learning_rate": 4.9946189605348775e-05,
+      "loss": 0.4281,
       "step": 6975
     },
     {
-      "epoch": 0.25,
-      "learning_rate": 4.9952124779620324e-05,
-      "loss": 0.3309,
+      "epoch": 0.2515587270695931,
+      "grad_norm": 0.16953638195991516,
+      "learning_rate": 4.9945998076164824e-05,
+      "loss": 0.4802,
       "step": 6980
     },
     {
-      "epoch": 0.25,
-      "learning_rate": 4.995194840284714e-05,
-      "loss": 0.2985,
+      "epoch": 0.25173892673081777,
+      "grad_norm": 0.22101880609989166,
+      "learning_rate": 4.994580620709648e-05,
+      "loss": 0.4825,
       "step": 6985
     },
     {
-      "epoch": 0.25,
-      "learning_rate": 4.995177170208935e-05,
-      "loss": 0.3145,
+      "epoch": 0.2519191263920424,
+      "grad_norm": 0.1716253012418747,
+      "learning_rate": 4.9945613998146356e-05,
+      "loss": 0.4562,
       "step": 6990
     },
     {
-      "epoch": 0.25,
-      "learning_rate": 4.995159467734925e-05,
-      "loss": 0.3284,
+      "epoch": 0.252099326053267,
+      "grad_norm": 0.19001296162605286,
+      "learning_rate": 4.9945421449317065e-05,
+      "loss": 0.4797,
       "step": 6995
     },
     {
-      "epoch": 0.25,
-      "learning_rate": 4.9951417328629136e-05,
-      "loss": 0.3456,
+      "epoch": 0.25227952571449164,
+      "grad_norm": 0.1752406656742096,
+      "learning_rate": 4.9945228560611244e-05,
+      "loss": 0.4398,
       "step": 7000
     },
     {
-      "epoch": 0.25,
-      "eval_loss": 0.3314030170440674,
-      "eval_runtime": 10.5578,
-      "eval_samples_per_second": 9.472,
-      "eval_steps_per_second": 9.472,
+      "epoch": 0.25227952571449164,
+      "eval_loss": 0.49143746495246887,
+      "eval_runtime": 3.5206,
+      "eval_samples_per_second": 28.404,
+      "eval_steps_per_second": 7.101,
       "step": 7000
     },
     {
-      "epoch": 0.25,
-      "learning_rate": 4.9951239655931314e-05,
-      "loss": 0.3508,
+      "epoch": 0.2524597253757163,
+      "grad_norm": 0.15203414857387543,
+      "learning_rate": 4.994503533203151e-05,
+      "loss": 0.4602,
       "step": 7005
     },
     {
-      "epoch": 0.25,
-      "learning_rate": 4.995106165925809e-05,
-      "loss": 0.3357,
+      "epoch": 0.25263992503694094,
+      "grad_norm": 0.1956150233745575,
+      "learning_rate": 4.9944841763580505e-05,
+      "loss": 0.4639,
       "step": 7010
     },
     {
-      "epoch": 0.25,
-      "learning_rate": 4.9950883338611765e-05,
-      "loss": 0.3471,
+      "epoch": 0.25282012469816556,
+      "grad_norm": 0.18748140335083008,
+      "learning_rate": 4.9944647855260854e-05,
+      "loss": 0.4597,
       "step": 7015
     },
     {
-      "epoch": 0.25,
-      "learning_rate": 4.9950704693994673e-05,
-      "loss": 0.2958,
+      "epoch": 0.2530003243593902,
+      "grad_norm": 0.19696061313152313,
+      "learning_rate": 4.99444536070752e-05,
+      "loss": 0.449,
       "step": 7020
     },
     {
-      "epoch": 0.25,
-      "learning_rate": 4.995052572540912e-05,
-      "loss": 0.3492,
+      "epoch": 0.25318052402061486,
+      "grad_norm": 0.1847335398197174,
+      "learning_rate": 4.9944259019026207e-05,
+      "loss": 0.4875,
       "step": 7025
     },
     {
-      "epoch": 0.25,
-      "learning_rate": 4.995034643285743e-05,
-      "loss": 0.3428,
+      "epoch": 0.2533607236818395,
+      "grad_norm": 0.17691460251808167,
+      "learning_rate": 4.99440640911165e-05,
+      "loss": 0.4793,
       "step": 7030
     },
     {
-      "epoch": 0.25,
-      "learning_rate": 4.995016681634194e-05,
-      "loss": 0.3175,
+      "epoch": 0.2535409233430641,
+      "grad_norm": 0.18585610389709473,
+      "learning_rate": 4.994386882334877e-05,
+      "loss": 0.4646,
       "step": 7035
     },
     {
-      "epoch": 0.25,
-      "learning_rate": 4.9949986875864976e-05,
-      "loss": 0.3196,
+      "epoch": 0.25372112300428873,
+      "grad_norm": 0.2865478992462158,
+      "learning_rate": 4.994367321572564e-05,
+      "loss": 0.491,
       "step": 7040
     },
     {
-      "epoch": 0.25,
-      "learning_rate": 4.994980661142887e-05,
-      "loss": 0.3415,
+      "epoch": 0.2539013226655134,
+      "grad_norm": 0.20224061608314514,
+      "learning_rate": 4.9943477268249796e-05,
+      "loss": 0.5204,
       "step": 7045
     },
     {
-      "epoch": 0.25,
-      "learning_rate": 4.9949626023035974e-05,
-      "loss": 0.3353,
+      "epoch": 0.25408152232673803,
+      "grad_norm": 0.181314155459404,
+      "learning_rate": 4.99432809809239e-05,
+      "loss": 0.4963,
       "step": 7050
     },
     {
-      "epoch": 0.25,
-      "learning_rate": 4.994944511068862e-05,
-      "loss": 0.3419,
+      "epoch": 0.25426172198796265,
+      "grad_norm": 0.22092019021511078,
+      "learning_rate": 4.9943084353750635e-05,
+      "loss": 0.4489,
       "step": 7055
     },
     {
-      "epoch": 0.25,
-      "learning_rate": 4.9949263874389176e-05,
-      "loss": 0.3174,
+      "epoch": 0.2544419216491873,
+      "grad_norm": 0.19176073372364044,
+      "learning_rate": 4.9942887386732676e-05,
+      "loss": 0.4806,
       "step": 7060
     },
     {
-      "epoch": 0.25,
-      "learning_rate": 4.994908231413997e-05,
-      "loss": 0.3193,
+      "epoch": 0.25462212131041195,
+      "grad_norm": 0.2104840874671936,
+      "learning_rate": 4.99426900798727e-05,
+      "loss": 0.4741,
       "step": 7065
     },
     {
-      "epoch": 0.25,
-      "learning_rate": 4.994890042994337e-05,
-      "loss": 0.351,
+      "epoch": 0.2548023209716366,
+      "grad_norm": 0.14947918057441711,
+      "learning_rate": 4.9942492433173405e-05,
+      "loss": 0.5013,
       "step": 7070
     },
     {
-      "epoch": 0.25,
-      "learning_rate": 4.994871822180175e-05,
-      "loss": 0.3311,
+      "epoch": 0.2549825206328612,
+      "grad_norm": 0.1898050606250763,
+      "learning_rate": 4.9942294446637486e-05,
+      "loss": 0.4995,
       "step": 7075
     },
     {
-      "epoch": 0.25,
-      "learning_rate": 4.994853568971746e-05,
-      "loss": 0.3302,
+      "epoch": 0.2551627202940858,
+      "grad_norm": 0.18579892814159393,
+      "learning_rate": 4.9942096120267626e-05,
+      "loss": 0.4634,
       "step": 7080
     },
     {
-      "epoch": 0.25,
-      "learning_rate": 4.994835283369288e-05,
-      "loss": 0.372,
+      "epoch": 0.2553429199553105,
+      "grad_norm": 0.2083805650472641,
+      "learning_rate": 4.9941897454066535e-05,
+      "loss": 0.4602,
       "step": 7085
     },
     {
-      "epoch": 0.25,
-      "learning_rate": 4.994816965373038e-05,
-      "loss": 0.3315,
+      "epoch": 0.2555231196165351,
+      "grad_norm": 0.22849465906620026,
+      "learning_rate": 4.9941698448036916e-05,
+      "loss": 0.4593,
       "step": 7090
     },
     {
-      "epoch": 0.25,
-      "learning_rate": 4.994798614983234e-05,
-      "loss": 0.3142,
+      "epoch": 0.25570331927775974,
+      "grad_norm": 0.2611923813819885,
+      "learning_rate": 4.994149910218149e-05,
+      "loss": 0.4715,
       "step": 7095
     },
     {
-      "epoch": 0.25,
-      "learning_rate": 4.994780232200113e-05,
-      "loss": 0.3445,
+      "epoch": 0.2558835189389844,
+      "grad_norm": 0.20796480774879456,
+      "learning_rate": 4.994129941650296e-05,
+      "loss": 0.4678,
       "step": 7100
     },
     {
-      "epoch": 0.25,
-      "learning_rate": 4.9947618170239154e-05,
-      "loss": 0.3185,
+      "epoch": 0.25606371860020904,
+      "grad_norm": 0.16699600219726562,
+      "learning_rate": 4.994109939100406e-05,
+      "loss": 0.4886,
       "step": 7105
     },
     {
-      "epoch": 0.25,
-      "learning_rate": 4.99474336945488e-05,
-      "loss": 0.3178,
+      "epoch": 0.25624391826143367,
+      "grad_norm": 0.18829451501369476,
+      "learning_rate": 4.994089902568751e-05,
+      "loss": 0.4502,
       "step": 7110
     },
     {
-      "epoch": 0.25,
-      "learning_rate": 4.9947248894932464e-05,
-      "loss": 0.3449,
+      "epoch": 0.2564241179226583,
+      "grad_norm": 0.20573166012763977,
+      "learning_rate": 4.994069832055604e-05,
+      "loss": 0.4985,
       "step": 7115
     },
     {
-      "epoch": 0.25,
-      "learning_rate": 4.994706377139253e-05,
-      "loss": 0.3358,
+      "epoch": 0.25660431758388297,
+      "grad_norm": 0.22231273353099823,
+      "learning_rate": 4.994049727561239e-05,
+      "loss": 0.4926,
       "step": 7120
     },
     {
-      "epoch": 0.25,
-      "learning_rate": 4.994687832393142e-05,
-      "loss": 0.3166,
+      "epoch": 0.2567845172451076,
+      "grad_norm": 0.19772711396217346,
+      "learning_rate": 4.994029589085929e-05,
+      "loss": 0.4781,
       "step": 7125
     },
     {
-      "epoch": 0.25,
-      "learning_rate": 4.994669255255154e-05,
-      "loss": 0.3177,
+      "epoch": 0.2569647169063322,
+      "grad_norm": 0.24097611010074615,
+      "learning_rate": 4.9940094166299486e-05,
+      "loss": 0.4474,
       "step": 7130
     },
     {
-      "epoch": 0.25,
-      "learning_rate": 4.994650645725529e-05,
-      "loss": 0.3748,
+      "epoch": 0.25714491656755684,
+      "grad_norm": 0.2717370390892029,
+      "learning_rate": 4.9939892101935723e-05,
+      "loss": 0.473,
       "step": 7135
     },
     {
-      "epoch": 0.25,
-      "learning_rate": 4.99463200380451e-05,
-      "loss": 0.3148,
+      "epoch": 0.2573251162287815,
+      "grad_norm": 0.16408710181713104,
+      "learning_rate": 4.993968969777076e-05,
+      "loss": 0.4656,
       "step": 7140
     },
     {
-      "epoch": 0.25,
-      "learning_rate": 4.9946133294923384e-05,
-      "loss": 0.3303,
+      "epoch": 0.25750531589000614,
+      "grad_norm": 0.19900165498256683,
+      "learning_rate": 4.993948695380736e-05,
+      "loss": 0.4309,
       "step": 7145
     },
     {
-      "epoch": 0.25,
-      "learning_rate": 4.9945946227892566e-05,
-      "loss": 0.3862,
+      "epoch": 0.25768551555123076,
+      "grad_norm": 0.19811104238033295,
+      "learning_rate": 4.993928387004827e-05,
+      "loss": 0.4194,
       "step": 7150
     },
     {
-      "epoch": 0.25,
-      "learning_rate": 4.9945758836955075e-05,
-      "loss": 0.334,
+      "epoch": 0.2578657152124554,
+      "grad_norm": 0.20136548578739166,
+      "learning_rate": 4.9939080446496264e-05,
+      "loss": 0.4728,
       "step": 7155
     },
     {
-      "epoch": 0.25,
-      "learning_rate": 4.9945571122113344e-05,
-      "loss": 0.3192,
+      "epoch": 0.25804591487368006,
+      "grad_norm": 0.20853722095489502,
+      "learning_rate": 4.993887668315413e-05,
+      "loss": 0.436,
       "step": 7160
     },
     {
-      "epoch": 0.25,
-      "learning_rate": 4.994538308336981e-05,
-      "loss": 0.3359,
+      "epoch": 0.2582261145349047,
+      "grad_norm": 0.18556027114391327,
+      "learning_rate": 4.9938672580024615e-05,
+      "loss": 0.4321,
       "step": 7165
     },
     {
-      "epoch": 0.25,
-      "learning_rate": 4.9945194720726926e-05,
-      "loss": 0.3084,
+      "epoch": 0.2584063141961293,
+      "grad_norm": 0.24166804552078247,
+      "learning_rate": 4.993846813711052e-05,
+      "loss": 0.4749,
       "step": 7170
     },
     {
-      "epoch": 0.25,
-      "learning_rate": 4.9945006034187124e-05,
-      "loss": 0.3323,
+      "epoch": 0.25858651385735393,
+      "grad_norm": 0.17454582452774048,
+      "learning_rate": 4.9938263354414626e-05,
+      "loss": 0.4782,
       "step": 7175
     },
     {
-      "epoch": 0.25,
-      "learning_rate": 4.9944817023752853e-05,
-      "loss": 0.3196,
+      "epoch": 0.2587667135185786,
+      "grad_norm": 0.18085728585720062,
+      "learning_rate": 4.993805823193972e-05,
+      "loss": 0.4631,
       "step": 7180
     },
     {
-      "epoch": 0.25,
-      "learning_rate": 4.9944627689426574e-05,
-      "loss": 0.3358,
+      "epoch": 0.25894691317980323,
+      "grad_norm": 0.168624609708786,
+      "learning_rate": 4.99378527696886e-05,
+      "loss": 0.4665,
       "step": 7185
     },
     {
-      "epoch": 0.25,
-      "learning_rate": 4.994443803121075e-05,
-      "loss": 0.3596,
+      "epoch": 0.25912711284102785,
+      "grad_norm": 0.17559649050235748,
+      "learning_rate": 4.9937646967664066e-05,
+      "loss": 0.4262,
       "step": 7190
     },
     {
-      "epoch": 0.25,
-      "learning_rate": 4.994424804910783e-05,
-      "loss": 0.3423,
+      "epoch": 0.2593073125022525,
+      "grad_norm": 0.1460302621126175,
+      "learning_rate": 4.993744082586891e-05,
+      "loss": 0.4484,
       "step": 7195
     },
     {
-      "epoch": 0.25,
-      "learning_rate": 4.9944057743120284e-05,
-      "loss": 0.3223,
+      "epoch": 0.25948751216347715,
+      "grad_norm": 0.250219464302063,
+      "learning_rate": 4.9937234344305964e-05,
+      "loss": 0.477,
       "step": 7200
     },
     {
-      "epoch": 0.25,
-      "learning_rate": 4.99438671132506e-05,
-      "loss": 0.3438,
+      "epoch": 0.2596677118247018,
+      "grad_norm": 0.17748834192752838,
+      "learning_rate": 4.993702752297802e-05,
+      "loss": 0.4931,
       "step": 7205
     },
     {
-      "epoch": 0.25,
-      "learning_rate": 4.994367615950123e-05,
-      "loss": 0.2991,
+      "epoch": 0.2598479114859264,
+      "grad_norm": 0.21896113455295563,
+      "learning_rate": 4.99368203618879e-05,
+      "loss": 0.4621,
       "step": 7210
     },
     {
-      "epoch": 0.25,
-      "learning_rate": 4.994348488187467e-05,
-      "loss": 0.3617,
+      "epoch": 0.260028111147151,
+      "grad_norm": 0.25870126485824585,
+      "learning_rate": 4.9936612861038446e-05,
+      "loss": 0.4634,
       "step": 7215
     },
     {
-      "epoch": 0.25,
-      "learning_rate": 4.9943293280373395e-05,
-      "loss": 0.3359,
+      "epoch": 0.2602083108083757,
+      "grad_norm": 0.1914265900850296,
+      "learning_rate": 4.993640502043246e-05,
+      "loss": 0.4604,
       "step": 7220
     },
     {
-      "epoch": 0.25,
-      "learning_rate": 4.994310135499989e-05,
-      "loss": 0.3333,
+      "epoch": 0.2603885104696003,
+      "grad_norm": 0.16129115223884583,
+      "learning_rate": 4.993619684007278e-05,
+      "loss": 0.4598,
       "step": 7225
     },
     {
-      "epoch": 0.25,
-      "learning_rate": 4.994290910575666e-05,
-      "loss": 0.3275,
+      "epoch": 0.26056871013082494,
+      "grad_norm": 0.19293855130672455,
+      "learning_rate": 4.993598831996225e-05,
+      "loss": 0.464,
       "step": 7230
     },
     {
-      "epoch": 0.25,
-      "learning_rate": 4.994271653264619e-05,
-      "loss": 0.3242,
+      "epoch": 0.26074890979204957,
+      "grad_norm": 0.1789807379245758,
+      "learning_rate": 4.993577946010371e-05,
+      "loss": 0.4816,
       "step": 7235
     },
     {
-      "epoch": 0.25,
-      "learning_rate": 4.994252363567098e-05,
-      "loss": 0.3308,
+      "epoch": 0.26092910945327424,
+      "grad_norm": 0.19150225818157196,
+      "learning_rate": 4.99355702605e-05,
+      "loss": 0.4875,
       "step": 7240
     },
     {
-      "epoch": 0.25,
-      "learning_rate": 4.9942330414833546e-05,
-      "loss": 0.3246,
+      "epoch": 0.26110930911449887,
+      "grad_norm": 0.2224695235490799,
+      "learning_rate": 4.9935360721153965e-05,
+      "loss": 0.4727,
       "step": 7245
     },
     {
-      "epoch": 0.26,
-      "learning_rate": 4.9942136870136383e-05,
-      "loss": 0.3392,
+      "epoch": 0.2612895087757235,
+      "grad_norm": 0.17574471235275269,
+      "learning_rate": 4.993515084206848e-05,
+      "loss": 0.4247,
       "step": 7250
     },
     {
-      "epoch": 0.26,
-      "learning_rate": 4.9941943001582015e-05,
-      "loss": 0.331,
+      "epoch": 0.2614697084369481,
+      "grad_norm": 0.16948924958705902,
+      "learning_rate": 4.9934940623246387e-05,
+      "loss": 0.4704,
       "step": 7255
     },
     {
-      "epoch": 0.26,
-      "learning_rate": 4.994174880917295e-05,
-      "loss": 0.3131,
+      "epoch": 0.2616499080981728,
+      "grad_norm": 0.17520290613174438,
+      "learning_rate": 4.993473006469055e-05,
+      "loss": 0.4646,
       "step": 7260
     },
     {
-      "epoch": 0.26,
-      "learning_rate": 4.9941554292911715e-05,
-      "loss": 0.3576,
+      "epoch": 0.2618301077593974,
+      "grad_norm": 0.20636965334415436,
+      "learning_rate": 4.993451916640386e-05,
+      "loss": 0.4463,
       "step": 7265
     },
     {
-      "epoch": 0.26,
-      "learning_rate": 4.9941359452800826e-05,
-      "loss": 0.3267,
+      "epoch": 0.26201030742062204,
+      "grad_norm": 0.1967248171567917,
+      "learning_rate": 4.9934307928389154e-05,
+      "loss": 0.4631,
       "step": 7270
     },
     {
-      "epoch": 0.26,
-      "learning_rate": 4.994116428884283e-05,
-      "loss": 0.3251,
+      "epoch": 0.2621905070818467,
+      "grad_norm": 0.22940480709075928,
+      "learning_rate": 4.993409635064934e-05,
+      "loss": 0.4354,
       "step": 7275
     },
     {
-      "epoch": 0.26,
-      "learning_rate": 4.994096880104026e-05,
-      "loss": 0.3205,
+      "epoch": 0.26237070674307134,
+      "grad_norm": 0.2190198004245758,
+      "learning_rate": 4.9933884433187295e-05,
+      "loss": 0.4826,
       "step": 7280
     },
     {
-      "epoch": 0.26,
-      "learning_rate": 4.994077298939564e-05,
-      "loss": 0.3455,
+      "epoch": 0.26255090640429596,
+      "grad_norm": 0.16079477965831757,
+      "learning_rate": 4.9933672176005894e-05,
+      "loss": 0.4499,
       "step": 7285
     },
     {
-      "epoch": 0.26,
-      "learning_rate": 4.994057685391152e-05,
-      "loss": 0.3158,
+      "epoch": 0.2627311060655206,
+      "grad_norm": 0.24084745347499847,
+      "learning_rate": 4.993345957910804e-05,
+      "loss": 0.5024,
       "step": 7290
     },
     {
-      "epoch": 0.26,
-      "learning_rate": 4.994038039459045e-05,
-      "loss": 0.3288,
+      "epoch": 0.26291130572674526,
+      "grad_norm": 0.16274696588516235,
+      "learning_rate": 4.9933246642496626e-05,
+      "loss": 0.4463,
       "step": 7295
     },
     {
-      "epoch": 0.26,
-      "learning_rate": 4.994018361143497e-05,
-      "loss": 0.3069,
+      "epoch": 0.2630915053879699,
+      "grad_norm": 0.21243201196193695,
+      "learning_rate": 4.9933033366174554e-05,
+      "loss": 0.4764,
       "step": 7300
     },
     {
-      "epoch": 0.26,
-      "learning_rate": 4.993998650444764e-05,
-      "loss": 0.3741,
+      "epoch": 0.2632717050491945,
+      "grad_norm": 0.20116925239562988,
+      "learning_rate": 4.9932819750144734e-05,
+      "loss": 0.484,
       "step": 7305
     },
     {
-      "epoch": 0.26,
-      "learning_rate": 4.9939789073631025e-05,
-      "loss": 0.3298,
+      "epoch": 0.2634519047104191,
+      "grad_norm": 0.20278726518154144,
+      "learning_rate": 4.993260579441006e-05,
+      "loss": 0.4757,
       "step": 7310
     },
     {
-      "epoch": 0.26,
-      "learning_rate": 4.993959131898769e-05,
-      "loss": 0.3459,
+      "epoch": 0.2636321043716438,
+      "grad_norm": 0.17847001552581787,
+      "learning_rate": 4.993239149897347e-05,
+      "loss": 0.4309,
       "step": 7315
     },
     {
-      "epoch": 0.26,
-      "learning_rate": 4.993939324052019e-05,
-      "loss": 0.3345,
+      "epoch": 0.2638123040328684,
+      "grad_norm": 0.1934574991464615,
+      "learning_rate": 4.993217686383787e-05,
+      "loss": 0.4267,
       "step": 7320
     },
     {
-      "epoch": 0.26,
-      "learning_rate": 4.9939194838231104e-05,
-      "loss": 0.3655,
+      "epoch": 0.26399250369409305,
+      "grad_norm": 0.1816255897283554,
+      "learning_rate": 4.993196188900618e-05,
+      "loss": 0.4676,
       "step": 7325
     },
     {
-      "epoch": 0.26,
-      "learning_rate": 4.993899611212302e-05,
-      "loss": 0.3267,
+      "epoch": 0.2641727033553177,
+      "grad_norm": 0.1440691500902176,
+      "learning_rate": 4.993174657448135e-05,
+      "loss": 0.4451,
       "step": 7330
     },
     {
-      "epoch": 0.26,
-      "learning_rate": 4.993879706219849e-05,
-      "loss": 0.3244,
+      "epoch": 0.26435290301654235,
+      "grad_norm": 0.18416348099708557,
+      "learning_rate": 4.993153092026629e-05,
+      "loss": 0.4692,
       "step": 7335
     },
     {
-      "epoch": 0.26,
-      "learning_rate": 4.993859768846013e-05,
-      "loss": 0.3159,
+      "epoch": 0.264533102677767,
+      "grad_norm": 0.1908058077096939,
+      "learning_rate": 4.9931314926363945e-05,
+      "loss": 0.438,
       "step": 7340
     },
     {
-      "epoch": 0.26,
-      "learning_rate": 4.99383979909105e-05,
-      "loss": 0.3231,
+      "epoch": 0.2647133023389916,
+      "grad_norm": 0.17665372788906097,
+      "learning_rate": 4.993109859277727e-05,
+      "loss": 0.4475,
       "step": 7345
     },
     {
-      "epoch": 0.26,
-      "learning_rate": 4.993819796955221e-05,
-      "loss": 0.3526,
+      "epoch": 0.2648935020002162,
+      "grad_norm": 0.20020587742328644,
+      "learning_rate": 4.99308819195092e-05,
+      "loss": 0.466,
       "step": 7350
     },
     {
-      "epoch": 0.26,
-      "learning_rate": 4.9937997624387864e-05,
-      "loss": 0.329,
+      "epoch": 0.2650737016614409,
+      "grad_norm": 0.20081447064876556,
+      "learning_rate": 4.9930664906562695e-05,
+      "loss": 0.4793,
       "step": 7355
     },
     {
-      "epoch": 0.26,
-      "learning_rate": 4.993779695542005e-05,
-      "loss": 0.3234,
+      "epoch": 0.2652539013226655,
+      "grad_norm": 0.1721320003271103,
+      "learning_rate": 4.993049105163899e-05,
+      "loss": 0.4658,
       "step": 7360
     },
     {
-      "epoch": 0.26,
-      "learning_rate": 4.993759596265138e-05,
-      "loss": 0.3473,
+      "epoch": 0.26543410098389014,
+      "grad_norm": 0.24417546391487122,
+      "learning_rate": 4.993027342727875e-05,
+      "loss": 0.4484,
       "step": 7365
     },
     {
-      "epoch": 0.26,
-      "learning_rate": 4.9937394646084454e-05,
-      "loss": 0.3387,
+      "epoch": 0.26561430064511476,
+      "grad_norm": 0.2776060700416565,
+      "learning_rate": 4.993005546324836e-05,
+      "loss": 0.4846,
       "step": 7370
     },
     {
-      "epoch": 0.26,
-      "learning_rate": 4.993719300572189e-05,
-      "loss": 0.3333,
+      "epoch": 0.26579450030633944,
+      "grad_norm": 0.19747385382652283,
+      "learning_rate": 4.9929837159550784e-05,
+      "loss": 0.4603,
       "step": 7375
     },
     {
-      "epoch": 0.26,
-      "learning_rate": 4.993699104156632e-05,
-      "loss": 0.3357,
+      "epoch": 0.26597469996756407,
+      "grad_norm": 0.1888158768415451,
+      "learning_rate": 4.9929618516189e-05,
+      "loss": 0.4405,
       "step": 7380
     },
     {
-      "epoch": 0.26,
-      "learning_rate": 4.993678875362036e-05,
-      "loss": 0.338,
+      "epoch": 0.2661548996287887,
+      "grad_norm": 0.20889566838741302,
+      "learning_rate": 4.9929399533166e-05,
+      "loss": 0.4846,
       "step": 7385
     },
     {
-      "epoch": 0.26,
-      "learning_rate": 4.993658614188662e-05,
-      "loss": 0.3469,
+      "epoch": 0.2663350992900133,
+      "grad_norm": 0.19640065729618073,
+      "learning_rate": 4.992918021048475e-05,
+      "loss": 0.4725,
       "step": 7390
     },
     {
-      "epoch": 0.26,
-      "learning_rate": 4.9936383206367745e-05,
-      "loss": 0.3189,
+      "epoch": 0.266515298951238,
+      "grad_norm": 0.20840995013713837,
+      "learning_rate": 4.992896054814825e-05,
+      "loss": 0.4594,
       "step": 7395
     },
     {
-      "epoch": 0.26,
-      "learning_rate": 4.993617994706637e-05,
-      "loss": 0.3178,
+      "epoch": 0.2666954986124626,
+      "grad_norm": 0.2129368633031845,
+      "learning_rate": 4.992874054615949e-05,
+      "loss": 0.4721,
       "step": 7400
     },
     {
-      "epoch": 0.26,
-      "learning_rate": 4.9935976363985126e-05,
-      "loss": 0.3372,
+      "epoch": 0.26687569827368723,
+      "grad_norm": 0.13430319726467133,
+      "learning_rate": 4.992852020452147e-05,
+      "loss": 0.4479,
       "step": 7405
     },
     {
-      "epoch": 0.26,
-      "learning_rate": 4.993577245712666e-05,
-      "loss": 0.3349,
+      "epoch": 0.26705589793491186,
+      "grad_norm": 0.2261311113834381,
+      "learning_rate": 4.992829952323718e-05,
+      "loss": 0.4972,
       "step": 7410
     },
     {
-      "epoch": 0.26,
-      "learning_rate": 4.993556822649363e-05,
-      "loss": 0.3583,
+      "epoch": 0.26723609759613653,
+      "grad_norm": 0.21001654863357544,
+      "learning_rate": 4.992807850230964e-05,
+      "loss": 0.5047,
       "step": 7415
     },
     {
-      "epoch": 0.26,
-      "learning_rate": 4.993536367208868e-05,
-      "loss": 0.318,
+      "epoch": 0.26741629725736116,
+      "grad_norm": 0.2056044489145279,
+      "learning_rate": 4.992785714174185e-05,
+      "loss": 0.4793,
       "step": 7420
     },
     {
-      "epoch": 0.26,
-      "learning_rate": 4.993515879391447e-05,
-      "loss": 0.3181,
+      "epoch": 0.2675964969185858,
+      "grad_norm": 0.156636044383049,
+      "learning_rate": 4.9927635441536844e-05,
+      "loss": 0.4914,
       "step": 7425
     },
     {
-      "epoch": 0.26,
-      "learning_rate": 4.993495359197364e-05,
-      "loss": 0.3264,
+      "epoch": 0.2677766965798104,
+      "grad_norm": 0.18658673763275146,
+      "learning_rate": 4.9927413401697625e-05,
+      "loss": 0.4451,
       "step": 7430
     },
     {
-      "epoch": 0.26,
-      "learning_rate": 4.993474806626889e-05,
-      "loss": 0.2904,
+      "epoch": 0.2679568962410351,
+      "grad_norm": 0.14643876254558563,
+      "learning_rate": 4.992719102222723e-05,
+      "loss": 0.4108,
       "step": 7435
     },
     {
-      "epoch": 0.26,
-      "learning_rate": 4.9934542216802856e-05,
-      "loss": 0.3285,
+      "epoch": 0.2681370959022597,
+      "grad_norm": 0.2012728452682495,
+      "learning_rate": 4.9926968303128674e-05,
+      "loss": 0.4497,
       "step": 7440
     },
     {
-      "epoch": 0.26,
-      "learning_rate": 4.9934336043578225e-05,
-      "loss": 0.3649,
+      "epoch": 0.2683172955634843,
+      "grad_norm": 0.19251321256160736,
+      "learning_rate": 4.9926745244405e-05,
+      "loss": 0.4938,
       "step": 7445
     },
     {
-      "epoch": 0.26,
-      "learning_rate": 4.993412954659767e-05,
-      "loss": 0.3389,
+      "epoch": 0.268497495224709,
+      "grad_norm": 0.19038186967372894,
+      "learning_rate": 4.992652184605926e-05,
+      "loss": 0.4739,
       "step": 7450
     },
     {
-      "epoch": 0.26,
-      "learning_rate": 4.993392272586388e-05,
-      "loss": 0.3359,
+      "epoch": 0.2686776948859336,
+      "grad_norm": 0.21992941200733185,
+      "learning_rate": 4.992629810809448e-05,
+      "loss": 0.4839,
       "step": 7455
     },
     {
-      "epoch": 0.26,
-      "learning_rate": 4.9933715581379536e-05,
-      "loss": 0.35,
+      "epoch": 0.26885789454715825,
+      "grad_norm": 0.208626389503479,
+      "learning_rate": 4.992607403051371e-05,
+      "loss": 0.4498,
       "step": 7460
     },
     {
-      "epoch": 0.26,
-      "learning_rate": 4.993350811314733e-05,
-      "loss": 0.3076,
+      "epoch": 0.26903809420838287,
+      "grad_norm": 0.14760567247867584,
+      "learning_rate": 4.9925849613320006e-05,
+      "loss": 0.4185,
       "step": 7465
     },
     {
-      "epoch": 0.26,
-      "learning_rate": 4.993330032116995e-05,
-      "loss": 0.353,
+      "epoch": 0.26921829386960755,
+      "grad_norm": 0.2208520770072937,
+      "learning_rate": 4.992562485651644e-05,
+      "loss": 0.4734,
       "step": 7470
     },
     {
-      "epoch": 0.26,
-      "learning_rate": 4.9933092205450094e-05,
-      "loss": 0.3349,
+      "epoch": 0.2693984935308322,
+      "grad_norm": 0.18603351712226868,
+      "learning_rate": 4.992539976010605e-05,
+      "loss": 0.5101,
       "step": 7475
     },
     {
-      "epoch": 0.26,
-      "learning_rate": 4.9932883765990466e-05,
-      "loss": 0.3446,
+      "epoch": 0.2695786931920568,
+      "grad_norm": 0.2119520753622055,
+      "learning_rate": 4.992517432409192e-05,
+      "loss": 0.4816,
       "step": 7480
     },
     {
-      "epoch": 0.26,
-      "learning_rate": 4.993267500279378e-05,
-      "loss": 0.3393,
+      "epoch": 0.2697588928532814,
+      "grad_norm": 0.17017479240894318,
+      "learning_rate": 4.99249485484771e-05,
+      "loss": 0.4189,
       "step": 7485
     },
     {
-      "epoch": 0.26,
-      "learning_rate": 4.993246591586274e-05,
-      "loss": 0.3423,
+      "epoch": 0.2699390925145061,
+      "grad_norm": 0.2112009972333908,
+      "learning_rate": 4.99247224332647e-05,
+      "loss": 0.4506,
       "step": 7490
     },
     {
-      "epoch": 0.26,
-      "learning_rate": 4.993225650520006e-05,
-      "loss": 0.3387,
+      "epoch": 0.2701192921757307,
+      "grad_norm": 0.22888222336769104,
+      "learning_rate": 4.992449597845777e-05,
+      "loss": 0.444,
       "step": 7495
     },
     {
-      "epoch": 0.26,
-      "learning_rate": 4.9932046770808454e-05,
-      "loss": 0.3369,
+      "epoch": 0.27029949183695534,
+      "grad_norm": 0.19122794270515442,
+      "learning_rate": 4.992426918405941e-05,
+      "loss": 0.4645,
       "step": 7500
     },
     {
-      "epoch": 0.26,
-      "eval_loss": 0.3308403789997101,
-      "eval_runtime": 10.5623,
-      "eval_samples_per_second": 9.468,
-      "eval_steps_per_second": 9.468,
+      "epoch": 0.27029949183695534,
+      "eval_loss": 0.48843687772750854,
+      "eval_runtime": 3.5217,
+      "eval_samples_per_second": 28.395,
+      "eval_steps_per_second": 7.099,
       "step": 7500
     },
     {
-      "epoch": 0.26,
-      "learning_rate": 4.993183671269066e-05,
-      "loss": 0.3551,
+      "epoch": 0.27047969149817996,
+      "grad_norm": 0.18610242009162903,
+      "learning_rate": 4.992404205007272e-05,
+      "loss": 0.4546,
       "step": 7505
     },
     {
-      "epoch": 0.26,
-      "learning_rate": 4.993162633084939e-05,
-      "loss": 0.3412,
+      "epoch": 0.27065989115940464,
+      "grad_norm": 0.23450550436973572,
+      "learning_rate": 4.992381457650077e-05,
+      "loss": 0.4374,
       "step": 7510
     },
     {
-      "epoch": 0.26,
-      "learning_rate": 4.993141562528739e-05,
-      "loss": 0.3525,
+      "epoch": 0.27084009082062926,
+      "grad_norm": 0.2130429446697235,
+      "learning_rate": 4.9923586763346674e-05,
+      "loss": 0.4876,
       "step": 7515
     },
     {
-      "epoch": 0.26,
-      "learning_rate": 4.9931204596007386e-05,
-      "loss": 0.314,
+      "epoch": 0.2710202904818539,
+      "grad_norm": 0.18153907358646393,
+      "learning_rate": 4.992335861061354e-05,
+      "loss": 0.4945,
       "step": 7520
     },
     {
-      "epoch": 0.26,
-      "learning_rate": 4.9930993243012114e-05,
-      "loss": 0.3427,
+      "epoch": 0.2712004901430785,
+      "grad_norm": 0.16226467490196228,
+      "learning_rate": 4.992313011830446e-05,
+      "loss": 0.4293,
       "step": 7525
     },
     {
-      "epoch": 0.26,
-      "learning_rate": 4.993078156630433e-05,
-      "loss": 0.3562,
+      "epoch": 0.2713806898043032,
+      "grad_norm": 0.18704953789710999,
+      "learning_rate": 4.992290128642257e-05,
+      "loss": 0.4861,
       "step": 7530
     },
     {
-      "epoch": 0.27,
-      "learning_rate": 4.9930569565886775e-05,
-      "loss": 0.3455,
+      "epoch": 0.2715608894655278,
+      "grad_norm": 0.24006149172782898,
+      "learning_rate": 4.992267211497097e-05,
+      "loss": 0.5093,
       "step": 7535
     },
     {
-      "epoch": 0.27,
-      "learning_rate": 4.993035724176221e-05,
-      "loss": 0.3324,
+      "epoch": 0.27174108912675243,
+      "grad_norm": 0.17990513145923615,
+      "learning_rate": 4.992244260395278e-05,
+      "loss": 0.5006,
       "step": 7540
     },
     {
-      "epoch": 0.27,
-      "learning_rate": 4.993014459393338e-05,
-      "loss": 0.3426,
+      "epoch": 0.27192128878797706,
+      "grad_norm": 0.19324089586734772,
+      "learning_rate": 4.992221275337115e-05,
+      "loss": 0.4631,
       "step": 7545
     },
     {
-      "epoch": 0.27,
-      "learning_rate": 4.992993162240306e-05,
-      "loss": 0.328,
+      "epoch": 0.27210148844920173,
+      "grad_norm": 0.1453637182712555,
+      "learning_rate": 4.992198256322918e-05,
+      "loss": 0.4758,
       "step": 7550
     },
     {
-      "epoch": 0.27,
-      "learning_rate": 4.9929718327174004e-05,
-      "loss": 0.3342,
+      "epoch": 0.27228168811042636,
+      "grad_norm": 0.18295510113239288,
+      "learning_rate": 4.992175203353003e-05,
+      "loss": 0.4832,
       "step": 7555
     },
     {
-      "epoch": 0.27,
-      "learning_rate": 4.992950470824898e-05,
-      "loss": 0.3309,
+      "epoch": 0.272461887771651,
+      "grad_norm": 0.20734231173992157,
+      "learning_rate": 4.992152116427683e-05,
+      "loss": 0.4669,
       "step": 7560
     },
     {
-      "epoch": 0.27,
-      "learning_rate": 4.9929290765630766e-05,
-      "loss": 0.3365,
+      "epoch": 0.2726420874328756,
+      "grad_norm": 0.17811109125614166,
+      "learning_rate": 4.992128995547274e-05,
+      "loss": 0.4197,
       "step": 7565
     },
     {
-      "epoch": 0.27,
-      "learning_rate": 4.9929076499322145e-05,
-      "loss": 0.342,
+      "epoch": 0.2728222870941003,
+      "grad_norm": 0.16642598807811737,
+      "learning_rate": 4.992105840712089e-05,
+      "loss": 0.4575,
       "step": 7570
     },
     {
-      "epoch": 0.27,
-      "learning_rate": 4.9928861909325894e-05,
-      "loss": 0.3079,
+      "epoch": 0.2730024867553249,
+      "grad_norm": 0.2581879794597626,
+      "learning_rate": 4.992082651922444e-05,
+      "loss": 0.4631,
       "step": 7575
     },
     {
-      "epoch": 0.27,
-      "learning_rate": 4.9928646995644804e-05,
-      "loss": 0.3414,
+      "epoch": 0.2731826864165495,
+      "grad_norm": 0.17311778664588928,
+      "learning_rate": 4.992059429178656e-05,
+      "loss": 0.4674,
       "step": 7580
     },
     {
-      "epoch": 0.27,
-      "learning_rate": 4.992843175828165e-05,
-      "loss": 0.3426,
+      "epoch": 0.27336288607777415,
+      "grad_norm": 0.17181158065795898,
+      "learning_rate": 4.992036172481041e-05,
+      "loss": 0.4439,
       "step": 7585
     },
     {
-      "epoch": 0.27,
-      "learning_rate": 4.992821619723924e-05,
-      "loss": 0.3287,
+      "epoch": 0.2735430857389988,
+      "grad_norm": 0.21754592657089233,
+      "learning_rate": 4.992012881829915e-05,
+      "loss": 0.4399,
       "step": 7590
     },
     {
-      "epoch": 0.27,
-      "learning_rate": 4.992800031252038e-05,
-      "loss": 0.3117,
+      "epoch": 0.27372328540022345,
+      "grad_norm": 0.18678173422813416,
+      "learning_rate": 4.9919895572255956e-05,
+      "loss": 0.4932,
       "step": 7595
     },
     {
-      "epoch": 0.27,
-      "learning_rate": 4.9927784104127864e-05,
-      "loss": 0.3147,
+      "epoch": 0.27390348506144807,
+      "grad_norm": 0.16362018883228302,
+      "learning_rate": 4.9919661986684024e-05,
+      "loss": 0.4359,
       "step": 7600
     },
     {
-      "epoch": 0.27,
-      "learning_rate": 4.992756757206449e-05,
-      "loss": 0.3535,
+      "epoch": 0.27408368472267275,
+      "grad_norm": 0.2467060089111328,
+      "learning_rate": 4.991942806158652e-05,
+      "loss": 0.4648,
       "step": 7605
     },
     {
-      "epoch": 0.27,
-      "learning_rate": 4.992735071633308e-05,
-      "loss": 0.3252,
+      "epoch": 0.27426388438389737,
+      "grad_norm": 0.22692877054214478,
+      "learning_rate": 4.991919379696662e-05,
+      "loss": 0.5261,
       "step": 7610
     },
     {
-      "epoch": 0.27,
-      "learning_rate": 4.992713353693646e-05,
-      "loss": 0.3412,
+      "epoch": 0.274444084045122,
+      "grad_norm": 0.1850813329219818,
+      "learning_rate": 4.9918959192827534e-05,
+      "loss": 0.4305,
       "step": 7615
     },
     {
-      "epoch": 0.27,
-      "learning_rate": 4.992691603387742e-05,
-      "loss": 0.3453,
+      "epoch": 0.2746242837063466,
+      "grad_norm": 0.183277890086174,
+      "learning_rate": 4.9918724249172454e-05,
+      "loss": 0.4175,
       "step": 7620
     },
     {
-      "epoch": 0.27,
-      "learning_rate": 4.992669820715882e-05,
-      "loss": 0.3312,
+      "epoch": 0.2748044833675713,
+      "grad_norm": 0.18612994253635406,
+      "learning_rate": 4.9918488966004587e-05,
+      "loss": 0.4802,
       "step": 7625
     },
     {
-      "epoch": 0.27,
-      "learning_rate": 4.9926480056783465e-05,
-      "loss": 0.3548,
+      "epoch": 0.2749846830287959,
+      "grad_norm": 0.1732574850320816,
+      "learning_rate": 4.9918253343327123e-05,
+      "loss": 0.4322,
       "step": 7630
     },
     {
-      "epoch": 0.27,
-      "learning_rate": 4.99262615827542e-05,
-      "loss": 0.3218,
+      "epoch": 0.27516488269002054,
+      "grad_norm": 0.20043861865997314,
+      "learning_rate": 4.991801738114329e-05,
+      "loss": 0.4696,
       "step": 7635
     },
     {
-      "epoch": 0.27,
-      "learning_rate": 4.992604278507385e-05,
-      "loss": 0.3271,
+      "epoch": 0.27534508235124516,
+      "grad_norm": 0.2102193683385849,
+      "learning_rate": 4.991778107945629e-05,
+      "loss": 0.4652,
       "step": 7640
     },
     {
-      "epoch": 0.27,
-      "learning_rate": 4.9925823663745255e-05,
-      "loss": 0.3429,
+      "epoch": 0.27552528201246984,
+      "grad_norm": 0.20501478016376495,
+      "learning_rate": 4.9917544438269346e-05,
+      "loss": 0.4636,
       "step": 7645
     },
     {
-      "epoch": 0.27,
-      "learning_rate": 4.992560421877127e-05,
-      "loss": 0.3242,
+      "epoch": 0.27570548167369446,
+      "grad_norm": 0.17633172869682312,
+      "learning_rate": 4.991730745758568e-05,
+      "loss": 0.4995,
       "step": 7650
     },
     {
-      "epoch": 0.27,
-      "learning_rate": 4.992538445015474e-05,
-      "loss": 0.3192,
+      "epoch": 0.2758856813349191,
+      "grad_norm": 0.18572551012039185,
+      "learning_rate": 4.991707013740853e-05,
+      "loss": 0.4752,
       "step": 7655
     },
     {
-      "epoch": 0.27,
-      "learning_rate": 4.992516435789852e-05,
-      "loss": 0.3249,
+      "epoch": 0.2760658809961437,
+      "grad_norm": 0.14463745057582855,
+      "learning_rate": 4.991683247774113e-05,
+      "loss": 0.4448,
       "step": 7660
     },
     {
-      "epoch": 0.27,
-      "learning_rate": 4.992494394200548e-05,
-      "loss": 0.3245,
+      "epoch": 0.2762460806573684,
+      "grad_norm": 0.2664778232574463,
+      "learning_rate": 4.99165944785867e-05,
+      "loss": 0.4605,
       "step": 7665
     },
     {
-      "epoch": 0.27,
-      "learning_rate": 4.992472320247845e-05,
-      "loss": 0.3199,
+      "epoch": 0.276426280318593,
+      "grad_norm": 0.20322363078594208,
+      "learning_rate": 4.991635613994849e-05,
+      "loss": 0.4676,
       "step": 7670
     },
     {
-      "epoch": 0.27,
-      "learning_rate": 4.992450213932033e-05,
-      "loss": 0.3396,
+      "epoch": 0.27660647997981763,
+      "grad_norm": 0.21015766263008118,
+      "learning_rate": 4.991611746182977e-05,
+      "loss": 0.4449,
       "step": 7675
     },
     {
-      "epoch": 0.27,
-      "learning_rate": 4.9924280752533965e-05,
-      "loss": 0.3112,
+      "epoch": 0.27678667964104225,
+      "grad_norm": 0.23488974571228027,
+      "learning_rate": 4.991587844423376e-05,
+      "loss": 0.4613,
       "step": 7680
     },
     {
-      "epoch": 0.27,
-      "learning_rate": 4.992405904212224e-05,
-      "loss": 0.3247,
+      "epoch": 0.27696687930226693,
+      "grad_norm": 0.2240959256887436,
+      "learning_rate": 4.9915639087163736e-05,
+      "loss": 0.4601,
       "step": 7685
     },
     {
-      "epoch": 0.27,
-      "learning_rate": 4.9923837008088046e-05,
-      "loss": 0.3239,
+      "epoch": 0.27714707896349156,
+      "grad_norm": 0.18855224549770355,
+      "learning_rate": 4.991539939062295e-05,
+      "loss": 0.4584,
       "step": 7690
     },
     {
-      "epoch": 0.27,
-      "learning_rate": 4.9923614650434245e-05,
-      "loss": 0.3494,
+      "epoch": 0.2773272786247162,
+      "grad_norm": 0.20485761761665344,
+      "learning_rate": 4.9915159354614674e-05,
+      "loss": 0.4444,
       "step": 7695
     },
     {
-      "epoch": 0.27,
-      "learning_rate": 4.992339196916373e-05,
-      "loss": 0.3242,
+      "epoch": 0.2775074782859408,
+      "grad_norm": 0.18977908790111542,
+      "learning_rate": 4.9914918979142163e-05,
+      "loss": 0.4626,
       "step": 7700
     },
     {
-      "epoch": 0.27,
-      "learning_rate": 4.99231689642794e-05,
-      "loss": 0.3402,
+      "epoch": 0.2776876779471655,
+      "grad_norm": 0.15290701389312744,
+      "learning_rate": 4.991467826420872e-05,
+      "loss": 0.4448,
       "step": 7705
     },
     {
-      "epoch": 0.27,
-      "learning_rate": 4.992294563578415e-05,
-      "loss": 0.3341,
+      "epoch": 0.2778678776083901,
+      "grad_norm": 0.22318267822265625,
+      "learning_rate": 4.99144372098176e-05,
+      "loss": 0.4937,
       "step": 7710
     },
     {
-      "epoch": 0.27,
-      "learning_rate": 4.992272198368086e-05,
-      "loss": 0.3508,
+      "epoch": 0.2780480772696147,
+      "grad_norm": 0.2257704883813858,
+      "learning_rate": 4.9914195815972104e-05,
+      "loss": 0.4445,
       "step": 7715
     },
     {
-      "epoch": 0.27,
-      "learning_rate": 4.992249800797247e-05,
-      "loss": 0.3563,
+      "epoch": 0.27822827693083935,
+      "grad_norm": 0.2357822060585022,
+      "learning_rate": 4.991395408267551e-05,
+      "loss": 0.4546,
       "step": 7720
     },
     {
-      "epoch": 0.27,
-      "learning_rate": 4.992227370866186e-05,
-      "loss": 0.3281,
+      "epoch": 0.278408476592064,
+      "grad_norm": 0.17252907156944275,
+      "learning_rate": 4.991371200993111e-05,
+      "loss": 0.445,
       "step": 7725
     },
     {
-      "epoch": 0.27,
-      "learning_rate": 4.992204908575194e-05,
-      "loss": 0.3573,
+      "epoch": 0.27858867625328865,
+      "grad_norm": 0.16278617084026337,
+      "learning_rate": 4.991346959774221e-05,
+      "loss": 0.4839,
       "step": 7730
     },
     {
-      "epoch": 0.27,
-      "learning_rate": 4.9921824139245656e-05,
-      "loss": 0.3089,
+      "epoch": 0.27876887591451327,
+      "grad_norm": 0.1887577474117279,
+      "learning_rate": 4.9913226846112114e-05,
+      "loss": 0.4232,
       "step": 7735
     },
     {
-      "epoch": 0.27,
-      "learning_rate": 4.992159886914589e-05,
-      "loss": 0.3237,
+      "epoch": 0.2789490755757379,
+      "grad_norm": 0.17083100974559784,
+      "learning_rate": 4.991298375504413e-05,
+      "loss": 0.4863,
       "step": 7740
     },
     {
-      "epoch": 0.27,
-      "learning_rate": 4.9921373275455594e-05,
-      "loss": 0.3421,
+      "epoch": 0.27912927523696257,
+      "grad_norm": 0.1833745837211609,
+      "learning_rate": 4.991274032454156e-05,
+      "loss": 0.4544,
       "step": 7745
     },
     {
-      "epoch": 0.27,
-      "learning_rate": 4.99211473581777e-05,
-      "loss": 0.3322,
+      "epoch": 0.2793094748981872,
+      "grad_norm": 0.15337827801704407,
+      "learning_rate": 4.991249655460773e-05,
+      "loss": 0.4621,
       "step": 7750
     },
     {
-      "epoch": 0.27,
-      "learning_rate": 4.992092111731512e-05,
-      "loss": 0.3562,
+      "epoch": 0.2794896745594118,
+      "grad_norm": 0.22126014530658722,
+      "learning_rate": 4.991225244524595e-05,
+      "loss": 0.4861,
       "step": 7755
     },
     {
-      "epoch": 0.27,
-      "learning_rate": 4.992069455287081e-05,
-      "loss": 0.3329,
+      "epoch": 0.27966987422063644,
+      "grad_norm": 0.203162282705307,
+      "learning_rate": 4.991200799645955e-05,
+      "loss": 0.4432,
       "step": 7760
     },
     {
-      "epoch": 0.27,
-      "learning_rate": 4.99204676648477e-05,
-      "loss": 0.3719,
+      "epoch": 0.2798500738818611,
+      "grad_norm": 0.1990215927362442,
+      "learning_rate": 4.991176320825188e-05,
+      "loss": 0.4583,
       "step": 7765
     },
     {
-      "epoch": 0.27,
-      "learning_rate": 4.992024045324874e-05,
-      "loss": 0.3369,
+      "epoch": 0.28003027354308574,
+      "grad_norm": 0.18323101103305817,
+      "learning_rate": 4.991151808062625e-05,
+      "loss": 0.4382,
       "step": 7770
     },
     {
-      "epoch": 0.27,
-      "learning_rate": 4.992001291807688e-05,
-      "loss": 0.3362,
+      "epoch": 0.28021047320431036,
+      "grad_norm": 0.1830572485923767,
+      "learning_rate": 4.9911272613586006e-05,
+      "loss": 0.4663,
       "step": 7775
     },
     {
-      "epoch": 0.27,
-      "learning_rate": 4.991978505933508e-05,
-      "loss": 0.3308,
+      "epoch": 0.28039067286553504,
+      "grad_norm": 0.1886363923549652,
+      "learning_rate": 4.99110268071345e-05,
+      "loss": 0.4462,
       "step": 7780
     },
     {
-      "epoch": 0.27,
-      "learning_rate": 4.9919556877026284e-05,
-      "loss": 0.3392,
+      "epoch": 0.28057087252675966,
+      "grad_norm": 0.15756884217262268,
+      "learning_rate": 4.991078066127508e-05,
+      "loss": 0.4707,
       "step": 7785
     },
     {
-      "epoch": 0.27,
-      "learning_rate": 4.991932837115347e-05,
-      "loss": 0.338,
+      "epoch": 0.2807510721879843,
+      "grad_norm": 0.17909304797649384,
+      "learning_rate": 4.991053417601109e-05,
+      "loss": 0.497,
       "step": 7790
     },
     {
-      "epoch": 0.27,
-      "learning_rate": 4.991909954171961e-05,
-      "loss": 0.3218,
+      "epoch": 0.2809312718492089,
+      "grad_norm": 0.22559069097042084,
+      "learning_rate": 4.99102873513459e-05,
+      "loss": 0.5026,
       "step": 7795
     },
     {
-      "epoch": 0.27,
-      "learning_rate": 4.991887038872766e-05,
-      "loss": 0.3144,
+      "epoch": 0.2811114715104336,
+      "grad_norm": 0.19916412234306335,
+      "learning_rate": 4.991004018728286e-05,
+      "loss": 0.4825,
       "step": 7800
     },
     {
-      "epoch": 0.27,
-      "learning_rate": 4.991864091218059e-05,
-      "loss": 0.3482,
+      "epoch": 0.2812916711716582,
+      "grad_norm": 0.23998232185840607,
+      "learning_rate": 4.990979268382535e-05,
+      "loss": 0.5141,
       "step": 7805
     },
     {
-      "epoch": 0.27,
-      "learning_rate": 4.9918411112081395e-05,
-      "loss": 0.3249,
+      "epoch": 0.28147187083288283,
+      "grad_norm": 0.17623983323574066,
+      "learning_rate": 4.9909544840976744e-05,
+      "loss": 0.4464,
       "step": 7810
     },
     {
-      "epoch": 0.27,
-      "learning_rate": 4.991818098843305e-05,
-      "loss": 0.3433,
+      "epoch": 0.28165207049410745,
+      "grad_norm": 0.2640053927898407,
+      "learning_rate": 4.99092966587404e-05,
+      "loss": 0.4738,
       "step": 7815
     },
     {
-      "epoch": 0.28,
-      "learning_rate": 4.991795054123856e-05,
-      "loss": 0.3122,
+      "epoch": 0.28183227015533213,
+      "grad_norm": 0.15781162679195404,
+      "learning_rate": 4.990904813711972e-05,
+      "loss": 0.4283,
       "step": 7820
     },
     {
-      "epoch": 0.28,
-      "learning_rate": 4.991771977050089e-05,
-      "loss": 0.3412,
+      "epoch": 0.28201246981655675,
+      "grad_norm": 0.15870793163776398,
+      "learning_rate": 4.990879927611808e-05,
+      "loss": 0.4477,
       "step": 7825
     },
     {
-      "epoch": 0.28,
-      "learning_rate": 4.991748867622306e-05,
-      "loss": 0.3673,
+      "epoch": 0.2821926694777814,
+      "grad_norm": 0.23389936983585358,
+      "learning_rate": 4.990855007573887e-05,
+      "loss": 0.4772,
       "step": 7830
     },
     {
-      "epoch": 0.28,
-      "learning_rate": 4.991725725840805e-05,
-      "loss": 0.3109,
+      "epoch": 0.282372869139006,
+      "grad_norm": 0.24188333749771118,
+      "learning_rate": 4.9908300535985486e-05,
+      "loss": 0.4548,
       "step": 7835
     },
     {
-      "epoch": 0.28,
-      "learning_rate": 4.991702551705888e-05,
-      "loss": 0.3558,
+      "epoch": 0.2825530688002307,
+      "grad_norm": 0.1669192612171173,
+      "learning_rate": 4.990805065686133e-05,
+      "loss": 0.4476,
       "step": 7840
     },
     {
-      "epoch": 0.28,
-      "learning_rate": 4.9916793452178554e-05,
-      "loss": 0.3088,
+      "epoch": 0.2827332684614553,
+      "grad_norm": 0.15673232078552246,
+      "learning_rate": 4.990780043836981e-05,
+      "loss": 0.4181,
       "step": 7845
     },
     {
-      "epoch": 0.28,
-      "learning_rate": 4.9916561063770084e-05,
-      "loss": 0.3375,
+      "epoch": 0.2829134681226799,
+      "grad_norm": 0.146693155169487,
+      "learning_rate": 4.9907549880514334e-05,
+      "loss": 0.4416,
       "step": 7850
     },
     {
-      "epoch": 0.28,
-      "learning_rate": 4.991632835183649e-05,
-      "loss": 0.3374,
+      "epoch": 0.28309366778390455,
+      "grad_norm": 0.24487614631652832,
+      "learning_rate": 4.990729898329831e-05,
+      "loss": 0.4169,
       "step": 7855
     },
     {
-      "epoch": 0.28,
-      "learning_rate": 4.99160953163808e-05,
-      "loss": 0.3386,
+      "epoch": 0.2832738674451292,
+      "grad_norm": 0.2468927651643753,
+      "learning_rate": 4.9907047746725154e-05,
+      "loss": 0.4804,
       "step": 7860
     },
     {
-      "epoch": 0.28,
-      "learning_rate": 4.991586195740603e-05,
-      "loss": 0.3391,
+      "epoch": 0.28345406710635385,
+      "grad_norm": 0.22746756672859192,
+      "learning_rate": 4.99067961707983e-05,
+      "loss": 0.4499,
       "step": 7865
     },
     {
-      "epoch": 0.28,
-      "learning_rate": 4.991562827491521e-05,
-      "loss": 0.2986,
+      "epoch": 0.28363426676757847,
+      "grad_norm": 0.17610248923301697,
+      "learning_rate": 4.990654425552117e-05,
+      "loss": 0.4824,
       "step": 7870
     },
     {
-      "epoch": 0.28,
-      "learning_rate": 4.9915394268911375e-05,
-      "loss": 0.3754,
+      "epoch": 0.2838144664288031,
+      "grad_norm": 0.20157018303871155,
+      "learning_rate": 4.9906292000897196e-05,
+      "loss": 0.4641,
       "step": 7875
     },
     {
-      "epoch": 0.28,
-      "learning_rate": 4.9915159939397574e-05,
-      "loss": 0.315,
+      "epoch": 0.28399466609002777,
+      "grad_norm": 0.23462137579917908,
+      "learning_rate": 4.990603940692982e-05,
+      "loss": 0.4813,
       "step": 7880
     },
     {
-      "epoch": 0.28,
-      "learning_rate": 4.991492528637683e-05,
-      "loss": 0.3335,
+      "epoch": 0.2841748657512524,
+      "grad_norm": 0.19886912405490875,
+      "learning_rate": 4.990578647362247e-05,
+      "loss": 0.4413,
       "step": 7885
     },
     {
-      "epoch": 0.28,
-      "learning_rate": 4.9914690309852205e-05,
-      "loss": 0.3211,
+      "epoch": 0.284355065412477,
+      "grad_norm": 0.17550869286060333,
+      "learning_rate": 4.9905533200978606e-05,
+      "loss": 0.4518,
       "step": 7890
     },
     {
-      "epoch": 0.28,
-      "learning_rate": 4.991445500982675e-05,
-      "loss": 0.3403,
+      "epoch": 0.28453526507370164,
+      "grad_norm": 0.17420583963394165,
+      "learning_rate": 4.9905279589001674e-05,
+      "loss": 0.4741,
       "step": 7895
     },
     {
-      "epoch": 0.28,
-      "learning_rate": 4.9914219386303513e-05,
-      "loss": 0.3454,
+      "epoch": 0.2847154647349263,
+      "grad_norm": 0.2519500255584717,
+      "learning_rate": 4.990502563769514e-05,
+      "loss": 0.4974,
       "step": 7900
     },
     {
-      "epoch": 0.28,
-      "learning_rate": 4.991398343928555e-05,
-      "loss": 0.3164,
+      "epoch": 0.28489566439615094,
+      "grad_norm": 0.21269458532333374,
+      "learning_rate": 4.990477134706244e-05,
+      "loss": 0.527,
       "step": 7905
     },
     {
-      "epoch": 0.28,
-      "learning_rate": 4.9913747168775934e-05,
-      "loss": 0.3341,
+      "epoch": 0.28507586405737556,
+      "grad_norm": 0.2123083919286728,
+      "learning_rate": 4.990451671710705e-05,
+      "loss": 0.4385,
       "step": 7910
     },
     {
-      "epoch": 0.28,
-      "learning_rate": 4.991351057477774e-05,
-      "loss": 0.3248,
+      "epoch": 0.2852560637186002,
+      "grad_norm": 0.17294515669345856,
+      "learning_rate": 4.990426174783245e-05,
+      "loss": 0.4442,
       "step": 7915
     },
     {
-      "epoch": 0.28,
-      "learning_rate": 4.9913273657294025e-05,
-      "loss": 0.3214,
+      "epoch": 0.28543626337982486,
+      "grad_norm": 0.14767326414585114,
+      "learning_rate": 4.99040064392421e-05,
+      "loss": 0.4443,
       "step": 7920
     },
     {
-      "epoch": 0.28,
-      "learning_rate": 4.991303641632786e-05,
-      "loss": 0.3329,
+      "epoch": 0.2856164630410495,
+      "grad_norm": 0.19468598067760468,
+      "learning_rate": 4.9903750791339485e-05,
+      "loss": 0.475,
       "step": 7925
     },
     {
-      "epoch": 0.28,
-      "learning_rate": 4.991279885188235e-05,
-      "loss": 0.3253,
+      "epoch": 0.2857966627022741,
+      "grad_norm": 0.20397444069385529,
+      "learning_rate": 4.990349480412809e-05,
+      "loss": 0.4989,
       "step": 7930
     },
     {
-      "epoch": 0.28,
-      "learning_rate": 4.991256096396055e-05,
-      "loss": 0.3295,
+      "epoch": 0.28597686236349873,
+      "grad_norm": 0.18751677870750427,
+      "learning_rate": 4.990323847761139e-05,
+      "loss": 0.46,
       "step": 7935
     },
     {
-      "epoch": 0.28,
-      "learning_rate": 4.991232275256558e-05,
-      "loss": 0.3242,
+      "epoch": 0.2861570620247234,
+      "grad_norm": 0.20503003895282745,
+      "learning_rate": 4.99029818117929e-05,
+      "loss": 0.4274,
       "step": 7940
     },
     {
-      "epoch": 0.28,
-      "learning_rate": 4.991208421770051e-05,
-      "loss": 0.3328,
+      "epoch": 0.28633726168594803,
+      "grad_norm": 0.20516186952590942,
+      "learning_rate": 4.9902724806676094e-05,
+      "loss": 0.4503,
       "step": 7945
     },
     {
-      "epoch": 0.28,
-      "learning_rate": 4.9911845359368434e-05,
-      "loss": 0.3206,
+      "epoch": 0.28651746134717265,
+      "grad_norm": 0.21163448691368103,
+      "learning_rate": 4.990246746226449e-05,
+      "loss": 0.5434,
       "step": 7950
     },
     {
-      "epoch": 0.28,
-      "learning_rate": 4.9911606177572476e-05,
-      "loss": 0.3497,
+      "epoch": 0.28669766100839733,
+      "grad_norm": 0.21273358166217804,
+      "learning_rate": 4.9902209778561585e-05,
+      "loss": 0.4581,
       "step": 7955
     },
     {
-      "epoch": 0.28,
-      "learning_rate": 4.991136667231572e-05,
-      "loss": 0.3293,
+      "epoch": 0.28687786066962195,
+      "grad_norm": 0.2407582700252533,
+      "learning_rate": 4.9901951755570896e-05,
+      "loss": 0.4555,
       "step": 7960
     },
     {
-      "epoch": 0.28,
-      "learning_rate": 4.991112684360129e-05,
-      "loss": 0.3348,
+      "epoch": 0.2870580603308466,
+      "grad_norm": 0.19239749014377594,
+      "learning_rate": 4.9901693393295935e-05,
+      "loss": 0.4809,
       "step": 7965
     },
     {
-      "epoch": 0.28,
-      "learning_rate": 4.99108866914323e-05,
-      "loss": 0.3222,
+      "epoch": 0.2872382599920712,
+      "grad_norm": 0.20676837861537933,
+      "learning_rate": 4.990143469174022e-05,
+      "loss": 0.4315,
       "step": 7970
     },
     {
-      "epoch": 0.28,
-      "learning_rate": 4.9910646215811854e-05,
-      "loss": 0.3057,
+      "epoch": 0.2874184596532959,
+      "grad_norm": 0.26478490233421326,
+      "learning_rate": 4.990117565090728e-05,
+      "loss": 0.4561,
       "step": 7975
     },
     {
-      "epoch": 0.28,
-      "learning_rate": 4.9910405416743086e-05,
-      "loss": 0.3665,
+      "epoch": 0.2875986593145205,
+      "grad_norm": 0.19174893200397491,
+      "learning_rate": 4.990091627080065e-05,
+      "loss": 0.525,
       "step": 7980
     },
     {
-      "epoch": 0.28,
-      "learning_rate": 4.991016429422912e-05,
-      "loss": 0.329,
+      "epoch": 0.2877788589757451,
+      "grad_norm": 0.18685384094715118,
+      "learning_rate": 4.9900656551423844e-05,
+      "loss": 0.4683,
       "step": 7985
     },
     {
-      "epoch": 0.28,
-      "learning_rate": 4.990992284827309e-05,
-      "loss": 0.3296,
+      "epoch": 0.28795905863696974,
+      "grad_norm": 0.17788560688495636,
+      "learning_rate": 4.990039649278042e-05,
+      "loss": 0.4609,
       "step": 7990
     },
     {
-      "epoch": 0.28,
-      "learning_rate": 4.9909681078878136e-05,
-      "loss": 0.3494,
+      "epoch": 0.2881392582981944,
+      "grad_norm": 0.2256275713443756,
+      "learning_rate": 4.990013609487391e-05,
+      "loss": 0.4687,
       "step": 7995
     },
     {
-      "epoch": 0.28,
-      "learning_rate": 4.9909438986047375e-05,
-      "loss": 0.3159,
+      "epoch": 0.28831945795941905,
+      "grad_norm": 0.2363923341035843,
+      "learning_rate": 4.989987535770787e-05,
+      "loss": 0.4783,
       "step": 8000
     },
     {
-      "epoch": 0.28,
-      "eval_loss": 0.3288015127182007,
-      "eval_runtime": 10.5609,
-      "eval_samples_per_second": 9.469,
-      "eval_steps_per_second": 9.469,
+      "epoch": 0.28831945795941905,
+      "eval_loss": 0.48639407753944397,
+      "eval_runtime": 3.5338,
+      "eval_samples_per_second": 28.298,
+      "eval_steps_per_second": 7.075,
       "step": 8000
     },
     {
-      "epoch": 0.28,
-      "learning_rate": 4.9909196569783975e-05,
-      "loss": 0.3127,
+      "epoch": 0.28849965762064367,
+      "grad_norm": 0.20103095471858978,
+      "learning_rate": 4.9899614281285856e-05,
+      "loss": 0.4766,
       "step": 8005
     },
     {
-      "epoch": 0.28,
-      "learning_rate": 4.9908953830091066e-05,
-      "loss": 0.3201,
+      "epoch": 0.2886798572818683,
+      "grad_norm": 0.19780664145946503,
+      "learning_rate": 4.98993528656114e-05,
+      "loss": 0.5028,
       "step": 8010
     },
     {
-      "epoch": 0.28,
-      "learning_rate": 4.990871076697181e-05,
-      "loss": 0.3433,
+      "epoch": 0.28886005694309297,
+      "grad_norm": 0.19557078182697296,
+      "learning_rate": 4.9899091110688104e-05,
+      "loss": 0.4693,
       "step": 8015
     },
     {
-      "epoch": 0.28,
-      "learning_rate": 4.990846738042936e-05,
-      "loss": 0.3307,
+      "epoch": 0.2890402566043176,
+      "grad_norm": 0.18514655530452728,
+      "learning_rate": 4.98988290165195e-05,
+      "loss": 0.4637,
       "step": 8020
     },
     {
-      "epoch": 0.28,
-      "learning_rate": 4.990822367046688e-05,
-      "loss": 0.3218,
+      "epoch": 0.2892204562655422,
+      "grad_norm": 0.14113621413707733,
+      "learning_rate": 4.9898566583109174e-05,
+      "loss": 0.4324,
       "step": 8025
     },
     {
-      "epoch": 0.28,
-      "learning_rate": 4.990797963708752e-05,
-      "loss": 0.3419,
+      "epoch": 0.28940065592676684,
+      "grad_norm": 0.20794130861759186,
+      "learning_rate": 4.989830381046069e-05,
+      "loss": 0.5302,
       "step": 8030
     },
     {
-      "epoch": 0.28,
-      "learning_rate": 4.990773528029446e-05,
-      "loss": 0.338,
+      "epoch": 0.2895808555879915,
+      "grad_norm": 0.18446312844753265,
+      "learning_rate": 4.9898040698577655e-05,
+      "loss": 0.4497,
       "step": 8035
     },
     {
-      "epoch": 0.28,
-      "learning_rate": 4.990749060009089e-05,
-      "loss": 0.3297,
+      "epoch": 0.28976105524921614,
+      "grad_norm": 0.20031079649925232,
+      "learning_rate": 4.9897777247463615e-05,
+      "loss": 0.4522,
       "step": 8040
     },
     {
-      "epoch": 0.28,
-      "learning_rate": 4.9907245596479956e-05,
-      "loss": 0.3106,
+      "epoch": 0.28994125491044076,
+      "grad_norm": 0.21210455894470215,
+      "learning_rate": 4.989751345712219e-05,
+      "loss": 0.4408,
       "step": 8045
     },
     {
-      "epoch": 0.28,
-      "learning_rate": 4.990700026946485e-05,
-      "loss": 0.3051,
+      "epoch": 0.2901214545716654,
+      "grad_norm": 0.18559224903583527,
+      "learning_rate": 4.989724932755697e-05,
+      "loss": 0.4555,
       "step": 8050
     },
     {
-      "epoch": 0.28,
-      "learning_rate": 4.9906754619048767e-05,
-      "loss": 0.3461,
+      "epoch": 0.29030165423289006,
+      "grad_norm": 0.21420493721961975,
+      "learning_rate": 4.9896984858771546e-05,
+      "loss": 0.4763,
       "step": 8055
     },
     {
-      "epoch": 0.28,
-      "learning_rate": 4.9906508645234885e-05,
-      "loss": 0.323,
+      "epoch": 0.2904818538941147,
+      "grad_norm": 0.1991835981607437,
+      "learning_rate": 4.9896720050769516e-05,
+      "loss": 0.4475,
       "step": 8060
     },
     {
-      "epoch": 0.28,
-      "learning_rate": 4.990626234802641e-05,
-      "loss": 0.3196,
+      "epoch": 0.2906620535553393,
+      "grad_norm": 0.23717185854911804,
+      "learning_rate": 4.98964549035545e-05,
+      "loss": 0.4635,
       "step": 8065
     },
     {
-      "epoch": 0.28,
-      "learning_rate": 4.990601572742652e-05,
-      "loss": 0.3257,
+      "epoch": 0.29084225321656393,
+      "grad_norm": 0.2004196047782898,
+      "learning_rate": 4.989618941713011e-05,
+      "loss": 0.4411,
       "step": 8070
     },
     {
-      "epoch": 0.28,
-      "learning_rate": 4.990576878343844e-05,
-      "loss": 0.3501,
+      "epoch": 0.2910224528777886,
+      "grad_norm": 0.21253125369548798,
+      "learning_rate": 4.9895923591499954e-05,
+      "loss": 0.4925,
       "step": 8075
     },
     {
-      "epoch": 0.28,
-      "learning_rate": 4.9905521516065356e-05,
-      "loss": 0.3417,
+      "epoch": 0.29120265253901323,
+      "grad_norm": 0.253714382648468,
+      "learning_rate": 4.9895657426667666e-05,
+      "loss": 0.4723,
       "step": 8080
     },
     {
-      "epoch": 0.28,
-      "learning_rate": 4.99052739253105e-05,
-      "loss": 0.3415,
+      "epoch": 0.29138285220023785,
+      "grad_norm": 0.21997293829917908,
+      "learning_rate": 4.9895390922636854e-05,
+      "loss": 0.4838,
       "step": 8085
     },
     {
-      "epoch": 0.28,
-      "learning_rate": 4.990502601117707e-05,
-      "loss": 0.3378,
+      "epoch": 0.2915630518614625,
+      "grad_norm": 0.17473578453063965,
+      "learning_rate": 4.989512407941117e-05,
+      "loss": 0.4264,
       "step": 8090
     },
     {
-      "epoch": 0.28,
-      "learning_rate": 4.990477777366829e-05,
-      "loss": 0.3518,
+      "epoch": 0.29174325152268715,
+      "grad_norm": 0.1931363344192505,
+      "learning_rate": 4.989485689699423e-05,
+      "loss": 0.485,
       "step": 8095
     },
     {
-      "epoch": 0.28,
-      "learning_rate": 4.990452921278738e-05,
-      "loss": 0.35,
+      "epoch": 0.2919234511839118,
+      "grad_norm": 0.16745588183403015,
+      "learning_rate": 4.98945893753897e-05,
+      "loss": 0.4865,
       "step": 8100
     },
     {
-      "epoch": 0.29,
-      "learning_rate": 4.990428032853757e-05,
-      "loss": 0.3323,
+      "epoch": 0.2921036508451364,
+      "grad_norm": 0.19875845313072205,
+      "learning_rate": 4.98943215146012e-05,
+      "loss": 0.4319,
       "step": 8105
     },
     {
-      "epoch": 0.29,
-      "learning_rate": 4.99040311209221e-05,
-      "loss": 0.3318,
+      "epoch": 0.292283850506361,
+      "grad_norm": 0.21367180347442627,
+      "learning_rate": 4.989405331463239e-05,
+      "loss": 0.4779,
       "step": 8110
     },
     {
-      "epoch": 0.29,
-      "learning_rate": 4.9903781589944193e-05,
-      "loss": 0.3153,
+      "epoch": 0.2924640501675857,
+      "grad_norm": 0.16339702904224396,
+      "learning_rate": 4.989378477548692e-05,
+      "loss": 0.4386,
       "step": 8115
     },
     {
-      "epoch": 0.29,
-      "learning_rate": 4.99035317356071e-05,
-      "loss": 0.312,
+      "epoch": 0.2926442498288103,
+      "grad_norm": 0.15709401667118073,
+      "learning_rate": 4.9893515897168455e-05,
+      "loss": 0.4246,
       "step": 8120
     },
     {
-      "epoch": 0.29,
-      "learning_rate": 4.990328155791405e-05,
-      "loss": 0.3254,
+      "epoch": 0.29282444949003494,
+      "grad_norm": 0.23660311102867126,
+      "learning_rate": 4.989324667968066e-05,
+      "loss": 0.4658,
       "step": 8125
     },
     {
-      "epoch": 0.29,
-      "learning_rate": 4.9903031056868306e-05,
-      "loss": 0.3438,
+      "epoch": 0.2930046491512596,
+      "grad_norm": 0.17152658104896545,
+      "learning_rate": 4.9892977123027194e-05,
+      "loss": 0.4546,
       "step": 8130
     },
     {
-      "epoch": 0.29,
-      "learning_rate": 4.990278023247311e-05,
-      "loss": 0.331,
+      "epoch": 0.29318484881248424,
+      "grad_norm": 0.25571581721305847,
+      "learning_rate": 4.989270722721173e-05,
+      "loss": 0.4507,
       "step": 8135
     },
     {
-      "epoch": 0.29,
-      "learning_rate": 4.990252908473174e-05,
-      "loss": 0.3616,
+      "epoch": 0.29336504847370887,
+      "grad_norm": 0.16954413056373596,
+      "learning_rate": 4.989243699223796e-05,
+      "loss": 0.4685,
       "step": 8140
     },
     {
-      "epoch": 0.29,
-      "learning_rate": 4.990227761364743e-05,
-      "loss": 0.3103,
+      "epoch": 0.2935452481349335,
+      "grad_norm": 0.19579976797103882,
+      "learning_rate": 4.989216641810955e-05,
+      "loss": 0.5074,
       "step": 8145
     },
     {
-      "epoch": 0.29,
-      "learning_rate": 4.990202581922345e-05,
-      "loss": 0.3174,
+      "epoch": 0.29372544779615817,
+      "grad_norm": 0.16534093022346497,
+      "learning_rate": 4.989189550483019e-05,
+      "loss": 0.4594,
       "step": 8150
     },
     {
-      "epoch": 0.29,
-      "learning_rate": 4.990177370146308e-05,
-      "loss": 0.3259,
+      "epoch": 0.2939056474573828,
+      "grad_norm": 0.19394879043102264,
+      "learning_rate": 4.9891624252403574e-05,
+      "loss": 0.4696,
       "step": 8155
     },
     {
-      "epoch": 0.29,
-      "learning_rate": 4.99015212603696e-05,
-      "loss": 0.3287,
+      "epoch": 0.2940858471186074,
+      "grad_norm": 0.2109389305114746,
+      "learning_rate": 4.98913526608334e-05,
+      "loss": 0.4606,
       "step": 8160
     },
     {
-      "epoch": 0.29,
-      "learning_rate": 4.990126849594627e-05,
-      "loss": 0.3591,
+      "epoch": 0.29426604677983204,
+      "grad_norm": 0.19720369577407837,
+      "learning_rate": 4.9891080730123365e-05,
+      "loss": 0.4608,
       "step": 8165
     },
     {
-      "epoch": 0.29,
-      "learning_rate": 4.9901015408196375e-05,
-      "loss": 0.3171,
+      "epoch": 0.2944462464410567,
+      "grad_norm": 0.16822995245456696,
+      "learning_rate": 4.9890808460277163e-05,
+      "loss": 0.4925,
       "step": 8170
     },
     {
-      "epoch": 0.29,
-      "learning_rate": 4.9900761997123216e-05,
-      "loss": 0.3008,
+      "epoch": 0.29462644610228134,
+      "grad_norm": 0.1735120564699173,
+      "learning_rate": 4.9890535851298526e-05,
+      "loss": 0.4465,
       "step": 8175
     },
     {
-      "epoch": 0.29,
-      "learning_rate": 4.990050826273006e-05,
-      "loss": 0.3335,
+      "epoch": 0.29480664576350596,
+      "grad_norm": 0.16720019280910492,
+      "learning_rate": 4.989026290319115e-05,
+      "loss": 0.4391,
       "step": 8180
     },
     {
-      "epoch": 0.29,
-      "learning_rate": 4.990025420502022e-05,
-      "loss": 0.3139,
+      "epoch": 0.2949868454247306,
+      "grad_norm": 0.19565287232398987,
+      "learning_rate": 4.988998961595876e-05,
+      "loss": 0.4996,
       "step": 8185
     },
     {
-      "epoch": 0.29,
-      "learning_rate": 4.989999982399699e-05,
-      "loss": 0.3203,
+      "epoch": 0.29516704508595526,
+      "grad_norm": 0.17741116881370544,
+      "learning_rate": 4.988971598960509e-05,
+      "loss": 0.4652,
       "step": 8190
     },
     {
-      "epoch": 0.29,
-      "learning_rate": 4.9899745119663674e-05,
-      "loss": 0.3375,
+      "epoch": 0.2953472447471799,
+      "grad_norm": 0.16898617148399353,
+      "learning_rate": 4.988944202413386e-05,
+      "loss": 0.478,
       "step": 8195
     },
     {
-      "epoch": 0.29,
-      "learning_rate": 4.989949009202357e-05,
-      "loss": 0.362,
+      "epoch": 0.2955274444084045,
+      "grad_norm": 0.2120852917432785,
+      "learning_rate": 4.988916771954879e-05,
+      "loss": 0.4719,
       "step": 8200
     },
     {
-      "epoch": 0.29,
-      "learning_rate": 4.9899234741080006e-05,
-      "loss": 0.313,
+      "epoch": 0.2957076440696291,
+      "grad_norm": 0.16278114914894104,
+      "learning_rate": 4.988889307585364e-05,
+      "loss": 0.4799,
       "step": 8205
     },
     {
-      "epoch": 0.29,
-      "learning_rate": 4.989897906683628e-05,
-      "loss": 0.3655,
+      "epoch": 0.2958878437308538,
+      "grad_norm": 0.18301096558570862,
+      "learning_rate": 4.988861809305213e-05,
+      "loss": 0.4712,
       "step": 8210
     },
     {
-      "epoch": 0.29,
-      "learning_rate": 4.9898723069295725e-05,
-      "loss": 0.349,
+      "epoch": 0.29606804339207843,
+      "grad_norm": 0.15939167141914368,
+      "learning_rate": 4.988834277114802e-05,
+      "loss": 0.4389,
       "step": 8215
     },
     {
-      "epoch": 0.29,
-      "learning_rate": 4.9898466748461653e-05,
-      "loss": 0.3174,
+      "epoch": 0.29624824305330305,
+      "grad_norm": 0.1933128535747528,
+      "learning_rate": 4.988806711014505e-05,
+      "loss": 0.4845,
       "step": 8220
     },
     {
-      "epoch": 0.29,
-      "learning_rate": 4.989821010433741e-05,
-      "loss": 0.3315,
+      "epoch": 0.2964284427145277,
+      "grad_norm": 0.1766856461763382,
+      "learning_rate": 4.9887791110047e-05,
+      "loss": 0.4751,
       "step": 8225
     },
     {
-      "epoch": 0.29,
-      "learning_rate": 4.98979531369263e-05,
-      "loss": 0.3352,
+      "epoch": 0.29660864237575235,
+      "grad_norm": 0.1650475561618805,
+      "learning_rate": 4.9887514770857605e-05,
+      "loss": 0.416,
       "step": 8230
     },
     {
-      "epoch": 0.29,
-      "learning_rate": 4.9897695846231696e-05,
-      "loss": 0.2966,
+      "epoch": 0.296788842036977,
+      "grad_norm": 0.16024361550807953,
+      "learning_rate": 4.988723809258064e-05,
+      "loss": 0.4831,
       "step": 8235
     },
     {
-      "epoch": 0.29,
-      "learning_rate": 4.989743823225691e-05,
-      "loss": 0.3223,
+      "epoch": 0.2969690416982016,
+      "grad_norm": 0.1936086267232895,
+      "learning_rate": 4.9886961075219885e-05,
+      "loss": 0.4589,
       "step": 8240
     },
     {
-      "epoch": 0.29,
-      "learning_rate": 4.98971802950053e-05,
-      "loss": 0.3506,
+      "epoch": 0.2971492413594262,
+      "grad_norm": 0.20437081158161163,
+      "learning_rate": 4.988668371877909e-05,
+      "loss": 0.5085,
       "step": 8245
     },
     {
-      "epoch": 0.29,
-      "learning_rate": 4.989692203448021e-05,
-      "loss": 0.334,
+      "epoch": 0.2973294410206509,
+      "grad_norm": 0.2507721185684204,
+      "learning_rate": 4.988640602326205e-05,
+      "loss": 0.4888,
       "step": 8250
     },
     {
-      "epoch": 0.29,
-      "learning_rate": 4.9896663450684986e-05,
-      "loss": 0.3311,
+      "epoch": 0.2975096406818755,
+      "grad_norm": 0.2515529692173004,
+      "learning_rate": 4.9886127988672554e-05,
+      "loss": 0.4617,
       "step": 8255
     },
     {
-      "epoch": 0.29,
-      "learning_rate": 4.989640454362301e-05,
-      "loss": 0.3512,
+      "epoch": 0.29768984034310014,
+      "grad_norm": 0.18619805574417114,
+      "learning_rate": 4.988584961501438e-05,
+      "loss": 0.4843,
       "step": 8260
     },
     {
-      "epoch": 0.29,
-      "learning_rate": 4.9896145313297614e-05,
-      "loss": 0.3332,
+      "epoch": 0.29787004000432477,
+      "grad_norm": 0.18921491503715515,
+      "learning_rate": 4.988557090229132e-05,
+      "loss": 0.4308,
       "step": 8265
     },
     {
-      "epoch": 0.29,
-      "learning_rate": 4.989588575971219e-05,
-      "loss": 0.3331,
+      "epoch": 0.29805023966554944,
+      "grad_norm": 0.17561018466949463,
+      "learning_rate": 4.988529185050717e-05,
+      "loss": 0.4686,
       "step": 8270
     },
     {
-      "epoch": 0.29,
-      "learning_rate": 4.989562588287009e-05,
-      "loss": 0.3218,
+      "epoch": 0.29823043932677407,
+      "grad_norm": 0.18587404489517212,
+      "learning_rate": 4.988501245966574e-05,
+      "loss": 0.4476,
       "step": 8275
     },
     {
-      "epoch": 0.29,
-      "learning_rate": 4.98953656827747e-05,
-      "loss": 0.3283,
+      "epoch": 0.2984106389879987,
+      "grad_norm": 0.22094377875328064,
+      "learning_rate": 4.988473272977083e-05,
+      "loss": 0.4372,
       "step": 8280
     },
     {
-      "epoch": 0.29,
-      "learning_rate": 4.989510515942939e-05,
-      "loss": 0.3352,
+      "epoch": 0.29859083864922337,
+      "grad_norm": 0.23441436886787415,
+      "learning_rate": 4.988445266082626e-05,
+      "loss": 0.4697,
       "step": 8285
     },
     {
-      "epoch": 0.29,
-      "learning_rate": 4.989484431283755e-05,
-      "loss": 0.3431,
+      "epoch": 0.298771038310448,
+      "grad_norm": 0.2000618875026703,
+      "learning_rate": 4.988417225283585e-05,
+      "loss": 0.4548,
       "step": 8290
     },
     {
-      "epoch": 0.29,
-      "learning_rate": 4.989458314300256e-05,
-      "loss": 0.3274,
+      "epoch": 0.2989512379716726,
+      "grad_norm": 0.16407370567321777,
+      "learning_rate": 4.9883891505803394e-05,
+      "loss": 0.4583,
       "step": 8295
     },
     {
-      "epoch": 0.29,
-      "learning_rate": 4.989432164992781e-05,
-      "loss": 0.291,
+      "epoch": 0.29913143763289723,
+      "grad_norm": 0.1786169558763504,
+      "learning_rate": 4.988361041973274e-05,
+      "loss": 0.4534,
       "step": 8300
     },
     {
-      "epoch": 0.29,
-      "learning_rate": 4.98940598336167e-05,
-      "loss": 0.3318,
+      "epoch": 0.2993116372941219,
+      "grad_norm": 0.22652758657932281,
+      "learning_rate": 4.988332899462771e-05,
+      "loss": 0.4826,
       "step": 8305
     },
     {
-      "epoch": 0.29,
-      "learning_rate": 4.989379769407263e-05,
-      "loss": 0.3251,
+      "epoch": 0.29949183695534654,
+      "grad_norm": 0.23459400236606598,
+      "learning_rate": 4.9883047230492144e-05,
+      "loss": 0.4353,
       "step": 8310
     },
     {
-      "epoch": 0.29,
-      "learning_rate": 4.9893535231299004e-05,
-      "loss": 0.334,
+      "epoch": 0.29967203661657116,
+      "grad_norm": 0.19752709567546844,
+      "learning_rate": 4.988276512732987e-05,
+      "loss": 0.4551,
       "step": 8315
     },
     {
-      "epoch": 0.29,
-      "learning_rate": 4.9893272445299236e-05,
-      "loss": 0.3066,
+      "epoch": 0.2998522362777958,
+      "grad_norm": 0.184610515832901,
+      "learning_rate": 4.988248268514475e-05,
+      "loss": 0.4713,
       "step": 8320
     },
     {
-      "epoch": 0.29,
-      "learning_rate": 4.9893009336076716e-05,
-      "loss": 0.3556,
+      "epoch": 0.30003243593902046,
+      "grad_norm": 0.16262328624725342,
+      "learning_rate": 4.988219990394061e-05,
+      "loss": 0.472,
       "step": 8325
     },
     {
-      "epoch": 0.29,
-      "learning_rate": 4.989274590363489e-05,
-      "loss": 0.3614,
+      "epoch": 0.3002126356002451,
+      "grad_norm": 0.1739642322063446,
+      "learning_rate": 4.988191678372132e-05,
+      "loss": 0.4622,
       "step": 8330
     },
     {
-      "epoch": 0.29,
-      "learning_rate": 4.989248214797715e-05,
-      "loss": 0.3364,
+      "epoch": 0.3003928352614697,
+      "grad_norm": 0.15458573400974274,
+      "learning_rate": 4.988163332449073e-05,
+      "loss": 0.4829,
       "step": 8335
     },
     {
-      "epoch": 0.29,
-      "learning_rate": 4.9892218069106944e-05,
-      "loss": 0.3103,
+      "epoch": 0.3005730349226943,
+      "grad_norm": 0.18645693361759186,
+      "learning_rate": 4.9881349526252694e-05,
+      "loss": 0.4414,
       "step": 8340
     },
     {
-      "epoch": 0.29,
-      "learning_rate": 4.989195366702768e-05,
-      "loss": 0.3487,
+      "epoch": 0.300753234583919,
+      "grad_norm": 0.17546844482421875,
+      "learning_rate": 4.988106538901109e-05,
+      "loss": 0.4612,
       "step": 8345
     },
     {
-      "epoch": 0.29,
-      "learning_rate": 4.989168894174281e-05,
-      "loss": 0.3232,
+      "epoch": 0.3009334342451436,
+      "grad_norm": 0.1612381488084793,
+      "learning_rate": 4.9880780912769796e-05,
+      "loss": 0.4759,
       "step": 8350
     },
     {
-      "epoch": 0.29,
-      "learning_rate": 4.989142389325576e-05,
-      "loss": 0.3302,
+      "epoch": 0.30111363390636825,
+      "grad_norm": 0.179169699549675,
+      "learning_rate": 4.988049609753268e-05,
+      "loss": 0.4535,
       "step": 8355
     },
     {
-      "epoch": 0.29,
-      "learning_rate": 4.989115852156998e-05,
-      "loss": 0.3242,
+      "epoch": 0.3012938335675929,
+      "grad_norm": 0.1993793547153473,
+      "learning_rate": 4.988021094330362e-05,
+      "loss": 0.4463,
       "step": 8360
     },
     {
-      "epoch": 0.29,
-      "learning_rate": 4.989089282668891e-05,
-      "loss": 0.3235,
+      "epoch": 0.30147403322881755,
+      "grad_norm": 0.22224834561347961,
+      "learning_rate": 4.987992545008649e-05,
+      "loss": 0.4579,
       "step": 8365
     },
     {
-      "epoch": 0.29,
-      "learning_rate": 4.989062680861599e-05,
-      "loss": 0.3434,
+      "epoch": 0.3016542328900422,
+      "grad_norm": 0.17759408056735992,
+      "learning_rate": 4.987963961788521e-05,
+      "loss": 0.4718,
       "step": 8370
     },
     {
-      "epoch": 0.29,
-      "learning_rate": 4.9890360467354694e-05,
-      "loss": 0.3097,
+      "epoch": 0.3018344325512668,
+      "grad_norm": 0.1582093983888626,
+      "learning_rate": 4.9879353446703655e-05,
+      "loss": 0.4568,
       "step": 8375
     },
     {
-      "epoch": 0.29,
-      "learning_rate": 4.9890093802908466e-05,
-      "loss": 0.2919,
+      "epoch": 0.3020146322124914,
+      "grad_norm": 0.20277364552021027,
+      "learning_rate": 4.987906693654572e-05,
+      "loss": 0.4607,
       "step": 8380
     },
     {
-      "epoch": 0.3,
-      "learning_rate": 4.988982681528078e-05,
-      "loss": 0.3215,
+      "epoch": 0.3021948318737161,
+      "grad_norm": 0.20033441483974457,
+      "learning_rate": 4.987878008741531e-05,
+      "loss": 0.4503,
       "step": 8385
     },
     {
-      "epoch": 0.3,
-      "learning_rate": 4.988955950447509e-05,
-      "loss": 0.3283,
+      "epoch": 0.3023750315349407,
+      "grad_norm": 0.2945716977119446,
+      "learning_rate": 4.9878492899316346e-05,
+      "loss": 0.5069,
       "step": 8390
     },
     {
-      "epoch": 0.3,
-      "learning_rate": 4.988929187049488e-05,
-      "loss": 0.3286,
+      "epoch": 0.30255523119616534,
+      "grad_norm": 0.16416752338409424,
+      "learning_rate": 4.987820537225273e-05,
+      "loss": 0.4901,
       "step": 8395
     },
     {
-      "epoch": 0.3,
-      "learning_rate": 4.9889023913343605e-05,
-      "loss": 0.3343,
+      "epoch": 0.30273543085738996,
+      "grad_norm": 0.25771409273147583,
+      "learning_rate": 4.9877917506228386e-05,
+      "loss": 0.471,
       "step": 8400
     },
     {
-      "epoch": 0.3,
-      "learning_rate": 4.988875563302477e-05,
-      "loss": 0.3434,
+      "epoch": 0.30291563051861464,
+      "grad_norm": 0.18886184692382812,
+      "learning_rate": 4.987762930124723e-05,
+      "loss": 0.4543,
       "step": 8405
     },
     {
-      "epoch": 0.3,
-      "learning_rate": 4.988848702954184e-05,
-      "loss": 0.3349,
+      "epoch": 0.30309583017983927,
+      "grad_norm": 0.17397350072860718,
+      "learning_rate": 4.987734075731319e-05,
+      "loss": 0.4644,
       "step": 8410
     },
     {
-      "epoch": 0.3,
-      "learning_rate": 4.9888218102898305e-05,
-      "loss": 0.3262,
+      "epoch": 0.3032760298410639,
+      "grad_norm": 0.19836938381195068,
+      "learning_rate": 4.9877051874430204e-05,
+      "loss": 0.4784,
       "step": 8415
     },
     {
-      "epoch": 0.3,
-      "learning_rate": 4.988794885309766e-05,
-      "loss": 0.3224,
+      "epoch": 0.3034562295022885,
+      "grad_norm": 0.1647557020187378,
+      "learning_rate": 4.987676265260219e-05,
+      "loss": 0.4518,
       "step": 8420
     },
     {
-      "epoch": 0.3,
-      "learning_rate": 4.9887679280143404e-05,
-      "loss": 0.3409,
+      "epoch": 0.3036364291635132,
+      "grad_norm": 0.21178211271762848,
+      "learning_rate": 4.987647309183311e-05,
+      "loss": 0.4945,
       "step": 8425
     },
     {
-      "epoch": 0.3,
-      "learning_rate": 4.988740938403903e-05,
-      "loss": 0.3289,
+      "epoch": 0.3038166288247378,
+      "grad_norm": 0.19628405570983887,
+      "learning_rate": 4.98761831921269e-05,
+      "loss": 0.4792,
       "step": 8430
     },
     {
-      "epoch": 0.3,
-      "learning_rate": 4.9887139164788056e-05,
-      "loss": 0.331,
+      "epoch": 0.30399682848596243,
+      "grad_norm": 0.20497670769691467,
+      "learning_rate": 4.987589295348751e-05,
+      "loss": 0.4518,
       "step": 8435
     },
     {
-      "epoch": 0.3,
-      "learning_rate": 4.988686862239397e-05,
-      "loss": 0.3153,
+      "epoch": 0.30417702814718706,
+      "grad_norm": 0.2078750729560852,
+      "learning_rate": 4.987560237591889e-05,
+      "loss": 0.5137,
       "step": 8440
     },
     {
-      "epoch": 0.3,
-      "learning_rate": 4.98865977568603e-05,
-      "loss": 0.3435,
+      "epoch": 0.30435722780841173,
+      "grad_norm": 0.15747526288032532,
+      "learning_rate": 4.987531145942501e-05,
+      "loss": 0.4622,
       "step": 8445
     },
     {
-      "epoch": 0.3,
-      "learning_rate": 4.988632656819057e-05,
-      "loss": 0.3379,
+      "epoch": 0.30453742746963636,
+      "grad_norm": 0.19545340538024902,
+      "learning_rate": 4.987502020400983e-05,
+      "loss": 0.46,
       "step": 8450
     },
     {
-      "epoch": 0.3,
-      "learning_rate": 4.988605505638827e-05,
-      "loss": 0.3532,
+      "epoch": 0.304717627130861,
+      "grad_norm": 0.19771957397460938,
+      "learning_rate": 4.9874728609677316e-05,
+      "loss": 0.4655,
       "step": 8455
     },
     {
-      "epoch": 0.3,
-      "learning_rate": 4.9885783221456963e-05,
-      "loss": 0.3551,
+      "epoch": 0.30489782679208566,
+      "grad_norm": 0.1792868971824646,
+      "learning_rate": 4.9874436676431435e-05,
+      "loss": 0.452,
       "step": 8460
     },
     {
-      "epoch": 0.3,
-      "learning_rate": 4.988551106340015e-05,
-      "loss": 0.3478,
+      "epoch": 0.3050780264533103,
+      "grad_norm": 0.2185121327638626,
+      "learning_rate": 4.9874144404276165e-05,
+      "loss": 0.4803,
       "step": 8465
     },
     {
-      "epoch": 0.3,
-      "learning_rate": 4.988523858222138e-05,
-      "loss": 0.3282,
+      "epoch": 0.3052582261145349,
+      "grad_norm": 0.18078531324863434,
+      "learning_rate": 4.98738517932155e-05,
+      "loss": 0.4395,
       "step": 8470
     },
     {
-      "epoch": 0.3,
-      "learning_rate": 4.9884965777924185e-05,
-      "loss": 0.3283,
+      "epoch": 0.3054384257757595,
+      "grad_norm": 0.14936190843582153,
+      "learning_rate": 4.987355884325342e-05,
+      "loss": 0.4672,
       "step": 8475
     },
     {
-      "epoch": 0.3,
-      "learning_rate": 4.988469265051211e-05,
-      "loss": 0.3209,
+      "epoch": 0.3056186254369842,
+      "grad_norm": 0.19642381370067596,
+      "learning_rate": 4.987326555439392e-05,
+      "loss": 0.4507,
       "step": 8480
     },
     {
-      "epoch": 0.3,
-      "learning_rate": 4.9884419199988705e-05,
-      "loss": 0.3086,
+      "epoch": 0.3057988250982088,
+      "grad_norm": 0.1511949896812439,
+      "learning_rate": 4.987297192664099e-05,
+      "loss": 0.426,
       "step": 8485
     },
     {
-      "epoch": 0.3,
-      "learning_rate": 4.988414542635751e-05,
-      "loss": 0.2962,
+      "epoch": 0.30597902475943345,
+      "grad_norm": 0.20524610579013824,
+      "learning_rate": 4.9872677959998626e-05,
+      "loss": 0.4387,
       "step": 8490
     },
     {
-      "epoch": 0.3,
-      "learning_rate": 4.988387132962209e-05,
-      "loss": 0.314,
+      "epoch": 0.30615922442065807,
+      "grad_norm": 0.19021369516849518,
+      "learning_rate": 4.987238365447084e-05,
+      "loss": 0.49,
       "step": 8495
     },
     {
-      "epoch": 0.3,
-      "learning_rate": 4.9883596909786e-05,
-      "loss": 0.351,
+      "epoch": 0.30633942408188275,
+      "grad_norm": 0.2122432291507721,
+      "learning_rate": 4.987208901006165e-05,
+      "loss": 0.4536,
       "step": 8500
     },
     {
-      "epoch": 0.3,
-      "eval_loss": 0.32749566435813904,
-      "eval_runtime": 10.5451,
-      "eval_samples_per_second": 9.483,
-      "eval_steps_per_second": 9.483,
+      "epoch": 0.30633942408188275,
+      "eval_loss": 0.4835846722126007,
+      "eval_runtime": 3.5343,
+      "eval_samples_per_second": 28.294,
+      "eval_steps_per_second": 7.074,
       "step": 8500
     },
     {
-      "epoch": 0.3,
-      "learning_rate": 4.98833221668528e-05,
-      "loss": 0.3367,
+      "epoch": 0.3065196237431074,
+      "grad_norm": 0.2556203603744507,
+      "learning_rate": 4.9871794026775067e-05,
+      "loss": 0.4393,
       "step": 8505
     },
     {
-      "epoch": 0.3,
-      "learning_rate": 4.9883047100826064e-05,
-      "loss": 0.3108,
+      "epoch": 0.306699823404332,
+      "grad_norm": 0.16816875338554382,
+      "learning_rate": 4.98714987046151e-05,
+      "loss": 0.4726,
       "step": 8510
     },
     {
-      "epoch": 0.3,
-      "learning_rate": 4.988277171170936e-05,
-      "loss": 0.3493,
+      "epoch": 0.3068800230655566,
+      "grad_norm": 0.2908528745174408,
+      "learning_rate": 4.9871203043585774e-05,
+      "loss": 0.4778,
       "step": 8515
     },
     {
-      "epoch": 0.3,
-      "learning_rate": 4.988249599950626e-05,
-      "loss": 0.3309,
+      "epoch": 0.3070602227267813,
+      "grad_norm": 0.2010086327791214,
+      "learning_rate": 4.987090704369112e-05,
+      "loss": 0.4649,
       "step": 8520
     },
     {
-      "epoch": 0.3,
-      "learning_rate": 4.988221996422035e-05,
-      "loss": 0.3633,
+      "epoch": 0.3072404223880059,
+      "grad_norm": 0.19618774950504303,
+      "learning_rate": 4.9870610704935185e-05,
+      "loss": 0.5132,
       "step": 8525
     },
     {
-      "epoch": 0.3,
-      "learning_rate": 4.9881943605855215e-05,
-      "loss": 0.3119,
+      "epoch": 0.30742062204923054,
+      "grad_norm": 0.15902240574359894,
+      "learning_rate": 4.9870314027321984e-05,
+      "loss": 0.4017,
       "step": 8530
     },
     {
-      "epoch": 0.3,
-      "learning_rate": 4.9881666924414436e-05,
-      "loss": 0.3409,
+      "epoch": 0.30760082171045516,
+      "grad_norm": 0.22079305350780487,
+      "learning_rate": 4.9870017010855576e-05,
+      "loss": 0.4651,
       "step": 8535
     },
     {
-      "epoch": 0.3,
-      "learning_rate": 4.988138991990161e-05,
-      "loss": 0.3129,
+      "epoch": 0.30778102137167984,
+      "grad_norm": 0.19805465638637543,
+      "learning_rate": 4.9869719655539994e-05,
+      "loss": 0.4438,
       "step": 8540
     },
     {
-      "epoch": 0.3,
-      "learning_rate": 4.9881112592320333e-05,
-      "loss": 0.3077,
+      "epoch": 0.30796122103290446,
+      "grad_norm": 0.17793111503124237,
+      "learning_rate": 4.98694219613793e-05,
+      "loss": 0.4655,
       "step": 8545
     },
     {
-      "epoch": 0.3,
-      "learning_rate": 4.988083494167421e-05,
-      "loss": 0.3065,
+      "epoch": 0.3081414206941291,
+      "grad_norm": 0.16948306560516357,
+      "learning_rate": 4.986912392837755e-05,
+      "loss": 0.4485,
       "step": 8550
     },
     {
-      "epoch": 0.3,
-      "learning_rate": 4.988055696796683e-05,
-      "loss": 0.3331,
+      "epoch": 0.3083216203553537,
+      "grad_norm": 0.16313959658145905,
+      "learning_rate": 4.98688255565388e-05,
+      "loss": 0.4846,
       "step": 8555
     },
     {
-      "epoch": 0.3,
-      "learning_rate": 4.988027867120183e-05,
-      "loss": 0.3345,
+      "epoch": 0.3085018200165784,
+      "grad_norm": 0.20121654868125916,
+      "learning_rate": 4.9868526845867115e-05,
+      "loss": 0.4565,
       "step": 8560
     },
     {
-      "epoch": 0.3,
-      "learning_rate": 4.9880000051382804e-05,
-      "loss": 0.315,
+      "epoch": 0.308682019677803,
+      "grad_norm": 0.17432835698127747,
+      "learning_rate": 4.9868227796366566e-05,
+      "loss": 0.4505,
       "step": 8565
     },
     {
-      "epoch": 0.3,
-      "learning_rate": 4.9879721108513364e-05,
-      "loss": 0.331,
+      "epoch": 0.30886221933902763,
+      "grad_norm": 0.2507334053516388,
+      "learning_rate": 4.986792840804122e-05,
+      "loss": 0.4353,
       "step": 8570
     },
     {
-      "epoch": 0.3,
-      "learning_rate": 4.987944184259715e-05,
-      "loss": 0.3345,
+      "epoch": 0.30904241900025226,
+      "grad_norm": 0.12277240306138992,
+      "learning_rate": 4.986762868089517e-05,
+      "loss": 0.4637,
       "step": 8575
     },
     {
-      "epoch": 0.3,
-      "learning_rate": 4.987916225363778e-05,
-      "loss": 0.3136,
+      "epoch": 0.30922261866147693,
+      "grad_norm": 0.20205271244049072,
+      "learning_rate": 4.98673286149325e-05,
+      "loss": 0.4189,
       "step": 8580
     },
     {
-      "epoch": 0.3,
-      "learning_rate": 4.9878882341638886e-05,
-      "loss": 0.3176,
+      "epoch": 0.30940281832270156,
+      "grad_norm": 0.16052600741386414,
+      "learning_rate": 4.986702821015729e-05,
+      "loss": 0.4499,
       "step": 8585
     },
     {
-      "epoch": 0.3,
-      "learning_rate": 4.9878602106604097e-05,
-      "loss": 0.3392,
+      "epoch": 0.3095830179839262,
+      "grad_norm": 0.16914494335651398,
+      "learning_rate": 4.9866727466573634e-05,
+      "loss": 0.4447,
       "step": 8590
     },
     {
-      "epoch": 0.3,
-      "learning_rate": 4.987832154853706e-05,
-      "loss": 0.3551,
+      "epoch": 0.3097632176451508,
+      "grad_norm": 0.19733648002147675,
+      "learning_rate": 4.986642638418563e-05,
+      "loss": 0.4586,
       "step": 8595
     },
     {
-      "epoch": 0.3,
-      "learning_rate": 4.9878040667441405e-05,
-      "loss": 0.3136,
+      "epoch": 0.3099434173063755,
+      "grad_norm": 0.19025246798992157,
+      "learning_rate": 4.986612496299738e-05,
+      "loss": 0.5077,
       "step": 8600
     },
     {
-      "epoch": 0.3,
-      "learning_rate": 4.9877759463320795e-05,
-      "loss": 0.3382,
+      "epoch": 0.3101236169676001,
+      "grad_norm": 0.1786302626132965,
+      "learning_rate": 4.986582320301299e-05,
+      "loss": 0.4415,
       "step": 8605
     },
     {
-      "epoch": 0.3,
-      "learning_rate": 4.9877477936178864e-05,
-      "loss": 0.3103,
+      "epoch": 0.3103038166288247,
+      "grad_norm": 0.18163448572158813,
+      "learning_rate": 4.9865521104236575e-05,
+      "loss": 0.465,
       "step": 8610
     },
     {
-      "epoch": 0.3,
-      "learning_rate": 4.9877196086019286e-05,
-      "loss": 0.321,
+      "epoch": 0.31048401629004935,
+      "grad_norm": 0.19647279381752014,
+      "learning_rate": 4.986521866667225e-05,
+      "loss": 0.4357,
       "step": 8615
     },
     {
-      "epoch": 0.3,
-      "learning_rate": 4.9876913912845705e-05,
-      "loss": 0.3064,
+      "epoch": 0.310664215951274,
+      "grad_norm": 0.15463513135910034,
+      "learning_rate": 4.9864915890324136e-05,
+      "loss": 0.3975,
       "step": 8620
     },
     {
-      "epoch": 0.3,
-      "learning_rate": 4.9876631416661796e-05,
-      "loss": 0.3302,
+      "epoch": 0.31084441561249865,
+      "grad_norm": 0.1965053528547287,
+      "learning_rate": 4.986461277519635e-05,
+      "loss": 0.463,
       "step": 8625
     },
     {
-      "epoch": 0.3,
-      "learning_rate": 4.9876348597471224e-05,
-      "loss": 0.3565,
+      "epoch": 0.31102461527372327,
+      "grad_norm": 0.20269662141799927,
+      "learning_rate": 4.9864309321293035e-05,
+      "loss": 0.441,
       "step": 8630
     },
     {
-      "epoch": 0.3,
-      "learning_rate": 4.9876065455277654e-05,
-      "loss": 0.3379,
+      "epoch": 0.31120481493494795,
+      "grad_norm": 0.2272135317325592,
+      "learning_rate": 4.986400552861832e-05,
+      "loss": 0.5056,
       "step": 8635
     },
     {
-      "epoch": 0.3,
-      "learning_rate": 4.9875781990084766e-05,
-      "loss": 0.3173,
+      "epoch": 0.31138501459617257,
+      "grad_norm": 0.1746290922164917,
+      "learning_rate": 4.986370139717634e-05,
+      "loss": 0.4583,
       "step": 8640
     },
     {
-      "epoch": 0.3,
-      "learning_rate": 4.9875498201896254e-05,
-      "loss": 0.3347,
+      "epoch": 0.3115652142573972,
+      "grad_norm": 0.17649775743484497,
+      "learning_rate": 4.9863396926971245e-05,
+      "loss": 0.4873,
       "step": 8645
     },
     {
-      "epoch": 0.3,
-      "learning_rate": 4.987521409071578e-05,
-      "loss": 0.3462,
+      "epoch": 0.3117454139186218,
+      "grad_norm": 0.185885488986969,
+      "learning_rate": 4.9863092118007185e-05,
+      "loss": 0.4692,
       "step": 8650
     },
     {
-      "epoch": 0.3,
-      "learning_rate": 4.9874929656547054e-05,
-      "loss": 0.3155,
+      "epoch": 0.3119256135798465,
+      "grad_norm": 0.1990884244441986,
+      "learning_rate": 4.98627869702883e-05,
+      "loss": 0.4578,
       "step": 8655
     },
     {
-      "epoch": 0.3,
-      "learning_rate": 4.987464489939375e-05,
-      "loss": 0.3135,
+      "epoch": 0.3121058132410711,
+      "grad_norm": 0.22869646549224854,
+      "learning_rate": 4.9862481483818755e-05,
+      "loss": 0.4932,
       "step": 8660
     },
     {
-      "epoch": 0.3,
-      "learning_rate": 4.987435981925958e-05,
-      "loss": 0.3295,
+      "epoch": 0.31228601290229574,
+      "grad_norm": 0.1766314059495926,
+      "learning_rate": 4.986217565860272e-05,
+      "loss": 0.439,
       "step": 8665
     },
     {
-      "epoch": 0.31,
-      "learning_rate": 4.987407441614824e-05,
-      "loss": 0.3148,
+      "epoch": 0.31246621256352036,
+      "grad_norm": 0.14868828654289246,
+      "learning_rate": 4.986186949464435e-05,
+      "loss": 0.4357,
       "step": 8670
     },
     {
-      "epoch": 0.31,
-      "learning_rate": 4.987378869006344e-05,
-      "loss": 0.3619,
+      "epoch": 0.31264641222474504,
+      "grad_norm": 0.17688503861427307,
+      "learning_rate": 4.986156299194783e-05,
+      "loss": 0.4194,
       "step": 8675
     },
     {
-      "epoch": 0.31,
-      "learning_rate": 4.987350264100889e-05,
-      "loss": 0.3541,
+      "epoch": 0.31282661188596966,
+      "grad_norm": 0.215143084526062,
+      "learning_rate": 4.9861256150517324e-05,
+      "loss": 0.4795,
       "step": 8680
     },
     {
-      "epoch": 0.31,
-      "learning_rate": 4.987321626898829e-05,
-      "loss": 0.3266,
+      "epoch": 0.3130068115471943,
+      "grad_norm": 0.18795055150985718,
+      "learning_rate": 4.9860948970357014e-05,
+      "loss": 0.4408,
       "step": 8685
     },
     {
-      "epoch": 0.31,
-      "learning_rate": 4.987292957400537e-05,
-      "loss": 0.3196,
+      "epoch": 0.3131870112084189,
+      "grad_norm": 0.14594529569149017,
+      "learning_rate": 4.986064145147108e-05,
+      "loss": 0.4296,
       "step": 8690
     },
     {
-      "epoch": 0.31,
-      "learning_rate": 4.987264255606385e-05,
-      "loss": 0.3261,
+      "epoch": 0.3133672108696436,
+      "grad_norm": 0.19025376439094543,
+      "learning_rate": 4.986033359386373e-05,
+      "loss": 0.4736,
       "step": 8695
     },
     {
-      "epoch": 0.31,
-      "learning_rate": 4.987235521516747e-05,
-      "loss": 0.3328,
+      "epoch": 0.3135474105308682,
+      "grad_norm": 0.18264278769493103,
+      "learning_rate": 4.986002539753915e-05,
+      "loss": 0.4456,
       "step": 8700
     },
     {
-      "epoch": 0.31,
-      "learning_rate": 4.987206755131993e-05,
-      "loss": 0.3345,
+      "epoch": 0.31372761019209283,
+      "grad_norm": 0.14919918775558472,
+      "learning_rate": 4.985971686250153e-05,
+      "loss": 0.4644,
       "step": 8705
     },
     {
-      "epoch": 0.31,
-      "learning_rate": 4.9871779564525e-05,
-      "loss": 0.3279,
+      "epoch": 0.31390780985331745,
+      "grad_norm": 0.2042977660894394,
+      "learning_rate": 4.985940798875508e-05,
+      "loss": 0.4554,
       "step": 8710
     },
     {
-      "epoch": 0.31,
-      "learning_rate": 4.987149125478639e-05,
-      "loss": 0.3352,
+      "epoch": 0.31408800951454213,
+      "grad_norm": 0.1687832921743393,
+      "learning_rate": 4.9859098776304015e-05,
+      "loss": 0.4685,
       "step": 8715
     },
     {
-      "epoch": 0.31,
-      "learning_rate": 4.987120262210787e-05,
-      "loss": 0.3469,
+      "epoch": 0.31426820917576676,
+      "grad_norm": 0.19032564759254456,
+      "learning_rate": 4.985878922515253e-05,
+      "loss": 0.451,
       "step": 8720
     },
     {
-      "epoch": 0.31,
-      "learning_rate": 4.9870913666493166e-05,
-      "loss": 0.3329,
+      "epoch": 0.3144484088369914,
+      "grad_norm": 0.15537263453006744,
+      "learning_rate": 4.985847933530486e-05,
+      "loss": 0.4373,
       "step": 8725
     },
     {
-      "epoch": 0.31,
-      "learning_rate": 4.9870624387946046e-05,
-      "loss": 0.3516,
+      "epoch": 0.314628608498216,
+      "grad_norm": 0.24856919050216675,
+      "learning_rate": 4.985816910676523e-05,
+      "loss": 0.5004,
       "step": 8730
     },
     {
-      "epoch": 0.31,
-      "learning_rate": 4.9870334786470256e-05,
-      "loss": 0.3184,
+      "epoch": 0.3148088081594407,
+      "grad_norm": 0.18003836274147034,
+      "learning_rate": 4.985785853953786e-05,
+      "loss": 0.4703,
       "step": 8735
     },
     {
-      "epoch": 0.31,
-      "learning_rate": 4.987004486206956e-05,
-      "loss": 0.3499,
+      "epoch": 0.3149890078206653,
+      "grad_norm": 0.1739412099123001,
+      "learning_rate": 4.9857547633626964e-05,
+      "loss": 0.4538,
       "step": 8740
     },
     {
-      "epoch": 0.31,
-      "learning_rate": 4.986975461474772e-05,
-      "loss": 0.2988,
+      "epoch": 0.3151692074818899,
+      "grad_norm": 0.19691774249076843,
+      "learning_rate": 4.985723638903681e-05,
+      "loss": 0.4671,
       "step": 8745
     },
     {
-      "epoch": 0.31,
-      "learning_rate": 4.986946404450851e-05,
-      "loss": 0.349,
+      "epoch": 0.31534940714311455,
+      "grad_norm": 0.2076934427022934,
+      "learning_rate": 4.9856924805771614e-05,
+      "loss": 0.4796,
       "step": 8750
     },
     {
-      "epoch": 0.31,
-      "learning_rate": 4.986917315135569e-05,
-      "loss": 0.3331,
+      "epoch": 0.3155296068043392,
+      "grad_norm": 0.16024084389209747,
+      "learning_rate": 4.9856612883835633e-05,
+      "loss": 0.4567,
       "step": 8755
     },
     {
-      "epoch": 0.31,
-      "learning_rate": 4.986888193529305e-05,
-      "loss": 0.3303,
+      "epoch": 0.31570980646556385,
+      "grad_norm": 0.23982550203800201,
+      "learning_rate": 4.985630062323311e-05,
+      "loss": 0.4944,
       "step": 8760
     },
     {
-      "epoch": 0.31,
-      "learning_rate": 4.986859039632436e-05,
-      "loss": 0.3464,
+      "epoch": 0.31589000612678847,
+      "grad_norm": 0.16979862749576569,
+      "learning_rate": 4.9855988023968314e-05,
+      "loss": 0.481,
       "step": 8765
     },
     {
-      "epoch": 0.31,
-      "learning_rate": 4.986829853445343e-05,
-      "loss": 0.3315,
+      "epoch": 0.3160702057880131,
+      "grad_norm": 0.19428572058677673,
+      "learning_rate": 4.9855675086045486e-05,
+      "loss": 0.4769,
       "step": 8770
     },
     {
-      "epoch": 0.31,
-      "learning_rate": 4.986800634968402e-05,
-      "loss": 0.3099,
+      "epoch": 0.31625040544923777,
+      "grad_norm": 0.21432949602603912,
+      "learning_rate": 4.985536180946889e-05,
+      "loss": 0.4535,
       "step": 8775
     },
     {
-      "epoch": 0.31,
-      "learning_rate": 4.9867713842019934e-05,
-      "loss": 0.3392,
+      "epoch": 0.3164306051104624,
+      "grad_norm": 0.1915997415781021,
+      "learning_rate": 4.9855048194242816e-05,
+      "loss": 0.4451,
       "step": 8780
     },
     {
-      "epoch": 0.31,
-      "learning_rate": 4.9867421011464984e-05,
-      "loss": 0.3358,
+      "epoch": 0.316610804771687,
+      "grad_norm": 0.18941251933574677,
+      "learning_rate": 4.985473424037151e-05,
+      "loss": 0.4671,
       "step": 8785
     },
     {
-      "epoch": 0.31,
-      "learning_rate": 4.986712785802295e-05,
-      "loss": 0.3171,
+      "epoch": 0.3167910044329117,
+      "grad_norm": 0.17575441300868988,
+      "learning_rate": 4.985441994785927e-05,
+      "loss": 0.4611,
       "step": 8790
     },
     {
-      "epoch": 0.31,
-      "learning_rate": 4.986683438169766e-05,
-      "loss": 0.3225,
+      "epoch": 0.3169712040941363,
+      "grad_norm": 0.1792481243610382,
+      "learning_rate": 4.9854105316710364e-05,
+      "loss": 0.4631,
       "step": 8795
     },
     {
-      "epoch": 0.31,
-      "learning_rate": 4.9866540582492907e-05,
-      "loss": 0.2957,
+      "epoch": 0.31715140375536094,
+      "grad_norm": 0.20034508407115936,
+      "learning_rate": 4.9853790346929096e-05,
+      "loss": 0.4751,
       "step": 8800
     },
     {
-      "epoch": 0.31,
-      "learning_rate": 4.98662464604125e-05,
-      "loss": 0.3151,
+      "epoch": 0.31733160341658556,
+      "grad_norm": 0.21609006822109222,
+      "learning_rate": 4.9853475038519736e-05,
+      "loss": 0.516,
       "step": 8805
     },
     {
-      "epoch": 0.31,
-      "learning_rate": 4.9865952015460285e-05,
-      "loss": 0.3267,
+      "epoch": 0.31751180307781024,
+      "grad_norm": 0.17463666200637817,
+      "learning_rate": 4.9853159391486594e-05,
+      "loss": 0.4715,
       "step": 8810
     },
     {
-      "epoch": 0.31,
-      "learning_rate": 4.9865657247640065e-05,
-      "loss": 0.327,
+      "epoch": 0.31769200273903486,
+      "grad_norm": 0.1856069266796112,
+      "learning_rate": 4.9852843405833965e-05,
+      "loss": 0.4599,
       "step": 8815
     },
     {
-      "epoch": 0.31,
-      "learning_rate": 4.9865362156955674e-05,
-      "loss": 0.3355,
+      "epoch": 0.3178722024002595,
+      "grad_norm": 0.2578841745853424,
+      "learning_rate": 4.985252708156616e-05,
+      "loss": 0.4756,
       "step": 8820
     },
     {
-      "epoch": 0.31,
-      "learning_rate": 4.9865066743410946e-05,
-      "loss": 0.3326,
+      "epoch": 0.3180524020614841,
+      "grad_norm": 0.19595186412334442,
+      "learning_rate": 4.985221041868748e-05,
+      "loss": 0.4905,
       "step": 8825
     },
     {
-      "epoch": 0.31,
-      "learning_rate": 4.986477100700972e-05,
-      "loss": 0.3034,
+      "epoch": 0.3182326017227088,
+      "grad_norm": 0.1569208949804306,
+      "learning_rate": 4.9851893417202247e-05,
+      "loss": 0.4481,
       "step": 8830
     },
     {
-      "epoch": 0.31,
-      "learning_rate": 4.9864474947755816e-05,
-      "loss": 0.3149,
+      "epoch": 0.3184128013839334,
+      "grad_norm": 0.1994239240884781,
+      "learning_rate": 4.9851576077114784e-05,
+      "loss": 0.4628,
       "step": 8835
     },
     {
-      "epoch": 0.31,
-      "learning_rate": 4.9864178565653084e-05,
-      "loss": 0.3197,
+      "epoch": 0.31859300104515803,
+      "grad_norm": 0.15360099077224731,
+      "learning_rate": 4.98512583984294e-05,
+      "loss": 0.4508,
       "step": 8840
     },
     {
-      "epoch": 0.31,
-      "learning_rate": 4.98638818607054e-05,
-      "loss": 0.3333,
+      "epoch": 0.31877320070638265,
+      "grad_norm": 0.16948221623897552,
+      "learning_rate": 4.985094038115043e-05,
+      "loss": 0.4559,
       "step": 8845
     },
     {
-      "epoch": 0.31,
-      "learning_rate": 4.986358483291658e-05,
-      "loss": 0.3281,
+      "epoch": 0.31895340036760733,
+      "grad_norm": 0.1563452035188675,
+      "learning_rate": 4.985062202528221e-05,
+      "loss": 0.4163,
       "step": 8850
     },
     {
-      "epoch": 0.31,
-      "learning_rate": 4.98632874822905e-05,
-      "loss": 0.3363,
+      "epoch": 0.31913360002883195,
+      "grad_norm": 0.20190612971782684,
+      "learning_rate": 4.985030333082908e-05,
+      "loss": 0.4037,
       "step": 8855
     },
     {
-      "epoch": 0.31,
-      "learning_rate": 4.986298980883102e-05,
-      "loss": 0.3327,
+      "epoch": 0.3193137996900566,
+      "grad_norm": 0.17092286050319672,
+      "learning_rate": 4.984998429779538e-05,
+      "loss": 0.4339,
       "step": 8860
     },
     {
-      "epoch": 0.31,
-      "learning_rate": 4.9862691812542e-05,
-      "loss": 0.3381,
+      "epoch": 0.3194939993512812,
+      "grad_norm": 0.1695345640182495,
+      "learning_rate": 4.9849664926185445e-05,
+      "loss": 0.4485,
       "step": 8865
     },
     {
-      "epoch": 0.31,
-      "learning_rate": 4.986239349342732e-05,
-      "loss": 0.3479,
+      "epoch": 0.3196741990125059,
+      "grad_norm": 0.17635978758335114,
+      "learning_rate": 4.9849345216003654e-05,
+      "loss": 0.4437,
       "step": 8870
     },
     {
-      "epoch": 0.31,
-      "learning_rate": 4.9862094851490836e-05,
-      "loss": 0.3281,
+      "epoch": 0.3198543986737305,
+      "grad_norm": 0.17225243151187897,
+      "learning_rate": 4.9849025167254324e-05,
+      "loss": 0.4684,
       "step": 8875
     },
     {
-      "epoch": 0.31,
-      "learning_rate": 4.986179588673644e-05,
-      "loss": 0.3317,
+      "epoch": 0.3200345983349551,
+      "grad_norm": 0.21078833937644958,
+      "learning_rate": 4.984870477994186e-05,
+      "loss": 0.473,
       "step": 8880
     },
     {
-      "epoch": 0.31,
-      "learning_rate": 4.9861496599168006e-05,
-      "loss": 0.3445,
+      "epoch": 0.32021479799617975,
+      "grad_norm": 0.16367143392562866,
+      "learning_rate": 4.9848384054070584e-05,
+      "loss": 0.4629,
       "step": 8885
     },
     {
-      "epoch": 0.31,
-      "learning_rate": 4.9861196988789425e-05,
-      "loss": 0.3412,
+      "epoch": 0.3203949976574044,
+      "grad_norm": 0.1702301949262619,
+      "learning_rate": 4.9848062989644894e-05,
+      "loss": 0.445,
       "step": 8890
     },
     {
-      "epoch": 0.31,
-      "learning_rate": 4.986089705560458e-05,
-      "loss": 0.3177,
+      "epoch": 0.32057519731862905,
+      "grad_norm": 0.1748560667037964,
+      "learning_rate": 4.984774158666916e-05,
+      "loss": 0.4815,
       "step": 8895
     },
     {
-      "epoch": 0.31,
-      "learning_rate": 4.9860596799617375e-05,
-      "loss": 0.3395,
+      "epoch": 0.32075539697985367,
+      "grad_norm": 0.20928102731704712,
+      "learning_rate": 4.9847419845147755e-05,
+      "loss": 0.4564,
       "step": 8900
     },
     {
-      "epoch": 0.31,
-      "learning_rate": 4.986029622083171e-05,
-      "loss": 0.3146,
+      "epoch": 0.3209355966410783,
+      "grad_norm": 0.28013554215431213,
+      "learning_rate": 4.984709776508506e-05,
+      "loss": 0.473,
       "step": 8905
     },
     {
-      "epoch": 0.31,
-      "learning_rate": 4.9859995319251475e-05,
-      "loss": 0.3429,
+      "epoch": 0.32111579630230297,
+      "grad_norm": 0.14769336581230164,
+      "learning_rate": 4.984677534648548e-05,
+      "loss": 0.4582,
       "step": 8910
     },
     {
-      "epoch": 0.31,
-      "learning_rate": 4.985969409488059e-05,
-      "loss": 0.3353,
+      "epoch": 0.3212959959635276,
+      "grad_norm": 0.2364906519651413,
+      "learning_rate": 4.984645258935339e-05,
+      "loss": 0.468,
       "step": 8915
     },
     {
-      "epoch": 0.31,
-      "learning_rate": 4.985939254772295e-05,
-      "loss": 0.325,
+      "epoch": 0.3214761956247522,
+      "grad_norm": 0.13622739911079407,
+      "learning_rate": 4.9846129493693183e-05,
+      "loss": 0.4214,
       "step": 8920
     },
     {
-      "epoch": 0.31,
-      "learning_rate": 4.9859090677782494e-05,
-      "loss": 0.3591,
+      "epoch": 0.32165639528597684,
+      "grad_norm": 0.2001640349626541,
+      "learning_rate": 4.984580605950929e-05,
+      "loss": 0.4832,
       "step": 8925
     },
     {
-      "epoch": 0.31,
-      "learning_rate": 4.985878848506312e-05,
-      "loss": 0.3047,
+      "epoch": 0.3218365949472015,
+      "grad_norm": 0.18895497918128967,
+      "learning_rate": 4.984548228680609e-05,
+      "loss": 0.498,
       "step": 8930
     },
     {
-      "epoch": 0.31,
-      "learning_rate": 4.9858485969568755e-05,
-      "loss": 0.3302,
+      "epoch": 0.32201679460842614,
+      "grad_norm": 0.18323996663093567,
+      "learning_rate": 4.9845158175588006e-05,
+      "loss": 0.4827,
       "step": 8935
     },
     {
-      "epoch": 0.31,
-      "learning_rate": 4.9858183131303344e-05,
-      "loss": 0.2988,
+      "epoch": 0.32219699426965076,
+      "grad_norm": 0.16754454374313354,
+      "learning_rate": 4.9844833725859454e-05,
+      "loss": 0.4472,
       "step": 8940
     },
     {
-      "epoch": 0.31,
-      "learning_rate": 4.9857879970270805e-05,
-      "loss": 0.3423,
+      "epoch": 0.3223771939308754,
+      "grad_norm": 0.1655125617980957,
+      "learning_rate": 4.9844508937624844e-05,
+      "loss": 0.4637,
       "step": 8945
     },
     {
-      "epoch": 0.31,
-      "learning_rate": 4.985757648647507e-05,
-      "loss": 0.3119,
+      "epoch": 0.32255739359210006,
+      "grad_norm": 0.20413149893283844,
+      "learning_rate": 4.984418381088862e-05,
+      "loss": 0.444,
       "step": 8950
     },
     {
-      "epoch": 0.32,
-      "learning_rate": 4.985727267992009e-05,
-      "loss": 0.3219,
+      "epoch": 0.3227375932533247,
+      "grad_norm": 0.16138513386249542,
+      "learning_rate": 4.98438583456552e-05,
+      "loss": 0.4388,
       "step": 8955
     },
     {
-      "epoch": 0.32,
-      "learning_rate": 4.98569685506098e-05,
-      "loss": 0.3093,
+      "epoch": 0.3229177929145493,
+      "grad_norm": 0.194508358836174,
+      "learning_rate": 4.9843532541929016e-05,
+      "loss": 0.4823,
       "step": 8960
     },
     {
-      "epoch": 0.32,
-      "learning_rate": 4.9856664098548165e-05,
-      "loss": 0.3414,
+      "epoch": 0.323097992575774,
+      "grad_norm": 0.19578173756599426,
+      "learning_rate": 4.9843206399714516e-05,
+      "loss": 0.4623,
       "step": 8965
     },
     {
-      "epoch": 0.32,
-      "learning_rate": 4.9856359323739124e-05,
-      "loss": 0.3321,
+      "epoch": 0.3232781922369986,
+      "grad_norm": 0.19344592094421387,
+      "learning_rate": 4.984287991901613e-05,
+      "loss": 0.4333,
       "step": 8970
     },
     {
-      "epoch": 0.32,
-      "learning_rate": 4.985605422618663e-05,
-      "loss": 0.3158,
+      "epoch": 0.32345839189822323,
+      "grad_norm": 0.20126448571681976,
+      "learning_rate": 4.9842553099838324e-05,
+      "loss": 0.4837,
       "step": 8975
     },
     {
-      "epoch": 0.32,
-      "learning_rate": 4.985574880589467e-05,
-      "loss": 0.3373,
+      "epoch": 0.32363859155944785,
+      "grad_norm": 0.1657496839761734,
+      "learning_rate": 4.9842225942185536e-05,
+      "loss": 0.4983,
       "step": 8980
     },
     {
-      "epoch": 0.32,
-      "learning_rate": 4.985544306286717e-05,
-      "loss": 0.341,
+      "epoch": 0.32381879122067253,
+      "grad_norm": 0.14652882516384125,
+      "learning_rate": 4.984189844606223e-05,
+      "loss": 0.4724,
       "step": 8985
     },
     {
-      "epoch": 0.32,
-      "learning_rate": 4.9855136997108146e-05,
-      "loss": 0.334,
+      "epoch": 0.32399899088189715,
+      "grad_norm": 0.15776591002941132,
+      "learning_rate": 4.984157061147287e-05,
+      "loss": 0.4545,
       "step": 8990
     },
     {
-      "epoch": 0.32,
-      "learning_rate": 4.985483060862153e-05,
-      "loss": 0.3127,
+      "epoch": 0.3241791905431218,
+      "grad_norm": 0.1731184720993042,
+      "learning_rate": 4.984124243842192e-05,
+      "loss": 0.4698,
       "step": 8995
     },
     {
-      "epoch": 0.32,
-      "learning_rate": 4.985452389741132e-05,
-      "loss": 0.3007,
+      "epoch": 0.3243593902043464,
+      "grad_norm": 0.16583359241485596,
+      "learning_rate": 4.984091392691385e-05,
+      "loss": 0.4673,
       "step": 9000
     },
     {
-      "epoch": 0.32,
-      "eval_loss": 0.32597243785858154,
-      "eval_runtime": 10.5417,
-      "eval_samples_per_second": 9.486,
-      "eval_steps_per_second": 9.486,
+      "epoch": 0.3243593902043464,
+      "eval_loss": 0.48290419578552246,
+      "eval_runtime": 3.5289,
+      "eval_samples_per_second": 28.337,
+      "eval_steps_per_second": 7.084,
       "step": 9000
     },
     {
-      "epoch": 0.32,
-      "learning_rate": 4.9854216863481505e-05,
-      "loss": 0.357,
+      "epoch": 0.3245395898655711,
+      "grad_norm": 0.22497062385082245,
+      "learning_rate": 4.984058507695314e-05,
+      "loss": 0.4875,
       "step": 9005
     },
     {
-      "epoch": 0.32,
-      "learning_rate": 4.985390950683606e-05,
-      "loss": 0.3185,
+      "epoch": 0.3247197895267957,
+      "grad_norm": 0.1878102570772171,
+      "learning_rate": 4.9840255888544265e-05,
+      "loss": 0.4555,
       "step": 9010
     },
     {
-      "epoch": 0.32,
-      "learning_rate": 4.985360182747898e-05,
-      "loss": 0.3264,
+      "epoch": 0.3248999891880203,
+      "grad_norm": 0.2143104523420334,
+      "learning_rate": 4.983992636169171e-05,
+      "loss": 0.4425,
       "step": 9015
     },
     {
-      "epoch": 0.32,
-      "learning_rate": 4.985329382541426e-05,
-      "loss": 0.333,
+      "epoch": 0.32508018884924494,
+      "grad_norm": 0.18537423014640808,
+      "learning_rate": 4.9839596496399964e-05,
+      "loss": 0.4809,
       "step": 9020
     },
     {
-      "epoch": 0.32,
-      "learning_rate": 4.98529855006459e-05,
-      "loss": 0.3089,
+      "epoch": 0.3252603885104696,
+      "grad_norm": 0.1459311693906784,
+      "learning_rate": 4.983926629267353e-05,
+      "loss": 0.4394,
       "step": 9025
     },
     {
-      "epoch": 0.32,
-      "learning_rate": 4.985267685317789e-05,
-      "loss": 0.3267,
+      "epoch": 0.32544058817169425,
+      "grad_norm": 0.16225488483905792,
+      "learning_rate": 4.98389357505169e-05,
+      "loss": 0.4756,
       "step": 9030
     },
     {
-      "epoch": 0.32,
-      "learning_rate": 4.985236788301427e-05,
-      "loss": 0.3334,
+      "epoch": 0.32562078783291887,
+      "grad_norm": 0.21932868659496307,
+      "learning_rate": 4.983860486993458e-05,
+      "loss": 0.4869,
       "step": 9035
     },
     {
-      "epoch": 0.32,
-      "learning_rate": 4.985205859015902e-05,
-      "loss": 0.3135,
+      "epoch": 0.3258009874941435,
+      "grad_norm": 0.17709140479564667,
+      "learning_rate": 4.983827365093109e-05,
+      "loss": 0.4538,
       "step": 9040
     },
     {
-      "epoch": 0.32,
-      "learning_rate": 4.985174897461616e-05,
-      "loss": 0.338,
+      "epoch": 0.32598118715536817,
+      "grad_norm": 0.17748679220676422,
+      "learning_rate": 4.9837942093510914e-05,
+      "loss": 0.4759,
       "step": 9045
     },
     {
-      "epoch": 0.32,
-      "learning_rate": 4.985143903638973e-05,
-      "loss": 0.3489,
+      "epoch": 0.3261613868165928,
+      "grad_norm": 0.17787578701972961,
+      "learning_rate": 4.983761019767859e-05,
+      "loss": 0.4355,
       "step": 9050
     },
     {
-      "epoch": 0.32,
-      "learning_rate": 4.985112877548373e-05,
-      "loss": 0.3145,
+      "epoch": 0.3263415864778174,
+      "grad_norm": 0.21222223341464996,
+      "learning_rate": 4.983727796343864e-05,
+      "loss": 0.4351,
       "step": 9055
     },
     {
-      "epoch": 0.32,
-      "learning_rate": 4.9850818191902216e-05,
-      "loss": 0.3318,
+      "epoch": 0.32652178613904204,
+      "grad_norm": 0.18843764066696167,
+      "learning_rate": 4.983694539079558e-05,
+      "loss": 0.4608,
       "step": 9060
     },
     {
-      "epoch": 0.32,
-      "learning_rate": 4.9850507285649196e-05,
-      "loss": 0.316,
+      "epoch": 0.3267019858002667,
+      "grad_norm": 0.2049497812986374,
+      "learning_rate": 4.9836612479753955e-05,
+      "loss": 0.468,
       "step": 9065
     },
     {
-      "epoch": 0.32,
-      "learning_rate": 4.985019605672871e-05,
-      "loss": 0.3234,
+      "epoch": 0.32688218546149134,
+      "grad_norm": 0.1821064054965973,
+      "learning_rate": 4.9836279230318286e-05,
+      "loss": 0.4842,
       "step": 9070
     },
     {
-      "epoch": 0.32,
-      "learning_rate": 4.984988450514482e-05,
-      "loss": 0.3117,
+      "epoch": 0.32706238512271596,
+      "grad_norm": 0.18487945199012756,
+      "learning_rate": 4.983594564249312e-05,
+      "loss": 0.4394,
       "step": 9075
     },
     {
-      "epoch": 0.32,
-      "learning_rate": 4.984957263090154e-05,
-      "loss": 0.3166,
+      "epoch": 0.3272425847839406,
+      "grad_norm": 0.19593387842178345,
+      "learning_rate": 4.9835611716283015e-05,
+      "loss": 0.4812,
       "step": 9080
     },
     {
-      "epoch": 0.32,
-      "learning_rate": 4.984926043400295e-05,
-      "loss": 0.3375,
+      "epoch": 0.32742278444516526,
+      "grad_norm": 0.1479068547487259,
+      "learning_rate": 4.98352774516925e-05,
+      "loss": 0.4608,
       "step": 9085
     },
     {
-      "epoch": 0.32,
-      "learning_rate": 4.984894791445308e-05,
-      "loss": 0.32,
+      "epoch": 0.3276029841063899,
+      "grad_norm": 0.21857762336730957,
+      "learning_rate": 4.983494284872614e-05,
+      "loss": 0.4715,
       "step": 9090
     },
     {
-      "epoch": 0.32,
-      "learning_rate": 4.984863507225601e-05,
-      "loss": 0.3319,
+      "epoch": 0.3277831837676145,
+      "grad_norm": 0.18269813060760498,
+      "learning_rate": 4.983460790738849e-05,
+      "loss": 0.4397,
       "step": 9095
     },
     {
-      "epoch": 0.32,
-      "learning_rate": 4.984832190741577e-05,
-      "loss": 0.3137,
+      "epoch": 0.32796338342883913,
+      "grad_norm": 0.1688721776008606,
+      "learning_rate": 4.983427262768411e-05,
+      "loss": 0.4727,
       "step": 9100
     },
     {
-      "epoch": 0.32,
-      "learning_rate": 4.984800841993647e-05,
-      "loss": 0.3157,
+      "epoch": 0.3281435830900638,
+      "grad_norm": 0.2021578699350357,
+      "learning_rate": 4.983393700961758e-05,
+      "loss": 0.4429,
       "step": 9105
     },
     {
-      "epoch": 0.32,
-      "learning_rate": 4.9847694609822133e-05,
-      "loss": 0.3358,
+      "epoch": 0.32832378275128843,
+      "grad_norm": 0.16843554377555847,
+      "learning_rate": 4.9833601053193465e-05,
+      "loss": 0.4566,
       "step": 9110
     },
     {
-      "epoch": 0.32,
-      "learning_rate": 4.9847380477076866e-05,
-      "loss": 0.3316,
+      "epoch": 0.32850398241251305,
+      "grad_norm": 0.19692714512348175,
+      "learning_rate": 4.983326475841635e-05,
+      "loss": 0.4815,
       "step": 9115
     },
     {
-      "epoch": 0.32,
-      "learning_rate": 4.984706602170474e-05,
-      "loss": 0.3314,
+      "epoch": 0.3286841820737377,
+      "grad_norm": 0.20175185799598694,
+      "learning_rate": 4.983292812529081e-05,
+      "loss": 0.4429,
       "step": 9120
     },
     {
-      "epoch": 0.32,
-      "learning_rate": 4.9846751243709835e-05,
-      "loss": 0.3301,
+      "epoch": 0.32886438173496235,
+      "grad_norm": 0.16880235075950623,
+      "learning_rate": 4.9832591153821424e-05,
+      "loss": 0.4417,
       "step": 9125
     },
     {
-      "epoch": 0.32,
-      "learning_rate": 4.9846436143096245e-05,
-      "loss": 0.3047,
+      "epoch": 0.329044581396187,
+      "grad_norm": 0.1737411469221115,
+      "learning_rate": 4.983225384401279e-05,
+      "loss": 0.4304,
       "step": 9130
     },
     {
-      "epoch": 0.32,
-      "learning_rate": 4.9846120719868045e-05,
-      "loss": 0.324,
+      "epoch": 0.3292247810574116,
+      "grad_norm": 0.20997703075408936,
+      "learning_rate": 4.983191619586951e-05,
+      "loss": 0.4775,
       "step": 9135
     },
     {
-      "epoch": 0.32,
-      "learning_rate": 4.984580497402935e-05,
-      "loss": 0.3417,
+      "epoch": 0.3294049807186363,
+      "grad_norm": 0.20304125547409058,
+      "learning_rate": 4.9831578209396186e-05,
+      "loss": 0.4416,
       "step": 9140
     },
     {
-      "epoch": 0.32,
-      "learning_rate": 4.9845488905584245e-05,
-      "loss": 0.3142,
+      "epoch": 0.3295851803798609,
+      "grad_norm": 0.22506339848041534,
+      "learning_rate": 4.9831239884597407e-05,
+      "loss": 0.531,
       "step": 9145
     },
     {
-      "epoch": 0.32,
-      "learning_rate": 4.9845172514536844e-05,
-      "loss": 0.3444,
+      "epoch": 0.3297653800410855,
+      "grad_norm": 0.16725526750087738,
+      "learning_rate": 4.983090122147779e-05,
+      "loss": 0.4536,
       "step": 9150
     },
     {
-      "epoch": 0.32,
-      "learning_rate": 4.984485580089125e-05,
-      "loss": 0.3562,
+      "epoch": 0.32994557970231014,
+      "grad_norm": 0.1725703775882721,
+      "learning_rate": 4.983056222004196e-05,
+      "loss": 0.4525,
       "step": 9155
     },
     {
-      "epoch": 0.32,
-      "learning_rate": 4.9844538764651584e-05,
-      "loss": 0.3122,
+      "epoch": 0.3301257793635348,
+      "grad_norm": 0.13673150539398193,
+      "learning_rate": 4.9830222880294525e-05,
+      "loss": 0.4515,
       "step": 9160
     },
     {
-      "epoch": 0.32,
-      "learning_rate": 4.9844221405821946e-05,
-      "loss": 0.3444,
+      "epoch": 0.33030597902475944,
+      "grad_norm": 0.19130565226078033,
+      "learning_rate": 4.982988320224011e-05,
+      "loss": 0.4683,
       "step": 9165
     },
     {
-      "epoch": 0.32,
-      "learning_rate": 4.984390372440647e-05,
-      "loss": 0.2972,
+      "epoch": 0.33048617868598407,
+      "grad_norm": 0.2034883350133896,
+      "learning_rate": 4.9829543185883344e-05,
+      "loss": 0.4899,
       "step": 9170
     },
     {
-      "epoch": 0.32,
-      "learning_rate": 4.984358572040928e-05,
-      "loss": 0.3356,
+      "epoch": 0.3306663783472087,
+      "grad_norm": 0.22771266102790833,
+      "learning_rate": 4.982920283122885e-05,
+      "loss": 0.4736,
       "step": 9175
     },
     {
-      "epoch": 0.32,
-      "learning_rate": 4.9843267393834495e-05,
-      "loss": 0.3507,
+      "epoch": 0.33084657800843337,
+      "grad_norm": 0.18065966665744781,
+      "learning_rate": 4.982886213828128e-05,
+      "loss": 0.4542,
       "step": 9180
     },
     {
-      "epoch": 0.32,
-      "learning_rate": 4.9842948744686256e-05,
-      "loss": 0.3003,
+      "epoch": 0.331026777669658,
+      "grad_norm": 0.1368507593870163,
+      "learning_rate": 4.9828521107045276e-05,
+      "loss": 0.4624,
       "step": 9185
     },
     {
-      "epoch": 0.32,
-      "learning_rate": 4.984262977296871e-05,
-      "loss": 0.3545,
+      "epoch": 0.3312069773308826,
+      "grad_norm": 0.16015149652957916,
+      "learning_rate": 4.982817973752548e-05,
+      "loss": 0.4361,
       "step": 9190
     },
     {
-      "epoch": 0.32,
-      "learning_rate": 4.9842310478685975e-05,
-      "loss": 0.3167,
+      "epoch": 0.33138717699210724,
+      "grad_norm": 0.19293028116226196,
+      "learning_rate": 4.9827838029726535e-05,
+      "loss": 0.4403,
       "step": 9195
     },
     {
-      "epoch": 0.32,
-      "learning_rate": 4.984199086184222e-05,
-      "loss": 0.3063,
+      "epoch": 0.3315673766533319,
+      "grad_norm": 0.18895958364009857,
+      "learning_rate": 4.9827495983653104e-05,
+      "loss": 0.4999,
       "step": 9200
     },
     {
-      "epoch": 0.32,
-      "learning_rate": 4.984167092244158e-05,
-      "loss": 0.3186,
+      "epoch": 0.33174757631455654,
+      "grad_norm": 0.16563822329044342,
+      "learning_rate": 4.982715359930985e-05,
+      "loss": 0.436,
       "step": 9205
     },
     {
-      "epoch": 0.32,
-      "learning_rate": 4.9841350660488215e-05,
-      "loss": 0.3231,
+      "epoch": 0.33192777597578116,
+      "grad_norm": 0.21397612988948822,
+      "learning_rate": 4.982681087670144e-05,
+      "loss": 0.4835,
       "step": 9210
     },
     {
-      "epoch": 0.32,
-      "learning_rate": 4.9841030075986284e-05,
-      "loss": 0.3199,
+      "epoch": 0.3321079756370058,
+      "grad_norm": 0.18013893067836761,
+      "learning_rate": 4.982646781583252e-05,
+      "loss": 0.4622,
       "step": 9215
     },
     {
-      "epoch": 0.32,
-      "learning_rate": 4.9840709168939945e-05,
-      "loss": 0.3169,
+      "epoch": 0.33228817529823046,
+      "grad_norm": 0.19024062156677246,
+      "learning_rate": 4.98261244167078e-05,
+      "loss": 0.4419,
       "step": 9220
     },
     {
-      "epoch": 0.32,
-      "learning_rate": 4.9840387939353375e-05,
-      "loss": 0.3304,
+      "epoch": 0.3324683749594551,
+      "grad_norm": 0.18336628377437592,
+      "learning_rate": 4.9825780679331935e-05,
+      "loss": 0.4525,
       "step": 9225
     },
     {
-      "epoch": 0.32,
-      "learning_rate": 4.984006638723073e-05,
-      "loss": 0.3147,
+      "epoch": 0.3326485746206797,
+      "grad_norm": 0.17188718914985657,
+      "learning_rate": 4.982543660370962e-05,
+      "loss": 0.4737,
       "step": 9230
     },
     {
-      "epoch": 0.32,
-      "learning_rate": 4.9839744512576205e-05,
-      "loss": 0.329,
+      "epoch": 0.3328287742819043,
+      "grad_norm": 0.1877148300409317,
+      "learning_rate": 4.982509218984553e-05,
+      "loss": 0.4557,
       "step": 9235
     },
     {
-      "epoch": 0.33,
-      "learning_rate": 4.9839422315393954e-05,
-      "loss": 0.3072,
+      "epoch": 0.333008973943129,
+      "grad_norm": 0.2199414074420929,
+      "learning_rate": 4.982474743774437e-05,
+      "loss": 0.4892,
       "step": 9240
     },
     {
-      "epoch": 0.33,
-      "learning_rate": 4.983909979568818e-05,
-      "loss": 0.3553,
+      "epoch": 0.33318917360435363,
+      "grad_norm": 0.17190028727054596,
+      "learning_rate": 4.982440234741082e-05,
+      "loss": 0.487,
       "step": 9245
     },
     {
-      "epoch": 0.33,
-      "learning_rate": 4.983877695346306e-05,
-      "loss": 0.332,
+      "epoch": 0.33336937326557825,
+      "grad_norm": 0.2042684108018875,
+      "learning_rate": 4.9824056918849614e-05,
+      "loss": 0.4648,
       "step": 9250
     },
     {
-      "epoch": 0.33,
-      "learning_rate": 4.983845378872279e-05,
-      "loss": 0.3178,
+      "epoch": 0.3335495729268029,
+      "grad_norm": 0.15054477751255035,
+      "learning_rate": 4.9823711152065425e-05,
+      "loss": 0.4622,
       "step": 9255
     },
     {
-      "epoch": 0.33,
-      "learning_rate": 4.983813030147158e-05,
-      "loss": 0.328,
+      "epoch": 0.33372977258802755,
+      "grad_norm": 0.20531079173088074,
+      "learning_rate": 4.9823365047062986e-05,
+      "loss": 0.466,
       "step": 9260
     },
     {
-      "epoch": 0.33,
-      "learning_rate": 4.9837806491713606e-05,
-      "loss": 0.3063,
+      "epoch": 0.3339099722492522,
+      "grad_norm": 0.17524632811546326,
+      "learning_rate": 4.9823018603847e-05,
+      "loss": 0.4271,
       "step": 9265
     },
     {
-      "epoch": 0.33,
-      "learning_rate": 4.983748235945309e-05,
-      "loss": 0.3263,
+      "epoch": 0.3340901719104768,
+      "grad_norm": 0.21310240030288696,
+      "learning_rate": 4.9822671822422195e-05,
+      "loss": 0.4699,
       "step": 9270
     },
     {
-      "epoch": 0.33,
-      "learning_rate": 4.9837157904694234e-05,
-      "loss": 0.3139,
+      "epoch": 0.3342703715717014,
+      "grad_norm": 0.19141936302185059,
+      "learning_rate": 4.982232470279329e-05,
+      "loss": 0.4508,
       "step": 9275
     },
     {
-      "epoch": 0.33,
-      "learning_rate": 4.983683312744124e-05,
-      "loss": 0.3351,
+      "epoch": 0.3344505712329261,
+      "grad_norm": 0.21620051562786102,
+      "learning_rate": 4.9821977244965014e-05,
+      "loss": 0.502,
       "step": 9280
     },
     {
-      "epoch": 0.33,
-      "learning_rate": 4.983650802769834e-05,
-      "loss": 0.3058,
+      "epoch": 0.3346307708941507,
+      "grad_norm": 0.16024553775787354,
+      "learning_rate": 4.98216294489421e-05,
+      "loss": 0.442,
       "step": 9285
     },
     {
-      "epoch": 0.33,
-      "learning_rate": 4.983618260546976e-05,
-      "loss": 0.3224,
+      "epoch": 0.33481097055537534,
+      "grad_norm": 0.21437714993953705,
+      "learning_rate": 4.982128131472929e-05,
+      "loss": 0.4775,
       "step": 9290
     },
     {
-      "epoch": 0.33,
-      "learning_rate": 4.9835856860759714e-05,
-      "loss": 0.3122,
+      "epoch": 0.3349911702166,
+      "grad_norm": 0.18475022912025452,
+      "learning_rate": 4.982093284233134e-05,
+      "loss": 0.4903,
       "step": 9295
     },
     {
-      "epoch": 0.33,
-      "learning_rate": 4.9835530793572425e-05,
-      "loss": 0.2956,
+      "epoch": 0.33517136987782464,
+      "grad_norm": 0.13329213857650757,
+      "learning_rate": 4.982058403175298e-05,
+      "loss": 0.443,
       "step": 9300
     },
     {
-      "epoch": 0.33,
-      "learning_rate": 4.983520440391214e-05,
-      "loss": 0.3209,
+      "epoch": 0.33535156953904927,
+      "grad_norm": 0.16213421523571014,
+      "learning_rate": 4.982023488299897e-05,
+      "loss": 0.4502,
       "step": 9305
     },
     {
-      "epoch": 0.33,
-      "learning_rate": 4.9834877691783094e-05,
-      "loss": 0.3261,
+      "epoch": 0.3355317692002739,
+      "grad_norm": 0.19696162641048431,
+      "learning_rate": 4.981988539607406e-05,
+      "loss": 0.4939,
       "step": 9310
     },
     {
-      "epoch": 0.33,
-      "learning_rate": 4.983455065718953e-05,
-      "loss": 0.3342,
+      "epoch": 0.33571196886149857,
+      "grad_norm": 0.153734490275383,
+      "learning_rate": 4.981953557098302e-05,
+      "loss": 0.4557,
       "step": 9315
     },
     {
-      "epoch": 0.33,
-      "learning_rate": 4.983422330013569e-05,
-      "loss": 0.3251,
+      "epoch": 0.3358921685227232,
+      "grad_norm": 0.16626779735088348,
+      "learning_rate": 4.981918540773061e-05,
+      "loss": 0.4774,
       "step": 9320
     },
     {
-      "epoch": 0.33,
-      "learning_rate": 4.983389562062583e-05,
-      "loss": 0.3228,
+      "epoch": 0.3360723681839478,
+      "grad_norm": 0.2404673844575882,
+      "learning_rate": 4.981883490632161e-05,
+      "loss": 0.4717,
       "step": 9325
     },
     {
-      "epoch": 0.33,
-      "learning_rate": 4.98335676186642e-05,
-      "loss": 0.3524,
+      "epoch": 0.33625256784517243,
+      "grad_norm": 0.19677095115184784,
+      "learning_rate": 4.9818484066760786e-05,
+      "loss": 0.4442,
       "step": 9330
     },
     {
-      "epoch": 0.33,
-      "learning_rate": 4.9833239294255054e-05,
-      "loss": 0.342,
+      "epoch": 0.3364327675063971,
+      "grad_norm": 0.17953583598136902,
+      "learning_rate": 4.9818132889052914e-05,
+      "loss": 0.4613,
       "step": 9335
     },
     {
-      "epoch": 0.33,
-      "learning_rate": 4.9832910647402666e-05,
-      "loss": 0.3058,
+      "epoch": 0.33661296716762173,
+      "grad_norm": 0.19360299408435822,
+      "learning_rate": 4.98177813732028e-05,
+      "loss": 0.4416,
       "step": 9340
     },
     {
-      "epoch": 0.33,
-      "learning_rate": 4.9832581678111304e-05,
-      "loss": 0.3527,
+      "epoch": 0.33679316682884636,
+      "grad_norm": 0.16744250059127808,
+      "learning_rate": 4.9817429519215206e-05,
+      "loss": 0.454,
       "step": 9345
     },
     {
-      "epoch": 0.33,
-      "learning_rate": 4.983225238638523e-05,
-      "loss": 0.3186,
+      "epoch": 0.336973366490071,
+      "grad_norm": 0.15772564709186554,
+      "learning_rate": 4.981707732709495e-05,
+      "loss": 0.4651,
       "step": 9350
     },
     {
-      "epoch": 0.33,
-      "learning_rate": 4.983192277222872e-05,
-      "loss": 0.3057,
+      "epoch": 0.33715356615129566,
+      "grad_norm": 0.16711197793483734,
+      "learning_rate": 4.9816724796846814e-05,
+      "loss": 0.4269,
       "step": 9355
     },
     {
-      "epoch": 0.33,
-      "learning_rate": 4.9831592835646064e-05,
-      "loss": 0.3024,
+      "epoch": 0.3373337658125203,
+      "grad_norm": 0.14946970343589783,
+      "learning_rate": 4.981637192847561e-05,
+      "loss": 0.4646,
       "step": 9360
     },
     {
-      "epoch": 0.33,
-      "learning_rate": 4.983126257664153e-05,
-      "loss": 0.3142,
+      "epoch": 0.3375139654737449,
+      "grad_norm": 0.20291027426719666,
+      "learning_rate": 4.9816018721986145e-05,
+      "loss": 0.4598,
       "step": 9365
     },
     {
-      "epoch": 0.33,
-      "learning_rate": 4.9830931995219426e-05,
-      "loss": 0.3291,
+      "epoch": 0.3376941651349695,
+      "grad_norm": 0.1772489696741104,
+      "learning_rate": 4.981566517738323e-05,
+      "loss": 0.4794,
       "step": 9370
     },
     {
-      "epoch": 0.33,
-      "learning_rate": 4.983060109138403e-05,
-      "loss": 0.3053,
+      "epoch": 0.3378743647961942,
+      "grad_norm": 0.1733308732509613,
+      "learning_rate": 4.981531129467168e-05,
+      "loss": 0.4581,
       "step": 9375
     },
     {
-      "epoch": 0.33,
-      "learning_rate": 4.9830269865139643e-05,
-      "loss": 0.34,
+      "epoch": 0.3380545644574188,
+      "grad_norm": 0.13439474999904633,
+      "learning_rate": 4.981495707385632e-05,
+      "loss": 0.4519,
       "step": 9380
     },
     {
-      "epoch": 0.33,
-      "learning_rate": 4.9829938316490574e-05,
-      "loss": 0.3537,
+      "epoch": 0.33823476411864345,
+      "grad_norm": 0.15566691756248474,
+      "learning_rate": 4.9814602514941965e-05,
+      "loss": 0.452,
       "step": 9385
     },
     {
-      "epoch": 0.33,
-      "learning_rate": 4.9829606445441106e-05,
-      "loss": 0.3217,
+      "epoch": 0.33841496377986807,
+      "grad_norm": 0.1827496588230133,
+      "learning_rate": 4.981424761793346e-05,
+      "loss": 0.4634,
       "step": 9390
     },
     {
-      "epoch": 0.33,
-      "learning_rate": 4.982927425199557e-05,
-      "loss": 0.2932,
+      "epoch": 0.33859516344109275,
+      "grad_norm": 0.16756999492645264,
+      "learning_rate": 4.9813892382835635e-05,
+      "loss": 0.485,
       "step": 9395
     },
     {
-      "epoch": 0.33,
-      "learning_rate": 4.982894173615827e-05,
-      "loss": 0.3026,
+      "epoch": 0.3387753631023174,
+      "grad_norm": 0.17952053248882294,
+      "learning_rate": 4.981353680965334e-05,
+      "loss": 0.4769,
       "step": 9400
     },
     {
-      "epoch": 0.33,
-      "learning_rate": 4.982860889793352e-05,
-      "loss": 0.3515,
+      "epoch": 0.338955562763542,
+      "grad_norm": 0.14793772995471954,
+      "learning_rate": 4.98131808983914e-05,
+      "loss": 0.4435,
       "step": 9405
     },
     {
-      "epoch": 0.33,
-      "learning_rate": 4.982827573732565e-05,
-      "loss": 0.3247,
+      "epoch": 0.3391357624247666,
+      "grad_norm": 0.19632504880428314,
+      "learning_rate": 4.9812824649054674e-05,
+      "loss": 0.3992,
       "step": 9410
     },
     {
-      "epoch": 0.33,
-      "learning_rate": 4.9827942254338975e-05,
-      "loss": 0.3391,
+      "epoch": 0.3393159620859913,
+      "grad_norm": 0.19574934244155884,
+      "learning_rate": 4.9812468061648024e-05,
+      "loss": 0.489,
       "step": 9415
     },
     {
-      "epoch": 0.33,
-      "learning_rate": 4.982760844897785e-05,
-      "loss": 0.3342,
+      "epoch": 0.3394961617472159,
+      "grad_norm": 0.22474491596221924,
+      "learning_rate": 4.981211113617629e-05,
+      "loss": 0.4526,
       "step": 9420
     },
     {
-      "epoch": 0.33,
-      "learning_rate": 4.9827274321246575e-05,
-      "loss": 0.3341,
+      "epoch": 0.33967636140844054,
+      "grad_norm": 0.18208986520767212,
+      "learning_rate": 4.981175387264435e-05,
+      "loss": 0.4648,
       "step": 9425
     },
     {
-      "epoch": 0.33,
-      "learning_rate": 4.98269398711495e-05,
-      "loss": 0.3457,
+      "epoch": 0.33985656106966516,
+      "grad_norm": 0.16022849082946777,
+      "learning_rate": 4.9811396271057067e-05,
+      "loss": 0.4527,
       "step": 9430
     },
     {
-      "epoch": 0.33,
-      "learning_rate": 4.982660509869099e-05,
-      "loss": 0.3443,
+      "epoch": 0.34003676073088984,
+      "grad_norm": 0.21129754185676575,
+      "learning_rate": 4.981103833141931e-05,
+      "loss": 0.4382,
       "step": 9435
     },
     {
-      "epoch": 0.33,
-      "learning_rate": 4.9826270003875364e-05,
-      "loss": 0.3341,
+      "epoch": 0.34021696039211446,
+      "grad_norm": 0.20417428016662598,
+      "learning_rate": 4.981068005373597e-05,
+      "loss": 0.4474,
       "step": 9440
     },
     {
-      "epoch": 0.33,
-      "learning_rate": 4.9825934586706986e-05,
-      "loss": 0.309,
+      "epoch": 0.3403971600533391,
+      "grad_norm": 0.21465253829956055,
+      "learning_rate": 4.981032143801191e-05,
+      "loss": 0.4521,
       "step": 9445
     },
     {
-      "epoch": 0.33,
-      "learning_rate": 4.9825598847190205e-05,
-      "loss": 0.3391,
+      "epoch": 0.3405773597145637,
+      "grad_norm": 0.19207169115543365,
+      "learning_rate": 4.980996248425202e-05,
+      "loss": 0.456,
       "step": 9450
     },
     {
-      "epoch": 0.33,
-      "learning_rate": 4.9825262785329384e-05,
-      "loss": 0.3325,
+      "epoch": 0.3407575593757884,
+      "grad_norm": 0.1812460720539093,
+      "learning_rate": 4.98096031924612e-05,
+      "loss": 0.469,
       "step": 9455
     },
     {
-      "epoch": 0.33,
-      "learning_rate": 4.9824926401128894e-05,
-      "loss": 0.2993,
+      "epoch": 0.340937759037013,
+      "grad_norm": 0.21918563544750214,
+      "learning_rate": 4.9809243562644334e-05,
+      "loss": 0.4829,
       "step": 9460
     },
     {
-      "epoch": 0.33,
-      "learning_rate": 4.9824589694593086e-05,
-      "loss": 0.3034,
+      "epoch": 0.34111795869823763,
+      "grad_norm": 0.18760140240192413,
+      "learning_rate": 4.980888359480634e-05,
+      "loss": 0.4243,
       "step": 9465
     },
     {
-      "epoch": 0.33,
-      "learning_rate": 4.982425266572634e-05,
-      "loss": 0.3118,
+      "epoch": 0.3412981583594623,
+      "grad_norm": 0.1885528415441513,
+      "learning_rate": 4.98085232889521e-05,
+      "loss": 0.4846,
       "step": 9470
     },
     {
-      "epoch": 0.33,
-      "learning_rate": 4.982391531453304e-05,
-      "loss": 0.3312,
+      "epoch": 0.34147835802068693,
+      "grad_norm": 0.16368678212165833,
+      "learning_rate": 4.980816264508654e-05,
+      "loss": 0.4488,
       "step": 9475
     },
     {
-      "epoch": 0.33,
-      "learning_rate": 4.982357764101755e-05,
-      "loss": 0.3397,
+      "epoch": 0.34165855768191156,
+      "grad_norm": 0.21680162847042084,
+      "learning_rate": 4.980780166321456e-05,
+      "loss": 0.4932,
       "step": 9480
     },
     {
-      "epoch": 0.33,
-      "learning_rate": 4.982323964518427e-05,
-      "loss": 0.3235,
+      "epoch": 0.3418387573431362,
+      "grad_norm": 0.19433487951755524,
+      "learning_rate": 4.9807440343341095e-05,
+      "loss": 0.4472,
       "step": 9485
     },
     {
-      "epoch": 0.33,
-      "learning_rate": 4.982290132703759e-05,
-      "loss": 0.3259,
+      "epoch": 0.34201895700436086,
+      "grad_norm": 0.16287468373775482,
+      "learning_rate": 4.980707868547105e-05,
+      "loss": 0.4767,
       "step": 9490
     },
     {
-      "epoch": 0.33,
-      "learning_rate": 4.9822562686581884e-05,
-      "loss": 0.336,
+      "epoch": 0.3421991566655855,
+      "grad_norm": 0.15700995922088623,
+      "learning_rate": 4.9806716689609356e-05,
+      "loss": 0.4508,
       "step": 9495
     },
     {
-      "epoch": 0.33,
-      "learning_rate": 4.982222372382156e-05,
-      "loss": 0.3457,
+      "epoch": 0.3423793563268101,
+      "grad_norm": 0.2554052472114563,
+      "learning_rate": 4.980635435576096e-05,
+      "loss": 0.4737,
       "step": 9500
     },
     {
-      "epoch": 0.33,
-      "eval_loss": 0.32349956035614014,
-      "eval_runtime": 10.5421,
-      "eval_samples_per_second": 9.486,
-      "eval_steps_per_second": 9.486,
+      "epoch": 0.3423793563268101,
+      "eval_loss": 0.4815308451652527,
+      "eval_runtime": 3.5226,
+      "eval_samples_per_second": 28.388,
+      "eval_steps_per_second": 7.097,
       "step": 9500
     },
     {
-      "epoch": 0.33,
-      "learning_rate": 4.982188443876102e-05,
-      "loss": 0.3081,
+      "epoch": 0.3425595559880347,
+      "grad_norm": 0.21748638153076172,
+      "learning_rate": 4.980599168393079e-05,
+      "loss": 0.4881,
       "step": 9505
     },
     {
-      "epoch": 0.33,
-      "learning_rate": 4.982154483140468e-05,
-      "loss": 0.3298,
+      "epoch": 0.3427397556492594,
+      "grad_norm": 0.18842586874961853,
+      "learning_rate": 4.9805628674123774e-05,
+      "loss": 0.4897,
       "step": 9510
     },
     {
-      "epoch": 0.33,
-      "learning_rate": 4.982120490175692e-05,
-      "loss": 0.3431,
+      "epoch": 0.342919955310484,
+      "grad_norm": 0.1802000105381012,
+      "learning_rate": 4.9805265326344874e-05,
+      "loss": 0.4584,
       "step": 9515
     },
     {
-      "epoch": 0.33,
-      "learning_rate": 4.982086464982219e-05,
-      "loss": 0.303,
+      "epoch": 0.34310015497170865,
+      "grad_norm": 0.1921650469303131,
+      "learning_rate": 4.980490164059904e-05,
+      "loss": 0.4693,
       "step": 9520
     },
     {
-      "epoch": 0.34,
-      "learning_rate": 4.982052407560488e-05,
-      "loss": 0.3386,
+      "epoch": 0.34328035463293327,
+      "grad_norm": 0.1768704205751419,
+      "learning_rate": 4.9804537616891224e-05,
+      "loss": 0.4649,
       "step": 9525
     },
     {
-      "epoch": 0.34,
-      "learning_rate": 4.982018317910943e-05,
-      "loss": 0.3359,
+      "epoch": 0.34346055429415795,
+      "grad_norm": 0.1596553921699524,
+      "learning_rate": 4.980417325522638e-05,
+      "loss": 0.4141,
       "step": 9530
     },
     {
-      "epoch": 0.34,
-      "learning_rate": 4.9819841960340246e-05,
-      "loss": 0.3161,
+      "epoch": 0.34364075395538257,
+      "grad_norm": 0.18743056058883667,
+      "learning_rate": 4.980380855560949e-05,
+      "loss": 0.4305,
       "step": 9535
     },
     {
-      "epoch": 0.34,
-      "learning_rate": 4.981950041930178e-05,
-      "loss": 0.3158,
+      "epoch": 0.3438209536166072,
+      "grad_norm": 0.16719353199005127,
+      "learning_rate": 4.98034435180455e-05,
+      "loss": 0.4731,
       "step": 9540
     },
     {
-      "epoch": 0.34,
-      "learning_rate": 4.9819158555998455e-05,
-      "loss": 0.3248,
+      "epoch": 0.3440011532778318,
+      "grad_norm": 0.17522072792053223,
+      "learning_rate": 4.980307814253939e-05,
+      "loss": 0.4487,
       "step": 9545
     },
     {
-      "epoch": 0.34,
-      "learning_rate": 4.981881637043471e-05,
-      "loss": 0.3429,
+      "epoch": 0.3441813529390565,
+      "grad_norm": 0.15065240859985352,
+      "learning_rate": 4.9802712429096154e-05,
+      "loss": 0.4602,
       "step": 9550
     },
     {
-      "epoch": 0.34,
-      "learning_rate": 4.9818473862614996e-05,
-      "loss": 0.3222,
+      "epoch": 0.3443615526002811,
+      "grad_norm": 0.19820521771907806,
+      "learning_rate": 4.980234637772075e-05,
+      "loss": 0.4411,
       "step": 9555
     },
     {
-      "epoch": 0.34,
-      "learning_rate": 4.9818131032543746e-05,
-      "loss": 0.3426,
+      "epoch": 0.34454175226150574,
+      "grad_norm": 0.20624442398548126,
+      "learning_rate": 4.980197998841819e-05,
+      "loss": 0.4873,
       "step": 9560
     },
     {
-      "epoch": 0.34,
-      "learning_rate": 4.981778788022543e-05,
-      "loss": 0.3337,
+      "epoch": 0.34472195192273036,
+      "grad_norm": 0.18093755841255188,
+      "learning_rate": 4.9801613261193455e-05,
+      "loss": 0.46,
       "step": 9565
     },
     {
-      "epoch": 0.34,
-      "learning_rate": 4.981744440566449e-05,
-      "loss": 0.2884,
+      "epoch": 0.34490215158395504,
+      "grad_norm": 0.17363637685775757,
+      "learning_rate": 4.9801246196051535e-05,
+      "loss": 0.4306,
       "step": 9570
     },
     {
-      "epoch": 0.34,
-      "learning_rate": 4.981710060886539e-05,
-      "loss": 0.319,
+      "epoch": 0.34508235124517966,
+      "grad_norm": 0.1665421426296234,
+      "learning_rate": 4.980087879299744e-05,
+      "loss": 0.4478,
       "step": 9575
     },
     {
-      "epoch": 0.34,
-      "learning_rate": 4.9816756489832594e-05,
-      "loss": 0.3413,
+      "epoch": 0.3452625509064043,
+      "grad_norm": 0.18756218254566193,
+      "learning_rate": 4.980051105203617e-05,
+      "loss": 0.5001,
       "step": 9580
     },
     {
-      "epoch": 0.34,
-      "learning_rate": 4.981641204857057e-05,
-      "loss": 0.3181,
+      "epoch": 0.3454427505676289,
+      "grad_norm": 0.16964589059352875,
+      "learning_rate": 4.980014297317274e-05,
+      "loss": 0.4586,
       "step": 9585
     },
     {
-      "epoch": 0.34,
-      "learning_rate": 4.981606728508379e-05,
-      "loss": 0.3521,
+      "epoch": 0.3456229502288536,
+      "grad_norm": 0.24527285993099213,
+      "learning_rate": 4.979977455641217e-05,
+      "loss": 0.456,
       "step": 9590
     },
     {
-      "epoch": 0.34,
-      "learning_rate": 4.9815722199376734e-05,
-      "loss": 0.3322,
+      "epoch": 0.3458031498900782,
+      "grad_norm": 0.22246739268302917,
+      "learning_rate": 4.9799405801759466e-05,
+      "loss": 0.4951,
       "step": 9595
     },
     {
-      "epoch": 0.34,
-      "learning_rate": 4.9815376791453876e-05,
-      "loss": 0.3417,
+      "epoch": 0.34598334955130283,
+      "grad_norm": 0.18261441588401794,
+      "learning_rate": 4.979903670921966e-05,
+      "loss": 0.4364,
       "step": 9600
     },
     {
-      "epoch": 0.34,
-      "learning_rate": 4.9815031061319705e-05,
-      "loss": 0.3309,
+      "epoch": 0.34616354921252745,
+      "grad_norm": 0.15440581738948822,
+      "learning_rate": 4.9798667278797775e-05,
+      "loss": 0.4383,
       "step": 9605
     },
     {
-      "epoch": 0.34,
-      "learning_rate": 4.9814685008978713e-05,
-      "loss": 0.3283,
+      "epoch": 0.34634374887375213,
+      "grad_norm": 0.15516167879104614,
+      "learning_rate": 4.979829751049886e-05,
+      "loss": 0.4524,
       "step": 9610
     },
     {
-      "epoch": 0.34,
-      "learning_rate": 4.981433863443538e-05,
-      "loss": 0.3128,
+      "epoch": 0.34652394853497676,
+      "grad_norm": 0.13671459257602692,
+      "learning_rate": 4.979792740432794e-05,
+      "loss": 0.4862,
       "step": 9615
     },
     {
-      "epoch": 0.34,
-      "learning_rate": 4.981399193769422e-05,
-      "loss": 0.3348,
+      "epoch": 0.3467041481962014,
+      "grad_norm": 0.155983105301857,
+      "learning_rate": 4.9797556960290047e-05,
+      "loss": 0.4566,
       "step": 9620
     },
     {
-      "epoch": 0.34,
-      "learning_rate": 4.9813644918759725e-05,
-      "loss": 0.3077,
+      "epoch": 0.346884347857426,
+      "grad_norm": 0.14357610046863556,
+      "learning_rate": 4.9797186178390255e-05,
+      "loss": 0.4731,
       "step": 9625
     },
     {
-      "epoch": 0.34,
-      "learning_rate": 4.981329757763641e-05,
-      "loss": 0.3369,
+      "epoch": 0.3470645475186507,
+      "grad_norm": 0.19745934009552002,
+      "learning_rate": 4.97968150586336e-05,
+      "loss": 0.5,
       "step": 9630
     },
     {
-      "epoch": 0.34,
-      "learning_rate": 4.9812949914328766e-05,
-      "loss": 0.3688,
+      "epoch": 0.3472447471798753,
+      "grad_norm": 0.19887088239192963,
+      "learning_rate": 4.9796443601025144e-05,
+      "loss": 0.43,
       "step": 9635
     },
     {
-      "epoch": 0.34,
-      "learning_rate": 4.981260192884133e-05,
-      "loss": 0.3196,
+      "epoch": 0.3474249468410999,
+      "grad_norm": 0.17473819851875305,
+      "learning_rate": 4.9796071805569936e-05,
+      "loss": 0.4736,
       "step": 9640
     },
     {
-      "epoch": 0.34,
-      "learning_rate": 4.981225362117861e-05,
-      "loss": 0.3187,
+      "epoch": 0.3476051465023246,
+      "grad_norm": 0.17900219559669495,
+      "learning_rate": 4.9795699672273054e-05,
+      "loss": 0.4633,
       "step": 9645
     },
     {
-      "epoch": 0.34,
-      "learning_rate": 4.981190499134513e-05,
-      "loss": 0.3324,
+      "epoch": 0.3477853461635492,
+      "grad_norm": 0.21898692846298218,
+      "learning_rate": 4.979532720113956e-05,
+      "loss": 0.4815,
       "step": 9650
     },
     {
-      "epoch": 0.34,
-      "learning_rate": 4.981155603934541e-05,
-      "loss": 0.3113,
+      "epoch": 0.34796554582477385,
+      "grad_norm": 0.18762876093387604,
+      "learning_rate": 4.979495439217454e-05,
+      "loss": 0.4904,
       "step": 9655
     },
     {
-      "epoch": 0.34,
-      "learning_rate": 4.981120676518398e-05,
-      "loss": 0.3361,
+      "epoch": 0.34814574548599847,
+      "grad_norm": 0.2149989753961563,
+      "learning_rate": 4.9794581245383074e-05,
+      "loss": 0.468,
       "step": 9660
     },
     {
-      "epoch": 0.34,
-      "learning_rate": 4.9810857168865386e-05,
-      "loss": 0.3341,
+      "epoch": 0.34832594514722315,
+      "grad_norm": 0.1882586032152176,
+      "learning_rate": 4.979420776077023e-05,
+      "loss": 0.4934,
       "step": 9665
     },
     {
-      "epoch": 0.34,
-      "learning_rate": 4.981050725039417e-05,
-      "loss": 0.3221,
+      "epoch": 0.34850614480844777,
+      "grad_norm": 0.19263622164726257,
+      "learning_rate": 4.979383393834111e-05,
+      "loss": 0.4851,
       "step": 9670
     },
     {
-      "epoch": 0.34,
-      "learning_rate": 4.981015700977486e-05,
-      "loss": 0.3095,
+      "epoch": 0.3486863444696724,
+      "grad_norm": 0.190067321062088,
+      "learning_rate": 4.9793459778100794e-05,
+      "loss": 0.4344,
       "step": 9675
     },
     {
-      "epoch": 0.34,
-      "learning_rate": 4.980980644701202e-05,
-      "loss": 0.3289,
+      "epoch": 0.348866544130897,
+      "grad_norm": 0.16212069988250732,
+      "learning_rate": 4.979308528005439e-05,
+      "loss": 0.4459,
       "step": 9680
     },
     {
-      "epoch": 0.34,
-      "learning_rate": 4.980945556211019e-05,
-      "loss": 0.2993,
+      "epoch": 0.3490467437921217,
+      "grad_norm": 0.20930610597133636,
+      "learning_rate": 4.9792710444207004e-05,
+      "loss": 0.4726,
       "step": 9685
     },
     {
-      "epoch": 0.34,
-      "learning_rate": 4.980910435507393e-05,
-      "loss": 0.3396,
+      "epoch": 0.3492269434533463,
+      "grad_norm": 0.20407001674175262,
+      "learning_rate": 4.979233527056374e-05,
+      "loss": 0.47,
       "step": 9690
     },
     {
-      "epoch": 0.34,
-      "learning_rate": 4.9808752825907795e-05,
-      "loss": 0.3295,
+      "epoch": 0.34940714311457094,
+      "grad_norm": 0.16390258073806763,
+      "learning_rate": 4.9791959759129706e-05,
+      "loss": 0.4299,
       "step": 9695
     },
     {
-      "epoch": 0.34,
-      "learning_rate": 4.980840097461635e-05,
-      "loss": 0.3072,
+      "epoch": 0.34958734277579556,
+      "grad_norm": 0.18050611019134521,
+      "learning_rate": 4.979158390991002e-05,
+      "loss": 0.4971,
       "step": 9700
     },
     {
-      "epoch": 0.34,
-      "learning_rate": 4.980804880120418e-05,
-      "loss": 0.3135,
+      "epoch": 0.34976754243702024,
+      "grad_norm": 0.172181636095047,
+      "learning_rate": 4.9791207722909794e-05,
+      "loss": 0.4547,
       "step": 9705
     },
     {
-      "epoch": 0.34,
-      "learning_rate": 4.9807696305675844e-05,
-      "loss": 0.335,
+      "epoch": 0.34994774209824486,
+      "grad_norm": 0.12145093083381653,
+      "learning_rate": 4.9790831198134175e-05,
+      "loss": 0.471,
       "step": 9710
     },
     {
-      "epoch": 0.34,
-      "learning_rate": 4.980734348803592e-05,
-      "loss": 0.3319,
+      "epoch": 0.3501279417594695,
+      "grad_norm": 0.17209351062774658,
+      "learning_rate": 4.979045433558828e-05,
+      "loss": 0.4392,
       "step": 9715
     },
     {
-      "epoch": 0.34,
-      "learning_rate": 4.980699034828898e-05,
-      "loss": 0.3338,
+      "epoch": 0.3503081414206941,
+      "grad_norm": 0.21002522110939026,
+      "learning_rate": 4.979007713527723e-05,
+      "loss": 0.4976,
       "step": 9720
     },
     {
-      "epoch": 0.34,
-      "learning_rate": 4.980663688643963e-05,
-      "loss": 0.3719,
+      "epoch": 0.3504883410819188,
+      "grad_norm": 0.16209115087985992,
+      "learning_rate": 4.9789699597206196e-05,
+      "loss": 0.4635,
       "step": 9725
     },
     {
-      "epoch": 0.34,
-      "learning_rate": 4.980628310249244e-05,
-      "loss": 0.3326,
+      "epoch": 0.3506685407431434,
+      "grad_norm": 0.14901262521743774,
+      "learning_rate": 4.97893217213803e-05,
+      "loss": 0.4632,
       "step": 9730
     },
     {
-      "epoch": 0.34,
-      "learning_rate": 4.980592899645201e-05,
-      "loss": 0.3386,
+      "epoch": 0.35084874040436803,
+      "grad_norm": 0.16062456369400024,
+      "learning_rate": 4.9788943507804686e-05,
+      "loss": 0.4482,
       "step": 9735
     },
     {
-      "epoch": 0.34,
-      "learning_rate": 4.980557456832295e-05,
-      "loss": 0.3241,
+      "epoch": 0.35102894006559265,
+      "grad_norm": 0.14877210557460785,
+      "learning_rate": 4.9788564956484527e-05,
+      "loss": 0.4823,
       "step": 9740
     },
     {
-      "epoch": 0.34,
-      "learning_rate": 4.980521981810985e-05,
-      "loss": 0.3372,
+      "epoch": 0.35120913972681733,
+      "grad_norm": 0.1947886049747467,
+      "learning_rate": 4.978818606742496e-05,
+      "loss": 0.4606,
       "step": 9745
     },
     {
-      "epoch": 0.34,
-      "learning_rate": 4.980486474581732e-05,
-      "loss": 0.3224,
+      "epoch": 0.35138933938804195,
+      "grad_norm": 0.18804532289505005,
+      "learning_rate": 4.978780684063116e-05,
+      "loss": 0.4878,
       "step": 9750
     },
     {
-      "epoch": 0.34,
-      "learning_rate": 4.980450935144996e-05,
-      "loss": 0.2923,
+      "epoch": 0.3515695390492666,
+      "grad_norm": 0.179021954536438,
+      "learning_rate": 4.97874272761083e-05,
+      "loss": 0.4619,
       "step": 9755
     },
     {
-      "epoch": 0.34,
-      "learning_rate": 4.980415363501239e-05,
-      "loss": 0.3383,
+      "epoch": 0.3517497387104912,
+      "grad_norm": 0.1584351509809494,
+      "learning_rate": 4.978704737386153e-05,
+      "loss": 0.4884,
       "step": 9760
     },
     {
-      "epoch": 0.34,
-      "learning_rate": 4.9803797596509235e-05,
-      "loss": 0.3293,
+      "epoch": 0.3519299383717159,
+      "grad_norm": 0.18059109151363373,
+      "learning_rate": 4.9786667133896046e-05,
+      "loss": 0.4296,
       "step": 9765
     },
     {
-      "epoch": 0.34,
-      "learning_rate": 4.980344123594512e-05,
-      "loss": 0.3373,
+      "epoch": 0.3521101380329405,
+      "grad_norm": 0.16413117945194244,
+      "learning_rate": 4.978628655621702e-05,
+      "loss": 0.4785,
       "step": 9770
     },
     {
-      "epoch": 0.34,
-      "learning_rate": 4.9803084553324666e-05,
-      "loss": 0.3358,
+      "epoch": 0.3522903376941651,
+      "grad_norm": 0.17627866566181183,
+      "learning_rate": 4.9785905640829635e-05,
+      "loss": 0.4539,
       "step": 9775
     },
     {
-      "epoch": 0.34,
-      "learning_rate": 4.98027275486525e-05,
-      "loss": 0.3051,
+      "epoch": 0.35247053735538975,
+      "grad_norm": 0.19507911801338196,
+      "learning_rate": 4.978552438773909e-05,
+      "loss": 0.4753,
       "step": 9780
     },
     {
-      "epoch": 0.34,
-      "learning_rate": 4.980237022193326e-05,
-      "loss": 0.3373,
+      "epoch": 0.3526507370166144,
+      "grad_norm": 0.23072102665901184,
+      "learning_rate": 4.9785142796950566e-05,
+      "loss": 0.4645,
       "step": 9785
     },
     {
-      "epoch": 0.34,
-      "learning_rate": 4.98020125731716e-05,
-      "loss": 0.3485,
+      "epoch": 0.35283093667783905,
+      "grad_norm": 0.16471746563911438,
+      "learning_rate": 4.978476086846928e-05,
+      "loss": 0.4756,
       "step": 9790
     },
     {
-      "epoch": 0.34,
-      "learning_rate": 4.980165460237214e-05,
-      "loss": 0.3667,
+      "epoch": 0.35301113633906367,
+      "grad_norm": 0.17672182619571686,
+      "learning_rate": 4.978437860230042e-05,
+      "loss": 0.459,
       "step": 9795
     },
     {
-      "epoch": 0.34,
-      "learning_rate": 4.980129630953955e-05,
-      "loss": 0.3064,
+      "epoch": 0.3531913360002883,
+      "grad_norm": 0.17026053369045258,
+      "learning_rate": 4.97839959984492e-05,
+      "loss": 0.4961,
       "step": 9800
     },
     {
-      "epoch": 0.34,
-      "learning_rate": 4.980093769467846e-05,
-      "loss": 0.326,
+      "epoch": 0.35337153566151297,
+      "grad_norm": 0.19481323659420013,
+      "learning_rate": 4.978361305692083e-05,
+      "loss": 0.4534,
       "step": 9805
     },
     {
-      "epoch": 0.35,
-      "learning_rate": 4.9800578757793546e-05,
-      "loss": 0.304,
+      "epoch": 0.3535517353227376,
+      "grad_norm": 0.17143213748931885,
+      "learning_rate": 4.978322977772053e-05,
+      "loss": 0.4726,
       "step": 9810
     },
     {
-      "epoch": 0.35,
-      "learning_rate": 4.980021949888946e-05,
-      "loss": 0.3163,
+      "epoch": 0.3537319349839622,
+      "grad_norm": 0.20486843585968018,
+      "learning_rate": 4.978284616085352e-05,
+      "loss": 0.5067,
       "step": 9815
     },
     {
-      "epoch": 0.35,
-      "learning_rate": 4.979985991797087e-05,
-      "loss": 0.3347,
+      "epoch": 0.3539121346451869,
+      "grad_norm": 0.15131095051765442,
+      "learning_rate": 4.9782462206325045e-05,
+      "loss": 0.4922,
       "step": 9820
     },
     {
-      "epoch": 0.35,
-      "learning_rate": 4.979950001504244e-05,
-      "loss": 0.3157,
+      "epoch": 0.3540923343064115,
+      "grad_norm": 0.20476940274238586,
+      "learning_rate": 4.978207791414031e-05,
+      "loss": 0.47,
       "step": 9825
     },
     {
-      "epoch": 0.35,
-      "learning_rate": 4.9799139790108846e-05,
-      "loss": 0.2742,
+      "epoch": 0.35427253396763614,
+      "grad_norm": 0.16714629530906677,
+      "learning_rate": 4.978169328430456e-05,
+      "loss": 0.454,
       "step": 9830
     },
     {
-      "epoch": 0.35,
-      "learning_rate": 4.979877924317477e-05,
-      "loss": 0.3241,
+      "epoch": 0.35445273362886076,
+      "grad_norm": 0.18224307894706726,
+      "learning_rate": 4.978130831682304e-05,
+      "loss": 0.4527,
       "step": 9835
     },
     {
-      "epoch": 0.35,
-      "learning_rate": 4.979841837424488e-05,
-      "loss": 0.3275,
+      "epoch": 0.35463293329008544,
+      "grad_norm": 0.1600581556558609,
+      "learning_rate": 4.978092301170099e-05,
+      "loss": 0.4877,
       "step": 9840
     },
     {
-      "epoch": 0.35,
-      "learning_rate": 4.979805718332388e-05,
-      "loss": 0.331,
+      "epoch": 0.35481313295131006,
+      "grad_norm": 0.21684806048870087,
+      "learning_rate": 4.9780537368943655e-05,
+      "loss": 0.4997,
       "step": 9845
     },
     {
-      "epoch": 0.35,
-      "learning_rate": 4.979769567041643e-05,
-      "loss": 0.327,
+      "epoch": 0.3549933326125347,
+      "grad_norm": 0.15424644947052002,
+      "learning_rate": 4.978015138855631e-05,
+      "loss": 0.4398,
       "step": 9850
     },
     {
-      "epoch": 0.35,
-      "learning_rate": 4.979733383552727e-05,
-      "loss": 0.3175,
+      "epoch": 0.3551735322737593,
+      "grad_norm": 0.16304361820220947,
+      "learning_rate": 4.9779765070544195e-05,
+      "loss": 0.4756,
       "step": 9855
     },
     {
-      "epoch": 0.35,
-      "learning_rate": 4.979697167866105e-05,
-      "loss": 0.318,
+      "epoch": 0.355353731934984,
+      "grad_norm": 0.1486636996269226,
+      "learning_rate": 4.977937841491257e-05,
+      "loss": 0.4152,
       "step": 9860
     },
     {
-      "epoch": 0.35,
-      "learning_rate": 4.979660919982251e-05,
-      "loss": 0.3552,
+      "epoch": 0.3555339315962086,
+      "grad_norm": 0.1545153707265854,
+      "learning_rate": 4.9778991421666724e-05,
+      "loss": 0.447,
       "step": 9865
     },
     {
-      "epoch": 0.35,
-      "learning_rate": 4.979624639901633e-05,
-      "loss": 0.316,
+      "epoch": 0.35571413125743323,
+      "grad_norm": 0.14595122635364532,
+      "learning_rate": 4.977860409081191e-05,
+      "loss": 0.4589,
       "step": 9870
     },
     {
-      "epoch": 0.35,
-      "learning_rate": 4.979588327624725e-05,
-      "loss": 0.3402,
+      "epoch": 0.35589433091865785,
+      "grad_norm": 0.17412835359573364,
+      "learning_rate": 4.977821642235341e-05,
+      "loss": 0.4367,
       "step": 9875
     },
     {
-      "epoch": 0.35,
-      "learning_rate": 4.979551983151995e-05,
-      "loss": 0.3271,
+      "epoch": 0.35607453057988253,
+      "grad_norm": 0.20736642181873322,
+      "learning_rate": 4.9777828416296513e-05,
+      "loss": 0.4574,
       "step": 9880
     },
     {
-      "epoch": 0.35,
-      "learning_rate": 4.9795156064839166e-05,
-      "loss": 0.3719,
+      "epoch": 0.35625473024110715,
+      "grad_norm": 0.20509152114391327,
+      "learning_rate": 4.9777440072646504e-05,
+      "loss": 0.4694,
       "step": 9885
     },
     {
-      "epoch": 0.35,
-      "learning_rate": 4.9794791976209634e-05,
-      "loss": 0.3154,
+      "epoch": 0.3564349299023318,
+      "grad_norm": 0.1846247762441635,
+      "learning_rate": 4.977705139140867e-05,
+      "loss": 0.4674,
       "step": 9890
     },
     {
-      "epoch": 0.35,
-      "learning_rate": 4.979442756563605e-05,
-      "loss": 0.3374,
+      "epoch": 0.3566151295635564,
+      "grad_norm": 0.2067510038614273,
+      "learning_rate": 4.97766623725883e-05,
+      "loss": 0.4276,
       "step": 9895
     },
     {
-      "epoch": 0.35,
-      "learning_rate": 4.9794062833123176e-05,
-      "loss": 0.328,
+      "epoch": 0.3567953292247811,
+      "grad_norm": 0.2130732536315918,
+      "learning_rate": 4.977627301619071e-05,
+      "loss": 0.4933,
       "step": 9900
     },
     {
-      "epoch": 0.35,
-      "learning_rate": 4.979369777867574e-05,
-      "loss": 0.3276,
+      "epoch": 0.3569755288860057,
+      "grad_norm": 0.2266158163547516,
+      "learning_rate": 4.97758833222212e-05,
+      "loss": 0.4332,
       "step": 9905
     },
     {
-      "epoch": 0.35,
-      "learning_rate": 4.979333240229847e-05,
-      "loss": 0.3205,
+      "epoch": 0.3571557285472303,
+      "grad_norm": 0.18190445005893707,
+      "learning_rate": 4.977549329068506e-05,
+      "loss": 0.4467,
       "step": 9910
     },
     {
-      "epoch": 0.35,
-      "learning_rate": 4.9792966703996116e-05,
-      "loss": 0.309,
+      "epoch": 0.35733592820845494,
+      "grad_norm": 0.14068225026130676,
+      "learning_rate": 4.977510292158764e-05,
+      "loss": 0.4436,
       "step": 9915
     },
     {
-      "epoch": 0.35,
-      "learning_rate": 4.979260068377344e-05,
-      "loss": 0.3277,
+      "epoch": 0.3575161278696796,
+      "grad_norm": 0.1720120906829834,
+      "learning_rate": 4.977471221493423e-05,
+      "loss": 0.4622,
       "step": 9920
     },
     {
-      "epoch": 0.35,
-      "learning_rate": 4.979223434163517e-05,
-      "loss": 0.3036,
+      "epoch": 0.35769632753090425,
+      "grad_norm": 0.1588047444820404,
+      "learning_rate": 4.977432117073016e-05,
+      "loss": 0.4633,
       "step": 9925
     },
     {
-      "epoch": 0.35,
-      "learning_rate": 4.979186767758608e-05,
-      "loss": 0.3299,
+      "epoch": 0.35787652719212887,
+      "grad_norm": 0.168239563703537,
+      "learning_rate": 4.9773929788980766e-05,
+      "loss": 0.4476,
       "step": 9930
     },
     {
-      "epoch": 0.35,
-      "learning_rate": 4.9791500691630936e-05,
-      "loss": 0.335,
+      "epoch": 0.3580567268533535,
+      "grad_norm": 0.16691339015960693,
+      "learning_rate": 4.9773538069691375e-05,
+      "loss": 0.4693,
       "step": 9935
     },
     {
-      "epoch": 0.35,
-      "learning_rate": 4.979113338377448e-05,
-      "loss": 0.3341,
+      "epoch": 0.35823692651457817,
+      "grad_norm": 0.15054184198379517,
+      "learning_rate": 4.977314601286732e-05,
+      "loss": 0.4347,
       "step": 9940
     },
     {
-      "epoch": 0.35,
-      "learning_rate": 4.979076575402151e-05,
-      "loss": 0.3399,
+      "epoch": 0.3584171261758028,
+      "grad_norm": 0.147248312830925,
+      "learning_rate": 4.9772753618513945e-05,
+      "loss": 0.444,
       "step": 9945
     },
     {
-      "epoch": 0.35,
-      "learning_rate": 4.979039780237677e-05,
-      "loss": 0.3285,
+      "epoch": 0.3585973258370274,
+      "grad_norm": 0.20914539694786072,
+      "learning_rate": 4.9772360886636605e-05,
+      "loss": 0.4573,
       "step": 9950
     },
     {
-      "epoch": 0.35,
-      "learning_rate": 4.979002952884506e-05,
-      "loss": 0.3308,
+      "epoch": 0.35877752549825204,
+      "grad_norm": 0.16650797426700592,
+      "learning_rate": 4.977196781724064e-05,
+      "loss": 0.4585,
       "step": 9955
     },
     {
-      "epoch": 0.35,
-      "learning_rate": 4.978966093343115e-05,
-      "loss": 0.3288,
+      "epoch": 0.3589577251594767,
+      "grad_norm": 0.2121131271123886,
+      "learning_rate": 4.9771574410331415e-05,
+      "loss": 0.4204,
       "step": 9960
     },
     {
-      "epoch": 0.35,
-      "learning_rate": 4.978929201613983e-05,
-      "loss": 0.341,
+      "epoch": 0.35913792482070134,
+      "grad_norm": 0.1681807041168213,
+      "learning_rate": 4.977118066591427e-05,
+      "loss": 0.5008,
       "step": 9965
     },
     {
-      "epoch": 0.35,
-      "learning_rate": 4.978892277697589e-05,
-      "loss": 0.3037,
+      "epoch": 0.35931812448192596,
+      "grad_norm": 0.18898646533489227,
+      "learning_rate": 4.97707865839946e-05,
+      "loss": 0.5016,
       "step": 9970
     },
     {
-      "epoch": 0.35,
-      "learning_rate": 4.978855321594413e-05,
-      "loss": 0.315,
+      "epoch": 0.35949832414315064,
+      "grad_norm": 0.1509980857372284,
+      "learning_rate": 4.977039216457775e-05,
+      "loss": 0.4251,
       "step": 9975
     },
     {
-      "epoch": 0.35,
-      "learning_rate": 4.978818333304933e-05,
-      "loss": 0.3041,
+      "epoch": 0.35967852380437526,
+      "grad_norm": 0.15001198649406433,
+      "learning_rate": 4.976999740766911e-05,
+      "loss": 0.4745,
       "step": 9980
     },
     {
-      "epoch": 0.35,
-      "learning_rate": 4.978781312829633e-05,
-      "loss": 0.32,
+      "epoch": 0.3598587234655999,
+      "grad_norm": 0.19364404678344727,
+      "learning_rate": 4.976960231327404e-05,
+      "loss": 0.4624,
       "step": 9985
     },
     {
-      "epoch": 0.35,
-      "learning_rate": 4.978744260168988e-05,
-      "loss": 0.3322,
+      "epoch": 0.3600389231268245,
+      "grad_norm": 0.1906237006187439,
+      "learning_rate": 4.976920688139794e-05,
+      "loss": 0.4492,
       "step": 9990
     },
     {
-      "epoch": 0.35,
-      "learning_rate": 4.978707175323485e-05,
-      "loss": 0.336,
+      "epoch": 0.3602191227880492,
+      "grad_norm": 0.1953149288892746,
+      "learning_rate": 4.9768811112046196e-05,
+      "loss": 0.4655,
       "step": 9995
     },
     {
-      "epoch": 0.35,
-      "learning_rate": 4.978670058293601e-05,
-      "loss": 0.2962,
+      "epoch": 0.3603993224492738,
+      "grad_norm": 0.1708211600780487,
+      "learning_rate": 4.976841500522419e-05,
+      "loss": 0.4667,
       "step": 10000
     },
     {
-      "epoch": 0.35,
-      "eval_loss": 0.3222596049308777,
-      "eval_runtime": 10.5408,
-      "eval_samples_per_second": 9.487,
-      "eval_steps_per_second": 9.487,
+      "epoch": 0.3603993224492738,
+      "eval_loss": 0.4796702265739441,
+      "eval_runtime": 3.5164,
+      "eval_samples_per_second": 28.438,
+      "eval_steps_per_second": 7.11,
       "step": 10000
     },
     {
-      "epoch": 0.35,
-      "learning_rate": 4.978632909079821e-05,
-      "loss": 0.3262,
+      "epoch": 0.36057952211049843,
+      "grad_norm": 0.20447692275047302,
+      "learning_rate": 4.976801856093732e-05,
+      "loss": 0.4598,
       "step": 10005
     },
     {
-      "epoch": 0.35,
-      "learning_rate": 4.9785957276826254e-05,
-      "loss": 0.3352,
+      "epoch": 0.36075972177172305,
+      "grad_norm": 0.16265089809894562,
+      "learning_rate": 4.976762177919101e-05,
+      "loss": 0.4668,
       "step": 10010
     },
     {
-      "epoch": 0.35,
-      "learning_rate": 4.978558514102498e-05,
-      "loss": 0.3371,
+      "epoch": 0.36093992143294773,
+      "grad_norm": 0.1456652581691742,
+      "learning_rate": 4.976722465999063e-05,
+      "loss": 0.4301,
       "step": 10015
     },
     {
-      "epoch": 0.35,
-      "learning_rate": 4.978521268339922e-05,
-      "loss": 0.3208,
+      "epoch": 0.36112012109417235,
+      "grad_norm": 0.17092099785804749,
+      "learning_rate": 4.976682720334162e-05,
+      "loss": 0.4681,
       "step": 10020
     },
     {
-      "epoch": 0.35,
-      "learning_rate": 4.97848399039538e-05,
-      "loss": 0.3262,
+      "epoch": 0.361300320755397,
+      "grad_norm": 0.1890965700149536,
+      "learning_rate": 4.976642940924938e-05,
+      "loss": 0.477,
       "step": 10025
     },
     {
-      "epoch": 0.35,
-      "learning_rate": 4.978446680269358e-05,
-      "loss": 0.3099,
+      "epoch": 0.3614805204166216,
+      "grad_norm": 0.16893772780895233,
+      "learning_rate": 4.976603127771934e-05,
+      "loss": 0.4796,
       "step": 10030
     },
     {
-      "epoch": 0.35,
-      "learning_rate": 4.9784093379623386e-05,
-      "loss": 0.3112,
+      "epoch": 0.3616607200778463,
+      "grad_norm": 0.16961359977722168,
+      "learning_rate": 4.9765632808756915e-05,
+      "loss": 0.4314,
       "step": 10035
     },
     {
-      "epoch": 0.35,
-      "learning_rate": 4.9783719634748074e-05,
-      "loss": 0.3207,
+      "epoch": 0.3618409197390709,
+      "grad_norm": 0.1630767434835434,
+      "learning_rate": 4.9765234002367534e-05,
+      "loss": 0.4594,
       "step": 10040
     },
     {
-      "epoch": 0.35,
-      "learning_rate": 4.978334556807249e-05,
-      "loss": 0.3204,
+      "epoch": 0.3620211194002955,
+      "grad_norm": 0.16475652158260345,
+      "learning_rate": 4.9764834858556635e-05,
+      "loss": 0.4489,
       "step": 10045
     },
     {
-      "epoch": 0.35,
-      "learning_rate": 4.97829711796015e-05,
-      "loss": 0.327,
+      "epoch": 0.36220131906152014,
+      "grad_norm": 0.1670771986246109,
+      "learning_rate": 4.9764435377329654e-05,
+      "loss": 0.4278,
       "step": 10050
     },
     {
-      "epoch": 0.35,
-      "learning_rate": 4.978259646933996e-05,
-      "loss": 0.34,
+      "epoch": 0.3623815187227448,
+      "grad_norm": 0.1597401350736618,
+      "learning_rate": 4.9764035558692045e-05,
+      "loss": 0.4035,
       "step": 10055
     },
     {
-      "epoch": 0.35,
-      "learning_rate": 4.978222143729274e-05,
-      "loss": 0.338,
+      "epoch": 0.36256171838396944,
+      "grad_norm": 0.2012094110250473,
+      "learning_rate": 4.9763635402649236e-05,
+      "loss": 0.4649,
       "step": 10060
     },
     {
-      "epoch": 0.35,
-      "learning_rate": 4.978184608346471e-05,
-      "loss": 0.3158,
+      "epoch": 0.36274191804519407,
+      "grad_norm": 0.16659463942050934,
+      "learning_rate": 4.9763234909206695e-05,
+      "loss": 0.447,
       "step": 10065
     },
     {
-      "epoch": 0.35,
-      "learning_rate": 4.978147040786073e-05,
-      "loss": 0.3068,
+      "epoch": 0.3629221177064187,
+      "grad_norm": 0.15925155580043793,
+      "learning_rate": 4.976283407836987e-05,
+      "loss": 0.4575,
       "step": 10070
     },
     {
-      "epoch": 0.35,
-      "learning_rate": 4.97810944104857e-05,
-      "loss": 0.3244,
+      "epoch": 0.36310231736764337,
+      "grad_norm": 0.15115337073802948,
+      "learning_rate": 4.976243291014423e-05,
+      "loss": 0.4517,
       "step": 10075
     },
     {
-      "epoch": 0.35,
-      "learning_rate": 4.978071809134448e-05,
-      "loss": 0.3069,
+      "epoch": 0.363282517028868,
+      "grad_norm": 0.1577247679233551,
+      "learning_rate": 4.976203140453523e-05,
+      "loss": 0.4539,
       "step": 10080
     },
     {
-      "epoch": 0.35,
-      "learning_rate": 4.978034145044198e-05,
-      "loss": 0.3123,
+      "epoch": 0.3634627166900926,
+      "grad_norm": 0.16622968018054962,
+      "learning_rate": 4.9761629561548354e-05,
+      "loss": 0.4357,
       "step": 10085
     },
     {
-      "epoch": 0.35,
-      "learning_rate": 4.977996448778307e-05,
-      "loss": 0.3258,
+      "epoch": 0.36364291635131724,
+      "grad_norm": 0.1571248173713684,
+      "learning_rate": 4.976122738118906e-05,
+      "loss": 0.4368,
       "step": 10090
     },
     {
-      "epoch": 0.36,
-      "learning_rate": 4.977958720337265e-05,
-      "loss": 0.3306,
+      "epoch": 0.3638231160125419,
+      "grad_norm": 0.17340044677257538,
+      "learning_rate": 4.976082486346284e-05,
+      "loss": 0.4284,
       "step": 10095
     },
     {
-      "epoch": 0.36,
-      "learning_rate": 4.977920959721561e-05,
-      "loss": 0.3117,
+      "epoch": 0.36400331567376654,
+      "grad_norm": 0.20559607446193695,
+      "learning_rate": 4.9760422008375176e-05,
+      "loss": 0.4247,
       "step": 10100
     },
     {
-      "epoch": 0.36,
-      "learning_rate": 4.977883166931688e-05,
-      "loss": 0.2943,
+      "epoch": 0.36418351533499116,
+      "grad_norm": 0.159358948469162,
+      "learning_rate": 4.976001881593155e-05,
+      "loss": 0.4389,
       "step": 10105
     },
     {
-      "epoch": 0.36,
-      "learning_rate": 4.977845341968135e-05,
-      "loss": 0.3441,
+      "epoch": 0.3643637149962158,
+      "grad_norm": 0.19919893145561218,
+      "learning_rate": 4.9759615286137465e-05,
+      "loss": 0.4724,
       "step": 10110
     },
     {
-      "epoch": 0.36,
-      "learning_rate": 4.9778074848313925e-05,
-      "loss": 0.3405,
+      "epoch": 0.36454391465744046,
+      "grad_norm": 0.1773001104593277,
+      "learning_rate": 4.975921141899842e-05,
+      "loss": 0.4847,
       "step": 10115
     },
     {
-      "epoch": 0.36,
-      "learning_rate": 4.977769595521953e-05,
-      "loss": 0.3054,
+      "epoch": 0.3647241143186651,
+      "grad_norm": 0.17267437279224396,
+      "learning_rate": 4.975880721451991e-05,
+      "loss": 0.4398,
       "step": 10120
     },
     {
-      "epoch": 0.36,
-      "learning_rate": 4.977731674040308e-05,
-      "loss": 0.3339,
+      "epoch": 0.3649043139798897,
+      "grad_norm": 0.19149023294448853,
+      "learning_rate": 4.975840267270744e-05,
+      "loss": 0.4703,
       "step": 10125
     },
     {
-      "epoch": 0.36,
-      "learning_rate": 4.97769372038695e-05,
-      "loss": 0.3445,
+      "epoch": 0.3650845136411143,
+      "grad_norm": 0.1706707626581192,
+      "learning_rate": 4.975799779356653e-05,
+      "loss": 0.5034,
       "step": 10130
     },
     {
-      "epoch": 0.36,
-      "learning_rate": 4.977655734562373e-05,
-      "loss": 0.3327,
+      "epoch": 0.365264713302339,
+      "grad_norm": 0.18125681579113007,
+      "learning_rate": 4.97575925771027e-05,
+      "loss": 0.473,
       "step": 10135
     },
     {
-      "epoch": 0.36,
-      "learning_rate": 4.9776177165670676e-05,
-      "loss": 0.3147,
+      "epoch": 0.36544491296356363,
+      "grad_norm": 0.17481379210948944,
+      "learning_rate": 4.975718702332146e-05,
+      "loss": 0.4341,
       "step": 10140
     },
     {
-      "epoch": 0.36,
-      "learning_rate": 4.97757966640153e-05,
-      "loss": 0.3163,
+      "epoch": 0.36562511262478825,
+      "grad_norm": 0.19114582240581512,
+      "learning_rate": 4.9756781132228334e-05,
+      "loss": 0.4629,
       "step": 10145
     },
     {
-      "epoch": 0.36,
-      "learning_rate": 4.977541584066253e-05,
-      "loss": 0.3185,
+      "epoch": 0.36580531228601293,
+      "grad_norm": 0.1782168298959732,
+      "learning_rate": 4.975637490382887e-05,
+      "loss": 0.4375,
       "step": 10150
     },
     {
-      "epoch": 0.36,
-      "learning_rate": 4.9775034695617314e-05,
-      "loss": 0.3235,
+      "epoch": 0.36598551194723755,
+      "grad_norm": 0.21205340325832367,
+      "learning_rate": 4.975596833812858e-05,
+      "loss": 0.4514,
       "step": 10155
     },
     {
-      "epoch": 0.36,
-      "learning_rate": 4.9774653228884594e-05,
-      "loss": 0.3276,
+      "epoch": 0.3661657116084622,
+      "grad_norm": 0.19303537905216217,
+      "learning_rate": 4.9755561435133024e-05,
+      "loss": 0.4596,
       "step": 10160
     },
     {
-      "epoch": 0.36,
-      "learning_rate": 4.977427144046933e-05,
-      "loss": 0.3118,
+      "epoch": 0.3663459112696868,
+      "grad_norm": 0.16598139703273773,
+      "learning_rate": 4.9755154194847734e-05,
+      "loss": 0.4418,
       "step": 10165
     },
     {
-      "epoch": 0.36,
-      "learning_rate": 4.977388933037649e-05,
-      "loss": 0.3274,
+      "epoch": 0.3665261109309115,
+      "grad_norm": 0.19510379433631897,
+      "learning_rate": 4.9754746617278254e-05,
+      "loss": 0.4699,
       "step": 10170
     },
     {
-      "epoch": 0.36,
-      "learning_rate": 4.977350689861101e-05,
-      "loss": 0.3286,
+      "epoch": 0.3667063105921361,
+      "grad_norm": 0.16283252835273743,
+      "learning_rate": 4.975433870243015e-05,
+      "loss": 0.4337,
       "step": 10175
     },
     {
-      "epoch": 0.36,
-      "learning_rate": 4.9773124145177875e-05,
-      "loss": 0.308,
+      "epoch": 0.3668865102533607,
+      "grad_norm": 0.20761866867542267,
+      "learning_rate": 4.975393045030897e-05,
+      "loss": 0.4831,
       "step": 10180
     },
     {
-      "epoch": 0.36,
-      "learning_rate": 4.977274107008205e-05,
-      "loss": 0.3173,
+      "epoch": 0.36706670991458534,
+      "grad_norm": 0.15632273256778717,
+      "learning_rate": 4.9753521860920284e-05,
+      "loss": 0.4706,
       "step": 10185
     },
     {
-      "epoch": 0.36,
-      "learning_rate": 4.9772357673328506e-05,
-      "loss": 0.3216,
+      "epoch": 0.36724690957581,
+      "grad_norm": 0.19553816318511963,
+      "learning_rate": 4.975311293426965e-05,
+      "loss": 0.4547,
       "step": 10190
     },
     {
-      "epoch": 0.36,
-      "learning_rate": 4.977197395492223e-05,
-      "loss": 0.3111,
+      "epoch": 0.36742710923703464,
+      "grad_norm": 0.15499721467494965,
+      "learning_rate": 4.975270367036264e-05,
+      "loss": 0.4329,
       "step": 10195
     },
     {
-      "epoch": 0.36,
-      "learning_rate": 4.977158991486818e-05,
-      "loss": 0.315,
+      "epoch": 0.36760730889825927,
+      "grad_norm": 0.19358859956264496,
+      "learning_rate": 4.975229406920485e-05,
+      "loss": 0.4449,
       "step": 10200
     },
     {
-      "epoch": 0.36,
-      "learning_rate": 4.9771205553171376e-05,
-      "loss": 0.283,
+      "epoch": 0.3677875085594839,
+      "grad_norm": 0.19693566858768463,
+      "learning_rate": 4.975188413080184e-05,
+      "loss": 0.4487,
       "step": 10205
     },
     {
-      "epoch": 0.36,
-      "learning_rate": 4.9770820869836784e-05,
-      "loss": 0.3386,
+      "epoch": 0.36796770822070857,
+      "grad_norm": 0.16378581523895264,
+      "learning_rate": 4.97514738551592e-05,
+      "loss": 0.4419,
       "step": 10210
     },
     {
-      "epoch": 0.36,
-      "learning_rate": 4.977043586486941e-05,
-      "loss": 0.3396,
+      "epoch": 0.3681479078819332,
+      "grad_norm": 0.22160831093788147,
+      "learning_rate": 4.975106324228252e-05,
+      "loss": 0.4503,
       "step": 10215
     },
     {
-      "epoch": 0.36,
-      "learning_rate": 4.977005053827425e-05,
-      "loss": 0.3357,
+      "epoch": 0.3683281075431578,
+      "grad_norm": 0.18501290678977966,
+      "learning_rate": 4.975065229217739e-05,
+      "loss": 0.4692,
       "step": 10220
     },
     {
-      "epoch": 0.36,
-      "learning_rate": 4.976966489005631e-05,
-      "loss": 0.3087,
+      "epoch": 0.36850830720438243,
+      "grad_norm": 0.17586058378219604,
+      "learning_rate": 4.9750241004849415e-05,
+      "loss": 0.4593,
       "step": 10225
     },
     {
-      "epoch": 0.36,
-      "learning_rate": 4.9769278920220596e-05,
-      "loss": 0.3442,
+      "epoch": 0.3686885068656071,
+      "grad_norm": 0.17529945075511932,
+      "learning_rate": 4.974982938030421e-05,
+      "loss": 0.4924,
       "step": 10230
     },
     {
-      "epoch": 0.36,
-      "learning_rate": 4.9768892628772115e-05,
-      "loss": 0.343,
+      "epoch": 0.36886870652683174,
+      "grad_norm": 0.22143632173538208,
+      "learning_rate": 4.974941741854736e-05,
+      "loss": 0.4816,
       "step": 10235
     },
     {
-      "epoch": 0.36,
-      "learning_rate": 4.9768506015715896e-05,
-      "loss": 0.3256,
+      "epoch": 0.36904890618805636,
+      "grad_norm": 0.16022737324237823,
+      "learning_rate": 4.974900511958449e-05,
+      "loss": 0.469,
       "step": 10240
     },
     {
-      "epoch": 0.36,
-      "learning_rate": 4.976811908105694e-05,
-      "loss": 0.3094,
+      "epoch": 0.369229105849281,
+      "grad_norm": 0.1857171654701233,
+      "learning_rate": 4.974859248342122e-05,
+      "loss": 0.4561,
       "step": 10245
     },
     {
-      "epoch": 0.36,
-      "learning_rate": 4.9767731824800277e-05,
-      "loss": 0.3192,
+      "epoch": 0.36940930551050566,
+      "grad_norm": 0.15506871044635773,
+      "learning_rate": 4.974817951006318e-05,
+      "loss": 0.484,
       "step": 10250
     },
     {
-      "epoch": 0.36,
-      "learning_rate": 4.976734424695095e-05,
-      "loss": 0.3161,
+      "epoch": 0.3695895051717303,
+      "grad_norm": 0.16695240139961243,
+      "learning_rate": 4.9747766199515967e-05,
+      "loss": 0.4345,
       "step": 10255
     },
     {
-      "epoch": 0.36,
-      "learning_rate": 4.976695634751397e-05,
-      "loss": 0.3199,
+      "epoch": 0.3697697048329549,
+      "grad_norm": 0.1858297437429428,
+      "learning_rate": 4.9747352551785234e-05,
+      "loss": 0.4694,
       "step": 10260
     },
     {
-      "epoch": 0.36,
-      "learning_rate": 4.9766568126494394e-05,
-      "loss": 0.2714,
+      "epoch": 0.3699499044941795,
+      "grad_norm": 0.18636928498744965,
+      "learning_rate": 4.9746938566876624e-05,
+      "loss": 0.4836,
       "step": 10265
     },
     {
-      "epoch": 0.36,
-      "learning_rate": 4.976617958389724e-05,
-      "loss": 0.2952,
+      "epoch": 0.3701301041554042,
+      "grad_norm": 0.14627771079540253,
+      "learning_rate": 4.9746524244795755e-05,
+      "loss": 0.4755,
       "step": 10270
     },
     {
-      "epoch": 0.36,
-      "learning_rate": 4.976579071972758e-05,
-      "loss": 0.3468,
+      "epoch": 0.3703103038166288,
+      "grad_norm": 0.2531489431858063,
+      "learning_rate": 4.974610958554829e-05,
+      "loss": 0.4655,
       "step": 10275
     },
     {
-      "epoch": 0.36,
-      "learning_rate": 4.976540153399044e-05,
-      "loss": 0.3307,
+      "epoch": 0.37049050347785345,
+      "grad_norm": 0.14432822167873383,
+      "learning_rate": 4.974569458913988e-05,
+      "loss": 0.4514,
       "step": 10280
     },
     {
-      "epoch": 0.36,
-      "learning_rate": 4.976501202669089e-05,
-      "loss": 0.3136,
+      "epoch": 0.3706707031390781,
+      "grad_norm": 0.200893372297287,
+      "learning_rate": 4.974527925557616e-05,
+      "loss": 0.4709,
       "step": 10285
     },
     {
-      "epoch": 0.36,
-      "learning_rate": 4.976462219783397e-05,
-      "loss": 0.3317,
+      "epoch": 0.37085090280030275,
+      "grad_norm": 0.1737281084060669,
+      "learning_rate": 4.974486358486281e-05,
+      "loss": 0.4521,
       "step": 10290
     },
     {
-      "epoch": 0.36,
-      "learning_rate": 4.976423204742475e-05,
-      "loss": 0.2882,
+      "epoch": 0.3710311024615274,
+      "grad_norm": 0.18836097419261932,
+      "learning_rate": 4.9744447577005484e-05,
+      "loss": 0.4333,
       "step": 10295
     },
     {
-      "epoch": 0.36,
-      "learning_rate": 4.976384157546831e-05,
-      "loss": 0.3635,
+      "epoch": 0.371211302122752,
+      "grad_norm": 0.16161343455314636,
+      "learning_rate": 4.974403123200984e-05,
+      "loss": 0.4779,
       "step": 10300
     },
     {
-      "epoch": 0.36,
-      "learning_rate": 4.97634507819697e-05,
-      "loss": 0.307,
+      "epoch": 0.3713915017839766,
+      "grad_norm": 0.1600758135318756,
+      "learning_rate": 4.9743614549881566e-05,
+      "loss": 0.4377,
       "step": 10305
     },
     {
-      "epoch": 0.36,
-      "learning_rate": 4.9763059666933996e-05,
-      "loss": 0.2976,
+      "epoch": 0.3715717014452013,
+      "grad_norm": 0.2167365998029709,
+      "learning_rate": 4.974319753062634e-05,
+      "loss": 0.4531,
       "step": 10310
     },
     {
-      "epoch": 0.36,
-      "learning_rate": 4.9762668230366286e-05,
-      "loss": 0.3068,
+      "epoch": 0.3717519011064259,
+      "grad_norm": 0.20788373053073883,
+      "learning_rate": 4.9742780174249835e-05,
+      "loss": 0.4682,
       "step": 10315
     },
     {
-      "epoch": 0.36,
-      "learning_rate": 4.9762276472271653e-05,
-      "loss": 0.3155,
+      "epoch": 0.37193210076765054,
+      "grad_norm": 0.16195015609264374,
+      "learning_rate": 4.974236248075774e-05,
+      "loss": 0.4617,
       "step": 10320
     },
     {
-      "epoch": 0.36,
-      "learning_rate": 4.976188439265517e-05,
-      "loss": 0.3258,
+      "epoch": 0.3721123004288752,
+      "grad_norm": 0.21307627856731415,
+      "learning_rate": 4.974194445015574e-05,
+      "loss": 0.474,
       "step": 10325
     },
     {
-      "epoch": 0.36,
-      "learning_rate": 4.9761491991521945e-05,
-      "loss": 0.3216,
+      "epoch": 0.37229250009009984,
+      "grad_norm": 0.3052423894405365,
+      "learning_rate": 4.974152608244955e-05,
+      "loss": 0.4861,
       "step": 10330
     },
     {
-      "epoch": 0.36,
-      "learning_rate": 4.976109926887706e-05,
-      "loss": 0.3007,
+      "epoch": 0.37247269975132447,
+      "grad_norm": 0.16430044174194336,
+      "learning_rate": 4.9741107377644845e-05,
+      "loss": 0.4265,
       "step": 10335
     },
     {
-      "epoch": 0.36,
-      "learning_rate": 4.9760706224725616e-05,
-      "loss": 0.3354,
+      "epoch": 0.3726528994125491,
+      "grad_norm": 0.19519920647144318,
+      "learning_rate": 4.974068833574736e-05,
+      "loss": 0.447,
       "step": 10340
     },
     {
-      "epoch": 0.36,
-      "learning_rate": 4.9760312859072726e-05,
-      "loss": 0.3335,
+      "epoch": 0.37283309907377377,
+      "grad_norm": 0.22070425748825073,
+      "learning_rate": 4.974026895676277e-05,
+      "loss": 0.51,
       "step": 10345
     },
     {
-      "epoch": 0.36,
-      "learning_rate": 4.975991917192349e-05,
-      "loss": 0.3095,
+      "epoch": 0.3730132987349984,
+      "grad_norm": 0.17484411597251892,
+      "learning_rate": 4.973984924069681e-05,
+      "loss": 0.4552,
       "step": 10350
     },
     {
-      "epoch": 0.36,
-      "learning_rate": 4.975952516328302e-05,
-      "loss": 0.3143,
+      "epoch": 0.373193498396223,
+      "grad_norm": 0.15123674273490906,
+      "learning_rate": 4.9739429187555185e-05,
+      "loss": 0.4425,
       "step": 10355
     },
     {
-      "epoch": 0.36,
-      "learning_rate": 4.975913083315643e-05,
-      "loss": 0.3287,
+      "epoch": 0.37337369805744763,
+      "grad_norm": 0.17658279836177826,
+      "learning_rate": 4.973900879734364e-05,
+      "loss": 0.4305,
       "step": 10360
     },
     {
-      "epoch": 0.36,
-      "learning_rate": 4.975873618154885e-05,
-      "loss": 0.3102,
+      "epoch": 0.3735538977186723,
+      "grad_norm": 0.1411803662776947,
+      "learning_rate": 4.973858807006788e-05,
+      "loss": 0.4365,
       "step": 10365
     },
     {
-      "epoch": 0.36,
-      "learning_rate": 4.975834120846539e-05,
-      "loss": 0.3033,
+      "epoch": 0.37373409737989693,
+      "grad_norm": 0.1503876894712448,
+      "learning_rate": 4.973816700573366e-05,
+      "loss": 0.4826,
       "step": 10370
     },
     {
-      "epoch": 0.37,
-      "learning_rate": 4.975794591391119e-05,
-      "loss": 0.3203,
+      "epoch": 0.37391429704112156,
+      "grad_norm": 0.14822295308113098,
+      "learning_rate": 4.97377456043467e-05,
+      "loss": 0.4837,
       "step": 10375
     },
     {
-      "epoch": 0.37,
-      "learning_rate": 4.975755029789138e-05,
-      "loss": 0.342,
+      "epoch": 0.3740944967023462,
+      "grad_norm": 0.1543811559677124,
+      "learning_rate": 4.9737323865912734e-05,
+      "loss": 0.4731,
       "step": 10380
     },
     {
-      "epoch": 0.37,
-      "learning_rate": 4.9757154360411095e-05,
-      "loss": 0.2972,
+      "epoch": 0.37427469636357086,
+      "grad_norm": 0.18722733855247498,
+      "learning_rate": 4.973690179043753e-05,
+      "loss": 0.4697,
       "step": 10385
     },
     {
-      "epoch": 0.37,
-      "learning_rate": 4.975675810147547e-05,
-      "loss": 0.3124,
+      "epoch": 0.3744548960247955,
+      "grad_norm": 0.20646195113658905,
+      "learning_rate": 4.9736479377926826e-05,
+      "loss": 0.4538,
       "step": 10390
     },
     {
-      "epoch": 0.37,
-      "learning_rate": 4.975636152108966e-05,
-      "loss": 0.3097,
+      "epoch": 0.3746350956860201,
+      "grad_norm": 0.19029046595096588,
+      "learning_rate": 4.9736056628386374e-05,
+      "loss": 0.4714,
       "step": 10395
     },
     {
-      "epoch": 0.37,
-      "learning_rate": 4.975596461925881e-05,
-      "loss": 0.3229,
+      "epoch": 0.3748152953472447,
+      "grad_norm": 0.18136341869831085,
+      "learning_rate": 4.973563354182195e-05,
+      "loss": 0.4703,
       "step": 10400
     },
     {
-      "epoch": 0.37,
-      "learning_rate": 4.975556739598808e-05,
-      "loss": 0.3471,
+      "epoch": 0.3749954950084694,
+      "grad_norm": 0.20808909833431244,
+      "learning_rate": 4.97352101182393e-05,
+      "loss": 0.5067,
       "step": 10405
     },
     {
-      "epoch": 0.37,
-      "learning_rate": 4.975516985128261e-05,
-      "loss": 0.3405,
+      "epoch": 0.375175694669694,
+      "grad_norm": 0.18707624077796936,
+      "learning_rate": 4.9734786357644204e-05,
+      "loss": 0.4373,
       "step": 10410
     },
     {
-      "epoch": 0.37,
-      "learning_rate": 4.9754771985147586e-05,
-      "loss": 0.316,
+      "epoch": 0.37535589433091865,
+      "grad_norm": 0.19006437063217163,
+      "learning_rate": 4.9734362260042434e-05,
+      "loss": 0.4335,
       "step": 10415
     },
     {
-      "epoch": 0.37,
-      "learning_rate": 4.975437379758816e-05,
-      "loss": 0.3192,
+      "epoch": 0.37553609399214327,
+      "grad_norm": 0.22536343336105347,
+      "learning_rate": 4.973393782543976e-05,
+      "loss": 0.4656,
       "step": 10420
     },
     {
-      "epoch": 0.37,
-      "learning_rate": 4.97539752886095e-05,
-      "loss": 0.305,
+      "epoch": 0.37571629365336795,
+      "grad_norm": 0.20031480491161346,
+      "learning_rate": 4.9733513053841984e-05,
+      "loss": 0.4998,
       "step": 10425
     },
     {
-      "epoch": 0.37,
-      "learning_rate": 4.975357645821678e-05,
-      "loss": 0.3185,
+      "epoch": 0.3758964933145926,
+      "grad_norm": 0.17395085096359253,
+      "learning_rate": 4.973308794525487e-05,
+      "loss": 0.4831,
       "step": 10430
     },
     {
-      "epoch": 0.37,
-      "learning_rate": 4.9753177306415185e-05,
-      "loss": 0.3341,
+      "epoch": 0.3760766929758172,
+      "grad_norm": 0.20083628594875336,
+      "learning_rate": 4.973266249968423e-05,
+      "loss": 0.4584,
       "step": 10435
     },
     {
-      "epoch": 0.37,
-      "learning_rate": 4.9752777833209904e-05,
-      "loss": 0.3518,
+      "epoch": 0.3762568926370418,
+      "grad_norm": 0.1963411122560501,
+      "learning_rate": 4.973223671713585e-05,
+      "loss": 0.4865,
       "step": 10440
     },
     {
-      "epoch": 0.37,
-      "learning_rate": 4.975237803860611e-05,
-      "loss": 0.3262,
+      "epoch": 0.3764370922982665,
+      "grad_norm": 0.20614828169345856,
+      "learning_rate": 4.973181059761552e-05,
+      "loss": 0.4628,
       "step": 10445
     },
     {
-      "epoch": 0.37,
-      "learning_rate": 4.9751977922609e-05,
-      "loss": 0.3294,
+      "epoch": 0.3766172919594911,
+      "grad_norm": 0.2048252522945404,
+      "learning_rate": 4.973138414112908e-05,
+      "loss": 0.4685,
       "step": 10450
     },
     {
-      "epoch": 0.37,
-      "learning_rate": 4.975157748522376e-05,
-      "loss": 0.3283,
+      "epoch": 0.37679749162071574,
+      "grad_norm": 0.1850869357585907,
+      "learning_rate": 4.97309573476823e-05,
+      "loss": 0.4674,
       "step": 10455
     },
     {
-      "epoch": 0.37,
-      "learning_rate": 4.975117672645562e-05,
-      "loss": 0.369,
+      "epoch": 0.37697769128194036,
+      "grad_norm": 0.19031380116939545,
+      "learning_rate": 4.9730530217281023e-05,
+      "loss": 0.4621,
       "step": 10460
     },
     {
-      "epoch": 0.37,
-      "learning_rate": 4.9750775646309744e-05,
-      "loss": 0.3316,
+      "epoch": 0.37715789094316504,
+      "grad_norm": 0.15056711435317993,
+      "learning_rate": 4.973010274993106e-05,
+      "loss": 0.4718,
       "step": 10465
     },
     {
-      "epoch": 0.37,
-      "learning_rate": 4.975037424479136e-05,
-      "loss": 0.3258,
+      "epoch": 0.37733809060438966,
+      "grad_norm": 0.19555266201496124,
+      "learning_rate": 4.9729674945638236e-05,
+      "loss": 0.4505,
       "step": 10470
     },
     {
-      "epoch": 0.37,
-      "learning_rate": 4.974997252190568e-05,
-      "loss": 0.3497,
+      "epoch": 0.3775182902656143,
+      "grad_norm": 0.18490497767925262,
+      "learning_rate": 4.972924680440838e-05,
+      "loss": 0.5174,
       "step": 10475
     },
     {
-      "epoch": 0.37,
-      "learning_rate": 4.974957047765792e-05,
-      "loss": 0.2926,
+      "epoch": 0.37769848992683897,
+      "grad_norm": 0.19573722779750824,
+      "learning_rate": 4.9728818326247316e-05,
+      "loss": 0.4594,
       "step": 10480
     },
     {
-      "epoch": 0.37,
-      "learning_rate": 4.974916811205329e-05,
-      "loss": 0.3198,
+      "epoch": 0.3778786895880636,
+      "grad_norm": 0.17318245768547058,
+      "learning_rate": 4.97283895111609e-05,
+      "loss": 0.4886,
       "step": 10485
     },
     {
-      "epoch": 0.37,
-      "learning_rate": 4.9748765425097035e-05,
-      "loss": 0.3363,
+      "epoch": 0.3780588892492882,
+      "grad_norm": 0.17160874605178833,
+      "learning_rate": 4.972796035915496e-05,
+      "loss": 0.4565,
       "step": 10490
     },
     {
-      "epoch": 0.37,
-      "learning_rate": 4.974836241679436e-05,
-      "loss": 0.3114,
+      "epoch": 0.37823908891051283,
+      "grad_norm": 0.16455820202827454,
+      "learning_rate": 4.9727530870235345e-05,
+      "loss": 0.4022,
       "step": 10495
     },
     {
-      "epoch": 0.37,
-      "learning_rate": 4.974795908715051e-05,
-      "loss": 0.3215,
+      "epoch": 0.3784192885717375,
+      "grad_norm": 0.21172361075878143,
+      "learning_rate": 4.972710104440791e-05,
+      "loss": 0.449,
       "step": 10500
     },
     {
-      "epoch": 0.37,
-      "eval_loss": 0.31911852955818176,
-      "eval_runtime": 10.5355,
-      "eval_samples_per_second": 9.492,
-      "eval_steps_per_second": 9.492,
+      "epoch": 0.3784192885717375,
+      "eval_loss": 0.47920936346054077,
+      "eval_runtime": 3.5282,
+      "eval_samples_per_second": 28.343,
+      "eval_steps_per_second": 7.086,
       "step": 10500
     },
     {
-      "epoch": 0.37,
-      "learning_rate": 4.974755543617072e-05,
-      "loss": 0.312,
+      "epoch": 0.37859948823296213,
+      "grad_norm": 0.15352605283260345,
+      "learning_rate": 4.9726670881678517e-05,
+      "loss": 0.446,
       "step": 10505
     },
     {
-      "epoch": 0.37,
-      "learning_rate": 4.974715146386024e-05,
-      "loss": 0.328,
+      "epoch": 0.37877968789418676,
+      "grad_norm": 0.18436992168426514,
+      "learning_rate": 4.972624038205301e-05,
+      "loss": 0.4641,
       "step": 10510
     },
     {
-      "epoch": 0.37,
-      "learning_rate": 4.9746747170224306e-05,
-      "loss": 0.3036,
+      "epoch": 0.3789598875554114,
+      "grad_norm": 0.1690305769443512,
+      "learning_rate": 4.972580954553727e-05,
+      "loss": 0.4275,
       "step": 10515
     },
     {
-      "epoch": 0.37,
-      "learning_rate": 4.974634255526817e-05,
-      "loss": 0.3429,
+      "epoch": 0.37914008721663606,
+      "grad_norm": 0.20340761542320251,
+      "learning_rate": 4.9725378372137166e-05,
+      "loss": 0.4284,
       "step": 10520
     },
     {
-      "epoch": 0.37,
-      "learning_rate": 4.974593761899708e-05,
-      "loss": 0.31,
+      "epoch": 0.3793202868778607,
+      "grad_norm": 0.17345954477787018,
+      "learning_rate": 4.9724946861858566e-05,
+      "loss": 0.4662,
       "step": 10525
     },
     {
-      "epoch": 0.37,
-      "learning_rate": 4.9745532361416295e-05,
-      "loss": 0.3408,
+      "epoch": 0.3795004865390853,
+      "grad_norm": 0.16937977075576782,
+      "learning_rate": 4.9724515014707354e-05,
+      "loss": 0.4486,
       "step": 10530
     },
     {
-      "epoch": 0.37,
-      "learning_rate": 4.974512678253108e-05,
-      "loss": 0.3312,
+      "epoch": 0.3796806862003099,
+      "grad_norm": 0.1675933450460434,
+      "learning_rate": 4.9724082830689404e-05,
+      "loss": 0.466,
       "step": 10535
     },
     {
-      "epoch": 0.37,
-      "learning_rate": 4.97447208823467e-05,
-      "loss": 0.3257,
+      "epoch": 0.3798608858615346,
+      "grad_norm": 0.14321444928646088,
+      "learning_rate": 4.972365030981062e-05,
+      "loss": 0.4187,
       "step": 10540
     },
     {
-      "epoch": 0.37,
-      "learning_rate": 4.974431466086843e-05,
-      "loss": 0.3016,
+      "epoch": 0.3800410855227592,
+      "grad_norm": 0.19194680452346802,
+      "learning_rate": 4.972321745207688e-05,
+      "loss": 0.4069,
       "step": 10545
     },
     {
-      "epoch": 0.37,
-      "learning_rate": 4.9743908118101545e-05,
-      "loss": 0.3356,
+      "epoch": 0.38022128518398385,
+      "grad_norm": 0.16965104639530182,
+      "learning_rate": 4.972278425749409e-05,
+      "loss": 0.4158,
       "step": 10550
     },
     {
-      "epoch": 0.37,
-      "learning_rate": 4.9743501254051315e-05,
-      "loss": 0.3142,
+      "epoch": 0.38040148484520847,
+      "grad_norm": 0.18481513857841492,
+      "learning_rate": 4.972235072606816e-05,
+      "loss": 0.472,
       "step": 10555
     },
     {
-      "epoch": 0.37,
-      "learning_rate": 4.974309406872303e-05,
-      "loss": 0.3145,
+      "epoch": 0.38058168450643315,
+      "grad_norm": 0.17054641246795654,
+      "learning_rate": 4.972191685780498e-05,
+      "loss": 0.4005,
       "step": 10560
     },
     {
-      "epoch": 0.37,
-      "learning_rate": 4.974268656212198e-05,
-      "loss": 0.3287,
+      "epoch": 0.38076188416765777,
+      "grad_norm": 0.2037338763475418,
+      "learning_rate": 4.972148265271047e-05,
+      "loss": 0.4741,
       "step": 10565
     },
     {
-      "epoch": 0.37,
-      "learning_rate": 4.9742278734253444e-05,
-      "loss": 0.3307,
+      "epoch": 0.3809420838288824,
+      "grad_norm": 0.17392988502979279,
+      "learning_rate": 4.9721048110790546e-05,
+      "loss": 0.4823,
       "step": 10570
     },
     {
-      "epoch": 0.37,
-      "learning_rate": 4.974187058512273e-05,
-      "loss": 0.3203,
+      "epoch": 0.381122283490107,
+      "grad_norm": 0.1915632039308548,
+      "learning_rate": 4.972061323205113e-05,
+      "loss": 0.4084,
       "step": 10575
     },
     {
-      "epoch": 0.37,
-      "learning_rate": 4.9741462114735126e-05,
-      "loss": 0.3111,
+      "epoch": 0.3813024831513317,
+      "grad_norm": 0.16601555049419403,
+      "learning_rate": 4.972017801649814e-05,
+      "loss": 0.4264,
       "step": 10580
     },
     {
-      "epoch": 0.37,
-      "learning_rate": 4.974105332309594e-05,
-      "loss": 0.3086,
+      "epoch": 0.3814826828125563,
+      "grad_norm": 0.15968888998031616,
+      "learning_rate": 4.971974246413752e-05,
+      "loss": 0.4563,
       "step": 10585
     },
     {
-      "epoch": 0.37,
-      "learning_rate": 4.974064421021049e-05,
-      "loss": 0.2876,
+      "epoch": 0.38166288247378094,
+      "grad_norm": 0.17778979241847992,
+      "learning_rate": 4.971930657497518e-05,
+      "loss": 0.451,
       "step": 10590
     },
     {
-      "epoch": 0.37,
-      "learning_rate": 4.974023477608407e-05,
-      "loss": 0.3169,
+      "epoch": 0.38184308213500556,
+      "grad_norm": 0.16911092400550842,
+      "learning_rate": 4.971887034901708e-05,
+      "loss": 0.4147,
       "step": 10595
     },
     {
-      "epoch": 0.37,
-      "learning_rate": 4.973982502072201e-05,
-      "loss": 0.3043,
+      "epoch": 0.38202328179623024,
+      "grad_norm": 0.24606285989284515,
+      "learning_rate": 4.971843378626916e-05,
+      "loss": 0.4428,
       "step": 10600
     },
     {
-      "epoch": 0.37,
-      "learning_rate": 4.9739414944129633e-05,
-      "loss": 0.3481,
+      "epoch": 0.38220348145745486,
+      "grad_norm": 0.1590912640094757,
+      "learning_rate": 4.971799688673737e-05,
+      "loss": 0.441,
       "step": 10605
     },
     {
-      "epoch": 0.37,
-      "learning_rate": 4.973900454631225e-05,
-      "loss": 0.3462,
+      "epoch": 0.3823836811186795,
+      "grad_norm": 0.1825900822877884,
+      "learning_rate": 4.971755965042765e-05,
+      "loss": 0.4634,
       "step": 10610
     },
     {
-      "epoch": 0.37,
-      "learning_rate": 4.973859382727519e-05,
-      "loss": 0.3047,
+      "epoch": 0.3825638807799041,
+      "grad_norm": 0.14857780933380127,
+      "learning_rate": 4.9717122077345965e-05,
+      "loss": 0.4584,
       "step": 10615
     },
     {
-      "epoch": 0.37,
-      "learning_rate": 4.97381827870238e-05,
-      "loss": 0.3369,
+      "epoch": 0.3827440804411288,
+      "grad_norm": 0.18574172258377075,
+      "learning_rate": 4.971668416749828e-05,
+      "loss": 0.4721,
       "step": 10620
     },
     {
-      "epoch": 0.37,
-      "learning_rate": 4.973777142556342e-05,
-      "loss": 0.3028,
+      "epoch": 0.3829242801023534,
+      "grad_norm": 0.21157976984977722,
+      "learning_rate": 4.971624592089056e-05,
+      "loss": 0.464,
       "step": 10625
     },
     {
-      "epoch": 0.37,
-      "learning_rate": 4.973735974289937e-05,
-      "loss": 0.325,
+      "epoch": 0.38310447976357803,
+      "grad_norm": 0.15750320255756378,
+      "learning_rate": 4.971580733752877e-05,
+      "loss": 0.4572,
       "step": 10630
     },
     {
-      "epoch": 0.37,
-      "learning_rate": 4.9736947739037e-05,
-      "loss": 0.3325,
+      "epoch": 0.38328467942480265,
+      "grad_norm": 0.188838928937912,
+      "learning_rate": 4.9715368417418894e-05,
+      "loss": 0.4838,
       "step": 10635
     },
     {
-      "epoch": 0.37,
-      "learning_rate": 4.973653541398168e-05,
-      "loss": 0.3084,
+      "epoch": 0.38346487908602733,
+      "grad_norm": 0.19151824712753296,
+      "learning_rate": 4.9714929160566906e-05,
+      "loss": 0.4321,
       "step": 10640
     },
     {
-      "epoch": 0.37,
-      "learning_rate": 4.973612276773874e-05,
-      "loss": 0.3269,
+      "epoch": 0.38364507874725196,
+      "grad_norm": 0.17638704180717468,
+      "learning_rate": 4.971448956697879e-05,
+      "loss": 0.4894,
       "step": 10645
     },
     {
-      "epoch": 0.37,
-      "learning_rate": 4.973570980031355e-05,
-      "loss": 0.3139,
+      "epoch": 0.3838252784084766,
+      "grad_norm": 0.191755011677742,
+      "learning_rate": 4.9714049636660544e-05,
+      "loss": 0.4614,
       "step": 10650
     },
     {
-      "epoch": 0.37,
-      "learning_rate": 4.973529651171147e-05,
-      "loss": 0.3211,
+      "epoch": 0.38400547806970126,
+      "grad_norm": 0.1991795152425766,
+      "learning_rate": 4.971360936961815e-05,
+      "loss": 0.4343,
       "step": 10655
     },
     {
-      "epoch": 0.38,
-      "learning_rate": 4.973488290193786e-05,
-      "loss": 0.3487,
+      "epoch": 0.3841856777309259,
+      "grad_norm": 0.1604321449995041,
+      "learning_rate": 4.971316876585762e-05,
+      "loss": 0.4477,
       "step": 10660
     },
     {
-      "epoch": 0.38,
-      "learning_rate": 4.9734468970998105e-05,
-      "loss": 0.3103,
+      "epoch": 0.3843658773921505,
+      "grad_norm": 0.19753023982048035,
+      "learning_rate": 4.971272782538495e-05,
+      "loss": 0.4803,
       "step": 10665
     },
     {
-      "epoch": 0.38,
-      "learning_rate": 4.9734054718897574e-05,
-      "loss": 0.3237,
+      "epoch": 0.3845460770533751,
+      "grad_norm": 0.1573752909898758,
+      "learning_rate": 4.971228654820615e-05,
+      "loss": 0.4378,
       "step": 10670
     },
     {
-      "epoch": 0.38,
-      "learning_rate": 4.973364014564163e-05,
-      "loss": 0.2995,
+      "epoch": 0.3847262767145998,
+      "grad_norm": 0.19612035155296326,
+      "learning_rate": 4.971184493432722e-05,
+      "loss": 0.4939,
       "step": 10675
     },
     {
-      "epoch": 0.38,
-      "learning_rate": 4.973322525123568e-05,
-      "loss": 0.3166,
+      "epoch": 0.3849064763758244,
+      "grad_norm": 0.1873956173658371,
+      "learning_rate": 4.9711402983754194e-05,
+      "loss": 0.4378,
       "step": 10680
     },
     {
-      "epoch": 0.38,
-      "learning_rate": 4.973281003568509e-05,
-      "loss": 0.2974,
+      "epoch": 0.38508667603704905,
+      "grad_norm": 0.1653887778520584,
+      "learning_rate": 4.971096069649309e-05,
+      "loss": 0.4403,
       "step": 10685
     },
     {
-      "epoch": 0.38,
-      "learning_rate": 4.973239449899527e-05,
-      "loss": 0.3526,
+      "epoch": 0.38526687569827367,
+      "grad_norm": 0.1948213279247284,
+      "learning_rate": 4.971051807254993e-05,
+      "loss": 0.3959,
       "step": 10690
     },
     {
-      "epoch": 0.38,
-      "learning_rate": 4.97319786411716e-05,
-      "loss": 0.3314,
+      "epoch": 0.38544707535949835,
+      "grad_norm": 0.17649932205677032,
+      "learning_rate": 4.9710075111930744e-05,
+      "loss": 0.4329,
       "step": 10695
     },
     {
-      "epoch": 0.38,
-      "learning_rate": 4.9731562462219495e-05,
-      "loss": 0.3269,
+      "epoch": 0.38562727502072297,
+      "grad_norm": 0.16458600759506226,
+      "learning_rate": 4.970963181464157e-05,
+      "loss": 0.4233,
       "step": 10700
     },
     {
-      "epoch": 0.38,
-      "learning_rate": 4.9731145962144344e-05,
-      "loss": 0.3073,
+      "epoch": 0.3858074746819476,
+      "grad_norm": 0.1625463217496872,
+      "learning_rate": 4.970918818068844e-05,
+      "loss": 0.4972,
       "step": 10705
     },
     {
-      "epoch": 0.38,
-      "learning_rate": 4.973072914095156e-05,
-      "loss": 0.3118,
+      "epoch": 0.3859876743431722,
+      "grad_norm": 0.12830519676208496,
+      "learning_rate": 4.9708744210077406e-05,
+      "loss": 0.4458,
       "step": 10710
     },
     {
-      "epoch": 0.38,
-      "learning_rate": 4.973031199864656e-05,
-      "loss": 0.3227,
+      "epoch": 0.3861678740043969,
+      "grad_norm": 0.20030345022678375,
+      "learning_rate": 4.9708299902814516e-05,
+      "loss": 0.4751,
       "step": 10715
     },
     {
-      "epoch": 0.38,
-      "learning_rate": 4.9729894535234756e-05,
-      "loss": 0.3423,
+      "epoch": 0.3863480736656215,
+      "grad_norm": 0.15342171490192413,
+      "learning_rate": 4.970785525890582e-05,
+      "loss": 0.4467,
       "step": 10720
     },
     {
-      "epoch": 0.38,
-      "learning_rate": 4.972947675072157e-05,
-      "loss": 0.3177,
+      "epoch": 0.38652827332684614,
+      "grad_norm": 0.19483308494091034,
+      "learning_rate": 4.9707410278357393e-05,
+      "loss": 0.4594,
       "step": 10725
     },
     {
-      "epoch": 0.38,
-      "learning_rate": 4.972905864511241e-05,
-      "loss": 0.3132,
+      "epoch": 0.38670847298807076,
+      "grad_norm": 0.22155672311782837,
+      "learning_rate": 4.970696496117527e-05,
+      "loss": 0.4764,
       "step": 10730
     },
     {
-      "epoch": 0.38,
-      "learning_rate": 4.9728640218412744e-05,
-      "loss": 0.3445,
+      "epoch": 0.38688867264929544,
+      "grad_norm": 0.22407083213329315,
+      "learning_rate": 4.970651930736554e-05,
+      "loss": 0.4337,
       "step": 10735
     },
     {
-      "epoch": 0.38,
-      "learning_rate": 4.9728221470627965e-05,
-      "loss": 0.3371,
+      "epoch": 0.38706887231052006,
+      "grad_norm": 0.14635224640369415,
+      "learning_rate": 4.970607331693427e-05,
+      "loss": 0.4752,
       "step": 10740
     },
     {
-      "epoch": 0.38,
-      "learning_rate": 4.9727802401763537e-05,
-      "loss": 0.3068,
+      "epoch": 0.3872490719717447,
+      "grad_norm": 0.18128085136413574,
+      "learning_rate": 4.970562698988753e-05,
+      "loss": 0.4927,
       "step": 10745
     },
     {
-      "epoch": 0.38,
-      "learning_rate": 4.972738301182489e-05,
-      "loss": 0.308,
+      "epoch": 0.3874292716329693,
+      "grad_norm": 0.1800365000963211,
+      "learning_rate": 4.970518032623141e-05,
+      "loss": 0.4195,
       "step": 10750
     },
     {
-      "epoch": 0.38,
-      "learning_rate": 4.972696330081746e-05,
-      "loss": 0.3264,
+      "epoch": 0.387609471294194,
+      "grad_norm": 0.18861335515975952,
+      "learning_rate": 4.9704733325971986e-05,
+      "loss": 0.4563,
       "step": 10755
     },
     {
-      "epoch": 0.38,
-      "learning_rate": 4.9726543268746725e-05,
-      "loss": 0.3176,
+      "epoch": 0.3877896709554186,
+      "grad_norm": 0.18732388317584991,
+      "learning_rate": 4.9704285989115355e-05,
+      "loss": 0.4643,
       "step": 10760
     },
     {
-      "epoch": 0.38,
-      "learning_rate": 4.972612291561811e-05,
-      "loss": 0.3149,
+      "epoch": 0.38796987061664323,
+      "grad_norm": 0.17132775485515594,
+      "learning_rate": 4.970383831566762e-05,
+      "loss": 0.4932,
       "step": 10765
     },
     {
-      "epoch": 0.38,
-      "learning_rate": 4.972570224143708e-05,
-      "loss": 0.3294,
+      "epoch": 0.38815007027786785,
+      "grad_norm": 0.17382220923900604,
+      "learning_rate": 4.970339030563485e-05,
+      "loss": 0.4402,
       "step": 10770
     },
     {
-      "epoch": 0.38,
-      "learning_rate": 4.972528124620911e-05,
-      "loss": 0.341,
+      "epoch": 0.38833026993909253,
+      "grad_norm": 0.15054017305374146,
+      "learning_rate": 4.9702941959023185e-05,
+      "loss": 0.4593,
       "step": 10775
     },
     {
-      "epoch": 0.38,
-      "learning_rate": 4.9724859929939655e-05,
-      "loss": 0.331,
+      "epoch": 0.38851046960031715,
+      "grad_norm": 0.19927044212818146,
+      "learning_rate": 4.9702493275838713e-05,
+      "loss": 0.4606,
       "step": 10780
     },
     {
-      "epoch": 0.38,
-      "learning_rate": 4.972443829263419e-05,
-      "loss": 0.316,
+      "epoch": 0.3886906692615418,
+      "grad_norm": 0.16357219219207764,
+      "learning_rate": 4.970204425608756e-05,
+      "loss": 0.4324,
       "step": 10785
     },
     {
-      "epoch": 0.38,
-      "learning_rate": 4.9724016334298186e-05,
-      "loss": 0.3089,
+      "epoch": 0.3888708689227664,
+      "grad_norm": 0.1704314649105072,
+      "learning_rate": 4.970159489977583e-05,
+      "loss": 0.4257,
       "step": 10790
     },
     {
-      "epoch": 0.38,
-      "learning_rate": 4.972359405493712e-05,
-      "loss": 0.3033,
+      "epoch": 0.3890510685839911,
+      "grad_norm": 0.17188839614391327,
+      "learning_rate": 4.970114520690965e-05,
+      "loss": 0.4469,
       "step": 10795
     },
     {
-      "epoch": 0.38,
-      "learning_rate": 4.972317145455648e-05,
-      "loss": 0.3289,
+      "epoch": 0.3892312682452157,
+      "grad_norm": 0.17841334640979767,
+      "learning_rate": 4.9700695177495154e-05,
+      "loss": 0.3993,
       "step": 10800
     },
     {
-      "epoch": 0.38,
-      "learning_rate": 4.9722748533161753e-05,
-      "loss": 0.3353,
+      "epoch": 0.3894114679064403,
+      "grad_norm": 0.21081610023975372,
+      "learning_rate": 4.970024481153847e-05,
+      "loss": 0.4844,
       "step": 10805
     },
     {
-      "epoch": 0.38,
-      "learning_rate": 4.972232529075843e-05,
-      "loss": 0.3247,
+      "epoch": 0.38959166756766495,
+      "grad_norm": 0.18982069194316864,
+      "learning_rate": 4.9699794109045726e-05,
+      "loss": 0.4578,
       "step": 10810
     },
     {
-      "epoch": 0.38,
-      "learning_rate": 4.972190172735201e-05,
-      "loss": 0.3195,
+      "epoch": 0.3897718672288896,
+      "grad_norm": 0.18108953535556793,
+      "learning_rate": 4.969934307002307e-05,
+      "loss": 0.4661,
       "step": 10815
     },
     {
-      "epoch": 0.38,
-      "learning_rate": 4.9721477842947984e-05,
-      "loss": 0.3657,
+      "epoch": 0.38995206689011425,
+      "grad_norm": 0.1565956175327301,
+      "learning_rate": 4.969889169447664e-05,
+      "loss": 0.4347,
       "step": 10820
     },
     {
-      "epoch": 0.38,
-      "learning_rate": 4.972105363755186e-05,
-      "loss": 0.3068,
+      "epoch": 0.39013226655133887,
+      "grad_norm": 0.16742676496505737,
+      "learning_rate": 4.9698439982412616e-05,
+      "loss": 0.459,
       "step": 10825
     },
     {
-      "epoch": 0.38,
-      "learning_rate": 4.972062911116914e-05,
-      "loss": 0.3295,
+      "epoch": 0.39031246621256355,
+      "grad_norm": 0.15715833008289337,
+      "learning_rate": 4.969798793383711e-05,
+      "loss": 0.4821,
       "step": 10830
     },
     {
-      "epoch": 0.38,
-      "learning_rate": 4.9720204263805345e-05,
-      "loss": 0.321,
+      "epoch": 0.39049266587378817,
+      "grad_norm": 0.21100378036499023,
+      "learning_rate": 4.9697535548756304e-05,
+      "loss": 0.4679,
       "step": 10835
     },
     {
-      "epoch": 0.38,
-      "learning_rate": 4.9719779095466e-05,
-      "loss": 0.3057,
+      "epoch": 0.3906728655350128,
+      "grad_norm": 0.17519085109233856,
+      "learning_rate": 4.969708282717635e-05,
+      "loss": 0.4637,
       "step": 10840
     },
     {
-      "epoch": 0.38,
-      "learning_rate": 4.9719353606156606e-05,
-      "loss": 0.3415,
+      "epoch": 0.3908530651962374,
+      "grad_norm": 0.15852008759975433,
+      "learning_rate": 4.969662976910344e-05,
+      "loss": 0.4755,
       "step": 10845
     },
     {
-      "epoch": 0.38,
-      "learning_rate": 4.97189277958827e-05,
-      "loss": 0.3129,
+      "epoch": 0.3910332648574621,
+      "grad_norm": 0.22311712801456451,
+      "learning_rate": 4.969617637454373e-05,
+      "loss": 0.4252,
       "step": 10850
     },
     {
-      "epoch": 0.38,
-      "learning_rate": 4.97185016646498e-05,
-      "loss": 0.3154,
+      "epoch": 0.3912134645186867,
+      "grad_norm": 0.17925986647605896,
+      "learning_rate": 4.9695722643503384e-05,
+      "loss": 0.4975,
       "step": 10855
     },
     {
-      "epoch": 0.38,
-      "learning_rate": 4.971807521246345e-05,
-      "loss": 0.3251,
+      "epoch": 0.39139366417991134,
+      "grad_norm": 0.17491289973258972,
+      "learning_rate": 4.969526857598861e-05,
+      "loss": 0.4618,
       "step": 10860
     },
     {
-      "epoch": 0.38,
-      "learning_rate": 4.971764843932919e-05,
-      "loss": 0.3323,
+      "epoch": 0.39157386384113596,
+      "grad_norm": 0.2179514467716217,
+      "learning_rate": 4.969481417200558e-05,
+      "loss": 0.4999,
       "step": 10865
     },
     {
-      "epoch": 0.38,
-      "learning_rate": 4.971722134525254e-05,
-      "loss": 0.3168,
+      "epoch": 0.39175406350236064,
+      "grad_norm": 0.21115978062152863,
+      "learning_rate": 4.969435943156048e-05,
+      "loss": 0.4691,
       "step": 10870
     },
     {
-      "epoch": 0.38,
-      "learning_rate": 4.9716793930239066e-05,
-      "loss": 0.3006,
+      "epoch": 0.39193426316358526,
+      "grad_norm": 0.16908550262451172,
+      "learning_rate": 4.969390435465952e-05,
+      "loss": 0.4589,
       "step": 10875
     },
     {
-      "epoch": 0.38,
-      "learning_rate": 4.971636619429432e-05,
-      "loss": 0.3111,
+      "epoch": 0.3921144628248099,
+      "grad_norm": 0.16399134695529938,
+      "learning_rate": 4.969344894130889e-05,
+      "loss": 0.4356,
       "step": 10880
     },
     {
-      "epoch": 0.38,
-      "learning_rate": 4.971593813742383e-05,
-      "loss": 0.3157,
+      "epoch": 0.3922946624860345,
+      "grad_norm": 0.15767602622509003,
+      "learning_rate": 4.96929931915148e-05,
+      "loss": 0.4667,
       "step": 10885
     },
     {
-      "epoch": 0.38,
-      "learning_rate": 4.9715509759633194e-05,
-      "loss": 0.3366,
+      "epoch": 0.3924748621472592,
+      "grad_norm": 0.20866945385932922,
+      "learning_rate": 4.9692537105283465e-05,
+      "loss": 0.4452,
       "step": 10890
     },
     {
-      "epoch": 0.38,
-      "learning_rate": 4.9715081060927936e-05,
-      "loss": 0.3153,
+      "epoch": 0.3926550618084838,
+      "grad_norm": 0.15087029337882996,
+      "learning_rate": 4.969208068262109e-05,
+      "loss": 0.442,
       "step": 10895
     },
     {
-      "epoch": 0.38,
-      "learning_rate": 4.971465204131365e-05,
-      "loss": 0.3141,
+      "epoch": 0.39283526146970843,
+      "grad_norm": 0.1792287528514862,
+      "learning_rate": 4.969162392353389e-05,
+      "loss": 0.4511,
       "step": 10900
     },
     {
-      "epoch": 0.38,
-      "learning_rate": 4.971422270079588e-05,
-      "loss": 0.342,
+      "epoch": 0.39301546113093305,
+      "grad_norm": 0.16663479804992676,
+      "learning_rate": 4.96911668280281e-05,
+      "loss": 0.4726,
       "step": 10905
     },
     {
-      "epoch": 0.38,
-      "learning_rate": 4.971379303938023e-05,
-      "loss": 0.3119,
+      "epoch": 0.39319566079215773,
+      "grad_norm": 0.16242770850658417,
+      "learning_rate": 4.969070939610995e-05,
+      "loss": 0.4885,
       "step": 10910
     },
     {
-      "epoch": 0.38,
-      "learning_rate": 4.9713363057072256e-05,
-      "loss": 0.2966,
+      "epoch": 0.39337586045338235,
+      "grad_norm": 0.14637549221515656,
+      "learning_rate": 4.969025162778566e-05,
+      "loss": 0.4273,
       "step": 10915
     },
     {
-      "epoch": 0.38,
-      "learning_rate": 4.971293275387755e-05,
-      "loss": 0.3179,
+      "epoch": 0.393556060114607,
+      "grad_norm": 0.1730150431394577,
+      "learning_rate": 4.968979352306146e-05,
+      "loss": 0.4698,
       "step": 10920
     },
     {
-      "epoch": 0.38,
-      "learning_rate": 4.9712502129801695e-05,
-      "loss": 0.315,
+      "epoch": 0.3937362597758316,
+      "grad_norm": 0.15149009227752686,
+      "learning_rate": 4.968933508194361e-05,
+      "loss": 0.4338,
       "step": 10925
     },
     {
-      "epoch": 0.38,
-      "learning_rate": 4.971207118485029e-05,
-      "loss": 0.3437,
+      "epoch": 0.3939164594370563,
+      "grad_norm": 0.16374550759792328,
+      "learning_rate": 4.968887630443836e-05,
+      "loss": 0.4735,
       "step": 10930
     },
     {
-      "epoch": 0.38,
-      "learning_rate": 4.971163991902893e-05,
-      "loss": 0.3267,
+      "epoch": 0.3940966590982809,
+      "grad_norm": 0.16097819805145264,
+      "learning_rate": 4.968841719055194e-05,
+      "loss": 0.434,
       "step": 10935
     },
     {
-      "epoch": 0.38,
-      "learning_rate": 4.971120833234321e-05,
-      "loss": 0.3061,
+      "epoch": 0.3942768587595055,
+      "grad_norm": 0.18901294469833374,
+      "learning_rate": 4.968795774029061e-05,
+      "loss": 0.4419,
       "step": 10940
     },
     {
-      "epoch": 0.39,
-      "learning_rate": 4.971077642479873e-05,
-      "loss": 0.3452,
+      "epoch": 0.39445705842073014,
+      "grad_norm": 0.20963457226753235,
+      "learning_rate": 4.9687497953660646e-05,
+      "loss": 0.455,
       "step": 10945
     },
     {
-      "epoch": 0.39,
-      "learning_rate": 4.9710344196401114e-05,
-      "loss": 0.2993,
+      "epoch": 0.3946372580819548,
+      "grad_norm": 0.1951969563961029,
+      "learning_rate": 4.9687037830668306e-05,
+      "loss": 0.4736,
       "step": 10950
     },
     {
-      "epoch": 0.39,
-      "learning_rate": 4.9709911647155953e-05,
-      "loss": 0.3241,
+      "epoch": 0.39481745774317945,
+      "grad_norm": 0.17209535837173462,
+      "learning_rate": 4.968657737131984e-05,
+      "loss": 0.4568,
       "step": 10955
     },
     {
-      "epoch": 0.39,
-      "learning_rate": 4.9709478777068875e-05,
-      "loss": 0.3116,
+      "epoch": 0.39499765740440407,
+      "grad_norm": 0.20058324933052063,
+      "learning_rate": 4.968611657562154e-05,
+      "loss": 0.4182,
       "step": 10960
     },
     {
-      "epoch": 0.39,
-      "learning_rate": 4.97090455861455e-05,
-      "loss": 0.3161,
+      "epoch": 0.3951778570656287,
+      "grad_norm": 0.15861481428146362,
+      "learning_rate": 4.968565544357969e-05,
+      "loss": 0.4993,
       "step": 10965
     },
     {
-      "epoch": 0.39,
-      "learning_rate": 4.970861207439146e-05,
-      "loss": 0.3234,
+      "epoch": 0.39535805672685337,
+      "grad_norm": 0.18526114523410797,
+      "learning_rate": 4.968519397520056e-05,
+      "loss": 0.4578,
       "step": 10970
     },
     {
-      "epoch": 0.39,
-      "learning_rate": 4.970817824181236e-05,
-      "loss": 0.2984,
+      "epoch": 0.395538256388078,
+      "grad_norm": 0.14980530738830566,
+      "learning_rate": 4.968473217049044e-05,
+      "loss": 0.4595,
       "step": 10975
     },
     {
-      "epoch": 0.39,
-      "learning_rate": 4.970774408841387e-05,
-      "loss": 0.3408,
+      "epoch": 0.3957184560493026,
+      "grad_norm": 0.186975359916687,
+      "learning_rate": 4.9684270029455624e-05,
+      "loss": 0.45,
       "step": 10980
     },
     {
-      "epoch": 0.39,
-      "learning_rate": 4.970730961420159e-05,
-      "loss": 0.29,
+      "epoch": 0.3958986557105273,
+      "grad_norm": 0.15820086002349854,
+      "learning_rate": 4.968380755210241e-05,
+      "loss": 0.4645,
       "step": 10985
     },
     {
-      "epoch": 0.39,
-      "learning_rate": 4.970687481918118e-05,
-      "loss": 0.3175,
+      "epoch": 0.3960788553717519,
+      "grad_norm": 0.14868688583374023,
+      "learning_rate": 4.9683344738437096e-05,
+      "loss": 0.4714,
       "step": 10990
     },
     {
-      "epoch": 0.39,
-      "learning_rate": 4.970643970335828e-05,
-      "loss": 0.2969,
+      "epoch": 0.39625905503297654,
+      "grad_norm": 0.18641044199466705,
+      "learning_rate": 4.968288158846599e-05,
+      "loss": 0.4926,
       "step": 10995
     },
     {
-      "epoch": 0.39,
-      "learning_rate": 4.970600426673855e-05,
-      "loss": 0.3276,
+      "epoch": 0.39643925469420116,
+      "grad_norm": 0.17056040465831757,
+      "learning_rate": 4.968241810219539e-05,
+      "loss": 0.4738,
       "step": 11000
     },
     {
-      "epoch": 0.39,
-      "eval_loss": 0.31932923197746277,
-      "eval_runtime": 10.5374,
-      "eval_samples_per_second": 9.49,
-      "eval_steps_per_second": 9.49,
+      "epoch": 0.39643925469420116,
+      "eval_loss": 0.4763174057006836,
+      "eval_runtime": 3.5834,
+      "eval_samples_per_second": 27.907,
+      "eval_steps_per_second": 6.977,
       "step": 11000
     },
     {
-      "epoch": 0.39,
-      "learning_rate": 4.970556850932763e-05,
-      "loss": 0.3115,
+      "epoch": 0.39661945435542584,
+      "grad_norm": 0.16105006635189056,
+      "learning_rate": 4.9681954279631635e-05,
+      "loss": 0.4637,
       "step": 11005
     },
     {
-      "epoch": 0.39,
-      "learning_rate": 4.970513243113119e-05,
-      "loss": 0.3083,
+      "epoch": 0.39679965401665046,
+      "grad_norm": 0.17040333151817322,
+      "learning_rate": 4.968149012078103e-05,
+      "loss": 0.4618,
       "step": 11010
     },
     {
-      "epoch": 0.39,
-      "learning_rate": 4.970469603215488e-05,
-      "loss": 0.3105,
+      "epoch": 0.3969798536778751,
+      "grad_norm": 0.16445614397525787,
+      "learning_rate": 4.9681025625649905e-05,
+      "loss": 0.4516,
       "step": 11015
     },
     {
-      "epoch": 0.39,
-      "learning_rate": 4.9704259312404385e-05,
-      "loss": 0.3308,
+      "epoch": 0.3971600533390997,
+      "grad_norm": 0.16184879839420319,
+      "learning_rate": 4.968056079424457e-05,
+      "loss": 0.4915,
       "step": 11020
     },
     {
-      "epoch": 0.39,
-      "learning_rate": 4.9703822271885355e-05,
-      "loss": 0.332,
+      "epoch": 0.3973402530003244,
+      "grad_norm": 0.15629048645496368,
+      "learning_rate": 4.9680095626571384e-05,
+      "loss": 0.4608,
       "step": 11025
     },
     {
-      "epoch": 0.39,
-      "learning_rate": 4.9703384910603473e-05,
-      "loss": 0.3089,
+      "epoch": 0.397520452661549,
+      "grad_norm": 0.20805750787258148,
+      "learning_rate": 4.967963012263667e-05,
+      "loss": 0.4356,
       "step": 11030
     },
     {
-      "epoch": 0.39,
-      "learning_rate": 4.970294722856442e-05,
-      "loss": 0.3084,
+      "epoch": 0.39770065232277363,
+      "grad_norm": 0.16487817466259003,
+      "learning_rate": 4.967916428244677e-05,
+      "loss": 0.4734,
       "step": 11035
     },
     {
-      "epoch": 0.39,
-      "learning_rate": 4.970250922577387e-05,
-      "loss": 0.3292,
+      "epoch": 0.39788085198399825,
+      "grad_norm": 0.1743331402540207,
+      "learning_rate": 4.9678698106008034e-05,
+      "loss": 0.4458,
       "step": 11040
     },
     {
-      "epoch": 0.39,
-      "learning_rate": 4.970207090223753e-05,
-      "loss": 0.3208,
+      "epoch": 0.39806105164522293,
+      "grad_norm": 0.17199203372001648,
+      "learning_rate": 4.967823159332682e-05,
+      "loss": 0.4683,
       "step": 11045
     },
     {
-      "epoch": 0.39,
-      "learning_rate": 4.970163225796107e-05,
-      "loss": 0.3034,
+      "epoch": 0.39824125130644755,
+      "grad_norm": 0.17917218804359436,
+      "learning_rate": 4.967776474440948e-05,
+      "loss": 0.4884,
       "step": 11050
     },
     {
-      "epoch": 0.39,
-      "learning_rate": 4.97011932929502e-05,
-      "loss": 0.3415,
+      "epoch": 0.3984214509676722,
+      "grad_norm": 0.16735875606536865,
+      "learning_rate": 4.967729755926237e-05,
+      "loss": 0.4558,
       "step": 11055
     },
     {
-      "epoch": 0.39,
-      "learning_rate": 4.9700754007210606e-05,
-      "loss": 0.3013,
+      "epoch": 0.3986016506288968,
+      "grad_norm": 0.23970329761505127,
+      "learning_rate": 4.967683003789185e-05,
+      "loss": 0.4259,
       "step": 11060
     },
     {
-      "epoch": 0.39,
-      "learning_rate": 4.9700314400748005e-05,
-      "loss": 0.339,
+      "epoch": 0.3987818502901215,
+      "grad_norm": 0.17850400507450104,
+      "learning_rate": 4.967636218030431e-05,
+      "loss": 0.4225,
       "step": 11065
     },
     {
-      "epoch": 0.39,
-      "learning_rate": 4.9699874473568095e-05,
-      "loss": 0.3431,
+      "epoch": 0.3989620499513461,
+      "grad_norm": 0.14817573130130768,
+      "learning_rate": 4.967589398650611e-05,
+      "loss": 0.4513,
       "step": 11070
     },
     {
-      "epoch": 0.39,
-      "learning_rate": 4.96994342256766e-05,
-      "loss": 0.2995,
+      "epoch": 0.3991422496125707,
+      "grad_norm": 0.15426768362522125,
+      "learning_rate": 4.9675425456503634e-05,
+      "loss": 0.4577,
       "step": 11075
     },
     {
-      "epoch": 0.39,
-      "learning_rate": 4.969899365707922e-05,
-      "loss": 0.3354,
+      "epoch": 0.39932244927379534,
+      "grad_norm": 0.18102167546749115,
+      "learning_rate": 4.967495659030326e-05,
+      "loss": 0.4538,
       "step": 11080
     },
     {
-      "epoch": 0.39,
-      "learning_rate": 4.969855276778169e-05,
-      "loss": 0.3413,
+      "epoch": 0.39950264893502,
+      "grad_norm": 0.21321766078472137,
+      "learning_rate": 4.9674487387911374e-05,
+      "loss": 0.4502,
       "step": 11085
     },
     {
-      "epoch": 0.39,
-      "learning_rate": 4.969811155778974e-05,
-      "loss": 0.3404,
+      "epoch": 0.39968284859624464,
+      "grad_norm": 0.17416805028915405,
+      "learning_rate": 4.967401784933439e-05,
+      "loss": 0.4346,
       "step": 11090
     },
     {
-      "epoch": 0.39,
-      "learning_rate": 4.969767002710907e-05,
-      "loss": 0.3439,
+      "epoch": 0.39986304825746927,
+      "grad_norm": 0.20586751401424408,
+      "learning_rate": 4.9673547974578674e-05,
+      "loss": 0.4706,
       "step": 11095
     },
     {
-      "epoch": 0.39,
-      "learning_rate": 4.969722817574544e-05,
-      "loss": 0.3299,
+      "epoch": 0.4000432479186939,
+      "grad_norm": 0.1607694774866104,
+      "learning_rate": 4.967307776365065e-05,
+      "loss": 0.4478,
       "step": 11100
     },
     {
-      "epoch": 0.39,
-      "learning_rate": 4.969678600370457e-05,
-      "loss": 0.3264,
+      "epoch": 0.40022344757991857,
+      "grad_norm": 0.2245350033044815,
+      "learning_rate": 4.967260721655672e-05,
+      "loss": 0.4874,
       "step": 11105
     },
     {
-      "epoch": 0.39,
-      "learning_rate": 4.969634351099222e-05,
-      "loss": 0.3363,
+      "epoch": 0.4004036472411432,
+      "grad_norm": 0.18260863423347473,
+      "learning_rate": 4.967213633330329e-05,
+      "loss": 0.4176,
       "step": 11110
     },
     {
-      "epoch": 0.39,
-      "learning_rate": 4.9695900697614126e-05,
-      "loss": 0.3401,
+      "epoch": 0.4005838469023678,
+      "grad_norm": 0.1575452834367752,
+      "learning_rate": 4.967166511389678e-05,
+      "loss": 0.4249,
       "step": 11115
     },
     {
-      "epoch": 0.39,
-      "learning_rate": 4.969545756357603e-05,
-      "loss": 0.3148,
+      "epoch": 0.40076404656359244,
+      "grad_norm": 0.18237917125225067,
+      "learning_rate": 4.967119355834361e-05,
+      "loss": 0.4794,
       "step": 11120
     },
     {
-      "epoch": 0.39,
-      "learning_rate": 4.969501410888369e-05,
-      "loss": 0.2961,
+      "epoch": 0.4009442462248171,
+      "grad_norm": 0.2220340073108673,
+      "learning_rate": 4.96707216666502e-05,
+      "loss": 0.4514,
       "step": 11125
     },
     {
-      "epoch": 0.39,
-      "learning_rate": 4.9694570333542874e-05,
-      "loss": 0.2927,
+      "epoch": 0.40112444588604174,
+      "grad_norm": 0.17675887048244476,
+      "learning_rate": 4.9670249438822994e-05,
+      "loss": 0.4648,
       "step": 11130
     },
     {
-      "epoch": 0.39,
-      "learning_rate": 4.9694126237559335e-05,
-      "loss": 0.3554,
+      "epoch": 0.40130464554726636,
+      "grad_norm": 0.19447670876979828,
+      "learning_rate": 4.966977687486841e-05,
+      "loss": 0.4771,
       "step": 11135
     },
     {
-      "epoch": 0.39,
-      "learning_rate": 4.969368182093883e-05,
-      "loss": 0.3212,
+      "epoch": 0.401484845208491,
+      "grad_norm": 0.1338052898645401,
+      "learning_rate": 4.966930397479289e-05,
+      "loss": 0.416,
       "step": 11140
     },
     {
-      "epoch": 0.39,
-      "learning_rate": 4.9693237083687155e-05,
-      "loss": 0.3346,
+      "epoch": 0.40166504486971566,
+      "grad_norm": 0.19171032309532166,
+      "learning_rate": 4.966883073860288e-05,
+      "loss": 0.4714,
       "step": 11145
     },
     {
-      "epoch": 0.39,
-      "learning_rate": 4.969279202581007e-05,
-      "loss": 0.341,
+      "epoch": 0.4018452445309403,
+      "grad_norm": 0.17477290332317352,
+      "learning_rate": 4.966835716630483e-05,
+      "loss": 0.4862,
       "step": 11150
     },
     {
-      "epoch": 0.39,
-      "learning_rate": 4.9692346647313343e-05,
-      "loss": 0.3208,
+      "epoch": 0.4020254441921649,
+      "grad_norm": 0.22172503173351288,
+      "learning_rate": 4.966788325790519e-05,
+      "loss": 0.4264,
       "step": 11155
     },
     {
-      "epoch": 0.39,
-      "learning_rate": 4.969190094820278e-05,
-      "loss": 0.3564,
+      "epoch": 0.4022056438533896,
+      "grad_norm": 0.1821010559797287,
+      "learning_rate": 4.966740901341042e-05,
+      "loss": 0.4472,
       "step": 11160
     },
     {
-      "epoch": 0.39,
-      "learning_rate": 4.969145492848414e-05,
-      "loss": 0.3175,
+      "epoch": 0.4023858435146142,
+      "grad_norm": 0.21025283634662628,
+      "learning_rate": 4.9666934432826975e-05,
+      "loss": 0.446,
       "step": 11165
     },
     {
-      "epoch": 0.39,
-      "learning_rate": 4.969100858816324e-05,
-      "loss": 0.3286,
+      "epoch": 0.40256604317583883,
+      "grad_norm": 0.16647890210151672,
+      "learning_rate": 4.9666459516161316e-05,
+      "loss": 0.4754,
       "step": 11170
     },
     {
-      "epoch": 0.39,
-      "learning_rate": 4.969056192724586e-05,
-      "loss": 0.3217,
+      "epoch": 0.40274624283706345,
+      "grad_norm": 0.15028271079063416,
+      "learning_rate": 4.9665984263419926e-05,
+      "loss": 0.4176,
       "step": 11175
     },
     {
-      "epoch": 0.39,
-      "learning_rate": 4.969011494573781e-05,
-      "loss": 0.3129,
+      "epoch": 0.40292644249828813,
+      "grad_norm": 0.1495635062456131,
+      "learning_rate": 4.9665508674609277e-05,
+      "loss": 0.4659,
       "step": 11180
     },
     {
-      "epoch": 0.39,
-      "learning_rate": 4.968966764364489e-05,
-      "loss": 0.3072,
+      "epoch": 0.40310664215951275,
+      "grad_norm": 0.17869426310062408,
+      "learning_rate": 4.966503274973585e-05,
+      "loss": 0.453,
       "step": 11185
     },
     {
-      "epoch": 0.39,
-      "learning_rate": 4.96892200209729e-05,
-      "loss": 0.3381,
+      "epoch": 0.4032868418207374,
+      "grad_norm": 0.1512574553489685,
+      "learning_rate": 4.9664556488806124e-05,
+      "loss": 0.4645,
       "step": 11190
     },
     {
-      "epoch": 0.39,
-      "learning_rate": 4.968877207772766e-05,
-      "loss": 0.297,
+      "epoch": 0.403467041481962,
+      "grad_norm": 0.18554282188415527,
+      "learning_rate": 4.96640798918266e-05,
+      "loss": 0.4688,
       "step": 11195
     },
     {
-      "epoch": 0.39,
-      "learning_rate": 4.9688323813914995e-05,
-      "loss": 0.3213,
+      "epoch": 0.4036472411431867,
+      "grad_norm": 0.16681110858917236,
+      "learning_rate": 4.966360295880375e-05,
+      "loss": 0.4333,
       "step": 11200
     },
     {
-      "epoch": 0.39,
-      "learning_rate": 4.968787522954071e-05,
-      "loss": 0.3401,
+      "epoch": 0.4038274408044113,
+      "grad_norm": 0.2254079133272171,
+      "learning_rate": 4.966312568974409e-05,
+      "loss": 0.4501,
       "step": 11205
     },
     {
-      "epoch": 0.39,
-      "learning_rate": 4.968742632461063e-05,
-      "loss": 0.3256,
+      "epoch": 0.4040076404656359,
+      "grad_norm": 0.14422693848609924,
+      "learning_rate": 4.966264808465412e-05,
+      "loss": 0.4039,
       "step": 11210
     },
     {
-      "epoch": 0.39,
-      "learning_rate": 4.9686977099130594e-05,
-      "loss": 0.315,
+      "epoch": 0.40418784012686054,
+      "grad_norm": 0.17305003106594086,
+      "learning_rate": 4.9662170143540336e-05,
+      "loss": 0.4803,
       "step": 11215
     },
     {
-      "epoch": 0.39,
-      "learning_rate": 4.968652755310643e-05,
-      "loss": 0.3294,
+      "epoch": 0.4043680397880852,
+      "grad_norm": 0.18754348158836365,
+      "learning_rate": 4.966169186640927e-05,
+      "loss": 0.4577,
       "step": 11220
     },
     {
-      "epoch": 0.39,
-      "learning_rate": 4.968607768654398e-05,
-      "loss": 0.3661,
+      "epoch": 0.40454823944930984,
+      "grad_norm": 0.15006259083747864,
+      "learning_rate": 4.966121325326742e-05,
+      "loss": 0.4578,
       "step": 11225
     },
     {
-      "epoch": 0.4,
-      "learning_rate": 4.968562749944906e-05,
-      "loss": 0.3357,
+      "epoch": 0.40472843911053447,
+      "grad_norm": 0.1722080409526825,
+      "learning_rate": 4.9660734304121315e-05,
+      "loss": 0.4343,
       "step": 11230
     },
     {
-      "epoch": 0.4,
-      "learning_rate": 4.9685176991827555e-05,
-      "loss": 0.3292,
+      "epoch": 0.4049086387717591,
+      "grad_norm": 0.13911651074886322,
+      "learning_rate": 4.9660255018977475e-05,
+      "loss": 0.4435,
       "step": 11235
     },
     {
-      "epoch": 0.4,
-      "learning_rate": 4.968472616368529e-05,
-      "loss": 0.3061,
+      "epoch": 0.40508883843298377,
+      "grad_norm": 0.20192518830299377,
+      "learning_rate": 4.9659775397842444e-05,
+      "loss": 0.4632,
       "step": 11240
     },
     {
-      "epoch": 0.4,
-      "learning_rate": 4.968427501502812e-05,
-      "loss": 0.3282,
+      "epoch": 0.4052690380942084,
+      "grad_norm": 0.17392924427986145,
+      "learning_rate": 4.965929544072274e-05,
+      "loss": 0.4377,
       "step": 11245
     },
     {
-      "epoch": 0.4,
-      "learning_rate": 4.968382354586191e-05,
-      "loss": 0.324,
+      "epoch": 0.405449237755433,
+      "grad_norm": 0.17670120298862457,
+      "learning_rate": 4.9658815147624914e-05,
+      "loss": 0.4451,
       "step": 11250
     },
     {
-      "epoch": 0.4,
-      "learning_rate": 4.968337175619251e-05,
-      "loss": 0.3114,
+      "epoch": 0.40562943741665763,
+      "grad_norm": 0.22466880083084106,
+      "learning_rate": 4.9658334518555507e-05,
+      "loss": 0.5001,
       "step": 11255
     },
     {
-      "epoch": 0.4,
-      "learning_rate": 4.9682919646025805e-05,
-      "loss": 0.3184,
+      "epoch": 0.4058096370778823,
+      "grad_norm": 0.18523770570755005,
+      "learning_rate": 4.965785355352106e-05,
+      "loss": 0.4358,
       "step": 11260
     },
     {
-      "epoch": 0.4,
-      "learning_rate": 4.9682467215367655e-05,
-      "loss": 0.3407,
+      "epoch": 0.40598983673910694,
+      "grad_norm": 0.20358721911907196,
+      "learning_rate": 4.965737225252814e-05,
+      "loss": 0.4626,
       "step": 11265
     },
     {
-      "epoch": 0.4,
-      "learning_rate": 4.968201446422393e-05,
-      "loss": 0.2974,
+      "epoch": 0.40617003640033156,
+      "grad_norm": 0.2111283838748932,
+      "learning_rate": 4.9656890615583297e-05,
+      "loss": 0.4708,
       "step": 11270
     },
     {
-      "epoch": 0.4,
-      "learning_rate": 4.968156139260052e-05,
-      "loss": 0.3225,
+      "epoch": 0.4063502360615562,
+      "grad_norm": 0.200625941157341,
+      "learning_rate": 4.965640864269309e-05,
+      "loss": 0.4916,
       "step": 11275
     },
     {
-      "epoch": 0.4,
-      "learning_rate": 4.968110800050329e-05,
-      "loss": 0.328,
+      "epoch": 0.40653043572278086,
+      "grad_norm": 0.17772261798381805,
+      "learning_rate": 4.965592633386408e-05,
+      "loss": 0.4205,
       "step": 11280
     },
     {
-      "epoch": 0.4,
-      "learning_rate": 4.9680654287938145e-05,
-      "loss": 0.3285,
+      "epoch": 0.4067106353840055,
+      "grad_norm": 0.16529619693756104,
+      "learning_rate": 4.965544368910285e-05,
+      "loss": 0.4354,
       "step": 11285
     },
     {
-      "epoch": 0.4,
-      "learning_rate": 4.9680200254910966e-05,
-      "loss": 0.3313,
+      "epoch": 0.4068908350452301,
+      "grad_norm": 0.12482757121324539,
+      "learning_rate": 4.965496070841599e-05,
+      "loss": 0.4189,
       "step": 11290
     },
     {
-      "epoch": 0.4,
-      "learning_rate": 4.9679745901427656e-05,
-      "loss": 0.3458,
+      "epoch": 0.4070710347064547,
+      "grad_norm": 0.16134874522686005,
+      "learning_rate": 4.965447739181005e-05,
+      "loss": 0.4405,
       "step": 11295
     },
     {
-      "epoch": 0.4,
-      "learning_rate": 4.967929122749411e-05,
-      "loss": 0.3087,
+      "epoch": 0.4072512343676794,
+      "grad_norm": 0.1545429825782776,
+      "learning_rate": 4.965399373929163e-05,
+      "loss": 0.4762,
       "step": 11300
     },
     {
-      "epoch": 0.4,
-      "learning_rate": 4.967883623311622e-05,
-      "loss": 0.3174,
+      "epoch": 0.407431434028904,
+      "grad_norm": 0.16678392887115479,
+      "learning_rate": 4.9653509750867324e-05,
+      "loss": 0.4442,
       "step": 11305
     },
     {
-      "epoch": 0.4,
-      "learning_rate": 4.9678380918299915e-05,
-      "loss": 0.3439,
+      "epoch": 0.40761163369012865,
+      "grad_norm": 0.19490495324134827,
+      "learning_rate": 4.965302542654371e-05,
+      "loss": 0.436,
       "step": 11310
     },
     {
-      "epoch": 0.4,
-      "learning_rate": 4.9677925283051105e-05,
-      "loss": 0.3138,
+      "epoch": 0.4077918333513533,
+      "grad_norm": 0.19900862872600555,
+      "learning_rate": 4.9652540766327406e-05,
+      "loss": 0.5119,
       "step": 11315
     },
     {
-      "epoch": 0.4,
-      "learning_rate": 4.967746932737569e-05,
-      "loss": 0.3158,
+      "epoch": 0.40797203301257795,
+      "grad_norm": 0.1822587251663208,
+      "learning_rate": 4.9652055770225005e-05,
+      "loss": 0.4696,
       "step": 11320
     },
     {
-      "epoch": 0.4,
-      "learning_rate": 4.96770130512796e-05,
-      "loss": 0.3368,
+      "epoch": 0.4081522326738026,
+      "grad_norm": 0.1810266673564911,
+      "learning_rate": 4.965157043824311e-05,
+      "loss": 0.4568,
       "step": 11325
     },
     {
-      "epoch": 0.4,
-      "learning_rate": 4.9676556454768755e-05,
-      "loss": 0.3406,
+      "epoch": 0.4083324323350272,
+      "grad_norm": 0.19422537088394165,
+      "learning_rate": 4.965108477038835e-05,
+      "loss": 0.4705,
       "step": 11330
     },
     {
-      "epoch": 0.4,
-      "learning_rate": 4.9676099537849083e-05,
-      "loss": 0.3389,
+      "epoch": 0.4085126319962519,
+      "grad_norm": 0.15788643062114716,
+      "learning_rate": 4.965059876666733e-05,
+      "loss": 0.4606,
       "step": 11335
     },
     {
-      "epoch": 0.4,
-      "learning_rate": 4.967564230052653e-05,
-      "loss": 0.31,
+      "epoch": 0.4086928316574765,
+      "grad_norm": 0.15164151787757874,
+      "learning_rate": 4.965011242708667e-05,
+      "loss": 0.465,
       "step": 11340
     },
     {
-      "epoch": 0.4,
-      "learning_rate": 4.9675184742807025e-05,
-      "loss": 0.3074,
+      "epoch": 0.4088730313187011,
+      "grad_norm": 0.19079901278018951,
+      "learning_rate": 4.964962575165301e-05,
+      "loss": 0.4909,
       "step": 11345
     },
     {
-      "epoch": 0.4,
-      "learning_rate": 4.96747268646965e-05,
-      "loss": 0.3408,
+      "epoch": 0.40905323097992574,
+      "grad_norm": 0.1349342167377472,
+      "learning_rate": 4.964913874037296e-05,
+      "loss": 0.4571,
       "step": 11350
     },
     {
-      "epoch": 0.4,
-      "learning_rate": 4.967426866620091e-05,
-      "loss": 0.3086,
+      "epoch": 0.4092334306411504,
+      "grad_norm": 0.1663023829460144,
+      "learning_rate": 4.9648651393253176e-05,
+      "loss": 0.461,
       "step": 11355
     },
     {
-      "epoch": 0.4,
-      "learning_rate": 4.967381014732621e-05,
-      "loss": 0.329,
+      "epoch": 0.40941363030237504,
+      "grad_norm": 0.16813883185386658,
+      "learning_rate": 4.964816371030029e-05,
+      "loss": 0.408,
       "step": 11360
     },
     {
-      "epoch": 0.4,
-      "learning_rate": 4.9673351308078335e-05,
-      "loss": 0.3225,
+      "epoch": 0.40959382996359966,
+      "grad_norm": 0.15600088238716125,
+      "learning_rate": 4.964767569152093e-05,
+      "loss": 0.4358,
       "step": 11365
     },
     {
-      "epoch": 0.4,
-      "learning_rate": 4.9672892148463254e-05,
-      "loss": 0.3171,
+      "epoch": 0.4097740296248243,
+      "grad_norm": 0.14908957481384277,
+      "learning_rate": 4.964718733692178e-05,
+      "loss": 0.4046,
       "step": 11370
     },
     {
-      "epoch": 0.4,
-      "learning_rate": 4.967243266848694e-05,
-      "loss": 0.2968,
+      "epoch": 0.40995422928604897,
+      "grad_norm": 0.1925799548625946,
+      "learning_rate": 4.9646698646509465e-05,
+      "loss": 0.4426,
       "step": 11375
     },
     {
-      "epoch": 0.4,
-      "learning_rate": 4.967197286815534e-05,
-      "loss": 0.3255,
+      "epoch": 0.4101344289472736,
+      "grad_norm": 0.15564826130867004,
+      "learning_rate": 4.9646209620290654e-05,
+      "loss": 0.4438,
       "step": 11380
     },
     {
-      "epoch": 0.4,
-      "learning_rate": 4.9671512747474436e-05,
-      "loss": 0.313,
+      "epoch": 0.4103146286084982,
+      "grad_norm": 0.19330483675003052,
+      "learning_rate": 4.9645720258272014e-05,
+      "loss": 0.4536,
       "step": 11385
     },
     {
-      "epoch": 0.4,
-      "learning_rate": 4.967105230645019e-05,
-      "loss": 0.3233,
+      "epoch": 0.41049482826972283,
+      "grad_norm": 0.18074195086956024,
+      "learning_rate": 4.96452305604602e-05,
+      "loss": 0.4674,
       "step": 11390
     },
     {
-      "epoch": 0.4,
-      "learning_rate": 4.96705915450886e-05,
-      "loss": 0.3151,
+      "epoch": 0.4106750279309475,
+      "grad_norm": 0.20103107392787933,
+      "learning_rate": 4.964474052686189e-05,
+      "loss": 0.4257,
       "step": 11395
     },
     {
-      "epoch": 0.4,
-      "learning_rate": 4.9670130463395624e-05,
-      "loss": 0.3142,
+      "epoch": 0.41085522759217213,
+      "grad_norm": 0.14884856343269348,
+      "learning_rate": 4.9644250157483765e-05,
+      "loss": 0.4443,
       "step": 11400
     },
     {
-      "epoch": 0.4,
-      "learning_rate": 4.9669669061377266e-05,
-      "loss": 0.3363,
+      "epoch": 0.41103542725339676,
+      "grad_norm": 0.1957724243402481,
+      "learning_rate": 4.96437594523325e-05,
+      "loss": 0.4462,
       "step": 11405
     },
     {
-      "epoch": 0.4,
-      "learning_rate": 4.9669207339039514e-05,
-      "loss": 0.3107,
+      "epoch": 0.4112156269146214,
+      "grad_norm": 0.15725165605545044,
+      "learning_rate": 4.964326841141479e-05,
+      "loss": 0.4184,
       "step": 11410
     },
     {
-      "epoch": 0.4,
-      "learning_rate": 4.9668745296388366e-05,
-      "loss": 0.3451,
+      "epoch": 0.41139582657584606,
+      "grad_norm": 0.16227923333644867,
+      "learning_rate": 4.964277703473731e-05,
+      "loss": 0.4157,
       "step": 11415
     },
     {
-      "epoch": 0.4,
-      "learning_rate": 4.966828293342982e-05,
-      "loss": 0.3532,
+      "epoch": 0.4115760262370707,
+      "grad_norm": 0.1838981658220291,
+      "learning_rate": 4.9642285322306766e-05,
+      "loss": 0.4706,
       "step": 11420
     },
     {
-      "epoch": 0.4,
-      "learning_rate": 4.9667820250169874e-05,
-      "loss": 0.33,
+      "epoch": 0.4117562258982953,
+      "grad_norm": 0.19162507355213165,
+      "learning_rate": 4.9641793274129864e-05,
+      "loss": 0.4535,
       "step": 11425
     },
     {
-      "epoch": 0.4,
-      "learning_rate": 4.966735724661453e-05,
-      "loss": 0.3177,
+      "epoch": 0.4119364255595199,
+      "grad_norm": 0.14456330239772797,
+      "learning_rate": 4.964130089021329e-05,
+      "loss": 0.4752,
       "step": 11430
     },
     {
-      "epoch": 0.4,
-      "learning_rate": 4.966689392276982e-05,
-      "loss": 0.3225,
+      "epoch": 0.4121166252207446,
+      "grad_norm": 0.17111894488334656,
+      "learning_rate": 4.964080817056377e-05,
+      "loss": 0.4575,
       "step": 11435
     },
     {
-      "epoch": 0.4,
-      "learning_rate": 4.966643027864174e-05,
-      "loss": 0.3229,
+      "epoch": 0.4122968248819692,
+      "grad_norm": 0.16951219737529755,
+      "learning_rate": 4.9640315115188004e-05,
+      "loss": 0.4254,
       "step": 11440
     },
     {
-      "epoch": 0.4,
-      "learning_rate": 4.9665966314236324e-05,
-      "loss": 0.3148,
+      "epoch": 0.41247702454319385,
+      "grad_norm": 0.14794260263442993,
+      "learning_rate": 4.963982172409272e-05,
+      "loss": 0.4429,
       "step": 11445
     },
     {
-      "epoch": 0.4,
-      "learning_rate": 4.966550202955959e-05,
-      "loss": 0.3068,
+      "epoch": 0.41265722420441847,
+      "grad_norm": 0.21253348886966705,
+      "learning_rate": 4.963932799728462e-05,
+      "loss": 0.4842,
       "step": 11450
     },
     {
-      "epoch": 0.4,
-      "learning_rate": 4.966503742461756e-05,
-      "loss": 0.3421,
+      "epoch": 0.41283742386564315,
+      "grad_norm": 0.19845059514045715,
+      "learning_rate": 4.963883393477046e-05,
+      "loss": 0.4506,
       "step": 11455
     },
     {
-      "epoch": 0.4,
-      "learning_rate": 4.966457249941629e-05,
-      "loss": 0.3407,
+      "epoch": 0.41301762352686777,
+      "grad_norm": 0.16377714276313782,
+      "learning_rate": 4.963833953655696e-05,
+      "loss": 0.43,
       "step": 11460
     },
     {
-      "epoch": 0.4,
-      "learning_rate": 4.9664107253961786e-05,
-      "loss": 0.3071,
+      "epoch": 0.4131978231880924,
+      "grad_norm": 0.13513842225074768,
+      "learning_rate": 4.963784480265085e-05,
+      "loss": 0.4142,
       "step": 11465
     },
     {
-      "epoch": 0.4,
-      "learning_rate": 4.966364168826011e-05,
-      "loss": 0.3285,
+      "epoch": 0.413378022849317,
+      "grad_norm": 0.20631147921085358,
+      "learning_rate": 4.963734973305887e-05,
+      "loss": 0.4484,
       "step": 11470
     },
     {
-      "epoch": 0.4,
-      "learning_rate": 4.9663175802317306e-05,
-      "loss": 0.3165,
+      "epoch": 0.4135582225105417,
+      "grad_norm": 0.14362797141075134,
+      "learning_rate": 4.963685432778777e-05,
+      "loss": 0.4563,
       "step": 11475
     },
     {
-      "epoch": 0.4,
-      "learning_rate": 4.966270959613941e-05,
-      "loss": 0.3275,
+      "epoch": 0.4137384221717663,
+      "grad_norm": 0.17292509973049164,
+      "learning_rate": 4.963635858684431e-05,
+      "loss": 0.4537,
       "step": 11480
     },
     {
-      "epoch": 0.4,
-      "learning_rate": 4.966224306973249e-05,
-      "loss": 0.3074,
+      "epoch": 0.41391862183299094,
+      "grad_norm": 0.19188092648983002,
+      "learning_rate": 4.963586251023523e-05,
+      "loss": 0.469,
       "step": 11485
     },
     {
-      "epoch": 0.4,
-      "learning_rate": 4.9661776223102583e-05,
-      "loss": 0.3402,
+      "epoch": 0.41409882149421556,
+      "grad_norm": 0.17507772147655487,
+      "learning_rate": 4.963536609796729e-05,
+      "loss": 0.424,
       "step": 11490
     },
     {
-      "epoch": 0.4,
-      "learning_rate": 4.966130905625578e-05,
-      "loss": 0.3486,
+      "epoch": 0.41427902115544024,
+      "grad_norm": 0.13874538242816925,
+      "learning_rate": 4.963486935004725e-05,
+      "loss": 0.457,
       "step": 11495
     },
     {
-      "epoch": 0.4,
-      "learning_rate": 4.966084156919813e-05,
-      "loss": 0.3687,
+      "epoch": 0.41445922081666486,
+      "grad_norm": 0.1583535075187683,
+      "learning_rate": 4.96343722664819e-05,
+      "loss": 0.4558,
       "step": 11500
     },
     {
-      "epoch": 0.4,
-      "eval_loss": 0.31692856550216675,
-      "eval_runtime": 10.5329,
-      "eval_samples_per_second": 9.494,
-      "eval_steps_per_second": 9.494,
+      "epoch": 0.41445922081666486,
+      "eval_loss": 0.4750808775424957,
+      "eval_runtime": 3.597,
+      "eval_samples_per_second": 27.801,
+      "eval_steps_per_second": 6.95,
       "step": 11500
     },
     {
-      "epoch": 0.4,
-      "learning_rate": 4.966037376193569e-05,
-      "loss": 0.2997,
+      "epoch": 0.4146394204778895,
+      "grad_norm": 0.1701897829771042,
+      "learning_rate": 4.9633874847277985e-05,
+      "loss": 0.509,
       "step": 11505
     },
     {
-      "epoch": 0.4,
-      "learning_rate": 4.965990563447456e-05,
-      "loss": 0.3563,
+      "epoch": 0.41481962013911416,
+      "grad_norm": 0.20449738204479218,
+      "learning_rate": 4.9633377092442305e-05,
+      "loss": 0.4207,
       "step": 11510
     },
     {
-      "epoch": 0.41,
-      "learning_rate": 4.96594371868208e-05,
-      "loss": 0.2938,
+      "epoch": 0.4149998198003388,
+      "grad_norm": 0.21852335333824158,
+      "learning_rate": 4.9632879001981616e-05,
+      "loss": 0.4482,
       "step": 11515
     },
     {
-      "epoch": 0.41,
-      "learning_rate": 4.96589684189805e-05,
-      "loss": 0.309,
+      "epoch": 0.4151800194615634,
+      "grad_norm": 0.1580123007297516,
+      "learning_rate": 4.963238057590273e-05,
+      "loss": 0.4244,
       "step": 11520
     },
     {
-      "epoch": 0.41,
-      "learning_rate": 4.965849933095976e-05,
-      "loss": 0.3217,
+      "epoch": 0.41536021912278803,
+      "grad_norm": 0.18125052750110626,
+      "learning_rate": 4.963188181421243e-05,
+      "loss": 0.4599,
       "step": 11525
     },
     {
-      "epoch": 0.41,
-      "learning_rate": 4.965802992276464e-05,
-      "loss": 0.3492,
+      "epoch": 0.4155404187840127,
+      "grad_norm": 0.18217013776302338,
+      "learning_rate": 4.9631382716917504e-05,
+      "loss": 0.4347,
       "step": 11530
     },
     {
-      "epoch": 0.41,
-      "learning_rate": 4.9657560194401266e-05,
-      "loss": 0.3112,
+      "epoch": 0.41572061844523733,
+      "grad_norm": 0.1741897016763687,
+      "learning_rate": 4.9630883284024756e-05,
+      "loss": 0.428,
       "step": 11535
     },
     {
-      "epoch": 0.41,
-      "learning_rate": 4.9657090145875704e-05,
-      "loss": 0.3295,
+      "epoch": 0.41590081810646196,
+      "grad_norm": 0.20398588478565216,
+      "learning_rate": 4.9630383515541e-05,
+      "loss": 0.4587,
       "step": 11540
     },
     {
-      "epoch": 0.41,
-      "learning_rate": 4.965661977719409e-05,
-      "loss": 0.3108,
+      "epoch": 0.4160810177676866,
+      "grad_norm": 0.1447395384311676,
+      "learning_rate": 4.9629883411473025e-05,
+      "loss": 0.4406,
       "step": 11545
     },
     {
-      "epoch": 0.41,
-      "learning_rate": 4.965614908836251e-05,
-      "loss": 0.3004,
+      "epoch": 0.41626121742891126,
+      "grad_norm": 0.21012569963932037,
+      "learning_rate": 4.962938297182767e-05,
+      "loss": 0.4901,
       "step": 11550
     },
     {
-      "epoch": 0.41,
-      "learning_rate": 4.965567807938709e-05,
-      "loss": 0.3019,
+      "epoch": 0.4164414170901359,
+      "grad_norm": 0.17350612580776215,
+      "learning_rate": 4.962888219661173e-05,
+      "loss": 0.4447,
       "step": 11555
     },
     {
-      "epoch": 0.41,
-      "learning_rate": 4.965520675027394e-05,
-      "loss": 0.319,
+      "epoch": 0.4166216167513605,
+      "grad_norm": 0.167585551738739,
+      "learning_rate": 4.9628381085832046e-05,
+      "loss": 0.4391,
       "step": 11560
     },
     {
-      "epoch": 0.41,
-      "learning_rate": 4.965473510102918e-05,
-      "loss": 0.3118,
+      "epoch": 0.4168018164125851,
+      "grad_norm": 0.15864285826683044,
+      "learning_rate": 4.962787963949543e-05,
+      "loss": 0.4621,
       "step": 11565
     },
     {
-      "epoch": 0.41,
-      "learning_rate": 4.965426313165892e-05,
-      "loss": 0.3351,
+      "epoch": 0.4169820160738098,
+      "grad_norm": 0.20318114757537842,
+      "learning_rate": 4.9627377857608725e-05,
+      "loss": 0.4257,
       "step": 11570
     },
     {
-      "epoch": 0.41,
-      "learning_rate": 4.965379084216931e-05,
-      "loss": 0.3064,
+      "epoch": 0.4171622157350344,
+      "grad_norm": 0.1912168711423874,
+      "learning_rate": 4.962687574017877e-05,
+      "loss": 0.43,
       "step": 11575
     },
     {
-      "epoch": 0.41,
-      "learning_rate": 4.965331823256648e-05,
-      "loss": 0.3145,
+      "epoch": 0.41734241539625905,
+      "grad_norm": 0.1616230010986328,
+      "learning_rate": 4.962637328721239e-05,
+      "loss": 0.446,
       "step": 11580
     },
     {
-      "epoch": 0.41,
-      "learning_rate": 4.9652845302856545e-05,
-      "loss": 0.3336,
+      "epoch": 0.41752261505748367,
+      "grad_norm": 0.21484871208667755,
+      "learning_rate": 4.962587049871645e-05,
+      "loss": 0.4384,
       "step": 11585
     },
     {
-      "epoch": 0.41,
-      "learning_rate": 4.965237205304567e-05,
-      "loss": 0.298,
+      "epoch": 0.41770281471870835,
+      "grad_norm": 0.1611786037683487,
+      "learning_rate": 4.9625367374697795e-05,
+      "loss": 0.4294,
       "step": 11590
     },
     {
-      "epoch": 0.41,
-      "learning_rate": 4.965189848313999e-05,
-      "loss": 0.2922,
+      "epoch": 0.41788301437993297,
+      "grad_norm": 0.17779512703418732,
+      "learning_rate": 4.9624863915163275e-05,
+      "loss": 0.487,
       "step": 11595
     },
     {
-      "epoch": 0.41,
-      "learning_rate": 4.965142459314565e-05,
-      "loss": 0.2873,
+      "epoch": 0.4180632140411576,
+      "grad_norm": 0.16837961971759796,
+      "learning_rate": 4.962436012011975e-05,
+      "loss": 0.4363,
       "step": 11600
     },
     {
-      "epoch": 0.41,
-      "learning_rate": 4.9650950383068813e-05,
-      "loss": 0.3168,
+      "epoch": 0.4182434137023822,
+      "grad_norm": 0.13357023894786835,
+      "learning_rate": 4.9623855989574086e-05,
+      "loss": 0.4507,
       "step": 11605
     },
     {
-      "epoch": 0.41,
-      "learning_rate": 4.9650475852915625e-05,
-      "loss": 0.3136,
+      "epoch": 0.4184236133636069,
+      "grad_norm": 0.13329064846038818,
+      "learning_rate": 4.9623351523533144e-05,
+      "loss": 0.4462,
       "step": 11610
     },
     {
-      "epoch": 0.41,
-      "learning_rate": 4.965000100269226e-05,
-      "loss": 0.3312,
+      "epoch": 0.4186038130248315,
+      "grad_norm": 0.16751320660114288,
+      "learning_rate": 4.962284672200381e-05,
+      "loss": 0.4459,
       "step": 11615
     },
     {
-      "epoch": 0.41,
-      "learning_rate": 4.964952583240487e-05,
-      "loss": 0.3107,
+      "epoch": 0.41878401268605614,
+      "grad_norm": 0.14165711402893066,
+      "learning_rate": 4.962234158499296e-05,
+      "loss": 0.4635,
       "step": 11620
     },
     {
-      "epoch": 0.41,
-      "learning_rate": 4.9649050342059635e-05,
-      "loss": 0.324,
+      "epoch": 0.41896421234728076,
+      "grad_norm": 0.20206506550312042,
+      "learning_rate": 4.9621836112507475e-05,
+      "loss": 0.487,
       "step": 11625
     },
     {
-      "epoch": 0.41,
-      "learning_rate": 4.9648574531662726e-05,
-      "loss": 0.3329,
+      "epoch": 0.41914441200850544,
+      "grad_norm": 0.20387201011180878,
+      "learning_rate": 4.9621330304554234e-05,
+      "loss": 0.4851,
       "step": 11630
     },
     {
-      "epoch": 0.41,
-      "learning_rate": 4.964809840122032e-05,
-      "loss": 0.3211,
+      "epoch": 0.41932461166973006,
+      "grad_norm": 0.154767245054245,
+      "learning_rate": 4.962082416114014e-05,
+      "loss": 0.4584,
       "step": 11635
     },
     {
-      "epoch": 0.41,
-      "learning_rate": 4.96476219507386e-05,
-      "loss": 0.3055,
+      "epoch": 0.4195048113309547,
+      "grad_norm": 0.1658494919538498,
+      "learning_rate": 4.962031768227208e-05,
+      "loss": 0.4511,
       "step": 11640
     },
     {
-      "epoch": 0.41,
-      "learning_rate": 4.964714518022375e-05,
-      "loss": 0.3302,
+      "epoch": 0.4196850109921793,
+      "grad_norm": 0.16036735475063324,
+      "learning_rate": 4.9619810867956954e-05,
+      "loss": 0.4439,
       "step": 11645
     },
     {
-      "epoch": 0.41,
-      "learning_rate": 4.964666808968197e-05,
-      "loss": 0.3407,
+      "epoch": 0.419865210653404,
+      "grad_norm": 0.18504783511161804,
+      "learning_rate": 4.9619303718201685e-05,
+      "loss": 0.4651,
       "step": 11650
     },
     {
-      "epoch": 0.41,
-      "learning_rate": 4.9646190679119445e-05,
-      "loss": 0.3278,
+      "epoch": 0.4200454103146286,
+      "grad_norm": 0.213784322142601,
+      "learning_rate": 4.9618796233013155e-05,
+      "loss": 0.4465,
       "step": 11655
     },
     {
-      "epoch": 0.41,
-      "learning_rate": 4.964571294854237e-05,
-      "loss": 0.3406,
+      "epoch": 0.42022560997585323,
+      "grad_norm": 0.15321993827819824,
+      "learning_rate": 4.96182884123983e-05,
+      "loss": 0.4583,
       "step": 11660
     },
     {
-      "epoch": 0.41,
-      "learning_rate": 4.964523489795696e-05,
-      "loss": 0.3247,
+      "epoch": 0.4204058096370779,
+      "grad_norm": 0.23537690937519073,
+      "learning_rate": 4.961778025636402e-05,
+      "loss": 0.5085,
       "step": 11665
     },
     {
-      "epoch": 0.41,
-      "learning_rate": 4.964475652736942e-05,
-      "loss": 0.3085,
+      "epoch": 0.42058600929830253,
+      "grad_norm": 0.17639757692813873,
+      "learning_rate": 4.961727176491726e-05,
+      "loss": 0.449,
       "step": 11670
     },
     {
-      "epoch": 0.41,
-      "learning_rate": 4.964427783678595e-05,
-      "loss": 0.3015,
+      "epoch": 0.42076620895952715,
+      "grad_norm": 0.14851510524749756,
+      "learning_rate": 4.9616762938064945e-05,
+      "loss": 0.4373,
       "step": 11675
     },
     {
-      "epoch": 0.41,
-      "learning_rate": 4.964379882621278e-05,
-      "loss": 0.3264,
+      "epoch": 0.4209464086207518,
+      "grad_norm": 0.19301354885101318,
+      "learning_rate": 4.961625377581399e-05,
+      "loss": 0.4564,
       "step": 11680
     },
     {
-      "epoch": 0.41,
-      "learning_rate": 4.964331949565611e-05,
-      "loss": 0.3442,
+      "epoch": 0.42112660828197646,
+      "grad_norm": 0.1511390209197998,
+      "learning_rate": 4.961574427817135e-05,
+      "loss": 0.4522,
       "step": 11685
     },
     {
-      "epoch": 0.41,
-      "learning_rate": 4.9642839845122195e-05,
-      "loss": 0.3321,
+      "epoch": 0.4213068079432011,
+      "grad_norm": 0.16378375887870789,
+      "learning_rate": 4.9615234445143954e-05,
+      "loss": 0.4472,
       "step": 11690
     },
     {
-      "epoch": 0.41,
-      "learning_rate": 4.964235987461724e-05,
-      "loss": 0.3283,
+      "epoch": 0.4214870076044257,
+      "grad_norm": 0.14959801733493805,
+      "learning_rate": 4.961472427673875e-05,
+      "loss": 0.4596,
       "step": 11695
     },
     {
-      "epoch": 0.41,
-      "learning_rate": 4.9641879584147476e-05,
-      "loss": 0.2993,
+      "epoch": 0.4216672072656503,
+      "grad_norm": 0.16105423867702484,
+      "learning_rate": 4.961421377296271e-05,
+      "loss": 0.4472,
       "step": 11700
     },
     {
-      "epoch": 0.41,
-      "learning_rate": 4.964139897371915e-05,
-      "loss": 0.3119,
+      "epoch": 0.421847406926875,
+      "grad_norm": 0.18409523367881775,
+      "learning_rate": 4.9613702933822756e-05,
+      "loss": 0.4599,
       "step": 11705
     },
     {
-      "epoch": 0.41,
-      "learning_rate": 4.964091804333849e-05,
-      "loss": 0.3216,
+      "epoch": 0.4220276065880996,
+      "grad_norm": 0.215390145778656,
+      "learning_rate": 4.961319175932588e-05,
+      "loss": 0.4891,
       "step": 11710
     },
     {
-      "epoch": 0.41,
-      "learning_rate": 4.964043679301176e-05,
-      "loss": 0.313,
+      "epoch": 0.42220780624932425,
+      "grad_norm": 0.22909221053123474,
+      "learning_rate": 4.961268024947902e-05,
+      "loss": 0.4779,
       "step": 11715
     },
     {
-      "epoch": 0.41,
-      "learning_rate": 4.963995522274519e-05,
-      "loss": 0.333,
+      "epoch": 0.42238800591054887,
+      "grad_norm": 0.1938232034444809,
+      "learning_rate": 4.961216840428916e-05,
+      "loss": 0.4564,
       "step": 11720
     },
     {
-      "epoch": 0.41,
-      "learning_rate": 4.963947333254504e-05,
-      "loss": 0.3288,
+      "epoch": 0.42256820557177355,
+      "grad_norm": 0.18986928462982178,
+      "learning_rate": 4.961165622376327e-05,
+      "loss": 0.485,
       "step": 11725
     },
     {
-      "epoch": 0.41,
-      "learning_rate": 4.9638991122417564e-05,
-      "loss": 0.3167,
+      "epoch": 0.42274840523299817,
+      "grad_norm": 0.1635420173406601,
+      "learning_rate": 4.9611143707908336e-05,
+      "loss": 0.4641,
       "step": 11730
     },
     {
-      "epoch": 0.41,
-      "learning_rate": 4.963850859236904e-05,
-      "loss": 0.2891,
+      "epoch": 0.4229286048942228,
+      "grad_norm": 0.2149038463830948,
+      "learning_rate": 4.961063085673132e-05,
+      "loss": 0.5116,
       "step": 11735
     },
     {
-      "epoch": 0.41,
-      "learning_rate": 4.96380257424057e-05,
-      "loss": 0.3347,
+      "epoch": 0.4231088045554474,
+      "grad_norm": 0.15662051737308502,
+      "learning_rate": 4.9610117670239235e-05,
+      "loss": 0.4396,
       "step": 11740
     },
     {
-      "epoch": 0.41,
-      "learning_rate": 4.9637542572533844e-05,
-      "loss": 0.2952,
+      "epoch": 0.4232890042166721,
+      "grad_norm": 0.1819683313369751,
+      "learning_rate": 4.960960414843906e-05,
+      "loss": 0.4357,
       "step": 11745
     },
     {
-      "epoch": 0.41,
-      "learning_rate": 4.963705908275973e-05,
-      "loss": 0.2992,
+      "epoch": 0.4234692038778967,
+      "grad_norm": 0.17882205545902252,
+      "learning_rate": 4.960909029133779e-05,
+      "loss": 0.4215,
       "step": 11750
     },
     {
-      "epoch": 0.41,
-      "learning_rate": 4.9636575273089646e-05,
-      "loss": 0.3329,
+      "epoch": 0.42364940353912134,
+      "grad_norm": 0.1997404247522354,
+      "learning_rate": 4.9608576098942426e-05,
+      "loss": 0.4522,
       "step": 11755
     },
     {
-      "epoch": 0.41,
-      "learning_rate": 4.963609114352986e-05,
-      "loss": 0.2935,
+      "epoch": 0.42382960320034596,
+      "grad_norm": 0.18902002274990082,
+      "learning_rate": 4.960806157125998e-05,
+      "loss": 0.4362,
       "step": 11760
     },
     {
-      "epoch": 0.41,
-      "learning_rate": 4.963560669408668e-05,
-      "loss": 0.3331,
+      "epoch": 0.42400980286157064,
+      "grad_norm": 0.18085581064224243,
+      "learning_rate": 4.960754670829746e-05,
+      "loss": 0.4583,
       "step": 11765
     },
     {
-      "epoch": 0.41,
-      "learning_rate": 4.963512192476637e-05,
-      "loss": 0.3127,
+      "epoch": 0.42419000252279526,
+      "grad_norm": 0.2524605989456177,
+      "learning_rate": 4.960703151006189e-05,
+      "loss": 0.4756,
       "step": 11770
     },
     {
-      "epoch": 0.41,
-      "learning_rate": 4.963463683557524e-05,
-      "loss": 0.3097,
+      "epoch": 0.4243702021840199,
+      "grad_norm": 0.1703946441411972,
+      "learning_rate": 4.9606515976560265e-05,
+      "loss": 0.4533,
       "step": 11775
     },
     {
-      "epoch": 0.41,
-      "learning_rate": 4.9634151426519584e-05,
-      "loss": 0.3086,
+      "epoch": 0.4245504018452445,
+      "grad_norm": 0.16210578382015228,
+      "learning_rate": 4.960600010779963e-05,
+      "loss": 0.4632,
       "step": 11780
     },
     {
-      "epoch": 0.41,
-      "learning_rate": 4.9633665697605707e-05,
-      "loss": 0.3225,
+      "epoch": 0.4247306015064692,
+      "grad_norm": 0.17518621683120728,
+      "learning_rate": 4.9605483903787006e-05,
+      "loss": 0.4919,
       "step": 11785
     },
     {
-      "epoch": 0.41,
-      "learning_rate": 4.9633179648839916e-05,
-      "loss": 0.3424,
+      "epoch": 0.4249108011676938,
+      "grad_norm": 0.15813525021076202,
+      "learning_rate": 4.960496736452943e-05,
+      "loss": 0.4522,
       "step": 11790
     },
     {
-      "epoch": 0.41,
-      "learning_rate": 4.963269328022853e-05,
-      "loss": 0.3198,
+      "epoch": 0.42509100082891843,
+      "grad_norm": 0.19452042877674103,
+      "learning_rate": 4.9604450490033936e-05,
+      "loss": 0.4526,
       "step": 11795
     },
     {
-      "epoch": 0.42,
-      "learning_rate": 4.963220659177784e-05,
-      "loss": 0.3151,
+      "epoch": 0.42527120049014305,
+      "grad_norm": 0.21348360180854797,
+      "learning_rate": 4.960393328030757e-05,
+      "loss": 0.4682,
       "step": 11800
     },
     {
-      "epoch": 0.42,
-      "learning_rate": 4.9631719583494184e-05,
-      "loss": 0.3088,
+      "epoch": 0.42545140015136773,
+      "grad_norm": 0.1495480090379715,
+      "learning_rate": 4.9603415735357374e-05,
+      "loss": 0.4668,
       "step": 11805
     },
     {
-      "epoch": 0.42,
-      "learning_rate": 4.9631232255383884e-05,
-      "loss": 0.3407,
+      "epoch": 0.42563159981259235,
+      "grad_norm": 0.1405077874660492,
+      "learning_rate": 4.96028978551904e-05,
+      "loss": 0.3944,
       "step": 11810
     },
     {
-      "epoch": 0.42,
-      "learning_rate": 4.963074460745327e-05,
-      "loss": 0.3167,
+      "epoch": 0.425811799473817,
+      "grad_norm": 0.18521526455879211,
+      "learning_rate": 4.96023796398137e-05,
+      "loss": 0.4723,
       "step": 11815
     },
     {
-      "epoch": 0.42,
-      "learning_rate": 4.9630256639708675e-05,
-      "loss": 0.3313,
+      "epoch": 0.4259919991350416,
+      "grad_norm": 0.22655785083770752,
+      "learning_rate": 4.9601861089234355e-05,
+      "loss": 0.4484,
       "step": 11820
     },
     {
-      "epoch": 0.42,
-      "learning_rate": 4.962976835215641e-05,
-      "loss": 0.3076,
+      "epoch": 0.4261721987962663,
+      "grad_norm": 0.1640099287033081,
+      "learning_rate": 4.9601342203459405e-05,
+      "loss": 0.4423,
       "step": 11825
     },
     {
-      "epoch": 0.42,
-      "learning_rate": 4.9629279744802845e-05,
-      "loss": 0.3118,
+      "epoch": 0.4263523984574909,
+      "grad_norm": 0.17364855110645294,
+      "learning_rate": 4.960082298249593e-05,
+      "loss": 0.4574,
       "step": 11830
     },
     {
-      "epoch": 0.42,
-      "learning_rate": 4.962879081765432e-05,
-      "loss": 0.3026,
+      "epoch": 0.4265325981187155,
+      "grad_norm": 0.18359048664569855,
+      "learning_rate": 4.9600303426351013e-05,
+      "loss": 0.4274,
       "step": 11835
     },
     {
-      "epoch": 0.42,
-      "learning_rate": 4.9628301570717176e-05,
-      "loss": 0.3041,
+      "epoch": 0.4267127977799402,
+      "grad_norm": 0.1665671169757843,
+      "learning_rate": 4.959978353503172e-05,
+      "loss": 0.4399,
       "step": 11840
     },
     {
-      "epoch": 0.42,
-      "learning_rate": 4.962781200399776e-05,
-      "loss": 0.3037,
+      "epoch": 0.4268929974411648,
+      "grad_norm": 0.1284245401620865,
+      "learning_rate": 4.959926330854514e-05,
+      "loss": 0.4541,
       "step": 11845
     },
     {
-      "epoch": 0.42,
-      "learning_rate": 4.962732211750244e-05,
-      "loss": 0.3158,
+      "epoch": 0.42707319710238945,
+      "grad_norm": 0.19221022725105286,
+      "learning_rate": 4.9598742746898364e-05,
+      "loss": 0.4692,
       "step": 11850
     },
     {
-      "epoch": 0.42,
-      "learning_rate": 4.962683191123757e-05,
-      "loss": 0.307,
+      "epoch": 0.42725339676361407,
+      "grad_norm": 0.21203385293483734,
+      "learning_rate": 4.959822185009847e-05,
+      "loss": 0.4709,
       "step": 11855
     },
     {
-      "epoch": 0.42,
-      "learning_rate": 4.962634138520952e-05,
-      "loss": 0.3302,
+      "epoch": 0.42743359642483875,
+      "grad_norm": 0.15997375547885895,
+      "learning_rate": 4.959770061815258e-05,
+      "loss": 0.4242,
       "step": 11860
     },
     {
-      "epoch": 0.42,
-      "learning_rate": 4.9625850539424644e-05,
-      "loss": 0.3337,
+      "epoch": 0.42761379608606337,
+      "grad_norm": 0.19017674028873444,
+      "learning_rate": 4.959717905106777e-05,
+      "loss": 0.4596,
       "step": 11865
     },
     {
-      "epoch": 0.42,
-      "learning_rate": 4.9625359373889344e-05,
-      "loss": 0.3347,
+      "epoch": 0.427793995747288,
+      "grad_norm": 0.19879359006881714,
+      "learning_rate": 4.9596657148851154e-05,
+      "loss": 0.4626,
       "step": 11870
     },
     {
-      "epoch": 0.42,
-      "learning_rate": 4.962486788860997e-05,
-      "loss": 0.3301,
+      "epoch": 0.4279741954085126,
+      "grad_norm": 0.14506494998931885,
+      "learning_rate": 4.9596134911509865e-05,
+      "loss": 0.4507,
       "step": 11875
     },
     {
-      "epoch": 0.42,
-      "learning_rate": 4.962437608359293e-05,
-      "loss": 0.3024,
+      "epoch": 0.4281543950697373,
+      "grad_norm": 0.16767846047878265,
+      "learning_rate": 4.959561233905098e-05,
+      "loss": 0.4191,
       "step": 11880
     },
     {
-      "epoch": 0.42,
-      "learning_rate": 4.9623883958844574e-05,
-      "loss": 0.3066,
+      "epoch": 0.4283345947309619,
+      "grad_norm": 0.1973155289888382,
+      "learning_rate": 4.9595089431481645e-05,
+      "loss": 0.4705,
       "step": 11885
     },
     {
-      "epoch": 0.42,
-      "learning_rate": 4.9623391514371334e-05,
-      "loss": 0.3476,
+      "epoch": 0.42851479439218654,
+      "grad_norm": 0.16755364835262299,
+      "learning_rate": 4.9594566188808985e-05,
+      "loss": 0.421,
       "step": 11890
     },
     {
-      "epoch": 0.42,
-      "learning_rate": 4.962289875017957e-05,
-      "loss": 0.2938,
+      "epoch": 0.42869499405341116,
+      "grad_norm": 0.18635892868041992,
+      "learning_rate": 4.959404261104012e-05,
+      "loss": 0.4883,
       "step": 11895
     },
     {
-      "epoch": 0.42,
-      "learning_rate": 4.962240566627569e-05,
-      "loss": 0.3341,
+      "epoch": 0.42887519371463584,
+      "grad_norm": 0.18180792033672333,
+      "learning_rate": 4.959351869818218e-05,
+      "loss": 0.4775,
       "step": 11900
     },
     {
-      "epoch": 0.42,
-      "learning_rate": 4.96219122626661e-05,
-      "loss": 0.3252,
+      "epoch": 0.42905539337586046,
+      "grad_norm": 0.1798887848854065,
+      "learning_rate": 4.9592994450242316e-05,
+      "loss": 0.4618,
       "step": 11905
     },
     {
-      "epoch": 0.42,
-      "learning_rate": 4.962141853935721e-05,
-      "loss": 0.3255,
+      "epoch": 0.4292355930370851,
+      "grad_norm": 0.23241207003593445,
+      "learning_rate": 4.9592469867227655e-05,
+      "loss": 0.4479,
       "step": 11910
     },
     {
-      "epoch": 0.42,
-      "learning_rate": 4.962092449635543e-05,
-      "loss": 0.333,
+      "epoch": 0.4294157926983097,
+      "grad_norm": 0.18085531890392303,
+      "learning_rate": 4.959194494914537e-05,
+      "loss": 0.4617,
       "step": 11915
     },
     {
-      "epoch": 0.42,
-      "learning_rate": 4.9620430133667165e-05,
-      "loss": 0.3133,
+      "epoch": 0.4295959923595344,
+      "grad_norm": 0.1726156622171402,
+      "learning_rate": 4.9591419696002575e-05,
+      "loss": 0.4501,
       "step": 11920
     },
     {
-      "epoch": 0.42,
-      "learning_rate": 4.961993545129884e-05,
-      "loss": 0.3345,
+      "epoch": 0.429776192020759,
+      "grad_norm": 0.16227740049362183,
+      "learning_rate": 4.9590894107806454e-05,
+      "loss": 0.4358,
       "step": 11925
     },
     {
-      "epoch": 0.42,
-      "learning_rate": 4.961944044925688e-05,
-      "loss": 0.307,
+      "epoch": 0.42995639168198363,
+      "grad_norm": 0.15402673184871674,
+      "learning_rate": 4.959036818456417e-05,
+      "loss": 0.44,
       "step": 11930
     },
     {
-      "epoch": 0.42,
-      "learning_rate": 4.9618945127547715e-05,
-      "loss": 0.3097,
+      "epoch": 0.43013659134320825,
+      "grad_norm": 0.1438482403755188,
+      "learning_rate": 4.958984192628288e-05,
+      "loss": 0.463,
       "step": 11935
     },
     {
-      "epoch": 0.42,
-      "learning_rate": 4.961844948617777e-05,
-      "loss": 0.3506,
+      "epoch": 0.43031679100443293,
+      "grad_norm": 0.18575410544872284,
+      "learning_rate": 4.958931533296975e-05,
+      "loss": 0.4778,
       "step": 11940
     },
     {
-      "epoch": 0.42,
-      "learning_rate": 4.9617953525153485e-05,
-      "loss": 0.3049,
+      "epoch": 0.43049699066565755,
+      "grad_norm": 0.17367704212665558,
+      "learning_rate": 4.958878840463196e-05,
+      "loss": 0.4159,
       "step": 11945
     },
     {
-      "epoch": 0.42,
-      "learning_rate": 4.9617457244481294e-05,
-      "loss": 0.3365,
+      "epoch": 0.4306771903268822,
+      "grad_norm": 0.1562519371509552,
+      "learning_rate": 4.958826114127668e-05,
+      "loss": 0.4685,
       "step": 11950
     },
     {
-      "epoch": 0.42,
-      "learning_rate": 4.961696064416764e-05,
-      "loss": 0.3425,
+      "epoch": 0.4308573899881068,
+      "grad_norm": 0.13806965947151184,
+      "learning_rate": 4.958773354291111e-05,
+      "loss": 0.4313,
       "step": 11955
     },
     {
-      "epoch": 0.42,
-      "learning_rate": 4.961646372421898e-05,
-      "loss": 0.3385,
+      "epoch": 0.4310375896493315,
+      "grad_norm": 0.19402550160884857,
+      "learning_rate": 4.958720560954243e-05,
+      "loss": 0.4232,
       "step": 11960
     },
     {
-      "epoch": 0.42,
-      "learning_rate": 4.961596648464176e-05,
-      "loss": 0.3063,
+      "epoch": 0.4312177893105561,
+      "grad_norm": 0.18233174085617065,
+      "learning_rate": 4.958667734117784e-05,
+      "loss": 0.4742,
       "step": 11965
     },
     {
-      "epoch": 0.42,
-      "learning_rate": 4.9615468925442433e-05,
-      "loss": 0.306,
+      "epoch": 0.4313979889717807,
+      "grad_norm": 0.16293558478355408,
+      "learning_rate": 4.958614873782452e-05,
+      "loss": 0.4507,
       "step": 11970
     },
     {
-      "epoch": 0.42,
-      "learning_rate": 4.961497104662747e-05,
-      "loss": 0.3161,
+      "epoch": 0.43157818863300534,
+      "grad_norm": 0.23064178228378296,
+      "learning_rate": 4.95856197994897e-05,
+      "loss": 0.4566,
       "step": 11975
     },
     {
-      "epoch": 0.42,
-      "learning_rate": 4.961447284820333e-05,
-      "loss": 0.3016,
+      "epoch": 0.43175838829423,
+      "grad_norm": 0.14595326781272888,
+      "learning_rate": 4.958509052618055e-05,
+      "loss": 0.4589,
       "step": 11980
     },
     {
-      "epoch": 0.42,
-      "learning_rate": 4.9613974330176484e-05,
-      "loss": 0.3038,
+      "epoch": 0.43193858795545464,
+      "grad_norm": 0.13431671261787415,
+      "learning_rate": 4.958456091790431e-05,
+      "loss": 0.4459,
       "step": 11985
     },
     {
-      "epoch": 0.42,
-      "learning_rate": 4.96134754925534e-05,
-      "loss": 0.3315,
+      "epoch": 0.43211878761667927,
+      "grad_norm": 0.17889688909053802,
+      "learning_rate": 4.9584030974668195e-05,
+      "loss": 0.441,
       "step": 11990
     },
     {
-      "epoch": 0.42,
-      "learning_rate": 4.9612976335340556e-05,
-      "loss": 0.3475,
+      "epoch": 0.4322989872779039,
+      "grad_norm": 0.18637695908546448,
+      "learning_rate": 4.958350069647941e-05,
+      "loss": 0.4372,
       "step": 11995
     },
     {
-      "epoch": 0.42,
-      "learning_rate": 4.961247685854443e-05,
-      "loss": 0.3028,
+      "epoch": 0.43247918693912857,
+      "grad_norm": 0.20668061077594757,
+      "learning_rate": 4.958297008334519e-05,
+      "loss": 0.4777,
       "step": 12000
     },
     {
-      "epoch": 0.42,
-      "eval_loss": 0.31634584069252014,
-      "eval_runtime": 10.5394,
-      "eval_samples_per_second": 9.488,
-      "eval_steps_per_second": 9.488,
+      "epoch": 0.43247918693912857,
+      "eval_loss": 0.474940687417984,
+      "eval_runtime": 3.5638,
+      "eval_samples_per_second": 28.06,
+      "eval_steps_per_second": 7.015,
       "step": 12000
     },
     {
-      "epoch": 0.42,
-      "learning_rate": 4.961197706217152e-05,
-      "loss": 0.3324,
+      "epoch": 0.4326593866003532,
+      "grad_norm": 0.2014397382736206,
+      "learning_rate": 4.958243913527276e-05,
+      "loss": 0.4991,
       "step": 12005
     },
     {
-      "epoch": 0.42,
-      "learning_rate": 4.961147694622831e-05,
-      "loss": 0.3144,
+      "epoch": 0.4328395862615778,
+      "grad_norm": 0.17615807056427002,
+      "learning_rate": 4.958190785226936e-05,
+      "loss": 0.47,
       "step": 12010
     },
     {
-      "epoch": 0.42,
-      "learning_rate": 4.961097651072128e-05,
-      "loss": 0.3181,
+      "epoch": 0.4330197859228025,
+      "grad_norm": 0.20073723793029785,
+      "learning_rate": 4.958137623434222e-05,
+      "loss": 0.4421,
       "step": 12015
     },
     {
-      "epoch": 0.42,
-      "learning_rate": 4.961047575565695e-05,
-      "loss": 0.3045,
+      "epoch": 0.4331999855840271,
+      "grad_norm": 0.15735125541687012,
+      "learning_rate": 4.958084428149859e-05,
+      "loss": 0.4472,
       "step": 12020
     },
     {
-      "epoch": 0.42,
-      "learning_rate": 4.96099746810418e-05,
-      "loss": 0.3209,
+      "epoch": 0.43338018524525174,
+      "grad_norm": 0.21121670305728912,
+      "learning_rate": 4.9580311993745715e-05,
+      "loss": 0.451,
       "step": 12025
     },
     {
-      "epoch": 0.42,
-      "learning_rate": 4.960947328688236e-05,
-      "loss": 0.3149,
+      "epoch": 0.43356038490647636,
+      "grad_norm": 0.17626595497131348,
+      "learning_rate": 4.957977937109085e-05,
+      "loss": 0.4332,
       "step": 12030
     },
     {
-      "epoch": 0.42,
-      "learning_rate": 4.960897157318512e-05,
-      "loss": 0.3207,
+      "epoch": 0.43374058456770104,
+      "grad_norm": 0.18630550801753998,
+      "learning_rate": 4.9579246413541245e-05,
+      "loss": 0.4434,
       "step": 12035
     },
     {
-      "epoch": 0.42,
-      "learning_rate": 4.96084695399566e-05,
-      "loss": 0.3323,
+      "epoch": 0.43392078422892566,
+      "grad_norm": 0.17722322046756744,
+      "learning_rate": 4.957871312110417e-05,
+      "loss": 0.4437,
       "step": 12040
     },
     {
-      "epoch": 0.42,
-      "learning_rate": 4.960796718720333e-05,
-      "loss": 0.3058,
+      "epoch": 0.4341009838901503,
+      "grad_norm": 0.1969458907842636,
+      "learning_rate": 4.9578179493786884e-05,
+      "loss": 0.4555,
       "step": 12045
     },
     {
-      "epoch": 0.42,
-      "learning_rate": 4.9607464514931824e-05,
-      "loss": 0.3216,
+      "epoch": 0.4342811835513749,
+      "grad_norm": 0.19826200604438782,
+      "learning_rate": 4.9577645531596666e-05,
+      "loss": 0.4186,
       "step": 12050
     },
     {
-      "epoch": 0.42,
-      "learning_rate": 4.9606961523148597e-05,
-      "loss": 0.3235,
+      "epoch": 0.4344613832125996,
+      "grad_norm": 0.2032472789287567,
+      "learning_rate": 4.957711123454079e-05,
+      "loss": 0.4377,
       "step": 12055
     },
     {
-      "epoch": 0.42,
-      "learning_rate": 4.96064582118602e-05,
-      "loss": 0.3093,
+      "epoch": 0.4346415828738242,
+      "grad_norm": 0.20564451813697815,
+      "learning_rate": 4.957657660262652e-05,
+      "loss": 0.4529,
       "step": 12060
     },
     {
-      "epoch": 0.42,
-      "learning_rate": 4.960595458107315e-05,
-      "loss": 0.3025,
+      "epoch": 0.43482178253504883,
+      "grad_norm": 0.2119935005903244,
+      "learning_rate": 4.957604163586116e-05,
+      "loss": 0.4644,
       "step": 12065
     },
     {
-      "epoch": 0.42,
-      "learning_rate": 4.9605450630794e-05,
-      "loss": 0.3243,
+      "epoch": 0.43500198219627345,
+      "grad_norm": 0.15134219825267792,
+      "learning_rate": 4.9575506334251984e-05,
+      "loss": 0.4497,
       "step": 12070
     },
     {
-      "epoch": 0.42,
-      "learning_rate": 4.960494636102929e-05,
-      "loss": 0.334,
+      "epoch": 0.43518218185749813,
+      "grad_norm": 0.19318123161792755,
+      "learning_rate": 4.95749706978063e-05,
+      "loss": 0.4525,
       "step": 12075
     },
     {
-      "epoch": 0.43,
-      "learning_rate": 4.960444177178557e-05,
-      "loss": 0.336,
+      "epoch": 0.43536238151872275,
+      "grad_norm": 0.19543355703353882,
+      "learning_rate": 4.9574434726531395e-05,
+      "loss": 0.4887,
       "step": 12080
     },
     {
-      "epoch": 0.43,
-      "learning_rate": 4.960393686306939e-05,
-      "loss": 0.3214,
+      "epoch": 0.4355425811799474,
+      "grad_norm": 0.17110168933868408,
+      "learning_rate": 4.957389842043457e-05,
+      "loss": 0.4431,
       "step": 12085
     },
     {
-      "epoch": 0.43,
-      "learning_rate": 4.9603431634887296e-05,
-      "loss": 0.3246,
+      "epoch": 0.435722780841172,
+      "grad_norm": 0.18978413939476013,
+      "learning_rate": 4.957336177952314e-05,
+      "loss": 0.46,
       "step": 12090
     },
     {
-      "epoch": 0.43,
-      "learning_rate": 4.960292608724587e-05,
-      "loss": 0.3268,
+      "epoch": 0.4359029805023967,
+      "grad_norm": 0.22278402745723724,
+      "learning_rate": 4.957282480380442e-05,
+      "loss": 0.4595,
       "step": 12095
     },
     {
-      "epoch": 0.43,
-      "learning_rate": 4.960242022015165e-05,
-      "loss": 0.3246,
+      "epoch": 0.4360831801636213,
+      "grad_norm": 0.17783726751804352,
+      "learning_rate": 4.957228749328571e-05,
+      "loss": 0.4289,
       "step": 12100
     },
     {
-      "epoch": 0.43,
-      "learning_rate": 4.960191403361122e-05,
-      "loss": 0.3349,
+      "epoch": 0.4362633798248459,
+      "grad_norm": 0.2311362475156784,
+      "learning_rate": 4.957174984797434e-05,
+      "loss": 0.4326,
       "step": 12105
     },
     {
-      "epoch": 0.43,
-      "learning_rate": 4.960140752763115e-05,
-      "loss": 0.3367,
+      "epoch": 0.43644357948607054,
+      "grad_norm": 0.1933164894580841,
+      "learning_rate": 4.957121186787764e-05,
+      "loss": 0.4614,
       "step": 12110
     },
     {
-      "epoch": 0.43,
-      "learning_rate": 4.9600900702218016e-05,
-      "loss": 0.3581,
+      "epoch": 0.4366237791472952,
+      "grad_norm": 0.1860789954662323,
+      "learning_rate": 4.957067355300293e-05,
+      "loss": 0.4762,
       "step": 12115
     },
     {
-      "epoch": 0.43,
-      "learning_rate": 4.96003935573784e-05,
-      "loss": 0.3233,
+      "epoch": 0.43680397880851984,
+      "grad_norm": 0.18945522606372833,
+      "learning_rate": 4.9570134903357556e-05,
+      "loss": 0.4207,
       "step": 12120
     },
     {
-      "epoch": 0.43,
-      "learning_rate": 4.9599886093118887e-05,
-      "loss": 0.3331,
+      "epoch": 0.43698417846974447,
+      "grad_norm": 0.21275950968265533,
+      "learning_rate": 4.956959591894885e-05,
+      "loss": 0.4836,
       "step": 12125
     },
     {
-      "epoch": 0.43,
-      "learning_rate": 4.9599378309446063e-05,
-      "loss": 0.3189,
+      "epoch": 0.4371643781309691,
+      "grad_norm": 0.18478412926197052,
+      "learning_rate": 4.956905659978416e-05,
+      "loss": 0.4406,
       "step": 12130
     },
     {
-      "epoch": 0.43,
-      "learning_rate": 4.959887020636652e-05,
-      "loss": 0.3251,
+      "epoch": 0.43734457779219377,
+      "grad_norm": 0.16761377453804016,
+      "learning_rate": 4.9568516945870825e-05,
+      "loss": 0.4556,
       "step": 12135
     },
     {
-      "epoch": 0.43,
-      "learning_rate": 4.959836178388687e-05,
-      "loss": 0.3252,
+      "epoch": 0.4375247774534184,
+      "grad_norm": 0.14763802289962769,
+      "learning_rate": 4.9567976957216204e-05,
+      "loss": 0.4445,
       "step": 12140
     },
     {
-      "epoch": 0.43,
-      "learning_rate": 4.9597853042013694e-05,
-      "loss": 0.3224,
+      "epoch": 0.437704977114643,
+      "grad_norm": 0.1819683164358139,
+      "learning_rate": 4.956743663382766e-05,
+      "loss": 0.4411,
       "step": 12145
     },
     {
-      "epoch": 0.43,
-      "learning_rate": 4.959734398075361e-05,
-      "loss": 0.3192,
+      "epoch": 0.43788517677586763,
+      "grad_norm": 0.1549748033285141,
+      "learning_rate": 4.9566895975712533e-05,
+      "loss": 0.444,
       "step": 12150
     },
     {
-      "epoch": 0.43,
-      "learning_rate": 4.959683460011322e-05,
-      "loss": 0.3215,
+      "epoch": 0.4380653764370923,
+      "grad_norm": 0.16789652407169342,
+      "learning_rate": 4.9566354982878215e-05,
+      "loss": 0.4636,
       "step": 12155
     },
     {
-      "epoch": 0.43,
-      "learning_rate": 4.9596324900099145e-05,
-      "loss": 0.3254,
+      "epoch": 0.43824557609831694,
+      "grad_norm": 0.15570096671581268,
+      "learning_rate": 4.956581365533207e-05,
+      "loss": 0.4223,
       "step": 12160
     },
     {
-      "epoch": 0.43,
-      "learning_rate": 4.9595814880718e-05,
-      "loss": 0.3287,
+      "epoch": 0.43842577575954156,
+      "grad_norm": 0.20014168322086334,
+      "learning_rate": 4.956527199308146e-05,
+      "loss": 0.429,
       "step": 12165
     },
     {
-      "epoch": 0.43,
-      "learning_rate": 4.959530454197641e-05,
-      "loss": 0.3454,
+      "epoch": 0.43860597542076624,
+      "grad_norm": 0.2379717230796814,
+      "learning_rate": 4.956472999613379e-05,
+      "loss": 0.4862,
       "step": 12170
     },
     {
-      "epoch": 0.43,
-      "learning_rate": 4.9594793883881e-05,
-      "loss": 0.3208,
+      "epoch": 0.43878617508199086,
+      "grad_norm": 0.1970072239637375,
+      "learning_rate": 4.956418766449642e-05,
+      "loss": 0.4299,
       "step": 12175
     },
     {
-      "epoch": 0.43,
-      "learning_rate": 4.95942829064384e-05,
-      "loss": 0.3358,
+      "epoch": 0.4389663747432155,
+      "grad_norm": 0.19804427027702332,
+      "learning_rate": 4.956364499817674e-05,
+      "loss": 0.4662,
       "step": 12180
     },
     {
-      "epoch": 0.43,
-      "learning_rate": 4.9593771609655244e-05,
-      "loss": 0.3224,
+      "epoch": 0.4391465744044401,
+      "grad_norm": 0.18238474428653717,
+      "learning_rate": 4.956310199718217e-05,
+      "loss": 0.4628,
       "step": 12185
     },
     {
-      "epoch": 0.43,
-      "learning_rate": 4.9593259993538164e-05,
-      "loss": 0.3418,
+      "epoch": 0.4393267740656648,
+      "grad_norm": 0.1626834273338318,
+      "learning_rate": 4.956255866152008e-05,
+      "loss": 0.445,
       "step": 12190
     },
     {
-      "epoch": 0.43,
-      "learning_rate": 4.9592748058093824e-05,
-      "loss": 0.329,
+      "epoch": 0.4395069737268894,
+      "grad_norm": 0.18369080126285553,
+      "learning_rate": 4.956201499119788e-05,
+      "loss": 0.4526,
       "step": 12195
     },
     {
-      "epoch": 0.43,
-      "learning_rate": 4.959223580332885e-05,
-      "loss": 0.3275,
+      "epoch": 0.439687173388114,
+      "grad_norm": 0.1799646019935608,
+      "learning_rate": 4.956147098622299e-05,
+      "loss": 0.4295,
       "step": 12200
     },
     {
-      "epoch": 0.43,
-      "learning_rate": 4.9591723229249905e-05,
-      "loss": 0.3173,
+      "epoch": 0.43986737304933865,
+      "grad_norm": 0.13832755386829376,
+      "learning_rate": 4.9560926646602813e-05,
+      "loss": 0.4452,
       "step": 12205
     },
     {
-      "epoch": 0.43,
-      "learning_rate": 4.9591210335863635e-05,
-      "loss": 0.3148,
+      "epoch": 0.44004757271056333,
+      "grad_norm": 0.1626981645822525,
+      "learning_rate": 4.9560381972344765e-05,
+      "loss": 0.4387,
       "step": 12210
     },
     {
-      "epoch": 0.43,
-      "learning_rate": 4.959069712317671e-05,
-      "loss": 0.3237,
+      "epoch": 0.44022777237178795,
+      "grad_norm": 0.15181361138820648,
+      "learning_rate": 4.955983696345626e-05,
+      "loss": 0.4615,
       "step": 12215
     },
     {
-      "epoch": 0.43,
-      "learning_rate": 4.9590183591195785e-05,
-      "loss": 0.3236,
+      "epoch": 0.4404079720330126,
+      "grad_norm": 0.1977403461933136,
+      "learning_rate": 4.955929161994474e-05,
+      "loss": 0.429,
       "step": 12220
     },
     {
-      "epoch": 0.43,
-      "learning_rate": 4.9589669739927534e-05,
-      "loss": 0.3257,
+      "epoch": 0.4405881716942372,
+      "grad_norm": 0.1452832669019699,
+      "learning_rate": 4.955874594181763e-05,
+      "loss": 0.4529,
       "step": 12225
     },
     {
-      "epoch": 0.43,
-      "learning_rate": 4.958915556937862e-05,
-      "loss": 0.314,
+      "epoch": 0.4407683713554619,
+      "grad_norm": 0.15825548768043518,
+      "learning_rate": 4.955819992908235e-05,
+      "loss": 0.4748,
       "step": 12230
     },
     {
-      "epoch": 0.43,
-      "learning_rate": 4.9588641079555734e-05,
-      "loss": 0.3264,
+      "epoch": 0.4409485710166865,
+      "grad_norm": 0.17782393097877502,
+      "learning_rate": 4.9557653581746355e-05,
+      "loss": 0.4872,
       "step": 12235
     },
     {
-      "epoch": 0.43,
-      "learning_rate": 4.958812627046555e-05,
-      "loss": 0.341,
+      "epoch": 0.4411287706779111,
+      "grad_norm": 0.18894019722938538,
+      "learning_rate": 4.955710689981708e-05,
+      "loss": 0.4215,
       "step": 12240
     },
     {
-      "epoch": 0.43,
-      "learning_rate": 4.9587611142114744e-05,
-      "loss": 0.3338,
+      "epoch": 0.44130897033913574,
+      "grad_norm": 0.18643000721931458,
+      "learning_rate": 4.955655988330199e-05,
+      "loss": 0.4314,
       "step": 12245
     },
     {
-      "epoch": 0.43,
-      "learning_rate": 4.958709569451001e-05,
-      "loss": 0.3163,
+      "epoch": 0.4414891700003604,
+      "grad_norm": 0.17689816653728485,
+      "learning_rate": 4.955601253220852e-05,
+      "loss": 0.4323,
       "step": 12250
     },
     {
-      "epoch": 0.43,
-      "learning_rate": 4.958657992765805e-05,
-      "loss": 0.317,
+      "epoch": 0.44166936966158504,
+      "grad_norm": 0.19838203489780426,
+      "learning_rate": 4.955546484654413e-05,
+      "loss": 0.47,
       "step": 12255
     },
     {
-      "epoch": 0.43,
-      "learning_rate": 4.958606384156554e-05,
-      "loss": 0.3226,
+      "epoch": 0.44184956932280967,
+      "grad_norm": 0.17451122403144836,
+      "learning_rate": 4.955491682631629e-05,
+      "loss": 0.4202,
       "step": 12260
     },
     {
-      "epoch": 0.43,
-      "learning_rate": 4.958554743623921e-05,
-      "loss": 0.33,
+      "epoch": 0.4420297689840343,
+      "grad_norm": 0.16732750833034515,
+      "learning_rate": 4.955436847153246e-05,
+      "loss": 0.4592,
       "step": 12265
     },
     {
-      "epoch": 0.43,
-      "learning_rate": 4.9585030711685745e-05,
-      "loss": 0.318,
+      "epoch": 0.44220996864525897,
+      "grad_norm": 0.17255975306034088,
+      "learning_rate": 4.955381978220011e-05,
+      "loss": 0.4607,
       "step": 12270
     },
     {
-      "epoch": 0.43,
-      "learning_rate": 4.958451366791185e-05,
-      "loss": 0.3325,
+      "epoch": 0.4423901683064836,
+      "grad_norm": 0.20615766942501068,
+      "learning_rate": 4.955327075832672e-05,
+      "loss": 0.4488,
       "step": 12275
     },
     {
-      "epoch": 0.43,
-      "learning_rate": 4.958399630492425e-05,
-      "loss": 0.3203,
+      "epoch": 0.4425703679677082,
+      "grad_norm": 0.16766251623630524,
+      "learning_rate": 4.955272139991978e-05,
+      "loss": 0.4407,
       "step": 12280
     },
     {
-      "epoch": 0.43,
-      "learning_rate": 4.958347862272966e-05,
-      "loss": 0.3243,
+      "epoch": 0.44275056762893283,
+      "grad_norm": 0.16281658411026,
+      "learning_rate": 4.955217170698675e-05,
+      "loss": 0.4417,
       "step": 12285
     },
     {
-      "epoch": 0.43,
-      "learning_rate": 4.9582960621334805e-05,
-      "loss": 0.355,
+      "epoch": 0.4429307672901575,
+      "grad_norm": 0.1925809234380722,
+      "learning_rate": 4.955162167953514e-05,
+      "loss": 0.4254,
       "step": 12290
     },
     {
-      "epoch": 0.43,
-      "learning_rate": 4.95824423007464e-05,
-      "loss": 0.3308,
+      "epoch": 0.44311096695138213,
+      "grad_norm": 0.15888631343841553,
+      "learning_rate": 4.955107131757244e-05,
+      "loss": 0.4305,
       "step": 12295
     },
     {
-      "epoch": 0.43,
-      "learning_rate": 4.9581923660971186e-05,
-      "loss": 0.3271,
+      "epoch": 0.44329116661260676,
+      "grad_norm": 0.153702974319458,
+      "learning_rate": 4.955052062110615e-05,
+      "loss": 0.4306,
       "step": 12300
     },
     {
-      "epoch": 0.43,
-      "learning_rate": 4.9581404702015896e-05,
-      "loss": 0.3112,
+      "epoch": 0.4434713662738314,
+      "grad_norm": 0.190146803855896,
+      "learning_rate": 4.9549969590143765e-05,
+      "loss": 0.4711,
       "step": 12305
     },
     {
-      "epoch": 0.43,
-      "learning_rate": 4.958088542388726e-05,
-      "loss": 0.3161,
+      "epoch": 0.44365156593505606,
+      "grad_norm": 0.150401771068573,
+      "learning_rate": 4.9549418224692795e-05,
+      "loss": 0.4323,
       "step": 12310
     },
     {
-      "epoch": 0.43,
-      "learning_rate": 4.958036582659204e-05,
-      "loss": 0.3227,
+      "epoch": 0.4438317655962807,
+      "grad_norm": 0.17060506343841553,
+      "learning_rate": 4.954886652476076e-05,
+      "loss": 0.4777,
       "step": 12315
     },
     {
-      "epoch": 0.43,
-      "learning_rate": 4.9579845910136955e-05,
-      "loss": 0.3271,
+      "epoch": 0.4440119652575053,
+      "grad_norm": 0.16213923692703247,
+      "learning_rate": 4.9548314490355165e-05,
+      "loss": 0.4736,
       "step": 12320
     },
     {
-      "epoch": 0.43,
-      "learning_rate": 4.957932567452876e-05,
-      "loss": 0.3227,
+      "epoch": 0.4441921649187299,
+      "grad_norm": 0.17394715547561646,
+      "learning_rate": 4.954776212148354e-05,
+      "loss": 0.4731,
       "step": 12325
     },
     {
-      "epoch": 0.43,
-      "learning_rate": 4.957880511977424e-05,
-      "loss": 0.3203,
+      "epoch": 0.4443723645799546,
+      "grad_norm": 0.16405607759952545,
+      "learning_rate": 4.95472094181534e-05,
+      "loss": 0.42,
       "step": 12330
     },
     {
-      "epoch": 0.43,
-      "learning_rate": 4.957828424588012e-05,
-      "loss": 0.3467,
+      "epoch": 0.4445525642411792,
+      "grad_norm": 0.13627347350120544,
+      "learning_rate": 4.9546656380372306e-05,
+      "loss": 0.4613,
       "step": 12335
     },
     {
-      "epoch": 0.43,
-      "learning_rate": 4.9577763052853186e-05,
-      "loss": 0.3006,
+      "epoch": 0.44473276390240385,
+      "grad_norm": 0.19392867386341095,
+      "learning_rate": 4.954610300814776e-05,
+      "loss": 0.4315,
       "step": 12340
     },
     {
-      "epoch": 0.43,
-      "learning_rate": 4.957724154070019e-05,
-      "loss": 0.3177,
+      "epoch": 0.4449129635636285,
+      "grad_norm": 0.14533652365207672,
+      "learning_rate": 4.9545549301487306e-05,
+      "loss": 0.467,
       "step": 12345
     },
     {
-      "epoch": 0.43,
-      "learning_rate": 4.957671970942792e-05,
-      "loss": 0.3384,
+      "epoch": 0.44509316322485315,
+      "grad_norm": 0.17351950705051422,
+      "learning_rate": 4.95449952603985e-05,
+      "loss": 0.432,
       "step": 12350
     },
     {
-      "epoch": 0.43,
-      "learning_rate": 4.9576197559043134e-05,
-      "loss": 0.3048,
+      "epoch": 0.4452733628860778,
+      "grad_norm": 0.1487894058227539,
+      "learning_rate": 4.9544440884888885e-05,
+      "loss": 0.4856,
       "step": 12355
     },
     {
-      "epoch": 0.43,
-      "learning_rate": 4.957567508955262e-05,
-      "loss": 0.3106,
+      "epoch": 0.4454535625473024,
+      "grad_norm": 0.16079506278038025,
+      "learning_rate": 4.954388617496602e-05,
+      "loss": 0.4793,
       "step": 12360
     },
     {
-      "epoch": 0.44,
-      "learning_rate": 4.9575152300963154e-05,
-      "loss": 0.2756,
+      "epoch": 0.4456337622085271,
+      "grad_norm": 0.16144710779190063,
+      "learning_rate": 4.954333113063745e-05,
+      "loss": 0.4313,
       "step": 12365
     },
     {
-      "epoch": 0.44,
-      "learning_rate": 4.957462919328154e-05,
-      "loss": 0.3165,
+      "epoch": 0.4458139618697517,
+      "grad_norm": 0.1664080023765564,
+      "learning_rate": 4.9542775751910755e-05,
+      "loss": 0.4207,
       "step": 12370
     },
     {
-      "epoch": 0.44,
-      "learning_rate": 4.9574105766514555e-05,
-      "loss": 0.3283,
+      "epoch": 0.4459941615309763,
+      "grad_norm": 0.16213445365428925,
+      "learning_rate": 4.954222003879349e-05,
+      "loss": 0.4896,
       "step": 12375
     },
     {
-      "epoch": 0.44,
-      "learning_rate": 4.9573582020669007e-05,
-      "loss": 0.3237,
+      "epoch": 0.44617436119220094,
+      "grad_norm": 0.15756379067897797,
+      "learning_rate": 4.954166399129322e-05,
+      "loss": 0.4438,
       "step": 12380
     },
     {
-      "epoch": 0.44,
-      "learning_rate": 4.957305795575169e-05,
-      "loss": 0.3246,
+      "epoch": 0.4463545608534256,
+      "grad_norm": 0.21012753248214722,
+      "learning_rate": 4.954110760941754e-05,
+      "loss": 0.4689,
       "step": 12385
     },
     {
-      "epoch": 0.44,
-      "learning_rate": 4.957253357176941e-05,
-      "loss": 0.3304,
+      "epoch": 0.44653476051465024,
+      "grad_norm": 0.1980627030134201,
+      "learning_rate": 4.954055089317401e-05,
+      "loss": 0.4605,
       "step": 12390
     },
     {
-      "epoch": 0.44,
-      "learning_rate": 4.957200886872897e-05,
-      "loss": 0.3393,
+      "epoch": 0.44671496017587486,
+      "grad_norm": 0.19075866043567657,
+      "learning_rate": 4.9539993842570226e-05,
+      "loss": 0.4795,
       "step": 12395
     },
     {
-      "epoch": 0.44,
-      "learning_rate": 4.957148384663719e-05,
-      "loss": 0.307,
+      "epoch": 0.4468951598370995,
+      "grad_norm": 0.19187109172344208,
+      "learning_rate": 4.953943645761378e-05,
+      "loss": 0.4509,
       "step": 12400
     },
     {
-      "epoch": 0.44,
-      "learning_rate": 4.957095850550089e-05,
-      "loss": 0.3274,
+      "epoch": 0.44707535949832417,
+      "grad_norm": 0.16127610206604004,
+      "learning_rate": 4.9538878738312265e-05,
+      "loss": 0.4858,
       "step": 12405
     },
     {
-      "epoch": 0.44,
-      "learning_rate": 4.9570432845326884e-05,
-      "loss": 0.3076,
+      "epoch": 0.4472555591595488,
+      "grad_norm": 0.1766381561756134,
+      "learning_rate": 4.953832068467328e-05,
+      "loss": 0.4753,
       "step": 12410
     },
     {
-      "epoch": 0.44,
-      "learning_rate": 4.9569906866122e-05,
-      "loss": 0.3041,
+      "epoch": 0.4474357588207734,
+      "grad_norm": 0.16287901997566223,
+      "learning_rate": 4.953776229670442e-05,
+      "loss": 0.4756,
       "step": 12415
     },
     {
-      "epoch": 0.44,
-      "learning_rate": 4.956938056789306e-05,
-      "loss": 0.309,
+      "epoch": 0.44761595848199803,
+      "grad_norm": 0.15042343735694885,
+      "learning_rate": 4.9537203574413305e-05,
+      "loss": 0.408,
       "step": 12420
     },
     {
-      "epoch": 0.44,
-      "learning_rate": 4.956885395064692e-05,
-      "loss": 0.3154,
+      "epoch": 0.4477961581432227,
+      "grad_norm": 0.18103763461112976,
+      "learning_rate": 4.953664451780754e-05,
+      "loss": 0.4455,
       "step": 12425
     },
     {
-      "epoch": 0.44,
-      "learning_rate": 4.956832701439039e-05,
-      "loss": 0.3421,
+      "epoch": 0.44797635780444733,
+      "grad_norm": 0.23670852184295654,
+      "learning_rate": 4.953608512689474e-05,
+      "loss": 0.4823,
       "step": 12430
     },
     {
-      "epoch": 0.44,
-      "learning_rate": 4.9567799759130335e-05,
-      "loss": 0.3171,
+      "epoch": 0.44815655746567196,
+      "grad_norm": 0.21633177995681763,
+      "learning_rate": 4.9535525401682535e-05,
+      "loss": 0.4432,
       "step": 12435
     },
     {
-      "epoch": 0.44,
-      "learning_rate": 4.9567272184873586e-05,
-      "loss": 0.3307,
+      "epoch": 0.4483367571268966,
+      "grad_norm": 0.178372323513031,
+      "learning_rate": 4.9534965342178546e-05,
+      "loss": 0.4945,
       "step": 12440
     },
     {
-      "epoch": 0.44,
-      "learning_rate": 4.9566744291627e-05,
-      "loss": 0.3093,
+      "epoch": 0.44851695678812126,
+      "grad_norm": 0.2000657618045807,
+      "learning_rate": 4.95344049483904e-05,
+      "loss": 0.5071,
       "step": 12445
     },
     {
-      "epoch": 0.44,
-      "learning_rate": 4.956621607939742e-05,
-      "loss": 0.2988,
+      "epoch": 0.4486971564493459,
+      "grad_norm": 0.15917164087295532,
+      "learning_rate": 4.953384422032574e-05,
+      "loss": 0.4075,
       "step": 12450
     },
     {
-      "epoch": 0.44,
-      "learning_rate": 4.956568754819173e-05,
-      "loss": 0.2914,
+      "epoch": 0.4488773561105705,
+      "grad_norm": 0.14177782833576202,
+      "learning_rate": 4.9533283157992206e-05,
+      "loss": 0.4403,
       "step": 12455
     },
     {
-      "epoch": 0.44,
-      "learning_rate": 4.956515869801677e-05,
-      "loss": 0.3039,
+      "epoch": 0.4490575557717951,
+      "grad_norm": 0.17211756110191345,
+      "learning_rate": 4.953272176139744e-05,
+      "loss": 0.4176,
       "step": 12460
     },
     {
-      "epoch": 0.44,
-      "learning_rate": 4.9564629528879406e-05,
-      "loss": 0.3196,
+      "epoch": 0.4492377554330198,
+      "grad_norm": 0.14136864244937897,
+      "learning_rate": 4.953216003054908e-05,
+      "loss": 0.4492,
       "step": 12465
     },
     {
-      "epoch": 0.44,
-      "learning_rate": 4.956410004078652e-05,
-      "loss": 0.2953,
+      "epoch": 0.4494179550942444,
+      "grad_norm": 0.19541148841381073,
+      "learning_rate": 4.95315979654548e-05,
+      "loss": 0.4768,
       "step": 12470
     },
     {
-      "epoch": 0.44,
-      "learning_rate": 4.9563570233744995e-05,
-      "loss": 0.2914,
+      "epoch": 0.44959815475546905,
+      "grad_norm": 0.15044446289539337,
+      "learning_rate": 4.953103556612224e-05,
+      "loss": 0.4469,
       "step": 12475
     },
     {
-      "epoch": 0.44,
-      "learning_rate": 4.956304010776168e-05,
-      "loss": 0.302,
+      "epoch": 0.44977835441669367,
+      "grad_norm": 0.20979291200637817,
+      "learning_rate": 4.953047283255907e-05,
+      "loss": 0.4432,
       "step": 12480
     },
     {
-      "epoch": 0.44,
-      "learning_rate": 4.9562509662843495e-05,
-      "loss": 0.3218,
+      "epoch": 0.44995855407791835,
+      "grad_norm": 0.20431764423847198,
+      "learning_rate": 4.9529909764772956e-05,
+      "loss": 0.4589,
       "step": 12485
     },
     {
-      "epoch": 0.44,
-      "learning_rate": 4.95619788989973e-05,
-      "loss": 0.3176,
+      "epoch": 0.45013875373914297,
+      "grad_norm": 0.2022426575422287,
+      "learning_rate": 4.952934636277158e-05,
+      "loss": 0.4544,
       "step": 12490
     },
     {
-      "epoch": 0.44,
-      "learning_rate": 4.956144781622999e-05,
-      "loss": 0.3164,
+      "epoch": 0.4503189534003676,
+      "grad_norm": 0.22232398390769958,
+      "learning_rate": 4.95287826265626e-05,
+      "loss": 0.4909,
       "step": 12495
     },
     {
-      "epoch": 0.44,
-      "learning_rate": 4.9560916414548474e-05,
-      "loss": 0.2955,
+      "epoch": 0.4504991530615922,
+      "grad_norm": 0.18882164359092712,
+      "learning_rate": 4.95282185561537e-05,
+      "loss": 0.4937,
       "step": 12500
     },
     {
-      "epoch": 0.44,
-      "eval_loss": 0.31484097242355347,
-      "eval_runtime": 10.5384,
-      "eval_samples_per_second": 9.489,
-      "eval_steps_per_second": 9.489,
+      "epoch": 0.4504991530615922,
+      "eval_loss": 0.47478610277175903,
+      "eval_runtime": 3.606,
+      "eval_samples_per_second": 27.732,
+      "eval_steps_per_second": 6.933,
       "step": 12500
     },
     {
-      "epoch": 0.44,
-      "learning_rate": 4.9560384693959644e-05,
-      "loss": 0.3188,
+      "epoch": 0.4506793527228169,
+      "grad_norm": 0.15758134424686432,
+      "learning_rate": 4.952765415155258e-05,
+      "loss": 0.4397,
       "step": 12505
     },
     {
-      "epoch": 0.44,
-      "learning_rate": 4.955985265447041e-05,
-      "loss": 0.3233,
+      "epoch": 0.4508595523840415,
+      "grad_norm": 0.15819260478019714,
+      "learning_rate": 4.9527089412766926e-05,
+      "loss": 0.4245,
       "step": 12510
     },
     {
-      "epoch": 0.44,
-      "learning_rate": 4.9559320296087664e-05,
-      "loss": 0.3287,
+      "epoch": 0.45103975204526614,
+      "grad_norm": 0.20191067457199097,
+      "learning_rate": 4.952652433980442e-05,
+      "loss": 0.4507,
       "step": 12515
     },
     {
-      "epoch": 0.44,
-      "learning_rate": 4.955878761881834e-05,
-      "loss": 0.3187,
+      "epoch": 0.4512199517064908,
+      "grad_norm": 0.14131148159503937,
+      "learning_rate": 4.952595893267277e-05,
+      "loss": 0.4334,
       "step": 12520
     },
     {
-      "epoch": 0.44,
-      "learning_rate": 4.955825462266933e-05,
-      "loss": 0.3355,
+      "epoch": 0.45140015136771544,
+      "grad_norm": 0.1871136575937271,
+      "learning_rate": 4.9525393191379674e-05,
+      "loss": 0.4563,
       "step": 12525
     },
     {
-      "epoch": 0.44,
-      "learning_rate": 4.9557721307647576e-05,
-      "loss": 0.3231,
+      "epoch": 0.45158035102894006,
+      "grad_norm": 0.1495848298072815,
+      "learning_rate": 4.952482711593285e-05,
+      "loss": 0.4611,
       "step": 12530
     },
     {
-      "epoch": 0.44,
-      "learning_rate": 4.9557187673759996e-05,
-      "loss": 0.3486,
+      "epoch": 0.4517605506901647,
+      "grad_norm": 0.1738033890724182,
+      "learning_rate": 4.9524260706339996e-05,
+      "loss": 0.4509,
       "step": 12535
     },
     {
-      "epoch": 0.44,
-      "learning_rate": 4.955665372101351e-05,
-      "loss": 0.3138,
+      "epoch": 0.45194075035138936,
+      "grad_norm": 0.2945556938648224,
+      "learning_rate": 4.952369396260884e-05,
+      "loss": 0.4461,
       "step": 12540
     },
     {
-      "epoch": 0.44,
-      "learning_rate": 4.955611944941507e-05,
-      "loss": 0.3292,
+      "epoch": 0.452120950012614,
+      "grad_norm": 0.19618237018585205,
+      "learning_rate": 4.952312688474711e-05,
+      "loss": 0.4466,
       "step": 12545
     },
     {
-      "epoch": 0.44,
-      "learning_rate": 4.95555848589716e-05,
-      "loss": 0.3262,
+      "epoch": 0.4523011496738386,
+      "grad_norm": 0.1678617298603058,
+      "learning_rate": 4.952255947276252e-05,
+      "loss": 0.474,
       "step": 12550
     },
     {
-      "epoch": 0.44,
-      "learning_rate": 4.9555049949690044e-05,
-      "loss": 0.3101,
+      "epoch": 0.45248134933506323,
+      "grad_norm": 0.1514994353055954,
+      "learning_rate": 4.95219917266628e-05,
+      "loss": 0.4271,
       "step": 12555
     },
     {
-      "epoch": 0.44,
-      "learning_rate": 4.955451472157734e-05,
-      "loss": 0.3335,
+      "epoch": 0.4526615489962879,
+      "grad_norm": 0.14985314011573792,
+      "learning_rate": 4.95214236464557e-05,
+      "loss": 0.4505,
       "step": 12560
     },
     {
-      "epoch": 0.44,
-      "learning_rate": 4.9554086309533434e-05,
-      "loss": 0.3185,
+      "epoch": 0.45284174865751253,
+      "grad_norm": 0.19544857740402222,
+      "learning_rate": 4.952085523214894e-05,
+      "loss": 0.4893,
       "step": 12565
     },
     {
-      "epoch": 0.44,
-      "learning_rate": 4.955355050754219e-05,
-      "loss": 0.3242,
+      "epoch": 0.45302194831873716,
+      "grad_norm": 0.19242407381534576,
+      "learning_rate": 4.9520286483750277e-05,
+      "loss": 0.4616,
       "step": 12570
     },
     {
-      "epoch": 0.44,
-      "learning_rate": 4.955301438673928e-05,
-      "loss": 0.3294,
+      "epoch": 0.4532021479799618,
+      "grad_norm": 0.14265184104442596,
+      "learning_rate": 4.9519717401267465e-05,
+      "loss": 0.4539,
       "step": 12575
     },
     {
-      "epoch": 0.44,
-      "learning_rate": 4.9552477947131656e-05,
-      "loss": 0.3064,
+      "epoch": 0.45338234764118646,
+      "grad_norm": 0.181315615773201,
+      "learning_rate": 4.9519147984708246e-05,
+      "loss": 0.4585,
       "step": 12580
     },
     {
-      "epoch": 0.44,
-      "learning_rate": 4.9551941188726294e-05,
-      "loss": 0.3415,
+      "epoch": 0.4535625473024111,
+      "grad_norm": 0.16452257335186005,
+      "learning_rate": 4.9518578234080384e-05,
+      "loss": 0.4469,
       "step": 12585
     },
     {
-      "epoch": 0.44,
-      "learning_rate": 4.955140411153015e-05,
-      "loss": 0.3077,
+      "epoch": 0.4537427469636357,
+      "grad_norm": 0.14965125918388367,
+      "learning_rate": 4.951800814939164e-05,
+      "loss": 0.4432,
       "step": 12590
     },
     {
-      "epoch": 0.44,
-      "learning_rate": 4.9550866715550205e-05,
-      "loss": 0.3585,
+      "epoch": 0.4539229466248603,
+      "grad_norm": 0.16924121975898743,
+      "learning_rate": 4.951743773064978e-05,
+      "loss": 0.4248,
       "step": 12595
     },
     {
-      "epoch": 0.44,
-      "learning_rate": 4.955032900079343e-05,
-      "loss": 0.3242,
+      "epoch": 0.454103146286085,
+      "grad_norm": 0.21651718020439148,
+      "learning_rate": 4.951686697786258e-05,
+      "loss": 0.4606,
       "step": 12600
     },
     {
-      "epoch": 0.44,
-      "learning_rate": 4.954979096726682e-05,
-      "loss": 0.3211,
+      "epoch": 0.4542833459473096,
+      "grad_norm": 0.1570291370153427,
+      "learning_rate": 4.951629589103781e-05,
+      "loss": 0.4161,
       "step": 12605
     },
     {
-      "epoch": 0.44,
-      "learning_rate": 4.954925261497735e-05,
-      "loss": 0.2988,
+      "epoch": 0.45446354560853425,
+      "grad_norm": 0.18217509984970093,
+      "learning_rate": 4.951572447018326e-05,
+      "loss": 0.4096,
       "step": 12610
     },
     {
-      "epoch": 0.44,
-      "learning_rate": 4.954871394393201e-05,
-      "loss": 0.3153,
+      "epoch": 0.45464374526975887,
+      "grad_norm": 0.1948910504579544,
+      "learning_rate": 4.951515271530671e-05,
+      "loss": 0.4347,
       "step": 12615
     },
     {
-      "epoch": 0.44,
-      "learning_rate": 4.954817495413781e-05,
-      "loss": 0.3298,
+      "epoch": 0.45482394493098355,
+      "grad_norm": 0.1557484269142151,
+      "learning_rate": 4.951458062641595e-05,
+      "loss": 0.44,
       "step": 12620
     },
     {
-      "epoch": 0.44,
-      "learning_rate": 4.954763564560172e-05,
-      "loss": 0.3225,
+      "epoch": 0.45500414459220817,
+      "grad_norm": 0.21191979944705963,
+      "learning_rate": 4.951400820351877e-05,
+      "loss": 0.4676,
       "step": 12625
     },
     {
-      "epoch": 0.44,
-      "learning_rate": 4.954709601833078e-05,
-      "loss": 0.3155,
+      "epoch": 0.4551843442534328,
+      "grad_norm": 0.20376695692539215,
+      "learning_rate": 4.951343544662298e-05,
+      "loss": 0.4121,
       "step": 12630
     },
     {
-      "epoch": 0.44,
-      "learning_rate": 4.954655607233195e-05,
-      "loss": 0.3131,
+      "epoch": 0.4553645439146574,
+      "grad_norm": 0.18317648768424988,
+      "learning_rate": 4.9512862355736376e-05,
+      "loss": 0.4806,
       "step": 12635
     },
     {
-      "epoch": 0.44,
-      "learning_rate": 4.9546015807612275e-05,
-      "loss": 0.3121,
+      "epoch": 0.4555447435758821,
+      "grad_norm": 0.15849465131759644,
+      "learning_rate": 4.951228893086677e-05,
+      "loss": 0.4107,
       "step": 12640
     },
     {
-      "epoch": 0.44,
-      "learning_rate": 4.954547522417877e-05,
-      "loss": 0.3246,
+      "epoch": 0.4557249432371067,
+      "grad_norm": 0.1734628528356552,
+      "learning_rate": 4.951171517202197e-05,
+      "loss": 0.4431,
       "step": 12645
     },
     {
-      "epoch": 0.45,
-      "learning_rate": 4.954493432203844e-05,
-      "loss": 0.3338,
+      "epoch": 0.45590514289833134,
+      "grad_norm": 0.19202375411987305,
+      "learning_rate": 4.95111410792098e-05,
+      "loss": 0.4589,
       "step": 12650
     },
     {
-      "epoch": 0.45,
-      "learning_rate": 4.954439310119831e-05,
-      "loss": 0.3075,
+      "epoch": 0.45608534255955596,
+      "grad_norm": 0.1690249890089035,
+      "learning_rate": 4.951056665243807e-05,
+      "loss": 0.4878,
       "step": 12655
     },
     {
-      "epoch": 0.45,
-      "learning_rate": 4.954385156166541e-05,
-      "loss": 0.2985,
+      "epoch": 0.45626554222078064,
+      "grad_norm": 0.13854099810123444,
+      "learning_rate": 4.950999189171463e-05,
+      "loss": 0.4422,
       "step": 12660
     },
     {
-      "epoch": 0.45,
-      "learning_rate": 4.9543309703446783e-05,
-      "loss": 0.3016,
+      "epoch": 0.45644574188200526,
+      "grad_norm": 0.16640843451023102,
+      "learning_rate": 4.9509416797047284e-05,
+      "loss": 0.4411,
       "step": 12665
     },
     {
-      "epoch": 0.45,
-      "learning_rate": 4.9542767526549445e-05,
-      "loss": 0.3416,
+      "epoch": 0.4566259415432299,
+      "grad_norm": 0.19313612580299377,
+      "learning_rate": 4.9508841368443884e-05,
+      "loss": 0.4788,
       "step": 12670
     },
     {
-      "epoch": 0.45,
-      "learning_rate": 4.954222503098045e-05,
-      "loss": 0.3014,
+      "epoch": 0.45680614120445456,
+      "grad_norm": 0.19194231927394867,
+      "learning_rate": 4.9508265605912265e-05,
+      "loss": 0.4654,
       "step": 12675
     },
     {
-      "epoch": 0.45,
-      "learning_rate": 4.954168221674683e-05,
-      "loss": 0.313,
+      "epoch": 0.4569863408656792,
+      "grad_norm": 0.2459125965833664,
+      "learning_rate": 4.950768950946026e-05,
+      "loss": 0.4259,
       "step": 12680
     },
     {
-      "epoch": 0.45,
-      "learning_rate": 4.954113908385565e-05,
-      "loss": 0.3281,
+      "epoch": 0.4571665405269038,
+      "grad_norm": 0.17043115198612213,
+      "learning_rate": 4.950711307909575e-05,
+      "loss": 0.4381,
       "step": 12685
     },
     {
-      "epoch": 0.45,
-      "learning_rate": 4.954059563231394e-05,
-      "loss": 0.3349,
+      "epoch": 0.45734674018812843,
+      "grad_norm": 0.1516115516424179,
+      "learning_rate": 4.950653631482656e-05,
+      "loss": 0.4153,
       "step": 12690
     },
     {
-      "epoch": 0.45,
-      "learning_rate": 4.954005186212878e-05,
-      "loss": 0.3061,
+      "epoch": 0.4575269398493531,
+      "grad_norm": 0.18203213810920715,
+      "learning_rate": 4.9505959216660556e-05,
+      "loss": 0.4587,
       "step": 12695
     },
     {
-      "epoch": 0.45,
-      "learning_rate": 4.9539507773307215e-05,
-      "loss": 0.3097,
+      "epoch": 0.45770713951057773,
+      "grad_norm": 0.16077850759029388,
+      "learning_rate": 4.9505381784605606e-05,
+      "loss": 0.4684,
       "step": 12700
     },
     {
-      "epoch": 0.45,
-      "learning_rate": 4.9538963365856305e-05,
-      "loss": 0.3377,
+      "epoch": 0.45788733917180235,
+      "grad_norm": 0.16512863337993622,
+      "learning_rate": 4.9504804018669574e-05,
+      "loss": 0.4445,
       "step": 12705
     },
     {
-      "epoch": 0.45,
-      "learning_rate": 4.953841863978314e-05,
-      "loss": 0.3343,
+      "epoch": 0.458067538833027,
+      "grad_norm": 0.16586919128894806,
+      "learning_rate": 4.9504225918860326e-05,
+      "loss": 0.4602,
       "step": 12710
     },
     {
-      "epoch": 0.45,
-      "learning_rate": 4.953787359509477e-05,
-      "loss": 0.3359,
+      "epoch": 0.45824773849425166,
+      "grad_norm": 0.1363137811422348,
+      "learning_rate": 4.9503647485185744e-05,
+      "loss": 0.4658,
       "step": 12715
     },
     {
-      "epoch": 0.45,
-      "learning_rate": 4.953732823179829e-05,
-      "loss": 0.3121,
+      "epoch": 0.4584279381554763,
+      "grad_norm": 0.18424583971500397,
+      "learning_rate": 4.950306871765371e-05,
+      "loss": 0.4875,
       "step": 12720
     },
     {
-      "epoch": 0.45,
-      "learning_rate": 4.9536782549900776e-05,
-      "loss": 0.3339,
+      "epoch": 0.4586081378167009,
+      "grad_norm": 0.21040771901607513,
+      "learning_rate": 4.9502489616272115e-05,
+      "loss": 0.492,
       "step": 12725
     },
     {
-      "epoch": 0.45,
-      "learning_rate": 4.953623654940931e-05,
-      "loss": 0.3041,
+      "epoch": 0.4587883374779255,
+      "grad_norm": 0.15792599320411682,
+      "learning_rate": 4.9501910181048836e-05,
+      "loss": 0.4733,
       "step": 12730
     },
     {
-      "epoch": 0.45,
-      "learning_rate": 4.953569023033098e-05,
-      "loss": 0.3408,
+      "epoch": 0.4589685371391502,
+      "grad_norm": 0.1881541758775711,
+      "learning_rate": 4.950133041199177e-05,
+      "loss": 0.4476,
       "step": 12735
     },
     {
-      "epoch": 0.45,
-      "learning_rate": 4.9535143592672874e-05,
-      "loss": 0.315,
+      "epoch": 0.4591487368003748,
+      "grad_norm": 0.14302214980125427,
+      "learning_rate": 4.950075030910883e-05,
+      "loss": 0.457,
       "step": 12740
     },
     {
-      "epoch": 0.45,
-      "learning_rate": 4.9534596636442106e-05,
-      "loss": 0.3189,
+      "epoch": 0.45932893646159945,
+      "grad_norm": 0.16461120545864105,
+      "learning_rate": 4.9500169872407906e-05,
+      "loss": 0.4546,
       "step": 12745
     },
     {
-      "epoch": 0.45,
-      "learning_rate": 4.953404936164577e-05,
-      "loss": 0.3107,
+      "epoch": 0.45950913612282407,
+      "grad_norm": 0.18678593635559082,
+      "learning_rate": 4.949958910189692e-05,
+      "loss": 0.4279,
       "step": 12750
     },
     {
-      "epoch": 0.45,
-      "learning_rate": 4.953350176829097e-05,
-      "loss": 0.3081,
+      "epoch": 0.45968933578404875,
+      "grad_norm": 0.18721038103103638,
+      "learning_rate": 4.949900799758377e-05,
+      "loss": 0.443,
       "step": 12755
     },
     {
-      "epoch": 0.45,
-      "learning_rate": 4.9532953856384814e-05,
-      "loss": 0.3367,
+      "epoch": 0.45986953544527337,
+      "grad_norm": 0.1702054888010025,
+      "learning_rate": 4.949842655947637e-05,
+      "loss": 0.4364,
       "step": 12760
     },
     {
-      "epoch": 0.45,
-      "learning_rate": 4.9532405625934425e-05,
-      "loss": 0.3211,
+      "epoch": 0.460049735106498,
+      "grad_norm": 0.18055669963359833,
+      "learning_rate": 4.9497844787582665e-05,
+      "loss": 0.4579,
       "step": 12765
     },
     {
-      "epoch": 0.45,
-      "learning_rate": 4.9531857076946914e-05,
-      "loss": 0.3377,
+      "epoch": 0.4602299347677226,
+      "grad_norm": 0.18560905754566193,
+      "learning_rate": 4.949726268191056e-05,
+      "loss": 0.4471,
       "step": 12770
     },
     {
-      "epoch": 0.45,
-      "learning_rate": 4.9531308209429404e-05,
-      "loss": 0.3257,
+      "epoch": 0.4604101344289473,
+      "grad_norm": 0.16065679490566254,
+      "learning_rate": 4.9496680242467994e-05,
+      "loss": 0.4403,
       "step": 12775
     },
     {
-      "epoch": 0.45,
-      "learning_rate": 4.953075902338903e-05,
-      "loss": 0.3615,
+      "epoch": 0.4605903340901719,
+      "grad_norm": 0.1695624440908432,
+      "learning_rate": 4.94960974692629e-05,
+      "loss": 0.4245,
       "step": 12780
     },
     {
-      "epoch": 0.45,
-      "learning_rate": 4.9530209518832915e-05,
-      "loss": 0.307,
+      "epoch": 0.46077053375139654,
+      "grad_norm": 0.1966797411441803,
+      "learning_rate": 4.949551436230323e-05,
+      "loss": 0.4664,
       "step": 12785
     },
     {
-      "epoch": 0.45,
-      "learning_rate": 4.952965969576819e-05,
-      "loss": 0.3345,
+      "epoch": 0.46095073341262116,
+      "grad_norm": 0.19263668358325958,
+      "learning_rate": 4.9494930921596913e-05,
+      "loss": 0.4823,
       "step": 12790
     },
     {
-      "epoch": 0.45,
-      "learning_rate": 4.9529109554202e-05,
-      "loss": 0.3315,
+      "epoch": 0.46113093307384584,
+      "grad_norm": 0.16463421285152435,
+      "learning_rate": 4.949434714715191e-05,
+      "loss": 0.4511,
       "step": 12795
     },
     {
-      "epoch": 0.45,
-      "learning_rate": 4.95285590941415e-05,
-      "loss": 0.3312,
+      "epoch": 0.46131113273507046,
+      "grad_norm": 0.18872933089733124,
+      "learning_rate": 4.949376303897616e-05,
+      "loss": 0.4424,
       "step": 12800
     },
     {
-      "epoch": 0.45,
-      "learning_rate": 4.952800831559381e-05,
-      "loss": 0.3103,
+      "epoch": 0.4614913323962951,
+      "grad_norm": 0.14205656945705414,
+      "learning_rate": 4.949317859707764e-05,
+      "loss": 0.4761,
       "step": 12805
     },
     {
-      "epoch": 0.45,
-      "learning_rate": 4.9527457218566104e-05,
-      "loss": 0.3058,
+      "epoch": 0.4616715320575197,
+      "grad_norm": 0.16842302680015564,
+      "learning_rate": 4.94925938214643e-05,
+      "loss": 0.4496,
       "step": 12810
     },
     {
-      "epoch": 0.45,
-      "learning_rate": 4.952690580306553e-05,
-      "loss": 0.295,
+      "epoch": 0.4618517317187444,
+      "grad_norm": 0.15166863799095154,
+      "learning_rate": 4.9492008712144104e-05,
+      "loss": 0.4274,
       "step": 12815
     },
     {
-      "epoch": 0.45,
-      "learning_rate": 4.952635406909925e-05,
-      "loss": 0.3236,
+      "epoch": 0.462031931379969,
+      "grad_norm": 0.17034749686717987,
+      "learning_rate": 4.9491423269125035e-05,
+      "loss": 0.4313,
       "step": 12820
     },
     {
-      "epoch": 0.45,
-      "learning_rate": 4.952580201667442e-05,
-      "loss": 0.3461,
+      "epoch": 0.46221213104119363,
+      "grad_norm": 0.1309286206960678,
+      "learning_rate": 4.9490837492415074e-05,
+      "loss": 0.4346,
       "step": 12825
     },
     {
-      "epoch": 0.45,
-      "learning_rate": 4.952524964579821e-05,
-      "loss": 0.3296,
+      "epoch": 0.46239233070241825,
+      "grad_norm": 0.18976598978042603,
+      "learning_rate": 4.949025138202218e-05,
+      "loss": 0.4361,
       "step": 12830
     },
     {
-      "epoch": 0.45,
-      "learning_rate": 4.952469695647781e-05,
-      "loss": 0.3343,
+      "epoch": 0.46257253036364293,
+      "grad_norm": 0.1648472547531128,
+      "learning_rate": 4.948966493795437e-05,
+      "loss": 0.4638,
       "step": 12835
     },
     {
-      "epoch": 0.45,
-      "learning_rate": 4.952414394872037e-05,
-      "loss": 0.3293,
+      "epoch": 0.46275273002486755,
+      "grad_norm": 0.18952538073062897,
+      "learning_rate": 4.94890781602196e-05,
+      "loss": 0.4521,
       "step": 12840
     },
     {
-      "epoch": 0.45,
-      "learning_rate": 4.952359062253309e-05,
-      "loss": 0.3196,
+      "epoch": 0.4629329296860922,
+      "grad_norm": 0.15713319182395935,
+      "learning_rate": 4.9488491048825894e-05,
+      "loss": 0.4122,
       "step": 12845
     },
     {
-      "epoch": 0.45,
-      "learning_rate": 4.9523036977923145e-05,
-      "loss": 0.2872,
+      "epoch": 0.46311312934731685,
+      "grad_norm": 0.16628234088420868,
+      "learning_rate": 4.9487903603781225e-05,
+      "loss": 0.4013,
       "step": 12850
     },
     {
-      "epoch": 0.45,
-      "learning_rate": 4.952248301489772e-05,
-      "loss": 0.3446,
+      "epoch": 0.4632933290085415,
+      "grad_norm": 0.17581500113010406,
+      "learning_rate": 4.948731582509362e-05,
+      "loss": 0.4685,
       "step": 12855
     },
     {
-      "epoch": 0.45,
-      "learning_rate": 4.952192873346403e-05,
-      "loss": 0.3047,
+      "epoch": 0.4634735286697661,
+      "grad_norm": 0.1562207192182541,
+      "learning_rate": 4.948672771277107e-05,
+      "loss": 0.4381,
       "step": 12860
     },
     {
-      "epoch": 0.45,
-      "learning_rate": 4.952137413362924e-05,
-      "loss": 0.3364,
+      "epoch": 0.4636537283309907,
+      "grad_norm": 0.16530804336071014,
+      "learning_rate": 4.9486139266821606e-05,
+      "loss": 0.4846,
       "step": 12865
     },
     {
-      "epoch": 0.45,
-      "learning_rate": 4.9520819215400574e-05,
-      "loss": 0.3326,
+      "epoch": 0.4638339279922154,
+      "grad_norm": 0.17569473385810852,
+      "learning_rate": 4.948555048725323e-05,
+      "loss": 0.4499,
       "step": 12870
     },
     {
-      "epoch": 0.45,
-      "learning_rate": 4.952026397878523e-05,
-      "loss": 0.3088,
+      "epoch": 0.46401412765344,
+      "grad_norm": 0.16137835383415222,
+      "learning_rate": 4.948496137407397e-05,
+      "loss": 0.4263,
       "step": 12875
     },
     {
-      "epoch": 0.45,
-      "learning_rate": 4.951970842379042e-05,
-      "loss": 0.3121,
+      "epoch": 0.46419432731466465,
+      "grad_norm": 0.18716813623905182,
+      "learning_rate": 4.948437192729186e-05,
+      "loss": 0.4634,
       "step": 12880
     },
     {
-      "epoch": 0.45,
-      "learning_rate": 4.9519152550423355e-05,
-      "loss": 0.3348,
+      "epoch": 0.46437452697588927,
+      "grad_norm": 0.17155876755714417,
+      "learning_rate": 4.948378214691491e-05,
+      "loss": 0.4479,
       "step": 12885
     },
     {
-      "epoch": 0.45,
-      "learning_rate": 4.9518596358691246e-05,
-      "loss": 0.2744,
+      "epoch": 0.46455472663711395,
+      "grad_norm": 0.16690602898597717,
+      "learning_rate": 4.948319203295117e-05,
+      "loss": 0.409,
       "step": 12890
     },
     {
-      "epoch": 0.45,
-      "learning_rate": 4.951803984860133e-05,
-      "loss": 0.3146,
+      "epoch": 0.46473492629833857,
+      "grad_norm": 0.18938426673412323,
+      "learning_rate": 4.9482601585408684e-05,
+      "loss": 0.4384,
       "step": 12895
     },
     {
-      "epoch": 0.45,
-      "learning_rate": 4.9517483020160816e-05,
-      "loss": 0.3183,
+      "epoch": 0.4649151259595632,
+      "grad_norm": 0.1318012923002243,
+      "learning_rate": 4.9482010804295485e-05,
+      "loss": 0.4067,
       "step": 12900
     },
     {
-      "epoch": 0.45,
-      "learning_rate": 4.9516925873376955e-05,
-      "loss": 0.3227,
+      "epoch": 0.4650953256207878,
+      "grad_norm": 0.15733405947685242,
+      "learning_rate": 4.9481419689619635e-05,
+      "loss": 0.4644,
       "step": 12905
     },
     {
-      "epoch": 0.45,
-      "learning_rate": 4.951636840825695e-05,
-      "loss": 0.3254,
+      "epoch": 0.4652755252820125,
+      "grad_norm": 0.18979695439338684,
+      "learning_rate": 4.9480828241389176e-05,
+      "loss": 0.469,
       "step": 12910
     },
     {
-      "epoch": 0.45,
-      "learning_rate": 4.951581062480807e-05,
-      "loss": 0.353,
+      "epoch": 0.4654557249432371,
+      "grad_norm": 0.23890700936317444,
+      "learning_rate": 4.948023645961218e-05,
+      "loss": 0.468,
       "step": 12915
     },
     {
-      "epoch": 0.45,
-      "learning_rate": 4.9515252523037545e-05,
-      "loss": 0.3276,
+      "epoch": 0.46563592460446174,
+      "grad_norm": 0.13900893926620483,
+      "learning_rate": 4.9479644344296694e-05,
+      "loss": 0.4231,
       "step": 12920
     },
     {
-      "epoch": 0.45,
-      "learning_rate": 4.9514694102952625e-05,
-      "loss": 0.3128,
+      "epoch": 0.46581612426568636,
+      "grad_norm": 0.16857506334781647,
+      "learning_rate": 4.94790518954508e-05,
+      "loss": 0.3977,
       "step": 12925
     },
     {
-      "epoch": 0.45,
-      "learning_rate": 4.951413536456055e-05,
-      "loss": 0.3235,
+      "epoch": 0.46599632392691104,
+      "grad_norm": 0.2040061503648758,
+      "learning_rate": 4.9478459113082556e-05,
+      "loss": 0.4529,
       "step": 12930
     },
     {
-      "epoch": 0.46,
-      "learning_rate": 4.951357630786859e-05,
-      "loss": 0.3151,
+      "epoch": 0.46617652358813566,
+      "grad_norm": 0.1292639821767807,
+      "learning_rate": 4.9477865997200044e-05,
+      "loss": 0.4345,
       "step": 12935
     },
     {
-      "epoch": 0.46,
-      "learning_rate": 4.9513016932883994e-05,
-      "loss": 0.319,
+      "epoch": 0.4663567232493603,
+      "grad_norm": 0.1888824999332428,
+      "learning_rate": 4.9477272547811354e-05,
+      "loss": 0.436,
       "step": 12940
     },
     {
-      "epoch": 0.46,
-      "learning_rate": 4.9512457239614026e-05,
-      "loss": 0.3616,
+      "epoch": 0.4665369229105849,
+      "grad_norm": 0.18811088800430298,
+      "learning_rate": 4.947667876492457e-05,
+      "loss": 0.4403,
       "step": 12945
     },
     {
-      "epoch": 0.46,
-      "learning_rate": 4.951189722806596e-05,
-      "loss": 0.3197,
+      "epoch": 0.4667171225718096,
+      "grad_norm": 0.17005407810211182,
+      "learning_rate": 4.947608464854776e-05,
+      "loss": 0.4412,
       "step": 12950
     },
     {
-      "epoch": 0.46,
-      "learning_rate": 4.951133689824706e-05,
-      "loss": 0.304,
+      "epoch": 0.4668973222330342,
+      "grad_norm": 0.17521242797374725,
+      "learning_rate": 4.9475490198689044e-05,
+      "loss": 0.4371,
       "step": 12955
     },
     {
-      "epoch": 0.46,
-      "learning_rate": 4.95107762501646e-05,
-      "loss": 0.3244,
+      "epoch": 0.46707752189425883,
+      "grad_norm": 0.21031251549720764,
+      "learning_rate": 4.947489541535651e-05,
+      "loss": 0.4775,
       "step": 12960
     },
     {
-      "epoch": 0.46,
-      "learning_rate": 4.951021528382587e-05,
-      "loss": 0.2945,
+      "epoch": 0.46725772155548345,
+      "grad_norm": 0.19381965696811676,
+      "learning_rate": 4.947430029855827e-05,
+      "loss": 0.4625,
       "step": 12965
     },
     {
-      "epoch": 0.46,
-      "learning_rate": 4.9509653999238143e-05,
-      "loss": 0.3401,
+      "epoch": 0.46743792121670813,
+      "grad_norm": 0.19049398601055145,
+      "learning_rate": 4.9473704848302424e-05,
+      "loss": 0.4648,
       "step": 12970
     },
     {
-      "epoch": 0.46,
-      "learning_rate": 4.950909239640871e-05,
-      "loss": 0.3175,
+      "epoch": 0.46761812087793275,
+      "grad_norm": 0.17350056767463684,
+      "learning_rate": 4.947310906459709e-05,
+      "loss": 0.4143,
       "step": 12975
     },
     {
-      "epoch": 0.46,
-      "learning_rate": 4.950853047534487e-05,
-      "loss": 0.3314,
+      "epoch": 0.4677983205391574,
+      "grad_norm": 0.20652742683887482,
+      "learning_rate": 4.947251294745038e-05,
+      "loss": 0.4555,
       "step": 12980
     },
     {
-      "epoch": 0.46,
-      "learning_rate": 4.950796823605391e-05,
-      "loss": 0.3377,
+      "epoch": 0.467978520200382,
+      "grad_norm": 0.1818598359823227,
+      "learning_rate": 4.947191649687042e-05,
+      "loss": 0.4312,
       "step": 12985
     },
     {
-      "epoch": 0.46,
-      "learning_rate": 4.950740567854313e-05,
-      "loss": 0.3175,
+      "epoch": 0.4681587198616067,
+      "grad_norm": 0.17591311037540436,
+      "learning_rate": 4.947131971286534e-05,
+      "loss": 0.4311,
       "step": 12990
     },
     {
-      "epoch": 0.46,
-      "learning_rate": 4.950684280281985e-05,
-      "loss": 0.3056,
+      "epoch": 0.4683389195228313,
+      "grad_norm": 0.14075951278209686,
+      "learning_rate": 4.947072259544326e-05,
+      "loss": 0.4396,
       "step": 12995
     },
     {
-      "epoch": 0.46,
-      "learning_rate": 4.950627960889136e-05,
-      "loss": 0.306,
+      "epoch": 0.4685191191840559,
+      "grad_norm": 0.20111533999443054,
+      "learning_rate": 4.947012514461232e-05,
+      "loss": 0.4697,
       "step": 13000
     },
     {
-      "epoch": 0.46,
-      "eval_loss": 0.31242769956588745,
-      "eval_runtime": 10.5492,
-      "eval_samples_per_second": 9.479,
-      "eval_steps_per_second": 9.479,
+      "epoch": 0.4685191191840559,
+      "eval_loss": 0.4741581976413727,
+      "eval_runtime": 3.5647,
+      "eval_samples_per_second": 28.053,
+      "eval_steps_per_second": 7.013,
       "step": 13000
     },
     {
-      "epoch": 0.46,
-      "learning_rate": 4.9505716096764976e-05,
-      "loss": 0.3369,
+      "epoch": 0.46869931884528054,
+      "grad_norm": 0.1940506100654602,
+      "learning_rate": 4.9469527360380676e-05,
+      "loss": 0.4683,
       "step": 13005
     },
     {
-      "epoch": 0.46,
-      "learning_rate": 4.950515226644802e-05,
-      "loss": 0.3456,
+      "epoch": 0.4688795185065052,
+      "grad_norm": 0.16118963062763214,
+      "learning_rate": 4.946892924275645e-05,
+      "loss": 0.4534,
       "step": 13010
     },
     {
-      "epoch": 0.46,
-      "learning_rate": 4.9504588117947814e-05,
-      "loss": 0.3514,
+      "epoch": 0.46905971816772984,
+      "grad_norm": 0.1373254507780075,
+      "learning_rate": 4.9468330791747795e-05,
+      "loss": 0.4503,
       "step": 13015
     },
     {
-      "epoch": 0.46,
-      "learning_rate": 4.950402365127168e-05,
-      "loss": 0.3113,
+      "epoch": 0.46923991782895447,
+      "grad_norm": 0.1803598254919052,
+      "learning_rate": 4.946773200736288e-05,
+      "loss": 0.481,
       "step": 13020
     },
     {
-      "epoch": 0.46,
-      "learning_rate": 4.950345886642694e-05,
-      "loss": 0.3034,
+      "epoch": 0.46942011749017915,
+      "grad_norm": 0.18587857484817505,
+      "learning_rate": 4.9467132889609845e-05,
+      "loss": 0.4849,
       "step": 13025
     },
     {
-      "epoch": 0.46,
-      "learning_rate": 4.950289376342094e-05,
-      "loss": 0.3103,
+      "epoch": 0.46960031715140377,
+      "grad_norm": 0.17999742925167084,
+      "learning_rate": 4.946653343849686e-05,
+      "loss": 0.4621,
       "step": 13030
     },
     {
-      "epoch": 0.46,
-      "learning_rate": 4.9502328342261015e-05,
-      "loss": 0.3041,
+      "epoch": 0.4697805168126284,
+      "grad_norm": 0.17007941007614136,
+      "learning_rate": 4.9465933654032106e-05,
+      "loss": 0.441,
       "step": 13035
     },
     {
-      "epoch": 0.46,
-      "learning_rate": 4.9501762602954505e-05,
-      "loss": 0.3191,
+      "epoch": 0.469960716473853,
+      "grad_norm": 0.1974543035030365,
+      "learning_rate": 4.9465333536223734e-05,
+      "loss": 0.4479,
       "step": 13040
     },
     {
-      "epoch": 0.46,
-      "learning_rate": 4.9501196545508756e-05,
-      "loss": 0.3235,
+      "epoch": 0.4701409161350777,
+      "grad_norm": 0.1567683070898056,
+      "learning_rate": 4.946473308507993e-05,
+      "loss": 0.4881,
       "step": 13045
     },
     {
-      "epoch": 0.46,
-      "learning_rate": 4.950063016993112e-05,
-      "loss": 0.324,
+      "epoch": 0.4703211157963023,
+      "grad_norm": 0.17787274718284607,
+      "learning_rate": 4.9464132300608876e-05,
+      "loss": 0.4597,
       "step": 13050
     },
     {
-      "epoch": 0.46,
-      "learning_rate": 4.9500063476228936e-05,
-      "loss": 0.3255,
+      "epoch": 0.47050131545752694,
+      "grad_norm": 0.2110513597726822,
+      "learning_rate": 4.9463531182818756e-05,
+      "loss": 0.4353,
       "step": 13055
     },
     {
-      "epoch": 0.46,
-      "learning_rate": 4.9499496464409576e-05,
-      "loss": 0.3304,
+      "epoch": 0.47068151511875156,
+      "grad_norm": 0.19464978575706482,
+      "learning_rate": 4.946292973171777e-05,
+      "loss": 0.419,
       "step": 13060
     },
     {
-      "epoch": 0.46,
-      "learning_rate": 4.9498929134480406e-05,
-      "loss": 0.3112,
+      "epoch": 0.47086171477997624,
+      "grad_norm": 0.18189744651317596,
+      "learning_rate": 4.946232794731408e-05,
+      "loss": 0.463,
       "step": 13065
     },
     {
-      "epoch": 0.46,
-      "learning_rate": 4.9498361486448786e-05,
-      "loss": 0.3035,
+      "epoch": 0.47104191444120086,
+      "grad_norm": 0.21045507490634918,
+      "learning_rate": 4.946172582961593e-05,
+      "loss": 0.5024,
       "step": 13070
     },
     {
-      "epoch": 0.46,
-      "learning_rate": 4.949779352032208e-05,
-      "loss": 0.3178,
+      "epoch": 0.4712221141024255,
+      "grad_norm": 0.16427326202392578,
+      "learning_rate": 4.946112337863148e-05,
+      "loss": 0.4796,
       "step": 13075
     },
     {
-      "epoch": 0.46,
-      "learning_rate": 4.949722523610768e-05,
-      "loss": 0.309,
+      "epoch": 0.4714023137636501,
+      "grad_norm": 0.18981902301311493,
+      "learning_rate": 4.9460520594368975e-05,
+      "loss": 0.4395,
       "step": 13080
     },
     {
-      "epoch": 0.46,
-      "learning_rate": 4.949665663381295e-05,
-      "loss": 0.3165,
+      "epoch": 0.4715825134248748,
+      "grad_norm": 0.20156720280647278,
+      "learning_rate": 4.945991747683661e-05,
+      "loss": 0.4865,
       "step": 13085
     },
     {
-      "epoch": 0.46,
-      "learning_rate": 4.949608771344528e-05,
-      "loss": 0.3333,
+      "epoch": 0.4717627130860994,
+      "grad_norm": 0.14787843823432922,
+      "learning_rate": 4.9459314026042605e-05,
+      "loss": 0.4583,
       "step": 13090
     },
     {
-      "epoch": 0.46,
-      "learning_rate": 4.949551847501205e-05,
-      "loss": 0.3362,
+      "epoch": 0.47194291274732403,
+      "grad_norm": 0.1643127053976059,
+      "learning_rate": 4.9458710241995174e-05,
+      "loss": 0.4685,
       "step": 13095
     },
     {
-      "epoch": 0.46,
-      "learning_rate": 4.9494948918520655e-05,
-      "loss": 0.2987,
+      "epoch": 0.47212311240854865,
+      "grad_norm": 0.19920234382152557,
+      "learning_rate": 4.9458106124702565e-05,
+      "loss": 0.453,
       "step": 13100
     },
     {
-      "epoch": 0.46,
-      "learning_rate": 4.9494379043978494e-05,
-      "loss": 0.2885,
+      "epoch": 0.47230331206977333,
+      "grad_norm": 0.16437123715877533,
+      "learning_rate": 4.945750167417299e-05,
+      "loss": 0.4052,
       "step": 13105
     },
     {
-      "epoch": 0.46,
-      "learning_rate": 4.9493808851392964e-05,
-      "loss": 0.3404,
+      "epoch": 0.47248351173099795,
+      "grad_norm": 0.1786852478981018,
+      "learning_rate": 4.945689689041468e-05,
+      "loss": 0.4609,
       "step": 13110
     },
     {
-      "epoch": 0.46,
-      "learning_rate": 4.9493238340771465e-05,
-      "loss": 0.3182,
+      "epoch": 0.4726637113922226,
+      "grad_norm": NaN,
+      "learning_rate": 4.9456412823488897e-05,
+      "loss": 0.4437,
       "step": 13115
     },
     {
-      "epoch": 0.46,
-      "learning_rate": 4.949266751212142e-05,
-      "loss": 0.3212,
+      "epoch": 0.4728439110534472,
+      "grad_norm": 0.18749703466892242,
+      "learning_rate": 4.9455807439939664e-05,
+      "loss": 0.4415,
       "step": 13120
     },
     {
-      "epoch": 0.46,
-      "learning_rate": 4.949209636545021e-05,
-      "loss": 0.3207,
+      "epoch": 0.4730241107146719,
+      "grad_norm": 0.15940409898757935,
+      "learning_rate": 4.9455201723184773e-05,
+      "loss": 0.4648,
       "step": 13125
     },
     {
-      "epoch": 0.46,
-      "learning_rate": 4.949152490076528e-05,
-      "loss": 0.3215,
+      "epoch": 0.4732043103758965,
+      "grad_norm": 0.1593160629272461,
+      "learning_rate": 4.9454595673232505e-05,
+      "loss": 0.4266,
       "step": 13130
     },
     {
-      "epoch": 0.46,
-      "learning_rate": 4.949095311807404e-05,
-      "loss": 0.3127,
+      "epoch": 0.4733845100371211,
+      "grad_norm": 0.1828412562608719,
+      "learning_rate": 4.9453989290091106e-05,
+      "loss": 0.404,
       "step": 13135
     },
     {
-      "epoch": 0.46,
-      "learning_rate": 4.9490381017383906e-05,
-      "loss": 0.3429,
+      "epoch": 0.47356470969834574,
+      "grad_norm": 0.16431830823421478,
+      "learning_rate": 4.945338257376884e-05,
+      "loss": 0.487,
       "step": 13140
     },
     {
-      "epoch": 0.46,
-      "learning_rate": 4.9489808598702325e-05,
-      "loss": 0.3377,
+      "epoch": 0.4737449093595704,
+      "grad_norm": 0.17641369998455048,
+      "learning_rate": 4.9452775524273963e-05,
+      "loss": 0.4485,
       "step": 13145
     },
     {
-      "epoch": 0.46,
-      "learning_rate": 4.948923586203671e-05,
-      "loss": 0.3528,
+      "epoch": 0.47392510902079504,
+      "grad_norm": 0.1741415560245514,
+      "learning_rate": 4.9452168141614754e-05,
+      "loss": 0.4163,
       "step": 13150
     },
     {
-      "epoch": 0.46,
-      "learning_rate": 4.948866280739452e-05,
-      "loss": 0.3036,
+      "epoch": 0.47410530868201967,
+      "grad_norm": 0.20057284832000732,
+      "learning_rate": 4.9451560425799495e-05,
+      "loss": 0.4494,
       "step": 13155
     },
     {
-      "epoch": 0.46,
-      "learning_rate": 4.9488089434783165e-05,
-      "loss": 0.3039,
+      "epoch": 0.4742855083432443,
+      "grad_norm": 0.16417837142944336,
+      "learning_rate": 4.9450952376836454e-05,
+      "loss": 0.4283,
       "step": 13160
     },
     {
-      "epoch": 0.46,
-      "learning_rate": 4.948751574421011e-05,
-      "loss": 0.3316,
+      "epoch": 0.47446570800446897,
+      "grad_norm": 0.14967764914035797,
+      "learning_rate": 4.945034399473392e-05,
+      "loss": 0.441,
       "step": 13165
     },
     {
-      "epoch": 0.46,
-      "learning_rate": 4.948694173568281e-05,
-      "loss": 0.3309,
+      "epoch": 0.4746459076656936,
+      "grad_norm": 0.17966294288635254,
+      "learning_rate": 4.944973527950019e-05,
+      "loss": 0.4765,
       "step": 13170
     },
     {
-      "epoch": 0.46,
-      "learning_rate": 4.94863674092087e-05,
-      "loss": 0.3295,
+      "epoch": 0.4748261073269182,
+      "grad_norm": 0.18716269731521606,
+      "learning_rate": 4.944912623114354e-05,
+      "loss": 0.4723,
       "step": 13175
     },
     {
-      "epoch": 0.46,
-      "learning_rate": 4.9485792764795255e-05,
-      "loss": 0.3174,
+      "epoch": 0.47500630698814283,
+      "grad_norm": 0.16850250959396362,
+      "learning_rate": 4.9448516849672285e-05,
+      "loss": 0.4771,
       "step": 13180
     },
     {
-      "epoch": 0.46,
-      "learning_rate": 4.948521780244992e-05,
-      "loss": 0.3137,
+      "epoch": 0.4751865066493675,
+      "grad_norm": 0.1683826893568039,
+      "learning_rate": 4.944790713509472e-05,
+      "loss": 0.4295,
       "step": 13185
     },
     {
-      "epoch": 0.46,
-      "learning_rate": 4.948464252218017e-05,
-      "loss": 0.3127,
+      "epoch": 0.47536670631059214,
+      "grad_norm": 0.1629522442817688,
+      "learning_rate": 4.9447297087419155e-05,
+      "loss": 0.4116,
       "step": 13190
     },
     {
-      "epoch": 0.46,
-      "learning_rate": 4.9484066923993476e-05,
-      "loss": 0.3373,
+      "epoch": 0.47554690597181676,
+      "grad_norm": 0.16249153017997742,
+      "learning_rate": 4.9446686706653896e-05,
+      "loss": 0.431,
       "step": 13195
     },
     {
-      "epoch": 0.46,
-      "learning_rate": 4.948349100789731e-05,
-      "loss": 0.3312,
+      "epoch": 0.47572710563304144,
+      "grad_norm": 0.16128148138523102,
+      "learning_rate": 4.944607599280726e-05,
+      "loss": 0.4552,
       "step": 13200
     },
     {
-      "epoch": 0.46,
-      "learning_rate": 4.9483030046130576e-05,
-      "loss": 0.3225,
+      "epoch": 0.47590730529426606,
+      "grad_norm": 0.1829308420419693,
+      "learning_rate": 4.944546494588758e-05,
+      "loss": 0.4313,
       "step": 13205
     },
     {
-      "epoch": 0.46,
-      "learning_rate": 4.94824535578162e-05,
-      "loss": 0.3187,
+      "epoch": 0.4760875049554907,
+      "grad_norm": 0.1804686188697815,
+      "learning_rate": 4.944485356590317e-05,
+      "loss": 0.4562,
       "step": 13210
     },
     {
-      "epoch": 0.46,
-      "learning_rate": 4.94818767516133e-05,
-      "loss": 0.3131,
+      "epoch": 0.4762677046167153,
+      "grad_norm": 0.17455480992794037,
+      "learning_rate": 4.9444241852862366e-05,
+      "loss": 0.4468,
       "step": 13215
     },
     {
-      "epoch": 0.47,
-      "learning_rate": 4.948129962752937e-05,
-      "loss": 0.3162,
+      "epoch": 0.47644790427794,
+      "grad_norm": 0.18920566141605377,
+      "learning_rate": 4.944362980677349e-05,
+      "loss": 0.4687,
       "step": 13220
     },
     {
-      "epoch": 0.47,
-      "learning_rate": 4.948072218557189e-05,
-      "loss": 0.3216,
+      "epoch": 0.4766281039391646,
+      "grad_norm": 0.19825154542922974,
+      "learning_rate": 4.9443017427644906e-05,
+      "loss": 0.4629,
       "step": 13225
     },
     {
-      "epoch": 0.47,
-      "learning_rate": 4.948014442574837e-05,
-      "loss": 0.3128,
+      "epoch": 0.4768083036003892,
+      "grad_norm": 0.1756822019815445,
+      "learning_rate": 4.944240471548493e-05,
+      "loss": 0.4885,
       "step": 13230
     },
     {
-      "epoch": 0.47,
-      "learning_rate": 4.947956634806631e-05,
-      "loss": 0.2955,
+      "epoch": 0.47698850326161385,
+      "grad_norm": 0.20122404396533966,
+      "learning_rate": 4.944179167030193e-05,
+      "loss": 0.4641,
       "step": 13235
     },
     {
-      "epoch": 0.47,
-      "learning_rate": 4.947898795253321e-05,
-      "loss": 0.3057,
+      "epoch": 0.47716870292283853,
+      "grad_norm": 0.17638641595840454,
+      "learning_rate": 4.944117829210424e-05,
+      "loss": 0.4641,
       "step": 13240
     },
     {
-      "epoch": 0.47,
-      "learning_rate": 4.947840923915659e-05,
-      "loss": 0.315,
+      "epoch": 0.47734890258406315,
+      "grad_norm": 0.20242132246494293,
+      "learning_rate": 4.944056458090024e-05,
+      "loss": 0.4815,
       "step": 13245
     },
     {
-      "epoch": 0.47,
-      "learning_rate": 4.9477830207943944e-05,
-      "loss": 0.3163,
+      "epoch": 0.4775291022452878,
+      "grad_norm": 0.18638722598552704,
+      "learning_rate": 4.943995053669827e-05,
+      "loss": 0.4504,
       "step": 13250
     },
     {
-      "epoch": 0.47,
-      "learning_rate": 4.947725085890281e-05,
-      "loss": 0.3074,
+      "epoch": 0.4777093019065124,
+      "grad_norm": 0.16893352568149567,
+      "learning_rate": 4.943933615950671e-05,
+      "loss": 0.4275,
       "step": 13255
     },
     {
-      "epoch": 0.47,
-      "learning_rate": 4.947667119204071e-05,
-      "loss": 0.3027,
+      "epoch": 0.4778895015677371,
+      "grad_norm": 0.16085629165172577,
+      "learning_rate": 4.9438721449333925e-05,
+      "loss": 0.4648,
       "step": 13260
     },
     {
-      "epoch": 0.47,
-      "learning_rate": 4.947609120736516e-05,
-      "loss": 0.3092,
+      "epoch": 0.4780697012289617,
+      "grad_norm": 0.1869862973690033,
+      "learning_rate": 4.943810640618829e-05,
+      "loss": 0.4415,
       "step": 13265
     },
     {
-      "epoch": 0.47,
-      "learning_rate": 4.94755109048837e-05,
-      "loss": 0.3223,
+      "epoch": 0.4782499008901863,
+      "grad_norm": 0.16547581553459167,
+      "learning_rate": 4.9437491030078185e-05,
+      "loss": 0.5078,
       "step": 13270
     },
     {
-      "epoch": 0.47,
-      "learning_rate": 4.947493028460386e-05,
-      "loss": 0.317,
+      "epoch": 0.47843010055141094,
+      "grad_norm": 0.19443146884441376,
+      "learning_rate": 4.9436875321012e-05,
+      "loss": 0.4237,
       "step": 13275
     },
     {
-      "epoch": 0.47,
-      "learning_rate": 4.9474349346533175e-05,
-      "loss": 0.3389,
+      "epoch": 0.4786103002126356,
+      "grad_norm": 0.1601037085056305,
+      "learning_rate": 4.943625927899812e-05,
+      "loss": 0.4692,
       "step": 13280
     },
     {
-      "epoch": 0.47,
-      "learning_rate": 4.947376809067921e-05,
-      "loss": 0.3291,
+      "epoch": 0.47879049987386024,
+      "grad_norm": 0.17408448457717896,
+      "learning_rate": 4.943564290404494e-05,
+      "loss": 0.4617,
       "step": 13285
     },
     {
-      "epoch": 0.47,
-      "learning_rate": 4.947318651704948e-05,
-      "loss": 0.278,
+      "epoch": 0.47897069953508487,
+      "grad_norm": 0.17495988309383392,
+      "learning_rate": 4.943502619616085e-05,
+      "loss": 0.4349,
       "step": 13290
     },
     {
-      "epoch": 0.47,
-      "learning_rate": 4.947260462565155e-05,
-      "loss": 0.3282,
+      "epoch": 0.4791508991963095,
+      "grad_norm": 0.18326228857040405,
+      "learning_rate": 4.9434409155354266e-05,
+      "loss": 0.4518,
       "step": 13295
     },
     {
-      "epoch": 0.47,
-      "learning_rate": 4.947202241649298e-05,
-      "loss": 0.3176,
+      "epoch": 0.47933109885753417,
+      "grad_norm": 0.14436647295951843,
+      "learning_rate": 4.9433791781633584e-05,
+      "loss": 0.4689,
       "step": 13300
     },
     {
-      "epoch": 0.47,
-      "learning_rate": 4.9471439889581325e-05,
-      "loss": 0.3368,
+      "epoch": 0.4795112985187588,
+      "grad_norm": 0.1709493100643158,
+      "learning_rate": 4.9433174075007216e-05,
+      "loss": 0.4747,
       "step": 13305
     },
     {
-      "epoch": 0.47,
-      "learning_rate": 4.947085704492416e-05,
-      "loss": 0.3488,
+      "epoch": 0.4796914981799834,
+      "grad_norm": 0.19501015543937683,
+      "learning_rate": 4.943255603548359e-05,
+      "loss": 0.4446,
       "step": 13310
     },
     {
-      "epoch": 0.47,
-      "learning_rate": 4.947027388252903e-05,
-      "loss": 0.3198,
+      "epoch": 0.47987169784120803,
+      "grad_norm": 0.16329129040241241,
+      "learning_rate": 4.943193766307111e-05,
+      "loss": 0.4545,
       "step": 13315
     },
     {
-      "epoch": 0.47,
-      "learning_rate": 4.9469690402403525e-05,
-      "loss": 0.3155,
+      "epoch": 0.4800518975024327,
+      "grad_norm": 0.16876693069934845,
+      "learning_rate": 4.9431318957778214e-05,
+      "loss": 0.4874,
       "step": 13320
     },
     {
-      "epoch": 0.47,
-      "learning_rate": 4.946910660455522e-05,
-      "loss": 0.3389,
+      "epoch": 0.48023209716365733,
+      "grad_norm": 0.18165095150470734,
+      "learning_rate": 4.943069991961333e-05,
+      "loss": 0.4349,
       "step": 13325
     },
     {
-      "epoch": 0.47,
-      "learning_rate": 4.946852248899169e-05,
-      "loss": 0.3143,
+      "epoch": 0.48041229682488196,
+      "grad_norm": 0.17460200190544128,
+      "learning_rate": 4.9430080548584884e-05,
+      "loss": 0.4678,
       "step": 13330
     },
     {
-      "epoch": 0.47,
-      "learning_rate": 4.946793805572052e-05,
-      "loss": 0.3292,
+      "epoch": 0.4805924964861066,
+      "grad_norm": 0.14548338949680328,
+      "learning_rate": 4.9429460844701325e-05,
+      "loss": 0.4717,
       "step": 13335
     },
     {
-      "epoch": 0.47,
-      "learning_rate": 4.94673533047493e-05,
-      "loss": 0.3163,
+      "epoch": 0.48077269614733126,
+      "grad_norm": 0.17974305152893066,
+      "learning_rate": 4.942884080797109e-05,
+      "loss": 0.4617,
       "step": 13340
     },
     {
-      "epoch": 0.47,
-      "learning_rate": 4.946676823608562e-05,
-      "loss": 0.325,
+      "epoch": 0.4809528958085559,
+      "grad_norm": 0.194953054189682,
+      "learning_rate": 4.942822043840262e-05,
+      "loss": 0.4724,
       "step": 13345
     },
     {
-      "epoch": 0.47,
-      "learning_rate": 4.946618284973708e-05,
-      "loss": 0.3158,
+      "epoch": 0.4811330954697805,
+      "grad_norm": 0.16658879816532135,
+      "learning_rate": 4.942759973600439e-05,
+      "loss": 0.4236,
       "step": 13350
     },
     {
-      "epoch": 0.47,
-      "learning_rate": 4.946559714571127e-05,
-      "loss": 0.3372,
+      "epoch": 0.4813132951310052,
+      "grad_norm": 0.18458372354507446,
+      "learning_rate": 4.9426978700784834e-05,
+      "loss": 0.4801,
       "step": 13355
     },
     {
-      "epoch": 0.47,
-      "learning_rate": 4.946501112401582e-05,
-      "loss": 0.3333,
+      "epoch": 0.4814934947922298,
+      "grad_norm": 0.18043503165245056,
+      "learning_rate": 4.942635733275243e-05,
+      "loss": 0.4592,
       "step": 13360
     },
     {
-      "epoch": 0.47,
-      "learning_rate": 4.9464424784658314e-05,
-      "loss": 0.3234,
+      "epoch": 0.4816736944534544,
+      "grad_norm": 0.2545734643936157,
+      "learning_rate": 4.942573563191563e-05,
+      "loss": 0.4656,
       "step": 13365
     },
     {
-      "epoch": 0.47,
-      "learning_rate": 4.946383812764638e-05,
-      "loss": 0.2952,
+      "epoch": 0.48185389411467905,
+      "grad_norm": 0.18358397483825684,
+      "learning_rate": 4.9425113598282916e-05,
+      "loss": 0.4401,
       "step": 13370
     },
     {
-      "epoch": 0.47,
-      "learning_rate": 4.946325115298763e-05,
-      "loss": 0.3085,
+      "epoch": 0.4820340937759037,
+      "grad_norm": 0.17365938425064087,
+      "learning_rate": 4.942449123186274e-05,
+      "loss": 0.4459,
       "step": 13375
     },
     {
-      "epoch": 0.47,
-      "learning_rate": 4.946266386068968e-05,
-      "loss": 0.2821,
+      "epoch": 0.48221429343712835,
+      "grad_norm": 0.15696145594120026,
+      "learning_rate": 4.942386853266362e-05,
+      "loss": 0.4618,
       "step": 13380
     },
     {
-      "epoch": 0.47,
-      "learning_rate": 4.946207625076017e-05,
-      "loss": 0.2997,
+      "epoch": 0.48239449309835297,
+      "grad_norm": 0.16444018483161926,
+      "learning_rate": 4.942324550069402e-05,
+      "loss": 0.4885,
       "step": 13385
     },
     {
-      "epoch": 0.47,
-      "learning_rate": 4.946148832320672e-05,
-      "loss": 0.3096,
+      "epoch": 0.4825746927595776,
+      "grad_norm": 0.14012210071086884,
+      "learning_rate": 4.942262213596241e-05,
+      "loss": 0.4398,
       "step": 13390
     },
     {
-      "epoch": 0.47,
-      "learning_rate": 4.9460900078036955e-05,
-      "loss": 0.3095,
+      "epoch": 0.4827548924208023,
+      "grad_norm": 0.23050503432750702,
+      "learning_rate": 4.942199843847732e-05,
+      "loss": 0.4899,
       "step": 13395
     },
     {
-      "epoch": 0.47,
-      "learning_rate": 4.9460311515258526e-05,
-      "loss": 0.3151,
+      "epoch": 0.4829350920820269,
+      "grad_norm": 0.15428675711154938,
+      "learning_rate": 4.942137440824723e-05,
+      "loss": 0.4707,
       "step": 13400
     },
     {
-      "epoch": 0.47,
-      "learning_rate": 4.945972263487908e-05,
-      "loss": 0.341,
+      "epoch": 0.4831152917432515,
+      "grad_norm": 0.2107590287923813,
+      "learning_rate": 4.9420750045280636e-05,
+      "loss": 0.4507,
       "step": 13405
     },
     {
-      "epoch": 0.47,
-      "learning_rate": 4.945913343690625e-05,
-      "loss": 0.3015,
+      "epoch": 0.48329549140447614,
+      "grad_norm": 0.1626407951116562,
+      "learning_rate": 4.942012534958605e-05,
+      "loss": 0.4398,
       "step": 13410
     },
     {
-      "epoch": 0.47,
-      "learning_rate": 4.9458543921347686e-05,
-      "loss": 0.339,
+      "epoch": 0.4834756910657008,
+      "grad_norm": 0.14003750681877136,
+      "learning_rate": 4.9419500321171987e-05,
+      "loss": 0.4246,
       "step": 13415
     },
     {
-      "epoch": 0.47,
-      "learning_rate": 4.9457954088211055e-05,
-      "loss": 0.3195,
+      "epoch": 0.48365589072692544,
+      "grad_norm": 0.16057373583316803,
+      "learning_rate": 4.9418874960046954e-05,
+      "loss": 0.4636,
       "step": 13420
     },
     {
-      "epoch": 0.47,
-      "learning_rate": 4.945736393750401e-05,
-      "loss": 0.3124,
+      "epoch": 0.48383609038815006,
+      "grad_norm": 0.1851695328950882,
+      "learning_rate": 4.941824926621948e-05,
+      "loss": 0.4292,
       "step": 13425
     },
     {
-      "epoch": 0.47,
-      "learning_rate": 4.9456773469234206e-05,
-      "loss": 0.3075,
+      "epoch": 0.4840162900493747,
+      "grad_norm": 0.1717042773962021,
+      "learning_rate": 4.941762323969809e-05,
+      "loss": 0.4782,
       "step": 13430
     },
     {
-      "epoch": 0.47,
-      "learning_rate": 4.9456182683409316e-05,
-      "loss": 0.3203,
+      "epoch": 0.48419648971059936,
+      "grad_norm": 0.16300024092197418,
+      "learning_rate": 4.9416996880491305e-05,
+      "loss": 0.4498,
       "step": 13435
     },
     {
-      "epoch": 0.47,
-      "learning_rate": 4.945559158003702e-05,
-      "loss": 0.3248,
+      "epoch": 0.484376689371824,
+      "grad_norm": 0.1653520166873932,
+      "learning_rate": 4.941637018860767e-05,
+      "loss": 0.417,
       "step": 13440
     },
     {
-      "epoch": 0.47,
-      "learning_rate": 4.945500015912498e-05,
-      "loss": 0.3142,
+      "epoch": 0.4845568890330486,
+      "grad_norm": 0.14384515583515167,
+      "learning_rate": 4.941574316405572e-05,
+      "loss": 0.4253,
       "step": 13445
     },
     {
-      "epoch": 0.47,
-      "learning_rate": 4.945440842068087e-05,
-      "loss": 0.3309,
+      "epoch": 0.48473708869427323,
+      "grad_norm": 0.18884818255901337,
+      "learning_rate": 4.9415115806843993e-05,
+      "loss": 0.4521,
       "step": 13450
     },
     {
-      "epoch": 0.47,
-      "learning_rate": 4.94538163647124e-05,
-      "loss": 0.3139,
+      "epoch": 0.4849172883554979,
+      "grad_norm": 0.14921367168426514,
+      "learning_rate": 4.941448811698104e-05,
+      "loss": 0.4797,
       "step": 13455
     },
     {
-      "epoch": 0.47,
-      "learning_rate": 4.945322399122723e-05,
-      "loss": 0.2987,
+      "epoch": 0.48509748801672253,
+      "grad_norm": 0.16940531134605408,
+      "learning_rate": 4.9413860094475414e-05,
+      "loss": 0.4575,
       "step": 13460
     },
     {
-      "epoch": 0.47,
-      "learning_rate": 4.945263130023306e-05,
-      "loss": 0.3205,
+      "epoch": 0.48527768767794716,
+      "grad_norm": 0.23463140428066254,
+      "learning_rate": 4.9413231739335664e-05,
+      "loss": 0.4568,
       "step": 13465
     },
     {
-      "epoch": 0.47,
-      "learning_rate": 4.945203829173759e-05,
-      "loss": 0.3159,
+      "epoch": 0.4854578873391718,
+      "grad_norm": 0.14103110134601593,
+      "learning_rate": 4.9412603051570364e-05,
+      "loss": 0.4384,
       "step": 13470
     },
     {
-      "epoch": 0.47,
-      "learning_rate": 4.945144496574852e-05,
-      "loss": 0.3238,
+      "epoch": 0.48563808700039646,
+      "grad_norm": 0.1720830202102661,
+      "learning_rate": 4.941197403118808e-05,
+      "loss": 0.4834,
       "step": 13475
     },
     {
-      "epoch": 0.47,
-      "learning_rate": 4.945085132227354e-05,
-      "loss": 0.3118,
+      "epoch": 0.4858182866616211,
+      "grad_norm": 0.1815214455127716,
+      "learning_rate": 4.9411344678197366e-05,
+      "loss": 0.4557,
       "step": 13480
     },
     {
-      "epoch": 0.47,
-      "learning_rate": 4.945025736132038e-05,
-      "loss": 0.3195,
+      "epoch": 0.4859984863228457,
+      "grad_norm": 0.18314550817012787,
+      "learning_rate": 4.941071499260681e-05,
+      "loss": 0.4589,
       "step": 13485
     },
     {
-      "epoch": 0.47,
-      "learning_rate": 4.9449663082896744e-05,
-      "loss": 0.2916,
+      "epoch": 0.4861786859840703,
+      "grad_norm": 0.1780148446559906,
+      "learning_rate": 4.9410084974424994e-05,
+      "loss": 0.4338,
       "step": 13490
     },
     {
-      "epoch": 0.47,
-      "learning_rate": 4.944906848701034e-05,
-      "loss": 0.3271,
+      "epoch": 0.486358885645295,
+      "grad_norm": 0.17320701479911804,
+      "learning_rate": 4.940945462366049e-05,
+      "loss": 0.4356,
       "step": 13495
     },
     {
-      "epoch": 0.47,
-      "learning_rate": 4.94484735736689e-05,
-      "loss": 0.3491,
+      "epoch": 0.4865390853065196,
+      "grad_norm": 0.1759449988603592,
+      "learning_rate": 4.94088239403219e-05,
+      "loss": 0.4147,
       "step": 13500
     },
     {
-      "epoch": 0.47,
-      "eval_loss": 0.31191739439964294,
-      "eval_runtime": 10.5398,
-      "eval_samples_per_second": 9.488,
-      "eval_steps_per_second": 9.488,
+      "epoch": 0.4865390853065196,
+      "eval_loss": 0.47342750430107117,
+      "eval_runtime": 3.5547,
+      "eval_samples_per_second": 28.131,
+      "eval_steps_per_second": 7.033,
       "step": 13500
     },
     {
-      "epoch": 0.48,
-      "learning_rate": 4.9447878342880134e-05,
-      "loss": 0.3164,
+      "epoch": 0.48671928496774425,
+      "grad_norm": 0.1460382640361786,
+      "learning_rate": 4.94081929244178e-05,
+      "loss": 0.4134,
       "step": 13505
     },
     {
-      "epoch": 0.48,
-      "learning_rate": 4.944728279465179e-05,
-      "loss": 0.31,
+      "epoch": 0.48689948462896887,
+      "grad_norm": 0.20045733451843262,
+      "learning_rate": 4.94075615759568e-05,
+      "loss": 0.4641,
       "step": 13510
     },
     {
-      "epoch": 0.48,
-      "learning_rate": 4.9446686928991574e-05,
-      "loss": 0.2992,
+      "epoch": 0.48707968429019355,
+      "grad_norm": 0.1686198115348816,
+      "learning_rate": 4.94069298949475e-05,
+      "loss": 0.4628,
       "step": 13515
     },
     {
-      "epoch": 0.48,
-      "learning_rate": 4.944609074590725e-05,
-      "loss": 0.3281,
+      "epoch": 0.48725988395141817,
+      "grad_norm": 0.16252323985099792,
+      "learning_rate": 4.9406297881398504e-05,
+      "loss": 0.4763,
       "step": 13520
     },
     {
-      "epoch": 0.48,
-      "learning_rate": 4.944549424540655e-05,
-      "loss": 0.3215,
+      "epoch": 0.4874400836126428,
+      "grad_norm": 0.2166275978088379,
+      "learning_rate": 4.940566553531843e-05,
+      "loss": 0.4497,
       "step": 13525
     },
     {
-      "epoch": 0.48,
-      "learning_rate": 4.944489742749721e-05,
-      "loss": 0.3348,
+      "epoch": 0.48762028327386747,
+      "grad_norm": 0.22283247113227844,
+      "learning_rate": 4.940503285671588e-05,
+      "loss": 0.465,
       "step": 13530
     },
     {
-      "epoch": 0.48,
-      "learning_rate": 4.9444300292186994e-05,
-      "loss": 0.323,
+      "epoch": 0.4878004829350921,
+      "grad_norm": 0.11601988226175308,
+      "learning_rate": 4.940439984559949e-05,
+      "loss": 0.3932,
       "step": 13535
     },
     {
-      "epoch": 0.48,
-      "learning_rate": 4.944370283948365e-05,
-      "loss": 0.3194,
+      "epoch": 0.4879806825963167,
+      "grad_norm": 0.17983366549015045,
+      "learning_rate": 4.940376650197787e-05,
+      "loss": 0.4426,
       "step": 13540
     },
     {
-      "epoch": 0.48,
-      "learning_rate": 4.9443105069394924e-05,
-      "loss": 0.34,
+      "epoch": 0.48816088225754134,
+      "grad_norm": 0.2088562697172165,
+      "learning_rate": 4.940313282585967e-05,
+      "loss": 0.4687,
       "step": 13545
     },
     {
-      "epoch": 0.48,
-      "learning_rate": 4.944250698192859e-05,
-      "loss": 0.3085,
+      "epoch": 0.488341081918766,
+      "grad_norm": 0.17913992702960968,
+      "learning_rate": 4.940249881725349e-05,
+      "loss": 0.4504,
       "step": 13550
     },
     {
-      "epoch": 0.48,
-      "learning_rate": 4.944190857709242e-05,
-      "loss": 0.314,
+      "epoch": 0.48852128157999064,
+      "grad_norm": 0.18297719955444336,
+      "learning_rate": 4.9401864476168e-05,
+      "loss": 0.4756,
       "step": 13555
     },
     {
-      "epoch": 0.48,
-      "learning_rate": 4.944130985489417e-05,
-      "loss": 0.3261,
+      "epoch": 0.48870148124121526,
+      "grad_norm": 0.18946805596351624,
+      "learning_rate": 4.9401229802611826e-05,
+      "loss": 0.4461,
       "step": 13560
     },
     {
-      "epoch": 0.48,
-      "learning_rate": 4.944071081534161e-05,
-      "loss": 0.3415,
+      "epoch": 0.4888816809024399,
+      "grad_norm": 0.19642286002635956,
+      "learning_rate": 4.9400594796593626e-05,
+      "loss": 0.4813,
       "step": 13565
     },
     {
-      "epoch": 0.48,
-      "learning_rate": 4.944011145844254e-05,
-      "loss": 0.3244,
+      "epoch": 0.48906188056366456,
+      "grad_norm": 0.21718788146972656,
+      "learning_rate": 4.940008655241221e-05,
+      "loss": 0.4587,
       "step": 13570
     },
     {
-      "epoch": 0.48,
-      "learning_rate": 4.9439511784204715e-05,
-      "loss": 0.3109,
+      "epoch": 0.4892420802248892,
+      "grad_norm": 0.2410500943660736,
+      "learning_rate": 4.939945094798416e-05,
+      "loss": 0.4804,
       "step": 13575
     },
     {
-      "epoch": 0.48,
-      "learning_rate": 4.943891179263595e-05,
-      "loss": 0.3116,
+      "epoch": 0.4894222798861138,
+      "grad_norm": 0.1772763729095459,
+      "learning_rate": 4.93988150111183e-05,
+      "loss": 0.4455,
       "step": 13580
     },
     {
-      "epoch": 0.48,
-      "learning_rate": 4.9438311483744013e-05,
-      "loss": 0.3322,
+      "epoch": 0.48960247954733843,
+      "grad_norm": 0.16895850002765656,
+      "learning_rate": 4.939817874182333e-05,
+      "loss": 0.4365,
       "step": 13585
     },
     {
-      "epoch": 0.48,
-      "learning_rate": 4.9437710857536715e-05,
-      "loss": 0.3271,
+      "epoch": 0.4897826792085631,
+      "grad_norm": 0.2072199136018753,
+      "learning_rate": 4.939754214010788e-05,
+      "loss": 0.4403,
       "step": 13590
     },
     {
-      "epoch": 0.48,
-      "learning_rate": 4.943710991402184e-05,
-      "loss": 0.3172,
+      "epoch": 0.48996287886978773,
+      "grad_norm": 0.16987258195877075,
+      "learning_rate": 4.939690520598065e-05,
+      "loss": 0.449,
       "step": 13595
     },
     {
-      "epoch": 0.48,
-      "learning_rate": 4.94365086532072e-05,
-      "loss": 0.2977,
+      "epoch": 0.49014307853101236,
+      "grad_norm": 0.16706013679504395,
+      "learning_rate": 4.9396267939450316e-05,
+      "loss": 0.4429,
       "step": 13600
     },
     {
-      "epoch": 0.48,
-      "learning_rate": 4.9435907075100595e-05,
-      "loss": 0.2983,
+      "epoch": 0.490323278192237,
+      "grad_norm": 0.15251483023166656,
+      "learning_rate": 4.939563034052555e-05,
+      "loss": 0.4386,
       "step": 13605
     },
     {
-      "epoch": 0.48,
-      "learning_rate": 4.943530517970985e-05,
-      "loss": 0.3503,
+      "epoch": 0.49050347785346166,
+      "grad_norm": 0.17916177213191986,
+      "learning_rate": 4.9394992409215036e-05,
+      "loss": 0.4344,
       "step": 13610
     },
     {
-      "epoch": 0.48,
-      "learning_rate": 4.943470296704276e-05,
-      "loss": 0.326,
+      "epoch": 0.4906836775146863,
+      "grad_norm": 0.13827458024024963,
+      "learning_rate": 4.939435414552748e-05,
+      "loss": 0.4667,
       "step": 13615
     },
     {
-      "epoch": 0.48,
-      "learning_rate": 4.943410043710715e-05,
-      "loss": 0.3147,
+      "epoch": 0.4908638771759109,
+      "grad_norm": 0.1721857637166977,
+      "learning_rate": 4.939371554947156e-05,
+      "loss": 0.4614,
       "step": 13620
     },
     {
-      "epoch": 0.48,
-      "learning_rate": 4.9433497589910864e-05,
-      "loss": 0.3018,
+      "epoch": 0.4910440768371355,
+      "grad_norm": 0.19055204093456268,
+      "learning_rate": 4.9393076621056e-05,
+      "loss": 0.4408,
       "step": 13625
     },
     {
-      "epoch": 0.48,
-      "learning_rate": 4.9432894425461704e-05,
-      "loss": 0.3277,
+      "epoch": 0.4912242764983602,
+      "grad_norm": 0.1874268501996994,
+      "learning_rate": 4.93924373602895e-05,
+      "loss": 0.4647,
       "step": 13630
     },
     {
-      "epoch": 0.48,
-      "learning_rate": 4.943229094376752e-05,
-      "loss": 0.3264,
+      "epoch": 0.4914044761595848,
+      "grad_norm": 0.1848086565732956,
+      "learning_rate": 4.9391797767180755e-05,
+      "loss": 0.4801,
       "step": 13635
     },
     {
-      "epoch": 0.48,
-      "learning_rate": 4.9431687144836136e-05,
-      "loss": 0.3253,
+      "epoch": 0.49158467582080945,
+      "grad_norm": 0.1567952185869217,
+      "learning_rate": 4.939115784173849e-05,
+      "loss": 0.4194,
       "step": 13640
     },
     {
-      "epoch": 0.48,
-      "learning_rate": 4.9431083028675394e-05,
-      "loss": 0.3422,
+      "epoch": 0.49176487548203407,
+      "grad_norm": 0.16139036417007446,
+      "learning_rate": 4.9390517583971416e-05,
+      "loss": 0.473,
       "step": 13645
     },
     {
-      "epoch": 0.48,
-      "learning_rate": 4.9430478595293136e-05,
-      "loss": 0.3188,
+      "epoch": 0.49194507514325875,
+      "grad_norm": 0.14114044606685638,
+      "learning_rate": 4.938987699388827e-05,
+      "loss": 0.4326,
       "step": 13650
     },
     {
-      "epoch": 0.48,
-      "learning_rate": 4.942987384469723e-05,
-      "loss": 0.3083,
+      "epoch": 0.49212527480448337,
+      "grad_norm": 0.20750269293785095,
+      "learning_rate": 4.938923607149777e-05,
+      "loss": 0.4792,
       "step": 13655
     },
     {
-      "epoch": 0.48,
-      "learning_rate": 4.942926877689549e-05,
-      "loss": 0.3319,
+      "epoch": 0.492305474465708,
+      "grad_norm": 0.1520867943763733,
+      "learning_rate": 4.938859481680865e-05,
+      "loss": 0.453,
       "step": 13660
     },
     {
-      "epoch": 0.48,
-      "learning_rate": 4.942866339189581e-05,
-      "loss": 0.2984,
+      "epoch": 0.4924856741269326,
+      "grad_norm": 0.17390841245651245,
+      "learning_rate": 4.9387953229829644e-05,
+      "loss": 0.4304,
       "step": 13665
     },
     {
-      "epoch": 0.48,
-      "learning_rate": 4.942805768970603e-05,
-      "loss": 0.2981,
+      "epoch": 0.4926658737881573,
+      "grad_norm": 0.14850212633609772,
+      "learning_rate": 4.938731131056949e-05,
+      "loss": 0.4328,
       "step": 13670
     },
     {
-      "epoch": 0.48,
-      "learning_rate": 4.9427451670334014e-05,
-      "loss": 0.3058,
+      "epoch": 0.4928460734493819,
+      "grad_norm": 0.15505053102970123,
+      "learning_rate": 4.938666905903696e-05,
+      "loss": 0.4195,
       "step": 13675
     },
     {
-      "epoch": 0.48,
-      "learning_rate": 4.942684533378764e-05,
-      "loss": 0.3059,
+      "epoch": 0.49302627311060654,
+      "grad_norm": 0.1582787036895752,
+      "learning_rate": 4.938602647524077e-05,
+      "loss": 0.4119,
       "step": 13680
     },
     {
-      "epoch": 0.48,
-      "learning_rate": 4.942623868007478e-05,
-      "loss": 0.3228,
+      "epoch": 0.49320647277183116,
+      "grad_norm": 0.2123645544052124,
+      "learning_rate": 4.93853835591897e-05,
+      "loss": 0.4687,
       "step": 13685
     },
     {
-      "epoch": 0.48,
-      "learning_rate": 4.9425631709203314e-05,
-      "loss": 0.3068,
+      "epoch": 0.49338667243305584,
+      "grad_norm": 0.19299419224262238,
+      "learning_rate": 4.93847403108925e-05,
+      "loss": 0.4948,
       "step": 13690
     },
     {
-      "epoch": 0.48,
-      "learning_rate": 4.94250244211811e-05,
-      "loss": 0.3145,
+      "epoch": 0.49356687209428046,
+      "grad_norm": 0.19016866385936737,
+      "learning_rate": 4.938409673035793e-05,
+      "loss": 0.4477,
       "step": 13695
     },
     {
-      "epoch": 0.48,
-      "learning_rate": 4.9424416816016055e-05,
-      "loss": 0.3242,
+      "epoch": 0.4937470717555051,
+      "grad_norm": 0.1360589861869812,
+      "learning_rate": 4.938345281759476e-05,
+      "loss": 0.4734,
       "step": 13700
     },
     {
-      "epoch": 0.48,
-      "learning_rate": 4.942380889371605e-05,
-      "loss": 0.3058,
+      "epoch": 0.49392727141672976,
+      "grad_norm": 0.15430140495300293,
+      "learning_rate": 4.9382808572611775e-05,
+      "loss": 0.4432,
       "step": 13705
     },
     {
-      "epoch": 0.48,
-      "learning_rate": 4.942320065428898e-05,
-      "loss": 0.3208,
+      "epoch": 0.4941074710779544,
+      "grad_norm": 0.17401854693889618,
+      "learning_rate": 4.938216399541773e-05,
+      "loss": 0.4308,
       "step": 13710
     },
     {
-      "epoch": 0.48,
-      "learning_rate": 4.9422592097742745e-05,
-      "loss": 0.332,
+      "epoch": 0.494287670739179,
+      "grad_norm": 0.17819242179393768,
+      "learning_rate": 4.9381519086021434e-05,
+      "loss": 0.449,
       "step": 13715
     },
     {
-      "epoch": 0.48,
-      "learning_rate": 4.942198322408526e-05,
-      "loss": 0.3246,
+      "epoch": 0.49446787040040363,
+      "grad_norm": 0.1599666327238083,
+      "learning_rate": 4.9380873844431654e-05,
+      "loss": 0.4336,
       "step": 13720
     },
     {
-      "epoch": 0.48,
-      "learning_rate": 4.94213740333244e-05,
-      "loss": 0.324,
+      "epoch": 0.4946480700616283,
+      "grad_norm": 0.13881821930408478,
+      "learning_rate": 4.938022827065719e-05,
+      "loss": 0.4488,
       "step": 13725
     },
     {
-      "epoch": 0.48,
-      "learning_rate": 4.94207645254681e-05,
-      "loss": 0.3244,
+      "epoch": 0.49482826972285293,
+      "grad_norm": 0.1781865805387497,
+      "learning_rate": 4.937958236470684e-05,
+      "loss": 0.4461,
       "step": 13730
     },
     {
-      "epoch": 0.48,
-      "learning_rate": 4.942015470052426e-05,
-      "loss": 0.2947,
+      "epoch": 0.49500846938407755,
+      "grad_norm": 0.1789911985397339,
+      "learning_rate": 4.93789361265894e-05,
+      "loss": 0.4537,
       "step": 13735
     },
     {
-      "epoch": 0.48,
-      "learning_rate": 4.9419544558500813e-05,
-      "loss": 0.331,
+      "epoch": 0.4951886690453022,
+      "grad_norm": 0.1587119698524475,
+      "learning_rate": 4.9378289556313673e-05,
+      "loss": 0.4544,
       "step": 13740
     },
     {
-      "epoch": 0.48,
-      "learning_rate": 4.941893409940567e-05,
-      "loss": 0.3408,
+      "epoch": 0.49536886870652685,
+      "grad_norm": 0.1573282778263092,
+      "learning_rate": 4.9377642653888464e-05,
+      "loss": 0.396,
       "step": 13745
     },
     {
-      "epoch": 0.48,
-      "learning_rate": 4.9418323323246764e-05,
-      "loss": 0.3019,
+      "epoch": 0.4955490683677515,
+      "grad_norm": 0.13113892078399658,
+      "learning_rate": 4.93769954193226e-05,
+      "loss": 0.4652,
       "step": 13750
     },
     {
-      "epoch": 0.48,
-      "learning_rate": 4.9417712230032016e-05,
-      "loss": 0.3216,
+      "epoch": 0.4957292680289761,
+      "grad_norm": 0.1477142572402954,
+      "learning_rate": 4.9376347852624895e-05,
+      "loss": 0.4209,
       "step": 13755
     },
     {
-      "epoch": 0.48,
-      "learning_rate": 4.941710081976937e-05,
-      "loss": 0.3146,
+      "epoch": 0.4959094676902007,
+      "grad_norm": 0.1767813116312027,
+      "learning_rate": 4.937569995380417e-05,
+      "loss": 0.4656,
       "step": 13760
     },
     {
-      "epoch": 0.48,
-      "learning_rate": 4.941648909246676e-05,
-      "loss": 0.2873,
+      "epoch": 0.4960896673514254,
+      "grad_norm": 0.14544841647148132,
+      "learning_rate": 4.937505172286925e-05,
+      "loss": 0.453,
       "step": 13765
     },
     {
-      "epoch": 0.48,
-      "learning_rate": 4.941587704813213e-05,
-      "loss": 0.3238,
+      "epoch": 0.49626986701265,
+      "grad_norm": 0.16864298284053802,
+      "learning_rate": 4.9374403159828965e-05,
+      "loss": 0.4242,
       "step": 13770
     },
     {
-      "epoch": 0.48,
-      "learning_rate": 4.941526468677343e-05,
-      "loss": 0.3233,
+      "epoch": 0.49645006667387465,
+      "grad_norm": 0.16903114318847656,
+      "learning_rate": 4.9373754264692164e-05,
+      "loss": 0.4519,
       "step": 13775
     },
     {
-      "epoch": 0.48,
-      "learning_rate": 4.9414652008398605e-05,
-      "loss": 0.3045,
+      "epoch": 0.49663026633509927,
+      "grad_norm": 0.21901170909404755,
+      "learning_rate": 4.937310503746767e-05,
+      "loss": 0.4414,
       "step": 13780
     },
     {
-      "epoch": 0.48,
-      "learning_rate": 4.9414039013015615e-05,
-      "loss": 0.3186,
+      "epoch": 0.49681046599632395,
+      "grad_norm": 0.18153007328510284,
+      "learning_rate": 4.937245547816435e-05,
+      "loss": 0.4262,
       "step": 13785
     },
     {
-      "epoch": 0.49,
-      "learning_rate": 4.9413425700632415e-05,
-      "loss": 0.3047,
+      "epoch": 0.49699066565754857,
+      "grad_norm": 0.16006413102149963,
+      "learning_rate": 4.937180558679104e-05,
+      "loss": 0.4462,
       "step": 13790
     },
     {
-      "epoch": 0.49,
-      "learning_rate": 4.9412812071256974e-05,
-      "loss": 0.3117,
+      "epoch": 0.4971708653187732,
+      "grad_norm": 0.17855258285999298,
+      "learning_rate": 4.937115536335659e-05,
+      "loss": 0.4327,
       "step": 13795
     },
     {
-      "epoch": 0.49,
-      "learning_rate": 4.941219812489726e-05,
-      "loss": 0.3269,
+      "epoch": 0.4973510649799978,
+      "grad_norm": 0.1678069531917572,
+      "learning_rate": 4.937050480786987e-05,
+      "loss": 0.4655,
       "step": 13800
     },
     {
-      "epoch": 0.49,
-      "learning_rate": 4.9411583861561236e-05,
-      "loss": 0.303,
+      "epoch": 0.4975312646412225,
+      "grad_norm": 0.13157707452774048,
+      "learning_rate": 4.936985392033975e-05,
+      "loss": 0.4951,
       "step": 13805
     },
     {
-      "epoch": 0.49,
-      "learning_rate": 4.941096928125688e-05,
-      "loss": 0.3261,
+      "epoch": 0.4977114643024471,
+      "grad_norm": 0.20059573650360107,
+      "learning_rate": 4.936920270077508e-05,
+      "loss": 0.456,
       "step": 13810
     },
     {
-      "epoch": 0.49,
-      "learning_rate": 4.9410354383992186e-05,
-      "loss": 0.2925,
+      "epoch": 0.49789166396367174,
+      "grad_norm": 0.13859055936336517,
+      "learning_rate": 4.936855114918474e-05,
+      "loss": 0.4359,
       "step": 13815
     },
     {
-      "epoch": 0.49,
-      "learning_rate": 4.940973916977512e-05,
-      "loss": 0.3165,
+      "epoch": 0.49807186362489636,
+      "grad_norm": 0.165902242064476,
+      "learning_rate": 4.936789926557761e-05,
+      "loss": 0.4371,
       "step": 13820
     },
     {
-      "epoch": 0.49,
-      "learning_rate": 4.940912363861368e-05,
-      "loss": 0.3106,
+      "epoch": 0.49825206328612104,
+      "grad_norm": 0.1760404258966446,
+      "learning_rate": 4.936724704996257e-05,
+      "loss": 0.4465,
       "step": 13825
     },
     {
-      "epoch": 0.49,
-      "learning_rate": 4.940850779051586e-05,
-      "loss": 0.3322,
+      "epoch": 0.49843226294734566,
+      "grad_norm": 0.16966862976551056,
+      "learning_rate": 4.936659450234851e-05,
+      "loss": 0.4454,
       "step": 13830
     },
     {
-      "epoch": 0.49,
-      "learning_rate": 4.9407891625489644e-05,
-      "loss": 0.3183,
+      "epoch": 0.4986124626085703,
+      "grad_norm": 0.16548436880111694,
+      "learning_rate": 4.936594162274431e-05,
+      "loss": 0.4679,
       "step": 13835
     },
     {
-      "epoch": 0.49,
-      "learning_rate": 4.9407275143543044e-05,
-      "loss": 0.3155,
+      "epoch": 0.4987926622697949,
+      "grad_norm": 0.19853737950325012,
+      "learning_rate": 4.936528841115887e-05,
+      "loss": 0.4545,
       "step": 13840
     },
     {
-      "epoch": 0.49,
-      "learning_rate": 4.940665834468407e-05,
-      "loss": 0.3208,
+      "epoch": 0.4989728619310196,
+      "grad_norm": 0.19320517778396606,
+      "learning_rate": 4.936463486760111e-05,
+      "loss": 0.4758,
       "step": 13845
     },
     {
-      "epoch": 0.49,
-      "learning_rate": 4.940604122892071e-05,
-      "loss": 0.306,
+      "epoch": 0.4991530615922442,
+      "grad_norm": 0.2133074700832367,
+      "learning_rate": 4.936398099207991e-05,
+      "loss": 0.4675,
       "step": 13850
     },
     {
-      "epoch": 0.49,
-      "learning_rate": 4.940542379626099e-05,
-      "loss": 0.3462,
+      "epoch": 0.49933326125346883,
+      "grad_norm": 0.16722871363162994,
+      "learning_rate": 4.936332678460417e-05,
+      "loss": 0.4416,
       "step": 13855
     },
     {
-      "epoch": 0.49,
-      "learning_rate": 4.9404806046712927e-05,
-      "loss": 0.3338,
+      "epoch": 0.4995134609146935,
+      "grad_norm": 0.1773030161857605,
+      "learning_rate": 4.936267224518284e-05,
+      "loss": 0.475,
       "step": 13860
     },
     {
-      "epoch": 0.49,
-      "learning_rate": 4.940418798028455e-05,
-      "loss": 0.3077,
+      "epoch": 0.49969366057591813,
+      "grad_norm": 0.13764159381389618,
+      "learning_rate": 4.936201737382481e-05,
+      "loss": 0.4597,
       "step": 13865
     },
     {
-      "epoch": 0.49,
-      "learning_rate": 4.940356959698386e-05,
-      "loss": 0.3194,
+      "epoch": 0.49987386023714275,
+      "grad_norm": 0.19111090898513794,
+      "learning_rate": 4.9361362170539006e-05,
+      "loss": 0.4601,
       "step": 13870
     },
     {
-      "epoch": 0.49,
-      "learning_rate": 4.9402950896818906e-05,
-      "loss": 0.2823,
+      "epoch": 0.5000540598983674,
+      "grad_norm": 0.13824209570884705,
+      "learning_rate": 4.936070663533436e-05,
+      "loss": 0.4103,
       "step": 13875
     },
     {
-      "epoch": 0.49,
-      "learning_rate": 4.940233187979772e-05,
-      "loss": 0.3186,
+      "epoch": 0.500234259559592,
+      "grad_norm": 0.17304690182209015,
+      "learning_rate": 4.936005076821981e-05,
+      "loss": 0.4446,
       "step": 13880
     },
     {
-      "epoch": 0.49,
-      "learning_rate": 4.940171254592834e-05,
-      "loss": 0.3073,
+      "epoch": 0.5004144592208166,
+      "grad_norm": 0.1318497657775879,
+      "learning_rate": 4.9359394569204274e-05,
+      "loss": 0.4341,
       "step": 13885
     },
     {
-      "epoch": 0.49,
-      "learning_rate": 4.940109289521879e-05,
-      "loss": 0.3308,
+      "epoch": 0.5005946588820414,
+      "grad_norm": 0.1591554433107376,
+      "learning_rate": 4.9358738038296714e-05,
+      "loss": 0.4414,
       "step": 13890
     },
     {
-      "epoch": 0.49,
-      "learning_rate": 4.9400472927677144e-05,
-      "loss": 0.284,
+      "epoch": 0.500774858543266,
+      "grad_norm": 0.18022438883781433,
+      "learning_rate": 4.935808117550605e-05,
+      "loss": 0.4741,
       "step": 13895
     },
     {
-      "epoch": 0.49,
-      "learning_rate": 4.939985264331143e-05,
-      "loss": 0.3212,
+      "epoch": 0.5009550582044906,
+      "grad_norm": 0.16793961822986603,
+      "learning_rate": 4.935742398084127e-05,
+      "loss": 0.4567,
       "step": 13900
     },
     {
-      "epoch": 0.49,
-      "learning_rate": 4.939923204212972e-05,
-      "loss": 0.316,
+      "epoch": 0.5011352578657152,
+      "grad_norm": 0.13464850187301636,
+      "learning_rate": 4.935676645431128e-05,
+      "loss": 0.447,
       "step": 13905
     },
     {
-      "epoch": 0.49,
-      "learning_rate": 4.9398611124140046e-05,
-      "loss": 0.2989,
+      "epoch": 0.5013154575269398,
+      "grad_norm": 0.1593722254037857,
+      "learning_rate": 4.935610859592508e-05,
+      "loss": 0.425,
       "step": 13910
     },
     {
-      "epoch": 0.49,
-      "learning_rate": 4.939798988935049e-05,
-      "loss": 0.3339,
+      "epoch": 0.5014956571881645,
+      "grad_norm": 0.1893470287322998,
+      "learning_rate": 4.93554504056916e-05,
+      "loss": 0.4611,
       "step": 13915
     },
     {
-      "epoch": 0.49,
-      "learning_rate": 4.939736833776912e-05,
-      "loss": 0.3322,
+      "epoch": 0.5016758568493891,
+      "grad_norm": 0.17857231199741364,
+      "learning_rate": 4.935479188361983e-05,
+      "loss": 0.4438,
       "step": 13920
     },
     {
-      "epoch": 0.49,
-      "learning_rate": 4.9396746469403996e-05,
-      "loss": 0.3308,
+      "epoch": 0.5018560565106137,
+      "grad_norm": 0.15780913829803467,
+      "learning_rate": 4.935413302971874e-05,
+      "loss": 0.4474,
       "step": 13925
     },
     {
-      "epoch": 0.49,
-      "learning_rate": 4.93961242842632e-05,
-      "loss": 0.3017,
+      "epoch": 0.5020362561718384,
+      "grad_norm": 0.17613886296749115,
+      "learning_rate": 4.93534738439973e-05,
+      "loss": 0.4686,
       "step": 13930
     },
     {
-      "epoch": 0.49,
-      "learning_rate": 4.9395501782354813e-05,
-      "loss": 0.3068,
+      "epoch": 0.5022164558330631,
+      "grad_norm": 0.16873963177204132,
+      "learning_rate": 4.9352814326464493e-05,
+      "loss": 0.4534,
       "step": 13935
     },
     {
-      "epoch": 0.49,
-      "learning_rate": 4.939487896368691e-05,
-      "loss": 0.3179,
+      "epoch": 0.5023966554942877,
+      "grad_norm": 0.1617756485939026,
+      "learning_rate": 4.93521544771293e-05,
+      "loss": 0.4431,
       "step": 13940
     },
     {
-      "epoch": 0.49,
-      "learning_rate": 4.939425582826757e-05,
-      "loss": 0.3163,
+      "epoch": 0.5025768551555123,
+      "grad_norm": 0.1517868936061859,
+      "learning_rate": 4.9351494296000726e-05,
+      "loss": 0.4691,
       "step": 13945
     },
     {
-      "epoch": 0.49,
-      "learning_rate": 4.9393632376104904e-05,
-      "loss": 0.3401,
+      "epoch": 0.5027570548167369,
+      "grad_norm": 0.18268106877803802,
+      "learning_rate": 4.935083378308776e-05,
+      "loss": 0.4257,
       "step": 13950
     },
     {
-      "epoch": 0.49,
-      "learning_rate": 4.9393008607206995e-05,
-      "loss": 0.32,
+      "epoch": 0.5029372544779616,
+      "grad_norm": 0.16596758365631104,
+      "learning_rate": 4.935017293839939e-05,
+      "loss": 0.408,
       "step": 13955
     },
     {
-      "epoch": 0.49,
-      "learning_rate": 4.9392384521581945e-05,
-      "loss": 0.3022,
+      "epoch": 0.5031174541391862,
+      "grad_norm": 0.2062714397907257,
+      "learning_rate": 4.934951176194462e-05,
+      "loss": 0.4706,
       "step": 13960
     },
     {
-      "epoch": 0.49,
-      "learning_rate": 4.939176011923786e-05,
-      "loss": 0.3167,
+      "epoch": 0.5032976538004108,
+      "grad_norm": 0.1877586394548416,
+      "learning_rate": 4.934885025373248e-05,
+      "loss": 0.4726,
       "step": 13965
     },
     {
-      "epoch": 0.49,
-      "learning_rate": 4.939113540018284e-05,
-      "loss": 0.3418,
+      "epoch": 0.5034778534616355,
+      "grad_norm": 0.17197106778621674,
+      "learning_rate": 4.9348188413771966e-05,
+      "loss": 0.4465,
       "step": 13970
     },
     {
-      "epoch": 0.49,
-      "learning_rate": 4.9390510364425004e-05,
-      "loss": 0.3078,
+      "epoch": 0.5036580531228602,
+      "grad_norm": 0.16831296682357788,
+      "learning_rate": 4.93475262420721e-05,
+      "loss": 0.489,
       "step": 13975
     },
     {
-      "epoch": 0.49,
-      "learning_rate": 4.9389885011972464e-05,
-      "loss": 0.3185,
+      "epoch": 0.5038382527840848,
+      "grad_norm": 0.18322400748729706,
+      "learning_rate": 4.93468637386419e-05,
+      "loss": 0.4343,
       "step": 13980
     },
     {
-      "epoch": 0.49,
-      "learning_rate": 4.938925934283334e-05,
-      "loss": 0.3481,
+      "epoch": 0.5040184524453094,
+      "grad_norm": 0.21650521457195282,
+      "learning_rate": 4.93462009034904e-05,
+      "loss": 0.4735,
       "step": 13985
     },
     {
-      "epoch": 0.49,
-      "learning_rate": 4.9388633357015755e-05,
-      "loss": 0.317,
+      "epoch": 0.504198652106534,
+      "grad_norm": 0.17821024358272552,
+      "learning_rate": 4.9345537736626626e-05,
+      "loss": 0.4375,
       "step": 13990
     },
     {
-      "epoch": 0.49,
-      "learning_rate": 4.938800705452784e-05,
-      "loss": 0.3323,
+      "epoch": 0.5043788517677587,
+      "grad_norm": 0.15985864400863647,
+      "learning_rate": 4.934487423805961e-05,
+      "loss": 0.4419,
       "step": 13995
     },
     {
-      "epoch": 0.49,
-      "learning_rate": 4.9387380435377726e-05,
-      "loss": 0.316,
+      "epoch": 0.5045590514289833,
+      "grad_norm": 0.1555582582950592,
+      "learning_rate": 4.93442104077984e-05,
+      "loss": 0.4606,
       "step": 14000
     },
     {
-      "epoch": 0.49,
-      "eval_loss": 0.3117991089820862,
-      "eval_runtime": 10.538,
-      "eval_samples_per_second": 9.489,
-      "eval_steps_per_second": 9.489,
+      "epoch": 0.5045590514289833,
+      "eval_loss": 0.4720987379550934,
+      "eval_runtime": 3.5871,
+      "eval_samples_per_second": 27.878,
+      "eval_steps_per_second": 6.969,
       "step": 14000
     },
     {
-      "epoch": 0.49,
-      "learning_rate": 4.938675349957356e-05,
-      "loss": 0.3099,
+      "epoch": 0.5047392510902079,
+      "grad_norm": 0.1948181539773941,
+      "learning_rate": 4.934354624585202e-05,
+      "loss": 0.4392,
       "step": 14005
     },
     {
-      "epoch": 0.49,
-      "learning_rate": 4.938612624712346e-05,
-      "loss": 0.3418,
+      "epoch": 0.5049194507514326,
+      "grad_norm": 0.1451338827610016,
+      "learning_rate": 4.934288175222955e-05,
+      "loss": 0.4184,
       "step": 14010
     },
     {
-      "epoch": 0.49,
-      "learning_rate": 4.938549867803558e-05,
-      "loss": 0.3279,
+      "epoch": 0.5050996504126573,
+      "grad_norm": 0.1611844301223755,
+      "learning_rate": 4.934221692694003e-05,
+      "loss": 0.4428,
       "step": 14015
     },
     {
-      "epoch": 0.49,
-      "learning_rate": 4.9384870792318075e-05,
-      "loss": 0.3224,
+      "epoch": 0.5052798500738819,
+      "grad_norm": 0.19232018291950226,
+      "learning_rate": 4.934155176999252e-05,
+      "loss": 0.4512,
       "step": 14020
     },
     {
-      "epoch": 0.49,
-      "learning_rate": 4.938424258997909e-05,
-      "loss": 0.3266,
+      "epoch": 0.5054600497351065,
+      "grad_norm": 0.24862439930438995,
+      "learning_rate": 4.934088628139607e-05,
+      "loss": 0.4715,
       "step": 14025
     },
     {
-      "epoch": 0.49,
-      "learning_rate": 4.9383614071026787e-05,
-      "loss": 0.2839,
+      "epoch": 0.5056402493963311,
+      "grad_norm": 0.15161451697349548,
+      "learning_rate": 4.9340220461159757e-05,
+      "loss": 0.4179,
       "step": 14030
     },
     {
-      "epoch": 0.49,
-      "learning_rate": 4.938298523546932e-05,
-      "loss": 0.3114,
+      "epoch": 0.5058204490575557,
+      "grad_norm": 0.19004757702350616,
+      "learning_rate": 4.933955430929266e-05,
+      "loss": 0.4474,
       "step": 14035
     },
     {
-      "epoch": 0.49,
-      "learning_rate": 4.938235608331486e-05,
-      "loss": 0.3205,
+      "epoch": 0.5060006487187804,
+      "grad_norm": 0.23055416345596313,
+      "learning_rate": 4.933888782580385e-05,
+      "loss": 0.4602,
       "step": 14040
     },
     {
-      "epoch": 0.49,
-      "learning_rate": 4.938172661457158e-05,
-      "loss": 0.3337,
+      "epoch": 0.5061808483800051,
+      "grad_norm": 0.18940792977809906,
+      "learning_rate": 4.93382210107024e-05,
+      "loss": 0.4346,
       "step": 14045
     },
     {
-      "epoch": 0.49,
-      "learning_rate": 4.9381096829247644e-05,
-      "loss": 0.318,
+      "epoch": 0.5063610480412297,
+      "grad_norm": 0.1852693110704422,
+      "learning_rate": 4.9337553863997396e-05,
+      "loss": 0.4499,
       "step": 14050
     },
     {
-      "epoch": 0.49,
-      "learning_rate": 4.938046672735123e-05,
-      "loss": 0.316,
+      "epoch": 0.5065412477024543,
+      "grad_norm": 0.1449885368347168,
+      "learning_rate": 4.933688638569794e-05,
+      "loss": 0.4657,
       "step": 14055
     },
     {
-      "epoch": 0.49,
-      "learning_rate": 4.937983630889053e-05,
-      "loss": 0.3182,
+      "epoch": 0.506721447363679,
+      "grad_norm": 0.1560732126235962,
+      "learning_rate": 4.933621857581312e-05,
+      "loss": 0.4184,
       "step": 14060
     },
     {
-      "epoch": 0.49,
-      "learning_rate": 4.937920557387372e-05,
-      "loss": 0.3189,
+      "epoch": 0.5069016470249036,
+      "grad_norm": 0.1511814445257187,
+      "learning_rate": 4.933555043435203e-05,
+      "loss": 0.4409,
       "step": 14065
     },
     {
-      "epoch": 0.5,
-      "learning_rate": 4.937857452230899e-05,
-      "loss": 0.3122,
+      "epoch": 0.5070818466861282,
+      "grad_norm": 0.1766042411327362,
+      "learning_rate": 4.9334881961323776e-05,
+      "loss": 0.4559,
       "step": 14070
     },
     {
-      "epoch": 0.5,
-      "learning_rate": 4.937794315420454e-05,
-      "loss": 0.2821,
+      "epoch": 0.5072620463473528,
+      "grad_norm": 0.14052477478981018,
+      "learning_rate": 4.9334213156737465e-05,
+      "loss": 0.4476,
       "step": 14075
     },
     {
-      "epoch": 0.5,
-      "learning_rate": 4.9377311469568566e-05,
-      "loss": 0.3035,
+      "epoch": 0.5074422460085775,
+      "grad_norm": 0.1610243022441864,
+      "learning_rate": 4.933354402060221e-05,
+      "loss": 0.434,
       "step": 14080
     },
     {
-      "epoch": 0.5,
-      "learning_rate": 4.937667946840926e-05,
-      "loss": 0.3213,
+      "epoch": 0.5076224456698022,
+      "grad_norm": 0.1901833862066269,
+      "learning_rate": 4.9332874552927135e-05,
+      "loss": 0.4489,
       "step": 14085
     },
     {
-      "epoch": 0.5,
-      "learning_rate": 4.937604715073484e-05,
-      "loss": 0.3141,
+      "epoch": 0.5078026453310268,
+      "grad_norm": 0.16567949950695038,
+      "learning_rate": 4.933220475372136e-05,
+      "loss": 0.4473,
       "step": 14090
     },
     {
-      "epoch": 0.5,
-      "learning_rate": 4.937541451655351e-05,
-      "loss": 0.3298,
+      "epoch": 0.5079828449922514,
+      "grad_norm": 0.2085447609424591,
+      "learning_rate": 4.933153462299399e-05,
+      "loss": 0.4374,
       "step": 14095
     },
     {
-      "epoch": 0.5,
-      "learning_rate": 4.937478156587349e-05,
-      "loss": 0.3311,
+      "epoch": 0.5081630446534761,
+      "grad_norm": 0.1837865263223648,
+      "learning_rate": 4.933086416075418e-05,
+      "loss": 0.4615,
       "step": 14100
     },
     {
-      "epoch": 0.5,
-      "learning_rate": 4.9374148298702994e-05,
-      "loss": 0.3197,
+      "epoch": 0.5083432443147007,
+      "grad_norm": 0.1936039924621582,
+      "learning_rate": 4.933019336701106e-05,
+      "loss": 0.4533,
       "step": 14105
     },
     {
-      "epoch": 0.5,
-      "learning_rate": 4.937351471505024e-05,
-      "loss": 0.3159,
+      "epoch": 0.5085234439759253,
+      "grad_norm": 0.38860994577407837,
+      "learning_rate": 4.932952224177376e-05,
+      "loss": 0.4412,
       "step": 14110
     },
     {
-      "epoch": 0.5,
-      "learning_rate": 4.937288081492347e-05,
-      "loss": 0.3294,
+      "epoch": 0.5087036436371499,
+      "grad_norm": 0.18047145009040833,
+      "learning_rate": 4.932885078505143e-05,
+      "loss": 0.4725,
       "step": 14115
     },
     {
-      "epoch": 0.5,
-      "learning_rate": 4.9372246598330895e-05,
-      "loss": 0.3011,
+      "epoch": 0.5088838432983745,
+      "grad_norm": 0.16559572517871857,
+      "learning_rate": 4.932817899685323e-05,
+      "loss": 0.452,
       "step": 14120
     },
     {
-      "epoch": 0.5,
-      "learning_rate": 4.937161206528077e-05,
-      "loss": 0.3129,
+      "epoch": 0.5090640429595993,
+      "grad_norm": 0.1487295925617218,
+      "learning_rate": 4.9327506877188284e-05,
+      "loss": 0.4177,
       "step": 14125
     },
     {
-      "epoch": 0.5,
-      "learning_rate": 4.9370977215781314e-05,
-      "loss": 0.3069,
+      "epoch": 0.5092442426208239,
+      "grad_norm": 0.14861665666103363,
+      "learning_rate": 4.9326834426065775e-05,
+      "loss": 0.4183,
       "step": 14130
     },
     {
-      "epoch": 0.5,
-      "learning_rate": 4.9370342049840786e-05,
-      "loss": 0.3171,
+      "epoch": 0.5094244422820485,
+      "grad_norm": 0.13988757133483887,
+      "learning_rate": 4.9326161643494856e-05,
+      "loss": 0.4216,
       "step": 14135
     },
     {
-      "epoch": 0.5,
-      "learning_rate": 4.9369706567467425e-05,
-      "loss": 0.3133,
+      "epoch": 0.5096046419432732,
+      "grad_norm": 0.1480373740196228,
+      "learning_rate": 4.93254885294847e-05,
+      "loss": 0.4165,
       "step": 14140
     },
     {
-      "epoch": 0.5,
-      "learning_rate": 4.9369070768669477e-05,
-      "loss": 0.3231,
+      "epoch": 0.5097848416044978,
+      "grad_norm": 0.1930646002292633,
+      "learning_rate": 4.932481508404446e-05,
+      "loss": 0.426,
       "step": 14145
     },
     {
-      "epoch": 0.5,
-      "learning_rate": 4.936843465345522e-05,
-      "loss": 0.3313,
+      "epoch": 0.5099650412657224,
+      "grad_norm": 0.18709403276443481,
+      "learning_rate": 4.932414130718334e-05,
+      "loss": 0.4608,
       "step": 14150
     },
     {
-      "epoch": 0.5,
-      "learning_rate": 4.936779822183289e-05,
-      "loss": 0.3145,
+      "epoch": 0.510145240926947,
+      "grad_norm": 0.17628611624240875,
+      "learning_rate": 4.932346719891049e-05,
+      "loss": 0.4455,
       "step": 14155
     },
     {
-      "epoch": 0.5,
-      "learning_rate": 4.9367161473810766e-05,
-      "loss": 0.3093,
+      "epoch": 0.5103254405881716,
+      "grad_norm": 0.1651298552751541,
+      "learning_rate": 4.9322792759235115e-05,
+      "loss": 0.4711,
       "step": 14160
     },
     {
-      "epoch": 0.5,
-      "learning_rate": 4.93665244093971e-05,
-      "loss": 0.3168,
+      "epoch": 0.5105056402493964,
+      "grad_norm": 0.16614636778831482,
+      "learning_rate": 4.93221179881664e-05,
+      "loss": 0.4232,
       "step": 14165
     },
     {
-      "epoch": 0.5,
-      "learning_rate": 4.9365887028600175e-05,
-      "loss": 0.3277,
+      "epoch": 0.510685839910621,
+      "grad_norm": 0.16470062732696533,
+      "learning_rate": 4.932144288571353e-05,
+      "loss": 0.4558,
       "step": 14170
     },
     {
-      "epoch": 0.5,
-      "learning_rate": 4.9365249331428265e-05,
-      "loss": 0.3193,
+      "epoch": 0.5108660395718456,
+      "grad_norm": 0.20364750921726227,
+      "learning_rate": 4.932076745188571e-05,
+      "loss": 0.4421,
       "step": 14175
     },
     {
-      "epoch": 0.5,
-      "learning_rate": 4.9364611317889644e-05,
-      "loss": 0.315,
+      "epoch": 0.5110462392330702,
+      "grad_norm": 0.19398504495620728,
+      "learning_rate": 4.9320091686692136e-05,
+      "loss": 0.4581,
       "step": 14180
     },
     {
-      "epoch": 0.5,
-      "learning_rate": 4.936397298799261e-05,
-      "loss": 0.3124,
+      "epoch": 0.5112264388942949,
+      "grad_norm": 0.1814093142747879,
+      "learning_rate": 4.931941559014204e-05,
+      "loss": 0.4678,
       "step": 14185
     },
     {
-      "epoch": 0.5,
-      "learning_rate": 4.9363334341745446e-05,
-      "loss": 0.3186,
+      "epoch": 0.5114066385555195,
+      "grad_norm": 0.15303486585617065,
+      "learning_rate": 4.93187391622446e-05,
+      "loss": 0.436,
       "step": 14190
     },
     {
-      "epoch": 0.5,
-      "learning_rate": 4.936269537915643e-05,
-      "loss": 0.3363,
+      "epoch": 0.5115868382167441,
+      "grad_norm": 0.1580374389886856,
+      "learning_rate": 4.931806240300905e-05,
+      "loss": 0.4715,
       "step": 14195
     },
     {
-      "epoch": 0.5,
-      "learning_rate": 4.936205610023388e-05,
-      "loss": 0.3318,
+      "epoch": 0.5117670378779688,
+      "grad_norm": 0.18431012332439423,
+      "learning_rate": 4.931738531244461e-05,
+      "loss": 0.4241,
       "step": 14200
     },
     {
-      "epoch": 0.5,
-      "learning_rate": 4.936141650498608e-05,
-      "loss": 0.2952,
+      "epoch": 0.5119472375391935,
+      "grad_norm": 0.20473140478134155,
+      "learning_rate": 4.93167078905605e-05,
+      "loss": 0.4226,
       "step": 14205
     },
     {
-      "epoch": 0.5,
-      "learning_rate": 4.9360776593421347e-05,
-      "loss": 0.3331,
+      "epoch": 0.5121274372004181,
+      "grad_norm": 0.1783171445131302,
+      "learning_rate": 4.931603013736595e-05,
+      "loss": 0.4515,
       "step": 14210
     },
     {
-      "epoch": 0.5,
-      "learning_rate": 4.9360136365547985e-05,
-      "loss": 0.3029,
+      "epoch": 0.5123076368616427,
+      "grad_norm": 0.1747739315032959,
+      "learning_rate": 4.931535205287021e-05,
+      "loss": 0.4537,
       "step": 14215
     },
     {
-      "epoch": 0.5,
-      "learning_rate": 4.93594958213743e-05,
-      "loss": 0.2965,
+      "epoch": 0.5124878365228673,
+      "grad_norm": 0.17563074827194214,
+      "learning_rate": 4.93146736370825e-05,
+      "loss": 0.459,
       "step": 14220
     },
     {
-      "epoch": 0.5,
-      "learning_rate": 4.935885496090862e-05,
-      "loss": 0.3002,
+      "epoch": 0.512668036184092,
+      "grad_norm": 0.16567584872245789,
+      "learning_rate": 4.931399489001206e-05,
+      "loss": 0.4407,
       "step": 14225
     },
     {
-      "epoch": 0.5,
-      "learning_rate": 4.9358213784159264e-05,
-      "loss": 0.3356,
+      "epoch": 0.5128482358453166,
+      "grad_norm": 0.20225879549980164,
+      "learning_rate": 4.931331581166816e-05,
+      "loss": 0.5078,
       "step": 14230
     },
     {
-      "epoch": 0.5,
-      "learning_rate": 4.935757229113455e-05,
-      "loss": 0.3206,
+      "epoch": 0.5130284355065412,
+      "grad_norm": 0.17894190549850464,
+      "learning_rate": 4.931263640206003e-05,
+      "loss": 0.429,
       "step": 14235
     },
     {
-      "epoch": 0.5,
-      "learning_rate": 4.9356930481842814e-05,
-      "loss": 0.3251,
+      "epoch": 0.5132086351677659,
+      "grad_norm": 0.2424396276473999,
+      "learning_rate": 4.9311956661196945e-05,
+      "loss": 0.4652,
       "step": 14240
     },
     {
-      "epoch": 0.5,
-      "learning_rate": 4.935628835629238e-05,
-      "loss": 0.3013,
+      "epoch": 0.5133888348289906,
+      "grad_norm": 0.17801520228385925,
+      "learning_rate": 4.931127658908815e-05,
+      "loss": 0.4764,
       "step": 14245
     },
     {
-      "epoch": 0.5,
-      "learning_rate": 4.93556459144916e-05,
-      "loss": 0.2957,
+      "epoch": 0.5135690344902152,
+      "grad_norm": 0.16889409720897675,
+      "learning_rate": 4.931059618574292e-05,
+      "loss": 0.4521,
       "step": 14250
     },
     {
-      "epoch": 0.5,
-      "learning_rate": 4.935500315644881e-05,
-      "loss": 0.2915,
+      "epoch": 0.5137492341514398,
+      "grad_norm": 0.17756950855255127,
+      "learning_rate": 4.930991545117052e-05,
+      "loss": 0.4439,
       "step": 14255
     },
     {
-      "epoch": 0.5,
-      "learning_rate": 4.935436008217236e-05,
-      "loss": 0.3131,
+      "epoch": 0.5139294338126644,
+      "grad_norm": 0.13921865820884705,
+      "learning_rate": 4.930923438538024e-05,
+      "loss": 0.4176,
       "step": 14260
     },
     {
-      "epoch": 0.5,
-      "learning_rate": 4.935371669167058e-05,
-      "loss": 0.3186,
+      "epoch": 0.514109633473889,
+      "grad_norm": 0.21671532094478607,
+      "learning_rate": 4.930855298838134e-05,
+      "loss": 0.4944,
       "step": 14265
     },
     {
-      "epoch": 0.5,
-      "learning_rate": 4.935307298495185e-05,
-      "loss": 0.2998,
+      "epoch": 0.5142898331351137,
+      "grad_norm": 0.1547165960073471,
+      "learning_rate": 4.930787126018311e-05,
+      "loss": 0.4394,
       "step": 14270
     },
     {
-      "epoch": 0.5,
-      "learning_rate": 4.935242896202451e-05,
-      "loss": 0.3304,
+      "epoch": 0.5144700327963383,
+      "grad_norm": 0.16631481051445007,
+      "learning_rate": 4.930718920079484e-05,
+      "loss": 0.4397,
       "step": 14275
     },
     {
-      "epoch": 0.5,
-      "learning_rate": 4.935178462289694e-05,
-      "loss": 0.3435,
+      "epoch": 0.514650232457563,
+      "grad_norm": 0.14847290515899658,
+      "learning_rate": 4.9306506810225824e-05,
+      "loss": 0.4688,
       "step": 14280
     },
     {
-      "epoch": 0.5,
-      "learning_rate": 4.9351139967577475e-05,
-      "loss": 0.2953,
+      "epoch": 0.5148304321187877,
+      "grad_norm": 0.1879221498966217,
+      "learning_rate": 4.930582408848536e-05,
+      "loss": 0.4581,
       "step": 14285
     },
     {
-      "epoch": 0.5,
-      "learning_rate": 4.935049499607452e-05,
-      "loss": 0.3056,
+      "epoch": 0.5150106317800123,
+      "grad_norm": 0.15472771227359772,
+      "learning_rate": 4.930514103558275e-05,
+      "loss": 0.4211,
       "step": 14290
     },
     {
-      "epoch": 0.5,
-      "learning_rate": 4.9349849708396436e-05,
-      "loss": 0.302,
+      "epoch": 0.5151908314412369,
+      "grad_norm": 0.22856763005256653,
+      "learning_rate": 4.9304457651527305e-05,
+      "loss": 0.4399,
       "step": 14295
     },
     {
-      "epoch": 0.5,
-      "learning_rate": 4.93492041045516e-05,
-      "loss": 0.3219,
+      "epoch": 0.5153710311024615,
+      "grad_norm": 0.13668227195739746,
+      "learning_rate": 4.930377393632832e-05,
+      "loss": 0.4319,
       "step": 14300
     },
     {
-      "epoch": 0.5,
-      "learning_rate": 4.934855818454839e-05,
-      "loss": 0.3243,
+      "epoch": 0.5155512307636861,
+      "grad_norm": 0.1626194417476654,
+      "learning_rate": 4.9303089889995125e-05,
+      "loss": 0.4161,
       "step": 14305
     },
     {
-      "epoch": 0.5,
-      "learning_rate": 4.93479119483952e-05,
-      "loss": 0.3482,
+      "epoch": 0.5157314304249108,
+      "grad_norm": 0.17859403789043427,
+      "learning_rate": 4.930240551253703e-05,
+      "loss": 0.4305,
       "step": 14310
     },
     {
-      "epoch": 0.5,
-      "learning_rate": 4.9347265396100414e-05,
-      "loss": 0.3026,
+      "epoch": 0.5159116300861354,
+      "grad_norm": 0.16735269129276276,
+      "learning_rate": 4.930172080396337e-05,
+      "loss": 0.4726,
       "step": 14315
     },
     {
-      "epoch": 0.5,
-      "learning_rate": 4.934661852767244e-05,
-      "loss": 0.32,
+      "epoch": 0.5160918297473601,
+      "grad_norm": 0.1954079419374466,
+      "learning_rate": 4.930103576428346e-05,
+      "loss": 0.4316,
       "step": 14320
     },
     {
-      "epoch": 0.5,
-      "learning_rate": 4.934597134311967e-05,
-      "loss": 0.3109,
+      "epoch": 0.5162720294085847,
+      "grad_norm": 0.17768363654613495,
+      "learning_rate": 4.9300350393506655e-05,
+      "loss": 0.4013,
       "step": 14325
     },
     {
-      "epoch": 0.5,
-      "learning_rate": 4.93453238424505e-05,
-      "loss": 0.3353,
+      "epoch": 0.5164522290698094,
+      "grad_norm": 0.1929953694343567,
+      "learning_rate": 4.929966469164228e-05,
+      "loss": 0.4537,
       "step": 14330
     },
     {
-      "epoch": 0.5,
-      "learning_rate": 4.934467602567335e-05,
-      "loss": 0.2842,
+      "epoch": 0.516632428731034,
+      "grad_norm": 0.17287170886993408,
+      "learning_rate": 4.9298978658699674e-05,
+      "loss": 0.4316,
       "step": 14335
     },
     {
-      "epoch": 0.5,
-      "learning_rate": 4.934402789279662e-05,
-      "loss": 0.304,
+      "epoch": 0.5168126283922586,
+      "grad_norm": 0.18152733147144318,
+      "learning_rate": 4.9298292294688183e-05,
+      "loss": 0.4502,
       "step": 14340
     },
     {
-      "epoch": 0.5,
-      "learning_rate": 4.934337944382874e-05,
-      "loss": 0.3147,
+      "epoch": 0.5169928280534832,
+      "grad_norm": 0.19768409430980682,
+      "learning_rate": 4.929760559961717e-05,
+      "loss": 0.4375,
       "step": 14345
     },
     {
-      "epoch": 0.5,
-      "learning_rate": 4.934273067877811e-05,
-      "loss": 0.2964,
+      "epoch": 0.5171730277147079,
+      "grad_norm": 0.18449656665325165,
+      "learning_rate": 4.9296918573495984e-05,
+      "loss": 0.4628,
       "step": 14350
     },
     {
-      "epoch": 0.51,
-      "learning_rate": 4.934208159765317e-05,
-      "loss": 0.3063,
+      "epoch": 0.5173532273759325,
+      "grad_norm": 0.13669456541538239,
+      "learning_rate": 4.9296231216333986e-05,
+      "loss": 0.4446,
       "step": 14355
     },
     {
-      "epoch": 0.51,
-      "learning_rate": 4.9341432200462343e-05,
-      "loss": 0.3435,
+      "epoch": 0.5175334270371572,
+      "grad_norm": 0.17907440662384033,
+      "learning_rate": 4.929554352814055e-05,
+      "loss": 0.4351,
       "step": 14360
     },
     {
-      "epoch": 0.51,
-      "learning_rate": 4.934078248721407e-05,
-      "loss": 0.306,
+      "epoch": 0.5177136266983818,
+      "grad_norm": 0.2036353498697281,
+      "learning_rate": 4.9294855508925026e-05,
+      "loss": 0.4293,
       "step": 14365
     },
     {
-      "epoch": 0.51,
-      "learning_rate": 4.934013245791676e-05,
-      "loss": 0.3341,
+      "epoch": 0.5178938263596065,
+      "grad_norm": 0.14478379487991333,
+      "learning_rate": 4.92941671586968e-05,
+      "loss": 0.4088,
       "step": 14370
     },
     {
-      "epoch": 0.51,
-      "learning_rate": 4.933948211257888e-05,
-      "loss": 0.3356,
+      "epoch": 0.5180740260208311,
+      "grad_norm": 0.19350174069404602,
+      "learning_rate": 4.9293478477465254e-05,
+      "loss": 0.4735,
       "step": 14375
     },
     {
-      "epoch": 0.51,
-      "learning_rate": 4.933883145120886e-05,
-      "loss": 0.3242,
+      "epoch": 0.5182542256820557,
+      "grad_norm": 0.15223993360996246,
+      "learning_rate": 4.9292789465239765e-05,
+      "loss": 0.42,
       "step": 14380
     },
     {
-      "epoch": 0.51,
-      "learning_rate": 4.933818047381516e-05,
-      "loss": 0.3052,
+      "epoch": 0.5184344253432803,
+      "grad_norm": 0.17773163318634033,
+      "learning_rate": 4.929210012202973e-05,
+      "loss": 0.4626,
       "step": 14385
     },
     {
-      "epoch": 0.51,
-      "learning_rate": 4.933752918040622e-05,
-      "loss": 0.3118,
+      "epoch": 0.518614625004505,
+      "grad_norm": 0.19823719561100006,
+      "learning_rate": 4.929141044784452e-05,
+      "loss": 0.4291,
       "step": 14390
     },
     {
-      "epoch": 0.51,
-      "learning_rate": 4.933687757099051e-05,
-      "loss": 0.3243,
+      "epoch": 0.5187948246657297,
+      "grad_norm": 0.163058340549469,
+      "learning_rate": 4.929072044269356e-05,
+      "loss": 0.4824,
       "step": 14395
     },
     {
-      "epoch": 0.51,
-      "learning_rate": 4.933622564557648e-05,
-      "loss": 0.2979,
+      "epoch": 0.5189750243269543,
+      "grad_norm": 0.17107297480106354,
+      "learning_rate": 4.929003010658623e-05,
+      "loss": 0.4953,
       "step": 14400
     },
     {
-      "epoch": 0.51,
-      "learning_rate": 4.9335573404172594e-05,
-      "loss": 0.3052,
+      "epoch": 0.5191552239881789,
+      "grad_norm": 0.17323638498783112,
+      "learning_rate": 4.928933943953193e-05,
+      "loss": 0.4444,
       "step": 14405
     },
     {
-      "epoch": 0.51,
-      "learning_rate": 4.9334920846787324e-05,
-      "loss": 0.3168,
+      "epoch": 0.5193354236494035,
+      "grad_norm": 0.19337867200374603,
+      "learning_rate": 4.92886484415401e-05,
+      "loss": 0.4295,
       "step": 14410
     },
     {
-      "epoch": 0.51,
-      "learning_rate": 4.933426797342915e-05,
-      "loss": 0.3032,
+      "epoch": 0.5195156233106282,
+      "grad_norm": 0.250715434551239,
+      "learning_rate": 4.9287957112620134e-05,
+      "loss": 0.5048,
       "step": 14415
     },
     {
-      "epoch": 0.51,
-      "learning_rate": 4.933361478410654e-05,
-      "loss": 0.319,
+      "epoch": 0.5196958229718528,
+      "grad_norm": 0.19110535085201263,
+      "learning_rate": 4.928726545278145e-05,
+      "loss": 0.4508,
       "step": 14420
     },
     {
-      "epoch": 0.51,
-      "learning_rate": 4.933296127882798e-05,
-      "loss": 0.3071,
+      "epoch": 0.5198760226330774,
+      "grad_norm": 0.2401556819677353,
+      "learning_rate": 4.9286573462033484e-05,
+      "loss": 0.52,
       "step": 14425
     },
     {
-      "epoch": 0.51,
-      "learning_rate": 4.9332307457601955e-05,
-      "loss": 0.2847,
+      "epoch": 0.520056222294302,
+      "grad_norm": 0.145725280046463,
+      "learning_rate": 4.9285881140385645e-05,
+      "loss": 0.4753,
       "step": 14430
     },
     {
-      "epoch": 0.51,
-      "learning_rate": 4.9331653320436954e-05,
-      "loss": 0.2941,
+      "epoch": 0.5202364219555268,
+      "grad_norm": 0.17486672103405,
+      "learning_rate": 4.928518848784739e-05,
+      "loss": 0.4178,
       "step": 14435
     },
     {
-      "epoch": 0.51,
-      "learning_rate": 4.933099886734147e-05,
-      "loss": 0.2967,
+      "epoch": 0.5204166216167514,
+      "grad_norm": 0.1457482874393463,
+      "learning_rate": 4.928449550442814e-05,
+      "loss": 0.4379,
       "step": 14440
     },
     {
-      "epoch": 0.51,
-      "learning_rate": 4.9330344098324e-05,
-      "loss": 0.304,
+      "epoch": 0.520596821277976,
+      "grad_norm": 0.21257434785366058,
+      "learning_rate": 4.928380219013734e-05,
+      "loss": 0.465,
       "step": 14445
     },
     {
-      "epoch": 0.51,
-      "learning_rate": 4.932968901339304e-05,
-      "loss": 0.3298,
+      "epoch": 0.5207770209392006,
+      "grad_norm": 0.17879636585712433,
+      "learning_rate": 4.928310854498444e-05,
+      "loss": 0.4134,
       "step": 14450
     },
     {
-      "epoch": 0.51,
-      "learning_rate": 4.932903361255711e-05,
-      "loss": 0.2917,
+      "epoch": 0.5209572206004253,
+      "grad_norm": 0.20684200525283813,
+      "learning_rate": 4.928241456897887e-05,
+      "loss": 0.4,
       "step": 14455
     },
     {
-      "epoch": 0.51,
-      "learning_rate": 4.9328377895824705e-05,
-      "loss": 0.2729,
+      "epoch": 0.5211374202616499,
+      "grad_norm": 0.15960803627967834,
+      "learning_rate": 4.928172026213012e-05,
+      "loss": 0.4198,
       "step": 14460
     },
     {
-      "epoch": 0.51,
-      "learning_rate": 4.9327721863204354e-05,
-      "loss": 0.316,
+      "epoch": 0.5213176199228745,
+      "grad_norm": 0.1935255080461502,
+      "learning_rate": 4.928102562444763e-05,
+      "loss": 0.4503,
       "step": 14465
     },
     {
-      "epoch": 0.51,
-      "learning_rate": 4.932706551470456e-05,
-      "loss": 0.2947,
+      "epoch": 0.5214978195840991,
+      "grad_norm": 0.18149693310260773,
+      "learning_rate": 4.928033065594086e-05,
+      "loss": 0.4337,
       "step": 14470
     },
     {
-      "epoch": 0.51,
-      "learning_rate": 4.932640885033386e-05,
-      "loss": 0.3385,
+      "epoch": 0.5216780192453239,
+      "grad_norm": 0.21554191410541534,
+      "learning_rate": 4.927963535661929e-05,
+      "loss": 0.3948,
       "step": 14475
     },
     {
-      "epoch": 0.51,
-      "learning_rate": 4.9325751870100765e-05,
-      "loss": 0.2871,
+      "epoch": 0.5218582189065485,
+      "grad_norm": 0.14424189925193787,
+      "learning_rate": 4.927893972649239e-05,
+      "loss": 0.4235,
       "step": 14480
     },
     {
-      "epoch": 0.51,
-      "learning_rate": 4.932509457401382e-05,
-      "loss": 0.3188,
+      "epoch": 0.5220384185677731,
+      "grad_norm": 0.1991516500711441,
+      "learning_rate": 4.927824376556964e-05,
+      "loss": 0.4594,
       "step": 14485
     },
     {
-      "epoch": 0.51,
-      "learning_rate": 4.9324436962081546e-05,
-      "loss": 0.302,
+      "epoch": 0.5222186182289977,
+      "grad_norm": 0.12362515181303024,
+      "learning_rate": 4.927754747386051e-05,
+      "loss": 0.4268,
       "step": 14490
     },
     {
-      "epoch": 0.51,
-      "learning_rate": 4.932377903431249e-05,
-      "loss": 0.3015,
+      "epoch": 0.5223988178902224,
+      "grad_norm": 0.219016894698143,
+      "learning_rate": 4.92768508513745e-05,
+      "loss": 0.4574,
       "step": 14495
     },
     {
-      "epoch": 0.51,
-      "learning_rate": 4.9323120790715194e-05,
-      "loss": 0.3215,
+      "epoch": 0.522579017551447,
+      "grad_norm": 0.1444326788187027,
+      "learning_rate": 4.92761538981211e-05,
+      "loss": 0.4237,
       "step": 14500
     },
     {
-      "epoch": 0.51,
-      "eval_loss": 0.3104207515716553,
-      "eval_runtime": 10.5492,
-      "eval_samples_per_second": 9.479,
-      "eval_steps_per_second": 9.479,
+      "epoch": 0.522579017551447,
+      "eval_loss": 0.47057196497917175,
+      "eval_runtime": 3.5636,
+      "eval_samples_per_second": 28.062,
+      "eval_steps_per_second": 7.015,
       "step": 14500
     },
     {
-      "epoch": 0.51,
-      "learning_rate": 4.9322462231298206e-05,
-      "loss": 0.3215,
+      "epoch": 0.5227592172126716,
+      "grad_norm": 0.23088853061199188,
+      "learning_rate": 4.92754566141098e-05,
+      "loss": 0.4652,
       "step": 14505
     },
     {
-      "epoch": 0.51,
-      "learning_rate": 4.932180335607007e-05,
-      "loss": 0.3366,
+      "epoch": 0.5229394168738962,
+      "grad_norm": 0.20848578214645386,
+      "learning_rate": 4.9274758999350115e-05,
+      "loss": 0.4217,
       "step": 14510
     },
     {
-      "epoch": 0.51,
-      "learning_rate": 4.932114416503936e-05,
-      "loss": 0.3001,
+      "epoch": 0.523119616535121,
+      "grad_norm": 0.1422787457704544,
+      "learning_rate": 4.9274061053851525e-05,
+      "loss": 0.438,
       "step": 14515
     },
     {
-      "epoch": 0.51,
-      "learning_rate": 4.9320484658214605e-05,
-      "loss": 0.2973,
+      "epoch": 0.5232998161963456,
+      "grad_norm": 0.21294035017490387,
+      "learning_rate": 4.9273362777623555e-05,
+      "loss": 0.4632,
       "step": 14520
     },
     {
-      "epoch": 0.51,
-      "learning_rate": 4.9319824835604386e-05,
-      "loss": 0.3296,
+      "epoch": 0.5234800158575702,
+      "grad_norm": 0.1711554378271103,
+      "learning_rate": 4.927266417067572e-05,
+      "loss": 0.469,
       "step": 14525
     },
     {
-      "epoch": 0.51,
-      "learning_rate": 4.9319164697217276e-05,
-      "loss": 0.2947,
+      "epoch": 0.5236602155187948,
+      "grad_norm": 0.17742256820201874,
+      "learning_rate": 4.9271965233017527e-05,
+      "loss": 0.4506,
       "step": 14530
     },
     {
-      "epoch": 0.51,
-      "learning_rate": 4.931850424306183e-05,
-      "loss": 0.327,
+      "epoch": 0.5238404151800194,
+      "grad_norm": 0.17392541468143463,
+      "learning_rate": 4.9271265964658517e-05,
+      "loss": 0.4368,
       "step": 14535
     },
     {
-      "epoch": 0.51,
-      "learning_rate": 4.931784347314665e-05,
-      "loss": 0.3126,
+      "epoch": 0.5240206148412441,
+      "grad_norm": 0.15873579680919647,
+      "learning_rate": 4.92705663656082e-05,
+      "loss": 0.4176,
       "step": 14540
     },
     {
-      "epoch": 0.51,
-      "learning_rate": 4.931718238748029e-05,
-      "loss": 0.311,
+      "epoch": 0.5242008145024687,
+      "grad_norm": 0.19420740008354187,
+      "learning_rate": 4.926986643587612e-05,
+      "loss": 0.4758,
       "step": 14545
     },
     {
-      "epoch": 0.51,
-      "learning_rate": 4.9316520986071335e-05,
-      "loss": 0.3362,
+      "epoch": 0.5243810141636934,
+      "grad_norm": 0.16508692502975464,
+      "learning_rate": 4.9269166175471806e-05,
+      "loss": 0.4577,
       "step": 14550
     },
     {
-      "epoch": 0.51,
-      "learning_rate": 4.931585926892839e-05,
-      "loss": 0.3107,
+      "epoch": 0.524561213824918,
+      "grad_norm": 0.1393464207649231,
+      "learning_rate": 4.92684655844048e-05,
+      "loss": 0.4444,
       "step": 14555
     },
     {
-      "epoch": 0.51,
-      "learning_rate": 4.931519723606003e-05,
-      "loss": 0.3039,
+      "epoch": 0.5247414134861427,
+      "grad_norm": 0.16871263086795807,
+      "learning_rate": 4.9267764662684654e-05,
+      "loss": 0.4719,
       "step": 14560
     },
     {
-      "epoch": 0.51,
-      "learning_rate": 4.9314534887474864e-05,
-      "loss": 0.3238,
+      "epoch": 0.5249216131473673,
+      "grad_norm": 0.16914774477481842,
+      "learning_rate": 4.9267063410320907e-05,
+      "loss": 0.4371,
       "step": 14565
     },
     {
-      "epoch": 0.51,
-      "learning_rate": 4.9313872223181485e-05,
-      "loss": 0.3258,
+      "epoch": 0.5251018128085919,
+      "grad_norm": 0.15868036448955536,
+      "learning_rate": 4.926636182732313e-05,
+      "loss": 0.4464,
       "step": 14570
     },
     {
-      "epoch": 0.51,
-      "learning_rate": 4.93132092431885e-05,
-      "loss": 0.3171,
+      "epoch": 0.5252820124698165,
+      "grad_norm": 0.20540837943553925,
+      "learning_rate": 4.926565991370086e-05,
+      "loss": 0.4362,
       "step": 14575
     },
     {
-      "epoch": 0.51,
-      "learning_rate": 4.931254594750451e-05,
-      "loss": 0.2988,
+      "epoch": 0.5254622121310412,
+      "grad_norm": 0.20324106514453888,
+      "learning_rate": 4.926495766946368e-05,
+      "loss": 0.4428,
       "step": 14580
     },
     {
-      "epoch": 0.51,
-      "learning_rate": 4.9311882336138136e-05,
-      "loss": 0.3502,
+      "epoch": 0.5256424117922658,
+      "grad_norm": 0.190400630235672,
+      "learning_rate": 4.9264255094621135e-05,
+      "loss": 0.4453,
       "step": 14585
     },
     {
-      "epoch": 0.51,
-      "learning_rate": 4.931121840909799e-05,
-      "loss": 0.3244,
+      "epoch": 0.5258226114534905,
+      "grad_norm": 0.19358183443546295,
+      "learning_rate": 4.9263552189182826e-05,
+      "loss": 0.454,
       "step": 14590
     },
     {
-      "epoch": 0.51,
-      "learning_rate": 4.93105541663927e-05,
-      "loss": 0.3312,
+      "epoch": 0.5260028111147151,
+      "grad_norm": 0.18700343370437622,
+      "learning_rate": 4.926284895315831e-05,
+      "loss": 0.486,
       "step": 14595
     },
     {
-      "epoch": 0.51,
-      "learning_rate": 4.930988960803088e-05,
-      "loss": 0.2911,
+      "epoch": 0.5261830107759398,
+      "grad_norm": 0.1986667960882187,
+      "learning_rate": 4.926214538655718e-05,
+      "loss": 0.4768,
       "step": 14600
     },
     {
-      "epoch": 0.51,
-      "learning_rate": 4.930922473402117e-05,
-      "loss": 0.3002,
+      "epoch": 0.5263632104371644,
+      "grad_norm": 0.16327090561389923,
+      "learning_rate": 4.926144148938901e-05,
+      "loss": 0.4532,
       "step": 14605
     },
     {
-      "epoch": 0.51,
-      "learning_rate": 4.930855954437219e-05,
-      "loss": 0.2905,
+      "epoch": 0.526543410098389,
+      "grad_norm": 0.18385253846645355,
+      "learning_rate": 4.92607372616634e-05,
+      "loss": 0.4345,
       "step": 14610
     },
     {
-      "epoch": 0.51,
-      "learning_rate": 4.9307894039092596e-05,
-      "loss": 0.3078,
+      "epoch": 0.5267236097596136,
+      "grad_norm": 0.16108223795890808,
+      "learning_rate": 4.9260032703389936e-05,
+      "loss": 0.4524,
       "step": 14615
     },
     {
-      "epoch": 0.51,
-      "learning_rate": 4.9307228218191e-05,
-      "loss": 0.3168,
+      "epoch": 0.5269038094208383,
+      "grad_norm": 0.22687973082065582,
+      "learning_rate": 4.9259327814578234e-05,
+      "loss": 0.4706,
       "step": 14620
     },
     {
-      "epoch": 0.51,
-      "learning_rate": 4.9306562081676076e-05,
-      "loss": 0.3214,
+      "epoch": 0.5270840090820629,
+      "grad_norm": 0.20703905820846558,
+      "learning_rate": 4.925862259523788e-05,
+      "loss": 0.4411,
       "step": 14625
     },
     {
-      "epoch": 0.51,
-      "learning_rate": 4.9305895629556464e-05,
-      "loss": 0.3257,
+      "epoch": 0.5272642087432876,
+      "grad_norm": 0.17245493829250336,
+      "learning_rate": 4.925791704537849e-05,
+      "loss": 0.4483,
       "step": 14630
     },
     {
-      "epoch": 0.51,
-      "learning_rate": 4.930522886184081e-05,
-      "loss": 0.2989,
+      "epoch": 0.5274444084045122,
+      "grad_norm": 0.21393609046936035,
+      "learning_rate": 4.925721116500968e-05,
+      "loss": 0.4409,
       "step": 14635
     },
     {
-      "epoch": 0.52,
-      "learning_rate": 4.9304561778537786e-05,
-      "loss": 0.3381,
+      "epoch": 0.5276246080657369,
+      "grad_norm": 0.16517458856105804,
+      "learning_rate": 4.9256504954141066e-05,
+      "loss": 0.4424,
       "step": 14640
     },
     {
-      "epoch": 0.52,
-      "learning_rate": 4.9303894379656035e-05,
-      "loss": 0.3171,
+      "epoch": 0.5278048077269615,
+      "grad_norm": 0.1912750005722046,
+      "learning_rate": 4.925579841278226e-05,
+      "loss": 0.4668,
       "step": 14645
     },
     {
-      "epoch": 0.52,
-      "learning_rate": 4.930322666520423e-05,
-      "loss": 0.3156,
+      "epoch": 0.5279850073881861,
+      "grad_norm": 0.19278542697429657,
+      "learning_rate": 4.9255091540942905e-05,
+      "loss": 0.4311,
       "step": 14650
     },
     {
-      "epoch": 0.52,
-      "learning_rate": 4.9302558635191054e-05,
-      "loss": 0.3314,
+      "epoch": 0.5281652070494107,
+      "grad_norm": 0.14870208501815796,
+      "learning_rate": 4.925438433863262e-05,
+      "loss": 0.4558,
       "step": 14655
     },
     {
-      "epoch": 0.52,
-      "learning_rate": 4.9301890289625165e-05,
-      "loss": 0.316,
+      "epoch": 0.5283454067106353,
+      "grad_norm": 0.15407264232635498,
+      "learning_rate": 4.925367680586104e-05,
+      "loss": 0.4101,
       "step": 14660
     },
     {
-      "epoch": 0.52,
-      "learning_rate": 4.930122162851524e-05,
-      "loss": 0.3382,
+      "epoch": 0.52852560637186,
+      "grad_norm": 0.183024600148201,
+      "learning_rate": 4.925296894263782e-05,
+      "loss": 0.4485,
       "step": 14665
     },
     {
-      "epoch": 0.52,
-      "learning_rate": 4.930055265186997e-05,
-      "loss": 0.2921,
+      "epoch": 0.5287058060330847,
+      "grad_norm": 0.15193411707878113,
+      "learning_rate": 4.925226074897259e-05,
+      "loss": 0.4471,
       "step": 14670
     },
     {
-      "epoch": 0.52,
-      "learning_rate": 4.929988335969805e-05,
-      "loss": 0.3044,
+      "epoch": 0.5288860056943093,
+      "grad_norm": 0.18354560434818268,
+      "learning_rate": 4.9251552224875e-05,
+      "loss": 0.4246,
       "step": 14675
     },
     {
-      "epoch": 0.52,
-      "learning_rate": 4.929921375200814e-05,
-      "loss": 0.2977,
+      "epoch": 0.529066205355534,
+      "grad_norm": 0.16226573288440704,
+      "learning_rate": 4.9250843370354704e-05,
+      "loss": 0.4549,
       "step": 14680
     },
     {
-      "epoch": 0.52,
-      "learning_rate": 4.9298543828808966e-05,
-      "loss": 0.3022,
+      "epoch": 0.5292464050167586,
+      "grad_norm": 0.18447591364383698,
+      "learning_rate": 4.925013418542136e-05,
+      "loss": 0.4531,
       "step": 14685
     },
     {
-      "epoch": 0.52,
-      "learning_rate": 4.9297873590109213e-05,
-      "loss": 0.3132,
+      "epoch": 0.5294266046779832,
+      "grad_norm": 0.12299656867980957,
+      "learning_rate": 4.9249424670084636e-05,
+      "loss": 0.4247,
       "step": 14690
     },
     {
-      "epoch": 0.52,
-      "learning_rate": 4.9297203035917586e-05,
-      "loss": 0.3095,
+      "epoch": 0.5296068043392078,
+      "grad_norm": 0.19318439066410065,
+      "learning_rate": 4.92487148243542e-05,
+      "loss": 0.4277,
       "step": 14695
     },
     {
-      "epoch": 0.52,
-      "learning_rate": 4.9296532166242785e-05,
-      "loss": 0.3058,
+      "epoch": 0.5297870040004324,
+      "grad_norm": 0.22538743913173676,
+      "learning_rate": 4.924800464823971e-05,
+      "loss": 0.4561,
       "step": 14700
     },
     {
-      "epoch": 0.52,
-      "learning_rate": 4.929586098109352e-05,
-      "loss": 0.2842,
+      "epoch": 0.5299672036616572,
+      "grad_norm": 0.15232376754283905,
+      "learning_rate": 4.9247294141750864e-05,
+      "loss": 0.4437,
       "step": 14705
     },
     {
-      "epoch": 0.52,
-      "learning_rate": 4.929518948047852e-05,
-      "loss": 0.2967,
+      "epoch": 0.5301474033228818,
+      "grad_norm": 0.2040311098098755,
+      "learning_rate": 4.924658330489732e-05,
+      "loss": 0.451,
       "step": 14710
     },
     {
-      "epoch": 0.52,
-      "learning_rate": 4.92945176644065e-05,
-      "loss": 0.3238,
+      "epoch": 0.5303276029841064,
+      "grad_norm": 0.1730954647064209,
+      "learning_rate": 4.9245872137688776e-05,
+      "loss": 0.4292,
       "step": 14715
     },
     {
-      "epoch": 0.52,
-      "learning_rate": 4.9293845532886176e-05,
-      "loss": 0.3106,
+      "epoch": 0.530507802645331,
+      "grad_norm": 0.16043753921985626,
+      "learning_rate": 4.9245160640134916e-05,
+      "loss": 0.4814,
       "step": 14720
     },
     {
-      "epoch": 0.52,
-      "learning_rate": 4.929317308592627e-05,
-      "loss": 0.3143,
+      "epoch": 0.5306880023065557,
+      "grad_norm": 0.17977766692638397,
+      "learning_rate": 4.924444881224544e-05,
+      "loss": 0.4409,
       "step": 14725
     },
     {
-      "epoch": 0.52,
-      "learning_rate": 4.9292500323535526e-05,
-      "loss": 0.3155,
+      "epoch": 0.5308682019677803,
+      "grad_norm": 0.21359415352344513,
+      "learning_rate": 4.924373665403004e-05,
+      "loss": 0.4522,
       "step": 14730
     },
     {
-      "epoch": 0.52,
-      "learning_rate": 4.9291827245722674e-05,
-      "loss": 0.2988,
+      "epoch": 0.5310484016290049,
+      "grad_norm": 0.1500881314277649,
+      "learning_rate": 4.924302416549842e-05,
+      "loss": 0.4555,
       "step": 14735
     },
     {
-      "epoch": 0.52,
-      "learning_rate": 4.929115385249645e-05,
-      "loss": 0.3148,
+      "epoch": 0.5312286012902295,
+      "grad_norm": 0.18042995035648346,
+      "learning_rate": 4.92423113466603e-05,
+      "loss": 0.467,
       "step": 14740
     },
     {
-      "epoch": 0.52,
-      "learning_rate": 4.9290480143865616e-05,
-      "loss": 0.3172,
+      "epoch": 0.5314088009514543,
+      "grad_norm": 0.14702163636684418,
+      "learning_rate": 4.9241598197525374e-05,
+      "loss": 0.465,
       "step": 14745
     },
     {
-      "epoch": 0.52,
-      "learning_rate": 4.928980611983889e-05,
-      "loss": 0.3113,
+      "epoch": 0.5315890006126789,
+      "grad_norm": 0.14355307817459106,
+      "learning_rate": 4.9240884718103366e-05,
+      "loss": 0.4373,
       "step": 14750
     },
     {
-      "epoch": 0.52,
-      "learning_rate": 4.928913178042505e-05,
-      "loss": 0.3493,
+      "epoch": 0.5317692002739035,
+      "grad_norm": 0.17057915031909943,
+      "learning_rate": 4.9240170908403996e-05,
+      "loss": 0.4753,
       "step": 14755
     },
     {
-      "epoch": 0.52,
-      "learning_rate": 4.928845712563284e-05,
-      "loss": 0.3253,
+      "epoch": 0.5319493999351281,
+      "grad_norm": 0.16943103075027466,
+      "learning_rate": 4.9239456768436985e-05,
+      "loss": 0.442,
       "step": 14760
     },
     {
-      "epoch": 0.52,
-      "learning_rate": 4.928778215547101e-05,
-      "loss": 0.3245,
+      "epoch": 0.5321295995963528,
+      "grad_norm": 0.21389932930469513,
+      "learning_rate": 4.923874229821208e-05,
+      "loss": 0.4608,
       "step": 14765
     },
     {
-      "epoch": 0.52,
-      "learning_rate": 4.928710686994835e-05,
-      "loss": 0.3032,
+      "epoch": 0.5323097992575774,
+      "grad_norm": 0.17261211574077606,
+      "learning_rate": 4.9238027497738995e-05,
+      "loss": 0.4512,
       "step": 14770
     },
     {
-      "epoch": 0.52,
-      "learning_rate": 4.9286431269073596e-05,
-      "loss": 0.2691,
+      "epoch": 0.532489998918802,
+      "grad_norm": 0.1695832759141922,
+      "learning_rate": 4.9237312367027484e-05,
+      "loss": 0.4398,
       "step": 14775
     },
     {
-      "epoch": 0.52,
-      "learning_rate": 4.928575535285555e-05,
-      "loss": 0.2994,
+      "epoch": 0.5326701985800266,
+      "grad_norm": 0.17145270109176636,
+      "learning_rate": 4.923659690608728e-05,
+      "loss": 0.4597,
       "step": 14780
     },
     {
-      "epoch": 0.52,
-      "learning_rate": 4.9285079121302974e-05,
-      "loss": 0.3101,
+      "epoch": 0.5328503982412514,
+      "grad_norm": 0.193446546792984,
+      "learning_rate": 4.923588111492814e-05,
+      "loss": 0.4286,
       "step": 14785
     },
     {
-      "epoch": 0.52,
-      "learning_rate": 4.928440257442465e-05,
-      "loss": 0.3046,
+      "epoch": 0.533030597902476,
+      "grad_norm": 0.19881367683410645,
+      "learning_rate": 4.923516499355981e-05,
+      "loss": 0.4713,
       "step": 14790
     },
     {
-      "epoch": 0.52,
-      "learning_rate": 4.928372571222935e-05,
-      "loss": 0.3185,
+      "epoch": 0.5332107975637006,
+      "grad_norm": 0.16294337809085846,
+      "learning_rate": 4.9234448541992045e-05,
+      "loss": 0.3998,
       "step": 14795
     },
     {
-      "epoch": 0.52,
-      "learning_rate": 4.9283048534725896e-05,
-      "loss": 0.3414,
+      "epoch": 0.5333909972249252,
+      "grad_norm": 0.1420733481645584,
+      "learning_rate": 4.9233731760234616e-05,
+      "loss": 0.4301,
       "step": 14800
     },
     {
-      "epoch": 0.52,
-      "learning_rate": 4.928237104192305e-05,
-      "loss": 0.3148,
+      "epoch": 0.5335711968861498,
+      "grad_norm": 0.17047178745269775,
+      "learning_rate": 4.923301464829728e-05,
+      "loss": 0.4181,
       "step": 14805
     },
     {
-      "epoch": 0.52,
-      "learning_rate": 4.928169323382962e-05,
-      "loss": 0.3139,
+      "epoch": 0.5337513965473745,
+      "grad_norm": 0.13956919312477112,
+      "learning_rate": 4.923229720618981e-05,
+      "loss": 0.455,
       "step": 14810
     },
     {
-      "epoch": 0.52,
-      "learning_rate": 4.92810151104544e-05,
-      "loss": 0.3235,
+      "epoch": 0.5339315962085991,
+      "grad_norm": 0.17539288103580475,
+      "learning_rate": 4.923157943392199e-05,
+      "loss": 0.4681,
       "step": 14815
     },
     {
-      "epoch": 0.52,
-      "learning_rate": 4.9280336671806206e-05,
-      "loss": 0.3351,
+      "epoch": 0.5341117958698237,
+      "grad_norm": 0.19295528531074524,
+      "learning_rate": 4.9230861331503586e-05,
+      "loss": 0.4691,
       "step": 14820
     },
     {
-      "epoch": 0.52,
-      "learning_rate": 4.927965791789384e-05,
-      "loss": 0.3248,
+      "epoch": 0.5342919955310484,
+      "grad_norm": 0.15335458517074585,
+      "learning_rate": 4.923014289894439e-05,
+      "loss": 0.4094,
       "step": 14825
     },
     {
-      "epoch": 0.52,
-      "learning_rate": 4.927897884872612e-05,
-      "loss": 0.3035,
+      "epoch": 0.5344721951922731,
+      "grad_norm": 0.17784667015075684,
+      "learning_rate": 4.922942413625418e-05,
+      "loss": 0.4263,
       "step": 14830
     },
     {
-      "epoch": 0.52,
-      "learning_rate": 4.927829946431186e-05,
-      "loss": 0.2935,
+      "epoch": 0.5346523948534977,
+      "grad_norm": 0.14942409098148346,
+      "learning_rate": 4.922870504344276e-05,
+      "loss": 0.432,
       "step": 14835
     },
     {
-      "epoch": 0.52,
-      "learning_rate": 4.927761976465988e-05,
-      "loss": 0.3099,
+      "epoch": 0.5348325945147223,
+      "grad_norm": 0.1628066897392273,
+      "learning_rate": 4.9227985620519934e-05,
+      "loss": 0.4321,
       "step": 14840
     },
     {
-      "epoch": 0.52,
-      "learning_rate": 4.9276939749779005e-05,
-      "loss": 0.3211,
+      "epoch": 0.5350127941759469,
+      "grad_norm": 0.17270931601524353,
+      "learning_rate": 4.922726586749549e-05,
+      "loss": 0.4449,
       "step": 14845
     },
     {
-      "epoch": 0.52,
-      "learning_rate": 4.927625941967807e-05,
-      "loss": 0.3079,
+      "epoch": 0.5351929938371716,
+      "grad_norm": 0.1773119568824768,
+      "learning_rate": 4.922654578437923e-05,
+      "loss": 0.4791,
       "step": 14850
     },
     {
-      "epoch": 0.52,
-      "learning_rate": 4.92755787743659e-05,
-      "loss": 0.3197,
+      "epoch": 0.5353731934983962,
+      "grad_norm": 0.1501081883907318,
+      "learning_rate": 4.9225825371180985e-05,
+      "loss": 0.4242,
       "step": 14855
     },
     {
-      "epoch": 0.52,
-      "learning_rate": 4.927489781385134e-05,
-      "loss": 0.3212,
+      "epoch": 0.5355533931596208,
+      "grad_norm": 0.16684913635253906,
+      "learning_rate": 4.9225104627910553e-05,
+      "loss": 0.4395,
       "step": 14860
     },
     {
-      "epoch": 0.52,
-      "learning_rate": 4.927421653814323e-05,
-      "loss": 0.3044,
+      "epoch": 0.5357335928208455,
+      "grad_norm": 0.2014274299144745,
+      "learning_rate": 4.922438355457777e-05,
+      "loss": 0.4403,
       "step": 14865
     },
     {
-      "epoch": 0.52,
-      "learning_rate": 4.927353494725041e-05,
-      "loss": 0.3069,
+      "epoch": 0.5359137924820702,
+      "grad_norm": 0.14999887347221375,
+      "learning_rate": 4.922366215119244e-05,
+      "loss": 0.4354,
       "step": 14870
     },
     {
-      "epoch": 0.52,
-      "learning_rate": 4.9272853041181744e-05,
-      "loss": 0.3359,
+      "epoch": 0.5360939921432948,
+      "grad_norm": 0.17910043895244598,
+      "learning_rate": 4.922294041776441e-05,
+      "loss": 0.4549,
       "step": 14875
     },
     {
-      "epoch": 0.52,
-      "learning_rate": 4.927217081994607e-05,
-      "loss": 0.3104,
+      "epoch": 0.5362741918045194,
+      "grad_norm": 0.1802271157503128,
+      "learning_rate": 4.922221835430351e-05,
+      "loss": 0.4787,
       "step": 14880
     },
     {
-      "epoch": 0.52,
-      "learning_rate": 4.9271488283552266e-05,
-      "loss": 0.3353,
+      "epoch": 0.536454391465744,
+      "grad_norm": 0.15048733353614807,
+      "learning_rate": 4.922149596081956e-05,
+      "loss": 0.4334,
       "step": 14885
     },
     {
-      "epoch": 0.52,
-      "learning_rate": 4.927080543200917e-05,
-      "loss": 0.3059,
+      "epoch": 0.5366345911269687,
+      "grad_norm": 0.17902320623397827,
+      "learning_rate": 4.9220773237322424e-05,
+      "loss": 0.4446,
       "step": 14890
     },
     {
-      "epoch": 0.52,
-      "learning_rate": 4.9270122265325666e-05,
-      "loss": 0.3228,
+      "epoch": 0.5368147907881933,
+      "grad_norm": 0.2159227430820465,
+      "learning_rate": 4.922005018382195e-05,
+      "loss": 0.398,
       "step": 14895
     },
     {
-      "epoch": 0.52,
-      "learning_rate": 4.926943878351062e-05,
-      "loss": 0.3114,
+      "epoch": 0.536994990449418,
+      "grad_norm": 0.17634055018424988,
+      "learning_rate": 4.9219326800327967e-05,
+      "loss": 0.4409,
       "step": 14900
     },
     {
-      "epoch": 0.52,
-      "learning_rate": 4.926875498657291e-05,
-      "loss": 0.3156,
+      "epoch": 0.5371751901106426,
+      "grad_norm": 0.15450234711170197,
+      "learning_rate": 4.921860308685036e-05,
+      "loss": 0.482,
       "step": 14905
     },
     {
-      "epoch": 0.52,
-      "learning_rate": 4.92680708745214e-05,
-      "loss": 0.324,
+      "epoch": 0.5373553897718673,
+      "grad_norm": 0.18871042132377625,
+      "learning_rate": 4.921787904339897e-05,
+      "loss": 0.4748,
       "step": 14910
     },
     {
-      "epoch": 0.52,
-      "learning_rate": 4.926738644736498e-05,
-      "loss": 0.3226,
+      "epoch": 0.5375355894330919,
+      "grad_norm": 0.15520833432674408,
+      "learning_rate": 4.921715466998366e-05,
+      "loss": 0.4209,
       "step": 14915
     },
     {
-      "epoch": 0.52,
-      "learning_rate": 4.926670170511255e-05,
-      "loss": 0.3128,
+      "epoch": 0.5377157890943165,
+      "grad_norm": 0.17904292047023773,
+      "learning_rate": 4.921642996661431e-05,
+      "loss": 0.4665,
       "step": 14920
     },
     {
-      "epoch": 0.53,
-      "learning_rate": 4.9266016647772996e-05,
-      "loss": 0.3139,
+      "epoch": 0.5378959887555411,
+      "grad_norm": 0.23425260186195374,
+      "learning_rate": 4.9215704933300795e-05,
+      "loss": 0.4156,
       "step": 14925
     },
     {
-      "epoch": 0.53,
-      "learning_rate": 4.92653312753552e-05,
-      "loss": 0.2981,
+      "epoch": 0.5380761884167657,
+      "grad_norm": 0.18364645540714264,
+      "learning_rate": 4.921497957005299e-05,
+      "loss": 0.4648,
       "step": 14930
     },
     {
-      "epoch": 0.53,
-      "learning_rate": 4.9264645587868064e-05,
-      "loss": 0.3413,
+      "epoch": 0.5382563880779904,
+      "grad_norm": 0.13957159221172333,
+      "learning_rate": 4.921425387688077e-05,
+      "loss": 0.4577,
       "step": 14935
     },
     {
-      "epoch": 0.53,
-      "learning_rate": 4.92639595853205e-05,
-      "loss": 0.3121,
+      "epoch": 0.5384365877392151,
+      "grad_norm": 0.1416681855916977,
+      "learning_rate": 4.9213527853794025e-05,
+      "loss": 0.4146,
       "step": 14940
     },
     {
-      "epoch": 0.53,
-      "learning_rate": 4.926327326772141e-05,
-      "loss": 0.3223,
+      "epoch": 0.5386167874004397,
+      "grad_norm": 0.18616509437561035,
+      "learning_rate": 4.921280150080266e-05,
+      "loss": 0.4689,
       "step": 14945
     },
     {
-      "epoch": 0.53,
-      "learning_rate": 4.926258663507971e-05,
-      "loss": 0.3568,
+      "epoch": 0.5387969870616643,
+      "grad_norm": 0.2297711819410324,
+      "learning_rate": 4.9212074817916554e-05,
+      "loss": 0.4326,
       "step": 14950
     },
     {
-      "epoch": 0.53,
-      "learning_rate": 4.926189968740431e-05,
-      "loss": 0.3296,
+      "epoch": 0.538977186722889,
+      "grad_norm": 0.17574380338191986,
+      "learning_rate": 4.9211347805145626e-05,
+      "loss": 0.4521,
       "step": 14955
     },
     {
-      "epoch": 0.53,
-      "learning_rate": 4.926121242470413e-05,
-      "loss": 0.3021,
+      "epoch": 0.5391573863841136,
+      "grad_norm": 0.15670587122440338,
+      "learning_rate": 4.921062046249976e-05,
+      "loss": 0.4515,
       "step": 14960
     },
     {
-      "epoch": 0.53,
-      "learning_rate": 4.92605248469881e-05,
-      "loss": 0.3203,
+      "epoch": 0.5393375860453382,
+      "grad_norm": 0.17424379289150238,
+      "learning_rate": 4.9209892789988886e-05,
+      "loss": 0.4088,
       "step": 14965
     },
     {
-      "epoch": 0.53,
-      "learning_rate": 4.925983695426514e-05,
-      "loss": 0.3302,
+      "epoch": 0.5395177857065628,
+      "grad_norm": 0.16352404654026031,
+      "learning_rate": 4.920916478762291e-05,
+      "loss": 0.4369,
       "step": 14970
     },
     {
-      "epoch": 0.53,
-      "learning_rate": 4.925914874654418e-05,
-      "loss": 0.3069,
+      "epoch": 0.5396979853677875,
+      "grad_norm": 0.1390729397535324,
+      "learning_rate": 4.920843645541174e-05,
+      "loss": 0.442,
       "step": 14975
     },
     {
-      "epoch": 0.53,
-      "learning_rate": 4.925846022383417e-05,
-      "loss": 0.2889,
+      "epoch": 0.5398781850290122,
+      "grad_norm": 0.20522846281528473,
+      "learning_rate": 4.9207707793365325e-05,
+      "loss": 0.4439,
       "step": 14980
     },
     {
-      "epoch": 0.53,
-      "learning_rate": 4.9257771386144033e-05,
-      "loss": 0.3106,
+      "epoch": 0.5400583846902368,
+      "grad_norm": 0.18843060731887817,
+      "learning_rate": 4.9206978801493574e-05,
+      "loss": 0.4427,
       "step": 14985
     },
     {
-      "epoch": 0.53,
-      "learning_rate": 4.925708223348273e-05,
-      "loss": 0.2993,
+      "epoch": 0.5402385843514614,
+      "grad_norm": 0.1719266027212143,
+      "learning_rate": 4.920624947980642e-05,
+      "loss": 0.4301,
       "step": 14990
     },
     {
-      "epoch": 0.53,
-      "learning_rate": 4.9256392765859194e-05,
-      "loss": 0.3323,
+      "epoch": 0.5404187840126861,
+      "grad_norm": 0.21065543591976166,
+      "learning_rate": 4.9205519828313804e-05,
+      "loss": 0.4552,
       "step": 14995
     },
     {
-      "epoch": 0.53,
-      "learning_rate": 4.9255702983282394e-05,
-      "loss": 0.3239,
+      "epoch": 0.5405989836739107,
+      "grad_norm": 0.16618047654628754,
+      "learning_rate": 4.9204789847025666e-05,
+      "loss": 0.4583,
       "step": 15000
     },
     {
-      "epoch": 0.53,
-      "eval_loss": 0.3090079426765442,
-      "eval_runtime": 10.5523,
-      "eval_samples_per_second": 9.477,
-      "eval_steps_per_second": 9.477,
+      "epoch": 0.5405989836739107,
+      "eval_loss": 0.4699234068393707,
+      "eval_runtime": 3.5836,
+      "eval_samples_per_second": 27.905,
+      "eval_steps_per_second": 6.976,
       "step": 15000
     },
     {
-      "epoch": 0.53,
-      "learning_rate": 4.925501288576127e-05,
-      "loss": 0.3297,
+      "epoch": 0.5407791833351353,
+      "grad_norm": 0.16865772008895874,
+      "learning_rate": 4.920405953595196e-05,
+      "loss": 0.4513,
       "step": 15005
     },
     {
-      "epoch": 0.53,
-      "learning_rate": 4.925432247330477e-05,
-      "loss": 0.3281,
+      "epoch": 0.5409593829963599,
+      "grad_norm": 0.18350325524806976,
+      "learning_rate": 4.9203328895102616e-05,
+      "loss": 0.4655,
       "step": 15010
     },
     {
-      "epoch": 0.53,
-      "learning_rate": 4.92536317459219e-05,
-      "loss": 0.3328,
+      "epoch": 0.5411395826575846,
+      "grad_norm": 0.1351919025182724,
+      "learning_rate": 4.920259792448761e-05,
+      "loss": 0.4242,
       "step": 15015
     },
     {
-      "epoch": 0.53,
-      "learning_rate": 4.9252940703621596e-05,
-      "loss": 0.3205,
+      "epoch": 0.5413197823188093,
+      "grad_norm": 0.18003039062023163,
+      "learning_rate": 4.9201866624116896e-05,
+      "loss": 0.4616,
       "step": 15020
     },
     {
-      "epoch": 0.53,
-      "learning_rate": 4.925224934641284e-05,
-      "loss": 0.3209,
+      "epoch": 0.5414999819800339,
+      "grad_norm": 0.19233831763267517,
+      "learning_rate": 4.920113499400043e-05,
+      "loss": 0.4467,
       "step": 15025
     },
     {
-      "epoch": 0.53,
-      "learning_rate": 4.925155767430462e-05,
-      "loss": 0.3308,
+      "epoch": 0.5416801816412585,
+      "grad_norm": 0.16504104435443878,
+      "learning_rate": 4.9200403034148186e-05,
+      "loss": 0.4578,
       "step": 15030
     },
     {
-      "epoch": 0.53,
-      "learning_rate": 4.925086568730589e-05,
-      "loss": 0.3384,
+      "epoch": 0.5418603813024832,
+      "grad_norm": 0.2171589583158493,
+      "learning_rate": 4.919967074457014e-05,
+      "loss": 0.4737,
       "step": 15035
     },
     {
-      "epoch": 0.53,
-      "learning_rate": 4.925017338542566e-05,
-      "loss": 0.318,
+      "epoch": 0.5420405809637078,
+      "grad_norm": 0.17700451612472534,
+      "learning_rate": 4.919893812527626e-05,
+      "loss": 0.4803,
       "step": 15040
     },
     {
-      "epoch": 0.53,
-      "learning_rate": 4.924948076867291e-05,
-      "loss": 0.3091,
+      "epoch": 0.5422207806249324,
+      "grad_norm": 0.15637724101543427,
+      "learning_rate": 4.919820517627653e-05,
+      "loss": 0.4535,
       "step": 15045
     },
     {
-      "epoch": 0.53,
-      "learning_rate": 4.9248787837056633e-05,
-      "loss": 0.2807,
+      "epoch": 0.542400980286157,
+      "grad_norm": 0.18739740550518036,
+      "learning_rate": 4.9197471897580945e-05,
+      "loss": 0.4589,
       "step": 15050
     },
     {
-      "epoch": 0.53,
-      "learning_rate": 4.9248094590585825e-05,
-      "loss": 0.3351,
+      "epoch": 0.5425811799473818,
+      "grad_norm": 0.1649918407201767,
+      "learning_rate": 4.9196738289199484e-05,
+      "loss": 0.4825,
       "step": 15055
     },
     {
-      "epoch": 0.53,
-      "learning_rate": 4.924740102926948e-05,
-      "loss": 0.3397,
+      "epoch": 0.5427613796086064,
+      "grad_norm": 0.15141309797763824,
+      "learning_rate": 4.9196004351142156e-05,
+      "loss": 0.4802,
       "step": 15060
     },
     {
-      "epoch": 0.53,
-      "learning_rate": 4.9246707153116624e-05,
-      "loss": 0.286,
+      "epoch": 0.542941579269831,
+      "grad_norm": 0.18119317293167114,
+      "learning_rate": 4.9195270083418946e-05,
+      "loss": 0.469,
       "step": 15065
     },
     {
-      "epoch": 0.53,
-      "learning_rate": 4.9246012962136256e-05,
-      "loss": 0.3172,
+      "epoch": 0.5431217789310556,
+      "grad_norm": 0.145644873380661,
+      "learning_rate": 4.919453548603987e-05,
+      "loss": 0.4697,
       "step": 15070
     },
     {
-      "epoch": 0.53,
-      "learning_rate": 4.9245318456337384e-05,
-      "loss": 0.3121,
+      "epoch": 0.5433019785922802,
+      "grad_norm": 0.19345882534980774,
+      "learning_rate": 4.9193800559014935e-05,
+      "loss": 0.4733,
       "step": 15075
     },
     {
-      "epoch": 0.53,
-      "learning_rate": 4.924462363572903e-05,
-      "loss": 0.3049,
+      "epoch": 0.5434821782535049,
+      "grad_norm": 0.16904519498348236,
+      "learning_rate": 4.919306530235415e-05,
+      "loss": 0.4677,
       "step": 15080
     },
     {
-      "epoch": 0.53,
-      "learning_rate": 4.924392850032021e-05,
-      "loss": 0.3186,
+      "epoch": 0.5436623779147295,
+      "grad_norm": 0.15902064740657806,
+      "learning_rate": 4.919232971606753e-05,
+      "loss": 0.4743,
       "step": 15085
     },
     {
-      "epoch": 0.53,
-      "learning_rate": 4.924323305011996e-05,
-      "loss": 0.2936,
+      "epoch": 0.5438425775759541,
+      "grad_norm": 0.1665191799402237,
+      "learning_rate": 4.91915938001651e-05,
+      "loss": 0.4044,
       "step": 15090
     },
     {
-      "epoch": 0.53,
-      "learning_rate": 4.9242537285137314e-05,
-      "loss": 0.3289,
+      "epoch": 0.5440227772371788,
+      "grad_norm": 0.19838020205497742,
+      "learning_rate": 4.91908575546569e-05,
+      "loss": 0.4449,
       "step": 15095
     },
     {
-      "epoch": 0.53,
-      "learning_rate": 4.924184120538129e-05,
-      "loss": 0.3353,
+      "epoch": 0.5442029768984035,
+      "grad_norm": 0.1867869794368744,
+      "learning_rate": 4.919012097955294e-05,
+      "loss": 0.4659,
       "step": 15100
     },
     {
-      "epoch": 0.53,
-      "learning_rate": 4.924114481086094e-05,
-      "loss": 0.3201,
+      "epoch": 0.5443831765596281,
+      "grad_norm": 0.1696668267250061,
+      "learning_rate": 4.918938407486326e-05,
+      "loss": 0.4349,
       "step": 15105
     },
     {
-      "epoch": 0.53,
-      "learning_rate": 4.924044810158529e-05,
-      "loss": 0.3327,
+      "epoch": 0.5445633762208527,
+      "grad_norm": 0.15441665053367615,
+      "learning_rate": 4.918864684059792e-05,
+      "loss": 0.4425,
       "step": 15110
     },
     {
-      "epoch": 0.53,
-      "learning_rate": 4.9239751077563405e-05,
-      "loss": 0.3331,
+      "epoch": 0.5447435758820773,
+      "grad_norm": 0.1583663374185562,
+      "learning_rate": 4.918790927676694e-05,
+      "loss": 0.3933,
       "step": 15115
     },
     {
-      "epoch": 0.53,
-      "learning_rate": 4.923905373880433e-05,
-      "loss": 0.3006,
+      "epoch": 0.544923775543302,
+      "grad_norm": 0.16343651711940765,
+      "learning_rate": 4.918717138338038e-05,
+      "loss": 0.4006,
       "step": 15120
     },
     {
-      "epoch": 0.53,
-      "learning_rate": 4.923835608531711e-05,
-      "loss": 0.3307,
+      "epoch": 0.5451039752045266,
+      "grad_norm": 0.14582295715808868,
+      "learning_rate": 4.918643316044829e-05,
+      "loss": 0.4413,
       "step": 15125
     },
     {
-      "epoch": 0.53,
-      "learning_rate": 4.9237658117110805e-05,
-      "loss": 0.3225,
+      "epoch": 0.5452841748657512,
+      "grad_norm": 0.15008191764354706,
+      "learning_rate": 4.9185694607980737e-05,
+      "loss": 0.4231,
       "step": 15130
     },
     {
-      "epoch": 0.53,
-      "learning_rate": 4.923695983419449e-05,
-      "loss": 0.3058,
+      "epoch": 0.5454643745269759,
+      "grad_norm": 0.16880400478839874,
+      "learning_rate": 4.918495572598777e-05,
+      "loss": 0.4398,
       "step": 15135
     },
     {
-      "epoch": 0.53,
-      "learning_rate": 4.923626123657722e-05,
-      "loss": 0.307,
+      "epoch": 0.5456445741882006,
+      "grad_norm": 0.14454464614391327,
+      "learning_rate": 4.918421651447948e-05,
+      "loss": 0.4116,
       "step": 15140
     },
     {
-      "epoch": 0.53,
-      "learning_rate": 4.923556232426807e-05,
-      "loss": 0.3158,
+      "epoch": 0.5458247738494252,
+      "grad_norm": 0.22429820895195007,
+      "learning_rate": 4.9183476973465905e-05,
+      "loss": 0.4625,
       "step": 15145
     },
     {
-      "epoch": 0.53,
-      "learning_rate": 4.923486309727612e-05,
-      "loss": 0.3194,
+      "epoch": 0.5460049735106498,
+      "grad_norm": 0.17825621366500854,
+      "learning_rate": 4.9182737102957147e-05,
+      "loss": 0.4437,
       "step": 15150
     },
     {
-      "epoch": 0.53,
-      "learning_rate": 4.923416355561044e-05,
-      "loss": 0.2916,
+      "epoch": 0.5461851731718744,
+      "grad_norm": 0.16203615069389343,
+      "learning_rate": 4.9181996902963265e-05,
+      "loss": 0.4262,
       "step": 15155
     },
     {
-      "epoch": 0.53,
-      "learning_rate": 4.923346369928012e-05,
-      "loss": 0.3148,
+      "epoch": 0.546365372833099,
+      "grad_norm": 0.2014383226633072,
+      "learning_rate": 4.918125637349437e-05,
+      "loss": 0.479,
       "step": 15160
     },
     {
-      "epoch": 0.53,
-      "learning_rate": 4.923276352829424e-05,
-      "loss": 0.3263,
+      "epoch": 0.5465455724943237,
+      "grad_norm": 0.16691508889198303,
+      "learning_rate": 4.918051551456053e-05,
+      "loss": 0.4402,
       "step": 15165
     },
     {
-      "epoch": 0.53,
-      "learning_rate": 4.92320630426619e-05,
-      "loss": 0.3075,
+      "epoch": 0.5467257721555483,
+      "grad_norm": 0.18141350150108337,
+      "learning_rate": 4.917977432617186e-05,
+      "loss": 0.4648,
       "step": 15170
     },
     {
-      "epoch": 0.53,
-      "learning_rate": 4.9231362242392187e-05,
-      "loss": 0.3038,
+      "epoch": 0.546905971816773,
+      "grad_norm": 0.15297532081604004,
+      "learning_rate": 4.9179032808338435e-05,
+      "loss": 0.4106,
       "step": 15175
     },
     {
-      "epoch": 0.53,
-      "learning_rate": 4.92306611274942e-05,
-      "loss": 0.293,
+      "epoch": 0.5470861714779977,
+      "grad_norm": 0.14727525413036346,
+      "learning_rate": 4.917829096107037e-05,
+      "loss": 0.4584,
       "step": 15180
     },
     {
-      "epoch": 0.53,
-      "learning_rate": 4.9229959697977065e-05,
-      "loss": 0.3103,
+      "epoch": 0.5472663711392223,
+      "grad_norm": 0.18853066861629486,
+      "learning_rate": 4.917754878437778e-05,
+      "loss": 0.482,
       "step": 15185
     },
     {
-      "epoch": 0.53,
-      "learning_rate": 4.9229257953849855e-05,
-      "loss": 0.324,
+      "epoch": 0.5474465708004469,
+      "grad_norm": 0.196510449051857,
+      "learning_rate": 4.9176806278270757e-05,
+      "loss": 0.4466,
       "step": 15190
     },
     {
-      "epoch": 0.53,
-      "learning_rate": 4.9228555895121706e-05,
-      "loss": 0.2869,
+      "epoch": 0.5476267704616715,
+      "grad_norm": 0.16554872691631317,
+      "learning_rate": 4.917606344275944e-05,
+      "loss": 0.4687,
       "step": 15195
     },
     {
-      "epoch": 0.53,
-      "learning_rate": 4.9227853521801726e-05,
-      "loss": 0.3081,
+      "epoch": 0.5478069701228961,
+      "grad_norm": 0.15905487537384033,
+      "learning_rate": 4.917532027785394e-05,
+      "loss": 0.4855,
       "step": 15200
     },
     {
-      "epoch": 0.53,
-      "learning_rate": 4.922715083389903e-05,
-      "loss": 0.3291,
+      "epoch": 0.5479871697841208,
+      "grad_norm": 0.17025235295295715,
+      "learning_rate": 4.917457678356437e-05,
+      "loss": 0.4565,
       "step": 15205
     },
     {
-      "epoch": 0.54,
-      "learning_rate": 4.922644783142274e-05,
-      "loss": 0.3184,
+      "epoch": 0.5481673694453455,
+      "grad_norm": 0.18130546808242798,
+      "learning_rate": 4.917383295990088e-05,
+      "loss": 0.456,
       "step": 15210
     },
     {
-      "epoch": 0.54,
-      "learning_rate": 4.9225744514382007e-05,
-      "loss": 0.3145,
+      "epoch": 0.5483475691065701,
+      "grad_norm": 0.13437901437282562,
+      "learning_rate": 4.9173088806873596e-05,
+      "loss": 0.4207,
       "step": 15215
     },
     {
-      "epoch": 0.54,
-      "learning_rate": 4.922504088278593e-05,
-      "loss": 0.3139,
+      "epoch": 0.5485277687677947,
+      "grad_norm": 0.1458578109741211,
+      "learning_rate": 4.917234432449266e-05,
+      "loss": 0.4671,
       "step": 15220
     },
     {
-      "epoch": 0.54,
-      "learning_rate": 4.922433693664368e-05,
-      "loss": 0.3348,
+      "epoch": 0.5487079684290194,
+      "grad_norm": 0.16906622052192688,
+      "learning_rate": 4.9171599512768206e-05,
+      "loss": 0.4346,
       "step": 15225
     },
     {
-      "epoch": 0.54,
-      "learning_rate": 4.922363267596437e-05,
-      "loss": 0.3065,
+      "epoch": 0.548888168090244,
+      "grad_norm": 0.17966735363006592,
+      "learning_rate": 4.917085437171038e-05,
+      "loss": 0.4734,
       "step": 15230
     },
     {
-      "epoch": 0.54,
-      "learning_rate": 4.922292810075715e-05,
-      "loss": 0.3147,
+      "epoch": 0.5490683677514686,
+      "grad_norm": 0.135285422205925,
+      "learning_rate": 4.917010890132936e-05,
+      "loss": 0.4293,
       "step": 15235
     },
     {
-      "epoch": 0.54,
-      "learning_rate": 4.9222223211031185e-05,
-      "loss": 0.3264,
+      "epoch": 0.5492485674126932,
+      "grad_norm": 0.15000662207603455,
+      "learning_rate": 4.916936310163528e-05,
+      "loss": 0.4076,
       "step": 15240
     },
     {
-      "epoch": 0.54,
-      "learning_rate": 4.9221518006795595e-05,
-      "loss": 0.2945,
+      "epoch": 0.5494287670739179,
+      "grad_norm": 0.17892701923847198,
+      "learning_rate": 4.9168616972638304e-05,
+      "loss": 0.4461,
       "step": 15245
     },
     {
-      "epoch": 0.54,
-      "learning_rate": 4.922081248805958e-05,
-      "loss": 0.2911,
+      "epoch": 0.5496089667351426,
+      "grad_norm": 0.16764943301677704,
+      "learning_rate": 4.91678705143486e-05,
+      "loss": 0.4662,
       "step": 15250
     },
     {
-      "epoch": 0.54,
-      "learning_rate": 4.922010665483226e-05,
-      "loss": 0.3106,
+      "epoch": 0.5497891663963672,
+      "grad_norm": 0.19673018157482147,
+      "learning_rate": 4.916712372677635e-05,
+      "loss": 0.4367,
       "step": 15255
     },
     {
-      "epoch": 0.54,
-      "learning_rate": 4.921940050712281e-05,
-      "loss": 0.3063,
+      "epoch": 0.5499693660575918,
+      "grad_norm": 0.13246646523475647,
+      "learning_rate": 4.916637660993171e-05,
+      "loss": 0.4468,
       "step": 15260
     },
     {
-      "epoch": 0.54,
-      "learning_rate": 4.9218694044940424e-05,
-      "loss": 0.3207,
+      "epoch": 0.5501495657188165,
+      "grad_norm": 0.1905840039253235,
+      "learning_rate": 4.916562916382487e-05,
+      "loss": 0.4495,
       "step": 15265
     },
     {
-      "epoch": 0.54,
-      "learning_rate": 4.9217987268294244e-05,
-      "loss": 0.2964,
+      "epoch": 0.5503297653800411,
+      "grad_norm": 0.18271660804748535,
+      "learning_rate": 4.916488138846601e-05,
+      "loss": 0.4694,
       "step": 15270
     },
     {
-      "epoch": 0.54,
-      "learning_rate": 4.921728017719347e-05,
-      "loss": 0.344,
+      "epoch": 0.5505099650412657,
+      "grad_norm": 0.15053506195545197,
+      "learning_rate": 4.916413328386531e-05,
+      "loss": 0.4371,
       "step": 15275
     },
     {
-      "epoch": 0.54,
-      "learning_rate": 4.921657277164726e-05,
-      "loss": 0.2759,
+      "epoch": 0.5506901647024903,
+      "grad_norm": 0.17273497581481934,
+      "learning_rate": 4.916338485003298e-05,
+      "loss": 0.4595,
       "step": 15280
     },
     {
-      "epoch": 0.54,
-      "learning_rate": 4.9215865051664824e-05,
-      "loss": 0.3429,
+      "epoch": 0.550870364363715,
+      "grad_norm": 0.1739252358675003,
+      "learning_rate": 4.916263608697921e-05,
+      "loss": 0.4688,
       "step": 15285
     },
     {
-      "epoch": 0.54,
-      "learning_rate": 4.9215157017255333e-05,
-      "loss": 0.3455,
+      "epoch": 0.5510505640249397,
+      "grad_norm": 0.16885848343372345,
+      "learning_rate": 4.916188699471421e-05,
+      "loss": 0.4307,
       "step": 15290
     },
     {
-      "epoch": 0.54,
-      "learning_rate": 4.921444866842799e-05,
-      "loss": 0.3217,
+      "epoch": 0.5512307636861643,
+      "grad_norm": 0.15223874151706696,
+      "learning_rate": 4.916113757324817e-05,
+      "loss": 0.4597,
       "step": 15295
     },
     {
-      "epoch": 0.54,
-      "learning_rate": 4.921374000519199e-05,
-      "loss": 0.3146,
+      "epoch": 0.5514109633473889,
+      "grad_norm": 0.2001715898513794,
+      "learning_rate": 4.9160387822591306e-05,
+      "loss": 0.4417,
       "step": 15300
     },
     {
-      "epoch": 0.54,
-      "learning_rate": 4.9213031027556526e-05,
-      "loss": 0.294,
+      "epoch": 0.5515911630086135,
+      "grad_norm": 0.1790451556444168,
+      "learning_rate": 4.915963774275384e-05,
+      "loss": 0.447,
       "step": 15305
     },
     {
-      "epoch": 0.54,
-      "learning_rate": 4.921232173553082e-05,
-      "loss": 0.3116,
+      "epoch": 0.5517713626698382,
+      "grad_norm": 0.16498644649982452,
+      "learning_rate": 4.915888733374598e-05,
+      "loss": 0.4523,
       "step": 15310
     },
     {
-      "epoch": 0.54,
-      "learning_rate": 4.921161212912406e-05,
-      "loss": 0.332,
+      "epoch": 0.5519515623310628,
+      "grad_norm": 0.13701501488685608,
+      "learning_rate": 4.915813659557796e-05,
+      "loss": 0.4299,
       "step": 15315
     },
     {
-      "epoch": 0.54,
-      "learning_rate": 4.9210902208345486e-05,
-      "loss": 0.3257,
+      "epoch": 0.5521317619922874,
+      "grad_norm": 0.16090936958789825,
+      "learning_rate": 4.9157385528260016e-05,
+      "loss": 0.4038,
       "step": 15320
     },
     {
-      "epoch": 0.54,
-      "learning_rate": 4.921019197320429e-05,
-      "loss": 0.3238,
+      "epoch": 0.552311961653512,
+      "grad_norm": 0.156123086810112,
+      "learning_rate": 4.915663413180236e-05,
+      "loss": 0.4675,
       "step": 15325
     },
     {
-      "epoch": 0.54,
-      "learning_rate": 4.920948142370972e-05,
-      "loss": 0.2883,
+      "epoch": 0.5524921613147368,
+      "grad_norm": 0.17523333430290222,
+      "learning_rate": 4.915588240621524e-05,
+      "loss": 0.4965,
       "step": 15330
     },
     {
-      "epoch": 0.54,
-      "learning_rate": 4.9208770559870984e-05,
-      "loss": 0.3216,
+      "epoch": 0.5526723609759614,
+      "grad_norm": 0.1823972910642624,
+      "learning_rate": 4.9155130351508904e-05,
+      "loss": 0.4191,
       "step": 15335
     },
     {
-      "epoch": 0.54,
-      "learning_rate": 4.9208059381697314e-05,
-      "loss": 0.3331,
+      "epoch": 0.552852560637186,
+      "grad_norm": 0.19572138786315918,
+      "learning_rate": 4.915437796769359e-05,
+      "loss": 0.4305,
       "step": 15340
     },
     {
-      "epoch": 0.54,
-      "learning_rate": 4.920734788919794e-05,
-      "loss": 0.3136,
+      "epoch": 0.5530327602984106,
+      "grad_norm": 0.16662746667861938,
+      "learning_rate": 4.915362525477955e-05,
+      "loss": 0.4268,
       "step": 15345
     },
     {
-      "epoch": 0.54,
-      "learning_rate": 4.9206636082382115e-05,
-      "loss": 0.2992,
+      "epoch": 0.5532129599596353,
+      "grad_norm": 0.22275352478027344,
+      "learning_rate": 4.915287221277706e-05,
+      "loss": 0.4884,
       "step": 15350
     },
     {
-      "epoch": 0.54,
-      "learning_rate": 4.920592396125907e-05,
-      "loss": 0.3019,
+      "epoch": 0.5533931596208599,
+      "grad_norm": 0.19144576787948608,
+      "learning_rate": 4.915211884169635e-05,
+      "loss": 0.4326,
       "step": 15355
     },
     {
-      "epoch": 0.54,
-      "learning_rate": 4.920521152583806e-05,
-      "loss": 0.3563,
+      "epoch": 0.5535733592820845,
+      "grad_norm": 0.19240762293338776,
+      "learning_rate": 4.915136514154769e-05,
+      "loss": 0.4484,
       "step": 15360
     },
     {
-      "epoch": 0.54,
-      "learning_rate": 4.920449877612833e-05,
-      "loss": 0.3187,
+      "epoch": 0.5537535589433091,
+      "grad_norm": 0.13195356726646423,
+      "learning_rate": 4.915061111234136e-05,
+      "loss": 0.4294,
       "step": 15365
     },
     {
-      "epoch": 0.54,
-      "learning_rate": 4.9203785712139126e-05,
-      "loss": 0.3117,
+      "epoch": 0.5539337586045339,
+      "grad_norm": 0.1802310198545456,
+      "learning_rate": 4.914985675408763e-05,
+      "loss": 0.4616,
       "step": 15370
     },
     {
-      "epoch": 0.54,
-      "learning_rate": 4.920307233387973e-05,
-      "loss": 0.3155,
+      "epoch": 0.5541139582657585,
+      "grad_norm": 0.12225214391946793,
+      "learning_rate": 4.914910206679678e-05,
+      "loss": 0.3886,
       "step": 15375
     },
     {
-      "epoch": 0.54,
-      "learning_rate": 4.9202358641359375e-05,
-      "loss": 0.305,
+      "epoch": 0.5542941579269831,
+      "grad_norm": 0.139492928981781,
+      "learning_rate": 4.914834705047909e-05,
+      "loss": 0.447,
       "step": 15380
     },
     {
-      "epoch": 0.54,
-      "learning_rate": 4.9201644634587354e-05,
-      "loss": 0.29,
+      "epoch": 0.5544743575882077,
+      "grad_norm": 0.14158743619918823,
+      "learning_rate": 4.9147591705144844e-05,
+      "loss": 0.4547,
       "step": 15385
     },
     {
-      "epoch": 0.54,
-      "learning_rate": 4.920093031357292e-05,
-      "loss": 0.3041,
+      "epoch": 0.5546545572494324,
+      "grad_norm": 0.164401575922966,
+      "learning_rate": 4.9146836030804346e-05,
+      "loss": 0.4442,
       "step": 15390
     },
     {
-      "epoch": 0.54,
-      "learning_rate": 4.9200215678325354e-05,
-      "loss": 0.3318,
+      "epoch": 0.554834756910657,
+      "grad_norm": 0.17058567702770233,
+      "learning_rate": 4.914608002746787e-05,
+      "loss": 0.4267,
       "step": 15395
     },
     {
-      "epoch": 0.54,
-      "learning_rate": 4.9199500728853944e-05,
-      "loss": 0.3196,
+      "epoch": 0.5550149565718816,
+      "grad_norm": 0.18380822241306305,
+      "learning_rate": 4.914532369514573e-05,
+      "loss": 0.4541,
       "step": 15400
     },
     {
-      "epoch": 0.54,
-      "learning_rate": 4.919878546516795e-05,
-      "loss": 0.3327,
+      "epoch": 0.5551951562331063,
+      "grad_norm": 0.1776258945465088,
+      "learning_rate": 4.914456703384823e-05,
+      "loss": 0.4566,
       "step": 15405
     },
     {
-      "epoch": 0.54,
-      "learning_rate": 4.919806988727669e-05,
-      "loss": 0.2934,
+      "epoch": 0.555375355894331,
+      "grad_norm": 0.13963723182678223,
+      "learning_rate": 4.914381004358568e-05,
+      "loss": 0.4172,
       "step": 15410
     },
     {
-      "epoch": 0.54,
-      "learning_rate": 4.9197353995189434e-05,
-      "loss": 0.2936,
+      "epoch": 0.5555555555555556,
+      "grad_norm": 0.1411467045545578,
+      "learning_rate": 4.9143052724368396e-05,
+      "loss": 0.3885,
       "step": 15415
     },
     {
-      "epoch": 0.54,
-      "learning_rate": 4.9196637788915486e-05,
-      "loss": 0.3291,
+      "epoch": 0.5557357552167802,
+      "grad_norm": 0.1662912219762802,
+      "learning_rate": 4.914229507620669e-05,
+      "loss": 0.4943,
       "step": 15420
     },
     {
-      "epoch": 0.54,
-      "learning_rate": 4.9195921268464143e-05,
-      "loss": 0.3002,
+      "epoch": 0.5559159548780048,
+      "grad_norm": 0.2055153101682663,
+      "learning_rate": 4.914153709911088e-05,
+      "loss": 0.4344,
       "step": 15425
     },
     {
-      "epoch": 0.54,
-      "learning_rate": 4.91952044338447e-05,
-      "loss": 0.3251,
+      "epoch": 0.5560961545392294,
+      "grad_norm": 0.17844471335411072,
+      "learning_rate": 4.9140778793091316e-05,
+      "loss": 0.4778,
       "step": 15430
     },
     {
-      "epoch": 0.54,
-      "learning_rate": 4.9194487285066474e-05,
-      "loss": 0.3033,
+      "epoch": 0.5562763542004541,
+      "grad_norm": 0.17044708132743835,
+      "learning_rate": 4.9140020158158305e-05,
+      "loss": 0.4267,
       "step": 15435
     },
     {
-      "epoch": 0.54,
-      "learning_rate": 4.919376982213878e-05,
-      "loss": 0.284,
+      "epoch": 0.5564565538616787,
+      "grad_norm": 0.1435849368572235,
+      "learning_rate": 4.91392611943222e-05,
+      "loss": 0.4465,
       "step": 15440
     },
     {
-      "epoch": 0.54,
-      "learning_rate": 4.919305204507093e-05,
-      "loss": 0.3229,
+      "epoch": 0.5566367535229034,
+      "grad_norm": 0.19326213002204895,
+      "learning_rate": 4.913850190159333e-05,
+      "loss": 0.4709,
       "step": 15445
     },
     {
-      "epoch": 0.54,
-      "learning_rate": 4.919233395387224e-05,
-      "loss": 0.3252,
+      "epoch": 0.556816953184128,
+      "grad_norm": 0.17136943340301514,
+      "learning_rate": 4.9137742279982035e-05,
+      "loss": 0.4668,
       "step": 15450
     },
     {
-      "epoch": 0.54,
-      "learning_rate": 4.919161554855204e-05,
-      "loss": 0.3176,
+      "epoch": 0.5569971528453527,
+      "grad_norm": 0.16831465065479279,
+      "learning_rate": 4.913698232949868e-05,
+      "loss": 0.471,
       "step": 15455
     },
     {
-      "epoch": 0.54,
-      "learning_rate": 4.919089682911965e-05,
-      "loss": 0.3115,
+      "epoch": 0.5571773525065773,
+      "grad_norm": 0.20569349825382233,
+      "learning_rate": 4.9136222050153626e-05,
+      "loss": 0.4573,
       "step": 15460
     },
     {
-      "epoch": 0.54,
-      "learning_rate": 4.919017779558441e-05,
-      "loss": 0.3387,
+      "epoch": 0.5573575521678019,
+      "grad_norm": 0.1882450133562088,
+      "learning_rate": 4.913546144195721e-05,
+      "loss": 0.4867,
       "step": 15465
     },
     {
-      "epoch": 0.54,
-      "learning_rate": 4.918945844795565e-05,
-      "loss": 0.301,
+      "epoch": 0.5575377518290265,
+      "grad_norm": 0.15884269773960114,
+      "learning_rate": 4.9134700504919805e-05,
+      "loss": 0.4389,
       "step": 15470
     },
     {
-      "epoch": 0.54,
-      "learning_rate": 4.918873878624272e-05,
-      "loss": 0.3064,
+      "epoch": 0.5577179514902512,
+      "grad_norm": 0.16877703368663788,
+      "learning_rate": 4.913393923905178e-05,
+      "loss": 0.4443,
       "step": 15475
     },
     {
-      "epoch": 0.54,
-      "learning_rate": 4.918801881045495e-05,
-      "loss": 0.3127,
+      "epoch": 0.5578981511514758,
+      "grad_norm": 0.18422040343284607,
+      "learning_rate": 4.9133177644363506e-05,
+      "loss": 0.4407,
       "step": 15480
     },
     {
-      "epoch": 0.54,
-      "learning_rate": 4.91872985206017e-05,
-      "loss": 0.3176,
+      "epoch": 0.5580783508127005,
+      "grad_norm": 0.1388750970363617,
+      "learning_rate": 4.9132415720865355e-05,
+      "loss": 0.4309,
       "step": 15485
     },
     {
-      "epoch": 0.54,
-      "learning_rate": 4.9186577916692325e-05,
-      "loss": 0.3175,
+      "epoch": 0.5582585504739251,
+      "grad_norm": 0.16202110052108765,
+      "learning_rate": 4.913165346856772e-05,
+      "loss": 0.4428,
       "step": 15490
     },
     {
-      "epoch": 0.55,
-      "learning_rate": 4.918585699873617e-05,
-      "loss": 0.3359,
+      "epoch": 0.5584387501351498,
+      "grad_norm": 0.17716795206069946,
+      "learning_rate": 4.9130890887480966e-05,
+      "loss": 0.4904,
       "step": 15495
     },
     {
-      "epoch": 0.55,
-      "learning_rate": 4.918513576674259e-05,
-      "loss": 0.3018,
+      "epoch": 0.5586189497963744,
+      "grad_norm": 0.14335161447525024,
+      "learning_rate": 4.91301279776155e-05,
+      "loss": 0.4346,
       "step": 15500
     },
     {
-      "epoch": 0.55,
-      "eval_loss": 0.309041827917099,
-      "eval_runtime": 10.5379,
-      "eval_samples_per_second": 9.49,
-      "eval_steps_per_second": 9.49,
+      "epoch": 0.5586189497963744,
+      "eval_loss": 0.4687560200691223,
+      "eval_runtime": 3.5676,
+      "eval_samples_per_second": 28.03,
+      "eval_steps_per_second": 7.008,
       "step": 15500
     },
     {
-      "epoch": 0.55,
-      "learning_rate": 4.918441422072098e-05,
-      "loss": 0.3475,
+      "epoch": 0.558799149457599,
+      "grad_norm": 0.1963035613298416,
+      "learning_rate": 4.912936473898172e-05,
+      "loss": 0.4499,
       "step": 15505
     },
     {
-      "epoch": 0.55,
-      "learning_rate": 4.9183692360680676e-05,
-      "loss": 0.3085,
+      "epoch": 0.5589793491188236,
+      "grad_norm": 0.1802736520767212,
+      "learning_rate": 4.912860117159001e-05,
+      "loss": 0.4784,
       "step": 15510
     },
     {
-      "epoch": 0.55,
-      "learning_rate": 4.918297018663107e-05,
-      "loss": 0.3378,
+      "epoch": 0.5591595487800483,
+      "grad_norm": 0.16385473310947418,
+      "learning_rate": 4.9127837275450786e-05,
+      "loss": 0.428,
       "step": 15515
     },
     {
-      "epoch": 0.55,
-      "learning_rate": 4.918224769858153e-05,
-      "loss": 0.3255,
+      "epoch": 0.5593397484412729,
+      "grad_norm": 0.1578623354434967,
+      "learning_rate": 4.9127073050574445e-05,
+      "loss": 0.4125,
       "step": 15520
     },
     {
-      "epoch": 0.55,
-      "learning_rate": 4.918152489654145e-05,
-      "loss": 0.3167,
+      "epoch": 0.5595199481024976,
+      "grad_norm": 0.2023286670446396,
+      "learning_rate": 4.912630849697141e-05,
+      "loss": 0.4491,
       "step": 15525
     },
     {
-      "epoch": 0.55,
-      "learning_rate": 4.918080178052019e-05,
-      "loss": 0.3193,
+      "epoch": 0.5597001477637222,
+      "grad_norm": 0.17793312668800354,
+      "learning_rate": 4.912554361465208e-05,
+      "loss": 0.435,
       "step": 15530
     },
     {
-      "epoch": 0.55,
-      "learning_rate": 4.918007835052717e-05,
-      "loss": 0.311,
+      "epoch": 0.5598803474249469,
+      "grad_norm": 0.18119443953037262,
+      "learning_rate": 4.91247784036269e-05,
+      "loss": 0.4516,
       "step": 15535
     },
     {
-      "epoch": 0.55,
-      "learning_rate": 4.917935460657176e-05,
-      "loss": 0.3121,
+      "epoch": 0.5600605470861715,
+      "grad_norm": 0.15839527547359467,
+      "learning_rate": 4.912401286390629e-05,
+      "loss": 0.4549,
       "step": 15540
     },
     {
-      "epoch": 0.55,
-      "learning_rate": 4.917863054866336e-05,
-      "loss": 0.3069,
+      "epoch": 0.5602407467473961,
+      "grad_norm": 0.19357532262802124,
+      "learning_rate": 4.912324699550066e-05,
+      "loss": 0.4639,
       "step": 15545
     },
     {
-      "epoch": 0.55,
-      "learning_rate": 4.9177906176811384e-05,
-      "loss": 0.2918,
+      "epoch": 0.5604209464086207,
+      "grad_norm": 0.16731002926826477,
+      "learning_rate": 4.9122480798420474e-05,
+      "loss": 0.4079,
       "step": 15550
     },
     {
-      "epoch": 0.55,
-      "learning_rate": 4.917718149102523e-05,
-      "loss": 0.2929,
+      "epoch": 0.5606011460698453,
+      "grad_norm": 0.16960617899894714,
+      "learning_rate": 4.912171427267615e-05,
+      "loss": 0.4516,
       "step": 15555
     },
     {
-      "epoch": 0.55,
-      "learning_rate": 4.9176456491314304e-05,
-      "loss": 0.3343,
+      "epoch": 0.5607813457310701,
+      "grad_norm": 0.16323991119861603,
+      "learning_rate": 4.912094741827814e-05,
+      "loss": 0.4396,
       "step": 15560
     },
     {
-      "epoch": 0.55,
-      "learning_rate": 4.917573117768803e-05,
-      "loss": 0.2854,
+      "epoch": 0.5609615453922947,
+      "grad_norm": 0.17487618327140808,
+      "learning_rate": 4.9120180235236895e-05,
+      "loss": 0.4219,
       "step": 15565
     },
     {
-      "epoch": 0.55,
-      "learning_rate": 4.917500555015582e-05,
-      "loss": 0.3236,
+      "epoch": 0.5611417450535193,
+      "grad_norm": 0.15167950093746185,
+      "learning_rate": 4.911941272356286e-05,
+      "loss": 0.4486,
       "step": 15570
     },
     {
-      "epoch": 0.55,
-      "learning_rate": 4.917427960872708e-05,
-      "loss": 0.3021,
+      "epoch": 0.561321944714744,
+      "grad_norm": 0.17126460373401642,
+      "learning_rate": 4.91186448832665e-05,
+      "loss": 0.4338,
       "step": 15575
     },
     {
-      "epoch": 0.55,
-      "learning_rate": 4.917355335341126e-05,
-      "loss": 0.3098,
+      "epoch": 0.5615021443759686,
+      "grad_norm": 0.18867506086826324,
+      "learning_rate": 4.911787671435827e-05,
+      "loss": 0.4422,
       "step": 15580
     },
     {
-      "epoch": 0.55,
-      "learning_rate": 4.9172826784217786e-05,
-      "loss": 0.3106,
+      "epoch": 0.5616823440371932,
+      "grad_norm": 0.1628337949514389,
+      "learning_rate": 4.911710821684864e-05,
+      "loss": 0.4467,
       "step": 15585
     },
     {
-      "epoch": 0.55,
-      "learning_rate": 4.9172099901156076e-05,
-      "loss": 0.3122,
+      "epoch": 0.5618625436984178,
+      "grad_norm": 0.1885543018579483,
+      "learning_rate": 4.911633939074808e-05,
+      "loss": 0.4332,
       "step": 15590
     },
     {
-      "epoch": 0.55,
-      "learning_rate": 4.9171372704235584e-05,
-      "loss": 0.2936,
+      "epoch": 0.5620427433596424,
+      "grad_norm": 0.20939011871814728,
+      "learning_rate": 4.9115570236067065e-05,
+      "loss": 0.4422,
       "step": 15595
     },
     {
-      "epoch": 0.55,
-      "learning_rate": 4.917064519346575e-05,
-      "loss": 0.329,
+      "epoch": 0.5622229430208672,
+      "grad_norm": 0.15987086296081543,
+      "learning_rate": 4.911480075281607e-05,
+      "loss": 0.4408,
       "step": 15600
     },
     {
-      "epoch": 0.55,
-      "learning_rate": 4.916991736885602e-05,
-      "loss": 0.285,
+      "epoch": 0.5624031426820918,
+      "grad_norm": 0.13029427826404572,
+      "learning_rate": 4.911403094100559e-05,
+      "loss": 0.4403,
       "step": 15605
     },
     {
-      "epoch": 0.55,
-      "learning_rate": 4.916918923041584e-05,
-      "loss": 0.3033,
+      "epoch": 0.5625833423433164,
+      "grad_norm": 0.18694214522838593,
+      "learning_rate": 4.91132608006461e-05,
+      "loss": 0.4415,
       "step": 15610
     },
     {
-      "epoch": 0.55,
-      "learning_rate": 4.916846077815466e-05,
-      "loss": 0.3181,
+      "epoch": 0.562763542004541,
+      "grad_norm": 0.1758272796869278,
+      "learning_rate": 4.9112490331748105e-05,
+      "loss": 0.4124,
       "step": 15615
     },
     {
-      "epoch": 0.55,
-      "learning_rate": 4.9167732012081944e-05,
-      "loss": 0.3313,
+      "epoch": 0.5629437416657657,
+      "grad_norm": 0.16514959931373596,
+      "learning_rate": 4.91117195343221e-05,
+      "loss": 0.4524,
       "step": 15620
     },
     {
-      "epoch": 0.55,
-      "learning_rate": 4.9167002932207156e-05,
-      "loss": 0.2979,
+      "epoch": 0.5631239413269903,
+      "grad_norm": 0.14684879779815674,
+      "learning_rate": 4.911094840837859e-05,
+      "loss": 0.4564,
       "step": 15625
     },
     {
-      "epoch": 0.55,
-      "learning_rate": 4.916627353853976e-05,
-      "loss": 0.3045,
+      "epoch": 0.5633041409882149,
+      "grad_norm": 0.17186830937862396,
+      "learning_rate": 4.911017695392807e-05,
+      "loss": 0.4371,
       "step": 15630
     },
     {
-      "epoch": 0.55,
-      "learning_rate": 4.916554383108923e-05,
-      "loss": 0.3124,
+      "epoch": 0.5634843406494395,
+      "grad_norm": 0.1487584263086319,
+      "learning_rate": 4.910940517098105e-05,
+      "loss": 0.4323,
       "step": 15635
     },
     {
-      "epoch": 0.55,
-      "learning_rate": 4.916481380986504e-05,
-      "loss": 0.3113,
+      "epoch": 0.5636645403106643,
+      "grad_norm": 0.1303669810295105,
+      "learning_rate": 4.910863305954805e-05,
+      "loss": 0.4372,
       "step": 15640
     },
     {
-      "epoch": 0.55,
-      "learning_rate": 4.9164083474876675e-05,
-      "loss": 0.3283,
+      "epoch": 0.5638447399718889,
+      "grad_norm": 0.16420136392116547,
+      "learning_rate": 4.91078606196396e-05,
+      "loss": 0.4775,
       "step": 15645
     },
     {
-      "epoch": 0.55,
-      "learning_rate": 4.91633528261336e-05,
-      "loss": 0.3191,
+      "epoch": 0.5640249396331135,
+      "grad_norm": 0.18261446058750153,
+      "learning_rate": 4.910708785126622e-05,
+      "loss": 0.4135,
       "step": 15650
     },
     {
-      "epoch": 0.55,
-      "learning_rate": 4.9162621863645317e-05,
-      "loss": 0.3268,
+      "epoch": 0.5642051392943381,
+      "grad_norm": 0.1571108102798462,
+      "learning_rate": 4.910631475443843e-05,
+      "loss": 0.4463,
       "step": 15655
     },
     {
-      "epoch": 0.55,
-      "learning_rate": 4.916189058742131e-05,
-      "loss": 0.2921,
+      "epoch": 0.5643853389555628,
+      "grad_norm": 0.18948879837989807,
+      "learning_rate": 4.910554132916677e-05,
+      "loss": 0.4544,
       "step": 15660
     },
     {
-      "epoch": 0.55,
-      "learning_rate": 4.9161158997471086e-05,
-      "loss": 0.3122,
+      "epoch": 0.5645655386167874,
+      "grad_norm": 0.19807395339012146,
+      "learning_rate": 4.910476757546177e-05,
+      "loss": 0.4835,
       "step": 15665
     },
     {
-      "epoch": 0.55,
-      "learning_rate": 4.916042709380413e-05,
-      "loss": 0.3073,
+      "epoch": 0.564745738278012,
+      "grad_norm": 0.1863390952348709,
+      "learning_rate": 4.910399349333399e-05,
+      "loss": 0.4522,
       "step": 15670
     },
     {
-      "epoch": 0.55,
-      "learning_rate": 4.915969487642995e-05,
-      "loss": 0.2936,
+      "epoch": 0.5649259379392366,
+      "grad_norm": 0.17297495901584625,
+      "learning_rate": 4.910321908279396e-05,
+      "loss": 0.4493,
       "step": 15675
     },
     {
-      "epoch": 0.55,
-      "learning_rate": 4.9159108876667794e-05,
-      "loss": 0.3335,
+      "epoch": 0.5651061376004614,
+      "grad_norm": 0.1477319598197937,
+      "learning_rate": 4.910244434385224e-05,
+      "loss": 0.4326,
       "step": 15680
     },
     {
-      "epoch": 0.55,
-      "learning_rate": 4.915837609464457e-05,
-      "loss": 0.3175,
+      "epoch": 0.565286337261686,
+      "grad_norm": 0.18947510421276093,
+      "learning_rate": 4.9101669276519374e-05,
+      "loss": 0.4266,
       "step": 15685
     },
     {
-      "epoch": 0.55,
-      "learning_rate": 4.915764299894076e-05,
-      "loss": 0.2781,
+      "epoch": 0.5654665369229106,
+      "grad_norm": 0.1564638763666153,
+      "learning_rate": 4.910089388080593e-05,
+      "loss": 0.4275,
       "step": 15690
     },
     {
-      "epoch": 0.55,
-      "learning_rate": 4.9156909589565866e-05,
-      "loss": 0.3107,
+      "epoch": 0.5656467365841352,
+      "grad_norm": 0.1595679074525833,
+      "learning_rate": 4.9100118156722485e-05,
+      "loss": 0.4342,
       "step": 15695
     },
     {
-      "epoch": 0.55,
-      "learning_rate": 4.915617586652943e-05,
-      "loss": 0.3128,
+      "epoch": 0.5658269362453598,
+      "grad_norm": 0.16428592801094055,
+      "learning_rate": 4.909934210427959e-05,
+      "loss": 0.4652,
       "step": 15700
     },
     {
-      "epoch": 0.55,
-      "learning_rate": 4.915544182984096e-05,
-      "loss": 0.3059,
+      "epoch": 0.5660071359065845,
+      "grad_norm": 0.15228533744812012,
+      "learning_rate": 4.909856572348782e-05,
+      "loss": 0.3994,
       "step": 15705
     },
     {
-      "epoch": 0.55,
-      "learning_rate": 4.915470747951001e-05,
-      "loss": 0.3112,
+      "epoch": 0.5661873355678091,
+      "grad_norm": 0.18478292226791382,
+      "learning_rate": 4.9097789014357763e-05,
+      "loss": 0.4288,
       "step": 15710
     },
     {
-      "epoch": 0.55,
-      "learning_rate": 4.91539728155461e-05,
-      "loss": 0.3301,
+      "epoch": 0.5663675352290337,
+      "grad_norm": 0.1817883402109146,
+      "learning_rate": 4.9097011976899996e-05,
+      "loss": 0.4205,
       "step": 15715
     },
     {
-      "epoch": 0.55,
-      "learning_rate": 4.915323783795876e-05,
-      "loss": 0.3488,
+      "epoch": 0.5665477348902584,
+      "grad_norm": 0.16065534949302673,
+      "learning_rate": 4.90962346111251e-05,
+      "loss": 0.4548,
       "step": 15720
     },
     {
-      "epoch": 0.55,
-      "learning_rate": 4.915250254675755e-05,
-      "loss": 0.3322,
+      "epoch": 0.5667279345514831,
+      "grad_norm": 0.1838037371635437,
+      "learning_rate": 4.909545691704368e-05,
+      "loss": 0.4084,
       "step": 15725
     },
     {
-      "epoch": 0.55,
-      "learning_rate": 4.9151766941952015e-05,
-      "loss": 0.3113,
+      "epoch": 0.5669081342127077,
+      "grad_norm": 0.21379324793815613,
+      "learning_rate": 4.909467889466632e-05,
+      "loss": 0.4492,
       "step": 15730
     },
     {
-      "epoch": 0.55,
-      "learning_rate": 4.91510310235517e-05,
-      "loss": 0.3019,
+      "epoch": 0.5670883338739323,
+      "grad_norm": 0.18225577473640442,
+      "learning_rate": 4.9093900544003625e-05,
+      "loss": 0.4403,
       "step": 15735
     },
     {
-      "epoch": 0.55,
-      "learning_rate": 4.9150294791566167e-05,
-      "loss": 0.2977,
+      "epoch": 0.5672685335351569,
+      "grad_norm": 0.16789789497852325,
+      "learning_rate": 4.909327762711516e-05,
+      "loss": 0.4592,
       "step": 15740
     },
     {
-      "epoch": 0.55,
-      "learning_rate": 4.9149558246004964e-05,
-      "loss": 0.329,
+      "epoch": 0.5674487331963816,
+      "grad_norm": 0.16580015420913696,
+      "learning_rate": 4.909249868556559e-05,
+      "loss": 0.4639,
       "step": 15745
     },
     {
-      "epoch": 0.55,
-      "learning_rate": 4.914882138687767e-05,
-      "loss": 0.3104,
+      "epoch": 0.5676289328576062,
+      "grad_norm": 0.16387851536273956,
+      "learning_rate": 4.909171941576039e-05,
+      "loss": 0.3932,
       "step": 15750
     },
     {
-      "epoch": 0.55,
-      "learning_rate": 4.914808421419384e-05,
-      "loss": 0.3201,
+      "epoch": 0.5678091325188309,
+      "grad_norm": 0.1637287735939026,
+      "learning_rate": 4.909093981771018e-05,
+      "loss": 0.4378,
       "step": 15755
     },
     {
-      "epoch": 0.55,
-      "learning_rate": 4.914734672796305e-05,
-      "loss": 0.2997,
+      "epoch": 0.5679893321800555,
+      "grad_norm": 0.1793089210987091,
+      "learning_rate": 4.909015989142557e-05,
+      "loss": 0.4265,
       "step": 15760
     },
     {
-      "epoch": 0.55,
-      "learning_rate": 4.914660892819488e-05,
-      "loss": 0.3156,
+      "epoch": 0.5681695318412802,
+      "grad_norm": 0.17664669454097748,
+      "learning_rate": 4.908937963691721e-05,
+      "loss": 0.4682,
       "step": 15765
     },
     {
-      "epoch": 0.55,
-      "learning_rate": 4.914587081489891e-05,
-      "loss": 0.3139,
+      "epoch": 0.5683497315025048,
+      "grad_norm": 0.17455190420150757,
+      "learning_rate": 4.90885990541957e-05,
+      "loss": 0.4658,
       "step": 15770
     },
     {
-      "epoch": 0.56,
-      "learning_rate": 4.914513238808471e-05,
-      "loss": 0.3022,
+      "epoch": 0.5685299311637294,
+      "grad_norm": 0.17027586698532104,
+      "learning_rate": 4.9087818143271703e-05,
+      "loss": 0.4062,
       "step": 15775
     },
     {
-      "epoch": 0.56,
-      "learning_rate": 4.914439364776189e-05,
-      "loss": 0.2884,
+      "epoch": 0.568710130824954,
+      "grad_norm": 0.23216940462589264,
+      "learning_rate": 4.9087036904155844e-05,
+      "loss": 0.4304,
       "step": 15780
     },
     {
-      "epoch": 0.56,
-      "learning_rate": 4.914365459394002e-05,
-      "loss": 0.3486,
+      "epoch": 0.5688903304861787,
+      "grad_norm": 0.14484082162380219,
+      "learning_rate": 4.908625533685878e-05,
+      "loss": 0.4446,
       "step": 15785
     },
     {
-      "epoch": 0.56,
-      "learning_rate": 4.914291522662871e-05,
-      "loss": 0.3116,
+      "epoch": 0.5690705301474033,
+      "grad_norm": 0.18710222840309143,
+      "learning_rate": 4.908547344139114e-05,
+      "loss": 0.4319,
       "step": 15790
     },
     {
-      "epoch": 0.56,
-      "learning_rate": 4.9142175545837555e-05,
-      "loss": 0.3308,
+      "epoch": 0.569250729808628,
+      "grad_norm": 0.16132237017154694,
+      "learning_rate": 4.908469121776358e-05,
+      "loss": 0.4643,
       "step": 15795
     },
     {
-      "epoch": 0.56,
-      "learning_rate": 4.9141435551576165e-05,
-      "loss": 0.319,
+      "epoch": 0.5694309294698526,
+      "grad_norm": 0.16530324518680573,
+      "learning_rate": 4.908390866598678e-05,
+      "loss": 0.4599,
       "step": 15800
     },
     {
-      "epoch": 0.56,
-      "learning_rate": 4.914069524385414e-05,
-      "loss": 0.3092,
+      "epoch": 0.5696111291310773,
+      "grad_norm": 0.16622117161750793,
+      "learning_rate": 4.908312578607138e-05,
+      "loss": 0.4115,
       "step": 15805
     },
     {
-      "epoch": 0.56,
-      "learning_rate": 4.9139954622681094e-05,
-      "loss": 0.3191,
+      "epoch": 0.5697913287923019,
+      "grad_norm": 0.19069482386112213,
+      "learning_rate": 4.9082342578028054e-05,
+      "loss": 0.4046,
       "step": 15810
     },
     {
-      "epoch": 0.56,
-      "learning_rate": 4.9139213688066656e-05,
-      "loss": 0.328,
+      "epoch": 0.5699715284535265,
+      "grad_norm": 0.15032880008220673,
+      "learning_rate": 4.908155904186747e-05,
+      "loss": 0.4386,
       "step": 15815
     },
     {
-      "epoch": 0.56,
-      "learning_rate": 4.913847244002043e-05,
-      "loss": 0.3296,
+      "epoch": 0.5701517281147511,
+      "grad_norm": 0.18905720114707947,
+      "learning_rate": 4.9080775177600316e-05,
+      "loss": 0.4328,
       "step": 15820
     },
     {
-      "epoch": 0.56,
-      "learning_rate": 4.9137730878552044e-05,
-      "loss": 0.3267,
+      "epoch": 0.5703319277759757,
+      "grad_norm": 0.1644611656665802,
+      "learning_rate": 4.907999098523726e-05,
+      "loss": 0.4288,
       "step": 15825
     },
     {
-      "epoch": 0.56,
-      "learning_rate": 4.913698900367114e-05,
-      "loss": 0.3038,
+      "epoch": 0.5705121274372004,
+      "grad_norm": 0.21615582704544067,
+      "learning_rate": 4.9079206464788986e-05,
+      "loss": 0.4575,
       "step": 15830
     },
     {
-      "epoch": 0.56,
-      "learning_rate": 4.9136246815387334e-05,
-      "loss": 0.3019,
+      "epoch": 0.5706923270984251,
+      "grad_norm": 0.15707986056804657,
+      "learning_rate": 4.907842161626618e-05,
+      "loss": 0.4264,
       "step": 15835
     },
     {
-      "epoch": 0.56,
-      "learning_rate": 4.913550431371027e-05,
-      "loss": 0.2831,
+      "epoch": 0.5708725267596497,
+      "grad_norm": 0.22739115357398987,
+      "learning_rate": 4.9077636439679554e-05,
+      "loss": 0.4573,
       "step": 15840
     },
     {
-      "epoch": 0.56,
-      "learning_rate": 4.9134761498649586e-05,
-      "loss": 0.3151,
+      "epoch": 0.5710527264208743,
+      "grad_norm": 0.16067542135715485,
+      "learning_rate": 4.907685093503978e-05,
+      "loss": 0.4617,
       "step": 15845
     },
     {
-      "epoch": 0.56,
-      "learning_rate": 4.913401837021493e-05,
-      "loss": 0.3058,
+      "epoch": 0.571232926082099,
+      "grad_norm": 0.17511925101280212,
+      "learning_rate": 4.907606510235758e-05,
+      "loss": 0.4553,
       "step": 15850
     },
     {
-      "epoch": 0.56,
-      "learning_rate": 4.913327492841595e-05,
-      "loss": 0.2843,
+      "epoch": 0.5714131257433236,
+      "grad_norm": 0.15019887685775757,
+      "learning_rate": 4.9075278941643655e-05,
+      "loss": 0.4247,
       "step": 15855
     },
     {
-      "epoch": 0.56,
-      "learning_rate": 4.9132531173262306e-05,
-      "loss": 0.2963,
+      "epoch": 0.5715933254045482,
+      "grad_norm": 0.2024499922990799,
+      "learning_rate": 4.907449245290872e-05,
+      "loss": 0.4579,
       "step": 15860
     },
     {
-      "epoch": 0.56,
-      "learning_rate": 4.913178710476365e-05,
-      "loss": 0.3378,
+      "epoch": 0.5717735250657728,
+      "grad_norm": 0.1599128544330597,
+      "learning_rate": 4.907370563616347e-05,
+      "loss": 0.4592,
       "step": 15865
     },
     {
-      "epoch": 0.56,
-      "learning_rate": 4.913104272292963e-05,
-      "loss": 0.3084,
+      "epoch": 0.5719537247269975,
+      "grad_norm": 0.18762356042861938,
+      "learning_rate": 4.907291849141865e-05,
+      "loss": 0.4446,
       "step": 15870
     },
     {
-      "epoch": 0.56,
-      "learning_rate": 4.913029802776993e-05,
-      "loss": 0.3009,
+      "epoch": 0.5721339243882222,
+      "grad_norm": 0.20976999402046204,
+      "learning_rate": 4.907213101868498e-05,
+      "loss": 0.4023,
       "step": 15875
     },
     {
-      "epoch": 0.56,
-      "learning_rate": 4.9129553019294214e-05,
-      "loss": 0.3213,
+      "epoch": 0.5723141240494468,
+      "grad_norm": 0.1850195974111557,
+      "learning_rate": 4.9071343217973185e-05,
+      "loss": 0.3954,
       "step": 15880
     },
     {
-      "epoch": 0.56,
-      "learning_rate": 4.9128807697512146e-05,
-      "loss": 0.3199,
+      "epoch": 0.5724943237106714,
+      "grad_norm": 0.1581835001707077,
+      "learning_rate": 4.9070555089294004e-05,
+      "loss": 0.4613,
       "step": 15885
     },
     {
-      "epoch": 0.56,
-      "learning_rate": 4.912806206243342e-05,
-      "loss": 0.3262,
+      "epoch": 0.5726745233718961,
+      "grad_norm": 0.13934732973575592,
+      "learning_rate": 4.906976663265816e-05,
+      "loss": 0.3904,
       "step": 15890
     },
     {
-      "epoch": 0.56,
-      "learning_rate": 4.9127316114067715e-05,
-      "loss": 0.3089,
+      "epoch": 0.5728547230331207,
+      "grad_norm": 0.18741364777088165,
+      "learning_rate": 4.906897784807641e-05,
+      "loss": 0.4099,
       "step": 15895
     },
     {
-      "epoch": 0.56,
-      "learning_rate": 4.91265698524247e-05,
-      "loss": 0.3164,
+      "epoch": 0.5730349226943453,
+      "grad_norm": 0.24076466262340546,
+      "learning_rate": 4.9068188735559494e-05,
+      "loss": 0.4394,
       "step": 15900
     },
     {
-      "epoch": 0.56,
-      "learning_rate": 4.9125823277514075e-05,
-      "loss": 0.322,
+      "epoch": 0.5732151223555699,
+      "grad_norm": 0.14442354440689087,
+      "learning_rate": 4.9067399295118166e-05,
+      "loss": 0.389,
       "step": 15905
     },
     {
-      "epoch": 0.56,
-      "learning_rate": 4.9125076389345534e-05,
-      "loss": 0.3247,
+      "epoch": 0.5733953220167947,
+      "grad_norm": 0.15537168085575104,
+      "learning_rate": 4.906660952676318e-05,
+      "loss": 0.4864,
       "step": 15910
     },
     {
-      "epoch": 0.56,
-      "learning_rate": 4.912432918792878e-05,
-      "loss": 0.2992,
+      "epoch": 0.5735755216780193,
+      "grad_norm": 0.1750095635652542,
+      "learning_rate": 4.906581943050531e-05,
+      "loss": 0.453,
       "step": 15915
     },
     {
-      "epoch": 0.56,
-      "learning_rate": 4.912358167327351e-05,
-      "loss": 0.3167,
+      "epoch": 0.5737557213392439,
+      "grad_norm": 0.13674676418304443,
+      "learning_rate": 4.90650290063553e-05,
+      "loss": 0.4463,
       "step": 15920
     },
     {
-      "epoch": 0.56,
-      "learning_rate": 4.9122833845389426e-05,
-      "loss": 0.3208,
+      "epoch": 0.5739359210004685,
+      "grad_norm": 0.21467788517475128,
+      "learning_rate": 4.9064238254323934e-05,
+      "loss": 0.4174,
       "step": 15925
     },
     {
-      "epoch": 0.56,
-      "learning_rate": 4.9122085704286246e-05,
-      "loss": 0.2921,
+      "epoch": 0.5741161206616932,
+      "grad_norm": 0.19165554642677307,
+      "learning_rate": 4.906344717442198e-05,
+      "loss": 0.4429,
       "step": 15930
     },
     {
-      "epoch": 0.56,
-      "learning_rate": 4.9121337249973675e-05,
-      "loss": 0.3092,
+      "epoch": 0.5742963203229178,
+      "grad_norm": 0.183024063706398,
+      "learning_rate": 4.906265576666022e-05,
+      "loss": 0.4571,
       "step": 15935
     },
     {
-      "epoch": 0.56,
-      "learning_rate": 4.912058848246144e-05,
-      "loss": 0.3083,
+      "epoch": 0.5744765199841424,
+      "grad_norm": 0.18742796778678894,
+      "learning_rate": 4.906186403104942e-05,
+      "loss": 0.4394,
       "step": 15940
     },
     {
-      "epoch": 0.56,
-      "learning_rate": 4.911983940175926e-05,
-      "loss": 0.2973,
+      "epoch": 0.574656719645367,
+      "grad_norm": 0.17034278810024261,
+      "learning_rate": 4.9061071967600394e-05,
+      "loss": 0.4422,
       "step": 15945
     },
     {
-      "epoch": 0.56,
-      "learning_rate": 4.9119090007876866e-05,
-      "loss": 0.326,
+      "epoch": 0.5748369193065918,
+      "grad_norm": 0.19667524099349976,
+      "learning_rate": 4.906027957632392e-05,
+      "loss": 0.4346,
       "step": 15950
     },
     {
-      "epoch": 0.56,
-      "learning_rate": 4.911834030082397e-05,
-      "loss": 0.3251,
+      "epoch": 0.5750171189678164,
+      "grad_norm": 0.16463510692119598,
+      "learning_rate": 4.9059486857230785e-05,
+      "loss": 0.4951,
       "step": 15955
     },
     {
-      "epoch": 0.56,
-      "learning_rate": 4.911759028061033e-05,
-      "loss": 0.3269,
+      "epoch": 0.575197318629041,
+      "grad_norm": 0.1695079207420349,
+      "learning_rate": 4.9058693810331815e-05,
+      "loss": 0.4675,
       "step": 15960
     },
     {
-      "epoch": 0.56,
-      "learning_rate": 4.9116839947245677e-05,
-      "loss": 0.3348,
+      "epoch": 0.5753775182902656,
+      "grad_norm": 0.13872882723808289,
+      "learning_rate": 4.905790043563779e-05,
+      "loss": 0.4046,
       "step": 15965
     },
     {
-      "epoch": 0.56,
-      "learning_rate": 4.9116089300739745e-05,
-      "loss": 0.3269,
+      "epoch": 0.5755577179514902,
+      "grad_norm": 0.196579709649086,
+      "learning_rate": 4.905710673315953e-05,
+      "loss": 0.452,
       "step": 15970
     },
     {
-      "epoch": 0.56,
-      "learning_rate": 4.9115338341102287e-05,
-      "loss": 0.306,
+      "epoch": 0.5757379176127149,
+      "grad_norm": 0.195785254240036,
+      "learning_rate": 4.9056312702907844e-05,
+      "loss": 0.4594,
       "step": 15975
     },
     {
-      "epoch": 0.56,
-      "learning_rate": 4.911458706834305e-05,
-      "loss": 0.2972,
+      "epoch": 0.5759181172739395,
+      "grad_norm": 0.1784789115190506,
+      "learning_rate": 4.905551834489356e-05,
+      "loss": 0.4217,
       "step": 15980
     },
     {
-      "epoch": 0.56,
-      "learning_rate": 4.9113835482471796e-05,
-      "loss": 0.3247,
+      "epoch": 0.5760983169351641,
+      "grad_norm": 0.1556771695613861,
+      "learning_rate": 4.9054723659127496e-05,
+      "loss": 0.4264,
       "step": 15985
     },
     {
-      "epoch": 0.56,
-      "learning_rate": 4.911308358349828e-05,
-      "loss": 0.336,
+      "epoch": 0.5762785165963888,
+      "grad_norm": 0.1567842811346054,
+      "learning_rate": 4.9053928645620484e-05,
+      "loss": 0.449,
       "step": 15990
     },
     {
-      "epoch": 0.56,
-      "learning_rate": 4.911233137143226e-05,
-      "loss": 0.3204,
+      "epoch": 0.5764587162576135,
+      "grad_norm": 0.19066265225410461,
+      "learning_rate": 4.9053133304383346e-05,
+      "loss": 0.4512,
       "step": 15995
     },
     {
-      "epoch": 0.56,
-      "learning_rate": 4.911157884628352e-05,
-      "loss": 0.3203,
+      "epoch": 0.5766389159188381,
+      "grad_norm": 0.17999373376369476,
+      "learning_rate": 4.9052337635426925e-05,
+      "loss": 0.4467,
       "step": 16000
     },
     {
-      "epoch": 0.56,
-      "eval_loss": 0.30903559923171997,
-      "eval_runtime": 10.5428,
-      "eval_samples_per_second": 9.485,
-      "eval_steps_per_second": 9.485,
+      "epoch": 0.5766389159188381,
+      "eval_loss": 0.4678036570549011,
+      "eval_runtime": 3.5673,
+      "eval_samples_per_second": 28.032,
+      "eval_steps_per_second": 7.008,
       "step": 16000
     },
     {
-      "epoch": 0.56,
-      "learning_rate": 4.9110826008061804e-05,
-      "loss": 0.2991,
+      "epoch": 0.5768191155800627,
+      "grad_norm": 0.21432943642139435,
+      "learning_rate": 4.9051541638762055e-05,
+      "loss": 0.4813,
       "step": 16005
     },
     {
-      "epoch": 0.56,
-      "learning_rate": 4.911007285677691e-05,
-      "loss": 0.3346,
+      "epoch": 0.5769993152412873,
+      "grad_norm": 0.1877758800983429,
+      "learning_rate": 4.905074531439959e-05,
+      "loss": 0.4132,
       "step": 16010
     },
     {
-      "epoch": 0.56,
-      "learning_rate": 4.910931939243861e-05,
-      "loss": 0.3237,
+      "epoch": 0.577179514902512,
+      "grad_norm": 0.1863996535539627,
+      "learning_rate": 4.904994866235038e-05,
+      "loss": 0.4336,
       "step": 16015
     },
     {
-      "epoch": 0.56,
-      "learning_rate": 4.910856561505668e-05,
-      "loss": 0.2991,
+      "epoch": 0.5773597145637366,
+      "grad_norm": 0.16214637458324432,
+      "learning_rate": 4.904915168262527e-05,
+      "loss": 0.4463,
       "step": 16020
     },
     {
-      "epoch": 0.56,
-      "learning_rate": 4.9107811524640926e-05,
-      "loss": 0.3116,
+      "epoch": 0.5775399142249612,
+      "grad_norm": 0.17068390548229218,
+      "learning_rate": 4.9048354375235125e-05,
+      "loss": 0.4509,
       "step": 16025
     },
     {
-      "epoch": 0.56,
-      "learning_rate": 4.9107057121201116e-05,
-      "loss": 0.3334,
+      "epoch": 0.5777201138861859,
+      "grad_norm": 0.1428307294845581,
+      "learning_rate": 4.9047556740190814e-05,
+      "loss": 0.4417,
       "step": 16030
     },
     {
-      "epoch": 0.56,
-      "learning_rate": 4.9106302404747064e-05,
-      "loss": 0.2986,
+      "epoch": 0.5779003135474106,
+      "grad_norm": 0.19168958067893982,
+      "learning_rate": 4.90467587775032e-05,
+      "loss": 0.4479,
       "step": 16035
     },
     {
-      "epoch": 0.56,
-      "learning_rate": 4.9105547375288553e-05,
-      "loss": 0.3229,
+      "epoch": 0.5780805132086352,
+      "grad_norm": 0.15157431364059448,
+      "learning_rate": 4.9045960487183144e-05,
+      "loss": 0.4559,
       "step": 16040
     },
     {
-      "epoch": 0.56,
-      "learning_rate": 4.91047920328354e-05,
-      "loss": 0.3129,
+      "epoch": 0.5782607128698598,
+      "grad_norm": 0.17288236320018768,
+      "learning_rate": 4.904516186924154e-05,
+      "loss": 0.4378,
       "step": 16045
     },
     {
-      "epoch": 0.56,
-      "learning_rate": 4.910403637739741e-05,
-      "loss": 0.337,
+      "epoch": 0.5784409125310844,
+      "grad_norm": 0.15723341703414917,
+      "learning_rate": 4.904436292368925e-05,
+      "loss": 0.4208,
       "step": 16050
     },
     {
-      "epoch": 0.56,
-      "learning_rate": 4.910328040898439e-05,
-      "loss": 0.3069,
+      "epoch": 0.578621112192309,
+      "grad_norm": 0.13947317004203796,
+      "learning_rate": 4.9043563650537185e-05,
+      "loss": 0.445,
       "step": 16055
     },
     {
-      "epoch": 0.57,
-      "learning_rate": 4.9102524127606156e-05,
-      "loss": 0.3133,
+      "epoch": 0.5788013118535337,
+      "grad_norm": 0.18408885598182678,
+      "learning_rate": 4.9042764049796205e-05,
+      "loss": 0.4146,
       "step": 16060
     },
     {
-      "epoch": 0.57,
-      "learning_rate": 4.910176753327254e-05,
-      "loss": 0.3323,
+      "epoch": 0.5789815115147584,
+      "grad_norm": 0.17368938028812408,
+      "learning_rate": 4.904196412147723e-05,
+      "loss": 0.4689,
       "step": 16065
     },
     {
-      "epoch": 0.57,
-      "learning_rate": 4.910101062599334e-05,
-      "loss": 0.3099,
+      "epoch": 0.579161711175983,
+      "grad_norm": 0.14086973667144775,
+      "learning_rate": 4.904116386559115e-05,
+      "loss": 0.4167,
       "step": 16070
     },
     {
-      "epoch": 0.57,
-      "learning_rate": 4.910025340577841e-05,
-      "loss": 0.2945,
+      "epoch": 0.5793419108372077,
+      "grad_norm": 0.15786948800086975,
+      "learning_rate": 4.9040363282148854e-05,
+      "loss": 0.4391,
       "step": 16075
     },
     {
-      "epoch": 0.57,
-      "learning_rate": 4.9099495872637574e-05,
-      "loss": 0.3001,
+      "epoch": 0.5795221104984323,
+      "grad_norm": 0.15535566210746765,
+      "learning_rate": 4.903956237116127e-05,
+      "loss": 0.4552,
       "step": 16080
     },
     {
-      "epoch": 0.57,
-      "learning_rate": 4.909873802658067e-05,
-      "loss": 0.3157,
+      "epoch": 0.5797023101596569,
+      "grad_norm": 0.3356935679912567,
+      "learning_rate": 4.9038761132639304e-05,
+      "loss": 0.4424,
       "step": 16085
     },
     {
-      "epoch": 0.57,
-      "learning_rate": 4.909797986761753e-05,
-      "loss": 0.3145,
+      "epoch": 0.5798825098208815,
+      "grad_norm": 0.1919674128293991,
+      "learning_rate": 4.903795956659387e-05,
+      "loss": 0.4388,
       "step": 16090
     },
     {
-      "epoch": 0.57,
-      "learning_rate": 4.9097221395758e-05,
-      "loss": 0.3083,
+      "epoch": 0.5800627094821061,
+      "grad_norm": 0.18245883285999298,
+      "learning_rate": 4.9037157673035894e-05,
+      "loss": 0.452,
       "step": 16095
     },
     {
-      "epoch": 0.57,
-      "learning_rate": 4.909646261101193e-05,
-      "loss": 0.3221,
+      "epoch": 0.5802429091433308,
+      "grad_norm": 0.18774840235710144,
+      "learning_rate": 4.903635545197629e-05,
+      "loss": 0.4434,
       "step": 16100
     },
     {
-      "epoch": 0.57,
-      "learning_rate": 4.909570351338917e-05,
-      "loss": 0.3278,
+      "epoch": 0.5804231088045555,
+      "grad_norm": 0.1595427542924881,
+      "learning_rate": 4.9035552903426006e-05,
+      "loss": 0.4435,
       "step": 16105
     },
     {
-      "epoch": 0.57,
-      "learning_rate": 4.9094944102899585e-05,
-      "loss": 0.322,
+      "epoch": 0.5806033084657801,
+      "grad_norm": 0.1629880964756012,
+      "learning_rate": 4.903475002739596e-05,
+      "loss": 0.4895,
       "step": 16110
     },
     {
-      "epoch": 0.57,
-      "learning_rate": 4.909418437955303e-05,
-      "loss": 0.3157,
+      "epoch": 0.5807835081270047,
+      "grad_norm": 0.18076345324516296,
+      "learning_rate": 4.903394682389711e-05,
+      "loss": 0.4733,
       "step": 16115
     },
     {
-      "epoch": 0.57,
-      "learning_rate": 4.909342434335937e-05,
-      "loss": 0.3012,
+      "epoch": 0.5809637077882294,
+      "grad_norm": 0.16395287215709686,
+      "learning_rate": 4.9033143292940376e-05,
+      "loss": 0.4474,
       "step": 16120
     },
     {
-      "epoch": 0.57,
-      "learning_rate": 4.909266399432847e-05,
-      "loss": 0.3021,
+      "epoch": 0.581143907449454,
+      "grad_norm": 0.21613916754722595,
+      "learning_rate": 4.903233943453672e-05,
+      "loss": 0.4297,
       "step": 16125
     },
     {
-      "epoch": 0.57,
-      "learning_rate": 4.9091903332470215e-05,
-      "loss": 0.3079,
+      "epoch": 0.5813241071106786,
+      "grad_norm": 0.15141765773296356,
+      "learning_rate": 4.9031535248697095e-05,
+      "loss": 0.4635,
       "step": 16130
     },
     {
-      "epoch": 0.57,
-      "learning_rate": 4.909114235779446e-05,
-      "loss": 0.3236,
+      "epoch": 0.5815043067719032,
+      "grad_norm": 0.18841727077960968,
+      "learning_rate": 4.9030730735432453e-05,
+      "loss": 0.4198,
       "step": 16135
     },
     {
-      "epoch": 0.57,
-      "learning_rate": 4.909038107031111e-05,
-      "loss": 0.3187,
+      "epoch": 0.5816845064331279,
+      "grad_norm": 0.1524806171655655,
+      "learning_rate": 4.902992589475376e-05,
+      "loss": 0.431,
       "step": 16140
     },
     {
-      "epoch": 0.57,
-      "learning_rate": 4.9089619470030025e-05,
-      "loss": 0.3149,
+      "epoch": 0.5818647060943526,
+      "grad_norm": 0.21377435326576233,
+      "learning_rate": 4.9029120726671974e-05,
+      "loss": 0.4369,
       "step": 16145
     },
     {
-      "epoch": 0.57,
-      "learning_rate": 4.908885755696111e-05,
-      "loss": 0.2911,
+      "epoch": 0.5820449057555772,
+      "grad_norm": 0.17817193269729614,
+      "learning_rate": 4.902831523119808e-05,
+      "loss": 0.4394,
       "step": 16150
     },
     {
-      "epoch": 0.57,
-      "learning_rate": 4.908809533111427e-05,
-      "loss": 0.3365,
+      "epoch": 0.5822251054168018,
+      "grad_norm": 0.14097528159618378,
+      "learning_rate": 4.902750940834303e-05,
+      "loss": 0.4648,
       "step": 16155
     },
     {
-      "epoch": 0.57,
-      "learning_rate": 4.908733279249937e-05,
-      "loss": 0.272,
+      "epoch": 0.5824053050780265,
+      "grad_norm": 0.1693866103887558,
+      "learning_rate": 4.9026703258117825e-05,
+      "loss": 0.4576,
       "step": 16160
     },
     {
-      "epoch": 0.57,
-      "learning_rate": 4.9086569941126335e-05,
-      "loss": 0.3254,
+      "epoch": 0.5825855047392511,
+      "grad_norm": 0.17576813697814941,
+      "learning_rate": 4.902589678053343e-05,
+      "loss": 0.4165,
       "step": 16165
     },
     {
-      "epoch": 0.57,
-      "learning_rate": 4.9085806777005054e-05,
-      "loss": 0.2899,
+      "epoch": 0.5827657044004757,
+      "grad_norm": 0.16277313232421875,
+      "learning_rate": 4.9025089975600855e-05,
+      "loss": 0.4695,
       "step": 16170
     },
     {
-      "epoch": 0.57,
-      "learning_rate": 4.9085043300145453e-05,
-      "loss": 0.3273,
+      "epoch": 0.5829459040617003,
+      "grad_norm": 0.17352886497974396,
+      "learning_rate": 4.902428284333107e-05,
+      "loss": 0.4715,
       "step": 16175
     },
     {
-      "epoch": 0.57,
-      "learning_rate": 4.908427951055744e-05,
-      "loss": 0.2825,
+      "epoch": 0.583126103722925,
+      "grad_norm": 0.16284751892089844,
+      "learning_rate": 4.902347538373509e-05,
+      "loss": 0.4683,
       "step": 16180
     },
     {
-      "epoch": 0.57,
-      "learning_rate": 4.908351540825093e-05,
-      "loss": 0.304,
+      "epoch": 0.5833063033841497,
+      "grad_norm": 0.17194764316082,
+      "learning_rate": 4.9022667596823904e-05,
+      "loss": 0.4208,
       "step": 16185
     },
     {
-      "epoch": 0.57,
-      "learning_rate": 4.9082750993235835e-05,
-      "loss": 0.294,
+      "epoch": 0.5834865030453743,
+      "grad_norm": 0.1685447096824646,
+      "learning_rate": 4.902185948260853e-05,
+      "loss": 0.4543,
       "step": 16190
     },
     {
-      "epoch": 0.57,
-      "learning_rate": 4.90819862655221e-05,
-      "loss": 0.3118,
+      "epoch": 0.5836667027065989,
+      "grad_norm": 0.18297097086906433,
+      "learning_rate": 4.9021051041099966e-05,
+      "loss": 0.4427,
       "step": 16195
     },
     {
-      "epoch": 0.57,
-      "learning_rate": 4.908122122511963e-05,
-      "loss": 0.338,
+      "epoch": 0.5838469023678236,
+      "grad_norm": 0.20284834504127502,
+      "learning_rate": 4.902024227230924e-05,
+      "loss": 0.4038,
       "step": 16200
     },
     {
-      "epoch": 0.57,
-      "learning_rate": 4.9080455872038385e-05,
-      "loss": 0.3199,
+      "epoch": 0.5840271020290482,
+      "grad_norm": 0.1463736742734909,
+      "learning_rate": 4.9019433176247353e-05,
+      "loss": 0.4688,
       "step": 16205
     },
     {
-      "epoch": 0.57,
-      "learning_rate": 4.9079690206288284e-05,
-      "loss": 0.3157,
+      "epoch": 0.5842073016902728,
+      "grad_norm": 0.14948871731758118,
+      "learning_rate": 4.901862375292534e-05,
+      "loss": 0.4649,
       "step": 16210
     },
     {
-      "epoch": 0.57,
-      "learning_rate": 4.9078924227879274e-05,
-      "loss": 0.2938,
+      "epoch": 0.5843875013514974,
+      "grad_norm": 0.16728514432907104,
+      "learning_rate": 4.901781400235423e-05,
+      "loss": 0.4369,
       "step": 16215
     },
     {
-      "epoch": 0.57,
-      "learning_rate": 4.907815793682131e-05,
-      "loss": 0.3026,
+      "epoch": 0.584567701012722,
+      "grad_norm": 0.1717401146888733,
+      "learning_rate": 4.901700392454506e-05,
+      "loss": 0.4683,
       "step": 16220
     },
     {
-      "epoch": 0.57,
-      "learning_rate": 4.9077391333124323e-05,
-      "loss": 0.2991,
+      "epoch": 0.5847479006739468,
+      "grad_norm": 0.18547184765338898,
+      "learning_rate": 4.9016193519508855e-05,
+      "loss": 0.4614,
       "step": 16225
     },
     {
-      "epoch": 0.57,
-      "learning_rate": 4.9076624416798285e-05,
-      "loss": 0.3261,
+      "epoch": 0.5849281003351714,
+      "grad_norm": 0.1905558854341507,
+      "learning_rate": 4.901538278725666e-05,
+      "loss": 0.4074,
       "step": 16230
     },
     {
-      "epoch": 0.57,
-      "learning_rate": 4.907585718785314e-05,
-      "loss": 0.3206,
+      "epoch": 0.585108299996396,
+      "grad_norm": 0.169610396027565,
+      "learning_rate": 4.901457172779953e-05,
+      "loss": 0.4232,
       "step": 16235
     },
     {
-      "epoch": 0.57,
-      "learning_rate": 4.907508964629886e-05,
-      "loss": 0.3247,
+      "epoch": 0.5852884996576206,
+      "grad_norm": 0.14899376034736633,
+      "learning_rate": 4.9013760341148515e-05,
+      "loss": 0.4495,
       "step": 16240
     },
     {
-      "epoch": 0.57,
-      "learning_rate": 4.9074321792145405e-05,
-      "loss": 0.2934,
+      "epoch": 0.5854686993188453,
+      "grad_norm": 0.1737765222787857,
+      "learning_rate": 4.901294862731466e-05,
+      "loss": 0.4513,
       "step": 16245
     },
     {
-      "epoch": 0.57,
-      "learning_rate": 4.907355362540275e-05,
-      "loss": 0.3117,
+      "epoch": 0.5856488989800699,
+      "grad_norm": 0.18790043890476227,
+      "learning_rate": 4.901213658630902e-05,
+      "loss": 0.4812,
       "step": 16250
     },
     {
-      "epoch": 0.57,
-      "learning_rate": 4.907278514608086e-05,
-      "loss": 0.3057,
+      "epoch": 0.5858290986412945,
+      "grad_norm": 0.13040530681610107,
+      "learning_rate": 4.901132421814267e-05,
+      "loss": 0.4453,
       "step": 16255
     },
     {
-      "epoch": 0.57,
-      "learning_rate": 4.9072016354189725e-05,
-      "loss": 0.3288,
+      "epoch": 0.5860092983025192,
+      "grad_norm": 0.1362471878528595,
+      "learning_rate": 4.901051152282669e-05,
+      "loss": 0.433,
       "step": 16260
     },
     {
-      "epoch": 0.57,
-      "learning_rate": 4.9071247249739324e-05,
-      "loss": 0.3158,
+      "epoch": 0.5861894979637439,
+      "grad_norm": 0.1804387867450714,
+      "learning_rate": 4.9009698500372124e-05,
+      "loss": 0.4489,
       "step": 16265
     },
     {
-      "epoch": 0.57,
-      "learning_rate": 4.907047783273963e-05,
-      "loss": 0.3155,
+      "epoch": 0.5863696976249685,
+      "grad_norm": 0.1898290514945984,
+      "learning_rate": 4.9008885150790076e-05,
+      "loss": 0.4161,
       "step": 16270
     },
     {
-      "epoch": 0.57,
-      "learning_rate": 4.9069708103200655e-05,
-      "loss": 0.3297,
+      "epoch": 0.5865498972861931,
+      "grad_norm": 0.16169433295726776,
+      "learning_rate": 4.900807147409161e-05,
+      "loss": 0.4563,
       "step": 16275
     },
     {
-      "epoch": 0.57,
-      "learning_rate": 4.906893806113238e-05,
-      "loss": 0.3117,
+      "epoch": 0.5867300969474177,
+      "grad_norm": 0.13549362123012543,
+      "learning_rate": 4.900725747028782e-05,
+      "loss": 0.4627,
       "step": 16280
     },
     {
-      "epoch": 0.57,
-      "learning_rate": 4.906816770654481e-05,
-      "loss": 0.3229,
+      "epoch": 0.5869102966086424,
+      "grad_norm": 0.15982681512832642,
+      "learning_rate": 4.900644313938979e-05,
+      "loss": 0.4483,
       "step": 16285
     },
     {
-      "epoch": 0.57,
-      "learning_rate": 4.906739703944794e-05,
-      "loss": 0.2878,
+      "epoch": 0.587090496269867,
+      "grad_norm": 0.14519517123699188,
+      "learning_rate": 4.900562848140863e-05,
+      "loss": 0.4483,
       "step": 16290
     },
     {
-      "epoch": 0.57,
-      "learning_rate": 4.9066626059851775e-05,
-      "loss": 0.3256,
+      "epoch": 0.5872706959310916,
+      "grad_norm": 0.21161200106143951,
+      "learning_rate": 4.900481349635542e-05,
+      "loss": 0.4173,
       "step": 16295
     },
     {
-      "epoch": 0.57,
-      "learning_rate": 4.906585476776634e-05,
-      "loss": 0.31,
+      "epoch": 0.5874508955923163,
+      "grad_norm": 0.14361968636512756,
+      "learning_rate": 4.9003998184241275e-05,
+      "loss": 0.477,
       "step": 16300
     },
     {
-      "epoch": 0.57,
-      "learning_rate": 4.906508316320163e-05,
-      "loss": 0.3165,
+      "epoch": 0.587631095253541,
+      "grad_norm": 0.16247104108333588,
+      "learning_rate": 4.9003182545077305e-05,
+      "loss": 0.4436,
       "step": 16305
     },
     {
-      "epoch": 0.57,
-      "learning_rate": 4.9064311246167684e-05,
-      "loss": 0.3078,
+      "epoch": 0.5878112949147656,
+      "grad_norm": 0.21194829046726227,
+      "learning_rate": 4.9002366578874626e-05,
+      "loss": 0.4429,
       "step": 16310
     },
     {
-      "epoch": 0.57,
-      "learning_rate": 4.9063539016674517e-05,
-      "loss": 0.3042,
+      "epoch": 0.5879914945759902,
+      "grad_norm": 0.15204815566539764,
+      "learning_rate": 4.9001550285644336e-05,
+      "loss": 0.4554,
       "step": 16315
     },
     {
-      "epoch": 0.57,
-      "learning_rate": 4.9062766474732146e-05,
-      "loss": 0.2994,
+      "epoch": 0.5881716942372148,
+      "grad_norm": 0.16609960794448853,
+      "learning_rate": 4.900073366539758e-05,
+      "loss": 0.4564,
       "step": 16320
     },
     {
-      "epoch": 0.57,
-      "learning_rate": 4.906199362035062e-05,
-      "loss": 0.293,
+      "epoch": 0.5883518938984394,
+      "grad_norm": 0.1630159169435501,
+      "learning_rate": 4.8999916718145475e-05,
+      "loss": 0.4304,
       "step": 16325
     },
     {
-      "epoch": 0.57,
-      "learning_rate": 4.9061220453539955e-05,
-      "loss": 0.3131,
+      "epoch": 0.5885320935596641,
+      "grad_norm": 0.1939852088689804,
+      "learning_rate": 4.899909944389914e-05,
+      "loss": 0.4236,
       "step": 16330
     },
     {
-      "epoch": 0.57,
-      "learning_rate": 4.9060446974310206e-05,
-      "loss": 0.3289,
+      "epoch": 0.5887122932208887,
+      "grad_norm": 0.19106720387935638,
+      "learning_rate": 4.8998281842669734e-05,
+      "loss": 0.4602,
       "step": 16335
     },
     {
-      "epoch": 0.57,
-      "learning_rate": 4.90596731826714e-05,
-      "loss": 0.2934,
+      "epoch": 0.5888924928821134,
+      "grad_norm": 0.19278432428836823,
+      "learning_rate": 4.8997463914468376e-05,
+      "loss": 0.4576,
       "step": 16340
     },
     {
-      "epoch": 0.58,
-      "learning_rate": 4.905889907863361e-05,
-      "loss": 0.302,
+      "epoch": 0.589072692543338,
+      "grad_norm": 0.20767638087272644,
+      "learning_rate": 4.899664565930623e-05,
+      "loss": 0.484,
       "step": 16345
     },
     {
-      "epoch": 0.58,
-      "learning_rate": 4.9058124662206864e-05,
-      "loss": 0.2956,
+      "epoch": 0.5892528922045627,
+      "grad_norm": 0.1724158674478531,
+      "learning_rate": 4.8995827077194425e-05,
+      "loss": 0.4261,
       "step": 16350
     },
     {
-      "epoch": 0.58,
-      "learning_rate": 4.9057349933401213e-05,
-      "loss": 0.3129,
+      "epoch": 0.5894330918657873,
+      "grad_norm": 0.1584085077047348,
+      "learning_rate": 4.899500816814412e-05,
+      "loss": 0.4627,
       "step": 16355
     },
     {
-      "epoch": 0.58,
-      "learning_rate": 4.905657489222674e-05,
-      "loss": 0.3028,
+      "epoch": 0.5896132915270119,
+      "grad_norm": 0.159664586186409,
+      "learning_rate": 4.8994188932166473e-05,
+      "loss": 0.4483,
       "step": 16360
     },
     {
-      "epoch": 0.58,
-      "learning_rate": 4.905579953869349e-05,
-      "loss": 0.2844,
+      "epoch": 0.5897934911882365,
+      "grad_norm": 0.14001238346099854,
+      "learning_rate": 4.8993369369272646e-05,
+      "loss": 0.4758,
       "step": 16365
     },
     {
-      "epoch": 0.58,
-      "learning_rate": 4.905502387281154e-05,
-      "loss": 0.2775,
+      "epoch": 0.5899736908494612,
+      "grad_norm": 0.1366659700870514,
+      "learning_rate": 4.899254947947382e-05,
+      "loss": 0.441,
       "step": 16370
     },
     {
-      "epoch": 0.58,
-      "learning_rate": 4.905424789459096e-05,
-      "loss": 0.3449,
+      "epoch": 0.5901538905106858,
+      "grad_norm": 0.13994751870632172,
+      "learning_rate": 4.899172926278113e-05,
+      "loss": 0.3618,
       "step": 16375
     },
     {
-      "epoch": 0.58,
-      "learning_rate": 4.9053471604041815e-05,
-      "loss": 0.3173,
+      "epoch": 0.5903340901719105,
+      "grad_norm": 0.18011939525604248,
+      "learning_rate": 4.899090871920579e-05,
+      "loss": 0.444,
       "step": 16380
     },
     {
-      "epoch": 0.58,
-      "learning_rate": 4.9052695001174196e-05,
-      "loss": 0.286,
+      "epoch": 0.5905142898331351,
+      "grad_norm": 0.18300089240074158,
+      "learning_rate": 4.899008784875896e-05,
+      "loss": 0.4418,
       "step": 16385
     },
     {
-      "epoch": 0.58,
-      "learning_rate": 4.9051918085998184e-05,
-      "loss": 0.3029,
+      "epoch": 0.5906944894943598,
+      "grad_norm": 0.16376063227653503,
+      "learning_rate": 4.898926665145183e-05,
+      "loss": 0.4708,
       "step": 16390
     },
     {
-      "epoch": 0.58,
-      "learning_rate": 4.905114085852387e-05,
-      "loss": 0.2968,
+      "epoch": 0.5908746891555844,
+      "grad_norm": 0.20874595642089844,
+      "learning_rate": 4.898844512729558e-05,
+      "loss": 0.4593,
       "step": 16395
     },
     {
-      "epoch": 0.58,
-      "learning_rate": 4.905036331876134e-05,
-      "loss": 0.3236,
+      "epoch": 0.591054888816809,
+      "grad_norm": 0.1489773839712143,
+      "learning_rate": 4.898762327630142e-05,
+      "loss": 0.4314,
       "step": 16400
     },
     {
-      "epoch": 0.58,
-      "learning_rate": 4.9049585466720695e-05,
-      "loss": 0.2975,
+      "epoch": 0.5912350884780336,
+      "grad_norm": 0.1538877636194229,
+      "learning_rate": 4.898680109848053e-05,
+      "loss": 0.4327,
       "step": 16405
     },
     {
-      "epoch": 0.58,
-      "learning_rate": 4.904880730241202e-05,
-      "loss": 0.3245,
+      "epoch": 0.5914152881392583,
+      "grad_norm": 0.16070400178432465,
+      "learning_rate": 4.8985978593844115e-05,
+      "loss": 0.4837,
       "step": 16410
     },
     {
-      "epoch": 0.58,
-      "learning_rate": 4.904802882584545e-05,
-      "loss": 0.3016,
+      "epoch": 0.591595487800483,
+      "grad_norm": 0.15595030784606934,
+      "learning_rate": 4.898515576240339e-05,
+      "loss": 0.4543,
       "step": 16415
     },
     {
-      "epoch": 0.58,
-      "learning_rate": 4.9047250037031064e-05,
-      "loss": 0.3133,
+      "epoch": 0.5917756874617076,
+      "grad_norm": 0.1852567046880722,
+      "learning_rate": 4.898433260416956e-05,
+      "loss": 0.4185,
       "step": 16420
     },
     {
-      "epoch": 0.58,
-      "learning_rate": 4.9046470935978986e-05,
-      "loss": 0.3239,
+      "epoch": 0.5919558871229322,
+      "grad_norm": 0.18064150214195251,
+      "learning_rate": 4.898350911915385e-05,
+      "loss": 0.4524,
       "step": 16425
     },
     {
-      "epoch": 0.58,
-      "learning_rate": 4.904569152269932e-05,
-      "loss": 0.3015,
+      "epoch": 0.5921360867841569,
+      "grad_norm": 0.1406656950712204,
+      "learning_rate": 4.898268530736746e-05,
+      "loss": 0.4357,
       "step": 16430
     },
     {
-      "epoch": 0.58,
-      "learning_rate": 4.904491179720221e-05,
-      "loss": 0.3449,
+      "epoch": 0.5923162864453815,
+      "grad_norm": 0.1857532560825348,
+      "learning_rate": 4.898186116882163e-05,
+      "loss": 0.4287,
       "step": 16435
     },
     {
-      "epoch": 0.58,
-      "learning_rate": 4.9044131759497755e-05,
-      "loss": 0.2949,
+      "epoch": 0.5924964861066061,
+      "grad_norm": 0.2215130776166916,
+      "learning_rate": 4.8981036703527584e-05,
+      "loss": 0.4508,
       "step": 16440
     },
     {
-      "epoch": 0.58,
-      "learning_rate": 4.904335140959611e-05,
-      "loss": 0.2984,
+      "epoch": 0.5926766857678307,
+      "grad_norm": 0.16718199849128723,
+      "learning_rate": 4.898021191149655e-05,
+      "loss": 0.4264,
       "step": 16445
     },
     {
-      "epoch": 0.58,
-      "learning_rate": 4.904257074750738e-05,
-      "loss": 0.3246,
+      "epoch": 0.5928568854290553,
+      "grad_norm": 0.16508464515209198,
+      "learning_rate": 4.897938679273979e-05,
+      "loss": 0.4389,
       "step": 16450
     },
     {
-      "epoch": 0.58,
-      "learning_rate": 4.904178977324171e-05,
-      "loss": 0.3193,
+      "epoch": 0.5930370850902801,
+      "grad_norm": 0.16538691520690918,
+      "learning_rate": 4.897856134726851e-05,
+      "loss": 0.4461,
       "step": 16455
     },
     {
-      "epoch": 0.58,
-      "learning_rate": 4.904100848680925e-05,
-      "loss": 0.3002,
+      "epoch": 0.5932172847515047,
+      "grad_norm": 0.16483493149280548,
+      "learning_rate": 4.897773557509398e-05,
+      "loss": 0.4803,
       "step": 16460
     },
     {
-      "epoch": 0.58,
-      "learning_rate": 4.9040226888220144e-05,
-      "loss": 0.2971,
+      "epoch": 0.5933974844127293,
+      "grad_norm": 0.17354772984981537,
+      "learning_rate": 4.897690947622745e-05,
+      "loss": 0.4594,
       "step": 16465
     },
     {
-      "epoch": 0.58,
-      "learning_rate": 4.903944497748453e-05,
-      "loss": 0.3081,
+      "epoch": 0.593577684073954,
+      "grad_norm": 0.18836447596549988,
+      "learning_rate": 4.8976083050680164e-05,
+      "loss": 0.4594,
       "step": 16470
     },
     {
-      "epoch": 0.58,
-      "learning_rate": 4.9038662754612564e-05,
-      "loss": 0.3071,
+      "epoch": 0.5937578837351786,
+      "grad_norm": 0.17023785412311554,
+      "learning_rate": 4.897525629846339e-05,
+      "loss": 0.4418,
       "step": 16475
     },
     {
-      "epoch": 0.58,
-      "learning_rate": 4.9037880219614405e-05,
-      "loss": 0.3106,
+      "epoch": 0.5939380833964032,
+      "grad_norm": 0.13566334545612335,
+      "learning_rate": 4.897442921958839e-05,
+      "loss": 0.4338,
       "step": 16480
     },
     {
-      "epoch": 0.58,
-      "learning_rate": 4.903709737250021e-05,
-      "loss": 0.3069,
+      "epoch": 0.5941182830576278,
+      "grad_norm": 0.18862979114055634,
+      "learning_rate": 4.8973601814066436e-05,
+      "loss": 0.4653,
       "step": 16485
     },
     {
-      "epoch": 0.58,
-      "learning_rate": 4.9036314213280153e-05,
-      "loss": 0.3297,
+      "epoch": 0.5942984827188524,
+      "grad_norm": 0.16662077605724335,
+      "learning_rate": 4.897277408190879e-05,
+      "loss": 0.4596,
       "step": 16490
     },
     {
-      "epoch": 0.58,
-      "learning_rate": 4.903553074196439e-05,
-      "loss": 0.3066,
+      "epoch": 0.5944786823800772,
+      "grad_norm": 0.20535314083099365,
+      "learning_rate": 4.897194602312675e-05,
+      "loss": 0.4829,
       "step": 16495
     },
     {
-      "epoch": 0.58,
-      "learning_rate": 4.90347469585631e-05,
-      "loss": 0.2976,
+      "epoch": 0.5946588820413018,
+      "grad_norm": 0.16183775663375854,
+      "learning_rate": 4.8971117637731576e-05,
+      "loss": 0.4483,
       "step": 16500
     },
     {
-      "epoch": 0.58,
-      "eval_loss": 0.30791518092155457,
-      "eval_runtime": 10.5594,
-      "eval_samples_per_second": 9.47,
-      "eval_steps_per_second": 9.47,
+      "epoch": 0.5946588820413018,
+      "eval_loss": 0.4669727683067322,
+      "eval_runtime": 3.6173,
+      "eval_samples_per_second": 27.645,
+      "eval_steps_per_second": 6.911,
       "step": 16500
     },
     {
-      "epoch": 0.58,
-      "learning_rate": 4.9033962863086466e-05,
-      "loss": 0.304,
+      "epoch": 0.5948390817025264,
+      "grad_norm": 0.18323639035224915,
+      "learning_rate": 4.8970288925734575e-05,
+      "loss": 0.4502,
       "step": 16505
     },
     {
-      "epoch": 0.58,
-      "learning_rate": 4.903317845554466e-05,
-      "loss": 0.3337,
+      "epoch": 0.595019281363751,
+      "grad_norm": 0.1653471291065216,
+      "learning_rate": 4.8969459887147025e-05,
+      "loss": 0.4789,
       "step": 16510
     },
     {
-      "epoch": 0.58,
-      "learning_rate": 4.903239373594788e-05,
-      "loss": 0.3248,
+      "epoch": 0.5951994810249757,
+      "grad_norm": 0.14966969192028046,
+      "learning_rate": 4.896863052198022e-05,
+      "loss": 0.4468,
       "step": 16515
     },
     {
-      "epoch": 0.58,
-      "learning_rate": 4.90316087043063e-05,
-      "loss": 0.2933,
+      "epoch": 0.5953796806862003,
+      "grad_norm": 0.16616958379745483,
+      "learning_rate": 4.896780083024547e-05,
+      "loss": 0.49,
       "step": 16520
     },
     {
-      "epoch": 0.58,
-      "learning_rate": 4.9030823360630106e-05,
-      "loss": 0.3117,
+      "epoch": 0.5955598803474249,
+      "grad_norm": 0.17226427793502808,
+      "learning_rate": 4.896697081195407e-05,
+      "loss": 0.4843,
       "step": 16525
     },
     {
-      "epoch": 0.58,
-      "learning_rate": 4.903003770492952e-05,
-      "loss": 0.3175,
+      "epoch": 0.5957400800086495,
+      "grad_norm": 0.21469104290008545,
+      "learning_rate": 4.896614046711734e-05,
+      "loss": 0.4526,
       "step": 16530
     },
     {
-      "epoch": 0.58,
-      "learning_rate": 4.902925173721473e-05,
-      "loss": 0.3121,
+      "epoch": 0.5959202796698743,
+      "grad_norm": 0.17571860551834106,
+      "learning_rate": 4.896530979574658e-05,
+      "loss": 0.4803,
       "step": 16535
     },
     {
-      "epoch": 0.58,
-      "learning_rate": 4.9028465457495936e-05,
-      "loss": 0.2972,
+      "epoch": 0.5961004793310989,
+      "grad_norm": 0.16215255856513977,
+      "learning_rate": 4.896447879785311e-05,
+      "loss": 0.4276,
       "step": 16540
     },
     {
-      "epoch": 0.58,
-      "learning_rate": 4.9027678865783354e-05,
-      "loss": 0.3076,
+      "epoch": 0.5962806789923235,
+      "grad_norm": 0.17751629650592804,
+      "learning_rate": 4.896364747344827e-05,
+      "loss": 0.4628,
       "step": 16545
     },
     {
-      "epoch": 0.58,
-      "learning_rate": 4.90268919620872e-05,
-      "loss": 0.3243,
+      "epoch": 0.5964608786535481,
+      "grad_norm": 0.1444987803697586,
+      "learning_rate": 4.896281582254336e-05,
+      "loss": 0.4643,
       "step": 16550
     },
     {
-      "epoch": 0.58,
-      "learning_rate": 4.902610474641769e-05,
-      "loss": 0.3134,
+      "epoch": 0.5966410783147728,
+      "grad_norm": 0.17875704169273376,
+      "learning_rate": 4.8961983845149725e-05,
+      "loss": 0.4847,
       "step": 16555
     },
     {
-      "epoch": 0.58,
-      "learning_rate": 4.902531721878504e-05,
-      "loss": 0.3299,
+      "epoch": 0.5968212779759974,
+      "grad_norm": 0.20280811190605164,
+      "learning_rate": 4.8961151541278706e-05,
+      "loss": 0.4574,
       "step": 16560
     },
     {
-      "epoch": 0.58,
-      "learning_rate": 4.902452937919948e-05,
-      "loss": 0.3056,
+      "epoch": 0.597001477637222,
+      "grad_norm": 0.1767544448375702,
+      "learning_rate": 4.8960318910941626e-05,
+      "loss": 0.4164,
       "step": 16565
     },
     {
-      "epoch": 0.58,
-      "learning_rate": 4.9023741227671236e-05,
-      "loss": 0.336,
+      "epoch": 0.5971816772984467,
+      "grad_norm": 0.17916817963123322,
+      "learning_rate": 4.8959485954149855e-05,
+      "loss": 0.4439,
       "step": 16570
     },
     {
-      "epoch": 0.58,
-      "learning_rate": 4.902295276421055e-05,
-      "loss": 0.3167,
+      "epoch": 0.5973618769596714,
+      "grad_norm": 0.16673676669597626,
+      "learning_rate": 4.895865267091471e-05,
+      "loss": 0.4553,
       "step": 16575
     },
     {
-      "epoch": 0.58,
-      "learning_rate": 4.902216398882766e-05,
-      "loss": 0.29,
+      "epoch": 0.597542076620896,
+      "grad_norm": 0.1650109440088272,
+      "learning_rate": 4.895781906124757e-05,
+      "loss": 0.4454,
       "step": 16580
     },
     {
-      "epoch": 0.58,
-      "learning_rate": 4.902137490153279e-05,
-      "loss": 0.3181,
+      "epoch": 0.5977222762821206,
+      "grad_norm": 0.1822521984577179,
+      "learning_rate": 4.895698512515978e-05,
+      "loss": 0.4683,
       "step": 16585
     },
     {
-      "epoch": 0.58,
-      "learning_rate": 4.9020585502336205e-05,
-      "loss": 0.2852,
+      "epoch": 0.5979024759433452,
+      "grad_norm": 0.1935487687587738,
+      "learning_rate": 4.895615086266271e-05,
+      "loss": 0.4437,
       "step": 16590
     },
     {
-      "epoch": 0.58,
-      "learning_rate": 4.9019795791248144e-05,
-      "loss": 0.3185,
+      "epoch": 0.5980826756045698,
+      "grad_norm": 0.15225858986377716,
+      "learning_rate": 4.895531627376772e-05,
+      "loss": 0.4567,
       "step": 16595
     },
     {
-      "epoch": 0.58,
-      "learning_rate": 4.901900576827886e-05,
-      "loss": 0.3138,
+      "epoch": 0.5982628752657945,
+      "grad_norm": 0.1286819726228714,
+      "learning_rate": 4.8954481358486185e-05,
+      "loss": 0.4368,
       "step": 16600
     },
     {
-      "epoch": 0.58,
-      "learning_rate": 4.9018215433438616e-05,
-      "loss": 0.2948,
+      "epoch": 0.5984430749270191,
+      "grad_norm": 0.1508607119321823,
+      "learning_rate": 4.8953646116829477e-05,
+      "loss": 0.4622,
       "step": 16605
     },
     {
-      "epoch": 0.58,
-      "learning_rate": 4.9017424786737674e-05,
-      "loss": 0.3113,
+      "epoch": 0.5986232745882438,
+      "grad_norm": 0.1493672877550125,
+      "learning_rate": 4.895281054880898e-05,
+      "loss": 0.4192,
       "step": 16610
     },
     {
-      "epoch": 0.58,
-      "learning_rate": 4.9016633828186306e-05,
-      "loss": 0.2955,
+      "epoch": 0.5988034742494684,
+      "grad_norm": 0.17611226439476013,
+      "learning_rate": 4.895197465443608e-05,
+      "loss": 0.4706,
       "step": 16615
     },
     {
-      "epoch": 0.58,
-      "learning_rate": 4.901584255779477e-05,
-      "loss": 0.3217,
+      "epoch": 0.5989836739106931,
+      "grad_norm": 0.1974530965089798,
+      "learning_rate": 4.8951138433722154e-05,
+      "loss": 0.4066,
       "step": 16620
     },
     {
-      "epoch": 0.58,
-      "learning_rate": 4.9015209316963526e-05,
-      "loss": 0.3111,
+      "epoch": 0.5991638735719177,
+      "grad_norm": 0.14464324712753296,
+      "learning_rate": 4.895030188667861e-05,
+      "loss": 0.4478,
       "step": 16625
     },
     {
-      "epoch": 0.59,
-      "learning_rate": 4.9014417485285595e-05,
-      "loss": 0.2994,
+      "epoch": 0.5993440732331423,
+      "grad_norm": 0.13125242292881012,
+      "learning_rate": 4.894946501331684e-05,
+      "loss": 0.4564,
       "step": 16630
     },
     {
-      "epoch": 0.59,
-      "learning_rate": 4.9013625341796276e-05,
-      "loss": 0.3065,
+      "epoch": 0.5995242728943669,
+      "grad_norm": 0.20426687598228455,
+      "learning_rate": 4.894862781364824e-05,
+      "loss": 0.4608,
       "step": 16635
     },
     {
-      "epoch": 0.59,
-      "learning_rate": 4.9012832886505855e-05,
-      "loss": 0.3198,
+      "epoch": 0.5997044725555916,
+      "grad_norm": 0.17692770063877106,
+      "learning_rate": 4.894779028768423e-05,
+      "loss": 0.469,
       "step": 16640
     },
     {
-      "epoch": 0.59,
-      "learning_rate": 4.901204011942463e-05,
-      "loss": 0.3359,
+      "epoch": 0.5998846722168162,
+      "grad_norm": 0.18528582155704498,
+      "learning_rate": 4.894695243543621e-05,
+      "loss": 0.4498,
       "step": 16645
     },
     {
-      "epoch": 0.59,
-      "learning_rate": 4.901124704056289e-05,
-      "loss": 0.3179,
+      "epoch": 0.6000648718780409,
+      "grad_norm": 0.19530199468135834,
+      "learning_rate": 4.8946114256915597e-05,
+      "loss": 0.4055,
       "step": 16650
     },
     {
-      "epoch": 0.59,
-      "learning_rate": 4.901045364993092e-05,
-      "loss": 0.2868,
+      "epoch": 0.6002450715392655,
+      "grad_norm": 0.1795613169670105,
+      "learning_rate": 4.894527575213382e-05,
+      "loss": 0.4876,
       "step": 16655
     },
     {
-      "epoch": 0.59,
-      "learning_rate": 4.900965994753904e-05,
-      "loss": 0.3325,
+      "epoch": 0.6004252712004902,
+      "grad_norm": 0.16493387520313263,
+      "learning_rate": 4.8944436921102296e-05,
+      "loss": 0.4382,
       "step": 16660
     },
     {
-      "epoch": 0.59,
-      "learning_rate": 4.9008865933397547e-05,
-      "loss": 0.3052,
+      "epoch": 0.6006054708617148,
+      "grad_norm": 0.1708437204360962,
+      "learning_rate": 4.894359776383245e-05,
+      "loss": 0.4134,
       "step": 16665
     },
     {
-      "epoch": 0.59,
-      "learning_rate": 4.9008071607516746e-05,
-      "loss": 0.2971,
+      "epoch": 0.6007856705229394,
+      "grad_norm": 0.14613789319992065,
+      "learning_rate": 4.894275828033572e-05,
+      "loss": 0.469,
       "step": 16670
     },
     {
-      "epoch": 0.59,
-      "learning_rate": 4.900727696990697e-05,
-      "loss": 0.3172,
+      "epoch": 0.600965870184164,
+      "grad_norm": 0.15820561349391937,
+      "learning_rate": 4.894191847062355e-05,
+      "loss": 0.435,
       "step": 16675
     },
     {
-      "epoch": 0.59,
-      "learning_rate": 4.900648202057851e-05,
-      "loss": 0.3044,
+      "epoch": 0.6011460698453887,
+      "grad_norm": 0.16094128787517548,
+      "learning_rate": 4.894107833470737e-05,
+      "loss": 0.4457,
       "step": 16680
     },
     {
-      "epoch": 0.59,
-      "learning_rate": 4.9005686759541714e-05,
-      "loss": 0.3029,
+      "epoch": 0.6013262695066133,
+      "grad_norm": 0.21351011097431183,
+      "learning_rate": 4.8940237872598635e-05,
+      "loss": 0.4868,
       "step": 16685
     },
     {
-      "epoch": 0.59,
-      "learning_rate": 4.9004891186806894e-05,
-      "loss": 0.3203,
+      "epoch": 0.601506469167838,
+      "grad_norm": 0.1446036994457245,
+      "learning_rate": 4.8939397084308794e-05,
+      "loss": 0.4357,
       "step": 16690
     },
     {
-      "epoch": 0.59,
-      "learning_rate": 4.900409530238438e-05,
-      "loss": 0.301,
+      "epoch": 0.6016866688290626,
+      "grad_norm": 0.16345882415771484,
+      "learning_rate": 4.8938555969849306e-05,
+      "loss": 0.4722,
       "step": 16695
     },
     {
-      "epoch": 0.59,
-      "learning_rate": 4.9003299106284504e-05,
-      "loss": 0.3246,
+      "epoch": 0.6018668684902873,
+      "grad_norm": 0.1572277694940567,
+      "learning_rate": 4.893771452923162e-05,
+      "loss": 0.4265,
       "step": 16700
     },
     {
-      "epoch": 0.59,
-      "learning_rate": 4.900250259851762e-05,
-      "loss": 0.3039,
+      "epoch": 0.6020470681515119,
+      "grad_norm": 0.188238263130188,
+      "learning_rate": 4.893687276246721e-05,
+      "loss": 0.4051,
       "step": 16705
     },
     {
-      "epoch": 0.59,
-      "learning_rate": 4.900170577909404e-05,
-      "loss": 0.2734,
+      "epoch": 0.6022272678127365,
+      "grad_norm": 0.15958626568317413,
+      "learning_rate": 4.893603066956755e-05,
+      "loss": 0.4454,
       "step": 16710
     },
     {
-      "epoch": 0.59,
-      "learning_rate": 4.900090864802414e-05,
-      "loss": 0.2907,
+      "epoch": 0.6024074674739611,
+      "grad_norm": 0.18399354815483093,
+      "learning_rate": 4.89351882505441e-05,
+      "loss": 0.4459,
       "step": 16715
     },
     {
-      "epoch": 0.59,
-      "learning_rate": 4.9000111205318254e-05,
-      "loss": 0.3238,
+      "epoch": 0.6025876671351857,
+      "grad_norm": 0.16669590771198273,
+      "learning_rate": 4.8934345505408344e-05,
+      "loss": 0.4402,
       "step": 16720
     },
     {
-      "epoch": 0.59,
-      "learning_rate": 4.899931345098674e-05,
-      "loss": 0.3159,
+      "epoch": 0.6027678667964104,
+      "grad_norm": 0.16864614188671112,
+      "learning_rate": 4.893350243417177e-05,
+      "loss": 0.4374,
       "step": 16725
     },
     {
-      "epoch": 0.59,
-      "learning_rate": 4.899851538503996e-05,
-      "loss": 0.3073,
+      "epoch": 0.6029480664576351,
+      "grad_norm": 0.1995852291584015,
+      "learning_rate": 4.893265903684585e-05,
+      "loss": 0.4212,
       "step": 16730
     },
     {
-      "epoch": 0.59,
-      "learning_rate": 4.899771700748827e-05,
-      "loss": 0.3185,
+      "epoch": 0.6031282661188597,
+      "grad_norm": 0.17713283002376556,
+      "learning_rate": 4.8931815313442095e-05,
+      "loss": 0.4731,
       "step": 16735
     },
     {
-      "epoch": 0.59,
-      "learning_rate": 4.899691831834205e-05,
-      "loss": 0.2971,
+      "epoch": 0.6033084657800843,
+      "grad_norm": 0.16761018335819244,
+      "learning_rate": 4.893097126397198e-05,
+      "loss": 0.4195,
       "step": 16740
     },
     {
-      "epoch": 0.59,
-      "learning_rate": 4.899611931761164e-05,
-      "loss": 0.31,
+      "epoch": 0.603488665441309,
+      "grad_norm": 0.14120902121067047,
+      "learning_rate": 4.893012688844702e-05,
+      "loss": 0.4436,
       "step": 16745
     },
     {
-      "epoch": 0.59,
-      "learning_rate": 4.899532000530746e-05,
-      "loss": 0.3042,
+      "epoch": 0.6036688651025336,
+      "grad_norm": 0.1854945719242096,
+      "learning_rate": 4.8929282186878714e-05,
+      "loss": 0.4699,
       "step": 16750
     },
     {
-      "epoch": 0.59,
-      "learning_rate": 4.899452038143985e-05,
-      "loss": 0.3021,
+      "epoch": 0.6038490647637582,
+      "grad_norm": 0.15078695118427277,
+      "learning_rate": 4.892843715927857e-05,
+      "loss": 0.4322,
       "step": 16755
     },
     {
-      "epoch": 0.59,
-      "learning_rate": 4.89937204460192e-05,
-      "loss": 0.3235,
+      "epoch": 0.6040292644249828,
+      "grad_norm": 0.17069613933563232,
+      "learning_rate": 4.89275918056581e-05,
+      "loss": 0.4183,
       "step": 16760
     },
     {
-      "epoch": 0.59,
-      "learning_rate": 4.899292019905591e-05,
-      "loss": 0.2933,
+      "epoch": 0.6042094640862076,
+      "grad_norm": 0.1491325944662094,
+      "learning_rate": 4.892674612602882e-05,
+      "loss": 0.4358,
       "step": 16765
     },
     {
-      "epoch": 0.59,
-      "learning_rate": 4.899211964056036e-05,
-      "loss": 0.3161,
+      "epoch": 0.6043896637474322,
+      "grad_norm": 0.17146345973014832,
+      "learning_rate": 4.892590012040227e-05,
+      "loss": 0.4274,
       "step": 16770
     },
     {
-      "epoch": 0.59,
-      "learning_rate": 4.899131877054295e-05,
-      "loss": 0.3139,
+      "epoch": 0.6045698634086568,
+      "grad_norm": 0.1850728988647461,
+      "learning_rate": 4.8925053788789954e-05,
+      "loss": 0.4698,
       "step": 16775
     },
     {
-      "epoch": 0.59,
-      "learning_rate": 4.899051758901407e-05,
-      "loss": 0.3109,
+      "epoch": 0.6047500630698814,
+      "grad_norm": 0.1922415941953659,
+      "learning_rate": 4.892420713120341e-05,
+      "loss": 0.4253,
       "step": 16780
     },
     {
-      "epoch": 0.59,
-      "learning_rate": 4.898971609598414e-05,
-      "loss": 0.3411,
+      "epoch": 0.6049302627311061,
+      "grad_norm": 0.1879766434431076,
+      "learning_rate": 4.8923360147654184e-05,
+      "loss": 0.4515,
       "step": 16785
     },
     {
-      "epoch": 0.59,
-      "learning_rate": 4.898891429146355e-05,
-      "loss": 0.3293,
+      "epoch": 0.6051104623923307,
+      "grad_norm": 0.19672177731990814,
+      "learning_rate": 4.89225128381538e-05,
+      "loss": 0.4909,
       "step": 16790
     },
     {
-      "epoch": 0.59,
-      "learning_rate": 4.898811217546272e-05,
-      "loss": 0.327,
+      "epoch": 0.6052906620535553,
+      "grad_norm": 0.16113631427288055,
+      "learning_rate": 4.8921665202713816e-05,
+      "loss": 0.4287,
       "step": 16795
     },
     {
-      "epoch": 0.59,
-      "learning_rate": 4.8987309747992064e-05,
-      "loss": 0.3188,
+      "epoch": 0.6054708617147799,
+      "grad_norm": 0.1752663552761078,
+      "learning_rate": 4.8920817241345776e-05,
+      "loss": 0.4615,
       "step": 16800
     },
     {
-      "epoch": 0.59,
-      "learning_rate": 4.8986507009061994e-05,
-      "loss": 0.2961,
+      "epoch": 0.6056510613760047,
+      "grad_norm": 0.15758828818798065,
+      "learning_rate": 4.8919968954061223e-05,
+      "loss": 0.4314,
       "step": 16805
     },
     {
-      "epoch": 0.59,
-      "learning_rate": 4.898570395868294e-05,
-      "loss": 0.3104,
+      "epoch": 0.6058312610372293,
+      "grad_norm": 0.22554457187652588,
+      "learning_rate": 4.891912034087173e-05,
+      "loss": 0.48,
       "step": 16810
     },
     {
-      "epoch": 0.59,
-      "learning_rate": 4.898490059686532e-05,
-      "loss": 0.3094,
+      "epoch": 0.6060114606984539,
+      "grad_norm": 0.16206426918506622,
+      "learning_rate": 4.8918271401788855e-05,
+      "loss": 0.4626,
       "step": 16815
     },
     {
-      "epoch": 0.59,
-      "learning_rate": 4.8984096923619586e-05,
-      "loss": 0.3213,
+      "epoch": 0.6061916603596785,
+      "grad_norm": 0.16757212579250336,
+      "learning_rate": 4.8917422136824167e-05,
+      "loss": 0.4692,
       "step": 16820
     },
     {
-      "epoch": 0.59,
-      "learning_rate": 4.898329293895616e-05,
-      "loss": 0.3141,
+      "epoch": 0.6063718600209032,
+      "grad_norm": 0.19474922120571136,
+      "learning_rate": 4.891657254598922e-05,
+      "loss": 0.4177,
       "step": 16825
     },
     {
-      "epoch": 0.59,
-      "learning_rate": 4.898248864288547e-05,
-      "loss": 0.3198,
+      "epoch": 0.6065520596821278,
+      "grad_norm": 0.15436606109142303,
+      "learning_rate": 4.891572262929561e-05,
+      "loss": 0.4621,
       "step": 16830
     },
     {
-      "epoch": 0.59,
-      "learning_rate": 4.898168403541797e-05,
-      "loss": 0.3185,
+      "epoch": 0.6067322593433524,
+      "grad_norm": 0.15269695222377777,
+      "learning_rate": 4.89148723867549e-05,
+      "loss": 0.4116,
       "step": 16835
     },
     {
-      "epoch": 0.59,
-      "learning_rate": 4.8980879116564114e-05,
-      "loss": 0.3084,
+      "epoch": 0.606912459004577,
+      "grad_norm": 0.20142877101898193,
+      "learning_rate": 4.89140218183787e-05,
+      "loss": 0.4751,
       "step": 16840
     },
     {
-      "epoch": 0.59,
-      "learning_rate": 4.898007388633434e-05,
-      "loss": 0.3077,
+      "epoch": 0.6070926586658018,
+      "grad_norm": 0.14995922148227692,
+      "learning_rate": 4.8913170924178576e-05,
+      "loss": 0.3919,
       "step": 16845
     },
     {
-      "epoch": 0.59,
-      "learning_rate": 4.897926834473912e-05,
-      "loss": 0.3165,
+      "epoch": 0.6072728583270264,
+      "grad_norm": 0.179569274187088,
+      "learning_rate": 4.8912319704166124e-05,
+      "loss": 0.4542,
       "step": 16850
     },
     {
-      "epoch": 0.59,
-      "learning_rate": 4.89784624917889e-05,
-      "loss": 0.3233,
+      "epoch": 0.607453057988251,
+      "grad_norm": 0.15274690091609955,
+      "learning_rate": 4.891146815835295e-05,
+      "loss": 0.4364,
       "step": 16855
     },
     {
-      "epoch": 0.59,
-      "learning_rate": 4.8977656327494145e-05,
-      "loss": 0.3296,
+      "epoch": 0.6076332576494756,
+      "grad_norm": 0.13948167860507965,
+      "learning_rate": 4.8910616286750654e-05,
+      "loss": 0.4454,
       "step": 16860
     },
     {
-      "epoch": 0.59,
-      "learning_rate": 4.897684985186533e-05,
-      "loss": 0.2964,
+      "epoch": 0.6078134573107002,
+      "grad_norm": 0.18718601763248444,
+      "learning_rate": 4.890976408937084e-05,
+      "loss": 0.4327,
       "step": 16865
     },
     {
-      "epoch": 0.59,
-      "learning_rate": 4.8976043064912916e-05,
-      "loss": 0.3053,
+      "epoch": 0.6079936569719249,
+      "grad_norm": 0.2123367339372635,
+      "learning_rate": 4.890891156622511e-05,
+      "loss": 0.419,
       "step": 16870
     },
     {
-      "epoch": 0.59,
-      "learning_rate": 4.8975235966647376e-05,
-      "loss": 0.2974,
+      "epoch": 0.6081738566331495,
+      "grad_norm": 0.18205073475837708,
+      "learning_rate": 4.8908058717325097e-05,
+      "loss": 0.4819,
       "step": 16875
     },
     {
-      "epoch": 0.59,
-      "learning_rate": 4.8974428557079204e-05,
-      "loss": 0.3186,
+      "epoch": 0.6083540562943741,
+      "grad_norm": 0.1736924946308136,
+      "learning_rate": 4.8907205542682414e-05,
+      "loss": 0.4596,
       "step": 16880
     },
     {
-      "epoch": 0.59,
-      "learning_rate": 4.897362083621888e-05,
-      "loss": 0.294,
+      "epoch": 0.6085342559555988,
+      "grad_norm": 0.18796013295650482,
+      "learning_rate": 4.890635204230868e-05,
+      "loss": 0.4273,
       "step": 16885
     },
     {
-      "epoch": 0.59,
-      "learning_rate": 4.8972812804076884e-05,
-      "loss": 0.3269,
+      "epoch": 0.6087144556168235,
+      "grad_norm": 0.19030854105949402,
+      "learning_rate": 4.890549821621553e-05,
+      "loss": 0.4642,
       "step": 16890
     },
     {
-      "epoch": 0.59,
-      "learning_rate": 4.897200446066371e-05,
-      "loss": 0.3195,
+      "epoch": 0.6088946552780481,
+      "grad_norm": 0.16654013097286224,
+      "learning_rate": 4.8904644064414585e-05,
+      "loss": 0.4615,
       "step": 16895
     },
     {
-      "epoch": 0.59,
-      "learning_rate": 4.897119580598987e-05,
-      "loss": 0.3059,
+      "epoch": 0.6090748549392727,
+      "grad_norm": 0.20467707514762878,
+      "learning_rate": 4.8903789586917505e-05,
+      "loss": 0.4292,
       "step": 16900
     },
     {
-      "epoch": 0.59,
-      "learning_rate": 4.897038684006583e-05,
-      "loss": 0.2914,
+      "epoch": 0.6092550546004973,
+      "grad_norm": 0.14566195011138916,
+      "learning_rate": 4.890293478373592e-05,
+      "loss": 0.483,
       "step": 16905
     },
     {
-      "epoch": 0.59,
-      "learning_rate": 4.896957756290213e-05,
-      "loss": 0.3063,
+      "epoch": 0.609435254261722,
+      "grad_norm": 0.19296708703041077,
+      "learning_rate": 4.8902079654881466e-05,
+      "loss": 0.4536,
       "step": 16910
     },
     {
-      "epoch": 0.6,
-      "learning_rate": 4.896876797450926e-05,
-      "loss": 0.3383,
+      "epoch": 0.6096154539229466,
+      "grad_norm": 0.15972061455249786,
+      "learning_rate": 4.890122420036581e-05,
+      "loss": 0.4392,
       "step": 16915
     },
     {
-      "epoch": 0.6,
-      "learning_rate": 4.8967958074897724e-05,
-      "loss": 0.3196,
+      "epoch": 0.6097956535841713,
+      "grad_norm": 0.18985244631767273,
+      "learning_rate": 4.890036842020061e-05,
+      "loss": 0.4542,
       "step": 16920
     },
     {
-      "epoch": 0.6,
-      "learning_rate": 4.896714786407806e-05,
-      "loss": 0.3065,
+      "epoch": 0.6099758532453959,
+      "grad_norm": 0.1495533585548401,
+      "learning_rate": 4.88995123143975e-05,
+      "loss": 0.4318,
       "step": 16925
     },
     {
-      "epoch": 0.6,
-      "learning_rate": 4.8966337342060765e-05,
-      "loss": 0.3096,
+      "epoch": 0.6101560529066206,
+      "grad_norm": 0.21450436115264893,
+      "learning_rate": 4.8898655882968175e-05,
+      "loss": 0.4514,
       "step": 16930
     },
     {
-      "epoch": 0.6,
-      "learning_rate": 4.8965526508856387e-05,
-      "loss": 0.2989,
+      "epoch": 0.6103362525678452,
+      "grad_norm": 0.2090601772069931,
+      "learning_rate": 4.889779912592429e-05,
+      "loss": 0.4552,
       "step": 16935
     },
     {
-      "epoch": 0.6,
-      "learning_rate": 4.896471536447543e-05,
-      "loss": 0.3001,
+      "epoch": 0.6105164522290698,
+      "grad_norm": 0.1768369823694229,
+      "learning_rate": 4.889694204327751e-05,
+      "loss": 0.4526,
       "step": 16940
     },
     {
-      "epoch": 0.6,
-      "learning_rate": 4.896390390892844e-05,
-      "loss": 0.3072,
+      "epoch": 0.6106966518902944,
+      "grad_norm": 0.16384200751781464,
+      "learning_rate": 4.889608463503953e-05,
+      "loss": 0.4559,
       "step": 16945
     },
     {
-      "epoch": 0.6,
-      "learning_rate": 4.896309214222595e-05,
-      "loss": 0.2968,
+      "epoch": 0.610876851551519,
+      "grad_norm": 0.14642253518104553,
+      "learning_rate": 4.8895226901222026e-05,
+      "loss": 0.4559,
       "step": 16950
     },
     {
-      "epoch": 0.6,
-      "learning_rate": 4.8962280064378504e-05,
-      "loss": 0.2986,
+      "epoch": 0.6110570512127437,
+      "grad_norm": 0.14469105005264282,
+      "learning_rate": 4.889436884183667e-05,
+      "loss": 0.4116,
       "step": 16955
     },
     {
-      "epoch": 0.6,
-      "learning_rate": 4.896146767539664e-05,
-      "loss": 0.2951,
+      "epoch": 0.6112372508739684,
+      "grad_norm": 0.15812519192695618,
+      "learning_rate": 4.889351045689518e-05,
+      "loss": 0.4722,
       "step": 16960
     },
     {
-      "epoch": 0.6,
-      "learning_rate": 4.896065497529091e-05,
-      "loss": 0.3021,
+      "epoch": 0.611417450535193,
+      "grad_norm": 0.18630902469158173,
+      "learning_rate": 4.889265174640922e-05,
+      "loss": 0.4555,
       "step": 16965
     },
     {
-      "epoch": 0.6,
-      "learning_rate": 4.8959841964071864e-05,
-      "loss": 0.3164,
+      "epoch": 0.6115976501964177,
+      "grad_norm": 0.1799592524766922,
+      "learning_rate": 4.889179271039052e-05,
+      "loss": 0.4385,
       "step": 16970
     },
     {
-      "epoch": 0.6,
-      "learning_rate": 4.895902864175006e-05,
-      "loss": 0.329,
+      "epoch": 0.6117778498576423,
+      "grad_norm": 0.17888915538787842,
+      "learning_rate": 4.8890933348850757e-05,
+      "loss": 0.4545,
       "step": 16975
     },
     {
-      "epoch": 0.6,
-      "learning_rate": 4.895821500833606e-05,
-      "loss": 0.3042,
+      "epoch": 0.6119580495188669,
+      "grad_norm": 0.16347283124923706,
+      "learning_rate": 4.8890073661801655e-05,
+      "loss": 0.3919,
       "step": 16980
     },
     {
-      "epoch": 0.6,
-      "learning_rate": 4.8957401063840435e-05,
-      "loss": 0.3218,
+      "epoch": 0.6121382491800915,
+      "grad_norm": 0.22050714492797852,
+      "learning_rate": 4.888921364925493e-05,
+      "loss": 0.4469,
       "step": 16985
     },
     {
-      "epoch": 0.6,
-      "learning_rate": 4.895658680827374e-05,
-      "loss": 0.315,
+      "epoch": 0.6123184488413161,
+      "grad_norm": 0.2015305459499359,
+      "learning_rate": 4.888835331122229e-05,
+      "loss": 0.4552,
       "step": 16990
     },
     {
-      "epoch": 0.6,
-      "learning_rate": 4.8955772241646546e-05,
-      "loss": 0.3186,
+      "epoch": 0.6124986485025408,
+      "grad_norm": 0.1845419853925705,
+      "learning_rate": 4.888749264771546e-05,
+      "loss": 0.4543,
       "step": 16995
     },
     {
-      "epoch": 0.6,
-      "learning_rate": 4.8954957363969445e-05,
-      "loss": 0.3001,
+      "epoch": 0.6126788481637655,
+      "grad_norm": 0.2000664323568344,
+      "learning_rate": 4.8886631658746175e-05,
+      "loss": 0.4648,
       "step": 17000
     },
     {
-      "epoch": 0.6,
-      "eval_loss": 0.30690938234329224,
-      "eval_runtime": 10.5382,
-      "eval_samples_per_second": 9.489,
-      "eval_steps_per_second": 9.489,
+      "epoch": 0.6126788481637655,
+      "eval_loss": 0.4671698808670044,
+      "eval_runtime": 3.5317,
+      "eval_samples_per_second": 28.315,
+      "eval_steps_per_second": 7.079,
       "step": 17000
     },
     {
-      "epoch": 0.6,
-      "learning_rate": 4.895414217525301e-05,
-      "loss": 0.3164,
+      "epoch": 0.6128590478249901,
+      "grad_norm": 0.17209337651729584,
+      "learning_rate": 4.8885770344326154e-05,
+      "loss": 0.3971,
       "step": 17005
     },
     {
-      "epoch": 0.6,
-      "learning_rate": 4.895332667550783e-05,
-      "loss": 0.2993,
+      "epoch": 0.6130392474862147,
+      "grad_norm": 0.17581631243228912,
+      "learning_rate": 4.8884908704467137e-05,
+      "loss": 0.4352,
       "step": 17010
     },
     {
-      "epoch": 0.6,
-      "learning_rate": 4.895251086474447e-05,
-      "loss": 0.3298,
+      "epoch": 0.6132194471474394,
+      "grad_norm": 0.19064390659332275,
+      "learning_rate": 4.888404673918085e-05,
+      "loss": 0.4705,
       "step": 17015
     },
     {
-      "epoch": 0.6,
-      "learning_rate": 4.895169474297356e-05,
-      "loss": 0.3277,
+      "epoch": 0.613399646808664,
+      "grad_norm": 0.1969899833202362,
+      "learning_rate": 4.8883184448479066e-05,
+      "loss": 0.4157,
       "step": 17020
     },
     {
-      "epoch": 0.6,
-      "learning_rate": 4.895087831020567e-05,
-      "loss": 0.3215,
+      "epoch": 0.6135798464698886,
+      "grad_norm": 0.17131738364696503,
+      "learning_rate": 4.888232183237352e-05,
+      "loss": 0.436,
       "step": 17025
     },
     {
-      "epoch": 0.6,
-      "learning_rate": 4.895006156645141e-05,
-      "loss": 0.332,
+      "epoch": 0.6137600461311132,
+      "grad_norm": 0.17256931960582733,
+      "learning_rate": 4.888145889087595e-05,
+      "loss": 0.4619,
       "step": 17030
     },
     {
-      "epoch": 0.6,
-      "learning_rate": 4.8949244511721385e-05,
-      "loss": 0.3016,
+      "epoch": 0.6139402457923379,
+      "grad_norm": 0.17418308556079865,
+      "learning_rate": 4.888059562399814e-05,
+      "loss": 0.4318,
       "step": 17035
     },
     {
-      "epoch": 0.6,
-      "learning_rate": 4.894842714602621e-05,
-      "loss": 0.3298,
+      "epoch": 0.6141204454535626,
+      "grad_norm": 0.1743146926164627,
+      "learning_rate": 4.887973203175183e-05,
+      "loss": 0.4002,
       "step": 17040
     },
     {
-      "epoch": 0.6,
-      "learning_rate": 4.894760946937648e-05,
-      "loss": 0.3033,
+      "epoch": 0.6143006451147872,
+      "grad_norm": 0.168928861618042,
+      "learning_rate": 4.88788681141488e-05,
+      "loss": 0.4291,
       "step": 17045
     },
     {
-      "epoch": 0.6,
-      "learning_rate": 4.8946791481782826e-05,
-      "loss": 0.3404,
+      "epoch": 0.6144808447760118,
+      "grad_norm": 0.1870613992214203,
+      "learning_rate": 4.8878003871200807e-05,
+      "loss": 0.4425,
       "step": 17050
     },
     {
-      "epoch": 0.6,
-      "learning_rate": 4.894597318325587e-05,
-      "loss": 0.3124,
+      "epoch": 0.6146610444372365,
+      "grad_norm": 0.1746305525302887,
+      "learning_rate": 4.8877139302919636e-05,
+      "loss": 0.4185,
       "step": 17055
     },
     {
-      "epoch": 0.6,
-      "learning_rate": 4.8945154573806235e-05,
-      "loss": 0.3361,
+      "epoch": 0.6148412440984611,
+      "grad_norm": 0.16369090974330902,
+      "learning_rate": 4.887627440931707e-05,
+      "loss": 0.4495,
       "step": 17060
     },
     {
-      "epoch": 0.6,
-      "learning_rate": 4.8944335653444544e-05,
-      "loss": 0.2948,
+      "epoch": 0.6150214437596857,
+      "grad_norm": 0.16876505315303802,
+      "learning_rate": 4.887540919040488e-05,
+      "loss": 0.4685,
       "step": 17065
     },
     {
-      "epoch": 0.6,
-      "learning_rate": 4.8943516422181435e-05,
-      "loss": 0.3212,
+      "epoch": 0.6152016434209103,
+      "grad_norm": 0.17735637724399567,
+      "learning_rate": 4.887454364619487e-05,
+      "loss": 0.3699,
       "step": 17070
     },
     {
-      "epoch": 0.6,
-      "learning_rate": 4.894269688002755e-05,
-      "loss": 0.3018,
+      "epoch": 0.6153818430821351,
+      "grad_norm": 0.17825958132743835,
+      "learning_rate": 4.8873677776698824e-05,
+      "loss": 0.4413,
       "step": 17075
     },
     {
-      "epoch": 0.6,
-      "learning_rate": 4.8941877026993514e-05,
-      "loss": 0.3198,
+      "epoch": 0.6155620427433597,
+      "grad_norm": 0.21326866745948792,
+      "learning_rate": 4.8872811581928536e-05,
+      "loss": 0.4284,
       "step": 17080
     },
     {
-      "epoch": 0.6,
-      "learning_rate": 4.894105686308999e-05,
-      "loss": 0.2923,
+      "epoch": 0.6157422424045843,
+      "grad_norm": 0.16894589364528656,
+      "learning_rate": 4.887194506189581e-05,
+      "loss": 0.4346,
       "step": 17085
     },
     {
-      "epoch": 0.6,
-      "learning_rate": 4.894023638832762e-05,
-      "loss": 0.3264,
+      "epoch": 0.6159224420658089,
+      "grad_norm": 0.18406248092651367,
+      "learning_rate": 4.887107821661247e-05,
+      "loss": 0.4555,
       "step": 17090
     },
     {
-      "epoch": 0.6,
-      "learning_rate": 4.8939415602717056e-05,
-      "loss": 0.3026,
+      "epoch": 0.6161026417270336,
+      "grad_norm": 0.17447605729103088,
+      "learning_rate": 4.887021104609029e-05,
+      "loss": 0.4917,
       "step": 17095
     },
     {
-      "epoch": 0.6,
-      "learning_rate": 4.893859450626896e-05,
-      "loss": 0.3145,
+      "epoch": 0.6162828413882582,
+      "grad_norm": 0.1875600963830948,
+      "learning_rate": 4.886934355034112e-05,
+      "loss": 0.4652,
       "step": 17100
     },
     {
-      "epoch": 0.6,
-      "learning_rate": 4.893777309899399e-05,
-      "loss": 0.3186,
+      "epoch": 0.6164630410494828,
+      "grad_norm": 0.1392199695110321,
+      "learning_rate": 4.886847572937676e-05,
+      "loss": 0.4185,
       "step": 17105
     },
     {
-      "epoch": 0.6,
-      "learning_rate": 4.8936951380902804e-05,
-      "loss": 0.3439,
+      "epoch": 0.6166432407107074,
+      "grad_norm": 0.21595901250839233,
+      "learning_rate": 4.886760758320904e-05,
+      "loss": 0.445,
       "step": 17110
     },
     {
-      "epoch": 0.6,
-      "learning_rate": 4.893612935200609e-05,
-      "loss": 0.3224,
+      "epoch": 0.6168234403719322,
+      "grad_norm": 0.19852277636528015,
+      "learning_rate": 4.886673911184978e-05,
+      "loss": 0.4681,
       "step": 17115
     },
     {
-      "epoch": 0.6,
-      "learning_rate": 4.89353070123145e-05,
-      "loss": 0.3126,
+      "epoch": 0.6170036400331568,
+      "grad_norm": 0.15305131673812866,
+      "learning_rate": 4.8865870315310834e-05,
+      "loss": 0.4507,
       "step": 17120
     },
     {
-      "epoch": 0.6,
-      "learning_rate": 4.893448436183873e-05,
-      "loss": 0.314,
+      "epoch": 0.6171838396943814,
+      "grad_norm": 0.1263405680656433,
+      "learning_rate": 4.886500119360402e-05,
+      "loss": 0.4257,
       "step": 17125
     },
     {
-      "epoch": 0.6,
-      "learning_rate": 4.893366140058946e-05,
-      "loss": 0.3017,
+      "epoch": 0.617364039355606,
+      "grad_norm": 0.17302808165550232,
+      "learning_rate": 4.886413174674118e-05,
+      "loss": 0.4358,
       "step": 17130
     },
     {
-      "epoch": 0.6,
-      "learning_rate": 4.893283812857736e-05,
-      "loss": 0.3289,
+      "epoch": 0.6175442390168306,
+      "grad_norm": 0.157601997256279,
+      "learning_rate": 4.886326197473417e-05,
+      "loss": 0.4449,
       "step": 17135
     },
     {
-      "epoch": 0.6,
-      "learning_rate": 4.8932014545813134e-05,
-      "loss": 0.3609,
+      "epoch": 0.6177244386780553,
+      "grad_norm": 0.20116667449474335,
+      "learning_rate": 4.8862391877594835e-05,
+      "loss": 0.4696,
       "step": 17140
     },
     {
-      "epoch": 0.6,
-      "learning_rate": 4.893119065230747e-05,
-      "loss": 0.2998,
+      "epoch": 0.6179046383392799,
+      "grad_norm": 0.14023269712924957,
+      "learning_rate": 4.886152145533503e-05,
+      "loss": 0.4229,
       "step": 17145
     },
     {
-      "epoch": 0.6,
-      "learning_rate": 4.8930366448071066e-05,
-      "loss": 0.351,
+      "epoch": 0.6180848380005045,
+      "grad_norm": 0.18752652406692505,
+      "learning_rate": 4.886065070796662e-05,
+      "loss": 0.4556,
       "step": 17150
     },
     {
-      "epoch": 0.6,
-      "learning_rate": 4.892954193311462e-05,
-      "loss": 0.3124,
+      "epoch": 0.6182650376617292,
+      "grad_norm": 0.1420794129371643,
+      "learning_rate": 4.8859779635501456e-05,
+      "loss": 0.4229,
       "step": 17155
     },
     {
-      "epoch": 0.6,
-      "learning_rate": 4.892871710744884e-05,
-      "loss": 0.3311,
+      "epoch": 0.6184452373229539,
+      "grad_norm": 0.13428914546966553,
+      "learning_rate": 4.885890823795142e-05,
+      "loss": 0.4079,
       "step": 17160
     },
     {
-      "epoch": 0.6,
-      "learning_rate": 4.892789197108445e-05,
-      "loss": 0.3135,
+      "epoch": 0.6186254369841785,
+      "grad_norm": 0.18215683102607727,
+      "learning_rate": 4.885803651532838e-05,
+      "loss": 0.4227,
       "step": 17165
     },
     {
-      "epoch": 0.6,
-      "learning_rate": 4.892706652403215e-05,
-      "loss": 0.3331,
+      "epoch": 0.6188056366454031,
+      "grad_norm": 0.14137986302375793,
+      "learning_rate": 4.88571644676442e-05,
+      "loss": 0.3939,
       "step": 17170
     },
     {
-      "epoch": 0.6,
-      "learning_rate": 4.892624076630265e-05,
-      "loss": 0.3063,
+      "epoch": 0.6189858363066277,
+      "grad_norm": 0.16938935220241547,
+      "learning_rate": 4.885629209491078e-05,
+      "loss": 0.4753,
       "step": 17175
     },
     {
-      "epoch": 0.6,
-      "learning_rate": 4.892541469790669e-05,
-      "loss": 0.3259,
+      "epoch": 0.6191660359678524,
+      "grad_norm": 0.16520722210407257,
+      "learning_rate": 4.885541939714e-05,
+      "loss": 0.4466,
       "step": 17180
     },
     {
-      "epoch": 0.6,
-      "learning_rate": 4.892458831885498e-05,
-      "loss": 0.321,
+      "epoch": 0.619346235629077,
+      "grad_norm": 0.16338035464286804,
+      "learning_rate": 4.885454637434375e-05,
+      "loss": 0.4746,
       "step": 17185
     },
     {
-      "epoch": 0.6,
-      "learning_rate": 4.892376162915826e-05,
-      "loss": 0.2919,
+      "epoch": 0.6195264352903016,
+      "grad_norm": 0.1479274183511734,
+      "learning_rate": 4.8853673026533926e-05,
+      "loss": 0.4363,
       "step": 17190
     },
     {
-      "epoch": 0.6,
-      "learning_rate": 4.8922934628827263e-05,
-      "loss": 0.314,
+      "epoch": 0.6197066349515263,
+      "grad_norm": 0.1792302280664444,
+      "learning_rate": 4.885279935372242e-05,
+      "loss": 0.4159,
       "step": 17195
     },
     {
-      "epoch": 0.61,
-      "learning_rate": 4.892210731787273e-05,
-      "loss": 0.3051,
+      "epoch": 0.619886834612751,
+      "grad_norm": 0.1646362543106079,
+      "learning_rate": 4.8851925355921144e-05,
+      "loss": 0.4298,
       "step": 17200
     },
     {
-      "epoch": 0.61,
-      "learning_rate": 4.89212796963054e-05,
-      "loss": 0.316,
+      "epoch": 0.6200670342739756,
+      "grad_norm": 0.20265613496303558,
+      "learning_rate": 4.8851051033142004e-05,
+      "loss": 0.4525,
       "step": 17205
     },
     {
-      "epoch": 0.61,
-      "learning_rate": 4.892045176413601e-05,
-      "loss": 0.305,
+      "epoch": 0.6202472339352002,
+      "grad_norm": 0.18897970020771027,
+      "learning_rate": 4.8850176385396904e-05,
+      "loss": 0.4509,
       "step": 17210
     },
     {
-      "epoch": 0.61,
-      "learning_rate": 4.8919623521375316e-05,
-      "loss": 0.307,
+      "epoch": 0.6204274335964248,
+      "grad_norm": 0.2247178852558136,
+      "learning_rate": 4.884930141269778e-05,
+      "loss": 0.4515,
       "step": 17215
     },
     {
-      "epoch": 0.61,
-      "learning_rate": 4.8918794968034084e-05,
-      "loss": 0.3026,
+      "epoch": 0.6206076332576494,
+      "grad_norm": 0.22851864993572235,
+      "learning_rate": 4.884842611505653e-05,
+      "loss": 0.4337,
       "step": 17220
     },
     {
-      "epoch": 0.61,
-      "learning_rate": 4.891796610412306e-05,
-      "loss": 0.2894,
+      "epoch": 0.6207878329188741,
+      "grad_norm": 0.17176374793052673,
+      "learning_rate": 4.8847550492485094e-05,
+      "loss": 0.4607,
       "step": 17225
     },
     {
-      "epoch": 0.61,
-      "learning_rate": 4.8917136929653007e-05,
-      "loss": 0.2895,
+      "epoch": 0.6209680325800987,
+      "grad_norm": 0.16532184183597565,
+      "learning_rate": 4.884667454499541e-05,
+      "loss": 0.4572,
       "step": 17230
     },
     {
-      "epoch": 0.61,
-      "learning_rate": 4.891630744463469e-05,
-      "loss": 0.3233,
+      "epoch": 0.6211482322413234,
+      "grad_norm": 0.16950088739395142,
+      "learning_rate": 4.884579827259939e-05,
+      "loss": 0.4596,
       "step": 17235
     },
     {
-      "epoch": 0.61,
-      "learning_rate": 4.891547764907888e-05,
-      "loss": 0.3337,
+      "epoch": 0.621328431902548,
+      "grad_norm": 0.13479334115982056,
+      "learning_rate": 4.8844921675308985e-05,
+      "loss": 0.4187,
       "step": 17240
     },
     {
-      "epoch": 0.61,
-      "learning_rate": 4.891464754299636e-05,
-      "loss": 0.2812,
+      "epoch": 0.6215086315637727,
+      "grad_norm": 0.1946098804473877,
+      "learning_rate": 4.884404475313614e-05,
+      "loss": 0.4806,
       "step": 17245
     },
     {
-      "epoch": 0.61,
-      "learning_rate": 4.89138171263979e-05,
-      "loss": 0.3134,
+      "epoch": 0.6216888312249973,
+      "grad_norm": 0.20148830115795135,
+      "learning_rate": 4.884316750609281e-05,
+      "loss": 0.4833,
       "step": 17250
     },
     {
-      "epoch": 0.61,
-      "learning_rate": 4.8912986399294285e-05,
-      "loss": 0.3149,
+      "epoch": 0.6218690308862219,
+      "grad_norm": 0.18048211932182312,
+      "learning_rate": 4.884228993419093e-05,
+      "loss": 0.4797,
       "step": 17255
     },
     {
-      "epoch": 0.61,
-      "learning_rate": 4.89121553616963e-05,
-      "loss": 0.3306,
+      "epoch": 0.6220492305474465,
+      "grad_norm": 0.17456752061843872,
+      "learning_rate": 4.884141203744248e-05,
+      "loss": 0.4505,
       "step": 17260
     },
     {
-      "epoch": 0.61,
-      "learning_rate": 4.891132401361473e-05,
-      "loss": 0.3212,
+      "epoch": 0.6222294302086712,
+      "grad_norm": 0.19601954519748688,
+      "learning_rate": 4.884053381585939e-05,
+      "loss": 0.4685,
       "step": 17265
     },
     {
-      "epoch": 0.61,
-      "learning_rate": 4.8910492355060386e-05,
-      "loss": 0.3005,
+      "epoch": 0.6224096298698959,
+      "grad_norm": 0.191552072763443,
+      "learning_rate": 4.883965526945365e-05,
+      "loss": 0.4564,
       "step": 17270
     },
     {
-      "epoch": 0.61,
-      "learning_rate": 4.8909660386044055e-05,
-      "loss": 0.3292,
+      "epoch": 0.6225898295311205,
+      "grad_norm": 0.17397992312908173,
+      "learning_rate": 4.883877639823722e-05,
+      "loss": 0.4623,
       "step": 17275
     },
     {
-      "epoch": 0.61,
-      "learning_rate": 4.890882810657654e-05,
-      "loss": 0.337,
+      "epoch": 0.6227700291923451,
+      "grad_norm": 0.1623370498418808,
+      "learning_rate": 4.8837897202222076e-05,
+      "loss": 0.4449,
       "step": 17280
     },
     {
-      "epoch": 0.61,
-      "learning_rate": 4.890799551666865e-05,
-      "loss": 0.3018,
+      "epoch": 0.6229502288535698,
+      "grad_norm": 0.15014895796775818,
+      "learning_rate": 4.88370176814202e-05,
+      "loss": 0.443,
       "step": 17285
     },
     {
-      "epoch": 0.61,
-      "learning_rate": 4.890716261633119e-05,
-      "loss": 0.3002,
+      "epoch": 0.6231304285147944,
+      "grad_norm": 0.17987573146820068,
+      "learning_rate": 4.8836137835843574e-05,
+      "loss": 0.4186,
       "step": 17290
     },
     {
-      "epoch": 0.61,
-      "learning_rate": 4.890632940557498e-05,
-      "loss": 0.2851,
+      "epoch": 0.623310628176019,
+      "grad_norm": 0.16012683510780334,
+      "learning_rate": 4.8835257665504184e-05,
+      "loss": 0.4429,
       "step": 17295
     },
     {
-      "epoch": 0.61,
-      "learning_rate": 4.890549588441083e-05,
-      "loss": 0.3158,
+      "epoch": 0.6234908278372436,
+      "grad_norm": 0.17883889377117157,
+      "learning_rate": 4.883437717041403e-05,
+      "loss": 0.4408,
       "step": 17300
     },
     {
-      "epoch": 0.61,
-      "learning_rate": 4.8904662052849586e-05,
-      "loss": 0.3095,
+      "epoch": 0.6236710274984683,
+      "grad_norm": 0.17255452275276184,
+      "learning_rate": 4.883349635058508e-05,
+      "loss": 0.4326,
       "step": 17305
     },
     {
-      "epoch": 0.61,
-      "learning_rate": 4.890382791090205e-05,
-      "loss": 0.3175,
+      "epoch": 0.623851227159693,
+      "grad_norm": 0.14830821752548218,
+      "learning_rate": 4.883261520602937e-05,
+      "loss": 0.4488,
       "step": 17310
     },
     {
-      "epoch": 0.61,
-      "learning_rate": 4.890299345857907e-05,
-      "loss": 0.3159,
+      "epoch": 0.6240314268209176,
+      "grad_norm": 0.16953101754188538,
+      "learning_rate": 4.88317337367589e-05,
+      "loss": 0.4381,
       "step": 17315
     },
     {
-      "epoch": 0.61,
-      "learning_rate": 4.8902158695891464e-05,
-      "loss": 0.3371,
+      "epoch": 0.6242116264821422,
+      "grad_norm": 0.16917838156223297,
+      "learning_rate": 4.883085194278566e-05,
+      "loss": 0.4023,
       "step": 17320
     },
     {
-      "epoch": 0.61,
-      "learning_rate": 4.890132362285008e-05,
-      "loss": 0.3326,
+      "epoch": 0.6243918261433669,
+      "grad_norm": 0.14477120339870453,
+      "learning_rate": 4.882996982412168e-05,
+      "loss": 0.416,
       "step": 17325
     },
     {
-      "epoch": 0.61,
-      "learning_rate": 4.890048823946576e-05,
-      "loss": 0.3218,
+      "epoch": 0.6245720258045915,
+      "grad_norm": 0.1996786743402481,
+      "learning_rate": 4.882908738077898e-05,
+      "loss": 0.4539,
       "step": 17330
     },
     {
-      "epoch": 0.61,
-      "learning_rate": 4.889965254574935e-05,
-      "loss": 0.3009,
+      "epoch": 0.6247522254658161,
+      "grad_norm": 0.18695120513439178,
+      "learning_rate": 4.8828204612769566e-05,
+      "loss": 0.4737,
       "step": 17335
     },
     {
-      "epoch": 0.61,
-      "learning_rate": 4.8898816541711706e-05,
-      "loss": 0.3419,
+      "epoch": 0.6249324251270407,
+      "grad_norm": 0.17920991778373718,
+      "learning_rate": 4.882732152010549e-05,
+      "loss": 0.4147,
       "step": 17340
     },
     {
-      "epoch": 0.61,
-      "learning_rate": 4.8897980227363684e-05,
-      "loss": 0.3228,
+      "epoch": 0.6251126247882653,
+      "grad_norm": 0.1538057178258896,
+      "learning_rate": 4.8826438102798766e-05,
+      "loss": 0.4651,
       "step": 17345
     },
     {
-      "epoch": 0.61,
-      "learning_rate": 4.889714360271612e-05,
-      "loss": 0.3057,
+      "epoch": 0.6252928244494901,
+      "grad_norm": 0.17022547125816345,
+      "learning_rate": 4.8825554360861436e-05,
+      "loss": 0.4641,
       "step": 17350
     },
     {
-      "epoch": 0.61,
-      "learning_rate": 4.889630666777991e-05,
-      "loss": 0.2896,
+      "epoch": 0.6254730241107147,
+      "grad_norm": 0.17763908207416534,
+      "learning_rate": 4.882467029430554e-05,
+      "loss": 0.3932,
       "step": 17355
     },
     {
-      "epoch": 0.61,
-      "learning_rate": 4.889546942256591e-05,
-      "loss": 0.3151,
+      "epoch": 0.6256532237719393,
+      "grad_norm": 0.1638226956129074,
+      "learning_rate": 4.8823785903143124e-05,
+      "loss": 0.4439,
       "step": 17360
     },
     {
-      "epoch": 0.61,
-      "learning_rate": 4.889463186708498e-05,
-      "loss": 0.3091,
+      "epoch": 0.625833423433164,
+      "grad_norm": 0.1570916473865509,
+      "learning_rate": 4.882290118738624e-05,
+      "loss": 0.4014,
       "step": 17365
     },
     {
-      "epoch": 0.61,
-      "learning_rate": 4.8893794001348e-05,
-      "loss": 0.3189,
+      "epoch": 0.6260136230943886,
+      "grad_norm": 0.16065819561481476,
+      "learning_rate": 4.882201614704694e-05,
+      "loss": 0.4374,
       "step": 17370
     },
     {
-      "epoch": 0.61,
-      "learning_rate": 4.8892955825365856e-05,
-      "loss": 0.3118,
+      "epoch": 0.6261938227556132,
+      "grad_norm": 0.1514330506324768,
+      "learning_rate": 4.8821130782137275e-05,
+      "loss": 0.4195,
       "step": 17375
     },
     {
-      "epoch": 0.61,
-      "learning_rate": 4.889211733914942e-05,
-      "loss": 0.2891,
+      "epoch": 0.6263740224168378,
+      "grad_norm": 0.14630937576293945,
+      "learning_rate": 4.882024509266932e-05,
+      "loss": 0.4394,
       "step": 17380
     },
     {
-      "epoch": 0.61,
-      "learning_rate": 4.88912785427096e-05,
-      "loss": 0.2917,
+      "epoch": 0.6265542220780624,
+      "grad_norm": 0.15218989551067352,
+      "learning_rate": 4.881935907865514e-05,
+      "loss": 0.4132,
       "step": 17385
     },
     {
-      "epoch": 0.61,
-      "learning_rate": 4.889043943605726e-05,
-      "loss": 0.3112,
+      "epoch": 0.6267344217392872,
+      "grad_norm": 0.20841450989246368,
+      "learning_rate": 4.88184727401068e-05,
+      "loss": 0.4669,
       "step": 17390
     },
     {
-      "epoch": 0.61,
-      "learning_rate": 4.888960001920331e-05,
-      "loss": 0.3095,
+      "epoch": 0.6269146214005118,
+      "grad_norm": 0.18255776166915894,
+      "learning_rate": 4.881758607703638e-05,
+      "loss": 0.4354,
       "step": 17395
     },
     {
-      "epoch": 0.61,
-      "learning_rate": 4.8888760292158654e-05,
-      "loss": 0.3276,
+      "epoch": 0.6270948210617364,
+      "grad_norm": 0.12769578397274017,
+      "learning_rate": 4.881669908945596e-05,
+      "loss": 0.4514,
       "step": 17400
     },
     {
-      "epoch": 0.61,
-      "learning_rate": 4.888792025493418e-05,
-      "loss": 0.3031,
+      "epoch": 0.627275020722961,
+      "grad_norm": 0.18235896527767181,
+      "learning_rate": 4.881581177737763e-05,
+      "loss": 0.4523,
       "step": 17405
     },
     {
-      "epoch": 0.61,
-      "learning_rate": 4.888707990754081e-05,
-      "loss": 0.315,
+      "epoch": 0.6274552203841857,
+      "grad_norm": 0.13113798201084137,
+      "learning_rate": 4.8814924140813466e-05,
+      "loss": 0.4246,
       "step": 17410
     },
     {
-      "epoch": 0.61,
-      "learning_rate": 4.8886239249989455e-05,
-      "loss": 0.3281,
+      "epoch": 0.6276354200454103,
+      "grad_norm": 0.18605348467826843,
+      "learning_rate": 4.881403617977558e-05,
+      "loss": 0.4335,
       "step": 17415
     },
     {
-      "epoch": 0.61,
-      "learning_rate": 4.8885398282291015e-05,
-      "loss": 0.2979,
+      "epoch": 0.6278156197066349,
+      "grad_norm": 0.19018565118312836,
+      "learning_rate": 4.8813147894276065e-05,
+      "loss": 0.4198,
       "step": 17420
     },
     {
-      "epoch": 0.61,
-      "learning_rate": 4.888455700445642e-05,
-      "loss": 0.3312,
+      "epoch": 0.6279958193678596,
+      "grad_norm": 0.22579999268054962,
+      "learning_rate": 4.8812259284327013e-05,
+      "loss": 0.4842,
       "step": 17425
     },
     {
-      "epoch": 0.61,
-      "learning_rate": 4.88837154164966e-05,
-      "loss": 0.3159,
+      "epoch": 0.6281760190290843,
+      "grad_norm": 0.13994662463665009,
+      "learning_rate": 4.8811370349940545e-05,
+      "loss": 0.4695,
       "step": 17430
     },
     {
-      "epoch": 0.61,
-      "learning_rate": 4.888287351842247e-05,
-      "loss": 0.3123,
+      "epoch": 0.6283562186903089,
+      "grad_norm": 0.1737351268529892,
+      "learning_rate": 4.8810481091128756e-05,
+      "loss": 0.4521,
       "step": 17435
     },
     {
-      "epoch": 0.61,
-      "learning_rate": 4.888203131024497e-05,
-      "loss": 0.3222,
+      "epoch": 0.6285364183515335,
+      "grad_norm": 0.1526714414358139,
+      "learning_rate": 4.880959150790379e-05,
+      "loss": 0.428,
       "step": 17440
     },
     {
-      "epoch": 0.61,
-      "learning_rate": 4.888118879197502e-05,
-      "loss": 0.313,
+      "epoch": 0.6287166180127581,
+      "grad_norm": 0.16444048285484314,
+      "learning_rate": 4.880870160027773e-05,
+      "loss": 0.4125,
       "step": 17445
     },
     {
-      "epoch": 0.61,
-      "learning_rate": 4.888034596362359e-05,
-      "loss": 0.3172,
+      "epoch": 0.6288968176739828,
+      "grad_norm": 0.2083633989095688,
+      "learning_rate": 4.8807811368262726e-05,
+      "loss": 0.4616,
       "step": 17450
     },
     {
-      "epoch": 0.61,
-      "learning_rate": 4.887950282520159e-05,
-      "loss": 0.3118,
+      "epoch": 0.6290770173352074,
+      "grad_norm": 0.2831677496433258,
+      "learning_rate": 4.8806920811870905e-05,
+      "loss": 0.4404,
       "step": 17455
     },
     {
-      "epoch": 0.61,
-      "learning_rate": 4.8878659376719995e-05,
-      "loss": 0.2852,
+      "epoch": 0.629257216996432,
+      "grad_norm": 0.17356400191783905,
+      "learning_rate": 4.88060299311144e-05,
+      "loss": 0.4132,
       "step": 17460
     },
     {
-      "epoch": 0.61,
-      "learning_rate": 4.887781561818974e-05,
-      "loss": 0.3064,
+      "epoch": 0.6294374166576567,
+      "grad_norm": 0.1940339356660843,
+      "learning_rate": 4.880513872600534e-05,
+      "loss": 0.422,
       "step": 17465
     },
     {
-      "epoch": 0.61,
-      "learning_rate": 4.8876971549621784e-05,
-      "loss": 0.3038,
+      "epoch": 0.6296176163188814,
+      "grad_norm": 0.15967166423797607,
+      "learning_rate": 4.880424719655588e-05,
+      "loss": 0.4363,
       "step": 17470
     },
     {
-      "epoch": 0.61,
-      "learning_rate": 4.8876127171027084e-05,
-      "loss": 0.3113,
+      "epoch": 0.629797815980106,
+      "grad_norm": 0.16466853022575378,
+      "learning_rate": 4.8803355342778145e-05,
+      "loss": 0.4705,
       "step": 17475
     },
     {
-      "epoch": 0.61,
-      "learning_rate": 4.887528248241662e-05,
-      "loss": 0.3153,
+      "epoch": 0.6299780156413306,
+      "grad_norm": 0.17861898243427277,
+      "learning_rate": 4.880246316468432e-05,
+      "loss": 0.4117,
       "step": 17480
     },
     {
-      "epoch": 0.62,
-      "learning_rate": 4.8874437483801335e-05,
-      "loss": 0.3126,
+      "epoch": 0.6301582153025552,
+      "grad_norm": 0.13602939248085022,
+      "learning_rate": 4.880157066228653e-05,
+      "loss": 0.4148,
       "step": 17485
     },
     {
-      "epoch": 0.62,
-      "learning_rate": 4.887359217519222e-05,
-      "loss": 0.3134,
+      "epoch": 0.6303384149637798,
+      "grad_norm": 0.1881345808506012,
+      "learning_rate": 4.880067783559695e-05,
+      "loss": 0.4365,
       "step": 17490
     },
     {
-      "epoch": 0.62,
-      "learning_rate": 4.887274655660025e-05,
-      "loss": 0.2875,
+      "epoch": 0.6305186146250045,
+      "grad_norm": 0.23560486733913422,
+      "learning_rate": 4.8799784684627745e-05,
+      "loss": 0.459,
       "step": 17495
     },
     {
-      "epoch": 0.62,
-      "learning_rate": 4.887190062803638e-05,
-      "loss": 0.3005,
+      "epoch": 0.6306988142862291,
+      "grad_norm": 0.1441272646188736,
+      "learning_rate": 4.8798891209391085e-05,
+      "loss": 0.4468,
       "step": 17500
     },
     {
-      "epoch": 0.62,
-      "eval_loss": 0.3056062161922455,
-      "eval_runtime": 10.5404,
-      "eval_samples_per_second": 9.487,
-      "eval_steps_per_second": 9.487,
+      "epoch": 0.6306988142862291,
+      "eval_loss": 0.4662560522556305,
+      "eval_runtime": 3.533,
+      "eval_samples_per_second": 28.305,
+      "eval_steps_per_second": 7.076,
       "step": 17500
     },
     {
-      "epoch": 0.62,
-      "learning_rate": 4.8871054389511626e-05,
-      "loss": 0.3079,
+      "epoch": 0.6308790139474538,
+      "grad_norm": 0.18622781336307526,
+      "learning_rate": 4.879799740989914e-05,
+      "loss": 0.4294,
       "step": 17505
     },
     {
-      "epoch": 0.62,
-      "learning_rate": 4.887020784103696e-05,
-      "loss": 0.3054,
+      "epoch": 0.6310592136086784,
+      "grad_norm": 0.17122696340084076,
+      "learning_rate": 4.879710328616408e-05,
+      "loss": 0.463,
       "step": 17510
     },
     {
-      "epoch": 0.62,
-      "learning_rate": 4.886936098262338e-05,
-      "loss": 0.3035,
+      "epoch": 0.6312394132699031,
+      "grad_norm": 0.18917937576770782,
+      "learning_rate": 4.87962088381981e-05,
+      "loss": 0.4626,
       "step": 17515
     },
     {
-      "epoch": 0.62,
-      "learning_rate": 4.8868513814281866e-05,
-      "loss": 0.3326,
+      "epoch": 0.6314196129311277,
+      "grad_norm": 0.1895674765110016,
+      "learning_rate": 4.8795314066013386e-05,
+      "loss": 0.4487,
       "step": 17520
     },
     {
-      "epoch": 0.62,
-      "learning_rate": 4.8867666336023445e-05,
-      "loss": 0.3312,
+      "epoch": 0.6315998125923523,
+      "grad_norm": 0.17945843935012817,
+      "learning_rate": 4.879441896962212e-05,
+      "loss": 0.4573,
       "step": 17525
     },
     {
-      "epoch": 0.62,
-      "learning_rate": 4.8866818547859095e-05,
-      "loss": 0.3105,
+      "epoch": 0.6317800122535769,
+      "grad_norm": 0.16244640946388245,
+      "learning_rate": 4.87935235490365e-05,
+      "loss": 0.415,
       "step": 17530
     },
     {
-      "epoch": 0.62,
-      "learning_rate": 4.886597044979984e-05,
-      "loss": 0.3202,
+      "epoch": 0.6319602119148016,
+      "grad_norm": 0.16950534284114838,
+      "learning_rate": 4.879262780426873e-05,
+      "loss": 0.4423,
       "step": 17535
     },
     {
-      "epoch": 0.62,
-      "learning_rate": 4.886512204185669e-05,
-      "loss": 0.3186,
+      "epoch": 0.6321404115760262,
+      "grad_norm": 0.15685759484767914,
+      "learning_rate": 4.879173173533101e-05,
+      "loss": 0.4512,
       "step": 17540
     },
     {
-      "epoch": 0.62,
-      "learning_rate": 4.886427332404065e-05,
-      "loss": 0.3076,
+      "epoch": 0.6323206112372509,
+      "grad_norm": 0.18144509196281433,
+      "learning_rate": 4.879083534223556e-05,
+      "loss": 0.4618,
       "step": 17545
     },
     {
-      "epoch": 0.62,
-      "learning_rate": 4.886342429636276e-05,
-      "loss": 0.3072,
+      "epoch": 0.6325008108984755,
+      "grad_norm": 0.14772038161754608,
+      "learning_rate": 4.878993862499458e-05,
+      "loss": 0.4673,
       "step": 17550
     },
     {
-      "epoch": 0.62,
-      "learning_rate": 4.8862574958834016e-05,
-      "loss": 0.2953,
+      "epoch": 0.6326810105597002,
+      "grad_norm": 0.17589062452316284,
+      "learning_rate": 4.8789041583620285e-05,
+      "loss": 0.4252,
       "step": 17555
     },
     {
-      "epoch": 0.62,
-      "learning_rate": 4.8861725311465475e-05,
-      "loss": 0.313,
+      "epoch": 0.6328612102209248,
+      "grad_norm": 0.1934499889612198,
+      "learning_rate": 4.878814421812492e-05,
+      "loss": 0.4261,
       "step": 17560
     },
     {
-      "epoch": 0.62,
-      "learning_rate": 4.886087535426815e-05,
-      "loss": 0.3168,
+      "epoch": 0.6330414098821494,
+      "grad_norm": 0.16541697084903717,
+      "learning_rate": 4.878724652852068e-05,
+      "loss": 0.4186,
       "step": 17565
     },
     {
-      "epoch": 0.62,
-      "learning_rate": 4.8860025087253086e-05,
-      "loss": 0.301,
+      "epoch": 0.633221609543374,
+      "grad_norm": 0.18114767968654633,
+      "learning_rate": 4.878634851481982e-05,
+      "loss": 0.4684,
       "step": 17570
     },
     {
-      "epoch": 0.62,
-      "learning_rate": 4.8859174510431314e-05,
-      "loss": 0.3209,
+      "epoch": 0.6334018092045987,
+      "grad_norm": 0.17742811143398285,
+      "learning_rate": 4.878545017703457e-05,
+      "loss": 0.4497,
       "step": 17575
     },
     {
-      "epoch": 0.62,
-      "learning_rate": 4.8858323623813896e-05,
-      "loss": 0.3257,
+      "epoch": 0.6335820088658234,
+      "grad_norm": 0.16837285459041595,
+      "learning_rate": 4.8784551515177154e-05,
+      "loss": 0.4176,
       "step": 17580
     },
     {
-      "epoch": 0.62,
-      "learning_rate": 4.8857472427411864e-05,
-      "loss": 0.3161,
+      "epoch": 0.633762208527048,
+      "grad_norm": 0.14889617264270782,
+      "learning_rate": 4.878365252925984e-05,
+      "loss": 0.4337,
       "step": 17585
     },
     {
-      "epoch": 0.62,
-      "learning_rate": 4.8856620921236276e-05,
-      "loss": 0.2901,
+      "epoch": 0.6339424081882726,
+      "grad_norm": 0.1742192506790161,
+      "learning_rate": 4.878275321929485e-05,
+      "loss": 0.4535,
       "step": 17590
     },
     {
-      "epoch": 0.62,
-      "learning_rate": 4.8855769105298185e-05,
-      "loss": 0.3008,
+      "epoch": 0.6341226078494973,
+      "grad_norm": 0.20734204351902008,
+      "learning_rate": 4.878185358529447e-05,
+      "loss": 0.4681,
       "step": 17595
     },
     {
-      "epoch": 0.62,
-      "learning_rate": 4.8854916979608656e-05,
-      "loss": 0.3401,
+      "epoch": 0.6343028075107219,
+      "grad_norm": 0.16586579382419586,
+      "learning_rate": 4.8780953627270924e-05,
+      "loss": 0.4205,
       "step": 17600
     },
     {
-      "epoch": 0.62,
-      "learning_rate": 4.885406454417875e-05,
-      "loss": 0.3126,
+      "epoch": 0.6344830071719465,
+      "grad_norm": 0.1419735550880432,
+      "learning_rate": 4.8780053345236496e-05,
+      "loss": 0.4481,
       "step": 17605
     },
     {
-      "epoch": 0.62,
-      "learning_rate": 4.8853211799019534e-05,
-      "loss": 0.3105,
+      "epoch": 0.6346632068331711,
+      "grad_norm": 0.1975310742855072,
+      "learning_rate": 4.877915273920345e-05,
+      "loss": 0.4853,
       "step": 17610
     },
     {
-      "epoch": 0.62,
-      "learning_rate": 4.8852358744142085e-05,
-      "loss": 0.3052,
+      "epoch": 0.6348434064943957,
+      "grad_norm": 0.19924567639827728,
+      "learning_rate": 4.8778251809184045e-05,
+      "loss": 0.4531,
       "step": 17615
     },
     {
-      "epoch": 0.62,
-      "learning_rate": 4.8851505379557475e-05,
-      "loss": 0.3248,
+      "epoch": 0.6350236061556205,
+      "grad_norm": 0.17732134461402893,
+      "learning_rate": 4.877735055519056e-05,
+      "loss": 0.4414,
       "step": 17620
     },
     {
-      "epoch": 0.62,
-      "learning_rate": 4.885065170527679e-05,
-      "loss": 0.3126,
+      "epoch": 0.6352038058168451,
+      "grad_norm": 0.2488788664340973,
+      "learning_rate": 4.877644897723528e-05,
+      "loss": 0.4517,
       "step": 17625
     },
     {
-      "epoch": 0.62,
-      "learning_rate": 4.8849797721311106e-05,
-      "loss": 0.3261,
+      "epoch": 0.6353840054780697,
+      "grad_norm": 0.1604602336883545,
+      "learning_rate": 4.877554707533049e-05,
+      "loss": 0.4141,
       "step": 17630
     },
     {
-      "epoch": 0.62,
-      "learning_rate": 4.8848943427671526e-05,
-      "loss": 0.3238,
+      "epoch": 0.6355642051392943,
+      "grad_norm": 0.1665060818195343,
+      "learning_rate": 4.877464484948847e-05,
+      "loss": 0.4129,
       "step": 17635
     },
     {
-      "epoch": 0.62,
-      "learning_rate": 4.884808882436912e-05,
-      "loss": 0.3123,
+      "epoch": 0.635744404800519,
+      "grad_norm": 0.17682184278964996,
+      "learning_rate": 4.877374229972152e-05,
+      "loss": 0.4501,
       "step": 17640
     },
     {
-      "epoch": 0.62,
-      "learning_rate": 4.8847233911415e-05,
-      "loss": 0.3282,
+      "epoch": 0.6359246044617436,
+      "grad_norm": 0.17669498920440674,
+      "learning_rate": 4.8772839426041935e-05,
+      "loss": 0.4363,
       "step": 17645
     },
     {
-      "epoch": 0.62,
-      "learning_rate": 4.884637868882027e-05,
-      "loss": 0.3058,
+      "epoch": 0.6361048041229682,
+      "grad_norm": 0.17283748090267181,
+      "learning_rate": 4.877193622846201e-05,
+      "loss": 0.4117,
       "step": 17650
     },
     {
-      "epoch": 0.62,
-      "learning_rate": 4.884552315659603e-05,
-      "loss": 0.3324,
+      "epoch": 0.6362850037841928,
+      "grad_norm": 0.18358486890792847,
+      "learning_rate": 4.877103270699406e-05,
+      "loss": 0.4261,
       "step": 17655
     },
     {
-      "epoch": 0.62,
-      "learning_rate": 4.884466731475338e-05,
-      "loss": 0.3031,
+      "epoch": 0.6364652034454176,
+      "grad_norm": 0.15538670122623444,
+      "learning_rate": 4.877012886165039e-05,
+      "loss": 0.4387,
       "step": 17660
     },
     {
-      "epoch": 0.62,
-      "learning_rate": 4.884381116330343e-05,
-      "loss": 0.3154,
+      "epoch": 0.6366454031066422,
+      "grad_norm": 0.17753304541110992,
+      "learning_rate": 4.8769224692443315e-05,
+      "loss": 0.4024,
       "step": 17665
     },
     {
-      "epoch": 0.62,
-      "learning_rate": 4.8842954702257324e-05,
-      "loss": 0.3194,
+      "epoch": 0.6368256027678668,
+      "grad_norm": 0.20612354576587677,
+      "learning_rate": 4.8768320199385166e-05,
+      "loss": 0.465,
       "step": 17670
     },
     {
-      "epoch": 0.62,
-      "learning_rate": 4.884209793162615e-05,
-      "loss": 0.3127,
+      "epoch": 0.6370058024290914,
+      "grad_norm": 0.21157251298427582,
+      "learning_rate": 4.876741538248825e-05,
+      "loss": 0.4686,
       "step": 17675
     },
     {
-      "epoch": 0.62,
-      "learning_rate": 4.884124085142106e-05,
-      "loss": 0.3082,
+      "epoch": 0.6371860020903161,
+      "grad_norm": 0.22163233160972595,
+      "learning_rate": 4.876651024176489e-05,
+      "loss": 0.4685,
       "step": 17680
     },
     {
-      "epoch": 0.62,
-      "learning_rate": 4.8840383461653164e-05,
-      "loss": 0.3087,
+      "epoch": 0.6373662017515407,
+      "grad_norm": 0.1729961782693863,
+      "learning_rate": 4.8765604777227434e-05,
+      "loss": 0.4498,
       "step": 17685
     },
     {
-      "epoch": 0.62,
-      "learning_rate": 4.883952576233359e-05,
-      "loss": 0.3167,
+      "epoch": 0.6375464014127653,
+      "grad_norm": 0.16511957347393036,
+      "learning_rate": 4.8764698988888223e-05,
+      "loss": 0.4789,
       "step": 17690
     },
     {
-      "epoch": 0.62,
-      "learning_rate": 4.8838667753473496e-05,
-      "loss": 0.3426,
+      "epoch": 0.6377266010739899,
+      "grad_norm": 0.15577971935272217,
+      "learning_rate": 4.876379287675958e-05,
+      "loss": 0.4499,
       "step": 17695
     },
     {
-      "epoch": 0.62,
-      "learning_rate": 4.8837809435084016e-05,
-      "loss": 0.3001,
+      "epoch": 0.6379068007352147,
+      "grad_norm": 0.23532330989837646,
+      "learning_rate": 4.8762886440853865e-05,
+      "loss": 0.458,
       "step": 17700
     },
     {
-      "epoch": 0.62,
-      "learning_rate": 4.8836950807176276e-05,
-      "loss": 0.2905,
+      "epoch": 0.6380870003964393,
+      "grad_norm": 0.1516086459159851,
+      "learning_rate": 4.876197968118342e-05,
+      "loss": 0.4136,
       "step": 17705
     },
     {
-      "epoch": 0.62,
-      "learning_rate": 4.8836091869761444e-05,
-      "loss": 0.3193,
+      "epoch": 0.6382672000576639,
+      "grad_norm": 0.16963382065296173,
+      "learning_rate": 4.8761072597760604e-05,
+      "loss": 0.4761,
       "step": 17710
     },
     {
-      "epoch": 0.62,
-      "learning_rate": 4.883523262285068e-05,
-      "loss": 0.3067,
+      "epoch": 0.6384473997188885,
+      "grad_norm": 0.17492809891700745,
+      "learning_rate": 4.876016519059778e-05,
+      "loss": 0.4299,
       "step": 17715
     },
     {
-      "epoch": 0.62,
-      "learning_rate": 4.8834373066455116e-05,
-      "loss": 0.3145,
+      "epoch": 0.6386275993801132,
+      "grad_norm": 0.21216866374015808,
+      "learning_rate": 4.87592574597073e-05,
+      "loss": 0.4399,
       "step": 17720
     },
     {
-      "epoch": 0.62,
-      "learning_rate": 4.883351320058592e-05,
-      "loss": 0.3245,
+      "epoch": 0.6388077990413378,
+      "grad_norm": 0.1838618665933609,
+      "learning_rate": 4.8758349405101535e-05,
+      "loss": 0.4794,
       "step": 17725
     },
     {
-      "epoch": 0.62,
-      "learning_rate": 4.883265302525427e-05,
-      "loss": 0.3328,
+      "epoch": 0.6389879987025624,
+      "grad_norm": 0.20086225867271423,
+      "learning_rate": 4.875744102679286e-05,
+      "loss": 0.4475,
       "step": 17730
     },
     {
-      "epoch": 0.62,
-      "learning_rate": 4.8831792540471334e-05,
-      "loss": 0.3123,
+      "epoch": 0.639168198363787,
+      "grad_norm": 0.18566478788852692,
+      "learning_rate": 4.8756532324793646e-05,
+      "loss": 0.4394,
       "step": 17735
     },
     {
-      "epoch": 0.62,
-      "learning_rate": 4.883093174624826e-05,
-      "loss": 0.3107,
+      "epoch": 0.6393483980250118,
+      "grad_norm": 0.1914960891008377,
+      "learning_rate": 4.875562329911629e-05,
+      "loss": 0.4574,
       "step": 17740
     },
     {
-      "epoch": 0.62,
-      "learning_rate": 4.8830070642596256e-05,
-      "loss": 0.3077,
+      "epoch": 0.6395285976862364,
+      "grad_norm": 0.14925505220890045,
+      "learning_rate": 4.8754713949773156e-05,
+      "loss": 0.4712,
       "step": 17745
     },
     {
-      "epoch": 0.62,
-      "learning_rate": 4.882920922952648e-05,
-      "loss": 0.3102,
+      "epoch": 0.639708797347461,
+      "grad_norm": 0.1491568684577942,
+      "learning_rate": 4.875380427677665e-05,
+      "loss": 0.4545,
       "step": 17750
     },
     {
-      "epoch": 0.62,
-      "learning_rate": 4.882834750705014e-05,
-      "loss": 0.3418,
+      "epoch": 0.6398889970086856,
+      "grad_norm": 0.16431868076324463,
+      "learning_rate": 4.8752894280139155e-05,
+      "loss": 0.4363,
       "step": 17755
     },
     {
-      "epoch": 0.62,
-      "learning_rate": 4.88274854751784e-05,
-      "loss": 0.2913,
+      "epoch": 0.6400691966699102,
+      "grad_norm": 0.16594944894313812,
+      "learning_rate": 4.8751983959873074e-05,
+      "loss": 0.443,
       "step": 17760
     },
     {
-      "epoch": 0.63,
-      "learning_rate": 4.882662313392247e-05,
-      "loss": 0.3132,
+      "epoch": 0.6402493963311349,
+      "grad_norm": 0.17833413183689117,
+      "learning_rate": 4.875107331599081e-05,
+      "loss": 0.4451,
       "step": 17765
     },
     {
-      "epoch": 0.63,
-      "learning_rate": 4.8825760483293533e-05,
-      "loss": 0.2979,
+      "epoch": 0.6404295959923595,
+      "grad_norm": 0.1914624124765396,
+      "learning_rate": 4.875016234850478e-05,
+      "loss": 0.4021,
       "step": 17770
     },
     {
-      "epoch": 0.63,
-      "learning_rate": 4.88248975233028e-05,
-      "loss": 0.3141,
+      "epoch": 0.6406097956535842,
+      "grad_norm": 0.16454005241394043,
+      "learning_rate": 4.8749251057427374e-05,
+      "loss": 0.4073,
       "step": 17775
     },
     {
-      "epoch": 0.63,
-      "learning_rate": 4.882403425396148e-05,
-      "loss": 0.3054,
+      "epoch": 0.6407899953148088,
+      "grad_norm": 0.1951947659254074,
+      "learning_rate": 4.874833944277103e-05,
+      "loss": 0.5004,
       "step": 17780
     },
     {
-      "epoch": 0.63,
-      "learning_rate": 4.882317067528077e-05,
-      "loss": 0.328,
+      "epoch": 0.6409701949760335,
+      "grad_norm": 0.1762334704399109,
+      "learning_rate": 4.8747427504548146e-05,
+      "loss": 0.4096,
       "step": 17785
     },
     {
-      "epoch": 0.63,
-      "learning_rate": 4.88223067872719e-05,
-      "loss": 0.3238,
+      "epoch": 0.6411503946372581,
+      "grad_norm": 0.18907704949378967,
+      "learning_rate": 4.8746515242771175e-05,
+      "loss": 0.4241,
       "step": 17790
     },
     {
-      "epoch": 0.63,
-      "learning_rate": 4.882144258994606e-05,
-      "loss": 0.3282,
+      "epoch": 0.6413305942984827,
+      "grad_norm": 0.20052656531333923,
+      "learning_rate": 4.874560265745252e-05,
+      "loss": 0.4246,
       "step": 17795
     },
     {
-      "epoch": 0.63,
-      "learning_rate": 4.88205780833145e-05,
-      "loss": 0.3001,
+      "epoch": 0.6415107939597073,
+      "grad_norm": 0.19393053650856018,
+      "learning_rate": 4.874468974860463e-05,
+      "loss": 0.4644,
       "step": 17800
     },
     {
-      "epoch": 0.63,
-      "learning_rate": 4.8819713267388425e-05,
-      "loss": 0.307,
+      "epoch": 0.641690993620932,
+      "grad_norm": 0.2143370658159256,
+      "learning_rate": 4.874377651623994e-05,
+      "loss": 0.462,
       "step": 17805
     },
     {
-      "epoch": 0.63,
-      "learning_rate": 4.881884814217907e-05,
-      "loss": 0.3345,
+      "epoch": 0.6418711932821566,
+      "grad_norm": 0.20955999195575714,
+      "learning_rate": 4.8742862960370895e-05,
+      "loss": 0.4356,
       "step": 17810
     },
     {
-      "epoch": 0.63,
-      "learning_rate": 4.881798270769767e-05,
-      "loss": 0.3091,
+      "epoch": 0.6420513929433813,
+      "grad_norm": 0.1920044720172882,
+      "learning_rate": 4.874194908100993e-05,
+      "loss": 0.4294,
       "step": 17815
     },
     {
-      "epoch": 0.63,
-      "learning_rate": 4.8817116963955465e-05,
-      "loss": 0.2985,
+      "epoch": 0.6422315926046059,
+      "grad_norm": 0.20551706850528717,
+      "learning_rate": 4.874103487816951e-05,
+      "loss": 0.4574,
       "step": 17820
     },
     {
-      "epoch": 0.63,
-      "learning_rate": 4.881625091096369e-05,
-      "loss": 0.2963,
+      "epoch": 0.6424117922658306,
+      "grad_norm": 0.1954600214958191,
+      "learning_rate": 4.8740120351862087e-05,
+      "loss": 0.439,
       "step": 17825
     },
     {
-      "epoch": 0.63,
-      "learning_rate": 4.881538454873359e-05,
-      "loss": 0.34,
+      "epoch": 0.6425919919270552,
+      "grad_norm": 0.17379389703273773,
+      "learning_rate": 4.873920550210012e-05,
+      "loss": 0.4056,
       "step": 17830
     },
     {
-      "epoch": 0.63,
-      "learning_rate": 4.881451787727642e-05,
-      "loss": 0.3245,
+      "epoch": 0.6427721915882798,
+      "grad_norm": 0.17289118468761444,
+      "learning_rate": 4.8738290328896076e-05,
+      "loss": 0.4321,
       "step": 17835
     },
     {
-      "epoch": 0.63,
-      "learning_rate": 4.881365089660342e-05,
-      "loss": 0.3092,
+      "epoch": 0.6429523912495044,
+      "grad_norm": 0.181446373462677,
+      "learning_rate": 4.8737374832262415e-05,
+      "loss": 0.4323,
       "step": 17840
     },
     {
-      "epoch": 0.63,
-      "learning_rate": 4.881278360672587e-05,
-      "loss": 0.3129,
+      "epoch": 0.643132590910729,
+      "grad_norm": 0.25358590483665466,
+      "learning_rate": 4.8736459012211624e-05,
+      "loss": 0.4085,
       "step": 17845
     },
     {
-      "epoch": 0.63,
-      "learning_rate": 4.881191600765501e-05,
-      "loss": 0.3273,
+      "epoch": 0.6433127905719537,
+      "grad_norm": 0.23851445317268372,
+      "learning_rate": 4.873554286875618e-05,
+      "loss": 0.4697,
       "step": 17850
     },
     {
-      "epoch": 0.63,
-      "learning_rate": 4.8811048099402124e-05,
-      "loss": 0.305,
+      "epoch": 0.6434929902331784,
+      "grad_norm": 0.18585999310016632,
+      "learning_rate": 4.873462640190855e-05,
+      "loss": 0.457,
       "step": 17855
     },
     {
-      "epoch": 0.63,
-      "learning_rate": 4.8810179881978455e-05,
-      "loss": 0.2934,
+      "epoch": 0.643673189894403,
+      "grad_norm": 0.16106140613555908,
+      "learning_rate": 4.873370961168123e-05,
+      "loss": 0.449,
       "step": 17860
     },
     {
-      "epoch": 0.63,
-      "learning_rate": 4.88093113553953e-05,
-      "loss": 0.3206,
+      "epoch": 0.6438533895556277,
+      "grad_norm": 0.18756996095180511,
+      "learning_rate": 4.873279249808672e-05,
+      "loss": 0.4517,
       "step": 17865
     },
     {
-      "epoch": 0.63,
-      "learning_rate": 4.8808442519663936e-05,
-      "loss": 0.3365,
+      "epoch": 0.6440335892168523,
+      "grad_norm": 0.26352357864379883,
+      "learning_rate": 4.87318750611375e-05,
+      "loss": 0.4494,
       "step": 17870
     },
     {
-      "epoch": 0.63,
-      "learning_rate": 4.880757337479562e-05,
-      "loss": 0.3166,
+      "epoch": 0.6442137888780769,
+      "grad_norm": 0.18100088834762573,
+      "learning_rate": 4.873095730084608e-05,
+      "loss": 0.442,
       "step": 17875
     },
     {
-      "epoch": 0.63,
-      "learning_rate": 4.8806703920801666e-05,
-      "loss": 0.3241,
+      "epoch": 0.6443939885393015,
+      "grad_norm": 0.1749761700630188,
+      "learning_rate": 4.873003921722496e-05,
+      "loss": 0.4248,
       "step": 17880
     },
     {
-      "epoch": 0.63,
-      "learning_rate": 4.880583415769334e-05,
-      "loss": 0.299,
+      "epoch": 0.6445741882005261,
+      "grad_norm": 0.16620847582817078,
+      "learning_rate": 4.8729120810286654e-05,
+      "loss": 0.4626,
       "step": 17885
     },
     {
-      "epoch": 0.63,
-      "learning_rate": 4.880496408548196e-05,
-      "loss": 0.3023,
+      "epoch": 0.6447543878617508,
+      "grad_norm": 0.18539084494113922,
+      "learning_rate": 4.872820208004367e-05,
+      "loss": 0.42,
       "step": 17890
     },
     {
-      "epoch": 0.63,
-      "learning_rate": 4.88040937041788e-05,
-      "loss": 0.3261,
+      "epoch": 0.6449345875229755,
+      "grad_norm": 0.15285968780517578,
+      "learning_rate": 4.8727283026508516e-05,
+      "loss": 0.428,
       "step": 17895
     },
     {
-      "epoch": 0.63,
-      "learning_rate": 4.8803223013795175e-05,
-      "loss": 0.3378,
+      "epoch": 0.6451147871842001,
+      "grad_norm": 0.16171573102474213,
+      "learning_rate": 4.872636364969373e-05,
+      "loss": 0.4222,
       "step": 17900
     },
     {
-      "epoch": 0.63,
-      "learning_rate": 4.8802352014342376e-05,
-      "loss": 0.2997,
+      "epoch": 0.6452949868454247,
+      "grad_norm": 0.2132413238286972,
+      "learning_rate": 4.872544394961184e-05,
+      "loss": 0.4037,
       "step": 17905
     },
     {
-      "epoch": 0.63,
-      "learning_rate": 4.880148070583174e-05,
-      "loss": 0.3037,
+      "epoch": 0.6454751865066494,
+      "grad_norm": 0.16865713894367218,
+      "learning_rate": 4.872452392627537e-05,
+      "loss": 0.4025,
       "step": 17910
     },
     {
-      "epoch": 0.63,
-      "learning_rate": 4.880060908827455e-05,
-      "loss": 0.3039,
+      "epoch": 0.645655386167874,
+      "grad_norm": 0.18928374350070953,
+      "learning_rate": 4.8723603579696844e-05,
+      "loss": 0.4316,
       "step": 17915
     },
     {
-      "epoch": 0.63,
-      "learning_rate": 4.879973716168214e-05,
-      "loss": 0.2984,
+      "epoch": 0.6458355858290986,
+      "grad_norm": 0.14977779984474182,
+      "learning_rate": 4.872268290988882e-05,
+      "loss": 0.4344,
       "step": 17920
     },
     {
-      "epoch": 0.63,
-      "learning_rate": 4.8798864926065823e-05,
-      "loss": 0.3049,
+      "epoch": 0.6460157854903232,
+      "grad_norm": 0.15273813903331757,
+      "learning_rate": 4.8721761916863826e-05,
+      "loss": 0.4632,
       "step": 17925
     },
     {
-      "epoch": 0.63,
-      "learning_rate": 4.879799238143693e-05,
-      "loss": 0.2784,
+      "epoch": 0.646195985151548,
+      "grad_norm": 0.16625703871250153,
+      "learning_rate": 4.872084060063443e-05,
+      "loss": 0.4079,
       "step": 17930
     },
     {
-      "epoch": 0.63,
-      "learning_rate": 4.879711952780679e-05,
-      "loss": 0.2891,
+      "epoch": 0.6463761848127726,
+      "grad_norm": 0.16598527133464813,
+      "learning_rate": 4.871991896121317e-05,
+      "loss": 0.4384,
       "step": 17935
     },
     {
-      "epoch": 0.63,
-      "learning_rate": 4.879624636518674e-05,
-      "loss": 0.297,
+      "epoch": 0.6465563844739972,
+      "grad_norm": 0.14382204413414001,
+      "learning_rate": 4.8718996998612605e-05,
+      "loss": 0.4601,
       "step": 17940
     },
     {
-      "epoch": 0.63,
-      "learning_rate": 4.879537289358811e-05,
-      "loss": 0.2995,
+      "epoch": 0.6467365841352218,
+      "grad_norm": 0.169216588139534,
+      "learning_rate": 4.8718074712845296e-05,
+      "loss": 0.4329,
       "step": 17945
     },
     {
-      "epoch": 0.63,
-      "learning_rate": 4.879449911302225e-05,
-      "loss": 0.3221,
+      "epoch": 0.6469167837964465,
+      "grad_norm": 0.21951742470264435,
+      "learning_rate": 4.871715210392381e-05,
+      "loss": 0.4817,
       "step": 17950
     },
     {
-      "epoch": 0.63,
-      "learning_rate": 4.879362502350049e-05,
-      "loss": 0.3157,
+      "epoch": 0.6470969834576711,
+      "grad_norm": 0.14136020839214325,
+      "learning_rate": 4.871622917186073e-05,
+      "loss": 0.4133,
       "step": 17955
     },
     {
-      "epoch": 0.63,
-      "learning_rate": 4.87927506250342e-05,
-      "loss": 0.3068,
+      "epoch": 0.6472771831188957,
+      "grad_norm": 0.22290566563606262,
+      "learning_rate": 4.871530591666861e-05,
+      "loss": 0.4538,
       "step": 17960
     },
     {
-      "epoch": 0.63,
-      "learning_rate": 4.879187591763472e-05,
-      "loss": 0.3526,
+      "epoch": 0.6474573827801203,
+      "grad_norm": 0.1538752168416977,
+      "learning_rate": 4.871438233836003e-05,
+      "loss": 0.4566,
       "step": 17965
     },
     {
-      "epoch": 0.63,
-      "learning_rate": 4.879100090131341e-05,
-      "loss": 0.2967,
+      "epoch": 0.6476375824413451,
+      "grad_norm": 0.13945698738098145,
+      "learning_rate": 4.871345843694759e-05,
+      "loss": 0.4536,
       "step": 17970
     },
     {
-      "epoch": 0.63,
-      "learning_rate": 4.879012557608162e-05,
-      "loss": 0.2943,
+      "epoch": 0.6478177821025697,
+      "grad_norm": 0.19078098237514496,
+      "learning_rate": 4.871253421244387e-05,
+      "loss": 0.4743,
       "step": 17975
     },
     {
-      "epoch": 0.63,
-      "learning_rate": 4.8789249941950746e-05,
-      "loss": 0.32,
+      "epoch": 0.6479979817637943,
+      "grad_norm": 0.17093034088611603,
+      "learning_rate": 4.871160966486147e-05,
+      "loss": 0.4625,
       "step": 17980
     },
     {
-      "epoch": 0.63,
-      "learning_rate": 4.878837399893213e-05,
-      "loss": 0.334,
+      "epoch": 0.6481781814250189,
+      "grad_norm": 0.20973354578018188,
+      "learning_rate": 4.8710684794212966e-05,
+      "loss": 0.4308,
       "step": 17985
     },
     {
-      "epoch": 0.63,
-      "learning_rate": 4.8787497747037156e-05,
-      "loss": 0.3192,
+      "epoch": 0.6483583810862436,
+      "grad_norm": 0.19713053107261658,
+      "learning_rate": 4.870975960051098e-05,
+      "loss": 0.4244,
       "step": 17990
     },
     {
-      "epoch": 0.63,
-      "learning_rate": 4.8786621186277204e-05,
-      "loss": 0.2938,
+      "epoch": 0.6485385807474682,
+      "grad_norm": 0.18588168919086456,
+      "learning_rate": 4.8708834083768106e-05,
+      "loss": 0.4199,
       "step": 17995
     },
     {
-      "epoch": 0.63,
-      "learning_rate": 4.878574431666365e-05,
-      "loss": 0.314,
+      "epoch": 0.6487187804086928,
+      "grad_norm": 0.15173941850662231,
+      "learning_rate": 4.870790824399696e-05,
+      "loss": 0.3994,
       "step": 18000
     },
     {
-      "epoch": 0.63,
-      "eval_loss": 0.3048493266105652,
-      "eval_runtime": 10.5346,
-      "eval_samples_per_second": 9.493,
-      "eval_steps_per_second": 9.493,
+      "epoch": 0.6487187804086928,
+      "eval_loss": 0.4651120901107788,
+      "eval_runtime": 3.5665,
+      "eval_samples_per_second": 28.038,
+      "eval_steps_per_second": 7.01,
       "step": 18000
     },
     {
-      "epoch": 0.63,
-      "learning_rate": 4.8784867138207877e-05,
-      "loss": 0.2992,
+      "epoch": 0.6488989800699174,
+      "grad_norm": 0.1769859790802002,
+      "learning_rate": 4.870698208121015e-05,
+      "loss": 0.4507,
       "step": 18005
     },
     {
-      "epoch": 0.63,
-      "learning_rate": 4.878398965092129e-05,
-      "loss": 0.3117,
+      "epoch": 0.6490791797311422,
+      "grad_norm": 0.19280551373958588,
+      "learning_rate": 4.870605559542031e-05,
+      "loss": 0.4139,
       "step": 18010
     },
     {
-      "epoch": 0.63,
-      "learning_rate": 4.878311185481526e-05,
-      "loss": 0.3081,
+      "epoch": 0.6492593793923668,
+      "grad_norm": 0.1606101095676422,
+      "learning_rate": 4.870512878664004e-05,
+      "loss": 0.4447,
       "step": 18015
     },
     {
-      "epoch": 0.63,
-      "learning_rate": 4.8782233749901205e-05,
-      "loss": 0.3108,
+      "epoch": 0.6494395790535914,
+      "grad_norm": 0.16047056019306183,
+      "learning_rate": 4.870420165488199e-05,
+      "loss": 0.4453,
       "step": 18020
     },
     {
-      "epoch": 0.63,
-      "learning_rate": 4.878135533619051e-05,
-      "loss": 0.3003,
+      "epoch": 0.649619778714816,
+      "grad_norm": 0.1686955988407135,
+      "learning_rate": 4.8703274200158765e-05,
+      "loss": 0.4609,
       "step": 18025
     },
     {
-      "epoch": 0.63,
-      "learning_rate": 4.8780476613694594e-05,
-      "loss": 0.3065,
+      "epoch": 0.6497999783760406,
+      "grad_norm": 0.15635234117507935,
+      "learning_rate": 4.8702346422483035e-05,
+      "loss": 0.4296,
       "step": 18030
     },
     {
-      "epoch": 0.63,
-      "learning_rate": 4.877959758242486e-05,
-      "loss": 0.307,
+      "epoch": 0.6499801780372653,
+      "grad_norm": 0.158394455909729,
+      "learning_rate": 4.870141832186742e-05,
+      "loss": 0.447,
       "step": 18035
     },
     {
-      "epoch": 0.63,
-      "learning_rate": 4.877871824239272e-05,
-      "loss": 0.3209,
+      "epoch": 0.6501603776984899,
+      "grad_norm": 0.16858071088790894,
+      "learning_rate": 4.870048989832456e-05,
+      "loss": 0.4233,
       "step": 18040
     },
     {
-      "epoch": 0.63,
-      "learning_rate": 4.877783859360959e-05,
-      "loss": 0.2935,
+      "epoch": 0.6503405773597145,
+      "grad_norm": 0.11584287881851196,
+      "learning_rate": 4.869956115186712e-05,
+      "loss": 0.3873,
       "step": 18045
     },
     {
-      "epoch": 0.64,
-      "learning_rate": 4.87769586360869e-05,
-      "loss": 0.3417,
+      "epoch": 0.6505207770209392,
+      "grad_norm": 0.15428020060062408,
+      "learning_rate": 4.8698632082507754e-05,
+      "loss": 0.4426,
       "step": 18050
     },
     {
-      "epoch": 0.64,
-      "learning_rate": 4.877607836983608e-05,
-      "loss": 0.3528,
+      "epoch": 0.6507009766821639,
+      "grad_norm": 0.2866957187652588,
+      "learning_rate": 4.869770269025911e-05,
+      "loss": 0.4639,
       "step": 18055
     },
     {
-      "epoch": 0.64,
-      "learning_rate": 4.8775197794868544e-05,
-      "loss": 0.3162,
+      "epoch": 0.6508811763433885,
+      "grad_norm": 0.21071985363960266,
+      "learning_rate": 4.8696772975133854e-05,
+      "loss": 0.4842,
       "step": 18060
     },
     {
-      "epoch": 0.64,
-      "learning_rate": 4.8774316911195736e-05,
-      "loss": 0.302,
+      "epoch": 0.6510613760046131,
+      "grad_norm": 0.17984484136104584,
+      "learning_rate": 4.869584293714465e-05,
+      "loss": 0.4272,
       "step": 18065
     },
     {
-      "epoch": 0.64,
-      "learning_rate": 4.877343571882909e-05,
-      "loss": 0.3097,
+      "epoch": 0.6512415756658377,
+      "grad_norm": 0.20826445519924164,
+      "learning_rate": 4.8694912576304175e-05,
+      "loss": 0.4561,
       "step": 18070
     },
     {
-      "epoch": 0.64,
-      "learning_rate": 4.877255421778004e-05,
-      "loss": 0.3214,
+      "epoch": 0.6514217753270624,
+      "grad_norm": 0.14839668571949005,
+      "learning_rate": 4.8693981892625105e-05,
+      "loss": 0.445,
       "step": 18075
     },
     {
-      "epoch": 0.64,
-      "learning_rate": 4.8771672408060056e-05,
-      "loss": 0.3215,
+      "epoch": 0.651601974988287,
+      "grad_norm": 0.15876393020153046,
+      "learning_rate": 4.869305088612012e-05,
+      "loss": 0.4377,
       "step": 18080
     },
     {
-      "epoch": 0.64,
-      "learning_rate": 4.8770790289680555e-05,
-      "loss": 0.321,
+      "epoch": 0.6517821746495117,
+      "grad_norm": 0.20614424347877502,
+      "learning_rate": 4.8692119556801905e-05,
+      "loss": 0.4499,
       "step": 18085
     },
     {
-      "epoch": 0.64,
-      "learning_rate": 4.876990786265302e-05,
-      "loss": 0.3224,
+      "epoch": 0.6519623743107363,
+      "grad_norm": 0.17227613925933838,
+      "learning_rate": 4.8691187904683145e-05,
+      "loss": 0.4511,
       "step": 18090
     },
     {
-      "epoch": 0.64,
-      "learning_rate": 4.8769025126988896e-05,
-      "loss": 0.3153,
+      "epoch": 0.652142573971961,
+      "grad_norm": 0.15382865071296692,
+      "learning_rate": 4.869025592977654e-05,
+      "loss": 0.4734,
       "step": 18095
     },
     {
-      "epoch": 0.64,
-      "learning_rate": 4.876814208269964e-05,
-      "loss": 0.33,
+      "epoch": 0.6523227736331856,
+      "grad_norm": 0.1578940749168396,
+      "learning_rate": 4.8689323632094784e-05,
+      "loss": 0.4353,
       "step": 18100
     },
     {
-      "epoch": 0.64,
-      "learning_rate": 4.8767258729796725e-05,
-      "loss": 0.3092,
+      "epoch": 0.6525029732944102,
+      "grad_norm": 0.17838072776794434,
+      "learning_rate": 4.868839101165058e-05,
+      "loss": 0.4476,
       "step": 18105
     },
     {
-      "epoch": 0.64,
-      "learning_rate": 4.876637506829162e-05,
-      "loss": 0.3139,
+      "epoch": 0.6526831729556348,
+      "grad_norm": 0.16446293890476227,
+      "learning_rate": 4.868745806845663e-05,
+      "loss": 0.4486,
       "step": 18110
     },
     {
-      "epoch": 0.64,
-      "learning_rate": 4.876549109819579e-05,
-      "loss": 0.3209,
+      "epoch": 0.6528633726168594,
+      "grad_norm": 0.16984452307224274,
+      "learning_rate": 4.8686524802525656e-05,
+      "loss": 0.4135,
       "step": 18115
     },
     {
-      "epoch": 0.64,
-      "learning_rate": 4.876460681952073e-05,
-      "loss": 0.2928,
+      "epoch": 0.6530435722780841,
+      "grad_norm": 0.18492572009563446,
+      "learning_rate": 4.868559121387036e-05,
+      "loss": 0.465,
       "step": 18120
     },
     {
-      "epoch": 0.64,
-      "learning_rate": 4.876372223227791e-05,
-      "loss": 0.2933,
+      "epoch": 0.6532237719393088,
+      "grad_norm": 0.18226665258407593,
+      "learning_rate": 4.868465730250348e-05,
+      "loss": 0.452,
       "step": 18125
     },
     {
-      "epoch": 0.64,
-      "learning_rate": 4.876283733647881e-05,
-      "loss": 0.3381,
+      "epoch": 0.6534039716005334,
+      "grad_norm": 0.18231280148029327,
+      "learning_rate": 4.868372306843772e-05,
+      "loss": 0.4465,
       "step": 18130
     },
     {
-      "epoch": 0.64,
-      "learning_rate": 4.876195213213494e-05,
-      "loss": 0.331,
+      "epoch": 0.653584171261758,
+      "grad_norm": 0.22582538425922394,
+      "learning_rate": 4.868278851168582e-05,
+      "loss": 0.4061,
       "step": 18135
     },
     {
-      "epoch": 0.64,
-      "learning_rate": 4.876106661925777e-05,
-      "loss": 0.3399,
+      "epoch": 0.6537643709229827,
+      "grad_norm": 0.16542305052280426,
+      "learning_rate": 4.8681853632260524e-05,
+      "loss": 0.4386,
       "step": 18140
     },
     {
-      "epoch": 0.64,
-      "learning_rate": 4.8760180797858815e-05,
-      "loss": 0.3081,
+      "epoch": 0.6539445705842073,
+      "grad_norm": 0.21034276485443115,
+      "learning_rate": 4.868091843017454e-05,
+      "loss": 0.4536,
       "step": 18145
     },
     {
-      "epoch": 0.64,
-      "learning_rate": 4.8759294667949565e-05,
-      "loss": 0.3229,
+      "epoch": 0.6541247702454319,
+      "grad_norm": 0.20631127059459686,
+      "learning_rate": 4.867998290544064e-05,
+      "loss": 0.4516,
       "step": 18150
     },
     {
-      "epoch": 0.64,
-      "learning_rate": 4.875840822954154e-05,
-      "loss": 0.3046,
+      "epoch": 0.6543049699066565,
+      "grad_norm": 0.1725880652666092,
+      "learning_rate": 4.867904705807155e-05,
+      "loss": 0.4615,
       "step": 18155
     },
     {
-      "epoch": 0.64,
-      "learning_rate": 4.875752148264624e-05,
-      "loss": 0.3098,
+      "epoch": 0.6544851695678812,
+      "grad_norm": 0.14972399175167084,
+      "learning_rate": 4.8678110888080026e-05,
+      "loss": 0.4562,
       "step": 18160
     },
     {
-      "epoch": 0.64,
-      "learning_rate": 4.875663442727517e-05,
-      "loss": 0.3071,
+      "epoch": 0.6546653692291059,
+      "grad_norm": 0.1854199767112732,
+      "learning_rate": 4.8677174395478834e-05,
+      "loss": 0.4723,
       "step": 18165
     },
     {
-      "epoch": 0.64,
-      "learning_rate": 4.875574706343987e-05,
-      "loss": 0.305,
+      "epoch": 0.6548455688903305,
+      "grad_norm": 0.16778244078159332,
+      "learning_rate": 4.867623758028072e-05,
+      "loss": 0.4472,
       "step": 18170
     },
     {
-      "epoch": 0.64,
-      "learning_rate": 4.8754859391151844e-05,
-      "loss": 0.2952,
+      "epoch": 0.6550257685515551,
+      "grad_norm": 0.14149196445941925,
+      "learning_rate": 4.867530044249845e-05,
+      "loss": 0.4036,
       "step": 18175
     },
     {
-      "epoch": 0.64,
-      "learning_rate": 4.875397141042263e-05,
-      "loss": 0.3284,
+      "epoch": 0.6552059682127798,
+      "grad_norm": 0.15938983857631683,
+      "learning_rate": 4.86743629821448e-05,
+      "loss": 0.431,
       "step": 18180
     },
     {
-      "epoch": 0.64,
-      "learning_rate": 4.875308312126374e-05,
-      "loss": 0.3252,
+      "epoch": 0.6553861678740044,
+      "grad_norm": 0.16331376135349274,
+      "learning_rate": 4.867342519923254e-05,
+      "loss": 0.4418,
       "step": 18185
     },
     {
-      "epoch": 0.64,
-      "learning_rate": 4.8752194523686726e-05,
-      "loss": 0.2949,
+      "epoch": 0.655566367535229,
+      "grad_norm": 0.15382960438728333,
+      "learning_rate": 4.867248709377443e-05,
+      "loss": 0.4471,
       "step": 18190
     },
     {
-      "epoch": 0.64,
-      "learning_rate": 4.875130561770312e-05,
-      "loss": 0.3216,
+      "epoch": 0.6557465671964536,
+      "grad_norm": 0.1844521015882492,
+      "learning_rate": 4.8671548665783287e-05,
+      "loss": 0.4734,
       "step": 18195
     },
     {
-      "epoch": 0.64,
-      "learning_rate": 4.875041640332446e-05,
-      "loss": 0.2878,
+      "epoch": 0.6559267668576783,
+      "grad_norm": 0.21769674122333527,
+      "learning_rate": 4.867060991527186e-05,
+      "loss": 0.4561,
       "step": 18200
     },
     {
-      "epoch": 0.64,
-      "learning_rate": 4.87495268805623e-05,
-      "loss": 0.2957,
+      "epoch": 0.656106966518903,
+      "grad_norm": 0.21248985826969147,
+      "learning_rate": 4.866967084225297e-05,
+      "loss": 0.4137,
       "step": 18205
     },
     {
-      "epoch": 0.64,
-      "learning_rate": 4.8748637049428185e-05,
-      "loss": 0.3343,
+      "epoch": 0.6562871661801276,
+      "grad_norm": 0.18385855853557587,
+      "learning_rate": 4.866873144673939e-05,
+      "loss": 0.4094,
       "step": 18210
     },
     {
-      "epoch": 0.64,
-      "learning_rate": 4.874774690993367e-05,
-      "loss": 0.2956,
+      "epoch": 0.6564673658413522,
+      "grad_norm": 0.13242194056510925,
+      "learning_rate": 4.866779172874392e-05,
+      "loss": 0.4741,
       "step": 18215
     },
     {
-      "epoch": 0.64,
-      "learning_rate": 4.8746856462090305e-05,
-      "loss": 0.3084,
+      "epoch": 0.6566475655025769,
+      "grad_norm": 0.17850540578365326,
+      "learning_rate": 4.866685168827938e-05,
+      "loss": 0.4539,
       "step": 18220
     },
     {
-      "epoch": 0.64,
-      "learning_rate": 4.874596570590967e-05,
-      "loss": 0.3132,
+      "epoch": 0.6568277651638015,
+      "grad_norm": 0.1694021075963974,
+      "learning_rate": 4.8665911325358556e-05,
+      "loss": 0.4531,
       "step": 18225
     },
     {
-      "epoch": 0.64,
-      "learning_rate": 4.874507464140331e-05,
-      "loss": 0.3079,
+      "epoch": 0.6570079648250261,
+      "grad_norm": 0.13871118426322937,
+      "learning_rate": 4.866497063999429e-05,
+      "loss": 0.4149,
       "step": 18230
     },
     {
-      "epoch": 0.64,
-      "learning_rate": 4.8744183268582804e-05,
-      "loss": 0.3015,
+      "epoch": 0.6571881644862507,
+      "grad_norm": 0.13802644610404968,
+      "learning_rate": 4.866402963219937e-05,
+      "loss": 0.4539,
       "step": 18235
     },
     {
-      "epoch": 0.64,
-      "learning_rate": 4.874329158745973e-05,
-      "loss": 0.3592,
+      "epoch": 0.6573683641474753,
+      "grad_norm": 0.17566382884979248,
+      "learning_rate": 4.8663088301986625e-05,
+      "loss": 0.4297,
       "step": 18240
     },
     {
-      "epoch": 0.64,
-      "learning_rate": 4.874239959804566e-05,
-      "loss": 0.3334,
+      "epoch": 0.6575485638087001,
+      "grad_norm": 0.21962518990039825,
+      "learning_rate": 4.866214664936889e-05,
+      "loss": 0.4451,
       "step": 18245
     },
     {
-      "epoch": 0.64,
-      "learning_rate": 4.874150730035218e-05,
-      "loss": 0.3284,
+      "epoch": 0.6577287634699247,
+      "grad_norm": 0.12450151145458221,
+      "learning_rate": 4.8661204674358984e-05,
+      "loss": 0.4547,
       "step": 18250
     },
     {
-      "epoch": 0.64,
-      "learning_rate": 4.874061469439087e-05,
-      "loss": 0.3008,
+      "epoch": 0.6579089631311493,
+      "grad_norm": 0.16591274738311768,
+      "learning_rate": 4.8660262376969745e-05,
+      "loss": 0.4615,
       "step": 18255
     },
     {
-      "epoch": 0.64,
-      "learning_rate": 4.873972178017333e-05,
-      "loss": 0.317,
+      "epoch": 0.658089162792374,
+      "grad_norm": 0.1811162382364273,
+      "learning_rate": 4.8659319757214016e-05,
+      "loss": 0.443,
       "step": 18260
     },
     {
-      "epoch": 0.64,
-      "learning_rate": 4.873882855771115e-05,
-      "loss": 0.3178,
+      "epoch": 0.6582693624535986,
+      "grad_norm": 0.17516489326953888,
+      "learning_rate": 4.865837681510463e-05,
+      "loss": 0.4069,
       "step": 18265
     },
     {
-      "epoch": 0.64,
-      "learning_rate": 4.8737935027015915e-05,
-      "loss": 0.3101,
+      "epoch": 0.6584495621148232,
+      "grad_norm": 0.16720084846019745,
+      "learning_rate": 4.865743355065444e-05,
+      "loss": 0.4127,
       "step": 18270
     },
     {
-      "epoch": 0.64,
-      "learning_rate": 4.873704118809924e-05,
-      "loss": 0.3289,
+      "epoch": 0.6586297617760478,
+      "grad_norm": 0.13735070824623108,
+      "learning_rate": 4.8656489963876305e-05,
+      "loss": 0.4367,
       "step": 18275
     },
     {
-      "epoch": 0.64,
-      "learning_rate": 4.8736147040972736e-05,
-      "loss": 0.2987,
+      "epoch": 0.6588099614372726,
+      "grad_norm": 0.15320982038974762,
+      "learning_rate": 4.865554605478308e-05,
+      "loss": 0.4592,
       "step": 18280
     },
     {
-      "epoch": 0.64,
-      "learning_rate": 4.8735252585647995e-05,
-      "loss": 0.3261,
+      "epoch": 0.6589901610984972,
+      "grad_norm": 0.2053997814655304,
+      "learning_rate": 4.8654601823387605e-05,
+      "loss": 0.4266,
       "step": 18285
     },
     {
-      "epoch": 0.64,
-      "learning_rate": 4.873435782213665e-05,
-      "loss": 0.3093,
+      "epoch": 0.6591703607597218,
+      "grad_norm": 0.1573738306760788,
+      "learning_rate": 4.8653657269702765e-05,
+      "loss": 0.4257,
       "step": 18290
     },
     {
-      "epoch": 0.64,
-      "learning_rate": 4.8733462750450306e-05,
-      "loss": 0.284,
+      "epoch": 0.6593505604209464,
+      "grad_norm": 0.1637280136346817,
+      "learning_rate": 4.865271239374142e-05,
+      "loss": 0.4559,
       "step": 18295
     },
     {
-      "epoch": 0.64,
-      "learning_rate": 4.873256737060059e-05,
-      "loss": 0.3229,
+      "epoch": 0.659530760082171,
+      "grad_norm": 0.2082606703042984,
+      "learning_rate": 4.865176719551645e-05,
+      "loss": 0.4498,
       "step": 18300
     },
     {
-      "epoch": 0.64,
-      "learning_rate": 4.8731671682599126e-05,
-      "loss": 0.3229,
+      "epoch": 0.6597109597433957,
+      "grad_norm": 0.1640523225069046,
+      "learning_rate": 4.8650821675040735e-05,
+      "loss": 0.4573,
       "step": 18305
     },
     {
-      "epoch": 0.64,
-      "learning_rate": 4.873077568645755e-05,
-      "loss": 0.3249,
+      "epoch": 0.6598911594046203,
+      "grad_norm": 0.130737766623497,
+      "learning_rate": 4.864987583232715e-05,
+      "loss": 0.4277,
       "step": 18310
     },
     {
-      "epoch": 0.64,
-      "learning_rate": 4.8729879382187485e-05,
-      "loss": 0.3393,
+      "epoch": 0.6600713590658449,
+      "grad_norm": 0.17735107243061066,
+      "learning_rate": 4.8648929667388596e-05,
+      "loss": 0.4758,
       "step": 18315
     },
     {
-      "epoch": 0.64,
-      "learning_rate": 4.872898276980058e-05,
-      "loss": 0.33,
+      "epoch": 0.6602515587270696,
+      "grad_norm": 0.180114284157753,
+      "learning_rate": 4.8647983180237945e-05,
+      "loss": 0.4557,
       "step": 18320
     },
     {
-      "epoch": 0.64,
-      "learning_rate": 4.872808584930847e-05,
-      "loss": 0.3207,
+      "epoch": 0.6604317583882943,
+      "grad_norm": 0.19766369462013245,
+      "learning_rate": 4.8647036370888096e-05,
+      "loss": 0.4227,
       "step": 18325
     },
     {
-      "epoch": 0.64,
-      "learning_rate": 4.87271886207228e-05,
-      "loss": 0.2978,
+      "epoch": 0.6606119580495189,
+      "grad_norm": 0.18029169738292694,
+      "learning_rate": 4.8646089239351966e-05,
+      "loss": 0.4243,
       "step": 18330
     },
     {
-      "epoch": 0.65,
-      "learning_rate": 4.872629108405522e-05,
-      "loss": 0.3273,
+      "epoch": 0.6607921577107435,
+      "grad_norm": 0.1884010136127472,
+      "learning_rate": 4.8645141785642444e-05,
+      "loss": 0.469,
       "step": 18335
     },
     {
-      "epoch": 0.65,
-      "learning_rate": 4.872539323931739e-05,
-      "loss": 0.2903,
+      "epoch": 0.6609723573719681,
+      "grad_norm": 0.17342260479927063,
+      "learning_rate": 4.864419400977244e-05,
+      "loss": 0.4331,
       "step": 18340
     },
     {
-      "epoch": 0.65,
-      "learning_rate": 4.8724495086520964e-05,
-      "loss": 0.3063,
+      "epoch": 0.6611525570331928,
+      "grad_norm": 0.15091748535633087,
+      "learning_rate": 4.864324591175487e-05,
+      "loss": 0.4205,
       "step": 18345
     },
     {
-      "epoch": 0.65,
-      "learning_rate": 4.872359662567761e-05,
-      "loss": 0.3177,
+      "epoch": 0.6613327566944174,
+      "grad_norm": 0.20120559632778168,
+      "learning_rate": 4.8642297491602654e-05,
+      "loss": 0.441,
       "step": 18350
     },
     {
-      "epoch": 0.65,
-      "learning_rate": 4.8722697856798985e-05,
-      "loss": 0.293,
+      "epoch": 0.661512956355642,
+      "grad_norm": 0.2035011202096939,
+      "learning_rate": 4.8641348749328716e-05,
+      "loss": 0.4648,
       "step": 18355
     },
     {
-      "epoch": 0.65,
-      "learning_rate": 4.8721798779896756e-05,
-      "loss": 0.3201,
+      "epoch": 0.6616931560168667,
+      "grad_norm": 0.1479252576828003,
+      "learning_rate": 4.864039968494597e-05,
+      "loss": 0.4378,
       "step": 18360
     },
     {
-      "epoch": 0.65,
-      "learning_rate": 4.872089939498261e-05,
-      "loss": 0.3389,
+      "epoch": 0.6618733556780914,
+      "grad_norm": 0.15773577988147736,
+      "learning_rate": 4.863945029846736e-05,
+      "loss": 0.4031,
       "step": 18365
     },
     {
-      "epoch": 0.65,
-      "learning_rate": 4.871999970206822e-05,
-      "loss": 0.297,
+      "epoch": 0.662053555339316,
+      "grad_norm": 0.20971988141536713,
+      "learning_rate": 4.863850058990582e-05,
+      "loss": 0.4773,
       "step": 18370
     },
     {
-      "epoch": 0.65,
-      "learning_rate": 4.871909970116526e-05,
-      "loss": 0.3382,
+      "epoch": 0.6622337550005406,
+      "grad_norm": 0.17060470581054688,
+      "learning_rate": 4.863755055927428e-05,
+      "loss": 0.4178,
       "step": 18375
     },
     {
-      "epoch": 0.65,
-      "learning_rate": 4.871819939228542e-05,
-      "loss": 0.3001,
+      "epoch": 0.6624139546617652,
+      "grad_norm": 0.22001157701015472,
+      "learning_rate": 4.863660020658569e-05,
+      "loss": 0.4637,
       "step": 18380
     },
     {
-      "epoch": 0.65,
-      "learning_rate": 4.871729877544039e-05,
-      "loss": 0.3313,
+      "epoch": 0.6625941543229898,
+      "grad_norm": 0.15214625000953674,
+      "learning_rate": 4.8635649531853e-05,
+      "loss": 0.4174,
       "step": 18385
     },
     {
-      "epoch": 0.65,
-      "learning_rate": 4.871639785064187e-05,
-      "loss": 0.3017,
+      "epoch": 0.6627743539842145,
+      "grad_norm": 0.15511523187160492,
+      "learning_rate": 4.863469853508916e-05,
+      "loss": 0.4156,
       "step": 18390
     },
     {
-      "epoch": 0.65,
-      "learning_rate": 4.871549661790156e-05,
-      "loss": 0.3424,
+      "epoch": 0.6629545536454391,
+      "grad_norm": 0.18479692935943604,
+      "learning_rate": 4.863374721630713e-05,
+      "loss": 0.4064,
       "step": 18395
     },
     {
-      "epoch": 0.65,
-      "learning_rate": 4.871459507723115e-05,
-      "loss": 0.3293,
+      "epoch": 0.6631347533066638,
+      "grad_norm": 0.3170281946659088,
+      "learning_rate": 4.8632795575519876e-05,
+      "loss": 0.4339,
       "step": 18400
     },
     {
-      "epoch": 0.65,
-      "learning_rate": 4.8713693228642356e-05,
-      "loss": 0.3195,
+      "epoch": 0.6633149529678884,
+      "grad_norm": 0.1785048097372055,
+      "learning_rate": 4.863184361274035e-05,
+      "loss": 0.4805,
       "step": 18405
     },
     {
-      "epoch": 0.65,
-      "learning_rate": 4.871279107214687e-05,
-      "loss": 0.3149,
+      "epoch": 0.6634951526291131,
+      "grad_norm": 0.1605699211359024,
+      "learning_rate": 4.863089132798153e-05,
+      "loss": 0.4641,
       "step": 18410
     },
     {
-      "epoch": 0.65,
-      "learning_rate": 4.871188860775643e-05,
-      "loss": 0.3221,
+      "epoch": 0.6636753522903377,
+      "grad_norm": 0.20723554491996765,
+      "learning_rate": 4.86299387212564e-05,
+      "loss": 0.4315,
       "step": 18415
     },
     {
-      "epoch": 0.65,
-      "learning_rate": 4.871098583548274e-05,
-      "loss": 0.2928,
+      "epoch": 0.6638555519515623,
+      "grad_norm": 0.14896468818187714,
+      "learning_rate": 4.862898579257792e-05,
+      "loss": 0.4462,
       "step": 18420
     },
     {
-      "epoch": 0.65,
-      "learning_rate": 4.871008275533753e-05,
-      "loss": 0.2977,
+      "epoch": 0.6640357516127869,
+      "grad_norm": 0.16965442895889282,
+      "learning_rate": 4.8628032541959086e-05,
+      "loss": 0.4679,
       "step": 18425
     },
     {
-      "epoch": 0.65,
-      "learning_rate": 4.8709179367332514e-05,
-      "loss": 0.3271,
+      "epoch": 0.6642159512740116,
+      "grad_norm": 0.1498372107744217,
+      "learning_rate": 4.8627078969412885e-05,
+      "loss": 0.43,
       "step": 18430
     },
     {
-      "epoch": 0.65,
-      "learning_rate": 4.8708275671479434e-05,
-      "loss": 0.3082,
+      "epoch": 0.6643961509352363,
+      "grad_norm": 0.1531539112329483,
+      "learning_rate": 4.86261250749523e-05,
+      "loss": 0.4028,
       "step": 18435
     },
     {
-      "epoch": 0.65,
-      "learning_rate": 4.8707371667790016e-05,
-      "loss": 0.3163,
+      "epoch": 0.6645763505964609,
+      "grad_norm": 0.1347551792860031,
+      "learning_rate": 4.862517085859034e-05,
+      "loss": 0.3987,
       "step": 18440
     },
     {
-      "epoch": 0.65,
-      "learning_rate": 4.8706467356276e-05,
-      "loss": 0.2826,
+      "epoch": 0.6647565502576855,
+      "grad_norm": 0.2017221748828888,
+      "learning_rate": 4.862421632034e-05,
+      "loss": 0.4778,
       "step": 18445
     },
     {
-      "epoch": 0.65,
-      "learning_rate": 4.870556273694913e-05,
-      "loss": 0.3069,
+      "epoch": 0.6649367499189102,
+      "grad_norm": 0.16347235441207886,
+      "learning_rate": 4.8623261460214284e-05,
+      "loss": 0.4565,
       "step": 18450
     },
     {
-      "epoch": 0.65,
-      "learning_rate": 4.8704657809821154e-05,
-      "loss": 0.3196,
+      "epoch": 0.6651169495801348,
+      "grad_norm": 0.18009866774082184,
+      "learning_rate": 4.862230627822621e-05,
+      "loss": 0.4521,
       "step": 18455
     },
     {
-      "epoch": 0.65,
-      "learning_rate": 4.8703752574903814e-05,
-      "loss": 0.3224,
+      "epoch": 0.6652971492413594,
+      "grad_norm": 0.18370333313941956,
+      "learning_rate": 4.8621350774388784e-05,
+      "loss": 0.4353,
       "step": 18460
     },
     {
-      "epoch": 0.65,
-      "learning_rate": 4.870284703220887e-05,
-      "loss": 0.331,
+      "epoch": 0.665477348902584,
+      "grad_norm": 0.17444929480552673,
+      "learning_rate": 4.8620394948715034e-05,
+      "loss": 0.4615,
       "step": 18465
     },
     {
-      "epoch": 0.65,
-      "learning_rate": 4.870194118174807e-05,
-      "loss": 0.2804,
+      "epoch": 0.6656575485638087,
+      "grad_norm": 0.17793411016464233,
+      "learning_rate": 4.8619438801217966e-05,
+      "loss": 0.4321,
       "step": 18470
     },
     {
-      "epoch": 0.65,
-      "learning_rate": 4.8701035023533193e-05,
-      "loss": 0.3272,
+      "epoch": 0.6658377482250334,
+      "grad_norm": 0.1900653839111328,
+      "learning_rate": 4.861848233191062e-05,
+      "loss": 0.4412,
       "step": 18475
     },
     {
-      "epoch": 0.65,
-      "learning_rate": 4.8700128557576e-05,
-      "loss": 0.2817,
+      "epoch": 0.666017947886258,
+      "grad_norm": 0.191162571310997,
+      "learning_rate": 4.861752554080602e-05,
+      "loss": 0.433,
       "step": 18480
     },
     {
-      "epoch": 0.65,
-      "learning_rate": 4.869922178388824e-05,
-      "loss": 0.301,
+      "epoch": 0.6661981475474826,
+      "grad_norm": 0.1730828732252121,
+      "learning_rate": 4.861656842791722e-05,
+      "loss": 0.457,
       "step": 18485
     },
     {
-      "epoch": 0.65,
-      "learning_rate": 4.869831470248172e-05,
-      "loss": 0.3247,
+      "epoch": 0.6663783472087073,
+      "grad_norm": 0.1662280112504959,
+      "learning_rate": 4.861561099325723e-05,
+      "loss": 0.4409,
       "step": 18490
     },
     {
-      "epoch": 0.65,
-      "learning_rate": 4.869740731336819e-05,
-      "loss": 0.3042,
+      "epoch": 0.6665585468699319,
+      "grad_norm": 0.16253672540187836,
+      "learning_rate": 4.8614653236839125e-05,
+      "loss": 0.4723,
       "step": 18495
     },
     {
-      "epoch": 0.65,
-      "learning_rate": 4.869649961655946e-05,
-      "loss": 0.302,
+      "epoch": 0.6667387465311565,
+      "grad_norm": 0.17181946337223053,
+      "learning_rate": 4.861369515867594e-05,
+      "loss": 0.4322,
       "step": 18500
     },
     {
-      "epoch": 0.65,
-      "eval_loss": 0.3054150938987732,
-      "eval_runtime": 10.5494,
-      "eval_samples_per_second": 9.479,
-      "eval_steps_per_second": 9.479,
+      "epoch": 0.6667387465311565,
+      "eval_loss": 0.4643363654613495,
+      "eval_runtime": 3.5289,
+      "eval_samples_per_second": 28.337,
+      "eval_steps_per_second": 7.084,
       "step": 18500
     },
     {
-      "epoch": 0.65,
-      "learning_rate": 4.869559161206728e-05,
-      "loss": 0.3064,
+      "epoch": 0.6669189461923811,
+      "grad_norm": 0.2049456685781479,
+      "learning_rate": 4.861273675878073e-05,
+      "loss": 0.4311,
       "step": 18505
     },
     {
-      "epoch": 0.65,
-      "learning_rate": 4.869468329990347e-05,
-      "loss": 0.3255,
+      "epoch": 0.6670991458536057,
+      "grad_norm": 0.1761494129896164,
+      "learning_rate": 4.861177803716655e-05,
+      "loss": 0.4485,
       "step": 18510
     },
     {
-      "epoch": 0.65,
-      "learning_rate": 4.8693774680079806e-05,
-      "loss": 0.3148,
+      "epoch": 0.6672793455148305,
+      "grad_norm": 0.17736728489398956,
+      "learning_rate": 4.861081899384647e-05,
+      "loss": 0.4739,
       "step": 18515
     },
     {
-      "epoch": 0.65,
-      "learning_rate": 4.86928657526081e-05,
-      "loss": 0.302,
+      "epoch": 0.6674595451760551,
+      "grad_norm": 0.19387714564800262,
+      "learning_rate": 4.860985962883355e-05,
+      "loss": 0.4827,
       "step": 18520
     },
     {
-      "epoch": 0.65,
-      "learning_rate": 4.869195651750014e-05,
-      "loss": 0.3206,
+      "epoch": 0.6676397448372797,
+      "grad_norm": 0.16782864928245544,
+      "learning_rate": 4.860889994214086e-05,
+      "loss": 0.4753,
       "step": 18525
     },
     {
-      "epoch": 0.65,
-      "learning_rate": 4.869104697476774e-05,
-      "loss": 0.3158,
+      "epoch": 0.6678199444985043,
+      "grad_norm": 0.17974188923835754,
+      "learning_rate": 4.860793993378149e-05,
+      "loss": 0.4338,
       "step": 18530
     },
     {
-      "epoch": 0.65,
-      "learning_rate": 4.869013712442271e-05,
-      "loss": 0.2796,
+      "epoch": 0.668000144159729,
+      "grad_norm": 0.1873011589050293,
+      "learning_rate": 4.86069796037685e-05,
+      "loss": 0.4763,
       "step": 18535
     },
     {
-      "epoch": 0.65,
-      "learning_rate": 4.868922696647685e-05,
-      "loss": 0.321,
+      "epoch": 0.6681803438209536,
+      "grad_norm": 0.1606231927871704,
+      "learning_rate": 4.8606018952114985e-05,
+      "loss": 0.4097,
       "step": 18540
     },
     {
-      "epoch": 0.65,
-      "learning_rate": 4.8688316500942e-05,
-      "loss": 0.3074,
+      "epoch": 0.6683605434821782,
+      "grad_norm": 0.19824917614459991,
+      "learning_rate": 4.8605057978834034e-05,
+      "loss": 0.4387,
       "step": 18545
     },
     {
-      "epoch": 0.65,
-      "learning_rate": 4.868740572782997e-05,
-      "loss": 0.2912,
+      "epoch": 0.6685407431434028,
+      "grad_norm": 0.16622477769851685,
+      "learning_rate": 4.860409668393873e-05,
+      "loss": 0.4453,
       "step": 18550
     },
     {
-      "epoch": 0.65,
-      "learning_rate": 4.868649464715259e-05,
-      "loss": 0.3335,
+      "epoch": 0.6687209428046276,
+      "grad_norm": 0.1473122388124466,
+      "learning_rate": 4.8603135067442184e-05,
+      "loss": 0.4171,
       "step": 18555
     },
     {
-      "epoch": 0.65,
-      "learning_rate": 4.868558325892168e-05,
-      "loss": 0.3184,
+      "epoch": 0.6689011424658522,
+      "grad_norm": 0.15606799721717834,
+      "learning_rate": 4.8602173129357496e-05,
+      "loss": 0.4392,
       "step": 18560
     },
     {
-      "epoch": 0.65,
-      "learning_rate": 4.8684671563149095e-05,
-      "loss": 0.3027,
+      "epoch": 0.6690813421270768,
+      "grad_norm": 0.2004939466714859,
+      "learning_rate": 4.8601210869697766e-05,
+      "loss": 0.4659,
       "step": 18565
     },
     {
-      "epoch": 0.65,
-      "learning_rate": 4.868375955984665e-05,
-      "loss": 0.3115,
+      "epoch": 0.6692615417883014,
+      "grad_norm": 0.17140799760818481,
+      "learning_rate": 4.8600248288476105e-05,
+      "loss": 0.4405,
       "step": 18570
     },
     {
-      "epoch": 0.65,
-      "learning_rate": 4.868284724902619e-05,
-      "loss": 0.3293,
+      "epoch": 0.6694417414495261,
+      "grad_norm": 0.20640818774700165,
+      "learning_rate": 4.859928538570564e-05,
+      "loss": 0.4663,
       "step": 18575
     },
     {
-      "epoch": 0.65,
-      "learning_rate": 4.868193463069958e-05,
-      "loss": 0.3297,
+      "epoch": 0.6696219411107507,
+      "grad_norm": 0.15983538329601288,
+      "learning_rate": 4.859832216139947e-05,
+      "loss": 0.4411,
       "step": 18580
     },
     {
-      "epoch": 0.65,
-      "learning_rate": 4.8681021704878634e-05,
-      "loss": 0.3219,
+      "epoch": 0.6698021407719753,
+      "grad_norm": 0.18739761412143707,
+      "learning_rate": 4.8597358615570734e-05,
+      "loss": 0.4453,
       "step": 18585
     },
     {
-      "epoch": 0.65,
-      "learning_rate": 4.868010847157525e-05,
-      "loss": 0.3026,
+      "epoch": 0.6699823404332,
+      "grad_norm": 0.2296430617570877,
+      "learning_rate": 4.8596394748232556e-05,
+      "loss": 0.4441,
       "step": 18590
     },
     {
-      "epoch": 0.65,
-      "learning_rate": 4.867919493080125e-05,
-      "loss": 0.3164,
+      "epoch": 0.6701625400944247,
+      "grad_norm": 0.1832958459854126,
+      "learning_rate": 4.8595430559398056e-05,
+      "loss": 0.4621,
       "step": 18595
     },
     {
-      "epoch": 0.65,
-      "learning_rate": 4.867828108256851e-05,
-      "loss": 0.3117,
+      "epoch": 0.6703427397556493,
+      "grad_norm": 0.2158745378255844,
+      "learning_rate": 4.85944660490804e-05,
+      "loss": 0.4708,
       "step": 18600
     },
     {
-      "epoch": 0.65,
-      "learning_rate": 4.86773669268889e-05,
-      "loss": 0.2914,
+      "epoch": 0.6705229394168739,
+      "grad_norm": 0.16886022686958313,
+      "learning_rate": 4.85935012172927e-05,
+      "loss": 0.4215,
       "step": 18605
     },
     {
-      "epoch": 0.65,
-      "learning_rate": 4.867645246377428e-05,
-      "loss": 0.3312,
+      "epoch": 0.6707031390780985,
+      "grad_norm": 0.15665309131145477,
+      "learning_rate": 4.859253606404812e-05,
+      "loss": 0.4263,
       "step": 18610
     },
     {
-      "epoch": 0.65,
-      "learning_rate": 4.867553769323653e-05,
-      "loss": 0.3089,
+      "epoch": 0.6708833387393232,
+      "grad_norm": 0.1781836897134781,
+      "learning_rate": 4.85915705893598e-05,
+      "loss": 0.4506,
       "step": 18615
     },
     {
-      "epoch": 0.66,
-      "learning_rate": 4.867462261528752e-05,
-      "loss": 0.3158,
+      "epoch": 0.6710635384005478,
+      "grad_norm": 0.18428124487400055,
+      "learning_rate": 4.85906047932409e-05,
+      "loss": 0.4865,
       "step": 18620
     },
     {
-      "epoch": 0.66,
-      "learning_rate": 4.867370722993915e-05,
-      "loss": 0.3053,
+      "epoch": 0.6712437380617724,
+      "grad_norm": 0.1660621613264084,
+      "learning_rate": 4.858963867570458e-05,
+      "loss": 0.4609,
       "step": 18625
     },
     {
-      "epoch": 0.66,
-      "learning_rate": 4.8672791537203286e-05,
-      "loss": 0.3143,
+      "epoch": 0.6714239377229971,
+      "grad_norm": 0.1916099488735199,
+      "learning_rate": 4.8588672236764e-05,
+      "loss": 0.4376,
       "step": 18630
     },
     {
-      "epoch": 0.66,
-      "learning_rate": 4.8671875537091824e-05,
-      "loss": 0.333,
+      "epoch": 0.6716041373842218,
+      "grad_norm": 0.1742095947265625,
+      "learning_rate": 4.858770547643232e-05,
+      "loss": 0.4813,
       "step": 18635
     },
     {
-      "epoch": 0.66,
-      "learning_rate": 4.867095922961666e-05,
-      "loss": 0.2965,
+      "epoch": 0.6717843370454464,
+      "grad_norm": 0.1986844837665558,
+      "learning_rate": 4.858673839472273e-05,
+      "loss": 0.447,
       "step": 18640
     },
     {
-      "epoch": 0.66,
-      "learning_rate": 4.867004261478969e-05,
-      "loss": 0.2995,
+      "epoch": 0.671964536706671,
+      "grad_norm": 0.168625608086586,
+      "learning_rate": 4.858577099164839e-05,
+      "loss": 0.4426,
       "step": 18645
     },
     {
-      "epoch": 0.66,
-      "learning_rate": 4.866912569262282e-05,
-      "loss": 0.3155,
+      "epoch": 0.6721447363678956,
+      "grad_norm": 0.18472012877464294,
+      "learning_rate": 4.858480326722249e-05,
+      "loss": 0.4643,
       "step": 18650
     },
     {
-      "epoch": 0.66,
-      "learning_rate": 4.866820846312794e-05,
-      "loss": 0.3045,
+      "epoch": 0.6723249360291202,
+      "grad_norm": 0.1810387521982193,
+      "learning_rate": 4.858383522145821e-05,
+      "loss": 0.4517,
       "step": 18655
     },
     {
-      "epoch": 0.66,
-      "learning_rate": 4.866729092631698e-05,
-      "loss": 0.2974,
+      "epoch": 0.6725051356903449,
+      "grad_norm": 0.18452127277851105,
+      "learning_rate": 4.858286685436873e-05,
+      "loss": 0.4554,
       "step": 18660
     },
     {
-      "epoch": 0.66,
-      "learning_rate": 4.866637308220184e-05,
-      "loss": 0.3212,
+      "epoch": 0.6726853353515695,
+      "grad_norm": 0.15459772944450378,
+      "learning_rate": 4.8581898165967274e-05,
+      "loss": 0.4355,
       "step": 18665
     },
     {
-      "epoch": 0.66,
-      "learning_rate": 4.8665454930794454e-05,
-      "loss": 0.3561,
+      "epoch": 0.6728655350127942,
+      "grad_norm": 0.1914420872926712,
+      "learning_rate": 4.858092915626701e-05,
+      "loss": 0.3937,
       "step": 18670
     },
     {
-      "epoch": 0.66,
-      "learning_rate": 4.8664536472106715e-05,
-      "loss": 0.2988,
+      "epoch": 0.6730457346740188,
+      "grad_norm": 0.15395872294902802,
+      "learning_rate": 4.8579959825281155e-05,
+      "loss": 0.4635,
       "step": 18675
     },
     {
-      "epoch": 0.66,
-      "learning_rate": 4.866361770615058e-05,
-      "loss": 0.2998,
+      "epoch": 0.6732259343352435,
+      "grad_norm": 0.13936087489128113,
+      "learning_rate": 4.857899017302291e-05,
+      "loss": 0.4779,
       "step": 18680
     },
     {
-      "epoch": 0.66,
-      "learning_rate": 4.8662698632937955e-05,
-      "loss": 0.3376,
+      "epoch": 0.6734061339964681,
+      "grad_norm": 0.14681056141853333,
+      "learning_rate": 4.8578020199505495e-05,
+      "loss": 0.4428,
       "step": 18685
     },
     {
-      "epoch": 0.66,
-      "learning_rate": 4.866177925248078e-05,
-      "loss": 0.3156,
+      "epoch": 0.6735863336576927,
+      "grad_norm": 0.13621117174625397,
+      "learning_rate": 4.857704990474211e-05,
+      "loss": 0.4088,
       "step": 18690
     },
     {
-      "epoch": 0.66,
-      "learning_rate": 4.8660859564791006e-05,
-      "loss": 0.3197,
+      "epoch": 0.6737665333189173,
+      "grad_norm": 0.1684703528881073,
+      "learning_rate": 4.857607928874599e-05,
+      "loss": 0.4154,
       "step": 18695
     },
     {
-      "epoch": 0.66,
-      "learning_rate": 4.865993956988055e-05,
-      "loss": 0.3498,
+      "epoch": 0.673946732980142,
+      "grad_norm": 0.1323196440935135,
+      "learning_rate": 4.857510835153035e-05,
+      "loss": 0.4315,
       "step": 18700
     },
     {
-      "epoch": 0.66,
-      "learning_rate": 4.865901926776139e-05,
-      "loss": 0.3049,
+      "epoch": 0.6741269326413666,
+      "grad_norm": 0.16398736834526062,
+      "learning_rate": 4.857413709310843e-05,
+      "loss": 0.4505,
       "step": 18705
     },
     {
-      "epoch": 0.66,
-      "learning_rate": 4.865809865844544e-05,
-      "loss": 0.3001,
+      "epoch": 0.6743071323025913,
+      "grad_norm": 0.18038709461688995,
+      "learning_rate": 4.857316551349345e-05,
+      "loss": 0.4387,
       "step": 18710
     },
     {
-      "epoch": 0.66,
-      "learning_rate": 4.865717774194467e-05,
-      "loss": 0.3121,
+      "epoch": 0.6744873319638159,
+      "grad_norm": 0.169601172208786,
+      "learning_rate": 4.8572193612698656e-05,
+      "loss": 0.4301,
       "step": 18715
     },
     {
-      "epoch": 0.66,
-      "learning_rate": 4.865625651827105e-05,
-      "loss": 0.308,
+      "epoch": 0.6746675316250406,
+      "grad_norm": 0.19574612379074097,
+      "learning_rate": 4.857122139073729e-05,
+      "loss": 0.4735,
       "step": 18720
     },
     {
-      "epoch": 0.66,
-      "learning_rate": 4.865533498743652e-05,
-      "loss": 0.3274,
+      "epoch": 0.6748477312862652,
+      "grad_norm": 0.19240935146808624,
+      "learning_rate": 4.857024884762259e-05,
+      "loss": 0.4467,
       "step": 18725
     },
     {
-      "epoch": 0.66,
-      "learning_rate": 4.8654413149453056e-05,
-      "loss": 0.3408,
+      "epoch": 0.6750279309474898,
+      "grad_norm": 0.16014917194843292,
+      "learning_rate": 4.856927598336782e-05,
+      "loss": 0.4609,
       "step": 18730
     },
     {
-      "epoch": 0.66,
-      "learning_rate": 4.865349100433263e-05,
-      "loss": 0.3402,
+      "epoch": 0.6752081306087144,
+      "grad_norm": 0.1911192089319229,
+      "learning_rate": 4.856830279798623e-05,
+      "loss": 0.5029,
       "step": 18735
     },
     {
-      "epoch": 0.66,
-      "learning_rate": 4.8652568552087206e-05,
-      "loss": 0.3118,
+      "epoch": 0.675388330269939,
+      "grad_norm": 0.1548725664615631,
+      "learning_rate": 4.856732929149107e-05,
+      "loss": 0.4556,
       "step": 18740
     },
     {
-      "epoch": 0.66,
-      "learning_rate": 4.865164579272877e-05,
-      "loss": 0.3159,
+      "epoch": 0.6755685299311637,
+      "grad_norm": 0.16711898148059845,
+      "learning_rate": 4.856635546389562e-05,
+      "loss": 0.4449,
       "step": 18745
     },
     {
-      "epoch": 0.66,
-      "learning_rate": 4.86507227262693e-05,
-      "loss": 0.3268,
+      "epoch": 0.6757487295923884,
+      "grad_norm": 0.18358518183231354,
+      "learning_rate": 4.856538131521313e-05,
+      "loss": 0.4313,
       "step": 18750
     },
     {
-      "epoch": 0.66,
-      "learning_rate": 4.864979935272078e-05,
-      "loss": 0.313,
+      "epoch": 0.675928929253613,
+      "grad_norm": 0.18038628995418549,
+      "learning_rate": 4.8564406845456885e-05,
+      "loss": 0.4427,
       "step": 18755
     },
     {
-      "epoch": 0.66,
-      "learning_rate": 4.8648875672095205e-05,
-      "loss": 0.274,
+      "epoch": 0.6761091289148377,
+      "grad_norm": 0.16218414902687073,
+      "learning_rate": 4.8563432054640155e-05,
+      "loss": 0.4352,
       "step": 18760
     },
     {
-      "epoch": 0.66,
-      "learning_rate": 4.864795168440456e-05,
-      "loss": 0.2841,
+      "epoch": 0.6762893285760623,
+      "grad_norm": 0.1689649075269699,
+      "learning_rate": 4.856245694277623e-05,
+      "loss": 0.4558,
       "step": 18765
     },
     {
-      "epoch": 0.66,
-      "learning_rate": 4.864702738966085e-05,
-      "loss": 0.3536,
+      "epoch": 0.6764695282372869,
+      "grad_norm": 0.2027473747730255,
+      "learning_rate": 4.8561481509878395e-05,
+      "loss": 0.4304,
       "step": 18770
     },
     {
-      "epoch": 0.66,
-      "learning_rate": 4.864610278787607e-05,
-      "loss": 0.3068,
+      "epoch": 0.6766497278985115,
+      "grad_norm": 0.18545962870121002,
+      "learning_rate": 4.856050575595993e-05,
+      "loss": 0.4456,
       "step": 18775
     },
     {
-      "epoch": 0.66,
-      "learning_rate": 4.8645177879062235e-05,
-      "loss": 0.3262,
+      "epoch": 0.6768299275597361,
+      "grad_norm": 0.17125312983989716,
+      "learning_rate": 4.8559529681034135e-05,
+      "loss": 0.4816,
       "step": 18780
     },
     {
-      "epoch": 0.66,
-      "learning_rate": 4.8644252663231344e-05,
-      "loss": 0.2929,
+      "epoch": 0.6770101272209609,
+      "grad_norm": 0.14758311212062836,
+      "learning_rate": 4.855855328511432e-05,
+      "loss": 0.458,
       "step": 18785
     },
     {
-      "epoch": 0.66,
-      "learning_rate": 4.864332714039541e-05,
-      "loss": 0.3021,
+      "epoch": 0.6771903268821855,
+      "grad_norm": 0.14200688898563385,
+      "learning_rate": 4.8557576568213755e-05,
+      "loss": 0.4096,
       "step": 18790
     },
     {
-      "epoch": 0.66,
-      "learning_rate": 4.864240131056646e-05,
-      "loss": 0.3201,
+      "epoch": 0.6773705265434101,
+      "grad_norm": 0.15297238528728485,
+      "learning_rate": 4.855659953034579e-05,
+      "loss": 0.3955,
       "step": 18795
     },
     {
-      "epoch": 0.66,
-      "learning_rate": 4.8641475173756514e-05,
-      "loss": 0.307,
+      "epoch": 0.6775507262046347,
+      "grad_norm": 0.17643597722053528,
+      "learning_rate": 4.855562217152371e-05,
+      "loss": 0.4581,
       "step": 18800
     },
     {
-      "epoch": 0.66,
-      "learning_rate": 4.864054872997759e-05,
-      "loss": 0.3242,
+      "epoch": 0.6777309258658594,
+      "grad_norm": 0.15438847243785858,
+      "learning_rate": 4.8554644491760845e-05,
+      "loss": 0.4511,
       "step": 18805
     },
     {
-      "epoch": 0.66,
-      "learning_rate": 4.863962197924172e-05,
-      "loss": 0.3069,
+      "epoch": 0.677911125527084,
+      "grad_norm": 0.17449019849300385,
+      "learning_rate": 4.8553666491070505e-05,
+      "loss": 0.4128,
       "step": 18810
     },
     {
-      "epoch": 0.66,
-      "learning_rate": 4.863869492156093e-05,
-      "loss": 0.3135,
+      "epoch": 0.6780913251883086,
+      "grad_norm": 0.180333212018013,
+      "learning_rate": 4.855268816946601e-05,
+      "loss": 0.4442,
       "step": 18815
     },
     {
-      "epoch": 0.66,
-      "learning_rate": 4.863776755694728e-05,
-      "loss": 0.324,
+      "epoch": 0.6782715248495332,
+      "grad_norm": 0.1781253069639206,
+      "learning_rate": 4.855170952696071e-05,
+      "loss": 0.4484,
       "step": 18820
     },
     {
-      "epoch": 0.66,
-      "learning_rate": 4.863683988541279e-05,
-      "loss": 0.3285,
+      "epoch": 0.678451724510758,
+      "grad_norm": 0.15952306985855103,
+      "learning_rate": 4.855073056356793e-05,
+      "loss": 0.4409,
       "step": 18825
     },
     {
-      "epoch": 0.66,
-      "learning_rate": 4.86359119069695e-05,
-      "loss": 0.2937,
+      "epoch": 0.6786319241719826,
+      "grad_norm": 0.18420594930648804,
+      "learning_rate": 4.854975127930099e-05,
+      "loss": 0.4467,
       "step": 18830
     },
     {
-      "epoch": 0.66,
-      "learning_rate": 4.863498362162948e-05,
-      "loss": 0.3272,
+      "epoch": 0.6788121238332072,
+      "grad_norm": 0.15397235751152039,
+      "learning_rate": 4.854877167417327e-05,
+      "loss": 0.4439,
       "step": 18835
     },
     {
-      "epoch": 0.66,
-      "learning_rate": 4.863405502940478e-05,
-      "loss": 0.2852,
+      "epoch": 0.6789923234944318,
+      "grad_norm": 0.15009304881095886,
+      "learning_rate": 4.854779174819807e-05,
+      "loss": 0.4126,
       "step": 18840
     },
     {
-      "epoch": 0.66,
-      "learning_rate": 4.863312613030744e-05,
-      "loss": 0.2904,
+      "epoch": 0.6791725231556565,
+      "grad_norm": 0.17719674110412598,
+      "learning_rate": 4.8546811501388784e-05,
+      "loss": 0.4565,
       "step": 18845
     },
     {
-      "epoch": 0.66,
-      "learning_rate": 4.863219692434954e-05,
-      "loss": 0.3309,
+      "epoch": 0.6793527228168811,
+      "grad_norm": 0.1648426502943039,
+      "learning_rate": 4.854583093375875e-05,
+      "loss": 0.4347,
       "step": 18850
     },
     {
-      "epoch": 0.66,
-      "learning_rate": 4.863126741154313e-05,
-      "loss": 0.3213,
+      "epoch": 0.6795329224781057,
+      "grad_norm": 0.14466212689876556,
+      "learning_rate": 4.8544850045321324e-05,
+      "loss": 0.3926,
       "step": 18855
     },
     {
-      "epoch": 0.66,
-      "learning_rate": 4.863033759190029e-05,
-      "loss": 0.3233,
+      "epoch": 0.6797131221393303,
+      "grad_norm": 0.1948980838060379,
+      "learning_rate": 4.8543868836089865e-05,
+      "loss": 0.4354,
       "step": 18860
     },
     {
-      "epoch": 0.66,
-      "learning_rate": 4.862940746543308e-05,
-      "loss": 0.3006,
+      "epoch": 0.6798933218005551,
+      "grad_norm": 0.1797015517950058,
+      "learning_rate": 4.8542887306077765e-05,
+      "loss": 0.4419,
       "step": 18865
     },
     {
-      "epoch": 0.66,
-      "learning_rate": 4.862847703215361e-05,
-      "loss": 0.3136,
+      "epoch": 0.6800735214617797,
+      "grad_norm": 0.12938088178634644,
+      "learning_rate": 4.8541905455298373e-05,
+      "loss": 0.4491,
       "step": 18870
     },
     {
-      "epoch": 0.66,
-      "learning_rate": 4.8627546292073915e-05,
-      "loss": 0.3159,
+      "epoch": 0.6802537211230043,
+      "grad_norm": 0.21112160384655,
+      "learning_rate": 4.854092328376509e-05,
+      "loss": 0.4418,
       "step": 18875
     },
     {
-      "epoch": 0.66,
-      "learning_rate": 4.862661524520611e-05,
-      "loss": 0.2867,
+      "epoch": 0.6804339207842289,
+      "grad_norm": 0.18489904701709747,
+      "learning_rate": 4.853994079149128e-05,
+      "loss": 0.4395,
       "step": 18880
     },
     {
-      "epoch": 0.66,
-      "learning_rate": 4.862568389156228e-05,
-      "loss": 0.3126,
+      "epoch": 0.6806141204454536,
+      "grad_norm": 0.15210841596126556,
+      "learning_rate": 4.853895797849033e-05,
+      "loss": 0.4479,
       "step": 18885
     },
     {
-      "epoch": 0.66,
-      "learning_rate": 4.8624752231154517e-05,
-      "loss": 0.322,
+      "epoch": 0.6807943201066782,
+      "grad_norm": 0.15210148692131042,
+      "learning_rate": 4.8537974844775636e-05,
+      "loss": 0.3912,
       "step": 18890
     },
     {
-      "epoch": 0.66,
-      "learning_rate": 4.86238202639949e-05,
-      "loss": 0.311,
+      "epoch": 0.6809745197679028,
+      "grad_norm": 0.1489754021167755,
+      "learning_rate": 4.85369913903606e-05,
+      "loss": 0.465,
       "step": 18895
     },
     {
-      "epoch": 0.66,
-      "learning_rate": 4.862288799009556e-05,
-      "loss": 0.3173,
+      "epoch": 0.6811547194291274,
+      "grad_norm": 0.20377914607524872,
+      "learning_rate": 4.85360076152586e-05,
+      "loss": 0.454,
       "step": 18900
     },
     {
-      "epoch": 0.67,
-      "learning_rate": 4.862195540946858e-05,
-      "loss": 0.2893,
+      "epoch": 0.6813349190903522,
+      "grad_norm": 0.20608647167682648,
+      "learning_rate": 4.8535023519483055e-05,
+      "loss": 0.4495,
       "step": 18905
     },
     {
-      "epoch": 0.67,
-      "learning_rate": 4.862102252212608e-05,
-      "loss": 0.305,
+      "epoch": 0.6815151187515768,
+      "grad_norm": 0.1873638778924942,
+      "learning_rate": 4.853403910304738e-05,
+      "loss": 0.4436,
       "step": 18910
     },
     {
-      "epoch": 0.67,
-      "learning_rate": 4.8620089328080176e-05,
-      "loss": 0.299,
+      "epoch": 0.6816953184128014,
+      "grad_norm": 0.16861487925052643,
+      "learning_rate": 4.853305436596497e-05,
+      "loss": 0.4342,
       "step": 18915
     },
     {
-      "epoch": 0.67,
-      "learning_rate": 4.861915582734297e-05,
-      "loss": 0.2985,
+      "epoch": 0.681875518074026,
+      "grad_norm": 0.14870557188987732,
+      "learning_rate": 4.853206930824925e-05,
+      "loss": 0.4825,
       "step": 18920
     },
     {
-      "epoch": 0.67,
-      "learning_rate": 4.861822201992659e-05,
-      "loss": 0.3384,
+      "epoch": 0.6820557177352506,
+      "grad_norm": 0.16530665755271912,
+      "learning_rate": 4.853108392991366e-05,
+      "loss": 0.4248,
       "step": 18925
     },
     {
-      "epoch": 0.67,
-      "learning_rate": 4.861728790584317e-05,
-      "loss": 0.3381,
+      "epoch": 0.6822359173964753,
+      "grad_norm": 0.15075165033340454,
+      "learning_rate": 4.8530098230971586e-05,
+      "loss": 0.4616,
       "step": 18930
     },
     {
-      "epoch": 0.67,
-      "learning_rate": 4.861635348510482e-05,
-      "loss": 0.3576,
+      "epoch": 0.6824161170576999,
+      "grad_norm": 0.16559144854545593,
+      "learning_rate": 4.852911221143649e-05,
+      "loss": 0.4233,
       "step": 18935
     },
     {
-      "epoch": 0.67,
-      "learning_rate": 4.861541875772369e-05,
-      "loss": 0.2984,
+      "epoch": 0.6825963167189246,
+      "grad_norm": 0.17002665996551514,
+      "learning_rate": 4.852812587132178e-05,
+      "loss": 0.448,
       "step": 18940
     },
     {
-      "epoch": 0.67,
-      "learning_rate": 4.861448372371191e-05,
-      "loss": 0.2902,
+      "epoch": 0.6827765163801492,
+      "grad_norm": 0.14457158744335175,
+      "learning_rate": 4.8527139210640924e-05,
+      "loss": 0.4475,
       "step": 18945
     },
     {
-      "epoch": 0.67,
-      "learning_rate": 4.861354838308162e-05,
-      "loss": 0.2975,
+      "epoch": 0.6829567160413739,
+      "grad_norm": 0.18576878309249878,
+      "learning_rate": 4.852615222940735e-05,
+      "loss": 0.432,
       "step": 18950
     },
     {
-      "epoch": 0.67,
-      "learning_rate": 4.861261273584497e-05,
-      "loss": 0.3032,
+      "epoch": 0.6831369157025985,
+      "grad_norm": 0.20716910064220428,
+      "learning_rate": 4.852516492763451e-05,
+      "loss": 0.4773,
       "step": 18955
     },
     {
-      "epoch": 0.67,
-      "learning_rate": 4.8611676782014104e-05,
-      "loss": 0.3307,
+      "epoch": 0.6833171153638231,
+      "grad_norm": 0.1403721570968628,
+      "learning_rate": 4.852417730533585e-05,
+      "loss": 0.4248,
       "step": 18960
     },
     {
-      "epoch": 0.67,
-      "learning_rate": 4.861074052160117e-05,
-      "loss": 0.3202,
+      "epoch": 0.6834973150250477,
+      "grad_norm": 0.1828588992357254,
+      "learning_rate": 4.852318936252482e-05,
+      "loss": 0.4335,
       "step": 18965
     },
     {
-      "epoch": 0.67,
-      "learning_rate": 4.860980395461834e-05,
-      "loss": 0.3167,
+      "epoch": 0.6836775146862724,
+      "grad_norm": 0.18035824596881866,
+      "learning_rate": 4.8522201099214904e-05,
+      "loss": 0.4083,
       "step": 18970
     },
     {
-      "epoch": 0.67,
-      "learning_rate": 4.860886708107776e-05,
-      "loss": 0.3472,
+      "epoch": 0.683857714347497,
+      "grad_norm": 0.16249680519104004,
+      "learning_rate": 4.852121251541955e-05,
+      "loss": 0.4238,
       "step": 18975
     },
     {
-      "epoch": 0.67,
-      "learning_rate": 4.8607929900991594e-05,
-      "loss": 0.3341,
+      "epoch": 0.6840379140087217,
+      "grad_norm": 0.1821296066045761,
+      "learning_rate": 4.8520223611152215e-05,
+      "loss": 0.4195,
       "step": 18980
     },
     {
-      "epoch": 0.67,
-      "learning_rate": 4.860699241437202e-05,
-      "loss": 0.3142,
+      "epoch": 0.6842181136699463,
+      "grad_norm": 0.16044920682907104,
+      "learning_rate": 4.851923438642639e-05,
+      "loss": 0.4541,
       "step": 18985
     },
     {
-      "epoch": 0.67,
-      "learning_rate": 4.860605462123121e-05,
-      "loss": 0.327,
+      "epoch": 0.684398313331171,
+      "grad_norm": 0.169897198677063,
+      "learning_rate": 4.851824484125556e-05,
+      "loss": 0.4616,
       "step": 18990
     },
     {
-      "epoch": 0.67,
-      "learning_rate": 4.860511652158134e-05,
-      "loss": 0.3138,
+      "epoch": 0.6845785129923956,
+      "grad_norm": 0.17817699909210205,
+      "learning_rate": 4.851725497565319e-05,
+      "loss": 0.4653,
       "step": 18995
     },
     {
-      "epoch": 0.67,
-      "learning_rate": 4.860417811543458e-05,
-      "loss": 0.3034,
+      "epoch": 0.6847587126536202,
+      "grad_norm": 0.19420120120048523,
+      "learning_rate": 4.851626478963278e-05,
+      "loss": 0.4647,
       "step": 19000
     },
     {
-      "epoch": 0.67,
-      "eval_loss": 0.30461186170578003,
-      "eval_runtime": 10.5328,
-      "eval_samples_per_second": 9.494,
-      "eval_steps_per_second": 9.494,
+      "epoch": 0.6847587126536202,
+      "eval_loss": 0.46408936381340027,
+      "eval_runtime": 3.5332,
+      "eval_samples_per_second": 28.303,
+      "eval_steps_per_second": 7.076,
       "step": 19000
     },
     {
-      "epoch": 0.67,
-      "learning_rate": 4.860323940280313e-05,
-      "loss": 0.3134,
+      "epoch": 0.6849389123148448,
+      "grad_norm": 0.18936601281166077,
+      "learning_rate": 4.851527428320781e-05,
+      "loss": 0.4437,
       "step": 19005
     },
     {
-      "epoch": 0.67,
-      "learning_rate": 4.860230038369917e-05,
-      "loss": 0.3235,
+      "epoch": 0.6851191119760695,
+      "grad_norm": 0.16841618716716766,
+      "learning_rate": 4.8514283456391785e-05,
+      "loss": 0.4336,
       "step": 19010
     },
     {
-      "epoch": 0.67,
-      "learning_rate": 4.860136105813489e-05,
-      "loss": 0.3028,
+      "epoch": 0.6852993116372941,
+      "grad_norm": 0.16009177267551422,
+      "learning_rate": 4.8513292309198197e-05,
+      "loss": 0.4376,
       "step": 19015
     },
     {
-      "epoch": 0.67,
-      "learning_rate": 4.86004214261225e-05,
-      "loss": 0.3016,
+      "epoch": 0.6854795112985188,
+      "grad_norm": 0.16848459839820862,
+      "learning_rate": 4.851230084164056e-05,
+      "loss": 0.4548,
       "step": 19020
     },
     {
-      "epoch": 0.67,
-      "learning_rate": 4.859948148767418e-05,
-      "loss": 0.2898,
+      "epoch": 0.6856597109597434,
+      "grad_norm": 0.1737171709537506,
+      "learning_rate": 4.851130905373237e-05,
+      "loss": 0.4012,
       "step": 19025
     },
     {
-      "epoch": 0.67,
-      "learning_rate": 4.8598541242802155e-05,
-      "loss": 0.2882,
+      "epoch": 0.685839910620968,
+      "grad_norm": 0.1548008769750595,
+      "learning_rate": 4.851031694548714e-05,
+      "loss": 0.4344,
       "step": 19030
     },
     {
-      "epoch": 0.67,
-      "learning_rate": 4.859760069151862e-05,
-      "loss": 0.2995,
+      "epoch": 0.6860201102821927,
+      "grad_norm": 0.17847006022930145,
+      "learning_rate": 4.85093245169184e-05,
+      "loss": 0.4582,
       "step": 19035
     },
     {
-      "epoch": 0.67,
-      "learning_rate": 4.859665983383579e-05,
-      "loss": 0.309,
+      "epoch": 0.6862003099434173,
+      "grad_norm": 0.17521251738071442,
+      "learning_rate": 4.8508331768039674e-05,
+      "loss": 0.436,
       "step": 19040
     },
     {
-      "epoch": 0.67,
-      "learning_rate": 4.8595718669765874e-05,
-      "loss": 0.3045,
+      "epoch": 0.6863805096046419,
+      "grad_norm": 0.1900710016489029,
+      "learning_rate": 4.850733869886446e-05,
+      "loss": 0.4429,
       "step": 19045
     },
     {
-      "epoch": 0.67,
-      "learning_rate": 4.859477719932112e-05,
-      "loss": 0.2974,
+      "epoch": 0.6865607092658665,
+      "grad_norm": 0.22452399134635925,
+      "learning_rate": 4.850634530940632e-05,
+      "loss": 0.4393,
       "step": 19050
     },
     {
-      "epoch": 0.67,
-      "learning_rate": 4.859383542251372e-05,
-      "loss": 0.322,
+      "epoch": 0.6867409089270912,
+      "grad_norm": 0.1778751015663147,
+      "learning_rate": 4.850535159967877e-05,
+      "loss": 0.4614,
       "step": 19055
     },
     {
-      "epoch": 0.67,
-      "learning_rate": 4.8592893339355925e-05,
-      "loss": 0.3175,
+      "epoch": 0.6869211085883159,
+      "grad_norm": 0.1997576355934143,
+      "learning_rate": 4.8504357569695365e-05,
+      "loss": 0.4381,
       "step": 19060
     },
     {
-      "epoch": 0.67,
-      "learning_rate": 4.859195094985996e-05,
-      "loss": 0.3287,
+      "epoch": 0.6871013082495405,
+      "grad_norm": 0.17467275261878967,
+      "learning_rate": 4.850336321946963e-05,
+      "loss": 0.4544,
       "step": 19065
     },
     {
-      "epoch": 0.67,
-      "learning_rate": 4.859100825403805e-05,
-      "loss": 0.311,
+      "epoch": 0.6872815079107651,
+      "grad_norm": 0.15613585710525513,
+      "learning_rate": 4.850236854901513e-05,
+      "loss": 0.42,
       "step": 19070
     },
     {
-      "epoch": 0.67,
-      "learning_rate": 4.859006525190244e-05,
-      "loss": 0.3208,
+      "epoch": 0.6874617075719898,
+      "grad_norm": 0.18194815516471863,
+      "learning_rate": 4.85013735583454e-05,
+      "loss": 0.475,
       "step": 19075
     },
     {
-      "epoch": 0.67,
-      "learning_rate": 4.858912194346539e-05,
-      "loss": 0.3133,
+      "epoch": 0.6876419072332144,
+      "grad_norm": 0.17985866963863373,
+      "learning_rate": 4.850037824747401e-05,
+      "loss": 0.4177,
       "step": 19080
     },
     {
-      "epoch": 0.67,
-      "learning_rate": 4.858817832873913e-05,
-      "loss": 0.315,
+      "epoch": 0.687822106894439,
+      "grad_norm": 0.19839969277381897,
+      "learning_rate": 4.8499382616414515e-05,
+      "loss": 0.4429,
       "step": 19085
     },
     {
-      "epoch": 0.67,
-      "learning_rate": 4.8587234407735926e-05,
-      "loss": 0.2925,
+      "epoch": 0.6880023065556636,
+      "grad_norm": 0.1918293535709381,
+      "learning_rate": 4.8498386665180474e-05,
+      "loss": 0.4412,
       "step": 19090
     },
     {
-      "epoch": 0.67,
-      "learning_rate": 4.858629018046803e-05,
-      "loss": 0.3254,
+      "epoch": 0.6881825062168883,
+      "grad_norm": 0.20238642394542694,
+      "learning_rate": 4.8497390393785475e-05,
+      "loss": 0.4255,
       "step": 19095
     },
     {
-      "epoch": 0.67,
-      "learning_rate": 4.858534564694769e-05,
-      "loss": 0.2914,
+      "epoch": 0.688362705878113,
+      "grad_norm": 0.15424714982509613,
+      "learning_rate": 4.849639380224308e-05,
+      "loss": 0.4449,
       "step": 19100
     },
     {
-      "epoch": 0.67,
-      "learning_rate": 4.858440080718719e-05,
-      "loss": 0.3161,
+      "epoch": 0.6885429055393376,
+      "grad_norm": 0.15987904369831085,
+      "learning_rate": 4.8495396890566855e-05,
+      "loss": 0.4287,
       "step": 19105
     },
     {
-      "epoch": 0.67,
-      "learning_rate": 4.858345566119879e-05,
-      "loss": 0.3113,
+      "epoch": 0.6887231052005622,
+      "grad_norm": 0.19473698735237122,
+      "learning_rate": 4.849439965877041e-05,
+      "loss": 0.4664,
       "step": 19110
     },
     {
-      "epoch": 0.67,
-      "learning_rate": 4.8582510208994746e-05,
-      "loss": 0.3259,
+      "epoch": 0.6889033048617869,
+      "grad_norm": 0.15366004407405853,
+      "learning_rate": 4.849340210686732e-05,
+      "loss": 0.4612,
       "step": 19115
     },
     {
-      "epoch": 0.67,
-      "learning_rate": 4.858156445058736e-05,
-      "loss": 0.3197,
+      "epoch": 0.6890835045230115,
+      "grad_norm": 0.20622040331363678,
+      "learning_rate": 4.849240423487117e-05,
+      "loss": 0.4397,
       "step": 19120
     },
     {
-      "epoch": 0.67,
-      "learning_rate": 4.85806183859889e-05,
-      "loss": 0.3311,
+      "epoch": 0.6892637041842361,
+      "grad_norm": 0.1534797102212906,
+      "learning_rate": 4.8491406042795565e-05,
+      "loss": 0.4492,
       "step": 19125
     },
     {
-      "epoch": 0.67,
-      "learning_rate": 4.857967201521165e-05,
-      "loss": 0.3248,
+      "epoch": 0.6894439038454607,
+      "grad_norm": 0.16360458731651306,
+      "learning_rate": 4.849040753065409e-05,
+      "loss": 0.4591,
       "step": 19130
     },
     {
-      "epoch": 0.67,
-      "learning_rate": 4.857872533826789e-05,
-      "loss": 0.3058,
+      "epoch": 0.6896241035066855,
+      "grad_norm": 0.15157437324523926,
+      "learning_rate": 4.848940869846037e-05,
+      "loss": 0.4604,
       "step": 19135
     },
     {
-      "epoch": 0.67,
-      "learning_rate": 4.8577778355169935e-05,
-      "loss": 0.3079,
+      "epoch": 0.6898043031679101,
+      "grad_norm": 0.2104407548904419,
+      "learning_rate": 4.8488409546228e-05,
+      "loss": 0.4153,
       "step": 19140
     },
     {
-      "epoch": 0.67,
-      "learning_rate": 4.857683106593005e-05,
-      "loss": 0.3095,
+      "epoch": 0.6899845028291347,
+      "grad_norm": 0.2011989951133728,
+      "learning_rate": 4.8487410073970594e-05,
+      "loss": 0.4757,
       "step": 19145
     },
     {
-      "epoch": 0.67,
-      "learning_rate": 4.8575883470560565e-05,
-      "loss": 0.2988,
+      "epoch": 0.6901647024903593,
+      "grad_norm": 0.18023835122585297,
+      "learning_rate": 4.848641028170178e-05,
+      "loss": 0.4305,
       "step": 19150
     },
     {
-      "epoch": 0.67,
-      "learning_rate": 4.857493556907376e-05,
-      "loss": 0.3,
+      "epoch": 0.690344902151584,
+      "grad_norm": 0.1514541208744049,
+      "learning_rate": 4.848541016943516e-05,
+      "loss": 0.4403,
       "step": 19155
     },
     {
-      "epoch": 0.67,
-      "learning_rate": 4.857398736148196e-05,
-      "loss": 0.3329,
+      "epoch": 0.6905251018128086,
+      "grad_norm": 0.1530730277299881,
+      "learning_rate": 4.8484409737184386e-05,
+      "loss": 0.4532,
       "step": 19160
     },
     {
-      "epoch": 0.67,
-      "learning_rate": 4.857303884779747e-05,
-      "loss": 0.3193,
+      "epoch": 0.6907053014740332,
+      "grad_norm": 0.18207421898841858,
+      "learning_rate": 4.848340898496308e-05,
+      "loss": 0.4431,
       "step": 19165
     },
     {
-      "epoch": 0.67,
-      "learning_rate": 4.8572090028032604e-05,
-      "loss": 0.3163,
+      "epoch": 0.6908855011352578,
+      "grad_norm": 0.17817628383636475,
+      "learning_rate": 4.848240791278486e-05,
+      "loss": 0.425,
       "step": 19170
     },
     {
-      "epoch": 0.67,
-      "learning_rate": 4.8571140902199684e-05,
-      "loss": 0.3055,
+      "epoch": 0.6910657007964826,
+      "grad_norm": 0.13877823948860168,
+      "learning_rate": 4.848140652066339e-05,
+      "loss": 0.4346,
       "step": 19175
     },
     {
-      "epoch": 0.67,
-      "learning_rate": 4.857019147031103e-05,
-      "loss": 0.3248,
+      "epoch": 0.6912459004577072,
+      "grad_norm": 0.15629951655864716,
+      "learning_rate": 4.84804048086123e-05,
+      "loss": 0.4438,
       "step": 19180
     },
     {
-      "epoch": 0.67,
-      "learning_rate": 4.8569241732378975e-05,
-      "loss": 0.3135,
+      "epoch": 0.6914261001189318,
+      "grad_norm": 0.15390846133232117,
+      "learning_rate": 4.8479402776645235e-05,
+      "loss": 0.4709,
       "step": 19185
     },
     {
-      "epoch": 0.68,
-      "learning_rate": 4.856829168841585e-05,
-      "loss": 0.3074,
+      "epoch": 0.6916062997801564,
+      "grad_norm": 0.17919474840164185,
+      "learning_rate": 4.847840042477586e-05,
+      "loss": 0.4731,
       "step": 19190
     },
     {
-      "epoch": 0.68,
-      "learning_rate": 4.856734133843399e-05,
-      "loss": 0.3359,
+      "epoch": 0.691786499441381,
+      "grad_norm": 0.170258030295372,
+      "learning_rate": 4.8477397753017816e-05,
+      "loss": 0.468,
       "step": 19195
     },
     {
-      "epoch": 0.68,
-      "learning_rate": 4.8566390682445734e-05,
-      "loss": 0.3111,
+      "epoch": 0.6919666991026057,
+      "grad_norm": 0.19165678322315216,
+      "learning_rate": 4.847639476138478e-05,
+      "loss": 0.5018,
       "step": 19200
     },
     {
-      "epoch": 0.68,
-      "learning_rate": 4.8565439720463424e-05,
-      "loss": 0.3306,
+      "epoch": 0.6921468987638303,
+      "grad_norm": 0.15710307657718658,
+      "learning_rate": 4.8475391449890405e-05,
+      "loss": 0.4447,
       "step": 19205
     },
     {
-      "epoch": 0.68,
-      "learning_rate": 4.856448845249941e-05,
-      "loss": 0.3012,
+      "epoch": 0.6923270984250549,
+      "grad_norm": 0.1523151695728302,
+      "learning_rate": 4.847438781854837e-05,
+      "loss": 0.4627,
       "step": 19210
     },
     {
-      "epoch": 0.68,
-      "learning_rate": 4.856353687856604e-05,
-      "loss": 0.3008,
+      "epoch": 0.6925072980862796,
+      "grad_norm": 0.14756174385547638,
+      "learning_rate": 4.8473383867372345e-05,
+      "loss": 0.4621,
       "step": 19215
     },
     {
-      "epoch": 0.68,
-      "learning_rate": 4.856258499867568e-05,
-      "loss": 0.3193,
+      "epoch": 0.6926874977475043,
+      "grad_norm": 0.2157120406627655,
+      "learning_rate": 4.847237959637602e-05,
+      "loss": 0.4199,
       "step": 19220
     },
     {
-      "epoch": 0.68,
-      "learning_rate": 4.8561632812840674e-05,
-      "loss": 0.3251,
+      "epoch": 0.6928676974087289,
+      "grad_norm": 0.14184360206127167,
+      "learning_rate": 4.847137500557305e-05,
+      "loss": 0.4347,
       "step": 19225
     },
     {
-      "epoch": 0.68,
-      "learning_rate": 4.8560680321073405e-05,
-      "loss": 0.2841,
+      "epoch": 0.6930478970699535,
+      "grad_norm": 0.13508060574531555,
+      "learning_rate": 4.847037009497715e-05,
+      "loss": 0.4535,
       "step": 19230
     },
     {
-      "epoch": 0.68,
-      "learning_rate": 4.8559727523386214e-05,
-      "loss": 0.3252,
+      "epoch": 0.6932280967311781,
+      "grad_norm": 0.14974738657474518,
+      "learning_rate": 4.8469364864602e-05,
+      "loss": 0.4337,
       "step": 19235
     },
     {
-      "epoch": 0.68,
-      "learning_rate": 4.8558774419791496e-05,
-      "loss": 0.3157,
+      "epoch": 0.6934082963924028,
+      "grad_norm": 0.2109774798154831,
+      "learning_rate": 4.846835931446129e-05,
+      "loss": 0.4732,
       "step": 19240
     },
     {
-      "epoch": 0.68,
-      "learning_rate": 4.855782101030162e-05,
-      "loss": 0.3063,
+      "epoch": 0.6935884960536274,
+      "grad_norm": 0.1753171980381012,
+      "learning_rate": 4.846735344456873e-05,
+      "loss": 0.4438,
       "step": 19245
     },
     {
-      "epoch": 0.68,
-      "learning_rate": 4.855686729492896e-05,
-      "loss": 0.284,
+      "epoch": 0.693768695714852,
+      "grad_norm": 0.20404259860515594,
+      "learning_rate": 4.8466347254938034e-05,
+      "loss": 0.4431,
       "step": 19250
     },
     {
-      "epoch": 0.68,
-      "learning_rate": 4.85559132736859e-05,
-      "loss": 0.3275,
+      "epoch": 0.6939488953760767,
+      "grad_norm": 0.1709374636411667,
+      "learning_rate": 4.846534074558289e-05,
+      "loss": 0.4206,
       "step": 19255
     },
     {
-      "epoch": 0.68,
-      "learning_rate": 4.855495894658484e-05,
-      "loss": 0.295,
+      "epoch": 0.6941290950373014,
+      "grad_norm": 0.13313916325569153,
+      "learning_rate": 4.8464333916517025e-05,
+      "loss": 0.4372,
       "step": 19260
     },
     {
-      "epoch": 0.68,
-      "learning_rate": 4.8554004313638154e-05,
-      "loss": 0.3287,
+      "epoch": 0.694309294698526,
+      "grad_norm": 0.1679733395576477,
+      "learning_rate": 4.8463326767754145e-05,
+      "loss": 0.4753,
       "step": 19265
     },
     {
-      "epoch": 0.68,
-      "learning_rate": 4.855304937485825e-05,
-      "loss": 0.3224,
+      "epoch": 0.6944894943597506,
+      "grad_norm": 0.14599260687828064,
+      "learning_rate": 4.846231929930799e-05,
+      "loss": 0.423,
       "step": 19270
     },
     {
-      "epoch": 0.68,
-      "learning_rate": 4.8552094130257515e-05,
-      "loss": 0.3019,
+      "epoch": 0.6946696940209752,
+      "grad_norm": 0.17458441853523254,
+      "learning_rate": 4.846131151119228e-05,
+      "loss": 0.4576,
       "step": 19275
     },
     {
-      "epoch": 0.68,
-      "learning_rate": 4.855113857984836e-05,
-      "loss": 0.334,
+      "epoch": 0.6948498936821998,
+      "grad_norm": 0.14298151433467865,
+      "learning_rate": 4.8460303403420735e-05,
+      "loss": 0.4628,
       "step": 19280
     },
     {
-      "epoch": 0.68,
-      "learning_rate": 4.85501827236432e-05,
-      "loss": 0.295,
+      "epoch": 0.6950300933434245,
+      "grad_norm": 0.18267078697681427,
+      "learning_rate": 4.84592949760071e-05,
+      "loss": 0.4505,
       "step": 19285
     },
     {
-      "epoch": 0.68,
-      "learning_rate": 4.8549226561654426e-05,
-      "loss": 0.2808,
+      "epoch": 0.6952102930046492,
+      "grad_norm": 0.18791311979293823,
+      "learning_rate": 4.845828622896511e-05,
+      "loss": 0.3925,
       "step": 19290
     },
     {
-      "epoch": 0.68,
-      "learning_rate": 4.8548270093894465e-05,
-      "loss": 0.3072,
+      "epoch": 0.6953904926658738,
+      "grad_norm": 0.1504139006137848,
+      "learning_rate": 4.8457277162308526e-05,
+      "loss": 0.4281,
       "step": 19295
     },
     {
-      "epoch": 0.68,
-      "learning_rate": 4.854731332037574e-05,
-      "loss": 0.3187,
+      "epoch": 0.6955706923270984,
+      "grad_norm": 0.1674523800611496,
+      "learning_rate": 4.8456267776051066e-05,
+      "loss": 0.4462,
       "step": 19300
     },
     {
-      "epoch": 0.68,
-      "learning_rate": 4.8546356241110666e-05,
-      "loss": 0.3158,
+      "epoch": 0.6957508919883231,
+      "grad_norm": 0.1913890391588211,
+      "learning_rate": 4.84552580702065e-05,
+      "loss": 0.4054,
       "step": 19305
     },
     {
-      "epoch": 0.68,
-      "learning_rate": 4.8545398856111676e-05,
-      "loss": 0.3075,
+      "epoch": 0.6959310916495477,
+      "grad_norm": 0.18787144124507904,
+      "learning_rate": 4.8454248044788594e-05,
+      "loss": 0.4503,
       "step": 19310
     },
     {
-      "epoch": 0.68,
-      "learning_rate": 4.854444116539121e-05,
-      "loss": 0.3375,
+      "epoch": 0.6961112913107723,
+      "grad_norm": 0.15281574428081512,
+      "learning_rate": 4.845323769981109e-05,
+      "loss": 0.4506,
       "step": 19315
     },
     {
-      "epoch": 0.68,
-      "learning_rate": 4.854348316896168e-05,
-      "loss": 0.3051,
+      "epoch": 0.6962914909719969,
+      "grad_norm": 0.17581430077552795,
+      "learning_rate": 4.845222703528777e-05,
+      "loss": 0.4174,
       "step": 19320
     },
     {
-      "epoch": 0.68,
-      "learning_rate": 4.854252486683554e-05,
-      "loss": 0.293,
+      "epoch": 0.6964716906332216,
+      "grad_norm": 0.14308330416679382,
+      "learning_rate": 4.845121605123239e-05,
+      "loss": 0.4106,
       "step": 19325
     },
     {
-      "epoch": 0.68,
-      "learning_rate": 4.8541566259025226e-05,
-      "loss": 0.3089,
+      "epoch": 0.6966518902944463,
+      "grad_norm": 0.1626398116350174,
+      "learning_rate": 4.8450204747658734e-05,
+      "loss": 0.4736,
       "step": 19330
     },
     {
-      "epoch": 0.68,
-      "learning_rate": 4.8540607345543194e-05,
-      "loss": 0.312,
+      "epoch": 0.6968320899556709,
+      "grad_norm": 0.14919036626815796,
+      "learning_rate": 4.844919312458058e-05,
+      "loss": 0.4042,
       "step": 19335
     },
     {
-      "epoch": 0.68,
-      "learning_rate": 4.853964812640188e-05,
-      "loss": 0.3218,
+      "epoch": 0.6970122896168955,
+      "grad_norm": 0.15542307496070862,
+      "learning_rate": 4.844818118201171e-05,
+      "loss": 0.4378,
       "step": 19340
     },
     {
-      "epoch": 0.68,
-      "learning_rate": 4.853868860161376e-05,
-      "loss": 0.2937,
+      "epoch": 0.6971924892781202,
+      "grad_norm": 0.18490786850452423,
+      "learning_rate": 4.844716891996591e-05,
+      "loss": 0.4315,
       "step": 19345
     },
     {
-      "epoch": 0.68,
-      "learning_rate": 4.8537728771191275e-05,
-      "loss": 0.3022,
+      "epoch": 0.6973726889393448,
+      "grad_norm": 0.20370669662952423,
+      "learning_rate": 4.8446156338456975e-05,
+      "loss": 0.4472,
       "step": 19350
     },
     {
-      "epoch": 0.68,
-      "learning_rate": 4.8536768635146895e-05,
-      "loss": 0.3084,
+      "epoch": 0.6975528886005694,
+      "grad_norm": 0.1456826627254486,
+      "learning_rate": 4.8445143437498696e-05,
+      "loss": 0.4498,
       "step": 19355
     },
     {
-      "epoch": 0.68,
-      "learning_rate": 4.853580819349309e-05,
-      "loss": 0.3086,
+      "epoch": 0.697733088261794,
+      "grad_norm": 0.159530371427536,
+      "learning_rate": 4.844413021710488e-05,
+      "loss": 0.4576,
       "step": 19360
     },
     {
-      "epoch": 0.68,
-      "learning_rate": 4.853484744624232e-05,
-      "loss": 0.3378,
+      "epoch": 0.6979132879230187,
+      "grad_norm": 0.1378137171268463,
+      "learning_rate": 4.844311667728932e-05,
+      "loss": 0.4643,
       "step": 19365
     },
     {
-      "epoch": 0.68,
-      "learning_rate": 4.853388639340707e-05,
-      "loss": 0.2914,
+      "epoch": 0.6980934875842434,
+      "grad_norm": 0.17406761646270752,
+      "learning_rate": 4.844210281806585e-05,
+      "loss": 0.4662,
       "step": 19370
     },
     {
-      "epoch": 0.68,
-      "learning_rate": 4.853292503499981e-05,
-      "loss": 0.293,
+      "epoch": 0.698273687245468,
+      "grad_norm": 0.14517958462238312,
+      "learning_rate": 4.844108863944826e-05,
+      "loss": 0.4136,
       "step": 19375
     },
     {
-      "epoch": 0.68,
-      "learning_rate": 4.8531963371033026e-05,
-      "loss": 0.2818,
+      "epoch": 0.6984538869066926,
+      "grad_norm": 0.16255003213882446,
+      "learning_rate": 4.844007414145037e-05,
+      "loss": 0.4686,
       "step": 19380
     },
     {
-      "epoch": 0.68,
-      "learning_rate": 4.8531001401519214e-05,
-      "loss": 0.3251,
+      "epoch": 0.6986340865679173,
+      "grad_norm": 0.14890755712985992,
+      "learning_rate": 4.843905932408601e-05,
+      "loss": 0.4148,
       "step": 19385
     },
     {
-      "epoch": 0.68,
-      "learning_rate": 4.853003912647085e-05,
-      "loss": 0.2689,
+      "epoch": 0.6988142862291419,
+      "grad_norm": 0.1591944694519043,
+      "learning_rate": 4.843804418736901e-05,
+      "loss": 0.4243,
       "step": 19390
     },
     {
-      "epoch": 0.68,
-      "learning_rate": 4.852907654590044e-05,
-      "loss": 0.3265,
+      "epoch": 0.6989944858903665,
+      "grad_norm": 0.15910327434539795,
+      "learning_rate": 4.84370287313132e-05,
+      "loss": 0.4091,
       "step": 19395
     },
     {
-      "epoch": 0.68,
-      "learning_rate": 4.8528113659820474e-05,
-      "loss": 0.306,
+      "epoch": 0.6991746855515911,
+      "grad_norm": 0.19701680541038513,
+      "learning_rate": 4.84360129559324e-05,
+      "loss": 0.4521,
       "step": 19400
     },
     {
-      "epoch": 0.68,
-      "learning_rate": 4.8527150468243454e-05,
-      "loss": 0.2849,
+      "epoch": 0.6993548852128157,
+      "grad_norm": 0.16682995855808258,
+      "learning_rate": 4.8434996861240457e-05,
+      "loss": 0.4676,
       "step": 19405
     },
     {
-      "epoch": 0.68,
-      "learning_rate": 4.8526186971181893e-05,
-      "loss": 0.3083,
+      "epoch": 0.6995350848740405,
+      "grad_norm": 0.14017580449581146,
+      "learning_rate": 4.843398044725123e-05,
+      "loss": 0.4273,
       "step": 19410
     },
     {
-      "epoch": 0.68,
-      "learning_rate": 4.8525223168648296e-05,
-      "loss": 0.3484,
+      "epoch": 0.6997152845352651,
+      "grad_norm": 0.19788876175880432,
+      "learning_rate": 4.843296371397855e-05,
+      "loss": 0.4129,
       "step": 19415
     },
     {
-      "epoch": 0.68,
-      "learning_rate": 4.852425906065519e-05,
-      "loss": 0.3175,
+      "epoch": 0.6998954841964897,
+      "grad_norm": 0.21277104318141937,
+      "learning_rate": 4.843194666143628e-05,
+      "loss": 0.5118,
       "step": 19420
     },
     {
-      "epoch": 0.68,
-      "learning_rate": 4.852329464721507e-05,
-      "loss": 0.3194,
+      "epoch": 0.7000756838577143,
+      "grad_norm": 0.19250738620758057,
+      "learning_rate": 4.843092928963827e-05,
+      "loss": 0.4856,
       "step": 19425
     },
     {
-      "epoch": 0.68,
-      "learning_rate": 4.852232992834048e-05,
-      "loss": 0.2969,
+      "epoch": 0.700255883518939,
+      "grad_norm": 0.15623074769973755,
+      "learning_rate": 4.8429911598598386e-05,
+      "loss": 0.4348,
       "step": 19430
     },
     {
-      "epoch": 0.68,
-      "learning_rate": 4.852136490404393e-05,
-      "loss": 0.3438,
+      "epoch": 0.7004360831801636,
+      "grad_norm": 0.13664281368255615,
+      "learning_rate": 4.84288935883305e-05,
+      "loss": 0.4232,
       "step": 19435
     },
     {
-      "epoch": 0.68,
-      "learning_rate": 4.852039957433797e-05,
-      "loss": 0.3239,
+      "epoch": 0.7006162828413882,
+      "grad_norm": 0.1700459122657776,
+      "learning_rate": 4.842787525884847e-05,
+      "loss": 0.4598,
       "step": 19440
     },
     {
-      "epoch": 0.68,
-      "learning_rate": 4.851943393923512e-05,
-      "loss": 0.3088,
+      "epoch": 0.700796482502613,
+      "grad_norm": 0.15901970863342285,
+      "learning_rate": 4.842685661016617e-05,
+      "loss": 0.4267,
       "step": 19445
     },
     {
-      "epoch": 0.68,
-      "learning_rate": 4.851846799874791e-05,
-      "loss": 0.3085,
+      "epoch": 0.7009766821638376,
+      "grad_norm": 0.20481829345226288,
+      "learning_rate": 4.842583764229749e-05,
+      "loss": 0.4309,
       "step": 19450
     },
     {
-      "epoch": 0.68,
-      "learning_rate": 4.8517501752888895e-05,
-      "loss": 0.274,
+      "epoch": 0.7011568818250622,
+      "grad_norm": 0.19733159244060516,
+      "learning_rate": 4.8424818355256304e-05,
+      "loss": 0.4754,
       "step": 19455
     },
     {
-      "epoch": 0.68,
-      "learning_rate": 4.851653520167062e-05,
-      "loss": 0.3079,
+      "epoch": 0.7013370814862868,
+      "grad_norm": 0.1566084325313568,
+      "learning_rate": 4.84237987490565e-05,
+      "loss": 0.4086,
       "step": 19460
     },
     {
-      "epoch": 0.68,
-      "learning_rate": 4.851556834510564e-05,
-      "loss": 0.3116,
+      "epoch": 0.7015172811475114,
+      "grad_norm": 0.16943992674350739,
+      "learning_rate": 4.842277882371198e-05,
+      "loss": 0.4623,
       "step": 19465
     },
     {
-      "epoch": 0.69,
-      "learning_rate": 4.851460118320649e-05,
-      "loss": 0.2865,
+      "epoch": 0.7016974808087361,
+      "grad_norm": 0.1365501880645752,
+      "learning_rate": 4.842175857923663e-05,
+      "loss": 0.4053,
       "step": 19470
     },
     {
-      "epoch": 0.69,
-      "learning_rate": 4.851363371598575e-05,
-      "loss": 0.3095,
+      "epoch": 0.7018776804699607,
+      "grad_norm": 0.1714967042207718,
+      "learning_rate": 4.842073801564436e-05,
+      "loss": 0.4687,
       "step": 19475
     },
     {
-      "epoch": 0.69,
-      "learning_rate": 4.851266594345597e-05,
-      "loss": 0.2859,
+      "epoch": 0.7020578801311853,
+      "grad_norm": 0.1788230538368225,
+      "learning_rate": 4.841971713294906e-05,
+      "loss": 0.4179,
       "step": 19480
     },
     {
-      "epoch": 0.69,
-      "learning_rate": 4.8511697865629715e-05,
-      "loss": 0.2908,
+      "epoch": 0.70223807979241,
+      "grad_norm": 0.16317056119441986,
+      "learning_rate": 4.841869593116466e-05,
+      "loss": 0.4278,
       "step": 19485
     },
     {
-      "epoch": 0.69,
-      "learning_rate": 4.851072948251957e-05,
-      "loss": 0.3073,
+      "epoch": 0.7024182794536347,
+      "grad_norm": 0.149271622300148,
+      "learning_rate": 4.841767441030505e-05,
+      "loss": 0.4528,
       "step": 19490
     },
     {
-      "epoch": 0.69,
-      "learning_rate": 4.850976079413808e-05,
-      "loss": 0.3105,
+      "epoch": 0.7025984791148593,
+      "grad_norm": 0.17781810462474823,
+      "learning_rate": 4.841665257038416e-05,
+      "loss": 0.421,
       "step": 19495
     },
     {
-      "epoch": 0.69,
-      "learning_rate": 4.850879180049785e-05,
-      "loss": 0.3014,
+      "epoch": 0.7027786787760839,
+      "grad_norm": 0.16509835422039032,
+      "learning_rate": 4.841563041141592e-05,
+      "loss": 0.4252,
       "step": 19500
     },
     {
-      "epoch": 0.69,
-      "eval_loss": 0.30417418479919434,
-      "eval_runtime": 10.5331,
-      "eval_samples_per_second": 9.494,
-      "eval_steps_per_second": 9.494,
+      "epoch": 0.7027786787760839,
+      "eval_loss": 0.4628825783729553,
+      "eval_runtime": 3.5462,
+      "eval_samples_per_second": 28.199,
+      "eval_steps_per_second": 7.05,
       "step": 19500
     },
     {
-      "epoch": 0.69,
-      "learning_rate": 4.8507822501611444e-05,
-      "loss": 0.3377,
+      "epoch": 0.7029588784373085,
+      "grad_norm": 0.16837020218372345,
+      "learning_rate": 4.841460793341425e-05,
+      "loss": 0.457,
       "step": 19505
     },
     {
-      "epoch": 0.69,
-      "learning_rate": 4.850685289749146e-05,
-      "loss": 0.2887,
+      "epoch": 0.7031390780985332,
+      "grad_norm": 0.18554091453552246,
+      "learning_rate": 4.8413585136393066e-05,
+      "loss": 0.421,
       "step": 19510
     },
     {
-      "epoch": 0.69,
-      "learning_rate": 4.850588298815048e-05,
-      "loss": 0.3038,
+      "epoch": 0.7033192777597578,
+      "grad_norm": 0.20547594130039215,
+      "learning_rate": 4.841256202036634e-05,
+      "loss": 0.4686,
       "step": 19515
     },
     {
-      "epoch": 0.69,
-      "learning_rate": 4.850491277360111e-05,
-      "loss": 0.2823,
+      "epoch": 0.7034994774209824,
+      "grad_norm": 0.18383750319480896,
+      "learning_rate": 4.841153858534797e-05,
+      "loss": 0.4405,
       "step": 19520
     },
     {
-      "epoch": 0.69,
-      "learning_rate": 4.8503942253855927e-05,
-      "loss": 0.3096,
+      "epoch": 0.7036796770822071,
+      "grad_norm": 0.18066047132015228,
+      "learning_rate": 4.8410514831351926e-05,
+      "loss": 0.3865,
       "step": 19525
     },
     {
-      "epoch": 0.69,
-      "learning_rate": 4.8502971428927544e-05,
-      "loss": 0.295,
+      "epoch": 0.7038598767434318,
+      "grad_norm": 0.144275963306427,
+      "learning_rate": 4.840949075839215e-05,
+      "loss": 0.4067,
       "step": 19530
     },
     {
-      "epoch": 0.69,
-      "learning_rate": 4.850200029882857e-05,
-      "loss": 0.2925,
+      "epoch": 0.7040400764046564,
+      "grad_norm": 0.19798412919044495,
+      "learning_rate": 4.84084663664826e-05,
+      "loss": 0.4514,
       "step": 19535
     },
     {
-      "epoch": 0.69,
-      "learning_rate": 4.850102886357161e-05,
-      "loss": 0.3171,
+      "epoch": 0.704220276065881,
+      "grad_norm": 0.16639620065689087,
+      "learning_rate": 4.840744165563721e-05,
+      "loss": 0.4901,
       "step": 19540
     },
     {
-      "epoch": 0.69,
-      "learning_rate": 4.850005712316928e-05,
-      "loss": 0.3163,
+      "epoch": 0.7044004757271056,
+      "grad_norm": 0.17380645871162415,
+      "learning_rate": 4.8406416625869977e-05,
+      "loss": 0.4401,
       "step": 19545
     },
     {
-      "epoch": 0.69,
-      "learning_rate": 4.849908507763419e-05,
-      "loss": 0.3025,
+      "epoch": 0.7045806753883302,
+      "grad_norm": 0.16596467792987823,
+      "learning_rate": 4.840539127719484e-05,
+      "loss": 0.4576,
       "step": 19550
     },
     {
-      "epoch": 0.69,
-      "learning_rate": 4.849811272697897e-05,
-      "loss": 0.3011,
+      "epoch": 0.7047608750495549,
+      "grad_norm": 0.14880597591400146,
+      "learning_rate": 4.8404365609625785e-05,
+      "loss": 0.4128,
       "step": 19555
     },
     {
-      "epoch": 0.69,
-      "learning_rate": 4.849714007121624e-05,
-      "loss": 0.3052,
+      "epoch": 0.7049410747107795,
+      "grad_norm": 0.21893320977687836,
+      "learning_rate": 4.840333962317678e-05,
+      "loss": 0.4075,
       "step": 19560
     },
     {
-      "epoch": 0.69,
-      "learning_rate": 4.8496167110358637e-05,
-      "loss": 0.304,
+      "epoch": 0.7051212743720042,
+      "grad_norm": 0.16774235665798187,
+      "learning_rate": 4.8402313317861797e-05,
+      "loss": 0.4624,
       "step": 19565
     },
     {
-      "epoch": 0.69,
-      "learning_rate": 4.8495193844418784e-05,
-      "loss": 0.3287,
+      "epoch": 0.7053014740332288,
+      "grad_norm": 0.1787051409482956,
+      "learning_rate": 4.840128669369483e-05,
+      "loss": 0.441,
       "step": 19570
     },
     {
-      "epoch": 0.69,
-      "learning_rate": 4.849422027340932e-05,
-      "loss": 0.3088,
+      "epoch": 0.7054816736944535,
+      "grad_norm": 0.1392039656639099,
+      "learning_rate": 4.840025975068986e-05,
+      "loss": 0.4667,
       "step": 19575
     },
     {
-      "epoch": 0.69,
-      "learning_rate": 4.8493246397342896e-05,
-      "loss": 0.3191,
+      "epoch": 0.7056618733556781,
+      "grad_norm": 0.17485755681991577,
+      "learning_rate": 4.839923248886089e-05,
+      "loss": 0.4378,
       "step": 19580
     },
     {
-      "epoch": 0.69,
-      "learning_rate": 4.849227221623215e-05,
-      "loss": 0.2954,
+      "epoch": 0.7058420730169027,
+      "grad_norm": 0.14245180785655975,
+      "learning_rate": 4.8398204908221894e-05,
+      "loss": 0.445,
       "step": 19585
     },
     {
-      "epoch": 0.69,
-      "learning_rate": 4.849129773008972e-05,
-      "loss": 0.2997,
+      "epoch": 0.7060222726781273,
+      "grad_norm": 0.19247664511203766,
+      "learning_rate": 4.839717700878689e-05,
+      "loss": 0.4163,
       "step": 19590
     },
     {
-      "epoch": 0.69,
-      "learning_rate": 4.849032293892828e-05,
-      "loss": 0.3194,
+      "epoch": 0.706202472339352,
+      "grad_norm": 0.14661888778209686,
+      "learning_rate": 4.839614879056989e-05,
+      "loss": 0.4428,
       "step": 19595
     },
     {
-      "epoch": 0.69,
-      "learning_rate": 4.848934784276048e-05,
-      "loss": 0.2893,
+      "epoch": 0.7063826720005766,
+      "grad_norm": 0.2213519662618637,
+      "learning_rate": 4.839512025358488e-05,
+      "loss": 0.4248,
       "step": 19600
     },
     {
-      "epoch": 0.69,
-      "learning_rate": 4.848837244159897e-05,
-      "loss": 0.3139,
+      "epoch": 0.7065628716618013,
+      "grad_norm": 0.13426418602466583,
+      "learning_rate": 4.839409139784589e-05,
+      "loss": 0.4201,
       "step": 19605
     },
     {
-      "epoch": 0.69,
-      "learning_rate": 4.848739673545643e-05,
-      "loss": 0.3236,
+      "epoch": 0.7067430713230259,
+      "grad_norm": 0.17198221385478973,
+      "learning_rate": 4.839306222336694e-05,
+      "loss": 0.4788,
       "step": 19610
     },
     {
-      "epoch": 0.69,
-      "learning_rate": 4.8486420724345516e-05,
-      "loss": 0.3074,
+      "epoch": 0.7069232709842506,
+      "grad_norm": 0.1814018040895462,
+      "learning_rate": 4.839203273016203e-05,
+      "loss": 0.4323,
       "step": 19615
     },
     {
-      "epoch": 0.69,
-      "learning_rate": 4.8485444408278915e-05,
-      "loss": 0.3413,
+      "epoch": 0.7071034706454752,
+      "grad_norm": 0.20377859473228455,
+      "learning_rate": 4.839100291824522e-05,
+      "loss": 0.4059,
       "step": 19620
     },
     {
-      "epoch": 0.69,
-      "learning_rate": 4.8484467787269294e-05,
-      "loss": 0.3285,
+      "epoch": 0.7072836703066998,
+      "grad_norm": 0.16725854575634003,
+      "learning_rate": 4.838997278763051e-05,
+      "loss": 0.4669,
       "step": 19625
     },
     {
-      "epoch": 0.69,
-      "learning_rate": 4.848349086132933e-05,
-      "loss": 0.3101,
+      "epoch": 0.7074638699679244,
+      "grad_norm": 0.15145759284496307,
+      "learning_rate": 4.838894233833196e-05,
+      "loss": 0.4384,
       "step": 19630
     },
     {
-      "epoch": 0.69,
-      "learning_rate": 4.848251363047171e-05,
-      "loss": 0.3094,
+      "epoch": 0.707644069629149,
+      "grad_norm": 0.16167926788330078,
+      "learning_rate": 4.8387911570363596e-05,
+      "loss": 0.4536,
       "step": 19635
     },
     {
-      "epoch": 0.69,
-      "learning_rate": 4.848153609470914e-05,
-      "loss": 0.3325,
+      "epoch": 0.7078242692903738,
+      "grad_norm": 0.15461744368076324,
+      "learning_rate": 4.838688048373946e-05,
+      "loss": 0.4833,
       "step": 19640
     },
     {
-      "epoch": 0.69,
-      "learning_rate": 4.848055825405429e-05,
-      "loss": 0.2964,
+      "epoch": 0.7080044689515984,
+      "grad_norm": 0.18322111666202545,
+      "learning_rate": 4.838584907847361e-05,
+      "loss": 0.4641,
       "step": 19645
     },
     {
-      "epoch": 0.69,
-      "learning_rate": 4.847958010851986e-05,
-      "loss": 0.3134,
+      "epoch": 0.708184668612823,
+      "grad_norm": 0.15840373933315277,
+      "learning_rate": 4.83848173545801e-05,
+      "loss": 0.4341,
       "step": 19650
     },
     {
-      "epoch": 0.69,
-      "learning_rate": 4.847860165811857e-05,
-      "loss": 0.2776,
+      "epoch": 0.7083648682740477,
+      "grad_norm": 0.1756211519241333,
+      "learning_rate": 4.8383785312072974e-05,
+      "loss": 0.4358,
       "step": 19655
     },
     {
-      "epoch": 0.69,
-      "learning_rate": 4.84776229028631e-05,
-      "loss": 0.3465,
+      "epoch": 0.7085450679352723,
+      "grad_norm": 0.17935514450073242,
+      "learning_rate": 4.83827529509663e-05,
+      "loss": 0.4682,
       "step": 19660
     },
     {
-      "epoch": 0.69,
-      "learning_rate": 4.847664384276617e-05,
-      "loss": 0.3151,
+      "epoch": 0.7087252675964969,
+      "grad_norm": 0.20242036879062653,
+      "learning_rate": 4.8381720271274146e-05,
+      "loss": 0.4391,
       "step": 19665
     },
     {
-      "epoch": 0.69,
-      "learning_rate": 4.84756644778405e-05,
-      "loss": 0.3081,
+      "epoch": 0.7089054672577215,
+      "grad_norm": 0.15790332853794098,
+      "learning_rate": 4.838068727301057e-05,
+      "loss": 0.4066,
       "step": 19670
     },
     {
-      "epoch": 0.69,
-      "learning_rate": 4.847468480809879e-05,
-      "loss": 0.3069,
+      "epoch": 0.7090856669189461,
+      "grad_norm": 0.14953632652759552,
+      "learning_rate": 4.8379653956189675e-05,
+      "loss": 0.4043,
       "step": 19675
     },
     {
-      "epoch": 0.69,
-      "learning_rate": 4.847370483355377e-05,
-      "loss": 0.3152,
+      "epoch": 0.7092658665801709,
+      "grad_norm": 0.20323163270950317,
+      "learning_rate": 4.8378620320825505e-05,
+      "loss": 0.4517,
       "step": 19680
     },
     {
-      "epoch": 0.69,
-      "learning_rate": 4.847272455421816e-05,
-      "loss": 0.2902,
+      "epoch": 0.7094460662413955,
+      "grad_norm": 0.17088757455348969,
+      "learning_rate": 4.837758636693217e-05,
+      "loss": 0.4426,
       "step": 19685
     },
     {
-      "epoch": 0.69,
-      "learning_rate": 4.847174397010469e-05,
-      "loss": 0.3075,
+      "epoch": 0.7096262659026201,
+      "grad_norm": 0.15249846875667572,
+      "learning_rate": 4.837655209452374e-05,
+      "loss": 0.4364,
       "step": 19690
     },
     {
-      "epoch": 0.69,
-      "learning_rate": 4.84707630812261e-05,
-      "loss": 0.2934,
+      "epoch": 0.7098064655638447,
+      "grad_norm": 0.17006030678749084,
+      "learning_rate": 4.837551750361432e-05,
+      "loss": 0.4031,
       "step": 19695
     },
     {
-      "epoch": 0.69,
-      "learning_rate": 4.8469781887595115e-05,
-      "loss": 0.3203,
+      "epoch": 0.7099866652250694,
+      "grad_norm": 0.17397870123386383,
+      "learning_rate": 4.837448259421799e-05,
+      "loss": 0.4184,
       "step": 19700
     },
     {
-      "epoch": 0.69,
-      "learning_rate": 4.846880038922447e-05,
-      "loss": 0.3175,
+      "epoch": 0.710166864886294,
+      "grad_norm": 0.16801589727401733,
+      "learning_rate": 4.837344736634887e-05,
+      "loss": 0.4629,
       "step": 19705
     },
     {
-      "epoch": 0.69,
-      "learning_rate": 4.846781858612693e-05,
-      "loss": 0.2932,
+      "epoch": 0.7103470645475186,
+      "grad_norm": 0.18342332541942596,
+      "learning_rate": 4.8372411820021054e-05,
+      "loss": 0.4433,
       "step": 19710
     },
     {
-      "epoch": 0.69,
-      "learning_rate": 4.846683647831522e-05,
-      "loss": 0.311,
+      "epoch": 0.7105272642087432,
+      "grad_norm": 0.15994971990585327,
+      "learning_rate": 4.8371375955248644e-05,
+      "loss": 0.444,
       "step": 19715
     },
     {
-      "epoch": 0.69,
-      "learning_rate": 4.846585406580212e-05,
-      "loss": 0.3067,
+      "epoch": 0.710707463869968,
+      "grad_norm": 0.1518271565437317,
+      "learning_rate": 4.837033977204577e-05,
+      "loss": 0.4278,
       "step": 19720
     },
     {
-      "epoch": 0.69,
-      "learning_rate": 4.846487134860036e-05,
-      "loss": 0.3117,
+      "epoch": 0.7108876635311926,
+      "grad_norm": 0.18965473771095276,
+      "learning_rate": 4.836930327042654e-05,
+      "loss": 0.4386,
       "step": 19725
     },
     {
-      "epoch": 0.69,
-      "learning_rate": 4.8463888326722714e-05,
-      "loss": 0.3373,
+      "epoch": 0.7110678631924172,
+      "grad_norm": 0.20504818856716156,
+      "learning_rate": 4.8368266450405077e-05,
+      "loss": 0.4627,
       "step": 19730
     },
     {
-      "epoch": 0.69,
-      "learning_rate": 4.846290500018194e-05,
-      "loss": 0.3003,
+      "epoch": 0.7112480628536418,
+      "grad_norm": 0.1648981124162674,
+      "learning_rate": 4.83672293119955e-05,
+      "loss": 0.4572,
       "step": 19735
     },
     {
-      "epoch": 0.69,
-      "learning_rate": 4.84619213689908e-05,
-      "loss": 0.3069,
+      "epoch": 0.7114282625148665,
+      "grad_norm": 0.2660510241985321,
+      "learning_rate": 4.836619185521196e-05,
+      "loss": 0.4748,
       "step": 19740
     },
     {
-      "epoch": 0.69,
-      "learning_rate": 4.846093743316208e-05,
-      "loss": 0.3144,
+      "epoch": 0.7116084621760911,
+      "grad_norm": 0.14745326340198517,
+      "learning_rate": 4.836515408006857e-05,
+      "loss": 0.4311,
       "step": 19745
     },
     {
-      "epoch": 0.69,
-      "learning_rate": 4.845995319270855e-05,
-      "loss": 0.3183,
+      "epoch": 0.7117886618373157,
+      "grad_norm": 0.13908466696739197,
+      "learning_rate": 4.8364115986579485e-05,
+      "loss": 0.436,
       "step": 19750
     },
     {
-      "epoch": 0.7,
-      "learning_rate": 4.845896864764299e-05,
-      "loss": 0.3005,
+      "epoch": 0.7119688614985403,
+      "grad_norm": 0.16212470829486847,
+      "learning_rate": 4.8363077574758836e-05,
+      "loss": 0.4103,
       "step": 19755
     },
     {
-      "epoch": 0.7,
-      "learning_rate": 4.845798379797818e-05,
-      "loss": 0.3114,
+      "epoch": 0.7121490611597651,
+      "grad_norm": 0.17893613874912262,
+      "learning_rate": 4.836203884462078e-05,
+      "loss": 0.4971,
       "step": 19760
     },
     {
-      "epoch": 0.7,
-      "learning_rate": 4.84569986437269e-05,
-      "loss": 0.3258,
+      "epoch": 0.7123292608209897,
+      "grad_norm": 0.18436241149902344,
+      "learning_rate": 4.836099979617947e-05,
+      "loss": 0.4828,
       "step": 19765
     },
     {
-      "epoch": 0.7,
-      "learning_rate": 4.845601318490196e-05,
-      "loss": 0.302,
+      "epoch": 0.7125094604822143,
+      "grad_norm": 0.16563914716243744,
+      "learning_rate": 4.835996042944907e-05,
+      "loss": 0.4026,
       "step": 19770
     },
     {
-      "epoch": 0.7,
-      "learning_rate": 4.845502742151615e-05,
-      "loss": 0.3038,
+      "epoch": 0.7126896601434389,
+      "grad_norm": 0.1609710305929184,
+      "learning_rate": 4.835892074444372e-05,
+      "loss": 0.4159,
       "step": 19775
     },
     {
-      "epoch": 0.7,
-      "learning_rate": 4.845404135358226e-05,
-      "loss": 0.3218,
+      "epoch": 0.7128698598046636,
+      "grad_norm": 0.14441636204719543,
+      "learning_rate": 4.8357880741177605e-05,
+      "loss": 0.4171,
       "step": 19780
     },
     {
-      "epoch": 0.7,
-      "learning_rate": 4.8453054981113106e-05,
-      "loss": 0.3271,
+      "epoch": 0.7130500594658882,
+      "grad_norm": 0.14850714802742004,
+      "learning_rate": 4.835684041966488e-05,
+      "loss": 0.4341,
       "step": 19785
     },
     {
-      "epoch": 0.7,
-      "learning_rate": 4.845206830412149e-05,
-      "loss": 0.3154,
+      "epoch": 0.7132302591271128,
+      "grad_norm": 0.17348629236221313,
+      "learning_rate": 4.835579977991973e-05,
+      "loss": 0.4328,
       "step": 19790
     },
     {
-      "epoch": 0.7,
-      "learning_rate": 4.8451081322620215e-05,
-      "loss": 0.3069,
+      "epoch": 0.7134104587883375,
+      "grad_norm": 0.18963472545146942,
+      "learning_rate": 4.8354758821956325e-05,
+      "loss": 0.418,
       "step": 19795
     },
     {
-      "epoch": 0.7,
-      "learning_rate": 4.845009403662211e-05,
-      "loss": 0.3261,
+      "epoch": 0.7135906584495622,
+      "grad_norm": 0.15153972804546356,
+      "learning_rate": 4.8353717545788855e-05,
+      "loss": 0.4423,
       "step": 19800
     },
     {
-      "epoch": 0.7,
-      "learning_rate": 4.844910644613998e-05,
-      "loss": 0.2997,
+      "epoch": 0.7137708581107868,
+      "grad_norm": 0.17662297189235687,
+      "learning_rate": 4.835267595143151e-05,
+      "loss": 0.4625,
       "step": 19805
     },
     {
-      "epoch": 0.7,
-      "learning_rate": 4.8448118551186664e-05,
-      "loss": 0.2808,
+      "epoch": 0.7139510577720114,
+      "grad_norm": 0.1881992071866989,
+      "learning_rate": 4.8351842446858454e-05,
+      "loss": 0.4305,
       "step": 19810
     },
     {
-      "epoch": 0.7,
-      "learning_rate": 4.8447130351774986e-05,
-      "loss": 0.2954,
+      "epoch": 0.714131257433236,
+      "grad_norm": 0.1803097277879715,
+      "learning_rate": 4.8350800279795086e-05,
+      "loss": 0.4069,
       "step": 19815
     },
     {
-      "epoch": 0.7,
-      "learning_rate": 4.844614184791776e-05,
-      "loss": 0.3093,
+      "epoch": 0.7143114570944606,
+      "grad_norm": 0.14879196882247925,
+      "learning_rate": 4.834975779458158e-05,
+      "loss": 0.447,
       "step": 19820
     },
     {
-      "epoch": 0.7,
-      "learning_rate": 4.844515303962784e-05,
-      "loss": 0.311,
+      "epoch": 0.7144916567556853,
+      "grad_norm": 0.1836402863264084,
+      "learning_rate": 4.834871499123216e-05,
+      "loss": 0.4518,
       "step": 19825
     },
     {
-      "epoch": 0.7,
-      "learning_rate": 4.844416392691806e-05,
-      "loss": 0.2939,
+      "epoch": 0.7146718564169099,
+      "grad_norm": 0.16930605471134186,
+      "learning_rate": 4.8347671869761e-05,
+      "loss": 0.4471,
       "step": 19830
     },
     {
-      "epoch": 0.7,
-      "learning_rate": 4.844317450980126e-05,
-      "loss": 0.3139,
+      "epoch": 0.7148520560781346,
+      "grad_norm": 0.14788581430912018,
+      "learning_rate": 4.8346837143546e-05,
+      "loss": 0.4639,
       "step": 19835
     },
     {
-      "epoch": 0.7,
-      "learning_rate": 4.844218478829028e-05,
-      "loss": 0.2842,
+      "epoch": 0.7150322557393592,
+      "grad_norm": 0.16612641513347626,
+      "learning_rate": 4.834579344949157e-05,
+      "loss": 0.4454,
       "step": 19840
     },
     {
-      "epoch": 0.7,
-      "learning_rate": 4.844119476239799e-05,
-      "loss": 0.2998,
+      "epoch": 0.7152124554005839,
+      "grad_norm": 0.17854836583137512,
+      "learning_rate": 4.834474943735522e-05,
+      "loss": 0.3823,
       "step": 19845
     },
     {
-      "epoch": 0.7,
-      "learning_rate": 4.844020443213724e-05,
-      "loss": 0.3093,
+      "epoch": 0.7153926550618085,
+      "grad_norm": 0.1602153182029724,
+      "learning_rate": 4.834370510715118e-05,
+      "loss": 0.4561,
       "step": 19850
     },
     {
-      "epoch": 0.7,
-      "learning_rate": 4.8439213797520874e-05,
-      "loss": 0.2712,
+      "epoch": 0.7155728547230331,
+      "grad_norm": 0.1520301103591919,
+      "learning_rate": 4.8342660458893677e-05,
+      "loss": 0.4599,
       "step": 19855
     },
     {
-      "epoch": 0.7,
-      "learning_rate": 4.8438222858561764e-05,
-      "loss": 0.3019,
+      "epoch": 0.7157530543842577,
+      "grad_norm": 0.16628147661685944,
+      "learning_rate": 4.834161549259695e-05,
+      "loss": 0.4178,
       "step": 19860
     },
     {
-      "epoch": 0.7,
-      "learning_rate": 4.8437231615272775e-05,
-      "loss": 0.3105,
+      "epoch": 0.7159332540454824,
+      "grad_norm": 0.19706813991069794,
+      "learning_rate": 4.8340570208275224e-05,
+      "loss": 0.4677,
       "step": 19865
     },
     {
-      "epoch": 0.7,
-      "learning_rate": 4.8436240067666784e-05,
-      "loss": 0.3299,
+      "epoch": 0.716113453706707,
+      "grad_norm": 0.1859341561794281,
+      "learning_rate": 4.833952460594275e-05,
+      "loss": 0.4665,
       "step": 19870
     },
     {
-      "epoch": 0.7,
-      "learning_rate": 4.843524821575666e-05,
-      "loss": 0.308,
+      "epoch": 0.7162936533679317,
+      "grad_norm": 0.16585981845855713,
+      "learning_rate": 4.8338478685613775e-05,
+      "loss": 0.4378,
       "step": 19875
     },
     {
-      "epoch": 0.7,
-      "learning_rate": 4.8434256059555284e-05,
-      "loss": 0.2992,
+      "epoch": 0.7164738530291563,
+      "grad_norm": 0.15817083418369293,
+      "learning_rate": 4.8337432447302544e-05,
+      "loss": 0.453,
       "step": 19880
     },
     {
-      "epoch": 0.7,
-      "learning_rate": 4.843326359907553e-05,
-      "loss": 0.3057,
+      "epoch": 0.716654052690381,
+      "grad_norm": 0.1971365064382553,
+      "learning_rate": 4.833638589102332e-05,
+      "loss": 0.4405,
       "step": 19885
     },
     {
-      "epoch": 0.7,
-      "learning_rate": 4.84322708343303e-05,
-      "loss": 0.2805,
+      "epoch": 0.7168342523516056,
+      "grad_norm": 0.16757121682167053,
+      "learning_rate": 4.8335339016790346e-05,
+      "loss": 0.4392,
       "step": 19890
     },
     {
-      "epoch": 0.7,
-      "learning_rate": 4.843127776533247e-05,
-      "loss": 0.2985,
+      "epoch": 0.7170144520128302,
+      "grad_norm": 0.20326150953769684,
+      "learning_rate": 4.8334291824617905e-05,
+      "loss": 0.4335,
       "step": 19895
     },
     {
-      "epoch": 0.7,
-      "learning_rate": 4.843028439209494e-05,
-      "loss": 0.3125,
+      "epoch": 0.7171946516740548,
+      "grad_norm": 0.16170883178710938,
+      "learning_rate": 4.8333244314520254e-05,
+      "loss": 0.4238,
       "step": 19900
     },
     {
-      "epoch": 0.7,
-      "learning_rate": 4.8429290714630605e-05,
-      "loss": 0.3265,
+      "epoch": 0.7173748513352795,
+      "grad_norm": 0.17413537204265594,
+      "learning_rate": 4.833219648651166e-05,
+      "loss": 0.479,
       "step": 19905
     },
     {
-      "epoch": 0.7,
-      "learning_rate": 4.8428296732952376e-05,
-      "loss": 0.302,
+      "epoch": 0.7175550509965041,
+      "grad_norm": 0.17065051198005676,
+      "learning_rate": 4.8331148340606416e-05,
+      "loss": 0.4408,
       "step": 19910
     },
     {
-      "epoch": 0.7,
-      "learning_rate": 4.842730244707315e-05,
-      "loss": 0.3318,
+      "epoch": 0.7177352506577288,
+      "grad_norm": 0.22111043334007263,
+      "learning_rate": 4.833009987681878e-05,
+      "loss": 0.4456,
       "step": 19915
     },
     {
-      "epoch": 0.7,
-      "learning_rate": 4.8426307857005845e-05,
-      "loss": 0.2854,
+      "epoch": 0.7179154503189534,
+      "grad_norm": 0.14916540682315826,
+      "learning_rate": 4.832905109516306e-05,
+      "loss": 0.4503,
       "step": 19920
     },
     {
-      "epoch": 0.7,
-      "learning_rate": 4.842531296276337e-05,
-      "loss": 0.3413,
+      "epoch": 0.718095649980178,
+      "grad_norm": 0.16950784623622894,
+      "learning_rate": 4.832800199565353e-05,
+      "loss": 0.4433,
       "step": 19925
     },
     {
-      "epoch": 0.7,
-      "learning_rate": 4.8424317764358645e-05,
-      "loss": 0.3107,
+      "epoch": 0.7182758496414027,
+      "grad_norm": 0.15380850434303284,
+      "learning_rate": 4.8326952578304496e-05,
+      "loss": 0.4298,
       "step": 19930
     },
     {
-      "epoch": 0.7,
-      "learning_rate": 4.842332226180459e-05,
-      "loss": 0.2836,
+      "epoch": 0.7184560493026273,
+      "grad_norm": 0.16172832250595093,
+      "learning_rate": 4.832590284313024e-05,
+      "loss": 0.4582,
       "step": 19935
     },
     {
-      "epoch": 0.7,
-      "learning_rate": 4.842232645511413e-05,
-      "loss": 0.3137,
+      "epoch": 0.7186362489638519,
+      "grad_norm": 0.17244853079319,
+      "learning_rate": 4.832485279014508e-05,
+      "loss": 0.471,
       "step": 19940
     },
     {
-      "epoch": 0.7,
-      "learning_rate": 4.8421330344300196e-05,
-      "loss": 0.3102,
+      "epoch": 0.7188164486250765,
+      "grad_norm": 0.14485011994838715,
+      "learning_rate": 4.832380241936332e-05,
+      "loss": 0.4233,
       "step": 19945
     },
     {
-      "epoch": 0.7,
-      "learning_rate": 4.842033392937573e-05,
-      "loss": 0.3285,
+      "epoch": 0.7189966482863013,
+      "grad_norm": 0.1884828507900238,
+      "learning_rate": 4.832275173079926e-05,
+      "loss": 0.4718,
       "step": 19950
     },
     {
-      "epoch": 0.7,
-      "learning_rate": 4.841933721035365e-05,
-      "loss": 0.3212,
+      "epoch": 0.7191768479475259,
+      "grad_norm": 0.17681987583637238,
+      "learning_rate": 4.832170072446723e-05,
+      "loss": 0.4592,
       "step": 19955
     },
     {
-      "epoch": 0.7,
-      "learning_rate": 4.8418340187246926e-05,
-      "loss": 0.2831,
+      "epoch": 0.7193570476087505,
+      "grad_norm": 0.1624554544687271,
+      "learning_rate": 4.832064940038154e-05,
+      "loss": 0.4429,
       "step": 19960
     },
     {
-      "epoch": 0.7,
-      "learning_rate": 4.8417342860068476e-05,
-      "loss": 0.3049,
+      "epoch": 0.7195372472699751,
+      "grad_norm": 0.19783754646778107,
+      "learning_rate": 4.831959775855651e-05,
+      "loss": 0.4366,
       "step": 19965
     },
     {
-      "epoch": 0.7,
-      "learning_rate": 4.8416345228831266e-05,
-      "loss": 0.3054,
+      "epoch": 0.7197174469311998,
+      "grad_norm": 0.15661893784999847,
+      "learning_rate": 4.831854579900649e-05,
+      "loss": 0.4336,
       "step": 19970
     },
     {
-      "epoch": 0.7,
-      "learning_rate": 4.8415347293548245e-05,
-      "loss": 0.2954,
+      "epoch": 0.7198976465924244,
+      "grad_norm": 0.1397693157196045,
+      "learning_rate": 4.831749352174578e-05,
+      "loss": 0.427,
       "step": 19975
     },
     {
-      "epoch": 0.7,
-      "learning_rate": 4.841434905423237e-05,
-      "loss": 0.2984,
+      "epoch": 0.720077846253649,
+      "grad_norm": 0.175705686211586,
+      "learning_rate": 4.831644092678875e-05,
+      "loss": 0.4108,
       "step": 19980
     },
     {
-      "epoch": 0.7,
-      "learning_rate": 4.8413350510896605e-05,
-      "loss": 0.3073,
+      "epoch": 0.7202580459148736,
+      "grad_norm": 0.1558927595615387,
+      "learning_rate": 4.831538801414972e-05,
+      "loss": 0.4495,
       "step": 19985
     },
     {
-      "epoch": 0.7,
-      "learning_rate": 4.841235166355391e-05,
-      "loss": 0.3117,
+      "epoch": 0.7204382455760984,
+      "grad_norm": 0.17159906029701233,
+      "learning_rate": 4.831433478384304e-05,
+      "loss": 0.4014,
       "step": 19990
     },
     {
-      "epoch": 0.7,
-      "learning_rate": 4.841135251221726e-05,
-      "loss": 0.317,
+      "epoch": 0.720618445237323,
+      "grad_norm": 0.13353735208511353,
+      "learning_rate": 4.831328123588307e-05,
+      "loss": 0.3997,
       "step": 19995
     },
     {
-      "epoch": 0.7,
-      "learning_rate": 4.8410353056899626e-05,
-      "loss": 0.3122,
+      "epoch": 0.7207986448985476,
+      "grad_norm": 0.1744467169046402,
+      "learning_rate": 4.8312227370284155e-05,
+      "loss": 0.4517,
       "step": 20000
     },
     {
-      "epoch": 0.7,
-      "eval_loss": 0.3004976212978363,
-      "eval_runtime": 10.5473,
-      "eval_samples_per_second": 9.481,
-      "eval_steps_per_second": 9.481,
+      "epoch": 0.7207986448985476,
+      "eval_loss": 0.46302729845046997,
+      "eval_runtime": 3.5385,
+      "eval_samples_per_second": 28.261,
+      "eval_steps_per_second": 7.065,
       "step": 20000
     },
     {
-      "epoch": 0.7,
-      "learning_rate": 4.8409353297613994e-05,
-      "loss": 0.3367,
+      "epoch": 0.7209788445597722,
+      "grad_norm": 0.14821864664554596,
+      "learning_rate": 4.831117318706065e-05,
+      "loss": 0.4504,
       "step": 20005
     },
     {
-      "epoch": 0.7,
-      "learning_rate": 4.840835323437332e-05,
-      "loss": 0.2945,
+      "epoch": 0.7211590442209969,
+      "grad_norm": 0.16435876488685608,
+      "learning_rate": 4.8310118686226926e-05,
+      "loss": 0.4398,
       "step": 20010
     },
     {
-      "epoch": 0.7,
-      "learning_rate": 4.840735286719061e-05,
-      "loss": 0.2952,
+      "epoch": 0.7213392438822215,
+      "grad_norm": 0.17198477685451508,
+      "learning_rate": 4.830906386779735e-05,
+      "loss": 0.4684,
       "step": 20015
     },
     {
-      "epoch": 0.7,
-      "learning_rate": 4.8406352196078864e-05,
-      "loss": 0.2975,
+      "epoch": 0.7215194435434461,
+      "grad_norm": 0.1611550748348236,
+      "learning_rate": 4.830800873178629e-05,
+      "loss": 0.4502,
       "step": 20020
     },
     {
-      "epoch": 0.7,
-      "learning_rate": 4.8405351221051044e-05,
-      "loss": 0.3112,
+      "epoch": 0.7216996432046707,
+      "grad_norm": 0.18277272582054138,
+      "learning_rate": 4.830695327820812e-05,
+      "loss": 0.4442,
       "step": 20025
     },
     {
-      "epoch": 0.7,
-      "learning_rate": 4.840434994212017e-05,
-      "loss": 0.2915,
+      "epoch": 0.7218798428658955,
+      "grad_norm": 0.21301540732383728,
+      "learning_rate": 4.8305897507077236e-05,
+      "loss": 0.4286,
       "step": 20030
     },
     {
-      "epoch": 0.7,
-      "learning_rate": 4.8403348359299236e-05,
-      "loss": 0.3102,
+      "epoch": 0.7220600425271201,
+      "grad_norm": 0.19387184083461761,
+      "learning_rate": 4.8304841418407995e-05,
+      "loss": 0.4271,
       "step": 20035
     },
     {
-      "epoch": 0.71,
-      "learning_rate": 4.840234647260125e-05,
-      "loss": 0.3047,
+      "epoch": 0.7222402421883447,
+      "grad_norm": 0.18802416324615479,
+      "learning_rate": 4.8303785012214814e-05,
+      "loss": 0.4276,
       "step": 20040
     },
     {
-      "epoch": 0.71,
-      "learning_rate": 4.8401344282039204e-05,
-      "loss": 0.3044,
+      "epoch": 0.7224204418495693,
+      "grad_norm": 0.18947114050388336,
+      "learning_rate": 4.8302728288512064e-05,
+      "loss": 0.4585,
       "step": 20045
     },
     {
-      "epoch": 0.71,
-      "learning_rate": 4.840034178762613e-05,
-      "loss": 0.282,
+      "epoch": 0.722600641510794,
+      "grad_norm": 0.18048980832099915,
+      "learning_rate": 4.8301671247314165e-05,
+      "loss": 0.4562,
       "step": 20050
     },
     {
-      "epoch": 0.71,
-      "learning_rate": 4.839933898937505e-05,
-      "loss": 0.3284,
+      "epoch": 0.7227808411720186,
+      "grad_norm": 0.16718988120555878,
+      "learning_rate": 4.830061388863549e-05,
+      "loss": 0.4233,
       "step": 20055
     },
     {
-      "epoch": 0.71,
-      "learning_rate": 4.8398335887298965e-05,
-      "loss": 0.2998,
+      "epoch": 0.7229610408332432,
+      "grad_norm": 0.17260010540485382,
+      "learning_rate": 4.8299556212490474e-05,
+      "loss": 0.3985,
       "step": 20060
     },
     {
-      "epoch": 0.71,
-      "learning_rate": 4.839733248141092e-05,
-      "loss": 0.3012,
+      "epoch": 0.7231412404944678,
+      "grad_norm": 0.1411719173192978,
+      "learning_rate": 4.829849821889352e-05,
+      "loss": 0.4143,
       "step": 20065
     },
     {
-      "epoch": 0.71,
-      "learning_rate": 4.839632877172392e-05,
-      "loss": 0.291,
+      "epoch": 0.7233214401556926,
+      "grad_norm": 0.19284233450889587,
+      "learning_rate": 4.829743990785903e-05,
+      "loss": 0.4585,
       "step": 20070
     },
     {
-      "epoch": 0.71,
-      "learning_rate": 4.839532475825101e-05,
-      "loss": 0.3247,
+      "epoch": 0.7235016398169172,
+      "grad_norm": 0.13794685900211334,
+      "learning_rate": 4.829638127940143e-05,
+      "loss": 0.4115,
       "step": 20075
     },
     {
-      "epoch": 0.71,
-      "learning_rate": 4.839432044100523e-05,
-      "loss": 0.3074,
+      "epoch": 0.7236818394781418,
+      "grad_norm": 0.17087949812412262,
+      "learning_rate": 4.8295322333535146e-05,
+      "loss": 0.4631,
       "step": 20080
     },
     {
-      "epoch": 0.71,
-      "learning_rate": 4.839331581999962e-05,
-      "loss": 0.2962,
+      "epoch": 0.7238620391393664,
+      "grad_norm": 0.18118272721767426,
+      "learning_rate": 4.829426307027461e-05,
+      "loss": 0.4688,
       "step": 20085
     },
     {
-      "epoch": 0.71,
-      "learning_rate": 4.839231089524721e-05,
-      "loss": 0.3216,
+      "epoch": 0.724042238800591,
+      "grad_norm": 0.1501915603876114,
+      "learning_rate": 4.829320348963425e-05,
+      "loss": 0.4322,
       "step": 20090
     },
     {
-      "epoch": 0.71,
-      "learning_rate": 4.8391305666761065e-05,
-      "loss": 0.3007,
+      "epoch": 0.7242224384618157,
+      "grad_norm": 0.1526506394147873,
+      "learning_rate": 4.8292143591628494e-05,
+      "loss": 0.4439,
       "step": 20095
     },
     {
-      "epoch": 0.71,
-      "learning_rate": 4.8390300134554236e-05,
-      "loss": 0.2944,
+      "epoch": 0.7244026381230403,
+      "grad_norm": 0.16467134654521942,
+      "learning_rate": 4.82910833762718e-05,
+      "loss": 0.4476,
       "step": 20100
     },
     {
-      "epoch": 0.71,
-      "learning_rate": 4.8389294298639766e-05,
-      "loss": 0.2958,
+      "epoch": 0.7245828377842649,
+      "grad_norm": 0.199048712849617,
+      "learning_rate": 4.82900228435786e-05,
+      "loss": 0.4174,
       "step": 20105
     },
     {
-      "epoch": 0.71,
-      "learning_rate": 4.8388288159030727e-05,
-      "loss": 0.3113,
+      "epoch": 0.7247630374454896,
+      "grad_norm": 0.22329410910606384,
+      "learning_rate": 4.828896199356335e-05,
+      "loss": 0.4558,
       "step": 20110
     },
     {
-      "epoch": 0.71,
-      "learning_rate": 4.838728171574018e-05,
-      "loss": 0.2786,
+      "epoch": 0.7249432371067143,
+      "grad_norm": 0.15609164535999298,
+      "learning_rate": 4.82879008262405e-05,
+      "loss": 0.4342,
       "step": 20115
     },
     {
-      "epoch": 0.71,
-      "learning_rate": 4.83862749687812e-05,
-      "loss": 0.3319,
+      "epoch": 0.7251234367679389,
+      "grad_norm": 0.15675166249275208,
+      "learning_rate": 4.8286839341624515e-05,
+      "loss": 0.4524,
       "step": 20120
     },
     {
-      "epoch": 0.71,
-      "learning_rate": 4.8385267918166844e-05,
-      "loss": 0.2875,
+      "epoch": 0.7253036364291635,
+      "grad_norm": 0.16207316517829895,
+      "learning_rate": 4.828577753972984e-05,
+      "loss": 0.4452,
       "step": 20125
     },
     {
-      "epoch": 0.71,
-      "learning_rate": 4.838426056391019e-05,
-      "loss": 0.32,
+      "epoch": 0.7254838360903881,
+      "grad_norm": 0.15296733379364014,
+      "learning_rate": 4.8284715420570964e-05,
+      "loss": 0.3956,
       "step": 20130
     },
     {
-      "epoch": 0.71,
-      "learning_rate": 4.838325290602434e-05,
-      "loss": 0.3277,
+      "epoch": 0.7256640357516128,
+      "grad_norm": 0.16610366106033325,
+      "learning_rate": 4.8283652984162345e-05,
+      "loss": 0.4279,
       "step": 20135
     },
     {
-      "epoch": 0.71,
-      "learning_rate": 4.8382244944522346e-05,
-      "loss": 0.3221,
+      "epoch": 0.7258442354128374,
+      "grad_norm": 0.1927533745765686,
+      "learning_rate": 4.828259023051847e-05,
+      "loss": 0.4534,
       "step": 20140
     },
     {
-      "epoch": 0.71,
-      "learning_rate": 4.838123667941731e-05,
-      "loss": 0.3474,
+      "epoch": 0.7260244350740621,
+      "grad_norm": 0.13150177896022797,
+      "learning_rate": 4.82815271596538e-05,
+      "loss": 0.4348,
       "step": 20145
     },
     {
-      "epoch": 0.71,
-      "learning_rate": 4.8380228110722336e-05,
-      "loss": 0.2811,
+      "epoch": 0.7262046347352867,
+      "grad_norm": 0.15792779624462128,
+      "learning_rate": 4.8280463771582835e-05,
+      "loss": 0.4647,
       "step": 20150
     },
     {
-      "epoch": 0.71,
-      "learning_rate": 4.83792192384505e-05,
-      "loss": 0.284,
+      "epoch": 0.7263848343965114,
+      "grad_norm": 0.16151747107505798,
+      "learning_rate": 4.8279400066320055e-05,
+      "loss": 0.4718,
       "step": 20155
     },
     {
-      "epoch": 0.71,
-      "learning_rate": 4.8378210062614916e-05,
-      "loss": 0.295,
+      "epoch": 0.726565034057736,
+      "grad_norm": 0.14716650545597076,
+      "learning_rate": 4.827833604387996e-05,
+      "loss": 0.4125,
       "step": 20160
     },
     {
-      "epoch": 0.71,
-      "learning_rate": 4.837720058322868e-05,
-      "loss": 0.3091,
+      "epoch": 0.7267452337189606,
+      "grad_norm": 0.16372278332710266,
+      "learning_rate": 4.827727170427704e-05,
+      "loss": 0.4419,
       "step": 20165
     },
     {
-      "epoch": 0.71,
-      "learning_rate": 4.8376190800304896e-05,
-      "loss": 0.3085,
+      "epoch": 0.7269254333801852,
+      "grad_norm": 0.15306620299816132,
+      "learning_rate": 4.82762070475258e-05,
+      "loss": 0.4368,
       "step": 20170
     },
     {
-      "epoch": 0.71,
-      "learning_rate": 4.8375180713856684e-05,
-      "loss": 0.3105,
+      "epoch": 0.7271056330414098,
+      "grad_norm": 0.13666532933712006,
+      "learning_rate": 4.827514207364075e-05,
+      "loss": 0.4406,
       "step": 20175
     },
     {
-      "epoch": 0.71,
-      "learning_rate": 4.837417032389715e-05,
-      "loss": 0.3035,
+      "epoch": 0.7272858327026345,
+      "grad_norm": 0.19249454140663147,
+      "learning_rate": 4.8274076782636393e-05,
+      "loss": 0.4606,
       "step": 20180
     },
     {
-      "epoch": 0.71,
-      "learning_rate": 4.837315963043943e-05,
-      "loss": 0.2796,
+      "epoch": 0.7274660323638592,
+      "grad_norm": 0.1536799520254135,
+      "learning_rate": 4.827301117452725e-05,
+      "loss": 0.4276,
       "step": 20185
     },
     {
-      "epoch": 0.71,
-      "learning_rate": 4.8372148633496614e-05,
-      "loss": 0.299,
+      "epoch": 0.7276462320250838,
+      "grad_norm": 0.15559114515781403,
+      "learning_rate": 4.8271945249327825e-05,
+      "loss": 0.4438,
       "step": 20190
     },
     {
-      "epoch": 0.71,
-      "learning_rate": 4.8371137333081866e-05,
-      "loss": 0.3032,
+      "epoch": 0.7278264316863085,
+      "grad_norm": 0.15832901000976562,
+      "learning_rate": 4.827087900705266e-05,
+      "loss": 0.4691,
       "step": 20195
     },
     {
-      "epoch": 0.71,
-      "learning_rate": 4.837012572920831e-05,
-      "loss": 0.33,
+      "epoch": 0.7280066313475331,
+      "grad_norm": 0.20250186324119568,
+      "learning_rate": 4.826981244771627e-05,
+      "loss": 0.4298,
       "step": 20200
     },
     {
-      "epoch": 0.71,
-      "learning_rate": 4.836911382188906e-05,
-      "loss": 0.2876,
+      "epoch": 0.7281868310087577,
+      "grad_norm": 0.17052537202835083,
+      "learning_rate": 4.826874557133319e-05,
+      "loss": 0.4224,
       "step": 20205
     },
     {
-      "epoch": 0.71,
-      "learning_rate": 4.8368101611137265e-05,
-      "loss": 0.3226,
+      "epoch": 0.7283670306699823,
+      "grad_norm": 0.15635935962200165,
+      "learning_rate": 4.826767837791796e-05,
+      "loss": 0.4561,
       "step": 20210
     },
     {
-      "epoch": 0.71,
-      "learning_rate": 4.836708909696608e-05,
-      "loss": 0.3189,
+      "epoch": 0.7285472303312069,
+      "grad_norm": 0.15289193391799927,
+      "learning_rate": 4.826661086748512e-05,
+      "loss": 0.4283,
       "step": 20215
     },
     {
-      "epoch": 0.71,
-      "learning_rate": 4.836607627938864e-05,
-      "loss": 0.3081,
+      "epoch": 0.7287274299924316,
+      "grad_norm": 0.16655376553535461,
+      "learning_rate": 4.82655430400492e-05,
+      "loss": 0.4535,
       "step": 20220
     },
     {
-      "epoch": 0.71,
-      "learning_rate": 4.836506315841809e-05,
-      "loss": 0.307,
+      "epoch": 0.7289076296536563,
+      "grad_norm": 0.18925489485263824,
+      "learning_rate": 4.826447489562477e-05,
+      "loss": 0.4096,
       "step": 20225
     },
     {
-      "epoch": 0.71,
-      "learning_rate": 4.8364049734067606e-05,
-      "loss": 0.3143,
+      "epoch": 0.7290878293148809,
+      "grad_norm": 0.15604831278324127,
+      "learning_rate": 4.826340643422637e-05,
+      "loss": 0.4379,
       "step": 20230
     },
     {
-      "epoch": 0.71,
-      "learning_rate": 4.8363036006350324e-05,
-      "loss": 0.3044,
+      "epoch": 0.7292680289761055,
+      "grad_norm": 0.16162985563278198,
+      "learning_rate": 4.826233765586856e-05,
+      "loss": 0.4017,
       "step": 20235
     },
     {
-      "epoch": 0.71,
-      "learning_rate": 4.836202197527943e-05,
-      "loss": 0.3096,
+      "epoch": 0.7294482286373302,
+      "grad_norm": 0.17815397679805756,
+      "learning_rate": 4.826126856056591e-05,
+      "loss": 0.4229,
       "step": 20240
     },
     {
-      "epoch": 0.71,
-      "learning_rate": 4.836100764086806e-05,
-      "loss": 0.2831,
+      "epoch": 0.7296284282985548,
+      "grad_norm": 0.17258299887180328,
+      "learning_rate": 4.826019914833297e-05,
+      "loss": 0.4471,
       "step": 20245
     },
     {
-      "epoch": 0.71,
-      "learning_rate": 4.835999300312941e-05,
-      "loss": 0.288,
+      "epoch": 0.7298086279597794,
+      "grad_norm": 0.1543063521385193,
+      "learning_rate": 4.8259129419184326e-05,
+      "loss": 0.4246,
       "step": 20250
     },
     {
-      "epoch": 0.71,
-      "learning_rate": 4.835897806207664e-05,
-      "loss": 0.3392,
+      "epoch": 0.729988827621004,
+      "grad_norm": 0.2001371681690216,
+      "learning_rate": 4.8258059373134546e-05,
+      "loss": 0.437,
       "step": 20255
     },
     {
-      "epoch": 0.71,
-      "learning_rate": 4.835796281772295e-05,
-      "loss": 0.3071,
+      "epoch": 0.7301690272822287,
+      "grad_norm": 0.16070033609867096,
+      "learning_rate": 4.8256989010198215e-05,
+      "loss": 0.4449,
       "step": 20260
     },
     {
-      "epoch": 0.71,
-      "learning_rate": 4.835694727008149e-05,
-      "loss": 0.3295,
+      "epoch": 0.7303492269434534,
+      "grad_norm": 0.16648970544338226,
+      "learning_rate": 4.8255918330389906e-05,
+      "loss": 0.4399,
       "step": 20265
     },
     {
-      "epoch": 0.71,
-      "learning_rate": 4.835593141916546e-05,
-      "loss": 0.3131,
+      "epoch": 0.730529426604678,
+      "grad_norm": 0.2026122361421585,
+      "learning_rate": 4.8254847333724204e-05,
+      "loss": 0.446,
       "step": 20270
     },
     {
-      "epoch": 0.71,
-      "learning_rate": 4.8354915264988065e-05,
-      "loss": 0.3186,
+      "epoch": 0.7307096262659026,
+      "grad_norm": 0.1755245178937912,
+      "learning_rate": 4.8253776020215725e-05,
+      "loss": 0.4188,
       "step": 20275
     },
     {
-      "epoch": 0.71,
-      "learning_rate": 4.835389880756248e-05,
-      "loss": 0.3357,
+      "epoch": 0.7308898259271273,
+      "grad_norm": 0.15573327243328094,
+      "learning_rate": 4.825270438987904e-05,
+      "loss": 0.4301,
       "step": 20280
     },
     {
-      "epoch": 0.71,
-      "learning_rate": 4.8352882046901915e-05,
-      "loss": 0.291,
+      "epoch": 0.7310700255883519,
+      "grad_norm": 0.17587211728096008,
+      "learning_rate": 4.825163244272876e-05,
+      "loss": 0.4094,
       "step": 20285
     },
     {
-      "epoch": 0.71,
-      "learning_rate": 4.8351864983019556e-05,
-      "loss": 0.3085,
+      "epoch": 0.7312502252495765,
+      "grad_norm": 0.15777091681957245,
+      "learning_rate": 4.825056017877949e-05,
+      "loss": 0.446,
       "step": 20290
     },
     {
-      "epoch": 0.71,
-      "learning_rate": 4.835084761592863e-05,
-      "loss": 0.3044,
+      "epoch": 0.7314304249108011,
+      "grad_norm": 0.1694624423980713,
+      "learning_rate": 4.824948759804584e-05,
+      "loss": 0.4262,
       "step": 20295
     },
     {
-      "epoch": 0.71,
-      "learning_rate": 4.8349829945642336e-05,
-      "loss": 0.3227,
+      "epoch": 0.7316106245720259,
+      "grad_norm": 0.15097464621067047,
+      "learning_rate": 4.824841470054242e-05,
+      "loss": 0.4455,
       "step": 20300
     },
     {
-      "epoch": 0.71,
-      "learning_rate": 4.8348811972173877e-05,
-      "loss": 0.3336,
+      "epoch": 0.7317908242332505,
+      "grad_norm": 0.1695692390203476,
+      "learning_rate": 4.824734148628386e-05,
+      "loss": 0.4678,
       "step": 20305
     },
     {
-      "epoch": 0.71,
-      "learning_rate": 4.834779369553649e-05,
-      "loss": 0.286,
+      "epoch": 0.7319710238944751,
+      "grad_norm": 0.18447819352149963,
+      "learning_rate": 4.824626795528476e-05,
+      "loss": 0.4814,
       "step": 20310
     },
     {
-      "epoch": 0.71,
-      "learning_rate": 4.8346775115743394e-05,
-      "loss": 0.2999,
+      "epoch": 0.7321512235556997,
+      "grad_norm": 0.15823550522327423,
+      "learning_rate": 4.8245194107559774e-05,
+      "loss": 0.4266,
       "step": 20315
     },
     {
-      "epoch": 0.71,
-      "learning_rate": 4.8345756232807805e-05,
-      "loss": 0.3127,
+      "epoch": 0.7323314232169243,
+      "grad_norm": 0.17037831246852875,
+      "learning_rate": 4.824411994312351e-05,
+      "loss": 0.4264,
       "step": 20320
     },
     {
-      "epoch": 0.72,
-      "learning_rate": 4.834473704674296e-05,
-      "loss": 0.3193,
+      "epoch": 0.732511622878149,
+      "grad_norm": 0.17536459863185883,
+      "learning_rate": 4.824304546199061e-05,
+      "loss": 0.4311,
       "step": 20325
     },
     {
-      "epoch": 0.72,
-      "learning_rate": 4.8343717557562076e-05,
-      "loss": 0.3123,
+      "epoch": 0.7326918225393736,
+      "grad_norm": 0.1618126928806305,
+      "learning_rate": 4.824197066417572e-05,
+      "loss": 0.4535,
       "step": 20330
     },
     {
-      "epoch": 0.72,
-      "learning_rate": 4.8342697765278424e-05,
-      "loss": 0.2938,
+      "epoch": 0.7328720222005982,
+      "grad_norm": 0.19258461892604828,
+      "learning_rate": 4.824089554969348e-05,
+      "loss": 0.4267,
       "step": 20335
     },
     {
-      "epoch": 0.72,
-      "learning_rate": 4.834167766990521e-05,
-      "loss": 0.3127,
+      "epoch": 0.733052221861823,
+      "grad_norm": 0.19731229543685913,
+      "learning_rate": 4.823982011855854e-05,
+      "loss": 0.4058,
       "step": 20340
     },
     {
-      "epoch": 0.72,
-      "learning_rate": 4.8340657271455694e-05,
-      "loss": 0.3214,
+      "epoch": 0.7332324215230476,
+      "grad_norm": 0.1781267374753952,
+      "learning_rate": 4.8238744370785545e-05,
+      "loss": 0.4357,
       "step": 20345
     },
     {
-      "epoch": 0.72,
-      "learning_rate": 4.833963656994313e-05,
-      "loss": 0.325,
+      "epoch": 0.7334126211842722,
+      "grad_norm": 0.11069954186677933,
+      "learning_rate": 4.823766830638916e-05,
+      "loss": 0.3899,
       "step": 20350
     },
     {
-      "epoch": 0.72,
-      "learning_rate": 4.833861556538076e-05,
-      "loss": 0.3356,
+      "epoch": 0.7335928208454968,
+      "grad_norm": 0.19684797525405884,
+      "learning_rate": 4.823659192538404e-05,
+      "loss": 0.4537,
       "step": 20355
     },
     {
-      "epoch": 0.72,
-      "learning_rate": 4.833759425778186e-05,
-      "loss": 0.3004,
+      "epoch": 0.7337730205067214,
+      "grad_norm": 0.13966168463230133,
+      "learning_rate": 4.8235515227784856e-05,
+      "loss": 0.407,
       "step": 20360
     },
     {
-      "epoch": 0.72,
-      "learning_rate": 4.833657264715967e-05,
-      "loss": 0.3253,
+      "epoch": 0.7339532201679461,
+      "grad_norm": 0.17704619467258453,
+      "learning_rate": 4.823443821360627e-05,
+      "loss": 0.4499,
       "step": 20365
     },
     {
-      "epoch": 0.72,
-      "learning_rate": 4.833555073352747e-05,
-      "loss": 0.3046,
+      "epoch": 0.7341334198291707,
+      "grad_norm": 0.16276319324970245,
+      "learning_rate": 4.8233360882862965e-05,
+      "loss": 0.4285,
       "step": 20370
     },
     {
-      "epoch": 0.72,
-      "learning_rate": 4.8334528516898515e-05,
-      "loss": 0.3052,
+      "epoch": 0.7343136194903953,
+      "grad_norm": 0.1888384073972702,
+      "learning_rate": 4.823228323556962e-05,
+      "loss": 0.4563,
       "step": 20375
     },
     {
-      "epoch": 0.72,
-      "learning_rate": 4.833350599728609e-05,
-      "loss": 0.3069,
+      "epoch": 0.73449381915162,
+      "grad_norm": 0.17120252549648285,
+      "learning_rate": 4.8231205271740916e-05,
+      "loss": 0.4775,
       "step": 20380
     },
     {
-      "epoch": 0.72,
-      "learning_rate": 4.8332483174703465e-05,
-      "loss": 0.2918,
+      "epoch": 0.7346740188128447,
+      "grad_norm": 0.1270987093448639,
+      "learning_rate": 4.8230126991391534e-05,
+      "loss": 0.4342,
       "step": 20385
     },
     {
-      "epoch": 0.72,
-      "learning_rate": 4.833146004916392e-05,
-      "loss": 0.2913,
+      "epoch": 0.7348542184740693,
+      "grad_norm": 0.16976775228977203,
+      "learning_rate": 4.822904839453617e-05,
+      "loss": 0.4159,
       "step": 20390
     },
     {
-      "epoch": 0.72,
-      "learning_rate": 4.8330436620680745e-05,
-      "loss": 0.2978,
+      "epoch": 0.7350344181352939,
+      "grad_norm": 0.1889120191335678,
+      "learning_rate": 4.822796948118952e-05,
+      "loss": 0.4339,
       "step": 20395
     },
     {
-      "epoch": 0.72,
-      "learning_rate": 4.832941288926723e-05,
-      "loss": 0.3143,
+      "epoch": 0.7352146177965185,
+      "grad_norm": 0.18324951827526093,
+      "learning_rate": 4.822689025136627e-05,
+      "loss": 0.4625,
       "step": 20400
     },
     {
-      "epoch": 0.72,
-      "learning_rate": 4.8328388854936656e-05,
-      "loss": 0.288,
+      "epoch": 0.7353948174577432,
+      "grad_norm": 0.19597896933555603,
+      "learning_rate": 4.822581070508115e-05,
+      "loss": 0.4828,
       "step": 20405
     },
     {
-      "epoch": 0.72,
-      "learning_rate": 4.832736451770233e-05,
-      "loss": 0.302,
+      "epoch": 0.7355750171189678,
+      "grad_norm": 0.20112694799900055,
+      "learning_rate": 4.8224730842348856e-05,
+      "loss": 0.4774,
       "step": 20410
     },
     {
-      "epoch": 0.72,
-      "learning_rate": 4.832633987757755e-05,
-      "loss": 0.3215,
+      "epoch": 0.7357552167801924,
+      "grad_norm": 0.16758687794208527,
+      "learning_rate": 4.8223650663184094e-05,
+      "loss": 0.4098,
       "step": 20415
     },
     {
-      "epoch": 0.72,
-      "learning_rate": 4.832531493457561e-05,
-      "loss": 0.3222,
+      "epoch": 0.7359354164414171,
+      "grad_norm": 0.16736017167568207,
+      "learning_rate": 4.82225701676016e-05,
+      "loss": 0.452,
       "step": 20420
     },
     {
-      "epoch": 0.72,
-      "learning_rate": 4.832428968870984e-05,
-      "loss": 0.3041,
+      "epoch": 0.7361156161026418,
+      "grad_norm": 0.20359498262405396,
+      "learning_rate": 4.822148935561607e-05,
+      "loss": 0.419,
       "step": 20425
     },
     {
-      "epoch": 0.72,
-      "learning_rate": 4.832326413999354e-05,
-      "loss": 0.3167,
+      "epoch": 0.7362958157638664,
+      "grad_norm": 0.17675939202308655,
+      "learning_rate": 4.8220408227242255e-05,
+      "loss": 0.4267,
       "step": 20430
     },
     {
-      "epoch": 0.72,
-      "learning_rate": 4.832223828844003e-05,
-      "loss": 0.288,
+      "epoch": 0.736476015425091,
+      "grad_norm": 0.18879984319210052,
+      "learning_rate": 4.821932678249487e-05,
+      "loss": 0.4493,
       "step": 20435
     },
     {
-      "epoch": 0.72,
-      "learning_rate": 4.832121213406261e-05,
-      "loss": 0.3283,
+      "epoch": 0.7366562150863156,
+      "grad_norm": 0.17970100045204163,
+      "learning_rate": 4.821824502138864e-05,
+      "loss": 0.4387,
       "step": 20440
     },
     {
-      "epoch": 0.72,
-      "learning_rate": 4.832018567687463e-05,
-      "loss": 0.3207,
+      "epoch": 0.7368364147475402,
+      "grad_norm": 0.16193163394927979,
+      "learning_rate": 4.8217162943938333e-05,
+      "loss": 0.418,
       "step": 20445
     },
     {
-      "epoch": 0.72,
-      "learning_rate": 4.83191589168894e-05,
-      "loss": 0.3025,
+      "epoch": 0.7370166144087649,
+      "grad_norm": 0.14100541174411774,
+      "learning_rate": 4.821608055015867e-05,
+      "loss": 0.4138,
       "step": 20450
     },
     {
-      "epoch": 0.72,
-      "learning_rate": 4.831813185412026e-05,
-      "loss": 0.2808,
+      "epoch": 0.7371968140699896,
+      "grad_norm": 0.16195495426654816,
+      "learning_rate": 4.8214997840064404e-05,
+      "loss": 0.4636,
       "step": 20455
     },
     {
-      "epoch": 0.72,
-      "learning_rate": 4.8317104488580553e-05,
-      "loss": 0.3145,
+      "epoch": 0.7373770137312142,
+      "grad_norm": 0.15169329941272736,
+      "learning_rate": 4.821391481367029e-05,
+      "loss": 0.4275,
       "step": 20460
     },
     {
-      "epoch": 0.72,
-      "learning_rate": 4.8316076820283604e-05,
-      "loss": 0.2893,
+      "epoch": 0.7375572133924388,
+      "grad_norm": 0.14933131635189056,
+      "learning_rate": 4.8212831470991076e-05,
+      "loss": 0.4403,
       "step": 20465
     },
     {
-      "epoch": 0.72,
-      "learning_rate": 4.8315048849242774e-05,
-      "loss": 0.3034,
+      "epoch": 0.7377374130536635,
+      "grad_norm": 0.19773240387439728,
+      "learning_rate": 4.821174781204153e-05,
+      "loss": 0.4501,
       "step": 20470
     },
     {
-      "epoch": 0.72,
-      "learning_rate": 4.8314020575471386e-05,
-      "loss": 0.2834,
+      "epoch": 0.7379176127148881,
+      "grad_norm": 0.18996679782867432,
+      "learning_rate": 4.821066383683641e-05,
+      "loss": 0.4403,
       "step": 20475
     },
     {
-      "epoch": 0.72,
-      "learning_rate": 4.8312991998982816e-05,
-      "loss": 0.3003,
+      "epoch": 0.7380978123761127,
+      "grad_norm": 0.17520780861377716,
+      "learning_rate": 4.82095795453905e-05,
+      "loss": 0.4348,
       "step": 20480
     },
     {
-      "epoch": 0.72,
-      "learning_rate": 4.83119631197904e-05,
-      "loss": 0.3166,
+      "epoch": 0.7382780120373373,
+      "grad_norm": 0.22540989518165588,
+      "learning_rate": 4.820849493771855e-05,
+      "loss": 0.4144,
       "step": 20485
     },
     {
-      "epoch": 0.72,
-      "learning_rate": 4.831093393790751e-05,
-      "loss": 0.2674,
+      "epoch": 0.738458211698562,
+      "grad_norm": 0.1550588756799698,
+      "learning_rate": 4.820741001383536e-05,
+      "loss": 0.3918,
       "step": 20490
     },
     {
-      "epoch": 0.72,
-      "learning_rate": 4.8309904453347504e-05,
-      "loss": 0.2891,
+      "epoch": 0.7386384113597867,
+      "grad_norm": 0.16206443309783936,
+      "learning_rate": 4.82063247737557e-05,
+      "loss": 0.4357,
       "step": 20495
     },
     {
-      "epoch": 0.72,
-      "learning_rate": 4.830887466612375e-05,
-      "loss": 0.2972,
+      "epoch": 0.7388186110210113,
+      "grad_norm": 0.20156899094581604,
+      "learning_rate": 4.820523921749435e-05,
+      "loss": 0.4141,
       "step": 20500
     },
     {
-      "epoch": 0.72,
-      "eval_loss": 0.3011039197444916,
-      "eval_runtime": 10.5518,
-      "eval_samples_per_second": 9.477,
-      "eval_steps_per_second": 9.477,
+      "epoch": 0.7388186110210113,
+      "eval_loss": 0.4630433917045593,
+      "eval_runtime": 3.5384,
+      "eval_samples_per_second": 28.261,
+      "eval_steps_per_second": 7.065,
       "step": 20500
     },
     {
-      "epoch": 0.72,
-      "learning_rate": 4.8307844576249625e-05,
-      "loss": 0.2883,
+      "epoch": 0.7389988106822359,
+      "grad_norm": 0.15963950753211975,
+      "learning_rate": 4.820415334506611e-05,
+      "loss": 0.433,
       "step": 20505
     },
     {
-      "epoch": 0.72,
-      "learning_rate": 4.830681418373849e-05,
-      "loss": 0.2932,
+      "epoch": 0.7391790103434606,
+      "grad_norm": 0.14632678031921387,
+      "learning_rate": 4.8203067156485775e-05,
+      "loss": 0.401,
       "step": 20510
     },
     {
-      "epoch": 0.72,
-      "learning_rate": 4.830578348860374e-05,
-      "loss": 0.303,
+      "epoch": 0.7393592100046852,
+      "grad_norm": 0.13777978718280792,
+      "learning_rate": 4.820198065176814e-05,
+      "loss": 0.4334,
       "step": 20515
     },
     {
-      "epoch": 0.72,
-      "learning_rate": 4.830475249085875e-05,
-      "loss": 0.3148,
+      "epoch": 0.7395394096659098,
+      "grad_norm": 0.1814904361963272,
+      "learning_rate": 4.820089383092802e-05,
+      "loss": 0.4499,
       "step": 20520
     },
     {
-      "epoch": 0.72,
-      "learning_rate": 4.83037211905169e-05,
-      "loss": 0.291,
+      "epoch": 0.7397196093271344,
+      "grad_norm": 0.17688485980033875,
+      "learning_rate": 4.819980669398021e-05,
+      "loss": 0.4625,
       "step": 20525
     },
     {
-      "epoch": 0.72,
-      "learning_rate": 4.83026895875916e-05,
-      "loss": 0.2962,
+      "epoch": 0.739899808988359,
+      "grad_norm": 0.17957401275634766,
+      "learning_rate": 4.819871924093951e-05,
+      "loss": 0.4324,
       "step": 20530
     },
     {
-      "epoch": 0.72,
-      "learning_rate": 4.8301657682096226e-05,
-      "loss": 0.3105,
+      "epoch": 0.7400800086495838,
+      "grad_norm": 0.17410768568515778,
+      "learning_rate": 4.819763147182077e-05,
+      "loss": 0.4289,
       "step": 20535
     },
     {
-      "epoch": 0.72,
-      "learning_rate": 4.830062547404418e-05,
-      "loss": 0.2883,
+      "epoch": 0.7402602083108084,
+      "grad_norm": 0.15775148570537567,
+      "learning_rate": 4.819654338663879e-05,
+      "loss": 0.423,
       "step": 20540
     },
     {
-      "epoch": 0.72,
-      "learning_rate": 4.829959296344887e-05,
-      "loss": 0.2962,
+      "epoch": 0.740440407972033,
+      "grad_norm": 0.2120039314031601,
+      "learning_rate": 4.8195454985408394e-05,
+      "loss": 0.4377,
       "step": 20545
     },
     {
-      "epoch": 0.72,
-      "learning_rate": 4.82985601503237e-05,
-      "loss": 0.3058,
+      "epoch": 0.7406206076332577,
+      "grad_norm": 0.15215463936328888,
+      "learning_rate": 4.8194366268144415e-05,
+      "loss": 0.4067,
       "step": 20550
     },
     {
-      "epoch": 0.72,
-      "learning_rate": 4.8297527034682086e-05,
-      "loss": 0.3153,
+      "epoch": 0.7408008072944823,
+      "grad_norm": 0.14415352046489716,
+      "learning_rate": 4.8193277234861686e-05,
+      "loss": 0.4525,
       "step": 20555
     },
     {
-      "epoch": 0.72,
-      "learning_rate": 4.8296493616537435e-05,
-      "loss": 0.2977,
+      "epoch": 0.7409810069557069,
+      "grad_norm": 0.18399272859096527,
+      "learning_rate": 4.819218788557505e-05,
+      "loss": 0.4216,
       "step": 20560
     },
     {
-      "epoch": 0.72,
-      "learning_rate": 4.8295459895903164e-05,
-      "loss": 0.3376,
+      "epoch": 0.7411612066169315,
+      "grad_norm": 0.20752954483032227,
+      "learning_rate": 4.819109822029933e-05,
+      "loss": 0.4236,
       "step": 20565
     },
     {
-      "epoch": 0.72,
-      "learning_rate": 4.82944258727927e-05,
-      "loss": 0.3026,
+      "epoch": 0.7413414062781561,
+      "grad_norm": 0.1981305330991745,
+      "learning_rate": 4.81900082390494e-05,
+      "loss": 0.4087,
       "step": 20570
     },
     {
-      "epoch": 0.72,
-      "learning_rate": 4.8293391547219465e-05,
-      "loss": 0.2868,
+      "epoch": 0.7415216059393809,
+      "grad_norm": 0.22662696242332458,
+      "learning_rate": 4.818891794184009e-05,
+      "loss": 0.4559,
       "step": 20575
     },
     {
-      "epoch": 0.72,
-      "learning_rate": 4.82923569191969e-05,
-      "loss": 0.3052,
+      "epoch": 0.7417018056006055,
+      "grad_norm": 0.16101478040218353,
+      "learning_rate": 4.818782732868627e-05,
+      "loss": 0.4777,
       "step": 20580
     },
     {
-      "epoch": 0.72,
-      "learning_rate": 4.829132198873842e-05,
-      "loss": 0.2877,
+      "epoch": 0.7418820052618301,
+      "grad_norm": 0.1838495433330536,
+      "learning_rate": 4.8186736399602784e-05,
+      "loss": 0.4384,
       "step": 20585
     },
     {
-      "epoch": 0.72,
-      "learning_rate": 4.829028675585748e-05,
-      "loss": 0.3021,
+      "epoch": 0.7420622049230547,
+      "grad_norm": 0.13098390400409698,
+      "learning_rate": 4.818564515460451e-05,
+      "loss": 0.4176,
       "step": 20590
     },
     {
-      "epoch": 0.72,
-      "learning_rate": 4.828925122056752e-05,
-      "loss": 0.2839,
+      "epoch": 0.7422424045842794,
+      "grad_norm": 0.18415838479995728,
+      "learning_rate": 4.818455359370631e-05,
+      "loss": 0.4396,
       "step": 20595
     },
     {
-      "epoch": 0.72,
-      "learning_rate": 4.8288215382881976e-05,
-      "loss": 0.3139,
+      "epoch": 0.742422604245504,
+      "grad_norm": 0.19082091748714447,
+      "learning_rate": 4.818346171692305e-05,
+      "loss": 0.4251,
       "step": 20600
     },
     {
-      "epoch": 0.72,
-      "learning_rate": 4.8287179242814303e-05,
-      "loss": 0.3174,
+      "epoch": 0.7426028039067286,
+      "grad_norm": 0.1550193876028061,
+      "learning_rate": 4.8182369524269616e-05,
+      "loss": 0.4532,
       "step": 20605
     },
     {
-      "epoch": 0.73,
-      "learning_rate": 4.828614280037795e-05,
-      "loss": 0.3124,
+      "epoch": 0.7427830035679532,
+      "grad_norm": 0.18546633422374725,
+      "learning_rate": 4.818127701576089e-05,
+      "loss": 0.4659,
       "step": 20610
     },
     {
-      "epoch": 0.73,
-      "learning_rate": 4.828510605558638e-05,
-      "loss": 0.3055,
+      "epoch": 0.742963203229178,
+      "grad_norm": 0.1380166858434677,
+      "learning_rate": 4.818018419141174e-05,
+      "loss": 0.4328,
       "step": 20615
     },
     {
-      "epoch": 0.73,
-      "learning_rate": 4.828406900845306e-05,
-      "loss": 0.3304,
+      "epoch": 0.7431434028904026,
+      "grad_norm": 0.15850889682769775,
+      "learning_rate": 4.817909105123708e-05,
+      "loss": 0.4383,
       "step": 20620
     },
     {
-      "epoch": 0.73,
-      "learning_rate": 4.8283031658991446e-05,
-      "loss": 0.2789,
+      "epoch": 0.7433236025516272,
+      "grad_norm": 0.14858169853687286,
+      "learning_rate": 4.817799759525179e-05,
+      "loss": 0.4636,
       "step": 20625
     },
     {
-      "epoch": 0.73,
-      "learning_rate": 4.8281994007215006e-05,
-      "loss": 0.3249,
+      "epoch": 0.7435038022128518,
+      "grad_norm": 0.15085402131080627,
+      "learning_rate": 4.8176903823470765e-05,
+      "loss": 0.4968,
       "step": 20630
     },
     {
-      "epoch": 0.73,
-      "learning_rate": 4.828095605313722e-05,
-      "loss": 0.2882,
+      "epoch": 0.7436840018740765,
+      "grad_norm": 0.16185277700424194,
+      "learning_rate": 4.817580973590892e-05,
+      "loss": 0.4329,
       "step": 20635
     },
     {
-      "epoch": 0.73,
-      "learning_rate": 4.827991779677157e-05,
-      "loss": 0.3259,
+      "epoch": 0.7438642015353011,
+      "grad_norm": 0.13464045524597168,
+      "learning_rate": 4.817471533258114e-05,
+      "loss": 0.4278,
       "step": 20640
     },
     {
-      "epoch": 0.73,
-      "learning_rate": 4.8278879238131514e-05,
-      "loss": 0.3139,
+      "epoch": 0.7440444011965257,
+      "grad_norm": 0.18652474880218506,
+      "learning_rate": 4.817362061350237e-05,
+      "loss": 0.4305,
       "step": 20645
     },
     {
-      "epoch": 0.73,
-      "learning_rate": 4.8277840377230565e-05,
-      "loss": 0.2858,
+      "epoch": 0.7442246008577504,
+      "grad_norm": 0.17949911952018738,
+      "learning_rate": 4.81725255786875e-05,
+      "loss": 0.4353,
       "step": 20650
     },
     {
-      "epoch": 0.73,
-      "learning_rate": 4.827680121408219e-05,
-      "loss": 0.3077,
+      "epoch": 0.7444048005189751,
+      "grad_norm": 0.1657877266407013,
+      "learning_rate": 4.817143022815145e-05,
+      "loss": 0.4464,
       "step": 20655
     },
     {
-      "epoch": 0.73,
-      "learning_rate": 4.82757617486999e-05,
-      "loss": 0.3209,
+      "epoch": 0.7445850001801997,
+      "grad_norm": 0.14313843846321106,
+      "learning_rate": 4.817033456190915e-05,
+      "loss": 0.4019,
       "step": 20660
     },
     {
-      "epoch": 0.73,
-      "learning_rate": 4.827472198109717e-05,
-      "loss": 0.295,
+      "epoch": 0.7447651998414243,
+      "grad_norm": 0.2136252373456955,
+      "learning_rate": 4.816923857997553e-05,
+      "loss": 0.4717,
       "step": 20665
     },
     {
-      "epoch": 0.73,
-      "learning_rate": 4.8273681911287525e-05,
-      "loss": 0.3056,
+      "epoch": 0.7449453995026489,
+      "grad_norm": 0.19905641674995422,
+      "learning_rate": 4.816814228236551e-05,
+      "loss": 0.4624,
       "step": 20670
     },
     {
-      "epoch": 0.73,
-      "learning_rate": 4.827264153928446e-05,
-      "loss": 0.319,
+      "epoch": 0.7451255991638736,
+      "grad_norm": 0.18663300573825836,
+      "learning_rate": 4.8167045669094044e-05,
+      "loss": 0.4294,
       "step": 20675
     },
     {
-      "epoch": 0.73,
-      "learning_rate": 4.8271600865101475e-05,
-      "loss": 0.317,
+      "epoch": 0.7453057988250982,
+      "grad_norm": 0.1515270322561264,
+      "learning_rate": 4.816594874017607e-05,
+      "loss": 0.4471,
       "step": 20680
     },
     {
-      "epoch": 0.73,
-      "learning_rate": 4.827055988875209e-05,
-      "loss": 0.3192,
+      "epoch": 0.7454859984863228,
+      "grad_norm": 0.15805648267269135,
+      "learning_rate": 4.8164851495626526e-05,
+      "loss": 0.4288,
       "step": 20685
     },
     {
-      "epoch": 0.73,
-      "learning_rate": 4.8269518610249825e-05,
-      "loss": 0.3178,
+      "epoch": 0.7456661981475475,
+      "grad_norm": 0.15768414735794067,
+      "learning_rate": 4.816375393546037e-05,
+      "loss": 0.4443,
       "step": 20690
     },
     {
-      "epoch": 0.73,
-      "learning_rate": 4.82684770296082e-05,
-      "loss": 0.3107,
+      "epoch": 0.7458463978087722,
+      "grad_norm": 0.15465152263641357,
+      "learning_rate": 4.8162656059692545e-05,
+      "loss": 0.4479,
       "step": 20695
     },
     {
-      "epoch": 0.73,
-      "learning_rate": 4.8267435146840724e-05,
-      "loss": 0.308,
+      "epoch": 0.7460265974699968,
+      "grad_norm": 0.17679835855960846,
+      "learning_rate": 4.816155786833802e-05,
+      "loss": 0.4345,
       "step": 20700
     },
     {
-      "epoch": 0.73,
-      "learning_rate": 4.826639296196094e-05,
-      "loss": 0.3015,
+      "epoch": 0.7462067971312214,
+      "grad_norm": 0.1588921993970871,
+      "learning_rate": 4.816045936141175e-05,
+      "loss": 0.4298,
       "step": 20705
     },
     {
-      "epoch": 0.73,
-      "learning_rate": 4.826535047498238e-05,
-      "loss": 0.3189,
+      "epoch": 0.746386996792446,
+      "grad_norm": 0.18818655610084534,
+      "learning_rate": 4.815936053892871e-05,
+      "loss": 0.4402,
       "step": 20710
     },
     {
-      "epoch": 0.73,
-      "learning_rate": 4.826430768591858e-05,
-      "loss": 0.3386,
+      "epoch": 0.7465671964536706,
+      "grad_norm": 0.11379419267177582,
+      "learning_rate": 4.815826140090386e-05,
+      "loss": 0.4041,
       "step": 20715
     },
     {
-      "epoch": 0.73,
-      "learning_rate": 4.8263264594783063e-05,
-      "loss": 0.3106,
+      "epoch": 0.7467473961148953,
+      "grad_norm": 0.22116567194461823,
+      "learning_rate": 4.815716194735218e-05,
+      "loss": 0.4516,
       "step": 20720
     },
     {
-      "epoch": 0.73,
-      "learning_rate": 4.8262221201589394e-05,
-      "loss": 0.2998,
+      "epoch": 0.7469275957761199,
+      "grad_norm": 0.20635400712490082,
+      "learning_rate": 4.8156062178288666e-05,
+      "loss": 0.4372,
       "step": 20725
     },
     {
-      "epoch": 0.73,
-      "learning_rate": 4.826117750635112e-05,
-      "loss": 0.3047,
+      "epoch": 0.7471077954373446,
+      "grad_norm": 0.19008758664131165,
+      "learning_rate": 4.815496209372827e-05,
+      "loss": 0.4649,
       "step": 20730
     },
     {
-      "epoch": 0.73,
-      "learning_rate": 4.8260133509081776e-05,
-      "loss": 0.3295,
+      "epoch": 0.7472879950985692,
+      "grad_norm": 0.1746254712343216,
+      "learning_rate": 4.815386169368601e-05,
+      "loss": 0.454,
       "step": 20735
     },
     {
-      "epoch": 0.73,
-      "learning_rate": 4.8259089209794926e-05,
-      "loss": 0.2965,
+      "epoch": 0.7474681947597939,
+      "grad_norm": 0.16542162001132965,
+      "learning_rate": 4.8152760978176864e-05,
+      "loss": 0.4185,
       "step": 20740
     },
     {
-      "epoch": 0.73,
-      "learning_rate": 4.825804460850414e-05,
-      "loss": 0.3049,
+      "epoch": 0.7476483944210185,
+      "grad_norm": 0.17639301717281342,
+      "learning_rate": 4.815165994721583e-05,
+      "loss": 0.4053,
       "step": 20745
     },
     {
-      "epoch": 0.73,
-      "learning_rate": 4.825699970522297e-05,
-      "loss": 0.3185,
+      "epoch": 0.7478285940822431,
+      "grad_norm": 0.13610634207725525,
+      "learning_rate": 4.8150558600817916e-05,
+      "loss": 0.4214,
       "step": 20750
     },
     {
-      "epoch": 0.73,
-      "learning_rate": 4.825595449996498e-05,
-      "loss": 0.3247,
+      "epoch": 0.7480087937434677,
+      "grad_norm": 0.17135217785835266,
+      "learning_rate": 4.814945693899812e-05,
+      "loss": 0.4546,
       "step": 20755
     },
     {
-      "epoch": 0.73,
-      "learning_rate": 4.8254908992743756e-05,
-      "loss": 0.2992,
+      "epoch": 0.7481889934046924,
+      "grad_norm": 0.18286027014255524,
+      "learning_rate": 4.8148354961771457e-05,
+      "loss": 0.4598,
       "step": 20760
     },
     {
-      "epoch": 0.73,
-      "learning_rate": 4.825386318357285e-05,
-      "loss": 0.2991,
+      "epoch": 0.748369193065917,
+      "grad_norm": 0.17390398681163788,
+      "learning_rate": 4.814725266915294e-05,
+      "loss": 0.4174,
       "step": 20765
     },
     {
-      "epoch": 0.73,
-      "learning_rate": 4.8252817072465864e-05,
-      "loss": 0.3054,
+      "epoch": 0.7485493927271417,
+      "grad_norm": 0.18580150604248047,
+      "learning_rate": 4.814615006115759e-05,
+      "loss": 0.4566,
       "step": 20770
     },
     {
-      "epoch": 0.73,
-      "learning_rate": 4.8251770659436375e-05,
-      "loss": 0.3022,
+      "epoch": 0.7487295923883663,
+      "grad_norm": 0.1448039710521698,
+      "learning_rate": 4.814504713780041e-05,
+      "loss": 0.4102,
       "step": 20775
     },
     {
-      "epoch": 0.73,
-      "learning_rate": 4.8250723944497964e-05,
-      "loss": 0.3082,
+      "epoch": 0.748909792049591,
+      "grad_norm": 0.14498092234134674,
+      "learning_rate": 4.814394389909647e-05,
+      "loss": 0.4198,
       "step": 20780
     },
     {
-      "epoch": 0.73,
-      "learning_rate": 4.824988635518195e-05,
-      "loss": 0.2986,
+      "epoch": 0.7490899917108156,
+      "grad_norm": 0.14252233505249023,
+      "learning_rate": 4.814284034506076e-05,
+      "loss": 0.4443,
       "step": 20785
     },
     {
-      "epoch": 0.73,
-      "learning_rate": 4.8248839096841736e-05,
-      "loss": 0.2889,
+      "epoch": 0.7492701913720402,
+      "grad_norm": 0.15442687273025513,
+      "learning_rate": 4.8141736475708325e-05,
+      "loss": 0.4487,
       "step": 20790
     },
     {
-      "epoch": 0.73,
-      "learning_rate": 4.8247791536630674e-05,
-      "loss": 0.3032,
+      "epoch": 0.7494503910332648,
+      "grad_norm": 0.15849895775318146,
+      "learning_rate": 4.814063229105422e-05,
+      "loss": 0.418,
       "step": 20795
     },
     {
-      "epoch": 0.73,
-      "learning_rate": 4.824674367456235e-05,
-      "loss": 0.3027,
+      "epoch": 0.7496305906944895,
+      "grad_norm": 0.16628096997737885,
+      "learning_rate": 4.813952779111348e-05,
+      "loss": 0.4183,
       "step": 20800
     },
     {
-      "epoch": 0.73,
-      "learning_rate": 4.824569551065039e-05,
-      "loss": 0.3137,
+      "epoch": 0.7498107903557142,
+      "grad_norm": 0.16264590620994568,
+      "learning_rate": 4.813842297590115e-05,
+      "loss": 0.4466,
       "step": 20805
     },
     {
-      "epoch": 0.73,
-      "learning_rate": 4.824464704490838e-05,
-      "loss": 0.3047,
+      "epoch": 0.7499909900169388,
+      "grad_norm": 0.16959542036056519,
+      "learning_rate": 4.813731784543229e-05,
+      "loss": 0.4476,
       "step": 20810
     },
     {
-      "epoch": 0.73,
-      "learning_rate": 4.824359827734997e-05,
-      "loss": 0.3319,
+      "epoch": 0.7501711896781634,
+      "grad_norm": 0.14073093235492706,
+      "learning_rate": 4.813621239972195e-05,
+      "loss": 0.4246,
       "step": 20815
     },
     {
-      "epoch": 0.73,
-      "learning_rate": 4.8242549207988735e-05,
-      "loss": 0.2975,
+      "epoch": 0.750351389339388,
+      "grad_norm": 0.21107293665409088,
+      "learning_rate": 4.81351066387852e-05,
+      "loss": 0.4279,
       "step": 20820
     },
     {
-      "epoch": 0.73,
-      "learning_rate": 4.824149983683832e-05,
-      "loss": 0.3049,
+      "epoch": 0.7505315890006127,
+      "grad_norm": 0.1864260733127594,
+      "learning_rate": 4.813400056263709e-05,
+      "loss": 0.4295,
       "step": 20825
     },
     {
-      "epoch": 0.73,
-      "learning_rate": 4.824045016391235e-05,
-      "loss": 0.2976,
+      "epoch": 0.7507117886618373,
+      "grad_norm": 0.13779747486114502,
+      "learning_rate": 4.813289417129272e-05,
+      "loss": 0.4166,
       "step": 20830
     },
     {
-      "epoch": 0.73,
-      "learning_rate": 4.823940018922445e-05,
-      "loss": 0.3072,
+      "epoch": 0.7508919883230619,
+      "grad_norm": 0.17728812992572784,
+      "learning_rate": 4.813178746476713e-05,
+      "loss": 0.4764,
       "step": 20835
     },
     {
-      "epoch": 0.73,
-      "learning_rate": 4.823834991278825e-05,
-      "loss": 0.3261,
+      "epoch": 0.7510721879842865,
+      "grad_norm": 0.1685779094696045,
+      "learning_rate": 4.813068044307543e-05,
+      "loss": 0.4457,
       "step": 20840
     },
     {
-      "epoch": 0.73,
-      "learning_rate": 4.8237299334617395e-05,
-      "loss": 0.3124,
+      "epoch": 0.7512523876455113,
+      "grad_norm": 0.17559683322906494,
+      "learning_rate": 4.812957310623267e-05,
+      "loss": 0.4377,
       "step": 20845
     },
     {
-      "epoch": 0.73,
-      "learning_rate": 4.823624845472552e-05,
-      "loss": 0.3198,
+      "epoch": 0.7514325873067359,
+      "grad_norm": 0.18856866657733917,
+      "learning_rate": 4.812846545425396e-05,
+      "loss": 0.4306,
       "step": 20850
     },
     {
-      "epoch": 0.73,
-      "learning_rate": 4.823519727312627e-05,
-      "loss": 0.3064,
+      "epoch": 0.7516127869679605,
+      "grad_norm": 0.1679704487323761,
+      "learning_rate": 4.812735748715439e-05,
+      "loss": 0.4218,
       "step": 20855
     },
     {
-      "epoch": 0.73,
-      "learning_rate": 4.8234145789833296e-05,
-      "loss": 0.2938,
+      "epoch": 0.7517929866291851,
+      "grad_norm": 0.20436346530914307,
+      "learning_rate": 4.812624920494905e-05,
+      "loss": 0.4475,
       "step": 20860
     },
     {
-      "epoch": 0.73,
-      "learning_rate": 4.8233094004860256e-05,
-      "loss": 0.3166,
+      "epoch": 0.7519731862904098,
+      "grad_norm": 0.16926033794879913,
+      "learning_rate": 4.812514060765304e-05,
+      "loss": 0.4159,
       "step": 20865
     },
     {
-      "epoch": 0.73,
-      "learning_rate": 4.823225235968135e-05,
-      "loss": 0.3031,
+      "epoch": 0.7521533859516344,
+      "grad_norm": 0.18585766851902008,
+      "learning_rate": 4.8124031695281465e-05,
+      "loss": 0.4935,
       "step": 20870
     },
     {
-      "epoch": 0.73,
-      "learning_rate": 4.823120003171859e-05,
-      "loss": 0.2976,
+      "epoch": 0.752333585612859,
+      "grad_norm": 0.15015818178653717,
+      "learning_rate": 4.812292246784944e-05,
+      "loss": 0.4683,
       "step": 20875
     },
     {
-      "epoch": 0.73,
-      "learning_rate": 4.8230147402114024e-05,
-      "loss": 0.3039,
+      "epoch": 0.7525137852740836,
+      "grad_norm": 0.18261437118053436,
+      "learning_rate": 4.8121812925372074e-05,
+      "loss": 0.4706,
       "step": 20880
     },
     {
-      "epoch": 0.73,
-      "learning_rate": 4.822909447088129e-05,
-      "loss": 0.2846,
+      "epoch": 0.7526939849353084,
+      "grad_norm": 0.15235644578933716,
+      "learning_rate": 4.812070306786448e-05,
+      "loss": 0.493,
       "step": 20885
     },
     {
-      "epoch": 0.73,
-      "learning_rate": 4.822804123803407e-05,
-      "loss": 0.3226,
+      "epoch": 0.752874184596533,
+      "grad_norm": 0.1405205875635147,
+      "learning_rate": 4.811959289534178e-05,
+      "loss": 0.4798,
       "step": 20890
     },
     {
-      "epoch": 0.74,
-      "learning_rate": 4.8226987703586056e-05,
-      "loss": 0.2975,
+      "epoch": 0.7530543842577576,
+      "grad_norm": 0.16182038187980652,
+      "learning_rate": 4.811848240781911e-05,
+      "loss": 0.4287,
       "step": 20895
     },
     {
-      "epoch": 0.74,
-      "learning_rate": 4.82259338675509e-05,
-      "loss": 0.3286,
+      "epoch": 0.7532345839189822,
+      "grad_norm": 0.17305776476860046,
+      "learning_rate": 4.811737160531159e-05,
+      "loss": 0.4401,
       "step": 20900
     },
     {
-      "epoch": 0.74,
-      "learning_rate": 4.8224879729942307e-05,
-      "loss": 0.2886,
+      "epoch": 0.7534147835802069,
+      "grad_norm": 0.1699703484773636,
+      "learning_rate": 4.811626048783435e-05,
+      "loss": 0.4514,
       "step": 20905
     },
     {
-      "epoch": 0.74,
-      "learning_rate": 4.822382529077396e-05,
-      "loss": 0.2958,
+      "epoch": 0.7535949832414315,
+      "grad_norm": 0.1501130610704422,
+      "learning_rate": 4.8115149055402545e-05,
+      "loss": 0.4463,
       "step": 20910
     },
     {
-      "epoch": 0.74,
-      "learning_rate": 4.8222770550059537e-05,
-      "loss": 0.3194,
+      "epoch": 0.7537751829026561,
+      "grad_norm": 0.18883809447288513,
+      "learning_rate": 4.811403730803131e-05,
+      "loss": 0.4432,
       "step": 20915
     },
     {
-      "epoch": 0.74,
-      "learning_rate": 4.8221715507812746e-05,
-      "loss": 0.3349,
+      "epoch": 0.7539553825638807,
+      "grad_norm": 0.1885019838809967,
+      "learning_rate": 4.811292524573579e-05,
+      "loss": 0.4145,
       "step": 20920
     },
     {
-      "epoch": 0.74,
-      "learning_rate": 4.822066016404729e-05,
-      "loss": 0.2894,
+      "epoch": 0.7541355822251055,
+      "grad_norm": 0.2025119662284851,
+      "learning_rate": 4.811181286853113e-05,
+      "loss": 0.4805,
       "step": 20925
     },
     {
-      "epoch": 0.74,
-      "learning_rate": 4.821960451877686e-05,
-      "loss": 0.2959,
+      "epoch": 0.7543157818863301,
+      "grad_norm": 0.22509577870368958,
+      "learning_rate": 4.811070017643251e-05,
+      "loss": 0.4674,
       "step": 20930
     },
     {
-      "epoch": 0.74,
-      "learning_rate": 4.821854857201516e-05,
-      "loss": 0.2829,
+      "epoch": 0.7544959815475547,
+      "grad_norm": 0.18206484615802765,
+      "learning_rate": 4.810958716945507e-05,
+      "loss": 0.4384,
       "step": 20935
     },
     {
-      "epoch": 0.74,
-      "learning_rate": 4.821749232377592e-05,
-      "loss": 0.3081,
+      "epoch": 0.7546761812087793,
+      "grad_norm": 0.1454673856496811,
+      "learning_rate": 4.810847384761397e-05,
+      "loss": 0.4224,
       "step": 20940
     },
     {
-      "epoch": 0.74,
-      "learning_rate": 4.821643577407284e-05,
-      "loss": 0.3231,
+      "epoch": 0.754856380870004,
+      "grad_norm": 0.15570025146007538,
+      "learning_rate": 4.810736021092439e-05,
+      "loss": 0.471,
       "step": 20945
     },
     {
-      "epoch": 0.74,
-      "learning_rate": 4.821537892291964e-05,
-      "loss": 0.3026,
+      "epoch": 0.7550365805312286,
+      "grad_norm": 0.18441855907440186,
+      "learning_rate": 4.810624625940151e-05,
+      "loss": 0.4365,
       "step": 20950
     },
     {
-      "epoch": 0.74,
-      "learning_rate": 4.821432177033004e-05,
-      "loss": 0.3115,
+      "epoch": 0.7552167801924532,
+      "grad_norm": 0.20488105714321136,
+      "learning_rate": 4.810513199306049e-05,
+      "loss": 0.458,
       "step": 20955
     },
     {
-      "epoch": 0.74,
-      "learning_rate": 4.8213264316317776e-05,
-      "loss": 0.3078,
+      "epoch": 0.7553969798536779,
+      "grad_norm": 0.16301538050174713,
+      "learning_rate": 4.8104017411916526e-05,
+      "loss": 0.4031,
       "step": 20960
     },
     {
-      "epoch": 0.74,
-      "learning_rate": 4.821220656089657e-05,
-      "loss": 0.3144,
+      "epoch": 0.7555771795149026,
+      "grad_norm": 0.1542273312807083,
+      "learning_rate": 4.810290251598479e-05,
+      "loss": 0.4097,
       "step": 20965
     },
     {
-      "epoch": 0.74,
-      "learning_rate": 4.8211148504080164e-05,
-      "loss": 0.3014,
+      "epoch": 0.7557573791761272,
+      "grad_norm": 0.1671588271856308,
+      "learning_rate": 4.8101787305280485e-05,
+      "loss": 0.3957,
       "step": 20970
     },
     {
-      "epoch": 0.74,
-      "learning_rate": 4.821009014588229e-05,
-      "loss": 0.2936,
+      "epoch": 0.7559375788373518,
+      "grad_norm": 0.1824352741241455,
+      "learning_rate": 4.8100671779818795e-05,
+      "loss": 0.4619,
       "step": 20975
     },
     {
-      "epoch": 0.74,
-      "learning_rate": 4.820903148631669e-05,
-      "loss": 0.2923,
+      "epoch": 0.7561177784985764,
+      "grad_norm": 0.18826597929000854,
+      "learning_rate": 4.809955593961493e-05,
+      "loss": 0.4472,
       "step": 20980
     },
     {
-      "epoch": 0.74,
-      "learning_rate": 4.820797252539711e-05,
-      "loss": 0.3053,
+      "epoch": 0.756297978159801,
+      "grad_norm": 0.14293283224105835,
+      "learning_rate": 4.809843978468409e-05,
+      "loss": 0.4693,
       "step": 20985
     },
     {
-      "epoch": 0.74,
-      "learning_rate": 4.82069132631373e-05,
-      "loss": 0.3627,
+      "epoch": 0.7564781778210257,
+      "grad_norm": 0.13552211225032806,
+      "learning_rate": 4.809732331504148e-05,
+      "loss": 0.39,
       "step": 20990
     },
     {
-      "epoch": 0.74,
-      "learning_rate": 4.8205853699551016e-05,
-      "loss": 0.291,
+      "epoch": 0.7566583774822503,
+      "grad_norm": 0.1683761477470398,
+      "learning_rate": 4.8096206530702305e-05,
+      "loss": 0.4059,
       "step": 20995
     },
     {
-      "epoch": 0.74,
-      "learning_rate": 4.820479383465202e-05,
-      "loss": 0.3204,
+      "epoch": 0.756838577143475,
+      "grad_norm": 0.17060311138629913,
+      "learning_rate": 4.809508943168179e-05,
+      "loss": 0.4352,
       "step": 21000
     },
     {
-      "epoch": 0.74,
-      "eval_loss": 0.29989683628082275,
-      "eval_runtime": 10.5493,
-      "eval_samples_per_second": 9.479,
-      "eval_steps_per_second": 9.479,
+      "epoch": 0.756838577143475,
+      "eval_loss": 0.462345689535141,
+      "eval_runtime": 3.5382,
+      "eval_samples_per_second": 28.263,
+      "eval_steps_per_second": 7.066,
       "step": 21000
     },
     {
-      "epoch": 0.74,
-      "learning_rate": 4.8203733668454066e-05,
-      "loss": 0.298,
+      "epoch": 0.7570187768046996,
+      "grad_norm": 0.179361492395401,
+      "learning_rate": 4.8093972017995155e-05,
+      "loss": 0.4466,
       "step": 21005
     },
     {
-      "epoch": 0.74,
-      "learning_rate": 4.8202673200970925e-05,
-      "loss": 0.2888,
+      "epoch": 0.7571989764659243,
+      "grad_norm": 0.19304408133029938,
+      "learning_rate": 4.8092854289657617e-05,
+      "loss": 0.4786,
       "step": 21010
     },
     {
-      "epoch": 0.74,
-      "learning_rate": 4.820161243221636e-05,
-      "loss": 0.2988,
+      "epoch": 0.7573791761271489,
+      "grad_norm": 0.16757160425186157,
+      "learning_rate": 4.809173624668442e-05,
+      "loss": 0.4614,
       "step": 21015
     },
     {
-      "epoch": 0.74,
-      "learning_rate": 4.820055136220415e-05,
-      "loss": 0.3334,
+      "epoch": 0.7575593757883735,
+      "grad_norm": 0.16519469022750854,
+      "learning_rate": 4.8090617889090786e-05,
+      "loss": 0.4297,
       "step": 21020
     },
     {
-      "epoch": 0.74,
-      "learning_rate": 4.819948999094807e-05,
-      "loss": 0.3501,
+      "epoch": 0.7577395754495981,
+      "grad_norm": 0.1686154156923294,
+      "learning_rate": 4.808949921689194e-05,
+      "loss": 0.387,
       "step": 21025
     },
     {
-      "epoch": 0.74,
-      "learning_rate": 4.81984283184619e-05,
-      "loss": 0.3129,
+      "epoch": 0.7579197751108228,
+      "grad_norm": 0.1524842232465744,
+      "learning_rate": 4.8088380230103145e-05,
+      "loss": 0.4291,
       "step": 21030
     },
     {
-      "epoch": 0.74,
-      "learning_rate": 4.819736634475942e-05,
-      "loss": 0.2927,
+      "epoch": 0.7580999747720474,
+      "grad_norm": 0.18697072565555573,
+      "learning_rate": 4.808726092873964e-05,
+      "loss": 0.4262,
       "step": 21035
     },
     {
-      "epoch": 0.74,
-      "learning_rate": 4.819630406985444e-05,
-      "loss": 0.2925,
+      "epoch": 0.7582801744332721,
+      "grad_norm": 0.21223722398281097,
+      "learning_rate": 4.808614131281668e-05,
+      "loss": 0.4522,
       "step": 21040
     },
     {
-      "epoch": 0.74,
-      "learning_rate": 4.819524149376072e-05,
-      "loss": 0.3117,
+      "epoch": 0.7584603740944967,
+      "grad_norm": 0.16958850622177124,
+      "learning_rate": 4.808502138234951e-05,
+      "loss": 0.4667,
       "step": 21045
     },
     {
-      "epoch": 0.74,
-      "learning_rate": 4.8194178616492094e-05,
-      "loss": 0.2986,
+      "epoch": 0.7586405737557214,
+      "grad_norm": 0.1471187025308609,
+      "learning_rate": 4.808390113735339e-05,
+      "loss": 0.4847,
       "step": 21050
     },
     {
-      "epoch": 0.74,
-      "learning_rate": 4.819311543806233e-05,
-      "loss": 0.3166,
+      "epoch": 0.758820773416946,
+      "grad_norm": 0.156825989484787,
+      "learning_rate": 4.808278057784359e-05,
+      "loss": 0.4648,
       "step": 21055
     },
     {
-      "epoch": 0.74,
-      "learning_rate": 4.819205195848526e-05,
-      "loss": 0.3084,
+      "epoch": 0.7590009730781706,
+      "grad_norm": 0.15920226275920868,
+      "learning_rate": 4.808165970383538e-05,
+      "loss": 0.4215,
       "step": 21060
     },
     {
-      "epoch": 0.74,
-      "learning_rate": 4.8190988177774666e-05,
-      "loss": 0.3021,
+      "epoch": 0.7591811727393952,
+      "grad_norm": 0.1420614868402481,
+      "learning_rate": 4.808053851534401e-05,
+      "loss": 0.411,
       "step": 21065
     },
     {
-      "epoch": 0.74,
-      "learning_rate": 4.818992409594438e-05,
-      "loss": 0.289,
+      "epoch": 0.7593613724006198,
+      "grad_norm": 0.21698161959648132,
+      "learning_rate": 4.8079417012384786e-05,
+      "loss": 0.4415,
       "step": 21070
     },
     {
-      "epoch": 0.74,
-      "learning_rate": 4.8188859713008215e-05,
-      "loss": 0.309,
+      "epoch": 0.7595415720618445,
+      "grad_norm": 0.18251734972000122,
+      "learning_rate": 4.807829519497297e-05,
+      "loss": 0.4644,
       "step": 21075
     },
     {
-      "epoch": 0.74,
-      "learning_rate": 4.8187795028979985e-05,
-      "loss": 0.3262,
+      "epoch": 0.7597217717230692,
+      "grad_norm": 0.18656039237976074,
+      "learning_rate": 4.8077173063123843e-05,
+      "loss": 0.456,
       "step": 21080
     },
     {
-      "epoch": 0.74,
-      "learning_rate": 4.818673004387352e-05,
-      "loss": 0.3161,
+      "epoch": 0.7599019713842938,
+      "grad_norm": 0.17298544943332672,
+      "learning_rate": 4.807605061685271e-05,
+      "loss": 0.4445,
       "step": 21085
     },
     {
-      "epoch": 0.74,
-      "learning_rate": 4.8185664757702654e-05,
-      "loss": 0.2769,
+      "epoch": 0.7600821710455185,
+      "grad_norm": 0.1439000368118286,
+      "learning_rate": 4.807492785617484e-05,
+      "loss": 0.4311,
       "step": 21090
     },
     {
-      "epoch": 0.74,
-      "learning_rate": 4.81845991704812e-05,
-      "loss": 0.3011,
+      "epoch": 0.7602623707067431,
+      "grad_norm": 0.14586910605430603,
+      "learning_rate": 4.807380478110556e-05,
+      "loss": 0.4383,
       "step": 21095
     },
     {
-      "epoch": 0.74,
-      "learning_rate": 4.8183533282223e-05,
-      "loss": 0.3161,
+      "epoch": 0.7604425703679677,
+      "grad_norm": 0.18215352296829224,
+      "learning_rate": 4.8072681391660153e-05,
+      "loss": 0.4222,
       "step": 21100
     },
     {
-      "epoch": 0.74,
-      "learning_rate": 4.8182467092941915e-05,
-      "loss": 0.298,
+      "epoch": 0.7606227700291923,
+      "grad_norm": 0.1855933964252472,
+      "learning_rate": 4.807155768785393e-05,
+      "loss": 0.4237,
       "step": 21105
     },
     {
-      "epoch": 0.74,
-      "learning_rate": 4.818140060265176e-05,
-      "loss": 0.3351,
+      "epoch": 0.7608029696904169,
+      "grad_norm": 0.22114041447639465,
+      "learning_rate": 4.80704336697022e-05,
+      "loss": 0.4643,
       "step": 21110
     },
     {
-      "epoch": 0.74,
-      "learning_rate": 4.81803338113664e-05,
-      "loss": 0.3028,
+      "epoch": 0.7609831693516416,
+      "grad_norm": 0.1789657175540924,
+      "learning_rate": 4.806930933722027e-05,
+      "loss": 0.4599,
       "step": 21115
     },
     {
-      "epoch": 0.74,
-      "learning_rate": 4.817926671909968e-05,
-      "loss": 0.3052,
+      "epoch": 0.7611633690128663,
+      "grad_norm": 0.17131780087947845,
+      "learning_rate": 4.806818469042348e-05,
+      "loss": 0.4171,
       "step": 21120
     },
     {
-      "epoch": 0.74,
-      "learning_rate": 4.8178199325865457e-05,
-      "loss": 0.3117,
+      "epoch": 0.7613435686740909,
+      "grad_norm": 0.20241133868694305,
+      "learning_rate": 4.806705972932713e-05,
+      "loss": 0.4611,
       "step": 21125
     },
     {
-      "epoch": 0.74,
-      "learning_rate": 4.817713163167759e-05,
-      "loss": 0.2919,
+      "epoch": 0.7615237683353155,
+      "grad_norm": 0.19503600895404816,
+      "learning_rate": 4.806593445394656e-05,
+      "loss": 0.4813,
       "step": 21130
     },
     {
-      "epoch": 0.74,
-      "learning_rate": 4.817606363654995e-05,
-      "loss": 0.3248,
+      "epoch": 0.7617039679965402,
+      "grad_norm": 0.1657884567975998,
+      "learning_rate": 4.8064808864297094e-05,
+      "loss": 0.4784,
       "step": 21135
     },
     {
-      "epoch": 0.74,
-      "learning_rate": 4.81749953404964e-05,
-      "loss": 0.2991,
+      "epoch": 0.7618841676577648,
+      "grad_norm": 0.18670715391635895,
+      "learning_rate": 4.806368296039409e-05,
+      "loss": 0.42,
       "step": 21140
     },
     {
-      "epoch": 0.74,
-      "learning_rate": 4.8173926743530794e-05,
-      "loss": 0.3388,
+      "epoch": 0.7620643673189894,
+      "grad_norm": 0.16101418435573578,
+      "learning_rate": 4.806255674225285e-05,
+      "loss": 0.4258,
       "step": 21145
     },
     {
-      "epoch": 0.74,
-      "learning_rate": 4.817285784566703e-05,
-      "loss": 0.3201,
+      "epoch": 0.762244566980214,
+      "grad_norm": 0.16799142956733704,
+      "learning_rate": 4.806143020988875e-05,
+      "loss": 0.4105,
       "step": 21150
     },
     {
-      "epoch": 0.74,
-      "learning_rate": 4.817178864691898e-05,
-      "loss": 0.3244,
+      "epoch": 0.7624247666414388,
+      "grad_norm": 0.15857410430908203,
+      "learning_rate": 4.806030336331713e-05,
+      "loss": 0.4251,
       "step": 21155
     },
     {
-      "epoch": 0.74,
-      "learning_rate": 4.817071914730052e-05,
-      "loss": 0.3065,
+      "epoch": 0.7626049663026634,
+      "grad_norm": 0.16540922224521637,
+      "learning_rate": 4.8059176202553337e-05,
+      "loss": 0.4593,
       "step": 21160
     },
     {
-      "epoch": 0.74,
-      "learning_rate": 4.816964934682555e-05,
-      "loss": 0.2745,
+      "epoch": 0.762785165963888,
+      "grad_norm": 0.1826573759317398,
+      "learning_rate": 4.8058048727612724e-05,
+      "loss": 0.4513,
       "step": 21165
     },
     {
-      "epoch": 0.74,
-      "learning_rate": 4.816857924550794e-05,
-      "loss": 0.3042,
+      "epoch": 0.7629653656251126,
+      "grad_norm": 0.1935107260942459,
+      "learning_rate": 4.8056920938510675e-05,
+      "loss": 0.4293,
       "step": 21170
     },
     {
-      "epoch": 0.74,
-      "learning_rate": 4.8167508843361606e-05,
-      "loss": 0.3095,
+      "epoch": 0.7631455652863373,
+      "grad_norm": 0.18235011398792267,
+      "learning_rate": 4.8055792835262536e-05,
+      "loss": 0.4956,
       "step": 21175
     },
     {
-      "epoch": 0.75,
-      "learning_rate": 4.816643814040044e-05,
-      "loss": 0.289,
+      "epoch": 0.7633257649475619,
+      "grad_norm": 0.168792262673378,
+      "learning_rate": 4.8054664417883685e-05,
+      "loss": 0.4623,
       "step": 21180
     },
     {
-      "epoch": 0.75,
-      "learning_rate": 4.8165367136638335e-05,
-      "loss": 0.3259,
+      "epoch": 0.7635059646087865,
+      "grad_norm": 0.17810115218162537,
+      "learning_rate": 4.8053535686389495e-05,
+      "loss": 0.4265,
       "step": 21185
     },
     {
-      "epoch": 0.75,
-      "learning_rate": 4.81642958320892e-05,
-      "loss": 0.313,
+      "epoch": 0.7636861642700111,
+      "grad_norm": 0.20032259821891785,
+      "learning_rate": 4.805240664079534e-05,
+      "loss": 0.4372,
       "step": 21190
     },
     {
-      "epoch": 0.75,
-      "learning_rate": 4.8163224226766955e-05,
-      "loss": 0.2789,
+      "epoch": 0.7638663639312359,
+      "grad_norm": 0.147112175822258,
+      "learning_rate": 4.805127728111662e-05,
+      "loss": 0.4288,
       "step": 21195
     },
     {
-      "epoch": 0.75,
-      "learning_rate": 4.8162152320685515e-05,
-      "loss": 0.3211,
+      "epoch": 0.7640465635924605,
+      "grad_norm": 0.1480092704296112,
+      "learning_rate": 4.80501476073687e-05,
+      "loss": 0.4137,
       "step": 21200
     },
     {
-      "epoch": 0.75,
-      "learning_rate": 4.8161080113858786e-05,
-      "loss": 0.3257,
+      "epoch": 0.7642267632536851,
+      "grad_norm": 0.17761540412902832,
+      "learning_rate": 4.8049017619566986e-05,
+      "loss": 0.4727,
       "step": 21205
     },
     {
-      "epoch": 0.75,
-      "learning_rate": 4.8160007606300694e-05,
-      "loss": 0.3075,
+      "epoch": 0.7644069629149097,
+      "grad_norm": 0.14854562282562256,
+      "learning_rate": 4.8047887317726865e-05,
+      "loss": 0.4549,
       "step": 21210
     },
     {
-      "epoch": 0.75,
-      "learning_rate": 4.815893479802517e-05,
-      "loss": 0.3072,
+      "epoch": 0.7645871625761343,
+      "grad_norm": 0.1848391890525818,
+      "learning_rate": 4.804675670186374e-05,
+      "loss": 0.4468,
       "step": 21215
     },
     {
-      "epoch": 0.75,
-      "learning_rate": 4.815786168904613e-05,
-      "loss": 0.3119,
+      "epoch": 0.764767362237359,
+      "grad_norm": 0.18001249432563782,
+      "learning_rate": 4.804562577199302e-05,
+      "loss": 0.4149,
       "step": 21220
     },
     {
-      "epoch": 0.75,
-      "learning_rate": 4.8156788279377534e-05,
-      "loss": 0.3135,
+      "epoch": 0.7649475618985836,
+      "grad_norm": 0.16468872129917145,
+      "learning_rate": 4.804449452813011e-05,
+      "loss": 0.4224,
       "step": 21225
     },
     {
-      "epoch": 0.75,
-      "learning_rate": 4.8155714569033294e-05,
-      "loss": 0.2867,
+      "epoch": 0.7651277615598082,
+      "grad_norm": 0.17478923499584198,
+      "learning_rate": 4.804336297029043e-05,
+      "loss": 0.4528,
       "step": 21230
     },
     {
-      "epoch": 0.75,
-      "learning_rate": 4.815464055802736e-05,
-      "loss": 0.2794,
+      "epoch": 0.765307961221033,
+      "grad_norm": 0.15894725918769836,
+      "learning_rate": 4.804223109848939e-05,
+      "loss": 0.4523,
       "step": 21235
     },
     {
-      "epoch": 0.75,
-      "learning_rate": 4.815356624637368e-05,
-      "loss": 0.3002,
+      "epoch": 0.7654881608822576,
+      "grad_norm": 0.14751605689525604,
+      "learning_rate": 4.80410989127424e-05,
+      "loss": 0.4273,
       "step": 21240
     },
     {
-      "epoch": 0.75,
-      "learning_rate": 4.815249163408621e-05,
-      "loss": 0.3174,
+      "epoch": 0.7656683605434822,
+      "grad_norm": 0.2025453895330429,
+      "learning_rate": 4.803996641306491e-05,
+      "loss": 0.44,
       "step": 21245
     },
     {
-      "epoch": 0.75,
-      "learning_rate": 4.815141672117888e-05,
-      "loss": 0.2666,
+      "epoch": 0.7658485602047068,
+      "grad_norm": 0.14769452810287476,
+      "learning_rate": 4.803883359947233e-05,
+      "loss": 0.4416,
       "step": 21250
     },
     {
-      "epoch": 0.75,
-      "learning_rate": 4.8150341507665664e-05,
-      "loss": 0.3275,
+      "epoch": 0.7660287598659314,
+      "grad_norm": 0.15946455299854279,
+      "learning_rate": 4.803770047198011e-05,
+      "loss": 0.4366,
       "step": 21255
     },
     {
-      "epoch": 0.75,
-      "learning_rate": 4.814926599356052e-05,
-      "loss": 0.2986,
+      "epoch": 0.7662089595271561,
+      "grad_norm": 0.1485140323638916,
+      "learning_rate": 4.8036567030603676e-05,
+      "loss": 0.4624,
       "step": 21260
     },
     {
-      "epoch": 0.75,
-      "learning_rate": 4.8148190178877414e-05,
-      "loss": 0.3171,
+      "epoch": 0.7663891591883807,
+      "grad_norm": 0.15414132177829742,
+      "learning_rate": 4.803543327535848e-05,
+      "loss": 0.4321,
       "step": 21265
     },
     {
-      "epoch": 0.75,
-      "learning_rate": 4.814711406363032e-05,
-      "loss": 0.3302,
+      "epoch": 0.7665693588496053,
+      "grad_norm": 0.17065320909023285,
+      "learning_rate": 4.803429920625996e-05,
+      "loss": 0.4396,
       "step": 21270
     },
     {
-      "epoch": 0.75,
-      "learning_rate": 4.81460376478332e-05,
-      "loss": 0.3228,
+      "epoch": 0.76674955851083,
+      "grad_norm": 0.15544413030147552,
+      "learning_rate": 4.803316482332358e-05,
+      "loss": 0.4714,
       "step": 21275
     },
     {
-      "epoch": 0.75,
-      "learning_rate": 4.8144960931500026e-05,
-      "loss": 0.3102,
+      "epoch": 0.7669297581720547,
+      "grad_norm": 0.22066541016101837,
+      "learning_rate": 4.803203012656479e-05,
+      "loss": 0.4651,
       "step": 21280
     },
     {
-      "epoch": 0.75,
-      "learning_rate": 4.8143883914644796e-05,
-      "loss": 0.2783,
+      "epoch": 0.7671099578332793,
+      "grad_norm": 0.21565388143062592,
+      "learning_rate": 4.803089511599904e-05,
+      "loss": 0.4593,
       "step": 21285
     },
     {
-      "epoch": 0.75,
-      "learning_rate": 4.814280659728149e-05,
-      "loss": 0.2945,
+      "epoch": 0.7672901574945039,
+      "grad_norm": 0.15955856442451477,
+      "learning_rate": 4.8029759791641804e-05,
+      "loss": 0.4406,
       "step": 21290
     },
     {
-      "epoch": 0.75,
-      "learning_rate": 4.8141728979424085e-05,
-      "loss": 0.2845,
+      "epoch": 0.7674703571557285,
+      "grad_norm": 0.19206349551677704,
+      "learning_rate": 4.8028624153508555e-05,
+      "loss": 0.4127,
       "step": 21295
     },
     {
-      "epoch": 0.75,
-      "learning_rate": 4.8140651061086584e-05,
-      "loss": 0.309,
+      "epoch": 0.7676505568169532,
+      "grad_norm": 0.15551474690437317,
+      "learning_rate": 4.8027488201614754e-05,
+      "loss": 0.4212,
       "step": 21300
     },
     {
-      "epoch": 0.75,
-      "learning_rate": 4.8139572842282975e-05,
-      "loss": 0.3261,
+      "epoch": 0.7678307564781778,
+      "grad_norm": 0.16938930749893188,
+      "learning_rate": 4.802635193597589e-05,
+      "loss": 0.4515,
       "step": 21305
     },
     {
-      "epoch": 0.75,
-      "learning_rate": 4.813849432302726e-05,
-      "loss": 0.3056,
+      "epoch": 0.7680109561394025,
+      "grad_norm": 0.16218577325344086,
+      "learning_rate": 4.802521535660744e-05,
+      "loss": 0.4228,
       "step": 21310
     },
     {
-      "epoch": 0.75,
-      "learning_rate": 4.813741550333345e-05,
-      "loss": 0.2973,
+      "epoch": 0.7681911558006271,
+      "grad_norm": 0.14813528954982758,
+      "learning_rate": 4.802407846352488e-05,
+      "loss": 0.4586,
       "step": 21315
     },
     {
-      "epoch": 0.75,
-      "learning_rate": 4.8136336383215554e-05,
-      "loss": 0.3224,
+      "epoch": 0.7683713554618518,
+      "grad_norm": 0.17613081634044647,
+      "learning_rate": 4.802294125674372e-05,
+      "loss": 0.4574,
       "step": 21320
     },
     {
-      "epoch": 0.75,
-      "learning_rate": 4.813525696268757e-05,
-      "loss": 0.2816,
+      "epoch": 0.7685515551230764,
+      "grad_norm": 0.1591581255197525,
+      "learning_rate": 4.8021803736279435e-05,
+      "loss": 0.4451,
       "step": 21325
     },
     {
-      "epoch": 0.75,
-      "learning_rate": 4.813417724176352e-05,
-      "loss": 0.291,
+      "epoch": 0.768731754784301,
+      "grad_norm": 0.16665023565292358,
+      "learning_rate": 4.8020665902147535e-05,
+      "loss": 0.4057,
       "step": 21330
     },
     {
-      "epoch": 0.75,
-      "learning_rate": 4.813309722045742e-05,
-      "loss": 0.3231,
+      "epoch": 0.7689119544455256,
+      "grad_norm": 0.17413721978664398,
+      "learning_rate": 4.801952775436352e-05,
+      "loss": 0.4655,
       "step": 21335
     },
     {
-      "epoch": 0.75,
-      "learning_rate": 4.813201689878331e-05,
-      "loss": 0.2945,
+      "epoch": 0.7690921541067502,
+      "grad_norm": 0.1840430349111557,
+      "learning_rate": 4.8018389292942886e-05,
+      "loss": 0.4568,
       "step": 21340
     },
     {
-      "epoch": 0.75,
-      "learning_rate": 4.8130936276755204e-05,
-      "loss": 0.3041,
+      "epoch": 0.7692723537679749,
+      "grad_norm": 0.158147931098938,
+      "learning_rate": 4.801725051790117e-05,
+      "loss": 0.4459,
       "step": 21345
     },
     {
-      "epoch": 0.75,
-      "learning_rate": 4.8129855354387125e-05,
-      "loss": 0.2851,
+      "epoch": 0.7694525534291996,
+      "grad_norm": 0.19126906991004944,
+      "learning_rate": 4.801611142925386e-05,
+      "loss": 0.4429,
       "step": 21350
     },
     {
-      "epoch": 0.75,
-      "learning_rate": 4.812877413169312e-05,
-      "loss": 0.3145,
+      "epoch": 0.7696327530904242,
+      "grad_norm": 0.22057095170021057,
+      "learning_rate": 4.801497202701649e-05,
+      "loss": 0.4369,
       "step": 21355
     },
     {
-      "epoch": 0.75,
-      "learning_rate": 4.812769260868723e-05,
-      "loss": 0.3013,
+      "epoch": 0.7698129527516488,
+      "grad_norm": 0.16785407066345215,
+      "learning_rate": 4.8013832311204586e-05,
+      "loss": 0.4594,
       "step": 21360
     },
     {
-      "epoch": 0.75,
-      "learning_rate": 4.812661078538349e-05,
-      "loss": 0.2727,
+      "epoch": 0.7699931524128735,
+      "grad_norm": 0.17427514493465424,
+      "learning_rate": 4.801269228183367e-05,
+      "loss": 0.4408,
       "step": 21365
     },
     {
-      "epoch": 0.75,
-      "learning_rate": 4.812552866179595e-05,
-      "loss": 0.3139,
+      "epoch": 0.7701733520740981,
+      "grad_norm": 0.14857777953147888,
+      "learning_rate": 4.8011551938919283e-05,
+      "loss": 0.4515,
       "step": 21370
     },
     {
-      "epoch": 0.75,
-      "learning_rate": 4.8124446237938665e-05,
-      "loss": 0.3225,
+      "epoch": 0.7703535517353227,
+      "grad_norm": 0.19615520536899567,
+      "learning_rate": 4.801041128247695e-05,
+      "loss": 0.4395,
       "step": 21375
     },
     {
-      "epoch": 0.75,
-      "learning_rate": 4.812336351382568e-05,
-      "loss": 0.3011,
+      "epoch": 0.7705337513965473,
+      "grad_norm": 0.46812736988067627,
+      "learning_rate": 4.800927031252222e-05,
+      "loss": 0.3949,
       "step": 21380
     },
     {
-      "epoch": 0.75,
-      "learning_rate": 4.8122280489471064e-05,
-      "loss": 0.3018,
+      "epoch": 0.770713951057772,
+      "grad_norm": 0.18027018010616302,
+      "learning_rate": 4.800812902907063e-05,
+      "loss": 0.446,
       "step": 21385
     },
     {
-      "epoch": 0.75,
-      "learning_rate": 4.812119716488886e-05,
-      "loss": 0.2842,
+      "epoch": 0.7708941507189967,
+      "grad_norm": 0.1679840236902237,
+      "learning_rate": 4.800698743213774e-05,
+      "loss": 0.4554,
       "step": 21390
     },
     {
-      "epoch": 0.75,
-      "learning_rate": 4.812011354009317e-05,
-      "loss": 0.2962,
+      "epoch": 0.7710743503802213,
+      "grad_norm": 0.19741559028625488,
+      "learning_rate": 4.80058455217391e-05,
+      "loss": 0.4695,
       "step": 21395
     },
     {
-      "epoch": 0.75,
-      "learning_rate": 4.8119029615098024e-05,
-      "loss": 0.3168,
+      "epoch": 0.7712545500414459,
+      "grad_norm": 0.19206707179546356,
+      "learning_rate": 4.800470329789027e-05,
+      "loss": 0.4572,
       "step": 21400
     },
     {
-      "epoch": 0.75,
-      "learning_rate": 4.811794538991752e-05,
-      "loss": 0.2889,
+      "epoch": 0.7714347497026706,
+      "grad_norm": 0.17079663276672363,
+      "learning_rate": 4.800356076060682e-05,
+      "loss": 0.4186,
       "step": 21405
     },
     {
-      "epoch": 0.75,
-      "learning_rate": 4.811686086456574e-05,
-      "loss": 0.3059,
+      "epoch": 0.7716149493638952,
+      "grad_norm": 0.20327655971050262,
+      "learning_rate": 4.800241790990429e-05,
+      "loss": 0.4195,
       "step": 21410
     },
     {
-      "epoch": 0.75,
-      "learning_rate": 4.811577603905674e-05,
-      "loss": 0.2992,
+      "epoch": 0.7717951490251198,
+      "grad_norm": 0.1791381984949112,
+      "learning_rate": 4.8001274745798286e-05,
+      "loss": 0.457,
       "step": 21415
     },
     {
-      "epoch": 0.75,
-      "learning_rate": 4.8114690913404634e-05,
-      "loss": 0.3163,
+      "epoch": 0.7719753486863444,
+      "grad_norm": 0.1447453647851944,
+      "learning_rate": 4.800013126830437e-05,
+      "loss": 0.4155,
       "step": 21420
     },
     {
-      "epoch": 0.75,
-      "learning_rate": 4.811360548762349e-05,
-      "loss": 0.2916,
+      "epoch": 0.772155548347569,
+      "grad_norm": 0.17693381011486053,
+      "learning_rate": 4.799898747743811e-05,
+      "loss": 0.3966,
       "step": 21425
     },
     {
-      "epoch": 0.75,
-      "learning_rate": 4.811251976172742e-05,
-      "loss": 0.2905,
+      "epoch": 0.7723357480087938,
+      "grad_norm": 0.2063083052635193,
+      "learning_rate": 4.799784337321509e-05,
+      "loss": 0.4712,
       "step": 21430
     },
     {
-      "epoch": 0.75,
-      "learning_rate": 4.811143373573051e-05,
-      "loss": 0.313,
+      "epoch": 0.7725159476700184,
+      "grad_norm": 0.217716246843338,
+      "learning_rate": 4.799669895565092e-05,
+      "loss": 0.4261,
       "step": 21435
     },
     {
-      "epoch": 0.75,
-      "learning_rate": 4.811034740964686e-05,
-      "loss": 0.3108,
+      "epoch": 0.772696147331243,
+      "grad_norm": 0.13754284381866455,
+      "learning_rate": 4.799555422476117e-05,
+      "loss": 0.4144,
       "step": 21440
     },
     {
-      "epoch": 0.75,
-      "learning_rate": 4.810926078349058e-05,
-      "loss": 0.3084,
+      "epoch": 0.7728763469924677,
+      "grad_norm": 0.1717095524072647,
+      "learning_rate": 4.799440918056145e-05,
+      "loss": 0.4505,
       "step": 21445
     },
     {
-      "epoch": 0.75,
-      "learning_rate": 4.8108173857275785e-05,
-      "loss": 0.3066,
+      "epoch": 0.7730565466536923,
+      "grad_norm": 0.16796359419822693,
+      "learning_rate": 4.7993263823067355e-05,
+      "loss": 0.4523,
       "step": 21450
     },
     {
-      "epoch": 0.75,
-      "learning_rate": 4.810708663101658e-05,
-      "loss": 0.3084,
+      "epoch": 0.7732367463149169,
+      "grad_norm": 0.1903134435415268,
+      "learning_rate": 4.79921181522945e-05,
+      "loss": 0.4197,
       "step": 21455
     },
     {
-      "epoch": 0.76,
-      "learning_rate": 4.8105999104727086e-05,
-      "loss": 0.2944,
+      "epoch": 0.7734169459761415,
+      "grad_norm": 0.1834080070257187,
+      "learning_rate": 4.799097216825847e-05,
+      "loss": 0.4232,
       "step": 21460
     },
     {
-      "epoch": 0.76,
-      "learning_rate": 4.810491127842141e-05,
-      "loss": 0.317,
+      "epoch": 0.7735971456373663,
+      "grad_norm": 0.17639869451522827,
+      "learning_rate": 4.7989825870974904e-05,
+      "loss": 0.4085,
       "step": 21465
     },
     {
-      "epoch": 0.76,
-      "learning_rate": 4.8103823152113695e-05,
-      "loss": 0.3204,
+      "epoch": 0.7737773452985909,
+      "grad_norm": 0.18790148198604584,
+      "learning_rate": 4.798867926045941e-05,
+      "loss": 0.4522,
       "step": 21470
     },
     {
-      "epoch": 0.76,
-      "learning_rate": 4.8102734725818056e-05,
-      "loss": 0.3106,
+      "epoch": 0.7739575449598155,
+      "grad_norm": 0.17099972069263458,
+      "learning_rate": 4.798753233672762e-05,
+      "loss": 0.4409,
       "step": 21475
     },
     {
-      "epoch": 0.76,
-      "learning_rate": 4.810164599954864e-05,
-      "loss": 0.3186,
+      "epoch": 0.7741377446210401,
+      "grad_norm": 0.15639808773994446,
+      "learning_rate": 4.798638509979514e-05,
+      "loss": 0.4576,
       "step": 21480
     },
     {
-      "epoch": 0.76,
-      "learning_rate": 4.810055697331957e-05,
-      "loss": 0.2905,
+      "epoch": 0.7743179442822647,
+      "grad_norm": 0.19810642302036285,
+      "learning_rate": 4.7985237549677624e-05,
+      "loss": 0.4336,
       "step": 21485
     },
     {
-      "epoch": 0.76,
-      "learning_rate": 4.809946764714499e-05,
-      "loss": 0.3014,
+      "epoch": 0.7744981439434894,
+      "grad_norm": 0.192165344953537,
+      "learning_rate": 4.79840896863907e-05,
+      "loss": 0.4481,
       "step": 21490
     },
     {
-      "epoch": 0.76,
-      "learning_rate": 4.809837802103905e-05,
-      "loss": 0.2906,
+      "epoch": 0.774678343604714,
+      "grad_norm": 0.17463374137878418,
+      "learning_rate": 4.798294150994999e-05,
+      "loss": 0.4444,
       "step": 21495
     },
     {
-      "epoch": 0.76,
-      "learning_rate": 4.80972880950159e-05,
-      "loss": 0.2881,
+      "epoch": 0.7748585432659386,
+      "grad_norm": 0.1958521455526352,
+      "learning_rate": 4.798179302037116e-05,
+      "loss": 0.4435,
       "step": 21500
     },
     {
-      "epoch": 0.76,
-      "eval_loss": 0.30054381489753723,
-      "eval_runtime": 10.5336,
-      "eval_samples_per_second": 9.493,
-      "eval_steps_per_second": 9.493,
+      "epoch": 0.7748585432659386,
+      "eval_loss": 0.4604308009147644,
+      "eval_runtime": 3.5333,
+      "eval_samples_per_second": 28.302,
+      "eval_steps_per_second": 7.075,
       "step": 21500
     },
     {
-      "epoch": 0.76,
-      "learning_rate": 4.809619786908968e-05,
-      "loss": 0.2831,
+      "epoch": 0.7750387429271633,
+      "grad_norm": 0.18340334296226501,
+      "learning_rate": 4.798064421766985e-05,
+      "loss": 0.4139,
       "step": 21505
     },
     {
-      "epoch": 0.76,
-      "learning_rate": 4.809510734327455e-05,
-      "loss": 0.3034,
+      "epoch": 0.775218942588388,
+      "grad_norm": 0.18603767454624176,
+      "learning_rate": 4.7979495101861705e-05,
+      "loss": 0.418,
       "step": 21510
     },
     {
-      "epoch": 0.76,
-      "learning_rate": 4.809401651758466e-05,
-      "loss": 0.3227,
+      "epoch": 0.7753991422496126,
+      "grad_norm": 0.21015454828739166,
+      "learning_rate": 4.7978345672962395e-05,
+      "loss": 0.4884,
       "step": 21515
     },
     {
-      "epoch": 0.76,
-      "learning_rate": 4.809292539203421e-05,
-      "loss": 0.3268,
+      "epoch": 0.7755793419108372,
+      "grad_norm": 0.16696617007255554,
+      "learning_rate": 4.797719593098757e-05,
+      "loss": 0.4452,
       "step": 21520
     },
     {
-      "epoch": 0.76,
-      "learning_rate": 4.809183396663732e-05,
-      "loss": 0.3147,
+      "epoch": 0.7757595415720618,
+      "grad_norm": 0.19136182963848114,
+      "learning_rate": 4.79760458759529e-05,
+      "loss": 0.449,
       "step": 21525
     },
     {
-      "epoch": 0.76,
-      "learning_rate": 4.809074224140819e-05,
-      "loss": 0.2981,
+      "epoch": 0.7759397412332865,
+      "grad_norm": 0.14051967859268188,
+      "learning_rate": 4.797489550787405e-05,
+      "loss": 0.4396,
       "step": 21530
     },
     {
-      "epoch": 0.76,
-      "learning_rate": 4.808965021636099e-05,
-      "loss": 0.2945,
+      "epoch": 0.7761199408945111,
+      "grad_norm": 0.15521185100078583,
+      "learning_rate": 4.7973744826766706e-05,
+      "loss": 0.479,
       "step": 21535
     },
     {
-      "epoch": 0.76,
-      "learning_rate": 4.808855789150991e-05,
-      "loss": 0.3062,
+      "epoch": 0.7763001405557357,
+      "grad_norm": 0.1503034085035324,
+      "learning_rate": 4.797259383264653e-05,
+      "loss": 0.4187,
       "step": 21540
     },
     {
-      "epoch": 0.76,
-      "learning_rate": 4.8087465266869115e-05,
-      "loss": 0.3048,
+      "epoch": 0.7764803402169604,
+      "grad_norm": 0.18658557534217834,
+      "learning_rate": 4.7971442525529206e-05,
+      "loss": 0.4604,
       "step": 21545
     },
     {
-      "epoch": 0.76,
-      "learning_rate": 4.808637234245279e-05,
-      "loss": 0.3082,
+      "epoch": 0.7766605398781851,
+      "grad_norm": 0.19798524677753448,
+      "learning_rate": 4.797029090543044e-05,
+      "loss": 0.447,
       "step": 21550
     },
     {
-      "epoch": 0.76,
-      "learning_rate": 4.808527911827514e-05,
-      "loss": 0.2964,
+      "epoch": 0.7768407395394097,
+      "grad_norm": 0.21234340965747833,
+      "learning_rate": 4.796913897236589e-05,
+      "loss": 0.4392,
       "step": 21555
     },
     {
-      "epoch": 0.76,
-      "learning_rate": 4.808418559435036e-05,
-      "loss": 0.2923,
+      "epoch": 0.7770209392006343,
+      "grad_norm": 0.16527117788791656,
+      "learning_rate": 4.796798672635128e-05,
+      "loss": 0.4516,
       "step": 21560
     },
     {
-      "epoch": 0.76,
-      "learning_rate": 4.8083091770692635e-05,
-      "loss": 0.3136,
+      "epoch": 0.7772011388618589,
+      "grad_norm": 0.20785808563232422,
+      "learning_rate": 4.7966834167402295e-05,
+      "loss": 0.4954,
       "step": 21565
     },
     {
-      "epoch": 0.76,
-      "learning_rate": 4.808199764731618e-05,
-      "loss": 0.3084,
+      "epoch": 0.7773813385230836,
+      "grad_norm": 0.15618237853050232,
+      "learning_rate": 4.7965681295534635e-05,
+      "loss": 0.399,
       "step": 21570
     },
     {
-      "epoch": 0.76,
-      "learning_rate": 4.80809032242352e-05,
-      "loss": 0.3067,
+      "epoch": 0.7775615381843082,
+      "grad_norm": 0.1682867556810379,
+      "learning_rate": 4.7964528110764026e-05,
+      "loss": 0.4513,
       "step": 21575
     },
     {
-      "epoch": 0.76,
-      "learning_rate": 4.8079808501463896e-05,
-      "loss": 0.2922,
+      "epoch": 0.7777417378455328,
+      "grad_norm": 0.1370386779308319,
+      "learning_rate": 4.796337461310616e-05,
+      "loss": 0.4479,
       "step": 21580
     },
     {
-      "epoch": 0.76,
-      "learning_rate": 4.807871347901649e-05,
-      "loss": 0.2992,
+      "epoch": 0.7779219375067575,
+      "grad_norm": 0.41439640522003174,
+      "learning_rate": 4.796222080257676e-05,
+      "loss": 0.4443,
       "step": 21585
     },
     {
-      "epoch": 0.76,
-      "learning_rate": 4.807761815690721e-05,
-      "loss": 0.3009,
+      "epoch": 0.7781021371679822,
+      "grad_norm": 0.1332816630601883,
+      "learning_rate": 4.7961066679191544e-05,
+      "loss": 0.429,
       "step": 21590
     },
     {
-      "epoch": 0.76,
-      "learning_rate": 4.807652253515025e-05,
-      "loss": 0.2782,
+      "epoch": 0.7782823368292068,
+      "grad_norm": 0.20065148174762726,
+      "learning_rate": 4.7959912242966245e-05,
+      "loss": 0.4144,
       "step": 21595
     },
     {
-      "epoch": 0.76,
-      "learning_rate": 4.8075426613759865e-05,
-      "loss": 0.2967,
+      "epoch": 0.7784625364904314,
+      "grad_norm": 0.18139199912548065,
+      "learning_rate": 4.795875749391659e-05,
+      "loss": 0.4245,
       "step": 21600
     },
     {
-      "epoch": 0.76,
-      "learning_rate": 4.807433039275027e-05,
-      "loss": 0.3211,
+      "epoch": 0.778642736151656,
+      "grad_norm": 0.1368846297264099,
+      "learning_rate": 4.79576024320583e-05,
+      "loss": 0.4475,
       "step": 21605
     },
     {
-      "epoch": 0.76,
-      "learning_rate": 4.80732338721357e-05,
-      "loss": 0.3165,
+      "epoch": 0.7788229358128806,
+      "grad_norm": 0.21975240111351013,
+      "learning_rate": 4.7956447057407125e-05,
+      "loss": 0.4563,
       "step": 21610
     },
     {
-      "epoch": 0.76,
-      "learning_rate": 4.807213705193039e-05,
-      "loss": 0.3044,
+      "epoch": 0.7790031354741053,
+      "grad_norm": 0.16079792380332947,
+      "learning_rate": 4.795529136997881e-05,
+      "loss": 0.4176,
       "step": 21615
     },
     {
-      "epoch": 0.76,
-      "learning_rate": 4.8071039932148585e-05,
-      "loss": 0.3058,
+      "epoch": 0.7791833351353299,
+      "grad_norm": 0.16716989874839783,
+      "learning_rate": 4.795413536978909e-05,
+      "loss": 0.4083,
       "step": 21620
     },
     {
-      "epoch": 0.76,
-      "learning_rate": 4.806994251280454e-05,
-      "loss": 0.3035,
+      "epoch": 0.7793635347965546,
+      "grad_norm": 0.21081286668777466,
+      "learning_rate": 4.795297905685372e-05,
+      "loss": 0.4439,
       "step": 21625
     },
     {
-      "epoch": 0.76,
-      "learning_rate": 4.806884479391249e-05,
-      "loss": 0.3142,
+      "epoch": 0.7795437344577792,
+      "grad_norm": 0.17637106776237488,
+      "learning_rate": 4.7951822431188455e-05,
+      "loss": 0.4057,
       "step": 21630
     },
     {
-      "epoch": 0.76,
-      "learning_rate": 4.8067746775486696e-05,
-      "loss": 0.3171,
+      "epoch": 0.7797239341190039,
+      "grad_norm": 0.1675841063261032,
+      "learning_rate": 4.7950665492809047e-05,
+      "loss": 0.467,
       "step": 21635
     },
     {
-      "epoch": 0.76,
-      "learning_rate": 4.806664845754141e-05,
-      "loss": 0.3367,
+      "epoch": 0.7799041337802285,
+      "grad_norm": 0.1935756951570511,
+      "learning_rate": 4.794950824173127e-05,
+      "loss": 0.3958,
       "step": 21640
     },
     {
-      "epoch": 0.76,
-      "learning_rate": 4.806554984009089e-05,
-      "loss": 0.3105,
+      "epoch": 0.7800843334414531,
+      "grad_norm": 0.18767641484737396,
+      "learning_rate": 4.794835067797089e-05,
+      "loss": 0.4528,
       "step": 21645
     },
     {
-      "epoch": 0.76,
-      "learning_rate": 4.806445092314941e-05,
-      "loss": 0.3224,
+      "epoch": 0.7802645331026777,
+      "grad_norm": 0.1749449372291565,
+      "learning_rate": 4.794719280154367e-05,
+      "loss": 0.4452,
       "step": 21650
     },
     {
-      "epoch": 0.76,
-      "learning_rate": 4.806335170673124e-05,
-      "loss": 0.2948,
+      "epoch": 0.7804447327639024,
+      "grad_norm": 0.13821455836296082,
+      "learning_rate": 4.794603461246539e-05,
+      "loss": 0.4487,
       "step": 21655
     },
     {
-      "epoch": 0.76,
-      "learning_rate": 4.806225219085064e-05,
-      "loss": 0.2976,
+      "epoch": 0.7806249324251271,
+      "grad_norm": 0.15566319227218628,
+      "learning_rate": 4.794487611075184e-05,
+      "loss": 0.4393,
       "step": 21660
     },
     {
-      "epoch": 0.76,
-      "learning_rate": 4.80611523755219e-05,
-      "loss": 0.3046,
+      "epoch": 0.7808051320863517,
+      "grad_norm": 0.17537739872932434,
+      "learning_rate": 4.794371729641878e-05,
+      "loss": 0.4475,
       "step": 21665
     },
     {
-      "epoch": 0.76,
-      "learning_rate": 4.806005226075929e-05,
-      "loss": 0.3117,
+      "epoch": 0.7809853317475763,
+      "grad_norm": 0.1799578070640564,
+      "learning_rate": 4.794255816948202e-05,
+      "loss": 0.4405,
       "step": 21670
     },
     {
-      "epoch": 0.76,
-      "learning_rate": 4.80589518465771e-05,
-      "loss": 0.3219,
+      "epoch": 0.781165531408801,
+      "grad_norm": 0.1784065216779709,
+      "learning_rate": 4.794139872995736e-05,
+      "loss": 0.4588,
       "step": 21675
     },
     {
-      "epoch": 0.76,
-      "learning_rate": 4.805785113298962e-05,
-      "loss": 0.3039,
+      "epoch": 0.7813457310700256,
+      "grad_norm": 0.18928910791873932,
+      "learning_rate": 4.7940238977860563e-05,
+      "loss": 0.4349,
       "step": 21680
     },
     {
-      "epoch": 0.76,
-      "learning_rate": 4.805675012001113e-05,
-      "loss": 0.3166,
+      "epoch": 0.7815259307312502,
+      "grad_norm": 0.15001779794692993,
+      "learning_rate": 4.793907891320746e-05,
+      "loss": 0.3882,
       "step": 21685
     },
     {
-      "epoch": 0.76,
-      "learning_rate": 4.805564880765594e-05,
-      "loss": 0.299,
+      "epoch": 0.7817061303924748,
+      "grad_norm": 0.21316440403461456,
+      "learning_rate": 4.793791853601385e-05,
+      "loss": 0.4141,
       "step": 21690
     },
     {
-      "epoch": 0.76,
-      "learning_rate": 4.8054547195938345e-05,
-      "loss": 0.3175,
+      "epoch": 0.7818863300536995,
+      "grad_norm": 0.1485886126756668,
+      "learning_rate": 4.793675784629554e-05,
+      "loss": 0.444,
       "step": 21695
     },
     {
-      "epoch": 0.76,
-      "learning_rate": 4.805344528487266e-05,
-      "loss": 0.2838,
+      "epoch": 0.7820665297149242,
+      "grad_norm": 0.17598822712898254,
+      "learning_rate": 4.7935596844068343e-05,
+      "loss": 0.4348,
       "step": 21700
     },
     {
-      "epoch": 0.76,
-      "learning_rate": 4.805234307447316e-05,
-      "loss": 0.2707,
+      "epoch": 0.7822467293761488,
+      "grad_norm": 0.1468224823474884,
+      "learning_rate": 4.793443552934808e-05,
+      "loss": 0.4099,
       "step": 21705
     },
     {
-      "epoch": 0.76,
-      "learning_rate": 4.8051240564754184e-05,
-      "loss": 0.3001,
+      "epoch": 0.7824269290373734,
+      "grad_norm": 0.21252159774303436,
+      "learning_rate": 4.793327390215058e-05,
+      "loss": 0.4587,
       "step": 21710
     },
     {
-      "epoch": 0.76,
-      "learning_rate": 4.805013775573004e-05,
-      "loss": 0.3248,
+      "epoch": 0.782607128698598,
+      "grad_norm": 0.1317516267299652,
+      "learning_rate": 4.7932111962491654e-05,
+      "loss": 0.4466,
       "step": 21715
     },
     {
-      "epoch": 0.76,
-      "learning_rate": 4.804903464741506e-05,
-      "loss": 0.2908,
+      "epoch": 0.7827873283598227,
+      "grad_norm": 0.17863036692142487,
+      "learning_rate": 4.7930949710387145e-05,
+      "loss": 0.436,
       "step": 21720
     },
     {
-      "epoch": 0.76,
-      "learning_rate": 4.804793123982354e-05,
-      "loss": 0.3068,
+      "epoch": 0.7829675280210473,
+      "grad_norm": 0.1505396068096161,
+      "learning_rate": 4.792978714585289e-05,
+      "loss": 0.4427,
       "step": 21725
     },
     {
-      "epoch": 0.76,
-      "learning_rate": 4.804682753296983e-05,
-      "loss": 0.3034,
+      "epoch": 0.7831477276822719,
+      "grad_norm": 0.17207299172878265,
+      "learning_rate": 4.7928624268904724e-05,
+      "loss": 0.3839,
       "step": 21730
     },
     {
-      "epoch": 0.76,
-      "learning_rate": 4.8045723526868245e-05,
-      "loss": 0.3117,
+      "epoch": 0.7833279273434965,
+      "grad_norm": 0.1621370017528534,
+      "learning_rate": 4.7927461079558476e-05,
+      "loss": 0.4446,
       "step": 21735
     },
     {
-      "epoch": 0.76,
-      "learning_rate": 4.8044619221533135e-05,
-      "loss": 0.2958,
+      "epoch": 0.7835081270047213,
+      "grad_norm": 0.1606713980436325,
+      "learning_rate": 4.792629757783003e-05,
+      "loss": 0.4394,
       "step": 21740
     },
     {
-      "epoch": 0.77,
-      "learning_rate": 4.804351461697883e-05,
-      "loss": 0.3404,
+      "epoch": 0.7836883266659459,
+      "grad_norm": 0.15808479487895966,
+      "learning_rate": 4.792513376373521e-05,
+      "loss": 0.4333,
       "step": 21745
     },
     {
-      "epoch": 0.77,
-      "learning_rate": 4.804240971321967e-05,
-      "loss": 0.2699,
+      "epoch": 0.7838685263271705,
+      "grad_norm": 0.19757013022899628,
+      "learning_rate": 4.7923969637289875e-05,
+      "loss": 0.4498,
       "step": 21750
     },
     {
-      "epoch": 0.77,
-      "learning_rate": 4.8041304510270016e-05,
-      "loss": 0.3103,
+      "epoch": 0.7840487259883951,
+      "grad_norm": 0.1744808852672577,
+      "learning_rate": 4.7922805198509905e-05,
+      "loss": 0.4234,
       "step": 21755
     },
     {
-      "epoch": 0.77,
-      "learning_rate": 4.804019900814419e-05,
-      "loss": 0.2829,
+      "epoch": 0.7842289256496198,
+      "grad_norm": 0.14034539461135864,
+      "learning_rate": 4.7921640447411146e-05,
+      "loss": 0.428,
       "step": 21760
     },
     {
-      "epoch": 0.77,
-      "learning_rate": 4.803909320685658e-05,
-      "loss": 0.3489,
+      "epoch": 0.7844091253108444,
+      "grad_norm": 0.12732501327991486,
+      "learning_rate": 4.792047538400947e-05,
+      "loss": 0.415,
       "step": 21765
     },
     {
-      "epoch": 0.77,
-      "learning_rate": 4.8037987106421525e-05,
-      "loss": 0.3017,
+      "epoch": 0.784589324972069,
+      "grad_norm": 0.13818371295928955,
+      "learning_rate": 4.791931000832076e-05,
+      "loss": 0.4057,
       "step": 21770
     },
     {
-      "epoch": 0.77,
-      "learning_rate": 4.8036880706853385e-05,
-      "loss": 0.3193,
+      "epoch": 0.7847695246332936,
+      "grad_norm": 0.2030201405286789,
+      "learning_rate": 4.791814432036088e-05,
+      "loss": 0.4127,
       "step": 21775
     },
     {
-      "epoch": 0.77,
-      "learning_rate": 4.803577400816654e-05,
-      "loss": 0.2851,
+      "epoch": 0.7849497242945184,
+      "grad_norm": 0.13716308772563934,
+      "learning_rate": 4.791697832014573e-05,
+      "loss": 0.4666,
       "step": 21780
     },
     {
-      "epoch": 0.77,
-      "learning_rate": 4.803466701037534e-05,
-      "loss": 0.3184,
+      "epoch": 0.785129923955743,
+      "grad_norm": 0.21387408673763275,
+      "learning_rate": 4.791581200769118e-05,
+      "loss": 0.4192,
       "step": 21785
     },
     {
-      "epoch": 0.77,
-      "learning_rate": 4.803355971349417e-05,
-      "loss": 0.3362,
+      "epoch": 0.7853101236169676,
+      "grad_norm": 0.21067731082439423,
+      "learning_rate": 4.7914645383013134e-05,
+      "loss": 0.4534,
       "step": 21790
     },
     {
-      "epoch": 0.77,
-      "learning_rate": 4.803245211753741e-05,
-      "loss": 0.2961,
+      "epoch": 0.7854903232781922,
+      "grad_norm": 0.18593548238277435,
+      "learning_rate": 4.791347844612748e-05,
+      "loss": 0.4512,
       "step": 21795
     },
     {
-      "epoch": 0.77,
-      "learning_rate": 4.8031344222519434e-05,
-      "loss": 0.2965,
+      "epoch": 0.7856705229394169,
+      "grad_norm": 0.18970215320587158,
+      "learning_rate": 4.7912311197050115e-05,
+      "loss": 0.4188,
       "step": 21800
     },
     {
-      "epoch": 0.77,
-      "learning_rate": 4.803023602845465e-05,
-      "loss": 0.331,
+      "epoch": 0.7858507226006415,
+      "grad_norm": 0.15977688133716583,
+      "learning_rate": 4.791114363579695e-05,
+      "loss": 0.4136,
       "step": 21805
     },
     {
-      "epoch": 0.77,
-      "learning_rate": 4.802912753535741e-05,
-      "loss": 0.3147,
+      "epoch": 0.7860309222618661,
+      "grad_norm": 0.17266367375850677,
+      "learning_rate": 4.790997576238389e-05,
+      "loss": 0.4518,
       "step": 21810
     },
     {
-      "epoch": 0.77,
-      "learning_rate": 4.802801874324213e-05,
-      "loss": 0.3112,
+      "epoch": 0.7862111219230908,
+      "grad_norm": 0.20080499351024628,
+      "learning_rate": 4.790880757682684e-05,
+      "loss": 0.427,
       "step": 21815
     },
     {
-      "epoch": 0.77,
-      "learning_rate": 4.8026909652123196e-05,
-      "loss": 0.316,
+      "epoch": 0.7863913215843155,
+      "grad_norm": 0.18386758863925934,
+      "learning_rate": 4.790763907914172e-05,
+      "loss": 0.4013,
       "step": 21820
     },
     {
-      "epoch": 0.77,
-      "learning_rate": 4.802580026201502e-05,
-      "loss": 0.3191,
+      "epoch": 0.7865715212455401,
+      "grad_norm": 0.17184041440486908,
+      "learning_rate": 4.790647026934446e-05,
+      "loss": 0.4037,
       "step": 21825
     },
     {
-      "epoch": 0.77,
-      "learning_rate": 4.802469057293201e-05,
-      "loss": 0.3106,
+      "epoch": 0.7867517209067647,
+      "grad_norm": 0.15970267355442047,
+      "learning_rate": 4.790530114745097e-05,
+      "loss": 0.4333,
       "step": 21830
     },
     {
-      "epoch": 0.77,
-      "learning_rate": 4.802358058488856e-05,
-      "loss": 0.3008,
+      "epoch": 0.7869319205679893,
+      "grad_norm": 0.17542044818401337,
+      "learning_rate": 4.7904131713477196e-05,
+      "loss": 0.4588,
       "step": 21835
     },
     {
-      "epoch": 0.77,
-      "learning_rate": 4.802247029789909e-05,
-      "loss": 0.3206,
+      "epoch": 0.787112120229214,
+      "grad_norm": 0.16546791791915894,
+      "learning_rate": 4.790296196743905e-05,
+      "loss": 0.4421,
       "step": 21840
     },
     {
-      "epoch": 0.77,
-      "learning_rate": 4.802135971197802e-05,
-      "loss": 0.3182,
+      "epoch": 0.7872923198904386,
+      "grad_norm": 0.1802026927471161,
+      "learning_rate": 4.790179190935249e-05,
+      "loss": 0.4335,
       "step": 21845
     },
     {
-      "epoch": 0.77,
-      "learning_rate": 4.802024882713976e-05,
-      "loss": 0.3298,
+      "epoch": 0.7874725195516632,
+      "grad_norm": 0.19843994081020355,
+      "learning_rate": 4.790062153923345e-05,
+      "loss": 0.4394,
       "step": 21850
     },
     {
-      "epoch": 0.77,
-      "learning_rate": 4.8019137643398736e-05,
-      "loss": 0.3019,
+      "epoch": 0.7876527192128879,
+      "grad_norm": 0.1890823394060135,
+      "learning_rate": 4.7899450857097875e-05,
+      "loss": 0.4763,
       "step": 21855
     },
     {
-      "epoch": 0.77,
-      "learning_rate": 4.801802616076939e-05,
-      "loss": 0.2883,
+      "epoch": 0.7878329188741126,
+      "grad_norm": 0.15386788547039032,
+      "learning_rate": 4.789827986296172e-05,
+      "loss": 0.4183,
       "step": 21860
     },
     {
-      "epoch": 0.77,
-      "learning_rate": 4.801691437926614e-05,
-      "loss": 0.2949,
+      "epoch": 0.7880131185353372,
+      "grad_norm": 0.17523469030857086,
+      "learning_rate": 4.789710855684092e-05,
+      "loss": 0.4289,
       "step": 21865
     },
     {
-      "epoch": 0.77,
-      "learning_rate": 4.801580229890342e-05,
-      "loss": 0.2882,
+      "epoch": 0.7881933181965618,
+      "grad_norm": 0.1903219223022461,
+      "learning_rate": 4.789593693875146e-05,
+      "loss": 0.4049,
       "step": 21870
     },
     {
-      "epoch": 0.77,
-      "learning_rate": 4.8014689919695685e-05,
-      "loss": 0.301,
+      "epoch": 0.7883735178577864,
+      "grad_norm": 0.21866121888160706,
+      "learning_rate": 4.7894765008709286e-05,
+      "loss": 0.4868,
       "step": 21875
     },
     {
-      "epoch": 0.77,
-      "learning_rate": 4.801357724165736e-05,
-      "loss": 0.3056,
+      "epoch": 0.788553717519011,
+      "grad_norm": 0.13972169160842896,
+      "learning_rate": 4.789359276673038e-05,
+      "loss": 0.4237,
       "step": 21880
     },
     {
-      "epoch": 0.77,
-      "learning_rate": 4.8012464264802906e-05,
-      "loss": 0.3166,
+      "epoch": 0.7887339171802357,
+      "grad_norm": 0.18218587338924408,
+      "learning_rate": 4.789242021283069e-05,
+      "loss": 0.4031,
       "step": 21885
     },
     {
-      "epoch": 0.77,
-      "learning_rate": 4.801135098914677e-05,
-      "loss": 0.2869,
+      "epoch": 0.7889141168414603,
+      "grad_norm": 0.1976306140422821,
+      "learning_rate": 4.789124734702622e-05,
+      "loss": 0.4757,
       "step": 21890
     },
     {
-      "epoch": 0.77,
-      "learning_rate": 4.80102374147034e-05,
-      "loss": 0.2947,
+      "epoch": 0.789094316502685,
+      "grad_norm": 0.15920376777648926,
+      "learning_rate": 4.789007416933293e-05,
+      "loss": 0.4281,
       "step": 21895
     },
     {
-      "epoch": 0.77,
-      "learning_rate": 4.800912354148727e-05,
-      "loss": 0.2838,
+      "epoch": 0.7892745161639096,
+      "grad_norm": 0.1940070241689682,
+      "learning_rate": 4.788890067976682e-05,
+      "loss": 0.4296,
       "step": 21900
     },
     {
-      "epoch": 0.77,
-      "learning_rate": 4.800800936951283e-05,
-      "loss": 0.308,
+      "epoch": 0.7894547158251343,
+      "grad_norm": 0.1800207644701004,
+      "learning_rate": 4.788772687834386e-05,
+      "loss": 0.4521,
       "step": 21905
     },
     {
-      "epoch": 0.77,
-      "learning_rate": 4.8006894898794554e-05,
-      "loss": 0.3103,
+      "epoch": 0.7896349154863589,
+      "grad_norm": 0.166525736451149,
+      "learning_rate": 4.7886552765080055e-05,
+      "loss": 0.4653,
       "step": 21910
     },
     {
-      "epoch": 0.77,
-      "learning_rate": 4.80057801293469e-05,
-      "loss": 0.3056,
+      "epoch": 0.7898151151475835,
+      "grad_norm": 0.19525845348834991,
+      "learning_rate": 4.78853783399914e-05,
+      "loss": 0.4835,
       "step": 21915
     },
     {
-      "epoch": 0.77,
-      "learning_rate": 4.800466506118436e-05,
-      "loss": 0.3012,
+      "epoch": 0.7899953148088081,
+      "grad_norm": 0.19068776071071625,
+      "learning_rate": 4.78842036030939e-05,
+      "loss": 0.427,
       "step": 21920
     },
     {
-      "epoch": 0.77,
-      "learning_rate": 4.800354969432141e-05,
-      "loss": 0.2974,
+      "epoch": 0.7901755144700328,
+      "grad_norm": 0.1702924221754074,
+      "learning_rate": 4.7883028554403554e-05,
+      "loss": 0.4455,
       "step": 21925
     },
     {
-      "epoch": 0.77,
-      "learning_rate": 4.800243402877252e-05,
-      "loss": 0.2985,
+      "epoch": 0.7903557141312574,
+      "grad_norm": 0.14877758920192719,
+      "learning_rate": 4.788185319393637e-05,
+      "loss": 0.4352,
       "step": 21930
     },
     {
-      "epoch": 0.77,
-      "learning_rate": 4.8001318064552184e-05,
-      "loss": 0.2776,
+      "epoch": 0.7905359137924821,
+      "grad_norm": 0.1357695311307907,
+      "learning_rate": 4.788067752170837e-05,
+      "loss": 0.4631,
       "step": 21935
     },
     {
-      "epoch": 0.77,
-      "learning_rate": 4.800020180167489e-05,
-      "loss": 0.2897,
+      "epoch": 0.7907161134537067,
+      "grad_norm": 0.194273442029953,
+      "learning_rate": 4.787950153773557e-05,
+      "loss": 0.4193,
       "step": 21940
     },
     {
-      "epoch": 0.77,
-      "learning_rate": 4.7999085240155137e-05,
-      "loss": 0.2792,
+      "epoch": 0.7908963131149314,
+      "grad_norm": 0.15530972182750702,
+      "learning_rate": 4.7878325242033987e-05,
+      "loss": 0.4188,
       "step": 21945
     },
     {
-      "epoch": 0.77,
-      "learning_rate": 4.7997968380007416e-05,
-      "loss": 0.3092,
+      "epoch": 0.791076512776156,
+      "grad_norm": 0.14824257791042328,
+      "learning_rate": 4.7877148634619657e-05,
+      "loss": 0.4546,
       "step": 21950
     },
     {
-      "epoch": 0.77,
-      "learning_rate": 4.799685122124624e-05,
-      "loss": 0.3157,
+      "epoch": 0.7912567124373806,
+      "grad_norm": 0.19110989570617676,
+      "learning_rate": 4.7875971715508606e-05,
+      "loss": 0.426,
       "step": 21955
     },
     {
-      "epoch": 0.77,
-      "learning_rate": 4.7995733763886095e-05,
-      "loss": 0.3111,
+      "epoch": 0.7914369120986052,
+      "grad_norm": 0.1911250203847885,
+      "learning_rate": 4.787479448471686e-05,
+      "loss": 0.4591,
       "step": 21960
     },
     {
-      "epoch": 0.77,
-      "learning_rate": 4.79946160079415e-05,
-      "loss": 0.2896,
+      "epoch": 0.7916171117598299,
+      "grad_norm": 0.16899849474430084,
+      "learning_rate": 4.787361694226048e-05,
+      "loss": 0.4502,
       "step": 21965
     },
     {
-      "epoch": 0.77,
-      "learning_rate": 4.7993497953426986e-05,
-      "loss": 0.3364,
+      "epoch": 0.7917973114210546,
+      "grad_norm": 0.15524151921272278,
+      "learning_rate": 4.787243908815548e-05,
+      "loss": 0.4257,
       "step": 21970
     },
     {
-      "epoch": 0.77,
-      "learning_rate": 4.799237960035704e-05,
-      "loss": 0.3244,
+      "epoch": 0.7919775110822792,
+      "grad_norm": 0.16246455907821655,
+      "learning_rate": 4.787126092241795e-05,
+      "loss": 0.4214,
       "step": 21975
     },
     {
-      "epoch": 0.77,
-      "learning_rate": 4.799126094874621e-05,
-      "loss": 0.3001,
+      "epoch": 0.7921577107435038,
+      "grad_norm": 0.1904502511024475,
+      "learning_rate": 4.78700824450639e-05,
+      "loss": 0.4643,
       "step": 21980
     },
     {
-      "epoch": 0.77,
-      "learning_rate": 4.7990141998609e-05,
-      "loss": 0.3243,
+      "epoch": 0.7923379104047285,
+      "grad_norm": 0.17854608595371246,
+      "learning_rate": 4.786890365610941e-05,
+      "loss": 0.4244,
       "step": 21985
     },
     {
-      "epoch": 0.77,
-      "learning_rate": 4.798902274995995e-05,
-      "loss": 0.3208,
+      "epoch": 0.7925181100659531,
+      "grad_norm": 0.1559765338897705,
+      "learning_rate": 4.786772455557054e-05,
+      "loss": 0.4281,
       "step": 21990
     },
     {
-      "epoch": 0.77,
-      "learning_rate": 4.7987903202813586e-05,
-      "loss": 0.318,
+      "epoch": 0.7926983097271777,
+      "grad_norm": 0.16417820751667023,
+      "learning_rate": 4.786654514346335e-05,
+      "loss": 0.4106,
       "step": 21995
     },
     {
-      "epoch": 0.77,
-      "learning_rate": 4.798678335718445e-05,
-      "loss": 0.3,
+      "epoch": 0.7928785093884023,
+      "grad_norm": 0.20214970409870148,
+      "learning_rate": 4.7865365419803896e-05,
+      "loss": 0.4704,
       "step": 22000
     },
     {
-      "epoch": 0.77,
-      "eval_loss": 0.30042049288749695,
-      "eval_runtime": 10.5417,
-      "eval_samples_per_second": 9.486,
-      "eval_steps_per_second": 9.486,
+      "epoch": 0.7928785093884023,
+      "eval_loss": 0.4604956805706024,
+      "eval_runtime": 3.5345,
+      "eval_samples_per_second": 28.293,
+      "eval_steps_per_second": 7.073,
       "step": 22000
     },
     {
-      "epoch": 0.77,
-      "learning_rate": 4.7985663213087085e-05,
-      "loss": 0.3042,
+      "epoch": 0.7930587090496269,
+      "grad_norm": 0.14078626036643982,
+      "learning_rate": 4.786418538460828e-05,
+      "loss": 0.4562,
       "step": 22005
     },
     {
-      "epoch": 0.77,
-      "learning_rate": 4.798454277053602e-05,
-      "loss": 0.3175,
+      "epoch": 0.7932389087108517,
+      "grad_norm": 0.15193304419517517,
+      "learning_rate": 4.7863005037892554e-05,
+      "loss": 0.482,
       "step": 22010
     },
     {
-      "epoch": 0.77,
-      "learning_rate": 4.798342202954583e-05,
-      "loss": 0.3416,
+      "epoch": 0.7934191083720763,
+      "grad_norm": 0.1598198413848877,
+      "learning_rate": 4.786182437967282e-05,
+      "loss": 0.4144,
       "step": 22015
     },
     {
-      "epoch": 0.77,
-      "learning_rate": 4.798230099013104e-05,
-      "loss": 0.3022,
+      "epoch": 0.7935993080333009,
+      "grad_norm": 0.1988956481218338,
+      "learning_rate": 4.786064340996515e-05,
+      "loss": 0.4673,
       "step": 22020
     },
     {
-      "epoch": 0.77,
-      "learning_rate": 4.7981179652306215e-05,
-      "loss": 0.302,
+      "epoch": 0.7937795076945255,
+      "grad_norm": 0.19150066375732422,
+      "learning_rate": 4.7859462128785635e-05,
+      "loss": 0.4278,
       "step": 22025
     },
     {
-      "epoch": 0.78,
-      "learning_rate": 4.7980058016085925e-05,
-      "loss": 0.29,
+      "epoch": 0.7939597073557502,
+      "grad_norm": 0.18495355546474457,
+      "learning_rate": 4.785828053615038e-05,
+      "loss": 0.4411,
       "step": 22030
     },
     {
-      "epoch": 0.78,
-      "learning_rate": 4.797893608148473e-05,
-      "loss": 0.3239,
+      "epoch": 0.7941399070169748,
+      "grad_norm": 0.1614546775817871,
+      "learning_rate": 4.785709863207548e-05,
+      "loss": 0.3946,
       "step": 22035
     },
     {
-      "epoch": 0.78,
-      "learning_rate": 4.797781384851719e-05,
-      "loss": 0.3278,
+      "epoch": 0.7943201066781994,
+      "grad_norm": 0.1417195051908493,
+      "learning_rate": 4.785591641657704e-05,
+      "loss": 0.4045,
       "step": 22040
     },
     {
-      "epoch": 0.78,
-      "learning_rate": 4.7976691317197874e-05,
-      "loss": 0.3006,
+      "epoch": 0.794500306339424,
+      "grad_norm": 0.16670504212379456,
+      "learning_rate": 4.7854733889671154e-05,
+      "loss": 0.4343,
       "step": 22045
     },
     {
-      "epoch": 0.78,
-      "learning_rate": 4.797556848754137e-05,
-      "loss": 0.3105,
+      "epoch": 0.7946805060006488,
+      "grad_norm": 0.17932060360908508,
+      "learning_rate": 4.785355105137395e-05,
+      "loss": 0.4841,
       "step": 22050
     },
     {
-      "epoch": 0.78,
-      "learning_rate": 4.7974445359562256e-05,
-      "loss": 0.3179,
+      "epoch": 0.7948607056618734,
+      "grad_norm": 0.1873452365398407,
+      "learning_rate": 4.785236790170153e-05,
+      "loss": 0.432,
       "step": 22055
     },
     {
-      "epoch": 0.78,
-      "learning_rate": 4.797332193327509e-05,
-      "loss": 0.3234,
+      "epoch": 0.795040905323098,
+      "grad_norm": 0.16498543322086334,
+      "learning_rate": 4.7851184440670026e-05,
+      "loss": 0.4354,
       "step": 22060
     },
     {
-      "epoch": 0.78,
-      "learning_rate": 4.7972198208694494e-05,
-      "loss": 0.3119,
+      "epoch": 0.7952211049843226,
+      "grad_norm": 0.15727977454662323,
+      "learning_rate": 4.785000066829556e-05,
+      "loss": 0.4322,
       "step": 22065
     },
     {
-      "epoch": 0.78,
-      "learning_rate": 4.7971074185835045e-05,
-      "loss": 0.3022,
+      "epoch": 0.7954013046455473,
+      "grad_norm": 0.17471718788146973,
+      "learning_rate": 4.784881658459426e-05,
+      "loss": 0.422,
       "step": 22070
     },
     {
-      "epoch": 0.78,
-      "learning_rate": 4.7969949864711335e-05,
-      "loss": 0.2895,
+      "epoch": 0.7955815043067719,
+      "grad_norm": 0.19372014701366425,
+      "learning_rate": 4.784763218958226e-05,
+      "loss": 0.4249,
       "step": 22075
     },
     {
-      "epoch": 0.78,
-      "learning_rate": 4.796882524533796e-05,
-      "loss": 0.344,
+      "epoch": 0.7957617039679965,
+      "grad_norm": 0.1373492181301117,
+      "learning_rate": 4.784644748327568e-05,
+      "loss": 0.4094,
       "step": 22080
     },
     {
-      "epoch": 0.78,
-      "learning_rate": 4.796770032772953e-05,
-      "loss": 0.3162,
+      "epoch": 0.7959419036292211,
+      "grad_norm": 0.16347211599349976,
+      "learning_rate": 4.7845262465690695e-05,
+      "loss": 0.4199,
       "step": 22085
     },
     {
-      "epoch": 0.78,
-      "learning_rate": 4.796657511190064e-05,
-      "loss": 0.293,
+      "epoch": 0.7961221032904459,
+      "grad_norm": 0.16035209596157074,
+      "learning_rate": 4.7844077136843426e-05,
+      "loss": 0.442,
       "step": 22090
     },
     {
-      "epoch": 0.78,
-      "learning_rate": 4.7965449597865914e-05,
-      "loss": 0.3211,
+      "epoch": 0.7963023029516705,
+      "grad_norm": 0.19094820320606232,
+      "learning_rate": 4.784289149675002e-05,
+      "loss": 0.4381,
       "step": 22095
     },
     {
-      "epoch": 0.78,
-      "learning_rate": 4.796432378563996e-05,
-      "loss": 0.3143,
+      "epoch": 0.7964825026128951,
+      "grad_norm": 0.17128083109855652,
+      "learning_rate": 4.784170554542665e-05,
+      "loss": 0.4571,
       "step": 22100
     },
     {
-      "epoch": 0.78,
-      "learning_rate": 4.796319767523739e-05,
-      "loss": 0.3107,
+      "epoch": 0.7966627022741197,
+      "grad_norm": 0.16196797788143158,
+      "learning_rate": 4.784051928288946e-05,
+      "loss": 0.4337,
       "step": 22105
     },
     {
-      "epoch": 0.78,
-      "learning_rate": 4.796207126667284e-05,
-      "loss": 0.2938,
+      "epoch": 0.7968429019353443,
+      "grad_norm": 0.16744011640548706,
+      "learning_rate": 4.7839332709154613e-05,
+      "loss": 0.4011,
       "step": 22110
     },
     {
-      "epoch": 0.78,
-      "learning_rate": 4.7960944559960915e-05,
-      "loss": 0.3009,
+      "epoch": 0.797023101596569,
+      "grad_norm": 0.1657259315252304,
+      "learning_rate": 4.783814582423829e-05,
+      "loss": 0.4232,
       "step": 22115
     },
     {
-      "epoch": 0.78,
-      "learning_rate": 4.795981755511627e-05,
-      "loss": 0.3159,
+      "epoch": 0.7972033012577936,
+      "grad_norm": 0.17152729630470276,
+      "learning_rate": 4.783695862815664e-05,
+      "loss": 0.4448,
       "step": 22120
     },
     {
-      "epoch": 0.78,
-      "learning_rate": 4.7958690252153514e-05,
-      "loss": 0.3194,
+      "epoch": 0.7973835009190182,
+      "grad_norm": 0.17332147061824799,
+      "learning_rate": 4.783577112092585e-05,
+      "loss": 0.4088,
       "step": 22125
     },
     {
-      "epoch": 0.78,
-      "learning_rate": 4.7957562651087304e-05,
-      "loss": 0.3359,
+      "epoch": 0.797563700580243,
+      "grad_norm": 0.1492212414741516,
+      "learning_rate": 4.783458330256211e-05,
+      "loss": 0.4846,
       "step": 22130
     },
     {
-      "epoch": 0.78,
-      "learning_rate": 4.795643475193227e-05,
-      "loss": 0.3258,
+      "epoch": 0.7977439002414676,
+      "grad_norm": 0.21426644921302795,
+      "learning_rate": 4.783339517308159e-05,
+      "loss": 0.3948,
       "step": 22135
     },
     {
-      "epoch": 0.78,
-      "learning_rate": 4.795530655470306e-05,
-      "loss": 0.304,
+      "epoch": 0.7979240999026922,
+      "grad_norm": 0.18977008759975433,
+      "learning_rate": 4.783220673250048e-05,
+      "loss": 0.4355,
       "step": 22140
     },
     {
-      "epoch": 0.78,
-      "learning_rate": 4.795417805941432e-05,
-      "loss": 0.2929,
+      "epoch": 0.7981042995639168,
+      "grad_norm": 0.1697508692741394,
+      "learning_rate": 4.783101798083498e-05,
+      "loss": 0.3963,
       "step": 22145
     },
     {
-      "epoch": 0.78,
-      "learning_rate": 4.795304926608071e-05,
-      "loss": 0.3054,
+      "epoch": 0.7982844992251414,
+      "grad_norm": 0.1844054013490677,
+      "learning_rate": 4.782982891810127e-05,
+      "loss": 0.4502,
       "step": 22150
     },
     {
-      "epoch": 0.78,
-      "learning_rate": 4.795192017471688e-05,
-      "loss": 0.3056,
+      "epoch": 0.7984646988863661,
+      "grad_norm": 0.19432871043682098,
+      "learning_rate": 4.782863954431557e-05,
+      "loss": 0.4602,
       "step": 22155
     },
     {
-      "epoch": 0.78,
-      "learning_rate": 4.795079078533749e-05,
-      "loss": 0.2981,
+      "epoch": 0.7986448985475907,
+      "grad_norm": 0.16126291453838348,
+      "learning_rate": 4.7827449859494065e-05,
+      "loss": 0.4409,
       "step": 22160
     },
     {
-      "epoch": 0.78,
-      "learning_rate": 4.7949661097957205e-05,
-      "loss": 0.2938,
+      "epoch": 0.7988250982088154,
+      "grad_norm": 0.17771978676319122,
+      "learning_rate": 4.7826259863652975e-05,
+      "loss": 0.4657,
       "step": 22165
     },
     {
-      "epoch": 0.78,
-      "learning_rate": 4.79485311125907e-05,
-      "loss": 0.2924,
+      "epoch": 0.79900529787004,
+      "grad_norm": 0.18326117098331451,
+      "learning_rate": 4.7825069556808525e-05,
+      "loss": 0.4722,
       "step": 22170
     },
     {
-      "epoch": 0.78,
-      "learning_rate": 4.7947400829252636e-05,
-      "loss": 0.3178,
+      "epoch": 0.7991854975312647,
+      "grad_norm": 0.14180731773376465,
+      "learning_rate": 4.782387893897692e-05,
+      "loss": 0.4298,
       "step": 22175
     },
     {
-      "epoch": 0.78,
-      "learning_rate": 4.7946270247957704e-05,
-      "loss": 0.3124,
+      "epoch": 0.7993656971924893,
+      "grad_norm": 0.1825062483549118,
+      "learning_rate": 4.7822688010174376e-05,
+      "loss": 0.4577,
       "step": 22180
     },
     {
-      "epoch": 0.78,
-      "learning_rate": 4.7945139368720573e-05,
-      "loss": 0.3451,
+      "epoch": 0.7995458968537139,
+      "grad_norm": 0.19557367265224457,
+      "learning_rate": 4.782149677041713e-05,
+      "loss": 0.4641,
       "step": 22185
     },
     {
-      "epoch": 0.78,
-      "learning_rate": 4.794400819155592e-05,
-      "loss": 0.2928,
+      "epoch": 0.7997260965149385,
+      "grad_norm": 0.17543824017047882,
+      "learning_rate": 4.782030521972141e-05,
+      "loss": 0.4703,
       "step": 22190
     },
     {
-      "epoch": 0.78,
-      "learning_rate": 4.7942876716478444e-05,
-      "loss": 0.3024,
+      "epoch": 0.7999062961761632,
+      "grad_norm": 0.15802043676376343,
+      "learning_rate": 4.781911335810345e-05,
+      "loss": 0.4311,
       "step": 22195
     },
     {
-      "epoch": 0.78,
-      "learning_rate": 4.794174494350284e-05,
-      "loss": 0.3331,
+      "epoch": 0.8000864958373878,
+      "grad_norm": 0.14436450600624084,
+      "learning_rate": 4.781792118557948e-05,
+      "loss": 0.4534,
       "step": 22200
     },
     {
-      "epoch": 0.78,
-      "learning_rate": 4.79406128726438e-05,
-      "loss": 0.2988,
+      "epoch": 0.8002666954986125,
+      "grad_norm": 0.14026832580566406,
+      "learning_rate": 4.7816728702165765e-05,
+      "loss": 0.3809,
       "step": 22205
     },
     {
-      "epoch": 0.78,
-      "learning_rate": 4.793948050391601e-05,
-      "loss": 0.2974,
+      "epoch": 0.8004468951598371,
+      "grad_norm": 0.1877908557653427,
+      "learning_rate": 4.781553590787853e-05,
+      "loss": 0.4493,
       "step": 22210
     },
     {
-      "epoch": 0.78,
-      "learning_rate": 4.7938347837334186e-05,
-      "loss": 0.2996,
+      "epoch": 0.8006270948210618,
+      "grad_norm": 0.18697671592235565,
+      "learning_rate": 4.7814342802734034e-05,
+      "loss": 0.4576,
       "step": 22215
     },
     {
-      "epoch": 0.78,
-      "learning_rate": 4.793721487291303e-05,
-      "loss": 0.3041,
+      "epoch": 0.8008072944822864,
+      "grad_norm": 0.1507420688867569,
+      "learning_rate": 4.781314938674855e-05,
+      "loss": 0.472,
       "step": 22220
     },
     {
-      "epoch": 0.78,
-      "learning_rate": 4.793608161066726e-05,
-      "loss": 0.2966,
+      "epoch": 0.800987494143511,
+      "grad_norm": 0.11861146241426468,
+      "learning_rate": 4.78119556599383e-05,
+      "loss": 0.4527,
       "step": 22225
     },
     {
-      "epoch": 0.78,
-      "learning_rate": 4.7934948050611586e-05,
-      "loss": 0.3155,
+      "epoch": 0.8011676938047356,
+      "grad_norm": 0.24812163412570953,
+      "learning_rate": 4.781076162231959e-05,
+      "loss": 0.4276,
       "step": 22230
     },
     {
-      "epoch": 0.78,
-      "learning_rate": 4.793381419276073e-05,
-      "loss": 0.3225,
+      "epoch": 0.8013478934659602,
+      "grad_norm": 0.13593071699142456,
+      "learning_rate": 4.7809567273908656e-05,
+      "loss": 0.4523,
       "step": 22235
     },
     {
-      "epoch": 0.78,
-      "learning_rate": 4.79326800371294e-05,
-      "loss": 0.2849,
+      "epoch": 0.8015280931271849,
+      "grad_norm": 0.16076035797595978,
+      "learning_rate": 4.7808372614721786e-05,
+      "loss": 0.401,
       "step": 22240
     },
     {
-      "epoch": 0.78,
-      "learning_rate": 4.793154558373234e-05,
-      "loss": 0.2952,
+      "epoch": 0.8017082927884096,
+      "grad_norm": 0.2315647453069687,
+      "learning_rate": 4.780717764477526e-05,
+      "loss": 0.3886,
       "step": 22245
     },
     {
-      "epoch": 0.78,
-      "learning_rate": 4.793041083258427e-05,
-      "loss": 0.296,
+      "epoch": 0.8018884924496342,
+      "grad_norm": 0.15930241346359253,
+      "learning_rate": 4.780598236408535e-05,
+      "loss": 0.4124,
       "step": 22250
     },
     {
-      "epoch": 0.78,
-      "learning_rate": 4.7929275783699935e-05,
-      "loss": 0.3116,
+      "epoch": 0.8020686921108588,
+      "grad_norm": 0.18420235812664032,
+      "learning_rate": 4.780478677266835e-05,
+      "loss": 0.4208,
       "step": 22255
     },
     {
-      "epoch": 0.78,
-      "learning_rate": 4.7928140437094055e-05,
-      "loss": 0.2974,
+      "epoch": 0.8022488917720835,
+      "grad_norm": 0.17045623064041138,
+      "learning_rate": 4.780359087054054e-05,
+      "loss": 0.4138,
       "step": 22260
     },
     {
-      "epoch": 0.78,
-      "learning_rate": 4.792700479278138e-05,
-      "loss": 0.2694,
+      "epoch": 0.8024290914333081,
+      "grad_norm": 0.165228009223938,
+      "learning_rate": 4.780239465771822e-05,
+      "loss": 0.4348,
       "step": 22265
     },
     {
-      "epoch": 0.78,
-      "learning_rate": 4.7925868850776664e-05,
-      "loss": 0.3148,
+      "epoch": 0.8026092910945327,
+      "grad_norm": 0.20021255314350128,
+      "learning_rate": 4.78011981342177e-05,
+      "loss": 0.4712,
       "step": 22270
     },
     {
-      "epoch": 0.78,
-      "learning_rate": 4.792473261109465e-05,
-      "loss": 0.3181,
+      "epoch": 0.8027894907557573,
+      "grad_norm": 0.18207870423793793,
+      "learning_rate": 4.7800001300055254e-05,
+      "loss": 0.4717,
       "step": 22275
     },
     {
-      "epoch": 0.78,
-      "learning_rate": 4.792359607375009e-05,
-      "loss": 0.2937,
+      "epoch": 0.802969690416982,
+      "grad_norm": 0.16366621851921082,
+      "learning_rate": 4.7798804155247205e-05,
+      "loss": 0.4682,
       "step": 22280
     },
     {
-      "epoch": 0.78,
-      "learning_rate": 4.7922459238757746e-05,
-      "loss": 0.294,
+      "epoch": 0.8031498900782067,
+      "grad_norm": 0.15080103278160095,
+      "learning_rate": 4.7797606699809874e-05,
+      "loss": 0.409,
       "step": 22285
     },
     {
-      "epoch": 0.78,
-      "learning_rate": 4.7921322106132366e-05,
-      "loss": 0.3054,
+      "epoch": 0.8033300897394313,
+      "grad_norm": 0.1815640777349472,
+      "learning_rate": 4.779640893375956e-05,
+      "loss": 0.456,
       "step": 22290
     },
     {
-      "epoch": 0.78,
-      "learning_rate": 4.792018467588873e-05,
-      "loss": 0.3309,
+      "epoch": 0.8035102894006559,
+      "grad_norm": 0.17141170799732208,
+      "learning_rate": 4.7795210857112585e-05,
+      "loss": 0.4179,
       "step": 22295
     },
     {
-      "epoch": 0.78,
-      "learning_rate": 4.7919046948041605e-05,
-      "loss": 0.2885,
+      "epoch": 0.8036904890618806,
+      "grad_norm": 0.1825985610485077,
+      "learning_rate": 4.7794012469885276e-05,
+      "loss": 0.4053,
       "step": 22300
     },
     {
-      "epoch": 0.78,
-      "learning_rate": 4.791790892260575e-05,
-      "loss": 0.3064,
+      "epoch": 0.8038706887231052,
+      "grad_norm": 0.17858922481536865,
+      "learning_rate": 4.779281377209396e-05,
+      "loss": 0.4419,
       "step": 22305
     },
     {
-      "epoch": 0.78,
-      "learning_rate": 4.791677059959595e-05,
-      "loss": 0.3486,
+      "epoch": 0.8040508883843298,
+      "grad_norm": 0.18266747891902924,
+      "learning_rate": 4.779161476375497e-05,
+      "loss": 0.4313,
       "step": 22310
     },
     {
-      "epoch": 0.79,
-      "learning_rate": 4.7915631979027e-05,
-      "loss": 0.3017,
+      "epoch": 0.8042310880455544,
+      "grad_norm": 0.15619586408138275,
+      "learning_rate": 4.7790415444884645e-05,
+      "loss": 0.4456,
       "step": 22315
     },
     {
-      "epoch": 0.79,
-      "learning_rate": 4.7914493060913654e-05,
-      "loss": 0.2882,
+      "epoch": 0.8044112877067792,
+      "grad_norm": 0.14940191805362701,
+      "learning_rate": 4.778921581549932e-05,
+      "loss": 0.4155,
       "step": 22320
     },
     {
-      "epoch": 0.79,
-      "learning_rate": 4.7913353845270725e-05,
-      "loss": 0.3062,
+      "epoch": 0.8045914873680038,
+      "grad_norm": 0.18177048861980438,
+      "learning_rate": 4.778801587561535e-05,
+      "loss": 0.4538,
       "step": 22325
     },
     {
-      "epoch": 0.79,
-      "learning_rate": 4.7912214332113e-05,
-      "loss": 0.3241,
+      "epoch": 0.8047716870292284,
+      "grad_norm": 0.19796302914619446,
+      "learning_rate": 4.778681562524906e-05,
+      "loss": 0.4435,
       "step": 22330
     },
     {
-      "epoch": 0.79,
-      "learning_rate": 4.7911074521455266e-05,
-      "loss": 0.2649,
+      "epoch": 0.804951886690453,
+      "grad_norm": 0.17838706076145172,
+      "learning_rate": 4.778561506441682e-05,
+      "loss": 0.4309,
       "step": 22335
     },
     {
-      "epoch": 0.79,
-      "learning_rate": 4.7909934413312335e-05,
-      "loss": 0.32,
+      "epoch": 0.8051320863516777,
+      "grad_norm": 0.14537879824638367,
+      "learning_rate": 4.7784414193135e-05,
+      "loss": 0.4494,
       "step": 22340
     },
     {
-      "epoch": 0.79,
-      "learning_rate": 4.7908794007698996e-05,
-      "loss": 0.2956,
+      "epoch": 0.8053122860129023,
+      "grad_norm": 0.2153496891260147,
+      "learning_rate": 4.778321301141994e-05,
+      "loss": 0.4376,
       "step": 22345
     },
     {
-      "epoch": 0.79,
-      "learning_rate": 4.7907653304630074e-05,
-      "loss": 0.3181,
+      "epoch": 0.8054924856741269,
+      "grad_norm": 0.17818701267242432,
+      "learning_rate": 4.7782011519288e-05,
+      "loss": 0.4245,
       "step": 22350
     },
     {
-      "epoch": 0.79,
-      "learning_rate": 4.790651230412037e-05,
-      "loss": 0.2962,
+      "epoch": 0.8056726853353515,
+      "grad_norm": 0.15808773040771484,
+      "learning_rate": 4.778080971675558e-05,
+      "loss": 0.42,
       "step": 22355
     },
     {
-      "epoch": 0.79,
-      "learning_rate": 4.790537100618469e-05,
-      "loss": 0.3048,
+      "epoch": 0.8058528849965763,
+      "grad_norm": 0.16724058985710144,
+      "learning_rate": 4.777960760383904e-05,
+      "loss": 0.4454,
       "step": 22360
     },
     {
-      "epoch": 0.79,
-      "learning_rate": 4.790422941083786e-05,
-      "loss": 0.3162,
+      "epoch": 0.8060330846578009,
+      "grad_norm": 0.14002850651741028,
+      "learning_rate": 4.777840518055475e-05,
+      "loss": 0.4447,
       "step": 22365
     },
     {
-      "epoch": 0.79,
-      "learning_rate": 4.790308751809471e-05,
-      "loss": 0.3088,
+      "epoch": 0.8062132843190255,
+      "grad_norm": 0.1951669454574585,
+      "learning_rate": 4.77772024469191e-05,
+      "loss": 0.4589,
       "step": 22370
     },
     {
-      "epoch": 0.79,
-      "learning_rate": 4.790194532797007e-05,
-      "loss": 0.3218,
+      "epoch": 0.8063934839802501,
+      "grad_norm": 0.17650310695171356,
+      "learning_rate": 4.7775999402948476e-05,
+      "loss": 0.3997,
       "step": 22375
     },
     {
-      "epoch": 0.79,
-      "learning_rate": 4.790080284047876e-05,
-      "loss": 0.2781,
+      "epoch": 0.8065736836414747,
+      "grad_norm": 0.17684021592140198,
+      "learning_rate": 4.7774796048659276e-05,
+      "loss": 0.4246,
       "step": 22380
     },
     {
-      "epoch": 0.79,
-      "learning_rate": 4.789966005563561e-05,
-      "loss": 0.2965,
+      "epoch": 0.8067538833026994,
+      "grad_norm": 0.18748098611831665,
+      "learning_rate": 4.7773592384067884e-05,
+      "loss": 0.4371,
       "step": 22385
     },
     {
-      "epoch": 0.79,
-      "learning_rate": 4.7898516973455474e-05,
-      "loss": 0.3253,
+      "epoch": 0.806934082963924,
+      "grad_norm": 0.19001294672489166,
+      "learning_rate": 4.7772388409190704e-05,
+      "loss": 0.4281,
       "step": 22390
     },
     {
-      "epoch": 0.79,
-      "learning_rate": 4.789737359395319e-05,
-      "loss": 0.2921,
+      "epoch": 0.8071142826251486,
+      "grad_norm": 0.1857748180627823,
+      "learning_rate": 4.7771184124044144e-05,
+      "loss": 0.4078,
       "step": 22395
     },
     {
-      "epoch": 0.79,
-      "learning_rate": 4.789622991714359e-05,
-      "loss": 0.2875,
+      "epoch": 0.8072944822863733,
+      "grad_norm": 0.1612103283405304,
+      "learning_rate": 4.776997952864461e-05,
+      "loss": 0.4418,
       "step": 22400
     },
     {
-      "epoch": 0.79,
-      "learning_rate": 4.789508594304154e-05,
-      "loss": 0.2885,
+      "epoch": 0.807474681947598,
+      "grad_norm": 0.1528964787721634,
+      "learning_rate": 4.7768774623008506e-05,
+      "loss": 0.4683,
       "step": 22405
     },
     {
-      "epoch": 0.79,
-      "learning_rate": 4.789394167166188e-05,
-      "loss": 0.3002,
+      "epoch": 0.8076548816088226,
+      "grad_norm": 0.15798471868038177,
+      "learning_rate": 4.776756940715226e-05,
+      "loss": 0.4182,
       "step": 22410
     },
     {
-      "epoch": 0.79,
-      "learning_rate": 4.7892797103019476e-05,
-      "loss": 0.3167,
+      "epoch": 0.8078350812700472,
+      "grad_norm": 0.16511400043964386,
+      "learning_rate": 4.7766363881092294e-05,
+      "loss": 0.4249,
       "step": 22415
     },
     {
-      "epoch": 0.79,
-      "learning_rate": 4.7891652237129194e-05,
-      "loss": 0.3041,
+      "epoch": 0.8080152809312718,
+      "grad_norm": 0.17201100289821625,
+      "learning_rate": 4.776515804484502e-05,
+      "loss": 0.4506,
       "step": 22420
     },
     {
-      "epoch": 0.79,
-      "learning_rate": 4.789050707400588e-05,
-      "loss": 0.3157,
+      "epoch": 0.8081954805924965,
+      "grad_norm": 0.2050049602985382,
+      "learning_rate": 4.776395189842688e-05,
+      "loss": 0.4595,
       "step": 22425
     },
     {
-      "epoch": 0.79,
-      "learning_rate": 4.7889361613664436e-05,
-      "loss": 0.3145,
+      "epoch": 0.8083756802537211,
+      "grad_norm": 0.16215232014656067,
+      "learning_rate": 4.7762745441854296e-05,
+      "loss": 0.4377,
       "step": 22430
     },
     {
-      "epoch": 0.79,
-      "learning_rate": 4.7888215856119714e-05,
-      "loss": 0.2968,
+      "epoch": 0.8085558799149457,
+      "grad_norm": 0.1810922622680664,
+      "learning_rate": 4.776153867514372e-05,
+      "loss": 0.4346,
       "step": 22435
     },
     {
-      "epoch": 0.79,
-      "learning_rate": 4.7887069801386585e-05,
-      "loss": 0.3106,
+      "epoch": 0.8087360795761704,
+      "grad_norm": 0.1510697603225708,
+      "learning_rate": 4.7760331598311584e-05,
+      "loss": 0.4279,
       "step": 22440
     },
     {
-      "epoch": 0.79,
-      "learning_rate": 4.7885923449479936e-05,
-      "loss": 0.3012,
+      "epoch": 0.8089162792373951,
+      "grad_norm": 0.21630224585533142,
+      "learning_rate": 4.7759124211374335e-05,
+      "loss": 0.4654,
       "step": 22445
     },
     {
-      "epoch": 0.79,
-      "learning_rate": 4.788477680041465e-05,
-      "loss": 0.28,
+      "epoch": 0.8090964788986197,
+      "grad_norm": 0.1879904568195343,
+      "learning_rate": 4.775791651434843e-05,
+      "loss": 0.3944,
       "step": 22450
     },
     {
-      "epoch": 0.79,
-      "learning_rate": 4.788362985420564e-05,
-      "loss": 0.2875,
+      "epoch": 0.8092766785598443,
+      "grad_norm": 0.20264166593551636,
+      "learning_rate": 4.7756708507250314e-05,
+      "loss": 0.4777,
       "step": 22455
     },
     {
-      "epoch": 0.79,
-      "learning_rate": 4.788248261086776e-05,
-      "loss": 0.3307,
+      "epoch": 0.8094568782210689,
+      "grad_norm": 0.1782803237438202,
+      "learning_rate": 4.775550019009645e-05,
+      "loss": 0.4346,
       "step": 22460
     },
     {
-      "epoch": 0.79,
-      "learning_rate": 4.788133507041592e-05,
-      "loss": 0.2875,
+      "epoch": 0.8096370778822936,
+      "grad_norm": 0.16567277908325195,
+      "learning_rate": 4.77542915629033e-05,
+      "loss": 0.4615,
       "step": 22465
     },
     {
-      "epoch": 0.79,
-      "learning_rate": 4.788018723286504e-05,
-      "loss": 0.2809,
+      "epoch": 0.8098172775435182,
+      "grad_norm": 0.17552456259727478,
+      "learning_rate": 4.7753082625687334e-05,
+      "loss": 0.4801,
       "step": 22470
     },
     {
-      "epoch": 0.79,
-      "learning_rate": 4.787903909822999e-05,
-      "loss": 0.2954,
+      "epoch": 0.8099974772047429,
+      "grad_norm": 0.21018646657466888,
+      "learning_rate": 4.7751873378465026e-05,
+      "loss": 0.4294,
       "step": 22475
     },
     {
-      "epoch": 0.79,
-      "learning_rate": 4.7877890666525703e-05,
-      "loss": 0.2881,
+      "epoch": 0.8101776768659675,
+      "grad_norm": 0.16765153408050537,
+      "learning_rate": 4.7750663821252844e-05,
+      "loss": 0.4349,
       "step": 22480
     },
     {
-      "epoch": 0.79,
-      "learning_rate": 4.787674193776709e-05,
-      "loss": 0.2969,
+      "epoch": 0.8103578765271922,
+      "grad_norm": 0.18359611928462982,
+      "learning_rate": 4.7749453954067275e-05,
+      "loss": 0.4012,
       "step": 22485
     },
     {
-      "epoch": 0.79,
-      "learning_rate": 4.7875592911969045e-05,
-      "loss": 0.3009,
+      "epoch": 0.8105380761884168,
+      "grad_norm": 0.19209226965904236,
+      "learning_rate": 4.774824377692479e-05,
+      "loss": 0.4422,
       "step": 22490
     },
     {
-      "epoch": 0.79,
-      "learning_rate": 4.787444358914651e-05,
-      "loss": 0.3125,
+      "epoch": 0.8107182758496414,
+      "grad_norm": 0.18595997989177704,
+      "learning_rate": 4.77470332898419e-05,
+      "loss": 0.4376,
       "step": 22495
     },
     {
-      "epoch": 0.79,
-      "learning_rate": 4.78732939693144e-05,
-      "loss": 0.3025,
+      "epoch": 0.810898475510866,
+      "grad_norm": 0.12643958628177643,
+      "learning_rate": 4.774582249283509e-05,
+      "loss": 0.3845,
       "step": 22500
     },
     {
-      "epoch": 0.79,
-      "eval_loss": 0.2996106445789337,
-      "eval_runtime": 10.5303,
-      "eval_samples_per_second": 9.496,
-      "eval_steps_per_second": 9.496,
+      "epoch": 0.810898475510866,
+      "eval_loss": 0.46091198921203613,
+      "eval_runtime": 3.5434,
+      "eval_samples_per_second": 28.221,
+      "eval_steps_per_second": 7.055,
       "step": 22500
     },
     {
-      "epoch": 0.79,
-      "learning_rate": 4.7872144052487636e-05,
-      "loss": 0.2975,
+      "epoch": 0.8110786751720906,
+      "grad_norm": 0.18900783360004425,
+      "learning_rate": 4.774461138592085e-05,
+      "loss": 0.4837,
       "step": 22505
     },
     {
-      "epoch": 0.79,
-      "learning_rate": 4.7870993838681166e-05,
-      "loss": 0.3188,
+      "epoch": 0.8112588748333153,
+      "grad_norm": 0.1918240487575531,
+      "learning_rate": 4.774339996911567e-05,
+      "loss": 0.4452,
       "step": 22510
     },
     {
-      "epoch": 0.79,
-      "learning_rate": 4.7869843327909904e-05,
-      "loss": 0.3126,
+      "epoch": 0.81143907449454,
+      "grad_norm": 0.18408475816249847,
+      "learning_rate": 4.7742188242436075e-05,
+      "loss": 0.4227,
       "step": 22515
     },
     {
-      "epoch": 0.79,
-      "learning_rate": 4.78686925201888e-05,
-      "loss": 0.3167,
+      "epoch": 0.8116192741557646,
+      "grad_norm": 0.1483888477087021,
+      "learning_rate": 4.774097620589857e-05,
+      "loss": 0.465,
       "step": 22520
     },
     {
-      "epoch": 0.79,
-      "learning_rate": 4.7867541415532796e-05,
-      "loss": 0.3221,
+      "epoch": 0.8117994738169892,
+      "grad_norm": 0.1650637537240982,
+      "learning_rate": 4.773976385951967e-05,
+      "loss": 0.4325,
       "step": 22525
     },
     {
-      "epoch": 0.79,
-      "learning_rate": 4.786639001395684e-05,
-      "loss": 0.2836,
+      "epoch": 0.8119796734782139,
+      "grad_norm": 0.19324971735477448,
+      "learning_rate": 4.773855120331588e-05,
+      "loss": 0.4326,
       "step": 22530
     },
     {
-      "epoch": 0.79,
-      "learning_rate": 4.786523831547587e-05,
-      "loss": 0.2995,
+      "epoch": 0.8121598731394385,
+      "grad_norm": 0.16592490673065186,
+      "learning_rate": 4.773733823730374e-05,
+      "loss": 0.4163,
       "step": 22535
     },
     {
-      "epoch": 0.79,
-      "learning_rate": 4.7864086320104853e-05,
-      "loss": 0.3091,
+      "epoch": 0.8123400728006631,
+      "grad_norm": 0.17151491343975067,
+      "learning_rate": 4.773612496149977e-05,
+      "loss": 0.4414,
       "step": 22540
     },
     {
-      "epoch": 0.79,
-      "learning_rate": 4.786293402785875e-05,
-      "loss": 0.2919,
+      "epoch": 0.8125202724618877,
+      "grad_norm": 0.1465132236480713,
+      "learning_rate": 4.77349113759205e-05,
+      "loss": 0.4609,
       "step": 22545
     },
     {
-      "epoch": 0.79,
-      "learning_rate": 4.786178143875251e-05,
-      "loss": 0.2988,
+      "epoch": 0.8127004721231124,
+      "grad_norm": 0.1558566689491272,
+      "learning_rate": 4.7733697480582464e-05,
+      "loss": 0.4365,
       "step": 22550
     },
     {
-      "epoch": 0.79,
-      "learning_rate": 4.786062855280111e-05,
-      "loss": 0.3053,
+      "epoch": 0.8128806717843371,
+      "grad_norm": 0.16560520231723785,
+      "learning_rate": 4.7732483275502194e-05,
+      "loss": 0.438,
       "step": 22555
     },
     {
-      "epoch": 0.79,
-      "learning_rate": 4.78594753700195e-05,
-      "loss": 0.3258,
+      "epoch": 0.8130608714455617,
+      "grad_norm": 0.14548031985759735,
+      "learning_rate": 4.773126876069625e-05,
+      "loss": 0.4234,
       "step": 22560
     },
     {
-      "epoch": 0.79,
-      "learning_rate": 4.785832189042268e-05,
-      "loss": 0.303,
+      "epoch": 0.8132410711067863,
+      "grad_norm": 0.16290007531642914,
+      "learning_rate": 4.773005393618116e-05,
+      "loss": 0.407,
       "step": 22565
     },
     {
-      "epoch": 0.79,
-      "learning_rate": 4.785716811402561e-05,
-      "loss": 0.2957,
+      "epoch": 0.813421270768011,
+      "grad_norm": 0.16611815989017487,
+      "learning_rate": 4.7728838801973485e-05,
+      "loss": 0.3838,
       "step": 22570
     },
     {
-      "epoch": 0.79,
-      "learning_rate": 4.785601404084328e-05,
-      "loss": 0.3172,
+      "epoch": 0.8136014704292356,
+      "grad_norm": 0.1640740931034088,
+      "learning_rate": 4.772762335808979e-05,
+      "loss": 0.4413,
       "step": 22575
     },
     {
-      "epoch": 0.79,
-      "learning_rate": 4.785485967089066e-05,
-      "loss": 0.3203,
+      "epoch": 0.8137816700904602,
+      "grad_norm": 0.1402037888765335,
+      "learning_rate": 4.772640760454663e-05,
+      "loss": 0.4131,
       "step": 22580
     },
     {
-      "epoch": 0.79,
-      "learning_rate": 4.7853705004182756e-05,
-      "loss": 0.3185,
+      "epoch": 0.8139618697516848,
+      "grad_norm": 0.15909187495708466,
+      "learning_rate": 4.772519154136056e-05,
+      "loss": 0.4276,
       "step": 22585
     },
     {
-      "epoch": 0.79,
-      "learning_rate": 4.785255004073455e-05,
-      "loss": 0.3293,
+      "epoch": 0.8141420694129095,
+      "grad_norm": 0.16949698328971863,
+      "learning_rate": 4.772397516854815e-05,
+      "loss": 0.4531,
       "step": 22590
     },
     {
-      "epoch": 0.79,
-      "learning_rate": 4.785139478056105e-05,
-      "loss": 0.2998,
+      "epoch": 0.8143222690741342,
+      "grad_norm": 0.19324524700641632,
+      "learning_rate": 4.7722758486125986e-05,
+      "loss": 0.4291,
       "step": 22595
     },
     {
-      "epoch": 0.8,
-      "learning_rate": 4.785023922367725e-05,
-      "loss": 0.3287,
+      "epoch": 0.8145024687353588,
+      "grad_norm": 0.17862281203269958,
+      "learning_rate": 4.772154149411063e-05,
+      "loss": 0.469,
       "step": 22600
     },
     {
-      "epoch": 0.8,
-      "learning_rate": 4.784908337009815e-05,
-      "loss": 0.2895,
+      "epoch": 0.8146826683965834,
+      "grad_norm": 0.19284473359584808,
+      "learning_rate": 4.772032419251868e-05,
+      "loss": 0.4235,
       "step": 22605
     },
     {
-      "epoch": 0.8,
-      "learning_rate": 4.784792721983875e-05,
-      "loss": 0.3014,
+      "epoch": 0.814862868057808,
+      "grad_norm": 0.14753711223602295,
+      "learning_rate": 4.77191065813667e-05,
+      "loss": 0.4528,
       "step": 22610
     },
     {
-      "epoch": 0.8,
-      "learning_rate": 4.784677077291408e-05,
-      "loss": 0.3052,
+      "epoch": 0.8150430677190327,
+      "grad_norm": 0.1892748773097992,
+      "learning_rate": 4.7717888660671306e-05,
+      "loss": 0.4374,
       "step": 22615
     },
     {
-      "epoch": 0.8,
-      "learning_rate": 4.784561402933915e-05,
-      "loss": 0.2981,
+      "epoch": 0.8152232673802573,
+      "grad_norm": 0.15627089142799377,
+      "learning_rate": 4.771667043044906e-05,
+      "loss": 0.4639,
       "step": 22620
     },
     {
-      "epoch": 0.8,
-      "learning_rate": 4.784445698912898e-05,
-      "loss": 0.2791,
+      "epoch": 0.8154034670414819,
+      "grad_norm": 0.1614178568124771,
+      "learning_rate": 4.771545189071659e-05,
+      "loss": 0.449,
       "step": 22625
     },
     {
-      "epoch": 0.8,
-      "learning_rate": 4.784329965229858e-05,
-      "loss": 0.3382,
+      "epoch": 0.8155836667027065,
+      "grad_norm": 0.18917955458164215,
+      "learning_rate": 4.771423304149049e-05,
+      "loss": 0.4695,
       "step": 22630
     },
     {
-      "epoch": 0.8,
-      "learning_rate": 4.7842142018863e-05,
-      "loss": 0.3277,
+      "epoch": 0.8157638663639313,
+      "grad_norm": 0.17164795100688934,
+      "learning_rate": 4.771301388278735e-05,
+      "loss": 0.4405,
       "step": 22635
     },
     {
-      "epoch": 0.8,
-      "learning_rate": 4.784098408883725e-05,
-      "loss": 0.2887,
+      "epoch": 0.8159440660251559,
+      "grad_norm": 0.18476401269435883,
+      "learning_rate": 4.7711794414623796e-05,
+      "loss": 0.4584,
       "step": 22640
     },
     {
-      "epoch": 0.8,
-      "learning_rate": 4.783982586223638e-05,
-      "loss": 0.3109,
+      "epoch": 0.8161242656863805,
+      "grad_norm": 0.1459241509437561,
+      "learning_rate": 4.771057463701644e-05,
+      "loss": 0.4258,
       "step": 22645
     },
     {
-      "epoch": 0.8,
-      "learning_rate": 4.783866733907543e-05,
-      "loss": 0.2923,
+      "epoch": 0.8163044653476051,
+      "grad_norm": 0.16828380525112152,
+      "learning_rate": 4.77093545499819e-05,
+      "loss": 0.4332,
       "step": 22650
     },
     {
-      "epoch": 0.8,
-      "learning_rate": 4.7837508519369425e-05,
-      "loss": 0.3152,
+      "epoch": 0.8164846650088298,
+      "grad_norm": 0.14585858583450317,
+      "learning_rate": 4.770813415353681e-05,
+      "loss": 0.4004,
       "step": 22655
     },
     {
-      "epoch": 0.8,
-      "learning_rate": 4.783634940313342e-05,
-      "loss": 0.3044,
+      "epoch": 0.8166648646700544,
+      "grad_norm": 0.16096548736095428,
+      "learning_rate": 4.7706913447697785e-05,
+      "loss": 0.4693,
       "step": 22660
     },
     {
-      "epoch": 0.8,
-      "learning_rate": 4.7835189990382475e-05,
-      "loss": 0.3201,
+      "epoch": 0.816845064331279,
+      "grad_norm": 0.19337119162082672,
+      "learning_rate": 4.7705692432481455e-05,
+      "loss": 0.4197,
       "step": 22665
     },
     {
-      "epoch": 0.8,
-      "learning_rate": 4.7834030281131636e-05,
-      "loss": 0.3069,
+      "epoch": 0.8170252639925037,
+      "grad_norm": 0.1468539834022522,
+      "learning_rate": 4.770447110790447e-05,
+      "loss": 0.4135,
       "step": 22670
     },
     {
-      "epoch": 0.8,
-      "learning_rate": 4.7832870275395956e-05,
-      "loss": 0.3071,
+      "epoch": 0.8172054636537284,
+      "grad_norm": 0.13654807209968567,
+      "learning_rate": 4.770324947398346e-05,
+      "loss": 0.4459,
       "step": 22675
     },
     {
-      "epoch": 0.8,
-      "learning_rate": 4.783170997319051e-05,
-      "loss": 0.306,
+      "epoch": 0.817385663314953,
+      "grad_norm": 0.16397765278816223,
+      "learning_rate": 4.770202753073506e-05,
+      "loss": 0.4188,
       "step": 22680
     },
     {
-      "epoch": 0.8,
-      "learning_rate": 4.783054937453035e-05,
-      "loss": 0.3189,
+      "epoch": 0.8175658629761776,
+      "grad_norm": 0.20366117358207703,
+      "learning_rate": 4.770080527817594e-05,
+      "loss": 0.4387,
       "step": 22685
     },
     {
-      "epoch": 0.8,
-      "learning_rate": 4.782938847943056e-05,
-      "loss": 0.3142,
+      "epoch": 0.8177460626374022,
+      "grad_norm": 0.1833733767271042,
+      "learning_rate": 4.7699582716322743e-05,
+      "loss": 0.4509,
       "step": 22690
     },
     {
-      "epoch": 0.8,
-      "learning_rate": 4.7828227287906205e-05,
-      "loss": 0.2958,
+      "epoch": 0.8179262622986269,
+      "grad_norm": 0.18486203253269196,
+      "learning_rate": 4.7698359845192126e-05,
+      "loss": 0.4042,
       "step": 22695
     },
     {
-      "epoch": 0.8,
-      "learning_rate": 4.782706579997236e-05,
-      "loss": 0.3093,
+      "epoch": 0.8181064619598515,
+      "grad_norm": 0.17110316455364227,
+      "learning_rate": 4.769713666480075e-05,
+      "loss": 0.406,
       "step": 22700
     },
     {
-      "epoch": 0.8,
-      "learning_rate": 4.7825904015644106e-05,
-      "loss": 0.3015,
+      "epoch": 0.8182866616210761,
+      "grad_norm": 0.15279895067214966,
+      "learning_rate": 4.769591317516528e-05,
+      "loss": 0.4436,
       "step": 22705
     },
     {
-      "epoch": 0.8,
-      "learning_rate": 4.782474193493654e-05,
-      "loss": 0.3103,
+      "epoch": 0.8184668612823008,
+      "grad_norm": 0.15406733751296997,
+      "learning_rate": 4.769468937630239e-05,
+      "loss": 0.4464,
       "step": 22710
     },
     {
-      "epoch": 0.8,
-      "learning_rate": 4.7823579557864725e-05,
-      "loss": 0.274,
+      "epoch": 0.8186470609435255,
+      "grad_norm": 0.16378438472747803,
+      "learning_rate": 4.769346526822874e-05,
+      "loss": 0.4166,
       "step": 22715
     },
     {
-      "epoch": 0.8,
-      "learning_rate": 4.7822416884443786e-05,
-      "loss": 0.2857,
+      "epoch": 0.8188272606047501,
+      "grad_norm": 0.16179268062114716,
+      "learning_rate": 4.769224085096103e-05,
+      "loss": 0.3881,
       "step": 22720
     },
     {
-      "epoch": 0.8,
-      "learning_rate": 4.78212539146888e-05,
-      "loss": 0.3126,
+      "epoch": 0.8190074602659747,
+      "grad_norm": 0.18712985515594482,
+      "learning_rate": 4.769101612451593e-05,
+      "loss": 0.4326,
       "step": 22725
     },
     {
-      "epoch": 0.8,
-      "learning_rate": 4.782009064861488e-05,
-      "loss": 0.3157,
+      "epoch": 0.8191876599271993,
+      "grad_norm": 0.1865232139825821,
+      "learning_rate": 4.768979108891013e-05,
+      "loss": 0.4499,
       "step": 22730
     },
     {
-      "epoch": 0.8,
-      "learning_rate": 4.7818927086237106e-05,
-      "loss": 0.3064,
+      "epoch": 0.819367859588424,
+      "grad_norm": 0.1458199918270111,
+      "learning_rate": 4.7688565744160315e-05,
+      "loss": 0.4193,
       "step": 22735
     },
     {
-      "epoch": 0.8,
-      "learning_rate": 4.7817763227570614e-05,
-      "loss": 0.3148,
+      "epoch": 0.8195480592496486,
+      "grad_norm": 0.17637084424495697,
+      "learning_rate": 4.768734009028319e-05,
+      "loss": 0.4239,
       "step": 22740
     },
     {
-      "epoch": 0.8,
-      "learning_rate": 4.78165990726305e-05,
-      "loss": 0.2939,
+      "epoch": 0.8197282589108732,
+      "grad_norm": 0.14654532074928284,
+      "learning_rate": 4.768611412729545e-05,
+      "loss": 0.4284,
       "step": 22745
     },
     {
-      "epoch": 0.8,
-      "learning_rate": 4.7815434621431884e-05,
-      "loss": 0.2772,
+      "epoch": 0.8199084585720979,
+      "grad_norm": 0.19782017171382904,
+      "learning_rate": 4.768488785521379e-05,
+      "loss": 0.4591,
       "step": 22750
     },
     {
-      "epoch": 0.8,
-      "learning_rate": 4.7814269873989895e-05,
-      "loss": 0.2954,
+      "epoch": 0.8200886582333226,
+      "grad_norm": 0.126174658536911,
+      "learning_rate": 4.768366127405493e-05,
+      "loss": 0.4302,
       "step": 22755
     },
     {
-      "epoch": 0.8,
-      "learning_rate": 4.7813104830319634e-05,
-      "loss": 0.3113,
+      "epoch": 0.8202688578945472,
+      "grad_norm": 0.17687900364398956,
+      "learning_rate": 4.768243438383557e-05,
+      "loss": 0.446,
       "step": 22760
     },
     {
-      "epoch": 0.8,
-      "learning_rate": 4.781193949043624e-05,
-      "loss": 0.3328,
+      "epoch": 0.8204490575557718,
+      "grad_norm": 0.17083390057086945,
+      "learning_rate": 4.768120718457244e-05,
+      "loss": 0.4627,
       "step": 22765
     },
     {
-      "epoch": 0.8,
-      "learning_rate": 4.781077385435485e-05,
-      "loss": 0.3097,
+      "epoch": 0.8206292572169964,
+      "grad_norm": 0.16993774473667145,
+      "learning_rate": 4.767997967628225e-05,
+      "loss": 0.4264,
       "step": 22770
     },
     {
-      "epoch": 0.8,
-      "learning_rate": 4.780960792209059e-05,
-      "loss": 0.3012,
+      "epoch": 0.820809456878221,
+      "grad_norm": 0.19042931497097015,
+      "learning_rate": 4.767875185898173e-05,
+      "loss": 0.4299,
       "step": 22775
     },
     {
-      "epoch": 0.8,
-      "learning_rate": 4.780844169365861e-05,
-      "loss": 0.3298,
+      "epoch": 0.8209896565394457,
+      "grad_norm": 0.19879989326000214,
+      "learning_rate": 4.767752373268761e-05,
+      "loss": 0.4124,
       "step": 22780
     },
     {
-      "epoch": 0.8,
-      "learning_rate": 4.780727516907405e-05,
-      "loss": 0.2949,
+      "epoch": 0.8211698562006703,
+      "grad_norm": 0.17709743976593018,
+      "learning_rate": 4.7676295297416615e-05,
+      "loss": 0.4415,
       "step": 22785
     },
     {
-      "epoch": 0.8,
-      "learning_rate": 4.780610834835204e-05,
-      "loss": 0.321,
+      "epoch": 0.821350055861895,
+      "grad_norm": 0.1532343327999115,
+      "learning_rate": 4.767506655318549e-05,
+      "loss": 0.3988,
       "step": 22790
     },
     {
-      "epoch": 0.8,
-      "learning_rate": 4.780494123150775e-05,
-      "loss": 0.3075,
+      "epoch": 0.8215302555231196,
+      "grad_norm": 0.20457062125205994,
+      "learning_rate": 4.767383750001097e-05,
+      "loss": 0.4313,
       "step": 22795
     },
     {
-      "epoch": 0.8,
-      "learning_rate": 4.780377381855633e-05,
-      "loss": 0.3277,
+      "epoch": 0.8217104551843443,
+      "grad_norm": 0.1625969558954239,
+      "learning_rate": 4.76726081379098e-05,
+      "loss": 0.4156,
       "step": 22800
     },
     {
-      "epoch": 0.8,
-      "learning_rate": 4.780260610951293e-05,
-      "loss": 0.3016,
+      "epoch": 0.8218906548455689,
+      "grad_norm": 0.17059491574764252,
+      "learning_rate": 4.7671378466898735e-05,
+      "loss": 0.4303,
       "step": 22805
     },
     {
-      "epoch": 0.8,
-      "learning_rate": 4.7801438104392715e-05,
-      "loss": 0.3104,
+      "epoch": 0.8220708545067935,
+      "grad_norm": 0.18344855308532715,
+      "learning_rate": 4.767014848699453e-05,
+      "loss": 0.4356,
       "step": 22810
     },
     {
-      "epoch": 0.8,
-      "learning_rate": 4.7800269803210856e-05,
-      "loss": 0.3077,
+      "epoch": 0.8222510541680181,
+      "grad_norm": 0.21067775785923004,
+      "learning_rate": 4.766891819821394e-05,
+      "loss": 0.4577,
       "step": 22815
     },
     {
-      "epoch": 0.8,
-      "learning_rate": 4.779910120598252e-05,
-      "loss": 0.3111,
+      "epoch": 0.8224312538292428,
+      "grad_norm": 0.1624084860086441,
+      "learning_rate": 4.766768760057374e-05,
+      "loss": 0.441,
       "step": 22820
     },
     {
-      "epoch": 0.8,
-      "learning_rate": 4.779793231272288e-05,
-      "loss": 0.2961,
+      "epoch": 0.8226114534904675,
+      "grad_norm": 0.16547471284866333,
+      "learning_rate": 4.766645669409067e-05,
+      "loss": 0.4524,
       "step": 22825
     },
     {
-      "epoch": 0.8,
-      "learning_rate": 4.779676312344711e-05,
-      "loss": 0.3126,
+      "epoch": 0.8227916531516921,
+      "grad_norm": 0.15320567786693573,
+      "learning_rate": 4.766522547878152e-05,
+      "loss": 0.3913,
       "step": 22830
     },
     {
-      "epoch": 0.8,
-      "learning_rate": 4.77955936381704e-05,
-      "loss": 0.3014,
+      "epoch": 0.8229718528129167,
+      "grad_norm": 0.19396758079528809,
+      "learning_rate": 4.766399395466307e-05,
+      "loss": 0.4019,
       "step": 22835
     },
     {
-      "epoch": 0.8,
-      "learning_rate": 4.779442385690793e-05,
-      "loss": 0.3251,
+      "epoch": 0.8231520524741414,
+      "grad_norm": 0.16974592208862305,
+      "learning_rate": 4.766276212175207e-05,
+      "loss": 0.4173,
       "step": 22840
     },
     {
-      "epoch": 0.8,
-      "learning_rate": 4.779325377967488e-05,
-      "loss": 0.3128,
+      "epoch": 0.823332252135366,
+      "grad_norm": 0.17741717398166656,
+      "learning_rate": 4.766152998006534e-05,
+      "loss": 0.4183,
       "step": 22845
     },
     {
-      "epoch": 0.8,
-      "learning_rate": 4.779208340648646e-05,
-      "loss": 0.3177,
+      "epoch": 0.8235124517965906,
+      "grad_norm": 0.20168620347976685,
+      "learning_rate": 4.766029752961965e-05,
+      "loss": 0.4499,
       "step": 22850
     },
     {
-      "epoch": 0.8,
-      "learning_rate": 4.779091273735785e-05,
-      "loss": 0.324,
+      "epoch": 0.8236926514578152,
+      "grad_norm": 0.1789730191230774,
+      "learning_rate": 4.765906477043179e-05,
+      "loss": 0.4358,
       "step": 22855
     },
     {
-      "epoch": 0.8,
-      "learning_rate": 4.7789741772304265e-05,
-      "loss": 0.2845,
+      "epoch": 0.8238728511190399,
+      "grad_norm": 0.1608288586139679,
+      "learning_rate": 4.765783170251856e-05,
+      "loss": 0.4312,
       "step": 22860
     },
     {
-      "epoch": 0.8,
-      "learning_rate": 4.7788570511340905e-05,
-      "loss": 0.2829,
+      "epoch": 0.8240530507802646,
+      "grad_norm": 0.16104714572429657,
+      "learning_rate": 4.7656598325896755e-05,
+      "loss": 0.4676,
       "step": 22865
     },
     {
-      "epoch": 0.8,
-      "learning_rate": 4.778739895448296e-05,
-      "loss": 0.2705,
+      "epoch": 0.8242332504414892,
+      "grad_norm": 0.17927196621894836,
+      "learning_rate": 4.765536464058319e-05,
+      "loss": 0.4367,
       "step": 22870
     },
     {
-      "epoch": 0.8,
-      "learning_rate": 4.7786227101745666e-05,
-      "loss": 0.3156,
+      "epoch": 0.8244134501027138,
+      "grad_norm": 0.1429985910654068,
+      "learning_rate": 4.7654130646594666e-05,
+      "loss": 0.4697,
       "step": 22875
     },
     {
-      "epoch": 0.8,
-      "learning_rate": 4.778505495314423e-05,
-      "loss": 0.2754,
+      "epoch": 0.8245936497639385,
+      "grad_norm": 0.1574939787387848,
+      "learning_rate": 4.7652896343948e-05,
+      "loss": 0.4621,
       "step": 22880
     },
     {
-      "epoch": 0.81,
-      "learning_rate": 4.7783882508693876e-05,
-      "loss": 0.2873,
+      "epoch": 0.8247738494251631,
+      "grad_norm": 0.15647585690021515,
+      "learning_rate": 4.765166173266001e-05,
+      "loss": 0.4169,
       "step": 22885
     },
     {
-      "epoch": 0.81,
-      "learning_rate": 4.7782709768409815e-05,
-      "loss": 0.3089,
+      "epoch": 0.8249540490863877,
+      "grad_norm": 0.15062108635902405,
+      "learning_rate": 4.7650426812747505e-05,
+      "loss": 0.4242,
       "step": 22890
     },
     {
-      "epoch": 0.81,
-      "learning_rate": 4.778153673230728e-05,
-      "loss": 0.3223,
+      "epoch": 0.8251342487476123,
+      "grad_norm": 0.15938465297222137,
+      "learning_rate": 4.764919158422733e-05,
+      "loss": 0.44,
       "step": 22895
     },
     {
-      "epoch": 0.81,
-      "learning_rate": 4.778036340040152e-05,
-      "loss": 0.3066,
+      "epoch": 0.8253144484088369,
+      "grad_norm": 0.19385161995887756,
+      "learning_rate": 4.76479560471163e-05,
+      "loss": 0.4743,
       "step": 22900
     },
     {
-      "epoch": 0.81,
-      "learning_rate": 4.777918977270773e-05,
-      "loss": 0.2998,
+      "epoch": 0.8254946480700617,
+      "grad_norm": 0.1362161636352539,
+      "learning_rate": 4.764672020143125e-05,
+      "loss": 0.417,
       "step": 22905
     },
     {
-      "epoch": 0.81,
-      "learning_rate": 4.777801584924119e-05,
-      "loss": 0.289,
+      "epoch": 0.8256748477312863,
+      "grad_norm": 0.1929071843624115,
+      "learning_rate": 4.7645484047189025e-05,
+      "loss": 0.4504,
       "step": 22910
     },
     {
-      "epoch": 0.81,
-      "learning_rate": 4.777684163001712e-05,
-      "loss": 0.3045,
+      "epoch": 0.8258550473925109,
+      "grad_norm": 0.17340517044067383,
+      "learning_rate": 4.764424758440647e-05,
+      "loss": 0.4313,
       "step": 22915
     },
     {
-      "epoch": 0.81,
-      "learning_rate": 4.777566711505078e-05,
-      "loss": 0.315,
+      "epoch": 0.8260352470537355,
+      "grad_norm": 0.13786520063877106,
+      "learning_rate": 4.764301081310042e-05,
+      "loss": 0.4284,
       "step": 22920
     },
     {
-      "epoch": 0.81,
-      "learning_rate": 4.7774492304357396e-05,
-      "loss": 0.2783,
+      "epoch": 0.8262154467149602,
+      "grad_norm": 0.22200287878513336,
+      "learning_rate": 4.764177373328773e-05,
+      "loss": 0.4638,
       "step": 22925
     },
     {
-      "epoch": 0.81,
-      "learning_rate": 4.7773317197952256e-05,
-      "loss": 0.3309,
+      "epoch": 0.8263956463761848,
+      "grad_norm": 0.1627381592988968,
+      "learning_rate": 4.764053634498526e-05,
+      "loss": 0.473,
       "step": 22930
     },
     {
-      "epoch": 0.81,
-      "learning_rate": 4.777214179585059e-05,
-      "loss": 0.3209,
+      "epoch": 0.8265758460374094,
+      "grad_norm": 0.18759234249591827,
+      "learning_rate": 4.763929864820986e-05,
+      "loss": 0.4233,
       "step": 22935
     },
     {
-      "epoch": 0.81,
-      "learning_rate": 4.777096609806767e-05,
-      "loss": 0.2981,
+      "epoch": 0.826756045698634,
+      "grad_norm": 0.16739268600940704,
+      "learning_rate": 4.7638060642978405e-05,
+      "loss": 0.4639,
       "step": 22940
     },
     {
-      "epoch": 0.81,
-      "learning_rate": 4.7769790104618764e-05,
-      "loss": 0.3165,
+      "epoch": 0.8269362453598588,
+      "grad_norm": 0.13698381185531616,
+      "learning_rate": 4.763682232930776e-05,
+      "loss": 0.4376,
       "step": 22945
     },
     {
-      "epoch": 0.81,
-      "learning_rate": 4.7768613815519145e-05,
-      "loss": 0.2977,
+      "epoch": 0.8271164450210834,
+      "grad_norm": 0.16585668921470642,
+      "learning_rate": 4.763558370721478e-05,
+      "loss": 0.3934,
       "step": 22950
     },
     {
-      "epoch": 0.81,
-      "learning_rate": 4.7767437230784075e-05,
-      "loss": 0.3124,
+      "epoch": 0.827296644682308,
+      "grad_norm": 0.1562475860118866,
+      "learning_rate": 4.7634344776716364e-05,
+      "loss": 0.4305,
       "step": 22955
     },
     {
-      "epoch": 0.81,
-      "learning_rate": 4.776626035042884e-05,
-      "loss": 0.2972,
+      "epoch": 0.8274768443435326,
+      "grad_norm": 0.18153679370880127,
+      "learning_rate": 4.763310553782938e-05,
+      "loss": 0.4395,
       "step": 22960
     },
     {
-      "epoch": 0.81,
-      "learning_rate": 4.776508317446871e-05,
-      "loss": 0.2934,
+      "epoch": 0.8276570440047573,
+      "grad_norm": 0.21862560510635376,
+      "learning_rate": 4.7631865990570715e-05,
+      "loss": 0.4454,
       "step": 22965
     },
     {
-      "epoch": 0.81,
-      "learning_rate": 4.7763905702918984e-05,
-      "loss": 0.2941,
+      "epoch": 0.8278372436659819,
+      "grad_norm": 0.16682131588459015,
+      "learning_rate": 4.763062613495726e-05,
+      "loss": 0.4646,
       "step": 22970
     },
     {
-      "epoch": 0.81,
-      "learning_rate": 4.776272793579494e-05,
-      "loss": 0.3158,
+      "epoch": 0.8280174433272065,
+      "grad_norm": 0.19076871871948242,
+      "learning_rate": 4.76293859710059e-05,
+      "loss": 0.4251,
       "step": 22975
     },
     {
-      "epoch": 0.81,
-      "learning_rate": 4.776154987311188e-05,
-      "loss": 0.3337,
+      "epoch": 0.8281976429884311,
+      "grad_norm": 0.14495553076267242,
+      "learning_rate": 4.7628145498733543e-05,
+      "loss": 0.4469,
       "step": 22980
     },
     {
-      "epoch": 0.81,
-      "learning_rate": 4.7760371514885096e-05,
-      "loss": 0.3142,
+      "epoch": 0.8283778426496559,
+      "grad_norm": 0.15680228173732758,
+      "learning_rate": 4.762690471815708e-05,
+      "loss": 0.4153,
       "step": 22985
     },
     {
-      "epoch": 0.81,
-      "learning_rate": 4.7759192861129886e-05,
-      "loss": 0.308,
+      "epoch": 0.8285580423108805,
+      "grad_norm": 0.15533293783664703,
+      "learning_rate": 4.762566362929343e-05,
+      "loss": 0.426,
       "step": 22990
     },
     {
-      "epoch": 0.81,
-      "learning_rate": 4.775801391186156e-05,
-      "loss": 0.3155,
+      "epoch": 0.8287382419721051,
+      "grad_norm": 0.16373470425605774,
+      "learning_rate": 4.762442223215949e-05,
+      "loss": 0.4247,
       "step": 22995
     },
     {
-      "epoch": 0.81,
-      "learning_rate": 4.775683466709541e-05,
-      "loss": 0.3034,
+      "epoch": 0.8289184416333297,
+      "grad_norm": 0.15683810412883759,
+      "learning_rate": 4.762318052677217e-05,
+      "loss": 0.4058,
       "step": 23000
     },
     {
-      "epoch": 0.81,
-      "eval_loss": 0.2990756332874298,
-      "eval_runtime": 10.544,
-      "eval_samples_per_second": 9.484,
-      "eval_steps_per_second": 9.484,
+      "epoch": 0.8289184416333297,
+      "eval_loss": 0.4584188461303711,
+      "eval_runtime": 3.5371,
+      "eval_samples_per_second": 28.271,
+      "eval_steps_per_second": 7.068,
       "step": 23000
     },
     {
-      "epoch": 0.81,
-      "learning_rate": 4.7755655126846766e-05,
-      "loss": 0.3311,
+      "epoch": 0.8290986412945544,
+      "grad_norm": 0.20195883512496948,
+      "learning_rate": 4.76219385131484e-05,
+      "loss": 0.4855,
       "step": 23005
     },
     {
-      "epoch": 0.81,
-      "learning_rate": 4.775447529113093e-05,
-      "loss": 0.3021,
+      "epoch": 0.829278840955779,
+      "grad_norm": 0.1833004504442215,
+      "learning_rate": 4.76206961913051e-05,
+      "loss": 0.4323,
       "step": 23010
     },
     {
-      "epoch": 0.81,
-      "learning_rate": 4.775329515996324e-05,
-      "loss": 0.3034,
+      "epoch": 0.8294590406170036,
+      "grad_norm": 0.19199278950691223,
+      "learning_rate": 4.7619453561259195e-05,
+      "loss": 0.4518,
       "step": 23015
     },
     {
-      "epoch": 0.81,
-      "learning_rate": 4.7752114733358996e-05,
-      "loss": 0.3255,
+      "epoch": 0.8296392402782283,
+      "grad_norm": 0.16530929505825043,
+      "learning_rate": 4.76182106230276e-05,
+      "loss": 0.4471,
       "step": 23020
     },
     {
-      "epoch": 0.81,
-      "learning_rate": 4.775093401133355e-05,
-      "loss": 0.3309,
+      "epoch": 0.829819439939453,
+      "grad_norm": 0.1725630909204483,
+      "learning_rate": 4.761696737662728e-05,
+      "loss": 0.4395,
       "step": 23025
     },
     {
-      "epoch": 0.81,
-      "learning_rate": 4.774975299390221e-05,
-      "loss": 0.3222,
+      "epoch": 0.8299996396006776,
+      "grad_norm": 0.17936821281909943,
+      "learning_rate": 4.761572382207515e-05,
+      "loss": 0.4192,
       "step": 23030
     },
     {
-      "epoch": 0.81,
-      "learning_rate": 4.774857168108032e-05,
-      "loss": 0.3248,
+      "epoch": 0.8301798392619022,
+      "grad_norm": 0.18170322477817535,
+      "learning_rate": 4.761447995938817e-05,
+      "loss": 0.4352,
       "step": 23035
     },
     {
-      "epoch": 0.81,
-      "learning_rate": 4.7747390072883217e-05,
-      "loss": 0.2998,
+      "epoch": 0.8303600389231268,
+      "grad_norm": 0.18376685678958893,
+      "learning_rate": 4.761323578858327e-05,
+      "loss": 0.4149,
       "step": 23040
     },
     {
-      "epoch": 0.81,
-      "learning_rate": 4.7746208169326246e-05,
-      "loss": 0.2918,
+      "epoch": 0.8305402385843514,
+      "grad_norm": 0.16382743418216705,
+      "learning_rate": 4.761199130967742e-05,
+      "loss": 0.4194,
       "step": 23045
     },
     {
-      "epoch": 0.81,
-      "learning_rate": 4.7745025970424756e-05,
-      "loss": 0.3339,
+      "epoch": 0.8307204382455761,
+      "grad_norm": 0.2041010707616806,
+      "learning_rate": 4.761074652268755e-05,
+      "loss": 0.457,
       "step": 23050
     },
     {
-      "epoch": 0.81,
-      "learning_rate": 4.7743843476194096e-05,
-      "loss": 0.2757,
+      "epoch": 0.8309006379068007,
+      "grad_norm": 0.15971267223358154,
+      "learning_rate": 4.7609501427630657e-05,
+      "loss": 0.41,
       "step": 23055
     },
     {
-      "epoch": 0.81,
-      "learning_rate": 4.774266068664961e-05,
-      "loss": 0.3043,
+      "epoch": 0.8310808375680254,
+      "grad_norm": 0.16985180974006653,
+      "learning_rate": 4.7608256024523666e-05,
+      "loss": 0.4303,
       "step": 23060
     },
     {
-      "epoch": 0.81,
-      "learning_rate": 4.7741477601806664e-05,
-      "loss": 0.3021,
+      "epoch": 0.83126103722925,
+      "grad_norm": 0.1820903718471527,
+      "learning_rate": 4.760701031338358e-05,
+      "loss": 0.4765,
       "step": 23065
     },
     {
-      "epoch": 0.81,
-      "learning_rate": 4.7740294221680616e-05,
-      "loss": 0.2808,
+      "epoch": 0.8314412368904747,
+      "grad_norm": 0.14193592965602875,
+      "learning_rate": 4.760576429422734e-05,
+      "loss": 0.4375,
       "step": 23070
     },
     {
-      "epoch": 0.81,
-      "learning_rate": 4.7739110546286844e-05,
-      "loss": 0.2895,
+      "epoch": 0.8316214365516993,
+      "grad_norm": 0.1694120466709137,
+      "learning_rate": 4.760451796707195e-05,
+      "loss": 0.4215,
       "step": 23075
     },
     {
-      "epoch": 0.81,
-      "learning_rate": 4.7737926575640705e-05,
-      "loss": 0.3191,
+      "epoch": 0.8318016362129239,
+      "grad_norm": 0.172800675034523,
+      "learning_rate": 4.7603271331934376e-05,
+      "loss": 0.4102,
       "step": 23080
     },
     {
-      "epoch": 0.81,
-      "learning_rate": 4.7736742309757565e-05,
-      "loss": 0.3305,
+      "epoch": 0.8319818358741485,
+      "grad_norm": 0.15721376240253448,
+      "learning_rate": 4.76020243888316e-05,
+      "loss": 0.4472,
       "step": 23085
     },
     {
-      "epoch": 0.81,
-      "learning_rate": 4.7735557748652815e-05,
-      "loss": 0.3037,
+      "epoch": 0.8321620355353732,
+      "grad_norm": 0.15566657483577728,
+      "learning_rate": 4.760077713778062e-05,
+      "loss": 0.3864,
       "step": 23090
     },
     {
-      "epoch": 0.81,
-      "learning_rate": 4.773437289234183e-05,
-      "loss": 0.2879,
+      "epoch": 0.8323422351965978,
+      "grad_norm": 0.18366749584674835,
+      "learning_rate": 4.759952957879843e-05,
+      "loss": 0.4561,
       "step": 23095
     },
     {
-      "epoch": 0.81,
-      "learning_rate": 4.7733187740839996e-05,
-      "loss": 0.3036,
+      "epoch": 0.8325224348578225,
+      "grad_norm": 0.15724977850914001,
+      "learning_rate": 4.759828171190202e-05,
+      "loss": 0.428,
       "step": 23100
     },
     {
-      "epoch": 0.81,
-      "learning_rate": 4.7732002294162706e-05,
-      "loss": 0.3107,
+      "epoch": 0.8327026345190471,
+      "grad_norm": 0.2003716081380844,
+      "learning_rate": 4.7597033537108405e-05,
+      "loss": 0.4906,
       "step": 23105
     },
     {
-      "epoch": 0.81,
-      "learning_rate": 4.773081655232534e-05,
-      "loss": 0.3171,
+      "epoch": 0.8328828341802718,
+      "grad_norm": 0.1818516105413437,
+      "learning_rate": 4.759578505443458e-05,
+      "loss": 0.4225,
       "step": 23110
     },
     {
-      "epoch": 0.81,
-      "learning_rate": 4.77296305153433e-05,
-      "loss": 0.3143,
+      "epoch": 0.8330630338414964,
+      "grad_norm": 0.21417291462421417,
+      "learning_rate": 4.759453626389756e-05,
+      "loss": 0.4341,
       "step": 23115
     },
     {
-      "epoch": 0.81,
-      "learning_rate": 4.7728444183232e-05,
-      "loss": 0.3031,
+      "epoch": 0.833243233502721,
+      "grad_norm": 0.16403871774673462,
+      "learning_rate": 4.759328716551435e-05,
+      "loss": 0.4445,
       "step": 23120
     },
     {
-      "epoch": 0.81,
-      "learning_rate": 4.772725755600682e-05,
-      "loss": 0.3024,
+      "epoch": 0.8334234331639456,
+      "grad_norm": 0.16569961607456207,
+      "learning_rate": 4.759203775930198e-05,
+      "loss": 0.4387,
       "step": 23125
     },
     {
-      "epoch": 0.81,
-      "learning_rate": 4.772607063368318e-05,
-      "loss": 0.2787,
+      "epoch": 0.8336036328251702,
+      "grad_norm": 0.16702339053153992,
+      "learning_rate": 4.7590788045277474e-05,
+      "loss": 0.3888,
       "step": 23130
     },
     {
-      "epoch": 0.81,
-      "learning_rate": 4.772488341627649e-05,
-      "loss": 0.3211,
+      "epoch": 0.8337838324863949,
+      "grad_norm": 0.16976197063922882,
+      "learning_rate": 4.758953802345785e-05,
+      "loss": 0.4482,
       "step": 23135
     },
     {
-      "epoch": 0.81,
-      "learning_rate": 4.772369590380217e-05,
-      "loss": 0.2972,
+      "epoch": 0.8339640321476196,
+      "grad_norm": 0.16811056435108185,
+      "learning_rate": 4.758828769386015e-05,
+      "loss": 0.4274,
       "step": 23140
     },
     {
-      "epoch": 0.81,
-      "learning_rate": 4.7722508096275626e-05,
-      "loss": 0.2936,
+      "epoch": 0.8341442318088442,
+      "grad_norm": 0.13746735453605652,
+      "learning_rate": 4.758703705650139e-05,
+      "loss": 0.4351,
       "step": 23145
     },
     {
-      "epoch": 0.81,
-      "learning_rate": 4.7721319993712286e-05,
-      "loss": 0.3176,
+      "epoch": 0.8343244314700689,
+      "grad_norm": 0.16167333722114563,
+      "learning_rate": 4.758578611139864e-05,
+      "loss": 0.4553,
       "step": 23150
     },
     {
-      "epoch": 0.81,
-      "learning_rate": 4.7720131596127594e-05,
-      "loss": 0.2945,
+      "epoch": 0.8345046311312935,
+      "grad_norm": 0.16944856941699982,
+      "learning_rate": 4.758453485856892e-05,
+      "loss": 0.3807,
       "step": 23155
     },
     {
-      "epoch": 0.81,
-      "learning_rate": 4.771894290353695e-05,
-      "loss": 0.2996,
+      "epoch": 0.8346848307925181,
+      "grad_norm": 0.17565567791461945,
+      "learning_rate": 4.758328329802928e-05,
+      "loss": 0.4513,
       "step": 23160
     },
     {
-      "epoch": 0.82,
-      "learning_rate": 4.771775391595582e-05,
-      "loss": 0.3121,
+      "epoch": 0.8348650304537427,
+      "grad_norm": 0.1966121941804886,
+      "learning_rate": 4.758203142979678e-05,
+      "loss": 0.4304,
       "step": 23165
     },
     {
-      "epoch": 0.82,
-      "learning_rate": 4.771656463339962e-05,
-      "loss": 0.3092,
+      "epoch": 0.8350452301149673,
+      "grad_norm": 0.15772442519664764,
+      "learning_rate": 4.7580779253888476e-05,
+      "loss": 0.4289,
       "step": 23170
     },
     {
-      "epoch": 0.82,
-      "learning_rate": 4.7715375055883804e-05,
-      "loss": 0.2792,
+      "epoch": 0.8352254297761921,
+      "grad_norm": 0.16442222893238068,
+      "learning_rate": 4.757952677032143e-05,
+      "loss": 0.4614,
       "step": 23175
     },
     {
-      "epoch": 0.82,
-      "learning_rate": 4.771418518342381e-05,
-      "loss": 0.2955,
+      "epoch": 0.8354056294374167,
+      "grad_norm": 0.15727081894874573,
+      "learning_rate": 4.7578273979112696e-05,
+      "loss": 0.4287,
       "step": 23180
     },
     {
-      "epoch": 0.82,
-      "learning_rate": 4.771299501603509e-05,
-      "loss": 0.3136,
+      "epoch": 0.8355858290986413,
+      "grad_norm": 0.15362627804279327,
+      "learning_rate": 4.757702088027935e-05,
+      "loss": 0.4199,
       "step": 23185
     },
     {
-      "epoch": 0.82,
-      "learning_rate": 4.7711804553733095e-05,
-      "loss": 0.3088,
+      "epoch": 0.8357660287598659,
+      "grad_norm": 0.1784539520740509,
+      "learning_rate": 4.757576747383847e-05,
+      "loss": 0.4247,
       "step": 23190
     },
     {
-      "epoch": 0.82,
-      "learning_rate": 4.771061379653329e-05,
-      "loss": 0.2912,
+      "epoch": 0.8359462284210906,
+      "grad_norm": 0.1273956149816513,
+      "learning_rate": 4.757451375980713e-05,
+      "loss": 0.4282,
       "step": 23195
     },
     {
-      "epoch": 0.82,
-      "learning_rate": 4.7709422744451126e-05,
-      "loss": 0.3338,
+      "epoch": 0.8361264280823152,
+      "grad_norm": 0.10995710641145706,
+      "learning_rate": 4.757325973820241e-05,
+      "loss": 0.393,
       "step": 23200
     },
     {
-      "epoch": 0.82,
-      "learning_rate": 4.7708231397502076e-05,
-      "loss": 0.2899,
+      "epoch": 0.8363066277435398,
+      "grad_norm": 0.1851365864276886,
+      "learning_rate": 4.75720054090414e-05,
+      "loss": 0.4583,
       "step": 23205
     },
     {
-      "epoch": 0.82,
-      "learning_rate": 4.770703975570161e-05,
-      "loss": 0.3269,
+      "epoch": 0.8364868274047644,
+      "grad_norm": 0.16413268446922302,
+      "learning_rate": 4.7570750772341174e-05,
+      "loss": 0.4797,
       "step": 23210
     },
     {
-      "epoch": 0.82,
-      "learning_rate": 4.7705847819065195e-05,
-      "loss": 0.3311,
+      "epoch": 0.8366670270659892,
+      "grad_norm": 0.1683937907218933,
+      "learning_rate": 4.756949582811885e-05,
+      "loss": 0.4372,
       "step": 23215
     },
     {
-      "epoch": 0.82,
-      "learning_rate": 4.770465558760831e-05,
-      "loss": 0.3322,
+      "epoch": 0.8368472267272138,
+      "grad_norm": 0.1728678047657013,
+      "learning_rate": 4.7568240576391507e-05,
+      "loss": 0.4376,
       "step": 23220
     },
     {
-      "epoch": 0.82,
-      "learning_rate": 4.770346306134643e-05,
-      "loss": 0.3201,
+      "epoch": 0.8370274263884384,
+      "grad_norm": 0.20877417922019958,
+      "learning_rate": 4.7566985017176255e-05,
+      "loss": 0.4624,
       "step": 23225
     },
     {
-      "epoch": 0.82,
-      "learning_rate": 4.770227024029506e-05,
-      "loss": 0.2873,
+      "epoch": 0.837207626049663,
+      "grad_norm": 0.18232528865337372,
+      "learning_rate": 4.756572915049021e-05,
+      "loss": 0.4714,
       "step": 23230
     },
     {
-      "epoch": 0.82,
-      "learning_rate": 4.7701077124469655e-05,
-      "loss": 0.2996,
+      "epoch": 0.8373878257108877,
+      "grad_norm": 0.1842896044254303,
+      "learning_rate": 4.756447297635047e-05,
+      "loss": 0.4341,
       "step": 23235
     },
     {
-      "epoch": 0.82,
-      "learning_rate": 4.769988371388573e-05,
-      "loss": 0.3076,
+      "epoch": 0.8375680253721123,
+      "grad_norm": 0.17772360146045685,
+      "learning_rate": 4.756321649477415e-05,
+      "loss": 0.4225,
       "step": 23240
     },
     {
-      "epoch": 0.82,
-      "learning_rate": 4.769869000855878e-05,
-      "loss": 0.3316,
+      "epoch": 0.8377482250333369,
+      "grad_norm": 0.16995792090892792,
+      "learning_rate": 4.756195970577838e-05,
+      "loss": 0.4395,
       "step": 23245
     },
     {
-      "epoch": 0.82,
-      "learning_rate": 4.7697496008504297e-05,
-      "loss": 0.3065,
+      "epoch": 0.8379284246945615,
+      "grad_norm": 0.1724376529455185,
+      "learning_rate": 4.7560702609380275e-05,
+      "loss": 0.428,
       "step": 23250
     },
     {
-      "epoch": 0.82,
-      "learning_rate": 4.769630171373778e-05,
-      "loss": 0.2948,
+      "epoch": 0.8381086243557863,
+      "grad_norm": 0.1635865867137909,
+      "learning_rate": 4.755944520559697e-05,
+      "loss": 0.424,
       "step": 23255
     },
     {
-      "epoch": 0.82,
-      "learning_rate": 4.769510712427475e-05,
-      "loss": 0.3184,
+      "epoch": 0.8382888240170109,
+      "grad_norm": 0.14121320843696594,
+      "learning_rate": 4.755818749444558e-05,
+      "loss": 0.4165,
       "step": 23260
     },
     {
-      "epoch": 0.82,
-      "learning_rate": 4.7693912240130703e-05,
-      "loss": 0.2942,
+      "epoch": 0.8384690236782355,
+      "grad_norm": 0.19829101860523224,
+      "learning_rate": 4.755692947594326e-05,
+      "loss": 0.4237,
       "step": 23265
     },
     {
-      "epoch": 0.82,
-      "learning_rate": 4.7692717061321165e-05,
-      "loss": 0.3225,
+      "epoch": 0.8386492233394601,
+      "grad_norm": 0.17292529344558716,
+      "learning_rate": 4.755567115010714e-05,
+      "loss": 0.4166,
       "step": 23270
     },
     {
-      "epoch": 0.82,
-      "learning_rate": 4.7691521587861655e-05,
-      "loss": 0.328,
+      "epoch": 0.8388294230006847,
+      "grad_norm": 0.174628347158432,
+      "learning_rate": 4.755441251695437e-05,
+      "loss": 0.4377,
       "step": 23275
     },
     {
-      "epoch": 0.82,
-      "learning_rate": 4.769032581976769e-05,
-      "loss": 0.3131,
+      "epoch": 0.8390096226619094,
+      "grad_norm": 0.1492447406053543,
+      "learning_rate": 4.75531535765021e-05,
+      "loss": 0.4591,
       "step": 23280
     },
     {
-      "epoch": 0.82,
-      "learning_rate": 4.7689129757054795e-05,
-      "loss": 0.3128,
+      "epoch": 0.839189822323134,
+      "grad_norm": 0.20081254839897156,
+      "learning_rate": 4.755189432876747e-05,
+      "loss": 0.4029,
       "step": 23285
     },
     {
-      "epoch": 0.82,
-      "learning_rate": 4.76879333997385e-05,
-      "loss": 0.2992,
+      "epoch": 0.8393700219843586,
+      "grad_norm": 0.19274207949638367,
+      "learning_rate": 4.755063477376766e-05,
+      "loss": 0.475,
       "step": 23290
     },
     {
-      "epoch": 0.82,
-      "learning_rate": 4.768673674783435e-05,
-      "loss": 0.3081,
+      "epoch": 0.8395502216455833,
+      "grad_norm": 0.16275402903556824,
+      "learning_rate": 4.75493749115198e-05,
+      "loss": 0.456,
       "step": 23295
     },
     {
-      "epoch": 0.82,
-      "learning_rate": 4.768553980135787e-05,
-      "loss": 0.2858,
+      "epoch": 0.839730421306808,
+      "grad_norm": 0.2030659317970276,
+      "learning_rate": 4.7548114742041084e-05,
+      "loss": 0.4322,
       "step": 23300
     },
     {
-      "epoch": 0.82,
-      "learning_rate": 4.768434256032461e-05,
-      "loss": 0.332,
+      "epoch": 0.8399106209680326,
+      "grad_norm": 0.15633316338062286,
+      "learning_rate": 4.754685426534866e-05,
+      "loss": 0.4611,
       "step": 23305
     },
     {
-      "epoch": 0.82,
-      "learning_rate": 4.768314502475011e-05,
-      "loss": 0.3253,
+      "epoch": 0.8400908206292572,
+      "grad_norm": 0.17400497198104858,
+      "learning_rate": 4.754559348145972e-05,
+      "loss": 0.4066,
       "step": 23310
     },
     {
-      "epoch": 0.82,
-      "learning_rate": 4.768194719464991e-05,
-      "loss": 0.3177,
+      "epoch": 0.8402710202904818,
+      "grad_norm": 0.157339945435524,
+      "learning_rate": 4.754433239039143e-05,
+      "loss": 0.4289,
       "step": 23315
     },
     {
-      "epoch": 0.82,
-      "learning_rate": 4.768074907003959e-05,
-      "loss": 0.314,
+      "epoch": 0.8404512199517065,
+      "grad_norm": 0.1500500738620758,
+      "learning_rate": 4.7543070992160984e-05,
+      "loss": 0.4215,
       "step": 23320
     },
     {
-      "epoch": 0.82,
-      "learning_rate": 4.7679550650934676e-05,
-      "loss": 0.2749,
+      "epoch": 0.8406314196129311,
+      "grad_norm": 0.21953436732292175,
+      "learning_rate": 4.754180928678555e-05,
+      "loss": 0.4825,
       "step": 23325
     },
     {
-      "epoch": 0.82,
-      "learning_rate": 4.767835193735075e-05,
-      "loss": 0.2995,
+      "epoch": 0.8408116192741558,
+      "grad_norm": 0.1740504950284958,
+      "learning_rate": 4.754054727428233e-05,
+      "loss": 0.4395,
       "step": 23330
     },
     {
-      "epoch": 0.82,
-      "learning_rate": 4.767715292930337e-05,
-      "loss": 0.3238,
+      "epoch": 0.8409918189353804,
+      "grad_norm": 0.17277638614177704,
+      "learning_rate": 4.753928495466853e-05,
+      "loss": 0.431,
       "step": 23335
     },
     {
-      "epoch": 0.82,
-      "learning_rate": 4.76759536268081e-05,
-      "loss": 0.312,
+      "epoch": 0.8411720185966051,
+      "grad_norm": 0.196901336312294,
+      "learning_rate": 4.7538022327961316e-05,
+      "loss": 0.4826,
       "step": 23340
     },
     {
-      "epoch": 0.82,
-      "learning_rate": 4.7674754029880524e-05,
-      "loss": 0.2929,
+      "epoch": 0.8413522182578297,
+      "grad_norm": 0.2405674159526825,
+      "learning_rate": 4.7536759394177925e-05,
+      "loss": 0.4524,
       "step": 23345
     },
     {
-      "epoch": 0.82,
-      "learning_rate": 4.76735541385362e-05,
-      "loss": 0.3103,
+      "epoch": 0.8415324179190543,
+      "grad_norm": 0.1807735562324524,
+      "learning_rate": 4.7535496153335544e-05,
+      "loss": 0.4144,
       "step": 23350
     },
     {
-      "epoch": 0.82,
-      "learning_rate": 4.767235395279073e-05,
-      "loss": 0.3366,
+      "epoch": 0.8417126175802789,
+      "grad_norm": 0.1897483915090561,
+      "learning_rate": 4.75342326054514e-05,
+      "loss": 0.431,
       "step": 23355
     },
     {
-      "epoch": 0.82,
-      "learning_rate": 4.767115347265968e-05,
-      "loss": 0.3116,
+      "epoch": 0.8418928172415036,
+      "grad_norm": 0.17141766846179962,
+      "learning_rate": 4.7532968750542694e-05,
+      "loss": 0.4671,
       "step": 23360
     },
     {
-      "epoch": 0.82,
-      "learning_rate": 4.766995269815864e-05,
-      "loss": 0.3052,
+      "epoch": 0.8420730169027282,
+      "grad_norm": 0.17842607200145721,
+      "learning_rate": 4.753170458862665e-05,
+      "loss": 0.471,
       "step": 23365
     },
     {
-      "epoch": 0.82,
-      "learning_rate": 4.766875162930321e-05,
-      "loss": 0.3002,
+      "epoch": 0.8422532165639529,
+      "grad_norm": 0.14743296802043915,
+      "learning_rate": 4.75304401197205e-05,
+      "loss": 0.4735,
       "step": 23370
     },
     {
-      "epoch": 0.82,
-      "learning_rate": 4.7667550266108974e-05,
-      "loss": 0.3176,
+      "epoch": 0.8424334162251775,
+      "grad_norm": 0.1760016530752182,
+      "learning_rate": 4.7529175343841455e-05,
+      "loss": 0.431,
       "step": 23375
     },
     {
-      "epoch": 0.82,
-      "learning_rate": 4.766634860859154e-05,
-      "loss": 0.2901,
+      "epoch": 0.8426136158864022,
+      "grad_norm": 0.16929085552692413,
+      "learning_rate": 4.7527910261006755e-05,
+      "loss": 0.4272,
       "step": 23380
     },
     {
-      "epoch": 0.82,
-      "learning_rate": 4.76651466567665e-05,
-      "loss": 0.2979,
+      "epoch": 0.8427938155476268,
+      "grad_norm": 0.167909175157547,
+      "learning_rate": 4.752664487123365e-05,
+      "loss": 0.4715,
       "step": 23385
     },
     {
-      "epoch": 0.82,
-      "learning_rate": 4.766394441064948e-05,
-      "loss": 0.3301,
+      "epoch": 0.8429740152088514,
+      "grad_norm": 0.17611196637153625,
+      "learning_rate": 4.752537917453937e-05,
+      "loss": 0.4361,
       "step": 23390
     },
     {
-      "epoch": 0.82,
-      "learning_rate": 4.766274187025607e-05,
-      "loss": 0.3071,
+      "epoch": 0.843154214870076,
+      "grad_norm": 0.1789659559726715,
+      "learning_rate": 4.752411317094115e-05,
+      "loss": 0.4534,
       "step": 23395
     },
     {
-      "epoch": 0.82,
-      "learning_rate": 4.766153903560189e-05,
-      "loss": 0.3096,
+      "epoch": 0.8433344145313006,
+      "grad_norm": 0.17830581963062286,
+      "learning_rate": 4.752284686045626e-05,
+      "loss": 0.4091,
       "step": 23400
     },
     {
-      "epoch": 0.82,
-      "learning_rate": 4.766033590670256e-05,
-      "loss": 0.313,
+      "epoch": 0.8435146141925253,
+      "grad_norm": 0.18294012546539307,
+      "learning_rate": 4.752158024310194e-05,
+      "loss": 0.4673,
       "step": 23405
     },
     {
-      "epoch": 0.82,
-      "learning_rate": 4.765913248357371e-05,
-      "loss": 0.3152,
+      "epoch": 0.84369481385375,
+      "grad_norm": 0.15136177837848663,
+      "learning_rate": 4.752031331889545e-05,
+      "loss": 0.4319,
       "step": 23410
     },
     {
-      "epoch": 0.82,
-      "learning_rate": 4.7657928766230944e-05,
-      "loss": 0.2958,
+      "epoch": 0.8438750135149746,
+      "grad_norm": 0.20712289214134216,
+      "learning_rate": 4.751904608785405e-05,
+      "loss": 0.4255,
       "step": 23415
     },
     {
-      "epoch": 0.82,
-      "learning_rate": 4.7656724754689916e-05,
-      "loss": 0.3091,
+      "epoch": 0.8440552131761992,
+      "grad_norm": 0.1756359040737152,
+      "learning_rate": 4.7517778549994994e-05,
+      "loss": 0.4445,
       "step": 23420
     },
     {
-      "epoch": 0.82,
-      "learning_rate": 4.765552044896625e-05,
-      "loss": 0.3127,
+      "epoch": 0.8442354128374239,
+      "grad_norm": 0.15145882964134216,
+      "learning_rate": 4.751651070533558e-05,
+      "loss": 0.4436,
       "step": 23425
     },
     {
-      "epoch": 0.82,
-      "learning_rate": 4.7654315849075574e-05,
-      "loss": 0.3271,
+      "epoch": 0.8444156124986485,
+      "grad_norm": 0.18007196485996246,
+      "learning_rate": 4.7515242553893056e-05,
+      "loss": 0.4498,
       "step": 23430
     },
     {
-      "epoch": 0.82,
-      "learning_rate": 4.7653110955033545e-05,
-      "loss": 0.2826,
+      "epoch": 0.8445958121598731,
+      "grad_norm": 0.17577087879180908,
+      "learning_rate": 4.751397409568472e-05,
+      "loss": 0.4107,
       "step": 23435
     },
     {
-      "epoch": 0.82,
-      "learning_rate": 4.765190576685579e-05,
-      "loss": 0.3016,
+      "epoch": 0.8447760118210977,
+      "grad_norm": 0.1596294343471527,
+      "learning_rate": 4.7512705330727847e-05,
+      "loss": 0.4042,
       "step": 23440
     },
     {
-      "epoch": 0.82,
-      "learning_rate": 4.765070028455797e-05,
-      "loss": 0.308,
+      "epoch": 0.8449562114823224,
+      "grad_norm": 0.14652547240257263,
+      "learning_rate": 4.751143625903972e-05,
+      "loss": 0.4424,
       "step": 23445
     },
     {
-      "epoch": 0.83,
-      "learning_rate": 4.7649494508155735e-05,
-      "loss": 0.2744,
+      "epoch": 0.8451364111435471,
+      "grad_norm": 0.14940057694911957,
+      "learning_rate": 4.751016688063763e-05,
+      "loss": 0.4387,
       "step": 23450
     },
     {
-      "epoch": 0.83,
-      "learning_rate": 4.7648288437664746e-05,
-      "loss": 0.3216,
+      "epoch": 0.8453166108047717,
+      "grad_norm": 0.16962915658950806,
+      "learning_rate": 4.750889719553888e-05,
+      "loss": 0.447,
       "step": 23455
     },
     {
-      "epoch": 0.83,
-      "learning_rate": 4.764708207310065e-05,
-      "loss": 0.2988,
+      "epoch": 0.8454968104659963,
+      "grad_norm": 0.193365216255188,
+      "learning_rate": 4.7507627203760754e-05,
+      "loss": 0.4235,
       "step": 23460
     },
     {
-      "epoch": 0.83,
-      "learning_rate": 4.7645875414479115e-05,
-      "loss": 0.3203,
+      "epoch": 0.845677010127221,
+      "grad_norm": 0.24096348881721497,
+      "learning_rate": 4.7506356905320574e-05,
+      "loss": 0.4828,
       "step": 23465
     },
     {
-      "epoch": 0.83,
-      "learning_rate": 4.764466846181582e-05,
-      "loss": 0.3257,
+      "epoch": 0.8458572097884456,
+      "grad_norm": 0.17045122385025024,
+      "learning_rate": 4.7505086300235635e-05,
+      "loss": 0.3911,
       "step": 23470
     },
     {
-      "epoch": 0.83,
-      "learning_rate": 4.7643461215126425e-05,
-      "loss": 0.3117,
+      "epoch": 0.8460374094496702,
+      "grad_norm": 0.18029987812042236,
+      "learning_rate": 4.750381538852325e-05,
+      "loss": 0.4495,
       "step": 23475
     },
     {
-      "epoch": 0.83,
-      "learning_rate": 4.764225367442661e-05,
-      "loss": 0.3221,
+      "epoch": 0.8462176091108948,
+      "grad_norm": 0.1832016408443451,
+      "learning_rate": 4.7502544170200735e-05,
+      "loss": 0.4109,
       "step": 23480
     },
     {
-      "epoch": 0.83,
-      "learning_rate": 4.764104583973205e-05,
-      "loss": 0.3148,
+      "epoch": 0.8463978087721195,
+      "grad_norm": 0.14569438993930817,
+      "learning_rate": 4.750127264528542e-05,
+      "loss": 0.4221,
       "step": 23485
     },
     {
-      "epoch": 0.83,
-      "learning_rate": 4.7639837711058435e-05,
-      "loss": 0.3044,
+      "epoch": 0.8465780084333442,
+      "grad_norm": 0.16590553522109985,
+      "learning_rate": 4.750000081379462e-05,
+      "loss": 0.4384,
       "step": 23490
     },
     {
-      "epoch": 0.83,
-      "learning_rate": 4.763862928842144e-05,
-      "loss": 0.3233,
+      "epoch": 0.8467582080945688,
+      "grad_norm": 0.201650008559227,
+      "learning_rate": 4.749872867574566e-05,
+      "loss": 0.4363,
       "step": 23495
     },
     {
-      "epoch": 0.83,
-      "learning_rate": 4.7637420571836775e-05,
-      "loss": 0.3088,
+      "epoch": 0.8469384077557934,
+      "grad_norm": 0.20361419022083282,
+      "learning_rate": 4.7497456231155884e-05,
+      "loss": 0.407,
       "step": 23500
     },
     {
-      "epoch": 0.83,
-      "eval_loss": 0.29781535267829895,
-      "eval_runtime": 10.5453,
-      "eval_samples_per_second": 9.483,
-      "eval_steps_per_second": 9.483,
+      "epoch": 0.8469384077557934,
+      "eval_loss": 0.4585105776786804,
+      "eval_runtime": 3.5434,
+      "eval_samples_per_second": 28.221,
+      "eval_steps_per_second": 7.055,
       "step": 23500
     },
     {
-      "epoch": 0.83,
-      "learning_rate": 4.763621156132011e-05,
-      "loss": 0.2941,
+      "epoch": 0.847118607417018,
+      "grad_norm": 0.1582178771495819,
+      "learning_rate": 4.749618348004262e-05,
+      "loss": 0.441,
       "step": 23505
     },
     {
-      "epoch": 0.83,
-      "learning_rate": 4.763500225688716e-05,
-      "loss": 0.3113,
+      "epoch": 0.8472988070782427,
+      "grad_norm": 0.1565631479024887,
+      "learning_rate": 4.749491042242321e-05,
+      "loss": 0.4264,
       "step": 23510
     },
     {
-      "epoch": 0.83,
-      "learning_rate": 4.763379265855363e-05,
-      "loss": 0.3271,
+      "epoch": 0.8474790067394673,
+      "grad_norm": 0.15611840784549713,
+      "learning_rate": 4.7493637058314996e-05,
+      "loss": 0.4848,
       "step": 23515
     },
     {
-      "epoch": 0.83,
-      "learning_rate": 4.763258276633521e-05,
-      "loss": 0.3066,
+      "epoch": 0.8476592064006919,
+      "grad_norm": 0.17525018751621246,
+      "learning_rate": 4.749236338773535e-05,
+      "loss": 0.4368,
       "step": 23520
     },
     {
-      "epoch": 0.83,
-      "learning_rate": 4.7631372580247624e-05,
-      "loss": 0.2862,
+      "epoch": 0.8478394060619167,
+      "grad_norm": 0.18957500159740448,
+      "learning_rate": 4.749108941070159e-05,
+      "loss": 0.4456,
       "step": 23525
     },
     {
-      "epoch": 0.83,
-      "learning_rate": 4.7630162100306576e-05,
-      "loss": 0.3283,
+      "epoch": 0.8480196057231413,
+      "grad_norm": 0.1911652833223343,
+      "learning_rate": 4.74898151272311e-05,
+      "loss": 0.4294,
       "step": 23530
     },
     {
-      "epoch": 0.83,
-      "learning_rate": 4.7628951326527786e-05,
-      "loss": 0.3109,
+      "epoch": 0.8481998053843659,
+      "grad_norm": 0.17784404754638672,
+      "learning_rate": 4.748854053734122e-05,
+      "loss": 0.4079,
       "step": 23535
     },
     {
-      "epoch": 0.83,
-      "learning_rate": 4.762774025892698e-05,
-      "loss": 0.2908,
+      "epoch": 0.8483800050455905,
+      "grad_norm": 0.16650652885437012,
+      "learning_rate": 4.748726564104935e-05,
+      "loss": 0.4114,
       "step": 23540
     },
     {
-      "epoch": 0.83,
-      "learning_rate": 4.762652889751988e-05,
-      "loss": 0.3238,
+      "epoch": 0.8485602047068151,
+      "grad_norm": 0.15656359493732452,
+      "learning_rate": 4.748599043837282e-05,
+      "loss": 0.4926,
       "step": 23545
     },
     {
-      "epoch": 0.83,
-      "learning_rate": 4.762531724232221e-05,
-      "loss": 0.3091,
+      "epoch": 0.8487404043680398,
+      "grad_norm": 0.1699918955564499,
+      "learning_rate": 4.748471492932903e-05,
+      "loss": 0.4273,
       "step": 23550
     },
     {
-      "epoch": 0.83,
-      "learning_rate": 4.7624105293349706e-05,
-      "loss": 0.3094,
+      "epoch": 0.8489206040292644,
+      "grad_norm": 0.15680480003356934,
+      "learning_rate": 4.748343911393536e-05,
+      "loss": 0.3945,
       "step": 23555
     },
     {
-      "epoch": 0.83,
-      "learning_rate": 4.762289305061811e-05,
-      "loss": 0.316,
+      "epoch": 0.849100803690489,
+      "grad_norm": 0.14087727665901184,
+      "learning_rate": 4.748216299220918e-05,
+      "loss": 0.4801,
       "step": 23560
     },
     {
-      "epoch": 0.83,
-      "learning_rate": 4.762168051414315e-05,
-      "loss": 0.327,
+      "epoch": 0.8492810033517137,
+      "grad_norm": 0.1624569147825241,
+      "learning_rate": 4.7480886564167877e-05,
+      "loss": 0.3957,
       "step": 23565
     },
     {
-      "epoch": 0.83,
-      "learning_rate": 4.7620467683940574e-05,
-      "loss": 0.2885,
+      "epoch": 0.8494612030129384,
+      "grad_norm": 0.15580429136753082,
+      "learning_rate": 4.747960982982885e-05,
+      "loss": 0.425,
       "step": 23570
     },
     {
-      "epoch": 0.83,
-      "learning_rate": 4.761925456002614e-05,
-      "loss": 0.3026,
+      "epoch": 0.849641402674163,
+      "grad_norm": 0.15668196976184845,
+      "learning_rate": 4.747833278920949e-05,
+      "loss": 0.4421,
       "step": 23575
     },
     {
-      "epoch": 0.83,
-      "learning_rate": 4.7618041142415596e-05,
-      "loss": 0.3034,
+      "epoch": 0.8498216023353876,
+      "grad_norm": 0.14758959412574768,
+      "learning_rate": 4.74770554423272e-05,
+      "loss": 0.4473,
       "step": 23580
     },
     {
-      "epoch": 0.83,
-      "learning_rate": 4.7616827431124674e-05,
-      "loss": 0.2888,
+      "epoch": 0.8500018019966122,
+      "grad_norm": 0.15379658341407776,
+      "learning_rate": 4.747577778919938e-05,
+      "loss": 0.4246,
       "step": 23585
     },
     {
-      "epoch": 0.83,
-      "learning_rate": 4.761561342616917e-05,
-      "loss": 0.3184,
+      "epoch": 0.8501820016578369,
+      "grad_norm": 0.14333048462867737,
+      "learning_rate": 4.747449982984344e-05,
+      "loss": 0.4508,
       "step": 23590
     },
     {
-      "epoch": 0.83,
-      "learning_rate": 4.761439912756483e-05,
-      "loss": 0.314,
+      "epoch": 0.8503622013190615,
+      "grad_norm": 0.18294481933116913,
+      "learning_rate": 4.747322156427679e-05,
+      "loss": 0.4115,
       "step": 23595
     },
     {
-      "epoch": 0.83,
-      "learning_rate": 4.7613184535327414e-05,
-      "loss": 0.3091,
+      "epoch": 0.8505424009802861,
+      "grad_norm": 0.18898597359657288,
+      "learning_rate": 4.747194299251685e-05,
+      "loss": 0.4568,
       "step": 23600
     },
     {
-      "epoch": 0.83,
-      "learning_rate": 4.76119696494727e-05,
-      "loss": 0.2994,
+      "epoch": 0.8507226006415108,
+      "grad_norm": 0.1791786253452301,
+      "learning_rate": 4.7470664114581034e-05,
+      "loss": 0.4312,
       "step": 23605
     },
     {
-      "epoch": 0.83,
-      "learning_rate": 4.761075447001647e-05,
-      "loss": 0.3185,
+      "epoch": 0.8509028003027355,
+      "grad_norm": 0.2092021107673645,
+      "learning_rate": 4.746938493048677e-05,
+      "loss": 0.4392,
       "step": 23610
     },
     {
-      "epoch": 0.83,
-      "learning_rate": 4.760953899697449e-05,
-      "loss": 0.3372,
+      "epoch": 0.8510829999639601,
+      "grad_norm": 0.17262564599514008,
+      "learning_rate": 4.7468105440251494e-05,
+      "loss": 0.4335,
       "step": 23615
     },
     {
-      "epoch": 0.83,
-      "learning_rate": 4.760832323036254e-05,
-      "loss": 0.3087,
+      "epoch": 0.8512631996251847,
+      "grad_norm": 0.1536947339773178,
+      "learning_rate": 4.746682564389262e-05,
+      "loss": 0.3972,
       "step": 23620
     },
     {
-      "epoch": 0.83,
-      "learning_rate": 4.7607107170196417e-05,
-      "loss": 0.2997,
+      "epoch": 0.8514433992864093,
+      "grad_norm": 0.16581547260284424,
+      "learning_rate": 4.7465545541427605e-05,
+      "loss": 0.4377,
       "step": 23625
     },
     {
-      "epoch": 0.83,
-      "learning_rate": 4.7605890816491904e-05,
-      "loss": 0.2975,
+      "epoch": 0.851623598947634,
+      "grad_norm": 0.1816360354423523,
+      "learning_rate": 4.746426513287387e-05,
+      "loss": 0.4616,
       "step": 23630
     },
     {
-      "epoch": 0.83,
-      "learning_rate": 4.760467416926481e-05,
-      "loss": 0.3262,
+      "epoch": 0.8518037986088586,
+      "grad_norm": 0.1770544946193695,
+      "learning_rate": 4.746298441824889e-05,
+      "loss": 0.4234,
       "step": 23635
     },
     {
-      "epoch": 0.83,
-      "learning_rate": 4.76034572285309e-05,
-      "loss": 0.3111,
+      "epoch": 0.8519839982700832,
+      "grad_norm": 0.15067480504512787,
+      "learning_rate": 4.746170339757009e-05,
+      "loss": 0.4579,
       "step": 23640
     },
     {
-      "epoch": 0.83,
-      "learning_rate": 4.7602239994306e-05,
-      "loss": 0.2784,
+      "epoch": 0.8521641979313079,
+      "grad_norm": 0.12504181265830994,
+      "learning_rate": 4.746042207085492e-05,
+      "loss": 0.4412,
       "step": 23645
     },
     {
-      "epoch": 0.83,
-      "learning_rate": 4.7601022466605914e-05,
-      "loss": 0.2921,
+      "epoch": 0.8523443975925326,
+      "grad_norm": 0.1552511602640152,
+      "learning_rate": 4.745914043812085e-05,
+      "loss": 0.4427,
       "step": 23650
     },
     {
-      "epoch": 0.83,
-      "learning_rate": 4.759980464544645e-05,
-      "loss": 0.2982,
+      "epoch": 0.8525245972537572,
+      "grad_norm": 0.21108126640319824,
+      "learning_rate": 4.745785849938535e-05,
+      "loss": 0.4025,
       "step": 23655
     },
     {
-      "epoch": 0.83,
-      "learning_rate": 4.7598586530843405e-05,
-      "loss": 0.294,
+      "epoch": 0.8527047969149818,
+      "grad_norm": 0.17185276746749878,
+      "learning_rate": 4.7456576254665866e-05,
+      "loss": 0.4341,
       "step": 23660
     },
     {
-      "epoch": 0.83,
-      "learning_rate": 4.7597368122812607e-05,
-      "loss": 0.3147,
+      "epoch": 0.8528849965762064,
+      "grad_norm": 0.21345165371894836,
+      "learning_rate": 4.7455293703979884e-05,
+      "loss": 0.4513,
       "step": 23665
     },
     {
-      "epoch": 0.83,
-      "learning_rate": 4.7596149421369886e-05,
-      "loss": 0.3126,
+      "epoch": 0.853065196237431,
+      "grad_norm": 0.16104549169540405,
+      "learning_rate": 4.745401084734487e-05,
+      "loss": 0.4178,
       "step": 23670
     },
     {
-      "epoch": 0.83,
-      "learning_rate": 4.7594930426531046e-05,
-      "loss": 0.3015,
+      "epoch": 0.8532453958986557,
+      "grad_norm": 0.1738523244857788,
+      "learning_rate": 4.7452727684778305e-05,
+      "loss": 0.4372,
       "step": 23675
     },
     {
-      "epoch": 0.83,
-      "learning_rate": 4.759371113831193e-05,
-      "loss": 0.3026,
+      "epoch": 0.8534255955598804,
+      "grad_norm": 0.19414255023002625,
+      "learning_rate": 4.7451444216297674e-05,
+      "loss": 0.4444,
       "step": 23680
     },
     {
-      "epoch": 0.83,
-      "learning_rate": 4.759249155672837e-05,
-      "loss": 0.3211,
+      "epoch": 0.853605795221105,
+      "grad_norm": 0.16047124564647675,
+      "learning_rate": 4.7450160441920466e-05,
+      "loss": 0.4439,
       "step": 23685
     },
     {
-      "epoch": 0.83,
-      "learning_rate": 4.759127168179619e-05,
-      "loss": 0.2899,
+      "epoch": 0.8537859948823296,
+      "grad_norm": 0.1373186558485031,
+      "learning_rate": 4.744887636166416e-05,
+      "loss": 0.396,
       "step": 23690
     },
     {
-      "epoch": 0.83,
-      "learning_rate": 4.759005151353123e-05,
-      "loss": 0.3028,
+      "epoch": 0.8539661945435543,
+      "grad_norm": 0.15237957239151,
+      "learning_rate": 4.744759197554626e-05,
+      "loss": 0.4106,
       "step": 23695
     },
     {
-      "epoch": 0.83,
-      "learning_rate": 4.758883105194934e-05,
-      "loss": 0.3102,
+      "epoch": 0.8541463942047789,
+      "grad_norm": 0.18528686463832855,
+      "learning_rate": 4.744630728358427e-05,
+      "loss": 0.4458,
       "step": 23700
     },
     {
-      "epoch": 0.83,
-      "learning_rate": 4.758761029706637e-05,
-      "loss": 0.2816,
+      "epoch": 0.8543265938660035,
+      "grad_norm": 0.16544073820114136,
+      "learning_rate": 4.744502228579569e-05,
+      "loss": 0.4257,
       "step": 23705
     },
     {
-      "epoch": 0.83,
-      "learning_rate": 4.758638924889817e-05,
-      "loss": 0.2997,
+      "epoch": 0.8545067935272281,
+      "grad_norm": 0.16749577224254608,
+      "learning_rate": 4.7443736982198026e-05,
+      "loss": 0.4441,
       "step": 23710
     },
     {
-      "epoch": 0.83,
-      "learning_rate": 4.758516790746059e-05,
-      "loss": 0.3077,
+      "epoch": 0.8546869931884528,
+      "grad_norm": 0.19285382330417633,
+      "learning_rate": 4.7442451372808797e-05,
+      "loss": 0.4527,
       "step": 23715
     },
     {
-      "epoch": 0.83,
-      "learning_rate": 4.758394627276949e-05,
-      "loss": 0.3017,
+      "epoch": 0.8548671928496775,
+      "grad_norm": 0.17987944185733795,
+      "learning_rate": 4.74411654576455e-05,
+      "loss": 0.4205,
       "step": 23720
     },
     {
-      "epoch": 0.83,
-      "learning_rate": 4.7582724344840726e-05,
-      "loss": 0.2612,
+      "epoch": 0.8550473925109021,
+      "grad_norm": 0.14557349681854248,
+      "learning_rate": 4.7439879236725674e-05,
+      "loss": 0.4193,
       "step": 23725
     },
     {
-      "epoch": 0.83,
-      "learning_rate": 4.7581502123690165e-05,
-      "loss": 0.3053,
+      "epoch": 0.8552275921721267,
+      "grad_norm": 0.184067040681839,
+      "learning_rate": 4.743859271006684e-05,
+      "loss": 0.4426,
       "step": 23730
     },
     {
-      "epoch": 0.84,
-      "learning_rate": 4.758027960933368e-05,
-      "loss": 0.3,
+      "epoch": 0.8554077918333514,
+      "grad_norm": 0.14480531215667725,
+      "learning_rate": 4.743730587768652e-05,
+      "loss": 0.4561,
       "step": 23735
     },
     {
-      "epoch": 0.84,
-      "learning_rate": 4.757905680178715e-05,
-      "loss": 0.323,
+      "epoch": 0.855587991494576,
+      "grad_norm": 0.16141416132450104,
+      "learning_rate": 4.7436018739602255e-05,
+      "loss": 0.4452,
       "step": 23740
     },
     {
-      "epoch": 0.84,
-      "learning_rate": 4.757783370106646e-05,
-      "loss": 0.3073,
+      "epoch": 0.8557681911558006,
+      "grad_norm": 0.1394343227148056,
+      "learning_rate": 4.743473129583158e-05,
+      "loss": 0.4205,
       "step": 23745
     },
     {
-      "epoch": 0.84,
-      "learning_rate": 4.757661030718746e-05,
-      "loss": 0.3241,
+      "epoch": 0.8559483908170252,
+      "grad_norm": 0.18532565236091614,
+      "learning_rate": 4.743344354639203e-05,
+      "loss": 0.4093,
       "step": 23750
     },
     {
-      "epoch": 0.84,
-      "learning_rate": 4.757538662016606e-05,
-      "loss": 0.2923,
+      "epoch": 0.8561285904782499,
+      "grad_norm": 0.17660491168498993,
+      "learning_rate": 4.743215549130115e-05,
+      "loss": 0.4328,
       "step": 23755
     },
     {
-      "epoch": 0.84,
-      "learning_rate": 4.757416264001814e-05,
-      "loss": 0.3052,
+      "epoch": 0.8563087901394746,
+      "grad_norm": 0.20785756409168243,
+      "learning_rate": 4.743086713057651e-05,
+      "loss": 0.427,
       "step": 23760
     },
     {
-      "epoch": 0.84,
-      "learning_rate": 4.75729383667596e-05,
-      "loss": 0.2748,
+      "epoch": 0.8564889898006992,
+      "grad_norm": 0.19306261837482452,
+      "learning_rate": 4.7429578464235635e-05,
+      "loss": 0.4162,
       "step": 23765
     },
     {
-      "epoch": 0.84,
-      "learning_rate": 4.7571713800406324e-05,
-      "loss": 0.3155,
+      "epoch": 0.8566691894619238,
+      "grad_norm": 0.1779967099428177,
+      "learning_rate": 4.742828949229611e-05,
+      "loss": 0.4251,
       "step": 23770
     },
     {
-      "epoch": 0.84,
-      "learning_rate": 4.7570488940974236e-05,
-      "loss": 0.3109,
+      "epoch": 0.8568493891231485,
+      "grad_norm": 0.22586210072040558,
+      "learning_rate": 4.742700021477547e-05,
+      "loss": 0.4546,
       "step": 23775
     },
     {
-      "epoch": 0.84,
-      "learning_rate": 4.7569263788479204e-05,
-      "loss": 0.2879,
+      "epoch": 0.8570295887843731,
+      "grad_norm": 0.1746557056903839,
+      "learning_rate": 4.74257106316913e-05,
+      "loss": 0.3891,
       "step": 23780
     },
     {
-      "epoch": 0.84,
-      "learning_rate": 4.756803834293717e-05,
-      "loss": 0.3166,
+      "epoch": 0.8572097884455977,
+      "grad_norm": 0.1889764368534088,
+      "learning_rate": 4.742442074306116e-05,
+      "loss": 0.4304,
       "step": 23785
     },
     {
-      "epoch": 0.84,
-      "learning_rate": 4.756681260436402e-05,
-      "loss": 0.3123,
+      "epoch": 0.8573899881068223,
+      "grad_norm": 0.16799265146255493,
+      "learning_rate": 4.742313054890263e-05,
+      "loss": 0.3993,
       "step": 23790
     },
     {
-      "epoch": 0.84,
-      "learning_rate": 4.7565586572775686e-05,
-      "loss": 0.2884,
+      "epoch": 0.8575701877680469,
+      "grad_norm": 0.17759232223033905,
+      "learning_rate": 4.742184004923329e-05,
+      "loss": 0.4243,
       "step": 23795
     },
     {
-      "epoch": 0.84,
-      "learning_rate": 4.7564360248188086e-05,
-      "loss": 0.308,
+      "epoch": 0.8577503874292717,
+      "grad_norm": 0.16500405967235565,
+      "learning_rate": 4.742054924407072e-05,
+      "loss": 0.4649,
       "step": 23800
     },
     {
-      "epoch": 0.84,
-      "learning_rate": 4.756313363061713e-05,
-      "loss": 0.3314,
+      "epoch": 0.8579305870904963,
+      "grad_norm": 0.20370623469352722,
+      "learning_rate": 4.7419258133432504e-05,
+      "loss": 0.4699,
       "step": 23805
     },
     {
-      "epoch": 0.84,
-      "learning_rate": 4.756190672007876e-05,
-      "loss": 0.3089,
+      "epoch": 0.8581107867517209,
+      "grad_norm": 0.18750417232513428,
+      "learning_rate": 4.741796671733624e-05,
+      "loss": 0.4242,
       "step": 23810
     },
     {
-      "epoch": 0.84,
-      "learning_rate": 4.756067951658889e-05,
-      "loss": 0.3132,
+      "epoch": 0.8582909864129455,
+      "grad_norm": 0.1584675908088684,
+      "learning_rate": 4.741667499579952e-05,
+      "loss": 0.4649,
       "step": 23815
     },
     {
-      "epoch": 0.84,
-      "learning_rate": 4.755945202016348e-05,
-      "loss": 0.2933,
+      "epoch": 0.8584711860741702,
+      "grad_norm": 0.2016519159078598,
+      "learning_rate": 4.741538296883994e-05,
+      "loss": 0.4401,
       "step": 23820
     },
     {
-      "epoch": 0.84,
-      "learning_rate": 4.755822423081844e-05,
-      "loss": 0.2671,
+      "epoch": 0.8586513857353948,
+      "grad_norm": 0.14827871322631836,
+      "learning_rate": 4.741409063647511e-05,
+      "loss": 0.4315,
       "step": 23825
     },
     {
-      "epoch": 0.84,
-      "learning_rate": 4.755699614856973e-05,
-      "loss": 0.3004,
+      "epoch": 0.8588315853966194,
+      "grad_norm": 0.22020429372787476,
+      "learning_rate": 4.741279799872263e-05,
+      "loss": 0.4144,
       "step": 23830
     },
     {
-      "epoch": 0.84,
-      "learning_rate": 4.755576777343329e-05,
-      "loss": 0.2954,
+      "epoch": 0.8590117850578441,
+      "grad_norm": 0.15583039820194244,
+      "learning_rate": 4.7411505055600115e-05,
+      "loss": 0.4558,
       "step": 23835
     },
     {
-      "epoch": 0.84,
-      "learning_rate": 4.7554539105425075e-05,
-      "loss": 0.3193,
+      "epoch": 0.8591919847190688,
+      "grad_norm": 0.17358680069446564,
+      "learning_rate": 4.741021180712519e-05,
+      "loss": 0.441,
       "step": 23840
     },
     {
-      "epoch": 0.84,
-      "learning_rate": 4.7553310144561024e-05,
-      "loss": 0.3082,
+      "epoch": 0.8593721843802934,
+      "grad_norm": 0.2088773101568222,
+      "learning_rate": 4.7408918253315464e-05,
+      "loss": 0.4209,
       "step": 23845
     },
     {
-      "epoch": 0.84,
-      "learning_rate": 4.755208089085711e-05,
-      "loss": 0.2875,
+      "epoch": 0.859552384041518,
+      "grad_norm": 0.1951671689748764,
+      "learning_rate": 4.740762439418856e-05,
+      "loss": 0.4292,
       "step": 23850
     },
     {
-      "epoch": 0.84,
-      "learning_rate": 4.7550851344329286e-05,
-      "loss": 0.3087,
+      "epoch": 0.8597325837027426,
+      "grad_norm": 0.14672303199768066,
+      "learning_rate": 4.740633022976213e-05,
+      "loss": 0.4192,
       "step": 23855
     },
     {
-      "epoch": 0.84,
-      "learning_rate": 4.754962150499352e-05,
-      "loss": 0.3027,
+      "epoch": 0.8599127833639673,
+      "grad_norm": 0.21486572921276093,
+      "learning_rate": 4.740503576005377e-05,
+      "loss": 0.4734,
       "step": 23860
     },
     {
-      "epoch": 0.84,
-      "learning_rate": 4.7548391372865776e-05,
-      "loss": 0.3188,
+      "epoch": 0.8600929830251919,
+      "grad_norm": 0.15998758375644684,
+      "learning_rate": 4.740374098508115e-05,
+      "loss": 0.4318,
       "step": 23865
     },
     {
-      "epoch": 0.84,
-      "learning_rate": 4.7547160947962034e-05,
-      "loss": 0.3187,
+      "epoch": 0.8602731826864165,
+      "grad_norm": 0.17849770188331604,
+      "learning_rate": 4.740244590486188e-05,
+      "loss": 0.4296,
       "step": 23870
     },
     {
-      "epoch": 0.84,
-      "learning_rate": 4.754593023029827e-05,
-      "loss": 0.3014,
+      "epoch": 0.8604533823476412,
+      "grad_norm": 0.18155914545059204,
+      "learning_rate": 4.740115051941363e-05,
+      "loss": 0.4598,
       "step": 23875
     },
     {
-      "epoch": 0.84,
-      "learning_rate": 4.7544699219890454e-05,
-      "loss": 0.2919,
+      "epoch": 0.8606335820088659,
+      "grad_norm": 0.15962116420269012,
+      "learning_rate": 4.7399854828754045e-05,
+      "loss": 0.417,
       "step": 23880
     },
     {
-      "epoch": 0.84,
-      "learning_rate": 4.754346791675458e-05,
-      "loss": 0.3071,
+      "epoch": 0.8608137816700905,
+      "grad_norm": 0.15650032460689545,
+      "learning_rate": 4.7398558832900774e-05,
+      "loss": 0.4187,
       "step": 23885
     },
     {
-      "epoch": 0.84,
-      "learning_rate": 4.754223632090663e-05,
-      "loss": 0.3336,
+      "epoch": 0.8609939813313151,
+      "grad_norm": 0.1548149287700653,
+      "learning_rate": 4.7397262531871466e-05,
+      "loss": 0.4128,
       "step": 23890
     },
     {
-      "epoch": 0.84,
-      "learning_rate": 4.7541004432362596e-05,
-      "loss": 0.3384,
+      "epoch": 0.8611741809925397,
+      "grad_norm": 0.1666049212217331,
+      "learning_rate": 4.739596592568381e-05,
+      "loss": 0.4565,
       "step": 23895
     },
     {
-      "epoch": 0.84,
-      "learning_rate": 4.753977225113847e-05,
-      "loss": 0.305,
+      "epoch": 0.8613543806537644,
+      "grad_norm": 0.14310970902442932,
+      "learning_rate": 4.7394669014355444e-05,
+      "loss": 0.4583,
       "step": 23900
     },
     {
-      "epoch": 0.84,
-      "learning_rate": 4.7538539777250266e-05,
-      "loss": 0.3004,
+      "epoch": 0.861534580314989,
+      "grad_norm": 0.15406273305416107,
+      "learning_rate": 4.739337179790404e-05,
+      "loss": 0.4391,
       "step": 23905
     },
     {
-      "epoch": 0.84,
-      "learning_rate": 4.753730701071397e-05,
-      "loss": 0.3281,
+      "epoch": 0.8617147799762136,
+      "grad_norm": 0.15325191617012024,
+      "learning_rate": 4.739207427634729e-05,
+      "loss": 0.4413,
       "step": 23910
     },
     {
-      "epoch": 0.84,
-      "learning_rate": 4.75360739515456e-05,
-      "loss": 0.2932,
+      "epoch": 0.8618949796374383,
+      "grad_norm": 0.14947649836540222,
+      "learning_rate": 4.7390776449702864e-05,
+      "loss": 0.439,
       "step": 23915
     },
     {
-      "epoch": 0.84,
-      "learning_rate": 4.7534840599761154e-05,
-      "loss": 0.3116,
+      "epoch": 0.862075179298663,
+      "grad_norm": 0.16361840069293976,
+      "learning_rate": 4.738947831798844e-05,
+      "loss": 0.4578,
       "step": 23920
     },
     {
-      "epoch": 0.84,
-      "learning_rate": 4.7533606955376665e-05,
-      "loss": 0.3204,
+      "epoch": 0.8622553789598876,
+      "grad_norm": 0.20598135888576508,
+      "learning_rate": 4.738817988122171e-05,
+      "loss": 0.4678,
       "step": 23925
     },
     {
-      "epoch": 0.84,
-      "learning_rate": 4.7532373018408125e-05,
-      "loss": 0.3119,
+      "epoch": 0.8624355786211122,
+      "grad_norm": 0.17498494684696198,
+      "learning_rate": 4.738688113942036e-05,
+      "loss": 0.4866,
       "step": 23930
     },
     {
-      "epoch": 0.84,
-      "learning_rate": 4.753113878887158e-05,
-      "loss": 0.2632,
+      "epoch": 0.8626157782823368,
+      "grad_norm": 0.1920240968465805,
+      "learning_rate": 4.73855820926021e-05,
+      "loss": 0.4274,
       "step": 23935
     },
     {
-      "epoch": 0.84,
-      "learning_rate": 4.752990426678304e-05,
-      "loss": 0.3219,
+      "epoch": 0.8627959779435614,
+      "grad_norm": 0.14575999975204468,
+      "learning_rate": 4.738428274078461e-05,
+      "loss": 0.4185,
       "step": 23940
     },
     {
-      "epoch": 0.84,
-      "learning_rate": 4.7528669452158554e-05,
-      "loss": 0.3126,
+      "epoch": 0.8629761776047861,
+      "grad_norm": 0.14960631728172302,
+      "learning_rate": 4.73829830839856e-05,
+      "loss": 0.4148,
       "step": 23945
     },
     {
-      "epoch": 0.84,
-      "learning_rate": 4.752743434501413e-05,
-      "loss": 0.3073,
+      "epoch": 0.8631563772660107,
+      "grad_norm": 0.15745064616203308,
+      "learning_rate": 4.738168312222278e-05,
+      "loss": 0.4097,
       "step": 23950
     },
     {
-      "epoch": 0.84,
-      "learning_rate": 4.752619894536582e-05,
-      "loss": 0.2941,
+      "epoch": 0.8633365769272354,
+      "grad_norm": 0.1760377436876297,
+      "learning_rate": 4.738038285551386e-05,
+      "loss": 0.4678,
       "step": 23955
     },
     {
-      "epoch": 0.84,
-      "learning_rate": 4.752496325322967e-05,
-      "loss": 0.3093,
+      "epoch": 0.86351677658846,
+      "grad_norm": 0.15912814438343048,
+      "learning_rate": 4.7379082283876566e-05,
+      "loss": 0.4585,
       "step": 23960
     },
     {
-      "epoch": 0.84,
-      "learning_rate": 4.752372726862171e-05,
-      "loss": 0.3013,
+      "epoch": 0.8636969762496847,
+      "grad_norm": 0.13206642866134644,
+      "learning_rate": 4.737778140732859e-05,
+      "loss": 0.4157,
       "step": 23965
     },
     {
-      "epoch": 0.84,
-      "learning_rate": 4.752249099155799e-05,
-      "loss": 0.2963,
+      "epoch": 0.8638771759109093,
+      "grad_norm": 0.21865518391132355,
+      "learning_rate": 4.737648022588769e-05,
+      "loss": 0.4373,
       "step": 23970
     },
     {
-      "epoch": 0.84,
-      "learning_rate": 4.752125442205457e-05,
-      "loss": 0.3125,
+      "epoch": 0.8640573755721339,
+      "grad_norm": 0.15692706406116486,
+      "learning_rate": 4.737517873957158e-05,
+      "loss": 0.3932,
       "step": 23975
     },
     {
-      "epoch": 0.84,
-      "learning_rate": 4.752001756012751e-05,
-      "loss": 0.3002,
+      "epoch": 0.8642375752333585,
+      "grad_norm": 0.19991809129714966,
+      "learning_rate": 4.737387694839798e-05,
+      "loss": 0.4778,
       "step": 23980
     },
     {
-      "epoch": 0.84,
-      "learning_rate": 4.751878040579286e-05,
-      "loss": 0.3147,
+      "epoch": 0.8644177748945832,
+      "grad_norm": 0.1880597621202469,
+      "learning_rate": 4.737257485238465e-05,
+      "loss": 0.3943,
       "step": 23985
     },
     {
-      "epoch": 0.84,
-      "learning_rate": 4.7517542959066684e-05,
-      "loss": 0.3145,
+      "epoch": 0.8645979745558078,
+      "grad_norm": 0.21935416758060455,
+      "learning_rate": 4.737127245154931e-05,
+      "loss": 0.4207,
       "step": 23990
     },
     {
-      "epoch": 0.84,
-      "learning_rate": 4.751630521996505e-05,
-      "loss": 0.2935,
+      "epoch": 0.8647781742170325,
+      "grad_norm": 0.16751424968242645,
+      "learning_rate": 4.736996974590972e-05,
+      "loss": 0.4191,
       "step": 23995
     },
     {
-      "epoch": 0.84,
-      "learning_rate": 4.751506718850404e-05,
-      "loss": 0.3306,
+      "epoch": 0.8649583738782571,
+      "grad_norm": 0.18504785001277924,
+      "learning_rate": 4.736866673548362e-05,
+      "loss": 0.4088,
       "step": 24000
     },
     {
-      "epoch": 0.84,
-      "eval_loss": 0.2971259653568268,
-      "eval_runtime": 10.5298,
-      "eval_samples_per_second": 9.497,
-      "eval_steps_per_second": 9.497,
+      "epoch": 0.8649583738782571,
+      "eval_loss": 0.45828837156295776,
+      "eval_runtime": 3.5369,
+      "eval_samples_per_second": 28.273,
+      "eval_steps_per_second": 7.068,
       "step": 24000
     },
     {
-      "epoch": 0.84,
-      "learning_rate": 4.751382886469972e-05,
-      "loss": 0.3354,
+      "epoch": 0.8651385735394818,
+      "grad_norm": 0.18134254217147827,
+      "learning_rate": 4.7367363420288765e-05,
+      "loss": 0.4821,
       "step": 24005
     },
     {
-      "epoch": 0.84,
-      "learning_rate": 4.7512590248568163e-05,
-      "loss": 0.2793,
+      "epoch": 0.8653187732007064,
+      "grad_norm": 0.15396596491336823,
+      "learning_rate": 4.736605980034292e-05,
+      "loss": 0.433,
       "step": 24010
     },
     {
-      "epoch": 0.84,
-      "learning_rate": 4.7511351340125456e-05,
-      "loss": 0.3002,
+      "epoch": 0.865498972861931,
+      "grad_norm": 0.1897013932466507,
+      "learning_rate": 4.7364755875663834e-05,
+      "loss": 0.4416,
       "step": 24015
     },
     {
-      "epoch": 0.85,
-      "learning_rate": 4.75101121393877e-05,
-      "loss": 0.2878,
+      "epoch": 0.8656791725231556,
+      "grad_norm": 0.15314961969852448,
+      "learning_rate": 4.736345164626929e-05,
+      "loss": 0.4014,
       "step": 24020
     },
     {
-      "epoch": 0.85,
-      "learning_rate": 4.750887264637096e-05,
-      "loss": 0.2873,
+      "epoch": 0.8658593721843802,
+      "grad_norm": 0.17612630128860474,
+      "learning_rate": 4.736214711217703e-05,
+      "loss": 0.4435,
       "step": 24025
     },
     {
-      "epoch": 0.85,
-      "learning_rate": 4.750763286109134e-05,
-      "loss": 0.3029,
+      "epoch": 0.866039571845605,
+      "grad_norm": 0.15653395652770996,
+      "learning_rate": 4.736084227340486e-05,
+      "loss": 0.4473,
       "step": 24030
     },
     {
-      "epoch": 0.85,
-      "learning_rate": 4.750639278356496e-05,
-      "loss": 0.317,
+      "epoch": 0.8662197715068296,
+      "grad_norm": 0.14282876253128052,
+      "learning_rate": 4.735953712997053e-05,
+      "loss": 0.4484,
       "step": 24035
     },
     {
-      "epoch": 0.85,
-      "learning_rate": 4.7505152413807886e-05,
-      "loss": 0.2921,
+      "epoch": 0.8663999711680542,
+      "grad_norm": 0.184425950050354,
+      "learning_rate": 4.7358231681891855e-05,
+      "loss": 0.4453,
       "step": 24040
     },
     {
-      "epoch": 0.85,
-      "learning_rate": 4.750391175183624e-05,
-      "loss": 0.3101,
+      "epoch": 0.8665801708292789,
+      "grad_norm": 0.1802368015050888,
+      "learning_rate": 4.735692592918658e-05,
+      "loss": 0.4524,
       "step": 24045
     },
     {
-      "epoch": 0.85,
-      "learning_rate": 4.750267079766614e-05,
-      "loss": 0.315,
+      "epoch": 0.8667603704905035,
+      "grad_norm": 0.15805204212665558,
+      "learning_rate": 4.735561987187253e-05,
+      "loss": 0.4139,
       "step": 24050
     },
     {
-      "epoch": 0.85,
-      "learning_rate": 4.750142955131368e-05,
-      "loss": 0.3072,
+      "epoch": 0.8669405701517281,
+      "grad_norm": 0.16449551284313202,
+      "learning_rate": 4.7354313509967486e-05,
+      "loss": 0.4236,
       "step": 24055
     },
     {
-      "epoch": 0.85,
-      "learning_rate": 4.7500188012794985e-05,
-      "loss": 0.3223,
+      "epoch": 0.8671207698129527,
+      "grad_norm": 0.1771911382675171,
+      "learning_rate": 4.735300684348925e-05,
+      "loss": 0.4665,
       "step": 24060
     },
     {
-      "epoch": 0.85,
-      "learning_rate": 4.749894618212618e-05,
-      "loss": 0.2978,
+      "epoch": 0.8673009694741773,
+      "grad_norm": 0.1740954965353012,
+      "learning_rate": 4.735169987245561e-05,
+      "loss": 0.4394,
       "step": 24065
     },
     {
-      "epoch": 0.85,
-      "learning_rate": 4.749770405932339e-05,
-      "loss": 0.2927,
+      "epoch": 0.8674811691354021,
+      "grad_norm": 0.18734583258628845,
+      "learning_rate": 4.735039259688441e-05,
+      "loss": 0.4227,
       "step": 24070
     },
     {
-      "epoch": 0.85,
-      "learning_rate": 4.749646164440273e-05,
-      "loss": 0.2951,
+      "epoch": 0.8676613687966267,
+      "grad_norm": 0.1404682993888855,
+      "learning_rate": 4.734908501679342e-05,
+      "loss": 0.4358,
       "step": 24075
     },
     {
-      "epoch": 0.85,
-      "learning_rate": 4.7495218937380356e-05,
-      "loss": 0.3082,
+      "epoch": 0.8678415684578513,
+      "grad_norm": 0.16821545362472534,
+      "learning_rate": 4.7347777132200475e-05,
+      "loss": 0.4389,
       "step": 24080
     },
     {
-      "epoch": 0.85,
-      "learning_rate": 4.7493975938272376e-05,
-      "loss": 0.3136,
+      "epoch": 0.8680217681190759,
+      "grad_norm": 0.17715495824813843,
+      "learning_rate": 4.73464689431234e-05,
+      "loss": 0.4148,
       "step": 24085
     },
     {
-      "epoch": 0.85,
-      "learning_rate": 4.749273264709495e-05,
-      "loss": 0.2949,
+      "epoch": 0.8682019677803006,
+      "grad_norm": 0.16599716246128082,
+      "learning_rate": 4.734516044958001e-05,
+      "loss": 0.4243,
       "step": 24090
     },
     {
-      "epoch": 0.85,
-      "learning_rate": 4.749148906386421e-05,
-      "loss": 0.2751,
+      "epoch": 0.8683821674415252,
+      "grad_norm": 0.15897536277770996,
+      "learning_rate": 4.7343851651588137e-05,
+      "loss": 0.431,
       "step": 24095
     },
     {
-      "epoch": 0.85,
-      "learning_rate": 4.749024518859631e-05,
-      "loss": 0.3013,
+      "epoch": 0.8685623671027498,
+      "grad_norm": 0.14617212116718292,
+      "learning_rate": 4.73425425491656e-05,
+      "loss": 0.4395,
       "step": 24100
     },
     {
-      "epoch": 0.85,
-      "learning_rate": 4.74890010213074e-05,
-      "loss": 0.3143,
+      "epoch": 0.8687425667639744,
+      "grad_norm": 0.1834208071231842,
+      "learning_rate": 4.734123314233026e-05,
+      "loss": 0.4274,
       "step": 24105
     },
     {
-      "epoch": 0.85,
-      "learning_rate": 4.748775656201363e-05,
-      "loss": 0.3332,
+      "epoch": 0.8689227664251992,
+      "grad_norm": 0.22381344437599182,
+      "learning_rate": 4.733992343109994e-05,
+      "loss": 0.4279,
       "step": 24110
     },
     {
-      "epoch": 0.85,
-      "learning_rate": 4.748651181073115e-05,
-      "loss": 0.2902,
+      "epoch": 0.8691029660864238,
+      "grad_norm": 0.14666973054409027,
+      "learning_rate": 4.7338613415492486e-05,
+      "loss": 0.4255,
       "step": 24115
     },
     {
-      "epoch": 0.85,
-      "learning_rate": 4.748526676747614e-05,
-      "loss": 0.2741,
+      "epoch": 0.8692831657476484,
+      "grad_norm": 0.17257143557071686,
+      "learning_rate": 4.733730309552575e-05,
+      "loss": 0.4548,
       "step": 24120
     },
     {
-      "epoch": 0.85,
-      "learning_rate": 4.748402143226477e-05,
-      "loss": 0.3071,
+      "epoch": 0.869463365408873,
+      "grad_norm": 0.17522400617599487,
+      "learning_rate": 4.733599247121758e-05,
+      "loss": 0.4354,
       "step": 24125
     },
     {
-      "epoch": 0.85,
-      "learning_rate": 4.748277580511319e-05,
-      "loss": 0.3265,
+      "epoch": 0.8696435650700977,
+      "grad_norm": 0.19729764759540558,
+      "learning_rate": 4.733468154258585e-05,
+      "loss": 0.46,
       "step": 24130
     },
     {
-      "epoch": 0.85,
-      "learning_rate": 4.748152988603758e-05,
-      "loss": 0.3057,
+      "epoch": 0.8698237647313223,
+      "grad_norm": 0.15954478085041046,
+      "learning_rate": 4.733337030964839e-05,
+      "loss": 0.4451,
       "step": 24135
     },
     {
-      "epoch": 0.85,
-      "learning_rate": 4.748028367505413e-05,
-      "loss": 0.3049,
+      "epoch": 0.8700039643925469,
+      "grad_norm": 0.16600678861141205,
+      "learning_rate": 4.73320587724231e-05,
+      "loss": 0.4397,
       "step": 24140
     },
     {
-      "epoch": 0.85,
-      "learning_rate": 4.747903717217901e-05,
-      "loss": 0.3015,
+      "epoch": 0.8701841640537715,
+      "grad_norm": 0.16135752201080322,
+      "learning_rate": 4.733074693092783e-05,
+      "loss": 0.4045,
       "step": 24145
     },
     {
-      "epoch": 0.85,
-      "learning_rate": 4.74777903774284e-05,
-      "loss": 0.3069,
+      "epoch": 0.8703643637149963,
+      "grad_norm": 0.16556213796138763,
+      "learning_rate": 4.732943478518045e-05,
+      "loss": 0.4408,
       "step": 24150
     },
     {
-      "epoch": 0.85,
-      "learning_rate": 4.74765432908185e-05,
-      "loss": 0.3236,
+      "epoch": 0.8705445633762209,
+      "grad_norm": 0.15862923860549927,
+      "learning_rate": 4.732812233519884e-05,
+      "loss": 0.4239,
       "step": 24155
     },
     {
-      "epoch": 0.85,
-      "learning_rate": 4.747529591236549e-05,
-      "loss": 0.3162,
+      "epoch": 0.8707247630374455,
+      "grad_norm": 0.19938628375530243,
+      "learning_rate": 4.73268095810009e-05,
+      "loss": 0.4735,
       "step": 24160
     },
     {
-      "epoch": 0.85,
-      "learning_rate": 4.7474048242085585e-05,
-      "loss": 0.2927,
+      "epoch": 0.8709049626986701,
+      "grad_norm": 0.19490042328834534,
+      "learning_rate": 4.732549652260449e-05,
+      "loss": 0.4184,
       "step": 24165
     },
     {
-      "epoch": 0.85,
-      "learning_rate": 4.7472800279994964e-05,
-      "loss": 0.3169,
+      "epoch": 0.8710851623598947,
+      "grad_norm": 0.19054916501045227,
+      "learning_rate": 4.732418316002751e-05,
+      "loss": 0.4391,
       "step": 24170
     },
     {
-      "epoch": 0.85,
-      "learning_rate": 4.7471552026109855e-05,
-      "loss": 0.3284,
+      "epoch": 0.8712653620211194,
+      "grad_norm": 0.15895889699459076,
+      "learning_rate": 4.732286949328787e-05,
+      "loss": 0.448,
       "step": 24175
     },
     {
-      "epoch": 0.85,
-      "learning_rate": 4.7470303480446446e-05,
-      "loss": 0.3145,
+      "epoch": 0.871445561682344,
+      "grad_norm": 0.16004562377929688,
+      "learning_rate": 4.732155552240345e-05,
+      "loss": 0.3978,
       "step": 24180
     },
     {
-      "epoch": 0.85,
-      "learning_rate": 4.746905464302095e-05,
-      "loss": 0.2979,
+      "epoch": 0.8716257613435687,
+      "grad_norm": 0.2268204391002655,
+      "learning_rate": 4.732024124739215e-05,
+      "loss": 0.4128,
       "step": 24185
     },
     {
-      "epoch": 0.85,
-      "learning_rate": 4.74678055138496e-05,
-      "loss": 0.308,
+      "epoch": 0.8718059610047934,
+      "grad_norm": 0.18399815261363983,
+      "learning_rate": 4.731892666827189e-05,
+      "loss": 0.4377,
       "step": 24190
     },
     {
-      "epoch": 0.85,
-      "learning_rate": 4.746655609294858e-05,
-      "loss": 0.3221,
+      "epoch": 0.871986160666018,
+      "grad_norm": 0.13044115900993347,
+      "learning_rate": 4.731761178506058e-05,
+      "loss": 0.4663,
       "step": 24195
     },
     {
-      "epoch": 0.85,
-      "learning_rate": 4.746530638033415e-05,
-      "loss": 0.3295,
+      "epoch": 0.8721663603272426,
+      "grad_norm": 0.1856122612953186,
+      "learning_rate": 4.7316296597776123e-05,
+      "loss": 0.4289,
       "step": 24200
     },
     {
-      "epoch": 0.85,
-      "learning_rate": 4.746405637602252e-05,
-      "loss": 0.3207,
+      "epoch": 0.8723465599884672,
+      "grad_norm": 0.17390163242816925,
+      "learning_rate": 4.731498110643645e-05,
+      "loss": 0.4633,
       "step": 24205
     },
     {
-      "epoch": 0.85,
-      "learning_rate": 4.746280608002992e-05,
-      "loss": 0.3052,
+      "epoch": 0.8725267596496918,
+      "grad_norm": 0.15870732069015503,
+      "learning_rate": 4.731366531105947e-05,
+      "loss": 0.436,
       "step": 24210
     },
     {
-      "epoch": 0.85,
-      "learning_rate": 4.746155549237259e-05,
-      "loss": 0.2973,
+      "epoch": 0.8727069593109165,
+      "grad_norm": 0.18578395247459412,
+      "learning_rate": 4.731234921166313e-05,
+      "loss": 0.4246,
       "step": 24215
     },
     {
-      "epoch": 0.85,
-      "learning_rate": 4.746030461306676e-05,
-      "loss": 0.2646,
+      "epoch": 0.8728871589721411,
+      "grad_norm": 0.19250373542308807,
+      "learning_rate": 4.7311032808265356e-05,
+      "loss": 0.4511,
       "step": 24220
     },
     {
-      "epoch": 0.85,
-      "learning_rate": 4.745905344212869e-05,
-      "loss": 0.3076,
+      "epoch": 0.8730673586333658,
+      "grad_norm": 0.18087449669837952,
+      "learning_rate": 4.730971610088407e-05,
+      "loss": 0.4365,
       "step": 24225
     },
     {
-      "epoch": 0.85,
-      "learning_rate": 4.7457801979574604e-05,
-      "loss": 0.2814,
+      "epoch": 0.8732475582945904,
+      "grad_norm": 0.15141281485557556,
+      "learning_rate": 4.7308399089537224e-05,
+      "loss": 0.4196,
       "step": 24230
     },
     {
-      "epoch": 0.85,
-      "learning_rate": 4.745655022542075e-05,
-      "loss": 0.3061,
+      "epoch": 0.8734277579558151,
+      "grad_norm": 0.16130012273788452,
+      "learning_rate": 4.730708177424276e-05,
+      "loss": 0.4339,
       "step": 24235
     },
     {
-      "epoch": 0.85,
-      "learning_rate": 4.74552981796834e-05,
-      "loss": 0.3201,
+      "epoch": 0.8736079576170397,
+      "grad_norm": 0.21587517857551575,
+      "learning_rate": 4.730576415501863e-05,
+      "loss": 0.4417,
       "step": 24240
     },
     {
-      "epoch": 0.85,
-      "learning_rate": 4.7454045842378806e-05,
-      "loss": 0.2877,
+      "epoch": 0.8737881572782643,
+      "grad_norm": 0.1740473508834839,
+      "learning_rate": 4.730444623188278e-05,
+      "loss": 0.4073,
       "step": 24245
     },
     {
-      "epoch": 0.85,
-      "learning_rate": 4.745279321352321e-05,
-      "loss": 0.2999,
+      "epoch": 0.8739683569394889,
+      "grad_norm": 0.1535811573266983,
+      "learning_rate": 4.730312800485316e-05,
+      "loss": 0.4328,
       "step": 24250
     },
     {
-      "epoch": 0.85,
-      "learning_rate": 4.74515402931329e-05,
-      "loss": 0.292,
+      "epoch": 0.8741485566007136,
+      "grad_norm": 0.20786762237548828,
+      "learning_rate": 4.7301809473947744e-05,
+      "loss": 0.4224,
       "step": 24255
     },
     {
-      "epoch": 0.85,
-      "learning_rate": 4.745028708122414e-05,
-      "loss": 0.2876,
+      "epoch": 0.8743287562619382,
+      "grad_norm": 0.19472654163837433,
+      "learning_rate": 4.730049063918449e-05,
+      "loss": 0.4272,
       "step": 24260
     },
     {
-      "epoch": 0.85,
-      "learning_rate": 4.744903357781319e-05,
-      "loss": 0.3125,
+      "epoch": 0.8745089559231629,
+      "grad_norm": 0.2173738032579422,
+      "learning_rate": 4.729917150058137e-05,
+      "loss": 0.4626,
       "step": 24265
     },
     {
-      "epoch": 0.85,
-      "learning_rate": 4.7447779782916333e-05,
-      "loss": 0.3157,
+      "epoch": 0.8746891555843875,
+      "grad_norm": 0.2047669142484665,
+      "learning_rate": 4.729785205815637e-05,
+      "loss": 0.4195,
       "step": 24270
     },
     {
-      "epoch": 0.85,
-      "learning_rate": 4.7446525696549855e-05,
-      "loss": 0.3172,
+      "epoch": 0.8748693552456122,
+      "grad_norm": 0.19127419590950012,
+      "learning_rate": 4.7296532311927436e-05,
+      "loss": 0.4646,
       "step": 24275
     },
     {
-      "epoch": 0.85,
-      "learning_rate": 4.744527131873003e-05,
-      "loss": 0.2945,
+      "epoch": 0.8750495549068368,
+      "grad_norm": 0.22150792181491852,
+      "learning_rate": 4.729521226191257e-05,
+      "loss": 0.426,
       "step": 24280
     },
     {
-      "epoch": 0.85,
-      "learning_rate": 4.744401664947315e-05,
-      "loss": 0.2963,
+      "epoch": 0.8752297545680614,
+      "grad_norm": 0.2369442582130432,
+      "learning_rate": 4.729389190812975e-05,
+      "loss": 0.4241,
       "step": 24285
     },
     {
-      "epoch": 0.85,
-      "learning_rate": 4.744276168879551e-05,
-      "loss": 0.2887,
+      "epoch": 0.875409954229286,
+      "grad_norm": 0.17270110547542572,
+      "learning_rate": 4.729257125059697e-05,
+      "loss": 0.4575,
       "step": 24290
     },
     {
-      "epoch": 0.85,
-      "learning_rate": 4.744150643671339e-05,
-      "loss": 0.2818,
+      "epoch": 0.8755901538905106,
+      "grad_norm": 0.16564151644706726,
+      "learning_rate": 4.729125028933222e-05,
+      "loss": 0.4041,
       "step": 24295
     },
     {
-      "epoch": 0.85,
-      "learning_rate": 4.7440250893243104e-05,
-      "loss": 0.3011,
+      "epoch": 0.8757703535517353,
+      "grad_norm": 0.1850225180387497,
+      "learning_rate": 4.728992902435351e-05,
+      "loss": 0.4068,
       "step": 24300
     },
     {
-      "epoch": 0.86,
-      "learning_rate": 4.743899505840095e-05,
-      "loss": 0.3359,
+      "epoch": 0.87595055321296,
+      "grad_norm": 0.17433954775333405,
+      "learning_rate": 4.728860745567883e-05,
+      "loss": 0.4035,
       "step": 24305
     },
     {
-      "epoch": 0.86,
-      "learning_rate": 4.743773893220323e-05,
-      "loss": 0.2925,
+      "epoch": 0.8761307528741846,
+      "grad_norm": 0.1604575216770172,
+      "learning_rate": 4.728728558332618e-05,
+      "loss": 0.4102,
       "step": 24310
     },
     {
-      "epoch": 0.86,
-      "learning_rate": 4.743648251466626e-05,
-      "loss": 0.2962,
+      "epoch": 0.8763109525354092,
+      "grad_norm": 0.18941442668437958,
+      "learning_rate": 4.7285963407313594e-05,
+      "loss": 0.4188,
       "step": 24315
     },
     {
-      "epoch": 0.86,
-      "learning_rate": 4.743522580580634e-05,
-      "loss": 0.3224,
+      "epoch": 0.8764911521966339,
+      "grad_norm": 0.15453235805034637,
+      "learning_rate": 4.728464092765906e-05,
+      "loss": 0.4138,
       "step": 24320
     },
     {
-      "epoch": 0.86,
-      "learning_rate": 4.743396880563981e-05,
-      "loss": 0.3107,
+      "epoch": 0.8766713518578585,
+      "grad_norm": 0.16290608048439026,
+      "learning_rate": 4.7283318144380606e-05,
+      "loss": 0.4676,
       "step": 24325
     },
     {
-      "epoch": 0.86,
-      "learning_rate": 4.7432711514182976e-05,
-      "loss": 0.3132,
+      "epoch": 0.8768515515190831,
+      "grad_norm": 0.15486080944538116,
+      "learning_rate": 4.728199505749626e-05,
+      "loss": 0.4694,
       "step": 24330
     },
     {
-      "epoch": 0.86,
-      "learning_rate": 4.743145393145217e-05,
-      "loss": 0.3075,
+      "epoch": 0.8770317511803077,
+      "grad_norm": 0.14686860144138336,
+      "learning_rate": 4.728067166702404e-05,
+      "loss": 0.4268,
       "step": 24335
     },
     {
-      "epoch": 0.86,
-      "learning_rate": 4.743019605746372e-05,
-      "loss": 0.2577,
+      "epoch": 0.8772119508415325,
+      "grad_norm": 0.18616507947444916,
+      "learning_rate": 4.7279347972982e-05,
+      "loss": 0.4561,
       "step": 24340
     },
     {
-      "epoch": 0.86,
-      "learning_rate": 4.742893789223394e-05,
-      "loss": 0.3243,
+      "epoch": 0.8773921505027571,
+      "grad_norm": 0.17040805518627167,
+      "learning_rate": 4.727802397538814e-05,
+      "loss": 0.4399,
       "step": 24345
     },
     {
-      "epoch": 0.86,
-      "learning_rate": 4.7427679435779196e-05,
-      "loss": 0.2945,
+      "epoch": 0.8775723501639817,
+      "grad_norm": 0.18918024003505707,
+      "learning_rate": 4.7276699674260525e-05,
+      "loss": 0.4592,
       "step": 24350
     },
     {
-      "epoch": 0.86,
-      "learning_rate": 4.7426420688115816e-05,
-      "loss": 0.3102,
+      "epoch": 0.8777525498252063,
+      "grad_norm": 0.1710633933544159,
+      "learning_rate": 4.727537506961719e-05,
+      "loss": 0.4546,
       "step": 24355
     },
     {
-      "epoch": 0.86,
-      "learning_rate": 4.742516164926014e-05,
-      "loss": 0.3217,
+      "epoch": 0.877932749486431,
+      "grad_norm": 0.16070812940597534,
+      "learning_rate": 4.727405016147618e-05,
+      "loss": 0.443,
       "step": 24360
     },
     {
-      "epoch": 0.86,
-      "learning_rate": 4.742390231922852e-05,
-      "loss": 0.2823,
+      "epoch": 0.8781129491476556,
+      "grad_norm": 0.20280031859874725,
+      "learning_rate": 4.727272494985554e-05,
+      "loss": 0.4375,
       "step": 24365
     },
     {
-      "epoch": 0.86,
-      "learning_rate": 4.74226426980373e-05,
-      "loss": 0.3064,
+      "epoch": 0.8782931488088802,
+      "grad_norm": 0.1605139523744583,
+      "learning_rate": 4.7271399434773345e-05,
+      "loss": 0.44,
       "step": 24370
     },
     {
-      "epoch": 0.86,
-      "learning_rate": 4.742138278570285e-05,
-      "loss": 0.3351,
+      "epoch": 0.8784733484701048,
+      "grad_norm": 0.19736407697200775,
+      "learning_rate": 4.7270073616247646e-05,
+      "loss": 0.4579,
       "step": 24375
     },
     {
-      "epoch": 0.86,
-      "learning_rate": 4.7420122582241514e-05,
-      "loss": 0.341,
+      "epoch": 0.8786535481313296,
+      "grad_norm": 0.19877927005290985,
+      "learning_rate": 4.72687474942965e-05,
+      "loss": 0.4602,
       "step": 24380
     },
     {
-      "epoch": 0.86,
-      "learning_rate": 4.741886208766966e-05,
-      "loss": 0.3188,
+      "epoch": 0.8788337477925542,
+      "grad_norm": 0.16131383180618286,
+      "learning_rate": 4.7267421068937984e-05,
+      "loss": 0.4178,
       "step": 24385
     },
     {
-      "epoch": 0.86,
-      "learning_rate": 4.741760130200367e-05,
-      "loss": 0.3149,
+      "epoch": 0.8790139474537788,
+      "grad_norm": 0.19206862151622772,
+      "learning_rate": 4.7266094340190165e-05,
+      "loss": 0.4558,
       "step": 24390
     },
     {
-      "epoch": 0.86,
-      "learning_rate": 4.7416340225259886e-05,
-      "loss": 0.3037,
+      "epoch": 0.8791941471150034,
+      "grad_norm": 0.20475110411643982,
+      "learning_rate": 4.7264767308071126e-05,
+      "loss": 0.417,
       "step": 24395
     },
     {
-      "epoch": 0.86,
-      "learning_rate": 4.7415078857454704e-05,
-      "loss": 0.3078,
+      "epoch": 0.879374346776228,
+      "grad_norm": 0.21035943925380707,
+      "learning_rate": 4.726343997259893e-05,
+      "loss": 0.4433,
       "step": 24400
     },
     {
-      "epoch": 0.86,
-      "learning_rate": 4.7413817198604496e-05,
-      "loss": 0.287,
+      "epoch": 0.8795545464374527,
+      "grad_norm": 0.18233457207679749,
+      "learning_rate": 4.7262112333791685e-05,
+      "loss": 0.3937,
       "step": 24405
     },
     {
-      "epoch": 0.86,
-      "learning_rate": 4.7412555248725646e-05,
-      "loss": 0.295,
+      "epoch": 0.8797347460986773,
+      "grad_norm": 0.18523730337619781,
+      "learning_rate": 4.7260784391667475e-05,
+      "loss": 0.4456,
       "step": 24410
     },
     {
-      "epoch": 0.86,
-      "learning_rate": 4.7411293007834525e-05,
-      "loss": 0.3377,
+      "epoch": 0.8799149457599019,
+      "grad_norm": 0.19751974940299988,
+      "learning_rate": 4.725945614624438e-05,
+      "loss": 0.4533,
       "step": 24415
     },
     {
-      "epoch": 0.86,
-      "learning_rate": 4.741003047594754e-05,
-      "loss": 0.2956,
+      "epoch": 0.8800951454211267,
+      "grad_norm": 0.15248170495033264,
+      "learning_rate": 4.7258127597540505e-05,
+      "loss": 0.4397,
       "step": 24420
     },
     {
-      "epoch": 0.86,
-      "learning_rate": 4.740876765308107e-05,
-      "loss": 0.3259,
+      "epoch": 0.8802753450823513,
+      "grad_norm": 0.1679372489452362,
+      "learning_rate": 4.725679874557395e-05,
+      "loss": 0.4419,
       "step": 24425
     },
     {
-      "epoch": 0.86,
-      "learning_rate": 4.740750453925153e-05,
-      "loss": 0.2787,
+      "epoch": 0.8804555447435759,
+      "grad_norm": 0.256967157125473,
+      "learning_rate": 4.7255469590362825e-05,
+      "loss": 0.432,
       "step": 24430
     },
     {
-      "epoch": 0.86,
-      "learning_rate": 4.7406241134475316e-05,
-      "loss": 0.3286,
+      "epoch": 0.8806357444048005,
+      "grad_norm": 0.18858453631401062,
+      "learning_rate": 4.725414013192523e-05,
+      "loss": 0.4169,
       "step": 24435
     },
     {
-      "epoch": 0.86,
-      "learning_rate": 4.740497743876882e-05,
-      "loss": 0.3364,
+      "epoch": 0.8808159440660251,
+      "grad_norm": 0.18355509638786316,
+      "learning_rate": 4.725281037027929e-05,
+      "loss": 0.4668,
       "step": 24440
     },
     {
-      "epoch": 0.86,
-      "learning_rate": 4.7403713452148454e-05,
-      "loss": 0.2879,
+      "epoch": 0.8809961437272498,
+      "grad_norm": 0.16345731914043427,
+      "learning_rate": 4.725148030544311e-05,
+      "loss": 0.4466,
       "step": 24445
     },
     {
-      "epoch": 0.86,
-      "learning_rate": 4.740244917463064e-05,
-      "loss": 0.3261,
+      "epoch": 0.8811763433884744,
+      "grad_norm": 0.19114868342876434,
+      "learning_rate": 4.7250149937434826e-05,
+      "loss": 0.3905,
       "step": 24450
     },
     {
-      "epoch": 0.86,
-      "learning_rate": 4.7401184606231786e-05,
-      "loss": 0.2984,
+      "epoch": 0.881356543049699,
+      "grad_norm": 0.19460555911064148,
+      "learning_rate": 4.724881926627255e-05,
+      "loss": 0.437,
       "step": 24455
     },
     {
-      "epoch": 0.86,
-      "learning_rate": 4.7399919746968316e-05,
-      "loss": 0.3239,
+      "epoch": 0.8815367427109237,
+      "grad_norm": Infinity,
+      "learning_rate": 4.7247754511084054e-05,
+      "loss": 0.4409,
       "step": 24460
     },
     {
-      "epoch": 0.86,
-      "learning_rate": 4.739865459685664e-05,
-      "loss": 0.2893,
+      "epoch": 0.8817169423721484,
+      "grad_norm": 0.14812225103378296,
+      "learning_rate": 4.72464232942903e-05,
+      "loss": 0.3929,
       "step": 24465
     },
     {
-      "epoch": 0.86,
-      "learning_rate": 4.7397389155913197e-05,
-      "loss": 0.3265,
+      "epoch": 0.881897142033373,
+      "grad_norm": 0.1727660894393921,
+      "learning_rate": 4.724509177439333e-05,
+      "loss": 0.4262,
       "step": 24470
     },
     {
-      "epoch": 0.86,
-      "learning_rate": 4.7396123424154426e-05,
-      "loss": 0.2947,
+      "epoch": 0.8820773416945976,
+      "grad_norm": 0.16415224969387054,
+      "learning_rate": 4.724375995141129e-05,
+      "loss": 0.4162,
       "step": 24475
     },
     {
-      "epoch": 0.86,
-      "learning_rate": 4.739485740159674e-05,
-      "loss": 0.2713,
+      "epoch": 0.8822575413558222,
+      "grad_norm": 0.16160134971141815,
+      "learning_rate": 4.724242782536234e-05,
+      "loss": 0.4698,
       "step": 24480
     },
     {
-      "epoch": 0.86,
-      "learning_rate": 4.739359108825659e-05,
-      "loss": 0.318,
+      "epoch": 0.8824377410170469,
+      "grad_norm": 0.15631291270256042,
+      "learning_rate": 4.724109539626461e-05,
+      "loss": 0.4455,
       "step": 24485
     },
     {
-      "epoch": 0.86,
-      "learning_rate": 4.739232448415043e-05,
-      "loss": 0.3234,
+      "epoch": 0.8826179406782715,
+      "grad_norm": 0.16617485880851746,
+      "learning_rate": 4.7239762664136264e-05,
+      "loss": 0.4638,
       "step": 24490
     },
     {
-      "epoch": 0.86,
-      "learning_rate": 4.739105758929468e-05,
-      "loss": 0.3022,
+      "epoch": 0.8827981403394961,
+      "grad_norm": 0.1969350427389145,
+      "learning_rate": 4.7238429628995456e-05,
+      "loss": 0.4458,
       "step": 24495
     },
     {
-      "epoch": 0.86,
-      "learning_rate": 4.738979040370582e-05,
-      "loss": 0.3255,
+      "epoch": 0.8829783400007208,
+      "grad_norm": 0.12761616706848145,
+      "learning_rate": 4.723709629086035e-05,
+      "loss": 0.4276,
       "step": 24500
     },
     {
-      "epoch": 0.86,
-      "eval_loss": 0.297638863325119,
-      "eval_runtime": 10.5374,
-      "eval_samples_per_second": 9.49,
-      "eval_steps_per_second": 9.49,
+      "epoch": 0.8829783400007208,
+      "eval_loss": 0.4570537507534027,
+      "eval_runtime": 3.5263,
+      "eval_samples_per_second": 28.359,
+      "eval_steps_per_second": 7.09,
       "step": 24500
     },
     {
-      "epoch": 0.86,
-      "learning_rate": 4.7388522927400276e-05,
-      "loss": 0.2881,
+      "epoch": 0.8831585396619455,
+      "grad_norm": 0.20359250903129578,
+      "learning_rate": 4.723576264974911e-05,
+      "loss": 0.4593,
       "step": 24505
     },
     {
-      "epoch": 0.86,
-      "learning_rate": 4.7387255160394514e-05,
-      "loss": 0.32,
+      "epoch": 0.8833387393231701,
+      "grad_norm": 0.1633104681968689,
+      "learning_rate": 4.723442870567991e-05,
+      "loss": 0.4538,
       "step": 24510
     },
     {
-      "epoch": 0.86,
-      "learning_rate": 4.7385987102705e-05,
-      "loss": 0.3053,
+      "epoch": 0.8835189389843947,
+      "grad_norm": 0.19287440180778503,
+      "learning_rate": 4.7233094458670926e-05,
+      "loss": 0.4838,
       "step": 24515
     },
     {
-      "epoch": 0.86,
-      "learning_rate": 4.73847187543482e-05,
-      "loss": 0.294,
+      "epoch": 0.8836991386456193,
+      "grad_norm": 0.16194848716259003,
+      "learning_rate": 4.723175990874034e-05,
+      "loss": 0.4247,
       "step": 24520
     },
     {
-      "epoch": 0.86,
-      "learning_rate": 4.738345011534058e-05,
-      "loss": 0.2906,
+      "epoch": 0.883879338306844,
+      "grad_norm": 0.15081371366977692,
+      "learning_rate": 4.723042505590631e-05,
+      "loss": 0.4416,
       "step": 24525
     },
     {
-      "epoch": 0.86,
-      "learning_rate": 4.738218118569861e-05,
-      "loss": 0.2912,
+      "epoch": 0.8840595379680686,
+      "grad_norm": 0.15278756618499756,
+      "learning_rate": 4.7229089900187065e-05,
+      "loss": 0.4031,
       "step": 24530
     },
     {
-      "epoch": 0.86,
-      "learning_rate": 4.738091196543876e-05,
-      "loss": 0.2966,
+      "epoch": 0.8842397376292933,
+      "grad_norm": 0.19165587425231934,
+      "learning_rate": 4.722775444160076e-05,
+      "loss": 0.4159,
       "step": 24535
     },
     {
-      "epoch": 0.86,
-      "learning_rate": 4.737964245457752e-05,
-      "loss": 0.3212,
+      "epoch": 0.8844199372905179,
+      "grad_norm": 0.17740246653556824,
+      "learning_rate": 4.722641868016561e-05,
+      "loss": 0.4277,
       "step": 24540
     },
     {
-      "epoch": 0.86,
-      "learning_rate": 4.7378372653131375e-05,
-      "loss": 0.3061,
+      "epoch": 0.8846001369517426,
+      "grad_norm": 0.14334651827812195,
+      "learning_rate": 4.72250826158998e-05,
+      "loss": 0.4596,
       "step": 24545
     },
     {
-      "epoch": 0.86,
-      "learning_rate": 4.7377102561116816e-05,
-      "loss": 0.3129,
+      "epoch": 0.8847803366129672,
+      "grad_norm": 0.1924460381269455,
+      "learning_rate": 4.722374624882155e-05,
+      "loss": 0.4452,
       "step": 24550
     },
     {
-      "epoch": 0.86,
-      "learning_rate": 4.737583217855032e-05,
-      "loss": 0.3163,
+      "epoch": 0.8849605362741918,
+      "grad_norm": 0.18541020154953003,
+      "learning_rate": 4.7222409578949054e-05,
+      "loss": 0.4299,
       "step": 24555
     },
     {
-      "epoch": 0.86,
-      "learning_rate": 4.7374561505448384e-05,
-      "loss": 0.3168,
+      "epoch": 0.8851407359354164,
+      "grad_norm": 0.14057400822639465,
+      "learning_rate": 4.7221072606300543e-05,
+      "loss": 0.4235,
       "step": 24560
     },
     {
-      "epoch": 0.86,
-      "learning_rate": 4.737329054182752e-05,
-      "loss": 0.3179,
+      "epoch": 0.885320935596641,
+      "grad_norm": 0.17242777347564697,
+      "learning_rate": 4.721973533089421e-05,
+      "loss": 0.4452,
       "step": 24565
     },
     {
-      "epoch": 0.86,
-      "learning_rate": 4.737201928770422e-05,
-      "loss": 0.2977,
+      "epoch": 0.8855011352578657,
+      "grad_norm": 0.14792689681053162,
+      "learning_rate": 4.721866529259576e-05,
+      "loss": 0.4112,
       "step": 24570
     },
     {
-      "epoch": 0.86,
-      "learning_rate": 4.737074774309499e-05,
-      "loss": 0.2845,
+      "epoch": 0.8856813349190904,
+      "grad_norm": 0.20297929644584656,
+      "learning_rate": 4.7217327472271283e-05,
+      "loss": 0.4249,
       "step": 24575
     },
     {
-      "epoch": 0.86,
-      "learning_rate": 4.736947590801635e-05,
-      "loss": 0.2973,
+      "epoch": 0.885861534580315,
+      "grad_norm": 0.19619841873645782,
+      "learning_rate": 4.7215989349240026e-05,
+      "loss": 0.4288,
       "step": 24580
     },
     {
-      "epoch": 0.86,
-      "learning_rate": 4.73682037824848e-05,
-      "loss": 0.3149,
+      "epoch": 0.8860417342415396,
+      "grad_norm": 0.15129470825195312,
+      "learning_rate": 4.721465092352021e-05,
+      "loss": 0.4335,
       "step": 24585
     },
     {
-      "epoch": 0.87,
-      "learning_rate": 4.7366931366516876e-05,
-      "loss": 0.3216,
+      "epoch": 0.8862219339027643,
+      "grad_norm": 0.18301598727703094,
+      "learning_rate": 4.7213312195130076e-05,
+      "loss": 0.4433,
       "step": 24590
     },
     {
-      "epoch": 0.87,
-      "learning_rate": 4.736565866012908e-05,
-      "loss": 0.3094,
+      "epoch": 0.8864021335639889,
+      "grad_norm": 0.16560180485248566,
+      "learning_rate": 4.721197316408787e-05,
+      "loss": 0.4086,
       "step": 24595
     },
     {
-      "epoch": 0.87,
-      "learning_rate": 4.736438566333795e-05,
-      "loss": 0.2993,
+      "epoch": 0.8865823332252135,
+      "grad_norm": 0.1662699282169342,
+      "learning_rate": 4.721063383041182e-05,
+      "loss": 0.4438,
       "step": 24600
     },
     {
-      "epoch": 0.87,
-      "learning_rate": 4.736311237616001e-05,
-      "loss": 0.3066,
+      "epoch": 0.8867625328864381,
+      "grad_norm": 0.2002149224281311,
+      "learning_rate": 4.720929419412019e-05,
+      "loss": 0.4215,
       "step": 24605
     },
     {
-      "epoch": 0.87,
-      "learning_rate": 4.736183879861179e-05,
-      "loss": 0.2802,
+      "epoch": 0.8869427325476628,
+      "grad_norm": 0.15055419504642487,
+      "learning_rate": 4.720795425523122e-05,
+      "loss": 0.4198,
       "step": 24610
     },
     {
-      "epoch": 0.87,
-      "learning_rate": 4.736056493070983e-05,
-      "loss": 0.2907,
+      "epoch": 0.8871229322088875,
+      "grad_norm": 0.18256331980228424,
+      "learning_rate": 4.720661401376318e-05,
+      "loss": 0.4618,
       "step": 24615
     },
     {
-      "epoch": 0.87,
-      "learning_rate": 4.7359290772470675e-05,
-      "loss": 0.312,
+      "epoch": 0.8873031318701121,
+      "grad_norm": 0.15107060968875885,
+      "learning_rate": 4.7205273469734325e-05,
+      "loss": 0.4164,
       "step": 24620
     },
     {
-      "epoch": 0.87,
-      "learning_rate": 4.735801632391087e-05,
-      "loss": 0.2899,
+      "epoch": 0.8874833315313367,
+      "grad_norm": 0.16118554770946503,
+      "learning_rate": 4.7203932623162917e-05,
+      "loss": 0.4657,
       "step": 24625
     },
     {
-      "epoch": 0.87,
-      "learning_rate": 4.735674158504695e-05,
-      "loss": 0.3086,
+      "epoch": 0.8876635311925614,
+      "grad_norm": 0.1474160999059677,
+      "learning_rate": 4.720259147406722e-05,
+      "loss": 0.4628,
       "step": 24630
     },
     {
-      "epoch": 0.87,
-      "learning_rate": 4.735546655589547e-05,
-      "loss": 0.2949,
+      "epoch": 0.887843730853786,
+      "grad_norm": 0.18264853954315186,
+      "learning_rate": 4.720125002246552e-05,
+      "loss": 0.4558,
       "step": 24635
     },
     {
-      "epoch": 0.87,
-      "learning_rate": 4.7354191236472995e-05,
-      "loss": 0.3053,
+      "epoch": 0.8880239305150106,
+      "grad_norm": 0.2043534368276596,
+      "learning_rate": 4.719990826837608e-05,
+      "loss": 0.4473,
       "step": 24640
     },
     {
-      "epoch": 0.87,
-      "learning_rate": 4.735291562679608e-05,
-      "loss": 0.3149,
+      "epoch": 0.8882041301762352,
+      "grad_norm": 0.1613204926252365,
+      "learning_rate": 4.719856621181719e-05,
+      "loss": 0.4262,
       "step": 24645
     },
     {
-      "epoch": 0.87,
-      "learning_rate": 4.735163972688128e-05,
-      "loss": 0.2791,
+      "epoch": 0.8883843298374599,
+      "grad_norm": 0.13441753387451172,
+      "learning_rate": 4.7197223852807136e-05,
+      "loss": 0.4354,
       "step": 24650
     },
     {
-      "epoch": 0.87,
-      "learning_rate": 4.7350363536745174e-05,
-      "loss": 0.2994,
+      "epoch": 0.8885645294986846,
+      "grad_norm": 0.221405029296875,
+      "learning_rate": 4.71958811913642e-05,
+      "loss": 0.4739,
       "step": 24655
     },
     {
-      "epoch": 0.87,
-      "learning_rate": 4.734908705640433e-05,
-      "loss": 0.3109,
+      "epoch": 0.8887447291599092,
+      "grad_norm": 0.18559223413467407,
+      "learning_rate": 4.719453822750669e-05,
+      "loss": 0.4838,
       "step": 24660
     },
     {
-      "epoch": 0.87,
-      "learning_rate": 4.7347810285875315e-05,
-      "loss": 0.3036,
+      "epoch": 0.8889249288211338,
+      "grad_norm": 0.16596554219722748,
+      "learning_rate": 4.7193194961252885e-05,
+      "loss": 0.4106,
       "step": 24665
     },
     {
-      "epoch": 0.87,
-      "learning_rate": 4.7346533225174705e-05,
-      "loss": 0.2824,
+      "epoch": 0.8891051284823585,
+      "grad_norm": 0.18130573630332947,
+      "learning_rate": 4.7191851392621086e-05,
+      "loss": 0.4411,
       "step": 24670
     },
     {
-      "epoch": 0.87,
-      "learning_rate": 4.7345255874319096e-05,
-      "loss": 0.3066,
+      "epoch": 0.8892853281435831,
+      "grad_norm": 0.19347991049289703,
+      "learning_rate": 4.719050752162962e-05,
+      "loss": 0.4076,
       "step": 24675
     },
     {
-      "epoch": 0.87,
-      "learning_rate": 4.7343978233325064e-05,
-      "loss": 0.3002,
+      "epoch": 0.8894655278048077,
+      "grad_norm": 0.18345235288143158,
+      "learning_rate": 4.7189163348296794e-05,
+      "loss": 0.3907,
       "step": 24680
     },
     {
-      "epoch": 0.87,
-      "learning_rate": 4.7342700302209195e-05,
-      "loss": 0.3059,
+      "epoch": 0.8896457274660323,
+      "grad_norm": 0.18901300430297852,
+      "learning_rate": 4.71878188726409e-05,
+      "loss": 0.4677,
       "step": 24685
     },
     {
-      "epoch": 0.87,
-      "learning_rate": 4.7341422080988094e-05,
-      "loss": 0.2977,
+      "epoch": 0.889825927127257,
+      "grad_norm": 0.17101964354515076,
+      "learning_rate": 4.718647409468028e-05,
+      "loss": 0.4077,
       "step": 24690
     },
     {
-      "epoch": 0.87,
-      "learning_rate": 4.734014356967835e-05,
-      "loss": 0.3186,
+      "epoch": 0.8900061267884817,
+      "grad_norm": 0.2040674090385437,
+      "learning_rate": 4.7185129014433234e-05,
+      "loss": 0.4288,
       "step": 24695
     },
     {
-      "epoch": 0.87,
-      "learning_rate": 4.733886476829656e-05,
-      "loss": 0.2897,
+      "epoch": 0.8901863264497063,
+      "grad_norm": 0.13120241463184357,
+      "learning_rate": 4.71837836319181e-05,
+      "loss": 0.4134,
       "step": 24700
     },
     {
-      "epoch": 0.87,
-      "learning_rate": 4.7337585676859344e-05,
-      "loss": 0.2943,
+      "epoch": 0.8903665261109309,
+      "grad_norm": 0.18854455649852753,
+      "learning_rate": 4.7182437947153216e-05,
+      "loss": 0.4561,
       "step": 24705
     },
     {
-      "epoch": 0.87,
-      "learning_rate": 4.7336306295383285e-05,
-      "loss": 0.3145,
+      "epoch": 0.8905467257721555,
+      "grad_norm": 0.14998573064804077,
+      "learning_rate": 4.718109196015691e-05,
+      "loss": 0.4087,
       "step": 24710
     },
     {
-      "epoch": 0.87,
-      "learning_rate": 4.7335026623885015e-05,
-      "loss": 0.3302,
+      "epoch": 0.8907269254333802,
+      "grad_norm": 0.1941816210746765,
+      "learning_rate": 4.717974567094752e-05,
+      "loss": 0.4604,
       "step": 24715
     },
     {
-      "epoch": 0.87,
-      "learning_rate": 4.733374666238115e-05,
-      "loss": 0.2812,
+      "epoch": 0.8909071250946048,
+      "grad_norm": 0.16004863381385803,
+      "learning_rate": 4.7178399079543386e-05,
+      "loss": 0.4399,
       "step": 24720
     },
     {
-      "epoch": 0.87,
-      "learning_rate": 4.733246641088829e-05,
-      "loss": 0.2879,
+      "epoch": 0.8910873247558294,
+      "grad_norm": 0.192369744181633,
+      "learning_rate": 4.717705218596286e-05,
+      "loss": 0.4439,
       "step": 24725
     },
     {
-      "epoch": 0.87,
-      "learning_rate": 4.7331185869423075e-05,
-      "loss": 0.3062,
+      "epoch": 0.8912675244170541,
+      "grad_norm": 0.2026439905166626,
+      "learning_rate": 4.717570499022429e-05,
+      "loss": 0.4401,
       "step": 24730
     },
     {
-      "epoch": 0.87,
-      "learning_rate": 4.7329905038002136e-05,
-      "loss": 0.285,
+      "epoch": 0.8914477240782788,
+      "grad_norm": 0.1843568980693817,
+      "learning_rate": 4.7174357492346035e-05,
+      "loss": 0.4492,
       "step": 24735
     },
     {
-      "epoch": 0.87,
-      "learning_rate": 4.732862391664209e-05,
-      "loss": 0.3035,
+      "epoch": 0.8916279237395034,
+      "grad_norm": 0.17085200548171997,
+      "learning_rate": 4.717300969234645e-05,
+      "loss": 0.4401,
       "step": 24740
     },
     {
-      "epoch": 0.87,
-      "learning_rate": 4.732734250535959e-05,
-      "loss": 0.2817,
+      "epoch": 0.891808123400728,
+      "grad_norm": 0.15291930735111237,
+      "learning_rate": 4.717166159024391e-05,
+      "loss": 0.3833,
       "step": 24745
     },
     {
-      "epoch": 0.87,
-      "learning_rate": 4.732606080417124e-05,
-      "loss": 0.3166,
+      "epoch": 0.8919883230619526,
+      "grad_norm": 0.21157534420490265,
+      "learning_rate": 4.717031318605676e-05,
+      "loss": 0.4876,
       "step": 24750
     },
     {
-      "epoch": 0.87,
-      "learning_rate": 4.732477881309372e-05,
-      "loss": 0.3061,
+      "epoch": 0.8921685227231773,
+      "grad_norm": 0.1500784456729889,
+      "learning_rate": 4.716896447980339e-05,
+      "loss": 0.4665,
       "step": 24755
     },
     {
-      "epoch": 0.87,
-      "learning_rate": 4.732349653214365e-05,
-      "loss": 0.3076,
+      "epoch": 0.8923487223844019,
+      "grad_norm": 0.20307470858097076,
+      "learning_rate": 4.716761547150218e-05,
+      "loss": 0.4211,
       "step": 24760
     },
     {
-      "epoch": 0.87,
-      "learning_rate": 4.732221396133769e-05,
-      "loss": 0.3195,
+      "epoch": 0.8925289220456265,
+      "grad_norm": 0.21808555722236633,
+      "learning_rate": 4.716626616117149e-05,
+      "loss": 0.4722,
       "step": 24765
     },
     {
-      "epoch": 0.87,
-      "learning_rate": 4.73209311006925e-05,
-      "loss": 0.3247,
+      "epoch": 0.8927091217068512,
+      "grad_norm": 0.16036725044250488,
+      "learning_rate": 4.7164916548829716e-05,
+      "loss": 0.4207,
       "step": 24770
     },
     {
-      "epoch": 0.87,
-      "learning_rate": 4.731964795022473e-05,
-      "loss": 0.3167,
+      "epoch": 0.8928893213680759,
+      "grad_norm": 0.1857818365097046,
+      "learning_rate": 4.716356663449525e-05,
+      "loss": 0.4423,
       "step": 24775
     },
     {
-      "epoch": 0.87,
-      "learning_rate": 4.731836450995103e-05,
-      "loss": 0.3,
+      "epoch": 0.8930695210293005,
+      "grad_norm": 0.16408823430538177,
+      "learning_rate": 4.716221641818648e-05,
+      "loss": 0.3968,
       "step": 24780
     },
     {
-      "epoch": 0.87,
-      "learning_rate": 4.7317080779888076e-05,
-      "loss": 0.2758,
+      "epoch": 0.8932497206905251,
+      "grad_norm": 0.1698572039604187,
+      "learning_rate": 4.716086589992179e-05,
+      "loss": 0.4458,
       "step": 24785
     },
     {
-      "epoch": 0.87,
-      "learning_rate": 4.7315796760052544e-05,
-      "loss": 0.3273,
+      "epoch": 0.8934299203517497,
+      "grad_norm": 0.17390035092830658,
+      "learning_rate": 4.7159515079719606e-05,
+      "loss": 0.4407,
       "step": 24790
     },
     {
-      "epoch": 0.87,
-      "learning_rate": 4.73145124504611e-05,
-      "loss": 0.2699,
+      "epoch": 0.8936101200129744,
+      "grad_norm": 0.321679025888443,
+      "learning_rate": 4.715816395759832e-05,
+      "loss": 0.4106,
       "step": 24795
     },
     {
-      "epoch": 0.87,
-      "learning_rate": 4.731322785113041e-05,
-      "loss": 0.285,
+      "epoch": 0.893790319674199,
+      "grad_norm": 0.14596271514892578,
+      "learning_rate": 4.715681253357633e-05,
+      "loss": 0.4319,
       "step": 24800
     },
     {
-      "epoch": 0.87,
-      "learning_rate": 4.7311942962077164e-05,
-      "loss": 0.3223,
+      "epoch": 0.8939705193354236,
+      "grad_norm": 0.1981964260339737,
+      "learning_rate": 4.715546080767207e-05,
+      "loss": 0.4163,
       "step": 24805
     },
     {
-      "epoch": 0.87,
-      "learning_rate": 4.731065778331805e-05,
-      "loss": 0.3058,
+      "epoch": 0.8941507189966483,
+      "grad_norm": 0.13960348069667816,
+      "learning_rate": 4.715410877990394e-05,
+      "loss": 0.4762,
       "step": 24810
     },
     {
-      "epoch": 0.87,
-      "learning_rate": 4.730937231486974e-05,
-      "loss": 0.2904,
+      "epoch": 0.894330918657873,
+      "grad_norm": 0.13228678703308105,
+      "learning_rate": 4.7152756450290365e-05,
+      "loss": 0.4419,
       "step": 24815
     },
     {
-      "epoch": 0.87,
-      "learning_rate": 4.7308086556748926e-05,
-      "loss": 0.2698,
+      "epoch": 0.8945111183190976,
+      "grad_norm": 0.13322246074676514,
+      "learning_rate": 4.715140381884977e-05,
+      "loss": 0.4458,
       "step": 24820
     },
     {
-      "epoch": 0.87,
-      "learning_rate": 4.730680050897231e-05,
-      "loss": 0.3074,
+      "epoch": 0.8946913179803222,
+      "grad_norm": 0.190599724650383,
+      "learning_rate": 4.715005088560059e-05,
+      "loss": 0.4297,
       "step": 24825
     },
     {
-      "epoch": 0.87,
-      "learning_rate": 4.730551417155661e-05,
-      "loss": 0.2964,
+      "epoch": 0.8948715176415468,
+      "grad_norm": 0.2033384144306183,
+      "learning_rate": 4.714869765056126e-05,
+      "loss": 0.4869,
       "step": 24830
     },
     {
-      "epoch": 0.87,
-      "learning_rate": 4.730422754451849e-05,
-      "loss": 0.2897,
+      "epoch": 0.8950517173027714,
+      "grad_norm": 0.1728903204202652,
+      "learning_rate": 4.714734411375021e-05,
+      "loss": 0.4397,
       "step": 24835
     },
     {
-      "epoch": 0.87,
-      "learning_rate": 4.730294062787468e-05,
-      "loss": 0.3129,
+      "epoch": 0.8952319169639961,
+      "grad_norm": 0.1565089374780655,
+      "learning_rate": 4.714599027518588e-05,
+      "loss": 0.4149,
       "step": 24840
     },
     {
-      "epoch": 0.87,
-      "learning_rate": 4.730165342164188e-05,
-      "loss": 0.3175,
+      "epoch": 0.8954121166252208,
+      "grad_norm": 0.14530473947525024,
+      "learning_rate": 4.7144636134886725e-05,
+      "loss": 0.4479,
       "step": 24845
     },
     {
-      "epoch": 0.87,
-      "learning_rate": 4.730036592583682e-05,
-      "loss": 0.2899,
+      "epoch": 0.8955923162864454,
+      "grad_norm": 0.18022358417510986,
+      "learning_rate": 4.714328169287119e-05,
+      "loss": 0.4322,
       "step": 24850
     },
     {
-      "epoch": 0.87,
-      "learning_rate": 4.7299078140476195e-05,
-      "loss": 0.3284,
+      "epoch": 0.89577251594767,
+      "grad_norm": 0.20143041014671326,
+      "learning_rate": 4.714192694915772e-05,
+      "loss": 0.4065,
       "step": 24855
     },
     {
-      "epoch": 0.87,
-      "learning_rate": 4.7297790065576734e-05,
-      "loss": 0.3055,
+      "epoch": 0.8959527156088947,
+      "grad_norm": 0.1598571240901947,
+      "learning_rate": 4.7140571903764796e-05,
+      "loss": 0.4355,
       "step": 24860
     },
     {
-      "epoch": 0.87,
-      "learning_rate": 4.729650170115517e-05,
-      "loss": 0.3146,
+      "epoch": 0.8961329152701193,
+      "grad_norm": 0.14883951842784882,
+      "learning_rate": 4.713921655671086e-05,
+      "loss": 0.4303,
       "step": 24865
     },
     {
-      "epoch": 0.87,
-      "learning_rate": 4.729521304722823e-05,
-      "loss": 0.307,
+      "epoch": 0.8963131149313439,
+      "grad_norm": 0.22746653854846954,
+      "learning_rate": 4.713786090801438e-05,
+      "loss": 0.414,
       "step": 24870
     },
     {
-      "epoch": 0.88,
-      "learning_rate": 4.7293924103812626e-05,
-      "loss": 0.2946,
+      "epoch": 0.8964933145925685,
+      "grad_norm": 0.1794048696756363,
+      "learning_rate": 4.713650495769384e-05,
+      "loss": 0.441,
       "step": 24875
     },
     {
-      "epoch": 0.88,
-      "learning_rate": 4.7292634870925126e-05,
-      "loss": 0.3226,
+      "epoch": 0.8966735142537932,
+      "grad_norm": 0.1713542640209198,
+      "learning_rate": 4.713514870576769e-05,
+      "loss": 0.4191,
       "step": 24880
     },
     {
-      "epoch": 0.88,
-      "learning_rate": 4.729134534858244e-05,
-      "loss": 0.3063,
+      "epoch": 0.8968537139150179,
+      "grad_norm": 0.189683198928833,
+      "learning_rate": 4.713379215225444e-05,
+      "loss": 0.4236,
       "step": 24885
     },
     {
-      "epoch": 0.88,
-      "learning_rate": 4.729005553680133e-05,
-      "loss": 0.315,
+      "epoch": 0.8970339135762425,
+      "grad_norm": 0.13822448253631592,
+      "learning_rate": 4.713243529717256e-05,
+      "loss": 0.4481,
       "step": 24890
     },
     {
-      "epoch": 0.88,
-      "learning_rate": 4.7288765435598536e-05,
-      "loss": 0.3197,
+      "epoch": 0.8972141132374671,
+      "grad_norm": 0.1470467448234558,
+      "learning_rate": 4.713107814054052e-05,
+      "loss": 0.4471,
       "step": 24895
     },
     {
-      "epoch": 0.88,
-      "learning_rate": 4.7287475044990815e-05,
-      "loss": 0.306,
+      "epoch": 0.8973943128986918,
+      "grad_norm": 0.23228789865970612,
+      "learning_rate": 4.7129720682376835e-05,
+      "loss": 0.4273,
       "step": 24900
     },
     {
-      "epoch": 0.88,
-      "learning_rate": 4.7286184364994924e-05,
-      "loss": 0.3047,
+      "epoch": 0.8975745125599164,
+      "grad_norm": 0.16581259667873383,
+      "learning_rate": 4.712836292269999e-05,
+      "loss": 0.4401,
       "step": 24905
     },
     {
-      "epoch": 0.88,
-      "learning_rate": 4.728489339562761e-05,
-      "loss": 0.3094,
+      "epoch": 0.897754712221141,
+      "grad_norm": 0.20045721530914307,
+      "learning_rate": 4.712700486152848e-05,
+      "loss": 0.4381,
       "step": 24910
     },
     {
-      "epoch": 0.88,
-      "learning_rate": 4.728360213690564e-05,
-      "loss": 0.3052,
+      "epoch": 0.8979349118823656,
+      "grad_norm": 0.1778937131166458,
+      "learning_rate": 4.712564649888081e-05,
+      "loss": 0.4466,
       "step": 24915
     },
     {
-      "epoch": 0.88,
-      "learning_rate": 4.7282310588845786e-05,
-      "loss": 0.2992,
+      "epoch": 0.8981151115435902,
+      "grad_norm": 0.22372958064079285,
+      "learning_rate": 4.7124287834775496e-05,
+      "loss": 0.4296,
       "step": 24920
     },
     {
-      "epoch": 0.88,
-      "learning_rate": 4.7281018751464814e-05,
-      "loss": 0.3046,
+      "epoch": 0.898295311204815,
+      "grad_norm": 0.15579721331596375,
+      "learning_rate": 4.7122928869231044e-05,
+      "loss": 0.4306,
       "step": 24925
     },
     {
-      "epoch": 0.88,
-      "learning_rate": 4.727972662477949e-05,
-      "loss": 0.2983,
+      "epoch": 0.8984755108660396,
+      "grad_norm": 0.1593790054321289,
+      "learning_rate": 4.712156960226597e-05,
+      "loss": 0.4518,
       "step": 24930
     },
     {
-      "epoch": 0.88,
-      "learning_rate": 4.727843420880661e-05,
-      "loss": 0.318,
+      "epoch": 0.8986557105272642,
+      "grad_norm": 0.18624834716320038,
+      "learning_rate": 4.7120210033898784e-05,
+      "loss": 0.4181,
       "step": 24935
     },
     {
-      "epoch": 0.88,
-      "learning_rate": 4.727714150356294e-05,
-      "loss": 0.2807,
+      "epoch": 0.8988359101884889,
+      "grad_norm": 0.19123771786689758,
+      "learning_rate": 4.711885016414802e-05,
+      "loss": 0.4106,
       "step": 24940
     },
     {
-      "epoch": 0.88,
-      "learning_rate": 4.727584850906526e-05,
-      "loss": 0.3074,
+      "epoch": 0.8990161098497135,
+      "grad_norm": 0.1681036353111267,
+      "learning_rate": 4.7117489993032216e-05,
+      "loss": 0.4661,
       "step": 24945
     },
     {
-      "epoch": 0.88,
-      "learning_rate": 4.7274555225330374e-05,
-      "loss": 0.2902,
+      "epoch": 0.8991963095109381,
+      "grad_norm": 0.16487166285514832,
+      "learning_rate": 4.711612952056988e-05,
+      "loss": 0.4357,
       "step": 24950
     },
     {
-      "epoch": 0.88,
-      "learning_rate": 4.7273261652375066e-05,
-      "loss": 0.2953,
+      "epoch": 0.8993765091721627,
+      "grad_norm": 0.1486726552248001,
+      "learning_rate": 4.711476874677957e-05,
+      "loss": 0.4281,
       "step": 24955
     },
     {
-      "epoch": 0.88,
-      "learning_rate": 4.727196779021614e-05,
-      "loss": 0.3285,
+      "epoch": 0.8995567088333873,
+      "grad_norm": 0.1914074718952179,
+      "learning_rate": 4.711340767167982e-05,
+      "loss": 0.4557,
       "step": 24960
     },
     {
-      "epoch": 0.88,
-      "learning_rate": 4.7270673638870386e-05,
-      "loss": 0.3097,
+      "epoch": 0.8997369084946121,
+      "grad_norm": 0.17185872793197632,
+      "learning_rate": 4.711204629528917e-05,
+      "loss": 0.4334,
       "step": 24965
     },
     {
-      "epoch": 0.88,
-      "learning_rate": 4.7269379198354606e-05,
-      "loss": 0.3233,
+      "epoch": 0.8999171081558367,
+      "grad_norm": 0.15413062274456024,
+      "learning_rate": 4.711068461762617e-05,
+      "loss": 0.4027,
       "step": 24970
     },
     {
-      "epoch": 0.88,
-      "learning_rate": 4.726808446868562e-05,
-      "loss": 0.3141,
+      "epoch": 0.9000973078170613,
+      "grad_norm": 0.2285022735595703,
+      "learning_rate": 4.710932263870936e-05,
+      "loss": 0.423,
       "step": 24975
     },
     {
-      "epoch": 0.88,
-      "learning_rate": 4.7266789449880235e-05,
-      "loss": 0.2786,
+      "epoch": 0.9002775074782859,
+      "grad_norm": 0.18323256075382233,
+      "learning_rate": 4.710796035855732e-05,
+      "loss": 0.4656,
       "step": 24980
     },
     {
-      "epoch": 0.88,
-      "learning_rate": 4.7265494141955255e-05,
-      "loss": 0.2944,
+      "epoch": 0.9004577071395106,
+      "grad_norm": 0.19098562002182007,
+      "learning_rate": 4.71065977771886e-05,
+      "loss": 0.4367,
       "step": 24985
     },
     {
-      "epoch": 0.88,
-      "learning_rate": 4.7264198544927515e-05,
-      "loss": 0.313,
+      "epoch": 0.9006379068007352,
+      "grad_norm": 0.14779239892959595,
+      "learning_rate": 4.710523489462177e-05,
+      "loss": 0.4426,
       "step": 24990
     },
     {
-      "epoch": 0.88,
-      "learning_rate": 4.726290265881382e-05,
-      "loss": 0.2895,
+      "epoch": 0.9008181064619598,
+      "grad_norm": 0.1570071429014206,
+      "learning_rate": 4.710387171087539e-05,
+      "loss": 0.4674,
       "step": 24995
     },
     {
-      "epoch": 0.88,
-      "learning_rate": 4.7261606483631017e-05,
-      "loss": 0.3045,
+      "epoch": 0.9009983061231844,
+      "grad_norm": 0.1645687371492386,
+      "learning_rate": 4.7102508225968035e-05,
+      "loss": 0.4766,
       "step": 25000
     },
     {
-      "epoch": 0.88,
-      "eval_loss": 0.2971786856651306,
-      "eval_runtime": 10.5292,
-      "eval_samples_per_second": 9.497,
-      "eval_steps_per_second": 9.497,
+      "epoch": 0.9009983061231844,
+      "eval_loss": 0.4570350646972656,
+      "eval_runtime": 3.5365,
+      "eval_samples_per_second": 28.276,
+      "eval_steps_per_second": 7.069,
       "step": 25000
     },
     {
-      "epoch": 0.88,
-      "learning_rate": 4.726031001939592e-05,
-      "loss": 0.3188,
+      "epoch": 0.9011785057844092,
+      "grad_norm": 0.15871667861938477,
+      "learning_rate": 4.7101144439918287e-05,
+      "loss": 0.4376,
       "step": 25005
     },
     {
-      "epoch": 0.88,
-      "learning_rate": 4.725901326612536e-05,
-      "loss": 0.3274,
+      "epoch": 0.9013587054456338,
+      "grad_norm": 0.1973850131034851,
+      "learning_rate": 4.709978035274473e-05,
+      "loss": 0.446,
       "step": 25010
     },
     {
-      "epoch": 0.88,
-      "learning_rate": 4.72577162238362e-05,
-      "loss": 0.3033,
+      "epoch": 0.9015389051068584,
+      "grad_norm": 0.15039071440696716,
+      "learning_rate": 4.709841596446594e-05,
+      "loss": 0.4431,
       "step": 25015
     },
     {
-      "epoch": 0.88,
-      "learning_rate": 4.725641889254525e-05,
-      "loss": 0.3326,
+      "epoch": 0.901719104768083,
+      "grad_norm": 0.16307663917541504,
+      "learning_rate": 4.70970512751005e-05,
+      "loss": 0.4178,
       "step": 25020
     },
     {
-      "epoch": 0.88,
-      "learning_rate": 4.725512127226937e-05,
-      "loss": 0.3117,
+      "epoch": 0.9018993044293077,
+      "grad_norm": 0.17655940353870392,
+      "learning_rate": 4.709568628466703e-05,
+      "loss": 0.4112,
       "step": 25025
     },
     {
-      "epoch": 0.88,
-      "learning_rate": 4.725382336302541e-05,
-      "loss": 0.3072,
+      "epoch": 0.9020795040905323,
+      "grad_norm": 0.21620622277259827,
+      "learning_rate": 4.709432099318411e-05,
+      "loss": 0.4354,
       "step": 25030
     },
     {
-      "epoch": 0.88,
-      "learning_rate": 4.725252516483021e-05,
-      "loss": 0.2899,
+      "epoch": 0.9022597037517569,
+      "grad_norm": 0.17951133847236633,
+      "learning_rate": 4.7092955400670336e-05,
+      "loss": 0.4332,
       "step": 25035
     },
     {
-      "epoch": 0.88,
-      "learning_rate": 4.725122667770064e-05,
-      "loss": 0.305,
+      "epoch": 0.9024399034129816,
+      "grad_norm": 0.15933279693126678,
+      "learning_rate": 4.7091589507144326e-05,
+      "loss": 0.3898,
       "step": 25040
     },
     {
-      "epoch": 0.88,
-      "learning_rate": 4.7249927901653566e-05,
-      "loss": 0.2833,
+      "epoch": 0.9026201030742063,
+      "grad_norm": 0.1578451693058014,
+      "learning_rate": 4.7090223312624683e-05,
+      "loss": 0.4215,
       "step": 25045
     },
     {
-      "epoch": 0.88,
-      "learning_rate": 4.724862883670583e-05,
-      "loss": 0.3178,
+      "epoch": 0.9028003027354309,
+      "grad_norm": 0.1939324289560318,
+      "learning_rate": 4.708885681713003e-05,
+      "loss": 0.4807,
       "step": 25050
     },
     {
-      "epoch": 0.88,
-      "learning_rate": 4.724732948287432e-05,
-      "loss": 0.2996,
+      "epoch": 0.9029805023966555,
+      "grad_norm": 0.1886306256055832,
+      "learning_rate": 4.708749002067897e-05,
+      "loss": 0.4057,
       "step": 25055
     },
     {
-      "epoch": 0.88,
-      "learning_rate": 4.7246029840175884e-05,
-      "loss": 0.2997,
+      "epoch": 0.9031607020578801,
+      "grad_norm": 0.1769237220287323,
+      "learning_rate": 4.708612292329015e-05,
+      "loss": 0.4619,
       "step": 25060
     },
     {
-      "epoch": 0.88,
-      "learning_rate": 4.7244729908627426e-05,
-      "loss": 0.3063,
+      "epoch": 0.9033409017191047,
+      "grad_norm": 0.17207685112953186,
+      "learning_rate": 4.7084755524982175e-05,
+      "loss": 0.4632,
       "step": 25065
     },
     {
-      "epoch": 0.88,
-      "learning_rate": 4.72434296882458e-05,
-      "loss": 0.2872,
+      "epoch": 0.9035211013803294,
+      "grad_norm": 0.17112012207508087,
+      "learning_rate": 4.7083387825773676e-05,
+      "loss": 0.4322,
       "step": 25070
     },
     {
-      "epoch": 0.88,
-      "learning_rate": 4.7242129179047904e-05,
-      "loss": 0.2907,
+      "epoch": 0.903701301041554,
+      "grad_norm": 0.1867535561323166,
+      "learning_rate": 4.70820198256833e-05,
+      "loss": 0.4359,
       "step": 25075
     },
     {
-      "epoch": 0.88,
-      "learning_rate": 4.724082838105062e-05,
-      "loss": 0.3003,
+      "epoch": 0.9038815007027787,
+      "grad_norm": 0.21555501222610474,
+      "learning_rate": 4.708065152472967e-05,
+      "loss": 0.4375,
       "step": 25080
     },
     {
-      "epoch": 0.88,
-      "learning_rate": 4.723952729427083e-05,
-      "loss": 0.3175,
+      "epoch": 0.9040617003640034,
+      "grad_norm": 0.16034695506095886,
+      "learning_rate": 4.707928292293144e-05,
+      "loss": 0.4587,
       "step": 25085
     },
     {
-      "epoch": 0.88,
-      "learning_rate": 4.723822591872544e-05,
-      "loss": 0.3222,
+      "epoch": 0.904241900025228,
+      "grad_norm": 0.15322132408618927,
+      "learning_rate": 4.7077914020307266e-05,
+      "loss": 0.4422,
       "step": 25090
     },
     {
-      "epoch": 0.88,
-      "learning_rate": 4.7236924254431344e-05,
-      "loss": 0.3183,
+      "epoch": 0.9044220996864526,
+      "grad_norm": 0.16674070060253143,
+      "learning_rate": 4.707654481687578e-05,
+      "loss": 0.4216,
       "step": 25095
     },
     {
-      "epoch": 0.88,
-      "learning_rate": 4.723562230140544e-05,
-      "loss": 0.2934,
+      "epoch": 0.9046022993476772,
+      "grad_norm": 0.16037312150001526,
+      "learning_rate": 4.707517531265565e-05,
+      "loss": 0.4055,
       "step": 25100
     },
     {
-      "epoch": 0.88,
-      "learning_rate": 4.7234320059664626e-05,
-      "loss": 0.3129,
+      "epoch": 0.9047824990089018,
+      "grad_norm": 0.17139527201652527,
+      "learning_rate": 4.707380550766553e-05,
+      "loss": 0.4565,
       "step": 25105
     },
     {
-      "epoch": 0.88,
-      "learning_rate": 4.7233017529225826e-05,
-      "loss": 0.2844,
+      "epoch": 0.9049626986701265,
+      "grad_norm": 0.1893669068813324,
+      "learning_rate": 4.7072435401924075e-05,
+      "loss": 0.4536,
       "step": 25110
     },
     {
-      "epoch": 0.88,
-      "learning_rate": 4.723171471010595e-05,
-      "loss": 0.2877,
+      "epoch": 0.9051428983313511,
+      "grad_norm": 0.17662879824638367,
+      "learning_rate": 4.7071064995449964e-05,
+      "loss": 0.4653,
       "step": 25115
     },
     {
-      "epoch": 0.88,
-      "learning_rate": 4.72304116023219e-05,
-      "loss": 0.2901,
+      "epoch": 0.9053230979925758,
+      "grad_norm": 0.16549894213676453,
+      "learning_rate": 4.7069694288261864e-05,
+      "loss": 0.402,
       "step": 25120
     },
     {
-      "epoch": 0.88,
-      "learning_rate": 4.72291082058906e-05,
-      "loss": 0.2915,
+      "epoch": 0.9055032976538004,
+      "grad_norm": 0.1662409007549286,
+      "learning_rate": 4.706832328037846e-05,
+      "loss": 0.4282,
       "step": 25125
     },
     {
-      "epoch": 0.88,
-      "learning_rate": 4.7227804520829e-05,
-      "loss": 0.3152,
+      "epoch": 0.9056834973150251,
+      "grad_norm": 0.21818654239177704,
+      "learning_rate": 4.706695197181842e-05,
+      "loss": 0.4141,
       "step": 25130
     },
     {
-      "epoch": 0.88,
-      "learning_rate": 4.7226500547153996e-05,
-      "loss": 0.301,
+      "epoch": 0.9058636969762497,
+      "grad_norm": 0.14861366152763367,
+      "learning_rate": 4.706558036260042e-05,
+      "loss": 0.4126,
       "step": 25135
     },
     {
-      "epoch": 0.88,
-      "learning_rate": 4.722519628488253e-05,
-      "loss": 0.3223,
+      "epoch": 0.9060438966374743,
+      "grad_norm": 0.17982593178749084,
+      "learning_rate": 4.7064208452743174e-05,
+      "loss": 0.4186,
       "step": 25140
     },
     {
-      "epoch": 0.88,
-      "learning_rate": 4.722389173403153e-05,
-      "loss": 0.3069,
+      "epoch": 0.9062240962986989,
+      "grad_norm": 0.19204726815223694,
+      "learning_rate": 4.706283624226536e-05,
+      "loss": 0.4466,
       "step": 25145
     },
     {
-      "epoch": 0.88,
-      "learning_rate": 4.7222586894617956e-05,
-      "loss": 0.3253,
+      "epoch": 0.9064042959599236,
+      "grad_norm": 0.20537865161895752,
+      "learning_rate": 4.7061463731185676e-05,
+      "loss": 0.4208,
       "step": 25150
     },
     {
-      "epoch": 0.89,
-      "learning_rate": 4.722128176665873e-05,
-      "loss": 0.2785,
+      "epoch": 0.9065844956211482,
+      "grad_norm": 0.12418222427368164,
+      "learning_rate": 4.7060090919522806e-05,
+      "loss": 0.4412,
       "step": 25155
     },
     {
-      "epoch": 0.89,
-      "learning_rate": 4.72199763501708e-05,
-      "loss": 0.3071,
+      "epoch": 0.9067646952823729,
+      "grad_norm": 0.15775543451309204,
+      "learning_rate": 4.705871780729548e-05,
+      "loss": 0.4425,
       "step": 25160
     },
     {
-      "epoch": 0.89,
-      "learning_rate": 4.7218670645171125e-05,
-      "loss": 0.3289,
+      "epoch": 0.9069448949435975,
+      "grad_norm": 0.15570229291915894,
+      "learning_rate": 4.705734439452239e-05,
+      "loss": 0.445,
       "step": 25165
     },
     {
-      "epoch": 0.89,
-      "learning_rate": 4.721736465167666e-05,
-      "loss": 0.3063,
+      "epoch": 0.9071250946048222,
+      "grad_norm": 0.18232616782188416,
+      "learning_rate": 4.705597068122225e-05,
+      "loss": 0.4503,
       "step": 25170
     },
     {
-      "epoch": 0.89,
-      "learning_rate": 4.721605836970435e-05,
-      "loss": 0.3212,
+      "epoch": 0.9073052942660468,
+      "grad_norm": 0.20230647921562195,
+      "learning_rate": 4.705459666741379e-05,
+      "loss": 0.4519,
       "step": 25175
     },
     {
-      "epoch": 0.89,
-      "learning_rate": 4.7214751799271165e-05,
-      "loss": 0.2908,
+      "epoch": 0.9074854939272714,
+      "grad_norm": 0.1548745334148407,
+      "learning_rate": 4.705322235311571e-05,
+      "loss": 0.3896,
       "step": 25180
     },
     {
-      "epoch": 0.89,
-      "learning_rate": 4.721344494039407e-05,
-      "loss": 0.3022,
+      "epoch": 0.907665693588496,
+      "grad_norm": 0.19131234288215637,
+      "learning_rate": 4.705184773834675e-05,
+      "loss": 0.4694,
       "step": 25185
     },
     {
-      "epoch": 0.89,
-      "learning_rate": 4.721213779309003e-05,
-      "loss": 0.3016,
+      "epoch": 0.9078458932497206,
+      "grad_norm": 0.17528343200683594,
+      "learning_rate": 4.705047282312563e-05,
+      "loss": 0.4463,
       "step": 25190
     },
     {
-      "epoch": 0.89,
-      "learning_rate": 4.7210830357376015e-05,
-      "loss": 0.3272,
+      "epoch": 0.9080260929109454,
+      "grad_norm": 0.16300806403160095,
+      "learning_rate": 4.704909760747109e-05,
+      "loss": 0.4469,
       "step": 25195
     },
     {
-      "epoch": 0.89,
-      "learning_rate": 4.720952263326901e-05,
-      "loss": 0.3067,
+      "epoch": 0.90820629257217,
+      "grad_norm": 0.14464612305164337,
+      "learning_rate": 4.704772209140186e-05,
+      "loss": 0.4079,
       "step": 25200
     },
     {
-      "epoch": 0.89,
-      "learning_rate": 4.7208214620785986e-05,
-      "loss": 0.3113,
+      "epoch": 0.9083864922333946,
+      "grad_norm": 0.19416652619838715,
+      "learning_rate": 4.704634627493669e-05,
+      "loss": 0.4484,
       "step": 25205
     },
     {
-      "epoch": 0.89,
-      "learning_rate": 4.7206906319943934e-05,
-      "loss": 0.277,
+      "epoch": 0.9085666918946192,
+      "grad_norm": 0.18099582195281982,
+      "learning_rate": 4.704497015809432e-05,
+      "loss": 0.4251,
       "step": 25210
     },
     {
-      "epoch": 0.89,
-      "learning_rate": 4.720559773075984e-05,
-      "loss": 0.3084,
+      "epoch": 0.9087468915558439,
+      "grad_norm": 0.19250603020191193,
+      "learning_rate": 4.70435937408935e-05,
+      "loss": 0.434,
       "step": 25215
     },
     {
-      "epoch": 0.89,
-      "learning_rate": 4.720428885325069e-05,
-      "loss": 0.2904,
+      "epoch": 0.9089270912170685,
+      "grad_norm": 0.17500755190849304,
+      "learning_rate": 4.704221702335298e-05,
+      "loss": 0.3891,
       "step": 25220
     },
     {
-      "epoch": 0.89,
-      "learning_rate": 4.720297968743348e-05,
-      "loss": 0.3018,
+      "epoch": 0.9091072908782931,
+      "grad_norm": 0.1807602047920227,
+      "learning_rate": 4.7040840005491526e-05,
+      "loss": 0.4505,
       "step": 25225
     },
     {
-      "epoch": 0.89,
-      "learning_rate": 4.720167023332522e-05,
-      "loss": 0.3067,
+      "epoch": 0.9092874905395177,
+      "grad_norm": 0.1530625820159912,
+      "learning_rate": 4.7039462687327885e-05,
+      "loss": 0.4264,
       "step": 25230
     },
     {
-      "epoch": 0.89,
-      "learning_rate": 4.72003604909429e-05,
-      "loss": 0.3084,
+      "epoch": 0.9094676902007425,
+      "grad_norm": 0.1661711484193802,
+      "learning_rate": 4.703808506888084e-05,
+      "loss": 0.4535,
       "step": 25235
     },
     {
-      "epoch": 0.89,
-      "learning_rate": 4.7199050460303525e-05,
-      "loss": 0.2959,
+      "epoch": 0.9096478898619671,
+      "grad_norm": 0.14613687992095947,
+      "learning_rate": 4.7036707150169145e-05,
+      "loss": 0.4152,
       "step": 25240
     },
     {
-      "epoch": 0.89,
-      "learning_rate": 4.719774014142411e-05,
-      "loss": 0.2769,
+      "epoch": 0.9098280895231917,
+      "grad_norm": 0.18025970458984375,
+      "learning_rate": 4.703532893121159e-05,
+      "loss": 0.4294,
       "step": 25245
     },
     {
-      "epoch": 0.89,
-      "learning_rate": 4.719642953432167e-05,
-      "loss": 0.2959,
+      "epoch": 0.9100082891844163,
+      "grad_norm": 0.14726756513118744,
+      "learning_rate": 4.703395041202694e-05,
+      "loss": 0.4612,
       "step": 25250
     },
     {
-      "epoch": 0.89,
-      "learning_rate": 4.719511863901322e-05,
-      "loss": 0.2857,
+      "epoch": 0.910188488845641,
+      "grad_norm": 0.19874265789985657,
+      "learning_rate": 4.703257159263398e-05,
+      "loss": 0.4415,
       "step": 25255
     },
     {
-      "epoch": 0.89,
-      "learning_rate": 4.719380745551578e-05,
-      "loss": 0.2813,
+      "epoch": 0.9103686885068656,
+      "grad_norm": 0.17537064850330353,
+      "learning_rate": 4.70311924730515e-05,
+      "loss": 0.4454,
       "step": 25260
     },
     {
-      "epoch": 0.89,
-      "learning_rate": 4.7192495983846385e-05,
-      "loss": 0.3073,
+      "epoch": 0.9105488881680902,
+      "grad_norm": 0.1721314936876297,
+      "learning_rate": 4.702981305329829e-05,
+      "loss": 0.4314,
       "step": 25265
     },
     {
-      "epoch": 0.89,
-      "learning_rate": 4.719118422402205e-05,
-      "loss": 0.2863,
+      "epoch": 0.9107290878293148,
+      "grad_norm": 0.16811582446098328,
+      "learning_rate": 4.702843333339314e-05,
+      "loss": 0.4317,
       "step": 25270
     },
     {
-      "epoch": 0.89,
-      "learning_rate": 4.718987217605981e-05,
-      "loss": 0.3157,
+      "epoch": 0.9109092874905396,
+      "grad_norm": 0.19957655668258667,
+      "learning_rate": 4.702705331335485e-05,
+      "loss": 0.4278,
       "step": 25275
     },
     {
-      "epoch": 0.89,
-      "learning_rate": 4.71885598399767e-05,
-      "loss": 0.2739,
+      "epoch": 0.9110894871517642,
+      "grad_norm": 0.17485888302326202,
+      "learning_rate": 4.702567299320223e-05,
+      "loss": 0.4501,
       "step": 25280
     },
     {
-      "epoch": 0.89,
-      "learning_rate": 4.718724721578978e-05,
-      "loss": 0.2985,
+      "epoch": 0.9112696868129888,
+      "grad_norm": 0.13707765936851501,
+      "learning_rate": 4.702429237295407e-05,
+      "loss": 0.4126,
       "step": 25285
     },
     {
-      "epoch": 0.89,
-      "learning_rate": 4.7185934303516054e-05,
-      "loss": 0.2974,
+      "epoch": 0.9114498864742134,
+      "grad_norm": 0.182311549782753,
+      "learning_rate": 4.702291145262919e-05,
+      "loss": 0.4308,
       "step": 25290
     },
     {
-      "epoch": 0.89,
-      "learning_rate": 4.718462110317261e-05,
-      "loss": 0.3091,
+      "epoch": 0.9116300861354381,
+      "grad_norm": 0.16993586719036102,
+      "learning_rate": 4.702153023224641e-05,
+      "loss": 0.4155,
       "step": 25295
     },
     {
-      "epoch": 0.89,
-      "learning_rate": 4.718330761477647e-05,
-      "loss": 0.2901,
+      "epoch": 0.9118102857966627,
+      "grad_norm": 0.16655287146568298,
+      "learning_rate": 4.7020148711824546e-05,
+      "loss": 0.4481,
       "step": 25300
     },
     {
-      "epoch": 0.89,
-      "learning_rate": 4.7181993838344704e-05,
-      "loss": 0.2989,
+      "epoch": 0.9119904854578873,
+      "grad_norm": 0.1623249053955078,
+      "learning_rate": 4.701876689138242e-05,
+      "loss": 0.4569,
       "step": 25305
     },
     {
-      "epoch": 0.89,
-      "learning_rate": 4.718067977389437e-05,
-      "loss": 0.2995,
+      "epoch": 0.9121706851191119,
+      "grad_norm": 0.15807580947875977,
+      "learning_rate": 4.701738477093885e-05,
+      "loss": 0.4459,
       "step": 25310
     },
     {
-      "epoch": 0.89,
-      "learning_rate": 4.7179365421442526e-05,
-      "loss": 0.3258,
+      "epoch": 0.9123508847803367,
+      "grad_norm": 0.20741410553455353,
+      "learning_rate": 4.701600235051268e-05,
+      "loss": 0.4314,
       "step": 25315
     },
     {
-      "epoch": 0.89,
-      "learning_rate": 4.717805078100623e-05,
-      "loss": 0.3229,
+      "epoch": 0.9125310844415613,
+      "grad_norm": 0.1682579219341278,
+      "learning_rate": 4.701461963012274e-05,
+      "loss": 0.4152,
       "step": 25320
     },
     {
-      "epoch": 0.89,
-      "learning_rate": 4.717673585260257e-05,
-      "loss": 0.2945,
+      "epoch": 0.9127112841027859,
+      "grad_norm": 0.15219363570213318,
+      "learning_rate": 4.701323660978787e-05,
+      "loss": 0.415,
       "step": 25325
     },
     {
-      "epoch": 0.89,
-      "learning_rate": 4.717542063624861e-05,
-      "loss": 0.3246,
+      "epoch": 0.9128914837640105,
+      "grad_norm": 0.1439961940050125,
+      "learning_rate": 4.701185328952692e-05,
+      "loss": 0.4282,
       "step": 25330
     },
     {
-      "epoch": 0.89,
-      "learning_rate": 4.7174105131961424e-05,
-      "loss": 0.2893,
+      "epoch": 0.9130716834252351,
+      "grad_norm": 0.18314993381500244,
+      "learning_rate": 4.701046966935872e-05,
+      "loss": 0.4325,
       "step": 25335
     },
     {
-      "epoch": 0.89,
-      "learning_rate": 4.7172789339758095e-05,
-      "loss": 0.3087,
+      "epoch": 0.9132518830864598,
+      "grad_norm": 0.19595153629779816,
+      "learning_rate": 4.700908574930213e-05,
+      "loss": 0.4474,
       "step": 25340
     },
     {
-      "epoch": 0.89,
-      "learning_rate": 4.717147325965571e-05,
-      "loss": 0.2858,
+      "epoch": 0.9134320827476844,
+      "grad_norm": 0.14837300777435303,
+      "learning_rate": 4.700770152937601e-05,
+      "loss": 0.4244,
       "step": 25345
     },
     {
-      "epoch": 0.89,
-      "learning_rate": 4.717015689167136e-05,
-      "loss": 0.2646,
+      "epoch": 0.9136122824089091,
+      "grad_norm": 0.19343934953212738,
+      "learning_rate": 4.700631700959923e-05,
+      "loss": 0.4336,
       "step": 25350
     },
     {
-      "epoch": 0.89,
-      "learning_rate": 4.716884023582213e-05,
-      "loss": 0.3022,
+      "epoch": 0.9137924820701337,
+      "grad_norm": 0.15232184529304504,
+      "learning_rate": 4.700493218999063e-05,
+      "loss": 0.4504,
       "step": 25355
     },
     {
-      "epoch": 0.89,
-      "learning_rate": 4.716752329212512e-05,
-      "loss": 0.3013,
+      "epoch": 0.9139726817313584,
+      "grad_norm": 0.15530715882778168,
+      "learning_rate": 4.700354707056909e-05,
+      "loss": 0.4485,
       "step": 25360
     },
     {
-      "epoch": 0.89,
-      "learning_rate": 4.716620606059742e-05,
-      "loss": 0.3055,
+      "epoch": 0.914152881392583,
+      "grad_norm": 0.19562314450740814,
+      "learning_rate": 4.7002161651353485e-05,
+      "loss": 0.46,
       "step": 25365
     },
     {
-      "epoch": 0.89,
-      "learning_rate": 4.7164888541256154e-05,
-      "loss": 0.289,
+      "epoch": 0.9143330810538076,
+      "grad_norm": 0.17155522108078003,
+      "learning_rate": 4.7000775932362684e-05,
+      "loss": 0.4414,
       "step": 25370
     },
     {
-      "epoch": 0.89,
-      "learning_rate": 4.716357073411841e-05,
-      "loss": 0.3143,
+      "epoch": 0.9145132807150322,
+      "grad_norm": 0.17853830754756927,
+      "learning_rate": 4.699938991361558e-05,
+      "loss": 0.4158,
       "step": 25375
     },
     {
-      "epoch": 0.89,
-      "learning_rate": 4.7162252639201314e-05,
-      "loss": 0.3089,
+      "epoch": 0.9146934803762569,
+      "grad_norm": 0.17986997961997986,
+      "learning_rate": 4.6998003595131035e-05,
+      "loss": 0.4345,
       "step": 25380
     },
     {
-      "epoch": 0.89,
-      "learning_rate": 4.7160934256521974e-05,
-      "loss": 0.2892,
+      "epoch": 0.9148736800374815,
+      "grad_norm": 0.16720250248908997,
+      "learning_rate": 4.699661697692796e-05,
+      "loss": 0.4374,
       "step": 25385
     },
     {
-      "epoch": 0.89,
-      "learning_rate": 4.71596155860975e-05,
-      "loss": 0.3338,
+      "epoch": 0.9150538796987062,
+      "grad_norm": 0.18035994470119476,
+      "learning_rate": 4.699523005902522e-05,
+      "loss": 0.423,
       "step": 25390
     },
     {
-      "epoch": 0.89,
-      "learning_rate": 4.715829662794502e-05,
-      "loss": 0.291,
+      "epoch": 0.9152340793599308,
+      "grad_norm": 0.17745894193649292,
+      "learning_rate": 4.699384284144174e-05,
+      "loss": 0.4273,
       "step": 25395
     },
     {
-      "epoch": 0.89,
-      "learning_rate": 4.715697738208167e-05,
-      "loss": 0.2916,
+      "epoch": 0.9154142790211555,
+      "grad_norm": 0.1389733850955963,
+      "learning_rate": 4.699245532419642e-05,
+      "loss": 0.4206,
       "step": 25400
     },
     {
-      "epoch": 0.89,
-      "learning_rate": 4.715565784852457e-05,
-      "loss": 0.2713,
+      "epoch": 0.9155944786823801,
+      "grad_norm": 0.16255025565624237,
+      "learning_rate": 4.699106750730814e-05,
+      "loss": 0.4705,
       "step": 25405
     },
     {
-      "epoch": 0.89,
-      "learning_rate": 4.715433802729085e-05,
-      "loss": 0.3044,
+      "epoch": 0.9157746783436047,
+      "grad_norm": 0.16465483605861664,
+      "learning_rate": 4.6989679390795826e-05,
+      "loss": 0.4509,
       "step": 25410
     },
     {
-      "epoch": 0.89,
-      "learning_rate": 4.715301791839765e-05,
-      "loss": 0.3076,
+      "epoch": 0.9159548780048293,
+      "grad_norm": 0.16897955536842346,
+      "learning_rate": 4.6988290974678384e-05,
+      "loss": 0.4106,
       "step": 25415
     },
     {
-      "epoch": 0.89,
-      "learning_rate": 4.7151697521862104e-05,
-      "loss": 0.3235,
+      "epoch": 0.916135077666054,
+      "grad_norm": 0.17163941264152527,
+      "learning_rate": 4.698690225897474e-05,
+      "loss": 0.4134,
       "step": 25420
     },
     {
-      "epoch": 0.89,
-      "learning_rate": 4.715037683770138e-05,
-      "loss": 0.362,
+      "epoch": 0.9163152773272786,
+      "grad_norm": 0.16389816999435425,
+      "learning_rate": 4.698551324370381e-05,
+      "loss": 0.4388,
       "step": 25425
     },
     {
-      "epoch": 0.89,
-      "learning_rate": 4.71490558659326e-05,
-      "loss": 0.2861,
+      "epoch": 0.9164954769885033,
+      "grad_norm": 0.17674943804740906,
+      "learning_rate": 4.698412392888452e-05,
+      "loss": 0.4547,
       "step": 25430
     },
     {
-      "epoch": 0.89,
-      "learning_rate": 4.714773460657293e-05,
-      "loss": 0.2833,
+      "epoch": 0.9166756766497279,
+      "grad_norm": 0.1594432145357132,
+      "learning_rate": 4.698273431453579e-05,
+      "loss": 0.4396,
       "step": 25435
     },
     {
-      "epoch": 0.9,
-      "learning_rate": 4.7146413059639524e-05,
-      "loss": 0.3223,
+      "epoch": 0.9168558763109526,
+      "grad_norm": 0.17538687586784363,
+      "learning_rate": 4.6981344400676566e-05,
+      "loss": 0.4724,
       "step": 25440
     },
     {
-      "epoch": 0.9,
-      "learning_rate": 4.714509122514953e-05,
-      "loss": 0.2809,
+      "epoch": 0.9170360759721772,
+      "grad_norm": 0.1810724139213562,
+      "learning_rate": 4.697995418732578e-05,
+      "loss": 0.4075,
       "step": 25445
     },
     {
-      "epoch": 0.9,
-      "learning_rate": 4.7143769103120125e-05,
-      "loss": 0.3196,
+      "epoch": 0.9172162756334018,
+      "grad_norm": 0.15371228754520416,
+      "learning_rate": 4.6978563674502375e-05,
+      "loss": 0.4578,
       "step": 25450
     },
     {
-      "epoch": 0.9,
-      "learning_rate": 4.714244669356848e-05,
-      "loss": 0.2941,
+      "epoch": 0.9173964752946264,
+      "grad_norm": 0.14616931974887848,
+      "learning_rate": 4.6977172862225294e-05,
+      "loss": 0.4127,
       "step": 25455
     },
     {
-      "epoch": 0.9,
-      "learning_rate": 4.714112399651175e-05,
-      "loss": 0.3171,
+      "epoch": 0.917576674955851,
+      "grad_norm": 0.19712692499160767,
+      "learning_rate": 4.697578175051348e-05,
+      "loss": 0.4611,
       "step": 25460
     },
     {
-      "epoch": 0.9,
-      "learning_rate": 4.713980101196712e-05,
-      "loss": 0.2729,
+      "epoch": 0.9177568746170757,
+      "grad_norm": 0.16714803874492645,
+      "learning_rate": 4.69743903393859e-05,
+      "loss": 0.4397,
       "step": 25465
     },
     {
-      "epoch": 0.9,
-      "learning_rate": 4.713847773995176e-05,
-      "loss": 0.2862,
+      "epoch": 0.9179370742783004,
+      "grad_norm": 0.15617181360721588,
+      "learning_rate": 4.6972998628861506e-05,
+      "loss": 0.3992,
       "step": 25470
     },
     {
-      "epoch": 0.9,
-      "learning_rate": 4.7137154180482854e-05,
-      "loss": 0.2835,
+      "epoch": 0.918117273939525,
+      "grad_norm": 0.17703229188919067,
+      "learning_rate": 4.697160661895927e-05,
+      "loss": 0.4722,
       "step": 25475
     },
     {
-      "epoch": 0.9,
-      "learning_rate": 4.71358303335776e-05,
-      "loss": 0.31,
+      "epoch": 0.9182974736007496,
+      "grad_norm": 0.17935994267463684,
+      "learning_rate": 4.6970214309698134e-05,
+      "loss": 0.4434,
       "step": 25480
     },
     {
-      "epoch": 0.9,
-      "learning_rate": 4.713450619925317e-05,
-      "loss": 0.2939,
+      "epoch": 0.9184776732619743,
+      "grad_norm": 0.22057127952575684,
+      "learning_rate": 4.6968821701097086e-05,
+      "loss": 0.4237,
       "step": 25485
     },
     {
-      "epoch": 0.9,
-      "learning_rate": 4.713318177752677e-05,
-      "loss": 0.2882,
+      "epoch": 0.9186578729231989,
+      "grad_norm": 0.17813293635845184,
+      "learning_rate": 4.69674287931751e-05,
+      "loss": 0.4612,
       "step": 25490
     },
     {
-      "epoch": 0.9,
-      "learning_rate": 4.713185706841559e-05,
-      "loss": 0.3074,
+      "epoch": 0.9188380725844235,
+      "grad_norm": 0.20391541719436646,
+      "learning_rate": 4.696603558595115e-05,
+      "loss": 0.4252,
       "step": 25495
     },
     {
-      "epoch": 0.9,
-      "learning_rate": 4.7130532071936814e-05,
-      "loss": 0.3055,
+      "epoch": 0.9190182722456481,
+      "grad_norm": 0.1561935395002365,
+      "learning_rate": 4.696464207944421e-05,
+      "loss": 0.4011,
       "step": 25500
     },
     {
-      "epoch": 0.9,
-      "eval_loss": 0.29553505778312683,
-      "eval_runtime": 10.5452,
-      "eval_samples_per_second": 9.483,
-      "eval_steps_per_second": 9.483,
+      "epoch": 0.9190182722456481,
+      "eval_loss": 0.4568355083465576,
+      "eval_runtime": 3.5356,
+      "eval_samples_per_second": 28.284,
+      "eval_steps_per_second": 7.071,
       "step": 25500
     },
     {
-      "epoch": 0.9,
-      "learning_rate": 4.712920678810768e-05,
-      "loss": 0.2973,
+      "epoch": 0.9191984719068728,
+      "grad_norm": 0.18729767203330994,
+      "learning_rate": 4.696324827367328e-05,
+      "loss": 0.4199,
       "step": 25505
     },
     {
-      "epoch": 0.9,
-      "learning_rate": 4.712788121694538e-05,
-      "loss": 0.2921,
+      "epoch": 0.9193786715680975,
+      "grad_norm": 0.19635900855064392,
+      "learning_rate": 4.696185416865734e-05,
+      "loss": 0.4135,
       "step": 25510
     },
     {
-      "epoch": 0.9,
-      "learning_rate": 4.712655535846713e-05,
-      "loss": 0.2836,
+      "epoch": 0.9195588712293221,
+      "grad_norm": 0.1535876840353012,
+      "learning_rate": 4.6960459764415386e-05,
+      "loss": 0.4719,
       "step": 25515
     },
     {
-      "epoch": 0.9,
-      "learning_rate": 4.712522921269012e-05,
-      "loss": 0.313,
+      "epoch": 0.9197390708905467,
+      "grad_norm": 0.16010642051696777,
+      "learning_rate": 4.695906506096643e-05,
+      "loss": 0.432,
       "step": 25520
     },
     {
-      "epoch": 0.9,
-      "learning_rate": 4.712390277963161e-05,
-      "loss": 0.3145,
+      "epoch": 0.9199192705517714,
+      "grad_norm": 0.1595637947320938,
+      "learning_rate": 4.6957670058329464e-05,
+      "loss": 0.4195,
       "step": 25525
     },
     {
-      "epoch": 0.9,
-      "learning_rate": 4.7122576059308796e-05,
-      "loss": 0.2959,
+      "epoch": 0.920099470212996,
+      "grad_norm": 0.15828733146190643,
+      "learning_rate": 4.6956274756523484e-05,
+      "loss": 0.4074,
       "step": 25530
     },
     {
-      "epoch": 0.9,
-      "learning_rate": 4.712124905173891e-05,
-      "loss": 0.3051,
+      "epoch": 0.9202796698742206,
+      "grad_norm": 0.16714178025722504,
+      "learning_rate": 4.695487915556752e-05,
+      "loss": 0.4126,
       "step": 25535
     },
     {
-      "epoch": 0.9,
-      "learning_rate": 4.711992175693918e-05,
-      "loss": 0.3089,
+      "epoch": 0.9204598695354452,
+      "grad_norm": 0.1568571776151657,
+      "learning_rate": 4.695348325548057e-05,
+      "loss": 0.451,
       "step": 25540
     },
     {
-      "epoch": 0.9,
-      "learning_rate": 4.7118594174926856e-05,
-      "loss": 0.325,
+      "epoch": 0.92064006919667,
+      "grad_norm": 0.1514507383108139,
+      "learning_rate": 4.695208705628167e-05,
+      "loss": 0.4052,
       "step": 25545
     },
     {
-      "epoch": 0.9,
-      "learning_rate": 4.711726630571915e-05,
-      "loss": 0.3128,
+      "epoch": 0.9208202688578946,
+      "grad_norm": 0.15713444352149963,
+      "learning_rate": 4.695069055798983e-05,
+      "loss": 0.4466,
       "step": 25550
     },
     {
-      "epoch": 0.9,
-      "learning_rate": 4.711593814933333e-05,
-      "loss": 0.3181,
+      "epoch": 0.9210004685191192,
+      "grad_norm": 0.1931026428937912,
+      "learning_rate": 4.694929376062408e-05,
+      "loss": 0.4462,
       "step": 25555
     },
     {
-      "epoch": 0.9,
-      "learning_rate": 4.7114609705786614e-05,
-      "loss": 0.2796,
+      "epoch": 0.9211806681803438,
+      "grad_norm": 0.1435370147228241,
+      "learning_rate": 4.694789666420345e-05,
+      "loss": 0.4197,
       "step": 25560
     },
     {
-      "epoch": 0.9,
-      "learning_rate": 4.711328097509627e-05,
-      "loss": 0.3029,
+      "epoch": 0.9213608678415685,
+      "grad_norm": 0.15287569165229797,
+      "learning_rate": 4.694649926874698e-05,
+      "loss": 0.4207,
       "step": 25565
     },
     {
-      "epoch": 0.9,
-      "learning_rate": 4.711195195727955e-05,
-      "loss": 0.2966,
+      "epoch": 0.9215410675027931,
+      "grad_norm": 0.17821337282657623,
+      "learning_rate": 4.694510157427371e-05,
+      "loss": 0.4436,
       "step": 25570
     },
     {
-      "epoch": 0.9,
-      "learning_rate": 4.7110622652353696e-05,
-      "loss": 0.3051,
+      "epoch": 0.9217212671640177,
+      "grad_norm": 0.1588643342256546,
+      "learning_rate": 4.694370358080267e-05,
+      "loss": 0.4427,
       "step": 25575
     },
     {
-      "epoch": 0.9,
-      "learning_rate": 4.7109293060335994e-05,
-      "loss": 0.3295,
+      "epoch": 0.9219014668252423,
+      "grad_norm": 0.2273692488670349,
+      "learning_rate": 4.6942305288352926e-05,
+      "loss": 0.4394,
       "step": 25580
     },
     {
-      "epoch": 0.9,
-      "learning_rate": 4.710796318124368e-05,
-      "loss": 0.2901,
+      "epoch": 0.922081666486467,
+      "grad_norm": 0.1667856127023697,
+      "learning_rate": 4.694090669694351e-05,
+      "loss": 0.4425,
       "step": 25585
     },
     {
-      "epoch": 0.9,
-      "learning_rate": 4.7106633015094037e-05,
-      "loss": 0.3316,
+      "epoch": 0.9222618661476917,
+      "grad_norm": 0.17408248782157898,
+      "learning_rate": 4.69395078065935e-05,
+      "loss": 0.4393,
       "step": 25590
     },
     {
-      "epoch": 0.9,
-      "learning_rate": 4.710530256190433e-05,
-      "loss": 0.2916,
+      "epoch": 0.9224420658089163,
+      "grad_norm": 0.17394454777240753,
+      "learning_rate": 4.693810861732194e-05,
+      "loss": 0.4212,
       "step": 25595
     },
     {
-      "epoch": 0.9,
-      "learning_rate": 4.7103971821691845e-05,
-      "loss": 0.2884,
+      "epoch": 0.9226222654701409,
+      "grad_norm": 0.21937377750873566,
+      "learning_rate": 4.69367091291479e-05,
+      "loss": 0.4295,
       "step": 25600
     },
     {
-      "epoch": 0.9,
-      "learning_rate": 4.710264079447384e-05,
-      "loss": 0.2916,
+      "epoch": 0.9228024651313655,
+      "grad_norm": 0.17998327314853668,
+      "learning_rate": 4.693530934209044e-05,
+      "loss": 0.4028,
       "step": 25605
     },
     {
-      "epoch": 0.9,
-      "learning_rate": 4.710130948026763e-05,
-      "loss": 0.2877,
+      "epoch": 0.9229826647925902,
+      "grad_norm": 0.14169585704803467,
+      "learning_rate": 4.693390925616864e-05,
+      "loss": 0.4253,
       "step": 25610
     },
     {
-      "epoch": 0.9,
-      "learning_rate": 4.7099977879090464e-05,
-      "loss": 0.3071,
+      "epoch": 0.9231628644538148,
+      "grad_norm": 0.16097652912139893,
+      "learning_rate": 4.693250887140157e-05,
+      "loss": 0.4052,
       "step": 25615
     },
     {
-      "epoch": 0.9,
-      "learning_rate": 4.7098645990959654e-05,
-      "loss": 0.3166,
+      "epoch": 0.9233430641150394,
+      "grad_norm": 0.14775985479354858,
+      "learning_rate": 4.6931108187808316e-05,
+      "loss": 0.4093,
       "step": 25620
     },
     {
-      "epoch": 0.9,
-      "learning_rate": 4.7097313815892495e-05,
-      "loss": 0.3184,
+      "epoch": 0.9235232637762641,
+      "grad_norm": 0.1724456250667572,
+      "learning_rate": 4.692970720540796e-05,
+      "loss": 0.4383,
       "step": 25625
     },
     {
-      "epoch": 0.9,
-      "learning_rate": 4.7095981353906275e-05,
-      "loss": 0.2892,
+      "epoch": 0.9237034634374888,
+      "grad_norm": 0.14798258244991302,
+      "learning_rate": 4.6928305924219587e-05,
+      "loss": 0.441,
       "step": 25630
     },
     {
-      "epoch": 0.9,
-      "learning_rate": 4.70946486050183e-05,
-      "loss": 0.2976,
+      "epoch": 0.9238836630987134,
+      "grad_norm": 0.16991879045963287,
+      "learning_rate": 4.692690434426229e-05,
+      "loss": 0.4178,
       "step": 25635
     },
     {
-      "epoch": 0.9,
-      "learning_rate": 4.7093315569245876e-05,
-      "loss": 0.3174,
+      "epoch": 0.924063862759938,
+      "grad_norm": 0.1819402128458023,
+      "learning_rate": 4.692550246555517e-05,
+      "loss": 0.4375,
       "step": 25640
     },
     {
-      "epoch": 0.9,
-      "learning_rate": 4.7091982246606304e-05,
-      "loss": 0.3334,
+      "epoch": 0.9242440624211626,
+      "grad_norm": 0.17619052529335022,
+      "learning_rate": 4.692410028811732e-05,
+      "loss": 0.4423,
       "step": 25645
     },
     {
-      "epoch": 0.9,
-      "learning_rate": 4.709064863711691e-05,
-      "loss": 0.3152,
+      "epoch": 0.9244242620823873,
+      "grad_norm": 0.17652717232704163,
+      "learning_rate": 4.692269781196785e-05,
+      "loss": 0.4498,
       "step": 25650
     },
     {
-      "epoch": 0.9,
-      "learning_rate": 4.708931474079499e-05,
-      "loss": 0.2769,
+      "epoch": 0.9246044617436119,
+      "grad_norm": 0.1433330625295639,
+      "learning_rate": 4.692129503712587e-05,
+      "loss": 0.4225,
       "step": 25655
     },
     {
-      "epoch": 0.9,
-      "learning_rate": 4.708798055765788e-05,
-      "loss": 0.3005,
+      "epoch": 0.9247846614048365,
+      "grad_norm": 0.18287613987922668,
+      "learning_rate": 4.6919891963610485e-05,
+      "loss": 0.4516,
       "step": 25660
     },
     {
-      "epoch": 0.9,
-      "learning_rate": 4.70866460877229e-05,
-      "loss": 0.3029,
+      "epoch": 0.9249648610660612,
+      "grad_norm": 0.17915308475494385,
+      "learning_rate": 4.691848859144081e-05,
+      "loss": 0.4276,
       "step": 25665
     },
     {
-      "epoch": 0.9,
-      "learning_rate": 4.7085311331007375e-05,
-      "loss": 0.2968,
+      "epoch": 0.9251450607272859,
+      "grad_norm": 0.16471678018569946,
+      "learning_rate": 4.691708492063598e-05,
+      "loss": 0.4445,
       "step": 25670
     },
     {
-      "epoch": 0.9,
-      "learning_rate": 4.708397628752864e-05,
-      "loss": 0.2918,
+      "epoch": 0.9253252603885105,
+      "grad_norm": 0.18665508925914764,
+      "learning_rate": 4.6915680951215114e-05,
+      "loss": 0.4602,
       "step": 25675
     },
     {
-      "epoch": 0.9,
-      "learning_rate": 4.7082640957304025e-05,
-      "loss": 0.3066,
+      "epoch": 0.9255054600497351,
+      "grad_norm": 0.1585237830877304,
+      "learning_rate": 4.6914276683197334e-05,
+      "loss": 0.4243,
       "step": 25680
     },
     {
-      "epoch": 0.9,
-      "learning_rate": 4.7081305340350875e-05,
-      "loss": 0.3111,
+      "epoch": 0.9256856597109597,
+      "grad_norm": 0.1394348442554474,
+      "learning_rate": 4.6912872116601776e-05,
+      "loss": 0.3881,
       "step": 25685
     },
     {
-      "epoch": 0.9,
-      "learning_rate": 4.707996943668652e-05,
-      "loss": 0.2707,
+      "epoch": 0.9258658593721844,
+      "grad_norm": 0.14394313097000122,
+      "learning_rate": 4.6911467251447574e-05,
+      "loss": 0.424,
       "step": 25690
     },
     {
-      "epoch": 0.9,
-      "learning_rate": 4.707863324632832e-05,
-      "loss": 0.3182,
+      "epoch": 0.926046059033409,
+      "grad_norm": 0.20186100900173187,
+      "learning_rate": 4.691006208775388e-05,
+      "loss": 0.4345,
       "step": 25695
     },
     {
-      "epoch": 0.9,
-      "learning_rate": 4.707729676929361e-05,
-      "loss": 0.2845,
+      "epoch": 0.9262262586946337,
+      "grad_norm": 0.16559197008609772,
+      "learning_rate": 4.690865662553983e-05,
+      "loss": 0.4314,
       "step": 25700
     },
     {
-      "epoch": 0.9,
-      "learning_rate": 4.7075960005599754e-05,
-      "loss": 0.289,
+      "epoch": 0.9264064583558583,
+      "grad_norm": 0.2052149623632431,
+      "learning_rate": 4.690725086482457e-05,
+      "loss": 0.4128,
       "step": 25705
     },
     {
-      "epoch": 0.9,
-      "learning_rate": 4.707462295526411e-05,
-      "loss": 0.294,
+      "epoch": 0.926586658017083,
+      "grad_norm": 0.1739431619644165,
+      "learning_rate": 4.690584480562726e-05,
+      "loss": 0.3859,
       "step": 25710
     },
     {
-      "epoch": 0.9,
-      "learning_rate": 4.7073285618304034e-05,
-      "loss": 0.2875,
+      "epoch": 0.9267668576783076,
+      "grad_norm": 0.17957325279712677,
+      "learning_rate": 4.6904438447967064e-05,
+      "loss": 0.429,
       "step": 25715
     },
     {
-      "epoch": 0.9,
-      "learning_rate": 4.7071947994736887e-05,
-      "loss": 0.2998,
+      "epoch": 0.9269470573395322,
+      "grad_norm": 0.16392971575260162,
+      "learning_rate": 4.690303179186313e-05,
+      "loss": 0.4385,
       "step": 25720
     },
     {
-      "epoch": 0.91,
-      "learning_rate": 4.707061008458004e-05,
-      "loss": 0.3073,
+      "epoch": 0.9271272570007568,
+      "grad_norm": 0.20912306010723114,
+      "learning_rate": 4.690162483733462e-05,
+      "loss": 0.4398,
       "step": 25725
     },
     {
-      "epoch": 0.91,
-      "learning_rate": 4.706927188785087e-05,
-      "loss": 0.2987,
+      "epoch": 0.9273074566619814,
+      "grad_norm": 0.18170976638793945,
+      "learning_rate": 4.690021758440072e-05,
+      "loss": 0.4031,
       "step": 25730
     },
     {
-      "epoch": 0.91,
-      "learning_rate": 4.706793340456674e-05,
-      "loss": 0.3124,
+      "epoch": 0.9274876563232061,
+      "grad_norm": 0.24062508344650269,
+      "learning_rate": 4.68988100330806e-05,
+      "loss": 0.4034,
       "step": 25735
     },
     {
-      "epoch": 0.91,
-      "learning_rate": 4.7066594634745055e-05,
-      "loss": 0.2986,
+      "epoch": 0.9276678559844308,
+      "grad_norm": 0.14883898198604584,
+      "learning_rate": 4.689740218339342e-05,
+      "loss": 0.4644,
       "step": 25740
     },
     {
-      "epoch": 0.91,
-      "learning_rate": 4.7065255578403165e-05,
-      "loss": 0.3156,
+      "epoch": 0.9278480556456554,
+      "grad_norm": 0.19792114198207855,
+      "learning_rate": 4.689599403535839e-05,
+      "loss": 0.4129,
       "step": 25745
     },
     {
-      "epoch": 0.91,
-      "learning_rate": 4.706391623555848e-05,
-      "loss": 0.3243,
+      "epoch": 0.92802825530688,
+      "grad_norm": 0.16648323833942413,
+      "learning_rate": 4.6894585588994676e-05,
+      "loss": 0.4379,
       "step": 25750
     },
     {
-      "epoch": 0.91,
-      "learning_rate": 4.7062576606228385e-05,
-      "loss": 0.2935,
+      "epoch": 0.9282084549681047,
+      "grad_norm": 0.18078407645225525,
+      "learning_rate": 4.689317684432147e-05,
+      "loss": 0.4586,
       "step": 25755
     },
     {
-      "epoch": 0.91,
-      "learning_rate": 4.706123669043027e-05,
-      "loss": 0.2948,
+      "epoch": 0.9283886546293293,
+      "grad_norm": 0.14701730012893677,
+      "learning_rate": 4.689176780135797e-05,
+      "loss": 0.4344,
       "step": 25760
     },
     {
-      "epoch": 0.91,
-      "learning_rate": 4.705989648818154e-05,
-      "loss": 0.3191,
+      "epoch": 0.9285688542905539,
+      "grad_norm": 0.1665094792842865,
+      "learning_rate": 4.689035846012336e-05,
+      "loss": 0.4549,
       "step": 25765
     },
     {
-      "epoch": 0.91,
-      "learning_rate": 4.705855599949959e-05,
-      "loss": 0.3178,
+      "epoch": 0.9287490539517785,
+      "grad_norm": 0.19282084703445435,
+      "learning_rate": 4.688894882063687e-05,
+      "loss": 0.4313,
       "step": 25770
     },
     {
-      "epoch": 0.91,
-      "learning_rate": 4.705721522440182e-05,
-      "loss": 0.2991,
+      "epoch": 0.9289292536130032,
+      "grad_norm": 0.18455705046653748,
+      "learning_rate": 4.688753888291768e-05,
+      "loss": 0.4461,
       "step": 25775
     },
     {
-      "epoch": 0.91,
-      "learning_rate": 4.705587416290565e-05,
-      "loss": 0.3112,
+      "epoch": 0.9291094532742279,
+      "grad_norm": 0.1790475845336914,
+      "learning_rate": 4.688612864698502e-05,
+      "loss": 0.4302,
       "step": 25780
     },
     {
-      "epoch": 0.91,
-      "learning_rate": 4.7054532815028494e-05,
-      "loss": 0.3057,
+      "epoch": 0.9292896529354525,
+      "grad_norm": 0.16020038723945618,
+      "learning_rate": 4.6884718112858085e-05,
+      "loss": 0.4316,
       "step": 25785
     },
     {
-      "epoch": 0.91,
-      "learning_rate": 4.705319118078776e-05,
-      "loss": 0.3145,
+      "epoch": 0.9294698525966771,
+      "grad_norm": 0.22309835255146027,
+      "learning_rate": 4.688330728055611e-05,
+      "loss": 0.4452,
       "step": 25790
     },
     {
-      "epoch": 0.91,
-      "learning_rate": 4.7051849260200874e-05,
-      "loss": 0.2953,
+      "epoch": 0.9296500522579018,
+      "grad_norm": 0.1724500060081482,
+      "learning_rate": 4.68818961500983e-05,
+      "loss": 0.4117,
       "step": 25795
     },
     {
-      "epoch": 0.91,
-      "learning_rate": 4.7050507053285256e-05,
-      "loss": 0.3187,
+      "epoch": 0.9298302519191264,
+      "grad_norm": 0.1488490104675293,
+      "learning_rate": 4.68804847215039e-05,
+      "loss": 0.4567,
       "step": 25800
     },
     {
-      "epoch": 0.91,
-      "learning_rate": 4.704916456005834e-05,
-      "loss": 0.3019,
+      "epoch": 0.930010451580351,
+      "grad_norm": 0.18638138473033905,
+      "learning_rate": 4.6879072994792126e-05,
+      "loss": 0.4115,
       "step": 25805
     },
     {
-      "epoch": 0.91,
-      "learning_rate": 4.704782178053755e-05,
-      "loss": 0.3052,
+      "epoch": 0.9301906512415756,
+      "grad_norm": 0.1752612441778183,
+      "learning_rate": 4.687766096998223e-05,
+      "loss": 0.4133,
       "step": 25810
     },
     {
-      "epoch": 0.91,
-      "learning_rate": 4.704647871474033e-05,
-      "loss": 0.3106,
+      "epoch": 0.9303708509028003,
+      "grad_norm": 0.17065556347370148,
+      "learning_rate": 4.6876248647093424e-05,
+      "loss": 0.4351,
       "step": 25815
     },
     {
-      "epoch": 0.91,
-      "learning_rate": 4.70451353626841e-05,
-      "loss": 0.3061,
+      "epoch": 0.930551050564025,
+      "grad_norm": 0.15603967010974884,
+      "learning_rate": 4.687483602614497e-05,
+      "loss": 0.4381,
       "step": 25820
     },
     {
-      "epoch": 0.91,
-      "learning_rate": 4.7043791724386324e-05,
-      "loss": 0.2958,
+      "epoch": 0.9307312502252496,
+      "grad_norm": 0.1808996945619583,
+      "learning_rate": 4.687342310715612e-05,
+      "loss": 0.4155,
       "step": 25825
     },
     {
-      "epoch": 0.91,
-      "learning_rate": 4.7042447799864444e-05,
-      "loss": 0.284,
+      "epoch": 0.9309114498864742,
+      "grad_norm": 0.17924173176288605,
+      "learning_rate": 4.687200989014611e-05,
+      "loss": 0.4348,
       "step": 25830
     },
     {
-      "epoch": 0.91,
-      "learning_rate": 4.70411035891359e-05,
-      "loss": 0.3004,
+      "epoch": 0.9310916495476989,
+      "grad_norm": 0.17122434079647064,
+      "learning_rate": 4.687059637513419e-05,
+      "loss": 0.4378,
       "step": 25835
     },
     {
-      "epoch": 0.91,
-      "learning_rate": 4.703975909221815e-05,
-      "loss": 0.3132,
+      "epoch": 0.9312718492089235,
+      "grad_norm": 0.1833307445049286,
+      "learning_rate": 4.686918256213964e-05,
+      "loss": 0.4553,
       "step": 25840
     },
     {
-      "epoch": 0.91,
-      "learning_rate": 4.703841430912865e-05,
-      "loss": 0.3095,
+      "epoch": 0.9314520488701481,
+      "grad_norm": 0.14123891294002533,
+      "learning_rate": 4.6867768451181706e-05,
+      "loss": 0.4414,
       "step": 25845
     },
     {
-      "epoch": 0.91,
-      "learning_rate": 4.703706923988488e-05,
-      "loss": 0.3004,
+      "epoch": 0.9316322485313727,
+      "grad_norm": 0.18901309370994568,
+      "learning_rate": 4.6866354042279666e-05,
+      "loss": 0.4157,
       "step": 25850
     },
     {
-      "epoch": 0.91,
-      "learning_rate": 4.703572388450427e-05,
-      "loss": 0.3101,
+      "epoch": 0.9318124481925975,
+      "grad_norm": 0.1698739230632782,
+      "learning_rate": 4.686493933545278e-05,
+      "loss": 0.4427,
       "step": 25855
     },
     {
-      "epoch": 0.91,
-      "learning_rate": 4.703437824300432e-05,
-      "loss": 0.3068,
+      "epoch": 0.9319926478538221,
+      "grad_norm": 0.14457178115844727,
+      "learning_rate": 4.686352433072033e-05,
+      "loss": 0.4243,
       "step": 25860
     },
     {
-      "epoch": 0.91,
-      "learning_rate": 4.703303231540248e-05,
-      "loss": 0.2825,
+      "epoch": 0.9321728475150467,
+      "grad_norm": 0.18088513612747192,
+      "learning_rate": 4.6862109028101596e-05,
+      "loss": 0.4398,
       "step": 25865
     },
     {
-      "epoch": 0.91,
-      "learning_rate": 4.7031686101716244e-05,
-      "loss": 0.3267,
+      "epoch": 0.9323530471762713,
+      "grad_norm": 0.17031700909137726,
+      "learning_rate": 4.686069342761585e-05,
+      "loss": 0.4415,
       "step": 25870
     },
     {
-      "epoch": 0.91,
-      "learning_rate": 4.703033960196308e-05,
-      "loss": 0.2971,
+      "epoch": 0.9325332468374959,
+      "grad_norm": 0.16327226161956787,
+      "learning_rate": 4.6859277529282406e-05,
+      "loss": 0.4316,
       "step": 25875
     },
     {
-      "epoch": 0.91,
-      "learning_rate": 4.702899281616048e-05,
-      "loss": 0.3243,
+      "epoch": 0.9327134464987206,
+      "grad_norm": 0.15043266117572784,
+      "learning_rate": 4.6857861333120525e-05,
+      "loss": 0.4595,
       "step": 25880
     },
     {
-      "epoch": 0.91,
-      "learning_rate": 4.702764574432592e-05,
-      "loss": 0.3012,
+      "epoch": 0.9328936461599452,
+      "grad_norm": 0.1847275346517563,
+      "learning_rate": 4.685644483914952e-05,
+      "loss": 0.4269,
       "step": 25885
     },
     {
-      "epoch": 0.91,
-      "learning_rate": 4.70262983864769e-05,
-      "loss": 0.3297,
+      "epoch": 0.9330738458211698,
+      "grad_norm": 0.1813531517982483,
+      "learning_rate": 4.685502804738868e-05,
+      "loss": 0.4171,
       "step": 25890
     },
     {
-      "epoch": 0.91,
-      "learning_rate": 4.7024950742630906e-05,
-      "loss": 0.2954,
+      "epoch": 0.9332540454823945,
+      "grad_norm": 0.16259461641311646,
+      "learning_rate": 4.685361095785732e-05,
+      "loss": 0.4319,
       "step": 25895
     },
     {
-      "epoch": 0.91,
-      "learning_rate": 4.7023602812805444e-05,
-      "loss": 0.2899,
+      "epoch": 0.9334342451436192,
+      "grad_norm": 0.17456987500190735,
+      "learning_rate": 4.685219357057474e-05,
+      "loss": 0.4362,
       "step": 25900
     },
     {
-      "epoch": 0.91,
-      "learning_rate": 4.702225459701802e-05,
-      "loss": 0.319,
+      "epoch": 0.9336144448048438,
+      "grad_norm": 0.1771409511566162,
+      "learning_rate": 4.6850775885560255e-05,
+      "loss": 0.4237,
       "step": 25905
     },
     {
-      "epoch": 0.91,
-      "learning_rate": 4.702090609528612e-05,
-      "loss": 0.3112,
+      "epoch": 0.9337946444660684,
+      "grad_norm": 0.14823564887046814,
+      "learning_rate": 4.684935790283318e-05,
+      "loss": 0.4253,
       "step": 25910
     },
     {
-      "epoch": 0.91,
-      "learning_rate": 4.7019557307627284e-05,
-      "loss": 0.2806,
+      "epoch": 0.933974844127293,
+      "grad_norm": 0.17045463621616364,
+      "learning_rate": 4.684793962241283e-05,
+      "loss": 0.4249,
       "step": 25915
     },
     {
-      "epoch": 0.91,
-      "learning_rate": 4.701820823405899e-05,
-      "loss": 0.2969,
+      "epoch": 0.9341550437885177,
+      "grad_norm": 0.2057085633277893,
+      "learning_rate": 4.684652104431852e-05,
+      "loss": 0.4473,
       "step": 25920
     },
     {
-      "epoch": 0.91,
-      "learning_rate": 4.7016858874598786e-05,
-      "loss": 0.2954,
+      "epoch": 0.9343352434497423,
+      "grad_norm": 0.1884680688381195,
+      "learning_rate": 4.684510216856961e-05,
+      "loss": 0.4832,
       "step": 25925
     },
     {
-      "epoch": 0.91,
-      "learning_rate": 4.7015509229264174e-05,
-      "loss": 0.3091,
+      "epoch": 0.9345154431109669,
+      "grad_norm": 0.15199509263038635,
+      "learning_rate": 4.684368299518541e-05,
+      "loss": 0.4648,
       "step": 25930
     },
     {
-      "epoch": 0.91,
-      "learning_rate": 4.701415929807268e-05,
-      "loss": 0.3093,
+      "epoch": 0.9346956427721916,
+      "grad_norm": 0.2007426768541336,
+      "learning_rate": 4.684226352418525e-05,
+      "loss": 0.465,
       "step": 25935
     },
     {
-      "epoch": 0.91,
-      "learning_rate": 4.7012809081041845e-05,
-      "loss": 0.2918,
+      "epoch": 0.9348758424334163,
+      "grad_norm": 0.19974127411842346,
+      "learning_rate": 4.684084375558848e-05,
+      "loss": 0.4523,
       "step": 25940
     },
     {
-      "epoch": 0.91,
-      "learning_rate": 4.701145857818918e-05,
-      "loss": 0.2836,
+      "epoch": 0.9350560420946409,
+      "grad_norm": 0.1712539941072464,
+      "learning_rate": 4.6839423689414455e-05,
+      "loss": 0.4529,
       "step": 25945
     },
     {
-      "epoch": 0.91,
-      "learning_rate": 4.7010107789532236e-05,
-      "loss": 0.2904,
+      "epoch": 0.9352362417558655,
+      "grad_norm": 0.19021165370941162,
+      "learning_rate": 4.68380033256825e-05,
+      "loss": 0.409,
       "step": 25950
     },
     {
-      "epoch": 0.91,
-      "learning_rate": 4.7008756715088554e-05,
-      "loss": 0.3015,
+      "epoch": 0.9354164414170901,
+      "grad_norm": 0.1501091569662094,
+      "learning_rate": 4.6836582664411975e-05,
+      "loss": 0.4259,
       "step": 25955
     },
     {
-      "epoch": 0.91,
-      "learning_rate": 4.700740535487566e-05,
-      "loss": 0.3074,
+      "epoch": 0.9355966410783148,
+      "grad_norm": 0.16499082744121552,
+      "learning_rate": 4.683516170562224e-05,
+      "loss": 0.4451,
       "step": 25960
     },
     {
-      "epoch": 0.91,
-      "learning_rate": 4.7006053708911115e-05,
-      "loss": 0.2716,
+      "epoch": 0.9357768407395394,
+      "grad_norm": 0.12097856402397156,
+      "learning_rate": 4.683374044933266e-05,
+      "loss": 0.4216,
       "step": 25965
     },
     {
-      "epoch": 0.91,
-      "learning_rate": 4.7004701777212465e-05,
-      "loss": 0.3119,
+      "epoch": 0.935957040400764,
+      "grad_norm": 0.17843008041381836,
+      "learning_rate": 4.683231889556259e-05,
+      "loss": 0.4188,
       "step": 25970
     },
     {
-      "epoch": 0.91,
-      "learning_rate": 4.700334955979726e-05,
-      "loss": 0.3131,
+      "epoch": 0.9361372400619887,
+      "grad_norm": 0.18024946749210358,
+      "learning_rate": 4.68308970443314e-05,
+      "loss": 0.4292,
       "step": 25975
     },
     {
-      "epoch": 0.91,
-      "learning_rate": 4.700199705668307e-05,
-      "loss": 0.2855,
+      "epoch": 0.9363174397232134,
+      "grad_norm": 0.18790417909622192,
+      "learning_rate": 4.6829474895658464e-05,
+      "loss": 0.4282,
       "step": 25980
     },
     {
-      "epoch": 0.91,
-      "learning_rate": 4.700064426788744e-05,
-      "loss": 0.3164,
+      "epoch": 0.936497639384438,
+      "grad_norm": 0.16022486984729767,
+      "learning_rate": 4.682805244956316e-05,
+      "loss": 0.4387,
       "step": 25985
     },
     {
-      "epoch": 0.91,
-      "learning_rate": 4.6999291193427954e-05,
-      "loss": 0.2818,
+      "epoch": 0.9366778390456626,
+      "grad_norm": 0.15988902747631073,
+      "learning_rate": 4.682662970606487e-05,
+      "loss": 0.4369,
       "step": 25990
     },
     {
-      "epoch": 0.91,
-      "learning_rate": 4.699793783332216e-05,
-      "loss": 0.3179,
+      "epoch": 0.9368580387068872,
+      "grad_norm": 0.1928686499595642,
+      "learning_rate": 4.682520666518297e-05,
+      "loss": 0.4324,
       "step": 25995
     },
     {
-      "epoch": 0.91,
-      "learning_rate": 4.699658418758764e-05,
-      "loss": 0.2999,
+      "epoch": 0.9370382383681118,
+      "grad_norm": 0.18302272260189056,
+      "learning_rate": 4.682378332693686e-05,
+      "loss": 0.4272,
       "step": 26000
     },
     {
-      "epoch": 0.91,
-      "eval_loss": 0.2956346273422241,
-      "eval_runtime": 10.5342,
-      "eval_samples_per_second": 9.493,
-      "eval_steps_per_second": 9.493,
+      "epoch": 0.9370382383681118,
+      "eval_loss": 0.45559266209602356,
+      "eval_runtime": 3.5347,
+      "eval_samples_per_second": 28.291,
+      "eval_steps_per_second": 7.073,
       "step": 26000
     },
     {
-      "epoch": 0.91,
-      "learning_rate": 4.6995230256241984e-05,
-      "loss": 0.3159,
+      "epoch": 0.9372184380293365,
+      "grad_norm": 0.15167242288589478,
+      "learning_rate": 4.682235969134593e-05,
+      "loss": 0.4238,
       "step": 26005
     },
     {
-      "epoch": 0.92,
-      "learning_rate": 4.699387603930274e-05,
-      "loss": 0.3136,
+      "epoch": 0.9373986376905611,
+      "grad_norm": 0.16070915758609772,
+      "learning_rate": 4.682093575842957e-05,
+      "loss": 0.4412,
       "step": 26010
     },
     {
-      "epoch": 0.92,
-      "learning_rate": 4.6992521536787525e-05,
-      "loss": 0.3187,
+      "epoch": 0.9375788373517858,
+      "grad_norm": 0.15172840654850006,
+      "learning_rate": 4.681951152820718e-05,
+      "loss": 0.4573,
       "step": 26015
     },
     {
-      "epoch": 0.92,
-      "learning_rate": 4.699116674871391e-05,
-      "loss": 0.3006,
+      "epoch": 0.9377590370130104,
+      "grad_norm": 0.1980069875717163,
+      "learning_rate": 4.6818087000698175e-05,
+      "loss": 0.4407,
       "step": 26020
     },
     {
-      "epoch": 0.92,
-      "learning_rate": 4.698981167509948e-05,
-      "loss": 0.3087,
+      "epoch": 0.9379392366742351,
+      "grad_norm": 0.17154526710510254,
+      "learning_rate": 4.6816662175921965e-05,
+      "loss": 0.4435,
       "step": 26025
     },
     {
-      "epoch": 0.92,
-      "learning_rate": 4.6988456315961835e-05,
-      "loss": 0.2858,
+      "epoch": 0.9381194363354597,
+      "grad_norm": 0.21349559724330902,
+      "learning_rate": 4.6815237053897946e-05,
+      "loss": 0.4316,
       "step": 26030
     },
     {
-      "epoch": 0.92,
-      "learning_rate": 4.698710067131859e-05,
-      "loss": 0.3003,
+      "epoch": 0.9382996359966843,
+      "grad_norm": 0.2062828242778778,
+      "learning_rate": 4.6813811634645554e-05,
+      "loss": 0.4443,
       "step": 26035
     },
     {
-      "epoch": 0.92,
-      "learning_rate": 4.6985744741187324e-05,
-      "loss": 0.2924,
+      "epoch": 0.9384798356579089,
+      "grad_norm": 0.16467832028865814,
+      "learning_rate": 4.68123859181842e-05,
+      "loss": 0.4363,
       "step": 26040
     },
     {
-      "epoch": 0.92,
-      "learning_rate": 4.698438852558564e-05,
-      "loss": 0.3158,
+      "epoch": 0.9386600353191336,
+      "grad_norm": 0.16752782464027405,
+      "learning_rate": 4.68109599045333e-05,
+      "loss": 0.4136,
       "step": 26045
     },
     {
-      "epoch": 0.92,
-      "learning_rate": 4.6983032024531173e-05,
-      "loss": 0.3382,
+      "epoch": 0.9388402349803583,
+      "grad_norm": 0.2054414600133896,
+      "learning_rate": 4.6809533593712305e-05,
+      "loss": 0.4687,
       "step": 26050
     },
     {
-      "epoch": 0.92,
-      "learning_rate": 4.698167523804152e-05,
-      "loss": 0.2996,
+      "epoch": 0.9390204346415829,
+      "grad_norm": 0.20668792724609375,
+      "learning_rate": 4.680810698574064e-05,
+      "loss": 0.4421,
       "step": 26055
     },
     {
-      "epoch": 0.92,
-      "learning_rate": 4.698031816613431e-05,
-      "loss": 0.3003,
+      "epoch": 0.9392006343028075,
+      "grad_norm": 0.1917664259672165,
+      "learning_rate": 4.680668008063773e-05,
+      "loss": 0.4454,
       "step": 26060
     },
     {
-      "epoch": 0.92,
-      "learning_rate": 4.697896080882714e-05,
-      "loss": 0.2955,
+      "epoch": 0.9393808339640322,
+      "grad_norm": 0.13623978197574615,
+      "learning_rate": 4.680525287842303e-05,
+      "loss": 0.4358,
       "step": 26065
     },
     {
-      "epoch": 0.92,
-      "learning_rate": 4.697760316613765e-05,
-      "loss": 0.3226,
+      "epoch": 0.9395610336252568,
+      "grad_norm": 0.1284467577934265,
+      "learning_rate": 4.6803825379115985e-05,
+      "loss": 0.4677,
       "step": 26070
     },
     {
-      "epoch": 0.92,
-      "learning_rate": 4.6976245238083474e-05,
-      "loss": 0.3099,
+      "epoch": 0.9397412332864814,
+      "grad_norm": 0.16711989045143127,
+      "learning_rate": 4.680239758273604e-05,
+      "loss": 0.4238,
       "step": 26075
     },
     {
-      "epoch": 0.92,
-      "learning_rate": 4.6974887024682236e-05,
-      "loss": 0.3085,
+      "epoch": 0.939921432947706,
+      "grad_norm": 0.1910637617111206,
+      "learning_rate": 4.6800969489302646e-05,
+      "loss": 0.4641,
       "step": 26080
     },
     {
-      "epoch": 0.92,
-      "learning_rate": 4.697352852595156e-05,
-      "loss": 0.3156,
+      "epoch": 0.9401016326089306,
+      "grad_norm": 0.16136784851551056,
+      "learning_rate": 4.6799541098835264e-05,
+      "loss": 0.4006,
       "step": 26085
     },
     {
-      "epoch": 0.92,
-      "learning_rate": 4.697216974190911e-05,
-      "loss": 0.3053,
+      "epoch": 0.9402818322701554,
+      "grad_norm": 0.1435597538948059,
+      "learning_rate": 4.679811241135335e-05,
+      "loss": 0.4049,
       "step": 26090
     },
     {
-      "epoch": 0.92,
-      "learning_rate": 4.6970810672572514e-05,
-      "loss": 0.2997,
+      "epoch": 0.94046203193138,
+      "grad_norm": 0.16800041496753693,
+      "learning_rate": 4.679668342687638e-05,
+      "loss": 0.411,
       "step": 26095
     },
     {
-      "epoch": 0.92,
-      "learning_rate": 4.6969451317959415e-05,
-      "loss": 0.3295,
+      "epoch": 0.9406422315926046,
+      "grad_norm": 0.19089345633983612,
+      "learning_rate": 4.679525414542382e-05,
+      "loss": 0.435,
       "step": 26100
     },
     {
-      "epoch": 0.92,
-      "learning_rate": 4.696809167808748e-05,
-      "loss": 0.3008,
+      "epoch": 0.9408224312538292,
+      "grad_norm": 0.22108210623264313,
+      "learning_rate": 4.6793824567015135e-05,
+      "loss": 0.438,
       "step": 26105
     },
     {
-      "epoch": 0.92,
-      "learning_rate": 4.696673175297433e-05,
-      "loss": 0.3188,
+      "epoch": 0.9410026309150539,
+      "grad_norm": 0.15207885205745697,
+      "learning_rate": 4.679239469166982e-05,
+      "loss": 0.4272,
       "step": 26110
     },
     {
-      "epoch": 0.92,
-      "learning_rate": 4.696537154263767e-05,
-      "loss": 0.3141,
+      "epoch": 0.9411828305762785,
+      "grad_norm": 0.19638319313526154,
+      "learning_rate": 4.679096451940734e-05,
+      "loss": 0.4027,
       "step": 26115
     },
     {
-      "epoch": 0.92,
-      "learning_rate": 4.696401104709512e-05,
-      "loss": 0.293,
+      "epoch": 0.9413630302375031,
+      "grad_norm": 0.22127874195575714,
+      "learning_rate": 4.678953405024718e-05,
+      "loss": 0.4261,
       "step": 26120
     },
     {
-      "epoch": 0.92,
-      "learning_rate": 4.696265026636437e-05,
-      "loss": 0.2927,
+      "epoch": 0.9415432298987277,
+      "grad_norm": 0.17142963409423828,
+      "learning_rate": 4.678810328420885e-05,
+      "loss": 0.4106,
       "step": 26125
     },
     {
-      "epoch": 0.92,
-      "learning_rate": 4.6961289200463077e-05,
-      "loss": 0.2989,
+      "epoch": 0.9417234295599525,
+      "grad_norm": 0.15688422322273254,
+      "learning_rate": 4.678667222131183e-05,
+      "loss": 0.4312,
       "step": 26130
     },
     {
-      "epoch": 0.92,
-      "learning_rate": 4.695992784940892e-05,
-      "loss": 0.2919,
+      "epoch": 0.9419036292211771,
+      "grad_norm": 0.1551523059606552,
+      "learning_rate": 4.678524086157561e-05,
+      "loss": 0.4178,
       "step": 26135
     },
     {
-      "epoch": 0.92,
-      "learning_rate": 4.695856621321957e-05,
-      "loss": 0.3212,
+      "epoch": 0.9420838288824017,
+      "grad_norm": 0.1784355640411377,
+      "learning_rate": 4.678380920501971e-05,
+      "loss": 0.4415,
       "step": 26140
     },
     {
-      "epoch": 0.92,
-      "learning_rate": 4.69572042919127e-05,
-      "loss": 0.3111,
+      "epoch": 0.9422640285436263,
+      "grad_norm": 0.14627361297607422,
+      "learning_rate": 4.6782377251663624e-05,
+      "loss": 0.4048,
       "step": 26145
     },
     {
-      "epoch": 0.92,
-      "learning_rate": 4.695584208550602e-05,
-      "loss": 0.2652,
+      "epoch": 0.942444228204851,
+      "grad_norm": 0.14835497736930847,
+      "learning_rate": 4.678094500152686e-05,
+      "loss": 0.4471,
       "step": 26150
     },
     {
-      "epoch": 0.92,
-      "learning_rate": 4.695447959401719e-05,
-      "loss": 0.2956,
+      "epoch": 0.9426244278660756,
+      "grad_norm": 0.15386229753494263,
+      "learning_rate": 4.677951245462895e-05,
+      "loss": 0.4297,
       "step": 26155
     },
     {
-      "epoch": 0.92,
-      "learning_rate": 4.695311681746391e-05,
-      "loss": 0.3111,
+      "epoch": 0.9428046275273002,
+      "grad_norm": 0.16428400576114655,
+      "learning_rate": 4.677807961098939e-05,
+      "loss": 0.4265,
       "step": 26160
     },
     {
-      "epoch": 0.92,
-      "learning_rate": 4.695175375586388e-05,
-      "loss": 0.3066,
+      "epoch": 0.9429848271885248,
+      "grad_norm": 0.18662281334400177,
+      "learning_rate": 4.677664647062771e-05,
+      "loss": 0.4396,
       "step": 26165
     },
     {
-      "epoch": 0.92,
-      "learning_rate": 4.695039040923479e-05,
-      "loss": 0.3114,
+      "epoch": 0.9431650268497496,
+      "grad_norm": 0.17339280247688293,
+      "learning_rate": 4.6775213033563445e-05,
+      "loss": 0.4346,
       "step": 26170
     },
     {
-      "epoch": 0.92,
-      "learning_rate": 4.694902677759435e-05,
-      "loss": 0.3438,
+      "epoch": 0.9433452265109742,
+      "grad_norm": 0.1801465004682541,
+      "learning_rate": 4.677377929981611e-05,
+      "loss": 0.4492,
       "step": 26175
     },
     {
-      "epoch": 0.92,
-      "learning_rate": 4.6947662860960257e-05,
-      "loss": 0.3166,
+      "epoch": 0.9435254261721988,
+      "grad_norm": 0.14782361686229706,
+      "learning_rate": 4.6772345269405255e-05,
+      "loss": 0.4003,
       "step": 26180
     },
     {
-      "epoch": 0.92,
-      "learning_rate": 4.694629865935023e-05,
-      "loss": 0.2927,
+      "epoch": 0.9437056258334234,
+      "grad_norm": 0.20001257956027985,
+      "learning_rate": 4.677091094235041e-05,
+      "loss": 0.4631,
       "step": 26185
     },
     {
-      "epoch": 0.92,
-      "learning_rate": 4.694493417278197e-05,
-      "loss": 0.3508,
+      "epoch": 0.9438858254946481,
+      "grad_norm": 0.16880610585212708,
+      "learning_rate": 4.6769476318671116e-05,
+      "loss": 0.4457,
       "step": 26190
     },
     {
-      "epoch": 0.92,
-      "learning_rate": 4.694356940127321e-05,
-      "loss": 0.3214,
+      "epoch": 0.9440660251558727,
+      "grad_norm": 0.18512383103370667,
+      "learning_rate": 4.676804139838692e-05,
+      "loss": 0.4106,
       "step": 26195
     },
     {
-      "epoch": 0.92,
-      "learning_rate": 4.6942204344841665e-05,
-      "loss": 0.3333,
+      "epoch": 0.9442462248170973,
+      "grad_norm": 0.19107241928577423,
+      "learning_rate": 4.6766606181517375e-05,
+      "loss": 0.417,
       "step": 26200
     },
     {
-      "epoch": 0.92,
-      "learning_rate": 4.694083900350505e-05,
-      "loss": 0.2991,
+      "epoch": 0.944426424478322,
+      "grad_norm": 0.18682074546813965,
+      "learning_rate": 4.676517066808204e-05,
+      "loss": 0.4276,
       "step": 26205
     },
     {
-      "epoch": 0.92,
-      "learning_rate": 4.6939473377281105e-05,
-      "loss": 0.2859,
+      "epoch": 0.9446066241395467,
+      "grad_norm": 0.19449475407600403,
+      "learning_rate": 4.676373485810046e-05,
+      "loss": 0.4347,
       "step": 26210
     },
     {
-      "epoch": 0.92,
-      "learning_rate": 4.6938107466187564e-05,
-      "loss": 0.3082,
+      "epoch": 0.9447868238007713,
+      "grad_norm": 0.16231991350650787,
+      "learning_rate": 4.6762298751592215e-05,
+      "loss": 0.4527,
       "step": 26215
     },
     {
-      "epoch": 0.92,
-      "learning_rate": 4.693674127024215e-05,
-      "loss": 0.2948,
+      "epoch": 0.9449670234619959,
+      "grad_norm": 0.19462081789970398,
+      "learning_rate": 4.676086234857686e-05,
+      "loss": 0.4579,
       "step": 26220
     },
     {
-      "epoch": 0.92,
-      "learning_rate": 4.69353747894626e-05,
-      "loss": 0.3076,
+      "epoch": 0.9451472231232205,
+      "grad_norm": 0.14375783503055573,
+      "learning_rate": 4.675942564907396e-05,
+      "loss": 0.4156,
       "step": 26225
     },
     {
-      "epoch": 0.92,
-      "learning_rate": 4.693400802386667e-05,
-      "loss": 0.31,
+      "epoch": 0.9453274227844451,
+      "grad_norm": 0.19837602972984314,
+      "learning_rate": 4.675798865310311e-05,
+      "loss": 0.4444,
       "step": 26230
     },
     {
-      "epoch": 0.92,
-      "learning_rate": 4.69326409734721e-05,
-      "loss": 0.287,
+      "epoch": 0.9455076224456698,
+      "grad_norm": 0.2307501882314682,
+      "learning_rate": 4.675655136068387e-05,
+      "loss": 0.4416,
       "step": 26235
     },
     {
-      "epoch": 0.92,
-      "learning_rate": 4.6931273638296635e-05,
-      "loss": 0.2953,
+      "epoch": 0.9456878221068944,
+      "grad_norm": 0.19053228199481964,
+      "learning_rate": 4.675511377183583e-05,
+      "loss": 0.4442,
       "step": 26240
     },
     {
-      "epoch": 0.92,
-      "learning_rate": 4.6929906018358046e-05,
-      "loss": 0.2955,
+      "epoch": 0.9458680217681191,
+      "grad_norm": 0.1901927888393402,
+      "learning_rate": 4.675367588657858e-05,
+      "loss": 0.4622,
       "step": 26245
     },
     {
-      "epoch": 0.92,
-      "learning_rate": 4.692853811367407e-05,
-      "loss": 0.302,
+      "epoch": 0.9460482214293437,
+      "grad_norm": 0.14448778331279755,
+      "learning_rate": 4.675223770493171e-05,
+      "loss": 0.4176,
       "step": 26250
     },
     {
-      "epoch": 0.92,
-      "learning_rate": 4.6927169924262486e-05,
-      "loss": 0.3084,
+      "epoch": 0.9462284210905684,
+      "grad_norm": 0.1635839194059372,
+      "learning_rate": 4.675079922691481e-05,
+      "loss": 0.4253,
       "step": 26255
     },
     {
-      "epoch": 0.92,
-      "learning_rate": 4.692580145014105e-05,
-      "loss": 0.3084,
+      "epoch": 0.946408620751793,
+      "grad_norm": 0.18405769765377045,
+      "learning_rate": 4.6749360452547485e-05,
+      "loss": 0.4426,
       "step": 26260
     },
     {
-      "epoch": 0.92,
-      "learning_rate": 4.6924432691327524e-05,
-      "loss": 0.2952,
+      "epoch": 0.9465888204130176,
+      "grad_norm": 0.15434348583221436,
+      "learning_rate": 4.674792138184933e-05,
+      "loss": 0.4832,
       "step": 26265
     },
     {
-      "epoch": 0.92,
-      "learning_rate": 4.6923063647839696e-05,
-      "loss": 0.2839,
+      "epoch": 0.9467690200742422,
+      "grad_norm": 0.14686670899391174,
+      "learning_rate": 4.674648201483995e-05,
+      "loss": 0.4396,
       "step": 26270
     },
     {
-      "epoch": 0.92,
-      "learning_rate": 4.692169431969533e-05,
-      "loss": 0.3139,
+      "epoch": 0.9469492197354669,
+      "grad_norm": 0.16869889199733734,
+      "learning_rate": 4.674504235153898e-05,
+      "loss": 0.4205,
       "step": 26275
     },
     {
-      "epoch": 0.92,
-      "learning_rate": 4.6920324706912215e-05,
-      "loss": 0.2949,
+      "epoch": 0.9471294193966915,
+      "grad_norm": 0.14260682463645935,
+      "learning_rate": 4.6743602391966004e-05,
+      "loss": 0.4266,
       "step": 26280
     },
     {
-      "epoch": 0.92,
-      "learning_rate": 4.691895480950813e-05,
-      "loss": 0.2875,
+      "epoch": 0.9473096190579162,
+      "grad_norm": 0.19560997188091278,
+      "learning_rate": 4.674216213614066e-05,
+      "loss": 0.4319,
       "step": 26285
     },
     {
-      "epoch": 0.92,
-      "learning_rate": 4.6917584627500855e-05,
-      "loss": 0.2957,
+      "epoch": 0.9474898187191408,
+      "grad_norm": 0.19733212888240814,
+      "learning_rate": 4.674072158408257e-05,
+      "loss": 0.4331,
       "step": 26290
     },
     {
-      "epoch": 0.93,
-      "learning_rate": 4.6916214160908194e-05,
-      "loss": 0.2871,
+      "epoch": 0.9476700183803655,
+      "grad_norm": 0.1900247186422348,
+      "learning_rate": 4.6739280735811355e-05,
+      "loss": 0.4258,
       "step": 26295
     },
     {
-      "epoch": 0.93,
-      "learning_rate": 4.691484340974793e-05,
-      "loss": 0.2999,
+      "epoch": 0.9478502180415901,
+      "grad_norm": 0.2038991004228592,
+      "learning_rate": 4.6737839591346645e-05,
+      "loss": 0.446,
       "step": 26300
     },
     {
-      "epoch": 0.93,
-      "learning_rate": 4.691347237403787e-05,
-      "loss": 0.3261,
+      "epoch": 0.9480304177028147,
+      "grad_norm": 0.19306516647338867,
+      "learning_rate": 4.6736398150708076e-05,
+      "loss": 0.4093,
       "step": 26305
     },
     {
-      "epoch": 0.93,
-      "learning_rate": 4.691210105379582e-05,
-      "loss": 0.3266,
+      "epoch": 0.9482106173640393,
+      "grad_norm": 0.15221892297267914,
+      "learning_rate": 4.67349564139153e-05,
+      "loss": 0.403,
       "step": 26310
     },
     {
-      "epoch": 0.93,
-      "learning_rate": 4.691072944903957e-05,
-      "loss": 0.2979,
+      "epoch": 0.948390817025264,
+      "grad_norm": 0.14861221611499786,
+      "learning_rate": 4.673351438098794e-05,
+      "loss": 0.4573,
       "step": 26315
     },
     {
-      "epoch": 0.93,
-      "learning_rate": 4.6909357559786935e-05,
-      "loss": 0.3069,
+      "epoch": 0.9485710166864886,
+      "grad_norm": 0.15452301502227783,
+      "learning_rate": 4.673207205194566e-05,
+      "loss": 0.4291,
       "step": 26320
     },
     {
-      "epoch": 0.93,
-      "learning_rate": 4.6907985386055735e-05,
-      "loss": 0.2832,
+      "epoch": 0.9487512163477133,
+      "grad_norm": 0.1777305006980896,
+      "learning_rate": 4.6730629426808114e-05,
+      "loss": 0.4437,
       "step": 26325
     },
     {
-      "epoch": 0.93,
-      "learning_rate": 4.6906612927863786e-05,
-      "loss": 0.2859,
+      "epoch": 0.9489314160089379,
+      "grad_norm": 0.17141470313072205,
+      "learning_rate": 4.6729186505594943e-05,
+      "loss": 0.4247,
       "step": 26330
     },
     {
-      "epoch": 0.93,
-      "learning_rate": 4.69052401852289e-05,
-      "loss": 0.3004,
+      "epoch": 0.9491116156701626,
+      "grad_norm": 0.21854913234710693,
+      "learning_rate": 4.672774328832581e-05,
+      "loss": 0.4053,
       "step": 26335
     },
     {
-      "epoch": 0.93,
-      "learning_rate": 4.690386715816891e-05,
-      "loss": 0.2796,
+      "epoch": 0.9492918153313872,
+      "grad_norm": 0.2022477239370346,
+      "learning_rate": 4.6726299775020385e-05,
+      "loss": 0.4414,
       "step": 26340
     },
     {
-      "epoch": 0.93,
-      "learning_rate": 4.6902493846701637e-05,
-      "loss": 0.2625,
+      "epoch": 0.9494720149926118,
+      "grad_norm": 0.12977053225040436,
+      "learning_rate": 4.672485596569833e-05,
+      "loss": 0.4409,
       "step": 26345
     },
     {
-      "epoch": 0.93,
-      "learning_rate": 4.690112025084492e-05,
-      "loss": 0.2902,
+      "epoch": 0.9496522146538364,
+      "grad_norm": 0.15741246938705444,
+      "learning_rate": 4.672341186037932e-05,
+      "loss": 0.4438,
       "step": 26350
     },
     {
-      "epoch": 0.93,
-      "learning_rate": 4.689974637061658e-05,
-      "loss": 0.294,
+      "epoch": 0.949832414315061,
+      "grad_norm": 0.1705687791109085,
+      "learning_rate": 4.672196745908303e-05,
+      "loss": 0.407,
       "step": 26355
     },
     {
-      "epoch": 0.93,
-      "learning_rate": 4.689837220603448e-05,
-      "loss": 0.297,
+      "epoch": 0.9500126139762857,
+      "grad_norm": 0.17730578780174255,
+      "learning_rate": 4.672052276182913e-05,
+      "loss": 0.4412,
       "step": 26360
     },
     {
-      "epoch": 0.93,
-      "learning_rate": 4.689699775711644e-05,
-      "loss": 0.2752,
+      "epoch": 0.9501928136375104,
+      "grad_norm": 0.14052340388298035,
+      "learning_rate": 4.671907776863732e-05,
+      "loss": 0.408,
       "step": 26365
     },
     {
-      "epoch": 0.93,
-      "learning_rate": 4.6895623023880313e-05,
-      "loss": 0.3214,
+      "epoch": 0.950373013298735,
+      "grad_norm": 0.15693028271198273,
+      "learning_rate": 4.671763247952728e-05,
+      "loss": 0.3939,
       "step": 26370
     },
     {
-      "epoch": 0.93,
-      "learning_rate": 4.689424800634395e-05,
-      "loss": 0.2921,
+      "epoch": 0.9505532129599596,
+      "grad_norm": 0.19423072040081024,
+      "learning_rate": 4.67161868945187e-05,
+      "loss": 0.4108,
       "step": 26375
     },
     {
-      "epoch": 0.93,
-      "learning_rate": 4.6892872704525213e-05,
-      "loss": 0.2828,
+      "epoch": 0.9507334126211843,
+      "grad_norm": 0.15230786800384521,
+      "learning_rate": 4.671474101363128e-05,
+      "loss": 0.4263,
       "step": 26380
     },
     {
-      "epoch": 0.93,
-      "learning_rate": 4.6891497118441954e-05,
-      "loss": 0.2992,
+      "epoch": 0.9509136122824089,
+      "grad_norm": 0.19921550154685974,
+      "learning_rate": 4.6713294836884716e-05,
+      "loss": 0.4178,
       "step": 26385
     },
     {
-      "epoch": 0.93,
-      "learning_rate": 4.689012124811202e-05,
-      "loss": 0.3126,
+      "epoch": 0.9510938119436335,
+      "grad_norm": 0.16791215538978577,
+      "learning_rate": 4.671184836429871e-05,
+      "loss": 0.4573,
       "step": 26390
     },
     {
-      "epoch": 0.93,
-      "learning_rate": 4.68887450935533e-05,
-      "loss": 0.3218,
+      "epoch": 0.9512740116048581,
+      "grad_norm": 0.17958621680736542,
+      "learning_rate": 4.6710401595892986e-05,
+      "loss": 0.4352,
       "step": 26395
     },
     {
-      "epoch": 0.93,
-      "learning_rate": 4.6887368654783634e-05,
-      "loss": 0.3072,
+      "epoch": 0.9514542112660829,
+      "grad_norm": 0.16500048339366913,
+      "learning_rate": 4.670895453168724e-05,
+      "loss": 0.4216,
       "step": 26400
     },
     {
-      "epoch": 0.93,
-      "learning_rate": 4.6885991931820925e-05,
-      "loss": 0.296,
+      "epoch": 0.9516344109273075,
+      "grad_norm": 0.1619076132774353,
+      "learning_rate": 4.670750717170119e-05,
+      "loss": 0.4489,
       "step": 26405
     },
     {
-      "epoch": 0.93,
-      "learning_rate": 4.6884614924683026e-05,
-      "loss": 0.2905,
+      "epoch": 0.9518146105885321,
+      "grad_norm": 0.14866675436496735,
+      "learning_rate": 4.6706059515954546e-05,
+      "loss": 0.4267,
       "step": 26410
     },
     {
-      "epoch": 0.93,
-      "learning_rate": 4.688323763338783e-05,
-      "loss": 0.2966,
+      "epoch": 0.9519948102497567,
+      "grad_norm": 0.17370857298374176,
+      "learning_rate": 4.670461156446706e-05,
+      "loss": 0.4023,
       "step": 26415
     },
     {
-      "epoch": 0.93,
-      "learning_rate": 4.6881860057953215e-05,
-      "loss": 0.3035,
+      "epoch": 0.9521750099109814,
+      "grad_norm": 0.23588383197784424,
+      "learning_rate": 4.6703163317258436e-05,
+      "loss": 0.4328,
       "step": 26420
     },
     {
-      "epoch": 0.93,
-      "learning_rate": 4.688048219839707e-05,
-      "loss": 0.3109,
+      "epoch": 0.952355209572206,
+      "grad_norm": 0.16146144270896912,
+      "learning_rate": 4.670171477434841e-05,
+      "loss": 0.402,
       "step": 26425
     },
     {
-      "epoch": 0.93,
-      "learning_rate": 4.6879104054737276e-05,
-      "loss": 0.3058,
+      "epoch": 0.9525354092334306,
+      "grad_norm": 0.1492733508348465,
+      "learning_rate": 4.670026593575673e-05,
+      "loss": 0.4807,
       "step": 26430
     },
     {
-      "epoch": 0.93,
-      "learning_rate": 4.687772562699174e-05,
-      "loss": 0.2764,
+      "epoch": 0.9527156088946552,
+      "grad_norm": 0.15443341434001923,
+      "learning_rate": 4.669881680150312e-05,
+      "loss": 0.4296,
       "step": 26435
     },
     {
-      "epoch": 0.93,
-      "learning_rate": 4.687634691517835e-05,
-      "loss": 0.3059,
+      "epoch": 0.95289580855588,
+      "grad_norm": 0.14959372580051422,
+      "learning_rate": 4.6697367371607334e-05,
+      "loss": 0.4148,
       "step": 26440
     },
     {
-      "epoch": 0.93,
-      "learning_rate": 4.6874967919315016e-05,
-      "loss": 0.2983,
+      "epoch": 0.9530760082171046,
+      "grad_norm": 0.18775780498981476,
+      "learning_rate": 4.669591764608913e-05,
+      "loss": 0.4036,
       "step": 26445
     },
     {
-      "epoch": 0.93,
-      "learning_rate": 4.687358863941964e-05,
-      "loss": 0.3109,
+      "epoch": 0.9532562078783292,
+      "grad_norm": 0.20374846458435059,
+      "learning_rate": 4.669446762496823e-05,
+      "loss": 0.4418,
       "step": 26450
     },
     {
-      "epoch": 0.93,
-      "learning_rate": 4.6872209075510134e-05,
-      "loss": 0.2926,
+      "epoch": 0.9534364075395538,
+      "grad_norm": 0.17194941639900208,
+      "learning_rate": 4.669301730826442e-05,
+      "loss": 0.4616,
       "step": 26455
     },
     {
-      "epoch": 0.93,
-      "learning_rate": 4.68708292276044e-05,
-      "loss": 0.3072,
+      "epoch": 0.9536166072007785,
+      "grad_norm": Infinity,
+      "learning_rate": 4.669185684209495e-05,
+      "loss": 0.4184,
       "step": 26460
     },
     {
-      "epoch": 0.93,
-      "learning_rate": 4.6869449095720364e-05,
-      "loss": 0.2999,
+      "epoch": 0.9537968068620031,
+      "grad_norm": 0.26384392380714417,
+      "learning_rate": 4.669040599339167e-05,
+      "loss": 0.4399,
       "step": 26465
     },
     {
-      "epoch": 0.93,
-      "learning_rate": 4.686806867987594e-05,
-      "loss": 0.2932,
+      "epoch": 0.9539770065232277,
+      "grad_norm": 0.16087359189987183,
+      "learning_rate": 4.6688954849160817e-05,
+      "loss": 0.4367,
       "step": 26470
     },
     {
-      "epoch": 0.93,
-      "learning_rate": 4.6866687980089064e-05,
-      "loss": 0.3021,
+      "epoch": 0.9541572061844523,
+      "grad_norm": 0.16282179951667786,
+      "learning_rate": 4.668750340942215e-05,
+      "loss": 0.4427,
       "step": 26475
     },
     {
-      "epoch": 0.93,
-      "learning_rate": 4.686530699637765e-05,
-      "loss": 0.3005,
+      "epoch": 0.9543374058456771,
+      "grad_norm": 0.14763139188289642,
+      "learning_rate": 4.6686051674195454e-05,
+      "loss": 0.4148,
       "step": 26480
     },
     {
-      "epoch": 0.93,
-      "learning_rate": 4.686392572875964e-05,
-      "loss": 0.2977,
+      "epoch": 0.9545176055069017,
+      "grad_norm": 0.17169588804244995,
+      "learning_rate": 4.66845996435005e-05,
+      "loss": 0.3915,
       "step": 26485
     },
     {
-      "epoch": 0.93,
-      "learning_rate": 4.686254417725296e-05,
-      "loss": 0.2818,
+      "epoch": 0.9546978051681263,
+      "grad_norm": 0.1894696205854416,
+      "learning_rate": 4.668314731735707e-05,
+      "loss": 0.4249,
       "step": 26490
     },
     {
-      "epoch": 0.93,
-      "learning_rate": 4.6861162341875556e-05,
-      "loss": 0.3014,
+      "epoch": 0.9548780048293509,
+      "grad_norm": 0.14295339584350586,
+      "learning_rate": 4.668169469578496e-05,
+      "loss": 0.4194,
       "step": 26495
     },
     {
-      "epoch": 0.93,
-      "learning_rate": 4.6859780222645355e-05,
-      "loss": 0.3317,
+      "epoch": 0.9550582044905755,
+      "grad_norm": 0.1540336161851883,
+      "learning_rate": 4.6680241778803955e-05,
+      "loss": 0.4442,
       "step": 26500
     },
     {
-      "epoch": 0.93,
-      "eval_loss": 0.29478785395622253,
-      "eval_runtime": 10.5324,
-      "eval_samples_per_second": 9.494,
-      "eval_steps_per_second": 9.494,
+      "epoch": 0.9550582044905755,
+      "eval_loss": 0.45503300428390503,
+      "eval_runtime": 3.5368,
+      "eval_samples_per_second": 28.274,
+      "eval_steps_per_second": 7.069,
       "step": 26500
     },
     {
-      "epoch": 0.93,
-      "learning_rate": 4.685839781958032e-05,
-      "loss": 0.3181,
+      "epoch": 0.9552384041518002,
+      "grad_norm": 0.12401163578033447,
+      "learning_rate": 4.6678788566433854e-05,
+      "loss": 0.4263,
       "step": 26505
     },
     {
-      "epoch": 0.93,
-      "learning_rate": 4.68570151326984e-05,
-      "loss": 0.3067,
+      "epoch": 0.9554186038130248,
+      "grad_norm": 0.1782900094985962,
+      "learning_rate": 4.6677335058694454e-05,
+      "loss": 0.4176,
       "step": 26510
     },
     {
-      "epoch": 0.93,
-      "learning_rate": 4.6855632162017535e-05,
-      "loss": 0.3111,
+      "epoch": 0.9555988034742494,
+      "grad_norm": 0.18796321749687195,
+      "learning_rate": 4.667588125560556e-05,
+      "loss": 0.4209,
       "step": 26515
     },
     {
-      "epoch": 0.93,
-      "learning_rate": 4.6854248907555695e-05,
-      "loss": 0.3103,
+      "epoch": 0.9557790031354741,
+      "grad_norm": 0.17212247848510742,
+      "learning_rate": 4.667442715718698e-05,
+      "loss": 0.4083,
       "step": 26520
     },
     {
-      "epoch": 0.93,
-      "learning_rate": 4.685286536933083e-05,
-      "loss": 0.3183,
+      "epoch": 0.9559592027966988,
+      "grad_norm": 0.1399925798177719,
+      "learning_rate": 4.667297276345853e-05,
+      "loss": 0.4253,
       "step": 26525
     },
     {
-      "epoch": 0.93,
-      "learning_rate": 4.685148154736091e-05,
-      "loss": 0.3011,
+      "epoch": 0.9561394024579234,
+      "grad_norm": 0.16920849680900574,
+      "learning_rate": 4.6671518074440025e-05,
+      "loss": 0.4156,
       "step": 26530
     },
     {
-      "epoch": 0.93,
-      "learning_rate": 4.685009744166391e-05,
-      "loss": 0.3188,
+      "epoch": 0.956319602119148,
+      "grad_norm": 0.22345153987407684,
+      "learning_rate": 4.6670063090151286e-05,
+      "loss": 0.4336,
       "step": 26535
     },
     {
-      "epoch": 0.93,
-      "learning_rate": 4.6848713052257786e-05,
-      "loss": 0.303,
+      "epoch": 0.9564998017803726,
+      "grad_norm": 0.1682891845703125,
+      "learning_rate": 4.666860781061212e-05,
+      "loss": 0.446,
       "step": 26540
     },
     {
-      "epoch": 0.93,
-      "learning_rate": 4.684732837916053e-05,
-      "loss": 0.3034,
+      "epoch": 0.9566800014415973,
+      "grad_norm": 0.15830697119235992,
+      "learning_rate": 4.666715223584237e-05,
+      "loss": 0.4206,
       "step": 26545
     },
     {
-      "epoch": 0.93,
-      "learning_rate": 4.6845943422390105e-05,
-      "loss": 0.2979,
+      "epoch": 0.9568602011028219,
+      "grad_norm": 0.16852952539920807,
+      "learning_rate": 4.6665696365861874e-05,
+      "loss": 0.4387,
       "step": 26550
     },
     {
-      "epoch": 0.93,
-      "learning_rate": 4.684455818196451e-05,
-      "loss": 0.2811,
+      "epoch": 0.9570404007640466,
+      "grad_norm": 0.13287803530693054,
+      "learning_rate": 4.666424020069045e-05,
+      "loss": 0.4461,
       "step": 26555
     },
     {
-      "epoch": 0.93,
-      "learning_rate": 4.6843172657901714e-05,
-      "loss": 0.2723,
+      "epoch": 0.9572206004252712,
+      "grad_norm": 0.15766523778438568,
+      "learning_rate": 4.666278374034795e-05,
+      "loss": 0.422,
       "step": 26560
     },
     {
-      "epoch": 0.93,
-      "learning_rate": 4.684178685021973e-05,
-      "loss": 0.3109,
+      "epoch": 0.9574008000864959,
+      "grad_norm": 0.1761545091867447,
+      "learning_rate": 4.6661326984854225e-05,
+      "loss": 0.4452,
       "step": 26565
     },
     {
-      "epoch": 0.93,
-      "learning_rate": 4.684040075893652e-05,
-      "loss": 0.2876,
+      "epoch": 0.9575809997477205,
+      "grad_norm": 0.17396709322929382,
+      "learning_rate": 4.6659869934229106e-05,
+      "loss": 0.4474,
       "step": 26570
     },
     {
-      "epoch": 0.93,
-      "learning_rate": 4.683901438407011e-05,
-      "loss": 0.2866,
+      "epoch": 0.9577611994089451,
+      "grad_norm": 0.16844242811203003,
+      "learning_rate": 4.665841258849245e-05,
+      "loss": 0.4374,
       "step": 26575
     },
     {
-      "epoch": 0.94,
-      "learning_rate": 4.683762772563849e-05,
-      "loss": 0.2905,
+      "epoch": 0.9579413990701697,
+      "grad_norm": 0.1774928718805313,
+      "learning_rate": 4.6656954947664125e-05,
+      "loss": 0.404,
       "step": 26580
     },
     {
-      "epoch": 0.94,
-      "learning_rate": 4.683624078365967e-05,
-      "loss": 0.2961,
+      "epoch": 0.9581215987313944,
+      "grad_norm": 0.16202402114868164,
+      "learning_rate": 4.665549701176397e-05,
+      "loss": 0.4328,
       "step": 26585
     },
     {
-      "epoch": 0.94,
-      "learning_rate": 4.683485355815165e-05,
-      "loss": 0.2906,
+      "epoch": 0.958301798392619,
+      "grad_norm": 0.2063705176115036,
+      "learning_rate": 4.6654038780811866e-05,
+      "loss": 0.458,
       "step": 26590
     },
     {
-      "epoch": 0.94,
-      "learning_rate": 4.683346604913244e-05,
-      "loss": 0.3036,
+      "epoch": 0.9584819980538437,
+      "grad_norm": 0.14139318466186523,
+      "learning_rate": 4.665258025482767e-05,
+      "loss": 0.3822,
       "step": 26595
     },
     {
-      "epoch": 0.94,
-      "learning_rate": 4.6832078256620065e-05,
-      "loss": 0.2968,
+      "epoch": 0.9586621977150683,
+      "grad_norm": 0.1882011443376541,
+      "learning_rate": 4.665112143383127e-05,
+      "loss": 0.424,
       "step": 26600
     },
     {
-      "epoch": 0.94,
-      "learning_rate": 4.683069018063254e-05,
-      "loss": 0.3245,
+      "epoch": 0.958842397376293,
+      "grad_norm": 0.1827850192785263,
+      "learning_rate": 4.664966231784253e-05,
+      "loss": 0.4365,
       "step": 26605
     },
     {
-      "epoch": 0.94,
-      "learning_rate": 4.68293018211879e-05,
-      "loss": 0.2988,
+      "epoch": 0.9590225970375176,
+      "grad_norm": 0.20430363714694977,
+      "learning_rate": 4.664820290688133e-05,
+      "loss": 0.4152,
       "step": 26610
     },
     {
-      "epoch": 0.94,
-      "learning_rate": 4.682791317830416e-05,
-      "loss": 0.2597,
+      "epoch": 0.9592027966987422,
+      "grad_norm": 0.21347634494304657,
+      "learning_rate": 4.664674320096756e-05,
+      "loss": 0.4451,
       "step": 26615
     },
     {
-      "epoch": 0.94,
-      "learning_rate": 4.6826524251999344e-05,
-      "loss": 0.3083,
+      "epoch": 0.9593829963599668,
+      "grad_norm": 0.17805293202400208,
+      "learning_rate": 4.66452832001211e-05,
+      "loss": 0.4693,
       "step": 26620
     },
     {
-      "epoch": 0.94,
-      "learning_rate": 4.6825135042291504e-05,
-      "loss": 0.2865,
+      "epoch": 0.9595631960211914,
+      "grad_norm": 0.17523930966854095,
+      "learning_rate": 4.664382290436185e-05,
+      "loss": 0.4101,
       "step": 26625
     },
     {
-      "epoch": 0.94,
-      "learning_rate": 4.682374554919866e-05,
-      "loss": 0.2915,
+      "epoch": 0.9597433956824161,
+      "grad_norm": 0.17870113253593445,
+      "learning_rate": 4.6642362313709706e-05,
+      "loss": 0.4164,
       "step": 26630
     },
     {
-      "epoch": 0.94,
-      "learning_rate": 4.6822355772738865e-05,
-      "loss": 0.2914,
+      "epoch": 0.9599235953436408,
+      "grad_norm": 0.20278626680374146,
+      "learning_rate": 4.664090142818456e-05,
+      "loss": 0.457,
       "step": 26635
     },
     {
-      "epoch": 0.94,
-      "learning_rate": 4.682124374755895e-05,
-      "loss": 0.3391,
+      "epoch": 0.9601037950048654,
+      "grad_norm": 0.2184595763683319,
+      "learning_rate": 4.663944024780632e-05,
+      "loss": 0.4481,
       "step": 26640
     },
     {
-      "epoch": 0.94,
-      "learning_rate": 4.681985346108412e-05,
-      "loss": 0.3162,
+      "epoch": 0.96028399466609,
+      "grad_norm": 0.174354687333107,
+      "learning_rate": 4.66379787725949e-05,
+      "loss": 0.443,
       "step": 26645
     },
     {
-      "epoch": 0.94,
-      "learning_rate": 4.681846289129287e-05,
-      "loss": 0.3289,
+      "epoch": 0.9604641943273147,
+      "grad_norm": 0.20273347198963165,
+      "learning_rate": 4.663651700257021e-05,
+      "loss": 0.4258,
       "step": 26650
     },
     {
-      "epoch": 0.94,
-      "learning_rate": 4.681707203820325e-05,
-      "loss": 0.2981,
+      "epoch": 0.9606443939885393,
+      "grad_norm": 0.1632155179977417,
+      "learning_rate": 4.6635054937752166e-05,
+      "loss": 0.4303,
       "step": 26655
     },
     {
-      "epoch": 0.94,
-      "learning_rate": 4.6815680901833334e-05,
-      "loss": 0.2916,
+      "epoch": 0.9608245936497639,
+      "grad_norm": 0.16507910192012787,
+      "learning_rate": 4.6633592578160687e-05,
+      "loss": 0.4055,
       "step": 26660
     },
     {
-      "epoch": 0.94,
-      "learning_rate": 4.681428948220118e-05,
-      "loss": 0.3133,
+      "epoch": 0.9610047933109885,
+      "grad_norm": 0.1875597983598709,
+      "learning_rate": 4.6632129923815694e-05,
+      "loss": 0.4617,
       "step": 26665
     },
     {
-      "epoch": 0.94,
-      "learning_rate": 4.681289777932485e-05,
-      "loss": 0.2867,
+      "epoch": 0.9611849929722132,
+      "grad_norm": 0.16827437281608582,
+      "learning_rate": 4.663066697473711e-05,
+      "loss": 0.4083,
       "step": 26670
     },
     {
-      "epoch": 0.94,
-      "learning_rate": 4.6811505793222426e-05,
-      "loss": 0.2839,
+      "epoch": 0.9613651926334379,
+      "grad_norm": 0.1666988730430603,
+      "learning_rate": 4.662920373094489e-05,
+      "loss": 0.4235,
       "step": 26675
     },
     {
-      "epoch": 0.94,
-      "learning_rate": 4.681011352391197e-05,
-      "loss": 0.2843,
+      "epoch": 0.9615453922946625,
+      "grad_norm": 0.18340694904327393,
+      "learning_rate": 4.662774019245896e-05,
+      "loss": 0.4255,
       "step": 26680
     },
     {
-      "epoch": 0.94,
-      "learning_rate": 4.6808720971411556e-05,
-      "loss": 0.299,
+      "epoch": 0.9617255919558871,
+      "grad_norm": 0.18287798762321472,
+      "learning_rate": 4.6626276359299245e-05,
+      "loss": 0.3855,
       "step": 26685
     },
     {
-      "epoch": 0.94,
-      "learning_rate": 4.6807328135739284e-05,
-      "loss": 0.3218,
+      "epoch": 0.9619057916171118,
+      "grad_norm": 0.1787051558494568,
+      "learning_rate": 4.662481223148571e-05,
+      "loss": 0.4594,
       "step": 26690
     },
     {
-      "epoch": 0.94,
-      "learning_rate": 4.6805935016913216e-05,
-      "loss": 0.3144,
+      "epoch": 0.9620859912783364,
+      "grad_norm": 0.1771918535232544,
+      "learning_rate": 4.662334780903829e-05,
+      "loss": 0.415,
       "step": 26695
     },
     {
-      "epoch": 0.94,
-      "learning_rate": 4.680454161495146e-05,
-      "loss": 0.3157,
+      "epoch": 0.962266190939561,
+      "grad_norm": 0.15356236696243286,
+      "learning_rate": 4.6621883091976945e-05,
+      "loss": 0.4285,
       "step": 26700
     },
     {
-      "epoch": 0.94,
-      "learning_rate": 4.68031479298721e-05,
-      "loss": 0.2977,
+      "epoch": 0.9624463906007856,
+      "grad_norm": 0.1645512878894806,
+      "learning_rate": 4.662041808032163e-05,
+      "loss": 0.465,
       "step": 26705
     },
     {
-      "epoch": 0.94,
-      "learning_rate": 4.680175396169323e-05,
-      "loss": 0.2742,
+      "epoch": 0.9626265902620104,
+      "grad_norm": 0.20833955705165863,
+      "learning_rate": 4.661895277409231e-05,
+      "loss": 0.4383,
       "step": 26710
     },
     {
-      "epoch": 0.94,
-      "learning_rate": 4.6800359710432955e-05,
-      "loss": 0.3116,
+      "epoch": 0.962806789923235,
+      "grad_norm": 0.15828882157802582,
+      "learning_rate": 4.661748717330893e-05,
+      "loss": 0.4052,
       "step": 26715
     },
     {
-      "epoch": 0.94,
-      "learning_rate": 4.679896517610938e-05,
-      "loss": 0.2944,
+      "epoch": 0.9629869895844596,
+      "grad_norm": 0.15166039764881134,
+      "learning_rate": 4.6616021277991476e-05,
+      "loss": 0.4495,
       "step": 26720
     },
     {
-      "epoch": 0.94,
-      "learning_rate": 4.67975703587406e-05,
-      "loss": 0.3056,
+      "epoch": 0.9631671892456842,
+      "grad_norm": 0.18399517238140106,
+      "learning_rate": 4.6614555088159924e-05,
+      "loss": 0.4607,
       "step": 26725
     },
     {
-      "epoch": 0.94,
-      "learning_rate": 4.679617525834474e-05,
-      "loss": 0.2851,
+      "epoch": 0.9633473889069089,
+      "grad_norm": 0.16991674900054932,
+      "learning_rate": 4.661308860383424e-05,
+      "loss": 0.4267,
       "step": 26730
     },
     {
-      "epoch": 0.94,
-      "learning_rate": 4.67947798749399e-05,
-      "loss": 0.29,
+      "epoch": 0.9635275885681335,
+      "grad_norm": 0.18746553361415863,
+      "learning_rate": 4.661162182503441e-05,
+      "loss": 0.456,
       "step": 26735
     },
     {
-      "epoch": 0.94,
-      "learning_rate": 4.679338420854422e-05,
-      "loss": 0.2968,
+      "epoch": 0.9637077882293581,
+      "grad_norm": 0.18917517364025116,
+      "learning_rate": 4.661015475178041e-05,
+      "loss": 0.4372,
       "step": 26740
     },
     {
-      "epoch": 0.94,
-      "learning_rate": 4.67919882591758e-05,
-      "loss": 0.3058,
+      "epoch": 0.9638879878905827,
+      "grad_norm": 0.16291776299476624,
+      "learning_rate": 4.6608687384092244e-05,
+      "loss": 0.4211,
       "step": 26745
     },
     {
-      "epoch": 0.94,
-      "learning_rate": 4.679059202685278e-05,
-      "loss": 0.3168,
+      "epoch": 0.9640681875518075,
+      "grad_norm": 0.17410635948181152,
+      "learning_rate": 4.660721972198989e-05,
+      "loss": 0.4218,
       "step": 26750
     },
     {
-      "epoch": 0.94,
-      "learning_rate": 4.678919551159327e-05,
-      "loss": 0.2959,
+      "epoch": 0.9642483872130321,
+      "grad_norm": 0.17787551879882812,
+      "learning_rate": 4.6605751765493354e-05,
+      "loss": 0.4457,
       "step": 26755
     },
     {
-      "epoch": 0.94,
-      "learning_rate": 4.678779871341543e-05,
-      "loss": 0.2937,
+      "epoch": 0.9644285868742567,
+      "grad_norm": 0.17140106856822968,
+      "learning_rate": 4.660428351462263e-05,
+      "loss": 0.4006,
       "step": 26760
     },
     {
-      "epoch": 0.94,
-      "learning_rate": 4.6786401632337376e-05,
-      "loss": 0.2977,
+      "epoch": 0.9646087865354813,
+      "grad_norm": 0.17782960832118988,
+      "learning_rate": 4.660281496939773e-05,
+      "loss": 0.4501,
       "step": 26765
     },
     {
-      "epoch": 0.94,
-      "learning_rate": 4.678500426837725e-05,
-      "loss": 0.3166,
+      "epoch": 0.9647889861967059,
+      "grad_norm": 0.15600888431072235,
+      "learning_rate": 4.6601346129838655e-05,
+      "loss": 0.4476,
       "step": 26770
     },
     {
-      "epoch": 0.94,
-      "learning_rate": 4.678360662155321e-05,
-      "loss": 0.3044,
+      "epoch": 0.9649691858579306,
+      "grad_norm": 0.18162627518177032,
+      "learning_rate": 4.6599876995965424e-05,
+      "loss": 0.418,
       "step": 26775
     },
     {
-      "epoch": 0.94,
-      "learning_rate": 4.678220869188339e-05,
-      "loss": 0.3031,
+      "epoch": 0.9651493855191552,
+      "grad_norm": 0.20125152170658112,
+      "learning_rate": 4.659840756779805e-05,
+      "loss": 0.4183,
       "step": 26780
     },
     {
-      "epoch": 0.94,
-      "learning_rate": 4.6780810479385945e-05,
-      "loss": 0.303,
+      "epoch": 0.9653295851803798,
+      "grad_norm": 0.16319221258163452,
+      "learning_rate": 4.6596937845356556e-05,
+      "loss": 0.4437,
       "step": 26785
     },
     {
-      "epoch": 0.94,
-      "learning_rate": 4.677941198407903e-05,
-      "loss": 0.307,
+      "epoch": 0.9655097848416045,
+      "grad_norm": 0.2335626482963562,
+      "learning_rate": 4.659546782866096e-05,
+      "loss": 0.4571,
       "step": 26790
     },
     {
-      "epoch": 0.94,
-      "learning_rate": 4.677801320598081e-05,
-      "loss": 0.3096,
+      "epoch": 0.9656899845028292,
+      "grad_norm": 0.17783339321613312,
+      "learning_rate": 4.6593997517731305e-05,
+      "loss": 0.4296,
       "step": 26795
     },
     {
-      "epoch": 0.94,
-      "learning_rate": 4.677661414510943e-05,
-      "loss": 0.3295,
+      "epoch": 0.9658701841640538,
+      "grad_norm": 0.16893276572227478,
+      "learning_rate": 4.65925269125876e-05,
+      "loss": 0.4179,
       "step": 26800
     },
     {
-      "epoch": 0.94,
-      "learning_rate": 4.677521480148307e-05,
-      "loss": 0.3012,
+      "epoch": 0.9660503838252784,
+      "grad_norm": 0.160037562251091,
+      "learning_rate": 4.6591056013249914e-05,
+      "loss": 0.4114,
       "step": 26805
     },
     {
-      "epoch": 0.94,
-      "learning_rate": 4.67738151751199e-05,
-      "loss": 0.2984,
+      "epoch": 0.966230583486503,
+      "grad_norm": 0.18261387944221497,
+      "learning_rate": 4.6589584819738254e-05,
+      "loss": 0.4149,
       "step": 26810
     },
     {
-      "epoch": 0.94,
-      "learning_rate": 4.677241526603808e-05,
-      "loss": 0.3075,
+      "epoch": 0.9664107831477277,
+      "grad_norm": 0.18064436316490173,
+      "learning_rate": 4.658811333207269e-05,
+      "loss": 0.4456,
       "step": 26815
     },
     {
-      "epoch": 0.94,
-      "learning_rate": 4.677101507425581e-05,
-      "loss": 0.2996,
+      "epoch": 0.9665909828089523,
+      "grad_norm": 0.16004979610443115,
+      "learning_rate": 4.658664155027326e-05,
+      "loss": 0.4362,
       "step": 26820
     },
     {
-      "epoch": 0.94,
-      "learning_rate": 4.6769614599791245e-05,
-      "loss": 0.3274,
+      "epoch": 0.9667711824701769,
+      "grad_norm": 0.18018919229507446,
+      "learning_rate": 4.658516947436001e-05,
+      "loss": 0.4137,
       "step": 26825
     },
     {
-      "epoch": 0.94,
-      "learning_rate": 4.676821384266259e-05,
-      "loss": 0.3016,
+      "epoch": 0.9669513821314016,
+      "grad_norm": 0.1572667360305786,
+      "learning_rate": 4.658369710435302e-05,
+      "loss": 0.4265,
       "step": 26830
     },
     {
-      "epoch": 0.94,
-      "learning_rate": 4.6766812802888016e-05,
-      "loss": 0.2927,
+      "epoch": 0.9671315817926263,
+      "grad_norm": 0.14630809426307678,
+      "learning_rate": 4.6582224440272325e-05,
+      "loss": 0.4567,
       "step": 26835
     },
     {
-      "epoch": 0.94,
-      "learning_rate": 4.676541148048573e-05,
-      "loss": 0.3203,
+      "epoch": 0.9673117814538509,
+      "grad_norm": 0.15697656571865082,
+      "learning_rate": 4.6580751482138e-05,
+      "loss": 0.4598,
       "step": 26840
     },
     {
-      "epoch": 0.94,
-      "learning_rate": 4.6764009875473914e-05,
-      "loss": 0.2974,
+      "epoch": 0.9674919811150755,
+      "grad_norm": 0.1883528083562851,
+      "learning_rate": 4.657927822997012e-05,
+      "loss": 0.4144,
       "step": 26845
     },
     {
-      "epoch": 0.94,
-      "learning_rate": 4.676260798787078e-05,
-      "loss": 0.3076,
+      "epoch": 0.9676721807763001,
+      "grad_norm": 0.1584809124469757,
+      "learning_rate": 4.657780468378875e-05,
+      "loss": 0.421,
       "step": 26850
     },
     {
-      "epoch": 0.94,
-      "learning_rate": 4.6761205817694514e-05,
-      "loss": 0.3237,
+      "epoch": 0.9678523804375248,
+      "grad_norm": 0.1841040551662445,
+      "learning_rate": 4.657633084361397e-05,
+      "loss": 0.4483,
       "step": 26855
     },
     {
-      "epoch": 0.95,
-      "learning_rate": 4.6759803364963336e-05,
-      "loss": 0.3246,
+      "epoch": 0.9680325800987494,
+      "grad_norm": 0.1718447059392929,
+      "learning_rate": 4.657485670946585e-05,
+      "loss": 0.4267,
       "step": 26860
     },
     {
-      "epoch": 0.95,
-      "learning_rate": 4.6758400629695456e-05,
-      "loss": 0.3098,
+      "epoch": 0.968212779759974,
+      "grad_norm": 0.16161082684993744,
+      "learning_rate": 4.65733822813645e-05,
+      "loss": 0.4503,
       "step": 26865
     },
     {
-      "epoch": 0.95,
-      "learning_rate": 4.6756997611909074e-05,
-      "loss": 0.3105,
+      "epoch": 0.9683929794211987,
+      "grad_norm": 0.175176739692688,
+      "learning_rate": 4.657190755932999e-05,
+      "loss": 0.4539,
       "step": 26870
     },
     {
-      "epoch": 0.95,
-      "learning_rate": 4.675559431162242e-05,
-      "loss": 0.2933,
+      "epoch": 0.9685731790824234,
+      "grad_norm": 0.16012953221797943,
+      "learning_rate": 4.65704325433824e-05,
+      "loss": 0.4453,
       "step": 26875
     },
     {
-      "epoch": 0.95,
-      "learning_rate": 4.675419072885372e-05,
-      "loss": 0.2906,
+      "epoch": 0.968753378743648,
+      "grad_norm": 0.16824325919151306,
+      "learning_rate": 4.6568957233541854e-05,
+      "loss": 0.4654,
       "step": 26880
     },
     {
-      "epoch": 0.95,
-      "learning_rate": 4.675278686362118e-05,
-      "loss": 0.2907,
+      "epoch": 0.9689335784048726,
+      "grad_norm": 0.15294811129570007,
+      "learning_rate": 4.6567481629828443e-05,
+      "loss": 0.4517,
       "step": 26885
     },
     {
-      "epoch": 0.95,
-      "learning_rate": 4.675138271594304e-05,
-      "loss": 0.3234,
+      "epoch": 0.9691137780660972,
+      "grad_norm": 0.19897432625293732,
+      "learning_rate": 4.6566005732262275e-05,
+      "loss": 0.4646,
       "step": 26890
     },
     {
-      "epoch": 0.95,
-      "learning_rate": 4.674997828583753e-05,
-      "loss": 0.3188,
+      "epoch": 0.9692939777273218,
+      "grad_norm": 0.16633781790733337,
+      "learning_rate": 4.6564529540863446e-05,
+      "loss": 0.3853,
       "step": 26895
     },
     {
-      "epoch": 0.95,
-      "learning_rate": 4.6748573573322884e-05,
-      "loss": 0.3043,
+      "epoch": 0.9694741773885465,
+      "grad_norm": 0.13114029169082642,
+      "learning_rate": 4.656305305565208e-05,
+      "loss": 0.4149,
       "step": 26900
     },
     {
-      "epoch": 0.95,
-      "learning_rate": 4.674716857841735e-05,
-      "loss": 0.2734,
+      "epoch": 0.9696543770497712,
+      "grad_norm": 0.17593474686145782,
+      "learning_rate": 4.656157627664829e-05,
+      "loss": 0.4126,
       "step": 26905
     },
     {
-      "epoch": 0.95,
-      "learning_rate": 4.6745763301139153e-05,
-      "loss": 0.2985,
+      "epoch": 0.9698345767109958,
+      "grad_norm": 0.1551685929298401,
+      "learning_rate": 4.6560099203872196e-05,
+      "loss": 0.4258,
       "step": 26910
     },
     {
-      "epoch": 0.95,
-      "learning_rate": 4.674435774150656e-05,
-      "loss": 0.3309,
+      "epoch": 0.9700147763722204,
+      "grad_norm": 0.18880245089530945,
+      "learning_rate": 4.655862183734392e-05,
+      "loss": 0.4148,
       "step": 26915
     },
     {
-      "epoch": 0.95,
-      "learning_rate": 4.674295189953781e-05,
-      "loss": 0.3117,
+      "epoch": 0.9701949760334451,
+      "grad_norm": 0.1571785807609558,
+      "learning_rate": 4.6557144177083604e-05,
+      "loss": 0.4471,
       "step": 26920
     },
     {
-      "epoch": 0.95,
-      "learning_rate": 4.674154577525116e-05,
-      "loss": 0.3057,
+      "epoch": 0.9703751756946697,
+      "grad_norm": 0.18303348124027252,
+      "learning_rate": 4.655566622311137e-05,
+      "loss": 0.4398,
       "step": 26925
     },
     {
-      "epoch": 0.95,
-      "learning_rate": 4.6740139368664866e-05,
-      "loss": 0.3029,
+      "epoch": 0.9705553753558943,
+      "grad_norm": 0.19541716575622559,
+      "learning_rate": 4.6554187975447364e-05,
+      "loss": 0.4099,
       "step": 26930
     },
     {
-      "epoch": 0.95,
-      "learning_rate": 4.673873267979718e-05,
-      "loss": 0.3176,
+      "epoch": 0.9707355750171189,
+      "grad_norm": 0.17658253014087677,
+      "learning_rate": 4.655270943411171e-05,
+      "loss": 0.4512,
       "step": 26935
     },
     {
-      "epoch": 0.95,
-      "learning_rate": 4.6737325708666394e-05,
-      "loss": 0.3005,
+      "epoch": 0.9709157746783436,
+      "grad_norm": 0.18457411229610443,
+      "learning_rate": 4.655123059912456e-05,
+      "loss": 0.4632,
       "step": 26940
     },
     {
-      "epoch": 0.95,
-      "learning_rate": 4.673591845529074e-05,
-      "loss": 0.3125,
+      "epoch": 0.9710959743395683,
+      "grad_norm": 0.15944646298885345,
+      "learning_rate": 4.654975147050607e-05,
+      "loss": 0.4436,
       "step": 26945
     },
     {
-      "epoch": 0.95,
-      "learning_rate": 4.6734510919688525e-05,
-      "loss": 0.3098,
+      "epoch": 0.9712761740007929,
+      "grad_norm": 0.17047551274299622,
+      "learning_rate": 4.654827204827639e-05,
+      "loss": 0.4476,
       "step": 26950
     },
     {
-      "epoch": 0.95,
-      "learning_rate": 4.6733103101878e-05,
-      "loss": 0.3302,
+      "epoch": 0.9714563736620175,
+      "grad_norm": 0.18987303972244263,
+      "learning_rate": 4.654679233245568e-05,
+      "loss": 0.4751,
       "step": 26955
     },
     {
-      "epoch": 0.95,
-      "learning_rate": 4.673169500187746e-05,
-      "loss": 0.2737,
+      "epoch": 0.9716365733232422,
+      "grad_norm": 0.18949106335639954,
+      "learning_rate": 4.654531232306409e-05,
+      "loss": 0.432,
       "step": 26960
     },
     {
-      "epoch": 0.95,
-      "learning_rate": 4.673028661970518e-05,
-      "loss": 0.2857,
+      "epoch": 0.9718167729844668,
+      "grad_norm": 0.15808527171611786,
+      "learning_rate": 4.654383202012179e-05,
+      "loss": 0.3919,
       "step": 26965
     },
     {
-      "epoch": 0.95,
-      "learning_rate": 4.672887795537945e-05,
-      "loss": 0.3199,
+      "epoch": 0.9719969726456914,
+      "grad_norm": 0.1749500036239624,
+      "learning_rate": 4.654235142364895e-05,
+      "loss": 0.4509,
       "step": 26970
     },
     {
-      "epoch": 0.95,
-      "learning_rate": 4.672746900891856e-05,
-      "loss": 0.2955,
+      "epoch": 0.972177172306916,
+      "grad_norm": 0.1732529252767563,
+      "learning_rate": 4.654087053366575e-05,
+      "loss": 0.4012,
       "step": 26975
     },
     {
-      "epoch": 0.95,
-      "learning_rate": 4.672605978034079e-05,
-      "loss": 0.2969,
+      "epoch": 0.9723573719681406,
+      "grad_norm": 0.17838901281356812,
+      "learning_rate": 4.653938935019235e-05,
+      "loss": 0.4284,
       "step": 26980
     },
     {
-      "epoch": 0.95,
-      "learning_rate": 4.6724650269664465e-05,
-      "loss": 0.3046,
+      "epoch": 0.9725375716293654,
+      "grad_norm": 0.16471531987190247,
+      "learning_rate": 4.653790787324894e-05,
+      "loss": 0.4306,
       "step": 26985
     },
     {
-      "epoch": 0.95,
-      "learning_rate": 4.672324047690787e-05,
-      "loss": 0.3196,
+      "epoch": 0.97271777129059,
+      "grad_norm": 0.19036082923412323,
+      "learning_rate": 4.6536426102855714e-05,
+      "loss": 0.4657,
       "step": 26990
     },
     {
-      "epoch": 0.95,
-      "learning_rate": 4.672183040208932e-05,
-      "loss": 0.2903,
+      "epoch": 0.9728979709518146,
+      "grad_norm": 0.15139730274677277,
+      "learning_rate": 4.6534944039032845e-05,
+      "loss": 0.4456,
       "step": 26995
     },
     {
-      "epoch": 0.95,
-      "learning_rate": 4.672042004522711e-05,
-      "loss": 0.3032,
+      "epoch": 0.9730781706130393,
+      "grad_norm": 0.17167389392852783,
+      "learning_rate": 4.6533461681800534e-05,
+      "loss": 0.4286,
       "step": 27000
     },
     {
-      "epoch": 0.95,
-      "eval_loss": 0.29417115449905396,
-      "eval_runtime": 10.5309,
-      "eval_samples_per_second": 9.496,
-      "eval_steps_per_second": 9.496,
+      "epoch": 0.9730781706130393,
+      "eval_loss": 0.45476090908050537,
+      "eval_runtime": 3.5374,
+      "eval_samples_per_second": 28.269,
+      "eval_steps_per_second": 7.067,
       "step": 27000
     },
     {
-      "epoch": 0.95,
-      "learning_rate": 4.671900940633956e-05,
-      "loss": 0.3257,
+      "epoch": 0.9732583702742639,
+      "grad_norm": 0.17848706245422363,
+      "learning_rate": 4.6531979031178975e-05,
+      "loss": 0.3916,
       "step": 27005
     },
     {
-      "epoch": 0.95,
-      "learning_rate": 4.671759848544499e-05,
-      "loss": 0.3167,
+      "epoch": 0.9734385699354885,
+      "grad_norm": 0.15405860543251038,
+      "learning_rate": 4.6530496087188374e-05,
+      "loss": 0.4158,
       "step": 27010
     },
     {
-      "epoch": 0.95,
-      "learning_rate": 4.671618728256172e-05,
-      "loss": 0.2893,
+      "epoch": 0.9736187695967131,
+      "grad_norm": 0.17654937505722046,
+      "learning_rate": 4.652901284984893e-05,
+      "loss": 0.4398,
       "step": 27015
     },
     {
-      "epoch": 0.95,
-      "learning_rate": 4.671477579770806e-05,
-      "loss": 0.2664,
+      "epoch": 0.9737989692579377,
+      "grad_norm": 0.17176496982574463,
+      "learning_rate": 4.652752931918085e-05,
+      "loss": 0.4457,
       "step": 27020
     },
     {
-      "epoch": 0.95,
-      "learning_rate": 4.671336403090235e-05,
-      "loss": 0.3067,
+      "epoch": 0.9739791689191625,
+      "grad_norm": 0.1502595841884613,
+      "learning_rate": 4.652604549520436e-05,
+      "loss": 0.445,
       "step": 27025
     },
     {
-      "epoch": 0.95,
-      "learning_rate": 4.671195198216293e-05,
-      "loss": 0.3055,
+      "epoch": 0.9741593685803871,
+      "grad_norm": 0.14152762293815613,
+      "learning_rate": 4.652456137793966e-05,
+      "loss": 0.3946,
       "step": 27030
     },
     {
-      "epoch": 0.95,
-      "learning_rate": 4.67105396515081e-05,
-      "loss": 0.2689,
+      "epoch": 0.9743395682416117,
+      "grad_norm": 0.18833976984024048,
+      "learning_rate": 4.6523076967406984e-05,
+      "loss": 0.4336,
       "step": 27035
     },
     {
-      "epoch": 0.95,
-      "learning_rate": 4.6709127038956235e-05,
-      "loss": 0.299,
+      "epoch": 0.9745197679028363,
+      "grad_norm": 0.14718258380889893,
+      "learning_rate": 4.652159226362655e-05,
+      "loss": 0.4079,
       "step": 27040
     },
     {
-      "epoch": 0.95,
-      "learning_rate": 4.670771414452566e-05,
-      "loss": 0.3221,
+      "epoch": 0.974699967564061,
+      "grad_norm": 0.16922341287136078,
+      "learning_rate": 4.652010726661858e-05,
+      "loss": 0.4368,
       "step": 27045
     },
     {
-      "epoch": 0.95,
-      "learning_rate": 4.670630096823472e-05,
-      "loss": 0.3033,
+      "epoch": 0.9748801672252856,
+      "grad_norm": 0.18971046805381775,
+      "learning_rate": 4.6518621976403335e-05,
+      "loss": 0.4514,
       "step": 27050
     },
     {
-      "epoch": 0.95,
-      "learning_rate": 4.670488751010177e-05,
-      "loss": 0.2964,
+      "epoch": 0.9750603668865102,
+      "grad_norm": 0.15908949077129364,
+      "learning_rate": 4.6517136393001015e-05,
+      "loss": 0.4215,
       "step": 27055
     },
     {
-      "epoch": 0.95,
-      "learning_rate": 4.6703473770145155e-05,
-      "loss": 0.3276,
+      "epoch": 0.9752405665477349,
+      "grad_norm": 0.15814034640789032,
+      "learning_rate": 4.651565051643188e-05,
+      "loss": 0.4227,
       "step": 27060
     },
     {
-      "epoch": 0.95,
-      "learning_rate": 4.6702059748383244e-05,
-      "loss": 0.2896,
+      "epoch": 0.9754207662089596,
+      "grad_norm": 0.19080659747123718,
+      "learning_rate": 4.651416434671617e-05,
+      "loss": 0.4194,
       "step": 27065
     },
     {
-      "epoch": 0.95,
-      "learning_rate": 4.670064544483438e-05,
-      "loss": 0.3221,
+      "epoch": 0.9756009658701842,
+      "grad_norm": 0.17519068717956543,
+      "learning_rate": 4.651267788387415e-05,
+      "loss": 0.423,
       "step": 27070
     },
     {
-      "epoch": 0.95,
-      "learning_rate": 4.669923085951694e-05,
-      "loss": 0.3254,
+      "epoch": 0.9757811655314088,
+      "grad_norm": 0.18204259872436523,
+      "learning_rate": 4.651119112792604e-05,
+      "loss": 0.4161,
       "step": 27075
     },
     {
-      "epoch": 0.95,
-      "learning_rate": 4.6697815992449287e-05,
-      "loss": 0.2988,
+      "epoch": 0.9759613651926334,
+      "grad_norm": 0.17226997017860413,
+      "learning_rate": 4.6509704078892124e-05,
+      "loss": 0.4583,
       "step": 27080
     },
     {
-      "epoch": 0.95,
-      "learning_rate": 4.66964008436498e-05,
-      "loss": 0.2927,
+      "epoch": 0.9761415648538581,
+      "grad_norm": 0.15808525681495667,
+      "learning_rate": 4.650821673679265e-05,
+      "loss": 0.4533,
       "step": 27085
     },
     {
-      "epoch": 0.95,
-      "learning_rate": 4.6694985413136835e-05,
-      "loss": 0.3356,
+      "epoch": 0.9763217645150827,
+      "grad_norm": 0.15862226486206055,
+      "learning_rate": 4.6506729101647897e-05,
+      "loss": 0.3829,
       "step": 27090
     },
     {
-      "epoch": 0.95,
-      "learning_rate": 4.6693569700928795e-05,
-      "loss": 0.3185,
+      "epoch": 0.9765019641763073,
+      "grad_norm": 0.18071158230304718,
+      "learning_rate": 4.650524117347812e-05,
+      "loss": 0.4365,
       "step": 27095
     },
     {
-      "epoch": 0.95,
-      "learning_rate": 4.669215370704404e-05,
-      "loss": 0.2984,
+      "epoch": 0.976682163837532,
+      "grad_norm": 0.2250252366065979,
+      "learning_rate": 4.650375295230359e-05,
+      "loss": 0.453,
       "step": 27100
     },
     {
-      "epoch": 0.95,
-      "learning_rate": 4.669073743150096e-05,
-      "loss": 0.3017,
+      "epoch": 0.9768623634987567,
+      "grad_norm": 0.182828888297081,
+      "learning_rate": 4.6502264438144596e-05,
+      "loss": 0.4589,
       "step": 27105
     },
     {
-      "epoch": 0.95,
-      "learning_rate": 4.668932087431797e-05,
-      "loss": 0.3103,
+      "epoch": 0.9770425631599813,
+      "grad_norm": 0.18622487783432007,
+      "learning_rate": 4.650077563102141e-05,
+      "loss": 0.4099,
       "step": 27110
     },
     {
-      "epoch": 0.95,
-      "learning_rate": 4.6687904035513416e-05,
-      "loss": 0.2788,
+      "epoch": 0.9772227628212059,
+      "grad_norm": 0.16842947900295258,
+      "learning_rate": 4.6499286530954314e-05,
+      "loss": 0.4102,
       "step": 27115
     },
     {
-      "epoch": 0.95,
-      "learning_rate": 4.668648691510574e-05,
-      "loss": 0.2997,
+      "epoch": 0.9774029624824305,
+      "grad_norm": 0.17270709574222565,
+      "learning_rate": 4.649779713796361e-05,
+      "loss": 0.4312,
       "step": 27120
     },
     {
-      "epoch": 0.95,
-      "learning_rate": 4.6685069513113315e-05,
-      "loss": 0.2898,
+      "epoch": 0.9775831621436551,
+      "grad_norm": 0.1607503890991211,
+      "learning_rate": 4.649630745206958e-05,
+      "loss": 0.4738,
       "step": 27125
     },
     {
-      "epoch": 0.95,
-      "learning_rate": 4.668365182955455e-05,
-      "loss": 0.2964,
+      "epoch": 0.9777633618048798,
+      "grad_norm": 0.13477887213230133,
+      "learning_rate": 4.649481747329252e-05,
+      "loss": 0.4357,
       "step": 27130
     },
     {
-      "epoch": 0.95,
-      "learning_rate": 4.6682233864447865e-05,
-      "loss": 0.2835,
+      "epoch": 0.9779435614661044,
+      "grad_norm": 0.16214247047901154,
+      "learning_rate": 4.6493327201652725e-05,
+      "loss": 0.3851,
       "step": 27135
     },
     {
-      "epoch": 0.95,
-      "learning_rate": 4.668081561781166e-05,
-      "loss": 0.3101,
+      "epoch": 0.9781237611273291,
+      "grad_norm": 0.19764074683189392,
+      "learning_rate": 4.649183663717052e-05,
+      "loss": 0.4212,
       "step": 27140
     },
     {
-      "epoch": 0.96,
-      "learning_rate": 4.667939708966435e-05,
-      "loss": 0.3146,
+      "epoch": 0.9783039607885538,
+      "grad_norm": 0.169803187251091,
+      "learning_rate": 4.6490345779866197e-05,
+      "loss": 0.4267,
       "step": 27145
     },
     {
-      "epoch": 0.96,
-      "learning_rate": 4.667797828002436e-05,
-      "loss": 0.2676,
+      "epoch": 0.9784841604497784,
+      "grad_norm": 0.16970309615135193,
+      "learning_rate": 4.6488854629760074e-05,
+      "loss": 0.4518,
       "step": 27150
     },
     {
-      "epoch": 0.96,
-      "learning_rate": 4.66765591889101e-05,
-      "loss": 0.2785,
+      "epoch": 0.978664360111003,
+      "grad_norm": 0.15592272579669952,
+      "learning_rate": 4.648736318687247e-05,
+      "loss": 0.4588,
       "step": 27155
     },
     {
-      "epoch": 0.96,
-      "learning_rate": 4.667513981634001e-05,
-      "loss": 0.3098,
+      "epoch": 0.9788445597722276,
+      "grad_norm": 0.17874795198440552,
+      "learning_rate": 4.64858714512237e-05,
+      "loss": 0.4404,
       "step": 27160
     },
     {
-      "epoch": 0.96,
-      "learning_rate": 4.667372016233251e-05,
-      "loss": 0.2952,
+      "epoch": 0.9790247594334522,
+      "grad_norm": 0.15960319340229034,
+      "learning_rate": 4.648437942283409e-05,
+      "loss": 0.408,
       "step": 27165
     },
     {
-      "epoch": 0.96,
-      "learning_rate": 4.667230022690605e-05,
-      "loss": 0.3123,
+      "epoch": 0.9792049590946769,
+      "grad_norm": 0.17572569847106934,
+      "learning_rate": 4.6482887101723974e-05,
+      "loss": 0.4118,
       "step": 27170
     },
     {
-      "epoch": 0.96,
-      "learning_rate": 4.667088001007904e-05,
-      "loss": 0.2965,
+      "epoch": 0.9793851587559015,
+      "grad_norm": 0.20097064971923828,
+      "learning_rate": 4.6481394487913673e-05,
+      "loss": 0.4195,
       "step": 27175
     },
     {
-      "epoch": 0.96,
-      "learning_rate": 4.666945951186994e-05,
-      "loss": 0.3047,
+      "epoch": 0.9795653584171262,
+      "grad_norm": 0.18549709022045135,
+      "learning_rate": 4.647990158142354e-05,
+      "loss": 0.4298,
       "step": 27180
     },
     {
-      "epoch": 0.96,
-      "learning_rate": 4.666803873229719e-05,
-      "loss": 0.2921,
+      "epoch": 0.9797455580783508,
+      "grad_norm": 0.17674687504768372,
+      "learning_rate": 4.6478408382273905e-05,
+      "loss": 0.441,
       "step": 27185
     },
     {
-      "epoch": 0.96,
-      "learning_rate": 4.666661767137923e-05,
-      "loss": 0.3198,
+      "epoch": 0.9799257577395755,
+      "grad_norm": 0.14485035836696625,
+      "learning_rate": 4.6476914890485114e-05,
+      "loss": 0.4422,
       "step": 27190
     },
     {
-      "epoch": 0.96,
-      "learning_rate": 4.666519632913453e-05,
-      "loss": 0.2968,
+      "epoch": 0.9801059574008001,
+      "grad_norm": 0.1638946533203125,
+      "learning_rate": 4.647542110607751e-05,
+      "loss": 0.4341,
       "step": 27195
     },
     {
-      "epoch": 0.96,
-      "learning_rate": 4.666377470558152e-05,
-      "loss": 0.2936,
+      "epoch": 0.9802861570620247,
+      "grad_norm": 0.1771542876958847,
+      "learning_rate": 4.6473927029071454e-05,
+      "loss": 0.4298,
       "step": 27200
     },
     {
-      "epoch": 0.96,
-      "learning_rate": 4.6662352800738684e-05,
-      "loss": 0.281,
+      "epoch": 0.9804663567232493,
+      "grad_norm": 0.16256387531757355,
+      "learning_rate": 4.6472432659487296e-05,
+      "loss": 0.4271,
       "step": 27205
     },
     {
-      "epoch": 0.96,
-      "learning_rate": 4.666093061462447e-05,
-      "loss": 0.2865,
+      "epoch": 0.980646556384474,
+      "grad_norm": 0.15982799232006073,
+      "learning_rate": 4.647093799734541e-05,
+      "loss": 0.4554,
       "step": 27210
     },
     {
-      "epoch": 0.96,
-      "learning_rate": 4.665950814725734e-05,
-      "loss": 0.2759,
+      "epoch": 0.9808267560456987,
+      "grad_norm": 0.19146406650543213,
+      "learning_rate": 4.646944304266615e-05,
+      "loss": 0.4195,
       "step": 27215
     },
     {
-      "epoch": 0.96,
-      "learning_rate": 4.6658085398655784e-05,
-      "loss": 0.2907,
+      "epoch": 0.9810069557069233,
+      "grad_norm": 0.17436739802360535,
+      "learning_rate": 4.646794779546988e-05,
+      "loss": 0.4558,
       "step": 27220
     },
     {
-      "epoch": 0.96,
-      "learning_rate": 4.6656662368838247e-05,
-      "loss": 0.291,
+      "epoch": 0.9811871553681479,
+      "grad_norm": 0.17052249610424042,
+      "learning_rate": 4.6466452255776976e-05,
+      "loss": 0.4448,
       "step": 27225
     },
     {
-      "epoch": 0.96,
-      "learning_rate": 4.665523905782323e-05,
-      "loss": 0.3372,
+      "epoch": 0.9813673550293726,
+      "grad_norm": 0.1632729321718216,
+      "learning_rate": 4.646495642360782e-05,
+      "loss": 0.4211,
       "step": 27230
     },
     {
-      "epoch": 0.96,
-      "learning_rate": 4.665381546562921e-05,
-      "loss": 0.2823,
+      "epoch": 0.9815475546905972,
+      "grad_norm": 0.2103988230228424,
+      "learning_rate": 4.6463460298982787e-05,
+      "loss": 0.431,
       "step": 27235
     },
     {
-      "epoch": 0.96,
-      "learning_rate": 4.665239159227466e-05,
-      "loss": 0.3084,
+      "epoch": 0.9817277543518218,
+      "grad_norm": 0.19348739087581635,
+      "learning_rate": 4.646196388192226e-05,
+      "loss": 0.4097,
       "step": 27240
     },
     {
-      "epoch": 0.96,
-      "learning_rate": 4.665096743777807e-05,
-      "loss": 0.3115,
+      "epoch": 0.9819079540130464,
+      "grad_norm": 0.1999654471874237,
+      "learning_rate": 4.646046717244663e-05,
+      "loss": 0.435,
       "step": 27245
     },
     {
-      "epoch": 0.96,
-      "learning_rate": 4.664954300215795e-05,
-      "loss": 0.294,
+      "epoch": 0.982088153674271,
+      "grad_norm": 0.16381704807281494,
+      "learning_rate": 4.6458970170576296e-05,
+      "loss": 0.3906,
       "step": 27250
     },
     {
-      "epoch": 0.96,
-      "learning_rate": 4.664811828543277e-05,
-      "loss": 0.3071,
+      "epoch": 0.9822683533354958,
+      "grad_norm": 0.15966589748859406,
+      "learning_rate": 4.6457472876331644e-05,
+      "loss": 0.4372,
       "step": 27255
     },
     {
-      "epoch": 0.96,
-      "learning_rate": 4.6646693287621045e-05,
-      "loss": 0.287,
+      "epoch": 0.9824485529967204,
+      "grad_norm": 0.17224369943141937,
+      "learning_rate": 4.6455975289733077e-05,
+      "loss": 0.4433,
       "step": 27260
     },
     {
-      "epoch": 0.96,
-      "learning_rate": 4.664526800874127e-05,
-      "loss": 0.2981,
+      "epoch": 0.982628752657945,
+      "grad_norm": 0.1269349455833435,
+      "learning_rate": 4.6454477410801e-05,
+      "loss": 0.3989,
       "step": 27265
     },
     {
-      "epoch": 0.96,
-      "learning_rate": 4.664384244881196e-05,
-      "loss": 0.2819,
+      "epoch": 0.9828089523191696,
+      "grad_norm": 0.1905910223722458,
+      "learning_rate": 4.6452979239555825e-05,
+      "loss": 0.4544,
       "step": 27270
     },
     {
-      "epoch": 0.96,
-      "learning_rate": 4.6642416607851615e-05,
-      "loss": 0.3413,
+      "epoch": 0.9829891519803943,
+      "grad_norm": 0.17192591726779938,
+      "learning_rate": 4.645148077601796e-05,
+      "loss": 0.4112,
       "step": 27275
     },
     {
-      "epoch": 0.96,
-      "learning_rate": 4.6640990485878755e-05,
-      "loss": 0.2848,
+      "epoch": 0.9831693516416189,
+      "grad_norm": 0.1754426509141922,
+      "learning_rate": 4.644998202020783e-05,
+      "loss": 0.471,
       "step": 27280
     },
     {
-      "epoch": 0.96,
-      "learning_rate": 4.6639564082911896e-05,
-      "loss": 0.3091,
+      "epoch": 0.9833495513028435,
+      "grad_norm": 0.16824094951152802,
+      "learning_rate": 4.644848297214584e-05,
+      "loss": 0.4322,
       "step": 27285
     },
     {
-      "epoch": 0.96,
-      "learning_rate": 4.663813739896956e-05,
-      "loss": 0.2923,
+      "epoch": 0.9835297509640681,
+      "grad_norm": 0.15957225859165192,
+      "learning_rate": 4.6446983631852424e-05,
+      "loss": 0.4364,
       "step": 27290
     },
     {
-      "epoch": 0.96,
-      "learning_rate": 4.663671043407026e-05,
-      "loss": 0.3156,
+      "epoch": 0.9837099506252929,
+      "grad_norm": 0.16426332294940948,
+      "learning_rate": 4.6445483999348006e-05,
+      "loss": 0.4243,
       "step": 27295
     },
     {
-      "epoch": 0.96,
-      "learning_rate": 4.663528318823255e-05,
-      "loss": 0.2914,
+      "epoch": 0.9838901502865175,
+      "grad_norm": 0.1607593595981598,
+      "learning_rate": 4.6443984074653026e-05,
+      "loss": 0.4289,
       "step": 27300
     },
     {
-      "epoch": 0.96,
-      "learning_rate": 4.6633855661474935e-05,
-      "loss": 0.2957,
+      "epoch": 0.9840703499477421,
+      "grad_norm": 0.15784458816051483,
+      "learning_rate": 4.644248385778791e-05,
+      "loss": 0.396,
       "step": 27305
     },
     {
-      "epoch": 0.96,
-      "learning_rate": 4.6632427853815965e-05,
-      "loss": 0.2945,
+      "epoch": 0.9842505496089667,
+      "grad_norm": 0.21137282252311707,
+      "learning_rate": 4.6440983348773105e-05,
+      "loss": 0.4444,
       "step": 27310
     },
     {
-      "epoch": 0.96,
-      "learning_rate": 4.6630999765274175e-05,
-      "loss": 0.294,
+      "epoch": 0.9844307492701914,
+      "grad_norm": 0.21864286065101624,
+      "learning_rate": 4.6439482547629046e-05,
+      "loss": 0.4633,
       "step": 27315
     },
     {
-      "epoch": 0.96,
-      "learning_rate": 4.662957139586811e-05,
-      "loss": 0.2932,
+      "epoch": 0.984610948931416,
+      "grad_norm": 0.1778162717819214,
+      "learning_rate": 4.6437981454376194e-05,
+      "loss": 0.3835,
       "step": 27320
     },
     {
-      "epoch": 0.96,
-      "learning_rate": 4.6628142745616313e-05,
-      "loss": 0.2979,
+      "epoch": 0.9847911485926406,
+      "grad_norm": 0.1485341191291809,
+      "learning_rate": 4.6436480069034995e-05,
+      "loss": 0.4085,
       "step": 27325
     },
     {
-      "epoch": 0.96,
-      "learning_rate": 4.6626713814537335e-05,
-      "loss": 0.2806,
+      "epoch": 0.9849713482538652,
+      "grad_norm": 0.14085252583026886,
+      "learning_rate": 4.6434978391625905e-05,
+      "loss": 0.4097,
       "step": 27330
     },
     {
-      "epoch": 0.96,
-      "learning_rate": 4.662528460264973e-05,
-      "loss": 0.289,
+      "epoch": 0.98515154791509,
+      "grad_norm": 0.12692780792713165,
+      "learning_rate": 4.643347642216939e-05,
+      "loss": 0.4053,
       "step": 27335
     },
     {
-      "epoch": 0.96,
-      "learning_rate": 4.662385510997206e-05,
-      "loss": 0.3088,
+      "epoch": 0.9853317475763146,
+      "grad_norm": 0.18438071012496948,
+      "learning_rate": 4.64319741606859e-05,
+      "loss": 0.43,
       "step": 27340
     },
     {
-      "epoch": 0.96,
-      "learning_rate": 4.662242533652288e-05,
-      "loss": 0.2632,
+      "epoch": 0.9855119472375392,
+      "grad_norm": 0.21135936677455902,
+      "learning_rate": 4.6430471607195917e-05,
+      "loss": 0.4677,
       "step": 27345
     },
     {
-      "epoch": 0.96,
-      "learning_rate": 4.6620995282320756e-05,
-      "loss": 0.2826,
+      "epoch": 0.9856921468987638,
+      "grad_norm": 0.16982705891132355,
+      "learning_rate": 4.64289687617199e-05,
+      "loss": 0.4531,
       "step": 27350
     },
     {
-      "epoch": 0.96,
-      "learning_rate": 4.661956494738426e-05,
-      "loss": 0.2917,
+      "epoch": 0.9858723465599885,
+      "grad_norm": 0.16164273023605347,
+      "learning_rate": 4.642746562427834e-05,
+      "loss": 0.4122,
       "step": 27355
     },
     {
-      "epoch": 0.96,
-      "learning_rate": 4.6618134331731956e-05,
-      "loss": 0.3059,
+      "epoch": 0.9860525462212131,
+      "grad_norm": 0.20969519019126892,
+      "learning_rate": 4.6425962194891705e-05,
+      "loss": 0.4288,
       "step": 27360
     },
     {
-      "epoch": 0.96,
-      "learning_rate": 4.661670343538243e-05,
-      "loss": 0.2964,
+      "epoch": 0.9862327458824377,
+      "grad_norm": 0.14550209045410156,
+      "learning_rate": 4.6424458473580486e-05,
+      "loss": 0.4714,
       "step": 27365
     },
     {
-      "epoch": 0.96,
-      "learning_rate": 4.6615272258354255e-05,
-      "loss": 0.3051,
+      "epoch": 0.9864129455436623,
+      "grad_norm": 0.18479301035404205,
+      "learning_rate": 4.6422954460365165e-05,
+      "loss": 0.4166,
       "step": 27370
     },
     {
-      "epoch": 0.96,
-      "learning_rate": 4.661384080066601e-05,
-      "loss": 0.3155,
+      "epoch": 0.9865931452048871,
+      "grad_norm": 0.15923672914505005,
+      "learning_rate": 4.642145015526624e-05,
+      "loss": 0.4384,
       "step": 27375
     },
     {
-      "epoch": 0.96,
-      "learning_rate": 4.661240906233629e-05,
-      "loss": 0.3184,
+      "epoch": 0.9867733448661117,
+      "grad_norm": 0.1702156960964203,
+      "learning_rate": 4.64199455583042e-05,
+      "loss": 0.427,
       "step": 27380
     },
     {
-      "epoch": 0.96,
-      "learning_rate": 4.661097704338368e-05,
-      "loss": 0.2967,
+      "epoch": 0.9869535445273363,
+      "grad_norm": 0.19491025805473328,
+      "learning_rate": 4.641844066949955e-05,
+      "loss": 0.4257,
       "step": 27385
     },
     {
-      "epoch": 0.96,
-      "learning_rate": 4.6609544743826775e-05,
-      "loss": 0.3054,
+      "epoch": 0.9871337441885609,
+      "grad_norm": 0.21191620826721191,
+      "learning_rate": 4.6416935488872806e-05,
+      "loss": 0.4846,
       "step": 27390
     },
     {
-      "epoch": 0.96,
-      "learning_rate": 4.6608112163684175e-05,
-      "loss": 0.3207,
+      "epoch": 0.9873139438497855,
+      "grad_norm": 0.1783130019903183,
+      "learning_rate": 4.6415430016444445e-05,
+      "loss": 0.4066,
       "step": 27395
     },
     {
-      "epoch": 0.96,
-      "learning_rate": 4.6606679302974474e-05,
-      "loss": 0.2907,
+      "epoch": 0.9874941435110102,
+      "grad_norm": 0.16882510483264923,
+      "learning_rate": 4.6413924252235006e-05,
+      "loss": 0.4118,
       "step": 27400
     },
     {
-      "epoch": 0.96,
-      "learning_rate": 4.660524616171629e-05,
-      "loss": 0.3062,
+      "epoch": 0.9876743431722348,
+      "grad_norm": 0.18911772966384888,
+      "learning_rate": 4.641241819626499e-05,
+      "loss": 0.4378,
       "step": 27405
     },
     {
-      "epoch": 0.96,
-      "learning_rate": 4.660381273992822e-05,
-      "loss": 0.2796,
+      "epoch": 0.9878545428334595,
+      "grad_norm": 0.15727469325065613,
+      "learning_rate": 4.641091184855492e-05,
+      "loss": 0.4031,
       "step": 27410
     },
     {
-      "epoch": 0.96,
-      "learning_rate": 4.660237903762887e-05,
-      "loss": 0.3121,
+      "epoch": 0.9880347424946841,
+      "grad_norm": 0.1802516132593155,
+      "learning_rate": 4.640940520912532e-05,
+      "loss": 0.4541,
       "step": 27415
     },
     {
-      "epoch": 0.96,
-      "learning_rate": 4.660094505483688e-05,
-      "loss": 0.2687,
+      "epoch": 0.9882149421559088,
+      "grad_norm": 0.18460440635681152,
+      "learning_rate": 4.640789827799673e-05,
+      "loss": 0.409,
       "step": 27420
     },
     {
-      "epoch": 0.96,
-      "learning_rate": 4.659951079157084e-05,
-      "loss": 0.319,
+      "epoch": 0.9883951418171334,
+      "grad_norm": 0.18090972304344177,
+      "learning_rate": 4.640639105518966e-05,
+      "loss": 0.4232,
       "step": 27425
     },
     {
-      "epoch": 0.97,
-      "learning_rate": 4.65980762478494e-05,
-      "loss": 0.3121,
+      "epoch": 0.988575341478358,
+      "grad_norm": 0.15177130699157715,
+      "learning_rate": 4.6404883540724665e-05,
+      "loss": 0.3694,
       "step": 27430
     },
     {
-      "epoch": 0.97,
-      "learning_rate": 4.659664142369116e-05,
-      "loss": 0.3029,
+      "epoch": 0.9887555411395826,
+      "grad_norm": 0.19808749854564667,
+      "learning_rate": 4.6403375734622265e-05,
+      "loss": 0.416,
       "step": 27435
     },
     {
-      "epoch": 0.97,
-      "learning_rate": 4.659520631911477e-05,
-      "loss": 0.2954,
+      "epoch": 0.9889357408008073,
+      "grad_norm": 0.18107527494430542,
+      "learning_rate": 4.640186763690302e-05,
+      "loss": 0.4578,
       "step": 27440
     },
     {
-      "epoch": 0.97,
-      "learning_rate": 4.659377093413886e-05,
-      "loss": 0.2916,
+      "epoch": 0.9891159404620319,
+      "grad_norm": 0.18067267537117004,
+      "learning_rate": 4.640035924758748e-05,
+      "loss": 0.4572,
       "step": 27445
     },
     {
-      "epoch": 0.97,
-      "learning_rate": 4.659233526878206e-05,
-      "loss": 0.2977,
+      "epoch": 0.9892961401232566,
+      "grad_norm": 0.14522598683834076,
+      "learning_rate": 4.6398850566696176e-05,
+      "loss": 0.4252,
       "step": 27450
     },
     {
-      "epoch": 0.97,
-      "learning_rate": 4.659089932306302e-05,
-      "loss": 0.2922,
+      "epoch": 0.9894763397844812,
+      "grad_norm": 0.1613176465034485,
+      "learning_rate": 4.6397341594249675e-05,
+      "loss": 0.3907,
       "step": 27455
     },
     {
-      "epoch": 0.97,
-      "learning_rate": 4.658946309700039e-05,
-      "loss": 0.3278,
+      "epoch": 0.9896565394457059,
+      "grad_norm": 0.1852157860994339,
+      "learning_rate": 4.639583233026855e-05,
+      "loss": 0.4324,
       "step": 27460
     },
     {
-      "epoch": 0.97,
-      "learning_rate": 4.658802659061279e-05,
-      "loss": 0.2658,
+      "epoch": 0.9898367391069305,
+      "grad_norm": 0.16514049470424652,
+      "learning_rate": 4.639432277477335e-05,
+      "loss": 0.4413,
       "step": 27465
     },
     {
-      "epoch": 0.97,
-      "learning_rate": 4.6586589803918905e-05,
-      "loss": 0.2951,
+      "epoch": 0.9900169387681551,
+      "grad_norm": 0.14694753289222717,
+      "learning_rate": 4.639281292778464e-05,
+      "loss": 0.4236,
       "step": 27470
     },
     {
-      "epoch": 0.97,
-      "learning_rate": 4.658515273693738e-05,
-      "loss": 0.3163,
+      "epoch": 0.9901971384293797,
+      "grad_norm": 0.18619796633720398,
+      "learning_rate": 4.639130278932299e-05,
+      "loss": 0.4442,
       "step": 27475
     },
     {
-      "epoch": 0.97,
-      "learning_rate": 4.658371538968686e-05,
-      "loss": 0.2748,
+      "epoch": 0.9903773380906044,
+      "grad_norm": 0.15744751691818237,
+      "learning_rate": 4.638979235940899e-05,
+      "loss": 0.4074,
       "step": 27480
     },
     {
-      "epoch": 0.97,
-      "learning_rate": 4.658227776218603e-05,
-      "loss": 0.3057,
+      "epoch": 0.990557537751829,
+      "grad_norm": 0.17762751877307892,
+      "learning_rate": 4.63882816380632e-05,
+      "loss": 0.4212,
       "step": 27485
     },
     {
-      "epoch": 0.97,
-      "learning_rate": 4.658083985445354e-05,
-      "loss": 0.2993,
+      "epoch": 0.9907377374130537,
+      "grad_norm": 0.18236143887043,
+      "learning_rate": 4.638677062530622e-05,
+      "loss": 0.4089,
       "step": 27490
     },
     {
-      "epoch": 0.97,
-      "learning_rate": 4.6579401666508074e-05,
-      "loss": 0.3059,
+      "epoch": 0.9909179370742783,
+      "grad_norm": 0.1682094931602478,
+      "learning_rate": 4.638525932115863e-05,
+      "loss": 0.4288,
       "step": 27495
     },
     {
-      "epoch": 0.97,
-      "learning_rate": 4.657796319836829e-05,
-      "loss": 0.3165,
+      "epoch": 0.991098136735503,
+      "grad_norm": 0.17078998684883118,
+      "learning_rate": 4.6383747725641027e-05,
+      "loss": 0.4328,
       "step": 27500
     },
     {
-      "epoch": 0.97,
-      "eval_loss": 0.293122798204422,
-      "eval_runtime": 10.5363,
-      "eval_samples_per_second": 9.491,
-      "eval_steps_per_second": 9.491,
+      "epoch": 0.991098136735503,
+      "eval_loss": 0.45491889119148254,
+      "eval_runtime": 3.529,
+      "eval_samples_per_second": 28.337,
+      "eval_steps_per_second": 7.084,
       "step": 27500
     },
     {
-      "epoch": 0.97,
-      "learning_rate": 4.657652445005288e-05,
-      "loss": 0.2904,
+      "epoch": 0.9912783363967276,
+      "grad_norm": 0.185177743434906,
+      "learning_rate": 4.6382235838774e-05,
+      "loss": 0.4294,
       "step": 27505
     },
     {
-      "epoch": 0.97,
-      "learning_rate": 4.657508542158052e-05,
-      "loss": 0.2853,
+      "epoch": 0.9914585360579522,
+      "grad_norm": 0.1422509402036667,
+      "learning_rate": 4.6380723660578144e-05,
+      "loss": 0.3911,
       "step": 27510
     },
     {
-      "epoch": 0.97,
-      "learning_rate": 4.657364611296989e-05,
-      "loss": 0.2936,
+      "epoch": 0.9916387357191768,
+      "grad_norm": 0.1539120376110077,
+      "learning_rate": 4.6379211191074066e-05,
+      "loss": 0.4519,
       "step": 27515
     },
     {
-      "epoch": 0.97,
-      "learning_rate": 4.6572206524239686e-05,
-      "loss": 0.2737,
+      "epoch": 0.9918189353804014,
+      "grad_norm": 0.19037242233753204,
+      "learning_rate": 4.637769843028238e-05,
+      "loss": 0.4822,
       "step": 27520
     },
     {
-      "epoch": 0.97,
-      "learning_rate": 4.6570766655408594e-05,
-      "loss": 0.2966,
+      "epoch": 0.9919991350416261,
+      "grad_norm": 0.2115742713212967,
+      "learning_rate": 4.637618537822369e-05,
+      "loss": 0.4498,
       "step": 27525
     },
     {
-      "epoch": 0.97,
-      "learning_rate": 4.656932650649532e-05,
-      "loss": 0.3009,
+      "epoch": 0.9921793347028508,
+      "grad_norm": 0.17919567227363586,
+      "learning_rate": 4.6374672034918606e-05,
+      "loss": 0.4453,
       "step": 27530
     },
     {
-      "epoch": 0.97,
-      "learning_rate": 4.656788607751855e-05,
-      "loss": 0.3059,
+      "epoch": 0.9923595343640754,
+      "grad_norm": 0.13095134496688843,
+      "learning_rate": 4.6373158400387775e-05,
+      "loss": 0.4289,
       "step": 27535
     },
     {
-      "epoch": 0.97,
-      "learning_rate": 4.656644536849699e-05,
-      "loss": 0.2924,
+      "epoch": 0.9925397340253,
+      "grad_norm": 0.18032823503017426,
+      "learning_rate": 4.6371644474651773e-05,
+      "loss": 0.4262,
       "step": 27540
     },
     {
-      "epoch": 0.97,
-      "learning_rate": 4.656500437944936e-05,
-      "loss": 0.3073,
+      "epoch": 0.9927199336865247,
+      "grad_norm": 0.14844289422035217,
+      "learning_rate": 4.637013025773127e-05,
+      "loss": 0.4254,
       "step": 27545
     },
     {
-      "epoch": 0.97,
-      "learning_rate": 4.656356311039435e-05,
-      "loss": 0.295,
+      "epoch": 0.9929001333477493,
+      "grad_norm": 0.182834193110466,
+      "learning_rate": 4.636861574964687e-05,
+      "loss": 0.4526,
       "step": 27550
     },
     {
-      "epoch": 0.97,
-      "learning_rate": 4.6562121561350686e-05,
-      "loss": 0.2916,
+      "epoch": 0.9930803330089739,
+      "grad_norm": 0.18387030065059662,
+      "learning_rate": 4.6367100950419226e-05,
+      "loss": 0.4377,
       "step": 27555
     },
     {
-      "epoch": 0.97,
-      "learning_rate": 4.656067973233709e-05,
-      "loss": 0.3131,
+      "epoch": 0.9932605326701985,
+      "grad_norm": 0.200217604637146,
+      "learning_rate": 4.636558586006896e-05,
+      "loss": 0.4289,
       "step": 27560
     },
     {
-      "epoch": 0.97,
-      "learning_rate": 4.6559237623372265e-05,
-      "loss": 0.2935,
+      "epoch": 0.9934407323314233,
+      "grad_norm": 0.16323430836200714,
+      "learning_rate": 4.636407047861673e-05,
+      "loss": 0.41,
       "step": 27565
     },
     {
-      "epoch": 0.97,
-      "learning_rate": 4.655779523447495e-05,
-      "loss": 0.3054,
+      "epoch": 0.9936209319926479,
+      "grad_norm": 0.14790096879005432,
+      "learning_rate": 4.6362554806083176e-05,
+      "loss": 0.4418,
       "step": 27570
     },
     {
-      "epoch": 0.97,
-      "learning_rate": 4.6556352565663885e-05,
-      "loss": 0.3058,
+      "epoch": 0.9938011316538725,
+      "grad_norm": 0.18364425003528595,
+      "learning_rate": 4.6361038842488944e-05,
+      "loss": 0.4132,
       "step": 27575
     },
     {
-      "epoch": 0.97,
-      "learning_rate": 4.6554909616957774e-05,
-      "loss": 0.2799,
+      "epoch": 0.9939813313150971,
+      "grad_norm": 0.15414080023765564,
+      "learning_rate": 4.63595225878547e-05,
+      "loss": 0.4203,
       "step": 27580
     },
     {
-      "epoch": 0.97,
-      "learning_rate": 4.6553466388375375e-05,
-      "loss": 0.3288,
+      "epoch": 0.9941615309763218,
+      "grad_norm": 0.17159409821033478,
+      "learning_rate": 4.635800604220109e-05,
+      "loss": 0.4208,
       "step": 27585
     },
     {
-      "epoch": 0.97,
-      "learning_rate": 4.655202287993541e-05,
-      "loss": 0.3051,
+      "epoch": 0.9943417306375464,
+      "grad_norm": 0.13989026844501495,
+      "learning_rate": 4.635648920554878e-05,
+      "loss": 0.4302,
       "step": 27590
     },
     {
-      "epoch": 0.97,
-      "learning_rate": 4.6550579091656646e-05,
-      "loss": 0.2936,
+      "epoch": 0.994521930298771,
+      "grad_norm": 0.19220641255378723,
+      "learning_rate": 4.635497207791845e-05,
+      "loss": 0.4385,
       "step": 27595
     },
     {
-      "epoch": 0.97,
-      "learning_rate": 4.65491350235578e-05,
-      "loss": 0.3018,
+      "epoch": 0.9947021299599956,
+      "grad_norm": 0.20542806386947632,
+      "learning_rate": 4.635345465933075e-05,
+      "loss": 0.4185,
       "step": 27600
     },
     {
-      "epoch": 0.97,
-      "learning_rate": 4.654769067565764e-05,
-      "loss": 0.29,
+      "epoch": 0.9948823296212204,
+      "grad_norm": 0.15193575620651245,
+      "learning_rate": 4.635193694980636e-05,
+      "loss": 0.4152,
       "step": 27605
     },
     {
-      "epoch": 0.97,
-      "learning_rate": 4.654624604797492e-05,
-      "loss": 0.2886,
+      "epoch": 0.995062529282445,
+      "grad_norm": 0.174140065908432,
+      "learning_rate": 4.635041894936598e-05,
+      "loss": 0.4468,
       "step": 27610
     },
     {
-      "epoch": 0.97,
-      "learning_rate": 4.65448011405284e-05,
-      "loss": 0.3219,
+      "epoch": 0.9952427289436696,
+      "grad_norm": 0.18024513125419617,
+      "learning_rate": 4.6348900658030263e-05,
+      "loss": 0.4326,
       "step": 27615
     },
     {
-      "epoch": 0.97,
-      "learning_rate": 4.654335595333683e-05,
-      "loss": 0.2825,
+      "epoch": 0.9954229286048942,
+      "grad_norm": 0.18026669323444366,
+      "learning_rate": 4.634738207581991e-05,
+      "loss": 0.4329,
       "step": 27620
     },
     {
-      "epoch": 0.97,
-      "learning_rate": 4.654191048641898e-05,
-      "loss": 0.2895,
+      "epoch": 0.9956031282661189,
+      "grad_norm": 0.15884456038475037,
+      "learning_rate": 4.634586320275561e-05,
+      "loss": 0.4202,
       "step": 27625
     },
     {
-      "epoch": 0.97,
-      "learning_rate": 4.6540464739793624e-05,
-      "loss": 0.3029,
+      "epoch": 0.9957833279273435,
+      "grad_norm": 0.1757356971502304,
+      "learning_rate": 4.634434403885805e-05,
+      "loss": 0.4291,
       "step": 27630
     },
     {
-      "epoch": 0.97,
-      "learning_rate": 4.653901871347952e-05,
-      "loss": 0.3063,
+      "epoch": 0.9959635275885681,
+      "grad_norm": 0.16361142694950104,
+      "learning_rate": 4.634282458414795e-05,
+      "loss": 0.4216,
       "step": 27635
     },
     {
-      "epoch": 0.97,
-      "learning_rate": 4.653757240749546e-05,
-      "loss": 0.3149,
+      "epoch": 0.9961437272497927,
+      "grad_norm": 0.21342548727989197,
+      "learning_rate": 4.634130483864598e-05,
+      "loss": 0.4669,
       "step": 27640
     },
     {
-      "epoch": 0.97,
-      "learning_rate": 4.653612582186021e-05,
-      "loss": 0.3045,
+      "epoch": 0.9963239269110175,
+      "grad_norm": 0.2188446968793869,
+      "learning_rate": 4.6339784802372874e-05,
+      "loss": 0.4772,
       "step": 27645
     },
     {
-      "epoch": 0.97,
-      "learning_rate": 4.653467895659256e-05,
-      "loss": 0.2987,
+      "epoch": 0.9965041265722421,
+      "grad_norm": 0.1779574155807495,
+      "learning_rate": 4.633826447534934e-05,
+      "loss": 0.4633,
       "step": 27650
     },
     {
-      "epoch": 0.97,
-      "learning_rate": 4.6533231811711296e-05,
-      "loss": 0.3002,
+      "epoch": 0.9966843262334667,
+      "grad_norm": 0.19599558413028717,
+      "learning_rate": 4.633674385759607e-05,
+      "loss": 0.4405,
       "step": 27655
     },
     {
-      "epoch": 0.97,
-      "learning_rate": 4.653178438723521e-05,
-      "loss": 0.3108,
+      "epoch": 0.9968645258946913,
+      "grad_norm": 0.15029726922512054,
+      "learning_rate": 4.6335222949133794e-05,
+      "loss": 0.4461,
       "step": 27660
     },
     {
-      "epoch": 0.97,
-      "learning_rate": 4.653033668318309e-05,
-      "loss": 0.2842,
+      "epoch": 0.997044725555916,
+      "grad_norm": 0.17829090356826782,
+      "learning_rate": 4.633370174998324e-05,
+      "loss": 0.4567,
       "step": 27665
     },
     {
-      "epoch": 0.97,
-      "learning_rate": 4.652888869957375e-05,
-      "loss": 0.2661,
+      "epoch": 0.9972249252171406,
+      "grad_norm": 0.1724914163351059,
+      "learning_rate": 4.6332180260165135e-05,
+      "loss": 0.4121,
       "step": 27670
     },
     {
-      "epoch": 0.97,
-      "learning_rate": 4.652744043642597e-05,
-      "loss": 0.312,
+      "epoch": 0.9974051248783652,
+      "grad_norm": 0.18331369757652283,
+      "learning_rate": 4.63306584797002e-05,
+      "loss": 0.4185,
       "step": 27675
     },
     {
-      "epoch": 0.97,
-      "learning_rate": 4.652599189375856e-05,
-      "loss": 0.3114,
+      "epoch": 0.9975853245395898,
+      "grad_norm": 0.2127685546875,
+      "learning_rate": 4.632913640860918e-05,
+      "loss": 0.465,
       "step": 27680
     },
     {
-      "epoch": 0.97,
-      "learning_rate": 4.652454307159033e-05,
-      "loss": 0.3029,
+      "epoch": 0.9977655242008145,
+      "grad_norm": 0.18392300605773926,
+      "learning_rate": 4.6327614046912796e-05,
+      "loss": 0.4276,
       "step": 27685
     },
     {
-      "epoch": 0.97,
-      "learning_rate": 4.65230939699401e-05,
-      "loss": 0.326,
+      "epoch": 0.9979457238620392,
+      "grad_norm": 0.18069952726364136,
+      "learning_rate": 4.63260913946318e-05,
+      "loss": 0.4089,
       "step": 27690
     },
     {
-      "epoch": 0.97,
-      "learning_rate": 4.6521644588826675e-05,
-      "loss": 0.3124,
+      "epoch": 0.9981259235232638,
+      "grad_norm": 0.17359404265880585,
+      "learning_rate": 4.632456845178694e-05,
+      "loss": 0.4294,
       "step": 27695
     },
     {
-      "epoch": 0.97,
-      "learning_rate": 4.652019492826888e-05,
-      "loss": 0.2903,
+      "epoch": 0.9983061231844884,
+      "grad_norm": 0.1843539923429489,
+      "learning_rate": 4.632304521839896e-05,
+      "loss": 0.4795,
       "step": 27700
     },
     {
-      "epoch": 0.97,
-      "learning_rate": 4.6518744988285536e-05,
-      "loss": 0.3094,
+      "epoch": 0.998486322845713,
+      "grad_norm": 0.17044410109519958,
+      "learning_rate": 4.6321521694488627e-05,
+      "loss": 0.4345,
       "step": 27705
     },
     {
-      "epoch": 0.97,
-      "learning_rate": 4.651729476889547e-05,
-      "loss": 0.2816,
+      "epoch": 0.9986665225069377,
+      "grad_norm": 0.24153609573841095,
+      "learning_rate": 4.631999788007668e-05,
+      "loss": 0.437,
       "step": 27710
     },
     {
-      "epoch": 0.98,
-      "learning_rate": 4.651584427011751e-05,
-      "loss": 0.3151,
+      "epoch": 0.9988467221681623,
+      "grad_norm": 0.16202819347381592,
+      "learning_rate": 4.631847377518389e-05,
+      "loss": 0.4272,
       "step": 27715
     },
     {
-      "epoch": 0.98,
-      "learning_rate": 4.6514393491970496e-05,
-      "loss": 0.291,
+      "epoch": 0.999026921829387,
+      "grad_norm": 0.16928540170192719,
+      "learning_rate": 4.6316949379831025e-05,
+      "loss": 0.4584,
       "step": 27720
     },
     {
-      "epoch": 0.98,
-      "learning_rate": 4.651294243447326e-05,
-      "loss": 0.2991,
+      "epoch": 0.9992071214906116,
+      "grad_norm": 0.20613226294517517,
+      "learning_rate": 4.6315424694038854e-05,
+      "loss": 0.4415,
       "step": 27725
     },
     {
-      "epoch": 0.98,
-      "learning_rate": 4.651149109764464e-05,
-      "loss": 0.301,
+      "epoch": 0.9993873211518363,
+      "grad_norm": 0.17505665123462677,
+      "learning_rate": 4.631389971782815e-05,
+      "loss": 0.4521,
       "step": 27730
     },
     {
-      "epoch": 0.98,
-      "learning_rate": 4.651003948150349e-05,
-      "loss": 0.2971,
+      "epoch": 0.9995675208130609,
+      "grad_norm": 0.18743610382080078,
+      "learning_rate": 4.631237445121968e-05,
+      "loss": 0.4555,
       "step": 27735
     },
     {
-      "epoch": 0.98,
-      "learning_rate": 4.6508587586068644e-05,
-      "loss": 0.2991,
+      "epoch": 0.9997477204742855,
+      "grad_norm": 0.17451363801956177,
+      "learning_rate": 4.631084889423424e-05,
+      "loss": 0.4042,
       "step": 27740
     },
     {
-      "epoch": 0.98,
-      "learning_rate": 4.650713541135897e-05,
-      "loss": 0.3036,
+      "epoch": 0.9999279201355101,
+      "grad_norm": 0.17138254642486572,
+      "learning_rate": 4.630932304689261e-05,
+      "loss": 0.411,
       "step": 27745
     },
     {
-      "epoch": 0.98,
-      "learning_rate": 4.650568295739331e-05,
-      "loss": 0.3455,
+      "epoch": 1.0001081197967348,
+      "grad_norm": 0.18573379516601562,
+      "learning_rate": 4.6307796909215574e-05,
+      "loss": 0.4382,
       "step": 27750
     },
     {
-      "epoch": 0.98,
-      "learning_rate": 4.650423022419054e-05,
-      "loss": 0.3175,
+      "epoch": 1.0002883194579595,
+      "grad_norm": 0.18925435841083527,
+      "learning_rate": 4.630627048122393e-05,
+      "loss": 0.4086,
       "step": 27755
     },
     {
-      "epoch": 0.98,
-      "learning_rate": 4.650277721176951e-05,
-      "loss": 0.3293,
+      "epoch": 1.000468519119184,
+      "grad_norm": 0.15313346683979034,
+      "learning_rate": 4.630474376293849e-05,
+      "loss": 0.435,
       "step": 27760
     },
     {
-      "epoch": 0.98,
-      "learning_rate": 4.6501323920149086e-05,
-      "loss": 0.2894,
+      "epoch": 1.0006487187804087,
+      "grad_norm": 0.16638581454753876,
+      "learning_rate": 4.630321675438002e-05,
+      "loss": 0.3956,
       "step": 27765
     },
     {
-      "epoch": 0.98,
-      "learning_rate": 4.649987034934814e-05,
-      "loss": 0.3213,
+      "epoch": 1.0008289184416332,
+      "grad_norm": 0.23420801758766174,
+      "learning_rate": 4.630168945556937e-05,
+      "loss": 0.4247,
       "step": 27770
     },
     {
-      "epoch": 0.98,
-      "learning_rate": 4.6498416499385546e-05,
-      "loss": 0.2862,
+      "epoch": 1.001009118102858,
+      "grad_norm": 0.19458545744419098,
+      "learning_rate": 4.63001618665273e-05,
+      "loss": 0.4022,
       "step": 27775
     },
     {
-      "epoch": 0.98,
-      "learning_rate": 4.6496962370280183e-05,
-      "loss": 0.3092,
+      "epoch": 1.0011893177640827,
+      "grad_norm": 0.15574969351291656,
+      "learning_rate": 4.629863398727466e-05,
+      "loss": 0.4313,
       "step": 27780
     },
     {
-      "epoch": 0.98,
-      "learning_rate": 4.6495507962050936e-05,
-      "loss": 0.2987,
+      "epoch": 1.0013695174253072,
+      "grad_norm": 0.14644834399223328,
+      "learning_rate": 4.629710581783226e-05,
+      "loss": 0.4278,
       "step": 27785
     },
     {
-      "epoch": 0.98,
-      "learning_rate": 4.649405327471667e-05,
-      "loss": 0.3139,
+      "epoch": 1.001549717086532,
+      "grad_norm": 0.17037297785282135,
+      "learning_rate": 4.6295577358220914e-05,
+      "loss": 0.4216,
       "step": 27790
     },
     {
-      "epoch": 0.98,
-      "learning_rate": 4.64925983082963e-05,
-      "loss": 0.2881,
+      "epoch": 1.0017299167477565,
+      "grad_norm": 0.18147750198841095,
+      "learning_rate": 4.629404860846145e-05,
+      "loss": 0.4044,
       "step": 27795
     },
     {
-      "epoch": 0.98,
-      "learning_rate": 4.6491143062808696e-05,
-      "loss": 0.3101,
+      "epoch": 1.0019101164089812,
+      "grad_norm": 0.17213241755962372,
+      "learning_rate": 4.629251956857469e-05,
+      "loss": 0.4361,
       "step": 27800
     },
     {
-      "epoch": 0.98,
-      "learning_rate": 4.648968753827277e-05,
-      "loss": 0.291,
+      "epoch": 1.0020903160702057,
+      "grad_norm": 0.16381776332855225,
+      "learning_rate": 4.629099023858148e-05,
+      "loss": 0.4295,
       "step": 27805
     },
     {
-      "epoch": 0.98,
-      "learning_rate": 4.648823173470741e-05,
-      "loss": 0.2974,
+      "epoch": 1.0022705157314304,
+      "grad_norm": 0.19637686014175415,
+      "learning_rate": 4.628946061850265e-05,
+      "loss": 0.4377,
       "step": 27810
     },
     {
-      "epoch": 0.98,
-      "learning_rate": 4.648677565213152e-05,
-      "loss": 0.3195,
+      "epoch": 1.002450715392655,
+      "grad_norm": 0.19968008995056152,
+      "learning_rate": 4.628793070835904e-05,
+      "loss": 0.406,
       "step": 27815
     },
     {
-      "epoch": 0.98,
-      "learning_rate": 4.648531929056401e-05,
-      "loss": 0.2996,
+      "epoch": 1.0026309150538797,
+      "grad_norm": 0.19794724881649017,
+      "learning_rate": 4.628640050817149e-05,
+      "loss": 0.4257,
       "step": 27820
     },
     {
-      "epoch": 0.98,
-      "learning_rate": 4.64838626500238e-05,
-      "loss": 0.2848,
+      "epoch": 1.0028111147151044,
+      "grad_norm": 0.14951010048389435,
+      "learning_rate": 4.628487001796086e-05,
+      "loss": 0.4467,
       "step": 27825
     },
     {
-      "epoch": 0.98,
-      "learning_rate": 4.6482405730529773e-05,
-      "loss": 0.3081,
+      "epoch": 1.002991314376329,
+      "grad_norm": 0.169041708111763,
+      "learning_rate": 4.628333923774799e-05,
+      "loss": 0.4444,
       "step": 27830
     },
     {
-      "epoch": 0.98,
-      "learning_rate": 4.6480948532100875e-05,
-      "loss": 0.3071,
+      "epoch": 1.0031715140375537,
+      "grad_norm": 0.1682281494140625,
+      "learning_rate": 4.628180816755375e-05,
+      "loss": 0.4206,
       "step": 27835
     },
     {
-      "epoch": 0.98,
-      "learning_rate": 4.647949105475602e-05,
-      "loss": 0.3182,
+      "epoch": 1.0033517136987782,
+      "grad_norm": 0.18344461917877197,
+      "learning_rate": 4.6280276807398994e-05,
+      "loss": 0.4229,
       "step": 27840
     },
     {
-      "epoch": 0.98,
-      "learning_rate": 4.6478033298514124e-05,
-      "loss": 0.3004,
+      "epoch": 1.003531913360003,
+      "grad_norm": 0.14406928420066833,
+      "learning_rate": 4.627874515730459e-05,
+      "loss": 0.4192,
       "step": 27845
     },
     {
-      "epoch": 0.98,
-      "learning_rate": 4.6476575263394124e-05,
-      "loss": 0.2944,
+      "epoch": 1.0037121130212274,
+      "grad_norm": 0.1358763873577118,
+      "learning_rate": 4.6277213217291395e-05,
+      "loss": 0.4126,
       "step": 27850
     },
     {
-      "epoch": 0.98,
-      "learning_rate": 4.647511694941494e-05,
-      "loss": 0.305,
+      "epoch": 1.0038923126824522,
+      "grad_norm": 0.15834762156009674,
+      "learning_rate": 4.6275680987380296e-05,
+      "loss": 0.4679,
       "step": 27855
     },
     {
-      "epoch": 0.98,
-      "learning_rate": 4.6473658356595525e-05,
-      "loss": 0.2991,
+      "epoch": 1.004072512343677,
+      "grad_norm": 0.1732066124677658,
+      "learning_rate": 4.627414846759216e-05,
+      "loss": 0.3903,
       "step": 27860
     },
     {
-      "epoch": 0.98,
-      "learning_rate": 4.6472199484954805e-05,
-      "loss": 0.2867,
+      "epoch": 1.0042527120049014,
+      "grad_norm": 0.17274266481399536,
+      "learning_rate": 4.627261565794787e-05,
+      "loss": 0.4179,
       "step": 27865
     },
     {
-      "epoch": 0.98,
-      "learning_rate": 4.647074033451172e-05,
-      "loss": 0.3062,
+      "epoch": 1.0044329116661261,
+      "grad_norm": 0.1775139421224594,
+      "learning_rate": 4.6271082558468306e-05,
+      "loss": 0.4045,
       "step": 27870
     },
     {
-      "epoch": 0.98,
-      "learning_rate": 4.6469280905285224e-05,
-      "loss": 0.2917,
+      "epoch": 1.0046131113273506,
+      "grad_norm": 0.17623652517795563,
+      "learning_rate": 4.626954916917436e-05,
+      "loss": 0.3981,
       "step": 27875
     },
     {
-      "epoch": 0.98,
-      "learning_rate": 4.646782119729426e-05,
-      "loss": 0.3011,
+      "epoch": 1.0047933109885754,
+      "grad_norm": 0.18118661642074585,
+      "learning_rate": 4.626801549008693e-05,
+      "loss": 0.432,
       "step": 27880
     },
     {
-      "epoch": 0.98,
-      "learning_rate": 4.6466361210557794e-05,
-      "loss": 0.2834,
+      "epoch": 1.0049735106498,
+      "grad_norm": 0.1708979606628418,
+      "learning_rate": 4.6266481521226904e-05,
+      "loss": 0.4477,
       "step": 27885
     },
     {
-      "epoch": 0.98,
-      "learning_rate": 4.646490094509477e-05,
-      "loss": 0.3241,
+      "epoch": 1.0051537103110246,
+      "grad_norm": 0.17798757553100586,
+      "learning_rate": 4.6264947262615186e-05,
+      "loss": 0.4128,
       "step": 27890
     },
     {
-      "epoch": 0.98,
-      "learning_rate": 4.6463440400924156e-05,
-      "loss": 0.3104,
+      "epoch": 1.0053339099722494,
+      "grad_norm": 0.16669988632202148,
+      "learning_rate": 4.626341271427268e-05,
+      "loss": 0.4202,
       "step": 27895
     },
     {
-      "epoch": 0.98,
-      "learning_rate": 4.646197957806491e-05,
-      "loss": 0.3255,
+      "epoch": 1.0055141096334739,
+      "grad_norm": 0.1673901528120041,
+      "learning_rate": 4.626187787622029e-05,
+      "loss": 0.4245,
       "step": 27900
     },
     {
-      "epoch": 0.98,
-      "learning_rate": 4.646051847653601e-05,
-      "loss": 0.3279,
+      "epoch": 1.0056943092946986,
+      "grad_norm": 0.1705399453639984,
+      "learning_rate": 4.626034274847893e-05,
+      "loss": 0.4076,
       "step": 27905
     },
     {
-      "epoch": 0.98,
-      "learning_rate": 4.6459057096356414e-05,
-      "loss": 0.2838,
+      "epoch": 1.0058745089559231,
+      "grad_norm": 0.17363837361335754,
+      "learning_rate": 4.625880733106951e-05,
+      "loss": 0.4292,
       "step": 27910
     },
     {
-      "epoch": 0.98,
-      "learning_rate": 4.64575954375451e-05,
-      "loss": 0.3108,
+      "epoch": 1.0060547086171479,
+      "grad_norm": 0.15469084680080414,
+      "learning_rate": 4.625727162401296e-05,
+      "loss": 0.3902,
       "step": 27915
     },
     {
-      "epoch": 0.98,
-      "learning_rate": 4.645613350012106e-05,
-      "loss": 0.3067,
+      "epoch": 1.0062349082783724,
+      "grad_norm": 0.1932782232761383,
+      "learning_rate": 4.625573562733021e-05,
+      "loss": 0.4187,
       "step": 27920
     },
     {
-      "epoch": 0.98,
-      "learning_rate": 4.6454671284103255e-05,
-      "loss": 0.3262,
+      "epoch": 1.006415107939597,
+      "grad_norm": 0.2232416868209839,
+      "learning_rate": 4.625419934104217e-05,
+      "loss": 0.459,
       "step": 27925
     },
     {
-      "epoch": 0.98,
-      "learning_rate": 4.6453208789510695e-05,
-      "loss": 0.3154,
+      "epoch": 1.0065953076008216,
+      "grad_norm": 0.1562723070383072,
+      "learning_rate": 4.625266276516978e-05,
+      "loss": 0.42,
       "step": 27930
     },
     {
-      "epoch": 0.98,
-      "learning_rate": 4.645174601636235e-05,
-      "loss": 0.3076,
+      "epoch": 1.0067755072620463,
+      "grad_norm": 0.16613322496414185,
+      "learning_rate": 4.625112589973397e-05,
+      "loss": 0.4205,
       "step": 27935
     },
     {
-      "epoch": 0.98,
-      "learning_rate": 4.645028296467722e-05,
-      "loss": 0.3014,
+      "epoch": 1.006955706923271,
+      "grad_norm": 0.15974131226539612,
+      "learning_rate": 4.624958874475569e-05,
+      "loss": 0.4098,
       "step": 27940
     },
     {
-      "epoch": 0.98,
-      "learning_rate": 4.6448819634474305e-05,
-      "loss": 0.279,
+      "epoch": 1.0071359065844956,
+      "grad_norm": 0.12717698514461517,
+      "learning_rate": 4.624805130025588e-05,
+      "loss": 0.4076,
       "step": 27945
     },
     {
-      "epoch": 0.98,
-      "learning_rate": 4.64473560257726e-05,
-      "loss": 0.2873,
+      "epoch": 1.0073161062457203,
+      "grad_norm": 0.20461279153823853,
+      "learning_rate": 4.624651356625548e-05,
+      "loss": 0.438,
       "step": 27950
     },
     {
-      "epoch": 0.98,
-      "learning_rate": 4.644589213859112e-05,
-      "loss": 0.2604,
+      "epoch": 1.0074963059069448,
+      "grad_norm": 0.16563838720321655,
+      "learning_rate": 4.624497554277544e-05,
+      "loss": 0.3795,
       "step": 27955
     },
     {
-      "epoch": 0.98,
-      "learning_rate": 4.644442797294886e-05,
-      "loss": 0.3031,
+      "epoch": 1.0076765055681696,
+      "grad_norm": 0.1689777672290802,
+      "learning_rate": 4.6243437229836725e-05,
+      "loss": 0.425,
       "step": 27960
     },
     {
-      "epoch": 0.98,
-      "learning_rate": 4.644296352886482e-05,
-      "loss": 0.3184,
+      "epoch": 1.007856705229394,
+      "grad_norm": 0.1565127670764923,
+      "learning_rate": 4.624189862746029e-05,
+      "loss": 0.3809,
       "step": 27965
     },
     {
-      "epoch": 0.98,
-      "learning_rate": 4.644149880635804e-05,
-      "loss": 0.3232,
+      "epoch": 1.0080369048906188,
+      "grad_norm": 0.17497403919696808,
+      "learning_rate": 4.62403597356671e-05,
+      "loss": 0.4502,
       "step": 27970
     },
     {
-      "epoch": 0.98,
-      "learning_rate": 4.644003380544754e-05,
-      "loss": 0.3124,
+      "epoch": 1.0082171045518435,
+      "grad_norm": 0.1647554188966751,
+      "learning_rate": 4.623882055447813e-05,
+      "loss": 0.4453,
       "step": 27975
     },
     {
-      "epoch": 0.98,
-      "learning_rate": 4.6438568526152315e-05,
-      "loss": 0.281,
+      "epoch": 1.008397304213068,
+      "grad_norm": 0.18609805405139923,
+      "learning_rate": 4.623728108391433e-05,
+      "loss": 0.4263,
       "step": 27980
     },
     {
-      "epoch": 0.98,
-      "learning_rate": 4.6437102968491416e-05,
-      "loss": 0.3167,
+      "epoch": 1.0085775038742928,
+      "grad_norm": 0.15984831750392914,
+      "learning_rate": 4.6235741323996696e-05,
+      "loss": 0.4564,
       "step": 27985
     },
     {
-      "epoch": 0.98,
-      "learning_rate": 4.6435637132483855e-05,
-      "loss": 0.2948,
+      "epoch": 1.0087577035355173,
+      "grad_norm": 0.1694299876689911,
+      "learning_rate": 4.623420127474619e-05,
+      "loss": 0.4075,
       "step": 27990
     },
     {
-      "epoch": 0.98,
-      "learning_rate": 4.6434171018148674e-05,
-      "loss": 0.2974,
+      "epoch": 1.008937903196742,
+      "grad_norm": 0.18183918297290802,
+      "learning_rate": 4.62326609361838e-05,
+      "loss": 0.3977,
       "step": 27995
     },
     {
-      "epoch": 0.99,
-      "learning_rate": 4.6432704625504915e-05,
-      "loss": 0.31,
+      "epoch": 1.0091181028579665,
+      "grad_norm": 0.18939772248268127,
+      "learning_rate": 4.623112030833052e-05,
+      "loss": 0.4101,
       "step": 28000
     },
     {
-      "epoch": 0.99,
-      "eval_loss": 0.2925254702568054,
-      "eval_runtime": 10.5771,
-      "eval_samples_per_second": 9.454,
-      "eval_steps_per_second": 9.454,
+      "epoch": 1.0091181028579665,
+      "eval_loss": 0.4545483887195587,
+      "eval_runtime": 3.5395,
+      "eval_samples_per_second": 28.252,
+      "eval_steps_per_second": 7.063,
       "step": 28000
     },
     {
-      "epoch": 0.99,
-      "learning_rate": 4.6431237954571605e-05,
-      "loss": 0.3,
+      "epoch": 1.0092983025191913,
+      "grad_norm": 0.15189491212368011,
+      "learning_rate": 4.622957939120734e-05,
+      "loss": 0.416,
       "step": 28005
     },
     {
-      "epoch": 0.99,
-      "learning_rate": 4.64297710053678e-05,
-      "loss": 0.3034,
+      "epoch": 1.0094785021804158,
+      "grad_norm": 0.1771238148212433,
+      "learning_rate": 4.6228038184835256e-05,
+      "loss": 0.4021,
       "step": 28010
     },
     {
-      "epoch": 0.99,
-      "learning_rate": 4.642830377791253e-05,
-      "loss": 0.2771,
+      "epoch": 1.0096587018416405,
+      "grad_norm": 0.18271222710609436,
+      "learning_rate": 4.622649668923525e-05,
+      "loss": 0.4491,
       "step": 28015
     },
     {
-      "epoch": 0.99,
-      "learning_rate": 4.642683627222486e-05,
-      "loss": 0.2768,
+      "epoch": 1.0098389015028653,
+      "grad_norm": 0.19635730981826782,
+      "learning_rate": 4.622495490442834e-05,
+      "loss": 0.408,
       "step": 28020
     },
     {
-      "epoch": 0.99,
-      "learning_rate": 4.642536848832384e-05,
-      "loss": 0.2962,
+      "epoch": 1.0100191011640898,
+      "grad_norm": 0.1971067190170288,
+      "learning_rate": 4.622341283043553e-05,
+      "loss": 0.4363,
       "step": 28025
     },
     {
-      "epoch": 0.99,
-      "learning_rate": 4.642390042622854e-05,
-      "loss": 0.2933,
+      "epoch": 1.0101993008253145,
+      "grad_norm": 0.1596841812133789,
+      "learning_rate": 4.622187046727783e-05,
+      "loss": 0.4135,
       "step": 28030
     },
     {
-      "epoch": 0.99,
-      "learning_rate": 4.6422432085958014e-05,
-      "loss": 0.282,
+      "epoch": 1.010379500486539,
+      "grad_norm": 0.20057372748851776,
+      "learning_rate": 4.622032781497625e-05,
+      "loss": 0.4142,
       "step": 28035
     },
     {
-      "epoch": 0.99,
-      "learning_rate": 4.642096346753131e-05,
-      "loss": 0.3212,
+      "epoch": 1.0105597001477638,
+      "grad_norm": 0.21350523829460144,
+      "learning_rate": 4.6218784873551816e-05,
+      "loss": 0.4545,
       "step": 28040
     },
     {
-      "epoch": 0.99,
-      "learning_rate": 4.6419494570967524e-05,
-      "loss": 0.2929,
+      "epoch": 1.0107398998089883,
+      "grad_norm": 0.21572305262088776,
+      "learning_rate": 4.621724164302554e-05,
+      "loss": 0.4428,
       "step": 28045
     },
     {
-      "epoch": 0.99,
-      "learning_rate": 4.6418025396285716e-05,
-      "loss": 0.291,
+      "epoch": 1.010920099470213,
+      "grad_norm": 0.15721938014030457,
+      "learning_rate": 4.621569812341846e-05,
+      "loss": 0.4368,
       "step": 28050
     },
     {
-      "epoch": 0.99,
-      "learning_rate": 4.641655594350496e-05,
-      "loss": 0.2882,
+      "epoch": 1.0111002991314377,
+      "grad_norm": 0.18772733211517334,
+      "learning_rate": 4.62141543147516e-05,
+      "loss": 0.4255,
       "step": 28055
     },
     {
-      "epoch": 0.99,
-      "learning_rate": 4.641508621264434e-05,
-      "loss": 0.3456,
+      "epoch": 1.0112804987926622,
+      "grad_norm": 0.15941715240478516,
+      "learning_rate": 4.6212610217045985e-05,
+      "loss": 0.4191,
       "step": 28060
     },
     {
-      "epoch": 0.99,
-      "learning_rate": 4.641361620372294e-05,
-      "loss": 0.2915,
+      "epoch": 1.011460698453887,
+      "grad_norm": 0.191365584731102,
+      "learning_rate": 4.6211065830322674e-05,
+      "loss": 0.4096,
       "step": 28065
     },
     {
-      "epoch": 0.99,
-      "learning_rate": 4.6412145916759844e-05,
-      "loss": 0.3201,
+      "epoch": 1.0116408981151115,
+      "grad_norm": 0.15204831957817078,
+      "learning_rate": 4.62095211546027e-05,
+      "loss": 0.4178,
       "step": 28070
     },
     {
-      "epoch": 0.99,
-      "learning_rate": 4.641067535177414e-05,
-      "loss": 0.3373,
+      "epoch": 1.0118210977763362,
+      "grad_norm": 0.1380978375673294,
+      "learning_rate": 4.62079761899071e-05,
+      "loss": 0.3991,
       "step": 28075
     },
     {
-      "epoch": 0.99,
-      "learning_rate": 4.640920450878493e-05,
-      "loss": 0.303,
+      "epoch": 1.0120012974375607,
+      "grad_norm": 0.1999700665473938,
+      "learning_rate": 4.6206430936256925e-05,
+      "loss": 0.4427,
       "step": 28080
     },
     {
-      "epoch": 0.99,
-      "learning_rate": 4.6407733387811314e-05,
-      "loss": 0.278,
+      "epoch": 1.0121814970987855,
+      "grad_norm": 0.16516053676605225,
+      "learning_rate": 4.620488539367325e-05,
+      "loss": 0.4186,
       "step": 28085
     },
     {
-      "epoch": 0.99,
-      "learning_rate": 4.640626198887238e-05,
-      "loss": 0.2931,
+      "epoch": 1.0123616967600102,
+      "grad_norm": 0.1769554764032364,
+      "learning_rate": 4.62033395621771e-05,
+      "loss": 0.3821,
       "step": 28090
     },
     {
-      "epoch": 0.99,
-      "learning_rate": 4.6404790311987245e-05,
-      "loss": 0.2942,
+      "epoch": 1.0125418964212347,
+      "grad_norm": 0.15776276588439941,
+      "learning_rate": 4.6201793441789566e-05,
+      "loss": 0.4336,
       "step": 28095
     },
     {
-      "epoch": 0.99,
-      "learning_rate": 4.6403318357175016e-05,
-      "loss": 0.2935,
+      "epoch": 1.0127220960824594,
+      "grad_norm": 0.16874942183494568,
+      "learning_rate": 4.620024703253169e-05,
+      "loss": 0.4315,
       "step": 28100
     },
     {
-      "epoch": 0.99,
-      "learning_rate": 4.640184612445481e-05,
-      "loss": 0.2873,
+      "epoch": 1.012902295743684,
+      "grad_norm": 0.18879170715808868,
+      "learning_rate": 4.619870033442456e-05,
+      "loss": 0.4295,
       "step": 28105
     },
     {
-      "epoch": 0.99,
-      "learning_rate": 4.640037361384573e-05,
-      "loss": 0.3022,
+      "epoch": 1.0130824954049087,
+      "grad_norm": 0.17838184535503387,
+      "learning_rate": 4.619715334748924e-05,
+      "loss": 0.3818,
       "step": 28110
     },
     {
-      "epoch": 0.99,
-      "learning_rate": 4.6398900825366895e-05,
-      "loss": 0.2989,
+      "epoch": 1.0132626950661332,
+      "grad_norm": 0.1531950831413269,
+      "learning_rate": 4.619560607174681e-05,
+      "loss": 0.4307,
       "step": 28115
     },
     {
-      "epoch": 0.99,
-      "learning_rate": 4.6397427759037437e-05,
-      "loss": 0.3209,
+      "epoch": 1.013442894727358,
+      "grad_norm": 0.15358266234397888,
+      "learning_rate": 4.619405850721835e-05,
+      "loss": 0.3952,
       "step": 28120
     },
     {
-      "epoch": 0.99,
-      "learning_rate": 4.639595441487649e-05,
-      "loss": 0.312,
+      "epoch": 1.0136230943885824,
+      "grad_norm": 0.16065557301044464,
+      "learning_rate": 4.6192510653924956e-05,
+      "loss": 0.4233,
       "step": 28125
     },
     {
-      "epoch": 0.99,
-      "learning_rate": 4.639448079290317e-05,
-      "loss": 0.2821,
+      "epoch": 1.0138032940498072,
+      "grad_norm": 0.2141285538673401,
+      "learning_rate": 4.6190962511887694e-05,
+      "loss": 0.3981,
       "step": 28130
     },
     {
-      "epoch": 0.99,
-      "learning_rate": 4.6393006893136616e-05,
-      "loss": 0.2885,
+      "epoch": 1.013983493711032,
+      "grad_norm": 0.23794370889663696,
+      "learning_rate": 4.618941408112768e-05,
+      "loss": 0.3997,
       "step": 28135
     },
     {
-      "epoch": 0.99,
-      "learning_rate": 4.6391532715595964e-05,
-      "loss": 0.2991,
+      "epoch": 1.0141636933722564,
+      "grad_norm": 0.17404253780841827,
+      "learning_rate": 4.6187865361665995e-05,
+      "loss": 0.4309,
       "step": 28140
     },
     {
-      "epoch": 0.99,
-      "learning_rate": 4.639005826030035e-05,
-      "loss": 0.326,
+      "epoch": 1.0143438930334812,
+      "grad_norm": 0.1899445503950119,
+      "learning_rate": 4.618631635352375e-05,
+      "loss": 0.4306,
       "step": 28145
     },
     {
-      "epoch": 0.99,
-      "learning_rate": 4.638858352726895e-05,
-      "loss": 0.2807,
+      "epoch": 1.0145240926947057,
+      "grad_norm": 0.14172546565532684,
+      "learning_rate": 4.6184767056722044e-05,
+      "loss": 0.4384,
       "step": 28150
     },
     {
-      "epoch": 0.99,
-      "learning_rate": 4.638710851652087e-05,
-      "loss": 0.2922,
+      "epoch": 1.0147042923559304,
+      "grad_norm": 0.18066340684890747,
+      "learning_rate": 4.618321747128199e-05,
+      "loss": 0.4248,
       "step": 28155
     },
     {
-      "epoch": 0.99,
-      "learning_rate": 4.638563322807528e-05,
-      "loss": 0.2952,
+      "epoch": 1.014884492017155,
+      "grad_norm": 0.18051232397556305,
+      "learning_rate": 4.61816675972247e-05,
+      "loss": 0.4058,
       "step": 28160
     },
     {
-      "epoch": 0.99,
-      "learning_rate": 4.638415766195135e-05,
-      "loss": 0.2912,
+      "epoch": 1.0150646916783796,
+      "grad_norm": 0.21809528768062592,
+      "learning_rate": 4.6180117434571285e-05,
+      "loss": 0.3864,
       "step": 28165
     },
     {
-      "epoch": 0.99,
-      "learning_rate": 4.638268181816821e-05,
-      "loss": 0.285,
+      "epoch": 1.0152448913396044,
+      "grad_norm": 0.1755502074956894,
+      "learning_rate": 4.6178566983342875e-05,
+      "loss": 0.411,
       "step": 28170
     },
     {
-      "epoch": 0.99,
-      "learning_rate": 4.6381205696745046e-05,
-      "loss": 0.2991,
+      "epoch": 1.015425091000829,
+      "grad_norm": 0.2000172734260559,
+      "learning_rate": 4.617701624356059e-05,
+      "loss": 0.4297,
       "step": 28175
     },
     {
-      "epoch": 0.99,
-      "learning_rate": 4.637972929770101e-05,
-      "loss": 0.3202,
+      "epoch": 1.0156052906620536,
+      "grad_norm": 0.15490785241127014,
+      "learning_rate": 4.6175465215245556e-05,
+      "loss": 0.4408,
       "step": 28180
     },
     {
-      "epoch": 0.99,
-      "learning_rate": 4.637825262105528e-05,
-      "loss": 0.2775,
+      "epoch": 1.0157854903232781,
+      "grad_norm": 0.1819140464067459,
+      "learning_rate": 4.617391389841891e-05,
+      "loss": 0.3964,
       "step": 28185
     },
     {
-      "epoch": 0.99,
-      "learning_rate": 4.6376775666827045e-05,
-      "loss": 0.3135,
+      "epoch": 1.0159656899845029,
+      "grad_norm": 0.1457093507051468,
+      "learning_rate": 4.617236229310179e-05,
+      "loss": 0.3747,
       "step": 28190
     },
     {
-      "epoch": 0.99,
-      "learning_rate": 4.6375298435035455e-05,
-      "loss": 0.3039,
+      "epoch": 1.0161458896457274,
+      "grad_norm": 0.18323257565498352,
+      "learning_rate": 4.617081039931532e-05,
+      "loss": 0.4372,
       "step": 28195
     },
     {
-      "epoch": 0.99,
-      "learning_rate": 4.63738209256997e-05,
-      "loss": 0.3044,
+      "epoch": 1.0163260893069521,
+      "grad_norm": 0.16842052340507507,
+      "learning_rate": 4.616925821708067e-05,
+      "loss": 0.4285,
       "step": 28200
     },
     {
-      "epoch": 0.99,
-      "learning_rate": 4.637234313883896e-05,
-      "loss": 0.3249,
+      "epoch": 1.0165062889681766,
+      "grad_norm": 0.1632702648639679,
+      "learning_rate": 4.6167705746418974e-05,
+      "loss": 0.395,
       "step": 28205
     },
     {
-      "epoch": 0.99,
-      "learning_rate": 4.6370865074472435e-05,
-      "loss": 0.3282,
+      "epoch": 1.0166864886294014,
+      "grad_norm": 0.15500997006893158,
+      "learning_rate": 4.616615298735138e-05,
+      "loss": 0.4169,
       "step": 28210
     },
     {
-      "epoch": 0.99,
-      "learning_rate": 4.636938673261931e-05,
-      "loss": 0.3199,
+      "epoch": 1.016866688290626,
+      "grad_norm": 0.18930770456790924,
+      "learning_rate": 4.616459993989906e-05,
+      "loss": 0.4304,
       "step": 28215
     },
     {
-      "epoch": 0.99,
-      "learning_rate": 4.6367908113298784e-05,
-      "loss": 0.2861,
+      "epoch": 1.0170468879518506,
+      "grad_norm": 0.19972793757915497,
+      "learning_rate": 4.616304660408315e-05,
+      "loss": 0.4025,
       "step": 28220
     },
     {
-      "epoch": 0.99,
-      "learning_rate": 4.636642921653006e-05,
-      "loss": 0.2998,
+      "epoch": 1.0172270876130753,
+      "grad_norm": 0.1533392369747162,
+      "learning_rate": 4.6161492979924834e-05,
+      "loss": 0.4091,
       "step": 28225
     },
     {
-      "epoch": 0.99,
-      "learning_rate": 4.6364950042332326e-05,
-      "loss": 0.2819,
+      "epoch": 1.0174072872742999,
+      "grad_norm": 0.17162777483463287,
+      "learning_rate": 4.615993906744528e-05,
+      "loss": 0.4054,
       "step": 28230
     },
     {
-      "epoch": 0.99,
-      "learning_rate": 4.636347059072479e-05,
-      "loss": 0.312,
+      "epoch": 1.0175874869355246,
+      "grad_norm": 0.2027086317539215,
+      "learning_rate": 4.615838486666564e-05,
+      "loss": 0.3837,
       "step": 28235
     },
     {
-      "epoch": 0.99,
-      "learning_rate": 4.636199086172667e-05,
-      "loss": 0.2945,
+      "epoch": 1.017767686596749,
+      "grad_norm": 0.1900579333305359,
+      "learning_rate": 4.6156830377607105e-05,
+      "loss": 0.4096,
       "step": 28240
     },
     {
-      "epoch": 0.99,
-      "learning_rate": 4.636051085535718e-05,
-      "loss": 0.2942,
+      "epoch": 1.0179478862579738,
+      "grad_norm": 0.15131549537181854,
+      "learning_rate": 4.615527560029086e-05,
+      "loss": 0.4474,
       "step": 28245
     },
     {
-      "epoch": 0.99,
-      "learning_rate": 4.635903057163553e-05,
-      "loss": 0.3138,
+      "epoch": 1.0181280859191986,
+      "grad_norm": 0.1787458062171936,
+      "learning_rate": 4.615372053473808e-05,
+      "loss": 0.3827,
       "step": 28250
     },
     {
-      "epoch": 0.99,
-      "learning_rate": 4.6357550010580954e-05,
-      "loss": 0.2827,
+      "epoch": 1.018308285580423,
+      "grad_norm": 0.18287619948387146,
+      "learning_rate": 4.6152165180969944e-05,
+      "loss": 0.4215,
       "step": 28255
     },
     {
-      "epoch": 0.99,
-      "learning_rate": 4.635606917221266e-05,
-      "loss": 0.2791,
+      "epoch": 1.0184884852416478,
+      "grad_norm": 0.15783603489398956,
+      "learning_rate": 4.6150609539007664e-05,
+      "loss": 0.3793,
       "step": 28260
     },
     {
-      "epoch": 0.99,
-      "learning_rate": 4.6354588056549883e-05,
-      "loss": 0.3206,
+      "epoch": 1.0186686849028723,
+      "grad_norm": 0.2039651870727539,
+      "learning_rate": 4.614905360887241e-05,
+      "loss": 0.4043,
       "step": 28265
     },
     {
-      "epoch": 0.99,
-      "learning_rate": 4.6353106663611843e-05,
-      "loss": 0.3046,
+      "epoch": 1.018848884564097,
+      "grad_norm": 0.18400070071220398,
+      "learning_rate": 4.61474973905854e-05,
+      "loss": 0.4042,
       "step": 28270
     },
     {
-      "epoch": 0.99,
-      "learning_rate": 4.6351624993417795e-05,
-      "loss": 0.3276,
+      "epoch": 1.0190290842253216,
+      "grad_norm": 0.19877344369888306,
+      "learning_rate": 4.614594088416784e-05,
+      "loss": 0.4057,
       "step": 28275
     },
     {
-      "epoch": 0.99,
-      "learning_rate": 4.635014304598697e-05,
-      "loss": 0.3149,
+      "epoch": 1.0192092838865463,
+      "grad_norm": 0.19915084540843964,
+      "learning_rate": 4.614438408964092e-05,
+      "loss": 0.4553,
       "step": 28280
     },
     {
-      "epoch": 1.0,
-      "learning_rate": 4.6348660821338605e-05,
-      "loss": 0.2965,
+      "epoch": 1.019389483547771,
+      "grad_norm": 0.19474834203720093,
+      "learning_rate": 4.614282700702587e-05,
+      "loss": 0.4374,
       "step": 28285
     },
     {
-      "epoch": 1.0,
-      "learning_rate": 4.6347178319491946e-05,
-      "loss": 0.2983,
+      "epoch": 1.0195696832089955,
+      "grad_norm": 0.1851280927658081,
+      "learning_rate": 4.614126963634389e-05,
+      "loss": 0.4065,
       "step": 28290
     },
     {
-      "epoch": 1.0,
-      "learning_rate": 4.634569554046625e-05,
-      "loss": 0.3216,
+      "epoch": 1.0197498828702203,
+      "grad_norm": 0.1995769590139389,
+      "learning_rate": 4.6139711977616207e-05,
+      "loss": 0.4195,
       "step": 28295
     },
     {
-      "epoch": 1.0,
-      "learning_rate": 4.6344212484280756e-05,
-      "loss": 0.3091,
+      "epoch": 1.0199300825314448,
+      "grad_norm": 0.17774897813796997,
+      "learning_rate": 4.6138154030864036e-05,
+      "loss": 0.4494,
       "step": 28300
     },
     {
-      "epoch": 1.0,
-      "learning_rate": 4.634272915095475e-05,
-      "loss": 0.3023,
+      "epoch": 1.0201102821926695,
+      "grad_norm": 0.1672256588935852,
+      "learning_rate": 4.613659579610861e-05,
+      "loss": 0.413,
       "step": 28305
     },
     {
-      "epoch": 1.0,
-      "learning_rate": 4.634124554050745e-05,
-      "loss": 0.3142,
+      "epoch": 1.020290481853894,
+      "grad_norm": 0.17119145393371582,
+      "learning_rate": 4.613503727337116e-05,
+      "loss": 0.3927,
       "step": 28310
     },
     {
-      "epoch": 1.0,
-      "learning_rate": 4.633976165295815e-05,
-      "loss": 0.2853,
+      "epoch": 1.0204706815151188,
+      "grad_norm": 0.16596178710460663,
+      "learning_rate": 4.613347846267292e-05,
+      "loss": 0.4003,
       "step": 28315
     },
     {
-      "epoch": 1.0,
-      "learning_rate": 4.6338277488326095e-05,
-      "loss": 0.3037,
+      "epoch": 1.0206508811763433,
+      "grad_norm": 0.1888241469860077,
+      "learning_rate": 4.6131919364035126e-05,
+      "loss": 0.4401,
       "step": 28320
     },
     {
-      "epoch": 1.0,
-      "learning_rate": 4.633679304663059e-05,
-      "loss": 0.3117,
+      "epoch": 1.020831080837568,
+      "grad_norm": 0.2028852254152298,
+      "learning_rate": 4.613035997747902e-05,
+      "loss": 0.4529,
       "step": 28325
     },
     {
-      "epoch": 1.0,
-      "learning_rate": 4.633530832789088e-05,
-      "loss": 0.2757,
+      "epoch": 1.0210112804987928,
+      "grad_norm": 0.18602454662322998,
+      "learning_rate": 4.612880030302585e-05,
+      "loss": 0.4408,
       "step": 28330
     },
     {
-      "epoch": 1.0,
-      "learning_rate": 4.633382333212625e-05,
-      "loss": 0.3027,
+      "epoch": 1.0211914801600173,
+      "grad_norm": 0.2020941823720932,
+      "learning_rate": 4.6127240340696876e-05,
+      "loss": 0.4332,
       "step": 28335
     },
     {
-      "epoch": 1.0,
-      "learning_rate": 4.633233805935598e-05,
-      "loss": 0.2905,
+      "epoch": 1.021371679821242,
+      "grad_norm": 0.17948292195796967,
+      "learning_rate": 4.6125680090513334e-05,
+      "loss": 0.4235,
       "step": 28340
     },
     {
-      "epoch": 1.0,
-      "learning_rate": 4.6330852509599364e-05,
-      "loss": 0.3119,
+      "epoch": 1.0215518794824665,
+      "grad_norm": 0.17429165542125702,
+      "learning_rate": 4.612411955249649e-05,
+      "loss": 0.4238,
       "step": 28345
     },
     {
-      "epoch": 1.0,
-      "learning_rate": 4.632936668287568e-05,
-      "loss": 0.2912,
+      "epoch": 1.0217320791436912,
+      "grad_norm": 0.18674267828464508,
+      "learning_rate": 4.612255872666762e-05,
+      "loss": 0.3849,
       "step": 28350
     },
     {
-      "epoch": 1.0,
-      "learning_rate": 4.632788057920423e-05,
-      "loss": 0.3334,
+      "epoch": 1.0219122788049158,
+      "grad_norm": 0.13337041437625885,
+      "learning_rate": 4.612099761304796e-05,
+      "loss": 0.4294,
       "step": 28355
     },
     {
-      "epoch": 1.0,
-      "learning_rate": 4.632639419860431e-05,
-      "loss": 0.3038,
+      "epoch": 1.0220924784661405,
+      "grad_norm": 0.15359793603420258,
+      "learning_rate": 4.6119436211658805e-05,
+      "loss": 0.4145,
       "step": 28360
     },
     {
-      "epoch": 1.0,
-      "learning_rate": 4.632490754109521e-05,
-      "loss": 0.2781,
+      "epoch": 1.0222726781273652,
+      "grad_norm": 0.17645424604415894,
+      "learning_rate": 4.611787452252142e-05,
+      "loss": 0.4163,
       "step": 28365
     },
     {
-      "epoch": 1.0,
-      "learning_rate": 4.6323420606696245e-05,
-      "loss": 0.2914,
+      "epoch": 1.0224528777885897,
+      "grad_norm": 0.14870162308216095,
+      "learning_rate": 4.6116312545657083e-05,
+      "loss": 0.382,
       "step": 28370
     },
     {
-      "epoch": 1.0,
-      "learning_rate": 4.632193339542671e-05,
-      "loss": 0.304,
+      "epoch": 1.0226330774498145,
+      "grad_norm": 0.18462905287742615,
+      "learning_rate": 4.611475028108707e-05,
+      "loss": 0.4238,
       "step": 28375
     },
     {
-      "epoch": 1.0,
-      "learning_rate": 4.632044590730593e-05,
-      "loss": 0.2986,
+      "epoch": 1.022813277111039,
+      "grad_norm": 0.1805817037820816,
+      "learning_rate": 4.611318772883268e-05,
+      "loss": 0.4203,
       "step": 28380
     },
     {
-      "epoch": 1.0,
-      "learning_rate": 4.631895814235321e-05,
-      "loss": 0.3096,
+      "epoch": 1.0229934767722637,
+      "grad_norm": 0.18119722604751587,
+      "learning_rate": 4.6111624888915196e-05,
+      "loss": 0.4058,
       "step": 28385
     },
     {
-      "epoch": 1.0,
-      "learning_rate": 4.6317470100587856e-05,
-      "loss": 0.3176,
+      "epoch": 1.0231736764334882,
+      "grad_norm": 0.14671674370765686,
+      "learning_rate": 4.611006176135591e-05,
+      "loss": 0.4227,
       "step": 28390
     },
     {
-      "epoch": 1.0,
-      "learning_rate": 4.631598178202921e-05,
-      "loss": 0.3122,
+      "epoch": 1.023353876094713,
+      "grad_norm": 0.13989455997943878,
+      "learning_rate": 4.610849834617611e-05,
+      "loss": 0.3985,
       "step": 28395
     },
     {
-      "epoch": 1.0,
-      "learning_rate": 4.631449318669659e-05,
-      "loss": 0.2804,
+      "epoch": 1.0235340757559377,
+      "grad_norm": 0.17615246772766113,
+      "learning_rate": 4.610693464339711e-05,
+      "loss": 0.3773,
       "step": 28400
     },
     {
-      "epoch": 1.0,
-      "learning_rate": 4.631300431460932e-05,
-      "loss": 0.2944,
+      "epoch": 1.0237142754171622,
+      "grad_norm": 0.17603693902492523,
+      "learning_rate": 4.6105370653040216e-05,
+      "loss": 0.3988,
       "step": 28405
     },
     {
-      "epoch": 1.0,
-      "learning_rate": 4.6311515165786736e-05,
-      "loss": 0.3003,
+      "epoch": 1.023894475078387,
+      "grad_norm": 0.15467150509357452,
+      "learning_rate": 4.6103806375126735e-05,
+      "loss": 0.4082,
       "step": 28410
     },
     {
-      "epoch": 1.0,
-      "learning_rate": 4.631002574024818e-05,
-      "loss": 0.2809,
+      "epoch": 1.0240746747396114,
+      "grad_norm": 0.19032159447669983,
+      "learning_rate": 4.6102241809677974e-05,
+      "loss": 0.4251,
       "step": 28415
     },
     {
-      "epoch": 1.0,
-      "learning_rate": 4.630853603801296e-05,
-      "loss": 0.3108,
+      "epoch": 1.0242548744008362,
+      "grad_norm": 0.19951966404914856,
+      "learning_rate": 4.610067695671525e-05,
+      "loss": 0.4239,
       "step": 28420
     },
     {
-      "epoch": 1.0,
-      "learning_rate": 4.6307046059100465e-05,
-      "loss": 0.2931,
+      "epoch": 1.0244350740620607,
+      "grad_norm": 0.1830962896347046,
+      "learning_rate": 4.609911181625989e-05,
+      "loss": 0.4016,
       "step": 28425
     },
     {
-      "epoch": 1.0,
-      "learning_rate": 4.630555580353001e-05,
-      "loss": 0.2833,
+      "epoch": 1.0246152737232854,
+      "grad_norm": 0.18551193177700043,
+      "learning_rate": 4.609754638833322e-05,
+      "loss": 0.3889,
       "step": 28430
     },
     {
-      "epoch": 1.0,
-      "learning_rate": 4.630406527132095e-05,
-      "loss": 0.2914,
+      "epoch": 1.02479547338451,
+      "grad_norm": 0.18302911520004272,
+      "learning_rate": 4.609598067295656e-05,
+      "loss": 0.4453,
       "step": 28435
     },
     {
-      "epoch": 1.0,
-      "learning_rate": 4.6302574462492656e-05,
-      "loss": 0.2863,
+      "epoch": 1.0249756730457347,
+      "grad_norm": 0.15674737095832825,
+      "learning_rate": 4.6094414670151253e-05,
+      "loss": 0.4165,
       "step": 28440
     },
     {
-      "epoch": 1.0,
-      "learning_rate": 4.630108337706446e-05,
-      "loss": 0.2728,
+      "epoch": 1.0251558727069594,
+      "grad_norm": 0.18824201822280884,
+      "learning_rate": 4.609284837993863e-05,
+      "loss": 0.4216,
       "step": 28445
     },
     {
-      "epoch": 1.0,
-      "learning_rate": 4.629959201505574e-05,
-      "loss": 0.2981,
+      "epoch": 1.025336072368184,
+      "grad_norm": 0.13510337471961975,
+      "learning_rate": 4.609128180234003e-05,
+      "loss": 0.4233,
       "step": 28450
     },
     {
-      "epoch": 1.0,
-      "learning_rate": 4.629810037648585e-05,
-      "loss": 0.2896,
+      "epoch": 1.0255162720294086,
+      "grad_norm": 0.17444869875907898,
+      "learning_rate": 4.60897149373768e-05,
+      "loss": 0.4555,
       "step": 28455
     },
     {
-      "epoch": 1.0,
-      "learning_rate": 4.6296608461374166e-05,
-      "loss": 0.2883,
+      "epoch": 1.0256964716906332,
+      "grad_norm": 0.14340052008628845,
+      "learning_rate": 4.6088147785070284e-05,
+      "loss": 0.3925,
       "step": 28460
     },
     {
-      "epoch": 1.0,
-      "learning_rate": 4.6295116269740055e-05,
-      "loss": 0.2993,
+      "epoch": 1.025876671351858,
+      "grad_norm": 0.15294180810451508,
+      "learning_rate": 4.608658034544184e-05,
+      "loss": 0.4104,
       "step": 28465
     },
     {
-      "epoch": 1.0,
-      "learning_rate": 4.6293623801602894e-05,
-      "loss": 0.2843,
+      "epoch": 1.0260568710130824,
+      "grad_norm": 0.21026860177516937,
+      "learning_rate": 4.608501261851282e-05,
+      "loss": 0.4463,
       "step": 28470
     },
     {
-      "epoch": 1.0,
-      "learning_rate": 4.6292131056982056e-05,
-      "loss": 0.2819,
+      "epoch": 1.0262370706743071,
+      "grad_norm": 0.18391993641853333,
+      "learning_rate": 4.6083444604304584e-05,
+      "loss": 0.4285,
       "step": 28475
     },
     {
-      "epoch": 1.0,
-      "learning_rate": 4.6290638035896935e-05,
-      "loss": 0.2676,
+      "epoch": 1.0264172703355319,
+      "grad_norm": 0.16915670037269592,
+      "learning_rate": 4.60818763028385e-05,
+      "loss": 0.3984,
       "step": 28480
     },
     {
-      "epoch": 1.0,
-      "learning_rate": 4.6289144738366905e-05,
-      "loss": 0.2986,
+      "epoch": 1.0265974699967564,
+      "grad_norm": 0.15066103637218475,
+      "learning_rate": 4.608030771413593e-05,
+      "loss": 0.4118,
       "step": 28485
     },
     {
-      "epoch": 1.0,
-      "learning_rate": 4.628765116441136e-05,
-      "loss": 0.2851,
+      "epoch": 1.0267776696579811,
+      "grad_norm": 0.1730443239212036,
+      "learning_rate": 4.607873883821825e-05,
+      "loss": 0.4169,
       "step": 28490
     },
     {
-      "epoch": 1.0,
-      "learning_rate": 4.62861573140497e-05,
-      "loss": 0.3139,
+      "epoch": 1.0269578693192056,
+      "grad_norm": 0.2241186797618866,
+      "learning_rate": 4.6077169675106836e-05,
+      "loss": 0.4354,
       "step": 28495
     },
     {
-      "epoch": 1.0,
-      "learning_rate": 4.628466318730131e-05,
-      "loss": 0.2856,
+      "epoch": 1.0271380689804304,
+      "grad_norm": 0.20758524537086487,
+      "learning_rate": 4.6075600224823066e-05,
+      "loss": 0.4531,
       "step": 28500
     },
     {
-      "epoch": 1.0,
-      "eval_loss": 0.2926534414291382,
-      "eval_runtime": 10.5975,
-      "eval_samples_per_second": 9.436,
-      "eval_steps_per_second": 9.436,
+      "epoch": 1.0271380689804304,
+      "eval_loss": 0.45461568236351013,
+      "eval_runtime": 3.545,
+      "eval_samples_per_second": 28.209,
+      "eval_steps_per_second": 7.052,
       "step": 28500
     },
     {
-      "epoch": 1.0,
-      "learning_rate": 4.62831687841856e-05,
-      "loss": 0.2937,
+      "epoch": 1.0273182686416549,
+      "grad_norm": 0.143527090549469,
+      "learning_rate": 4.607403048738832e-05,
+      "loss": 0.3849,
       "step": 28505
     },
     {
-      "epoch": 1.0,
-      "learning_rate": 4.6281674104721966e-05,
-      "loss": 0.2959,
+      "epoch": 1.0274984683028796,
+      "grad_norm": 0.17763735353946686,
+      "learning_rate": 4.607246046282399e-05,
+      "loss": 0.4174,
       "step": 28510
     },
     {
-      "epoch": 1.0,
-      "learning_rate": 4.628017914892982e-05,
-      "loss": 0.2815,
+      "epoch": 1.0276786679641041,
+      "grad_norm": 0.2017643004655838,
+      "learning_rate": 4.607089015115147e-05,
+      "loss": 0.4237,
       "step": 28515
     },
     {
-      "epoch": 1.0,
-      "learning_rate": 4.627868391682857e-05,
-      "loss": 0.2922,
+      "epoch": 1.0278588676253289,
+      "grad_norm": 0.16140924394130707,
+      "learning_rate": 4.6069319552392145e-05,
+      "loss": 0.4097,
       "step": 28520
     },
     {
-      "epoch": 1.0,
-      "learning_rate": 4.627718840843763e-05,
-      "loss": 0.3035,
+      "epoch": 1.0280390672865536,
+      "grad_norm": 0.16550087928771973,
+      "learning_rate": 4.6067748666567425e-05,
+      "loss": 0.3789,
       "step": 28525
     },
     {
-      "epoch": 1.0,
-      "learning_rate": 4.6275692623776434e-05,
-      "loss": 0.2818,
+      "epoch": 1.028219266947778,
+      "grad_norm": 0.1847558319568634,
+      "learning_rate": 4.60661774936987e-05,
+      "loss": 0.4227,
       "step": 28530
     },
     {
-      "epoch": 1.0,
-      "learning_rate": 4.627419656286438e-05,
-      "loss": 0.3022,
+      "epoch": 1.0283994666090028,
+      "grad_norm": 0.18655303120613098,
+      "learning_rate": 4.606460603380739e-05,
+      "loss": 0.4278,
       "step": 28535
     },
     {
-      "epoch": 1.0,
-      "learning_rate": 4.6272700225720906e-05,
-      "loss": 0.2987,
+      "epoch": 1.0285796662702273,
+      "grad_norm": 0.17594356834888458,
+      "learning_rate": 4.606303428691491e-05,
+      "loss": 0.4417,
       "step": 28540
     },
     {
-      "epoch": 1.0,
-      "learning_rate": 4.627120361236544e-05,
-      "loss": 0.2933,
+      "epoch": 1.028759865931452,
+      "grad_norm": 0.18802598118782043,
+      "learning_rate": 4.606146225304265e-05,
+      "loss": 0.4576,
       "step": 28545
     },
     {
-      "epoch": 1.0,
-      "learning_rate": 4.626970672281741e-05,
-      "loss": 0.2709,
+      "epoch": 1.0289400655926766,
+      "grad_norm": 0.190033420920372,
+      "learning_rate": 4.605988993221206e-05,
+      "loss": 0.3783,
       "step": 28550
     },
     {
-      "epoch": 1.0,
-      "learning_rate": 4.6268209557096256e-05,
-      "loss": 0.2913,
+      "epoch": 1.0291202652539013,
+      "grad_norm": 0.16919276118278503,
+      "learning_rate": 4.605831732444453e-05,
+      "loss": 0.4453,
       "step": 28555
     },
     {
-      "epoch": 1.0,
-      "learning_rate": 4.626671211522142e-05,
-      "loss": 0.2927,
+      "epoch": 1.029300464915126,
+      "grad_norm": 0.17263488471508026,
+      "learning_rate": 4.605674442976152e-05,
+      "loss": 0.3981,
       "step": 28560
     },
     {
-      "epoch": 1.0,
-      "learning_rate": 4.6265214397212344e-05,
-      "loss": 0.27,
+      "epoch": 1.0294806645763506,
+      "grad_norm": 0.18015895783901215,
+      "learning_rate": 4.6055171248184434e-05,
+      "loss": 0.3846,
       "step": 28565
     },
     {
-      "epoch": 1.01,
-      "learning_rate": 4.6263716403088464e-05,
-      "loss": 0.3177,
+      "epoch": 1.0296608642375753,
+      "grad_norm": 0.18624891340732574,
+      "learning_rate": 4.605359777973472e-05,
+      "loss": 0.4022,
       "step": 28570
     },
     {
-      "epoch": 1.01,
-      "learning_rate": 4.626221813286925e-05,
-      "loss": 0.3077,
+      "epoch": 1.0298410638987998,
+      "grad_norm": 0.18329830467700958,
+      "learning_rate": 4.6052024024433815e-05,
+      "loss": 0.4198,
       "step": 28575
     },
     {
-      "epoch": 1.01,
-      "learning_rate": 4.626071958657414e-05,
-      "loss": 0.2766,
+      "epoch": 1.0300212635600245,
+      "grad_norm": 0.17667533457279205,
+      "learning_rate": 4.605044998230315e-05,
+      "loss": 0.4046,
       "step": 28580
     },
     {
-      "epoch": 1.01,
-      "learning_rate": 4.62592207642226e-05,
-      "loss": 0.3328,
+      "epoch": 1.030201463221249,
+      "grad_norm": 0.1662374883890152,
+      "learning_rate": 4.604887565336419e-05,
+      "loss": 0.4322,
       "step": 28585
     },
     {
-      "epoch": 1.01,
-      "learning_rate": 4.625772166583409e-05,
-      "loss": 0.2749,
+      "epoch": 1.0303816628824738,
+      "grad_norm": 0.20289361476898193,
+      "learning_rate": 4.604730103763837e-05,
+      "loss": 0.4173,
       "step": 28590
     },
     {
-      "epoch": 1.01,
-      "learning_rate": 4.625622229142806e-05,
-      "loss": 0.266,
+      "epoch": 1.0305618625436985,
+      "grad_norm": 0.17283061146736145,
+      "learning_rate": 4.604572613514714e-05,
+      "loss": 0.4174,
       "step": 28595
     },
     {
-      "epoch": 1.01,
-      "learning_rate": 4.625472264102401e-05,
-      "loss": 0.2969,
+      "epoch": 1.030742062204923,
+      "grad_norm": 0.21077826619148254,
+      "learning_rate": 4.6044150945911974e-05,
+      "loss": 0.4365,
       "step": 28600
     },
     {
-      "epoch": 1.01,
-      "learning_rate": 4.6253222714641386e-05,
-      "loss": 0.2833,
+      "epoch": 1.0309222618661478,
+      "grad_norm": 0.19256427884101868,
+      "learning_rate": 4.604257546995433e-05,
+      "loss": 0.4491,
       "step": 28605
     },
     {
-      "epoch": 1.01,
-      "learning_rate": 4.625172251229968e-05,
-      "loss": 0.2772,
+      "epoch": 1.0311024615273723,
+      "grad_norm": 0.15783824026584625,
+      "learning_rate": 4.6040999707295665e-05,
+      "loss": 0.4072,
       "step": 28610
     },
     {
-      "epoch": 1.01,
-      "learning_rate": 4.6250222034018354e-05,
-      "loss": 0.3067,
+      "epoch": 1.031282661188597,
+      "grad_norm": 0.15280930697917938,
+      "learning_rate": 4.603942365795745e-05,
+      "loss": 0.4217,
       "step": 28615
     },
     {
-      "epoch": 1.01,
-      "learning_rate": 4.62487212798169e-05,
-      "loss": 0.2986,
+      "epoch": 1.0314628608498215,
+      "grad_norm": 0.16231854259967804,
+      "learning_rate": 4.603784732196116e-05,
+      "loss": 0.3995,
       "step": 28620
     },
     {
-      "epoch": 1.01,
-      "learning_rate": 4.62472202497148e-05,
-      "loss": 0.3008,
+      "epoch": 1.0316430605110463,
+      "grad_norm": 0.1602470427751541,
+      "learning_rate": 4.603627069932827e-05,
+      "loss": 0.4217,
       "step": 28625
     },
     {
-      "epoch": 1.01,
-      "learning_rate": 4.6245718943731556e-05,
-      "loss": 0.2753,
+      "epoch": 1.0318232601722708,
+      "grad_norm": 0.2087296098470688,
+      "learning_rate": 4.603469379008028e-05,
+      "loss": 0.4469,
       "step": 28630
     },
     {
-      "epoch": 1.01,
-      "learning_rate": 4.624421736188664e-05,
-      "loss": 0.2953,
+      "epoch": 1.0320034598334955,
+      "grad_norm": 0.18025125563144684,
+      "learning_rate": 4.603311659423864e-05,
+      "loss": 0.4198,
       "step": 28635
     },
     {
-      "epoch": 1.01,
-      "learning_rate": 4.624271550419957e-05,
-      "loss": 0.2925,
+      "epoch": 1.0321836594947202,
+      "grad_norm": 0.19022351503372192,
+      "learning_rate": 4.603153911182487e-05,
+      "loss": 0.4439,
       "step": 28640
     },
     {
-      "epoch": 1.01,
-      "learning_rate": 4.624121337068985e-05,
-      "loss": 0.302,
+      "epoch": 1.0323638591559448,
+      "grad_norm": 0.17020079493522644,
+      "learning_rate": 4.602996134286045e-05,
+      "loss": 0.3931,
       "step": 28645
     },
     {
-      "epoch": 1.01,
-      "learning_rate": 4.623971096137696e-05,
-      "loss": 0.3007,
+      "epoch": 1.0325440588171695,
+      "grad_norm": 0.17506608366966248,
+      "learning_rate": 4.602838328736688e-05,
+      "loss": 0.4457,
       "step": 28650
     },
     {
-      "epoch": 1.01,
-      "learning_rate": 4.6238208276280425e-05,
-      "loss": 0.2937,
+      "epoch": 1.032724258478394,
+      "grad_norm": 0.22321075201034546,
+      "learning_rate": 4.6026804945365655e-05,
+      "loss": 0.4347,
       "step": 28655
     },
     {
-      "epoch": 1.01,
-      "learning_rate": 4.6236705315419746e-05,
-      "loss": 0.2881,
+      "epoch": 1.0329044581396187,
+      "grad_norm": 0.16402897238731384,
+      "learning_rate": 4.602522631687829e-05,
+      "loss": 0.4257,
       "step": 28660
     },
     {
-      "epoch": 1.01,
-      "learning_rate": 4.623520207881445e-05,
-      "loss": 0.2761,
+      "epoch": 1.0330846578008432,
+      "grad_norm": 0.18853814899921417,
+      "learning_rate": 4.602364740192628e-05,
+      "loss": 0.4312,
       "step": 28665
     },
     {
-      "epoch": 1.01,
-      "learning_rate": 4.623369856648405e-05,
-      "loss": 0.2717,
+      "epoch": 1.033264857462068,
+      "grad_norm": 0.1805652529001236,
+      "learning_rate": 4.602206820053114e-05,
+      "loss": 0.4294,
       "step": 28670
     },
     {
-      "epoch": 1.01,
-      "learning_rate": 4.6232194778448064e-05,
-      "loss": 0.2863,
+      "epoch": 1.0334450571232927,
+      "grad_norm": 0.15662455558776855,
+      "learning_rate": 4.6020488712714404e-05,
+      "loss": 0.4338,
       "step": 28675
     },
     {
-      "epoch": 1.01,
-      "learning_rate": 4.623069071472602e-05,
-      "loss": 0.2899,
+      "epoch": 1.0336252567845172,
+      "grad_norm": 0.157413512468338,
+      "learning_rate": 4.6018908938497573e-05,
+      "loss": 0.4187,
       "step": 28680
     },
     {
-      "epoch": 1.01,
-      "learning_rate": 4.622918637533745e-05,
-      "loss": 0.2708,
+      "epoch": 1.033805456445742,
+      "grad_norm": 0.1799081712961197,
+      "learning_rate": 4.6017328877902176e-05,
+      "loss": 0.4412,
       "step": 28685
     },
     {
-      "epoch": 1.01,
-      "learning_rate": 4.622768176030189e-05,
-      "loss": 0.3391,
+      "epoch": 1.0339856561069665,
+      "grad_norm": 0.14759212732315063,
+      "learning_rate": 4.601574853094974e-05,
+      "loss": 0.407,
       "step": 28690
     },
     {
-      "epoch": 1.01,
-      "learning_rate": 4.622617686963887e-05,
-      "loss": 0.3078,
+      "epoch": 1.0341658557681912,
+      "grad_norm": 0.1873723268508911,
+      "learning_rate": 4.60141678976618e-05,
+      "loss": 0.4277,
       "step": 28695
     },
     {
-      "epoch": 1.01,
-      "learning_rate": 4.6224671703367925e-05,
-      "loss": 0.2943,
+      "epoch": 1.0343460554294157,
+      "grad_norm": 0.16125832498073578,
+      "learning_rate": 4.6012586978059893e-05,
+      "loss": 0.4602,
       "step": 28700
     },
     {
-      "epoch": 1.01,
-      "learning_rate": 4.622316626150861e-05,
-      "loss": 0.3048,
+      "epoch": 1.0345262550906404,
+      "grad_norm": 0.1734544336795807,
+      "learning_rate": 4.601100577216556e-05,
+      "loss": 0.4233,
       "step": 28705
     },
     {
-      "epoch": 1.01,
-      "learning_rate": 4.622166054408047e-05,
-      "loss": 0.2815,
+      "epoch": 1.034706454751865,
+      "grad_norm": 0.1755981296300888,
+      "learning_rate": 4.600942428000033e-05,
+      "loss": 0.4266,
       "step": 28710
     },
     {
-      "epoch": 1.01,
-      "learning_rate": 4.622015455110306e-05,
-      "loss": 0.2984,
+      "epoch": 1.0348866544130897,
+      "grad_norm": 0.16786262392997742,
+      "learning_rate": 4.600784250158577e-05,
+      "loss": 0.4204,
       "step": 28715
     },
     {
-      "epoch": 1.01,
-      "learning_rate": 4.621864828259592e-05,
-      "loss": 0.2911,
+      "epoch": 1.0350668540743144,
+      "grad_norm": 0.14706341922283173,
+      "learning_rate": 4.600626043694343e-05,
+      "loss": 0.4241,
       "step": 28720
     },
     {
-      "epoch": 1.01,
-      "learning_rate": 4.6217141738578605e-05,
-      "loss": 0.3086,
+      "epoch": 1.035247053735539,
+      "grad_norm": 0.1602163165807724,
+      "learning_rate": 4.600467808609485e-05,
+      "loss": 0.4105,
       "step": 28725
     },
     {
-      "epoch": 1.01,
-      "learning_rate": 4.6215634919070706e-05,
-      "loss": 0.2838,
+      "epoch": 1.0354272533967637,
+      "grad_norm": 0.1791827380657196,
+      "learning_rate": 4.60030954490616e-05,
+      "loss": 0.4406,
       "step": 28730
     },
     {
-      "epoch": 1.01,
-      "learning_rate": 4.621412782409176e-05,
-      "loss": 0.2952,
+      "epoch": 1.0356074530579882,
+      "grad_norm": 0.17273133993148804,
+      "learning_rate": 4.600151252586524e-05,
+      "loss": 0.4023,
       "step": 28735
     },
     {
-      "epoch": 1.01,
-      "learning_rate": 4.621262045366134e-05,
-      "loss": 0.2666,
+      "epoch": 1.035787652719213,
+      "grad_norm": 0.16618365049362183,
+      "learning_rate": 4.5999929316527335e-05,
+      "loss": 0.4238,
       "step": 28740
     },
     {
-      "epoch": 1.01,
-      "learning_rate": 4.621111280779903e-05,
-      "loss": 0.2786,
+      "epoch": 1.0359678523804374,
+      "grad_norm": 0.147443950176239,
+      "learning_rate": 4.599834582106946e-05,
+      "loss": 0.4491,
       "step": 28745
     },
     {
-      "epoch": 1.01,
-      "learning_rate": 4.62096048865244e-05,
-      "loss": 0.284,
+      "epoch": 1.0361480520416622,
+      "grad_norm": 0.1795797199010849,
+      "learning_rate": 4.599676203951319e-05,
+      "loss": 0.4178,
       "step": 28750
     },
     {
-      "epoch": 1.01,
-      "learning_rate": 4.620809668985703e-05,
-      "loss": 0.2934,
+      "epoch": 1.036328251702887,
+      "grad_norm": 0.15541589260101318,
+      "learning_rate": 4.59951779718801e-05,
+      "loss": 0.4204,
       "step": 28755
     },
     {
-      "epoch": 1.01,
-      "learning_rate": 4.6206588217816494e-05,
-      "loss": 0.3039,
+      "epoch": 1.0365084513641114,
+      "grad_norm": 0.17285758256912231,
+      "learning_rate": 4.599359361819178e-05,
+      "loss": 0.3788,
       "step": 28760
     },
     {
-      "epoch": 1.01,
-      "learning_rate": 4.6205079470422385e-05,
-      "loss": 0.2786,
+      "epoch": 1.0366886510253361,
+      "grad_norm": 0.13503800332546234,
+      "learning_rate": 4.5992008978469806e-05,
+      "loss": 0.3833,
       "step": 28765
     },
     {
-      "epoch": 1.01,
-      "learning_rate": 4.62035704476943e-05,
-      "loss": 0.2848,
+      "epoch": 1.0368688506865607,
+      "grad_norm": 0.19433645904064178,
+      "learning_rate": 4.599042405273578e-05,
+      "loss": 0.4648,
       "step": 28770
     },
     {
-      "epoch": 1.01,
-      "learning_rate": 4.620206114965182e-05,
-      "loss": 0.2921,
+      "epoch": 1.0370490503477854,
+      "grad_norm": 0.18631264567375183,
+      "learning_rate": 4.598883884101128e-05,
+      "loss": 0.4023,
       "step": 28775
     },
     {
-      "epoch": 1.01,
-      "learning_rate": 4.6200551576314555e-05,
-      "loss": 0.2966,
+      "epoch": 1.03722925000901,
+      "grad_norm": 0.1467059701681137,
+      "learning_rate": 4.598725334331793e-05,
+      "loss": 0.4018,
       "step": 28780
     },
     {
-      "epoch": 1.01,
-      "learning_rate": 4.61990417277021e-05,
-      "loss": 0.2852,
+      "epoch": 1.0374094496702346,
+      "grad_norm": 0.160639688372612,
+      "learning_rate": 4.5985667559677303e-05,
+      "loss": 0.3925,
       "step": 28785
     },
     {
-      "epoch": 1.01,
-      "learning_rate": 4.619753160383405e-05,
-      "loss": 0.2852,
+      "epoch": 1.0375896493314594,
+      "grad_norm": 0.1615612506866455,
+      "learning_rate": 4.598408149011102e-05,
+      "loss": 0.4166,
       "step": 28790
     },
     {
-      "epoch": 1.01,
-      "learning_rate": 4.619602120473003e-05,
-      "loss": 0.2634,
+      "epoch": 1.0377698489926839,
+      "grad_norm": 0.24002741277217865,
+      "learning_rate": 4.598249513464069e-05,
+      "loss": 0.3697,
       "step": 28795
     },
     {
-      "epoch": 1.01,
-      "learning_rate": 4.619451053040964e-05,
-      "loss": 0.2672,
+      "epoch": 1.0379500486539086,
+      "grad_norm": 0.18659855425357819,
+      "learning_rate": 4.5980908493287933e-05,
+      "loss": 0.4156,
       "step": 28800
     },
     {
-      "epoch": 1.01,
-      "learning_rate": 4.6192999580892505e-05,
-      "loss": 0.2912,
+      "epoch": 1.0381302483151331,
+      "grad_norm": 0.18351595103740692,
+      "learning_rate": 4.5979321566074354e-05,
+      "loss": 0.4355,
       "step": 28805
     },
     {
-      "epoch": 1.01,
-      "learning_rate": 4.619148835619823e-05,
-      "loss": 0.2948,
+      "epoch": 1.0383104479763579,
+      "grad_norm": 0.20558714866638184,
+      "learning_rate": 4.5977734353021586e-05,
+      "loss": 0.4162,
       "step": 28810
     },
     {
-      "epoch": 1.01,
-      "learning_rate": 4.6189976856346445e-05,
-      "loss": 0.3218,
+      "epoch": 1.0384906476375824,
+      "grad_norm": 0.15841884911060333,
+      "learning_rate": 4.5976146854151244e-05,
+      "loss": 0.3912,
       "step": 28815
     },
     {
-      "epoch": 1.01,
-      "learning_rate": 4.618846508135678e-05,
-      "loss": 0.2818,
+      "epoch": 1.038670847298807,
+      "grad_norm": 0.17524218559265137,
+      "learning_rate": 4.597455906948496e-05,
+      "loss": 0.419,
       "step": 28820
     },
     {
-      "epoch": 1.01,
-      "learning_rate": 4.6186953031248856e-05,
-      "loss": 0.2722,
+      "epoch": 1.0388510469600316,
+      "grad_norm": 0.1817321926355362,
+      "learning_rate": 4.597297099904437e-05,
+      "loss": 0.4137,
       "step": 28825
     },
     {
-      "epoch": 1.01,
-      "learning_rate": 4.6185440706042303e-05,
-      "loss": 0.2779,
+      "epoch": 1.0390312466212563,
+      "grad_norm": 0.18538501858711243,
+      "learning_rate": 4.597138264285111e-05,
+      "loss": 0.3848,
       "step": 28830
     },
     {
-      "epoch": 1.01,
-      "learning_rate": 4.618392810575677e-05,
-      "loss": 0.3148,
+      "epoch": 1.039211446282481,
+      "grad_norm": 0.16245108842849731,
+      "learning_rate": 4.596979400092683e-05,
+      "loss": 0.4461,
       "step": 28835
     },
     {
-      "epoch": 1.01,
-      "learning_rate": 4.6182415230411896e-05,
-      "loss": 0.2818,
+      "epoch": 1.0393916459437056,
+      "grad_norm": 0.2564604878425598,
+      "learning_rate": 4.5968205073293156e-05,
+      "loss": 0.4233,
       "step": 28840
     },
     {
-      "epoch": 1.01,
-      "learning_rate": 4.618090208002731e-05,
-      "loss": 0.3025,
+      "epoch": 1.0395718456049303,
+      "grad_norm": 0.1610010415315628,
+      "learning_rate": 4.596661585997176e-05,
+      "loss": 0.4268,
       "step": 28845
     },
     {
-      "epoch": 1.02,
-      "learning_rate": 4.617938865462268e-05,
-      "loss": 0.279,
+      "epoch": 1.0397520452661548,
+      "grad_norm": 0.19828878343105316,
+      "learning_rate": 4.596502636098427e-05,
+      "loss": 0.43,
       "step": 28850
     },
     {
-      "epoch": 1.02,
-      "learning_rate": 4.617787495421764e-05,
-      "loss": 0.2818,
+      "epoch": 1.0399322449273796,
+      "grad_norm": 0.17699094116687775,
+      "learning_rate": 4.596343657635236e-05,
+      "loss": 0.4678,
       "step": 28855
     },
     {
-      "epoch": 1.02,
-      "learning_rate": 4.6176360978831854e-05,
-      "loss": 0.2788,
+      "epoch": 1.040112444588604,
+      "grad_norm": 0.1853751391172409,
+      "learning_rate": 4.596184650609768e-05,
+      "loss": 0.4363,
       "step": 28860
     },
     {
-      "epoch": 1.02,
-      "learning_rate": 4.617484672848498e-05,
-      "loss": 0.2834,
+      "epoch": 1.0402926442498288,
+      "grad_norm": 0.17727990448474884,
+      "learning_rate": 4.596025615024191e-05,
+      "loss": 0.4541,
       "step": 28865
     },
     {
-      "epoch": 1.02,
-      "learning_rate": 4.617333220319666e-05,
-      "loss": 0.3027,
+      "epoch": 1.0404728439110535,
+      "grad_norm": 0.16852405667304993,
+      "learning_rate": 4.5958665508806696e-05,
+      "loss": 0.3916,
       "step": 28870
     },
     {
-      "epoch": 1.02,
-      "learning_rate": 4.617181740298659e-05,
-      "loss": 0.2951,
+      "epoch": 1.040653043572278,
+      "grad_norm": 0.1764410436153412,
+      "learning_rate": 4.595707458181373e-05,
+      "loss": 0.4453,
       "step": 28875
     },
     {
-      "epoch": 1.02,
-      "learning_rate": 4.617030232787441e-05,
-      "loss": 0.3074,
+      "epoch": 1.0408332432335028,
+      "grad_norm": 0.19933126866817474,
+      "learning_rate": 4.595548336928468e-05,
+      "loss": 0.4561,
       "step": 28880
     },
     {
-      "epoch": 1.02,
-      "learning_rate": 4.616878697787982e-05,
-      "loss": 0.2972,
+      "epoch": 1.0410134428947273,
+      "grad_norm": 0.2222566306591034,
+      "learning_rate": 4.5953891871241226e-05,
+      "loss": 0.418,
       "step": 28885
     },
     {
-      "epoch": 1.02,
-      "learning_rate": 4.616727135302248e-05,
-      "loss": 0.3025,
+      "epoch": 1.041193642555952,
+      "grad_norm": 0.16706469655036926,
+      "learning_rate": 4.5952300087705055e-05,
+      "loss": 0.453,
       "step": 28890
     },
     {
-      "epoch": 1.02,
-      "learning_rate": 4.616575545332207e-05,
-      "loss": 0.2992,
+      "epoch": 1.0413738422171765,
+      "grad_norm": 0.15461257100105286,
+      "learning_rate": 4.595070801869784e-05,
+      "loss": 0.4037,
       "step": 28895
     },
     {
-      "epoch": 1.02,
-      "learning_rate": 4.616423927879827e-05,
-      "loss": 0.2954,
+      "epoch": 1.0415540418784013,
+      "grad_norm": 0.16241170465946198,
+      "learning_rate": 4.594911566424129e-05,
+      "loss": 0.4157,
       "step": 28900
     },
     {
-      "epoch": 1.02,
-      "learning_rate": 4.6162722829470774e-05,
-      "loss": 0.2932,
+      "epoch": 1.041734241539626,
+      "grad_norm": 0.18162140250205994,
+      "learning_rate": 4.5947523024357106e-05,
+      "loss": 0.3942,
       "step": 28905
     },
     {
-      "epoch": 1.02,
-      "learning_rate": 4.616120610535927e-05,
-      "loss": 0.3215,
+      "epoch": 1.0419144412008505,
+      "grad_norm": 0.21681730449199677,
+      "learning_rate": 4.5945930099066966e-05,
+      "loss": 0.4247,
       "step": 28910
     },
     {
-      "epoch": 1.02,
-      "learning_rate": 4.615968910648345e-05,
-      "loss": 0.3091,
+      "epoch": 1.0420946408620753,
+      "grad_norm": 0.16923941671848297,
+      "learning_rate": 4.594433688839258e-05,
+      "loss": 0.4093,
       "step": 28915
     },
     {
-      "epoch": 1.02,
-      "learning_rate": 4.615817183286301e-05,
-      "loss": 0.3195,
+      "epoch": 1.0422748405232998,
+      "grad_norm": 0.17253339290618896,
+      "learning_rate": 4.594274339235567e-05,
+      "loss": 0.4216,
       "step": 28920
     },
     {
-      "epoch": 1.02,
-      "learning_rate": 4.615665428451765e-05,
-      "loss": 0.2676,
+      "epoch": 1.0424550401845245,
+      "grad_norm": 0.17402751743793488,
+      "learning_rate": 4.594114961097793e-05,
+      "loss": 0.421,
       "step": 28925
     },
     {
-      "epoch": 1.02,
-      "learning_rate": 4.615513646146707e-05,
-      "loss": 0.2835,
+      "epoch": 1.042635239845749,
+      "grad_norm": 0.17757736146450043,
+      "learning_rate": 4.593955554428108e-05,
+      "loss": 0.416,
       "step": 28930
     },
     {
-      "epoch": 1.02,
-      "learning_rate": 4.6153618363731e-05,
-      "loss": 0.2917,
+      "epoch": 1.0428154395069738,
+      "grad_norm": 0.19702796638011932,
+      "learning_rate": 4.593796119228684e-05,
+      "loss": 0.4186,
       "step": 28935
     },
     {
-      "epoch": 1.02,
-      "learning_rate": 4.6152099991329135e-05,
-      "loss": 0.3195,
+      "epoch": 1.0429956391681983,
+      "grad_norm": 0.17790283262729645,
+      "learning_rate": 4.593636655501693e-05,
+      "loss": 0.4218,
       "step": 28940
     },
     {
-      "epoch": 1.02,
-      "learning_rate": 4.615058134428118e-05,
-      "loss": 0.2848,
+      "epoch": 1.043175838829423,
+      "grad_norm": 0.18331415951251984,
+      "learning_rate": 4.593477163249308e-05,
+      "loss": 0.4208,
       "step": 28945
     },
     {
-      "epoch": 1.02,
-      "learning_rate": 4.614906242260687e-05,
-      "loss": 0.3069,
+      "epoch": 1.0433560384906477,
+      "grad_norm": 0.19453075528144836,
+      "learning_rate": 4.5933176424737025e-05,
+      "loss": 0.4193,
       "step": 28950
     },
     {
-      "epoch": 1.02,
-      "learning_rate": 4.614754322632593e-05,
-      "loss": 0.2777,
+      "epoch": 1.0435362381518722,
+      "grad_norm": 0.2180250734090805,
+      "learning_rate": 4.5931580931770494e-05,
+      "loss": 0.406,
       "step": 28955
     },
     {
-      "epoch": 1.02,
-      "learning_rate": 4.6146023755458065e-05,
-      "loss": 0.3258,
+      "epoch": 1.043716437813097,
+      "grad_norm": 0.16775591671466827,
+      "learning_rate": 4.592998515361522e-05,
+      "loss": 0.4246,
       "step": 28960
     },
     {
-      "epoch": 1.02,
-      "learning_rate": 4.614450401002303e-05,
-      "loss": 0.2998,
+      "epoch": 1.0438966374743215,
+      "grad_norm": 0.1662595570087433,
+      "learning_rate": 4.5928389090292955e-05,
+      "loss": 0.4263,
       "step": 28965
     },
     {
-      "epoch": 1.02,
-      "learning_rate": 4.6143288015999886e-05,
-      "loss": 0.2975,
+      "epoch": 1.0440768371355462,
+      "grad_norm": 0.17564332485198975,
+      "learning_rate": 4.592679274182544e-05,
+      "loss": 0.4001,
       "step": 28970
     },
     {
-      "epoch": 1.02,
-      "learning_rate": 4.614176777639365e-05,
-      "loss": 0.2925,
+      "epoch": 1.0442570367967707,
+      "grad_norm": 0.18523907661437988,
+      "learning_rate": 4.592519610823442e-05,
+      "loss": 0.4231,
       "step": 28975
     },
     {
-      "epoch": 1.02,
-      "learning_rate": 4.614024726227548e-05,
-      "loss": 0.3256,
+      "epoch": 1.0444372364579955,
+      "grad_norm": 0.17822639644145966,
+      "learning_rate": 4.592359918954165e-05,
+      "loss": 0.4239,
       "step": 28980
     },
     {
-      "epoch": 1.02,
-      "learning_rate": 4.6138726473665144e-05,
-      "loss": 0.296,
+      "epoch": 1.0446174361192202,
+      "grad_norm": 0.19693171977996826,
+      "learning_rate": 4.59220019857689e-05,
+      "loss": 0.4082,
       "step": 28985
     },
     {
-      "epoch": 1.02,
-      "learning_rate": 4.613720541058237e-05,
-      "loss": 0.2818,
+      "epoch": 1.0447976357804447,
+      "grad_norm": 0.16676361858844757,
+      "learning_rate": 4.592040449693793e-05,
+      "loss": 0.4109,
       "step": 28990
     },
     {
-      "epoch": 1.02,
-      "learning_rate": 4.613568407304692e-05,
-      "loss": 0.2902,
+      "epoch": 1.0449778354416694,
+      "grad_norm": 0.16993066668510437,
+      "learning_rate": 4.59188067230705e-05,
+      "loss": 0.4093,
       "step": 28995
     },
     {
-      "epoch": 1.02,
-      "learning_rate": 4.6134162461078535e-05,
-      "loss": 0.3066,
+      "epoch": 1.045158035102894,
+      "grad_norm": 0.20470210909843445,
+      "learning_rate": 4.591720866418836e-05,
+      "loss": 0.4041,
       "step": 29000
     },
     {
-      "epoch": 1.02,
-      "eval_loss": 0.2924649715423584,
-      "eval_runtime": 10.5475,
-      "eval_samples_per_second": 9.481,
-      "eval_steps_per_second": 9.481,
+      "epoch": 1.045158035102894,
+      "eval_loss": 0.4535163640975952,
+      "eval_runtime": 3.5354,
+      "eval_samples_per_second": 28.285,
+      "eval_steps_per_second": 7.071,
       "step": 29000
     },
     {
-      "epoch": 1.02,
-      "learning_rate": 4.613264057469697e-05,
-      "loss": 0.3196,
+      "epoch": 1.0453382347641187,
+      "grad_norm": 0.1358274519443512,
+      "learning_rate": 4.591561032031332e-05,
+      "loss": 0.366,
       "step": 29005
     },
     {
-      "epoch": 1.02,
-      "learning_rate": 4.6131118413922015e-05,
-      "loss": 0.2835,
+      "epoch": 1.0455184344253432,
+      "grad_norm": 0.1652381271123886,
+      "learning_rate": 4.591401169146713e-05,
+      "loss": 0.4371,
       "step": 29010
     },
     {
-      "epoch": 1.02,
-      "learning_rate": 4.61295959787734e-05,
-      "loss": 0.2965,
+      "epoch": 1.045698634086568,
+      "grad_norm": 0.2014123499393463,
+      "learning_rate": 4.591241277767158e-05,
+      "loss": 0.4396,
       "step": 29015
     },
     {
-      "epoch": 1.02,
-      "learning_rate": 4.61280732692709e-05,
-      "loss": 0.2943,
+      "epoch": 1.0458788337477924,
+      "grad_norm": 0.23837004601955414,
+      "learning_rate": 4.591081357894845e-05,
+      "loss": 0.4038,
       "step": 29020
     },
     {
-      "epoch": 1.02,
-      "learning_rate": 4.61265502854343e-05,
-      "loss": 0.2936,
+      "epoch": 1.0460590334090172,
+      "grad_norm": 0.167801171541214,
+      "learning_rate": 4.590921409531954e-05,
+      "loss": 0.4502,
       "step": 29025
     },
     {
-      "epoch": 1.02,
-      "learning_rate": 4.6125027027283375e-05,
-      "loss": 0.2984,
+      "epoch": 1.046239233070242,
+      "grad_norm": 0.14784109592437744,
+      "learning_rate": 4.5907614326806635e-05,
+      "loss": 0.4006,
       "step": 29030
     },
     {
-      "epoch": 1.02,
-      "learning_rate": 4.612350349483788e-05,
-      "loss": 0.2713,
+      "epoch": 1.0464194327314664,
+      "grad_norm": 0.17408788204193115,
+      "learning_rate": 4.5906014273431535e-05,
+      "loss": 0.3958,
       "step": 29035
     },
     {
-      "epoch": 1.02,
-      "learning_rate": 4.6121979688117626e-05,
-      "loss": 0.293,
+      "epoch": 1.0465996323926912,
+      "grad_norm": 0.1565713882446289,
+      "learning_rate": 4.590441393521604e-05,
+      "loss": 0.4062,
       "step": 29040
     },
     {
-      "epoch": 1.02,
-      "learning_rate": 4.6120455607142385e-05,
-      "loss": 0.2994,
+      "epoch": 1.0467798320539157,
+      "grad_norm": 0.20593391358852386,
+      "learning_rate": 4.590281331218195e-05,
+      "loss": 0.4573,
       "step": 29045
     },
     {
-      "epoch": 1.02,
-      "learning_rate": 4.611893125193194e-05,
-      "loss": 0.29,
+      "epoch": 1.0469600317151404,
+      "grad_norm": 0.1602008193731308,
+      "learning_rate": 4.5901212404351065e-05,
+      "loss": 0.4379,
       "step": 29050
     },
     {
-      "epoch": 1.02,
-      "learning_rate": 4.61174066225061e-05,
-      "loss": 0.2924,
+      "epoch": 1.047140231376365,
+      "grad_norm": 0.19146746397018433,
+      "learning_rate": 4.589961121174522e-05,
+      "loss": 0.3882,
       "step": 29055
     },
     {
-      "epoch": 1.02,
-      "learning_rate": 4.611588171888464e-05,
-      "loss": 0.3052,
+      "epoch": 1.0473204310375896,
+      "grad_norm": 0.18531233072280884,
+      "learning_rate": 4.5898009734386216e-05,
+      "loss": 0.4533,
       "step": 29060
     },
     {
-      "epoch": 1.02,
-      "learning_rate": 4.611435654108738e-05,
-      "loss": 0.2863,
+      "epoch": 1.0475006306988144,
+      "grad_norm": 0.15970437228679657,
+      "learning_rate": 4.5896407972295875e-05,
+      "loss": 0.3773,
       "step": 29065
     },
     {
-      "epoch": 1.02,
-      "learning_rate": 4.611283108913411e-05,
-      "loss": 0.2687,
+      "epoch": 1.047680830360039,
+      "grad_norm": 0.17048677802085876,
+      "learning_rate": 4.589480592549602e-05,
+      "loss": 0.4305,
       "step": 29070
     },
     {
-      "epoch": 1.02,
-      "learning_rate": 4.6111305363044644e-05,
-      "loss": 0.2993,
+      "epoch": 1.0478610300212636,
+      "grad_norm": 0.1856071949005127,
+      "learning_rate": 4.5893203594008476e-05,
+      "loss": 0.4541,
       "step": 29075
     },
     {
-      "epoch": 1.02,
-      "learning_rate": 4.6109779362838794e-05,
-      "loss": 0.293,
+      "epoch": 1.0480412296824881,
+      "grad_norm": 0.19591088593006134,
+      "learning_rate": 4.5891600977855085e-05,
+      "loss": 0.3656,
       "step": 29080
     },
     {
-      "epoch": 1.02,
-      "learning_rate": 4.610825308853637e-05,
-      "loss": 0.2927,
+      "epoch": 1.0482214293437129,
+      "grad_norm": 0.15485340356826782,
+      "learning_rate": 4.5889998077057674e-05,
+      "loss": 0.4005,
       "step": 29085
     },
     {
-      "epoch": 1.02,
-      "learning_rate": 4.610672654015719e-05,
-      "loss": 0.3053,
+      "epoch": 1.0484016290049374,
+      "grad_norm": 0.17994803190231323,
+      "learning_rate": 4.588839489163808e-05,
+      "loss": 0.406,
       "step": 29090
     },
     {
-      "epoch": 1.02,
-      "learning_rate": 4.610519971772107e-05,
-      "loss": 0.3047,
+      "epoch": 1.0485818286661621,
+      "grad_norm": 0.1988825649023056,
+      "learning_rate": 4.5886791421618155e-05,
+      "loss": 0.402,
       "step": 29095
     },
     {
-      "epoch": 1.02,
-      "learning_rate": 4.610367262124785e-05,
-      "loss": 0.3033,
+      "epoch": 1.0487620283273866,
+      "grad_norm": 0.18159984052181244,
+      "learning_rate": 4.5885187667019733e-05,
+      "loss": 0.4012,
       "step": 29100
     },
     {
-      "epoch": 1.02,
-      "learning_rate": 4.6102145250757345e-05,
-      "loss": 0.3131,
+      "epoch": 1.0489422279886114,
+      "grad_norm": 0.17193853855133057,
+      "learning_rate": 4.588358362786468e-05,
+      "loss": 0.4378,
       "step": 29105
     },
     {
-      "epoch": 1.02,
-      "learning_rate": 4.6100617606269385e-05,
-      "loss": 0.3028,
+      "epoch": 1.049122427649836,
+      "grad_norm": 0.1589495688676834,
+      "learning_rate": 4.588197930417484e-05,
+      "loss": 0.4416,
       "step": 29110
     },
     {
-      "epoch": 1.02,
-      "learning_rate": 4.609908968780382e-05,
-      "loss": 0.2893,
+      "epoch": 1.0493026273110606,
+      "grad_norm": 0.1768224835395813,
+      "learning_rate": 4.588037469597207e-05,
+      "loss": 0.4335,
       "step": 29115
     },
     {
-      "epoch": 1.02,
-      "learning_rate": 4.6097561495380474e-05,
-      "loss": 0.275,
+      "epoch": 1.0494828269722853,
+      "grad_norm": 0.207745760679245,
+      "learning_rate": 4.587876980327824e-05,
+      "loss": 0.4003,
       "step": 29120
     },
     {
-      "epoch": 1.02,
-      "learning_rate": 4.60960330290192e-05,
-      "loss": 0.2819,
+      "epoch": 1.0496630266335099,
+      "grad_norm": 0.20602521300315857,
+      "learning_rate": 4.587716462611522e-05,
+      "loss": 0.453,
       "step": 29125
     },
     {
-      "epoch": 1.02,
-      "learning_rate": 4.609450428873984e-05,
-      "loss": 0.2878,
+      "epoch": 1.0498432262947346,
+      "grad_norm": 0.1503795087337494,
+      "learning_rate": 4.587555916450487e-05,
+      "loss": 0.3976,
       "step": 29130
     },
     {
-      "epoch": 1.03,
-      "learning_rate": 4.609297527456224e-05,
-      "loss": 0.3103,
+      "epoch": 1.050023425955959,
+      "grad_norm": 0.21797312796115875,
+      "learning_rate": 4.587395341846907e-05,
+      "loss": 0.4621,
       "step": 29135
     },
     {
-      "epoch": 1.03,
-      "learning_rate": 4.609144598650625e-05,
-      "loss": 0.2843,
+      "epoch": 1.0502036256171838,
+      "grad_norm": 0.17475320398807526,
+      "learning_rate": 4.587234738802969e-05,
+      "loss": 0.4362,
       "step": 29140
     },
     {
-      "epoch": 1.03,
-      "learning_rate": 4.608991642459175e-05,
-      "loss": 0.3022,
+      "epoch": 1.0503838252784086,
+      "grad_norm": 0.15241824090480804,
+      "learning_rate": 4.5870741073208624e-05,
+      "loss": 0.4161,
       "step": 29145
     },
     {
-      "epoch": 1.03,
-      "learning_rate": 4.6088386588838574e-05,
-      "loss": 0.3251,
+      "epoch": 1.050564024939633,
+      "grad_norm": 0.17844238877296448,
+      "learning_rate": 4.5869134474027756e-05,
+      "loss": 0.4955,
       "step": 29150
     },
     {
-      "epoch": 1.03,
-      "learning_rate": 4.60868564792666e-05,
-      "loss": 0.3163,
+      "epoch": 1.0507442246008578,
+      "grad_norm": 0.1921539604663849,
+      "learning_rate": 4.586752759050896e-05,
+      "loss": 0.4587,
       "step": 29155
     },
     {
-      "epoch": 1.03,
-      "learning_rate": 4.6085326095895685e-05,
-      "loss": 0.2952,
+      "epoch": 1.0509244242620823,
+      "grad_norm": 0.20518042147159576,
+      "learning_rate": 4.586592042267415e-05,
+      "loss": 0.4486,
       "step": 29160
     },
     {
-      "epoch": 1.03,
-      "learning_rate": 4.608379543874571e-05,
-      "loss": 0.3182,
+      "epoch": 1.051104623923307,
+      "grad_norm": 0.15805669128894806,
+      "learning_rate": 4.586431297054521e-05,
+      "loss": 0.4031,
       "step": 29165
     },
     {
-      "epoch": 1.03,
-      "learning_rate": 4.608226450783655e-05,
-      "loss": 0.2704,
+      "epoch": 1.0512848235845316,
+      "grad_norm": 0.20354914665222168,
+      "learning_rate": 4.586270523414404e-05,
+      "loss": 0.4089,
       "step": 29170
     },
     {
-      "epoch": 1.03,
-      "learning_rate": 4.6080733303188075e-05,
-      "loss": 0.3,
+      "epoch": 1.0514650232457563,
+      "grad_norm": 0.18959026038646698,
+      "learning_rate": 4.586109721349255e-05,
+      "loss": 0.4437,
       "step": 29175
     },
     {
-      "epoch": 1.03,
-      "learning_rate": 4.607920182482017e-05,
-      "loss": 0.2958,
+      "epoch": 1.051645222906981,
+      "grad_norm": 0.2117196023464203,
+      "learning_rate": 4.585948890861266e-05,
+      "loss": 0.4373,
       "step": 29180
     },
     {
-      "epoch": 1.03,
-      "learning_rate": 4.607767007275273e-05,
-      "loss": 0.2918,
+      "epoch": 1.0518254225682055,
+      "grad_norm": 0.18057450652122498,
+      "learning_rate": 4.585788031952627e-05,
+      "loss": 0.4321,
       "step": 29185
     },
     {
-      "epoch": 1.03,
-      "learning_rate": 4.607613804700562e-05,
-      "loss": 0.307,
+      "epoch": 1.0520056222294303,
+      "grad_norm": 0.1702229529619217,
+      "learning_rate": 4.5856271446255294e-05,
+      "loss": 0.3866,
       "step": 29190
     },
     {
-      "epoch": 1.03,
-      "learning_rate": 4.607460574759876e-05,
-      "loss": 0.3087,
+      "epoch": 1.0521858218906548,
+      "grad_norm": 0.17912250757217407,
+      "learning_rate": 4.5854662288821656e-05,
+      "loss": 0.3849,
       "step": 29195
     },
     {
-      "epoch": 1.03,
-      "learning_rate": 4.607307317455203e-05,
-      "loss": 0.2707,
+      "epoch": 1.0523660215518795,
+      "grad_norm": 0.19952471554279327,
+      "learning_rate": 4.5853052847247286e-05,
+      "loss": 0.4252,
       "step": 29200
     },
     {
-      "epoch": 1.03,
-      "learning_rate": 4.607154032788533e-05,
-      "loss": 0.3145,
+      "epoch": 1.052546221213104,
+      "grad_norm": 0.17590820789337158,
+      "learning_rate": 4.5851443121554104e-05,
+      "loss": 0.4243,
       "step": 29205
     },
     {
-      "epoch": 1.03,
-      "learning_rate": 4.607000720761856e-05,
-      "loss": 0.29,
+      "epoch": 1.0527264208743288,
+      "grad_norm": 0.18067772686481476,
+      "learning_rate": 4.584983311176405e-05,
+      "loss": 0.4142,
       "step": 29210
     },
     {
-      "epoch": 1.03,
-      "learning_rate": 4.606847381377164e-05,
-      "loss": 0.2782,
+      "epoch": 1.0529066205355533,
+      "grad_norm": 0.17660415172576904,
+      "learning_rate": 4.584822281789906e-05,
+      "loss": 0.4109,
       "step": 29215
     },
     {
-      "epoch": 1.03,
-      "learning_rate": 4.606694014636447e-05,
-      "loss": 0.3053,
+      "epoch": 1.053086820196778,
+      "grad_norm": 0.20561589300632477,
+      "learning_rate": 4.5846612239981065e-05,
+      "loss": 0.4339,
       "step": 29220
     },
     {
-      "epoch": 1.03,
-      "learning_rate": 4.6065406205416966e-05,
-      "loss": 0.2973,
+      "epoch": 1.0532670198580028,
+      "grad_norm": 0.14123573899269104,
+      "learning_rate": 4.584500137803201e-05,
+      "loss": 0.4289,
       "step": 29225
     },
     {
-      "epoch": 1.03,
-      "learning_rate": 4.606387199094905e-05,
-      "loss": 0.2895,
+      "epoch": 1.0534472195192273,
+      "grad_norm": 0.17180036008358002,
+      "learning_rate": 4.5843390232073856e-05,
+      "loss": 0.4204,
       "step": 29230
     },
     {
-      "epoch": 1.03,
-      "learning_rate": 4.606233750298063e-05,
-      "loss": 0.2861,
+      "epoch": 1.053627419180452,
+      "grad_norm": 0.18303899466991425,
+      "learning_rate": 4.5841778802128544e-05,
+      "loss": 0.425,
       "step": 29235
     },
     {
-      "epoch": 1.03,
-      "learning_rate": 4.6060802741531636e-05,
-      "loss": 0.2772,
+      "epoch": 1.0538076188416765,
+      "grad_norm": 0.15086086094379425,
+      "learning_rate": 4.5840167088218026e-05,
+      "loss": 0.3899,
       "step": 29240
     },
     {
-      "epoch": 1.03,
-      "learning_rate": 4.605926770662201e-05,
-      "loss": 0.3141,
+      "epoch": 1.0539878185029012,
+      "grad_norm": 0.15532901883125305,
+      "learning_rate": 4.583855509036427e-05,
+      "loss": 0.4025,
       "step": 29245
     },
     {
-      "epoch": 1.03,
-      "learning_rate": 4.605773239827166e-05,
-      "loss": 0.2708,
+      "epoch": 1.0541680181641258,
+      "grad_norm": 0.15308701992034912,
+      "learning_rate": 4.5836942808589235e-05,
+      "loss": 0.4168,
       "step": 29250
     },
     {
-      "epoch": 1.03,
-      "learning_rate": 4.605619681650054e-05,
-      "loss": 0.2624,
+      "epoch": 1.0543482178253505,
+      "grad_norm": 0.1958785504102707,
+      "learning_rate": 4.583533024291489e-05,
+      "loss": 0.4647,
       "step": 29255
     },
     {
-      "epoch": 1.03,
-      "learning_rate": 4.605466096132858e-05,
-      "loss": 0.3042,
+      "epoch": 1.0545284174865752,
+      "grad_norm": 0.15900172293186188,
+      "learning_rate": 4.583371739336319e-05,
+      "loss": 0.4076,
       "step": 29260
     },
     {
-      "epoch": 1.03,
-      "learning_rate": 4.605312483277571e-05,
-      "loss": 0.3024,
+      "epoch": 1.0547086171477997,
+      "grad_norm": 0.19627580046653748,
+      "learning_rate": 4.583210425995614e-05,
+      "loss": 0.4619,
       "step": 29265
     },
     {
-      "epoch": 1.03,
-      "learning_rate": 4.60515884308619e-05,
-      "loss": 0.3008,
+      "epoch": 1.0548888168090245,
+      "grad_norm": 0.17516663670539856,
+      "learning_rate": 4.58304908427157e-05,
+      "loss": 0.4433,
       "step": 29270
     },
     {
-      "epoch": 1.03,
-      "learning_rate": 4.605005175560709e-05,
-      "loss": 0.3072,
+      "epoch": 1.055069016470249,
+      "grad_norm": 0.1998327523469925,
+      "learning_rate": 4.582887714166386e-05,
+      "loss": 0.4309,
       "step": 29275
     },
     {
-      "epoch": 1.03,
-      "learning_rate": 4.604851480703123e-05,
-      "loss": 0.2906,
+      "epoch": 1.0552492161314737,
+      "grad_norm": 0.15834058821201324,
+      "learning_rate": 4.582726315682259e-05,
+      "loss": 0.4016,
       "step": 29280
     },
     {
-      "epoch": 1.03,
-      "learning_rate": 4.604697758515427e-05,
-      "loss": 0.2937,
+      "epoch": 1.0554294157926982,
+      "grad_norm": 0.16241154074668884,
+      "learning_rate": 4.58256488882139e-05,
+      "loss": 0.4014,
       "step": 29285
     },
     {
-      "epoch": 1.03,
-      "learning_rate": 4.604544008999618e-05,
-      "loss": 0.2759,
+      "epoch": 1.055609615453923,
+      "grad_norm": 0.18402625620365143,
+      "learning_rate": 4.582403433585978e-05,
+      "loss": 0.4168,
       "step": 29290
     },
     {
-      "epoch": 1.03,
-      "learning_rate": 4.6043902321576926e-05,
-      "loss": 0.3239,
+      "epoch": 1.0557898151151477,
+      "grad_norm": 0.17826689779758453,
+      "learning_rate": 4.582241949978221e-05,
+      "loss": 0.4416,
       "step": 29295
     },
     {
-      "epoch": 1.03,
-      "learning_rate": 4.6042364279916464e-05,
-      "loss": 0.2735,
+      "epoch": 1.0559700147763722,
+      "grad_norm": 0.18548059463500977,
+      "learning_rate": 4.582080438000321e-05,
+      "loss": 0.422,
       "step": 29300
     },
     {
-      "epoch": 1.03,
-      "learning_rate": 4.6040825965034764e-05,
-      "loss": 0.3082,
+      "epoch": 1.056150214437597,
+      "grad_norm": 0.1519528478384018,
+      "learning_rate": 4.581918897654479e-05,
+      "loss": 0.4051,
       "step": 29305
     },
     {
-      "epoch": 1.03,
-      "learning_rate": 4.603928737695181e-05,
-      "loss": 0.2771,
+      "epoch": 1.0563304140988214,
+      "grad_norm": 0.14724068343639374,
+      "learning_rate": 4.581757328942894e-05,
+      "loss": 0.3999,
       "step": 29310
     },
     {
-      "epoch": 1.03,
-      "learning_rate": 4.6037748515687576e-05,
-      "loss": 0.3013,
+      "epoch": 1.0565106137600462,
+      "grad_norm": 0.18171782791614532,
+      "learning_rate": 4.5815957318677693e-05,
+      "loss": 0.4232,
       "step": 29315
     },
     {
-      "epoch": 1.03,
-      "learning_rate": 4.603620938126204e-05,
-      "loss": 0.2855,
+      "epoch": 1.0566908134212707,
+      "grad_norm": 0.20490474998950958,
+      "learning_rate": 4.5814341064313055e-05,
+      "loss": 0.4333,
       "step": 29320
     },
     {
-      "epoch": 1.03,
-      "learning_rate": 4.603466997369519e-05,
-      "loss": 0.3087,
+      "epoch": 1.0568710130824954,
+      "grad_norm": 0.18439970910549164,
+      "learning_rate": 4.581272452635705e-05,
+      "loss": 0.3899,
       "step": 29325
     },
     {
-      "epoch": 1.03,
-      "learning_rate": 4.603313029300701e-05,
-      "loss": 0.2584,
+      "epoch": 1.05705121274372,
+      "grad_norm": 0.14571373164653778,
+      "learning_rate": 4.58111077048317e-05,
+      "loss": 0.4089,
       "step": 29330
     },
     {
-      "epoch": 1.03,
-      "learning_rate": 4.603159033921749e-05,
-      "loss": 0.294,
+      "epoch": 1.0572314124049447,
+      "grad_norm": 0.15669065713882446,
+      "learning_rate": 4.5809490599759034e-05,
+      "loss": 0.3999,
       "step": 29335
     },
     {
-      "epoch": 1.03,
-      "learning_rate": 4.6030050112346643e-05,
-      "loss": 0.3152,
+      "epoch": 1.0574116120661694,
+      "grad_norm": 0.16960056126117706,
+      "learning_rate": 4.58078732111611e-05,
+      "loss": 0.4173,
       "step": 29340
     },
     {
-      "epoch": 1.03,
-      "learning_rate": 4.602850961241445e-05,
-      "loss": 0.2669,
+      "epoch": 1.057591811727394,
+      "grad_norm": 0.17213228344917297,
+      "learning_rate": 4.5806255539059905e-05,
+      "loss": 0.4012,
       "step": 29345
     },
     {
-      "epoch": 1.03,
-      "learning_rate": 4.602696883944091e-05,
-      "loss": 0.2989,
+      "epoch": 1.0577720113886186,
+      "grad_norm": 0.15806017816066742,
+      "learning_rate": 4.5804637583477514e-05,
+      "loss": 0.4421,
       "step": 29350
     },
     {
-      "epoch": 1.03,
-      "learning_rate": 4.602542779344605e-05,
-      "loss": 0.2704,
+      "epoch": 1.0579522110498432,
+      "grad_norm": 0.19483284652233124,
+      "learning_rate": 4.5803019344435974e-05,
+      "loss": 0.4188,
       "step": 29355
     },
     {
-      "epoch": 1.03,
-      "learning_rate": 4.602388647444985e-05,
-      "loss": 0.2651,
+      "epoch": 1.058132410711068,
+      "grad_norm": 0.18508678674697876,
+      "learning_rate": 4.580140082195731e-05,
+      "loss": 0.4134,
       "step": 29360
     },
     {
-      "epoch": 1.03,
-      "learning_rate": 4.6022344882472354e-05,
-      "loss": 0.2857,
+      "epoch": 1.0583126103722924,
+      "grad_norm": 0.1570422649383545,
+      "learning_rate": 4.579978201606359e-05,
+      "loss": 0.4138,
       "step": 29365
     },
     {
-      "epoch": 1.03,
-      "learning_rate": 4.6020803017533555e-05,
-      "loss": 0.319,
+      "epoch": 1.0584928100335171,
+      "grad_norm": 0.14425136148929596,
+      "learning_rate": 4.5798162926776865e-05,
+      "loss": 0.4375,
       "step": 29370
     },
     {
-      "epoch": 1.03,
-      "learning_rate": 4.6019260879653486e-05,
-      "loss": 0.2883,
+      "epoch": 1.0586730096947419,
+      "grad_norm": 0.17451147735118866,
+      "learning_rate": 4.57965435541192e-05,
+      "loss": 0.4292,
       "step": 29375
     },
     {
-      "epoch": 1.03,
-      "learning_rate": 4.6017718468852165e-05,
-      "loss": 0.2926,
+      "epoch": 1.0588532093559664,
+      "grad_norm": 0.16206900775432587,
+      "learning_rate": 4.579492389811266e-05,
+      "loss": 0.4548,
       "step": 29380
     },
     {
-      "epoch": 1.03,
-      "learning_rate": 4.601617578514962e-05,
-      "loss": 0.2898,
+      "epoch": 1.0590334090171911,
+      "grad_norm": 0.1823441982269287,
+      "learning_rate": 4.57933039587793e-05,
+      "loss": 0.409,
       "step": 29385
     },
     {
-      "epoch": 1.03,
-      "learning_rate": 4.6014632828565887e-05,
-      "loss": 0.323,
+      "epoch": 1.0592136086784156,
+      "grad_norm": 0.18232353031635284,
+      "learning_rate": 4.579168373614121e-05,
+      "loss": 0.3947,
       "step": 29390
     },
     {
-      "epoch": 1.03,
-      "learning_rate": 4.6013089599121e-05,
-      "loss": 0.2977,
+      "epoch": 1.0593938083396404,
+      "grad_norm": 0.16864867508411407,
+      "learning_rate": 4.5790063230220444e-05,
+      "loss": 0.4212,
       "step": 29395
     },
     {
-      "epoch": 1.03,
-      "learning_rate": 4.601154609683498e-05,
-      "loss": 0.3198,
+      "epoch": 1.0595740080008649,
+      "grad_norm": 0.19608835875988007,
+      "learning_rate": 4.578844244103909e-05,
+      "loss": 0.42,
       "step": 29400
     },
     {
-      "epoch": 1.03,
-      "learning_rate": 4.601000232172789e-05,
-      "loss": 0.2892,
+      "epoch": 1.0597542076620896,
+      "grad_norm": 0.1760438084602356,
+      "learning_rate": 4.5786821368619236e-05,
+      "loss": 0.4085,
       "step": 29405
     },
     {
-      "epoch": 1.03,
-      "learning_rate": 4.600845827381976e-05,
-      "loss": 0.2732,
+      "epoch": 1.0599344073233143,
+      "grad_norm": 0.17386280000209808,
+      "learning_rate": 4.578520001298297e-05,
+      "loss": 0.4454,
       "step": 29410
     },
     {
-      "epoch": 1.03,
-      "learning_rate": 4.600691395313065e-05,
-      "loss": 0.3062,
+      "epoch": 1.0601146069845389,
+      "grad_norm": 0.14320212602615356,
+      "learning_rate": 4.5783578374152376e-05,
+      "loss": 0.389,
       "step": 29415
     },
     {
-      "epoch": 1.04,
-      "learning_rate": 4.600536935968061e-05,
-      "loss": 0.2863,
+      "epoch": 1.0602948066457636,
+      "grad_norm": 0.20935241878032684,
+      "learning_rate": 4.578195645214955e-05,
+      "loss": 0.4252,
       "step": 29420
     },
     {
-      "epoch": 1.04,
-      "learning_rate": 4.6003824493489676e-05,
-      "loss": 0.2974,
+      "epoch": 1.060475006306988,
+      "grad_norm": 0.17992691695690155,
+      "learning_rate": 4.578033424699659e-05,
+      "loss": 0.4362,
       "step": 29425
     },
     {
-      "epoch": 1.04,
-      "learning_rate": 4.600227935457794e-05,
-      "loss": 0.2896,
+      "epoch": 1.0606552059682128,
+      "grad_norm": 0.18552446365356445,
+      "learning_rate": 4.57787117587156e-05,
+      "loss": 0.4269,
       "step": 29430
     },
     {
-      "epoch": 1.04,
-      "learning_rate": 4.600073394296544e-05,
-      "loss": 0.3128,
+      "epoch": 1.0608354056294373,
+      "grad_norm": 0.18173004686832428,
+      "learning_rate": 4.577708898732868e-05,
+      "loss": 0.4358,
       "step": 29435
     },
     {
-      "epoch": 1.04,
-      "learning_rate": 4.5999188258672246e-05,
-      "loss": 0.3325,
+      "epoch": 1.061015605290662,
+      "grad_norm": 0.19340266287326813,
+      "learning_rate": 4.577546593285795e-05,
+      "loss": 0.4268,
       "step": 29440
     },
     {
-      "epoch": 1.04,
-      "learning_rate": 4.599764230171844e-05,
-      "loss": 0.296,
+      "epoch": 1.0611958049518866,
+      "grad_norm": 0.17074435949325562,
+      "learning_rate": 4.577384259532552e-05,
+      "loss": 0.4342,
       "step": 29445
     },
     {
-      "epoch": 1.04,
-      "learning_rate": 4.599609607212408e-05,
-      "loss": 0.2799,
+      "epoch": 1.0613760046131113,
+      "grad_norm": 0.17281962931156158,
+      "learning_rate": 4.57722189747535e-05,
+      "loss": 0.3953,
       "step": 29450
     },
     {
-      "epoch": 1.04,
-      "learning_rate": 4.599454956990925e-05,
-      "loss": 0.3323,
+      "epoch": 1.061556204274336,
+      "grad_norm": 0.16818857192993164,
+      "learning_rate": 4.577059507116403e-05,
+      "loss": 0.3934,
       "step": 29455
     },
     {
-      "epoch": 1.04,
-      "learning_rate": 4.599300279509403e-05,
-      "loss": 0.3046,
+      "epoch": 1.0617364039355606,
+      "grad_norm": 0.15899808704853058,
+      "learning_rate": 4.576897088457921e-05,
+      "loss": 0.4223,
       "step": 29460
     },
     {
-      "epoch": 1.04,
-      "learning_rate": 4.5991455747698505e-05,
-      "loss": 0.2976,
+      "epoch": 1.0619166035967853,
+      "grad_norm": 0.17966797947883606,
+      "learning_rate": 4.576734641502119e-05,
+      "loss": 0.4265,
       "step": 29465
     },
     {
-      "epoch": 1.04,
-      "learning_rate": 4.598990842774276e-05,
-      "loss": 0.2796,
+      "epoch": 1.0620968032580098,
+      "grad_norm": 0.198250874876976,
+      "learning_rate": 4.57657216625121e-05,
+      "loss": 0.4491,
       "step": 29470
     },
     {
-      "epoch": 1.04,
-      "learning_rate": 4.598836083524689e-05,
-      "loss": 0.2941,
+      "epoch": 1.0622770029192345,
+      "grad_norm": 0.15183819830417633,
+      "learning_rate": 4.576409662707406e-05,
+      "loss": 0.4185,
       "step": 29475
     },
     {
-      "epoch": 1.04,
-      "learning_rate": 4.598681297023099e-05,
-      "loss": 0.2621,
+      "epoch": 1.062457202580459,
+      "grad_norm": 0.1668313890695572,
+      "learning_rate": 4.576247130872924e-05,
+      "loss": 0.3849,
       "step": 29480
     },
     {
-      "epoch": 1.04,
-      "learning_rate": 4.598526483271514e-05,
-      "loss": 0.313,
+      "epoch": 1.0626374022416838,
+      "grad_norm": 0.15694022178649902,
+      "learning_rate": 4.5760845707499753e-05,
+      "loss": 0.4084,
       "step": 29485
     },
     {
-      "epoch": 1.04,
-      "learning_rate": 4.598371642271947e-05,
-      "loss": 0.3199,
+      "epoch": 1.0628176019029085,
+      "grad_norm": 0.17181478440761566,
+      "learning_rate": 4.575921982340777e-05,
+      "loss": 0.4529,
       "step": 29490
     },
     {
-      "epoch": 1.04,
-      "learning_rate": 4.598216774026406e-05,
-      "loss": 0.3125,
+      "epoch": 1.062997801564133,
+      "grad_norm": 0.18744981288909912,
+      "learning_rate": 4.575759365647543e-05,
+      "loss": 0.4149,
       "step": 29495
     },
     {
-      "epoch": 1.04,
-      "learning_rate": 4.598061878536904e-05,
-      "loss": 0.3023,
+      "epoch": 1.0631780012253578,
+      "grad_norm": 0.1948675960302353,
+      "learning_rate": 4.5755967206724906e-05,
+      "loss": 0.4541,
       "step": 29500
     },
     {
-      "epoch": 1.04,
-      "eval_loss": 0.2920066714286804,
-      "eval_runtime": 10.538,
-      "eval_samples_per_second": 9.489,
-      "eval_steps_per_second": 9.489,
+      "epoch": 1.0631780012253578,
+      "eval_loss": 0.45399028062820435,
+      "eval_runtime": 3.5401,
+      "eval_samples_per_second": 28.248,
+      "eval_steps_per_second": 7.062,
       "step": 29500
     },
     {
-      "epoch": 1.04,
-      "learning_rate": 4.597906955805451e-05,
-      "loss": 0.2885,
+      "epoch": 1.0633582008865823,
+      "grad_norm": 0.2056550681591034,
+      "learning_rate": 4.5754340474178334e-05,
+      "loss": 0.4456,
       "step": 29505
     },
     {
-      "epoch": 1.04,
-      "learning_rate": 4.5977520058340586e-05,
-      "loss": 0.2782,
+      "epoch": 1.063538400547807,
+      "grad_norm": 0.14791448414325714,
+      "learning_rate": 4.57527134588579e-05,
+      "loss": 0.429,
       "step": 29510
     },
     {
-      "epoch": 1.04,
-      "learning_rate": 4.5975970286247384e-05,
-      "loss": 0.2928,
+      "epoch": 1.0637186002090315,
+      "grad_norm": 0.16541562974452972,
+      "learning_rate": 4.5751086160785764e-05,
+      "loss": 0.4213,
       "step": 29515
     },
     {
-      "epoch": 1.04,
-      "learning_rate": 4.5974420241795034e-05,
-      "loss": 0.2862,
+      "epoch": 1.0638987998702563,
+      "grad_norm": 0.18588918447494507,
+      "learning_rate": 4.57494585799841e-05,
+      "loss": 0.4371,
       "step": 29520
     },
     {
-      "epoch": 1.04,
-      "learning_rate": 4.597286992500366e-05,
-      "loss": 0.2886,
+      "epoch": 1.064078999531481,
+      "grad_norm": 0.20339952409267426,
+      "learning_rate": 4.574783071647507e-05,
+      "loss": 0.4157,
       "step": 29525
     },
     {
-      "epoch": 1.04,
-      "learning_rate": 4.597131933589339e-05,
-      "loss": 0.2839,
+      "epoch": 1.0642591991927055,
+      "grad_norm": 0.1556108593940735,
+      "learning_rate": 4.574620257028087e-05,
+      "loss": 0.4185,
       "step": 29530
     },
     {
-      "epoch": 1.04,
-      "learning_rate": 4.5969768474484364e-05,
-      "loss": 0.2895,
+      "epoch": 1.0644393988539302,
+      "grad_norm": 0.2084941267967224,
+      "learning_rate": 4.574457414142367e-05,
+      "loss": 0.4225,
       "step": 29535
     },
     {
-      "epoch": 1.04,
-      "learning_rate": 4.596821734079672e-05,
-      "loss": 0.2762,
+      "epoch": 1.0646195985151548,
+      "grad_norm": 0.1711416095495224,
+      "learning_rate": 4.574294542992566e-05,
+      "loss": 0.4142,
       "step": 29540
     },
     {
-      "epoch": 1.04,
-      "learning_rate": 4.596666593485058e-05,
-      "loss": 0.3067,
+      "epoch": 1.0647997981763795,
+      "grad_norm": 0.17444539070129395,
+      "learning_rate": 4.574131643580905e-05,
+      "loss": 0.4169,
       "step": 29545
     },
     {
-      "epoch": 1.04,
-      "learning_rate": 4.5965114256666104e-05,
-      "loss": 0.2999,
+      "epoch": 1.064979997837604,
+      "grad_norm": 0.2214493304491043,
+      "learning_rate": 4.573968715909601e-05,
+      "loss": 0.4479,
       "step": 29550
     },
     {
-      "epoch": 1.04,
-      "learning_rate": 4.5963562306263435e-05,
-      "loss": 0.2703,
+      "epoch": 1.0651601974988287,
+      "grad_norm": 0.16091366112232208,
+      "learning_rate": 4.573805759980875e-05,
+      "loss": 0.4392,
       "step": 29555
     },
     {
-      "epoch": 1.04,
-      "learning_rate": 4.5962010083662724e-05,
-      "loss": 0.2873,
+      "epoch": 1.0653403971600532,
+      "grad_norm": 0.1976114958524704,
+      "learning_rate": 4.573642775796947e-05,
+      "loss": 0.4436,
       "step": 29560
     },
     {
-      "epoch": 1.04,
-      "learning_rate": 4.596045758888413e-05,
-      "loss": 0.3129,
+      "epoch": 1.065520596821278,
+      "grad_norm": 0.14705707132816315,
+      "learning_rate": 4.573479763360038e-05,
+      "loss": 0.4187,
       "step": 29565
     },
     {
-      "epoch": 1.04,
-      "learning_rate": 4.5958904821947805e-05,
-      "loss": 0.303,
+      "epoch": 1.0657007964825027,
+      "grad_norm": 0.20177073776721954,
+      "learning_rate": 4.573316722672368e-05,
+      "loss": 0.4562,
       "step": 29570
     },
     {
-      "epoch": 1.04,
-      "learning_rate": 4.5957351782873916e-05,
-      "loss": 0.2996,
+      "epoch": 1.0658809961437272,
+      "grad_norm": 0.16343575716018677,
+      "learning_rate": 4.573153653736159e-05,
+      "loss": 0.4348,
       "step": 29575
     },
     {
-      "epoch": 1.04,
-      "learning_rate": 4.595579847168263e-05,
-      "loss": 0.28,
+      "epoch": 1.066061195804952,
+      "grad_norm": 0.19736631214618683,
+      "learning_rate": 4.572990556553634e-05,
+      "loss": 0.4196,
       "step": 29580
     },
     {
-      "epoch": 1.04,
-      "learning_rate": 4.59542448883941e-05,
-      "loss": 0.2625,
+      "epoch": 1.0662413954661765,
+      "grad_norm": 0.1972283273935318,
+      "learning_rate": 4.572827431127012e-05,
+      "loss": 0.4352,
       "step": 29585
     },
     {
-      "epoch": 1.04,
-      "learning_rate": 4.595269103302852e-05,
-      "loss": 0.2995,
+      "epoch": 1.0664215951274012,
+      "grad_norm": 0.17068789899349213,
+      "learning_rate": 4.5726642774585195e-05,
+      "loss": 0.4277,
       "step": 29590
     },
     {
-      "epoch": 1.04,
-      "learning_rate": 4.595113690560605e-05,
-      "loss": 0.2965,
+      "epoch": 1.0666017947886257,
+      "grad_norm": 0.22462329268455505,
+      "learning_rate": 4.5725010955503764e-05,
+      "loss": 0.4148,
       "step": 29595
     },
     {
-      "epoch": 1.04,
-      "learning_rate": 4.594958250614687e-05,
-      "loss": 0.2867,
+      "epoch": 1.0667819944498504,
+      "grad_norm": 0.14778359234333038,
+      "learning_rate": 4.572337885404807e-05,
+      "loss": 0.4104,
       "step": 29600
     },
     {
-      "epoch": 1.04,
-      "learning_rate": 4.594802783467117e-05,
-      "loss": 0.2938,
+      "epoch": 1.066962194111075,
+      "grad_norm": 0.1986793428659439,
+      "learning_rate": 4.5721746470240355e-05,
+      "loss": 0.4351,
       "step": 29605
     },
     {
-      "epoch": 1.04,
-      "learning_rate": 4.594647289119914e-05,
-      "loss": 0.2893,
+      "epoch": 1.0671423937722997,
+      "grad_norm": 0.20412606000900269,
+      "learning_rate": 4.572011380410286e-05,
+      "loss": 0.3812,
       "step": 29610
     },
     {
-      "epoch": 1.04,
-      "learning_rate": 4.594491767575096e-05,
-      "loss": 0.2957,
+      "epoch": 1.0673225934335244,
+      "grad_norm": 0.17064638435840607,
+      "learning_rate": 4.571848085565783e-05,
+      "loss": 0.4169,
       "step": 29615
     },
     {
-      "epoch": 1.04,
-      "learning_rate": 4.594336218834683e-05,
-      "loss": 0.26,
+      "epoch": 1.067502793094749,
+      "grad_norm": 0.20599807798862457,
+      "learning_rate": 4.5716847624927496e-05,
+      "loss": 0.4217,
       "step": 29620
     },
     {
-      "epoch": 1.04,
-      "learning_rate": 4.594180642900694e-05,
-      "loss": 0.2826,
+      "epoch": 1.0676829927559737,
+      "grad_norm": 0.1645248532295227,
+      "learning_rate": 4.5715214111934134e-05,
+      "loss": 0.3981,
       "step": 29625
     },
     {
-      "epoch": 1.04,
-      "learning_rate": 4.5940250397751494e-05,
-      "loss": 0.2641,
+      "epoch": 1.0678631924171982,
+      "grad_norm": 0.16410000622272491,
+      "learning_rate": 4.571358031669999e-05,
+      "loss": 0.4021,
       "step": 29630
     },
     {
-      "epoch": 1.04,
-      "learning_rate": 4.5938694094600697e-05,
-      "loss": 0.3104,
+      "epoch": 1.068043392078423,
+      "grad_norm": 0.19438453018665314,
+      "learning_rate": 4.571194623924731e-05,
+      "loss": 0.4029,
       "step": 29635
     },
     {
-      "epoch": 1.04,
-      "learning_rate": 4.593713751957476e-05,
-      "loss": 0.2913,
+      "epoch": 1.0682235917396474,
+      "grad_norm": 0.18021252751350403,
+      "learning_rate": 4.571031187959839e-05,
+      "loss": 0.3966,
       "step": 29640
     },
     {
-      "epoch": 1.04,
-      "learning_rate": 4.593558067269389e-05,
-      "loss": 0.3053,
+      "epoch": 1.0684037914008722,
+      "grad_norm": 0.17498481273651123,
+      "learning_rate": 4.570867723777548e-05,
+      "loss": 0.3946,
       "step": 29645
     },
     {
-      "epoch": 1.04,
-      "learning_rate": 4.59340235539783e-05,
-      "loss": 0.301,
+      "epoch": 1.068583991062097,
+      "grad_norm": 0.16093195974826813,
+      "learning_rate": 4.570704231380084e-05,
+      "loss": 0.4181,
       "step": 29650
     },
     {
-      "epoch": 1.04,
-      "learning_rate": 4.593246616344821e-05,
-      "loss": 0.2842,
+      "epoch": 1.0687641907233214,
+      "grad_norm": 0.20035532116889954,
+      "learning_rate": 4.570540710769676e-05,
+      "loss": 0.4173,
       "step": 29655
     },
     {
-      "epoch": 1.04,
-      "learning_rate": 4.593090850112384e-05,
-      "loss": 0.3104,
+      "epoch": 1.0689443903845461,
+      "grad_norm": 0.16321523487567902,
+      "learning_rate": 4.5703771619485523e-05,
+      "loss": 0.4154,
       "step": 29660
     },
     {
-      "epoch": 1.04,
-      "learning_rate": 4.592935056702542e-05,
-      "loss": 0.2777,
+      "epoch": 1.0691245900457707,
+      "grad_norm": 0.17585666477680206,
+      "learning_rate": 4.570213584918941e-05,
+      "loss": 0.4138,
       "step": 29665
     },
     {
-      "epoch": 1.04,
-      "learning_rate": 4.592779236117318e-05,
-      "loss": 0.3036,
+      "epoch": 1.0693047897069954,
+      "grad_norm": 0.14717882871627808,
+      "learning_rate": 4.570049979683069e-05,
+      "loss": 0.4144,
       "step": 29670
     },
     {
-      "epoch": 1.04,
-      "learning_rate": 4.592623388358734e-05,
-      "loss": 0.3212,
+      "epoch": 1.06948498936822,
+      "grad_norm": 0.16287022829055786,
+      "learning_rate": 4.569886346243167e-05,
+      "loss": 0.4082,
       "step": 29675
     },
     {
-      "epoch": 1.04,
-      "learning_rate": 4.592467513428815e-05,
-      "loss": 0.2755,
+      "epoch": 1.0696651890294446,
+      "grad_norm": 0.1774132400751114,
+      "learning_rate": 4.569722684601465e-05,
+      "loss": 0.443,
       "step": 29680
     },
     {
-      "epoch": 1.04,
-      "learning_rate": 4.592311611329584e-05,
-      "loss": 0.2793,
+      "epoch": 1.0698453886906694,
+      "grad_norm": 0.17134885489940643,
+      "learning_rate": 4.569558994760192e-05,
+      "loss": 0.423,
       "step": 29685
     },
     {
-      "epoch": 1.04,
-      "learning_rate": 4.5921556820630656e-05,
-      "loss": 0.2717,
+      "epoch": 1.0700255883518939,
+      "grad_norm": 0.20845064520835876,
+      "learning_rate": 4.569395276721579e-05,
+      "loss": 0.4343,
       "step": 29690
     },
     {
-      "epoch": 1.04,
-      "learning_rate": 4.591999725631285e-05,
-      "loss": 0.2906,
+      "epoch": 1.0702057880131186,
+      "grad_norm": 0.19249121844768524,
+      "learning_rate": 4.569231530487855e-05,
+      "loss": 0.4168,
       "step": 29695
     },
     {
-      "epoch": 1.04,
-      "learning_rate": 4.591843742036266e-05,
-      "loss": 0.3059,
+      "epoch": 1.0703859876743431,
+      "grad_norm": 0.21180908381938934,
+      "learning_rate": 4.569067756061253e-05,
+      "loss": 0.4245,
       "step": 29700
     },
     {
-      "epoch": 1.05,
-      "learning_rate": 4.591687731280034e-05,
-      "loss": 0.3108,
+      "epoch": 1.0705661873355679,
+      "grad_norm": 0.199473574757576,
+      "learning_rate": 4.568903953444003e-05,
+      "loss": 0.4199,
       "step": 29705
     },
     {
-      "epoch": 1.05,
-      "learning_rate": 4.591531693364617e-05,
-      "loss": 0.2996,
+      "epoch": 1.0707463869967924,
+      "grad_norm": 0.19441378116607666,
+      "learning_rate": 4.568740122638337e-05,
+      "loss": 0.4198,
       "step": 29710
     },
     {
-      "epoch": 1.05,
-      "learning_rate": 4.591375628292038e-05,
-      "loss": 0.315,
+      "epoch": 1.070926586658017,
+      "grad_norm": 0.17675237357616425,
+      "learning_rate": 4.568576263646487e-05,
+      "loss": 0.3893,
       "step": 29715
     },
     {
-      "epoch": 1.05,
-      "learning_rate": 4.591219536064325e-05,
-      "loss": 0.2997,
+      "epoch": 1.0711067863192416,
+      "grad_norm": 0.2046016901731491,
+      "learning_rate": 4.5684123764706855e-05,
+      "loss": 0.4285,
       "step": 29720
     },
     {
-      "epoch": 1.05,
-      "learning_rate": 4.5910634166835045e-05,
-      "loss": 0.2891,
+      "epoch": 1.0712869859804663,
+      "grad_norm": 0.1835736781358719,
+      "learning_rate": 4.568248461113167e-05,
+      "loss": 0.4359,
       "step": 29725
     },
     {
-      "epoch": 1.05,
-      "learning_rate": 4.5909072701516035e-05,
-      "loss": 0.2988,
+      "epoch": 1.071467185641691,
+      "grad_norm": 0.1686883121728897,
+      "learning_rate": 4.5680845175761635e-05,
+      "loss": 0.4405,
       "step": 29730
     },
     {
-      "epoch": 1.05,
-      "learning_rate": 4.59075109647065e-05,
-      "loss": 0.2868,
+      "epoch": 1.0716473853029156,
+      "grad_norm": 0.15105289220809937,
+      "learning_rate": 4.567920545861908e-05,
+      "loss": 0.4067,
       "step": 29735
     },
     {
-      "epoch": 1.05,
-      "learning_rate": 4.590594895642671e-05,
-      "loss": 0.2998,
+      "epoch": 1.0718275849641403,
+      "grad_norm": 0.21068331599235535,
+      "learning_rate": 4.5677565459726355e-05,
+      "loss": 0.4408,
       "step": 29740
     },
     {
-      "epoch": 1.05,
-      "learning_rate": 4.5904386676696955e-05,
-      "loss": 0.2672,
+      "epoch": 1.0720077846253648,
+      "grad_norm": 0.16096103191375732,
+      "learning_rate": 4.567592517910582e-05,
+      "loss": 0.4391,
       "step": 29745
     },
     {
-      "epoch": 1.05,
-      "learning_rate": 4.5902824125537516e-05,
-      "loss": 0.3028,
+      "epoch": 1.0721879842865896,
+      "grad_norm": 0.18630751967430115,
+      "learning_rate": 4.567428461677979e-05,
+      "loss": 0.4219,
       "step": 29750
     },
     {
-      "epoch": 1.05,
-      "learning_rate": 4.5901261302968676e-05,
-      "loss": 0.3042,
+      "epoch": 1.072368183947814,
+      "grad_norm": 0.18894068896770477,
+      "learning_rate": 4.567264377277064e-05,
+      "loss": 0.3977,
       "step": 29755
     },
     {
-      "epoch": 1.05,
-      "learning_rate": 4.589969820901073e-05,
-      "loss": 0.2895,
+      "epoch": 1.0725483836090388,
+      "grad_norm": 0.2151520699262619,
+      "learning_rate": 4.5671002647100716e-05,
+      "loss": 0.4539,
       "step": 29760
     },
     {
-      "epoch": 1.05,
-      "learning_rate": 4.589813484368398e-05,
-      "loss": 0.2965,
+      "epoch": 1.0727285832702635,
+      "grad_norm": 0.1710183173418045,
+      "learning_rate": 4.566936123979239e-05,
+      "loss": 0.4098,
       "step": 29765
     },
     {
-      "epoch": 1.05,
-      "learning_rate": 4.589657120700872e-05,
-      "loss": 0.2915,
+      "epoch": 1.072908782931488,
+      "grad_norm": 0.17231152951717377,
+      "learning_rate": 4.566771955086802e-05,
+      "loss": 0.4076,
       "step": 29770
     },
     {
-      "epoch": 1.05,
-      "learning_rate": 4.589500729900525e-05,
-      "loss": 0.2911,
+      "epoch": 1.0730889825927128,
+      "grad_norm": 0.18413393199443817,
+      "learning_rate": 4.566607758034997e-05,
+      "loss": 0.4398,
       "step": 29775
     },
     {
-      "epoch": 1.05,
-      "learning_rate": 4.5893443119693895e-05,
-      "loss": 0.2829,
+      "epoch": 1.0732691822539373,
+      "grad_norm": 0.18336208164691925,
+      "learning_rate": 4.566443532826061e-05,
+      "loss": 0.4552,
       "step": 29780
     },
     {
-      "epoch": 1.05,
-      "learning_rate": 4.589187866909495e-05,
-      "loss": 0.2855,
+      "epoch": 1.073449381915162,
+      "grad_norm": 0.18550068140029907,
+      "learning_rate": 4.566279279462232e-05,
+      "loss": 0.4303,
       "step": 29785
     },
     {
-      "epoch": 1.05,
-      "learning_rate": 4.589031394722871e-05,
-      "loss": 0.2959,
+      "epoch": 1.0736295815763865,
+      "grad_norm": 0.1502627283334732,
+      "learning_rate": 4.5661149979457485e-05,
+      "loss": 0.4258,
       "step": 29790
     },
     {
-      "epoch": 1.05,
-      "learning_rate": 4.588874895411552e-05,
-      "loss": 0.2871,
+      "epoch": 1.0738097812376113,
+      "grad_norm": 0.17316554486751556,
+      "learning_rate": 4.565950688278848e-05,
+      "loss": 0.414,
       "step": 29795
     },
     {
-      "epoch": 1.05,
-      "learning_rate": 4.5887183689775694e-05,
-      "loss": 0.2746,
+      "epoch": 1.073989980898836,
+      "grad_norm": 0.19836434721946716,
+      "learning_rate": 4.5657863504637694e-05,
+      "loss": 0.3891,
       "step": 29800
     },
     {
-      "epoch": 1.05,
-      "learning_rate": 4.588561815422955e-05,
-      "loss": 0.2942,
+      "epoch": 1.0741701805600605,
+      "grad_norm": 0.17455318570137024,
+      "learning_rate": 4.5656219845027516e-05,
+      "loss": 0.3748,
       "step": 29805
     },
     {
-      "epoch": 1.05,
-      "learning_rate": 4.588405234749742e-05,
-      "loss": 0.3044,
+      "epoch": 1.0743503802212853,
+      "grad_norm": 0.2048964947462082,
+      "learning_rate": 4.565457590398034e-05,
+      "loss": 0.431,
       "step": 29810
     },
     {
-      "epoch": 1.05,
-      "learning_rate": 4.588248626959963e-05,
-      "loss": 0.2734,
+      "epoch": 1.0745305798825098,
+      "grad_norm": 0.1792224943637848,
+      "learning_rate": 4.5652931681518565e-05,
+      "loss": 0.4157,
       "step": 29815
     },
     {
-      "epoch": 1.05,
-      "learning_rate": 4.588091992055653e-05,
-      "loss": 0.2695,
+      "epoch": 1.0747107795437345,
+      "grad_norm": 0.22106477618217468,
+      "learning_rate": 4.56512871776646e-05,
+      "loss": 0.4332,
       "step": 29820
     },
     {
-      "epoch": 1.05,
-      "learning_rate": 4.587935330038843e-05,
-      "loss": 0.3035,
+      "epoch": 1.074890979204959,
+      "grad_norm": 0.19627970457077026,
+      "learning_rate": 4.564964239244084e-05,
+      "loss": 0.4105,
       "step": 29825
     },
     {
-      "epoch": 1.05,
-      "learning_rate": 4.58777864091157e-05,
-      "loss": 0.3184,
+      "epoch": 1.0750711788661838,
+      "grad_norm": 0.1940174698829651,
+      "learning_rate": 4.564799732586971e-05,
+      "loss": 0.4369,
       "step": 29830
     },
     {
-      "epoch": 1.05,
-      "learning_rate": 4.587621924675866e-05,
-      "loss": 0.3016,
+      "epoch": 1.0752513785274083,
+      "grad_norm": 0.19587305188179016,
+      "learning_rate": 4.56463519779736e-05,
+      "loss": 0.3994,
       "step": 29835
     },
     {
-      "epoch": 1.05,
-      "learning_rate": 4.5874651813337674e-05,
-      "loss": 0.3051,
+      "epoch": 1.075431578188633,
+      "grad_norm": 0.14254744350910187,
+      "learning_rate": 4.564470634877495e-05,
+      "loss": 0.4357,
       "step": 29840
     },
     {
-      "epoch": 1.05,
-      "learning_rate": 4.58730841088731e-05,
-      "loss": 0.2728,
+      "epoch": 1.0756117778498577,
+      "grad_norm": 0.2016969919204712,
+      "learning_rate": 4.564306043829617e-05,
+      "loss": 0.4505,
       "step": 29845
     },
     {
-      "epoch": 1.05,
-      "learning_rate": 4.587151613338528e-05,
-      "loss": 0.2969,
+      "epoch": 1.0757919775110822,
+      "grad_norm": 0.17493173480033875,
+      "learning_rate": 4.564141424655969e-05,
+      "loss": 0.3733,
       "step": 29850
     },
     {
-      "epoch": 1.05,
-      "learning_rate": 4.5869947886894577e-05,
-      "loss": 0.2884,
+      "epoch": 1.075972177172307,
+      "grad_norm": 0.16929864883422852,
+      "learning_rate": 4.563976777358793e-05,
+      "loss": 0.3971,
       "step": 29855
     },
     {
-      "epoch": 1.05,
-      "learning_rate": 4.586837936942136e-05,
-      "loss": 0.2899,
+      "epoch": 1.0761523768335315,
+      "grad_norm": 0.18552954494953156,
+      "learning_rate": 4.563812101940334e-05,
+      "loss": 0.4258,
       "step": 29860
     },
     {
-      "epoch": 1.05,
-      "learning_rate": 4.586681058098598e-05,
-      "loss": 0.2928,
+      "epoch": 1.0763325764947562,
+      "grad_norm": 0.1586046814918518,
+      "learning_rate": 4.563647398402834e-05,
+      "loss": 0.4375,
       "step": 29865
     },
     {
-      "epoch": 1.05,
-      "learning_rate": 4.586524152160881e-05,
-      "loss": 0.2819,
+      "epoch": 1.0765127761559807,
+      "grad_norm": 0.1995263397693634,
+      "learning_rate": 4.5634826667485385e-05,
+      "loss": 0.4137,
       "step": 29870
     },
     {
-      "epoch": 1.05,
-      "learning_rate": 4.586367219131025e-05,
-      "loss": 0.2921,
+      "epoch": 1.0766929758172055,
+      "grad_norm": 0.16926749050617218,
+      "learning_rate": 4.56331790697969e-05,
+      "loss": 0.4065,
       "step": 29875
     },
     {
-      "epoch": 1.05,
-      "learning_rate": 4.5862102590110644e-05,
-      "loss": 0.3056,
+      "epoch": 1.0768731754784302,
+      "grad_norm": 0.18349143862724304,
+      "learning_rate": 4.563153119098535e-05,
+      "loss": 0.4106,
       "step": 29880
     },
     {
-      "epoch": 1.05,
-      "learning_rate": 4.586053271803039e-05,
-      "loss": 0.2808,
+      "epoch": 1.0770533751396547,
+      "grad_norm": 0.18791143596172333,
+      "learning_rate": 4.562988303107319e-05,
+      "loss": 0.4174,
       "step": 29885
     },
     {
-      "epoch": 1.05,
-      "learning_rate": 4.585896257508986e-05,
-      "loss": 0.3111,
+      "epoch": 1.0772335748008794,
+      "grad_norm": 0.1980997622013092,
+      "learning_rate": 4.5628234590082864e-05,
+      "loss": 0.4179,
       "step": 29890
     },
     {
-      "epoch": 1.05,
-      "learning_rate": 4.585739216130945e-05,
-      "loss": 0.3143,
+      "epoch": 1.077413774462104,
+      "grad_norm": 0.19258050620555878,
+      "learning_rate": 4.562658586803683e-05,
+      "loss": 0.4205,
       "step": 29895
     },
     {
-      "epoch": 1.05,
-      "learning_rate": 4.5855821476709546e-05,
-      "loss": 0.2987,
+      "epoch": 1.0775939741233287,
+      "grad_norm": 0.1779131144285202,
+      "learning_rate": 4.5624936864957556e-05,
+      "loss": 0.3996,
       "step": 29900
     },
     {
-      "epoch": 1.05,
-      "learning_rate": 4.5854250521310556e-05,
-      "loss": 0.2665,
+      "epoch": 1.0777741737845532,
+      "grad_norm": 0.14219936728477478,
+      "learning_rate": 4.562328758086752e-05,
+      "loss": 0.4039,
       "step": 29905
     },
     {
-      "epoch": 1.05,
-      "learning_rate": 4.5852679295132853e-05,
-      "loss": 0.2807,
+      "epoch": 1.077954373445778,
+      "grad_norm": 0.18294230103492737,
+      "learning_rate": 4.562163801578918e-05,
+      "loss": 0.4341,
       "step": 29910
     },
     {
-      "epoch": 1.05,
-      "learning_rate": 4.5851107798196854e-05,
-      "loss": 0.279,
+      "epoch": 1.0781345731070027,
+      "grad_norm": 0.18481683731079102,
+      "learning_rate": 4.561998816974501e-05,
+      "loss": 0.4289,
       "step": 29915
     },
     {
-      "epoch": 1.05,
-      "learning_rate": 4.584953603052297e-05,
-      "loss": 0.3015,
+      "epoch": 1.0783147727682272,
+      "grad_norm": 0.17565883696079254,
+      "learning_rate": 4.561833804275749e-05,
+      "loss": 0.4318,
       "step": 29920
     },
     {
-      "epoch": 1.05,
-      "learning_rate": 4.584796399213159e-05,
-      "loss": 0.3045,
+      "epoch": 1.078494972429452,
+      "grad_norm": 0.17877116799354553,
+      "learning_rate": 4.561668763484911e-05,
+      "loss": 0.4196,
       "step": 29925
     },
     {
-      "epoch": 1.05,
-      "learning_rate": 4.584639168304315e-05,
-      "loss": 0.2951,
+      "epoch": 1.0786751720906764,
+      "grad_norm": 0.1558140218257904,
+      "learning_rate": 4.561503694604236e-05,
+      "loss": 0.3991,
       "step": 29930
     },
     {
-      "epoch": 1.05,
-      "learning_rate": 4.584481910327804e-05,
-      "loss": 0.2686,
+      "epoch": 1.0788553717519012,
+      "grad_norm": 0.19004541635513306,
+      "learning_rate": 4.561338597635972e-05,
+      "loss": 0.468,
       "step": 29935
     },
     {
-      "epoch": 1.05,
-      "learning_rate": 4.5843246252856697e-05,
-      "loss": 0.2661,
+      "epoch": 1.0790355714131257,
+      "grad_norm": 0.14542394876480103,
+      "learning_rate": 4.561173472582368e-05,
+      "loss": 0.448,
       "step": 29940
     },
     {
-      "epoch": 1.05,
-      "learning_rate": 4.584167313179954e-05,
-      "loss": 0.2954,
+      "epoch": 1.0792157710743504,
+      "grad_norm": 0.19577011466026306,
+      "learning_rate": 4.5610083194456754e-05,
+      "loss": 0.4157,
       "step": 29945
     },
     {
-      "epoch": 1.05,
-      "learning_rate": 4.584009974012699e-05,
-      "loss": 0.2725,
+      "epoch": 1.079395970735575,
+      "grad_norm": 0.21589916944503784,
+      "learning_rate": 4.5608431382281426e-05,
+      "loss": 0.4265,
       "step": 29950
     },
     {
-      "epoch": 1.05,
-      "learning_rate": 4.583852607785948e-05,
-      "loss": 0.3023,
+      "epoch": 1.0795761703967997,
+      "grad_norm": 0.14693371951580048,
+      "learning_rate": 4.560677928932021e-05,
+      "loss": 0.411,
       "step": 29955
     },
     {
-      "epoch": 1.05,
-      "learning_rate": 4.583695214501743e-05,
-      "loss": 0.3105,
+      "epoch": 1.0797563700580244,
+      "grad_norm": 0.19670480489730835,
+      "learning_rate": 4.560512691559562e-05,
+      "loss": 0.4504,
       "step": 29960
     },
     {
-      "epoch": 1.05,
-      "learning_rate": 4.58353779416213e-05,
-      "loss": 0.279,
+      "epoch": 1.079936569719249,
+      "grad_norm": 0.18953409790992737,
+      "learning_rate": 4.560347426113017e-05,
+      "loss": 0.4325,
       "step": 29965
     },
     {
-      "epoch": 1.05,
-      "learning_rate": 4.5833803467691515e-05,
-      "loss": 0.3067,
+      "epoch": 1.0801167693804736,
+      "grad_norm": 0.18052424490451813,
+      "learning_rate": 4.560182132594637e-05,
+      "loss": 0.4092,
       "step": 29970
     },
     {
-      "epoch": 1.05,
-      "learning_rate": 4.583222872324853e-05,
-      "loss": 0.2747,
+      "epoch": 1.0802969690416981,
+      "grad_norm": 0.19230318069458008,
+      "learning_rate": 4.560016811006673e-05,
+      "loss": 0.4343,
       "step": 29975
     },
     {
-      "epoch": 1.05,
-      "learning_rate": 4.5830653708312775e-05,
-      "loss": 0.32,
+      "epoch": 1.0804771687029229,
+      "grad_norm": 0.2135818749666214,
+      "learning_rate": 4.5598514613513796e-05,
+      "loss": 0.4111,
       "step": 29980
     },
     {
-      "epoch": 1.05,
-      "learning_rate": 4.582907842290472e-05,
-      "loss": 0.3127,
+      "epoch": 1.0806573683641474,
+      "grad_norm": 0.1807836890220642,
+      "learning_rate": 4.5596860836310094e-05,
+      "loss": 0.435,
       "step": 29985
     },
     {
-      "epoch": 1.06,
-      "learning_rate": 4.582750286704479e-05,
-      "loss": 0.2701,
+      "epoch": 1.0808375680253721,
+      "grad_norm": 0.18161781132221222,
+      "learning_rate": 4.559520677847815e-05,
+      "loss": 0.4094,
       "step": 29990
     },
     {
-      "epoch": 1.06,
-      "learning_rate": 4.582592704075348e-05,
-      "loss": 0.2854,
+      "epoch": 1.0810177676865969,
+      "grad_norm": 0.19649210572242737,
+      "learning_rate": 4.55935524400405e-05,
+      "loss": 0.3864,
       "step": 29995
     },
     {
-      "epoch": 1.06,
-      "learning_rate": 4.582435094405122e-05,
-      "loss": 0.3255,
+      "epoch": 1.0811979673478214,
+      "grad_norm": 0.15059545636177063,
+      "learning_rate": 4.559189782101968e-05,
+      "loss": 0.3919,
       "step": 30000
     },
     {
-      "epoch": 1.06,
-      "eval_loss": 0.29047587513923645,
-      "eval_runtime": 10.5326,
-      "eval_samples_per_second": 9.494,
-      "eval_steps_per_second": 9.494,
+      "epoch": 1.0811979673478214,
+      "eval_loss": 0.4527240991592407,
+      "eval_runtime": 3.5376,
+      "eval_samples_per_second": 28.268,
+      "eval_steps_per_second": 7.067,
       "step": 30000
     },
     {
-      "epoch": 1.06,
-      "learning_rate": 4.582277457695849e-05,
-      "loss": 0.3098,
+      "epoch": 1.081378167009046,
+      "grad_norm": 0.16200894117355347,
+      "learning_rate": 4.559024292143824e-05,
+      "loss": 0.4298,
       "step": 30005
     },
     {
-      "epoch": 1.06,
-      "learning_rate": 4.582119793949576e-05,
-      "loss": 0.2848,
+      "epoch": 1.0815583666702706,
+      "grad_norm": 0.19894878566265106,
+      "learning_rate": 4.558858774131873e-05,
+      "loss": 0.4217,
       "step": 30010
     },
     {
-      "epoch": 1.06,
-      "learning_rate": 4.581962103168349e-05,
-      "loss": 0.2573,
+      "epoch": 1.0817385663314953,
+      "grad_norm": 0.18201377987861633,
+      "learning_rate": 4.5586932280683696e-05,
+      "loss": 0.4006,
       "step": 30015
     },
     {
-      "epoch": 1.06,
-      "learning_rate": 4.581804385354217e-05,
-      "loss": 0.3003,
+      "epoch": 1.0819187659927199,
+      "grad_norm": 0.22016799449920654,
+      "learning_rate": 4.55852765395557e-05,
+      "loss": 0.4411,
       "step": 30020
     },
     {
-      "epoch": 1.06,
-      "learning_rate": 4.5816466405092265e-05,
-      "loss": 0.2979,
+      "epoch": 1.0820989656539446,
+      "grad_norm": 0.21885946393013,
+      "learning_rate": 4.5583620517957296e-05,
+      "loss": 0.4293,
       "step": 30025
     },
     {
-      "epoch": 1.06,
-      "learning_rate": 4.5814888686354254e-05,
-      "loss": 0.2854,
+      "epoch": 1.0822791653151693,
+      "grad_norm": 0.21789535880088806,
+      "learning_rate": 4.5581964215911036e-05,
+      "loss": 0.4038,
       "step": 30030
     },
     {
-      "epoch": 1.06,
-      "learning_rate": 4.5813310697348644e-05,
-      "loss": 0.3119,
+      "epoch": 1.0824593649763938,
+      "grad_norm": 0.14212708175182343,
+      "learning_rate": 4.558030763343951e-05,
+      "loss": 0.4043,
       "step": 30035
     },
     {
-      "epoch": 1.06,
-      "learning_rate": 4.58117324380959e-05,
-      "loss": 0.2668,
+      "epoch": 1.0826395646376186,
+      "grad_norm": 0.1961928755044937,
+      "learning_rate": 4.5578650770565274e-05,
+      "loss": 0.4439,
       "step": 30040
     },
     {
-      "epoch": 1.06,
-      "learning_rate": 4.5810153908616534e-05,
-      "loss": 0.3009,
+      "epoch": 1.082819764298843,
+      "grad_norm": 0.18222855031490326,
+      "learning_rate": 4.55769936273109e-05,
+      "loss": 0.4137,
       "step": 30045
     },
     {
-      "epoch": 1.06,
-      "learning_rate": 4.580857510893103e-05,
-      "loss": 0.2783,
+      "epoch": 1.0829999639600678,
+      "grad_norm": 0.18094827234745026,
+      "learning_rate": 4.5575336203698985e-05,
+      "loss": 0.3873,
       "step": 30050
     },
     {
-      "epoch": 1.06,
-      "learning_rate": 4.5806996039059894e-05,
-      "loss": 0.2799,
+      "epoch": 1.0831801636212923,
+      "grad_norm": 0.14984826743602753,
+      "learning_rate": 4.557367849975208e-05,
+      "loss": 0.4094,
       "step": 30055
     },
     {
-      "epoch": 1.06,
-      "learning_rate": 4.580541669902362e-05,
-      "loss": 0.3114,
+      "epoch": 1.083360363282517,
+      "grad_norm": 0.18728286027908325,
+      "learning_rate": 4.5572020515492805e-05,
+      "loss": 0.4556,
       "step": 30060
     },
     {
-      "epoch": 1.06,
-      "learning_rate": 4.5803837088842736e-05,
-      "loss": 0.2727,
+      "epoch": 1.0835405629437416,
+      "grad_norm": 0.1479881852865219,
+      "learning_rate": 4.5570362250943725e-05,
+      "loss": 0.3622,
       "step": 30065
     },
     {
-      "epoch": 1.06,
-      "learning_rate": 4.5802257208537725e-05,
-      "loss": 0.2785,
+      "epoch": 1.0837207626049663,
+      "grad_norm": 0.2368391901254654,
+      "learning_rate": 4.556870370612744e-05,
+      "loss": 0.4298,
       "step": 30070
     },
     {
-      "epoch": 1.06,
-      "learning_rate": 4.5800677058129114e-05,
-      "loss": 0.2772,
+      "epoch": 1.083900962266191,
+      "grad_norm": 0.1753835380077362,
+      "learning_rate": 4.5567044881066555e-05,
+      "loss": 0.4145,
       "step": 30075
     },
     {
-      "epoch": 1.06,
-      "learning_rate": 4.579909663763742e-05,
-      "loss": 0.2984,
+      "epoch": 1.0840811619274155,
+      "grad_norm": 0.17474301159381866,
+      "learning_rate": 4.556538577578366e-05,
+      "loss": 0.4013,
       "step": 30080
     },
     {
-      "epoch": 1.06,
-      "learning_rate": 4.579751594708317e-05,
-      "loss": 0.3128,
+      "epoch": 1.0842613615886403,
+      "grad_norm": 0.2075207531452179,
+      "learning_rate": 4.5563726390301374e-05,
+      "loss": 0.3922,
       "step": 30085
     },
     {
-      "epoch": 1.06,
-      "learning_rate": 4.579593498648688e-05,
-      "loss": 0.3048,
+      "epoch": 1.0844415612498648,
+      "grad_norm": 0.1771383434534073,
+      "learning_rate": 4.5562066724642286e-05,
+      "loss": 0.3737,
       "step": 30090
     },
     {
-      "epoch": 1.06,
-      "learning_rate": 4.5794353755869076e-05,
-      "loss": 0.2876,
+      "epoch": 1.0846217609110895,
+      "grad_norm": 0.1907252073287964,
+      "learning_rate": 4.556040677882903e-05,
+      "loss": 0.4369,
       "step": 30095
     },
     {
-      "epoch": 1.06,
-      "learning_rate": 4.5792772255250285e-05,
-      "loss": 0.3073,
+      "epoch": 1.084801960572314,
+      "grad_norm": 0.1812266856431961,
+      "learning_rate": 4.55587465528842e-05,
+      "loss": 0.4365,
       "step": 30100
     },
     {
-      "epoch": 1.06,
-      "learning_rate": 4.579119048465106e-05,
-      "loss": 0.3059,
+      "epoch": 1.0849821602335388,
+      "grad_norm": 0.17140717804431915,
+      "learning_rate": 4.5557086046830446e-05,
+      "loss": 0.4023,
       "step": 30105
     },
     {
-      "epoch": 1.06,
-      "learning_rate": 4.578960844409192e-05,
-      "loss": 0.2862,
+      "epoch": 1.0851623598947633,
+      "grad_norm": 0.20850670337677002,
+      "learning_rate": 4.555542526069036e-05,
+      "loss": 0.4173,
       "step": 30110
     },
     {
-      "epoch": 1.06,
-      "learning_rate": 4.5788026133593434e-05,
-      "loss": 0.2938,
+      "epoch": 1.085342559555988,
+      "grad_norm": 0.21336951851844788,
+      "learning_rate": 4.5553764194486583e-05,
+      "loss": 0.431,
       "step": 30115
     },
     {
-      "epoch": 1.06,
-      "learning_rate": 4.578644355317611e-05,
-      "loss": 0.2748,
+      "epoch": 1.0855227592172128,
+      "grad_norm": 0.16418558359146118,
+      "learning_rate": 4.5552102848241764e-05,
+      "loss": 0.4199,
       "step": 30120
     },
     {
-      "epoch": 1.06,
-      "learning_rate": 4.578486070286052e-05,
-      "loss": 0.3168,
+      "epoch": 1.0857029588784373,
+      "grad_norm": 0.15260203182697296,
+      "learning_rate": 4.5550441221978505e-05,
+      "loss": 0.4031,
       "step": 30125
     },
     {
-      "epoch": 1.06,
-      "learning_rate": 4.578327758266721e-05,
-      "loss": 0.2866,
+      "epoch": 1.085883158539662,
+      "grad_norm": 0.17375853657722473,
+      "learning_rate": 4.554877931571947e-05,
+      "loss": 0.4301,
       "step": 30130
     },
     {
-      "epoch": 1.06,
-      "learning_rate": 4.578169419261674e-05,
-      "loss": 0.293,
+      "epoch": 1.0860633582008865,
+      "grad_norm": 0.1743849366903305,
+      "learning_rate": 4.5547117129487305e-05,
+      "loss": 0.4394,
       "step": 30135
     },
     {
-      "epoch": 1.06,
-      "learning_rate": 4.578011053272967e-05,
-      "loss": 0.2965,
+      "epoch": 1.0862435578621112,
+      "grad_norm": 0.1563817262649536,
+      "learning_rate": 4.554545466330463e-05,
+      "loss": 0.423,
       "step": 30140
     },
     {
-      "epoch": 1.06,
-      "learning_rate": 4.577852660302655e-05,
-      "loss": 0.2518,
+      "epoch": 1.0864237575233358,
+      "grad_norm": 0.22052206099033356,
+      "learning_rate": 4.554379191719412e-05,
+      "loss": 0.4591,
       "step": 30145
     },
     {
-      "epoch": 1.06,
-      "learning_rate": 4.577694240352797e-05,
-      "loss": 0.2861,
+      "epoch": 1.0866039571845605,
+      "grad_norm": 0.23430973291397095,
+      "learning_rate": 4.554212889117842e-05,
+      "loss": 0.4231,
       "step": 30150
     },
     {
-      "epoch": 1.06,
-      "learning_rate": 4.577535793425447e-05,
-      "loss": 0.2965,
+      "epoch": 1.0867841568457852,
+      "grad_norm": 0.1482902467250824,
+      "learning_rate": 4.55404655852802e-05,
+      "loss": 0.4042,
       "step": 30155
     },
     {
-      "epoch": 1.06,
-      "learning_rate": 4.5773773195226646e-05,
-      "loss": 0.2779,
+      "epoch": 1.0869643565070097,
+      "grad_norm": 0.19089116156101227,
+      "learning_rate": 4.5538801999522115e-05,
+      "loss": 0.3929,
       "step": 30160
     },
     {
-      "epoch": 1.06,
-      "learning_rate": 4.577218818646506e-05,
-      "loss": 0.2751,
+      "epoch": 1.0871445561682345,
+      "grad_norm": 0.1974368840456009,
+      "learning_rate": 4.553713813392682e-05,
+      "loss": 0.393,
       "step": 30165
     },
     {
-      "epoch": 1.06,
-      "learning_rate": 4.5770602907990314e-05,
-      "loss": 0.298,
+      "epoch": 1.087324755829459,
+      "grad_norm": 0.20312896370887756,
+      "learning_rate": 4.5535473988517e-05,
+      "loss": 0.3933,
       "step": 30170
     },
     {
-      "epoch": 1.06,
-      "learning_rate": 4.5769017359822965e-05,
-      "loss": 0.3175,
+      "epoch": 1.0875049554906837,
+      "grad_norm": 0.16208785772323608,
+      "learning_rate": 4.553380956331531e-05,
+      "loss": 0.4074,
       "step": 30175
     },
     {
-      "epoch": 1.06,
-      "learning_rate": 4.5767431541983616e-05,
-      "loss": 0.3072,
+      "epoch": 1.0876851551519082,
+      "grad_norm": 0.16057203710079193,
+      "learning_rate": 4.5532144858344446e-05,
+      "loss": 0.4167,
       "step": 30180
     },
     {
-      "epoch": 1.06,
-      "learning_rate": 4.576584545449286e-05,
-      "loss": 0.2959,
+      "epoch": 1.087865354813133,
+      "grad_norm": 0.19308283925056458,
+      "learning_rate": 4.5530479873627095e-05,
+      "loss": 0.4052,
       "step": 30185
     },
     {
-      "epoch": 1.06,
-      "learning_rate": 4.576425909737128e-05,
-      "loss": 0.2951,
+      "epoch": 1.0880455544743577,
+      "grad_norm": 0.15638864040374756,
+      "learning_rate": 4.552881460918592e-05,
+      "loss": 0.4148,
       "step": 30190
     },
     {
-      "epoch": 1.06,
-      "learning_rate": 4.576267247063948e-05,
-      "loss": 0.2759,
+      "epoch": 1.0882257541355822,
+      "grad_norm": 0.1577041745185852,
+      "learning_rate": 4.552714906504362e-05,
+      "loss": 0.4295,
       "step": 30195
     },
     {
-      "epoch": 1.06,
-      "learning_rate": 4.576108557431806e-05,
-      "loss": 0.2547,
+      "epoch": 1.088405953796807,
+      "grad_norm": 0.18273350596427917,
+      "learning_rate": 4.552548324122289e-05,
+      "loss": 0.4369,
       "step": 30200
     },
     {
-      "epoch": 1.06,
-      "learning_rate": 4.5759498408427634e-05,
-      "loss": 0.3103,
+      "epoch": 1.0885861534580314,
+      "grad_norm": 0.18305066227912903,
+      "learning_rate": 4.552381713774643e-05,
+      "loss": 0.3818,
       "step": 30205
     },
     {
-      "epoch": 1.06,
-      "learning_rate": 4.57579109729888e-05,
-      "loss": 0.2763,
+      "epoch": 1.0887663531192562,
+      "grad_norm": 0.1907605528831482,
+      "learning_rate": 4.5522150754636926e-05,
+      "loss": 0.4491,
       "step": 30210
     },
     {
-      "epoch": 1.06,
-      "learning_rate": 4.575632326802217e-05,
-      "loss": 0.2865,
+      "epoch": 1.0889465527804807,
+      "grad_norm": 0.18072152137756348,
+      "learning_rate": 4.55204840919171e-05,
+      "loss": 0.3953,
       "step": 30215
     },
     {
-      "epoch": 1.06,
-      "learning_rate": 4.5754735293548355e-05,
-      "loss": 0.2729,
+      "epoch": 1.0891267524417054,
+      "grad_norm": 0.15544374287128448,
+      "learning_rate": 4.551881714960965e-05,
+      "loss": 0.4226,
       "step": 30220
     },
     {
-      "epoch": 1.06,
-      "learning_rate": 4.575314704958799e-05,
-      "loss": 0.2916,
+      "epoch": 1.08930695210293,
+      "grad_norm": 0.19354958832263947,
+      "learning_rate": 4.5517149927737276e-05,
+      "loss": 0.3948,
       "step": 30225
     },
     {
-      "epoch": 1.06,
-      "learning_rate": 4.575155853616168e-05,
-      "loss": 0.3036,
+      "epoch": 1.0894871517641547,
+      "grad_norm": 0.21312229335308075,
+      "learning_rate": 4.551548242632272e-05,
+      "loss": 0.4241,
       "step": 30230
     },
     {
-      "epoch": 1.06,
-      "learning_rate": 4.5749969753290065e-05,
-      "loss": 0.2789,
+      "epoch": 1.0896673514253794,
+      "grad_norm": 0.19396165013313293,
+      "learning_rate": 4.5513814645388686e-05,
+      "loss": 0.435,
       "step": 30235
     },
     {
-      "epoch": 1.06,
-      "learning_rate": 4.574838070099376e-05,
-      "loss": 0.2888,
+      "epoch": 1.089847551086604,
+      "grad_norm": 0.18126662075519562,
+      "learning_rate": 4.55121465849579e-05,
+      "loss": 0.4313,
       "step": 30240
     },
     {
-      "epoch": 1.06,
-      "learning_rate": 4.574679137929341e-05,
-      "loss": 0.3114,
+      "epoch": 1.0900277507478286,
+      "grad_norm": 0.1702282428741455,
+      "learning_rate": 4.551047824505308e-05,
+      "loss": 0.4408,
       "step": 30245
     },
     {
-      "epoch": 1.06,
-      "learning_rate": 4.5745201788209646e-05,
-      "loss": 0.3183,
+      "epoch": 1.0902079504090532,
+      "grad_norm": 0.17638468742370605,
+      "learning_rate": 4.550880962569697e-05,
+      "loss": 0.4091,
       "step": 30250
     },
     {
-      "epoch": 1.06,
-      "learning_rate": 4.5743611927763106e-05,
-      "loss": 0.2891,
+      "epoch": 1.090388150070278,
+      "grad_norm": 0.20289641618728638,
+      "learning_rate": 4.55071407269123e-05,
+      "loss": 0.4385,
       "step": 30255
     },
     {
-      "epoch": 1.06,
-      "learning_rate": 4.5742021797974447e-05,
-      "loss": 0.3024,
+      "epoch": 1.0905683497315024,
+      "grad_norm": 0.16934141516685486,
+      "learning_rate": 4.5505471548721815e-05,
+      "loss": 0.4698,
       "step": 30260
     },
     {
-      "epoch": 1.06,
-      "learning_rate": 4.574043139886429e-05,
-      "loss": 0.3048,
+      "epoch": 1.0907485493927271,
+      "grad_norm": 0.1499088704586029,
+      "learning_rate": 4.550380209114824e-05,
+      "loss": 0.4283,
       "step": 30265
     },
     {
-      "epoch": 1.06,
-      "learning_rate": 4.57388407304533e-05,
-      "loss": 0.3093,
+      "epoch": 1.0909287490539519,
+      "grad_norm": 0.1858983188867569,
+      "learning_rate": 4.550213235421433e-05,
+      "loss": 0.4332,
       "step": 30270
     },
     {
-      "epoch": 1.07,
-      "learning_rate": 4.573724979276214e-05,
-      "loss": 0.2744,
+      "epoch": 1.0911089487151764,
+      "grad_norm": 0.1682644486427307,
+      "learning_rate": 4.550046233794284e-05,
+      "loss": 0.4435,
       "step": 30275
     },
     {
-      "epoch": 1.07,
-      "learning_rate": 4.573565858581145e-05,
-      "loss": 0.3016,
+      "epoch": 1.0912891483764011,
+      "grad_norm": 0.21229910850524902,
+      "learning_rate": 4.5498792042356516e-05,
+      "loss": 0.4552,
       "step": 30280
     },
     {
-      "epoch": 1.07,
-      "learning_rate": 4.57340671096219e-05,
-      "loss": 0.2876,
+      "epoch": 1.0914693480376256,
+      "grad_norm": 0.18798302114009857,
+      "learning_rate": 4.549712146747812e-05,
+      "loss": 0.4752,
       "step": 30285
     },
     {
-      "epoch": 1.07,
-      "learning_rate": 4.573247536421415e-05,
-      "loss": 0.2879,
+      "epoch": 1.0916495476988504,
+      "grad_norm": 0.192665234208107,
+      "learning_rate": 4.549545061333042e-05,
+      "loss": 0.4404,
       "step": 30290
     },
     {
-      "epoch": 1.07,
-      "learning_rate": 4.5730883349608885e-05,
-      "loss": 0.3148,
+      "epoch": 1.0918297473600749,
+      "grad_norm": 0.18173715472221375,
+      "learning_rate": 4.549377947993617e-05,
+      "loss": 0.4051,
       "step": 30295
     },
     {
-      "epoch": 1.07,
-      "learning_rate": 4.572929106582675e-05,
-      "loss": 0.3076,
+      "epoch": 1.0920099470212996,
+      "grad_norm": 0.17702946066856384,
+      "learning_rate": 4.549210806731814e-05,
+      "loss": 0.4092,
       "step": 30300
     },
     {
-      "epoch": 1.07,
-      "learning_rate": 4.572769851288843e-05,
-      "loss": 0.2992,
+      "epoch": 1.0921901466825243,
+      "grad_norm": 0.19283123314380646,
+      "learning_rate": 4.54904363754991e-05,
+      "loss": 0.4396,
       "step": 30305
     },
     {
-      "epoch": 1.07,
-      "learning_rate": 4.5726105690814616e-05,
-      "loss": 0.2713,
+      "epoch": 1.0923703463437489,
+      "grad_norm": 0.16534939408302307,
+      "learning_rate": 4.5488764404501836e-05,
+      "loss": 0.4085,
       "step": 30310
     },
     {
-      "epoch": 1.07,
-      "learning_rate": 4.5724512599625965e-05,
-      "loss": 0.311,
+      "epoch": 1.0925505460049736,
+      "grad_norm": 0.1978139728307724,
+      "learning_rate": 4.5487092154349134e-05,
+      "loss": 0.4261,
       "step": 30315
     },
     {
-      "epoch": 1.07,
-      "learning_rate": 4.572291923934319e-05,
-      "loss": 0.276,
+      "epoch": 1.092730745666198,
+      "grad_norm": 0.17723916471004486,
+      "learning_rate": 4.548541962506375e-05,
+      "loss": 0.4394,
       "step": 30320
     },
     {
-      "epoch": 1.07,
-      "learning_rate": 4.5721325609986956e-05,
-      "loss": 0.2811,
+      "epoch": 1.0929109453274228,
+      "grad_norm": 0.15324194729328156,
+      "learning_rate": 4.54837468166685e-05,
+      "loss": 0.4128,
       "step": 30325
     },
     {
-      "epoch": 1.07,
-      "learning_rate": 4.571973171157796e-05,
-      "loss": 0.3065,
+      "epoch": 1.0930911449886473,
+      "grad_norm": 0.14871060848236084,
+      "learning_rate": 4.548207372918617e-05,
+      "loss": 0.4189,
       "step": 30330
     },
     {
-      "epoch": 1.07,
-      "learning_rate": 4.5718137544136906e-05,
-      "loss": 0.2823,
+      "epoch": 1.093271344649872,
+      "grad_norm": 0.1577349752187729,
+      "learning_rate": 4.5480400362639544e-05,
+      "loss": 0.3968,
       "step": 30335
     },
     {
-      "epoch": 1.07,
-      "learning_rate": 4.57165431076845e-05,
-      "loss": 0.2673,
+      "epoch": 1.0934515443110966,
+      "grad_norm": 0.18540742993354797,
+      "learning_rate": 4.5478726717051425e-05,
+      "loss": 0.4243,
       "step": 30340
     },
     {
-      "epoch": 1.07,
-      "learning_rate": 4.571494840224143e-05,
-      "loss": 0.3034,
+      "epoch": 1.0936317439723213,
+      "grad_norm": 0.17788349092006683,
+      "learning_rate": 4.547705279244462e-05,
+      "loss": 0.4177,
       "step": 30345
     },
     {
-      "epoch": 1.07,
-      "learning_rate": 4.57133534278284e-05,
-      "loss": 0.2886,
+      "epoch": 1.093811943633546,
+      "grad_norm": 0.17169490456581116,
+      "learning_rate": 4.5475378588841945e-05,
+      "loss": 0.4525,
       "step": 30350
     },
     {
-      "epoch": 1.07,
-      "learning_rate": 4.571175818446613e-05,
-      "loss": 0.3097,
+      "epoch": 1.0939921432947706,
+      "grad_norm": 0.17133040726184845,
+      "learning_rate": 4.547370410626619e-05,
+      "loss": 0.4083,
       "step": 30355
     },
     {
-      "epoch": 1.07,
-      "learning_rate": 4.5710162672175326e-05,
-      "loss": 0.301,
+      "epoch": 1.0941723429559953,
+      "grad_norm": 0.2028420865535736,
+      "learning_rate": 4.547202934474019e-05,
+      "loss": 0.448,
       "step": 30360
     },
     {
-      "epoch": 1.07,
-      "learning_rate": 4.570856689097672e-05,
-      "loss": 0.3069,
+      "epoch": 1.0943525426172198,
+      "grad_norm": 0.2202211320400238,
+      "learning_rate": 4.5470354304286746e-05,
+      "loss": 0.4242,
       "step": 30365
     },
     {
-      "epoch": 1.07,
-      "learning_rate": 4.570697084089101e-05,
-      "loss": 0.2981,
+      "epoch": 1.0945327422784445,
+      "grad_norm": 0.19225327670574188,
+      "learning_rate": 4.546867898492869e-05,
+      "loss": 0.4575,
       "step": 30370
     },
     {
-      "epoch": 1.07,
-      "learning_rate": 4.570537452193893e-05,
-      "loss": 0.321,
+      "epoch": 1.094712941939669,
+      "grad_norm": 0.18303290009498596,
+      "learning_rate": 4.546700338668884e-05,
+      "loss": 0.4284,
       "step": 30375
     },
     {
-      "epoch": 1.07,
-      "learning_rate": 4.570377793414121e-05,
-      "loss": 0.3171,
+      "epoch": 1.0948931416008938,
+      "grad_norm": 0.19182315468788147,
+      "learning_rate": 4.546532750959004e-05,
+      "loss": 0.4241,
       "step": 30380
     },
     {
-      "epoch": 1.07,
-      "learning_rate": 4.570218107751858e-05,
-      "loss": 0.3056,
+      "epoch": 1.0950733412621185,
+      "grad_norm": 0.15474554896354675,
+      "learning_rate": 4.546365135365511e-05,
+      "loss": 0.4373,
       "step": 30385
     },
     {
-      "epoch": 1.07,
-      "learning_rate": 4.5700583952091766e-05,
-      "loss": 0.2731,
+      "epoch": 1.095253540923343,
+      "grad_norm": 0.17440365254878998,
+      "learning_rate": 4.54619749189069e-05,
+      "loss": 0.4173,
       "step": 30390
     },
     {
-      "epoch": 1.07,
-      "learning_rate": 4.569898655788152e-05,
-      "loss": 0.2996,
+      "epoch": 1.0954337405845678,
+      "grad_norm": 0.18450212478637695,
+      "learning_rate": 4.546029820536824e-05,
+      "loss": 0.4307,
       "step": 30395
     },
     {
-      "epoch": 1.07,
-      "learning_rate": 4.5697388894908565e-05,
-      "loss": 0.3146,
+      "epoch": 1.0956139402457923,
+      "grad_norm": 0.15440581738948822,
+      "learning_rate": 4.545862121306197e-05,
+      "loss": 0.4153,
       "step": 30400
     },
     {
-      "epoch": 1.07,
-      "learning_rate": 4.5695790963193666e-05,
-      "loss": 0.2897,
+      "epoch": 1.095794139907017,
+      "grad_norm": 0.19222205877304077,
+      "learning_rate": 4.5456943942010954e-05,
+      "loss": 0.4095,
       "step": 30405
     },
     {
-      "epoch": 1.07,
-      "learning_rate": 4.569419276275755e-05,
-      "loss": 0.2906,
+      "epoch": 1.0959743395682415,
+      "grad_norm": 0.24602346122264862,
+      "learning_rate": 4.545526639223804e-05,
+      "loss": 0.4373,
       "step": 30410
     },
     {
-      "epoch": 1.07,
-      "learning_rate": 4.569259429362098e-05,
-      "loss": 0.2822,
+      "epoch": 1.0961545392294663,
+      "grad_norm": 0.12822362780570984,
+      "learning_rate": 4.545358856376608e-05,
+      "loss": 0.3651,
       "step": 30415
     },
     {
-      "epoch": 1.07,
-      "learning_rate": 4.569131532486135e-05,
-      "loss": 0.3105,
+      "epoch": 1.096334738890691,
+      "grad_norm": 0.1669246107339859,
+      "learning_rate": 4.545191045661793e-05,
+      "loss": 0.383,
       "step": 30420
     },
     {
-      "epoch": 1.07,
-      "learning_rate": 4.5689716372116266e-05,
-      "loss": 0.2918,
+      "epoch": 1.0965149385519155,
+      "grad_norm": 0.18753668665885925,
+      "learning_rate": 4.5450232070816455e-05,
+      "loss": 0.4247,
       "step": 30425
     },
     {
-      "epoch": 1.07,
-      "learning_rate": 4.5688117150728856e-05,
-      "loss": 0.2943,
+      "epoch": 1.0966951382131402,
+      "grad_norm": 0.19645550847053528,
+      "learning_rate": 4.544855340638454e-05,
+      "loss": 0.3919,
       "step": 30430
     },
     {
-      "epoch": 1.07,
-      "learning_rate": 4.5686517660719875e-05,
-      "loss": 0.3087,
+      "epoch": 1.0968753378743648,
+      "grad_norm": 0.16774851083755493,
+      "learning_rate": 4.544687446334504e-05,
+      "loss": 0.4473,
       "step": 30435
     },
     {
-      "epoch": 1.07,
-      "learning_rate": 4.56849179021101e-05,
-      "loss": 0.2929,
+      "epoch": 1.0970555375355895,
+      "grad_norm": 0.16702622175216675,
+      "learning_rate": 4.544519524172083e-05,
+      "loss": 0.4206,
       "step": 30440
     },
     {
-      "epoch": 1.07,
-      "learning_rate": 4.568331787492029e-05,
-      "loss": 0.2754,
+      "epoch": 1.097235737196814,
+      "grad_norm": 0.18590982258319855,
+      "learning_rate": 4.5443515741534805e-05,
+      "loss": 0.374,
       "step": 30445
     },
     {
-      "epoch": 1.07,
-      "learning_rate": 4.568171757917123e-05,
-      "loss": 0.3089,
+      "epoch": 1.0974159368580387,
+      "grad_norm": 0.18357446789741516,
+      "learning_rate": 4.544183596280982e-05,
+      "loss": 0.4525,
       "step": 30450
     },
     {
-      "epoch": 1.07,
-      "learning_rate": 4.568011701488371e-05,
-      "loss": 0.289,
+      "epoch": 1.0975961365192632,
+      "grad_norm": 0.196320578455925,
+      "learning_rate": 4.544015590556879e-05,
+      "loss": 0.4196,
       "step": 30455
     },
     {
-      "epoch": 1.07,
-      "learning_rate": 4.567851618207848e-05,
-      "loss": 0.281,
+      "epoch": 1.097776336180488,
+      "grad_norm": 0.21362504363059998,
+      "learning_rate": 4.5438475569834585e-05,
+      "loss": 0.4115,
       "step": 30460
     },
     {
-      "epoch": 1.07,
-      "learning_rate": 4.567691508077635e-05,
-      "loss": 0.2843,
+      "epoch": 1.0979565358417127,
+      "grad_norm": 0.18946920335292816,
+      "learning_rate": 4.5436794955630115e-05,
+      "loss": 0.4346,
       "step": 30465
     },
     {
-      "epoch": 1.07,
-      "learning_rate": 4.5675313710998116e-05,
-      "loss": 0.2716,
+      "epoch": 1.0981367355029372,
+      "grad_norm": 0.1708582490682602,
+      "learning_rate": 4.5435114062978255e-05,
+      "loss": 0.4377,
       "step": 30470
     },
     {
-      "epoch": 1.07,
-      "learning_rate": 4.5673712072764544e-05,
-      "loss": 0.2786,
+      "epoch": 1.098316935164162,
+      "grad_norm": 0.1605166643857956,
+      "learning_rate": 4.543343289190194e-05,
+      "loss": 0.4377,
       "step": 30475
     },
     {
-      "epoch": 1.07,
-      "learning_rate": 4.567211016609645e-05,
-      "loss": 0.2728,
+      "epoch": 1.0984971348253865,
+      "grad_norm": 0.18001148104667664,
+      "learning_rate": 4.543175144242405e-05,
+      "loss": 0.4053,
       "step": 30480
     },
     {
-      "epoch": 1.07,
-      "learning_rate": 4.567050799101463e-05,
-      "loss": 0.2905,
+      "epoch": 1.0986773344866112,
+      "grad_norm": 0.16023088991641998,
+      "learning_rate": 4.5430069714567503e-05,
+      "loss": 0.4372,
       "step": 30485
     },
     {
-      "epoch": 1.07,
-      "learning_rate": 4.566890554753989e-05,
-      "loss": 0.2941,
+      "epoch": 1.0988575341478357,
+      "grad_norm": 0.1586974859237671,
+      "learning_rate": 4.5428387708355214e-05,
+      "loss": 0.3954,
       "step": 30490
     },
     {
-      "epoch": 1.07,
-      "learning_rate": 4.5667302835693034e-05,
-      "loss": 0.2961,
+      "epoch": 1.0990377338090604,
+      "grad_norm": 0.1917564570903778,
+      "learning_rate": 4.542670542381009e-05,
+      "loss": 0.4374,
       "step": 30495
     },
     {
-      "epoch": 1.07,
-      "learning_rate": 4.5665699855494863e-05,
-      "loss": 0.3006,
+      "epoch": 1.099217933470285,
+      "grad_norm": 0.16604667901992798,
+      "learning_rate": 4.542502286095507e-05,
+      "loss": 0.4139,
       "step": 30500
     },
     {
-      "epoch": 1.07,
-      "eval_loss": 0.2904548943042755,
-      "eval_runtime": 10.5425,
-      "eval_samples_per_second": 9.485,
-      "eval_steps_per_second": 9.485,
+      "epoch": 1.099217933470285,
+      "eval_loss": 0.4521316885948181,
+      "eval_runtime": 3.5361,
+      "eval_samples_per_second": 28.28,
+      "eval_steps_per_second": 7.07,
       "step": 30500
     },
     {
-      "epoch": 1.07,
-      "learning_rate": 4.56640966069662e-05,
-      "loss": 0.2878,
+      "epoch": 1.0993981331315097,
+      "grad_norm": 0.16811847686767578,
+      "learning_rate": 4.542334001981307e-05,
+      "loss": 0.4095,
       "step": 30505
     },
     {
-      "epoch": 1.07,
-      "learning_rate": 4.566249309012787e-05,
-      "loss": 0.3108,
+      "epoch": 1.0995783327927344,
+      "grad_norm": 0.1503792107105255,
+      "learning_rate": 4.5421656900407e-05,
+      "loss": 0.442,
       "step": 30510
     },
     {
-      "epoch": 1.07,
-      "learning_rate": 4.566088930500068e-05,
-      "loss": 0.3057,
+      "epoch": 1.099758532453959,
+      "grad_norm": 0.18034709990024567,
+      "learning_rate": 4.5419973502759816e-05,
+      "loss": 0.4725,
       "step": 30515
     },
     {
-      "epoch": 1.07,
-      "learning_rate": 4.565928525160545e-05,
-      "loss": 0.3067,
+      "epoch": 1.0999387321151837,
+      "grad_norm": 0.16529808938503265,
+      "learning_rate": 4.541828982689445e-05,
+      "loss": 0.3991,
       "step": 30520
     },
     {
-      "epoch": 1.07,
-      "learning_rate": 4.565768092996302e-05,
-      "loss": 0.2811,
+      "epoch": 1.1001189317764082,
+      "grad_norm": 0.17672580480575562,
+      "learning_rate": 4.541660587283384e-05,
+      "loss": 0.4159,
       "step": 30525
     },
     {
-      "epoch": 1.07,
-      "learning_rate": 4.5656076340094224e-05,
-      "loss": 0.2812,
+      "epoch": 1.100299131437633,
+      "grad_norm": 0.17020630836486816,
+      "learning_rate": 4.541492164060092e-05,
+      "loss": 0.4095,
       "step": 30530
     },
     {
-      "epoch": 1.07,
-      "learning_rate": 4.5654471482019876e-05,
-      "loss": 0.2949,
+      "epoch": 1.1004793310988576,
+      "grad_norm": 0.19656726717948914,
+      "learning_rate": 4.541323713021865e-05,
+      "loss": 0.4345,
       "step": 30535
     },
     {
-      "epoch": 1.07,
-      "learning_rate": 4.565286635576084e-05,
-      "loss": 0.2894,
+      "epoch": 1.1006595307600822,
+      "grad_norm": 0.15002837777137756,
+      "learning_rate": 4.541155234170997e-05,
+      "loss": 0.4065,
       "step": 30540
     },
     {
-      "epoch": 1.07,
-      "learning_rate": 4.565126096133794e-05,
-      "loss": 0.288,
+      "epoch": 1.100839730421307,
+      "grad_norm": 0.15486060082912445,
+      "learning_rate": 4.540986727509785e-05,
+      "loss": 0.3887,
       "step": 30545
     },
     {
-      "epoch": 1.07,
-      "learning_rate": 4.564965529877203e-05,
-      "loss": 0.3034,
+      "epoch": 1.1010199300825314,
+      "grad_norm": 0.16535469889640808,
+      "learning_rate": 4.540818193040523e-05,
+      "loss": 0.395,
       "step": 30550
     },
     {
-      "epoch": 1.08,
-      "learning_rate": 4.5648049368083956e-05,
-      "loss": 0.2972,
+      "epoch": 1.1012001297437561,
+      "grad_norm": 0.14959602057933807,
+      "learning_rate": 4.54064963076551e-05,
+      "loss": 0.4287,
       "step": 30555
     },
     {
-      "epoch": 1.08,
-      "learning_rate": 4.564644316929456e-05,
-      "loss": 0.2935,
+      "epoch": 1.1013803294049807,
+      "grad_norm": 0.19328902661800385,
+      "learning_rate": 4.5404810406870396e-05,
+      "loss": 0.4119,
       "step": 30560
     },
     {
-      "epoch": 1.08,
-      "learning_rate": 4.564483670242471e-05,
-      "loss": 0.2799,
+      "epoch": 1.1015605290662054,
+      "grad_norm": 0.18112468719482422,
+      "learning_rate": 4.54031242280741e-05,
+      "loss": 0.4082,
       "step": 30565
     },
     {
-      "epoch": 1.08,
-      "learning_rate": 4.564322996749526e-05,
-      "loss": 0.2615,
+      "epoch": 1.10174072872743,
+      "grad_norm": 0.2009975016117096,
+      "learning_rate": 4.540143777128919e-05,
+      "loss": 0.4164,
       "step": 30570
     },
     {
-      "epoch": 1.08,
-      "learning_rate": 4.5641622964527076e-05,
-      "loss": 0.3086,
+      "epoch": 1.1019209283886546,
+      "grad_norm": 0.15006959438323975,
+      "learning_rate": 4.539975103653864e-05,
+      "loss": 0.37,
       "step": 30575
     },
     {
-      "epoch": 1.08,
-      "learning_rate": 4.564001569354102e-05,
-      "loss": 0.3051,
+      "epoch": 1.1021011280498794,
+      "grad_norm": 0.15584491193294525,
+      "learning_rate": 4.5398064023845424e-05,
+      "loss": 0.4106,
       "step": 30580
     },
     {
-      "epoch": 1.08,
-      "learning_rate": 4.563840815455797e-05,
-      "loss": 0.2889,
+      "epoch": 1.1022813277111039,
+      "grad_norm": 0.17378970980644226,
+      "learning_rate": 4.539637673323255e-05,
+      "loss": 0.4463,
       "step": 30585
     },
     {
-      "epoch": 1.08,
-      "learning_rate": 4.563680034759877e-05,
-      "loss": 0.2883,
+      "epoch": 1.1024615273723286,
+      "grad_norm": 0.21250377595424652,
+      "learning_rate": 4.539468916472298e-05,
+      "loss": 0.4063,
       "step": 30590
     },
     {
-      "epoch": 1.08,
-      "learning_rate": 4.563519227268433e-05,
-      "loss": 0.2804,
+      "epoch": 1.1026417270335531,
+      "grad_norm": 0.19015663862228394,
+      "learning_rate": 4.539300131833972e-05,
+      "loss": 0.4417,
       "step": 30595
     },
     {
-      "epoch": 1.08,
-      "learning_rate": 4.563358392983553e-05,
-      "loss": 0.2779,
+      "epoch": 1.1028219266947779,
+      "grad_norm": 0.15822429955005646,
+      "learning_rate": 4.539131319410577e-05,
+      "loss": 0.4316,
       "step": 30600
     },
     {
-      "epoch": 1.08,
-      "learning_rate": 4.563197531907323e-05,
-      "loss": 0.2925,
+      "epoch": 1.1030021263560024,
+      "grad_norm": 0.17139464616775513,
+      "learning_rate": 4.538962479204412e-05,
+      "loss": 0.4495,
       "step": 30605
     },
     {
-      "epoch": 1.08,
-      "learning_rate": 4.563036644041833e-05,
-      "loss": 0.312,
+      "epoch": 1.103182326017227,
+      "grad_norm": 0.14621873199939728,
+      "learning_rate": 4.538793611217778e-05,
+      "loss": 0.4219,
       "step": 30610
     },
     {
-      "epoch": 1.08,
-      "learning_rate": 4.56287572938917e-05,
-      "loss": 0.3245,
+      "epoch": 1.1033625256784516,
+      "grad_norm": 0.15971679985523224,
+      "learning_rate": 4.538624715452976e-05,
+      "loss": 0.4193,
       "step": 30615
     },
     {
-      "epoch": 1.08,
-      "learning_rate": 4.562714787951427e-05,
-      "loss": 0.303,
+      "epoch": 1.1035427253396763,
+      "grad_norm": 0.16001123189926147,
+      "learning_rate": 4.538455791912307e-05,
+      "loss": 0.4119,
       "step": 30620
     },
     {
-      "epoch": 1.08,
-      "learning_rate": 4.5625538197306916e-05,
-      "loss": 0.2757,
+      "epoch": 1.103722925000901,
+      "grad_norm": 0.17780780792236328,
+      "learning_rate": 4.5382868405980724e-05,
+      "loss": 0.4474,
       "step": 30625
     },
     {
-      "epoch": 1.08,
-      "learning_rate": 4.5623928247290536e-05,
-      "loss": 0.2908,
+      "epoch": 1.1039031246621256,
+      "grad_norm": 0.17984482645988464,
+      "learning_rate": 4.5381178615125746e-05,
+      "loss": 0.4399,
       "step": 30630
     },
     {
-      "epoch": 1.08,
-      "learning_rate": 4.5622318029486045e-05,
-      "loss": 0.2996,
+      "epoch": 1.1040833243233503,
+      "grad_norm": 0.1892779916524887,
+      "learning_rate": 4.537948854658115e-05,
+      "loss": 0.4361,
       "step": 30635
     },
     {
-      "epoch": 1.08,
-      "learning_rate": 4.562070754391434e-05,
-      "loss": 0.2964,
+      "epoch": 1.1042635239845748,
+      "grad_norm": 0.13633422553539276,
+      "learning_rate": 4.537779820036997e-05,
+      "loss": 0.4032,
       "step": 30640
     },
     {
-      "epoch": 1.08,
-      "learning_rate": 4.561909679059635e-05,
-      "loss": 0.2906,
+      "epoch": 1.1044437236457996,
+      "grad_norm": 0.1995527446269989,
+      "learning_rate": 4.5376107576515235e-05,
+      "loss": 0.4478,
       "step": 30645
     },
     {
-      "epoch": 1.08,
-      "learning_rate": 4.561748576955297e-05,
-      "loss": 0.2945,
+      "epoch": 1.104623923307024,
+      "grad_norm": 0.21352799236774445,
+      "learning_rate": 4.537441667503998e-05,
+      "loss": 0.4282,
       "step": 30650
     },
     {
-      "epoch": 1.08,
-      "learning_rate": 4.5615874480805114e-05,
-      "loss": 0.2942,
+      "epoch": 1.1048041229682488,
+      "grad_norm": 0.2059534788131714,
+      "learning_rate": 4.537272549596724e-05,
+      "loss": 0.4291,
       "step": 30655
     },
     {
-      "epoch": 1.08,
-      "learning_rate": 4.561426292437372e-05,
-      "loss": 0.2847,
+      "epoch": 1.1049843226294735,
+      "grad_norm": 0.17112842202186584,
+      "learning_rate": 4.5371034039320065e-05,
+      "loss": 0.3752,
       "step": 30660
     },
     {
-      "epoch": 1.08,
-      "learning_rate": 4.561265110027971e-05,
-      "loss": 0.2865,
+      "epoch": 1.105164522290698,
+      "grad_norm": 0.18487350642681122,
+      "learning_rate": 4.536934230512149e-05,
+      "loss": 0.4198,
       "step": 30665
     },
     {
-      "epoch": 1.08,
-      "learning_rate": 4.561103900854401e-05,
-      "loss": 0.2858,
+      "epoch": 1.1053447219519228,
+      "grad_norm": 0.20532426238059998,
+      "learning_rate": 4.5367650293394574e-05,
+      "loss": 0.4549,
       "step": 30670
     },
     {
-      "epoch": 1.08,
-      "learning_rate": 4.560942664918755e-05,
-      "loss": 0.2842,
+      "epoch": 1.1055249216131473,
+      "grad_norm": 0.1958533376455307,
+      "learning_rate": 4.536595800416236e-05,
+      "loss": 0.424,
       "step": 30675
     },
     {
-      "epoch": 1.08,
-      "learning_rate": 4.5607814022231266e-05,
-      "loss": 0.3133,
+      "epoch": 1.105705121274372,
+      "grad_norm": 0.18324759602546692,
+      "learning_rate": 4.5364265437447915e-05,
+      "loss": 0.4424,
       "step": 30680
     },
     {
-      "epoch": 1.08,
-      "learning_rate": 4.5606201127696105e-05,
-      "loss": 0.2843,
+      "epoch": 1.1058853209355965,
+      "grad_norm": 0.16815263032913208,
+      "learning_rate": 4.536257259327429e-05,
+      "loss": 0.3943,
       "step": 30685
     },
     {
-      "epoch": 1.08,
-      "learning_rate": 4.5604587965602996e-05,
-      "loss": 0.3128,
+      "epoch": 1.1060655205968213,
+      "grad_norm": 0.16847245395183563,
+      "learning_rate": 4.536087947166456e-05,
+      "loss": 0.4235,
       "step": 30690
     },
     {
-      "epoch": 1.08,
-      "learning_rate": 4.560297453597288e-05,
-      "loss": 0.2892,
+      "epoch": 1.106245720258046,
+      "grad_norm": 0.18886929750442505,
+      "learning_rate": 4.5359186072641796e-05,
+      "loss": 0.431,
       "step": 30695
     },
     {
-      "epoch": 1.08,
-      "learning_rate": 4.560136083882674e-05,
-      "loss": 0.3015,
+      "epoch": 1.1064259199192705,
+      "grad_norm": 0.19722402095794678,
+      "learning_rate": 4.535749239622906e-05,
+      "loss": 0.4551,
       "step": 30700
     },
     {
-      "epoch": 1.08,
-      "learning_rate": 4.559974687418549e-05,
-      "loss": 0.3036,
+      "epoch": 1.1066061195804953,
+      "grad_norm": 0.1684960126876831,
+      "learning_rate": 4.535579844244943e-05,
+      "loss": 0.4292,
       "step": 30705
     },
     {
-      "epoch": 1.08,
-      "learning_rate": 4.559813264207011e-05,
-      "loss": 0.2806,
+      "epoch": 1.1067863192417198,
+      "grad_norm": 0.16538777947425842,
+      "learning_rate": 4.535410421132598e-05,
+      "loss": 0.4574,
       "step": 30710
     },
     {
-      "epoch": 1.08,
-      "learning_rate": 4.5596518142501556e-05,
-      "loss": 0.289,
+      "epoch": 1.1069665189029445,
+      "grad_norm": 0.21061739325523376,
+      "learning_rate": 4.535240970288181e-05,
+      "loss": 0.402,
       "step": 30715
     },
     {
-      "epoch": 1.08,
-      "learning_rate": 4.5594903375500786e-05,
-      "loss": 0.2716,
+      "epoch": 1.107146718564169,
+      "grad_norm": 0.18841953575611115,
+      "learning_rate": 4.535071491713999e-05,
+      "loss": 0.4477,
       "step": 30720
     },
     {
-      "epoch": 1.08,
-      "learning_rate": 4.559328834108876e-05,
-      "loss": 0.2837,
+      "epoch": 1.1073269182253938,
+      "grad_norm": 0.15210363268852234,
+      "learning_rate": 4.534901985412363e-05,
+      "loss": 0.4299,
       "step": 30725
     },
     {
-      "epoch": 1.08,
-      "learning_rate": 4.559167303928646e-05,
-      "loss": 0.2884,
+      "epoch": 1.1075071178866183,
+      "grad_norm": 0.18912887573242188,
+      "learning_rate": 4.53473245138558e-05,
+      "loss": 0.4201,
       "step": 30730
     },
     {
-      "epoch": 1.08,
-      "learning_rate": 4.559005747011486e-05,
-      "loss": 0.2818,
+      "epoch": 1.107687317547843,
+      "grad_norm": 0.1371183693408966,
+      "learning_rate": 4.5345628896359625e-05,
+      "loss": 0.4116,
       "step": 30735
     },
     {
-      "epoch": 1.08,
-      "learning_rate": 4.5588441633594926e-05,
-      "loss": 0.304,
+      "epoch": 1.1078675172090677,
+      "grad_norm": 0.15921291708946228,
+      "learning_rate": 4.5343933001658194e-05,
+      "loss": 0.4222,
       "step": 30740
     },
     {
-      "epoch": 1.08,
-      "learning_rate": 4.558682552974765e-05,
-      "loss": 0.306,
+      "epoch": 1.1080477168702922,
+      "grad_norm": 0.23919863998889923,
+      "learning_rate": 4.5342236829774617e-05,
+      "loss": 0.453,
       "step": 30745
     },
     {
-      "epoch": 1.08,
-      "learning_rate": 4.5585209158594014e-05,
-      "loss": 0.2925,
+      "epoch": 1.108227916531517,
+      "grad_norm": 0.2175007313489914,
+      "learning_rate": 4.534054038073199e-05,
+      "loss": 0.4146,
       "step": 30750
     },
     {
-      "epoch": 1.08,
-      "learning_rate": 4.558359252015499e-05,
-      "loss": 0.2952,
+      "epoch": 1.1084081161927415,
+      "grad_norm": 0.16294798254966736,
+      "learning_rate": 4.533884365455345e-05,
+      "loss": 0.3633,
       "step": 30755
     },
     {
-      "epoch": 1.08,
-      "learning_rate": 4.55819756144516e-05,
-      "loss": 0.2705,
+      "epoch": 1.1085883158539662,
+      "grad_norm": 0.15248903632164001,
+      "learning_rate": 4.5337146651262094e-05,
+      "loss": 0.4316,
       "step": 30760
     },
     {
-      "epoch": 1.08,
-      "learning_rate": 4.558035844150481e-05,
-      "loss": 0.2831,
+      "epoch": 1.1087685155151907,
+      "grad_norm": 0.18344025313854218,
+      "learning_rate": 4.533544937088106e-05,
+      "loss": 0.3941,
       "step": 30765
     },
     {
-      "epoch": 1.08,
-      "learning_rate": 4.557874100133563e-05,
-      "loss": 0.2785,
+      "epoch": 1.1089487151764155,
+      "grad_norm": 0.2048654705286026,
+      "learning_rate": 4.533375181343346e-05,
+      "loss": 0.4528,
       "step": 30770
     },
     {
-      "epoch": 1.08,
-      "learning_rate": 4.557712329396507e-05,
-      "loss": 0.3199,
+      "epoch": 1.1091289148376402,
+      "grad_norm": 0.20084497332572937,
+      "learning_rate": 4.5332053978942436e-05,
+      "loss": 0.4366,
       "step": 30775
     },
     {
-      "epoch": 1.08,
-      "learning_rate": 4.5575505319414117e-05,
-      "loss": 0.3042,
+      "epoch": 1.1093091144988647,
+      "grad_norm": 0.13244780898094177,
+      "learning_rate": 4.5330355867431106e-05,
+      "loss": 0.4239,
       "step": 30780
     },
     {
-      "epoch": 1.08,
-      "learning_rate": 4.5573887077703786e-05,
-      "loss": 0.291,
+      "epoch": 1.1094893141600894,
+      "grad_norm": 0.21777132153511047,
+      "learning_rate": 4.532865747892261e-05,
+      "loss": 0.4454,
       "step": 30785
     },
     {
-      "epoch": 1.08,
-      "learning_rate": 4.557226856885509e-05,
-      "loss": 0.2976,
+      "epoch": 1.109669513821314,
+      "grad_norm": 0.19449090957641602,
+      "learning_rate": 4.53269588134401e-05,
+      "loss": 0.4324,
       "step": 30790
     },
     {
-      "epoch": 1.08,
-      "learning_rate": 4.557064979288905e-05,
-      "loss": 0.3074,
+      "epoch": 1.1098497134825387,
+      "grad_norm": 0.16358676552772522,
+      "learning_rate": 4.532525987100671e-05,
+      "loss": 0.4083,
       "step": 30795
     },
     {
-      "epoch": 1.08,
-      "learning_rate": 4.5569030749826666e-05,
-      "loss": 0.28,
+      "epoch": 1.1100299131437632,
+      "grad_norm": 0.1827852874994278,
+      "learning_rate": 4.532356065164558e-05,
+      "loss": 0.411,
       "step": 30800
     },
     {
-      "epoch": 1.08,
-      "learning_rate": 4.556741143968899e-05,
-      "loss": 0.2845,
+      "epoch": 1.110210112804988,
+      "grad_norm": 0.16371290385723114,
+      "learning_rate": 4.5321861155379884e-05,
+      "loss": 0.4314,
       "step": 30805
     },
     {
-      "epoch": 1.08,
-      "learning_rate": 4.556579186249702e-05,
-      "loss": 0.303,
+      "epoch": 1.1103903124662127,
+      "grad_norm": 0.18602146208286285,
+      "learning_rate": 4.532016138223276e-05,
+      "loss": 0.4158,
       "step": 30810
     },
     {
-      "epoch": 1.08,
-      "learning_rate": 4.55641720182718e-05,
-      "loss": 0.3016,
+      "epoch": 1.1105705121274372,
+      "grad_norm": 0.14630118012428284,
+      "learning_rate": 4.5318461332227365e-05,
+      "loss": 0.3876,
       "step": 30815
     },
     {
-      "epoch": 1.08,
-      "learning_rate": 4.556255190703436e-05,
-      "loss": 0.2929,
+      "epoch": 1.110750711788662,
+      "grad_norm": 0.1606942117214203,
+      "learning_rate": 4.531676100538688e-05,
+      "loss": 0.4428,
       "step": 30820
     },
     {
-      "epoch": 1.08,
-      "learning_rate": 4.556093152880573e-05,
-      "loss": 0.28,
+      "epoch": 1.1109309114498864,
+      "grad_norm": 0.16617447137832642,
+      "learning_rate": 4.5315060401734445e-05,
+      "loss": 0.4363,
       "step": 30825
     },
     {
-      "epoch": 1.08,
-      "learning_rate": 4.555931088360695e-05,
-      "loss": 0.3086,
+      "epoch": 1.1111111111111112,
+      "grad_norm": 0.1629336029291153,
+      "learning_rate": 4.5313359521293254e-05,
+      "loss": 0.403,
       "step": 30830
     },
     {
-      "epoch": 1.08,
-      "learning_rate": 4.555768997145908e-05,
-      "loss": 0.3224,
+      "epoch": 1.1112913107723357,
+      "grad_norm": 0.1905163824558258,
+      "learning_rate": 4.5311658364086474e-05,
+      "loss": 0.4358,
       "step": 30835
     },
     {
-      "epoch": 1.09,
-      "learning_rate": 4.555606879238314e-05,
-      "loss": 0.3191,
+      "epoch": 1.1114715104335604,
+      "grad_norm": 0.17347170412540436,
+      "learning_rate": 4.531029723906539e-05,
+      "loss": 0.4344,
       "step": 30840
     },
     {
-      "epoch": 1.09,
-      "learning_rate": 4.555444734640021e-05,
-      "loss": 0.2812,
+      "epoch": 1.111651710094785,
+      "grad_norm": 0.1710389107465744,
+      "learning_rate": 4.530859558373896e-05,
+      "loss": 0.4036,
       "step": 30845
     },
     {
-      "epoch": 1.09,
-      "learning_rate": 4.555282563353132e-05,
-      "loss": 0.2703,
+      "epoch": 1.1118319097560097,
+      "grad_norm": 0.15823177993297577,
+      "learning_rate": 4.530689365171184e-05,
+      "loss": 0.3619,
       "step": 30850
     },
     {
-      "epoch": 1.09,
-      "learning_rate": 4.5551203653797525e-05,
-      "loss": 0.2794,
+      "epoch": 1.1120121094172344,
+      "grad_norm": 0.17477105557918549,
+      "learning_rate": 4.530519144300722e-05,
+      "loss": 0.4299,
       "step": 30855
     },
     {
-      "epoch": 1.09,
-      "learning_rate": 4.55495814072199e-05,
-      "loss": 0.287,
+      "epoch": 1.112192309078459,
+      "grad_norm": 0.15027157962322235,
+      "learning_rate": 4.530348895764831e-05,
+      "loss": 0.4332,
       "step": 30860
     },
     {
-      "epoch": 1.09,
-      "learning_rate": 4.554795889381951e-05,
-      "loss": 0.2831,
+      "epoch": 1.1123725087396836,
+      "grad_norm": 0.1768927425146103,
+      "learning_rate": 4.530178619565829e-05,
+      "loss": 0.434,
       "step": 30865
     },
     {
-      "epoch": 1.09,
-      "learning_rate": 4.554633611361741e-05,
-      "loss": 0.2908,
+      "epoch": 1.1125527084009081,
+      "grad_norm": 0.17332357168197632,
+      "learning_rate": 4.5300083157060356e-05,
+      "loss": 0.4292,
       "step": 30870
     },
     {
-      "epoch": 1.09,
-      "learning_rate": 4.554471306663467e-05,
-      "loss": 0.2641,
+      "epoch": 1.1127329080621329,
+      "grad_norm": 0.22046640515327454,
+      "learning_rate": 4.529837984187773e-05,
+      "loss": 0.4375,
       "step": 30875
     },
     {
-      "epoch": 1.09,
-      "learning_rate": 4.5543089752892375e-05,
-      "loss": 0.3155,
+      "epoch": 1.1129131077233574,
+      "grad_norm": 0.15392176806926727,
+      "learning_rate": 4.529667625013361e-05,
+      "loss": 0.4253,
       "step": 30880
     },
     {
-      "epoch": 1.09,
-      "learning_rate": 4.554146617241161e-05,
-      "loss": 0.2905,
+      "epoch": 1.1130933073845821,
+      "grad_norm": 0.1741206794977188,
+      "learning_rate": 4.52949723818512e-05,
+      "loss": 0.3929,
       "step": 30885
     },
     {
-      "epoch": 1.09,
-      "learning_rate": 4.553984232521342e-05,
-      "loss": 0.3115,
+      "epoch": 1.1132735070458069,
+      "grad_norm": 0.1908791959285736,
+      "learning_rate": 4.529326823705372e-05,
+      "loss": 0.4403,
       "step": 30890
     },
     {
-      "epoch": 1.09,
-      "learning_rate": 4.553821821131893e-05,
-      "loss": 0.3043,
+      "epoch": 1.1134537067070314,
+      "grad_norm": 0.20315563678741455,
+      "learning_rate": 4.5291563815764384e-05,
+      "loss": 0.4401,
       "step": 30895
     },
     {
-      "epoch": 1.09,
-      "learning_rate": 4.553659383074921e-05,
-      "loss": 0.2819,
+      "epoch": 1.113633906368256,
+      "grad_norm": 0.16982267796993256,
+      "learning_rate": 4.528985911800643e-05,
+      "loss": 0.4372,
       "step": 30900
     },
     {
-      "epoch": 1.09,
-      "learning_rate": 4.5534969183525354e-05,
-      "loss": 0.2957,
+      "epoch": 1.1138141060294806,
+      "grad_norm": 0.1527600884437561,
+      "learning_rate": 4.5288154143803066e-05,
+      "loss": 0.3923,
       "step": 30905
     },
     {
-      "epoch": 1.09,
-      "learning_rate": 4.553334426966846e-05,
-      "loss": 0.2755,
+      "epoch": 1.1139943056907053,
+      "grad_norm": 0.17226542532444,
+      "learning_rate": 4.528644889317753e-05,
+      "loss": 0.4158,
       "step": 30910
     },
     {
-      "epoch": 1.09,
-      "learning_rate": 4.553171908919961e-05,
-      "loss": 0.2928,
+      "epoch": 1.1141745053519299,
+      "grad_norm": 0.178422212600708,
+      "learning_rate": 4.528474336615306e-05,
+      "loss": 0.4268,
       "step": 30915
     },
     {
-      "epoch": 1.09,
-      "learning_rate": 4.5530093642139927e-05,
-      "loss": 0.2944,
+      "epoch": 1.1143547050131546,
+      "grad_norm": 0.1679055094718933,
+      "learning_rate": 4.528303756275288e-05,
+      "loss": 0.4216,
       "step": 30920
     },
     {
-      "epoch": 1.09,
-      "learning_rate": 4.55284679285105e-05,
-      "loss": 0.2891,
+      "epoch": 1.1145349046743793,
+      "grad_norm": 0.21220412850379944,
+      "learning_rate": 4.528133148300026e-05,
+      "loss": 0.4276,
       "step": 30925
     },
     {
-      "epoch": 1.09,
-      "learning_rate": 4.5526841948332455e-05,
-      "loss": 0.2939,
+      "epoch": 1.1147151043356038,
+      "grad_norm": 0.16649261116981506,
+      "learning_rate": 4.52796251269184e-05,
+      "loss": 0.4486,
       "step": 30930
     },
     {
-      "epoch": 1.09,
-      "learning_rate": 4.55252157016269e-05,
-      "loss": 0.2685,
+      "epoch": 1.1148953039968286,
+      "grad_norm": 0.19539640843868256,
+      "learning_rate": 4.527791849453059e-05,
+      "loss": 0.4124,
       "step": 30935
     },
     {
-      "epoch": 1.09,
-      "learning_rate": 4.5523589188414926e-05,
-      "loss": 0.2761,
+      "epoch": 1.115075503658053,
+      "grad_norm": 0.17714886367321014,
+      "learning_rate": 4.5276211585860064e-05,
+      "loss": 0.4119,
       "step": 30940
     },
     {
-      "epoch": 1.09,
-      "learning_rate": 4.552196240871769e-05,
-      "loss": 0.2695,
+      "epoch": 1.1152557033192778,
+      "grad_norm": 0.17066708207130432,
+      "learning_rate": 4.527450440093008e-05,
+      "loss": 0.4194,
       "step": 30945
     },
     {
-      "epoch": 1.09,
-      "learning_rate": 4.5520335362556286e-05,
-      "loss": 0.2848,
+      "epoch": 1.1154359029805023,
+      "grad_norm": 0.1813109666109085,
+      "learning_rate": 4.527279693976389e-05,
+      "loss": 0.4032,
       "step": 30950
     },
     {
-      "epoch": 1.09,
-      "learning_rate": 4.551870804995186e-05,
-      "loss": 0.2786,
+      "epoch": 1.115616102641727,
+      "grad_norm": 0.1931227445602417,
+      "learning_rate": 4.527108920238478e-05,
+      "loss": 0.4249,
       "step": 30955
     },
     {
-      "epoch": 1.09,
-      "learning_rate": 4.551708047092552e-05,
-      "loss": 0.2955,
+      "epoch": 1.1157963023029516,
+      "grad_norm": 0.16914038360118866,
+      "learning_rate": 4.5269381188815996e-05,
+      "loss": 0.381,
       "step": 30960
     },
     {
-      "epoch": 1.09,
-      "learning_rate": 4.551545262549842e-05,
-      "loss": 0.2877,
+      "epoch": 1.1159765019641763,
+      "grad_norm": 0.1841052621603012,
+      "learning_rate": 4.526767289908083e-05,
+      "loss": 0.4565,
       "step": 30965
     },
     {
-      "epoch": 1.09,
-      "learning_rate": 4.5513824513691686e-05,
-      "loss": 0.2801,
+      "epoch": 1.116156701625401,
+      "grad_norm": 0.1776377409696579,
+      "learning_rate": 4.5265964333202526e-05,
+      "loss": 0.4371,
       "step": 30970
     },
     {
-      "epoch": 1.09,
-      "learning_rate": 4.551219613552646e-05,
-      "loss": 0.2774,
+      "epoch": 1.1163369012866255,
+      "grad_norm": 0.17121796309947968,
+      "learning_rate": 4.526425549120439e-05,
+      "loss": 0.4383,
       "step": 30975
     },
     {
-      "epoch": 1.09,
-      "learning_rate": 4.551056749102389e-05,
-      "loss": 0.2859,
+      "epoch": 1.1165171009478503,
+      "grad_norm": 0.16264458000659943,
+      "learning_rate": 4.526254637310971e-05,
+      "loss": 0.4006,
       "step": 30980
     },
     {
-      "epoch": 1.09,
-      "learning_rate": 4.550893858020511e-05,
-      "loss": 0.2759,
+      "epoch": 1.1166973006090748,
+      "grad_norm": 0.19844268262386322,
+      "learning_rate": 4.526083697894173e-05,
+      "loss": 0.4223,
       "step": 30985
     },
     {
-      "epoch": 1.09,
-      "learning_rate": 4.550730940309128e-05,
-      "loss": 0.3017,
+      "epoch": 1.1168775002702995,
+      "grad_norm": 0.19541612267494202,
+      "learning_rate": 4.525912730872379e-05,
+      "loss": 0.42,
       "step": 30990
     },
     {
-      "epoch": 1.09,
-      "learning_rate": 4.5505679959703565e-05,
-      "loss": 0.3095,
+      "epoch": 1.117057699931524,
+      "grad_norm": 0.19237811863422394,
+      "learning_rate": 4.525741736247916e-05,
+      "loss": 0.3994,
       "step": 30995
     },
     {
-      "epoch": 1.09,
-      "learning_rate": 4.550405025006309e-05,
-      "loss": 0.2957,
+      "epoch": 1.1172378995927488,
+      "grad_norm": 0.15686160326004028,
+      "learning_rate": 4.5255707140231136e-05,
+      "loss": 0.4275,
       "step": 31000
     },
     {
-      "epoch": 1.09,
-      "eval_loss": 0.2897602617740631,
-      "eval_runtime": 10.5439,
-      "eval_samples_per_second": 9.484,
-      "eval_steps_per_second": 9.484,
+      "epoch": 1.1172378995927488,
+      "eval_loss": 0.4531842768192291,
+      "eval_runtime": 3.5417,
+      "eval_samples_per_second": 28.235,
+      "eval_steps_per_second": 7.059,
       "step": 31000
     },
     {
-      "epoch": 1.09,
-      "learning_rate": 4.550242027419105e-05,
-      "loss": 0.285,
+      "epoch": 1.1174180992539733,
+      "grad_norm": 0.16874189674854279,
+      "learning_rate": 4.5253996642003025e-05,
+      "loss": 0.4781,
       "step": 31005
     },
     {
-      "epoch": 1.09,
-      "learning_rate": 4.550079003210859e-05,
-      "loss": 0.2983,
+      "epoch": 1.117598298915198,
+      "grad_norm": 0.17031274735927582,
+      "learning_rate": 4.525228586781813e-05,
+      "loss": 0.4237,
       "step": 31010
     },
     {
-      "epoch": 1.09,
-      "learning_rate": 4.5499159523836875e-05,
-      "loss": 0.2819,
+      "epoch": 1.1177784985764228,
+      "grad_norm": 0.19716662168502808,
+      "learning_rate": 4.525057481769976e-05,
+      "loss": 0.415,
       "step": 31015
     },
     {
-      "epoch": 1.09,
-      "learning_rate": 4.549752874939709e-05,
-      "loss": 0.3313,
+      "epoch": 1.1179586982376473,
+      "grad_norm": 0.15895715355873108,
+      "learning_rate": 4.524886349167124e-05,
+      "loss": 0.3885,
       "step": 31020
     },
     {
-      "epoch": 1.09,
-      "learning_rate": 4.5495897708810395e-05,
-      "loss": 0.266,
+      "epoch": 1.118138897898872,
+      "grad_norm": 0.1836315244436264,
+      "learning_rate": 4.5247151889755855e-05,
+      "loss": 0.4201,
       "step": 31025
     },
     {
-      "epoch": 1.09,
-      "learning_rate": 4.549426640209799e-05,
-      "loss": 0.3073,
+      "epoch": 1.1183190975600965,
+      "grad_norm": 0.17792055010795593,
+      "learning_rate": 4.5245440011976966e-05,
+      "loss": 0.4348,
       "step": 31030
     },
     {
-      "epoch": 1.09,
-      "learning_rate": 4.5492634829281026e-05,
-      "loss": 0.3227,
+      "epoch": 1.1184992972213212,
+      "grad_norm": 0.17918603122234344,
+      "learning_rate": 4.524372785835787e-05,
+      "loss": 0.4039,
       "step": 31035
     },
     {
-      "epoch": 1.09,
-      "learning_rate": 4.549100299038072e-05,
-      "loss": 0.2894,
+      "epoch": 1.118679496882546,
+      "grad_norm": 0.1892392635345459,
+      "learning_rate": 4.524201542892189e-05,
+      "loss": 0.4318,
       "step": 31040
     },
     {
-      "epoch": 1.09,
-      "learning_rate": 4.548937088541823e-05,
-      "loss": 0.2956,
+      "epoch": 1.1188596965437705,
+      "grad_norm": 0.17238296568393707,
+      "learning_rate": 4.524030272369238e-05,
+      "loss": 0.4143,
       "step": 31045
     },
     {
-      "epoch": 1.09,
-      "learning_rate": 4.548773851441477e-05,
-      "loss": 0.2959,
+      "epoch": 1.1190398962049952,
+      "grad_norm": 0.1613350510597229,
+      "learning_rate": 4.523858974269266e-05,
+      "loss": 0.4531,
       "step": 31050
     },
     {
-      "epoch": 1.09,
-      "learning_rate": 4.548610587739152e-05,
-      "loss": 0.2917,
+      "epoch": 1.1192200958662197,
+      "grad_norm": 0.20245467126369476,
+      "learning_rate": 4.5236876485946076e-05,
+      "loss": 0.4655,
       "step": 31055
     },
     {
-      "epoch": 1.09,
-      "learning_rate": 4.54844729743697e-05,
-      "loss": 0.3038,
+      "epoch": 1.1194002955274445,
+      "grad_norm": 0.16436737775802612,
+      "learning_rate": 4.5235162953475966e-05,
+      "loss": 0.4167,
       "step": 31060
     },
     {
-      "epoch": 1.09,
-      "learning_rate": 4.548283980537048e-05,
-      "loss": 0.2948,
+      "epoch": 1.119580495188669,
+      "grad_norm": 0.15972478687763214,
+      "learning_rate": 4.523344914530568e-05,
+      "loss": 0.4051,
       "step": 31065
     },
     {
-      "epoch": 1.09,
-      "learning_rate": 4.5481206370415094e-05,
-      "loss": 0.2723,
+      "epoch": 1.1197606948498937,
+      "grad_norm": 0.15999290347099304,
+      "learning_rate": 4.523173506145856e-05,
+      "loss": 0.3884,
       "step": 31070
     },
     {
-      "epoch": 1.09,
-      "learning_rate": 4.547989943097659e-05,
-      "loss": 0.2976,
+      "epoch": 1.1199408945111182,
+      "grad_norm": 0.15831471979618073,
+      "learning_rate": 4.5230020701957976e-05,
+      "loss": 0.3952,
       "step": 31075
     },
     {
-      "epoch": 1.09,
-      "learning_rate": 4.5478265517353534e-05,
-      "loss": 0.3196,
+      "epoch": 1.120121094172343,
+      "grad_norm": 0.17444223165512085,
+      "learning_rate": 4.522830606682727e-05,
+      "loss": 0.4452,
       "step": 31080
     },
     {
-      "epoch": 1.09,
-      "learning_rate": 4.54766313378337e-05,
-      "loss": 0.2942,
+      "epoch": 1.1203012938335677,
+      "grad_norm": 0.16463379561901093,
+      "learning_rate": 4.5226591156089816e-05,
+      "loss": 0.4349,
       "step": 31085
     },
     {
-      "epoch": 1.09,
-      "learning_rate": 4.5474996892438294e-05,
-      "loss": 0.2788,
+      "epoch": 1.1204814934947922,
+      "grad_norm": 0.16084179282188416,
+      "learning_rate": 4.522487596976897e-05,
+      "loss": 0.4415,
       "step": 31090
     },
     {
-      "epoch": 1.09,
-      "learning_rate": 4.5473362181188554e-05,
-      "loss": 0.269,
+      "epoch": 1.120661693156017,
+      "grad_norm": 0.16912023723125458,
+      "learning_rate": 4.5223160507888106e-05,
+      "loss": 0.4269,
       "step": 31095
     },
     {
-      "epoch": 1.09,
-      "learning_rate": 4.5471727204105695e-05,
-      "loss": 0.2889,
+      "epoch": 1.1208418928172414,
+      "grad_norm": 0.1586826741695404,
+      "learning_rate": 4.5221444770470595e-05,
+      "loss": 0.4002,
       "step": 31100
     },
     {
-      "epoch": 1.09,
-      "learning_rate": 4.5470091961210955e-05,
-      "loss": 0.268,
+      "epoch": 1.1210220924784662,
+      "grad_norm": 0.1658790409564972,
+      "learning_rate": 4.521972875753981e-05,
+      "loss": 0.41,
       "step": 31105
     },
     {
-      "epoch": 1.09,
-      "learning_rate": 4.5468456452525556e-05,
-      "loss": 0.2981,
+      "epoch": 1.1212022921396907,
+      "grad_norm": 0.18843096494674683,
+      "learning_rate": 4.521801246911914e-05,
+      "loss": 0.4006,
       "step": 31110
     },
     {
-      "epoch": 1.09,
-      "learning_rate": 4.546682067807074e-05,
-      "loss": 0.2968,
+      "epoch": 1.1213824918009154,
+      "grad_norm": 0.1666111797094345,
+      "learning_rate": 4.521629590523197e-05,
+      "loss": 0.3906,
       "step": 31115
     },
     {
-      "epoch": 1.09,
-      "learning_rate": 4.546518463786775e-05,
-      "loss": 0.2593,
+      "epoch": 1.12156269146214,
+      "grad_norm": 0.15296362340450287,
+      "learning_rate": 4.521457906590167e-05,
+      "loss": 0.4152,
       "step": 31120
     },
     {
-      "epoch": 1.1,
-      "learning_rate": 4.546354833193782e-05,
-      "loss": 0.3151,
+      "epoch": 1.1217428911233647,
+      "grad_norm": 0.20949307084083557,
+      "learning_rate": 4.521286195115165e-05,
+      "loss": 0.4181,
       "step": 31125
     },
     {
-      "epoch": 1.1,
-      "learning_rate": 4.54619117603022e-05,
-      "loss": 0.2953,
+      "epoch": 1.1219230907845894,
+      "grad_norm": 0.1738746613264084,
+      "learning_rate": 4.5211144561005295e-05,
+      "loss": 0.4134,
       "step": 31130
     },
     {
-      "epoch": 1.1,
-      "learning_rate": 4.546027492298215e-05,
-      "loss": 0.2878,
+      "epoch": 1.122103290445814,
+      "grad_norm": 0.199616476893425,
+      "learning_rate": 4.520942689548601e-05,
+      "loss": 0.4121,
       "step": 31135
     },
     {
-      "epoch": 1.1,
-      "learning_rate": 4.54586378199989e-05,
-      "loss": 0.2767,
+      "epoch": 1.1222834901070387,
+      "grad_norm": 0.13446038961410522,
+      "learning_rate": 4.52077089546172e-05,
+      "loss": 0.4055,
       "step": 31140
     },
     {
-      "epoch": 1.1,
-      "learning_rate": 4.545700045137373e-05,
-      "loss": 0.2915,
+      "epoch": 1.1224636897682632,
+      "grad_norm": 0.15235783159732819,
+      "learning_rate": 4.520599073842226e-05,
+      "loss": 0.416,
       "step": 31145
     },
     {
-      "epoch": 1.1,
-      "learning_rate": 4.545536281712789e-05,
-      "loss": 0.2771,
+      "epoch": 1.122643889429488,
+      "grad_norm": 0.1988137811422348,
+      "learning_rate": 4.520427224692462e-05,
+      "loss": 0.4166,
       "step": 31150
     },
     {
-      "epoch": 1.1,
-      "learning_rate": 4.5453724917282635e-05,
-      "loss": 0.3109,
+      "epoch": 1.1228240890907124,
+      "grad_norm": 0.19993172585964203,
+      "learning_rate": 4.520255348014768e-05,
+      "loss": 0.4663,
       "step": 31155
     },
     {
-      "epoch": 1.1,
-      "learning_rate": 4.5452086751859254e-05,
-      "loss": 0.2773,
+      "epoch": 1.1230042887519371,
+      "grad_norm": 0.2015010416507721,
+      "learning_rate": 4.520083443811485e-05,
+      "loss": 0.3944,
       "step": 31160
     },
     {
-      "epoch": 1.1,
-      "learning_rate": 4.5450448320879004e-05,
-      "loss": 0.283,
+      "epoch": 1.1231844884131619,
+      "grad_norm": 0.17933201789855957,
+      "learning_rate": 4.519911512084957e-05,
+      "loss": 0.4343,
       "step": 31165
     },
     {
-      "epoch": 1.1,
-      "learning_rate": 4.544880962436315e-05,
-      "loss": 0.2982,
+      "epoch": 1.1233646880743864,
+      "grad_norm": 0.15731388330459595,
+      "learning_rate": 4.519739552837526e-05,
+      "loss": 0.4049,
       "step": 31170
     },
     {
-      "epoch": 1.1,
-      "learning_rate": 4.544717066233298e-05,
-      "loss": 0.2781,
+      "epoch": 1.1235448877356111,
+      "grad_norm": 0.17282412946224213,
+      "learning_rate": 4.519567566071534e-05,
+      "loss": 0.4227,
       "step": 31175
     },
     {
-      "epoch": 1.1,
-      "learning_rate": 4.544553143480978e-05,
-      "loss": 0.2913,
+      "epoch": 1.1237250873968356,
+      "grad_norm": 0.165944442152977,
+      "learning_rate": 4.519395551789325e-05,
+      "loss": 0.4523,
       "step": 31180
     },
     {
-      "epoch": 1.1,
-      "learning_rate": 4.544389194181483e-05,
-      "loss": 0.28,
+      "epoch": 1.1239052870580604,
+      "grad_norm": 0.1765991449356079,
+      "learning_rate": 4.5192235099932425e-05,
+      "loss": 0.4304,
       "step": 31185
     },
     {
-      "epoch": 1.1,
-      "learning_rate": 4.544225218336942e-05,
-      "loss": 0.2506,
+      "epoch": 1.1240854867192849,
+      "grad_norm": 0.17275585234165192,
+      "learning_rate": 4.519051440685631e-05,
+      "loss": 0.4245,
       "step": 31190
     },
     {
-      "epoch": 1.1,
-      "learning_rate": 4.544061215949483e-05,
-      "loss": 0.2924,
+      "epoch": 1.1242656863805096,
+      "grad_norm": 0.18365871906280518,
+      "learning_rate": 4.518879343868834e-05,
+      "loss": 0.4363,
       "step": 31195
     },
     {
-      "epoch": 1.1,
-      "learning_rate": 4.543897187021236e-05,
-      "loss": 0.2654,
+      "epoch": 1.1244458860417343,
+      "grad_norm": 0.17454713582992554,
+      "learning_rate": 4.5187072195451975e-05,
+      "loss": 0.4329,
       "step": 31200
     },
     {
-      "epoch": 1.1,
-      "learning_rate": 4.543733131554332e-05,
-      "loss": 0.2989,
+      "epoch": 1.1246260857029589,
+      "grad_norm": 0.1820690780878067,
+      "learning_rate": 4.518535067717066e-05,
+      "loss": 0.4217,
       "step": 31205
     },
     {
-      "epoch": 1.1,
-      "learning_rate": 4.5435690495508984e-05,
-      "loss": 0.2829,
+      "epoch": 1.1248062853641836,
+      "grad_norm": 0.16349612176418304,
+      "learning_rate": 4.518362888386784e-05,
+      "loss": 0.4456,
       "step": 31210
     },
     {
-      "epoch": 1.1,
-      "learning_rate": 4.543404941013069e-05,
-      "loss": 0.2898,
+      "epoch": 1.124986485025408,
+      "grad_norm": 0.17479030787944794,
+      "learning_rate": 4.5181906815566986e-05,
+      "loss": 0.4347,
       "step": 31215
     },
     {
-      "epoch": 1.1,
-      "learning_rate": 4.543240805942973e-05,
-      "loss": 0.2847,
+      "epoch": 1.1251666846866328,
+      "grad_norm": 0.14878250658512115,
+      "learning_rate": 4.5180184472291566e-05,
+      "loss": 0.4315,
       "step": 31220
     },
     {
-      "epoch": 1.1,
-      "learning_rate": 4.5430766443427405e-05,
-      "loss": 0.3066,
+      "epoch": 1.1253468843478573,
+      "grad_norm": 0.18707720935344696,
+      "learning_rate": 4.5178461854065036e-05,
+      "loss": 0.3981,
       "step": 31225
     },
     {
-      "epoch": 1.1,
-      "learning_rate": 4.542912456214505e-05,
-      "loss": 0.3007,
+      "epoch": 1.125527084009082,
+      "grad_norm": 0.19858436286449432,
+      "learning_rate": 4.517673896091087e-05,
+      "loss": 0.405,
       "step": 31230
     },
     {
-      "epoch": 1.1,
-      "learning_rate": 4.542748241560398e-05,
-      "loss": 0.2982,
+      "epoch": 1.1257072836703066,
+      "grad_norm": 0.19132393598556519,
+      "learning_rate": 4.517501579285255e-05,
+      "loss": 0.4229,
       "step": 31235
     },
     {
-      "epoch": 1.1,
-      "learning_rate": 4.542584000382552e-05,
-      "loss": 0.3101,
+      "epoch": 1.1258874833315313,
+      "grad_norm": 0.1902177333831787,
+      "learning_rate": 4.5173292349913534e-05,
+      "loss": 0.3822,
       "step": 31240
     },
     {
-      "epoch": 1.1,
-      "learning_rate": 4.542419732683097e-05,
-      "loss": 0.2978,
+      "epoch": 1.126067682992756,
+      "grad_norm": 0.1548781394958496,
+      "learning_rate": 4.517156863211732e-05,
+      "loss": 0.4134,
       "step": 31245
     },
     {
-      "epoch": 1.1,
-      "learning_rate": 4.542255438464169e-05,
-      "loss": 0.2544,
+      "epoch": 1.1262478826539806,
+      "grad_norm": 0.20979823172092438,
+      "learning_rate": 4.517018945999895e-05,
+      "loss": 0.4261,
       "step": 31250
     },
     {
-      "epoch": 1.1,
-      "learning_rate": 4.542091117727899e-05,
-      "loss": 0.3042,
+      "epoch": 1.1264280823152053,
+      "grad_norm": 0.17488163709640503,
+      "learning_rate": 4.5168465247518955e-05,
+      "loss": 0.4078,
       "step": 31255
     },
     {
-      "epoch": 1.1,
-      "learning_rate": 4.5419267704764227e-05,
-      "loss": 0.3146,
+      "epoch": 1.1266082819764298,
+      "grad_norm": 0.16054439544677734,
+      "learning_rate": 4.516674076024752e-05,
+      "loss": 0.4304,
       "step": 31260
     },
     {
-      "epoch": 1.1,
-      "learning_rate": 4.541762396711872e-05,
-      "loss": 0.3158,
+      "epoch": 1.1267884816376545,
+      "grad_norm": 0.16727547347545624,
+      "learning_rate": 4.516501599820816e-05,
+      "loss": 0.4086,
       "step": 31265
     },
     {
-      "epoch": 1.1,
-      "learning_rate": 4.5415979964363825e-05,
-      "loss": 0.2852,
+      "epoch": 1.1269686812988793,
+      "grad_norm": 0.17022760212421417,
+      "learning_rate": 4.5163290961424355e-05,
+      "loss": 0.4378,
       "step": 31270
     },
     {
-      "epoch": 1.1,
-      "learning_rate": 4.541433569652089e-05,
-      "loss": 0.2918,
+      "epoch": 1.1271488809601038,
+      "grad_norm": 0.1838621199131012,
+      "learning_rate": 4.5161565649919614e-05,
+      "loss": 0.4176,
       "step": 31275
     },
     {
-      "epoch": 1.1,
-      "learning_rate": 4.541269116361125e-05,
-      "loss": 0.3133,
+      "epoch": 1.1273290806213285,
+      "grad_norm": 0.16698120534420013,
+      "learning_rate": 4.515984006371744e-05,
+      "loss": 0.4196,
       "step": 31280
     },
     {
-      "epoch": 1.1,
-      "learning_rate": 4.541104636565627e-05,
-      "loss": 0.2998,
+      "epoch": 1.127509280282553,
+      "grad_norm": 0.19196081161499023,
+      "learning_rate": 4.5158114202841354e-05,
+      "loss": 0.3924,
       "step": 31285
     },
     {
-      "epoch": 1.1,
-      "learning_rate": 4.540940130267731e-05,
-      "loss": 0.2574,
+      "epoch": 1.1276894799437778,
+      "grad_norm": 0.17519588768482208,
+      "learning_rate": 4.515638806731486e-05,
+      "loss": 0.4042,
       "step": 31290
     },
     {
-      "epoch": 1.1,
-      "learning_rate": 4.5407755974695726e-05,
-      "loss": 0.305,
+      "epoch": 1.1278696796050023,
+      "grad_norm": 0.1803540289402008,
+      "learning_rate": 4.515466165716149e-05,
+      "loss": 0.3852,
       "step": 31295
     },
     {
-      "epoch": 1.1,
-      "learning_rate": 4.540611038173288e-05,
-      "loss": 0.2872,
+      "epoch": 1.128049879266227,
+      "grad_norm": 0.19045265018939972,
+      "learning_rate": 4.5152934972404746e-05,
+      "loss": 0.4543,
       "step": 31300
     },
     {
-      "epoch": 1.1,
-      "learning_rate": 4.540446452381013e-05,
-      "loss": 0.2743,
+      "epoch": 1.1282300789274515,
+      "grad_norm": 0.2028571367263794,
+      "learning_rate": 4.515120801306818e-05,
+      "loss": 0.4327,
       "step": 31305
     },
     {
-      "epoch": 1.1,
-      "learning_rate": 4.540281840094886e-05,
-      "loss": 0.2836,
+      "epoch": 1.1284102785886763,
+      "grad_norm": 0.18844293057918549,
+      "learning_rate": 4.514948077917529e-05,
+      "loss": 0.4408,
       "step": 31310
     },
     {
-      "epoch": 1.1,
-      "learning_rate": 4.540117201317044e-05,
-      "loss": 0.2902,
+      "epoch": 1.128590478249901,
+      "grad_norm": 0.17702354490756989,
+      "learning_rate": 4.514775327074963e-05,
+      "loss": 0.4077,
       "step": 31315
     },
     {
-      "epoch": 1.1,
-      "learning_rate": 4.539952536049624e-05,
-      "loss": 0.2857,
+      "epoch": 1.1287706779111255,
+      "grad_norm": 0.1727331429719925,
+      "learning_rate": 4.514602548781474e-05,
+      "loss": 0.4235,
       "step": 31320
     },
     {
-      "epoch": 1.1,
-      "learning_rate": 4.539787844294766e-05,
-      "loss": 0.2743,
+      "epoch": 1.1289508775723502,
+      "grad_norm": 0.18904700875282288,
+      "learning_rate": 4.514429743039414e-05,
+      "loss": 0.4351,
       "step": 31325
     },
     {
-      "epoch": 1.1,
-      "learning_rate": 4.539623126054606e-05,
-      "loss": 0.2676,
+      "epoch": 1.1291310772335748,
+      "grad_norm": 0.15556100010871887,
+      "learning_rate": 4.51425690985114e-05,
+      "loss": 0.4397,
       "step": 31330
     },
     {
-      "epoch": 1.1,
-      "learning_rate": 4.539458381331284e-05,
-      "loss": 0.2786,
+      "epoch": 1.1293112768947995,
+      "grad_norm": 0.16988936066627502,
+      "learning_rate": 4.514084049219005e-05,
+      "loss": 0.3974,
       "step": 31335
     },
     {
-      "epoch": 1.1,
-      "learning_rate": 4.53929361012694e-05,
-      "loss": 0.2938,
+      "epoch": 1.129491476556024,
+      "grad_norm": 0.16999125480651855,
+      "learning_rate": 4.513911161145365e-05,
+      "loss": 0.3972,
       "step": 31340
     },
     {
-      "epoch": 1.1,
-      "learning_rate": 4.5391288124437116e-05,
-      "loss": 0.3097,
+      "epoch": 1.1296716762172487,
+      "grad_norm": 0.20566906034946442,
+      "learning_rate": 4.513738245632575e-05,
+      "loss": 0.4368,
       "step": 31345
     },
     {
-      "epoch": 1.1,
-      "learning_rate": 4.5389639882837386e-05,
-      "loss": 0.2933,
+      "epoch": 1.1298518758784732,
+      "grad_norm": 0.2229662835597992,
+      "learning_rate": 4.513565302682992e-05,
+      "loss": 0.4017,
       "step": 31350
     },
     {
-      "epoch": 1.1,
-      "learning_rate": 4.538799137649163e-05,
-      "loss": 0.2772,
+      "epoch": 1.130032075539698,
+      "grad_norm": 0.1942766159772873,
+      "learning_rate": 4.513392332298971e-05,
+      "loss": 0.4144,
       "step": 31355
     },
     {
-      "epoch": 1.1,
-      "learning_rate": 4.538634260542124e-05,
-      "loss": 0.2808,
+      "epoch": 1.1302122752009227,
+      "grad_norm": 0.1725579798221588,
+      "learning_rate": 4.513219334482869e-05,
+      "loss": 0.386,
       "step": 31360
     },
     {
-      "epoch": 1.1,
-      "learning_rate": 4.538469356964763e-05,
-      "loss": 0.3264,
+      "epoch": 1.1303924748621472,
+      "grad_norm": 0.1760469377040863,
+      "learning_rate": 4.513046309237044e-05,
+      "loss": 0.4181,
       "step": 31365
     },
     {
-      "epoch": 1.1,
-      "learning_rate": 4.5383044269192196e-05,
-      "loss": 0.2675,
+      "epoch": 1.130572674523372,
+      "grad_norm": 0.1780589520931244,
+      "learning_rate": 4.512873256563852e-05,
+      "loss": 0.4206,
       "step": 31370
     },
     {
-      "epoch": 1.1,
-      "learning_rate": 4.5381394704076374e-05,
-      "loss": 0.3035,
+      "epoch": 1.1307528741845965,
+      "grad_norm": 0.16992972791194916,
+      "learning_rate": 4.5127001764656526e-05,
+      "loss": 0.4126,
       "step": 31375
     },
     {
-      "epoch": 1.1,
-      "learning_rate": 4.537974487432157e-05,
-      "loss": 0.3087,
+      "epoch": 1.1309330738458212,
+      "grad_norm": 0.1983700394630432,
+      "learning_rate": 4.5125270689448015e-05,
+      "loss": 0.4072,
       "step": 31380
     },
     {
-      "epoch": 1.1,
-      "learning_rate": 4.537809477994922e-05,
-      "loss": 0.3148,
+      "epoch": 1.1311132735070457,
+      "grad_norm": 0.17569278180599213,
+      "learning_rate": 4.51235393400366e-05,
+      "loss": 0.4082,
       "step": 31385
     },
     {
-      "epoch": 1.1,
-      "learning_rate": 4.537644442098073e-05,
-      "loss": 0.3031,
+      "epoch": 1.1312934731682704,
+      "grad_norm": 0.22225306928157806,
+      "learning_rate": 4.512180771644585e-05,
+      "loss": 0.4641,
       "step": 31390
     },
     {
-      "epoch": 1.1,
-      "learning_rate": 4.537479379743754e-05,
-      "loss": 0.2633,
+      "epoch": 1.131473672829495,
+      "grad_norm": 0.1481614112854004,
+      "learning_rate": 4.512007581869937e-05,
+      "loss": 0.3931,
       "step": 31395
     },
     {
-      "epoch": 1.1,
-      "learning_rate": 4.537314290934108e-05,
-      "loss": 0.2997,
+      "epoch": 1.1316538724907197,
+      "grad_norm": 0.21973121166229248,
+      "learning_rate": 4.511834364682076e-05,
+      "loss": 0.4273,
       "step": 31400
     },
     {
-      "epoch": 1.1,
-      "learning_rate": 4.537149175671278e-05,
-      "loss": 0.3034,
+      "epoch": 1.1318340721519444,
+      "grad_norm": 0.21641644835472107,
+      "learning_rate": 4.51166112008336e-05,
+      "loss": 0.4151,
       "step": 31405
     },
     {
-      "epoch": 1.11,
-      "learning_rate": 4.5369840339574096e-05,
-      "loss": 0.2862,
+      "epoch": 1.132014271813169,
+      "grad_norm": 0.18511059880256653,
+      "learning_rate": 4.511487848076151e-05,
+      "loss": 0.4423,
       "step": 31410
     },
     {
-      "epoch": 1.11,
-      "learning_rate": 4.536818865794645e-05,
-      "loss": 0.2872,
+      "epoch": 1.1321944714743937,
+      "grad_norm": 0.2063097059726715,
+      "learning_rate": 4.5113145486628095e-05,
+      "loss": 0.4257,
       "step": 31415
     },
     {
-      "epoch": 1.11,
-      "learning_rate": 4.53665367118513e-05,
-      "loss": 0.2752,
+      "epoch": 1.1323746711356182,
+      "grad_norm": 0.17430323362350464,
+      "learning_rate": 4.511141221845697e-05,
+      "loss": 0.4076,
       "step": 31420
     },
     {
-      "epoch": 1.11,
-      "learning_rate": 4.5364884501310086e-05,
-      "loss": 0.3041,
+      "epoch": 1.132554870796843,
+      "grad_norm": 0.1422608643770218,
+      "learning_rate": 4.510967867627175e-05,
+      "loss": 0.3761,
       "step": 31425
     },
     {
-      "epoch": 1.11,
-      "learning_rate": 4.5363232026344284e-05,
-      "loss": 0.2794,
+      "epoch": 1.1327350704580676,
+      "grad_norm": 0.2072693109512329,
+      "learning_rate": 4.510794486009604e-05,
+      "loss": 0.4133,
       "step": 31430
     },
     {
-      "epoch": 1.11,
-      "learning_rate": 4.536157928697532e-05,
-      "loss": 0.295,
+      "epoch": 1.1329152701192922,
+      "grad_norm": 0.1433665156364441,
+      "learning_rate": 4.5106210769953484e-05,
+      "loss": 0.3992,
       "step": 31435
     },
     {
-      "epoch": 1.11,
-      "learning_rate": 4.5359926283224674e-05,
-      "loss": 0.2901,
+      "epoch": 1.133095469780517,
+      "grad_norm": 0.20627710223197937,
+      "learning_rate": 4.5104476405867705e-05,
+      "loss": 0.4233,
       "step": 31440
     },
     {
-      "epoch": 1.11,
-      "learning_rate": 4.535827301511381e-05,
-      "loss": 0.2754,
+      "epoch": 1.1332756694417414,
+      "grad_norm": 0.1617184430360794,
+      "learning_rate": 4.510274176786231e-05,
+      "loss": 0.4564,
       "step": 31445
     },
     {
-      "epoch": 1.11,
-      "learning_rate": 4.535661948266418e-05,
-      "loss": 0.2856,
+      "epoch": 1.1334558691029661,
+      "grad_norm": 0.1667255014181137,
+      "learning_rate": 4.5101006855960956e-05,
+      "loss": 0.4173,
       "step": 31450
     },
     {
-      "epoch": 1.11,
-      "learning_rate": 4.5354965685897255e-05,
-      "loss": 0.2911,
+      "epoch": 1.1336360687641907,
+      "grad_norm": 0.17955441772937775,
+      "learning_rate": 4.5099271670187285e-05,
+      "loss": 0.4203,
       "step": 31455
     },
     {
-      "epoch": 1.11,
-      "learning_rate": 4.5353311624834516e-05,
-      "loss": 0.2583,
+      "epoch": 1.1338162684254154,
+      "grad_norm": 0.1967443972826004,
+      "learning_rate": 4.5097536210564915e-05,
+      "loss": 0.4261,
       "step": 31460
     },
     {
-      "epoch": 1.11,
-      "learning_rate": 4.5351657299497444e-05,
-      "loss": 0.2752,
+      "epoch": 1.13399646808664,
+      "grad_norm": 0.1767064779996872,
+      "learning_rate": 4.5095800477117514e-05,
+      "loss": 0.423,
       "step": 31465
     },
     {
-      "epoch": 1.11,
-      "learning_rate": 4.535000270990751e-05,
-      "loss": 0.2796,
+      "epoch": 1.1341766677478646,
+      "grad_norm": 0.15333469212055206,
+      "learning_rate": 4.5094064469868726e-05,
+      "loss": 0.4139,
       "step": 31470
     },
     {
-      "epoch": 1.11,
-      "learning_rate": 4.5348347856086214e-05,
-      "loss": 0.2898,
+      "epoch": 1.1343568674090894,
+      "grad_norm": 0.1856624186038971,
+      "learning_rate": 4.509232818884219e-05,
+      "loss": 0.4326,
       "step": 31475
     },
     {
-      "epoch": 1.11,
-      "learning_rate": 4.534669273805502e-05,
-      "loss": 0.2915,
+      "epoch": 1.1345370670703139,
+      "grad_norm": 0.20386086404323578,
+      "learning_rate": 4.5090591634061577e-05,
+      "loss": 0.4424,
       "step": 31480
     },
     {
-      "epoch": 1.11,
-      "learning_rate": 4.534503735583543e-05,
-      "loss": 0.2943,
+      "epoch": 1.1347172667315386,
+      "grad_norm": 0.2295321226119995,
+      "learning_rate": 4.508885480555055e-05,
+      "loss": 0.4369,
       "step": 31485
     },
     {
-      "epoch": 1.11,
-      "learning_rate": 4.5343381709448943e-05,
-      "loss": 0.2976,
+      "epoch": 1.1348974663927631,
+      "grad_norm": 0.2000851184129715,
+      "learning_rate": 4.5087117703332755e-05,
+      "loss": 0.3865,
       "step": 31490
     },
     {
-      "epoch": 1.11,
-      "learning_rate": 4.534172579891705e-05,
-      "loss": 0.3062,
+      "epoch": 1.1350776660539879,
+      "grad_norm": 0.19415894150733948,
+      "learning_rate": 4.508538032743187e-05,
+      "loss": 0.3817,
       "step": 31495
     },
     {
-      "epoch": 1.11,
-      "learning_rate": 4.534006962426125e-05,
-      "loss": 0.2871,
+      "epoch": 1.1352578657152124,
+      "grad_norm": 0.23722659051418304,
+      "learning_rate": 4.5083642677871575e-05,
+      "loss": 0.4131,
       "step": 31500
     },
     {
-      "epoch": 1.11,
-      "eval_loss": 0.28987738490104675,
-      "eval_runtime": 10.5447,
-      "eval_samples_per_second": 9.483,
-      "eval_steps_per_second": 9.483,
+      "epoch": 1.1352578657152124,
+      "eval_loss": 0.45174121856689453,
+      "eval_runtime": 3.5299,
+      "eval_samples_per_second": 28.33,
+      "eval_steps_per_second": 7.082,
       "step": 31500
     },
     {
-      "epoch": 1.11,
-      "learning_rate": 4.5338413185503056e-05,
-      "loss": 0.2818,
+      "epoch": 1.135438065376437,
+      "grad_norm": 0.1844346821308136,
+      "learning_rate": 4.508190475467553e-05,
+      "loss": 0.4107,
       "step": 31505
     },
     {
-      "epoch": 1.11,
-      "learning_rate": 4.533675648266397e-05,
-      "loss": 0.2519,
+      "epoch": 1.1356182650376616,
+      "grad_norm": 0.16183345019817352,
+      "learning_rate": 4.508016655786742e-05,
+      "loss": 0.4014,
       "step": 31510
     },
     {
-      "epoch": 1.11,
-      "learning_rate": 4.533509951576549e-05,
-      "loss": 0.2737,
+      "epoch": 1.1357984646988863,
+      "grad_norm": 0.20634831488132477,
+      "learning_rate": 4.507842808747093e-05,
+      "loss": 0.4184,
       "step": 31515
     },
     {
-      "epoch": 1.11,
-      "learning_rate": 4.5333442284829164e-05,
-      "loss": 0.3031,
+      "epoch": 1.135978664360111,
+      "grad_norm": 0.15508055686950684,
+      "learning_rate": 4.507668934350975e-05,
+      "loss": 0.4209,
       "step": 31520
     },
     {
-      "epoch": 1.11,
-      "learning_rate": 4.533178478987647e-05,
-      "loss": 0.3129,
+      "epoch": 1.1361588640213356,
+      "grad_norm": 0.1634722650051117,
+      "learning_rate": 4.507495032600756e-05,
+      "loss": 0.3867,
       "step": 31525
     },
     {
-      "epoch": 1.11,
-      "learning_rate": 4.533012703092897e-05,
-      "loss": 0.2801,
+      "epoch": 1.1363390636825603,
+      "grad_norm": 0.1492082178592682,
+      "learning_rate": 4.5073211034988055e-05,
+      "loss": 0.3896,
       "step": 31530
     },
     {
-      "epoch": 1.11,
-      "learning_rate": 4.5328469008008154e-05,
-      "loss": 0.2885,
+      "epoch": 1.1365192633437848,
+      "grad_norm": 0.17373375594615936,
+      "learning_rate": 4.507147147047494e-05,
+      "loss": 0.4057,
       "step": 31535
     },
     {
-      "epoch": 1.11,
-      "learning_rate": 4.532681072113557e-05,
-      "loss": 0.2672,
+      "epoch": 1.1366994630050096,
+      "grad_norm": 0.178725004196167,
+      "learning_rate": 4.5069731632491914e-05,
+      "loss": 0.4372,
       "step": 31540
     },
     {
-      "epoch": 1.11,
-      "learning_rate": 4.532515217033275e-05,
-      "loss": 0.3211,
+      "epoch": 1.1368796626662343,
+      "grad_norm": 0.17013704776763916,
+      "learning_rate": 4.506799152106268e-05,
+      "loss": 0.441,
       "step": 31545
     },
     {
-      "epoch": 1.11,
-      "learning_rate": 4.5323493355621214e-05,
-      "loss": 0.3024,
+      "epoch": 1.1370598623274588,
+      "grad_norm": 0.1601504683494568,
+      "learning_rate": 4.506625113621094e-05,
+      "loss": 0.4496,
       "step": 31550
     },
     {
-      "epoch": 1.11,
-      "learning_rate": 4.5321834277022515e-05,
-      "loss": 0.3081,
+      "epoch": 1.1372400619886835,
+      "grad_norm": 0.16774295270442963,
+      "learning_rate": 4.506451047796042e-05,
+      "loss": 0.4396,
       "step": 31555
     },
     {
-      "epoch": 1.11,
-      "learning_rate": 4.532017493455819e-05,
-      "loss": 0.3033,
+      "epoch": 1.137420261649908,
+      "grad_norm": 0.18739162385463715,
+      "learning_rate": 4.506276954633483e-05,
+      "loss": 0.4404,
       "step": 31560
     },
     {
-      "epoch": 1.11,
-      "learning_rate": 4.531851532824979e-05,
-      "loss": 0.3097,
+      "epoch": 1.1376004613111328,
+      "grad_norm": 0.2417900562286377,
+      "learning_rate": 4.506102834135788e-05,
+      "loss": 0.4549,
       "step": 31565
     },
     {
-      "epoch": 1.11,
-      "learning_rate": 4.531685545811885e-05,
-      "loss": 0.2925,
+      "epoch": 1.1377806609723573,
+      "grad_norm": 0.182041198015213,
+      "learning_rate": 4.5059286863053314e-05,
+      "loss": 0.4239,
       "step": 31570
     },
     {
-      "epoch": 1.11,
-      "learning_rate": 4.5315195324186935e-05,
-      "loss": 0.2856,
+      "epoch": 1.137960860633582,
+      "grad_norm": 0.18919362127780914,
+      "learning_rate": 4.5057545111444846e-05,
+      "loss": 0.4025,
       "step": 31575
     },
     {
-      "epoch": 1.11,
-      "learning_rate": 4.53135349264756e-05,
-      "loss": 0.2676,
+      "epoch": 1.1381410602948066,
+      "grad_norm": 0.17599248886108398,
+      "learning_rate": 4.50558030865562e-05,
+      "loss": 0.4041,
       "step": 31580
     },
     {
-      "epoch": 1.11,
-      "learning_rate": 4.531187426500639e-05,
-      "loss": 0.279,
+      "epoch": 1.1383212599560313,
+      "grad_norm": 0.15774129331111908,
+      "learning_rate": 4.505406078841113e-05,
+      "loss": 0.4701,
       "step": 31585
     },
     {
-      "epoch": 1.11,
-      "learning_rate": 4.5310213339800895e-05,
-      "loss": 0.3094,
+      "epoch": 1.138501459617256,
+      "grad_norm": 0.17122262716293335,
+      "learning_rate": 4.505231821703336e-05,
+      "loss": 0.4054,
       "step": 31590
     },
     {
-      "epoch": 1.11,
-      "learning_rate": 4.5308552150880655e-05,
-      "loss": 0.2832,
+      "epoch": 1.1386816592784805,
+      "grad_norm": 0.15675261616706848,
+      "learning_rate": 4.505057537244664e-05,
+      "loss": 0.4538,
       "step": 31595
     },
     {
-      "epoch": 1.11,
-      "learning_rate": 4.5306890698267254e-05,
-      "loss": 0.2985,
+      "epoch": 1.1388618589397053,
+      "grad_norm": 0.1774090677499771,
+      "learning_rate": 4.504883225467471e-05,
+      "loss": 0.4258,
       "step": 31600
     },
     {
-      "epoch": 1.11,
-      "learning_rate": 4.5305228981982254e-05,
-      "loss": 0.2949,
+      "epoch": 1.1390420586009298,
+      "grad_norm": 0.13083837926387787,
+      "learning_rate": 4.5047088863741314e-05,
+      "loss": 0.3618,
       "step": 31605
     },
     {
-      "epoch": 1.11,
-      "learning_rate": 4.5303567002047246e-05,
-      "loss": 0.2908,
+      "epoch": 1.1392222582621545,
+      "grad_norm": 0.1870373785495758,
+      "learning_rate": 4.5045345199670227e-05,
+      "loss": 0.4286,
       "step": 31610
     },
     {
-      "epoch": 1.11,
-      "learning_rate": 4.530190475848379e-05,
-      "loss": 0.3151,
+      "epoch": 1.139402457923379,
+      "grad_norm": 0.1769174039363861,
+      "learning_rate": 4.504360126248518e-05,
+      "loss": 0.4424,
       "step": 31615
     },
     {
-      "epoch": 1.11,
-      "learning_rate": 4.530024225131349e-05,
-      "loss": 0.2759,
+      "epoch": 1.1395826575846038,
+      "grad_norm": 0.19420921802520752,
+      "learning_rate": 4.5041857052209954e-05,
+      "loss": 0.4086,
       "step": 31620
     },
     {
-      "epoch": 1.11,
-      "learning_rate": 4.529857948055791e-05,
-      "loss": 0.2868,
+      "epoch": 1.1397628572458283,
+      "grad_norm": 0.19101811945438385,
+      "learning_rate": 4.504011256886831e-05,
+      "loss": 0.4177,
       "step": 31625
     },
     {
-      "epoch": 1.11,
-      "learning_rate": 4.529691644623866e-05,
-      "loss": 0.2885,
+      "epoch": 1.139943056907053,
+      "grad_norm": 0.20886962115764618,
+      "learning_rate": 4.503836781248401e-05,
+      "loss": 0.4463,
       "step": 31630
     },
     {
-      "epoch": 1.11,
-      "learning_rate": 4.5295253148377334e-05,
-      "loss": 0.2702,
+      "epoch": 1.1401232565682777,
+      "grad_norm": 0.20521090924739838,
+      "learning_rate": 4.503662278308082e-05,
+      "loss": 0.408,
       "step": 31635
     },
     {
-      "epoch": 1.11,
-      "learning_rate": 4.529358958699551e-05,
-      "loss": 0.2905,
+      "epoch": 1.1403034562295022,
+      "grad_norm": 0.16079290211200714,
+      "learning_rate": 4.503487748068253e-05,
+      "loss": 0.437,
       "step": 31640
     },
     {
-      "epoch": 1.11,
-      "learning_rate": 4.5291925762114805e-05,
-      "loss": 0.3137,
+      "epoch": 1.140483655890727,
+      "grad_norm": 0.18102915585041046,
+      "learning_rate": 4.5033131905312906e-05,
+      "loss": 0.4308,
       "step": 31645
     },
     {
-      "epoch": 1.11,
-      "learning_rate": 4.529026167375681e-05,
-      "loss": 0.2886,
+      "epoch": 1.1406638555519515,
+      "grad_norm": 0.1715448796749115,
+      "learning_rate": 4.503138605699575e-05,
+      "loss": 0.3977,
       "step": 31650
     },
     {
-      "epoch": 1.11,
-      "learning_rate": 4.5288597321943146e-05,
-      "loss": 0.3019,
+      "epoch": 1.1408440552131762,
+      "grad_norm": 0.18790487945079803,
+      "learning_rate": 4.502963993575483e-05,
+      "loss": 0.4169,
       "step": 31655
     },
     {
-      "epoch": 1.11,
-      "learning_rate": 4.528693270669542e-05,
-      "loss": 0.2911,
+      "epoch": 1.141024254874401,
+      "grad_norm": 0.15706005692481995,
+      "learning_rate": 4.502789354161395e-05,
+      "loss": 0.416,
       "step": 31660
     },
     {
-      "epoch": 1.11,
-      "learning_rate": 4.528526782803523e-05,
-      "loss": 0.2665,
+      "epoch": 1.1412044545356255,
+      "grad_norm": 0.1588856726884842,
+      "learning_rate": 4.5026146874596895e-05,
+      "loss": 0.4035,
       "step": 31665
     },
     {
-      "epoch": 1.11,
-      "learning_rate": 4.5283602685984215e-05,
-      "loss": 0.2793,
+      "epoch": 1.1413846541968502,
+      "grad_norm": 0.2316998541355133,
+      "learning_rate": 4.502439993472746e-05,
+      "loss": 0.4545,
       "step": 31670
     },
     {
-      "epoch": 1.11,
-      "learning_rate": 4.528193728056398e-05,
-      "loss": 0.2908,
+      "epoch": 1.1415648538580747,
+      "grad_norm": 0.20132675766944885,
+      "learning_rate": 4.502265272202946e-05,
+      "loss": 0.4461,
       "step": 31675
     },
     {
-      "epoch": 1.11,
-      "learning_rate": 4.5280271611796165e-05,
-      "loss": 0.2861,
+      "epoch": 1.1417450535192994,
+      "grad_norm": 0.1755920946598053,
+      "learning_rate": 4.502090523652669e-05,
+      "loss": 0.4417,
       "step": 31680
     },
     {
-      "epoch": 1.11,
-      "learning_rate": 4.527860567970238e-05,
-      "loss": 0.301,
+      "epoch": 1.141925253180524,
+      "grad_norm": 0.19191905856132507,
+      "learning_rate": 4.501915747824296e-05,
+      "loss": 0.4183,
       "step": 31685
     },
     {
-      "epoch": 1.11,
-      "learning_rate": 4.527693948430427e-05,
-      "loss": 0.2825,
+      "epoch": 1.1421054528417487,
+      "grad_norm": 0.16491484642028809,
+      "learning_rate": 4.501740944720209e-05,
+      "loss": 0.4373,
       "step": 31690
     },
     {
-      "epoch": 1.12,
-      "learning_rate": 4.527527302562346e-05,
-      "loss": 0.298,
+      "epoch": 1.1422856525029732,
+      "grad_norm": 0.16868291795253754,
+      "learning_rate": 4.501566114342789e-05,
+      "loss": 0.4038,
       "step": 31695
     },
     {
-      "epoch": 1.12,
-      "learning_rate": 4.5273606303681596e-05,
-      "loss": 0.2914,
+      "epoch": 1.142465852164198,
+      "grad_norm": 0.1392037272453308,
+      "learning_rate": 4.501391256694418e-05,
+      "loss": 0.3929,
       "step": 31700
     },
     {
-      "epoch": 1.12,
-      "learning_rate": 4.527193931850031e-05,
-      "loss": 0.2769,
+      "epoch": 1.1426460518254227,
+      "grad_norm": 0.1932985484600067,
+      "learning_rate": 4.501216371777479e-05,
+      "loss": 0.4323,
       "step": 31705
     },
     {
-      "epoch": 1.12,
-      "learning_rate": 4.527027207010125e-05,
-      "loss": 0.3404,
+      "epoch": 1.1428262514866472,
+      "grad_norm": 0.17676390707492828,
+      "learning_rate": 4.501041459594354e-05,
+      "loss": 0.419,
       "step": 31710
     },
     {
-      "epoch": 1.12,
-      "learning_rate": 4.5268604558506075e-05,
-      "loss": 0.2766,
+      "epoch": 1.143006451147872,
+      "grad_norm": 0.20211122930049896,
+      "learning_rate": 4.5008665201474273e-05,
+      "loss": 0.4147,
       "step": 31715
     },
     {
-      "epoch": 1.12,
-      "learning_rate": 4.526693678373642e-05,
-      "loss": 0.2803,
+      "epoch": 1.1431866508090964,
+      "grad_norm": 0.16196738183498383,
+      "learning_rate": 4.50069155343908e-05,
+      "loss": 0.4052,
       "step": 31720
     },
     {
-      "epoch": 1.12,
-      "learning_rate": 4.526526874581395e-05,
-      "loss": 0.2841,
+      "epoch": 1.1433668504703212,
+      "grad_norm": 0.16365759074687958,
+      "learning_rate": 4.500516559471699e-05,
+      "loss": 0.424,
       "step": 31725
     },
     {
-      "epoch": 1.12,
-      "learning_rate": 4.5263600444760316e-05,
-      "loss": 0.2872,
+      "epoch": 1.1435470501315457,
+      "grad_norm": 0.16799362003803253,
+      "learning_rate": 4.500341538247667e-05,
+      "loss": 0.4237,
       "step": 31730
     },
     {
-      "epoch": 1.12,
-      "learning_rate": 4.526193188059718e-05,
-      "loss": 0.2733,
+      "epoch": 1.1437272497927704,
+      "grad_norm": 0.1993078589439392,
+      "learning_rate": 4.500166489769369e-05,
+      "loss": 0.4194,
       "step": 31735
     },
     {
-      "epoch": 1.12,
-      "learning_rate": 4.526026305334623e-05,
-      "loss": 0.2901,
+      "epoch": 1.143907449453995,
+      "grad_norm": 0.15421169996261597,
+      "learning_rate": 4.49999141403919e-05,
+      "loss": 0.3786,
       "step": 31740
     },
     {
-      "epoch": 1.12,
-      "learning_rate": 4.52585939630291e-05,
-      "loss": 0.2933,
+      "epoch": 1.1440876491152197,
+      "grad_norm": 0.17669521272182465,
+      "learning_rate": 4.499816311059514e-05,
+      "loss": 0.3992,
       "step": 31745
     },
     {
-      "epoch": 1.12,
-      "learning_rate": 4.5256924609667484e-05,
-      "loss": 0.3002,
+      "epoch": 1.1442678487764444,
+      "grad_norm": 0.18035532534122467,
+      "learning_rate": 4.499641180832729e-05,
+      "loss": 0.433,
       "step": 31750
     },
     {
-      "epoch": 1.12,
-      "learning_rate": 4.525525499328306e-05,
-      "loss": 0.2962,
+      "epoch": 1.144448048437669,
+      "grad_norm": 0.15400253236293793,
+      "learning_rate": 4.49946602336122e-05,
+      "loss": 0.4367,
       "step": 31755
     },
     {
-      "epoch": 1.12,
-      "learning_rate": 4.525358511389749e-05,
-      "loss": 0.2934,
+      "epoch": 1.1446282480988936,
+      "grad_norm": 0.16569051146507263,
+      "learning_rate": 4.4992908386473727e-05,
+      "loss": 0.3999,
       "step": 31760
     },
     {
-      "epoch": 1.12,
-      "learning_rate": 4.525191497153246e-05,
-      "loss": 0.3064,
+      "epoch": 1.1448084477601181,
+      "grad_norm": 0.16333115100860596,
+      "learning_rate": 4.4991156266935755e-05,
+      "loss": 0.4458,
       "step": 31765
     },
     {
-      "epoch": 1.12,
-      "learning_rate": 4.5250244566209666e-05,
-      "loss": 0.2713,
+      "epoch": 1.1449886474213429,
+      "grad_norm": 0.17189821600914001,
+      "learning_rate": 4.498940387502214e-05,
+      "loss": 0.399,
       "step": 31770
     },
     {
-      "epoch": 1.12,
-      "learning_rate": 4.52485738979508e-05,
-      "loss": 0.3043,
+      "epoch": 1.1451688470825676,
+      "grad_norm": 0.2206844836473465,
+      "learning_rate": 4.498765121075678e-05,
+      "loss": 0.4515,
       "step": 31775
     },
     {
-      "epoch": 1.12,
-      "learning_rate": 4.5246902966777535e-05,
-      "loss": 0.2654,
+      "epoch": 1.1453490467437921,
+      "grad_norm": 0.18912701308727264,
+      "learning_rate": 4.498589827416354e-05,
+      "loss": 0.4375,
       "step": 31780
     },
     {
-      "epoch": 1.12,
-      "learning_rate": 4.524523177271158e-05,
-      "loss": 0.2602,
+      "epoch": 1.1455292464050169,
+      "grad_norm": 0.17603172361850739,
+      "learning_rate": 4.49841450652663e-05,
+      "loss": 0.407,
       "step": 31785
     },
     {
-      "epoch": 1.12,
-      "learning_rate": 4.5243560315774634e-05,
-      "loss": 0.2898,
+      "epoch": 1.1457094460662414,
+      "grad_norm": 0.1851511150598526,
+      "learning_rate": 4.498239158408896e-05,
+      "loss": 0.4273,
       "step": 31790
     },
     {
-      "epoch": 1.12,
-      "learning_rate": 4.52418885959884e-05,
-      "loss": 0.2941,
+      "epoch": 1.145889645727466,
+      "grad_norm": 0.13465797901153564,
+      "learning_rate": 4.498063783065539e-05,
+      "loss": 0.4008,
       "step": 31795
     },
     {
-      "epoch": 1.12,
-      "learning_rate": 4.524021661337458e-05,
-      "loss": 0.3295,
+      "epoch": 1.1460698453886906,
+      "grad_norm": 0.16686324775218964,
+      "learning_rate": 4.4978883804989516e-05,
+      "loss": 0.4083,
       "step": 31800
     },
     {
-      "epoch": 1.12,
-      "learning_rate": 4.5238544367954884e-05,
-      "loss": 0.2934,
+      "epoch": 1.1462500450499153,
+      "grad_norm": 0.16204862296581268,
+      "learning_rate": 4.49771295071152e-05,
+      "loss": 0.4026,
       "step": 31805
     },
     {
-      "epoch": 1.12,
-      "learning_rate": 4.5236871859751026e-05,
-      "loss": 0.2955,
+      "epoch": 1.1464302447111399,
+      "grad_norm": 0.19073139131069183,
+      "learning_rate": 4.4975374937056366e-05,
+      "loss": 0.4004,
       "step": 31810
     },
     {
-      "epoch": 1.12,
-      "learning_rate": 4.523519908878473e-05,
-      "loss": 0.2837,
+      "epoch": 1.1466104443723646,
+      "grad_norm": 0.20284846425056458,
+      "learning_rate": 4.4973620094836926e-05,
+      "loss": 0.4195,
       "step": 31815
     },
     {
-      "epoch": 1.12,
-      "learning_rate": 4.5233526055077705e-05,
-      "loss": 0.3061,
+      "epoch": 1.1467906440335893,
+      "grad_norm": 0.18009348213672638,
+      "learning_rate": 4.497186498048077e-05,
+      "loss": 0.4421,
       "step": 31820
     },
     {
-      "epoch": 1.12,
-      "learning_rate": 4.5231852758651685e-05,
-      "loss": 0.2806,
+      "epoch": 1.1469708436948138,
+      "grad_norm": 0.18369971215724945,
+      "learning_rate": 4.497010959401183e-05,
+      "loss": 0.417,
       "step": 31825
     },
     {
-      "epoch": 1.12,
-      "learning_rate": 4.5230179199528384e-05,
-      "loss": 0.3167,
+      "epoch": 1.1471510433560386,
+      "grad_norm": 0.17240361869335175,
+      "learning_rate": 4.4968353935454004e-05,
+      "loss": 0.3953,
       "step": 31830
     },
     {
-      "epoch": 1.12,
-      "learning_rate": 4.5228505377729534e-05,
-      "loss": 0.2961,
+      "epoch": 1.147331243017263,
+      "grad_norm": 0.182241752743721,
+      "learning_rate": 4.496659800483123e-05,
+      "loss": 0.4441,
       "step": 31835
     },
     {
-      "epoch": 1.12,
-      "learning_rate": 4.522683129327688e-05,
-      "loss": 0.3087,
+      "epoch": 1.1475114426784878,
+      "grad_norm": 0.20979800820350647,
+      "learning_rate": 4.4964841802167414e-05,
+      "loss": 0.4349,
       "step": 31840
     },
     {
-      "epoch": 1.12,
-      "learning_rate": 4.522515694619215e-05,
-      "loss": 0.2866,
+      "epoch": 1.1476916423397123,
+      "grad_norm": 0.1706288456916809,
+      "learning_rate": 4.49630853274865e-05,
+      "loss": 0.3891,
       "step": 31845
     },
     {
-      "epoch": 1.12,
-      "learning_rate": 4.522348233649709e-05,
-      "loss": 0.2903,
+      "epoch": 1.147871842000937,
+      "grad_norm": 0.17474350333213806,
+      "learning_rate": 4.496132858081241e-05,
+      "loss": 0.4184,
       "step": 31850
     },
     {
-      "epoch": 1.12,
-      "learning_rate": 4.522180746421343e-05,
-      "loss": 0.2939,
+      "epoch": 1.1480520416621616,
+      "grad_norm": 0.16004018485546112,
+      "learning_rate": 4.495957156216908e-05,
+      "loss": 0.4254,
       "step": 31855
     },
     {
-      "epoch": 1.12,
-      "learning_rate": 4.522013232936293e-05,
-      "loss": 0.2891,
+      "epoch": 1.1482322413233863,
+      "grad_norm": 0.20258988440036774,
+      "learning_rate": 4.495781427158046e-05,
+      "loss": 0.4425,
       "step": 31860
     },
     {
-      "epoch": 1.12,
-      "learning_rate": 4.521845693196735e-05,
-      "loss": 0.3175,
+      "epoch": 1.148412440984611,
+      "grad_norm": 0.18658919632434845,
+      "learning_rate": 4.4956056709070485e-05,
+      "loss": 0.4369,
       "step": 31865
     },
     {
-      "epoch": 1.12,
-      "learning_rate": 4.5216781272048414e-05,
-      "loss": 0.3182,
+      "epoch": 1.1485926406458355,
+      "grad_norm": 0.16538068652153015,
+      "learning_rate": 4.495429887466309e-05,
+      "loss": 0.4099,
       "step": 31870
     },
     {
-      "epoch": 1.12,
-      "learning_rate": 4.5215105349627906e-05,
-      "loss": 0.2788,
+      "epoch": 1.1487728403070603,
+      "grad_norm": 0.16032880544662476,
+      "learning_rate": 4.495254076838225e-05,
+      "loss": 0.3942,
       "step": 31875
     },
     {
-      "epoch": 1.12,
-      "learning_rate": 4.5213429164727575e-05,
-      "loss": 0.2906,
+      "epoch": 1.1489530399682848,
+      "grad_norm": 0.14562112092971802,
+      "learning_rate": 4.495078239025189e-05,
+      "loss": 0.4076,
       "step": 31880
     },
     {
-      "epoch": 1.12,
-      "learning_rate": 4.521175271736918e-05,
-      "loss": 0.2794,
+      "epoch": 1.1491332396295095,
+      "grad_norm": 0.20045393705368042,
+      "learning_rate": 4.494902374029599e-05,
+      "loss": 0.4033,
       "step": 31885
     },
     {
-      "epoch": 1.12,
-      "learning_rate": 4.521007600757451e-05,
-      "loss": 0.2859,
+      "epoch": 1.149313439290734,
+      "grad_norm": 0.1704903542995453,
+      "learning_rate": 4.4947264818538505e-05,
+      "loss": 0.4255,
       "step": 31890
     },
     {
-      "epoch": 1.12,
-      "learning_rate": 4.520839903536531e-05,
-      "loss": 0.2792,
+      "epoch": 1.1494936389519588,
+      "grad_norm": 0.18456216156482697,
+      "learning_rate": 4.49455056250034e-05,
+      "loss": 0.4368,
       "step": 31895
     },
     {
-      "epoch": 1.12,
-      "learning_rate": 4.520672180076337e-05,
-      "loss": 0.2994,
+      "epoch": 1.1496738386131833,
+      "grad_norm": 0.18716961145401,
+      "learning_rate": 4.494374615971464e-05,
+      "loss": 0.4268,
       "step": 31900
     },
     {
-      "epoch": 1.12,
-      "learning_rate": 4.520504430379047e-05,
-      "loss": 0.2815,
+      "epoch": 1.149854038274408,
+      "grad_norm": 0.1741458624601364,
+      "learning_rate": 4.49419864226962e-05,
+      "loss": 0.3761,
       "step": 31905
     },
     {
-      "epoch": 1.12,
-      "learning_rate": 4.520336654446838e-05,
-      "loss": 0.2966,
+      "epoch": 1.1500342379356328,
+      "grad_norm": 0.17461341619491577,
+      "learning_rate": 4.494022641397205e-05,
+      "loss": 0.4028,
       "step": 31910
     },
     {
-      "epoch": 1.12,
-      "learning_rate": 4.5201688522818895e-05,
-      "loss": 0.3142,
+      "epoch": 1.1502144375968573,
+      "grad_norm": 0.16203176975250244,
+      "learning_rate": 4.493846613356619e-05,
+      "loss": 0.4028,
       "step": 31915
     },
     {
-      "epoch": 1.12,
-      "learning_rate": 4.5200010238863794e-05,
-      "loss": 0.2811,
+      "epoch": 1.150394637258082,
+      "grad_norm": 0.16421377658843994,
+      "learning_rate": 4.493670558150258e-05,
+      "loss": 0.4365,
       "step": 31920
     },
     {
-      "epoch": 1.12,
-      "learning_rate": 4.5198331692624874e-05,
-      "loss": 0.2741,
+      "epoch": 1.1505748369193065,
+      "grad_norm": 0.17700999975204468,
+      "learning_rate": 4.493494475780521e-05,
+      "loss": 0.4262,
       "step": 31925
     },
     {
-      "epoch": 1.12,
-      "learning_rate": 4.5196652884123926e-05,
-      "loss": 0.3182,
+      "epoch": 1.1507550365805312,
+      "grad_norm": 0.1917300820350647,
+      "learning_rate": 4.493318366249809e-05,
+      "loss": 0.4477,
       "step": 31930
     },
     {
-      "epoch": 1.12,
-      "learning_rate": 4.5194973813382744e-05,
-      "loss": 0.2964,
+      "epoch": 1.150935236241756,
+      "grad_norm": 0.24866171181201935,
+      "learning_rate": 4.4931422295605196e-05,
+      "loss": 0.3896,
       "step": 31935
     },
     {
-      "epoch": 1.12,
-      "learning_rate": 4.5193294480423146e-05,
-      "loss": 0.3003,
+      "epoch": 1.1511154359029805,
+      "grad_norm": 0.16452571749687195,
+      "learning_rate": 4.492966065715053e-05,
+      "loss": 0.4155,
       "step": 31940
     },
     {
-      "epoch": 1.12,
-      "learning_rate": 4.5191614885266925e-05,
-      "loss": 0.2881,
+      "epoch": 1.1512956355642052,
+      "grad_norm": 0.15788938105106354,
+      "learning_rate": 4.4927898747158095e-05,
+      "loss": 0.424,
       "step": 31945
     },
     {
-      "epoch": 1.12,
-      "learning_rate": 4.5189935027935894e-05,
-      "loss": 0.2914,
+      "epoch": 1.1514758352254297,
+      "grad_norm": 0.16340243816375732,
+      "learning_rate": 4.4926136565651904e-05,
+      "loss": 0.4527,
       "step": 31950
     },
     {
-      "epoch": 1.12,
-      "learning_rate": 4.518825490845185e-05,
-      "loss": 0.2768,
+      "epoch": 1.1516560348866545,
+      "grad_norm": 0.16550558805465698,
+      "learning_rate": 4.492437411265596e-05,
+      "loss": 0.4677,
       "step": 31955
     },
     {
-      "epoch": 1.12,
-      "learning_rate": 4.518657452683663e-05,
-      "loss": 0.2805,
+      "epoch": 1.151836234547879,
+      "grad_norm": 0.14602138102054596,
+      "learning_rate": 4.492261138819427e-05,
+      "loss": 0.4225,
       "step": 31960
     },
     {
-      "epoch": 1.12,
-      "learning_rate": 4.518489388311205e-05,
-      "loss": 0.3325,
+      "epoch": 1.1520164342091037,
+      "grad_norm": 0.19070053100585938,
+      "learning_rate": 4.492084839229086e-05,
+      "loss": 0.4458,
       "step": 31965
     },
     {
-      "epoch": 1.12,
-      "learning_rate": 4.518321297729991e-05,
-      "loss": 0.3182,
+      "epoch": 1.1521966338703282,
+      "grad_norm": 0.18122611939907074,
+      "learning_rate": 4.491908512496975e-05,
+      "loss": 0.4202,
       "step": 31970
     },
     {
-      "epoch": 1.12,
-      "learning_rate": 4.518153180942206e-05,
-      "loss": 0.2958,
+      "epoch": 1.152376833531553,
+      "grad_norm": 0.1686059683561325,
+      "learning_rate": 4.491732158625496e-05,
+      "loss": 0.4264,
       "step": 31975
     },
     {
-      "epoch": 1.13,
-      "learning_rate": 4.517985037950032e-05,
-      "loss": 0.2981,
+      "epoch": 1.1525570331927777,
+      "grad_norm": 0.15016207098960876,
+      "learning_rate": 4.491555777617051e-05,
+      "loss": 0.4277,
       "step": 31980
     },
     {
-      "epoch": 1.13,
-      "learning_rate": 4.5178168687556525e-05,
-      "loss": 0.2792,
+      "epoch": 1.1527372328540022,
+      "grad_norm": 0.15645290911197662,
+      "learning_rate": 4.491379369474046e-05,
+      "loss": 0.4294,
       "step": 31985
     },
     {
-      "epoch": 1.13,
-      "learning_rate": 4.5176486733612497e-05,
-      "loss": 0.2914,
+      "epoch": 1.152917432515227,
+      "grad_norm": 0.14731116592884064,
+      "learning_rate": 4.491202934198881e-05,
+      "loss": 0.3946,
       "step": 31990
     },
     {
-      "epoch": 1.13,
-      "learning_rate": 4.517480451769009e-05,
-      "loss": 0.3029,
+      "epoch": 1.1530976321764514,
+      "grad_norm": 0.1814972460269928,
+      "learning_rate": 4.491026471793962e-05,
+      "loss": 0.4306,
       "step": 31995
     },
     {
-      "epoch": 1.13,
-      "learning_rate": 4.517312203981114e-05,
-      "loss": 0.2882,
+      "epoch": 1.1532778318376762,
+      "grad_norm": 0.2180231809616089,
+      "learning_rate": 4.4908499822616934e-05,
+      "loss": 0.421,
       "step": 32000
     },
     {
-      "epoch": 1.13,
-      "eval_loss": 0.2893707752227783,
-      "eval_runtime": 10.5373,
-      "eval_samples_per_second": 9.49,
-      "eval_steps_per_second": 9.49,
+      "epoch": 1.1532778318376762,
+      "eval_loss": 0.4514605402946472,
+      "eval_runtime": 3.5405,
+      "eval_samples_per_second": 28.244,
+      "eval_steps_per_second": 7.061,
       "step": 32000
     },
     {
-      "epoch": 1.13,
-      "learning_rate": 4.51714392999975e-05,
-      "loss": 0.2729,
+      "epoch": 1.1534580314989007,
+      "grad_norm": 0.18036217987537384,
+      "learning_rate": 4.4906734656044786e-05,
+      "loss": 0.4261,
       "step": 32005
     },
     {
-      "epoch": 1.13,
-      "learning_rate": 4.516975629827101e-05,
-      "loss": 0.3085,
+      "epoch": 1.1536382311601254,
+      "grad_norm": 0.2106885462999344,
+      "learning_rate": 4.4904969218247234e-05,
+      "loss": 0.3771,
       "step": 32010
     },
     {
-      "epoch": 1.13,
-      "learning_rate": 4.516807303465353e-05,
-      "loss": 0.2839,
+      "epoch": 1.15381843082135,
+      "grad_norm": 0.15591959655284882,
+      "learning_rate": 4.490320350924833e-05,
+      "loss": 0.4009,
       "step": 32015
     },
     {
-      "epoch": 1.13,
-      "learning_rate": 4.516638950916691e-05,
-      "loss": 0.2747,
+      "epoch": 1.1539986304825747,
+      "grad_norm": 0.14869220554828644,
+      "learning_rate": 4.490143752907213e-05,
+      "loss": 0.4101,
       "step": 32020
     },
     {
-      "epoch": 1.13,
-      "learning_rate": 4.516470572183302e-05,
-      "loss": 0.2748,
+      "epoch": 1.1541788301437994,
+      "grad_norm": 0.1626279205083847,
+      "learning_rate": 4.489967127774271e-05,
+      "loss": 0.4174,
       "step": 32025
     },
     {
-      "epoch": 1.13,
-      "learning_rate": 4.5163021672673714e-05,
-      "loss": 0.2887,
+      "epoch": 1.154359029805024,
+      "grad_norm": 0.1796058565378189,
+      "learning_rate": 4.48979047552841e-05,
+      "loss": 0.4132,
       "step": 32030
     },
     {
-      "epoch": 1.13,
-      "learning_rate": 4.516133736171084e-05,
-      "loss": 0.3063,
+      "epoch": 1.1545392294662487,
+      "grad_norm": 0.16110989451408386,
+      "learning_rate": 4.48961379617204e-05,
+      "loss": 0.4238,
       "step": 32035
     },
     {
-      "epoch": 1.13,
-      "learning_rate": 4.5159652788966314e-05,
-      "loss": 0.2861,
+      "epoch": 1.1547194291274732,
+      "grad_norm": 0.17553108930587769,
+      "learning_rate": 4.489437089707568e-05,
+      "loss": 0.3808,
       "step": 32040
     },
     {
-      "epoch": 1.13,
-      "learning_rate": 4.5157967954461967e-05,
-      "loss": 0.2965,
+      "epoch": 1.154899628788698,
+      "grad_norm": 0.289981484413147,
+      "learning_rate": 4.489260356137399e-05,
+      "loss": 0.4202,
       "step": 32045
     },
     {
-      "epoch": 1.13,
-      "learning_rate": 4.515628285821969e-05,
-      "loss": 0.3038,
+      "epoch": 1.1550798284499226,
+      "grad_norm": 0.162216454744339,
+      "learning_rate": 4.489083595463944e-05,
+      "loss": 0.409,
       "step": 32050
     },
     {
-      "epoch": 1.13,
-      "learning_rate": 4.515459750026137e-05,
-      "loss": 0.288,
+      "epoch": 1.1552600281111471,
+      "grad_norm": 0.18123595416545868,
+      "learning_rate": 4.48890680768961e-05,
+      "loss": 0.4316,
       "step": 32055
     },
     {
-      "epoch": 1.13,
-      "learning_rate": 4.515291188060888e-05,
-      "loss": 0.2806,
+      "epoch": 1.1554402277723719,
+      "grad_norm": 0.22130350768566132,
+      "learning_rate": 4.488729992816806e-05,
+      "loss": 0.4034,
       "step": 32060
     },
     {
-      "epoch": 1.13,
-      "learning_rate": 4.5151225999284115e-05,
-      "loss": 0.3051,
+      "epoch": 1.1556204274335964,
+      "grad_norm": 0.15505443513393402,
+      "learning_rate": 4.488553150847941e-05,
+      "loss": 0.3919,
       "step": 32065
     },
     {
-      "epoch": 1.13,
-      "learning_rate": 4.514953985630895e-05,
-      "loss": 0.3021,
+      "epoch": 1.1558006270948211,
+      "grad_norm": 0.17959943413734436,
+      "learning_rate": 4.4883762817854236e-05,
+      "loss": 0.4241,
       "step": 32070
     },
     {
-      "epoch": 1.13,
-      "learning_rate": 4.51478534517053e-05,
-      "loss": 0.2987,
+      "epoch": 1.1559808267560456,
+      "grad_norm": 0.18637138605117798,
+      "learning_rate": 4.488199385631665e-05,
+      "loss": 0.4389,
       "step": 32075
     },
     {
-      "epoch": 1.13,
-      "learning_rate": 4.514616678549505e-05,
-      "loss": 0.2964,
+      "epoch": 1.1561610264172704,
+      "grad_norm": 0.16619151830673218,
+      "learning_rate": 4.488022462389074e-05,
+      "loss": 0.4582,
       "step": 32080
     },
     {
-      "epoch": 1.13,
-      "learning_rate": 4.51444798577001e-05,
-      "loss": 0.273,
+      "epoch": 1.1563412260784949,
+      "grad_norm": 0.22853337228298187,
+      "learning_rate": 4.487845512060063e-05,
+      "loss": 0.4317,
       "step": 32085
     },
     {
-      "epoch": 1.13,
-      "learning_rate": 4.514279266834235e-05,
-      "loss": 0.2763,
+      "epoch": 1.1565214257397196,
+      "grad_norm": 0.15449535846710205,
+      "learning_rate": 4.487668534647041e-05,
+      "loss": 0.4514,
       "step": 32090
     },
     {
-      "epoch": 1.13,
-      "learning_rate": 4.514110521744371e-05,
-      "loss": 0.2772,
+      "epoch": 1.1567016254009443,
+      "grad_norm": 0.18602608144283295,
+      "learning_rate": 4.4874915301524194e-05,
+      "loss": 0.4577,
       "step": 32095
     },
     {
-      "epoch": 1.13,
-      "learning_rate": 4.5139417505026105e-05,
-      "loss": 0.2844,
+      "epoch": 1.1568818250621689,
+      "grad_norm": 0.178580179810524,
+      "learning_rate": 4.487314498578611e-05,
+      "loss": 0.4175,
       "step": 32100
     },
     {
-      "epoch": 1.13,
-      "learning_rate": 4.513772953111143e-05,
-      "loss": 0.2938,
+      "epoch": 1.1570620247233936,
+      "grad_norm": 0.17809760570526123,
+      "learning_rate": 4.4871374399280273e-05,
+      "loss": 0.4256,
       "step": 32105
     },
     {
-      "epoch": 1.13,
-      "learning_rate": 4.5136041295721606e-05,
-      "loss": 0.3007,
+      "epoch": 1.157242224384618,
+      "grad_norm": 0.16492916643619537,
+      "learning_rate": 4.4869603542030806e-05,
+      "loss": 0.4098,
       "step": 32110
     },
     {
-      "epoch": 1.13,
-      "learning_rate": 4.5134352798878556e-05,
-      "loss": 0.3043,
+      "epoch": 1.1574224240458428,
+      "grad_norm": 0.1905374526977539,
+      "learning_rate": 4.486783241406184e-05,
+      "loss": 0.4009,
       "step": 32115
     },
     {
-      "epoch": 1.13,
-      "learning_rate": 4.51326640406042e-05,
-      "loss": 0.2907,
+      "epoch": 1.1576026237070673,
+      "grad_norm": 0.16066482663154602,
+      "learning_rate": 4.486606101539751e-05,
+      "loss": 0.398,
       "step": 32120
     },
     {
-      "epoch": 1.13,
-      "learning_rate": 4.513097502092047e-05,
-      "loss": 0.3222,
+      "epoch": 1.157782823368292,
+      "grad_norm": 0.1904597133398056,
+      "learning_rate": 4.486428934606194e-05,
+      "loss": 0.405,
       "step": 32125
     },
     {
-      "epoch": 1.13,
-      "learning_rate": 4.5129285739849304e-05,
-      "loss": 0.3052,
+      "epoch": 1.1579630230295166,
+      "grad_norm": 0.16945073008537292,
+      "learning_rate": 4.486251740607927e-05,
+      "loss": 0.4205,
       "step": 32130
     },
     {
-      "epoch": 1.13,
-      "learning_rate": 4.512759619741262e-05,
-      "loss": 0.2914,
+      "epoch": 1.1581432226907413,
+      "grad_norm": 0.19391489028930664,
+      "learning_rate": 4.486074519547365e-05,
+      "loss": 0.4073,
       "step": 32135
     },
     {
-      "epoch": 1.13,
-      "learning_rate": 4.5125906393632376e-05,
-      "loss": 0.3115,
+      "epoch": 1.158323422351966,
+      "grad_norm": 0.13818098604679108,
+      "learning_rate": 4.4858972714269215e-05,
+      "loss": 0.3963,
       "step": 32140
     },
     {
-      "epoch": 1.13,
-      "learning_rate": 4.512421632853049e-05,
-      "loss": 0.2945,
+      "epoch": 1.1585036220131906,
+      "grad_norm": 0.1501578539609909,
+      "learning_rate": 4.485719996249012e-05,
+      "loss": 0.4193,
       "step": 32145
     },
     {
-      "epoch": 1.13,
-      "learning_rate": 4.5122526002128924e-05,
-      "loss": 0.3217,
+      "epoch": 1.1586838216744153,
+      "grad_norm": 0.19974437355995178,
+      "learning_rate": 4.4855426940160536e-05,
+      "loss": 0.4491,
       "step": 32150
     },
     {
-      "epoch": 1.13,
-      "learning_rate": 4.5120835414449615e-05,
-      "loss": 0.3123,
+      "epoch": 1.1588640213356398,
+      "grad_norm": 0.1437712460756302,
+      "learning_rate": 4.4853653647304596e-05,
+      "loss": 0.4116,
       "step": 32155
     },
     {
-      "epoch": 1.13,
-      "learning_rate": 4.5119144565514514e-05,
-      "loss": 0.2885,
+      "epoch": 1.1590442209968645,
+      "grad_norm": 0.22510062158107758,
+      "learning_rate": 4.485188008394646e-05,
+      "loss": 0.419,
       "step": 32160
     },
     {
-      "epoch": 1.13,
-      "learning_rate": 4.5117453455345594e-05,
-      "loss": 0.2797,
+      "epoch": 1.1592244206580893,
+      "grad_norm": 0.16433167457580566,
+      "learning_rate": 4.485010625011031e-05,
+      "loss": 0.4199,
       "step": 32165
     },
     {
-      "epoch": 1.13,
-      "learning_rate": 4.5115762083964786e-05,
-      "loss": 0.2813,
+      "epoch": 1.1594046203193138,
+      "grad_norm": 0.18741507828235626,
+      "learning_rate": 4.48483321458203e-05,
+      "loss": 0.4125,
       "step": 32170
     },
     {
-      "epoch": 1.13,
-      "learning_rate": 4.511407045139406e-05,
-      "loss": 0.3063,
+      "epoch": 1.1595848199805385,
+      "grad_norm": 0.171631321310997,
+      "learning_rate": 4.484655777110062e-05,
+      "loss": 0.4034,
       "step": 32175
     },
     {
-      "epoch": 1.13,
-      "learning_rate": 4.51123785576554e-05,
-      "loss": 0.2874,
+      "epoch": 1.159765019641763,
+      "grad_norm": 0.20612791180610657,
+      "learning_rate": 4.484478312597542e-05,
+      "loss": 0.4072,
       "step": 32180
     },
     {
-      "epoch": 1.13,
-      "learning_rate": 4.511068640277075e-05,
-      "loss": 0.2988,
+      "epoch": 1.1599452193029878,
+      "grad_norm": 0.162516251206398,
+      "learning_rate": 4.4843008210468896e-05,
+      "loss": 0.4092,
       "step": 32185
     },
     {
-      "epoch": 1.13,
-      "learning_rate": 4.5108993986762096e-05,
-      "loss": 0.2857,
+      "epoch": 1.1601254189642123,
+      "grad_norm": 0.17401380836963654,
+      "learning_rate": 4.484123302460523e-05,
+      "loss": 0.3899,
       "step": 32190
     },
     {
-      "epoch": 1.13,
-      "learning_rate": 4.5107301309651404e-05,
-      "loss": 0.2855,
+      "epoch": 1.160305618625437,
+      "grad_norm": 0.1766970008611679,
+      "learning_rate": 4.48394575684086e-05,
+      "loss": 0.4243,
       "step": 32195
     },
     {
-      "epoch": 1.13,
-      "learning_rate": 4.5105608371460654e-05,
-      "loss": 0.3222,
+      "epoch": 1.1604858182866615,
+      "grad_norm": 0.17945876717567444,
+      "learning_rate": 4.483768184190321e-05,
+      "loss": 0.4148,
       "step": 32200
     },
     {
-      "epoch": 1.13,
-      "learning_rate": 4.510391517221183e-05,
-      "loss": 0.289,
+      "epoch": 1.1606660179478863,
+      "grad_norm": 0.17272137105464935,
+      "learning_rate": 4.4835905845113234e-05,
+      "loss": 0.4174,
       "step": 32205
     },
     {
-      "epoch": 1.13,
-      "learning_rate": 4.510222171192692e-05,
-      "loss": 0.2758,
+      "epoch": 1.160846217609111,
+      "grad_norm": 0.14284390211105347,
+      "learning_rate": 4.483412957806289e-05,
+      "loss": 0.4218,
       "step": 32210
     },
     {
-      "epoch": 1.13,
-      "learning_rate": 4.51005279906279e-05,
-      "loss": 0.3126,
+      "epoch": 1.1610264172703355,
+      "grad_norm": 0.18068622052669525,
+      "learning_rate": 4.483235304077636e-05,
+      "loss": 0.4128,
       "step": 32215
     },
     {
-      "epoch": 1.13,
-      "learning_rate": 4.509883400833678e-05,
-      "loss": 0.2726,
+      "epoch": 1.1612066169315602,
+      "grad_norm": 0.17758698761463165,
+      "learning_rate": 4.483057623327787e-05,
+      "loss": 0.4533,
       "step": 32220
     },
     {
-      "epoch": 1.13,
-      "learning_rate": 4.509713976507554e-05,
-      "loss": 0.2743,
+      "epoch": 1.1613868165927848,
+      "grad_norm": 0.17569084465503693,
+      "learning_rate": 4.4828799155591615e-05,
+      "loss": 0.448,
       "step": 32225
     },
     {
-      "epoch": 1.13,
-      "learning_rate": 4.509544526086619e-05,
-      "loss": 0.2863,
+      "epoch": 1.1615670162540095,
+      "grad_norm": 0.18562377989292145,
+      "learning_rate": 4.4827021807741806e-05,
+      "loss": 0.4072,
       "step": 32230
     },
     {
-      "epoch": 1.13,
-      "learning_rate": 4.509375049573072e-05,
-      "loss": 0.2754,
+      "epoch": 1.161747215915234,
+      "grad_norm": 0.19902653992176056,
+      "learning_rate": 4.482524418975267e-05,
+      "loss": 0.4286,
       "step": 32235
     },
     {
-      "epoch": 1.13,
-      "learning_rate": 4.5092055469691145e-05,
-      "loss": 0.3209,
+      "epoch": 1.1619274155764587,
+      "grad_norm": 0.17017509043216705,
+      "learning_rate": 4.482346630164842e-05,
+      "loss": 0.4172,
       "step": 32240
     },
     {
-      "epoch": 1.13,
-      "learning_rate": 4.509036018276948e-05,
-      "loss": 0.2833,
+      "epoch": 1.1621076152376832,
+      "grad_norm": 0.14589948952198029,
+      "learning_rate": 4.4821688143453275e-05,
+      "loss": 0.4425,
       "step": 32245
     },
     {
-      "epoch": 1.13,
-      "learning_rate": 4.508866463498772e-05,
-      "loss": 0.2931,
+      "epoch": 1.162287814898908,
+      "grad_norm": 0.17183731496334076,
+      "learning_rate": 4.4819909715191475e-05,
+      "loss": 0.4243,
       "step": 32250
     },
     {
-      "epoch": 1.13,
-      "learning_rate": 4.508696882636788e-05,
-      "loss": 0.3042,
+      "epoch": 1.1624680145601327,
+      "grad_norm": 0.14236903190612793,
+      "learning_rate": 4.481813101688723e-05,
+      "loss": 0.4096,
       "step": 32255
     },
     {
-      "epoch": 1.13,
-      "learning_rate": 4.5085272756932e-05,
-      "loss": 0.282,
+      "epoch": 1.1626482142213572,
+      "grad_norm": 0.21945489943027496,
+      "learning_rate": 4.4816352048564806e-05,
+      "loss": 0.4207,
       "step": 32260
     },
     {
-      "epoch": 1.14,
-      "learning_rate": 4.508357642670209e-05,
-      "loss": 0.2859,
+      "epoch": 1.162828413882582,
+      "grad_norm": 0.21350398659706116,
+      "learning_rate": 4.48145728102484e-05,
+      "loss": 0.4225,
       "step": 32265
     },
     {
-      "epoch": 1.14,
-      "learning_rate": 4.508187983570017e-05,
-      "loss": 0.2901,
+      "epoch": 1.1630086135438065,
+      "grad_norm": 0.11839442700147629,
+      "learning_rate": 4.481279330196229e-05,
+      "loss": 0.3818,
       "step": 32270
     },
     {
-      "epoch": 1.14,
-      "learning_rate": 4.5080182983948274e-05,
-      "loss": 0.3145,
+      "epoch": 1.1631888132050312,
+      "grad_norm": 0.16868416965007782,
+      "learning_rate": 4.481101352373071e-05,
+      "loss": 0.4231,
       "step": 32275
     },
     {
-      "epoch": 1.14,
-      "learning_rate": 4.5078485871468436e-05,
-      "loss": 0.2789,
+      "epoch": 1.163369012866256,
+      "grad_norm": 0.1809322088956833,
+      "learning_rate": 4.480923347557789e-05,
+      "loss": 0.4284,
       "step": 32280
     },
     {
-      "epoch": 1.14,
-      "learning_rate": 4.50767884982827e-05,
-      "loss": 0.3104,
+      "epoch": 1.1635492125274804,
+      "grad_norm": 0.2081749439239502,
+      "learning_rate": 4.48074531575281e-05,
+      "loss": 0.4275,
       "step": 32285
     },
     {
-      "epoch": 1.14,
-      "learning_rate": 4.5075090864413085e-05,
-      "loss": 0.2759,
+      "epoch": 1.1637294121887052,
+      "grad_norm": 0.17321893572807312,
+      "learning_rate": 4.480567256960561e-05,
+      "loss": 0.4338,
       "step": 32290
     },
     {
-      "epoch": 1.14,
-      "learning_rate": 4.507339296988165e-05,
-      "loss": 0.3297,
+      "epoch": 1.1639096118499297,
+      "grad_norm": 0.19908635318279266,
+      "learning_rate": 4.480389171183466e-05,
+      "loss": 0.3936,
       "step": 32295
     },
     {
-      "epoch": 1.14,
-      "learning_rate": 4.507169481471044e-05,
-      "loss": 0.2993,
+      "epoch": 1.1640898115111544,
+      "grad_norm": 0.19130414724349976,
+      "learning_rate": 4.480211058423952e-05,
+      "loss": 0.4122,
       "step": 32300
     },
     {
-      "epoch": 1.14,
-      "learning_rate": 4.506999639892149e-05,
-      "loss": 0.3079,
+      "epoch": 1.164270011172379,
+      "grad_norm": 0.20319342613220215,
+      "learning_rate": 4.480032918684446e-05,
+      "loss": 0.3795,
       "step": 32305
     },
     {
-      "epoch": 1.14,
-      "learning_rate": 4.506829772253687e-05,
-      "loss": 0.3061,
+      "epoch": 1.1644502108336037,
+      "grad_norm": 0.19457338750362396,
+      "learning_rate": 4.479854751967374e-05,
+      "loss": 0.4287,
       "step": 32310
     },
     {
-      "epoch": 1.14,
-      "learning_rate": 4.506659878557863e-05,
-      "loss": 0.2899,
+      "epoch": 1.1646304104948282,
+      "grad_norm": 0.14478561282157898,
+      "learning_rate": 4.479676558275164e-05,
+      "loss": 0.413,
       "step": 32315
     },
     {
-      "epoch": 1.14,
-      "learning_rate": 4.506489958806883e-05,
-      "loss": 0.3119,
+      "epoch": 1.164810610156053,
+      "grad_norm": 0.19652974605560303,
+      "learning_rate": 4.479498337610246e-05,
+      "loss": 0.4123,
       "step": 32320
     },
     {
-      "epoch": 1.14,
-      "learning_rate": 4.506320013002953e-05,
-      "loss": 0.2938,
+      "epoch": 1.1649908098172777,
+      "grad_norm": 0.14270542562007904,
+      "learning_rate": 4.4793200899750445e-05,
+      "loss": 0.4011,
       "step": 32325
     },
     {
-      "epoch": 1.14,
-      "learning_rate": 4.506150041148281e-05,
-      "loss": 0.3046,
+      "epoch": 1.1651710094785022,
+      "grad_norm": 0.16907188296318054,
+      "learning_rate": 4.4791418153719914e-05,
+      "loss": 0.4427,
       "step": 32330
     },
     {
-      "epoch": 1.14,
-      "learning_rate": 4.5059800432450705e-05,
-      "loss": 0.2817,
+      "epoch": 1.165351209139727,
+      "grad_norm": 0.19798724353313446,
+      "learning_rate": 4.4789635138035137e-05,
+      "loss": 0.4178,
       "step": 32335
     },
     {
-      "epoch": 1.14,
-      "learning_rate": 4.5058100192955327e-05,
-      "loss": 0.279,
+      "epoch": 1.1655314088009514,
+      "grad_norm": 0.1947140246629715,
+      "learning_rate": 4.4787851852720405e-05,
+      "loss": 0.405,
       "step": 32340
     },
     {
-      "epoch": 1.14,
-      "learning_rate": 4.505639969301874e-05,
-      "loss": 0.2616,
+      "epoch": 1.1657116084621761,
+      "grad_norm": 0.19974881410598755,
+      "learning_rate": 4.4786068297800034e-05,
+      "loss": 0.4248,
       "step": 32345
     },
     {
-      "epoch": 1.14,
-      "learning_rate": 4.5054698932663006e-05,
-      "loss": 0.287,
+      "epoch": 1.1658918081234007,
+      "grad_norm": 0.23895248770713806,
+      "learning_rate": 4.478428447329831e-05,
+      "loss": 0.4351,
       "step": 32350
     },
     {
-      "epoch": 1.14,
-      "learning_rate": 4.505299791191023e-05,
-      "loss": 0.292,
+      "epoch": 1.1660720077846254,
+      "grad_norm": 0.1716042160987854,
+      "learning_rate": 4.4782500379239534e-05,
+      "loss": 0.4104,
       "step": 32355
     },
     {
-      "epoch": 1.14,
-      "learning_rate": 4.50512966307825e-05,
-      "loss": 0.3002,
+      "epoch": 1.16625220744585,
+      "grad_norm": 0.14665000140666962,
+      "learning_rate": 4.4780716015648026e-05,
+      "loss": 0.4237,
       "step": 32360
     },
     {
-      "epoch": 1.14,
-      "learning_rate": 4.504959508930189e-05,
-      "loss": 0.2942,
+      "epoch": 1.1664324071070746,
+      "grad_norm": 0.1704597920179367,
+      "learning_rate": 4.4778931382548096e-05,
+      "loss": 0.4007,
       "step": 32365
     },
     {
-      "epoch": 1.14,
-      "learning_rate": 4.50478932874905e-05,
-      "loss": 0.3056,
+      "epoch": 1.1666126067682994,
+      "grad_norm": 0.17646756768226624,
+      "learning_rate": 4.477714647996405e-05,
+      "loss": 0.4092,
       "step": 32370
     },
     {
-      "epoch": 1.14,
-      "learning_rate": 4.504619122537043e-05,
-      "loss": 0.2633,
+      "epoch": 1.1667928064295239,
+      "grad_norm": 0.13408124446868896,
+      "learning_rate": 4.4775361307920205e-05,
+      "loss": 0.4013,
       "step": 32375
     },
     {
-      "epoch": 1.14,
-      "learning_rate": 4.504448890296377e-05,
-      "loss": 0.2736,
+      "epoch": 1.1669730060907486,
+      "grad_norm": 0.1914110779762268,
+      "learning_rate": 4.47735758664409e-05,
+      "loss": 0.4323,
       "step": 32380
     },
     {
-      "epoch": 1.14,
-      "learning_rate": 4.504278632029264e-05,
-      "loss": 0.2809,
+      "epoch": 1.1671532057519731,
+      "grad_norm": 0.19718259572982788,
+      "learning_rate": 4.4771790155550455e-05,
+      "loss": 0.4199,
       "step": 32385
     },
     {
-      "epoch": 1.14,
-      "learning_rate": 4.5041083477379134e-05,
-      "loss": 0.3094,
+      "epoch": 1.1673334054131979,
+      "grad_norm": 0.19468161463737488,
+      "learning_rate": 4.477000417527319e-05,
+      "loss": 0.4024,
       "step": 32390
     },
     {
-      "epoch": 1.14,
-      "learning_rate": 4.503938037424537e-05,
-      "loss": 0.2886,
+      "epoch": 1.1675136050744224,
+      "grad_norm": 0.16953888535499573,
+      "learning_rate": 4.476821792563345e-05,
+      "loss": 0.4586,
       "step": 32395
     },
     {
-      "epoch": 1.14,
-      "learning_rate": 4.503767701091346e-05,
-      "loss": 0.2882,
+      "epoch": 1.167693804735647,
+      "grad_norm": 0.18917879462242126,
+      "learning_rate": 4.4766788731997023e-05,
+      "loss": 0.4336,
       "step": 32400
     },
     {
-      "epoch": 1.14,
-      "learning_rate": 4.5035973387405514e-05,
-      "loss": 0.3187,
+      "epoch": 1.1678740043968716,
+      "grad_norm": 0.16631363332271576,
+      "learning_rate": 4.476500199756615e-05,
+      "loss": 0.3989,
       "step": 32405
     },
     {
-      "epoch": 1.14,
-      "learning_rate": 4.5034269503743665e-05,
-      "loss": 0.2848,
+      "epoch": 1.1680542040580963,
+      "grad_norm": 0.17123185098171234,
+      "learning_rate": 4.476321499384096e-05,
+      "loss": 0.4394,
       "step": 32410
     },
     {
-      "epoch": 1.14,
-      "learning_rate": 4.5032565359950016e-05,
-      "loss": 0.3036,
+      "epoch": 1.168234403719321,
+      "grad_norm": 0.14406786859035492,
+      "learning_rate": 4.476142772084578e-05,
+      "loss": 0.4602,
       "step": 32415
     },
     {
-      "epoch": 1.14,
-      "learning_rate": 4.5030860956046725e-05,
-      "loss": 0.3061,
+      "epoch": 1.1684146033805456,
+      "grad_norm": 0.17469120025634766,
+      "learning_rate": 4.475964017860498e-05,
+      "loss": 0.4058,
       "step": 32420
     },
     {
-      "epoch": 1.14,
-      "learning_rate": 4.5029156292055895e-05,
-      "loss": 0.2894,
+      "epoch": 1.1685948030417703,
+      "grad_norm": 0.15574227273464203,
+      "learning_rate": 4.47578523671429e-05,
+      "loss": 0.4586,
       "step": 32425
     },
     {
-      "epoch": 1.14,
-      "learning_rate": 4.502745136799967e-05,
-      "loss": 0.2715,
+      "epoch": 1.1687750027029948,
+      "grad_norm": 0.22920864820480347,
+      "learning_rate": 4.4756064286483915e-05,
+      "loss": 0.4453,
       "step": 32430
     },
     {
-      "epoch": 1.14,
-      "learning_rate": 4.5025746183900195e-05,
-      "loss": 0.3355,
+      "epoch": 1.1689552023642196,
+      "grad_norm": 0.19061683118343353,
+      "learning_rate": 4.475427593665238e-05,
+      "loss": 0.4672,
       "step": 32435
     },
     {
-      "epoch": 1.14,
-      "learning_rate": 4.502404073977959e-05,
-      "loss": 0.2947,
+      "epoch": 1.1691354020254443,
+      "grad_norm": 0.20378006994724274,
+      "learning_rate": 4.475248731767265e-05,
+      "loss": 0.4179,
       "step": 32440
     },
     {
-      "epoch": 1.14,
-      "learning_rate": 4.5022335035660026e-05,
-      "loss": 0.2973,
+      "epoch": 1.1693156016866688,
+      "grad_norm": 0.16971811652183533,
+      "learning_rate": 4.475069842956911e-05,
+      "loss": 0.397,
       "step": 32445
     },
     {
-      "epoch": 1.14,
-      "learning_rate": 4.502062907156363e-05,
-      "loss": 0.304,
+      "epoch": 1.1694958013478935,
+      "grad_norm": 0.17052094638347626,
+      "learning_rate": 4.4748909272366133e-05,
+      "loss": 0.3946,
       "step": 32450
     },
     {
-      "epoch": 1.14,
-      "learning_rate": 4.5018922847512555e-05,
-      "loss": 0.3056,
+      "epoch": 1.169676001009118,
+      "grad_norm": 0.14758871495723724,
+      "learning_rate": 4.474711984608809e-05,
+      "loss": 0.3939,
       "step": 32455
     },
     {
-      "epoch": 1.14,
-      "learning_rate": 4.501721636352897e-05,
-      "loss": 0.2712,
+      "epoch": 1.1698562006703428,
+      "grad_norm": 0.18905098736286163,
+      "learning_rate": 4.474533015075936e-05,
+      "loss": 0.4125,
       "step": 32460
     },
     {
-      "epoch": 1.14,
-      "learning_rate": 4.501550961963502e-05,
-      "loss": 0.275,
+      "epoch": 1.1700364003315673,
+      "grad_norm": 0.1888333261013031,
+      "learning_rate": 4.474354018640432e-05,
+      "loss": 0.4127,
       "step": 32465
     },
     {
-      "epoch": 1.14,
-      "learning_rate": 4.501380261585286e-05,
-      "loss": 0.2669,
+      "epoch": 1.170216599992792,
+      "grad_norm": 0.15173350274562836,
+      "learning_rate": 4.4741749953047374e-05,
+      "loss": 0.4341,
       "step": 32470
     },
     {
-      "epoch": 1.14,
-      "learning_rate": 4.501209535220468e-05,
-      "loss": 0.3307,
+      "epoch": 1.1703967996540166,
+      "grad_norm": 0.18733574450016022,
+      "learning_rate": 4.473995945071291e-05,
+      "loss": 0.4164,
       "step": 32475
     },
     {
-      "epoch": 1.14,
-      "learning_rate": 4.501038782871261e-05,
-      "loss": 0.2873,
+      "epoch": 1.1705769993152413,
+      "grad_norm": 0.18692295253276825,
+      "learning_rate": 4.473816867942532e-05,
+      "loss": 0.4247,
       "step": 32480
     },
     {
-      "epoch": 1.14,
-      "learning_rate": 4.500868004539885e-05,
-      "loss": 0.3082,
+      "epoch": 1.170757198976466,
+      "grad_norm": 0.22081652283668518,
+      "learning_rate": 4.473637763920901e-05,
+      "loss": 0.4228,
       "step": 32485
     },
     {
-      "epoch": 1.14,
-      "learning_rate": 4.5006972002285574e-05,
-      "loss": 0.2735,
+      "epoch": 1.1709373986376905,
+      "grad_norm": 0.15377368032932281,
+      "learning_rate": 4.4734586330088365e-05,
+      "loss": 0.381,
       "step": 32490
     },
     {
-      "epoch": 1.14,
-      "learning_rate": 4.500526369939494e-05,
-      "loss": 0.3092,
+      "epoch": 1.1711175982989153,
+      "grad_norm": 0.17810076475143433,
+      "learning_rate": 4.47327947520878e-05,
+      "loss": 0.413,
       "step": 32495
     },
     {
-      "epoch": 1.14,
-      "learning_rate": 4.500355513674914e-05,
-      "loss": 0.297,
+      "epoch": 1.1712977979601398,
+      "grad_norm": 0.17749875783920288,
+      "learning_rate": 4.4731002905231735e-05,
+      "loss": 0.4012,
       "step": 32500
     },
     {
-      "epoch": 1.14,
-      "eval_loss": 0.28957316279411316,
-      "eval_runtime": 10.558,
-      "eval_samples_per_second": 9.471,
-      "eval_steps_per_second": 9.471,
+      "epoch": 1.1712977979601398,
+      "eval_loss": 0.4519113600254059,
+      "eval_runtime": 3.5356,
+      "eval_samples_per_second": 28.284,
+      "eval_steps_per_second": 7.071,
       "step": 32500
     },
     {
-      "epoch": 1.14,
-      "learning_rate": 4.500184631437037e-05,
-      "loss": 0.2703,
+      "epoch": 1.1714779976213645,
+      "grad_norm": 0.20969074964523315,
+      "learning_rate": 4.4729210789544576e-05,
+      "loss": 0.4399,
       "step": 32505
     },
     {
-      "epoch": 1.14,
-      "learning_rate": 4.5000137232280806e-05,
-      "loss": 0.2789,
+      "epoch": 1.171658197282589,
+      "grad_norm": 0.17028845846652985,
+      "learning_rate": 4.472741840505073e-05,
+      "loss": 0.4074,
       "step": 32510
     },
     {
-      "epoch": 1.14,
-      "learning_rate": 4.499842789050263e-05,
-      "loss": 0.2644,
+      "epoch": 1.1718383969438138,
+      "grad_norm": 0.19405877590179443,
+      "learning_rate": 4.472562575177464e-05,
+      "loss": 0.4364,
       "step": 32515
     },
     {
-      "epoch": 1.14,
-      "learning_rate": 4.499671828905806e-05,
-      "loss": 0.286,
+      "epoch": 1.1720185966050383,
+      "grad_norm": 0.20871876180171967,
+      "learning_rate": 4.472383282974071e-05,
+      "loss": 0.4526,
       "step": 32520
     },
     {
-      "epoch": 1.14,
-      "learning_rate": 4.499500842796927e-05,
-      "loss": 0.285,
+      "epoch": 1.172198796266263,
+      "grad_norm": 0.16674499213695526,
+      "learning_rate": 4.4722039638973374e-05,
+      "loss": 0.4057,
       "step": 32525
     },
     {
-      "epoch": 1.14,
-      "learning_rate": 4.499329830725848e-05,
-      "loss": 0.2909,
+      "epoch": 1.1723789959274877,
+      "grad_norm": 0.17715777456760406,
+      "learning_rate": 4.4720246179497074e-05,
+      "loss": 0.4047,
       "step": 32530
     },
     {
-      "epoch": 1.14,
-      "learning_rate": 4.4991587926947886e-05,
-      "loss": 0.2899,
+      "epoch": 1.1725591955887122,
+      "grad_norm": 0.15576201677322388,
+      "learning_rate": 4.471845245133623e-05,
+      "loss": 0.4307,
       "step": 32535
     },
     {
-      "epoch": 1.14,
-      "learning_rate": 4.4989877287059694e-05,
-      "loss": 0.2982,
+      "epoch": 1.172739395249937,
+      "grad_norm": 0.19848684966564178,
+      "learning_rate": 4.4716658454515294e-05,
+      "loss": 0.4309,
       "step": 32540
     },
     {
-      "epoch": 1.15,
-      "learning_rate": 4.4988166387616125e-05,
-      "loss": 0.2857,
+      "epoch": 1.1729195949111615,
+      "grad_norm": 0.21592764556407928,
+      "learning_rate": 4.4714864189058705e-05,
+      "loss": 0.4146,
       "step": 32545
     },
     {
-      "epoch": 1.15,
-      "learning_rate": 4.498645522863938e-05,
-      "loss": 0.2861,
+      "epoch": 1.1730997945723862,
+      "grad_norm": 0.17954400181770325,
+      "learning_rate": 4.47130696549909e-05,
+      "loss": 0.4367,
       "step": 32550
     },
     {
-      "epoch": 1.15,
-      "learning_rate": 4.49847438101517e-05,
-      "loss": 0.2805,
+      "epoch": 1.173279994233611,
+      "grad_norm": 0.17298442125320435,
+      "learning_rate": 4.471127485233635e-05,
+      "loss": 0.3967,
       "step": 32555
     },
     {
-      "epoch": 1.15,
-      "learning_rate": 4.498303213217527e-05,
-      "loss": 0.3046,
+      "epoch": 1.1734601938948355,
+      "grad_norm": 0.17591506242752075,
+      "learning_rate": 4.4709479781119485e-05,
+      "loss": 0.4144,
       "step": 32560
     },
     {
-      "epoch": 1.15,
-      "learning_rate": 4.498132019473236e-05,
-      "loss": 0.2983,
+      "epoch": 1.1736403935560602,
+      "grad_norm": 0.1792951375246048,
+      "learning_rate": 4.470768444136478e-05,
+      "loss": 0.429,
       "step": 32565
     },
     {
-      "epoch": 1.15,
-      "learning_rate": 4.497960799784516e-05,
-      "loss": 0.2732,
+      "epoch": 1.1738205932172847,
+      "grad_norm": 0.19630871713161469,
+      "learning_rate": 4.470588883309669e-05,
+      "loss": 0.4472,
       "step": 32570
     },
     {
-      "epoch": 1.15,
-      "learning_rate": 4.497789554153593e-05,
-      "loss": 0.295,
+      "epoch": 1.1740007928785094,
+      "grad_norm": 0.1824900358915329,
+      "learning_rate": 4.470409295633968e-05,
+      "loss": 0.4047,
       "step": 32575
     },
     {
-      "epoch": 1.15,
-      "learning_rate": 4.4976182825826877e-05,
-      "loss": 0.2888,
+      "epoch": 1.174180992539734,
+      "grad_norm": 0.21130244433879852,
+      "learning_rate": 4.470229681111822e-05,
+      "loss": 0.4292,
       "step": 32580
     },
     {
-      "epoch": 1.15,
-      "learning_rate": 4.4974469850740264e-05,
-      "loss": 0.2696,
+      "epoch": 1.1743611922009587,
+      "grad_norm": 0.22852078080177307,
+      "learning_rate": 4.470050039745677e-05,
+      "loss": 0.4134,
       "step": 32585
     },
     {
-      "epoch": 1.15,
-      "learning_rate": 4.4972756616298323e-05,
-      "loss": 0.2795,
+      "epoch": 1.1745413918621832,
+      "grad_norm": 0.1561303436756134,
+      "learning_rate": 4.4698703715379823e-05,
+      "loss": 0.4225,
       "step": 32590
     },
     {
-      "epoch": 1.15,
-      "learning_rate": 4.49710431225233e-05,
-      "loss": 0.2967,
+      "epoch": 1.174721591523408,
+      "grad_norm": 0.19721920788288116,
+      "learning_rate": 4.4696906764911855e-05,
+      "loss": 0.4508,
       "step": 32595
     },
     {
-      "epoch": 1.15,
-      "learning_rate": 4.496932936943744e-05,
-      "loss": 0.2909,
+      "epoch": 1.1749017911846327,
+      "grad_norm": 0.19223082065582275,
+      "learning_rate": 4.469510954607734e-05,
+      "loss": 0.4363,
       "step": 32600
     },
     {
-      "epoch": 1.15,
-      "learning_rate": 4.4967615357063006e-05,
-      "loss": 0.2927,
+      "epoch": 1.1750819908458572,
+      "grad_norm": 0.15646933019161224,
+      "learning_rate": 4.4693312058900774e-05,
+      "loss": 0.4086,
       "step": 32605
     },
     {
-      "epoch": 1.15,
-      "learning_rate": 4.4965901085422235e-05,
-      "loss": 0.2756,
+      "epoch": 1.175262190507082,
+      "grad_norm": 0.19095897674560547,
+      "learning_rate": 4.469151430340665e-05,
+      "loss": 0.4511,
       "step": 32610
     },
     {
-      "epoch": 1.15,
-      "learning_rate": 4.49641865545374e-05,
-      "loss": 0.2814,
+      "epoch": 1.1754423901683064,
+      "grad_norm": 0.1620524823665619,
+      "learning_rate": 4.4689716279619444e-05,
+      "loss": 0.4414,
       "step": 32615
     },
     {
-      "epoch": 1.15,
-      "learning_rate": 4.496247176443077e-05,
-      "loss": 0.3063,
+      "epoch": 1.1756225898295312,
+      "grad_norm": 0.14250415563583374,
+      "learning_rate": 4.468791798756367e-05,
+      "loss": 0.4214,
       "step": 32620
     },
     {
-      "epoch": 1.15,
-      "learning_rate": 4.4960756715124586e-05,
-      "loss": 0.3235,
+      "epoch": 1.1758027894907557,
+      "grad_norm": 0.16954278945922852,
+      "learning_rate": 4.4686119427263826e-05,
+      "loss": 0.4095,
       "step": 32625
     },
     {
-      "epoch": 1.15,
-      "learning_rate": 4.495904140664114e-05,
-      "loss": 0.2832,
+      "epoch": 1.1759829891519804,
+      "grad_norm": 0.19082613289356232,
+      "learning_rate": 4.4684320598744426e-05,
+      "loss": 0.4455,
       "step": 32630
     },
     {
-      "epoch": 1.15,
-      "learning_rate": 4.495732583900269e-05,
-      "loss": 0.2814,
+      "epoch": 1.176163188813205,
+      "grad_norm": 0.1630188673734665,
+      "learning_rate": 4.468252150202995e-05,
+      "loss": 0.4229,
       "step": 32635
     },
     {
-      "epoch": 1.15,
-      "learning_rate": 4.495561001223152e-05,
-      "loss": 0.299,
+      "epoch": 1.1763433884744297,
+      "grad_norm": 0.20702870190143585,
+      "learning_rate": 4.4680722137144935e-05,
+      "loss": 0.4234,
       "step": 32640
     },
     {
-      "epoch": 1.15,
-      "learning_rate": 4.4953893926349896e-05,
-      "loss": 0.2991,
+      "epoch": 1.1765235881356544,
+      "grad_norm": 0.1635637730360031,
+      "learning_rate": 4.46789225041139e-05,
+      "loss": 0.3663,
       "step": 32645
     },
     {
-      "epoch": 1.15,
-      "learning_rate": 4.495217758138012e-05,
-      "loss": 0.2541,
+      "epoch": 1.176703787796879,
+      "grad_norm": 0.18059010803699493,
+      "learning_rate": 4.467712260296135e-05,
+      "loss": 0.3952,
       "step": 32650
     },
     {
-      "epoch": 1.15,
-      "learning_rate": 4.495046097734447e-05,
-      "loss": 0.2964,
+      "epoch": 1.1768839874581036,
+      "grad_norm": 0.18861469626426697,
+      "learning_rate": 4.467532243371181e-05,
+      "loss": 0.436,
       "step": 32655
     },
     {
-      "epoch": 1.15,
-      "learning_rate": 4.494874411426522e-05,
-      "loss": 0.2943,
+      "epoch": 1.1770641871193281,
+      "grad_norm": 0.1615755259990692,
+      "learning_rate": 4.467352199638982e-05,
+      "loss": 0.4543,
       "step": 32660
     },
     {
-      "epoch": 1.15,
-      "learning_rate": 4.4947026992164677e-05,
-      "loss": 0.2958,
+      "epoch": 1.1772443867805529,
+      "grad_norm": 0.19088678061962128,
+      "learning_rate": 4.46717212910199e-05,
+      "loss": 0.4356,
       "step": 32665
     },
     {
-      "epoch": 1.15,
-      "learning_rate": 4.4945309611065136e-05,
-      "loss": 0.3157,
+      "epoch": 1.1774245864417776,
+      "grad_norm": 0.1913667619228363,
+      "learning_rate": 4.466992031762658e-05,
+      "loss": 0.4021,
       "step": 32670
     },
     {
-      "epoch": 1.15,
-      "learning_rate": 4.49435919709889e-05,
-      "loss": 0.3251,
+      "epoch": 1.1776047861030021,
+      "grad_norm": 0.1861899197101593,
+      "learning_rate": 4.466811907623441e-05,
+      "loss": 0.4698,
       "step": 32675
     },
     {
-      "epoch": 1.15,
-      "learning_rate": 4.4941874071958254e-05,
-      "loss": 0.2771,
+      "epoch": 1.1777849857642269,
+      "grad_norm": 0.21820510923862457,
+      "learning_rate": 4.466631756686792e-05,
+      "loss": 0.4119,
       "step": 32680
     },
     {
-      "epoch": 1.15,
-      "learning_rate": 4.494015591399553e-05,
-      "loss": 0.2911,
+      "epoch": 1.1779651854254514,
+      "grad_norm": 0.20619237422943115,
+      "learning_rate": 4.4664515789551665e-05,
+      "loss": 0.4625,
       "step": 32685
     },
     {
-      "epoch": 1.15,
-      "learning_rate": 4.493843749712301e-05,
-      "loss": 0.3239,
+      "epoch": 1.178145385086676,
+      "grad_norm": 0.19361612200737,
+      "learning_rate": 4.466271374431019e-05,
+      "loss": 0.4151,
       "step": 32690
     },
     {
-      "epoch": 1.15,
-      "learning_rate": 4.493671882136302e-05,
-      "loss": 0.3032,
+      "epoch": 1.1783255847479006,
+      "grad_norm": 0.2008073925971985,
+      "learning_rate": 4.466091143116804e-05,
+      "loss": 0.4312,
       "step": 32695
     },
     {
-      "epoch": 1.15,
-      "learning_rate": 4.493499988673788e-05,
-      "loss": 0.3173,
+      "epoch": 1.1785057844091253,
+      "grad_norm": 0.15440912544727325,
+      "learning_rate": 4.465910885014979e-05,
+      "loss": 0.3915,
       "step": 32700
     },
     {
-      "epoch": 1.15,
-      "learning_rate": 4.493328069326991e-05,
-      "loss": 0.286,
+      "epoch": 1.1786859840703499,
+      "grad_norm": 0.37958988547325134,
+      "learning_rate": 4.4657306001279975e-05,
+      "loss": 0.4194,
       "step": 32705
     },
     {
-      "epoch": 1.15,
-      "learning_rate": 4.493156124098141e-05,
-      "loss": 0.2682,
+      "epoch": 1.1788661837315746,
+      "grad_norm": 0.19730152189731598,
+      "learning_rate": 4.4655502884583177e-05,
+      "loss": 0.4029,
       "step": 32710
     },
     {
-      "epoch": 1.15,
-      "learning_rate": 4.4929841529894735e-05,
-      "loss": 0.2971,
+      "epoch": 1.1790463833927993,
+      "grad_norm": 0.19884982705116272,
+      "learning_rate": 4.465369950008396e-05,
+      "loss": 0.4183,
       "step": 32715
     },
     {
-      "epoch": 1.15,
-      "learning_rate": 4.4928121560032196e-05,
-      "loss": 0.2826,
+      "epoch": 1.1792265830540238,
+      "grad_norm": 0.22939956188201904,
+      "learning_rate": 4.465189584780689e-05,
+      "loss": 0.4118,
       "step": 32720
     },
     {
-      "epoch": 1.15,
-      "learning_rate": 4.4926401331416135e-05,
-      "loss": 0.2779,
+      "epoch": 1.1794067827152486,
+      "grad_norm": 0.19102820754051208,
+      "learning_rate": 4.465009192777655e-05,
+      "loss": 0.4277,
       "step": 32725
     },
     {
-      "epoch": 1.15,
-      "learning_rate": 4.4924680844068887e-05,
-      "loss": 0.2962,
+      "epoch": 1.179586982376473,
+      "grad_norm": 0.19602443277835846,
+      "learning_rate": 4.4648287740017506e-05,
+      "loss": 0.4216,
       "step": 32730
     },
     {
-      "epoch": 1.15,
-      "learning_rate": 4.492296009801279e-05,
-      "loss": 0.277,
+      "epoch": 1.1797671820376978,
+      "grad_norm": 0.1469677835702896,
+      "learning_rate": 4.464648328455434e-05,
+      "loss": 0.4232,
       "step": 32735
     },
     {
-      "epoch": 1.15,
-      "learning_rate": 4.492123909327018e-05,
-      "loss": 0.3074,
+      "epoch": 1.1799473816989223,
+      "grad_norm": 0.15973035991191864,
+      "learning_rate": 4.464467856141166e-05,
+      "loss": 0.4254,
       "step": 32740
     },
     {
-      "epoch": 1.15,
-      "learning_rate": 4.491951782986341e-05,
-      "loss": 0.3198,
+      "epoch": 1.180127581360147,
+      "grad_norm": 0.18778270483016968,
+      "learning_rate": 4.4642873570614016e-05,
+      "loss": 0.4131,
       "step": 32745
     },
     {
-      "epoch": 1.15,
-      "learning_rate": 4.491779630781483e-05,
-      "loss": 0.3118,
+      "epoch": 1.1803077810213716,
+      "grad_norm": 0.16484954953193665,
+      "learning_rate": 4.464106831218604e-05,
+      "loss": 0.4222,
       "step": 32750
     },
     {
-      "epoch": 1.15,
-      "learning_rate": 4.491607452714679e-05,
-      "loss": 0.298,
+      "epoch": 1.1804879806825963,
+      "grad_norm": 0.17252211272716522,
+      "learning_rate": 4.4639262786152306e-05,
+      "loss": 0.4331,
       "step": 32755
     },
     {
-      "epoch": 1.15,
-      "learning_rate": 4.491435248788165e-05,
-      "loss": 0.293,
+      "epoch": 1.180668180343821,
+      "grad_norm": 0.16965141892433167,
+      "learning_rate": 4.463745699253742e-05,
+      "loss": 0.4342,
       "step": 32760
     },
     {
-      "epoch": 1.15,
-      "learning_rate": 4.4912630190041765e-05,
-      "loss": 0.2902,
+      "epoch": 1.1808483800050456,
+      "grad_norm": 0.1783820539712906,
+      "learning_rate": 4.463565093136598e-05,
+      "loss": 0.4189,
       "step": 32765
     },
     {
-      "epoch": 1.15,
-      "learning_rate": 4.49109076336495e-05,
-      "loss": 0.2658,
+      "epoch": 1.1810285796662703,
+      "grad_norm": 0.2075047343969345,
+      "learning_rate": 4.46338446026626e-05,
+      "loss": 0.4277,
       "step": 32770
     },
     {
-      "epoch": 1.15,
-      "learning_rate": 4.4909184818727224e-05,
-      "loss": 0.2954,
+      "epoch": 1.1812087793274948,
+      "grad_norm": 0.21429748833179474,
+      "learning_rate": 4.463203800645188e-05,
+      "loss": 0.4372,
       "step": 32775
     },
     {
-      "epoch": 1.15,
-      "learning_rate": 4.4907461745297294e-05,
-      "loss": 0.2952,
+      "epoch": 1.1813889789887195,
+      "grad_norm": 0.1777476817369461,
+      "learning_rate": 4.4630231142758455e-05,
+      "loss": 0.4113,
       "step": 32780
     },
     {
-      "epoch": 1.15,
-      "learning_rate": 4.49057384133821e-05,
-      "loss": 0.2875,
+      "epoch": 1.1815691786499443,
+      "grad_norm": 0.1785525679588318,
+      "learning_rate": 4.4628424011606927e-05,
+      "loss": 0.4065,
       "step": 32785
     },
     {
-      "epoch": 1.15,
-      "learning_rate": 4.490401482300401e-05,
-      "loss": 0.3302,
+      "epoch": 1.1817493783111688,
+      "grad_norm": 0.18620218336582184,
+      "learning_rate": 4.462661661302192e-05,
+      "loss": 0.4236,
       "step": 32790
     },
     {
-      "epoch": 1.15,
-      "learning_rate": 4.490229097418541e-05,
-      "loss": 0.32,
+      "epoch": 1.1819295779723935,
+      "grad_norm": 0.16326984763145447,
+      "learning_rate": 4.462480894702806e-05,
+      "loss": 0.3865,
       "step": 32795
     },
     {
-      "epoch": 1.15,
-      "learning_rate": 4.490056686694867e-05,
-      "loss": 0.3094,
+      "epoch": 1.182109777633618,
+      "grad_norm": 0.16543501615524292,
+      "learning_rate": 4.462300101364999e-05,
+      "loss": 0.4208,
       "step": 32800
     },
     {
-      "epoch": 1.15,
-      "learning_rate": 4.489884250131619e-05,
-      "loss": 0.3041,
+      "epoch": 1.1822899772948428,
+      "grad_norm": 0.1949724704027176,
+      "learning_rate": 4.4621192812912316e-05,
+      "loss": 0.4216,
       "step": 32805
     },
     {
-      "epoch": 1.15,
-      "learning_rate": 4.489711787731036e-05,
-      "loss": 0.3089,
+      "epoch": 1.1824701769560673,
+      "grad_norm": 0.19553053379058838,
+      "learning_rate": 4.46193843448397e-05,
+      "loss": 0.4616,
       "step": 32810
     },
     {
-      "epoch": 1.15,
-      "learning_rate": 4.489539299495356e-05,
-      "loss": 0.286,
+      "epoch": 1.182650376617292,
+      "grad_norm": 0.17806103825569153,
+      "learning_rate": 4.461757560945676e-05,
+      "loss": 0.4171,
       "step": 32815
     },
     {
-      "epoch": 1.15,
-      "learning_rate": 4.4893667854268184e-05,
-      "loss": 0.3166,
+      "epoch": 1.1828305762785165,
+      "grad_norm": 0.21131014823913574,
+      "learning_rate": 4.461576660678815e-05,
+      "loss": 0.4127,
       "step": 32820
     },
     {
-      "epoch": 1.15,
-      "learning_rate": 4.489194245527665e-05,
-      "loss": 0.308,
+      "epoch": 1.1830107759397412,
+      "grad_norm": 0.16493719816207886,
+      "learning_rate": 4.461395733685853e-05,
+      "loss": 0.4213,
       "step": 32825
     },
     {
-      "epoch": 1.16,
-      "learning_rate": 4.489021679800135e-05,
-      "loss": 0.2992,
+      "epoch": 1.183190975600966,
+      "grad_norm": 0.21300533413887024,
+      "learning_rate": 4.461214779969253e-05,
+      "loss": 0.4309,
       "step": 32830
     },
     {
-      "epoch": 1.16,
-      "learning_rate": 4.4888490882464696e-05,
-      "loss": 0.2923,
+      "epoch": 1.1833711752621905,
+      "grad_norm": 0.17048995196819305,
+      "learning_rate": 4.4610337995314816e-05,
+      "loss": 0.4518,
       "step": 32835
     },
     {
-      "epoch": 1.16,
-      "learning_rate": 4.48867647086891e-05,
-      "loss": 0.312,
+      "epoch": 1.1835513749234152,
+      "grad_norm": 0.17380398511886597,
+      "learning_rate": 4.460852792375004e-05,
+      "loss": 0.4386,
       "step": 32840
     },
     {
-      "epoch": 1.16,
-      "learning_rate": 4.488503827669696e-05,
-      "loss": 0.2999,
+      "epoch": 1.1837315745846397,
+      "grad_norm": 0.16541922092437744,
+      "learning_rate": 4.4606717585022876e-05,
+      "loss": 0.4223,
       "step": 32845
     },
     {
-      "epoch": 1.16,
-      "learning_rate": 4.4883311586510706e-05,
-      "loss": 0.3034,
+      "epoch": 1.1839117742458645,
+      "grad_norm": 0.17973947525024414,
+      "learning_rate": 4.460490697915797e-05,
+      "loss": 0.4302,
       "step": 32850
     },
     {
-      "epoch": 1.16,
-      "learning_rate": 4.4881584638152756e-05,
-      "loss": 0.2815,
+      "epoch": 1.184091973907089,
+      "grad_norm": 0.24981901049613953,
+      "learning_rate": 4.460309610618001e-05,
+      "loss": 0.4425,
       "step": 32855
     },
     {
-      "epoch": 1.16,
-      "learning_rate": 4.4879857431645536e-05,
-      "loss": 0.289,
+      "epoch": 1.1842721735683137,
+      "grad_norm": 0.23169773817062378,
+      "learning_rate": 4.4601284966113656e-05,
+      "loss": 0.4564,
       "step": 32860
     },
     {
-      "epoch": 1.16,
-      "learning_rate": 4.4878129967011464e-05,
-      "loss": 0.3124,
+      "epoch": 1.1844523732295382,
+      "grad_norm": 0.15463878214359283,
+      "learning_rate": 4.4599473558983596e-05,
+      "loss": 0.3629,
       "step": 32865
     },
     {
-      "epoch": 1.16,
-      "learning_rate": 4.4876402244272974e-05,
-      "loss": 0.2979,
+      "epoch": 1.184632572890763,
+      "grad_norm": 0.17254050076007843,
+      "learning_rate": 4.459766188481449e-05,
+      "loss": 0.414,
       "step": 32870
     },
     {
-      "epoch": 1.16,
-      "learning_rate": 4.4874674263452496e-05,
-      "loss": 0.2549,
+      "epoch": 1.1848127725519877,
+      "grad_norm": 0.1448955237865448,
+      "learning_rate": 4.459584994363105e-05,
+      "loss": 0.4048,
       "step": 32875
     },
     {
-      "epoch": 1.16,
-      "learning_rate": 4.4872946024572474e-05,
-      "loss": 0.3019,
+      "epoch": 1.1849929722132122,
+      "grad_norm": 0.15114130079746246,
+      "learning_rate": 4.459403773545794e-05,
+      "loss": 0.4154,
       "step": 32880
     },
     {
-      "epoch": 1.16,
-      "learning_rate": 4.4871217527655344e-05,
-      "loss": 0.2989,
+      "epoch": 1.185173171874437,
+      "grad_norm": 0.16808383166790009,
+      "learning_rate": 4.459222526031987e-05,
+      "loss": 0.4226,
       "step": 32885
     },
     {
-      "epoch": 1.16,
-      "learning_rate": 4.4869488772723555e-05,
-      "loss": 0.2659,
+      "epoch": 1.1853533715356614,
+      "grad_norm": 0.18770961463451385,
+      "learning_rate": 4.45904125182415e-05,
+      "loss": 0.4326,
       "step": 32890
     },
     {
-      "epoch": 1.16,
-      "learning_rate": 4.4867759759799544e-05,
-      "loss": 0.2789,
+      "epoch": 1.1855335711968862,
+      "grad_norm": 0.18759319186210632,
+      "learning_rate": 4.458859950924758e-05,
+      "loss": 0.4325,
       "step": 32895
     },
     {
-      "epoch": 1.16,
-      "learning_rate": 4.4866030488905766e-05,
-      "loss": 0.2879,
+      "epoch": 1.1857137708581107,
+      "grad_norm": 0.19363155961036682,
+      "learning_rate": 4.458678623336277e-05,
+      "loss": 0.4474,
       "step": 32900
     },
     {
-      "epoch": 1.16,
-      "learning_rate": 4.486430096006467e-05,
-      "loss": 0.3129,
+      "epoch": 1.1858939705193354,
+      "grad_norm": 0.13609229028224945,
+      "learning_rate": 4.4584972690611784e-05,
+      "loss": 0.3956,
       "step": 32905
     },
     {
-      "epoch": 1.16,
-      "learning_rate": 4.4862571173298727e-05,
-      "loss": 0.2921,
+      "epoch": 1.18607417018056,
+      "grad_norm": 0.20877060294151306,
+      "learning_rate": 4.458315888101935e-05,
+      "loss": 0.4139,
       "step": 32910
     },
     {
-      "epoch": 1.16,
-      "learning_rate": 4.4860841128630384e-05,
-      "loss": 0.3155,
+      "epoch": 1.1862543698417847,
+      "grad_norm": 0.1942521333694458,
+      "learning_rate": 4.4581344804610156e-05,
+      "loss": 0.4022,
       "step": 32915
     },
     {
-      "epoch": 1.16,
-      "learning_rate": 4.485911082608211e-05,
-      "loss": 0.2996,
+      "epoch": 1.1864345695030094,
+      "grad_norm": 0.19765913486480713,
+      "learning_rate": 4.457953046140894e-05,
+      "loss": 0.4402,
       "step": 32920
     },
     {
-      "epoch": 1.16,
-      "learning_rate": 4.4857380265676363e-05,
-      "loss": 0.282,
+      "epoch": 1.186614769164234,
+      "grad_norm": 0.15261493623256683,
+      "learning_rate": 4.4577715851440405e-05,
+      "loss": 0.4131,
       "step": 32925
     },
     {
-      "epoch": 1.16,
-      "learning_rate": 4.485564944743562e-05,
-      "loss": 0.2971,
+      "epoch": 1.1867949688254587,
+      "grad_norm": 0.17721693217754364,
+      "learning_rate": 4.4575900974729284e-05,
+      "loss": 0.434,
       "step": 32930
     },
     {
-      "epoch": 1.16,
-      "learning_rate": 4.4853918371382365e-05,
-      "loss": 0.2768,
+      "epoch": 1.1869751684866832,
+      "grad_norm": 0.1746007800102234,
+      "learning_rate": 4.45740858313003e-05,
+      "loss": 0.4198,
       "step": 32935
     },
     {
-      "epoch": 1.16,
-      "learning_rate": 4.485218703753905e-05,
-      "loss": 0.3019,
+      "epoch": 1.187155368147908,
+      "grad_norm": 0.18980930745601654,
+      "learning_rate": 4.45722704211782e-05,
+      "loss": 0.4456,
       "step": 32940
     },
     {
-      "epoch": 1.16,
-      "learning_rate": 4.485045544592818e-05,
-      "loss": 0.2769,
+      "epoch": 1.1873355678091326,
+      "grad_norm": 0.15787626802921295,
+      "learning_rate": 4.457045474438769e-05,
+      "loss": 0.3863,
       "step": 32945
     },
     {
-      "epoch": 1.16,
-      "learning_rate": 4.484872359657223e-05,
-      "loss": 0.2651,
+      "epoch": 1.1875157674703571,
+      "grad_norm": 0.2203458696603775,
+      "learning_rate": 4.4568638800953524e-05,
+      "loss": 0.3924,
       "step": 32950
     },
     {
-      "epoch": 1.16,
-      "learning_rate": 4.4846991489493674e-05,
-      "loss": 0.3109,
+      "epoch": 1.1876959671315819,
+      "grad_norm": 0.18546192348003387,
+      "learning_rate": 4.4566822590900445e-05,
+      "loss": 0.4048,
       "step": 32955
     },
     {
-      "epoch": 1.16,
-      "learning_rate": 4.4845259124715014e-05,
-      "loss": 0.2813,
+      "epoch": 1.1878761667928064,
+      "grad_norm": 0.19496969878673553,
+      "learning_rate": 4.45650061142532e-05,
+      "loss": 0.3955,
       "step": 32960
     },
     {
-      "epoch": 1.16,
-      "learning_rate": 4.4843526502258747e-05,
-      "loss": 0.2924,
+      "epoch": 1.1880563664540311,
+      "grad_norm": 0.1746780127286911,
+      "learning_rate": 4.456318937103653e-05,
+      "loss": 0.4654,
       "step": 32965
     },
     {
-      "epoch": 1.16,
-      "learning_rate": 4.484179362214736e-05,
-      "loss": 0.2847,
+      "epoch": 1.1882365661152556,
+      "grad_norm": 0.17759990692138672,
+      "learning_rate": 4.45613723612752e-05,
+      "loss": 0.4105,
       "step": 32970
     },
     {
-      "epoch": 1.16,
-      "learning_rate": 4.484006048440337e-05,
-      "loss": 0.2839,
+      "epoch": 1.1884167657764804,
+      "grad_norm": 0.21956616640090942,
+      "learning_rate": 4.455955508499395e-05,
+      "loss": 0.4118,
       "step": 32975
     },
     {
-      "epoch": 1.16,
-      "learning_rate": 4.483832708904927e-05,
-      "loss": 0.2725,
+      "epoch": 1.1885969654377049,
+      "grad_norm": 0.1860971599817276,
+      "learning_rate": 4.455773754221755e-05,
+      "loss": 0.4431,
       "step": 32980
     },
     {
-      "epoch": 1.16,
-      "learning_rate": 4.4836593436107555e-05,
-      "loss": 0.3009,
+      "epoch": 1.1887771650989296,
+      "grad_norm": 0.17252309620380402,
+      "learning_rate": 4.455591973297077e-05,
+      "loss": 0.4097,
       "step": 32985
     },
     {
-      "epoch": 1.16,
-      "learning_rate": 4.483485952560076e-05,
-      "loss": 0.2911,
+      "epoch": 1.1889573647601543,
+      "grad_norm": 0.16842429339885712,
+      "learning_rate": 4.455410165727836e-05,
+      "loss": 0.4194,
       "step": 32990
     },
     {
-      "epoch": 1.16,
-      "learning_rate": 4.4833125357551375e-05,
-      "loss": 0.2847,
+      "epoch": 1.1891375644213789,
+      "grad_norm": 0.16480675339698792,
+      "learning_rate": 4.4552283315165114e-05,
+      "loss": 0.4398,
       "step": 32995
     },
     {
-      "epoch": 1.16,
-      "learning_rate": 4.4831390931981934e-05,
-      "loss": 0.3052,
+      "epoch": 1.1893177640826036,
+      "grad_norm": 0.1965269148349762,
+      "learning_rate": 4.455046470665578e-05,
+      "loss": 0.435,
       "step": 33000
     },
     {
-      "epoch": 1.16,
-      "eval_loss": 0.28887951374053955,
-      "eval_runtime": 10.5389,
-      "eval_samples_per_second": 9.489,
-      "eval_steps_per_second": 9.489,
+      "epoch": 1.1893177640826036,
+      "eval_loss": 0.4501653015613556,
+      "eval_runtime": 3.5279,
+      "eval_samples_per_second": 28.345,
+      "eval_steps_per_second": 7.086,
       "step": 33000
     },
     {
-      "epoch": 1.16,
-      "learning_rate": 4.4829656248914945e-05,
-      "loss": 0.3001,
+      "epoch": 1.189497963743828,
+      "grad_norm": 0.16842001676559448,
+      "learning_rate": 4.454864583177515e-05,
+      "loss": 0.4186,
       "step": 33005
     },
     {
-      "epoch": 1.16,
-      "learning_rate": 4.4827921308372945e-05,
-      "loss": 0.2907,
+      "epoch": 1.1896781634050528,
+      "grad_norm": 0.17782749235630035,
+      "learning_rate": 4.454682669054801e-05,
+      "loss": 0.4134,
       "step": 33010
     },
     {
-      "epoch": 1.16,
-      "learning_rate": 4.482618611037845e-05,
-      "loss": 0.2813,
+      "epoch": 1.1898583630662773,
+      "grad_norm": 0.20939677953720093,
+      "learning_rate": 4.454500728299914e-05,
+      "loss": 0.4094,
       "step": 33015
     },
     {
-      "epoch": 1.16,
-      "learning_rate": 4.482445065495399e-05,
-      "loss": 0.2769,
+      "epoch": 1.190038562727502,
+      "grad_norm": 0.22223712503910065,
+      "learning_rate": 4.454318760915333e-05,
+      "loss": 0.4279,
       "step": 33020
     },
     {
-      "epoch": 1.16,
-      "learning_rate": 4.4822714942122105e-05,
-      "loss": 0.3057,
+      "epoch": 1.1902187623887266,
+      "grad_norm": 0.19512085616588593,
+      "learning_rate": 4.4541367669035373e-05,
+      "loss": 0.4125,
       "step": 33025
     },
     {
-      "epoch": 1.16,
-      "learning_rate": 4.482097897190534e-05,
-      "loss": 0.2956,
+      "epoch": 1.1903989620499513,
+      "grad_norm": 0.17766328155994415,
+      "learning_rate": 4.453954746267006e-05,
+      "loss": 0.4323,
       "step": 33030
     },
     {
-      "epoch": 1.16,
-      "learning_rate": 4.481924274432622e-05,
-      "loss": 0.282,
+      "epoch": 1.190579161711176,
+      "grad_norm": 0.17056851089000702,
+      "learning_rate": 4.45377269900822e-05,
+      "loss": 0.4184,
       "step": 33035
     },
     {
-      "epoch": 1.16,
-      "learning_rate": 4.481750625940729e-05,
-      "loss": 0.2964,
+      "epoch": 1.1907593613724006,
+      "grad_norm": 0.18727493286132812,
+      "learning_rate": 4.45359062512966e-05,
+      "loss": 0.42,
       "step": 33040
     },
     {
-      "epoch": 1.16,
-      "learning_rate": 4.4815769517171105e-05,
-      "loss": 0.3054,
+      "epoch": 1.1909395610336253,
+      "grad_norm": 0.16119909286499023,
+      "learning_rate": 4.453408524633805e-05,
+      "loss": 0.3888,
       "step": 33045
     },
     {
-      "epoch": 1.16,
-      "learning_rate": 4.481403251764021e-05,
-      "loss": 0.2802,
+      "epoch": 1.1911197606948498,
+      "grad_norm": 0.16698874533176422,
+      "learning_rate": 4.453226397523137e-05,
+      "loss": 0.4427,
       "step": 33050
     },
     {
-      "epoch": 1.16,
-      "learning_rate": 4.481229526083715e-05,
-      "loss": 0.302,
+      "epoch": 1.1912999603560745,
+      "grad_norm": 0.1766866147518158,
+      "learning_rate": 4.453044243800137e-05,
+      "loss": 0.4434,
       "step": 33055
     },
     {
-      "epoch": 1.16,
-      "learning_rate": 4.4810557746784514e-05,
-      "loss": 0.2758,
+      "epoch": 1.1914801600172993,
+      "grad_norm": 0.19949309527873993,
+      "learning_rate": 4.452862063467289e-05,
+      "loss": 0.4333,
       "step": 33060
     },
     {
-      "epoch": 1.16,
-      "learning_rate": 4.4808819975504835e-05,
-      "loss": 0.2762,
+      "epoch": 1.1916603596785238,
+      "grad_norm": 0.18705400824546814,
+      "learning_rate": 4.452679856527072e-05,
+      "loss": 0.4109,
       "step": 33065
     },
     {
-      "epoch": 1.16,
-      "learning_rate": 4.480708194702068e-05,
-      "loss": 0.3195,
+      "epoch": 1.1918405593397485,
+      "grad_norm": 0.19949448108673096,
+      "learning_rate": 4.45249762298197e-05,
+      "loss": 0.3773,
       "step": 33070
     },
     {
-      "epoch": 1.16,
-      "learning_rate": 4.480534366135462e-05,
-      "loss": 0.2812,
+      "epoch": 1.192020759000973,
+      "grad_norm": 0.1769210696220398,
+      "learning_rate": 4.452315362834467e-05,
+      "loss": 0.4424,
       "step": 33075
     },
     {
-      "epoch": 1.16,
-      "learning_rate": 4.480360511852922e-05,
-      "loss": 0.2912,
+      "epoch": 1.1922009586621978,
+      "grad_norm": 0.19094079732894897,
+      "learning_rate": 4.452133076087045e-05,
+      "loss": 0.3962,
       "step": 33080
     },
     {
-      "epoch": 1.16,
-      "learning_rate": 4.480186631856707e-05,
-      "loss": 0.2869,
+      "epoch": 1.1923811583234223,
+      "grad_norm": 0.2079572230577469,
+      "learning_rate": 4.4519507627421873e-05,
+      "loss": 0.4552,
       "step": 33085
     },
     {
-      "epoch": 1.16,
-      "learning_rate": 4.480012726149072e-05,
-      "loss": 0.2889,
+      "epoch": 1.192561357984647,
+      "grad_norm": 0.24924804270267487,
+      "learning_rate": 4.451768422802378e-05,
+      "loss": 0.3995,
       "step": 33090
     },
     {
-      "epoch": 1.16,
-      "learning_rate": 4.479838794732277e-05,
-      "loss": 0.302,
+      "epoch": 1.1927415576458715,
+      "grad_norm": 0.1824015974998474,
+      "learning_rate": 4.451586056270103e-05,
+      "loss": 0.4448,
       "step": 33095
     },
     {
-      "epoch": 1.16,
-      "learning_rate": 4.4796648376085815e-05,
-      "loss": 0.2824,
+      "epoch": 1.1929217573070963,
+      "grad_norm": 0.1912459284067154,
+      "learning_rate": 4.4514036631478444e-05,
+      "loss": 0.4076,
       "step": 33100
     },
     {
-      "epoch": 1.16,
-      "learning_rate": 4.479490854780242e-05,
-      "loss": 0.2879,
+      "epoch": 1.193101956968321,
+      "grad_norm": 0.1705092191696167,
+      "learning_rate": 4.4512212434380894e-05,
+      "loss": 0.4206,
       "step": 33105
     },
     {
-      "epoch": 1.16,
-      "learning_rate": 4.479316846249517e-05,
-      "loss": 0.31,
+      "epoch": 1.1932821566295455,
+      "grad_norm": 0.17675426602363586,
+      "learning_rate": 4.4510387971433225e-05,
+      "loss": 0.4243,
       "step": 33110
     },
     {
-      "epoch": 1.17,
-      "learning_rate": 4.479142812018669e-05,
-      "loss": 0.2899,
+      "epoch": 1.1934623562907702,
+      "grad_norm": 0.21491315960884094,
+      "learning_rate": 4.45085632426603e-05,
+      "loss": 0.4422,
       "step": 33115
     },
     {
-      "epoch": 1.17,
-      "learning_rate": 4.478968752089955e-05,
-      "loss": 0.3027,
+      "epoch": 1.1936425559519948,
+      "grad_norm": 0.2086547613143921,
+      "learning_rate": 4.4506738248086974e-05,
+      "loss": 0.4128,
       "step": 33120
     },
     {
-      "epoch": 1.17,
-      "learning_rate": 4.478794666465636e-05,
-      "loss": 0.2736,
+      "epoch": 1.1938227556132195,
+      "grad_norm": 0.1773192584514618,
+      "learning_rate": 4.4504912987738124e-05,
+      "loss": 0.4244,
       "step": 33125
     },
     {
-      "epoch": 1.17,
-      "learning_rate": 4.478620555147972e-05,
-      "loss": 0.2686,
+      "epoch": 1.194002955274444,
+      "grad_norm": 0.19399385154247284,
+      "learning_rate": 4.45030874616386e-05,
+      "loss": 0.4028,
       "step": 33130
     },
     {
-      "epoch": 1.17,
-      "learning_rate": 4.478446418139224e-05,
-      "loss": 0.3045,
+      "epoch": 1.1941831549356687,
+      "grad_norm": 0.22621554136276245,
+      "learning_rate": 4.45012616698133e-05,
+      "loss": 0.4292,
       "step": 33135
     },
     {
-      "epoch": 1.17,
-      "learning_rate": 4.478272255441653e-05,
-      "loss": 0.273,
+      "epoch": 1.1943633545968932,
+      "grad_norm": 0.17529787123203278,
+      "learning_rate": 4.449943561228707e-05,
+      "loss": 0.3879,
       "step": 33140
     },
     {
-      "epoch": 1.17,
-      "learning_rate": 4.478098067057522e-05,
-      "loss": 0.2836,
+      "epoch": 1.194543554258118,
+      "grad_norm": 0.18595170974731445,
+      "learning_rate": 4.449760928908481e-05,
+      "loss": 0.4049,
       "step": 33145
     },
     {
-      "epoch": 1.17,
-      "learning_rate": 4.47792385298909e-05,
-      "loss": 0.2809,
+      "epoch": 1.1947237539193427,
+      "grad_norm": 0.17124305665493011,
+      "learning_rate": 4.44957827002314e-05,
+      "loss": 0.3847,
       "step": 33150
     },
     {
-      "epoch": 1.17,
-      "learning_rate": 4.477749613238621e-05,
-      "loss": 0.3096,
+      "epoch": 1.1949039535805672,
+      "grad_norm": 0.1618146151304245,
+      "learning_rate": 4.449395584575172e-05,
+      "loss": 0.4069,
       "step": 33155
     },
     {
-      "epoch": 1.17,
-      "learning_rate": 4.4775753478083756e-05,
-      "loss": 0.3059,
+      "epoch": 1.195084153241792,
+      "grad_norm": 0.15872132778167725,
+      "learning_rate": 4.449212872567068e-05,
+      "loss": 0.4154,
       "step": 33160
     },
     {
-      "epoch": 1.17,
-      "learning_rate": 4.477401056700618e-05,
-      "loss": 0.2698,
+      "epoch": 1.1952643529030165,
+      "grad_norm": 0.18358559906482697,
+      "learning_rate": 4.4490301340013146e-05,
+      "loss": 0.4185,
       "step": 33165
     },
     {
-      "epoch": 1.17,
-      "learning_rate": 4.477226739917611e-05,
-      "loss": 0.2845,
+      "epoch": 1.1954445525642412,
+      "grad_norm": 0.139307901263237,
+      "learning_rate": 4.4488473688804034e-05,
+      "loss": 0.3846,
       "step": 33170
     },
     {
-      "epoch": 1.17,
-      "learning_rate": 4.477052397461617e-05,
-      "loss": 0.3061,
+      "epoch": 1.195624752225466,
+      "grad_norm": 0.16078798472881317,
+      "learning_rate": 4.448664577206824e-05,
+      "loss": 0.4072,
       "step": 33175
     },
     {
-      "epoch": 1.17,
-      "learning_rate": 4.476878029334902e-05,
-      "loss": 0.2727,
+      "epoch": 1.1958049518866904,
+      "grad_norm": 0.20551976561546326,
+      "learning_rate": 4.448481758983067e-05,
+      "loss": 0.4151,
       "step": 33180
     },
     {
-      "epoch": 1.17,
-      "learning_rate": 4.4767036355397276e-05,
-      "loss": 0.2941,
+      "epoch": 1.1959851515479152,
+      "grad_norm": 0.17032408714294434,
+      "learning_rate": 4.4482989142116236e-05,
+      "loss": 0.4053,
       "step": 33185
     },
     {
-      "epoch": 1.17,
-      "learning_rate": 4.47652921607836e-05,
-      "loss": 0.284,
+      "epoch": 1.1961653512091397,
+      "grad_norm": 0.19917094707489014,
+      "learning_rate": 4.4481160428949845e-05,
+      "loss": 0.3863,
       "step": 33190
     },
     {
-      "epoch": 1.17,
-      "learning_rate": 4.476354770953062e-05,
-      "loss": 0.2993,
+      "epoch": 1.1963455508703644,
+      "grad_norm": 0.17138609290122986,
+      "learning_rate": 4.447933145035642e-05,
+      "loss": 0.4144,
       "step": 33195
     },
     {
-      "epoch": 1.17,
-      "learning_rate": 4.476180300166101e-05,
-      "loss": 0.2959,
+      "epoch": 1.196525750531589,
+      "grad_norm": 0.16464346647262573,
+      "learning_rate": 4.447750220636086e-05,
+      "loss": 0.4573,
       "step": 33200
     },
     {
-      "epoch": 1.17,
-      "learning_rate": 4.476005803719741e-05,
-      "loss": 0.2831,
+      "epoch": 1.1967059501928137,
+      "grad_norm": 0.18972603976726532,
+      "learning_rate": 4.4475672696988117e-05,
+      "loss": 0.4093,
       "step": 33205
     },
     {
-      "epoch": 1.17,
-      "learning_rate": 4.4758312816162466e-05,
-      "loss": 0.3011,
+      "epoch": 1.1968861498540382,
+      "grad_norm": 0.1480892151594162,
+      "learning_rate": 4.44738429222631e-05,
+      "loss": 0.3913,
       "step": 33210
     },
     {
-      "epoch": 1.17,
-      "learning_rate": 4.4756567338578856e-05,
-      "loss": 0.3106,
+      "epoch": 1.197066349515263,
+      "grad_norm": 0.1947953850030899,
+      "learning_rate": 4.4472012882210744e-05,
+      "loss": 0.4281,
       "step": 33215
     },
     {
-      "epoch": 1.17,
-      "learning_rate": 4.4754821604469245e-05,
-      "loss": 0.2985,
+      "epoch": 1.1972465491764877,
+      "grad_norm": 0.1418490707874298,
+      "learning_rate": 4.4470182576855984e-05,
+      "loss": 0.4228,
       "step": 33220
     },
     {
-      "epoch": 1.17,
-      "learning_rate": 4.47530756138563e-05,
-      "loss": 0.3007,
+      "epoch": 1.1974267488377122,
+      "grad_norm": 0.21313992142677307,
+      "learning_rate": 4.446835200622376e-05,
+      "loss": 0.4354,
       "step": 33225
     },
     {
-      "epoch": 1.17,
-      "learning_rate": 4.4751329366762677e-05,
-      "loss": 0.3026,
+      "epoch": 1.197606948498937,
+      "grad_norm": 0.16371306777000427,
+      "learning_rate": 4.4466521170339e-05,
+      "loss": 0.4339,
       "step": 33230
     },
     {
-      "epoch": 1.17,
-      "learning_rate": 4.474958286321105e-05,
-      "loss": 0.2842,
+      "epoch": 1.1977871481601614,
+      "grad_norm": 0.17602042853832245,
+      "learning_rate": 4.446469006922666e-05,
+      "loss": 0.4083,
       "step": 33235
     },
     {
-      "epoch": 1.17,
-      "learning_rate": 4.4747836103224115e-05,
-      "loss": 0.2882,
+      "epoch": 1.1979673478213861,
+      "grad_norm": 0.2149427831172943,
+      "learning_rate": 4.446285870291168e-05,
+      "loss": 0.4426,
       "step": 33240
     },
     {
-      "epoch": 1.17,
-      "learning_rate": 4.474608908682455e-05,
-      "loss": 0.2853,
+      "epoch": 1.1981475474826107,
+      "grad_norm": 0.19372370839118958,
+      "learning_rate": 4.4461027071419034e-05,
+      "loss": 0.4215,
       "step": 33245
     },
     {
-      "epoch": 1.17,
-      "learning_rate": 4.474434181403502e-05,
-      "loss": 0.3047,
+      "epoch": 1.1983277471438354,
+      "grad_norm": 0.1823178380727768,
+      "learning_rate": 4.445919517477365e-05,
+      "loss": 0.4066,
       "step": 33250
     },
     {
-      "epoch": 1.17,
-      "learning_rate": 4.4742594284878225e-05,
-      "loss": 0.2728,
+      "epoch": 1.19850794680506,
+      "grad_norm": 0.2029699683189392,
+      "learning_rate": 4.445736301300051e-05,
+      "loss": 0.4223,
       "step": 33255
     },
     {
-      "epoch": 1.17,
-      "learning_rate": 4.4740846499376856e-05,
-      "loss": 0.2977,
+      "epoch": 1.1986881464662846,
+      "grad_norm": 0.16662827134132385,
+      "learning_rate": 4.445553058612455e-05,
+      "loss": 0.4364,
       "step": 33260
     },
     {
-      "epoch": 1.17,
-      "learning_rate": 4.4739098457553606e-05,
-      "loss": 0.3427,
+      "epoch": 1.1988683461275094,
+      "grad_norm": 0.1548469513654709,
+      "learning_rate": 4.445369789417077e-05,
+      "loss": 0.4124,
       "step": 33265
     },
     {
-      "epoch": 1.17,
-      "learning_rate": 4.473735015943117e-05,
-      "loss": 0.2878,
+      "epoch": 1.1990485457887339,
+      "grad_norm": 0.15706267952919006,
+      "learning_rate": 4.445186493716411e-05,
+      "loss": 0.3692,
       "step": 33270
     },
     {
-      "epoch": 1.17,
-      "learning_rate": 4.473560160503225e-05,
-      "loss": 0.2875,
+      "epoch": 1.1992287454499586,
+      "grad_norm": 0.19710016250610352,
+      "learning_rate": 4.4450031715129556e-05,
+      "loss": 0.4461,
       "step": 33275
     },
     {
-      "epoch": 1.17,
-      "learning_rate": 4.473385279437955e-05,
-      "loss": 0.2912,
+      "epoch": 1.1994089451111831,
+      "grad_norm": 0.16916310787200928,
+      "learning_rate": 4.4448198228092095e-05,
+      "loss": 0.3842,
       "step": 33280
     },
     {
-      "epoch": 1.17,
-      "learning_rate": 4.4732103727495775e-05,
-      "loss": 0.2663,
+      "epoch": 1.1995891447724079,
+      "grad_norm": 0.24299342930316925,
+      "learning_rate": 4.444636447607669e-05,
+      "loss": 0.4048,
       "step": 33285
     },
     {
-      "epoch": 1.17,
-      "learning_rate": 4.473035440440364e-05,
-      "loss": 0.2863,
+      "epoch": 1.1997693444336326,
+      "grad_norm": 0.1836657077074051,
+      "learning_rate": 4.444453045910834e-05,
+      "loss": 0.4123,
       "step": 33290
     },
     {
-      "epoch": 1.17,
-      "learning_rate": 4.472860482512585e-05,
-      "loss": 0.2773,
+      "epoch": 1.199949544094857,
+      "grad_norm": 0.2232903391122818,
+      "learning_rate": 4.444269617721202e-05,
+      "loss": 0.4128,
       "step": 33295
     },
     {
-      "epoch": 1.17,
-      "learning_rate": 4.472685498968513e-05,
-      "loss": 0.2813,
+      "epoch": 1.2001297437560818,
+      "grad_norm": 0.150344580411911,
+      "learning_rate": 4.444086163041273e-05,
+      "loss": 0.4184,
       "step": 33300
     },
     {
-      "epoch": 1.17,
-      "learning_rate": 4.47251048981042e-05,
-      "loss": 0.302,
+      "epoch": 1.2003099434173063,
+      "grad_norm": 0.16514082252979279,
+      "learning_rate": 4.443902681873547e-05,
+      "loss": 0.4122,
       "step": 33305
     },
     {
-      "epoch": 1.17,
-      "learning_rate": 4.4723354550405784e-05,
-      "loss": 0.2719,
+      "epoch": 1.200490143078531,
+      "grad_norm": 0.16925464570522308,
+      "learning_rate": 4.443719174220523e-05,
+      "loss": 0.4041,
       "step": 33310
     },
     {
-      "epoch": 1.17,
-      "learning_rate": 4.4721603946612604e-05,
-      "loss": 0.2746,
+      "epoch": 1.2006703427397556,
+      "grad_norm": 0.2036067545413971,
+      "learning_rate": 4.443535640084702e-05,
+      "loss": 0.4335,
       "step": 33315
     },
     {
-      "epoch": 1.17,
-      "learning_rate": 4.47198530867474e-05,
-      "loss": 0.3027,
+      "epoch": 1.2008505424009803,
+      "grad_norm": 0.19253499805927277,
+      "learning_rate": 4.443352079468583e-05,
+      "loss": 0.4581,
       "step": 33320
     },
     {
-      "epoch": 1.17,
-      "learning_rate": 4.4718101970832895e-05,
-      "loss": 0.2748,
+      "epoch": 1.2010307420622048,
+      "grad_norm": 0.16766758263111115,
+      "learning_rate": 4.4431684923746695e-05,
+      "loss": 0.4244,
       "step": 33325
     },
     {
-      "epoch": 1.17,
-      "learning_rate": 4.4716350598891834e-05,
-      "loss": 0.2948,
+      "epoch": 1.2012109417234296,
+      "grad_norm": 0.14643500745296478,
+      "learning_rate": 4.442984878805461e-05,
+      "loss": 0.4007,
       "step": 33330
     },
     {
-      "epoch": 1.17,
-      "learning_rate": 4.4714598970946944e-05,
-      "loss": 0.3105,
+      "epoch": 1.2013911413846543,
+      "grad_norm": 0.19802068173885345,
+      "learning_rate": 4.44280123876346e-05,
+      "loss": 0.4579,
       "step": 33335
     },
     {
-      "epoch": 1.17,
-      "learning_rate": 4.471284708702098e-05,
-      "loss": 0.3052,
+      "epoch": 1.2015713410458788,
+      "grad_norm": 0.2025717943906784,
+      "learning_rate": 4.4426175722511674e-05,
+      "loss": 0.4056,
       "step": 33340
     },
     {
-      "epoch": 1.17,
-      "learning_rate": 4.47110949471367e-05,
-      "loss": 0.3049,
+      "epoch": 1.2017515407071035,
+      "grad_norm": 0.181138813495636,
+      "learning_rate": 4.442433879271087e-05,
+      "loss": 0.4282,
       "step": 33345
     },
     {
-      "epoch": 1.17,
-      "learning_rate": 4.4709342551316836e-05,
-      "loss": 0.2955,
+      "epoch": 1.201931740368328,
+      "grad_norm": 0.1720162183046341,
+      "learning_rate": 4.4422501598257216e-05,
+      "loss": 0.4089,
       "step": 33350
     },
     {
-      "epoch": 1.17,
-      "learning_rate": 4.470758989958415e-05,
-      "loss": 0.276,
+      "epoch": 1.2021119400295528,
+      "grad_norm": 0.23699156939983368,
+      "learning_rate": 4.442066413917573e-05,
+      "loss": 0.4118,
       "step": 33355
     },
     {
-      "epoch": 1.17,
-      "learning_rate": 4.470583699196139e-05,
-      "loss": 0.2762,
+      "epoch": 1.2022921396907773,
+      "grad_norm": 0.16807523369789124,
+      "learning_rate": 4.441882641549145e-05,
+      "loss": 0.4238,
       "step": 33360
     },
     {
-      "epoch": 1.17,
-      "learning_rate": 4.470408382847133e-05,
-      "loss": 0.3054,
+      "epoch": 1.202472339352002,
+      "grad_norm": 0.16875354945659637,
+      "learning_rate": 4.441698842722943e-05,
+      "loss": 0.41,
       "step": 33365
     },
     {
-      "epoch": 1.17,
-      "learning_rate": 4.4702330409136716e-05,
-      "loss": 0.2919,
+      "epoch": 1.2026525390132266,
+      "grad_norm": 0.15263767540454865,
+      "learning_rate": 4.44151501744147e-05,
+      "loss": 0.3937,
       "step": 33370
     },
     {
-      "epoch": 1.17,
-      "learning_rate": 4.470057673398034e-05,
-      "loss": 0.2929,
+      "epoch": 1.2028327386744513,
+      "grad_norm": 0.19248110055923462,
+      "learning_rate": 4.441331165707231e-05,
+      "loss": 0.4074,
       "step": 33375
     },
     {
-      "epoch": 1.17,
-      "learning_rate": 4.469882280302495e-05,
-      "loss": 0.3022,
+      "epoch": 1.203012938335676,
+      "grad_norm": 0.1765611320734024,
+      "learning_rate": 4.44114728752273e-05,
+      "loss": 0.4095,
       "step": 33380
     },
     {
-      "epoch": 1.17,
-      "learning_rate": 4.469706861629333e-05,
-      "loss": 0.3106,
+      "epoch": 1.2031931379969005,
+      "grad_norm": 0.1671658307313919,
+      "learning_rate": 4.440963382890474e-05,
+      "loss": 0.4064,
       "step": 33385
     },
     {
-      "epoch": 1.17,
-      "learning_rate": 4.469531417380826e-05,
-      "loss": 0.2885,
+      "epoch": 1.2033733376581253,
+      "grad_norm": 0.1801038384437561,
+      "learning_rate": 4.440779451812967e-05,
+      "loss": 0.391,
       "step": 33390
     },
     {
-      "epoch": 1.17,
-      "learning_rate": 4.46935594755925e-05,
-      "loss": 0.2992,
+      "epoch": 1.2035535373193498,
+      "grad_norm": 0.18195895850658417,
+      "learning_rate": 4.4405954942927155e-05,
+      "loss": 0.4165,
       "step": 33395
     },
     {
-      "epoch": 1.18,
-      "learning_rate": 4.4691804521668865e-05,
-      "loss": 0.2822,
+      "epoch": 1.2037337369805745,
+      "grad_norm": 0.22114919126033783,
+      "learning_rate": 4.440411510332226e-05,
+      "loss": 0.4255,
       "step": 33400
     },
     {
-      "epoch": 1.18,
-      "learning_rate": 4.4690049312060126e-05,
-      "loss": 0.3032,
+      "epoch": 1.203913936641799,
+      "grad_norm": 0.18893185257911682,
+      "learning_rate": 4.4402274999340065e-05,
+      "loss": 0.4319,
       "step": 33405
     },
     {
-      "epoch": 1.18,
-      "learning_rate": 4.468829384678907e-05,
-      "loss": 0.2915,
+      "epoch": 1.2040941363030238,
+      "grad_norm": 0.19673962891101837,
+      "learning_rate": 4.4400434631005626e-05,
+      "loss": 0.4201,
       "step": 33410
     },
     {
-      "epoch": 1.18,
-      "learning_rate": 4.4686538125878485e-05,
-      "loss": 0.2831,
+      "epoch": 1.2042743359642483,
+      "grad_norm": 0.21751564741134644,
+      "learning_rate": 4.439859399834402e-05,
+      "loss": 0.4748,
       "step": 33415
     },
     {
-      "epoch": 1.18,
-      "learning_rate": 4.468478214935118e-05,
-      "loss": 0.2915,
+      "epoch": 1.204454535625473,
+      "grad_norm": 0.17174293100833893,
+      "learning_rate": 4.4396753101380316e-05,
+      "loss": 0.3964,
       "step": 33420
     },
     {
-      "epoch": 1.18,
-      "learning_rate": 4.4683025917229956e-05,
-      "loss": 0.3001,
+      "epoch": 1.2046347352866977,
+      "grad_norm": 0.18520551919937134,
+      "learning_rate": 4.4394911940139616e-05,
+      "loss": 0.4367,
       "step": 33425
     },
     {
-      "epoch": 1.18,
-      "learning_rate": 4.468126942953761e-05,
-      "loss": 0.3102,
+      "epoch": 1.2048149349479222,
+      "grad_norm": 0.16532555222511292,
+      "learning_rate": 4.4393070514647e-05,
+      "loss": 0.4135,
       "step": 33430
     },
     {
-      "epoch": 1.18,
-      "learning_rate": 4.467951268629695e-05,
-      "loss": 0.2899,
+      "epoch": 1.204995134609147,
+      "grad_norm": 0.1470910906791687,
+      "learning_rate": 4.439122882492754e-05,
+      "loss": 0.4098,
       "step": 33435
     },
     {
-      "epoch": 1.18,
-      "learning_rate": 4.4677755687530787e-05,
-      "loss": 0.2786,
+      "epoch": 1.2051753342703715,
+      "grad_norm": 0.18562817573547363,
+      "learning_rate": 4.438938687100636e-05,
+      "loss": 0.4498,
       "step": 33440
     },
     {
-      "epoch": 1.18,
-      "learning_rate": 4.467599843326193e-05,
-      "loss": 0.2695,
+      "epoch": 1.2053555339315962,
+      "grad_norm": 0.14606645703315735,
+      "learning_rate": 4.438754465290852e-05,
+      "loss": 0.4095,
       "step": 33445
     },
     {
-      "epoch": 1.18,
-      "learning_rate": 4.467424092351321e-05,
-      "loss": 0.3129,
+      "epoch": 1.205535733592821,
+      "grad_norm": 0.15203700959682465,
+      "learning_rate": 4.4385702170659144e-05,
+      "loss": 0.3668,
       "step": 33450
     },
     {
-      "epoch": 1.18,
-      "learning_rate": 4.467248315830743e-05,
-      "loss": 0.273,
+      "epoch": 1.2057159332540455,
+      "grad_norm": 0.17846938967704773,
+      "learning_rate": 4.4383859424283325e-05,
+      "loss": 0.4517,
       "step": 33455
     },
     {
-      "epoch": 1.18,
-      "learning_rate": 4.4670725137667425e-05,
-      "loss": 0.2893,
+      "epoch": 1.2058961329152702,
+      "grad_norm": 0.16596698760986328,
+      "learning_rate": 4.438201641380618e-05,
+      "loss": 0.3891,
       "step": 33460
     },
     {
-      "epoch": 1.18,
-      "learning_rate": 4.466896686161601e-05,
-      "loss": 0.2975,
+      "epoch": 1.2060763325764947,
+      "grad_norm": 0.14986424148082733,
+      "learning_rate": 4.438017313925281e-05,
+      "loss": 0.3901,
       "step": 33465
     },
     {
-      "epoch": 1.18,
-      "learning_rate": 4.466720833017604e-05,
-      "loss": 0.2891,
+      "epoch": 1.2062565322377194,
+      "grad_norm": 0.18405571579933167,
+      "learning_rate": 4.4378329600648337e-05,
+      "loss": 0.4215,
       "step": 33470
     },
     {
-      "epoch": 1.18,
-      "learning_rate": 4.4665449543370316e-05,
-      "loss": 0.2937,
+      "epoch": 1.206436731898944,
+      "grad_norm": 0.16309815645217896,
+      "learning_rate": 4.437648579801788e-05,
+      "loss": 0.4015,
       "step": 33475
     },
     {
-      "epoch": 1.18,
-      "learning_rate": 4.466369050122169e-05,
-      "loss": 0.2957,
+      "epoch": 1.2066169315601687,
+      "grad_norm": 0.16883867979049683,
+      "learning_rate": 4.437464173138655e-05,
+      "loss": 0.4407,
       "step": 33480
     },
     {
-      "epoch": 1.18,
-      "learning_rate": 4.466193120375301e-05,
-      "loss": 0.2962,
+      "epoch": 1.2067971312213932,
+      "grad_norm": 0.18659423291683197,
+      "learning_rate": 4.437279740077947e-05,
+      "loss": 0.4177,
       "step": 33485
     },
     {
-      "epoch": 1.18,
-      "learning_rate": 4.46601716509871e-05,
-      "loss": 0.3038,
+      "epoch": 1.206977330882618,
+      "grad_norm": 0.2079888880252838,
+      "learning_rate": 4.437095280622178e-05,
+      "loss": 0.4201,
       "step": 33490
     },
     {
-      "epoch": 1.18,
-      "learning_rate": 4.465841184294682e-05,
-      "loss": 0.2816,
+      "epoch": 1.2071575305438427,
+      "grad_norm": 0.15758785605430603,
+      "learning_rate": 4.4369107947738606e-05,
+      "loss": 0.4037,
       "step": 33495
     },
     {
-      "epoch": 1.18,
-      "learning_rate": 4.465665177965502e-05,
-      "loss": 0.3047,
+      "epoch": 1.2073377302050672,
+      "grad_norm": 0.18078656494617462,
+      "learning_rate": 4.436726282535509e-05,
+      "loss": 0.4328,
       "step": 33500
     },
     {
-      "epoch": 1.18,
-      "eval_loss": 0.2884669601917267,
-      "eval_runtime": 10.5481,
-      "eval_samples_per_second": 9.48,
-      "eval_steps_per_second": 9.48,
+      "epoch": 1.2073377302050672,
+      "eval_loss": 0.45027703046798706,
+      "eval_runtime": 3.5418,
+      "eval_samples_per_second": 28.234,
+      "eval_steps_per_second": 7.058,
       "step": 33500
     },
     {
-      "epoch": 1.18,
-      "learning_rate": 4.4654891461134556e-05,
-      "loss": 0.2963,
+      "epoch": 1.207517929866292,
+      "grad_norm": 0.18480472266674042,
+      "learning_rate": 4.436541743909637e-05,
+      "loss": 0.4078,
       "step": 33505
     },
     {
-      "epoch": 1.18,
-      "learning_rate": 4.465313088740828e-05,
-      "loss": 0.2943,
+      "epoch": 1.2076981295275164,
+      "grad_norm": 0.24424877762794495,
+      "learning_rate": 4.436357178898759e-05,
+      "loss": 0.4427,
       "step": 33510
     },
     {
-      "epoch": 1.18,
-      "learning_rate": 4.4651370058499045e-05,
-      "loss": 0.3003,
+      "epoch": 1.2078783291887412,
+      "grad_norm": 0.15822868049144745,
+      "learning_rate": 4.436172587505389e-05,
+      "loss": 0.4074,
       "step": 33515
     },
     {
-      "epoch": 1.18,
-      "learning_rate": 4.464960897442973e-05,
-      "loss": 0.2907,
+      "epoch": 1.2080585288499657,
+      "grad_norm": 0.1882435828447342,
+      "learning_rate": 4.435987969732042e-05,
+      "loss": 0.4272,
       "step": 33520
     },
     {
-      "epoch": 1.18,
-      "learning_rate": 4.4647847635223175e-05,
-      "loss": 0.2912,
+      "epoch": 1.2082387285111904,
+      "grad_norm": 0.18066862225532532,
+      "learning_rate": 4.4358033255812334e-05,
+      "loss": 0.3978,
       "step": 33525
     },
     {
-      "epoch": 1.18,
-      "learning_rate": 4.464608604090228e-05,
-      "loss": 0.2866,
+      "epoch": 1.208418928172415,
+      "grad_norm": 0.18334537744522095,
+      "learning_rate": 4.43561865505548e-05,
+      "loss": 0.4683,
       "step": 33530
     },
     {
-      "epoch": 1.18,
-      "learning_rate": 4.46443241914899e-05,
-      "loss": 0.2882,
+      "epoch": 1.2085991278336397,
+      "grad_norm": 0.17291270196437836,
+      "learning_rate": 4.4354339581572974e-05,
+      "loss": 0.4201,
       "step": 33535
     },
     {
-      "epoch": 1.18,
-      "learning_rate": 4.464256208700892e-05,
-      "loss": 0.2638,
+      "epoch": 1.2087793274948644,
+      "grad_norm": 0.19278649985790253,
+      "learning_rate": 4.4352492348892015e-05,
+      "loss": 0.4244,
       "step": 33540
     },
     {
-      "epoch": 1.18,
-      "learning_rate": 4.464079972748221e-05,
-      "loss": 0.3069,
+      "epoch": 1.208959527156089,
+      "grad_norm": 0.17571572959423065,
+      "learning_rate": 4.435064485253709e-05,
+      "loss": 0.4039,
       "step": 33545
     },
     {
-      "epoch": 1.18,
-      "learning_rate": 4.463903711293267e-05,
-      "loss": 0.283,
+      "epoch": 1.2091397268173136,
+      "grad_norm": 0.21595270931720734,
+      "learning_rate": 4.434879709253338e-05,
+      "loss": 0.4619,
       "step": 33550
     },
     {
-      "epoch": 1.18,
-      "learning_rate": 4.463727424338316e-05,
-      "loss": 0.2771,
+      "epoch": 1.2093199264785381,
+      "grad_norm": 0.16592943668365479,
+      "learning_rate": 4.434694906890605e-05,
+      "loss": 0.4259,
       "step": 33555
     },
     {
-      "epoch": 1.18,
-      "learning_rate": 4.46355111188566e-05,
-      "loss": 0.2906,
+      "epoch": 1.2095001261397629,
+      "grad_norm": 0.1708192229270935,
+      "learning_rate": 4.434510078168029e-05,
+      "loss": 0.4389,
       "step": 33560
     },
     {
-      "epoch": 1.18,
-      "learning_rate": 4.463374773937587e-05,
-      "loss": 0.3029,
+      "epoch": 1.2096803258009876,
+      "grad_norm": 0.13730975985527039,
+      "learning_rate": 4.434325223088128e-05,
+      "loss": 0.3832,
       "step": 33565
     },
     {
-      "epoch": 1.18,
-      "learning_rate": 4.4631984104963854e-05,
-      "loss": 0.2964,
+      "epoch": 1.2098605254622121,
+      "grad_norm": 0.1926315575838089,
+      "learning_rate": 4.43414034165342e-05,
+      "loss": 0.3929,
       "step": 33570
     },
     {
-      "epoch": 1.18,
-      "learning_rate": 4.463022021564346e-05,
-      "loss": 0.2757,
+      "epoch": 1.2100407251234369,
+      "grad_norm": 0.16301114857196808,
+      "learning_rate": 4.433955433866424e-05,
+      "loss": 0.3981,
       "step": 33575
     },
     {
-      "epoch": 1.18,
-      "learning_rate": 4.46284560714376e-05,
-      "loss": 0.2749,
+      "epoch": 1.2102209247846614,
+      "grad_norm": 0.16956037282943726,
+      "learning_rate": 4.4337704997296604e-05,
+      "loss": 0.4255,
       "step": 33580
     },
     {
-      "epoch": 1.18,
-      "learning_rate": 4.462669167236917e-05,
-      "loss": 0.3006,
+      "epoch": 1.210401124445886,
+      "grad_norm": 0.23975764214992523,
+      "learning_rate": 4.4335855392456474e-05,
+      "loss": 0.4243,
       "step": 33585
     },
     {
-      "epoch": 1.18,
-      "learning_rate": 4.462492701846109e-05,
-      "loss": 0.2939,
+      "epoch": 1.2105813241071106,
+      "grad_norm": 0.1944325715303421,
+      "learning_rate": 4.4334005524169066e-05,
+      "loss": 0.4094,
       "step": 33590
     },
     {
-      "epoch": 1.18,
-      "learning_rate": 4.462316210973626e-05,
-      "loss": 0.2729,
+      "epoch": 1.2107615237683353,
+      "grad_norm": 0.17085929214954376,
+      "learning_rate": 4.433215539245956e-05,
+      "loss": 0.4084,
       "step": 33595
     },
     {
-      "epoch": 1.18,
-      "learning_rate": 4.46213969462176e-05,
-      "loss": 0.2856,
+      "epoch": 1.2109417234295599,
+      "grad_norm": 0.21070963144302368,
+      "learning_rate": 4.43303049973532e-05,
+      "loss": 0.4491,
       "step": 33600
     },
     {
-      "epoch": 1.18,
-      "learning_rate": 4.4619631527928043e-05,
-      "loss": 0.3168,
+      "epoch": 1.2111219230907846,
+      "grad_norm": 0.158106729388237,
+      "learning_rate": 4.4328454338875156e-05,
+      "loss": 0.3992,
       "step": 33605
     },
     {
-      "epoch": 1.18,
-      "learning_rate": 4.4617865854890495e-05,
-      "loss": 0.2675,
+      "epoch": 1.2113021227520093,
+      "grad_norm": 0.19057874381542206,
+      "learning_rate": 4.4326603417050676e-05,
+      "loss": 0.3716,
       "step": 33610
     },
     {
-      "epoch": 1.18,
-      "learning_rate": 4.461609992712788e-05,
-      "loss": 0.2795,
+      "epoch": 1.2114823224132338,
+      "grad_norm": 0.1798359900712967,
+      "learning_rate": 4.4324752231904965e-05,
+      "loss": 0.4243,
       "step": 33615
     },
     {
-      "epoch": 1.18,
-      "learning_rate": 4.461433374466314e-05,
-      "loss": 0.2885,
+      "epoch": 1.2116625220744586,
+      "grad_norm": 0.20672789216041565,
+      "learning_rate": 4.432290078346324e-05,
+      "loss": 0.4384,
       "step": 33620
     },
     {
-      "epoch": 1.18,
-      "learning_rate": 4.4612567307519205e-05,
-      "loss": 0.2828,
+      "epoch": 1.211842721735683,
+      "grad_norm": 0.18144282698631287,
+      "learning_rate": 4.4321049071750743e-05,
+      "loss": 0.3966,
       "step": 33625
     },
     {
-      "epoch": 1.18,
-      "learning_rate": 4.4610800615719007e-05,
-      "loss": 0.299,
+      "epoch": 1.2120229213969078,
+      "grad_norm": 0.2156475931406021,
+      "learning_rate": 4.431919709679269e-05,
+      "loss": 0.4488,
       "step": 33630
     },
     {
-      "epoch": 1.18,
-      "learning_rate": 4.4609033669285485e-05,
-      "loss": 0.2943,
+      "epoch": 1.2122031210581323,
+      "grad_norm": 0.1874171942472458,
+      "learning_rate": 4.431734485861432e-05,
+      "loss": 0.4277,
       "step": 33635
     },
     {
-      "epoch": 1.18,
-      "learning_rate": 4.460726646824159e-05,
-      "loss": 0.2989,
+      "epoch": 1.212383320719357,
+      "grad_norm": 0.20030780136585236,
+      "learning_rate": 4.431549235724086e-05,
+      "loss": 0.3958,
       "step": 33640
     },
     {
-      "epoch": 1.18,
-      "learning_rate": 4.460549901261025e-05,
-      "loss": 0.2855,
+      "epoch": 1.2125635203805816,
+      "grad_norm": 0.16703949868679047,
+      "learning_rate": 4.431363959269755e-05,
+      "loss": 0.4372,
       "step": 33645
     },
     {
-      "epoch": 1.18,
-      "learning_rate": 4.4603731302414445e-05,
-      "loss": 0.2958,
+      "epoch": 1.2127437200418063,
+      "grad_norm": 0.1703837811946869,
+      "learning_rate": 4.431178656500965e-05,
+      "loss": 0.404,
       "step": 33650
     },
     {
-      "epoch": 1.18,
-      "learning_rate": 4.460196333767709e-05,
-      "loss": 0.2847,
+      "epoch": 1.212923919703031,
+      "grad_norm": 0.1517026126384735,
+      "learning_rate": 4.43099332742024e-05,
+      "loss": 0.4263,
       "step": 33655
     },
     {
-      "epoch": 1.18,
-      "learning_rate": 4.460019511842117e-05,
-      "loss": 0.2869,
+      "epoch": 1.2131041193642556,
+      "grad_norm": 0.21069294214248657,
+      "learning_rate": 4.430807972030104e-05,
+      "loss": 0.3994,
       "step": 33660
     },
     {
-      "epoch": 1.18,
-      "learning_rate": 4.459842664466962e-05,
-      "loss": 0.2747,
+      "epoch": 1.2132843190254803,
+      "grad_norm": 0.1706804782152176,
+      "learning_rate": 4.430622590333083e-05,
+      "loss": 0.4113,
       "step": 33665
     },
     {
-      "epoch": 1.18,
-      "learning_rate": 4.459665791644543e-05,
-      "loss": 0.2878,
+      "epoch": 1.2134645186867048,
+      "grad_norm": 0.20271262526512146,
+      "learning_rate": 4.430437182331704e-05,
+      "loss": 0.4189,
       "step": 33670
     },
     {
-      "epoch": 1.18,
-      "learning_rate": 4.459488893377155e-05,
-      "loss": 0.3042,
+      "epoch": 1.2136447183479295,
+      "grad_norm": 0.1703236848115921,
+      "learning_rate": 4.430251748028492e-05,
+      "loss": 0.4197,
       "step": 33675
     },
     {
-      "epoch": 1.18,
-      "learning_rate": 4.459311969667095e-05,
-      "loss": 0.2931,
+      "epoch": 1.2138249180091543,
+      "grad_norm": 0.16944654285907745,
+      "learning_rate": 4.430066287425973e-05,
+      "loss": 0.4716,
       "step": 33680
     },
     {
-      "epoch": 1.19,
-      "learning_rate": 4.459135020516659e-05,
-      "loss": 0.2853,
+      "epoch": 1.2140051176703788,
+      "grad_norm": 0.17633916437625885,
+      "learning_rate": 4.429880800526675e-05,
+      "loss": 0.3921,
       "step": 33685
     },
     {
-      "epoch": 1.19,
-      "learning_rate": 4.4589580459281474e-05,
-      "loss": 0.287,
+      "epoch": 1.2141853173316035,
+      "grad_norm": 0.2126268744468689,
+      "learning_rate": 4.4296952873331235e-05,
+      "loss": 0.4292,
       "step": 33690
     },
     {
-      "epoch": 1.19,
-      "learning_rate": 4.4587810459038556e-05,
-      "loss": 0.2946,
+      "epoch": 1.214365516992828,
+      "grad_norm": 0.1921420842409134,
+      "learning_rate": 4.4295097478478484e-05,
+      "loss": 0.4163,
       "step": 33695
     },
     {
-      "epoch": 1.19,
-      "learning_rate": 4.458604020446084e-05,
-      "loss": 0.292,
+      "epoch": 1.2145457166540528,
+      "grad_norm": 0.19828033447265625,
+      "learning_rate": 4.4293241820733764e-05,
+      "loss": 0.4112,
       "step": 33700
     },
     {
-      "epoch": 1.19,
-      "learning_rate": 4.458426969557129e-05,
-      "loss": 0.3181,
+      "epoch": 1.2147259163152773,
+      "grad_norm": 0.171315535902977,
+      "learning_rate": 4.429138590012236e-05,
+      "loss": 0.4352,
       "step": 33705
     },
     {
-      "epoch": 1.19,
-      "learning_rate": 4.4582498932392906e-05,
-      "loss": 0.2827,
+      "epoch": 1.214906115976502,
+      "grad_norm": 0.15419135987758636,
+      "learning_rate": 4.428952971666956e-05,
+      "loss": 0.4077,
       "step": 33710
     },
     {
-      "epoch": 1.19,
-      "learning_rate": 4.4580727914948685e-05,
-      "loss": 0.2826,
+      "epoch": 1.2150863156377265,
+      "grad_norm": 0.16436229646205902,
+      "learning_rate": 4.428767327040065e-05,
+      "loss": 0.4068,
       "step": 33715
     },
     {
-      "epoch": 1.19,
-      "learning_rate": 4.45789566432616e-05,
-      "loss": 0.2858,
+      "epoch": 1.2152665152989512,
+      "grad_norm": 0.17633068561553955,
+      "learning_rate": 4.428581656134092e-05,
+      "loss": 0.4457,
       "step": 33720
     },
     {
-      "epoch": 1.19,
-      "learning_rate": 4.457718511735468e-05,
-      "loss": 0.2845,
+      "epoch": 1.215446714960176,
+      "grad_norm": 0.1803673654794693,
+      "learning_rate": 4.4283959589515686e-05,
+      "loss": 0.4286,
       "step": 33725
     },
     {
-      "epoch": 1.19,
-      "learning_rate": 4.4575413337250914e-05,
-      "loss": 0.3087,
+      "epoch": 1.2156269146214005,
+      "grad_norm": 0.14655368030071259,
+      "learning_rate": 4.428210235495023e-05,
+      "loss": 0.4373,
       "step": 33730
     },
     {
-      "epoch": 1.19,
-      "learning_rate": 4.45736413029733e-05,
-      "loss": 0.2915,
+      "epoch": 1.2158071142826252,
+      "grad_norm": 0.18454967439174652,
+      "learning_rate": 4.428024485766986e-05,
+      "loss": 0.4165,
       "step": 33735
     },
     {
-      "epoch": 1.19,
-      "learning_rate": 4.457186901454485e-05,
-      "loss": 0.2721,
+      "epoch": 1.2159873139438497,
+      "grad_norm": 0.1603035181760788,
+      "learning_rate": 4.427838709769989e-05,
+      "loss": 0.3995,
       "step": 33740
     },
     {
-      "epoch": 1.19,
-      "learning_rate": 4.4570096471988596e-05,
-      "loss": 0.3164,
+      "epoch": 1.2161675136050745,
+      "grad_norm": 0.1812167763710022,
+      "learning_rate": 4.427652907506562e-05,
+      "loss": 0.4046,
       "step": 33745
     },
     {
-      "epoch": 1.19,
-      "learning_rate": 4.4568323675327525e-05,
-      "loss": 0.2786,
+      "epoch": 1.216347713266299,
+      "grad_norm": 0.16783830523490906,
+      "learning_rate": 4.427467078979238e-05,
+      "loss": 0.4103,
       "step": 33750
     },
     {
-      "epoch": 1.19,
-      "learning_rate": 4.456655062458467e-05,
-      "loss": 0.2895,
+      "epoch": 1.2165279129275237,
+      "grad_norm": 0.20408573746681213,
+      "learning_rate": 4.427281224190548e-05,
+      "loss": 0.4424,
       "step": 33755
     },
     {
-      "epoch": 1.19,
-      "learning_rate": 4.4564777319783047e-05,
-      "loss": 0.2568,
+      "epoch": 1.2167081125887482,
+      "grad_norm": 0.13193145394325256,
+      "learning_rate": 4.427095343143025e-05,
+      "loss": 0.4065,
       "step": 33760
     },
     {
-      "epoch": 1.19,
-      "learning_rate": 4.456300376094568e-05,
-      "loss": 0.2961,
+      "epoch": 1.216888312249973,
+      "grad_norm": 0.1420491337776184,
+      "learning_rate": 4.4269094358392e-05,
+      "loss": 0.4043,
       "step": 33765
     },
     {
-      "epoch": 1.19,
-      "learning_rate": 4.456122994809561e-05,
-      "loss": 0.3038,
+      "epoch": 1.2170685119111977,
+      "grad_norm": 0.19709770381450653,
+      "learning_rate": 4.4267235022816084e-05,
+      "loss": 0.4288,
       "step": 33770
     },
     {
-      "epoch": 1.19,
-      "learning_rate": 4.455945588125585e-05,
-      "loss": 0.2928,
+      "epoch": 1.2172487115724222,
+      "grad_norm": 0.17064298689365387,
+      "learning_rate": 4.4265375424727815e-05,
+      "loss": 0.4346,
       "step": 33775
     },
     {
-      "epoch": 1.19,
-      "learning_rate": 4.4557681560449456e-05,
-      "loss": 0.2999,
+      "epoch": 1.217428911233647,
+      "grad_norm": 0.16574378311634064,
+      "learning_rate": 4.4263515564152534e-05,
+      "loss": 0.4228,
       "step": 33780
     },
     {
-      "epoch": 1.19,
-      "learning_rate": 4.455590698569945e-05,
-      "loss": 0.3193,
+      "epoch": 1.2176091108948714,
+      "grad_norm": 0.20119209587574005,
+      "learning_rate": 4.426165544111558e-05,
+      "loss": 0.4042,
       "step": 33785
     },
     {
-      "epoch": 1.19,
-      "learning_rate": 4.4554132157028885e-05,
-      "loss": 0.2992,
+      "epoch": 1.2177893105560962,
+      "grad_norm": 0.16486892104148865,
+      "learning_rate": 4.4259795055642305e-05,
+      "loss": 0.4075,
       "step": 33790
     },
     {
-      "epoch": 1.19,
-      "learning_rate": 4.4552357074460796e-05,
-      "loss": 0.3181,
+      "epoch": 1.2179695102173207,
+      "grad_norm": 0.17339356243610382,
+      "learning_rate": 4.425793440775805e-05,
+      "loss": 0.4147,
       "step": 33795
     },
     {
-      "epoch": 1.19,
-      "learning_rate": 4.4550581738018236e-05,
-      "loss": 0.2944,
+      "epoch": 1.2181497098785454,
+      "grad_norm": 0.16329234838485718,
+      "learning_rate": 4.425607349748816e-05,
+      "loss": 0.411,
       "step": 33800
     },
     {
-      "epoch": 1.19,
-      "learning_rate": 4.454880614772426e-05,
-      "loss": 0.2957,
+      "epoch": 1.2183299095397702,
+      "grad_norm": 0.17299103736877441,
+      "learning_rate": 4.425421232485801e-05,
+      "loss": 0.4047,
       "step": 33805
     },
     {
-      "epoch": 1.19,
-      "learning_rate": 4.454703030360192e-05,
-      "loss": 0.3041,
+      "epoch": 1.2185101092009947,
+      "grad_norm": 0.18187008798122406,
+      "learning_rate": 4.4252350889892936e-05,
+      "loss": 0.4287,
       "step": 33810
     },
     {
-      "epoch": 1.19,
-      "learning_rate": 4.4545254205674266e-05,
-      "loss": 0.2972,
+      "epoch": 1.2186903088622194,
+      "grad_norm": 0.19990107417106628,
+      "learning_rate": 4.425048919261832e-05,
+      "loss": 0.4089,
       "step": 33815
     },
     {
-      "epoch": 1.19,
-      "learning_rate": 4.4543477853964375e-05,
-      "loss": 0.2742,
+      "epoch": 1.218870508523444,
+      "grad_norm": 0.1538524031639099,
+      "learning_rate": 4.4248627233059505e-05,
+      "loss": 0.4271,
       "step": 33820
     },
     {
-      "epoch": 1.19,
-      "learning_rate": 4.45417012484953e-05,
-      "loss": 0.279,
+      "epoch": 1.2190507081846687,
+      "grad_norm": 0.18129165470600128,
+      "learning_rate": 4.4246765011241864e-05,
+      "loss": 0.3741,
       "step": 33825
     },
     {
-      "epoch": 1.19,
-      "learning_rate": 4.453992438929011e-05,
-      "loss": 0.2805,
+      "epoch": 1.2192309078458932,
+      "grad_norm": 0.20447011291980743,
+      "learning_rate": 4.4244902527190785e-05,
+      "loss": 0.4431,
       "step": 33830
     },
     {
-      "epoch": 1.19,
-      "learning_rate": 4.4538147276371884e-05,
-      "loss": 0.2776,
+      "epoch": 1.219411107507118,
+      "grad_norm": 0.148980513215065,
+      "learning_rate": 4.424303978093163e-05,
+      "loss": 0.3938,
       "step": 33835
     },
     {
-      "epoch": 1.19,
-      "learning_rate": 4.4536369909763683e-05,
-      "loss": 0.2843,
+      "epoch": 1.2195913071683426,
+      "grad_norm": 0.14329323172569275,
+      "learning_rate": 4.424117677248979e-05,
+      "loss": 0.4257,
       "step": 33840
     },
     {
-      "epoch": 1.19,
-      "learning_rate": 4.453459228948861e-05,
-      "loss": 0.2555,
+      "epoch": 1.2197715068295671,
+      "grad_norm": 0.22545427083969116,
+      "learning_rate": 4.423931350189065e-05,
+      "loss": 0.4461,
       "step": 33845
     },
     {
-      "epoch": 1.19,
-      "learning_rate": 4.453281441556972e-05,
-      "loss": 0.3132,
+      "epoch": 1.2199517064907919,
+      "grad_norm": 0.17364872992038727,
+      "learning_rate": 4.423744996915957e-05,
+      "loss": 0.445,
       "step": 33850
     },
     {
-      "epoch": 1.19,
-      "learning_rate": 4.45310362880301e-05,
-      "loss": 0.3091,
+      "epoch": 1.2201319061520164,
+      "grad_norm": 0.2247709333896637,
+      "learning_rate": 4.4235586174321964e-05,
+      "loss": 0.4471,
       "step": 33855
     },
     {
-      "epoch": 1.19,
-      "learning_rate": 4.4529257906892864e-05,
-      "loss": 0.2793,
+      "epoch": 1.2203121058132411,
+      "grad_norm": 0.18981455266475677,
+      "learning_rate": 4.423372211740323e-05,
+      "loss": 0.4089,
       "step": 33860
     },
     {
-      "epoch": 1.19,
-      "learning_rate": 4.4527479272181074e-05,
-      "loss": 0.2814,
+      "epoch": 1.2204923054744656,
+      "grad_norm": 0.15154671669006348,
+      "learning_rate": 4.423185779842874e-05,
+      "loss": 0.4149,
       "step": 33865
     },
     {
-      "epoch": 1.19,
-      "learning_rate": 4.452570038391783e-05,
-      "loss": 0.3016,
+      "epoch": 1.2206725051356904,
+      "grad_norm": 0.1997554749250412,
+      "learning_rate": 4.422999321742393e-05,
+      "loss": 0.3822,
       "step": 33870
     },
     {
-      "epoch": 1.19,
-      "learning_rate": 4.452392124212624e-05,
-      "loss": 0.2923,
+      "epoch": 1.2208527047969149,
+      "grad_norm": 0.21992482244968414,
+      "learning_rate": 4.422812837441417e-05,
+      "loss": 0.4158,
       "step": 33875
     },
     {
-      "epoch": 1.19,
-      "learning_rate": 4.45221418468294e-05,
-      "loss": 0.2731,
+      "epoch": 1.2210329044581396,
+      "grad_norm": 0.19677501916885376,
+      "learning_rate": 4.422626326942489e-05,
+      "loss": 0.4037,
       "step": 33880
     },
     {
-      "epoch": 1.19,
-      "learning_rate": 4.452036219805041e-05,
-      "loss": 0.3242,
+      "epoch": 1.2212131041193643,
+      "grad_norm": 0.1691318303346634,
+      "learning_rate": 4.422439790248149e-05,
+      "loss": 0.4369,
       "step": 33885
     },
     {
-      "epoch": 1.19,
-      "learning_rate": 4.451858229581238e-05,
-      "loss": 0.2567,
+      "epoch": 1.2213933037805889,
+      "grad_norm": 0.15583454072475433,
+      "learning_rate": 4.4222532273609396e-05,
+      "loss": 0.3988,
       "step": 33890
     },
     {
-      "epoch": 1.19,
-      "learning_rate": 4.451680214013842e-05,
-      "loss": 0.3106,
+      "epoch": 1.2215735034418136,
+      "grad_norm": 0.1846490055322647,
+      "learning_rate": 4.422066638283402e-05,
+      "loss": 0.4234,
       "step": 33895
     },
     {
-      "epoch": 1.19,
-      "learning_rate": 4.4515021731051654e-05,
-      "loss": 0.3036,
+      "epoch": 1.221753703103038,
+      "grad_norm": 0.18129578232765198,
+      "learning_rate": 4.421880023018079e-05,
+      "loss": 0.4312,
       "step": 33900
     },
     {
-      "epoch": 1.19,
-      "learning_rate": 4.4513241068575196e-05,
-      "loss": 0.2825,
+      "epoch": 1.2219339027642628,
+      "grad_norm": 0.15360191464424133,
+      "learning_rate": 4.421693381567512e-05,
+      "loss": 0.4302,
       "step": 33905
     },
     {
-      "epoch": 1.19,
-      "learning_rate": 4.4511460152732156e-05,
-      "loss": 0.2796,
+      "epoch": 1.2221141024254873,
+      "grad_norm": 0.2045409232378006,
+      "learning_rate": 4.421506713934245e-05,
+      "loss": 0.4176,
       "step": 33910
     },
     {
-      "epoch": 1.19,
-      "learning_rate": 4.4509678983545654e-05,
-      "loss": 0.2792,
+      "epoch": 1.222294302086712,
+      "grad_norm": 0.15887264907360077,
+      "learning_rate": 4.421320020120821e-05,
+      "loss": 0.4114,
       "step": 33915
     },
     {
-      "epoch": 1.19,
-      "learning_rate": 4.4507897561038845e-05,
-      "loss": 0.266,
+      "epoch": 1.2224745017479366,
+      "grad_norm": 0.18806084990501404,
+      "learning_rate": 4.4211333001297836e-05,
+      "loss": 0.4254,
       "step": 33920
     },
     {
-      "epoch": 1.19,
-      "learning_rate": 4.4506115885234836e-05,
-      "loss": 0.2884,
+      "epoch": 1.2226547014091613,
+      "grad_norm": 0.21067681908607483,
+      "learning_rate": 4.420946553963677e-05,
+      "loss": 0.4186,
       "step": 33925
     },
     {
-      "epoch": 1.19,
-      "learning_rate": 4.450433395615676e-05,
-      "loss": 0.283,
+      "epoch": 1.222834901070386,
+      "grad_norm": 0.15859293937683105,
+      "learning_rate": 4.4207597816250454e-05,
+      "loss": 0.4144,
       "step": 33930
     },
     {
-      "epoch": 1.19,
-      "learning_rate": 4.4502551773827775e-05,
-      "loss": 0.2885,
+      "epoch": 1.2230151007316106,
+      "grad_norm": 0.1685830056667328,
+      "learning_rate": 4.420572983116434e-05,
+      "loss": 0.4313,
       "step": 33935
     },
     {
-      "epoch": 1.19,
-      "learning_rate": 4.4500769338271e-05,
-      "loss": 0.2917,
+      "epoch": 1.2231953003928353,
+      "grad_norm": 0.17942924797534943,
+      "learning_rate": 4.420386158440388e-05,
+      "loss": 0.4175,
       "step": 33940
     },
     {
-      "epoch": 1.19,
-      "learning_rate": 4.4498986649509584e-05,
-      "loss": 0.2898,
+      "epoch": 1.2233755000540598,
+      "grad_norm": 0.19803176820278168,
+      "learning_rate": 4.420199307599452e-05,
+      "loss": 0.4151,
       "step": 33945
     },
     {
-      "epoch": 1.19,
-      "learning_rate": 4.449720370756668e-05,
-      "loss": 0.2922,
+      "epoch": 1.2235556997152846,
+      "grad_norm": 0.18446727097034454,
+      "learning_rate": 4.420012430596172e-05,
+      "loss": 0.4478,
       "step": 33950
     },
     {
-      "epoch": 1.19,
-      "learning_rate": 4.449542051246544e-05,
-      "loss": 0.3001,
+      "epoch": 1.2237358993765093,
+      "grad_norm": 0.15331429243087769,
+      "learning_rate": 4.419825527433095e-05,
+      "loss": 0.424,
       "step": 33955
     },
     {
-      "epoch": 1.19,
-      "learning_rate": 4.4493637064229e-05,
-      "loss": 0.2668,
+      "epoch": 1.2239160990377338,
+      "grad_norm": 0.21690736711025238,
+      "learning_rate": 4.419638598112765e-05,
+      "loss": 0.4068,
       "step": 33960
     },
     {
-      "epoch": 1.19,
-      "learning_rate": 4.449185336288054e-05,
-      "loss": 0.2862,
+      "epoch": 1.2240962986989585,
+      "grad_norm": 0.18475008010864258,
+      "learning_rate": 4.4194516426377326e-05,
+      "loss": 0.4087,
       "step": 33965
     },
     {
-      "epoch": 1.2,
-      "learning_rate": 4.44900694084432e-05,
-      "loss": 0.2975,
+      "epoch": 1.224276498360183,
+      "grad_norm": 0.19182796776294708,
+      "learning_rate": 4.4192646610105425e-05,
+      "loss": 0.4304,
       "step": 33970
     },
     {
-      "epoch": 1.2,
-      "learning_rate": 4.448828520094017e-05,
-      "loss": 0.293,
+      "epoch": 1.2244566980214078,
+      "grad_norm": 0.15352657437324524,
+      "learning_rate": 4.419077653233743e-05,
+      "loss": 0.4087,
       "step": 33975
     },
     {
-      "epoch": 1.2,
-      "learning_rate": 4.448650074039459e-05,
-      "loss": 0.2925,
+      "epoch": 1.2246368976826323,
+      "grad_norm": 0.18293914198875427,
+      "learning_rate": 4.418890619309882e-05,
+      "loss": 0.4337,
       "step": 33980
     },
     {
-      "epoch": 1.2,
-      "learning_rate": 4.4484716026829635e-05,
-      "loss": 0.2922,
+      "epoch": 1.224817097343857,
+      "grad_norm": 0.22879280149936676,
+      "learning_rate": 4.4187035592415085e-05,
+      "loss": 0.4208,
       "step": 33985
     },
     {
-      "epoch": 1.2,
-      "learning_rate": 4.448293106026849e-05,
-      "loss": 0.2837,
+      "epoch": 1.2249972970050815,
+      "grad_norm": 0.18083889782428741,
+      "learning_rate": 4.41851647303117e-05,
+      "loss": 0.4357,
       "step": 33990
     },
     {
-      "epoch": 1.2,
-      "learning_rate": 4.448114584073432e-05,
-      "loss": 0.308,
+      "epoch": 1.2251774966663063,
+      "grad_norm": 0.17108914256095886,
+      "learning_rate": 4.4183293606814155e-05,
+      "loss": 0.3936,
       "step": 33995
     },
     {
-      "epoch": 1.2,
-      "learning_rate": 4.447936036825031e-05,
-      "loss": 0.3173,
+      "epoch": 1.225357696327531,
+      "grad_norm": 0.15355034172534943,
+      "learning_rate": 4.418142222194795e-05,
+      "loss": 0.3946,
       "step": 34000
     },
     {
-      "epoch": 1.2,
-      "eval_loss": 0.28807368874549866,
-      "eval_runtime": 10.551,
-      "eval_samples_per_second": 9.478,
-      "eval_steps_per_second": 9.478,
+      "epoch": 1.225357696327531,
+      "eval_loss": 0.45026490092277527,
+      "eval_runtime": 3.5278,
+      "eval_samples_per_second": 28.346,
+      "eval_steps_per_second": 7.086,
       "step": 34000
     },
     {
-      "epoch": 1.2,
-      "learning_rate": 4.447757464283965e-05,
-      "loss": 0.2874,
+      "epoch": 1.2255378959887555,
+      "grad_norm": 0.18301257491111755,
+      "learning_rate": 4.4179550575738584e-05,
+      "loss": 0.4336,
       "step": 34005
     },
     {
-      "epoch": 1.2,
-      "learning_rate": 4.447578866452552e-05,
-      "loss": 0.2941,
+      "epoch": 1.2257180956499802,
+      "grad_norm": 0.16814136505126953,
+      "learning_rate": 4.4177678668211555e-05,
+      "loss": 0.4443,
       "step": 34010
     },
     {
-      "epoch": 1.2,
-      "learning_rate": 4.44740024333311e-05,
-      "loss": 0.2793,
+      "epoch": 1.2258982953112048,
+      "grad_norm": 0.22709350287914276,
+      "learning_rate": 4.417580649939237e-05,
+      "loss": 0.4527,
       "step": 34015
     },
     {
-      "epoch": 1.2,
-      "learning_rate": 4.4472215949279594e-05,
-      "loss": 0.2921,
+      "epoch": 1.2260784949724295,
+      "grad_norm": 0.15572187304496765,
+      "learning_rate": 4.417393406930652e-05,
+      "loss": 0.3646,
       "step": 34020
     },
     {
-      "epoch": 1.2,
-      "learning_rate": 4.44704292123942e-05,
-      "loss": 0.2979,
+      "epoch": 1.226258694633654,
+      "grad_norm": 0.17748330533504486,
+      "learning_rate": 4.4172061377979545e-05,
+      "loss": 0.4317,
       "step": 34025
     },
     {
-      "epoch": 1.2,
-      "learning_rate": 4.446864222269812e-05,
-      "loss": 0.3073,
+      "epoch": 1.2264388942948787,
+      "grad_norm": 0.18140548467636108,
+      "learning_rate": 4.417018842543694e-05,
+      "loss": 0.4283,
       "step": 34030
     },
     {
-      "epoch": 1.2,
-      "learning_rate": 4.446685498021455e-05,
-      "loss": 0.3038,
+      "epoch": 1.2266190939561032,
+      "grad_norm": 0.1755266636610031,
+      "learning_rate": 4.416831521170424e-05,
+      "loss": 0.4276,
       "step": 34035
     },
     {
-      "epoch": 1.2,
-      "learning_rate": 4.446506748496668e-05,
-      "loss": 0.2936,
+      "epoch": 1.226799293617328,
+      "grad_norm": 0.1404469907283783,
+      "learning_rate": 4.416644173680694e-05,
+      "loss": 0.3802,
       "step": 34040
     },
     {
-      "epoch": 1.2,
-      "learning_rate": 4.4463279736977754e-05,
-      "loss": 0.3041,
+      "epoch": 1.2269794932785527,
+      "grad_norm": 0.18569141626358032,
+      "learning_rate": 4.416456800077059e-05,
+      "loss": 0.4175,
       "step": 34045
     },
     {
-      "epoch": 1.2,
-      "learning_rate": 4.446149173627096e-05,
-      "loss": 0.3029,
+      "epoch": 1.2271596929397772,
+      "grad_norm": 0.20885035395622253,
+      "learning_rate": 4.416269400362071e-05,
+      "loss": 0.4017,
       "step": 34050
     },
     {
-      "epoch": 1.2,
-      "learning_rate": 4.445970348286952e-05,
-      "loss": 0.3014,
+      "epoch": 1.227339892601002,
+      "grad_norm": 0.18960420787334442,
+      "learning_rate": 4.416081974538283e-05,
+      "loss": 0.4186,
       "step": 34055
     },
     {
-      "epoch": 1.2,
-      "learning_rate": 4.445791497679666e-05,
-      "loss": 0.2813,
+      "epoch": 1.2275200922622265,
+      "grad_norm": 0.15756478905677795,
+      "learning_rate": 4.41589452260825e-05,
+      "loss": 0.3757,
       "step": 34060
     },
     {
-      "epoch": 1.2,
-      "learning_rate": 4.445612621807559e-05,
-      "loss": 0.2693,
+      "epoch": 1.2277002919234512,
+      "grad_norm": 0.16769862174987793,
+      "learning_rate": 4.415707044574524e-05,
+      "loss": 0.3859,
       "step": 34065
     },
     {
-      "epoch": 1.2,
-      "learning_rate": 4.445433720672955e-05,
-      "loss": 0.2814,
+      "epoch": 1.227880491584676,
+      "grad_norm": 0.21524082124233246,
+      "learning_rate": 4.415519540439661e-05,
+      "loss": 0.46,
       "step": 34070
     },
     {
-      "epoch": 1.2,
-      "learning_rate": 4.4452547942781765e-05,
-      "loss": 0.2809,
+      "epoch": 1.2280606912459004,
+      "grad_norm": 0.1766262948513031,
+      "learning_rate": 4.4153320102062155e-05,
+      "loss": 0.3977,
       "step": 34075
     },
     {
-      "epoch": 1.2,
-      "learning_rate": 4.4450758426255456e-05,
-      "loss": 0.285,
+      "epoch": 1.2282408909071252,
+      "grad_norm": 0.15509352087974548,
+      "learning_rate": 4.4151444538767414e-05,
+      "loss": 0.4288,
       "step": 34080
     },
     {
-      "epoch": 1.2,
-      "learning_rate": 4.444896865717387e-05,
-      "loss": 0.2821,
+      "epoch": 1.2284210905683497,
+      "grad_norm": 0.14794203639030457,
+      "learning_rate": 4.414956871453796e-05,
+      "loss": 0.4274,
       "step": 34085
     },
     {
-      "epoch": 1.2,
-      "learning_rate": 4.444717863556025e-05,
-      "loss": 0.3087,
+      "epoch": 1.2286012902295744,
+      "grad_norm": 0.1935451328754425,
+      "learning_rate": 4.4147692629399326e-05,
+      "loss": 0.3819,
       "step": 34090
     },
     {
-      "epoch": 1.2,
-      "learning_rate": 4.444538836143782e-05,
-      "loss": 0.2888,
+      "epoch": 1.228781489890799,
+      "grad_norm": 0.14869855344295502,
+      "learning_rate": 4.414581628337709e-05,
+      "loss": 0.4268,
       "step": 34095
     },
     {
-      "epoch": 1.2,
-      "learning_rate": 4.444359783482985e-05,
-      "loss": 0.2941,
+      "epoch": 1.2289616895520237,
+      "grad_norm": 0.17884673178195953,
+      "learning_rate": 4.4143939676496825e-05,
+      "loss": 0.4162,
       "step": 34100
     },
     {
-      "epoch": 1.2,
-      "learning_rate": 4.4441807055759574e-05,
-      "loss": 0.3051,
+      "epoch": 1.2291418892132482,
+      "grad_norm": 0.19956587255001068,
+      "learning_rate": 4.414206280878408e-05,
+      "loss": 0.4309,
       "step": 34105
     },
     {
-      "epoch": 1.2,
-      "learning_rate": 4.444001602425024e-05,
-      "loss": 0.2595,
+      "epoch": 1.229322088874473,
+      "grad_norm": 0.1776251643896103,
+      "learning_rate": 4.414018568026443e-05,
+      "loss": 0.4051,
       "step": 34110
     },
     {
-      "epoch": 1.2,
-      "learning_rate": 4.443822474032511e-05,
-      "loss": 0.2836,
+      "epoch": 1.2295022885356977,
+      "grad_norm": 0.16299593448638916,
+      "learning_rate": 4.413830829096347e-05,
+      "loss": 0.4617,
       "step": 34115
     },
     {
-      "epoch": 1.2,
-      "learning_rate": 4.443643320400744e-05,
-      "loss": 0.3046,
+      "epoch": 1.2296824881969222,
+      "grad_norm": 0.17670361697673798,
+      "learning_rate": 4.413643064090675e-05,
+      "loss": 0.399,
       "step": 34120
     },
     {
-      "epoch": 1.2,
-      "learning_rate": 4.44346414153205e-05,
-      "loss": 0.2819,
+      "epoch": 1.229862687858147,
+      "grad_norm": 0.16211190819740295,
+      "learning_rate": 4.4134552730119874e-05,
+      "loss": 0.4074,
       "step": 34125
     },
     {
-      "epoch": 1.2,
-      "learning_rate": 4.443284937428754e-05,
-      "loss": 0.2821,
+      "epoch": 1.2300428875193714,
+      "grad_norm": 0.1310056746006012,
+      "learning_rate": 4.413267455862842e-05,
+      "loss": 0.4109,
       "step": 34130
     },
     {
-      "epoch": 1.2,
-      "learning_rate": 4.443105708093185e-05,
-      "loss": 0.3211,
+      "epoch": 1.2302230871805961,
+      "grad_norm": 0.1611911952495575,
+      "learning_rate": 4.4130796126457984e-05,
+      "loss": 0.3891,
       "step": 34135
     },
     {
-      "epoch": 1.2,
-      "learning_rate": 4.4429264535276684e-05,
-      "loss": 0.2882,
+      "epoch": 1.2304032868418207,
+      "grad_norm": 0.17254267632961273,
+      "learning_rate": 4.412891743363416e-05,
+      "loss": 0.3842,
       "step": 34140
     },
     {
-      "epoch": 1.2,
-      "learning_rate": 4.442747173734532e-05,
-      "loss": 0.3001,
+      "epoch": 1.2305834865030454,
+      "grad_norm": 0.2114710956811905,
+      "learning_rate": 4.412703848018253e-05,
+      "loss": 0.421,
       "step": 34145
     },
     {
-      "epoch": 1.2,
-      "learning_rate": 4.4425678687161046e-05,
-      "loss": 0.2723,
+      "epoch": 1.23076368616427,
+      "grad_norm": 0.1724889725446701,
+      "learning_rate": 4.4125159266128696e-05,
+      "loss": 0.4537,
       "step": 34150
     },
     {
-      "epoch": 1.2,
-      "learning_rate": 4.4423885384747124e-05,
-      "loss": 0.2883,
+      "epoch": 1.2309438858254946,
+      "grad_norm": 0.17160102725028992,
+      "learning_rate": 4.412327979149828e-05,
+      "loss": 0.4127,
       "step": 34155
     },
     {
-      "epoch": 1.2,
-      "learning_rate": 4.442209183012685e-05,
-      "loss": 0.2654,
+      "epoch": 1.2311240854867194,
+      "grad_norm": 0.15789945423603058,
+      "learning_rate": 4.412140005631688e-05,
+      "loss": 0.3848,
       "step": 34160
     },
     {
-      "epoch": 1.2,
-      "learning_rate": 4.442029802332353e-05,
-      "loss": 0.2737,
+      "epoch": 1.2313042851479439,
+      "grad_norm": 0.1946706771850586,
+      "learning_rate": 4.4119520060610105e-05,
+      "loss": 0.4288,
       "step": 34165
     },
     {
-      "epoch": 1.2,
-      "learning_rate": 4.4418503964360424e-05,
-      "loss": 0.29,
+      "epoch": 1.2314844848091686,
+      "grad_norm": 0.1840618997812271,
+      "learning_rate": 4.411763980440357e-05,
+      "loss": 0.3903,
       "step": 34170
     },
     {
-      "epoch": 1.2,
-      "learning_rate": 4.441670965326085e-05,
-      "loss": 0.2884,
+      "epoch": 1.2316646844703931,
+      "grad_norm": 0.16430874168872833,
+      "learning_rate": 4.411575928772289e-05,
+      "loss": 0.4252,
       "step": 34175
     },
     {
-      "epoch": 1.2,
-      "learning_rate": 4.4414915090048095e-05,
-      "loss": 0.3039,
+      "epoch": 1.2318448841316179,
+      "grad_norm": 0.16133743524551392,
+      "learning_rate": 4.41138785105937e-05,
+      "loss": 0.4184,
       "step": 34180
     },
     {
-      "epoch": 1.2,
-      "learning_rate": 4.4413120274745455e-05,
-      "loss": 0.2871,
+      "epoch": 1.2320250837928426,
+      "grad_norm": 0.18589156866073608,
+      "learning_rate": 4.411199747304161e-05,
+      "loss": 0.3987,
       "step": 34185
     },
     {
-      "epoch": 1.2,
-      "learning_rate": 4.441132520737625e-05,
-      "loss": 0.2728,
+      "epoch": 1.232205283454067,
+      "grad_norm": 0.19146478176116943,
+      "learning_rate": 4.4110116175092254e-05,
+      "loss": 0.437,
       "step": 34190
     },
     {
-      "epoch": 1.2,
-      "learning_rate": 4.440952988796378e-05,
-      "loss": 0.3025,
+      "epoch": 1.2323854831152918,
+      "grad_norm": 0.1762928068637848,
+      "learning_rate": 4.410823461677126e-05,
+      "loss": 0.3935,
       "step": 34195
     },
     {
-      "epoch": 1.2,
-      "learning_rate": 4.440773431653136e-05,
-      "loss": 0.2866,
+      "epoch": 1.2325656827765163,
+      "grad_norm": 0.17464253306388855,
+      "learning_rate": 4.4106352798104276e-05,
+      "loss": 0.4299,
       "step": 34200
     },
     {
-      "epoch": 1.2,
-      "learning_rate": 4.440593849310229e-05,
-      "loss": 0.2843,
+      "epoch": 1.232745882437741,
+      "grad_norm": 0.15182888507843018,
+      "learning_rate": 4.410447071911693e-05,
+      "loss": 0.4296,
       "step": 34205
     },
     {
-      "epoch": 1.2,
-      "learning_rate": 4.4404142417699904e-05,
-      "loss": 0.2922,
+      "epoch": 1.2329260820989656,
+      "grad_norm": 0.20655085146427155,
+      "learning_rate": 4.410258837983488e-05,
+      "loss": 0.4122,
       "step": 34210
     },
     {
-      "epoch": 1.2,
-      "learning_rate": 4.4402346090347515e-05,
-      "loss": 0.2985,
+      "epoch": 1.2331062817601903,
+      "grad_norm": 0.15774188935756683,
+      "learning_rate": 4.4100705780283746e-05,
+      "loss": 0.3955,
       "step": 34215
     },
     {
-      "epoch": 1.2,
-      "learning_rate": 4.4400549511068445e-05,
-      "loss": 0.2766,
+      "epoch": 1.2332864814214148,
+      "grad_norm": 0.16499753296375275,
+      "learning_rate": 4.40988229204892e-05,
+      "loss": 0.4257,
       "step": 34220
     },
     {
-      "epoch": 1.2,
-      "learning_rate": 4.4398752679886026e-05,
-      "loss": 0.2857,
+      "epoch": 1.2334666810826396,
+      "grad_norm": 0.21794357895851135,
+      "learning_rate": 4.4096939800476894e-05,
+      "loss": 0.4236,
       "step": 34225
     },
     {
-      "epoch": 1.2,
-      "learning_rate": 4.4396955596823594e-05,
-      "loss": 0.3026,
+      "epoch": 1.2336468807438643,
+      "grad_norm": 0.16437794268131256,
+      "learning_rate": 4.409505642027248e-05,
+      "loss": 0.4395,
       "step": 34230
     },
     {
-      "epoch": 1.2,
-      "learning_rate": 4.439515826190447e-05,
-      "loss": 0.2888,
+      "epoch": 1.2338270804050888,
+      "grad_norm": 0.152201309800148,
+      "learning_rate": 4.409317277990161e-05,
+      "loss": 0.4321,
       "step": 34235
     },
     {
-      "epoch": 1.2,
-      "learning_rate": 4.4393360675151996e-05,
-      "loss": 0.3029,
+      "epoch": 1.2340072800663135,
+      "grad_norm": 0.1949222981929779,
+      "learning_rate": 4.409128887938997e-05,
+      "loss": 0.4265,
       "step": 34240
     },
     {
-      "epoch": 1.2,
-      "learning_rate": 4.439156283658952e-05,
-      "loss": 0.2839,
+      "epoch": 1.234187479727538,
+      "grad_norm": 0.18767017126083374,
+      "learning_rate": 4.408940471876321e-05,
+      "loss": 0.4111,
       "step": 34245
     },
     {
-      "epoch": 1.21,
-      "learning_rate": 4.438976474624038e-05,
-      "loss": 0.2773,
+      "epoch": 1.2343676793887628,
+      "grad_norm": 0.22580619156360626,
+      "learning_rate": 4.4087520298047003e-05,
+      "loss": 0.416,
       "step": 34250
     },
     {
-      "epoch": 1.21,
-      "learning_rate": 4.4387966404127926e-05,
-      "loss": 0.3019,
+      "epoch": 1.2345478790499873,
+      "grad_norm": 0.14515608549118042,
+      "learning_rate": 4.4085635617267026e-05,
+      "loss": 0.4118,
       "step": 34255
     },
     {
-      "epoch": 1.21,
-      "learning_rate": 4.4386167810275493e-05,
-      "loss": 0.3107,
+      "epoch": 1.234728078711212,
+      "grad_norm": 0.17137964069843292,
+      "learning_rate": 4.408375067644897e-05,
+      "loss": 0.3908,
       "step": 34260
     },
     {
-      "epoch": 1.21,
-      "learning_rate": 4.438436896470646e-05,
-      "loss": 0.2845,
+      "epoch": 1.2349082783724366,
+      "grad_norm": 0.17997115850448608,
+      "learning_rate": 4.40818654756185e-05,
+      "loss": 0.4645,
       "step": 34265
     },
     {
-      "epoch": 1.21,
-      "learning_rate": 4.4382569867444164e-05,
-      "loss": 0.2895,
+      "epoch": 1.2350884780336613,
+      "grad_norm": 0.16378553211688995,
+      "learning_rate": 4.40799800148013e-05,
+      "loss": 0.4137,
       "step": 34270
     },
     {
-      "epoch": 1.21,
-      "learning_rate": 4.438077051851197e-05,
-      "loss": 0.2912,
+      "epoch": 1.235268677694886,
+      "grad_norm": 0.1839524507522583,
+      "learning_rate": 4.407809429402308e-05,
+      "loss": 0.4448,
       "step": 34275
     },
     {
-      "epoch": 1.21,
-      "learning_rate": 4.437897091793325e-05,
-      "loss": 0.2887,
+      "epoch": 1.2354488773561105,
+      "grad_norm": 0.1332423835992813,
+      "learning_rate": 4.407620831330951e-05,
+      "loss": 0.4172,
       "step": 34280
     },
     {
-      "epoch": 1.21,
-      "learning_rate": 4.437717106573136e-05,
-      "loss": 0.2649,
+      "epoch": 1.2356290770173353,
+      "grad_norm": 0.18333803117275238,
+      "learning_rate": 4.407432207268629e-05,
+      "loss": 0.4224,
       "step": 34285
     },
     {
-      "epoch": 1.21,
-      "learning_rate": 4.437537096192967e-05,
-      "loss": 0.2903,
+      "epoch": 1.2358092766785598,
+      "grad_norm": 0.2039324939250946,
+      "learning_rate": 4.4072435572179136e-05,
+      "loss": 0.4289,
       "step": 34290
     },
     {
-      "epoch": 1.21,
-      "learning_rate": 4.437357060655156e-05,
-      "loss": 0.2955,
+      "epoch": 1.2359894763397845,
+      "grad_norm": 0.17370297014713287,
+      "learning_rate": 4.407054881181373e-05,
+      "loss": 0.4226,
       "step": 34295
     },
     {
-      "epoch": 1.21,
-      "learning_rate": 4.4371769999620406e-05,
-      "loss": 0.2914,
+      "epoch": 1.236169676001009,
+      "grad_norm": 0.16956277191638947,
+      "learning_rate": 4.40686617916158e-05,
+      "loss": 0.4354,
       "step": 34300
     },
     {
-      "epoch": 1.21,
-      "learning_rate": 4.4369969141159585e-05,
-      "loss": 0.2676,
+      "epoch": 1.2363498756622338,
+      "grad_norm": 0.16114924848079681,
+      "learning_rate": 4.406677451161103e-05,
+      "loss": 0.4433,
       "step": 34305
     },
     {
-      "epoch": 1.21,
-      "learning_rate": 4.436816803119248e-05,
-      "loss": 0.2737,
+      "epoch": 1.2365300753234585,
+      "grad_norm": 0.15897580981254578,
+      "learning_rate": 4.406488697182516e-05,
+      "loss": 0.3884,
       "step": 34310
     },
     {
-      "epoch": 1.21,
-      "learning_rate": 4.4366366669742476e-05,
-      "loss": 0.3046,
+      "epoch": 1.236710274984683,
+      "grad_norm": 0.17210477590560913,
+      "learning_rate": 4.406299917228389e-05,
+      "loss": 0.4553,
       "step": 34315
     },
     {
-      "epoch": 1.21,
-      "learning_rate": 4.4364565056832966e-05,
-      "loss": 0.3038,
+      "epoch": 1.2368904746459077,
+      "grad_norm": 0.17768850922584534,
+      "learning_rate": 4.406111111301295e-05,
+      "loss": 0.4372,
       "step": 34320
     },
     {
-      "epoch": 1.21,
-      "learning_rate": 4.4362763192487336e-05,
-      "loss": 0.2929,
+      "epoch": 1.2370706743071322,
+      "grad_norm": 0.20289890468120575,
+      "learning_rate": 4.405922279403807e-05,
+      "loss": 0.4266,
       "step": 34325
     },
     {
-      "epoch": 1.21,
-      "learning_rate": 4.436096107672899e-05,
-      "loss": 0.2979,
+      "epoch": 1.237250873968357,
+      "grad_norm": 0.18634875118732452,
+      "learning_rate": 4.405733421538496e-05,
+      "loss": 0.39,
       "step": 34330
     },
     {
-      "epoch": 1.21,
-      "learning_rate": 4.435915870958132e-05,
-      "loss": 0.3026,
+      "epoch": 1.2374310736295815,
+      "grad_norm": 0.16916252672672272,
+      "learning_rate": 4.4055445377079364e-05,
+      "loss": 0.3812,
       "step": 34335
     },
     {
-      "epoch": 1.21,
-      "learning_rate": 4.4357356091067734e-05,
-      "loss": 0.3056,
+      "epoch": 1.2376112732908062,
+      "grad_norm": 0.17390431463718414,
+      "learning_rate": 4.405355627914701e-05,
+      "loss": 0.4682,
       "step": 34340
     },
     {
-      "epoch": 1.21,
-      "learning_rate": 4.435555322121164e-05,
-      "loss": 0.3093,
+      "epoch": 1.237791472952031,
+      "grad_norm": 0.17190834879875183,
+      "learning_rate": 4.405166692161365e-05,
+      "loss": 0.4097,
       "step": 34345
     },
     {
-      "epoch": 1.21,
-      "learning_rate": 4.435375010003644e-05,
-      "loss": 0.2713,
+      "epoch": 1.2379716726132555,
+      "grad_norm": 0.15604445338249207,
+      "learning_rate": 4.4049777304505e-05,
+      "loss": 0.4193,
       "step": 34350
     },
     {
-      "epoch": 1.21,
-      "learning_rate": 4.435194672756555e-05,
-      "loss": 0.295,
+      "epoch": 1.2381518722744802,
+      "grad_norm": 0.19317157566547394,
+      "learning_rate": 4.404788742784683e-05,
+      "loss": 0.4087,
       "step": 34355
     },
     {
-      "epoch": 1.21,
-      "learning_rate": 4.4350143103822385e-05,
-      "loss": 0.2824,
+      "epoch": 1.2383320719357047,
+      "grad_norm": 0.15867052972316742,
+      "learning_rate": 4.404599729166489e-05,
+      "loss": 0.4123,
       "step": 34360
     },
     {
-      "epoch": 1.21,
-      "learning_rate": 4.434833922883036e-05,
-      "loss": 0.2898,
+      "epoch": 1.2385122715969294,
+      "grad_norm": 0.18117307126522064,
+      "learning_rate": 4.404410689598491e-05,
+      "loss": 0.4088,
       "step": 34365
     },
     {
-      "epoch": 1.21,
-      "learning_rate": 4.434653510261291e-05,
-      "loss": 0.2953,
+      "epoch": 1.238692471258154,
+      "grad_norm": 0.20402644574642181,
+      "learning_rate": 4.404221624083267e-05,
+      "loss": 0.4284,
       "step": 34370
     },
     {
-      "epoch": 1.21,
-      "learning_rate": 4.434473072519344e-05,
-      "loss": 0.3097,
+      "epoch": 1.2388726709193787,
+      "grad_norm": 0.20701655745506287,
+      "learning_rate": 4.4040325326233914e-05,
+      "loss": 0.386,
       "step": 34375
     },
     {
-      "epoch": 1.21,
-      "learning_rate": 4.4342926096595396e-05,
-      "loss": 0.301,
+      "epoch": 1.2390528705806032,
+      "grad_norm": 0.20911046862602234,
+      "learning_rate": 4.403843415221442e-05,
+      "loss": 0.3965,
       "step": 34380
     },
     {
-      "epoch": 1.21,
-      "learning_rate": 4.434112121684221e-05,
-      "loss": 0.2811,
+      "epoch": 1.239233070241828,
+      "grad_norm": 0.22463898360729218,
+      "learning_rate": 4.4036542718799944e-05,
+      "loss": 0.4292,
       "step": 34385
     },
     {
-      "epoch": 1.21,
-      "learning_rate": 4.4339316085957305e-05,
-      "loss": 0.2814,
+      "epoch": 1.2394132699030527,
+      "grad_norm": 0.1606704443693161,
+      "learning_rate": 4.403465102601626e-05,
+      "loss": 0.4427,
       "step": 34390
     },
     {
-      "epoch": 1.21,
-      "learning_rate": 4.433751070396412e-05,
-      "loss": 0.2852,
+      "epoch": 1.2395934695642772,
+      "grad_norm": 0.18783807754516602,
+      "learning_rate": 4.4032759073889134e-05,
+      "loss": 0.4545,
       "step": 34395
     },
     {
-      "epoch": 1.21,
-      "learning_rate": 4.433570507088611e-05,
-      "loss": 0.283,
+      "epoch": 1.239773669225502,
+      "grad_norm": 0.20419709384441376,
+      "learning_rate": 4.403086686244435e-05,
+      "loss": 0.4332,
       "step": 34400
     },
     {
-      "epoch": 1.21,
-      "learning_rate": 4.433389918674671e-05,
-      "loss": 0.2712,
+      "epoch": 1.2399538688867264,
+      "grad_norm": 0.16795575618743896,
+      "learning_rate": 4.40289743917077e-05,
+      "loss": 0.4359,
       "step": 34405
     },
     {
-      "epoch": 1.21,
-      "learning_rate": 4.4332093051569374e-05,
-      "loss": 0.3063,
+      "epoch": 1.2401340685479512,
+      "grad_norm": 0.157429039478302,
+      "learning_rate": 4.402708166170495e-05,
+      "loss": 0.4176,
       "step": 34410
     },
     {
-      "epoch": 1.21,
-      "learning_rate": 4.433028666537755e-05,
-      "loss": 0.2805,
+      "epoch": 1.2403142682091757,
+      "grad_norm": 0.1853724867105484,
+      "learning_rate": 4.4025188672461903e-05,
+      "loss": 0.4474,
       "step": 34415
     },
     {
-      "epoch": 1.21,
-      "learning_rate": 4.4328480028194684e-05,
-      "loss": 0.3026,
+      "epoch": 1.2404944678704004,
+      "grad_norm": 0.1714290827512741,
+      "learning_rate": 4.402329542400434e-05,
+      "loss": 0.4237,
       "step": 34420
     },
     {
-      "epoch": 1.21,
-      "learning_rate": 4.432667314004425e-05,
-      "loss": 0.2853,
+      "epoch": 1.240674667531625,
+      "grad_norm": 0.2036142498254776,
+      "learning_rate": 4.402140191635806e-05,
+      "loss": 0.4049,
       "step": 34425
     },
     {
-      "epoch": 1.21,
-      "learning_rate": 4.432486600094971e-05,
-      "loss": 0.3051,
+      "epoch": 1.2408548671928497,
+      "grad_norm": 0.1990082859992981,
+      "learning_rate": 4.401950814954886e-05,
+      "loss": 0.4252,
       "step": 34430
     },
     {
-      "epoch": 1.21,
-      "learning_rate": 4.432305861093451e-05,
-      "loss": 0.2682,
+      "epoch": 1.2410350668540744,
+      "grad_norm": 0.17865248024463654,
+      "learning_rate": 4.4017614123602546e-05,
+      "loss": 0.4072,
       "step": 34435
     },
     {
-      "epoch": 1.21,
-      "learning_rate": 4.4321250970022126e-05,
-      "loss": 0.2969,
+      "epoch": 1.241215266515299,
+      "grad_norm": 0.1914636194705963,
+      "learning_rate": 4.401571983854492e-05,
+      "loss": 0.4452,
       "step": 34440
     },
     {
-      "epoch": 1.21,
-      "learning_rate": 4.431944307823604e-05,
-      "loss": 0.2952,
+      "epoch": 1.2413954661765236,
+      "grad_norm": 0.1847231239080429,
+      "learning_rate": 4.401382529440179e-05,
+      "loss": 0.3943,
       "step": 34445
     },
     {
-      "epoch": 1.21,
-      "learning_rate": 4.431763493559971e-05,
-      "loss": 0.2963,
+      "epoch": 1.2415756658377481,
+      "grad_norm": 0.20364885032176971,
+      "learning_rate": 4.401193049119898e-05,
+      "loss": 0.4116,
       "step": 34450
     },
     {
-      "epoch": 1.21,
-      "learning_rate": 4.431582654213663e-05,
-      "loss": 0.296,
+      "epoch": 1.2417558654989729,
+      "grad_norm": 0.1959732621908188,
+      "learning_rate": 4.4010035428962295e-05,
+      "loss": 0.4072,
       "step": 34455
     },
     {
-      "epoch": 1.21,
-      "learning_rate": 4.431401789787026e-05,
-      "loss": 0.2775,
+      "epoch": 1.2419360651601976,
+      "grad_norm": 0.18121816217899323,
+      "learning_rate": 4.400814010771755e-05,
+      "loss": 0.4325,
       "step": 34460
     },
     {
-      "epoch": 1.21,
-      "learning_rate": 4.43122090028241e-05,
-      "loss": 0.2827,
+      "epoch": 1.2421162648214221,
+      "grad_norm": 0.23917900025844574,
+      "learning_rate": 4.400624452749058e-05,
+      "loss": 0.4233,
       "step": 34465
     },
     {
-      "epoch": 1.21,
-      "learning_rate": 4.431039985702164e-05,
-      "loss": 0.3046,
+      "epoch": 1.2422964644826469,
+      "grad_norm": 0.21127377450466156,
+      "learning_rate": 4.400434868830721e-05,
+      "loss": 0.4266,
       "step": 34470
     },
     {
-      "epoch": 1.21,
-      "learning_rate": 4.430859046048636e-05,
-      "loss": 0.2818,
+      "epoch": 1.2424766641438714,
+      "grad_norm": 0.21649089455604553,
+      "learning_rate": 4.4002452590193265e-05,
+      "loss": 0.4478,
       "step": 34475
     },
     {
-      "epoch": 1.21,
-      "learning_rate": 4.430678081324175e-05,
-      "loss": 0.2823,
+      "epoch": 1.242656863805096,
+      "grad_norm": 0.18373538553714752,
+      "learning_rate": 4.400055623317459e-05,
+      "loss": 0.3895,
       "step": 34480
     },
     {
-      "epoch": 1.21,
-      "learning_rate": 4.4304970915311326e-05,
-      "loss": 0.2867,
+      "epoch": 1.2428370634663206,
+      "grad_norm": 0.1825505495071411,
+      "learning_rate": 4.399865961727701e-05,
+      "loss": 0.3957,
       "step": 34485
     },
     {
-      "epoch": 1.21,
-      "learning_rate": 4.4303160766718574e-05,
-      "loss": 0.3043,
+      "epoch": 1.2430172631275453,
+      "grad_norm": 0.15860360860824585,
+      "learning_rate": 4.399676274252637e-05,
+      "loss": 0.4106,
       "step": 34490
     },
     {
-      "epoch": 1.21,
-      "learning_rate": 4.4301350367487005e-05,
-      "loss": 0.3266,
+      "epoch": 1.2431974627887699,
+      "grad_norm": 0.14994730055332184,
+      "learning_rate": 4.399486560894852e-05,
+      "loss": 0.3668,
       "step": 34495
     },
     {
-      "epoch": 1.21,
-      "learning_rate": 4.429953971764012e-05,
-      "loss": 0.287,
+      "epoch": 1.2433776624499946,
+      "grad_norm": 0.18823125958442688,
+      "learning_rate": 4.39929682165693e-05,
+      "loss": 0.419,
       "step": 34500
     },
     {
-      "epoch": 1.21,
-      "eval_loss": 0.2872275710105896,
-      "eval_runtime": 10.5666,
-      "eval_samples_per_second": 9.464,
-      "eval_steps_per_second": 9.464,
+      "epoch": 1.2433776624499946,
+      "eval_loss": 0.45045050978660583,
+      "eval_runtime": 3.5444,
+      "eval_samples_per_second": 28.213,
+      "eval_steps_per_second": 7.053,
       "step": 34500
     },
     {
-      "epoch": 1.21,
-      "learning_rate": 4.429772881720142e-05,
-      "loss": 0.2876,
+      "epoch": 1.2435578621112193,
+      "grad_norm": 0.20211850106716156,
+      "learning_rate": 4.399107056541456e-05,
+      "loss": 0.3954,
       "step": 34505
     },
     {
-      "epoch": 1.21,
-      "learning_rate": 4.429591766619444e-05,
-      "loss": 0.2798,
+      "epoch": 1.2437380617724438,
+      "grad_norm": 0.19048206508159637,
+      "learning_rate": 4.398917265551017e-05,
+      "loss": 0.4312,
       "step": 34510
     },
     {
-      "epoch": 1.21,
-      "learning_rate": 4.429410626464268e-05,
-      "loss": 0.285,
+      "epoch": 1.2439182614336686,
+      "grad_norm": 0.17858850955963135,
+      "learning_rate": 4.398727448688198e-05,
+      "loss": 0.4522,
       "step": 34515
     },
     {
-      "epoch": 1.21,
-      "learning_rate": 4.429229461256967e-05,
-      "loss": 0.2681,
+      "epoch": 1.244098461094893,
+      "grad_norm": 0.14461615681648254,
+      "learning_rate": 4.398537605955584e-05,
+      "loss": 0.4295,
       "step": 34520
     },
     {
-      "epoch": 1.21,
-      "learning_rate": 4.429048270999892e-05,
-      "loss": 0.2791,
+      "epoch": 1.2442786607561178,
+      "grad_norm": 0.16220664978027344,
+      "learning_rate": 4.398347737355764e-05,
+      "loss": 0.4194,
       "step": 34525
     },
     {
-      "epoch": 1.21,
-      "learning_rate": 4.428867055695398e-05,
-      "loss": 0.3229,
+      "epoch": 1.2444588604173423,
+      "grad_norm": 0.17861701548099518,
+      "learning_rate": 4.398157842891323e-05,
+      "loss": 0.4393,
       "step": 34530
     },
     {
-      "epoch": 1.22,
-      "learning_rate": 4.4286858153458344e-05,
-      "loss": 0.3223,
+      "epoch": 1.244639060078567,
+      "grad_norm": 0.2174644023180008,
+      "learning_rate": 4.3979679225648484e-05,
+      "loss": 0.4721,
       "step": 34535
     },
     {
-      "epoch": 1.22,
-      "learning_rate": 4.4285045499535584e-05,
-      "loss": 0.3156,
+      "epoch": 1.2448192597397916,
+      "grad_norm": 0.19384269416332245,
+      "learning_rate": 4.397777976378929e-05,
+      "loss": 0.4112,
       "step": 34540
     },
     {
-      "epoch": 1.22,
-      "learning_rate": 4.428323259520921e-05,
-      "loss": 0.2721,
+      "epoch": 1.2449994594010163,
+      "grad_norm": 0.16836018860340118,
+      "learning_rate": 4.397588004336152e-05,
+      "loss": 0.4547,
       "step": 34545
     },
     {
-      "epoch": 1.22,
-      "learning_rate": 4.428141944050277e-05,
-      "loss": 0.2948,
+      "epoch": 1.245179659062241,
+      "grad_norm": 0.163166344165802,
+      "learning_rate": 4.397398006439105e-05,
+      "loss": 0.4291,
       "step": 34550
     },
     {
-      "epoch": 1.22,
-      "learning_rate": 4.42796060354398e-05,
-      "loss": 0.2897,
+      "epoch": 1.2453598587234656,
+      "grad_norm": 0.17786313593387604,
+      "learning_rate": 4.397207982690378e-05,
+      "loss": 0.4256,
       "step": 34555
     },
     {
-      "epoch": 1.22,
-      "learning_rate": 4.427779238004386e-05,
-      "loss": 0.2914,
+      "epoch": 1.2455400583846903,
+      "grad_norm": 0.1747133880853653,
+      "learning_rate": 4.39701793309256e-05,
+      "loss": 0.4184,
       "step": 34560
     },
     {
-      "epoch": 1.22,
-      "learning_rate": 4.4275978474338485e-05,
-      "loss": 0.2896,
+      "epoch": 1.2457202580459148,
+      "grad_norm": 0.1763535887002945,
+      "learning_rate": 4.3968278576482394e-05,
+      "loss": 0.4077,
       "step": 34565
     },
     {
-      "epoch": 1.22,
-      "learning_rate": 4.4274164318347235e-05,
-      "loss": 0.2832,
+      "epoch": 1.2459004577071395,
+      "grad_norm": 0.22853651642799377,
+      "learning_rate": 4.396637756360007e-05,
+      "loss": 0.4367,
       "step": 34570
     },
     {
-      "epoch": 1.22,
-      "learning_rate": 4.427234991209366e-05,
-      "loss": 0.3122,
+      "epoch": 1.2460806573683643,
+      "grad_norm": 0.21163353323936462,
+      "learning_rate": 4.396447629230452e-05,
+      "loss": 0.4155,
       "step": 34575
     },
     {
-      "epoch": 1.22,
-      "learning_rate": 4.4270535255601324e-05,
-      "loss": 0.2831,
+      "epoch": 1.2462608570295888,
+      "grad_norm": 0.16000016033649445,
+      "learning_rate": 4.396257476262165e-05,
+      "loss": 0.3771,
       "step": 34580
     },
     {
-      "epoch": 1.22,
-      "learning_rate": 4.4268720348893786e-05,
-      "loss": 0.2953,
+      "epoch": 1.2464410566908135,
+      "grad_norm": 0.23140566051006317,
+      "learning_rate": 4.396067297457738e-05,
+      "loss": 0.4196,
       "step": 34585
     },
     {
-      "epoch": 1.22,
-      "learning_rate": 4.4266905191994615e-05,
-      "loss": 0.2978,
+      "epoch": 1.246621256352038,
+      "grad_norm": 0.16207090020179749,
+      "learning_rate": 4.3958770928197604e-05,
+      "loss": 0.4025,
       "step": 34590
     },
     {
-      "epoch": 1.22,
-      "learning_rate": 4.426508978492738e-05,
-      "loss": 0.2993,
+      "epoch": 1.2468014560132628,
+      "grad_norm": 0.18664728105068207,
+      "learning_rate": 4.395686862350824e-05,
+      "loss": 0.4154,
       "step": 34595
     },
     {
-      "epoch": 1.22,
-      "learning_rate": 4.426327412771565e-05,
-      "loss": 0.3028,
+      "epoch": 1.2469816556744873,
+      "grad_norm": 0.1926645040512085,
+      "learning_rate": 4.395496606053522e-05,
+      "loss": 0.4093,
       "step": 34600
     },
     {
-      "epoch": 1.22,
-      "learning_rate": 4.4261458220383e-05,
-      "loss": 0.2977,
+      "epoch": 1.247161855335712,
+      "grad_norm": 0.16075952351093292,
+      "learning_rate": 4.395306323930445e-05,
+      "loss": 0.4158,
       "step": 34605
     },
     {
-      "epoch": 1.22,
-      "learning_rate": 4.425964206295301e-05,
-      "loss": 0.2806,
+      "epoch": 1.2473420549969365,
+      "grad_norm": 0.1828758716583252,
+      "learning_rate": 4.3951160159841864e-05,
+      "loss": 0.3779,
       "step": 34610
     },
     {
-      "epoch": 1.22,
-      "learning_rate": 4.4257825655449257e-05,
-      "loss": 0.2709,
+      "epoch": 1.2475222546581612,
+      "grad_norm": 0.2084207683801651,
+      "learning_rate": 4.394925682217339e-05,
+      "loss": 0.4011,
       "step": 34615
     },
     {
-      "epoch": 1.22,
-      "learning_rate": 4.425600899789533e-05,
-      "loss": 0.2556,
+      "epoch": 1.247702454319386,
+      "grad_norm": 0.17692120373249054,
+      "learning_rate": 4.3947353226324964e-05,
+      "loss": 0.4318,
       "step": 34620
     },
     {
-      "epoch": 1.22,
-      "learning_rate": 4.425419209031482e-05,
-      "loss": 0.2851,
+      "epoch": 1.2478826539806105,
+      "grad_norm": 0.16040922701358795,
+      "learning_rate": 4.394544937232252e-05,
+      "loss": 0.396,
       "step": 34625
     },
     {
-      "epoch": 1.22,
-      "learning_rate": 4.42523749327313e-05,
-      "loss": 0.2911,
+      "epoch": 1.2480628536418352,
+      "grad_norm": 0.19275468587875366,
+      "learning_rate": 4.3943545260192e-05,
+      "loss": 0.4325,
       "step": 34630
     },
     {
-      "epoch": 1.22,
-      "learning_rate": 4.425055752516839e-05,
-      "loss": 0.2835,
+      "epoch": 1.2482430533030597,
+      "grad_norm": 0.1728275716304779,
+      "learning_rate": 4.394164088995933e-05,
+      "loss": 0.4156,
       "step": 34635
     },
     {
-      "epoch": 1.22,
-      "learning_rate": 4.424873986764968e-05,
-      "loss": 0.2993,
+      "epoch": 1.2484232529642845,
+      "grad_norm": 0.2221703678369522,
+      "learning_rate": 4.393973626165048e-05,
+      "loss": 0.3882,
       "step": 34640
     },
     {
-      "epoch": 1.22,
-      "learning_rate": 4.424692196019877e-05,
-      "loss": 0.289,
+      "epoch": 1.248603452625509,
+      "grad_norm": 0.18804728984832764,
+      "learning_rate": 4.393783137529139e-05,
+      "loss": 0.4215,
       "step": 34645
     },
     {
-      "epoch": 1.22,
-      "learning_rate": 4.4245103802839256e-05,
-      "loss": 0.2912,
+      "epoch": 1.2487836522867337,
+      "grad_norm": 0.15801158547401428,
+      "learning_rate": 4.393592623090801e-05,
+      "loss": 0.364,
       "step": 34650
     },
     {
-      "epoch": 1.22,
-      "learning_rate": 4.4243285395594754e-05,
-      "loss": 0.2923,
+      "epoch": 1.2489638519479582,
+      "grad_norm": 0.1544542908668518,
+      "learning_rate": 4.39340208285263e-05,
+      "loss": 0.4171,
       "step": 34655
     },
     {
-      "epoch": 1.22,
-      "learning_rate": 4.424146673848887e-05,
-      "loss": 0.291,
+      "epoch": 1.249144051609183,
+      "grad_norm": 0.19875742495059967,
+      "learning_rate": 4.3932115168172225e-05,
+      "loss": 0.4258,
       "step": 34660
     },
     {
-      "epoch": 1.22,
-      "learning_rate": 4.4239647831545225e-05,
-      "loss": 0.3019,
+      "epoch": 1.2493242512704077,
+      "grad_norm": 0.16528360545635223,
+      "learning_rate": 4.3930209249871744e-05,
+      "loss": 0.4562,
       "step": 34665
     },
     {
-      "epoch": 1.22,
-      "learning_rate": 4.423782867478743e-05,
-      "loss": 0.2934,
+      "epoch": 1.2495044509316322,
+      "grad_norm": 0.18268528580665588,
+      "learning_rate": 4.3928303073650835e-05,
+      "loss": 0.407,
       "step": 34670
     },
     {
-      "epoch": 1.22,
-      "learning_rate": 4.423600926823911e-05,
-      "loss": 0.2817,
+      "epoch": 1.249684650592857,
+      "grad_norm": 0.1825907826423645,
+      "learning_rate": 4.392639663953545e-05,
+      "loss": 0.4098,
       "step": 34675
     },
     {
-      "epoch": 1.22,
-      "learning_rate": 4.423418961192389e-05,
-      "loss": 0.3051,
+      "epoch": 1.2498648502540814,
+      "grad_norm": 0.17392845451831818,
+      "learning_rate": 4.3924489947551586e-05,
+      "loss": 0.4377,
       "step": 34680
     },
     {
-      "epoch": 1.22,
-      "learning_rate": 4.423236970586538e-05,
-      "loss": 0.2773,
+      "epoch": 1.2500450499153062,
+      "grad_norm": 0.1909123808145523,
+      "learning_rate": 4.39225829977252e-05,
+      "loss": 0.4298,
       "step": 34685
     },
     {
-      "epoch": 1.22,
-      "learning_rate": 4.423054955008724e-05,
-      "loss": 0.2957,
+      "epoch": 1.250225249576531,
+      "grad_norm": 0.13296571373939514,
+      "learning_rate": 4.392067579008229e-05,
+      "loss": 0.4111,
       "step": 34690
     },
     {
-      "epoch": 1.22,
-      "learning_rate": 4.422872914461308e-05,
-      "loss": 0.2919,
+      "epoch": 1.2504054492377554,
+      "grad_norm": 0.21893556416034698,
+      "learning_rate": 4.391876832464883e-05,
+      "loss": 0.4422,
       "step": 34695
     },
     {
-      "epoch": 1.22,
-      "learning_rate": 4.4226908489466544e-05,
-      "loss": 0.2951,
+      "epoch": 1.25058564889898,
+      "grad_norm": 0.18834765255451202,
+      "learning_rate": 4.3916860601450825e-05,
+      "loss": 0.4363,
       "step": 34700
     },
     {
-      "epoch": 1.22,
-      "learning_rate": 4.422508758467128e-05,
-      "loss": 0.278,
+      "epoch": 1.2507658485602047,
+      "grad_norm": 0.18092414736747742,
+      "learning_rate": 4.391495262051425e-05,
+      "loss": 0.4227,
       "step": 34705
     },
     {
-      "epoch": 1.22,
-      "learning_rate": 4.4223266430250915e-05,
-      "loss": 0.2756,
+      "epoch": 1.2509460482214294,
+      "grad_norm": 0.18126097321510315,
+      "learning_rate": 4.391304438186511e-05,
+      "loss": 0.3999,
       "step": 34710
     },
     {
-      "epoch": 1.22,
-      "learning_rate": 4.42214450262291e-05,
-      "loss": 0.2845,
+      "epoch": 1.251126247882654,
+      "grad_norm": 0.19100238382816315,
+      "learning_rate": 4.39111358855294e-05,
+      "loss": 0.3802,
       "step": 34715
     },
     {
-      "epoch": 1.22,
-      "learning_rate": 4.42196233726295e-05,
-      "loss": 0.2854,
+      "epoch": 1.2513064475438787,
+      "grad_norm": 0.18125472962856293,
+      "learning_rate": 4.390922713153312e-05,
+      "loss": 0.4184,
       "step": 34720
     },
     {
-      "epoch": 1.22,
-      "learning_rate": 4.421780146947575e-05,
-      "loss": 0.287,
+      "epoch": 1.2514866472051032,
+      "grad_norm": 0.2104109823703766,
+      "learning_rate": 4.3907318119902286e-05,
+      "loss": 0.3942,
       "step": 34725
     },
     {
-      "epoch": 1.22,
-      "learning_rate": 4.4215979316791516e-05,
-      "loss": 0.2951,
+      "epoch": 1.251666846866328,
+      "grad_norm": 0.18633076548576355,
+      "learning_rate": 4.390540885066291e-05,
+      "loss": 0.409,
       "step": 34730
     },
     {
-      "epoch": 1.22,
-      "learning_rate": 4.4214156914600454e-05,
-      "loss": 0.2786,
+      "epoch": 1.2518470465275526,
+      "grad_norm": 0.18093355000019073,
+      "learning_rate": 4.3903499323840985e-05,
+      "loss": 0.411,
       "step": 34735
     },
     {
-      "epoch": 1.22,
-      "learning_rate": 4.4212334262926226e-05,
-      "loss": 0.3042,
+      "epoch": 1.2520272461887771,
+      "grad_norm": 0.18245679140090942,
+      "learning_rate": 4.390158953946255e-05,
+      "loss": 0.4236,
       "step": 34740
     },
     {
-      "epoch": 1.22,
-      "learning_rate": 4.42105113617925e-05,
-      "loss": 0.2986,
+      "epoch": 1.2522074458500019,
+      "grad_norm": 0.16724418103694916,
+      "learning_rate": 4.3899679497553616e-05,
+      "loss": 0.4244,
       "step": 34745
     },
     {
-      "epoch": 1.22,
-      "learning_rate": 4.420868821122295e-05,
-      "loss": 0.2869,
+      "epoch": 1.2523876455112264,
+      "grad_norm": 0.1758909821510315,
+      "learning_rate": 4.3897769198140204e-05,
+      "loss": 0.438,
       "step": 34750
     },
     {
-      "epoch": 1.22,
-      "learning_rate": 4.420686481124123e-05,
-      "loss": 0.2745,
+      "epoch": 1.2525678451724511,
+      "grad_norm": 0.17863698303699493,
+      "learning_rate": 4.389585864124835e-05,
+      "loss": 0.4163,
       "step": 34755
     },
     {
-      "epoch": 1.22,
-      "learning_rate": 4.420504116187103e-05,
-      "loss": 0.2817,
+      "epoch": 1.2527480448336756,
+      "grad_norm": 0.19687694311141968,
+      "learning_rate": 4.389394782690408e-05,
+      "loss": 0.4004,
       "step": 34760
     },
     {
-      "epoch": 1.22,
-      "learning_rate": 4.4203217263136034e-05,
-      "loss": 0.2904,
+      "epoch": 1.2529282444949004,
+      "grad_norm": 0.17366833984851837,
+      "learning_rate": 4.389203675513343e-05,
+      "loss": 0.4307,
       "step": 34765
     },
     {
-      "epoch": 1.22,
-      "learning_rate": 4.420139311505991e-05,
-      "loss": 0.3057,
+      "epoch": 1.2531084441561249,
+      "grad_norm": 0.14887477457523346,
+      "learning_rate": 4.389012542596244e-05,
+      "loss": 0.435,
       "step": 34770
     },
     {
-      "epoch": 1.22,
-      "learning_rate": 4.419956871766636e-05,
-      "loss": 0.2846,
+      "epoch": 1.2532886438173496,
+      "grad_norm": 0.17449168860912323,
+      "learning_rate": 4.388821383941714e-05,
+      "loss": 0.4406,
       "step": 34775
     },
     {
-      "epoch": 1.22,
-      "learning_rate": 4.4197744070979056e-05,
-      "loss": 0.2672,
+      "epoch": 1.2534688434785743,
+      "grad_norm": 0.1995227187871933,
+      "learning_rate": 4.388630199552358e-05,
+      "loss": 0.4066,
       "step": 34780
     },
     {
-      "epoch": 1.22,
-      "learning_rate": 4.41959191750217e-05,
-      "loss": 0.3128,
+      "epoch": 1.2536490431397989,
+      "grad_norm": 0.19903253018856049,
+      "learning_rate": 4.388438989430782e-05,
+      "loss": 0.459,
       "step": 34785
     },
     {
-      "epoch": 1.22,
-      "learning_rate": 4.419409402981799e-05,
-      "loss": 0.2899,
+      "epoch": 1.2538292428010236,
+      "grad_norm": 0.23275376856327057,
+      "learning_rate": 4.3882477535795904e-05,
+      "loss": 0.4503,
       "step": 34790
     },
     {
-      "epoch": 1.22,
-      "learning_rate": 4.41922686353916e-05,
-      "loss": 0.3013,
+      "epoch": 1.254009442462248,
+      "grad_norm": 0.17660613358020782,
+      "learning_rate": 4.3880564920013885e-05,
+      "loss": 0.3901,
       "step": 34795
     },
     {
-      "epoch": 1.22,
-      "learning_rate": 4.4190442991766264e-05,
-      "loss": 0.2865,
+      "epoch": 1.2541896421234728,
+      "grad_norm": 0.1641213297843933,
+      "learning_rate": 4.3878652046987824e-05,
+      "loss": 0.4275,
       "step": 34800
     },
     {
-      "epoch": 1.22,
-      "learning_rate": 4.4188617098965675e-05,
-      "loss": 0.2979,
+      "epoch": 1.2543698417846976,
+      "grad_norm": 0.19240440428256989,
+      "learning_rate": 4.387673891674379e-05,
+      "loss": 0.4089,
       "step": 34805
     },
     {
-      "epoch": 1.22,
-      "learning_rate": 4.4186790957013536e-05,
-      "loss": 0.2925,
+      "epoch": 1.254550041445922,
+      "grad_norm": 0.18231160938739777,
+      "learning_rate": 4.387482552930783e-05,
+      "loss": 0.4007,
       "step": 34810
     },
     {
-      "epoch": 1.22,
-      "learning_rate": 4.418496456593356e-05,
-      "loss": 0.2984,
+      "epoch": 1.2547302411071466,
+      "grad_norm": 0.15601201355457306,
+      "learning_rate": 4.387291188470603e-05,
+      "loss": 0.4185,
       "step": 34815
     },
     {
-      "epoch": 1.23,
-      "learning_rate": 4.418313792574946e-05,
-      "loss": 0.2933,
+      "epoch": 1.2549104407683713,
+      "grad_norm": 0.1767847090959549,
+      "learning_rate": 4.387099798296447e-05,
+      "loss": 0.4317,
       "step": 34820
     },
     {
-      "epoch": 1.23,
-      "learning_rate": 4.4181311036484955e-05,
-      "loss": 0.2885,
+      "epoch": 1.255090640429596,
+      "grad_norm": 0.1789582371711731,
+      "learning_rate": 4.386908382410922e-05,
+      "loss": 0.4234,
       "step": 34825
     },
     {
-      "epoch": 1.23,
-      "learning_rate": 4.417948389816377e-05,
-      "loss": 0.3076,
+      "epoch": 1.2552708400908206,
+      "grad_norm": 0.1394336223602295,
+      "learning_rate": 4.3867169408166333e-05,
+      "loss": 0.3885,
       "step": 34830
     },
     {
-      "epoch": 1.23,
-      "learning_rate": 4.417765651080963e-05,
-      "loss": 0.271,
+      "epoch": 1.2554510397520453,
+      "grad_norm": 0.1963426172733307,
+      "learning_rate": 4.3865254735161934e-05,
+      "loss": 0.3865,
       "step": 34835
     },
     {
-      "epoch": 1.23,
-      "learning_rate": 4.417582887444625e-05,
-      "loss": 0.3028,
+      "epoch": 1.2556312394132698,
+      "grad_norm": 0.19364525377750397,
+      "learning_rate": 4.3863339805122086e-05,
+      "loss": 0.4117,
       "step": 34840
     },
     {
-      "epoch": 1.23,
-      "learning_rate": 4.4174000989097375e-05,
-      "loss": 0.3184,
+      "epoch": 1.2558114390744946,
+      "grad_norm": 0.19295062124729156,
+      "learning_rate": 4.386142461807288e-05,
+      "loss": 0.4395,
       "step": 34845
     },
     {
-      "epoch": 1.23,
-      "learning_rate": 4.417217285478673e-05,
-      "loss": 0.2935,
+      "epoch": 1.2559916387357193,
+      "grad_norm": 0.16617421805858612,
+      "learning_rate": 4.385950917404042e-05,
+      "loss": 0.3931,
       "step": 34850
     },
     {
-      "epoch": 1.23,
-      "learning_rate": 4.417034447153806e-05,
-      "loss": 0.2841,
+      "epoch": 1.2561718383969438,
+      "grad_norm": 0.16993173956871033,
+      "learning_rate": 4.3857593473050804e-05,
+      "loss": 0.4174,
       "step": 34855
     },
     {
-      "epoch": 1.23,
-      "learning_rate": 4.41685158393751e-05,
-      "loss": 0.2943,
+      "epoch": 1.2563520380581685,
+      "grad_norm": 0.1860906183719635,
+      "learning_rate": 4.3855677515130125e-05,
+      "loss": 0.4128,
       "step": 34860
     },
     {
-      "epoch": 1.23,
-      "learning_rate": 4.41666869583216e-05,
-      "loss": 0.2861,
+      "epoch": 1.256532237719393,
+      "grad_norm": 0.17956295609474182,
+      "learning_rate": 4.385376130030448e-05,
+      "loss": 0.4174,
       "step": 34865
     },
     {
-      "epoch": 1.23,
-      "learning_rate": 4.416485782840129e-05,
-      "loss": 0.2744,
+      "epoch": 1.2567124373806178,
+      "grad_norm": 0.17807215452194214,
+      "learning_rate": 4.38518448286e-05,
+      "loss": 0.4009,
       "step": 34870
     },
     {
-      "epoch": 1.23,
-      "learning_rate": 4.4163028449637933e-05,
-      "loss": 0.2977,
+      "epoch": 1.2568926370418423,
+      "grad_norm": 0.15387162566184998,
+      "learning_rate": 4.384992810004278e-05,
+      "loss": 0.4004,
       "step": 34875
     },
     {
-      "epoch": 1.23,
-      "learning_rate": 4.416119882205528e-05,
-      "loss": 0.3126,
+      "epoch": 1.257072836703067,
+      "grad_norm": 0.1546095311641693,
+      "learning_rate": 4.3848011114658934e-05,
+      "loss": 0.3967,
       "step": 34880
     },
     {
-      "epoch": 1.23,
-      "learning_rate": 4.41593689456771e-05,
-      "loss": 0.2921,
+      "epoch": 1.2572530363642915,
+      "grad_norm": 0.18044285476207733,
+      "learning_rate": 4.384609387247459e-05,
+      "loss": 0.4163,
       "step": 34885
     },
     {
-      "epoch": 1.23,
-      "learning_rate": 4.415753882052713e-05,
-      "loss": 0.3003,
+      "epoch": 1.2574332360255163,
+      "grad_norm": 0.2026679515838623,
+      "learning_rate": 4.384417637351587e-05,
+      "loss": 0.4258,
       "step": 34890
     },
     {
-      "epoch": 1.23,
-      "learning_rate": 4.415570844662914e-05,
-      "loss": 0.27,
+      "epoch": 1.257613435686741,
+      "grad_norm": 0.15723782777786255,
+      "learning_rate": 4.3842258617808895e-05,
+      "loss": 0.395,
       "step": 34895
     },
     {
-      "epoch": 1.23,
-      "learning_rate": 4.415387782400691e-05,
-      "loss": 0.2767,
+      "epoch": 1.2577936353479655,
+      "grad_norm": 0.19395868480205536,
+      "learning_rate": 4.384034060537979e-05,
+      "loss": 0.425,
       "step": 34900
     },
     {
-      "epoch": 1.23,
-      "learning_rate": 4.415204695268419e-05,
-      "loss": 0.2865,
+      "epoch": 1.2579738350091902,
+      "grad_norm": 0.1924101859331131,
+      "learning_rate": 4.383842233625469e-05,
+      "loss": 0.432,
       "step": 34905
     },
     {
-      "epoch": 1.23,
-      "learning_rate": 4.415021583268477e-05,
-      "loss": 0.3064,
+      "epoch": 1.2581540346704148,
+      "grad_norm": 0.13849791884422302,
+      "learning_rate": 4.383650381045974e-05,
+      "loss": 0.4305,
       "step": 34910
     },
     {
-      "epoch": 1.23,
-      "learning_rate": 4.414838446403242e-05,
-      "loss": 0.2653,
+      "epoch": 1.2583342343316395,
+      "grad_norm": 0.16291098296642303,
+      "learning_rate": 4.383458502802107e-05,
+      "loss": 0.4011,
       "step": 34915
     },
     {
-      "epoch": 1.23,
-      "learning_rate": 4.414655284675091e-05,
-      "loss": 0.2878,
+      "epoch": 1.2585144339928642,
+      "grad_norm": 0.19101671874523163,
+      "learning_rate": 4.383266598896482e-05,
+      "loss": 0.4103,
       "step": 34920
     },
     {
-      "epoch": 1.23,
-      "learning_rate": 4.414472098086403e-05,
-      "loss": 0.2892,
+      "epoch": 1.2586946336540887,
+      "grad_norm": 0.21897098422050476,
+      "learning_rate": 4.383074669331715e-05,
+      "loss": 0.4145,
       "step": 34925
     },
     {
-      "epoch": 1.23,
-      "learning_rate": 4.4142888866395566e-05,
-      "loss": 0.2682,
+      "epoch": 1.2588748333153132,
+      "grad_norm": 0.22563257813453674,
+      "learning_rate": 4.3828827141104186e-05,
+      "loss": 0.414,
       "step": 34930
     },
     {
-      "epoch": 1.23,
-      "learning_rate": 4.41410565033693e-05,
-      "loss": 0.3151,
+      "epoch": 1.259055032976538,
+      "grad_norm": 0.16169485449790955,
+      "learning_rate": 4.382690733235212e-05,
+      "loss": 0.4148,
       "step": 34935
     },
     {
-      "epoch": 1.23,
-      "learning_rate": 4.413922389180904e-05,
-      "loss": 0.2662,
+      "epoch": 1.2592352326377627,
+      "grad_norm": 0.20008553564548492,
+      "learning_rate": 4.382498726708707e-05,
+      "loss": 0.4406,
       "step": 34940
     },
     {
-      "epoch": 1.23,
-      "learning_rate": 4.413739103173856e-05,
-      "loss": 0.2859,
+      "epoch": 1.2594154322989872,
+      "grad_norm": 0.21007391810417175,
+      "learning_rate": 4.3823066945335225e-05,
+      "loss": 0.4308,
       "step": 34945
     },
     {
-      "epoch": 1.23,
-      "learning_rate": 4.4135557923181684e-05,
-      "loss": 0.302,
+      "epoch": 1.259595631960212,
+      "grad_norm": 0.17790304124355316,
+      "learning_rate": 4.3821146367122726e-05,
+      "loss": 0.3854,
       "step": 34950
     },
     {
-      "epoch": 1.23,
-      "learning_rate": 4.4133724566162195e-05,
-      "loss": 0.2936,
+      "epoch": 1.2597758316214365,
+      "grad_norm": 0.221872016787529,
+      "learning_rate": 4.381922553247576e-05,
+      "loss": 0.3961,
       "step": 34955
     },
     {
-      "epoch": 1.23,
-      "learning_rate": 4.413189096070389e-05,
-      "loss": 0.2965,
+      "epoch": 1.2599560312826612,
+      "grad_norm": 0.16703324019908905,
+      "learning_rate": 4.381730444142048e-05,
+      "loss": 0.4415,
       "step": 34960
     },
     {
-      "epoch": 1.23,
-      "learning_rate": 4.41300571068306e-05,
-      "loss": 0.2763,
+      "epoch": 1.260136230943886,
+      "grad_norm": 0.21276666224002838,
+      "learning_rate": 4.3815383093983084e-05,
+      "loss": 0.4316,
       "step": 34965
     },
     {
-      "epoch": 1.23,
-      "learning_rate": 4.412822300456613e-05,
-      "loss": 0.2747,
+      "epoch": 1.2603164306051104,
+      "grad_norm": 0.2033817023038864,
+      "learning_rate": 4.381346149018973e-05,
+      "loss": 0.409,
       "step": 34970
     },
     {
-      "epoch": 1.23,
-      "learning_rate": 4.4126388653934276e-05,
-      "loss": 0.2767,
+      "epoch": 1.260496630266335,
+      "grad_norm": 0.15015016496181488,
+      "learning_rate": 4.38115396300666e-05,
+      "loss": 0.4541,
       "step": 34975
     },
     {
-      "epoch": 1.23,
-      "learning_rate": 4.412455405495888e-05,
-      "loss": 0.295,
+      "epoch": 1.2606768299275597,
+      "grad_norm": 0.19646312296390533,
+      "learning_rate": 4.3809617513639886e-05,
+      "loss": 0.4066,
       "step": 34980
     },
     {
-      "epoch": 1.23,
-      "learning_rate": 4.412271920766375e-05,
-      "loss": 0.2844,
+      "epoch": 1.2608570295887844,
+      "grad_norm": 0.17837940156459808,
+      "learning_rate": 4.380769514093578e-05,
+      "loss": 0.4212,
       "step": 34985
     },
     {
-      "epoch": 1.23,
-      "learning_rate": 4.4120884112072714e-05,
-      "loss": 0.2889,
+      "epoch": 1.261037229250009,
+      "grad_norm": 0.20663444697856903,
+      "learning_rate": 4.380577251198047e-05,
+      "loss": 0.4459,
       "step": 34990
     },
     {
-      "epoch": 1.23,
-      "learning_rate": 4.41190487682096e-05,
-      "loss": 0.303,
+      "epoch": 1.2612174289112337,
+      "grad_norm": 0.20016391575336456,
+      "learning_rate": 4.380384962680015e-05,
+      "loss": 0.4125,
       "step": 34995
     },
     {
-      "epoch": 1.23,
-      "learning_rate": 4.411721317609824e-05,
-      "loss": 0.2971,
+      "epoch": 1.2613976285724582,
+      "grad_norm": 0.20151177048683167,
+      "learning_rate": 4.380192648542101e-05,
+      "loss": 0.4607,
       "step": 35000
     },
     {
-      "epoch": 1.23,
-      "eval_loss": 0.2867205739021301,
-      "eval_runtime": 10.5517,
-      "eval_samples_per_second": 9.477,
-      "eval_steps_per_second": 9.477,
+      "epoch": 1.2613976285724582,
+      "eval_loss": 0.44922810792922974,
+      "eval_runtime": 3.5387,
+      "eval_samples_per_second": 28.259,
+      "eval_steps_per_second": 7.065,
       "step": 35000
     },
     {
-      "epoch": 1.23,
-      "learning_rate": 4.4115377335762456e-05,
-      "loss": 0.3064,
+      "epoch": 1.261577828233683,
+      "grad_norm": 0.18609340488910675,
+      "learning_rate": 4.380000308786927e-05,
+      "loss": 0.4066,
       "step": 35005
     },
     {
-      "epoch": 1.23,
-      "learning_rate": 4.41135412472261e-05,
-      "loss": 0.2653,
+      "epoch": 1.2617580278949077,
+      "grad_norm": 0.18463589251041412,
+      "learning_rate": 4.3798079434171124e-05,
+      "loss": 0.4624,
       "step": 35010
     },
     {
-      "epoch": 1.23,
-      "learning_rate": 4.411170491051301e-05,
-      "loss": 0.295,
+      "epoch": 1.2619382275561322,
+      "grad_norm": 0.21645288169384003,
+      "learning_rate": 4.379615552435279e-05,
+      "loss": 0.4172,
       "step": 35015
     },
     {
-      "epoch": 1.23,
-      "learning_rate": 4.410986832564702e-05,
-      "loss": 0.2904,
+      "epoch": 1.262118427217357,
+      "grad_norm": 0.18142879009246826,
+      "learning_rate": 4.379423135844048e-05,
+      "loss": 0.445,
       "step": 35020
     },
     {
-      "epoch": 1.23,
-      "learning_rate": 4.410803149265199e-05,
-      "loss": 0.256,
+      "epoch": 1.2622986268785814,
+      "grad_norm": 0.17872460186481476,
+      "learning_rate": 4.379230693646039e-05,
+      "loss": 0.4329,
       "step": 35025
     },
     {
-      "epoch": 1.23,
-      "learning_rate": 4.4106194411551766e-05,
-      "loss": 0.2899,
+      "epoch": 1.2624788265398061,
+      "grad_norm": 0.16483642160892487,
+      "learning_rate": 4.3790382258438776e-05,
+      "loss": 0.436,
       "step": 35030
     },
     {
-      "epoch": 1.23,
-      "learning_rate": 4.410435708237019e-05,
-      "loss": 0.2897,
+      "epoch": 1.2626590262010309,
+      "grad_norm": 0.2143619805574417,
+      "learning_rate": 4.3788457324401826e-05,
+      "loss": 0.4184,
       "step": 35035
     },
     {
-      "epoch": 1.23,
-      "learning_rate": 4.410251950513113e-05,
-      "loss": 0.3124,
+      "epoch": 1.2628392258622554,
+      "grad_norm": 0.16062313318252563,
+      "learning_rate": 4.378653213437579e-05,
+      "loss": 0.4251,
       "step": 35040
     },
     {
-      "epoch": 1.23,
-      "learning_rate": 4.410068167985845e-05,
-      "loss": 0.2721,
+      "epoch": 1.26301942552348,
+      "grad_norm": 0.206187441945076,
+      "learning_rate": 4.3784606688386885e-05,
+      "loss": 0.4143,
       "step": 35045
     },
     {
-      "epoch": 1.23,
-      "learning_rate": 4.4098843606576004e-05,
-      "loss": 0.2894,
+      "epoch": 1.2631996251847046,
+      "grad_norm": 0.15511402487754822,
+      "learning_rate": 4.3782680986461356e-05,
+      "loss": 0.3726,
       "step": 35050
     },
     {
-      "epoch": 1.23,
-      "learning_rate": 4.4097005285307654e-05,
-      "loss": 0.2857,
+      "epoch": 1.2633798248459294,
+      "grad_norm": 0.15216435492038727,
+      "learning_rate": 4.3780755028625434e-05,
+      "loss": 0.407,
       "step": 35055
     },
     {
-      "epoch": 1.23,
-      "learning_rate": 4.409516671607729e-05,
-      "loss": 0.277,
+      "epoch": 1.2635600245071539,
+      "grad_norm": 0.1635829508304596,
+      "learning_rate": 4.377882881490536e-05,
+      "loss": 0.4166,
       "step": 35060
     },
     {
-      "epoch": 1.23,
-      "learning_rate": 4.409332789890876e-05,
-      "loss": 0.3018,
+      "epoch": 1.2637402241683786,
+      "grad_norm": 0.2377825379371643,
+      "learning_rate": 4.377690234532739e-05,
+      "loss": 0.4311,
       "step": 35065
     },
     {
-      "epoch": 1.23,
-      "learning_rate": 4.409148883382595e-05,
-      "loss": 0.2893,
+      "epoch": 1.2639204238296031,
+      "grad_norm": 0.1933947205543518,
+      "learning_rate": 4.3774975619917744e-05,
+      "loss": 0.4266,
       "step": 35070
     },
     {
-      "epoch": 1.23,
-      "learning_rate": 4.4089649520852736e-05,
-      "loss": 0.285,
+      "epoch": 1.2641006234908279,
+      "grad_norm": 0.1821034997701645,
+      "learning_rate": 4.3773048638702694e-05,
+      "loss": 0.4347,
       "step": 35075
     },
     {
-      "epoch": 1.23,
-      "learning_rate": 4.4087809960013013e-05,
-      "loss": 0.3,
+      "epoch": 1.2642808231520526,
+      "grad_norm": 0.21266570687294006,
+      "learning_rate": 4.3771121401708495e-05,
+      "loss": 0.4509,
       "step": 35080
     },
     {
-      "epoch": 1.23,
-      "learning_rate": 4.408597015133065e-05,
-      "loss": 0.2834,
+      "epoch": 1.264461022813277,
+      "grad_norm": 0.16414852440357208,
+      "learning_rate": 4.3769193908961405e-05,
+      "loss": 0.44,
       "step": 35085
     },
     {
-      "epoch": 1.23,
-      "learning_rate": 4.408413009482954e-05,
-      "loss": 0.3,
+      "epoch": 1.2646412224745016,
+      "grad_norm": 0.15532633662223816,
+      "learning_rate": 4.3767266160487675e-05,
+      "loss": 0.4466,
       "step": 35090
     },
     {
-      "epoch": 1.23,
-      "learning_rate": 4.4082289790533586e-05,
-      "loss": 0.2805,
+      "epoch": 1.2648214221357263,
+      "grad_norm": 0.1961180716753006,
+      "learning_rate": 4.376533815631357e-05,
+      "loss": 0.4613,
       "step": 35095
     },
     {
-      "epoch": 1.23,
-      "learning_rate": 4.408044923846667e-05,
-      "loss": 0.3079,
+      "epoch": 1.265001621796951,
+      "grad_norm": 0.18854406476020813,
+      "learning_rate": 4.3763409896465376e-05,
+      "loss": 0.4098,
       "step": 35100
     },
     {
-      "epoch": 1.24,
-      "learning_rate": 4.4078608438652705e-05,
-      "loss": 0.3199,
+      "epoch": 1.2651818214581756,
+      "grad_norm": 0.181373730301857,
+      "learning_rate": 4.376148138096936e-05,
+      "loss": 0.4146,
       "step": 35105
     },
     {
-      "epoch": 1.24,
-      "learning_rate": 4.407676739111557e-05,
-      "loss": 0.2876,
+      "epoch": 1.2653620211194003,
+      "grad_norm": 0.16534672677516937,
+      "learning_rate": 4.3759552609851785e-05,
+      "loss": 0.4341,
       "step": 35110
     },
     {
-      "epoch": 1.24,
-      "learning_rate": 4.407492609587919e-05,
-      "loss": 0.2647,
+      "epoch": 1.2655422207806248,
+      "grad_norm": 0.22548224031925201,
+      "learning_rate": 4.375762358313894e-05,
+      "loss": 0.3947,
       "step": 35115
     },
     {
-      "epoch": 1.24,
-      "learning_rate": 4.407308455296746e-05,
-      "loss": 0.3269,
+      "epoch": 1.2657224204418496,
+      "grad_norm": 0.18645992875099182,
+      "learning_rate": 4.37556943008571e-05,
+      "loss": 0.4356,
       "step": 35120
     },
     {
-      "epoch": 1.24,
-      "learning_rate": 4.4071242762404294e-05,
-      "loss": 0.2996,
+      "epoch": 1.2659026201030743,
+      "grad_norm": 0.1680716574192047,
+      "learning_rate": 4.375376476303256e-05,
+      "loss": 0.3941,
       "step": 35125
     },
     {
-      "epoch": 1.24,
-      "learning_rate": 4.406940072421362e-05,
-      "loss": 0.2991,
+      "epoch": 1.2660828197642988,
+      "grad_norm": 0.18578827381134033,
+      "learning_rate": 4.375183496969161e-05,
+      "loss": 0.4242,
       "step": 35130
     },
     {
-      "epoch": 1.24,
-      "learning_rate": 4.406755843841933e-05,
-      "loss": 0.2908,
+      "epoch": 1.2662630194255236,
+      "grad_norm": 0.1809617131948471,
+      "learning_rate": 4.374990492086053e-05,
+      "loss": 0.4275,
       "step": 35135
     },
     {
-      "epoch": 1.24,
-      "learning_rate": 4.406571590504537e-05,
-      "loss": 0.2929,
+      "epoch": 1.266443219086748,
+      "grad_norm": 0.21433116495609283,
+      "learning_rate": 4.3747974616565634e-05,
+      "loss": 0.4564,
       "step": 35140
     },
     {
-      "epoch": 1.24,
-      "learning_rate": 4.406387312411565e-05,
-      "loss": 0.3158,
+      "epoch": 1.2666234187479728,
+      "grad_norm": 0.18926386535167694,
+      "learning_rate": 4.3746044056833205e-05,
+      "loss": 0.4426,
       "step": 35145
     },
     {
-      "epoch": 1.24,
-      "learning_rate": 4.4062030095654104e-05,
-      "loss": 0.2799,
+      "epoch": 1.2668036184091975,
+      "grad_norm": 0.19316570460796356,
+      "learning_rate": 4.3744113241689565e-05,
+      "loss": 0.4352,
       "step": 35150
     },
     {
-      "epoch": 1.24,
-      "learning_rate": 4.406018681968466e-05,
-      "loss": 0.2741,
+      "epoch": 1.266983818070422,
+      "grad_norm": 0.20651282370090485,
+      "learning_rate": 4.3742182171161005e-05,
+      "loss": 0.4239,
       "step": 35155
     },
     {
-      "epoch": 1.24,
-      "learning_rate": 4.405834329623125e-05,
-      "loss": 0.3071,
+      "epoch": 1.2671640177316466,
+      "grad_norm": 0.15775299072265625,
+      "learning_rate": 4.3740250845273845e-05,
+      "loss": 0.3809,
       "step": 35160
     },
     {
-      "epoch": 1.24,
-      "learning_rate": 4.405649952531781e-05,
-      "loss": 0.2802,
+      "epoch": 1.2673442173928713,
+      "grad_norm": 0.1828775703907013,
+      "learning_rate": 4.373831926405439e-05,
+      "loss": 0.4036,
       "step": 35165
     },
     {
-      "epoch": 1.24,
-      "learning_rate": 4.405465550696829e-05,
-      "loss": 0.2824,
+      "epoch": 1.267524417054096,
+      "grad_norm": 0.17841170728206635,
+      "learning_rate": 4.373638742752897e-05,
+      "loss": 0.4488,
       "step": 35170
     },
     {
-      "epoch": 1.24,
-      "learning_rate": 4.4052811241206614e-05,
-      "loss": 0.2678,
+      "epoch": 1.2677046167153205,
+      "grad_norm": 0.1723710149526596,
+      "learning_rate": 4.373445533572389e-05,
+      "loss": 0.4132,
       "step": 35175
     },
     {
-      "epoch": 1.24,
-      "learning_rate": 4.405096672805675e-05,
-      "loss": 0.2921,
+      "epoch": 1.2678848163765453,
+      "grad_norm": 0.2052830010652542,
+      "learning_rate": 4.373252298866549e-05,
+      "loss": 0.4367,
       "step": 35180
     },
     {
-      "epoch": 1.24,
-      "learning_rate": 4.404912196754264e-05,
-      "loss": 0.2858,
+      "epoch": 1.2680650160377698,
+      "grad_norm": 0.17017756402492523,
+      "learning_rate": 4.3730590386380086e-05,
+      "loss": 0.4235,
       "step": 35185
     },
     {
-      "epoch": 1.24,
-      "learning_rate": 4.404727695968823e-05,
-      "loss": 0.306,
+      "epoch": 1.2682452156989945,
+      "grad_norm": 0.1878291219472885,
+      "learning_rate": 4.372865752889402e-05,
+      "loss": 0.4508,
       "step": 35190
     },
     {
-      "epoch": 1.24,
-      "learning_rate": 4.4045431704517474e-05,
-      "loss": 0.3048,
+      "epoch": 1.2684254153602192,
+      "grad_norm": 0.1998129040002823,
+      "learning_rate": 4.3726724416233625e-05,
+      "loss": 0.4286,
       "step": 35195
     },
     {
-      "epoch": 1.24,
-      "learning_rate": 4.404358620205435e-05,
-      "loss": 0.2906,
+      "epoch": 1.2686056150214438,
+      "grad_norm": 0.1536918580532074,
+      "learning_rate": 4.372479104842522e-05,
+      "loss": 0.4135,
       "step": 35200
     },
     {
-      "epoch": 1.24,
-      "learning_rate": 4.40417404523228e-05,
-      "loss": 0.2993,
+      "epoch": 1.2687858146826683,
+      "grad_norm": 0.2284386157989502,
+      "learning_rate": 4.372285742549517e-05,
+      "loss": 0.4026,
       "step": 35205
     },
     {
-      "epoch": 1.24,
-      "learning_rate": 4.403989445534681e-05,
-      "loss": 0.2896,
+      "epoch": 1.268966014343893,
+      "grad_norm": 0.1657029688358307,
+      "learning_rate": 4.372092354746982e-05,
+      "loss": 0.4326,
       "step": 35210
     },
     {
-      "epoch": 1.24,
-      "learning_rate": 4.403804821115033e-05,
-      "loss": 0.3059,
+      "epoch": 1.2691462140051177,
+      "grad_norm": 0.1616412103176117,
+      "learning_rate": 4.37189894143755e-05,
+      "loss": 0.424,
       "step": 35215
     },
     {
-      "epoch": 1.24,
-      "learning_rate": 4.403620171975734e-05,
-      "loss": 0.286,
+      "epoch": 1.2693264136663422,
+      "grad_norm": 0.18977968394756317,
+      "learning_rate": 4.371705502623858e-05,
+      "loss": 0.4001,
       "step": 35220
     },
     {
-      "epoch": 1.24,
-      "learning_rate": 4.403435498119182e-05,
-      "loss": 0.324,
+      "epoch": 1.269506613327567,
+      "grad_norm": 0.20277686417102814,
+      "learning_rate": 4.371512038308541e-05,
+      "loss": 0.4667,
       "step": 35225
     },
     {
-      "epoch": 1.24,
-      "learning_rate": 4.403250799547774e-05,
-      "loss": 0.2888,
+      "epoch": 1.2696868129887915,
+      "grad_norm": 0.16292032599449158,
+      "learning_rate": 4.371318548494234e-05,
+      "loss": 0.4241,
       "step": 35230
     },
     {
-      "epoch": 1.24,
-      "learning_rate": 4.40306607626391e-05,
-      "loss": 0.2763,
+      "epoch": 1.2698670126500162,
+      "grad_norm": 0.18846076726913452,
+      "learning_rate": 4.3711250331835754e-05,
+      "loss": 0.3928,
       "step": 35235
     },
     {
-      "epoch": 1.24,
-      "learning_rate": 4.402881328269985e-05,
-      "loss": 0.3169,
+      "epoch": 1.270047212311241,
+      "grad_norm": 0.17085112631320953,
+      "learning_rate": 4.370931492379199e-05,
+      "loss": 0.4118,
       "step": 35240
     },
     {
-      "epoch": 1.24,
-      "learning_rate": 4.402696555568402e-05,
-      "loss": 0.3118,
+      "epoch": 1.2702274119724655,
+      "grad_norm": 0.18065665662288666,
+      "learning_rate": 4.3707379260837444e-05,
+      "loss": 0.4184,
       "step": 35245
     },
     {
-      "epoch": 1.24,
-      "learning_rate": 4.402511758161557e-05,
-      "loss": 0.2793,
+      "epoch": 1.2704076116336902,
+      "grad_norm": 0.278125137090683,
+      "learning_rate": 4.370544334299847e-05,
+      "loss": 0.4567,
       "step": 35250
     },
     {
-      "epoch": 1.24,
-      "learning_rate": 4.402326936051852e-05,
-      "loss": 0.3025,
+      "epoch": 1.2705878112949147,
+      "grad_norm": 0.1816800832748413,
+      "learning_rate": 4.3703507170301454e-05,
+      "loss": 0.4395,
       "step": 35255
     },
     {
-      "epoch": 1.24,
-      "learning_rate": 4.402142089241684e-05,
-      "loss": 0.261,
+      "epoch": 1.2707680109561394,
+      "grad_norm": 0.16282232105731964,
+      "learning_rate": 4.370157074277278e-05,
+      "loss": 0.3854,
       "step": 35260
     },
     {
-      "epoch": 1.24,
-      "learning_rate": 4.4019572177334545e-05,
-      "loss": 0.2901,
+      "epoch": 1.270948210617364,
+      "grad_norm": 0.1948549896478653,
+      "learning_rate": 4.369963406043881e-05,
+      "loss": 0.4173,
       "step": 35265
     },
     {
-      "epoch": 1.24,
-      "learning_rate": 4.401772321529565e-05,
-      "loss": 0.2884,
+      "epoch": 1.2711284102785887,
+      "grad_norm": 0.18401391804218292,
+      "learning_rate": 4.3697697123325956e-05,
+      "loss": 0.3982,
       "step": 35270
     },
     {
-      "epoch": 1.24,
-      "learning_rate": 4.401587400632414e-05,
-      "loss": 0.3042,
+      "epoch": 1.2713086099398132,
+      "grad_norm": 0.17478926479816437,
+      "learning_rate": 4.36957599314606e-05,
+      "loss": 0.4542,
       "step": 35275
     },
     {
-      "epoch": 1.24,
-      "learning_rate": 4.4014024550444044e-05,
-      "loss": 0.2762,
+      "epoch": 1.271488809601038,
+      "grad_norm": 0.20537783205509186,
+      "learning_rate": 4.369382248486914e-05,
+      "loss": 0.4193,
       "step": 35280
     },
     {
-      "epoch": 1.24,
-      "learning_rate": 4.401217484767937e-05,
-      "loss": 0.2782,
+      "epoch": 1.2716690092622627,
+      "grad_norm": 0.1872226595878601,
+      "learning_rate": 4.3691884783577966e-05,
+      "loss": 0.4263,
       "step": 35285
     },
     {
-      "epoch": 1.24,
-      "learning_rate": 4.401032489805413e-05,
-      "loss": 0.2859,
+      "epoch": 1.2718492089234872,
+      "grad_norm": 0.18654055893421173,
+      "learning_rate": 4.368994682761347e-05,
+      "loss": 0.4464,
       "step": 35290
     },
     {
-      "epoch": 1.24,
-      "learning_rate": 4.400847470159236e-05,
-      "loss": 0.2922,
+      "epoch": 1.272029408584712,
+      "grad_norm": 0.17078270018100739,
+      "learning_rate": 4.3688008617002076e-05,
+      "loss": 0.4367,
       "step": 35295
     },
     {
-      "epoch": 1.24,
-      "learning_rate": 4.400662425831807e-05,
-      "loss": 0.2677,
+      "epoch": 1.2722096082459364,
+      "grad_norm": 0.1947806477546692,
+      "learning_rate": 4.368607015177018e-05,
+      "loss": 0.4056,
       "step": 35300
     },
     {
-      "epoch": 1.24,
-      "learning_rate": 4.4004773568255284e-05,
-      "loss": 0.2656,
+      "epoch": 1.2723898079071612,
+      "grad_norm": 0.17518731951713562,
+      "learning_rate": 4.368413143194419e-05,
+      "loss": 0.4531,
       "step": 35305
     },
     {
-      "epoch": 1.24,
-      "learning_rate": 4.400292263142805e-05,
-      "loss": 0.2953,
+      "epoch": 1.272570007568386,
+      "grad_norm": 0.19480538368225098,
+      "learning_rate": 4.368219245755053e-05,
+      "loss": 0.448,
       "step": 35310
     },
     {
-      "epoch": 1.24,
-      "learning_rate": 4.400107144786038e-05,
-      "loss": 0.2876,
+      "epoch": 1.2727502072296104,
+      "grad_norm": 0.15128688514232635,
+      "learning_rate": 4.368025322861562e-05,
+      "loss": 0.4321,
       "step": 35315
     },
     {
-      "epoch": 1.24,
-      "learning_rate": 4.399922001757631e-05,
-      "loss": 0.2982,
+      "epoch": 1.272930406890835,
+      "grad_norm": 0.16583947837352753,
+      "learning_rate": 4.367831374516588e-05,
+      "loss": 0.4399,
       "step": 35320
     },
     {
-      "epoch": 1.24,
-      "learning_rate": 4.3997368340599896e-05,
-      "loss": 0.2891,
+      "epoch": 1.2731106065520597,
+      "grad_norm": 0.22142437100410461,
+      "learning_rate": 4.3676374007227715e-05,
+      "loss": 0.4275,
       "step": 35325
     },
     {
-      "epoch": 1.24,
-      "learning_rate": 4.3995516416955187e-05,
-      "loss": 0.2719,
+      "epoch": 1.2732908062132844,
+      "grad_norm": 0.1897389143705368,
+      "learning_rate": 4.367443401482758e-05,
+      "loss": 0.4079,
       "step": 35330
     },
     {
-      "epoch": 1.24,
-      "learning_rate": 4.39936642466662e-05,
-      "loss": 0.3066,
+      "epoch": 1.273471005874509,
+      "grad_norm": 0.18949352204799652,
+      "learning_rate": 4.36724937679919e-05,
+      "loss": 0.4659,
       "step": 35335
     },
     {
-      "epoch": 1.24,
-      "learning_rate": 4.399181182975701e-05,
-      "loss": 0.2932,
+      "epoch": 1.2736512055357336,
+      "grad_norm": 0.1753978729248047,
+      "learning_rate": 4.367055326674711e-05,
+      "loss": 0.4273,
       "step": 35340
     },
     {
-      "epoch": 1.24,
-      "learning_rate": 4.398995916625165e-05,
-      "loss": 0.3074,
+      "epoch": 1.2738314051969581,
+      "grad_norm": 0.20065420866012573,
+      "learning_rate": 4.366861251111963e-05,
+      "loss": 0.4463,
       "step": 35345
     },
     {
-      "epoch": 1.24,
-      "learning_rate": 4.398810625617419e-05,
-      "loss": 0.294,
+      "epoch": 1.2740116048581829,
+      "grad_norm": 0.19253739714622498,
+      "learning_rate": 4.366667150113594e-05,
+      "loss": 0.4188,
       "step": 35350
     },
     {
-      "epoch": 1.24,
-      "learning_rate": 4.398625309954868e-05,
-      "loss": 0.2983,
+      "epoch": 1.2741918045194076,
+      "grad_norm": 0.1715114414691925,
+      "learning_rate": 4.366473023682245e-05,
+      "loss": 0.4355,
       "step": 35355
     },
     {
-      "epoch": 1.24,
-      "learning_rate": 4.398439969639919e-05,
-      "loss": 0.2994,
+      "epoch": 1.2743720041806321,
+      "grad_norm": 0.17242814600467682,
+      "learning_rate": 4.3662788718205625e-05,
+      "loss": 0.4026,
       "step": 35360
     },
     {
-      "epoch": 1.24,
-      "learning_rate": 4.398254604674978e-05,
-      "loss": 0.265,
+      "epoch": 1.2745522038418569,
+      "grad_norm": 0.20738811790943146,
+      "learning_rate": 4.366084694531192e-05,
+      "loss": 0.4046,
       "step": 35365
     },
     {
-      "epoch": 1.24,
-      "learning_rate": 4.3980692150624524e-05,
-      "loss": 0.2759,
+      "epoch": 1.2747324035030814,
+      "grad_norm": 0.19424347579479218,
+      "learning_rate": 4.365890491816779e-05,
+      "loss": 0.4434,
       "step": 35370
     },
     {
-      "epoch": 1.24,
-      "learning_rate": 4.397883800804749e-05,
-      "loss": 0.2845,
+      "epoch": 1.274912603164306,
+      "grad_norm": 0.21286608278751373,
+      "learning_rate": 4.365696263679969e-05,
+      "loss": 0.4375,
       "step": 35375
     },
     {
-      "epoch": 1.24,
-      "learning_rate": 4.397698361904274e-05,
-      "loss": 0.2766,
+      "epoch": 1.2750928028255306,
+      "grad_norm": 0.20882990956306458,
+      "learning_rate": 4.365502010123409e-05,
+      "loss": 0.4316,
       "step": 35380
     },
     {
-      "epoch": 1.24,
-      "learning_rate": 4.397512898363437e-05,
-      "loss": 0.2817,
+      "epoch": 1.2752730024867553,
+      "grad_norm": 0.21594053506851196,
+      "learning_rate": 4.365307731149745e-05,
+      "loss": 0.3852,
       "step": 35385
     },
     {
-      "epoch": 1.25,
-      "learning_rate": 4.397327410184646e-05,
-      "loss": 0.293,
+      "epoch": 1.2754532021479799,
+      "grad_norm": 0.1830471307039261,
+      "learning_rate": 4.3651134267616244e-05,
+      "loss": 0.4404,
       "step": 35390
     },
     {
-      "epoch": 1.25,
-      "learning_rate": 4.3971418973703085e-05,
-      "loss": 0.3059,
+      "epoch": 1.2756334018092046,
+      "grad_norm": 0.17027118802070618,
+      "learning_rate": 4.3649190969616946e-05,
+      "loss": 0.4043,
       "step": 35395
     },
     {
-      "epoch": 1.25,
-      "learning_rate": 4.396956359922834e-05,
-      "loss": 0.2993,
+      "epoch": 1.2758136014704293,
+      "grad_norm": 0.16624897718429565,
+      "learning_rate": 4.364724741752603e-05,
+      "loss": 0.4311,
       "step": 35400
     },
     {
-      "epoch": 1.25,
-      "learning_rate": 4.3967707978446316e-05,
-      "loss": 0.2904,
+      "epoch": 1.2759938011316538,
+      "grad_norm": 0.17142795026302338,
+      "learning_rate": 4.364530361136998e-05,
+      "loss": 0.4524,
       "step": 35405
     },
     {
-      "epoch": 1.25,
-      "learning_rate": 4.39658521113811e-05,
-      "loss": 0.2725,
+      "epoch": 1.2761740007928786,
+      "grad_norm": 0.197305366396904,
+      "learning_rate": 4.364335955117528e-05,
+      "loss": 0.4254,
       "step": 35410
     },
     {
-      "epoch": 1.25,
-      "learning_rate": 4.3963995998056795e-05,
-      "loss": 0.2539,
+      "epoch": 1.276354200454103,
+      "grad_norm": 0.1855136901140213,
+      "learning_rate": 4.364141523696841e-05,
+      "loss": 0.4182,
       "step": 35415
     },
     {
-      "epoch": 1.25,
-      "learning_rate": 4.396213963849749e-05,
-      "loss": 0.3026,
+      "epoch": 1.2765344001153278,
+      "grad_norm": 0.1479266881942749,
+      "learning_rate": 4.3639470668775865e-05,
+      "loss": 0.3851,
       "step": 35420
     },
     {
-      "epoch": 1.25,
-      "learning_rate": 4.3960283032727313e-05,
-      "loss": 0.3028,
+      "epoch": 1.2767145997765526,
+      "grad_norm": 0.18044587969779968,
+      "learning_rate": 4.363752584662415e-05,
+      "loss": 0.4205,
       "step": 35425
     },
     {
-      "epoch": 1.25,
-      "learning_rate": 4.395842618077034e-05,
-      "loss": 0.2885,
+      "epoch": 1.276894799437777,
+      "grad_norm": 0.2061435431241989,
+      "learning_rate": 4.3635580770539744e-05,
+      "loss": 0.4146,
       "step": 35430
     },
     {
-      "epoch": 1.25,
-      "learning_rate": 4.395656908265071e-05,
-      "loss": 0.2749,
+      "epoch": 1.2770749990990016,
+      "grad_norm": 0.2000425010919571,
+      "learning_rate": 4.363363544054916e-05,
+      "loss": 0.4167,
       "step": 35435
     },
     {
-      "epoch": 1.25,
-      "learning_rate": 4.3954711738392526e-05,
-      "loss": 0.3019,
+      "epoch": 1.2772551987602263,
+      "grad_norm": 0.18255402147769928,
+      "learning_rate": 4.3631689856678905e-05,
+      "loss": 0.393,
       "step": 35440
     },
     {
-      "epoch": 1.25,
-      "learning_rate": 4.395285414801989e-05,
-      "loss": 0.3046,
+      "epoch": 1.277435398421451,
+      "grad_norm": 0.20933406054973602,
+      "learning_rate": 4.362974401895547e-05,
+      "loss": 0.4177,
       "step": 35445
     },
     {
-      "epoch": 1.25,
-      "learning_rate": 4.395099631155694e-05,
-      "loss": 0.2637,
+      "epoch": 1.2776155980826756,
+      "grad_norm": 0.1986079066991806,
+      "learning_rate": 4.36277979274054e-05,
+      "loss": 0.4176,
       "step": 35450
     },
     {
-      "epoch": 1.25,
-      "learning_rate": 4.3949138229027795e-05,
-      "loss": 0.2905,
+      "epoch": 1.2777957977439003,
+      "grad_norm": 0.21557162702083588,
+      "learning_rate": 4.3625851582055174e-05,
+      "loss": 0.3766,
       "step": 35455
     },
     {
-      "epoch": 1.25,
-      "learning_rate": 4.394727990045657e-05,
-      "loss": 0.2931,
+      "epoch": 1.2779759974051248,
+      "grad_norm": 0.17857031524181366,
+      "learning_rate": 4.362390498293134e-05,
+      "loss": 0.4071,
       "step": 35460
     },
     {
-      "epoch": 1.25,
-      "learning_rate": 4.394542132586741e-05,
-      "loss": 0.3078,
+      "epoch": 1.2781561970663495,
+      "grad_norm": 0.19413699209690094,
+      "learning_rate": 4.362195813006039e-05,
+      "loss": 0.4288,
       "step": 35465
     },
     {
-      "epoch": 1.25,
-      "learning_rate": 4.394356250528443e-05,
-      "loss": 0.2762,
+      "epoch": 1.2783363967275743,
+      "grad_norm": 0.16761010885238647,
+      "learning_rate": 4.362001102346888e-05,
+      "loss": 0.4273,
       "step": 35470
     },
     {
-      "epoch": 1.25,
-      "learning_rate": 4.394170343873178e-05,
-      "loss": 0.2756,
+      "epoch": 1.2785165963887988,
+      "grad_norm": 0.20636701583862305,
+      "learning_rate": 4.3618063663183315e-05,
+      "loss": 0.4188,
       "step": 35475
     },
     {
-      "epoch": 1.25,
-      "learning_rate": 4.39398441262336e-05,
-      "loss": 0.2865,
+      "epoch": 1.2786967960500233,
+      "grad_norm": 0.22335347533226013,
+      "learning_rate": 4.361611604923025e-05,
+      "loss": 0.4317,
       "step": 35480
     },
     {
-      "epoch": 1.25,
-      "learning_rate": 4.393798456781402e-05,
-      "loss": 0.2863,
+      "epoch": 1.278876995711248,
+      "grad_norm": 0.23554205894470215,
+      "learning_rate": 4.361416818163619e-05,
+      "loss": 0.4478,
       "step": 35485
     },
     {
-      "epoch": 1.25,
-      "learning_rate": 4.393612476349719e-05,
-      "loss": 0.2765,
+      "epoch": 1.2790571953724728,
+      "grad_norm": 0.22054627537727356,
+      "learning_rate": 4.361222006042771e-05,
+      "loss": 0.3942,
       "step": 35490
     },
     {
-      "epoch": 1.25,
-      "learning_rate": 4.393426471330726e-05,
-      "loss": 0.2794,
+      "epoch": 1.2792373950336973,
+      "grad_norm": 0.19049589335918427,
+      "learning_rate": 4.361027168563132e-05,
+      "loss": 0.4183,
       "step": 35495
     },
     {
-      "epoch": 1.25,
-      "learning_rate": 4.393240441726838e-05,
-      "loss": 0.3137,
+      "epoch": 1.279417594694922,
+      "grad_norm": 0.1891845464706421,
+      "learning_rate": 4.360832305727359e-05,
+      "loss": 0.4341,
       "step": 35500
     },
     {
-      "epoch": 1.25,
-      "eval_loss": 0.285795658826828,
-      "eval_runtime": 10.5526,
-      "eval_samples_per_second": 9.476,
-      "eval_steps_per_second": 9.476,
+      "epoch": 1.279417594694922,
+      "eval_loss": 0.44991669058799744,
+      "eval_runtime": 3.8804,
+      "eval_samples_per_second": 25.771,
+      "eval_steps_per_second": 6.443,
       "step": 35500
     },
     {
-      "epoch": 1.25,
-      "learning_rate": 4.393054387540471e-05,
-      "loss": 0.2958,
+      "epoch": 1.2795977943561465,
+      "grad_norm": 0.16972802579402924,
+      "learning_rate": 4.360637417538106e-05,
+      "loss": 0.3922,
       "step": 35505
     },
     {
-      "epoch": 1.25,
-      "learning_rate": 4.3928683087740394e-05,
-      "loss": 0.27,
+      "epoch": 1.2797779940173712,
+      "grad_norm": 0.15279355645179749,
+      "learning_rate": 4.360442503998028e-05,
+      "loss": 0.387,
       "step": 35510
     },
     {
-      "epoch": 1.25,
-      "learning_rate": 4.392682205429962e-05,
-      "loss": 0.296,
+      "epoch": 1.279958193678596,
+      "grad_norm": 0.19364698231220245,
+      "learning_rate": 4.360247565109782e-05,
+      "loss": 0.4193,
       "step": 35515
     },
     {
-      "epoch": 1.25,
-      "learning_rate": 4.392496077510652e-05,
-      "loss": 0.3007,
+      "epoch": 1.2801383933398205,
+      "grad_norm": 0.13126322627067566,
+      "learning_rate": 4.3600526008760226e-05,
+      "loss": 0.3859,
       "step": 35520
     },
     {
-      "epoch": 1.25,
-      "learning_rate": 4.392309925018528e-05,
-      "loss": 0.3007,
+      "epoch": 1.2803185930010452,
+      "grad_norm": 0.18221980333328247,
+      "learning_rate": 4.359857611299406e-05,
+      "loss": 0.4055,
       "step": 35525
     },
     {
-      "epoch": 1.25,
-      "learning_rate": 4.3921237479560066e-05,
-      "loss": 0.2779,
+      "epoch": 1.2804987926622697,
+      "grad_norm": 0.13313446938991547,
+      "learning_rate": 4.35966259638259e-05,
+      "loss": 0.4021,
       "step": 35530
     },
     {
-      "epoch": 1.25,
-      "learning_rate": 4.3919375463255055e-05,
-      "loss": 0.2885,
+      "epoch": 1.2806789923234945,
+      "grad_norm": 0.20920483767986298,
+      "learning_rate": 4.359467556128232e-05,
+      "loss": 0.4436,
       "step": 35535
     },
     {
-      "epoch": 1.25,
-      "learning_rate": 4.391751320129442e-05,
-      "loss": 0.2823,
+      "epoch": 1.2808591919847192,
+      "grad_norm": 0.15556688606739044,
+      "learning_rate": 4.359272490538987e-05,
+      "loss": 0.3783,
       "step": 35540
     },
     {
-      "epoch": 1.25,
-      "learning_rate": 4.3915650693702345e-05,
-      "loss": 0.2783,
+      "epoch": 1.2810393916459437,
+      "grad_norm": 0.17486773431301117,
+      "learning_rate": 4.359077399617515e-05,
+      "loss": 0.4021,
       "step": 35545
     },
     {
-      "epoch": 1.25,
-      "learning_rate": 4.391378794050302e-05,
-      "loss": 0.29,
+      "epoch": 1.2812195913071682,
+      "grad_norm": 0.1680872142314911,
+      "learning_rate": 4.358882283366473e-05,
+      "loss": 0.3872,
       "step": 35550
     },
     {
-      "epoch": 1.25,
-      "learning_rate": 4.391192494172061e-05,
-      "loss": 0.2846,
+      "epoch": 1.281399790968393,
+      "grad_norm": 0.17405681312084198,
+      "learning_rate": 4.3586871417885204e-05,
+      "loss": 0.4278,
       "step": 35555
     },
     {
-      "epoch": 1.25,
-      "learning_rate": 4.391006169737932e-05,
-      "loss": 0.2778,
+      "epoch": 1.2815799906296177,
+      "grad_norm": 0.1495596170425415,
+      "learning_rate": 4.358491974886315e-05,
+      "loss": 0.4176,
       "step": 35560
     },
     {
-      "epoch": 1.25,
-      "learning_rate": 4.390819820750334e-05,
-      "loss": 0.2847,
+      "epoch": 1.2817601902908422,
+      "grad_norm": 0.188494473695755,
+      "learning_rate": 4.358296782662517e-05,
+      "loss": 0.4321,
       "step": 35565
     },
     {
-      "epoch": 1.25,
-      "learning_rate": 4.390633447211687e-05,
-      "loss": 0.3156,
+      "epoch": 1.281940389952067,
+      "grad_norm": 0.20558683574199677,
+      "learning_rate": 4.358101565119784e-05,
+      "loss": 0.4239,
       "step": 35570
     },
     {
-      "epoch": 1.25,
-      "learning_rate": 4.390447049124411e-05,
-      "loss": 0.3048,
+      "epoch": 1.2821205896132915,
+      "grad_norm": 0.16552236676216125,
+      "learning_rate": 4.3579063222607776e-05,
+      "loss": 0.4232,
       "step": 35575
     },
     {
-      "epoch": 1.25,
-      "learning_rate": 4.390260626490926e-05,
-      "loss": 0.2952,
+      "epoch": 1.2823007892745162,
+      "grad_norm": 0.17347842454910278,
+      "learning_rate": 4.357711054088157e-05,
+      "loss": 0.4082,
       "step": 35580
     },
     {
-      "epoch": 1.25,
-      "learning_rate": 4.390074179313652e-05,
-      "loss": 0.2893,
+      "epoch": 1.282480988935741,
+      "grad_norm": 0.17773760855197906,
+      "learning_rate": 4.357515760604583e-05,
+      "loss": 0.3855,
       "step": 35585
     },
     {
-      "epoch": 1.25,
-      "learning_rate": 4.389887707595011e-05,
-      "loss": 0.3191,
+      "epoch": 1.2826611885969654,
+      "grad_norm": 0.16329875588417053,
+      "learning_rate": 4.3573204418127165e-05,
+      "loss": 0.3963,
       "step": 35590
     },
     {
-      "epoch": 1.25,
-      "learning_rate": 4.3897012113374224e-05,
-      "loss": 0.2988,
+      "epoch": 1.28284138825819,
+      "grad_norm": 0.195121169090271,
+      "learning_rate": 4.357125097715218e-05,
+      "loss": 0.4459,
       "step": 35595
     },
     {
-      "epoch": 1.25,
-      "learning_rate": 4.3895146905433096e-05,
-      "loss": 0.3018,
+      "epoch": 1.2830215879194147,
+      "grad_norm": 0.22871240973472595,
+      "learning_rate": 4.35692972831475e-05,
+      "loss": 0.4106,
       "step": 35600
     },
     {
-      "epoch": 1.25,
-      "learning_rate": 4.389328145215094e-05,
-      "loss": 0.2972,
+      "epoch": 1.2832017875806394,
+      "grad_norm": 0.13964316248893738,
+      "learning_rate": 4.356734333613974e-05,
+      "loss": 0.4066,
       "step": 35605
     },
     {
-      "epoch": 1.25,
-      "learning_rate": 4.389141575355198e-05,
-      "loss": 0.2698,
+      "epoch": 1.283381987241864,
+      "grad_norm": 0.16373074054718018,
+      "learning_rate": 4.356538913615553e-05,
+      "loss": 0.4039,
       "step": 35610
     },
     {
-      "epoch": 1.25,
-      "learning_rate": 4.3889549809660425e-05,
-      "loss": 0.2763,
+      "epoch": 1.2835621869030887,
+      "grad_norm": 0.18498462438583374,
+      "learning_rate": 4.3563434683221475e-05,
+      "loss": 0.4177,
       "step": 35615
     },
     {
-      "epoch": 1.25,
-      "learning_rate": 4.388768362050052e-05,
-      "loss": 0.278,
+      "epoch": 1.2837423865643132,
+      "grad_norm": 0.2500152289867401,
+      "learning_rate": 4.356147997736422e-05,
+      "loss": 0.4178,
       "step": 35620
     },
     {
-      "epoch": 1.25,
-      "learning_rate": 4.388581718609649e-05,
-      "loss": 0.2908,
+      "epoch": 1.283922586225538,
+      "grad_norm": 0.17622318863868713,
+      "learning_rate": 4.3559525018610395e-05,
+      "loss": 0.3932,
       "step": 35625
     },
     {
-      "epoch": 1.25,
-      "learning_rate": 4.388395050647257e-05,
-      "loss": 0.2866,
+      "epoch": 1.2841027858867626,
+      "grad_norm": 0.16591259837150574,
+      "learning_rate": 4.355756980698664e-05,
+      "loss": 0.402,
       "step": 35630
     },
     {
-      "epoch": 1.25,
-      "learning_rate": 4.3882083581652996e-05,
-      "loss": 0.3039,
+      "epoch": 1.2842829855479871,
+      "grad_norm": 0.22070670127868652,
+      "learning_rate": 4.355561434251958e-05,
+      "loss": 0.4183,
       "step": 35635
     },
     {
-      "epoch": 1.25,
-      "learning_rate": 4.3880216411662015e-05,
-      "loss": 0.2785,
+      "epoch": 1.2844631852092119,
+      "grad_norm": 0.205961674451828,
+      "learning_rate": 4.3553658625235874e-05,
+      "loss": 0.4136,
       "step": 35640
     },
     {
-      "epoch": 1.25,
-      "learning_rate": 4.387834899652387e-05,
-      "loss": 0.2908,
+      "epoch": 1.2846433848704364,
+      "grad_norm": 0.1596502810716629,
+      "learning_rate": 4.355170265516216e-05,
+      "loss": 0.4191,
       "step": 35645
     },
     {
-      "epoch": 1.25,
-      "learning_rate": 4.38764813362628e-05,
-      "loss": 0.3119,
+      "epoch": 1.2848235845316611,
+      "grad_norm": 0.15234126150608063,
+      "learning_rate": 4.354974643232508e-05,
+      "loss": 0.4161,
       "step": 35650
     },
     {
-      "epoch": 1.25,
-      "learning_rate": 4.387461343090305e-05,
-      "loss": 0.3132,
+      "epoch": 1.2850037841928856,
+      "grad_norm": 0.181291401386261,
+      "learning_rate": 4.354778995675131e-05,
+      "loss": 0.4216,
       "step": 35655
     },
     {
-      "epoch": 1.25,
-      "learning_rate": 4.38727452804689e-05,
-      "loss": 0.2776,
+      "epoch": 1.2851839838541104,
+      "grad_norm": 0.16666920483112335,
+      "learning_rate": 4.354583322846748e-05,
+      "loss": 0.4508,
       "step": 35660
     },
     {
-      "epoch": 1.25,
-      "learning_rate": 4.387087688498458e-05,
-      "loss": 0.2854,
+      "epoch": 1.2853641835153349,
+      "grad_norm": 0.22122518718242645,
+      "learning_rate": 4.354387624750027e-05,
+      "loss": 0.413,
       "step": 35665
     },
     {
-      "epoch": 1.25,
-      "learning_rate": 4.386900824447436e-05,
-      "loss": 0.2701,
+      "epoch": 1.2855443831765596,
+      "grad_norm": 0.18632946908473969,
+      "learning_rate": 4.354191901387634e-05,
+      "loss": 0.42,
       "step": 35670
     },
     {
-      "epoch": 1.26,
-      "learning_rate": 4.3867139358962506e-05,
-      "loss": 0.2914,
+      "epoch": 1.2857245828377843,
+      "grad_norm": 0.16238003969192505,
+      "learning_rate": 4.3539961527622345e-05,
+      "loss": 0.3928,
       "step": 35675
     },
     {
-      "epoch": 1.26,
-      "learning_rate": 4.386527022847328e-05,
-      "loss": 0.2743,
+      "epoch": 1.2859047824990089,
+      "grad_norm": 0.21126984059810638,
+      "learning_rate": 4.353800378876497e-05,
+      "loss": 0.3968,
       "step": 35680
     },
     {
-      "epoch": 1.26,
-      "learning_rate": 4.3863400853030954e-05,
-      "loss": 0.3009,
+      "epoch": 1.2860849821602336,
+      "grad_norm": 0.17129042744636536,
+      "learning_rate": 4.3536045797330885e-05,
+      "loss": 0.4172,
       "step": 35685
     },
     {
-      "epoch": 1.26,
-      "learning_rate": 4.3861531232659796e-05,
-      "loss": 0.2845,
+      "epoch": 1.286265181821458,
+      "grad_norm": 0.18168915808200836,
+      "learning_rate": 4.353408755334676e-05,
+      "loss": 0.4172,
       "step": 35690
     },
     {
-      "epoch": 1.26,
-      "learning_rate": 4.3859661367384084e-05,
-      "loss": 0.2956,
+      "epoch": 1.2864453814826828,
+      "grad_norm": 0.15571817755699158,
+      "learning_rate": 4.3532129056839274e-05,
+      "loss": 0.3766,
       "step": 35695
     },
     {
-      "epoch": 1.26,
-      "learning_rate": 4.385779125722811e-05,
-      "loss": 0.2896,
+      "epoch": 1.2866255811439076,
+      "grad_norm": 0.2171332985162735,
+      "learning_rate": 4.353017030783513e-05,
+      "loss": 0.4286,
       "step": 35700
     },
     {
-      "epoch": 1.26,
-      "learning_rate": 4.385592090221613e-05,
-      "loss": 0.2897,
+      "epoch": 1.286805780805132,
+      "grad_norm": 0.1743398904800415,
+      "learning_rate": 4.3528211306360986e-05,
+      "loss": 0.4258,
       "step": 35705
     },
     {
-      "epoch": 1.26,
-      "learning_rate": 4.3854050302372453e-05,
-      "loss": 0.299,
+      "epoch": 1.2869859804663566,
+      "grad_norm": 0.1896362006664276,
+      "learning_rate": 4.352625205244357e-05,
+      "loss": 0.4056,
       "step": 35710
     },
     {
-      "epoch": 1.26,
-      "learning_rate": 4.3852179457721354e-05,
-      "loss": 0.2715,
+      "epoch": 1.2871661801275813,
+      "grad_norm": 0.1765444278717041,
+      "learning_rate": 4.352429254610955e-05,
+      "loss": 0.4137,
       "step": 35715
     },
     {
-      "epoch": 1.26,
-      "learning_rate": 4.385030836828713e-05,
-      "loss": 0.3069,
+      "epoch": 1.287346379788806,
+      "grad_norm": 0.20674960315227509,
+      "learning_rate": 4.352233278738562e-05,
+      "loss": 0.4382,
       "step": 35720
     },
     {
-      "epoch": 1.26,
-      "learning_rate": 4.384843703409407e-05,
-      "loss": 0.2927,
+      "epoch": 1.2875265794500306,
+      "grad_norm": 0.1722467541694641,
+      "learning_rate": 4.35203727762985e-05,
+      "loss": 0.429,
       "step": 35725
     },
     {
-      "epoch": 1.26,
-      "learning_rate": 4.384656545516649e-05,
-      "loss": 0.3173,
+      "epoch": 1.2877067791112553,
+      "grad_norm": 0.1646967977285385,
+      "learning_rate": 4.3518412512874885e-05,
+      "loss": 0.3841,
       "step": 35730
     },
     {
-      "epoch": 1.26,
-      "learning_rate": 4.384469363152867e-05,
-      "loss": 0.2901,
+      "epoch": 1.2878869787724798,
+      "grad_norm": 0.19040453433990479,
+      "learning_rate": 4.3516451997141485e-05,
+      "loss": 0.4182,
       "step": 35735
     },
     {
-      "epoch": 1.26,
-      "learning_rate": 4.3842821563204916e-05,
-      "loss": 0.2724,
+      "epoch": 1.2880671784337046,
+      "grad_norm": 0.1459205448627472,
+      "learning_rate": 4.3514491229125015e-05,
+      "loss": 0.3975,
       "step": 35740
     },
     {
-      "epoch": 1.26,
-      "learning_rate": 4.3840949250219546e-05,
-      "loss": 0.2853,
+      "epoch": 1.2882473780949293,
+      "grad_norm": 0.1318344622850418,
+      "learning_rate": 4.3512530208852185e-05,
+      "loss": 0.4007,
       "step": 35745
     },
     {
-      "epoch": 1.26,
-      "learning_rate": 4.383907669259687e-05,
-      "loss": 0.2798,
+      "epoch": 1.2884275777561538,
+      "grad_norm": 0.21362341940402985,
+      "learning_rate": 4.3510568936349714e-05,
+      "loss": 0.4328,
       "step": 35750
     },
     {
-      "epoch": 1.26,
-      "learning_rate": 4.383720389036119e-05,
-      "loss": 0.2963,
+      "epoch": 1.2886077774173785,
+      "grad_norm": 0.2101558893918991,
+      "learning_rate": 4.350860741164432e-05,
+      "loss": 0.468,
       "step": 35755
     },
     {
-      "epoch": 1.26,
-      "learning_rate": 4.383533084353684e-05,
-      "loss": 0.2845,
+      "epoch": 1.288787977078603,
+      "grad_norm": 0.1813066005706787,
+      "learning_rate": 4.350664563476274e-05,
+      "loss": 0.4452,
       "step": 35760
     },
     {
-      "epoch": 1.26,
-      "learning_rate": 4.383345755214813e-05,
-      "loss": 0.2914,
+      "epoch": 1.2889681767398278,
+      "grad_norm": 0.1613730937242508,
+      "learning_rate": 4.35046836057317e-05,
+      "loss": 0.4398,
       "step": 35765
     },
     {
-      "epoch": 1.26,
-      "learning_rate": 4.3831584016219385e-05,
-      "loss": 0.2825,
+      "epoch": 1.2891483764010523,
+      "grad_norm": 0.17162039875984192,
+      "learning_rate": 4.350272132457792e-05,
+      "loss": 0.4311,
       "step": 35770
     },
     {
-      "epoch": 1.26,
-      "learning_rate": 4.382971023577493e-05,
-      "loss": 0.3114,
+      "epoch": 1.289328576062277,
+      "grad_norm": 0.16017718613147736,
+      "learning_rate": 4.350075879132815e-05,
+      "loss": 0.413,
       "step": 35775
     },
     {
-      "epoch": 1.26,
-      "learning_rate": 4.382783621083911e-05,
-      "loss": 0.3057,
+      "epoch": 1.2895087757235015,
+      "grad_norm": 0.16685138642787933,
+      "learning_rate": 4.349879600600912e-05,
+      "loss": 0.4324,
       "step": 35780
     },
     {
-      "epoch": 1.26,
-      "learning_rate": 4.3825961941436236e-05,
-      "loss": 0.27,
+      "epoch": 1.2896889753847263,
+      "grad_norm": 0.18246233463287354,
+      "learning_rate": 4.349683296864758e-05,
+      "loss": 0.4127,
       "step": 35785
     },
     {
-      "epoch": 1.26,
-      "learning_rate": 4.3824087427590645e-05,
-      "loss": 0.2814,
+      "epoch": 1.289869175045951,
+      "grad_norm": 0.17236393690109253,
+      "learning_rate": 4.349486967927027e-05,
+      "loss": 0.4159,
       "step": 35790
     },
     {
-      "epoch": 1.26,
-      "learning_rate": 4.382221266932669e-05,
-      "loss": 0.2841,
+      "epoch": 1.2900493747071755,
+      "grad_norm": 0.17806027829647064,
+      "learning_rate": 4.349290613790393e-05,
+      "loss": 0.4077,
       "step": 35795
     },
     {
-      "epoch": 1.26,
-      "learning_rate": 4.382033766666871e-05,
-      "loss": 0.2814,
+      "epoch": 1.2902295743684002,
+      "grad_norm": 0.17833392322063446,
+      "learning_rate": 4.3490942344575336e-05,
+      "loss": 0.4335,
       "step": 35800
     },
     {
-      "epoch": 1.26,
-      "learning_rate": 4.381846241964105e-05,
-      "loss": 0.2871,
+      "epoch": 1.2904097740296248,
+      "grad_norm": 0.2036380022764206,
+      "learning_rate": 4.348897829931123e-05,
+      "loss": 0.4366,
       "step": 35805
     },
     {
-      "epoch": 1.26,
-      "learning_rate": 4.381658692826805e-05,
-      "loss": 0.2807,
+      "epoch": 1.2905899736908495,
+      "grad_norm": 0.1541140377521515,
+      "learning_rate": 4.348701400213838e-05,
+      "loss": 0.4208,
       "step": 35810
     },
     {
-      "epoch": 1.26,
-      "learning_rate": 4.3814711192574075e-05,
-      "loss": 0.2895,
+      "epoch": 1.2907701733520742,
+      "grad_norm": 0.19483250379562378,
+      "learning_rate": 4.3485049453083536e-05,
+      "loss": 0.3919,
       "step": 35815
     },
     {
-      "epoch": 1.26,
-      "learning_rate": 4.381283521258347e-05,
-      "loss": 0.3185,
+      "epoch": 1.2909503730132987,
+      "grad_norm": 0.1808329075574875,
+      "learning_rate": 4.348308465217348e-05,
+      "loss": 0.4261,
       "step": 35820
     },
     {
-      "epoch": 1.26,
-      "learning_rate": 4.38109589883206e-05,
-      "loss": 0.2811,
+      "epoch": 1.2911305726745232,
+      "grad_norm": 0.208456888794899,
+      "learning_rate": 4.348111959943496e-05,
+      "loss": 0.39,
       "step": 35825
     },
     {
-      "epoch": 1.26,
-      "learning_rate": 4.3809082519809826e-05,
-      "loss": 0.291,
+      "epoch": 1.291310772335748,
+      "grad_norm": 0.19897425174713135,
+      "learning_rate": 4.3479154294894774e-05,
+      "loss": 0.4752,
       "step": 35830
     },
     {
-      "epoch": 1.26,
-      "learning_rate": 4.380720580707551e-05,
-      "loss": 0.2967,
+      "epoch": 1.2914909719969727,
+      "grad_norm": 0.16709640622138977,
+      "learning_rate": 4.347718873857969e-05,
+      "loss": 0.4283,
       "step": 35835
     },
     {
-      "epoch": 1.26,
-      "learning_rate": 4.380532885014202e-05,
-      "loss": 0.2848,
+      "epoch": 1.2916711716581972,
+      "grad_norm": 0.19628377258777618,
+      "learning_rate": 4.347522293051648e-05,
+      "loss": 0.4476,
       "step": 35840
     },
     {
-      "epoch": 1.26,
-      "learning_rate": 4.380345164903373e-05,
-      "loss": 0.2688,
+      "epoch": 1.291851371319422,
+      "grad_norm": 0.19053253531455994,
+      "learning_rate": 4.3473256870731935e-05,
+      "loss": 0.4143,
       "step": 35845
     },
     {
-      "epoch": 1.26,
-      "learning_rate": 4.380157420377501e-05,
-      "loss": 0.2852,
+      "epoch": 1.2920315709806465,
+      "grad_norm": 0.16357672214508057,
+      "learning_rate": 4.347129055925285e-05,
+      "loss": 0.4099,
       "step": 35850
     },
     {
-      "epoch": 1.26,
-      "learning_rate": 4.379969651439024e-05,
-      "loss": 0.297,
+      "epoch": 1.2922117706418712,
+      "grad_norm": 0.17562684416770935,
+      "learning_rate": 4.3469323996106e-05,
+      "loss": 0.3779,
       "step": 35855
     },
     {
-      "epoch": 1.26,
-      "learning_rate": 4.3797818580903806e-05,
-      "loss": 0.2875,
+      "epoch": 1.292391970303096,
+      "grad_norm": 0.21830067038536072,
+      "learning_rate": 4.346735718131819e-05,
+      "loss": 0.4124,
       "step": 35860
     },
     {
-      "epoch": 1.26,
-      "learning_rate": 4.3795940403340076e-05,
-      "loss": 0.2677,
+      "epoch": 1.2925721699643204,
+      "grad_norm": 0.21300473809242249,
+      "learning_rate": 4.3465390114916206e-05,
+      "loss": 0.3939,
       "step": 35865
     },
     {
-      "epoch": 1.26,
-      "learning_rate": 4.379406198172345e-05,
-      "loss": 0.2658,
+      "epoch": 1.2927523696255452,
+      "grad_norm": 0.15682797133922577,
+      "learning_rate": 4.3463422796926864e-05,
+      "loss": 0.3946,
       "step": 35870
     },
     {
-      "epoch": 1.26,
-      "learning_rate": 4.3792183316078324e-05,
-      "loss": 0.2732,
+      "epoch": 1.2929325692867697,
+      "grad_norm": 0.1903832107782364,
+      "learning_rate": 4.3461455227376956e-05,
+      "loss": 0.4147,
       "step": 35875
     },
     {
-      "epoch": 1.26,
-      "learning_rate": 4.379030440642907e-05,
-      "loss": 0.3065,
+      "epoch": 1.2931127689479944,
+      "grad_norm": 0.1733843833208084,
+      "learning_rate": 4.3459487406293296e-05,
+      "loss": 0.4034,
       "step": 35880
     },
     {
-      "epoch": 1.26,
-      "learning_rate": 4.3788425252800104e-05,
-      "loss": 0.294,
+      "epoch": 1.293292968609219,
+      "grad_norm": 0.193365678191185,
+      "learning_rate": 4.34575193337027e-05,
+      "loss": 0.4185,
       "step": 35885
     },
     {
-      "epoch": 1.26,
-      "learning_rate": 4.378654585521582e-05,
-      "loss": 0.2804,
+      "epoch": 1.2934731682704437,
+      "grad_norm": 0.19296406209468842,
+      "learning_rate": 4.345555100963198e-05,
+      "loss": 0.4011,
       "step": 35890
     },
     {
-      "epoch": 1.26,
-      "learning_rate": 4.378466621370061e-05,
-      "loss": 0.2957,
+      "epoch": 1.2936533679316682,
+      "grad_norm": 0.16157175600528717,
+      "learning_rate": 4.3453582434107934e-05,
+      "loss": 0.4164,
       "step": 35895
     },
     {
-      "epoch": 1.26,
-      "learning_rate": 4.378278632827889e-05,
-      "loss": 0.2936,
+      "epoch": 1.293833567592893,
+      "grad_norm": 0.17810019850730896,
+      "learning_rate": 4.3451613607157416e-05,
+      "loss": 0.394,
       "step": 35900
     },
     {
-      "epoch": 1.26,
-      "learning_rate": 4.378090619897508e-05,
-      "loss": 0.2891,
+      "epoch": 1.2940137672541177,
+      "grad_norm": 0.1655871868133545,
+      "learning_rate": 4.344964452880723e-05,
+      "loss": 0.3975,
       "step": 35905
     },
     {
-      "epoch": 1.26,
-      "learning_rate": 4.377902582581357e-05,
-      "loss": 0.267,
+      "epoch": 1.2941939669153422,
+      "grad_norm": 0.24324651062488556,
+      "learning_rate": 4.3447675199084204e-05,
+      "loss": 0.4328,
       "step": 35910
     },
     {
-      "epoch": 1.26,
-      "learning_rate": 4.377714520881879e-05,
-      "loss": 0.3045,
+      "epoch": 1.294374166576567,
+      "grad_norm": 0.15802602469921112,
+      "learning_rate": 4.344570561801518e-05,
+      "loss": 0.4295,
       "step": 35915
     },
     {
-      "epoch": 1.26,
-      "learning_rate": 4.377526434801515e-05,
-      "loss": 0.3042,
+      "epoch": 1.2945543662377914,
+      "grad_norm": 0.1700577437877655,
+      "learning_rate": 4.344373578562698e-05,
+      "loss": 0.3671,
       "step": 35920
     },
     {
-      "epoch": 1.26,
-      "learning_rate": 4.377338324342708e-05,
-      "loss": 0.3083,
+      "epoch": 1.2947345658990161,
+      "grad_norm": 0.1449105143547058,
+      "learning_rate": 4.3441765701946455e-05,
+      "loss": 0.4034,
       "step": 35925
     },
     {
-      "epoch": 1.26,
-      "learning_rate": 4.377150189507899e-05,
-      "loss": 0.292,
+      "epoch": 1.2949147655602409,
+      "grad_norm": 0.18655553460121155,
+      "learning_rate": 4.343979536700045e-05,
+      "loss": 0.3888,
       "step": 35930
     },
     {
-      "epoch": 1.26,
-      "learning_rate": 4.376962030299533e-05,
-      "loss": 0.311,
+      "epoch": 1.2950949652214654,
+      "grad_norm": 0.1947435736656189,
+      "learning_rate": 4.34378247808158e-05,
+      "loss": 0.4335,
       "step": 35935
     },
     {
-      "epoch": 1.26,
-      "learning_rate": 4.376773846720051e-05,
-      "loss": 0.2774,
+      "epoch": 1.29527516488269,
+      "grad_norm": 0.1621570587158203,
+      "learning_rate": 4.343585394341936e-05,
+      "loss": 0.4024,
       "step": 35940
     },
     {
-      "epoch": 1.26,
-      "learning_rate": 4.3765856387718984e-05,
-      "loss": 0.2974,
+      "epoch": 1.2954553645439146,
+      "grad_norm": 0.15728051960468292,
+      "learning_rate": 4.343388285483797e-05,
+      "loss": 0.3716,
       "step": 35945
     },
     {
-      "epoch": 1.26,
-      "learning_rate": 4.3763974064575175e-05,
-      "loss": 0.2974,
+      "epoch": 1.2956355642051394,
+      "grad_norm": 0.17615413665771484,
+      "learning_rate": 4.34319115150985e-05,
+      "loss": 0.4256,
       "step": 35950
     },
     {
-      "epoch": 1.26,
-      "learning_rate": 4.376209149779353e-05,
-      "loss": 0.2679,
+      "epoch": 1.2958157638663639,
+      "grad_norm": 0.17523658275604248,
+      "learning_rate": 4.3429939924227806e-05,
+      "loss": 0.4552,
       "step": 35955
     },
     {
-      "epoch": 1.27,
-      "learning_rate": 4.37602086873985e-05,
-      "loss": 0.294,
+      "epoch": 1.2959959635275886,
+      "grad_norm": 0.15205228328704834,
+      "learning_rate": 4.3427968082252744e-05,
+      "loss": 0.4318,
       "step": 35960
     },
     {
-      "epoch": 1.27,
-      "learning_rate": 4.375832563341451e-05,
-      "loss": 0.3024,
+      "epoch": 1.2961761631888131,
+      "grad_norm": 0.18132284283638,
+      "learning_rate": 4.3425995989200184e-05,
+      "loss": 0.4419,
       "step": 35965
     },
     {
-      "epoch": 1.27,
-      "learning_rate": 4.3756442335866034e-05,
-      "loss": 0.267,
+      "epoch": 1.2963563628500379,
+      "grad_norm": 0.16100190579891205,
+      "learning_rate": 4.3424023645097e-05,
+      "loss": 0.4099,
       "step": 35970
     },
     {
-      "epoch": 1.27,
-      "learning_rate": 4.3754558794777514e-05,
-      "loss": 0.291,
+      "epoch": 1.2965365625112626,
+      "grad_norm": 0.22151976823806763,
+      "learning_rate": 4.342205104997006e-05,
+      "loss": 0.456,
       "step": 35975
     },
     {
-      "epoch": 1.27,
-      "learning_rate": 4.37526750101734e-05,
-      "loss": 0.2738,
+      "epoch": 1.296716762172487,
+      "grad_norm": 0.19415588676929474,
+      "learning_rate": 4.3420078203846245e-05,
+      "loss": 0.4543,
       "step": 35980
     },
     {
-      "epoch": 1.27,
-      "learning_rate": 4.375079098207816e-05,
-      "loss": 0.2847,
+      "epoch": 1.2968969618337116,
+      "grad_norm": 0.18258939683437347,
+      "learning_rate": 4.341810510675243e-05,
+      "loss": 0.429,
       "step": 35985
     },
     {
-      "epoch": 1.27,
-      "learning_rate": 4.374890671051626e-05,
-      "loss": 0.2949,
+      "epoch": 1.2970771614949363,
+      "grad_norm": 0.20076502859592438,
+      "learning_rate": 4.3416131758715496e-05,
+      "loss": 0.4105,
       "step": 35990
     },
     {
-      "epoch": 1.27,
-      "learning_rate": 4.3747022195512165e-05,
-      "loss": 0.3078,
+      "epoch": 1.297257361156161,
+      "grad_norm": 0.1802350878715515,
+      "learning_rate": 4.3414158159762334e-05,
+      "loss": 0.4423,
       "step": 35995
     },
     {
-      "epoch": 1.27,
-      "learning_rate": 4.374513743709034e-05,
-      "loss": 0.2833,
+      "epoch": 1.2974375608173856,
+      "grad_norm": 0.2174658179283142,
+      "learning_rate": 4.341218430991982e-05,
+      "loss": 0.4242,
       "step": 36000
     },
     {
-      "epoch": 1.27,
-      "eval_loss": 0.2862606644630432,
-      "eval_runtime": 10.5286,
-      "eval_samples_per_second": 9.498,
-      "eval_steps_per_second": 9.498,
+      "epoch": 1.2974375608173856,
+      "eval_loss": 0.4492260217666626,
+      "eval_runtime": 3.5485,
+      "eval_samples_per_second": 28.181,
+      "eval_steps_per_second": 7.045,
       "step": 36000
     },
     {
-      "epoch": 1.27,
-      "learning_rate": 4.374325243527526e-05,
-      "loss": 0.2925,
+      "epoch": 1.2976177604786103,
+      "grad_norm": 0.1733621507883072,
+      "learning_rate": 4.3410210209214875e-05,
+      "loss": 0.4091,
       "step": 36005
     },
     {
-      "epoch": 1.27,
-      "learning_rate": 4.3741367190091396e-05,
-      "loss": 0.2695,
+      "epoch": 1.2977979601398348,
+      "grad_norm": 0.20660637319087982,
+      "learning_rate": 4.3408235857674376e-05,
+      "loss": 0.4355,
       "step": 36010
     },
     {
-      "epoch": 1.27,
-      "learning_rate": 4.3739481701563236e-05,
-      "loss": 0.2741,
+      "epoch": 1.2979781598010596,
+      "grad_norm": 0.16919872164726257,
+      "learning_rate": 4.340626125532522e-05,
+      "loss": 0.4069,
       "step": 36015
     },
     {
-      "epoch": 1.27,
-      "learning_rate": 4.3737595969715254e-05,
-      "loss": 0.3076,
+      "epoch": 1.2981583594622843,
+      "grad_norm": 0.19883739948272705,
+      "learning_rate": 4.3404286402194326e-05,
+      "loss": 0.4177,
       "step": 36020
     },
     {
-      "epoch": 1.27,
-      "learning_rate": 4.3735709994571924e-05,
-      "loss": 0.2953,
+      "epoch": 1.2983385591235088,
+      "grad_norm": 0.13305337727069855,
+      "learning_rate": 4.340231129830859e-05,
+      "loss": 0.4168,
       "step": 36025
     },
     {
-      "epoch": 1.27,
-      "learning_rate": 4.3733823776157764e-05,
-      "loss": 0.2976,
+      "epoch": 1.2985187587847336,
+      "grad_norm": 0.19353942573070526,
+      "learning_rate": 4.3400335943694925e-05,
+      "loss": 0.3978,
       "step": 36030
     },
     {
-      "epoch": 1.27,
-      "learning_rate": 4.373193731449724e-05,
-      "loss": 0.3005,
+      "epoch": 1.298698958445958,
+      "grad_norm": 0.2054118812084198,
+      "learning_rate": 4.339836033838025e-05,
+      "loss": 0.4347,
       "step": 36035
     },
     {
-      "epoch": 1.27,
-      "learning_rate": 4.373005060961486e-05,
-      "loss": 0.3054,
+      "epoch": 1.2988791581071828,
+      "grad_norm": 0.16488704085350037,
+      "learning_rate": 4.339638448239147e-05,
+      "loss": 0.3811,
       "step": 36040
     },
     {
-      "epoch": 1.27,
-      "learning_rate": 4.372816366153511e-05,
-      "loss": 0.288,
+      "epoch": 1.2990593577684075,
+      "grad_norm": 0.1690049171447754,
+      "learning_rate": 4.3394408375755526e-05,
+      "loss": 0.4393,
       "step": 36045
     },
     {
-      "epoch": 1.27,
-      "learning_rate": 4.37262764702825e-05,
-      "loss": 0.2644,
+      "epoch": 1.299239557429632,
+      "grad_norm": 0.21596601605415344,
+      "learning_rate": 4.339243201849932e-05,
+      "loss": 0.4089,
       "step": 36050
     },
     {
-      "epoch": 1.27,
-      "learning_rate": 4.372438903588153e-05,
-      "loss": 0.3041,
+      "epoch": 1.2994197570908566,
+      "grad_norm": 0.17720189690589905,
+      "learning_rate": 4.339045541064978e-05,
+      "loss": 0.4117,
       "step": 36055
     },
     {
-      "epoch": 1.27,
-      "learning_rate": 4.3722501358356705e-05,
-      "loss": 0.2938,
+      "epoch": 1.2995999567520813,
+      "grad_norm": 0.16456130146980286,
+      "learning_rate": 4.3388478552233856e-05,
+      "loss": 0.3757,
       "step": 36060
     },
     {
-      "epoch": 1.27,
-      "learning_rate": 4.372061343773255e-05,
-      "loss": 0.2867,
+      "epoch": 1.299780156413306,
+      "grad_norm": 0.1337253451347351,
+      "learning_rate": 4.338650144327847e-05,
+      "loss": 0.4465,
       "step": 36065
     },
     {
-      "epoch": 1.27,
-      "learning_rate": 4.3718725274033556e-05,
-      "loss": 0.2751,
+      "epoch": 1.2999603560745305,
+      "grad_norm": 0.17448823153972626,
+      "learning_rate": 4.338452408381056e-05,
+      "loss": 0.4445,
       "step": 36070
     },
     {
-      "epoch": 1.27,
-      "learning_rate": 4.371683686728426e-05,
-      "loss": 0.3224,
+      "epoch": 1.3001405557357553,
+      "grad_norm": 0.1743369847536087,
+      "learning_rate": 4.338254647385708e-05,
+      "loss": 0.4016,
       "step": 36075
     },
     {
-      "epoch": 1.27,
-      "learning_rate": 4.371494821750916e-05,
-      "loss": 0.2743,
+      "epoch": 1.3003207553969798,
+      "grad_norm": 0.2010771483182907,
+      "learning_rate": 4.338056861344495e-05,
+      "loss": 0.4215,
       "step": 36080
     },
     {
-      "epoch": 1.27,
-      "learning_rate": 4.3713059324732796e-05,
-      "loss": 0.3075,
+      "epoch": 1.3005009550582045,
+      "grad_norm": 0.21414510905742645,
+      "learning_rate": 4.337859050260113e-05,
+      "loss": 0.4478,
       "step": 36085
     },
     {
-      "epoch": 1.27,
-      "learning_rate": 4.371117018897969e-05,
-      "loss": 0.2818,
+      "epoch": 1.3006811547194292,
+      "grad_norm": 0.20338131487369537,
+      "learning_rate": 4.337661214135258e-05,
+      "loss": 0.4265,
       "step": 36090
     },
     {
-      "epoch": 1.27,
-      "learning_rate": 4.370928081027437e-05,
-      "loss": 0.2828,
+      "epoch": 1.3008613543806538,
+      "grad_norm": 0.17728644609451294,
+      "learning_rate": 4.3374633529726247e-05,
+      "loss": 0.4039,
       "step": 36095
     },
     {
-      "epoch": 1.27,
-      "learning_rate": 4.370739118864137e-05,
-      "loss": 0.3118,
+      "epoch": 1.3010415540418783,
+      "grad_norm": 0.13623587787151337,
+      "learning_rate": 4.3372654667749086e-05,
+      "loss": 0.3892,
       "step": 36100
     },
     {
-      "epoch": 1.27,
-      "learning_rate": 4.3705501324105224e-05,
-      "loss": 0.2779,
+      "epoch": 1.301221753703103,
+      "grad_norm": 0.20998680591583252,
+      "learning_rate": 4.337067555544806e-05,
+      "loss": 0.4134,
       "step": 36105
     },
     {
-      "epoch": 1.27,
-      "learning_rate": 4.370361121669047e-05,
-      "loss": 0.2905,
+      "epoch": 1.3014019533643277,
+      "grad_norm": 0.1900775134563446,
+      "learning_rate": 4.336869619285014e-05,
+      "loss": 0.4382,
       "step": 36110
     },
     {
-      "epoch": 1.27,
-      "learning_rate": 4.370172086642165e-05,
-      "loss": 0.3131,
+      "epoch": 1.3015821530255522,
+      "grad_norm": 0.17663796246051788,
+      "learning_rate": 4.33667165799823e-05,
+      "loss": 0.4058,
       "step": 36115
     },
     {
-      "epoch": 1.27,
-      "learning_rate": 4.369983027332331e-05,
-      "loss": 0.2999,
+      "epoch": 1.301762352686777,
+      "grad_norm": 0.17739155888557434,
+      "learning_rate": 4.336473671687149e-05,
+      "loss": 0.3889,
       "step": 36120
     },
     {
-      "epoch": 1.27,
-      "learning_rate": 4.369793943742e-05,
-      "loss": 0.3009,
+      "epoch": 1.3019425523480015,
+      "grad_norm": 0.204684317111969,
+      "learning_rate": 4.33627566035447e-05,
+      "loss": 0.435,
       "step": 36125
     },
     {
-      "epoch": 1.27,
-      "learning_rate": 4.369604835873626e-05,
-      "loss": 0.281,
+      "epoch": 1.3021227520092262,
+      "grad_norm": 0.17078737914562225,
+      "learning_rate": 4.336077624002891e-05,
+      "loss": 0.3845,
       "step": 36130
     },
     {
-      "epoch": 1.27,
-      "learning_rate": 4.3694157037296665e-05,
-      "loss": 0.2928,
+      "epoch": 1.302302951670451,
+      "grad_norm": 0.20431089401245117,
+      "learning_rate": 4.33587956263511e-05,
+      "loss": 0.3966,
       "step": 36135
     },
     {
-      "epoch": 1.27,
-      "learning_rate": 4.3692265473125756e-05,
-      "loss": 0.3069,
+      "epoch": 1.3024831513316755,
+      "grad_norm": 0.18561053276062012,
+      "learning_rate": 4.335681476253824e-05,
+      "loss": 0.4092,
       "step": 36140
     },
     {
-      "epoch": 1.27,
-      "learning_rate": 4.3690373666248094e-05,
-      "loss": 0.3013,
+      "epoch": 1.3026633509929002,
+      "grad_norm": 0.1851317286491394,
+      "learning_rate": 4.335483364861734e-05,
+      "loss": 0.4035,
       "step": 36145
     },
     {
-      "epoch": 1.27,
-      "learning_rate": 4.3688481616688255e-05,
-      "loss": 0.2989,
+      "epoch": 1.3028435506541247,
+      "grad_norm": 0.18310926854610443,
+      "learning_rate": 4.3352852284615395e-05,
+      "loss": 0.4261,
       "step": 36150
     },
     {
-      "epoch": 1.27,
-      "learning_rate": 4.3686589324470796e-05,
-      "loss": 0.2663,
+      "epoch": 1.3030237503153494,
+      "grad_norm": 0.19984884560108185,
+      "learning_rate": 4.335087067055938e-05,
+      "loss": 0.4395,
       "step": 36155
     },
     {
-      "epoch": 1.27,
-      "learning_rate": 4.368469678962028e-05,
-      "loss": 0.2916,
+      "epoch": 1.303203949976574,
+      "grad_norm": 0.17451319098472595,
+      "learning_rate": 4.334888880647631e-05,
+      "loss": 0.4466,
       "step": 36160
     },
     {
-      "epoch": 1.27,
-      "learning_rate": 4.3682804012161305e-05,
-      "loss": 0.2982,
+      "epoch": 1.3033841496377987,
+      "grad_norm": 0.18329133093357086,
+      "learning_rate": 4.3346906692393184e-05,
+      "loss": 0.4218,
       "step": 36165
     },
     {
-      "epoch": 1.27,
-      "learning_rate": 4.368091099211843e-05,
-      "loss": 0.3024,
+      "epoch": 1.3035643492990232,
+      "grad_norm": 0.19069811701774597,
+      "learning_rate": 4.3344924328337e-05,
+      "loss": 0.4008,
       "step": 36170
     },
     {
-      "epoch": 1.27,
-      "learning_rate": 4.3679017729516226e-05,
-      "loss": 0.2766,
+      "epoch": 1.303744548960248,
+      "grad_norm": 0.18040628731250763,
+      "learning_rate": 4.334294171433478e-05,
+      "loss": 0.3912,
       "step": 36175
     },
     {
-      "epoch": 1.27,
-      "learning_rate": 4.367712422437929e-05,
-      "loss": 0.3051,
+      "epoch": 1.3039247486214727,
+      "grad_norm": 0.20285740494728088,
+      "learning_rate": 4.3340958850413526e-05,
+      "loss": 0.4664,
       "step": 36180
     },
     {
-      "epoch": 1.27,
-      "learning_rate": 4.3675230476732206e-05,
-      "loss": 0.2863,
+      "epoch": 1.3041049482826972,
+      "grad_norm": 0.2134329378604889,
+      "learning_rate": 4.3338975736600266e-05,
+      "loss": 0.4236,
       "step": 36185
     },
     {
-      "epoch": 1.27,
-      "learning_rate": 4.367333648659956e-05,
-      "loss": 0.3048,
+      "epoch": 1.304285147943922,
+      "grad_norm": 0.18839861452579498,
+      "learning_rate": 4.3336992372922e-05,
+      "loss": 0.4059,
       "step": 36190
     },
     {
-      "epoch": 1.27,
-      "learning_rate": 4.3671442254005955e-05,
-      "loss": 0.2668,
+      "epoch": 1.3044653476051464,
+      "grad_norm": 0.22417490184307098,
+      "learning_rate": 4.333500875940577e-05,
+      "loss": 0.4422,
       "step": 36195
     },
     {
-      "epoch": 1.27,
-      "learning_rate": 4.366954777897597e-05,
-      "loss": 0.2889,
+      "epoch": 1.3046455472663712,
+      "grad_norm": 0.1869698464870453,
+      "learning_rate": 4.33330248960786e-05,
+      "loss": 0.4479,
       "step": 36200
     },
     {
-      "epoch": 1.27,
-      "learning_rate": 4.366765306153421e-05,
-      "loss": 0.2853,
+      "epoch": 1.304825746927596,
+      "grad_norm": 0.14332622289657593,
+      "learning_rate": 4.33310407829675e-05,
+      "loss": 0.3966,
       "step": 36205
     },
     {
-      "epoch": 1.27,
-      "learning_rate": 4.366575810170528e-05,
-      "loss": 0.2799,
+      "epoch": 1.3050059465888204,
+      "grad_norm": 0.23910625278949738,
+      "learning_rate": 4.3329056420099534e-05,
+      "loss": 0.409,
       "step": 36210
     },
     {
-      "epoch": 1.27,
-      "learning_rate": 4.3663862899513784e-05,
-      "loss": 0.312,
+      "epoch": 1.305186146250045,
+      "grad_norm": 0.178507000207901,
+      "learning_rate": 4.332707180750172e-05,
+      "loss": 0.4216,
       "step": 36215
     },
     {
-      "epoch": 1.27,
-      "learning_rate": 4.366196745498432e-05,
-      "loss": 0.2744,
+      "epoch": 1.3053663459112697,
+      "grad_norm": 0.1948593556880951,
+      "learning_rate": 4.3325086945201096e-05,
+      "loss": 0.3968,
       "step": 36220
     },
     {
-      "epoch": 1.27,
-      "learning_rate": 4.366007176814152e-05,
-      "loss": 0.3083,
+      "epoch": 1.3055465455724944,
+      "grad_norm": 0.2139938473701477,
+      "learning_rate": 4.332310183322471e-05,
+      "loss": 0.4525,
       "step": 36225
     },
     {
-      "epoch": 1.27,
-      "learning_rate": 4.365817583900998e-05,
-      "loss": 0.3109,
+      "epoch": 1.305726745233719,
+      "grad_norm": 0.1733957827091217,
+      "learning_rate": 4.332111647159962e-05,
+      "loss": 0.4319,
       "step": 36230
     },
     {
-      "epoch": 1.27,
-      "learning_rate": 4.365627966761432e-05,
-      "loss": 0.2955,
+      "epoch": 1.3059069448949436,
+      "grad_norm": 0.18427705764770508,
+      "learning_rate": 4.331913086035285e-05,
+      "loss": 0.4355,
       "step": 36235
     },
     {
-      "epoch": 1.28,
-      "learning_rate": 4.3654383253979163e-05,
-      "loss": 0.2991,
+      "epoch": 1.3060871445561681,
+      "grad_norm": 0.17030684649944305,
+      "learning_rate": 4.3317144999511474e-05,
+      "loss": 0.4461,
       "step": 36240
     },
     {
-      "epoch": 1.28,
-      "learning_rate": 4.365248659812914e-05,
-      "loss": 0.2743,
+      "epoch": 1.3062673442173929,
+      "grad_norm": 0.17609108984470367,
+      "learning_rate": 4.3315158889102546e-05,
+      "loss": 0.4154,
       "step": 36245
     },
     {
-      "epoch": 1.28,
-      "learning_rate": 4.365058970008886e-05,
-      "loss": 0.2983,
+      "epoch": 1.3064475438786176,
+      "grad_norm": 0.22373346984386444,
+      "learning_rate": 4.3313172529153124e-05,
+      "loss": 0.394,
       "step": 36250
     },
     {
-      "epoch": 1.28,
-      "learning_rate": 4.364869255988298e-05,
-      "loss": 0.2837,
+      "epoch": 1.3066277435398421,
+      "grad_norm": 0.18238912522792816,
+      "learning_rate": 4.331118591969027e-05,
+      "loss": 0.4212,
       "step": 36255
     },
     {
-      "epoch": 1.28,
-      "learning_rate": 4.36467951775361e-05,
-      "loss": 0.2954,
+      "epoch": 1.3068079432010669,
+      "grad_norm": 0.17490154504776,
+      "learning_rate": 4.330919906074106e-05,
+      "loss": 0.3987,
       "step": 36260
     },
     {
-      "epoch": 1.28,
-      "learning_rate": 4.3644897553072885e-05,
-      "loss": 0.3065,
+      "epoch": 1.3069881428622914,
+      "grad_norm": 0.17882512509822845,
+      "learning_rate": 4.330721195233255e-05,
+      "loss": 0.4248,
       "step": 36265
     },
     {
-      "epoch": 1.28,
-      "learning_rate": 4.364299968651795e-05,
-      "loss": 0.2985,
+      "epoch": 1.307168342523516,
+      "grad_norm": 0.21549130976200104,
+      "learning_rate": 4.330522459449182e-05,
+      "loss": 0.4354,
       "step": 36270
     },
     {
-      "epoch": 1.28,
-      "learning_rate": 4.364110157789596e-05,
-      "loss": 0.2885,
+      "epoch": 1.3073485421847406,
+      "grad_norm": 0.19337189197540283,
+      "learning_rate": 4.330323698724596e-05,
+      "loss": 0.4282,
       "step": 36275
     },
     {
-      "epoch": 1.28,
-      "learning_rate": 4.363920322723155e-05,
-      "loss": 0.2916,
+      "epoch": 1.3075287418459653,
+      "grad_norm": 0.16519051790237427,
+      "learning_rate": 4.330124913062203e-05,
+      "loss": 0.4204,
       "step": 36280
     },
     {
-      "epoch": 1.28,
-      "learning_rate": 4.363730463454937e-05,
-      "loss": 0.301,
+      "epoch": 1.3077089415071899,
+      "grad_norm": 0.18474717438220978,
+      "learning_rate": 4.329926102464712e-05,
+      "loss": 0.4295,
       "step": 36285
     },
     {
-      "epoch": 1.28,
-      "learning_rate": 4.3635405799874076e-05,
-      "loss": 0.2864,
+      "epoch": 1.3078891411684146,
+      "grad_norm": 0.21671119332313538,
+      "learning_rate": 4.3297272669348325e-05,
+      "loss": 0.4407,
       "step": 36290
     },
     {
-      "epoch": 1.28,
-      "learning_rate": 4.363350672323031e-05,
-      "loss": 0.303,
+      "epoch": 1.3080693408296393,
+      "grad_norm": Infinity,
+      "learning_rate": 4.32956818056143e-05,
+      "loss": 0.4212,
       "step": 36295
     },
     {
-      "epoch": 1.28,
-      "learning_rate": 4.3631607404642747e-05,
-      "loss": 0.2987,
+      "epoch": 1.3082495404908638,
+      "grad_norm": 0.16523273289203644,
+      "learning_rate": 4.329369300160078e-05,
+      "loss": 0.4094,
       "step": 36300
     },
     {
-      "epoch": 1.28,
-      "learning_rate": 4.362970784413604e-05,
-      "loss": 0.2927,
+      "epoch": 1.3084297401520886,
+      "grad_norm": 0.18672731518745422,
+      "learning_rate": 4.329170394833923e-05,
+      "loss": 0.4582,
       "step": 36305
     },
     {
-      "epoch": 1.28,
-      "learning_rate": 4.362780804173484e-05,
-      "loss": 0.2928,
+      "epoch": 1.308609939813313,
+      "grad_norm": 0.19107048213481903,
+      "learning_rate": 4.328971464585676e-05,
+      "loss": 0.3931,
       "step": 36310
     },
     {
-      "epoch": 1.28,
-      "learning_rate": 4.362590799746384e-05,
-      "loss": 0.2916,
+      "epoch": 1.3087901394745378,
+      "grad_norm": 0.17582692205905914,
+      "learning_rate": 4.3287725094180466e-05,
+      "loss": 0.3489,
       "step": 36315
     },
     {
-      "epoch": 1.28,
-      "learning_rate": 4.36240077113477e-05,
-      "loss": 0.2699,
+      "epoch": 1.3089703391357626,
+      "grad_norm": 0.20227184891700745,
+      "learning_rate": 4.328573529333746e-05,
+      "loss": 0.4136,
       "step": 36320
     },
     {
-      "epoch": 1.28,
-      "learning_rate": 4.3622107183411096e-05,
-      "loss": 0.2861,
+      "epoch": 1.309150538796987,
+      "grad_norm": 0.17152424156665802,
+      "learning_rate": 4.328374524335485e-05,
+      "loss": 0.4413,
       "step": 36325
     },
     {
-      "epoch": 1.28,
-      "learning_rate": 4.36202064136787e-05,
-      "loss": 0.2675,
+      "epoch": 1.3093307384582116,
+      "grad_norm": 0.20434246957302094,
+      "learning_rate": 4.328175494425975e-05,
+      "loss": 0.4357,
       "step": 36330
     },
     {
-      "epoch": 1.28,
-      "learning_rate": 4.3618305402175196e-05,
-      "loss": 0.2824,
+      "epoch": 1.3095109381194363,
+      "grad_norm": 0.18232357501983643,
+      "learning_rate": 4.327976439607928e-05,
+      "loss": 0.4028,
       "step": 36335
     },
     {
-      "epoch": 1.28,
-      "learning_rate": 4.361640414892526e-05,
-      "loss": 0.2912,
+      "epoch": 1.309691137780661,
+      "grad_norm": 0.1636376529932022,
+      "learning_rate": 4.327777359884056e-05,
+      "loss": 0.3889,
       "step": 36340
     },
     {
-      "epoch": 1.28,
-      "learning_rate": 4.3614502653953596e-05,
-      "loss": 0.2873,
+      "epoch": 1.3098713374418856,
+      "grad_norm": 0.21091686189174652,
+      "learning_rate": 4.327578255257071e-05,
+      "loss": 0.4193,
       "step": 36345
     },
     {
-      "epoch": 1.28,
-      "learning_rate": 4.3612600917284874e-05,
-      "loss": 0.2911,
+      "epoch": 1.3100515371031103,
+      "grad_norm": 0.2029031366109848,
+      "learning_rate": 4.327379125729687e-05,
+      "loss": 0.4471,
       "step": 36350
     },
     {
-      "epoch": 1.28,
-      "learning_rate": 4.3610698938943795e-05,
-      "loss": 0.2893,
+      "epoch": 1.3102317367643348,
+      "grad_norm": 0.17957919836044312,
+      "learning_rate": 4.327179971304615e-05,
+      "loss": 0.384,
       "step": 36355
     },
     {
-      "epoch": 1.28,
-      "learning_rate": 4.360879671895506e-05,
-      "loss": 0.3058,
+      "epoch": 1.3104119364255595,
+      "grad_norm": 0.2107733190059662,
+      "learning_rate": 4.3269807919845705e-05,
+      "loss": 0.3978,
       "step": 36360
     },
     {
-      "epoch": 1.28,
-      "learning_rate": 4.3606894257343356e-05,
-      "loss": 0.2766,
+      "epoch": 1.3105921360867843,
+      "grad_norm": 0.16910696029663086,
+      "learning_rate": 4.326781587772266e-05,
+      "loss": 0.4386,
       "step": 36365
     },
     {
-      "epoch": 1.28,
-      "learning_rate": 4.36049915541334e-05,
-      "loss": 0.2626,
+      "epoch": 1.3107723357480088,
+      "grad_norm": 0.18517418205738068,
+      "learning_rate": 4.326582358670416e-05,
+      "loss": 0.3967,
       "step": 36370
     },
     {
-      "epoch": 1.28,
-      "learning_rate": 4.360308860934989e-05,
-      "loss": 0.2879,
+      "epoch": 1.3109525354092335,
+      "grad_norm": 0.21739526093006134,
+      "learning_rate": 4.326383104681735e-05,
+      "loss": 0.4325,
       "step": 36375
     },
     {
-      "epoch": 1.28,
-      "learning_rate": 4.360118542301753e-05,
-      "loss": 0.2785,
+      "epoch": 1.311132735070458,
+      "grad_norm": 0.18001195788383484,
+      "learning_rate": 4.3261838258089384e-05,
+      "loss": 0.4216,
       "step": 36380
     },
     {
-      "epoch": 1.28,
-      "learning_rate": 4.3599281995161035e-05,
-      "loss": 0.2919,
+      "epoch": 1.3113129347316828,
+      "grad_norm": 0.2097054123878479,
+      "learning_rate": 4.32598452205474e-05,
+      "loss": 0.4309,
       "step": 36385
     },
     {
-      "epoch": 1.28,
-      "learning_rate": 4.359737832580512e-05,
-      "loss": 0.2894,
+      "epoch": 1.3114931343929073,
+      "grad_norm": 0.15929162502288818,
+      "learning_rate": 4.325785193421856e-05,
+      "loss": 0.4344,
       "step": 36390
     },
     {
-      "epoch": 1.28,
-      "learning_rate": 4.359547441497451e-05,
-      "loss": 0.2673,
+      "epoch": 1.311673334054132,
+      "grad_norm": 0.17544053494930267,
+      "learning_rate": 4.325585839913003e-05,
+      "loss": 0.4344,
       "step": 36395
     },
     {
-      "epoch": 1.28,
-      "learning_rate": 4.359357026269392e-05,
-      "loss": 0.2778,
+      "epoch": 1.3118535337153565,
+      "grad_norm": 0.2082267850637436,
+      "learning_rate": 4.3253864615308956e-05,
+      "loss": 0.4096,
       "step": 36400
     },
     {
-      "epoch": 1.28,
-      "learning_rate": 4.3591665868988064e-05,
-      "loss": 0.3002,
+      "epoch": 1.3120337333765812,
+      "grad_norm": 0.1683303415775299,
+      "learning_rate": 4.3251870582782516e-05,
+      "loss": 0.4234,
       "step": 36405
     },
     {
-      "epoch": 1.28,
-      "learning_rate": 4.358976123388168e-05,
-      "loss": 0.3183,
+      "epoch": 1.312213933037806,
+      "grad_norm": 0.15545429289340973,
+      "learning_rate": 4.3249876301577877e-05,
+      "loss": 0.3982,
       "step": 36410
     },
     {
-      "epoch": 1.28,
-      "learning_rate": 4.3587856357399504e-05,
-      "loss": 0.276,
+      "epoch": 1.3123941326990305,
+      "grad_norm": 0.22306688129901886,
+      "learning_rate": 4.3247881771722195e-05,
+      "loss": 0.4617,
       "step": 36415
     },
     {
-      "epoch": 1.28,
-      "learning_rate": 4.358595123956626e-05,
-      "loss": 0.3022,
+      "epoch": 1.3125743323602552,
+      "grad_norm": 0.19142894446849823,
+      "learning_rate": 4.3245886993242666e-05,
+      "loss": 0.4333,
       "step": 36420
     },
     {
-      "epoch": 1.28,
-      "learning_rate": 4.358404588040669e-05,
-      "loss": 0.2946,
+      "epoch": 1.3127545320214797,
+      "grad_norm": 0.14523069560527802,
+      "learning_rate": 4.324389196616645e-05,
+      "loss": 0.394,
       "step": 36425
     },
     {
-      "epoch": 1.28,
-      "learning_rate": 4.358214027994553e-05,
-      "loss": 0.2858,
+      "epoch": 1.3129347316827045,
+      "grad_norm": 0.1527920961380005,
+      "learning_rate": 4.3241896690520746e-05,
+      "loss": 0.448,
       "step": 36430
     },
     {
-      "epoch": 1.28,
-      "learning_rate": 4.358023443820752e-05,
-      "loss": 0.3093,
+      "epoch": 1.3131149313439292,
+      "grad_norm": 0.1837022453546524,
+      "learning_rate": 4.323990116633273e-05,
+      "loss": 0.3912,
       "step": 36435
     },
     {
-      "epoch": 1.28,
-      "learning_rate": 4.357832835521742e-05,
-      "loss": 0.2731,
+      "epoch": 1.3132951310051537,
+      "grad_norm": 0.192708358168602,
+      "learning_rate": 4.323790539362958e-05,
+      "loss": 0.4349,
       "step": 36440
     },
     {
-      "epoch": 1.28,
-      "learning_rate": 4.357642203099996e-05,
-      "loss": 0.2918,
+      "epoch": 1.3134753306663782,
+      "grad_norm": 0.16803932189941406,
+      "learning_rate": 4.323590937243852e-05,
+      "loss": 0.4473,
       "step": 36445
     },
     {
-      "epoch": 1.28,
-      "learning_rate": 4.3574515465579915e-05,
-      "loss": 0.3219,
+      "epoch": 1.313655530327603,
+      "grad_norm": 0.18764148652553558,
+      "learning_rate": 4.323391310278672e-05,
+      "loss": 0.4169,
       "step": 36450
     },
     {
-      "epoch": 1.28,
-      "learning_rate": 4.3572608658982016e-05,
-      "loss": 0.2716,
+      "epoch": 1.3138357299888277,
+      "grad_norm": 0.14096736907958984,
+      "learning_rate": 4.3231916584701374e-05,
+      "loss": 0.389,
       "step": 36455
     },
     {
-      "epoch": 1.28,
-      "learning_rate": 4.3570701611231045e-05,
-      "loss": 0.2929,
+      "epoch": 1.3140159296500522,
+      "grad_norm": 0.16323328018188477,
+      "learning_rate": 4.32299198182097e-05,
+      "loss": 0.3844,
       "step": 36460
     },
     {
-      "epoch": 1.28,
-      "learning_rate": 4.356879432235175e-05,
-      "loss": 0.3035,
+      "epoch": 1.314196129311277,
+      "grad_norm": 0.2195003479719162,
+      "learning_rate": 4.32279228033389e-05,
+      "loss": 0.4448,
       "step": 36465
     },
     {
-      "epoch": 1.28,
-      "learning_rate": 4.356688679236889e-05,
-      "loss": 0.2904,
+      "epoch": 1.3143763289725015,
+      "grad_norm": 0.1823636144399643,
+      "learning_rate": 4.3225925540116174e-05,
+      "loss": 0.3907,
       "step": 36470
     },
     {
-      "epoch": 1.28,
-      "learning_rate": 4.356497902130724e-05,
-      "loss": 0.3162,
+      "epoch": 1.3145565286337262,
+      "grad_norm": 0.1974102109670639,
+      "learning_rate": 4.322392802856875e-05,
+      "loss": 0.4337,
       "step": 36475
     },
     {
-      "epoch": 1.28,
-      "learning_rate": 4.356307100919158e-05,
-      "loss": 0.3035,
+      "epoch": 1.314736728294951,
+      "grad_norm": 0.14405225217342377,
+      "learning_rate": 4.3221930268723834e-05,
+      "loss": 0.3845,
       "step": 36480
     },
     {
-      "epoch": 1.28,
-      "learning_rate": 4.356116275604667e-05,
-      "loss": 0.3243,
+      "epoch": 1.3149169279561754,
+      "grad_norm": 0.15730611979961395,
+      "learning_rate": 4.321993226060864e-05,
+      "loss": 0.4251,
       "step": 36485
     },
     {
-      "epoch": 1.28,
-      "learning_rate": 4.35592542618973e-05,
-      "loss": 0.3113,
+      "epoch": 1.3150971276174,
+      "grad_norm": 0.17731639742851257,
+      "learning_rate": 4.3217934004250396e-05,
+      "loss": 0.4128,
       "step": 36490
     },
     {
-      "epoch": 1.28,
-      "learning_rate": 4.355734552676823e-05,
-      "loss": 0.2986,
+      "epoch": 1.3152773272786247,
+      "grad_norm": 0.17964981496334076,
+      "learning_rate": 4.321593549967634e-05,
+      "loss": 0.4303,
       "step": 36495
     },
     {
-      "epoch": 1.28,
-      "learning_rate": 4.355543655068428e-05,
-      "loss": 0.2644,
+      "epoch": 1.3154575269398494,
+      "grad_norm": 0.17006689310073853,
+      "learning_rate": 4.3213936746913675e-05,
+      "loss": 0.418,
       "step": 36500
     },
     {
-      "epoch": 1.28,
-      "eval_loss": 0.28553855419158936,
-      "eval_runtime": 10.5423,
-      "eval_samples_per_second": 9.486,
-      "eval_steps_per_second": 9.486,
+      "epoch": 1.3154575269398494,
+      "eval_loss": 0.4494481384754181,
+      "eval_runtime": 3.5496,
+      "eval_samples_per_second": 28.172,
+      "eval_steps_per_second": 7.043,
       "step": 36500
     },
     {
-      "epoch": 1.28,
-      "learning_rate": 4.35535273336702e-05,
-      "loss": 0.2938,
+      "epoch": 1.315637726601074,
+      "grad_norm": 0.18565762042999268,
+      "learning_rate": 4.321193774598966e-05,
+      "loss": 0.4313,
       "step": 36505
     },
     {
-      "epoch": 1.28,
-      "learning_rate": 4.35516178757508e-05,
-      "loss": 0.2965,
+      "epoch": 1.3158179262622987,
+      "grad_norm": 0.18812930583953857,
+      "learning_rate": 4.3209938496931514e-05,
+      "loss": 0.4287,
       "step": 36510
     },
     {
-      "epoch": 1.28,
-      "learning_rate": 4.3549708176950864e-05,
-      "loss": 0.2701,
+      "epoch": 1.3159981259235232,
+      "grad_norm": 0.21864257752895355,
+      "learning_rate": 4.320793899976648e-05,
+      "loss": 0.4444,
       "step": 36515
     },
     {
-      "epoch": 1.28,
-      "learning_rate": 4.35477982372952e-05,
-      "loss": 0.2727,
+      "epoch": 1.316178325584748,
+      "grad_norm": 0.18846680223941803,
+      "learning_rate": 4.32059392545218e-05,
+      "loss": 0.4351,
       "step": 36520
     },
     {
-      "epoch": 1.29,
-      "learning_rate": 4.3545888056808595e-05,
-      "loss": 0.3053,
+      "epoch": 1.3163585252459726,
+      "grad_norm": 0.2109052538871765,
+      "learning_rate": 4.320393926122473e-05,
+      "loss": 0.3791,
       "step": 36525
     },
     {
-      "epoch": 1.29,
-      "learning_rate": 4.3543977635515856e-05,
-      "loss": 0.2859,
+      "epoch": 1.3165387249071971,
+      "grad_norm": 0.16090500354766846,
+      "learning_rate": 4.320193901990251e-05,
+      "loss": 0.4113,
       "step": 36530
     },
     {
-      "epoch": 1.29,
-      "learning_rate": 4.3542066973441784e-05,
-      "loss": 0.2769,
+      "epoch": 1.3167189245684219,
+      "grad_norm": 0.15768031775951385,
+      "learning_rate": 4.31999385305824e-05,
+      "loss": 0.3952,
       "step": 36535
     },
     {
-      "epoch": 1.29,
-      "learning_rate": 4.3540156070611194e-05,
-      "loss": 0.2967,
+      "epoch": 1.3168991242296464,
+      "grad_norm": 0.18743303418159485,
+      "learning_rate": 4.319793779329163e-05,
+      "loss": 0.4256,
       "step": 36540
     },
     {
-      "epoch": 1.29,
-      "learning_rate": 4.35382449270489e-05,
-      "loss": 0.2793,
+      "epoch": 1.3170793238908711,
+      "grad_norm": 0.2007930725812912,
+      "learning_rate": 4.31959368080575e-05,
+      "loss": 0.4352,
       "step": 36545
     },
     {
-      "epoch": 1.29,
-      "learning_rate": 4.3536333542779704e-05,
-      "loss": 0.3308,
+      "epoch": 1.3172595235520959,
+      "grad_norm": 0.21530933678150177,
+      "learning_rate": 4.319393557490725e-05,
+      "loss": 0.4444,
       "step": 36550
     },
     {
-      "epoch": 1.29,
-      "learning_rate": 4.353442191782844e-05,
-      "loss": 0.2875,
+      "epoch": 1.3174397232133204,
+      "grad_norm": 0.173873171210289,
+      "learning_rate": 4.3191934093868146e-05,
+      "loss": 0.4345,
       "step": 36555
     },
     {
-      "epoch": 1.29,
-      "learning_rate": 4.3532510052219924e-05,
-      "loss": 0.2916,
+      "epoch": 1.3176199228745449,
+      "grad_norm": 0.17242655158042908,
+      "learning_rate": 4.318993236496747e-05,
+      "loss": 0.3883,
       "step": 36560
     },
     {
-      "epoch": 1.29,
-      "learning_rate": 4.353059794597898e-05,
-      "loss": 0.3129,
+      "epoch": 1.3178001225357696,
+      "grad_norm": 0.14753524959087372,
+      "learning_rate": 4.318793038823248e-05,
+      "loss": 0.4319,
       "step": 36565
     },
     {
-      "epoch": 1.29,
-      "learning_rate": 4.3528685599130427e-05,
-      "loss": 0.2945,
+      "epoch": 1.3179803221969943,
+      "grad_norm": 0.18761220574378967,
+      "learning_rate": 4.318592816369046e-05,
+      "loss": 0.4511,
       "step": 36570
     },
     {
-      "epoch": 1.29,
-      "learning_rate": 4.35267730116991e-05,
-      "loss": 0.3076,
+      "epoch": 1.3181605218582189,
+      "grad_norm": 0.15937355160713196,
+      "learning_rate": 4.3183925691368695e-05,
+      "loss": 0.3816,
       "step": 36575
     },
     {
-      "epoch": 1.29,
-      "learning_rate": 4.3524860183709836e-05,
-      "loss": 0.2925,
+      "epoch": 1.3183407215194436,
+      "grad_norm": 0.18989646434783936,
+      "learning_rate": 4.318192297129446e-05,
+      "loss": 0.3873,
       "step": 36580
     },
     {
-      "epoch": 1.29,
-      "learning_rate": 4.352294711518748e-05,
-      "loss": 0.3183,
+      "epoch": 1.318520921180668,
+      "grad_norm": 0.1784542053937912,
+      "learning_rate": 4.3179920003495045e-05,
+      "loss": 0.4507,
       "step": 36585
     },
     {
-      "epoch": 1.29,
-      "learning_rate": 4.352103380615686e-05,
-      "loss": 0.3042,
+      "epoch": 1.3187011208418928,
+      "grad_norm": 0.162387415766716,
+      "learning_rate": 4.3177916787997735e-05,
+      "loss": 0.4457,
       "step": 36590
     },
     {
-      "epoch": 1.29,
-      "learning_rate": 4.3519120256642814e-05,
-      "loss": 0.2945,
+      "epoch": 1.3188813205031176,
+      "grad_norm": 0.1939140409231186,
+      "learning_rate": 4.3175913324829834e-05,
+      "loss": 0.377,
       "step": 36595
     },
     {
-      "epoch": 1.29,
-      "learning_rate": 4.35172064666702e-05,
-      "loss": 0.3112,
+      "epoch": 1.319061520164342,
+      "grad_norm": 0.19154348969459534,
+      "learning_rate": 4.317390961401862e-05,
+      "loss": 0.3741,
       "step": 36600
     },
     {
-      "epoch": 1.29,
-      "learning_rate": 4.3515292436263863e-05,
-      "loss": 0.2821,
+      "epoch": 1.3192417198255666,
+      "grad_norm": 0.18536292016506195,
+      "learning_rate": 4.3171905655591425e-05,
+      "loss": 0.3934,
       "step": 36605
     },
     {
-      "epoch": 1.29,
-      "learning_rate": 4.351337816544866e-05,
-      "loss": 0.2871,
+      "epoch": 1.3194219194867913,
+      "grad_norm": 0.1465887874364853,
+      "learning_rate": 4.316990144957553e-05,
+      "loss": 0.4129,
       "step": 36610
     },
     {
-      "epoch": 1.29,
-      "learning_rate": 4.351146365424944e-05,
-      "loss": 0.3181,
+      "epoch": 1.319602119148016,
+      "grad_norm": 0.21748924255371094,
+      "learning_rate": 4.316789699599824e-05,
+      "loss": 0.4058,
       "step": 36615
     },
     {
-      "epoch": 1.29,
-      "learning_rate": 4.350954890269106e-05,
-      "loss": 0.2962,
+      "epoch": 1.3197823188092406,
+      "grad_norm": 0.18273966014385223,
+      "learning_rate": 4.316589229488687e-05,
+      "loss": 0.429,
       "step": 36620
     },
     {
-      "epoch": 1.29,
-      "learning_rate": 4.3507633910798394e-05,
-      "loss": 0.2991,
+      "epoch": 1.3199625184704653,
+      "grad_norm": 0.21654346585273743,
+      "learning_rate": 4.3163887346268735e-05,
+      "loss": 0.4469,
       "step": 36625
     },
     {
-      "epoch": 1.29,
-      "learning_rate": 4.3505718678596294e-05,
-      "loss": 0.3041,
+      "epoch": 1.3201427181316898,
+      "grad_norm": 0.18244554102420807,
+      "learning_rate": 4.316188215017116e-05,
+      "loss": 0.4285,
       "step": 36630
     },
     {
-      "epoch": 1.29,
-      "learning_rate": 4.350380320610963e-05,
-      "loss": 0.301,
+      "epoch": 1.3203229177929146,
+      "grad_norm": 0.20105114579200745,
+      "learning_rate": 4.315987670662145e-05,
+      "loss": 0.4308,
       "step": 36635
     },
     {
-      "epoch": 1.29,
-      "learning_rate": 4.350188749336328e-05,
-      "loss": 0.2916,
+      "epoch": 1.3205031174541393,
+      "grad_norm": 0.20013387501239777,
+      "learning_rate": 4.315787101564693e-05,
+      "loss": 0.4312,
       "step": 36640
     },
     {
-      "epoch": 1.29,
-      "learning_rate": 4.349997154038211e-05,
-      "loss": 0.2933,
+      "epoch": 1.3206833171153638,
+      "grad_norm": 0.20948466658592224,
+      "learning_rate": 4.315586507727494e-05,
+      "loss": 0.4199,
       "step": 36645
     },
     {
-      "epoch": 1.29,
-      "learning_rate": 4.3498055347191005e-05,
-      "loss": 0.2878,
+      "epoch": 1.3208635167765885,
+      "grad_norm": 0.19690115749835968,
+      "learning_rate": 4.3153858891532804e-05,
+      "loss": 0.4365,
       "step": 36650
     },
     {
-      "epoch": 1.29,
-      "learning_rate": 4.3496138913814837e-05,
-      "loss": 0.3054,
+      "epoch": 1.321043716437813,
+      "grad_norm": 0.17856040596961975,
+      "learning_rate": 4.315185245844785e-05,
+      "loss": 0.4259,
       "step": 36655
     },
     {
-      "epoch": 1.29,
-      "learning_rate": 4.34942222402785e-05,
-      "loss": 0.2858,
+      "epoch": 1.3212239160990378,
+      "grad_norm": 0.16999214887619019,
+      "learning_rate": 4.314984577804743e-05,
+      "loss": 0.4338,
       "step": 36660
     },
     {
-      "epoch": 1.29,
-      "learning_rate": 4.3492305326606876e-05,
-      "loss": 0.2823,
+      "epoch": 1.3214041157602623,
+      "grad_norm": 0.1723853200674057,
+      "learning_rate": 4.314783885035887e-05,
+      "loss": 0.4113,
       "step": 36665
     },
     {
-      "epoch": 1.29,
-      "learning_rate": 4.349038817282485e-05,
-      "loss": 0.2597,
+      "epoch": 1.321584315421487,
+      "grad_norm": 0.19773799180984497,
+      "learning_rate": 4.314583167540952e-05,
+      "loss": 0.436,
       "step": 36670
     },
     {
-      "epoch": 1.29,
-      "learning_rate": 4.3488470778957316e-05,
-      "loss": 0.2985,
+      "epoch": 1.3217645150827115,
+      "grad_norm": 0.15514503419399261,
+      "learning_rate": 4.314382425322672e-05,
+      "loss": 0.4356,
       "step": 36675
     },
     {
-      "epoch": 1.29,
-      "learning_rate": 4.348655314502919e-05,
-      "loss": 0.2786,
+      "epoch": 1.3219447147439363,
+      "grad_norm": 0.20653900504112244,
+      "learning_rate": 4.314181658383783e-05,
+      "loss": 0.3912,
       "step": 36680
     },
     {
-      "epoch": 1.29,
-      "learning_rate": 4.348463527106533e-05,
-      "loss": 0.2555,
+      "epoch": 1.322124914405161,
+      "grad_norm": 0.15066716074943542,
+      "learning_rate": 4.313980866727021e-05,
+      "loss": 0.4433,
       "step": 36685
     },
     {
-      "epoch": 1.29,
-      "learning_rate": 4.348271715709068e-05,
-      "loss": 0.2963,
+      "epoch": 1.3223051140663855,
+      "grad_norm": 0.1296568661928177,
+      "learning_rate": 4.313780050355119e-05,
+      "loss": 0.3966,
       "step": 36690
     },
     {
-      "epoch": 1.29,
-      "learning_rate": 4.348079880313013e-05,
-      "loss": 0.2941,
+      "epoch": 1.3224853137276102,
+      "grad_norm": 0.2106575220823288,
+      "learning_rate": 4.313579209270817e-05,
+      "loss": 0.3952,
       "step": 36695
     },
     {
-      "epoch": 1.29,
-      "learning_rate": 4.347888020920858e-05,
-      "loss": 0.2925,
+      "epoch": 1.3226655133888348,
+      "grad_norm": 0.1847856044769287,
+      "learning_rate": 4.313378343476849e-05,
+      "loss": 0.4303,
       "step": 36700
     },
     {
-      "epoch": 1.29,
-      "learning_rate": 4.3476961375350945e-05,
-      "loss": 0.2928,
+      "epoch": 1.3228457130500595,
+      "grad_norm": 0.20785640180110931,
+      "learning_rate": 4.313177452975952e-05,
+      "loss": 0.4227,
       "step": 36705
     },
     {
-      "epoch": 1.29,
-      "learning_rate": 4.3475042301582145e-05,
-      "loss": 0.28,
+      "epoch": 1.3230259127112842,
+      "grad_norm": 0.14709509909152985,
+      "learning_rate": 4.312976537770863e-05,
+      "loss": 0.4353,
       "step": 36710
     },
     {
-      "epoch": 1.29,
-      "learning_rate": 4.347312298792711e-05,
-      "loss": 0.301,
+      "epoch": 1.3232061123725087,
+      "grad_norm": 0.19536839425563812,
+      "learning_rate": 4.312775597864319e-05,
+      "loss": 0.4323,
       "step": 36715
     },
     {
-      "epoch": 1.29,
-      "learning_rate": 4.3471203434410726e-05,
-      "loss": 0.2977,
+      "epoch": 1.3233863120337332,
+      "grad_norm": 0.15768684446811676,
+      "learning_rate": 4.31257463325906e-05,
+      "loss": 0.416,
       "step": 36720
     },
     {
-      "epoch": 1.29,
-      "learning_rate": 4.3469283641057954e-05,
-      "loss": 0.2834,
+      "epoch": 1.323566511694958,
+      "grad_norm": 0.20642277598381042,
+      "learning_rate": 4.312373643957821e-05,
+      "loss": 0.4052,
       "step": 36725
     },
     {
-      "epoch": 1.29,
-      "learning_rate": 4.3467363607893695e-05,
-      "loss": 0.2851,
+      "epoch": 1.3237467113561827,
+      "grad_norm": 0.1643398553133011,
+      "learning_rate": 4.312172629963343e-05,
+      "loss": 0.4166,
       "step": 36730
     },
     {
-      "epoch": 1.29,
-      "learning_rate": 4.34654433349429e-05,
-      "loss": 0.2715,
+      "epoch": 1.3239269110174072,
+      "grad_norm": 0.1844344139099121,
+      "learning_rate": 4.311971591278363e-05,
+      "loss": 0.4033,
       "step": 36735
     },
     {
-      "epoch": 1.29,
-      "learning_rate": 4.346352282223048e-05,
-      "loss": 0.2796,
+      "epoch": 1.324107110678632,
+      "grad_norm": 0.1502874493598938,
+      "learning_rate": 4.311770527905622e-05,
+      "loss": 0.3871,
       "step": 36740
     },
     {
-      "epoch": 1.29,
-      "learning_rate": 4.3461602069781396e-05,
-      "loss": 0.3012,
+      "epoch": 1.3242873103398565,
+      "grad_norm": 0.16832222044467926,
+      "learning_rate": 4.3115694398478574e-05,
+      "loss": 0.4425,
       "step": 36745
     },
     {
-      "epoch": 1.29,
-      "learning_rate": 4.3459681077620576e-05,
-      "loss": 0.2873,
+      "epoch": 1.3244675100010812,
+      "grad_norm": 0.14828276634216309,
+      "learning_rate": 4.31136832710781e-05,
+      "loss": 0.3985,
       "step": 36750
     },
     {
-      "epoch": 1.29,
-      "learning_rate": 4.345775984577295e-05,
-      "loss": 0.2934,
+      "epoch": 1.324647709662306,
+      "grad_norm": 0.20640629529953003,
+      "learning_rate": 4.31116718968822e-05,
+      "loss": 0.4414,
       "step": 36755
     },
     {
-      "epoch": 1.29,
-      "learning_rate": 4.345583837426349e-05,
-      "loss": 0.2894,
+      "epoch": 1.3248279093235305,
+      "grad_norm": 0.1458745151758194,
+      "learning_rate": 4.310966027591828e-05,
+      "loss": 0.4191,
       "step": 36760
     },
     {
-      "epoch": 1.29,
-      "learning_rate": 4.345391666311712e-05,
-      "loss": 0.2864,
+      "epoch": 1.3250081089847552,
+      "grad_norm": 0.2120623141527176,
+      "learning_rate": 4.3107648408213744e-05,
+      "loss": 0.4122,
       "step": 36765
     },
     {
-      "epoch": 1.29,
-      "learning_rate": 4.3451994712358815e-05,
-      "loss": 0.2974,
+      "epoch": 1.3251883086459797,
+      "grad_norm": 0.17051535844802856,
+      "learning_rate": 4.3105636293795995e-05,
+      "loss": 0.4658,
       "step": 36770
     },
     {
-      "epoch": 1.29,
-      "learning_rate": 4.345007252201351e-05,
-      "loss": 0.2784,
+      "epoch": 1.3253685083072044,
+      "grad_norm": 0.19851194322109222,
+      "learning_rate": 4.310362393269247e-05,
+      "loss": 0.4223,
       "step": 36775
     },
     {
-      "epoch": 1.29,
-      "learning_rate": 4.3448150092106185e-05,
-      "loss": 0.3068,
+      "epoch": 1.325548707968429,
+      "grad_norm": 0.2429044246673584,
+      "learning_rate": 4.310161132493057e-05,
+      "loss": 0.4051,
       "step": 36780
     },
     {
-      "epoch": 1.29,
-      "learning_rate": 4.344622742266178e-05,
-      "loss": 0.2844,
+      "epoch": 1.3257289076296537,
+      "grad_norm": 0.21794357895851135,
+      "learning_rate": 4.3099598470537716e-05,
+      "loss": 0.4102,
       "step": 36785
     },
     {
-      "epoch": 1.29,
-      "learning_rate": 4.3444689114656336e-05,
-      "loss": 0.2789,
+      "epoch": 1.3259091072908782,
+      "grad_norm": 0.17098356783390045,
+      "learning_rate": 4.3097585369541336e-05,
+      "loss": 0.4288,
       "step": 36790
     },
     {
-      "epoch": 1.29,
-      "learning_rate": 4.344276601410812e-05,
-      "loss": 0.2819,
+      "epoch": 1.326089306952103,
+      "grad_norm": 0.17869386076927185,
+      "learning_rate": 4.309557202196887e-05,
+      "loss": 0.4128,
       "step": 36795
     },
     {
-      "epoch": 1.29,
-      "learning_rate": 4.3440842674092734e-05,
-      "loss": 0.2916,
+      "epoch": 1.3262695066133277,
+      "grad_norm": 0.17495180666446686,
+      "learning_rate": 4.309355842784773e-05,
+      "loss": 0.4169,
       "step": 36800
     },
     {
-      "epoch": 1.29,
-      "learning_rate": 4.343891909463517e-05,
-      "loss": 0.3131,
+      "epoch": 1.3264497062745522,
+      "grad_norm": 0.15498143434524536,
+      "learning_rate": 4.309154458720536e-05,
+      "loss": 0.4238,
       "step": 36805
     },
     {
-      "epoch": 1.3,
-      "learning_rate": 4.3436995275760376e-05,
-      "loss": 0.2675,
+      "epoch": 1.326629905935777,
+      "grad_norm": 0.17504778504371643,
+      "learning_rate": 4.3089530500069194e-05,
+      "loss": 0.433,
       "step": 36810
     },
     {
-      "epoch": 1.3,
-      "learning_rate": 4.343507121749336e-05,
-      "loss": 0.2875,
+      "epoch": 1.3268101055970014,
+      "grad_norm": 0.1534842997789383,
+      "learning_rate": 4.308751616646668e-05,
+      "loss": 0.4162,
       "step": 36815
     },
     {
-      "epoch": 1.3,
-      "learning_rate": 4.3433146919859094e-05,
-      "loss": 0.2685,
+      "epoch": 1.3269903052582261,
+      "grad_norm": 0.14600083231925964,
+      "learning_rate": 4.308550158642526e-05,
+      "loss": 0.4286,
       "step": 36820
     },
     {
-      "epoch": 1.3,
-      "learning_rate": 4.343122238288257e-05,
-      "loss": 0.2749,
+      "epoch": 1.3271705049194509,
+      "grad_norm": 0.15580976009368896,
+      "learning_rate": 4.3083486759972384e-05,
+      "loss": 0.4622,
       "step": 36825
     },
     {
-      "epoch": 1.3,
-      "learning_rate": 4.3429297606588747e-05,
-      "loss": 0.2847,
+      "epoch": 1.3273507045806754,
+      "grad_norm": 0.14176791906356812,
+      "learning_rate": 4.30814716871355e-05,
+      "loss": 0.4127,
       "step": 36830
     },
     {
-      "epoch": 1.3,
-      "learning_rate": 4.342737259100266e-05,
-      "loss": 0.2959,
+      "epoch": 1.3275309042419,
+      "grad_norm": 0.1750148981809616,
+      "learning_rate": 4.3079456367942065e-05,
+      "loss": 0.4201,
       "step": 36835
     },
     {
-      "epoch": 1.3,
-      "learning_rate": 4.3425447336149275e-05,
-      "loss": 0.2964,
+      "epoch": 1.3277111039031246,
+      "grad_norm": 0.18358172476291656,
+      "learning_rate": 4.3077440802419544e-05,
+      "loss": 0.4328,
       "step": 36840
     },
     {
-      "epoch": 1.3,
-      "learning_rate": 4.3423521842053606e-05,
-      "loss": 0.2817,
+      "epoch": 1.3278913035643494,
+      "grad_norm": 0.20503877103328705,
+      "learning_rate": 4.307542499059538e-05,
+      "loss": 0.4361,
       "step": 36845
     },
     {
-      "epoch": 1.3,
-      "learning_rate": 4.342159610874063e-05,
-      "loss": 0.2893,
+      "epoch": 1.3280715032255739,
+      "grad_norm": 0.20601756870746613,
+      "learning_rate": 4.307340893249706e-05,
+      "loss": 0.4039,
       "step": 36850
     },
     {
-      "epoch": 1.3,
-      "learning_rate": 4.341967013623539e-05,
-      "loss": 0.2597,
+      "epoch": 1.3282517028867986,
+      "grad_norm": 0.1701328456401825,
+      "learning_rate": 4.307139262815204e-05,
+      "loss": 0.3766,
       "step": 36855
     },
     {
-      "epoch": 1.3,
-      "learning_rate": 4.341774392456286e-05,
-      "loss": 0.3009,
+      "epoch": 1.3284319025480231,
+      "grad_norm": 0.1465604156255722,
+      "learning_rate": 4.30693760775878e-05,
+      "loss": 0.4086,
       "step": 36860
     },
     {
-      "epoch": 1.3,
-      "learning_rate": 4.3415817473748074e-05,
-      "loss": 0.2799,
+      "epoch": 1.3286121022092479,
+      "grad_norm": 0.15594859421253204,
+      "learning_rate": 4.3067359280831797e-05,
+      "loss": 0.3908,
       "step": 36865
     },
     {
-      "epoch": 1.3,
-      "learning_rate": 4.3413890783816026e-05,
-      "loss": 0.2736,
+      "epoch": 1.3287923018704726,
+      "grad_norm": 0.15714693069458008,
+      "learning_rate": 4.306534223791153e-05,
+      "loss": 0.4307,
       "step": 36870
     },
     {
-      "epoch": 1.3,
-      "learning_rate": 4.3411963854791745e-05,
-      "loss": 0.3029,
+      "epoch": 1.328972501531697,
+      "grad_norm": 0.17976239323616028,
+      "learning_rate": 4.306332494885446e-05,
+      "loss": 0.4154,
       "step": 36875
     },
     {
-      "epoch": 1.3,
-      "learning_rate": 4.341003668670025e-05,
-      "loss": 0.2777,
+      "epoch": 1.3291527011929218,
+      "grad_norm": 0.17832331359386444,
+      "learning_rate": 4.30613074136881e-05,
+      "loss": 0.4376,
       "step": 36880
     },
     {
-      "epoch": 1.3,
-      "learning_rate": 4.340810927956656e-05,
-      "loss": 0.2807,
+      "epoch": 1.3293329008541463,
+      "grad_norm": 0.15694265067577362,
+      "learning_rate": 4.305928963243992e-05,
+      "loss": 0.4172,
       "step": 36885
     },
     {
-      "epoch": 1.3,
-      "learning_rate": 4.340618163341569e-05,
-      "loss": 0.2852,
+      "epoch": 1.329513100515371,
+      "grad_norm": 0.1645699441432953,
+      "learning_rate": 4.305727160513741e-05,
+      "loss": 0.459,
       "step": 36890
     },
     {
-      "epoch": 1.3,
-      "learning_rate": 4.340425374827269e-05,
-      "loss": 0.2986,
+      "epoch": 1.3296933001765956,
+      "grad_norm": 0.16790169477462769,
+      "learning_rate": 4.305525333180807e-05,
+      "loss": 0.3762,
       "step": 36895
     },
     {
-      "epoch": 1.3,
-      "learning_rate": 4.3402325624162585e-05,
-      "loss": 0.2926,
+      "epoch": 1.3298734998378203,
+      "grad_norm": 0.1775948405265808,
+      "learning_rate": 4.3053234812479406e-05,
+      "loss": 0.3997,
       "step": 36900
     },
     {
-      "epoch": 1.3,
-      "learning_rate": 4.340039726111041e-05,
-      "loss": 0.273,
+      "epoch": 1.3300536994990448,
+      "grad_norm": 0.19268429279327393,
+      "learning_rate": 4.305121604717891e-05,
+      "loss": 0.4283,
       "step": 36905
     },
     {
-      "epoch": 1.3,
-      "learning_rate": 4.3398468659141206e-05,
-      "loss": 0.2762,
+      "epoch": 1.3302338991602696,
+      "grad_norm": 0.1967247873544693,
+      "learning_rate": 4.304919703593409e-05,
+      "loss": 0.4265,
       "step": 36910
     },
     {
-      "epoch": 1.3,
-      "learning_rate": 4.3396539818280004e-05,
-      "loss": 0.2821,
+      "epoch": 1.3304140988214943,
+      "grad_norm": 0.1895369440317154,
+      "learning_rate": 4.304717777877246e-05,
+      "loss": 0.3815,
       "step": 36915
     },
     {
-      "epoch": 1.3,
-      "learning_rate": 4.339461073855186e-05,
-      "loss": 0.3022,
+      "epoch": 1.3305942984827188,
+      "grad_norm": 0.1666320413351059,
+      "learning_rate": 4.304515827572152e-05,
+      "loss": 0.4227,
       "step": 36920
     },
     {
-      "epoch": 1.3,
-      "learning_rate": 4.3392681419981815e-05,
-      "loss": 0.3041,
+      "epoch": 1.3307744981439436,
+      "grad_norm": 0.19722457230091095,
+      "learning_rate": 4.304313852680879e-05,
+      "loss": 0.4094,
       "step": 36925
     },
     {
-      "epoch": 1.3,
-      "learning_rate": 4.339075186259493e-05,
-      "loss": 0.2892,
+      "epoch": 1.330954697805168,
+      "grad_norm": 0.20568078756332397,
+      "learning_rate": 4.3041118532061794e-05,
+      "loss": 0.415,
       "step": 36930
     },
     {
-      "epoch": 1.3,
-      "learning_rate": 4.338882206641625e-05,
-      "loss": 0.2866,
+      "epoch": 1.3311348974663928,
+      "grad_norm": 0.16985160112380981,
+      "learning_rate": 4.303909829150805e-05,
+      "loss": 0.4198,
       "step": 36935
     },
     {
-      "epoch": 1.3,
-      "learning_rate": 4.3386892031470826e-05,
-      "loss": 0.2874,
+      "epoch": 1.3313150971276175,
+      "grad_norm": 0.1945793330669403,
+      "learning_rate": 4.3037077805175085e-05,
+      "loss": 0.4086,
       "step": 36940
     },
     {
-      "epoch": 1.3,
-      "learning_rate": 4.338496175778374e-05,
-      "loss": 0.2866,
+      "epoch": 1.331495296788842,
+      "grad_norm": 0.19523586332798004,
+      "learning_rate": 4.303505707309043e-05,
+      "loss": 0.4424,
       "step": 36945
     },
     {
-      "epoch": 1.3,
-      "learning_rate": 4.338303124538004e-05,
-      "loss": 0.2786,
+      "epoch": 1.3316754964500666,
+      "grad_norm": 0.18730002641677856,
+      "learning_rate": 4.303303609528161e-05,
+      "loss": 0.4097,
       "step": 36950
     },
     {
-      "epoch": 1.3,
-      "learning_rate": 4.338110049428478e-05,
-      "loss": 0.264,
+      "epoch": 1.3318556961112913,
+      "grad_norm": 0.22210468351840973,
+      "learning_rate": 4.303101487177616e-05,
+      "loss": 0.4432,
       "step": 36955
     },
     {
-      "epoch": 1.3,
-      "learning_rate": 4.3379169504523054e-05,
-      "loss": 0.2933,
+      "epoch": 1.332035895772516,
+      "grad_norm": 0.16588768362998962,
+      "learning_rate": 4.3028993402601636e-05,
+      "loss": 0.4628,
       "step": 36960
     },
     {
-      "epoch": 1.3,
-      "learning_rate": 4.337723827611992e-05,
-      "loss": 0.2882,
+      "epoch": 1.3322160954337405,
+      "grad_norm": 0.19048090279102325,
+      "learning_rate": 4.302697168778556e-05,
+      "loss": 0.4337,
       "step": 36965
     },
     {
-      "epoch": 1.3,
-      "learning_rate": 4.3375306809100455e-05,
-      "loss": 0.276,
+      "epoch": 1.3323962950949653,
+      "grad_norm": 0.18421435356140137,
+      "learning_rate": 4.302494972735549e-05,
+      "loss": 0.4059,
       "step": 36970
     },
     {
-      "epoch": 1.3,
-      "learning_rate": 4.337337510348975e-05,
-      "loss": 0.2903,
+      "epoch": 1.3325764947561898,
+      "grad_norm": 0.15311434864997864,
+      "learning_rate": 4.3022927521338965e-05,
+      "loss": 0.3721,
       "step": 36975
     },
     {
-      "epoch": 1.3,
-      "learning_rate": 4.337144315931288e-05,
-      "loss": 0.2877,
+      "epoch": 1.3327566944174145,
+      "grad_norm": 0.1969848871231079,
+      "learning_rate": 4.302090506976354e-05,
+      "loss": 0.4358,
       "step": 36980
     },
     {
-      "epoch": 1.3,
-      "learning_rate": 4.3369510976594917e-05,
-      "loss": 0.2836,
+      "epoch": 1.3329368940786392,
+      "grad_norm": 0.19193395972251892,
+      "learning_rate": 4.301888237265678e-05,
+      "loss": 0.4397,
       "step": 36985
     },
     {
-      "epoch": 1.3,
-      "learning_rate": 4.336757855536097e-05,
-      "loss": 0.3013,
+      "epoch": 1.3331170937398638,
+      "grad_norm": 0.19378533959388733,
+      "learning_rate": 4.301685943004622e-05,
+      "loss": 0.4522,
       "step": 36990
     },
     {
-      "epoch": 1.3,
-      "learning_rate": 4.336564589563611e-05,
-      "loss": 0.2935,
+      "epoch": 1.3332972934010883,
+      "grad_norm": 0.19879907369613647,
+      "learning_rate": 4.301483624195945e-05,
+      "loss": 0.4517,
       "step": 36995
     },
     {
-      "epoch": 1.3,
-      "learning_rate": 4.336371299744545e-05,
-      "loss": 0.295,
+      "epoch": 1.333477493062313,
+      "grad_norm": 0.16110192239284515,
+      "learning_rate": 4.301281280842403e-05,
+      "loss": 0.4311,
       "step": 37000
     },
     {
-      "epoch": 1.3,
-      "eval_loss": 0.2844620645046234,
-      "eval_runtime": 10.5304,
-      "eval_samples_per_second": 9.496,
-      "eval_steps_per_second": 9.496,
+      "epoch": 1.333477493062313,
+      "eval_loss": 0.4481649100780487,
+      "eval_runtime": 3.5338,
+      "eval_samples_per_second": 28.298,
+      "eval_steps_per_second": 7.075,
       "step": 37000
     },
     {
-      "epoch": 1.3,
-      "learning_rate": 4.336177986081407e-05,
-      "loss": 0.3063,
+      "epoch": 1.3336576927235377,
+      "grad_norm": 0.1890387237071991,
+      "learning_rate": 4.301078912946751e-05,
+      "loss": 0.4437,
       "step": 37005
     },
     {
-      "epoch": 1.3,
-      "learning_rate": 4.3359846485767083e-05,
-      "loss": 0.2739,
+      "epoch": 1.3338378923847622,
+      "grad_norm": 0.2009790688753128,
+      "learning_rate": 4.300876520511748e-05,
+      "loss": 0.4314,
       "step": 37010
     },
     {
-      "epoch": 1.3,
-      "learning_rate": 4.33579128723296e-05,
-      "loss": 0.2798,
+      "epoch": 1.334018092045987,
+      "grad_norm": 0.23660525679588318,
+      "learning_rate": 4.300674103540151e-05,
+      "loss": 0.4406,
       "step": 37015
     },
     {
-      "epoch": 1.3,
-      "learning_rate": 4.3355979020526704e-05,
-      "loss": 0.3129,
+      "epoch": 1.3341982917072115,
+      "grad_norm": 0.20665594935417175,
+      "learning_rate": 4.300471662034719e-05,
+      "loss": 0.3956,
       "step": 37020
     },
     {
-      "epoch": 1.3,
-      "learning_rate": 4.335404493038352e-05,
-      "loss": 0.2926,
+      "epoch": 1.3343784913684362,
+      "grad_norm": 0.19279490411281586,
+      "learning_rate": 4.3002691959982076e-05,
+      "loss": 0.4062,
       "step": 37025
     },
     {
-      "epoch": 1.3,
-      "learning_rate": 4.3352110601925155e-05,
-      "loss": 0.3111,
+      "epoch": 1.334558691029661,
+      "grad_norm": 0.20935142040252686,
+      "learning_rate": 4.3000667054333775e-05,
+      "loss": 0.4324,
       "step": 37030
     },
     {
-      "epoch": 1.3,
-      "learning_rate": 4.3350176035176734e-05,
-      "loss": 0.3063,
+      "epoch": 1.3347388906908855,
+      "grad_norm": 0.1611267477273941,
+      "learning_rate": 4.2998641903429875e-05,
+      "loss": 0.4183,
       "step": 37035
     },
     {
-      "epoch": 1.3,
-      "learning_rate": 4.3348241230163367e-05,
-      "loss": 0.2934,
+      "epoch": 1.3349190903521102,
+      "grad_norm": 0.2519780993461609,
+      "learning_rate": 4.299661650729796e-05,
+      "loss": 0.447,
       "step": 37040
     },
     {
-      "epoch": 1.3,
-      "learning_rate": 4.3346306186910185e-05,
-      "loss": 0.3079,
+      "epoch": 1.3350992900133347,
+      "grad_norm": 0.1985885351896286,
+      "learning_rate": 4.2994590865965634e-05,
+      "loss": 0.4577,
       "step": 37045
     },
     {
-      "epoch": 1.3,
-      "learning_rate": 4.33443709054423e-05,
-      "loss": 0.3001,
+      "epoch": 1.3352794896745594,
+      "grad_norm": 0.14682763814926147,
+      "learning_rate": 4.2992564979460484e-05,
+      "loss": 0.397,
       "step": 37050
     },
     {
-      "epoch": 1.3,
-      "learning_rate": 4.334243538578486e-05,
-      "loss": 0.303,
+      "epoch": 1.3354596893357842,
+      "grad_norm": 0.17361211776733398,
+      "learning_rate": 4.299053884781012e-05,
+      "loss": 0.3837,
       "step": 37055
     },
     {
-      "epoch": 1.3,
-      "learning_rate": 4.334049962796297e-05,
-      "loss": 0.3096,
+      "epoch": 1.3356398889970087,
+      "grad_norm": 0.17351488769054413,
+      "learning_rate": 4.2988512471042154e-05,
+      "loss": 0.4613,
       "step": 37060
     },
     {
-      "epoch": 1.3,
-      "learning_rate": 4.333856363200179e-05,
-      "loss": 0.2888,
+      "epoch": 1.3358200886582332,
+      "grad_norm": 0.16963542997837067,
+      "learning_rate": 4.2986485849184185e-05,
+      "loss": 0.4384,
       "step": 37065
     },
     {
-      "epoch": 1.3,
-      "learning_rate": 4.3336627397926444e-05,
-      "loss": 0.294,
+      "epoch": 1.336000288319458,
+      "grad_norm": 0.16560989618301392,
+      "learning_rate": 4.298445898226383e-05,
+      "loss": 0.4036,
       "step": 37070
     },
     {
-      "epoch": 1.3,
-      "learning_rate": 4.3334690925762074e-05,
-      "loss": 0.3129,
+      "epoch": 1.3361804879806827,
+      "grad_norm": 0.1843060851097107,
+      "learning_rate": 4.29824318703087e-05,
+      "loss": 0.384,
       "step": 37075
     },
     {
-      "epoch": 1.3,
-      "learning_rate": 4.333275421553382e-05,
-      "loss": 0.2793,
+      "epoch": 1.3363606876419072,
+      "grad_norm": 0.20342443883419037,
+      "learning_rate": 4.298040451334642e-05,
+      "loss": 0.4224,
       "step": 37080
     },
     {
-      "epoch": 1.3,
-      "learning_rate": 4.333081726726684e-05,
-      "loss": 0.275,
+      "epoch": 1.336540887303132,
+      "grad_norm": 0.17561762034893036,
+      "learning_rate": 4.297837691140461e-05,
+      "loss": 0.4238,
       "step": 37085
     },
     {
-      "epoch": 1.3,
-      "learning_rate": 4.332888008098629e-05,
-      "loss": 0.2776,
+      "epoch": 1.3367210869643564,
+      "grad_norm": 0.20807789266109467,
+      "learning_rate": 4.297634906451089e-05,
+      "loss": 0.4155,
       "step": 37090
     },
     {
-      "epoch": 1.31,
-      "learning_rate": 4.33269426567173e-05,
-      "loss": 0.286,
+      "epoch": 1.3369012866255812,
+      "grad_norm": 0.19749189913272858,
+      "learning_rate": 4.29743209726929e-05,
+      "loss": 0.4417,
       "step": 37095
     },
     {
-      "epoch": 1.31,
-      "learning_rate": 4.332500499448503e-05,
-      "loss": 0.2853,
+      "epoch": 1.337081486286806,
+      "grad_norm": 0.15271851420402527,
+      "learning_rate": 4.297229263597827e-05,
+      "loss": 0.3862,
       "step": 37100
     },
     {
-      "epoch": 1.31,
-      "learning_rate": 4.3323067094314654e-05,
-      "loss": 0.2795,
+      "epoch": 1.3372616859480304,
+      "grad_norm": 0.19541774690151215,
+      "learning_rate": 4.2970264054394625e-05,
+      "loss": 0.3878,
       "step": 37105
     },
     {
-      "epoch": 1.31,
-      "learning_rate": 4.332112895623133e-05,
-      "loss": 0.2884,
+      "epoch": 1.337441885609255,
+      "grad_norm": 0.20491370558738708,
+      "learning_rate": 4.296823522796961e-05,
+      "loss": 0.4285,
       "step": 37110
     },
     {
-      "epoch": 1.31,
-      "learning_rate": 4.3319190580260215e-05,
-      "loss": 0.2788,
+      "epoch": 1.3376220852704797,
+      "grad_norm": 0.19413931667804718,
+      "learning_rate": 4.2966206156730875e-05,
+      "loss": 0.4268,
       "step": 37115
     },
     {
-      "epoch": 1.31,
-      "learning_rate": 4.331725196642649e-05,
-      "loss": 0.2847,
+      "epoch": 1.3378022849317044,
+      "grad_norm": 0.17734622955322266,
+      "learning_rate": 4.296417684070606e-05,
+      "loss": 0.3678,
       "step": 37120
     },
     {
-      "epoch": 1.31,
-      "learning_rate": 4.331531311475531e-05,
-      "loss": 0.3001,
+      "epoch": 1.337982484592929,
+      "grad_norm": 0.1772347390651703,
+      "learning_rate": 4.29621472799228e-05,
+      "loss": 0.4416,
       "step": 37125
     },
     {
-      "epoch": 1.31,
-      "learning_rate": 4.331337402527187e-05,
-      "loss": 0.2949,
+      "epoch": 1.3381626842541536,
+      "grad_norm": 0.2039223164319992,
+      "learning_rate": 4.296011747440878e-05,
+      "loss": 0.4219,
       "step": 37130
     },
     {
-      "epoch": 1.31,
-      "learning_rate": 4.331143469800133e-05,
-      "loss": 0.3044,
+      "epoch": 1.3383428839153781,
+      "grad_norm": 0.17975406348705292,
+      "learning_rate": 4.295808742419163e-05,
+      "loss": 0.4512,
       "step": 37135
     },
     {
-      "epoch": 1.31,
-      "learning_rate": 4.330949513296888e-05,
-      "loss": 0.2916,
+      "epoch": 1.3385230835766029,
+      "grad_norm": 0.21606260538101196,
+      "learning_rate": 4.295605712929901e-05,
+      "loss": 0.4352,
       "step": 37140
     },
     {
-      "epoch": 1.31,
-      "learning_rate": 4.33075553301997e-05,
-      "loss": 0.2822,
+      "epoch": 1.3387032832378276,
+      "grad_norm": 0.16076348721981049,
+      "learning_rate": 4.295402658975859e-05,
+      "loss": 0.4171,
       "step": 37145
     },
     {
-      "epoch": 1.31,
-      "learning_rate": 4.330561528971898e-05,
-      "loss": 0.2897,
+      "epoch": 1.3388834828990521,
+      "grad_norm": 0.14485345780849457,
+      "learning_rate": 4.295199580559804e-05,
+      "loss": 0.4193,
       "step": 37150
     },
     {
-      "epoch": 1.31,
-      "learning_rate": 4.330367501155191e-05,
-      "loss": 0.2808,
+      "epoch": 1.3390636825602769,
+      "grad_norm": 0.17786848545074463,
+      "learning_rate": 4.2949964776845014e-05,
+      "loss": 0.3925,
       "step": 37155
     },
     {
-      "epoch": 1.31,
-      "learning_rate": 4.3301734495723687e-05,
-      "loss": 0.2806,
+      "epoch": 1.3392438822215014,
+      "grad_norm": 0.2276924103498459,
+      "learning_rate": 4.29479335035272e-05,
+      "loss": 0.4485,
       "step": 37160
     },
     {
-      "epoch": 1.31,
-      "learning_rate": 4.32997937422595e-05,
-      "loss": 0.2932,
+      "epoch": 1.339424081882726,
+      "grad_norm": 0.17060185968875885,
+      "learning_rate": 4.294590198567226e-05,
+      "loss": 0.4437,
       "step": 37165
     },
     {
-      "epoch": 1.31,
-      "learning_rate": 4.3297852751184555e-05,
-      "loss": 0.2838,
+      "epoch": 1.3396042815439506,
+      "grad_norm": 0.18180356919765472,
+      "learning_rate": 4.294387022330789e-05,
+      "loss": 0.3794,
       "step": 37170
     },
     {
-      "epoch": 1.31,
-      "learning_rate": 4.3295911522524044e-05,
-      "loss": 0.2988,
+      "epoch": 1.3397844812051753,
+      "grad_norm": 0.17206217348575592,
+      "learning_rate": 4.294183821646175e-05,
+      "loss": 0.4358,
       "step": 37175
     },
     {
-      "epoch": 1.31,
-      "learning_rate": 4.329397005630318e-05,
-      "loss": 0.2963,
+      "epoch": 1.3399646808663999,
+      "grad_norm": 0.19352799654006958,
+      "learning_rate": 4.293980596516155e-05,
+      "loss": 0.4043,
       "step": 37180
     },
     {
-      "epoch": 1.31,
-      "learning_rate": 4.3292028352547176e-05,
-      "loss": 0.2902,
+      "epoch": 1.3401448805276246,
+      "grad_norm": 0.2113627791404724,
+      "learning_rate": 4.2937773469434963e-05,
+      "loss": 0.3979,
       "step": 37185
     },
     {
-      "epoch": 1.31,
-      "learning_rate": 4.3290086411281234e-05,
-      "loss": 0.2953,
+      "epoch": 1.3403250801888493,
+      "grad_norm": 0.1969025582075119,
+      "learning_rate": 4.293574072930968e-05,
+      "loss": 0.4085,
       "step": 37190
     },
     {
-      "epoch": 1.31,
-      "learning_rate": 4.328814423253058e-05,
-      "loss": 0.2826,
+      "epoch": 1.3405052798500738,
+      "grad_norm": 0.1911691427230835,
+      "learning_rate": 4.29337077448134e-05,
+      "loss": 0.4124,
       "step": 37195
     },
     {
-      "epoch": 1.31,
-      "learning_rate": 4.3286201816320415e-05,
-      "loss": 0.2786,
+      "epoch": 1.3406854795112986,
+      "grad_norm": 0.16924938559532166,
+      "learning_rate": 4.293167451597383e-05,
+      "loss": 0.4439,
       "step": 37200
     },
     {
-      "epoch": 1.31,
-      "learning_rate": 4.328425916267598e-05,
-      "loss": 0.288,
+      "epoch": 1.340865679172523,
+      "grad_norm": 0.15736305713653564,
+      "learning_rate": 4.292964104281867e-05,
+      "loss": 0.3968,
       "step": 37205
     },
     {
-      "epoch": 1.31,
-      "learning_rate": 4.328231627162248e-05,
-      "loss": 0.285,
+      "epoch": 1.3410458788337478,
+      "grad_norm": 0.22912681102752686,
+      "learning_rate": 4.2927607325375616e-05,
+      "loss": 0.4219,
       "step": 37210
     },
     {
-      "epoch": 1.31,
-      "learning_rate": 4.328037314318516e-05,
-      "loss": 0.2882,
+      "epoch": 1.3412260784949726,
+      "grad_norm": 0.1928148865699768,
+      "learning_rate": 4.292557336367239e-05,
+      "loss": 0.3937,
       "step": 37215
     },
     {
-      "epoch": 1.31,
-      "learning_rate": 4.327842977738924e-05,
-      "loss": 0.287,
+      "epoch": 1.341406278156197,
+      "grad_norm": 0.19849920272827148,
+      "learning_rate": 4.2923539157736695e-05,
+      "loss": 0.4182,
       "step": 37220
     },
     {
-      "epoch": 1.31,
-      "learning_rate": 4.327648617425995e-05,
-      "loss": 0.2864,
+      "epoch": 1.3415864778174216,
+      "grad_norm": 0.14143873751163483,
+      "learning_rate": 4.292150470759624e-05,
+      "loss": 0.3873,
       "step": 37225
     },
     {
-      "epoch": 1.31,
-      "learning_rate": 4.3274542333822544e-05,
-      "loss": 0.3049,
+      "epoch": 1.3417666774786463,
+      "grad_norm": 0.17856481671333313,
+      "learning_rate": 4.291947001327876e-05,
+      "loss": 0.435,
       "step": 37230
     },
     {
-      "epoch": 1.31,
-      "learning_rate": 4.327259825610224e-05,
-      "loss": 0.2877,
+      "epoch": 1.341946877139871,
+      "grad_norm": 0.15934741497039795,
+      "learning_rate": 4.291743507481197e-05,
+      "loss": 0.4061,
       "step": 37235
     },
     {
-      "epoch": 1.31,
-      "learning_rate": 4.327065394112429e-05,
-      "loss": 0.2963,
+      "epoch": 1.3421270768010956,
+      "grad_norm": 0.193971648812294,
+      "learning_rate": 4.2915399892223595e-05,
+      "loss": 0.3881,
       "step": 37240
     },
     {
-      "epoch": 1.31,
-      "learning_rate": 4.3268709388913944e-05,
-      "loss": 0.2902,
+      "epoch": 1.3423072764623203,
+      "grad_norm": 0.22746896743774414,
+      "learning_rate": 4.2913364465541366e-05,
+      "loss": 0.411,
       "step": 37245
     },
     {
-      "epoch": 1.31,
-      "learning_rate": 4.326676459949644e-05,
-      "loss": 0.293,
+      "epoch": 1.3424874761235448,
+      "grad_norm": 0.21635814011096954,
+      "learning_rate": 4.291132879479302e-05,
+      "loss": 0.4091,
       "step": 37250
     },
     {
-      "epoch": 1.31,
-      "learning_rate": 4.326481957289704e-05,
-      "loss": 0.2932,
+      "epoch": 1.3426676757847695,
+      "grad_norm": 0.19370247423648834,
+      "learning_rate": 4.290929288000628e-05,
+      "loss": 0.4045,
       "step": 37255
     },
     {
-      "epoch": 1.31,
-      "learning_rate": 4.326287430914099e-05,
-      "loss": 0.3053,
+      "epoch": 1.3428478754459943,
+      "grad_norm": 0.14422830939292908,
+      "learning_rate": 4.290725672120889e-05,
+      "loss": 0.3829,
       "step": 37260
     },
     {
-      "epoch": 1.31,
-      "learning_rate": 4.326092880825355e-05,
-      "loss": 0.2989,
+      "epoch": 1.3430280751072188,
+      "grad_norm": 0.16739609837532043,
+      "learning_rate": 4.29052203184286e-05,
+      "loss": 0.4541,
       "step": 37265
     },
     {
-      "epoch": 1.31,
-      "learning_rate": 4.3258983070259994e-05,
-      "loss": 0.2813,
+      "epoch": 1.3432082747684435,
+      "grad_norm": 0.17968153953552246,
+      "learning_rate": 4.290318367169314e-05,
+      "loss": 0.4112,
       "step": 37270
     },
     {
-      "epoch": 1.31,
-      "learning_rate": 4.325703709518557e-05,
-      "loss": 0.2908,
+      "epoch": 1.343388474429668,
+      "grad_norm": 0.20018011331558228,
+      "learning_rate": 4.290114678103028e-05,
+      "loss": 0.4536,
       "step": 37275
     },
     {
-      "epoch": 1.31,
-      "learning_rate": 4.325509088305556e-05,
-      "loss": 0.2831,
+      "epoch": 1.3435686740908928,
+      "grad_norm": 0.1976526975631714,
+      "learning_rate": 4.289910964646776e-05,
+      "loss": 0.4011,
       "step": 37280
     },
     {
-      "epoch": 1.31,
-      "learning_rate": 4.325314443389522e-05,
-      "loss": 0.2986,
+      "epoch": 1.3437488737521173,
+      "grad_norm": 0.19108478724956512,
+      "learning_rate": 4.289707226803333e-05,
+      "loss": 0.4303,
       "step": 37285
     },
     {
-      "epoch": 1.31,
-      "learning_rate": 4.325119774772982e-05,
-      "loss": 0.2859,
+      "epoch": 1.343929073413342,
+      "grad_norm": 0.1997978538274765,
+      "learning_rate": 4.289503464575476e-05,
+      "loss": 0.4182,
       "step": 37290
     },
     {
-      "epoch": 1.31,
-      "learning_rate": 4.324925082458465e-05,
-      "loss": 0.2937,
+      "epoch": 1.3441092730745665,
+      "grad_norm": 0.20176932215690613,
+      "learning_rate": 4.28929967796598e-05,
+      "loss": 0.4345,
       "step": 37295
     },
     {
-      "epoch": 1.31,
-      "learning_rate": 4.3247303664485e-05,
-      "loss": 0.2865,
+      "epoch": 1.3442894727357912,
+      "grad_norm": 0.1834762841463089,
+      "learning_rate": 4.289095866977623e-05,
+      "loss": 0.4147,
       "step": 37300
     },
     {
-      "epoch": 1.31,
-      "learning_rate": 4.324535626745612e-05,
-      "loss": 0.2826,
+      "epoch": 1.344469672397016,
+      "grad_norm": 0.18414629995822906,
+      "learning_rate": 4.288892031613181e-05,
+      "loss": 0.4264,
       "step": 37305
     },
     {
-      "epoch": 1.31,
-      "learning_rate": 4.324340863352332e-05,
-      "loss": 0.2826,
+      "epoch": 1.3446498720582405,
+      "grad_norm": 0.175818532705307,
+      "learning_rate": 4.288688171875431e-05,
+      "loss": 0.379,
       "step": 37310
     },
     {
-      "epoch": 1.31,
-      "learning_rate": 4.3241460762711874e-05,
-      "loss": 0.2433,
+      "epoch": 1.3448300717194652,
+      "grad_norm": 0.16372618079185486,
+      "learning_rate": 4.288484287767152e-05,
+      "loss": 0.3926,
       "step": 37315
     },
     {
-      "epoch": 1.31,
-      "learning_rate": 4.3239512655047086e-05,
-      "loss": 0.2867,
+      "epoch": 1.3450102713806897,
+      "grad_norm": 0.17011037468910217,
+      "learning_rate": 4.2882803792911205e-05,
+      "loss": 0.4143,
       "step": 37320
     },
     {
-      "epoch": 1.31,
-      "learning_rate": 4.323756431055425e-05,
-      "loss": 0.2893,
+      "epoch": 1.3451904710419145,
+      "grad_norm": 0.13715289533138275,
+      "learning_rate": 4.288076446450115e-05,
+      "loss": 0.3912,
       "step": 37325
     },
     {
-      "epoch": 1.31,
-      "learning_rate": 4.323561572925866e-05,
-      "loss": 0.3029,
+      "epoch": 1.3453706707031392,
+      "grad_norm": 0.17241427302360535,
+      "learning_rate": 4.2878724892469135e-05,
+      "loss": 0.3965,
       "step": 37330
     },
     {
-      "epoch": 1.31,
-      "learning_rate": 4.323366691118561e-05,
-      "loss": 0.2687,
+      "epoch": 1.3455508703643637,
+      "grad_norm": 0.17453129589557648,
+      "learning_rate": 4.287668507684296e-05,
+      "loss": 0.4192,
       "step": 37335
     },
     {
-      "epoch": 1.31,
-      "learning_rate": 4.323171785636042e-05,
-      "loss": 0.2693,
+      "epoch": 1.3457310700255882,
+      "grad_norm": 0.22554276883602142,
+      "learning_rate": 4.287464501765041e-05,
+      "loss": 0.3842,
       "step": 37340
     },
     {
-      "epoch": 1.31,
-      "learning_rate": 4.322976856480839e-05,
-      "loss": 0.2726,
+      "epoch": 1.345911269686813,
+      "grad_norm": 0.1914464235305786,
+      "learning_rate": 4.2872604714919285e-05,
+      "loss": 0.4156,
       "step": 37345
     },
     {
-      "epoch": 1.31,
-      "learning_rate": 4.322781903655482e-05,
-      "loss": 0.2934,
+      "epoch": 1.3460914693480377,
+      "grad_norm": 0.16608940064907074,
+      "learning_rate": 4.287056416867738e-05,
+      "loss": 0.4284,
       "step": 37350
     },
     {
-      "epoch": 1.31,
-      "learning_rate": 4.322586927162504e-05,
-      "loss": 0.2632,
+      "epoch": 1.3462716690092622,
+      "grad_norm": 0.20731467008590698,
+      "learning_rate": 4.28685233789525e-05,
+      "loss": 0.4522,
       "step": 37355
     },
     {
-      "epoch": 1.31,
-      "learning_rate": 4.322391927004435e-05,
-      "loss": 0.3015,
+      "epoch": 1.346451868670487,
+      "grad_norm": 0.1691913902759552,
+      "learning_rate": 4.286648234577244e-05,
+      "loss": 0.4299,
       "step": 37360
     },
     {
-      "epoch": 1.31,
-      "learning_rate": 4.322196903183809e-05,
-      "loss": 0.2796,
+      "epoch": 1.3466320683317115,
+      "grad_norm": 0.19073499739170074,
+      "learning_rate": 4.286444106916503e-05,
+      "loss": 0.4108,
       "step": 37365
     },
     {
-      "epoch": 1.31,
-      "learning_rate": 4.322001855703156e-05,
-      "loss": 0.292,
+      "epoch": 1.3468122679929362,
+      "grad_norm": 0.1724918931722641,
+      "learning_rate": 4.286239954915806e-05,
+      "loss": 0.4408,
       "step": 37370
     },
     {
-      "epoch": 1.31,
-      "learning_rate": 4.3218067845650103e-05,
-      "loss": 0.2613,
+      "epoch": 1.346992467654161,
+      "grad_norm": 0.19788911938667297,
+      "learning_rate": 4.2860357785779356e-05,
+      "loss": 0.4212,
       "step": 37375
     },
     {
-      "epoch": 1.32,
-      "learning_rate": 4.321611689771904e-05,
-      "loss": 0.2833,
+      "epoch": 1.3471726673153854,
+      "grad_norm": 0.17659629881381989,
+      "learning_rate": 4.2858315779056734e-05,
+      "loss": 0.4535,
       "step": 37380
     },
     {
-      "epoch": 1.32,
-      "learning_rate": 4.321416571326371e-05,
-      "loss": 0.3139,
+      "epoch": 1.3473528669766102,
+      "grad_norm": 0.1844049096107483,
+      "learning_rate": 4.285627352901802e-05,
+      "loss": 0.388,
       "step": 37385
     },
     {
-      "epoch": 1.32,
-      "learning_rate": 4.3212214292309435e-05,
-      "loss": 0.2846,
+      "epoch": 1.3475330666378347,
+      "grad_norm": 0.19511617720127106,
+      "learning_rate": 4.285423103569103e-05,
+      "loss": 0.4225,
       "step": 37390
     },
     {
-      "epoch": 1.32,
-      "learning_rate": 4.321026263488156e-05,
-      "loss": 0.2884,
+      "epoch": 1.3477132662990594,
+      "grad_norm": 0.1395874172449112,
+      "learning_rate": 4.2852188299103614e-05,
+      "loss": 0.3867,
       "step": 37395
     },
     {
-      "epoch": 1.32,
-      "learning_rate": 4.320831074100542e-05,
-      "loss": 0.2962,
+      "epoch": 1.347893465960284,
+      "grad_norm": 0.1697840690612793,
+      "learning_rate": 4.2850145319283575e-05,
+      "loss": 0.3939,
       "step": 37400
     },
     {
-      "epoch": 1.32,
-      "learning_rate": 4.3206358610706376e-05,
-      "loss": 0.3143,
+      "epoch": 1.3480736656215087,
+      "grad_norm": 0.20716294646263123,
+      "learning_rate": 4.284810209625876e-05,
+      "loss": 0.4563,
       "step": 37405
     },
     {
-      "epoch": 1.32,
-      "learning_rate": 4.3204406244009754e-05,
-      "loss": 0.2929,
+      "epoch": 1.3482538652827332,
+      "grad_norm": 0.19542010128498077,
+      "learning_rate": 4.2846058630057016e-05,
+      "loss": 0.4322,
       "step": 37410
     },
     {
-      "epoch": 1.32,
-      "learning_rate": 4.3202453640940924e-05,
-      "loss": 0.2903,
+      "epoch": 1.348434064943958,
+      "grad_norm": 0.15762588381767273,
+      "learning_rate": 4.2844014920706176e-05,
+      "loss": 0.4279,
       "step": 37415
     },
     {
-      "epoch": 1.32,
-      "learning_rate": 4.320050080152522e-05,
-      "loss": 0.2909,
+      "epoch": 1.3486142646051826,
+      "grad_norm": 0.20713409781455994,
+      "learning_rate": 4.284197096823409e-05,
+      "loss": 0.4345,
       "step": 37420
     },
     {
-      "epoch": 1.32,
-      "learning_rate": 4.319854772578801e-05,
-      "loss": 0.2865,
+      "epoch": 1.3487944642664071,
+      "grad_norm": 0.18533405661582947,
+      "learning_rate": 4.2839926772668605e-05,
+      "loss": 0.4219,
       "step": 37425
     },
     {
-      "epoch": 1.32,
-      "learning_rate": 4.319659441375465e-05,
-      "loss": 0.3007,
+      "epoch": 1.3489746639276319,
+      "grad_norm": 0.19821499288082123,
+      "learning_rate": 4.283788233403757e-05,
+      "loss": 0.4249,
       "step": 37430
     },
     {
-      "epoch": 1.32,
-      "learning_rate": 4.319464086545052e-05,
-      "loss": 0.2722,
+      "epoch": 1.3491548635888564,
+      "grad_norm": 0.20067061483860016,
+      "learning_rate": 4.283583765236884e-05,
+      "loss": 0.4566,
       "step": 37435
     },
     {
-      "epoch": 1.32,
-      "learning_rate": 4.319268708090095e-05,
-      "loss": 0.3013,
+      "epoch": 1.3493350632500811,
+      "grad_norm": 0.16984498500823975,
+      "learning_rate": 4.2833792727690275e-05,
+      "loss": 0.38,
       "step": 37440
     },
     {
-      "epoch": 1.32,
-      "learning_rate": 4.3190733060131326e-05,
-      "loss": 0.2961,
+      "epoch": 1.3495152629113059,
+      "grad_norm": 0.1640443056821823,
+      "learning_rate": 4.283174756002973e-05,
+      "loss": 0.4147,
       "step": 37445
     },
     {
-      "epoch": 1.32,
-      "learning_rate": 4.318877880316703e-05,
-      "loss": 0.3062,
+      "epoch": 1.3496954625725304,
+      "grad_norm": 0.16982309520244598,
+      "learning_rate": 4.2829702149415094e-05,
+      "loss": 0.4274,
       "step": 37450
     },
     {
-      "epoch": 1.32,
-      "learning_rate": 4.318682431003343e-05,
-      "loss": 0.2742,
+      "epoch": 1.3498756622337549,
+      "grad_norm": 0.19028659164905548,
+      "learning_rate": 4.2827656495874205e-05,
+      "loss": 0.4304,
       "step": 37455
     },
     {
-      "epoch": 1.32,
-      "learning_rate": 4.31848695807559e-05,
-      "loss": 0.2973,
+      "epoch": 1.3500558618949796,
+      "grad_norm": 0.20685800909996033,
+      "learning_rate": 4.282561059943495e-05,
+      "loss": 0.4145,
       "step": 37460
     },
     {
-      "epoch": 1.32,
-      "learning_rate": 4.318291461535982e-05,
-      "loss": 0.2853,
+      "epoch": 1.3502360615562043,
+      "grad_norm": 0.18476605415344238,
+      "learning_rate": 4.2823564460125206e-05,
+      "loss": 0.4054,
       "step": 37465
     },
     {
-      "epoch": 1.32,
-      "learning_rate": 4.318095941387058e-05,
-      "loss": 0.2627,
+      "epoch": 1.3504162612174289,
+      "grad_norm": 0.17075756192207336,
+      "learning_rate": 4.2821518077972845e-05,
+      "loss": 0.3853,
       "step": 37470
     },
     {
-      "epoch": 1.32,
-      "learning_rate": 4.3179003976313556e-05,
-      "loss": 0.2873,
+      "epoch": 1.3505964608786536,
+      "grad_norm": 0.1620296984910965,
+      "learning_rate": 4.281947145300574e-05,
+      "loss": 0.3927,
       "step": 37475
     },
     {
-      "epoch": 1.32,
-      "learning_rate": 4.3177048302714154e-05,
-      "loss": 0.2737,
+      "epoch": 1.350776660539878,
+      "grad_norm": 0.1557290405035019,
+      "learning_rate": 4.2817424585251804e-05,
+      "loss": 0.4178,
       "step": 37480
     },
     {
-      "epoch": 1.32,
-      "learning_rate": 4.317509239309776e-05,
-      "loss": 0.2906,
+      "epoch": 1.3509568602011028,
+      "grad_norm": 0.17383769154548645,
+      "learning_rate": 4.2815377474738894e-05,
+      "loss": 0.4483,
       "step": 37485
     },
     {
-      "epoch": 1.32,
-      "learning_rate": 4.3173136247489766e-05,
-      "loss": 0.2583,
+      "epoch": 1.3511370598623276,
+      "grad_norm": 0.19885095953941345,
+      "learning_rate": 4.2813330121494924e-05,
+      "loss": 0.4472,
       "step": 37490
     },
     {
-      "epoch": 1.32,
-      "learning_rate": 4.317117986591557e-05,
-      "loss": 0.3009,
+      "epoch": 1.351317259523552,
+      "grad_norm": 0.18634334206581116,
+      "learning_rate": 4.281128252554778e-05,
+      "loss": 0.4156,
       "step": 37495
     },
     {
-      "epoch": 1.32,
-      "learning_rate": 4.316922324840058e-05,
-      "loss": 0.2795,
+      "epoch": 1.3514974591847766,
+      "grad_norm": 0.21078574657440186,
+      "learning_rate": 4.280923468692535e-05,
+      "loss": 0.4509,
       "step": 37500
     },
     {
-      "epoch": 1.32,
-      "eval_loss": 0.28416377305984497,
-      "eval_runtime": 10.5464,
-      "eval_samples_per_second": 9.482,
-      "eval_steps_per_second": 9.482,
+      "epoch": 1.3514974591847766,
+      "eval_loss": 0.44750291109085083,
+      "eval_runtime": 3.5455,
+      "eval_samples_per_second": 28.205,
+      "eval_steps_per_second": 7.051,
       "step": 37500
     },
     {
-      "epoch": 1.32,
-      "learning_rate": 4.3167266394970206e-05,
-      "loss": 0.2812,
+      "epoch": 1.3516776588460013,
+      "grad_norm": 0.21261395514011383,
+      "learning_rate": 4.280718660565556e-05,
+      "loss": 0.4191,
       "step": 37505
     },
     {
-      "epoch": 1.32,
-      "learning_rate": 4.316530930564985e-05,
-      "loss": 0.2841,
+      "epoch": 1.351857858507226,
+      "grad_norm": 0.1816295087337494,
+      "learning_rate": 4.280513828176629e-05,
+      "loss": 0.4367,
       "step": 37510
     },
     {
-      "epoch": 1.32,
-      "learning_rate": 4.3163351980464916e-05,
-      "loss": 0.2898,
+      "epoch": 1.3520380581684506,
+      "grad_norm": 0.16183680295944214,
+      "learning_rate": 4.280308971528546e-05,
+      "loss": 0.4168,
       "step": 37515
     },
     {
-      "epoch": 1.32,
-      "learning_rate": 4.3161394419440836e-05,
-      "loss": 0.2741,
+      "epoch": 1.3522182578296753,
+      "grad_norm": 0.19457785785198212,
+      "learning_rate": 4.280104090624097e-05,
+      "loss": 0.4259,
       "step": 37520
     },
     {
-      "epoch": 1.32,
-      "learning_rate": 4.3159436622603013e-05,
-      "loss": 0.3133,
+      "epoch": 1.3523984574908998,
+      "grad_norm": 0.16114507615566254,
+      "learning_rate": 4.279899185466077e-05,
+      "loss": 0.4355,
       "step": 37525
     },
     {
-      "epoch": 1.32,
-      "learning_rate": 4.3157478589976874e-05,
-      "loss": 0.3004,
+      "epoch": 1.3525786571521246,
+      "grad_norm": 0.17467471957206726,
+      "learning_rate": 4.2796942560572725e-05,
+      "loss": 0.4127,
       "step": 37530
     },
     {
-      "epoch": 1.32,
-      "learning_rate": 4.315552032158784e-05,
-      "loss": 0.3081,
+      "epoch": 1.3527588568133493,
+      "grad_norm": 0.18271581828594208,
+      "learning_rate": 4.279489302400479e-05,
+      "loss": 0.4033,
       "step": 37535
     },
     {
-      "epoch": 1.32,
-      "learning_rate": 4.315356181746134e-05,
-      "loss": 0.2947,
+      "epoch": 1.3529390564745738,
+      "grad_norm": 0.15047767758369446,
+      "learning_rate": 4.279284324498489e-05,
+      "loss": 0.3937,
       "step": 37540
     },
     {
-      "epoch": 1.32,
-      "learning_rate": 4.31516030776228e-05,
-      "loss": 0.3216,
+      "epoch": 1.3531192561357985,
+      "grad_norm": 0.1880033016204834,
+      "learning_rate": 4.2790793223540944e-05,
+      "loss": 0.4112,
       "step": 37545
     },
     {
-      "epoch": 1.32,
-      "learning_rate": 4.314964410209767e-05,
-      "loss": 0.2855,
+      "epoch": 1.353299455797023,
+      "grad_norm": 0.2034742534160614,
+      "learning_rate": 4.278874295970088e-05,
+      "loss": 0.4222,
       "step": 37550
     },
     {
-      "epoch": 1.32,
-      "learning_rate": 4.3147684890911365e-05,
-      "loss": 0.308,
+      "epoch": 1.3534796554582478,
+      "grad_norm": 0.14780563116073608,
+      "learning_rate": 4.278669245349264e-05,
+      "loss": 0.4239,
       "step": 37555
     },
     {
-      "epoch": 1.32,
-      "learning_rate": 4.314572544408932e-05,
-      "loss": 0.3049,
+      "epoch": 1.3536598551194725,
+      "grad_norm": 0.15537936985492706,
+      "learning_rate": 4.278464170494416e-05,
+      "loss": 0.3862,
       "step": 37560
     },
     {
-      "epoch": 1.32,
-      "learning_rate": 4.314376576165701e-05,
-      "loss": 0.2787,
+      "epoch": 1.353840054780697,
+      "grad_norm": 0.14298701286315918,
+      "learning_rate": 4.278259071408338e-05,
+      "loss": 0.4007,
       "step": 37565
     },
     {
-      "epoch": 1.32,
-      "learning_rate": 4.314180584363984e-05,
-      "loss": 0.2797,
+      "epoch": 1.3540202544419215,
+      "grad_norm": 0.1611703783273697,
+      "learning_rate": 4.278053948093824e-05,
+      "loss": 0.4007,
       "step": 37570
     },
     {
-      "epoch": 1.32,
-      "learning_rate": 4.313984569006329e-05,
-      "loss": 0.3024,
+      "epoch": 1.3542004541031463,
+      "grad_norm": 0.2151387333869934,
+      "learning_rate": 4.27784880055367e-05,
+      "loss": 0.4169,
       "step": 37575
     },
     {
-      "epoch": 1.32,
-      "learning_rate": 4.313788530095279e-05,
-      "loss": 0.2949,
+      "epoch": 1.354380653764371,
+      "grad_norm": 0.17721928656101227,
+      "learning_rate": 4.277643628790669e-05,
+      "loss": 0.4298,
       "step": 37580
     },
     {
-      "epoch": 1.32,
-      "learning_rate": 4.3135924676333806e-05,
-      "loss": 0.2951,
+      "epoch": 1.3545608534255955,
+      "grad_norm": 0.17061589658260345,
+      "learning_rate": 4.277438432807619e-05,
+      "loss": 0.4005,
       "step": 37585
     },
     {
-      "epoch": 1.32,
-      "learning_rate": 4.313396381623179e-05,
-      "loss": 0.2992,
+      "epoch": 1.3547410530868202,
+      "grad_norm": 0.159605473279953,
+      "learning_rate": 4.277233212607315e-05,
+      "loss": 0.4158,
       "step": 37590
     },
     {
-      "epoch": 1.32,
-      "learning_rate": 4.313200272067221e-05,
-      "loss": 0.2875,
+      "epoch": 1.3549212527480448,
+      "grad_norm": 0.15221120417118073,
+      "learning_rate": 4.2770279681925506e-05,
+      "loss": 0.4054,
       "step": 37595
     },
     {
-      "epoch": 1.32,
-      "learning_rate": 4.313004138968052e-05,
-      "loss": 0.2947,
+      "epoch": 1.3551014524092695,
+      "grad_norm": 0.1982593834400177,
+      "learning_rate": 4.2768226995661255e-05,
+      "loss": 0.3861,
       "step": 37600
     },
     {
-      "epoch": 1.32,
-      "learning_rate": 4.3128079823282185e-05,
-      "loss": 0.2795,
+      "epoch": 1.3552816520704942,
+      "grad_norm": 0.1528361439704895,
+      "learning_rate": 4.276617406730835e-05,
+      "loss": 0.4002,
       "step": 37605
     },
     {
-      "epoch": 1.32,
-      "learning_rate": 4.312611802150269e-05,
-      "loss": 0.29,
+      "epoch": 1.3554618517317187,
+      "grad_norm": 0.17486098408699036,
+      "learning_rate": 4.2764120896894755e-05,
+      "loss": 0.3899,
       "step": 37610
     },
     {
-      "epoch": 1.32,
-      "learning_rate": 4.31241559843675e-05,
-      "loss": 0.3024,
+      "epoch": 1.3556420513929432,
+      "grad_norm": 0.16037683188915253,
+      "learning_rate": 4.276206748444846e-05,
+      "loss": 0.4031,
       "step": 37615
     },
     {
-      "epoch": 1.32,
-      "learning_rate": 4.312219371190208e-05,
-      "loss": 0.2777,
+      "epoch": 1.355822251054168,
+      "grad_norm": 0.1568463146686554,
+      "learning_rate": 4.2760013829997434e-05,
+      "loss": 0.4136,
       "step": 37620
     },
     {
-      "epoch": 1.32,
-      "learning_rate": 4.3120231204131924e-05,
-      "loss": 0.2749,
+      "epoch": 1.3560024507153927,
+      "grad_norm": 0.18203102052211761,
+      "learning_rate": 4.2757959933569656e-05,
+      "loss": 0.3986,
       "step": 37625
     },
     {
-      "epoch": 1.32,
-      "learning_rate": 4.31182684610825e-05,
-      "loss": 0.2936,
+      "epoch": 1.3561826503766172,
+      "grad_norm": 0.16312682628631592,
+      "learning_rate": 4.275590579519311e-05,
+      "loss": 0.4104,
       "step": 37630
     },
     {
-      "epoch": 1.32,
-      "learning_rate": 4.311630548277931e-05,
-      "loss": 0.293,
+      "epoch": 1.356362850037842,
+      "grad_norm": 0.18882383406162262,
+      "learning_rate": 4.275385141489578e-05,
+      "loss": 0.4471,
       "step": 37635
     },
     {
-      "epoch": 1.32,
-      "learning_rate": 4.311434226924782e-05,
-      "loss": 0.305,
+      "epoch": 1.3565430496990665,
+      "grad_norm": 0.21008935570716858,
+      "learning_rate": 4.275179679270568e-05,
+      "loss": 0.4276,
       "step": 37640
     },
     {
-      "epoch": 1.32,
-      "learning_rate": 4.311237882051354e-05,
-      "loss": 0.2852,
+      "epoch": 1.3567232493602912,
+      "grad_norm": 0.21513846516609192,
+      "learning_rate": 4.274974192865077e-05,
+      "loss": 0.4229,
       "step": 37645
     },
     {
-      "epoch": 1.32,
-      "learning_rate": 4.311041513660195e-05,
-      "loss": 0.2913,
+      "epoch": 1.356903449021516,
+      "grad_norm": 0.2062099725008011,
+      "learning_rate": 4.274768682275907e-05,
+      "loss": 0.4215,
       "step": 37650
     },
     {
-      "epoch": 1.32,
-      "learning_rate": 4.310845121753857e-05,
-      "loss": 0.2762,
+      "epoch": 1.3570836486827405,
+      "grad_norm": 0.1856852024793625,
+      "learning_rate": 4.274563147505857e-05,
+      "loss": 0.4214,
       "step": 37655
     },
     {
-      "epoch": 1.32,
-      "learning_rate": 4.310648706334888e-05,
-      "loss": 0.3097,
+      "epoch": 1.3572638483439652,
+      "grad_norm": 0.2162654846906662,
+      "learning_rate": 4.2743575885577277e-05,
+      "loss": 0.415,
       "step": 37660
     },
     {
-      "epoch": 1.33,
-      "learning_rate": 4.310452267405839e-05,
-      "loss": 0.2954,
+      "epoch": 1.3574440480051897,
+      "grad_norm": 0.14745785295963287,
+      "learning_rate": 4.27415200543432e-05,
+      "loss": 0.4311,
       "step": 37665
     },
     {
-      "epoch": 1.33,
-      "learning_rate": 4.31025580496926e-05,
-      "loss": 0.2794,
+      "epoch": 1.3576242476664144,
+      "grad_norm": 0.16586333513259888,
+      "learning_rate": 4.2739463981384345e-05,
+      "loss": 0.4549,
       "step": 37670
     },
     {
-      "epoch": 1.33,
-      "learning_rate": 4.310059319027702e-05,
-      "loss": 0.3138,
+      "epoch": 1.357804447327639,
+      "grad_norm": 0.19415488839149475,
+      "learning_rate": 4.2737407666728724e-05,
+      "loss": 0.4612,
       "step": 37675
     },
     {
-      "epoch": 1.33,
-      "learning_rate": 4.309862809583717e-05,
-      "loss": 0.27,
+      "epoch": 1.3579846469888637,
+      "grad_norm": 0.20370012521743774,
+      "learning_rate": 4.2735351110404365e-05,
+      "loss": 0.3613,
       "step": 37680
     },
     {
-      "epoch": 1.33,
-      "learning_rate": 4.309666276639856e-05,
-      "loss": 0.3006,
+      "epoch": 1.3581648466500882,
+      "grad_norm": 0.16060635447502136,
+      "learning_rate": 4.273329431243927e-05,
+      "loss": 0.3881,
       "step": 37685
     },
     {
-      "epoch": 1.33,
-      "learning_rate": 4.309469720198671e-05,
-      "loss": 0.2863,
+      "epoch": 1.358345046311313,
+      "grad_norm": 0.1804780513048172,
+      "learning_rate": 4.273123727286148e-05,
+      "loss": 0.3816,
       "step": 37690
     },
     {
-      "epoch": 1.33,
-      "learning_rate": 4.309273140262715e-05,
-      "loss": 0.2867,
+      "epoch": 1.3585252459725377,
+      "grad_norm": 0.21636547148227692,
+      "learning_rate": 4.272917999169902e-05,
+      "loss": 0.4187,
       "step": 37695
     },
     {
-      "epoch": 1.33,
-      "learning_rate": 4.3090765368345384e-05,
-      "loss": 0.2901,
+      "epoch": 1.3587054456337622,
+      "grad_norm": 0.20415888726711273,
+      "learning_rate": 4.272712246897991e-05,
+      "loss": 0.4345,
       "step": 37700
     },
     {
-      "epoch": 1.33,
-      "learning_rate": 4.308879909916696e-05,
-      "loss": 0.2751,
+      "epoch": 1.358885645294987,
+      "grad_norm": 0.17378808557987213,
+      "learning_rate": 4.272506470473219e-05,
+      "loss": 0.4309,
       "step": 37705
     },
     {
-      "epoch": 1.33,
-      "learning_rate": 4.3086832595117397e-05,
-      "loss": 0.3008,
+      "epoch": 1.3590658449562114,
+      "grad_norm": 0.1880928874015808,
+      "learning_rate": 4.2723006698983894e-05,
+      "loss": 0.421,
       "step": 37710
     },
     {
-      "epoch": 1.33,
-      "learning_rate": 4.3084865856222234e-05,
-      "loss": 0.2783,
+      "epoch": 1.3592460446174361,
+      "grad_norm": 0.18104448914527893,
+      "learning_rate": 4.272094845176307e-05,
+      "loss": 0.4287,
       "step": 37715
     },
     {
-      "epoch": 1.33,
-      "learning_rate": 4.308289888250701e-05,
-      "loss": 0.2733,
+      "epoch": 1.3594262442786609,
+      "grad_norm": 0.17074403166770935,
+      "learning_rate": 4.2718889963097744e-05,
+      "loss": 0.4521,
       "step": 37720
     },
     {
-      "epoch": 1.33,
-      "learning_rate": 4.3080931673997256e-05,
-      "loss": 0.2934,
+      "epoch": 1.3596064439398854,
+      "grad_norm": 0.2121085524559021,
+      "learning_rate": 4.2716831233015974e-05,
+      "loss": 0.4466,
       "step": 37725
     },
     {
-      "epoch": 1.33,
-      "learning_rate": 4.307896423071852e-05,
-      "loss": 0.2764,
+      "epoch": 1.35978664360111,
+      "grad_norm": 0.14753827452659607,
+      "learning_rate": 4.2714772261545813e-05,
+      "loss": 0.4041,
       "step": 37730
     },
     {
-      "epoch": 1.33,
-      "learning_rate": 4.307699655269635e-05,
-      "loss": 0.2962,
+      "epoch": 1.3599668432623346,
+      "grad_norm": 0.18550336360931396,
+      "learning_rate": 4.27127130487153e-05,
+      "loss": 0.4358,
       "step": 37735
     },
     {
-      "epoch": 1.33,
-      "learning_rate": 4.3075028639956295e-05,
-      "loss": 0.2987,
+      "epoch": 1.3601470429235594,
+      "grad_norm": 0.1991180032491684,
+      "learning_rate": 4.2710653594552506e-05,
+      "loss": 0.4421,
       "step": 37740
     },
     {
-      "epoch": 1.33,
-      "learning_rate": 4.30730604925239e-05,
-      "loss": 0.298,
+      "epoch": 1.3603272425847839,
+      "grad_norm": 0.16525886952877045,
+      "learning_rate": 4.2708593899085494e-05,
+      "loss": 0.4052,
       "step": 37745
     },
     {
-      "epoch": 1.33,
-      "learning_rate": 4.3071092110424724e-05,
-      "loss": 0.2706,
+      "epoch": 1.3605074422460086,
+      "grad_norm": 0.17231962084770203,
+      "learning_rate": 4.270653396234231e-05,
+      "loss": 0.4103,
       "step": 37750
     },
     {
-      "epoch": 1.33,
-      "learning_rate": 4.306912349368433e-05,
-      "loss": 0.2879,
+      "epoch": 1.3606876419072331,
+      "grad_norm": 0.1659233570098877,
+      "learning_rate": 4.2704473784351036e-05,
+      "loss": 0.3799,
       "step": 37755
     },
     {
-      "epoch": 1.33,
-      "learning_rate": 4.306715464232828e-05,
-      "loss": 0.2762,
+      "epoch": 1.3608678415684579,
+      "grad_norm": 0.15923145413398743,
+      "learning_rate": 4.2702413365139724e-05,
+      "loss": 0.3827,
       "step": 37760
     },
     {
-      "epoch": 1.33,
-      "learning_rate": 4.3065185556382124e-05,
-      "loss": 0.2766,
+      "epoch": 1.3610480412296826,
+      "grad_norm": 0.13835641741752625,
+      "learning_rate": 4.270035270473647e-05,
+      "loss": 0.3854,
       "step": 37765
     },
     {
-      "epoch": 1.33,
-      "learning_rate": 4.306321623587144e-05,
-      "loss": 0.305,
+      "epoch": 1.361228240890907,
+      "grad_norm": 0.24770118296146393,
+      "learning_rate": 4.269829180316932e-05,
+      "loss": 0.4457,
       "step": 37770
     },
     {
-      "epoch": 1.33,
-      "learning_rate": 4.3061246680821806e-05,
-      "loss": 0.2775,
+      "epoch": 1.3614084405521318,
+      "grad_norm": 0.15679314732551575,
+      "learning_rate": 4.269623066046639e-05,
+      "loss": 0.421,
       "step": 37775
     },
     {
-      "epoch": 1.33,
-      "learning_rate": 4.3059276891258784e-05,
-      "loss": 0.2644,
+      "epoch": 1.3615886402133563,
+      "grad_norm": 0.20325350761413574,
+      "learning_rate": 4.269416927665573e-05,
+      "loss": 0.3852,
       "step": 37780
     },
     {
-      "epoch": 1.33,
-      "learning_rate": 4.3057306867207946e-05,
-      "loss": 0.2929,
+      "epoch": 1.361768839874581,
+      "grad_norm": 0.18375974893569946,
+      "learning_rate": 4.269210765176544e-05,
+      "loss": 0.4165,
       "step": 37785
     },
     {
-      "epoch": 1.33,
-      "learning_rate": 4.3055336608694884e-05,
-      "loss": 0.3116,
+      "epoch": 1.3619490395358056,
+      "grad_norm": 0.1880822330713272,
+      "learning_rate": 4.269004578582362e-05,
+      "loss": 0.4473,
       "step": 37790
     },
     {
-      "epoch": 1.33,
-      "learning_rate": 4.305336611574517e-05,
-      "loss": 0.2964,
+      "epoch": 1.3621292391970303,
+      "grad_norm": 0.16802115738391876,
+      "learning_rate": 4.2687983678858346e-05,
+      "loss": 0.4006,
       "step": 37795
     },
     {
-      "epoch": 1.33,
-      "learning_rate": 4.3051395388384405e-05,
-      "loss": 0.2832,
+      "epoch": 1.3623094388582548,
+      "grad_norm": 0.18482767045497894,
+      "learning_rate": 4.268592133089771e-05,
+      "loss": 0.4042,
       "step": 37800
     },
     {
-      "epoch": 1.33,
-      "learning_rate": 4.3049424426638156e-05,
-      "loss": 0.2944,
+      "epoch": 1.3624896385194796,
+      "grad_norm": 0.18209148943424225,
+      "learning_rate": 4.268385874196983e-05,
+      "loss": 0.419,
       "step": 37805
     },
     {
-      "epoch": 1.33,
-      "learning_rate": 4.304745323053203e-05,
-      "loss": 0.2928,
+      "epoch": 1.3626698381807043,
+      "grad_norm": 0.20778799057006836,
+      "learning_rate": 4.268179591210279e-05,
+      "loss": 0.4544,
       "step": 37810
     },
     {
-      "epoch": 1.33,
-      "learning_rate": 4.304548180009162e-05,
-      "loss": 0.2924,
+      "epoch": 1.3628500378419288,
+      "grad_norm": 0.19366760551929474,
+      "learning_rate": 4.267973284132471e-05,
+      "loss": 0.4157,
       "step": 37815
     },
     {
-      "epoch": 1.33,
-      "learning_rate": 4.304351013534252e-05,
-      "loss": 0.2803,
+      "epoch": 1.3630302375031536,
+      "grad_norm": 0.15186309814453125,
+      "learning_rate": 4.267766952966369e-05,
+      "loss": 0.3843,
       "step": 37820
     },
     {
-      "epoch": 1.33,
-      "learning_rate": 4.304153823631033e-05,
-      "loss": 0.3049,
+      "epoch": 1.363210437164378,
+      "grad_norm": 0.19888971745967865,
+      "learning_rate": 4.267560597714785e-05,
+      "loss": 0.4345,
       "step": 37825
     },
     {
-      "epoch": 1.33,
-      "learning_rate": 4.3039566103020654e-05,
-      "loss": 0.2881,
+      "epoch": 1.3633906368256028,
+      "grad_norm": 0.205244779586792,
+      "learning_rate": 4.2673542183805295e-05,
+      "loss": 0.437,
       "step": 37830
     },
     {
-      "epoch": 1.33,
-      "learning_rate": 4.30375937354991e-05,
-      "loss": 0.2739,
+      "epoch": 1.3635708364868275,
+      "grad_norm": 0.21422690153121948,
+      "learning_rate": 4.267147814966415e-05,
+      "loss": 0.4205,
       "step": 37835
     },
     {
-      "epoch": 1.33,
-      "learning_rate": 4.303562113377129e-05,
-      "loss": 0.2623,
+      "epoch": 1.363751036148052,
+      "grad_norm": 0.15978442132472992,
+      "learning_rate": 4.266941387475254e-05,
+      "loss": 0.4362,
       "step": 37840
     },
     {
-      "epoch": 1.33,
-      "learning_rate": 4.3033648297862814e-05,
-      "loss": 0.2819,
+      "epoch": 1.3639312358092766,
+      "grad_norm": 0.17365151643753052,
+      "learning_rate": 4.2667349359098586e-05,
+      "loss": 0.4162,
       "step": 37845
     },
     {
-      "epoch": 1.33,
-      "learning_rate": 4.303167522779931e-05,
-      "loss": 0.2592,
+      "epoch": 1.3641114354705013,
+      "grad_norm": 0.17353810369968414,
+      "learning_rate": 4.266528460273041e-05,
+      "loss": 0.3909,
       "step": 37850
     },
     {
-      "epoch": 1.33,
-      "learning_rate": 4.302970192360638e-05,
-      "loss": 0.3023,
+      "epoch": 1.364291635131726,
+      "grad_norm": 0.16570428013801575,
+      "learning_rate": 4.266321960567616e-05,
+      "loss": 0.3856,
       "step": 37855
     },
     {
-      "epoch": 1.33,
-      "learning_rate": 4.3027728385309655e-05,
-      "loss": 0.2909,
+      "epoch": 1.3644718347929505,
+      "grad_norm": 0.17112329602241516,
+      "learning_rate": 4.2661154367963965e-05,
+      "loss": 0.4471,
       "step": 37860
     },
     {
-      "epoch": 1.33,
-      "learning_rate": 4.302575461293476e-05,
-      "loss": 0.3025,
+      "epoch": 1.3646520344541753,
+      "grad_norm": 0.14121274650096893,
+      "learning_rate": 4.2659088889621954e-05,
+      "loss": 0.4194,
       "step": 37865
     },
     {
-      "epoch": 1.33,
-      "learning_rate": 4.302378060650731e-05,
-      "loss": 0.2759,
+      "epoch": 1.3648322341153998,
+      "grad_norm": 0.17881150543689728,
+      "learning_rate": 4.265702317067828e-05,
+      "loss": 0.4413,
       "step": 37870
     },
     {
-      "epoch": 1.33,
-      "learning_rate": 4.302180636605295e-05,
-      "loss": 0.2955,
+      "epoch": 1.3650124337766245,
+      "grad_norm": 0.22597812116146088,
+      "learning_rate": 4.2654957211161085e-05,
+      "loss": 0.4434,
       "step": 37875
     },
     {
-      "epoch": 1.33,
-      "learning_rate": 4.301983189159731e-05,
-      "loss": 0.2767,
+      "epoch": 1.3651926334378492,
+      "grad_norm": 0.16727763414382935,
+      "learning_rate": 4.2652891011098505e-05,
+      "loss": 0.4055,
       "step": 37880
     },
     {
-      "epoch": 1.33,
-      "learning_rate": 4.301785718316603e-05,
-      "loss": 0.3205,
+      "epoch": 1.3653728330990738,
+      "grad_norm": 0.19390372931957245,
+      "learning_rate": 4.265082457051872e-05,
+      "loss": 0.3988,
       "step": 37885
     },
     {
-      "epoch": 1.33,
-      "learning_rate": 4.3015882240784755e-05,
-      "loss": 0.3078,
+      "epoch": 1.3655530327602985,
+      "grad_norm": 0.177637979388237,
+      "learning_rate": 4.264875788944985e-05,
+      "loss": 0.411,
       "step": 37890
     },
     {
-      "epoch": 1.33,
-      "learning_rate": 4.301390706447912e-05,
-      "loss": 0.2932,
+      "epoch": 1.365733232421523,
+      "grad_norm": 0.17487314343452454,
+      "learning_rate": 4.2646690967920086e-05,
+      "loss": 0.4127,
       "step": 37895
     },
     {
-      "epoch": 1.33,
-      "learning_rate": 4.3011931654274776e-05,
-      "loss": 0.292,
+      "epoch": 1.3659134320827477,
+      "grad_norm": 0.203982412815094,
+      "learning_rate": 4.264462380595756e-05,
+      "loss": 0.4603,
       "step": 37900
     },
     {
-      "epoch": 1.33,
-      "learning_rate": 4.300995601019736e-05,
-      "loss": 0.2999,
+      "epoch": 1.3660936317439722,
+      "grad_norm": 0.23240825533866882,
+      "learning_rate": 4.264255640359046e-05,
+      "loss": 0.418,
       "step": 37905
     },
     {
-      "epoch": 1.33,
-      "learning_rate": 4.300798013227254e-05,
-      "loss": 0.2601,
+      "epoch": 1.366273831405197,
+      "grad_norm": 0.15722376108169556,
+      "learning_rate": 4.2640488760846945e-05,
+      "loss": 0.4449,
       "step": 37910
     },
     {
-      "epoch": 1.33,
-      "learning_rate": 4.300600402052597e-05,
-      "loss": 0.2811,
+      "epoch": 1.3664540310664215,
+      "grad_norm": 0.14210271835327148,
+      "learning_rate": 4.263842087775518e-05,
+      "loss": 0.3789,
       "step": 37915
     },
     {
-      "epoch": 1.33,
-      "learning_rate": 4.3004027674983294e-05,
-      "loss": 0.2905,
+      "epoch": 1.3666342307276462,
+      "grad_norm": 0.15911325812339783,
+      "learning_rate": 4.263635275434336e-05,
+      "loss": 0.4281,
       "step": 37920
     },
     {
-      "epoch": 1.33,
-      "learning_rate": 4.3002051095670195e-05,
-      "loss": 0.2913,
+      "epoch": 1.366814430388871,
+      "grad_norm": 0.17531517148017883,
+      "learning_rate": 4.263428439063963e-05,
+      "loss": 0.4233,
       "step": 37925
     },
     {
-      "epoch": 1.33,
-      "learning_rate": 4.300007428261231e-05,
-      "loss": 0.2848,
+      "epoch": 1.3669946300500955,
+      "grad_norm": 0.16672587394714355,
+      "learning_rate": 4.263221578667219e-05,
+      "loss": 0.4255,
       "step": 37930
     },
     {
-      "epoch": 1.33,
-      "learning_rate": 4.299809723583534e-05,
-      "loss": 0.2817,
+      "epoch": 1.3671748297113202,
+      "grad_norm": 0.21964746713638306,
+      "learning_rate": 4.263014694246924e-05,
+      "loss": 0.4298,
       "step": 37935
     },
     {
-      "epoch": 1.33,
-      "learning_rate": 4.2996119955364936e-05,
-      "loss": 0.2777,
+      "epoch": 1.3673550293725447,
+      "grad_norm": 0.161033034324646,
+      "learning_rate": 4.262807785805894e-05,
+      "loss": 0.3999,
       "step": 37940
     },
     {
-      "epoch": 1.34,
-      "learning_rate": 4.2994142441226764e-05,
-      "loss": 0.2785,
+      "epoch": 1.3675352290337695,
+      "grad_norm": 0.20498016476631165,
+      "learning_rate": 4.262600853346949e-05,
+      "loss": 0.4068,
       "step": 37945
     },
     {
-      "epoch": 1.34,
-      "learning_rate": 4.299216469344652e-05,
-      "loss": 0.2951,
+      "epoch": 1.3677154286949942,
+      "grad_norm": 0.20285950601100922,
+      "learning_rate": 4.262393896872909e-05,
+      "loss": 0.4023,
       "step": 37950
     },
     {
-      "epoch": 1.34,
-      "learning_rate": 4.2990186712049874e-05,
-      "loss": 0.3124,
+      "epoch": 1.3678956283562187,
+      "grad_norm": 0.16656413674354553,
+      "learning_rate": 4.262186916386594e-05,
+      "loss": 0.4094,
       "step": 37955
     },
     {
-      "epoch": 1.34,
-      "learning_rate": 4.2988208497062506e-05,
-      "loss": 0.3078,
+      "epoch": 1.3680758280174432,
+      "grad_norm": 0.1727702021598816,
+      "learning_rate": 4.261979911890822e-05,
+      "loss": 0.3972,
       "step": 37960
     },
     {
-      "epoch": 1.34,
-      "learning_rate": 4.29862300485101e-05,
-      "loss": 0.2748,
+      "epoch": 1.368256027678668,
+      "grad_norm": 0.14835266768932343,
+      "learning_rate": 4.261772883388416e-05,
+      "loss": 0.4069,
       "step": 37965
     },
     {
-      "epoch": 1.34,
-      "learning_rate": 4.298425136641836e-05,
-      "loss": 0.2856,
+      "epoch": 1.3684362273398927,
+      "grad_norm": 0.22121186554431915,
+      "learning_rate": 4.261565830882195e-05,
+      "loss": 0.4128,
       "step": 37970
     },
     {
-      "epoch": 1.34,
-      "learning_rate": 4.2982272450812964e-05,
-      "loss": 0.2934,
+      "epoch": 1.3686164270011172,
+      "grad_norm": 0.15944576263427734,
+      "learning_rate": 4.261358754374981e-05,
+      "loss": 0.4228,
       "step": 37975
     },
     {
-      "epoch": 1.34,
-      "learning_rate": 4.2980293301719614e-05,
-      "loss": 0.293,
+      "epoch": 1.368796626662342,
+      "grad_norm": 0.19683369994163513,
+      "learning_rate": 4.261151653869595e-05,
+      "loss": 0.436,
       "step": 37980
     },
     {
-      "epoch": 1.34,
-      "learning_rate": 4.297831391916401e-05,
-      "loss": 0.272,
+      "epoch": 1.3689768263235664,
+      "grad_norm": 0.16129469871520996,
+      "learning_rate": 4.260944529368858e-05,
+      "loss": 0.4114,
       "step": 37985
     },
     {
-      "epoch": 1.34,
-      "learning_rate": 4.297633430317183e-05,
-      "loss": 0.2812,
+      "epoch": 1.3691570259847912,
+      "grad_norm": 0.18280497193336487,
+      "learning_rate": 4.260737380875593e-05,
+      "loss": 0.4122,
       "step": 37990
     },
     {
-      "epoch": 1.34,
-      "learning_rate": 4.297435445376881e-05,
-      "loss": 0.2859,
+      "epoch": 1.369337225646016,
+      "grad_norm": 0.20879638195037842,
+      "learning_rate": 4.260530208392622e-05,
+      "loss": 0.3891,
       "step": 37995
     },
     {
-      "epoch": 1.34,
-      "learning_rate": 4.297237437098064e-05,
-      "loss": 0.3075,
+      "epoch": 1.3695174253072404,
+      "grad_norm": 0.21062108874320984,
+      "learning_rate": 4.260323011922768e-05,
+      "loss": 0.3986,
       "step": 38000
     },
     {
-      "epoch": 1.34,
-      "eval_loss": 0.28374922275543213,
-      "eval_runtime": 10.5368,
-      "eval_samples_per_second": 9.491,
-      "eval_steps_per_second": 9.491,
+      "epoch": 1.3695174253072404,
+      "eval_loss": 0.4474603235721588,
+      "eval_runtime": 3.5425,
+      "eval_samples_per_second": 28.229,
+      "eval_steps_per_second": 7.057,
       "step": 38000
     },
     {
-      "epoch": 1.34,
-      "learning_rate": 4.297039405483303e-05,
-      "loss": 0.2702,
+      "epoch": 1.369697624968465,
+      "grad_norm": 0.22914621233940125,
+      "learning_rate": 4.2601157914688535e-05,
+      "loss": 0.418,
       "step": 38005
     },
     {
-      "epoch": 1.34,
-      "learning_rate": 4.2968413505351704e-05,
-      "loss": 0.2736,
+      "epoch": 1.3698778246296897,
+      "grad_norm": 0.2014915496110916,
+      "learning_rate": 4.259908547033703e-05,
+      "loss": 0.4176,
       "step": 38010
     },
     {
-      "epoch": 1.34,
-      "learning_rate": 4.296643272256237e-05,
-      "loss": 0.2756,
+      "epoch": 1.3700580242909144,
+      "grad_norm": 0.18956822156906128,
+      "learning_rate": 4.2597012786201384e-05,
+      "loss": 0.414,
       "step": 38015
     },
     {
-      "epoch": 1.34,
-      "learning_rate": 4.296445170649074e-05,
-      "loss": 0.2682,
+      "epoch": 1.370238223952139,
+      "grad_norm": 0.18128299713134766,
+      "learning_rate": 4.2594939862309845e-05,
+      "loss": 0.4348,
       "step": 38020
     },
     {
-      "epoch": 1.34,
-      "learning_rate": 4.296247045716254e-05,
-      "loss": 0.2696,
+      "epoch": 1.3704184236133636,
+      "grad_norm": 0.22281454503536224,
+      "learning_rate": 4.259286669869066e-05,
+      "loss": 0.4026,
       "step": 38025
     },
     {
-      "epoch": 1.34,
-      "learning_rate": 4.2960488974603506e-05,
-      "loss": 0.2754,
+      "epoch": 1.3705986232745881,
+      "grad_norm": 0.21276813745498657,
+      "learning_rate": 4.259079329537208e-05,
+      "loss": 0.4402,
       "step": 38030
     },
     {
-      "epoch": 1.34,
-      "learning_rate": 4.295850725883936e-05,
-      "loss": 0.2919,
+      "epoch": 1.3707788229358129,
+      "grad_norm": 0.14526169002056122,
+      "learning_rate": 4.2588719652382336e-05,
+      "loss": 0.4397,
       "step": 38035
     },
     {
-      "epoch": 1.34,
-      "learning_rate": 4.295652530989583e-05,
-      "loss": 0.2839,
+      "epoch": 1.3709590225970376,
+      "grad_norm": 0.18278780579566956,
+      "learning_rate": 4.25866457697497e-05,
+      "loss": 0.393,
       "step": 38040
     },
     {
-      "epoch": 1.34,
-      "learning_rate": 4.295454312779865e-05,
-      "loss": 0.2828,
+      "epoch": 1.3711392222582621,
+      "grad_norm": 0.15444017946720123,
+      "learning_rate": 4.2584571647502417e-05,
+      "loss": 0.427,
       "step": 38045
     },
     {
-      "epoch": 1.34,
-      "learning_rate": 4.2952560712573565e-05,
-      "loss": 0.2947,
+      "epoch": 1.3713194219194869,
+      "grad_norm": 0.1620875895023346,
+      "learning_rate": 4.2582497285668746e-05,
+      "loss": 0.405,
       "step": 38050
     },
     {
-      "epoch": 1.34,
-      "learning_rate": 4.2950578064246305e-05,
-      "loss": 0.2927,
+      "epoch": 1.3714996215807114,
+      "grad_norm": 0.15654486417770386,
+      "learning_rate": 4.258042268427695e-05,
+      "loss": 0.3709,
       "step": 38055
     },
     {
-      "epoch": 1.34,
-      "learning_rate": 4.294859518284263e-05,
-      "loss": 0.3115,
+      "epoch": 1.371679821241936,
+      "grad_norm": 0.14211991429328918,
+      "learning_rate": 4.257834784335531e-05,
+      "loss": 0.4132,
       "step": 38060
     },
     {
-      "epoch": 1.34,
-      "learning_rate": 4.294661206838827e-05,
-      "loss": 0.3096,
+      "epoch": 1.3718600209031608,
+      "grad_norm": 0.17219579219818115,
+      "learning_rate": 4.257627276293208e-05,
+      "loss": 0.4233,
       "step": 38065
     },
     {
-      "epoch": 1.34,
-      "learning_rate": 4.294462872090897e-05,
-      "loss": 0.2757,
+      "epoch": 1.3720402205643853,
+      "grad_norm": 0.19686567783355713,
+      "learning_rate": 4.257419744303553e-05,
+      "loss": 0.4242,
       "step": 38070
     },
     {
-      "epoch": 1.34,
-      "learning_rate": 4.29426451404305e-05,
-      "loss": 0.2621,
+      "epoch": 1.3722204202256099,
+      "grad_norm": 0.1756589412689209,
+      "learning_rate": 4.257212188369395e-05,
+      "loss": 0.4201,
       "step": 38075
     },
     {
-      "epoch": 1.34,
-      "learning_rate": 4.2940661326978604e-05,
-      "loss": 0.2941,
+      "epoch": 1.3724006198868346,
+      "grad_norm": 0.17567414045333862,
+      "learning_rate": 4.257004608493561e-05,
+      "loss": 0.4198,
       "step": 38080
     },
     {
-      "epoch": 1.34,
-      "learning_rate": 4.293867728057905e-05,
-      "loss": 0.288,
+      "epoch": 1.3725808195480593,
+      "grad_norm": 0.2001543641090393,
+      "learning_rate": 4.256797004678879e-05,
+      "loss": 0.425,
       "step": 38085
     },
     {
-      "epoch": 1.34,
-      "learning_rate": 4.2936693001257586e-05,
-      "loss": 0.295,
+      "epoch": 1.3727610192092838,
+      "grad_norm": 0.20645958185195923,
+      "learning_rate": 4.2565893769281787e-05,
+      "loss": 0.4048,
       "step": 38090
     },
     {
-      "epoch": 1.34,
-      "learning_rate": 4.293470848903999e-05,
-      "loss": 0.3041,
+      "epoch": 1.3729412188705086,
+      "grad_norm": 0.18848280608654022,
+      "learning_rate": 4.256381725244287e-05,
+      "loss": 0.4125,
       "step": 38095
     },
     {
-      "epoch": 1.34,
-      "learning_rate": 4.293272374395202e-05,
-      "loss": 0.3014,
+      "epoch": 1.373121418531733,
+      "grad_norm": 0.16142655909061432,
+      "learning_rate": 4.2561740496300353e-05,
+      "loss": 0.3701,
       "step": 38100
     },
     {
-      "epoch": 1.34,
-      "learning_rate": 4.293073876601945e-05,
-      "loss": 0.245,
+      "epoch": 1.3733016181929578,
+      "grad_norm": 0.17975199222564697,
+      "learning_rate": 4.2559663500882515e-05,
+      "loss": 0.4448,
       "step": 38105
     },
     {
-      "epoch": 1.34,
-      "learning_rate": 4.292875355526806e-05,
-      "loss": 0.2874,
+      "epoch": 1.3734818178541826,
+      "grad_norm": 0.16210007667541504,
+      "learning_rate": 4.255758626621767e-05,
+      "loss": 0.4238,
       "step": 38110
     },
     {
-      "epoch": 1.34,
-      "learning_rate": 4.2926768111723615e-05,
-      "loss": 0.2582,
+      "epoch": 1.373662017515407,
+      "grad_norm": 0.1924877017736435,
+      "learning_rate": 4.2555508792334105e-05,
+      "loss": 0.42,
       "step": 38115
     },
     {
-      "epoch": 1.34,
-      "learning_rate": 4.29247824354119e-05,
-      "loss": 0.307,
+      "epoch": 1.3738422171766316,
+      "grad_norm": 0.18068957328796387,
+      "learning_rate": 4.255343107926013e-05,
+      "loss": 0.378,
       "step": 38120
     },
     {
-      "epoch": 1.34,
-      "learning_rate": 4.29227965263587e-05,
-      "loss": 0.2779,
+      "epoch": 1.3740224168378563,
+      "grad_norm": 0.21288029849529266,
+      "learning_rate": 4.255135312702406e-05,
+      "loss": 0.427,
       "step": 38125
     },
     {
-      "epoch": 1.34,
-      "learning_rate": 4.29208103845898e-05,
-      "loss": 0.3095,
+      "epoch": 1.374202616499081,
+      "grad_norm": 0.18913322687149048,
+      "learning_rate": 4.25492749356542e-05,
+      "loss": 0.4031,
       "step": 38130
     },
     {
-      "epoch": 1.34,
-      "learning_rate": 4.291882401013099e-05,
-      "loss": 0.2944,
+      "epoch": 1.3743828161603056,
+      "grad_norm": 0.1928529441356659,
+      "learning_rate": 4.2547196505178866e-05,
+      "loss": 0.3974,
       "step": 38135
     },
     {
-      "epoch": 1.34,
-      "learning_rate": 4.291683740300805e-05,
-      "loss": 0.2929,
+      "epoch": 1.3745630158215303,
+      "grad_norm": 0.186807319521904,
+      "learning_rate": 4.254511783562638e-05,
+      "loss": 0.4134,
       "step": 38140
     },
     {
-      "epoch": 1.34,
-      "learning_rate": 4.291485056324679e-05,
-      "loss": 0.2837,
+      "epoch": 1.3747432154827548,
+      "grad_norm": 0.19627200067043304,
+      "learning_rate": 4.254303892702506e-05,
+      "loss": 0.4457,
       "step": 38145
     },
     {
-      "epoch": 1.34,
-      "learning_rate": 4.2912863490873e-05,
-      "loss": 0.2728,
+      "epoch": 1.3749234151439795,
+      "grad_norm": 0.17085368931293488,
+      "learning_rate": 4.254095977940323e-05,
+      "loss": 0.4452,
       "step": 38150
     },
     {
-      "epoch": 1.34,
-      "learning_rate": 4.291087618591249e-05,
-      "loss": 0.2735,
+      "epoch": 1.3751036148052043,
+      "grad_norm": 0.15709124505519867,
+      "learning_rate": 4.2538880392789214e-05,
+      "loss": 0.4101,
       "step": 38155
     },
     {
-      "epoch": 1.34,
-      "learning_rate": 4.290888864839104e-05,
-      "loss": 0.2843,
+      "epoch": 1.3752838144664288,
+      "grad_norm": 0.2019108682870865,
+      "learning_rate": 4.253680076721136e-05,
+      "loss": 0.403,
       "step": 38160
     },
     {
-      "epoch": 1.34,
-      "learning_rate": 4.290690087833449e-05,
-      "loss": 0.267,
+      "epoch": 1.3754640141276535,
+      "grad_norm": 0.17493942379951477,
+      "learning_rate": 4.253472090269798e-05,
+      "loss": 0.4093,
       "step": 38165
     },
     {
-      "epoch": 1.34,
-      "learning_rate": 4.2904912875768624e-05,
-      "loss": 0.3167,
+      "epoch": 1.375644213788878,
+      "grad_norm": 0.18124589323997498,
+      "learning_rate": 4.253264079927742e-05,
+      "loss": 0.4191,
       "step": 38170
     },
     {
-      "epoch": 1.34,
-      "learning_rate": 4.2902924640719267e-05,
-      "loss": 0.2905,
+      "epoch": 1.3758244134501028,
+      "grad_norm": 0.20440387725830078,
+      "learning_rate": 4.253056045697803e-05,
+      "loss": 0.4026,
       "step": 38175
     },
     {
-      "epoch": 1.34,
-      "learning_rate": 4.290093617321223e-05,
-      "loss": 0.2805,
+      "epoch": 1.3760046131113273,
+      "grad_norm": 0.2788972854614258,
+      "learning_rate": 4.252847987582815e-05,
+      "loss": 0.4271,
       "step": 38180
     },
     {
-      "epoch": 1.34,
-      "learning_rate": 4.289894747327333e-05,
-      "loss": 0.2999,
+      "epoch": 1.376184812772552,
+      "grad_norm": 0.20534437894821167,
+      "learning_rate": 4.252639905585613e-05,
+      "loss": 0.4404,
       "step": 38185
     },
     {
-      "epoch": 1.34,
-      "learning_rate": 4.28969585409284e-05,
-      "loss": 0.2963,
+      "epoch": 1.3763650124337765,
+      "grad_norm": 0.18792025744915009,
+      "learning_rate": 4.2524317997090304e-05,
+      "loss": 0.427,
       "step": 38190
     },
     {
-      "epoch": 1.34,
-      "learning_rate": 4.2894969376203257e-05,
-      "loss": 0.2979,
+      "epoch": 1.3765452120950012,
+      "grad_norm": 0.2014329433441162,
+      "learning_rate": 4.2522236699559045e-05,
+      "loss": 0.4211,
       "step": 38195
     },
     {
-      "epoch": 1.34,
-      "learning_rate": 4.289297997912373e-05,
-      "loss": 0.2835,
+      "epoch": 1.376725411756226,
+      "grad_norm": 0.1882011443376541,
+      "learning_rate": 4.25201551632907e-05,
+      "loss": 0.431,
       "step": 38200
     },
     {
-      "epoch": 1.34,
-      "learning_rate": 4.289099034971565e-05,
-      "loss": 0.2844,
+      "epoch": 1.3769056114174505,
+      "grad_norm": 0.1633293330669403,
+      "learning_rate": 4.251807338831363e-05,
+      "loss": 0.4322,
       "step": 38205
     },
     {
-      "epoch": 1.34,
-      "learning_rate": 4.288900048800485e-05,
-      "loss": 0.303,
+      "epoch": 1.3770858110786752,
+      "grad_norm": 0.1637231707572937,
+      "learning_rate": 4.2515991374656204e-05,
+      "loss": 0.3997,
       "step": 38210
     },
     {
-      "epoch": 1.34,
-      "learning_rate": 4.288701039401717e-05,
-      "loss": 0.291,
+      "epoch": 1.3772660107398997,
+      "grad_norm": 0.19281378388404846,
+      "learning_rate": 4.251390912234679e-05,
+      "loss": 0.4273,
       "step": 38215
     },
     {
-      "epoch": 1.34,
-      "learning_rate": 4.288502006777844e-05,
-      "loss": 0.2979,
+      "epoch": 1.3774462104011245,
+      "grad_norm": 0.18837355077266693,
+      "learning_rate": 4.251182663141375e-05,
+      "loss": 0.4359,
       "step": 38220
     },
     {
-      "epoch": 1.34,
-      "learning_rate": 4.288302950931452e-05,
-      "loss": 0.2875,
+      "epoch": 1.3776264100623492,
+      "grad_norm": 0.17957091331481934,
+      "learning_rate": 4.2509743901885474e-05,
+      "loss": 0.4057,
       "step": 38225
     },
     {
-      "epoch": 1.35,
-      "learning_rate": 4.288103871865124e-05,
-      "loss": 0.286,
+      "epoch": 1.3778066097235737,
+      "grad_norm": 0.1847553253173828,
+      "learning_rate": 4.2507660933790314e-05,
+      "loss": 0.44,
       "step": 38230
     },
     {
-      "epoch": 1.35,
-      "learning_rate": 4.287904769581446e-05,
-      "loss": 0.2736,
+      "epoch": 1.3779868093847982,
+      "grad_norm": 0.16359250247478485,
+      "learning_rate": 4.250557772715667e-05,
+      "loss": 0.3645,
       "step": 38235
     },
     {
-      "epoch": 1.35,
-      "learning_rate": 4.287705644083003e-05,
-      "loss": 0.277,
+      "epoch": 1.378167009046023,
+      "grad_norm": 0.18574312329292297,
+      "learning_rate": 4.250349428201292e-05,
+      "loss": 0.4006,
       "step": 38240
     },
     {
-      "epoch": 1.35,
-      "learning_rate": 4.28750649537238e-05,
-      "loss": 0.2884,
+      "epoch": 1.3783472087072477,
+      "grad_norm": 0.16163067519664764,
+      "learning_rate": 4.250141059838745e-05,
+      "loss": 0.4349,
       "step": 38245
     },
     {
-      "epoch": 1.35,
-      "learning_rate": 4.287307323452163e-05,
-      "loss": 0.2813,
+      "epoch": 1.3785274083684722,
+      "grad_norm": 0.15968047082424164,
+      "learning_rate": 4.249932667630864e-05,
+      "loss": 0.4231,
       "step": 38250
     },
     {
-      "epoch": 1.35,
-      "learning_rate": 4.2871081283249395e-05,
-      "loss": 0.2952,
+      "epoch": 1.378707608029697,
+      "grad_norm": 0.17597581446170807,
+      "learning_rate": 4.24972425158049e-05,
+      "loss": 0.402,
       "step": 38255
     },
     {
-      "epoch": 1.35,
-      "learning_rate": 4.2869089099932936e-05,
-      "loss": 0.288,
+      "epoch": 1.3788878076909215,
+      "grad_norm": 0.19579049944877625,
+      "learning_rate": 4.249515811690461e-05,
+      "loss": 0.3995,
       "step": 38260
     },
     {
-      "epoch": 1.35,
-      "learning_rate": 4.2867096684598134e-05,
-      "loss": 0.2726,
+      "epoch": 1.3790680073521462,
+      "grad_norm": 0.1438038945198059,
+      "learning_rate": 4.249307347963619e-05,
+      "loss": 0.4186,
       "step": 38265
     },
     {
-      "epoch": 1.35,
-      "learning_rate": 4.286510403727086e-05,
-      "loss": 0.3139,
+      "epoch": 1.379248207013371,
+      "grad_norm": 0.19650869071483612,
+      "learning_rate": 4.249098860402802e-05,
+      "loss": 0.4032,
       "step": 38270
     },
     {
-      "epoch": 1.35,
-      "learning_rate": 4.2863111157976974e-05,
-      "loss": 0.3018,
+      "epoch": 1.3794284066745954,
+      "grad_norm": 0.16172637045383453,
+      "learning_rate": 4.2488903490108524e-05,
+      "loss": 0.4299,
       "step": 38275
     },
     {
-      "epoch": 1.35,
-      "learning_rate": 4.286111804674237e-05,
-      "loss": 0.2958,
+      "epoch": 1.3796086063358202,
+      "grad_norm": 0.21487955749034882,
+      "learning_rate": 4.2486818137906095e-05,
+      "loss": 0.4782,
       "step": 38280
     },
     {
-      "epoch": 1.35,
-      "learning_rate": 4.2859124703592915e-05,
-      "loss": 0.2771,
+      "epoch": 1.3797888059970447,
+      "grad_norm": 0.20923011004924774,
+      "learning_rate": 4.248473254744917e-05,
+      "loss": 0.442,
       "step": 38285
     },
     {
-      "epoch": 1.35,
-      "learning_rate": 4.28571311285545e-05,
-      "loss": 0.3014,
+      "epoch": 1.3799690056582694,
+      "grad_norm": 0.15938104689121246,
+      "learning_rate": 4.2482646718766136e-05,
+      "loss": 0.4118,
       "step": 38290
     },
     {
-      "epoch": 1.35,
-      "learning_rate": 4.2855137321653007e-05,
-      "loss": 0.2514,
+      "epoch": 1.380149205319494,
+      "grad_norm": 0.1880597025156021,
+      "learning_rate": 4.2480560651885425e-05,
+      "loss": 0.4324,
       "step": 38295
     },
     {
-      "epoch": 1.35,
-      "learning_rate": 4.285314328291431e-05,
-      "loss": 0.3112,
+      "epoch": 1.3803294049807187,
+      "grad_norm": 0.1785939335823059,
+      "learning_rate": 4.247847434683546e-05,
+      "loss": 0.4128,
       "step": 38300
     },
     {
-      "epoch": 1.35,
-      "learning_rate": 4.285114901236433e-05,
-      "loss": 0.2921,
+      "epoch": 1.3805096046419432,
+      "grad_norm": 0.2147594839334488,
+      "learning_rate": 4.247638780364468e-05,
+      "loss": 0.3871,
       "step": 38305
     },
     {
-      "epoch": 1.35,
-      "learning_rate": 4.284915451002893e-05,
-      "loss": 0.2843,
+      "epoch": 1.380689804303168,
+      "grad_norm": 0.13881224393844604,
+      "learning_rate": 4.247430102234149e-05,
+      "loss": 0.4052,
       "step": 38310
     },
     {
-      "epoch": 1.35,
-      "learning_rate": 4.284715977593403e-05,
-      "loss": 0.282,
+      "epoch": 1.3808700039643926,
+      "grad_norm": 0.18304836750030518,
+      "learning_rate": 4.2472214002954324e-05,
+      "loss": 0.3941,
       "step": 38315
     },
     {
-      "epoch": 1.35,
-      "learning_rate": 4.284516481010552e-05,
-      "loss": 0.2915,
+      "epoch": 1.3810502036256171,
+      "grad_norm": 0.1778329461812973,
+      "learning_rate": 4.247012674551163e-05,
+      "loss": 0.4741,
       "step": 38320
     },
     {
-      "epoch": 1.35,
-      "learning_rate": 4.2843169612569306e-05,
-      "loss": 0.2836,
+      "epoch": 1.3812304032868419,
+      "grad_norm": 0.18518294394016266,
+      "learning_rate": 4.246803925004185e-05,
+      "loss": 0.443,
       "step": 38325
     },
     {
-      "epoch": 1.35,
-      "learning_rate": 4.2841174183351294e-05,
-      "loss": 0.2788,
+      "epoch": 1.3814106029480664,
+      "grad_norm": 0.21796804666519165,
+      "learning_rate": 4.2465951516573406e-05,
+      "loss": 0.4183,
       "step": 38330
     },
     {
-      "epoch": 1.35,
-      "learning_rate": 4.2839178522477394e-05,
-      "loss": 0.295,
+      "epoch": 1.3815908026092911,
+      "grad_norm": 0.2137499749660492,
+      "learning_rate": 4.246386354513475e-05,
+      "loss": 0.4216,
       "step": 38335
     },
     {
-      "epoch": 1.35,
-      "learning_rate": 4.283718262997352e-05,
-      "loss": 0.2832,
+      "epoch": 1.3817710022705159,
+      "grad_norm": 0.18902204930782318,
+      "learning_rate": 4.246177533575435e-05,
+      "loss": 0.4524,
       "step": 38340
     },
     {
-      "epoch": 1.35,
-      "learning_rate": 4.2835186505865574e-05,
-      "loss": 0.3233,
+      "epoch": 1.3819512019317404,
+      "grad_norm": 0.171752467751503,
+      "learning_rate": 4.2459686888460635e-05,
+      "loss": 0.4111,
       "step": 38345
     },
     {
-      "epoch": 1.35,
-      "learning_rate": 4.283319015017949e-05,
-      "loss": 0.2912,
+      "epoch": 1.3821314015929649,
+      "grad_norm": 0.15336576104164124,
+      "learning_rate": 4.245759820328206e-05,
+      "loss": 0.3991,
       "step": 38350
     },
     {
-      "epoch": 1.35,
-      "learning_rate": 4.2831193562941185e-05,
-      "loss": 0.2738,
+      "epoch": 1.3823116012541896,
+      "grad_norm": 0.19525669515132904,
+      "learning_rate": 4.2455509280247097e-05,
+      "loss": 0.4658,
       "step": 38355
     },
     {
-      "epoch": 1.35,
-      "learning_rate": 4.282919674417658e-05,
-      "loss": 0.2797,
+      "epoch": 1.3824918009154143,
+      "grad_norm": 0.15848879516124725,
+      "learning_rate": 4.2453420119384195e-05,
+      "loss": 0.427,
       "step": 38360
     },
     {
-      "epoch": 1.35,
-      "learning_rate": 4.2827199693911616e-05,
-      "loss": 0.3054,
+      "epoch": 1.3826720005766389,
+      "grad_norm": 0.16411955654621124,
+      "learning_rate": 4.245133072072183e-05,
+      "loss": 0.4179,
       "step": 38365
     },
     {
-      "epoch": 1.35,
-      "learning_rate": 4.28252024121722e-05,
-      "loss": 0.2882,
+      "epoch": 1.3828522002378636,
+      "grad_norm": 0.21341267228126526,
+      "learning_rate": 4.244924108428846e-05,
+      "loss": 0.4251,
       "step": 38370
     },
     {
-      "epoch": 1.35,
-      "learning_rate": 4.282320489898428e-05,
-      "loss": 0.2655,
+      "epoch": 1.383032399899088,
+      "grad_norm": 0.16351541876792908,
+      "learning_rate": 4.2447151210112555e-05,
+      "loss": 0.4354,
       "step": 38375
     },
     {
-      "epoch": 1.35,
-      "learning_rate": 4.2821207154373796e-05,
-      "loss": 0.2844,
+      "epoch": 1.3832125995603128,
+      "grad_norm": 0.176330104470253,
+      "learning_rate": 4.2445061098222596e-05,
+      "loss": 0.4136,
       "step": 38380
     },
     {
-      "epoch": 1.35,
-      "learning_rate": 4.281920917836668e-05,
-      "loss": 0.2826,
+      "epoch": 1.3833927992215376,
+      "grad_norm": 0.16498027741909027,
+      "learning_rate": 4.244297074864705e-05,
+      "loss": 0.4119,
       "step": 38385
     },
     {
-      "epoch": 1.35,
-      "learning_rate": 4.2817210970988875e-05,
-      "loss": 0.2975,
+      "epoch": 1.383572998882762,
+      "grad_norm": 0.20347841084003448,
+      "learning_rate": 4.244088016141441e-05,
+      "loss": 0.4565,
       "step": 38390
     },
     {
-      "epoch": 1.35,
-      "learning_rate": 4.2815212532266325e-05,
-      "loss": 0.2714,
+      "epoch": 1.3837531985439868,
+      "grad_norm": 0.20916853845119476,
+      "learning_rate": 4.2438789336553154e-05,
+      "loss": 0.4209,
       "step": 38395
     },
     {
-      "epoch": 1.35,
-      "learning_rate": 4.2813213862224986e-05,
-      "loss": 0.2979,
+      "epoch": 1.3839333982052113,
+      "grad_norm": 0.20273546874523163,
+      "learning_rate": 4.2436698274091765e-05,
+      "loss": 0.4065,
       "step": 38400
     },
     {
-      "epoch": 1.35,
-      "learning_rate": 4.28112149608908e-05,
-      "loss": 0.2812,
+      "epoch": 1.384113597866436,
+      "grad_norm": 0.15745113790035248,
+      "learning_rate": 4.2434606974058756e-05,
+      "loss": 0.3952,
       "step": 38405
     },
     {
-      "epoch": 1.35,
-      "learning_rate": 4.2809215828289725e-05,
-      "loss": 0.2823,
+      "epoch": 1.3842937975276606,
+      "grad_norm": 0.20275966823101044,
+      "learning_rate": 4.243251543648258e-05,
+      "loss": 0.3806,
       "step": 38410
     },
     {
-      "epoch": 1.35,
-      "learning_rate": 4.2807216464447724e-05,
-      "loss": 0.282,
+      "epoch": 1.3844739971888853,
+      "grad_norm": 0.15785571932792664,
+      "learning_rate": 4.243042366139177e-05,
+      "loss": 0.4033,
       "step": 38415
     },
     {
-      "epoch": 1.35,
-      "learning_rate": 4.280521686939075e-05,
-      "loss": 0.3143,
+      "epoch": 1.3846541968501098,
+      "grad_norm": 0.19757528603076935,
+      "learning_rate": 4.24283316488148e-05,
+      "loss": 0.3708,
       "step": 38420
     },
     {
-      "epoch": 1.35,
-      "learning_rate": 4.280321704314477e-05,
-      "loss": 0.3048,
+      "epoch": 1.3848343965113346,
+      "grad_norm": 0.2014017403125763,
+      "learning_rate": 4.24262393987802e-05,
+      "loss": 0.4544,
       "step": 38425
     },
     {
-      "epoch": 1.35,
-      "learning_rate": 4.280121698573575e-05,
-      "loss": 0.3021,
+      "epoch": 1.3850145961725593,
+      "grad_norm": 0.20693495869636536,
+      "learning_rate": 4.242414691131645e-05,
+      "loss": 0.4509,
       "step": 38430
     },
     {
-      "epoch": 1.35,
-      "learning_rate": 4.279921669718966e-05,
-      "loss": 0.3082,
+      "epoch": 1.3851947958337838,
+      "grad_norm": 0.18149085342884064,
+      "learning_rate": 4.242205418645208e-05,
+      "loss": 0.4171,
       "step": 38435
     },
     {
-      "epoch": 1.35,
-      "learning_rate": 4.279721617753247e-05,
-      "loss": 0.2914,
+      "epoch": 1.3853749954950085,
+      "grad_norm": 0.2000938504934311,
+      "learning_rate": 4.2419961224215595e-05,
+      "loss": 0.3862,
       "step": 38440
     },
     {
-      "epoch": 1.35,
-      "learning_rate": 4.2795215426790155e-05,
-      "loss": 0.2763,
+      "epoch": 1.385555195156233,
+      "grad_norm": 0.19279134273529053,
+      "learning_rate": 4.2417868024635504e-05,
+      "loss": 0.4,
       "step": 38445
     },
     {
-      "epoch": 1.35,
-      "learning_rate": 4.279321444498869e-05,
-      "loss": 0.2776,
+      "epoch": 1.3857353948174578,
+      "grad_norm": 0.18606650829315186,
+      "learning_rate": 4.241577458774034e-05,
+      "loss": 0.4044,
       "step": 38450
     },
     {
-      "epoch": 1.35,
-      "learning_rate": 4.279121323215407e-05,
-      "loss": 0.2933,
+      "epoch": 1.3859155944786825,
+      "grad_norm": 0.1808106005191803,
+      "learning_rate": 4.241368091355862e-05,
+      "loss": 0.4383,
       "step": 38455
     },
     {
-      "epoch": 1.35,
-      "learning_rate": 4.278921178831227e-05,
-      "loss": 0.2893,
+      "epoch": 1.386095794139907,
+      "grad_norm": 0.15027718245983124,
+      "learning_rate": 4.241158700211886e-05,
+      "loss": 0.4218,
       "step": 38460
     },
     {
-      "epoch": 1.35,
-      "learning_rate": 4.278721011348927e-05,
-      "loss": 0.2816,
+      "epoch": 1.3862759938011315,
+      "grad_norm": 0.15103080868721008,
+      "learning_rate": 4.24094928534496e-05,
+      "loss": 0.416,
       "step": 38465
     },
     {
-      "epoch": 1.35,
-      "learning_rate": 4.278520820771108e-05,
-      "loss": 0.2731,
+      "epoch": 1.3864561934623563,
+      "grad_norm": 0.1560763716697693,
+      "learning_rate": 4.2407398467579376e-05,
+      "loss": 0.4568,
       "step": 38470
     },
     {
-      "epoch": 1.35,
-      "learning_rate": 4.278320607100368e-05,
-      "loss": 0.2938,
+      "epoch": 1.386636393123581,
+      "grad_norm": 0.17564137279987335,
+      "learning_rate": 4.2405303844536714e-05,
+      "loss": 0.3982,
       "step": 38475
     },
     {
-      "epoch": 1.35,
-      "learning_rate": 4.278120370339307e-05,
-      "loss": 0.2871,
+      "epoch": 1.3868165927848055,
+      "grad_norm": 0.1863342970609665,
+      "learning_rate": 4.2403208984350164e-05,
+      "loss": 0.4215,
       "step": 38480
     },
     {
-      "epoch": 1.35,
-      "learning_rate": 4.277920110490524e-05,
-      "loss": 0.2997,
+      "epoch": 1.3869967924460302,
+      "grad_norm": 0.2165493667125702,
+      "learning_rate": 4.240111388704825e-05,
+      "loss": 0.4043,
       "step": 38485
     },
     {
-      "epoch": 1.35,
-      "learning_rate": 4.2777198275566214e-05,
-      "loss": 0.2904,
+      "epoch": 1.3871769921072548,
+      "grad_norm": 0.17703358829021454,
+      "learning_rate": 4.2399018552659536e-05,
+      "loss": 0.4201,
       "step": 38490
     },
     {
-      "epoch": 1.35,
-      "learning_rate": 4.277519521540197e-05,
-      "loss": 0.2833,
+      "epoch": 1.3873571917684795,
+      "grad_norm": 0.12427543848752975,
+      "learning_rate": 4.239692298121256e-05,
+      "loss": 0.4103,
       "step": 38495
     },
     {
-      "epoch": 1.35,
-      "learning_rate": 4.277319192443854e-05,
-      "loss": 0.2809,
+      "epoch": 1.3875373914297042,
+      "grad_norm": 0.161130890250206,
+      "learning_rate": 4.239482717273587e-05,
+      "loss": 0.4157,
       "step": 38500
     },
     {
-      "epoch": 1.35,
-      "eval_loss": 0.28302499651908875,
-      "eval_runtime": 10.5824,
-      "eval_samples_per_second": 9.45,
-      "eval_steps_per_second": 9.45,
+      "epoch": 1.3875373914297042,
+      "eval_loss": 0.44684794545173645,
+      "eval_runtime": 3.5231,
+      "eval_samples_per_second": 28.384,
+      "eval_steps_per_second": 7.096,
       "step": 38500
     },
     {
-      "epoch": 1.35,
-      "learning_rate": 4.277118840270192e-05,
-      "loss": 0.3101,
+      "epoch": 1.3877175910909287,
+      "grad_norm": 0.20817585289478302,
+      "learning_rate": 4.2392731127258037e-05,
+      "loss": 0.4488,
       "step": 38505
     },
     {
-      "epoch": 1.35,
-      "learning_rate": 4.2769184650218133e-05,
-      "loss": 0.2771,
+      "epoch": 1.3878977907521532,
+      "grad_norm": 0.17625221610069275,
+      "learning_rate": 4.239063484480761e-05,
+      "loss": 0.4266,
       "step": 38510
     },
     {
-      "epoch": 1.36,
-      "learning_rate": 4.276718066701319e-05,
-      "loss": 0.278,
+      "epoch": 1.388077990413378,
+      "grad_norm": 0.2073509842157364,
+      "learning_rate": 4.238853832541315e-05,
+      "loss": 0.4415,
       "step": 38515
     },
     {
-      "epoch": 1.36,
-      "learning_rate": 4.276517645311312e-05,
-      "loss": 0.2896,
+      "epoch": 1.3882581900746027,
+      "grad_norm": 0.2082330286502838,
+      "learning_rate": 4.238644156910322e-05,
+      "loss": 0.388,
       "step": 38520
     },
     {
-      "epoch": 1.36,
-      "learning_rate": 4.276317200854394e-05,
-      "loss": 0.288,
+      "epoch": 1.3884383897358272,
+      "grad_norm": 0.19567419588565826,
+      "learning_rate": 4.238434457590639e-05,
+      "loss": 0.4268,
       "step": 38525
     },
     {
-      "epoch": 1.36,
-      "learning_rate": 4.276116733333167e-05,
-      "loss": 0.2893,
+      "epoch": 1.388618589397052,
+      "grad_norm": 0.19679030776023865,
+      "learning_rate": 4.238224734585123e-05,
+      "loss": 0.4052,
       "step": 38530
     },
     {
-      "epoch": 1.36,
-      "learning_rate": 4.275916242750236e-05,
-      "loss": 0.294,
+      "epoch": 1.3887987890582765,
+      "grad_norm": 0.219960555434227,
+      "learning_rate": 4.238014987896631e-05,
+      "loss": 0.4552,
       "step": 38535
     },
     {
-      "epoch": 1.36,
-      "learning_rate": 4.2757157291082026e-05,
-      "loss": 0.3026,
+      "epoch": 1.3889789887195012,
+      "grad_norm": 0.16919073462486267,
+      "learning_rate": 4.2378052175280216e-05,
+      "loss": 0.3934,
       "step": 38540
     },
     {
-      "epoch": 1.36,
-      "learning_rate": 4.27551519240967e-05,
-      "loss": 0.2939,
+      "epoch": 1.389159188380726,
+      "grad_norm": 0.15269456803798676,
+      "learning_rate": 4.237595423482153e-05,
+      "loss": 0.3468,
       "step": 38545
     },
     {
-      "epoch": 1.36,
-      "learning_rate": 4.2753146326572436e-05,
-      "loss": 0.2771,
+      "epoch": 1.3893393880419505,
+      "grad_norm": 0.18729466199874878,
+      "learning_rate": 4.237385605761883e-05,
+      "loss": 0.4197,
       "step": 38550
     },
     {
-      "epoch": 1.36,
-      "learning_rate": 4.275114049853526e-05,
-      "loss": 0.3041,
+      "epoch": 1.3895195877031752,
+      "grad_norm": 0.19090360403060913,
+      "learning_rate": 4.2371757643700705e-05,
+      "loss": 0.4401,
       "step": 38555
     },
     {
-      "epoch": 1.36,
-      "learning_rate": 4.274913444001123e-05,
-      "loss": 0.2886,
+      "epoch": 1.3896997873643997,
+      "grad_norm": 0.17647488415241241,
+      "learning_rate": 4.236965899309574e-05,
+      "loss": 0.4282,
       "step": 38560
     },
     {
-      "epoch": 1.36,
-      "learning_rate": 4.274712815102638e-05,
-      "loss": 0.2751,
+      "epoch": 1.3898799870256244,
+      "grad_norm": 0.2209496647119522,
+      "learning_rate": 4.236756010583254e-05,
+      "loss": 0.4473,
       "step": 38565
     },
     {
-      "epoch": 1.36,
-      "learning_rate": 4.274512163160677e-05,
-      "loss": 0.2692,
+      "epoch": 1.3900601866868492,
+      "grad_norm": 0.16806313395500183,
+      "learning_rate": 4.23654609819397e-05,
+      "loss": 0.4122,
       "step": 38570
     },
     {
-      "epoch": 1.36,
-      "learning_rate": 4.274311488177845e-05,
-      "loss": 0.3038,
+      "epoch": 1.3902403863480737,
+      "grad_norm": 0.20954453945159912,
+      "learning_rate": 4.236336162144581e-05,
+      "loss": 0.4201,
       "step": 38575
     },
     {
-      "epoch": 1.36,
-      "learning_rate": 4.274110790156747e-05,
-      "loss": 0.281,
+      "epoch": 1.3904205860092982,
+      "grad_norm": 0.14711293578147888,
+      "learning_rate": 4.236126202437948e-05,
+      "loss": 0.3762,
       "step": 38580
     },
     {
-      "epoch": 1.36,
-      "learning_rate": 4.273910069099991e-05,
-      "loss": 0.2563,
+      "epoch": 1.390600785670523,
+      "grad_norm": 0.23657748103141785,
+      "learning_rate": 4.235916219076931e-05,
+      "loss": 0.4146,
       "step": 38585
     },
     {
-      "epoch": 1.36,
-      "learning_rate": 4.273709325010181e-05,
-      "loss": 0.286,
+      "epoch": 1.3907809853317477,
+      "grad_norm": 0.16697244346141815,
+      "learning_rate": 4.235706212064392e-05,
+      "loss": 0.4094,
       "step": 38590
     },
     {
-      "epoch": 1.36,
-      "learning_rate": 4.2735085578899236e-05,
-      "loss": 0.2666,
+      "epoch": 1.3909611849929722,
+      "grad_norm": 0.18580500781536102,
+      "learning_rate": 4.2354961814031924e-05,
+      "loss": 0.402,
       "step": 38595
     },
     {
-      "epoch": 1.36,
-      "learning_rate": 4.2733077677418275e-05,
-      "loss": 0.2747,
+      "epoch": 1.391141384654197,
+      "grad_norm": 0.1857820749282837,
+      "learning_rate": 4.235286127096193e-05,
+      "loss": 0.3989,
       "step": 38600
     },
     {
-      "epoch": 1.36,
-      "learning_rate": 4.2731069545684974e-05,
-      "loss": 0.2701,
+      "epoch": 1.3913215843154214,
+      "grad_norm": 0.1920892894268036,
+      "learning_rate": 4.235076049146257e-05,
+      "loss": 0.4196,
       "step": 38605
     },
     {
-      "epoch": 1.36,
-      "learning_rate": 4.272906118372544e-05,
-      "loss": 0.3037,
+      "epoch": 1.3915017839766461,
+      "grad_norm": 0.18875955045223236,
+      "learning_rate": 4.2348659475562445e-05,
+      "loss": 0.4203,
       "step": 38610
     },
     {
-      "epoch": 1.36,
-      "learning_rate": 4.272705259156572e-05,
-      "loss": 0.2786,
+      "epoch": 1.3916819836378709,
+      "grad_norm": 0.15751686692237854,
+      "learning_rate": 4.23465582232902e-05,
+      "loss": 0.4026,
       "step": 38615
     },
     {
-      "epoch": 1.36,
-      "learning_rate": 4.27250437692319e-05,
-      "loss": 0.2938,
+      "epoch": 1.3918621832990954,
+      "grad_norm": 0.1680629700422287,
+      "learning_rate": 4.2344456734674454e-05,
+      "loss": 0.3999,
       "step": 38620
     },
     {
-      "epoch": 1.36,
-      "learning_rate": 4.272303471675007e-05,
-      "loss": 0.2668,
+      "epoch": 1.39204238296032,
+      "grad_norm": 0.14273864030838013,
+      "learning_rate": 4.234235500974384e-05,
+      "loss": 0.427,
       "step": 38625
     },
     {
-      "epoch": 1.36,
-      "learning_rate": 4.272102543414631e-05,
-      "loss": 0.2861,
+      "epoch": 1.3922225826215446,
+      "grad_norm": 0.17425033450126648,
+      "learning_rate": 4.2340253048527e-05,
+      "loss": 0.4206,
       "step": 38630
     },
     {
-      "epoch": 1.36,
-      "learning_rate": 4.2719015921446726e-05,
-      "loss": 0.2727,
+      "epoch": 1.3924027822827694,
+      "grad_norm": 0.14705787599086761,
+      "learning_rate": 4.233815085105257e-05,
+      "loss": 0.4026,
       "step": 38635
     },
     {
-      "epoch": 1.36,
-      "learning_rate": 4.2717006178677386e-05,
-      "loss": 0.2813,
+      "epoch": 1.3925829819439939,
+      "grad_norm": 0.19145555794239044,
+      "learning_rate": 4.233604841734919e-05,
+      "loss": 0.3902,
       "step": 38640
     },
     {
-      "epoch": 1.36,
-      "learning_rate": 4.271499620586441e-05,
-      "loss": 0.2772,
+      "epoch": 1.3927631816052186,
+      "grad_norm": 0.20708268880844116,
+      "learning_rate": 4.2333945747445516e-05,
+      "loss": 0.4036,
       "step": 38645
     },
     {
-      "epoch": 1.36,
-      "learning_rate": 4.271298600303387e-05,
-      "loss": 0.3153,
+      "epoch": 1.3929433812664431,
+      "grad_norm": 0.17952348291873932,
+      "learning_rate": 4.2331842841370175e-05,
+      "loss": 0.4346,
       "step": 38650
     },
     {
-      "epoch": 1.36,
-      "learning_rate": 4.2710975570211884e-05,
-      "loss": 0.3017,
+      "epoch": 1.3931235809276679,
+      "grad_norm": 0.21833153069019318,
+      "learning_rate": 4.232973969915184e-05,
+      "loss": 0.4475,
       "step": 38655
     },
     {
-      "epoch": 1.36,
-      "learning_rate": 4.2708964907424555e-05,
-      "loss": 0.3075,
+      "epoch": 1.3933037805888926,
+      "grad_norm": 0.18720868229866028,
+      "learning_rate": 4.232763632081915e-05,
+      "loss": 0.3696,
       "step": 38660
     },
     {
-      "epoch": 1.36,
-      "learning_rate": 4.2706954014697984e-05,
-      "loss": 0.2937,
+      "epoch": 1.393483980250117,
+      "grad_norm": 0.1900511085987091,
+      "learning_rate": 4.232553270640077e-05,
+      "loss": 0.3998,
       "step": 38665
     },
     {
-      "epoch": 1.36,
-      "learning_rate": 4.2704942892058285e-05,
-      "loss": 0.2905,
+      "epoch": 1.3936641799113418,
+      "grad_norm": 0.19259464740753174,
+      "learning_rate": 4.232342885592536e-05,
+      "loss": 0.3862,
       "step": 38670
     },
     {
-      "epoch": 1.36,
-      "learning_rate": 4.2702931539531567e-05,
-      "loss": 0.2858,
+      "epoch": 1.3938443795725663,
+      "grad_norm": 0.19600766897201538,
+      "learning_rate": 4.232132476942159e-05,
+      "loss": 0.3997,
       "step": 38675
     },
     {
-      "epoch": 1.36,
-      "learning_rate": 4.270091995714396e-05,
-      "loss": 0.2628,
+      "epoch": 1.394024579233791,
+      "grad_norm": 0.17231407761573792,
+      "learning_rate": 4.231922044691813e-05,
+      "loss": 0.406,
       "step": 38680
     },
     {
-      "epoch": 1.36,
-      "learning_rate": 4.269890814492156e-05,
-      "loss": 0.3051,
+      "epoch": 1.3942047788950156,
+      "grad_norm": 0.19340550899505615,
+      "learning_rate": 4.231711588844363e-05,
+      "loss": 0.4058,
       "step": 38685
     },
     {
-      "epoch": 1.36,
-      "learning_rate": 4.269689610289051e-05,
-      "loss": 0.3129,
+      "epoch": 1.3943849785562403,
+      "grad_norm": 0.19323815405368805,
+      "learning_rate": 4.2315011094026784e-05,
+      "loss": 0.4092,
       "step": 38690
     },
     {
-      "epoch": 1.36,
-      "learning_rate": 4.2694883831076924e-05,
-      "loss": 0.2792,
+      "epoch": 1.3945651782174648,
+      "grad_norm": 0.19978420436382294,
+      "learning_rate": 4.231290606369627e-05,
+      "loss": 0.442,
       "step": 38695
     },
     {
-      "epoch": 1.36,
-      "learning_rate": 4.2692871329506936e-05,
-      "loss": 0.3053,
+      "epoch": 1.3947453778786896,
+      "grad_norm": 0.1691822111606598,
+      "learning_rate": 4.2310800797480756e-05,
+      "loss": 0.4045,
       "step": 38700
     },
     {
-      "epoch": 1.36,
-      "learning_rate": 4.2690858598206665e-05,
-      "loss": 0.2971,
+      "epoch": 1.3949255775399143,
+      "grad_norm": 0.2136184573173523,
+      "learning_rate": 4.230869529540894e-05,
+      "loss": 0.3986,
       "step": 38705
     },
     {
-      "epoch": 1.36,
-      "learning_rate": 4.268884563720226e-05,
-      "loss": 0.2868,
+      "epoch": 1.3951057772011388,
+      "grad_norm": 0.1695156693458557,
+      "learning_rate": 4.230658955750949e-05,
+      "loss": 0.4319,
       "step": 38710
     },
     {
-      "epoch": 1.36,
-      "learning_rate": 4.2686832446519856e-05,
-      "loss": 0.2846,
+      "epoch": 1.3952859768623636,
+      "grad_norm": 0.17145298421382904,
+      "learning_rate": 4.230448358381112e-05,
+      "loss": 0.3985,
       "step": 38715
     },
     {
-      "epoch": 1.36,
-      "learning_rate": 4.2684819026185583e-05,
-      "loss": 0.2756,
+      "epoch": 1.395466176523588,
+      "grad_norm": 0.18987350165843964,
+      "learning_rate": 4.2302377374342505e-05,
+      "loss": 0.4378,
       "step": 38720
     },
     {
-      "epoch": 1.36,
-      "learning_rate": 4.268280537622559e-05,
-      "loss": 0.291,
+      "epoch": 1.3956463761848128,
+      "grad_norm": 0.17236654460430145,
+      "learning_rate": 4.2300270929132344e-05,
+      "loss": 0.4128,
       "step": 38725
     },
     {
-      "epoch": 1.36,
-      "learning_rate": 4.268079149666603e-05,
-      "loss": 0.3187,
+      "epoch": 1.3958265758460375,
+      "grad_norm": 0.18958859145641327,
+      "learning_rate": 4.229816424820935e-05,
+      "loss": 0.4303,
       "step": 38730
     },
     {
-      "epoch": 1.36,
-      "learning_rate": 4.267877738753303e-05,
-      "loss": 0.2888,
+      "epoch": 1.396006775507262,
+      "grad_norm": 0.19442634284496307,
+      "learning_rate": 4.229605733160221e-05,
+      "loss": 0.4405,
       "step": 38735
     },
     {
-      "epoch": 1.36,
-      "learning_rate": 4.2676763048852764e-05,
-      "loss": 0.2993,
+      "epoch": 1.3961869751684866,
+      "grad_norm": 0.16832728683948517,
+      "learning_rate": 4.2293950179339645e-05,
+      "loss": 0.4066,
       "step": 38740
     },
     {
-      "epoch": 1.36,
-      "learning_rate": 4.2674748480651385e-05,
-      "loss": 0.2921,
+      "epoch": 1.3963671748297113,
+      "grad_norm": 0.20120170712471008,
+      "learning_rate": 4.2291842791450356e-05,
+      "loss": 0.4036,
       "step": 38745
     },
     {
-      "epoch": 1.36,
-      "learning_rate": 4.267273368295504e-05,
-      "loss": 0.2735,
+      "epoch": 1.396547374490936,
+      "grad_norm": 0.17381571233272552,
+      "learning_rate": 4.2289735167963054e-05,
+      "loss": 0.4033,
       "step": 38750
     },
     {
-      "epoch": 1.36,
-      "learning_rate": 4.2670718655789896e-05,
-      "loss": 0.296,
+      "epoch": 1.3967275741521605,
+      "grad_norm": 0.17229481041431427,
+      "learning_rate": 4.228762730890645e-05,
+      "loss": 0.4159,
       "step": 38755
     },
     {
-      "epoch": 1.36,
-      "learning_rate": 4.266870339918211e-05,
-      "loss": 0.3009,
+      "epoch": 1.3969077738133853,
+      "grad_norm": 0.1683155596256256,
+      "learning_rate": 4.228551921430928e-05,
+      "loss": 0.3986,
       "step": 38760
     },
     {
-      "epoch": 1.36,
-      "learning_rate": 4.266668791315786e-05,
-      "loss": 0.2583,
+      "epoch": 1.3970879734746098,
+      "grad_norm": 0.1502419412136078,
+      "learning_rate": 4.228341088420026e-05,
+      "loss": 0.4202,
       "step": 38765
     },
     {
-      "epoch": 1.36,
-      "learning_rate": 4.266467219774331e-05,
-      "loss": 0.28,
+      "epoch": 1.3972681731358345,
+      "grad_norm": 0.16682939231395721,
+      "learning_rate": 4.2281302318608106e-05,
+      "loss": 0.4485,
       "step": 38770
     },
     {
-      "epoch": 1.36,
-      "learning_rate": 4.2662656252964635e-05,
-      "loss": 0.2812,
+      "epoch": 1.3974483727970592,
+      "grad_norm": 0.20583415031433105,
+      "learning_rate": 4.227919351756155e-05,
+      "loss": 0.4224,
       "step": 38775
     },
     {
-      "epoch": 1.36,
-      "learning_rate": 4.2660640078848e-05,
-      "loss": 0.3001,
+      "epoch": 1.3976285724582838,
+      "grad_norm": 0.1680290699005127,
+      "learning_rate": 4.227708448108934e-05,
+      "loss": 0.4034,
       "step": 38780
     },
     {
-      "epoch": 1.36,
-      "learning_rate": 4.26586236754196e-05,
-      "loss": 0.2902,
+      "epoch": 1.3978087721195085,
+      "grad_norm": 0.1613532453775406,
+      "learning_rate": 4.227497520922019e-05,
+      "loss": 0.4187,
       "step": 38785
     },
     {
-      "epoch": 1.36,
-      "learning_rate": 4.265660704270561e-05,
-      "loss": 0.2587,
+      "epoch": 1.397988971780733,
+      "grad_norm": 0.19061127305030823,
+      "learning_rate": 4.2272865701982855e-05,
+      "loss": 0.4305,
       "step": 38790
     },
     {
-      "epoch": 1.36,
-      "learning_rate": 4.2654590180732215e-05,
-      "loss": 0.3016,
+      "epoch": 1.3981691714419577,
+      "grad_norm": 0.1994105577468872,
+      "learning_rate": 4.227075595940606e-05,
+      "loss": 0.4196,
       "step": 38795
     },
     {
-      "epoch": 1.37,
-      "learning_rate": 4.2652573089525594e-05,
-      "loss": 0.2793,
+      "epoch": 1.3983493711031822,
+      "grad_norm": 0.16450746357440948,
+      "learning_rate": 4.226864598151857e-05,
+      "loss": 0.4488,
       "step": 38800
     },
     {
-      "epoch": 1.37,
-      "learning_rate": 4.2650555769111945e-05,
-      "loss": 0.2888,
+      "epoch": 1.398529570764407,
+      "grad_norm": 0.15390661358833313,
+      "learning_rate": 4.226653576834911e-05,
+      "loss": 0.4042,
       "step": 38805
     },
     {
-      "epoch": 1.37,
-      "learning_rate": 4.2648538219517466e-05,
-      "loss": 0.3073,
+      "epoch": 1.3987097704256315,
+      "grad_norm": 0.22952650487422943,
+      "learning_rate": 4.226442531992645e-05,
+      "loss": 0.4069,
       "step": 38810
     },
     {
-      "epoch": 1.37,
-      "learning_rate": 4.2646520440768346e-05,
-      "loss": 0.2823,
+      "epoch": 1.3988899700868562,
+      "grad_norm": 0.18804705142974854,
+      "learning_rate": 4.2262314636279334e-05,
+      "loss": 0.3962,
       "step": 38815
     },
     {
-      "epoch": 1.37,
-      "learning_rate": 4.264450243289079e-05,
-      "loss": 0.3026,
+      "epoch": 1.399070169748081,
+      "grad_norm": 0.15264515578746796,
+      "learning_rate": 4.226020371743653e-05,
+      "loss": 0.4237,
       "step": 38820
     },
     {
-      "epoch": 1.37,
-      "learning_rate": 4.2642484195911e-05,
-      "loss": 0.2741,
+      "epoch": 1.3992503694093055,
+      "grad_norm": 0.18798717856407166,
+      "learning_rate": 4.225809256342678e-05,
+      "loss": 0.4252,
       "step": 38825
     },
     {
-      "epoch": 1.37,
-      "learning_rate": 4.2640465729855186e-05,
-      "loss": 0.2949,
+      "epoch": 1.3994305690705302,
+      "grad_norm": 0.21027231216430664,
+      "learning_rate": 4.225598117427887e-05,
+      "loss": 0.4366,
       "step": 38830
     },
     {
-      "epoch": 1.37,
-      "learning_rate": 4.263844703474954e-05,
-      "loss": 0.2495,
+      "epoch": 1.3996107687317547,
+      "grad_norm": 0.17129452526569366,
+      "learning_rate": 4.225386955002155e-05,
+      "loss": 0.4157,
       "step": 38835
     },
     {
-      "epoch": 1.37,
-      "learning_rate": 4.2636428110620274e-05,
-      "loss": 0.2771,
+      "epoch": 1.3997909683929795,
+      "grad_norm": 0.19141530990600586,
+      "learning_rate": 4.2251757690683604e-05,
+      "loss": 0.3856,
       "step": 38840
     },
     {
-      "epoch": 1.37,
-      "learning_rate": 4.263440895749363e-05,
-      "loss": 0.2698,
+      "epoch": 1.3999711680542042,
+      "grad_norm": 0.1746709644794464,
+      "learning_rate": 4.22496455962938e-05,
+      "loss": 0.395,
       "step": 38845
     },
     {
-      "epoch": 1.37,
-      "learning_rate": 4.2632389575395795e-05,
-      "loss": 0.3005,
+      "epoch": 1.4001513677154287,
+      "grad_norm": 0.20320837199687958,
+      "learning_rate": 4.22475332668809e-05,
+      "loss": 0.4106,
       "step": 38850
     },
     {
-      "epoch": 1.37,
-      "learning_rate": 4.263036996435301e-05,
-      "loss": 0.2822,
+      "epoch": 1.4003315673766532,
+      "grad_norm": 0.17430174350738525,
+      "learning_rate": 4.224542070247371e-05,
+      "loss": 0.4004,
       "step": 38855
     },
     {
-      "epoch": 1.37,
-      "learning_rate": 4.262835012439148e-05,
-      "loss": 0.2735,
+      "epoch": 1.400511767037878,
+      "grad_norm": 0.1601121872663498,
+      "learning_rate": 4.224330790310101e-05,
+      "loss": 0.4212,
       "step": 38860
     },
     {
-      "epoch": 1.37,
-      "learning_rate": 4.262633005553745e-05,
-      "loss": 0.2999,
+      "epoch": 1.4006919666991027,
+      "grad_norm": 0.16500531136989594,
+      "learning_rate": 4.2241194868791565e-05,
+      "loss": 0.3749,
       "step": 38865
     },
     {
-      "epoch": 1.37,
-      "learning_rate": 4.262430975781714e-05,
-      "loss": 0.2825,
+      "epoch": 1.4008721663603272,
+      "grad_norm": 0.18439029157161713,
+      "learning_rate": 4.2239081599574184e-05,
+      "loss": 0.4193,
       "step": 38870
     },
     {
-      "epoch": 1.37,
-      "learning_rate": 4.2622289231256776e-05,
-      "loss": 0.2627,
+      "epoch": 1.401052366021552,
+      "grad_norm": 0.20299525558948517,
+      "learning_rate": 4.223696809547766e-05,
+      "loss": 0.4482,
       "step": 38875
     },
     {
-      "epoch": 1.37,
-      "learning_rate": 4.26202684758826e-05,
-      "loss": 0.2937,
+      "epoch": 1.4012325656827764,
+      "grad_norm": 0.17272146046161652,
+      "learning_rate": 4.2234854356530776e-05,
+      "loss": 0.4225,
       "step": 38880
     },
     {
-      "epoch": 1.37,
-      "learning_rate": 4.261824749172086e-05,
-      "loss": 0.283,
+      "epoch": 1.4014127653440012,
+      "grad_norm": 0.1982017606496811,
+      "learning_rate": 4.223274038276233e-05,
+      "loss": 0.4437,
       "step": 38885
     },
     {
-      "epoch": 1.37,
-      "learning_rate": 4.2616226278797776e-05,
-      "loss": 0.2989,
+      "epoch": 1.401592965005226,
+      "grad_norm": 0.16172081232070923,
+      "learning_rate": 4.223062617420114e-05,
+      "loss": 0.4323,
       "step": 38890
     },
     {
-      "epoch": 1.37,
-      "learning_rate": 4.2614204837139616e-05,
-      "loss": 0.2897,
+      "epoch": 1.4017731646664504,
+      "grad_norm": 0.1717664897441864,
+      "learning_rate": 4.2228511730876006e-05,
+      "loss": 0.3999,
       "step": 38895
     },
     {
-      "epoch": 1.37,
-      "learning_rate": 4.26121831667726e-05,
-      "loss": 0.3178,
+      "epoch": 1.4019533643276751,
+      "grad_norm": 0.18911094963550568,
+      "learning_rate": 4.2226397052815734e-05,
+      "loss": 0.409,
       "step": 38900
     },
     {
-      "epoch": 1.37,
-      "learning_rate": 4.2610161267723005e-05,
-      "loss": 0.281,
+      "epoch": 1.4021335639888997,
+      "grad_norm": 0.19404666125774384,
+      "learning_rate": 4.2224282140049145e-05,
+      "loss": 0.411,
       "step": 38905
     },
     {
-      "epoch": 1.37,
-      "learning_rate": 4.260813914001706e-05,
-      "loss": 0.2685,
+      "epoch": 1.4023137636501244,
+      "grad_norm": 0.19572977721691132,
+      "learning_rate": 4.2222166992605037e-05,
+      "loss": 0.4489,
       "step": 38910
     },
     {
-      "epoch": 1.37,
-      "learning_rate": 4.2606116783681045e-05,
-      "loss": 0.291,
+      "epoch": 1.402493963311349,
+      "grad_norm": 0.20973151922225952,
+      "learning_rate": 4.2220051610512236e-05,
+      "loss": 0.4269,
       "step": 38915
     },
     {
-      "epoch": 1.37,
-      "learning_rate": 4.2604094198741195e-05,
-      "loss": 0.2739,
+      "epoch": 1.4026741629725736,
+      "grad_norm": 0.17148537933826447,
+      "learning_rate": 4.2217935993799576e-05,
+      "loss": 0.3622,
       "step": 38920
     },
     {
-      "epoch": 1.37,
-      "learning_rate": 4.260207138522379e-05,
-      "loss": 0.3218,
+      "epoch": 1.4028543626337981,
+      "grad_norm": 0.1789887696504593,
+      "learning_rate": 4.221582014249586e-05,
+      "loss": 0.436,
       "step": 38925
     },
     {
-      "epoch": 1.37,
-      "learning_rate": 4.260004834315509e-05,
-      "loss": 0.2865,
+      "epoch": 1.4030345622950229,
+      "grad_norm": 0.17625224590301514,
+      "learning_rate": 4.2213704056629936e-05,
+      "loss": 0.4053,
       "step": 38930
     },
     {
-      "epoch": 1.37,
-      "learning_rate": 4.259802507256135e-05,
-      "loss": 0.3042,
+      "epoch": 1.4032147619562476,
+      "grad_norm": 0.1674962043762207,
+      "learning_rate": 4.2211587736230614e-05,
+      "loss": 0.4552,
       "step": 38935
     },
     {
-      "epoch": 1.37,
-      "learning_rate": 4.2596001573468864e-05,
-      "loss": 0.2767,
+      "epoch": 1.4033949616174721,
+      "grad_norm": 0.20774921774864197,
+      "learning_rate": 4.220947118132676e-05,
+      "loss": 0.4201,
       "step": 38940
     },
     {
-      "epoch": 1.37,
-      "learning_rate": 4.259397784590389e-05,
-      "loss": 0.2787,
+      "epoch": 1.4035751612786969,
+      "grad_norm": 0.15896476805210114,
+      "learning_rate": 4.220735439194718e-05,
+      "loss": 0.4324,
       "step": 38945
     },
     {
-      "epoch": 1.37,
-      "learning_rate": 4.2591953889892706e-05,
-      "loss": 0.2751,
+      "epoch": 1.4037553609399214,
+      "grad_norm": 0.21336419880390167,
+      "learning_rate": 4.220523736812073e-05,
+      "loss": 0.4731,
       "step": 38950
     },
     {
-      "epoch": 1.37,
-      "learning_rate": 4.25899297054616e-05,
-      "loss": 0.2917,
+      "epoch": 1.403935560601146,
+      "grad_norm": 0.23092104494571686,
+      "learning_rate": 4.220312010987626e-05,
+      "loss": 0.422,
       "step": 38955
     },
     {
-      "epoch": 1.37,
-      "learning_rate": 4.2587905292636845e-05,
-      "loss": 0.279,
+      "epoch": 1.4041157602623708,
+      "grad_norm": 0.20718926191329956,
+      "learning_rate": 4.22010026172426e-05,
+      "loss": 0.394,
       "step": 38960
     },
     {
-      "epoch": 1.37,
-      "learning_rate": 4.258588065144473e-05,
-      "loss": 0.2934,
+      "epoch": 1.4042959599235953,
+      "grad_norm": 0.18056684732437134,
+      "learning_rate": 4.219888489024861e-05,
+      "loss": 0.421,
       "step": 38965
     },
     {
-      "epoch": 1.37,
-      "learning_rate": 4.2583855781911544e-05,
-      "loss": 0.2789,
+      "epoch": 1.4044761595848199,
+      "grad_norm": 0.19858917593955994,
+      "learning_rate": 4.2196766928923147e-05,
+      "loss": 0.4036,
       "step": 38970
     },
     {
-      "epoch": 1.37,
-      "learning_rate": 4.258183068406358e-05,
-      "loss": 0.2685,
+      "epoch": 1.4046563592460446,
+      "grad_norm": 0.23689039051532745,
+      "learning_rate": 4.219464873329506e-05,
+      "loss": 0.4278,
       "step": 38975
     },
     {
-      "epoch": 1.37,
-      "learning_rate": 4.257980535792713e-05,
-      "loss": 0.3244,
+      "epoch": 1.4048365589072693,
+      "grad_norm": 0.1857483685016632,
+      "learning_rate": 4.219253030339322e-05,
+      "loss": 0.4156,
       "step": 38980
     },
     {
-      "epoch": 1.37,
-      "learning_rate": 4.2577779803528495e-05,
-      "loss": 0.3174,
+      "epoch": 1.4050167585684938,
+      "grad_norm": 0.1789952665567398,
+      "learning_rate": 4.2190411639246474e-05,
+      "loss": 0.4147,
       "step": 38985
     },
     {
-      "epoch": 1.37,
-      "learning_rate": 4.2575754020893966e-05,
-      "loss": 0.2866,
+      "epoch": 1.4051969582297186,
+      "grad_norm": 0.21986453235149384,
+      "learning_rate": 4.218829274088371e-05,
+      "loss": 0.454,
       "step": 38990
     },
     {
-      "epoch": 1.37,
-      "learning_rate": 4.257372801004986e-05,
-      "loss": 0.2718,
+      "epoch": 1.405377157890943,
+      "grad_norm": 0.16250832378864288,
+      "learning_rate": 4.2186173608333776e-05,
+      "loss": 0.3945,
       "step": 38995
     },
     {
-      "epoch": 1.37,
-      "learning_rate": 4.257170177102248e-05,
-      "loss": 0.2912,
+      "epoch": 1.4055573575521678,
+      "grad_norm": 0.16607147455215454,
+      "learning_rate": 4.2184054241625556e-05,
+      "loss": 0.429,
       "step": 39000
     },
     {
-      "epoch": 1.37,
-      "eval_loss": 0.28303414583206177,
-      "eval_runtime": 10.5465,
-      "eval_samples_per_second": 9.482,
-      "eval_steps_per_second": 9.482,
+      "epoch": 1.4055573575521678,
+      "eval_loss": 0.4471230208873749,
+      "eval_runtime": 3.5385,
+      "eval_samples_per_second": 28.261,
+      "eval_steps_per_second": 7.065,
       "step": 39000
     },
     {
-      "epoch": 1.37,
-      "learning_rate": 4.256967530383813e-05,
-      "loss": 0.3285,
+      "epoch": 1.4057375572133926,
+      "grad_norm": 0.18567611277103424,
+      "learning_rate": 4.218193464078792e-05,
+      "loss": 0.4589,
       "step": 39005
     },
     {
-      "epoch": 1.37,
-      "learning_rate": 4.2567648608523127e-05,
-      "loss": 0.2956,
+      "epoch": 1.405917756874617,
+      "grad_norm": 0.14248013496398926,
+      "learning_rate": 4.217981480584976e-05,
+      "loss": 0.4067,
       "step": 39010
     },
     {
-      "epoch": 1.37,
-      "learning_rate": 4.2565621685103784e-05,
-      "loss": 0.3043,
+      "epoch": 1.4060979565358416,
+      "grad_norm": 0.18082767724990845,
+      "learning_rate": 4.217769473683994e-05,
+      "loss": 0.4069,
       "step": 39015
     },
     {
-      "epoch": 1.37,
-      "learning_rate": 4.256359453360642e-05,
-      "loss": 0.2776,
+      "epoch": 1.4062781561970663,
+      "grad_norm": 0.2055099606513977,
+      "learning_rate": 4.217557443378736e-05,
+      "loss": 0.4208,
       "step": 39020
     },
     {
-      "epoch": 1.37,
-      "learning_rate": 4.2561567154057355e-05,
-      "loss": 0.2852,
+      "epoch": 1.406458355858291,
+      "grad_norm": 0.20879048109054565,
+      "learning_rate": 4.2173453896720906e-05,
+      "loss": 0.3962,
       "step": 39025
     },
     {
-      "epoch": 1.37,
-      "learning_rate": 4.255953954648291e-05,
-      "loss": 0.2754,
+      "epoch": 1.4066385555195156,
+      "grad_norm": 0.18732579052448273,
+      "learning_rate": 4.217133312566946e-05,
+      "loss": 0.4423,
       "step": 39030
     },
     {
-      "epoch": 1.37,
-      "learning_rate": 4.255751171090942e-05,
-      "loss": 0.3048,
+      "epoch": 1.4068187551807403,
+      "grad_norm": 0.199151873588562,
+      "learning_rate": 4.2169212120661926e-05,
+      "loss": 0.4394,
       "step": 39035
     },
     {
-      "epoch": 1.37,
-      "learning_rate": 4.255548364736321e-05,
-      "loss": 0.2933,
+      "epoch": 1.4069989548419648,
+      "grad_norm": 0.15201795101165771,
+      "learning_rate": 4.21670908817272e-05,
+      "loss": 0.4055,
       "step": 39040
     },
     {
-      "epoch": 1.37,
-      "learning_rate": 4.2553455355870615e-05,
-      "loss": 0.288,
+      "epoch": 1.4071791545031895,
+      "grad_norm": 0.15921539068222046,
+      "learning_rate": 4.2164969408894194e-05,
+      "loss": 0.4533,
       "step": 39045
     },
     {
-      "epoch": 1.37,
-      "learning_rate": 4.2551426836457976e-05,
-      "loss": 0.2834,
+      "epoch": 1.4073593541644143,
+      "grad_norm": 0.16528940200805664,
+      "learning_rate": 4.21628477021918e-05,
+      "loss": 0.4136,
       "step": 39050
     },
     {
-      "epoch": 1.37,
-      "learning_rate": 4.254939808915162e-05,
-      "loss": 0.3058,
+      "epoch": 1.4075395538256388,
+      "grad_norm": 0.20184743404388428,
+      "learning_rate": 4.216072576164892e-05,
+      "loss": 0.453,
       "step": 39055
     },
     {
-      "epoch": 1.37,
-      "learning_rate": 4.254736911397789e-05,
-      "loss": 0.275,
+      "epoch": 1.4077197534868635,
+      "grad_norm": 0.2002110779285431,
+      "learning_rate": 4.215860358729448e-05,
+      "loss": 0.4098,
       "step": 39060
     },
     {
-      "epoch": 1.37,
-      "learning_rate": 4.2545339910963146e-05,
-      "loss": 0.2764,
+      "epoch": 1.407899953148088,
+      "grad_norm": 0.16042189300060272,
+      "learning_rate": 4.215648117915739e-05,
+      "loss": 0.3906,
       "step": 39065
     },
     {
-      "epoch": 1.37,
-      "learning_rate": 4.2543310480133725e-05,
-      "loss": 0.2942,
+      "epoch": 1.4080801528093128,
+      "grad_norm": 0.21840621531009674,
+      "learning_rate": 4.2154358537266564e-05,
+      "loss": 0.4053,
       "step": 39070
     },
     {
-      "epoch": 1.37,
-      "learning_rate": 4.254128082151597e-05,
-      "loss": 0.2854,
+      "epoch": 1.4082603524705375,
+      "grad_norm": 0.15878549218177795,
+      "learning_rate": 4.2152235661650925e-05,
+      "loss": 0.4222,
       "step": 39075
     },
     {
-      "epoch": 1.37,
-      "learning_rate": 4.253925093513626e-05,
-      "loss": 0.2886,
+      "epoch": 1.408440552131762,
+      "grad_norm": 0.17753714323043823,
+      "learning_rate": 4.215011255233939e-05,
+      "loss": 0.4452,
       "step": 39080
     },
     {
-      "epoch": 1.38,
-      "learning_rate": 4.2537220821020916e-05,
-      "loss": 0.2735,
+      "epoch": 1.4086207517929865,
+      "grad_norm": 0.1962030678987503,
+      "learning_rate": 4.2147989209360903e-05,
+      "loss": 0.4588,
       "step": 39085
     },
     {
-      "epoch": 1.38,
-      "learning_rate": 4.253519047919633e-05,
-      "loss": 0.2643,
+      "epoch": 1.4088009514542112,
+      "grad_norm": 0.17643392086029053,
+      "learning_rate": 4.2145865632744376e-05,
+      "loss": 0.4102,
       "step": 39090
     },
     {
-      "epoch": 1.38,
-      "learning_rate": 4.2533159909688844e-05,
-      "loss": 0.3042,
+      "epoch": 1.408981151115436,
+      "grad_norm": 0.2033703774213791,
+      "learning_rate": 4.214374182251874e-05,
+      "loss": 0.3898,
       "step": 39095
     },
     {
-      "epoch": 1.38,
-      "learning_rate": 4.253112911252484e-05,
-      "loss": 0.2939,
+      "epoch": 1.4091613507766605,
+      "grad_norm": 0.2141045182943344,
+      "learning_rate": 4.214161777871296e-05,
+      "loss": 0.4343,
       "step": 39100
     },
     {
-      "epoch": 1.38,
-      "learning_rate": 4.2529098087730665e-05,
-      "loss": 0.2816,
+      "epoch": 1.4093415504378852,
+      "grad_norm": 0.19499428570270538,
+      "learning_rate": 4.213949350135594e-05,
+      "loss": 0.4263,
       "step": 39105
     },
     {
-      "epoch": 1.38,
-      "learning_rate": 4.252706683533272e-05,
-      "loss": 0.2897,
+      "epoch": 1.4095217500991097,
+      "grad_norm": 0.1868494302034378,
+      "learning_rate": 4.2137368990476655e-05,
+      "loss": 0.4148,
       "step": 39110
     },
     {
-      "epoch": 1.38,
-      "learning_rate": 4.252503535535735e-05,
-      "loss": 0.279,
+      "epoch": 1.4097019497603345,
+      "grad_norm": 0.16455568373203278,
+      "learning_rate": 4.213524424610402e-05,
+      "loss": 0.3985,
       "step": 39115
     },
     {
-      "epoch": 1.38,
-      "learning_rate": 4.252300364783095e-05,
-      "loss": 0.2804,
+      "epoch": 1.4098821494215592,
+      "grad_norm": 0.23864524066448212,
+      "learning_rate": 4.213311926826701e-05,
+      "loss": 0.4044,
       "step": 39120
     },
     {
-      "epoch": 1.38,
-      "learning_rate": 4.25209717127799e-05,
-      "loss": 0.2836,
+      "epoch": 1.4100623490827837,
+      "grad_norm": 0.20035329461097717,
+      "learning_rate": 4.213099405699457e-05,
+      "loss": 0.3905,
       "step": 39125
     },
     {
-      "epoch": 1.38,
-      "learning_rate": 4.2518939550230574e-05,
-      "loss": 0.2867,
+      "epoch": 1.4102425487440082,
+      "grad_norm": 0.18260158598423004,
+      "learning_rate": 4.212886861231564e-05,
+      "loss": 0.4173,
       "step": 39130
     },
     {
-      "epoch": 1.38,
-      "learning_rate": 4.251690716020936e-05,
-      "loss": 0.3086,
+      "epoch": 1.410422748405233,
+      "grad_norm": 0.14662878215312958,
+      "learning_rate": 4.212674293425919e-05,
+      "loss": 0.3954,
       "step": 39135
     },
     {
-      "epoch": 1.38,
-      "learning_rate": 4.2514874542742653e-05,
-      "loss": 0.2838,
+      "epoch": 1.4106029480664577,
+      "grad_norm": 0.2082986682653427,
+      "learning_rate": 4.2124617022854195e-05,
+      "loss": 0.4102,
       "step": 39140
     },
     {
-      "epoch": 1.38,
-      "learning_rate": 4.2512841697856855e-05,
-      "loss": 0.2926,
+      "epoch": 1.4107831477276822,
+      "grad_norm": 0.19178947806358337,
+      "learning_rate": 4.212249087812961e-05,
+      "loss": 0.4261,
       "step": 39145
     },
     {
-      "epoch": 1.38,
-      "learning_rate": 4.251080862557833e-05,
-      "loss": 0.293,
+      "epoch": 1.410963347388907,
+      "grad_norm": 0.18366897106170654,
+      "learning_rate": 4.2120364500114394e-05,
+      "loss": 0.3988,
       "step": 39150
     },
     {
-      "epoch": 1.38,
-      "learning_rate": 4.25087753259335e-05,
-      "loss": 0.2791,
+      "epoch": 1.4111435470501315,
+      "grad_norm": 0.20114761590957642,
+      "learning_rate": 4.2118237888837534e-05,
+      "loss": 0.4235,
       "step": 39155
     },
     {
-      "epoch": 1.38,
-      "learning_rate": 4.250674179894877e-05,
-      "loss": 0.2782,
+      "epoch": 1.4113237467113562,
+      "grad_norm": 0.17763690650463104,
+      "learning_rate": 4.2116111044327984e-05,
+      "loss": 0.4264,
       "step": 39160
     },
     {
-      "epoch": 1.38,
-      "learning_rate": 4.250470804465053e-05,
-      "loss": 0.3029,
+      "epoch": 1.411503946372581,
+      "grad_norm": 0.2246585637331009,
+      "learning_rate": 4.2113983966614745e-05,
+      "loss": 0.4089,
       "step": 39165
     },
     {
-      "epoch": 1.38,
-      "learning_rate": 4.250267406306519e-05,
-      "loss": 0.2718,
+      "epoch": 1.4116841460338054,
+      "grad_norm": 0.19135521352291107,
+      "learning_rate": 4.211185665572678e-05,
+      "loss": 0.4371,
       "step": 39170
     },
     {
-      "epoch": 1.38,
-      "learning_rate": 4.250063985421917e-05,
-      "loss": 0.2931,
+      "epoch": 1.4118643456950302,
+      "grad_norm": 0.19484780728816986,
+      "learning_rate": 4.2109729111693085e-05,
+      "loss": 0.3799,
       "step": 39175
     },
     {
-      "epoch": 1.38,
-      "learning_rate": 4.2498605418138865e-05,
-      "loss": 0.2912,
+      "epoch": 1.4120445453562547,
+      "grad_norm": 0.16387687623500824,
+      "learning_rate": 4.210760133454265e-05,
+      "loss": 0.3739,
       "step": 39180
     },
     {
-      "epoch": 1.38,
-      "learning_rate": 4.2496570754850704e-05,
-      "loss": 0.2946,
+      "epoch": 1.4122247450174794,
+      "grad_norm": 0.169037863612175,
+      "learning_rate": 4.210547332430446e-05,
+      "loss": 0.4141,
       "step": 39185
     },
     {
-      "epoch": 1.38,
-      "learning_rate": 4.2494535864381104e-05,
-      "loss": 0.2748,
+      "epoch": 1.412404944678704,
+      "grad_norm": 0.2004702091217041,
+      "learning_rate": 4.21033450810075e-05,
+      "loss": 0.4108,
       "step": 39190
     },
     {
-      "epoch": 1.38,
-      "learning_rate": 4.2492500746756484e-05,
-      "loss": 0.2726,
+      "epoch": 1.4125851443399287,
+      "grad_norm": 0.16014528274536133,
+      "learning_rate": 4.210121660468077e-05,
+      "loss": 0.3964,
       "step": 39195
     },
     {
-      "epoch": 1.38,
-      "learning_rate": 4.2490465402003273e-05,
-      "loss": 0.2799,
+      "epoch": 1.4127653440011532,
+      "grad_norm": 0.16173921525478363,
+      "learning_rate": 4.209908789535328e-05,
+      "loss": 0.4055,
       "step": 39200
     },
     {
-      "epoch": 1.38,
-      "learning_rate": 4.248842983014789e-05,
-      "loss": 0.2898,
+      "epoch": 1.412945543662378,
+      "grad_norm": 0.1848604828119278,
+      "learning_rate": 4.209695895305403e-05,
+      "loss": 0.4002,
       "step": 39205
     },
     {
-      "epoch": 1.38,
-      "learning_rate": 4.2486394031216766e-05,
-      "loss": 0.2735,
+      "epoch": 1.4131257433236026,
+      "grad_norm": 0.14384835958480835,
+      "learning_rate": 4.209482977781202e-05,
+      "loss": 0.4028,
       "step": 39210
     },
     {
-      "epoch": 1.38,
-      "learning_rate": 4.248435800523636e-05,
-      "loss": 0.2879,
+      "epoch": 1.4133059429848271,
+      "grad_norm": 0.15797820687294006,
+      "learning_rate": 4.209270036965627e-05,
+      "loss": 0.4436,
       "step": 39215
     },
     {
-      "epoch": 1.38,
-      "learning_rate": 4.248232175223307e-05,
-      "loss": 0.2592,
+      "epoch": 1.4134861426460519,
+      "grad_norm": 0.17530421912670135,
+      "learning_rate": 4.2090570728615774e-05,
+      "loss": 0.4092,
       "step": 39220
     },
     {
-      "epoch": 1.38,
-      "learning_rate": 4.248028527223335e-05,
-      "loss": 0.2795,
+      "epoch": 1.4136663423072764,
+      "grad_norm": 0.17297030985355377,
+      "learning_rate": 4.208844085471957e-05,
+      "loss": 0.4609,
       "step": 39225
     },
     {
-      "epoch": 1.38,
-      "learning_rate": 4.247824856526366e-05,
-      "loss": 0.2916,
+      "epoch": 1.4138465419685011,
+      "grad_norm": 0.1550433486700058,
+      "learning_rate": 4.2086310747996674e-05,
+      "loss": 0.4124,
       "step": 39230
     },
     {
-      "epoch": 1.38,
-      "learning_rate": 4.247621163135043e-05,
-      "loss": 0.2857,
+      "epoch": 1.4140267416297259,
+      "grad_norm": 0.19032645225524902,
+      "learning_rate": 4.2084180408476094e-05,
+      "loss": 0.4298,
       "step": 39235
     },
     {
-      "epoch": 1.38,
-      "learning_rate": 4.2474174470520103e-05,
-      "loss": 0.2693,
+      "epoch": 1.4142069412909504,
+      "grad_norm": 0.18603280186653137,
+      "learning_rate": 4.208204983618687e-05,
+      "loss": 0.4064,
       "step": 39240
     },
     {
-      "epoch": 1.38,
-      "learning_rate": 4.247213708279913e-05,
-      "loss": 0.2577,
+      "epoch": 1.4143871409521749,
+      "grad_norm": 0.21045638620853424,
+      "learning_rate": 4.2079919031158014e-05,
+      "loss": 0.435,
       "step": 39245
     },
     {
-      "epoch": 1.38,
-      "learning_rate": 4.247009946821398e-05,
-      "loss": 0.2838,
+      "epoch": 1.4145673406133996,
+      "grad_norm": 0.22773757576942444,
+      "learning_rate": 4.2077787993418574e-05,
+      "loss": 0.4148,
       "step": 39250
     },
     {
-      "epoch": 1.38,
-      "learning_rate": 4.24680616267911e-05,
-      "loss": 0.2802,
+      "epoch": 1.4147475402746243,
+      "grad_norm": 0.22381030023097992,
+      "learning_rate": 4.2075656722997583e-05,
+      "loss": 0.3881,
       "step": 39255
     },
     {
-      "epoch": 1.38,
-      "learning_rate": 4.246602355855696e-05,
-      "loss": 0.2706,
+      "epoch": 1.4149277399358489,
+      "grad_norm": 0.2066371738910675,
+      "learning_rate": 4.207352521992407e-05,
+      "loss": 0.4107,
       "step": 39260
     },
     {
-      "epoch": 1.38,
-      "learning_rate": 4.2463985263537995e-05,
-      "loss": 0.2977,
+      "epoch": 1.4151079395970736,
+      "grad_norm": 0.1946861892938614,
+      "learning_rate": 4.207139348422708e-05,
+      "loss": 0.4056,
       "step": 39265
     },
     {
-      "epoch": 1.38,
-      "learning_rate": 4.2461946741760704e-05,
-      "loss": 0.2698,
+      "epoch": 1.415288139258298,
+      "grad_norm": 0.16008664667606354,
+      "learning_rate": 4.2069261515935656e-05,
+      "loss": 0.4325,
       "step": 39270
     },
     {
-      "epoch": 1.38,
-      "learning_rate": 4.2459907993251545e-05,
-      "loss": 0.2963,
+      "epoch": 1.4154683389195228,
+      "grad_norm": 0.20454731583595276,
+      "learning_rate": 4.206712931507886e-05,
+      "loss": 0.3848,
       "step": 39275
     },
     {
-      "epoch": 1.38,
-      "learning_rate": 4.245786901803698e-05,
-      "loss": 0.2669,
+      "epoch": 1.4156485385807476,
+      "grad_norm": 0.19149497151374817,
+      "learning_rate": 4.206499688168572e-05,
+      "loss": 0.4211,
       "step": 39280
     },
     {
-      "epoch": 1.38,
-      "learning_rate": 4.245582981614349e-05,
-      "loss": 0.3193,
+      "epoch": 1.415828738241972,
+      "grad_norm": 0.18504855036735535,
+      "learning_rate": 4.2062864215785304e-05,
+      "loss": 0.4501,
       "step": 39285
     },
     {
-      "epoch": 1.38,
-      "learning_rate": 4.245379038759756e-05,
-      "loss": 0.2918,
+      "epoch": 1.4160089379031968,
+      "grad_norm": 0.18371212482452393,
+      "learning_rate": 4.206073131740668e-05,
+      "loss": 0.4172,
       "step": 39290
     },
     {
-      "epoch": 1.38,
-      "learning_rate": 4.2451750732425665e-05,
-      "loss": 0.2896,
+      "epoch": 1.4161891375644213,
+      "grad_norm": 0.17143423855304718,
+      "learning_rate": 4.205859818657888e-05,
+      "loss": 0.3977,
       "step": 39295
     },
     {
-      "epoch": 1.38,
-      "learning_rate": 4.244971085065429e-05,
-      "loss": 0.295,
+      "epoch": 1.416369337225646,
+      "grad_norm": 0.18957088887691498,
+      "learning_rate": 4.205646482333098e-05,
+      "loss": 0.4387,
       "step": 39300
     },
     {
-      "epoch": 1.38,
-      "learning_rate": 4.244767074230991e-05,
-      "loss": 0.2789,
+      "epoch": 1.4165495368868706,
+      "grad_norm": 0.18221431970596313,
+      "learning_rate": 4.205433122769206e-05,
+      "loss": 0.3961,
       "step": 39305
     },
     {
-      "epoch": 1.38,
-      "learning_rate": 4.244563040741903e-05,
-      "loss": 0.2951,
+      "epoch": 1.4167297365480953,
+      "grad_norm": 0.1832798719406128,
+      "learning_rate": 4.2052197399691174e-05,
+      "loss": 0.4188,
       "step": 39310
     },
     {
-      "epoch": 1.38,
-      "learning_rate": 4.244358984600813e-05,
-      "loss": 0.2768,
+      "epoch": 1.4169099362093198,
+      "grad_norm": 0.16603875160217285,
+      "learning_rate": 4.205006333935739e-05,
+      "loss": 0.3877,
       "step": 39315
     },
     {
-      "epoch": 1.38,
-      "learning_rate": 4.244154905810371e-05,
-      "loss": 0.302,
+      "epoch": 1.4170901358705446,
+      "grad_norm": 0.17114681005477905,
+      "learning_rate": 4.204792904671981e-05,
+      "loss": 0.4252,
       "step": 39320
     },
     {
-      "epoch": 1.38,
-      "learning_rate": 4.243950804373228e-05,
-      "loss": 0.2966,
+      "epoch": 1.4172703355317693,
+      "grad_norm": 0.16899357736110687,
+      "learning_rate": 4.204579452180749e-05,
+      "loss": 0.392,
       "step": 39325
     },
     {
-      "epoch": 1.38,
-      "learning_rate": 4.243746680292033e-05,
-      "loss": 0.2965,
+      "epoch": 1.4174505351929938,
+      "grad_norm": 0.20568254590034485,
+      "learning_rate": 4.2043659764649527e-05,
+      "loss": 0.4218,
       "step": 39330
     },
     {
-      "epoch": 1.38,
-      "learning_rate": 4.2435425335694355e-05,
-      "loss": 0.2845,
+      "epoch": 1.4176307348542185,
+      "grad_norm": 0.217452272772789,
+      "learning_rate": 4.2041524775274985e-05,
+      "loss": 0.3834,
       "step": 39335
     },
     {
-      "epoch": 1.38,
-      "learning_rate": 4.2433383642080884e-05,
-      "loss": 0.3013,
+      "epoch": 1.417810934515443,
+      "grad_norm": 0.18286210298538208,
+      "learning_rate": 4.2039389553712986e-05,
+      "loss": 0.4243,
       "step": 39340
     },
     {
-      "epoch": 1.38,
-      "learning_rate": 4.243134172210641e-05,
-      "loss": 0.2755,
+      "epoch": 1.4179911341766678,
+      "grad_norm": 0.176735058426857,
+      "learning_rate": 4.2037254099992584e-05,
+      "loss": 0.392,
       "step": 39345
     },
     {
-      "epoch": 1.38,
-      "learning_rate": 4.2429299575797454e-05,
-      "loss": 0.2901,
+      "epoch": 1.4181713338378925,
+      "grad_norm": 0.18584541976451874,
+      "learning_rate": 4.2035118414142905e-05,
+      "loss": 0.415,
       "step": 39350
     },
     {
-      "epoch": 1.38,
-      "learning_rate": 4.2427257203180526e-05,
-      "loss": 0.2814,
+      "epoch": 1.418351533499117,
+      "grad_norm": 0.20618844032287598,
+      "learning_rate": 4.203298249619303e-05,
+      "loss": 0.4303,
       "step": 39355
     },
     {
-      "epoch": 1.38,
-      "learning_rate": 4.242521460428215e-05,
-      "loss": 0.2813,
+      "epoch": 1.4185317331603415,
+      "grad_norm": 0.1963784545660019,
+      "learning_rate": 4.203084634617207e-05,
+      "loss": 0.418,
       "step": 39360
     },
     {
-      "epoch": 1.38,
-      "learning_rate": 4.242317177912885e-05,
-      "loss": 0.2902,
+      "epoch": 1.4187119328215663,
+      "grad_norm": 0.19487901031970978,
+      "learning_rate": 4.202870996410913e-05,
+      "loss": 0.4129,
       "step": 39365
     },
     {
-      "epoch": 1.39,
-      "learning_rate": 4.2421128727747136e-05,
-      "loss": 0.2907,
+      "epoch": 1.418892132482791,
+      "grad_norm": 0.19040003418922424,
+      "learning_rate": 4.2026573350033304e-05,
+      "loss": 0.4042,
       "step": 39370
     },
     {
-      "epoch": 1.39,
-      "learning_rate": 4.2419085450163555e-05,
-      "loss": 0.2991,
+      "epoch": 1.4190723321440155,
+      "grad_norm": 0.21092675626277924,
+      "learning_rate": 4.2024436503973716e-05,
+      "loss": 0.4216,
       "step": 39375
     },
     {
-      "epoch": 1.39,
-      "learning_rate": 4.2417041946404626e-05,
-      "loss": 0.2856,
+      "epoch": 1.4192525318052402,
+      "grad_norm": 0.17777122557163239,
+      "learning_rate": 4.2022299425959476e-05,
+      "loss": 0.4004,
       "step": 39380
     },
     {
-      "epoch": 1.39,
-      "learning_rate": 4.241499821649689e-05,
-      "loss": 0.2923,
+      "epoch": 1.4194327314664648,
+      "grad_norm": 0.1416267305612564,
+      "learning_rate": 4.2020162116019706e-05,
+      "loss": 0.3728,
       "step": 39385
     },
     {
-      "epoch": 1.39,
-      "learning_rate": 4.2412954260466876e-05,
-      "loss": 0.3112,
+      "epoch": 1.4196129311276895,
+      "grad_norm": 0.1654488891363144,
+      "learning_rate": 4.2018024574183525e-05,
+      "loss": 0.412,
       "step": 39390
     },
     {
-      "epoch": 1.39,
-      "learning_rate": 4.241091007834113e-05,
-      "loss": 0.2896,
+      "epoch": 1.4197931307889142,
+      "grad_norm": 0.19146573543548584,
+      "learning_rate": 4.2015886800480044e-05,
+      "loss": 0.393,
       "step": 39395
     },
     {
-      "epoch": 1.39,
-      "learning_rate": 4.240886567014618e-05,
-      "loss": 0.3001,
+      "epoch": 1.4199733304501387,
+      "grad_norm": 0.17153365910053253,
+      "learning_rate": 4.20137487949384e-05,
+      "loss": 0.4341,
       "step": 39400
     },
     {
-      "epoch": 1.39,
-      "learning_rate": 4.240682103590859e-05,
-      "loss": 0.291,
+      "epoch": 1.4201535301113635,
+      "grad_norm": 0.24907280504703522,
+      "learning_rate": 4.201161055758773e-05,
+      "loss": 0.4512,
       "step": 39405
     },
     {
-      "epoch": 1.39,
-      "learning_rate": 4.2404776175654906e-05,
-      "loss": 0.2911,
+      "epoch": 1.420333729772588,
+      "grad_norm": 0.1950104534626007,
+      "learning_rate": 4.200947208845716e-05,
+      "loss": 0.438,
       "step": 39410
     },
     {
-      "epoch": 1.39,
-      "learning_rate": 4.2402731089411664e-05,
-      "loss": 0.2845,
+      "epoch": 1.4205139294338127,
+      "grad_norm": 0.16665388643741608,
+      "learning_rate": 4.200733338757582e-05,
+      "loss": 0.4124,
       "step": 39415
     },
     {
-      "epoch": 1.39,
-      "learning_rate": 4.240068577720543e-05,
-      "loss": 0.3,
+      "epoch": 1.4206941290950372,
+      "grad_norm": 0.18930195271968842,
+      "learning_rate": 4.2005194454972864e-05,
+      "loss": 0.4393,
       "step": 39420
     },
     {
-      "epoch": 1.39,
-      "learning_rate": 4.239864023906276e-05,
-      "loss": 0.2983,
+      "epoch": 1.420874328756262,
+      "grad_norm": 0.1515640914440155,
+      "learning_rate": 4.2003055290677416e-05,
+      "loss": 0.3813,
       "step": 39425
     },
     {
-      "epoch": 1.39,
-      "learning_rate": 4.239659447501021e-05,
-      "loss": 0.277,
+      "epoch": 1.4210545284174865,
+      "grad_norm": 0.14799198508262634,
+      "learning_rate": 4.200091589471863e-05,
+      "loss": 0.4036,
       "step": 39430
     },
     {
-      "epoch": 1.39,
-      "learning_rate": 4.239454848507435e-05,
-      "loss": 0.2754,
+      "epoch": 1.4212347280787112,
+      "grad_norm": 0.18082498013973236,
+      "learning_rate": 4.199877626712567e-05,
+      "loss": 0.4354,
       "step": 39435
     },
     {
-      "epoch": 1.39,
-      "learning_rate": 4.239250226928174e-05,
-      "loss": 0.2756,
+      "epoch": 1.421414927739936,
+      "grad_norm": 0.1811361163854599,
+      "learning_rate": 4.199663640792767e-05,
+      "loss": 0.4331,
       "step": 39440
     },
     {
-      "epoch": 1.39,
-      "learning_rate": 4.239045582765895e-05,
-      "loss": 0.2923,
+      "epoch": 1.4215951274011605,
+      "grad_norm": 0.1772868037223816,
+      "learning_rate": 4.199449631715378e-05,
+      "loss": 0.3759,
       "step": 39445
     },
     {
-      "epoch": 1.39,
-      "learning_rate": 4.2388409160232545e-05,
-      "loss": 0.2859,
+      "epoch": 1.4217753270623852,
+      "grad_norm": 0.17968173325061798,
+      "learning_rate": 4.1992355994833175e-05,
+      "loss": 0.3889,
       "step": 39450
     },
     {
-      "epoch": 1.39,
-      "learning_rate": 4.2386362267029116e-05,
-      "loss": 0.2837,
+      "epoch": 1.4219555267236097,
+      "grad_norm": 0.2115190178155899,
+      "learning_rate": 4.199021544099501e-05,
+      "loss": 0.3891,
       "step": 39455
     },
     {
-      "epoch": 1.39,
-      "learning_rate": 4.2384315148075225e-05,
-      "loss": 0.2795,
+      "epoch": 1.4221357263848344,
+      "grad_norm": 0.2003733217716217,
+      "learning_rate": 4.1988074655668446e-05,
+      "loss": 0.4274,
       "step": 39460
     },
     {
-      "epoch": 1.39,
-      "learning_rate": 4.2382267803397454e-05,
-      "loss": 0.2934,
+      "epoch": 1.4223159260460592,
+      "grad_norm": 0.1704542636871338,
+      "learning_rate": 4.1985933638882655e-05,
+      "loss": 0.4479,
       "step": 39465
     },
     {
-      "epoch": 1.39,
-      "learning_rate": 4.23802202330224e-05,
-      "loss": 0.2652,
+      "epoch": 1.4224961257072837,
+      "grad_norm": 0.18098214268684387,
+      "learning_rate": 4.198379239066681e-05,
+      "loss": 0.4312,
       "step": 39470
     },
     {
-      "epoch": 1.39,
-      "learning_rate": 4.2378172436976644e-05,
-      "loss": 0.3227,
+      "epoch": 1.4226763253685082,
+      "grad_norm": 0.17561987042427063,
+      "learning_rate": 4.198165091105007e-05,
+      "loss": 0.396,
       "step": 39475
     },
     {
-      "epoch": 1.39,
-      "learning_rate": 4.2376124415286764e-05,
-      "loss": 0.2581,
+      "epoch": 1.422856525029733,
+      "grad_norm": 0.14922846853733063,
+      "learning_rate": 4.197950920006164e-05,
+      "loss": 0.4461,
       "step": 39480
     },
     {
-      "epoch": 1.39,
-      "learning_rate": 4.237407616797936e-05,
-      "loss": 0.2916,
+      "epoch": 1.4230367246909577,
+      "grad_norm": 0.189297616481781,
+      "learning_rate": 4.197736725773068e-05,
+      "loss": 0.3796,
       "step": 39485
     },
     {
-      "epoch": 1.39,
-      "learning_rate": 4.237202769508103e-05,
-      "loss": 0.2714,
+      "epoch": 1.4232169243521822,
+      "grad_norm": 0.18948666751384735,
+      "learning_rate": 4.197522508408637e-05,
+      "loss": 0.4309,
       "step": 39490
     },
     {
-      "epoch": 1.39,
-      "learning_rate": 4.2369978996618365e-05,
-      "loss": 0.2715,
+      "epoch": 1.423397124013407,
+      "grad_norm": 0.21630705893039703,
+      "learning_rate": 4.197308267915791e-05,
+      "loss": 0.4242,
       "step": 39495
     },
     {
-      "epoch": 1.39,
-      "learning_rate": 4.236793007261798e-05,
-      "loss": 0.2822,
+      "epoch": 1.4235773236746314,
+      "grad_norm": 0.2401585429906845,
+      "learning_rate": 4.1970940042974485e-05,
+      "loss": 0.4483,
       "step": 39500
     },
     {
-      "epoch": 1.39,
-      "eval_loss": 0.28314095735549927,
-      "eval_runtime": 10.5583,
-      "eval_samples_per_second": 9.471,
-      "eval_steps_per_second": 9.471,
+      "epoch": 1.4235773236746314,
+      "eval_loss": 0.44728443026542664,
+      "eval_runtime": 3.5187,
+      "eval_samples_per_second": 28.419,
+      "eval_steps_per_second": 7.105,
       "step": 39500
     },
     {
-      "epoch": 1.39,
-      "learning_rate": 4.236588092310646e-05,
-      "loss": 0.2802,
+      "epoch": 1.4237575233358561,
+      "grad_norm": 0.2015465348958969,
+      "learning_rate": 4.196879717556529e-05,
+      "loss": 0.4352,
       "step": 39505
     },
     {
-      "epoch": 1.39,
-      "learning_rate": 4.236383154811042e-05,
-      "loss": 0.284,
+      "epoch": 1.4239377229970809,
+      "grad_norm": 0.17503204941749573,
+      "learning_rate": 4.1966654076959516e-05,
+      "loss": 0.3889,
       "step": 39510
     },
     {
-      "epoch": 1.39,
-      "learning_rate": 4.236178194765647e-05,
-      "loss": 0.3279,
+      "epoch": 1.4241179226583054,
+      "grad_norm": 0.15715694427490234,
+      "learning_rate": 4.196451074718637e-05,
+      "loss": 0.4131,
       "step": 39515
     },
     {
-      "epoch": 1.39,
-      "learning_rate": 4.2359732121771226e-05,
-      "loss": 0.2795,
+      "epoch": 1.42429812231953,
+      "grad_norm": 0.20377200841903687,
+      "learning_rate": 4.196236718627504e-05,
+      "loss": 0.4144,
       "step": 39520
     },
     {
-      "epoch": 1.39,
-      "learning_rate": 4.23576820704813e-05,
-      "loss": 0.2839,
+      "epoch": 1.4244783219807546,
+      "grad_norm": 0.18509119749069214,
+      "learning_rate": 4.196022339425475e-05,
+      "loss": 0.4244,
       "step": 39525
     },
     {
-      "epoch": 1.39,
-      "learning_rate": 4.235563179381331e-05,
-      "loss": 0.271,
+      "epoch": 1.4246585216419794,
+      "grad_norm": 0.1515013426542282,
+      "learning_rate": 4.195807937115469e-05,
+      "loss": 0.4429,
       "step": 39530
     },
     {
-      "epoch": 1.39,
-      "learning_rate": 4.235358129179389e-05,
-      "loss": 0.2893,
+      "epoch": 1.4248387213032039,
+      "grad_norm": 0.2172863632440567,
+      "learning_rate": 4.1955935117004095e-05,
+      "loss": 0.3957,
       "step": 39535
     },
     {
-      "epoch": 1.39,
-      "learning_rate": 4.2351530564449635e-05,
-      "loss": 0.2983,
+      "epoch": 1.4250189209644286,
+      "grad_norm": 0.16076506674289703,
+      "learning_rate": 4.1953790631832156e-05,
+      "loss": 0.426,
       "step": 39540
     },
     {
-      "epoch": 1.39,
-      "learning_rate": 4.23494796118072e-05,
-      "loss": 0.2856,
+      "epoch": 1.4251991206256531,
+      "grad_norm": 0.24314238131046295,
+      "learning_rate": 4.1951645915668105e-05,
+      "loss": 0.4287,
       "step": 39545
     },
     {
-      "epoch": 1.39,
-      "learning_rate": 4.234742843389321e-05,
-      "loss": 0.2899,
+      "epoch": 1.4253793202868779,
+      "grad_norm": 0.22566382586956024,
+      "learning_rate": 4.1949500968541154e-05,
+      "loss": 0.4242,
       "step": 39550
     },
     {
-      "epoch": 1.39,
-      "learning_rate": 4.234537703073429e-05,
-      "loss": 0.2832,
+      "epoch": 1.4255595199481026,
+      "grad_norm": 0.1787545531988144,
+      "learning_rate": 4.194735579048055e-05,
+      "loss": 0.4171,
       "step": 39555
     },
     {
-      "epoch": 1.39,
-      "learning_rate": 4.234332540235708e-05,
-      "loss": 0.3104,
+      "epoch": 1.425739719609327,
+      "grad_norm": 0.142437145113945,
+      "learning_rate": 4.1945210381515485e-05,
+      "loss": 0.4205,
       "step": 39560
     },
     {
-      "epoch": 1.39,
-      "learning_rate": 4.234127354878822e-05,
-      "loss": 0.2856,
+      "epoch": 1.4259199192705518,
+      "grad_norm": 0.16321620345115662,
+      "learning_rate": 4.194306474167522e-05,
+      "loss": 0.4164,
       "step": 39565
     },
     {
-      "epoch": 1.39,
-      "learning_rate": 4.2339221470054355e-05,
-      "loss": 0.2836,
+      "epoch": 1.4261001189317764,
+      "grad_norm": 0.24486680328845978,
+      "learning_rate": 4.1940918870988976e-05,
+      "loss": 0.4363,
       "step": 39570
     },
     {
-      "epoch": 1.39,
-      "learning_rate": 4.2337169166182125e-05,
-      "loss": 0.2826,
+      "epoch": 1.426280318593001,
+      "grad_norm": 0.23074302077293396,
+      "learning_rate": 4.1938772769486e-05,
+      "loss": 0.4505,
       "step": 39575
     },
     {
-      "epoch": 1.39,
-      "learning_rate": 4.233511663719817e-05,
-      "loss": 0.2842,
+      "epoch": 1.4264605182542258,
+      "grad_norm": 0.1574547439813614,
+      "learning_rate": 4.193662643719552e-05,
+      "loss": 0.3959,
       "step": 39580
     },
     {
-      "epoch": 1.39,
-      "learning_rate": 4.2333063883129156e-05,
-      "loss": 0.2884,
+      "epoch": 1.4266407179154503,
+      "grad_norm": 0.18195997178554535,
+      "learning_rate": 4.193447987414678e-05,
+      "loss": 0.414,
       "step": 39585
     },
     {
-      "epoch": 1.39,
-      "learning_rate": 4.233101090400173e-05,
-      "loss": 0.2684,
+      "epoch": 1.4268209175766748,
+      "grad_norm": 0.19354818761348724,
+      "learning_rate": 4.1932333080369036e-05,
+      "loss": 0.4372,
       "step": 39590
     },
     {
-      "epoch": 1.39,
-      "learning_rate": 4.232895769984255e-05,
-      "loss": 0.2712,
+      "epoch": 1.4270011172378996,
+      "grad_norm": 0.1999470740556717,
+      "learning_rate": 4.1930186055891525e-05,
+      "loss": 0.4069,
       "step": 39595
     },
     {
-      "epoch": 1.39,
-      "learning_rate": 4.232690427067827e-05,
-      "loss": 0.2711,
+      "epoch": 1.4271813168991243,
+      "grad_norm": 0.18640463054180145,
+      "learning_rate": 4.1928038800743504e-05,
+      "loss": 0.4398,
       "step": 39600
     },
     {
-      "epoch": 1.39,
-      "learning_rate": 4.232485061653556e-05,
-      "loss": 0.3023,
+      "epoch": 1.4273615165603488,
+      "grad_norm": 0.18737342953681946,
+      "learning_rate": 4.192589131495424e-05,
+      "loss": 0.4102,
       "step": 39605
     },
     {
-      "epoch": 1.39,
-      "learning_rate": 4.232279673744108e-05,
-      "loss": 0.292,
+      "epoch": 1.4275417162215736,
+      "grad_norm": 0.16932520270347595,
+      "learning_rate": 4.1923743598552984e-05,
+      "loss": 0.4147,
       "step": 39610
     },
     {
-      "epoch": 1.39,
-      "learning_rate": 4.232074263342151e-05,
-      "loss": 0.2956,
+      "epoch": 1.427721915882798,
+      "grad_norm": 0.14350475370883942,
+      "learning_rate": 4.192159565156899e-05,
+      "loss": 0.4086,
       "step": 39615
     },
     {
-      "epoch": 1.39,
-      "learning_rate": 4.2318688304503505e-05,
-      "loss": 0.3008,
+      "epoch": 1.4279021155440228,
+      "grad_norm": 0.19839464128017426,
+      "learning_rate": 4.1919447474031546e-05,
+      "loss": 0.4067,
       "step": 39620
     },
     {
-      "epoch": 1.39,
-      "learning_rate": 4.231663375071375e-05,
-      "loss": 0.2826,
+      "epoch": 1.4280823152052475,
+      "grad_norm": 0.17800691723823547,
+      "learning_rate": 4.19172990659699e-05,
+      "loss": 0.4202,
       "step": 39625
     },
     {
-      "epoch": 1.39,
-      "learning_rate": 4.231457897207891e-05,
-      "loss": 0.2865,
+      "epoch": 1.428262514866472,
+      "grad_norm": 0.19023603200912476,
+      "learning_rate": 4.191515042741332e-05,
+      "loss": 0.394,
       "step": 39630
     },
     {
-      "epoch": 1.39,
-      "learning_rate": 4.231252396862568e-05,
-      "loss": 0.2763,
+      "epoch": 1.4284427145276966,
+      "grad_norm": 0.14899520576000214,
+      "learning_rate": 4.1913001558391095e-05,
+      "loss": 0.4398,
       "step": 39635
     },
     {
-      "epoch": 1.39,
-      "learning_rate": 4.2310468740380725e-05,
-      "loss": 0.315,
+      "epoch": 1.4286229141889213,
+      "grad_norm": 0.1971236616373062,
+      "learning_rate": 4.19108524589325e-05,
+      "loss": 0.407,
       "step": 39640
     },
     {
-      "epoch": 1.39,
-      "learning_rate": 4.230841328737075e-05,
-      "loss": 0.3025,
+      "epoch": 1.428803113850146,
+      "grad_norm": 0.16257238388061523,
+      "learning_rate": 4.190870312906682e-05,
+      "loss": 0.4173,
       "step": 39645
     },
     {
-      "epoch": 1.39,
-      "learning_rate": 4.230635760962244e-05,
-      "loss": 0.2954,
+      "epoch": 1.4289833135113705,
+      "grad_norm": 0.19401347637176514,
+      "learning_rate": 4.190655356882332e-05,
+      "loss": 0.4268,
       "step": 39650
     },
     {
-      "epoch": 1.4,
-      "learning_rate": 4.2304301707162475e-05,
-      "loss": 0.2741,
+      "epoch": 1.4291635131725953,
+      "grad_norm": 0.16988852620124817,
+      "learning_rate": 4.1904403778231313e-05,
+      "loss": 0.4242,
       "step": 39655
     },
     {
-      "epoch": 1.4,
-      "learning_rate": 4.230224558001755e-05,
-      "loss": 0.2777,
+      "epoch": 1.4293437128338198,
+      "grad_norm": 0.16938932240009308,
+      "learning_rate": 4.1902253757320074e-05,
+      "loss": 0.4149,
       "step": 39660
     },
     {
-      "epoch": 1.4,
-      "learning_rate": 4.230018922821438e-05,
-      "loss": 0.2835,
+      "epoch": 1.4295239124950445,
+      "grad_norm": 0.2068401277065277,
+      "learning_rate": 4.190010350611889e-05,
+      "loss": 0.4358,
       "step": 39665
     },
     {
-      "epoch": 1.4,
-      "learning_rate": 4.229813265177965e-05,
-      "loss": 0.2953,
+      "epoch": 1.4297041121562692,
+      "grad_norm": 0.1826406568288803,
+      "learning_rate": 4.1897953024657084e-05,
+      "loss": 0.4312,
       "step": 39670
     },
     {
-      "epoch": 1.4,
-      "learning_rate": 4.2296075850740066e-05,
-      "loss": 0.2712,
+      "epoch": 1.4298843118174938,
+      "grad_norm": 0.17122584581375122,
+      "learning_rate": 4.189580231296393e-05,
+      "loss": 0.44,
       "step": 39675
     },
     {
-      "epoch": 1.4,
-      "learning_rate": 4.229401882512234e-05,
-      "loss": 0.3011,
+      "epoch": 1.4300645114787185,
+      "grad_norm": 0.18926846981048584,
+      "learning_rate": 4.1893651371068743e-05,
+      "loss": 0.4346,
       "step": 39680
     },
     {
-      "epoch": 1.4,
-      "learning_rate": 4.229196157495318e-05,
-      "loss": 0.2996,
+      "epoch": 1.430244711139943,
+      "grad_norm": 0.16383080184459686,
+      "learning_rate": 4.1891500199000827e-05,
+      "loss": 0.3993,
       "step": 39685
     },
     {
-      "epoch": 1.4,
-      "learning_rate": 4.228990410025928e-05,
-      "loss": 0.2695,
+      "epoch": 1.4304249108011677,
+      "grad_norm": 0.1707577407360077,
+      "learning_rate": 4.1889348796789484e-05,
+      "loss": 0.3975,
       "step": 39690
     },
     {
-      "epoch": 1.4,
-      "learning_rate": 4.2287846401067374e-05,
-      "loss": 0.2822,
+      "epoch": 1.4306051104623922,
+      "grad_norm": 0.20824764668941498,
+      "learning_rate": 4.188719716446404e-05,
+      "loss": 0.417,
       "step": 39695
     },
     {
-      "epoch": 1.4,
-      "learning_rate": 4.228578847740419e-05,
-      "loss": 0.2804,
+      "epoch": 1.430785310123617,
+      "grad_norm": 0.15967229008674622,
+      "learning_rate": 4.188504530205381e-05,
+      "loss": 0.404,
       "step": 39700
     },
     {
-      "epoch": 1.4,
-      "learning_rate": 4.2283730329296417e-05,
-      "loss": 0.2755,
+      "epoch": 1.4309655097848415,
+      "grad_norm": 0.18324309587478638,
+      "learning_rate": 4.1882893209588104e-05,
+      "loss": 0.4453,
       "step": 39705
     },
     {
-      "epoch": 1.4,
-      "learning_rate": 4.228167195677081e-05,
-      "loss": 0.3029,
+      "epoch": 1.4311457094460662,
+      "grad_norm": 0.1739649921655655,
+      "learning_rate": 4.188074088709624e-05,
+      "loss": 0.4023,
       "step": 39710
     },
     {
-      "epoch": 1.4,
-      "learning_rate": 4.227961335985407e-05,
-      "loss": 0.2772,
+      "epoch": 1.431325909107291,
+      "grad_norm": 0.17934979498386383,
+      "learning_rate": 4.187858833460755e-05,
+      "loss": 0.4329,
       "step": 39715
     },
     {
-      "epoch": 1.4,
-      "learning_rate": 4.227755453857295e-05,
-      "loss": 0.2997,
+      "epoch": 1.4315061087685155,
+      "grad_norm": 0.19043073058128357,
+      "learning_rate": 4.187643555215137e-05,
+      "loss": 0.3626,
       "step": 39720
     },
     {
-      "epoch": 1.4,
-      "learning_rate": 4.227549549295415e-05,
-      "loss": 0.2893,
+      "epoch": 1.4316863084297402,
+      "grad_norm": 0.1956600397825241,
+      "learning_rate": 4.187428253975702e-05,
+      "loss": 0.4218,
       "step": 39725
     },
     {
-      "epoch": 1.4,
-      "learning_rate": 4.2273436223024445e-05,
-      "loss": 0.2749,
+      "epoch": 1.4318665080909647,
+      "grad_norm": 0.16748665273189545,
+      "learning_rate": 4.1872129297453835e-05,
+      "loss": 0.4228,
       "step": 39730
     },
     {
-      "epoch": 1.4,
-      "learning_rate": 4.227137672881055e-05,
-      "loss": 0.2954,
+      "epoch": 1.4320467077521895,
+      "grad_norm": 0.24490876495838165,
+      "learning_rate": 4.186997582527115e-05,
+      "loss": 0.4301,
       "step": 39735
     },
     {
-      "epoch": 1.4,
-      "learning_rate": 4.2269317010339215e-05,
-      "loss": 0.2996,
+      "epoch": 1.4322269074134142,
+      "grad_norm": 0.19111377000808716,
+      "learning_rate": 4.186782212323832e-05,
+      "loss": 0.4255,
       "step": 39740
     },
     {
-      "epoch": 1.4,
-      "learning_rate": 4.226725706763717e-05,
-      "loss": 0.2949,
+      "epoch": 1.4324071070746387,
+      "grad_norm": 0.236825630068779,
+      "learning_rate": 4.186566819138467e-05,
+      "loss": 0.4159,
       "step": 39745
     },
     {
-      "epoch": 1.4,
-      "learning_rate": 4.2265196900731176e-05,
-      "loss": 0.296,
+      "epoch": 1.4325873067358632,
+      "grad_norm": 0.17333528399467468,
+      "learning_rate": 4.186351402973956e-05,
+      "loss": 0.4423,
       "step": 39750
     },
     {
-      "epoch": 1.4,
-      "learning_rate": 4.226313650964798e-05,
-      "loss": 0.288,
+      "epoch": 1.432767506397088,
+      "grad_norm": 0.15992626547813416,
+      "learning_rate": 4.1861359638332343e-05,
+      "loss": 0.4119,
       "step": 39755
     },
     {
-      "epoch": 1.4,
-      "learning_rate": 4.226107589441432e-05,
-      "loss": 0.2819,
+      "epoch": 1.4329477060583127,
+      "grad_norm": 0.18964506685733795,
+      "learning_rate": 4.185920501719236e-05,
+      "loss": 0.4394,
       "step": 39760
     },
     {
-      "epoch": 1.4,
-      "learning_rate": 4.225901505505698e-05,
-      "loss": 0.2796,
+      "epoch": 1.4331279057195372,
+      "grad_norm": 0.1828775703907013,
+      "learning_rate": 4.185705016634897e-05,
+      "loss": 0.4115,
       "step": 39765
     },
     {
-      "epoch": 1.4,
-      "learning_rate": 4.2256953991602705e-05,
-      "loss": 0.262,
+      "epoch": 1.433308105380762,
+      "grad_norm": 0.17985029518604279,
+      "learning_rate": 4.1854895085831533e-05,
+      "loss": 0.4361,
       "step": 39770
     },
     {
-      "epoch": 1.4,
-      "learning_rate": 4.225489270407825e-05,
-      "loss": 0.2852,
+      "epoch": 1.4334883050419864,
+      "grad_norm": 0.19158676266670227,
+      "learning_rate": 4.1852739775669425e-05,
+      "loss": 0.4207,
       "step": 39775
     },
     {
-      "epoch": 1.4,
-      "learning_rate": 4.225283119251039e-05,
-      "loss": 0.2634,
+      "epoch": 1.4336685047032112,
+      "grad_norm": 0.1590370237827301,
+      "learning_rate": 4.185058423589199e-05,
+      "loss": 0.4402,
       "step": 39780
     },
     {
-      "epoch": 1.4,
-      "learning_rate": 4.225076945692587e-05,
-      "loss": 0.2904,
+      "epoch": 1.433848704364436,
+      "grad_norm": 0.17683321237564087,
+      "learning_rate": 4.184842846652861e-05,
+      "loss": 0.4283,
       "step": 39785
     },
     {
-      "epoch": 1.4,
-      "learning_rate": 4.224870749735149e-05,
-      "loss": 0.2756,
+      "epoch": 1.4340289040256604,
+      "grad_norm": 0.16492830216884613,
+      "learning_rate": 4.1846272467608655e-05,
+      "loss": 0.3904,
       "step": 39790
     },
     {
-      "epoch": 1.4,
-      "learning_rate": 4.224664531381402e-05,
-      "loss": 0.2935,
+      "epoch": 1.4342091036868851,
+      "grad_norm": 0.22784380614757538,
+      "learning_rate": 4.18441162391615e-05,
+      "loss": 0.4632,
       "step": 39795
     },
     {
-      "epoch": 1.4,
-      "learning_rate": 4.224458290634021e-05,
-      "loss": 0.2877,
+      "epoch": 1.4343893033481097,
+      "grad_norm": 0.17077568173408508,
+      "learning_rate": 4.184195978121652e-05,
+      "loss": 0.4248,
       "step": 39800
     },
     {
-      "epoch": 1.4,
-      "learning_rate": 4.224252027495686e-05,
-      "loss": 0.2858,
+      "epoch": 1.4345695030093344,
+      "grad_norm": 0.17400763928890228,
+      "learning_rate": 4.1839803093803106e-05,
+      "loss": 0.4172,
       "step": 39805
     },
     {
-      "epoch": 1.4,
-      "learning_rate": 4.224045741969075e-05,
-      "loss": 0.2794,
+      "epoch": 1.434749702670559,
+      "grad_norm": 0.1851196140050888,
+      "learning_rate": 4.183764617695063e-05,
+      "loss": 0.4743,
       "step": 39810
     },
     {
-      "epoch": 1.4,
-      "learning_rate": 4.223839434056866e-05,
-      "loss": 0.3027,
+      "epoch": 1.4349299023317836,
+      "grad_norm": 0.16248086094856262,
+      "learning_rate": 4.183548903068848e-05,
+      "loss": 0.4334,
       "step": 39815
     },
     {
-      "epoch": 1.4,
-      "learning_rate": 4.223633103761738e-05,
-      "loss": 0.2658,
+      "epoch": 1.4351101019930081,
+      "grad_norm": 0.19981388747692108,
+      "learning_rate": 4.1833331655046055e-05,
+      "loss": 0.4594,
       "step": 39820
     },
     {
-      "epoch": 1.4,
-      "learning_rate": 4.22342675108637e-05,
-      "loss": 0.2767,
+      "epoch": 1.4352903016542329,
+      "grad_norm": 0.21425208449363708,
+      "learning_rate": 4.1831174050052745e-05,
+      "loss": 0.4465,
       "step": 39825
     },
     {
-      "epoch": 1.4,
-      "learning_rate": 4.223220376033442e-05,
-      "loss": 0.2956,
+      "epoch": 1.4354705013154576,
+      "grad_norm": 0.16509120166301727,
+      "learning_rate": 4.182901621573795e-05,
+      "loss": 0.4481,
       "step": 39830
     },
     {
-      "epoch": 1.4,
-      "learning_rate": 4.223013978605633e-05,
-      "loss": 0.2936,
+      "epoch": 1.4356507009766821,
+      "grad_norm": 0.18470360338687897,
+      "learning_rate": 4.1826858152131064e-05,
+      "loss": 0.4352,
       "step": 39835
     },
     {
-      "epoch": 1.4,
-      "learning_rate": 4.222807558805622e-05,
-      "loss": 0.3004,
+      "epoch": 1.4358309006379069,
+      "grad_norm": 0.1776283085346222,
+      "learning_rate": 4.182469985926149e-05,
+      "loss": 0.3918,
       "step": 39840
     },
     {
-      "epoch": 1.4,
-      "learning_rate": 4.222601116636091e-05,
-      "loss": 0.2876,
+      "epoch": 1.4360111002991314,
+      "grad_norm": 0.16285589337348938,
+      "learning_rate": 4.1822541337158646e-05,
+      "loss": 0.4151,
       "step": 39845
     },
     {
-      "epoch": 1.4,
-      "learning_rate": 4.22239465209972e-05,
-      "loss": 0.2749,
+      "epoch": 1.436191299960356,
+      "grad_norm": 0.15581516921520233,
+      "learning_rate": 4.1820382585851925e-05,
+      "loss": 0.4004,
       "step": 39850
     },
     {
-      "epoch": 1.4,
-      "learning_rate": 4.222188165199189e-05,
-      "loss": 0.293,
+      "epoch": 1.4363714996215808,
+      "grad_norm": 0.15109944343566895,
+      "learning_rate": 4.1818223605370756e-05,
+      "loss": 0.4049,
       "step": 39855
     },
     {
-      "epoch": 1.4,
-      "learning_rate": 4.221981655937179e-05,
-      "loss": 0.3175,
+      "epoch": 1.4365516992828053,
+      "grad_norm": 0.1799648404121399,
+      "learning_rate": 4.1816064395744536e-05,
+      "loss": 0.4167,
       "step": 39860
     },
     {
-      "epoch": 1.4,
-      "learning_rate": 4.221775124316374e-05,
-      "loss": 0.2766,
+      "epoch": 1.4367318989440299,
+      "grad_norm": 0.19493626058101654,
+      "learning_rate": 4.181390495700271e-05,
+      "loss": 0.4467,
       "step": 39865
     },
     {
-      "epoch": 1.4,
-      "learning_rate": 4.2215685703394525e-05,
-      "loss": 0.2864,
+      "epoch": 1.4369120986052546,
+      "grad_norm": 0.21801850199699402,
+      "learning_rate": 4.181174528917468e-05,
+      "loss": 0.4298,
       "step": 39870
     },
     {
-      "epoch": 1.4,
-      "learning_rate": 4.221361994009098e-05,
-      "loss": 0.2821,
+      "epoch": 1.4370922982664793,
+      "grad_norm": 0.24591484665870667,
+      "learning_rate": 4.1809585392289865e-05,
+      "loss": 0.4069,
       "step": 39875
     },
     {
-      "epoch": 1.4,
-      "learning_rate": 4.2211553953279914e-05,
-      "loss": 0.2932,
+      "epoch": 1.4372724979277038,
+      "grad_norm": 0.16199952363967896,
+      "learning_rate": 4.180742526637771e-05,
+      "loss": 0.4273,
       "step": 39880
     },
     {
-      "epoch": 1.4,
-      "learning_rate": 4.220948774298817e-05,
-      "loss": 0.2625,
+      "epoch": 1.4374526975889286,
+      "grad_norm": 0.16908341646194458,
+      "learning_rate": 4.180526491146764e-05,
+      "loss": 0.4104,
       "step": 39885
     },
     {
-      "epoch": 1.4,
-      "learning_rate": 4.220742130924257e-05,
-      "loss": 0.2784,
+      "epoch": 1.437632897250153,
+      "grad_norm": 0.18602709472179413,
+      "learning_rate": 4.180310432758908e-05,
+      "loss": 0.4457,
       "step": 39890
     },
     {
-      "epoch": 1.4,
-      "learning_rate": 4.2205354652069946e-05,
-      "loss": 0.2753,
+      "epoch": 1.4378130969113778,
+      "grad_norm": 0.21784497797489166,
+      "learning_rate": 4.1800943514771486e-05,
+      "loss": 0.4454,
       "step": 39895
     },
     {
-      "epoch": 1.4,
-      "learning_rate": 4.2203287771497125e-05,
-      "loss": 0.3228,
+      "epoch": 1.4379932965726026,
+      "grad_norm": 0.166326642036438,
+      "learning_rate": 4.179878247304429e-05,
+      "loss": 0.4221,
       "step": 39900
     },
     {
-      "epoch": 1.4,
-      "learning_rate": 4.2201220667550955e-05,
-      "loss": 0.2661,
+      "epoch": 1.438173496233827,
+      "grad_norm": 0.17365773022174835,
+      "learning_rate": 4.1796621202436934e-05,
+      "loss": 0.3897,
       "step": 39905
     },
     {
-      "epoch": 1.4,
-      "learning_rate": 4.219915334025827e-05,
-      "loss": 0.2924,
+      "epoch": 1.4383536958950518,
+      "grad_norm": 0.231038898229599,
+      "learning_rate": 4.179445970297887e-05,
+      "loss": 0.4301,
       "step": 39910
     },
     {
-      "epoch": 1.4,
-      "learning_rate": 4.219708578964591e-05,
-      "loss": 0.2912,
+      "epoch": 1.4385338955562763,
+      "grad_norm": 0.19160155951976776,
+      "learning_rate": 4.179229797469954e-05,
+      "loss": 0.3991,
       "step": 39915
     },
     {
-      "epoch": 1.4,
-      "learning_rate": 4.2195018015740726e-05,
-      "loss": 0.2761,
+      "epoch": 1.438714095217501,
+      "grad_norm": 0.1742529571056366,
+      "learning_rate": 4.179013601762839e-05,
+      "loss": 0.4099,
       "step": 39920
     },
     {
-      "epoch": 1.4,
-      "learning_rate": 4.219295001856957e-05,
-      "loss": 0.2652,
+      "epoch": 1.4388942948787256,
+      "grad_norm": 0.17680296301841736,
+      "learning_rate": 4.17879738317949e-05,
+      "loss": 0.4349,
       "step": 39925
     },
     {
-      "epoch": 1.4,
-      "learning_rate": 4.219088179815928e-05,
-      "loss": 0.2966,
+      "epoch": 1.4390744945399503,
+      "grad_norm": 0.21508027613162994,
+      "learning_rate": 4.1785811417228513e-05,
+      "loss": 0.4284,
       "step": 39930
     },
     {
-      "epoch": 1.41,
-      "learning_rate": 4.218881335453673e-05,
-      "loss": 0.2974,
+      "epoch": 1.4392546942011748,
+      "grad_norm": 0.19093069434165955,
+      "learning_rate": 4.1783648773958706e-05,
+      "loss": 0.4541,
       "step": 39935
     },
     {
-      "epoch": 1.41,
-      "learning_rate": 4.218674468772875e-05,
-      "loss": 0.2852,
+      "epoch": 1.4394348938623995,
+      "grad_norm": 0.18556737899780273,
+      "learning_rate": 4.178148590201492e-05,
+      "loss": 0.416,
       "step": 39940
     },
     {
-      "epoch": 1.41,
-      "learning_rate": 4.218467579776223e-05,
-      "loss": 0.2864,
+      "epoch": 1.4396150935236243,
+      "grad_norm": 0.17068710923194885,
+      "learning_rate": 4.177932280142665e-05,
+      "loss": 0.4588,
       "step": 39945
     },
     {
-      "epoch": 1.41,
-      "learning_rate": 4.218260668466402e-05,
-      "loss": 0.2779,
+      "epoch": 1.4397952931848488,
+      "grad_norm": 0.195718452334404,
+      "learning_rate": 4.177715947222334e-05,
+      "loss": 0.4256,
       "step": 39950
     },
     {
-      "epoch": 1.41,
-      "learning_rate": 4.218053734846098e-05,
-      "loss": 0.2762,
+      "epoch": 1.4399754928460735,
+      "grad_norm": 0.21339944005012512,
+      "learning_rate": 4.177499591443449e-05,
+      "loss": 0.4219,
       "step": 39955
     },
     {
-      "epoch": 1.41,
-      "learning_rate": 4.2178467789179986e-05,
-      "loss": 0.2944,
+      "epoch": 1.440155692507298,
+      "grad_norm": 0.21166032552719116,
+      "learning_rate": 4.1772832128089564e-05,
+      "loss": 0.4104,
       "step": 39960
     },
     {
-      "epoch": 1.41,
-      "learning_rate": 4.217639800684791e-05,
-      "loss": 0.2905,
+      "epoch": 1.4403358921685228,
+      "grad_norm": 0.1573396623134613,
+      "learning_rate": 4.177066811321805e-05,
+      "loss": 0.4435,
       "step": 39965
     },
     {
-      "epoch": 1.41,
-      "learning_rate": 4.217432800149162e-05,
-      "loss": 0.2807,
+      "epoch": 1.4405160918297475,
+      "grad_norm": 0.2001795619726181,
+      "learning_rate": 4.176850386984943e-05,
+      "loss": 0.3743,
       "step": 39970
     },
     {
-      "epoch": 1.41,
-      "learning_rate": 4.217267183664723e-05,
-      "loss": 0.2647,
+      "epoch": 1.440696291490972,
+      "grad_norm": 0.19709444046020508,
+      "learning_rate": 4.176633939801319e-05,
+      "loss": 0.4321,
       "step": 39975
     },
     {
-      "epoch": 1.41,
-      "learning_rate": 4.2170601429915105e-05,
-      "loss": 0.3072,
+      "epoch": 1.4408764911521965,
+      "grad_norm": 0.1635955274105072,
+      "learning_rate": 4.176417469773882e-05,
+      "loss": 0.4251,
       "step": 39980
     },
     {
-      "epoch": 1.41,
-      "learning_rate": 4.2168530800234035e-05,
-      "loss": 0.2968,
+      "epoch": 1.4410566908134212,
+      "grad_norm": 0.15509602427482605,
+      "learning_rate": 4.1762009769055835e-05,
+      "loss": 0.4027,
       "step": 39985
     },
     {
-      "epoch": 1.41,
-      "learning_rate": 4.216645994763091e-05,
-      "loss": 0.272,
+      "epoch": 1.441236890474646,
+      "grad_norm": 0.1558351367712021,
+      "learning_rate": 4.1759844611993685e-05,
+      "loss": 0.4319,
       "step": 39990
     },
     {
-      "epoch": 1.41,
-      "learning_rate": 4.2164388872132607e-05,
-      "loss": 0.2814,
+      "epoch": 1.4414170901358705,
+      "grad_norm": 0.19999557733535767,
+      "learning_rate": 4.175767922658191e-05,
+      "loss": 0.3745,
       "step": 39995
     },
     {
-      "epoch": 1.41,
-      "learning_rate": 4.216231757376603e-05,
-      "loss": 0.3099,
+      "epoch": 1.4415972897970952,
+      "grad_norm": 0.16591276228427887,
+      "learning_rate": 4.1755513612849993e-05,
+      "loss": 0.4251,
       "step": 40000
     },
     {
-      "epoch": 1.41,
-      "eval_loss": 0.2824369966983795,
-      "eval_runtime": 10.5321,
-      "eval_samples_per_second": 9.495,
-      "eval_steps_per_second": 9.495,
+      "epoch": 1.4415972897970952,
+      "eval_loss": 0.4461284279823303,
+      "eval_runtime": 3.5457,
+      "eval_samples_per_second": 28.203,
+      "eval_steps_per_second": 7.051,
       "step": 40000
     },
     {
-      "epoch": 1.41,
-      "learning_rate": 4.216024605255807e-05,
-      "loss": 0.2774,
+      "epoch": 1.4417774894583197,
+      "grad_norm": 0.1718062311410904,
+      "learning_rate": 4.1753347770827454e-05,
+      "loss": 0.4254,
       "step": 40005
     },
     {
-      "epoch": 1.41,
-      "learning_rate": 4.215817430853563e-05,
-      "loss": 0.2988,
+      "epoch": 1.4419576891195445,
+      "grad_norm": 0.19657757878303528,
+      "learning_rate": 4.17511817005438e-05,
+      "loss": 0.4246,
       "step": 40010
     },
     {
-      "epoch": 1.41,
-      "learning_rate": 4.21561023417256e-05,
-      "loss": 0.3029,
+      "epoch": 1.4421378887807692,
+      "grad_norm": 0.18001805245876312,
+      "learning_rate": 4.1749015402028526e-05,
+      "loss": 0.409,
       "step": 40015
     },
     {
-      "epoch": 1.41,
-      "learning_rate": 4.215403015215489e-05,
-      "loss": 0.2871,
+      "epoch": 1.4423180884419937,
+      "grad_norm": 0.158418208360672,
+      "learning_rate": 4.174684887531116e-05,
+      "loss": 0.3819,
       "step": 40020
     },
     {
-      "epoch": 1.41,
-      "learning_rate": 4.21519577398504e-05,
-      "loss": 0.291,
+      "epoch": 1.4424982881032182,
+      "grad_norm": 0.23636119067668915,
+      "learning_rate": 4.174468212042123e-05,
+      "loss": 0.399,
       "step": 40025
     },
     {
-      "epoch": 1.41,
-      "learning_rate": 4.214988510483904e-05,
-      "loss": 0.2806,
+      "epoch": 1.442678487764443,
+      "grad_norm": 0.1882403939962387,
+      "learning_rate": 4.1742515137388246e-05,
+      "loss": 0.4072,
       "step": 40030
     },
     {
-      "epoch": 1.41,
-      "learning_rate": 4.214781224714773e-05,
-      "loss": 0.286,
+      "epoch": 1.4428586874256677,
+      "grad_norm": 0.17548276484012604,
+      "learning_rate": 4.174034792624173e-05,
+      "loss": 0.4078,
       "step": 40035
     },
     {
-      "epoch": 1.41,
-      "learning_rate": 4.214573916680338e-05,
-      "loss": 0.2788,
+      "epoch": 1.4430388870868922,
+      "grad_norm": 0.15663708746433258,
+      "learning_rate": 4.1738180487011214e-05,
+      "loss": 0.4162,
       "step": 40040
     },
     {
-      "epoch": 1.41,
-      "learning_rate": 4.2143665863832896e-05,
-      "loss": 0.2832,
+      "epoch": 1.443219086748117,
+      "grad_norm": 0.19377632439136505,
+      "learning_rate": 4.173601281972623e-05,
+      "loss": 0.4313,
       "step": 40045
     },
     {
-      "epoch": 1.41,
-      "learning_rate": 4.214159233826322e-05,
-      "loss": 0.2754,
+      "epoch": 1.4433992864093415,
+      "grad_norm": 0.22673170268535614,
+      "learning_rate": 4.173384492441632e-05,
+      "loss": 0.4314,
       "step": 40050
     },
     {
-      "epoch": 1.41,
-      "learning_rate": 4.2139518590121256e-05,
-      "loss": 0.2872,
+      "epoch": 1.4435794860705662,
+      "grad_norm": 0.15245167911052704,
+      "learning_rate": 4.173167680111101e-05,
+      "loss": 0.3985,
       "step": 40055
     },
     {
-      "epoch": 1.41,
-      "learning_rate": 4.213744461943395e-05,
-      "loss": 0.2831,
+      "epoch": 1.443759685731791,
+      "grad_norm": 0.1588246375322342,
+      "learning_rate": 4.1729508449839834e-05,
+      "loss": 0.42,
       "step": 40060
     },
     {
-      "epoch": 1.41,
-      "learning_rate": 4.213537042622821e-05,
-      "loss": 0.2739,
+      "epoch": 1.4439398853930154,
+      "grad_norm": 0.20271897315979004,
+      "learning_rate": 4.1727339870632345e-05,
+      "loss": 0.4485,
       "step": 40065
     },
     {
-      "epoch": 1.41,
-      "learning_rate": 4.213329601053098e-05,
-      "loss": 0.3059,
+      "epoch": 1.4441200850542402,
+      "grad_norm": 0.1787983775138855,
+      "learning_rate": 4.17251710635181e-05,
+      "loss": 0.3992,
       "step": 40070
     },
     {
-      "epoch": 1.41,
-      "learning_rate": 4.213122137236919e-05,
-      "loss": 0.2987,
+      "epoch": 1.4443002847154647,
+      "grad_norm": 0.17043855786323547,
+      "learning_rate": 4.1723002028526625e-05,
+      "loss": 0.4319,
       "step": 40075
     },
     {
-      "epoch": 1.41,
-      "learning_rate": 4.212914651176978e-05,
-      "loss": 0.2718,
+      "epoch": 1.4444804843766894,
+      "grad_norm": 0.1581939458847046,
+      "learning_rate": 4.172083276568749e-05,
+      "loss": 0.4217,
       "step": 40080
     },
     {
-      "epoch": 1.41,
-      "learning_rate": 4.212707142875969e-05,
-      "loss": 0.2723,
+      "epoch": 1.4446606840379141,
+      "grad_norm": 0.19269004464149475,
+      "learning_rate": 4.1718663275030246e-05,
+      "loss": 0.3794,
       "step": 40085
     },
     {
-      "epoch": 1.41,
-      "learning_rate": 4.212499612336587e-05,
-      "loss": 0.2886,
+      "epoch": 1.4448408836991387,
+      "grad_norm": 0.1823306381702423,
+      "learning_rate": 4.1716493556584455e-05,
+      "loss": 0.3788,
       "step": 40090
     },
     {
-      "epoch": 1.41,
-      "learning_rate": 4.212292059561526e-05,
-      "loss": 0.3042,
+      "epoch": 1.4450210833603632,
+      "grad_norm": 0.22795365750789642,
+      "learning_rate": 4.171432361037968e-05,
+      "loss": 0.4046,
       "step": 40095
     },
     {
-      "epoch": 1.41,
-      "learning_rate": 4.2120844845534816e-05,
-      "loss": 0.3038,
+      "epoch": 1.445201283021588,
+      "grad_norm": 0.1868722289800644,
+      "learning_rate": 4.1712153436445464e-05,
+      "loss": 0.4358,
       "step": 40100
     },
     {
-      "epoch": 1.41,
-      "learning_rate": 4.211876887315147e-05,
-      "loss": 0.293,
+      "epoch": 1.4453814826828126,
+      "grad_norm": 0.2065182775259018,
+      "learning_rate": 4.1709983034811406e-05,
+      "loss": 0.4051,
       "step": 40105
     },
     {
-      "epoch": 1.41,
-      "learning_rate": 4.211669267849221e-05,
-      "loss": 0.2877,
+      "epoch": 1.4455616823440371,
+      "grad_norm": 0.14879098534584045,
+      "learning_rate": 4.170781240550706e-05,
+      "loss": 0.3798,
       "step": 40110
     },
     {
-      "epoch": 1.41,
-      "learning_rate": 4.211461626158396e-05,
-      "loss": 0.2785,
+      "epoch": 1.4457418820052619,
+      "grad_norm": 0.15872523188591003,
+      "learning_rate": 4.170564154856201e-05,
+      "loss": 0.406,
       "step": 40115
     },
     {
-      "epoch": 1.41,
-      "learning_rate": 4.211253962245371e-05,
-      "loss": 0.297,
+      "epoch": 1.4459220816664864,
+      "grad_norm": 0.15561115741729736,
+      "learning_rate": 4.170347046400583e-05,
+      "loss": 0.3787,
       "step": 40120
     },
     {
-      "epoch": 1.41,
-      "learning_rate": 4.21104627611284e-05,
-      "loss": 0.2807,
+      "epoch": 1.4461022813277111,
+      "grad_norm": 0.22237730026245117,
+      "learning_rate": 4.170129915186809e-05,
+      "loss": 0.4211,
       "step": 40125
     },
     {
-      "epoch": 1.41,
-      "learning_rate": 4.210838567763501e-05,
-      "loss": 0.3068,
+      "epoch": 1.4462824809889359,
+      "grad_norm": 0.19073696434497833,
+      "learning_rate": 4.169912761217839e-05,
+      "loss": 0.4023,
       "step": 40130
     },
     {
-      "epoch": 1.41,
-      "learning_rate": 4.21063083720005e-05,
-      "loss": 0.2605,
+      "epoch": 1.4464626806501604,
+      "grad_norm": 0.18226531147956848,
+      "learning_rate": 4.16969558449663e-05,
+      "loss": 0.4144,
       "step": 40135
     },
     {
-      "epoch": 1.41,
-      "learning_rate": 4.210423084425186e-05,
-      "loss": 0.2947,
+      "epoch": 1.4466428803113849,
+      "grad_norm": 0.19188474118709564,
+      "learning_rate": 4.169478385026142e-05,
+      "loss": 0.4288,
       "step": 40140
     },
     {
-      "epoch": 1.41,
-      "learning_rate": 4.2102153094416055e-05,
-      "loss": 0.3167,
+      "epoch": 1.4468230799726096,
+      "grad_norm": 0.2194652110338211,
+      "learning_rate": 4.169261162809335e-05,
+      "loss": 0.4719,
       "step": 40145
     },
     {
-      "epoch": 1.41,
-      "learning_rate": 4.2100075122520055e-05,
-      "loss": 0.2757,
+      "epoch": 1.4470032796338343,
+      "grad_norm": 0.192906454205513,
+      "learning_rate": 4.169043917849168e-05,
+      "loss": 0.4207,
       "step": 40150
     },
     {
-      "epoch": 1.41,
-      "learning_rate": 4.2097996928590854e-05,
-      "loss": 0.2864,
+      "epoch": 1.4471834792950589,
+      "grad_norm": 0.17282317578792572,
+      "learning_rate": 4.1688266501486004e-05,
+      "loss": 0.388,
       "step": 40155
     },
     {
-      "epoch": 1.41,
-      "learning_rate": 4.209591851265543e-05,
-      "loss": 0.308,
+      "epoch": 1.4473636789562836,
+      "grad_norm": 0.2374262660741806,
+      "learning_rate": 4.168609359710592e-05,
+      "loss": 0.4231,
       "step": 40160
     },
     {
-      "epoch": 1.41,
-      "learning_rate": 4.209383987474077e-05,
-      "loss": 0.3065,
+      "epoch": 1.447543878617508,
+      "grad_norm": 0.14937429130077362,
+      "learning_rate": 4.1683920465381054e-05,
+      "loss": 0.4245,
       "step": 40165
     },
     {
-      "epoch": 1.41,
-      "learning_rate": 4.209176101487387e-05,
-      "loss": 0.3025,
+      "epoch": 1.4477240782787328,
+      "grad_norm": 0.15380804240703583,
+      "learning_rate": 4.1681747106340995e-05,
+      "loss": 0.4201,
       "step": 40170
     },
     {
-      "epoch": 1.41,
-      "learning_rate": 4.2089681933081716e-05,
-      "loss": 0.2801,
+      "epoch": 1.4479042779399576,
+      "grad_norm": 0.22721770405769348,
+      "learning_rate": 4.167957352001537e-05,
+      "loss": 0.3973,
       "step": 40175
     },
     {
-      "epoch": 1.41,
-      "learning_rate": 4.20876026293913e-05,
-      "loss": 0.308,
+      "epoch": 1.448084477601182,
+      "grad_norm": 0.21499106287956238,
+      "learning_rate": 4.167739970643377e-05,
+      "loss": 0.4453,
       "step": 40180
     },
     {
-      "epoch": 1.41,
-      "learning_rate": 4.208552310382963e-05,
-      "loss": 0.3071,
+      "epoch": 1.4482646772624068,
+      "grad_norm": 0.19283875823020935,
+      "learning_rate": 4.167522566562584e-05,
+      "loss": 0.4062,
       "step": 40185
     },
     {
-      "epoch": 1.41,
-      "learning_rate": 4.2083443356423694e-05,
-      "loss": 0.2772,
+      "epoch": 1.4484448769236313,
+      "grad_norm": 0.1777193248271942,
+      "learning_rate": 4.167305139762119e-05,
+      "loss": 0.4055,
       "step": 40190
     },
     {
-      "epoch": 1.41,
-      "learning_rate": 4.2081363387200504e-05,
-      "loss": 0.2632,
+      "epoch": 1.448625076584856,
+      "grad_norm": 0.2283918261528015,
+      "learning_rate": 4.167087690244943e-05,
+      "loss": 0.4338,
       "step": 40195
     },
     {
-      "epoch": 1.41,
-      "learning_rate": 4.2079283196187077e-05,
-      "loss": 0.2998,
+      "epoch": 1.4488052762460806,
+      "grad_norm": 0.17080770432949066,
+      "learning_rate": 4.1668702180140206e-05,
+      "loss": 0.4003,
       "step": 40200
     },
     {
-      "epoch": 1.41,
-      "learning_rate": 4.207720278341041e-05,
-      "loss": 0.2964,
+      "epoch": 1.4489854759073053,
+      "grad_norm": 0.17411410808563232,
+      "learning_rate": 4.166652723072314e-05,
+      "loss": 0.4115,
       "step": 40205
     },
     {
-      "epoch": 1.41,
-      "learning_rate": 4.2075122148897516e-05,
-      "loss": 0.3021,
+      "epoch": 1.4491656755685298,
+      "grad_norm": 0.18166224658489227,
+      "learning_rate": 4.166435205422787e-05,
+      "loss": 0.4184,
       "step": 40210
     },
     {
-      "epoch": 1.41,
-      "learning_rate": 4.207304129267542e-05,
-      "loss": 0.295,
+      "epoch": 1.4493458752297546,
+      "grad_norm": 0.17343877255916595,
+      "learning_rate": 4.166217665068403e-05,
+      "loss": 0.4373,
       "step": 40215
     },
     {
-      "epoch": 1.42,
-      "learning_rate": 4.207096021477113e-05,
-      "loss": 0.2857,
+      "epoch": 1.4495260748909793,
+      "grad_norm": 0.1936635822057724,
+      "learning_rate": 4.166000102012126e-05,
+      "loss": 0.387,
       "step": 40220
     },
     {
-      "epoch": 1.42,
-      "learning_rate": 4.2068878915211665e-05,
-      "loss": 0.2762,
+      "epoch": 1.4497062745522038,
+      "grad_norm": 0.1682083159685135,
+      "learning_rate": 4.16578251625692e-05,
+      "loss": 0.422,
       "step": 40225
     },
     {
-      "epoch": 1.42,
-      "learning_rate": 4.206679739402406e-05,
-      "loss": 0.2891,
+      "epoch": 1.4498864742134285,
+      "grad_norm": 0.20389603078365326,
+      "learning_rate": 4.16556490780575e-05,
+      "loss": 0.4434,
       "step": 40230
     },
     {
-      "epoch": 1.42,
-      "learning_rate": 4.2064715651235344e-05,
-      "loss": 0.2594,
+      "epoch": 1.450066673874653,
+      "grad_norm": 0.16445006430149078,
+      "learning_rate": 4.16534727666158e-05,
+      "loss": 0.4393,
       "step": 40235
     },
     {
-      "epoch": 1.42,
-      "learning_rate": 4.2062633686872545e-05,
-      "loss": 0.2997,
+      "epoch": 1.4502468735358778,
+      "grad_norm": 0.1689777970314026,
+      "learning_rate": 4.165129622827376e-05,
+      "loss": 0.3975,
       "step": 40240
     },
     {
-      "epoch": 1.42,
-      "learning_rate": 4.2060551500962676e-05,
-      "loss": 0.2933,
+      "epoch": 1.4504270731971025,
+      "grad_norm": 0.17561562359333038,
+      "learning_rate": 4.164911946306104e-05,
+      "loss": 0.4218,
       "step": 40245
     },
     {
-      "epoch": 1.42,
-      "learning_rate": 4.2058469093532806e-05,
-      "loss": 0.2704,
+      "epoch": 1.450607272858327,
+      "grad_norm": 0.18199895322322845,
+      "learning_rate": 4.164694247100728e-05,
+      "loss": 0.4344,
       "step": 40250
     },
     {
-      "epoch": 1.42,
-      "learning_rate": 4.205638646460994e-05,
-      "loss": 0.2835,
+      "epoch": 1.4507874725195515,
+      "grad_norm": 0.17951489984989166,
+      "learning_rate": 4.164476525214216e-05,
+      "loss": 0.4282,
       "step": 40255
     },
     {
-      "epoch": 1.42,
-      "learning_rate": 4.2054303614221154e-05,
-      "loss": 0.2882,
+      "epoch": 1.4509676721807763,
+      "grad_norm": 0.1799135059118271,
+      "learning_rate": 4.1642587806495324e-05,
+      "loss": 0.4252,
       "step": 40260
     },
     {
-      "epoch": 1.42,
-      "learning_rate": 4.205222054239346e-05,
-      "loss": 0.2911,
+      "epoch": 1.451147871842001,
+      "grad_norm": 0.172772616147995,
+      "learning_rate": 4.1640410134096465e-05,
+      "loss": 0.413,
       "step": 40265
     },
     {
-      "epoch": 1.42,
-      "learning_rate": 4.205013724915393e-05,
-      "loss": 0.2964,
+      "epoch": 1.4513280715032255,
+      "grad_norm": 0.15084902942180634,
+      "learning_rate": 4.1638232234975225e-05,
+      "loss": 0.4336,
       "step": 40270
     },
     {
-      "epoch": 1.42,
-      "learning_rate": 4.2048053734529604e-05,
-      "loss": 0.2897,
+      "epoch": 1.4515082711644502,
+      "grad_norm": 0.17858850955963135,
+      "learning_rate": 4.163605410916131e-05,
+      "loss": 0.443,
       "step": 40275
     },
     {
-      "epoch": 1.42,
-      "learning_rate": 4.204596999854753e-05,
-      "loss": 0.2759,
+      "epoch": 1.4516884708256748,
+      "grad_norm": 0.1724681556224823,
+      "learning_rate": 4.163387575668437e-05,
+      "loss": 0.3995,
       "step": 40280
     },
     {
-      "epoch": 1.42,
-      "learning_rate": 4.204388604123477e-05,
-      "loss": 0.2827,
+      "epoch": 1.4518686704868995,
+      "grad_norm": 0.15848155319690704,
+      "learning_rate": 4.163169717757409e-05,
+      "loss": 0.3968,
       "step": 40285
     },
     {
-      "epoch": 1.42,
-      "learning_rate": 4.204180186261839e-05,
-      "loss": 0.2804,
+      "epoch": 1.4520488701481242,
+      "grad_norm": 0.17804880440235138,
+      "learning_rate": 4.162951837186016e-05,
+      "loss": 0.4272,
       "step": 40290
     },
     {
-      "epoch": 1.42,
-      "learning_rate": 4.2039717462725435e-05,
-      "loss": 0.2762,
+      "epoch": 1.4522290698093487,
+      "grad_norm": 0.19191712141036987,
+      "learning_rate": 4.1627339339572256e-05,
+      "loss": 0.4096,
       "step": 40295
     },
     {
-      "epoch": 1.42,
-      "learning_rate": 4.203763284158299e-05,
-      "loss": 0.3067,
+      "epoch": 1.4524092694705735,
+      "grad_norm": 0.19533273577690125,
+      "learning_rate": 4.1625160080740075e-05,
+      "loss": 0.4129,
       "step": 40300
     },
     {
-      "epoch": 1.42,
-      "learning_rate": 4.203554799921809e-05,
-      "loss": 0.3022,
+      "epoch": 1.452589469131798,
+      "grad_norm": 0.28732794523239136,
+      "learning_rate": 4.162298059539331e-05,
+      "loss": 0.4749,
       "step": 40305
     },
     {
-      "epoch": 1.42,
-      "learning_rate": 4.203346293565784e-05,
-      "loss": 0.3145,
+      "epoch": 1.4527696687930227,
+      "grad_norm": 0.16327139735221863,
+      "learning_rate": 4.1620800883561656e-05,
+      "loss": 0.4015,
       "step": 40310
     },
     {
-      "epoch": 1.42,
-      "learning_rate": 4.203137765092929e-05,
-      "loss": 0.2803,
+      "epoch": 1.4529498684542472,
+      "grad_norm": 0.1888490468263626,
+      "learning_rate": 4.16186209452748e-05,
+      "loss": 0.4243,
       "step": 40315
     },
     {
-      "epoch": 1.42,
-      "learning_rate": 4.202929214505954e-05,
-      "loss": 0.306,
+      "epoch": 1.453130068115472,
+      "grad_norm": 0.20011357963085175,
+      "learning_rate": 4.161644078056246e-05,
+      "loss": 0.4201,
       "step": 40320
     },
     {
-      "epoch": 1.42,
-      "learning_rate": 4.202720641807565e-05,
-      "loss": 0.2882,
+      "epoch": 1.4533102677766965,
+      "grad_norm": 0.1774764508008957,
+      "learning_rate": 4.161426038945432e-05,
+      "loss": 0.3993,
       "step": 40325
     },
     {
-      "epoch": 1.42,
-      "learning_rate": 4.20251204700047e-05,
-      "loss": 0.2888,
+      "epoch": 1.4534904674379212,
+      "grad_norm": 0.18601520359516144,
+      "learning_rate": 4.1612079771980106e-05,
+      "loss": 0.382,
       "step": 40330
     },
     {
-      "epoch": 1.42,
-      "learning_rate": 4.202303430087379e-05,
-      "loss": 0.286,
+      "epoch": 1.453670667099146,
+      "grad_norm": 0.18623565137386322,
+      "learning_rate": 4.160989892816952e-05,
+      "loss": 0.4032,
       "step": 40335
     },
     {
-      "epoch": 1.42,
-      "learning_rate": 4.202094791070998e-05,
-      "loss": 0.2982,
+      "epoch": 1.4538508667603705,
+      "grad_norm": 0.16086627542972565,
+      "learning_rate": 4.160771785805228e-05,
+      "loss": 0.4054,
       "step": 40340
     },
     {
-      "epoch": 1.42,
-      "learning_rate": 4.201886129954039e-05,
-      "loss": 0.2865,
+      "epoch": 1.4540310664215952,
+      "grad_norm": 0.1630929857492447,
+      "learning_rate": 4.16055365616581e-05,
+      "loss": 0.4016,
       "step": 40345
     },
     {
-      "epoch": 1.42,
-      "learning_rate": 4.2016774467392106e-05,
-      "loss": 0.3012,
+      "epoch": 1.4542112660828197,
+      "grad_norm": 0.16197270154953003,
+      "learning_rate": 4.160335503901669e-05,
+      "loss": 0.4461,
       "step": 40350
     },
     {
-      "epoch": 1.42,
-      "learning_rate": 4.201468741429222e-05,
-      "loss": 0.2862,
+      "epoch": 1.4543914657440444,
+      "grad_norm": 0.160113126039505,
+      "learning_rate": 4.1601173290157794e-05,
+      "loss": 0.429,
       "step": 40355
     },
     {
-      "epoch": 1.42,
-      "learning_rate": 4.201260014026782e-05,
-      "loss": 0.2935,
+      "epoch": 1.4545716654052692,
+      "grad_norm": 0.20948779582977295,
+      "learning_rate": 4.159899131511111e-05,
+      "loss": 0.4058,
       "step": 40360
     },
     {
-      "epoch": 1.42,
-      "learning_rate": 4.2010512645346034e-05,
-      "loss": 0.2814,
+      "epoch": 1.4547518650664937,
+      "grad_norm": 0.19424784183502197,
+      "learning_rate": 4.15968091139064e-05,
+      "loss": 0.4113,
       "step": 40365
     },
     {
-      "epoch": 1.42,
-      "learning_rate": 4.2008424929553934e-05,
-      "loss": 0.3095,
+      "epoch": 1.4549320647277182,
+      "grad_norm": 0.19249606132507324,
+      "learning_rate": 4.159462668657337e-05,
+      "loss": 0.4407,
       "step": 40370
     },
     {
-      "epoch": 1.42,
-      "learning_rate": 4.200633699291866e-05,
-      "loss": 0.2908,
+      "epoch": 1.455112264388943,
+      "grad_norm": 0.13863718509674072,
+      "learning_rate": 4.159244403314176e-05,
+      "loss": 0.4143,
       "step": 40375
     },
     {
-      "epoch": 1.42,
-      "learning_rate": 4.20042488354673e-05,
-      "loss": 0.2814,
+      "epoch": 1.4552924640501677,
+      "grad_norm": 0.16577591001987457,
+      "learning_rate": 4.159026115364132e-05,
+      "loss": 0.4468,
       "step": 40380
     },
     {
-      "epoch": 1.42,
-      "learning_rate": 4.200216045722698e-05,
-      "loss": 0.2764,
+      "epoch": 1.4554726637113922,
+      "grad_norm": 0.16115480661392212,
+      "learning_rate": 4.1588078048101784e-05,
+      "loss": 0.4213,
       "step": 40385
     },
     {
-      "epoch": 1.42,
-      "learning_rate": 4.200007185822481e-05,
-      "loss": 0.2903,
+      "epoch": 1.455652863372617,
+      "grad_norm": 0.17219962179660797,
+      "learning_rate": 4.15858947165529e-05,
+      "loss": 0.4411,
       "step": 40390
     },
     {
-      "epoch": 1.42,
-      "learning_rate": 4.199798303848791e-05,
-      "loss": 0.2933,
+      "epoch": 1.4558330630338414,
+      "grad_norm": 0.17890863120555878,
+      "learning_rate": 4.15837111590244e-05,
+      "loss": 0.4052,
       "step": 40395
     },
     {
-      "epoch": 1.42,
-      "learning_rate": 4.19958939980434e-05,
-      "loss": 0.2782,
+      "epoch": 1.4560132626950661,
+      "grad_norm": 0.19050833582878113,
+      "learning_rate": 4.158152737554606e-05,
+      "loss": 0.4257,
       "step": 40400
     },
     {
-      "epoch": 1.42,
-      "learning_rate": 4.199380473691841e-05,
-      "loss": 0.2777,
+      "epoch": 1.4561934623562909,
+      "grad_norm": 0.1742664873600006,
+      "learning_rate": 4.1579343366147604e-05,
+      "loss": 0.4023,
       "step": 40405
     },
     {
-      "epoch": 1.42,
-      "learning_rate": 4.199171525514006e-05,
-      "loss": 0.2911,
+      "epoch": 1.4563736620175154,
+      "grad_norm": 0.19294539093971252,
+      "learning_rate": 4.157715913085881e-05,
+      "loss": 0.4072,
       "step": 40410
     },
     {
-      "epoch": 1.42,
-      "learning_rate": 4.198962555273549e-05,
-      "loss": 0.2774,
+      "epoch": 1.4565538616787401,
+      "grad_norm": 0.1636057049036026,
+      "learning_rate": 4.1574974669709435e-05,
+      "loss": 0.4023,
       "step": 40415
     },
     {
-      "epoch": 1.42,
-      "learning_rate": 4.1987535629731835e-05,
-      "loss": 0.3019,
+      "epoch": 1.4567340613399646,
+      "grad_norm": 0.15918225049972534,
+      "learning_rate": 4.1572789982729244e-05,
+      "loss": 0.3778,
       "step": 40420
     },
     {
-      "epoch": 1.42,
-      "learning_rate": 4.1985445486156216e-05,
-      "loss": 0.2958,
+      "epoch": 1.4569142610011894,
+      "grad_norm": 0.17120924592018127,
+      "learning_rate": 4.1570605069947986e-05,
+      "loss": 0.4285,
       "step": 40425
     },
     {
-      "epoch": 1.42,
-      "learning_rate": 4.198335512203578e-05,
-      "loss": 0.2622,
+      "epoch": 1.4570944606624139,
+      "grad_norm": 0.16076339781284332,
+      "learning_rate": 4.1568419931395456e-05,
+      "loss": 0.4201,
       "step": 40430
     },
     {
-      "epoch": 1.42,
-      "learning_rate": 4.1981264537397674e-05,
-      "loss": 0.2592,
+      "epoch": 1.4572746603236386,
+      "grad_norm": 0.14771609008312225,
+      "learning_rate": 4.15662345671014e-05,
+      "loss": 0.3823,
       "step": 40435
     },
     {
-      "epoch": 1.42,
-      "learning_rate": 4.197917373226904e-05,
-      "loss": 0.2674,
+      "epoch": 1.4574548599848631,
+      "grad_norm": 0.19097311794757843,
+      "learning_rate": 4.156404897709562e-05,
+      "loss": 0.4274,
       "step": 40440
     },
     {
-      "epoch": 1.42,
-      "learning_rate": 4.1977082706677026e-05,
-      "loss": 0.2983,
+      "epoch": 1.4576350596460879,
+      "grad_norm": 0.16327013075351715,
+      "learning_rate": 4.1561863161407866e-05,
+      "loss": 0.4176,
       "step": 40445
     },
     {
-      "epoch": 1.42,
-      "learning_rate": 4.197499146064877e-05,
-      "loss": 0.2961,
+      "epoch": 1.4578152593073126,
+      "grad_norm": 0.15996558964252472,
+      "learning_rate": 4.1559677120067935e-05,
+      "loss": 0.4001,
       "step": 40450
     },
     {
-      "epoch": 1.42,
-      "learning_rate": 4.197289999421145e-05,
-      "loss": 0.2746,
+      "epoch": 1.457995458968537,
+      "grad_norm": 0.1909141093492508,
+      "learning_rate": 4.1557490853105614e-05,
+      "loss": 0.3835,
       "step": 40455
     },
     {
-      "epoch": 1.42,
-      "learning_rate": 4.1970808307392204e-05,
-      "loss": 0.2803,
+      "epoch": 1.4581756586297618,
+      "grad_norm": 0.21793286502361298,
+      "learning_rate": 4.155530436055068e-05,
+      "loss": 0.3851,
       "step": 40460
     },
     {
-      "epoch": 1.42,
-      "learning_rate": 4.1968716400218196e-05,
-      "loss": 0.2771,
+      "epoch": 1.4583558582909864,
+      "grad_norm": 0.14985129237174988,
+      "learning_rate": 4.155311764243294e-05,
+      "loss": 0.385,
       "step": 40465
     },
     {
-      "epoch": 1.42,
-      "learning_rate": 4.1966624272716594e-05,
-      "loss": 0.2805,
+      "epoch": 1.458536057952211,
+      "grad_norm": 0.16198906302452087,
+      "learning_rate": 4.155093069878216e-05,
+      "loss": 0.3893,
       "step": 40470
     },
     {
-      "epoch": 1.42,
-      "learning_rate": 4.196453192491455e-05,
-      "loss": 0.2842,
+      "epoch": 1.4587162576134358,
+      "grad_norm": 0.20363600552082062,
+      "learning_rate": 4.154874352962816e-05,
+      "loss": 0.4146,
       "step": 40475
     },
     {
-      "epoch": 1.42,
-      "learning_rate": 4.1962439356839236e-05,
-      "loss": 0.2983,
+      "epoch": 1.4588964572746603,
+      "grad_norm": 0.1638229936361313,
+      "learning_rate": 4.154655613500075e-05,
+      "loss": 0.4149,
       "step": 40480
     },
     {
-      "epoch": 1.42,
-      "learning_rate": 4.196034656851783e-05,
-      "loss": 0.2932,
+      "epoch": 1.4590766569358848,
+      "grad_norm": 0.18865369260311127,
+      "learning_rate": 4.15443685149297e-05,
+      "loss": 0.3961,
       "step": 40485
     },
     {
-      "epoch": 1.42,
-      "learning_rate": 4.1958253559977506e-05,
-      "loss": 0.2799,
+      "epoch": 1.4592568565971096,
+      "grad_norm": 0.17428448796272278,
+      "learning_rate": 4.154218066944483e-05,
+      "loss": 0.4314,
       "step": 40490
     },
     {
-      "epoch": 1.42,
-      "learning_rate": 4.195616033124543e-05,
-      "loss": 0.2829,
+      "epoch": 1.4594370562583343,
+      "grad_norm": 0.18061710894107819,
+      "learning_rate": 4.1539992598575954e-05,
+      "loss": 0.4091,
       "step": 40495
     },
     {
-      "epoch": 1.42,
-      "learning_rate": 4.195406688234879e-05,
-      "loss": 0.3083,
+      "epoch": 1.4596172559195588,
+      "grad_norm": 0.14302043616771698,
+      "learning_rate": 4.153780430235289e-05,
+      "loss": 0.3872,
       "step": 40500
     },
     {
-      "epoch": 1.42,
-      "eval_loss": 0.2803879678249359,
-      "eval_runtime": 10.5501,
-      "eval_samples_per_second": 9.479,
-      "eval_steps_per_second": 9.479,
+      "epoch": 1.4596172559195588,
+      "eval_loss": 0.44571349024772644,
+      "eval_runtime": 3.5421,
+      "eval_samples_per_second": 28.232,
+      "eval_steps_per_second": 7.058,
       "step": 40500
     },
     {
-      "epoch": 1.43,
-      "learning_rate": 4.195197321331476e-05,
-      "loss": 0.3007,
+      "epoch": 1.4597974555807836,
+      "grad_norm": 0.1947942078113556,
+      "learning_rate": 4.153561578080543e-05,
+      "loss": 0.413,
       "step": 40505
     },
     {
-      "epoch": 1.43,
-      "learning_rate": 4.194987932417054e-05,
-      "loss": 0.3135,
+      "epoch": 1.459977655242008,
+      "grad_norm": 0.1369856297969818,
+      "learning_rate": 4.153342703396341e-05,
+      "loss": 0.3949,
       "step": 40510
     },
     {
-      "epoch": 1.43,
-      "learning_rate": 4.194778521494329e-05,
-      "loss": 0.2749,
+      "epoch": 1.4601578549032328,
+      "grad_norm": 0.17609436810016632,
+      "learning_rate": 4.153123806185666e-05,
+      "loss": 0.4322,
       "step": 40515
     },
     {
-      "epoch": 1.43,
-      "learning_rate": 4.194569088566023e-05,
-      "loss": 0.2535,
+      "epoch": 1.4603380545644575,
+      "grad_norm": 0.2042810618877411,
+      "learning_rate": 4.152904886451498e-05,
+      "loss": 0.4339,
       "step": 40520
     },
     {
-      "epoch": 1.43,
-      "learning_rate": 4.194359633634854e-05,
-      "loss": 0.2984,
+      "epoch": 1.460518254225682,
+      "grad_norm": 0.22225399315357208,
+      "learning_rate": 4.152685944196821e-05,
+      "loss": 0.4344,
       "step": 40525
     },
     {
-      "epoch": 1.43,
-      "learning_rate": 4.194150156703542e-05,
-      "loss": 0.2547,
+      "epoch": 1.4606984538869066,
+      "grad_norm": 0.47528132796287537,
+      "learning_rate": 4.152466979424619e-05,
+      "loss": 0.4446,
       "step": 40530
     },
     {
-      "epoch": 1.43,
-      "learning_rate": 4.1939406577748065e-05,
-      "loss": 0.3001,
+      "epoch": 1.4608786535481313,
+      "grad_norm": 0.16672192513942719,
+      "learning_rate": 4.1522479921378733e-05,
+      "loss": 0.4329,
       "step": 40535
     },
     {
-      "epoch": 1.43,
-      "learning_rate": 4.193731136851368e-05,
-      "loss": 0.2765,
+      "epoch": 1.461058853209356,
+      "grad_norm": 0.19320343434810638,
+      "learning_rate": 4.1520289823395686e-05,
+      "loss": 0.4102,
       "step": 40540
     },
     {
-      "epoch": 1.43,
-      "learning_rate": 4.193521593935946e-05,
-      "loss": 0.2817,
+      "epoch": 1.4612390528705805,
+      "grad_norm": 0.17399749159812927,
+      "learning_rate": 4.151809950032689e-05,
+      "loss": 0.4194,
       "step": 40545
     },
     {
-      "epoch": 1.43,
-      "learning_rate": 4.1933120290312636e-05,
-      "loss": 0.2948,
+      "epoch": 1.4614192525318053,
+      "grad_norm": 0.1759127378463745,
+      "learning_rate": 4.1515908952202186e-05,
+      "loss": 0.3828,
       "step": 40550
     },
     {
-      "epoch": 1.43,
-      "learning_rate": 4.19310244214004e-05,
-      "loss": 0.3034,
+      "epoch": 1.4615994521930298,
+      "grad_norm": 0.17152760922908783,
+      "learning_rate": 4.151371817905142e-05,
+      "loss": 0.4291,
       "step": 40555
     },
     {
-      "epoch": 1.43,
-      "learning_rate": 4.1928928332649965e-05,
-      "loss": 0.3064,
+      "epoch": 1.4617796518542545,
+      "grad_norm": 0.18539777398109436,
+      "learning_rate": 4.151152718090445e-05,
+      "loss": 0.376,
       "step": 40560
     },
     {
-      "epoch": 1.43,
-      "learning_rate": 4.192683202408855e-05,
-      "loss": 0.2673,
+      "epoch": 1.4619598515154792,
+      "grad_norm": 0.1786716729402542,
+      "learning_rate": 4.1509335957791106e-05,
+      "loss": 0.3922,
       "step": 40565
     },
     {
-      "epoch": 1.43,
-      "learning_rate": 4.1924735495743384e-05,
-      "loss": 0.2798,
+      "epoch": 1.4621400511767038,
+      "grad_norm": 0.15850621461868286,
+      "learning_rate": 4.150714450974126e-05,
+      "loss": 0.4005,
       "step": 40570
     },
     {
-      "epoch": 1.43,
-      "learning_rate": 4.1922638747641676e-05,
-      "loss": 0.2704,
+      "epoch": 1.4623202508379285,
+      "grad_norm": 0.19711154699325562,
+      "learning_rate": 4.150495283678477e-05,
+      "loss": 0.4157,
       "step": 40575
     },
     {
-      "epoch": 1.43,
-      "learning_rate": 4.192054177981065e-05,
-      "loss": 0.2958,
+      "epoch": 1.462500450499153,
+      "grad_norm": 0.2046595960855484,
+      "learning_rate": 4.150276093895149e-05,
+      "loss": 0.4197,
       "step": 40580
     },
     {
-      "epoch": 1.43,
-      "learning_rate": 4.191844459227756e-05,
-      "loss": 0.2943,
+      "epoch": 1.4626806501603777,
+      "grad_norm": 0.23600243031978607,
+      "learning_rate": 4.1500568816271285e-05,
+      "loss": 0.4453,
       "step": 40585
     },
     {
-      "epoch": 1.43,
-      "learning_rate": 4.1916347185069596e-05,
-      "loss": 0.2997,
+      "epoch": 1.4628608498216025,
+      "grad_norm": 0.18028078973293304,
+      "learning_rate": 4.149837646877402e-05,
+      "loss": 0.449,
       "step": 40590
     },
     {
-      "epoch": 1.43,
-      "learning_rate": 4.1914249558214025e-05,
-      "loss": 0.3073,
+      "epoch": 1.463041049482827,
+      "grad_norm": 0.21281985938549042,
+      "learning_rate": 4.149618389648958e-05,
+      "loss": 0.434,
       "step": 40595
     },
     {
-      "epoch": 1.43,
-      "learning_rate": 4.191215171173807e-05,
-      "loss": 0.3079,
+      "epoch": 1.4632212491440515,
+      "grad_norm": 0.18634112179279327,
+      "learning_rate": 4.149399109944783e-05,
+      "loss": 0.3694,
       "step": 40600
     },
     {
-      "epoch": 1.43,
-      "learning_rate": 4.1910053645668965e-05,
-      "loss": 0.2856,
+      "epoch": 1.4634014488052762,
+      "grad_norm": 0.2347174882888794,
+      "learning_rate": 4.1491798077678636e-05,
+      "loss": 0.4023,
       "step": 40605
     },
     {
-      "epoch": 1.43,
-      "learning_rate": 4.190795536003396e-05,
-      "loss": 0.2899,
+      "epoch": 1.463581648466501,
+      "grad_norm": 0.2317020446062088,
+      "learning_rate": 4.148960483121189e-05,
+      "loss": 0.4539,
       "step": 40610
     },
     {
-      "epoch": 1.43,
-      "learning_rate": 4.190585685486029e-05,
-      "loss": 0.2806,
+      "epoch": 1.4637618481277255,
+      "grad_norm": 0.16629308462142944,
+      "learning_rate": 4.148741136007747e-05,
+      "loss": 0.4403,
       "step": 40615
     },
     {
-      "epoch": 1.43,
-      "learning_rate": 4.190375813017522e-05,
-      "loss": 0.2648,
+      "epoch": 1.4639420477889502,
+      "grad_norm": 0.19857972860336304,
+      "learning_rate": 4.148521766430527e-05,
+      "loss": 0.3936,
       "step": 40620
     },
     {
-      "epoch": 1.43,
-      "learning_rate": 4.1901659186005996e-05,
-      "loss": 0.2655,
+      "epoch": 1.4641222474501747,
+      "grad_norm": 0.18918615579605103,
+      "learning_rate": 4.148302374392516e-05,
+      "loss": 0.4059,
       "step": 40625
     },
     {
-      "epoch": 1.43,
-      "learning_rate": 4.1899560022379856e-05,
-      "loss": 0.317,
+      "epoch": 1.4643024471113995,
+      "grad_norm": 0.1877117156982422,
+      "learning_rate": 4.148082959896704e-05,
+      "loss": 0.3964,
       "step": 40630
     },
     {
-      "epoch": 1.43,
-      "learning_rate": 4.189746063932407e-05,
-      "loss": 0.2878,
+      "epoch": 1.4644826467726242,
+      "grad_norm": 0.1804433912038803,
+      "learning_rate": 4.1478635229460814e-05,
+      "loss": 0.4587,
       "step": 40635
     },
     {
-      "epoch": 1.43,
-      "learning_rate": 4.189536103686589e-05,
-      "loss": 0.2867,
+      "epoch": 1.4646628464338487,
+      "grad_norm": 0.16584636270999908,
+      "learning_rate": 4.1476440635436376e-05,
+      "loss": 0.3789,
       "step": 40640
     },
     {
-      "epoch": 1.43,
-      "learning_rate": 4.189326121503258e-05,
-      "loss": 0.2737,
+      "epoch": 1.4648430460950732,
+      "grad_norm": 0.16475588083267212,
+      "learning_rate": 4.1474245816923616e-05,
+      "loss": 0.3847,
       "step": 40645
     },
     {
-      "epoch": 1.43,
-      "learning_rate": 4.1891161173851415e-05,
-      "loss": 0.2758,
+      "epoch": 1.465023245756298,
+      "grad_norm": 0.18502764403820038,
+      "learning_rate": 4.147205077395245e-05,
+      "loss": 0.4038,
       "step": 40650
     },
     {
-      "epoch": 1.43,
-      "learning_rate": 4.188906091334964e-05,
-      "loss": 0.2886,
+      "epoch": 1.4652034454175227,
+      "grad_norm": 0.14497093856334686,
+      "learning_rate": 4.146985550655279e-05,
+      "loss": 0.3932,
       "step": 40655
     },
     {
-      "epoch": 1.43,
-      "learning_rate": 4.188696043355455e-05,
-      "loss": 0.2923,
+      "epoch": 1.4653836450787472,
+      "grad_norm": 0.21168319880962372,
+      "learning_rate": 4.146766001475453e-05,
+      "loss": 0.411,
       "step": 40660
     },
     {
-      "epoch": 1.43,
-      "learning_rate": 4.188485973449341e-05,
-      "loss": 0.2712,
+      "epoch": 1.465563844739972,
+      "grad_norm": 0.17997999489307404,
+      "learning_rate": 4.146546429858759e-05,
+      "loss": 0.3992,
       "step": 40665
     },
     {
-      "epoch": 1.43,
-      "learning_rate": 4.1882758816193484e-05,
-      "loss": 0.3056,
+      "epoch": 1.4657440444011964,
+      "grad_norm": 0.1749783307313919,
+      "learning_rate": 4.146326835808188e-05,
+      "loss": 0.463,
       "step": 40670
     },
     {
-      "epoch": 1.43,
-      "learning_rate": 4.188065767868207e-05,
-      "loss": 0.3079,
+      "epoch": 1.4659242440624212,
+      "grad_norm": 0.19459916651248932,
+      "learning_rate": 4.1461072193267344e-05,
+      "loss": 0.4353,
       "step": 40675
     },
     {
-      "epoch": 1.43,
-      "learning_rate": 4.187855632198643e-05,
-      "loss": 0.279,
+      "epoch": 1.466104443723646,
+      "grad_norm": 0.18047486245632172,
+      "learning_rate": 4.145887580417387e-05,
+      "loss": 0.414,
       "step": 40680
     },
     {
-      "epoch": 1.43,
-      "learning_rate": 4.187645474613387e-05,
-      "loss": 0.2859,
+      "epoch": 1.4662846433848704,
+      "grad_norm": 0.21394000947475433,
+      "learning_rate": 4.145667919083141e-05,
+      "loss": 0.3916,
       "step": 40685
     },
     {
-      "epoch": 1.43,
-      "learning_rate": 4.187435295115166e-05,
-      "loss": 0.2993,
+      "epoch": 1.4664648430460951,
+      "grad_norm": 0.2331513613462448,
+      "learning_rate": 4.1454482353269875e-05,
+      "loss": 0.4231,
       "step": 40690
     },
     {
-      "epoch": 1.43,
-      "learning_rate": 4.18722509370671e-05,
-      "loss": 0.2918,
+      "epoch": 1.4666450427073197,
+      "grad_norm": 0.18889982998371124,
+      "learning_rate": 4.145228529151921e-05,
+      "loss": 0.4509,
       "step": 40695
     },
     {
-      "epoch": 1.43,
-      "learning_rate": 4.187014870390749e-05,
-      "loss": 0.2852,
+      "epoch": 1.4668252423685444,
+      "grad_norm": 0.1711355745792389,
+      "learning_rate": 4.1450088005609335e-05,
+      "loss": 0.4098,
       "step": 40700
     },
     {
-      "epoch": 1.43,
-      "learning_rate": 4.186804625170011e-05,
-      "loss": 0.2632,
+      "epoch": 1.467005442029769,
+      "grad_norm": 0.22893080115318298,
+      "learning_rate": 4.1447890495570205e-05,
+      "loss": 0.406,
       "step": 40705
     },
     {
-      "epoch": 1.43,
-      "learning_rate": 4.186594358047226e-05,
-      "loss": 0.3012,
+      "epoch": 1.4671856416909936,
+      "grad_norm": 0.1730494350194931,
+      "learning_rate": 4.1445692761431743e-05,
+      "loss": 0.4144,
       "step": 40710
     },
     {
-      "epoch": 1.43,
-      "learning_rate": 4.1863840690251255e-05,
-      "loss": 0.2882,
+      "epoch": 1.4673658413522181,
+      "grad_norm": 0.22228698432445526,
+      "learning_rate": 4.14434948032239e-05,
+      "loss": 0.4327,
       "step": 40715
     },
     {
-      "epoch": 1.43,
-      "learning_rate": 4.1861737581064386e-05,
-      "loss": 0.2928,
+      "epoch": 1.4675460410134429,
+      "grad_norm": 0.18846584856510162,
+      "learning_rate": 4.144129662097663e-05,
+      "loss": 0.4167,
       "step": 40720
     },
     {
-      "epoch": 1.43,
-      "learning_rate": 4.1859634252938974e-05,
-      "loss": 0.2965,
+      "epoch": 1.4677262406746676,
+      "grad_norm": 0.18671122193336487,
+      "learning_rate": 4.143909821471988e-05,
+      "loss": 0.4348,
       "step": 40725
     },
     {
-      "epoch": 1.43,
-      "learning_rate": 4.1857530705902316e-05,
-      "loss": 0.2917,
+      "epoch": 1.4679064403358921,
+      "grad_norm": 0.15065552294254303,
+      "learning_rate": 4.143689958448359e-05,
+      "loss": 0.4215,
       "step": 40730
     },
     {
-      "epoch": 1.43,
-      "learning_rate": 4.185542693998173e-05,
-      "loss": 0.2961,
+      "epoch": 1.4680866399971169,
+      "grad_norm": 0.19978941977024078,
+      "learning_rate": 4.143470073029774e-05,
+      "loss": 0.4065,
       "step": 40735
     },
     {
-      "epoch": 1.43,
-      "learning_rate": 4.1853322955204545e-05,
-      "loss": 0.2923,
+      "epoch": 1.4682668396583414,
+      "grad_norm": 0.1559489220380783,
+      "learning_rate": 4.143250165219226e-05,
+      "loss": 0.423,
       "step": 40740
     },
     {
-      "epoch": 1.43,
-      "learning_rate": 4.1851218751598056e-05,
-      "loss": 0.2859,
+      "epoch": 1.468447039319566,
+      "grad_norm": 0.1872124969959259,
+      "learning_rate": 4.143030235019713e-05,
+      "loss": 0.4196,
       "step": 40745
     },
     {
-      "epoch": 1.43,
-      "learning_rate": 4.18491143291896e-05,
-      "loss": 0.2739,
+      "epoch": 1.4686272389807908,
+      "grad_norm": 0.20608900487422943,
+      "learning_rate": 4.142810282434231e-05,
+      "loss": 0.419,
       "step": 40750
     },
     {
-      "epoch": 1.43,
-      "learning_rate": 4.18470096880065e-05,
-      "loss": 0.2854,
+      "epoch": 1.4688074386420154,
+      "grad_norm": 0.16788601875305176,
+      "learning_rate": 4.1425903074657776e-05,
+      "loss": 0.4231,
       "step": 40755
     },
     {
-      "epoch": 1.43,
-      "learning_rate": 4.1844904828076086e-05,
-      "loss": 0.2799,
+      "epoch": 1.4689876383032399,
+      "grad_norm": 0.305698037147522,
+      "learning_rate": 4.142370310117348e-05,
+      "loss": 0.3935,
       "step": 40760
     },
     {
-      "epoch": 1.43,
-      "learning_rate": 4.184279974942568e-05,
-      "loss": 0.3104,
+      "epoch": 1.4691678379644646,
+      "grad_norm": 0.13668397068977356,
+      "learning_rate": 4.142150290391943e-05,
+      "loss": 0.387,
       "step": 40765
     },
     {
-      "epoch": 1.43,
-      "learning_rate": 4.184069445208262e-05,
-      "loss": 0.2642,
+      "epoch": 1.4693480376256893,
+      "grad_norm": 0.15131601691246033,
+      "learning_rate": 4.141930248292557e-05,
+      "loss": 0.4168,
       "step": 40770
     },
     {
-      "epoch": 1.43,
-      "learning_rate": 4.183858893607423e-05,
-      "loss": 0.2641,
+      "epoch": 1.4695282372869138,
+      "grad_norm": 0.19707903265953064,
+      "learning_rate": 4.141710183822189e-05,
+      "loss": 0.4628,
       "step": 40775
     },
     {
-      "epoch": 1.43,
-      "learning_rate": 4.1836483201427876e-05,
-      "loss": 0.3153,
+      "epoch": 1.4697084369481386,
+      "grad_norm": 0.19335195422172546,
+      "learning_rate": 4.1414900969838375e-05,
+      "loss": 0.4157,
       "step": 40780
     },
     {
-      "epoch": 1.43,
-      "learning_rate": 4.183437724817087e-05,
-      "loss": 0.2718,
+      "epoch": 1.469888636609363,
+      "grad_norm": 0.20567668974399567,
+      "learning_rate": 4.141269987780502e-05,
+      "loss": 0.4535,
       "step": 40785
     },
     {
-      "epoch": 1.44,
-      "learning_rate": 4.183227107633058e-05,
-      "loss": 0.2915,
+      "epoch": 1.4700688362705878,
+      "grad_norm": 0.17595280706882477,
+      "learning_rate": 4.14104985621518e-05,
+      "loss": 0.4263,
       "step": 40790
     },
     {
-      "epoch": 1.44,
-      "learning_rate": 4.183016468593434e-05,
-      "loss": 0.2779,
+      "epoch": 1.4702490359318126,
+      "grad_norm": 0.18598760664463043,
+      "learning_rate": 4.140829702290872e-05,
+      "loss": 0.4139,
       "step": 40795
     },
     {
-      "epoch": 1.44,
-      "learning_rate": 4.18280580770095e-05,
-      "loss": 0.2813,
+      "epoch": 1.470429235593037,
+      "grad_norm": 0.19767579436302185,
+      "learning_rate": 4.140609526010576e-05,
+      "loss": 0.3817,
       "step": 40800
     },
     {
-      "epoch": 1.44,
-      "learning_rate": 4.182595124958342e-05,
-      "loss": 0.2932,
+      "epoch": 1.4706094352542618,
+      "grad_norm": 0.18627217411994934,
+      "learning_rate": 4.140389327377294e-05,
+      "loss": 0.3926,
       "step": 40805
     },
     {
-      "epoch": 1.44,
-      "learning_rate": 4.1823844203683456e-05,
-      "loss": 0.2671,
+      "epoch": 1.4707896349154863,
+      "grad_norm": 0.20653992891311646,
+      "learning_rate": 4.140169106394024e-05,
+      "loss": 0.4532,
       "step": 40810
     },
     {
-      "epoch": 1.44,
-      "learning_rate": 4.182173693933695e-05,
-      "loss": 0.3033,
+      "epoch": 1.470969834576711,
+      "grad_norm": 0.16096976399421692,
+      "learning_rate": 4.139948863063768e-05,
+      "loss": 0.3833,
       "step": 40815
     },
     {
-      "epoch": 1.44,
-      "learning_rate": 4.1819629456571285e-05,
-      "loss": 0.3036,
+      "epoch": 1.4711500342379356,
+      "grad_norm": 0.14477817714214325,
+      "learning_rate": 4.1397285973895264e-05,
+      "loss": 0.4148,
       "step": 40820
     },
     {
-      "epoch": 1.44,
-      "learning_rate": 4.181752175541381e-05,
-      "loss": 0.3172,
+      "epoch": 1.4713302338991603,
+      "grad_norm": 0.17661945521831512,
+      "learning_rate": 4.1395083093743006e-05,
+      "loss": 0.4228,
       "step": 40825
     },
     {
-      "epoch": 1.44,
-      "learning_rate": 4.1815413835891904e-05,
-      "loss": 0.2817,
+      "epoch": 1.4715104335603848,
+      "grad_norm": 0.16946406662464142,
+      "learning_rate": 4.139287999021091e-05,
+      "loss": 0.4413,
       "step": 40830
     },
     {
-      "epoch": 1.44,
-      "learning_rate": 4.181330569803293e-05,
-      "loss": 0.2732,
+      "epoch": 1.4716906332216095,
+      "grad_norm": 0.2282014936208725,
+      "learning_rate": 4.1390676663328995e-05,
+      "loss": 0.4072,
       "step": 40835
     },
     {
-      "epoch": 1.44,
-      "learning_rate": 4.181119734186426e-05,
-      "loss": 0.2779,
+      "epoch": 1.4718708328828343,
+      "grad_norm": 0.1720079779624939,
+      "learning_rate": 4.138847311312728e-05,
+      "loss": 0.4107,
       "step": 40840
     },
     {
-      "epoch": 1.44,
-      "learning_rate": 4.180908876741327e-05,
-      "loss": 0.2961,
+      "epoch": 1.4720510325440588,
+      "grad_norm": 0.1870976686477661,
+      "learning_rate": 4.13862693396358e-05,
+      "loss": 0.4437,
       "step": 40845
     },
     {
-      "epoch": 1.44,
-      "learning_rate": 4.180697997470734e-05,
-      "loss": 0.2648,
+      "epoch": 1.4722312322052835,
+      "grad_norm": 0.14838680624961853,
+      "learning_rate": 4.138406534288457e-05,
+      "loss": 0.4164,
       "step": 40850
     },
     {
-      "epoch": 1.44,
-      "learning_rate": 4.1804870963773855e-05,
-      "loss": 0.2808,
+      "epoch": 1.472411431866508,
+      "grad_norm": 0.146432027220726,
+      "learning_rate": 4.138186112290362e-05,
+      "loss": 0.3791,
       "step": 40855
     },
     {
-      "epoch": 1.44,
-      "learning_rate": 4.180276173464019e-05,
-      "loss": 0.2897,
+      "epoch": 1.4725916315277328,
+      "grad_norm": 0.1663684993982315,
+      "learning_rate": 4.137965667972298e-05,
+      "loss": 0.3969,
       "step": 40860
     },
     {
-      "epoch": 1.44,
-      "learning_rate": 4.1800652287333745e-05,
-      "loss": 0.2875,
+      "epoch": 1.4727718311889575,
+      "grad_norm": 0.24084417521953583,
+      "learning_rate": 4.1377452013372695e-05,
+      "loss": 0.4105,
       "step": 40865
     },
     {
-      "epoch": 1.44,
-      "learning_rate": 4.17985426218819e-05,
-      "loss": 0.2783,
+      "epoch": 1.472952030850182,
+      "grad_norm": 0.21911956369876862,
+      "learning_rate": 4.13752471238828e-05,
+      "loss": 0.45,
       "step": 40870
     },
     {
-      "epoch": 1.44,
-      "learning_rate": 4.1796432738312046e-05,
-      "loss": 0.2906,
+      "epoch": 1.4731322305114065,
+      "grad_norm": 0.20108048617839813,
+      "learning_rate": 4.137304201128334e-05,
+      "loss": 0.432,
       "step": 40875
     },
     {
-      "epoch": 1.44,
-      "learning_rate": 4.179432263665159e-05,
-      "loss": 0.2797,
+      "epoch": 1.4733124301726312,
+      "grad_norm": 0.16961334645748138,
+      "learning_rate": 4.1370836675604326e-05,
+      "loss": 0.401,
       "step": 40880
     },
     {
-      "epoch": 1.44,
-      "learning_rate": 4.1792212316927917e-05,
-      "loss": 0.3023,
+      "epoch": 1.473492629833856,
+      "grad_norm": 0.16937093436717987,
+      "learning_rate": 4.1368631116875856e-05,
+      "loss": 0.4159,
       "step": 40885
     },
     {
-      "epoch": 1.44,
-      "learning_rate": 4.1790101779168425e-05,
-      "loss": 0.301,
+      "epoch": 1.4736728294950805,
+      "grad_norm": 0.19566501677036285,
+      "learning_rate": 4.136642533512795e-05,
+      "loss": 0.4196,
       "step": 40890
     },
     {
-      "epoch": 1.44,
-      "learning_rate": 4.1787991023400534e-05,
-      "loss": 0.2779,
+      "epoch": 1.4738530291563052,
+      "grad_norm": 0.17469613254070282,
+      "learning_rate": 4.136421933039066e-05,
+      "loss": 0.3932,
       "step": 40895
     },
     {
-      "epoch": 1.44,
-      "learning_rate": 4.178588004965165e-05,
-      "loss": 0.2833,
+      "epoch": 1.4740332288175297,
+      "grad_norm": 0.18725554645061493,
+      "learning_rate": 4.136201310269406e-05,
+      "loss": 0.4323,
       "step": 40900
     },
     {
-      "epoch": 1.44,
-      "learning_rate": 4.1783768857949166e-05,
-      "loss": 0.3119,
+      "epoch": 1.4742134284787545,
+      "grad_norm": 0.21251484751701355,
+      "learning_rate": 4.135980665206819e-05,
+      "loss": 0.4335,
       "step": 40905
     },
     {
-      "epoch": 1.44,
-      "learning_rate": 4.178165744832051e-05,
-      "loss": 0.293,
+      "epoch": 1.4743936281399792,
+      "grad_norm": 0.21791920065879822,
+      "learning_rate": 4.135759997854313e-05,
+      "loss": 0.4479,
       "step": 40910
     },
     {
-      "epoch": 1.44,
-      "learning_rate": 4.1779545820793085e-05,
-      "loss": 0.2687,
+      "epoch": 1.4745738278012037,
+      "grad_norm": 0.19280044734477997,
+      "learning_rate": 4.1355393082148936e-05,
+      "loss": 0.387,
       "step": 40915
     },
     {
-      "epoch": 1.44,
-      "learning_rate": 4.177743397539432e-05,
-      "loss": 0.2857,
+      "epoch": 1.4747540274624285,
+      "grad_norm": 0.14178399741649628,
+      "learning_rate": 4.1353185962915675e-05,
+      "loss": 0.4095,
       "step": 40920
     },
     {
-      "epoch": 1.44,
-      "learning_rate": 4.177532191215163e-05,
-      "loss": 0.3076,
+      "epoch": 1.474934227123653,
+      "grad_norm": 0.24102537333965302,
+      "learning_rate": 4.135097862087342e-05,
+      "loss": 0.4249,
       "step": 40925
     },
     {
-      "epoch": 1.44,
-      "learning_rate": 4.177320963109244e-05,
-      "loss": 0.2961,
+      "epoch": 1.4751144267848777,
+      "grad_norm": 0.2044147253036499,
+      "learning_rate": 4.134877105605225e-05,
+      "loss": 0.3982,
       "step": 40930
     },
     {
-      "epoch": 1.44,
-      "learning_rate": 4.177109713224418e-05,
-      "loss": 0.2973,
+      "epoch": 1.4752946264461022,
+      "grad_norm": 0.14155633747577667,
+      "learning_rate": 4.1346563268482245e-05,
+      "loss": 0.4237,
       "step": 40935
     },
     {
-      "epoch": 1.44,
-      "learning_rate": 4.176898441563428e-05,
-      "loss": 0.298,
+      "epoch": 1.475474826107327,
+      "grad_norm": 0.20636247098445892,
+      "learning_rate": 4.134435525819347e-05,
+      "loss": 0.414,
       "step": 40940
     },
     {
-      "epoch": 1.44,
-      "learning_rate": 4.176687148129016e-05,
-      "loss": 0.2911,
+      "epoch": 1.4756550257685515,
+      "grad_norm": 0.16069145500659943,
+      "learning_rate": 4.1342147025216015e-05,
+      "loss": 0.4105,
       "step": 40945
     },
     {
-      "epoch": 1.44,
-      "learning_rate": 4.176475832923927e-05,
-      "loss": 0.3035,
+      "epoch": 1.4758352254297762,
+      "grad_norm": 0.2055576741695404,
+      "learning_rate": 4.1339938569579985e-05,
+      "loss": 0.4265,
       "step": 40950
     },
     {
-      "epoch": 1.44,
-      "learning_rate": 4.1762644959509035e-05,
-      "loss": 0.2867,
+      "epoch": 1.476015425091001,
+      "grad_norm": 0.1672385036945343,
+      "learning_rate": 4.1337729891315445e-05,
+      "loss": 0.4226,
       "step": 40955
     },
     {
-      "epoch": 1.44,
-      "learning_rate": 4.176053137212691e-05,
-      "loss": 0.2858,
+      "epoch": 1.4761956247522254,
+      "grad_norm": 0.1980341076850891,
+      "learning_rate": 4.1335520990452504e-05,
+      "loss": 0.4252,
       "step": 40960
     },
     {
-      "epoch": 1.44,
-      "learning_rate": 4.175841756712032e-05,
-      "loss": 0.296,
+      "epoch": 1.4763758244134502,
+      "grad_norm": 0.19401240348815918,
+      "learning_rate": 4.1333311867021254e-05,
+      "loss": 0.4046,
       "step": 40965
     },
     {
-      "epoch": 1.44,
-      "learning_rate": 4.175630354451673e-05,
-      "loss": 0.2876,
+      "epoch": 1.4765560240746747,
+      "grad_norm": 0.2943877875804901,
+      "learning_rate": 4.133110252105178e-05,
+      "loss": 0.4149,
       "step": 40970
     },
     {
-      "epoch": 1.44,
-      "learning_rate": 4.1754189304343584e-05,
-      "loss": 0.2937,
+      "epoch": 1.4767362237358994,
+      "grad_norm": 0.15814274549484253,
+      "learning_rate": 4.13288929525742e-05,
+      "loss": 0.4066,
       "step": 40975
     },
     {
-      "epoch": 1.44,
-      "learning_rate": 4.175207484662833e-05,
-      "loss": 0.2828,
+      "epoch": 1.4769164233971241,
+      "grad_norm": 0.1717575490474701,
+      "learning_rate": 4.132668316161863e-05,
+      "loss": 0.4235,
       "step": 40980
     },
     {
-      "epoch": 1.44,
-      "learning_rate": 4.1749960171398413e-05,
-      "loss": 0.262,
+      "epoch": 1.4770966230583487,
+      "grad_norm": 0.1958007961511612,
+      "learning_rate": 4.1324473148215146e-05,
+      "loss": 0.4493,
       "step": 40985
     },
     {
-      "epoch": 1.44,
-      "learning_rate": 4.174784527868132e-05,
-      "loss": 0.3013,
+      "epoch": 1.4772768227195732,
+      "grad_norm": 0.17059576511383057,
+      "learning_rate": 4.1322262912393884e-05,
+      "loss": 0.3968,
       "step": 40990
     },
     {
-      "epoch": 1.44,
-      "learning_rate": 4.174573016850448e-05,
-      "loss": 0.3128,
+      "epoch": 1.477457022380798,
+      "grad_norm": 0.18559856712818146,
+      "learning_rate": 4.132005245418495e-05,
+      "loss": 0.4099,
       "step": 40995
     },
     {
-      "epoch": 1.44,
-      "learning_rate": 4.174361484089537e-05,
-      "loss": 0.2979,
+      "epoch": 1.4776372220420226,
+      "grad_norm": 0.206581249833107,
+      "learning_rate": 4.131784177361845e-05,
+      "loss": 0.4063,
       "step": 41000
     },
     {
-      "epoch": 1.44,
-      "eval_loss": 0.28040316700935364,
-      "eval_runtime": 10.542,
-      "eval_samples_per_second": 9.486,
-      "eval_steps_per_second": 9.486,
+      "epoch": 1.4776372220420226,
+      "eval_loss": 0.44472381472587585,
+      "eval_runtime": 3.5272,
+      "eval_samples_per_second": 28.351,
+      "eval_steps_per_second": 7.088,
       "step": 41000
     },
     {
-      "epoch": 1.44,
-      "learning_rate": 4.174149929588146e-05,
-      "loss": 0.2828,
+      "epoch": 1.4778174217032471,
+      "grad_norm": 0.2029978483915329,
+      "learning_rate": 4.131563087072453e-05,
+      "loss": 0.4428,
       "step": 41005
     },
     {
-      "epoch": 1.44,
-      "learning_rate": 4.173938353349022e-05,
-      "loss": 0.296,
+      "epoch": 1.4779976213644719,
+      "grad_norm": 0.1657671481370926,
+      "learning_rate": 4.131341974553329e-05,
+      "loss": 0.4282,
       "step": 41010
     },
     {
-      "epoch": 1.44,
-      "learning_rate": 4.1737267553749105e-05,
-      "loss": 0.3014,
+      "epoch": 1.4781778210256964,
+      "grad_norm": 0.22293435037136078,
+      "learning_rate": 4.131120839807487e-05,
+      "loss": 0.4178,
       "step": 41015
     },
     {
-      "epoch": 1.44,
-      "learning_rate": 4.173515135668561e-05,
-      "loss": 0.2585,
+      "epoch": 1.4783580206869211,
+      "grad_norm": 0.18376168608665466,
+      "learning_rate": 4.130899682837939e-05,
+      "loss": 0.4239,
       "step": 41020
     },
     {
-      "epoch": 1.44,
-      "learning_rate": 4.17330349423272e-05,
-      "loss": 0.2728,
+      "epoch": 1.4785382203481459,
+      "grad_norm": 0.20025895535945892,
+      "learning_rate": 4.130678503647698e-05,
+      "loss": 0.4262,
       "step": 41025
     },
     {
-      "epoch": 1.44,
-      "learning_rate": 4.173091831070135e-05,
-      "loss": 0.2747,
+      "epoch": 1.4787184200093704,
+      "grad_norm": 0.16391725838184357,
+      "learning_rate": 4.1304573022397784e-05,
+      "loss": 0.3756,
       "step": 41030
     },
     {
-      "epoch": 1.44,
-      "learning_rate": 4.172880146183556e-05,
-      "loss": 0.2934,
+      "epoch": 1.4788986196705949,
+      "grad_norm": 0.16140343248844147,
+      "learning_rate": 4.1302360786171946e-05,
+      "loss": 0.4036,
       "step": 41035
     },
     {
-      "epoch": 1.44,
-      "learning_rate": 4.172668439575731e-05,
-      "loss": 0.2874,
+      "epoch": 1.4790788193318196,
+      "grad_norm": 0.23130172491073608,
+      "learning_rate": 4.1300148327829593e-05,
+      "loss": 0.4511,
       "step": 41040
     },
     {
-      "epoch": 1.44,
-      "learning_rate": 4.1724567112494087e-05,
-      "loss": 0.2871,
+      "epoch": 1.4792590189930444,
+      "grad_norm": 0.17555512487888336,
+      "learning_rate": 4.1297935647400874e-05,
+      "loss": 0.4236,
       "step": 41045
     },
     {
-      "epoch": 1.44,
-      "learning_rate": 4.172244961207338e-05,
-      "loss": 0.2739,
+      "epoch": 1.4794392186542689,
+      "grad_norm": 0.18214865028858185,
+      "learning_rate": 4.1295722744915934e-05,
+      "loss": 0.4806,
       "step": 41050
     },
     {
-      "epoch": 1.44,
-      "learning_rate": 4.172033189452268e-05,
-      "loss": 0.2983,
+      "epoch": 1.4796194183154936,
+      "grad_norm": 0.17145216464996338,
+      "learning_rate": 4.129350962040494e-05,
+      "loss": 0.4215,
       "step": 41055
     },
     {
-      "epoch": 1.44,
-      "learning_rate": 4.17182139598695e-05,
-      "loss": 0.2867,
+      "epoch": 1.479799617976718,
+      "grad_norm": 0.1968265026807785,
+      "learning_rate": 4.1291296273898015e-05,
+      "loss": 0.4054,
       "step": 41060
     },
     {
-      "epoch": 1.44,
-      "learning_rate": 4.171609580814132e-05,
-      "loss": 0.2692,
+      "epoch": 1.4799798176379428,
+      "grad_norm": 0.14597874879837036,
+      "learning_rate": 4.1289082705425344e-05,
+      "loss": 0.4229,
       "step": 41065
     },
     {
-      "epoch": 1.44,
-      "learning_rate": 4.171397743936565e-05,
-      "loss": 0.2837,
+      "epoch": 1.4801600172991676,
+      "grad_norm": 0.1973959356546402,
+      "learning_rate": 4.1286868915017064e-05,
+      "loss": 0.3867,
       "step": 41070
     },
     {
-      "epoch": 1.45,
-      "learning_rate": 4.171185885357001e-05,
-      "loss": 0.2832,
+      "epoch": 1.480340216960392,
+      "grad_norm": 0.18178540468215942,
+      "learning_rate": 4.1284654902703356e-05,
+      "loss": 0.4134,
       "step": 41075
     },
     {
-      "epoch": 1.45,
-      "learning_rate": 4.170974005078189e-05,
-      "loss": 0.2724,
+      "epoch": 1.4805204166216168,
+      "grad_norm": 0.2181033045053482,
+      "learning_rate": 4.1282440668514376e-05,
+      "loss": 0.4203,
       "step": 41080
     },
     {
-      "epoch": 1.45,
-      "learning_rate": 4.170762103102881e-05,
-      "loss": 0.2777,
+      "epoch": 1.4807006162828413,
+      "grad_norm": 0.18515872955322266,
+      "learning_rate": 4.128022621248029e-05,
+      "loss": 0.4207,
       "step": 41085
     },
     {
-      "epoch": 1.45,
-      "learning_rate": 4.1705501794338274e-05,
-      "loss": 0.3046,
+      "epoch": 1.480880815944066,
+      "grad_norm": 0.16878053545951843,
+      "learning_rate": 4.1278011534631276e-05,
+      "loss": 0.4058,
       "step": 41090
     },
     {
-      "epoch": 1.45,
-      "learning_rate": 4.170338234073781e-05,
-      "loss": 0.2767,
+      "epoch": 1.4810610156052908,
+      "grad_norm": 0.1830851435661316,
+      "learning_rate": 4.127579663499752e-05,
+      "loss": 0.3987,
       "step": 41095
     },
     {
-      "epoch": 1.45,
-      "learning_rate": 4.170126267025494e-05,
-      "loss": 0.2689,
+      "epoch": 1.4812412152665153,
+      "grad_norm": 0.1524893045425415,
+      "learning_rate": 4.1273581513609173e-05,
+      "loss": 0.4154,
       "step": 41100
     },
     {
-      "epoch": 1.45,
-      "learning_rate": 4.169914278291718e-05,
-      "loss": 0.2693,
+      "epoch": 1.4814214149277398,
+      "grad_norm": 0.18265849351882935,
+      "learning_rate": 4.127136617049643e-05,
+      "loss": 0.4195,
       "step": 41105
     },
     {
-      "epoch": 1.45,
-      "learning_rate": 4.169702267875205e-05,
-      "loss": 0.2948,
+      "epoch": 1.4816016145889646,
+      "grad_norm": 0.18311630189418793,
+      "learning_rate": 4.126915060568947e-05,
+      "loss": 0.4548,
       "step": 41110
     },
     {
-      "epoch": 1.45,
-      "learning_rate": 4.169490235778709e-05,
-      "loss": 0.295,
+      "epoch": 1.4817818142501893,
+      "grad_norm": 0.1863049566745758,
+      "learning_rate": 4.126693481921848e-05,
+      "loss": 0.4483,
       "step": 41115
     },
     {
-      "epoch": 1.45,
-      "learning_rate": 4.169278182004982e-05,
-      "loss": 0.2893,
+      "epoch": 1.4819620139114138,
+      "grad_norm": 0.15636593103408813,
+      "learning_rate": 4.126471881111367e-05,
+      "loss": 0.4159,
       "step": 41120
     },
     {
-      "epoch": 1.45,
-      "learning_rate": 4.169066106556778e-05,
-      "loss": 0.2922,
+      "epoch": 1.4821422135726385,
+      "grad_norm": 0.1574103981256485,
+      "learning_rate": 4.1262502581405196e-05,
+      "loss": 0.4247,
       "step": 41125
     },
     {
-      "epoch": 1.45,
-      "learning_rate": 4.1688540094368505e-05,
-      "loss": 0.3208,
+      "epoch": 1.482322413233863,
+      "grad_norm": 0.1982761174440384,
+      "learning_rate": 4.1260286130123285e-05,
+      "loss": 0.4114,
       "step": 41130
     },
     {
-      "epoch": 1.45,
-      "learning_rate": 4.168641890647954e-05,
-      "loss": 0.2858,
+      "epoch": 1.4825026128950878,
+      "grad_norm": 0.17163503170013428,
+      "learning_rate": 4.125806945729812e-05,
+      "loss": 0.4174,
       "step": 41135
     },
     {
-      "epoch": 1.45,
-      "learning_rate": 4.1684297501928416e-05,
-      "loss": 0.2928,
+      "epoch": 1.4826828125563125,
+      "grad_norm": 0.1855458766222,
+      "learning_rate": 4.1255852562959904e-05,
+      "loss": 0.4327,
       "step": 41140
     },
     {
-      "epoch": 1.45,
-      "learning_rate": 4.1682175880742694e-05,
-      "loss": 0.2622,
+      "epoch": 1.482863012217537,
+      "grad_norm": 0.21407297253608704,
+      "learning_rate": 4.125363544713884e-05,
+      "loss": 0.4342,
       "step": 41145
     },
     {
-      "epoch": 1.45,
-      "learning_rate": 4.16800540429499e-05,
-      "loss": 0.2929,
+      "epoch": 1.4830432118787615,
+      "grad_norm": 0.20228859782218933,
+      "learning_rate": 4.1251418109865146e-05,
+      "loss": 0.4116,
       "step": 41150
     },
     {
-      "epoch": 1.45,
-      "learning_rate": 4.167793198857761e-05,
-      "loss": 0.2812,
+      "epoch": 1.4832234115399863,
+      "grad_norm": 0.19245915114879608,
+      "learning_rate": 4.124920055116903e-05,
+      "loss": 0.3937,
       "step": 41155
     },
     {
-      "epoch": 1.45,
-      "learning_rate": 4.167580971765335e-05,
-      "loss": 0.2923,
+      "epoch": 1.483403611201211,
+      "grad_norm": 0.1870017945766449,
+      "learning_rate": 4.12469827710807e-05,
+      "loss": 0.426,
       "step": 41160
     },
     {
-      "epoch": 1.45,
-      "learning_rate": 4.1673687230204695e-05,
-      "loss": 0.3185,
+      "epoch": 1.4835838108624355,
+      "grad_norm": 0.1986088752746582,
+      "learning_rate": 4.1244764769630375e-05,
+      "loss": 0.4125,
       "step": 41165
     },
     {
-      "epoch": 1.45,
-      "learning_rate": 4.16715645262592e-05,
-      "loss": 0.2985,
+      "epoch": 1.4837640105236602,
+      "grad_norm": 0.18899187445640564,
+      "learning_rate": 4.124254654684827e-05,
+      "loss": 0.4205,
       "step": 41170
     },
     {
-      "epoch": 1.45,
-      "learning_rate": 4.166944160584443e-05,
-      "loss": 0.3063,
+      "epoch": 1.4839442101848848,
+      "grad_norm": 0.1863652914762497,
+      "learning_rate": 4.1240328102764614e-05,
+      "loss": 0.4055,
       "step": 41175
     },
     {
-      "epoch": 1.45,
-      "learning_rate": 4.166731846898795e-05,
-      "loss": 0.2869,
+      "epoch": 1.4841244098461095,
+      "grad_norm": 0.2102532535791397,
+      "learning_rate": 4.1238109437409635e-05,
+      "loss": 0.4175,
       "step": 41180
     },
     {
-      "epoch": 1.45,
-      "learning_rate": 4.166519511571731e-05,
-      "loss": 0.3082,
+      "epoch": 1.4843046095073342,
+      "grad_norm": 0.17602580785751343,
+      "learning_rate": 4.123589055081356e-05,
+      "loss": 0.3868,
       "step": 41185
     },
     {
-      "epoch": 1.45,
-      "learning_rate": 4.166307154606011e-05,
-      "loss": 0.2986,
+      "epoch": 1.4844848091685587,
+      "grad_norm": 0.1657838225364685,
+      "learning_rate": 4.123367144300662e-05,
+      "loss": 0.3956,
       "step": 41190
     },
     {
-      "epoch": 1.45,
-      "learning_rate": 4.16609477600439e-05,
-      "loss": 0.2614,
+      "epoch": 1.4846650088297835,
+      "grad_norm": 0.15461941063404083,
+      "learning_rate": 4.123145211401904e-05,
+      "loss": 0.432,
       "step": 41195
     },
     {
-      "epoch": 1.45,
-      "learning_rate": 4.1658823757696265e-05,
-      "loss": 0.2923,
+      "epoch": 1.484845208491008,
+      "grad_norm": 0.17926329374313354,
+      "learning_rate": 4.1229232563881084e-05,
+      "loss": 0.4127,
       "step": 41200
     },
     {
-      "epoch": 1.45,
-      "learning_rate": 4.165669953904478e-05,
-      "loss": 0.2819,
+      "epoch": 1.4850254081522327,
+      "grad_norm": 0.15599684417247772,
+      "learning_rate": 4.122701279262296e-05,
+      "loss": 0.4382,
       "step": 41205
     },
     {
-      "epoch": 1.45,
-      "learning_rate": 4.165457510411703e-05,
-      "loss": 0.2844,
+      "epoch": 1.4852056078134572,
+      "grad_norm": 0.17323412001132965,
+      "learning_rate": 4.1224792800274936e-05,
+      "loss": 0.39,
       "step": 41210
     },
     {
-      "epoch": 1.45,
-      "learning_rate": 4.1652450452940594e-05,
-      "loss": 0.2923,
+      "epoch": 1.485385807474682,
+      "grad_norm": 0.19088657200336456,
+      "learning_rate": 4.122257258686725e-05,
+      "loss": 0.4044,
       "step": 41215
     },
     {
-      "epoch": 1.45,
-      "learning_rate": 4.1650325585543073e-05,
-      "loss": 0.2964,
+      "epoch": 1.4855660071359065,
+      "grad_norm": 0.1666693538427353,
+      "learning_rate": 4.122035215243015e-05,
+      "loss": 0.422,
       "step": 41220
     },
     {
-      "epoch": 1.45,
-      "learning_rate": 4.164820050195204e-05,
-      "loss": 0.2783,
+      "epoch": 1.4857462067971312,
+      "grad_norm": 0.23202933371067047,
+      "learning_rate": 4.12181314969939e-05,
+      "loss": 0.4344,
       "step": 41225
     },
     {
-      "epoch": 1.45,
-      "learning_rate": 4.164607520219509e-05,
-      "loss": 0.2901,
+      "epoch": 1.485926406458356,
+      "grad_norm": 0.19096602499485016,
+      "learning_rate": 4.1215910620588745e-05,
+      "loss": 0.4192,
       "step": 41230
     },
     {
-      "epoch": 1.45,
-      "learning_rate": 4.164394968629983e-05,
-      "loss": 0.2434,
+      "epoch": 1.4861066061195805,
+      "grad_norm": 0.21714234352111816,
+      "learning_rate": 4.1213689523244945e-05,
+      "loss": 0.397,
       "step": 41235
     },
     {
-      "epoch": 1.45,
-      "learning_rate": 4.1641823954293855e-05,
-      "loss": 0.2941,
+      "epoch": 1.4862868057808052,
+      "grad_norm": 0.22744521498680115,
+      "learning_rate": 4.121146820499277e-05,
+      "loss": 0.439,
       "step": 41240
     },
     {
-      "epoch": 1.45,
-      "learning_rate": 4.163969800620476e-05,
-      "loss": 0.3284,
+      "epoch": 1.4864670054420297,
+      "grad_norm": 0.1769517958164215,
+      "learning_rate": 4.120924666586248e-05,
+      "loss": 0.4124,
       "step": 41245
     },
     {
-      "epoch": 1.45,
-      "learning_rate": 4.1637571842060156e-05,
-      "loss": 0.3151,
+      "epoch": 1.4866472051032544,
+      "grad_norm": 0.16635587811470032,
+      "learning_rate": 4.1207024905884335e-05,
+      "loss": 0.4265,
       "step": 41250
     },
     {
-      "epoch": 1.45,
-      "learning_rate": 4.163544546188764e-05,
-      "loss": 0.2672,
+      "epoch": 1.4868274047644792,
+      "grad_norm": 0.1759326308965683,
+      "learning_rate": 4.120480292508861e-05,
+      "loss": 0.4143,
       "step": 41255
     },
     {
-      "epoch": 1.45,
-      "learning_rate": 4.163331886571483e-05,
-      "loss": 0.2914,
+      "epoch": 1.4870076044257037,
+      "grad_norm": 0.15628643333911896,
+      "learning_rate": 4.120258072350559e-05,
+      "loss": 0.4084,
       "step": 41260
     },
     {
-      "epoch": 1.45,
-      "learning_rate": 4.163119205356934e-05,
-      "loss": 0.2725,
+      "epoch": 1.4871878040869282,
+      "grad_norm": 0.1729431450366974,
+      "learning_rate": 4.1200358301165544e-05,
+      "loss": 0.4391,
       "step": 41265
     },
     {
-      "epoch": 1.45,
-      "learning_rate": 4.162906502547877e-05,
-      "loss": 0.2869,
+      "epoch": 1.487368003748153,
+      "grad_norm": 0.20765365660190582,
+      "learning_rate": 4.1198135658098755e-05,
+      "loss": 0.4579,
       "step": 41270
     },
     {
-      "epoch": 1.45,
-      "learning_rate": 4.162693778147077e-05,
-      "loss": 0.2658,
+      "epoch": 1.4875482034093777,
+      "grad_norm": 0.18058431148529053,
+      "learning_rate": 4.11959127943355e-05,
+      "loss": 0.4406,
       "step": 41275
     },
     {
-      "epoch": 1.45,
-      "learning_rate": 4.162481032157293e-05,
-      "loss": 0.2572,
+      "epoch": 1.4877284030706022,
+      "grad_norm": 0.16490638256072998,
+      "learning_rate": 4.119368970990607e-05,
+      "loss": 0.4603,
       "step": 41280
     },
     {
-      "epoch": 1.45,
-      "learning_rate": 4.1622682645812886e-05,
-      "loss": 0.2777,
+      "epoch": 1.487908602731827,
+      "grad_norm": 0.20363958179950714,
+      "learning_rate": 4.119146640484075e-05,
+      "loss": 0.3987,
       "step": 41285
     },
     {
-      "epoch": 1.45,
-      "learning_rate": 4.162055475421825e-05,
-      "loss": 0.307,
+      "epoch": 1.4880888023930514,
+      "grad_norm": 0.17337477207183838,
+      "learning_rate": 4.118924287916984e-05,
+      "loss": 0.4447,
       "step": 41290
     },
     {
-      "epoch": 1.45,
-      "learning_rate": 4.161842664681668e-05,
-      "loss": 0.2797,
+      "epoch": 1.4882690020542761,
+      "grad_norm": 0.1706027388572693,
+      "learning_rate": 4.118701913292363e-05,
+      "loss": 0.4281,
       "step": 41295
     },
     {
-      "epoch": 1.45,
-      "learning_rate": 4.1616298323635776e-05,
-      "loss": 0.2818,
+      "epoch": 1.4884492017155009,
+      "grad_norm": 0.22588284313678741,
+      "learning_rate": 4.118479516613242e-05,
+      "loss": 0.4214,
       "step": 41300
     },
     {
-      "epoch": 1.45,
-      "learning_rate": 4.161416978470321e-05,
-      "loss": 0.298,
+      "epoch": 1.4886294013767254,
+      "grad_norm": 0.17973540723323822,
+      "learning_rate": 4.1182570978826496e-05,
+      "loss": 0.4239,
       "step": 41305
     },
     {
-      "epoch": 1.45,
-      "learning_rate": 4.1612041030046576e-05,
-      "loss": 0.2989,
+      "epoch": 1.4888096010379501,
+      "grad_norm": 0.1527833491563797,
+      "learning_rate": 4.118034657103619e-05,
+      "loss": 0.3976,
       "step": 41310
     },
     {
-      "epoch": 1.45,
-      "learning_rate": 4.1609912059693545e-05,
-      "loss": 0.3038,
+      "epoch": 1.4889898006991746,
+      "grad_norm": 0.15691448748111725,
+      "learning_rate": 4.1178121942791786e-05,
+      "loss": 0.4011,
       "step": 41315
     },
     {
-      "epoch": 1.45,
-      "learning_rate": 4.1607782873671756e-05,
-      "loss": 0.2784,
+      "epoch": 1.4891700003603994,
+      "grad_norm": 0.20908258855342865,
+      "learning_rate": 4.117589709412361e-05,
+      "loss": 0.4172,
       "step": 41320
     },
     {
-      "epoch": 1.45,
-      "learning_rate": 4.160565347200885e-05,
-      "loss": 0.3006,
+      "epoch": 1.4893502000216239,
+      "grad_norm": 0.18858399987220764,
+      "learning_rate": 4.117367202506196e-05,
+      "loss": 0.4134,
       "step": 41325
     },
     {
-      "epoch": 1.45,
-      "learning_rate": 4.160352385473248e-05,
-      "loss": 0.2945,
+      "epoch": 1.4895303996828486,
+      "grad_norm": 0.2324933558702469,
+      "learning_rate": 4.117144673563717e-05,
+      "loss": 0.4321,
       "step": 41330
     },
     {
-      "epoch": 1.45,
-      "learning_rate": 4.160139402187029e-05,
-      "loss": 0.2757,
+      "epoch": 1.4897105993440731,
+      "grad_norm": 0.1979554146528244,
+      "learning_rate": 4.116922122587954e-05,
+      "loss": 0.3956,
       "step": 41335
     },
     {
-      "epoch": 1.45,
-      "learning_rate": 4.1599263973449944e-05,
-      "loss": 0.2844,
+      "epoch": 1.4898907990052979,
+      "grad_norm": 0.17453639209270477,
+      "learning_rate": 4.11669954958194e-05,
+      "loss": 0.4303,
       "step": 41340
     },
     {
-      "epoch": 1.45,
-      "learning_rate": 4.1597133709499095e-05,
-      "loss": 0.2848,
+      "epoch": 1.4900709986665226,
+      "grad_norm": 0.17398759722709656,
+      "learning_rate": 4.116476954548708e-05,
+      "loss": 0.4485,
       "step": 41345
     },
     {
-      "epoch": 1.45,
-      "learning_rate": 4.159500323004539e-05,
-      "loss": 0.2781,
+      "epoch": 1.490251198327747,
+      "grad_norm": 0.1593339741230011,
+      "learning_rate": 4.116254337491291e-05,
+      "loss": 0.4429,
       "step": 41350
     },
     {
-      "epoch": 1.45,
-      "learning_rate": 4.159287253511652e-05,
-      "loss": 0.2788,
+      "epoch": 1.4904313979889718,
+      "grad_norm": 0.15613697469234467,
+      "learning_rate": 4.1160316984127205e-05,
+      "loss": 0.3826,
       "step": 41355
     },
     {
-      "epoch": 1.46,
-      "learning_rate": 4.159074162474013e-05,
-      "loss": 0.2766,
+      "epoch": 1.4906115976501964,
+      "grad_norm": 0.1693560630083084,
+      "learning_rate": 4.115809037316032e-05,
+      "loss": 0.4813,
       "step": 41360
     },
     {
-      "epoch": 1.46,
-      "learning_rate": 4.1588610498943895e-05,
-      "loss": 0.3108,
+      "epoch": 1.490791797311421,
+      "grad_norm": 0.15100261569023132,
+      "learning_rate": 4.1155863542042575e-05,
+      "loss": 0.4032,
       "step": 41365
     },
     {
-      "epoch": 1.46,
-      "learning_rate": 4.1586479157755483e-05,
-      "loss": 0.3063,
+      "epoch": 1.4909719969726458,
+      "grad_norm": 0.2202003002166748,
+      "learning_rate": 4.115363649080432e-05,
+      "loss": 0.3974,
       "step": 41370
     },
     {
-      "epoch": 1.46,
-      "learning_rate": 4.158434760120257e-05,
-      "loss": 0.2931,
+      "epoch": 1.4911521966338703,
+      "grad_norm": 0.1720680594444275,
+      "learning_rate": 4.11514092194759e-05,
+      "loss": 0.3987,
       "step": 41375
     },
     {
-      "epoch": 1.46,
-      "learning_rate": 4.158221582931283e-05,
-      "loss": 0.2764,
+      "epoch": 1.4913323962950948,
+      "grad_norm": 0.16511781513690948,
+      "learning_rate": 4.114918172808765e-05,
+      "loss": 0.3955,
       "step": 41380
     },
     {
-      "epoch": 1.46,
-      "learning_rate": 4.158008384211395e-05,
-      "loss": 0.2896,
+      "epoch": 1.4915125959563196,
+      "grad_norm": 0.18977214395999908,
+      "learning_rate": 4.1146954016669925e-05,
+      "loss": 0.3936,
       "step": 41385
     },
     {
-      "epoch": 1.46,
-      "learning_rate": 4.15779516396336e-05,
-      "loss": 0.3045,
+      "epoch": 1.4916927956175443,
+      "grad_norm": 0.18784664571285248,
+      "learning_rate": 4.1144726085253084e-05,
+      "loss": 0.4507,
       "step": 41390
     },
     {
-      "epoch": 1.46,
-      "learning_rate": 4.157581922189949e-05,
-      "loss": 0.3006,
+      "epoch": 1.4918729952787688,
+      "grad_norm": 0.1858154982328415,
+      "learning_rate": 4.1142497933867465e-05,
+      "loss": 0.4643,
       "step": 41395
     },
     {
-      "epoch": 1.46,
-      "learning_rate": 4.157368658893928e-05,
-      "loss": 0.2764,
+      "epoch": 1.4920531949399936,
+      "grad_norm": 0.19693292677402496,
+      "learning_rate": 4.1140269562543445e-05,
+      "loss": 0.411,
       "step": 41400
     },
     {
-      "epoch": 1.46,
-      "learning_rate": 4.1571553740780665e-05,
-      "loss": 0.287,
+      "epoch": 1.492233394601218,
+      "grad_norm": 0.14796632528305054,
+      "learning_rate": 4.113804097131138e-05,
+      "loss": 0.4257,
       "step": 41405
     },
     {
-      "epoch": 1.46,
-      "learning_rate": 4.156942067745136e-05,
-      "loss": 0.3057,
+      "epoch": 1.4924135942624428,
+      "grad_norm": 0.2132243663072586,
+      "learning_rate": 4.1135812160201624e-05,
+      "loss": 0.4285,
       "step": 41410
     },
     {
-      "epoch": 1.46,
-      "learning_rate": 4.1567287398979036e-05,
-      "loss": 0.2763,
+      "epoch": 1.4925937939236675,
+      "grad_norm": 0.151437446475029,
+      "learning_rate": 4.113358312924455e-05,
+      "loss": 0.3778,
       "step": 41415
     },
     {
-      "epoch": 1.46,
-      "learning_rate": 4.156515390539142e-05,
-      "loss": 0.2769,
+      "epoch": 1.492773993584892,
+      "grad_norm": 0.1567225605249405,
+      "learning_rate": 4.1131353878470536e-05,
+      "loss": 0.365,
       "step": 41420
     },
     {
-      "epoch": 1.46,
-      "learning_rate": 4.156302019671618e-05,
-      "loss": 0.2767,
+      "epoch": 1.4929541932461168,
+      "grad_norm": 0.19728878140449524,
+      "learning_rate": 4.1129124407909944e-05,
+      "loss": 0.4408,
       "step": 41425
     },
     {
-      "epoch": 1.46,
-      "learning_rate": 4.1560886272981045e-05,
-      "loss": 0.2798,
+      "epoch": 1.4931343929073413,
+      "grad_norm": 0.1711881160736084,
+      "learning_rate": 4.112689471759316e-05,
+      "loss": 0.3945,
       "step": 41430
     },
     {
-      "epoch": 1.46,
-      "learning_rate": 4.155875213421372e-05,
-      "loss": 0.318,
+      "epoch": 1.493314592568566,
+      "grad_norm": 0.18180076777935028,
+      "learning_rate": 4.112466480755055e-05,
+      "loss": 0.4029,
       "step": 41435
     },
     {
-      "epoch": 1.46,
-      "learning_rate": 4.1556617780441895e-05,
-      "loss": 0.3104,
+      "epoch": 1.4934947922297905,
+      "grad_norm": 0.17946167290210724,
+      "learning_rate": 4.1122434677812506e-05,
+      "loss": 0.4141,
       "step": 41440
     },
     {
-      "epoch": 1.46,
-      "learning_rate": 4.155448321169332e-05,
-      "loss": 0.2767,
+      "epoch": 1.4936749918910153,
+      "grad_norm": 0.17549100518226624,
+      "learning_rate": 4.1120204328409416e-05,
+      "loss": 0.4119,
       "step": 41445
     },
     {
-      "epoch": 1.46,
-      "learning_rate": 4.155234842799568e-05,
-      "loss": 0.3217,
+      "epoch": 1.4938551915522398,
+      "grad_norm": 0.18163146078586578,
+      "learning_rate": 4.1117973759371666e-05,
+      "loss": 0.4345,
       "step": 41450
     },
     {
-      "epoch": 1.46,
-      "learning_rate": 4.1550213429376705e-05,
-      "loss": 0.2916,
+      "epoch": 1.4940353912134645,
+      "grad_norm": 0.1497826874256134,
+      "learning_rate": 4.111574297072963e-05,
+      "loss": 0.3586,
       "step": 41455
     },
     {
-      "epoch": 1.46,
-      "learning_rate": 4.154807821586412e-05,
-      "loss": 0.2939,
+      "epoch": 1.4942155908746892,
+      "grad_norm": 0.17524641752243042,
+      "learning_rate": 4.111351196251373e-05,
+      "loss": 0.3819,
       "step": 41460
     },
     {
-      "epoch": 1.46,
-      "learning_rate": 4.1545942787485636e-05,
-      "loss": 0.3086,
+      "epoch": 1.4943957905359138,
+      "grad_norm": 0.19001320004463196,
+      "learning_rate": 4.1111280734754345e-05,
+      "loss": 0.4198,
       "step": 41465
     },
     {
-      "epoch": 1.46,
-      "learning_rate": 4.1543807144268996e-05,
-      "loss": 0.266,
+      "epoch": 1.4945759901971385,
+      "grad_norm": 0.15616640448570251,
+      "learning_rate": 4.1109049287481874e-05,
+      "loss": 0.4113,
       "step": 41470
     },
     {
-      "epoch": 1.46,
-      "learning_rate": 4.154167128624191e-05,
-      "loss": 0.3067,
+      "epoch": 1.494756189858363,
+      "grad_norm": 0.1617770940065384,
+      "learning_rate": 4.110681762072672e-05,
+      "loss": 0.4383,
       "step": 41475
     },
     {
-      "epoch": 1.46,
-      "learning_rate": 4.153953521343214e-05,
-      "loss": 0.2628,
+      "epoch": 1.4949363895195877,
+      "grad_norm": 0.20571884512901306,
+      "learning_rate": 4.110458573451931e-05,
+      "loss": 0.4229,
       "step": 41480
     },
     {
-      "epoch": 1.46,
-      "learning_rate": 4.15373989258674e-05,
-      "loss": 0.2992,
+      "epoch": 1.4951165891808125,
+      "grad_norm": 0.18498611450195312,
+      "learning_rate": 4.110235362889003e-05,
+      "loss": 0.4011,
       "step": 41485
     },
     {
-      "epoch": 1.46,
-      "learning_rate": 4.153526242357543e-05,
-      "loss": 0.2798,
+      "epoch": 1.495296788842037,
+      "grad_norm": 0.19509319961071014,
+      "learning_rate": 4.1100121303869296e-05,
+      "loss": 0.41,
       "step": 41490
     },
     {
-      "epoch": 1.46,
-      "learning_rate": 4.1533125706583974e-05,
-      "loss": 0.3018,
+      "epoch": 1.4954769885032615,
+      "grad_norm": 0.21900251507759094,
+      "learning_rate": 4.109788875948754e-05,
+      "loss": 0.4099,
       "step": 41495
     },
     {
-      "epoch": 1.46,
-      "learning_rate": 4.153098877492079e-05,
-      "loss": 0.2966,
+      "epoch": 1.4956571881644862,
+      "grad_norm": 0.1650031954050064,
+      "learning_rate": 4.109565599577515e-05,
+      "loss": 0.386,
       "step": 41500
     },
     {
-      "epoch": 1.46,
-      "eval_loss": 0.28131207823753357,
-      "eval_runtime": 10.5489,
-      "eval_samples_per_second": 9.48,
-      "eval_steps_per_second": 9.48,
+      "epoch": 1.4956571881644862,
+      "eval_loss": 0.44493234157562256,
+      "eval_runtime": 3.5367,
+      "eval_samples_per_second": 28.275,
+      "eval_steps_per_second": 7.069,
       "step": 41500
     },
     {
-      "epoch": 1.46,
-      "learning_rate": 4.15288516286136e-05,
-      "loss": 0.2731,
+      "epoch": 1.495837387825711,
+      "grad_norm": 0.19879759848117828,
+      "learning_rate": 4.109342301276257e-05,
+      "loss": 0.4323,
       "step": 41505
     },
     {
-      "epoch": 1.46,
-      "learning_rate": 4.152671426769016e-05,
-      "loss": 0.2762,
+      "epoch": 1.4960175874869355,
+      "grad_norm": 0.22263413667678833,
+      "learning_rate": 4.109118981048022e-05,
+      "loss": 0.4187,
       "step": 41510
     },
     {
-      "epoch": 1.46,
-      "learning_rate": 4.1524576692178235e-05,
-      "loss": 0.2886,
+      "epoch": 1.4961977871481602,
+      "grad_norm": 0.18593116104602814,
+      "learning_rate": 4.1088956388958524e-05,
+      "loss": 0.4388,
       "step": 41515
     },
     {
-      "epoch": 1.46,
-      "learning_rate": 4.1522438902105564e-05,
-      "loss": 0.299,
+      "epoch": 1.4963779868093847,
+      "grad_norm": 0.16802819073200226,
+      "learning_rate": 4.1086722748227903e-05,
+      "loss": 0.4141,
       "step": 41520
     },
     {
-      "epoch": 1.46,
-      "learning_rate": 4.152030089749992e-05,
-      "loss": 0.282,
+      "epoch": 1.4965581864706095,
+      "grad_norm": 0.19732625782489777,
+      "learning_rate": 4.108448888831881e-05,
+      "loss": 0.4034,
       "step": 41525
     },
     {
-      "epoch": 1.46,
-      "learning_rate": 4.1518162678389054e-05,
-      "loss": 0.3056,
+      "epoch": 1.4967383861318342,
+      "grad_norm": 0.15761670470237732,
+      "learning_rate": 4.108225480926167e-05,
+      "loss": 0.4235,
       "step": 41530
     },
     {
-      "epoch": 1.46,
-      "learning_rate": 4.1516024244800734e-05,
-      "loss": 0.2771,
+      "epoch": 1.4969185857930587,
+      "grad_norm": 0.195955291390419,
+      "learning_rate": 4.108002051108691e-05,
+      "loss": 0.4061,
       "step": 41535
     },
     {
-      "epoch": 1.46,
-      "learning_rate": 4.151388559676272e-05,
-      "loss": 0.2835,
+      "epoch": 1.4970987854542832,
+      "grad_norm": 0.19495412707328796,
+      "learning_rate": 4.107778599382499e-05,
+      "loss": 0.4036,
       "step": 41540
     },
     {
-      "epoch": 1.46,
-      "learning_rate": 4.1511746734302786e-05,
-      "loss": 0.2921,
+      "epoch": 1.497278985115508,
+      "grad_norm": 0.2070777267217636,
+      "learning_rate": 4.1075551257506354e-05,
+      "loss": 0.4182,
       "step": 41545
     },
     {
-      "epoch": 1.46,
-      "learning_rate": 4.1509607657448704e-05,
-      "loss": 0.2958,
+      "epoch": 1.4974591847767327,
+      "grad_norm": 0.1688157469034195,
+      "learning_rate": 4.1073316302161435e-05,
+      "loss": 0.4271,
       "step": 41550
     },
     {
-      "epoch": 1.46,
-      "learning_rate": 4.150746836622824e-05,
-      "loss": 0.2752,
+      "epoch": 1.4976393844379572,
+      "grad_norm": 0.15762291848659515,
+      "learning_rate": 4.1071081127820696e-05,
+      "loss": 0.4158,
       "step": 41555
     },
     {
-      "epoch": 1.46,
-      "learning_rate": 4.150532886066918e-05,
-      "loss": 0.3075,
+      "epoch": 1.497819584099182,
+      "grad_norm": 0.19532495737075806,
+      "learning_rate": 4.1068845734514593e-05,
+      "loss": 0.4508,
       "step": 41560
     },
     {
-      "epoch": 1.46,
-      "learning_rate": 4.1503189140799315e-05,
-      "loss": 0.2905,
+      "epoch": 1.4979997837604064,
+      "grad_norm": 0.15557561814785004,
+      "learning_rate": 4.1066610122273575e-05,
+      "loss": 0.4036,
       "step": 41565
     },
     {
-      "epoch": 1.46,
-      "learning_rate": 4.15010492066464e-05,
-      "loss": 0.298,
+      "epoch": 1.4981799834216312,
+      "grad_norm": 0.18481487035751343,
+      "learning_rate": 4.1064374291128106e-05,
+      "loss": 0.4475,
       "step": 41570
     },
     {
-      "epoch": 1.46,
-      "learning_rate": 4.1498909058238246e-05,
-      "loss": 0.2838,
+      "epoch": 1.498360183082856,
+      "grad_norm": 0.17046202719211578,
+      "learning_rate": 4.1062138241108645e-05,
+      "loss": 0.4168,
       "step": 41575
     },
     {
-      "epoch": 1.46,
-      "learning_rate": 4.149676869560262e-05,
-      "loss": 0.2928,
+      "epoch": 1.4985403827440804,
+      "grad_norm": 0.18078990280628204,
+      "learning_rate": 4.105990197224566e-05,
+      "loss": 0.4325,
       "step": 41580
     },
     {
-      "epoch": 1.46,
-      "learning_rate": 4.149462811876733e-05,
-      "loss": 0.2737,
+      "epoch": 1.4987205824053051,
+      "grad_norm": 0.1715507060289383,
+      "learning_rate": 4.105766548456962e-05,
+      "loss": 0.422,
       "step": 41585
     },
     {
-      "epoch": 1.46,
-      "learning_rate": 4.1492487327760165e-05,
-      "loss": 0.2936,
+      "epoch": 1.4989007820665297,
+      "grad_norm": 0.17720800638198853,
+      "learning_rate": 4.1055428778111004e-05,
+      "loss": 0.4115,
       "step": 41590
     },
     {
-      "epoch": 1.46,
-      "learning_rate": 4.149034632260892e-05,
-      "loss": 0.2766,
+      "epoch": 1.4990809817277544,
+      "grad_norm": 0.16280311346054077,
+      "learning_rate": 4.105319185290027e-05,
+      "loss": 0.4068,
       "step": 41595
     },
     {
-      "epoch": 1.46,
-      "learning_rate": 4.1488205103341395e-05,
-      "loss": 0.279,
+      "epoch": 1.4992611813889791,
+      "grad_norm": 0.1896113157272339,
+      "learning_rate": 4.10509547089679e-05,
+      "loss": 0.4245,
       "step": 41600
     },
     {
-      "epoch": 1.46,
-      "learning_rate": 4.14860636699854e-05,
-      "loss": 0.2808,
+      "epoch": 1.4994413810502036,
+      "grad_norm": 0.17678354680538177,
+      "learning_rate": 4.10487173463444e-05,
+      "loss": 0.4236,
       "step": 41605
     },
     {
-      "epoch": 1.46,
-      "learning_rate": 4.148392202256872e-05,
-      "loss": 0.2937,
+      "epoch": 1.4996215807114281,
+      "grad_norm": 0.15229156613349915,
+      "learning_rate": 4.104647976506022e-05,
+      "loss": 0.3917,
       "step": 41610
     },
     {
-      "epoch": 1.46,
-      "learning_rate": 4.148178016111919e-05,
-      "loss": 0.2873,
+      "epoch": 1.4998017803726529,
+      "grad_norm": 0.2196597158908844,
+      "learning_rate": 4.104424196514586e-05,
+      "loss": 0.4439,
       "step": 41615
     },
     {
-      "epoch": 1.46,
-      "learning_rate": 4.147963808566459e-05,
-      "loss": 0.2778,
+      "epoch": 1.4999819800338776,
+      "grad_norm": 0.18850156664848328,
+      "learning_rate": 4.104200394663181e-05,
+      "loss": 0.4067,
       "step": 41620
     },
     {
-      "epoch": 1.46,
-      "learning_rate": 4.147749579623276e-05,
-      "loss": 0.2742,
+      "epoch": 1.5001621796951021,
+      "grad_norm": 0.19405673444271088,
+      "learning_rate": 4.103976570954856e-05,
+      "loss": 0.4496,
       "step": 41625
     },
     {
-      "epoch": 1.46,
-      "learning_rate": 4.14753532928515e-05,
-      "loss": 0.296,
+      "epoch": 1.5003423793563269,
+      "grad_norm": 0.20671923458576202,
+      "learning_rate": 4.103752725392661e-05,
+      "loss": 0.4071,
       "step": 41630
     },
     {
-      "epoch": 1.46,
-      "learning_rate": 4.1473210575548644e-05,
-      "loss": 0.2929,
+      "epoch": 1.5005225790175514,
+      "grad_norm": 0.2034417986869812,
+      "learning_rate": 4.103528857979646e-05,
+      "loss": 0.3851,
       "step": 41635
     },
     {
-      "epoch": 1.47,
-      "learning_rate": 4.1471067644352e-05,
-      "loss": 0.2728,
+      "epoch": 1.500702778678776,
+      "grad_norm": 0.1623309999704361,
+      "learning_rate": 4.10330496871886e-05,
+      "loss": 0.4161,
       "step": 41640
     },
     {
-      "epoch": 1.47,
-      "learning_rate": 4.1468924499289396e-05,
-      "loss": 0.2815,
+      "epoch": 1.5008829783400008,
+      "grad_norm": 0.1519644856452942,
+      "learning_rate": 4.103081057613355e-05,
+      "loss": 0.4051,
       "step": 41645
     },
     {
-      "epoch": 1.47,
-      "learning_rate": 4.1466781140388666e-05,
-      "loss": 0.3016,
+      "epoch": 1.5010631780012254,
+      "grad_norm": 0.14681312441825867,
+      "learning_rate": 4.1028571246661804e-05,
+      "loss": 0.4235,
       "step": 41650
     },
     {
-      "epoch": 1.47,
-      "learning_rate": 4.1464637567677636e-05,
-      "loss": 0.2837,
+      "epoch": 1.5012433776624499,
+      "grad_norm": 0.189476877450943,
+      "learning_rate": 4.102633169880388e-05,
+      "loss": 0.3896,
       "step": 41655
     },
     {
-      "epoch": 1.47,
-      "learning_rate": 4.146249378118413e-05,
-      "loss": 0.28,
+      "epoch": 1.5014235773236746,
+      "grad_norm": 0.17130400240421295,
+      "learning_rate": 4.102409193259029e-05,
+      "loss": 0.4145,
       "step": 41660
     },
     {
-      "epoch": 1.47,
-      "learning_rate": 4.1460349780936e-05,
-      "loss": 0.2909,
+      "epoch": 1.5016037769848993,
+      "grad_norm": 0.16844402253627777,
+      "learning_rate": 4.102185194805155e-05,
+      "loss": 0.445,
       "step": 41665
     },
     {
-      "epoch": 1.47,
-      "learning_rate": 4.145820556696107e-05,
-      "loss": 0.313,
+      "epoch": 1.501783976646124,
+      "grad_norm": 0.17054221034049988,
+      "learning_rate": 4.101961174521818e-05,
+      "loss": 0.4269,
       "step": 41670
     },
     {
-      "epoch": 1.47,
-      "learning_rate": 4.1456061139287186e-05,
-      "loss": 0.2919,
+      "epoch": 1.5019641763073486,
+      "grad_norm": 0.1802477240562439,
+      "learning_rate": 4.101737132412069e-05,
+      "loss": 0.4183,
       "step": 41675
     },
     {
-      "epoch": 1.47,
-      "learning_rate": 4.14539164979422e-05,
-      "loss": 0.2965,
+      "epoch": 1.502144375968573,
+      "grad_norm": 0.20035143196582794,
+      "learning_rate": 4.101513068478963e-05,
+      "loss": 0.4227,
       "step": 41680
     },
     {
-      "epoch": 1.47,
-      "learning_rate": 4.145177164295395e-05,
-      "loss": 0.2947,
+      "epoch": 1.5023245756297978,
+      "grad_norm": 0.165787011384964,
+      "learning_rate": 4.101288982725551e-05,
+      "loss": 0.4357,
       "step": 41685
     },
     {
-      "epoch": 1.47,
-      "learning_rate": 4.144962657435028e-05,
-      "loss": 0.2752,
+      "epoch": 1.5025047752910226,
+      "grad_norm": 0.1805458515882492,
+      "learning_rate": 4.1010648751548876e-05,
+      "loss": 0.4115,
       "step": 41690
     },
     {
-      "epoch": 1.47,
-      "learning_rate": 4.144748129215905e-05,
-      "loss": 0.2919,
+      "epoch": 1.502684974952247,
+      "grad_norm": 0.17907939851284027,
+      "learning_rate": 4.1008407457700234e-05,
+      "loss": 0.3908,
       "step": 41695
     },
     {
-      "epoch": 1.47,
-      "learning_rate": 4.144533579640812e-05,
-      "loss": 0.2917,
+      "epoch": 1.5028651746134716,
+      "grad_norm": 0.24330779910087585,
+      "learning_rate": 4.1006165945740154e-05,
+      "loss": 0.4361,
       "step": 41700
     },
     {
-      "epoch": 1.47,
-      "learning_rate": 4.1443190087125344e-05,
-      "loss": 0.2771,
+      "epoch": 1.5030453742746963,
+      "grad_norm": 0.21545155346393585,
+      "learning_rate": 4.100392421569916e-05,
+      "loss": 0.4375,
       "step": 41705
     },
     {
-      "epoch": 1.47,
-      "learning_rate": 4.144104416433858e-05,
-      "loss": 0.3088,
+      "epoch": 1.503225573935921,
+      "grad_norm": 0.18745043873786926,
+      "learning_rate": 4.1001682267607796e-05,
+      "loss": 0.3854,
       "step": 41710
     },
     {
-      "epoch": 1.47,
-      "learning_rate": 4.143889802807569e-05,
-      "loss": 0.3076,
+      "epoch": 1.5034057735971458,
+      "grad_norm": 0.16889239847660065,
+      "learning_rate": 4.0999440101496606e-05,
+      "loss": 0.3463,
       "step": 41715
     },
     {
-      "epoch": 1.47,
-      "learning_rate": 4.143675167836455e-05,
-      "loss": 0.2778,
+      "epoch": 1.5035859732583703,
+      "grad_norm": 0.18077971041202545,
+      "learning_rate": 4.099719771739614e-05,
+      "loss": 0.3698,
       "step": 41720
     },
     {
-      "epoch": 1.47,
-      "learning_rate": 4.143460511523302e-05,
-      "loss": 0.276,
+      "epoch": 1.5037661729195948,
+      "grad_norm": 0.20077045261859894,
+      "learning_rate": 4.099495511533696e-05,
+      "loss": 0.4117,
       "step": 41725
     },
     {
-      "epoch": 1.47,
-      "learning_rate": 4.1432458338708965e-05,
-      "loss": 0.2951,
+      "epoch": 1.5039463725808195,
+      "grad_norm": 0.1883951872587204,
+      "learning_rate": 4.099271229534961e-05,
+      "loss": 0.4191,
       "step": 41730
     },
     {
-      "epoch": 1.47,
-      "learning_rate": 4.143031134882027e-05,
-      "loss": 0.2916,
+      "epoch": 1.5041265722420443,
+      "grad_norm": 0.18416355550289154,
+      "learning_rate": 4.0990469257464645e-05,
+      "loss": 0.3896,
       "step": 41735
     },
     {
-      "epoch": 1.47,
-      "learning_rate": 4.142816414559481e-05,
-      "loss": 0.2642,
+      "epoch": 1.5043067719032688,
+      "grad_norm": 0.20206254720687866,
+      "learning_rate": 4.0988226001712646e-05,
+      "loss": 0.4027,
       "step": 41740
     },
     {
-      "epoch": 1.47,
-      "learning_rate": 4.142601672906047e-05,
-      "loss": 0.2762,
+      "epoch": 1.5044869715644935,
+      "grad_norm": 0.18239402770996094,
+      "learning_rate": 4.0985982528124156e-05,
+      "loss": 0.3764,
       "step": 41745
     },
     {
-      "epoch": 1.47,
-      "learning_rate": 4.142386909924512e-05,
-      "loss": 0.3077,
+      "epoch": 1.504667171225718,
+      "grad_norm": 0.1996680647134781,
+      "learning_rate": 4.098373883672974e-05,
+      "loss": 0.404,
       "step": 41750
     },
     {
-      "epoch": 1.47,
-      "learning_rate": 4.142172125617666e-05,
-      "loss": 0.2859,
+      "epoch": 1.5048473708869428,
+      "grad_norm": 0.1679215282201767,
+      "learning_rate": 4.0981494927559994e-05,
+      "loss": 0.4324,
       "step": 41755
     },
     {
-      "epoch": 1.47,
-      "learning_rate": 4.141957319988297e-05,
-      "loss": 0.2824,
+      "epoch": 1.5050275705481675,
+      "grad_norm": 0.19576683640480042,
+      "learning_rate": 4.0979250800645465e-05,
+      "loss": 0.4211,
       "step": 41760
     },
     {
-      "epoch": 1.47,
-      "learning_rate": 4.1417424930391946e-05,
-      "loss": 0.2843,
+      "epoch": 1.505207770209392,
+      "grad_norm": 0.20221199095249176,
+      "learning_rate": 4.097700645601673e-05,
+      "loss": 0.3923,
       "step": 41765
     },
     {
-      "epoch": 1.47,
-      "learning_rate": 4.1415276447731476e-05,
-      "loss": 0.3122,
+      "epoch": 1.5053879698706165,
+      "grad_norm": 0.14687031507492065,
+      "learning_rate": 4.097476189370439e-05,
+      "loss": 0.3818,
       "step": 41770
     },
     {
-      "epoch": 1.47,
-      "learning_rate": 4.141312775192946e-05,
-      "loss": 0.2629,
+      "epoch": 1.5055681695318412,
+      "grad_norm": 0.19657109677791595,
+      "learning_rate": 4.097251711373901e-05,
+      "loss": 0.417,
       "step": 41775
     },
     {
-      "epoch": 1.47,
-      "learning_rate": 4.141097884301379e-05,
-      "loss": 0.2667,
+      "epoch": 1.505748369193066,
+      "grad_norm": 0.2037065178155899,
+      "learning_rate": 4.097027211615117e-05,
+      "loss": 0.3996,
       "step": 41780
     },
     {
-      "epoch": 1.47,
-      "learning_rate": 4.140882972101239e-05,
-      "loss": 0.27,
+      "epoch": 1.5059285688542907,
+      "grad_norm": 0.22538067400455475,
+      "learning_rate": 4.096802690097146e-05,
+      "loss": 0.4108,
       "step": 41785
     },
     {
-      "epoch": 1.47,
-      "learning_rate": 4.1406680385953134e-05,
-      "loss": 0.304,
+      "epoch": 1.5061087685155152,
+      "grad_norm": 0.2207726240158081,
+      "learning_rate": 4.096578146823048e-05,
+      "loss": 0.4288,
       "step": 41790
     },
     {
-      "epoch": 1.47,
-      "learning_rate": 4.140453083786396e-05,
-      "loss": 0.3035,
+      "epoch": 1.5062889681767397,
+      "grad_norm": 0.16942410171031952,
+      "learning_rate": 4.096353581795882e-05,
+      "loss": 0.4209,
       "step": 41795
     },
     {
-      "epoch": 1.47,
-      "learning_rate": 4.140238107677276e-05,
-      "loss": 0.2816,
+      "epoch": 1.5064691678379645,
+      "grad_norm": 0.22281381487846375,
+      "learning_rate": 4.096128995018707e-05,
+      "loss": 0.4037,
       "step": 41800
     },
     {
-      "epoch": 1.47,
-      "learning_rate": 4.140023110270744e-05,
-      "loss": 0.2717,
+      "epoch": 1.5066493674991892,
+      "grad_norm": 0.19026844203472137,
+      "learning_rate": 4.095904386494585e-05,
+      "loss": 0.407,
       "step": 41805
     },
     {
-      "epoch": 1.47,
-      "learning_rate": 4.1398080915695936e-05,
-      "loss": 0.2917,
+      "epoch": 1.5068295671604137,
+      "grad_norm": 0.23003491759300232,
+      "learning_rate": 4.095679756226573e-05,
+      "loss": 0.3969,
       "step": 41810
     },
     {
-      "epoch": 1.47,
-      "learning_rate": 4.139593051576616e-05,
-      "loss": 0.3025,
+      "epoch": 1.5070097668216382,
+      "grad_norm": 0.19548535346984863,
+      "learning_rate": 4.0954551042177337e-05,
+      "loss": 0.403,
       "step": 41815
     },
     {
-      "epoch": 1.47,
-      "learning_rate": 4.1393779902946025e-05,
-      "loss": 0.2898,
+      "epoch": 1.507189966482863,
+      "grad_norm": 0.19245415925979614,
+      "learning_rate": 4.0952304304711275e-05,
+      "loss": 0.398,
       "step": 41820
     },
     {
-      "epoch": 1.47,
-      "learning_rate": 4.139162907726347e-05,
-      "loss": 0.2947,
+      "epoch": 1.5073701661440877,
+      "grad_norm": 0.18857969343662262,
+      "learning_rate": 4.0950057349898165e-05,
+      "loss": 0.4223,
       "step": 41825
     },
     {
-      "epoch": 1.47,
-      "learning_rate": 4.138947803874641e-05,
-      "loss": 0.3029,
+      "epoch": 1.5075503658053124,
+      "grad_norm": 0.19236567616462708,
+      "learning_rate": 4.0947810177768595e-05,
+      "loss": 0.4026,
       "step": 41830
     },
     {
-      "epoch": 1.47,
-      "learning_rate": 4.1387326787422774e-05,
-      "loss": 0.3071,
+      "epoch": 1.507730565466537,
+      "grad_norm": 0.14722810685634613,
+      "learning_rate": 4.094556278835321e-05,
+      "loss": 0.3917,
       "step": 41835
     },
     {
-      "epoch": 1.47,
-      "learning_rate": 4.138517532332051e-05,
-      "loss": 0.3212,
+      "epoch": 1.5079107651277615,
+      "grad_norm": 0.18496693670749664,
+      "learning_rate": 4.094331518168262e-05,
+      "loss": 0.4228,
       "step": 41840
     },
     {
-      "epoch": 1.47,
-      "learning_rate": 4.1383023646467536e-05,
-      "loss": 0.2873,
+      "epoch": 1.5080909647889862,
+      "grad_norm": 0.14098331332206726,
+      "learning_rate": 4.0941067357787445e-05,
+      "loss": 0.4242,
       "step": 41845
     },
     {
-      "epoch": 1.47,
-      "learning_rate": 4.138087175689179e-05,
-      "loss": 0.2885,
+      "epoch": 1.508271164450211,
+      "grad_norm": 0.16862405836582184,
+      "learning_rate": 4.093881931669831e-05,
+      "loss": 0.4082,
       "step": 41850
     },
     {
-      "epoch": 1.47,
-      "learning_rate": 4.137871965462124e-05,
-      "loss": 0.2966,
+      "epoch": 1.5084513641114354,
+      "grad_norm": 0.18131491541862488,
+      "learning_rate": 4.093657105844585e-05,
+      "loss": 0.4356,
       "step": 41855
     },
     {
-      "epoch": 1.47,
-      "learning_rate": 4.137656733968379e-05,
-      "loss": 0.2855,
+      "epoch": 1.50863156377266,
+      "grad_norm": 0.17331118881702423,
+      "learning_rate": 4.093432258306069e-05,
+      "loss": 0.4143,
       "step": 41860
     },
     {
-      "epoch": 1.47,
-      "learning_rate": 4.137441481210741e-05,
-      "loss": 0.2517,
+      "epoch": 1.5088117634338847,
+      "grad_norm": 0.17720840871334076,
+      "learning_rate": 4.093207389057348e-05,
+      "loss": 0.3942,
       "step": 41865
     },
     {
-      "epoch": 1.47,
-      "learning_rate": 4.137226207192005e-05,
-      "loss": 0.2752,
+      "epoch": 1.5089919630951094,
+      "grad_norm": 0.18100488185882568,
+      "learning_rate": 4.0929824981014845e-05,
+      "loss": 0.4303,
       "step": 41870
     },
     {
-      "epoch": 1.47,
-      "learning_rate": 4.137010911914965e-05,
-      "loss": 0.2747,
+      "epoch": 1.5091721627563341,
+      "grad_norm": 0.1756538450717926,
+      "learning_rate": 4.0927575854415425e-05,
+      "loss": 0.418,
       "step": 41875
     },
     {
-      "epoch": 1.47,
-      "learning_rate": 4.1367955953824176e-05,
-      "loss": 0.2791,
+      "epoch": 1.5093523624175587,
+      "grad_norm": 0.2303672730922699,
+      "learning_rate": 4.092532651080587e-05,
+      "loss": 0.4329,
       "step": 41880
     },
     {
-      "epoch": 1.47,
-      "learning_rate": 4.136580257597158e-05,
-      "loss": 0.2903,
+      "epoch": 1.5095325620787832,
+      "grad_norm": 0.15763244032859802,
+      "learning_rate": 4.092307695021682e-05,
+      "loss": 0.4029,
       "step": 41885
     },
     {
-      "epoch": 1.47,
-      "learning_rate": 4.136364898561982e-05,
-      "loss": 0.2824,
+      "epoch": 1.509712761740008,
+      "grad_norm": 0.16717827320098877,
+      "learning_rate": 4.092082717267893e-05,
+      "loss": 0.4265,
       "step": 41890
     },
     {
-      "epoch": 1.47,
-      "learning_rate": 4.1361495182796866e-05,
-      "loss": 0.2816,
+      "epoch": 1.5098929614012326,
+      "grad_norm": 0.2047724723815918,
+      "learning_rate": 4.091857717822286e-05,
+      "loss": 0.4463,
       "step": 41895
     },
     {
-      "epoch": 1.47,
-      "learning_rate": 4.1359341167530684e-05,
-      "loss": 0.2877,
+      "epoch": 1.5100731610624571,
+      "grad_norm": 0.18995271623134613,
+      "learning_rate": 4.091632696687925e-05,
+      "loss": 0.4023,
       "step": 41900
     },
     {
-      "epoch": 1.47,
-      "learning_rate": 4.1357186939849225e-05,
-      "loss": 0.2658,
+      "epoch": 1.5102533607236819,
+      "grad_norm": 0.19766134023666382,
+      "learning_rate": 4.091407653867877e-05,
+      "loss": 0.4182,
       "step": 41905
     },
     {
-      "epoch": 1.47,
-      "learning_rate": 4.135503249978049e-05,
-      "loss": 0.2948,
+      "epoch": 1.5104335603849064,
+      "grad_norm": 0.21319149434566498,
+      "learning_rate": 4.091182589365208e-05,
+      "loss": 0.4199,
       "step": 41910
     },
     {
-      "epoch": 1.47,
-      "learning_rate": 4.135287784735242e-05,
-      "loss": 0.3228,
+      "epoch": 1.5106137600461311,
+      "grad_norm": 0.17943358421325684,
+      "learning_rate": 4.090957503182984e-05,
+      "loss": 0.3911,
       "step": 41915
     },
     {
-      "epoch": 1.47,
-      "learning_rate": 4.135072298259301e-05,
-      "loss": 0.2755,
+      "epoch": 1.5107939597073559,
+      "grad_norm": 0.196610227227211,
+      "learning_rate": 4.0907323953242724e-05,
+      "loss": 0.4463,
       "step": 41920
     },
     {
-      "epoch": 1.48,
-      "learning_rate": 4.134856790553025e-05,
-      "loss": 0.2892,
+      "epoch": 1.5109741593685804,
+      "grad_norm": 0.18429423868656158,
+      "learning_rate": 4.0905072657921396e-05,
+      "loss": 0.4321,
       "step": 41925
     },
     {
-      "epoch": 1.48,
-      "learning_rate": 4.13464126161921e-05,
-      "loss": 0.2966,
+      "epoch": 1.5111543590298049,
+      "grad_norm": 0.16434848308563232,
+      "learning_rate": 4.090282114589653e-05,
+      "loss": 0.4202,
       "step": 41930
     },
     {
-      "epoch": 1.48,
-      "learning_rate": 4.134425711460656e-05,
-      "loss": 0.2837,
+      "epoch": 1.5113345586910296,
+      "grad_norm": 0.15789394080638885,
+      "learning_rate": 4.090056941719881e-05,
+      "loss": 0.3869,
       "step": 41935
     },
     {
-      "epoch": 1.48,
-      "learning_rate": 4.134210140080161e-05,
-      "loss": 0.2627,
+      "epoch": 1.5115147583522544,
+      "grad_norm": 0.16550783812999725,
+      "learning_rate": 4.08983174718589e-05,
+      "loss": 0.4141,
       "step": 41940
     },
     {
-      "epoch": 1.48,
-      "learning_rate": 4.1339945474805244e-05,
-      "loss": 0.3242,
+      "epoch": 1.511694958013479,
+      "grad_norm": 0.159341499209404,
+      "learning_rate": 4.08960653099075e-05,
+      "loss": 0.4122,
       "step": 41945
     },
     {
-      "epoch": 1.48,
-      "learning_rate": 4.133778933664546e-05,
-      "loss": 0.2873,
+      "epoch": 1.5118751576747036,
+      "grad_norm": 0.187036395072937,
+      "learning_rate": 4.089381293137529e-05,
+      "loss": 0.4435,
       "step": 41950
     },
     {
-      "epoch": 1.48,
-      "learning_rate": 4.1335632986350235e-05,
-      "loss": 0.29,
+      "epoch": 1.512055357335928,
+      "grad_norm": 0.18756411969661713,
+      "learning_rate": 4.0891560336292955e-05,
+      "loss": 0.4126,
       "step": 41955
     },
     {
-      "epoch": 1.48,
-      "learning_rate": 4.1333476423947594e-05,
-      "loss": 0.2837,
+      "epoch": 1.5122355569971528,
+      "grad_norm": 0.18429414927959442,
+      "learning_rate": 4.088930752469119e-05,
+      "loss": 0.4426,
       "step": 41960
     },
     {
-      "epoch": 1.48,
-      "learning_rate": 4.1331319649465515e-05,
-      "loss": 0.282,
+      "epoch": 1.5124157566583776,
+      "grad_norm": 0.16013193130493164,
+      "learning_rate": 4.0887054496600676e-05,
+      "loss": 0.3959,
       "step": 41965
     },
     {
-      "epoch": 1.48,
-      "learning_rate": 4.132916266293202e-05,
-      "loss": 0.3029,
+      "epoch": 1.512595956319602,
+      "grad_norm": 0.20077770948410034,
+      "learning_rate": 4.088480125205213e-05,
+      "loss": 0.3921,
       "step": 41970
     },
     {
-      "epoch": 1.48,
-      "learning_rate": 4.1327005464375106e-05,
-      "loss": 0.283,
+      "epoch": 1.5127761559808266,
+      "grad_norm": 0.20495128631591797,
+      "learning_rate": 4.088254779107624e-05,
+      "loss": 0.4392,
       "step": 41975
     },
     {
-      "epoch": 1.48,
-      "learning_rate": 4.13248480538228e-05,
-      "loss": 0.3077,
+      "epoch": 1.5129563556420513,
+      "grad_norm": 0.16244658827781677,
+      "learning_rate": 4.088074486648847e-05,
+      "loss": 0.4346,
       "step": 41980
     },
     {
-      "epoch": 1.48,
-      "learning_rate": 4.132269043130308e-05,
-      "loss": 0.286,
+      "epoch": 1.513136555303276,
+      "grad_norm": 0.2103343904018402,
+      "learning_rate": 4.0878491016020734e-05,
+      "loss": 0.4033,
       "step": 41985
     },
     {
-      "epoch": 1.48,
-      "learning_rate": 4.1320532596843985e-05,
-      "loss": 0.2866,
+      "epoch": 1.5133167549645008,
+      "grad_norm": 0.17164449393749237,
+      "learning_rate": 4.087623694921164e-05,
+      "loss": 0.4303,
       "step": 41990
     },
     {
-      "epoch": 1.48,
-      "learning_rate": 4.1318374550473536e-05,
-      "loss": 0.3032,
+      "epoch": 1.5134969546257253,
+      "grad_norm": 0.1736312210559845,
+      "learning_rate": 4.087398266609188e-05,
+      "loss": 0.4023,
       "step": 41995
     },
     {
-      "epoch": 1.48,
-      "learning_rate": 4.1316216292219746e-05,
-      "loss": 0.3057,
+      "epoch": 1.5136771542869498,
+      "grad_norm": 0.17446556687355042,
+      "learning_rate": 4.0871728166692195e-05,
+      "loss": 0.4104,
       "step": 42000
     },
     {
-      "epoch": 1.48,
-      "eval_loss": 0.2798173427581787,
-      "eval_runtime": 10.5336,
-      "eval_samples_per_second": 9.493,
-      "eval_steps_per_second": 9.493,
+      "epoch": 1.5136771542869498,
+      "eval_loss": 0.4447964131832123,
+      "eval_runtime": 3.5328,
+      "eval_samples_per_second": 28.306,
+      "eval_steps_per_second": 7.076,
       "step": 42000
     },
     {
-      "epoch": 1.48,
-      "learning_rate": 4.131405782211064e-05,
-      "loss": 0.2677,
+      "epoch": 1.5138573539481746,
+      "grad_norm": 0.16829872131347656,
+      "learning_rate": 4.0869473451043274e-05,
+      "loss": 0.4072,
       "step": 42005
     },
     {
-      "epoch": 1.48,
-      "learning_rate": 4.1311899140174246e-05,
-      "loss": 0.287,
+      "epoch": 1.5140375536093993,
+      "grad_norm": 0.1769193410873413,
+      "learning_rate": 4.086721851917585e-05,
+      "loss": 0.3981,
       "step": 42010
     },
     {
-      "epoch": 1.48,
-      "learning_rate": 4.1309740246438596e-05,
-      "loss": 0.2707,
+      "epoch": 1.5142177532706238,
+      "grad_norm": 0.13675101101398468,
+      "learning_rate": 4.086496337112064e-05,
+      "loss": 0.3936,
       "step": 42015
     },
     {
-      "epoch": 1.48,
-      "learning_rate": 4.1307581140931705e-05,
-      "loss": 0.2812,
+      "epoch": 1.5143979529318485,
+      "grad_norm": 0.16778729856014252,
+      "learning_rate": 4.0862708006908393e-05,
+      "loss": 0.3619,
       "step": 42020
     },
     {
-      "epoch": 1.48,
-      "learning_rate": 4.1305421823681635e-05,
-      "loss": 0.3009,
+      "epoch": 1.514578152593073,
+      "grad_norm": 0.21226206421852112,
+      "learning_rate": 4.086045242656982e-05,
+      "loss": 0.4095,
       "step": 42025
     },
     {
-      "epoch": 1.48,
-      "learning_rate": 4.13032622947164e-05,
-      "loss": 0.2792,
+      "epoch": 1.5147583522542978,
+      "grad_norm": 0.16903939843177795,
+      "learning_rate": 4.0858196630135636e-05,
+      "loss": 0.4046,
       "step": 42030
     },
     {
-      "epoch": 1.48,
-      "learning_rate": 4.1301102554064046e-05,
-      "loss": 0.2933,
+      "epoch": 1.5149385519155225,
+      "grad_norm": 0.18711115419864655,
+      "learning_rate": 4.0855940617636604e-05,
+      "loss": 0.4402,
       "step": 42035
     },
     {
-      "epoch": 1.48,
-      "learning_rate": 4.129894260175262e-05,
-      "loss": 0.2755,
+      "epoch": 1.515118751576747,
+      "grad_norm": 0.22342517971992493,
+      "learning_rate": 4.0853684389103444e-05,
+      "loss": 0.4173,
       "step": 42040
     },
     {
-      "epoch": 1.48,
-      "learning_rate": 4.1296782437810166e-05,
-      "loss": 0.2816,
+      "epoch": 1.5152989512379715,
+      "grad_norm": 0.2237049639225006,
+      "learning_rate": 4.0851427944566915e-05,
+      "loss": 0.441,
       "step": 42045
     },
     {
-      "epoch": 1.48,
-      "learning_rate": 4.129462206226473e-05,
-      "loss": 0.2795,
+      "epoch": 1.5154791508991963,
+      "grad_norm": 0.1923510730266571,
+      "learning_rate": 4.084917128405774e-05,
+      "loss": 0.4073,
       "step": 42050
     },
     {
-      "epoch": 1.48,
-      "learning_rate": 4.129246147514437e-05,
-      "loss": 0.2853,
+      "epoch": 1.515659350560421,
+      "grad_norm": 0.14314790070056915,
+      "learning_rate": 4.0846914407606676e-05,
+      "loss": 0.3933,
       "step": 42055
     },
     {
-      "epoch": 1.48,
-      "learning_rate": 4.129030067647713e-05,
-      "loss": 0.2732,
+      "epoch": 1.5158395502216457,
+      "grad_norm": 0.18931898474693298,
+      "learning_rate": 4.0844657315244475e-05,
+      "loss": 0.3878,
       "step": 42060
     },
     {
-      "epoch": 1.48,
-      "learning_rate": 4.128813966629108e-05,
-      "loss": 0.2671,
+      "epoch": 1.5160197498828702,
+      "grad_norm": 0.1530705988407135,
+      "learning_rate": 4.084240000700188e-05,
+      "loss": 0.4434,
       "step": 42065
     },
     {
-      "epoch": 1.48,
-      "learning_rate": 4.1285978444614266e-05,
-      "loss": 0.3028,
+      "epoch": 1.5161999495440948,
+      "grad_norm": 0.1672196239233017,
+      "learning_rate": 4.084014248290966e-05,
+      "loss": 0.423,
       "step": 42070
     },
     {
-      "epoch": 1.48,
-      "learning_rate": 4.1283817011474755e-05,
-      "loss": 0.2778,
+      "epoch": 1.5163801492053195,
+      "grad_norm": 0.17072351276874542,
+      "learning_rate": 4.083788474299856e-05,
+      "loss": 0.4333,
       "step": 42075
     },
     {
-      "epoch": 1.48,
-      "learning_rate": 4.128165536690061e-05,
-      "loss": 0.2685,
+      "epoch": 1.5165603488665442,
+      "grad_norm": 0.18741382658481598,
+      "learning_rate": 4.083562678729935e-05,
+      "loss": 0.4173,
       "step": 42080
     },
     {
-      "epoch": 1.48,
-      "learning_rate": 4.12794935109199e-05,
-      "loss": 0.289,
+      "epoch": 1.5167405485277687,
+      "grad_norm": 0.19706690311431885,
+      "learning_rate": 4.083336861584278e-05,
+      "loss": 0.4282,
       "step": 42085
     },
     {
-      "epoch": 1.48,
-      "learning_rate": 4.12773314435607e-05,
-      "loss": 0.2809,
+      "epoch": 1.5169207481889933,
+      "grad_norm": 0.1541031002998352,
+      "learning_rate": 4.0831110228659644e-05,
+      "loss": 0.4078,
       "step": 42090
     },
     {
-      "epoch": 1.48,
-      "learning_rate": 4.1275169164851066e-05,
-      "loss": 0.3002,
+      "epoch": 1.517100947850218,
+      "grad_norm": 0.16391873359680176,
+      "learning_rate": 4.082885162578069e-05,
+      "loss": 0.4351,
       "step": 42095
     },
     {
-      "epoch": 1.48,
-      "learning_rate": 4.127300667481909e-05,
-      "loss": 0.296,
+      "epoch": 1.5172811475114427,
+      "grad_norm": 0.1912384182214737,
+      "learning_rate": 4.0826592807236694e-05,
+      "loss": 0.3714,
       "step": 42100
     },
     {
-      "epoch": 1.48,
-      "learning_rate": 4.1270843973492854e-05,
-      "loss": 0.2916,
+      "epoch": 1.5174613471726675,
+      "grad_norm": 0.19222666323184967,
+      "learning_rate": 4.0824333773058434e-05,
+      "loss": 0.4148,
       "step": 42105
     },
     {
-      "epoch": 1.48,
-      "learning_rate": 4.126868106090043e-05,
-      "loss": 0.2953,
+      "epoch": 1.517641546833892,
+      "grad_norm": 0.18687984347343445,
+      "learning_rate": 4.08220745232767e-05,
+      "loss": 0.4279,
       "step": 42110
     },
     {
-      "epoch": 1.48,
-      "learning_rate": 4.12665179370699e-05,
-      "loss": 0.2817,
+      "epoch": 1.5178217464951165,
+      "grad_norm": 0.17974166572093964,
+      "learning_rate": 4.0819815057922254e-05,
+      "loss": 0.4572,
       "step": 42115
     },
     {
-      "epoch": 1.48,
-      "learning_rate": 4.1264354602029364e-05,
-      "loss": 0.2931,
+      "epoch": 1.5180019461563412,
+      "grad_norm": 0.18375541269779205,
+      "learning_rate": 4.0817555377025895e-05,
+      "loss": 0.4348,
       "step": 42120
     },
     {
-      "epoch": 1.48,
-      "learning_rate": 4.126219105580689e-05,
-      "loss": 0.2771,
+      "epoch": 1.518182145817566,
+      "grad_norm": 0.21198663115501404,
+      "learning_rate": 4.0815295480618395e-05,
+      "loss": 0.4156,
       "step": 42125
     },
     {
-      "epoch": 1.48,
-      "learning_rate": 4.1260027298430594e-05,
-      "loss": 0.3055,
+      "epoch": 1.5183623454787905,
+      "grad_norm": 0.18676543235778809,
+      "learning_rate": 4.081303536873058e-05,
+      "loss": 0.4357,
       "step": 42130
     },
     {
-      "epoch": 1.48,
-      "learning_rate": 4.125786332992855e-05,
-      "loss": 0.2824,
+      "epoch": 1.5185425451400152,
+      "grad_norm": 0.19135519862174988,
+      "learning_rate": 4.08107750413932e-05,
+      "loss": 0.3997,
       "step": 42135
     },
     {
-      "epoch": 1.48,
-      "learning_rate": 4.125613200313527e-05,
-      "loss": 0.2557,
+      "epoch": 1.5187227448012397,
+      "grad_norm": 0.16404955089092255,
+      "learning_rate": 4.080851449863708e-05,
+      "loss": 0.3918,
       "step": 42140
     },
     {
-      "epoch": 1.48,
-      "learning_rate": 4.1253967654677706e-05,
-      "loss": 0.3008,
+      "epoch": 1.5189029444624644,
+      "grad_norm": 0.17126797139644623,
+      "learning_rate": 4.080625374049301e-05,
+      "loss": 0.4189,
       "step": 42145
     },
     {
-      "epoch": 1.48,
-      "learning_rate": 4.125180309517308e-05,
-      "loss": 0.2772,
+      "epoch": 1.5190831441236892,
+      "grad_norm": 0.1819930374622345,
+      "learning_rate": 4.0803992766991785e-05,
+      "loss": 0.4446,
       "step": 42150
     },
     {
-      "epoch": 1.48,
-      "learning_rate": 4.124963832464951e-05,
-      "loss": 0.2838,
+      "epoch": 1.5192633437849137,
+      "grad_norm": 0.18088293075561523,
+      "learning_rate": 4.080173157816422e-05,
+      "loss": 0.4154,
       "step": 42155
     },
     {
-      "epoch": 1.48,
-      "learning_rate": 4.124747334313509e-05,
-      "loss": 0.2992,
+      "epoch": 1.5194435434461382,
+      "grad_norm": 0.18512316048145294,
+      "learning_rate": 4.079947017404113e-05,
+      "loss": 0.43,
       "step": 42160
     },
     {
-      "epoch": 1.48,
-      "learning_rate": 4.124530815065794e-05,
-      "loss": 0.3114,
+      "epoch": 1.519623743107363,
+      "grad_norm": 0.17444075644016266,
+      "learning_rate": 4.079720855465331e-05,
+      "loss": 0.4072,
       "step": 42165
     },
     {
-      "epoch": 1.48,
-      "learning_rate": 4.1243142747246164e-05,
-      "loss": 0.3013,
+      "epoch": 1.5198039427685877,
+      "grad_norm": 0.16421203315258026,
+      "learning_rate": 4.079494672003159e-05,
+      "loss": 0.4049,
       "step": 42170
     },
     {
-      "epoch": 1.48,
-      "learning_rate": 4.1240977132927886e-05,
-      "loss": 0.3294,
+      "epoch": 1.5199841424298124,
+      "grad_norm": 0.24568875133991241,
+      "learning_rate": 4.079268467020677e-05,
+      "loss": 0.4327,
       "step": 42175
     },
     {
-      "epoch": 1.48,
-      "learning_rate": 4.123881130773123e-05,
-      "loss": 0.2592,
+      "epoch": 1.520164342091037,
+      "grad_norm": 0.2001422941684723,
+      "learning_rate": 4.079042240520968e-05,
+      "loss": 0.4125,
       "step": 42180
     },
     {
-      "epoch": 1.48,
-      "learning_rate": 4.123664527168431e-05,
-      "loss": 0.302,
+      "epoch": 1.5203445417522614,
+      "grad_norm": 0.16765598952770233,
+      "learning_rate": 4.078815992507115e-05,
+      "loss": 0.3885,
       "step": 42185
     },
     {
-      "epoch": 1.48,
-      "learning_rate": 4.123447902481524e-05,
-      "loss": 0.2871,
+      "epoch": 1.5205247414134861,
+      "grad_norm": 0.20924493670463562,
+      "learning_rate": 4.078589722982199e-05,
+      "loss": 0.4214,
       "step": 42190
     },
     {
-      "epoch": 1.48,
-      "learning_rate": 4.123231256715217e-05,
-      "loss": 0.276,
+      "epoch": 1.5207049410747109,
+      "grad_norm": 0.183701753616333,
+      "learning_rate": 4.078363431949304e-05,
+      "loss": 0.4176,
       "step": 42195
     },
     {
-      "epoch": 1.48,
-      "learning_rate": 4.123014589872321e-05,
-      "loss": 0.2924,
+      "epoch": 1.5208851407359354,
+      "grad_norm": 0.1759697049856186,
+      "learning_rate": 4.0781371194115126e-05,
+      "loss": 0.4188,
       "step": 42200
     },
     {
-      "epoch": 1.48,
-      "learning_rate": 4.122797901955651e-05,
-      "loss": 0.2593,
+      "epoch": 1.52106534039716,
+      "grad_norm": 0.19455984234809875,
+      "learning_rate": 4.077910785371909e-05,
+      "loss": 0.4576,
       "step": 42205
     },
     {
-      "epoch": 1.49,
-      "learning_rate": 4.122581192968019e-05,
-      "loss": 0.276,
+      "epoch": 1.5212455400583846,
+      "grad_norm": 0.21152974665164948,
+      "learning_rate": 4.077684429833576e-05,
+      "loss": 0.4207,
       "step": 42210
     },
     {
-      "epoch": 1.49,
-      "learning_rate": 4.1223644629122395e-05,
-      "loss": 0.2861,
+      "epoch": 1.5214257397196094,
+      "grad_norm": 0.18577638268470764,
+      "learning_rate": 4.0774580527995975e-05,
+      "loss": 0.4593,
       "step": 42215
     },
     {
-      "epoch": 1.49,
-      "learning_rate": 4.122147711791128e-05,
-      "loss": 0.2924,
+      "epoch": 1.521605939380834,
+      "grad_norm": 0.18051621317863464,
+      "learning_rate": 4.0772316542730594e-05,
+      "loss": 0.4016,
       "step": 42220
     },
     {
-      "epoch": 1.49,
-      "learning_rate": 4.1219309396074956e-05,
-      "loss": 0.309,
+      "epoch": 1.5217861390420586,
+      "grad_norm": 0.1438736468553543,
+      "learning_rate": 4.077005234257045e-05,
+      "loss": 0.4366,
       "step": 42225
     },
     {
-      "epoch": 1.49,
-      "learning_rate": 4.12171414636416e-05,
-      "loss": 0.2984,
+      "epoch": 1.5219663387032831,
+      "grad_norm": 0.20258064568042755,
+      "learning_rate": 4.0767787927546394e-05,
+      "loss": 0.4339,
       "step": 42230
     },
     {
-      "epoch": 1.49,
-      "learning_rate": 4.121497332063935e-05,
-      "loss": 0.2866,
+      "epoch": 1.5221465383645079,
+      "grad_norm": 0.19496627151966095,
+      "learning_rate": 4.076552329768929e-05,
+      "loss": 0.4543,
       "step": 42235
     },
     {
-      "epoch": 1.49,
-      "learning_rate": 4.1212804967096355e-05,
-      "loss": 0.284,
+      "epoch": 1.5223267380257326,
+      "grad_norm": 0.1588614284992218,
+      "learning_rate": 4.0763258453029976e-05,
+      "loss": 0.4298,
       "step": 42240
     },
     {
-      "epoch": 1.49,
-      "learning_rate": 4.121063640304077e-05,
-      "loss": 0.3037,
+      "epoch": 1.522506937686957,
+      "grad_norm": 0.1556130349636078,
+      "learning_rate": 4.076099339359931e-05,
+      "loss": 0.4272,
       "step": 42245
     },
     {
-      "epoch": 1.49,
-      "learning_rate": 4.120846762850076e-05,
-      "loss": 0.3191,
+      "epoch": 1.5226871373481818,
+      "grad_norm": 0.20325186848640442,
+      "learning_rate": 4.0758728119428166e-05,
+      "loss": 0.4189,
       "step": 42250
     },
     {
-      "epoch": 1.49,
-      "learning_rate": 4.120629864350447e-05,
-      "loss": 0.2827,
+      "epoch": 1.5228673370094064,
+      "grad_norm": 0.20584464073181152,
+      "learning_rate": 4.075646263054741e-05,
+      "loss": 0.4432,
       "step": 42255
     },
     {
-      "epoch": 1.49,
-      "learning_rate": 4.120412944808008e-05,
-      "loss": 0.3025,
+      "epoch": 1.523047536670631,
+      "grad_norm": 0.15573151409626007,
+      "learning_rate": 4.0754196926987897e-05,
+      "loss": 0.3929,
       "step": 42260
     },
     {
-      "epoch": 1.49,
-      "learning_rate": 4.120196004225574e-05,
-      "loss": 0.276,
+      "epoch": 1.5232277363318558,
+      "grad_norm": 0.17175668478012085,
+      "learning_rate": 4.0751931008780496e-05,
+      "loss": 0.4094,
       "step": 42265
     },
     {
-      "epoch": 1.49,
-      "learning_rate": 4.119979042605964e-05,
-      "loss": 0.277,
+      "epoch": 1.5234079359930803,
+      "grad_norm": 0.16942541301250458,
+      "learning_rate": 4.074966487595608e-05,
+      "loss": 0.4101,
       "step": 42270
     },
     {
-      "epoch": 1.49,
-      "learning_rate": 4.1197620599519927e-05,
-      "loss": 0.306,
+      "epoch": 1.5235881356543048,
+      "grad_norm": 0.1620819866657257,
+      "learning_rate": 4.074739852854555e-05,
+      "loss": 0.4122,
       "step": 42275
     },
     {
-      "epoch": 1.49,
-      "learning_rate": 4.119545056266478e-05,
-      "loss": 0.2725,
+      "epoch": 1.5237683353155296,
+      "grad_norm": 0.22527901828289032,
+      "learning_rate": 4.0745131966579744e-05,
+      "loss": 0.3755,
       "step": 42280
     },
     {
-      "epoch": 1.49,
-      "learning_rate": 4.11932803155224e-05,
-      "loss": 0.279,
+      "epoch": 1.5239485349767543,
+      "grad_norm": 0.16794419288635254,
+      "learning_rate": 4.074286519008957e-05,
+      "loss": 0.4646,
       "step": 42285
     },
     {
-      "epoch": 1.49,
-      "learning_rate": 4.119110985812093e-05,
-      "loss": 0.2873,
+      "epoch": 1.524128734637979,
+      "grad_norm": 0.20727737247943878,
+      "learning_rate": 4.0740598199105904e-05,
+      "loss": 0.3832,
       "step": 42290
     },
     {
-      "epoch": 1.49,
-      "learning_rate": 4.118893919048858e-05,
-      "loss": 0.281,
+      "epoch": 1.5243089342992036,
+      "grad_norm": 0.1709243506193161,
+      "learning_rate": 4.073833099365965e-05,
+      "loss": 0.4257,
       "step": 42295
     },
     {
-      "epoch": 1.49,
-      "learning_rate": 4.118676831265351e-05,
-      "loss": 0.2841,
+      "epoch": 1.524489133960428,
+      "grad_norm": 0.16771294176578522,
+      "learning_rate": 4.073606357378166e-05,
+      "loss": 0.4417,
       "step": 42300
     },
     {
-      "epoch": 1.49,
-      "learning_rate": 4.1184597224643936e-05,
-      "loss": 0.2785,
+      "epoch": 1.5246693336216528,
+      "grad_norm": 0.2179926335811615,
+      "learning_rate": 4.073379593950286e-05,
+      "loss": 0.3852,
       "step": 42305
     },
     {
-      "epoch": 1.49,
-      "learning_rate": 4.118242592648802e-05,
-      "loss": 0.2958,
+      "epoch": 1.5248495332828775,
+      "grad_norm": 0.19868656992912292,
+      "learning_rate": 4.073152809085414e-05,
+      "loss": 0.4342,
       "step": 42310
     },
     {
-      "epoch": 1.49,
-      "learning_rate": 4.1180254418213976e-05,
-      "loss": 0.2794,
+      "epoch": 1.525029732944102,
+      "grad_norm": 0.14885884523391724,
+      "learning_rate": 4.0729260027866395e-05,
+      "loss": 0.4217,
       "step": 42315
     },
     {
-      "epoch": 1.49,
-      "learning_rate": 4.1178082699849984e-05,
-      "loss": 0.2772,
+      "epoch": 1.5252099326053266,
+      "grad_norm": 0.21947623789310455,
+      "learning_rate": 4.0726991750570525e-05,
+      "loss": 0.4223,
       "step": 42320
     },
     {
-      "epoch": 1.49,
-      "learning_rate": 4.1175910771424266e-05,
-      "loss": 0.2904,
+      "epoch": 1.5253901322665513,
+      "grad_norm": 0.2015712708234787,
+      "learning_rate": 4.072472325899743e-05,
+      "loss": 0.4617,
       "step": 42325
     },
     {
-      "epoch": 1.49,
-      "learning_rate": 4.1173738632964986e-05,
-      "loss": 0.2881,
+      "epoch": 1.525570331927776,
+      "grad_norm": 0.1568281650543213,
+      "learning_rate": 4.072245455317804e-05,
+      "loss": 0.3907,
       "step": 42330
     },
     {
-      "epoch": 1.49,
-      "learning_rate": 4.117156628450039e-05,
-      "loss": 0.2859,
+      "epoch": 1.5257505315890008,
+      "grad_norm": 0.16805681586265564,
+      "learning_rate": 4.0720185633143246e-05,
+      "loss": 0.421,
       "step": 42335
     },
     {
-      "epoch": 1.49,
-      "learning_rate": 4.1169393726058645e-05,
-      "loss": 0.2855,
+      "epoch": 1.5259307312502253,
+      "grad_norm": 0.16904395818710327,
+      "learning_rate": 4.071791649892396e-05,
+      "loss": 0.4295,
       "step": 42340
     },
     {
-      "epoch": 1.49,
-      "learning_rate": 4.116722095766798e-05,
-      "loss": 0.3053,
+      "epoch": 1.5261109309114498,
+      "grad_norm": 0.15109315514564514,
+      "learning_rate": 4.07156471505511e-05,
+      "loss": 0.3572,
       "step": 42345
     },
     {
-      "epoch": 1.49,
-      "learning_rate": 4.116504797935662e-05,
-      "loss": 0.2883,
+      "epoch": 1.5262911305726745,
+      "grad_norm": 0.1983298510313034,
+      "learning_rate": 4.0713377588055604e-05,
+      "loss": 0.4146,
       "step": 42350
     },
     {
-      "epoch": 1.49,
-      "learning_rate": 4.1162874791152755e-05,
-      "loss": 0.2792,
+      "epoch": 1.5264713302338992,
+      "grad_norm": 0.22432877123355865,
+      "learning_rate": 4.0711107811468375e-05,
+      "loss": 0.4239,
       "step": 42355
     },
     {
-      "epoch": 1.49,
-      "learning_rate": 4.116070139308461e-05,
-      "loss": 0.292,
+      "epoch": 1.5266515298951238,
+      "grad_norm": 0.18590804934501648,
+      "learning_rate": 4.070883782082035e-05,
+      "loss": 0.4095,
       "step": 42360
     },
     {
-      "epoch": 1.49,
-      "learning_rate": 4.115852778518041e-05,
-      "loss": 0.2843,
+      "epoch": 1.5268317295563483,
+      "grad_norm": 0.22011986374855042,
+      "learning_rate": 4.070656761614244e-05,
+      "loss": 0.4078,
       "step": 42365
     },
     {
-      "epoch": 1.49,
-      "learning_rate": 4.1156353967468384e-05,
-      "loss": 0.2932,
+      "epoch": 1.527011929217573,
+      "grad_norm": 0.19090569019317627,
+      "learning_rate": 4.0704297197465594e-05,
+      "loss": 0.4148,
       "step": 42370
     },
     {
-      "epoch": 1.49,
-      "learning_rate": 4.115417993997674e-05,
-      "loss": 0.3182,
+      "epoch": 1.5271921288787977,
+      "grad_norm": 0.14910888671875,
+      "learning_rate": 4.070202656482074e-05,
+      "loss": 0.41,
       "step": 42375
     },
     {
-      "epoch": 1.49,
-      "learning_rate": 4.115200570273372e-05,
-      "loss": 0.2978,
+      "epoch": 1.5273723285400225,
+      "grad_norm": 0.13588306307792664,
+      "learning_rate": 4.0699755718238806e-05,
+      "loss": 0.416,
       "step": 42380
     },
     {
-      "epoch": 1.49,
-      "learning_rate": 4.114983125576754e-05,
-      "loss": 0.3059,
+      "epoch": 1.527552528201247,
+      "grad_norm": 0.1951780468225479,
+      "learning_rate": 4.069748465775075e-05,
+      "loss": 0.4077,
       "step": 42385
     },
     {
-      "epoch": 1.49,
-      "learning_rate": 4.114765659910646e-05,
-      "loss": 0.2767,
+      "epoch": 1.5277327278624715,
+      "grad_norm": 0.1684001386165619,
+      "learning_rate": 4.06952133833875e-05,
+      "loss": 0.3969,
       "step": 42390
     },
     {
-      "epoch": 1.49,
-      "learning_rate": 4.1145481732778695e-05,
-      "loss": 0.3192,
+      "epoch": 1.5279129275236962,
+      "grad_norm": 0.15427611768245697,
+      "learning_rate": 4.0692941895180004e-05,
+      "loss": 0.4256,
       "step": 42395
     },
     {
-      "epoch": 1.49,
-      "learning_rate": 4.114330665681249e-05,
-      "loss": 0.2982,
+      "epoch": 1.528093127184921,
+      "grad_norm": 0.19290746748447418,
+      "learning_rate": 4.0690670193159214e-05,
+      "loss": 0.4083,
       "step": 42400
     },
     {
-      "epoch": 1.49,
-      "learning_rate": 4.114113137123608e-05,
-      "loss": 0.3278,
+      "epoch": 1.5282733268461455,
+      "grad_norm": 0.23477697372436523,
+      "learning_rate": 4.0688398277356076e-05,
+      "loss": 0.443,
       "step": 42405
     },
     {
-      "epoch": 1.49,
-      "learning_rate": 4.1138955876077725e-05,
-      "loss": 0.2876,
+      "epoch": 1.5284535265073702,
+      "grad_norm": 0.1695445477962494,
+      "learning_rate": 4.068612614780156e-05,
+      "loss": 0.4119,
       "step": 42410
     },
     {
-      "epoch": 1.49,
-      "learning_rate": 4.1136780171365656e-05,
-      "loss": 0.2876,
+      "epoch": 1.5286337261685947,
+      "grad_norm": 0.1924455612897873,
+      "learning_rate": 4.068385380452661e-05,
+      "loss": 0.3882,
       "step": 42415
     },
     {
-      "epoch": 1.49,
-      "learning_rate": 4.113460425712814e-05,
-      "loss": 0.2808,
+      "epoch": 1.5288139258298195,
+      "grad_norm": 0.17386910319328308,
+      "learning_rate": 4.0681581247562186e-05,
+      "loss": 0.4252,
       "step": 42420
     },
     {
-      "epoch": 1.49,
-      "learning_rate": 4.1132428133393414e-05,
-      "loss": 0.3196,
+      "epoch": 1.5289941254910442,
+      "grad_norm": 0.15284450352191925,
+      "learning_rate": 4.0679308476939245e-05,
+      "loss": 0.3943,
       "step": 42425
     },
     {
-      "epoch": 1.49,
-      "learning_rate": 4.113025180018974e-05,
-      "loss": 0.2973,
+      "epoch": 1.5291743251522687,
+      "grad_norm": 0.21759410202503204,
+      "learning_rate": 4.067703549268877e-05,
+      "loss": 0.4231,
       "step": 42430
     },
     {
-      "epoch": 1.49,
-      "learning_rate": 4.1128075257545384e-05,
-      "loss": 0.2506,
+      "epoch": 1.5293545248134932,
+      "grad_norm": 0.20288415253162384,
+      "learning_rate": 4.0674762294841715e-05,
+      "loss": 0.4339,
       "step": 42435
     },
     {
-      "epoch": 1.49,
-      "learning_rate": 4.1125898505488595e-05,
-      "loss": 0.3138,
+      "epoch": 1.529534724474718,
+      "grad_norm": 0.229291170835495,
+      "learning_rate": 4.067248888342907e-05,
+      "loss": 0.4271,
       "step": 42440
     },
     {
-      "epoch": 1.49,
-      "learning_rate": 4.112372154404765e-05,
-      "loss": 0.2755,
+      "epoch": 1.5297149241359427,
+      "grad_norm": 0.1930796056985855,
+      "learning_rate": 4.067021525848179e-05,
+      "loss": 0.4097,
       "step": 42445
     },
     {
-      "epoch": 1.49,
-      "learning_rate": 4.112154437325079e-05,
-      "loss": 0.2799,
+      "epoch": 1.5298951237971674,
+      "grad_norm": 0.17433331906795502,
+      "learning_rate": 4.066794142003086e-05,
+      "loss": 0.403,
       "step": 42450
     },
     {
-      "epoch": 1.49,
-      "learning_rate": 4.111936699312631e-05,
-      "loss": 0.3056,
+      "epoch": 1.530075323458392,
+      "grad_norm": 0.15750248730182648,
+      "learning_rate": 4.066566736810727e-05,
+      "loss": 0.3983,
       "step": 42455
     },
     {
-      "epoch": 1.49,
-      "learning_rate": 4.111718940370247e-05,
-      "loss": 0.2922,
+      "epoch": 1.5302555231196164,
+      "grad_norm": 0.22471831738948822,
+      "learning_rate": 4.0663393102741995e-05,
+      "loss": 0.4662,
       "step": 42460
     },
     {
-      "epoch": 1.49,
-      "learning_rate": 4.1115011605007545e-05,
-      "loss": 0.2564,
+      "epoch": 1.5304357227808412,
+      "grad_norm": 0.17949113249778748,
+      "learning_rate": 4.066111862396601e-05,
+      "loss": 0.4209,
       "step": 42465
     },
     {
-      "epoch": 1.49,
-      "learning_rate": 4.111283359706982e-05,
-      "loss": 0.2812,
+      "epoch": 1.530615922442066,
+      "grad_norm": 0.17478035390377045,
+      "learning_rate": 4.065884393181032e-05,
+      "loss": 0.3824,
       "step": 42470
     },
     {
-      "epoch": 1.49,
-      "learning_rate": 4.111065537991757e-05,
-      "loss": 0.3041,
+      "epoch": 1.5307961221032904,
+      "grad_norm": 0.14620402455329895,
+      "learning_rate": 4.065656902630592e-05,
+      "loss": 0.4105,
       "step": 42475
     },
     {
-      "epoch": 1.49,
-      "learning_rate": 4.1108476953579076e-05,
-      "loss": 0.2833,
+      "epoch": 1.530976321764515,
+      "grad_norm": 0.2226269692182541,
+      "learning_rate": 4.06542939074838e-05,
+      "loss": 0.4242,
       "step": 42480
     },
     {
-      "epoch": 1.49,
-      "learning_rate": 4.110629831808263e-05,
-      "loss": 0.2885,
+      "epoch": 1.5311565214257397,
+      "grad_norm": 0.1945657581090927,
+      "learning_rate": 4.0652018575374945e-05,
+      "loss": 0.4298,
       "step": 42485
     },
     {
-      "epoch": 1.49,
-      "learning_rate": 4.11041194734565e-05,
-      "loss": 0.2598,
+      "epoch": 1.5313367210869644,
+      "grad_norm": 0.18082071840763092,
+      "learning_rate": 4.0649743030010366e-05,
+      "loss": 0.4119,
       "step": 42490
     },
     {
-      "epoch": 1.5,
-      "learning_rate": 4.1101940419729e-05,
-      "loss": 0.2646,
+      "epoch": 1.5315169207481891,
+      "grad_norm": 0.1490999460220337,
+      "learning_rate": 4.064746727142108e-05,
+      "loss": 0.3611,
       "step": 42495
     },
     {
-      "epoch": 1.5,
-      "learning_rate": 4.109976115692842e-05,
-      "loss": 0.2901,
+      "epoch": 1.5316971204094136,
+      "grad_norm": 0.20667734742164612,
+      "learning_rate": 4.064519129963807e-05,
+      "loss": 0.3959,
       "step": 42500
     },
     {
-      "epoch": 1.5,
-      "eval_loss": 0.28041115403175354,
-      "eval_runtime": 10.5442,
-      "eval_samples_per_second": 9.484,
-      "eval_steps_per_second": 9.484,
+      "epoch": 1.5316971204094136,
+      "eval_loss": 0.4439639151096344,
+      "eval_runtime": 3.531,
+      "eval_samples_per_second": 28.32,
+      "eval_steps_per_second": 7.08,
       "step": 42500
     },
     {
-      "epoch": 1.5,
-      "learning_rate": 4.109758168508304e-05,
-      "loss": 0.285,
+      "epoch": 1.5318773200706381,
+      "grad_norm": 0.18733155727386475,
+      "learning_rate": 4.064291511469237e-05,
+      "loss": 0.4323,
       "step": 42505
     },
     {
-      "epoch": 1.5,
-      "learning_rate": 4.1095402004221184e-05,
-      "loss": 0.2893,
+      "epoch": 1.5320575197318629,
+      "grad_norm": 0.17432504892349243,
+      "learning_rate": 4.064063871661497e-05,
+      "loss": 0.3858,
       "step": 42510
     },
     {
-      "epoch": 1.5,
-      "learning_rate": 4.1093222114371134e-05,
-      "loss": 0.2786,
+      "epoch": 1.5322377193930876,
+      "grad_norm": 0.16605962812900543,
+      "learning_rate": 4.06383621054369e-05,
+      "loss": 0.4115,
       "step": 42515
     },
     {
-      "epoch": 1.5,
-      "learning_rate": 4.10910420155612e-05,
-      "loss": 0.2808,
+      "epoch": 1.5324179190543121,
+      "grad_norm": 0.16073060035705566,
+      "learning_rate": 4.0636085281189176e-05,
+      "loss": 0.428,
       "step": 42520
     },
     {
-      "epoch": 1.5,
-      "learning_rate": 4.108886170781969e-05,
-      "loss": 0.2984,
+      "epoch": 1.5325981187155369,
+      "grad_norm": 0.18236148357391357,
+      "learning_rate": 4.063380824390282e-05,
+      "loss": 0.4489,
       "step": 42525
     },
     {
-      "epoch": 1.5,
-      "learning_rate": 4.108668119117491e-05,
-      "loss": 0.2864,
+      "epoch": 1.5327783183767614,
+      "grad_norm": 0.16601453721523285,
+      "learning_rate": 4.063153099360884e-05,
+      "loss": 0.4184,
       "step": 42530
     },
     {
-      "epoch": 1.5,
-      "learning_rate": 4.1084500465655186e-05,
-      "loss": 0.2877,
+      "epoch": 1.532958518037986,
+      "grad_norm": 0.15551328659057617,
+      "learning_rate": 4.0629253530338284e-05,
+      "loss": 0.3619,
       "step": 42535
     },
     {
-      "epoch": 1.5,
-      "learning_rate": 4.1082319531288816e-05,
-      "loss": 0.2936,
+      "epoch": 1.5331387176992108,
+      "grad_norm": 0.2021326869726181,
+      "learning_rate": 4.062697585412218e-05,
+      "loss": 0.4062,
       "step": 42540
     },
     {
-      "epoch": 1.5,
-      "learning_rate": 4.108013838810413e-05,
-      "loss": 0.2693,
+      "epoch": 1.5333189173604354,
+      "grad_norm": 0.1801024228334427,
+      "learning_rate": 4.062469796499155e-05,
+      "loss": 0.398,
       "step": 42545
     },
     {
-      "epoch": 1.5,
-      "learning_rate": 4.1077957036129436e-05,
-      "loss": 0.2713,
+      "epoch": 1.5334991170216599,
+      "grad_norm": 0.1799817532300949,
+      "learning_rate": 4.062241986297743e-05,
+      "loss": 0.4576,
       "step": 42550
     },
     {
-      "epoch": 1.5,
-      "learning_rate": 4.1075775475393077e-05,
-      "loss": 0.2775,
+      "epoch": 1.5336793166828846,
+      "grad_norm": 0.22685213387012482,
+      "learning_rate": 4.062014154811087e-05,
+      "loss": 0.4469,
       "step": 42555
     },
     {
-      "epoch": 1.5,
-      "learning_rate": 4.107359370592335e-05,
-      "loss": 0.2905,
+      "epoch": 1.5338595163441093,
+      "grad_norm": 0.19450637698173523,
+      "learning_rate": 4.061786302042291e-05,
+      "loss": 0.4152,
       "step": 42560
     },
     {
-      "epoch": 1.5,
-      "learning_rate": 4.107141172774862e-05,
-      "loss": 0.2979,
+      "epoch": 1.534039716005334,
+      "grad_norm": 0.17256684601306915,
+      "learning_rate": 4.0615584279944586e-05,
+      "loss": 0.3805,
       "step": 42565
     },
     {
-      "epoch": 1.5,
-      "learning_rate": 4.106922954089719e-05,
-      "loss": 0.2906,
+      "epoch": 1.5342199156665586,
+      "grad_norm": 0.17024463415145874,
+      "learning_rate": 4.061330532670695e-05,
+      "loss": 0.4086,
       "step": 42570
     },
     {
-      "epoch": 1.5,
-      "learning_rate": 4.106704714539741e-05,
-      "loss": 0.2724,
+      "epoch": 1.534400115327783,
+      "grad_norm": 0.16583757102489471,
+      "learning_rate": 4.061102616074105e-05,
+      "loss": 0.4145,
       "step": 42575
     },
     {
-      "epoch": 1.5,
-      "learning_rate": 4.106486454127761e-05,
-      "loss": 0.2972,
+      "epoch": 1.5345803149890078,
+      "grad_norm": 0.1984453946352005,
+      "learning_rate": 4.060874678207794e-05,
+      "loss": 0.426,
       "step": 42580
     },
     {
-      "epoch": 1.5,
-      "learning_rate": 4.106268172856613e-05,
-      "loss": 0.2951,
+      "epoch": 1.5347605146502326,
+      "grad_norm": 0.192495658993721,
+      "learning_rate": 4.060646719074868e-05,
+      "loss": 0.4444,
       "step": 42585
     },
     {
-      "epoch": 1.5,
-      "learning_rate": 4.1060498707291314e-05,
-      "loss": 0.3015,
+      "epoch": 1.534940714311457,
+      "grad_norm": 0.1842796951532364,
+      "learning_rate": 4.060418738678432e-05,
+      "loss": 0.4038,
       "step": 42590
     },
     {
-      "epoch": 1.5,
-      "learning_rate": 4.1058315477481506e-05,
-      "loss": 0.2937,
+      "epoch": 1.5351209139726816,
+      "grad_norm": 0.17399680614471436,
+      "learning_rate": 4.060190737021594e-05,
+      "loss": 0.4175,
       "step": 42595
     },
     {
-      "epoch": 1.5,
-      "learning_rate": 4.105613203916505e-05,
-      "loss": 0.2863,
+      "epoch": 1.5353011136339063,
+      "grad_norm": 0.20878851413726807,
+      "learning_rate": 4.059962714107458e-05,
+      "loss": 0.4576,
       "step": 42600
     },
     {
-      "epoch": 1.5,
-      "learning_rate": 4.1053948392370315e-05,
-      "loss": 0.2809,
+      "epoch": 1.535481313295131,
+      "grad_norm": 0.19469203054904938,
+      "learning_rate": 4.059734669939133e-05,
+      "loss": 0.4215,
       "step": 42605
     },
     {
-      "epoch": 1.5,
-      "learning_rate": 4.105176453712563e-05,
-      "loss": 0.2942,
+      "epoch": 1.5356615129563558,
+      "grad_norm": 0.1677183359861374,
+      "learning_rate": 4.0595066045197245e-05,
+      "loss": 0.4248,
       "step": 42610
     },
     {
-      "epoch": 1.5,
-      "learning_rate": 4.104958047345936e-05,
-      "loss": 0.2738,
+      "epoch": 1.5358417126175803,
+      "grad_norm": 0.18572258949279785,
+      "learning_rate": 4.059278517852341e-05,
+      "loss": 0.395,
       "step": 42615
     },
     {
-      "epoch": 1.5,
-      "learning_rate": 4.104739620139987e-05,
-      "loss": 0.2744,
+      "epoch": 1.5360219122788048,
+      "grad_norm": 0.2249453216791153,
+      "learning_rate": 4.059050409940089e-05,
+      "loss": 0.4284,
       "step": 42620
     },
     {
-      "epoch": 1.5,
-      "learning_rate": 4.1045211720975506e-05,
-      "loss": 0.2867,
+      "epoch": 1.5362021119400295,
+      "grad_norm": 0.17428001761436462,
+      "learning_rate": 4.058822280786077e-05,
+      "loss": 0.3935,
       "step": 42625
     },
     {
-      "epoch": 1.5,
-      "learning_rate": 4.1043027032214656e-05,
-      "loss": 0.2808,
+      "epoch": 1.5363823116012543,
+      "grad_norm": 0.1760479211807251,
+      "learning_rate": 4.058594130393414e-05,
+      "loss": 0.4689,
       "step": 42630
     },
     {
-      "epoch": 1.5,
-      "learning_rate": 4.104084213514566e-05,
-      "loss": 0.2839,
+      "epoch": 1.5365625112624788,
+      "grad_norm": 0.2455560564994812,
+      "learning_rate": 4.058365958765207e-05,
+      "loss": 0.4436,
       "step": 42635
     },
     {
-      "epoch": 1.5,
-      "learning_rate": 4.1038657029796904e-05,
-      "loss": 0.2892,
+      "epoch": 1.5367427109237035,
+      "grad_norm": 0.18493212759494781,
+      "learning_rate": 4.058137765904565e-05,
+      "loss": 0.4125,
       "step": 42640
     },
     {
-      "epoch": 1.5,
-      "learning_rate": 4.1036471716196755e-05,
-      "loss": 0.293,
+      "epoch": 1.536922910584928,
+      "grad_norm": 0.16667693853378296,
+      "learning_rate": 4.057909551814599e-05,
+      "loss": 0.3845,
       "step": 42645
     },
     {
-      "epoch": 1.5,
-      "learning_rate": 4.103428619437359e-05,
-      "loss": 0.301,
+      "epoch": 1.5371031102461528,
+      "grad_norm": 0.1610030233860016,
+      "learning_rate": 4.057681316498416e-05,
+      "loss": 0.4306,
       "step": 42650
     },
     {
-      "epoch": 1.5,
-      "learning_rate": 4.103210046435579e-05,
-      "loss": 0.2923,
+      "epoch": 1.5372833099073775,
+      "grad_norm": 0.19051054120063782,
+      "learning_rate": 4.0574530599591264e-05,
+      "loss": 0.3968,
       "step": 42655
     },
     {
-      "epoch": 1.5,
-      "learning_rate": 4.1029914526171734e-05,
-      "loss": 0.2789,
+      "epoch": 1.537463509568602,
+      "grad_norm": 0.19822818040847778,
+      "learning_rate": 4.0572247821998414e-05,
+      "loss": 0.4072,
       "step": 42660
     },
     {
-      "epoch": 1.5,
-      "learning_rate": 4.1027728379849794e-05,
-      "loss": 0.3124,
+      "epoch": 1.5376437092298265,
+      "grad_norm": 0.16644228994846344,
+      "learning_rate": 4.056996483223669e-05,
+      "loss": 0.4174,
       "step": 42665
     },
     {
-      "epoch": 1.5,
-      "learning_rate": 4.1025542025418365e-05,
-      "loss": 0.2955,
+      "epoch": 1.5378239088910512,
+      "grad_norm": 0.14218945801258087,
+      "learning_rate": 4.056768163033722e-05,
+      "loss": 0.3884,
       "step": 42670
     },
     {
-      "epoch": 1.5,
-      "learning_rate": 4.102335546290583e-05,
-      "loss": 0.266,
+      "epoch": 1.538004108552276,
+      "grad_norm": 0.21810731291770935,
+      "learning_rate": 4.05653982163311e-05,
+      "loss": 0.3874,
       "step": 42675
     },
     {
-      "epoch": 1.5,
-      "learning_rate": 4.102116869234059e-05,
-      "loss": 0.291,
+      "epoch": 1.5381843082135007,
+      "grad_norm": 0.184495210647583,
+      "learning_rate": 4.056311459024944e-05,
+      "loss": 0.4172,
       "step": 42680
     },
     {
-      "epoch": 1.5,
-      "learning_rate": 4.1018981713751035e-05,
-      "loss": 0.2801,
+      "epoch": 1.5383645078747252,
+      "grad_norm": 0.23121629655361176,
+      "learning_rate": 4.0560830752123355e-05,
+      "loss": 0.4142,
       "step": 42685
     },
     {
-      "epoch": 1.5,
-      "learning_rate": 4.101679452716556e-05,
-      "loss": 0.2897,
+      "epoch": 1.5385447075359497,
+      "grad_norm": 0.1597655862569809,
+      "learning_rate": 4.0558546701983977e-05,
+      "loss": 0.4112,
       "step": 42690
     },
     {
-      "epoch": 1.5,
-      "learning_rate": 4.101460713261256e-05,
-      "loss": 0.2788,
+      "epoch": 1.5387249071971745,
+      "grad_norm": 0.17811793088912964,
+      "learning_rate": 4.05562624398624e-05,
+      "loss": 0.3963,
       "step": 42695
     },
     {
-      "epoch": 1.5,
-      "learning_rate": 4.101241953012044e-05,
-      "loss": 0.2689,
+      "epoch": 1.5389051068583992,
+      "grad_norm": 0.20575061440467834,
+      "learning_rate": 4.055397796578976e-05,
+      "loss": 0.4076,
       "step": 42700
     },
     {
-      "epoch": 1.5,
-      "learning_rate": 4.101023171971761e-05,
-      "loss": 0.2894,
+      "epoch": 1.5390853065196237,
+      "grad_norm": 0.1952335387468338,
+      "learning_rate": 4.055169327979719e-05,
+      "loss": 0.4558,
       "step": 42705
     },
     {
-      "epoch": 1.5,
-      "learning_rate": 4.1008043701432465e-05,
-      "loss": 0.2889,
+      "epoch": 1.5392655061808482,
+      "grad_norm": 0.1855412870645523,
+      "learning_rate": 4.0549408381915796e-05,
+      "loss": 0.4063,
       "step": 42710
     },
     {
-      "epoch": 1.5,
-      "learning_rate": 4.1005855475293426e-05,
-      "loss": 0.2879,
+      "epoch": 1.539445705842073,
+      "grad_norm": 0.16599588096141815,
+      "learning_rate": 4.054712327217673e-05,
+      "loss": 0.4303,
       "step": 42715
     },
     {
-      "epoch": 1.5,
-      "learning_rate": 4.10036670413289e-05,
-      "loss": 0.2864,
+      "epoch": 1.5396259055032977,
+      "grad_norm": 0.15664242208003998,
+      "learning_rate": 4.054483795061112e-05,
+      "loss": 0.3982,
       "step": 42720
     },
     {
-      "epoch": 1.5,
-      "learning_rate": 4.1001478399567305e-05,
-      "loss": 0.2618,
+      "epoch": 1.5398061051645224,
+      "grad_norm": 0.2023800015449524,
+      "learning_rate": 4.05425524172501e-05,
+      "loss": 0.4052,
       "step": 42725
     },
     {
-      "epoch": 1.5,
-      "learning_rate": 4.099928955003706e-05,
-      "loss": 0.2802,
+      "epoch": 1.539986304825747,
+      "grad_norm": 0.1691524088382721,
+      "learning_rate": 4.0540266672124814e-05,
+      "loss": 0.4235,
       "step": 42730
     },
     {
-      "epoch": 1.5,
-      "learning_rate": 4.099710049276658e-05,
-      "loss": 0.2785,
+      "epoch": 1.5401665044869715,
+      "grad_norm": 0.2135685384273529,
+      "learning_rate": 4.0537980715266404e-05,
+      "loss": 0.4041,
       "step": 42735
     },
     {
-      "epoch": 1.5,
-      "learning_rate": 4.0994911227784296e-05,
-      "loss": 0.2576,
+      "epoch": 1.5403467041481962,
+      "grad_norm": 0.16403359174728394,
+      "learning_rate": 4.0535694546706014e-05,
+      "loss": 0.3921,
       "step": 42740
     },
     {
-      "epoch": 1.5,
-      "learning_rate": 4.099272175511863e-05,
-      "loss": 0.3038,
+      "epoch": 1.540526903809421,
+      "grad_norm": 0.1725141555070877,
+      "learning_rate": 4.05334081664748e-05,
+      "loss": 0.4409,
       "step": 42745
     },
     {
-      "epoch": 1.5,
-      "learning_rate": 4.099053207479802e-05,
-      "loss": 0.312,
+      "epoch": 1.5407071034706454,
+      "grad_norm": 0.16590328514575958,
+      "learning_rate": 4.05311215746039e-05,
+      "loss": 0.4197,
       "step": 42750
     },
     {
-      "epoch": 1.5,
-      "learning_rate": 4.098834218685087e-05,
-      "loss": 0.3024,
+      "epoch": 1.5408873031318702,
+      "grad_norm": 0.20750023424625397,
+      "learning_rate": 4.0528834771124474e-05,
+      "loss": 0.4164,
       "step": 42755
     },
     {
-      "epoch": 1.5,
-      "learning_rate": 4.0986152091305645e-05,
-      "loss": 0.2952,
+      "epoch": 1.5410675027930947,
+      "grad_norm": 0.18303443491458893,
+      "learning_rate": 4.052654775606768e-05,
+      "loss": 0.4091,
       "step": 42760
     },
     {
-      "epoch": 1.5,
-      "learning_rate": 4.0983961788190774e-05,
-      "loss": 0.3006,
+      "epoch": 1.5412477024543194,
+      "grad_norm": 0.16997192800045013,
+      "learning_rate": 4.0524260529464695e-05,
+      "loss": 0.4316,
       "step": 42765
     },
     {
-      "epoch": 1.5,
-      "learning_rate": 4.0981771277534684e-05,
-      "loss": 0.322,
+      "epoch": 1.5414279021155441,
+      "grad_norm": 0.17433921992778778,
+      "learning_rate": 4.052197309134665e-05,
+      "loss": 0.3874,
       "step": 42770
     },
     {
-      "epoch": 1.5,
-      "learning_rate": 4.0979580559365835e-05,
-      "loss": 0.2838,
+      "epoch": 1.5416081017767687,
+      "grad_norm": 0.1725977659225464,
+      "learning_rate": 4.051968544174473e-05,
+      "loss": 0.4378,
       "step": 42775
     },
     {
-      "epoch": 1.51,
-      "learning_rate": 4.097738963371266e-05,
-      "loss": 0.2807,
+      "epoch": 1.5417883014379932,
+      "grad_norm": 0.23443584144115448,
+      "learning_rate": 4.0517397580690096e-05,
+      "loss": 0.4625,
       "step": 42780
     },
     {
-      "epoch": 1.51,
-      "learning_rate": 4.097519850060361e-05,
-      "loss": 0.2987,
+      "epoch": 1.541968501099218,
+      "grad_norm": 0.22204741835594177,
+      "learning_rate": 4.051510950821393e-05,
+      "loss": 0.4225,
       "step": 42785
     },
     {
-      "epoch": 1.51,
-      "learning_rate": 4.097300716006713e-05,
-      "loss": 0.2989,
+      "epoch": 1.5421487007604426,
+      "grad_norm": 0.160122349858284,
+      "learning_rate": 4.05128212243474e-05,
+      "loss": 0.3939,
       "step": 42790
     },
     {
-      "epoch": 1.51,
-      "learning_rate": 4.097081561213169e-05,
-      "loss": 0.2826,
+      "epoch": 1.5423289004216674,
+      "grad_norm": 0.21652203798294067,
+      "learning_rate": 4.0510532729121684e-05,
+      "loss": 0.4134,
       "step": 42795
     },
     {
-      "epoch": 1.51,
-      "learning_rate": 4.096862385682572e-05,
-      "loss": 0.277,
+      "epoch": 1.5425091000828919,
+      "grad_norm": 0.2179877758026123,
+      "learning_rate": 4.0508244022567966e-05,
+      "loss": 0.4251,
       "step": 42800
     },
     {
-      "epoch": 1.51,
-      "learning_rate": 4.09664318941777e-05,
-      "loss": 0.2774,
+      "epoch": 1.5426892997441164,
+      "grad_norm": 0.16457681357860565,
+      "learning_rate": 4.050595510471742e-05,
+      "loss": 0.4323,
       "step": 42805
     },
     {
-      "epoch": 1.51,
-      "learning_rate": 4.0964239724216094e-05,
-      "loss": 0.2809,
+      "epoch": 1.5428694994053411,
+      "grad_norm": 0.1653529703617096,
+      "learning_rate": 4.050366597560124e-05,
+      "loss": 0.41,
       "step": 42810
     },
     {
-      "epoch": 1.51,
-      "learning_rate": 4.096204734696935e-05,
-      "loss": 0.2972,
+      "epoch": 1.5430496990665659,
+      "grad_norm": 0.16849032044410706,
+      "learning_rate": 4.0501376635250606e-05,
+      "loss": 0.4431,
       "step": 42815
     },
     {
-      "epoch": 1.51,
-      "learning_rate": 4.095985476246593e-05,
-      "loss": 0.2467,
+      "epoch": 1.5432298987277904,
+      "grad_norm": 0.17099571228027344,
+      "learning_rate": 4.049908708369673e-05,
+      "loss": 0.3836,
       "step": 42820
     },
     {
-      "epoch": 1.51,
-      "learning_rate": 4.0957661970734326e-05,
-      "loss": 0.3159,
+      "epoch": 1.5434100983890149,
+      "grad_norm": 0.198577880859375,
+      "learning_rate": 4.049679732097079e-05,
+      "loss": 0.3952,
       "step": 42825
     },
     {
-      "epoch": 1.51,
-      "learning_rate": 4.095546897180299e-05,
-      "loss": 0.2756,
+      "epoch": 1.5435902980502396,
+      "grad_norm": 0.1876094788312912,
+      "learning_rate": 4.049450734710398e-05,
+      "loss": 0.4149,
       "step": 42830
     },
     {
-      "epoch": 1.51,
-      "learning_rate": 4.095327576570041e-05,
-      "loss": 0.2908,
+      "epoch": 1.5437704977114644,
+      "grad_norm": 0.14950236678123474,
+      "learning_rate": 4.049221716212751e-05,
+      "loss": 0.3671,
       "step": 42835
     },
     {
-      "epoch": 1.51,
-      "learning_rate": 4.095108235245505e-05,
-      "loss": 0.3008,
+      "epoch": 1.543950697372689,
+      "grad_norm": 0.19585910439491272,
+      "learning_rate": 4.048992676607258e-05,
+      "loss": 0.4084,
       "step": 42840
     },
     {
-      "epoch": 1.51,
-      "learning_rate": 4.0948888732095405e-05,
-      "loss": 0.2964,
+      "epoch": 1.5441308970339136,
+      "grad_norm": 0.16137710213661194,
+      "learning_rate": 4.04876361589704e-05,
+      "loss": 0.4119,
       "step": 42845
     },
     {
-      "epoch": 1.51,
-      "learning_rate": 4.0946694904649954e-05,
-      "loss": 0.3011,
+      "epoch": 1.544311096695138,
+      "grad_norm": 0.20326700806617737,
+      "learning_rate": 4.048534534085218e-05,
+      "loss": 0.4127,
       "step": 42850
     },
     {
-      "epoch": 1.51,
-      "learning_rate": 4.0944500870147176e-05,
-      "loss": 0.2992,
+      "epoch": 1.5444912963563628,
+      "grad_norm": 0.16420388221740723,
+      "learning_rate": 4.0483054311749114e-05,
+      "loss": 0.3519,
       "step": 42855
     },
     {
-      "epoch": 1.51,
-      "learning_rate": 4.094230662861555e-05,
-      "loss": 0.2891,
+      "epoch": 1.5446714960175876,
+      "grad_norm": 0.17619933187961578,
+      "learning_rate": 4.048076307169244e-05,
+      "loss": 0.4249,
       "step": 42860
     },
     {
-      "epoch": 1.51,
-      "learning_rate": 4.094011218008359e-05,
-      "loss": 0.2977,
+      "epoch": 1.544851695678812,
+      "grad_norm": 0.1833731085062027,
+      "learning_rate": 4.047847162071336e-05,
+      "loss": 0.4124,
       "step": 42865
     },
     {
-      "epoch": 1.51,
-      "learning_rate": 4.093791752457978e-05,
-      "loss": 0.2708,
+      "epoch": 1.5450318953400366,
+      "grad_norm": 0.18540017306804657,
+      "learning_rate": 4.0476179958843105e-05,
+      "loss": 0.3995,
       "step": 42870
     },
     {
-      "epoch": 1.51,
-      "learning_rate": 4.093572266213261e-05,
-      "loss": 0.2906,
+      "epoch": 1.5452120950012613,
+      "grad_norm": 0.19920703768730164,
+      "learning_rate": 4.04738880861129e-05,
+      "loss": 0.4219,
       "step": 42875
     },
     {
-      "epoch": 1.51,
-      "learning_rate": 4.093352759277058e-05,
-      "loss": 0.2713,
+      "epoch": 1.545392294662486,
+      "grad_norm": 0.19860808551311493,
+      "learning_rate": 4.0471596002553956e-05,
+      "loss": 0.4153,
       "step": 42880
     },
     {
-      "epoch": 1.51,
-      "learning_rate": 4.0931332316522206e-05,
-      "loss": 0.2765,
+      "epoch": 1.5455724943237108,
+      "grad_norm": 0.1732819527387619,
+      "learning_rate": 4.0469303708197515e-05,
+      "loss": 0.4229,
       "step": 42885
     },
     {
-      "epoch": 1.51,
-      "learning_rate": 4.092913683341597e-05,
-      "loss": 0.3059,
+      "epoch": 1.5457526939849353,
+      "grad_norm": 0.19622696936130524,
+      "learning_rate": 4.0467011203074815e-05,
+      "loss": 0.4149,
       "step": 42890
     },
     {
-      "epoch": 1.51,
-      "learning_rate": 4.0926941143480397e-05,
-      "loss": 0.2758,
+      "epoch": 1.5459328936461598,
+      "grad_norm": 0.1721510887145996,
+      "learning_rate": 4.0464718487217066e-05,
+      "loss": 0.4413,
       "step": 42895
     },
     {
-      "epoch": 1.51,
-      "learning_rate": 4.092474524674399e-05,
-      "loss": 0.267,
+      "epoch": 1.5461130933073846,
+      "grad_norm": 0.1426592469215393,
+      "learning_rate": 4.046242556065553e-05,
+      "loss": 0.4244,
       "step": 42900
     },
     {
-      "epoch": 1.51,
-      "learning_rate": 4.092254914323526e-05,
-      "loss": 0.2707,
+      "epoch": 1.5462932929686093,
+      "grad_norm": 0.20957881212234497,
+      "learning_rate": 4.046013242342144e-05,
+      "loss": 0.4354,
       "step": 42905
     },
     {
-      "epoch": 1.51,
-      "learning_rate": 4.0920352832982714e-05,
-      "loss": 0.3051,
+      "epoch": 1.5464734926298338,
+      "grad_norm": 0.15983013808727264,
+      "learning_rate": 4.045783907554604e-05,
+      "loss": 0.3761,
       "step": 42910
     },
     {
-      "epoch": 1.51,
-      "learning_rate": 4.0918156316014884e-05,
-      "loss": 0.3182,
+      "epoch": 1.5466536922910585,
+      "grad_norm": 0.1735716015100479,
+      "learning_rate": 4.045554551706057e-05,
+      "loss": 0.3935,
       "step": 42915
     },
     {
-      "epoch": 1.51,
-      "learning_rate": 4.091595959236029e-05,
-      "loss": 0.3193,
+      "epoch": 1.546833891952283,
+      "grad_norm": 0.17203421890735626,
+      "learning_rate": 4.045325174799629e-05,
+      "loss": 0.4176,
       "step": 42920
     },
     {
-      "epoch": 1.51,
-      "learning_rate": 4.091376266204744e-05,
-      "loss": 0.2697,
+      "epoch": 1.5470140916135078,
+      "grad_norm": 0.17394539713859558,
+      "learning_rate": 4.0450957768384446e-05,
+      "loss": 0.3769,
       "step": 42925
     },
     {
-      "epoch": 1.51,
-      "learning_rate": 4.0911565525104876e-05,
-      "loss": 0.2995,
+      "epoch": 1.5471942912747325,
+      "grad_norm": 0.18221496045589447,
+      "learning_rate": 4.044866357825629e-05,
+      "loss": 0.3931,
       "step": 42930
     },
     {
-      "epoch": 1.51,
-      "learning_rate": 4.090936818156112e-05,
-      "loss": 0.2751,
+      "epoch": 1.547374490935957,
+      "grad_norm": 0.24642367660999298,
+      "learning_rate": 4.0446369177643085e-05,
+      "loss": 0.4427,
       "step": 42935
     },
     {
-      "epoch": 1.51,
-      "learning_rate": 4.0907170631444703e-05,
-      "loss": 0.3034,
+      "epoch": 1.5475546905971815,
+      "grad_norm": 0.2149946391582489,
+      "learning_rate": 4.044407456657609e-05,
+      "loss": 0.4279,
       "step": 42940
     },
     {
-      "epoch": 1.51,
-      "learning_rate": 4.090497287478415e-05,
-      "loss": 0.2835,
+      "epoch": 1.5477348902584063,
+      "grad_norm": 0.147264301776886,
+      "learning_rate": 4.0441779745086575e-05,
+      "loss": 0.3768,
       "step": 42945
     },
     {
-      "epoch": 1.51,
-      "learning_rate": 4.090277491160802e-05,
-      "loss": 0.2839,
+      "epoch": 1.547915089919631,
+      "grad_norm": 0.17553412914276123,
+      "learning_rate": 4.0439484713205795e-05,
+      "loss": 0.3712,
       "step": 42950
     },
     {
-      "epoch": 1.51,
-      "learning_rate": 4.090057674194483e-05,
-      "loss": 0.3095,
+      "epoch": 1.5480952895808557,
+      "grad_norm": 0.1730952113866806,
+      "learning_rate": 4.0437189470965026e-05,
+      "loss": 0.4062,
       "step": 42955
     },
     {
-      "epoch": 1.51,
-      "learning_rate": 4.089837836582313e-05,
-      "loss": 0.2662,
+      "epoch": 1.5482754892420802,
+      "grad_norm": 0.21400214731693268,
+      "learning_rate": 4.043489401839554e-05,
+      "loss": 0.4106,
       "step": 42960
     },
     {
-      "epoch": 1.51,
-      "learning_rate": 4.089617978327146e-05,
-      "loss": 0.2959,
+      "epoch": 1.5484556889033048,
+      "grad_norm": 0.1784030795097351,
+      "learning_rate": 4.0432598355528606e-05,
+      "loss": 0.4188,
       "step": 42965
     },
     {
-      "epoch": 1.51,
-      "learning_rate": 4.089398099431838e-05,
-      "loss": 0.3037,
+      "epoch": 1.5486358885645295,
+      "grad_norm": 0.2129783183336258,
+      "learning_rate": 4.043030248239551e-05,
+      "loss": 0.4148,
       "step": 42970
     },
     {
-      "epoch": 1.51,
-      "learning_rate": 4.089178199899243e-05,
-      "loss": 0.3023,
+      "epoch": 1.5488160882257542,
+      "grad_norm": 0.23545123636722565,
+      "learning_rate": 4.042800639902754e-05,
+      "loss": 0.422,
       "step": 42975
     },
     {
-      "epoch": 1.51,
-      "learning_rate": 4.0889582797322155e-05,
-      "loss": 0.2804,
+      "epoch": 1.5489962878869787,
+      "grad_norm": 0.1820184886455536,
+      "learning_rate": 4.042571010545596e-05,
+      "loss": 0.371,
       "step": 42980
     },
     {
-      "epoch": 1.51,
-      "learning_rate": 4.088738338933613e-05,
-      "loss": 0.2966,
+      "epoch": 1.5491764875482033,
+      "grad_norm": 0.19500315189361572,
+      "learning_rate": 4.0423413601712065e-05,
+      "loss": 0.4332,
       "step": 42985
     },
     {
-      "epoch": 1.51,
-      "learning_rate": 4.0885183775062895e-05,
-      "loss": 0.269,
+      "epoch": 1.549356687209428,
+      "grad_norm": 0.18859097361564636,
+      "learning_rate": 4.042111688782715e-05,
+      "loss": 0.4262,
       "step": 42990
     },
     {
-      "epoch": 1.51,
-      "learning_rate": 4.088298395453102e-05,
-      "loss": 0.274,
+      "epoch": 1.5495368868706527,
+      "grad_norm": 0.16476643085479736,
+      "learning_rate": 4.04188199638325e-05,
+      "loss": 0.4073,
       "step": 42995
     },
     {
-      "epoch": 1.51,
-      "learning_rate": 4.088078392776906e-05,
-      "loss": 0.2803,
+      "epoch": 1.5497170865318775,
+      "grad_norm": 0.1720295399427414,
+      "learning_rate": 4.041652282975942e-05,
+      "loss": 0.4289,
       "step": 43000
     },
     {
-      "epoch": 1.51,
-      "eval_loss": 0.2803913950920105,
-      "eval_runtime": 10.5528,
-      "eval_samples_per_second": 9.476,
-      "eval_steps_per_second": 9.476,
+      "epoch": 1.5497170865318775,
+      "eval_loss": 0.4438794255256653,
+      "eval_runtime": 3.5307,
+      "eval_samples_per_second": 28.323,
+      "eval_steps_per_second": 7.081,
       "step": 43000
     },
     {
-      "epoch": 1.51,
-      "learning_rate": 4.08785836948056e-05,
-      "loss": 0.2946,
+      "epoch": 1.549897286193102,
+      "grad_norm": 0.17833761870861053,
+      "learning_rate": 4.041422548563919e-05,
+      "loss": 0.4653,
       "step": 43005
     },
     {
-      "epoch": 1.51,
-      "learning_rate": 4.0876383255669184e-05,
-      "loss": 0.2873,
+      "epoch": 1.5500774858543265,
+      "grad_norm": 0.1873827427625656,
+      "learning_rate": 4.041192793150314e-05,
+      "loss": 0.4042,
       "step": 43010
     },
     {
-      "epoch": 1.51,
-      "learning_rate": 4.08741826103884e-05,
-      "loss": 0.2766,
+      "epoch": 1.5502576855155512,
+      "grad_norm": 0.20229150354862213,
+      "learning_rate": 4.040963016738254e-05,
+      "loss": 0.446,
       "step": 43015
     },
     {
-      "epoch": 1.51,
-      "learning_rate": 4.087198175899181e-05,
-      "loss": 0.295,
+      "epoch": 1.550437885176776,
+      "grad_norm": 0.17710404098033905,
+      "learning_rate": 4.040733219330871e-05,
+      "loss": 0.4281,
       "step": 43020
     },
     {
-      "epoch": 1.51,
-      "learning_rate": 4.0869780701508e-05,
-      "loss": 0.3029,
+      "epoch": 1.5506180848380005,
+      "grad_norm": 0.15827186405658722,
+      "learning_rate": 4.040503400931297e-05,
+      "loss": 0.3829,
       "step": 43025
     },
     {
-      "epoch": 1.51,
-      "learning_rate": 4.086757943796555e-05,
-      "loss": 0.2777,
+      "epoch": 1.5507982844992252,
+      "grad_norm": 0.18000894784927368,
+      "learning_rate": 4.040273561542662e-05,
+      "loss": 0.4088,
       "step": 43030
     },
     {
-      "epoch": 1.51,
-      "learning_rate": 4.086537796839304e-05,
-      "loss": 0.3146,
+      "epoch": 1.5509784841604497,
+      "grad_norm": 0.18159720301628113,
+      "learning_rate": 4.0400437011680986e-05,
+      "loss": 0.4461,
       "step": 43035
     },
     {
-      "epoch": 1.51,
-      "learning_rate": 4.086317629281904e-05,
-      "loss": 0.2991,
+      "epoch": 1.5511586838216744,
+      "grad_norm": 0.19357514381408691,
+      "learning_rate": 4.039813819810737e-05,
+      "loss": 0.4084,
       "step": 43040
     },
     {
-      "epoch": 1.51,
-      "learning_rate": 4.086097441127216e-05,
-      "loss": 0.2706,
+      "epoch": 1.5513388834828992,
+      "grad_norm": 0.16021384298801422,
+      "learning_rate": 4.039583917473711e-05,
+      "loss": 0.4259,
       "step": 43045
     },
     {
-      "epoch": 1.51,
-      "learning_rate": 4.085877232378098e-05,
-      "loss": 0.2813,
+      "epoch": 1.5515190831441237,
+      "grad_norm": 0.22351530194282532,
+      "learning_rate": 4.039353994160152e-05,
+      "loss": 0.4508,
       "step": 43050
     },
     {
-      "epoch": 1.51,
-      "learning_rate": 4.085657003037409e-05,
-      "loss": 0.3014,
+      "epoch": 1.5516992828053482,
+      "grad_norm": 0.1790456771850586,
+      "learning_rate": 4.039124049873193e-05,
+      "loss": 0.435,
       "step": 43055
     },
     {
-      "epoch": 1.51,
-      "learning_rate": 4.0854367531080086e-05,
-      "loss": 0.2965,
+      "epoch": 1.551879482466573,
+      "grad_norm": 0.1445944756269455,
+      "learning_rate": 4.038894084615966e-05,
+      "loss": 0.4111,
       "step": 43060
     },
     {
-      "epoch": 1.52,
-      "learning_rate": 4.085216482592757e-05,
-      "loss": 0.2942,
+      "epoch": 1.5520596821277977,
+      "grad_norm": 0.1887555867433548,
+      "learning_rate": 4.038664098391606e-05,
+      "loss": 0.4236,
       "step": 43065
     },
     {
-      "epoch": 1.52,
-      "learning_rate": 4.0849961914945146e-05,
-      "loss": 0.291,
+      "epoch": 1.5522398817890224,
+      "grad_norm": 0.14658775925636292,
+      "learning_rate": 4.038434091203245e-05,
+      "loss": 0.4447,
       "step": 43070
     },
     {
-      "epoch": 1.52,
-      "learning_rate": 4.084775879816141e-05,
-      "loss": 0.2719,
+      "epoch": 1.552420081450247,
+      "grad_norm": 0.192291259765625,
+      "learning_rate": 4.038204063054017e-05,
+      "loss": 0.468,
       "step": 43075
     },
     {
-      "epoch": 1.52,
-      "learning_rate": 4.0845555475604966e-05,
-      "loss": 0.2727,
+      "epoch": 1.5526002811114714,
+      "grad_norm": 0.15899904072284698,
+      "learning_rate": 4.037974013947058e-05,
+      "loss": 0.377,
       "step": 43080
     },
     {
-      "epoch": 1.52,
-      "learning_rate": 4.0843351947304425e-05,
-      "loss": 0.2802,
+      "epoch": 1.5527804807726961,
+      "grad_norm": 0.182462677359581,
+      "learning_rate": 4.037743943885499e-05,
+      "loss": 0.4033,
       "step": 43085
     },
     {
-      "epoch": 1.52,
-      "learning_rate": 4.084114821328841e-05,
-      "loss": 0.2944,
+      "epoch": 1.5529606804339209,
+      "grad_norm": 0.17787639796733856,
+      "learning_rate": 4.037513852872478e-05,
+      "loss": 0.416,
       "step": 43090
     },
     {
-      "epoch": 1.52,
-      "learning_rate": 4.083894427358551e-05,
-      "loss": 0.2627,
+      "epoch": 1.5531408800951454,
+      "grad_norm": 0.1610783487558365,
+      "learning_rate": 4.037283740911128e-05,
+      "loss": 0.4039,
       "step": 43095
     },
     {
-      "epoch": 1.52,
-      "learning_rate": 4.0836740128224374e-05,
-      "loss": 0.3068,
+      "epoch": 1.55332107975637,
+      "grad_norm": 0.15985815227031708,
+      "learning_rate": 4.037053608004584e-05,
+      "loss": 0.4483,
       "step": 43100
     },
     {
-      "epoch": 1.52,
-      "learning_rate": 4.083453577723359e-05,
-      "loss": 0.2858,
+      "epoch": 1.5535012794175946,
+      "grad_norm": 0.21218977868556976,
+      "learning_rate": 4.036823454155982e-05,
+      "loss": 0.4047,
       "step": 43105
     },
     {
-      "epoch": 1.52,
-      "learning_rate": 4.083233122064179e-05,
-      "loss": 0.304,
+      "epoch": 1.5536814790788194,
+      "grad_norm": 0.1868336945772171,
+      "learning_rate": 4.036593279368458e-05,
+      "loss": 0.4103,
       "step": 43110
     },
     {
-      "epoch": 1.52,
-      "learning_rate": 4.083012645847762e-05,
-      "loss": 0.3016,
+      "epoch": 1.553861678740044,
+      "grad_norm": 0.24280469119548798,
+      "learning_rate": 4.0363630836451496e-05,
+      "loss": 0.4188,
       "step": 43115
     },
     {
-      "epoch": 1.52,
-      "learning_rate": 4.0827921490769683e-05,
-      "loss": 0.2655,
+      "epoch": 1.5540418784012686,
+      "grad_norm": 0.1968899369239807,
+      "learning_rate": 4.036132866989191e-05,
+      "loss": 0.419,
       "step": 43120
     },
     {
-      "epoch": 1.52,
-      "learning_rate": 4.0825716317546616e-05,
-      "loss": 0.3103,
+      "epoch": 1.5542220780624931,
+      "grad_norm": 0.19591078162193298,
+      "learning_rate": 4.035902629403718e-05,
+      "loss": 0.4057,
       "step": 43125
     },
     {
-      "epoch": 1.52,
-      "learning_rate": 4.0823510938837055e-05,
-      "loss": 0.2703,
+      "epoch": 1.5544022777237179,
+      "grad_norm": 0.1963011920452118,
+      "learning_rate": 4.0356723708918705e-05,
+      "loss": 0.394,
       "step": 43130
     },
     {
-      "epoch": 1.52,
-      "learning_rate": 4.0821305354669624e-05,
-      "loss": 0.2977,
+      "epoch": 1.5545824773849426,
+      "grad_norm": 0.1776949167251587,
+      "learning_rate": 4.035442091456784e-05,
+      "loss": 0.3855,
       "step": 43135
     },
     {
-      "epoch": 1.52,
-      "learning_rate": 4.081909956507298e-05,
-      "loss": 0.2772,
+      "epoch": 1.554762677046167,
+      "grad_norm": 0.21400536596775055,
+      "learning_rate": 4.035211791101596e-05,
+      "loss": 0.4177,
       "step": 43140
     },
     {
-      "epoch": 1.52,
-      "learning_rate": 4.081689357007575e-05,
-      "loss": 0.2902,
+      "epoch": 1.5549428767073918,
+      "grad_norm": 0.22430282831192017,
+      "learning_rate": 4.034981469829445e-05,
+      "loss": 0.444,
       "step": 43145
     },
     {
-      "epoch": 1.52,
-      "learning_rate": 4.081468736970658e-05,
-      "loss": 0.2871,
+      "epoch": 1.5551230763686164,
+      "grad_norm": 0.1932390183210373,
+      "learning_rate": 4.034751127643468e-05,
+      "loss": 0.3992,
       "step": 43150
     },
     {
-      "epoch": 1.52,
-      "learning_rate": 4.081248096399412e-05,
-      "loss": 0.2702,
+      "epoch": 1.555303276029841,
+      "grad_norm": 0.20954085886478424,
+      "learning_rate": 4.034520764546805e-05,
+      "loss": 0.4377,
       "step": 43155
     },
     {
-      "epoch": 1.52,
-      "learning_rate": 4.081027435296701e-05,
-      "loss": 0.2834,
+      "epoch": 1.5554834756910658,
+      "grad_norm": 0.21870118379592896,
+      "learning_rate": 4.034290380542593e-05,
+      "loss": 0.4406,
       "step": 43160
     },
     {
-      "epoch": 1.52,
-      "learning_rate": 4.08080675366539e-05,
-      "loss": 0.2735,
+      "epoch": 1.5556636753522903,
+      "grad_norm": 0.19834262132644653,
+      "learning_rate": 4.0340599756339715e-05,
+      "loss": 0.4218,
       "step": 43165
     },
     {
-      "epoch": 1.52,
-      "learning_rate": 4.080586051508347e-05,
-      "loss": 0.2994,
+      "epoch": 1.5558438750135148,
+      "grad_norm": 0.1412007212638855,
+      "learning_rate": 4.03382954982408e-05,
+      "loss": 0.3994,
       "step": 43170
     },
     {
-      "epoch": 1.52,
-      "learning_rate": 4.080365328828435e-05,
-      "loss": 0.2786,
+      "epoch": 1.5560240746747396,
+      "grad_norm": 0.15708720684051514,
+      "learning_rate": 4.0335991031160584e-05,
+      "loss": 0.3901,
       "step": 43175
     },
     {
-      "epoch": 1.52,
-      "learning_rate": 4.0801445856285205e-05,
-      "loss": 0.2774,
+      "epoch": 1.5562042743359643,
+      "grad_norm": 0.13734932243824005,
+      "learning_rate": 4.0333686355130454e-05,
+      "loss": 0.4384,
       "step": 43180
     },
     {
-      "epoch": 1.52,
-      "learning_rate": 4.07992382191147e-05,
-      "loss": 0.2731,
+      "epoch": 1.556384473997189,
+      "grad_norm": 0.2183203250169754,
+      "learning_rate": 4.033138147018181e-05,
+      "loss": 0.4417,
       "step": 43185
     },
     {
-      "epoch": 1.52,
-      "learning_rate": 4.079703037680149e-05,
-      "loss": 0.2878,
+      "epoch": 1.5565646736584136,
+      "grad_norm": 0.20702525973320007,
+      "learning_rate": 4.0329076376346063e-05,
+      "loss": 0.4323,
       "step": 43190
     },
     {
-      "epoch": 1.52,
-      "learning_rate": 4.079482232937426e-05,
-      "loss": 0.2821,
+      "epoch": 1.556744873319638,
+      "grad_norm": 0.20246562361717224,
+      "learning_rate": 4.032677107365463e-05,
+      "loss": 0.4611,
       "step": 43195
     },
     {
-      "epoch": 1.52,
-      "learning_rate": 4.079261407686167e-05,
-      "loss": 0.2589,
+      "epoch": 1.5569250729808628,
+      "grad_norm": 0.15538166463375092,
+      "learning_rate": 4.0324465562138905e-05,
+      "loss": 0.4066,
       "step": 43200
     },
     {
-      "epoch": 1.52,
-      "learning_rate": 4.079040561929239e-05,
-      "loss": 0.2859,
+      "epoch": 1.5571052726420875,
+      "grad_norm": 0.1648663729429245,
+      "learning_rate": 4.03221598418303e-05,
+      "loss": 0.4347,
       "step": 43205
     },
     {
-      "epoch": 1.52,
-      "learning_rate": 4.078819695669511e-05,
-      "loss": 0.2743,
+      "epoch": 1.557285472303312,
+      "grad_norm": 0.2193061113357544,
+      "learning_rate": 4.031985391276023e-05,
+      "loss": 0.412,
       "step": 43210
     },
     {
-      "epoch": 1.52,
-      "learning_rate": 4.0785988089098494e-05,
-      "loss": 0.2989,
+      "epoch": 1.5574656719645366,
+      "grad_norm": 0.1736992746591568,
+      "learning_rate": 4.031754777496012e-05,
+      "loss": 0.4344,
       "step": 43215
     },
     {
-      "epoch": 1.52,
-      "learning_rate": 4.078377901653122e-05,
-      "loss": 0.288,
+      "epoch": 1.5576458716257613,
+      "grad_norm": 0.18772049248218536,
+      "learning_rate": 4.031524142846139e-05,
+      "loss": 0.4455,
       "step": 43220
     },
     {
-      "epoch": 1.52,
-      "learning_rate": 4.078156973902198e-05,
-      "loss": 0.2737,
+      "epoch": 1.557826071286986,
+      "grad_norm": 0.185786172747612,
+      "learning_rate": 4.031293487329546e-05,
+      "loss": 0.4187,
       "step": 43225
     },
     {
-      "epoch": 1.52,
-      "learning_rate": 4.077936025659945e-05,
-      "loss": 0.3033,
+      "epoch": 1.5580062709482108,
+      "grad_norm": 0.21457712352275848,
+      "learning_rate": 4.031062810949375e-05,
+      "loss": 0.4008,
       "step": 43230
     },
     {
-      "epoch": 1.52,
-      "learning_rate": 4.0777150569292345e-05,
-      "loss": 0.2956,
+      "epoch": 1.5581864706094353,
+      "grad_norm": 0.16875572502613068,
+      "learning_rate": 4.03083211370877e-05,
+      "loss": 0.4041,
       "step": 43235
     },
     {
-      "epoch": 1.52,
-      "learning_rate": 4.0774940677129316e-05,
-      "loss": 0.2846,
+      "epoch": 1.5583666702706598,
+      "grad_norm": 0.1875239759683609,
+      "learning_rate": 4.0306013956108747e-05,
+      "loss": 0.4421,
       "step": 43240
     },
     {
-      "epoch": 1.52,
-      "learning_rate": 4.07727305801391e-05,
-      "loss": 0.3056,
+      "epoch": 1.5585468699318845,
+      "grad_norm": 0.1823228895664215,
+      "learning_rate": 4.030370656658831e-05,
+      "loss": 0.3946,
       "step": 43245
     },
     {
-      "epoch": 1.52,
-      "learning_rate": 4.0770520278350354e-05,
-      "loss": 0.2888,
+      "epoch": 1.5587270695931092,
+      "grad_norm": 0.1981915831565857,
+      "learning_rate": 4.030139896855783e-05,
+      "loss": 0.4127,
       "step": 43250
     },
     {
-      "epoch": 1.52,
-      "learning_rate": 4.07683097717918e-05,
-      "loss": 0.2872,
+      "epoch": 1.5589072692543338,
+      "grad_norm": 0.17064636945724487,
+      "learning_rate": 4.029909116204875e-05,
+      "loss": 0.3918,
       "step": 43255
     },
     {
-      "epoch": 1.52,
-      "learning_rate": 4.076609906049214e-05,
-      "loss": 0.2991,
+      "epoch": 1.5590874689155585,
+      "grad_norm": 0.21734385192394257,
+      "learning_rate": 4.0296783147092527e-05,
+      "loss": 0.4109,
       "step": 43260
     },
     {
-      "epoch": 1.52,
-      "learning_rate": 4.0763888144480075e-05,
-      "loss": 0.292,
+      "epoch": 1.559267668576783,
+      "grad_norm": 0.18934614956378937,
+      "learning_rate": 4.029447492372059e-05,
+      "loss": 0.407,
       "step": 43265
     },
     {
-      "epoch": 1.52,
-      "learning_rate": 4.076167702378431e-05,
-      "loss": 0.2684,
+      "epoch": 1.5594478682380077,
+      "grad_norm": 0.22457720339298248,
+      "learning_rate": 4.029216649196439e-05,
+      "loss": 0.428,
       "step": 43270
     },
     {
-      "epoch": 1.52,
-      "learning_rate": 4.0759465698433545e-05,
-      "loss": 0.2921,
+      "epoch": 1.5596280678992325,
+      "grad_norm": 0.16613884270191193,
+      "learning_rate": 4.028985785185538e-05,
+      "loss": 0.4337,
       "step": 43275
     },
     {
-      "epoch": 1.52,
-      "learning_rate": 4.0757254168456514e-05,
-      "loss": 0.2779,
+      "epoch": 1.559808267560457,
+      "grad_norm": 0.1653304249048233,
+      "learning_rate": 4.0287549003425026e-05,
+      "loss": 0.3973,
       "step": 43280
     },
     {
-      "epoch": 1.52,
-      "learning_rate": 4.075504243388192e-05,
-      "loss": 0.2974,
+      "epoch": 1.5599884672216815,
+      "grad_norm": 0.18372784554958344,
+      "learning_rate": 4.028523994670477e-05,
+      "loss": 0.4102,
       "step": 43285
     },
     {
-      "epoch": 1.52,
-      "learning_rate": 4.075283049473848e-05,
-      "loss": 0.2597,
+      "epoch": 1.5601686668829062,
+      "grad_norm": 0.1557554453611374,
+      "learning_rate": 4.028293068172608e-05,
+      "loss": 0.417,
       "step": 43290
     },
     {
-      "epoch": 1.52,
-      "learning_rate": 4.0750618351054924e-05,
-      "loss": 0.268,
+      "epoch": 1.560348866544131,
+      "grad_norm": 0.17378360033035278,
+      "learning_rate": 4.028062120852042e-05,
+      "loss": 0.4113,
       "step": 43295
     },
     {
-      "epoch": 1.52,
-      "learning_rate": 4.0748406002859963e-05,
-      "loss": 0.2818,
+      "epoch": 1.5605290662053557,
+      "grad_norm": 0.17926354706287384,
+      "learning_rate": 4.027831152711925e-05,
+      "loss": 0.402,
       "step": 43300
     },
     {
-      "epoch": 1.52,
-      "learning_rate": 4.074619345018233e-05,
-      "loss": 0.2867,
+      "epoch": 1.5607092658665802,
+      "grad_norm": 0.17073003947734833,
+      "learning_rate": 4.027600163755405e-05,
+      "loss": 0.4631,
       "step": 43305
     },
     {
-      "epoch": 1.52,
-      "learning_rate": 4.0743980693050753e-05,
-      "loss": 0.2596,
+      "epoch": 1.5608894655278047,
+      "grad_norm": 0.17934848368167877,
+      "learning_rate": 4.027369153985628e-05,
+      "loss": 0.4058,
       "step": 43310
     },
     {
-      "epoch": 1.52,
-      "learning_rate": 4.074176773149396e-05,
-      "loss": 0.3037,
+      "epoch": 1.5610696651890295,
+      "grad_norm": 0.16843603551387787,
+      "learning_rate": 4.0271381234057426e-05,
+      "loss": 0.4345,
       "step": 43315
     },
     {
-      "epoch": 1.52,
-      "learning_rate": 4.0739554565540697e-05,
-      "loss": 0.2669,
+      "epoch": 1.5612498648502542,
+      "grad_norm": 0.1804237961769104,
+      "learning_rate": 4.026907072018896e-05,
+      "loss": 0.4171,
       "step": 43320
     },
     {
-      "epoch": 1.52,
-      "learning_rate": 4.073734119521968e-05,
-      "loss": 0.2882,
+      "epoch": 1.5614300645114787,
+      "grad_norm": 0.18647703528404236,
+      "learning_rate": 4.0266759998282355e-05,
+      "loss": 0.4483,
       "step": 43325
     },
     {
-      "epoch": 1.52,
-      "learning_rate": 4.0735127620559664e-05,
-      "loss": 0.3036,
+      "epoch": 1.5616102641727032,
+      "grad_norm": 0.15589216351509094,
+      "learning_rate": 4.026444906836909e-05,
+      "loss": 0.3917,
       "step": 43330
     },
     {
-      "epoch": 1.52,
-      "learning_rate": 4.0732913841589385e-05,
-      "loss": 0.2811,
+      "epoch": 1.561790463833928,
+      "grad_norm": 0.2177332192659378,
+      "learning_rate": 4.026213793048068e-05,
+      "loss": 0.4652,
       "step": 43335
     },
     {
-      "epoch": 1.52,
-      "learning_rate": 4.073069985833758e-05,
-      "loss": 0.2856,
+      "epoch": 1.5619706634951527,
+      "grad_norm": 0.22540467977523804,
+      "learning_rate": 4.0259826584648596e-05,
+      "loss": 0.4222,
       "step": 43340
     },
     {
-      "epoch": 1.52,
-      "learning_rate": 4.072848567083302e-05,
-      "loss": 0.3084,
+      "epoch": 1.5621508631563774,
+      "grad_norm": 0.1936839520931244,
+      "learning_rate": 4.025751503090432e-05,
+      "loss": 0.4248,
       "step": 43345
     },
     {
-      "epoch": 1.53,
-      "learning_rate": 4.0726271279104435e-05,
-      "loss": 0.2959,
+      "epoch": 1.562331062817602,
+      "grad_norm": 0.20749831199645996,
+      "learning_rate": 4.025520326927936e-05,
+      "loss": 0.4026,
       "step": 43350
     },
     {
-      "epoch": 1.53,
-      "learning_rate": 4.072405668318057e-05,
-      "loss": 0.2694,
+      "epoch": 1.5625112624788264,
+      "grad_norm": 0.1811009794473648,
+      "learning_rate": 4.025289129980521e-05,
+      "loss": 0.3812,
       "step": 43355
     },
     {
-      "epoch": 1.53,
-      "learning_rate": 4.07218418830902e-05,
-      "loss": 0.3004,
+      "epoch": 1.5626914621400512,
+      "grad_norm": 0.1572161763906479,
+      "learning_rate": 4.025057912251337e-05,
+      "loss": 0.4306,
       "step": 43360
     },
     {
-      "epoch": 1.53,
-      "learning_rate": 4.0719626878862074e-05,
-      "loss": 0.246,
+      "epoch": 1.562871661801276,
+      "grad_norm": 0.18943502008914948,
+      "learning_rate": 4.024826673743533e-05,
+      "loss": 0.3895,
       "step": 43365
     },
     {
-      "epoch": 1.53,
-      "learning_rate": 4.071741167052495e-05,
-      "loss": 0.2825,
+      "epoch": 1.5630518614625004,
+      "grad_norm": 0.2089279443025589,
+      "learning_rate": 4.024595414460261e-05,
+      "loss": 0.4108,
       "step": 43370
     },
     {
-      "epoch": 1.53,
-      "learning_rate": 4.071519625810759e-05,
-      "loss": 0.2835,
+      "epoch": 1.563232061123725,
+      "grad_norm": 0.16804815828800201,
+      "learning_rate": 4.0243641344046725e-05,
+      "loss": 0.3943,
       "step": 43375
     },
     {
-      "epoch": 1.53,
-      "learning_rate": 4.0712980641638774e-05,
-      "loss": 0.2787,
+      "epoch": 1.5634122607849497,
+      "grad_norm": 0.19758044183254242,
+      "learning_rate": 4.0241328335799185e-05,
+      "loss": 0.4298,
       "step": 43380
     },
     {
-      "epoch": 1.53,
-      "learning_rate": 4.071076482114725e-05,
-      "loss": 0.27,
+      "epoch": 1.5635924604461744,
+      "grad_norm": 0.18327586352825165,
+      "learning_rate": 4.023901511989149e-05,
+      "loss": 0.4356,
       "step": 43385
     },
     {
-      "epoch": 1.53,
-      "learning_rate": 4.0708548796661805e-05,
-      "loss": 0.3059,
+      "epoch": 1.5637726601073991,
+      "grad_norm": 0.18740832805633545,
+      "learning_rate": 4.023670169635516e-05,
+      "loss": 0.4263,
       "step": 43390
     },
     {
-      "epoch": 1.53,
-      "learning_rate": 4.07063325682112e-05,
-      "loss": 0.2701,
+      "epoch": 1.5639528597686236,
+      "grad_norm": 0.18278846144676208,
+      "learning_rate": 4.0234388065221716e-05,
+      "loss": 0.3991,
       "step": 43395
     },
     {
-      "epoch": 1.53,
-      "learning_rate": 4.0704116135824226e-05,
-      "loss": 0.3022,
+      "epoch": 1.5641330594298481,
+      "grad_norm": 0.137952983379364,
+      "learning_rate": 4.02320742265227e-05,
+      "loss": 0.4308,
       "step": 43400
     },
     {
-      "epoch": 1.53,
-      "learning_rate": 4.070189949952965e-05,
-      "loss": 0.2763,
+      "epoch": 1.5643132590910729,
+      "grad_norm": 0.16237494349479675,
+      "learning_rate": 4.0229760180289604e-05,
+      "loss": 0.3995,
       "step": 43405
     },
     {
-      "epoch": 1.53,
-      "learning_rate": 4.0699682659356254e-05,
-      "loss": 0.3014,
+      "epoch": 1.5644934587522976,
+      "grad_norm": 0.1925591230392456,
+      "learning_rate": 4.022744592655398e-05,
+      "loss": 0.4095,
       "step": 43410
     },
     {
-      "epoch": 1.53,
-      "learning_rate": 4.069746561533282e-05,
-      "loss": 0.3238,
+      "epoch": 1.5646736584135221,
+      "grad_norm": 0.17975488305091858,
+      "learning_rate": 4.022513146534735e-05,
+      "loss": 0.397,
       "step": 43415
     },
     {
-      "epoch": 1.53,
-      "learning_rate": 4.069524836748815e-05,
-      "loss": 0.2848,
+      "epoch": 1.5648538580747469,
+      "grad_norm": 0.19634199142456055,
+      "learning_rate": 4.022281679670127e-05,
+      "loss": 0.4361,
       "step": 43420
     },
     {
-      "epoch": 1.53,
-      "learning_rate": 4.069303091585102e-05,
-      "loss": 0.2805,
+      "epoch": 1.5650340577359714,
+      "grad_norm": 0.22516696155071259,
+      "learning_rate": 4.022050192064724e-05,
+      "loss": 0.4091,
       "step": 43425
     },
     {
-      "epoch": 1.53,
-      "learning_rate": 4.069081326045023e-05,
-      "loss": 0.2921,
+      "epoch": 1.565214257397196,
+      "grad_norm": 0.19511666893959045,
+      "learning_rate": 4.021818683721682e-05,
+      "loss": 0.4301,
       "step": 43430
     },
     {
-      "epoch": 1.53,
-      "learning_rate": 4.068859540131457e-05,
-      "loss": 0.277,
+      "epoch": 1.5653944570584208,
+      "grad_norm": 0.19139276444911957,
+      "learning_rate": 4.021587154644156e-05,
+      "loss": 0.4132,
       "step": 43435
     },
     {
-      "epoch": 1.53,
-      "learning_rate": 4.068637733847283e-05,
-      "loss": 0.2988,
+      "epoch": 1.5655746567196454,
+      "grad_norm": 0.18717695772647858,
+      "learning_rate": 4.021355604835299e-05,
+      "loss": 0.4059,
       "step": 43440
     },
     {
-      "epoch": 1.53,
-      "learning_rate": 4.068415907195382e-05,
-      "loss": 0.295,
+      "epoch": 1.5657548563808699,
+      "grad_norm": 0.16392220556735992,
+      "learning_rate": 4.0211240342982656e-05,
+      "loss": 0.4349,
       "step": 43445
     },
     {
-      "epoch": 1.53,
-      "learning_rate": 4.0681940601786345e-05,
-      "loss": 0.2782,
+      "epoch": 1.5659350560420946,
+      "grad_norm": 0.23648183047771454,
+      "learning_rate": 4.0208924430362126e-05,
+      "loss": 0.4023,
       "step": 43450
     },
     {
-      "epoch": 1.53,
-      "learning_rate": 4.06797219279992e-05,
-      "loss": 0.2732,
+      "epoch": 1.5661152557033193,
+      "grad_norm": 0.23210176825523376,
+      "learning_rate": 4.020660831052295e-05,
+      "loss": 0.4303,
       "step": 43455
     },
     {
-      "epoch": 1.53,
-      "learning_rate": 4.06775030506212e-05,
-      "loss": 0.3081,
+      "epoch": 1.566295455364544,
+      "grad_norm": 0.18653087317943573,
+      "learning_rate": 4.020429198349667e-05,
+      "loss": 0.4452,
       "step": 43460
     },
     {
-      "epoch": 1.53,
-      "learning_rate": 4.067528396968116e-05,
-      "loss": 0.3017,
+      "epoch": 1.5664756550257686,
+      "grad_norm": 0.17263540625572205,
+      "learning_rate": 4.0201975449314865e-05,
+      "loss": 0.4231,
       "step": 43465
     },
     {
-      "epoch": 1.53,
-      "learning_rate": 4.067306468520788e-05,
-      "loss": 0.2942,
+      "epoch": 1.566655854686993,
+      "grad_norm": 0.2012946456670761,
+      "learning_rate": 4.019965870800908e-05,
+      "loss": 0.4403,
       "step": 43470
     },
     {
-      "epoch": 1.53,
-      "learning_rate": 4.06708451972302e-05,
-      "loss": 0.3089,
+      "epoch": 1.5668360543482178,
+      "grad_norm": 0.2045126110315323,
+      "learning_rate": 4.019734175961089e-05,
+      "loss": 0.4306,
       "step": 43475
     },
     {
-      "epoch": 1.53,
-      "learning_rate": 4.0668625505776906e-05,
-      "loss": 0.2851,
+      "epoch": 1.5670162540094426,
+      "grad_norm": 0.15547120571136475,
+      "learning_rate": 4.019502460415186e-05,
+      "loss": 0.4273,
       "step": 43480
     },
     {
-      "epoch": 1.53,
-      "learning_rate": 4.066640561087684e-05,
-      "loss": 0.2862,
+      "epoch": 1.567196453670667,
+      "grad_norm": 0.17727886140346527,
+      "learning_rate": 4.0192707241663567e-05,
+      "loss": 0.4196,
       "step": 43485
     },
     {
-      "epoch": 1.53,
-      "learning_rate": 4.066418551255883e-05,
-      "loss": 0.2883,
+      "epoch": 1.5673766533318916,
+      "grad_norm": 0.16581672430038452,
+      "learning_rate": 4.0190389672177575e-05,
+      "loss": 0.3979,
       "step": 43490
     },
     {
-      "epoch": 1.53,
-      "learning_rate": 4.066196521085169e-05,
-      "loss": 0.2865,
+      "epoch": 1.5675568529931163,
+      "grad_norm": 0.1664031445980072,
+      "learning_rate": 4.0188071895725466e-05,
+      "loss": 0.4289,
       "step": 43495
     },
     {
-      "epoch": 1.53,
-      "learning_rate": 4.0659744705784254e-05,
-      "loss": 0.2946,
+      "epoch": 1.567737052654341,
+      "grad_norm": 0.17638015747070312,
+      "learning_rate": 4.018575391233882e-05,
+      "loss": 0.4044,
       "step": 43500
     },
     {
-      "epoch": 1.53,
-      "eval_loss": 0.27933764457702637,
-      "eval_runtime": 10.6369,
-      "eval_samples_per_second": 9.401,
-      "eval_steps_per_second": 9.401,
+      "epoch": 1.567737052654341,
+      "eval_loss": 0.44283291697502136,
+      "eval_runtime": 3.524,
+      "eval_samples_per_second": 28.377,
+      "eval_steps_per_second": 7.094,
       "step": 43500
     },
     {
-      "epoch": 1.53,
-      "learning_rate": 4.065752399738535e-05,
-      "loss": 0.2491,
+      "epoch": 1.5679172523155658,
+      "grad_norm": 0.19836536049842834,
+      "learning_rate": 4.018343572204921e-05,
+      "loss": 0.4234,
       "step": 43505
     },
     {
-      "epoch": 1.53,
-      "learning_rate": 4.065530308568381e-05,
-      "loss": 0.2951,
+      "epoch": 1.5680974519767903,
+      "grad_norm": 0.1390867680311203,
+      "learning_rate": 4.018111732488823e-05,
+      "loss": 0.386,
       "step": 43510
     },
     {
-      "epoch": 1.53,
-      "learning_rate": 4.0653081970708486e-05,
-      "loss": 0.2764,
+      "epoch": 1.5682776516380148,
+      "grad_norm": 0.2248731255531311,
+      "learning_rate": 4.0178798720887465e-05,
+      "loss": 0.4143,
       "step": 43515
     },
     {
-      "epoch": 1.53,
-      "learning_rate": 4.065086065248821e-05,
-      "loss": 0.2887,
+      "epoch": 1.5684578512992395,
+      "grad_norm": 0.149456188082695,
+      "learning_rate": 4.017647991007851e-05,
+      "loss": 0.3873,
       "step": 43520
     },
     {
-      "epoch": 1.53,
-      "learning_rate": 4.0648639131051815e-05,
-      "loss": 0.272,
+      "epoch": 1.5686380509604643,
+      "grad_norm": 0.20427654683589935,
+      "learning_rate": 4.017416089249296e-05,
+      "loss": 0.4319,
       "step": 43525
     },
     {
-      "epoch": 1.53,
-      "learning_rate": 4.064641740642815e-05,
-      "loss": 0.2902,
+      "epoch": 1.5688182506216888,
+      "grad_norm": 0.18678858876228333,
+      "learning_rate": 4.017184166816239e-05,
+      "loss": 0.3849,
       "step": 43530
     },
     {
-      "epoch": 1.53,
-      "learning_rate": 4.064419547864607e-05,
-      "loss": 0.2942,
+      "epoch": 1.5689984502829135,
+      "grad_norm": 0.14671590924263,
+      "learning_rate": 4.0169522237118426e-05,
+      "loss": 0.3873,
       "step": 43535
     },
     {
-      "epoch": 1.53,
-      "learning_rate": 4.064197334773443e-05,
-      "loss": 0.2664,
+      "epoch": 1.569178649944138,
+      "grad_norm": 0.1654651165008545,
+      "learning_rate": 4.0167202599392656e-05,
+      "loss": 0.4051,
       "step": 43540
     },
     {
-      "epoch": 1.53,
-      "learning_rate": 4.063975101372207e-05,
-      "loss": 0.2929,
+      "epoch": 1.5693588496053628,
+      "grad_norm": 0.17216157913208008,
+      "learning_rate": 4.0164882755016685e-05,
+      "loss": 0.4207,
       "step": 43545
     },
     {
-      "epoch": 1.53,
-      "learning_rate": 4.063752847663785e-05,
-      "loss": 0.2745,
+      "epoch": 1.5695390492665875,
+      "grad_norm": 0.1685381382703781,
+      "learning_rate": 4.0162562704022124e-05,
+      "loss": 0.3873,
       "step": 43550
     },
     {
-      "epoch": 1.53,
-      "learning_rate": 4.063530573651063e-05,
-      "loss": 0.3009,
+      "epoch": 1.569719248927812,
+      "grad_norm": 0.16307340562343597,
+      "learning_rate": 4.0160242446440584e-05,
+      "loss": 0.3893,
       "step": 43555
     },
     {
-      "epoch": 1.53,
-      "learning_rate": 4.0633082793369256e-05,
-      "loss": 0.3144,
+      "epoch": 1.5698994485890365,
+      "grad_norm": 0.2021232396364212,
+      "learning_rate": 4.015792198230367e-05,
+      "loss": 0.4492,
       "step": 43560
     },
     {
-      "epoch": 1.53,
-      "learning_rate": 4.063085964724263e-05,
-      "loss": 0.2595,
+      "epoch": 1.5700796482502613,
+      "grad_norm": 0.1811605542898178,
+      "learning_rate": 4.0155601311643006e-05,
+      "loss": 0.3913,
       "step": 43565
     },
     {
-      "epoch": 1.53,
-      "learning_rate": 4.062863629815957e-05,
-      "loss": 0.2784,
+      "epoch": 1.570259847911486,
+      "grad_norm": 0.14910484850406647,
+      "learning_rate": 4.015328043449021e-05,
+      "loss": 0.4193,
       "step": 43570
     },
     {
-      "epoch": 1.53,
-      "learning_rate": 4.062641274614897e-05,
-      "loss": 0.2907,
+      "epoch": 1.5704400475727107,
+      "grad_norm": 0.24436582624912262,
+      "learning_rate": 4.0150959350876903e-05,
+      "loss": 0.432,
       "step": 43575
     },
     {
-      "epoch": 1.53,
-      "learning_rate": 4.062418899123971e-05,
-      "loss": 0.2899,
+      "epoch": 1.5706202472339352,
+      "grad_norm": 0.1627466231584549,
+      "learning_rate": 4.014863806083471e-05,
+      "loss": 0.4316,
       "step": 43580
     },
     {
-      "epoch": 1.53,
-      "learning_rate": 4.062196503346064e-05,
-      "loss": 0.2848,
+      "epoch": 1.5708004468951597,
+      "grad_norm": 0.18843139708042145,
+      "learning_rate": 4.0146316564395254e-05,
+      "loss": 0.4146,
       "step": 43585
     },
     {
-      "epoch": 1.53,
-      "learning_rate": 4.061974087284066e-05,
-      "loss": 0.2721,
+      "epoch": 1.5709806465563845,
+      "grad_norm": 0.17627915740013123,
+      "learning_rate": 4.014399486159016e-05,
+      "loss": 0.444,
       "step": 43590
     },
     {
-      "epoch": 1.53,
-      "learning_rate": 4.0617516509408636e-05,
-      "loss": 0.2915,
+      "epoch": 1.5711608462176092,
+      "grad_norm": 0.2867834270000458,
+      "learning_rate": 4.014167295245108e-05,
+      "loss": 0.4179,
       "step": 43595
     },
     {
-      "epoch": 1.53,
-      "learning_rate": 4.0615291943193446e-05,
-      "loss": 0.2946,
+      "epoch": 1.5713410458788337,
+      "grad_norm": 0.16591203212738037,
+      "learning_rate": 4.013935083700963e-05,
+      "loss": 0.4026,
       "step": 43600
     },
     {
-      "epoch": 1.53,
-      "learning_rate": 4.0613067174223987e-05,
-      "loss": 0.2876,
+      "epoch": 1.5715212455400582,
+      "grad_norm": 0.1895548552274704,
+      "learning_rate": 4.0137028515297456e-05,
+      "loss": 0.4287,
       "step": 43605
     },
     {
-      "epoch": 1.53,
-      "learning_rate": 4.0610842202529144e-05,
-      "loss": 0.2818,
+      "epoch": 1.571701445201283,
+      "grad_norm": 0.20505109429359436,
+      "learning_rate": 4.0134705987346206e-05,
+      "loss": 0.4474,
       "step": 43610
     },
     {
-      "epoch": 1.53,
-      "learning_rate": 4.0608617028137793e-05,
-      "loss": 0.2997,
+      "epoch": 1.5718816448625077,
+      "grad_norm": 0.18428920209407806,
+      "learning_rate": 4.013238325318751e-05,
+      "loss": 0.4338,
       "step": 43615
     },
     {
-      "epoch": 1.53,
-      "learning_rate": 4.060639165107885e-05,
-      "loss": 0.2639,
+      "epoch": 1.5720618445237324,
+      "grad_norm": 0.21533146500587463,
+      "learning_rate": 4.013006031285302e-05,
+      "loss": 0.4064,
       "step": 43620
     },
     {
-      "epoch": 1.53,
-      "learning_rate": 4.060416607138118e-05,
-      "loss": 0.289,
+      "epoch": 1.572242044184957,
+      "grad_norm": 0.17549021542072296,
+      "learning_rate": 4.012773716637439e-05,
+      "loss": 0.4381,
       "step": 43625
     },
     {
-      "epoch": 1.54,
-      "learning_rate": 4.06019402890737e-05,
-      "loss": 0.3012,
+      "epoch": 1.5724222438461815,
+      "grad_norm": 0.21111688017845154,
+      "learning_rate": 4.0125413813783275e-05,
+      "loss": 0.4279,
       "step": 43630
     },
     {
-      "epoch": 1.54,
-      "learning_rate": 4.0599714304185315e-05,
-      "loss": 0.3031,
+      "epoch": 1.5726024435074062,
+      "grad_norm": 0.23240859806537628,
+      "learning_rate": 4.0123090255111316e-05,
+      "loss": 0.435,
       "step": 43635
     },
     {
-      "epoch": 1.54,
-      "learning_rate": 4.059748811674492e-05,
-      "loss": 0.2745,
+      "epoch": 1.572782643168631,
+      "grad_norm": 0.1591193825006485,
+      "learning_rate": 4.0120766490390197e-05,
+      "loss": 0.4326,
       "step": 43640
     },
     {
-      "epoch": 1.54,
-      "learning_rate": 4.059526172678142e-05,
-      "loss": 0.2801,
+      "epoch": 1.5729628428298554,
+      "grad_norm": 0.19279992580413818,
+      "learning_rate": 4.011844251965154e-05,
+      "loss": 0.4427,
       "step": 43645
     },
     {
-      "epoch": 1.54,
-      "learning_rate": 4.059303513432372e-05,
-      "loss": 0.3029,
+      "epoch": 1.5731430424910802,
+      "grad_norm": 0.19141902029514313,
+      "learning_rate": 4.0116118342927045e-05,
+      "loss": 0.3833,
       "step": 43650
     },
     {
-      "epoch": 1.54,
-      "learning_rate": 4.0590808339400735e-05,
-      "loss": 0.2762,
+      "epoch": 1.5733232421523047,
+      "grad_norm": 0.1679369956254959,
+      "learning_rate": 4.0113793960248356e-05,
+      "loss": 0.4242,
       "step": 43655
     },
     {
-      "epoch": 1.54,
-      "learning_rate": 4.0588581342041385e-05,
-      "loss": 0.2592,
+      "epoch": 1.5735034418135294,
+      "grad_norm": 0.21762417256832123,
+      "learning_rate": 4.0111469371647156e-05,
+      "loss": 0.4086,
       "step": 43660
     },
     {
-      "epoch": 1.54,
-      "learning_rate": 4.0586354142274575e-05,
-      "loss": 0.3027,
+      "epoch": 1.5736836414747541,
+      "grad_norm": 0.1647619605064392,
+      "learning_rate": 4.010914457715511e-05,
+      "loss": 0.4345,
       "step": 43665
     },
     {
-      "epoch": 1.54,
-      "learning_rate": 4.0584126740129225e-05,
-      "loss": 0.2809,
+      "epoch": 1.5738638411359787,
+      "grad_norm": 0.27651816606521606,
+      "learning_rate": 4.01068195768039e-05,
+      "loss": 0.4154,
       "step": 43670
     },
     {
-      "epoch": 1.54,
-      "learning_rate": 4.0581899135634263e-05,
-      "loss": 0.2901,
+      "epoch": 1.5740440407972032,
+      "grad_norm": 0.20368202030658722,
+      "learning_rate": 4.010449437062519e-05,
+      "loss": 0.4557,
       "step": 43675
     },
     {
-      "epoch": 1.54,
-      "learning_rate": 4.057967132881861e-05,
-      "loss": 0.2773,
+      "epoch": 1.574224240458428,
+      "grad_norm": 0.17364609241485596,
+      "learning_rate": 4.0102168958650676e-05,
+      "loss": 0.3769,
       "step": 43680
     },
     {
-      "epoch": 1.54,
-      "learning_rate": 4.0577443319711197e-05,
-      "loss": 0.2837,
+      "epoch": 1.5744044401196526,
+      "grad_norm": 0.17485357820987701,
+      "learning_rate": 4.009984334091203e-05,
+      "loss": 0.3792,
       "step": 43685
     },
     {
-      "epoch": 1.54,
-      "learning_rate": 4.057521510834093e-05,
-      "loss": 0.2884,
+      "epoch": 1.5745846397808774,
+      "grad_norm": 0.20240798592567444,
+      "learning_rate": 4.009751751744094e-05,
+      "loss": 0.3928,
       "step": 43690
     },
     {
-      "epoch": 1.54,
-      "learning_rate": 4.0572986694736784e-05,
-      "loss": 0.2677,
+      "epoch": 1.5747648394421019,
+      "grad_norm": 0.21913383901119232,
+      "learning_rate": 4.009519148826909e-05,
+      "loss": 0.4397,
       "step": 43695
     },
     {
-      "epoch": 1.54,
-      "learning_rate": 4.057075807892765e-05,
-      "loss": 0.2645,
+      "epoch": 1.5749450391033264,
+      "grad_norm": 0.1732109934091568,
+      "learning_rate": 4.009286525342819e-05,
+      "loss": 0.3939,
       "step": 43700
     },
     {
-      "epoch": 1.54,
-      "learning_rate": 4.05685292609425e-05,
-      "loss": 0.2882,
+      "epoch": 1.5751252387645511,
+      "grad_norm": 0.20331084728240967,
+      "learning_rate": 4.0090538812949916e-05,
+      "loss": 0.4074,
       "step": 43705
     },
     {
-      "epoch": 1.54,
-      "learning_rate": 4.0566300240810245e-05,
-      "loss": 0.285,
+      "epoch": 1.5753054384257759,
+      "grad_norm": 0.18988221883773804,
+      "learning_rate": 4.008821216686598e-05,
+      "loss": 0.3991,
       "step": 43710
     },
     {
-      "epoch": 1.54,
-      "learning_rate": 4.056407101855985e-05,
-      "loss": 0.277,
+      "epoch": 1.5754856380870004,
+      "grad_norm": 0.17192235589027405,
+      "learning_rate": 4.008588531520807e-05,
+      "loss": 0.4486,
       "step": 43715
     },
     {
-      "epoch": 1.54,
-      "learning_rate": 4.056184159422024e-05,
-      "loss": 0.2951,
+      "epoch": 1.5756658377482249,
+      "grad_norm": 0.19935615360736847,
+      "learning_rate": 4.008355825800789e-05,
+      "loss": 0.4195,
       "step": 43720
     },
     {
-      "epoch": 1.54,
-      "learning_rate": 4.0559611967820375e-05,
-      "loss": 0.2924,
+      "epoch": 1.5758460374094496,
+      "grad_norm": 0.18798862397670746,
+      "learning_rate": 4.0081230995297154e-05,
+      "loss": 0.397,
       "step": 43725
     },
     {
-      "epoch": 1.54,
-      "learning_rate": 4.05573821393892e-05,
-      "loss": 0.3046,
+      "epoch": 1.5760262370706744,
+      "grad_norm": 0.16383147239685059,
+      "learning_rate": 4.007890352710757e-05,
+      "loss": 0.4065,
       "step": 43730
     },
     {
-      "epoch": 1.54,
-      "learning_rate": 4.055515210895568e-05,
-      "loss": 0.3116,
+      "epoch": 1.576206436731899,
+      "grad_norm": 0.18491680920124054,
+      "learning_rate": 4.007657585347083e-05,
+      "loss": 0.397,
       "step": 43735
     },
     {
-      "epoch": 1.54,
-      "learning_rate": 4.055292187654875e-05,
-      "loss": 0.2598,
+      "epoch": 1.5763866363931236,
+      "grad_norm": 0.1732751429080963,
+      "learning_rate": 4.007424797441868e-05,
+      "loss": 0.3996,
       "step": 43740
     },
     {
-      "epoch": 1.54,
-      "learning_rate": 4.055069144219739e-05,
-      "loss": 0.3018,
+      "epoch": 1.576566836054348,
+      "grad_norm": 0.17933018505573273,
+      "learning_rate": 4.00719198899828e-05,
+      "loss": 0.427,
       "step": 43745
     },
     {
-      "epoch": 1.54,
-      "learning_rate": 4.054846080593054e-05,
-      "loss": 0.3049,
+      "epoch": 1.5767470357155728,
+      "grad_norm": 0.17597095668315887,
+      "learning_rate": 4.006959160019495e-05,
+      "loss": 0.4075,
       "step": 43750
     },
     {
-      "epoch": 1.54,
-      "learning_rate": 4.054622996777718e-05,
-      "loss": 0.3057,
+      "epoch": 1.5769272353767976,
+      "grad_norm": 0.16884708404541016,
+      "learning_rate": 4.0067263105086825e-05,
+      "loss": 0.4244,
       "step": 43755
     },
     {
-      "epoch": 1.54,
-      "learning_rate": 4.054399892776627e-05,
-      "loss": 0.2754,
+      "epoch": 1.577107435038022,
+      "grad_norm": 0.21198032796382904,
+      "learning_rate": 4.0064934404690146e-05,
+      "loss": 0.4058,
       "step": 43760
     },
     {
-      "epoch": 1.54,
-      "learning_rate": 4.054176768592678e-05,
-      "loss": 0.2652,
+      "epoch": 1.5772876346992466,
+      "grad_norm": 0.16649512946605682,
+      "learning_rate": 4.006260549903666e-05,
+      "loss": 0.4174,
       "step": 43765
     },
     {
-      "epoch": 1.54,
-      "learning_rate": 4.053953624228767e-05,
-      "loss": 0.2693,
+      "epoch": 1.5774678343604713,
+      "grad_norm": 0.16471701860427856,
+      "learning_rate": 4.00602763881581e-05,
+      "loss": 0.3756,
       "step": 43770
     },
     {
-      "epoch": 1.54,
-      "learning_rate": 4.053730459687792e-05,
-      "loss": 0.297,
+      "epoch": 1.577648034021696,
+      "grad_norm": 0.1989121437072754,
+      "learning_rate": 4.005794707208618e-05,
+      "loss": 0.4276,
       "step": 43775
     },
     {
-      "epoch": 1.54,
-      "learning_rate": 4.053507274972652e-05,
-      "loss": 0.264,
+      "epoch": 1.5778282336829208,
+      "grad_norm": 0.14594782888889313,
+      "learning_rate": 4.005561755085265e-05,
+      "loss": 0.4355,
       "step": 43780
     },
     {
-      "epoch": 1.54,
-      "learning_rate": 4.053284070086244e-05,
-      "loss": 0.314,
+      "epoch": 1.5780084333441453,
+      "grad_norm": 0.19597959518432617,
+      "learning_rate": 4.0053287824489236e-05,
+      "loss": 0.4134,
       "step": 43785
     },
     {
-      "epoch": 1.54,
-      "learning_rate": 4.0530608450314654e-05,
-      "loss": 0.2932,
+      "epoch": 1.5781886330053698,
+      "grad_norm": 0.17923668026924133,
+      "learning_rate": 4.0050957893027706e-05,
+      "loss": 0.4061,
       "step": 43790
     },
     {
-      "epoch": 1.54,
-      "learning_rate": 4.052837599811216e-05,
-      "loss": 0.3142,
+      "epoch": 1.5783688326665946,
+      "grad_norm": 0.1501871943473816,
+      "learning_rate": 4.004862775649978e-05,
+      "loss": 0.3867,
       "step": 43795
     },
     {
-      "epoch": 1.54,
-      "learning_rate": 4.0526143344283916e-05,
-      "loss": 0.2902,
+      "epoch": 1.5785490323278193,
+      "grad_norm": 0.18202674388885498,
+      "learning_rate": 4.004629741493721e-05,
+      "loss": 0.4072,
       "step": 43800
     },
     {
-      "epoch": 1.54,
-      "learning_rate": 4.0523910488858954e-05,
-      "loss": 0.294,
+      "epoch": 1.578729231989044,
+      "grad_norm": 0.13299371302127838,
+      "learning_rate": 4.004396686837176e-05,
+      "loss": 0.4034,
       "step": 43805
     },
     {
-      "epoch": 1.54,
-      "learning_rate": 4.052167743186623e-05,
-      "loss": 0.3107,
+      "epoch": 1.5789094316502685,
+      "grad_norm": 0.17386196553707123,
+      "learning_rate": 4.004163611683517e-05,
+      "loss": 0.4056,
       "step": 43810
     },
     {
-      "epoch": 1.54,
-      "learning_rate": 4.0519444173334766e-05,
-      "loss": 0.2692,
+      "epoch": 1.579089631311493,
+      "grad_norm": 0.18044842779636383,
+      "learning_rate": 4.0039305160359195e-05,
+      "loss": 0.4121,
       "step": 43815
     },
     {
-      "epoch": 1.54,
-      "learning_rate": 4.051721071329354e-05,
-      "loss": 0.2847,
+      "epoch": 1.5792698309727178,
+      "grad_norm": 0.15430901944637299,
+      "learning_rate": 4.0036973998975604e-05,
+      "loss": 0.3994,
       "step": 43820
     },
     {
-      "epoch": 1.54,
-      "learning_rate": 4.0514977051771566e-05,
-      "loss": 0.2965,
+      "epoch": 1.5794500306339425,
+      "grad_norm": 0.17539086937904358,
+      "learning_rate": 4.0034642632716155e-05,
+      "loss": 0.4177,
       "step": 43825
     },
     {
-      "epoch": 1.54,
-      "learning_rate": 4.051274318879784e-05,
-      "loss": 0.2907,
+      "epoch": 1.579630230295167,
+      "grad_norm": 0.19086940586566925,
+      "learning_rate": 4.0032311061612604e-05,
+      "loss": 0.4033,
       "step": 43830
     },
     {
-      "epoch": 1.54,
-      "learning_rate": 4.051050912440135e-05,
-      "loss": 0.2911,
+      "epoch": 1.5798104299563915,
+      "grad_norm": 0.19829294085502625,
+      "learning_rate": 4.0029979285696736e-05,
+      "loss": 0.431,
       "step": 43835
     },
     {
-      "epoch": 1.54,
-      "learning_rate": 4.050827485861114e-05,
-      "loss": 0.2847,
+      "epoch": 1.5799906296176163,
+      "grad_norm": 0.21436335146427155,
+      "learning_rate": 4.0027647305000306e-05,
+      "loss": 0.4564,
       "step": 43840
     },
     {
-      "epoch": 1.54,
-      "learning_rate": 4.050604039145619e-05,
-      "loss": 0.2794,
+      "epoch": 1.580170829278841,
+      "grad_norm": 0.21377906203269958,
+      "learning_rate": 4.002531511955509e-05,
+      "loss": 0.441,
       "step": 43845
     },
     {
-      "epoch": 1.54,
-      "learning_rate": 4.0503805722965535e-05,
-      "loss": 0.2913,
+      "epoch": 1.5803510289400657,
+      "grad_norm": 0.1993681788444519,
+      "learning_rate": 4.0022982729392855e-05,
+      "loss": 0.4371,
       "step": 43850
     },
     {
-      "epoch": 1.54,
-      "learning_rate": 4.0501570853168166e-05,
-      "loss": 0.2668,
+      "epoch": 1.5805312286012903,
+      "grad_norm": 0.16495195031166077,
+      "learning_rate": 4.00206501345454e-05,
+      "loss": 0.436,
       "step": 43855
     },
     {
-      "epoch": 1.54,
-      "learning_rate": 4.049933578209313e-05,
-      "loss": 0.3133,
+      "epoch": 1.5807114282625148,
+      "grad_norm": 0.16763421893119812,
+      "learning_rate": 4.0018317335044495e-05,
+      "loss": 0.4238,
       "step": 43860
     },
     {
-      "epoch": 1.54,
-      "learning_rate": 4.049710050976943e-05,
-      "loss": 0.2685,
+      "epoch": 1.5808916279237395,
+      "grad_norm": 0.16144713759422302,
+      "learning_rate": 4.0015984330921916e-05,
+      "loss": 0.4115,
       "step": 43865
     },
     {
-      "epoch": 1.54,
-      "learning_rate": 4.049486503622609e-05,
-      "loss": 0.289,
+      "epoch": 1.5810718275849642,
+      "grad_norm": 0.20617428421974182,
+      "learning_rate": 4.0013651122209465e-05,
+      "loss": 0.4347,
       "step": 43870
     },
     {
-      "epoch": 1.54,
-      "learning_rate": 4.049262936149213e-05,
-      "loss": 0.2724,
+      "epoch": 1.5812520272461887,
+      "grad_norm": 0.1693793535232544,
+      "learning_rate": 4.0011317708938924e-05,
+      "loss": 0.3648,
       "step": 43875
     },
     {
-      "epoch": 1.54,
-      "learning_rate": 4.0490393485596605e-05,
-      "loss": 0.2723,
+      "epoch": 1.5814322269074133,
+      "grad_norm": 0.15378816425800323,
+      "learning_rate": 4.0008984091142086e-05,
+      "loss": 0.4536,
       "step": 43880
     },
     {
-      "epoch": 1.54,
-      "learning_rate": 4.0488157408568516e-05,
-      "loss": 0.2928,
+      "epoch": 1.581612426568638,
+      "grad_norm": 0.20745548605918884,
+      "learning_rate": 4.0006650268850745e-05,
+      "loss": 0.4071,
       "step": 43885
     },
     {
-      "epoch": 1.54,
-      "learning_rate": 4.048592113043692e-05,
-      "loss": 0.2693,
+      "epoch": 1.5817926262298627,
+      "grad_norm": 0.2162124514579773,
+      "learning_rate": 4.00043162420967e-05,
+      "loss": 0.4081,
       "step": 43890
     },
     {
-      "epoch": 1.54,
-      "learning_rate": 4.0483684651230835e-05,
-      "loss": 0.3124,
+      "epoch": 1.5819728258910875,
+      "grad_norm": 0.16745862364768982,
+      "learning_rate": 4.000198201091175e-05,
+      "loss": 0.4493,
       "step": 43895
     },
     {
-      "epoch": 1.54,
-      "learning_rate": 4.0481447970979316e-05,
-      "loss": 0.2454,
+      "epoch": 1.582153025552312,
+      "grad_norm": 0.16501225531101227,
+      "learning_rate": 3.999964757532769e-05,
+      "loss": 0.3873,
       "step": 43900
     },
     {
-      "epoch": 1.54,
-      "learning_rate": 4.047921108971139e-05,
-      "loss": 0.2742,
+      "epoch": 1.5823332252135365,
+      "grad_norm": 0.22815780341625214,
+      "learning_rate": 3.9997312935376346e-05,
+      "loss": 0.3807,
       "step": 43905
     },
     {
-      "epoch": 1.54,
-      "learning_rate": 4.0476974007456114e-05,
-      "loss": 0.2796,
+      "epoch": 1.5825134248747612,
+      "grad_norm": 0.17533724009990692,
+      "learning_rate": 3.9994978091089515e-05,
+      "loss": 0.4033,
       "step": 43910
     },
     {
-      "epoch": 1.55,
-      "learning_rate": 4.047473672424253e-05,
-      "loss": 0.2946,
+      "epoch": 1.582693624535986,
+      "grad_norm": 0.1684589833021164,
+      "learning_rate": 3.999264304249901e-05,
+      "loss": 0.4166,
       "step": 43915
     },
     {
-      "epoch": 1.55,
-      "learning_rate": 4.0472499240099695e-05,
-      "loss": 0.309,
+      "epoch": 1.5828738241972105,
+      "grad_norm": 0.16925841569900513,
+      "learning_rate": 3.999030778963665e-05,
+      "loss": 0.394,
       "step": 43920
     },
     {
-      "epoch": 1.55,
-      "learning_rate": 4.047026155505664e-05,
-      "loss": 0.288,
+      "epoch": 1.5830540238584352,
+      "grad_norm": 0.1653861552476883,
+      "learning_rate": 3.9987972332534246e-05,
+      "loss": 0.4332,
       "step": 43925
     },
     {
-      "epoch": 1.55,
-      "learning_rate": 4.046802366914244e-05,
-      "loss": 0.286,
+      "epoch": 1.5832342235196597,
+      "grad_norm": 0.1570800542831421,
+      "learning_rate": 3.998563667122362e-05,
+      "loss": 0.3974,
       "step": 43930
     },
     {
-      "epoch": 1.55,
-      "learning_rate": 4.046578558238615e-05,
-      "loss": 0.2997,
+      "epoch": 1.5834144231808844,
+      "grad_norm": 0.18440444767475128,
+      "learning_rate": 3.9983300805736595e-05,
+      "loss": 0.4235,
       "step": 43935
     },
     {
-      "epoch": 1.55,
-      "learning_rate": 4.0463547294816815e-05,
-      "loss": 0.2772,
+      "epoch": 1.5835946228421092,
+      "grad_norm": 0.15225686132907867,
+      "learning_rate": 3.9980964736104995e-05,
+      "loss": 0.4064,
       "step": 43940
     },
     {
-      "epoch": 1.55,
-      "learning_rate": 4.0461308806463525e-05,
-      "loss": 0.2691,
+      "epoch": 1.5837748225033337,
+      "grad_norm": 0.17495448887348175,
+      "learning_rate": 3.997862846236066e-05,
+      "loss": 0.4167,
       "step": 43945
     },
     {
-      "epoch": 1.55,
-      "learning_rate": 4.045907011735532e-05,
-      "loss": 0.2804,
+      "epoch": 1.5839550221645582,
+      "grad_norm": 0.1724550575017929,
+      "learning_rate": 3.9976291984535405e-05,
+      "loss": 0.4093,
       "step": 43950
     },
     {
-      "epoch": 1.55,
-      "learning_rate": 4.045683122752128e-05,
-      "loss": 0.2867,
+      "epoch": 1.584135221825783,
+      "grad_norm": 0.19340504705905914,
+      "learning_rate": 3.997395530266108e-05,
+      "loss": 0.4351,
       "step": 43955
     },
     {
-      "epoch": 1.55,
-      "learning_rate": 4.045459213699047e-05,
-      "loss": 0.2909,
+      "epoch": 1.5843154214870077,
+      "grad_norm": 0.18444287776947021,
+      "learning_rate": 3.9971618416769495e-05,
+      "loss": 0.3998,
       "step": 43960
     },
     {
-      "epoch": 1.55,
-      "learning_rate": 4.045235284579198e-05,
-      "loss": 0.3073,
+      "epoch": 1.5844956211482324,
+      "grad_norm": 0.13938504457473755,
+      "learning_rate": 3.9969281326892523e-05,
+      "loss": 0.3954,
       "step": 43965
     },
     {
-      "epoch": 1.55,
-      "learning_rate": 4.045011335395486e-05,
-      "loss": 0.3136,
+      "epoch": 1.584675820809457,
+      "grad_norm": 0.15549679100513458,
+      "learning_rate": 3.996694403306198e-05,
+      "loss": 0.4072,
       "step": 43970
     },
     {
-      "epoch": 1.55,
-      "learning_rate": 4.04478736615082e-05,
-      "loss": 0.288,
+      "epoch": 1.5848560204706814,
+      "grad_norm": 0.16593880951404572,
+      "learning_rate": 3.9964606535309735e-05,
+      "loss": 0.4273,
       "step": 43975
     },
     {
-      "epoch": 1.55,
-      "learning_rate": 4.044563376848107e-05,
-      "loss": 0.2732,
+      "epoch": 1.5850362201319061,
+      "grad_norm": 0.23627829551696777,
+      "learning_rate": 3.9962268833667615e-05,
+      "loss": 0.4176,
       "step": 43980
     },
     {
-      "epoch": 1.55,
-      "learning_rate": 4.0443393674902585e-05,
-      "loss": 0.3115,
+      "epoch": 1.5852164197931309,
+      "grad_norm": 0.1789269745349884,
+      "learning_rate": 3.9959930928167474e-05,
+      "loss": 0.4231,
       "step": 43985
     },
     {
-      "epoch": 1.55,
-      "learning_rate": 4.04411533808018e-05,
-      "loss": 0.2983,
+      "epoch": 1.5853966194543554,
+      "grad_norm": 0.20784516632556915,
+      "learning_rate": 3.995759281884118e-05,
+      "loss": 0.4042,
       "step": 43990
     },
     {
-      "epoch": 1.55,
-      "learning_rate": 4.043891288620781e-05,
-      "loss": 0.2848,
+      "epoch": 1.58557681911558,
+      "grad_norm": 0.18998631834983826,
+      "learning_rate": 3.995525450572059e-05,
+      "loss": 0.4483,
       "step": 43995
     },
     {
-      "epoch": 1.55,
-      "learning_rate": 4.0436672191149724e-05,
-      "loss": 0.2671,
+      "epoch": 1.5857570187768046,
+      "grad_norm": 0.2027607262134552,
+      "learning_rate": 3.9952915988837534e-05,
+      "loss": 0.3861,
       "step": 44000
     },
     {
-      "epoch": 1.55,
-      "eval_loss": 0.279005765914917,
-      "eval_runtime": 10.5327,
-      "eval_samples_per_second": 9.494,
-      "eval_steps_per_second": 9.494,
+      "epoch": 1.5857570187768046,
+      "eval_loss": 0.44330334663391113,
+      "eval_runtime": 3.5296,
+      "eval_samples_per_second": 28.332,
+      "eval_steps_per_second": 7.083,
       "step": 44000
     },
     {
-      "epoch": 1.55,
-      "learning_rate": 4.043443129565662e-05,
-      "loss": 0.2622,
+      "epoch": 1.5859372184380294,
+      "grad_norm": 0.18675218522548676,
+      "learning_rate": 3.99505772682239e-05,
+      "loss": 0.399,
       "step": 44005
     },
     {
-      "epoch": 1.55,
-      "learning_rate": 4.043219019975759e-05,
-      "loss": 0.2953,
+      "epoch": 1.586117418099254,
+      "grad_norm": 0.18582110106945038,
+      "learning_rate": 3.994823834391154e-05,
+      "loss": 0.3755,
       "step": 44010
     },
     {
-      "epoch": 1.55,
-      "learning_rate": 4.042994890348174e-05,
-      "loss": 0.2774,
+      "epoch": 1.5862976177604786,
+      "grad_norm": 0.18552209436893463,
+      "learning_rate": 3.994589921593233e-05,
+      "loss": 0.4288,
       "step": 44015
     },
     {
-      "epoch": 1.55,
-      "learning_rate": 4.042770740685818e-05,
-      "loss": 0.3034,
+      "epoch": 1.5864778174217031,
+      "grad_norm": 0.17207638919353485,
+      "learning_rate": 3.994355988431814e-05,
+      "loss": 0.3947,
       "step": 44020
     },
     {
-      "epoch": 1.55,
-      "learning_rate": 4.0425465709916e-05,
-      "loss": 0.2754,
+      "epoch": 1.5866580170829279,
+      "grad_norm": 0.22943609952926636,
+      "learning_rate": 3.994122034910083e-05,
+      "loss": 0.4271,
       "step": 44025
     },
     {
-      "epoch": 1.55,
-      "learning_rate": 4.042322381268432e-05,
-      "loss": 0.2879,
+      "epoch": 1.5868382167441526,
+      "grad_norm": 0.21029749512672424,
+      "learning_rate": 3.9938880610312294e-05,
+      "loss": 0.4071,
       "step": 44030
     },
     {
-      "epoch": 1.55,
-      "learning_rate": 4.0420981715192235e-05,
-      "loss": 0.297,
+      "epoch": 1.587018416405377,
+      "grad_norm": 0.1644810289144516,
+      "learning_rate": 3.99365406679844e-05,
+      "loss": 0.3875,
       "step": 44035
     },
     {
-      "epoch": 1.55,
-      "learning_rate": 4.041873941746887e-05,
-      "loss": 0.282,
+      "epoch": 1.5871986160666018,
+      "grad_norm": 0.1785242259502411,
+      "learning_rate": 3.993420052214904e-05,
+      "loss": 0.3676,
       "step": 44040
     },
     {
-      "epoch": 1.55,
-      "learning_rate": 4.0416496919543336e-05,
-      "loss": 0.3116,
+      "epoch": 1.5873788157278264,
+      "grad_norm": 0.17535260319709778,
+      "learning_rate": 3.9931860172838076e-05,
+      "loss": 0.3939,
       "step": 44045
     },
     {
-      "epoch": 1.55,
-      "learning_rate": 4.041425422144475e-05,
-      "loss": 0.2781,
+      "epoch": 1.587559015389051,
+      "grad_norm": 0.17083501815795898,
+      "learning_rate": 3.992951962008341e-05,
+      "loss": 0.4247,
       "step": 44050
     },
     {
-      "epoch": 1.55,
-      "learning_rate": 4.041201132320223e-05,
-      "loss": 0.2828,
+      "epoch": 1.5877392150502758,
+      "grad_norm": 0.230384960770607,
+      "learning_rate": 3.992717886391693e-05,
+      "loss": 0.4225,
       "step": 44055
     },
     {
-      "epoch": 1.55,
-      "learning_rate": 4.040976822484489e-05,
-      "loss": 0.3013,
+      "epoch": 1.5879194147115003,
+      "grad_norm": 0.1916215866804123,
+      "learning_rate": 3.992483790437054e-05,
+      "loss": 0.4315,
       "step": 44060
     },
     {
-      "epoch": 1.55,
-      "learning_rate": 4.0407524926401874e-05,
-      "loss": 0.3005,
+      "epoch": 1.5880996143727248,
+      "grad_norm": 0.17122019827365875,
+      "learning_rate": 3.992249674147611e-05,
+      "loss": 0.409,
       "step": 44065
     },
     {
-      "epoch": 1.55,
-      "learning_rate": 4.04052814279023e-05,
-      "loss": 0.3076,
+      "epoch": 1.5882798140339496,
+      "grad_norm": 0.17510074377059937,
+      "learning_rate": 3.9920155375265555e-05,
+      "loss": 0.4144,
       "step": 44070
     },
     {
-      "epoch": 1.55,
-      "learning_rate": 4.04030377293753e-05,
-      "loss": 0.3003,
+      "epoch": 1.5884600136951743,
+      "grad_norm": 0.1663195937871933,
+      "learning_rate": 3.991781380577076e-05,
+      "loss": 0.4265,
       "step": 44075
     },
     {
-      "epoch": 1.55,
-      "learning_rate": 4.040079383085001e-05,
-      "loss": 0.2927,
+      "epoch": 1.588640213356399,
+      "grad_norm": 0.18738073110580444,
+      "learning_rate": 3.991547203302366e-05,
+      "loss": 0.4012,
       "step": 44080
     },
     {
-      "epoch": 1.55,
-      "learning_rate": 4.039854973235556e-05,
-      "loss": 0.2855,
+      "epoch": 1.5888204130176236,
+      "grad_norm": 0.1917044073343277,
+      "learning_rate": 3.991313005705613e-05,
+      "loss": 0.4378,
       "step": 44085
     },
     {
-      "epoch": 1.55,
-      "learning_rate": 4.0396305433921086e-05,
-      "loss": 0.2697,
+      "epoch": 1.589000612678848,
+      "grad_norm": 0.2828647792339325,
+      "learning_rate": 3.991078787790009e-05,
+      "loss": 0.3974,
       "step": 44090
     },
     {
-      "epoch": 1.55,
-      "learning_rate": 4.039406093557573e-05,
-      "loss": 0.2696,
+      "epoch": 1.5891808123400728,
+      "grad_norm": 0.17092807590961456,
+      "learning_rate": 3.990844549558745e-05,
+      "loss": 0.4047,
       "step": 44095
     },
     {
-      "epoch": 1.55,
-      "learning_rate": 4.0391816237348646e-05,
-      "loss": 0.3029,
+      "epoch": 1.5893610120012975,
+      "grad_norm": 0.1954643726348877,
+      "learning_rate": 3.990610291015014e-05,
+      "loss": 0.4167,
       "step": 44100
     },
     {
-      "epoch": 1.55,
-      "learning_rate": 4.038957133926896e-05,
-      "loss": 0.2862,
+      "epoch": 1.589541211662522,
+      "grad_norm": 0.2312636822462082,
+      "learning_rate": 3.990376012162006e-05,
+      "loss": 0.4022,
       "step": 44105
     },
     {
-      "epoch": 1.55,
-      "learning_rate": 4.038732624136584e-05,
-      "loss": 0.2742,
+      "epoch": 1.5897214113237466,
+      "grad_norm": 0.1981632262468338,
+      "learning_rate": 3.990141713002912e-05,
+      "loss": 0.4631,
       "step": 44110
     },
     {
-      "epoch": 1.55,
-      "learning_rate": 4.038508094366843e-05,
-      "loss": 0.2863,
+      "epoch": 1.5899016109849713,
+      "grad_norm": 0.16944488883018494,
+      "learning_rate": 3.989907393540927e-05,
+      "loss": 0.3809,
       "step": 44115
     },
     {
-      "epoch": 1.55,
-      "learning_rate": 4.038283544620588e-05,
-      "loss": 0.2809,
+      "epoch": 1.590081810646196,
+      "grad_norm": 0.1728655993938446,
+      "learning_rate": 3.9896730537792415e-05,
+      "loss": 0.3891,
       "step": 44120
     },
     {
-      "epoch": 1.55,
-      "learning_rate": 4.0380589749007357e-05,
-      "loss": 0.2874,
+      "epoch": 1.5902620103074208,
+      "grad_norm": 0.19275811314582825,
+      "learning_rate": 3.989438693721049e-05,
+      "loss": 0.4316,
       "step": 44125
     },
     {
-      "epoch": 1.55,
-      "learning_rate": 4.037834385210201e-05,
-      "loss": 0.2926,
+      "epoch": 1.5904422099686453,
+      "grad_norm": 0.16822832822799683,
+      "learning_rate": 3.989204313369543e-05,
+      "loss": 0.4205,
       "step": 44130
     },
     {
-      "epoch": 1.55,
-      "learning_rate": 4.037609775551899e-05,
-      "loss": 0.2669,
+      "epoch": 1.5906224096298698,
+      "grad_norm": 0.16480205953121185,
+      "learning_rate": 3.9889699127279164e-05,
+      "loss": 0.435,
       "step": 44135
     },
     {
-      "epoch": 1.55,
-      "learning_rate": 4.037385145928749e-05,
-      "loss": 0.2645,
+      "epoch": 1.5908026092910945,
+      "grad_norm": 0.1605161726474762,
+      "learning_rate": 3.9887354917993635e-05,
+      "loss": 0.4301,
       "step": 44140
     },
     {
-      "epoch": 1.55,
-      "learning_rate": 4.0371604963436656e-05,
-      "loss": 0.2976,
+      "epoch": 1.5909828089523192,
+      "grad_norm": 0.16853894293308258,
+      "learning_rate": 3.988501050587078e-05,
+      "loss": 0.4255,
       "step": 44145
     },
     {
-      "epoch": 1.55,
-      "learning_rate": 4.036935826799566e-05,
-      "loss": 0.2779,
+      "epoch": 1.5911630086135438,
+      "grad_norm": 0.20577529072761536,
+      "learning_rate": 3.9882665890942526e-05,
+      "loss": 0.3954,
       "step": 44150
     },
     {
-      "epoch": 1.55,
-      "learning_rate": 4.0367111372993674e-05,
-      "loss": 0.2729,
+      "epoch": 1.5913432082747685,
+      "grad_norm": 0.15664778649806976,
+      "learning_rate": 3.988032107324084e-05,
+      "loss": 0.4088,
       "step": 44155
     },
     {
-      "epoch": 1.55,
-      "learning_rate": 4.036486427845987e-05,
-      "loss": 0.2845,
+      "epoch": 1.591523407935993,
+      "grad_norm": 0.22527268528938293,
+      "learning_rate": 3.987797605279766e-05,
+      "loss": 0.4338,
       "step": 44160
     },
     {
-      "epoch": 1.55,
-      "learning_rate": 4.036261698442344e-05,
-      "loss": 0.276,
+      "epoch": 1.5917036075972177,
+      "grad_norm": 0.15542291104793549,
+      "learning_rate": 3.987563082964493e-05,
+      "loss": 0.4333,
       "step": 44165
     },
     {
-      "epoch": 1.55,
-      "learning_rate": 4.036036949091354e-05,
-      "loss": 0.2767,
+      "epoch": 1.5918838072584425,
+      "grad_norm": 0.18337464332580566,
+      "learning_rate": 3.987328540381461e-05,
+      "loss": 0.4064,
       "step": 44170
     },
     {
-      "epoch": 1.55,
-      "learning_rate": 4.035812179795938e-05,
-      "loss": 0.2749,
+      "epoch": 1.592064006919667,
+      "grad_norm": 0.21346335113048553,
+      "learning_rate": 3.987093977533867e-05,
+      "loss": 0.4588,
       "step": 44175
     },
     {
-      "epoch": 1.55,
-      "learning_rate": 4.035587390559012e-05,
-      "loss": 0.2965,
+      "epoch": 1.5922442065808915,
+      "grad_norm": 0.1897256225347519,
+      "learning_rate": 3.986859394424905e-05,
+      "loss": 0.44,
       "step": 44180
     },
     {
-      "epoch": 1.55,
-      "learning_rate": 4.035362581383496e-05,
-      "loss": 0.3134,
+      "epoch": 1.5924244062421162,
+      "grad_norm": 0.1959022730588913,
+      "learning_rate": 3.98662479105777e-05,
+      "loss": 0.4221,
       "step": 44185
     },
     {
-      "epoch": 1.55,
-      "learning_rate": 4.035137752272309e-05,
-      "loss": 0.2744,
+      "epoch": 1.592604605903341,
+      "grad_norm": 0.20452988147735596,
+      "learning_rate": 3.986390167435661e-05,
+      "loss": 0.4131,
       "step": 44190
     },
     {
-      "epoch": 1.55,
-      "learning_rate": 4.0349129032283694e-05,
-      "loss": 0.2862,
+      "epoch": 1.5927848055645657,
+      "grad_norm": 0.18788976967334747,
+      "learning_rate": 3.9861555235617734e-05,
+      "loss": 0.4157,
       "step": 44195
     },
     {
-      "epoch": 1.56,
-      "learning_rate": 4.0346880342545975e-05,
-      "loss": 0.2865,
+      "epoch": 1.5929650052257902,
+      "grad_norm": 0.14749525487422943,
+      "learning_rate": 3.985920859439306e-05,
+      "loss": 0.371,
       "step": 44200
     },
     {
-      "epoch": 1.56,
-      "learning_rate": 4.034463145353913e-05,
-      "loss": 0.2837,
+      "epoch": 1.5931452048870147,
+      "grad_norm": 0.19985322654247284,
+      "learning_rate": 3.9856861750714535e-05,
+      "loss": 0.4088,
       "step": 44205
     },
     {
-      "epoch": 1.56,
-      "learning_rate": 4.034238236529235e-05,
-      "loss": 0.2748,
+      "epoch": 1.5933254045482395,
+      "grad_norm": 0.18450570106506348,
+      "learning_rate": 3.985451470461414e-05,
+      "loss": 0.3813,
       "step": 44210
     },
     {
-      "epoch": 1.56,
-      "learning_rate": 4.034013307783486e-05,
-      "loss": 0.3044,
+      "epoch": 1.5935056042094642,
+      "grad_norm": 0.26536038517951965,
+      "learning_rate": 3.985216745612387e-05,
+      "loss": 0.3779,
       "step": 44215
     },
     {
-      "epoch": 1.56,
-      "learning_rate": 4.033788359119585e-05,
-      "loss": 0.2856,
+      "epoch": 1.5936858038706887,
+      "grad_norm": 0.17288127541542053,
+      "learning_rate": 3.984982000527568e-05,
+      "loss": 0.4086,
       "step": 44220
     },
     {
-      "epoch": 1.56,
-      "learning_rate": 4.0335633905404526e-05,
-      "loss": 0.307,
+      "epoch": 1.5938660035319132,
+      "grad_norm": 0.19189390540122986,
+      "learning_rate": 3.984747235210158e-05,
+      "loss": 0.4554,
       "step": 44225
     },
     {
-      "epoch": 1.56,
-      "learning_rate": 4.033338402049009e-05,
-      "loss": 0.2752,
+      "epoch": 1.594046203193138,
+      "grad_norm": 0.17613013088703156,
+      "learning_rate": 3.984512449663353e-05,
+      "loss": 0.3713,
       "step": 44230
     },
     {
-      "epoch": 1.56,
-      "learning_rate": 4.033113393648178e-05,
-      "loss": 0.2831,
+      "epoch": 1.5942264028543627,
+      "grad_norm": 0.15496566891670227,
+      "learning_rate": 3.984277643890355e-05,
+      "loss": 0.3943,
       "step": 44235
     },
     {
-      "epoch": 1.56,
-      "learning_rate": 4.03288836534088e-05,
-      "loss": 0.2967,
+      "epoch": 1.5944066025155874,
+      "grad_norm": 0.16410352289676666,
+      "learning_rate": 3.98404281789436e-05,
+      "loss": 0.426,
       "step": 44240
     },
     {
-      "epoch": 1.56,
-      "learning_rate": 4.0326633171300363e-05,
-      "loss": 0.2885,
+      "epoch": 1.594586802176812,
+      "grad_norm": 0.19269990921020508,
+      "learning_rate": 3.9838079716785704e-05,
+      "loss": 0.3917,
       "step": 44245
     },
     {
-      "epoch": 1.56,
-      "learning_rate": 4.03243824901857e-05,
-      "loss": 0.2944,
+      "epoch": 1.5947670018380364,
+      "grad_norm": 0.19066692888736725,
+      "learning_rate": 3.983573105246183e-05,
+      "loss": 0.4285,
       "step": 44250
     },
     {
-      "epoch": 1.56,
-      "learning_rate": 4.032213161009403e-05,
-      "loss": 0.2789,
+      "epoch": 1.5949472014992612,
+      "grad_norm": 0.17595848441123962,
+      "learning_rate": 3.9833382186004005e-05,
+      "loss": 0.4524,
       "step": 44255
     },
     {
-      "epoch": 1.56,
-      "learning_rate": 4.031988053105457e-05,
-      "loss": 0.2865,
+      "epoch": 1.595127401160486,
+      "grad_norm": 0.15249063074588776,
+      "learning_rate": 3.983103311744421e-05,
+      "loss": 0.4218,
       "step": 44260
     },
     {
-      "epoch": 1.56,
-      "learning_rate": 4.031762925309657e-05,
-      "loss": 0.2864,
+      "epoch": 1.5953076008217104,
+      "grad_norm": 0.16133977472782135,
+      "learning_rate": 3.982868384681446e-05,
+      "loss": 0.4011,
       "step": 44265
     },
     {
-      "epoch": 1.56,
-      "learning_rate": 4.0315377776249244e-05,
-      "loss": 0.2824,
+      "epoch": 1.595487800482935,
+      "grad_norm": 0.16677658259868622,
+      "learning_rate": 3.982633437414677e-05,
+      "loss": 0.389,
       "step": 44270
     },
     {
-      "epoch": 1.56,
-      "learning_rate": 4.031312610054182e-05,
-      "loss": 0.2908,
+      "epoch": 1.5956680001441597,
+      "grad_norm": 0.18336910009384155,
+      "learning_rate": 3.9823984699473147e-05,
+      "loss": 0.3929,
       "step": 44275
     },
     {
-      "epoch": 1.56,
-      "learning_rate": 4.031087422600356e-05,
-      "loss": 0.2705,
+      "epoch": 1.5958481998053844,
+      "grad_norm": 0.2092888355255127,
+      "learning_rate": 3.982163482282559e-05,
+      "loss": 0.4169,
       "step": 44280
     },
     {
-      "epoch": 1.56,
-      "learning_rate": 4.030862215266368e-05,
-      "loss": 0.2942,
+      "epoch": 1.5960283994666091,
+      "grad_norm": 0.17717908322811127,
+      "learning_rate": 3.981928474423614e-05,
+      "loss": 0.4569,
       "step": 44285
     },
     {
-      "epoch": 1.56,
-      "learning_rate": 4.030636988055143e-05,
-      "loss": 0.2905,
+      "epoch": 1.5962085991278336,
+      "grad_norm": 0.15611782670021057,
+      "learning_rate": 3.98169344637368e-05,
+      "loss": 0.4195,
       "step": 44290
     },
     {
-      "epoch": 1.56,
-      "learning_rate": 4.030411740969606e-05,
-      "loss": 0.2792,
+      "epoch": 1.5963887987890581,
+      "grad_norm": 0.18297843635082245,
+      "learning_rate": 3.9814583981359596e-05,
+      "loss": 0.3639,
       "step": 44295
     },
     {
-      "epoch": 1.56,
-      "learning_rate": 4.030186474012681e-05,
-      "loss": 0.2969,
+      "epoch": 1.5965689984502829,
+      "grad_norm": 0.15773242712020874,
+      "learning_rate": 3.981223329713655e-05,
+      "loss": 0.3866,
       "step": 44300
     },
     {
-      "epoch": 1.56,
-      "learning_rate": 4.029961187187293e-05,
-      "loss": 0.2886,
+      "epoch": 1.5967491981115076,
+      "grad_norm": 0.20912377536296844,
+      "learning_rate": 3.98098824110997e-05,
+      "loss": 0.3917,
       "step": 44305
     },
     {
-      "epoch": 1.56,
-      "learning_rate": 4.029735880496367e-05,
-      "loss": 0.2571,
+      "epoch": 1.5969293977727324,
+      "grad_norm": 0.18566472828388214,
+      "learning_rate": 3.980753132328107e-05,
+      "loss": 0.4331,
       "step": 44310
     },
     {
-      "epoch": 1.56,
-      "learning_rate": 4.029510553942829e-05,
-      "loss": 0.3031,
+      "epoch": 1.5971095974339569,
+      "grad_norm": 0.17410017549991608,
+      "learning_rate": 3.9805180033712685e-05,
+      "loss": 0.4333,
       "step": 44315
     },
     {
-      "epoch": 1.56,
-      "learning_rate": 4.029285207529604e-05,
-      "loss": 0.2603,
+      "epoch": 1.5972897970951814,
+      "grad_norm": 0.17320142686367035,
+      "learning_rate": 3.980282854242659e-05,
+      "loss": 0.4056,
       "step": 44320
     },
     {
-      "epoch": 1.56,
-      "learning_rate": 4.029059841259619e-05,
-      "loss": 0.2936,
+      "epoch": 1.597469996756406,
+      "grad_norm": 0.19781510531902313,
+      "learning_rate": 3.9800476849454825e-05,
+      "loss": 0.4087,
       "step": 44325
     },
     {
-      "epoch": 1.56,
-      "learning_rate": 4.028834455135799e-05,
-      "loss": 0.2819,
+      "epoch": 1.5976501964176308,
+      "grad_norm": 0.18340526521205902,
+      "learning_rate": 3.979812495482943e-05,
+      "loss": 0.4255,
       "step": 44330
     },
     {
-      "epoch": 1.56,
-      "learning_rate": 4.028609049161072e-05,
-      "loss": 0.3126,
+      "epoch": 1.5978303960788554,
+      "grad_norm": 0.19130225479602814,
+      "learning_rate": 3.9795772858582444e-05,
+      "loss": 0.4059,
       "step": 44335
     },
     {
-      "epoch": 1.56,
-      "learning_rate": 4.028383623338363e-05,
-      "loss": 0.2795,
+      "epoch": 1.5980105957400799,
+      "grad_norm": 0.18067646026611328,
+      "learning_rate": 3.979342056074592e-05,
+      "loss": 0.4199,
       "step": 44340
     },
     {
-      "epoch": 1.56,
-      "learning_rate": 4.028158177670601e-05,
-      "loss": 0.2877,
+      "epoch": 1.5981907954013046,
+      "grad_norm": 0.20790977776050568,
+      "learning_rate": 3.97910680613519e-05,
+      "loss": 0.3784,
       "step": 44345
     },
     {
-      "epoch": 1.56,
-      "learning_rate": 4.027932712160711e-05,
-      "loss": 0.2691,
+      "epoch": 1.5983709950625293,
+      "grad_norm": 0.22750818729400635,
+      "learning_rate": 3.978871536043245e-05,
+      "loss": 0.4147,
       "step": 44350
     },
     {
-      "epoch": 1.56,
-      "learning_rate": 4.027707226811622e-05,
-      "loss": 0.2824,
+      "epoch": 1.598551194723754,
+      "grad_norm": 0.19239923357963562,
+      "learning_rate": 3.978636245801961e-05,
+      "loss": 0.4602,
       "step": 44355
     },
     {
-      "epoch": 1.56,
-      "learning_rate": 4.027481721626262e-05,
-      "loss": 0.2895,
+      "epoch": 1.5987313943849786,
+      "grad_norm": 0.13409259915351868,
+      "learning_rate": 3.9784009354145446e-05,
+      "loss": 0.4135,
       "step": 44360
     },
     {
-      "epoch": 1.56,
-      "learning_rate": 4.0272561966075586e-05,
-      "loss": 0.2697,
+      "epoch": 1.598911594046203,
+      "grad_norm": 0.18741188943386078,
+      "learning_rate": 3.978165604884201e-05,
+      "loss": 0.4095,
       "step": 44365
     },
     {
-      "epoch": 1.56,
-      "learning_rate": 4.027030651758439e-05,
-      "loss": 0.2739,
+      "epoch": 1.5990917937074278,
+      "grad_norm": 0.1525307148694992,
+      "learning_rate": 3.9779302542141384e-05,
+      "loss": 0.4054,
       "step": 44370
     },
     {
-      "epoch": 1.56,
-      "learning_rate": 4.026805087081833e-05,
-      "loss": 0.3059,
+      "epoch": 1.5992719933686526,
+      "grad_norm": 0.16644379496574402,
+      "learning_rate": 3.977694883407561e-05,
+      "loss": 0.4299,
       "step": 44375
     },
     {
-      "epoch": 1.56,
-      "learning_rate": 4.02657950258067e-05,
-      "loss": 0.2956,
+      "epoch": 1.599452193029877,
+      "grad_norm": 0.234503835439682,
+      "learning_rate": 3.977459492467678e-05,
+      "loss": 0.4382,
       "step": 44380
     },
     {
-      "epoch": 1.56,
-      "learning_rate": 4.0263538982578774e-05,
-      "loss": 0.2596,
+      "epoch": 1.5996323926911016,
+      "grad_norm": 0.1683819741010666,
+      "learning_rate": 3.977224081397696e-05,
+      "loss": 0.4252,
       "step": 44385
     },
     {
-      "epoch": 1.56,
-      "learning_rate": 4.0261282741163855e-05,
-      "loss": 0.2915,
+      "epoch": 1.5998125923523263,
+      "grad_norm": 0.15491560101509094,
+      "learning_rate": 3.97698865020082e-05,
+      "loss": 0.4087,
       "step": 44390
     },
     {
-      "epoch": 1.56,
-      "learning_rate": 4.025902630159124e-05,
-      "loss": 0.2936,
+      "epoch": 1.599992792013551,
+      "grad_norm": 0.18362046778202057,
+      "learning_rate": 3.9767531988802606e-05,
+      "loss": 0.3969,
       "step": 44395
     },
     {
-      "epoch": 1.56,
-      "learning_rate": 4.025676966389021e-05,
-      "loss": 0.2569,
+      "epoch": 1.6001729916747758,
+      "grad_norm": 0.178349107503891,
+      "learning_rate": 3.9765177274392244e-05,
+      "loss": 0.451,
       "step": 44400
     },
     {
-      "epoch": 1.56,
-      "learning_rate": 4.02545128280901e-05,
-      "loss": 0.2738,
+      "epoch": 1.6003531913360003,
+      "grad_norm": 0.1806674599647522,
+      "learning_rate": 3.9762822358809206e-05,
+      "loss": 0.4001,
       "step": 44405
     },
     {
-      "epoch": 1.56,
-      "learning_rate": 4.025225579422018e-05,
-      "loss": 0.3099,
+      "epoch": 1.6005333909972248,
+      "grad_norm": 0.16210544109344482,
+      "learning_rate": 3.976046724208557e-05,
+      "loss": 0.3909,
       "step": 44410
     },
     {
-      "epoch": 1.56,
-      "learning_rate": 4.025045002453369e-05,
-      "loss": 0.2883,
+      "epoch": 1.6007135906584495,
+      "grad_norm": 0.18187157809734344,
+      "learning_rate": 3.975811192425342e-05,
+      "loss": 0.4267,
       "step": 44415
     },
     {
-      "epoch": 1.56,
-      "learning_rate": 4.0248192634212e-05,
-      "loss": 0.291,
+      "epoch": 1.6008937903196743,
+      "grad_norm": 0.20049238204956055,
+      "learning_rate": 3.9755756405344855e-05,
+      "loss": 0.3962,
       "step": 44420
     },
     {
-      "epoch": 1.56,
-      "learning_rate": 4.0245935045902575e-05,
-      "loss": 0.2972,
+      "epoch": 1.6010739899808988,
+      "grad_norm": 0.1605527549982071,
+      "learning_rate": 3.9753400685391974e-05,
+      "loss": 0.4174,
       "step": 44425
     },
     {
-      "epoch": 1.56,
-      "learning_rate": 4.0243677259634734e-05,
-      "loss": 0.291,
+      "epoch": 1.6012541896421235,
+      "grad_norm": 0.16398051381111145,
+      "learning_rate": 3.975104476442686e-05,
+      "loss": 0.4539,
       "step": 44430
     },
     {
-      "epoch": 1.56,
-      "learning_rate": 4.0241419275437783e-05,
-      "loss": 0.2713,
+      "epoch": 1.601434389303348,
+      "grad_norm": 0.18974505364894867,
+      "learning_rate": 3.9748688642481614e-05,
+      "loss": 0.3968,
       "step": 44435
     },
     {
-      "epoch": 1.56,
-      "learning_rate": 4.023916109334105e-05,
-      "loss": 0.2927,
+      "epoch": 1.6016145889645728,
+      "grad_norm": 0.22180098295211792,
+      "learning_rate": 3.974633231958834e-05,
+      "loss": 0.4166,
       "step": 44440
     },
     {
-      "epoch": 1.56,
-      "learning_rate": 4.023690271337385e-05,
-      "loss": 0.3123,
+      "epoch": 1.6017947886257975,
+      "grad_norm": 0.15120428800582886,
+      "learning_rate": 3.974397579577914e-05,
+      "loss": 0.4171,
       "step": 44445
     },
     {
-      "epoch": 1.56,
-      "learning_rate": 4.0234644135565526e-05,
-      "loss": 0.2886,
+      "epoch": 1.601974988287022,
+      "grad_norm": 0.19408200681209564,
+      "learning_rate": 3.974161907108613e-05,
+      "loss": 0.4058,
       "step": 44450
     },
     {
-      "epoch": 1.56,
-      "learning_rate": 4.023238535994537e-05,
-      "loss": 0.2892,
+      "epoch": 1.6021551879482465,
+      "grad_norm": 0.1284191906452179,
+      "learning_rate": 3.973926214554142e-05,
+      "loss": 0.3842,
       "step": 44455
     },
     {
-      "epoch": 1.56,
-      "learning_rate": 4.023012638654273e-05,
-      "loss": 0.2858,
+      "epoch": 1.6023353876094713,
+      "grad_norm": 0.12077841907739639,
+      "learning_rate": 3.9736905019177106e-05,
+      "loss": 0.363,
       "step": 44460
     },
     {
-      "epoch": 1.56,
-      "learning_rate": 4.022786721538694e-05,
-      "loss": 0.2648,
+      "epoch": 1.602515587270696,
+      "grad_norm": 0.19127848744392395,
+      "learning_rate": 3.973454769202532e-05,
+      "loss": 0.3965,
       "step": 44465
     },
     {
-      "epoch": 1.56,
-      "learning_rate": 4.022560784650733e-05,
-      "loss": 0.2844,
+      "epoch": 1.6026957869319207,
+      "grad_norm": 0.16007506847381592,
+      "learning_rate": 3.9732190164118175e-05,
+      "loss": 0.3915,
       "step": 44470
     },
     {
-      "epoch": 1.56,
-      "learning_rate": 4.022334827993324e-05,
-      "loss": 0.2985,
+      "epoch": 1.6028759865931452,
+      "grad_norm": 0.16943296790122986,
+      "learning_rate": 3.972983243548779e-05,
+      "loss": 0.4289,
       "step": 44475
     },
     {
-      "epoch": 1.56,
-      "learning_rate": 4.0221088515693995e-05,
-      "loss": 0.2637,
+      "epoch": 1.6030561862543697,
+      "grad_norm": 0.15759488940238953,
+      "learning_rate": 3.972747450616629e-05,
+      "loss": 0.405,
       "step": 44480
     },
     {
-      "epoch": 1.57,
-      "learning_rate": 4.021882855381895e-05,
-      "loss": 0.2832,
+      "epoch": 1.6032363859155945,
+      "grad_norm": 0.178999662399292,
+      "learning_rate": 3.97251163761858e-05,
+      "loss": 0.3919,
       "step": 44485
     },
     {
-      "epoch": 1.57,
-      "learning_rate": 4.021656839433745e-05,
-      "loss": 0.3021,
+      "epoch": 1.6034165855768192,
+      "grad_norm": 0.26717543601989746,
+      "learning_rate": 3.9722758045578454e-05,
+      "loss": 0.4161,
       "step": 44490
     },
     {
-      "epoch": 1.57,
-      "learning_rate": 4.0214308037278836e-05,
-      "loss": 0.2737,
+      "epoch": 1.6035967852380437,
+      "grad_norm": 0.24966903030872345,
+      "learning_rate": 3.9720399514376374e-05,
+      "loss": 0.4346,
       "step": 44495
     },
     {
-      "epoch": 1.57,
-      "learning_rate": 4.021204748267246e-05,
-      "loss": 0.2878,
+      "epoch": 1.6037769848992682,
+      "grad_norm": 0.17381806671619415,
+      "learning_rate": 3.97180407826117e-05,
+      "loss": 0.4227,
       "step": 44500
     },
     {
-      "epoch": 1.57,
-      "eval_loss": 0.2789715826511383,
-      "eval_runtime": 10.538,
-      "eval_samples_per_second": 9.489,
-      "eval_steps_per_second": 9.489,
+      "epoch": 1.6037769848992682,
+      "eval_loss": 0.44284766912460327,
+      "eval_runtime": 3.5457,
+      "eval_samples_per_second": 28.203,
+      "eval_steps_per_second": 7.051,
       "step": 44500
     },
     {
-      "epoch": 1.57,
-      "learning_rate": 4.020978673054767e-05,
-      "loss": 0.2947,
+      "epoch": 1.603957184560493,
+      "grad_norm": 0.15787242352962494,
+      "learning_rate": 3.971568185031658e-05,
+      "loss": 0.3882,
       "step": 44505
     },
     {
-      "epoch": 1.57,
-      "learning_rate": 4.020752578093382e-05,
-      "loss": 0.2945,
+      "epoch": 1.6041373842217177,
+      "grad_norm": 0.168611541390419,
+      "learning_rate": 3.971332271752313e-05,
+      "loss": 0.3956,
       "step": 44510
     },
     {
-      "epoch": 1.57,
-      "learning_rate": 4.020526463386028e-05,
-      "loss": 0.2918,
+      "epoch": 1.6043175838829424,
+      "grad_norm": 0.17954233288764954,
+      "learning_rate": 3.9710963384263515e-05,
+      "loss": 0.4084,
       "step": 44515
     },
     {
-      "epoch": 1.57,
-      "learning_rate": 4.020300328935639e-05,
-      "loss": 0.3068,
+      "epoch": 1.604497783544167,
+      "grad_norm": 0.16351082921028137,
+      "learning_rate": 3.970860385056987e-05,
+      "loss": 0.4356,
       "step": 44520
     },
     {
-      "epoch": 1.57,
-      "learning_rate": 4.020074174745153e-05,
-      "loss": 0.2898,
+      "epoch": 1.6046779832053915,
+      "grad_norm": 0.19120116531848907,
+      "learning_rate": 3.9706244116474345e-05,
+      "loss": 0.418,
       "step": 44525
     },
     {
-      "epoch": 1.57,
-      "learning_rate": 4.0198480008175044e-05,
-      "loss": 0.2909,
+      "epoch": 1.6048581828666162,
+      "grad_norm": 0.15512365102767944,
+      "learning_rate": 3.97038841820091e-05,
+      "loss": 0.4185,
       "step": 44530
     },
     {
-      "epoch": 1.57,
-      "learning_rate": 4.019621807155632e-05,
-      "loss": 0.2707,
+      "epoch": 1.605038382527841,
+      "grad_norm": 0.1973215490579605,
+      "learning_rate": 3.970152404720627e-05,
+      "loss": 0.4046,
       "step": 44535
     },
     {
-      "epoch": 1.57,
-      "learning_rate": 4.019395593762472e-05,
-      "loss": 0.2816,
+      "epoch": 1.6052185821890654,
+      "grad_norm": 0.18126599490642548,
+      "learning_rate": 3.969916371209802e-05,
+      "loss": 0.4358,
       "step": 44540
     },
     {
-      "epoch": 1.57,
-      "learning_rate": 4.019169360640961e-05,
-      "loss": 0.2972,
+      "epoch": 1.6053987818502902,
+      "grad_norm": 0.21188373863697052,
+      "learning_rate": 3.9696803176716515e-05,
+      "loss": 0.472,
       "step": 44545
     },
     {
-      "epoch": 1.57,
-      "learning_rate": 4.018943107794038e-05,
-      "loss": 0.2918,
+      "epoch": 1.6055789815115147,
+      "grad_norm": 0.16535495221614838,
+      "learning_rate": 3.969444244109391e-05,
+      "loss": 0.4183,
       "step": 44550
     },
     {
-      "epoch": 1.57,
-      "learning_rate": 4.018716835224638e-05,
-      "loss": 0.2743,
+      "epoch": 1.6057591811727394,
+      "grad_norm": 0.21638596057891846,
+      "learning_rate": 3.969208150526237e-05,
+      "loss": 0.4231,
       "step": 44555
     },
     {
-      "epoch": 1.57,
-      "learning_rate": 4.0184905429357026e-05,
-      "loss": 0.2786,
+      "epoch": 1.6059393808339641,
+      "grad_norm": 0.21015167236328125,
+      "learning_rate": 3.968972036925407e-05,
+      "loss": 0.4411,
       "step": 44560
     },
     {
-      "epoch": 1.57,
-      "learning_rate": 4.018264230930167e-05,
-      "loss": 0.2878,
+      "epoch": 1.6061195804951887,
+      "grad_norm": 0.16333642601966858,
+      "learning_rate": 3.968735903310117e-05,
+      "loss": 0.4082,
       "step": 44565
     },
     {
-      "epoch": 1.57,
-      "learning_rate": 4.0180378992109714e-05,
-      "loss": 0.2979,
+      "epoch": 1.6062997801564132,
+      "grad_norm": 0.1700402796268463,
+      "learning_rate": 3.968499749683584e-05,
+      "loss": 0.406,
       "step": 44570
     },
     {
-      "epoch": 1.57,
-      "learning_rate": 4.017811547781054e-05,
-      "loss": 0.2852,
+      "epoch": 1.606479979817638,
+      "grad_norm": 0.16421562433242798,
+      "learning_rate": 3.968263576049027e-05,
+      "loss": 0.3949,
       "step": 44575
     },
     {
-      "epoch": 1.57,
-      "learning_rate": 4.0175851766433536e-05,
-      "loss": 0.2772,
+      "epoch": 1.6066601794788626,
+      "grad_norm": 0.18624955415725708,
+      "learning_rate": 3.968027382409663e-05,
+      "loss": 0.3847,
       "step": 44580
     },
     {
-      "epoch": 1.57,
-      "learning_rate": 4.01735878580081e-05,
-      "loss": 0.287,
+      "epoch": 1.6068403791400874,
+      "grad_norm": 0.23237396776676178,
+      "learning_rate": 3.9677911687687095e-05,
+      "loss": 0.4233,
       "step": 44585
     },
     {
-      "epoch": 1.57,
-      "learning_rate": 4.0171323752563636e-05,
-      "loss": 0.2819,
+      "epoch": 1.6070205788013119,
+      "grad_norm": 0.17745272815227509,
+      "learning_rate": 3.967554935129387e-05,
+      "loss": 0.405,
       "step": 44590
     },
     {
-      "epoch": 1.57,
-      "learning_rate": 4.016905945012952e-05,
-      "loss": 0.2993,
+      "epoch": 1.6072007784625364,
+      "grad_norm": 0.1729012280702591,
+      "learning_rate": 3.9673186814949115e-05,
+      "loss": 0.3969,
       "step": 44595
     },
     {
-      "epoch": 1.57,
-      "learning_rate": 4.0166794950735164e-05,
-      "loss": 0.2784,
+      "epoch": 1.6073809781237611,
+      "grad_norm": 0.2897031605243683,
+      "learning_rate": 3.967082407868503e-05,
+      "loss": 0.3892,
       "step": 44600
     },
     {
-      "epoch": 1.57,
-      "learning_rate": 4.016453025440997e-05,
-      "loss": 0.2872,
+      "epoch": 1.6075611777849859,
+      "grad_norm": 0.17418605089187622,
+      "learning_rate": 3.9668461142533807e-05,
+      "loss": 0.4135,
       "step": 44605
     },
     {
-      "epoch": 1.57,
-      "learning_rate": 4.016226536118335e-05,
-      "loss": 0.2724,
+      "epoch": 1.6077413774462104,
+      "grad_norm": 0.19740115106105804,
+      "learning_rate": 3.9666098006527653e-05,
+      "loss": 0.3922,
       "step": 44610
     },
     {
-      "epoch": 1.57,
-      "learning_rate": 4.016000027108471e-05,
-      "loss": 0.2988,
+      "epoch": 1.6079215771074349,
+      "grad_norm": 0.18826255202293396,
+      "learning_rate": 3.966373467069874e-05,
+      "loss": 0.4254,
       "step": 44615
     },
     {
-      "epoch": 1.57,
-      "learning_rate": 4.015773498414345e-05,
-      "loss": 0.2687,
+      "epoch": 1.6081017767686596,
+      "grad_norm": 0.18724419176578522,
+      "learning_rate": 3.9661371135079285e-05,
+      "loss": 0.4175,
       "step": 44620
     },
     {
-      "epoch": 1.57,
-      "learning_rate": 4.015546950038899e-05,
-      "loss": 0.2678,
+      "epoch": 1.6082819764298844,
+      "grad_norm": 0.20004336535930634,
+      "learning_rate": 3.9659007399701485e-05,
+      "loss": 0.4015,
       "step": 44625
     },
     {
-      "epoch": 1.57,
-      "learning_rate": 4.015320381985075e-05,
-      "loss": 0.2811,
+      "epoch": 1.608462176091109,
+      "grad_norm": 0.2576258182525635,
+      "learning_rate": 3.9656643464597545e-05,
+      "loss": 0.4176,
       "step": 44630
     },
     {
-      "epoch": 1.57,
-      "learning_rate": 4.015093794255814e-05,
-      "loss": 0.2692,
+      "epoch": 1.6086423757523336,
+      "grad_norm": 0.1877816915512085,
+      "learning_rate": 3.9654279329799684e-05,
+      "loss": 0.4314,
       "step": 44635
     },
     {
-      "epoch": 1.57,
-      "learning_rate": 4.0148671868540596e-05,
-      "loss": 0.2624,
+      "epoch": 1.608822575413558,
+      "grad_norm": 0.20604005455970764,
+      "learning_rate": 3.96519149953401e-05,
+      "loss": 0.4033,
       "step": 44640
     },
     {
-      "epoch": 1.57,
-      "learning_rate": 4.014640559782752e-05,
-      "loss": 0.3,
+      "epoch": 1.6090027750747828,
+      "grad_norm": 0.17732234299182892,
+      "learning_rate": 3.964955046125101e-05,
+      "loss": 0.4149,
       "step": 44645
     },
     {
-      "epoch": 1.57,
-      "learning_rate": 4.014413913044835e-05,
-      "loss": 0.2471,
+      "epoch": 1.6091829747360076,
+      "grad_norm": 0.22829070687294006,
+      "learning_rate": 3.964718572756463e-05,
+      "loss": 0.3822,
       "step": 44650
     },
     {
-      "epoch": 1.57,
-      "learning_rate": 4.0141872466432515e-05,
-      "loss": 0.2739,
+      "epoch": 1.609363174397232,
+      "grad_norm": 0.20384085178375244,
+      "learning_rate": 3.964482079431319e-05,
+      "loss": 0.411,
       "step": 44655
     },
     {
-      "epoch": 1.57,
-      "learning_rate": 4.013960560580945e-05,
-      "loss": 0.2639,
+      "epoch": 1.6095433740584568,
+      "grad_norm": 0.21670548617839813,
+      "learning_rate": 3.9642455661528885e-05,
+      "loss": 0.4065,
       "step": 44660
     },
     {
-      "epoch": 1.57,
-      "learning_rate": 4.013733854860858e-05,
-      "loss": 0.2871,
+      "epoch": 1.6097235737196813,
+      "grad_norm": 0.18137463927268982,
+      "learning_rate": 3.964009032924396e-05,
+      "loss": 0.3918,
       "step": 44665
     },
     {
-      "epoch": 1.57,
-      "learning_rate": 4.013507129485935e-05,
-      "loss": 0.2904,
+      "epoch": 1.609903773380906,
+      "grad_norm": 0.16431452333927155,
+      "learning_rate": 3.963772479749065e-05,
+      "loss": 0.3946,
       "step": 44670
     },
     {
-      "epoch": 1.57,
-      "learning_rate": 4.0132803844591186e-05,
-      "loss": 0.2661,
+      "epoch": 1.6100839730421308,
+      "grad_norm": 0.17995165288448334,
+      "learning_rate": 3.963535906630117e-05,
+      "loss": 0.4308,
       "step": 44675
     },
     {
-      "epoch": 1.57,
-      "learning_rate": 4.013053619783355e-05,
-      "loss": 0.2731,
+      "epoch": 1.6102641727033553,
+      "grad_norm": 0.17555540800094604,
+      "learning_rate": 3.9632993135707755e-05,
+      "loss": 0.4113,
       "step": 44680
     },
     {
-      "epoch": 1.57,
-      "learning_rate": 4.012826835461585e-05,
-      "loss": 0.2718,
+      "epoch": 1.6104443723645798,
+      "grad_norm": 0.21081358194351196,
+      "learning_rate": 3.963062700574264e-05,
+      "loss": 0.4261,
       "step": 44685
     },
     {
-      "epoch": 1.57,
-      "learning_rate": 4.012600031496757e-05,
-      "loss": 0.2877,
+      "epoch": 1.6106245720258046,
+      "grad_norm": 0.18575575947761536,
+      "learning_rate": 3.9628260676438064e-05,
+      "loss": 0.3961,
       "step": 44690
     },
     {
-      "epoch": 1.57,
-      "learning_rate": 4.012373207891814e-05,
-      "loss": 0.2779,
+      "epoch": 1.6108047716870293,
+      "grad_norm": 0.16622988879680634,
+      "learning_rate": 3.962589414782627e-05,
+      "loss": 0.4321,
       "step": 44695
     },
     {
-      "epoch": 1.57,
-      "learning_rate": 4.012146364649702e-05,
-      "loss": 0.288,
+      "epoch": 1.610984971348254,
+      "grad_norm": 0.19155137240886688,
+      "learning_rate": 3.96235274199395e-05,
+      "loss": 0.4032,
       "step": 44700
     },
     {
-      "epoch": 1.57,
-      "learning_rate": 4.011919501773366e-05,
-      "loss": 0.2802,
+      "epoch": 1.6111651710094785,
+      "grad_norm": 0.15774855017662048,
+      "learning_rate": 3.962116049280999e-05,
+      "loss": 0.4115,
       "step": 44705
     },
     {
-      "epoch": 1.57,
-      "learning_rate": 4.011692619265751e-05,
-      "loss": 0.2772,
+      "epoch": 1.611345370670703,
+      "grad_norm": 0.17993386089801788,
+      "learning_rate": 3.961879336647001e-05,
+      "loss": 0.4065,
       "step": 44710
     },
     {
-      "epoch": 1.57,
-      "learning_rate": 4.011465717129805e-05,
-      "loss": 0.2637,
+      "epoch": 1.6115255703319278,
+      "grad_norm": 0.18789418041706085,
+      "learning_rate": 3.961642604095181e-05,
+      "loss": 0.391,
       "step": 44715
     },
     {
-      "epoch": 1.57,
-      "learning_rate": 4.011238795368471e-05,
-      "loss": 0.2746,
+      "epoch": 1.6117057699931525,
+      "grad_norm": 0.1855124831199646,
+      "learning_rate": 3.9614058516287624e-05,
+      "loss": 0.4414,
       "step": 44720
     },
     {
-      "epoch": 1.57,
-      "learning_rate": 4.0110118539846986e-05,
-      "loss": 0.29,
+      "epoch": 1.611885969654377,
+      "grad_norm": 0.17999990284442902,
+      "learning_rate": 3.961169079250971e-05,
+      "loss": 0.385,
       "step": 44725
     },
     {
-      "epoch": 1.57,
-      "learning_rate": 4.0107848929814325e-05,
-      "loss": 0.3046,
+      "epoch": 1.6120661693156015,
+      "grad_norm": 0.1700844168663025,
+      "learning_rate": 3.9609322869650354e-05,
+      "loss": 0.4109,
       "step": 44730
     },
     {
-      "epoch": 1.57,
-      "learning_rate": 4.01055791236162e-05,
-      "loss": 0.3014,
+      "epoch": 1.6122463689768263,
+      "grad_norm": 0.1721271425485611,
+      "learning_rate": 3.9606954747741797e-05,
+      "loss": 0.4331,
       "step": 44735
     },
     {
-      "epoch": 1.57,
-      "learning_rate": 4.0103309121282085e-05,
-      "loss": 0.2705,
+      "epoch": 1.612426568638051,
+      "grad_norm": 0.17084349691867828,
+      "learning_rate": 3.960458642681631e-05,
+      "loss": 0.3995,
       "step": 44740
     },
     {
-      "epoch": 1.57,
-      "learning_rate": 4.010103892284146e-05,
-      "loss": 0.3123,
+      "epoch": 1.6126067682992757,
+      "grad_norm": 0.14765715599060059,
+      "learning_rate": 3.960221790690616e-05,
+      "loss": 0.4097,
       "step": 44745
     },
     {
-      "epoch": 1.57,
-      "learning_rate": 4.009876852832379e-05,
-      "loss": 0.2675,
+      "epoch": 1.6127869679605003,
+      "grad_norm": 0.16474126279354095,
+      "learning_rate": 3.959984918804361e-05,
+      "loss": 0.4378,
       "step": 44750
     },
     {
-      "epoch": 1.57,
-      "learning_rate": 4.009649793775857e-05,
-      "loss": 0.2937,
+      "epoch": 1.6129671676217248,
+      "grad_norm": 0.2174513041973114,
+      "learning_rate": 3.959748027026095e-05,
+      "loss": 0.3976,
       "step": 44755
     },
     {
-      "epoch": 1.57,
-      "learning_rate": 4.009422715117526e-05,
-      "loss": 0.2638,
+      "epoch": 1.6131473672829495,
+      "grad_norm": 0.15356197953224182,
+      "learning_rate": 3.959511115359045e-05,
+      "loss": 0.4151,
       "step": 44760
     },
     {
-      "epoch": 1.57,
-      "learning_rate": 4.009195616860337e-05,
-      "loss": 0.3036,
+      "epoch": 1.6133275669441742,
+      "grad_norm": 0.18707029521465302,
+      "learning_rate": 3.959274183806438e-05,
+      "loss": 0.4231,
       "step": 44765
     },
     {
-      "epoch": 1.58,
-      "learning_rate": 4.008968499007237e-05,
-      "loss": 0.2876,
+      "epoch": 1.6135077666053987,
+      "grad_norm": 0.20073364675045013,
+      "learning_rate": 3.959037232371503e-05,
+      "loss": 0.4256,
       "step": 44770
     },
     {
-      "epoch": 1.58,
-      "learning_rate": 4.0087413615611756e-05,
-      "loss": 0.2855,
+      "epoch": 1.6136879662666233,
+      "grad_norm": 0.20367778837680817,
+      "learning_rate": 3.9588002610574694e-05,
+      "loss": 0.4231,
       "step": 44775
     },
     {
-      "epoch": 1.58,
-      "learning_rate": 4.008514204525101e-05,
-      "loss": 0.3001,
+      "epoch": 1.613868165927848,
+      "grad_norm": 0.15489740669727325,
+      "learning_rate": 3.958563269867563e-05,
+      "loss": 0.3753,
       "step": 44780
     },
     {
-      "epoch": 1.58,
-      "learning_rate": 4.008287027901965e-05,
-      "loss": 0.3207,
+      "epoch": 1.6140483655890727,
+      "grad_norm": 0.18717733025550842,
+      "learning_rate": 3.958326258805015e-05,
+      "loss": 0.4145,
       "step": 44785
     },
     {
-      "epoch": 1.58,
-      "learning_rate": 4.008059831694716e-05,
-      "loss": 0.2864,
+      "epoch": 1.6142285652502975,
+      "grad_norm": 0.1864037811756134,
+      "learning_rate": 3.9580892278730534e-05,
+      "loss": 0.3988,
       "step": 44790
     },
     {
-      "epoch": 1.58,
-      "learning_rate": 4.0078326159063026e-05,
-      "loss": 0.2738,
+      "epoch": 1.614408764911522,
+      "grad_norm": 0.19516822695732117,
+      "learning_rate": 3.95785217707491e-05,
+      "loss": 0.4386,
       "step": 44795
     },
     {
-      "epoch": 1.58,
-      "learning_rate": 4.007605380539677e-05,
-      "loss": 0.2664,
+      "epoch": 1.6145889645727465,
+      "grad_norm": 0.17213517427444458,
+      "learning_rate": 3.957615106413811e-05,
+      "loss": 0.4205,
       "step": 44800
     },
     {
-      "epoch": 1.58,
-      "learning_rate": 4.0073781255977894e-05,
-      "loss": 0.2699,
+      "epoch": 1.6147691642339712,
+      "grad_norm": 0.16478149592876434,
+      "learning_rate": 3.95737801589299e-05,
+      "loss": 0.3819,
       "step": 44805
     },
     {
-      "epoch": 1.58,
-      "learning_rate": 4.00715085108359e-05,
-      "loss": 0.2927,
+      "epoch": 1.614949363895196,
+      "grad_norm": 0.15579113364219666,
+      "learning_rate": 3.957140905515674e-05,
+      "loss": 0.4181,
       "step": 44810
     },
     {
-      "epoch": 1.58,
-      "learning_rate": 4.00692355700003e-05,
-      "loss": 0.2615,
+      "epoch": 1.6151295635564207,
+      "grad_norm": 0.21928997337818146,
+      "learning_rate": 3.956903775285097e-05,
+      "loss": 0.4157,
       "step": 44815
     },
     {
-      "epoch": 1.58,
-      "learning_rate": 4.006696243350061e-05,
-      "loss": 0.2782,
+      "epoch": 1.6153097632176452,
+      "grad_norm": 0.14107581973075867,
+      "learning_rate": 3.956666625204487e-05,
+      "loss": 0.3837,
       "step": 44820
     },
     {
-      "epoch": 1.58,
-      "learning_rate": 4.0064689101366336e-05,
-      "loss": 0.2823,
+      "epoch": 1.6154899628788697,
+      "grad_norm": 0.19918832182884216,
+      "learning_rate": 3.956429455277077e-05,
+      "loss": 0.4018,
       "step": 44825
     },
     {
-      "epoch": 1.58,
-      "learning_rate": 4.006241557362701e-05,
-      "loss": 0.2789,
+      "epoch": 1.6156701625400944,
+      "grad_norm": 0.19882018864154816,
+      "learning_rate": 3.9561922655060965e-05,
+      "loss": 0.4323,
       "step": 44830
     },
     {
-      "epoch": 1.58,
-      "learning_rate": 4.0060141850312135e-05,
-      "loss": 0.2832,
+      "epoch": 1.6158503622013192,
+      "grad_norm": 0.17899902164936066,
+      "learning_rate": 3.95595505589478e-05,
+      "loss": 0.4351,
       "step": 44835
     },
     {
-      "epoch": 1.58,
-      "learning_rate": 4.005786793145126e-05,
-      "loss": 0.2645,
+      "epoch": 1.6160305618625437,
+      "grad_norm": 0.14967146515846252,
+      "learning_rate": 3.955717826446357e-05,
+      "loss": 0.4039,
       "step": 44840
     },
     {
-      "epoch": 1.58,
-      "learning_rate": 4.005559381707388e-05,
-      "loss": 0.275,
+      "epoch": 1.6162107615237682,
+      "grad_norm": 0.17827114462852478,
+      "learning_rate": 3.95548057716406e-05,
+      "loss": 0.3759,
       "step": 44845
     },
     {
-      "epoch": 1.58,
-      "learning_rate": 4.005331950720953e-05,
-      "loss": 0.283,
+      "epoch": 1.616390961184993,
+      "grad_norm": 0.19635485112667084,
+      "learning_rate": 3.955243308051122e-05,
+      "loss": 0.4571,
       "step": 44850
     },
     {
-      "epoch": 1.58,
-      "learning_rate": 4.0051045001887755e-05,
-      "loss": 0.283,
+      "epoch": 1.6165711608462177,
+      "grad_norm": 0.20336943864822388,
+      "learning_rate": 3.955006019110776e-05,
+      "loss": 0.4104,
       "step": 44855
     },
     {
-      "epoch": 1.58,
-      "learning_rate": 4.004877030113808e-05,
-      "loss": 0.3178,
+      "epoch": 1.6167513605074424,
+      "grad_norm": 0.16588455438613892,
+      "learning_rate": 3.954768710346255e-05,
+      "loss": 0.3808,
       "step": 44860
     },
     {
-      "epoch": 1.58,
-      "learning_rate": 4.0046495404990043e-05,
-      "loss": 0.2851,
+      "epoch": 1.616931560168667,
+      "grad_norm": 0.18041035532951355,
+      "learning_rate": 3.954531381760791e-05,
+      "loss": 0.3845,
       "step": 44865
     },
     {
-      "epoch": 1.58,
-      "learning_rate": 4.004422031347318e-05,
-      "loss": 0.2787,
+      "epoch": 1.6171117598298914,
+      "grad_norm": 0.19246827065944672,
+      "learning_rate": 3.95429403335762e-05,
+      "loss": 0.4226,
       "step": 44870
     },
     {
-      "epoch": 1.58,
-      "learning_rate": 4.0041945026617024e-05,
-      "loss": 0.274,
+      "epoch": 1.6172919594911161,
+      "grad_norm": 0.1802048683166504,
+      "learning_rate": 3.954056665139972e-05,
+      "loss": 0.4507,
       "step": 44875
     },
     {
-      "epoch": 1.58,
-      "learning_rate": 4.0039669544451126e-05,
-      "loss": 0.278,
+      "epoch": 1.6174721591523409,
+      "grad_norm": 0.1771692931652069,
+      "learning_rate": 3.9538192771110855e-05,
+      "loss": 0.3853,
       "step": 44880
     },
     {
-      "epoch": 1.58,
-      "learning_rate": 4.003739386700503e-05,
-      "loss": 0.2996,
+      "epoch": 1.6176523588135654,
+      "grad_norm": 0.19818015396595,
+      "learning_rate": 3.953581869274192e-05,
+      "loss": 0.4143,
       "step": 44885
     },
     {
-      "epoch": 1.58,
-      "learning_rate": 4.003511799430829e-05,
-      "loss": 0.3108,
+      "epoch": 1.61783255847479,
+      "grad_norm": 0.19741560518741608,
+      "learning_rate": 3.953344441632527e-05,
+      "loss": 0.4189,
       "step": 44890
     },
     {
-      "epoch": 1.58,
-      "learning_rate": 4.003284192639045e-05,
-      "loss": 0.28,
+      "epoch": 1.6180127581360146,
+      "grad_norm": 0.15774419903755188,
+      "learning_rate": 3.953106994189326e-05,
+      "loss": 0.4124,
       "step": 44895
     },
     {
-      "epoch": 1.58,
-      "learning_rate": 4.003056566328106e-05,
-      "loss": 0.2945,
+      "epoch": 1.6181929577972394,
+      "grad_norm": 0.1899198293685913,
+      "learning_rate": 3.952869526947823e-05,
+      "loss": 0.4436,
       "step": 44900
     },
     {
-      "epoch": 1.58,
-      "learning_rate": 4.002828920500968e-05,
-      "loss": 0.3199,
+      "epoch": 1.618373157458464,
+      "grad_norm": 0.19726921617984772,
+      "learning_rate": 3.9526320399112536e-05,
+      "loss": 0.3846,
       "step": 44905
     },
     {
-      "epoch": 1.58,
-      "learning_rate": 4.002601255160587e-05,
-      "loss": 0.2892,
+      "epoch": 1.6185533571196886,
+      "grad_norm": 0.18593871593475342,
+      "learning_rate": 3.952394533082855e-05,
+      "loss": 0.4035,
       "step": 44910
     },
     {
-      "epoch": 1.58,
-      "learning_rate": 4.00237357030992e-05,
-      "loss": 0.3298,
+      "epoch": 1.6187335567809131,
+      "grad_norm": 0.1598072648048401,
+      "learning_rate": 3.952157006465861e-05,
+      "loss": 0.4077,
       "step": 44915
     },
     {
-      "epoch": 1.58,
-      "learning_rate": 4.002145865951921e-05,
-      "loss": 0.2979,
+      "epoch": 1.6189137564421379,
+      "grad_norm": 0.2124381959438324,
+      "learning_rate": 3.95191946006351e-05,
+      "loss": 0.4266,
       "step": 44920
     },
     {
-      "epoch": 1.58,
-      "learning_rate": 4.001918142089548e-05,
-      "loss": 0.2916,
+      "epoch": 1.6190939561033626,
+      "grad_norm": 0.16838979721069336,
+      "learning_rate": 3.951681893879036e-05,
+      "loss": 0.4073,
       "step": 44925
     },
     {
-      "epoch": 1.58,
-      "learning_rate": 4.001690398725758e-05,
-      "loss": 0.2871,
+      "epoch": 1.619274155764587,
+      "grad_norm": 0.16557417809963226,
+      "learning_rate": 3.951444307915678e-05,
+      "loss": 0.4183,
       "step": 44930
     },
     {
-      "epoch": 1.58,
-      "learning_rate": 4.0014626358635084e-05,
-      "loss": 0.2833,
+      "epoch": 1.6194543554258118,
+      "grad_norm": 0.17037592828273773,
+      "learning_rate": 3.951206702176672e-05,
+      "loss": 0.3463,
       "step": 44935
     },
     {
-      "epoch": 1.58,
-      "learning_rate": 4.001234853505755e-05,
-      "loss": 0.2868,
+      "epoch": 1.6196345550870364,
+      "grad_norm": 0.16905367374420166,
+      "learning_rate": 3.950969076665256e-05,
+      "loss": 0.3978,
       "step": 44940
     },
     {
-      "epoch": 1.58,
-      "learning_rate": 4.001007051655457e-05,
-      "loss": 0.2803,
+      "epoch": 1.619814754748261,
+      "grad_norm": 0.16748353838920593,
+      "learning_rate": 3.950731431384668e-05,
+      "loss": 0.461,
       "step": 44945
     },
     {
-      "epoch": 1.58,
-      "learning_rate": 4.000779230315571e-05,
-      "loss": 0.2863,
+      "epoch": 1.6199949544094858,
+      "grad_norm": 0.1732574999332428,
+      "learning_rate": 3.950493766338144e-05,
+      "loss": 0.407,
       "step": 44950
     },
     {
-      "epoch": 1.58,
-      "learning_rate": 4.000551389489057e-05,
-      "loss": 0.2994,
+      "epoch": 1.6201751540707103,
+      "grad_norm": 0.12833158671855927,
+      "learning_rate": 3.9502560815289236e-05,
+      "loss": 0.3877,
       "step": 44955
     },
     {
-      "epoch": 1.58,
-      "learning_rate": 4.000323529178871e-05,
-      "loss": 0.3044,
+      "epoch": 1.6203553537319348,
+      "grad_norm": 0.17978282272815704,
+      "learning_rate": 3.950018376960245e-05,
+      "loss": 0.4099,
       "step": 44960
     },
     {
-      "epoch": 1.58,
-      "learning_rate": 4.000095649387973e-05,
-      "loss": 0.2818,
+      "epoch": 1.6205355533931596,
+      "grad_norm": 0.19126909971237183,
+      "learning_rate": 3.949780652635347e-05,
+      "loss": 0.4621,
       "step": 44965
     },
     {
-      "epoch": 1.58,
-      "learning_rate": 3.9998677501193214e-05,
-      "loss": 0.2852,
+      "epoch": 1.6207157530543843,
+      "grad_norm": 0.1761917769908905,
+      "learning_rate": 3.9495429085574675e-05,
+      "loss": 0.4437,
       "step": 44970
     },
     {
-      "epoch": 1.58,
-      "learning_rate": 3.999639831375876e-05,
-      "loss": 0.2878,
+      "epoch": 1.620895952715609,
+      "grad_norm": 0.21770428121089935,
+      "learning_rate": 3.949305144729847e-05,
+      "loss": 0.4273,
       "step": 44975
     },
     {
-      "epoch": 1.58,
-      "learning_rate": 3.999411893160595e-05,
-      "loss": 0.283,
+      "epoch": 1.6210761523768336,
+      "grad_norm": 0.21799218654632568,
+      "learning_rate": 3.949067361155725e-05,
+      "loss": 0.4328,
       "step": 44980
     },
     {
-      "epoch": 1.58,
-      "learning_rate": 3.9991839354764394e-05,
-      "loss": 0.2982,
+      "epoch": 1.621256352038058,
+      "grad_norm": 0.21713624894618988,
+      "learning_rate": 3.94882955783834e-05,
+      "loss": 0.4416,
       "step": 44985
     },
     {
-      "epoch": 1.58,
-      "learning_rate": 3.998955958326368e-05,
-      "loss": 0.2903,
+      "epoch": 1.6214365516992828,
+      "grad_norm": 0.1928412914276123,
+      "learning_rate": 3.9485917347809334e-05,
+      "loss": 0.4139,
       "step": 44990
     },
     {
-      "epoch": 1.58,
-      "learning_rate": 3.998727961713342e-05,
-      "loss": 0.305,
+      "epoch": 1.6216167513605075,
+      "grad_norm": 0.24302361905574799,
+      "learning_rate": 3.948353891986743e-05,
+      "loss": 0.4207,
       "step": 44995
     },
     {
-      "epoch": 1.58,
-      "learning_rate": 3.99849994564032e-05,
-      "loss": 0.2845,
+      "epoch": 1.621796951021732,
+      "grad_norm": 0.17859376966953278,
+      "learning_rate": 3.948116029459014e-05,
+      "loss": 0.4249,
       "step": 45000
     },
     {
-      "epoch": 1.58,
-      "eval_loss": 0.27851128578186035,
-      "eval_runtime": 10.5579,
-      "eval_samples_per_second": 9.472,
-      "eval_steps_per_second": 9.472,
+      "epoch": 1.621796951021732,
+      "eval_loss": 0.44367220997810364,
+      "eval_runtime": 3.5481,
+      "eval_samples_per_second": 28.184,
+      "eval_steps_per_second": 7.046,
       "step": 45000
     },
     {
-      "epoch": 1.58,
-      "learning_rate": 3.998271910110264e-05,
-      "loss": 0.3026,
+      "epoch": 1.6219771506829566,
+      "grad_norm": 0.18802441656589508,
+      "learning_rate": 3.947878147200983e-05,
+      "loss": 0.4004,
       "step": 45005
     },
     {
-      "epoch": 1.58,
-      "learning_rate": 3.998043855126135e-05,
-      "loss": 0.2869,
+      "epoch": 1.6221573503441813,
+      "grad_norm": 0.2054462730884552,
+      "learning_rate": 3.9476402452158926e-05,
+      "loss": 0.4113,
       "step": 45010
     },
     {
-      "epoch": 1.58,
-      "learning_rate": 3.997815780690895e-05,
-      "loss": 0.2622,
+      "epoch": 1.622337550005406,
+      "grad_norm": 0.17358975112438202,
+      "learning_rate": 3.947402323506984e-05,
+      "loss": 0.4125,
       "step": 45015
     },
     {
-      "epoch": 1.58,
-      "learning_rate": 3.997587686807502e-05,
-      "loss": 0.3,
+      "epoch": 1.6225177496666308,
+      "grad_norm": 0.2103101909160614,
+      "learning_rate": 3.9471643820774995e-05,
+      "loss": 0.4065,
       "step": 45020
     },
     {
-      "epoch": 1.58,
-      "learning_rate": 3.997359573478921e-05,
-      "loss": 0.2866,
+      "epoch": 1.6226979493278553,
+      "grad_norm": 0.18095333874225616,
+      "learning_rate": 3.94692642093068e-05,
+      "loss": 0.4137,
       "step": 45025
     },
     {
-      "epoch": 1.58,
-      "learning_rate": 3.997131440708112e-05,
-      "loss": 0.2857,
+      "epoch": 1.6228781489890798,
+      "grad_norm": 0.17532798647880554,
+      "learning_rate": 3.946688440069768e-05,
+      "loss": 0.4145,
       "step": 45030
     },
     {
-      "epoch": 1.58,
-      "learning_rate": 3.9969032884980386e-05,
-      "loss": 0.2889,
+      "epoch": 1.6230583486503045,
+      "grad_norm": 0.21436169743537903,
+      "learning_rate": 3.946450439498006e-05,
+      "loss": 0.432,
       "step": 45035
     },
     {
-      "epoch": 1.58,
-      "learning_rate": 3.9966751168516626e-05,
-      "loss": 0.2875,
+      "epoch": 1.6232385483115293,
+      "grad_norm": 0.15949729084968567,
+      "learning_rate": 3.946212419218638e-05,
+      "loss": 0.4074,
       "step": 45040
     },
     {
-      "epoch": 1.58,
-      "learning_rate": 3.996446925771946e-05,
-      "loss": 0.3088,
+      "epoch": 1.6234187479727538,
+      "grad_norm": 0.17256999015808105,
+      "learning_rate": 3.945974379234905e-05,
+      "loss": 0.4371,
       "step": 45045
     },
     {
-      "epoch": 1.58,
-      "learning_rate": 3.9962187152618525e-05,
-      "loss": 0.3145,
+      "epoch": 1.6235989476339785,
+      "grad_norm": 0.2011367380619049,
+      "learning_rate": 3.945736319550051e-05,
+      "loss": 0.443,
       "step": 45050
     },
     {
-      "epoch": 1.59,
-      "learning_rate": 3.995990485324345e-05,
-      "loss": 0.291,
+      "epoch": 1.623779147295203,
+      "grad_norm": 0.14212122559547424,
+      "learning_rate": 3.945498240167319e-05,
+      "loss": 0.4378,
       "step": 45055
     },
     {
-      "epoch": 1.59,
-      "learning_rate": 3.995762235962387e-05,
-      "loss": 0.2996,
+      "epoch": 1.6239593469564277,
+      "grad_norm": 0.21600797772407532,
+      "learning_rate": 3.9452601410899544e-05,
+      "loss": 0.4043,
       "step": 45060
     },
     {
-      "epoch": 1.59,
-      "learning_rate": 3.995533967178941e-05,
-      "loss": 0.2641,
+      "epoch": 1.6241395466176525,
+      "grad_norm": 0.16741763055324554,
+      "learning_rate": 3.9450220223212e-05,
+      "loss": 0.4199,
       "step": 45065
     },
     {
-      "epoch": 1.59,
-      "learning_rate": 3.995305678976974e-05,
-      "loss": 0.293,
+      "epoch": 1.624319746278877,
+      "grad_norm": 0.1476602405309677,
+      "learning_rate": 3.9447838838643e-05,
+      "loss": 0.3941,
       "step": 45070
     },
     {
-      "epoch": 1.59,
-      "learning_rate": 3.995077371359448e-05,
-      "loss": 0.2948,
+      "epoch": 1.6244999459401015,
+      "grad_norm": 0.17911425232887268,
+      "learning_rate": 3.9445457257224996e-05,
+      "loss": 0.4306,
       "step": 45075
     },
     {
-      "epoch": 1.59,
-      "learning_rate": 3.994849044329326e-05,
-      "loss": 0.2961,
+      "epoch": 1.6246801456013262,
+      "grad_norm": 0.17017336189746857,
+      "learning_rate": 3.944307547899042e-05,
+      "loss": 0.4172,
       "step": 45080
     },
     {
-      "epoch": 1.59,
-      "learning_rate": 3.9946206978895754e-05,
-      "loss": 0.2595,
+      "epoch": 1.624860345262551,
+      "grad_norm": 0.17422279715538025,
+      "learning_rate": 3.944069350397175e-05,
+      "loss": 0.3886,
       "step": 45085
     },
     {
-      "epoch": 1.59,
-      "learning_rate": 3.99439233204316e-05,
-      "loss": 0.294,
+      "epoch": 1.6250405449237757,
+      "grad_norm": 0.1593707799911499,
+      "learning_rate": 3.9438311332201424e-05,
+      "loss": 0.3895,
       "step": 45090
     },
     {
-      "epoch": 1.59,
-      "learning_rate": 3.994163946793045e-05,
-      "loss": 0.291,
+      "epoch": 1.6252207445850002,
+      "grad_norm": 0.1855667531490326,
+      "learning_rate": 3.94359289637119e-05,
+      "loss": 0.4233,
       "step": 45095
     },
     {
-      "epoch": 1.59,
-      "learning_rate": 3.993935542142196e-05,
-      "loss": 0.2901,
+      "epoch": 1.6254009442462247,
+      "grad_norm": 0.16529178619384766,
+      "learning_rate": 3.943354639853565e-05,
+      "loss": 0.3903,
       "step": 45100
     },
     {
-      "epoch": 1.59,
-      "learning_rate": 3.993707118093578e-05,
-      "loss": 0.2719,
+      "epoch": 1.6255811439074495,
+      "grad_norm": 0.15914855897426605,
+      "learning_rate": 3.9431163636705114e-05,
+      "loss": 0.4092,
       "step": 45105
     },
     {
-      "epoch": 1.59,
-      "learning_rate": 3.993478674650158e-05,
-      "loss": 0.2859,
+      "epoch": 1.6257613435686742,
+      "grad_norm": 0.21924299001693726,
+      "learning_rate": 3.942878067825277e-05,
+      "loss": 0.4527,
       "step": 45110
     },
     {
-      "epoch": 1.59,
-      "learning_rate": 3.993250211814901e-05,
-      "loss": 0.2932,
+      "epoch": 1.6259415432298987,
+      "grad_norm": 0.16359423100948334,
+      "learning_rate": 3.942639752321108e-05,
+      "loss": 0.3986,
       "step": 45115
     },
     {
-      "epoch": 1.59,
-      "learning_rate": 3.993021729590775e-05,
-      "loss": 0.2849,
+      "epoch": 1.6261217428911232,
+      "grad_norm": 0.23440222442150116,
+      "learning_rate": 3.942401417161252e-05,
+      "loss": 0.4406,
       "step": 45120
     },
     {
-      "epoch": 1.59,
-      "learning_rate": 3.992793227980744e-05,
-      "loss": 0.2834,
+      "epoch": 1.626301942552348,
+      "grad_norm": 0.2375732809305191,
+      "learning_rate": 3.942163062348956e-05,
+      "loss": 0.4347,
       "step": 45125
     },
     {
-      "epoch": 1.59,
-      "learning_rate": 3.992564706987779e-05,
-      "loss": 0.2918,
+      "epoch": 1.6264821422135727,
+      "grad_norm": 0.19271384179592133,
+      "learning_rate": 3.9419246878874674e-05,
+      "loss": 0.4476,
       "step": 45130
     },
     {
-      "epoch": 1.59,
-      "learning_rate": 3.992336166614844e-05,
-      "loss": 0.2689,
+      "epoch": 1.6266623418747974,
+      "grad_norm": 0.1916663497686386,
+      "learning_rate": 3.9416862937800335e-05,
+      "loss": 0.4026,
       "step": 45135
     },
     {
-      "epoch": 1.59,
-      "learning_rate": 3.9921076068649073e-05,
-      "loss": 0.3011,
+      "epoch": 1.626842541536022,
+      "grad_norm": 0.18741334974765778,
+      "learning_rate": 3.941447880029903e-05,
+      "loss": 0.3952,
       "step": 45140
     },
     {
-      "epoch": 1.59,
-      "learning_rate": 3.991879027740936e-05,
-      "loss": 0.2934,
+      "epoch": 1.6270227411972464,
+      "grad_norm": 0.16537748277187347,
+      "learning_rate": 3.941209446640325e-05,
+      "loss": 0.4506,
       "step": 45145
     },
     {
-      "epoch": 1.59,
-      "learning_rate": 3.9916504292459003e-05,
-      "loss": 0.2646,
+      "epoch": 1.6272029408584712,
+      "grad_norm": 0.17888925969600677,
+      "learning_rate": 3.940970993614547e-05,
+      "loss": 0.4331,
       "step": 45150
     },
     {
-      "epoch": 1.59,
-      "learning_rate": 3.991421811382765e-05,
-      "loss": 0.313,
+      "epoch": 1.627383140519696,
+      "grad_norm": 0.1680968850851059,
+      "learning_rate": 3.940732520955818e-05,
+      "loss": 0.4103,
       "step": 45155
     },
     {
-      "epoch": 1.59,
-      "learning_rate": 3.9911931741545014e-05,
-      "loss": 0.2736,
+      "epoch": 1.6275633401809204,
+      "grad_norm": 0.22438734769821167,
+      "learning_rate": 3.940494028667387e-05,
+      "loss": 0.4185,
       "step": 45160
     },
     {
-      "epoch": 1.59,
-      "learning_rate": 3.990964517564077e-05,
-      "loss": 0.288,
+      "epoch": 1.6277435398421451,
+      "grad_norm": 0.1568988710641861,
+      "learning_rate": 3.940255516752504e-05,
+      "loss": 0.404,
       "step": 45165
     },
     {
-      "epoch": 1.59,
-      "learning_rate": 3.9907358416144606e-05,
-      "loss": 0.315,
+      "epoch": 1.6279237395033697,
+      "grad_norm": 0.18171727657318115,
+      "learning_rate": 3.94001698521442e-05,
+      "loss": 0.3919,
       "step": 45170
     },
     {
-      "epoch": 1.59,
-      "learning_rate": 3.990507146308622e-05,
-      "loss": 0.2904,
+      "epoch": 1.6281039391645944,
+      "grad_norm": 0.20402833819389343,
+      "learning_rate": 3.9397784340563813e-05,
+      "loss": 0.3906,
       "step": 45175
     },
     {
-      "epoch": 1.59,
-      "learning_rate": 3.99027843164953e-05,
-      "loss": 0.2748,
+      "epoch": 1.6282841388258191,
+      "grad_norm": 0.2102474421262741,
+      "learning_rate": 3.939539863281641e-05,
+      "loss": 0.4252,
       "step": 45180
     },
     {
-      "epoch": 1.59,
-      "learning_rate": 3.990049697640156e-05,
-      "loss": 0.2799,
+      "epoch": 1.6284643384870436,
+      "grad_norm": 0.15504828095436096,
+      "learning_rate": 3.939301272893449e-05,
+      "loss": 0.394,
       "step": 45185
     },
     {
-      "epoch": 1.59,
-      "learning_rate": 3.9898209442834674e-05,
-      "loss": 0.2663,
+      "epoch": 1.6286445381482682,
+      "grad_norm": 0.19128242135047913,
+      "learning_rate": 3.939062662895055e-05,
+      "loss": 0.3909,
       "step": 45190
     },
     {
-      "epoch": 1.59,
-      "learning_rate": 3.9895921715824355e-05,
-      "loss": 0.2832,
+      "epoch": 1.6288247378094929,
+      "grad_norm": 0.18570300936698914,
+      "learning_rate": 3.938824033289712e-05,
+      "loss": 0.4437,
       "step": 45195
     },
     {
-      "epoch": 1.59,
-      "learning_rate": 3.989363379540032e-05,
-      "loss": 0.2778,
+      "epoch": 1.6290049374707176,
+      "grad_norm": 0.1933826506137848,
+      "learning_rate": 3.938585384080668e-05,
+      "loss": 0.4038,
       "step": 45200
     },
     {
-      "epoch": 1.59,
-      "learning_rate": 3.989134568159226e-05,
-      "loss": 0.3048,
+      "epoch": 1.6291851371319424,
+      "grad_norm": 0.17036275565624237,
+      "learning_rate": 3.93834671527118e-05,
+      "loss": 0.4106,
       "step": 45205
     },
     {
-      "epoch": 1.59,
-      "learning_rate": 3.988905737442989e-05,
-      "loss": 0.283,
+      "epoch": 1.6293653367931669,
+      "grad_norm": 0.2243850827217102,
+      "learning_rate": 3.9381080268644936e-05,
+      "loss": 0.4018,
       "step": 45210
     },
     {
-      "epoch": 1.59,
-      "learning_rate": 3.988676887394292e-05,
-      "loss": 0.2761,
+      "epoch": 1.6295455364543914,
+      "grad_norm": 0.2178872525691986,
+      "learning_rate": 3.9378693188638646e-05,
+      "loss": 0.4064,
       "step": 45215
     },
     {
-      "epoch": 1.59,
-      "learning_rate": 3.988448018016107e-05,
-      "loss": 0.2957,
+      "epoch": 1.629725736115616,
+      "grad_norm": 0.2029925286769867,
+      "learning_rate": 3.937630591272545e-05,
+      "loss": 0.4244,
       "step": 45220
     },
     {
-      "epoch": 1.59,
-      "learning_rate": 3.9882191293114055e-05,
-      "loss": 0.2997,
+      "epoch": 1.6299059357768408,
+      "grad_norm": 0.20879967510700226,
+      "learning_rate": 3.937391844093786e-05,
+      "loss": 0.4523,
       "step": 45225
     },
     {
-      "epoch": 1.59,
-      "learning_rate": 3.9879902212831586e-05,
-      "loss": 0.2857,
+      "epoch": 1.6300861354380654,
+      "grad_norm": 0.1707211285829544,
+      "learning_rate": 3.937153077330843e-05,
+      "loss": 0.4215,
       "step": 45230
     },
     {
-      "epoch": 1.59,
-      "learning_rate": 3.98776129393434e-05,
-      "loss": 0.2847,
+      "epoch": 1.6302663350992899,
+      "grad_norm": 0.17152579128742218,
+      "learning_rate": 3.936914290986966e-05,
+      "loss": 0.4481,
       "step": 45235
     },
     {
-      "epoch": 1.59,
-      "learning_rate": 3.987532347267922e-05,
-      "loss": 0.318,
+      "epoch": 1.6304465347605146,
+      "grad_norm": 0.17980226874351501,
+      "learning_rate": 3.9366754850654106e-05,
+      "loss": 0.3495,
       "step": 45240
     },
     {
-      "epoch": 1.59,
-      "learning_rate": 3.987303381286876e-05,
-      "loss": 0.2782,
+      "epoch": 1.6306267344217393,
+      "grad_norm": 0.18112671375274658,
+      "learning_rate": 3.9364366595694296e-05,
+      "loss": 0.3669,
       "step": 45245
     },
     {
-      "epoch": 1.59,
-      "learning_rate": 3.987074395994176e-05,
-      "loss": 0.2763,
+      "epoch": 1.630806934082964,
+      "grad_norm": 0.20279546082019806,
+      "learning_rate": 3.936197814502278e-05,
+      "loss": 0.3874,
       "step": 45250
     },
     {
-      "epoch": 1.59,
-      "learning_rate": 3.986845391392795e-05,
-      "loss": 0.2799,
+      "epoch": 1.6309871337441886,
+      "grad_norm": 0.15267014503479004,
+      "learning_rate": 3.9359589498672086e-05,
+      "loss": 0.4019,
       "step": 45255
     },
     {
-      "epoch": 1.59,
-      "learning_rate": 3.986616367485706e-05,
-      "loss": 0.2632,
+      "epoch": 1.631167333405413,
+      "grad_norm": 0.16568580269813538,
+      "learning_rate": 3.9357200656674764e-05,
+      "loss": 0.4233,
       "step": 45260
     },
     {
-      "epoch": 1.59,
-      "learning_rate": 3.986387324275884e-05,
-      "loss": 0.3109,
+      "epoch": 1.6313475330666378,
+      "grad_norm": 0.17788587510585785,
+      "learning_rate": 3.935481161906336e-05,
+      "loss": 0.3938,
       "step": 45265
     },
     {
-      "epoch": 1.59,
-      "learning_rate": 3.9861582617663014e-05,
-      "loss": 0.3049,
+      "epoch": 1.6315277327278626,
+      "grad_norm": 0.19289278984069824,
+      "learning_rate": 3.935242238587043e-05,
+      "loss": 0.4054,
       "step": 45270
     },
     {
-      "epoch": 1.59,
-      "learning_rate": 3.9859291799599334e-05,
-      "loss": 0.287,
+      "epoch": 1.631707932389087,
+      "grad_norm": 0.17799817025661469,
+      "learning_rate": 3.935003295712853e-05,
+      "loss": 0.4,
       "step": 45275
     },
     {
-      "epoch": 1.59,
-      "learning_rate": 3.985700078859754e-05,
-      "loss": 0.2702,
+      "epoch": 1.6318881320503116,
+      "grad_norm": 0.1564566195011139,
+      "learning_rate": 3.9347643332870206e-05,
+      "loss": 0.4249,
       "step": 45280
     },
     {
-      "epoch": 1.59,
-      "learning_rate": 3.985470958468739e-05,
-      "loss": 0.3012,
+      "epoch": 1.6320683317115363,
+      "grad_norm": 0.22281388938426971,
+      "learning_rate": 3.934525351312801e-05,
+      "loss": 0.396,
       "step": 45285
     },
     {
-      "epoch": 1.59,
-      "learning_rate": 3.985241818789862e-05,
-      "loss": 0.3009,
+      "epoch": 1.632248531372761,
+      "grad_norm": 0.17600180208683014,
+      "learning_rate": 3.934286349793452e-05,
+      "loss": 0.423,
       "step": 45290
     },
     {
-      "epoch": 1.59,
-      "learning_rate": 3.985012659826099e-05,
-      "loss": 0.3043,
+      "epoch": 1.6324287310339858,
+      "grad_norm": 0.19052568078041077,
+      "learning_rate": 3.93404732873223e-05,
+      "loss": 0.4046,
       "step": 45295
     },
     {
-      "epoch": 1.59,
-      "learning_rate": 3.984783481580425e-05,
-      "loss": 0.2747,
+      "epoch": 1.6326089306952103,
+      "grad_norm": 0.18254795670509338,
+      "learning_rate": 3.9338082881323896e-05,
+      "loss": 0.3924,
       "step": 45300
     },
     {
-      "epoch": 1.59,
-      "learning_rate": 3.984554284055816e-05,
-      "loss": 0.3035,
+      "epoch": 1.6327891303564348,
+      "grad_norm": 0.20248796045780182,
+      "learning_rate": 3.933569227997189e-05,
+      "loss": 0.4133,
       "step": 45305
     },
     {
-      "epoch": 1.59,
-      "learning_rate": 3.984325067255248e-05,
-      "loss": 0.2666,
+      "epoch": 1.6329693300176595,
+      "grad_norm": 0.1688736379146576,
+      "learning_rate": 3.9333301483298854e-05,
+      "loss": 0.3878,
       "step": 45310
     },
     {
-      "epoch": 1.59,
-      "learning_rate": 3.984095831181698e-05,
-      "loss": 0.3017,
+      "epoch": 1.6331495296788843,
+      "grad_norm": 0.17724154889583588,
+      "learning_rate": 3.9330910491337365e-05,
+      "loss": 0.3865,
       "step": 45315
     },
     {
-      "epoch": 1.59,
-      "learning_rate": 3.983866575838141e-05,
-      "loss": 0.2895,
+      "epoch": 1.633329729340109,
+      "grad_norm": 0.1554093062877655,
+      "learning_rate": 3.932851930411999e-05,
+      "loss": 0.4509,
       "step": 45320
     },
     {
-      "epoch": 1.59,
-      "learning_rate": 3.9836373012275543e-05,
-      "loss": 0.3074,
+      "epoch": 1.6335099290013335,
+      "grad_norm": 0.18391548097133636,
+      "learning_rate": 3.9326127921679315e-05,
+      "loss": 0.4208,
       "step": 45325
     },
     {
-      "epoch": 1.59,
-      "learning_rate": 3.983408007352916e-05,
-      "loss": 0.2672,
+      "epoch": 1.633690128662558,
+      "grad_norm": 0.18900564312934875,
+      "learning_rate": 3.932373634404793e-05,
+      "loss": 0.412,
       "step": 45330
     },
     {
-      "epoch": 1.6,
-      "learning_rate": 3.983178694217201e-05,
-      "loss": 0.2651,
+      "epoch": 1.6338703283237828,
+      "grad_norm": 0.17100130021572113,
+      "learning_rate": 3.932134457125839e-05,
+      "loss": 0.4065,
       "step": 45335
     },
     {
-      "epoch": 1.6,
-      "learning_rate": 3.982949361823388e-05,
-      "loss": 0.3096,
+      "epoch": 1.6340505279850075,
+      "grad_norm": 0.17979390919208527,
+      "learning_rate": 3.931895260334331e-05,
+      "loss": 0.3887,
       "step": 45340
     },
     {
-      "epoch": 1.6,
-      "learning_rate": 3.982720010174456e-05,
-      "loss": 0.2666,
+      "epoch": 1.634230727646232,
+      "grad_norm": 0.1904669851064682,
+      "learning_rate": 3.9316560440335275e-05,
+      "loss": 0.4119,
       "step": 45345
     },
     {
-      "epoch": 1.6,
-      "learning_rate": 3.982490639273381e-05,
-      "loss": 0.2852,
+      "epoch": 1.6344109273074565,
+      "grad_norm": 0.16896291077136993,
+      "learning_rate": 3.931416808226688e-05,
+      "loss": 0.4423,
       "step": 45350
     },
     {
-      "epoch": 1.6,
-      "learning_rate": 3.9822612491231425e-05,
-      "loss": 0.2856,
+      "epoch": 1.6345911269686813,
+      "grad_norm": 0.22345204651355743,
+      "learning_rate": 3.931177552917071e-05,
+      "loss": 0.4178,
       "step": 45355
     },
     {
-      "epoch": 1.6,
-      "learning_rate": 3.982031839726718e-05,
-      "loss": 0.2854,
+      "epoch": 1.634771326629906,
+      "grad_norm": 0.16224905848503113,
+      "learning_rate": 3.9309382781079375e-05,
+      "loss": 0.4169,
       "step": 45360
     },
     {
-      "epoch": 1.6,
-      "learning_rate": 3.981802411087087e-05,
-      "loss": 0.2934,
+      "epoch": 1.6349515262911307,
+      "grad_norm": 0.20584708452224731,
+      "learning_rate": 3.930698983802547e-05,
+      "loss": 0.4187,
       "step": 45365
     },
     {
-      "epoch": 1.6,
-      "learning_rate": 3.981572963207229e-05,
-      "loss": 0.3026,
+      "epoch": 1.6351317259523552,
+      "grad_norm": 0.17952153086662292,
+      "learning_rate": 3.930459670004159e-05,
+      "loss": 0.4389,
       "step": 45370
     },
     {
-      "epoch": 1.6,
-      "learning_rate": 3.981343496090121e-05,
-      "loss": 0.3136,
+      "epoch": 1.6353119256135797,
+      "grad_norm": 0.2221333533525467,
+      "learning_rate": 3.930220336716036e-05,
+      "loss": 0.4312,
       "step": 45375
     },
     {
-      "epoch": 1.6,
-      "learning_rate": 3.9811140097387456e-05,
-      "loss": 0.2767,
+      "epoch": 1.6354921252748045,
+      "grad_norm": 0.16802071034908295,
+      "learning_rate": 3.929980983941437e-05,
+      "loss": 0.4034,
       "step": 45380
     },
     {
-      "epoch": 1.6,
-      "learning_rate": 3.98088450415608e-05,
-      "loss": 0.299,
+      "epoch": 1.6356723249360292,
+      "grad_norm": 0.20047937333583832,
+      "learning_rate": 3.9297416116836246e-05,
+      "loss": 0.3942,
       "step": 45385
     },
     {
-      "epoch": 1.6,
-      "learning_rate": 3.9806549793451046e-05,
-      "loss": 0.2696,
+      "epoch": 1.6358525245972537,
+      "grad_norm": 0.19346770644187927,
+      "learning_rate": 3.9295022199458596e-05,
+      "loss": 0.4662,
       "step": 45390
     },
     {
-      "epoch": 1.6,
-      "learning_rate": 3.9804254353088003e-05,
-      "loss": 0.2684,
+      "epoch": 1.6360327242584782,
+      "grad_norm": 0.18764521181583405,
+      "learning_rate": 3.9292628087314034e-05,
+      "loss": 0.4294,
       "step": 45395
     },
     {
-      "epoch": 1.6,
-      "learning_rate": 3.980195872050147e-05,
-      "loss": 0.2764,
+      "epoch": 1.636212923919703,
+      "grad_norm": 0.1746283918619156,
+      "learning_rate": 3.9290233780435174e-05,
+      "loss": 0.4132,
       "step": 45400
     },
     {
-      "epoch": 1.6,
-      "learning_rate": 3.979966289572127e-05,
-      "loss": 0.269,
+      "epoch": 1.6363931235809277,
+      "grad_norm": 0.18788184225559235,
+      "learning_rate": 3.928783927885466e-05,
+      "loss": 0.4366,
       "step": 45405
     },
     {
-      "epoch": 1.6,
-      "learning_rate": 3.97973668787772e-05,
-      "loss": 0.269,
+      "epoch": 1.6365733232421524,
+      "grad_norm": 0.16621163487434387,
+      "learning_rate": 3.9285444582605086e-05,
+      "loss": 0.4393,
       "step": 45410
     },
     {
-      "epoch": 1.6,
-      "learning_rate": 3.9795070669699064e-05,
-      "loss": 0.2676,
+      "epoch": 1.636753522903377,
+      "grad_norm": 0.16439075767993927,
+      "learning_rate": 3.9283049691719106e-05,
+      "loss": 0.3832,
       "step": 45415
     },
     {
-      "epoch": 1.6,
-      "learning_rate": 3.9792774268516687e-05,
-      "loss": 0.2857,
+      "epoch": 1.6369337225646015,
+      "grad_norm": 0.15328091382980347,
+      "learning_rate": 3.928065460622933e-05,
+      "loss": 0.4351,
       "step": 45420
     },
     {
-      "epoch": 1.6,
-      "learning_rate": 3.979047767525989e-05,
-      "loss": 0.3097,
+      "epoch": 1.6371139222258262,
+      "grad_norm": 0.20914670825004578,
+      "learning_rate": 3.927825932616841e-05,
+      "loss": 0.3928,
       "step": 45425
     },
     {
-      "epoch": 1.6,
-      "learning_rate": 3.978818088995848e-05,
-      "loss": 0.2589,
+      "epoch": 1.637294121887051,
+      "grad_norm": 0.21218664944171906,
+      "learning_rate": 3.9275863851568964e-05,
+      "loss": 0.4199,
       "step": 45430
     },
     {
-      "epoch": 1.6,
-      "learning_rate": 3.97858839126423e-05,
-      "loss": 0.2776,
+      "epoch": 1.6374743215482754,
+      "grad_norm": 0.19584229588508606,
+      "learning_rate": 3.9273468182463645e-05,
+      "loss": 0.4132,
       "step": 45435
     },
     {
-      "epoch": 1.6,
-      "learning_rate": 3.978358674334115e-05,
-      "loss": 0.2829,
+      "epoch": 1.6376545212095002,
+      "grad_norm": 0.20541848242282867,
+      "learning_rate": 3.9271072318885076e-05,
+      "loss": 0.4287,
       "step": 45440
     },
     {
-      "epoch": 1.6,
-      "learning_rate": 3.9781289382084875e-05,
-      "loss": 0.3155,
+      "epoch": 1.6378347208707247,
+      "grad_norm": 0.20829689502716064,
+      "learning_rate": 3.9268676260865914e-05,
+      "loss": 0.4313,
       "step": 45445
     },
     {
-      "epoch": 1.6,
-      "learning_rate": 3.9778991828903286e-05,
-      "loss": 0.2622,
+      "epoch": 1.6380149205319494,
+      "grad_norm": 0.1655767560005188,
+      "learning_rate": 3.92662800084388e-05,
+      "loss": 0.4409,
       "step": 45450
     },
     {
-      "epoch": 1.6,
-      "learning_rate": 3.977669408382625e-05,
-      "loss": 0.2948,
+      "epoch": 1.6381951201931741,
+      "grad_norm": 0.17300541698932648,
+      "learning_rate": 3.9263883561636385e-05,
+      "loss": 0.4154,
       "step": 45455
     },
     {
-      "epoch": 1.6,
-      "learning_rate": 3.977439614688356e-05,
-      "loss": 0.2935,
+      "epoch": 1.6383753198543987,
+      "grad_norm": 0.220417782664299,
+      "learning_rate": 3.926148692049132e-05,
+      "loss": 0.418,
       "step": 45460
     },
     {
-      "epoch": 1.6,
-      "learning_rate": 3.9772098018105085e-05,
-      "loss": 0.259,
+      "epoch": 1.6385555195156232,
+      "grad_norm": 0.18571282923221588,
+      "learning_rate": 3.925909008503625e-05,
+      "loss": 0.4301,
       "step": 45465
     },
     {
-      "epoch": 1.6,
-      "learning_rate": 3.9769799697520645e-05,
-      "loss": 0.2842,
+      "epoch": 1.638735719176848,
+      "grad_norm": 0.18570081889629364,
+      "learning_rate": 3.925669305530384e-05,
+      "loss": 0.4454,
       "step": 45470
     },
     {
-      "epoch": 1.6,
-      "learning_rate": 3.976750118516009e-05,
-      "loss": 0.3044,
+      "epoch": 1.6389159188380726,
+      "grad_norm": 0.14735782146453857,
+      "learning_rate": 3.925429583132675e-05,
+      "loss": 0.4146,
       "step": 45475
     },
     {
-      "epoch": 1.6,
-      "learning_rate": 3.976520248105326e-05,
-      "loss": 0.2931,
+      "epoch": 1.6390961184992974,
+      "grad_norm": 0.17413799464702606,
+      "learning_rate": 3.925189841313764e-05,
+      "loss": 0.3904,
       "step": 45480
     },
     {
-      "epoch": 1.6,
-      "learning_rate": 3.976290358523002e-05,
-      "loss": 0.272,
+      "epoch": 1.6392763181605219,
+      "grad_norm": 0.27123942971229553,
+      "learning_rate": 3.924950080076917e-05,
+      "loss": 0.4315,
       "step": 45485
     },
     {
-      "epoch": 1.6,
-      "learning_rate": 3.9760604497720196e-05,
-      "loss": 0.2871,
+      "epoch": 1.6394565178217464,
+      "grad_norm": 0.1668565422296524,
+      "learning_rate": 3.924710299425401e-05,
+      "loss": 0.4216,
       "step": 45490
     },
     {
-      "epoch": 1.6,
-      "learning_rate": 3.975830521855366e-05,
-      "loss": 0.2799,
+      "epoch": 1.6396367174829711,
+      "grad_norm": 0.17070479691028595,
+      "learning_rate": 3.924470499362483e-05,
+      "loss": 0.423,
       "step": 45495
     },
     {
-      "epoch": 1.6,
-      "learning_rate": 3.975600574776024e-05,
-      "loss": 0.2895,
+      "epoch": 1.6398169171441959,
+      "grad_norm": 0.1955229640007019,
+      "learning_rate": 3.9242306798914305e-05,
+      "loss": 0.4145,
       "step": 45500
     },
     {
-      "epoch": 1.6,
-      "eval_loss": 0.2773810625076294,
-      "eval_runtime": 10.5526,
-      "eval_samples_per_second": 9.476,
-      "eval_steps_per_second": 9.476,
+      "epoch": 1.6398169171441959,
+      "eval_loss": 0.4432607591152191,
+      "eval_runtime": 3.5416,
+      "eval_samples_per_second": 28.236,
+      "eval_steps_per_second": 7.059,
       "step": 45500
     },
     {
-      "epoch": 1.6,
-      "learning_rate": 3.975370608536982e-05,
-      "loss": 0.2877,
+      "epoch": 1.6399971168054204,
+      "grad_norm": 0.1931176483631134,
+      "learning_rate": 3.9239908410155104e-05,
+      "loss": 0.412,
       "step": 45505
     },
     {
-      "epoch": 1.6,
-      "learning_rate": 3.975140623141226e-05,
-      "loss": 0.2763,
+      "epoch": 1.6401773164666449,
+      "grad_norm": 0.216093048453331,
+      "learning_rate": 3.923750982737991e-05,
+      "loss": 0.4194,
       "step": 45510
     },
     {
-      "epoch": 1.6,
-      "learning_rate": 3.9749106185917394e-05,
-      "loss": 0.2933,
+      "epoch": 1.6403575161278696,
+      "grad_norm": 0.17901061475276947,
+      "learning_rate": 3.92351110506214e-05,
+      "loss": 0.4208,
       "step": 45515
     },
     {
-      "epoch": 1.6,
-      "learning_rate": 3.974680594891511e-05,
-      "loss": 0.2566,
+      "epoch": 1.6405377157890944,
+      "grad_norm": 0.16819880902767181,
+      "learning_rate": 3.923271207991226e-05,
+      "loss": 0.3989,
       "step": 45520
     },
     {
-      "epoch": 1.6,
-      "learning_rate": 3.974450552043527e-05,
-      "loss": 0.2687,
+      "epoch": 1.640717915450319,
+      "grad_norm": 0.20534752309322357,
+      "learning_rate": 3.923031291528517e-05,
+      "loss": 0.4011,
       "step": 45525
     },
     {
-      "epoch": 1.6,
-      "learning_rate": 3.974220490050775e-05,
-      "loss": 0.2882,
+      "epoch": 1.6408981151115436,
+      "grad_norm": 0.16648423671722412,
+      "learning_rate": 3.922791355677282e-05,
+      "loss": 0.4435,
       "step": 45530
     },
     {
-      "epoch": 1.6,
-      "learning_rate": 3.97399040891624e-05,
-      "loss": 0.2873,
+      "epoch": 1.641078314772768,
+      "grad_norm": 0.20146596431732178,
+      "learning_rate": 3.9225514004407916e-05,
+      "loss": 0.4431,
       "step": 45535
     },
     {
-      "epoch": 1.6,
-      "learning_rate": 3.973760308642912e-05,
-      "loss": 0.2725,
+      "epoch": 1.6412585144339928,
+      "grad_norm": 0.18600507080554962,
+      "learning_rate": 3.922311425822313e-05,
+      "loss": 0.4459,
       "step": 45540
     },
     {
-      "epoch": 1.6,
-      "learning_rate": 3.973530189233777e-05,
-      "loss": 0.297,
+      "epoch": 1.6414387140952176,
+      "grad_norm": 0.19291111826896667,
+      "learning_rate": 3.922071431825116e-05,
+      "loss": 0.3978,
       "step": 45545
     },
     {
-      "epoch": 1.6,
-      "learning_rate": 3.9733000506918245e-05,
-      "loss": 0.2573,
+      "epoch": 1.641618913756442,
+      "grad_norm": 0.18025104701519012,
+      "learning_rate": 3.921831418452472e-05,
+      "loss": 0.4321,
       "step": 45550
     },
     {
-      "epoch": 1.6,
-      "learning_rate": 3.973069893020041e-05,
-      "loss": 0.284,
+      "epoch": 1.6417991134176668,
+      "grad_norm": 0.18391737341880798,
+      "learning_rate": 3.921591385707649e-05,
+      "loss": 0.4303,
       "step": 45555
     },
     {
-      "epoch": 1.6,
-      "learning_rate": 3.972839716221416e-05,
-      "loss": 0.2903,
+      "epoch": 1.6419793130788913,
+      "grad_norm": 0.23778654634952545,
+      "learning_rate": 3.921351333593919e-05,
+      "loss": 0.4226,
       "step": 45560
     },
     {
-      "epoch": 1.6,
-      "learning_rate": 3.972609520298938e-05,
-      "loss": 0.2788,
+      "epoch": 1.642159512740116,
+      "grad_norm": 0.1602877527475357,
+      "learning_rate": 3.921111262114553e-05,
+      "loss": 0.3877,
       "step": 45565
     },
     {
-      "epoch": 1.6,
-      "learning_rate": 3.9723793052555957e-05,
-      "loss": 0.2796,
+      "epoch": 1.6423397124013408,
+      "grad_norm": 0.17459672689437866,
+      "learning_rate": 3.920871171272821e-05,
+      "loss": 0.4258,
       "step": 45570
     },
     {
-      "epoch": 1.6,
-      "learning_rate": 3.9721490710943784e-05,
-      "loss": 0.2792,
+      "epoch": 1.6425199120625653,
+      "grad_norm": 0.16200929880142212,
+      "learning_rate": 3.920631061071994e-05,
+      "loss": 0.379,
       "step": 45575
     },
     {
-      "epoch": 1.6,
-      "learning_rate": 3.971918817818276e-05,
-      "loss": 0.3132,
+      "epoch": 1.6427001117237898,
+      "grad_norm": 0.2194150686264038,
+      "learning_rate": 3.9203909315153445e-05,
+      "loss": 0.4539,
       "step": 45580
     },
     {
-      "epoch": 1.6,
-      "learning_rate": 3.971688545430277e-05,
-      "loss": 0.2981,
+      "epoch": 1.6428803113850146,
+      "grad_norm": 0.15917415916919708,
+      "learning_rate": 3.920150782606142e-05,
+      "loss": 0.4171,
       "step": 45585
     },
     {
-      "epoch": 1.6,
-      "learning_rate": 3.971458253933373e-05,
-      "loss": 0.2772,
+      "epoch": 1.6430605110462393,
+      "grad_norm": 0.16585630178451538,
+      "learning_rate": 3.919910614347662e-05,
+      "loss": 0.4325,
       "step": 45590
     },
     {
-      "epoch": 1.6,
-      "learning_rate": 3.971227943330553e-05,
-      "loss": 0.3021,
+      "epoch": 1.643240710707464,
+      "grad_norm": 0.1744535267353058,
+      "learning_rate": 3.9196704267431735e-05,
+      "loss": 0.4188,
       "step": 45595
     },
     {
-      "epoch": 1.6,
-      "learning_rate": 3.9709976136248074e-05,
-      "loss": 0.2925,
+      "epoch": 1.6434209103686885,
+      "grad_norm": 0.17278815805912018,
+      "learning_rate": 3.91943021979595e-05,
+      "loss": 0.4041,
       "step": 45600
     },
     {
-      "epoch": 1.6,
-      "learning_rate": 3.9707672648191274e-05,
-      "loss": 0.2876,
+      "epoch": 1.643601110029913,
+      "grad_norm": 0.19840632379055023,
+      "learning_rate": 3.919189993509265e-05,
+      "loss": 0.4229,
       "step": 45605
     },
     {
-      "epoch": 1.6,
-      "learning_rate": 3.9705368969165036e-05,
-      "loss": 0.293,
+      "epoch": 1.6437813096911378,
+      "grad_norm": 0.178060382604599,
+      "learning_rate": 3.918949747886391e-05,
+      "loss": 0.4469,
       "step": 45610
     },
     {
-      "epoch": 1.6,
-      "learning_rate": 3.9703065099199275e-05,
-      "loss": 0.2589,
+      "epoch": 1.6439615093523625,
+      "grad_norm": 0.1763332337141037,
+      "learning_rate": 3.918709482930602e-05,
+      "loss": 0.3798,
       "step": 45615
     },
     {
-      "epoch": 1.61,
-      "learning_rate": 3.970076103832391e-05,
-      "loss": 0.296,
+      "epoch": 1.644141709013587,
+      "grad_norm": 0.17714865505695343,
+      "learning_rate": 3.918469198645171e-05,
+      "loss": 0.4069,
       "step": 45620
     },
     {
-      "epoch": 1.61,
-      "learning_rate": 3.969845678656884e-05,
-      "loss": 0.2924,
+      "epoch": 1.6443219086748115,
+      "grad_norm": 0.18300695717334747,
+      "learning_rate": 3.918228895033371e-05,
+      "loss": 0.3987,
       "step": 45625
     },
     {
-      "epoch": 1.61,
-      "learning_rate": 3.9696152343963996e-05,
-      "loss": 0.3112,
+      "epoch": 1.6445021083360363,
+      "grad_norm": 0.1943272352218628,
+      "learning_rate": 3.9179885720984775e-05,
+      "loss": 0.4106,
       "step": 45630
     },
     {
-      "epoch": 1.61,
-      "learning_rate": 3.9693847710539305e-05,
-      "loss": 0.3033,
+      "epoch": 1.644682307997261,
+      "grad_norm": 0.20393919944763184,
+      "learning_rate": 3.917748229843764e-05,
+      "loss": 0.4568,
       "step": 45635
     },
     {
-      "epoch": 1.61,
-      "learning_rate": 3.9691542886324684e-05,
-      "loss": 0.2816,
+      "epoch": 1.6448625076584857,
+      "grad_norm": 0.18924960494041443,
+      "learning_rate": 3.917507868272506e-05,
+      "loss": 0.4561,
       "step": 45640
     },
     {
-      "epoch": 1.61,
-      "learning_rate": 3.968923787135006e-05,
-      "loss": 0.2761,
+      "epoch": 1.6450427073197103,
+      "grad_norm": 0.1751289963722229,
+      "learning_rate": 3.9172674873879766e-05,
+      "loss": 0.4143,
       "step": 45645
     },
     {
-      "epoch": 1.61,
-      "learning_rate": 3.9686932665645357e-05,
-      "loss": 0.2845,
+      "epoch": 1.6452229069809348,
+      "grad_norm": 0.16475655138492584,
+      "learning_rate": 3.917027087193453e-05,
+      "loss": 0.3721,
       "step": 45650
     },
     {
-      "epoch": 1.61,
-      "learning_rate": 3.968462726924052e-05,
-      "loss": 0.2957,
+      "epoch": 1.6454031066421595,
+      "grad_norm": 0.2124844789505005,
+      "learning_rate": 3.9167866676922096e-05,
+      "loss": 0.4227,
       "step": 45655
     },
     {
-      "epoch": 1.61,
-      "learning_rate": 3.9682321682165474e-05,
-      "loss": 0.2873,
+      "epoch": 1.6455833063033842,
+      "grad_norm": 0.17801690101623535,
+      "learning_rate": 3.916546228887521e-05,
+      "loss": 0.4318,
       "step": 45660
     },
     {
-      "epoch": 1.61,
-      "learning_rate": 3.968001590445016e-05,
-      "loss": 0.2757,
+      "epoch": 1.6457635059646087,
+      "grad_norm": 0.17565682530403137,
+      "learning_rate": 3.9163057707826654e-05,
+      "loss": 0.3954,
       "step": 45665
     },
     {
-      "epoch": 1.61,
-      "learning_rate": 3.967770993612451e-05,
-      "loss": 0.2825,
+      "epoch": 1.6459437056258335,
+      "grad_norm": 0.17757514119148254,
+      "learning_rate": 3.916065293380917e-05,
+      "loss": 0.4042,
       "step": 45670
     },
     {
-      "epoch": 1.61,
-      "learning_rate": 3.967540377721847e-05,
-      "loss": 0.2928,
+      "epoch": 1.646123905287058,
+      "grad_norm": 0.20188312232494354,
+      "learning_rate": 3.9158247966855545e-05,
+      "loss": 0.4265,
       "step": 45675
     },
     {
-      "epoch": 1.61,
-      "learning_rate": 3.9673097427761975e-05,
-      "loss": 0.2877,
+      "epoch": 1.6463041049482827,
+      "grad_norm": 0.19370879232883453,
+      "learning_rate": 3.915584280699853e-05,
+      "loss": 0.4094,
       "step": 45680
     },
     {
-      "epoch": 1.61,
-      "learning_rate": 3.967079088778499e-05,
-      "loss": 0.2864,
+      "epoch": 1.6464843046095075,
+      "grad_norm": 0.15388374030590057,
+      "learning_rate": 3.915343745427089e-05,
+      "loss": 0.3697,
       "step": 45685
     },
     {
-      "epoch": 1.61,
-      "learning_rate": 3.9668484157317455e-05,
-      "loss": 0.2722,
+      "epoch": 1.646664504270732,
+      "grad_norm": 0.19175679981708527,
+      "learning_rate": 3.9151031908705406e-05,
+      "loss": 0.4483,
       "step": 45690
     },
     {
-      "epoch": 1.61,
-      "learning_rate": 3.966617723638931e-05,
-      "loss": 0.281,
+      "epoch": 1.6468447039319565,
+      "grad_norm": 0.19182166457176208,
+      "learning_rate": 3.9148626170334854e-05,
+      "loss": 0.3927,
       "step": 45695
     },
     {
-      "epoch": 1.61,
-      "learning_rate": 3.966387012503053e-05,
-      "loss": 0.2646,
+      "epoch": 1.6470249035931812,
+      "grad_norm": 0.1689678132534027,
+      "learning_rate": 3.914622023919201e-05,
+      "loss": 0.4264,
       "step": 45700
     },
     {
-      "epoch": 1.61,
-      "learning_rate": 3.966156282327106e-05,
-      "loss": 0.2854,
+      "epoch": 1.647205103254406,
+      "grad_norm": 0.17200689017772675,
+      "learning_rate": 3.914381411530965e-05,
+      "loss": 0.4504,
       "step": 45705
     },
     {
-      "epoch": 1.61,
-      "learning_rate": 3.965925533114085e-05,
-      "loss": 0.2768,
+      "epoch": 1.6473853029156307,
+      "grad_norm": 0.17576947808265686,
+      "learning_rate": 3.914140779872056e-05,
+      "loss": 0.4333,
       "step": 45710
     },
     {
-      "epoch": 1.61,
-      "learning_rate": 3.965694764866988e-05,
-      "loss": 0.2905,
+      "epoch": 1.6475655025768552,
+      "grad_norm": 0.18409965932369232,
+      "learning_rate": 3.913900128945753e-05,
+      "loss": 0.3983,
       "step": 45715
     },
     {
-      "epoch": 1.61,
-      "learning_rate": 3.9654639775888093e-05,
-      "loss": 0.2845,
+      "epoch": 1.6477457022380797,
+      "grad_norm": 0.1899709552526474,
+      "learning_rate": 3.913659458755335e-05,
+      "loss": 0.3916,
       "step": 45720
     },
     {
-      "epoch": 1.61,
-      "learning_rate": 3.9652331712825474e-05,
-      "loss": 0.2775,
+      "epoch": 1.6479259018993044,
+      "grad_norm": 0.1969635933637619,
+      "learning_rate": 3.9134187693040806e-05,
+      "loss": 0.4051,
       "step": 45725
     },
     {
-      "epoch": 1.61,
-      "learning_rate": 3.9650023459511983e-05,
-      "loss": 0.2828,
+      "epoch": 1.6481061015605292,
+      "grad_norm": 0.15576127171516418,
+      "learning_rate": 3.913178060595268e-05,
+      "loss": 0.4128,
       "step": 45730
     },
     {
-      "epoch": 1.61,
-      "learning_rate": 3.9647715015977594e-05,
-      "loss": 0.3034,
+      "epoch": 1.6482863012217537,
+      "grad_norm": 0.17115731537342072,
+      "learning_rate": 3.9129373326321785e-05,
+      "loss": 0.4159,
       "step": 45735
     },
     {
-      "epoch": 1.61,
-      "learning_rate": 3.964540638225227e-05,
-      "loss": 0.2993,
+      "epoch": 1.6484665008829782,
+      "grad_norm": 0.1695828139781952,
+      "learning_rate": 3.912696585418092e-05,
+      "loss": 0.4314,
       "step": 45740
     },
     {
-      "epoch": 1.61,
-      "learning_rate": 3.964309755836601e-05,
-      "loss": 0.3023,
+      "epoch": 1.648646700544203,
+      "grad_norm": 0.18110521137714386,
+      "learning_rate": 3.912455818956288e-05,
+      "loss": 0.4301,
       "step": 45745
     },
     {
-      "epoch": 1.61,
-      "learning_rate": 3.964078854434876e-05,
-      "loss": 0.2665,
+      "epoch": 1.6488269002054277,
+      "grad_norm": 0.15694189071655273,
+      "learning_rate": 3.912215033250046e-05,
+      "loss": 0.3813,
       "step": 45750
     },
     {
-      "epoch": 1.61,
-      "learning_rate": 3.9638479340230535e-05,
-      "loss": 0.2827,
+      "epoch": 1.6490070998666524,
+      "grad_norm": 0.1641474962234497,
+      "learning_rate": 3.911974228302647e-05,
+      "loss": 0.3951,
       "step": 45755
     },
     {
-      "epoch": 1.61,
-      "learning_rate": 3.96361699460413e-05,
-      "loss": 0.2882,
+      "epoch": 1.649187299527877,
+      "grad_norm": 0.20309121906757355,
+      "learning_rate": 3.911733404117375e-05,
+      "loss": 0.4113,
       "step": 45760
     },
     {
-      "epoch": 1.61,
-      "learning_rate": 3.9633860361811036e-05,
-      "loss": 0.2943,
+      "epoch": 1.6493674991891014,
+      "grad_norm": 0.219468891620636,
+      "learning_rate": 3.9114925606975064e-05,
+      "loss": 0.3907,
       "step": 45765
     },
     {
-      "epoch": 1.61,
-      "learning_rate": 3.963155058756975e-05,
-      "loss": 0.2901,
+      "epoch": 1.6495476988503261,
+      "grad_norm": 0.17919921875,
+      "learning_rate": 3.9112516980463255e-05,
+      "loss": 0.418,
       "step": 45770
     },
     {
-      "epoch": 1.61,
-      "learning_rate": 3.962924062334742e-05,
-      "loss": 0.283,
+      "epoch": 1.6497278985115509,
+      "grad_norm": 0.15530207753181458,
+      "learning_rate": 3.911010816167113e-05,
+      "loss": 0.3502,
       "step": 45775
     },
     {
-      "epoch": 1.61,
-      "learning_rate": 3.962693046917404e-05,
-      "loss": 0.2823,
+      "epoch": 1.6499080981727754,
+      "grad_norm": 0.15395016968250275,
+      "learning_rate": 3.910769915063153e-05,
+      "loss": 0.4206,
       "step": 45780
     },
     {
-      "epoch": 1.61,
-      "learning_rate": 3.96246201250796e-05,
-      "loss": 0.2808,
+      "epoch": 1.650088297834,
+      "grad_norm": 0.17081965506076813,
+      "learning_rate": 3.910528994737725e-05,
+      "loss": 0.4347,
       "step": 45785
     },
     {
-      "epoch": 1.61,
-      "learning_rate": 3.962230959109411e-05,
-      "loss": 0.2731,
+      "epoch": 1.6502684974952246,
+      "grad_norm": 0.2129909098148346,
+      "learning_rate": 3.910288055194112e-05,
+      "loss": 0.4404,
       "step": 45790
     },
     {
-      "epoch": 1.61,
-      "learning_rate": 3.961999886724756e-05,
-      "loss": 0.2726,
+      "epoch": 1.6504486971564494,
+      "grad_norm": 0.203140988945961,
+      "learning_rate": 3.910047096435598e-05,
+      "loss": 0.4398,
       "step": 45795
     },
     {
-      "epoch": 1.61,
-      "learning_rate": 3.961768795356998e-05,
-      "loss": 0.306,
+      "epoch": 1.650628896817674,
+      "grad_norm": 0.19998563826084137,
+      "learning_rate": 3.909806118465466e-05,
+      "loss": 0.4061,
       "step": 45800
     },
     {
-      "epoch": 1.61,
-      "learning_rate": 3.961537685009134e-05,
-      "loss": 0.2753,
+      "epoch": 1.6508090964788986,
+      "grad_norm": 0.1874360889196396,
+      "learning_rate": 3.909565121286997e-05,
+      "loss": 0.4194,
       "step": 45805
     },
     {
-      "epoch": 1.61,
-      "learning_rate": 3.961306555684167e-05,
-      "loss": 0.2986,
+      "epoch": 1.6509892961401231,
+      "grad_norm": 0.19713318347930908,
+      "learning_rate": 3.909324104903477e-05,
+      "loss": 0.4072,
       "step": 45810
     },
     {
-      "epoch": 1.61,
-      "learning_rate": 3.961075407385097e-05,
-      "loss": 0.2763,
+      "epoch": 1.6511694958013479,
+      "grad_norm": 0.17209462821483612,
+      "learning_rate": 3.9090830693181885e-05,
+      "loss": 0.417,
       "step": 45815
     },
     {
-      "epoch": 1.61,
-      "learning_rate": 3.960844240114926e-05,
-      "loss": 0.2872,
+      "epoch": 1.6513496954625726,
+      "grad_norm": 0.17033915221691132,
+      "learning_rate": 3.9088420145344164e-05,
+      "loss": 0.3954,
       "step": 45820
     },
     {
-      "epoch": 1.61,
-      "learning_rate": 3.960613053876655e-05,
-      "loss": 0.2956,
+      "epoch": 1.6515298951237973,
+      "grad_norm": 0.19625675678253174,
+      "learning_rate": 3.9086009405554445e-05,
+      "loss": 0.3812,
       "step": 45825
     },
     {
-      "epoch": 1.61,
-      "learning_rate": 3.960381848673286e-05,
-      "loss": 0.306,
+      "epoch": 1.6517100947850218,
+      "grad_norm": 0.16199907660484314,
+      "learning_rate": 3.908359847384557e-05,
+      "loss": 0.429,
       "step": 45830
     },
     {
-      "epoch": 1.61,
-      "learning_rate": 3.9601506245078216e-05,
-      "loss": 0.2837,
+      "epoch": 1.6518902944462464,
+      "grad_norm": 0.2132118195295334,
+      "learning_rate": 3.9081187350250406e-05,
+      "loss": 0.4157,
       "step": 45835
     },
     {
-      "epoch": 1.61,
-      "learning_rate": 3.959919381383262e-05,
-      "loss": 0.2831,
+      "epoch": 1.652070494107471,
+      "grad_norm": 0.17869247496128082,
+      "learning_rate": 3.9078776034801775e-05,
+      "loss": 0.399,
       "step": 45840
     },
     {
-      "epoch": 1.61,
-      "learning_rate": 3.959688119302614e-05,
-      "loss": 0.2889,
+      "epoch": 1.6522506937686958,
+      "grad_norm": 0.15512903034687042,
+      "learning_rate": 3.907636452753256e-05,
+      "loss": 0.4001,
       "step": 45845
     },
     {
-      "epoch": 1.61,
-      "learning_rate": 3.9594568382688755e-05,
-      "loss": 0.2778,
+      "epoch": 1.6524308934299203,
+      "grad_norm": 0.15080289542675018,
+      "learning_rate": 3.907395282847559e-05,
+      "loss": 0.4167,
       "step": 45850
     },
     {
-      "epoch": 1.61,
-      "learning_rate": 3.959225538285053e-05,
-      "loss": 0.2599,
+      "epoch": 1.6526110930911448,
+      "grad_norm": 0.19971080124378204,
+      "learning_rate": 3.907154093766375e-05,
+      "loss": 0.4297,
       "step": 45855
     },
     {
-      "epoch": 1.61,
-      "learning_rate": 3.9589942193541486e-05,
-      "loss": 0.2797,
+      "epoch": 1.6527912927523696,
+      "grad_norm": 0.19148339331150055,
+      "learning_rate": 3.9069128855129876e-05,
+      "loss": 0.411,
       "step": 45860
     },
     {
-      "epoch": 1.61,
-      "learning_rate": 3.958762881479165e-05,
-      "loss": 0.293,
+      "epoch": 1.6529714924135943,
+      "grad_norm": 0.15302042663097382,
+      "learning_rate": 3.906671658090686e-05,
+      "loss": 0.4408,
       "step": 45865
     },
     {
-      "epoch": 1.61,
-      "learning_rate": 3.9585315246631074e-05,
-      "loss": 0.283,
+      "epoch": 1.653151692074819,
+      "grad_norm": 0.1940101683139801,
+      "learning_rate": 3.9064304115027544e-05,
+      "loss": 0.3945,
       "step": 45870
     },
     {
-      "epoch": 1.61,
-      "learning_rate": 3.9583001489089786e-05,
-      "loss": 0.2734,
+      "epoch": 1.6533318917360436,
+      "grad_norm": 0.19384680688381195,
+      "learning_rate": 3.9061891457524814e-05,
+      "loss": 0.4195,
       "step": 45875
     },
     {
-      "epoch": 1.61,
-      "learning_rate": 3.9580687542197834e-05,
-      "loss": 0.3061,
+      "epoch": 1.653512091397268,
+      "grad_norm": 0.2088027447462082,
+      "learning_rate": 3.9059478608431526e-05,
+      "loss": 0.3886,
       "step": 45880
     },
     {
-      "epoch": 1.61,
-      "learning_rate": 3.957837340598527e-05,
-      "loss": 0.2923,
+      "epoch": 1.6536922910584928,
+      "grad_norm": 0.16105620563030243,
+      "learning_rate": 3.905706556778057e-05,
+      "loss": 0.4008,
       "step": 45885
     },
     {
-      "epoch": 1.61,
-      "learning_rate": 3.957605908048213e-05,
-      "loss": 0.3052,
+      "epoch": 1.6538724907197175,
+      "grad_norm": 0.1838972270488739,
+      "learning_rate": 3.9054652335604814e-05,
+      "loss": 0.428,
       "step": 45890
     },
     {
-      "epoch": 1.61,
-      "learning_rate": 3.9573744565718476e-05,
-      "loss": 0.2921,
+      "epoch": 1.654052690380942,
+      "grad_norm": 0.18796832859516144,
+      "learning_rate": 3.905223891193715e-05,
+      "loss": 0.4167,
       "step": 45895
     },
     {
-      "epoch": 1.61,
-      "learning_rate": 3.957142986172434e-05,
-      "loss": 0.2828,
+      "epoch": 1.6542328900421666,
+      "grad_norm": 0.18928878009319305,
+      "learning_rate": 3.9049825296810436e-05,
+      "loss": 0.3747,
       "step": 45900
     },
     {
-      "epoch": 1.62,
-      "learning_rate": 3.95691149685298e-05,
-      "loss": 0.29,
+      "epoch": 1.6544130897033913,
+      "grad_norm": 0.1923985332250595,
+      "learning_rate": 3.904741149025759e-05,
+      "loss": 0.3859,
       "step": 45905
     },
     {
-      "epoch": 1.62,
-      "learning_rate": 3.956679988616489e-05,
-      "loss": 0.313,
+      "epoch": 1.654593289364616,
+      "grad_norm": 0.2407243549823761,
+      "learning_rate": 3.904499749231147e-05,
+      "loss": 0.3979,
       "step": 45910
     },
     {
-      "epoch": 1.62,
-      "learning_rate": 3.9564484614659704e-05,
-      "loss": 0.2805,
+      "epoch": 1.6547734890258408,
+      "grad_norm": 0.1961093693971634,
+      "learning_rate": 3.904258330300498e-05,
+      "loss": 0.4073,
       "step": 45915
     },
     {
-      "epoch": 1.62,
-      "learning_rate": 3.956216915404427e-05,
-      "loss": 0.2935,
+      "epoch": 1.6549536886870653,
+      "grad_norm": 0.23786601424217224,
+      "learning_rate": 3.904016892237101e-05,
+      "loss": 0.4582,
       "step": 45920
     },
     {
-      "epoch": 1.62,
-      "learning_rate": 3.955985350434868e-05,
-      "loss": 0.3025,
+      "epoch": 1.6551338883482898,
+      "grad_norm": 0.1749258190393448,
+      "learning_rate": 3.903775435044246e-05,
+      "loss": 0.4478,
       "step": 45925
     },
     {
-      "epoch": 1.62,
-      "learning_rate": 3.955753766560297e-05,
-      "loss": 0.2691,
+      "epoch": 1.6553140880095145,
+      "grad_norm": 0.16315124928951263,
+      "learning_rate": 3.903533958725223e-05,
+      "loss": 0.3567,
       "step": 45930
     },
     {
-      "epoch": 1.62,
-      "learning_rate": 3.955522163783724e-05,
-      "loss": 0.2919,
+      "epoch": 1.6554942876707393,
+      "grad_norm": 0.14763472974300385,
+      "learning_rate": 3.9032924632833205e-05,
+      "loss": 0.3758,
       "step": 45935
     },
     {
-      "epoch": 1.62,
-      "learning_rate": 3.955290542108154e-05,
-      "loss": 0.289,
+      "epoch": 1.6556744873319638,
+      "grad_norm": 0.22486995160579681,
+      "learning_rate": 3.90305094872183e-05,
+      "loss": 0.4554,
       "step": 45940
     },
     {
-      "epoch": 1.62,
-      "learning_rate": 3.955058901536596e-05,
-      "loss": 0.2994,
+      "epoch": 1.6558546869931885,
+      "grad_norm": 0.19883042573928833,
+      "learning_rate": 3.902809415044043e-05,
+      "loss": 0.4338,
       "step": 45945
     },
     {
-      "epoch": 1.62,
-      "learning_rate": 3.9548272420720566e-05,
-      "loss": 0.2907,
+      "epoch": 1.656034886654413,
+      "grad_norm": 0.18573078513145447,
+      "learning_rate": 3.902567862253248e-05,
+      "loss": 0.4199,
       "step": 45950
     },
     {
-      "epoch": 1.62,
-      "learning_rate": 3.9545955637175446e-05,
-      "loss": 0.2685,
+      "epoch": 1.6562150863156377,
+      "grad_norm": 0.18856869637966156,
+      "learning_rate": 3.902326290352738e-05,
+      "loss": 0.3974,
       "step": 45955
     },
     {
-      "epoch": 1.62,
-      "learning_rate": 3.9543638664760675e-05,
-      "loss": 0.2935,
+      "epoch": 1.6563952859768625,
+      "grad_norm": 0.18239809572696686,
+      "learning_rate": 3.902084699345804e-05,
+      "loss": 0.433,
       "step": 45960
     },
     {
-      "epoch": 1.62,
-      "learning_rate": 3.9541321503506346e-05,
-      "loss": 0.279,
+      "epoch": 1.656575485638087,
+      "grad_norm": 0.23285090923309326,
+      "learning_rate": 3.9018430892357376e-05,
+      "loss": 0.4054,
       "step": 45965
     },
     {
-      "epoch": 1.62,
-      "learning_rate": 3.953900415344254e-05,
-      "loss": 0.3079,
+      "epoch": 1.6567556852993115,
+      "grad_norm": 0.16723239421844482,
+      "learning_rate": 3.90160146002583e-05,
+      "loss": 0.4016,
       "step": 45970
     },
     {
-      "epoch": 1.62,
-      "learning_rate": 3.9536686614599346e-05,
-      "loss": 0.2906,
+      "epoch": 1.6569358849605362,
+      "grad_norm": 0.14857235550880432,
+      "learning_rate": 3.901359811719374e-05,
+      "loss": 0.4149,
       "step": 45975
     },
     {
-      "epoch": 1.62,
-      "learning_rate": 3.953436888700684e-05,
-      "loss": 0.2907,
+      "epoch": 1.657116084621761,
+      "grad_norm": 0.1981445550918579,
+      "learning_rate": 3.901118144319662e-05,
+      "loss": 0.403,
       "step": 45980
     },
     {
-      "epoch": 1.62,
-      "learning_rate": 3.9532050970695153e-05,
-      "loss": 0.2623,
+      "epoch": 1.6572962842829857,
+      "grad_norm": 0.1867840737104416,
+      "learning_rate": 3.9008764578299866e-05,
+      "loss": 0.4106,
       "step": 45985
     },
     {
-      "epoch": 1.62,
-      "learning_rate": 3.952973286569436e-05,
-      "loss": 0.2741,
+      "epoch": 1.6574764839442102,
+      "grad_norm": 0.21596737205982208,
+      "learning_rate": 3.900634752253641e-05,
+      "loss": 0.4186,
       "step": 45990
     },
     {
-      "epoch": 1.62,
-      "learning_rate": 3.9527414572034555e-05,
-      "loss": 0.2883,
+      "epoch": 1.6576566836054347,
+      "grad_norm": 0.20319004356861115,
+      "learning_rate": 3.900393027593917e-05,
+      "loss": 0.4474,
       "step": 45995
     },
     {
-      "epoch": 1.62,
-      "learning_rate": 3.952509608974585e-05,
-      "loss": 0.2835,
+      "epoch": 1.6578368832666595,
+      "grad_norm": 0.16582396626472473,
+      "learning_rate": 3.90015128385411e-05,
+      "loss": 0.4064,
       "step": 46000
     },
     {
-      "epoch": 1.62,
-      "eval_loss": 0.2772733271121979,
-      "eval_runtime": 10.5485,
-      "eval_samples_per_second": 9.48,
-      "eval_steps_per_second": 9.48,
+      "epoch": 1.6578368832666595,
+      "eval_loss": 0.4423971474170685,
+      "eval_runtime": 3.531,
+      "eval_samples_per_second": 28.32,
+      "eval_steps_per_second": 7.08,
       "step": 46000
     },
     {
-      "epoch": 1.62,
-      "learning_rate": 3.952277741885834e-05,
-      "loss": 0.2889,
+      "epoch": 1.6580170829278842,
+      "grad_norm": 0.15383246541023254,
+      "learning_rate": 3.8999095210375124e-05,
+      "loss": 0.3594,
       "step": 46005
     },
     {
-      "epoch": 1.62,
-      "learning_rate": 3.9520458559402135e-05,
-      "loss": 0.298,
+      "epoch": 1.6581972825891087,
+      "grad_norm": 0.18127766251564026,
+      "learning_rate": 3.899667739147419e-05,
+      "loss": 0.431,
       "step": 46010
     },
     {
-      "epoch": 1.62,
-      "learning_rate": 3.951813951140735e-05,
-      "loss": 0.2769,
+      "epoch": 1.6583774822503332,
+      "grad_norm": 0.19410793483257294,
+      "learning_rate": 3.899425938187123e-05,
+      "loss": 0.3563,
       "step": 46015
     },
     {
-      "epoch": 1.62,
-      "learning_rate": 3.951582027490409e-05,
-      "loss": 0.2866,
+      "epoch": 1.658557681911558,
+      "grad_norm": 0.23459197580814362,
+      "learning_rate": 3.89918411815992e-05,
+      "loss": 0.4236,
       "step": 46020
     },
     {
-      "epoch": 1.62,
-      "learning_rate": 3.951350084992246e-05,
-      "loss": 0.3019,
+      "epoch": 1.6587378815727827,
+      "grad_norm": 0.1808062195777893,
+      "learning_rate": 3.898942279069104e-05,
+      "loss": 0.3991,
       "step": 46025
     },
     {
-      "epoch": 1.62,
-      "learning_rate": 3.95111812364926e-05,
-      "loss": 0.3141,
+      "epoch": 1.6589180812340074,
+      "grad_norm": 0.14679357409477234,
+      "learning_rate": 3.8987004209179715e-05,
+      "loss": 0.3997,
       "step": 46030
     },
     {
-      "epoch": 1.62,
-      "learning_rate": 3.950886143464461e-05,
-      "loss": 0.2843,
+      "epoch": 1.659098280895232,
+      "grad_norm": 0.18178507685661316,
+      "learning_rate": 3.898458543709815e-05,
+      "loss": 0.4321,
       "step": 46035
     },
     {
-      "epoch": 1.62,
-      "learning_rate": 3.950654144440862e-05,
-      "loss": 0.2648,
+      "epoch": 1.6592784805564564,
+      "grad_norm": 0.20688877999782562,
+      "learning_rate": 3.8982166474479323e-05,
+      "loss": 0.4243,
       "step": 46040
     },
     {
-      "epoch": 1.62,
-      "learning_rate": 3.9504221265814746e-05,
-      "loss": 0.2859,
+      "epoch": 1.6594586802176812,
+      "grad_norm": 0.21331140398979187,
+      "learning_rate": 3.8979747321356186e-05,
+      "loss": 0.4117,
       "step": 46045
     },
     {
-      "epoch": 1.62,
-      "learning_rate": 3.950190089889311e-05,
-      "loss": 0.3057,
+      "epoch": 1.659638879878906,
+      "grad_norm": 0.20621632039546967,
+      "learning_rate": 3.89773279777617e-05,
+      "loss": 0.4367,
       "step": 46050
     },
     {
-      "epoch": 1.62,
-      "learning_rate": 3.9499580343673856e-05,
-      "loss": 0.2912,
+      "epoch": 1.6598190795401304,
+      "grad_norm": 0.20351921021938324,
+      "learning_rate": 3.897490844372882e-05,
+      "loss": 0.4259,
       "step": 46055
     },
     {
-      "epoch": 1.62,
-      "learning_rate": 3.949725960018711e-05,
-      "loss": 0.2906,
+      "epoch": 1.6599992792013551,
+      "grad_norm": 0.17361485958099365,
+      "learning_rate": 3.897248871929052e-05,
+      "loss": 0.4149,
       "step": 46060
     },
     {
-      "epoch": 1.62,
-      "learning_rate": 3.9494938668463e-05,
-      "loss": 0.2849,
+      "epoch": 1.6601794788625797,
+      "grad_norm": 0.1759531944990158,
+      "learning_rate": 3.897006880447977e-05,
+      "loss": 0.3962,
       "step": 46065
     },
     {
-      "epoch": 1.62,
-      "learning_rate": 3.949261754853166e-05,
-      "loss": 0.2912,
+      "epoch": 1.6603596785238044,
+      "grad_norm": 0.17056173086166382,
+      "learning_rate": 3.896764869932953e-05,
+      "loss": 0.3861,
       "step": 46070
     },
     {
-      "epoch": 1.62,
-      "learning_rate": 3.949029624042323e-05,
-      "loss": 0.3034,
+      "epoch": 1.6605398781850291,
+      "grad_norm": 0.16253900527954102,
+      "learning_rate": 3.8965228403872784e-05,
+      "loss": 0.4117,
       "step": 46075
     },
     {
-      "epoch": 1.62,
-      "learning_rate": 3.948797474416786e-05,
-      "loss": 0.3001,
+      "epoch": 1.6607200778462536,
+      "grad_norm": 0.17955611646175385,
+      "learning_rate": 3.8962807918142507e-05,
+      "loss": 0.4292,
       "step": 46080
     },
     {
-      "epoch": 1.62,
-      "learning_rate": 3.948565305979568e-05,
-      "loss": 0.2599,
+      "epoch": 1.6609002775074782,
+      "grad_norm": 0.15855169296264648,
+      "learning_rate": 3.896038724217167e-05,
+      "loss": 0.3961,
       "step": 46085
     },
     {
-      "epoch": 1.62,
-      "learning_rate": 3.948333118733685e-05,
-      "loss": 0.3083,
+      "epoch": 1.6610804771687029,
+      "grad_norm": 0.192918598651886,
+      "learning_rate": 3.8957966375993266e-05,
+      "loss": 0.3705,
       "step": 46090
     },
     {
-      "epoch": 1.62,
-      "learning_rate": 3.94810091268215e-05,
-      "loss": 0.2711,
+      "epoch": 1.6612606768299276,
+      "grad_norm": Infinity,
+      "learning_rate": 3.895602954612325e-05,
+      "loss": 0.399,
       "step": 46095
     },
     {
-      "epoch": 1.62,
-      "learning_rate": 3.947868687827979e-05,
-      "loss": 0.2913,
+      "epoch": 1.6614408764911524,
+      "grad_norm": 0.15826788544654846,
+      "learning_rate": 3.895360833765434e-05,
+      "loss": 0.422,
       "step": 46100
     },
     {
-      "epoch": 1.62,
-      "learning_rate": 3.9476364441741873e-05,
-      "loss": 0.2639,
+      "epoch": 1.6616210761523769,
+      "grad_norm": 0.1578192263841629,
+      "learning_rate": 3.895118693907021e-05,
+      "loss": 0.4362,
       "step": 46105
     },
     {
-      "epoch": 1.62,
-      "learning_rate": 3.947404181723791e-05,
-      "loss": 0.2943,
+      "epoch": 1.6618012758136014,
+      "grad_norm": 0.18749912083148956,
+      "learning_rate": 3.8948765350403856e-05,
+      "loss": 0.4229,
       "step": 46110
     },
     {
-      "epoch": 1.62,
-      "learning_rate": 3.947171900479805e-05,
-      "loss": 0.293,
+      "epoch": 1.661981475474826,
+      "grad_norm": 0.2128848135471344,
+      "learning_rate": 3.8946343571688273e-05,
+      "loss": 0.3845,
       "step": 46115
     },
     {
-      "epoch": 1.62,
-      "learning_rate": 3.9469396004452453e-05,
-      "loss": 0.2659,
+      "epoch": 1.6621616751360508,
+      "grad_norm": 0.19334283471107483,
+      "learning_rate": 3.894392160295647e-05,
+      "loss": 0.4214,
       "step": 46120
     },
     {
-      "epoch": 1.62,
-      "learning_rate": 3.946707281623129e-05,
-      "loss": 0.271,
+      "epoch": 1.6623418747972754,
+      "grad_norm": 0.17643995583057404,
+      "learning_rate": 3.894149944424142e-05,
+      "loss": 0.3705,
       "step": 46125
     },
     {
-      "epoch": 1.62,
-      "learning_rate": 3.946474944016472e-05,
-      "loss": 0.2975,
+      "epoch": 1.6625220744584999,
+      "grad_norm": 0.17659886181354523,
+      "learning_rate": 3.8939077095576154e-05,
+      "loss": 0.3882,
       "step": 46130
     },
     {
-      "epoch": 1.62,
-      "learning_rate": 3.9462425876282904e-05,
-      "loss": 0.2955,
+      "epoch": 1.6627022741197246,
+      "grad_norm": 0.1546994149684906,
+      "learning_rate": 3.893665455699366e-05,
+      "loss": 0.3932,
       "step": 46135
     },
     {
-      "epoch": 1.62,
-      "learning_rate": 3.9460102124616027e-05,
-      "loss": 0.2788,
+      "epoch": 1.6628824737809493,
+      "grad_norm": 0.18664954602718353,
+      "learning_rate": 3.893423182852694e-05,
+      "loss": 0.43,
       "step": 46140
     },
     {
-      "epoch": 1.62,
-      "learning_rate": 3.945777818519425e-05,
-      "loss": 0.2802,
+      "epoch": 1.663062673442174,
+      "grad_norm": 0.2018563598394394,
+      "learning_rate": 3.893180891020901e-05,
+      "loss": 0.4061,
       "step": 46145
     },
     {
-      "epoch": 1.62,
-      "learning_rate": 3.9455454058047753e-05,
-      "loss": 0.2871,
+      "epoch": 1.6632428731033986,
+      "grad_norm": 0.17747381329536438,
+      "learning_rate": 3.8929385802072885e-05,
+      "loss": 0.3716,
       "step": 46150
     },
     {
-      "epoch": 1.62,
-      "learning_rate": 3.9453129743206714e-05,
-      "loss": 0.2792,
+      "epoch": 1.663423072764623,
+      "grad_norm": 0.1724482625722885,
+      "learning_rate": 3.892696250415156e-05,
+      "loss": 0.4176,
       "step": 46155
     },
     {
-      "epoch": 1.62,
-      "learning_rate": 3.9450805240701304e-05,
-      "loss": 0.301,
+      "epoch": 1.6636032724258478,
+      "grad_norm": 0.18025627732276917,
+      "learning_rate": 3.892453901647809e-05,
+      "loss": 0.4606,
       "step": 46160
     },
     {
-      "epoch": 1.62,
-      "learning_rate": 3.944848055056172e-05,
-      "loss": 0.2779,
+      "epoch": 1.6637834720870726,
+      "grad_norm": 0.18692156672477722,
+      "learning_rate": 3.8922115339085455e-05,
+      "loss": 0.3861,
       "step": 46165
     },
     {
-      "epoch": 1.62,
-      "learning_rate": 3.944615567281812e-05,
-      "loss": 0.2917,
+      "epoch": 1.663963671748297,
+      "grad_norm": 0.16012603044509888,
+      "learning_rate": 3.89196914720067e-05,
+      "loss": 0.3892,
       "step": 46170
     },
     {
-      "epoch": 1.62,
-      "learning_rate": 3.944383060750072e-05,
-      "loss": 0.2802,
+      "epoch": 1.6641438714095218,
+      "grad_norm": 0.20383629202842712,
+      "learning_rate": 3.891726741527484e-05,
+      "loss": 0.4128,
       "step": 46175
     },
     {
-      "epoch": 1.62,
-      "learning_rate": 3.944150535463971e-05,
-      "loss": 0.308,
+      "epoch": 1.6643240710707463,
+      "grad_norm": 0.2038664072751999,
+      "learning_rate": 3.891484316892291e-05,
+      "loss": 0.4766,
       "step": 46180
     },
     {
-      "epoch": 1.62,
-      "learning_rate": 3.9439179914265255e-05,
-      "loss": 0.2955,
+      "epoch": 1.664504270731971,
+      "grad_norm": 0.18376675248146057,
+      "learning_rate": 3.891241873298394e-05,
+      "loss": 0.3883,
       "step": 46185
     },
     {
-      "epoch": 1.63,
-      "learning_rate": 3.9436854286407565e-05,
-      "loss": 0.2685,
+      "epoch": 1.6646844703931958,
+      "grad_norm": 0.17446856200695038,
+      "learning_rate": 3.890999410749095e-05,
+      "loss": 0.4284,
       "step": 46190
     },
     {
-      "epoch": 1.63,
-      "learning_rate": 3.943452847109684e-05,
-      "loss": 0.277,
+      "epoch": 1.6648646700544203,
+      "grad_norm": 0.20632441341876984,
+      "learning_rate": 3.890756929247699e-05,
+      "loss": 0.4319,
       "step": 46195
     },
     {
-      "epoch": 1.63,
-      "learning_rate": 3.943220246836328e-05,
-      "loss": 0.3088,
+      "epoch": 1.6650448697156448,
+      "grad_norm": 0.1419346034526825,
+      "learning_rate": 3.890514428797508e-05,
+      "loss": 0.384,
       "step": 46200
     },
     {
-      "epoch": 1.63,
-      "learning_rate": 3.9429876278237083e-05,
-      "loss": 0.265,
+      "epoch": 1.6652250693768695,
+      "grad_norm": 0.17198680341243744,
+      "learning_rate": 3.890271909401828e-05,
+      "loss": 0.4073,
       "step": 46205
     },
     {
-      "epoch": 1.63,
-      "learning_rate": 3.9427549900748445e-05,
-      "loss": 0.2596,
+      "epoch": 1.6654052690380943,
+      "grad_norm": 0.17511259019374847,
+      "learning_rate": 3.890029371063962e-05,
+      "loss": 0.4082,
       "step": 46210
     },
     {
-      "epoch": 1.63,
-      "learning_rate": 3.942522333592759e-05,
-      "loss": 0.2739,
+      "epoch": 1.665585468699319,
+      "grad_norm": 0.16610261797904968,
+      "learning_rate": 3.8897868137872154e-05,
+      "loss": 0.4148,
       "step": 46215
     },
     {
-      "epoch": 1.63,
-      "learning_rate": 3.94228965838047e-05,
-      "loss": 0.2872,
+      "epoch": 1.6657656683605435,
+      "grad_norm": 0.15712276101112366,
+      "learning_rate": 3.889544237574893e-05,
+      "loss": 0.3822,
       "step": 46220
     },
     {
-      "epoch": 1.63,
-      "learning_rate": 3.942056964441001e-05,
-      "loss": 0.2797,
+      "epoch": 1.665945868021768,
+      "grad_norm": 0.15209558606147766,
+      "learning_rate": 3.889301642430299e-05,
+      "loss": 0.4053,
       "step": 46225
     },
     {
-      "epoch": 1.63,
-      "learning_rate": 3.941824251777373e-05,
-      "loss": 0.278,
+      "epoch": 1.6661260676829928,
+      "grad_norm": 0.1979793757200241,
+      "learning_rate": 3.889059028356738e-05,
+      "loss": 0.4097,
       "step": 46230
     },
     {
-      "epoch": 1.63,
-      "learning_rate": 3.9415915203926066e-05,
-      "loss": 0.291,
+      "epoch": 1.6663062673442175,
+      "grad_norm": 0.17566531896591187,
+      "learning_rate": 3.8888163953575174e-05,
+      "loss": 0.3878,
       "step": 46235
     },
     {
-      "epoch": 1.63,
-      "learning_rate": 3.941358770289725e-05,
-      "loss": 0.2806,
+      "epoch": 1.666486467005442,
+      "grad_norm": 0.23031200468540192,
+      "learning_rate": 3.8885737434359424e-05,
+      "loss": 0.4245,
       "step": 46240
     },
     {
-      "epoch": 1.63,
-      "learning_rate": 3.94112600147175e-05,
-      "loss": 0.2903,
+      "epoch": 1.6666666666666665,
+      "grad_norm": 0.16964511573314667,
+      "learning_rate": 3.8883310725953194e-05,
+      "loss": 0.3682,
       "step": 46245
     },
     {
-      "epoch": 1.63,
-      "learning_rate": 3.940893213941703e-05,
-      "loss": 0.2905,
+      "epoch": 1.6668468663278913,
+      "grad_norm": 0.20693713426589966,
+      "learning_rate": 3.888088382838954e-05,
+      "loss": 0.4083,
       "step": 46250
     },
     {
-      "epoch": 1.63,
-      "learning_rate": 3.9406604077026073e-05,
-      "loss": 0.2838,
+      "epoch": 1.667027065989116,
+      "grad_norm": 0.18024256825447083,
+      "learning_rate": 3.8878456741701524e-05,
+      "loss": 0.4068,
       "step": 46255
     },
     {
-      "epoch": 1.63,
-      "learning_rate": 3.9404275827574855e-05,
-      "loss": 0.2762,
+      "epoch": 1.6672072656503407,
+      "grad_norm": 0.1766664981842041,
+      "learning_rate": 3.887602946592223e-05,
+      "loss": 0.3993,
       "step": 46260
     },
     {
-      "epoch": 1.63,
-      "learning_rate": 3.9401947391093606e-05,
-      "loss": 0.2821,
+      "epoch": 1.6673874653115652,
+      "grad_norm": 0.18678483366966248,
+      "learning_rate": 3.8873602001084716e-05,
+      "loss": 0.4381,
       "step": 46265
     },
     {
-      "epoch": 1.63,
-      "learning_rate": 3.939961876761257e-05,
-      "loss": 0.3049,
+      "epoch": 1.6675676649727897,
+      "grad_norm": 0.13964805006980896,
+      "learning_rate": 3.887117434722206e-05,
+      "loss": 0.4004,
       "step": 46270
     },
     {
-      "epoch": 1.63,
-      "learning_rate": 3.939728995716196e-05,
-      "loss": 0.2822,
+      "epoch": 1.6677478646340145,
+      "grad_norm": 0.15428251028060913,
+      "learning_rate": 3.886874650436735e-05,
+      "loss": 0.3547,
       "step": 46275
     },
     {
-      "epoch": 1.63,
-      "learning_rate": 3.9394960959772045e-05,
-      "loss": 0.2667,
+      "epoch": 1.6679280642952392,
+      "grad_norm": 0.1429775059223175,
+      "learning_rate": 3.8866318472553644e-05,
+      "loss": 0.394,
       "step": 46280
     },
     {
-      "epoch": 1.63,
-      "learning_rate": 3.939263177547304e-05,
-      "loss": 0.2875,
+      "epoch": 1.6681082639564637,
+      "grad_norm": 0.16724984347820282,
+      "learning_rate": 3.886389025181404e-05,
+      "loss": 0.3977,
       "step": 46285
     },
     {
-      "epoch": 1.63,
-      "learning_rate": 3.9390302404295195e-05,
-      "loss": 0.2996,
+      "epoch": 1.6682884636176882,
+      "grad_norm": 0.1835627257823944,
+      "learning_rate": 3.886146184218161e-05,
+      "loss": 0.4373,
       "step": 46290
     },
     {
-      "epoch": 1.63,
-      "learning_rate": 3.938797284626876e-05,
-      "loss": 0.2871,
+      "epoch": 1.668468663278913,
+      "grad_norm": 0.17956386506557465,
+      "learning_rate": 3.8859033243689446e-05,
+      "loss": 0.4227,
       "step": 46295
     },
     {
-      "epoch": 1.63,
-      "learning_rate": 3.9385643101423985e-05,
-      "loss": 0.2833,
+      "epoch": 1.6686488629401377,
+      "grad_norm": 0.19547246396541595,
+      "learning_rate": 3.885660445637064e-05,
+      "loss": 0.383,
       "step": 46300
     },
     {
-      "epoch": 1.63,
-      "learning_rate": 3.9383313169791104e-05,
-      "loss": 0.2626,
+      "epoch": 1.6688290626013624,
+      "grad_norm": 0.2311650663614273,
+      "learning_rate": 3.885417548025828e-05,
+      "loss": 0.3988,
       "step": 46305
     },
     {
-      "epoch": 1.63,
-      "learning_rate": 3.9380983051400387e-05,
-      "loss": 0.2652,
+      "epoch": 1.669009262262587,
+      "grad_norm": 0.15516655147075653,
+      "learning_rate": 3.885174631538546e-05,
+      "loss": 0.4264,
       "step": 46310
     },
     {
-      "epoch": 1.63,
-      "learning_rate": 3.937865274628208e-05,
-      "loss": 0.2765,
+      "epoch": 1.6691894619238115,
+      "grad_norm": 0.19897285103797913,
+      "learning_rate": 3.8849316961785276e-05,
+      "loss": 0.4051,
       "step": 46315
     },
     {
-      "epoch": 1.63,
-      "learning_rate": 3.937632225446645e-05,
-      "loss": 0.2799,
+      "epoch": 1.6693696615850362,
+      "grad_norm": 0.22610118985176086,
+      "learning_rate": 3.884688741949084e-05,
+      "loss": 0.4286,
       "step": 46320
     },
     {
-      "epoch": 1.63,
-      "learning_rate": 3.937399157598374e-05,
-      "loss": 0.2789,
+      "epoch": 1.669549861246261,
+      "grad_norm": 0.16517026722431183,
+      "learning_rate": 3.884445768853524e-05,
+      "loss": 0.4167,
       "step": 46325
     },
     {
-      "epoch": 1.63,
-      "learning_rate": 3.937166071086422e-05,
-      "loss": 0.2795,
+      "epoch": 1.6697300609074857,
+      "grad_norm": 0.21267640590667725,
+      "learning_rate": 3.884202776895158e-05,
+      "loss": 0.3828,
       "step": 46330
     },
     {
-      "epoch": 1.63,
-      "learning_rate": 3.9369329659138165e-05,
-      "loss": 0.2931,
+      "epoch": 1.6699102605687102,
+      "grad_norm": 0.20657142996788025,
+      "learning_rate": 3.883959766077297e-05,
+      "loss": 0.4069,
       "step": 46335
     },
     {
-      "epoch": 1.63,
-      "learning_rate": 3.936699842083583e-05,
-      "loss": 0.2742,
+      "epoch": 1.6700904602299347,
+      "grad_norm": 0.16115951538085938,
+      "learning_rate": 3.8837167364032526e-05,
+      "loss": 0.4248,
       "step": 46340
     },
     {
-      "epoch": 1.63,
-      "learning_rate": 3.936466699598749e-05,
-      "loss": 0.2872,
+      "epoch": 1.6702706598911594,
+      "grad_norm": 0.2205209583044052,
+      "learning_rate": 3.883473687876336e-05,
+      "loss": 0.4162,
       "step": 46345
     },
     {
-      "epoch": 1.63,
-      "learning_rate": 3.9362335384623414e-05,
-      "loss": 0.2813,
+      "epoch": 1.6704508595523841,
+      "grad_norm": 0.18244054913520813,
+      "learning_rate": 3.883230620499857e-05,
+      "loss": 0.4127,
       "step": 46350
     },
     {
-      "epoch": 1.63,
-      "learning_rate": 3.9360003586773886e-05,
-      "loss": 0.2804,
+      "epoch": 1.6706310592136087,
+      "grad_norm": 0.14126257598400116,
+      "learning_rate": 3.8829875342771287e-05,
+      "loss": 0.4314,
       "step": 46355
     },
     {
-      "epoch": 1.63,
-      "learning_rate": 3.935767160246917e-05,
-      "loss": 0.2565,
+      "epoch": 1.6708112588748332,
+      "grad_norm": 0.1783919334411621,
+      "learning_rate": 3.8827444292114634e-05,
+      "loss": 0.4063,
       "step": 46360
     },
     {
-      "epoch": 1.63,
-      "learning_rate": 3.935533943173954e-05,
-      "loss": 0.2673,
+      "epoch": 1.670991458536058,
+      "grad_norm": 0.1652853786945343,
+      "learning_rate": 3.882501305306174e-05,
+      "loss": 0.4104,
       "step": 46365
     },
     {
-      "epoch": 1.63,
-      "learning_rate": 3.93530070746153e-05,
-      "loss": 0.2911,
+      "epoch": 1.6711716581972826,
+      "grad_norm": 0.16857433319091797,
+      "learning_rate": 3.8822581625645706e-05,
+      "loss": 0.3828,
       "step": 46370
     },
     {
-      "epoch": 1.63,
-      "learning_rate": 3.935067453112673e-05,
-      "loss": 0.2806,
+      "epoch": 1.6713518578585074,
+      "grad_norm": 0.2388414889574051,
+      "learning_rate": 3.882015000989968e-05,
+      "loss": 0.4411,
       "step": 46375
     },
     {
-      "epoch": 1.63,
-      "learning_rate": 3.934834180130409e-05,
-      "loss": 0.2451,
+      "epoch": 1.6715320575197319,
+      "grad_norm": 0.16864396631717682,
+      "learning_rate": 3.881771820585678e-05,
+      "loss": 0.4467,
       "step": 46380
     },
     {
-      "epoch": 1.63,
-      "learning_rate": 3.93460088851777e-05,
-      "loss": 0.2758,
+      "epoch": 1.6717122571809564,
+      "grad_norm": 0.1579165756702423,
+      "learning_rate": 3.881528621355015e-05,
+      "loss": 0.4215,
       "step": 46385
     },
     {
-      "epoch": 1.63,
-      "learning_rate": 3.934367578277783e-05,
-      "loss": 0.2824,
+      "epoch": 1.6718924568421811,
+      "grad_norm": 0.17890191078186035,
+      "learning_rate": 3.8812854033012916e-05,
+      "loss": 0.4252,
       "step": 46390
     },
     {
-      "epoch": 1.63,
-      "learning_rate": 3.934134249413479e-05,
-      "loss": 0.2827,
+      "epoch": 1.6720726565034059,
+      "grad_norm": 0.1696099191904068,
+      "learning_rate": 3.881042166427823e-05,
+      "loss": 0.3783,
       "step": 46395
     },
     {
-      "epoch": 1.63,
-      "learning_rate": 3.933900901927887e-05,
-      "loss": 0.2792,
+      "epoch": 1.6722528561646304,
+      "grad_norm": 0.13848702609539032,
+      "learning_rate": 3.880798910737921e-05,
+      "loss": 0.3871,
       "step": 46400
     },
     {
-      "epoch": 1.63,
-      "learning_rate": 3.933667535824036e-05,
-      "loss": 0.2811,
+      "epoch": 1.6724330558258549,
+      "grad_norm": 0.17733387649059296,
+      "learning_rate": 3.880555636234902e-05,
+      "loss": 0.4035,
       "step": 46405
     },
     {
-      "epoch": 1.63,
-      "learning_rate": 3.9334341511049576e-05,
-      "loss": 0.283,
+      "epoch": 1.6726132554870796,
+      "grad_norm": 0.21084022521972656,
+      "learning_rate": 3.880312342922079e-05,
+      "loss": 0.4408,
       "step": 46410
     },
     {
-      "epoch": 1.63,
-      "learning_rate": 3.9332007477736816e-05,
-      "loss": 0.2761,
+      "epoch": 1.6727934551483044,
+      "grad_norm": 0.21797522902488708,
+      "learning_rate": 3.880069030802768e-05,
+      "loss": 0.4113,
       "step": 46415
     },
     {
-      "epoch": 1.63,
-      "learning_rate": 3.932967325833238e-05,
-      "loss": 0.2941,
+      "epoch": 1.672973654809529,
+      "grad_norm": 0.19593912363052368,
+      "learning_rate": 3.879825699880284e-05,
+      "loss": 0.4092,
       "step": 46420
     },
     {
-      "epoch": 1.63,
-      "learning_rate": 3.932733885286658e-05,
-      "loss": 0.2831,
+      "epoch": 1.6731538544707536,
+      "grad_norm": 0.2192002683877945,
+      "learning_rate": 3.879582350157942e-05,
+      "loss": 0.4616,
       "step": 46425
     },
     {
-      "epoch": 1.63,
-      "learning_rate": 3.932500426136973e-05,
-      "loss": 0.2813,
+      "epoch": 1.673334054131978,
+      "grad_norm": 0.1574559360742569,
+      "learning_rate": 3.879338981639057e-05,
+      "loss": 0.3955,
       "step": 46430
     },
     {
-      "epoch": 1.63,
-      "learning_rate": 3.932266948387213e-05,
-      "loss": 0.2816,
+      "epoch": 1.6735142537932028,
+      "grad_norm": 0.20337316393852234,
+      "learning_rate": 3.8790955943269455e-05,
+      "loss": 0.4303,
       "step": 46435
     },
     {
-      "epoch": 1.63,
-      "learning_rate": 3.932033452040411e-05,
-      "loss": 0.3044,
+      "epoch": 1.6736944534544276,
+      "grad_norm": 0.15265347063541412,
+      "learning_rate": 3.878852188224924e-05,
+      "loss": 0.4054,
       "step": 46440
     },
     {
-      "epoch": 1.63,
-      "learning_rate": 3.931799937099599e-05,
-      "loss": 0.2882,
+      "epoch": 1.673874653115652,
+      "grad_norm": 0.1718795895576477,
+      "learning_rate": 3.8786087633363075e-05,
+      "loss": 0.3766,
       "step": 46445
     },
     {
-      "epoch": 1.63,
-      "learning_rate": 3.931566403567808e-05,
-      "loss": 0.2827,
+      "epoch": 1.6740548527768768,
+      "grad_norm": 0.18566109240055084,
+      "learning_rate": 3.8783653196644144e-05,
+      "loss": 0.3856,
       "step": 46450
     },
     {
-      "epoch": 1.63,
-      "learning_rate": 3.931332851448071e-05,
-      "loss": 0.2983,
+      "epoch": 1.6742350524381013,
+      "grad_norm": 0.18366526067256927,
+      "learning_rate": 3.87812185721256e-05,
+      "loss": 0.4811,
       "step": 46455
     },
     {
-      "epoch": 1.63,
-      "learning_rate": 3.931099280743419e-05,
-      "loss": 0.2995,
+      "epoch": 1.674415252099326,
+      "grad_norm": 0.1583663374185562,
+      "learning_rate": 3.8778783759840625e-05,
+      "loss": 0.437,
       "step": 46460
     },
     {
-      "epoch": 1.63,
-      "learning_rate": 3.9308656914568875e-05,
-      "loss": 0.2944,
+      "epoch": 1.6745954517605508,
+      "grad_norm": 0.16400283575057983,
+      "learning_rate": 3.877634875982239e-05,
+      "loss": 0.397,
       "step": 46465
     },
     {
-      "epoch": 1.63,
-      "learning_rate": 3.930632083591508e-05,
-      "loss": 0.2639,
+      "epoch": 1.6747756514217753,
+      "grad_norm": 0.18252485990524292,
+      "learning_rate": 3.877391357210407e-05,
+      "loss": 0.431,
       "step": 46470
     },
     {
-      "epoch": 1.64,
-      "learning_rate": 3.930398457150313e-05,
-      "loss": 0.2724,
+      "epoch": 1.6749558510829998,
+      "grad_norm": 0.22396881878376007,
+      "learning_rate": 3.877147819671884e-05,
+      "loss": 0.424,
       "step": 46475
     },
     {
-      "epoch": 1.64,
-      "learning_rate": 3.930164812136336e-05,
-      "loss": 0.2594,
+      "epoch": 1.6751360507442246,
+      "grad_norm": 0.16650210320949554,
+      "learning_rate": 3.876904263369989e-05,
+      "loss": 0.399,
       "step": 46480
     },
     {
-      "epoch": 1.64,
-      "learning_rate": 3.9299311485526124e-05,
-      "loss": 0.2702,
+      "epoch": 1.6753162504054493,
+      "grad_norm": 0.17203296720981598,
+      "learning_rate": 3.87666068830804e-05,
+      "loss": 0.4241,
       "step": 46485
     },
     {
-      "epoch": 1.64,
-      "learning_rate": 3.929697466402175e-05,
-      "loss": 0.2883,
+      "epoch": 1.675496450066674,
+      "grad_norm": 0.21028588712215424,
+      "learning_rate": 3.876417094489355e-05,
+      "loss": 0.3976,
       "step": 46490
     },
     {
-      "epoch": 1.64,
-      "learning_rate": 3.9294637656880586e-05,
-      "loss": 0.2708,
+      "epoch": 1.6756766497278985,
+      "grad_norm": 0.20283226668834686,
+      "learning_rate": 3.876173481917255e-05,
+      "loss": 0.4196,
       "step": 46495
     },
     {
-      "epoch": 1.64,
-      "learning_rate": 3.929230046413297e-05,
-      "loss": 0.2786,
+      "epoch": 1.675856849389123,
+      "grad_norm": 0.1598893254995346,
+      "learning_rate": 3.875929850595056e-05,
+      "loss": 0.3846,
       "step": 46500
     },
     {
-      "epoch": 1.64,
-      "eval_loss": 0.2775975167751312,
-      "eval_runtime": 10.5364,
-      "eval_samples_per_second": 9.491,
-      "eval_steps_per_second": 9.491,
+      "epoch": 1.675856849389123,
+      "eval_loss": 0.44327858090400696,
+      "eval_runtime": 3.5351,
+      "eval_samples_per_second": 28.288,
+      "eval_steps_per_second": 7.072,
       "step": 46500
     },
     {
-      "epoch": 1.64,
-      "learning_rate": 3.9289963085809246e-05,
-      "loss": 0.2771,
+      "epoch": 1.6760370490503478,
+      "grad_norm": 0.18217889964580536,
+      "learning_rate": 3.8756862005260804e-05,
+      "loss": 0.4492,
       "step": 46505
     },
     {
-      "epoch": 1.64,
-      "learning_rate": 3.928762552193977e-05,
-      "loss": 0.2936,
+      "epoch": 1.6762172487115725,
+      "grad_norm": 0.19530221819877625,
+      "learning_rate": 3.8754425317136465e-05,
+      "loss": 0.4021,
       "step": 46510
     },
     {
-      "epoch": 1.64,
-      "learning_rate": 3.9285287772554905e-05,
-      "loss": 0.2759,
+      "epoch": 1.676397448372797,
+      "grad_norm": 0.21097613871097565,
+      "learning_rate": 3.875198844161074e-05,
+      "loss": 0.4219,
       "step": 46515
     },
     {
-      "epoch": 1.64,
-      "learning_rate": 3.928294983768498e-05,
-      "loss": 0.2953,
+      "epoch": 1.6765776480340215,
+      "grad_norm": 0.14871633052825928,
+      "learning_rate": 3.874955137871684e-05,
+      "loss": 0.3881,
       "step": 46520
     },
     {
-      "epoch": 1.64,
-      "learning_rate": 3.928061171736036e-05,
-      "loss": 0.297,
+      "epoch": 1.6767578476952463,
+      "grad_norm": 0.1719123125076294,
+      "learning_rate": 3.874711412848796e-05,
+      "loss": 0.4257,
       "step": 46525
     },
     {
-      "epoch": 1.64,
-      "learning_rate": 3.9278273411611425e-05,
-      "loss": 0.2788,
+      "epoch": 1.676938047356471,
+      "grad_norm": 0.2002047300338745,
+      "learning_rate": 3.874467669095731e-05,
+      "loss": 0.3852,
       "step": 46530
     },
     {
-      "epoch": 1.64,
-      "learning_rate": 3.92759349204685e-05,
-      "loss": 0.2689,
+      "epoch": 1.6771182470176957,
+      "grad_norm": 0.18990401923656464,
+      "learning_rate": 3.87422390661581e-05,
+      "loss": 0.4062,
       "step": 46535
     },
     {
-      "epoch": 1.64,
-      "learning_rate": 3.927359624396199e-05,
-      "loss": 0.2782,
+      "epoch": 1.6772984466789203,
+      "grad_norm": 0.2010718584060669,
+      "learning_rate": 3.873980125412355e-05,
+      "loss": 0.4468,
       "step": 46540
     },
     {
-      "epoch": 1.64,
-      "learning_rate": 3.927125738212222e-05,
-      "loss": 0.2997,
+      "epoch": 1.6774786463401448,
+      "grad_norm": 0.20043112337589264,
+      "learning_rate": 3.873736325488687e-05,
+      "loss": 0.4265,
       "step": 46545
     },
     {
-      "epoch": 1.64,
-      "learning_rate": 3.9268918334979584e-05,
-      "loss": 0.2866,
+      "epoch": 1.6776588460013695,
+      "grad_norm": 0.208219975233078,
+      "learning_rate": 3.873492506848127e-05,
+      "loss": 0.4229,
       "step": 46550
     },
     {
-      "epoch": 1.64,
-      "learning_rate": 3.926657910256445e-05,
-      "loss": 0.2718,
+      "epoch": 1.6778390456625942,
+      "grad_norm": 0.1497962474822998,
+      "learning_rate": 3.873248669493997e-05,
+      "loss": 0.3987,
       "step": 46555
     },
     {
-      "epoch": 1.64,
-      "learning_rate": 3.926423968490719e-05,
-      "loss": 0.273,
+      "epoch": 1.6780192453238187,
+      "grad_norm": 0.1851639449596405,
+      "learning_rate": 3.87300481342962e-05,
+      "loss": 0.4378,
       "step": 46560
     },
     {
-      "epoch": 1.64,
-      "learning_rate": 3.9261900082038174e-05,
-      "loss": 0.2883,
+      "epoch": 1.6781994449850435,
+      "grad_norm": 0.18497171998023987,
+      "learning_rate": 3.8727609386583184e-05,
+      "loss": 0.3855,
       "step": 46565
     },
     {
-      "epoch": 1.64,
-      "learning_rate": 3.9259560293987785e-05,
-      "loss": 0.2827,
+      "epoch": 1.678379644646268,
+      "grad_norm": 0.19032762944698334,
+      "learning_rate": 3.872517045183416e-05,
+      "loss": 0.4086,
       "step": 46570
     },
     {
-      "epoch": 1.64,
-      "learning_rate": 3.92572203207864e-05,
-      "loss": 0.2716,
+      "epoch": 1.6785598443074927,
+      "grad_norm": 0.1807955503463745,
+      "learning_rate": 3.872273133008233e-05,
+      "loss": 0.4104,
       "step": 46575
     },
     {
-      "epoch": 1.64,
-      "learning_rate": 3.925488016246441e-05,
-      "loss": 0.3,
+      "epoch": 1.6787400439687175,
+      "grad_norm": 0.18649344146251678,
+      "learning_rate": 3.872029202136095e-05,
+      "loss": 0.4241,
       "step": 46580
     },
     {
-      "epoch": 1.64,
-      "learning_rate": 3.92525398190522e-05,
-      "loss": 0.3001,
+      "epoch": 1.678920243629942,
+      "grad_norm": 0.17632882297039032,
+      "learning_rate": 3.8717852525703246e-05,
+      "loss": 0.4656,
       "step": 46585
     },
     {
-      "epoch": 1.64,
-      "learning_rate": 3.9250199290580144e-05,
-      "loss": 0.2725,
+      "epoch": 1.6791004432911665,
+      "grad_norm": 0.1870754510164261,
+      "learning_rate": 3.871541284314245e-05,
+      "loss": 0.3766,
       "step": 46590
     },
     {
-      "epoch": 1.64,
-      "learning_rate": 3.924785857707865e-05,
-      "loss": 0.2713,
+      "epoch": 1.6792806429523912,
+      "grad_norm": 0.18941162526607513,
+      "learning_rate": 3.871297297371182e-05,
+      "loss": 0.4175,
       "step": 46595
     },
     {
-      "epoch": 1.64,
-      "learning_rate": 3.924551767857809e-05,
-      "loss": 0.2979,
+      "epoch": 1.679460842613616,
+      "grad_norm": 0.17592564225196838,
+      "learning_rate": 3.871053291744459e-05,
+      "loss": 0.3781,
       "step": 46600
     },
     {
-      "epoch": 1.64,
-      "learning_rate": 3.924317659510888e-05,
-      "loss": 0.2895,
+      "epoch": 1.6796410422748407,
+      "grad_norm": 0.17591671645641327,
+      "learning_rate": 3.870809267437398e-05,
+      "loss": 0.3759,
       "step": 46605
     },
     {
-      "epoch": 1.64,
-      "learning_rate": 3.924083532670141e-05,
-      "loss": 0.285,
+      "epoch": 1.6798212419360652,
+      "grad_norm": 0.17050381004810333,
+      "learning_rate": 3.870565224453329e-05,
+      "loss": 0.3951,
       "step": 46610
     },
     {
-      "epoch": 1.64,
-      "learning_rate": 3.9238493873386075e-05,
-      "loss": 0.2717,
+      "epoch": 1.6800014415972897,
+      "grad_norm": 0.20477713644504547,
+      "learning_rate": 3.870321162795573e-05,
+      "loss": 0.4295,
       "step": 46615
     },
     {
-      "epoch": 1.64,
-      "learning_rate": 3.923615223519328e-05,
-      "loss": 0.2959,
+      "epoch": 1.6801816412585144,
+      "grad_norm": 0.20674197375774384,
+      "learning_rate": 3.870077082467456e-05,
+      "loss": 0.4307,
       "step": 46620
     },
     {
-      "epoch": 1.64,
-      "learning_rate": 3.923381041215343e-05,
-      "loss": 0.2938,
+      "epoch": 1.6803618409197392,
+      "grad_norm": 0.1541663408279419,
+      "learning_rate": 3.8698329834723046e-05,
+      "loss": 0.4447,
       "step": 46625
     },
     {
-      "epoch": 1.64,
-      "learning_rate": 3.923146840429693e-05,
-      "loss": 0.2913,
+      "epoch": 1.6805420405809637,
+      "grad_norm": 0.1866021603345871,
+      "learning_rate": 3.8695888658134446e-05,
+      "loss": 0.3986,
       "step": 46630
     },
     {
-      "epoch": 1.64,
-      "learning_rate": 3.92291262116542e-05,
-      "loss": 0.2844,
+      "epoch": 1.6807222402421882,
+      "grad_norm": 0.17100781202316284,
+      "learning_rate": 3.8693447294942e-05,
+      "loss": 0.4129,
       "step": 46635
     },
     {
-      "epoch": 1.64,
-      "learning_rate": 3.9226783834255634e-05,
-      "loss": 0.2926,
+      "epoch": 1.680902439903413,
+      "grad_norm": 0.16829904913902283,
+      "learning_rate": 3.8691005745179e-05,
+      "loss": 0.436,
       "step": 46640
     },
     {
-      "epoch": 1.64,
-      "learning_rate": 3.9224441272131664e-05,
-      "loss": 0.2879,
+      "epoch": 1.6810826395646377,
+      "grad_norm": 0.17675045132637024,
+      "learning_rate": 3.868856400887868e-05,
+      "loss": 0.4185,
       "step": 46645
     },
     {
-      "epoch": 1.64,
-      "learning_rate": 3.922209852531269e-05,
-      "loss": 0.2757,
+      "epoch": 1.6812628392258624,
+      "grad_norm": 0.1742866337299347,
+      "learning_rate": 3.868612208607434e-05,
+      "loss": 0.4504,
       "step": 46650
     },
     {
-      "epoch": 1.64,
-      "learning_rate": 3.921975559382914e-05,
-      "loss": 0.2858,
+      "epoch": 1.681443038887087,
+      "grad_norm": 0.21476224064826965,
+      "learning_rate": 3.8683679976799235e-05,
+      "loss": 0.4158,
       "step": 46655
     },
     {
-      "epoch": 1.64,
-      "learning_rate": 3.921741247771144e-05,
-      "loss": 0.2691,
+      "epoch": 1.6816232385483114,
+      "grad_norm": 0.19867900013923645,
+      "learning_rate": 3.868123768108664e-05,
+      "loss": 0.3975,
       "step": 46660
     },
     {
-      "epoch": 1.64,
-      "learning_rate": 3.9215069176990006e-05,
-      "loss": 0.2943,
+      "epoch": 1.6818034382095362,
+      "grad_norm": 0.15411974489688873,
+      "learning_rate": 3.867879519896983e-05,
+      "loss": 0.4003,
       "step": 46665
     },
     {
-      "epoch": 1.64,
-      "learning_rate": 3.921272569169527e-05,
-      "loss": 0.2809,
+      "epoch": 1.6819836378707609,
+      "grad_norm": 0.20566970109939575,
+      "learning_rate": 3.8676352530482074e-05,
+      "loss": 0.4074,
       "step": 46670
     },
     {
-      "epoch": 1.64,
-      "learning_rate": 3.9210382021857655e-05,
-      "loss": 0.2826,
+      "epoch": 1.6821638375319854,
+      "grad_norm": 0.17975671589374542,
+      "learning_rate": 3.8673909675656675e-05,
+      "loss": 0.4181,
       "step": 46675
     },
     {
-      "epoch": 1.64,
-      "learning_rate": 3.920803816750759e-05,
-      "loss": 0.2968,
+      "epoch": 1.6823440371932101,
+      "grad_norm": 0.16022039949893951,
+      "learning_rate": 3.86714666345269e-05,
+      "loss": 0.4104,
       "step": 46680
     },
     {
-      "epoch": 1.64,
-      "learning_rate": 3.920569412867552e-05,
-      "loss": 0.2786,
+      "epoch": 1.6825242368544346,
+      "grad_norm": 0.22043001651763916,
+      "learning_rate": 3.8669023407126035e-05,
+      "loss": 0.4131,
       "step": 46685
     },
     {
-      "epoch": 1.64,
-      "learning_rate": 3.9203349905391876e-05,
-      "loss": 0.3007,
+      "epoch": 1.6827044365156594,
+      "grad_norm": 0.14950603246688843,
+      "learning_rate": 3.866657999348737e-05,
+      "loss": 0.3596,
       "step": 46690
     },
     {
-      "epoch": 1.64,
-      "learning_rate": 3.920100549768708e-05,
-      "loss": 0.3029,
+      "epoch": 1.682884636176884,
+      "grad_norm": 0.1892220824956894,
+      "learning_rate": 3.866413639364421e-05,
+      "loss": 0.3928,
       "step": 46695
     },
     {
-      "epoch": 1.64,
-      "learning_rate": 3.91986609055916e-05,
-      "loss": 0.2705,
+      "epoch": 1.6830648358381086,
+      "grad_norm": 0.15742996335029602,
+      "learning_rate": 3.8661692607629826e-05,
+      "loss": 0.41,
       "step": 46700
     },
     {
-      "epoch": 1.64,
-      "learning_rate": 3.919631612913586e-05,
-      "loss": 0.3101,
+      "epoch": 1.6832450354993331,
+      "grad_norm": 0.18481925129890442,
+      "learning_rate": 3.865924863547753e-05,
+      "loss": 0.4061,
       "step": 46705
     },
     {
-      "epoch": 1.64,
-      "learning_rate": 3.919397116835031e-05,
-      "loss": 0.2752,
+      "epoch": 1.6834252351605579,
+      "grad_norm": 0.2112767994403839,
+      "learning_rate": 3.865680447722062e-05,
+      "loss": 0.4467,
       "step": 46710
     },
     {
-      "epoch": 1.64,
-      "learning_rate": 3.91916260232654e-05,
-      "loss": 0.3007,
+      "epoch": 1.6836054348217826,
+      "grad_norm": 0.19676971435546875,
+      "learning_rate": 3.865436013289239e-05,
+      "loss": 0.4542,
       "step": 46715
     },
     {
-      "epoch": 1.64,
-      "learning_rate": 3.9189280693911576e-05,
-      "loss": 0.268,
+      "epoch": 1.6837856344830073,
+      "grad_norm": 0.18890570104122162,
+      "learning_rate": 3.865191560252614e-05,
+      "loss": 0.4311,
       "step": 46720
     },
     {
-      "epoch": 1.64,
-      "learning_rate": 3.918693518031929e-05,
-      "loss": 0.2829,
+      "epoch": 1.6839658341442318,
+      "grad_norm": 0.17850306630134583,
+      "learning_rate": 3.864947088615519e-05,
+      "loss": 0.424,
       "step": 46725
     },
     {
-      "epoch": 1.64,
-      "learning_rate": 3.918458948251902e-05,
-      "loss": 0.2913,
+      "epoch": 1.6841460338054564,
+      "grad_norm": 0.13823658227920532,
+      "learning_rate": 3.8647025983812844e-05,
+      "loss": 0.4182,
       "step": 46730
     },
     {
-      "epoch": 1.64,
-      "learning_rate": 3.9182243600541186e-05,
-      "loss": 0.2815,
+      "epoch": 1.684326233466681,
+      "grad_norm": 0.17909295856952667,
+      "learning_rate": 3.864458089553241e-05,
+      "loss": 0.4016,
       "step": 46735
     },
     {
-      "epoch": 1.64,
-      "learning_rate": 3.917989753441627e-05,
-      "loss": 0.2865,
+      "epoch": 1.6845064331279058,
+      "grad_norm": 0.20117956399917603,
+      "learning_rate": 3.8642135621347195e-05,
+      "loss": 0.3977,
       "step": 46740
     },
     {
-      "epoch": 1.64,
-      "learning_rate": 3.9177551284174724e-05,
-      "loss": 0.2695,
+      "epoch": 1.6846866327891303,
+      "grad_norm": 0.19415679574012756,
+      "learning_rate": 3.863969016129053e-05,
+      "loss": 0.3985,
       "step": 46745
     },
     {
-      "epoch": 1.64,
-      "learning_rate": 3.917520484984702e-05,
-      "loss": 0.2711,
+      "epoch": 1.6848668324503548,
+      "grad_norm": 0.2219560146331787,
+      "learning_rate": 3.8637244515395734e-05,
+      "loss": 0.4205,
       "step": 46750
     },
     {
-      "epoch": 1.64,
-      "learning_rate": 3.9173327569863296e-05,
-      "loss": 0.2899,
+      "epoch": 1.6850470321115796,
+      "grad_norm": 0.1750744730234146,
+      "learning_rate": 3.8634798683696114e-05,
+      "loss": 0.428,
       "step": 46755
     },
     {
-      "epoch": 1.65,
-      "learning_rate": 3.917098080425729e-05,
-      "loss": 0.2844,
+      "epoch": 1.6852272317728043,
+      "grad_norm": 0.19463662803173065,
+      "learning_rate": 3.8632352666225005e-05,
+      "loss": 0.4025,
       "step": 46760
     },
     {
-      "epoch": 1.65,
-      "learning_rate": 3.916863385465043e-05,
-      "loss": 0.2959,
+      "epoch": 1.685407431434029,
+      "grad_norm": 0.18877053260803223,
+      "learning_rate": 3.862990646301572e-05,
+      "loss": 0.4338,
       "step": 46765
     },
     {
-      "epoch": 1.65,
-      "learning_rate": 3.91662867210732e-05,
-      "loss": 0.276,
+      "epoch": 1.6855876310952536,
+      "grad_norm": 0.17238645255565643,
+      "learning_rate": 3.8627460074101606e-05,
+      "loss": 0.4193,
       "step": 46770
     },
     {
-      "epoch": 1.65,
-      "learning_rate": 3.9163939403556074e-05,
-      "loss": 0.2636,
+      "epoch": 1.685767830756478,
+      "grad_norm": 0.20663443207740784,
+      "learning_rate": 3.862501349951599e-05,
+      "loss": 0.4149,
       "step": 46775
     },
     {
-      "epoch": 1.65,
-      "learning_rate": 3.916159190212952e-05,
-      "loss": 0.2678,
+      "epoch": 1.6859480304177028,
+      "grad_norm": 0.17570826411247253,
+      "learning_rate": 3.86225667392922e-05,
+      "loss": 0.4042,
       "step": 46780
     },
     {
-      "epoch": 1.65,
-      "learning_rate": 3.915924421682404e-05,
-      "loss": 0.2807,
+      "epoch": 1.6861282300789275,
+      "grad_norm": 0.1705494225025177,
+      "learning_rate": 3.8620119793463573e-05,
+      "loss": 0.432,
       "step": 46785
     },
     {
-      "epoch": 1.65,
-      "learning_rate": 3.91568963476701e-05,
-      "loss": 0.299,
+      "epoch": 1.686308429740152,
+      "grad_norm": 0.18460239470005035,
+      "learning_rate": 3.861767266206345e-05,
+      "loss": 0.3972,
       "step": 46790
     },
     {
-      "epoch": 1.65,
-      "learning_rate": 3.915454829469819e-05,
-      "loss": 0.2929,
+      "epoch": 1.6864886294013766,
+      "grad_norm": 0.1540331393480301,
+      "learning_rate": 3.861522534512518e-05,
+      "loss": 0.4156,
       "step": 46795
     },
     {
-      "epoch": 1.65,
-      "learning_rate": 3.9152200057938796e-05,
-      "loss": 0.301,
+      "epoch": 1.6866688290626013,
+      "grad_norm": 0.2118932008743286,
+      "learning_rate": 3.861277784268209e-05,
+      "loss": 0.4159,
       "step": 46800
     },
     {
-      "epoch": 1.65,
-      "learning_rate": 3.9149851637422415e-05,
-      "loss": 0.2898,
+      "epoch": 1.686849028723826,
+      "grad_norm": 0.18069863319396973,
+      "learning_rate": 3.861033015476755e-05,
+      "loss": 0.4068,
       "step": 46805
     },
     {
-      "epoch": 1.65,
-      "learning_rate": 3.914750303317953e-05,
-      "loss": 0.3064,
+      "epoch": 1.6870292283850508,
+      "grad_norm": 0.2565334439277649,
+      "learning_rate": 3.860788228141489e-05,
+      "loss": 0.4314,
       "step": 46810
     },
     {
-      "epoch": 1.65,
-      "learning_rate": 3.914515424524065e-05,
-      "loss": 0.2903,
+      "epoch": 1.6872094280462753,
+      "grad_norm": 0.21841110289096832,
+      "learning_rate": 3.8605434222657465e-05,
+      "loss": 0.4269,
       "step": 46815
     },
     {
-      "epoch": 1.65,
-      "learning_rate": 3.9142805273636265e-05,
-      "loss": 0.2944,
+      "epoch": 1.6873896277074998,
+      "grad_norm": 0.1737854778766632,
+      "learning_rate": 3.860298597852864e-05,
+      "loss": 0.426,
       "step": 46820
     },
     {
-      "epoch": 1.65,
-      "learning_rate": 3.9140456118396865e-05,
-      "loss": 0.2811,
+      "epoch": 1.6875698273687245,
+      "grad_norm": 0.17643529176712036,
+      "learning_rate": 3.860053754906176e-05,
+      "loss": 0.4376,
       "step": 46825
     },
     {
-      "epoch": 1.65,
-      "learning_rate": 3.913810677955297e-05,
-      "loss": 0.3244,
+      "epoch": 1.6877500270299493,
+      "grad_norm": 0.18308429419994354,
+      "learning_rate": 3.859808893429019e-05,
+      "loss": 0.4082,
       "step": 46830
     },
     {
-      "epoch": 1.65,
-      "learning_rate": 3.913575725713507e-05,
-      "loss": 0.3122,
+      "epoch": 1.687930226691174,
+      "grad_norm": 0.1854483038187027,
+      "learning_rate": 3.859564013424729e-05,
+      "loss": 0.4091,
       "step": 46835
     },
     {
-      "epoch": 1.65,
-      "learning_rate": 3.9133407551173676e-05,
-      "loss": 0.2729,
+      "epoch": 1.6881104263523985,
+      "grad_norm": 0.2083449810743332,
+      "learning_rate": 3.859319114896643e-05,
+      "loss": 0.4431,
       "step": 46840
     },
     {
-      "epoch": 1.65,
-      "learning_rate": 3.91310576616993e-05,
-      "loss": 0.2804,
+      "epoch": 1.688290626013623,
+      "grad_norm": 0.15931734442710876,
+      "learning_rate": 3.859074197848097e-05,
+      "loss": 0.3969,
       "step": 46845
     },
     {
-      "epoch": 1.65,
-      "learning_rate": 3.912870758874246e-05,
-      "loss": 0.2933,
+      "epoch": 1.6884708256748477,
+      "grad_norm": 0.20284508168697357,
+      "learning_rate": 3.858829262282428e-05,
+      "loss": 0.4014,
       "step": 46850
     },
     {
-      "epoch": 1.65,
-      "learning_rate": 3.9126357332333656e-05,
-      "loss": 0.2887,
+      "epoch": 1.6886510253360725,
+      "grad_norm": 0.1814059317111969,
+      "learning_rate": 3.858584308202973e-05,
+      "loss": 0.4241,
       "step": 46855
     },
     {
-      "epoch": 1.65,
-      "learning_rate": 3.912400689250342e-05,
-      "loss": 0.2587,
+      "epoch": 1.688831224997297,
+      "grad_norm": 0.24571998417377472,
+      "learning_rate": 3.858339335613071e-05,
+      "loss": 0.4337,
       "step": 46860
     },
     {
-      "epoch": 1.65,
-      "learning_rate": 3.9121656269282245e-05,
-      "loss": 0.3161,
+      "epoch": 1.6890114246585215,
+      "grad_norm": 0.19344958662986755,
+      "learning_rate": 3.858094344516058e-05,
+      "loss": 0.4392,
       "step": 46865
     },
     {
-      "epoch": 1.65,
-      "learning_rate": 3.9119305462700696e-05,
-      "loss": 0.3236,
+      "epoch": 1.6891916243197462,
+      "grad_norm": 0.1718379110097885,
+      "learning_rate": 3.8578493349152725e-05,
+      "loss": 0.3957,
       "step": 46870
     },
     {
-      "epoch": 1.65,
-      "learning_rate": 3.911695447278925e-05,
-      "loss": 0.2894,
+      "epoch": 1.689371823980971,
+      "grad_norm": 0.18230155110359192,
+      "learning_rate": 3.857604306814052e-05,
+      "loss": 0.4316,
       "step": 46875
     },
     {
-      "epoch": 1.65,
-      "learning_rate": 3.911460329957846e-05,
-      "loss": 0.273,
+      "epoch": 1.6895520236421957,
+      "grad_norm": 0.2310900092124939,
+      "learning_rate": 3.857359260215736e-05,
+      "loss": 0.4239,
       "step": 46880
     },
     {
-      "epoch": 1.65,
-      "learning_rate": 3.9112251943098844e-05,
-      "loss": 0.2766,
+      "epoch": 1.6897322233034202,
+      "grad_norm": 0.14590436220169067,
+      "learning_rate": 3.8571141951236634e-05,
+      "loss": 0.3717,
       "step": 46885
     },
     {
-      "epoch": 1.65,
-      "learning_rate": 3.910990040338094e-05,
-      "loss": 0.2898,
+      "epoch": 1.6899124229646447,
+      "grad_norm": 0.20509690046310425,
+      "learning_rate": 3.8568691115411726e-05,
+      "loss": 0.4504,
       "step": 46890
     },
     {
-      "epoch": 1.65,
-      "learning_rate": 3.910754868045528e-05,
-      "loss": 0.2967,
+      "epoch": 1.6900926226258695,
+      "grad_norm": 0.22596745193004608,
+      "learning_rate": 3.856624009471602e-05,
+      "loss": 0.4144,
       "step": 46895
     },
     {
-      "epoch": 1.65,
-      "learning_rate": 3.910519677435239e-05,
-      "loss": 0.3025,
+      "epoch": 1.6902728222870942,
+      "grad_norm": 0.21356570720672607,
+      "learning_rate": 3.8563788889182925e-05,
+      "loss": 0.4174,
       "step": 46900
     },
     {
-      "epoch": 1.65,
-      "learning_rate": 3.910284468510283e-05,
-      "loss": 0.2868,
+      "epoch": 1.6904530219483187,
+      "grad_norm": 0.16195428371429443,
+      "learning_rate": 3.856133749884584e-05,
+      "loss": 0.4089,
       "step": 46905
     },
     {
-      "epoch": 1.65,
-      "learning_rate": 3.910049241273712e-05,
-      "loss": 0.3248,
+      "epoch": 1.6906332216095432,
+      "grad_norm": 0.1678587794303894,
+      "learning_rate": 3.8558885923738144e-05,
+      "loss": 0.4158,
       "step": 46910
     },
     {
-      "epoch": 1.65,
-      "learning_rate": 3.909813995728581e-05,
-      "loss": 0.2966,
+      "epoch": 1.690813421270768,
+      "grad_norm": 0.16259028017520905,
+      "learning_rate": 3.8556434163893254e-05,
+      "loss": 0.3828,
       "step": 46915
     },
     {
-      "epoch": 1.65,
-      "learning_rate": 3.909578731877943e-05,
-      "loss": 0.276,
+      "epoch": 1.6909936209319927,
+      "grad_norm": 0.16952233016490936,
+      "learning_rate": 3.8553982219344584e-05,
+      "loss": 0.4048,
       "step": 46920
     },
     {
-      "epoch": 1.65,
-      "learning_rate": 3.909343449724856e-05,
-      "loss": 0.2901,
+      "epoch": 1.6911738205932174,
+      "grad_norm": 0.20961052179336548,
+      "learning_rate": 3.855153009012552e-05,
+      "loss": 0.4091,
       "step": 46925
     },
     {
-      "epoch": 1.65,
-      "learning_rate": 3.909108149272371e-05,
-      "loss": 0.2952,
+      "epoch": 1.691354020254442,
+      "grad_norm": 0.17797677218914032,
+      "learning_rate": 3.854907777626948e-05,
+      "loss": 0.4102,
       "step": 46930
     },
     {
-      "epoch": 1.65,
-      "learning_rate": 3.908872830523547e-05,
-      "loss": 0.2676,
+      "epoch": 1.6915342199156664,
+      "grad_norm": 0.23650279641151428,
+      "learning_rate": 3.854662527780989e-05,
+      "loss": 0.4334,
       "step": 46935
     },
     {
-      "epoch": 1.65,
-      "learning_rate": 3.9086374934814375e-05,
-      "loss": 0.2738,
+      "epoch": 1.6917144195768912,
+      "grad_norm": 0.1639034003019333,
+      "learning_rate": 3.8544172594780145e-05,
+      "loss": 0.3668,
       "step": 46940
     },
     {
-      "epoch": 1.65,
-      "learning_rate": 3.9084021381490986e-05,
-      "loss": 0.2668,
+      "epoch": 1.691894619238116,
+      "grad_norm": 0.14335285127162933,
+      "learning_rate": 3.8541719727213675e-05,
+      "loss": 0.38,
       "step": 46945
     },
     {
-      "epoch": 1.65,
-      "learning_rate": 3.908166764529585e-05,
-      "loss": 0.2843,
+      "epoch": 1.6920748188993404,
+      "grad_norm": 0.2120550274848938,
+      "learning_rate": 3.853926667514389e-05,
+      "loss": 0.4064,
       "step": 46950
     },
     {
-      "epoch": 1.65,
-      "learning_rate": 3.907931372625955e-05,
-      "loss": 0.2964,
+      "epoch": 1.6922550185605651,
+      "grad_norm": 0.1797804832458496,
+      "learning_rate": 3.853681343860423e-05,
+      "loss": 0.4465,
       "step": 46955
     },
     {
-      "epoch": 1.65,
-      "learning_rate": 3.907695962441264e-05,
-      "loss": 0.2733,
+      "epoch": 1.6924352182217897,
+      "grad_norm": 0.17130543291568756,
+      "learning_rate": 3.8534360017628096e-05,
+      "loss": 0.3994,
       "step": 46960
     },
     {
-      "epoch": 1.65,
-      "learning_rate": 3.907460533978569e-05,
-      "loss": 0.2851,
+      "epoch": 1.6926154178830144,
+      "grad_norm": 0.17090466618537903,
+      "learning_rate": 3.853190641224893e-05,
+      "loss": 0.4236,
       "step": 46965
     },
     {
-      "epoch": 1.65,
-      "learning_rate": 3.907225087240925e-05,
-      "loss": 0.2835,
+      "epoch": 1.6927956175442391,
+      "grad_norm": 0.1600097268819809,
+      "learning_rate": 3.852945262250016e-05,
+      "loss": 0.4157,
       "step": 46970
     },
     {
-      "epoch": 1.65,
-      "learning_rate": 3.906989622231392e-05,
-      "loss": 0.2729,
+      "epoch": 1.6929758172054636,
+      "grad_norm": 0.16842542588710785,
+      "learning_rate": 3.8526998648415214e-05,
+      "loss": 0.4017,
       "step": 46975
     },
     {
-      "epoch": 1.65,
-      "learning_rate": 3.9067541389530256e-05,
-      "loss": 0.3028,
+      "epoch": 1.6931560168666882,
+      "grad_norm": 0.2014317363500595,
+      "learning_rate": 3.8524544490027534e-05,
+      "loss": 0.4108,
       "step": 46980
     },
     {
-      "epoch": 1.65,
-      "learning_rate": 3.906518637408884e-05,
-      "loss": 0.3031,
+      "epoch": 1.6933362165279129,
+      "grad_norm": 0.18398632109165192,
+      "learning_rate": 3.852209014737055e-05,
+      "loss": 0.3708,
       "step": 46985
     },
     {
-      "epoch": 1.65,
-      "learning_rate": 3.906283117602024e-05,
-      "loss": 0.3047,
+      "epoch": 1.6935164161891376,
+      "grad_norm": 0.217530757188797,
+      "learning_rate": 3.8519635620477714e-05,
+      "loss": 0.4193,
       "step": 46990
     },
     {
-      "epoch": 1.65,
-      "learning_rate": 3.906047579535504e-05,
-      "loss": 0.2737,
+      "epoch": 1.6936966158503624,
+      "grad_norm": 0.15751570463180542,
+      "learning_rate": 3.851718090938245e-05,
+      "loss": 0.4189,
       "step": 46995
     },
     {
-      "epoch": 1.65,
-      "learning_rate": 3.905812023212385e-05,
-      "loss": 0.3082,
+      "epoch": 1.6938768155115869,
+      "grad_norm": 0.21628709137439728,
+      "learning_rate": 3.851472601411822e-05,
+      "loss": 0.4166,
       "step": 47000
     },
     {
-      "epoch": 1.65,
-      "eval_loss": 0.27732065320014954,
-      "eval_runtime": 10.5336,
-      "eval_samples_per_second": 9.493,
-      "eval_steps_per_second": 9.493,
+      "epoch": 1.6938768155115869,
+      "eval_loss": 0.44244661927223206,
+      "eval_runtime": 3.5403,
+      "eval_samples_per_second": 28.246,
+      "eval_steps_per_second": 7.062,
       "step": 47000
     },
     {
-      "epoch": 1.65,
-      "learning_rate": 3.905576448635722e-05,
-      "loss": 0.3076,
+      "epoch": 1.6940570151728114,
+      "grad_norm": 0.19788464903831482,
+      "learning_rate": 3.851227093471847e-05,
+      "loss": 0.424,
       "step": 47005
     },
     {
-      "epoch": 1.65,
-      "learning_rate": 3.905340855808576e-05,
-      "loss": 0.2795,
+      "epoch": 1.694237214834036,
+      "grad_norm": 0.1924174278974533,
+      "learning_rate": 3.850981567121663e-05,
+      "loss": 0.4527,
       "step": 47010
     },
     {
-      "epoch": 1.65,
-      "learning_rate": 3.905105244734004e-05,
-      "loss": 0.2857,
+      "epoch": 1.6944174144952608,
+      "grad_norm": 0.16324442625045776,
+      "learning_rate": 3.850736022364617e-05,
+      "loss": 0.4156,
       "step": 47015
     },
     {
-      "epoch": 1.65,
-      "learning_rate": 3.9048696154150666e-05,
-      "loss": 0.2797,
+      "epoch": 1.6945976141564854,
+      "grad_norm": 0.23435580730438232,
+      "learning_rate": 3.8504904592040545e-05,
+      "loss": 0.4468,
       "step": 47020
     },
     {
-      "epoch": 1.65,
-      "learning_rate": 3.9046339678548234e-05,
-      "loss": 0.2599,
+      "epoch": 1.6947778138177099,
+      "grad_norm": 0.15442748367786407,
+      "learning_rate": 3.8502448776433216e-05,
+      "loss": 0.3801,
       "step": 47025
     },
     {
-      "epoch": 1.65,
-      "learning_rate": 3.904398302056333e-05,
-      "loss": 0.2587,
+      "epoch": 1.6949580134789346,
+      "grad_norm": 0.18875345587730408,
+      "learning_rate": 3.849999277685763e-05,
+      "loss": 0.3887,
       "step": 47030
     },
     {
-      "epoch": 1.65,
-      "learning_rate": 3.9041626180226575e-05,
-      "loss": 0.2786,
+      "epoch": 1.6951382131401593,
+      "grad_norm": 0.18448412418365479,
+      "learning_rate": 3.849753659334725e-05,
+      "loss": 0.4143,
       "step": 47035
     },
     {
-      "epoch": 1.65,
-      "learning_rate": 3.903926915756856e-05,
-      "loss": 0.2823,
+      "epoch": 1.695318412801384,
+      "grad_norm": 0.20887787640094757,
+      "learning_rate": 3.849508022593556e-05,
+      "loss": 0.4555,
       "step": 47040
     },
     {
-      "epoch": 1.66,
-      "learning_rate": 3.903691195261987e-05,
-      "loss": 0.2879,
+      "epoch": 1.6954986124626086,
+      "grad_norm": 0.16566239297389984,
+      "learning_rate": 3.849262367465601e-05,
+      "loss": 0.3988,
       "step": 47045
     },
     {
-      "epoch": 1.66,
-      "learning_rate": 3.903455456541114e-05,
-      "loss": 0.2644,
+      "epoch": 1.695678812123833,
+      "grad_norm": 0.2598322033882141,
+      "learning_rate": 3.849016693954207e-05,
+      "loss": 0.431,
       "step": 47050
     },
     {
-      "epoch": 1.66,
-      "learning_rate": 3.903219699597297e-05,
-      "loss": 0.2788,
+      "epoch": 1.6958590117850578,
+      "grad_norm": 0.1612170934677124,
+      "learning_rate": 3.848771002062722e-05,
+      "loss": 0.3811,
       "step": 47055
     },
     {
-      "epoch": 1.66,
-      "learning_rate": 3.902983924433596e-05,
-      "loss": 0.2764,
+      "epoch": 1.6960392114462826,
+      "grad_norm": 0.18904070556163788,
+      "learning_rate": 3.848525291794494e-05,
+      "loss": 0.4372,
       "step": 47060
     },
     {
-      "epoch": 1.66,
-      "learning_rate": 3.902748131053074e-05,
-      "loss": 0.2835,
+      "epoch": 1.696219411107507,
+      "grad_norm": 0.1836709976196289,
+      "learning_rate": 3.848279563152869e-05,
+      "loss": 0.3892,
       "step": 47065
     },
     {
-      "epoch": 1.66,
-      "learning_rate": 3.9025123194587927e-05,
-      "loss": 0.2733,
+      "epoch": 1.6963996107687318,
+      "grad_norm": 0.18471242487430573,
+      "learning_rate": 3.848033816141196e-05,
+      "loss": 0.3994,
       "step": 47070
     },
     {
-      "epoch": 1.66,
-      "learning_rate": 3.902276489653812e-05,
-      "loss": 0.2824,
+      "epoch": 1.6965798104299563,
+      "grad_norm": 0.18308402597904205,
+      "learning_rate": 3.847788050762824e-05,
+      "loss": 0.4139,
       "step": 47075
     },
     {
-      "epoch": 1.66,
-      "learning_rate": 3.902040641641196e-05,
-      "loss": 0.2928,
+      "epoch": 1.696760010091181,
+      "grad_norm": 0.1916336864233017,
+      "learning_rate": 3.8475422670211e-05,
+      "loss": 0.4577,
       "step": 47080
     },
     {
-      "epoch": 1.66,
-      "learning_rate": 3.9018047754240054e-05,
-      "loss": 0.2955,
+      "epoch": 1.6969402097524058,
+      "grad_norm": 0.17841115593910217,
+      "learning_rate": 3.8472964649193736e-05,
+      "loss": 0.4185,
       "step": 47085
     },
     {
-      "epoch": 1.66,
-      "learning_rate": 3.901568891005305e-05,
-      "loss": 0.284,
+      "epoch": 1.6971204094136303,
+      "grad_norm": 0.16689583659172058,
+      "learning_rate": 3.8470506444609946e-05,
+      "loss": 0.3826,
       "step": 47090
     },
     {
-      "epoch": 1.66,
-      "learning_rate": 3.9013329883881544e-05,
-      "loss": 0.316,
+      "epoch": 1.6973006090748548,
+      "grad_norm": 0.21197299659252167,
+      "learning_rate": 3.84680480564931e-05,
+      "loss": 0.4296,
       "step": 47095
     },
     {
-      "epoch": 1.66,
-      "learning_rate": 3.90109706757562e-05,
-      "loss": 0.297,
+      "epoch": 1.6974808087360795,
+      "grad_norm": 0.14996369183063507,
+      "learning_rate": 3.8465589484876716e-05,
+      "loss": 0.4002,
       "step": 47100
     },
     {
-      "epoch": 1.66,
-      "learning_rate": 3.900861128570763e-05,
-      "loss": 0.2845,
+      "epoch": 1.6976610083973043,
+      "grad_norm": 0.15397225320339203,
+      "learning_rate": 3.846313072979428e-05,
+      "loss": 0.3986,
       "step": 47105
     },
     {
-      "epoch": 1.66,
-      "learning_rate": 3.9006251713766474e-05,
-      "loss": 0.2836,
+      "epoch": 1.697841208058529,
+      "grad_norm": 0.1733284294605255,
+      "learning_rate": 3.846067179127929e-05,
+      "loss": 0.4098,
       "step": 47110
     },
     {
-      "epoch": 1.66,
-      "learning_rate": 3.900389195996337e-05,
-      "loss": 0.3025,
+      "epoch": 1.6980214077197535,
+      "grad_norm": 0.16923348605632782,
+      "learning_rate": 3.8458212669365256e-05,
+      "loss": 0.3978,
       "step": 47115
     },
     {
-      "epoch": 1.66,
-      "learning_rate": 3.900153202432895e-05,
-      "loss": 0.2968,
+      "epoch": 1.698201607380978,
+      "grad_norm": 0.16483382880687714,
+      "learning_rate": 3.845575336408568e-05,
+      "loss": 0.4234,
       "step": 47120
     },
     {
-      "epoch": 1.66,
-      "learning_rate": 3.8999171906893875e-05,
-      "loss": 0.2915,
+      "epoch": 1.6983818070422028,
+      "grad_norm": 0.18694746494293213,
+      "learning_rate": 3.845329387547407e-05,
+      "loss": 0.4383,
       "step": 47125
     },
     {
-      "epoch": 1.66,
-      "learning_rate": 3.8996811607688774e-05,
-      "loss": 0.2617,
+      "epoch": 1.6985620067034275,
+      "grad_norm": 0.20622876286506653,
+      "learning_rate": 3.845083420356393e-05,
+      "loss": 0.4094,
       "step": 47130
     },
     {
-      "epoch": 1.66,
-      "learning_rate": 3.8994451126744305e-05,
-      "loss": 0.2753,
+      "epoch": 1.698742206364652,
+      "grad_norm": 0.177239328622818,
+      "learning_rate": 3.8448374348388796e-05,
+      "loss": 0.4394,
       "step": 47135
     },
     {
-      "epoch": 1.66,
-      "learning_rate": 3.89920904640911e-05,
-      "loss": 0.3031,
+      "epoch": 1.6989224060258765,
+      "grad_norm": 0.21104155480861664,
+      "learning_rate": 3.8445914309982145e-05,
+      "loss": 0.4234,
       "step": 47140
     },
     {
-      "epoch": 1.66,
-      "learning_rate": 3.898972961975983e-05,
-      "loss": 0.3031,
+      "epoch": 1.6991026056871013,
+      "grad_norm": 0.19302360713481903,
+      "learning_rate": 3.844345408837753e-05,
+      "loss": 0.3983,
       "step": 47145
     },
     {
-      "epoch": 1.66,
-      "learning_rate": 3.898736859378114e-05,
-      "loss": 0.2961,
+      "epoch": 1.699282805348326,
+      "grad_norm": 0.18512195348739624,
+      "learning_rate": 3.844099368360845e-05,
+      "loss": 0.4173,
       "step": 47150
     },
     {
-      "epoch": 1.66,
-      "learning_rate": 3.898500738618568e-05,
-      "loss": 0.2889,
+      "epoch": 1.6994630050095507,
+      "grad_norm": 0.17956486344337463,
+      "learning_rate": 3.8438533095708426e-05,
+      "loss": 0.4372,
       "step": 47155
     },
     {
-      "epoch": 1.66,
-      "learning_rate": 3.898264599700412e-05,
-      "loss": 0.2899,
+      "epoch": 1.6996432046707752,
+      "grad_norm": 0.19222790002822876,
+      "learning_rate": 3.8436072324710995e-05,
+      "loss": 0.3728,
       "step": 47160
     },
     {
-      "epoch": 1.66,
-      "learning_rate": 3.898028442626712e-05,
-      "loss": 0.2805,
+      "epoch": 1.6998234043319997,
+      "grad_norm": 0.15799042582511902,
+      "learning_rate": 3.8433611370649686e-05,
+      "loss": 0.4133,
       "step": 47165
     },
     {
-      "epoch": 1.66,
-      "learning_rate": 3.897792267400533e-05,
-      "loss": 0.2881,
+      "epoch": 1.7000036039932245,
+      "grad_norm": 0.1787501573562622,
+      "learning_rate": 3.843115023355802e-05,
+      "loss": 0.4264,
       "step": 47170
     },
     {
-      "epoch": 1.66,
-      "learning_rate": 3.897556074024944e-05,
-      "loss": 0.2914,
+      "epoch": 1.7001838036544492,
+      "grad_norm": 0.17307919263839722,
+      "learning_rate": 3.842868891346952e-05,
+      "loss": 0.4198,
       "step": 47175
     },
     {
-      "epoch": 1.66,
-      "learning_rate": 3.8973198625030094e-05,
-      "loss": 0.3109,
+      "epoch": 1.7003640033156737,
+      "grad_norm": 0.1564173549413681,
+      "learning_rate": 3.8426227410417755e-05,
+      "loss": 0.3832,
       "step": 47180
     },
     {
-      "epoch": 1.66,
-      "learning_rate": 3.897083632837797e-05,
-      "loss": 0.2974,
+      "epoch": 1.7005442029768985,
+      "grad_norm": 0.21778932213783264,
+      "learning_rate": 3.842376572443623e-05,
+      "loss": 0.4172,
       "step": 47185
     },
     {
-      "epoch": 1.66,
-      "learning_rate": 3.896847385032375e-05,
-      "loss": 0.2844,
+      "epoch": 1.700724402638123,
+      "grad_norm": 0.16438071429729462,
+      "learning_rate": 3.8421303855558496e-05,
+      "loss": 0.4138,
       "step": 47190
     },
     {
-      "epoch": 1.66,
-      "learning_rate": 3.89661111908981e-05,
-      "loss": 0.2585,
+      "epoch": 1.7009046022993477,
+      "grad_norm": 0.2048167735338211,
+      "learning_rate": 3.8418841803818096e-05,
+      "loss": 0.4055,
       "step": 47195
     },
     {
-      "epoch": 1.66,
-      "learning_rate": 3.89637483501317e-05,
-      "loss": 0.2966,
+      "epoch": 1.7010848019605724,
+      "grad_norm": 0.15939681231975555,
+      "learning_rate": 3.841637956924857e-05,
+      "loss": 0.4179,
       "step": 47200
     },
     {
-      "epoch": 1.66,
-      "learning_rate": 3.896138532805523e-05,
-      "loss": 0.2913,
+      "epoch": 1.701265001621797,
+      "grad_norm": 0.16138513386249542,
+      "learning_rate": 3.841391715188348e-05,
+      "loss": 0.416,
       "step": 47205
     },
     {
-      "epoch": 1.66,
-      "learning_rate": 3.8959022124699375e-05,
-      "loss": 0.2973,
+      "epoch": 1.7014452012830215,
+      "grad_norm": 0.186210036277771,
+      "learning_rate": 3.841145455175635e-05,
+      "loss": 0.4287,
       "step": 47210
     },
     {
-      "epoch": 1.66,
-      "learning_rate": 3.895665874009482e-05,
-      "loss": 0.3064,
+      "epoch": 1.7016254009442462,
+      "grad_norm": 0.17103976011276245,
+      "learning_rate": 3.840899176890076e-05,
+      "loss": 0.3943,
       "step": 47215
     },
     {
-      "epoch": 1.66,
-      "learning_rate": 3.8954295174272245e-05,
-      "loss": 0.2975,
+      "epoch": 1.701805600605471,
+      "grad_norm": 0.15319930016994476,
+      "learning_rate": 3.840652880335025e-05,
+      "loss": 0.4086,
       "step": 47220
     },
     {
-      "epoch": 1.66,
-      "learning_rate": 3.895193142726234e-05,
-      "loss": 0.2924,
+      "epoch": 1.7019858002666957,
+      "grad_norm": 0.1645641177892685,
+      "learning_rate": 3.840406565513838e-05,
+      "loss": 0.3827,
       "step": 47225
     },
     {
-      "epoch": 1.66,
-      "learning_rate": 3.894956749909581e-05,
-      "loss": 0.2741,
+      "epoch": 1.7021659999279202,
+      "grad_norm": 0.1592930108308792,
+      "learning_rate": 3.840160232429872e-05,
+      "loss": 0.3865,
       "step": 47230
     },
     {
-      "epoch": 1.66,
-      "learning_rate": 3.894720338980332e-05,
-      "loss": 0.2908,
+      "epoch": 1.7023461995891447,
+      "grad_norm": 0.18125946819782257,
+      "learning_rate": 3.839913881086481e-05,
+      "loss": 0.428,
       "step": 47235
     },
     {
-      "epoch": 1.66,
-      "learning_rate": 3.89448390994156e-05,
-      "loss": 0.3165,
+      "epoch": 1.7025263992503694,
+      "grad_norm": 0.17202188074588776,
+      "learning_rate": 3.8396675114870234e-05,
+      "loss": 0.4201,
       "step": 47240
     },
     {
-      "epoch": 1.66,
-      "learning_rate": 3.894247462796333e-05,
-      "loss": 0.2858,
+      "epoch": 1.7027065989115941,
+      "grad_norm": 0.16342946887016296,
+      "learning_rate": 3.839421123634855e-05,
+      "loss": 0.389,
       "step": 47245
     },
     {
-      "epoch": 1.66,
-      "learning_rate": 3.894010997547722e-05,
-      "loss": 0.268,
+      "epoch": 1.7028867985728187,
+      "grad_norm": 0.17867057025432587,
+      "learning_rate": 3.8391747175333336e-05,
+      "loss": 0.4173,
       "step": 47250
     },
     {
-      "epoch": 1.66,
-      "learning_rate": 3.893774514198797e-05,
-      "loss": 0.2751,
+      "epoch": 1.7030669982340432,
+      "grad_norm": 0.20603960752487183,
+      "learning_rate": 3.838928293185815e-05,
+      "loss": 0.4412,
       "step": 47255
     },
     {
-      "epoch": 1.66,
-      "learning_rate": 3.8935380127526284e-05,
-      "loss": 0.2815,
+      "epoch": 1.703247197895268,
+      "grad_norm": 0.19826002418994904,
+      "learning_rate": 3.838681850595659e-05,
+      "loss": 0.3956,
       "step": 47260
     },
     {
-      "epoch": 1.66,
-      "learning_rate": 3.893301493212286e-05,
-      "loss": 0.3071,
+      "epoch": 1.7034273975564926,
+      "grad_norm": 0.18762384355068207,
+      "learning_rate": 3.83843538976622e-05,
+      "loss": 0.4009,
       "step": 47265
     },
     {
-      "epoch": 1.66,
-      "learning_rate": 3.8930649555808434e-05,
-      "loss": 0.2742,
+      "epoch": 1.7036075972177174,
+      "grad_norm": 0.21648670732975006,
+      "learning_rate": 3.838188910700861e-05,
+      "loss": 0.4453,
       "step": 47270
     },
     {
-      "epoch": 1.66,
-      "learning_rate": 3.89282839986137e-05,
-      "loss": 0.3003,
+      "epoch": 1.7037877968789419,
+      "grad_norm": 0.20147664844989777,
+      "learning_rate": 3.837942413402935e-05,
+      "loss": 0.4143,
       "step": 47275
     },
     {
-      "epoch": 1.66,
-      "learning_rate": 3.892591826056937e-05,
-      "loss": 0.2946,
+      "epoch": 1.7039679965401664,
+      "grad_norm": 0.16938365995883942,
+      "learning_rate": 3.837695897875803e-05,
+      "loss": 0.4284,
       "step": 47280
     },
     {
-      "epoch": 1.66,
-      "learning_rate": 3.892355234170617e-05,
-      "loss": 0.3135,
+      "epoch": 1.7041481962013911,
+      "grad_norm": 0.16605401039123535,
+      "learning_rate": 3.837449364122823e-05,
+      "loss": 0.4519,
       "step": 47285
     },
     {
-      "epoch": 1.66,
-      "learning_rate": 3.892118624205483e-05,
-      "loss": 0.2809,
+      "epoch": 1.7043283958626159,
+      "grad_norm": 0.15813526511192322,
+      "learning_rate": 3.837202812147355e-05,
+      "loss": 0.4502,
       "step": 47290
     },
     {
-      "epoch": 1.66,
-      "learning_rate": 3.891881996164605e-05,
-      "loss": 0.2887,
+      "epoch": 1.7045085955238404,
+      "grad_norm": 0.1578982025384903,
+      "learning_rate": 3.8369562419527574e-05,
+      "loss": 0.407,
       "step": 47295
     },
     {
-      "epoch": 1.66,
-      "learning_rate": 3.891645350051057e-05,
-      "loss": 0.2931,
+      "epoch": 1.7046887951850649,
+      "grad_norm": 0.20871727168560028,
+      "learning_rate": 3.8367096535423895e-05,
+      "loss": 0.4251,
       "step": 47300
     },
     {
-      "epoch": 1.66,
-      "learning_rate": 3.891408685867911e-05,
-      "loss": 0.3101,
+      "epoch": 1.7048689948462896,
+      "grad_norm": 0.1831165850162506,
+      "learning_rate": 3.836463046919612e-05,
+      "loss": 0.4263,
       "step": 47305
     },
     {
-      "epoch": 1.66,
-      "learning_rate": 3.891172003618241e-05,
-      "loss": 0.3029,
+      "epoch": 1.7050491945075144,
+      "grad_norm": 0.2124657928943634,
+      "learning_rate": 3.836216422087784e-05,
+      "loss": 0.3903,
       "step": 47310
     },
     {
-      "epoch": 1.66,
-      "learning_rate": 3.890935303305118e-05,
-      "loss": 0.279,
+      "epoch": 1.705229394168739,
+      "grad_norm": 0.2060711830854416,
+      "learning_rate": 3.8359697790502656e-05,
+      "loss": 0.4089,
       "step": 47315
     },
     {
-      "epoch": 1.66,
-      "learning_rate": 3.890698584931618e-05,
-      "loss": 0.2807,
+      "epoch": 1.7054095938299636,
+      "grad_norm": 0.1653200387954712,
+      "learning_rate": 3.835723117810418e-05,
+      "loss": 0.376,
       "step": 47320
     },
     {
-      "epoch": 1.67,
-      "learning_rate": 3.890461848500813e-05,
-      "loss": 0.2789,
+      "epoch": 1.705589793491188,
+      "grad_norm": 0.15278410911560059,
+      "learning_rate": 3.835476438371601e-05,
+      "loss": 0.4144,
       "step": 47325
     },
     {
-      "epoch": 1.67,
-      "learning_rate": 3.890225094015777e-05,
-      "loss": 0.2878,
+      "epoch": 1.7057699931524128,
+      "grad_norm": 0.21837785840034485,
+      "learning_rate": 3.835229740737176e-05,
+      "loss": 0.4179,
       "step": 47330
     },
     {
-      "epoch": 1.67,
-      "learning_rate": 3.889988321479584e-05,
-      "loss": 0.2847,
+      "epoch": 1.7059501928136376,
+      "grad_norm": 0.17335210740566254,
+      "learning_rate": 3.8349830249105057e-05,
+      "loss": 0.3892,
       "step": 47335
     },
     {
-      "epoch": 1.67,
-      "learning_rate": 3.889751530895309e-05,
-      "loss": 0.2574,
+      "epoch": 1.7061303924748623,
+      "grad_norm": 0.17480070888996124,
+      "learning_rate": 3.8347362908949484e-05,
+      "loss": 0.4117,
       "step": 47340
     },
     {
-      "epoch": 1.67,
-      "learning_rate": 3.8895147222660264e-05,
-      "loss": 0.2951,
+      "epoch": 1.7063105921360868,
+      "grad_norm": 0.2083815336227417,
+      "learning_rate": 3.834489538693868e-05,
+      "loss": 0.4644,
       "step": 47345
     },
     {
-      "epoch": 1.67,
-      "learning_rate": 3.8892778955948104e-05,
-      "loss": 0.3026,
+      "epoch": 1.7064907917973113,
+      "grad_norm": 0.1669875979423523,
+      "learning_rate": 3.8342427683106276e-05,
+      "loss": 0.3913,
       "step": 47350
     },
     {
-      "epoch": 1.67,
-      "learning_rate": 3.8890410508847376e-05,
-      "loss": 0.2767,
+      "epoch": 1.706670991458536,
+      "grad_norm": 0.16881945729255676,
+      "learning_rate": 3.833995979748587e-05,
+      "loss": 0.4141,
       "step": 47355
     },
     {
-      "epoch": 1.67,
-      "learning_rate": 3.888804188138882e-05,
-      "loss": 0.303,
+      "epoch": 1.7068511911197608,
+      "grad_norm": 0.17470820248126984,
+      "learning_rate": 3.833749173011108e-05,
+      "loss": 0.4245,
       "step": 47360
     },
     {
-      "epoch": 1.67,
-      "learning_rate": 3.888567307360318e-05,
-      "loss": 0.2889,
+      "epoch": 1.7070313907809853,
+      "grad_norm": 0.18530166149139404,
+      "learning_rate": 3.833502348101556e-05,
+      "loss": 0.4304,
       "step": 47365
     },
     {
-      "epoch": 1.67,
-      "learning_rate": 3.8883304085521236e-05,
-      "loss": 0.2902,
+      "epoch": 1.7072115904422098,
+      "grad_norm": 0.206429585814476,
+      "learning_rate": 3.833255505023292e-05,
+      "loss": 0.4014,
       "step": 47370
     },
     {
-      "epoch": 1.67,
-      "learning_rate": 3.8880934917173735e-05,
-      "loss": 0.2759,
+      "epoch": 1.7073917901034346,
+      "grad_norm": 0.172637477517128,
+      "learning_rate": 3.83300864377968e-05,
+      "loss": 0.3989,
       "step": 47375
     },
     {
-      "epoch": 1.67,
-      "learning_rate": 3.887856556859144e-05,
-      "loss": 0.3042,
+      "epoch": 1.7075719897646593,
+      "grad_norm": 0.18102556467056274,
+      "learning_rate": 3.832761764374084e-05,
+      "loss": 0.4066,
       "step": 47380
     },
     {
-      "epoch": 1.67,
-      "learning_rate": 3.887619603980512e-05,
-      "loss": 0.3202,
+      "epoch": 1.707752189425884,
+      "grad_norm": 0.16687656939029694,
+      "learning_rate": 3.832514866809866e-05,
+      "loss": 0.4241,
       "step": 47385
     },
     {
-      "epoch": 1.67,
-      "learning_rate": 3.8873826330845545e-05,
-      "loss": 0.2898,
+      "epoch": 1.7079323890871085,
+      "grad_norm": 0.17765361070632935,
+      "learning_rate": 3.832267951090392e-05,
+      "loss": 0.404,
       "step": 47390
     },
     {
-      "epoch": 1.67,
-      "learning_rate": 3.887145644174347e-05,
-      "loss": 0.2714,
+      "epoch": 1.708112588748333,
+      "grad_norm": 0.20449407398700714,
+      "learning_rate": 3.832021017219025e-05,
+      "loss": 0.4424,
       "step": 47395
     },
     {
-      "epoch": 1.67,
-      "learning_rate": 3.886908637252967e-05,
-      "loss": 0.2699,
+      "epoch": 1.7082927884095578,
+      "grad_norm": 0.20661450922489166,
+      "learning_rate": 3.831774065199128e-05,
+      "loss": 0.444,
       "step": 47400
     },
     {
-      "epoch": 1.67,
-      "learning_rate": 3.8866716123234934e-05,
-      "loss": 0.269,
+      "epoch": 1.7084729880707825,
+      "grad_norm": 0.1424778401851654,
+      "learning_rate": 3.8315270950340684e-05,
+      "loss": 0.3854,
       "step": 47405
     },
     {
-      "epoch": 1.67,
-      "learning_rate": 3.8864345693890024e-05,
-      "loss": 0.2756,
+      "epoch": 1.708653187732007,
+      "grad_norm": 0.1613854020833969,
+      "learning_rate": 3.831280106727211e-05,
+      "loss": 0.4014,
       "step": 47410
     },
     {
-      "epoch": 1.67,
-      "learning_rate": 3.886197508452572e-05,
-      "loss": 0.2592,
+      "epoch": 1.7088333873932315,
+      "grad_norm": 0.19576317071914673,
+      "learning_rate": 3.8310331002819186e-05,
+      "loss": 0.428,
       "step": 47415
     },
     {
-      "epoch": 1.67,
-      "learning_rate": 3.8859604295172816e-05,
-      "loss": 0.273,
+      "epoch": 1.7090135870544563,
+      "grad_norm": 0.19375436007976532,
+      "learning_rate": 3.830786075701558e-05,
+      "loss": 0.4368,
       "step": 47420
     },
     {
-      "epoch": 1.67,
-      "learning_rate": 3.8857233325862074e-05,
-      "loss": 0.2826,
+      "epoch": 1.709193786715681,
+      "grad_norm": 0.24213196337223053,
+      "learning_rate": 3.8305390329894945e-05,
+      "loss": 0.4112,
       "step": 47425
     },
     {
-      "epoch": 1.67,
-      "learning_rate": 3.885486217662428e-05,
-      "loss": 0.3095,
+      "epoch": 1.7093739863769057,
+      "grad_norm": 0.15626302361488342,
+      "learning_rate": 3.830291972149095e-05,
+      "loss": 0.4208,
       "step": 47430
     },
     {
-      "epoch": 1.67,
-      "learning_rate": 3.885249084749025e-05,
-      "loss": 0.2889,
+      "epoch": 1.7095541860381303,
+      "grad_norm": 0.18249692022800446,
+      "learning_rate": 3.8300448931837244e-05,
+      "loss": 0.3906,
       "step": 47435
     },
     {
-      "epoch": 1.67,
-      "learning_rate": 3.8850119338490745e-05,
-      "loss": 0.2691,
+      "epoch": 1.7097343856993548,
+      "grad_norm": 0.15607425570487976,
+      "learning_rate": 3.829797796096749e-05,
+      "loss": 0.4015,
       "step": 47440
     },
     {
-      "epoch": 1.67,
-      "learning_rate": 3.884774764965658e-05,
-      "loss": 0.279,
+      "epoch": 1.7099145853605795,
+      "grad_norm": 0.1599874198436737,
+      "learning_rate": 3.829550680891537e-05,
+      "loss": 0.4124,
       "step": 47445
     },
     {
-      "epoch": 1.67,
-      "learning_rate": 3.8845375781018524e-05,
-      "loss": 0.3029,
+      "epoch": 1.7100947850218042,
+      "grad_norm": 0.2017892599105835,
+      "learning_rate": 3.8293035475714543e-05,
+      "loss": 0.4204,
       "step": 47450
     },
     {
-      "epoch": 1.67,
-      "learning_rate": 3.884300373260739e-05,
-      "loss": 0.2701,
+      "epoch": 1.7102749846830287,
+      "grad_norm": 0.17167171835899353,
+      "learning_rate": 3.8290563961398686e-05,
+      "loss": 0.3933,
       "step": 47455
     },
     {
-      "epoch": 1.67,
-      "learning_rate": 3.8840631504453975e-05,
-      "loss": 0.2512,
+      "epoch": 1.7104551843442535,
+      "grad_norm": 0.18356946110725403,
+      "learning_rate": 3.828809226600146e-05,
+      "loss": 0.4294,
       "step": 47460
     },
     {
-      "epoch": 1.67,
-      "learning_rate": 3.883825909658908e-05,
-      "loss": 0.2756,
+      "epoch": 1.710635384005478,
+      "grad_norm": 0.20430004596710205,
+      "learning_rate": 3.828562038955655e-05,
+      "loss": 0.4215,
       "step": 47465
     },
     {
-      "epoch": 1.67,
-      "learning_rate": 3.883588650904352e-05,
-      "loss": 0.2821,
+      "epoch": 1.7108155836667027,
+      "grad_norm": 0.19897079467773438,
+      "learning_rate": 3.828314833209764e-05,
+      "loss": 0.4301,
       "step": 47470
     },
     {
-      "epoch": 1.67,
-      "learning_rate": 3.883351374184808e-05,
-      "loss": 0.2952,
+      "epoch": 1.7109957833279275,
+      "grad_norm": 0.17606034874916077,
+      "learning_rate": 3.828067609365841e-05,
+      "loss": 0.4146,
       "step": 47475
     },
     {
-      "epoch": 1.67,
-      "learning_rate": 3.883114079503359e-05,
-      "loss": 0.2917,
+      "epoch": 1.711175982989152,
+      "grad_norm": 0.18910571932792664,
+      "learning_rate": 3.827869817262381e-05,
+      "loss": 0.4191,
       "step": 47480
     },
     {
-      "epoch": 1.67,
-      "learning_rate": 3.882876766863085e-05,
-      "loss": 0.2885,
+      "epoch": 1.7113561826503765,
+      "grad_norm": 0.16573849320411682,
+      "learning_rate": 3.827622560850488e-05,
+      "loss": 0.4207,
       "step": 47485
     },
     {
-      "epoch": 1.67,
-      "learning_rate": 3.8826394362670684e-05,
-      "loss": 0.3046,
+      "epoch": 1.7115363823116012,
+      "grad_norm": 0.18701373040676117,
+      "learning_rate": 3.827375286349993e-05,
+      "loss": 0.4094,
       "step": 47490
     },
     {
-      "epoch": 1.67,
-      "learning_rate": 3.8824020877183886e-05,
-      "loss": 0.2945,
+      "epoch": 1.711716581972826,
+      "grad_norm": 0.16349554061889648,
+      "learning_rate": 3.827127993764269e-05,
+      "loss": 0.4338,
       "step": 47495
     },
     {
-      "epoch": 1.67,
-      "learning_rate": 3.8821647212201286e-05,
-      "loss": 0.275,
+      "epoch": 1.7118967816340507,
+      "grad_norm": 0.17863896489143372,
+      "learning_rate": 3.826880683096681e-05,
+      "loss": 0.4307,
       "step": 47500
     },
     {
-      "epoch": 1.67,
-      "eval_loss": 0.2766549289226532,
-      "eval_runtime": 10.5338,
-      "eval_samples_per_second": 9.493,
-      "eval_steps_per_second": 9.493,
+      "epoch": 1.7118967816340507,
+      "eval_loss": 0.4411923587322235,
+      "eval_runtime": 3.5372,
+      "eval_samples_per_second": 28.271,
+      "eval_steps_per_second": 7.068,
       "step": 47500
     },
     {
-      "epoch": 1.67,
-      "learning_rate": 3.881927336775371e-05,
-      "loss": 0.3027,
+      "epoch": 1.7120769812952752,
+      "grad_norm": 0.2336321473121643,
+      "learning_rate": 3.8266333543506016e-05,
+      "loss": 0.4195,
       "step": 47505
     },
     {
-      "epoch": 1.67,
-      "learning_rate": 3.881689934387198e-05,
-      "loss": 0.2904,
+      "epoch": 1.7122571809564997,
+      "grad_norm": 0.22593377530574799,
+      "learning_rate": 3.826386007529399e-05,
+      "loss": 0.4147,
       "step": 47510
     },
     {
-      "epoch": 1.67,
-      "learning_rate": 3.881452514058691e-05,
-      "loss": 0.2958,
+      "epoch": 1.7124373806177244,
+      "grad_norm": 0.23086853325366974,
+      "learning_rate": 3.826138642636444e-05,
+      "loss": 0.4112,
       "step": 47515
     },
     {
-      "epoch": 1.67,
-      "learning_rate": 3.881215075792934e-05,
-      "loss": 0.2804,
+      "epoch": 1.7126175802789492,
+      "grad_norm": 0.2238939106464386,
+      "learning_rate": 3.8258912596751076e-05,
+      "loss": 0.3977,
       "step": 47520
     },
     {
-      "epoch": 1.67,
-      "learning_rate": 3.8809776195930095e-05,
-      "loss": 0.2713,
+      "epoch": 1.7127977799401737,
+      "grad_norm": 0.1743045598268509,
+      "learning_rate": 3.8256438586487584e-05,
+      "loss": 0.429,
       "step": 47525
     },
     {
-      "epoch": 1.67,
-      "learning_rate": 3.8807401454620004e-05,
-      "loss": 0.2963,
+      "epoch": 1.7129779796013982,
+      "grad_norm": 0.1635396033525467,
+      "learning_rate": 3.825396439560769e-05,
+      "loss": 0.3928,
       "step": 47530
     },
     {
-      "epoch": 1.67,
-      "learning_rate": 3.880502653402991e-05,
-      "loss": 0.2641,
+      "epoch": 1.713158179262623,
+      "grad_norm": 0.1982748806476593,
+      "learning_rate": 3.8251490024145085e-05,
+      "loss": 0.4473,
       "step": 47535
     },
     {
-      "epoch": 1.67,
-      "learning_rate": 3.880265143419065e-05,
-      "loss": 0.2932,
+      "epoch": 1.7133383789238477,
+      "grad_norm": 0.17309851944446564,
+      "learning_rate": 3.824901547213351e-05,
+      "loss": 0.3952,
       "step": 47540
     },
     {
-      "epoch": 1.67,
-      "learning_rate": 3.8800276155133044e-05,
-      "loss": 0.2958,
+      "epoch": 1.7135185785850724,
+      "grad_norm": 0.17767395079135895,
+      "learning_rate": 3.824654073960665e-05,
+      "loss": 0.379,
       "step": 47545
     },
     {
-      "epoch": 1.67,
-      "learning_rate": 3.879790069688795e-05,
-      "loss": 0.2934,
+      "epoch": 1.713698778246297,
+      "grad_norm": 0.19817490875720978,
+      "learning_rate": 3.824406582659824e-05,
+      "loss": 0.4156,
       "step": 47550
     },
     {
-      "epoch": 1.67,
-      "learning_rate": 3.8795525059486216e-05,
-      "loss": 0.284,
+      "epoch": 1.7138789779075214,
+      "grad_norm": 0.20314614474773407,
+      "learning_rate": 3.8241590733142004e-05,
+      "loss": 0.3943,
       "step": 47555
     },
     {
-      "epoch": 1.67,
-      "learning_rate": 3.879314924295867e-05,
-      "loss": 0.2561,
+      "epoch": 1.7140591775687462,
+      "grad_norm": 0.17916688323020935,
+      "learning_rate": 3.823911545927166e-05,
+      "loss": 0.3994,
       "step": 47560
     },
     {
-      "epoch": 1.67,
-      "learning_rate": 3.8790773247336176e-05,
-      "loss": 0.2807,
+      "epoch": 1.7142393772299709,
+      "grad_norm": 0.1822926551103592,
+      "learning_rate": 3.823664000502093e-05,
+      "loss": 0.4236,
       "step": 47565
     },
     {
-      "epoch": 1.67,
-      "learning_rate": 3.8788397072649584e-05,
-      "loss": 0.2702,
+      "epoch": 1.7144195768911954,
+      "grad_norm": 0.1775011420249939,
+      "learning_rate": 3.823416437042353e-05,
+      "loss": 0.4446,
       "step": 47570
     },
     {
-      "epoch": 1.67,
-      "learning_rate": 3.8786020718929737e-05,
-      "loss": 0.2759,
+      "epoch": 1.7145997765524201,
+      "grad_norm": 0.1689702570438385,
+      "learning_rate": 3.823168855551321e-05,
+      "loss": 0.3789,
       "step": 47575
     },
     {
-      "epoch": 1.67,
-      "learning_rate": 3.8783644186207494e-05,
-      "loss": 0.2952,
+      "epoch": 1.7147799762136446,
+      "grad_norm": 0.17035454511642456,
+      "learning_rate": 3.82292125603237e-05,
+      "loss": 0.4214,
       "step": 47580
     },
     {
-      "epoch": 1.67,
-      "learning_rate": 3.878126747451373e-05,
-      "loss": 0.2847,
+      "epoch": 1.7149601758748694,
+      "grad_norm": 0.17879173159599304,
+      "learning_rate": 3.822673638488873e-05,
+      "loss": 0.4425,
       "step": 47585
     },
     {
-      "epoch": 1.67,
-      "learning_rate": 3.877889058387927e-05,
-      "loss": 0.2939,
+      "epoch": 1.715140375536094,
+      "grad_norm": 0.2069813311100006,
+      "learning_rate": 3.8224260029242034e-05,
+      "loss": 0.4517,
       "step": 47590
     },
     {
-      "epoch": 1.67,
-      "learning_rate": 3.8776513514335014e-05,
-      "loss": 0.3119,
+      "epoch": 1.7153205751973186,
+      "grad_norm": 0.16887204349040985,
+      "learning_rate": 3.822178349341735e-05,
+      "loss": 0.4043,
       "step": 47595
     },
     {
-      "epoch": 1.67,
-      "learning_rate": 3.8774136265911795e-05,
-      "loss": 0.2785,
+      "epoch": 1.7155007748585431,
+      "grad_norm": 0.16069333255290985,
+      "learning_rate": 3.821930677744843e-05,
+      "loss": 0.4107,
       "step": 47600
     },
     {
-      "epoch": 1.67,
-      "learning_rate": 3.87717588386405e-05,
-      "loss": 0.3148,
+      "epoch": 1.7156809745197679,
+      "grad_norm": 0.1842743456363678,
+      "learning_rate": 3.821682988136902e-05,
+      "loss": 0.4391,
       "step": 47605
     },
     {
-      "epoch": 1.68,
-      "learning_rate": 3.876938123255199e-05,
-      "loss": 0.2824,
+      "epoch": 1.7158611741809926,
+      "grad_norm": 0.2043282687664032,
+      "learning_rate": 3.821435280521286e-05,
+      "loss": 0.4216,
       "step": 47610
     },
     {
-      "epoch": 1.68,
-      "learning_rate": 3.8767003447677145e-05,
-      "loss": 0.3029,
+      "epoch": 1.7160413738422173,
+      "grad_norm": 0.14843593537807465,
+      "learning_rate": 3.82118755490137e-05,
+      "loss": 0.3988,
       "step": 47615
     },
     {
-      "epoch": 1.68,
-      "learning_rate": 3.876462548404683e-05,
-      "loss": 0.2829,
+      "epoch": 1.7162215735034418,
+      "grad_norm": 0.17393402755260468,
+      "learning_rate": 3.820939811280528e-05,
+      "loss": 0.4165,
       "step": 47620
     },
     {
-      "epoch": 1.68,
-      "learning_rate": 3.876224734169193e-05,
-      "loss": 0.2836,
+      "epoch": 1.7164017731646664,
+      "grad_norm": 0.20571184158325195,
+      "learning_rate": 3.820692049662139e-05,
+      "loss": 0.3767,
       "step": 47625
     },
     {
-      "epoch": 1.68,
-      "learning_rate": 3.87598690206433e-05,
-      "loss": 0.2844,
+      "epoch": 1.716581972825891,
+      "grad_norm": 0.18904875218868256,
+      "learning_rate": 3.820444270049576e-05,
+      "loss": 0.4407,
       "step": 47630
     },
     {
-      "epoch": 1.68,
-      "learning_rate": 3.875749052093186e-05,
-      "loss": 0.2848,
+      "epoch": 1.7167621724871158,
+      "grad_norm": 0.16259777545928955,
+      "learning_rate": 3.820196472446215e-05,
+      "loss": 0.4126,
       "step": 47635
     },
     {
-      "epoch": 1.68,
-      "learning_rate": 3.875511184258845e-05,
-      "loss": 0.287,
+      "epoch": 1.7169423721483403,
+      "grad_norm": 0.19046054780483246,
+      "learning_rate": 3.819948656855432e-05,
+      "loss": 0.3814,
       "step": 47640
     },
     {
-      "epoch": 1.68,
-      "learning_rate": 3.8752732985644e-05,
-      "loss": 0.2968,
+      "epoch": 1.7171225718095648,
+      "grad_norm": 0.17224107682704926,
+      "learning_rate": 3.819700823280605e-05,
+      "loss": 0.4453,
       "step": 47645
     },
     {
-      "epoch": 1.68,
-      "learning_rate": 3.875035395012936e-05,
-      "loss": 0.3014,
+      "epoch": 1.7173027714707896,
+      "grad_norm": 0.22785985469818115,
+      "learning_rate": 3.8194529717251095e-05,
+      "loss": 0.3798,
       "step": 47650
     },
     {
-      "epoch": 1.68,
-      "learning_rate": 3.8747974736075444e-05,
-      "loss": 0.2758,
+      "epoch": 1.7174829711320143,
+      "grad_norm": 0.17245115339756012,
+      "learning_rate": 3.819205102192323e-05,
+      "loss": 0.3987,
       "step": 47655
     },
     {
-      "epoch": 1.68,
-      "learning_rate": 3.874559534351314e-05,
-      "loss": 0.2948,
+      "epoch": 1.717663170793239,
+      "grad_norm": 0.17877335846424103,
+      "learning_rate": 3.818957214685622e-05,
+      "loss": 0.3945,
       "step": 47660
     },
     {
-      "epoch": 1.68,
-      "learning_rate": 3.8743215772473326e-05,
-      "loss": 0.3045,
+      "epoch": 1.7178433704544636,
+      "grad_norm": 0.20267999172210693,
+      "learning_rate": 3.8187093092083845e-05,
+      "loss": 0.4068,
       "step": 47665
     },
     {
-      "epoch": 1.68,
-      "learning_rate": 3.874083602298693e-05,
-      "loss": 0.2855,
+      "epoch": 1.718023570115688,
+      "grad_norm": 0.1948103904724121,
+      "learning_rate": 3.818461385763988e-05,
+      "loss": 0.4051,
       "step": 47670
     },
     {
-      "epoch": 1.68,
-      "learning_rate": 3.873845609508482e-05,
-      "loss": 0.2766,
+      "epoch": 1.7182037697769128,
+      "grad_norm": 0.1381908804178238,
+      "learning_rate": 3.81821344435581e-05,
+      "loss": 0.4079,
       "step": 47675
     },
     {
-      "epoch": 1.68,
-      "learning_rate": 3.873607598879791e-05,
-      "loss": 0.3062,
+      "epoch": 1.7183839694381375,
+      "grad_norm": 0.16724207997322083,
+      "learning_rate": 3.81796548498723e-05,
+      "loss": 0.387,
       "step": 47680
     },
     {
-      "epoch": 1.68,
-      "learning_rate": 3.8733695704157115e-05,
-      "loss": 0.2835,
+      "epoch": 1.718564169099362,
+      "grad_norm": 0.19141383469104767,
+      "learning_rate": 3.817717507661625e-05,
+      "loss": 0.4073,
       "step": 47685
     },
     {
-      "epoch": 1.68,
-      "learning_rate": 3.8731315241193324e-05,
-      "loss": 0.3016,
+      "epoch": 1.7187443687605868,
+      "grad_norm": 0.21919883787631989,
+      "learning_rate": 3.8174695123823734e-05,
+      "loss": 0.4621,
       "step": 47690
     },
     {
-      "epoch": 1.68,
-      "learning_rate": 3.872893459993745e-05,
-      "loss": 0.2987,
+      "epoch": 1.7189245684218113,
+      "grad_norm": 0.1990518867969513,
+      "learning_rate": 3.8172214991528554e-05,
+      "loss": 0.4047,
       "step": 47695
     },
     {
-      "epoch": 1.68,
-      "learning_rate": 3.872655378042043e-05,
-      "loss": 0.2697,
+      "epoch": 1.719104768083036,
+      "grad_norm": 0.18987984955310822,
+      "learning_rate": 3.8169734679764494e-05,
+      "loss": 0.4128,
       "step": 47700
     },
     {
-      "epoch": 1.68,
-      "learning_rate": 3.872417278267313e-05,
-      "loss": 0.2787,
+      "epoch": 1.7192849677442608,
+      "grad_norm": 0.18330170214176178,
+      "learning_rate": 3.816725418856535e-05,
+      "loss": 0.4213,
       "step": 47705
     },
     {
-      "epoch": 1.68,
-      "learning_rate": 3.87217916067265e-05,
-      "loss": 0.2992,
+      "epoch": 1.7194651674054853,
+      "grad_norm": 0.1505930870771408,
+      "learning_rate": 3.816477351796491e-05,
+      "loss": 0.4264,
       "step": 47710
     },
     {
-      "epoch": 1.68,
-      "learning_rate": 3.871941025261146e-05,
-      "loss": 0.2912,
+      "epoch": 1.7196453670667098,
+      "grad_norm": 0.18332305550575256,
+      "learning_rate": 3.8162292667996986e-05,
+      "loss": 0.4012,
       "step": 47715
     },
     {
-      "epoch": 1.68,
-      "learning_rate": 3.8717028720358904e-05,
-      "loss": 0.2638,
+      "epoch": 1.7198255667279345,
+      "grad_norm": 0.21521466970443726,
+      "learning_rate": 3.815981163869537e-05,
+      "loss": 0.4301,
       "step": 47720
     },
     {
-      "epoch": 1.68,
-      "learning_rate": 3.871464700999978e-05,
-      "loss": 0.3052,
+      "epoch": 1.7200057663891593,
+      "grad_norm": 0.21533189713954926,
+      "learning_rate": 3.815733043009387e-05,
+      "loss": 0.4403,
       "step": 47725
     },
     {
-      "epoch": 1.68,
-      "learning_rate": 3.871226512156499e-05,
-      "loss": 0.2839,
+      "epoch": 1.720185966050384,
+      "grad_norm": 0.22986893355846405,
+      "learning_rate": 3.815484904222629e-05,
+      "loss": 0.4242,
       "step": 47730
     },
     {
-      "epoch": 1.68,
-      "learning_rate": 3.870988305508549e-05,
-      "loss": 0.291,
+      "epoch": 1.7203661657116085,
+      "grad_norm": 0.1726514995098114,
+      "learning_rate": 3.8152367475126436e-05,
+      "loss": 0.4429,
       "step": 47735
     },
     {
-      "epoch": 1.68,
-      "learning_rate": 3.870750081059219e-05,
-      "loss": 0.2973,
+      "epoch": 1.720546365372833,
+      "grad_norm": 0.18250474333763123,
+      "learning_rate": 3.814988572882813e-05,
+      "loss": 0.4109,
       "step": 47740
     },
     {
-      "epoch": 1.68,
-      "learning_rate": 3.870511838811601e-05,
-      "loss": 0.2885,
+      "epoch": 1.7207265650340577,
+      "grad_norm": 0.15575775504112244,
+      "learning_rate": 3.814740380336517e-05,
+      "loss": 0.3745,
       "step": 47745
     },
     {
-      "epoch": 1.68,
-      "learning_rate": 3.870273578768792e-05,
-      "loss": 0.2954,
+      "epoch": 1.7209067646952825,
+      "grad_norm": 0.23164531588554382,
+      "learning_rate": 3.814492169877138e-05,
+      "loss": 0.4285,
       "step": 47750
     },
     {
-      "epoch": 1.68,
-      "learning_rate": 3.870035300933883e-05,
-      "loss": 0.2994,
+      "epoch": 1.721086964356507,
+      "grad_norm": 0.18944543600082397,
+      "learning_rate": 3.814243941508058e-05,
+      "loss": 0.4425,
       "step": 47755
     },
     {
-      "epoch": 1.68,
-      "learning_rate": 3.869797005309969e-05,
-      "loss": 0.3017,
+      "epoch": 1.7212671640177315,
+      "grad_norm": 0.17386077344417572,
+      "learning_rate": 3.813995695232658e-05,
+      "loss": 0.4277,
       "step": 47760
     },
     {
-      "epoch": 1.68,
-      "learning_rate": 3.869558691900142e-05,
-      "loss": 0.3055,
+      "epoch": 1.7214473636789562,
+      "grad_norm": 0.16808000206947327,
+      "learning_rate": 3.813747431054321e-05,
+      "loss": 0.4155,
       "step": 47765
     },
     {
-      "epoch": 1.68,
-      "learning_rate": 3.869320360707499e-05,
-      "loss": 0.2589,
+      "epoch": 1.721627563340181,
+      "grad_norm": 0.21575501561164856,
+      "learning_rate": 3.8134991489764305e-05,
+      "loss": 0.4243,
       "step": 47770
     },
     {
-      "epoch": 1.68,
-      "learning_rate": 3.869082011735133e-05,
-      "loss": 0.2804,
+      "epoch": 1.7218077630014057,
+      "grad_norm": 0.16547565162181854,
+      "learning_rate": 3.8132508490023674e-05,
+      "loss": 0.4241,
       "step": 47775
     },
     {
-      "epoch": 1.68,
-      "learning_rate": 3.8688436449861406e-05,
-      "loss": 0.259,
+      "epoch": 1.7219879626626302,
+      "grad_norm": 0.17828123271465302,
+      "learning_rate": 3.813002531135517e-05,
+      "loss": 0.4356,
       "step": 47780
     },
     {
-      "epoch": 1.68,
-      "learning_rate": 3.8686052604636144e-05,
-      "loss": 0.2794,
+      "epoch": 1.7221681623238547,
+      "grad_norm": 0.21467016637325287,
+      "learning_rate": 3.81275419537926e-05,
+      "loss": 0.399,
       "step": 47785
     },
     {
-      "epoch": 1.68,
-      "learning_rate": 3.868366858170651e-05,
-      "loss": 0.285,
+      "epoch": 1.7223483619850795,
+      "grad_norm": 0.16447187960147858,
+      "learning_rate": 3.8125058417369824e-05,
+      "loss": 0.4181,
       "step": 47790
     },
     {
-      "epoch": 1.68,
-      "learning_rate": 3.868128438110346e-05,
-      "loss": 0.3079,
+      "epoch": 1.7225285616463042,
+      "grad_norm": 0.17198550701141357,
+      "learning_rate": 3.812257470212066e-05,
+      "loss": 0.3742,
       "step": 47795
     },
     {
-      "epoch": 1.68,
-      "learning_rate": 3.867890000285795e-05,
-      "loss": 0.2741,
+      "epoch": 1.7227087613075287,
+      "grad_norm": 0.1887580305337906,
+      "learning_rate": 3.812009080807896e-05,
+      "loss": 0.3888,
       "step": 47800
     },
     {
-      "epoch": 1.68,
-      "learning_rate": 3.867651544700093e-05,
-      "loss": 0.2971,
+      "epoch": 1.7228889609687532,
+      "grad_norm": 0.18961556255817413,
+      "learning_rate": 3.8117606735278556e-05,
+      "loss": 0.44,
       "step": 47805
     },
     {
-      "epoch": 1.68,
-      "learning_rate": 3.8674130713563376e-05,
-      "loss": 0.285,
+      "epoch": 1.723069160629978,
+      "grad_norm": 0.18733921647071838,
+      "learning_rate": 3.811512248375332e-05,
+      "loss": 0.4275,
       "step": 47810
     },
     {
-      "epoch": 1.68,
-      "learning_rate": 3.8671745802576254e-05,
-      "loss": 0.2907,
+      "epoch": 1.7232493602912027,
+      "grad_norm": 0.17053326964378357,
+      "learning_rate": 3.811263805353705e-05,
+      "loss": 0.4142,
       "step": 47815
     },
     {
-      "epoch": 1.68,
-      "learning_rate": 3.866936071407051e-05,
-      "loss": 0.2883,
+      "epoch": 1.7234295599524274,
+      "grad_norm": 0.2137068808078766,
+      "learning_rate": 3.811015344466365e-05,
+      "loss": 0.4018,
       "step": 47820
     },
     {
-      "epoch": 1.68,
-      "learning_rate": 3.8666975448077135e-05,
-      "loss": 0.2868,
+      "epoch": 1.723609759613652,
+      "grad_norm": 0.21045798063278198,
+      "learning_rate": 3.810766865716693e-05,
+      "loss": 0.4379,
       "step": 47825
     },
     {
-      "epoch": 1.68,
-      "learning_rate": 3.866459000462708e-05,
-      "loss": 0.3158,
+      "epoch": 1.7237899592748764,
+      "grad_norm": 0.2026238739490509,
+      "learning_rate": 3.810518369108077e-05,
+      "loss": 0.4311,
       "step": 47830
     },
     {
-      "epoch": 1.68,
-      "learning_rate": 3.866220438375134e-05,
-      "loss": 0.3045,
+      "epoch": 1.7239701589361012,
+      "grad_norm": 0.1852739304304123,
+      "learning_rate": 3.8102698546439025e-05,
+      "loss": 0.397,
       "step": 47835
     },
     {
-      "epoch": 1.68,
-      "learning_rate": 3.865981858548087e-05,
-      "loss": 0.2977,
+      "epoch": 1.724150358597326,
+      "grad_norm": 0.20483826100826263,
+      "learning_rate": 3.810021322327554e-05,
+      "loss": 0.4485,
       "step": 47840
     },
     {
-      "epoch": 1.68,
-      "learning_rate": 3.865743260984667e-05,
-      "loss": 0.2843,
+      "epoch": 1.7243305582585506,
+      "grad_norm": 0.18066450953483582,
+      "learning_rate": 3.809772772162419e-05,
+      "loss": 0.4222,
       "step": 47845
     },
     {
-      "epoch": 1.68,
-      "learning_rate": 3.865504645687969e-05,
-      "loss": 0.3085,
+      "epoch": 1.7245107579197752,
+      "grad_norm": 0.1802929937839508,
+      "learning_rate": 3.809524204151883e-05,
+      "loss": 0.4147,
       "step": 47850
     },
     {
-      "epoch": 1.68,
-      "learning_rate": 3.865266012661095e-05,
-      "loss": 0.3038,
+      "epoch": 1.7246909575809997,
+      "grad_norm": 0.177862748503685,
+      "learning_rate": 3.809275618299335e-05,
+      "loss": 0.445,
       "step": 47855
     },
     {
-      "epoch": 1.68,
-      "learning_rate": 3.865027361907141e-05,
-      "loss": 0.2652,
+      "epoch": 1.7248711572422244,
+      "grad_norm": 0.1601111888885498,
+      "learning_rate": 3.809027014608159e-05,
+      "loss": 0.424,
       "step": 47860
     },
     {
-      "epoch": 1.68,
-      "learning_rate": 3.864788693429206e-05,
-      "loss": 0.2831,
+      "epoch": 1.7250513569034491,
+      "grad_norm": 0.1917547881603241,
+      "learning_rate": 3.808778393081742e-05,
+      "loss": 0.4559,
       "step": 47865
     },
     {
-      "epoch": 1.68,
-      "learning_rate": 3.8645500072303886e-05,
-      "loss": 0.2692,
+      "epoch": 1.7252315565646736,
+      "grad_norm": 0.1599884331226349,
+      "learning_rate": 3.808529753723475e-05,
+      "loss": 0.4261,
       "step": 47870
     },
     {
-      "epoch": 1.68,
-      "learning_rate": 3.86431130331379e-05,
-      "loss": 0.2932,
+      "epoch": 1.7254117562258982,
+      "grad_norm": 0.1725659966468811,
+      "learning_rate": 3.808281096536742e-05,
+      "loss": 0.4202,
       "step": 47875
     },
     {
-      "epoch": 1.68,
-      "learning_rate": 3.8640725816825075e-05,
-      "loss": 0.2795,
+      "epoch": 1.7255919558871229,
+      "grad_norm": 0.177949458360672,
+      "learning_rate": 3.808032421524933e-05,
+      "loss": 0.431,
       "step": 47880
     },
     {
-      "epoch": 1.68,
-      "learning_rate": 3.8638338423396415e-05,
-      "loss": 0.2969,
+      "epoch": 1.7257721555483476,
+      "grad_norm": 0.20164379477500916,
+      "learning_rate": 3.807783728691435e-05,
+      "loss": 0.3792,
       "step": 47885
     },
     {
-      "epoch": 1.68,
-      "learning_rate": 3.863595085288292e-05,
-      "loss": 0.2695,
+      "epoch": 1.7259523552095724,
+      "grad_norm": 0.16568000614643097,
+      "learning_rate": 3.8075350180396376e-05,
+      "loss": 0.4351,
       "step": 47890
     },
     {
-      "epoch": 1.69,
-      "learning_rate": 3.8633563105315585e-05,
-      "loss": 0.2813,
+      "epoch": 1.7261325548707969,
+      "grad_norm": 0.1772766262292862,
+      "learning_rate": 3.807286289572929e-05,
+      "loss": 0.4139,
       "step": 47895
     },
     {
-      "epoch": 1.69,
-      "learning_rate": 3.863117518072542e-05,
-      "loss": 0.2802,
+      "epoch": 1.7263127545320214,
+      "grad_norm": 0.20613396167755127,
+      "learning_rate": 3.8070375432946965e-05,
+      "loss": 0.4108,
       "step": 47900
     },
     {
-      "epoch": 1.69,
-      "learning_rate": 3.8628787079143434e-05,
-      "loss": 0.2588,
+      "epoch": 1.726492954193246,
+      "grad_norm": 0.1459551453590393,
+      "learning_rate": 3.806788779208331e-05,
+      "loss": 0.4489,
       "step": 47905
     },
     {
-      "epoch": 1.69,
-      "learning_rate": 3.862639880060063e-05,
-      "loss": 0.298,
+      "epoch": 1.7266731538544708,
+      "grad_norm": 0.20022889971733093,
+      "learning_rate": 3.806539997317221e-05,
+      "loss": 0.438,
       "step": 47910
     },
     {
-      "epoch": 1.69,
-      "learning_rate": 3.8624010345128015e-05,
-      "loss": 0.2723,
+      "epoch": 1.7268533535156954,
+      "grad_norm": 0.1898205578327179,
+      "learning_rate": 3.806291197624758e-05,
+      "loss": 0.4088,
       "step": 47915
     },
     {
-      "epoch": 1.69,
-      "learning_rate": 3.8621621712756596e-05,
-      "loss": 0.2967,
+      "epoch": 1.7270335531769199,
+      "grad_norm": 0.18108853697776794,
+      "learning_rate": 3.8060423801343294e-05,
+      "loss": 0.4036,
       "step": 47920
     },
     {
-      "epoch": 1.69,
-      "learning_rate": 3.86192329035174e-05,
-      "loss": 0.2764,
+      "epoch": 1.7272137528381446,
+      "grad_norm": 0.21101535856723785,
+      "learning_rate": 3.805793544849326e-05,
+      "loss": 0.4368,
       "step": 47925
     },
     {
-      "epoch": 1.69,
-      "learning_rate": 3.8616843917441447e-05,
-      "loss": 0.2625,
+      "epoch": 1.7273939524993693,
+      "grad_norm": 0.16363751888275146,
+      "learning_rate": 3.8055446917731386e-05,
+      "loss": 0.4097,
       "step": 47930
     },
     {
-      "epoch": 1.69,
-      "learning_rate": 3.861445475455974e-05,
-      "loss": 0.3208,
+      "epoch": 1.727574152160594,
+      "grad_norm": 0.21080221235752106,
+      "learning_rate": 3.805295820909158e-05,
+      "loss": 0.4326,
       "step": 47935
     },
     {
-      "epoch": 1.69,
-      "learning_rate": 3.861206541490332e-05,
-      "loss": 0.2718,
+      "epoch": 1.7277543518218186,
+      "grad_norm": 0.18741890788078308,
+      "learning_rate": 3.805046932260774e-05,
+      "loss": 0.4329,
       "step": 47940
     },
     {
-      "epoch": 1.69,
-      "learning_rate": 3.86096758985032e-05,
-      "loss": 0.2794,
+      "epoch": 1.727934551483043,
+      "grad_norm": 0.20189730823040009,
+      "learning_rate": 3.804798025831379e-05,
+      "loss": 0.4618,
       "step": 47945
     },
     {
-      "epoch": 1.69,
-      "learning_rate": 3.86072862053904e-05,
-      "loss": 0.2741,
+      "epoch": 1.7281147511442678,
+      "grad_norm": 0.16234025359153748,
+      "learning_rate": 3.804549101624362e-05,
+      "loss": 0.3968,
       "step": 47950
     },
     {
-      "epoch": 1.69,
-      "learning_rate": 3.860489633559596e-05,
-      "loss": 0.2908,
+      "epoch": 1.7282949508054926,
+      "grad_norm": 0.17763473093509674,
+      "learning_rate": 3.804300159643117e-05,
+      "loss": 0.3923,
       "step": 47955
     },
     {
-      "epoch": 1.69,
-      "learning_rate": 3.860250628915091e-05,
-      "loss": 0.2994,
+      "epoch": 1.728475150466717,
+      "grad_norm": 0.20635025203227997,
+      "learning_rate": 3.804051199891035e-05,
+      "loss": 0.3992,
       "step": 47960
     },
     {
-      "epoch": 1.69,
-      "learning_rate": 3.860011606608628e-05,
-      "loss": 0.3038,
+      "epoch": 1.7286553501279418,
+      "grad_norm": 0.1754833310842514,
+      "learning_rate": 3.803802222371507e-05,
+      "loss": 0.4374,
       "step": 47965
     },
     {
-      "epoch": 1.69,
-      "learning_rate": 3.85977256664331e-05,
-      "loss": 0.2753,
+      "epoch": 1.7288355497891663,
+      "grad_norm": 0.17620554566383362,
+      "learning_rate": 3.803553227087928e-05,
+      "loss": 0.409,
       "step": 47970
     },
     {
-      "epoch": 1.69,
-      "learning_rate": 3.8595335090222414e-05,
-      "loss": 0.286,
+      "epoch": 1.729015749450391,
+      "grad_norm": 0.1946227103471756,
+      "learning_rate": 3.803304214043687e-05,
+      "loss": 0.4026,
       "step": 47975
     },
     {
-      "epoch": 1.69,
-      "learning_rate": 3.859294433748527e-05,
-      "loss": 0.297,
+      "epoch": 1.7291959491116158,
+      "grad_norm": 0.22340406477451324,
+      "learning_rate": 3.803055183242179e-05,
+      "loss": 0.3881,
       "step": 47980
     },
     {
-      "epoch": 1.69,
-      "learning_rate": 3.85905534082527e-05,
-      "loss": 0.2689,
+      "epoch": 1.7293761487728403,
+      "grad_norm": 0.16771641373634338,
+      "learning_rate": 3.8028061346867963e-05,
+      "loss": 0.4304,
       "step": 47985
     },
     {
-      "epoch": 1.69,
-      "learning_rate": 3.8588162302555744e-05,
-      "loss": 0.2753,
+      "epoch": 1.7295563484340648,
+      "grad_norm": 0.16595138609409332,
+      "learning_rate": 3.802557068380932e-05,
+      "loss": 0.4088,
       "step": 47990
     },
     {
-      "epoch": 1.69,
-      "learning_rate": 3.8585771020425454e-05,
-      "loss": 0.2754,
+      "epoch": 1.7297365480952895,
+      "grad_norm": 0.22411175072193146,
+      "learning_rate": 3.80230798432798e-05,
+      "loss": 0.4289,
       "step": 47995
     },
     {
-      "epoch": 1.69,
-      "learning_rate": 3.858337956189289e-05,
-      "loss": 0.2888,
+      "epoch": 1.7299167477565143,
+      "grad_norm": 0.2403663694858551,
+      "learning_rate": 3.802058882531334e-05,
+      "loss": 0.4267,
       "step": 48000
     },
     {
-      "epoch": 1.69,
-      "eval_loss": 0.2754231095314026,
-      "eval_runtime": 10.5474,
-      "eval_samples_per_second": 9.481,
-      "eval_steps_per_second": 9.481,
+      "epoch": 1.7299167477565143,
+      "eval_loss": 0.44221970438957214,
+      "eval_runtime": 3.5221,
+      "eval_samples_per_second": 28.392,
+      "eval_steps_per_second": 7.098,
       "step": 48000
     },
     {
-      "epoch": 1.69,
-      "learning_rate": 3.858098792698909e-05,
-      "loss": 0.2768,
+      "epoch": 1.730096947417739,
+      "grad_norm": 0.1950840801000595,
+      "learning_rate": 3.801809762994387e-05,
+      "loss": 0.4326,
       "step": 48005
     },
     {
-      "epoch": 1.69,
-      "learning_rate": 3.8578596115745114e-05,
-      "loss": 0.2832,
+      "epoch": 1.7302771470789635,
+      "grad_norm": 0.18309806287288666,
+      "learning_rate": 3.801560625720535e-05,
+      "loss": 0.4377,
       "step": 48010
     },
     {
-      "epoch": 1.69,
-      "learning_rate": 3.857620412819201e-05,
-      "loss": 0.2764,
+      "epoch": 1.730457346740188,
+      "grad_norm": 0.17351049184799194,
+      "learning_rate": 3.8013114707131716e-05,
+      "loss": 0.3877,
       "step": 48015
     },
     {
-      "epoch": 1.69,
-      "learning_rate": 3.8573811964360844e-05,
-      "loss": 0.2743,
+      "epoch": 1.7306375464014128,
+      "grad_norm": 0.18052922189235687,
+      "learning_rate": 3.80106229797569e-05,
+      "loss": 0.3962,
       "step": 48020
     },
     {
-      "epoch": 1.69,
-      "learning_rate": 3.8571419624282676e-05,
-      "loss": 0.2823,
+      "epoch": 1.7308177460626375,
+      "grad_norm": 0.17976753413677216,
+      "learning_rate": 3.8008131075114886e-05,
+      "loss": 0.3876,
       "step": 48025
     },
     {
-      "epoch": 1.69,
-      "learning_rate": 3.856902710798857e-05,
-      "loss": 0.2791,
+      "epoch": 1.730997945723862,
+      "grad_norm": 0.1913975179195404,
+      "learning_rate": 3.800563899323959e-05,
+      "loss": 0.3877,
       "step": 48030
     },
     {
-      "epoch": 1.69,
-      "learning_rate": 3.8566634415509584e-05,
-      "loss": 0.2706,
+      "epoch": 1.7311781453850865,
+      "grad_norm": 0.19166803359985352,
+      "learning_rate": 3.800314673416498e-05,
+      "loss": 0.4104,
       "step": 48035
     },
     {
-      "epoch": 1.69,
-      "learning_rate": 3.85642415468768e-05,
-      "loss": 0.2822,
+      "epoch": 1.7313583450463113,
+      "grad_norm": 0.19192728400230408,
+      "learning_rate": 3.800065429792501e-05,
+      "loss": 0.3646,
       "step": 48040
     },
     {
-      "epoch": 1.69,
-      "learning_rate": 3.856184850212128e-05,
-      "loss": 0.2911,
+      "epoch": 1.731538544707536,
+      "grad_norm": 0.21600832045078278,
+      "learning_rate": 3.7998161684553656e-05,
+      "loss": 0.4372,
       "step": 48045
     },
     {
-      "epoch": 1.69,
-      "learning_rate": 3.855945528127408e-05,
-      "loss": 0.26,
+      "epoch": 1.7317187443687607,
+      "grad_norm": 0.20406126976013184,
+      "learning_rate": 3.799566889408486e-05,
+      "loss": 0.4218,
       "step": 48050
     },
     {
-      "epoch": 1.69,
-      "learning_rate": 3.8557061884366294e-05,
-      "loss": 0.2653,
+      "epoch": 1.7318989440299852,
+      "grad_norm": 0.17205806076526642,
+      "learning_rate": 3.7993175926552596e-05,
+      "loss": 0.3918,
       "step": 48055
     },
     {
-      "epoch": 1.69,
-      "learning_rate": 3.8554668311429e-05,
-      "loss": 0.3004,
+      "epoch": 1.7320791436912097,
+      "grad_norm": 0.15462535619735718,
+      "learning_rate": 3.799068278199081e-05,
+      "loss": 0.3719,
       "step": 48060
     },
     {
-      "epoch": 1.69,
-      "learning_rate": 3.855227456249327e-05,
-      "loss": 0.2668,
+      "epoch": 1.7322593433524345,
+      "grad_norm": 0.2172221541404724,
+      "learning_rate": 3.7988189460433496e-05,
+      "loss": 0.4315,
       "step": 48065
     },
     {
-      "epoch": 1.69,
-      "learning_rate": 3.854988063759019e-05,
-      "loss": 0.2933,
+      "epoch": 1.7324395430136592,
+      "grad_norm": 0.23806485533714294,
+      "learning_rate": 3.7985695961914614e-05,
+      "loss": 0.4457,
       "step": 48070
     },
     {
-      "epoch": 1.69,
-      "learning_rate": 3.854748653675083e-05,
-      "loss": 0.3116,
+      "epoch": 1.7326197426748837,
+      "grad_norm": 0.19144880771636963,
+      "learning_rate": 3.798320228646814e-05,
+      "loss": 0.4054,
       "step": 48075
     },
     {
-      "epoch": 1.69,
-      "learning_rate": 3.8545092260006287e-05,
-      "loss": 0.2593,
+      "epoch": 1.7327999423361085,
+      "grad_norm": 0.13344308733940125,
+      "learning_rate": 3.798070843412805e-05,
+      "loss": 0.4203,
       "step": 48080
     },
     {
-      "epoch": 1.69,
-      "learning_rate": 3.8542697807387656e-05,
-      "loss": 0.2967,
+      "epoch": 1.732980141997333,
+      "grad_norm": 0.17505653202533722,
+      "learning_rate": 3.7978214404928315e-05,
+      "loss": 0.4313,
       "step": 48085
     },
     {
-      "epoch": 1.69,
-      "learning_rate": 3.854030317892601e-05,
-      "loss": 0.2844,
+      "epoch": 1.7331603416585577,
+      "grad_norm": 0.15916766226291656,
+      "learning_rate": 3.7975720198902924e-05,
+      "loss": 0.3778,
       "step": 48090
     },
     {
-      "epoch": 1.69,
-      "learning_rate": 3.8537908374652455e-05,
-      "loss": 0.3032,
+      "epoch": 1.7333405413197824,
+      "grad_norm": 0.2202104777097702,
+      "learning_rate": 3.797322581608585e-05,
+      "loss": 0.4379,
       "step": 48095
     },
     {
-      "epoch": 1.69,
-      "learning_rate": 3.853551339459808e-05,
-      "loss": 0.2862,
+      "epoch": 1.733520740981007,
+      "grad_norm": 0.17521396279335022,
+      "learning_rate": 3.7970731256511104e-05,
+      "loss": 0.4141,
       "step": 48100
     },
     {
-      "epoch": 1.69,
-      "learning_rate": 3.8533118238793985e-05,
-      "loss": 0.2812,
+      "epoch": 1.7337009406422315,
+      "grad_norm": 0.1417665183544159,
+      "learning_rate": 3.796823652021265e-05,
+      "loss": 0.3653,
       "step": 48105
     },
     {
-      "epoch": 1.69,
-      "learning_rate": 3.853072290727127e-05,
-      "loss": 0.2666,
+      "epoch": 1.7338811403034562,
+      "grad_norm": 0.20344464480876923,
+      "learning_rate": 3.796574160722448e-05,
+      "loss": 0.4285,
       "step": 48110
     },
     {
-      "epoch": 1.69,
-      "learning_rate": 3.852832740006103e-05,
-      "loss": 0.2676,
+      "epoch": 1.734061339964681,
+      "grad_norm": 0.1791670322418213,
+      "learning_rate": 3.796324651758059e-05,
+      "loss": 0.4186,
       "step": 48115
     },
     {
-      "epoch": 1.69,
-      "learning_rate": 3.852593171719437e-05,
-      "loss": 0.2995,
+      "epoch": 1.7342415396259057,
+      "grad_norm": 0.19157084822654724,
+      "learning_rate": 3.796075125131498e-05,
+      "loss": 0.4068,
       "step": 48120
     },
     {
-      "epoch": 1.69,
-      "learning_rate": 3.852353585870242e-05,
-      "loss": 0.297,
+      "epoch": 1.7344217392871302,
+      "grad_norm": 0.19911465048789978,
+      "learning_rate": 3.795825580846164e-05,
+      "loss": 0.4383,
       "step": 48125
     },
     {
-      "epoch": 1.69,
-      "learning_rate": 3.852113982461625e-05,
-      "loss": 0.2822,
+      "epoch": 1.7346019389483547,
+      "grad_norm": 0.1901884824037552,
+      "learning_rate": 3.7955760189054566e-05,
+      "loss": 0.4327,
       "step": 48130
     },
     {
-      "epoch": 1.69,
-      "learning_rate": 3.8518743614967e-05,
-      "loss": 0.2708,
+      "epoch": 1.7347821386095794,
+      "grad_norm": 0.17567721009254456,
+      "learning_rate": 3.7953264393127774e-05,
+      "loss": 0.406,
       "step": 48135
     },
     {
-      "epoch": 1.69,
-      "learning_rate": 3.8516347229785765e-05,
-      "loss": 0.2885,
+      "epoch": 1.7349623382708041,
+      "grad_norm": 0.20942769944667816,
+      "learning_rate": 3.795076842071526e-05,
+      "loss": 0.4276,
       "step": 48140
     },
     {
-      "epoch": 1.69,
-      "learning_rate": 3.8513950669103676e-05,
-      "loss": 0.2671,
+      "epoch": 1.7351425379320287,
+      "grad_norm": 0.18025130033493042,
+      "learning_rate": 3.794827227185105e-05,
+      "loss": 0.4079,
       "step": 48145
     },
     {
-      "epoch": 1.69,
-      "learning_rate": 3.851155393295184e-05,
-      "loss": 0.2828,
+      "epoch": 1.7353227375932532,
+      "grad_norm": 0.21213695406913757,
+      "learning_rate": 3.794577594656912e-05,
+      "loss": 0.4179,
       "step": 48150
     },
     {
-      "epoch": 1.69,
-      "learning_rate": 3.850915702136139e-05,
-      "loss": 0.2785,
+      "epoch": 1.735502937254478,
+      "grad_norm": 0.18721318244934082,
+      "learning_rate": 3.79432794449035e-05,
+      "loss": 0.4373,
       "step": 48155
     },
     {
-      "epoch": 1.69,
-      "learning_rate": 3.850675993436342e-05,
-      "loss": 0.266,
+      "epoch": 1.7356831369157026,
+      "grad_norm": 0.2057793140411377,
+      "learning_rate": 3.794078276688822e-05,
+      "loss": 0.3952,
       "step": 48160
     },
     {
-      "epoch": 1.69,
-      "learning_rate": 3.850436267198909e-05,
-      "loss": 0.28,
+      "epoch": 1.7358633365769274,
+      "grad_norm": 0.1834101676940918,
+      "learning_rate": 3.7938285912557256e-05,
+      "loss": 0.4069,
       "step": 48165
     },
     {
-      "epoch": 1.69,
-      "learning_rate": 3.85019652342695e-05,
-      "loss": 0.2823,
+      "epoch": 1.7360435362381519,
+      "grad_norm": 0.21136592328548431,
+      "learning_rate": 3.793578888194467e-05,
+      "loss": 0.4132,
       "step": 48170
     },
     {
-      "epoch": 1.69,
-      "learning_rate": 3.84995676212358e-05,
-      "loss": 0.299,
+      "epoch": 1.7362237358993764,
+      "grad_norm": 0.1626768410205841,
+      "learning_rate": 3.793329167508445e-05,
+      "loss": 0.359,
       "step": 48175
     },
     {
-      "epoch": 1.7,
-      "learning_rate": 3.8497169832919105e-05,
-      "loss": 0.2796,
+      "epoch": 1.7364039355606011,
+      "grad_norm": 0.1921762079000473,
+      "learning_rate": 3.7930794292010654e-05,
+      "loss": 0.4221,
       "step": 48180
     },
     {
-      "epoch": 1.7,
-      "learning_rate": 3.8494771869350555e-05,
-      "loss": 0.2853,
+      "epoch": 1.7365841352218259,
+      "grad_norm": 0.17365401983261108,
+      "learning_rate": 3.7928296732757276e-05,
+      "loss": 0.4299,
       "step": 48185
     },
     {
-      "epoch": 1.7,
-      "learning_rate": 3.849237373056128e-05,
-      "loss": 0.2868,
+      "epoch": 1.7367643348830504,
+      "grad_norm": 0.18681730329990387,
+      "learning_rate": 3.7925798997358354e-05,
+      "loss": 0.4318,
       "step": 48190
     },
     {
-      "epoch": 1.7,
-      "learning_rate": 3.8489975416582436e-05,
-      "loss": 0.2774,
+      "epoch": 1.736944534544275,
+      "grad_norm": 0.15825678408145905,
+      "learning_rate": 3.792330108584793e-05,
+      "loss": 0.3875,
       "step": 48195
     },
     {
-      "epoch": 1.7,
-      "learning_rate": 3.848757692744515e-05,
-      "loss": 0.2843,
+      "epoch": 1.7371247342054996,
+      "grad_norm": 0.17392268776893616,
+      "learning_rate": 3.792080299826003e-05,
+      "loss": 0.4145,
       "step": 48200
     },
     {
-      "epoch": 1.7,
-      "learning_rate": 3.848517826318055e-05,
-      "loss": 0.2581,
+      "epoch": 1.7373049338667244,
+      "grad_norm": 0.17445778846740723,
+      "learning_rate": 3.79183047346287e-05,
+      "loss": 0.3871,
       "step": 48205
     },
     {
-      "epoch": 1.7,
-      "learning_rate": 3.848277942381981e-05,
-      "loss": 0.3146,
+      "epoch": 1.737485133527949,
+      "grad_norm": 0.18119192123413086,
+      "learning_rate": 3.7915806294987955e-05,
+      "loss": 0.3867,
       "step": 48210
     },
     {
-      "epoch": 1.7,
-      "learning_rate": 3.8480380409394066e-05,
-      "loss": 0.298,
+      "epoch": 1.7376653331891736,
+      "grad_norm": 0.20309992134571075,
+      "learning_rate": 3.7913307679371856e-05,
+      "loss": 0.4133,
       "step": 48215
     },
     {
-      "epoch": 1.7,
-      "learning_rate": 3.8477981219934464e-05,
-      "loss": 0.2889,
+      "epoch": 1.7378455328503981,
+      "grad_norm": 0.1670549362897873,
+      "learning_rate": 3.791080888781444e-05,
+      "loss": 0.405,
       "step": 48220
     },
     {
-      "epoch": 1.7,
-      "learning_rate": 3.8475581855472156e-05,
-      "loss": 0.2909,
+      "epoch": 1.7380257325116228,
+      "grad_norm": 0.162832111120224,
+      "learning_rate": 3.790830992034974e-05,
+      "loss": 0.3856,
       "step": 48225
     },
     {
-      "epoch": 1.7,
-      "learning_rate": 3.8473182316038306e-05,
-      "loss": 0.3038,
+      "epoch": 1.7382059321728476,
+      "grad_norm": 0.16759029030799866,
+      "learning_rate": 3.7905810777011837e-05,
+      "loss": 0.3806,
       "step": 48230
     },
     {
-      "epoch": 1.7,
-      "learning_rate": 3.847078260166406e-05,
-      "loss": 0.3052,
+      "epoch": 1.7383861318340723,
+      "grad_norm": 0.18382543325424194,
+      "learning_rate": 3.790331145783474e-05,
+      "loss": 0.3809,
       "step": 48235
     },
     {
-      "epoch": 1.7,
-      "learning_rate": 3.846838271238058e-05,
-      "loss": 0.2841,
+      "epoch": 1.7385663314952968,
+      "grad_norm": 0.18206432461738586,
+      "learning_rate": 3.7900811962852544e-05,
+      "loss": 0.4206,
       "step": 48240
     },
     {
-      "epoch": 1.7,
-      "learning_rate": 3.846598264821901e-05,
-      "loss": 0.2738,
+      "epoch": 1.7387465311565213,
+      "grad_norm": 0.17950773239135742,
+      "learning_rate": 3.789831229209927e-05,
+      "loss": 0.436,
       "step": 48245
     },
     {
-      "epoch": 1.7,
-      "learning_rate": 3.8463582409210554e-05,
-      "loss": 0.3019,
+      "epoch": 1.738926730817746,
+      "grad_norm": 0.16286823153495789,
+      "learning_rate": 3.7895812445608994e-05,
+      "loss": 0.4264,
       "step": 48250
     },
     {
-      "epoch": 1.7,
-      "learning_rate": 3.8461181995386335e-05,
-      "loss": 0.2845,
+      "epoch": 1.7391069304789708,
+      "grad_norm": 0.18310612440109253,
+      "learning_rate": 3.789331242341576e-05,
+      "loss": 0.4107,
       "step": 48255
     },
     {
-      "epoch": 1.7,
-      "learning_rate": 3.845878140677756e-05,
-      "loss": 0.3173,
+      "epoch": 1.7392871301401953,
+      "grad_norm": 0.17169633507728577,
+      "learning_rate": 3.789081222555365e-05,
+      "loss": 0.4303,
       "step": 48260
     },
     {
-      "epoch": 1.7,
-      "learning_rate": 3.845638064341536e-05,
-      "loss": 0.269,
+      "epoch": 1.7394673298014198,
+      "grad_norm": 0.20822210609912872,
+      "learning_rate": 3.7888311852056725e-05,
+      "loss": 0.409,
       "step": 48265
     },
     {
-      "epoch": 1.7,
-      "learning_rate": 3.8453979705330926e-05,
-      "loss": 0.2784,
+      "epoch": 1.7396475294626446,
+      "grad_norm": 0.18613308668136597,
+      "learning_rate": 3.788581130295903e-05,
+      "loss": 0.3899,
       "step": 48270
     },
     {
-      "epoch": 1.7,
-      "learning_rate": 3.845157859255543e-05,
-      "loss": 0.2945,
+      "epoch": 1.7398277291238693,
+      "grad_norm": 0.1604495495557785,
+      "learning_rate": 3.788331057829466e-05,
+      "loss": 0.413,
       "step": 48275
     },
     {
-      "epoch": 1.7,
-      "learning_rate": 3.844917730512005e-05,
-      "loss": 0.3041,
+      "epoch": 1.740007928785094,
+      "grad_norm": 0.18513786792755127,
+      "learning_rate": 3.788080967809767e-05,
+      "loss": 0.4236,
       "step": 48280
     },
     {
-      "epoch": 1.7,
-      "learning_rate": 3.844677584305598e-05,
-      "loss": 0.2842,
+      "epoch": 1.7401881284463185,
+      "grad_norm": 0.20122095942497253,
+      "learning_rate": 3.7878308602402156e-05,
+      "loss": 0.3951,
       "step": 48285
     },
     {
-      "epoch": 1.7,
-      "learning_rate": 3.844437420639439e-05,
-      "loss": 0.2791,
+      "epoch": 1.740368328107543,
+      "grad_norm": 0.22590294480323792,
+      "learning_rate": 3.787580735124217e-05,
+      "loss": 0.4287,
       "step": 48290
     },
     {
-      "epoch": 1.7,
-      "learning_rate": 3.844197239516645e-05,
-      "loss": 0.2651,
+      "epoch": 1.7405485277687678,
+      "grad_norm": 0.15910333395004272,
+      "learning_rate": 3.78733059246518e-05,
+      "loss": 0.4064,
       "step": 48295
     },
     {
-      "epoch": 1.7,
-      "learning_rate": 3.8439570409403355e-05,
-      "loss": 0.2748,
+      "epoch": 1.7407287274299925,
+      "grad_norm": 0.20325346291065216,
+      "learning_rate": 3.787080432266514e-05,
+      "loss": 0.415,
       "step": 48300
     },
     {
-      "epoch": 1.7,
-      "learning_rate": 3.8437168249136294e-05,
-      "loss": 0.302,
+      "epoch": 1.740908927091217,
+      "grad_norm": 0.16542761027812958,
+      "learning_rate": 3.786830254531626e-05,
+      "loss": 0.4409,
       "step": 48305
     },
     {
-      "epoch": 1.7,
-      "learning_rate": 3.8434765914396465e-05,
-      "loss": 0.259,
+      "epoch": 1.7410891267524415,
+      "grad_norm": 0.19882872700691223,
+      "learning_rate": 3.7865800592639245e-05,
+      "loss": 0.4755,
       "step": 48310
     },
     {
-      "epoch": 1.7,
-      "learning_rate": 3.843236340521505e-05,
-      "loss": 0.2918,
+      "epoch": 1.7412693264136663,
+      "grad_norm": 0.17635419964790344,
+      "learning_rate": 3.786329846466818e-05,
+      "loss": 0.3941,
       "step": 48315
     },
     {
-      "epoch": 1.7,
-      "learning_rate": 3.8429960721623255e-05,
-      "loss": 0.2759,
+      "epoch": 1.741449526074891,
+      "grad_norm": 0.22185000777244568,
+      "learning_rate": 3.786079616143718e-05,
+      "loss": 0.4116,
       "step": 48320
     },
     {
-      "epoch": 1.7,
-      "learning_rate": 3.842755786365227e-05,
-      "loss": 0.3081,
+      "epoch": 1.7416297257361157,
+      "grad_norm": 0.21445868909358978,
+      "learning_rate": 3.7858293682980315e-05,
+      "loss": 0.4251,
       "step": 48325
     },
     {
-      "epoch": 1.7,
-      "learning_rate": 3.8425154831333287e-05,
-      "loss": 0.2736,
+      "epoch": 1.7418099253973403,
+      "grad_norm": 0.18230614066123962,
+      "learning_rate": 3.785579102933168e-05,
+      "loss": 0.4337,
       "step": 48330
     },
     {
-      "epoch": 1.7,
-      "learning_rate": 3.8422751624697526e-05,
-      "loss": 0.2834,
+      "epoch": 1.7419901250585648,
+      "grad_norm": 0.18611697852611542,
+      "learning_rate": 3.7853288200525394e-05,
+      "loss": 0.4349,
       "step": 48335
     },
     {
-      "epoch": 1.7,
-      "learning_rate": 3.842034824377617e-05,
-      "loss": 0.2849,
+      "epoch": 1.7421703247197895,
+      "grad_norm": 0.17176128923892975,
+      "learning_rate": 3.785078519659554e-05,
+      "loss": 0.4123,
       "step": 48340
     },
     {
-      "epoch": 1.7,
-      "learning_rate": 3.8417944688600446e-05,
-      "loss": 0.2805,
+      "epoch": 1.7423505243810142,
+      "grad_norm": 0.18628238141536713,
+      "learning_rate": 3.784828201757623e-05,
+      "loss": 0.417,
       "step": 48345
     },
     {
-      "epoch": 1.7,
-      "learning_rate": 3.8415540959201544e-05,
-      "loss": 0.308,
+      "epoch": 1.742530724042239,
+      "grad_norm": 0.17633478343486786,
+      "learning_rate": 3.784577866350155e-05,
+      "loss": 0.4247,
       "step": 48350
     },
     {
-      "epoch": 1.7,
-      "learning_rate": 3.8413137055610696e-05,
-      "loss": 0.259,
+      "epoch": 1.7427109237034635,
+      "grad_norm": 0.17602220177650452,
+      "learning_rate": 3.7843275134405645e-05,
+      "loss": 0.4179,
       "step": 48355
     },
     {
-      "epoch": 1.7,
-      "learning_rate": 3.8410732977859086e-05,
-      "loss": 0.2816,
+      "epoch": 1.742891123364688,
+      "grad_norm": 0.15746374428272247,
+      "learning_rate": 3.7840771430322586e-05,
+      "loss": 0.4164,
       "step": 48360
     },
     {
-      "epoch": 1.7,
-      "learning_rate": 3.8408328725977966e-05,
-      "loss": 0.2959,
+      "epoch": 1.7430713230259127,
+      "grad_norm": 0.22847017645835876,
+      "learning_rate": 3.7838267551286504e-05,
+      "loss": 0.4151,
       "step": 48365
     },
     {
-      "epoch": 1.7,
-      "learning_rate": 3.840592429999852e-05,
-      "loss": 0.2828,
+      "epoch": 1.7432515226871375,
+      "grad_norm": 0.19369705021381378,
+      "learning_rate": 3.783576349733152e-05,
+      "loss": 0.4186,
       "step": 48370
     },
     {
-      "epoch": 1.7,
-      "learning_rate": 3.8403519699951996e-05,
-      "loss": 0.3202,
+      "epoch": 1.743431722348362,
+      "grad_norm": 0.1731320321559906,
+      "learning_rate": 3.7833259268491735e-05,
+      "loss": 0.3926,
       "step": 48375
     },
     {
-      "epoch": 1.7,
-      "learning_rate": 3.840111492586959e-05,
-      "loss": 0.2739,
+      "epoch": 1.7436119220095865,
+      "grad_norm": 0.17335784435272217,
+      "learning_rate": 3.7830754864801284e-05,
+      "loss": 0.4057,
       "step": 48380
     },
     {
-      "epoch": 1.7,
-      "learning_rate": 3.839870997778253e-05,
-      "loss": 0.2863,
+      "epoch": 1.7437921216708112,
+      "grad_norm": 0.1906140148639679,
+      "learning_rate": 3.782825028629428e-05,
+      "loss": 0.438,
       "step": 48385
     },
     {
-      "epoch": 1.7,
-      "learning_rate": 3.839630485572206e-05,
-      "loss": 0.2788,
+      "epoch": 1.743972321332036,
+      "grad_norm": 0.21084898710250854,
+      "learning_rate": 3.782574553300485e-05,
+      "loss": 0.4079,
       "step": 48390
     },
     {
-      "epoch": 1.7,
-      "learning_rate": 3.83938995597194e-05,
-      "loss": 0.2885,
+      "epoch": 1.7441525209932607,
+      "grad_norm": 0.2448464035987854,
+      "learning_rate": 3.7823240604967116e-05,
+      "loss": 0.4115,
       "step": 48395
     },
     {
-      "epoch": 1.7,
-      "learning_rate": 3.839149408980579e-05,
-      "loss": 0.2836,
+      "epoch": 1.7443327206544852,
+      "grad_norm": 0.185360386967659,
+      "learning_rate": 3.782073550221521e-05,
+      "loss": 0.4011,
       "step": 48400
     },
     {
-      "epoch": 1.7,
-      "learning_rate": 3.8389088446012444e-05,
-      "loss": 0.284,
+      "epoch": 1.7445129203157097,
+      "grad_norm": 0.16345396637916565,
+      "learning_rate": 3.781823022478327e-05,
+      "loss": 0.4431,
       "step": 48405
     },
     {
-      "epoch": 1.7,
-      "learning_rate": 3.838668262837061e-05,
-      "loss": 0.3085,
+      "epoch": 1.7446931199769344,
+      "grad_norm": 0.13991603255271912,
+      "learning_rate": 3.7815724772705423e-05,
+      "loss": 0.4147,
       "step": 48410
     },
     {
-      "epoch": 1.7,
-      "learning_rate": 3.838427663691153e-05,
-      "loss": 0.2612,
+      "epoch": 1.7448733196381592,
+      "grad_norm": 0.15714098513126373,
+      "learning_rate": 3.781321914601581e-05,
+      "loss": 0.3805,
       "step": 48415
     },
     {
-      "epoch": 1.7,
-      "learning_rate": 3.838187047166644e-05,
-      "loss": 0.2742,
+      "epoch": 1.7450535192993837,
+      "grad_norm": 0.20374760031700134,
+      "learning_rate": 3.781071334474856e-05,
+      "loss": 0.4347,
       "step": 48420
     },
     {
-      "epoch": 1.7,
-      "learning_rate": 3.837946413266657e-05,
-      "loss": 0.3048,
+      "epoch": 1.7452337189606082,
+      "grad_norm": 0.22393597662448883,
+      "learning_rate": 3.780820736893783e-05,
+      "loss": 0.441,
       "step": 48425
     },
     {
-      "epoch": 1.7,
-      "learning_rate": 3.837705761994318e-05,
-      "loss": 0.2909,
+      "epoch": 1.745413918621833,
+      "grad_norm": 0.23388418555259705,
+      "learning_rate": 3.780570121861775e-05,
+      "loss": 0.4262,
       "step": 48430
     },
     {
-      "epoch": 1.7,
-      "learning_rate": 3.8374650933527514e-05,
-      "loss": 0.2658,
+      "epoch": 1.7455941182830577,
+      "grad_norm": 0.23228850960731506,
+      "learning_rate": 3.7803194893822466e-05,
+      "loss": 0.4326,
       "step": 48435
     },
     {
-      "epoch": 1.7,
-      "learning_rate": 3.837224407345082e-05,
-      "loss": 0.2941,
+      "epoch": 1.7457743179442824,
+      "grad_norm": 0.24053829908370972,
+      "learning_rate": 3.780068839458614e-05,
+      "loss": 0.4341,
       "step": 48440
     },
     {
-      "epoch": 1.7,
-      "learning_rate": 3.836983703974435e-05,
-      "loss": 0.3045,
+      "epoch": 1.745954517605507,
+      "grad_norm": 0.1622568517923355,
+      "learning_rate": 3.779818172094291e-05,
+      "loss": 0.3959,
       "step": 48445
     },
     {
-      "epoch": 1.7,
-      "learning_rate": 3.8367429832439356e-05,
-      "loss": 0.277,
+      "epoch": 1.7461347172667314,
+      "grad_norm": 0.20720703899860382,
+      "learning_rate": 3.779567487292693e-05,
+      "loss": 0.4108,
       "step": 48450
     },
     {
-      "epoch": 1.7,
-      "learning_rate": 3.836502245156709e-05,
-      "loss": 0.2902,
+      "epoch": 1.7463149169279562,
+      "grad_norm": 0.20429673790931702,
+      "learning_rate": 3.779316785057235e-05,
+      "loss": 0.434,
       "step": 48455
     },
     {
-      "epoch": 1.7,
-      "learning_rate": 3.836261489715882e-05,
-      "loss": 0.2796,
+      "epoch": 1.7464951165891809,
+      "grad_norm": 0.23048081994056702,
+      "learning_rate": 3.7790660653913346e-05,
+      "loss": 0.4158,
       "step": 48460
     },
     {
-      "epoch": 1.71,
-      "learning_rate": 3.83602071692458e-05,
-      "loss": 0.2701,
+      "epoch": 1.7466753162504054,
+      "grad_norm": 0.19923274219036102,
+      "learning_rate": 3.778815328298406e-05,
+      "loss": 0.4146,
       "step": 48465
     },
     {
-      "epoch": 1.71,
-      "learning_rate": 3.83577992678593e-05,
-      "loss": 0.2802,
+      "epoch": 1.7468555159116301,
+      "grad_norm": 0.18411320447921753,
+      "learning_rate": 3.778564573781866e-05,
+      "loss": 0.4568,
       "step": 48470
     },
     {
-      "epoch": 1.71,
-      "learning_rate": 3.835539119303058e-05,
-      "loss": 0.2777,
+      "epoch": 1.7470357155728546,
+      "grad_norm": 0.20448732376098633,
+      "learning_rate": 3.778313801845132e-05,
+      "loss": 0.4085,
       "step": 48475
     },
     {
-      "epoch": 1.71,
-      "learning_rate": 3.8352982944790905e-05,
-      "loss": 0.3027,
+      "epoch": 1.7472159152340794,
+      "grad_norm": 0.18641623854637146,
+      "learning_rate": 3.7780630124916195e-05,
+      "loss": 0.41,
       "step": 48480
     },
     {
-      "epoch": 1.71,
-      "learning_rate": 3.835057452317154e-05,
-      "loss": 0.2844,
+      "epoch": 1.747396114895304,
+      "grad_norm": 0.1633073091506958,
+      "learning_rate": 3.7778122057247464e-05,
+      "loss": 0.4043,
       "step": 48485
     },
     {
-      "epoch": 1.71,
-      "learning_rate": 3.8348165928203766e-05,
-      "loss": 0.2777,
+      "epoch": 1.7475763145565286,
+      "grad_norm": 0.15964628756046295,
+      "learning_rate": 3.777561381547929e-05,
+      "loss": 0.3637,
       "step": 48490
     },
     {
-      "epoch": 1.71,
-      "learning_rate": 3.8345757159918855e-05,
-      "loss": 0.3131,
+      "epoch": 1.7477565142177531,
+      "grad_norm": 0.15613767504692078,
+      "learning_rate": 3.7773105399645845e-05,
+      "loss": 0.402,
       "step": 48495
     },
     {
-      "epoch": 1.71,
-      "learning_rate": 3.8343348218348086e-05,
-      "loss": 0.2874,
+      "epoch": 1.7479367138789779,
+      "grad_norm": 0.1990021914243698,
+      "learning_rate": 3.777059680978132e-05,
+      "loss": 0.4074,
       "step": 48500
     },
     {
-      "epoch": 1.71,
-      "eval_loss": 0.27594712376594543,
-      "eval_runtime": 10.5531,
-      "eval_samples_per_second": 9.476,
-      "eval_steps_per_second": 9.476,
+      "epoch": 1.7479367138789779,
+      "eval_loss": 0.4423659145832062,
+      "eval_runtime": 3.5418,
+      "eval_samples_per_second": 28.234,
+      "eval_steps_per_second": 7.059,
       "step": 48500
     },
     {
-      "epoch": 1.71,
-      "learning_rate": 3.834093910352273e-05,
-      "loss": 0.2925,
+      "epoch": 1.7481169135402026,
+      "grad_norm": 0.20178571343421936,
+      "learning_rate": 3.776808804591989e-05,
+      "loss": 0.4374,
       "step": 48505
     },
     {
-      "epoch": 1.71,
-      "learning_rate": 3.833852981547407e-05,
-      "loss": 0.2861,
+      "epoch": 1.7482971132014273,
+      "grad_norm": 0.16582556068897247,
+      "learning_rate": 3.776557910809572e-05,
+      "loss": 0.4201,
       "step": 48510
     },
     {
-      "epoch": 1.71,
-      "learning_rate": 3.8336120354233395e-05,
-      "loss": 0.2601,
+      "epoch": 1.7484773128626518,
+      "grad_norm": 0.2059725672006607,
+      "learning_rate": 3.7763069996343015e-05,
+      "loss": 0.4146,
       "step": 48515
     },
     {
-      "epoch": 1.71,
-      "learning_rate": 3.8333710719831986e-05,
-      "loss": 0.2842,
+      "epoch": 1.7486575125238764,
+      "grad_norm": 0.24231673777103424,
+      "learning_rate": 3.776056071069595e-05,
+      "loss": 0.4654,
       "step": 48520
     },
     {
-      "epoch": 1.71,
-      "learning_rate": 3.833130091230113e-05,
-      "loss": 0.2735,
+      "epoch": 1.748837712185101,
+      "grad_norm": 0.16822253167629242,
+      "learning_rate": 3.775805125118871e-05,
+      "loss": 0.4132,
       "step": 48525
     },
     {
-      "epoch": 1.71,
-      "learning_rate": 3.832889093167211e-05,
-      "loss": 0.2881,
+      "epoch": 1.7490179118463258,
+      "grad_norm": 0.21261481940746307,
+      "learning_rate": 3.7755541617855505e-05,
+      "loss": 0.4319,
       "step": 48530
     },
     {
-      "epoch": 1.71,
-      "learning_rate": 3.832648077797624e-05,
-      "loss": 0.3089,
+      "epoch": 1.7491981115075503,
+      "grad_norm": 0.15500931441783905,
+      "learning_rate": 3.77530318107305e-05,
+      "loss": 0.3957,
       "step": 48535
     },
     {
-      "epoch": 1.71,
-      "learning_rate": 3.832407045124479e-05,
-      "loss": 0.272,
+      "epoch": 1.7493783111687748,
+      "grad_norm": 0.24944768846035004,
+      "learning_rate": 3.775052182984792e-05,
+      "loss": 0.4093,
       "step": 48540
     },
     {
-      "epoch": 1.71,
-      "learning_rate": 3.832165995150907e-05,
-      "loss": 0.2908,
+      "epoch": 1.7495585108299996,
+      "grad_norm": 0.17092691361904144,
+      "learning_rate": 3.7748011675241935e-05,
+      "loss": 0.4218,
       "step": 48545
     },
     {
-      "epoch": 1.71,
-      "learning_rate": 3.831924927880037e-05,
-      "loss": 0.2877,
+      "epoch": 1.7497387104912243,
+      "grad_norm": 0.20408369600772858,
+      "learning_rate": 3.7745501346946764e-05,
+      "loss": 0.4123,
       "step": 48550
     },
     {
-      "epoch": 1.71,
-      "learning_rate": 3.831683843315e-05,
-      "loss": 0.2693,
+      "epoch": 1.749918910152449,
+      "grad_norm": 0.25046631693840027,
+      "learning_rate": 3.7742990844996606e-05,
+      "loss": 0.4413,
       "step": 48555
     },
     {
-      "epoch": 1.71,
-      "learning_rate": 3.8314427414589255e-05,
-      "loss": 0.2799,
+      "epoch": 1.7500991098136736,
+      "grad_norm": 0.1823837161064148,
+      "learning_rate": 3.7740480169425666e-05,
+      "loss": 0.4226,
       "step": 48560
     },
     {
-      "epoch": 1.71,
-      "learning_rate": 3.8312016223149446e-05,
-      "loss": 0.3004,
+      "epoch": 1.750279309474898,
+      "grad_norm": 0.19407956302165985,
+      "learning_rate": 3.7737969320268143e-05,
+      "loss": 0.4308,
       "step": 48565
     },
     {
-      "epoch": 1.71,
-      "learning_rate": 3.830960485886188e-05,
-      "loss": 0.2897,
+      "epoch": 1.7504595091361228,
+      "grad_norm": 0.23119215667247772,
+      "learning_rate": 3.7735458297558266e-05,
+      "loss": 0.4491,
       "step": 48570
     },
     {
-      "epoch": 1.71,
-      "learning_rate": 3.8307193321757865e-05,
-      "loss": 0.2824,
+      "epoch": 1.7506397087973475,
+      "grad_norm": 0.2448967546224594,
+      "learning_rate": 3.773294710133023e-05,
+      "loss": 0.4618,
       "step": 48575
     },
     {
-      "epoch": 1.71,
-      "learning_rate": 3.830478161186872e-05,
-      "loss": 0.3045,
+      "epoch": 1.750819908458572,
+      "grad_norm": 0.16788001358509064,
+      "learning_rate": 3.773043573161825e-05,
+      "loss": 0.3828,
       "step": 48580
     },
     {
-      "epoch": 1.71,
-      "learning_rate": 3.830236972922575e-05,
-      "loss": 0.281,
+      "epoch": 1.7510001081197968,
+      "grad_norm": 0.1867685317993164,
+      "learning_rate": 3.772792418845655e-05,
+      "loss": 0.3888,
       "step": 48585
     },
     {
-      "epoch": 1.71,
-      "learning_rate": 3.829995767386028e-05,
-      "loss": 0.2744,
+      "epoch": 1.7511803077810213,
+      "grad_norm": 0.2027750462293625,
+      "learning_rate": 3.7725412471879354e-05,
+      "loss": 0.4142,
       "step": 48590
     },
     {
-      "epoch": 1.71,
-      "learning_rate": 3.829754544580362e-05,
-      "loss": 0.3124,
+      "epoch": 1.751360507442246,
+      "grad_norm": 0.19157275557518005,
+      "learning_rate": 3.7722900581920875e-05,
+      "loss": 0.3886,
       "step": 48595
     },
     {
-      "epoch": 1.71,
-      "learning_rate": 3.829513304508709e-05,
-      "loss": 0.287,
+      "epoch": 1.7515407071034708,
+      "grad_norm": 0.19362173974514008,
+      "learning_rate": 3.7720388518615335e-05,
+      "loss": 0.4182,
       "step": 48600
     },
     {
-      "epoch": 1.71,
-      "learning_rate": 3.8292720471742025e-05,
-      "loss": 0.2996,
+      "epoch": 1.7517209067646953,
+      "grad_norm": 0.1794130504131317,
+      "learning_rate": 3.771787628199696e-05,
+      "loss": 0.3702,
       "step": 48605
     },
     {
-      "epoch": 1.71,
-      "learning_rate": 3.8290307725799745e-05,
-      "loss": 0.2858,
+      "epoch": 1.7519011064259198,
+      "grad_norm": 0.1916467696428299,
+      "learning_rate": 3.771536387209999e-05,
+      "loss": 0.3976,
       "step": 48610
     },
     {
-      "epoch": 1.71,
-      "learning_rate": 3.828789480729158e-05,
-      "loss": 0.2869,
+      "epoch": 1.7520813060871445,
+      "grad_norm": 0.2182328701019287,
+      "learning_rate": 3.771285128895865e-05,
+      "loss": 0.4402,
       "step": 48615
     },
     {
-      "epoch": 1.71,
-      "learning_rate": 3.828548171624886e-05,
-      "loss": 0.2766,
+      "epoch": 1.7522615057483693,
+      "grad_norm": 0.1866350769996643,
+      "learning_rate": 3.771033853260717e-05,
+      "loss": 0.4152,
       "step": 48620
     },
     {
-      "epoch": 1.71,
-      "learning_rate": 3.828306845270291e-05,
-      "loss": 0.2747,
+      "epoch": 1.752441705409594,
+      "grad_norm": 0.15760236978530884,
+      "learning_rate": 3.770782560307978e-05,
+      "loss": 0.4226,
       "step": 48625
     },
     {
-      "epoch": 1.71,
-      "learning_rate": 3.828065501668506e-05,
-      "loss": 0.2949,
+      "epoch": 1.7526219050708185,
+      "grad_norm": 0.2243875414133072,
+      "learning_rate": 3.770531250041074e-05,
+      "loss": 0.4394,
       "step": 48630
     },
     {
-      "epoch": 1.71,
-      "learning_rate": 3.827824140822667e-05,
-      "loss": 0.2672,
+      "epoch": 1.752802104732043,
+      "grad_norm": 0.16506066918373108,
+      "learning_rate": 3.770279922463428e-05,
+      "loss": 0.3997,
       "step": 48635
     },
     {
-      "epoch": 1.71,
-      "learning_rate": 3.827582762735907e-05,
-      "loss": 0.2871,
+      "epoch": 1.7529823043932677,
+      "grad_norm": 0.17547442018985748,
+      "learning_rate": 3.770028577578462e-05,
+      "loss": 0.4006,
       "step": 48640
     },
     {
-      "epoch": 1.71,
-      "learning_rate": 3.827341367411359e-05,
-      "loss": 0.2998,
+      "epoch": 1.7531625040544925,
+      "grad_norm": 0.18571364879608154,
+      "learning_rate": 3.769777215389604e-05,
+      "loss": 0.4179,
       "step": 48645
     },
     {
-      "epoch": 1.71,
-      "learning_rate": 3.8270999548521586e-05,
-      "loss": 0.2506,
+      "epoch": 1.753342703715717,
+      "grad_norm": 0.17581316828727722,
+      "learning_rate": 3.7695258359002775e-05,
+      "loss": 0.4329,
       "step": 48650
     },
     {
-      "epoch": 1.71,
-      "learning_rate": 3.8268585250614397e-05,
-      "loss": 0.3298,
+      "epoch": 1.7535229033769415,
+      "grad_norm": 0.22021692991256714,
+      "learning_rate": 3.769274439113906e-05,
+      "loss": 0.4256,
       "step": 48655
     },
     {
-      "epoch": 1.71,
-      "learning_rate": 3.826617078042337e-05,
-      "loss": 0.2921,
+      "epoch": 1.7537031030381662,
+      "grad_norm": 0.16692475974559784,
+      "learning_rate": 3.769023025033917e-05,
+      "loss": 0.4274,
       "step": 48660
     },
     {
-      "epoch": 1.71,
-      "learning_rate": 3.8263756137979855e-05,
-      "loss": 0.2699,
+      "epoch": 1.753883302699391,
+      "grad_norm": 0.16848334670066833,
+      "learning_rate": 3.768771593663735e-05,
+      "loss": 0.4117,
       "step": 48665
     },
     {
-      "epoch": 1.71,
-      "learning_rate": 3.826134132331521e-05,
-      "loss": 0.2849,
+      "epoch": 1.7540635023606157,
+      "grad_norm": 0.2044806033372879,
+      "learning_rate": 3.7685201450067845e-05,
+      "loss": 0.4221,
       "step": 48670
     },
     {
-      "epoch": 1.71,
-      "learning_rate": 3.825892633646079e-05,
-      "loss": 0.2732,
+      "epoch": 1.7542437020218402,
+      "grad_norm": 0.16209609806537628,
+      "learning_rate": 3.768268679066494e-05,
+      "loss": 0.389,
       "step": 48675
     },
     {
-      "epoch": 1.71,
-      "learning_rate": 3.825651117744796e-05,
-      "loss": 0.2994,
+      "epoch": 1.7544239016830647,
+      "grad_norm": 0.18497797846794128,
+      "learning_rate": 3.7680171958462875e-05,
+      "loss": 0.4079,
       "step": 48680
     },
     {
-      "epoch": 1.71,
-      "learning_rate": 3.8254095846308046e-05,
-      "loss": 0.2828,
+      "epoch": 1.7546041013442895,
+      "grad_norm": 0.15725673735141754,
+      "learning_rate": 3.767765695349592e-05,
+      "loss": 0.3912,
       "step": 48685
     },
     {
-      "epoch": 1.71,
-      "learning_rate": 3.825168034307245e-05,
-      "loss": 0.2586,
+      "epoch": 1.7547843010055142,
+      "grad_norm": 0.19795434176921844,
+      "learning_rate": 3.767514177579836e-05,
+      "loss": 0.4147,
       "step": 48690
     },
     {
-      "epoch": 1.71,
-      "learning_rate": 3.824926466777251e-05,
-      "loss": 0.2924,
+      "epoch": 1.7549645006667387,
+      "grad_norm": 0.17479757964611053,
+      "learning_rate": 3.7672626425404436e-05,
+      "loss": 0.4329,
       "step": 48695
     },
     {
-      "epoch": 1.71,
-      "learning_rate": 3.824684882043961e-05,
-      "loss": 0.2748,
+      "epoch": 1.7551447003279634,
+      "grad_norm": 0.20223133265972137,
+      "learning_rate": 3.767011090234842e-05,
+      "loss": 0.3739,
       "step": 48700
     },
     {
-      "epoch": 1.71,
-      "learning_rate": 3.824443280110509e-05,
-      "loss": 0.2621,
+      "epoch": 1.755324899989188,
+      "grad_norm": 0.2165326178073883,
+      "learning_rate": 3.7667595206664606e-05,
+      "loss": 0.4324,
       "step": 48705
     },
     {
-      "epoch": 1.71,
-      "learning_rate": 3.8242016609800345e-05,
-      "loss": 0.2377,
+      "epoch": 1.7555050996504127,
+      "grad_norm": 0.2068873792886734,
+      "learning_rate": 3.766507933838726e-05,
+      "loss": 0.3831,
       "step": 48710
     },
     {
-      "epoch": 1.71,
-      "learning_rate": 3.823960024655675e-05,
-      "loss": 0.2684,
+      "epoch": 1.7556852993116374,
+      "grad_norm": 0.1555756777524948,
+      "learning_rate": 3.7662563297550666e-05,
+      "loss": 0.4153,
       "step": 48715
     },
     {
-      "epoch": 1.71,
-      "learning_rate": 3.8237183711405666e-05,
-      "loss": 0.2967,
+      "epoch": 1.755865498972862,
+      "grad_norm": 0.21677540242671967,
+      "learning_rate": 3.766004708418909e-05,
+      "loss": 0.4148,
       "step": 48720
     },
     {
-      "epoch": 1.71,
-      "learning_rate": 3.823476700437847e-05,
-      "loss": 0.3064,
+      "epoch": 1.7560456986340864,
+      "grad_norm": 0.18430842459201813,
+      "learning_rate": 3.765753069833683e-05,
+      "loss": 0.4374,
       "step": 48725
     },
     {
-      "epoch": 1.71,
-      "learning_rate": 3.823235012550655e-05,
-      "loss": 0.2617,
+      "epoch": 1.7562258982953112,
+      "grad_norm": 0.16822421550750732,
+      "learning_rate": 3.7655014140028156e-05,
+      "loss": 0.3953,
       "step": 48730
     },
     {
-      "epoch": 1.71,
-      "learning_rate": 3.822993307482129e-05,
-      "loss": 0.2875,
+      "epoch": 1.756406097956536,
+      "grad_norm": 0.18837454915046692,
+      "learning_rate": 3.765249740929737e-05,
+      "loss": 0.3994,
       "step": 48735
     },
     {
-      "epoch": 1.71,
-      "learning_rate": 3.822751585235406e-05,
-      "loss": 0.2954,
+      "epoch": 1.7565862976177606,
+      "grad_norm": 0.21585826575756073,
+      "learning_rate": 3.764998050617876e-05,
+      "loss": 0.4374,
       "step": 48740
     },
     {
-      "epoch": 1.71,
-      "learning_rate": 3.822509845813626e-05,
-      "loss": 0.2939,
+      "epoch": 1.7567664972789852,
+      "grad_norm": 0.18385829031467438,
+      "learning_rate": 3.7647463430706605e-05,
+      "loss": 0.4148,
       "step": 48745
     },
     {
-      "epoch": 1.72,
-      "learning_rate": 3.8222680892199266e-05,
-      "loss": 0.2767,
+      "epoch": 1.7569466969402097,
+      "grad_norm": 0.17298060655593872,
+      "learning_rate": 3.7644946182915215e-05,
+      "loss": 0.4257,
       "step": 48750
     },
     {
-      "epoch": 1.72,
-      "learning_rate": 3.822026315457448e-05,
-      "loss": 0.2901,
+      "epoch": 1.7571268966014344,
+      "grad_norm": 0.17734064161777496,
+      "learning_rate": 3.764242876283888e-05,
+      "loss": 0.3634,
       "step": 48755
     },
     {
-      "epoch": 1.72,
-      "learning_rate": 3.821784524529328e-05,
-      "loss": 0.2757,
+      "epoch": 1.7573070962626591,
+      "grad_norm": 0.19183120131492615,
+      "learning_rate": 3.76399111705119e-05,
+      "loss": 0.3898,
       "step": 48760
     },
     {
-      "epoch": 1.72,
-      "learning_rate": 3.8215427164387086e-05,
-      "loss": 0.2656,
+      "epoch": 1.7574872959238836,
+      "grad_norm": 0.19240742921829224,
+      "learning_rate": 3.763739340596858e-05,
+      "loss": 0.4226,
       "step": 48765
     },
     {
-      "epoch": 1.72,
-      "learning_rate": 3.821300891188727e-05,
-      "loss": 0.2772,
+      "epoch": 1.7576674955851082,
+      "grad_norm": 0.19744504988193512,
+      "learning_rate": 3.763487546924322e-05,
+      "loss": 0.4182,
       "step": 48770
     },
     {
-      "epoch": 1.72,
-      "learning_rate": 3.821059048782524e-05,
-      "loss": 0.3002,
+      "epoch": 1.7578476952463329,
+      "grad_norm": 0.15184414386749268,
+      "learning_rate": 3.763235736037014e-05,
+      "loss": 0.368,
       "step": 48775
     },
     {
-      "epoch": 1.72,
-      "learning_rate": 3.8208171892232406e-05,
-      "loss": 0.2943,
+      "epoch": 1.7580278949075576,
+      "grad_norm": 0.1860179305076599,
+      "learning_rate": 3.762983907938362e-05,
+      "loss": 0.3983,
       "step": 48780
     },
     {
-      "epoch": 1.72,
-      "learning_rate": 3.820575312514015e-05,
-      "loss": 0.2894,
+      "epoch": 1.7582080945687824,
+      "grad_norm": 0.1840306669473648,
+      "learning_rate": 3.7627320626317994e-05,
+      "loss": 0.4329,
       "step": 48785
     },
     {
-      "epoch": 1.72,
-      "learning_rate": 3.820333418657991e-05,
-      "loss": 0.3017,
+      "epoch": 1.7583882942300069,
+      "grad_norm": 0.20998184382915497,
+      "learning_rate": 3.762480200120756e-05,
+      "loss": 0.3834,
       "step": 48790
     },
     {
-      "epoch": 1.72,
-      "learning_rate": 3.8200915076583064e-05,
-      "loss": 0.2886,
+      "epoch": 1.7585684938912314,
+      "grad_norm": 0.22181224822998047,
+      "learning_rate": 3.7622283204086653e-05,
+      "loss": 0.4068,
       "step": 48795
     },
     {
-      "epoch": 1.72,
-      "learning_rate": 3.8198495795181036e-05,
-      "loss": 0.2696,
+      "epoch": 1.758748693552456,
+      "grad_norm": 0.16627205908298492,
+      "learning_rate": 3.761976423498958e-05,
+      "loss": 0.3937,
       "step": 48800
     },
     {
-      "epoch": 1.72,
-      "learning_rate": 3.819607634240525e-05,
-      "loss": 0.2835,
+      "epoch": 1.7589288932136808,
+      "grad_norm": 0.15934346616268158,
+      "learning_rate": 3.761724509395066e-05,
+      "loss": 0.4018,
       "step": 48805
     },
     {
-      "epoch": 1.72,
-      "learning_rate": 3.8193656718287097e-05,
-      "loss": 0.2748,
+      "epoch": 1.7591090928749054,
+      "grad_norm": 0.1558726578950882,
+      "learning_rate": 3.761472578100422e-05,
+      "loss": 0.4015,
       "step": 48810
     },
     {
-      "epoch": 1.72,
-      "learning_rate": 3.819123692285802e-05,
-      "loss": 0.2899,
+      "epoch": 1.7592892925361299,
+      "grad_norm": 0.17845207452774048,
+      "learning_rate": 3.7612206296184594e-05,
+      "loss": 0.3892,
       "step": 48815
     },
     {
-      "epoch": 1.72,
-      "learning_rate": 3.818881695614941e-05,
-      "loss": 0.2798,
+      "epoch": 1.7594694921973546,
+      "grad_norm": 0.1628832072019577,
+      "learning_rate": 3.7609686639526086e-05,
+      "loss": 0.394,
       "step": 48820
     },
     {
-      "epoch": 1.72,
-      "learning_rate": 3.818639681819271e-05,
-      "loss": 0.286,
+      "epoch": 1.7596496918585793,
+      "grad_norm": 0.21811261773109436,
+      "learning_rate": 3.760716681106304e-05,
+      "loss": 0.4561,
       "step": 48825
     },
     {
-      "epoch": 1.72,
-      "learning_rate": 3.8183976509019334e-05,
-      "loss": 0.2705,
+      "epoch": 1.759829891519804,
+      "grad_norm": 0.2100425362586975,
+      "learning_rate": 3.760464681082979e-05,
+      "loss": 0.4025,
       "step": 48830
     },
     {
-      "epoch": 1.72,
-      "learning_rate": 3.818155602866072e-05,
-      "loss": 0.2901,
+      "epoch": 1.7600100911810286,
+      "grad_norm": 0.25296419858932495,
+      "learning_rate": 3.760212663886067e-05,
+      "loss": 0.4405,
       "step": 48835
     },
     {
-      "epoch": 1.72,
-      "learning_rate": 3.817913537714828e-05,
-      "loss": 0.285,
+      "epoch": 1.760190290842253,
+      "grad_norm": 0.22018037736415863,
+      "learning_rate": 3.759960629519e-05,
+      "loss": 0.4108,
       "step": 48840
     },
     {
-      "epoch": 1.72,
-      "learning_rate": 3.817671455451346e-05,
-      "loss": 0.2788,
+      "epoch": 1.7603704905034778,
+      "grad_norm": 0.17130319774150848,
+      "learning_rate": 3.759708577985215e-05,
+      "loss": 0.4407,
       "step": 48845
     },
     {
-      "epoch": 1.72,
-      "learning_rate": 3.8174293560787676e-05,
-      "loss": 0.281,
+      "epoch": 1.7605506901647026,
+      "grad_norm": 0.2034503072500229,
+      "learning_rate": 3.759456509288144e-05,
+      "loss": 0.4196,
       "step": 48850
     },
     {
-      "epoch": 1.72,
-      "learning_rate": 3.817187239600238e-05,
-      "loss": 0.2748,
+      "epoch": 1.7607308898259273,
+      "grad_norm": 0.17805613577365875,
+      "learning_rate": 3.759204423431222e-05,
+      "loss": 0.4007,
       "step": 48855
     },
     {
-      "epoch": 1.72,
-      "learning_rate": 3.8169451060188996e-05,
-      "loss": 0.2743,
+      "epoch": 1.7609110894871518,
+      "grad_norm": 0.1614164412021637,
+      "learning_rate": 3.7589523204178836e-05,
+      "loss": 0.4025,
       "step": 48860
     },
     {
-      "epoch": 1.72,
-      "learning_rate": 3.8167029553378966e-05,
-      "loss": 0.2853,
+      "epoch": 1.7610912891483763,
+      "grad_norm": 0.17002156376838684,
+      "learning_rate": 3.7587002002515623e-05,
+      "loss": 0.4165,
       "step": 48865
     },
     {
-      "epoch": 1.72,
-      "learning_rate": 3.8164607875603744e-05,
-      "loss": 0.2706,
+      "epoch": 1.761271488809601,
+      "grad_norm": 0.18788275122642517,
+      "learning_rate": 3.758448062935696e-05,
+      "loss": 0.3999,
       "step": 48870
     },
     {
-      "epoch": 1.72,
-      "learning_rate": 3.816218602689477e-05,
-      "loss": 0.2811,
+      "epoch": 1.7614516884708258,
+      "grad_norm": 0.2487161010503769,
+      "learning_rate": 3.758195908473717e-05,
+      "loss": 0.4245,
       "step": 48875
     },
     {
-      "epoch": 1.72,
-      "learning_rate": 3.815976400728346e-05,
-      "loss": 0.2945,
+      "epoch": 1.7616318881320503,
+      "grad_norm": 0.18604964017868042,
+      "learning_rate": 3.757943736869064e-05,
+      "loss": 0.4501,
       "step": 48880
     },
     {
-      "epoch": 1.72,
-      "learning_rate": 3.8157341816801305e-05,
-      "loss": 0.3165,
+      "epoch": 1.7618120877932748,
+      "grad_norm": 0.15470033884048462,
+      "learning_rate": 3.757691548125171e-05,
+      "loss": 0.3926,
       "step": 48885
     },
     {
-      "epoch": 1.72,
-      "learning_rate": 3.815491945547974e-05,
-      "loss": 0.2618,
+      "epoch": 1.7619922874544995,
+      "grad_norm": 0.1927955448627472,
+      "learning_rate": 3.757439342245473e-05,
+      "loss": 0.4354,
       "step": 48890
     },
     {
-      "epoch": 1.72,
-      "learning_rate": 3.815249692335021e-05,
-      "loss": 0.296,
+      "epoch": 1.7621724871157243,
+      "grad_norm": 0.20000748336315155,
+      "learning_rate": 3.757187119233408e-05,
+      "loss": 0.4443,
       "step": 48895
     },
     {
-      "epoch": 1.72,
-      "learning_rate": 3.8150074220444175e-05,
-      "loss": 0.288,
+      "epoch": 1.762352686776949,
+      "grad_norm": 0.18299151957035065,
+      "learning_rate": 3.756934879092412e-05,
+      "loss": 0.3823,
       "step": 48900
     },
     {
-      "epoch": 1.72,
-      "learning_rate": 3.814765134679308e-05,
-      "loss": 0.29,
+      "epoch": 1.7625328864381735,
+      "grad_norm": 0.17722856998443604,
+      "learning_rate": 3.756682621825922e-05,
+      "loss": 0.3988,
       "step": 48905
     },
     {
-      "epoch": 1.72,
-      "learning_rate": 3.8145228302428416e-05,
-      "loss": 0.2876,
+      "epoch": 1.762713086099398,
+      "grad_norm": 0.20240166783332825,
+      "learning_rate": 3.756430347437374e-05,
+      "loss": 0.4445,
       "step": 48910
     },
     {
-      "epoch": 1.72,
-      "learning_rate": 3.814280508738162e-05,
-      "loss": 0.2828,
+      "epoch": 1.7628932857606228,
+      "grad_norm": 0.17536598443984985,
+      "learning_rate": 3.7561780559302064e-05,
+      "loss": 0.4041,
       "step": 48915
     },
     {
-      "epoch": 1.72,
-      "learning_rate": 3.814038170168417e-05,
-      "loss": 0.2716,
+      "epoch": 1.7630734854218475,
+      "grad_norm": 0.19098421931266785,
+      "learning_rate": 3.7559257473078554e-05,
+      "loss": 0.4033,
       "step": 48920
     },
     {
-      "epoch": 1.72,
-      "learning_rate": 3.81379581453675e-05,
-      "loss": 0.275,
+      "epoch": 1.763253685083072,
+      "grad_norm": 0.17195981740951538,
+      "learning_rate": 3.755673421573759e-05,
+      "loss": 0.4251,
       "step": 48925
     },
     {
-      "epoch": 1.72,
-      "learning_rate": 3.813553441846312e-05,
-      "loss": 0.2987,
+      "epoch": 1.7634338847442965,
+      "grad_norm": 0.15965524315834045,
+      "learning_rate": 3.7554210787313554e-05,
+      "loss": 0.378,
       "step": 48930
     },
     {
-      "epoch": 1.72,
-      "learning_rate": 3.813311052100248e-05,
-      "loss": 0.2832,
+      "epoch": 1.7636140844055213,
+      "grad_norm": 0.20123085379600525,
+      "learning_rate": 3.755168718784083e-05,
+      "loss": 0.4222,
       "step": 48935
     },
     {
-      "epoch": 1.72,
-      "learning_rate": 3.813068645301705e-05,
-      "loss": 0.2787,
+      "epoch": 1.763794284066746,
+      "grad_norm": 0.1748918741941452,
+      "learning_rate": 3.75491634173538e-05,
+      "loss": 0.4014,
       "step": 48940
     },
     {
-      "epoch": 1.72,
-      "learning_rate": 3.8128262214538316e-05,
-      "loss": 0.2945,
+      "epoch": 1.7639744837279707,
+      "grad_norm": 0.1668645590543747,
+      "learning_rate": 3.7546639475886844e-05,
+      "loss": 0.4385,
       "step": 48945
     },
     {
-      "epoch": 1.72,
-      "learning_rate": 3.8125837805597743e-05,
-      "loss": 0.28,
+      "epoch": 1.7641546833891952,
+      "grad_norm": 0.17461763322353363,
+      "learning_rate": 3.754411536347435e-05,
+      "loss": 0.3705,
       "step": 48950
     },
     {
-      "epoch": 1.72,
-      "learning_rate": 3.8123413226226824e-05,
-      "loss": 0.2776,
+      "epoch": 1.7643348830504197,
+      "grad_norm": 0.19829685986042023,
+      "learning_rate": 3.7541591080150725e-05,
+      "loss": 0.4261,
       "step": 48955
     },
     {
-      "epoch": 1.72,
-      "learning_rate": 3.812098847645702e-05,
-      "loss": 0.2849,
+      "epoch": 1.7645150827116445,
+      "grad_norm": 0.16521485149860382,
+      "learning_rate": 3.753906662595035e-05,
+      "loss": 0.3963,
       "step": 48960
     },
     {
-      "epoch": 1.72,
-      "learning_rate": 3.811856355631984e-05,
-      "loss": 0.2916,
+      "epoch": 1.7646952823728692,
+      "grad_norm": 0.19236351549625397,
+      "learning_rate": 3.75365420009076e-05,
+      "loss": 0.454,
       "step": 48965
     },
     {
-      "epoch": 1.72,
-      "learning_rate": 3.811613846584675e-05,
-      "loss": 0.2675,
+      "epoch": 1.7648754820340937,
+      "grad_norm": 0.19585488736629486,
+      "learning_rate": 3.7534017205056915e-05,
+      "loss": 0.4209,
       "step": 48970
     },
     {
-      "epoch": 1.72,
-      "learning_rate": 3.8113713205069254e-05,
-      "loss": 0.3024,
+      "epoch": 1.7650556816953185,
+      "grad_norm": 0.13935211300849915,
+      "learning_rate": 3.7531492238432656e-05,
+      "loss": 0.386,
       "step": 48975
     },
     {
-      "epoch": 1.72,
-      "learning_rate": 3.811128777401882e-05,
-      "loss": 0.2783,
+      "epoch": 1.765235881356543,
+      "grad_norm": 0.182199165225029,
+      "learning_rate": 3.7528967101069254e-05,
+      "loss": 0.4283,
       "step": 48980
     },
     {
-      "epoch": 1.72,
-      "learning_rate": 3.810886217272697e-05,
-      "loss": 0.2888,
+      "epoch": 1.7654160810177677,
+      "grad_norm": 0.23854130506515503,
+      "learning_rate": 3.7526441793001094e-05,
+      "loss": 0.4379,
       "step": 48985
     },
     {
-      "epoch": 1.72,
-      "learning_rate": 3.810643640122518e-05,
-      "loss": 0.3066,
+      "epoch": 1.7655962806789924,
+      "grad_norm": 0.20375332236289978,
+      "learning_rate": 3.7523916314262585e-05,
+      "loss": 0.3951,
       "step": 48990
     },
     {
-      "epoch": 1.72,
-      "learning_rate": 3.8104010459544946e-05,
-      "loss": 0.3046,
+      "epoch": 1.765776480340217,
+      "grad_norm": 0.20095662772655487,
+      "learning_rate": 3.752139066488815e-05,
+      "loss": 0.4075,
       "step": 48995
     },
     {
-      "epoch": 1.72,
-      "learning_rate": 3.810158434771777e-05,
-      "loss": 0.311,
+      "epoch": 1.7659566800014415,
+      "grad_norm": 0.21261079609394073,
+      "learning_rate": 3.751886484491219e-05,
+      "loss": 0.4467,
       "step": 49000
     },
     {
-      "epoch": 1.72,
-      "eval_loss": 0.2752542495727539,
-      "eval_runtime": 10.5525,
-      "eval_samples_per_second": 9.476,
-      "eval_steps_per_second": 9.476,
+      "epoch": 1.7659566800014415,
+      "eval_loss": 0.4405106008052826,
+      "eval_runtime": 3.5203,
+      "eval_samples_per_second": 28.407,
+      "eval_steps_per_second": 7.102,
       "step": 49000
     },
     {
-      "epoch": 1.72,
-      "learning_rate": 3.8099158065775166e-05,
-      "loss": 0.2867,
+      "epoch": 1.7661368796626662,
+      "grad_norm": 0.13927486538887024,
+      "learning_rate": 3.751633885436912e-05,
+      "loss": 0.4111,
       "step": 49005
     },
     {
-      "epoch": 1.72,
-      "learning_rate": 3.8096731613748624e-05,
-      "loss": 0.2566,
+      "epoch": 1.766317079323891,
+      "grad_norm": 0.18607810139656067,
+      "learning_rate": 3.751381269329335e-05,
+      "loss": 0.4157,
       "step": 49010
     },
     {
-      "epoch": 1.72,
-      "learning_rate": 3.809430499166965e-05,
-      "loss": 0.2895,
+      "epoch": 1.7664972789851157,
+      "grad_norm": 0.20816226303577423,
+      "learning_rate": 3.751128636171931e-05,
+      "loss": 0.4413,
       "step": 49015
     },
     {
-      "epoch": 1.72,
-      "learning_rate": 3.809187819956975e-05,
-      "loss": 0.2768,
+      "epoch": 1.7666774786463402,
+      "grad_norm": 0.17857316136360168,
+      "learning_rate": 3.7508759859681416e-05,
+      "loss": 0.4119,
       "step": 49020
     },
     {
-      "epoch": 1.72,
-      "learning_rate": 3.808945123748045e-05,
-      "loss": 0.3111,
+      "epoch": 1.7668576783075647,
+      "grad_norm": 0.24799132347106934,
+      "learning_rate": 3.750623318721409e-05,
+      "loss": 0.3855,
       "step": 49025
     },
     {
-      "epoch": 1.73,
-      "learning_rate": 3.808702410543324e-05,
-      "loss": 0.2982,
+      "epoch": 1.7670378779687894,
+      "grad_norm": 0.15814100205898285,
+      "learning_rate": 3.7503706344351766e-05,
+      "loss": 0.3871,
       "step": 49030
     },
     {
-      "epoch": 1.73,
-      "learning_rate": 3.808459680345966e-05,
-      "loss": 0.2851,
+      "epoch": 1.7672180776300142,
+      "grad_norm": 0.19214874505996704,
+      "learning_rate": 3.7501179331128844e-05,
+      "loss": 0.391,
       "step": 49035
     },
     {
-      "epoch": 1.73,
-      "learning_rate": 3.808216933159121e-05,
-      "loss": 0.2602,
+      "epoch": 1.7673982772912387,
+      "grad_norm": 0.17339515686035156,
+      "learning_rate": 3.7498652147579786e-05,
+      "loss": 0.3669,
       "step": 49040
     },
     {
-      "epoch": 1.73,
-      "learning_rate": 3.8079741689859404e-05,
-      "loss": 0.2768,
+      "epoch": 1.7675784769524632,
+      "grad_norm": 0.1755550056695938,
+      "learning_rate": 3.749612479373902e-05,
+      "loss": 0.4258,
       "step": 49045
     },
     {
-      "epoch": 1.73,
-      "learning_rate": 3.807731387829578e-05,
-      "loss": 0.2631,
+      "epoch": 1.767758676613688,
+      "grad_norm": 0.18907277286052704,
+      "learning_rate": 3.749359726964096e-05,
+      "loss": 0.398,
       "step": 49050
     },
     {
-      "epoch": 1.73,
-      "learning_rate": 3.807488589693186e-05,
-      "loss": 0.2645,
+      "epoch": 1.7679388762749126,
+      "grad_norm": 0.1572108417749405,
+      "learning_rate": 3.749106957532006e-05,
+      "loss": 0.4311,
       "step": 49055
     },
     {
-      "epoch": 1.73,
-      "learning_rate": 3.807245774579915e-05,
-      "loss": 0.2747,
+      "epoch": 1.7681190759361374,
+      "grad_norm": 0.17797333002090454,
+      "learning_rate": 3.748854171081076e-05,
+      "loss": 0.3821,
       "step": 49060
     },
     {
-      "epoch": 1.73,
-      "learning_rate": 3.80700294249292e-05,
-      "loss": 0.293,
+      "epoch": 1.7682992755973619,
+      "grad_norm": 0.18352951109409332,
+      "learning_rate": 3.7486013676147495e-05,
+      "loss": 0.3894,
       "step": 49065
     },
     {
-      "epoch": 1.73,
-      "learning_rate": 3.806760093435352e-05,
-      "loss": 0.2619,
+      "epoch": 1.7684794752585864,
+      "grad_norm": 0.22129513323307037,
+      "learning_rate": 3.748348547136471e-05,
+      "loss": 0.4171,
       "step": 49070
     },
     {
-      "epoch": 1.73,
-      "learning_rate": 3.806517227410367e-05,
-      "loss": 0.2908,
+      "epoch": 1.7686596749198111,
+      "grad_norm": 0.17449241876602173,
+      "learning_rate": 3.748095709649685e-05,
+      "loss": 0.3955,
       "step": 49075
     },
     {
-      "epoch": 1.73,
-      "learning_rate": 3.806274344421115e-05,
-      "loss": 0.2813,
+      "epoch": 1.7688398745810359,
+      "grad_norm": 0.18049213290214539,
+      "learning_rate": 3.747842855157836e-05,
+      "loss": 0.3908,
       "step": 49080
     },
     {
-      "epoch": 1.73,
-      "learning_rate": 3.806031444470753e-05,
-      "loss": 0.2782,
+      "epoch": 1.7690200742422604,
+      "grad_norm": 0.22616390883922577,
+      "learning_rate": 3.747589983664371e-05,
+      "loss": 0.4427,
       "step": 49085
     },
     {
-      "epoch": 1.73,
-      "learning_rate": 3.805788527562433e-05,
-      "loss": 0.2899,
+      "epoch": 1.769200273903485,
+      "grad_norm": 0.17422033846378326,
+      "learning_rate": 3.7473370951727335e-05,
+      "loss": 0.4336,
       "step": 49090
     },
     {
-      "epoch": 1.73,
-      "learning_rate": 3.805545593699309e-05,
-      "loss": 0.2656,
+      "epoch": 1.7693804735647096,
+      "grad_norm": 0.20560961961746216,
+      "learning_rate": 3.74708418968637e-05,
+      "loss": 0.4182,
       "step": 49095
     },
     {
-      "epoch": 1.73,
-      "learning_rate": 3.805302642884536e-05,
-      "loss": 0.2963,
+      "epoch": 1.7695606732259344,
+      "grad_norm": 0.208149254322052,
+      "learning_rate": 3.7468312672087245e-05,
+      "loss": 0.3874,
       "step": 49100
     },
     {
-      "epoch": 1.73,
-      "learning_rate": 3.8050596751212684e-05,
-      "loss": 0.2814,
+      "epoch": 1.769740872887159,
+      "grad_norm": 0.18670162558555603,
+      "learning_rate": 3.746578327743246e-05,
+      "loss": 0.4345,
       "step": 49105
     },
     {
-      "epoch": 1.73,
-      "learning_rate": 3.804816690412661e-05,
-      "loss": 0.2822,
+      "epoch": 1.7699210725483836,
+      "grad_norm": 0.17205697298049927,
+      "learning_rate": 3.746325371293379e-05,
+      "loss": 0.4163,
       "step": 49110
     },
     {
-      "epoch": 1.73,
-      "learning_rate": 3.80457368876187e-05,
-      "loss": 0.2573,
+      "epoch": 1.7701012722096081,
+      "grad_norm": 0.16193123161792755,
+      "learning_rate": 3.746072397862569e-05,
+      "loss": 0.4363,
       "step": 49115
     },
     {
-      "epoch": 1.73,
-      "learning_rate": 3.804330670172048e-05,
-      "loss": 0.2755,
+      "epoch": 1.7702814718708328,
+      "grad_norm": 0.16992096602916718,
+      "learning_rate": 3.7458194074542643e-05,
+      "loss": 0.4117,
       "step": 49120
     },
     {
-      "epoch": 1.73,
-      "learning_rate": 3.804087634646352e-05,
-      "loss": 0.28,
+      "epoch": 1.7704616715320576,
+      "grad_norm": 0.19616004824638367,
+      "learning_rate": 3.7455664000719113e-05,
+      "loss": 0.4117,
       "step": 49125
     },
     {
-      "epoch": 1.73,
-      "learning_rate": 3.803844582187938e-05,
-      "loss": 0.2762,
+      "epoch": 1.7706418711932823,
+      "grad_norm": 0.21857887506484985,
+      "learning_rate": 3.745313375718957e-05,
+      "loss": 0.3996,
       "step": 49130
     },
     {
-      "epoch": 1.73,
-      "learning_rate": 3.8036015127999615e-05,
-      "loss": 0.2853,
+      "epoch": 1.7708220708545068,
+      "grad_norm": 0.2159646898508072,
+      "learning_rate": 3.74506033439885e-05,
+      "loss": 0.414,
       "step": 49135
     },
     {
-      "epoch": 1.73,
-      "learning_rate": 3.803358426485578e-05,
-      "loss": 0.2931,
+      "epoch": 1.7710022705157313,
+      "grad_norm": 0.18626536428928375,
+      "learning_rate": 3.744807276115036e-05,
+      "loss": 0.3965,
       "step": 49140
     },
     {
-      "epoch": 1.73,
-      "learning_rate": 3.803115323247944e-05,
-      "loss": 0.2888,
+      "epoch": 1.771182470176956,
+      "grad_norm": 0.19798725843429565,
+      "learning_rate": 3.744554200870965e-05,
+      "loss": 0.421,
       "step": 49145
     },
     {
-      "epoch": 1.73,
-      "learning_rate": 3.8028722030902165e-05,
-      "loss": 0.2877,
+      "epoch": 1.7713626698381808,
+      "grad_norm": 0.170378640294075,
+      "learning_rate": 3.744301108670083e-05,
+      "loss": 0.3878,
       "step": 49150
     },
     {
-      "epoch": 1.73,
-      "learning_rate": 3.802629066015552e-05,
-      "loss": 0.2623,
+      "epoch": 1.7715428694994053,
+      "grad_norm": 0.1625005006790161,
+      "learning_rate": 3.744047999515839e-05,
+      "loss": 0.3824,
       "step": 49155
     },
     {
-      "epoch": 1.73,
-      "learning_rate": 3.8023859120271084e-05,
-      "loss": 0.2574,
+      "epoch": 1.7717230691606298,
+      "grad_norm": 0.21197515726089478,
+      "learning_rate": 3.743794873411682e-05,
+      "loss": 0.4288,
       "step": 49160
     },
     {
-      "epoch": 1.73,
-      "learning_rate": 3.802142741128041e-05,
-      "loss": 0.2714,
+      "epoch": 1.7719032688218546,
+      "grad_norm": 0.21333040297031403,
+      "learning_rate": 3.743541730361062e-05,
+      "loss": 0.4542,
       "step": 49165
     },
     {
-      "epoch": 1.73,
-      "learning_rate": 3.801899553321509e-05,
-      "loss": 0.2799,
+      "epoch": 1.7720834684830793,
+      "grad_norm": 0.18313553929328918,
+      "learning_rate": 3.743288570367426e-05,
+      "loss": 0.4284,
       "step": 49170
     },
     {
-      "epoch": 1.73,
-      "learning_rate": 3.801656348610669e-05,
-      "loss": 0.2804,
+      "epoch": 1.772263668144304,
+      "grad_norm": 0.1629858762025833,
+      "learning_rate": 3.7430353934342235e-05,
+      "loss": 0.4091,
       "step": 49175
     },
     {
-      "epoch": 1.73,
-      "learning_rate": 3.801413126998679e-05,
-      "loss": 0.2926,
+      "epoch": 1.7724438678055285,
+      "grad_norm": 0.1397615224123001,
+      "learning_rate": 3.7427821995649044e-05,
+      "loss": 0.4355,
       "step": 49180
     },
     {
-      "epoch": 1.73,
-      "learning_rate": 3.8011698884886973e-05,
-      "loss": 0.2721,
+      "epoch": 1.772624067466753,
+      "grad_norm": 0.18151454627513885,
+      "learning_rate": 3.742528988762919e-05,
+      "loss": 0.4128,
       "step": 49185
     },
     {
-      "epoch": 1.73,
-      "learning_rate": 3.800926633083882e-05,
-      "loss": 0.2686,
+      "epoch": 1.7728042671279778,
+      "grad_norm": 0.17477920651435852,
+      "learning_rate": 3.742275761031716e-05,
+      "loss": 0.4322,
       "step": 49190
     },
     {
-      "epoch": 1.73,
-      "learning_rate": 3.800683360787391e-05,
-      "loss": 0.2771,
+      "epoch": 1.7729844667892025,
+      "grad_norm": 0.15377044677734375,
+      "learning_rate": 3.742022516374747e-05,
+      "loss": 0.3747,
       "step": 49195
     },
     {
-      "epoch": 1.73,
-      "learning_rate": 3.800440071602386e-05,
-      "loss": 0.2874,
+      "epoch": 1.773164666450427,
+      "grad_norm": 0.22250111401081085,
+      "learning_rate": 3.741769254795461e-05,
+      "loss": 0.4256,
       "step": 49200
     },
     {
-      "epoch": 1.73,
-      "learning_rate": 3.8001967655320216e-05,
-      "loss": 0.2632,
+      "epoch": 1.7733448661116518,
+      "grad_norm": 0.20984488725662231,
+      "learning_rate": 3.7415159762973094e-05,
+      "loss": 0.4101,
       "step": 49205
     },
     {
-      "epoch": 1.73,
-      "learning_rate": 3.799953442579459e-05,
-      "loss": 0.2797,
+      "epoch": 1.7735250657728763,
+      "grad_norm": 0.15702414512634277,
+      "learning_rate": 3.741262680883743e-05,
+      "loss": 0.4372,
       "step": 49210
     },
     {
-      "epoch": 1.73,
-      "learning_rate": 3.799710102747859e-05,
-      "loss": 0.3042,
+      "epoch": 1.773705265434101,
+      "grad_norm": 0.170258030295372,
+      "learning_rate": 3.7410093685582135e-05,
+      "loss": 0.3799,
       "step": 49215
     },
     {
-      "epoch": 1.73,
-      "learning_rate": 3.799466746040379e-05,
-      "loss": 0.2858,
+      "epoch": 1.7738854650953257,
+      "grad_norm": 0.15923821926116943,
+      "learning_rate": 3.740756039324171e-05,
+      "loss": 0.414,
       "step": 49220
     },
     {
-      "epoch": 1.73,
-      "learning_rate": 3.79922337246018e-05,
-      "loss": 0.2786,
+      "epoch": 1.7740656647565503,
+      "grad_norm": 0.15464822947978973,
+      "learning_rate": 3.7405026931850676e-05,
+      "loss": 0.4139,
       "step": 49225
     },
     {
-      "epoch": 1.73,
-      "learning_rate": 3.7989799820104215e-05,
-      "loss": 0.2892,
+      "epoch": 1.7742458644177748,
+      "grad_norm": 0.16088750958442688,
+      "learning_rate": 3.7402493301443556e-05,
+      "loss": 0.3763,
       "step": 49230
     },
     {
-      "epoch": 1.73,
-      "learning_rate": 3.798736574694264e-05,
-      "loss": 0.2848,
+      "epoch": 1.7744260640789995,
+      "grad_norm": 0.1694214642047882,
+      "learning_rate": 3.739995950205487e-05,
+      "loss": 0.3844,
       "step": 49235
     },
     {
-      "epoch": 1.73,
-      "learning_rate": 3.798493150514868e-05,
-      "loss": 0.2859,
+      "epoch": 1.7746062637402242,
+      "grad_norm": 0.16688519716262817,
+      "learning_rate": 3.739742553371913e-05,
+      "loss": 0.4343,
       "step": 49240
     },
     {
-      "epoch": 1.73,
-      "learning_rate": 3.7982497094753947e-05,
-      "loss": 0.3,
+      "epoch": 1.774786463401449,
+      "grad_norm": 0.1897389143705368,
+      "learning_rate": 3.7394891396470866e-05,
+      "loss": 0.4178,
       "step": 49245
     },
     {
-      "epoch": 1.73,
-      "learning_rate": 3.7980062515790036e-05,
-      "loss": 0.2706,
+      "epoch": 1.7749666630626735,
+      "grad_norm": 0.19561418890953064,
+      "learning_rate": 3.739235709034461e-05,
+      "loss": 0.3847,
       "step": 49250
     },
     {
-      "epoch": 1.73,
-      "learning_rate": 3.797762776828858e-05,
-      "loss": 0.2738,
+      "epoch": 1.775146862723898,
+      "grad_norm": 0.17393264174461365,
+      "learning_rate": 3.7389822615374884e-05,
+      "loss": 0.4522,
       "step": 49255
     },
     {
-      "epoch": 1.73,
-      "learning_rate": 3.797519285228117e-05,
-      "loss": 0.2571,
+      "epoch": 1.7753270623851227,
+      "grad_norm": 0.18216289579868317,
+      "learning_rate": 3.738728797159623e-05,
+      "loss": 0.4269,
       "step": 49260
     },
     {
-      "epoch": 1.73,
-      "learning_rate": 3.7972757767799424e-05,
-      "loss": 0.2982,
+      "epoch": 1.7755072620463475,
+      "grad_norm": 0.1977429836988449,
+      "learning_rate": 3.738475315904317e-05,
+      "loss": 0.3856,
       "step": 49265
     },
     {
-      "epoch": 1.73,
-      "learning_rate": 3.797032251487498e-05,
-      "loss": 0.2826,
+      "epoch": 1.775687461707572,
+      "grad_norm": 0.21220578253269196,
+      "learning_rate": 3.738221817775025e-05,
+      "loss": 0.4363,
       "step": 49270
     },
     {
-      "epoch": 1.73,
-      "learning_rate": 3.796788709353945e-05,
-      "loss": 0.2615,
+      "epoch": 1.7758676613687965,
+      "grad_norm": 0.1830170750617981,
+      "learning_rate": 3.7379683027752e-05,
+      "loss": 0.4057,
       "step": 49275
     },
     {
-      "epoch": 1.73,
-      "learning_rate": 3.796545150382445e-05,
-      "loss": 0.3103,
+      "epoch": 1.7760478610300212,
+      "grad_norm": 0.1913861781358719,
+      "learning_rate": 3.7377147709082966e-05,
+      "loss": 0.4363,
       "step": 49280
     },
     {
-      "epoch": 1.73,
-      "learning_rate": 3.7963015745761597e-05,
-      "loss": 0.2841,
+      "epoch": 1.776228060691246,
+      "grad_norm": 0.16212737560272217,
+      "learning_rate": 3.7374612221777694e-05,
+      "loss": 0.3936,
       "step": 49285
     },
     {
-      "epoch": 1.73,
-      "learning_rate": 3.796057981938253e-05,
-      "loss": 0.2699,
+      "epoch": 1.7764082603524707,
+      "grad_norm": 0.2292328029870987,
+      "learning_rate": 3.737207656587073e-05,
+      "loss": 0.417,
       "step": 49290
     },
     {
-      "epoch": 1.73,
-      "learning_rate": 3.795814372471888e-05,
-      "loss": 0.2865,
+      "epoch": 1.7765884600136952,
+      "grad_norm": 0.19917726516723633,
+      "learning_rate": 3.7369540741396614e-05,
+      "loss": 0.4154,
       "step": 49295
     },
     {
-      "epoch": 1.73,
-      "learning_rate": 3.795570746180227e-05,
-      "loss": 0.2647,
+      "epoch": 1.7767686596749197,
+      "grad_norm": 0.18198366463184357,
+      "learning_rate": 3.7367004748389897e-05,
+      "loss": 0.4043,
       "step": 49300
     },
     {
-      "epoch": 1.73,
-      "learning_rate": 3.795327103066434e-05,
-      "loss": 0.2766,
+      "epoch": 1.7769488593361444,
+      "grad_norm": 0.2522968053817749,
+      "learning_rate": 3.736446858688513e-05,
+      "loss": 0.4596,
       "step": 49305
     },
     {
-      "epoch": 1.73,
-      "learning_rate": 3.7950834431336725e-05,
-      "loss": 0.2839,
+      "epoch": 1.7771290589973692,
+      "grad_norm": 0.21617339551448822,
+      "learning_rate": 3.736193225691689e-05,
+      "loss": 0.4162,
       "step": 49310
     },
     {
-      "epoch": 1.74,
-      "learning_rate": 3.794839766385106e-05,
-      "loss": 0.3372,
+      "epoch": 1.7773092586585937,
+      "grad_norm": 0.1531769037246704,
+      "learning_rate": 3.73593957585197e-05,
+      "loss": 0.4273,
       "step": 49315
     },
     {
-      "epoch": 1.74,
-      "learning_rate": 3.794596072823897e-05,
-      "loss": 0.2916,
+      "epoch": 1.7774894583198182,
+      "grad_norm": 0.16418246924877167,
+      "learning_rate": 3.735685909172815e-05,
+      "loss": 0.3954,
       "step": 49320
     },
     {
-      "epoch": 1.74,
-      "learning_rate": 3.794352362453213e-05,
-      "loss": 0.2933,
+      "epoch": 1.777669657981043,
+      "grad_norm": 0.1850021332502365,
+      "learning_rate": 3.7354322256576765e-05,
+      "loss": 0.4432,
       "step": 49325
     },
     {
-      "epoch": 1.74,
-      "learning_rate": 3.7941086352762155e-05,
-      "loss": 0.3035,
+      "epoch": 1.7778498576422677,
+      "grad_norm": 0.14618489146232605,
+      "learning_rate": 3.735178525310015e-05,
+      "loss": 0.3764,
       "step": 49330
     },
     {
-      "epoch": 1.74,
-      "learning_rate": 3.79386489129607e-05,
-      "loss": 0.2965,
+      "epoch": 1.7780300573034924,
+      "grad_norm": 0.22677135467529297,
+      "learning_rate": 3.734924808133285e-05,
+      "loss": 0.4131,
       "step": 49335
     },
     {
-      "epoch": 1.74,
-      "learning_rate": 3.793621130515942e-05,
-      "loss": 0.2873,
+      "epoch": 1.778210256964717,
+      "grad_norm": 0.1776285022497177,
+      "learning_rate": 3.734671074130943e-05,
+      "loss": 0.4312,
       "step": 49340
     },
     {
-      "epoch": 1.74,
-      "learning_rate": 3.793377352938996e-05,
-      "loss": 0.2836,
+      "epoch": 1.7783904566259414,
+      "grad_norm": 0.1667148470878601,
+      "learning_rate": 3.734417323306447e-05,
+      "loss": 0.3951,
       "step": 49345
     },
     {
-      "epoch": 1.74,
-      "learning_rate": 3.793133558568398e-05,
-      "loss": 0.2836,
+      "epoch": 1.7785706562871662,
+      "grad_norm": 0.18254049122333527,
+      "learning_rate": 3.7341635556632544e-05,
+      "loss": 0.4046,
       "step": 49350
     },
     {
-      "epoch": 1.74,
-      "learning_rate": 3.792889747407312e-05,
-      "loss": 0.2956,
+      "epoch": 1.7787508559483909,
+      "grad_norm": 0.18212547898292542,
+      "learning_rate": 3.733909771204821e-05,
+      "loss": 0.4332,
       "step": 49355
     },
     {
-      "epoch": 1.74,
-      "learning_rate": 3.792645919458906e-05,
-      "loss": 0.2759,
+      "epoch": 1.7789310556096156,
+      "grad_norm": 0.19516201317310333,
+      "learning_rate": 3.733655969934607e-05,
+      "loss": 0.4102,
       "step": 49360
     },
     {
-      "epoch": 1.74,
-      "learning_rate": 3.7924020747263436e-05,
-      "loss": 0.2733,
+      "epoch": 1.7791112552708401,
+      "grad_norm": 0.17682771384716034,
+      "learning_rate": 3.733402151856069e-05,
+      "loss": 0.4214,
       "step": 49365
     },
     {
-      "epoch": 1.74,
-      "learning_rate": 3.792158213212793e-05,
-      "loss": 0.2935,
+      "epoch": 1.7792914549320646,
+      "grad_norm": 0.1816743165254593,
+      "learning_rate": 3.733148316972665e-05,
+      "loss": 0.4073,
       "step": 49370
     },
     {
-      "epoch": 1.74,
-      "learning_rate": 3.791914334921418e-05,
-      "loss": 0.2866,
+      "epoch": 1.7794716545932894,
+      "grad_norm": 0.20628765225410461,
+      "learning_rate": 3.732894465287854e-05,
+      "loss": 0.4256,
       "step": 49375
     },
     {
-      "epoch": 1.74,
-      "learning_rate": 3.7916704398553884e-05,
-      "loss": 0.2597,
+      "epoch": 1.779651854254514,
+      "grad_norm": 0.1907653659582138,
+      "learning_rate": 3.732640596805096e-05,
+      "loss": 0.4183,
       "step": 49380
     },
     {
-      "epoch": 1.74,
-      "learning_rate": 3.791426528017869e-05,
-      "loss": 0.281,
+      "epoch": 1.7798320539157386,
+      "grad_norm": 0.18405283987522125,
+      "learning_rate": 3.732386711527847e-05,
+      "loss": 0.4181,
       "step": 49385
     },
     {
-      "epoch": 1.74,
-      "learning_rate": 3.7911825994120266e-05,
-      "loss": 0.2893,
+      "epoch": 1.7800122535769631,
+      "grad_norm": 0.24257473647594452,
+      "learning_rate": 3.7321328094595685e-05,
+      "loss": 0.4125,
       "step": 49390
     },
     {
-      "epoch": 1.74,
-      "learning_rate": 3.790938654041029e-05,
-      "loss": 0.2988,
+      "epoch": 1.7801924532381879,
+      "grad_norm": 0.19096507132053375,
+      "learning_rate": 3.731878890603718e-05,
+      "loss": 0.3772,
       "step": 49395
     },
     {
-      "epoch": 1.74,
-      "learning_rate": 3.790694691908044e-05,
-      "loss": 0.2972,
+      "epoch": 1.7803726528994126,
+      "grad_norm": 0.19265110790729523,
+      "learning_rate": 3.731624954963757e-05,
+      "loss": 0.4049,
       "step": 49400
     },
     {
-      "epoch": 1.74,
-      "learning_rate": 3.790450713016239e-05,
-      "loss": 0.2926,
+      "epoch": 1.7805528525606373,
+      "grad_norm": 0.24484902620315552,
+      "learning_rate": 3.731371002543144e-05,
+      "loss": 0.4594,
       "step": 49405
     },
     {
-      "epoch": 1.74,
-      "learning_rate": 3.790206717368782e-05,
-      "loss": 0.3084,
+      "epoch": 1.7807330522218618,
+      "grad_norm": 0.19407442212104797,
+      "learning_rate": 3.73111703334534e-05,
+      "loss": 0.3982,
       "step": 49410
     },
     {
-      "epoch": 1.74,
-      "learning_rate": 3.7899627049688416e-05,
-      "loss": 0.2983,
+      "epoch": 1.7809132518830864,
+      "grad_norm": 0.18388943374156952,
+      "learning_rate": 3.7308630473738046e-05,
+      "loss": 0.3976,
       "step": 49415
     },
     {
-      "epoch": 1.74,
-      "learning_rate": 3.789718675819585e-05,
-      "loss": 0.2852,
+      "epoch": 1.781093451544311,
+      "grad_norm": 0.15484626591205597,
+      "learning_rate": 3.730609044631998e-05,
+      "loss": 0.4456,
       "step": 49420
     },
     {
-      "epoch": 1.74,
-      "learning_rate": 3.789474629924181e-05,
-      "loss": 0.278,
+      "epoch": 1.7812736512055358,
+      "grad_norm": 0.18442393839359283,
+      "learning_rate": 3.7303550251233824e-05,
+      "loss": 0.3943,
       "step": 49425
     },
     {
-      "epoch": 1.74,
-      "learning_rate": 3.7892305672857986e-05,
-      "loss": 0.3003,
+      "epoch": 1.7814538508667603,
+      "grad_norm": 0.18085482716560364,
+      "learning_rate": 3.730100988851417e-05,
+      "loss": 0.4069,
       "step": 49430
     },
     {
-      "epoch": 1.74,
-      "learning_rate": 3.7889864879076074e-05,
-      "loss": 0.2931,
+      "epoch": 1.7816340505279848,
+      "grad_norm": 0.21936087310314178,
+      "learning_rate": 3.7298469358195635e-05,
+      "loss": 0.4314,
       "step": 49435
     },
     {
-      "epoch": 1.74,
-      "learning_rate": 3.7887423917927764e-05,
-      "loss": 0.2539,
+      "epoch": 1.7818142501892096,
+      "grad_norm": 0.18822318315505981,
+      "learning_rate": 3.729592866031284e-05,
+      "loss": 0.4145,
       "step": 49440
     },
     {
-      "epoch": 1.74,
-      "learning_rate": 3.788498278944475e-05,
-      "loss": 0.2906,
+      "epoch": 1.7819944498504343,
+      "grad_norm": 0.2043982595205307,
+      "learning_rate": 3.729338779490039e-05,
+      "loss": 0.429,
       "step": 49445
     },
     {
-      "epoch": 1.74,
-      "learning_rate": 3.788254149365872e-05,
-      "loss": 0.2911,
+      "epoch": 1.782174649511659,
+      "grad_norm": 0.1583716869354248,
+      "learning_rate": 3.7290846761992924e-05,
+      "loss": 0.4125,
       "step": 49450
     },
     {
-      "epoch": 1.74,
-      "learning_rate": 3.788010003060137e-05,
-      "loss": 0.2989,
+      "epoch": 1.7823548491728836,
+      "grad_norm": 0.20296910405158997,
+      "learning_rate": 3.728830556162505e-05,
+      "loss": 0.4143,
       "step": 49455
     },
     {
-      "epoch": 1.74,
-      "learning_rate": 3.787765840030442e-05,
-      "loss": 0.2723,
+      "epoch": 1.782535048834108,
+      "grad_norm": 0.20263749361038208,
+      "learning_rate": 3.7285764193831384e-05,
+      "loss": 0.3639,
       "step": 49460
     },
     {
-      "epoch": 1.74,
-      "learning_rate": 3.7875216602799565e-05,
-      "loss": 0.3076,
+      "epoch": 1.7827152484953328,
+      "grad_norm": 0.15118259191513062,
+      "learning_rate": 3.728322265864656e-05,
+      "loss": 0.4173,
       "step": 49465
     },
     {
-      "epoch": 1.74,
-      "learning_rate": 3.78727746381185e-05,
-      "loss": 0.2934,
+      "epoch": 1.7828954481565575,
+      "grad_norm": 0.15944749116897583,
+      "learning_rate": 3.7280680956105206e-05,
+      "loss": 0.3725,
       "step": 49470
     },
     {
-      "epoch": 1.74,
-      "learning_rate": 3.787033250629295e-05,
-      "loss": 0.2802,
+      "epoch": 1.783075647817782,
+      "grad_norm": 0.1850617229938507,
+      "learning_rate": 3.727813908624196e-05,
+      "loss": 0.4237,
       "step": 49475
     },
     {
-      "epoch": 1.74,
-      "learning_rate": 3.7867890207354604e-05,
-      "loss": 0.2747,
+      "epoch": 1.7832558474790068,
+      "grad_norm": 0.17814558744430542,
+      "learning_rate": 3.727559704909144e-05,
+      "loss": 0.4384,
       "step": 49480
     },
     {
-      "epoch": 1.74,
-      "learning_rate": 3.786544774133519e-05,
-      "loss": 0.2802,
+      "epoch": 1.7834360471402313,
+      "grad_norm": 0.18063758313655853,
+      "learning_rate": 3.727305484468828e-05,
+      "loss": 0.4244,
       "step": 49485
     },
     {
-      "epoch": 1.74,
-      "learning_rate": 3.7863005108266417e-05,
-      "loss": 0.2939,
+      "epoch": 1.783616246801456,
+      "grad_norm": 0.21019330620765686,
+      "learning_rate": 3.727051247306713e-05,
+      "loss": 0.3935,
       "step": 49490
     },
     {
-      "epoch": 1.74,
-      "learning_rate": 3.786056230817999e-05,
-      "loss": 0.2919,
+      "epoch": 1.7837964464626808,
+      "grad_norm": 0.17506244778633118,
+      "learning_rate": 3.726796993426263e-05,
+      "loss": 0.4377,
       "step": 49495
     },
     {
-      "epoch": 1.74,
-      "learning_rate": 3.785811934110764e-05,
-      "loss": 0.2557,
+      "epoch": 1.7839766461239053,
+      "grad_norm": 0.17135731875896454,
+      "learning_rate": 3.726542722830941e-05,
+      "loss": 0.3957,
       "step": 49500
     },
     {
-      "epoch": 1.74,
-      "eval_loss": 0.27525606751441956,
-      "eval_runtime": 10.5363,
-      "eval_samples_per_second": 9.491,
-      "eval_steps_per_second": 9.491,
+      "epoch": 1.7839766461239053,
+      "eval_loss": 0.4405691623687744,
+      "eval_runtime": 3.5268,
+      "eval_samples_per_second": 28.354,
+      "eval_steps_per_second": 7.089,
       "step": 49500
     },
     {
-      "epoch": 1.74,
-      "learning_rate": 3.7855676207081095e-05,
-      "loss": 0.2857,
+      "epoch": 1.7841568457851298,
+      "grad_norm": 0.17667745053768158,
+      "learning_rate": 3.7262884355242116e-05,
+      "loss": 0.4162,
       "step": 49505
     },
     {
-      "epoch": 1.74,
-      "learning_rate": 3.7853232906132055e-05,
-      "loss": 0.2636,
+      "epoch": 1.7843370454463545,
+      "grad_norm": 0.2062585949897766,
+      "learning_rate": 3.7260341315095394e-05,
+      "loss": 0.3877,
       "step": 49510
     },
     {
-      "epoch": 1.74,
-      "learning_rate": 3.785078943829227e-05,
-      "loss": 0.2768,
+      "epoch": 1.7845172451075793,
+      "grad_norm": 0.1851799339056015,
+      "learning_rate": 3.72577981079039e-05,
+      "loss": 0.4212,
       "step": 49515
     },
     {
-      "epoch": 1.74,
-      "learning_rate": 3.7848345803593436e-05,
-      "loss": 0.3081,
+      "epoch": 1.784697444768804,
+      "grad_norm": 0.1553850919008255,
+      "learning_rate": 3.725525473370228e-05,
+      "loss": 0.4183,
       "step": 49520
     },
     {
-      "epoch": 1.74,
-      "learning_rate": 3.784590200206731e-05,
-      "loss": 0.3179,
+      "epoch": 1.7848776444300285,
+      "grad_norm": 0.18418540060520172,
+      "learning_rate": 3.725271119252519e-05,
+      "loss": 0.4034,
       "step": 49525
     },
     {
-      "epoch": 1.74,
-      "learning_rate": 3.7843458033745606e-05,
-      "loss": 0.2832,
+      "epoch": 1.785057844091253,
+      "grad_norm": 0.21145012974739075,
+      "learning_rate": 3.7250167484407274e-05,
+      "loss": 0.4491,
       "step": 49530
     },
     {
-      "epoch": 1.74,
-      "learning_rate": 3.784101389866007e-05,
-      "loss": 0.3037,
+      "epoch": 1.7852380437524777,
+      "grad_norm": 0.18339501321315765,
+      "learning_rate": 3.7247623609383206e-05,
+      "loss": 0.4113,
       "step": 49535
     },
     {
-      "epoch": 1.74,
-      "learning_rate": 3.783856959684242e-05,
-      "loss": 0.2964,
+      "epoch": 1.7854182434137025,
+      "grad_norm": 0.1943238079547882,
+      "learning_rate": 3.7245079567487635e-05,
+      "loss": 0.3917,
       "step": 49540
     },
     {
-      "epoch": 1.74,
-      "learning_rate": 3.783612512832442e-05,
-      "loss": 0.2684,
+      "epoch": 1.785598443074927,
+      "grad_norm": 0.15589359402656555,
+      "learning_rate": 3.724253535875522e-05,
+      "loss": 0.4486,
       "step": 49545
     },
     {
-      "epoch": 1.74,
-      "learning_rate": 3.7833680493137786e-05,
-      "loss": 0.2917,
+      "epoch": 1.7857786427361515,
+      "grad_norm": 0.25339236855506897,
+      "learning_rate": 3.723999098322064e-05,
+      "loss": 0.3768,
       "step": 49550
     },
     {
-      "epoch": 1.74,
-      "learning_rate": 3.783123569131427e-05,
-      "loss": 0.2968,
+      "epoch": 1.7859588423973762,
+      "grad_norm": 0.2046034038066864,
+      "learning_rate": 3.7237446440918545e-05,
+      "loss": 0.4172,
       "step": 49555
     },
     {
-      "epoch": 1.74,
-      "learning_rate": 3.782879072288561e-05,
-      "loss": 0.3113,
+      "epoch": 1.786139042058601,
+      "grad_norm": 0.19488854706287384,
+      "learning_rate": 3.723490173188362e-05,
+      "loss": 0.4283,
       "step": 49560
     },
     {
-      "epoch": 1.74,
-      "learning_rate": 3.782634558788356e-05,
-      "loss": 0.291,
+      "epoch": 1.7863192417198257,
+      "grad_norm": 0.16177228093147278,
+      "learning_rate": 3.723235685615052e-05,
+      "loss": 0.3981,
       "step": 49565
     },
     {
-      "epoch": 1.74,
-      "learning_rate": 3.782390028633986e-05,
-      "loss": 0.2844,
+      "epoch": 1.7864994413810502,
+      "grad_norm": 0.15859396755695343,
+      "learning_rate": 3.722981181375392e-05,
+      "loss": 0.399,
       "step": 49570
     },
     {
-      "epoch": 1.74,
-      "learning_rate": 3.782145481828628e-05,
-      "loss": 0.2514,
+      "epoch": 1.7866796410422747,
+      "grad_norm": 0.17841722071170807,
+      "learning_rate": 3.7227266604728516e-05,
+      "loss": 0.4186,
       "step": 49575
     },
     {
-      "epoch": 1.74,
-      "learning_rate": 3.781900918375455e-05,
-      "loss": 0.291,
+      "epoch": 1.7868598407034995,
+      "grad_norm": 0.22205208241939545,
+      "learning_rate": 3.722472122910896e-05,
+      "loss": 0.4336,
       "step": 49580
     },
     {
-      "epoch": 1.74,
-      "learning_rate": 3.781656338277644e-05,
-      "loss": 0.3004,
+      "epoch": 1.7870400403647242,
+      "grad_norm": 0.17837925255298615,
+      "learning_rate": 3.7222175686929947e-05,
+      "loss": 0.4137,
       "step": 49585
     },
     {
-      "epoch": 1.74,
-      "learning_rate": 3.781411741538369e-05,
-      "loss": 0.2854,
+      "epoch": 1.7872202400259487,
+      "grad_norm": 0.19479599595069885,
+      "learning_rate": 3.7219629978226165e-05,
+      "loss": 0.4216,
       "step": 49590
     },
     {
-      "epoch": 1.74,
-      "learning_rate": 3.7811671281608073e-05,
-      "loss": 0.2908,
+      "epoch": 1.7874004396871734,
+      "grad_norm": 0.19044643640518188,
+      "learning_rate": 3.7217084103032284e-05,
+      "loss": 0.4043,
       "step": 49595
     },
     {
-      "epoch": 1.75,
-      "learning_rate": 3.780922498148136e-05,
-      "loss": 0.3035,
+      "epoch": 1.787580639348398,
+      "grad_norm": 0.17032626271247864,
+      "learning_rate": 3.7214538061383e-05,
+      "loss": 0.4044,
       "step": 49600
     },
     {
-      "epoch": 1.75,
-      "learning_rate": 3.780677851503529e-05,
-      "loss": 0.291,
+      "epoch": 1.7877608390096227,
+      "grad_norm": 0.19500906765460968,
+      "learning_rate": 3.7211991853312996e-05,
+      "loss": 0.3888,
       "step": 49605
     },
     {
-      "epoch": 1.75,
-      "learning_rate": 3.780433188230165e-05,
-      "loss": 0.2509,
+      "epoch": 1.7879410386708474,
+      "grad_norm": 0.20601950585842133,
+      "learning_rate": 3.720944547885697e-05,
+      "loss": 0.3977,
       "step": 49610
     },
     {
-      "epoch": 1.75,
-      "learning_rate": 3.780188508331219e-05,
-      "loss": 0.2715,
+      "epoch": 1.788121238332072,
+      "grad_norm": 0.15395066142082214,
+      "learning_rate": 3.720689893804962e-05,
+      "loss": 0.4093,
       "step": 49615
     },
     {
-      "epoch": 1.75,
-      "learning_rate": 3.779943811809869e-05,
-      "loss": 0.2831,
+      "epoch": 1.7883014379932964,
+      "grad_norm": 0.21535900235176086,
+      "learning_rate": 3.720435223092562e-05,
+      "loss": 0.4189,
       "step": 49620
     },
     {
-      "epoch": 1.75,
-      "learning_rate": 3.779699098669292e-05,
-      "loss": 0.2855,
+      "epoch": 1.7884816376545212,
+      "grad_norm": 0.1768050193786621,
+      "learning_rate": 3.720180535751969e-05,
+      "loss": 0.4222,
       "step": 49625
     },
     {
-      "epoch": 1.75,
-      "learning_rate": 3.779454368912666e-05,
-      "loss": 0.2876,
+      "epoch": 1.788661837315746,
+      "grad_norm": 0.1698492169380188,
+      "learning_rate": 3.719925831786653e-05,
+      "loss": 0.3944,
       "step": 49630
     },
     {
-      "epoch": 1.75,
-      "learning_rate": 3.779209622543168e-05,
-      "loss": 0.2821,
+      "epoch": 1.7888420369769706,
+      "grad_norm": 0.1947326809167862,
+      "learning_rate": 3.7196711112000835e-05,
+      "loss": 0.43,
       "step": 49635
     },
     {
-      "epoch": 1.75,
-      "learning_rate": 3.778964859563977e-05,
-      "loss": 0.2485,
+      "epoch": 1.7890222366381952,
+      "grad_norm": 0.15801461040973663,
+      "learning_rate": 3.7194163739957306e-05,
+      "loss": 0.4035,
       "step": 49640
     },
     {
-      "epoch": 1.75,
-      "learning_rate": 3.778720079978269e-05,
-      "loss": 0.2964,
+      "epoch": 1.7892024362994197,
+      "grad_norm": 0.2287774682044983,
+      "learning_rate": 3.719161620177066e-05,
+      "loss": 0.4457,
       "step": 49645
     },
     {
-      "epoch": 1.75,
-      "learning_rate": 3.7784752837892235e-05,
-      "loss": 0.2808,
+      "epoch": 1.7893826359606444,
+      "grad_norm": 0.19257451593875885,
+      "learning_rate": 3.71890684974756e-05,
+      "loss": 0.4043,
       "step": 49650
     },
     {
-      "epoch": 1.75,
-      "learning_rate": 3.778230471000019e-05,
-      "loss": 0.2695,
+      "epoch": 1.7895628356218691,
+      "grad_norm": 0.22990871965885162,
+      "learning_rate": 3.7186520627106855e-05,
+      "loss": 0.4,
       "step": 49655
     },
     {
-      "epoch": 1.75,
-      "learning_rate": 3.777985641613835e-05,
-      "loss": 0.2931,
+      "epoch": 1.7897430352830936,
+      "grad_norm": 0.17866025865077972,
+      "learning_rate": 3.718397259069911e-05,
+      "loss": 0.418,
       "step": 49660
     },
     {
-      "epoch": 1.75,
-      "learning_rate": 3.7777407956338484e-05,
-      "loss": 0.2756,
+      "epoch": 1.7899232349443182,
+      "grad_norm": 0.18929140269756317,
+      "learning_rate": 3.718142438828711e-05,
+      "loss": 0.408,
       "step": 49665
     },
     {
-      "epoch": 1.75,
-      "learning_rate": 3.777495933063241e-05,
-      "loss": 0.2654,
+      "epoch": 1.7901034346055429,
+      "grad_norm": 0.19318966567516327,
+      "learning_rate": 3.717887601990555e-05,
+      "loss": 0.4305,
       "step": 49670
     },
     {
-      "epoch": 1.75,
-      "learning_rate": 3.7772510539051895e-05,
-      "loss": 0.269,
+      "epoch": 1.7902836342667676,
+      "grad_norm": 0.21049553155899048,
+      "learning_rate": 3.717632748558917e-05,
+      "loss": 0.4117,
       "step": 49675
     },
     {
-      "epoch": 1.75,
-      "learning_rate": 3.777006158162876e-05,
-      "loss": 0.2727,
+      "epoch": 1.7904638339279924,
+      "grad_norm": 0.20448461174964905,
+      "learning_rate": 3.717377878537269e-05,
+      "loss": 0.4479,
       "step": 49680
     },
     {
-      "epoch": 1.75,
-      "learning_rate": 3.776761245839478e-05,
-      "loss": 0.2853,
+      "epoch": 1.7906440335892169,
+      "grad_norm": 0.15858854353427887,
+      "learning_rate": 3.717122991929082e-05,
+      "loss": 0.3854,
       "step": 49685
     },
     {
-      "epoch": 1.75,
-      "learning_rate": 3.7765163169381776e-05,
-      "loss": 0.27,
+      "epoch": 1.7908242332504414,
+      "grad_norm": 0.18554039299488068,
+      "learning_rate": 3.7168680887378306e-05,
+      "loss": 0.4105,
       "step": 49690
     },
     {
-      "epoch": 1.75,
-      "learning_rate": 3.7762713714621544e-05,
-      "loss": 0.301,
+      "epoch": 1.7910044329116661,
+      "grad_norm": 0.21041366457939148,
+      "learning_rate": 3.716613168966986e-05,
+      "loss": 0.4343,
       "step": 49695
     },
     {
-      "epoch": 1.75,
-      "learning_rate": 3.7760264094145876e-05,
-      "loss": 0.3008,
+      "epoch": 1.7911846325728908,
+      "grad_norm": 0.18072649836540222,
+      "learning_rate": 3.716358232620024e-05,
+      "loss": 0.4132,
       "step": 49700
     },
     {
-      "epoch": 1.75,
-      "learning_rate": 3.775781430798659e-05,
-      "loss": 0.2869,
+      "epoch": 1.7913648322341154,
+      "grad_norm": 0.16237974166870117,
+      "learning_rate": 3.716103279700416e-05,
+      "loss": 0.4437,
       "step": 49705
     },
     {
-      "epoch": 1.75,
-      "learning_rate": 3.7755364356175493e-05,
-      "loss": 0.304,
+      "epoch": 1.79154503189534,
+      "grad_norm": 0.21563240885734558,
+      "learning_rate": 3.7158483102116354e-05,
+      "loss": 0.4425,
       "step": 49710
     },
     {
-      "epoch": 1.75,
-      "learning_rate": 3.7752914238744395e-05,
-      "loss": 0.257,
+      "epoch": 1.7917252315565646,
+      "grad_norm": 0.18444405496120453,
+      "learning_rate": 3.715593324157158e-05,
+      "loss": 0.4327,
       "step": 49715
     },
     {
-      "epoch": 1.75,
-      "learning_rate": 3.775046395572511e-05,
-      "loss": 0.2746,
+      "epoch": 1.7919054312177893,
+      "grad_norm": 0.20448465645313263,
+      "learning_rate": 3.715338321540457e-05,
+      "loss": 0.414,
       "step": 49720
     },
     {
-      "epoch": 1.75,
-      "learning_rate": 3.774801350714946e-05,
-      "loss": 0.2908,
+      "epoch": 1.792085630879014,
+      "grad_norm": 0.1820807009935379,
+      "learning_rate": 3.715083302365006e-05,
+      "loss": 0.407,
       "step": 49725
     },
     {
-      "epoch": 1.75,
-      "learning_rate": 3.774556289304925e-05,
-      "loss": 0.2751,
+      "epoch": 1.7922658305402386,
+      "grad_norm": 0.17193655669689178,
+      "learning_rate": 3.7148282666342804e-05,
+      "loss": 0.4177,
       "step": 49730
     },
     {
-      "epoch": 1.75,
-      "learning_rate": 3.77431121134563e-05,
-      "loss": 0.2916,
+      "epoch": 1.792446030201463,
+      "grad_norm": 0.1984802633523941,
+      "learning_rate": 3.714573214351754e-05,
+      "loss": 0.4338,
       "step": 49735
     },
     {
-      "epoch": 1.75,
-      "learning_rate": 3.7740661168402445e-05,
-      "loss": 0.2611,
+      "epoch": 1.7926262298626878,
+      "grad_norm": 0.18184858560562134,
+      "learning_rate": 3.714318145520905e-05,
+      "loss": 0.4176,
       "step": 49740
     },
     {
-      "epoch": 1.75,
-      "learning_rate": 3.773821005791951e-05,
-      "loss": 0.3029,
+      "epoch": 1.7928064295239126,
+      "grad_norm": 0.16971564292907715,
+      "learning_rate": 3.7140630601452045e-05,
+      "loss": 0.3892,
       "step": 49745
     },
     {
-      "epoch": 1.75,
-      "learning_rate": 3.77357587820393e-05,
-      "loss": 0.2782,
+      "epoch": 1.7929866291851373,
+      "grad_norm": 0.15988574922084808,
+      "learning_rate": 3.7138079582281306e-05,
+      "loss": 0.4246,
       "step": 49750
     },
     {
-      "epoch": 1.75,
-      "learning_rate": 3.7733307340793656e-05,
-      "loss": 0.2969,
+      "epoch": 1.7931668288463618,
+      "grad_norm": 0.1885288655757904,
+      "learning_rate": 3.713552839773158e-05,
+      "loss": 0.4083,
       "step": 49755
     },
     {
-      "epoch": 1.75,
-      "learning_rate": 3.7730855734214405e-05,
-      "loss": 0.2977,
+      "epoch": 1.7933470285075863,
+      "grad_norm": 0.15891823172569275,
+      "learning_rate": 3.713297704783763e-05,
+      "loss": 0.3774,
       "step": 49760
     },
     {
-      "epoch": 1.75,
-      "learning_rate": 3.7728403962333394e-05,
-      "loss": 0.3303,
+      "epoch": 1.793527228168811,
+      "grad_norm": 0.21999859809875488,
+      "learning_rate": 3.7130425532634214e-05,
+      "loss": 0.4091,
       "step": 49765
     },
     {
-      "epoch": 1.75,
-      "learning_rate": 3.772595202518243e-05,
-      "loss": 0.2923,
+      "epoch": 1.7937074278300358,
+      "grad_norm": 0.21735869348049164,
+      "learning_rate": 3.71278738521561e-05,
+      "loss": 0.4353,
       "step": 49770
     },
     {
-      "epoch": 1.75,
-      "learning_rate": 3.772349992279338e-05,
-      "loss": 0.2865,
+      "epoch": 1.7938876274912603,
+      "grad_norm": 0.17658278346061707,
+      "learning_rate": 3.7125322006438055e-05,
+      "loss": 0.3801,
       "step": 49775
     },
     {
-      "epoch": 1.75,
-      "learning_rate": 3.772104765519806e-05,
-      "loss": 0.2976,
+      "epoch": 1.7940678271524848,
+      "grad_norm": 0.17942824959754944,
+      "learning_rate": 3.712276999551485e-05,
+      "loss": 0.4222,
       "step": 49780
     },
     {
-      "epoch": 1.75,
-      "learning_rate": 3.771859522242832e-05,
-      "loss": 0.2694,
+      "epoch": 1.7942480268137095,
+      "grad_norm": 0.17095422744750977,
+      "learning_rate": 3.712021781942124e-05,
+      "loss": 0.3908,
       "step": 49785
     },
     {
-      "epoch": 1.75,
-      "learning_rate": 3.771614262451601e-05,
-      "loss": 0.3007,
+      "epoch": 1.7944282264749343,
+      "grad_norm": 0.20140354335308075,
+      "learning_rate": 3.711766547819201e-05,
+      "loss": 0.3921,
       "step": 49790
     },
     {
-      "epoch": 1.75,
-      "learning_rate": 3.771368986149296e-05,
-      "loss": 0.2718,
+      "epoch": 1.794608426136159,
+      "grad_norm": 0.1793317049741745,
+      "learning_rate": 3.711511297186194e-05,
+      "loss": 0.4273,
       "step": 49795
     },
     {
-      "epoch": 1.75,
-      "learning_rate": 3.771123693339103e-05,
-      "loss": 0.2804,
+      "epoch": 1.7947886257973835,
+      "grad_norm": 0.1982317417860031,
+      "learning_rate": 3.711256030046581e-05,
+      "loss": 0.3903,
       "step": 49800
     },
     {
-      "epoch": 1.75,
-      "learning_rate": 3.770878384024206e-05,
-      "loss": 0.2479,
+      "epoch": 1.794968825458608,
+      "grad_norm": 0.18103168904781342,
+      "learning_rate": 3.7110007464038375e-05,
+      "loss": 0.4436,
       "step": 49805
     },
     {
-      "epoch": 1.75,
-      "learning_rate": 3.7706330582077915e-05,
-      "loss": 0.2908,
+      "epoch": 1.7951490251198328,
+      "grad_norm": 0.20677992701530457,
+      "learning_rate": 3.710745446261444e-05,
+      "loss": 0.3946,
       "step": 49810
     },
     {
-      "epoch": 1.75,
-      "learning_rate": 3.770387715893043e-05,
-      "loss": 0.2954,
+      "epoch": 1.7953292247810575,
+      "grad_norm": 0.17719873785972595,
+      "learning_rate": 3.710490129622878e-05,
+      "loss": 0.4143,
       "step": 49815
     },
     {
-      "epoch": 1.75,
-      "learning_rate": 3.770142357083149e-05,
-      "loss": 0.3016,
+      "epoch": 1.795509424442282,
+      "grad_norm": 0.15315468609333038,
+      "learning_rate": 3.7102347964916183e-05,
+      "loss": 0.3798,
       "step": 49820
     },
     {
-      "epoch": 1.75,
-      "learning_rate": 3.769896981781292e-05,
-      "loss": 0.3029,
+      "epoch": 1.7956896241035065,
+      "grad_norm": 0.15976719558238983,
+      "learning_rate": 3.709979446871144e-05,
+      "loss": 0.4025,
       "step": 49825
     },
     {
-      "epoch": 1.75,
-      "learning_rate": 3.769651589990661e-05,
-      "loss": 0.2911,
+      "epoch": 1.7958698237647313,
+      "grad_norm": 0.19692429900169373,
+      "learning_rate": 3.7097240807649334e-05,
+      "loss": 0.4113,
       "step": 49830
     },
     {
-      "epoch": 1.75,
-      "learning_rate": 3.76940618171444e-05,
-      "loss": 0.2742,
+      "epoch": 1.796050023425956,
+      "grad_norm": 0.2119201272726059,
+      "learning_rate": 3.709468698176467e-05,
+      "loss": 0.4152,
       "step": 49835
     },
     {
-      "epoch": 1.75,
-      "learning_rate": 3.769160756955816e-05,
-      "loss": 0.2924,
+      "epoch": 1.7962302230871807,
+      "grad_norm": 0.17838400602340698,
+      "learning_rate": 3.7092132991092236e-05,
+      "loss": 0.4198,
       "step": 49840
     },
     {
-      "epoch": 1.75,
-      "learning_rate": 3.768915315717976e-05,
-      "loss": 0.2557,
+      "epoch": 1.7964104227484052,
+      "grad_norm": 0.17005471885204315,
+      "learning_rate": 3.7089578835666834e-05,
+      "loss": 0.3886,
       "step": 49845
     },
     {
-      "epoch": 1.75,
-      "learning_rate": 3.7686698580041075e-05,
-      "loss": 0.2791,
+      "epoch": 1.7965906224096297,
+      "grad_norm": 0.18709275126457214,
+      "learning_rate": 3.708702451552326e-05,
+      "loss": 0.4055,
       "step": 49850
     },
     {
-      "epoch": 1.75,
-      "learning_rate": 3.7684243838173964e-05,
-      "loss": 0.2808,
+      "epoch": 1.7967708220708545,
+      "grad_norm": 0.21572504937648773,
+      "learning_rate": 3.708447003069631e-05,
+      "loss": 0.4095,
       "step": 49855
     },
     {
-      "epoch": 1.75,
-      "learning_rate": 3.7681788931610315e-05,
-      "loss": 0.2843,
+      "epoch": 1.7969510217320792,
+      "grad_norm": 0.17971207201480865,
+      "learning_rate": 3.70819153812208e-05,
+      "loss": 0.4189,
       "step": 49860
     },
     {
-      "epoch": 1.75,
-      "learning_rate": 3.7679333860381984e-05,
-      "loss": 0.2857,
+      "epoch": 1.797131221393304,
+      "grad_norm": 0.20288041234016418,
+      "learning_rate": 3.707936056713154e-05,
+      "loss": 0.4218,
       "step": 49865
     },
     {
-      "epoch": 1.75,
-      "learning_rate": 3.767687862452086e-05,
-      "loss": 0.2519,
+      "epoch": 1.7973114210545285,
+      "grad_norm": 0.18527214229106903,
+      "learning_rate": 3.7076805588463324e-05,
+      "loss": 0.4071,
       "step": 49870
     },
     {
-      "epoch": 1.75,
-      "learning_rate": 3.7674423224058825e-05,
-      "loss": 0.2961,
+      "epoch": 1.797491620715753,
+      "grad_norm": 0.17355002462863922,
+      "learning_rate": 3.7074250445250954e-05,
+      "loss": 0.408,
       "step": 49875
     },
     {
-      "epoch": 1.75,
-      "learning_rate": 3.7671967659027755e-05,
-      "loss": 0.2871,
+      "epoch": 1.7976718203769777,
+      "grad_norm": 0.16921040415763855,
+      "learning_rate": 3.707169513752928e-05,
+      "loss": 0.4111,
       "step": 49880
     },
     {
-      "epoch": 1.76,
-      "learning_rate": 3.766951192945954e-05,
-      "loss": 0.3087,
+      "epoch": 1.7978520200382024,
+      "grad_norm": 0.1597539633512497,
+      "learning_rate": 3.70691396653331e-05,
+      "loss": 0.4486,
       "step": 49885
     },
     {
-      "epoch": 1.76,
-      "learning_rate": 3.766705603538606e-05,
-      "loss": 0.2804,
+      "epoch": 1.798032219699427,
+      "grad_norm": 0.1976759433746338,
+      "learning_rate": 3.7066584028697224e-05,
+      "loss": 0.3784,
       "step": 49890
     },
     {
-      "epoch": 1.76,
-      "learning_rate": 3.76645999768392e-05,
-      "loss": 0.2909,
+      "epoch": 1.7982124193606515,
+      "grad_norm": 0.1847647726535797,
+      "learning_rate": 3.706402822765647e-05,
+      "loss": 0.4264,
       "step": 49895
     },
     {
-      "epoch": 1.76,
-      "learning_rate": 3.7662143753850864e-05,
-      "loss": 0.2842,
+      "epoch": 1.7983926190218762,
+      "grad_norm": 0.2059379369020462,
+      "learning_rate": 3.7061472262245664e-05,
+      "loss": 0.3677,
       "step": 49900
     },
     {
-      "epoch": 1.76,
-      "learning_rate": 3.7659687366452934e-05,
-      "loss": 0.2629,
+      "epoch": 1.798572818683101,
+      "grad_norm": 0.18027707934379578,
+      "learning_rate": 3.7058916132499645e-05,
+      "loss": 0.4286,
       "step": 49905
     },
     {
-      "epoch": 1.76,
-      "learning_rate": 3.76572308146773e-05,
-      "loss": 0.2873,
+      "epoch": 1.7987530183443257,
+      "grad_norm": 0.1781684011220932,
+      "learning_rate": 3.705635983845322e-05,
+      "loss": 0.447,
       "step": 49910
     },
     {
-      "epoch": 1.76,
-      "learning_rate": 3.7654774098555874e-05,
-      "loss": 0.289,
+      "epoch": 1.7989332180055502,
+      "grad_norm": 0.14217472076416016,
+      "learning_rate": 3.7053803380141233e-05,
+      "loss": 0.3952,
       "step": 49915
     },
     {
-      "epoch": 1.76,
-      "learning_rate": 3.765231721812054e-05,
-      "loss": 0.2851,
+      "epoch": 1.7991134176667747,
+      "grad_norm": 0.17906373739242554,
+      "learning_rate": 3.70512467575985e-05,
+      "loss": 0.4536,
       "step": 49920
     },
     {
-      "epoch": 1.76,
-      "learning_rate": 3.764986017340321e-05,
-      "loss": 0.2801,
+      "epoch": 1.7992936173279994,
+      "grad_norm": 0.18036003410816193,
+      "learning_rate": 3.7048689970859874e-05,
+      "loss": 0.3784,
       "step": 49925
     },
     {
-      "epoch": 1.76,
-      "learning_rate": 3.764740296443577e-05,
-      "loss": 0.3052,
+      "epoch": 1.7994738169892242,
+      "grad_norm": 0.15445788204669952,
+      "learning_rate": 3.704613301996017e-05,
+      "loss": 0.4119,
       "step": 49930
     },
     {
-      "epoch": 1.76,
-      "learning_rate": 3.764494559125016e-05,
-      "loss": 0.2867,
+      "epoch": 1.7996540166504487,
+      "grad_norm": 0.20820528268814087,
+      "learning_rate": 3.7043575904934246e-05,
+      "loss": 0.4091,
       "step": 49935
     },
     {
-      "epoch": 1.76,
-      "learning_rate": 3.7642488053878246e-05,
-      "loss": 0.2873,
+      "epoch": 1.7998342163116732,
+      "grad_norm": 0.1830679327249527,
+      "learning_rate": 3.7041018625816926e-05,
+      "loss": 0.4069,
       "step": 49940
     },
     {
-      "epoch": 1.76,
-      "learning_rate": 3.7640030352351965e-05,
-      "loss": 0.3017,
+      "epoch": 1.800014415972898,
+      "grad_norm": 0.18425950407981873,
+      "learning_rate": 3.703846118264306e-05,
+      "loss": 0.4254,
       "step": 49945
     },
     {
-      "epoch": 1.76,
-      "learning_rate": 3.763757248670321e-05,
-      "loss": 0.2837,
+      "epoch": 1.8001946156341226,
+      "grad_norm": 0.19316412508487701,
+      "learning_rate": 3.703590357544749e-05,
+      "loss": 0.4153,
       "step": 49950
     },
     {
-      "epoch": 1.76,
-      "learning_rate": 3.7635114456963914e-05,
-      "loss": 0.2532,
+      "epoch": 1.8003748152953474,
+      "grad_norm": 0.17735867202281952,
+      "learning_rate": 3.7033345804265054e-05,
+      "loss": 0.3632,
       "step": 49955
     },
     {
-      "epoch": 1.76,
-      "learning_rate": 3.7632656263165984e-05,
-      "loss": 0.3039,
+      "epoch": 1.8005550149565719,
+      "grad_norm": 0.22728107869625092,
+      "learning_rate": 3.703078786913063e-05,
+      "loss": 0.3849,
       "step": 49960
     },
     {
-      "epoch": 1.76,
-      "learning_rate": 3.7630197905341336e-05,
-      "loss": 0.2665,
+      "epoch": 1.8007352146177964,
+      "grad_norm": 0.16942258179187775,
+      "learning_rate": 3.702822977007904e-05,
+      "loss": 0.3961,
       "step": 49965
     },
     {
-      "epoch": 1.76,
-      "learning_rate": 3.762773938352189e-05,
-      "loss": 0.2835,
+      "epoch": 1.8009154142790211,
+      "grad_norm": 0.1619909405708313,
+      "learning_rate": 3.7025671507145156e-05,
+      "loss": 0.4082,
       "step": 49970
     },
     {
-      "epoch": 1.76,
-      "learning_rate": 3.7625280697739574e-05,
-      "loss": 0.2875,
+      "epoch": 1.8010956139402459,
+      "grad_norm": 0.17971543967723846,
+      "learning_rate": 3.702311308036381e-05,
+      "loss": 0.4318,
       "step": 49975
     },
     {
-      "epoch": 1.76,
-      "learning_rate": 3.76228218480263e-05,
-      "loss": 0.2812,
+      "epoch": 1.8012758136014704,
+      "grad_norm": 0.19710752367973328,
+      "learning_rate": 3.702055448976989e-05,
+      "loss": 0.4109,
       "step": 49980
     },
     {
-      "epoch": 1.76,
-      "learning_rate": 3.762036283441401e-05,
-      "loss": 0.2623,
+      "epoch": 1.801456013262695,
+      "grad_norm": 0.21794439852237701,
+      "learning_rate": 3.7017995735398237e-05,
+      "loss": 0.3768,
       "step": 49985
     },
     {
-      "epoch": 1.76,
-      "learning_rate": 3.761790365693463e-05,
-      "loss": 0.2721,
+      "epoch": 1.8016362129239196,
+      "grad_norm": 0.21101748943328857,
+      "learning_rate": 3.7015436817283724e-05,
+      "loss": 0.4516,
       "step": 49990
     },
     {
-      "epoch": 1.76,
-      "learning_rate": 3.761544431562008e-05,
-      "loss": 0.2883,
+      "epoch": 1.8018164125851444,
+      "grad_norm": 0.21973469853401184,
+      "learning_rate": 3.70128777354612e-05,
+      "loss": 0.4238,
       "step": 49995
     },
     {
-      "epoch": 1.76,
-      "learning_rate": 3.76129848105023e-05,
-      "loss": 0.2767,
+      "epoch": 1.801996612246369,
+      "grad_norm": 0.1689206063747406,
+      "learning_rate": 3.701031848996555e-05,
+      "loss": 0.4313,
       "step": 50000
     },
     {
-      "epoch": 1.76,
-      "eval_loss": 0.2745961546897888,
-      "eval_runtime": 10.5442,
-      "eval_samples_per_second": 9.484,
-      "eval_steps_per_second": 9.484,
+      "epoch": 1.801996612246369,
+      "eval_loss": 0.4403511881828308,
+      "eval_runtime": 3.5415,
+      "eval_samples_per_second": 28.237,
+      "eval_steps_per_second": 7.059,
       "step": 50000
     },
     {
-      "epoch": 1.76,
-      "learning_rate": 3.7610525141613226e-05,
-      "loss": 0.2897,
+      "epoch": 1.8021768119075936,
+      "grad_norm": 0.17889846861362457,
+      "learning_rate": 3.700775908083164e-05,
+      "loss": 0.4117,
       "step": 50005
     },
     {
-      "epoch": 1.76,
-      "learning_rate": 3.7608065308984786e-05,
-      "loss": 0.3061,
+      "epoch": 1.8023570115688181,
+      "grad_norm": 0.18117192387580872,
+      "learning_rate": 3.7005199508094326e-05,
+      "loss": 0.4065,
       "step": 50010
     },
     {
-      "epoch": 1.76,
-      "learning_rate": 3.760560531264894e-05,
-      "loss": 0.3162,
+      "epoch": 1.8025372112300428,
+      "grad_norm": 0.18847104907035828,
+      "learning_rate": 3.700263977178851e-05,
+      "loss": 0.4004,
       "step": 50015
     },
     {
-      "epoch": 1.76,
-      "learning_rate": 3.760314515263761e-05,
-      "loss": 0.2929,
+      "epoch": 1.8027174108912676,
+      "grad_norm": 0.19416354596614838,
+      "learning_rate": 3.700007987194903e-05,
+      "loss": 0.4118,
       "step": 50020
     },
     {
-      "epoch": 1.76,
-      "learning_rate": 3.7600684828982747e-05,
-      "loss": 0.2934,
+      "epoch": 1.8028976105524923,
+      "grad_norm": 0.19143450260162354,
+      "learning_rate": 3.69975198086108e-05,
+      "loss": 0.4157,
       "step": 50025
     },
     {
-      "epoch": 1.76,
-      "learning_rate": 3.759822434171629e-05,
-      "loss": 0.2888,
+      "epoch": 1.8030778102137168,
+      "grad_norm": 0.17741017043590546,
+      "learning_rate": 3.699495958180868e-05,
+      "loss": 0.3936,
       "step": 50030
     },
     {
-      "epoch": 1.76,
-      "learning_rate": 3.759576369087019e-05,
-      "loss": 0.3187,
+      "epoch": 1.8032580098749413,
+      "grad_norm": 0.18034659326076508,
+      "learning_rate": 3.6992399191577554e-05,
+      "loss": 0.4407,
       "step": 50035
     },
     {
-      "epoch": 1.76,
-      "learning_rate": 3.759330287647641e-05,
-      "loss": 0.2742,
+      "epoch": 1.803438209536166,
+      "grad_norm": 0.17005987465381622,
+      "learning_rate": 3.698983863795232e-05,
+      "loss": 0.4033,
       "step": 50040
     },
     {
-      "epoch": 1.76,
-      "learning_rate": 3.7590841898566884e-05,
-      "loss": 0.2864,
+      "epoch": 1.8036184091973908,
+      "grad_norm": 0.19476930797100067,
+      "learning_rate": 3.698727792096785e-05,
+      "loss": 0.4351,
       "step": 50045
     },
     {
-      "epoch": 1.76,
-      "learning_rate": 3.7588380757173586e-05,
-      "loss": 0.2757,
+      "epoch": 1.8037986088586153,
+      "grad_norm": 0.2065165936946869,
+      "learning_rate": 3.698471704065904e-05,
+      "loss": 0.4046,
       "step": 50050
     },
     {
-      "epoch": 1.76,
-      "learning_rate": 3.758591945232845e-05,
-      "loss": 0.2854,
+      "epoch": 1.8039788085198398,
+      "grad_norm": 0.21584558486938477,
+      "learning_rate": 3.698215599706078e-05,
+      "loss": 0.3787,
       "step": 50055
     },
     {
-      "epoch": 1.76,
-      "learning_rate": 3.758345798406344e-05,
-      "loss": 0.2772,
+      "epoch": 1.8041590081810646,
+      "grad_norm": 0.19947922229766846,
+      "learning_rate": 3.6979594790207964e-05,
+      "loss": 0.4095,
       "step": 50060
     },
     {
-      "epoch": 1.76,
-      "learning_rate": 3.758099635241053e-05,
-      "loss": 0.3134,
+      "epoch": 1.8043392078422893,
+      "grad_norm": 0.1538824588060379,
+      "learning_rate": 3.697703342013549e-05,
+      "loss": 0.39,
       "step": 50065
     },
     {
-      "epoch": 1.76,
-      "learning_rate": 3.757853455740167e-05,
-      "loss": 0.2818,
+      "epoch": 1.804519407503514,
+      "grad_norm": 0.19563712179660797,
+      "learning_rate": 3.6974471886878255e-05,
+      "loss": 0.4194,
       "step": 50070
     },
     {
-      "epoch": 1.76,
-      "learning_rate": 3.7576072599068834e-05,
-      "loss": 0.2701,
+      "epoch": 1.8046996071647385,
+      "grad_norm": 0.18470360338687897,
+      "learning_rate": 3.697191019047116e-05,
+      "loss": 0.3901,
       "step": 50075
     },
     {
-      "epoch": 1.76,
-      "learning_rate": 3.7573610477443974e-05,
-      "loss": 0.2831,
+      "epoch": 1.804879806825963,
+      "grad_norm": 0.24419862031936646,
+      "learning_rate": 3.69693483309491e-05,
+      "loss": 0.4253,
       "step": 50080
     },
     {
-      "epoch": 1.76,
-      "learning_rate": 3.757114819255908e-05,
-      "loss": 0.2732,
+      "epoch": 1.8050600064871878,
+      "grad_norm": 0.1856248527765274,
+      "learning_rate": 3.696678630834699e-05,
+      "loss": 0.4468,
       "step": 50085
     },
     {
-      "epoch": 1.76,
-      "learning_rate": 3.7568685744446094e-05,
-      "loss": 0.2824,
+      "epoch": 1.8052402061484125,
+      "grad_norm": 0.2281470149755478,
+      "learning_rate": 3.696422412269974e-05,
+      "loss": 0.3804,
       "step": 50090
     },
     {
-      "epoch": 1.76,
-      "learning_rate": 3.756622313313702e-05,
-      "loss": 0.2878,
+      "epoch": 1.805420405809637,
+      "grad_norm": 0.17991898953914642,
+      "learning_rate": 3.696166177404224e-05,
+      "loss": 0.3965,
       "step": 50095
     },
     {
-      "epoch": 1.76,
-      "learning_rate": 3.7563760358663814e-05,
-      "loss": 0.287,
+      "epoch": 1.8056006054708618,
+      "grad_norm": 0.15711066126823425,
+      "learning_rate": 3.6959099262409425e-05,
+      "loss": 0.3968,
       "step": 50100
     },
     {
-      "epoch": 1.76,
-      "learning_rate": 3.756129742105846e-05,
-      "loss": 0.2736,
+      "epoch": 1.8057808051320863,
+      "grad_norm": 0.18689851462841034,
+      "learning_rate": 3.6956536587836186e-05,
+      "loss": 0.3862,
       "step": 50105
     },
     {
-      "epoch": 1.76,
-      "learning_rate": 3.755883432035294e-05,
-      "loss": 0.309,
+      "epoch": 1.805961004793311,
+      "grad_norm": 0.13950034976005554,
+      "learning_rate": 3.695397375035747e-05,
+      "loss": 0.4016,
       "step": 50110
     },
     {
-      "epoch": 1.76,
-      "learning_rate": 3.755637105657923e-05,
-      "loss": 0.2891,
+      "epoch": 1.8061412044545357,
+      "grad_norm": 0.19982732832431793,
+      "learning_rate": 3.695141075000816e-05,
+      "loss": 0.4065,
       "step": 50115
     },
     {
-      "epoch": 1.76,
-      "learning_rate": 3.755390762976933e-05,
-      "loss": 0.2989,
+      "epoch": 1.8063214041157603,
+      "grad_norm": 0.1736806035041809,
+      "learning_rate": 3.6948847586823196e-05,
+      "loss": 0.4303,
       "step": 50120
     },
     {
-      "epoch": 1.76,
-      "learning_rate": 3.7551444039955196e-05,
-      "loss": 0.29,
+      "epoch": 1.8065016037769848,
+      "grad_norm": 0.18712177872657776,
+      "learning_rate": 3.69462842608375e-05,
+      "loss": 0.4068,
       "step": 50125
     },
     {
-      "epoch": 1.76,
-      "learning_rate": 3.7548980287168836e-05,
-      "loss": 0.2877,
+      "epoch": 1.8066818034382095,
+      "grad_norm": 0.2055433690547943,
+      "learning_rate": 3.694372077208599e-05,
+      "loss": 0.4056,
       "step": 50130
     },
     {
-      "epoch": 1.76,
-      "learning_rate": 3.7546516371442244e-05,
-      "loss": 0.2558,
+      "epoch": 1.8068620030994342,
+      "grad_norm": 0.1870342344045639,
+      "learning_rate": 3.69411571206036e-05,
+      "loss": 0.4669,
       "step": 50135
     },
     {
-      "epoch": 1.76,
-      "learning_rate": 3.75440522928074e-05,
-      "loss": 0.286,
+      "epoch": 1.807042202760659,
+      "grad_norm": 0.224535271525383,
+      "learning_rate": 3.6938593306425255e-05,
+      "loss": 0.4317,
       "step": 50140
     },
     {
-      "epoch": 1.76,
-      "learning_rate": 3.75415880512963e-05,
-      "loss": 0.2894,
+      "epoch": 1.8072224024218835,
+      "grad_norm": 0.2587486505508423,
+      "learning_rate": 3.693602932958589e-05,
+      "loss": 0.4924,
       "step": 50145
     },
     {
-      "epoch": 1.76,
-      "learning_rate": 3.753912364694095e-05,
-      "loss": 0.2904,
+      "epoch": 1.807402602083108,
+      "grad_norm": 0.22826701402664185,
+      "learning_rate": 3.6933465190120434e-05,
+      "loss": 0.4257,
       "step": 50150
     },
     {
-      "epoch": 1.76,
-      "learning_rate": 3.7536659079773336e-05,
-      "loss": 0.2695,
+      "epoch": 1.8075828017443327,
+      "grad_norm": 0.1683250367641449,
+      "learning_rate": 3.693090088806383e-05,
+      "loss": 0.4344,
       "step": 50155
     },
     {
-      "epoch": 1.76,
-      "learning_rate": 3.753419434982548e-05,
-      "loss": 0.2806,
+      "epoch": 1.8077630014055575,
+      "grad_norm": 0.1752661168575287,
+      "learning_rate": 3.6928336423451e-05,
+      "loss": 0.4172,
       "step": 50160
     },
     {
-      "epoch": 1.76,
-      "learning_rate": 3.753172945712936e-05,
-      "loss": 0.2857,
+      "epoch": 1.807943201066782,
+      "grad_norm": 0.16361607611179352,
+      "learning_rate": 3.69257717963169e-05,
+      "loss": 0.4046,
       "step": 50165
     },
     {
-      "epoch": 1.77,
-      "learning_rate": 3.752926440171699e-05,
-      "loss": 0.3019,
+      "epoch": 1.8081234007280065,
+      "grad_norm": 0.18544571101665497,
+      "learning_rate": 3.692320700669648e-05,
+      "loss": 0.4402,
       "step": 50170
     },
     {
-      "epoch": 1.77,
-      "learning_rate": 3.7526799183620384e-05,
-      "loss": 0.2934,
+      "epoch": 1.8083036003892312,
+      "grad_norm": 0.16442377865314484,
+      "learning_rate": 3.6920642054624655e-05,
+      "loss": 0.3902,
       "step": 50175
     },
     {
-      "epoch": 1.77,
-      "learning_rate": 3.752433380287155e-05,
-      "loss": 0.2921,
+      "epoch": 1.808483800050456,
+      "grad_norm": 0.19460293650627136,
+      "learning_rate": 3.6918076940136406e-05,
+      "loss": 0.4246,
       "step": 50180
     },
     {
-      "epoch": 1.77,
-      "learning_rate": 3.752186825950249e-05,
-      "loss": 0.2819,
+      "epoch": 1.8086639997116807,
+      "grad_norm": 0.1653093844652176,
+      "learning_rate": 3.691551166326665e-05,
+      "loss": 0.4109,
       "step": 50185
     },
     {
-      "epoch": 1.77,
-      "learning_rate": 3.7519402553545224e-05,
-      "loss": 0.2756,
+      "epoch": 1.8088441993729052,
+      "grad_norm": 0.17277853190898895,
+      "learning_rate": 3.691294622405037e-05,
+      "loss": 0.3898,
       "step": 50190
     },
     {
-      "epoch": 1.77,
-      "learning_rate": 3.751693668503177e-05,
-      "loss": 0.2899,
+      "epoch": 1.8090243990341297,
+      "grad_norm": 0.19189266860485077,
+      "learning_rate": 3.69103806225225e-05,
+      "loss": 0.4178,
       "step": 50195
     },
     {
-      "epoch": 1.77,
-      "learning_rate": 3.7514470653994124e-05,
-      "loss": 0.3066,
+      "epoch": 1.8092045986953544,
+      "grad_norm": 0.18027593195438385,
+      "learning_rate": 3.690781485871799e-05,
+      "loss": 0.4507,
       "step": 50200
     },
     {
-      "epoch": 1.77,
-      "learning_rate": 3.751200446046434e-05,
-      "loss": 0.3111,
+      "epoch": 1.8093847983565792,
+      "grad_norm": 0.17248685657978058,
+      "learning_rate": 3.690524893267181e-05,
+      "loss": 0.3912,
       "step": 50205
     },
     {
-      "epoch": 1.77,
-      "learning_rate": 3.750953810447443e-05,
-      "loss": 0.2624,
+      "epoch": 1.8095649980178037,
+      "grad_norm": 0.21308831870555878,
+      "learning_rate": 3.690268284441893e-05,
+      "loss": 0.4291,
       "step": 50210
     },
     {
-      "epoch": 1.77,
-      "learning_rate": 3.750707158605641e-05,
-      "loss": 0.267,
+      "epoch": 1.8097451976790284,
+      "grad_norm": 0.19350473582744598,
+      "learning_rate": 3.6900116593994295e-05,
+      "loss": 0.4422,
       "step": 50215
     },
     {
-      "epoch": 1.77,
-      "learning_rate": 3.7504604905242296e-05,
-      "loss": 0.2832,
+      "epoch": 1.809925397340253,
+      "grad_norm": 0.17806485295295715,
+      "learning_rate": 3.6897550181432865e-05,
+      "loss": 0.3978,
       "step": 50220
     },
     {
-      "epoch": 1.77,
-      "learning_rate": 3.750213806206413e-05,
-      "loss": 0.271,
+      "epoch": 1.8101055970014777,
+      "grad_norm": 0.17208920419216156,
+      "learning_rate": 3.689498360676963e-05,
+      "loss": 0.3817,
       "step": 50225
     },
     {
-      "epoch": 1.77,
-      "learning_rate": 3.749967105655394e-05,
-      "loss": 0.3105,
+      "epoch": 1.8102857966627024,
+      "grad_norm": 0.21627385914325714,
+      "learning_rate": 3.689241687003955e-05,
+      "loss": 0.3997,
       "step": 50230
     },
     {
-      "epoch": 1.77,
-      "learning_rate": 3.749720388874377e-05,
-      "loss": 0.2642,
+      "epoch": 1.810465996323927,
+      "grad_norm": 0.16115298867225647,
+      "learning_rate": 3.688984997127758e-05,
+      "loss": 0.4023,
       "step": 50235
     },
     {
-      "epoch": 1.77,
-      "learning_rate": 3.749473655866563e-05,
-      "loss": 0.2783,
+      "epoch": 1.8106461959851514,
+      "grad_norm": 0.19665555655956268,
+      "learning_rate": 3.688728291051871e-05,
+      "loss": 0.4401,
       "step": 50240
     },
     {
-      "epoch": 1.77,
-      "learning_rate": 3.749226906635159e-05,
-      "loss": 0.2739,
+      "epoch": 1.8108263956463762,
+      "grad_norm": 0.1811586618423462,
+      "learning_rate": 3.688471568779791e-05,
+      "loss": 0.4217,
       "step": 50245
     },
     {
-      "epoch": 1.77,
-      "learning_rate": 3.7489801411833644e-05,
-      "loss": 0.2714,
+      "epoch": 1.8110065953076009,
+      "grad_norm": 0.19190897047519684,
+      "learning_rate": 3.6882148303150166e-05,
+      "loss": 0.4226,
       "step": 50250
     },
     {
-      "epoch": 1.77,
-      "learning_rate": 3.7487333595143875e-05,
-      "loss": 0.2904,
+      "epoch": 1.8111867949688256,
+      "grad_norm": 0.16638103127479553,
+      "learning_rate": 3.687958075661045e-05,
+      "loss": 0.3893,
       "step": 50255
     },
     {
-      "epoch": 1.77,
-      "learning_rate": 3.74848656163143e-05,
-      "loss": 0.293,
+      "epoch": 1.8113669946300501,
+      "grad_norm": 0.18631191551685333,
+      "learning_rate": 3.687701304821374e-05,
+      "loss": 0.3919,
       "step": 50260
     },
     {
-      "epoch": 1.77,
-      "learning_rate": 3.748239747537697e-05,
-      "loss": 0.2806,
+      "epoch": 1.8115471942912746,
+      "grad_norm": 0.16074827313423157,
+      "learning_rate": 3.687444517799503e-05,
+      "loss": 0.3798,
       "step": 50265
     },
     {
-      "epoch": 1.77,
-      "learning_rate": 3.747992917236395e-05,
-      "loss": 0.2623,
+      "epoch": 1.8117273939524994,
+      "grad_norm": 0.18698133528232574,
+      "learning_rate": 3.6871877145989305e-05,
+      "loss": 0.3859,
       "step": 50270
     },
     {
-      "epoch": 1.77,
-      "learning_rate": 3.7477460707307265e-05,
-      "loss": 0.298,
+      "epoch": 1.811907593613724,
+      "grad_norm": 0.1874270737171173,
+      "learning_rate": 3.686930895223156e-05,
+      "loss": 0.3785,
       "step": 50275
     },
     {
-      "epoch": 1.77,
-      "learning_rate": 3.747499208023897e-05,
-      "loss": 0.2629,
+      "epoch": 1.8120877932749486,
+      "grad_norm": 0.17781777679920197,
+      "learning_rate": 3.686674059675677e-05,
+      "loss": 0.3977,
       "step": 50280
     },
     {
-      "epoch": 1.77,
-      "learning_rate": 3.747252329119113e-05,
-      "loss": 0.2498,
+      "epoch": 1.8122679929361731,
+      "grad_norm": 0.17600201070308685,
+      "learning_rate": 3.686417207959994e-05,
+      "loss": 0.4425,
       "step": 50285
     },
     {
-      "epoch": 1.77,
-      "learning_rate": 3.74700543401958e-05,
-      "loss": 0.2924,
+      "epoch": 1.8124481925973979,
+      "grad_norm": 0.17503990232944489,
+      "learning_rate": 3.686160340079605e-05,
+      "loss": 0.3825,
       "step": 50290
     },
     {
-      "epoch": 1.77,
-      "learning_rate": 3.746758522728502e-05,
-      "loss": 0.2857,
+      "epoch": 1.8126283922586226,
+      "grad_norm": 0.18147842586040497,
+      "learning_rate": 3.6859034560380125e-05,
+      "loss": 0.4118,
       "step": 50295
     },
     {
-      "epoch": 1.77,
-      "learning_rate": 3.7465115952490876e-05,
-      "loss": 0.2857,
+      "epoch": 1.8128085919198473,
+      "grad_norm": 0.19389557838439941,
+      "learning_rate": 3.6856465558387144e-05,
+      "loss": 0.3737,
       "step": 50300
     },
     {
-      "epoch": 1.77,
-      "learning_rate": 3.746264651584541e-05,
-      "loss": 0.2751,
+      "epoch": 1.8129887915810718,
+      "grad_norm": 0.17562417685985565,
+      "learning_rate": 3.685389639485211e-05,
+      "loss": 0.3936,
       "step": 50305
     },
     {
-      "epoch": 1.77,
-      "learning_rate": 3.746017691738069e-05,
-      "loss": 0.2849,
+      "epoch": 1.8131689912422964,
+      "grad_norm": 0.16235031187534332,
+      "learning_rate": 3.685132706981004e-05,
+      "loss": 0.4133,
       "step": 50310
     },
     {
-      "epoch": 1.77,
-      "learning_rate": 3.7457707157128784e-05,
-      "loss": 0.2778,
+      "epoch": 1.813349190903521,
+      "grad_norm": 0.16100172698497772,
+      "learning_rate": 3.684875758329593e-05,
+      "loss": 0.401,
       "step": 50315
     },
     {
-      "epoch": 1.77,
-      "learning_rate": 3.7455237235121765e-05,
-      "loss": 0.2708,
+      "epoch": 1.8135293905647458,
+      "grad_norm": 0.1928766369819641,
+      "learning_rate": 3.68461879353448e-05,
+      "loss": 0.4323,
       "step": 50320
     },
     {
-      "epoch": 1.77,
-      "learning_rate": 3.74527671513917e-05,
-      "loss": 0.276,
+      "epoch": 1.8137095902259703,
+      "grad_norm": 0.1594439446926117,
+      "learning_rate": 3.684361812599164e-05,
+      "loss": 0.4094,
       "step": 50325
     },
     {
-      "epoch": 1.77,
-      "learning_rate": 3.745029690597065e-05,
-      "loss": 0.314,
+      "epoch": 1.8138897898871948,
+      "grad_norm": 0.24857677519321442,
+      "learning_rate": 3.684104815527149e-05,
+      "loss": 0.3897,
       "step": 50330
     },
     {
-      "epoch": 1.77,
-      "learning_rate": 3.7447826498890705e-05,
-      "loss": 0.2613,
+      "epoch": 1.8140699895484196,
+      "grad_norm": 0.17496556043624878,
+      "learning_rate": 3.6838478023219344e-05,
+      "loss": 0.3763,
       "step": 50335
     },
     {
-      "epoch": 1.77,
-      "learning_rate": 3.7445355930183945e-05,
-      "loss": 0.2961,
+      "epoch": 1.8142501892096443,
+      "grad_norm": 0.2180209904909134,
+      "learning_rate": 3.683590772987022e-05,
+      "loss": 0.3818,
       "step": 50340
     },
     {
-      "epoch": 1.77,
-      "learning_rate": 3.744288519988243e-05,
-      "loss": 0.297,
+      "epoch": 1.814430388870869,
+      "grad_norm": 0.18124771118164062,
+      "learning_rate": 3.683333727525916e-05,
+      "loss": 0.4156,
       "step": 50345
     },
     {
-      "epoch": 1.77,
-      "learning_rate": 3.744041430801826e-05,
-      "loss": 0.2742,
+      "epoch": 1.8146105885320936,
+      "grad_norm": 0.2114446610212326,
+      "learning_rate": 3.683076665942115e-05,
+      "loss": 0.4311,
       "step": 50350
     },
     {
-      "epoch": 1.77,
-      "learning_rate": 3.74379432546235e-05,
-      "loss": 0.2893,
+      "epoch": 1.814790788193318,
+      "grad_norm": 0.1923656463623047,
+      "learning_rate": 3.682819588239126e-05,
+      "loss": 0.4145,
       "step": 50355
     },
     {
-      "epoch": 1.77,
-      "learning_rate": 3.743547203973024e-05,
-      "loss": 0.2973,
+      "epoch": 1.8149709878545428,
+      "grad_norm": 0.16528750956058502,
+      "learning_rate": 3.682562494420447e-05,
+      "loss": 0.3791,
       "step": 50360
     },
     {
-      "epoch": 1.77,
-      "learning_rate": 3.7433000663370575e-05,
-      "loss": 0.2927,
+      "epoch": 1.8151511875157675,
+      "grad_norm": 0.18421389162540436,
+      "learning_rate": 3.682305384489585e-05,
+      "loss": 0.473,
       "step": 50365
     },
     {
-      "epoch": 1.77,
-      "learning_rate": 3.74305291255766e-05,
-      "loss": 0.2889,
+      "epoch": 1.8153313871769923,
+      "grad_norm": 0.18511919677257538,
+      "learning_rate": 3.682048258450039e-05,
+      "loss": 0.4415,
       "step": 50370
     },
     {
-      "epoch": 1.77,
-      "learning_rate": 3.742805742638038e-05,
-      "loss": 0.2961,
+      "epoch": 1.8155115868382168,
+      "grad_norm": 0.1795085072517395,
+      "learning_rate": 3.681791116305315e-05,
+      "loss": 0.4432,
       "step": 50375
     },
     {
-      "epoch": 1.77,
-      "learning_rate": 3.742558556581403e-05,
-      "loss": 0.3069,
+      "epoch": 1.8156917864994413,
+      "grad_norm": 0.16460351645946503,
+      "learning_rate": 3.6815339580589165e-05,
+      "loss": 0.3958,
       "step": 50380
     },
     {
-      "epoch": 1.77,
-      "learning_rate": 3.742311354390965e-05,
-      "loss": 0.2718,
+      "epoch": 1.815871986160666,
+      "grad_norm": 0.18292105197906494,
+      "learning_rate": 3.6812767837143455e-05,
+      "loss": 0.4311,
       "step": 50385
     },
     {
-      "epoch": 1.77,
-      "learning_rate": 3.7420641360699315e-05,
-      "loss": 0.2959,
+      "epoch": 1.8160521858218908,
+      "grad_norm": 0.19992749392986298,
+      "learning_rate": 3.681019593275108e-05,
+      "loss": 0.4179,
       "step": 50390
     },
     {
-      "epoch": 1.77,
-      "learning_rate": 3.741816901621515e-05,
-      "loss": 0.2958,
+      "epoch": 1.8162323854831153,
+      "grad_norm": 0.1999056190252304,
+      "learning_rate": 3.680762386744707e-05,
+      "loss": 0.4146,
       "step": 50395
     },
     {
-      "epoch": 1.77,
-      "learning_rate": 3.741569651048923e-05,
-      "loss": 0.2924,
+      "epoch": 1.8164125851443398,
+      "grad_norm": 0.16766919195652008,
+      "learning_rate": 3.6805051641266476e-05,
+      "loss": 0.3939,
       "step": 50400
     },
     {
-      "epoch": 1.77,
-      "learning_rate": 3.7413223843553694e-05,
-      "loss": 0.2873,
+      "epoch": 1.8165927848055645,
+      "grad_norm": 0.1673685610294342,
+      "learning_rate": 3.6802479254244327e-05,
+      "loss": 0.416,
       "step": 50405
     },
     {
-      "epoch": 1.77,
-      "learning_rate": 3.74107510154406e-05,
-      "loss": 0.2992,
+      "epoch": 1.8167729844667893,
+      "grad_norm": 0.2412656992673874,
+      "learning_rate": 3.679990670641569e-05,
+      "loss": 0.4164,
       "step": 50410
     },
     {
-      "epoch": 1.77,
-      "learning_rate": 3.7408278026182106e-05,
-      "loss": 0.2784,
+      "epoch": 1.816953184128014,
+      "grad_norm": 0.18288861215114594,
+      "learning_rate": 3.679733399781561e-05,
+      "loss": 0.4088,
       "step": 50415
     },
     {
-      "epoch": 1.77,
-      "learning_rate": 3.7405804875810286e-05,
-      "loss": 0.2861,
+      "epoch": 1.8171333837892385,
+      "grad_norm": 0.15355655550956726,
+      "learning_rate": 3.6794761128479125e-05,
+      "loss": 0.4283,
       "step": 50420
     },
     {
-      "epoch": 1.77,
-      "learning_rate": 3.740333156435727e-05,
-      "loss": 0.2779,
+      "epoch": 1.817313583450463,
+      "grad_norm": 0.1338133066892624,
+      "learning_rate": 3.679218809844132e-05,
+      "loss": 0.3979,
       "step": 50425
     },
     {
-      "epoch": 1.77,
-      "learning_rate": 3.740085809185517e-05,
-      "loss": 0.2882,
+      "epoch": 1.8174937831116877,
+      "grad_norm": 0.2555032968521118,
+      "learning_rate": 3.6789614907737226e-05,
+      "loss": 0.4351,
       "step": 50430
     },
     {
-      "epoch": 1.77,
-      "learning_rate": 3.73983844583361e-05,
-      "loss": 0.2531,
+      "epoch": 1.8176739827729125,
+      "grad_norm": 0.1840798258781433,
+      "learning_rate": 3.6787041556401914e-05,
+      "loss": 0.3992,
       "step": 50435
     },
     {
-      "epoch": 1.77,
-      "learning_rate": 3.7395910663832175e-05,
-      "loss": 0.2711,
+      "epoch": 1.817854182434137,
+      "grad_norm": 0.18239086866378784,
+      "learning_rate": 3.678446804447044e-05,
+      "loss": 0.3901,
       "step": 50440
     },
     {
-      "epoch": 1.77,
-      "learning_rate": 3.7393436708375516e-05,
-      "loss": 0.2866,
+      "epoch": 1.8180343820953615,
+      "grad_norm": 0.2021910846233368,
+      "learning_rate": 3.678189437197788e-05,
+      "loss": 0.4164,
       "step": 50445
     },
     {
-      "epoch": 1.77,
-      "learning_rate": 3.739096259199825e-05,
-      "loss": 0.2892,
+      "epoch": 1.8182145817565862,
+      "grad_norm": 0.22721827030181885,
+      "learning_rate": 3.6779320538959275e-05,
+      "loss": 0.4761,
       "step": 50450
     },
     {
-      "epoch": 1.78,
-      "learning_rate": 3.73884883147325e-05,
-      "loss": 0.2926,
+      "epoch": 1.818394781417811,
+      "grad_norm": 0.148165762424469,
+      "learning_rate": 3.6776746545449715e-05,
+      "loss": 0.3898,
       "step": 50455
     },
     {
-      "epoch": 1.78,
-      "learning_rate": 3.73860138766104e-05,
-      "loss": 0.3057,
+      "epoch": 1.8185749810790357,
+      "grad_norm": 0.1854783147573471,
+      "learning_rate": 3.677417239148428e-05,
+      "loss": 0.3955,
       "step": 50460
     },
     {
-      "epoch": 1.78,
-      "learning_rate": 3.7383539277664074e-05,
-      "loss": 0.2761,
+      "epoch": 1.8187551807402602,
+      "grad_norm": 0.18385660648345947,
+      "learning_rate": 3.6771598077098e-05,
+      "loss": 0.4202,
       "step": 50465
     },
     {
-      "epoch": 1.78,
-      "learning_rate": 3.738106451792564e-05,
-      "loss": 0.2693,
+      "epoch": 1.8189353804014847,
+      "grad_norm": 0.19717253744602203,
+      "learning_rate": 3.6769023602325985e-05,
+      "loss": 0.3799,
       "step": 50470
     },
     {
-      "epoch": 1.78,
-      "learning_rate": 3.737858959742726e-05,
-      "loss": 0.2797,
+      "epoch": 1.8191155800627095,
+      "grad_norm": 0.20641092956066132,
+      "learning_rate": 3.67664489672033e-05,
+      "loss": 0.4063,
       "step": 50475
     },
     {
-      "epoch": 1.78,
-      "learning_rate": 3.7376114516201036e-05,
-      "loss": 0.2759,
+      "epoch": 1.8192957797239342,
+      "grad_norm": 0.18371793627738953,
+      "learning_rate": 3.676387417176503e-05,
+      "loss": 0.4177,
       "step": 50480
     },
     {
-      "epoch": 1.78,
-      "learning_rate": 3.737363927427913e-05,
-      "loss": 0.2778,
+      "epoch": 1.8194759793851587,
+      "grad_norm": 0.16387097537517548,
+      "learning_rate": 3.676129921604625e-05,
+      "loss": 0.4037,
       "step": 50485
     },
     {
-      "epoch": 1.78,
-      "learning_rate": 3.737116387169367e-05,
-      "loss": 0.3123,
+      "epoch": 1.8196561790463834,
+      "grad_norm": 0.2013285756111145,
+      "learning_rate": 3.675872410008204e-05,
+      "loss": 0.4411,
       "step": 50490
     },
     {
-      "epoch": 1.78,
-      "learning_rate": 3.7368688308476796e-05,
-      "loss": 0.2828,
+      "epoch": 1.819836378707608,
+      "grad_norm": 0.17570237815380096,
+      "learning_rate": 3.675614882390751e-05,
+      "loss": 0.4142,
       "step": 50495
     },
     {
-      "epoch": 1.78,
-      "learning_rate": 3.736621258466066e-05,
-      "loss": 0.2872,
+      "epoch": 1.8200165783688327,
+      "grad_norm": 0.17556174099445343,
+      "learning_rate": 3.675357338755771e-05,
+      "loss": 0.4109,
       "step": 50500
     },
     {
-      "epoch": 1.78,
-      "eval_loss": 0.27529093623161316,
-      "eval_runtime": 10.5518,
-      "eval_samples_per_second": 9.477,
-      "eval_steps_per_second": 9.477,
+      "epoch": 1.8200165783688327,
+      "eval_loss": 0.4410496950149536,
+      "eval_runtime": 3.5389,
+      "eval_samples_per_second": 28.257,
+      "eval_steps_per_second": 7.064,
       "step": 50500
     },
     {
-      "epoch": 1.78,
-      "learning_rate": 3.73637367002774e-05,
-      "loss": 0.27,
+      "epoch": 1.8201967780300574,
+      "grad_norm": 0.18107643723487854,
+      "learning_rate": 3.675099779106775e-05,
+      "loss": 0.3993,
       "step": 50505
     },
     {
-      "epoch": 1.78,
-      "learning_rate": 3.736126065535918e-05,
-      "loss": 0.2908,
+      "epoch": 1.820376977691282,
+      "grad_norm": 0.20702654123306274,
+      "learning_rate": 3.6748422034472725e-05,
+      "loss": 0.4126,
       "step": 50510
     },
     {
-      "epoch": 1.78,
-      "learning_rate": 3.735878444993812e-05,
-      "loss": 0.3056,
+      "epoch": 1.8205571773525064,
+      "grad_norm": 0.15964628756046295,
+      "learning_rate": 3.674584611780772e-05,
+      "loss": 0.406,
       "step": 50515
     },
     {
-      "epoch": 1.78,
-      "learning_rate": 3.73563080840464e-05,
-      "loss": 0.2857,
+      "epoch": 1.8207373770137312,
+      "grad_norm": 0.16199252009391785,
+      "learning_rate": 3.6743270041107846e-05,
+      "loss": 0.4537,
       "step": 50520
     },
     {
-      "epoch": 1.78,
-      "learning_rate": 3.735383155771616e-05,
-      "loss": 0.3068,
+      "epoch": 1.820917576674956,
+      "grad_norm": 0.23481187224388123,
+      "learning_rate": 3.6740693804408175e-05,
+      "loss": 0.4111,
       "step": 50525
     },
     {
-      "epoch": 1.78,
-      "learning_rate": 3.735135487097955e-05,
-      "loss": 0.2963,
+      "epoch": 1.8210977763361806,
+      "grad_norm": 0.1678016185760498,
+      "learning_rate": 3.673811740774384e-05,
+      "loss": 0.4329,
       "step": 50530
     },
     {
-      "epoch": 1.78,
-      "learning_rate": 3.734887802386874e-05,
-      "loss": 0.3042,
+      "epoch": 1.8212779759974052,
+      "grad_norm": 0.16724498569965363,
+      "learning_rate": 3.673554085114991e-05,
+      "loss": 0.4037,
       "step": 50535
     },
     {
-      "epoch": 1.78,
-      "learning_rate": 3.734640101641589e-05,
-      "loss": 0.2815,
+      "epoch": 1.8214581756586297,
+      "grad_norm": 0.17332807183265686,
+      "learning_rate": 3.673296413466153e-05,
+      "loss": 0.3876,
       "step": 50540
     },
     {
-      "epoch": 1.78,
-      "learning_rate": 3.734392384865316e-05,
-      "loss": 0.2768,
+      "epoch": 1.8216383753198544,
+      "grad_norm": 0.20103994011878967,
+      "learning_rate": 3.673038725831377e-05,
+      "loss": 0.4242,
       "step": 50545
     },
     {
-      "epoch": 1.78,
-      "learning_rate": 3.734144652061271e-05,
-      "loss": 0.2734,
+      "epoch": 1.8218185749810791,
+      "grad_norm": 0.1513690948486328,
+      "learning_rate": 3.672781022214176e-05,
+      "loss": 0.3812,
       "step": 50550
     },
     {
-      "epoch": 1.78,
-      "learning_rate": 3.733896903232671e-05,
-      "loss": 0.2725,
+      "epoch": 1.8219987746423036,
+      "grad_norm": 0.14303380250930786,
+      "learning_rate": 3.67252330261806e-05,
+      "loss": 0.4238,
       "step": 50555
     },
     {
-      "epoch": 1.78,
-      "learning_rate": 3.733649138382734e-05,
-      "loss": 0.3065,
+      "epoch": 1.8221789743035282,
+      "grad_norm": 0.17810964584350586,
+      "learning_rate": 3.6722655670465416e-05,
+      "loss": 0.4074,
       "step": 50560
     },
     {
-      "epoch": 1.78,
-      "learning_rate": 3.7334013575146756e-05,
-      "loss": 0.2916,
+      "epoch": 1.8223591739647529,
+      "grad_norm": 0.15270282328128815,
+      "learning_rate": 3.672007815503132e-05,
+      "loss": 0.4043,
       "step": 50565
     },
     {
-      "epoch": 1.78,
-      "learning_rate": 3.733153560631712e-05,
-      "loss": 0.2769,
+      "epoch": 1.8225393736259776,
+      "grad_norm": 0.2379712015390396,
+      "learning_rate": 3.671750047991343e-05,
+      "loss": 0.4042,
       "step": 50570
     },
     {
-      "epoch": 1.78,
-      "learning_rate": 3.7329057477370636e-05,
-      "loss": 0.2781,
+      "epoch": 1.8227195732872024,
+      "grad_norm": 0.19908632338047028,
+      "learning_rate": 3.6714922645146856e-05,
+      "loss": 0.4034,
       "step": 50575
     },
     {
-      "epoch": 1.78,
-      "learning_rate": 3.732657918833946e-05,
-      "loss": 0.2919,
+      "epoch": 1.8228997729484269,
+      "grad_norm": 0.16394317150115967,
+      "learning_rate": 3.671234465076673e-05,
+      "loss": 0.4139,
       "step": 50580
     },
     {
-      "epoch": 1.78,
-      "learning_rate": 3.7324100739255774e-05,
-      "loss": 0.2488,
+      "epoch": 1.8230799726096514,
+      "grad_norm": 0.18769471347332,
+      "learning_rate": 3.670976649680819e-05,
+      "loss": 0.4194,
       "step": 50585
     },
     {
-      "epoch": 1.78,
-      "learning_rate": 3.7321622130151775e-05,
-      "loss": 0.2713,
+      "epoch": 1.8232601722708761,
+      "grad_norm": 0.18766231834888458,
+      "learning_rate": 3.6707188183306326e-05,
+      "loss": 0.4128,
       "step": 50590
     },
     {
-      "epoch": 1.78,
-      "learning_rate": 3.7319143361059616e-05,
-      "loss": 0.2853,
+      "epoch": 1.8234403719321008,
+      "grad_norm": 0.15433186292648315,
+      "learning_rate": 3.6704609710296295e-05,
+      "loss": 0.4008,
       "step": 50595
     },
     {
-      "epoch": 1.78,
-      "learning_rate": 3.731666443201151e-05,
-      "loss": 0.2938,
+      "epoch": 1.8236205715933254,
+      "grad_norm": 0.20321685075759888,
+      "learning_rate": 3.670203107781324e-05,
+      "loss": 0.3944,
       "step": 50600
     },
     {
-      "epoch": 1.78,
-      "learning_rate": 3.731418534303963e-05,
-      "loss": 0.2751,
+      "epoch": 1.82380077125455,
+      "grad_norm": 0.21775129437446594,
+      "learning_rate": 3.669945228589225e-05,
+      "loss": 0.4413,
       "step": 50605
     },
     {
-      "epoch": 1.78,
-      "learning_rate": 3.7311706094176166e-05,
-      "loss": 0.2899,
+      "epoch": 1.8239809709157746,
+      "grad_norm": 0.2147761881351471,
+      "learning_rate": 3.669687333456852e-05,
+      "loss": 0.418,
       "step": 50610
     },
     {
-      "epoch": 1.78,
-      "learning_rate": 3.730922668545331e-05,
-      "loss": 0.2622,
+      "epoch": 1.8241611705769993,
+      "grad_norm": 0.20212721824645996,
+      "learning_rate": 3.669429422387713e-05,
+      "loss": 0.3872,
       "step": 50615
     },
     {
-      "epoch": 1.78,
-      "learning_rate": 3.730674711690326e-05,
-      "loss": 0.2915,
+      "epoch": 1.824341370238224,
+      "grad_norm": 0.1724279671907425,
+      "learning_rate": 3.669171495385325e-05,
+      "loss": 0.3979,
       "step": 50620
     },
     {
-      "epoch": 1.78,
-      "learning_rate": 3.7304267388558215e-05,
-      "loss": 0.2604,
+      "epoch": 1.8245215698994486,
+      "grad_norm": 0.1798325926065445,
+      "learning_rate": 3.668913552453203e-05,
+      "loss": 0.404,
       "step": 50625
     },
     {
-      "epoch": 1.78,
-      "learning_rate": 3.730178750045036e-05,
-      "loss": 0.2886,
+      "epoch": 1.824701769560673,
+      "grad_norm": 0.1844148486852646,
+      "learning_rate": 3.668655593594858e-05,
+      "loss": 0.4004,
       "step": 50630
     },
     {
-      "epoch": 1.78,
-      "learning_rate": 3.729930745261191e-05,
-      "loss": 0.275,
+      "epoch": 1.8248819692218978,
+      "grad_norm": 0.22137939929962158,
+      "learning_rate": 3.6683976188138084e-05,
+      "loss": 0.3921,
       "step": 50635
     },
     {
-      "epoch": 1.78,
-      "learning_rate": 3.729682724507505e-05,
-      "loss": 0.3038,
+      "epoch": 1.8250621688831226,
+      "grad_norm": 0.1671324074268341,
+      "learning_rate": 3.6681396281135676e-05,
+      "loss": 0.3937,
       "step": 50640
     },
     {
-      "epoch": 1.78,
-      "learning_rate": 3.7294346877872e-05,
-      "loss": 0.2918,
+      "epoch": 1.8252423685443473,
+      "grad_norm": 0.21605345606803894,
+      "learning_rate": 3.6678816214976504e-05,
+      "loss": 0.4299,
       "step": 50645
     },
     {
-      "epoch": 1.78,
-      "learning_rate": 3.729186635103495e-05,
-      "loss": 0.2943,
+      "epoch": 1.8254225682055718,
+      "grad_norm": 0.16101548075675964,
+      "learning_rate": 3.667623598969572e-05,
+      "loss": 0.3997,
       "step": 50650
     },
     {
-      "epoch": 1.78,
-      "learning_rate": 3.728938566459612e-05,
-      "loss": 0.2683,
+      "epoch": 1.8256027678667963,
+      "grad_norm": 0.1385546624660492,
+      "learning_rate": 3.6673655605328475e-05,
+      "loss": 0.3785,
       "step": 50655
     },
     {
-      "epoch": 1.78,
-      "learning_rate": 3.728690481858771e-05,
-      "loss": 0.2774,
+      "epoch": 1.825782967528021,
+      "grad_norm": 0.1622258722782135,
+      "learning_rate": 3.667107506190993e-05,
+      "loss": 0.4387,
       "step": 50660
     },
     {
-      "epoch": 1.78,
-      "learning_rate": 3.728442381304194e-05,
-      "loss": 0.2812,
+      "epoch": 1.8259631671892458,
+      "grad_norm": 0.14885656535625458,
+      "learning_rate": 3.666849435947526e-05,
+      "loss": 0.4072,
       "step": 50665
     },
     {
-      "epoch": 1.78,
-      "learning_rate": 3.728194264799103e-05,
-      "loss": 0.2777,
+      "epoch": 1.8261433668504703,
+      "grad_norm": 0.20128627121448517,
+      "learning_rate": 3.6665913498059615e-05,
+      "loss": 0.4241,
       "step": 50670
     },
     {
-      "epoch": 1.78,
-      "learning_rate": 3.727946132346719e-05,
-      "loss": 0.3065,
+      "epoch": 1.8263235665116948,
+      "grad_norm": 0.17123094201087952,
+      "learning_rate": 3.666333247769814e-05,
+      "loss": 0.4182,
       "step": 50675
     },
     {
-      "epoch": 1.78,
-      "learning_rate": 3.7276979839502626e-05,
-      "loss": 0.2789,
+      "epoch": 1.8265037661729195,
+      "grad_norm": 0.19727930426597595,
+      "learning_rate": 3.666075129842603e-05,
+      "loss": 0.3987,
       "step": 50680
     },
     {
-      "epoch": 1.78,
-      "learning_rate": 3.727449819612958e-05,
-      "loss": 0.2794,
+      "epoch": 1.8266839658341443,
+      "grad_norm": 0.2271791249513626,
+      "learning_rate": 3.665816996027844e-05,
+      "loss": 0.4111,
       "step": 50685
     },
     {
-      "epoch": 1.78,
-      "learning_rate": 3.7272016393380256e-05,
-      "loss": 0.2922,
+      "epoch": 1.826864165495369,
+      "grad_norm": 0.18186520040035248,
+      "learning_rate": 3.665558846329055e-05,
+      "loss": 0.4242,
       "step": 50690
     },
     {
-      "epoch": 1.78,
-      "learning_rate": 3.726953443128689e-05,
-      "loss": 0.2913,
+      "epoch": 1.8270443651565935,
+      "grad_norm": 0.18122968077659607,
+      "learning_rate": 3.66530068074975e-05,
+      "loss": 0.4055,
       "step": 50695
     },
     {
-      "epoch": 1.78,
-      "learning_rate": 3.726705230988171e-05,
-      "loss": 0.2764,
+      "epoch": 1.827224564817818,
+      "grad_norm": 0.20953094959259033,
+      "learning_rate": 3.6650424992934504e-05,
+      "loss": 0.4515,
       "step": 50700
     },
     {
-      "epoch": 1.78,
-      "learning_rate": 3.7264570029196935e-05,
-      "loss": 0.3015,
+      "epoch": 1.8274047644790428,
+      "grad_norm": 0.1821168065071106,
+      "learning_rate": 3.664784301963673e-05,
+      "loss": 0.4363,
       "step": 50705
     },
     {
-      "epoch": 1.78,
-      "learning_rate": 3.72620875892648e-05,
-      "loss": 0.2876,
+      "epoch": 1.8275849641402675,
+      "grad_norm": 0.16025103628635406,
+      "learning_rate": 3.664526088763934e-05,
+      "loss": 0.3871,
       "step": 50710
     },
     {
-      "epoch": 1.78,
-      "learning_rate": 3.7259604990117536e-05,
-      "loss": 0.2714,
+      "epoch": 1.827765163801492,
+      "grad_norm": 0.1833401620388031,
+      "learning_rate": 3.6642678596977526e-05,
+      "loss": 0.4348,
       "step": 50715
     },
     {
-      "epoch": 1.78,
-      "learning_rate": 3.725712223178738e-05,
-      "loss": 0.2936,
+      "epoch": 1.8279453634627167,
+      "grad_norm": 0.14645916223526,
+      "learning_rate": 3.6640096147686467e-05,
+      "loss": 0.4141,
       "step": 50720
     },
     {
-      "epoch": 1.78,
-      "learning_rate": 3.7254639314306585e-05,
-      "loss": 0.2801,
+      "epoch": 1.8281255631239413,
+      "grad_norm": 0.2093980610370636,
+      "learning_rate": 3.663751353980136e-05,
+      "loss": 0.4121,
       "step": 50725
     },
     {
-      "epoch": 1.78,
-      "learning_rate": 3.725215623770736e-05,
-      "loss": 0.2894,
+      "epoch": 1.828305762785166,
+      "grad_norm": 0.19935870170593262,
+      "learning_rate": 3.663493077335738e-05,
+      "loss": 0.4305,
       "step": 50730
     },
     {
-      "epoch": 1.78,
-      "learning_rate": 3.724967300202196e-05,
-      "loss": 0.2929,
+      "epoch": 1.8284859624463907,
+      "grad_norm": 0.17722731828689575,
+      "learning_rate": 3.663234784838972e-05,
+      "loss": 0.4046,
       "step": 50735
     },
     {
-      "epoch": 1.79,
-      "learning_rate": 3.724718960728264e-05,
-      "loss": 0.2622,
+      "epoch": 1.8286661621076152,
+      "grad_norm": 0.1792575865983963,
+      "learning_rate": 3.662976476493357e-05,
+      "loss": 0.433,
       "step": 50740
     },
     {
-      "epoch": 1.79,
-      "learning_rate": 3.724470605352163e-05,
-      "loss": 0.2968,
+      "epoch": 1.8288463617688397,
+      "grad_norm": 0.15851470828056335,
+      "learning_rate": 3.662718152302413e-05,
+      "loss": 0.417,
       "step": 50745
     },
     {
-      "epoch": 1.79,
-      "learning_rate": 3.7242222340771174e-05,
-      "loss": 0.283,
+      "epoch": 1.8290265614300645,
+      "grad_norm": 0.15749147534370422,
+      "learning_rate": 3.6624598122696595e-05,
+      "loss": 0.4008,
       "step": 50750
     },
     {
-      "epoch": 1.79,
-      "learning_rate": 3.7239738469063535e-05,
-      "loss": 0.2829,
+      "epoch": 1.8292067610912892,
+      "grad_norm": 0.16477638483047485,
+      "learning_rate": 3.6622014563986155e-05,
+      "loss": 0.403,
       "step": 50755
     },
     {
-      "epoch": 1.79,
-      "learning_rate": 3.723725443843096e-05,
-      "loss": 0.3006,
+      "epoch": 1.829386960752514,
+      "grad_norm": 0.16740094125270844,
+      "learning_rate": 3.661943084692802e-05,
+      "loss": 0.4044,
       "step": 50760
     },
     {
-      "epoch": 1.79,
-      "learning_rate": 3.723477024890569e-05,
-      "loss": 0.2904,
+      "epoch": 1.8295671604137385,
+      "grad_norm": 0.2223920226097107,
+      "learning_rate": 3.661684697155739e-05,
+      "loss": 0.4402,
       "step": 50765
     },
     {
-      "epoch": 1.79,
-      "learning_rate": 3.7232285900520006e-05,
-      "loss": 0.2768,
+      "epoch": 1.829747360074963,
+      "grad_norm": 0.21728269755840302,
+      "learning_rate": 3.6614262937909474e-05,
+      "loss": 0.4326,
       "step": 50770
     },
     {
-      "epoch": 1.79,
-      "learning_rate": 3.722980139330614e-05,
-      "loss": 0.2764,
+      "epoch": 1.8299275597361877,
+      "grad_norm": 0.1693524718284607,
+      "learning_rate": 3.6611678746019464e-05,
+      "loss": 0.3812,
       "step": 50775
     },
     {
-      "epoch": 1.79,
-      "learning_rate": 3.722731672729637e-05,
-      "loss": 0.2917,
+      "epoch": 1.8301077593974124,
+      "grad_norm": 0.2092827707529068,
+      "learning_rate": 3.6609094395922585e-05,
+      "loss": 0.4101,
       "step": 50780
     },
     {
-      "epoch": 1.79,
-      "learning_rate": 3.7224831902522944e-05,
-      "loss": 0.2634,
+      "epoch": 1.830287959058637,
+      "grad_norm": 0.19062267243862152,
+      "learning_rate": 3.6606509887654046e-05,
+      "loss": 0.4504,
       "step": 50785
     },
     {
-      "epoch": 1.79,
-      "learning_rate": 3.722234691901814e-05,
-      "loss": 0.2761,
+      "epoch": 1.8304681587198615,
+      "grad_norm": 0.4250420928001404,
+      "learning_rate": 3.660392522124905e-05,
+      "loss": 0.4012,
       "step": 50790
     },
     {
-      "epoch": 1.79,
-      "learning_rate": 3.721986177681421e-05,
-      "loss": 0.2781,
+      "epoch": 1.8306483583810862,
+      "grad_norm": 0.18966785073280334,
+      "learning_rate": 3.660134039674282e-05,
+      "loss": 0.4067,
       "step": 50795
     },
     {
-      "epoch": 1.79,
-      "learning_rate": 3.721737647594343e-05,
-      "loss": 0.2898,
+      "epoch": 1.830828558042311,
+      "grad_norm": 0.17551647126674652,
+      "learning_rate": 3.659875541417057e-05,
+      "loss": 0.408,
       "step": 50800
     },
     {
-      "epoch": 1.79,
-      "learning_rate": 3.721489101643806e-05,
-      "loss": 0.2843,
+      "epoch": 1.8310087577035357,
+      "grad_norm": 0.16543635725975037,
+      "learning_rate": 3.659617027356753e-05,
+      "loss": 0.4234,
       "step": 50805
     },
     {
-      "epoch": 1.79,
-      "learning_rate": 3.72124053983304e-05,
-      "loss": 0.2757,
+      "epoch": 1.8311889573647602,
+      "grad_norm": 0.20097027719020844,
+      "learning_rate": 3.6593584974968916e-05,
+      "loss": 0.4096,
       "step": 50810
     },
     {
-      "epoch": 1.79,
-      "learning_rate": 3.720991962165269e-05,
-      "loss": 0.2847,
+      "epoch": 1.8313691570259847,
+      "grad_norm": 0.1641996055841446,
+      "learning_rate": 3.659099951840995e-05,
+      "loss": 0.4007,
       "step": 50815
     },
     {
-      "epoch": 1.79,
-      "learning_rate": 3.7207433686437224e-05,
-      "loss": 0.2749,
+      "epoch": 1.8315493566872094,
+      "grad_norm": 0.20720233023166656,
+      "learning_rate": 3.658841390392585e-05,
+      "loss": 0.4468,
       "step": 50820
     },
     {
-      "epoch": 1.79,
-      "learning_rate": 3.7204947592716274e-05,
-      "loss": 0.2849,
+      "epoch": 1.8317295563484342,
+      "grad_norm": 0.1876114308834076,
+      "learning_rate": 3.658582813155187e-05,
+      "loss": 0.404,
       "step": 50825
     },
     {
-      "epoch": 1.79,
-      "learning_rate": 3.720295860363726e-05,
-      "loss": 0.2891,
+      "epoch": 1.8319097560096587,
+      "grad_norm": 0.18697507679462433,
+      "learning_rate": 3.658324220132322e-05,
+      "loss": 0.3958,
       "step": 50830
     },
     {
-      "epoch": 1.79,
-      "learning_rate": 3.720047222468779e-05,
-      "loss": 0.2837,
+      "epoch": 1.8320899556708832,
+      "grad_norm": 0.1502731889486313,
+      "learning_rate": 3.658065611327513e-05,
+      "loss": 0.4068,
       "step": 50835
     },
     {
-      "epoch": 1.79,
-      "learning_rate": 3.719798568732323e-05,
-      "loss": 0.2903,
+      "epoch": 1.832270155332108,
+      "grad_norm": 0.18511810898780823,
+      "learning_rate": 3.6578069867442846e-05,
+      "loss": 0.426,
       "step": 50840
     },
     {
-      "epoch": 1.79,
-      "learning_rate": 3.719549899157586e-05,
-      "loss": 0.3028,
+      "epoch": 1.8324503549933326,
+      "grad_norm": 0.1467100828886032,
+      "learning_rate": 3.6575483463861604e-05,
+      "loss": 0.4337,
       "step": 50845
     },
     {
-      "epoch": 1.79,
-      "learning_rate": 3.719301213747798e-05,
-      "loss": 0.2797,
+      "epoch": 1.8326305546545574,
+      "grad_norm": 0.147383913397789,
+      "learning_rate": 3.657289690256664e-05,
+      "loss": 0.3969,
       "step": 50850
     },
     {
-      "epoch": 1.79,
-      "learning_rate": 3.719052512506187e-05,
-      "loss": 0.2833,
+      "epoch": 1.8328107543157819,
+      "grad_norm": 0.16182565689086914,
+      "learning_rate": 3.65703101835932e-05,
+      "loss": 0.4109,
       "step": 50855
     },
     {
-      "epoch": 1.79,
-      "learning_rate": 3.7188037954359826e-05,
-      "loss": 0.2882,
+      "epoch": 1.8329909539770064,
+      "grad_norm": 0.21511298418045044,
+      "learning_rate": 3.656772330697651e-05,
+      "loss": 0.3946,
       "step": 50860
     },
     {
-      "epoch": 1.79,
-      "learning_rate": 3.718555062540414e-05,
-      "loss": 0.2785,
+      "epoch": 1.8331711536382311,
+      "grad_norm": 0.2106020152568817,
+      "learning_rate": 3.6565136272751844e-05,
+      "loss": 0.3559,
       "step": 50865
     },
     {
-      "epoch": 1.79,
-      "learning_rate": 3.7183063138227115e-05,
-      "loss": 0.2619,
+      "epoch": 1.8333513532994559,
+      "grad_norm": 0.22178612649440765,
+      "learning_rate": 3.656254908095443e-05,
+      "loss": 0.4217,
       "step": 50870
     },
     {
-      "epoch": 1.79,
-      "learning_rate": 3.7180575492861034e-05,
-      "loss": 0.2815,
+      "epoch": 1.8335315529606806,
+      "grad_norm": 0.1671254187822342,
+      "learning_rate": 3.655996173161953e-05,
+      "loss": 0.4105,
       "step": 50875
     },
     {
-      "epoch": 1.79,
-      "learning_rate": 3.717808768933823e-05,
-      "loss": 0.2977,
+      "epoch": 1.8337117526219051,
+      "grad_norm": 0.18210411071777344,
+      "learning_rate": 3.6557374224782384e-05,
+      "loss": 0.4154,
       "step": 50880
     },
     {
-      "epoch": 1.79,
-      "learning_rate": 3.717559972769096e-05,
-      "loss": 0.2808,
+      "epoch": 1.8338919522831296,
+      "grad_norm": 0.19609753787517548,
+      "learning_rate": 3.6554786560478245e-05,
+      "loss": 0.4294,
       "step": 50885
     },
     {
-      "epoch": 1.79,
-      "learning_rate": 3.717311160795156e-05,
-      "loss": 0.2925,
+      "epoch": 1.8340721519443544,
+      "grad_norm": 0.16133646667003632,
+      "learning_rate": 3.655219873874238e-05,
+      "loss": 0.425,
       "step": 50890
     },
     {
-      "epoch": 1.79,
-      "learning_rate": 3.7170623330152324e-05,
-      "loss": 0.2829,
+      "epoch": 1.834252351605579,
+      "grad_norm": 0.19089598953723907,
+      "learning_rate": 3.654961075961005e-05,
+      "loss": 0.4355,
       "step": 50895
     },
     {
-      "epoch": 1.79,
-      "learning_rate": 3.716813489432557e-05,
-      "loss": 0.2828,
+      "epoch": 1.8344325512668036,
+      "grad_norm": 0.176603302359581,
+      "learning_rate": 3.654702262311651e-05,
+      "loss": 0.3786,
       "step": 50900
     },
     {
-      "epoch": 1.79,
-      "learning_rate": 3.71656463005036e-05,
-      "loss": 0.2682,
+      "epoch": 1.8346127509280281,
+      "grad_norm": 0.18047241866588593,
+      "learning_rate": 3.654443432929701e-05,
+      "loss": 0.3985,
       "step": 50905
     },
     {
-      "epoch": 1.79,
-      "learning_rate": 3.716315754871873e-05,
-      "loss": 0.3047,
+      "epoch": 1.8347929505892528,
+      "grad_norm": 0.17832064628601074,
+      "learning_rate": 3.654184587818684e-05,
+      "loss": 0.3826,
       "step": 50910
     },
     {
-      "epoch": 1.79,
-      "learning_rate": 3.716066863900327e-05,
-      "loss": 0.2797,
+      "epoch": 1.8349731502504776,
+      "grad_norm": 0.1874881535768509,
+      "learning_rate": 3.653925726982125e-05,
+      "loss": 0.4242,
       "step": 50915
     },
     {
-      "epoch": 1.79,
-      "learning_rate": 3.715817957138954e-05,
-      "loss": 0.2997,
+      "epoch": 1.8351533499117023,
+      "grad_norm": 0.16851532459259033,
+      "learning_rate": 3.653666850423551e-05,
+      "loss": 0.4039,
       "step": 50920
     },
     {
-      "epoch": 1.79,
-      "learning_rate": 3.715569034590988e-05,
-      "loss": 0.2846,
+      "epoch": 1.8353335495729268,
+      "grad_norm": 0.19694097340106964,
+      "learning_rate": 3.65340795814649e-05,
+      "loss": 0.429,
       "step": 50925
     },
     {
-      "epoch": 1.79,
-      "learning_rate": 3.715320096259658e-05,
-      "loss": 0.2702,
+      "epoch": 1.8355137492341513,
+      "grad_norm": 0.1746322214603424,
+      "learning_rate": 3.653149050154469e-05,
+      "loss": 0.406,
       "step": 50930
     },
     {
-      "epoch": 1.79,
-      "learning_rate": 3.7150711421481975e-05,
-      "loss": 0.3049,
+      "epoch": 1.835693948895376,
+      "grad_norm": 0.21489328145980835,
+      "learning_rate": 3.652890126451015e-05,
+      "loss": 0.4699,
       "step": 50935
     },
     {
-      "epoch": 1.79,
-      "learning_rate": 3.714822172259839e-05,
-      "loss": 0.2774,
+      "epoch": 1.8358741485566008,
+      "grad_norm": 0.2216479480266571,
+      "learning_rate": 3.652631187039657e-05,
+      "loss": 0.4144,
       "step": 50940
     },
     {
-      "epoch": 1.79,
-      "learning_rate": 3.714573186597815e-05,
-      "loss": 0.276,
+      "epoch": 1.8360543482178253,
+      "grad_norm": 0.21755515038967133,
+      "learning_rate": 3.6523722319239214e-05,
+      "loss": 0.43,
       "step": 50945
     },
     {
-      "epoch": 1.79,
-      "learning_rate": 3.7143241851653586e-05,
-      "loss": 0.2749,
+      "epoch": 1.8362345478790498,
+      "grad_norm": 0.22237977385520935,
+      "learning_rate": 3.652113261107338e-05,
+      "loss": 0.4274,
       "step": 50950
     },
     {
-      "epoch": 1.79,
-      "learning_rate": 3.714075167965703e-05,
-      "loss": 0.3144,
+      "epoch": 1.8364147475402746,
+      "grad_norm": 0.1792164444923401,
+      "learning_rate": 3.651854274593433e-05,
+      "loss": 0.4141,
       "step": 50955
     },
     {
-      "epoch": 1.79,
-      "learning_rate": 3.713826135002082e-05,
-      "loss": 0.2635,
+      "epoch": 1.8365949472014993,
+      "grad_norm": 0.16846583783626556,
+      "learning_rate": 3.651595272385738e-05,
+      "loss": 0.3788,
       "step": 50960
     },
     {
-      "epoch": 1.79,
-      "learning_rate": 3.7135770862777286e-05,
-      "loss": 0.2974,
+      "epoch": 1.836775146862724,
+      "grad_norm": 0.23637813329696655,
+      "learning_rate": 3.6513362544877794e-05,
+      "loss": 0.4256,
       "step": 50965
     },
     {
-      "epoch": 1.79,
-      "learning_rate": 3.713328021795875e-05,
-      "loss": 0.2894,
+      "epoch": 1.8369553465239485,
+      "grad_norm": 0.21190734207630157,
+      "learning_rate": 3.6510772209030885e-05,
+      "loss": 0.4456,
       "step": 50970
     },
     {
-      "epoch": 1.79,
-      "learning_rate": 3.7130789415597587e-05,
-      "loss": 0.2923,
+      "epoch": 1.837135546185173,
+      "grad_norm": 0.1966804713010788,
+      "learning_rate": 3.650818171635192e-05,
+      "loss": 0.3872,
       "step": 50975
     },
     {
-      "epoch": 1.79,
-      "learning_rate": 3.7128298455726104e-05,
-      "loss": 0.3008,
+      "epoch": 1.8373157458463978,
+      "grad_norm": 0.24481886625289917,
+      "learning_rate": 3.650559106687621e-05,
+      "loss": 0.4143,
       "step": 50980
     },
     {
-      "epoch": 1.79,
-      "learning_rate": 3.712580733837666e-05,
-      "loss": 0.2606,
+      "epoch": 1.8374959455076225,
+      "grad_norm": 0.2108885794878006,
+      "learning_rate": 3.650300026063905e-05,
+      "loss": 0.3876,
       "step": 50985
     },
     {
-      "epoch": 1.79,
-      "learning_rate": 3.71233160635816e-05,
-      "loss": 0.3075,
+      "epoch": 1.837676145168847,
+      "grad_norm": 0.17803071439266205,
+      "learning_rate": 3.650040929767575e-05,
+      "loss": 0.3957,
       "step": 50990
     },
     {
-      "epoch": 1.79,
-      "learning_rate": 3.7120824631373274e-05,
-      "loss": 0.2828,
+      "epoch": 1.8378563448300718,
+      "grad_norm": 0.1737845540046692,
+      "learning_rate": 3.649781817802159e-05,
+      "loss": 0.3848,
       "step": 50995
     },
     {
-      "epoch": 1.79,
-      "learning_rate": 3.7118333041784024e-05,
-      "loss": 0.2842,
+      "epoch": 1.8380365444912963,
+      "grad_norm": 0.16461852192878723,
+      "learning_rate": 3.649522690171188e-05,
+      "loss": 0.41,
       "step": 51000
     },
     {
-      "epoch": 1.79,
-      "eval_loss": 0.2750702500343323,
-      "eval_runtime": 10.538,
-      "eval_samples_per_second": 9.489,
-      "eval_steps_per_second": 9.489,
+      "epoch": 1.8380365444912963,
+      "eval_loss": 0.4396840035915375,
+      "eval_runtime": 3.5402,
+      "eval_samples_per_second": 28.247,
+      "eval_steps_per_second": 7.062,
       "step": 51000
     },
     {
-      "epoch": 1.79,
-      "learning_rate": 3.711584129484621e-05,
-      "loss": 0.2738,
+      "epoch": 1.838216744152521,
+      "grad_norm": 0.19908668100833893,
+      "learning_rate": 3.6492635468781934e-05,
+      "loss": 0.411,
       "step": 51005
     },
     {
-      "epoch": 1.79,
-      "learning_rate": 3.711334939059218e-05,
-      "loss": 0.2937,
+      "epoch": 1.8383969438137457,
+      "grad_norm": 0.23171748220920563,
+      "learning_rate": 3.649004387926705e-05,
+      "loss": 0.3974,
       "step": 51010
     },
     {
-      "epoch": 1.79,
-      "learning_rate": 3.7110857329054295e-05,
-      "loss": 0.29,
+      "epoch": 1.8385771434749703,
+      "grad_norm": 0.19724468886852264,
+      "learning_rate": 3.6487452133202546e-05,
+      "loss": 0.4076,
       "step": 51015
     },
     {
-      "epoch": 1.8,
-      "learning_rate": 3.7108365110264895e-05,
-      "loss": 0.2818,
+      "epoch": 1.8387573431361948,
+      "grad_norm": 0.1985427588224411,
+      "learning_rate": 3.648486023062373e-05,
+      "loss": 0.45,
       "step": 51020
     },
     {
-      "epoch": 1.8,
-      "learning_rate": 3.710587273425636e-05,
-      "loss": 0.2837,
+      "epoch": 1.8389375427974195,
+      "grad_norm": 0.16115103662014008,
+      "learning_rate": 3.6482268171565914e-05,
+      "loss": 0.3991,
       "step": 51025
     },
     {
-      "epoch": 1.8,
-      "learning_rate": 3.710338020106106e-05,
-      "loss": 0.2928,
+      "epoch": 1.8391177424586442,
+      "grad_norm": 0.20082534849643707,
+      "learning_rate": 3.647967595606443e-05,
+      "loss": 0.4553,
       "step": 51030
     },
     {
-      "epoch": 1.8,
-      "learning_rate": 3.7100887510711326e-05,
-      "loss": 0.3059,
+      "epoch": 1.839297942119869,
+      "grad_norm": 0.1648685485124588,
+      "learning_rate": 3.647708358415457e-05,
+      "loss": 0.4469,
       "step": 51035
     },
     {
-      "epoch": 1.8,
-      "learning_rate": 3.709839466323956e-05,
-      "loss": 0.288,
+      "epoch": 1.8394781417810935,
+      "grad_norm": 0.16533087193965912,
+      "learning_rate": 3.6474491055871675e-05,
+      "loss": 0.3725,
       "step": 51040
     },
     {
-      "epoch": 1.8,
-      "learning_rate": 3.709590165867809e-05,
-      "loss": 0.2917,
+      "epoch": 1.839658341442318,
+      "grad_norm": 0.18278469145298004,
+      "learning_rate": 3.647189837125106e-05,
+      "loss": 0.427,
       "step": 51045
     },
     {
-      "epoch": 1.8,
-      "learning_rate": 3.7093408497059325e-05,
-      "loss": 0.2823,
+      "epoch": 1.8398385411035427,
+      "grad_norm": 0.18159344792366028,
+      "learning_rate": 3.646930553032805e-05,
+      "loss": 0.4254,
       "step": 51050
     },
     {
-      "epoch": 1.8,
-      "learning_rate": 3.709091517841561e-05,
-      "loss": 0.2821,
+      "epoch": 1.8400187407647675,
+      "grad_norm": 0.16814720630645752,
+      "learning_rate": 3.646671253313797e-05,
+      "loss": 0.4013,
       "step": 51055
     },
     {
-      "epoch": 1.8,
-      "learning_rate": 3.708842170277934e-05,
-      "loss": 0.29,
+      "epoch": 1.840198940425992,
+      "grad_norm": 0.15060807764530182,
+      "learning_rate": 3.646411937971615e-05,
+      "loss": 0.3897,
       "step": 51060
     },
     {
-      "epoch": 1.8,
-      "learning_rate": 3.708592807018287e-05,
-      "loss": 0.2864,
+      "epoch": 1.8403791400872165,
+      "grad_norm": 0.19889427721500397,
+      "learning_rate": 3.646152607009793e-05,
+      "loss": 0.436,
       "step": 51065
     },
     {
-      "epoch": 1.8,
-      "learning_rate": 3.7083434280658605e-05,
-      "loss": 0.3002,
+      "epoch": 1.8405593397484412,
+      "grad_norm": 0.22724641859531403,
+      "learning_rate": 3.645893260431863e-05,
+      "loss": 0.4595,
       "step": 51070
     },
     {
-      "epoch": 1.8,
-      "learning_rate": 3.70809403342389e-05,
-      "loss": 0.2825,
+      "epoch": 1.840739539409666,
+      "grad_norm": 0.19296084344387054,
+      "learning_rate": 3.64563389824136e-05,
+      "loss": 0.3611,
       "step": 51075
     },
     {
-      "epoch": 1.8,
-      "learning_rate": 3.7078446230956143e-05,
-      "loss": 0.2741,
+      "epoch": 1.8409197390708907,
+      "grad_norm": 0.16096292436122894,
+      "learning_rate": 3.6453745204418164e-05,
+      "loss": 0.4039,
       "step": 51080
     },
     {
-      "epoch": 1.8,
-      "learning_rate": 3.707595197084272e-05,
-      "loss": 0.2888,
+      "epoch": 1.8410999387321152,
+      "grad_norm": 0.18182432651519775,
+      "learning_rate": 3.6451151270367665e-05,
+      "loss": 0.3959,
       "step": 51085
     },
     {
-      "epoch": 1.8,
-      "learning_rate": 3.707345755393103e-05,
-      "loss": 0.3029,
+      "epoch": 1.8412801383933397,
+      "grad_norm": 0.14909334480762482,
+      "learning_rate": 3.644855718029745e-05,
+      "loss": 0.4084,
       "step": 51090
     },
     {
-      "epoch": 1.8,
-      "learning_rate": 3.707096298025344e-05,
-      "loss": 0.2536,
+      "epoch": 1.8414603380545644,
+      "grad_norm": 0.2044646441936493,
+      "learning_rate": 3.644596293424286e-05,
+      "loss": 0.4397,
       "step": 51095
     },
     {
-      "epoch": 1.8,
-      "learning_rate": 3.706846824984236e-05,
-      "loss": 0.2956,
+      "epoch": 1.8416405377157892,
+      "grad_norm": 0.20742791891098022,
+      "learning_rate": 3.644336853223924e-05,
+      "loss": 0.4362,
       "step": 51100
     },
     {
-      "epoch": 1.8,
-      "learning_rate": 3.7065973362730164e-05,
-      "loss": 0.2647,
+      "epoch": 1.8418207373770137,
+      "grad_norm": 0.17310570180416107,
+      "learning_rate": 3.644077397432194e-05,
+      "loss": 0.4036,
       "step": 51105
     },
     {
-      "epoch": 1.8,
-      "learning_rate": 3.7063478318949255e-05,
-      "loss": 0.3026,
+      "epoch": 1.8420009370382384,
+      "grad_norm": 0.17185135185718536,
+      "learning_rate": 3.64381792605263e-05,
+      "loss": 0.4234,
       "step": 51110
     },
     {
-      "epoch": 1.8,
-      "learning_rate": 3.706098311853204e-05,
-      "loss": 0.2727,
+      "epoch": 1.842181136699463,
+      "grad_norm": 0.18162250518798828,
+      "learning_rate": 3.64355843908877e-05,
+      "loss": 0.396,
       "step": 51115
     },
     {
-      "epoch": 1.8,
-      "learning_rate": 3.70584877615109e-05,
-      "loss": 0.2703,
+      "epoch": 1.8423613363606877,
+      "grad_norm": 0.18723055720329285,
+      "learning_rate": 3.6432989365441474e-05,
+      "loss": 0.3928,
       "step": 51120
     },
     {
-      "epoch": 1.8,
-      "learning_rate": 3.7055992247918247e-05,
-      "loss": 0.2926,
+      "epoch": 1.8425415360219124,
+      "grad_norm": 0.21605227887630463,
+      "learning_rate": 3.643039418422297e-05,
+      "loss": 0.3746,
       "step": 51125
     },
     {
-      "epoch": 1.8,
-      "learning_rate": 3.7053496577786485e-05,
-      "loss": 0.3351,
+      "epoch": 1.842721735683137,
+      "grad_norm": 0.15051497519016266,
+      "learning_rate": 3.642779884726757e-05,
+      "loss": 0.4126,
       "step": 51130
     },
     {
-      "epoch": 1.8,
-      "learning_rate": 3.7051000751148e-05,
-      "loss": 0.2617,
+      "epoch": 1.8429019353443614,
+      "grad_norm": 0.19938229024410248,
+      "learning_rate": 3.642520335461061e-05,
+      "loss": 0.4506,
       "step": 51135
     },
     {
-      "epoch": 1.8,
-      "learning_rate": 3.704850476803522e-05,
-      "loss": 0.2807,
+      "epoch": 1.8430821350055862,
+      "grad_norm": 0.17573609948158264,
+      "learning_rate": 3.642260770628747e-05,
+      "loss": 0.4178,
       "step": 51140
     },
     {
-      "epoch": 1.8,
-      "learning_rate": 3.704600862848054e-05,
-      "loss": 0.2904,
+      "epoch": 1.8432623346668109,
+      "grad_norm": 0.16785641014575958,
+      "learning_rate": 3.6420011902333516e-05,
+      "loss": 0.4336,
       "step": 51145
     },
     {
-      "epoch": 1.8,
-      "learning_rate": 3.704351233251638e-05,
-      "loss": 0.3106,
+      "epoch": 1.8434425343280356,
+      "grad_norm": 0.21680709719657898,
+      "learning_rate": 3.641741594278411e-05,
+      "loss": 0.4317,
       "step": 51150
     },
     {
-      "epoch": 1.8,
-      "learning_rate": 3.704101588017516e-05,
-      "loss": 0.2817,
+      "epoch": 1.8436227339892601,
+      "grad_norm": 0.18228915333747864,
+      "learning_rate": 3.641481982767463e-05,
+      "loss": 0.4118,
       "step": 51155
     },
     {
-      "epoch": 1.8,
-      "learning_rate": 3.703851927148927e-05,
-      "loss": 0.2982,
+      "epoch": 1.8438029336504846,
+      "grad_norm": 0.2226298749446869,
+      "learning_rate": 3.641222355704042e-05,
+      "loss": 0.4553,
       "step": 51160
     },
     {
-      "epoch": 1.8,
-      "learning_rate": 3.7036022506491146e-05,
-      "loss": 0.2926,
+      "epoch": 1.8439831333117094,
+      "grad_norm": 0.20143426954746246,
+      "learning_rate": 3.640962713091689e-05,
+      "loss": 0.43,
       "step": 51165
     },
     {
-      "epoch": 1.8,
-      "learning_rate": 3.7033525585213205e-05,
-      "loss": 0.2752,
+      "epoch": 1.844163332972934,
+      "grad_norm": 0.1602238118648529,
+      "learning_rate": 3.64070305493394e-05,
+      "loss": 0.4138,
       "step": 51170
     },
     {
-      "epoch": 1.8,
-      "learning_rate": 3.703102850768786e-05,
-      "loss": 0.2834,
+      "epoch": 1.8443435326341586,
+      "grad_norm": 0.19529132544994354,
+      "learning_rate": 3.640443381234331e-05,
+      "loss": 0.4168,
       "step": 51175
     },
     {
-      "epoch": 1.8,
-      "learning_rate": 3.7028531273947536e-05,
-      "loss": 0.3048,
+      "epoch": 1.8445237322953831,
+      "grad_norm": 0.19434118270874023,
+      "learning_rate": 3.6401836919964035e-05,
+      "loss": 0.4032,
       "step": 51180
     },
     {
-      "epoch": 1.8,
-      "learning_rate": 3.702603388402467e-05,
-      "loss": 0.2871,
+      "epoch": 1.8447039319566079,
+      "grad_norm": 0.1953050196170807,
+      "learning_rate": 3.639923987223693e-05,
+      "loss": 0.4151,
       "step": 51185
     },
     {
-      "epoch": 1.8,
-      "learning_rate": 3.702353633795167e-05,
-      "loss": 0.28,
+      "epoch": 1.8448841316178326,
+      "grad_norm": 0.16159437596797943,
+      "learning_rate": 3.639664266919739e-05,
+      "loss": 0.4209,
       "step": 51190
     },
     {
-      "epoch": 1.8,
-      "learning_rate": 3.702103863576098e-05,
-      "loss": 0.2954,
+      "epoch": 1.8450643312790573,
+      "grad_norm": 0.22107087075710297,
+      "learning_rate": 3.639404531088081e-05,
+      "loss": 0.4232,
       "step": 51195
     },
     {
-      "epoch": 1.8,
-      "learning_rate": 3.7018540777485024e-05,
-      "loss": 0.2968,
+      "epoch": 1.8452445309402818,
+      "grad_norm": 0.1914900243282318,
+      "learning_rate": 3.6391447797322556e-05,
+      "loss": 0.4235,
       "step": 51200
     },
     {
-      "epoch": 1.8,
-      "learning_rate": 3.7016042763156236e-05,
-      "loss": 0.2944,
+      "epoch": 1.8454247306015064,
+      "grad_norm": 0.1916929930448532,
+      "learning_rate": 3.6388850128558036e-05,
+      "loss": 0.4633,
       "step": 51205
     },
     {
-      "epoch": 1.8,
-      "learning_rate": 3.7013544592807054e-05,
-      "loss": 0.2831,
+      "epoch": 1.845604930262731,
+      "grad_norm": 0.16584375500679016,
+      "learning_rate": 3.6386252304622636e-05,
+      "loss": 0.4198,
       "step": 51210
     },
     {
-      "epoch": 1.8,
-      "learning_rate": 3.7011046266469905e-05,
-      "loss": 0.2764,
+      "epoch": 1.8457851299239558,
+      "grad_norm": 0.19661644101142883,
+      "learning_rate": 3.6383654325551756e-05,
+      "loss": 0.3869,
       "step": 51215
     },
     {
-      "epoch": 1.8,
-      "learning_rate": 3.7008547784177246e-05,
-      "loss": 0.2949,
+      "epoch": 1.8459653295851803,
+      "grad_norm": 0.17381379008293152,
+      "learning_rate": 3.638105619138079e-05,
+      "loss": 0.3688,
       "step": 51220
     },
     {
-      "epoch": 1.8,
-      "learning_rate": 3.70060491459615e-05,
-      "loss": 0.2917,
+      "epoch": 1.846145529246405,
+      "grad_norm": 0.18140283226966858,
+      "learning_rate": 3.637845790214513e-05,
+      "loss": 0.464,
       "step": 51225
     },
     {
-      "epoch": 1.8,
-      "learning_rate": 3.7003550351855124e-05,
-      "loss": 0.2698,
+      "epoch": 1.8463257289076296,
+      "grad_norm": 0.18589642643928528,
+      "learning_rate": 3.637585945788019e-05,
+      "loss": 0.4818,
       "step": 51230
     },
     {
-      "epoch": 1.8,
-      "learning_rate": 3.700105140189056e-05,
-      "loss": 0.2764,
+      "epoch": 1.8465059285688543,
+      "grad_norm": 0.1755005270242691,
+      "learning_rate": 3.6373260858621364e-05,
+      "loss": 0.4101,
       "step": 51235
     },
     {
-      "epoch": 1.8,
-      "learning_rate": 3.699855229610024e-05,
-      "loss": 0.3163,
+      "epoch": 1.846686128230079,
+      "grad_norm": 0.14478670060634613,
+      "learning_rate": 3.637066210440407e-05,
+      "loss": 0.3988,
       "step": 51240
     },
     {
-      "epoch": 1.8,
-      "learning_rate": 3.6996053034516634e-05,
-      "loss": 0.2672,
+      "epoch": 1.8468663278913036,
+      "grad_norm": 0.1898811310529709,
+      "learning_rate": 3.6368063195263694e-05,
+      "loss": 0.3998,
       "step": 51245
     },
     {
-      "epoch": 1.8,
-      "learning_rate": 3.699355361717218e-05,
-      "loss": 0.2975,
+      "epoch": 1.847046527552528,
+      "grad_norm": 0.2270967662334442,
+      "learning_rate": 3.6365464131235664e-05,
+      "loss": 0.4444,
       "step": 51250
     },
     {
-      "epoch": 1.8,
-      "learning_rate": 3.699105404409935e-05,
-      "loss": 0.3058,
+      "epoch": 1.8472267272137528,
+      "grad_norm": 0.17820177972316742,
+      "learning_rate": 3.636286491235538e-05,
+      "loss": 0.4097,
       "step": 51255
     },
     {
-      "epoch": 1.8,
-      "learning_rate": 3.698855431533057e-05,
-      "loss": 0.2785,
+      "epoch": 1.8474069268749775,
+      "grad_norm": 0.19704335927963257,
+      "learning_rate": 3.636026553865828e-05,
+      "loss": 0.4587,
       "step": 51260
     },
     {
-      "epoch": 1.8,
-      "learning_rate": 3.698605443089833e-05,
-      "loss": 0.2886,
+      "epoch": 1.8475871265362023,
+      "grad_norm": 0.18560896813869476,
+      "learning_rate": 3.635766601017974e-05,
+      "loss": 0.4076,
       "step": 51265
     },
     {
-      "epoch": 1.8,
-      "learning_rate": 3.698355439083506e-05,
-      "loss": 0.2679,
+      "epoch": 1.8477673261974268,
+      "grad_norm": 0.19222392141819,
+      "learning_rate": 3.635506632695521e-05,
+      "loss": 0.4196,
       "step": 51270
     },
     {
-      "epoch": 1.8,
-      "learning_rate": 3.698105419517323e-05,
-      "loss": 0.2909,
+      "epoch": 1.8479475258586513,
+      "grad_norm": 0.1405927538871765,
+      "learning_rate": 3.6352466489020095e-05,
+      "loss": 0.4081,
       "step": 51275
     },
     {
-      "epoch": 1.8,
-      "learning_rate": 3.6978553843945306e-05,
-      "loss": 0.2745,
+      "epoch": 1.848127725519876,
+      "grad_norm": 0.16948674619197845,
+      "learning_rate": 3.634986649640982e-05,
+      "loss": 0.3944,
       "step": 51280
     },
     {
-      "epoch": 1.8,
-      "learning_rate": 3.697605333718376e-05,
-      "loss": 0.2649,
+      "epoch": 1.8483079251811008,
+      "grad_norm": 0.16560353338718414,
+      "learning_rate": 3.6347266349159826e-05,
+      "loss": 0.4439,
       "step": 51285
     },
     {
-      "epoch": 1.8,
-      "learning_rate": 3.6973552674921067e-05,
-      "loss": 0.2923,
+      "epoch": 1.8484881248423253,
+      "grad_norm": 0.15636521577835083,
+      "learning_rate": 3.634466604730551e-05,
+      "loss": 0.4157,
       "step": 51290
     },
     {
-      "epoch": 1.8,
-      "learning_rate": 3.697105185718967e-05,
-      "loss": 0.2965,
+      "epoch": 1.8486683245035498,
+      "grad_norm": 0.19085612893104553,
+      "learning_rate": 3.634206559088232e-05,
+      "loss": 0.4038,
       "step": 51295
     },
     {
-      "epoch": 1.8,
-      "learning_rate": 3.696855088402204e-05,
-      "loss": 0.2692,
+      "epoch": 1.8488485241647745,
+      "grad_norm": 0.1767541766166687,
+      "learning_rate": 3.633946497992568e-05,
+      "loss": 0.4058,
       "step": 51300
     },
     {
-      "epoch": 1.81,
-      "learning_rate": 3.696604975545068e-05,
-      "loss": 0.2756,
+      "epoch": 1.8490287238259993,
+      "grad_norm": 0.1703364998102188,
+      "learning_rate": 3.6336864214471035e-05,
+      "loss": 0.3893,
       "step": 51305
     },
     {
-      "epoch": 1.81,
-      "learning_rate": 3.696354847150806e-05,
-      "loss": 0.3012,
+      "epoch": 1.849208923487224,
+      "grad_norm": 0.19773104786872864,
+      "learning_rate": 3.63342632945538e-05,
+      "loss": 0.4547,
       "step": 51310
     },
     {
-      "epoch": 1.81,
-      "learning_rate": 3.696104703222663e-05,
-      "loss": 0.267,
+      "epoch": 1.8493891231484485,
+      "grad_norm": 0.20999063551425934,
+      "learning_rate": 3.6331662220209416e-05,
+      "loss": 0.4344,
       "step": 51315
     },
     {
-      "epoch": 1.81,
-      "learning_rate": 3.6958545437638896e-05,
-      "loss": 0.2901,
+      "epoch": 1.849569322809673,
+      "grad_norm": 0.16823193430900574,
+      "learning_rate": 3.6329060991473344e-05,
+      "loss": 0.4112,
       "step": 51320
     },
     {
-      "epoch": 1.81,
-      "learning_rate": 3.695604368777733e-05,
-      "loss": 0.2705,
+      "epoch": 1.8497495224708977,
+      "grad_norm": 0.21957112848758698,
+      "learning_rate": 3.6326459608380994e-05,
+      "loss": 0.3883,
       "step": 51325
     },
     {
-      "epoch": 1.81,
-      "learning_rate": 3.695354178267441e-05,
-      "loss": 0.2869,
+      "epoch": 1.8499297221321225,
+      "grad_norm": 0.17172305285930634,
+      "learning_rate": 3.6323858070967834e-05,
+      "loss": 0.3975,
       "step": 51330
     },
     {
-      "epoch": 1.81,
-      "learning_rate": 3.695103972236263e-05,
-      "loss": 0.2809,
+      "epoch": 1.850109921793347,
+      "grad_norm": 0.23369702696800232,
+      "learning_rate": 3.6321256379269296e-05,
+      "loss": 0.4123,
       "step": 51335
     },
     {
-      "epoch": 1.81,
-      "learning_rate": 3.694853750687448e-05,
-      "loss": 0.2787,
+      "epoch": 1.8502901214545715,
+      "grad_norm": 0.19112427532672882,
+      "learning_rate": 3.631865453332084e-05,
+      "loss": 0.4109,
       "step": 51340
     },
     {
-      "epoch": 1.81,
-      "learning_rate": 3.694603513624244e-05,
-      "loss": 0.2472,
+      "epoch": 1.8504703211157962,
+      "grad_norm": 0.19734065234661102,
+      "learning_rate": 3.63160525331579e-05,
+      "loss": 0.4006,
       "step": 51345
     },
     {
-      "epoch": 1.81,
-      "learning_rate": 3.6943532610499014e-05,
-      "loss": 0.2754,
+      "epoch": 1.850650520777021,
+      "grad_norm": 0.15451741218566895,
+      "learning_rate": 3.631345037881593e-05,
+      "loss": 0.4063,
       "step": 51350
     },
     {
-      "epoch": 1.81,
-      "learning_rate": 3.6941029929676675e-05,
-      "loss": 0.2844,
+      "epoch": 1.8508307204382457,
+      "grad_norm": 0.1700822114944458,
+      "learning_rate": 3.631084807033041e-05,
+      "loss": 0.4052,
       "step": 51355
     },
     {
-      "epoch": 1.81,
-      "learning_rate": 3.693852709380794e-05,
-      "loss": 0.2816,
+      "epoch": 1.8510109200994702,
+      "grad_norm": 0.1707632839679718,
+      "learning_rate": 3.630824560773676e-05,
+      "loss": 0.3775,
       "step": 51360
     },
     {
-      "epoch": 1.81,
-      "learning_rate": 3.69360241029253e-05,
-      "loss": 0.2789,
+      "epoch": 1.8511911197606947,
+      "grad_norm": 0.16627883911132812,
+      "learning_rate": 3.630564299107045e-05,
+      "loss": 0.4075,
       "step": 51365
     },
     {
-      "epoch": 1.81,
-      "learning_rate": 3.693352095706125e-05,
-      "loss": 0.3022,
+      "epoch": 1.8513713194219195,
+      "grad_norm": 0.24305926263332367,
+      "learning_rate": 3.630304022036694e-05,
+      "loss": 0.4367,
       "step": 51370
     },
     {
-      "epoch": 1.81,
-      "learning_rate": 3.69310176562483e-05,
-      "loss": 0.2998,
+      "epoch": 1.8515515190831442,
+      "grad_norm": 0.15088029205799103,
+      "learning_rate": 3.6300437295661706e-05,
+      "loss": 0.398,
       "step": 51375
     },
     {
-      "epoch": 1.81,
-      "learning_rate": 3.692851420051894e-05,
-      "loss": 0.2976,
+      "epoch": 1.851731718744369,
+      "grad_norm": 0.16519147157669067,
+      "learning_rate": 3.62978342169902e-05,
+      "loss": 0.4295,
       "step": 51380
     },
     {
-      "epoch": 1.81,
-      "learning_rate": 3.6926010589905694e-05,
-      "loss": 0.2608,
+      "epoch": 1.8519119184055934,
+      "grad_norm": 0.19771799445152283,
+      "learning_rate": 3.6295230984387884e-05,
+      "loss": 0.3928,
       "step": 51385
     },
     {
-      "epoch": 1.81,
-      "learning_rate": 3.692350682444104e-05,
-      "loss": 0.2994,
+      "epoch": 1.852092118066818,
+      "grad_norm": 0.164763942360878,
+      "learning_rate": 3.629262759789024e-05,
+      "loss": 0.4141,
       "step": 51390
     },
     {
-      "epoch": 1.81,
-      "learning_rate": 3.6921002904157533e-05,
-      "loss": 0.28,
+      "epoch": 1.8522723177280427,
+      "grad_norm": 0.18047907948493958,
+      "learning_rate": 3.6290024057532726e-05,
+      "loss": 0.4259,
       "step": 51395
     },
     {
-      "epoch": 1.81,
-      "learning_rate": 3.691849882908764e-05,
-      "loss": 0.2896,
+      "epoch": 1.8524525173892674,
+      "grad_norm": 0.21674178540706635,
+      "learning_rate": 3.6287420363350824e-05,
+      "loss": 0.3962,
       "step": 51400
     },
     {
-      "epoch": 1.81,
-      "learning_rate": 3.691599459926391e-05,
-      "loss": 0.2934,
+      "epoch": 1.852632717050492,
+      "grad_norm": 0.1720811426639557,
+      "learning_rate": 3.628481651538e-05,
+      "loss": 0.4129,
       "step": 51405
     },
     {
-      "epoch": 1.81,
-      "learning_rate": 3.691349021471884e-05,
-      "loss": 0.286,
+      "epoch": 1.8528129167117164,
+      "grad_norm": 0.201126828789711,
+      "learning_rate": 3.628221251365574e-05,
+      "loss": 0.4334,
       "step": 51410
     },
     {
-      "epoch": 1.81,
-      "learning_rate": 3.691098567548494e-05,
-      "loss": 0.2833,
+      "epoch": 1.8529931163729412,
+      "grad_norm": 0.17763501405715942,
+      "learning_rate": 3.627960835821351e-05,
+      "loss": 0.4179,
       "step": 51415
     },
     {
-      "epoch": 1.81,
-      "learning_rate": 3.690848098159475e-05,
-      "loss": 0.283,
+      "epoch": 1.853173316034166,
+      "grad_norm": 0.16252952814102173,
+      "learning_rate": 3.6277004049088815e-05,
+      "loss": 0.3806,
       "step": 51420
     },
     {
-      "epoch": 1.81,
-      "learning_rate": 3.6905976133080785e-05,
-      "loss": 0.266,
+      "epoch": 1.8533535156953906,
+      "grad_norm": 0.17400288581848145,
+      "learning_rate": 3.627439958631712e-05,
+      "loss": 0.3974,
       "step": 51425
     },
     {
-      "epoch": 1.81,
-      "learning_rate": 3.6903471129975556e-05,
-      "loss": 0.2611,
+      "epoch": 1.8535337153566152,
+      "grad_norm": 0.1547224074602127,
+      "learning_rate": 3.6271794969933895e-05,
+      "loss": 0.373,
       "step": 51430
     },
     {
-      "epoch": 1.81,
-      "learning_rate": 3.69009659723116e-05,
-      "loss": 0.2934,
+      "epoch": 1.8537139150178397,
+      "grad_norm": 0.17776109278202057,
+      "learning_rate": 3.626919019997467e-05,
+      "loss": 0.3768,
       "step": 51435
     },
     {
-      "epoch": 1.81,
-      "learning_rate": 3.689846066012145e-05,
-      "loss": 0.2724,
+      "epoch": 1.8538941146790644,
+      "grad_norm": 0.188429594039917,
+      "learning_rate": 3.6266585276474896e-05,
+      "loss": 0.4463,
       "step": 51440
     },
     {
-      "epoch": 1.81,
-      "learning_rate": 3.689595519343763e-05,
-      "loss": 0.2909,
+      "epoch": 1.8540743143402891,
+      "grad_norm": 0.18767085671424866,
+      "learning_rate": 3.626398019947008e-05,
+      "loss": 0.4271,
       "step": 51445
     },
     {
-      "epoch": 1.81,
-      "learning_rate": 3.689344957229267e-05,
-      "loss": 0.2729,
+      "epoch": 1.8542545140015136,
+      "grad_norm": 0.19436992704868317,
+      "learning_rate": 3.626137496899572e-05,
+      "loss": 0.399,
       "step": 51450
     },
     {
-      "epoch": 1.81,
-      "learning_rate": 3.689094379671911e-05,
-      "loss": 0.2945,
+      "epoch": 1.8544347136627382,
+      "grad_norm": 0.2516634166240692,
+      "learning_rate": 3.62587695850873e-05,
+      "loss": 0.4383,
       "step": 51455
     },
     {
-      "epoch": 1.81,
-      "learning_rate": 3.688843786674947e-05,
-      "loss": 0.2714,
+      "epoch": 1.8546149133239629,
+      "grad_norm": 0.18735727667808533,
+      "learning_rate": 3.625616404778033e-05,
+      "loss": 0.4246,
       "step": 51460
     },
     {
-      "epoch": 1.81,
-      "learning_rate": 3.6885931782416304e-05,
-      "loss": 0.2573,
+      "epoch": 1.8547951129851876,
+      "grad_norm": 0.19105082750320435,
+      "learning_rate": 3.62535583571103e-05,
+      "loss": 0.4179,
       "step": 51465
     },
     {
-      "epoch": 1.81,
-      "learning_rate": 3.6883425543752145e-05,
-      "loss": 0.2973,
+      "epoch": 1.8549753126464124,
+      "grad_norm": 0.20437583327293396,
+      "learning_rate": 3.625095251311272e-05,
+      "loss": 0.4275,
       "step": 51470
     },
     {
-      "epoch": 1.81,
-      "learning_rate": 3.688091915078955e-05,
-      "loss": 0.2655,
+      "epoch": 1.8551555123076369,
+      "grad_norm": 0.1884140968322754,
+      "learning_rate": 3.6248346515823084e-05,
+      "loss": 0.3885,
       "step": 51475
     },
     {
-      "epoch": 1.81,
-      "learning_rate": 3.687841260356103e-05,
-      "loss": 0.3057,
+      "epoch": 1.8553357119688614,
+      "grad_norm": 0.22424067556858063,
+      "learning_rate": 3.6245740365276914e-05,
+      "loss": 0.4377,
       "step": 51480
     },
     {
-      "epoch": 1.81,
-      "learning_rate": 3.6875905902099166e-05,
-      "loss": 0.2944,
+      "epoch": 1.8555159116300861,
+      "grad_norm": 0.1814982295036316,
+      "learning_rate": 3.62431340615097e-05,
+      "loss": 0.4373,
       "step": 51485
     },
     {
-      "epoch": 1.81,
-      "learning_rate": 3.687339904643648e-05,
-      "loss": 0.2956,
+      "epoch": 1.8556961112913108,
+      "grad_norm": 0.18222209811210632,
+      "learning_rate": 3.624052760455696e-05,
+      "loss": 0.3891,
       "step": 51490
     },
     {
-      "epoch": 1.81,
-      "learning_rate": 3.687089203660554e-05,
-      "loss": 0.2672,
+      "epoch": 1.8558763109525354,
+      "grad_norm": 0.21884121000766754,
+      "learning_rate": 3.6237920994454216e-05,
+      "loss": 0.436,
       "step": 51495
     },
     {
-      "epoch": 1.81,
-      "learning_rate": 3.6868384872638886e-05,
-      "loss": 0.3063,
+      "epoch": 1.85605651061376,
+      "grad_norm": 0.16417382657527924,
+      "learning_rate": 3.623531423123697e-05,
+      "loss": 0.4172,
       "step": 51500
     },
     {
-      "epoch": 1.81,
-      "eval_loss": 0.274717777967453,
-      "eval_runtime": 10.5391,
-      "eval_samples_per_second": 9.488,
-      "eval_steps_per_second": 9.488,
+      "epoch": 1.85605651061376,
+      "eval_loss": 0.43941619992256165,
+      "eval_runtime": 3.5304,
+      "eval_samples_per_second": 28.326,
+      "eval_steps_per_second": 7.081,
       "step": 51500
     },
     {
-      "epoch": 1.81,
-      "learning_rate": 3.686587755456907e-05,
-      "loss": 0.2998,
+      "epoch": 1.8562367102749846,
+      "grad_norm": 0.20107519626617432,
+      "learning_rate": 3.623270731494075e-05,
+      "loss": 0.4246,
       "step": 51505
     },
     {
-      "epoch": 1.81,
-      "learning_rate": 3.6863370082428674e-05,
-      "loss": 0.2863,
+      "epoch": 1.8564169099362093,
+      "grad_norm": 0.18021084368228912,
+      "learning_rate": 3.6230100245601056e-05,
+      "loss": 0.4206,
       "step": 51510
     },
     {
-      "epoch": 1.81,
-      "learning_rate": 3.686086245625022e-05,
-      "loss": 0.2824,
+      "epoch": 1.856597109597434,
+      "grad_norm": 0.16509023308753967,
+      "learning_rate": 3.6227493023253425e-05,
+      "loss": 0.4185,
       "step": 51515
     },
     {
-      "epoch": 1.81,
-      "learning_rate": 3.685835467606628e-05,
-      "loss": 0.3065,
+      "epoch": 1.8567773092586586,
+      "grad_norm": 0.2372467815876007,
+      "learning_rate": 3.622488564793337e-05,
+      "loss": 0.4057,
       "step": 51520
     },
     {
-      "epoch": 1.81,
-      "learning_rate": 3.685584674190942e-05,
-      "loss": 0.287,
+      "epoch": 1.856957508919883,
+      "grad_norm": 0.2038879245519638,
+      "learning_rate": 3.622227811967643e-05,
+      "loss": 0.3873,
       "step": 51525
     },
     {
-      "epoch": 1.81,
-      "learning_rate": 3.685333865381222e-05,
-      "loss": 0.2882,
+      "epoch": 1.8571377085811078,
+      "grad_norm": 0.1562618464231491,
+      "learning_rate": 3.6219670438518125e-05,
+      "loss": 0.3674,
       "step": 51530
     },
     {
-      "epoch": 1.81,
-      "learning_rate": 3.685133207251927e-05,
-      "loss": 0.2919,
+      "epoch": 1.8573179082423326,
+      "grad_norm": 0.20719584822654724,
+      "learning_rate": 3.621706260449397e-05,
+      "loss": 0.4226,
       "step": 51535
     },
     {
-      "epoch": 1.81,
-      "learning_rate": 3.684882370741149e-05,
-      "loss": 0.292,
+      "epoch": 1.8574981079035573,
+      "grad_norm": 0.15713489055633545,
+      "learning_rate": 3.621445461763952e-05,
+      "loss": 0.4243,
       "step": 51540
     },
     {
-      "epoch": 1.81,
-      "learning_rate": 3.684631518845454e-05,
-      "loss": 0.294,
+      "epoch": 1.8576783075647818,
+      "grad_norm": 0.19836826622486115,
+      "learning_rate": 3.62118464779903e-05,
+      "loss": 0.4348,
       "step": 51545
     },
     {
-      "epoch": 1.81,
-      "learning_rate": 3.684380651568099e-05,
-      "loss": 0.2997,
+      "epoch": 1.8578585072260063,
+      "grad_norm": 0.17970958352088928,
+      "learning_rate": 3.620923818558183e-05,
+      "loss": 0.4002,
       "step": 51550
     },
     {
-      "epoch": 1.81,
-      "learning_rate": 3.6841297689123415e-05,
-      "loss": 0.2699,
+      "epoch": 1.858038706887231,
+      "grad_norm": 0.2058597356081009,
+      "learning_rate": 3.6206629740449666e-05,
+      "loss": 0.3783,
       "step": 51555
     },
     {
-      "epoch": 1.81,
-      "learning_rate": 3.68387887088144e-05,
-      "loss": 0.3079,
+      "epoch": 1.8582189065484558,
+      "grad_norm": 0.17197424173355103,
+      "learning_rate": 3.620402114262934e-05,
+      "loss": 0.423,
       "step": 51560
     },
     {
-      "epoch": 1.81,
-      "learning_rate": 3.6836279574786516e-05,
-      "loss": 0.2889,
+      "epoch": 1.8583991062096803,
+      "grad_norm": 0.2183474451303482,
+      "learning_rate": 3.6201412392156395e-05,
+      "loss": 0.4269,
       "step": 51565
     },
     {
-      "epoch": 1.81,
-      "learning_rate": 3.683377028707233e-05,
-      "loss": 0.289,
+      "epoch": 1.8585793058709048,
+      "grad_norm": 0.1597500592470169,
+      "learning_rate": 3.619880348906638e-05,
+      "loss": 0.3744,
       "step": 51570
     },
     {
-      "epoch": 1.81,
-      "learning_rate": 3.6831260845704445e-05,
-      "loss": 0.3018,
+      "epoch": 1.8587595055321295,
+      "grad_norm": 0.14877550303936005,
+      "learning_rate": 3.619619443339483e-05,
+      "loss": 0.3873,
       "step": 51575
     },
     {
-      "epoch": 1.81,
-      "learning_rate": 3.682875125071544e-05,
-      "loss": 0.2832,
+      "epoch": 1.8589397051933543,
+      "grad_norm": 0.21616920828819275,
+      "learning_rate": 3.6193585225177296e-05,
+      "loss": 0.4411,
       "step": 51580
     },
     {
-      "epoch": 1.81,
-      "learning_rate": 3.6826241502137884e-05,
-      "loss": 0.2844,
+      "epoch": 1.859119904854579,
+      "grad_norm": 0.20081162452697754,
+      "learning_rate": 3.619097586444934e-05,
+      "loss": 0.4237,
       "step": 51585
     },
     {
-      "epoch": 1.82,
-      "learning_rate": 3.682373160000438e-05,
-      "loss": 0.2827,
+      "epoch": 1.8593001045158035,
+      "grad_norm": 0.1473800092935562,
+      "learning_rate": 3.618836635124649e-05,
+      "loss": 0.3905,
       "step": 51590
     },
     {
-      "epoch": 1.82,
-      "learning_rate": 3.6821221544347516e-05,
-      "loss": 0.291,
+      "epoch": 1.859480304177028,
+      "grad_norm": 0.18639175593852997,
+      "learning_rate": 3.618575668560433e-05,
+      "loss": 0.4168,
       "step": 51595
     },
     {
-      "epoch": 1.82,
-      "learning_rate": 3.681871133519988e-05,
-      "loss": 0.2522,
+      "epoch": 1.8596605038382528,
+      "grad_norm": 0.17644469439983368,
+      "learning_rate": 3.6183146867558394e-05,
+      "loss": 0.4074,
       "step": 51600
     },
     {
-      "epoch": 1.82,
-      "learning_rate": 3.681620097259407e-05,
-      "loss": 0.2753,
+      "epoch": 1.8598407034994775,
+      "grad_norm": 0.15397772192955017,
+      "learning_rate": 3.6180536897144245e-05,
+      "loss": 0.4132,
       "step": 51605
     },
     {
-      "epoch": 1.82,
-      "learning_rate": 3.6813690456562674e-05,
-      "loss": 0.2762,
+      "epoch": 1.860020903160702,
+      "grad_norm": 0.17601633071899414,
+      "learning_rate": 3.6177926774397455e-05,
+      "loss": 0.4003,
       "step": 51610
     },
     {
-      "epoch": 1.82,
-      "learning_rate": 3.681117978713829e-05,
-      "loss": 0.2705,
+      "epoch": 1.8602011028219267,
+      "grad_norm": 0.20962753891944885,
+      "learning_rate": 3.617531649935356e-05,
+      "loss": 0.431,
       "step": 51615
     },
     {
-      "epoch": 1.82,
-      "learning_rate": 3.680866896435353e-05,
-      "loss": 0.2775,
+      "epoch": 1.8603813024831513,
+      "grad_norm": 0.18711386620998383,
+      "learning_rate": 3.617270607204816e-05,
+      "loss": 0.4172,
       "step": 51620
     },
     {
-      "epoch": 1.82,
-      "learning_rate": 3.680615798824097e-05,
-      "loss": 0.3353,
+      "epoch": 1.860561502144376,
+      "grad_norm": 0.19060426950454712,
+      "learning_rate": 3.6170095492516795e-05,
+      "loss": 0.4218,
       "step": 51625
     },
     {
-      "epoch": 1.82,
-      "learning_rate": 3.680364685883325e-05,
-      "loss": 0.2734,
+      "epoch": 1.8607417018056007,
+      "grad_norm": 0.24624811112880707,
+      "learning_rate": 3.616748476079504e-05,
+      "loss": 0.4342,
       "step": 51630
     },
     {
-      "epoch": 1.82,
-      "learning_rate": 3.6801135576162935e-05,
-      "loss": 0.277,
+      "epoch": 1.8609219014668252,
+      "grad_norm": 0.20159181952476501,
+      "learning_rate": 3.616487387691847e-05,
+      "loss": 0.3987,
       "step": 51635
     },
     {
-      "epoch": 1.82,
-      "learning_rate": 3.679862414026266e-05,
-      "loss": 0.2991,
+      "epoch": 1.8611021011280497,
+      "grad_norm": 0.18336619436740875,
+      "learning_rate": 3.616226284092264e-05,
+      "loss": 0.4065,
       "step": 51640
     },
     {
-      "epoch": 1.82,
-      "learning_rate": 3.679611255116502e-05,
-      "loss": 0.3159,
+      "epoch": 1.8612823007892745,
+      "grad_norm": 0.21609511971473694,
+      "learning_rate": 3.615965165284316e-05,
+      "loss": 0.4113,
       "step": 51645
     },
     {
-      "epoch": 1.82,
-      "learning_rate": 3.679360080890264e-05,
-      "loss": 0.2819,
+      "epoch": 1.8614625004504992,
+      "grad_norm": 0.22596147656440735,
+      "learning_rate": 3.615704031271558e-05,
+      "loss": 0.4017,
       "step": 51650
     },
     {
-      "epoch": 1.82,
-      "learning_rate": 3.679108891350813e-05,
-      "loss": 0.2818,
+      "epoch": 1.861642700111724,
+      "grad_norm": 0.15656688809394836,
+      "learning_rate": 3.6154428820575484e-05,
+      "loss": 0.4422,
       "step": 51655
     },
     {
-      "epoch": 1.82,
-      "learning_rate": 3.6788576865014084e-05,
-      "loss": 0.264,
+      "epoch": 1.8618228997729485,
+      "grad_norm": 0.19237256050109863,
+      "learning_rate": 3.615181717645845e-05,
+      "loss": 0.4242,
       "step": 51660
     },
     {
-      "epoch": 1.82,
-      "learning_rate": 3.6786064663453143e-05,
-      "loss": 0.2892,
+      "epoch": 1.862003099434173,
+      "grad_norm": 0.18129973113536835,
+      "learning_rate": 3.6149205380400074e-05,
+      "loss": 0.4334,
       "step": 51665
     },
     {
-      "epoch": 1.82,
-      "learning_rate": 3.678355230885793e-05,
-      "loss": 0.2763,
+      "epoch": 1.8621832990953977,
+      "grad_norm": 0.19113698601722717,
+      "learning_rate": 3.614659343243594e-05,
+      "loss": 0.4155,
       "step": 51670
     },
     {
-      "epoch": 1.82,
-      "learning_rate": 3.6781039801261054e-05,
-      "loss": 0.2865,
+      "epoch": 1.8623634987566224,
+      "grad_norm": 0.20505818724632263,
+      "learning_rate": 3.6143981332601615e-05,
+      "loss": 0.4176,
       "step": 51675
     },
     {
-      "epoch": 1.82,
-      "learning_rate": 3.6778527140695135e-05,
-      "loss": 0.2993,
+      "epoch": 1.862543698417847,
+      "grad_norm": 0.16761788725852966,
+      "learning_rate": 3.6141369080932705e-05,
+      "loss": 0.422,
       "step": 51680
     },
     {
-      "epoch": 1.82,
-      "learning_rate": 3.67760143271928e-05,
-      "loss": 0.266,
+      "epoch": 1.8627238980790715,
+      "grad_norm": 0.18111896514892578,
+      "learning_rate": 3.61387566774648e-05,
+      "loss": 0.4167,
       "step": 51685
     },
     {
-      "epoch": 1.82,
-      "learning_rate": 3.677350136078668e-05,
-      "loss": 0.2778,
+      "epoch": 1.8629040977402962,
+      "grad_norm": 0.15977542102336884,
+      "learning_rate": 3.6136144122233497e-05,
+      "loss": 0.3934,
       "step": 51690
     },
     {
-      "epoch": 1.82,
-      "learning_rate": 3.6770988241509406e-05,
-      "loss": 0.2742,
+      "epoch": 1.863084297401521,
+      "grad_norm": 0.17929469048976898,
+      "learning_rate": 3.6133531415274376e-05,
+      "loss": 0.4395,
       "step": 51695
     },
     {
-      "epoch": 1.82,
-      "learning_rate": 3.6768474969393607e-05,
-      "loss": 0.2973,
+      "epoch": 1.8632644970627457,
+      "grad_norm": 0.19982290267944336,
+      "learning_rate": 3.613091855662305e-05,
+      "loss": 0.3916,
       "step": 51700
     },
     {
-      "epoch": 1.82,
-      "learning_rate": 3.6765961544471925e-05,
-      "loss": 0.3009,
+      "epoch": 1.8634446967239702,
+      "grad_norm": 0.16737203299999237,
+      "learning_rate": 3.6128305546315114e-05,
+      "loss": 0.4053,
       "step": 51705
     },
     {
-      "epoch": 1.82,
-      "learning_rate": 3.6763447966776975e-05,
-      "loss": 0.2739,
+      "epoch": 1.8636248963851947,
+      "grad_norm": 0.1759185642004013,
+      "learning_rate": 3.6125692384386164e-05,
+      "loss": 0.3961,
       "step": 51710
     },
     {
-      "epoch": 1.82,
-      "learning_rate": 3.676093423634142e-05,
-      "loss": 0.2892,
+      "epoch": 1.8638050960464194,
+      "grad_norm": 0.1947273313999176,
+      "learning_rate": 3.612307907087182e-05,
+      "loss": 0.4288,
       "step": 51715
     },
     {
-      "epoch": 1.82,
-      "learning_rate": 3.6758420353197863e-05,
-      "loss": 0.3095,
+      "epoch": 1.8639852957076442,
+      "grad_norm": 0.14398063719272614,
+      "learning_rate": 3.6120465605807666e-05,
+      "loss": 0.3918,
       "step": 51720
     },
     {
-      "epoch": 1.82,
-      "learning_rate": 3.675590631737898e-05,
-      "loss": 0.2716,
+      "epoch": 1.8641654953688687,
+      "grad_norm": 0.18036890029907227,
+      "learning_rate": 3.611785198922933e-05,
+      "loss": 0.422,
       "step": 51725
     },
     {
-      "epoch": 1.82,
-      "learning_rate": 3.6753392128917396e-05,
-      "loss": 0.2891,
+      "epoch": 1.8643456950300934,
+      "grad_norm": 0.18519078195095062,
+      "learning_rate": 3.611523822117241e-05,
+      "loss": 0.4418,
       "step": 51730
     },
     {
-      "epoch": 1.82,
-      "learning_rate": 3.675087778784577e-05,
-      "loss": 0.2994,
+      "epoch": 1.864525894691318,
+      "grad_norm": 0.20603454113006592,
+      "learning_rate": 3.611262430167253e-05,
+      "loss": 0.4232,
       "step": 51735
     },
     {
-      "epoch": 1.82,
-      "learning_rate": 3.674836329419673e-05,
-      "loss": 0.2867,
+      "epoch": 1.8647060943525426,
+      "grad_norm": 0.18835075199604034,
+      "learning_rate": 3.6110010230765276e-05,
+      "loss": 0.4057,
       "step": 51740
     },
     {
-      "epoch": 1.82,
-      "learning_rate": 3.674584864800294e-05,
-      "loss": 0.2896,
+      "epoch": 1.8648862940137674,
+      "grad_norm": 0.18818290531635284,
+      "learning_rate": 3.6107396008486296e-05,
+      "loss": 0.4408,
       "step": 51745
     },
     {
-      "epoch": 1.82,
-      "learning_rate": 3.674333384929705e-05,
-      "loss": 0.301,
+      "epoch": 1.8650664936749919,
+      "grad_norm": 0.16613999009132385,
+      "learning_rate": 3.61047816348712e-05,
+      "loss": 0.3983,
       "step": 51750
     },
     {
-      "epoch": 1.82,
-      "learning_rate": 3.67408188981117e-05,
-      "loss": 0.2757,
+      "epoch": 1.8652466933362164,
+      "grad_norm": 0.19603601098060608,
+      "learning_rate": 3.6102167109955594e-05,
+      "loss": 0.433,
       "step": 51755
     },
     {
-      "epoch": 1.82,
-      "learning_rate": 3.6738303794479556e-05,
-      "loss": 0.2815,
+      "epoch": 1.8654268929974411,
+      "grad_norm": 0.1884978711605072,
+      "learning_rate": 3.609955243377511e-05,
+      "loss": 0.4285,
       "step": 51760
     },
     {
-      "epoch": 1.82,
-      "learning_rate": 3.673578853843328e-05,
-      "loss": 0.3074,
+      "epoch": 1.8656070926586659,
+      "grad_norm": 0.19464312493801117,
+      "learning_rate": 3.609693760636538e-05,
+      "loss": 0.3915,
       "step": 51765
     },
     {
-      "epoch": 1.82,
-      "learning_rate": 3.673327313000552e-05,
-      "loss": 0.2867,
+      "epoch": 1.8657872923198906,
+      "grad_norm": 0.20593151450157166,
+      "learning_rate": 3.609432262776202e-05,
+      "loss": 0.4311,
       "step": 51770
     },
     {
-      "epoch": 1.82,
-      "learning_rate": 3.673075756922894e-05,
-      "loss": 0.2663,
+      "epoch": 1.8659674919811151,
+      "grad_norm": 0.1879161298274994,
+      "learning_rate": 3.6091707498000666e-05,
+      "loss": 0.4121,
       "step": 51775
     },
     {
-      "epoch": 1.82,
-      "learning_rate": 3.6728241856136204e-05,
-      "loss": 0.2919,
+      "epoch": 1.8661476916423396,
+      "grad_norm": 0.19726385176181793,
+      "learning_rate": 3.608909221711694e-05,
+      "loss": 0.3844,
       "step": 51780
     },
     {
-      "epoch": 1.82,
-      "learning_rate": 3.672572599075998e-05,
-      "loss": 0.3072,
+      "epoch": 1.8663278913035644,
+      "grad_norm": 0.1846427619457245,
+      "learning_rate": 3.6086476785146486e-05,
+      "loss": 0.4014,
       "step": 51785
     },
     {
-      "epoch": 1.82,
-      "learning_rate": 3.672320997313292e-05,
-      "loss": 0.2797,
+      "epoch": 1.866508090964789,
+      "grad_norm": 0.1984597146511078,
+      "learning_rate": 3.6083861202124926e-05,
+      "loss": 0.4117,
       "step": 51790
     },
     {
-      "epoch": 1.82,
-      "learning_rate": 3.672069380328772e-05,
-      "loss": 0.2797,
+      "epoch": 1.8666882906260136,
+      "grad_norm": 0.14796842634677887,
+      "learning_rate": 3.60812454680879e-05,
+      "loss": 0.3917,
       "step": 51795
     },
     {
-      "epoch": 1.82,
-      "learning_rate": 3.671817748125702e-05,
-      "loss": 0.3024,
+      "epoch": 1.8668684902872381,
+      "grad_norm": 0.21340517699718475,
+      "learning_rate": 3.607862958307106e-05,
+      "loss": 0.3993,
       "step": 51800
     },
     {
-      "epoch": 1.82,
-      "learning_rate": 3.671566100707352e-05,
-      "loss": 0.2841,
+      "epoch": 1.8670486899484628,
+      "grad_norm": 0.17566043138504028,
+      "learning_rate": 3.607601354711003e-05,
+      "loss": 0.4282,
       "step": 51805
     },
     {
-      "epoch": 1.82,
-      "learning_rate": 3.6713144380769885e-05,
-      "loss": 0.3027,
+      "epoch": 1.8672288896096876,
+      "grad_norm": 0.19569726288318634,
+      "learning_rate": 3.607339736024046e-05,
+      "loss": 0.3953,
       "step": 51810
     },
     {
-      "epoch": 1.82,
-      "learning_rate": 3.671062760237878e-05,
-      "loss": 0.2932,
+      "epoch": 1.8674090892709123,
+      "grad_norm": 0.18134260177612305,
+      "learning_rate": 3.6070781022497996e-05,
+      "loss": 0.4215,
       "step": 51815
     },
     {
-      "epoch": 1.82,
-      "learning_rate": 3.6708110671932896e-05,
-      "loss": 0.2705,
+      "epoch": 1.8675892889321368,
+      "grad_norm": 0.18229928612709045,
+      "learning_rate": 3.606816453391828e-05,
+      "loss": 0.3922,
       "step": 51820
     },
     {
-      "epoch": 1.82,
-      "learning_rate": 3.670559358946491e-05,
-      "loss": 0.2865,
+      "epoch": 1.8677694885933613,
+      "grad_norm": 0.16680172085762024,
+      "learning_rate": 3.606554789453697e-05,
+      "loss": 0.3905,
       "step": 51825
     },
     {
-      "epoch": 1.82,
-      "learning_rate": 3.6703076355007506e-05,
-      "loss": 0.2862,
+      "epoch": 1.867949688254586,
+      "grad_norm": 0.20481085777282715,
+      "learning_rate": 3.606293110438972e-05,
+      "loss": 0.4282,
       "step": 51830
     },
     {
-      "epoch": 1.82,
-      "learning_rate": 3.670055896859337e-05,
-      "loss": 0.3049,
+      "epoch": 1.8681298879158108,
+      "grad_norm": 0.17302840948104858,
+      "learning_rate": 3.6060314163512164e-05,
+      "loss": 0.4102,
       "step": 51835
     },
     {
-      "epoch": 1.82,
-      "learning_rate": 3.669804143025519e-05,
-      "loss": 0.2863,
+      "epoch": 1.8683100875770353,
+      "grad_norm": 0.1852729618549347,
+      "learning_rate": 3.605769707193997e-05,
+      "loss": 0.4104,
       "step": 51840
     },
     {
-      "epoch": 1.82,
-      "learning_rate": 3.669552374002564e-05,
-      "loss": 0.2776,
+      "epoch": 1.8684902872382598,
+      "grad_norm": 0.1932128518819809,
+      "learning_rate": 3.6055079829708795e-05,
+      "loss": 0.4389,
       "step": 51845
     },
     {
-      "epoch": 1.82,
-      "learning_rate": 3.6693005897937435e-05,
-      "loss": 0.2851,
+      "epoch": 1.8686704868994846,
+      "grad_norm": 0.20994549989700317,
+      "learning_rate": 3.60524624368543e-05,
+      "loss": 0.4512,
       "step": 51850
     },
     {
-      "epoch": 1.82,
-      "learning_rate": 3.669048790402324e-05,
-      "loss": 0.2902,
+      "epoch": 1.8688506865607093,
+      "grad_norm": 0.2091553956270218,
+      "learning_rate": 3.6049844893412143e-05,
+      "loss": 0.4099,
       "step": 51855
     },
     {
-      "epoch": 1.82,
-      "learning_rate": 3.6687969758315774e-05,
-      "loss": 0.2835,
+      "epoch": 1.869030886221934,
+      "grad_norm": 0.16755080223083496,
+      "learning_rate": 3.6047227199417987e-05,
+      "loss": 0.4143,
       "step": 51860
     },
     {
-      "epoch": 1.82,
-      "learning_rate": 3.668545146084772e-05,
-      "loss": 0.2973,
+      "epoch": 1.8692110858831585,
+      "grad_norm": 0.23165898025035858,
+      "learning_rate": 3.6044609354907505e-05,
+      "loss": 0.4135,
       "step": 51865
     },
     {
-      "epoch": 1.82,
-      "learning_rate": 3.668293301165179e-05,
-      "loss": 0.2554,
+      "epoch": 1.869391285544383,
+      "grad_norm": 0.20910170674324036,
+      "learning_rate": 3.604199135991635e-05,
+      "loss": 0.4071,
       "step": 51870
     },
     {
-      "epoch": 1.83,
-      "learning_rate": 3.668041441076066e-05,
-      "loss": 0.2768,
+      "epoch": 1.8695714852056078,
+      "grad_norm": 0.17225544154644012,
+      "learning_rate": 3.603937321448021e-05,
+      "loss": 0.4114,
       "step": 51875
     },
     {
-      "epoch": 1.83,
-      "learning_rate": 3.667789565820705e-05,
-      "loss": 0.2787,
+      "epoch": 1.8697516848668325,
+      "grad_norm": 0.17193418741226196,
+      "learning_rate": 3.6036754918634744e-05,
+      "loss": 0.4112,
       "step": 51880
     },
     {
-      "epoch": 1.83,
-      "learning_rate": 3.667537675402366e-05,
-      "loss": 0.2967,
+      "epoch": 1.8699318845280573,
+      "grad_norm": 0.17545709013938904,
+      "learning_rate": 3.6034136472415624e-05,
+      "loss": 0.4146,
       "step": 51885
     },
     {
-      "epoch": 1.83,
-      "learning_rate": 3.667285769824321e-05,
-      "loss": 0.2772,
+      "epoch": 1.8701120841892818,
+      "grad_norm": 0.17598995566368103,
+      "learning_rate": 3.603151787585853e-05,
+      "loss": 0.3975,
       "step": 51890
     },
     {
-      "epoch": 1.83,
-      "learning_rate": 3.667033849089838e-05,
-      "loss": 0.3012,
+      "epoch": 1.8702922838505063,
+      "grad_norm": 0.19214802980422974,
+      "learning_rate": 3.602889912899915e-05,
+      "loss": 0.4483,
       "step": 51895
     },
     {
-      "epoch": 1.83,
-      "learning_rate": 3.666781913202191e-05,
-      "loss": 0.2883,
+      "epoch": 1.870472483511731,
+      "grad_norm": 0.19266071915626526,
+      "learning_rate": 3.602628023187315e-05,
+      "loss": 0.4325,
       "step": 51900
     },
     {
-      "epoch": 1.83,
-      "learning_rate": 3.666529962164648e-05,
-      "loss": 0.2899,
+      "epoch": 1.8706526831729557,
+      "grad_norm": 0.2069164216518402,
+      "learning_rate": 3.602366118451621e-05,
+      "loss": 0.4078,
       "step": 51905
     },
     {
-      "epoch": 1.83,
-      "learning_rate": 3.6662779959804833e-05,
-      "loss": 0.2894,
+      "epoch": 1.8708328828341803,
+      "grad_norm": 0.18193595111370087,
+      "learning_rate": 3.6021041986964035e-05,
+      "loss": 0.4199,
       "step": 51910
     },
     {
-      "epoch": 1.83,
-      "learning_rate": 3.6660260146529676e-05,
-      "loss": 0.2798,
+      "epoch": 1.8710130824954048,
+      "grad_norm": 0.15143196284770966,
+      "learning_rate": 3.601842263925228e-05,
+      "loss": 0.3955,
       "step": 51915
     },
     {
-      "epoch": 1.83,
-      "learning_rate": 3.665774018185372e-05,
-      "loss": 0.2764,
+      "epoch": 1.8711932821566295,
+      "grad_norm": 0.1509236842393875,
+      "learning_rate": 3.601580314141666e-05,
+      "loss": 0.3925,
       "step": 51920
     },
     {
-      "epoch": 1.83,
-      "learning_rate": 3.665522006580969e-05,
-      "loss": 0.2734,
+      "epoch": 1.8713734818178542,
+      "grad_norm": 0.20824167132377625,
+      "learning_rate": 3.601318349349285e-05,
+      "loss": 0.4159,
       "step": 51925
     },
     {
-      "epoch": 1.83,
-      "learning_rate": 3.665269979843031e-05,
-      "loss": 0.3093,
+      "epoch": 1.871553681479079,
+      "grad_norm": 0.1683613657951355,
+      "learning_rate": 3.601056369551655e-05,
+      "loss": 0.3946,
       "step": 51930
     },
     {
-      "epoch": 1.83,
-      "learning_rate": 3.665017937974831e-05,
-      "loss": 0.2899,
+      "epoch": 1.8717338811403035,
+      "grad_norm": 0.20717251300811768,
+      "learning_rate": 3.600794374752346e-05,
+      "loss": 0.4327,
       "step": 51935
     },
     {
-      "epoch": 1.83,
-      "learning_rate": 3.664765880979639e-05,
-      "loss": 0.2641,
+      "epoch": 1.871914080801528,
+      "grad_norm": 0.1377881020307541,
+      "learning_rate": 3.600532364954926e-05,
+      "loss": 0.4057,
       "step": 51940
     },
     {
-      "epoch": 1.83,
-      "learning_rate": 3.664513808860732e-05,
-      "loss": 0.2931,
+      "epoch": 1.8720942804627527,
+      "grad_norm": 0.20057149231433868,
+      "learning_rate": 3.600270340162966e-05,
+      "loss": 0.4513,
       "step": 51945
     },
     {
-      "epoch": 1.83,
-      "learning_rate": 3.664261721621379e-05,
-      "loss": 0.2899,
+      "epoch": 1.8722744801239775,
+      "grad_norm": 0.1850452721118927,
+      "learning_rate": 3.600008300380035e-05,
+      "loss": 0.3977,
       "step": 51950
     },
     {
-      "epoch": 1.83,
-      "learning_rate": 3.6640096192648554e-05,
-      "loss": 0.3063,
+      "epoch": 1.872454679785202,
+      "grad_norm": 0.21717745065689087,
+      "learning_rate": 3.599746245609703e-05,
+      "loss": 0.3927,
       "step": 51955
     },
     {
-      "epoch": 1.83,
-      "learning_rate": 3.663757501794434e-05,
-      "loss": 0.2803,
+      "epoch": 1.8726348794464265,
+      "grad_norm": 0.17817428708076477,
+      "learning_rate": 3.599484175855543e-05,
+      "loss": 0.4003,
       "step": 51960
     },
     {
-      "epoch": 1.83,
-      "learning_rate": 3.663505369213388e-05,
-      "loss": 0.2783,
+      "epoch": 1.8728150791076512,
+      "grad_norm": 0.1645549237728119,
+      "learning_rate": 3.599222091121123e-05,
+      "loss": 0.4091,
       "step": 51965
     },
     {
-      "epoch": 1.83,
-      "learning_rate": 3.663253221524992e-05,
-      "loss": 0.2873,
+      "epoch": 1.872995278768876,
+      "grad_norm": 0.26613032817840576,
+      "learning_rate": 3.598959991410015e-05,
+      "loss": 0.4515,
       "step": 51970
     },
     {
-      "epoch": 1.83,
-      "learning_rate": 3.66300105873252e-05,
-      "loss": 0.306,
+      "epoch": 1.8731754784301007,
+      "grad_norm": 0.15055564045906067,
+      "learning_rate": 3.5986978767257906e-05,
+      "loss": 0.4026,
       "step": 51975
     },
     {
-      "epoch": 1.83,
-      "learning_rate": 3.6627488808392444e-05,
-      "loss": 0.2775,
+      "epoch": 1.8733556780913252,
+      "grad_norm": 0.19910356402397156,
+      "learning_rate": 3.59843574707202e-05,
+      "loss": 0.4153,
       "step": 51980
     },
     {
-      "epoch": 1.83,
-      "learning_rate": 3.662496687848442e-05,
-      "loss": 0.2764,
+      "epoch": 1.8735358777525497,
+      "grad_norm": 0.17660050094127655,
+      "learning_rate": 3.598173602452274e-05,
+      "loss": 0.4038,
       "step": 51985
     },
     {
-      "epoch": 1.83,
-      "learning_rate": 3.662244479763386e-05,
-      "loss": 0.3088,
+      "epoch": 1.8737160774137744,
+      "grad_norm": 0.19887585937976837,
+      "learning_rate": 3.5979114428701265e-05,
+      "loss": 0.4239,
       "step": 51990
     },
     {
-      "epoch": 1.83,
-      "learning_rate": 3.661992256587351e-05,
-      "loss": 0.2778,
+      "epoch": 1.8738962770749992,
+      "grad_norm": 0.1948709338903427,
+      "learning_rate": 3.597649268329148e-05,
+      "loss": 0.4,
       "step": 51995
     },
     {
-      "epoch": 1.83,
-      "learning_rate": 3.661740018323612e-05,
-      "loss": 0.3253,
+      "epoch": 1.8740764767362237,
+      "grad_norm": 0.2058577537536621,
+      "learning_rate": 3.59738707883291e-05,
+      "loss": 0.4223,
       "step": 52000
     },
     {
-      "epoch": 1.83,
-      "eval_loss": 0.2752087414264679,
-      "eval_runtime": 10.5359,
-      "eval_samples_per_second": 9.491,
-      "eval_steps_per_second": 9.491,
+      "epoch": 1.8740764767362237,
+      "eval_loss": 0.4389624297618866,
+      "eval_runtime": 3.5417,
+      "eval_samples_per_second": 28.235,
+      "eval_steps_per_second": 7.059,
       "step": 52000
     },
     {
-      "epoch": 1.83,
-      "learning_rate": 3.661487764975445e-05,
-      "loss": 0.2783,
+      "epoch": 1.8742566763974484,
+      "grad_norm": 0.19060571491718292,
+      "learning_rate": 3.5971248743849864e-05,
+      "loss": 0.4135,
       "step": 52005
     },
     {
-      "epoch": 1.83,
-      "learning_rate": 3.661235496546125e-05,
-      "loss": 0.2796,
+      "epoch": 1.874436876058673,
+      "grad_norm": 0.14245115220546722,
+      "learning_rate": 3.596862654988948e-05,
+      "loss": 0.3737,
       "step": 52010
     },
     {
-      "epoch": 1.83,
-      "learning_rate": 3.6609832130389274e-05,
-      "loss": 0.2624,
+      "epoch": 1.8746170757198977,
+      "grad_norm": 0.1959165632724762,
+      "learning_rate": 3.5966004206483687e-05,
+      "loss": 0.4575,
       "step": 52015
     },
     {
-      "epoch": 1.83,
-      "learning_rate": 3.6607309144571274e-05,
-      "loss": 0.307,
+      "epoch": 1.8747972753811224,
+      "grad_norm": 0.17014524340629578,
+      "learning_rate": 3.596338171366821e-05,
+      "loss": 0.3932,
       "step": 52020
     },
     {
-      "epoch": 1.83,
-      "learning_rate": 3.6604786008040024e-05,
-      "loss": 0.2823,
+      "epoch": 1.874977475042347,
+      "grad_norm": 0.18894478678703308,
+      "learning_rate": 3.596075907147878e-05,
+      "loss": 0.387,
       "step": 52025
     },
     {
-      "epoch": 1.83,
-      "learning_rate": 3.660226272082826e-05,
-      "loss": 0.2917,
+      "epoch": 1.8751576747035714,
+      "grad_norm": 0.1582806408405304,
+      "learning_rate": 3.595813627995113e-05,
+      "loss": 0.39,
       "step": 52030
     },
     {
-      "epoch": 1.83,
-      "learning_rate": 3.659973928296878e-05,
-      "loss": 0.2749,
+      "epoch": 1.8753378743647962,
+      "grad_norm": 0.20237891376018524,
+      "learning_rate": 3.595551333912099e-05,
+      "loss": 0.4171,
       "step": 52035
     },
     {
-      "epoch": 1.83,
-      "learning_rate": 3.659721569449431e-05,
-      "loss": 0.2948,
+      "epoch": 1.8755180740260209,
+      "grad_norm": 0.1860736459493637,
+      "learning_rate": 3.595289024902411e-05,
+      "loss": 0.406,
       "step": 52040
     },
     {
-      "epoch": 1.83,
-      "learning_rate": 3.6594691955437646e-05,
-      "loss": 0.2705,
+      "epoch": 1.8756982736872456,
+      "grad_norm": 0.191467747092247,
+      "learning_rate": 3.5950267009696206e-05,
+      "loss": 0.4538,
       "step": 52045
     },
     {
-      "epoch": 1.83,
-      "learning_rate": 3.659216806583155e-05,
-      "loss": 0.2612,
+      "epoch": 1.8758784733484701,
+      "grad_norm": 0.1970950961112976,
+      "learning_rate": 3.594764362117305e-05,
+      "loss": 0.3997,
       "step": 52050
     },
     {
-      "epoch": 1.83,
-      "learning_rate": 3.6589644025708785e-05,
-      "loss": 0.2771,
+      "epoch": 1.8760586730096946,
+      "grad_norm": 0.2063102424144745,
+      "learning_rate": 3.594502008349036e-05,
+      "loss": 0.4225,
       "step": 52055
     },
     {
-      "epoch": 1.83,
-      "learning_rate": 3.6587119835102135e-05,
-      "loss": 0.2768,
+      "epoch": 1.8762388726709194,
+      "grad_norm": 0.18709443509578705,
+      "learning_rate": 3.59423963966839e-05,
+      "loss": 0.3833,
       "step": 52060
     },
     {
-      "epoch": 1.83,
-      "learning_rate": 3.658459549404437e-05,
-      "loss": 0.2863,
+      "epoch": 1.8764190723321441,
+      "grad_norm": 0.1764218658208847,
+      "learning_rate": 3.59397725607894e-05,
+      "loss": 0.4111,
       "step": 52065
     },
     {
-      "epoch": 1.83,
-      "learning_rate": 3.658207100256827e-05,
-      "loss": 0.2957,
+      "epoch": 1.8765992719933686,
+      "grad_norm": 0.1787770837545395,
+      "learning_rate": 3.593714857584261e-05,
+      "loss": 0.4018,
       "step": 52070
     },
     {
-      "epoch": 1.83,
-      "learning_rate": 3.6579546360706606e-05,
-      "loss": 0.2587,
+      "epoch": 1.8767794716545931,
+      "grad_norm": 0.1761235147714615,
+      "learning_rate": 3.59345244418793e-05,
+      "loss": 0.4015,
       "step": 52075
     },
     {
-      "epoch": 1.83,
-      "learning_rate": 3.657702156849216e-05,
-      "loss": 0.2844,
+      "epoch": 1.8769596713158179,
+      "grad_norm": 0.1807931810617447,
+      "learning_rate": 3.593190015893521e-05,
+      "loss": 0.4127,
       "step": 52080
     },
     {
-      "epoch": 1.83,
-      "learning_rate": 3.657449662595773e-05,
-      "loss": 0.2673,
+      "epoch": 1.8771398709770426,
+      "grad_norm": 0.16054491698741913,
+      "learning_rate": 3.5929275727046095e-05,
+      "loss": 0.4037,
       "step": 52085
     },
     {
-      "epoch": 1.83,
-      "learning_rate": 3.6571971533136085e-05,
-      "loss": 0.2779,
+      "epoch": 1.8773200706382673,
+      "grad_norm": 0.23387478291988373,
+      "learning_rate": 3.5926651146247715e-05,
+      "loss": 0.429,
       "step": 52090
     },
     {
-      "epoch": 1.83,
-      "learning_rate": 3.656944629006001e-05,
-      "loss": 0.2795,
+      "epoch": 1.8775002702994918,
+      "grad_norm": 0.17151391506195068,
+      "learning_rate": 3.5924026416575826e-05,
+      "loss": 0.432,
       "step": 52095
     },
     {
-      "epoch": 1.83,
-      "learning_rate": 3.6566920896762304e-05,
-      "loss": 0.2798,
+      "epoch": 1.8776804699607164,
+      "grad_norm": 0.20386628806591034,
+      "learning_rate": 3.5921401538066195e-05,
+      "loss": 0.419,
       "step": 52100
     },
     {
-      "epoch": 1.83,
-      "learning_rate": 3.656439535327575e-05,
-      "loss": 0.281,
+      "epoch": 1.877860669621941,
+      "grad_norm": 0.16705821454524994,
+      "learning_rate": 3.591877651075458e-05,
+      "loss": 0.4308,
       "step": 52105
     },
     {
-      "epoch": 1.83,
-      "learning_rate": 3.656186965963315e-05,
-      "loss": 0.2881,
+      "epoch": 1.8780408692831658,
+      "grad_norm": 0.16681896150112152,
+      "learning_rate": 3.591615133467675e-05,
+      "loss": 0.4394,
       "step": 52110
     },
     {
-      "epoch": 1.83,
-      "learning_rate": 3.655934381586728e-05,
-      "loss": 0.2843,
+      "epoch": 1.8782210689443903,
+      "grad_norm": 0.16043955087661743,
+      "learning_rate": 3.591352600986847e-05,
+      "loss": 0.3904,
       "step": 52115
     },
     {
-      "epoch": 1.83,
-      "learning_rate": 3.6556817822010966e-05,
-      "loss": 0.2842,
+      "epoch": 1.878401268605615,
+      "grad_norm": 0.24545589089393616,
+      "learning_rate": 3.5910900536365517e-05,
+      "loss": 0.4429,
       "step": 52120
     },
     {
-      "epoch": 1.83,
-      "learning_rate": 3.6554291678096975e-05,
-      "loss": 0.2903,
+      "epoch": 1.8785814682668396,
+      "grad_norm": 0.18531683087348938,
+      "learning_rate": 3.590827491420365e-05,
+      "loss": 0.3819,
       "step": 52125
     },
     {
-      "epoch": 1.83,
-      "learning_rate": 3.655176538415813e-05,
-      "loss": 0.2705,
+      "epoch": 1.8787616679280643,
+      "grad_norm": 0.21406356990337372,
+      "learning_rate": 3.5905649143418654e-05,
+      "loss": 0.4126,
       "step": 52130
     },
     {
-      "epoch": 1.83,
-      "learning_rate": 3.6549238940227215e-05,
-      "loss": 0.2743,
+      "epoch": 1.878941867589289,
+      "grad_norm": 0.2111523151397705,
+      "learning_rate": 3.590302322404629e-05,
+      "loss": 0.4129,
       "step": 52135
     },
     {
-      "epoch": 1.83,
-      "learning_rate": 3.6546712346337054e-05,
-      "loss": 0.283,
+      "epoch": 1.8791220672505136,
+      "grad_norm": 0.19074268639087677,
+      "learning_rate": 3.590039715612236e-05,
+      "loss": 0.3669,
       "step": 52140
     },
     {
-      "epoch": 1.83,
-      "learning_rate": 3.654418560252043e-05,
-      "loss": 0.3023,
+      "epoch": 1.879302266911738,
+      "grad_norm": 0.1679174304008484,
+      "learning_rate": 3.5897770939682616e-05,
+      "loss": 0.4018,
       "step": 52145
     },
     {
-      "epoch": 1.83,
-      "learning_rate": 3.6541658708810175e-05,
-      "loss": 0.2607,
+      "epoch": 1.8794824665729628,
+      "grad_norm": 0.28113430738449097,
+      "learning_rate": 3.5895144574762855e-05,
+      "loss": 0.4003,
       "step": 52150
     },
     {
-      "epoch": 1.83,
-      "learning_rate": 3.6539131665239086e-05,
-      "loss": 0.2968,
+      "epoch": 1.8796626662341875,
+      "grad_norm": 0.1835562139749527,
+      "learning_rate": 3.589251806139887e-05,
+      "loss": 0.3801,
       "step": 52155
     },
     {
-      "epoch": 1.84,
-      "learning_rate": 3.6536604471839964e-05,
-      "loss": 0.2697,
+      "epoch": 1.8798428658954123,
+      "grad_norm": 0.21328282356262207,
+      "learning_rate": 3.588989139962642e-05,
+      "loss": 0.4041,
       "step": 52160
     },
     {
-      "epoch": 1.84,
-      "learning_rate": 3.6534077128645646e-05,
-      "loss": 0.2761,
+      "epoch": 1.8800230655566368,
+      "grad_norm": 0.15331655740737915,
+      "learning_rate": 3.5887264589481324e-05,
+      "loss": 0.399,
       "step": 52165
     },
     {
-      "epoch": 1.84,
-      "learning_rate": 3.6531549635688945e-05,
-      "loss": 0.2768,
+      "epoch": 1.8802032652178613,
+      "grad_norm": 0.20487786829471588,
+      "learning_rate": 3.588463763099934e-05,
+      "loss": 0.4275,
       "step": 52170
     },
     {
-      "epoch": 1.84,
-      "learning_rate": 3.652902199300266e-05,
-      "loss": 0.3106,
+      "epoch": 1.880383464879086,
+      "grad_norm": 0.17890450358390808,
+      "learning_rate": 3.5882010524216284e-05,
+      "loss": 0.4195,
       "step": 52175
     },
     {
-      "epoch": 1.84,
-      "learning_rate": 3.652649420061963e-05,
-      "loss": 0.2749,
+      "epoch": 1.8805636645403108,
+      "grad_norm": 0.1512042135000229,
+      "learning_rate": 3.587938326916794e-05,
+      "loss": 0.398,
       "step": 52180
     },
     {
-      "epoch": 1.84,
-      "learning_rate": 3.652396625857267e-05,
-      "loss": 0.2913,
+      "epoch": 1.8807438642015353,
+      "grad_norm": 0.17925597727298737,
+      "learning_rate": 3.587675586589011e-05,
+      "loss": 0.4313,
       "step": 52185
     },
     {
-      "epoch": 1.84,
-      "learning_rate": 3.65214381668946e-05,
-      "loss": 0.2954,
+      "epoch": 1.8809240638627598,
+      "grad_norm": 0.20325055718421936,
+      "learning_rate": 3.587412831441858e-05,
+      "loss": 0.3756,
       "step": 52190
     },
     {
-      "epoch": 1.84,
-      "learning_rate": 3.651890992561824e-05,
-      "loss": 0.2702,
+      "epoch": 1.8811042635239845,
+      "grad_norm": 0.1322261095046997,
+      "learning_rate": 3.5871500614789155e-05,
+      "loss": 0.3894,
       "step": 52195
     },
     {
-      "epoch": 1.84,
-      "learning_rate": 3.6516381534776436e-05,
-      "loss": 0.2934,
+      "epoch": 1.8812844631852093,
+      "grad_norm": 0.1917523741722107,
+      "learning_rate": 3.5868872767037646e-05,
+      "loss": 0.4009,
       "step": 52200
     },
     {
-      "epoch": 1.84,
-      "learning_rate": 3.651385299440201e-05,
-      "loss": 0.2758,
+      "epoch": 1.881464662846434,
+      "grad_norm": 0.174628347158432,
+      "learning_rate": 3.5866244771199855e-05,
+      "loss": 0.395,
       "step": 52205
     },
     {
-      "epoch": 1.84,
-      "learning_rate": 3.651132430452778e-05,
-      "loss": 0.2816,
+      "epoch": 1.8816448625076585,
+      "grad_norm": 0.17742396891117096,
+      "learning_rate": 3.586361662731157e-05,
+      "loss": 0.4242,
       "step": 52210
     },
     {
-      "epoch": 1.84,
-      "learning_rate": 3.6508795465186596e-05,
-      "loss": 0.2823,
+      "epoch": 1.881825062168883,
+      "grad_norm": 0.1793578863143921,
+      "learning_rate": 3.5860988335408616e-05,
+      "loss": 0.4331,
       "step": 52215
     },
     {
-      "epoch": 1.84,
-      "learning_rate": 3.650626647641129e-05,
-      "loss": 0.2805,
+      "epoch": 1.8820052618301077,
+      "grad_norm": 0.18306861817836761,
+      "learning_rate": 3.5858359895526807e-05,
+      "loss": 0.4215,
       "step": 52220
     },
     {
-      "epoch": 1.84,
-      "learning_rate": 3.65037373382347e-05,
-      "loss": 0.3046,
+      "epoch": 1.8821854614913325,
+      "grad_norm": 0.16969892382621765,
+      "learning_rate": 3.585573130770193e-05,
+      "loss": 0.4109,
       "step": 52225
     },
     {
-      "epoch": 1.84,
-      "learning_rate": 3.650120805068965e-05,
-      "loss": 0.2818,
+      "epoch": 1.882365661152557,
+      "grad_norm": 0.1748346984386444,
+      "learning_rate": 3.585310257196983e-05,
+      "loss": 0.3951,
       "step": 52230
     },
     {
-      "epoch": 1.84,
-      "learning_rate": 3.6498678613809e-05,
-      "loss": 0.2896,
+      "epoch": 1.8825458608137817,
+      "grad_norm": 0.20251350104808807,
+      "learning_rate": 3.5850473688366306e-05,
+      "loss": 0.4113,
       "step": 52235
     },
     {
-      "epoch": 1.84,
-      "learning_rate": 3.649614902762558e-05,
-      "loss": 0.2778,
+      "epoch": 1.8827260604750062,
+      "grad_norm": 0.19220057129859924,
+      "learning_rate": 3.5847844656927176e-05,
+      "loss": 0.3578,
       "step": 52240
     },
     {
-      "epoch": 1.84,
-      "learning_rate": 3.649361929217224e-05,
-      "loss": 0.2767,
+      "epoch": 1.882906260136231,
+      "grad_norm": 0.17758119106292725,
+      "learning_rate": 3.584521547768826e-05,
+      "loss": 0.3993,
       "step": 52245
     },
     {
-      "epoch": 1.84,
-      "learning_rate": 3.649108940748184e-05,
-      "loss": 0.2824,
+      "epoch": 1.8830864597974557,
+      "grad_norm": 0.17641225457191467,
+      "learning_rate": 3.584258615068539e-05,
+      "loss": 0.4343,
       "step": 52250
     },
     {
-      "epoch": 1.84,
-      "learning_rate": 3.648855937358721e-05,
-      "loss": 0.2869,
+      "epoch": 1.8832666594586802,
+      "grad_norm": 0.1818428784608841,
+      "learning_rate": 3.583995667595437e-05,
+      "loss": 0.3988,
       "step": 52255
     },
     {
-      "epoch": 1.84,
-      "learning_rate": 3.648602919052121e-05,
-      "loss": 0.3008,
+      "epoch": 1.8834468591199047,
+      "grad_norm": 0.16791364550590515,
+      "learning_rate": 3.583732705353105e-05,
+      "loss": 0.4238,
       "step": 52260
     },
     {
-      "epoch": 1.84,
-      "learning_rate": 3.648349885831669e-05,
-      "loss": 0.2764,
+      "epoch": 1.8836270587811295,
+      "grad_norm": 0.17646898329257965,
+      "learning_rate": 3.5834697283451244e-05,
+      "loss": 0.4154,
       "step": 52265
     },
     {
-      "epoch": 1.84,
-      "learning_rate": 3.6480968377006495e-05,
-      "loss": 0.2816,
+      "epoch": 1.8838072584423542,
+      "grad_norm": 0.1645815521478653,
+      "learning_rate": 3.5832067365750794e-05,
+      "loss": 0.4345,
       "step": 52270
     },
     {
-      "epoch": 1.84,
-      "learning_rate": 3.647843774662351e-05,
-      "loss": 0.2847,
+      "epoch": 1.883987458103579,
+      "grad_norm": 0.1957109421491623,
+      "learning_rate": 3.5829437300465504e-05,
+      "loss": 0.3884,
       "step": 52275
     },
     {
-      "epoch": 1.84,
-      "learning_rate": 3.647590696720057e-05,
-      "loss": 0.2593,
+      "epoch": 1.8841676577648034,
+      "grad_norm": 0.20072638988494873,
+      "learning_rate": 3.5826807087631243e-05,
+      "loss": 0.3996,
       "step": 52280
     },
     {
-      "epoch": 1.84,
-      "learning_rate": 3.647337603877054e-05,
-      "loss": 0.2581,
+      "epoch": 1.884347857426028,
+      "grad_norm": 0.1894032508134842,
+      "learning_rate": 3.582417672728383e-05,
+      "loss": 0.3905,
       "step": 52285
     },
     {
-      "epoch": 1.84,
-      "learning_rate": 3.647084496136627e-05,
-      "loss": 0.2778,
+      "epoch": 1.8845280570872527,
+      "grad_norm": 0.20433971285820007,
+      "learning_rate": 3.58215462194591e-05,
+      "loss": 0.4112,
       "step": 52290
     },
     {
-      "epoch": 1.84,
-      "learning_rate": 3.646831373502065e-05,
-      "loss": 0.2942,
+      "epoch": 1.8847082567484774,
+      "grad_norm": 0.19983576238155365,
+      "learning_rate": 3.58189155641929e-05,
+      "loss": 0.4133,
       "step": 52295
     },
     {
-      "epoch": 1.84,
-      "learning_rate": 3.6465782359766534e-05,
-      "loss": 0.3032,
+      "epoch": 1.884888456409702,
+      "grad_norm": 0.1698193997144699,
+      "learning_rate": 3.581628476152107e-05,
+      "loss": 0.4222,
       "step": 52300
     },
     {
-      "epoch": 1.84,
-      "learning_rate": 3.646325083563679e-05,
-      "loss": 0.2676,
+      "epoch": 1.8850686560709264,
+      "grad_norm": 0.18257997930049896,
+      "learning_rate": 3.581365381147946e-05,
+      "loss": 0.4192,
       "step": 52305
     },
     {
-      "epoch": 1.84,
-      "learning_rate": 3.646071916266429e-05,
-      "loss": 0.2821,
+      "epoch": 1.8852488557321512,
+      "grad_norm": 0.22537992894649506,
+      "learning_rate": 3.5811022714103906e-05,
+      "loss": 0.4401,
       "step": 52310
     },
     {
-      "epoch": 1.84,
-      "learning_rate": 3.64581873408819e-05,
-      "loss": 0.3013,
+      "epoch": 1.885429055393376,
+      "grad_norm": 0.19952908158302307,
+      "learning_rate": 3.580839146943026e-05,
+      "loss": 0.3866,
       "step": 52315
     },
     {
-      "epoch": 1.84,
-      "learning_rate": 3.64556553703225e-05,
-      "loss": 0.2605,
+      "epoch": 1.8856092550546006,
+      "grad_norm": 0.25064241886138916,
+      "learning_rate": 3.5805760077494366e-05,
+      "loss": 0.4678,
       "step": 52320
     },
     {
-      "epoch": 1.84,
-      "learning_rate": 3.645312325101896e-05,
-      "loss": 0.2636,
+      "epoch": 1.8857894547158252,
+      "grad_norm": 0.19262924790382385,
+      "learning_rate": 3.580312853833209e-05,
+      "loss": 0.4435,
       "step": 52325
     },
     {
-      "epoch": 1.84,
-      "learning_rate": 3.645059098300417e-05,
-      "loss": 0.2734,
+      "epoch": 1.8859696543770497,
+      "grad_norm": 0.12994927167892456,
+      "learning_rate": 3.580049685197928e-05,
+      "loss": 0.3827,
       "step": 52330
     },
     {
-      "epoch": 1.84,
-      "learning_rate": 3.644805856631101e-05,
-      "loss": 0.3045,
+      "epoch": 1.8861498540382744,
+      "grad_norm": 0.12715496122837067,
+      "learning_rate": 3.5797865018471785e-05,
+      "loss": 0.4211,
       "step": 52335
     },
     {
-      "epoch": 1.84,
-      "learning_rate": 3.644552600097234e-05,
-      "loss": 0.2675,
+      "epoch": 1.8863300536994991,
+      "grad_norm": 0.18410317599773407,
+      "learning_rate": 3.5795233037845475e-05,
+      "loss": 0.4057,
       "step": 52340
     },
     {
-      "epoch": 1.84,
-      "learning_rate": 3.644299328702106e-05,
-      "loss": 0.2913,
+      "epoch": 1.8865102533607236,
+      "grad_norm": 0.2159508764743805,
+      "learning_rate": 3.579260091013621e-05,
+      "loss": 0.4253,
       "step": 52345
     },
     {
-      "epoch": 1.84,
-      "learning_rate": 3.6440460424490055e-05,
-      "loss": 0.2898,
+      "epoch": 1.8866904530219482,
+      "grad_norm": 0.23478345572948456,
+      "learning_rate": 3.578996863537983e-05,
+      "loss": 0.4077,
       "step": 52350
     },
     {
-      "epoch": 1.84,
-      "learning_rate": 3.643792741341221e-05,
-      "loss": 0.2949,
+      "epoch": 1.8868706526831729,
+      "grad_norm": 0.13263140618801117,
+      "learning_rate": 3.578733621361223e-05,
+      "loss": 0.381,
       "step": 52355
     },
     {
-      "epoch": 1.84,
-      "learning_rate": 3.643539425382042e-05,
-      "loss": 0.2776,
+      "epoch": 1.8870508523443976,
+      "grad_norm": 0.19846735894680023,
+      "learning_rate": 3.578470364486926e-05,
+      "loss": 0.4142,
       "step": 52360
     },
     {
-      "epoch": 1.84,
-      "learning_rate": 3.643286094574757e-05,
-      "loss": 0.2794,
+      "epoch": 1.8872310520056224,
+      "grad_norm": 0.1773809790611267,
+      "learning_rate": 3.578259748407673e-05,
+      "loss": 0.4116,
       "step": 52365
     },
     {
-      "epoch": 1.84,
-      "learning_rate": 3.643032748922655e-05,
-      "loss": 0.3072,
+      "epoch": 1.8874112516668469,
+      "grad_norm": 0.1703425943851471,
+      "learning_rate": 3.577996465086848e-05,
+      "loss": 0.4306,
       "step": 52370
     },
     {
-      "epoch": 1.84,
-      "learning_rate": 3.6427793884290265e-05,
-      "loss": 0.2788,
+      "epoch": 1.8875914513280714,
+      "grad_norm": 0.1680893748998642,
+      "learning_rate": 3.5777331670785305e-05,
+      "loss": 0.4214,
       "step": 52375
     },
     {
-      "epoch": 1.84,
-      "learning_rate": 3.6425260130971604e-05,
-      "loss": 0.2757,
+      "epoch": 1.8877716509892961,
+      "grad_norm": 0.2334313541650772,
+      "learning_rate": 3.577469854386308e-05,
+      "loss": 0.4327,
       "step": 52380
     },
     {
-      "epoch": 1.84,
-      "learning_rate": 3.642272622930347e-05,
-      "loss": 0.3123,
+      "epoch": 1.8879518506505208,
+      "grad_norm": 0.19660332798957825,
+      "learning_rate": 3.5772065270137665e-05,
+      "loss": 0.4319,
       "step": 52385
     },
     {
-      "epoch": 1.84,
-      "learning_rate": 3.642019217931877e-05,
-      "loss": 0.2864,
+      "epoch": 1.8881320503117456,
+      "grad_norm": 0.20360225439071655,
+      "learning_rate": 3.5769431849644955e-05,
+      "loss": 0.4157,
       "step": 52390
     },
     {
-      "epoch": 1.84,
-      "learning_rate": 3.6417657981050396e-05,
-      "loss": 0.2776,
+      "epoch": 1.88831224997297,
+      "grad_norm": 0.1686132401227951,
+      "learning_rate": 3.5766798282420814e-05,
+      "loss": 0.3861,
       "step": 52395
     },
     {
-      "epoch": 1.84,
-      "learning_rate": 3.641512363453126e-05,
-      "loss": 0.2893,
+      "epoch": 1.8884924496341946,
+      "grad_norm": 0.2118513137102127,
+      "learning_rate": 3.576416456850113e-05,
+      "loss": 0.4282,
       "step": 52400
     },
     {
-      "epoch": 1.84,
-      "learning_rate": 3.6412589139794264e-05,
-      "loss": 0.303,
+      "epoch": 1.8886726492954193,
+      "grad_norm": 0.16424524784088135,
+      "learning_rate": 3.5761530707921794e-05,
+      "loss": 0.4134,
       "step": 52405
     },
     {
-      "epoch": 1.84,
-      "learning_rate": 3.641005449687231e-05,
-      "loss": 0.2639,
+      "epoch": 1.888852848956644,
+      "grad_norm": 0.17415526509284973,
+      "learning_rate": 3.575889670071868e-05,
+      "loss": 0.4505,
       "step": 52410
     },
     {
-      "epoch": 1.84,
-      "learning_rate": 3.640751970579834e-05,
-      "loss": 0.3054,
+      "epoch": 1.8890330486178686,
+      "grad_norm": 0.21902436017990112,
+      "learning_rate": 3.575626254692768e-05,
+      "loss": 0.3874,
       "step": 52415
     },
     {
-      "epoch": 1.84,
-      "learning_rate": 3.6404984766605235e-05,
-      "loss": 0.2825,
+      "epoch": 1.889213248279093,
+      "grad_norm": 0.16111133992671967,
+      "learning_rate": 3.5753628246584694e-05,
+      "loss": 0.4127,
       "step": 52420
     },
     {
-      "epoch": 1.84,
-      "learning_rate": 3.640244967932591e-05,
-      "loss": 0.2827,
+      "epoch": 1.8893934479403178,
+      "grad_norm": 0.1768888384103775,
+      "learning_rate": 3.57509937997256e-05,
+      "loss": 0.4033,
       "step": 52425
     },
     {
-      "epoch": 1.84,
-      "learning_rate": 3.63999144439933e-05,
-      "loss": 0.2741,
+      "epoch": 1.8895736476015426,
+      "grad_norm": 0.18009909987449646,
+      "learning_rate": 3.574835920638629e-05,
+      "loss": 0.4052,
       "step": 52430
     },
     {
-      "epoch": 1.84,
-      "learning_rate": 3.639737906064031e-05,
-      "loss": 0.2635,
+      "epoch": 1.8897538472627673,
+      "grad_norm": 0.21755820512771606,
+      "learning_rate": 3.574572446660268e-05,
+      "loss": 0.4571,
       "step": 52435
     },
     {
-      "epoch": 1.84,
-      "learning_rate": 3.639484352929986e-05,
-      "loss": 0.2647,
+      "epoch": 1.8899340469239918,
+      "grad_norm": 0.21082909405231476,
+      "learning_rate": 3.574308958041064e-05,
+      "loss": 0.3826,
       "step": 52440
     },
     {
-      "epoch": 1.85,
-      "learning_rate": 3.639230785000489e-05,
-      "loss": 0.268,
+      "epoch": 1.8901142465852163,
+      "grad_norm": 0.23044735193252563,
+      "learning_rate": 3.57404545478461e-05,
+      "loss": 0.4161,
       "step": 52445
     },
     {
-      "epoch": 1.85,
-      "learning_rate": 3.6389772022788295e-05,
-      "loss": 0.2776,
+      "epoch": 1.890294446246441,
+      "grad_norm": 0.21793107688426971,
+      "learning_rate": 3.573781936894493e-05,
+      "loss": 0.4393,
       "step": 52450
     },
     {
-      "epoch": 1.85,
-      "learning_rate": 3.638723604768302e-05,
-      "loss": 0.2842,
+      "epoch": 1.8904746459076658,
+      "grad_norm": 0.15808548033237457,
+      "learning_rate": 3.573518404374306e-05,
+      "loss": 0.3924,
       "step": 52455
     },
     {
-      "epoch": 1.85,
-      "learning_rate": 3.638469992472199e-05,
-      "loss": 0.2796,
+      "epoch": 1.8906548455688903,
+      "grad_norm": 0.18602894246578217,
+      "learning_rate": 3.5732548572276386e-05,
+      "loss": 0.4046,
       "step": 52460
     },
     {
-      "epoch": 1.85,
-      "learning_rate": 3.638216365393814e-05,
-      "loss": 0.2851,
+      "epoch": 1.8908350452301148,
+      "grad_norm": 0.22336095571517944,
+      "learning_rate": 3.572991295458081e-05,
+      "loss": 0.4219,
       "step": 52465
     },
     {
-      "epoch": 1.85,
-      "learning_rate": 3.6379627235364395e-05,
-      "loss": 0.3005,
+      "epoch": 1.8910152448913395,
+      "grad_norm": 0.1639654040336609,
+      "learning_rate": 3.572727719069225e-05,
+      "loss": 0.4163,
       "step": 52470
     },
     {
-      "epoch": 1.85,
-      "learning_rate": 3.637709066903368e-05,
-      "loss": 0.2855,
+      "epoch": 1.8911954445525643,
+      "grad_norm": 0.14199529588222504,
+      "learning_rate": 3.572464128064662e-05,
+      "loss": 0.4071,
       "step": 52475
     },
     {
-      "epoch": 1.85,
-      "learning_rate": 3.6374553954978955e-05,
-      "loss": 0.2723,
+      "epoch": 1.891375644213789,
+      "grad_norm": 0.22127693891525269,
+      "learning_rate": 3.5722005224479826e-05,
+      "loss": 0.3902,
       "step": 52480
     },
     {
-      "epoch": 1.85,
-      "learning_rate": 3.637201709323314e-05,
-      "loss": 0.2943,
+      "epoch": 1.8915558438750135,
+      "grad_norm": 0.19781959056854248,
+      "learning_rate": 3.571936902222778e-05,
+      "loss": 0.4002,
       "step": 52485
     },
     {
-      "epoch": 1.85,
-      "learning_rate": 3.636948008382917e-05,
-      "loss": 0.2891,
+      "epoch": 1.891736043536238,
+      "grad_norm": 0.19771304726600647,
+      "learning_rate": 3.571673267392642e-05,
+      "loss": 0.4017,
       "step": 52490
     },
     {
-      "epoch": 1.85,
-      "learning_rate": 3.63669429268e-05,
-      "loss": 0.2916,
+      "epoch": 1.8919162431974628,
+      "grad_norm": 0.22566063702106476,
+      "learning_rate": 3.571409617961164e-05,
+      "loss": 0.3925,
       "step": 52495
     },
     {
-      "epoch": 1.85,
-      "learning_rate": 3.636440562217856e-05,
-      "loss": 0.2823,
+      "epoch": 1.8920964428586875,
+      "grad_norm": 0.1664046347141266,
+      "learning_rate": 3.571145953931938e-05,
+      "loss": 0.3956,
       "step": 52500
     },
     {
-      "epoch": 1.85,
-      "eval_loss": 0.27478304505348206,
-      "eval_runtime": 10.5283,
-      "eval_samples_per_second": 9.498,
-      "eval_steps_per_second": 9.498,
+      "epoch": 1.8920964428586875,
+      "eval_loss": 0.4388228952884674,
+      "eval_runtime": 3.5346,
+      "eval_samples_per_second": 28.292,
+      "eval_steps_per_second": 7.073,
       "step": 52500
     },
     {
-      "epoch": 1.85,
-      "learning_rate": 3.636186816999781e-05,
-      "loss": 0.2729,
+      "epoch": 1.892276642519912,
+      "grad_norm": 0.18760746717453003,
+      "learning_rate": 3.5708822753085555e-05,
+      "loss": 0.407,
       "step": 52505
     },
     {
-      "epoch": 1.85,
-      "learning_rate": 3.6359330570290686e-05,
-      "loss": 0.2818,
+      "epoch": 1.8924568421811367,
+      "grad_norm": 0.19066941738128662,
+      "learning_rate": 3.5706185820946094e-05,
+      "loss": 0.3779,
       "step": 52510
     },
     {
-      "epoch": 1.85,
-      "learning_rate": 3.635679282309013e-05,
-      "loss": 0.2786,
+      "epoch": 1.8926370418423613,
+      "grad_norm": 0.18899337947368622,
+      "learning_rate": 3.570354874293692e-05,
+      "loss": 0.355,
       "step": 52515
     },
     {
-      "epoch": 1.85,
-      "learning_rate": 3.635425492842912e-05,
-      "loss": 0.2766,
+      "epoch": 1.892817241503586,
+      "grad_norm": 0.22324331104755402,
+      "learning_rate": 3.570091151909397e-05,
+      "loss": 0.406,
       "step": 52520
     },
     {
-      "epoch": 1.85,
-      "learning_rate": 3.6351716886340585e-05,
-      "loss": 0.287,
+      "epoch": 1.8929974411648107,
+      "grad_norm": 0.18958257138729095,
+      "learning_rate": 3.569827414945317e-05,
+      "loss": 0.426,
       "step": 52525
     },
     {
-      "epoch": 1.85,
-      "learning_rate": 3.6349178696857496e-05,
-      "loss": 0.2887,
+      "epoch": 1.8931776408260352,
+      "grad_norm": 0.20323504507541656,
+      "learning_rate": 3.5695636634050466e-05,
+      "loss": 0.4198,
       "step": 52530
     },
     {
-      "epoch": 1.85,
-      "learning_rate": 3.634664036001279e-05,
-      "loss": 0.2625,
+      "epoch": 1.8933578404872597,
+      "grad_norm": 0.2171027809381485,
+      "learning_rate": 3.569299897292177e-05,
+      "loss": 0.4075,
       "step": 52535
     },
     {
-      "epoch": 1.85,
-      "learning_rate": 3.634410187583944e-05,
-      "loss": 0.2751,
+      "epoch": 1.8935380401484845,
+      "grad_norm": 0.17089755833148956,
+      "learning_rate": 3.5690361166103045e-05,
+      "loss": 0.3932,
       "step": 52540
     },
     {
-      "epoch": 1.85,
-      "learning_rate": 3.6341563244370405e-05,
-      "loss": 0.303,
+      "epoch": 1.8937182398097092,
+      "grad_norm": 0.21072880923748016,
+      "learning_rate": 3.568772321363021e-05,
+      "loss": 0.4229,
       "step": 52545
     },
     {
-      "epoch": 1.85,
-      "learning_rate": 3.633902446563865e-05,
-      "loss": 0.2807,
+      "epoch": 1.893898439470934,
+      "grad_norm": 0.20696797966957092,
+      "learning_rate": 3.5685085115539225e-05,
+      "loss": 0.4261,
       "step": 52550
     },
     {
-      "epoch": 1.85,
-      "learning_rate": 3.633648553967712e-05,
-      "loss": 0.2919,
+      "epoch": 1.8940786391321585,
+      "grad_norm": 0.1554139405488968,
+      "learning_rate": 3.568244687186602e-05,
+      "loss": 0.388,
       "step": 52555
     },
     {
-      "epoch": 1.85,
-      "learning_rate": 3.633394646651881e-05,
-      "loss": 0.2839,
+      "epoch": 1.894258838793383,
+      "grad_norm": 0.19356495141983032,
+      "learning_rate": 3.5679808482646535e-05,
+      "loss": 0.4374,
       "step": 52560
     },
     {
-      "epoch": 1.85,
-      "learning_rate": 3.6331407246196674e-05,
-      "loss": 0.2871,
+      "epoch": 1.8944390384546077,
+      "grad_norm": 0.14796990156173706,
+      "learning_rate": 3.567716994791674e-05,
+      "loss": 0.3883,
       "step": 52565
     },
     {
-      "epoch": 1.85,
-      "learning_rate": 3.632886787874368e-05,
-      "loss": 0.2677,
+      "epoch": 1.8946192381158324,
+      "grad_norm": 0.21466509997844696,
+      "learning_rate": 3.5674531267712566e-05,
+      "loss": 0.435,
       "step": 52570
     },
     {
-      "epoch": 1.85,
-      "learning_rate": 3.6326328364192795e-05,
-      "loss": 0.2702,
+      "epoch": 1.894799437777057,
+      "grad_norm": 0.16786347329616547,
+      "learning_rate": 3.567189244206996e-05,
+      "loss": 0.3936,
       "step": 52575
     },
     {
-      "epoch": 1.85,
-      "learning_rate": 3.632378870257701e-05,
-      "loss": 0.28,
+      "epoch": 1.8949796374382815,
+      "grad_norm": 0.2334979772567749,
+      "learning_rate": 3.5669253471024904e-05,
+      "loss": 0.4226,
       "step": 52580
     },
     {
-      "epoch": 1.85,
-      "learning_rate": 3.632124889392929e-05,
-      "loss": 0.2936,
+      "epoch": 1.8951598370995062,
+      "grad_norm": 0.14717628061771393,
+      "learning_rate": 3.5666614354613325e-05,
+      "loss": 0.3976,
       "step": 52585
     },
     {
-      "epoch": 1.85,
-      "learning_rate": 3.631870893828262e-05,
-      "loss": 0.2671,
+      "epoch": 1.895340036760731,
+      "grad_norm": 0.17818215489387512,
+      "learning_rate": 3.5663975092871194e-05,
+      "loss": 0.3997,
       "step": 52590
     },
     {
-      "epoch": 1.85,
-      "learning_rate": 3.631616883566996e-05,
-      "loss": 0.2628,
+      "epoch": 1.8955202364219557,
+      "grad_norm": 0.20797117054462433,
+      "learning_rate": 3.5661335685834466e-05,
+      "loss": 0.4407,
       "step": 52595
     },
     {
-      "epoch": 1.85,
-      "learning_rate": 3.6313628586124317e-05,
-      "loss": 0.284,
+      "epoch": 1.8957004360831802,
+      "grad_norm": 0.17417512834072113,
+      "learning_rate": 3.56586961335391e-05,
+      "loss": 0.4273,
       "step": 52600
     },
     {
-      "epoch": 1.85,
-      "learning_rate": 3.631108818967865e-05,
-      "loss": 0.2837,
+      "epoch": 1.8958806357444047,
+      "grad_norm": 0.22176715731620789,
+      "learning_rate": 3.565605643602107e-05,
+      "loss": 0.4448,
       "step": 52605
     },
     {
-      "epoch": 1.85,
-      "learning_rate": 3.630854764636597e-05,
-      "loss": 0.2802,
+      "epoch": 1.8960608354056294,
+      "grad_norm": 0.1701037585735321,
+      "learning_rate": 3.565341659331633e-05,
+      "loss": 0.3807,
       "step": 52610
     },
     {
-      "epoch": 1.85,
-      "learning_rate": 3.630600695621924e-05,
-      "loss": 0.2782,
+      "epoch": 1.8962410350668542,
+      "grad_norm": 0.17541612684726715,
+      "learning_rate": 3.565077660546085e-05,
+      "loss": 0.4121,
       "step": 52615
     },
     {
-      "epoch": 1.85,
-      "learning_rate": 3.630346611927147e-05,
-      "loss": 0.2628,
+      "epoch": 1.8964212347280787,
+      "grad_norm": 0.17228418588638306,
+      "learning_rate": 3.5648136472490604e-05,
+      "loss": 0.4114,
       "step": 52620
     },
     {
-      "epoch": 1.85,
-      "learning_rate": 3.6300925135555634e-05,
-      "loss": 0.286,
+      "epoch": 1.8966014343893034,
+      "grad_norm": 0.1823887676000595,
+      "learning_rate": 3.5645496194441555e-05,
+      "loss": 0.422,
       "step": 52625
     },
     {
-      "epoch": 1.85,
-      "learning_rate": 3.6298384005104736e-05,
-      "loss": 0.2759,
+      "epoch": 1.896781634050528,
+      "grad_norm": 0.181901216506958,
+      "learning_rate": 3.564285577134969e-05,
+      "loss": 0.4025,
       "step": 52630
     },
     {
-      "epoch": 1.85,
-      "learning_rate": 3.629584272795177e-05,
-      "loss": 0.2929,
+      "epoch": 1.8969618337117526,
+      "grad_norm": 0.19372449815273285,
+      "learning_rate": 3.564021520325096e-05,
+      "loss": 0.4179,
       "step": 52635
     },
     {
-      "epoch": 1.85,
-      "learning_rate": 3.629330130412974e-05,
-      "loss": 0.285,
+      "epoch": 1.8971420333729774,
+      "grad_norm": 0.2167021781206131,
+      "learning_rate": 3.5637574490181376e-05,
+      "loss": 0.4143,
       "step": 52640
     },
     {
-      "epoch": 1.85,
-      "learning_rate": 3.629075973367162e-05,
-      "loss": 0.2624,
+      "epoch": 1.8973222330342019,
+      "grad_norm": 0.20177282392978668,
+      "learning_rate": 3.563493363217689e-05,
+      "loss": 0.3899,
       "step": 52645
     },
     {
-      "epoch": 1.85,
-      "learning_rate": 3.628821801661043e-05,
-      "loss": 0.2926,
+      "epoch": 1.8975024326954264,
+      "grad_norm": 0.18920306861400604,
+      "learning_rate": 3.563229262927349e-05,
+      "loss": 0.4006,
       "step": 52650
     },
     {
-      "epoch": 1.85,
-      "learning_rate": 3.628567615297916e-05,
-      "loss": 0.2976,
+      "epoch": 1.8976826323566511,
+      "grad_norm": 0.14058130979537964,
+      "learning_rate": 3.562965148150716e-05,
+      "loss": 0.3701,
       "step": 52655
     },
     {
-      "epoch": 1.85,
-      "learning_rate": 3.628313414281084e-05,
-      "loss": 0.3159,
+      "epoch": 1.8978628320178759,
+      "grad_norm": 0.20638848841190338,
+      "learning_rate": 3.56270101889139e-05,
+      "loss": 0.4196,
       "step": 52660
     },
     {
-      "epoch": 1.85,
-      "learning_rate": 3.6280591986138446e-05,
-      "loss": 0.2822,
+      "epoch": 1.8980430316791006,
+      "grad_norm": 0.22467444837093353,
+      "learning_rate": 3.562436875152967e-05,
+      "loss": 0.4454,
       "step": 52665
     },
     {
-      "epoch": 1.85,
-      "learning_rate": 3.6278049682995e-05,
-      "loss": 0.2769,
+      "epoch": 1.8982232313403251,
+      "grad_norm": 0.17210647463798523,
+      "learning_rate": 3.562172716939048e-05,
+      "loss": 0.4351,
       "step": 52670
     },
     {
-      "epoch": 1.85,
-      "learning_rate": 3.6275507233413516e-05,
-      "loss": 0.2939,
+      "epoch": 1.8984034310015496,
+      "grad_norm": 0.1578112095594406,
+      "learning_rate": 3.561908544253231e-05,
+      "loss": 0.4413,
       "step": 52675
     },
     {
-      "epoch": 1.85,
-      "learning_rate": 3.6272964637427e-05,
-      "loss": 0.2559,
+      "epoch": 1.8985836306627744,
+      "grad_norm": 0.15398284792900085,
+      "learning_rate": 3.561644357099116e-05,
+      "loss": 0.3976,
       "step": 52680
     },
     {
-      "epoch": 1.85,
-      "learning_rate": 3.6270421895068465e-05,
-      "loss": 0.2857,
+      "epoch": 1.898763830323999,
+      "grad_norm": 0.18252988159656525,
+      "learning_rate": 3.561380155480302e-05,
+      "loss": 0.4166,
       "step": 52685
     },
     {
-      "epoch": 1.85,
-      "learning_rate": 3.6267879006370936e-05,
-      "loss": 0.2951,
+      "epoch": 1.8989440299852236,
+      "grad_norm": 0.16655157506465912,
+      "learning_rate": 3.561115939400389e-05,
+      "loss": 0.4013,
       "step": 52690
     },
     {
-      "epoch": 1.85,
-      "learning_rate": 3.626533597136742e-05,
-      "loss": 0.2767,
+      "epoch": 1.8991242296464481,
+      "grad_norm": 0.1663331389427185,
+      "learning_rate": 3.560851708862977e-05,
+      "loss": 0.3884,
       "step": 52695
     },
     {
-      "epoch": 1.85,
-      "learning_rate": 3.626279279009094e-05,
-      "loss": 0.2571,
+      "epoch": 1.8993044293076728,
+      "grad_norm": 0.16247087717056274,
+      "learning_rate": 3.560587463871665e-05,
+      "loss": 0.3918,
       "step": 52700
     },
     {
-      "epoch": 1.85,
-      "learning_rate": 3.626024946257452e-05,
-      "loss": 0.3061,
+      "epoch": 1.8994846289688976,
+      "grad_norm": 0.18449072539806366,
+      "learning_rate": 3.560323204430055e-05,
+      "loss": 0.4069,
       "step": 52705
     },
     {
-      "epoch": 1.85,
-      "learning_rate": 3.6257705988851176e-05,
-      "loss": 0.2548,
+      "epoch": 1.8996648286301223,
+      "grad_norm": 0.19125069677829742,
+      "learning_rate": 3.560058930541746e-05,
+      "loss": 0.4051,
       "step": 52710
     },
     {
-      "epoch": 1.85,
-      "learning_rate": 3.6255162368953946e-05,
-      "loss": 0.2965,
+      "epoch": 1.8998450282913468,
+      "grad_norm": 0.15312950313091278,
+      "learning_rate": 3.55979464221034e-05,
+      "loss": 0.4207,
       "step": 52715
     },
     {
-      "epoch": 1.85,
-      "learning_rate": 3.6252618602915845e-05,
-      "loss": 0.307,
+      "epoch": 1.9000252279525713,
+      "grad_norm": 0.2004600614309311,
+      "learning_rate": 3.559530339439436e-05,
+      "loss": 0.4074,
       "step": 52720
     },
     {
-      "epoch": 1.86,
-      "learning_rate": 3.6250074690769906e-05,
-      "loss": 0.2867,
+      "epoch": 1.900205427613796,
+      "grad_norm": 0.1745137721300125,
+      "learning_rate": 3.5592660222326375e-05,
+      "loss": 0.3963,
       "step": 52725
     },
     {
-      "epoch": 1.86,
-      "learning_rate": 3.6247530632549165e-05,
-      "loss": 0.2846,
+      "epoch": 1.9003856272750208,
+      "grad_norm": 0.15934212505817413,
+      "learning_rate": 3.5590016905935436e-05,
+      "loss": 0.3571,
       "step": 52730
     },
     {
-      "epoch": 1.86,
-      "learning_rate": 3.6244986428286656e-05,
-      "loss": 0.2995,
+      "epoch": 1.9005658269362453,
+      "grad_norm": 0.20348936319351196,
+      "learning_rate": 3.558737344525758e-05,
+      "loss": 0.4435,
       "step": 52735
     },
     {
-      "epoch": 1.86,
-      "learning_rate": 3.62424420780154e-05,
-      "loss": 0.2915,
+      "epoch": 1.90074602659747,
+      "grad_norm": 0.20140963792800903,
+      "learning_rate": 3.558472984032879e-05,
+      "loss": 0.4502,
       "step": 52740
     },
     {
-      "epoch": 1.86,
-      "learning_rate": 3.623989758176844e-05,
-      "loss": 0.2876,
+      "epoch": 1.9009262262586946,
+      "grad_norm": 0.16998355090618134,
+      "learning_rate": 3.558208609118512e-05,
+      "loss": 0.4119,
       "step": 52745
     },
     {
-      "epoch": 1.86,
-      "learning_rate": 3.623735293957883e-05,
-      "loss": 0.2866,
+      "epoch": 1.9011064259199193,
+      "grad_norm": 0.17498774826526642,
+      "learning_rate": 3.5579442197862575e-05,
+      "loss": 0.3913,
       "step": 52750
     },
     {
-      "epoch": 1.86,
-      "learning_rate": 3.623480815147959e-05,
-      "loss": 0.299,
+      "epoch": 1.901286625581144,
+      "grad_norm": 0.1859028935432434,
+      "learning_rate": 3.557679816039717e-05,
+      "loss": 0.4298,
       "step": 52755
     },
     {
-      "epoch": 1.86,
-      "learning_rate": 3.623226321750377e-05,
-      "loss": 0.3092,
+      "epoch": 1.9014668252423685,
+      "grad_norm": 0.21767857670783997,
+      "learning_rate": 3.5574153978824945e-05,
+      "loss": 0.4364,
       "step": 52760
     },
     {
-      "epoch": 1.86,
-      "learning_rate": 3.622971813768441e-05,
-      "loss": 0.3142,
+      "epoch": 1.901647024903593,
+      "grad_norm": 0.1995837390422821,
+      "learning_rate": 3.557150965318192e-05,
+      "loss": 0.408,
       "step": 52765
     },
     {
-      "epoch": 1.86,
-      "learning_rate": 3.622717291205457e-05,
-      "loss": 0.2707,
+      "epoch": 1.9018272245648178,
+      "grad_norm": 0.22697186470031738,
+      "learning_rate": 3.5568865183504127e-05,
+      "loss": 0.4431,
       "step": 52770
     },
     {
-      "epoch": 1.86,
-      "learning_rate": 3.6224627540647286e-05,
-      "loss": 0.2993,
+      "epoch": 1.9020074242260425,
+      "grad_norm": 0.17042651772499084,
+      "learning_rate": 3.556622056982758e-05,
+      "loss": 0.3834,
       "step": 52775
     },
     {
-      "epoch": 1.86,
-      "learning_rate": 3.622208202349561e-05,
-      "loss": 0.2707,
+      "epoch": 1.9021876238872673,
+      "grad_norm": 0.16556668281555176,
+      "learning_rate": 3.556357581218833e-05,
+      "loss": 0.4083,
       "step": 52780
     },
     {
-      "epoch": 1.86,
-      "learning_rate": 3.621953636063259e-05,
-      "loss": 0.2887,
+      "epoch": 1.9023678235484918,
+      "grad_norm": 0.16349823772907257,
+      "learning_rate": 3.556093091062241e-05,
+      "loss": 0.3985,
       "step": 52785
     },
     {
-      "epoch": 1.86,
-      "learning_rate": 3.621699055209129e-05,
-      "loss": 0.2877,
+      "epoch": 1.9025480232097163,
+      "grad_norm": 0.18236073851585388,
+      "learning_rate": 3.555828586516584e-05,
+      "loss": 0.4089,
       "step": 52790
     },
     {
-      "epoch": 1.86,
-      "learning_rate": 3.621444459790476e-05,
-      "loss": 0.2692,
+      "epoch": 1.902728222870941,
+      "grad_norm": 0.15176883339881897,
+      "learning_rate": 3.5555640675854675e-05,
+      "loss": 0.4,
       "step": 52795
     },
     {
-      "epoch": 1.86,
-      "learning_rate": 3.621189849810606e-05,
-      "loss": 0.2881,
+      "epoch": 1.9029084225321657,
+      "grad_norm": 0.16588926315307617,
+      "learning_rate": 3.555299534272495e-05,
+      "loss": 0.4054,
       "step": 52800
     },
     {
-      "epoch": 1.86,
-      "learning_rate": 3.620935225272823e-05,
-      "loss": 0.2931,
+      "epoch": 1.9030886221933903,
+      "grad_norm": 0.17433229088783264,
+      "learning_rate": 3.55503498658127e-05,
+      "loss": 0.4082,
       "step": 52805
     },
     {
-      "epoch": 1.86,
-      "learning_rate": 3.620680586180437e-05,
-      "loss": 0.2877,
+      "epoch": 1.9032688218546148,
+      "grad_norm": 0.178829625248909,
+      "learning_rate": 3.5547704245153984e-05,
+      "loss": 0.4344,
       "step": 52810
     },
     {
-      "epoch": 1.86,
-      "learning_rate": 3.620425932536751e-05,
-      "loss": 0.2648,
+      "epoch": 1.9034490215158395,
+      "grad_norm": 0.259390652179718,
+      "learning_rate": 3.554505848078483e-05,
+      "loss": 0.407,
       "step": 52815
     },
     {
-      "epoch": 1.86,
-      "learning_rate": 3.620171264345073e-05,
-      "loss": 0.2814,
+      "epoch": 1.9036292211770642,
+      "grad_norm": 0.20816822350025177,
+      "learning_rate": 3.554241257274131e-05,
+      "loss": 0.4278,
       "step": 52820
     },
     {
-      "epoch": 1.86,
-      "learning_rate": 3.619916581608709e-05,
-      "loss": 0.2752,
+      "epoch": 1.903809420838289,
+      "grad_norm": 0.22725282609462738,
+      "learning_rate": 3.5539766521059455e-05,
+      "loss": 0.414,
       "step": 52825
     },
     {
-      "epoch": 1.86,
-      "learning_rate": 3.619661884330967e-05,
-      "loss": 0.2623,
+      "epoch": 1.9039896204995135,
+      "grad_norm": 0.20979876816272736,
+      "learning_rate": 3.553712032577532e-05,
+      "loss": 0.4301,
       "step": 52830
     },
     {
-      "epoch": 1.86,
-      "learning_rate": 3.619407172515153e-05,
-      "loss": 0.2936,
+      "epoch": 1.904169820160738,
+      "grad_norm": 0.16655777394771576,
+      "learning_rate": 3.5534473986924954e-05,
+      "loss": 0.4324,
       "step": 52835
     },
     {
-      "epoch": 1.86,
-      "learning_rate": 3.6191524461645736e-05,
-      "loss": 0.2566,
+      "epoch": 1.9043500198219627,
+      "grad_norm": 0.165932297706604,
+      "learning_rate": 3.553182750454442e-05,
+      "loss": 0.4295,
       "step": 52840
     },
     {
-      "epoch": 1.86,
-      "learning_rate": 3.618897705282538e-05,
-      "loss": 0.2904,
+      "epoch": 1.9045302194831875,
+      "grad_norm": 0.1914936900138855,
+      "learning_rate": 3.552918087866979e-05,
+      "loss": 0.4105,
       "step": 52845
     },
     {
-      "epoch": 1.86,
-      "learning_rate": 3.6186429498723535e-05,
-      "loss": 0.2571,
+      "epoch": 1.904710419144412,
+      "grad_norm": 0.15866640210151672,
+      "learning_rate": 3.552653410933709e-05,
+      "loss": 0.3874,
       "step": 52850
     },
     {
-      "epoch": 1.86,
-      "learning_rate": 3.618388179937327e-05,
-      "loss": 0.2748,
+      "epoch": 1.9048906188056365,
+      "grad_norm": 0.15294025838375092,
+      "learning_rate": 3.552388719658242e-05,
+      "loss": 0.3981,
       "step": 52855
     },
     {
-      "epoch": 1.86,
-      "learning_rate": 3.6181333954807674e-05,
-      "loss": 0.3076,
+      "epoch": 1.9050708184668612,
+      "grad_norm": 0.19104799628257751,
+      "learning_rate": 3.552124014044181e-05,
+      "loss": 0.4481,
       "step": 52860
     },
     {
-      "epoch": 1.86,
-      "learning_rate": 3.617878596505982e-05,
-      "loss": 0.2754,
+      "epoch": 1.905251018128086,
+      "grad_norm": 0.1697293370962143,
+      "learning_rate": 3.551859294095135e-05,
+      "loss": 0.4,
       "step": 52865
     },
     {
-      "epoch": 1.86,
-      "learning_rate": 3.6176237830162804e-05,
-      "loss": 0.2669,
+      "epoch": 1.9054312177893107,
+      "grad_norm": 0.17631365358829498,
+      "learning_rate": 3.551594559814709e-05,
+      "loss": 0.3584,
       "step": 52870
     },
     {
-      "epoch": 1.86,
-      "learning_rate": 3.61736895501497e-05,
-      "loss": 0.2724,
+      "epoch": 1.9056114174505352,
+      "grad_norm": 0.20345862209796906,
+      "learning_rate": 3.551329811206511e-05,
+      "loss": 0.4251,
       "step": 52875
     },
     {
-      "epoch": 1.86,
-      "learning_rate": 3.6171141125053595e-05,
-      "loss": 0.2857,
+      "epoch": 1.9057916171117597,
+      "grad_norm": 0.1694919317960739,
+      "learning_rate": 3.551065048274147e-05,
+      "loss": 0.4144,
       "step": 52880
     },
     {
-      "epoch": 1.86,
-      "learning_rate": 3.61685925549076e-05,
-      "loss": 0.2971,
+      "epoch": 1.9059718167729844,
+      "grad_norm": 0.1483502835035324,
+      "learning_rate": 3.550800271021226e-05,
+      "loss": 0.3912,
       "step": 52885
     },
     {
-      "epoch": 1.86,
-      "learning_rate": 3.616604383974478e-05,
-      "loss": 0.2753,
+      "epoch": 1.9061520164342092,
+      "grad_norm": 0.169077530503273,
+      "learning_rate": 3.550535479451356e-05,
+      "loss": 0.4053,
       "step": 52890
     },
     {
-      "epoch": 1.86,
-      "learning_rate": 3.616349497959824e-05,
-      "loss": 0.3067,
+      "epoch": 1.906332216095434,
+      "grad_norm": 0.20600876212120056,
+      "learning_rate": 3.550270673568141e-05,
+      "loss": 0.4135,
       "step": 52895
     },
     {
-      "epoch": 1.86,
-      "learning_rate": 3.6160945974501077e-05,
-      "loss": 0.2854,
+      "epoch": 1.9065124157566584,
+      "grad_norm": 0.2073904126882553,
+      "learning_rate": 3.550005853375193e-05,
+      "loss": 0.4526,
       "step": 52900
     },
     {
-      "epoch": 1.86,
-      "learning_rate": 3.6158396824486385e-05,
-      "loss": 0.287,
+      "epoch": 1.906692615417883,
+      "grad_norm": 0.17720499634742737,
+      "learning_rate": 3.549741018876118e-05,
+      "loss": 0.4018,
       "step": 52905
     },
     {
-      "epoch": 1.86,
-      "learning_rate": 3.6155847529587264e-05,
-      "loss": 0.3051,
+      "epoch": 1.9068728150791077,
+      "grad_norm": 0.1841294914484024,
+      "learning_rate": 3.549476170074526e-05,
+      "loss": 0.4054,
       "step": 52910
     },
     {
-      "epoch": 1.86,
-      "learning_rate": 3.6153298089836815e-05,
-      "loss": 0.2939,
+      "epoch": 1.9070530147403324,
+      "grad_norm": 0.20570293068885803,
+      "learning_rate": 3.549211306974023e-05,
+      "loss": 0.4207,
       "step": 52915
     },
     {
-      "epoch": 1.86,
-      "learning_rate": 3.615074850526815e-05,
-      "loss": 0.2884,
+      "epoch": 1.907233214401557,
+      "grad_norm": 0.1706700176000595,
+      "learning_rate": 3.548946429578219e-05,
+      "loss": 0.4166,
       "step": 52920
     },
     {
-      "epoch": 1.86,
-      "learning_rate": 3.614819877591435e-05,
-      "loss": 0.3121,
+      "epoch": 1.9074134140627814,
+      "grad_norm": 0.19486333429813385,
+      "learning_rate": 3.5486815378907234e-05,
+      "loss": 0.4408,
       "step": 52925
     },
     {
-      "epoch": 1.86,
-      "learning_rate": 3.6145648901808535e-05,
-      "loss": 0.2799,
+      "epoch": 1.9075936137240062,
+      "grad_norm": 0.15442302823066711,
+      "learning_rate": 3.548416631915146e-05,
+      "loss": 0.4376,
       "step": 52930
     },
     {
-      "epoch": 1.86,
-      "learning_rate": 3.614309888298383e-05,
-      "loss": 0.3066,
+      "epoch": 1.9077738133852309,
+      "grad_norm": 0.1942632794380188,
+      "learning_rate": 3.5481517116550936e-05,
+      "loss": 0.3904,
       "step": 52935
     },
     {
-      "epoch": 1.86,
-      "learning_rate": 3.614054871947332e-05,
-      "loss": 0.2771,
+      "epoch": 1.9079540130464556,
+      "grad_norm": 0.16237981617450714,
+      "learning_rate": 3.547886777114177e-05,
+      "loss": 0.353,
       "step": 52940
     },
     {
-      "epoch": 1.86,
-      "learning_rate": 3.6137998411310126e-05,
-      "loss": 0.2908,
+      "epoch": 1.9081342127076801,
+      "grad_norm": 0.2059374898672104,
+      "learning_rate": 3.5476218282960064e-05,
+      "loss": 0.4277,
       "step": 52945
     },
     {
-      "epoch": 1.86,
-      "learning_rate": 3.613544795852736e-05,
-      "loss": 0.2769,
+      "epoch": 1.9083144123689046,
+      "grad_norm": 0.1902727484703064,
+      "learning_rate": 3.547356865204191e-05,
+      "loss": 0.4187,
       "step": 52950
     },
     {
-      "epoch": 1.86,
-      "learning_rate": 3.6132897361158145e-05,
-      "loss": 0.2981,
+      "epoch": 1.9084946120301294,
+      "grad_norm": 0.19110523164272308,
+      "learning_rate": 3.5470918878423414e-05,
+      "loss": 0.4122,
       "step": 52955
     },
     {
-      "epoch": 1.86,
-      "learning_rate": 3.613034661923559e-05,
-      "loss": 0.2712,
+      "epoch": 1.9086748116913541,
+      "grad_norm": 0.19622482359409332,
+      "learning_rate": 3.546826896214067e-05,
+      "loss": 0.3987,
       "step": 52960
     },
     {
-      "epoch": 1.86,
-      "learning_rate": 3.612779573279283e-05,
-      "loss": 0.2836,
+      "epoch": 1.9088550113525786,
+      "grad_norm": 0.19886255264282227,
+      "learning_rate": 3.546561890322979e-05,
+      "loss": 0.4114,
       "step": 52965
     },
     {
-      "epoch": 1.86,
-      "learning_rate": 3.612524470186297e-05,
-      "loss": 0.2873,
+      "epoch": 1.9090352110138031,
+      "grad_norm": 0.1465664505958557,
+      "learning_rate": 3.546296870172689e-05,
+      "loss": 0.3839,
       "step": 52970
     },
     {
-      "epoch": 1.86,
-      "learning_rate": 3.6122693526479135e-05,
-      "loss": 0.2933,
+      "epoch": 1.9092154106750279,
+      "grad_norm": 0.19967186450958252,
+      "learning_rate": 3.546031835766806e-05,
+      "loss": 0.4433,
       "step": 52975
     },
     {
-      "epoch": 1.86,
-      "learning_rate": 3.6120142206674455e-05,
-      "loss": 0.2809,
+      "epoch": 1.9093956103362526,
+      "grad_norm": 0.17404790222644806,
+      "learning_rate": 3.545766787108941e-05,
+      "loss": 0.4215,
       "step": 52980
     },
     {
-      "epoch": 1.86,
-      "learning_rate": 3.611759074248206e-05,
-      "loss": 0.2955,
+      "epoch": 1.9095758099974773,
+      "grad_norm": 0.15910834074020386,
+      "learning_rate": 3.545501724202706e-05,
+      "loss": 0.3836,
       "step": 52985
     },
     {
-      "epoch": 1.86,
-      "learning_rate": 3.611503913393508e-05,
-      "loss": 0.2707,
+      "epoch": 1.9097560096587018,
+      "grad_norm": 0.17526118457317352,
+      "learning_rate": 3.545236647051713e-05,
+      "loss": 0.3867,
       "step": 52990
     },
     {
-      "epoch": 1.86,
-      "learning_rate": 3.611248738106663e-05,
-      "loss": 0.2893,
+      "epoch": 1.9099362093199264,
+      "grad_norm": 0.1866360753774643,
+      "learning_rate": 3.544971555659574e-05,
+      "loss": 0.3812,
       "step": 52995
     },
     {
-      "epoch": 1.86,
-      "learning_rate": 3.6109935483909876e-05,
-      "loss": 0.3082,
+      "epoch": 1.910116408981151,
+      "grad_norm": 0.1662186235189438,
+      "learning_rate": 3.544706450029898e-05,
+      "loss": 0.4114,
       "step": 53000
     },
     {
-      "epoch": 1.86,
-      "eval_loss": 0.27502691745758057,
-      "eval_runtime": 10.5394,
-      "eval_samples_per_second": 9.488,
-      "eval_steps_per_second": 9.488,
+      "epoch": 1.910116408981151,
+      "eval_loss": 0.4382114112377167,
+      "eval_runtime": 3.5422,
+      "eval_samples_per_second": 28.231,
+      "eval_steps_per_second": 7.058,
       "step": 53000
     },
     {
-      "epoch": 1.86,
-      "learning_rate": 3.610738344249791e-05,
-      "loss": 0.2816,
+      "epoch": 1.9102966086423758,
+      "grad_norm": 0.1859690397977829,
+      "learning_rate": 3.5444413301662996e-05,
+      "loss": 0.4291,
       "step": 53005
     },
     {
-      "epoch": 1.87,
-      "learning_rate": 3.6104831256863905e-05,
-      "loss": 0.2889,
+      "epoch": 1.9104768083036003,
+      "grad_norm": 0.2037961333990097,
+      "learning_rate": 3.544176196072391e-05,
+      "loss": 0.4178,
       "step": 53010
     },
     {
-      "epoch": 1.87,
-      "learning_rate": 3.610227892704097e-05,
-      "loss": 0.2816,
+      "epoch": 1.910657007964825,
+      "grad_norm": 0.19950434565544128,
+      "learning_rate": 3.543911047751783e-05,
+      "loss": 0.3977,
       "step": 53015
     },
     {
-      "epoch": 1.87,
-      "learning_rate": 3.609972645306227e-05,
-      "loss": 0.2639,
+      "epoch": 1.9108372076260496,
+      "grad_norm": 0.169523224234581,
+      "learning_rate": 3.5436458852080895e-05,
+      "loss": 0.4187,
       "step": 53020
     },
     {
-      "epoch": 1.87,
-      "learning_rate": 3.6097173834960935e-05,
-      "loss": 0.2738,
+      "epoch": 1.9110174072872743,
+      "grad_norm": 0.19487643241882324,
+      "learning_rate": 3.543380708444923e-05,
+      "loss": 0.4207,
       "step": 53025
     },
     {
-      "epoch": 1.87,
-      "learning_rate": 3.609462107277012e-05,
-      "loss": 0.2803,
+      "epoch": 1.911197606948499,
+      "grad_norm": 0.1871878057718277,
+      "learning_rate": 3.543115517465896e-05,
+      "loss": 0.4376,
       "step": 53030
     },
     {
-      "epoch": 1.87,
-      "learning_rate": 3.609206816652295e-05,
-      "loss": 0.2838,
+      "epoch": 1.9113778066097236,
+      "grad_norm": 0.17613734304904938,
+      "learning_rate": 3.542850312274622e-05,
+      "loss": 0.4136,
       "step": 53035
     },
     {
-      "epoch": 1.87,
-      "learning_rate": 3.6089515116252584e-05,
-      "loss": 0.2923,
+      "epoch": 1.911558006270948,
+      "grad_norm": 0.17650562524795532,
+      "learning_rate": 3.542585092874714e-05,
+      "loss": 0.3665,
       "step": 53040
     },
     {
-      "epoch": 1.87,
-      "learning_rate": 3.6086961921992173e-05,
-      "loss": 0.283,
+      "epoch": 1.9117382059321728,
+      "grad_norm": 0.16049924492835999,
+      "learning_rate": 3.542319859269787e-05,
+      "loss": 0.3824,
       "step": 53045
     },
     {
-      "epoch": 1.87,
-      "learning_rate": 3.6084408583774875e-05,
-      "loss": 0.2805,
+      "epoch": 1.9119184055933975,
+      "grad_norm": 0.1577320694923401,
+      "learning_rate": 3.5420546114634535e-05,
+      "loss": 0.3998,
       "step": 53050
     },
     {
-      "epoch": 1.87,
-      "learning_rate": 3.608185510163383e-05,
-      "loss": 0.2876,
+      "epoch": 1.9120986052546223,
+      "grad_norm": 0.22599272429943085,
+      "learning_rate": 3.541789349459327e-05,
+      "loss": 0.4059,
       "step": 53055
     },
     {
-      "epoch": 1.87,
-      "learning_rate": 3.60793014756022e-05,
-      "loss": 0.2843,
+      "epoch": 1.9122788049158468,
+      "grad_norm": 0.16319070756435394,
+      "learning_rate": 3.541524073261023e-05,
+      "loss": 0.3643,
       "step": 53060
     },
     {
-      "epoch": 1.87,
-      "learning_rate": 3.6076747705713145e-05,
-      "loss": 0.2778,
+      "epoch": 1.9124590045770713,
+      "grad_norm": 0.20491214096546173,
+      "learning_rate": 3.541258782872154e-05,
+      "loss": 0.412,
       "step": 53065
     },
     {
-      "epoch": 1.87,
-      "learning_rate": 3.607419379199982e-05,
-      "loss": 0.3115,
+      "epoch": 1.912639204238296,
+      "grad_norm": 0.14605240523815155,
+      "learning_rate": 3.540993478296337e-05,
+      "loss": 0.3939,
       "step": 53070
     },
     {
-      "epoch": 1.87,
-      "learning_rate": 3.6071639734495386e-05,
-      "loss": 0.28,
+      "epoch": 1.9128194038995208,
+      "grad_norm": 0.18107111752033234,
+      "learning_rate": 3.540728159537185e-05,
+      "loss": 0.4318,
       "step": 53075
     },
     {
-      "epoch": 1.87,
-      "learning_rate": 3.606908553323301e-05,
-      "loss": 0.2613,
+      "epoch": 1.9129996035607453,
+      "grad_norm": 0.16898123919963837,
+      "learning_rate": 3.540462826598313e-05,
+      "loss": 0.4414,
       "step": 53080
     },
     {
-      "epoch": 1.87,
-      "learning_rate": 3.606653118824586e-05,
-      "loss": 0.2805,
+      "epoch": 1.9131798032219698,
+      "grad_norm": 0.1803431212902069,
+      "learning_rate": 3.540197479483337e-05,
+      "loss": 0.4014,
       "step": 53085
     },
     {
-      "epoch": 1.87,
-      "learning_rate": 3.606397669956707e-05,
-      "loss": 0.2614,
+      "epoch": 1.9133600028831945,
+      "grad_norm": 0.17735476791858673,
+      "learning_rate": 3.539932118195871e-05,
+      "loss": 0.3945,
       "step": 53090
     },
     {
-      "epoch": 1.87,
-      "learning_rate": 3.606142206722985e-05,
-      "loss": 0.293,
+      "epoch": 1.9135402025444193,
+      "grad_norm": 0.1508413851261139,
+      "learning_rate": 3.539666742739532e-05,
+      "loss": 0.4031,
       "step": 53095
     },
     {
-      "epoch": 1.87,
-      "learning_rate": 3.6058867291267354e-05,
-      "loss": 0.2816,
+      "epoch": 1.913720402205644,
+      "grad_norm": 0.21895846724510193,
+      "learning_rate": 3.539401353117935e-05,
+      "loss": 0.4405,
       "step": 53100
     },
     {
-      "epoch": 1.87,
-      "learning_rate": 3.605631237171277e-05,
-      "loss": 0.287,
+      "epoch": 1.9139006018668685,
+      "grad_norm": 0.22758044302463531,
+      "learning_rate": 3.539135949334695e-05,
+      "loss": 0.4415,
       "step": 53105
     },
     {
-      "epoch": 1.87,
-      "learning_rate": 3.6053757308599235e-05,
-      "loss": 0.2713,
+      "epoch": 1.914080801528093,
+      "grad_norm": 0.17309550940990448,
+      "learning_rate": 3.538870531393429e-05,
+      "loss": 0.4228,
       "step": 53110
     },
     {
-      "epoch": 1.87,
-      "learning_rate": 3.6051202101959955e-05,
-      "loss": 0.3059,
+      "epoch": 1.9142610011893177,
+      "grad_norm": 0.2273998111486435,
+      "learning_rate": 3.538605099297754e-05,
+      "loss": 0.4158,
       "step": 53115
     },
     {
-      "epoch": 1.87,
-      "learning_rate": 3.604864675182809e-05,
-      "loss": 0.2723,
+      "epoch": 1.9144412008505425,
+      "grad_norm": 0.17294365167617798,
+      "learning_rate": 3.538339653051285e-05,
+      "loss": 0.3855,
       "step": 53120
     },
     {
-      "epoch": 1.87,
-      "learning_rate": 3.6046091258236834e-05,
-      "loss": 0.2759,
+      "epoch": 1.914621400511767,
+      "grad_norm": 0.2204235941171646,
+      "learning_rate": 3.5380741926576385e-05,
+      "loss": 0.4159,
       "step": 53125
     },
     {
-      "epoch": 1.87,
-      "learning_rate": 3.604353562121936e-05,
-      "loss": 0.2998,
+      "epoch": 1.9148016001729917,
+      "grad_norm": 0.18429414927959442,
+      "learning_rate": 3.537808718120433e-05,
+      "loss": 0.4255,
       "step": 53130
     },
     {
-      "epoch": 1.87,
-      "learning_rate": 3.6040979840808866e-05,
-      "loss": 0.2741,
+      "epoch": 1.9149817998342162,
+      "grad_norm": 0.18683667480945587,
+      "learning_rate": 3.537543229443285e-05,
+      "loss": 0.4042,
       "step": 53135
     },
     {
-      "epoch": 1.87,
-      "learning_rate": 3.603842391703851e-05,
-      "loss": 0.2959,
+      "epoch": 1.915161999495441,
+      "grad_norm": 0.20948028564453125,
+      "learning_rate": 3.537330828323237e-05,
+      "loss": 0.3947,
       "step": 53140
     },
     {
-      "epoch": 1.87,
-      "learning_rate": 3.60358678499415e-05,
-      "loss": 0.2795,
+      "epoch": 1.9153421991566657,
+      "grad_norm": 0.17518873512744904,
+      "learning_rate": 3.5370653142033075e-05,
+      "loss": 0.4003,
       "step": 53145
     },
     {
-      "epoch": 1.87,
-      "learning_rate": 3.603331163955102e-05,
-      "loss": 0.2791,
+      "epoch": 1.9155223988178902,
+      "grad_norm": 0.13344982266426086,
+      "learning_rate": 3.536799785953563e-05,
+      "loss": 0.4035,
       "step": 53150
     },
     {
-      "epoch": 1.87,
-      "learning_rate": 3.6030755285900255e-05,
-      "loss": 0.2649,
+      "epoch": 1.9157025984791147,
+      "grad_norm": 0.1912095695734024,
+      "learning_rate": 3.5365342435776225e-05,
+      "loss": 0.4021,
       "step": 53155
     },
     {
-      "epoch": 1.87,
-      "learning_rate": 3.602819878902241e-05,
-      "loss": 0.291,
+      "epoch": 1.9158827981403395,
+      "grad_norm": 0.19049333035945892,
+      "learning_rate": 3.536268687079104e-05,
+      "loss": 0.3966,
       "step": 53160
     },
     {
-      "epoch": 1.87,
-      "learning_rate": 3.602564214895067e-05,
-      "loss": 0.2913,
+      "epoch": 1.9160629978015642,
+      "grad_norm": 0.14792177081108093,
+      "learning_rate": 3.5360031164616244e-05,
+      "loss": 0.4132,
       "step": 53165
     },
     {
-      "epoch": 1.87,
-      "learning_rate": 3.602308536571822e-05,
-      "loss": 0.3051,
+      "epoch": 1.916243197462789,
+      "grad_norm": 0.1720864176750183,
+      "learning_rate": 3.535737531728803e-05,
+      "loss": 0.4254,
       "step": 53170
     },
     {
-      "epoch": 1.87,
-      "learning_rate": 3.602052843935828e-05,
-      "loss": 0.2615,
+      "epoch": 1.9164233971240134,
+      "grad_norm": 0.2024461328983307,
+      "learning_rate": 3.5354719328842585e-05,
+      "loss": 0.4462,
       "step": 53175
     },
     {
-      "epoch": 1.87,
-      "learning_rate": 3.601797136990404e-05,
-      "loss": 0.2853,
+      "epoch": 1.916603596785238,
+      "grad_norm": 0.19676236808300018,
+      "learning_rate": 3.53520631993161e-05,
+      "loss": 0.4637,
       "step": 53180
     },
     {
-      "epoch": 1.87,
-      "learning_rate": 3.601541415738871e-05,
-      "loss": 0.3029,
+      "epoch": 1.9167837964464627,
+      "grad_norm": 0.19073833525180817,
+      "learning_rate": 3.534940692874475e-05,
+      "loss": 0.4156,
       "step": 53185
     },
     {
-      "epoch": 1.87,
-      "learning_rate": 3.6012856801845476e-05,
-      "loss": 0.2798,
+      "epoch": 1.9169639961076874,
+      "grad_norm": 0.18771030008792877,
+      "learning_rate": 3.5346750517164736e-05,
+      "loss": 0.3749,
       "step": 53190
     },
     {
-      "epoch": 1.87,
-      "learning_rate": 3.6010299303307564e-05,
-      "loss": 0.2792,
+      "epoch": 1.917144195768912,
+      "grad_norm": 0.20612174272537231,
+      "learning_rate": 3.534409396461225e-05,
+      "loss": 0.4305,
       "step": 53195
     },
     {
-      "epoch": 1.87,
-      "learning_rate": 3.6007741661808156e-05,
-      "loss": 0.2701,
+      "epoch": 1.9173243954301364,
+      "grad_norm": 0.18494312465190887,
+      "learning_rate": 3.534143727112349e-05,
+      "loss": 0.4385,
       "step": 53200
     },
     {
-      "epoch": 1.87,
-      "learning_rate": 3.60051838773805e-05,
-      "loss": 0.2634,
+      "epoch": 1.9175045950913612,
+      "grad_norm": 0.22901670634746552,
+      "learning_rate": 3.533878043673464e-05,
+      "loss": 0.4131,
       "step": 53205
     },
     {
-      "epoch": 1.87,
-      "learning_rate": 3.600262595005777e-05,
-      "loss": 0.274,
+      "epoch": 1.917684794752586,
+      "grad_norm": 0.18775056302547455,
+      "learning_rate": 3.533612346148192e-05,
+      "loss": 0.4204,
       "step": 53210
     },
     {
-      "epoch": 1.87,
-      "learning_rate": 3.600006787987319e-05,
-      "loss": 0.276,
+      "epoch": 1.9178649944138106,
+      "grad_norm": 0.2237263172864914,
+      "learning_rate": 3.533346634540151e-05,
+      "loss": 0.4032,
       "step": 53215
     },
     {
-      "epoch": 1.87,
-      "learning_rate": 3.5997509666859986e-05,
-      "loss": 0.254,
+      "epoch": 1.9180451940750352,
+      "grad_norm": 0.24082402884960175,
+      "learning_rate": 3.533080908852962e-05,
+      "loss": 0.3854,
       "step": 53220
     },
     {
-      "epoch": 1.87,
-      "learning_rate": 3.599495131105136e-05,
-      "loss": 0.2867,
+      "epoch": 1.9182253937362597,
+      "grad_norm": 0.18689344823360443,
+      "learning_rate": 3.5328151690902465e-05,
+      "loss": 0.3829,
       "step": 53225
     },
     {
-      "epoch": 1.87,
-      "learning_rate": 3.599239281248054e-05,
-      "loss": 0.2812,
+      "epoch": 1.9184055933974844,
+      "grad_norm": 0.20460011065006256,
+      "learning_rate": 3.532549415255624e-05,
+      "loss": 0.4378,
       "step": 53230
     },
     {
-      "epoch": 1.87,
-      "learning_rate": 3.598983417118075e-05,
-      "loss": 0.2513,
+      "epoch": 1.9185857930587091,
+      "grad_norm": 0.18038150668144226,
+      "learning_rate": 3.5322836473527154e-05,
+      "loss": 0.4306,
       "step": 53235
     },
     {
-      "epoch": 1.87,
-      "learning_rate": 3.5987275387185196e-05,
-      "loss": 0.3058,
+      "epoch": 1.9187659927199336,
+      "grad_norm": 0.18723049759864807,
+      "learning_rate": 3.5320178653851425e-05,
+      "loss": 0.384,
       "step": 53240
     },
     {
-      "epoch": 1.87,
-      "learning_rate": 3.5984716460527116e-05,
-      "loss": 0.2823,
+      "epoch": 1.9189461923811584,
+      "grad_norm": 0.21049338579177856,
+      "learning_rate": 3.531752069356525e-05,
+      "loss": 0.4149,
       "step": 53245
     },
     {
-      "epoch": 1.87,
-      "learning_rate": 3.598215739123973e-05,
-      "loss": 0.2615,
+      "epoch": 1.919126392042383,
+      "grad_norm": 0.17135761678218842,
+      "learning_rate": 3.531486259270486e-05,
+      "loss": 0.4073,
       "step": 53250
     },
     {
-      "epoch": 1.87,
-      "learning_rate": 3.597959817935628e-05,
-      "loss": 0.2778,
+      "epoch": 1.9193065917036076,
+      "grad_norm": 0.1982400119304657,
+      "learning_rate": 3.531220435130646e-05,
+      "loss": 0.3981,
       "step": 53255
     },
     {
-      "epoch": 1.87,
-      "learning_rate": 3.5977038824909966e-05,
-      "loss": 0.272,
+      "epoch": 1.9194867913648324,
+      "grad_norm": 0.18633022904396057,
+      "learning_rate": 3.530954596940628e-05,
+      "loss": 0.3868,
       "step": 53260
     },
     {
-      "epoch": 1.87,
-      "learning_rate": 3.597447932793404e-05,
-      "loss": 0.2803,
+      "epoch": 1.9196669910260569,
+      "grad_norm": 0.20298829674720764,
+      "learning_rate": 3.530688744704053e-05,
+      "loss": 0.4442,
       "step": 53265
     },
     {
-      "epoch": 1.87,
-      "learning_rate": 3.5971919688461744e-05,
-      "loss": 0.2661,
+      "epoch": 1.9198471906872814,
+      "grad_norm": 0.17261956632137299,
+      "learning_rate": 3.530422878424543e-05,
+      "loss": 0.3916,
       "step": 53270
     },
     {
-      "epoch": 1.87,
-      "learning_rate": 3.596935990652629e-05,
-      "loss": 0.2634,
+      "epoch": 1.9200273903485061,
+      "grad_norm": 0.2012072056531906,
+      "learning_rate": 3.5301569981057205e-05,
+      "loss": 0.4336,
       "step": 53275
     },
     {
-      "epoch": 1.87,
-      "learning_rate": 3.596679998216093e-05,
-      "loss": 0.2915,
+      "epoch": 1.9202075900097308,
+      "grad_norm": 0.17168846726417542,
+      "learning_rate": 3.5298911037512086e-05,
+      "loss": 0.3852,
       "step": 53280
     },
     {
-      "epoch": 1.87,
-      "learning_rate": 3.59642399153989e-05,
-      "loss": 0.2551,
+      "epoch": 1.9203877896709556,
+      "grad_norm": 0.21444140374660492,
+      "learning_rate": 3.52962519536463e-05,
+      "loss": 0.4031,
       "step": 53285
     },
     {
-      "epoch": 1.87,
-      "learning_rate": 3.5961679706273447e-05,
-      "loss": 0.2735,
+      "epoch": 1.92056798933218,
+      "grad_norm": 0.206976979970932,
+      "learning_rate": 3.5293592729496076e-05,
+      "loss": 0.3876,
       "step": 53290
     },
     {
-      "epoch": 1.88,
-      "learning_rate": 3.59591193548178e-05,
-      "loss": 0.2879,
+      "epoch": 1.9207481889934046,
+      "grad_norm": 0.17099922895431519,
+      "learning_rate": 3.529093336509764e-05,
+      "loss": 0.4347,
       "step": 53295
     },
     {
-      "epoch": 1.88,
-      "learning_rate": 3.595655886106521e-05,
-      "loss": 0.3059,
+      "epoch": 1.9209283886546293,
+      "grad_norm": 0.19171853363513947,
+      "learning_rate": 3.528827386048723e-05,
+      "loss": 0.4066,
       "step": 53300
     },
     {
-      "epoch": 1.88,
-      "learning_rate": 3.5953998225048925e-05,
-      "loss": 0.2938,
+      "epoch": 1.921108588315854,
+      "grad_norm": 0.24429461359977722,
+      "learning_rate": 3.528561421570108e-05,
+      "loss": 0.4134,
       "step": 53305
     },
     {
-      "epoch": 1.88,
-      "learning_rate": 3.595143744680219e-05,
-      "loss": 0.2826,
+      "epoch": 1.9212887879770786,
+      "grad_norm": 0.1823347806930542,
+      "learning_rate": 3.528295443077543e-05,
+      "loss": 0.4363,
       "step": 53310
     },
     {
-      "epoch": 1.88,
-      "learning_rate": 3.594887652635826e-05,
-      "loss": 0.2837,
+      "epoch": 1.921468987638303,
+      "grad_norm": 0.22016625106334686,
+      "learning_rate": 3.528029450574651e-05,
+      "loss": 0.4154,
       "step": 53315
     },
     {
-      "epoch": 1.88,
-      "learning_rate": 3.594631546375038e-05,
-      "loss": 0.2787,
+      "epoch": 1.9216491872995278,
+      "grad_norm": 0.18516206741333008,
+      "learning_rate": 3.5277634440650584e-05,
+      "loss": 0.4497,
       "step": 53320
     },
     {
-      "epoch": 1.88,
-      "learning_rate": 3.5943754259011813e-05,
-      "loss": 0.2896,
+      "epoch": 1.9218293869607526,
+      "grad_norm": 0.18409641087055206,
+      "learning_rate": 3.527497423552386e-05,
+      "loss": 0.4074,
       "step": 53325
     },
     {
-      "epoch": 1.88,
-      "learning_rate": 3.594119291217581e-05,
-      "loss": 0.2584,
+      "epoch": 1.9220095866219773,
+      "grad_norm": 0.18460574746131897,
+      "learning_rate": 3.527231389040262e-05,
+      "loss": 0.385,
       "step": 53330
     },
     {
-      "epoch": 1.88,
-      "learning_rate": 3.5938631423275625e-05,
-      "loss": 0.2586,
+      "epoch": 1.9221897862832018,
+      "grad_norm": 0.18730106949806213,
+      "learning_rate": 3.526965340532308e-05,
+      "loss": 0.4215,
       "step": 53335
     },
     {
-      "epoch": 1.88,
-      "learning_rate": 3.5936069792344514e-05,
-      "loss": 0.2713,
+      "epoch": 1.9223699859444263,
+      "grad_norm": 0.20932596921920776,
+      "learning_rate": 3.52669927803215e-05,
+      "loss": 0.38,
       "step": 53340
     },
     {
-      "epoch": 1.88,
-      "learning_rate": 3.593350801941575e-05,
-      "loss": 0.286,
+      "epoch": 1.922550185605651,
+      "grad_norm": 0.20945709943771362,
+      "learning_rate": 3.5264332015434134e-05,
+      "loss": 0.386,
       "step": 53345
     },
     {
-      "epoch": 1.88,
-      "learning_rate": 3.593094610452259e-05,
-      "loss": 0.2882,
+      "epoch": 1.9227303852668758,
+      "grad_norm": 0.19625355303287506,
+      "learning_rate": 3.5261671110697234e-05,
+      "loss": 0.4177,
       "step": 53350
     },
     {
-      "epoch": 1.88,
-      "learning_rate": 3.59283840476983e-05,
-      "loss": 0.2778,
+      "epoch": 1.9229105849281003,
+      "grad_norm": 0.16801773011684418,
+      "learning_rate": 3.525901006614705e-05,
+      "loss": 0.4078,
       "step": 53355
     },
     {
-      "epoch": 1.88,
-      "learning_rate": 3.592582184897614e-05,
-      "loss": 0.2799,
+      "epoch": 1.9230907845893248,
+      "grad_norm": 0.23051504790782928,
+      "learning_rate": 3.525634888181983e-05,
+      "loss": 0.4157,
       "step": 53360
     },
     {
-      "epoch": 1.88,
-      "learning_rate": 3.592325950838938e-05,
-      "loss": 0.3066,
+      "epoch": 1.9232709842505495,
+      "grad_norm": 0.17896528542041779,
+      "learning_rate": 3.525368755775186e-05,
+      "loss": 0.4013,
       "step": 53365
     },
     {
-      "epoch": 1.88,
-      "learning_rate": 3.5920697025971306e-05,
-      "loss": 0.2768,
+      "epoch": 1.9234511839117743,
+      "grad_norm": 0.24281032383441925,
+      "learning_rate": 3.525102609397937e-05,
+      "loss": 0.4071,
       "step": 53370
     },
     {
-      "epoch": 1.88,
-      "learning_rate": 3.591813440175518e-05,
-      "loss": 0.2807,
+      "epoch": 1.923631383572999,
+      "grad_norm": 0.19194422662258148,
+      "learning_rate": 3.524836449053864e-05,
+      "loss": 0.4225,
       "step": 53375
     },
     {
-      "epoch": 1.88,
-      "learning_rate": 3.5915571635774256e-05,
-      "loss": 0.2799,
+      "epoch": 1.9238115832342235,
+      "grad_norm": 0.18676306307315826,
+      "learning_rate": 3.5245702747465924e-05,
+      "loss": 0.4121,
       "step": 53380
     },
     {
-      "epoch": 1.88,
-      "learning_rate": 3.5913008728061845e-05,
-      "loss": 0.2797,
+      "epoch": 1.923991782895448,
+      "grad_norm": 0.1551360785961151,
+      "learning_rate": 3.524304086479749e-05,
+      "loss": 0.4243,
       "step": 53385
     },
     {
-      "epoch": 1.88,
-      "learning_rate": 3.59104456786512e-05,
-      "loss": 0.2818,
+      "epoch": 1.9241719825566728,
+      "grad_norm": 0.1882885843515396,
+      "learning_rate": 3.5240378842569614e-05,
+      "loss": 0.3898,
       "step": 53390
     },
     {
-      "epoch": 1.88,
-      "learning_rate": 3.59078824875756e-05,
-      "loss": 0.2837,
+      "epoch": 1.9243521822178975,
+      "grad_norm": 0.22999207675457,
+      "learning_rate": 3.5237716680818554e-05,
+      "loss": 0.4386,
       "step": 53395
     },
     {
-      "epoch": 1.88,
-      "learning_rate": 3.590531915486835e-05,
-      "loss": 0.2849,
+      "epoch": 1.9245323818791222,
+      "grad_norm": 0.18411701917648315,
+      "learning_rate": 3.523505437958059e-05,
+      "loss": 0.3995,
       "step": 53400
     },
     {
-      "epoch": 1.88,
-      "learning_rate": 3.590275568056272e-05,
-      "loss": 0.2767,
+      "epoch": 1.9247125815403467,
+      "grad_norm": 0.1591692715883255,
+      "learning_rate": 3.5232391938891983e-05,
+      "loss": 0.3968,
       "step": 53405
     },
     {
-      "epoch": 1.88,
-      "learning_rate": 3.5900192064691976e-05,
-      "loss": 0.2593,
+      "epoch": 1.9248927812015713,
+      "grad_norm": 0.1875472217798233,
+      "learning_rate": 3.5229729358789026e-05,
+      "loss": 0.4125,
       "step": 53410
     },
     {
-      "epoch": 1.88,
-      "learning_rate": 3.5897628307289435e-05,
-      "loss": 0.2942,
+      "epoch": 1.925072980862796,
+      "grad_norm": 0.19325365126132965,
+      "learning_rate": 3.522706663930799e-05,
+      "loss": 0.404,
       "step": 53415
     },
     {
-      "epoch": 1.88,
-      "learning_rate": 3.589506440838836e-05,
-      "loss": 0.2855,
+      "epoch": 1.9252531805240207,
+      "grad_norm": 0.17323115468025208,
+      "learning_rate": 3.5224403780485136e-05,
+      "loss": 0.4047,
       "step": 53420
     },
     {
-      "epoch": 1.88,
-      "learning_rate": 3.589250036802206e-05,
-      "loss": 0.2643,
+      "epoch": 1.9254333801852452,
+      "grad_norm": 0.15711285173892975,
+      "learning_rate": 3.522174078235677e-05,
+      "loss": 0.3875,
       "step": 53425
     },
     {
-      "epoch": 1.88,
-      "learning_rate": 3.588993618622383e-05,
-      "loss": 0.3168,
+      "epoch": 1.9256135798464697,
+      "grad_norm": 0.1933702826499939,
+      "learning_rate": 3.521907764495917e-05,
+      "loss": 0.4531,
       "step": 53430
     },
     {
-      "epoch": 1.88,
-      "learning_rate": 3.588737186302695e-05,
-      "loss": 0.2947,
+      "epoch": 1.9257937795076945,
+      "grad_norm": 0.20678189396858215,
+      "learning_rate": 3.5216414368328607e-05,
+      "loss": 0.4352,
       "step": 53435
     },
     {
-      "epoch": 1.88,
-      "learning_rate": 3.5884807398464713e-05,
-      "loss": 0.271,
+      "epoch": 1.9259739791689192,
+      "grad_norm": 0.17905567586421967,
+      "learning_rate": 3.521375095250138e-05,
+      "loss": 0.4394,
       "step": 53440
     },
     {
-      "epoch": 1.88,
-      "learning_rate": 3.588224279257043e-05,
-      "loss": 0.2879,
+      "epoch": 1.926154178830144,
+      "grad_norm": 0.19163614511489868,
+      "learning_rate": 3.521108739751377e-05,
+      "loss": 0.3998,
       "step": 53445
     },
     {
-      "epoch": 1.88,
-      "learning_rate": 3.58796780453774e-05,
-      "loss": 0.2801,
+      "epoch": 1.9263343784913685,
+      "grad_norm": 0.1524534672498703,
+      "learning_rate": 3.5208423703402075e-05,
+      "loss": 0.3911,
       "step": 53450
     },
     {
-      "epoch": 1.88,
-      "learning_rate": 3.587711315691891e-05,
-      "loss": 0.288,
+      "epoch": 1.926514578152593,
+      "grad_norm": 0.15856072306632996,
+      "learning_rate": 3.520575987020258e-05,
+      "loss": 0.4365,
       "step": 53455
     },
     {
-      "epoch": 1.88,
-      "learning_rate": 3.587454812722829e-05,
-      "loss": 0.2895,
+      "epoch": 1.9266947778138177,
+      "grad_norm": 0.20642130076885223,
+      "learning_rate": 3.520309589795159e-05,
+      "loss": 0.4418,
       "step": 53460
     },
     {
-      "epoch": 1.88,
-      "learning_rate": 3.587198295633882e-05,
-      "loss": 0.2864,
+      "epoch": 1.9268749774750424,
+      "grad_norm": 0.1950474977493286,
+      "learning_rate": 3.520043178668538e-05,
+      "loss": 0.4325,
       "step": 53465
     },
     {
-      "epoch": 1.88,
-      "learning_rate": 3.586941764428381e-05,
-      "loss": 0.2726,
+      "epoch": 1.927055177136267,
+      "grad_norm": 0.1779826283454895,
+      "learning_rate": 3.519776753644028e-05,
+      "loss": 0.4068,
       "step": 53470
     },
     {
-      "epoch": 1.88,
-      "learning_rate": 3.586685219109659e-05,
-      "loss": 0.2633,
+      "epoch": 1.9272353767974915,
+      "grad_norm": 0.19303326308727264,
+      "learning_rate": 3.5195103147252564e-05,
+      "loss": 0.4287,
       "step": 53475
     },
     {
-      "epoch": 1.88,
-      "learning_rate": 3.586428659681044e-05,
-      "loss": 0.274,
+      "epoch": 1.9274155764587162,
+      "grad_norm": 0.20207709074020386,
+      "learning_rate": 3.5192438619158536e-05,
+      "loss": 0.4266,
       "step": 53480
     },
     {
-      "epoch": 1.88,
-      "learning_rate": 3.5861720861458695e-05,
-      "loss": 0.2746,
+      "epoch": 1.927595776119941,
+      "grad_norm": 0.16578246653079987,
+      "learning_rate": 3.5189773952194506e-05,
+      "loss": 0.377,
       "step": 53485
     },
     {
-      "epoch": 1.88,
-      "learning_rate": 3.585915498507467e-05,
-      "loss": 0.2978,
+      "epoch": 1.9277759757811657,
+      "grad_norm": 0.17136074602603912,
+      "learning_rate": 3.518710914639678e-05,
+      "loss": 0.3963,
       "step": 53490
     },
     {
-      "epoch": 1.88,
-      "learning_rate": 3.5856588967691664e-05,
-      "loss": 0.258,
+      "epoch": 1.9279561754423902,
+      "grad_norm": 0.1691044121980667,
+      "learning_rate": 3.518444420180167e-05,
+      "loss": 0.4504,
       "step": 53495
     },
     {
-      "epoch": 1.88,
-      "learning_rate": 3.585402280934301e-05,
-      "loss": 0.3014,
+      "epoch": 1.9281363751036147,
+      "grad_norm": 0.1666347235441208,
+      "learning_rate": 3.518177911844547e-05,
+      "loss": 0.4177,
       "step": 53500
     },
     {
-      "epoch": 1.88,
-      "eval_loss": 0.27490025758743286,
-      "eval_runtime": 10.529,
-      "eval_samples_per_second": 9.498,
-      "eval_steps_per_second": 9.498,
+      "epoch": 1.9281363751036147,
+      "eval_loss": 0.43853840231895447,
+      "eval_runtime": 3.5305,
+      "eval_samples_per_second": 28.325,
+      "eval_steps_per_second": 7.081,
       "step": 53500
     },
     {
-      "epoch": 1.88,
-      "learning_rate": 3.585145651006202e-05,
-      "loss": 0.2879,
+      "epoch": 1.9283165747648394,
+      "grad_norm": 0.18548129498958588,
+      "learning_rate": 3.5179113896364504e-05,
+      "loss": 0.4455,
       "step": 53505
     },
     {
-      "epoch": 1.88,
-      "learning_rate": 3.584889006988202e-05,
-      "loss": 0.2526,
+      "epoch": 1.9284967744260642,
+      "grad_norm": 0.16688363254070282,
+      "learning_rate": 3.517644853559509e-05,
+      "loss": 0.3915,
       "step": 53510
     },
     {
-      "epoch": 1.88,
-      "learning_rate": 3.584632348883633e-05,
-      "loss": 0.2781,
+      "epoch": 1.9286769740872887,
+      "grad_norm": 0.14306481182575226,
+      "learning_rate": 3.517378303617353e-05,
+      "loss": 0.425,
       "step": 53515
     },
     {
-      "epoch": 1.88,
-      "learning_rate": 3.5843756766958285e-05,
-      "loss": 0.3115,
+      "epoch": 1.9288571737485134,
+      "grad_norm": 0.13732267916202545,
+      "learning_rate": 3.517111739813615e-05,
+      "loss": 0.4324,
       "step": 53520
     },
     {
-      "epoch": 1.88,
-      "learning_rate": 3.58411899042812e-05,
-      "loss": 0.2705,
+      "epoch": 1.929037373409738,
+      "grad_norm": 0.1663437783718109,
+      "learning_rate": 3.516845162151925e-05,
+      "loss": 0.4314,
       "step": 53525
     },
     {
-      "epoch": 1.88,
-      "learning_rate": 3.583862290083841e-05,
-      "loss": 0.2759,
+      "epoch": 1.9292175730709626,
+      "grad_norm": 0.18951812386512756,
+      "learning_rate": 3.516578570635917e-05,
+      "loss": 0.4099,
       "step": 53530
     },
     {
-      "epoch": 1.88,
-      "learning_rate": 3.583605575666324e-05,
-      "loss": 0.3126,
+      "epoch": 1.9293977727321874,
+      "grad_norm": 0.17902812361717224,
+      "learning_rate": 3.5163119652692236e-05,
+      "loss": 0.4343,
       "step": 53535
     },
     {
-      "epoch": 1.88,
-      "learning_rate": 3.583348847178903e-05,
-      "loss": 0.3062,
+      "epoch": 1.9295779723934119,
+      "grad_norm": 0.1834252029657364,
+      "learning_rate": 3.5160453460554766e-05,
+      "loss": 0.3859,
       "step": 53540
     },
     {
-      "epoch": 1.88,
-      "learning_rate": 3.583092104624911e-05,
-      "loss": 0.28,
+      "epoch": 1.9297581720546364,
+      "grad_norm": 0.21351586282253265,
+      "learning_rate": 3.515778712998307e-05,
+      "loss": 0.4282,
       "step": 53545
     },
     {
-      "epoch": 1.88,
-      "learning_rate": 3.582835348007682e-05,
-      "loss": 0.3002,
+      "epoch": 1.9299383717158611,
+      "grad_norm": 0.18002468347549438,
+      "learning_rate": 3.515512066101351e-05,
+      "loss": 0.4183,
       "step": 53550
     },
     {
-      "epoch": 1.88,
-      "learning_rate": 3.5825785773305496e-05,
-      "loss": 0.2674,
+      "epoch": 1.9301185713770859,
+      "grad_norm": 0.18808430433273315,
+      "learning_rate": 3.515245405368238e-05,
+      "loss": 0.3664,
       "step": 53555
     },
     {
-      "epoch": 1.88,
-      "learning_rate": 3.582373150667954e-05,
-      "loss": 0.2849,
+      "epoch": 1.9302987710383106,
+      "grad_norm": 0.17705509066581726,
+      "learning_rate": 3.5149787308026036e-05,
+      "loss": 0.3765,
       "step": 53560
     },
     {
-      "epoch": 1.88,
-      "learning_rate": 3.582116354691397e-05,
-      "loss": 0.2912,
+      "epoch": 1.9304789706995351,
+      "grad_norm": 0.17073272168636322,
+      "learning_rate": 3.514712042408081e-05,
+      "loss": 0.4367,
       "step": 53565
     },
     {
-      "epoch": 1.88,
-      "learning_rate": 3.5818595446642725e-05,
-      "loss": 0.2874,
+      "epoch": 1.9306591703607596,
+      "grad_norm": 0.19067902863025665,
+      "learning_rate": 3.514445340188303e-05,
+      "loss": 0.4082,
       "step": 53570
     },
     {
-      "epoch": 1.88,
-      "learning_rate": 3.581602720589915e-05,
-      "loss": 0.2762,
+      "epoch": 1.9308393700219844,
+      "grad_norm": 0.19914193451404572,
+      "learning_rate": 3.5141786241469033e-05,
+      "loss": 0.4077,
       "step": 53575
     },
     {
-      "epoch": 1.89,
-      "learning_rate": 3.5813458824716594e-05,
-      "loss": 0.296,
+      "epoch": 1.931019569683209,
+      "grad_norm": 0.15693803131580353,
+      "learning_rate": 3.513911894287516e-05,
+      "loss": 0.3914,
       "step": 53580
     },
     {
-      "epoch": 1.89,
-      "learning_rate": 3.58108903031284e-05,
-      "loss": 0.2853,
+      "epoch": 1.9311997693444336,
+      "grad_norm": 0.18516100943088531,
+      "learning_rate": 3.5136451506137766e-05,
+      "loss": 0.4613,
       "step": 53585
     },
     {
-      "epoch": 1.89,
-      "learning_rate": 3.580832164116791e-05,
-      "loss": 0.2781,
+      "epoch": 1.9313799690056581,
+      "grad_norm": 0.18044835329055786,
+      "learning_rate": 3.513378393129317e-05,
+      "loss": 0.4471,
       "step": 53590
     },
     {
-      "epoch": 1.89,
-      "learning_rate": 3.580575283886849e-05,
-      "loss": 0.2912,
+      "epoch": 1.9315601686668828,
+      "grad_norm": 0.19149909913539886,
+      "learning_rate": 3.5131116218377735e-05,
+      "loss": 0.4116,
       "step": 53595
     },
     {
-      "epoch": 1.89,
-      "learning_rate": 3.5803183896263495e-05,
-      "loss": 0.2779,
+      "epoch": 1.9317403683281076,
+      "grad_norm": 0.1986377239227295,
+      "learning_rate": 3.51284483674278e-05,
+      "loss": 0.4369,
       "step": 53600
     },
     {
-      "epoch": 1.89,
-      "learning_rate": 3.5800614813386273e-05,
-      "loss": 0.2694,
+      "epoch": 1.9319205679893323,
+      "grad_norm": 0.16995567083358765,
+      "learning_rate": 3.5125780378479725e-05,
+      "loss": 0.4208,
       "step": 53605
     },
     {
-      "epoch": 1.89,
-      "learning_rate": 3.579804559027019e-05,
-      "loss": 0.2939,
+      "epoch": 1.9321007676505568,
+      "grad_norm": 0.16595984995365143,
+      "learning_rate": 3.5123112251569844e-05,
+      "loss": 0.3805,
       "step": 53610
     },
     {
-      "epoch": 1.89,
-      "learning_rate": 3.57954762269486e-05,
-      "loss": 0.2649,
+      "epoch": 1.9322809673117813,
+      "grad_norm": 0.2101103663444519,
+      "learning_rate": 3.5120443986734526e-05,
+      "loss": 0.3964,
       "step": 53615
     },
     {
-      "epoch": 1.89,
-      "learning_rate": 3.579290672345486e-05,
-      "loss": 0.2454,
+      "epoch": 1.932461166973006,
+      "grad_norm": 0.15714338421821594,
+      "learning_rate": 3.511777558401012e-05,
+      "loss": 0.3843,
       "step": 53620
     },
     {
-      "epoch": 1.89,
-      "learning_rate": 3.579033707982234e-05,
-      "loss": 0.3101,
+      "epoch": 1.9326413666342308,
+      "grad_norm": 0.14750663936138153,
+      "learning_rate": 3.511510704343297e-05,
+      "loss": 0.4092,
       "step": 53625
     },
     {
-      "epoch": 1.89,
-      "learning_rate": 3.578776729608441e-05,
-      "loss": 0.2627,
+      "epoch": 1.9328215662954553,
+      "grad_norm": 0.16553977131843567,
+      "learning_rate": 3.511243836503944e-05,
+      "loss": 0.419,
       "step": 53630
     },
     {
-      "epoch": 1.89,
-      "learning_rate": 3.578519737227443e-05,
-      "loss": 0.2664,
+      "epoch": 1.93300176595668,
+      "grad_norm": 0.23781326413154602,
+      "learning_rate": 3.5109769548865914e-05,
+      "loss": 0.3938,
       "step": 53635
     },
     {
-      "epoch": 1.89,
-      "learning_rate": 3.578262730842577e-05,
-      "loss": 0.2712,
+      "epoch": 1.9331819656179046,
+      "grad_norm": 0.19198031723499298,
+      "learning_rate": 3.510710059494871e-05,
+      "loss": 0.4057,
       "step": 53640
     },
     {
-      "epoch": 1.89,
-      "learning_rate": 3.5780057104571796e-05,
-      "loss": 0.3049,
+      "epoch": 1.9333621652791293,
+      "grad_norm": 0.1985791027545929,
+      "learning_rate": 3.5104431503324244e-05,
+      "loss": 0.4248,
       "step": 53645
     },
     {
-      "epoch": 1.89,
-      "learning_rate": 3.5777486760745887e-05,
-      "loss": 0.2892,
+      "epoch": 1.933542364940354,
+      "grad_norm": 0.18147054314613342,
+      "learning_rate": 3.510176227402884e-05,
+      "loss": 0.3811,
       "step": 53650
     },
     {
-      "epoch": 1.89,
-      "learning_rate": 3.577491627698142e-05,
-      "loss": 0.2862,
+      "epoch": 1.9337225646015785,
+      "grad_norm": 0.20852439105510712,
+      "learning_rate": 3.509909290709889e-05,
+      "loss": 0.4246,
       "step": 53655
     },
     {
-      "epoch": 1.89,
-      "learning_rate": 3.5772345653311755e-05,
-      "loss": 0.2894,
+      "epoch": 1.933902764262803,
+      "grad_norm": 0.19854693114757538,
+      "learning_rate": 3.5096423402570746e-05,
+      "loss": 0.4155,
       "step": 53660
     },
     {
-      "epoch": 1.89,
-      "learning_rate": 3.57697748897703e-05,
-      "loss": 0.2764,
+      "epoch": 1.9340829639240278,
+      "grad_norm": 0.15732385218143463,
+      "learning_rate": 3.5093753760480794e-05,
+      "loss": 0.4039,
       "step": 53665
     },
     {
-      "epoch": 1.89,
-      "learning_rate": 3.57672039863904e-05,
-      "loss": 0.2925,
+      "epoch": 1.9342631635852525,
+      "grad_norm": 0.17299677431583405,
+      "learning_rate": 3.509108398086539e-05,
+      "loss": 0.4189,
       "step": 53670
     },
     {
-      "epoch": 1.89,
-      "learning_rate": 3.576463294320545e-05,
-      "loss": 0.2872,
+      "epoch": 1.9344433632464773,
+      "grad_norm": 0.17880766093730927,
+      "learning_rate": 3.508841406376093e-05,
+      "loss": 0.4369,
       "step": 53675
     },
     {
-      "epoch": 1.89,
-      "learning_rate": 3.576206176024884e-05,
-      "loss": 0.2931,
+      "epoch": 1.9346235629077018,
+      "grad_norm": 0.23910261690616608,
+      "learning_rate": 3.508574400920379e-05,
+      "loss": 0.377,
       "step": 53680
     },
     {
-      "epoch": 1.89,
-      "learning_rate": 3.575949043755395e-05,
-      "loss": 0.2835,
+      "epoch": 1.9348037625689263,
+      "grad_norm": 0.165726900100708,
+      "learning_rate": 3.508307381723032e-05,
+      "loss": 0.4052,
       "step": 53685
     },
     {
-      "epoch": 1.89,
-      "learning_rate": 3.5756918975154174e-05,
-      "loss": 0.2891,
+      "epoch": 1.934983962230151,
+      "grad_norm": 0.18893593549728394,
+      "learning_rate": 3.508040348787694e-05,
+      "loss": 0.3806,
       "step": 53690
     },
     {
-      "epoch": 1.89,
-      "learning_rate": 3.575434737308289e-05,
-      "loss": 0.2525,
+      "epoch": 1.9351641618913757,
+      "grad_norm": 0.22258023917675018,
+      "learning_rate": 3.5077733021180006e-05,
+      "loss": 0.402,
       "step": 53695
     },
     {
-      "epoch": 1.89,
-      "learning_rate": 3.575177563137349e-05,
-      "loss": 0.3179,
+      "epoch": 1.9353443615526003,
+      "grad_norm": 0.17295654118061066,
+      "learning_rate": 3.5075062417175905e-05,
+      "loss": 0.4317,
       "step": 53700
     },
     {
-      "epoch": 1.89,
-      "learning_rate": 3.5749203750059365e-05,
-      "loss": 0.3078,
+      "epoch": 1.9355245612138248,
+      "grad_norm": 0.1404528021812439,
+      "learning_rate": 3.507239167590104e-05,
+      "loss": 0.4074,
       "step": 53705
     },
     {
-      "epoch": 1.89,
-      "learning_rate": 3.574663172917392e-05,
-      "loss": 0.2811,
+      "epoch": 1.9357047608750495,
+      "grad_norm": 0.16644835472106934,
+      "learning_rate": 3.5069720797391784e-05,
+      "loss": 0.4192,
       "step": 53710
     },
     {
-      "epoch": 1.89,
-      "learning_rate": 3.574405956875055e-05,
-      "loss": 0.2791,
+      "epoch": 1.9358849605362742,
+      "grad_norm": 0.21839553117752075,
+      "learning_rate": 3.506704978168453e-05,
+      "loss": 0.4101,
       "step": 53715
     },
     {
-      "epoch": 1.89,
-      "learning_rate": 3.574148726882264e-05,
-      "loss": 0.3034,
+      "epoch": 1.936065160197499,
+      "grad_norm": 0.21238647401332855,
+      "learning_rate": 3.506437862881567e-05,
+      "loss": 0.4314,
       "step": 53720
     },
     {
-      "epoch": 1.89,
-      "learning_rate": 3.5738914829423596e-05,
-      "loss": 0.2864,
+      "epoch": 1.9362453598587235,
+      "grad_norm": 0.19997538626194,
+      "learning_rate": 3.506170733882161e-05,
+      "loss": 0.371,
       "step": 53725
     },
     {
-      "epoch": 1.89,
-      "learning_rate": 3.573634225058682e-05,
-      "loss": 0.2926,
+      "epoch": 1.936425559519948,
+      "grad_norm": 0.18178488314151764,
+      "learning_rate": 3.505903591173872e-05,
+      "loss": 0.4023,
       "step": 53730
     },
     {
-      "epoch": 1.89,
-      "learning_rate": 3.5733769532345715e-05,
-      "loss": 0.2712,
+      "epoch": 1.9366057591811727,
+      "grad_norm": 0.161050945520401,
+      "learning_rate": 3.505636434760343e-05,
+      "loss": 0.4041,
       "step": 53735
     },
     {
-      "epoch": 1.89,
-      "learning_rate": 3.573119667473369e-05,
-      "loss": 0.2971,
+      "epoch": 1.9367859588423975,
+      "grad_norm": 0.22217293083667755,
+      "learning_rate": 3.505369264645211e-05,
+      "loss": 0.4249,
       "step": 53740
     },
     {
-      "epoch": 1.89,
-      "learning_rate": 3.572862367778415e-05,
-      "loss": 0.2853,
+      "epoch": 1.936966158503622,
+      "grad_norm": 0.1790328472852707,
+      "learning_rate": 3.505102080832118e-05,
+      "loss": 0.405,
       "step": 53745
     },
     {
-      "epoch": 1.89,
-      "learning_rate": 3.57260505415305e-05,
-      "loss": 0.2959,
+      "epoch": 1.9371463581648467,
+      "grad_norm": 0.17486228048801422,
+      "learning_rate": 3.504834883324704e-05,
+      "loss": 0.3947,
       "step": 53750
     },
     {
-      "epoch": 1.89,
-      "learning_rate": 3.5723477266006145e-05,
-      "loss": 0.2991,
+      "epoch": 1.9373265578260712,
+      "grad_norm": 0.1511772871017456,
+      "learning_rate": 3.504567672126608e-05,
+      "loss": 0.3633,
       "step": 53755
     },
     {
-      "epoch": 1.89,
-      "learning_rate": 3.572090385124451e-05,
-      "loss": 0.3107,
+      "epoch": 1.937506757487296,
+      "grad_norm": 0.18033269047737122,
+      "learning_rate": 3.504300447241473e-05,
+      "loss": 0.393,
       "step": 53760
     },
     {
-      "epoch": 1.89,
-      "learning_rate": 3.5718330297279e-05,
-      "loss": 0.2728,
+      "epoch": 1.9376869571485207,
+      "grad_norm": 0.20827585458755493,
+      "learning_rate": 3.504033208672939e-05,
+      "loss": 0.4061,
       "step": 53765
     },
     {
-      "epoch": 1.89,
-      "learning_rate": 3.5715756604143045e-05,
-      "loss": 0.2799,
+      "epoch": 1.9378671568097452,
+      "grad_norm": 0.22322550415992737,
+      "learning_rate": 3.5037659564246464e-05,
+      "loss": 0.4297,
       "step": 53770
     },
     {
-      "epoch": 1.89,
-      "learning_rate": 3.571318277187005e-05,
-      "loss": 0.2949,
+      "epoch": 1.9380473564709697,
+      "grad_norm": 0.22495147585868835,
+      "learning_rate": 3.5034986905002365e-05,
+      "loss": 0.396,
       "step": 53775
     },
     {
-      "epoch": 1.89,
-      "learning_rate": 3.5710608800493425e-05,
-      "loss": 0.2918,
+      "epoch": 1.9382275561321944,
+      "grad_norm": 0.1754370629787445,
+      "learning_rate": 3.503231410903352e-05,
+      "loss": 0.4142,
       "step": 53780
     },
     {
-      "epoch": 1.89,
-      "learning_rate": 3.570803469004661e-05,
-      "loss": 0.3069,
+      "epoch": 1.9384077557934192,
+      "grad_norm": 0.168187215924263,
+      "learning_rate": 3.5029641176376335e-05,
+      "loss": 0.413,
       "step": 53785
     },
     {
-      "epoch": 1.89,
-      "learning_rate": 3.570546044056303e-05,
-      "loss": 0.3087,
+      "epoch": 1.938587955454644,
+      "grad_norm": 0.17752215266227722,
+      "learning_rate": 3.502696810706723e-05,
+      "loss": 0.4108,
       "step": 53790
     },
     {
-      "epoch": 1.89,
-      "learning_rate": 3.570288605207609e-05,
-      "loss": 0.2924,
+      "epoch": 1.9387681551158684,
+      "grad_norm": 0.19909372925758362,
+      "learning_rate": 3.502429490114263e-05,
+      "loss": 0.4196,
       "step": 53795
     },
     {
-      "epoch": 1.89,
-      "learning_rate": 3.570031152461923e-05,
-      "loss": 0.2727,
+      "epoch": 1.938948354777093,
+      "grad_norm": 0.16391627490520477,
+      "learning_rate": 3.502162155863896e-05,
+      "loss": 0.429,
       "step": 53800
     },
     {
-      "epoch": 1.89,
-      "learning_rate": 3.5697736858225885e-05,
-      "loss": 0.3101,
+      "epoch": 1.9391285544383177,
+      "grad_norm": 0.19088739156723022,
+      "learning_rate": 3.5018948079592626e-05,
+      "loss": 0.356,
       "step": 53805
     },
     {
-      "epoch": 1.89,
-      "learning_rate": 3.569516205292947e-05,
-      "loss": 0.2856,
+      "epoch": 1.9393087540995424,
+      "grad_norm": 0.16824400424957275,
+      "learning_rate": 3.501627446404006e-05,
+      "loss": 0.3539,
       "step": 53810
     },
     {
-      "epoch": 1.89,
-      "learning_rate": 3.5692587108763424e-05,
-      "loss": 0.2822,
+      "epoch": 1.939488953760767,
+      "grad_norm": 0.19399607181549072,
+      "learning_rate": 3.5013600712017704e-05,
+      "loss": 0.3882,
       "step": 53815
     },
     {
-      "epoch": 1.89,
-      "learning_rate": 3.569001202576119e-05,
-      "loss": 0.289,
+      "epoch": 1.9396691534219914,
+      "grad_norm": 0.159646674990654,
+      "learning_rate": 3.501092682356197e-05,
+      "loss": 0.3816,
       "step": 53820
     },
     {
-      "epoch": 1.89,
-      "learning_rate": 3.5687436803956184e-05,
-      "loss": 0.2755,
+      "epoch": 1.9398493530832162,
+      "grad_norm": 0.22390936315059662,
+      "learning_rate": 3.5008252798709294e-05,
+      "loss": 0.4189,
       "step": 53825
     },
     {
-      "epoch": 1.89,
-      "learning_rate": 3.568486144338186e-05,
-      "loss": 0.2801,
+      "epoch": 1.9400295527444409,
+      "grad_norm": 0.18720601499080658,
+      "learning_rate": 3.5005578637496114e-05,
+      "loss": 0.3893,
       "step": 53830
     },
     {
-      "epoch": 1.89,
-      "learning_rate": 3.5682285944071645e-05,
-      "loss": 0.2956,
+      "epoch": 1.9402097524056656,
+      "grad_norm": 0.2625311613082886,
+      "learning_rate": 3.500290433995886e-05,
+      "loss": 0.386,
       "step": 53835
     },
     {
-      "epoch": 1.89,
-      "learning_rate": 3.567971030605899e-05,
-      "loss": 0.2879,
+      "epoch": 1.9403899520668901,
+      "grad_norm": 0.19546449184417725,
+      "learning_rate": 3.500022990613398e-05,
+      "loss": 0.4227,
       "step": 53840
     },
     {
-      "epoch": 1.89,
-      "learning_rate": 3.5677134529377334e-05,
-      "loss": 0.2941,
+      "epoch": 1.9405701517281146,
+      "grad_norm": 0.19765974581241608,
+      "learning_rate": 3.49975553360579e-05,
+      "loss": 0.3913,
       "step": 53845
     },
     {
-      "epoch": 1.89,
-      "learning_rate": 3.567455861406013e-05,
-      "loss": 0.2889,
+      "epoch": 1.9407503513893394,
+      "grad_norm": 0.17598837614059448,
+      "learning_rate": 3.4994880629767056e-05,
+      "loss": 0.3854,
       "step": 53850
     },
     {
-      "epoch": 1.89,
-      "learning_rate": 3.567198256014081e-05,
-      "loss": 0.2969,
+      "epoch": 1.9409305510505641,
+      "grad_norm": 0.17875751852989197,
+      "learning_rate": 3.499220578729791e-05,
+      "loss": 0.407,
       "step": 53855
     },
     {
-      "epoch": 1.89,
-      "learning_rate": 3.566940636765284e-05,
-      "loss": 0.2961,
+      "epoch": 1.9411107507117886,
+      "grad_norm": 0.20579689741134644,
+      "learning_rate": 3.498953080868689e-05,
+      "loss": 0.3687,
       "step": 53860
     },
     {
-      "epoch": 1.9,
-      "learning_rate": 3.566683003662964e-05,
-      "loss": 0.2817,
+      "epoch": 1.9412909503730131,
+      "grad_norm": 0.20091566443443298,
+      "learning_rate": 3.498685569397045e-05,
+      "loss": 0.4192,
       "step": 53865
     },
     {
-      "epoch": 1.9,
-      "learning_rate": 3.5664253567104694e-05,
-      "loss": 0.2542,
+      "epoch": 1.9414711500342379,
+      "grad_norm": 0.17543958127498627,
+      "learning_rate": 3.498418044318502e-05,
+      "loss": 0.3872,
       "step": 53870
     },
     {
-      "epoch": 1.9,
-      "learning_rate": 3.5661676959111444e-05,
-      "loss": 0.2545,
+      "epoch": 1.9416513496954626,
+      "grad_norm": 0.1831250786781311,
+      "learning_rate": 3.498150505636708e-05,
+      "loss": 0.3941,
       "step": 53875
     },
     {
-      "epoch": 1.9,
-      "learning_rate": 3.565910021268334e-05,
-      "loss": 0.2826,
+      "epoch": 1.9418315493566873,
+      "grad_norm": 0.18950533866882324,
+      "learning_rate": 3.4978829533553064e-05,
+      "loss": 0.4167,
       "step": 53880
     },
     {
-      "epoch": 1.9,
-      "learning_rate": 3.565652332785385e-05,
-      "loss": 0.2631,
+      "epoch": 1.9420117490179118,
+      "grad_norm": 0.2181074172258377,
+      "learning_rate": 3.497615387477942e-05,
+      "loss": 0.3822,
       "step": 53885
     },
     {
-      "epoch": 1.9,
-      "learning_rate": 3.5653946304656426e-05,
-      "loss": 0.2975,
+      "epoch": 1.9421919486791364,
+      "grad_norm": 0.20885571837425232,
+      "learning_rate": 3.497347808008262e-05,
+      "loss": 0.4137,
       "step": 53890
     },
     {
-      "epoch": 1.9,
-      "learning_rate": 3.565136914312453e-05,
-      "loss": 0.29,
+      "epoch": 1.942372148340361,
+      "grad_norm": 0.18355844914913177,
+      "learning_rate": 3.4970802149499106e-05,
+      "loss": 0.439,
       "step": 53895
     },
     {
-      "epoch": 1.9,
-      "learning_rate": 3.564879184329162e-05,
-      "loss": 0.277,
+      "epoch": 1.9425523480015858,
+      "grad_norm": 0.17857640981674194,
+      "learning_rate": 3.496812608306535e-05,
+      "loss": 0.4262,
       "step": 53900
     },
     {
-      "epoch": 1.9,
-      "learning_rate": 3.5646214405191176e-05,
-      "loss": 0.2721,
+      "epoch": 1.9427325476628103,
+      "grad_norm": 0.21502956748008728,
+      "learning_rate": 3.4965449880817795e-05,
+      "loss": 0.438,
       "step": 53905
     },
     {
-      "epoch": 1.9,
-      "learning_rate": 3.564363682885665e-05,
-      "loss": 0.283,
+      "epoch": 1.942912747324035,
+      "grad_norm": 0.17269867658615112,
+      "learning_rate": 3.4962773542792925e-05,
+      "loss": 0.3961,
       "step": 53910
     },
     {
-      "epoch": 1.9,
-      "learning_rate": 3.564105911432151e-05,
-      "loss": 0.2778,
+      "epoch": 1.9430929469852596,
+      "grad_norm": 0.1496947705745697,
+      "learning_rate": 3.4960097069027184e-05,
+      "loss": 0.4213,
       "step": 53915
     },
     {
-      "epoch": 1.9,
-      "learning_rate": 3.563848126161924e-05,
-      "loss": 0.2981,
+      "epoch": 1.9432731466464843,
+      "grad_norm": 0.19583889842033386,
+      "learning_rate": 3.495742045955706e-05,
+      "loss": 0.4293,
       "step": 53920
     },
     {
-      "epoch": 1.9,
-      "learning_rate": 3.563590327078329e-05,
-      "loss": 0.3012,
+      "epoch": 1.943453346307709,
+      "grad_norm": 0.1784418672323227,
+      "learning_rate": 3.4954743714419006e-05,
+      "loss": 0.3941,
       "step": 53925
     },
     {
-      "epoch": 1.9,
-      "learning_rate": 3.5633325141847144e-05,
-      "loss": 0.2836,
+      "epoch": 1.9436335459689336,
+      "grad_norm": 0.2068331241607666,
+      "learning_rate": 3.4952066833649495e-05,
+      "loss": 0.4088,
       "step": 53930
     },
     {
-      "epoch": 1.9,
-      "learning_rate": 3.563074687484429e-05,
-      "loss": 0.2967,
+      "epoch": 1.943813745630158,
+      "grad_norm": 0.18008601665496826,
+      "learning_rate": 3.4949389817285e-05,
+      "loss": 0.3759,
       "step": 53935
     },
     {
-      "epoch": 1.9,
-      "learning_rate": 3.5628168469808184e-05,
-      "loss": 0.2996,
+      "epoch": 1.9439939452913828,
+      "grad_norm": 0.17084579169750214,
+      "learning_rate": 3.494671266536199e-05,
+      "loss": 0.3866,
       "step": 53940
     },
     {
-      "epoch": 1.9,
-      "learning_rate": 3.562558992677233e-05,
-      "loss": 0.2953,
+      "epoch": 1.9441741449526075,
+      "grad_norm": 0.1742352992296219,
+      "learning_rate": 3.494403537791696e-05,
+      "loss": 0.4175,
       "step": 53945
     },
     {
-      "epoch": 1.9,
-      "learning_rate": 3.562301124577018e-05,
-      "loss": 0.2628,
+      "epoch": 1.9443543446138323,
+      "grad_norm": 0.2424996942281723,
+      "learning_rate": 3.494135795498636e-05,
+      "loss": 0.416,
       "step": 53950
     },
     {
-      "epoch": 1.9,
-      "learning_rate": 3.5620432426835226e-05,
-      "loss": 0.2767,
+      "epoch": 1.9445345442750568,
+      "grad_norm": 0.17355568706989288,
+      "learning_rate": 3.493868039660669e-05,
+      "loss": 0.4145,
       "step": 53955
     },
     {
-      "epoch": 1.9,
-      "learning_rate": 3.561785347000096e-05,
-      "loss": 0.2936,
+      "epoch": 1.9447147439362813,
+      "grad_norm": 0.15699560940265656,
+      "learning_rate": 3.493600270281442e-05,
+      "loss": 0.4125,
       "step": 53960
     },
     {
-      "epoch": 1.9,
-      "learning_rate": 3.561527437530088e-05,
-      "loss": 0.295,
+      "epoch": 1.944894943597506,
+      "grad_norm": 0.16016830503940582,
+      "learning_rate": 3.493332487364604e-05,
+      "loss": 0.4297,
       "step": 53965
     },
     {
-      "epoch": 1.9,
-      "learning_rate": 3.561269514276844e-05,
-      "loss": 0.2892,
+      "epoch": 1.9450751432587308,
+      "grad_norm": 0.23133453726768494,
+      "learning_rate": 3.4930646909138026e-05,
+      "loss": 0.4289,
       "step": 53970
     },
     {
-      "epoch": 1.9,
-      "learning_rate": 3.561011577243715e-05,
-      "loss": 0.2718,
+      "epoch": 1.9452553429199553,
+      "grad_norm": 0.23153449594974518,
+      "learning_rate": 3.492796880932687e-05,
+      "loss": 0.4518,
       "step": 53975
     },
     {
-      "epoch": 1.9,
-      "learning_rate": 3.5607536264340497e-05,
-      "loss": 0.2776,
+      "epoch": 1.9454355425811798,
+      "grad_norm": 0.1748327612876892,
+      "learning_rate": 3.492529057424907e-05,
+      "loss": 0.41,
       "step": 53980
     },
     {
-      "epoch": 1.9,
-      "learning_rate": 3.560495661851198e-05,
-      "loss": 0.2718,
+      "epoch": 1.9456157422424045,
+      "grad_norm": 0.2023671418428421,
+      "learning_rate": 3.49226122039411e-05,
+      "loss": 0.4217,
       "step": 53985
     },
     {
-      "epoch": 1.9,
-      "learning_rate": 3.5602376834985094e-05,
-      "loss": 0.2969,
+      "epoch": 1.9457959419036293,
+      "grad_norm": 0.15689332783222198,
+      "learning_rate": 3.491993369843946e-05,
+      "loss": 0.3638,
       "step": 53990
     },
     {
-      "epoch": 1.9,
-      "learning_rate": 3.5599796913793326e-05,
-      "loss": 0.2846,
+      "epoch": 1.945976141564854,
+      "grad_norm": 0.19829018414020538,
+      "learning_rate": 3.4917255057780646e-05,
+      "loss": 0.4338,
       "step": 53995
     },
     {
-      "epoch": 1.9,
-      "learning_rate": 3.559721685497019e-05,
-      "loss": 0.2723,
+      "epoch": 1.9461563412260785,
+      "grad_norm": 0.19208745658397675,
+      "learning_rate": 3.491457628200115e-05,
+      "loss": 0.4192,
       "step": 54000
     },
     {
-      "epoch": 1.9,
-      "eval_loss": 0.27377253770828247,
-      "eval_runtime": 10.5385,
-      "eval_samples_per_second": 9.489,
-      "eval_steps_per_second": 9.489,
+      "epoch": 1.9461563412260785,
+      "eval_loss": 0.4374231994152069,
+      "eval_runtime": 3.5248,
+      "eval_samples_per_second": 28.37,
+      "eval_steps_per_second": 7.093,
       "step": 54000
     },
     {
-      "epoch": 1.9,
-      "learning_rate": 3.559463665854916e-05,
-      "loss": 0.2901,
+      "epoch": 1.946336540887303,
+      "grad_norm": 0.20834042131900787,
+      "learning_rate": 3.491189737113748e-05,
+      "loss": 0.4054,
       "step": 54005
     },
     {
-      "epoch": 1.9,
-      "learning_rate": 3.5592056324563775e-05,
-      "loss": 0.2751,
+      "epoch": 1.9465167405485277,
+      "grad_norm": 0.2228306233882904,
+      "learning_rate": 3.490921832522612e-05,
+      "loss": 0.394,
       "step": 54010
     },
     {
-      "epoch": 1.9,
-      "learning_rate": 3.5589475853047505e-05,
-      "loss": 0.2912,
+      "epoch": 1.9466969402097525,
+      "grad_norm": 0.1861565113067627,
+      "learning_rate": 3.490653914430358e-05,
+      "loss": 0.4106,
       "step": 54015
     },
     {
-      "epoch": 1.9,
-      "learning_rate": 3.5586895244033874e-05,
-      "loss": 0.2607,
+      "epoch": 1.946877139870977,
+      "grad_norm": 0.1798921376466751,
+      "learning_rate": 3.490385982840636e-05,
+      "loss": 0.4095,
       "step": 54020
     },
     {
-      "epoch": 1.9,
-      "learning_rate": 3.5584314497556385e-05,
-      "loss": 0.2925,
+      "epoch": 1.9470573395322017,
+      "grad_norm": 0.15824037790298462,
+      "learning_rate": 3.490118037757097e-05,
+      "loss": 0.4057,
       "step": 54025
     },
     {
-      "epoch": 1.9,
-      "learning_rate": 3.558173361364855e-05,
-      "loss": 0.2835,
+      "epoch": 1.9472375391934262,
+      "grad_norm": 0.18199415504932404,
+      "learning_rate": 3.489850079183391e-05,
+      "loss": 0.3894,
       "step": 54030
     },
     {
-      "epoch": 1.9,
-      "learning_rate": 3.557915259234387e-05,
-      "loss": 0.275,
+      "epoch": 1.947417738854651,
+      "grad_norm": 0.18391664326190948,
+      "learning_rate": 3.489582107123169e-05,
+      "loss": 0.4136,
       "step": 54035
     },
     {
-      "epoch": 1.9,
-      "learning_rate": 3.5576571433675876e-05,
-      "loss": 0.2868,
+      "epoch": 1.9475979385158757,
+      "grad_norm": 0.1818101406097412,
+      "learning_rate": 3.489314121580084e-05,
+      "loss": 0.3785,
       "step": 54040
     },
     {
-      "epoch": 1.9,
-      "learning_rate": 3.5573990137678065e-05,
-      "loss": 0.2768,
+      "epoch": 1.9477781381771002,
+      "grad_norm": 0.2508971095085144,
+      "learning_rate": 3.4890461225577844e-05,
+      "loss": 0.4335,
       "step": 54045
     },
     {
-      "epoch": 1.9,
-      "learning_rate": 3.557140870438397e-05,
-      "loss": 0.2995,
+      "epoch": 1.9479583378383247,
+      "grad_norm": 0.17361201345920563,
+      "learning_rate": 3.488778110059924e-05,
+      "loss": 0.3918,
       "step": 54050
     },
     {
-      "epoch": 1.9,
-      "learning_rate": 3.556882713382709e-05,
-      "loss": 0.2717,
+      "epoch": 1.9481385374995495,
+      "grad_norm": 0.19530875980854034,
+      "learning_rate": 3.488510084090152e-05,
+      "loss": 0.4348,
       "step": 54055
     },
     {
-      "epoch": 1.9,
-      "learning_rate": 3.556624542604096e-05,
-      "loss": 0.2686,
+      "epoch": 1.9483187371607742,
+      "grad_norm": 0.19267144799232483,
+      "learning_rate": 3.488242044652122e-05,
+      "loss": 0.4229,
       "step": 54060
     },
     {
-      "epoch": 1.9,
-      "learning_rate": 3.5563663581059084e-05,
-      "loss": 0.2718,
+      "epoch": 1.948498936821999,
+      "grad_norm": 0.1983078569173813,
+      "learning_rate": 3.487973991749485e-05,
+      "loss": 0.4049,
       "step": 54065
     },
     {
-      "epoch": 1.9,
-      "learning_rate": 3.556108159891502e-05,
-      "loss": 0.3123,
+      "epoch": 1.9486791364832234,
+      "grad_norm": 0.156129390001297,
+      "learning_rate": 3.487705925385894e-05,
+      "loss": 0.4582,
       "step": 54070
     },
     {
-      "epoch": 1.9,
-      "learning_rate": 3.555849947964225e-05,
-      "loss": 0.2785,
+      "epoch": 1.948859336144448,
+      "grad_norm": 0.17622871696949005,
+      "learning_rate": 3.4874378455650016e-05,
+      "loss": 0.3908,
       "step": 54075
     },
     {
-      "epoch": 1.9,
-      "learning_rate": 3.555591722327434e-05,
-      "loss": 0.2834,
+      "epoch": 1.9490395358056727,
+      "grad_norm": 0.21170586347579956,
+      "learning_rate": 3.487169752290458e-05,
+      "loss": 0.4382,
       "step": 54080
     },
     {
-      "epoch": 1.9,
-      "learning_rate": 3.555333482984479e-05,
-      "loss": 0.277,
+      "epoch": 1.9492197354668974,
+      "grad_norm": 0.2054653912782669,
+      "learning_rate": 3.48690164556592e-05,
+      "loss": 0.4048,
       "step": 54085
     },
     {
-      "epoch": 1.9,
-      "learning_rate": 3.555075229938715e-05,
-      "loss": 0.3086,
+      "epoch": 1.949399935128122,
+      "grad_norm": 0.17421171069145203,
+      "learning_rate": 3.486633525395037e-05,
+      "loss": 0.3885,
       "step": 54090
     },
     {
-      "epoch": 1.9,
-      "learning_rate": 3.5548169631934936e-05,
-      "loss": 0.3118,
+      "epoch": 1.9495801347893464,
+      "grad_norm": 0.21200191974639893,
+      "learning_rate": 3.4863653917814626e-05,
+      "loss": 0.439,
       "step": 54095
     },
     {
-      "epoch": 1.9,
-      "learning_rate": 3.5545586827521695e-05,
-      "loss": 0.2825,
+      "epoch": 1.9497603344505712,
+      "grad_norm": 0.17257794737815857,
+      "learning_rate": 3.486097244728851e-05,
+      "loss": 0.3801,
       "step": 54100
     },
     {
-      "epoch": 1.9,
-      "learning_rate": 3.554300388618096e-05,
-      "loss": 0.3147,
+      "epoch": 1.949940534111796,
+      "grad_norm": 0.1847560554742813,
+      "learning_rate": 3.485829084240856e-05,
+      "loss": 0.3753,
       "step": 54105
     },
     {
-      "epoch": 1.9,
-      "learning_rate": 3.5540420807946264e-05,
-      "loss": 0.2986,
+      "epoch": 1.9501207337730206,
+      "grad_norm": 0.2439274787902832,
+      "learning_rate": 3.48556091032113e-05,
+      "loss": 0.4313,
       "step": 54110
     },
     {
-      "epoch": 1.9,
-      "learning_rate": 3.553783759285115e-05,
-      "loss": 0.2992,
+      "epoch": 1.9503009334342452,
+      "grad_norm": 0.1944105327129364,
+      "learning_rate": 3.485292722973327e-05,
+      "loss": 0.4109,
       "step": 54115
     },
     {
-      "epoch": 1.9,
-      "learning_rate": 3.553525424092916e-05,
-      "loss": 0.2876,
+      "epoch": 1.9504811330954697,
+      "grad_norm": 0.189129039645195,
+      "learning_rate": 3.4850245222011025e-05,
+      "loss": 0.3661,
       "step": 54120
     },
     {
-      "epoch": 1.9,
-      "learning_rate": 3.553267075221385e-05,
-      "loss": 0.2788,
+      "epoch": 1.9506613327566944,
+      "grad_norm": 0.2111881524324417,
+      "learning_rate": 3.484756308008109e-05,
+      "loss": 0.4375,
       "step": 54125
     },
     {
-      "epoch": 1.9,
-      "learning_rate": 3.553008712673873e-05,
-      "loss": 0.2929,
+      "epoch": 1.9508415324179191,
+      "grad_norm": 0.14287078380584717,
+      "learning_rate": 3.484488080398002e-05,
+      "loss": 0.399,
       "step": 54130
     },
     {
-      "epoch": 1.9,
-      "learning_rate": 3.552750336453738e-05,
-      "loss": 0.3044,
+      "epoch": 1.9510217320791436,
+      "grad_norm": 0.1772930771112442,
+      "learning_rate": 3.4842198393744354e-05,
+      "loss": 0.4539,
       "step": 54135
     },
     {
-      "epoch": 1.9,
-      "learning_rate": 3.552491946564334e-05,
-      "loss": 0.2799,
+      "epoch": 1.9512019317403684,
+      "grad_norm": 0.143804669380188,
+      "learning_rate": 3.483951584941063e-05,
+      "loss": 0.4164,
       "step": 54140
     },
     {
-      "epoch": 1.9,
-      "learning_rate": 3.5522335430090153e-05,
-      "loss": 0.2941,
+      "epoch": 1.951382131401593,
+      "grad_norm": 0.2153077870607376,
+      "learning_rate": 3.483683317101541e-05,
+      "loss": 0.4466,
       "step": 54145
     },
     {
-      "epoch": 1.91,
-      "learning_rate": 3.551975125791137e-05,
-      "loss": 0.2747,
+      "epoch": 1.9515623310628176,
+      "grad_norm": 0.2091062068939209,
+      "learning_rate": 3.483415035859525e-05,
+      "loss": 0.4344,
       "step": 54150
     },
     {
-      "epoch": 1.91,
-      "learning_rate": 3.551716694914056e-05,
-      "loss": 0.286,
+      "epoch": 1.9517425307240424,
+      "grad_norm": 0.19160053133964539,
+      "learning_rate": 3.48314674121867e-05,
+      "loss": 0.3617,
       "step": 54155
     },
     {
-      "epoch": 1.91,
-      "learning_rate": 3.551458250381126e-05,
-      "loss": 0.2917,
+      "epoch": 1.9519227303852669,
+      "grad_norm": 0.1973278522491455,
+      "learning_rate": 3.48287843318263e-05,
+      "loss": 0.3868,
       "step": 54160
     },
     {
-      "epoch": 1.91,
-      "learning_rate": 3.551199792195704e-05,
-      "loss": 0.3029,
+      "epoch": 1.9521029300464914,
+      "grad_norm": 0.24400056898593903,
+      "learning_rate": 3.482610111755062e-05,
+      "loss": 0.4246,
       "step": 54165
     },
     {
-      "epoch": 1.91,
-      "learning_rate": 3.5509413203611445e-05,
-      "loss": 0.2845,
+      "epoch": 1.9522831297077161,
+      "grad_norm": 0.18288040161132812,
+      "learning_rate": 3.4823417769396214e-05,
+      "loss": 0.4284,
       "step": 54170
     },
     {
-      "epoch": 1.91,
-      "learning_rate": 3.550682834880806e-05,
-      "loss": 0.2963,
+      "epoch": 1.9524633293689408,
+      "grad_norm": 0.18425007164478302,
+      "learning_rate": 3.482073428739964e-05,
+      "loss": 0.4324,
       "step": 54175
     },
     {
-      "epoch": 1.91,
-      "learning_rate": 3.550424335758042e-05,
-      "loss": 0.2871,
+      "epoch": 1.9526435290301656,
+      "grad_norm": 0.1836194396018982,
+      "learning_rate": 3.481805067159747e-05,
+      "loss": 0.4368,
       "step": 54180
     },
     {
-      "epoch": 1.91,
-      "learning_rate": 3.550165822996212e-05,
-      "loss": 0.2846,
+      "epoch": 1.95282372869139,
+      "grad_norm": 0.1682935655117035,
+      "learning_rate": 3.481536692202625e-05,
+      "loss": 0.4079,
       "step": 54185
     },
     {
-      "epoch": 1.91,
-      "learning_rate": 3.5499072965986694e-05,
-      "loss": 0.2951,
+      "epoch": 1.9530039283526146,
+      "grad_norm": 0.18854603171348572,
+      "learning_rate": 3.4812683038722565e-05,
+      "loss": 0.4287,
       "step": 54190
     },
     {
-      "epoch": 1.91,
-      "learning_rate": 3.549648756568773e-05,
-      "loss": 0.2784,
+      "epoch": 1.9531841280138393,
+      "grad_norm": 0.1703491359949112,
+      "learning_rate": 3.4809999021722964e-05,
+      "loss": 0.3948,
       "step": 54195
     },
     {
-      "epoch": 1.91,
-      "learning_rate": 3.549390202909879e-05,
-      "loss": 0.278,
+      "epoch": 1.953364327675064,
+      "grad_norm": 0.18725082278251648,
+      "learning_rate": 3.480731487106404e-05,
+      "loss": 0.4472,
       "step": 54200
     },
     {
-      "epoch": 1.91,
-      "learning_rate": 3.5491316356253446e-05,
-      "loss": 0.2768,
+      "epoch": 1.9535445273362886,
+      "grad_norm": 0.17395414412021637,
+      "learning_rate": 3.4804630586782336e-05,
+      "loss": 0.4191,
       "step": 54205
     },
     {
-      "epoch": 1.91,
-      "learning_rate": 3.5488730547185276e-05,
-      "loss": 0.2931,
+      "epoch": 1.953724726997513,
+      "grad_norm": 0.20272298157215118,
+      "learning_rate": 3.4801946168914445e-05,
+      "loss": 0.4106,
       "step": 54210
     },
     {
-      "epoch": 1.91,
-      "learning_rate": 3.548614460192785e-05,
-      "loss": 0.2693,
+      "epoch": 1.9539049266587378,
+      "grad_norm": 0.2384205460548401,
+      "learning_rate": 3.4799261617496936e-05,
+      "loss": 0.3815,
       "step": 54215
     },
     {
-      "epoch": 1.91,
-      "learning_rate": 3.5483558520514755e-05,
-      "loss": 0.2674,
+      "epoch": 1.9540851263199626,
+      "grad_norm": 0.1440509706735611,
+      "learning_rate": 3.4796576932566374e-05,
+      "loss": 0.4261,
       "step": 54220
     },
     {
-      "epoch": 1.91,
-      "learning_rate": 3.548097230297954e-05,
-      "loss": 0.2962,
+      "epoch": 1.9542653259811873,
+      "grad_norm": 0.20167045295238495,
+      "learning_rate": 3.4793892114159364e-05,
+      "loss": 0.4292,
       "step": 54225
     },
     {
-      "epoch": 1.91,
-      "learning_rate": 3.547838594935583e-05,
-      "loss": 0.2953,
+      "epoch": 1.9544455256424118,
+      "grad_norm": 0.179216668009758,
+      "learning_rate": 3.4791207162312464e-05,
+      "loss": 0.399,
       "step": 54230
     },
     {
-      "epoch": 1.91,
-      "learning_rate": 3.547579945967718e-05,
-      "loss": 0.2914,
+      "epoch": 1.9546257253036363,
+      "grad_norm": 0.18981371819972992,
+      "learning_rate": 3.478852207706226e-05,
+      "loss": 0.4152,
       "step": 54235
     },
     {
-      "epoch": 1.91,
-      "learning_rate": 3.547321283397718e-05,
-      "loss": 0.3063,
+      "epoch": 1.954805924964861,
+      "grad_norm": 0.1702110916376114,
+      "learning_rate": 3.4785836858445334e-05,
+      "loss": 0.4375,
       "step": 54240
     },
     {
-      "epoch": 1.91,
-      "learning_rate": 3.54706260722894e-05,
-      "loss": 0.2954,
+      "epoch": 1.9549861246260858,
+      "grad_norm": 0.17524947226047516,
+      "learning_rate": 3.478315150649829e-05,
+      "loss": 0.4262,
       "step": 54245
     },
     {
-      "epoch": 1.91,
-      "learning_rate": 3.5468039174647446e-05,
-      "loss": 0.298,
+      "epoch": 1.9551663242873103,
+      "grad_norm": 0.20712916553020477,
+      "learning_rate": 3.4780466021257685e-05,
+      "loss": 0.4056,
       "step": 54250
     },
     {
-      "epoch": 1.91,
-      "learning_rate": 3.54654521410849e-05,
-      "loss": 0.2783,
+      "epoch": 1.955346523948535,
+      "grad_norm": 0.171823188662529,
+      "learning_rate": 3.477778040276013e-05,
+      "loss": 0.3729,
       "step": 54255
     },
     {
-      "epoch": 1.91,
-      "learning_rate": 3.546286497163537e-05,
-      "loss": 0.2827,
+      "epoch": 1.9555267236097595,
+      "grad_norm": 0.16928203403949738,
+      "learning_rate": 3.4775094651042204e-05,
+      "loss": 0.4196,
       "step": 54260
     },
     {
-      "epoch": 1.91,
-      "learning_rate": 3.546027766633242e-05,
-      "loss": 0.2954,
+      "epoch": 1.9557069232709843,
+      "grad_norm": 0.19515909254550934,
+      "learning_rate": 3.4772408766140516e-05,
+      "loss": 0.4314,
       "step": 54265
     },
     {
-      "epoch": 1.91,
-      "learning_rate": 3.545769022520966e-05,
-      "loss": 0.2752,
+      "epoch": 1.955887122932209,
+      "grad_norm": 0.1939311921596527,
+      "learning_rate": 3.4769722748091646e-05,
+      "loss": 0.4058,
       "step": 54270
     },
     {
-      "epoch": 1.91,
-      "learning_rate": 3.5455102648300686e-05,
-      "loss": 0.2836,
+      "epoch": 1.9560673225934335,
+      "grad_norm": 0.2215987592935562,
+      "learning_rate": 3.47670365969322e-05,
+      "loss": 0.3656,
       "step": 54275
     },
     {
-      "epoch": 1.91,
-      "learning_rate": 3.545251493563909e-05,
-      "loss": 0.2787,
+      "epoch": 1.956247522254658,
+      "grad_norm": 0.18643298745155334,
+      "learning_rate": 3.476435031269876e-05,
+      "loss": 0.4374,
       "step": 54280
     },
     {
-      "epoch": 1.91,
-      "learning_rate": 3.544992708725848e-05,
-      "loss": 0.2801,
+      "epoch": 1.9564277219158828,
+      "grad_norm": 0.20477153360843658,
+      "learning_rate": 3.4761663895427946e-05,
+      "loss": 0.419,
       "step": 54285
     },
     {
-      "epoch": 1.91,
-      "learning_rate": 3.5447339103192454e-05,
-      "loss": 0.2571,
+      "epoch": 1.9566079215771075,
+      "grad_norm": 0.1482023149728775,
+      "learning_rate": 3.475897734515635e-05,
+      "loss": 0.4041,
       "step": 54290
     },
     {
-      "epoch": 1.91,
-      "learning_rate": 3.544475098347462e-05,
-      "loss": 0.2798,
+      "epoch": 1.9567881212383322,
+      "grad_norm": 0.17524316906929016,
+      "learning_rate": 3.4756290661920575e-05,
+      "loss": 0.4322,
       "step": 54295
     },
     {
-      "epoch": 1.91,
-      "learning_rate": 3.544216272813857e-05,
-      "loss": 0.2982,
+      "epoch": 1.9569683208995567,
+      "grad_norm": 0.21270814538002014,
+      "learning_rate": 3.4753603845757235e-05,
+      "loss": 0.373,
       "step": 54300
     },
     {
-      "epoch": 1.91,
-      "learning_rate": 3.543957433721792e-05,
-      "loss": 0.276,
+      "epoch": 1.9571485205607813,
+      "grad_norm": 0.1944366991519928,
+      "learning_rate": 3.475091689670292e-05,
+      "loss": 0.4083,
       "step": 54305
     },
     {
-      "epoch": 1.91,
-      "learning_rate": 3.543698581074628e-05,
-      "loss": 0.2727,
+      "epoch": 1.957328720222006,
+      "grad_norm": 0.2040632963180542,
+      "learning_rate": 3.4748229814794256e-05,
+      "loss": 0.386,
       "step": 54310
     },
     {
-      "epoch": 1.91,
-      "learning_rate": 3.543439714875725e-05,
-      "loss": 0.2706,
+      "epoch": 1.9575089198832307,
+      "grad_norm": 0.19412831962108612,
+      "learning_rate": 3.474554260006785e-05,
+      "loss": 0.4259,
       "step": 54315
     },
     {
-      "epoch": 1.91,
-      "learning_rate": 3.5431808351284455e-05,
-      "loss": 0.2784,
+      "epoch": 1.9576891195444552,
+      "grad_norm": 0.17978821694850922,
+      "learning_rate": 3.4742855252560315e-05,
+      "loss": 0.404,
       "step": 54320
     },
     {
-      "epoch": 1.91,
-      "learning_rate": 3.54292194183615e-05,
-      "loss": 0.295,
+      "epoch": 1.9578693192056797,
+      "grad_norm": 0.19575315713882446,
+      "learning_rate": 3.474016777230825e-05,
+      "loss": 0.4097,
       "step": 54325
     },
     {
-      "epoch": 1.91,
-      "learning_rate": 3.542663035002201e-05,
-      "loss": 0.2611,
+      "epoch": 1.9580495188669045,
+      "grad_norm": 0.16315071284770966,
+      "learning_rate": 3.473748015934829e-05,
+      "loss": 0.4187,
       "step": 54330
     },
     {
-      "epoch": 1.91,
-      "learning_rate": 3.542404114629959e-05,
-      "loss": 0.2928,
+      "epoch": 1.9582297185281292,
+      "grad_norm": 0.20219777524471283,
+      "learning_rate": 3.473479241371706e-05,
+      "loss": 0.4193,
       "step": 54335
     },
     {
-      "epoch": 1.91,
-      "learning_rate": 3.5421451807227866e-05,
-      "loss": 0.2922,
+      "epoch": 1.958409918189354,
+      "grad_norm": 0.17881831526756287,
+      "learning_rate": 3.473210453545116e-05,
+      "loss": 0.3867,
       "step": 54340
     },
     {
-      "epoch": 1.91,
-      "learning_rate": 3.541886233284047e-05,
-      "loss": 0.277,
+      "epoch": 1.9585901178505785,
+      "grad_norm": 0.17221704125404358,
+      "learning_rate": 3.472941652458722e-05,
+      "loss": 0.4314,
       "step": 54345
     },
     {
-      "epoch": 1.91,
-      "learning_rate": 3.5416272723171e-05,
-      "loss": 0.2538,
+      "epoch": 1.958770317511803,
+      "grad_norm": 0.15523254871368408,
+      "learning_rate": 3.472672838116187e-05,
+      "loss": 0.3914,
       "step": 54350
     },
     {
-      "epoch": 1.91,
-      "learning_rate": 3.5413682978253096e-05,
-      "loss": 0.2852,
+      "epoch": 1.9589505171730277,
+      "grad_norm": 0.19016127288341522,
+      "learning_rate": 3.472404010521172e-05,
+      "loss": 0.4245,
       "step": 54355
     },
     {
-      "epoch": 1.91,
-      "learning_rate": 3.541109309812038e-05,
-      "loss": 0.3012,
+      "epoch": 1.9591307168342524,
+      "grad_norm": 0.17222009599208832,
+      "learning_rate": 3.472135169677341e-05,
+      "loss": 0.4452,
       "step": 54360
     },
     {
-      "epoch": 1.91,
-      "learning_rate": 3.540850308280649e-05,
-      "loss": 0.2755,
+      "epoch": 1.959310916495477,
+      "grad_norm": 0.205901101231575,
+      "learning_rate": 3.471866315588356e-05,
+      "loss": 0.4402,
       "step": 54365
     },
     {
-      "epoch": 1.91,
-      "learning_rate": 3.5405912932345034e-05,
-      "loss": 0.2756,
+      "epoch": 1.9594911161567015,
+      "grad_norm": 0.21535156667232513,
+      "learning_rate": 3.4715974482578814e-05,
+      "loss": 0.4408,
       "step": 54370
     },
     {
-      "epoch": 1.91,
-      "learning_rate": 3.540332264676967e-05,
-      "loss": 0.2885,
+      "epoch": 1.9596713158179262,
+      "grad_norm": 0.18163524568080902,
+      "learning_rate": 3.471328567689579e-05,
+      "loss": 0.4052,
       "step": 54375
     },
     {
-      "epoch": 1.91,
-      "learning_rate": 3.540073222611401e-05,
-      "loss": 0.2794,
+      "epoch": 1.959851515479151,
+      "grad_norm": 0.20107874274253845,
+      "learning_rate": 3.471059673887114e-05,
+      "loss": 0.3852,
       "step": 54380
     },
     {
-      "epoch": 1.91,
-      "learning_rate": 3.53981416704117e-05,
-      "loss": 0.2606,
+      "epoch": 1.9600317151403757,
+      "grad_norm": 0.19046540558338165,
+      "learning_rate": 3.470790766854148e-05,
+      "loss": 0.4287,
       "step": 54385
     },
     {
-      "epoch": 1.91,
-      "learning_rate": 3.5395550979696366e-05,
-      "loss": 0.2811,
+      "epoch": 1.9602119148016002,
+      "grad_norm": 0.16740958392620087,
+      "learning_rate": 3.4705218465943456e-05,
+      "loss": 0.4202,
       "step": 54390
     },
     {
-      "epoch": 1.91,
-      "learning_rate": 3.539296015400166e-05,
-      "loss": 0.2918,
+      "epoch": 1.9603921144628247,
+      "grad_norm": 0.21119292080402374,
+      "learning_rate": 3.4702529131113715e-05,
+      "loss": 0.4048,
       "step": 54395
     },
     {
-      "epoch": 1.91,
-      "learning_rate": 3.5390369193361204e-05,
-      "loss": 0.2801,
+      "epoch": 1.9605723141240494,
+      "grad_norm": 0.20554935932159424,
+      "learning_rate": 3.469983966408889e-05,
+      "loss": 0.4192,
       "step": 54400
     },
     {
-      "epoch": 1.91,
-      "learning_rate": 3.538829632771053e-05,
-      "loss": 0.2677,
+      "epoch": 1.9607525137852742,
+      "grad_norm": 0.2031758427619934,
+      "learning_rate": 3.469715006490563e-05,
+      "loss": 0.4199,
       "step": 54405
     },
     {
-      "epoch": 1.91,
-      "learning_rate": 3.5385705124252526e-05,
-      "loss": 0.2817,
+      "epoch": 1.9609327134464987,
+      "grad_norm": 0.18050605058670044,
+      "learning_rate": 3.4694460333600574e-05,
+      "loss": 0.3987,
       "step": 54410
     },
     {
-      "epoch": 1.91,
-      "learning_rate": 3.5383113785943e-05,
-      "loss": 0.2995,
+      "epoch": 1.9611129131077234,
+      "grad_norm": 0.19206275045871735,
+      "learning_rate": 3.4691770470210374e-05,
+      "loss": 0.4103,
       "step": 54415
     },
     {
-      "epoch": 1.91,
-      "learning_rate": 3.5380522312815574e-05,
-      "loss": 0.264,
+      "epoch": 1.961293112768948,
+      "grad_norm": 0.16883911192417145,
+      "learning_rate": 3.4689080474771676e-05,
+      "loss": 0.4011,
       "step": 54420
     },
     {
-      "epoch": 1.91,
-      "learning_rate": 3.53779307049039e-05,
-      "loss": 0.2944,
+      "epoch": 1.9614733124301726,
+      "grad_norm": 0.1800912320613861,
+      "learning_rate": 3.4686390347321144e-05,
+      "loss": 0.3789,
       "step": 54425
     },
     {
-      "epoch": 1.91,
-      "learning_rate": 3.537533896224165e-05,
-      "loss": 0.2951,
+      "epoch": 1.9616535120913974,
+      "grad_norm": 0.1670406460762024,
+      "learning_rate": 3.46837000878954e-05,
+      "loss": 0.4023,
       "step": 54430
     },
     {
-      "epoch": 1.92,
-      "learning_rate": 3.537274708486246e-05,
-      "loss": 0.2744,
+      "epoch": 1.961833711752622,
+      "grad_norm": 0.14875911176204681,
+      "learning_rate": 3.468100969653114e-05,
+      "loss": 0.3879,
       "step": 54435
     },
     {
-      "epoch": 1.92,
-      "learning_rate": 3.5370155072799974e-05,
-      "loss": 0.3025,
+      "epoch": 1.9620139114138464,
+      "grad_norm": 0.1597927361726761,
+      "learning_rate": 3.4678319173264975e-05,
+      "loss": 0.4384,
       "step": 54440
     },
     {
-      "epoch": 1.92,
-      "learning_rate": 3.536756292608787e-05,
-      "loss": 0.3011,
+      "epoch": 1.9621941110750711,
+      "grad_norm": 0.15800374746322632,
+      "learning_rate": 3.46756285181336e-05,
+      "loss": 0.378,
       "step": 54445
     },
     {
-      "epoch": 1.92,
-      "learning_rate": 3.536497064475979e-05,
-      "loss": 0.2933,
+      "epoch": 1.9623743107362959,
+      "grad_norm": 0.1771484762430191,
+      "learning_rate": 3.467293773117365e-05,
+      "loss": 0.4328,
       "step": 54450
     },
     {
-      "epoch": 1.92,
-      "learning_rate": 3.536237822884939e-05,
-      "loss": 0.2763,
+      "epoch": 1.9625545103975206,
+      "grad_norm": 0.16408397257328033,
+      "learning_rate": 3.46702468124218e-05,
+      "loss": 0.3782,
       "step": 54455
     },
     {
-      "epoch": 1.92,
-      "learning_rate": 3.5359785678390346e-05,
-      "loss": 0.3083,
+      "epoch": 1.9627347100587451,
+      "grad_norm": 0.17288720607757568,
+      "learning_rate": 3.466755576191471e-05,
+      "loss": 0.4191,
       "step": 54460
     },
     {
-      "epoch": 1.92,
-      "learning_rate": 3.535719299341631e-05,
-      "loss": 0.289,
+      "epoch": 1.9629149097199696,
+      "grad_norm": 0.1751626580953598,
+      "learning_rate": 3.4664864579689036e-05,
+      "loss": 0.4097,
       "step": 54465
     },
     {
-      "epoch": 1.92,
-      "learning_rate": 3.535460017396095e-05,
-      "loss": 0.2785,
+      "epoch": 1.9630951093811944,
+      "grad_norm": 0.19084684550762177,
+      "learning_rate": 3.4662173265781464e-05,
+      "loss": 0.4035,
       "step": 54470
     },
     {
-      "epoch": 1.92,
-      "learning_rate": 3.535200722005793e-05,
-      "loss": 0.2717,
+      "epoch": 1.963275309042419,
+      "grad_norm": 0.1722564846277237,
+      "learning_rate": 3.4659481820228654e-05,
+      "loss": 0.4273,
       "step": 54475
     },
     {
-      "epoch": 1.92,
-      "learning_rate": 3.5349414131740914e-05,
-      "loss": 0.2753,
+      "epoch": 1.9634555087036436,
+      "grad_norm": 0.17883461713790894,
+      "learning_rate": 3.4656790243067274e-05,
+      "loss": 0.397,
       "step": 54480
     },
     {
-      "epoch": 1.92,
-      "learning_rate": 3.5346820909043585e-05,
-      "loss": 0.2864,
+      "epoch": 1.9636357083648681,
+      "grad_norm": 0.1756594032049179,
+      "learning_rate": 3.465409853433399e-05,
+      "loss": 0.39,
       "step": 54485
     },
     {
-      "epoch": 1.92,
-      "learning_rate": 3.53442275519996e-05,
-      "loss": 0.289,
+      "epoch": 1.9638159080260928,
+      "grad_norm": 0.1727340817451477,
+      "learning_rate": 3.465140669406548e-05,
+      "loss": 0.4083,
       "step": 54490
     },
     {
-      "epoch": 1.92,
-      "learning_rate": 3.534163406064264e-05,
-      "loss": 0.3057,
+      "epoch": 1.9639961076873176,
+      "grad_norm": 0.16905559599399567,
+      "learning_rate": 3.464871472229843e-05,
+      "loss": 0.4025,
       "step": 54495
     },
     {
-      "epoch": 1.92,
-      "learning_rate": 3.533904043500637e-05,
-      "loss": 0.293,
+      "epoch": 1.9641763073485423,
+      "grad_norm": 0.1733274906873703,
+      "learning_rate": 3.464602261906951e-05,
+      "loss": 0.4237,
       "step": 54500
     },
     {
-      "epoch": 1.92,
-      "eval_loss": 0.27375850081443787,
-      "eval_runtime": 10.5429,
-      "eval_samples_per_second": 9.485,
-      "eval_steps_per_second": 9.485,
+      "epoch": 1.9641763073485423,
+      "eval_loss": 0.4376771152019501,
+      "eval_runtime": 3.5433,
+      "eval_samples_per_second": 28.223,
+      "eval_steps_per_second": 7.056,
       "step": 54500
     },
     {
-      "epoch": 1.92,
-      "learning_rate": 3.533644667512448e-05,
-      "loss": 0.3125,
+      "epoch": 1.9643565070097668,
+      "grad_norm": 0.19322708249092102,
+      "learning_rate": 3.4643330384415396e-05,
+      "loss": 0.4047,
       "step": 54505
     },
     {
-      "epoch": 1.92,
-      "learning_rate": 3.5333852781030635e-05,
-      "loss": 0.3063,
+      "epoch": 1.9645367066709913,
+      "grad_norm": 0.18007676303386688,
+      "learning_rate": 3.464063801837277e-05,
+      "loss": 0.4018,
       "step": 54510
     },
     {
-      "epoch": 1.92,
-      "learning_rate": 3.5331258752758536e-05,
-      "loss": 0.2904,
+      "epoch": 1.964716906332216,
+      "grad_norm": 0.17371296882629395,
+      "learning_rate": 3.463794552097832e-05,
+      "loss": 0.3826,
       "step": 54515
     },
     {
-      "epoch": 1.92,
-      "learning_rate": 3.5328664590341836e-05,
-      "loss": 0.3004,
+      "epoch": 1.9648971059934408,
+      "grad_norm": 0.1630249321460724,
+      "learning_rate": 3.463525289226873e-05,
+      "loss": 0.3994,
       "step": 54520
     },
     {
-      "epoch": 1.92,
-      "learning_rate": 3.532607029381424e-05,
-      "loss": 0.2887,
+      "epoch": 1.9650773056546653,
+      "grad_norm": 0.15053841471672058,
+      "learning_rate": 3.463256013228068e-05,
+      "loss": 0.385,
       "step": 54525
     },
     {
-      "epoch": 1.92,
-      "learning_rate": 3.532347586320942e-05,
-      "loss": 0.2682,
+      "epoch": 1.96525750531589,
+      "grad_norm": 0.19676600396633148,
+      "learning_rate": 3.462986724105087e-05,
+      "loss": 0.3725,
       "step": 54530
     },
     {
-      "epoch": 1.92,
-      "learning_rate": 3.532088129856107e-05,
-      "loss": 0.2665,
+      "epoch": 1.9654377049771146,
+      "grad_norm": 0.1895761638879776,
+      "learning_rate": 3.462717421861597e-05,
+      "loss": 0.4309,
       "step": 54535
     },
     {
-      "epoch": 1.92,
-      "learning_rate": 3.531828659990289e-05,
-      "loss": 0.3099,
+      "epoch": 1.9656179046383393,
+      "grad_norm": 0.19437263906002045,
+      "learning_rate": 3.46244810650127e-05,
+      "loss": 0.3823,
       "step": 54540
     },
     {
-      "epoch": 1.92,
-      "learning_rate": 3.531569176726855e-05,
-      "loss": 0.2926,
+      "epoch": 1.965798104299564,
+      "grad_norm": 0.19000405073165894,
+      "learning_rate": 3.462178778027774e-05,
+      "loss": 0.4287,
       "step": 54545
     },
     {
-      "epoch": 1.92,
-      "learning_rate": 3.5313096800691756e-05,
-      "loss": 0.2811,
+      "epoch": 1.9659783039607885,
+      "grad_norm": 0.22047989070415497,
+      "learning_rate": 3.4619094364447776e-05,
+      "loss": 0.4186,
       "step": 54550
     },
     {
-      "epoch": 1.92,
-      "learning_rate": 3.5310501700206186e-05,
-      "loss": 0.2583,
+      "epoch": 1.966158503622013,
+      "grad_norm": 0.22895020246505737,
+      "learning_rate": 3.461640081755951e-05,
+      "loss": 0.4308,
       "step": 54555
     },
     {
-      "epoch": 1.92,
-      "learning_rate": 3.5307906465845554e-05,
-      "loss": 0.314,
+      "epoch": 1.9663387032832378,
+      "grad_norm": 0.1648799479007721,
+      "learning_rate": 3.4613707139649654e-05,
+      "loss": 0.4343,
       "step": 54560
     },
     {
-      "epoch": 1.92,
-      "learning_rate": 3.5305311097643545e-05,
-      "loss": 0.3002,
+      "epoch": 1.9665189029444625,
+      "grad_norm": 0.22859375178813934,
+      "learning_rate": 3.46110133307549e-05,
+      "loss": 0.4118,
       "step": 54565
     },
     {
-      "epoch": 1.92,
-      "learning_rate": 3.530271559563387e-05,
-      "loss": 0.32,
+      "epoch": 1.9666991026056873,
+      "grad_norm": 0.14714932441711426,
+      "learning_rate": 3.4608319390911937e-05,
+      "loss": 0.4129,
       "step": 54570
     },
     {
-      "epoch": 1.92,
-      "learning_rate": 3.530011995985022e-05,
-      "loss": 0.2996,
+      "epoch": 1.9668793022669118,
+      "grad_norm": 0.16554409265518188,
+      "learning_rate": 3.4605625320157496e-05,
+      "loss": 0.4315,
       "step": 54575
     },
     {
-      "epoch": 1.92,
-      "learning_rate": 3.52975241903263e-05,
-      "loss": 0.2864,
+      "epoch": 1.9670595019281363,
+      "grad_norm": 0.1864306479692459,
+      "learning_rate": 3.460293111852827e-05,
+      "loss": 0.4123,
       "step": 54580
     },
     {
-      "epoch": 1.92,
-      "learning_rate": 3.529492828709582e-05,
-      "loss": 0.2945,
+      "epoch": 1.967239701589361,
+      "grad_norm": 0.22947263717651367,
+      "learning_rate": 3.460023678606096e-05,
+      "loss": 0.3806,
       "step": 54585
     },
     {
-      "epoch": 1.92,
-      "learning_rate": 3.5292332250192474e-05,
-      "loss": 0.2948,
+      "epoch": 1.9674199012505857,
+      "grad_norm": 0.19999592006206512,
+      "learning_rate": 3.459754232279228e-05,
+      "loss": 0.3781,
       "step": 54590
     },
     {
-      "epoch": 1.92,
-      "learning_rate": 3.528973607964997e-05,
-      "loss": 0.3156,
+      "epoch": 1.9676001009118103,
+      "grad_norm": 0.15868158638477325,
+      "learning_rate": 3.459484772875895e-05,
+      "loss": 0.4076,
       "step": 54595
     },
     {
-      "epoch": 1.92,
-      "learning_rate": 3.5287139775502034e-05,
-      "loss": 0.2755,
+      "epoch": 1.9677803005730348,
+      "grad_norm": 0.18088065087795258,
+      "learning_rate": 3.459215300399768e-05,
+      "loss": 0.3592,
       "step": 54600
     },
     {
-      "epoch": 1.92,
-      "learning_rate": 3.528454333778237e-05,
-      "loss": 0.2495,
+      "epoch": 1.9679605002342595,
+      "grad_norm": 0.19539353251457214,
+      "learning_rate": 3.4589458148545174e-05,
+      "loss": 0.4784,
       "step": 54605
     },
     {
-      "epoch": 1.92,
-      "learning_rate": 3.528194676652468e-05,
-      "loss": 0.2742,
+      "epoch": 1.9681406998954842,
+      "grad_norm": 0.1806805580854416,
+      "learning_rate": 3.458676316243816e-05,
+      "loss": 0.3996,
       "step": 54610
     },
     {
-      "epoch": 1.92,
-      "learning_rate": 3.527935006176269e-05,
-      "loss": 0.2778,
+      "epoch": 1.968320899556709,
+      "grad_norm": 0.22320657968521118,
+      "learning_rate": 3.458406804571335e-05,
+      "loss": 0.3974,
       "step": 54615
     },
     {
-      "epoch": 1.92,
-      "learning_rate": 3.527675322353011e-05,
-      "loss": 0.2977,
+      "epoch": 1.9685010992179335,
+      "grad_norm": 0.21854592859745026,
+      "learning_rate": 3.458137279840749e-05,
+      "loss": 0.3921,
       "step": 54620
     },
     {
-      "epoch": 1.92,
-      "learning_rate": 3.527415625186067e-05,
-      "loss": 0.2769,
+      "epoch": 1.968681298879158,
+      "grad_norm": 0.20072756707668304,
+      "learning_rate": 3.457867742055725e-05,
+      "loss": 0.3974,
       "step": 54625
     },
     {
-      "epoch": 1.92,
-      "learning_rate": 3.527155914678808e-05,
-      "loss": 0.3208,
+      "epoch": 1.9688614985403827,
+      "grad_norm": 0.19224333763122559,
+      "learning_rate": 3.4575981912199404e-05,
+      "loss": 0.436,
       "step": 54630
     },
     {
-      "epoch": 1.92,
-      "learning_rate": 3.526896190834607e-05,
-      "loss": 0.2947,
+      "epoch": 1.9690416982016075,
+      "grad_norm": 0.1824558526277542,
+      "learning_rate": 3.457328627337065e-05,
+      "loss": 0.4161,
       "step": 54635
     },
     {
-      "epoch": 1.92,
-      "learning_rate": 3.526636453656834e-05,
-      "loss": 0.283,
+      "epoch": 1.969221897862832,
+      "grad_norm": 0.16836610436439514,
+      "learning_rate": 3.457059050410773e-05,
+      "loss": 0.4122,
       "step": 54640
     },
     {
-      "epoch": 1.92,
-      "learning_rate": 3.526376703148864e-05,
-      "loss": 0.2712,
+      "epoch": 1.9694020975240567,
+      "grad_norm": 0.17469918727874756,
+      "learning_rate": 3.456789460444737e-05,
+      "loss": 0.4447,
       "step": 54645
     },
     {
-      "epoch": 1.92,
-      "learning_rate": 3.526116939314069e-05,
-      "loss": 0.2911,
+      "epoch": 1.9695822971852812,
+      "grad_norm": 0.18369092047214508,
+      "learning_rate": 3.456519857442629e-05,
+      "loss": 0.4229,
       "step": 54650
     },
     {
-      "epoch": 1.92,
-      "learning_rate": 3.525857162155822e-05,
-      "loss": 0.2716,
+      "epoch": 1.969762496846506,
+      "grad_norm": 0.20481358468532562,
+      "learning_rate": 3.456250241408123e-05,
+      "loss": 0.3953,
       "step": 54655
     },
     {
-      "epoch": 1.92,
-      "learning_rate": 3.525597371677496e-05,
-      "loss": 0.274,
+      "epoch": 1.9699426965077307,
+      "grad_norm": 0.277506023645401,
+      "learning_rate": 3.455980612344894e-05,
+      "loss": 0.4109,
       "step": 54660
     },
     {
-      "epoch": 1.92,
-      "learning_rate": 3.525337567882464e-05,
-      "loss": 0.2763,
+      "epoch": 1.9701228961689552,
+      "grad_norm": 0.18882182240486145,
+      "learning_rate": 3.455710970256613e-05,
+      "loss": 0.3827,
       "step": 54665
     },
     {
-      "epoch": 1.92,
-      "learning_rate": 3.5250777507740986e-05,
-      "loss": 0.3088,
+      "epoch": 1.9703030958301797,
+      "grad_norm": 0.19233731925487518,
+      "learning_rate": 3.4554413151469553e-05,
+      "loss": 0.4228,
       "step": 54670
     },
     {
-      "epoch": 1.92,
-      "learning_rate": 3.524817920355775e-05,
-      "loss": 0.257,
+      "epoch": 1.9704832954914044,
+      "grad_norm": 0.19764070212841034,
+      "learning_rate": 3.455171647019595e-05,
+      "loss": 0.401,
       "step": 54675
     },
     {
-      "epoch": 1.92,
-      "learning_rate": 3.524558076630865e-05,
-      "loss": 0.2821,
+      "epoch": 1.9706634951526292,
+      "grad_norm": 0.17623180150985718,
+      "learning_rate": 3.454901965878205e-05,
+      "loss": 0.4398,
       "step": 54680
     },
     {
-      "epoch": 1.92,
-      "learning_rate": 3.524298219602745e-05,
-      "loss": 0.2752,
+      "epoch": 1.970843694813854,
+      "grad_norm": 0.16016143560409546,
+      "learning_rate": 3.454632271726461e-05,
+      "loss": 0.4032,
       "step": 54685
     },
     {
-      "epoch": 1.92,
-      "learning_rate": 3.5240383492747874e-05,
-      "loss": 0.293,
+      "epoch": 1.9710238944750784,
+      "grad_norm": 0.1639336496591568,
+      "learning_rate": 3.4543625645680375e-05,
+      "loss": 0.4031,
       "step": 54690
     },
     {
-      "epoch": 1.92,
-      "learning_rate": 3.523778465650366e-05,
-      "loss": 0.2954,
+      "epoch": 1.971204094136303,
+      "grad_norm": 0.1621045023202896,
+      "learning_rate": 3.454092844406609e-05,
+      "loss": 0.3798,
       "step": 54695
     },
     {
-      "epoch": 1.92,
-      "learning_rate": 3.523518568732856e-05,
-      "loss": 0.2765,
+      "epoch": 1.9713842937975277,
+      "grad_norm": 0.19391906261444092,
+      "learning_rate": 3.4538231112458497e-05,
+      "loss": 0.4085,
       "step": 54700
     },
     {
-      "epoch": 1.92,
-      "learning_rate": 3.5232586585256324e-05,
-      "loss": 0.2841,
+      "epoch": 1.9715644934587524,
+      "grad_norm": 0.2154162973165512,
+      "learning_rate": 3.4535533650894356e-05,
+      "loss": 0.4131,
       "step": 54705
     },
     {
-      "epoch": 1.92,
-      "learning_rate": 3.5229987350320695e-05,
-      "loss": 0.2696,
+      "epoch": 1.971744693119977,
+      "grad_norm": 0.16932529211044312,
+      "learning_rate": 3.453283605941041e-05,
+      "loss": 0.3887,
       "step": 54710
     },
     {
-      "epoch": 1.93,
-      "learning_rate": 3.522738798255542e-05,
-      "loss": 0.291,
+      "epoch": 1.9719248927812014,
+      "grad_norm": 0.18449749052524567,
+      "learning_rate": 3.453013833804342e-05,
+      "loss": 0.4192,
       "step": 54715
     },
     {
-      "epoch": 1.93,
-      "learning_rate": 3.522478848199425e-05,
-      "loss": 0.2888,
+      "epoch": 1.9721050924424262,
+      "grad_norm": 0.1401815116405487,
+      "learning_rate": 3.452744048683014e-05,
+      "loss": 0.4199,
       "step": 54720
     },
     {
-      "epoch": 1.93,
-      "learning_rate": 3.5222188848670936e-05,
-      "loss": 0.2965,
+      "epoch": 1.9722852921036509,
+      "grad_norm": 0.17515966296195984,
+      "learning_rate": 3.452474250580734e-05,
+      "loss": 0.3881,
       "step": 54725
     },
     {
-      "epoch": 1.93,
-      "learning_rate": 3.521958908261924e-05,
-      "loss": 0.2867,
+      "epoch": 1.9724654917648756,
+      "grad_norm": 0.19562271237373352,
+      "learning_rate": 3.452204439501175e-05,
+      "loss": 0.4005,
       "step": 54730
     },
     {
-      "epoch": 1.93,
-      "learning_rate": 3.5216989183872925e-05,
-      "loss": 0.2465,
+      "epoch": 1.9726456914261001,
+      "grad_norm": 0.16657692193984985,
+      "learning_rate": 3.451934615448016e-05,
+      "loss": 0.3684,
       "step": 54735
     },
     {
-      "epoch": 1.93,
-      "learning_rate": 3.521438915246572e-05,
-      "loss": 0.2976,
+      "epoch": 1.9728258910873246,
+      "grad_norm": 0.2561399042606354,
+      "learning_rate": 3.451664778424931e-05,
+      "loss": 0.4216,
       "step": 54740
     },
     {
-      "epoch": 1.93,
-      "learning_rate": 3.5211788988431424e-05,
-      "loss": 0.2708,
+      "epoch": 1.9730060907485494,
+      "grad_norm": 0.18860171735286713,
+      "learning_rate": 3.451394928435598e-05,
+      "loss": 0.4218,
       "step": 54745
     },
     {
-      "epoch": 1.93,
-      "learning_rate": 3.520918869180376e-05,
-      "loss": 0.2836,
+      "epoch": 1.9731862904097741,
+      "grad_norm": 0.17932887375354767,
+      "learning_rate": 3.451125065483695e-05,
+      "loss": 0.3872,
       "step": 54750
     },
     {
-      "epoch": 1.93,
-      "learning_rate": 3.520658826261651e-05,
-      "loss": 0.2953,
+      "epoch": 1.9733664900709986,
+      "grad_norm": 0.19358305633068085,
+      "learning_rate": 3.450855189572895e-05,
+      "loss": 0.4098,
       "step": 54755
     },
     {
-      "epoch": 1.93,
-      "learning_rate": 3.520398770090345e-05,
-      "loss": 0.2719,
+      "epoch": 1.9735466897322234,
+      "grad_norm": 0.1724853366613388,
+      "learning_rate": 3.450585300706878e-05,
+      "loss": 0.4266,
       "step": 54760
     },
     {
-      "epoch": 1.93,
-      "learning_rate": 3.5201387006698326e-05,
-      "loss": 0.3359,
+      "epoch": 1.9737268893934479,
+      "grad_norm": 0.17911700904369354,
+      "learning_rate": 3.45031539888932e-05,
+      "loss": 0.417,
       "step": 54765
     },
     {
-      "epoch": 1.93,
-      "learning_rate": 3.519878618003493e-05,
-      "loss": 0.2837,
+      "epoch": 1.9739070890546726,
+      "grad_norm": 0.16080361604690552,
+      "learning_rate": 3.450045484123899e-05,
+      "loss": 0.4101,
       "step": 54770
     },
     {
-      "epoch": 1.93,
-      "learning_rate": 3.5196185220947e-05,
-      "loss": 0.2681,
+      "epoch": 1.9740872887158973,
+      "grad_norm": 0.19026781618595123,
+      "learning_rate": 3.449775556414292e-05,
+      "loss": 0.4418,
       "step": 54775
     },
     {
-      "epoch": 1.93,
-      "learning_rate": 3.5193584129468326e-05,
-      "loss": 0.2896,
+      "epoch": 1.9742674883771218,
+      "grad_norm": 0.16302798688411713,
+      "learning_rate": 3.449505615764177e-05,
+      "loss": 0.4291,
       "step": 54780
     },
     {
-      "epoch": 1.93,
-      "learning_rate": 3.519098290563267e-05,
-      "loss": 0.3084,
+      "epoch": 1.9744476880383464,
+      "grad_norm": 0.13728991150856018,
+      "learning_rate": 3.449235662177233e-05,
+      "loss": 0.4054,
       "step": 54785
     },
     {
-      "epoch": 1.93,
-      "learning_rate": 3.518838154947384e-05,
-      "loss": 0.2752,
+      "epoch": 1.974627887699571,
+      "grad_norm": 0.16925320029258728,
+      "learning_rate": 3.4489656956571345e-05,
+      "loss": 0.3945,
       "step": 54790
     },
     {
-      "epoch": 1.93,
-      "learning_rate": 3.518578006102558e-05,
-      "loss": 0.2897,
+      "epoch": 1.9748080873607958,
+      "grad_norm": 0.20769056677818298,
+      "learning_rate": 3.448695716207564e-05,
+      "loss": 0.416,
       "step": 54795
     },
     {
-      "epoch": 1.93,
-      "learning_rate": 3.518317844032168e-05,
-      "loss": 0.2674,
+      "epoch": 1.9749882870220206,
+      "grad_norm": 0.20202518999576569,
+      "learning_rate": 3.448425723832197e-05,
+      "loss": 0.3943,
       "step": 54800
     },
     {
-      "epoch": 1.93,
-      "learning_rate": 3.5180576687395915e-05,
-      "loss": 0.2786,
+      "epoch": 1.975168486683245,
+      "grad_norm": 0.19320270419120789,
+      "learning_rate": 3.448155718534714e-05,
+      "loss": 0.4378,
       "step": 54805
     },
     {
-      "epoch": 1.93,
-      "learning_rate": 3.517797480228207e-05,
-      "loss": 0.2788,
+      "epoch": 1.9753486863444696,
+      "grad_norm": 0.15490759909152985,
+      "learning_rate": 3.447885700318792e-05,
+      "loss": 0.3863,
       "step": 54810
     },
     {
-      "epoch": 1.93,
-      "learning_rate": 3.517537278501393e-05,
-      "loss": 0.2804,
+      "epoch": 1.9755288860056943,
+      "grad_norm": 0.1758003681898117,
+      "learning_rate": 3.447615669188111e-05,
+      "loss": 0.4029,
       "step": 54815
     },
     {
-      "epoch": 1.93,
-      "learning_rate": 3.517277063562529e-05,
-      "loss": 0.2814,
+      "epoch": 1.975709085666919,
+      "grad_norm": 0.1570878028869629,
+      "learning_rate": 3.44734562514635e-05,
+      "loss": 0.386,
       "step": 54820
     },
     {
-      "epoch": 1.93,
-      "learning_rate": 3.517016835414992e-05,
-      "loss": 0.2579,
+      "epoch": 1.9758892853281436,
+      "grad_norm": 0.15345117449760437,
+      "learning_rate": 3.4470755681971886e-05,
+      "loss": 0.401,
       "step": 54825
     },
     {
-      "epoch": 1.93,
-      "learning_rate": 3.5167565940621614e-05,
-      "loss": 0.2763,
+      "epoch": 1.976069484989368,
+      "grad_norm": 0.201069176197052,
+      "learning_rate": 3.446805498344307e-05,
+      "loss": 0.4219,
       "step": 54830
     },
     {
-      "epoch": 1.93,
-      "learning_rate": 3.516496339507416e-05,
-      "loss": 0.2942,
+      "epoch": 1.9762496846505928,
+      "grad_norm": 0.16482049226760864,
+      "learning_rate": 3.446535415591382e-05,
+      "loss": 0.4091,
       "step": 54835
     },
     {
-      "epoch": 1.93,
-      "learning_rate": 3.5162360717541374e-05,
-      "loss": 0.2744,
+      "epoch": 1.9764298843118175,
+      "grad_norm": 0.19689258933067322,
+      "learning_rate": 3.446265319942096e-05,
+      "loss": 0.3863,
       "step": 54840
     },
     {
-      "epoch": 1.93,
-      "learning_rate": 3.515975790805701e-05,
-      "loss": 0.28,
+      "epoch": 1.9766100839730423,
+      "grad_norm": 0.14975933730602264,
+      "learning_rate": 3.4459952114001284e-05,
+      "loss": 0.426,
       "step": 54845
     },
     {
-      "epoch": 1.93,
-      "learning_rate": 3.51571549666549e-05,
-      "loss": 0.2785,
+      "epoch": 1.9767902836342668,
+      "grad_norm": 0.17771407961845398,
+      "learning_rate": 3.4457250899691586e-05,
+      "loss": 0.4053,
       "step": 54850
     },
     {
-      "epoch": 1.93,
-      "learning_rate": 3.5154551893368824e-05,
-      "loss": 0.2837,
+      "epoch": 1.9769704832954913,
+      "grad_norm": 0.16967308521270752,
+      "learning_rate": 3.4454549556528674e-05,
+      "loss": 0.3714,
       "step": 54855
     },
     {
-      "epoch": 1.93,
-      "learning_rate": 3.5151948688232575e-05,
-      "loss": 0.2921,
+      "epoch": 1.977150682956716,
+      "grad_norm": 0.19800063967704773,
+      "learning_rate": 3.445184808454936e-05,
+      "loss": 0.4534,
       "step": 54860
     },
     {
-      "epoch": 1.93,
-      "learning_rate": 3.5149345351279975e-05,
-      "loss": 0.2606,
+      "epoch": 1.9773308826179408,
+      "grad_norm": 0.19260716438293457,
+      "learning_rate": 3.444914648379045e-05,
+      "loss": 0.4158,
       "step": 54865
     },
     {
-      "epoch": 1.93,
-      "learning_rate": 3.51467418825448e-05,
-      "loss": 0.2747,
+      "epoch": 1.9775110822791653,
+      "grad_norm": 0.18546351790428162,
+      "learning_rate": 3.4446444754288745e-05,
+      "loss": 0.3974,
       "step": 54870
     },
     {
-      "epoch": 1.93,
-      "learning_rate": 3.5144138282060886e-05,
-      "loss": 0.2833,
+      "epoch": 1.9776912819403898,
+      "grad_norm": 0.17617270350456238,
+      "learning_rate": 3.444374289608105e-05,
+      "loss": 0.3895,
       "step": 54875
     },
     {
-      "epoch": 1.93,
-      "learning_rate": 3.514153454986201e-05,
-      "loss": 0.302,
+      "epoch": 1.9778714816016145,
+      "grad_norm": 0.20239382982254028,
+      "learning_rate": 3.444104090920419e-05,
+      "loss": 0.417,
       "step": 54880
     },
     {
-      "epoch": 1.93,
-      "learning_rate": 3.5138930685982e-05,
-      "loss": 0.2866,
+      "epoch": 1.9780516812628393,
+      "grad_norm": 0.15842540562152863,
+      "learning_rate": 3.443833879369499e-05,
+      "loss": 0.4095,
       "step": 54885
     },
     {
-      "epoch": 1.93,
-      "learning_rate": 3.513632669045464e-05,
-      "loss": 0.2823,
+      "epoch": 1.978231880924064,
+      "grad_norm": 0.1709548830986023,
+      "learning_rate": 3.443563654959024e-05,
+      "loss": 0.4075,
       "step": 54890
     },
     {
-      "epoch": 1.93,
-      "learning_rate": 3.513372256331378e-05,
-      "loss": 0.2745,
+      "epoch": 1.9784120805852885,
+      "grad_norm": 0.20768848061561584,
+      "learning_rate": 3.4432934176926765e-05,
+      "loss": 0.427,
       "step": 54895
     },
     {
-      "epoch": 1.93,
-      "learning_rate": 3.513111830459319e-05,
-      "loss": 0.2888,
+      "epoch": 1.978592280246513,
+      "grad_norm": 0.16096894443035126,
+      "learning_rate": 3.443023167574139e-05,
+      "loss": 0.4047,
       "step": 54900
     },
     {
-      "epoch": 1.93,
-      "learning_rate": 3.512851391432672e-05,
-      "loss": 0.2857,
+      "epoch": 1.9787724799077377,
+      "grad_norm": 0.17708556354045868,
+      "learning_rate": 3.4427529046070936e-05,
+      "loss": 0.4069,
       "step": 54905
     },
     {
-      "epoch": 1.93,
-      "learning_rate": 3.512590939254816e-05,
-      "loss": 0.2964,
+      "epoch": 1.9789526795689625,
+      "grad_norm": 0.1786879301071167,
+      "learning_rate": 3.442482628795223e-05,
+      "loss": 0.3877,
       "step": 54910
     },
     {
-      "epoch": 1.93,
-      "learning_rate": 3.512330473929135e-05,
-      "loss": 0.2619,
+      "epoch": 1.979132879230187,
+      "grad_norm": 0.2042560875415802,
+      "learning_rate": 3.442212340142209e-05,
+      "loss": 0.4307,
       "step": 54915
     },
     {
-      "epoch": 1.93,
-      "learning_rate": 3.51206999545901e-05,
-      "loss": 0.2906,
+      "epoch": 1.9793130788914117,
+      "grad_norm": 0.16689689457416534,
+      "learning_rate": 3.441942038651733e-05,
+      "loss": 0.4006,
       "step": 54920
     },
     {
-      "epoch": 1.93,
-      "learning_rate": 3.5118095038478225e-05,
-      "loss": 0.2607,
+      "epoch": 1.9794932785526362,
+      "grad_norm": 0.16733959317207336,
+      "learning_rate": 3.44167172432748e-05,
+      "loss": 0.4051,
       "step": 54925
     },
     {
-      "epoch": 1.93,
-      "learning_rate": 3.511548999098956e-05,
-      "loss": 0.2805,
+      "epoch": 1.979673478213861,
+      "grad_norm": 0.18891684710979462,
+      "learning_rate": 3.4414013971731323e-05,
+      "loss": 0.4374,
       "step": 54930
     },
     {
-      "epoch": 1.93,
-      "learning_rate": 3.511288481215792e-05,
-      "loss": 0.263,
+      "epoch": 1.9798536778750857,
+      "grad_norm": 0.18657353520393372,
+      "learning_rate": 3.441131057192373e-05,
+      "loss": 0.4465,
       "step": 54935
     },
     {
-      "epoch": 1.93,
-      "learning_rate": 3.5110279502017126e-05,
-      "loss": 0.2726,
+      "epoch": 1.9800338775363102,
+      "grad_norm": 0.1951577365398407,
+      "learning_rate": 3.440860704388884e-05,
+      "loss": 0.3765,
       "step": 54940
     },
     {
-      "epoch": 1.93,
-      "learning_rate": 3.510767406060103e-05,
-      "loss": 0.285,
+      "epoch": 1.9802140771975347,
+      "grad_norm": 0.21163040399551392,
+      "learning_rate": 3.440590338766351e-05,
+      "loss": 0.4099,
       "step": 54945
     },
     {
-      "epoch": 1.93,
-      "learning_rate": 3.5105068487943436e-05,
-      "loss": 0.2792,
+      "epoch": 1.9803942768587595,
+      "grad_norm": 0.19105908274650574,
+      "learning_rate": 3.4403199603284567e-05,
+      "loss": 0.4014,
       "step": 54950
     },
     {
-      "epoch": 1.93,
-      "learning_rate": 3.510246278407821e-05,
-      "loss": 0.2855,
+      "epoch": 1.9805744765199842,
+      "grad_norm": 0.18041996657848358,
+      "learning_rate": 3.440049569078885e-05,
+      "loss": 0.4271,
       "step": 54955
     },
     {
-      "epoch": 1.93,
-      "learning_rate": 3.509985694903913e-05,
-      "loss": 0.287,
+      "epoch": 1.980754676181209,
+      "grad_norm": 0.19663773477077484,
+      "learning_rate": 3.4397791650213193e-05,
+      "loss": 0.4073,
       "step": 54960
     },
     {
-      "epoch": 1.93,
-      "learning_rate": 3.509725098286009e-05,
-      "loss": 0.2961,
+      "epoch": 1.9809348758424334,
+      "grad_norm": 0.17666995525360107,
+      "learning_rate": 3.439508748159445e-05,
+      "loss": 0.3892,
       "step": 54965
     },
     {
-      "epoch": 1.93,
-      "learning_rate": 3.5094644885574886e-05,
-      "loss": 0.2824,
+      "epoch": 1.981115075503658,
+      "grad_norm": 0.14386975765228271,
+      "learning_rate": 3.4392383184969464e-05,
+      "loss": 0.3762,
       "step": 54970
     },
     {
-      "epoch": 1.93,
-      "learning_rate": 3.5092038657217376e-05,
-      "loss": 0.2817,
+      "epoch": 1.9812952751648827,
+      "grad_norm": 0.19409483671188354,
+      "learning_rate": 3.438967876037507e-05,
+      "loss": 0.375,
       "step": 54975
     },
     {
-      "epoch": 1.93,
-      "learning_rate": 3.508943229782139e-05,
-      "loss": 0.2871,
+      "epoch": 1.9814754748261074,
+      "grad_norm": 0.22416892647743225,
+      "learning_rate": 3.438697420784812e-05,
+      "loss": 0.4125,
       "step": 54980
     },
     {
-      "epoch": 1.93,
-      "learning_rate": 3.508682580742079e-05,
-      "loss": 0.2864,
+      "epoch": 1.981655674487332,
+      "grad_norm": 0.20359428226947784,
+      "learning_rate": 3.438426952742546e-05,
+      "loss": 0.4154,
       "step": 54985
     },
     {
-      "epoch": 1.93,
-      "learning_rate": 3.508421918604938e-05,
-      "loss": 0.2935,
+      "epoch": 1.9818358741485564,
+      "grad_norm": 0.20030748844146729,
+      "learning_rate": 3.4381564719143947e-05,
+      "loss": 0.3902,
       "step": 54990
     },
     {
-      "epoch": 1.93,
-      "learning_rate": 3.508161243374105e-05,
-      "loss": 0.2825,
+      "epoch": 1.9820160738097812,
+      "grad_norm": 0.15663334727287292,
+      "learning_rate": 3.437885978304043e-05,
+      "loss": 0.4205,
       "step": 54995
     },
     {
-      "epoch": 1.94,
-      "learning_rate": 3.5079005550529614e-05,
-      "loss": 0.2652,
+      "epoch": 1.982196273471006,
+      "grad_norm": 0.17842374742031097,
+      "learning_rate": 3.437615471915177e-05,
+      "loss": 0.4141,
       "step": 55000
     },
     {
-      "epoch": 1.94,
-      "eval_loss": 0.2729741930961609,
-      "eval_runtime": 10.5572,
-      "eval_samples_per_second": 9.472,
-      "eval_steps_per_second": 9.472,
+      "epoch": 1.982196273471006,
+      "eval_loss": 0.43720561265945435,
+      "eval_runtime": 3.5352,
+      "eval_samples_per_second": 28.287,
+      "eval_steps_per_second": 7.072,
       "step": 55000
     },
     {
-      "epoch": 1.94,
-      "learning_rate": 3.507639853644893e-05,
-      "loss": 0.2845,
+      "epoch": 1.9823764731322306,
+      "grad_norm": 0.155527725815773,
+      "learning_rate": 3.437344952751481e-05,
+      "loss": 0.4354,
       "step": 55005
     },
     {
-      "epoch": 1.94,
-      "learning_rate": 3.507379139153287e-05,
-      "loss": 0.2745,
+      "epoch": 1.9825566727934552,
+      "grad_norm": 0.16779151558876038,
+      "learning_rate": 3.4370744208166424e-05,
+      "loss": 0.4442,
       "step": 55010
     },
     {
-      "epoch": 1.94,
-      "learning_rate": 3.5071184115815255e-05,
-      "loss": 0.2683,
+      "epoch": 1.9827368724546797,
+      "grad_norm": 0.25572511553764343,
+      "learning_rate": 3.436803876114346e-05,
+      "loss": 0.4068,
       "step": 55015
     },
     {
-      "epoch": 1.94,
-      "learning_rate": 3.506857670932996e-05,
-      "loss": 0.3008,
+      "epoch": 1.9829170721159044,
+      "grad_norm": 0.1537223905324936,
+      "learning_rate": 3.436533318648277e-05,
+      "loss": 0.4256,
       "step": 55020
     },
     {
-      "epoch": 1.94,
-      "learning_rate": 3.506596917211082e-05,
-      "loss": 0.2836,
+      "epoch": 1.9830972717771291,
+      "grad_norm": 0.17157311737537384,
+      "learning_rate": 3.436262748422124e-05,
+      "loss": 0.3851,
       "step": 55025
     },
     {
-      "epoch": 1.94,
-      "learning_rate": 3.5063361504191714e-05,
-      "loss": 0.2907,
+      "epoch": 1.9832774714383536,
+      "grad_norm": 0.21722643077373505,
+      "learning_rate": 3.435992165439572e-05,
+      "loss": 0.4225,
       "step": 55030
     },
     {
-      "epoch": 1.94,
-      "learning_rate": 3.506075370560648e-05,
-      "loss": 0.2794,
+      "epoch": 1.9834576710995784,
+      "grad_norm": 0.20483043789863586,
+      "learning_rate": 3.435721569704308e-05,
+      "loss": 0.4113,
       "step": 55035
     },
     {
-      "epoch": 1.94,
-      "learning_rate": 3.5058145776389e-05,
-      "loss": 0.2697,
+      "epoch": 1.983637870760803,
+      "grad_norm": 0.2097228616476059,
+      "learning_rate": 3.4354509612200194e-05,
+      "loss": 0.407,
       "step": 55040
     },
     {
-      "epoch": 1.94,
-      "learning_rate": 3.505553771657312e-05,
-      "loss": 0.2841,
+      "epoch": 1.9838180704220276,
+      "grad_norm": 0.1592007428407669,
+      "learning_rate": 3.435180339990391e-05,
+      "loss": 0.3818,
       "step": 55045
     },
     {
-      "epoch": 1.94,
-      "learning_rate": 3.505292952619272e-05,
-      "loss": 0.2808,
+      "epoch": 1.9839982700832524,
+      "grad_norm": 0.1988215148448944,
+      "learning_rate": 3.4349097060191135e-05,
+      "loss": 0.3992,
       "step": 55050
     },
     {
-      "epoch": 1.94,
-      "learning_rate": 3.5050321205281636e-05,
-      "loss": 0.2969,
+      "epoch": 1.9841784697444769,
+      "grad_norm": 0.19926537573337555,
+      "learning_rate": 3.434639059309871e-05,
+      "loss": 0.3907,
       "step": 55055
     },
     {
-      "epoch": 1.94,
-      "learning_rate": 3.504771275387377e-05,
-      "loss": 0.2951,
+      "epoch": 1.9843586694057014,
+      "grad_norm": 0.1831171214580536,
+      "learning_rate": 3.434368399866352e-05,
+      "loss": 0.4176,
       "step": 55060
     },
     {
-      "epoch": 1.94,
-      "learning_rate": 3.504510417200297e-05,
-      "loss": 0.2714,
+      "epoch": 1.9845388690669261,
+      "grad_norm": 0.17820048332214355,
+      "learning_rate": 3.434097727692245e-05,
+      "loss": 0.4282,
       "step": 55065
     },
     {
-      "epoch": 1.94,
-      "learning_rate": 3.504249545970312e-05,
-      "loss": 0.2777,
+      "epoch": 1.9847190687281508,
+      "grad_norm": 0.21258282661437988,
+      "learning_rate": 3.433827042791238e-05,
+      "loss": 0.4548,
       "step": 55070
     },
     {
-      "epoch": 1.94,
-      "learning_rate": 3.503988661700807e-05,
-      "loss": 0.2983,
+      "epoch": 1.9848992683893756,
+      "grad_norm": 0.17607107758522034,
+      "learning_rate": 3.433556345167017e-05,
+      "loss": 0.4106,
       "step": 55075
     },
     {
-      "epoch": 1.94,
-      "learning_rate": 3.503727764395172e-05,
-      "loss": 0.2766,
+      "epoch": 1.9850794680506,
+      "grad_norm": 0.18194575607776642,
+      "learning_rate": 3.4332856348232725e-05,
+      "loss": 0.4119,
       "step": 55080
     },
     {
-      "epoch": 1.94,
-      "learning_rate": 3.503466854056793e-05,
-      "loss": 0.2603,
+      "epoch": 1.9852596677118246,
+      "grad_norm": 0.16195902228355408,
+      "learning_rate": 3.433014911763691e-05,
+      "loss": 0.3892,
       "step": 55085
     },
     {
-      "epoch": 1.94,
-      "learning_rate": 3.503205930689059e-05,
-      "loss": 0.3038,
+      "epoch": 1.9854398673730493,
+      "grad_norm": 0.17873816192150116,
+      "learning_rate": 3.432744175991963e-05,
+      "loss": 0.406,
       "step": 55090
     },
     {
-      "epoch": 1.94,
-      "learning_rate": 3.5029449942953555e-05,
-      "loss": 0.3096,
+      "epoch": 1.985620067034274,
+      "grad_norm": 0.17679697275161743,
+      "learning_rate": 3.432473427511776e-05,
+      "loss": 0.4588,
       "step": 55095
     },
     {
-      "epoch": 1.94,
-      "learning_rate": 3.5026840448790735e-05,
-      "loss": 0.2746,
+      "epoch": 1.9858002666954986,
+      "grad_norm": 0.1970539093017578,
+      "learning_rate": 3.4322026663268186e-05,
+      "loss": 0.3898,
       "step": 55100
     },
     {
-      "epoch": 1.94,
-      "learning_rate": 3.5024230824436e-05,
-      "loss": 0.2508,
+      "epoch": 1.985980466356723,
+      "grad_norm": 0.2121034860610962,
+      "learning_rate": 3.4319318924407806e-05,
+      "loss": 0.4556,
       "step": 55105
     },
     {
-      "epoch": 1.94,
-      "learning_rate": 3.502162106992322e-05,
-      "loss": 0.2831,
+      "epoch": 1.9861606660179478,
+      "grad_norm": 0.1846599280834198,
+      "learning_rate": 3.431661105857351e-05,
+      "loss": 0.4054,
       "step": 55110
     },
     {
-      "epoch": 1.94,
-      "learning_rate": 3.501901118528631e-05,
-      "loss": 0.2803,
+      "epoch": 1.9863408656791726,
+      "grad_norm": 0.17753110826015472,
+      "learning_rate": 3.4313903065802187e-05,
+      "loss": 0.3984,
       "step": 55115
     },
     {
-      "epoch": 1.94,
-      "learning_rate": 3.501640117055913e-05,
-      "loss": 0.2564,
+      "epoch": 1.9865210653403973,
+      "grad_norm": 0.1930995136499405,
+      "learning_rate": 3.431119494613075e-05,
+      "loss": 0.4278,
       "step": 55120
     },
     {
-      "epoch": 1.94,
-      "learning_rate": 3.50137910257756e-05,
-      "loss": 0.262,
+      "epoch": 1.9867012650016218,
+      "grad_norm": 0.17854326963424683,
+      "learning_rate": 3.430848669959607e-05,
+      "loss": 0.406,
       "step": 55125
     },
     {
-      "epoch": 1.94,
-      "learning_rate": 3.501118075096958e-05,
-      "loss": 0.2641,
+      "epoch": 1.9868814646628463,
+      "grad_norm": 0.16902780532836914,
+      "learning_rate": 3.430577832623507e-05,
+      "loss": 0.3918,
       "step": 55130
     },
     {
-      "epoch": 1.94,
-      "learning_rate": 3.500857034617497e-05,
-      "loss": 0.2845,
+      "epoch": 1.987061664324071,
+      "grad_norm": 0.23383301496505737,
+      "learning_rate": 3.4303069826084646e-05,
+      "loss": 0.3994,
       "step": 55135
     },
     {
-      "epoch": 1.94,
-      "learning_rate": 3.500595981142568e-05,
-      "loss": 0.2981,
+      "epoch": 1.9872418639852958,
+      "grad_norm": 0.19351869821548462,
+      "learning_rate": 3.430036119918168e-05,
+      "loss": 0.4034,
       "step": 55140
     },
     {
-      "epoch": 1.94,
-      "learning_rate": 3.50033491467556e-05,
-      "loss": 0.2878,
+      "epoch": 1.9874220636465203,
+      "grad_norm": 0.19689901173114777,
+      "learning_rate": 3.429765244556311e-05,
+      "loss": 0.4055,
       "step": 55145
     },
     {
-      "epoch": 1.94,
-      "learning_rate": 3.500073835219861e-05,
-      "loss": 0.2686,
+      "epoch": 1.987602263307745,
+      "grad_norm": 0.19208216667175293,
+      "learning_rate": 3.429494356526581e-05,
+      "loss": 0.3989,
       "step": 55150
     },
     {
-      "epoch": 1.94,
-      "learning_rate": 3.4998127427788644e-05,
-      "loss": 0.28,
+      "epoch": 1.9877824629689695,
+      "grad_norm": 0.19520142674446106,
+      "learning_rate": 3.4292234558326724e-05,
+      "loss": 0.3776,
       "step": 55155
     },
     {
-      "epoch": 1.94,
-      "learning_rate": 3.4995516373559566e-05,
-      "loss": 0.2787,
+      "epoch": 1.9879626626301943,
+      "grad_norm": 0.22114428877830505,
+      "learning_rate": 3.4289525424782726e-05,
+      "loss": 0.422,
       "step": 55160
     },
     {
-      "epoch": 1.94,
-      "learning_rate": 3.49929051895453e-05,
-      "loss": 0.2695,
+      "epoch": 1.988142862291419,
+      "grad_norm": 0.16735602915287018,
+      "learning_rate": 3.428681616467075e-05,
+      "loss": 0.4248,
       "step": 55165
     },
     {
-      "epoch": 1.94,
-      "learning_rate": 3.499029387577974e-05,
-      "loss": 0.2726,
+      "epoch": 1.9883230619526435,
+      "grad_norm": 0.18746241927146912,
+      "learning_rate": 3.4284106778027696e-05,
+      "loss": 0.4177,
       "step": 55170
     },
     {
-      "epoch": 1.94,
-      "learning_rate": 3.4987682432296805e-05,
-      "loss": 0.2867,
+      "epoch": 1.988503261613868,
+      "grad_norm": 0.24923117458820343,
+      "learning_rate": 3.4281397264890484e-05,
+      "loss": 0.4006,
       "step": 55175
     },
     {
-      "epoch": 1.94,
-      "learning_rate": 3.498507085913039e-05,
-      "loss": 0.2906,
+      "epoch": 1.9886834612750928,
+      "grad_norm": 0.147854283452034,
+      "learning_rate": 3.427868762529604e-05,
+      "loss": 0.4182,
       "step": 55180
     },
     {
-      "epoch": 1.94,
-      "learning_rate": 3.4982459156314415e-05,
-      "loss": 0.2819,
+      "epoch": 1.9888636609363175,
+      "grad_norm": 0.14005811512470245,
+      "learning_rate": 3.427597785928126e-05,
+      "loss": 0.3873,
       "step": 55185
     },
     {
-      "epoch": 1.94,
-      "learning_rate": 3.497984732388278e-05,
-      "loss": 0.2734,
+      "epoch": 1.9890438605975422,
+      "grad_norm": 0.18232989311218262,
+      "learning_rate": 3.4273267966883094e-05,
+      "loss": 0.374,
       "step": 55190
     },
     {
-      "epoch": 1.94,
-      "learning_rate": 3.497723536186941e-05,
-      "loss": 0.2846,
+      "epoch": 1.9892240602587667,
+      "grad_norm": 0.18193449079990387,
+      "learning_rate": 3.427055794813844e-05,
+      "loss": 0.3967,
       "step": 55195
     },
     {
-      "epoch": 1.94,
-      "learning_rate": 3.4974623270308215e-05,
-      "loss": 0.3048,
+      "epoch": 1.9894042599199913,
+      "grad_norm": 0.1848289668560028,
+      "learning_rate": 3.426784780308423e-05,
+      "loss": 0.3852,
       "step": 55200
     },
     {
-      "epoch": 1.94,
-      "learning_rate": 3.497201104923311e-05,
-      "loss": 0.3095,
+      "epoch": 1.989584459581216,
+      "grad_norm": 0.2115050107240677,
+      "learning_rate": 3.426513753175738e-05,
+      "loss": 0.4434,
       "step": 55205
     },
     {
-      "epoch": 1.94,
-      "learning_rate": 3.496939869867801e-05,
-      "loss": 0.2599,
+      "epoch": 1.9897646592424407,
+      "grad_norm": 0.20298829674720764,
+      "learning_rate": 3.426242713419483e-05,
+      "loss": 0.4076,
       "step": 55210
     },
     {
-      "epoch": 1.94,
-      "learning_rate": 3.4966786218676836e-05,
-      "loss": 0.2875,
+      "epoch": 1.9899448589036652,
+      "grad_norm": 0.20630566775798798,
+      "learning_rate": 3.425971661043351e-05,
+      "loss": 0.434,
       "step": 55215
     },
     {
-      "epoch": 1.94,
-      "learning_rate": 3.496417360926352e-05,
-      "loss": 0.2867,
+      "epoch": 1.9901250585648897,
+      "grad_norm": 0.1815614402294159,
+      "learning_rate": 3.425700596051033e-05,
+      "loss": 0.4045,
       "step": 55220
     },
     {
-      "epoch": 1.94,
-      "learning_rate": 3.496156087047197e-05,
-      "loss": 0.2812,
+      "epoch": 1.9903052582261145,
+      "grad_norm": 0.1584119349718094,
+      "learning_rate": 3.425429518446225e-05,
+      "loss": 0.4018,
       "step": 55225
     },
     {
-      "epoch": 1.94,
-      "learning_rate": 3.495894800233612e-05,
-      "loss": 0.2859,
+      "epoch": 1.9904854578873392,
+      "grad_norm": 0.17491492629051208,
+      "learning_rate": 3.425158428232618e-05,
+      "loss": 0.4208,
       "step": 55230
     },
     {
-      "epoch": 1.94,
-      "learning_rate": 3.495633500488989e-05,
-      "loss": 0.2807,
+      "epoch": 1.990665657548564,
+      "grad_norm": 0.16318289935588837,
+      "learning_rate": 3.4248873254139066e-05,
+      "loss": 0.4531,
       "step": 55235
     },
     {
-      "epoch": 1.94,
-      "learning_rate": 3.495372187816722e-05,
-      "loss": 0.2775,
+      "epoch": 1.9908458572097885,
+      "grad_norm": 0.18993204832077026,
+      "learning_rate": 3.424616209993785e-05,
+      "loss": 0.3925,
       "step": 55240
     },
     {
-      "epoch": 1.94,
-      "learning_rate": 3.4951108622202024e-05,
-      "loss": 0.2604,
+      "epoch": 1.991026056871013,
+      "grad_norm": 0.20490875840187073,
+      "learning_rate": 3.4243450819759464e-05,
+      "loss": 0.3857,
       "step": 55245
     },
     {
-      "epoch": 1.94,
-      "learning_rate": 3.494849523702824e-05,
-      "loss": 0.2818,
+      "epoch": 1.9912062565322377,
+      "grad_norm": 0.22841264307498932,
+      "learning_rate": 3.424073941364085e-05,
+      "loss": 0.4436,
       "step": 55250
     },
     {
-      "epoch": 1.94,
-      "learning_rate": 3.49458817226798e-05,
-      "loss": 0.2879,
+      "epoch": 1.9913864561934624,
+      "grad_norm": 0.18176591396331787,
+      "learning_rate": 3.423802788161895e-05,
+      "loss": 0.4367,
       "step": 55255
     },
     {
-      "epoch": 1.94,
-      "learning_rate": 3.494326807919065e-05,
-      "loss": 0.281,
+      "epoch": 1.991566655854687,
+      "grad_norm": 0.20801039040088654,
+      "learning_rate": 3.4235316223730706e-05,
+      "loss": 0.4058,
       "step": 55260
     },
     {
-      "epoch": 1.94,
-      "learning_rate": 3.4940654306594705e-05,
-      "loss": 0.2658,
+      "epoch": 1.9917468555159117,
+      "grad_norm": 0.19156023859977722,
+      "learning_rate": 3.423260444001307e-05,
+      "loss": 0.4368,
       "step": 55265
     },
     {
-      "epoch": 1.94,
-      "learning_rate": 3.493804040492592e-05,
-      "loss": 0.3026,
+      "epoch": 1.9919270551771362,
+      "grad_norm": 0.1914074718952179,
+      "learning_rate": 3.422989253050298e-05,
+      "loss": 0.4253,
       "step": 55270
     },
     {
-      "epoch": 1.94,
-      "learning_rate": 3.493542637421823e-05,
-      "loss": 0.2766,
+      "epoch": 1.992107254838361,
+      "grad_norm": 0.15868034958839417,
+      "learning_rate": 3.42271804952374e-05,
+      "loss": 0.3822,
       "step": 55275
     },
     {
-      "epoch": 1.94,
-      "learning_rate": 3.4932812214505576e-05,
-      "loss": 0.3027,
+      "epoch": 1.9922874544995857,
+      "grad_norm": 0.20103418827056885,
+      "learning_rate": 3.4224468334253265e-05,
+      "loss": 0.4241,
       "step": 55280
     },
     {
-      "epoch": 1.95,
-      "learning_rate": 3.49301979258219e-05,
-      "loss": 0.2719,
+      "epoch": 1.9924676541608102,
+      "grad_norm": 0.22683289647102356,
+      "learning_rate": 3.422175604758754e-05,
+      "loss": 0.4063,
       "step": 55285
     },
     {
-      "epoch": 1.95,
-      "learning_rate": 3.4927583508201155e-05,
-      "loss": 0.2899,
+      "epoch": 1.9926478538220347,
+      "grad_norm": 0.20226216316223145,
+      "learning_rate": 3.4219043635277173e-05,
+      "loss": 0.417,
       "step": 55290
     },
     {
-      "epoch": 1.95,
-      "learning_rate": 3.492496896167727e-05,
-      "loss": 0.2836,
+      "epoch": 1.9928280534832594,
+      "grad_norm": 0.20145370066165924,
+      "learning_rate": 3.4216331097359123e-05,
+      "loss": 0.4072,
       "step": 55295
     },
     {
-      "epoch": 1.95,
-      "learning_rate": 3.492235428628421e-05,
-      "loss": 0.2778,
+      "epoch": 1.9930082531444842,
+      "grad_norm": 0.1815030723810196,
+      "learning_rate": 3.421361843387034e-05,
+      "loss": 0.4481,
       "step": 55300
     },
     {
-      "epoch": 1.95,
-      "learning_rate": 3.4919739482055914e-05,
-      "loss": 0.3118,
+      "epoch": 1.9931884528057089,
+      "grad_norm": 0.16565322875976562,
+      "learning_rate": 3.42109056448478e-05,
+      "loss": 0.4235,
       "step": 55305
     },
     {
-      "epoch": 1.95,
-      "learning_rate": 3.491712454902635e-05,
-      "loss": 0.2832,
+      "epoch": 1.9933686524669334,
+      "grad_norm": 0.19166086614131927,
+      "learning_rate": 3.420819273032844e-05,
+      "loss": 0.4324,
       "step": 55310
     },
     {
-      "epoch": 1.95,
-      "learning_rate": 3.491450948722945e-05,
-      "loss": 0.2933,
+      "epoch": 1.993548852128158,
+      "grad_norm": 0.1797381490468979,
+      "learning_rate": 3.4205479690349246e-05,
+      "loss": 0.4356,
       "step": 55315
     },
     {
-      "epoch": 1.95,
-      "learning_rate": 3.491189429669918e-05,
-      "loss": 0.2857,
+      "epoch": 1.9937290517893826,
+      "grad_norm": 0.19711998105049133,
+      "learning_rate": 3.420276652494717e-05,
+      "loss": 0.3925,
       "step": 55320
     },
     {
-      "epoch": 1.95,
-      "learning_rate": 3.49092789774695e-05,
-      "loss": 0.2761,
+      "epoch": 1.9939092514506074,
+      "grad_norm": 0.157618910074234,
+      "learning_rate": 3.4200053234159185e-05,
+      "loss": 0.386,
       "step": 55325
     },
     {
-      "epoch": 1.95,
-      "learning_rate": 3.4906663529574344e-05,
-      "loss": 0.2938,
+      "epoch": 1.994089451111832,
+      "grad_norm": 0.2216421216726303,
+      "learning_rate": 3.419733981802226e-05,
+      "loss": 0.4304,
       "step": 55330
     },
     {
-      "epoch": 1.95,
-      "learning_rate": 3.490404795304771e-05,
-      "loss": 0.2615,
+      "epoch": 1.9942696507730564,
+      "grad_norm": 0.16847671568393707,
+      "learning_rate": 3.419462627657335e-05,
+      "loss": 0.4074,
       "step": 55335
     },
     {
-      "epoch": 1.95,
-      "learning_rate": 3.490143224792353e-05,
-      "loss": 0.2918,
+      "epoch": 1.9944498504342811,
+      "grad_norm": 0.19874538481235504,
+      "learning_rate": 3.4191912609849444e-05,
+      "loss": 0.386,
       "step": 55340
     },
     {
-      "epoch": 1.95,
-      "learning_rate": 3.489881641423579e-05,
-      "loss": 0.3023,
+      "epoch": 1.9946300500955059,
+      "grad_norm": 0.19660434126853943,
+      "learning_rate": 3.41891988178875e-05,
+      "loss": 0.3841,
       "step": 55345
     },
     {
-      "epoch": 1.95,
-      "learning_rate": 3.489620045201843e-05,
-      "loss": 0.304,
+      "epoch": 1.9948102497567306,
+      "grad_norm": 0.2128075808286667,
+      "learning_rate": 3.4186484900724514e-05,
+      "loss": 0.4227,
       "step": 55350
     },
     {
-      "epoch": 1.95,
-      "learning_rate": 3.489358436130543e-05,
-      "loss": 0.296,
+      "epoch": 1.9949904494179551,
+      "grad_norm": 0.1658129245042801,
+      "learning_rate": 3.418377085839744e-05,
+      "loss": 0.4055,
       "step": 55355
     },
     {
-      "epoch": 1.95,
-      "learning_rate": 3.489096814213076e-05,
-      "loss": 0.287,
+      "epoch": 1.9951706490791796,
+      "grad_norm": 0.1386236697435379,
+      "learning_rate": 3.418105669094327e-05,
+      "loss": 0.3855,
       "step": 55360
     },
     {
-      "epoch": 1.95,
-      "learning_rate": 3.488835179452839e-05,
-      "loss": 0.2749,
+      "epoch": 1.9953508487404044,
+      "grad_norm": 0.16813713312149048,
+      "learning_rate": 3.4178342398398985e-05,
+      "loss": 0.4049,
       "step": 55365
     },
     {
-      "epoch": 1.95,
-      "learning_rate": 3.488573531853229e-05,
-      "loss": 0.2931,
+      "epoch": 1.995531048401629,
+      "grad_norm": 0.19493988156318665,
+      "learning_rate": 3.417562798080155e-05,
+      "loss": 0.3933,
       "step": 55370
     },
     {
-      "epoch": 1.95,
-      "learning_rate": 3.4883118714176427e-05,
-      "loss": 0.2849,
+      "epoch": 1.9957112480628536,
+      "grad_norm": 0.19702892005443573,
+      "learning_rate": 3.417291343818797e-05,
+      "loss": 0.453,
       "step": 55375
     },
     {
-      "epoch": 1.95,
-      "learning_rate": 3.488050198149478e-05,
-      "loss": 0.2739,
+      "epoch": 1.9958914477240781,
+      "grad_norm": 0.18380559980869293,
+      "learning_rate": 3.417019877059522e-05,
+      "loss": 0.3968,
       "step": 55380
     },
     {
-      "epoch": 1.95,
-      "learning_rate": 3.487788512052132e-05,
-      "loss": 0.2894,
+      "epoch": 1.9960716473853028,
+      "grad_norm": 0.2419026494026184,
+      "learning_rate": 3.4167483978060285e-05,
+      "loss": 0.4219,
       "step": 55385
     },
     {
-      "epoch": 1.95,
-      "learning_rate": 3.4875268131290046e-05,
-      "loss": 0.2739,
+      "epoch": 1.9962518470465276,
+      "grad_norm": 0.19167737662792206,
+      "learning_rate": 3.416476906062015e-05,
+      "loss": 0.4485,
       "step": 55390
     },
     {
-      "epoch": 1.95,
-      "learning_rate": 3.4872651013834916e-05,
-      "loss": 0.2682,
+      "epoch": 1.9964320467077523,
+      "grad_norm": 0.18639305233955383,
+      "learning_rate": 3.416205401831182e-05,
+      "loss": 0.4201,
       "step": 55395
     },
     {
-      "epoch": 1.95,
-      "learning_rate": 3.487003376818992e-05,
-      "loss": 0.2523,
+      "epoch": 1.9966122463689768,
+      "grad_norm": 0.2547195851802826,
+      "learning_rate": 3.4159338851172276e-05,
+      "loss": 0.4503,
       "step": 55400
     },
     {
-      "epoch": 1.95,
-      "learning_rate": 3.486741639438904e-05,
-      "loss": 0.3135,
+      "epoch": 1.9967924460302013,
+      "grad_norm": 0.19804906845092773,
+      "learning_rate": 3.41566235592385e-05,
+      "loss": 0.4057,
       "step": 55405
     },
     {
-      "epoch": 1.95,
-      "learning_rate": 3.4864798892466256e-05,
-      "loss": 0.2938,
+      "epoch": 1.996972645691426,
+      "grad_norm": 0.21647123992443085,
+      "learning_rate": 3.415390814254752e-05,
+      "loss": 0.4265,
       "step": 55410
     },
     {
-      "epoch": 1.95,
-      "learning_rate": 3.4862181262455565e-05,
-      "loss": 0.263,
+      "epoch": 1.9971528453526508,
+      "grad_norm": 0.13523387908935547,
+      "learning_rate": 3.41511926011363e-05,
+      "loss": 0.4253,
       "step": 55415
     },
     {
-      "epoch": 1.95,
-      "learning_rate": 3.485956350439095e-05,
-      "loss": 0.2899,
+      "epoch": 1.9973330450138753,
+      "grad_norm": 0.16237081587314606,
+      "learning_rate": 3.414847693504186e-05,
+      "loss": 0.3982,
       "step": 55420
     },
     {
-      "epoch": 1.95,
-      "learning_rate": 3.48569456183064e-05,
-      "loss": 0.2736,
+      "epoch": 1.9975132446751,
+      "grad_norm": 0.12674693763256073,
+      "learning_rate": 3.414576114430119e-05,
+      "loss": 0.4324,
       "step": 55425
     },
     {
-      "epoch": 1.95,
-      "learning_rate": 3.485432760423591e-05,
-      "loss": 0.2928,
+      "epoch": 1.9976934443363246,
+      "grad_norm": 0.18315039575099945,
+      "learning_rate": 3.4143045228951296e-05,
+      "loss": 0.4158,
       "step": 55430
     },
     {
-      "epoch": 1.95,
-      "learning_rate": 3.485170946221347e-05,
-      "loss": 0.2833,
+      "epoch": 1.9978736439975493,
+      "grad_norm": 0.1764761209487915,
+      "learning_rate": 3.414032918902918e-05,
+      "loss": 0.3859,
       "step": 55435
     },
     {
-      "epoch": 1.95,
-      "learning_rate": 3.4849091192273064e-05,
-      "loss": 0.2792,
+      "epoch": 1.998053843658774,
+      "grad_norm": 0.19060806930065155,
+      "learning_rate": 3.413761302457185e-05,
+      "loss": 0.4238,
       "step": 55440
     },
     {
-      "epoch": 1.95,
-      "learning_rate": 3.484647279444872e-05,
-      "loss": 0.2718,
+      "epoch": 1.9982340433199985,
+      "grad_norm": 0.14990010857582092,
+      "learning_rate": 3.413489673561632e-05,
+      "loss": 0.4259,
       "step": 55445
     },
     {
-      "epoch": 1.95,
-      "learning_rate": 3.48438542687744e-05,
-      "loss": 0.2887,
+      "epoch": 1.998414242981223,
+      "grad_norm": 0.16765926778316498,
+      "learning_rate": 3.413218032219957e-05,
+      "loss": 0.4103,
       "step": 55450
     },
     {
-      "epoch": 1.95,
-      "learning_rate": 3.484123561528412e-05,
-      "loss": 0.274,
+      "epoch": 1.9985944426424478,
+      "grad_norm": 0.1568935662508011,
+      "learning_rate": 3.412946378435865e-05,
+      "loss": 0.4334,
       "step": 55455
     },
     {
-      "epoch": 1.95,
-      "learning_rate": 3.483861683401188e-05,
-      "loss": 0.2888,
+      "epoch": 1.9987746423036725,
+      "grad_norm": 0.1629842072725296,
+      "learning_rate": 3.4126747122130545e-05,
+      "loss": 0.4099,
       "step": 55460
     },
     {
-      "epoch": 1.95,
-      "learning_rate": 3.483599792499169e-05,
-      "loss": 0.2933,
+      "epoch": 1.9989548419648973,
+      "grad_norm": 0.1970033496618271,
+      "learning_rate": 3.4124030335552274e-05,
+      "loss": 0.4265,
       "step": 55465
     },
     {
-      "epoch": 1.95,
-      "learning_rate": 3.483337888825755e-05,
-      "loss": 0.2908,
+      "epoch": 1.9991350416261218,
+      "grad_norm": 0.2183963507413864,
+      "learning_rate": 3.4121313424660866e-05,
+      "loss": 0.4178,
       "step": 55470
     },
     {
-      "epoch": 1.95,
-      "learning_rate": 3.4830759723843464e-05,
-      "loss": 0.2973,
+      "epoch": 1.9993152412873463,
+      "grad_norm": 0.15539808571338654,
+      "learning_rate": 3.411859638949332e-05,
+      "loss": 0.4105,
       "step": 55475
     },
     {
-      "epoch": 1.95,
-      "learning_rate": 3.482814043178344e-05,
-      "loss": 0.2742,
+      "epoch": 1.999495440948571,
+      "grad_norm": 0.22591222822666168,
+      "learning_rate": 3.411642267190535e-05,
+      "loss": 0.4401,
       "step": 55480
     },
     {
-      "epoch": 1.95,
-      "learning_rate": 3.482552101211149e-05,
-      "loss": 0.285,
+      "epoch": 1.9996756406097957,
+      "grad_norm": 0.13716307282447815,
+      "learning_rate": 3.4113705413134064e-05,
+      "loss": 0.3836,
       "step": 55485
     },
     {
-      "epoch": 1.95,
-      "learning_rate": 3.4822901464861633e-05,
-      "loss": 0.287,
+      "epoch": 1.9998558402710203,
+      "grad_norm": 0.20352575182914734,
+      "learning_rate": 3.411098803019031e-05,
+      "loss": 0.4013,
       "step": 55490
     },
     {
-      "epoch": 1.95,
-      "learning_rate": 3.482028179006786e-05,
-      "loss": 0.2726,
+      "epoch": 2.0000360399322448,
+      "grad_norm": 0.2378479391336441,
+      "learning_rate": 3.41082705231111e-05,
+      "loss": 0.4204,
       "step": 55495
     },
     {
-      "epoch": 1.95,
-      "learning_rate": 3.48176619877642e-05,
-      "loss": 0.2736,
+      "epoch": 2.0002162395934695,
+      "grad_norm": 0.17885233461856842,
+      "learning_rate": 3.410555289193347e-05,
+      "loss": 0.3819,
       "step": 55500
     },
     {
-      "epoch": 1.95,
-      "eval_loss": 0.27318820357322693,
-      "eval_runtime": 10.5335,
-      "eval_samples_per_second": 9.494,
-      "eval_steps_per_second": 9.494,
+      "epoch": 2.0002162395934695,
+      "eval_loss": 0.43618687987327576,
+      "eval_runtime": 3.5358,
+      "eval_samples_per_second": 28.282,
+      "eval_steps_per_second": 7.071,
       "step": 55500
     },
     {
-      "epoch": 1.95,
-      "learning_rate": 3.4815042057984683e-05,
-      "loss": 0.2924,
+      "epoch": 2.0003964392546942,
+      "grad_norm": 0.21153418719768524,
+      "learning_rate": 3.4102835136694445e-05,
+      "loss": 0.4348,
       "step": 55505
     },
     {
-      "epoch": 1.95,
-      "learning_rate": 3.481242200076331e-05,
-      "loss": 0.2876,
+      "epoch": 2.000576638915919,
+      "grad_norm": 0.18446187674999237,
+      "learning_rate": 3.4100117257431055e-05,
+      "loss": 0.387,
       "step": 55510
     },
     {
-      "epoch": 1.95,
-      "learning_rate": 3.4809801816134096e-05,
-      "loss": 0.2819,
+      "epoch": 2.0007568385771433,
+      "grad_norm": 0.16465526819229126,
+      "learning_rate": 3.409739925418034e-05,
+      "loss": 0.3821,
       "step": 55515
     },
     {
-      "epoch": 1.95,
-      "learning_rate": 3.4807181504131075e-05,
-      "loss": 0.2526,
+      "epoch": 2.000937038238368,
+      "grad_norm": 0.19307419657707214,
+      "learning_rate": 3.40946811269793e-05,
+      "loss": 0.4331,
       "step": 55520
     },
     {
-      "epoch": 1.95,
-      "learning_rate": 3.480456106478826e-05,
-      "loss": 0.277,
+      "epoch": 2.0011172378995927,
+      "grad_norm": 0.18775279819965363,
+      "learning_rate": 3.4091962875865e-05,
+      "loss": 0.3975,
       "step": 55525
     },
     {
-      "epoch": 1.95,
-      "learning_rate": 3.480194049813969e-05,
-      "loss": 0.2633,
+      "epoch": 2.0012974375608175,
+      "grad_norm": 0.1970471441745758,
+      "learning_rate": 3.4089244500874474e-05,
+      "loss": 0.4205,
       "step": 55530
     },
     {
-      "epoch": 1.95,
-      "learning_rate": 3.479931980421938e-05,
-      "loss": 0.2715,
+      "epoch": 2.001477637222042,
+      "grad_norm": 0.15047413110733032,
+      "learning_rate": 3.408652600204474e-05,
+      "loss": 0.4004,
       "step": 55535
     },
     {
-      "epoch": 1.95,
-      "learning_rate": 3.479669898306136e-05,
-      "loss": 0.2846,
+      "epoch": 2.0016578368832665,
+      "grad_norm": 0.18766699731349945,
+      "learning_rate": 3.408380737941285e-05,
+      "loss": 0.3961,
       "step": 55540
     },
     {
-      "epoch": 1.95,
-      "learning_rate": 3.479407803469966e-05,
-      "loss": 0.2837,
+      "epoch": 2.001838036544491,
+      "grad_norm": 0.16125249862670898,
+      "learning_rate": 3.4081088633015845e-05,
+      "loss": 0.3797,
       "step": 55545
     },
     {
-      "epoch": 1.95,
-      "learning_rate": 3.47914569591683e-05,
-      "loss": 0.2764,
+      "epoch": 2.002018236205716,
+      "grad_norm": 0.23382239043712616,
+      "learning_rate": 3.407836976289077e-05,
+      "loss": 0.414,
       "step": 55550
     },
     {
-      "epoch": 1.95,
-      "learning_rate": 3.478883575650134e-05,
-      "loss": 0.2934,
+      "epoch": 2.0021984358669407,
+      "grad_norm": 0.17505022883415222,
+      "learning_rate": 3.4075650769074664e-05,
+      "loss": 0.3889,
       "step": 55555
     },
     {
-      "epoch": 1.95,
-      "learning_rate": 3.478621442673279e-05,
-      "loss": 0.2913,
+      "epoch": 2.0023786355281654,
+      "grad_norm": 0.21252861618995667,
+      "learning_rate": 3.4072931651604566e-05,
+      "loss": 0.4107,
       "step": 55560
     },
     {
-      "epoch": 1.95,
-      "learning_rate": 3.47835929698967e-05,
-      "loss": 0.2786,
+      "epoch": 2.0025588351893897,
+      "grad_norm": 0.1734471619129181,
+      "learning_rate": 3.407021241051754e-05,
+      "loss": 0.3442,
       "step": 55565
     },
     {
-      "epoch": 1.96,
-      "learning_rate": 3.47809713860271e-05,
-      "loss": 0.2932,
+      "epoch": 2.0027390348506144,
+      "grad_norm": 0.21779297292232513,
+      "learning_rate": 3.406749304585062e-05,
+      "loss": 0.3749,
       "step": 55570
     },
     {
-      "epoch": 1.96,
-      "learning_rate": 3.477834967515803e-05,
-      "loss": 0.2587,
+      "epoch": 2.002919234511839,
+      "grad_norm": 0.17229339480400085,
+      "learning_rate": 3.406477355764087e-05,
+      "loss": 0.3806,
       "step": 55575
     },
     {
-      "epoch": 1.96,
-      "learning_rate": 3.477572783732354e-05,
-      "loss": 0.303,
+      "epoch": 2.003099434173064,
+      "grad_norm": 0.18596251308918,
+      "learning_rate": 3.406205394592532e-05,
+      "loss": 0.3808,
       "step": 55580
     },
     {
-      "epoch": 1.96,
-      "learning_rate": 3.477310587255766e-05,
-      "loss": 0.2909,
+      "epoch": 2.003279633834288,
+      "grad_norm": 0.20308445394039154,
+      "learning_rate": 3.405933421074105e-05,
+      "loss": 0.395,
       "step": 55585
     },
     {
-      "epoch": 1.96,
-      "learning_rate": 3.477048378089444e-05,
-      "loss": 0.3092,
+      "epoch": 2.003459833495513,
+      "grad_norm": 0.26746681332588196,
+      "learning_rate": 3.40566143521251e-05,
+      "loss": 0.3969,
       "step": 55590
     },
     {
-      "epoch": 1.96,
-      "learning_rate": 3.476786156236794e-05,
-      "loss": 0.2843,
+      "epoch": 2.0036400331567377,
+      "grad_norm": 0.20009200274944305,
+      "learning_rate": 3.405389437011454e-05,
+      "loss": 0.3977,
       "step": 55595
     },
     {
-      "epoch": 1.96,
-      "learning_rate": 3.476523921701218e-05,
-      "loss": 0.2992,
+      "epoch": 2.0038202328179624,
+      "grad_norm": 0.17441947758197784,
+      "learning_rate": 3.405117426474642e-05,
+      "loss": 0.3884,
       "step": 55600
     },
     {
-      "epoch": 1.96,
-      "learning_rate": 3.4762616744861226e-05,
-      "loss": 0.2587,
+      "epoch": 2.004000432479187,
+      "grad_norm": 0.19016499817371368,
+      "learning_rate": 3.40484540360578e-05,
+      "loss": 0.3964,
       "step": 55605
     },
     {
-      "epoch": 1.96,
-      "learning_rate": 3.475999414594913e-05,
-      "loss": 0.273,
+      "epoch": 2.0041806321404114,
+      "grad_norm": 0.16852574050426483,
+      "learning_rate": 3.4045733684085745e-05,
+      "loss": 0.4337,
       "step": 55610
     },
     {
-      "epoch": 1.96,
-      "learning_rate": 3.475737142030995e-05,
-      "loss": 0.2752,
+      "epoch": 2.004360831801636,
+      "grad_norm": 0.20671193301677704,
+      "learning_rate": 3.404301320886732e-05,
+      "loss": 0.4207,
       "step": 55615
     },
     {
-      "epoch": 1.96,
-      "learning_rate": 3.4754748567977725e-05,
-      "loss": 0.2721,
+      "epoch": 2.004541031462861,
+      "grad_norm": 0.17593398690223694,
+      "learning_rate": 3.404029261043961e-05,
+      "loss": 0.3864,
       "step": 55620
     },
     {
-      "epoch": 1.96,
-      "learning_rate": 3.4752125588986525e-05,
-      "loss": 0.2776,
+      "epoch": 2.0047212311240856,
+      "grad_norm": 0.18823134899139404,
+      "learning_rate": 3.4037571888839635e-05,
+      "loss": 0.3806,
       "step": 55625
     },
     {
-      "epoch": 1.96,
-      "learning_rate": 3.474950248337039e-05,
-      "loss": 0.2845,
+      "epoch": 2.00490143078531,
+      "grad_norm": 0.16644109785556793,
+      "learning_rate": 3.403485104410451e-05,
+      "loss": 0.3839,
       "step": 55630
     },
     {
-      "epoch": 1.96,
-      "learning_rate": 3.4746879251163396e-05,
-      "loss": 0.2974,
+      "epoch": 2.0050816304465346,
+      "grad_norm": 0.16421851515769958,
+      "learning_rate": 3.403213007627128e-05,
+      "loss": 0.3643,
       "step": 55635
     },
     {
-      "epoch": 1.96,
-      "learning_rate": 3.4744255892399594e-05,
-      "loss": 0.2803,
+      "epoch": 2.0052618301077594,
+      "grad_norm": 0.14267808198928833,
+      "learning_rate": 3.402940898537703e-05,
+      "loss": 0.4057,
       "step": 55640
     },
     {
-      "epoch": 1.96,
-      "learning_rate": 3.4741632407113064e-05,
-      "loss": 0.2813,
+      "epoch": 2.005442029768984,
+      "grad_norm": 0.19570884108543396,
+      "learning_rate": 3.402668777145883e-05,
+      "loss": 0.3704,
       "step": 55645
     },
     {
-      "epoch": 1.96,
-      "learning_rate": 3.4739008795337845e-05,
-      "loss": 0.289,
+      "epoch": 2.005622229430209,
+      "grad_norm": 0.20084281265735626,
+      "learning_rate": 3.402396643455376e-05,
+      "loss": 0.43,
       "step": 55650
     },
     {
-      "epoch": 1.96,
-      "learning_rate": 3.473638505710802e-05,
-      "loss": 0.2605,
+      "epoch": 2.005802429091433,
+      "grad_norm": 0.17763961851596832,
+      "learning_rate": 3.402124497469889e-05,
+      "loss": 0.4088,
       "step": 55655
     },
     {
-      "epoch": 1.96,
-      "learning_rate": 3.4733761192457645e-05,
-      "loss": 0.2777,
+      "epoch": 2.005982628752658,
+      "grad_norm": 0.19127719104290009,
+      "learning_rate": 3.40185233919313e-05,
+      "loss": 0.3433,
       "step": 55660
     },
     {
-      "epoch": 1.96,
-      "learning_rate": 3.473113720142081e-05,
-      "loss": 0.2803,
+      "epoch": 2.0061628284138826,
+      "grad_norm": 0.20189210772514343,
+      "learning_rate": 3.4015801686288086e-05,
+      "loss": 0.3741,
       "step": 55665
     },
     {
-      "epoch": 1.96,
-      "learning_rate": 3.472851308403156e-05,
-      "loss": 0.2543,
+      "epoch": 2.0063430280751073,
+      "grad_norm": 0.16015605628490448,
+      "learning_rate": 3.401307985780631e-05,
+      "loss": 0.3657,
       "step": 55670
     },
     {
-      "epoch": 1.96,
-      "learning_rate": 3.472588884032398e-05,
-      "loss": 0.2648,
+      "epoch": 2.006523227736332,
+      "grad_norm": 0.16415055096149445,
+      "learning_rate": 3.401035790652307e-05,
+      "loss": 0.3841,
       "step": 55675
     },
     {
-      "epoch": 1.96,
-      "learning_rate": 3.472326447033214e-05,
-      "loss": 0.2898,
+      "epoch": 2.0067034273975564,
+      "grad_norm": 0.169295534491539,
+      "learning_rate": 3.400763583247545e-05,
+      "loss": 0.3687,
       "step": 55680
     },
     {
-      "epoch": 1.96,
-      "learning_rate": 3.4720639974090126e-05,
-      "loss": 0.2609,
+      "epoch": 2.006883627058781,
+      "grad_norm": 0.17803806066513062,
+      "learning_rate": 3.400491363570053e-05,
+      "loss": 0.3908,
       "step": 55685
     },
     {
-      "epoch": 1.96,
-      "learning_rate": 3.4718015351632e-05,
-      "loss": 0.2882,
+      "epoch": 2.007063826720006,
+      "grad_norm": 0.1836196929216385,
+      "learning_rate": 3.400219131623541e-05,
+      "loss": 0.4002,
       "step": 55690
     },
     {
-      "epoch": 1.96,
-      "learning_rate": 3.4715390602991855e-05,
-      "loss": 0.2933,
+      "epoch": 2.0072440263812306,
+      "grad_norm": 0.21328406035900116,
+      "learning_rate": 3.3999468874117176e-05,
+      "loss": 0.3604,
       "step": 55695
     },
     {
-      "epoch": 1.96,
-      "learning_rate": 3.471276572820377e-05,
-      "loss": 0.3122,
+      "epoch": 2.007424226042455,
+      "grad_norm": 0.1810711771249771,
+      "learning_rate": 3.399674630938292e-05,
+      "loss": 0.4015,
       "step": 55700
     },
     {
-      "epoch": 1.96,
-      "learning_rate": 3.471014072730181e-05,
-      "loss": 0.2917,
+      "epoch": 2.0076044257036796,
+      "grad_norm": 0.21469327807426453,
+      "learning_rate": 3.3994023622069734e-05,
+      "loss": 0.4499,
       "step": 55705
     },
     {
-      "epoch": 1.96,
-      "learning_rate": 3.470751560032009e-05,
-      "loss": 0.2758,
+      "epoch": 2.0077846253649043,
+      "grad_norm": 0.20458552241325378,
+      "learning_rate": 3.3991300812214724e-05,
+      "loss": 0.4261,
       "step": 55710
     },
     {
-      "epoch": 1.96,
-      "learning_rate": 3.4704890347292656e-05,
-      "loss": 0.2865,
+      "epoch": 2.007964825026129,
+      "grad_norm": 0.13773076236248016,
+      "learning_rate": 3.398857787985498e-05,
+      "loss": 0.3517,
       "step": 55715
     },
     {
-      "epoch": 1.96,
-      "learning_rate": 3.470226496825363e-05,
-      "loss": 0.2942,
+      "epoch": 2.008145024687354,
+      "grad_norm": 0.17209510505199432,
+      "learning_rate": 3.39858548250276e-05,
+      "loss": 0.3871,
       "step": 55720
     },
     {
-      "epoch": 1.96,
-      "learning_rate": 3.469963946323709e-05,
-      "loss": 0.2814,
+      "epoch": 2.008325224348578,
+      "grad_norm": 0.18736940622329712,
+      "learning_rate": 3.398313164776968e-05,
+      "loss": 0.3906,
       "step": 55725
     },
     {
-      "epoch": 1.96,
-      "learning_rate": 3.469701383227712e-05,
-      "loss": 0.2787,
+      "epoch": 2.008505424009803,
+      "grad_norm": 0.17324689030647278,
+      "learning_rate": 3.3980408348118337e-05,
+      "loss": 0.4141,
       "step": 55730
     },
     {
-      "epoch": 1.96,
-      "learning_rate": 3.469438807540781e-05,
-      "loss": 0.2709,
+      "epoch": 2.0086856236710275,
+      "grad_norm": 0.21437758207321167,
+      "learning_rate": 3.397768492611068e-05,
+      "loss": 0.3743,
       "step": 55735
     },
     {
-      "epoch": 1.96,
-      "learning_rate": 3.469176219266326e-05,
-      "loss": 0.2839,
+      "epoch": 2.0088658233322523,
+      "grad_norm": 0.20862148702144623,
+      "learning_rate": 3.397496138178379e-05,
+      "loss": 0.3897,
       "step": 55740
     },
     {
-      "epoch": 1.96,
-      "learning_rate": 3.468913618407758e-05,
-      "loss": 0.2847,
+      "epoch": 2.0090460229934766,
+      "grad_norm": 0.2141301929950714,
+      "learning_rate": 3.397223771517479e-05,
+      "loss": 0.4016,
       "step": 55745
     },
     {
-      "epoch": 1.96,
-      "learning_rate": 3.468651004968484e-05,
-      "loss": 0.2982,
+      "epoch": 2.0092262226547013,
+      "grad_norm": 0.21781469881534576,
+      "learning_rate": 3.39695139263208e-05,
+      "loss": 0.4502,
       "step": 55750
     },
     {
-      "epoch": 1.96,
-      "learning_rate": 3.468388378951915e-05,
-      "loss": 0.3017,
+      "epoch": 2.009406422315926,
+      "grad_norm": 0.1999913901090622,
+      "learning_rate": 3.39667900152589e-05,
+      "loss": 0.4362,
       "step": 55755
     },
     {
-      "epoch": 1.96,
-      "learning_rate": 3.468125740361461e-05,
-      "loss": 0.3006,
+      "epoch": 2.0095866219771508,
+      "grad_norm": 0.17094168066978455,
+      "learning_rate": 3.396406598202624e-05,
+      "loss": 0.4162,
       "step": 55760
     },
     {
-      "epoch": 1.96,
-      "learning_rate": 3.4678630892005334e-05,
-      "loss": 0.2897,
+      "epoch": 2.0097668216383755,
+      "grad_norm": 0.19397181272506714,
+      "learning_rate": 3.39613418266599e-05,
+      "loss": 0.4285,
       "step": 55765
     },
     {
-      "epoch": 1.96,
-      "learning_rate": 3.46760042547254e-05,
-      "loss": 0.3039,
+      "epoch": 2.0099470212996,
+      "grad_norm": 0.1974741369485855,
+      "learning_rate": 3.3958617549197024e-05,
+      "loss": 0.3795,
       "step": 55770
     },
     {
-      "epoch": 1.96,
-      "learning_rate": 3.4673377491808935e-05,
-      "loss": 0.2681,
+      "epoch": 2.0101272209608245,
+      "grad_norm": 0.16916494071483612,
+      "learning_rate": 3.395589314967472e-05,
+      "loss": 0.4219,
       "step": 55775
     },
     {
-      "epoch": 1.96,
-      "learning_rate": 3.467075060329004e-05,
-      "loss": 0.2855,
+      "epoch": 2.0103074206220493,
+      "grad_norm": 0.1723833829164505,
+      "learning_rate": 3.39531686281301e-05,
+      "loss": 0.3858,
       "step": 55780
     },
     {
-      "epoch": 1.96,
-      "learning_rate": 3.466812358920282e-05,
-      "loss": 0.2613,
+      "epoch": 2.010487620283274,
+      "grad_norm": 0.19679874181747437,
+      "learning_rate": 3.39504439846003e-05,
+      "loss": 0.4087,
       "step": 55785
     },
     {
-      "epoch": 1.96,
-      "learning_rate": 3.466549644958138e-05,
-      "loss": 0.2904,
+      "epoch": 2.0106678199444987,
+      "grad_norm": 0.18012060225009918,
+      "learning_rate": 3.3947719219122415e-05,
+      "loss": 0.4004,
       "step": 55790
     },
     {
-      "epoch": 1.96,
-      "learning_rate": 3.466286918445985e-05,
-      "loss": 0.2838,
+      "epoch": 2.010848019605723,
+      "grad_norm": 0.190474733710289,
+      "learning_rate": 3.39449943317336e-05,
+      "loss": 0.3732,
       "step": 55795
     },
     {
-      "epoch": 1.96,
-      "learning_rate": 3.466024179387232e-05,
-      "loss": 0.2562,
+      "epoch": 2.0110282192669477,
+      "grad_norm": 0.19163033366203308,
+      "learning_rate": 3.394226932247097e-05,
+      "loss": 0.4261,
       "step": 55800
     },
     {
-      "epoch": 1.96,
-      "learning_rate": 3.4657614277852926e-05,
-      "loss": 0.2805,
+      "epoch": 2.0112084189281725,
+      "grad_norm": 0.18268360197544098,
+      "learning_rate": 3.3939544191371654e-05,
+      "loss": 0.3969,
       "step": 55805
     },
     {
-      "epoch": 1.96,
-      "learning_rate": 3.465498663643576e-05,
-      "loss": 0.2851,
+      "epoch": 2.011388618589397,
+      "grad_norm": 0.1998191624879837,
+      "learning_rate": 3.393681893847277e-05,
+      "loss": 0.3728,
       "step": 55810
     },
     {
-      "epoch": 1.96,
-      "learning_rate": 3.465235886965498e-05,
-      "loss": 0.279,
+      "epoch": 2.0115688182506215,
+      "grad_norm": 0.20111285150051117,
+      "learning_rate": 3.393409356381147e-05,
+      "loss": 0.3964,
       "step": 55815
     },
     {
-      "epoch": 1.96,
-      "learning_rate": 3.464973097754466e-05,
-      "loss": 0.2746,
+      "epoch": 2.0117490179118462,
+      "grad_norm": 0.18400463461875916,
+      "learning_rate": 3.393136806742487e-05,
+      "loss": 0.3896,
       "step": 55820
     },
     {
-      "epoch": 1.96,
-      "learning_rate": 3.464710296013895e-05,
-      "loss": 0.2792,
+      "epoch": 2.011929217573071,
+      "grad_norm": 0.16936840116977692,
+      "learning_rate": 3.39286424493501e-05,
+      "loss": 0.3907,
       "step": 55825
     },
     {
-      "epoch": 1.96,
-      "learning_rate": 3.4644474817471964e-05,
-      "loss": 0.2919,
+      "epoch": 2.0121094172342957,
+      "grad_norm": 0.1799679696559906,
+      "learning_rate": 3.392591670962432e-05,
+      "loss": 0.3636,
       "step": 55830
     },
     {
-      "epoch": 1.96,
-      "learning_rate": 3.4641846549577836e-05,
-      "loss": 0.2743,
+      "epoch": 2.0122896168955204,
+      "grad_norm": 0.17350183427333832,
+      "learning_rate": 3.392319084828464e-05,
+      "loss": 0.3845,
       "step": 55835
     },
     {
-      "epoch": 1.96,
-      "learning_rate": 3.463921815649068e-05,
-      "loss": 0.2919,
+      "epoch": 2.0124698165567447,
+      "grad_norm": 0.18203559517860413,
+      "learning_rate": 3.3920464865368225e-05,
+      "loss": 0.4245,
       "step": 55840
     },
     {
-      "epoch": 1.96,
-      "learning_rate": 3.4636589638244635e-05,
-      "loss": 0.263,
+      "epoch": 2.0126500162179695,
+      "grad_norm": 0.18975740671157837,
+      "learning_rate": 3.3917738760912186e-05,
+      "loss": 0.3997,
       "step": 55845
     },
     {
-      "epoch": 1.96,
-      "learning_rate": 3.463396099487381e-05,
-      "loss": 0.3008,
+      "epoch": 2.012830215879194,
+      "grad_norm": 0.2037028819322586,
+      "learning_rate": 3.39150125349537e-05,
+      "loss": 0.3931,
       "step": 55850
     },
     {
-      "epoch": 1.97,
-      "learning_rate": 3.463133222641236e-05,
-      "loss": 0.2541,
+      "epoch": 2.013010415540419,
+      "grad_norm": 0.2006373107433319,
+      "learning_rate": 3.391228618752988e-05,
+      "loss": 0.3418,
       "step": 55855
     },
     {
-      "epoch": 1.97,
-      "learning_rate": 3.462870333289441e-05,
-      "loss": 0.2645,
+      "epoch": 2.013190615201643,
+      "grad_norm": 0.2143949270248413,
+      "learning_rate": 3.39095597186779e-05,
+      "loss": 0.3607,
       "step": 55860
     },
     {
-      "epoch": 1.97,
-      "learning_rate": 3.462607431435409e-05,
-      "loss": 0.283,
+      "epoch": 2.013370814862868,
+      "grad_norm": 0.18995679914951324,
+      "learning_rate": 3.390683312843489e-05,
+      "loss": 0.3774,
       "step": 55865
     },
     {
-      "epoch": 1.97,
-      "learning_rate": 3.462344517082553e-05,
-      "loss": 0.3032,
+      "epoch": 2.0135510145240927,
+      "grad_norm": 0.2137516885995865,
+      "learning_rate": 3.3904106416837994e-05,
+      "loss": 0.3916,
       "step": 55870
     },
     {
-      "epoch": 1.97,
-      "learning_rate": 3.462081590234289e-05,
-      "loss": 0.2615,
+      "epoch": 2.0137312141853174,
+      "grad_norm": 0.18241751194000244,
+      "learning_rate": 3.390137958392438e-05,
+      "loss": 0.3773,
       "step": 55875
     },
     {
-      "epoch": 1.97,
-      "learning_rate": 3.4618186508940283e-05,
-      "loss": 0.2786,
+      "epoch": 2.013911413846542,
+      "grad_norm": 0.19162051379680634,
+      "learning_rate": 3.3898652629731195e-05,
+      "loss": 0.3743,
       "step": 55880
     },
     {
-      "epoch": 1.97,
-      "learning_rate": 3.461555699065188e-05,
-      "loss": 0.2817,
+      "epoch": 2.0140916135077664,
+      "grad_norm": 0.17561465501785278,
+      "learning_rate": 3.389592555429558e-05,
+      "loss": 0.3987,
       "step": 55885
     },
     {
-      "epoch": 1.97,
-      "learning_rate": 3.461292734751179e-05,
-      "loss": 0.2885,
+      "epoch": 2.014271813168991,
+      "grad_norm": 0.20353040099143982,
+      "learning_rate": 3.389319835765471e-05,
+      "loss": 0.3807,
       "step": 55890
     },
     {
-      "epoch": 1.97,
-      "learning_rate": 3.461029757955419e-05,
-      "loss": 0.3076,
+      "epoch": 2.014452012830216,
+      "grad_norm": 0.1713477373123169,
+      "learning_rate": 3.389047103984573e-05,
+      "loss": 0.4177,
       "step": 55895
     },
     {
-      "epoch": 1.97,
-      "learning_rate": 3.4607667686813196e-05,
-      "loss": 0.2632,
+      "epoch": 2.0146322124914406,
+      "grad_norm": 0.18497739732265472,
+      "learning_rate": 3.388774360090581e-05,
+      "loss": 0.3734,
       "step": 55900
     },
     {
-      "epoch": 1.97,
-      "learning_rate": 3.460503766932297e-05,
-      "loss": 0.2721,
+      "epoch": 2.014812412152665,
+      "grad_norm": 0.1842721551656723,
+      "learning_rate": 3.38850160408721e-05,
+      "loss": 0.367,
       "step": 55905
     },
     {
-      "epoch": 1.97,
-      "learning_rate": 3.460240752711766e-05,
-      "loss": 0.2715,
+      "epoch": 2.0149926118138897,
+      "grad_norm": 0.14854677021503448,
+      "learning_rate": 3.3882288359781764e-05,
+      "loss": 0.4289,
       "step": 55910
     },
     {
-      "epoch": 1.97,
-      "learning_rate": 3.459977726023142e-05,
-      "loss": 0.2827,
+      "epoch": 2.0151728114751144,
+      "grad_norm": 0.2103215456008911,
+      "learning_rate": 3.387956055767197e-05,
+      "loss": 0.3974,
       "step": 55915
     },
     {
-      "epoch": 1.97,
-      "learning_rate": 3.45971468686984e-05,
-      "loss": 0.3024,
+      "epoch": 2.015353011136339,
+      "grad_norm": 0.2002989947795868,
+      "learning_rate": 3.387683263457989e-05,
+      "loss": 0.3743,
       "step": 55920
     },
     {
-      "epoch": 1.97,
-      "learning_rate": 3.459451635255275e-05,
-      "loss": 0.2799,
+      "epoch": 2.015533210797564,
+      "grad_norm": 0.1742120236158371,
+      "learning_rate": 3.3874104590542676e-05,
+      "loss": 0.386,
       "step": 55925
     },
     {
-      "epoch": 1.97,
-      "learning_rate": 3.459188571182863e-05,
-      "loss": 0.2953,
+      "epoch": 2.015713410458788,
+      "grad_norm": 0.19794981181621552,
+      "learning_rate": 3.38713764255975e-05,
+      "loss": 0.4048,
       "step": 55930
     },
     {
-      "epoch": 1.97,
-      "learning_rate": 3.458925494656019e-05,
-      "loss": 0.2639,
+      "epoch": 2.015893610120013,
+      "grad_norm": 0.21913039684295654,
+      "learning_rate": 3.3868648139781545e-05,
+      "loss": 0.4255,
       "step": 55935
     },
     {
-      "epoch": 1.97,
-      "learning_rate": 3.45866240567816e-05,
-      "loss": 0.2732,
+      "epoch": 2.0160738097812376,
+      "grad_norm": 0.19685392081737518,
+      "learning_rate": 3.3865919733131975e-05,
+      "loss": 0.3708,
       "step": 55940
     },
     {
-      "epoch": 1.97,
-      "learning_rate": 3.458399304252701e-05,
-      "loss": 0.2735,
+      "epoch": 2.0162540094424624,
+      "grad_norm": 0.23018737137317657,
+      "learning_rate": 3.386319120568597e-05,
+      "loss": 0.411,
       "step": 55945
     },
     {
-      "epoch": 1.97,
-      "learning_rate": 3.458136190383059e-05,
-      "loss": 0.266,
+      "epoch": 2.016434209103687,
+      "grad_norm": 0.196548730134964,
+      "learning_rate": 3.38604625574807e-05,
+      "loss": 0.3945,
       "step": 55950
     },
     {
-      "epoch": 1.97,
-      "learning_rate": 3.4578730640726496e-05,
-      "loss": 0.2605,
+      "epoch": 2.0166144087649114,
+      "grad_norm": 0.24264010787010193,
+      "learning_rate": 3.3857733788553335e-05,
+      "loss": 0.4077,
       "step": 55955
     },
     {
-      "epoch": 1.97,
-      "learning_rate": 3.4576099253248905e-05,
-      "loss": 0.2745,
+      "epoch": 2.016794608426136,
+      "grad_norm": 0.24170154333114624,
+      "learning_rate": 3.385500489894107e-05,
+      "loss": 0.3743,
       "step": 55960
     },
     {
-      "epoch": 1.97,
-      "learning_rate": 3.457346774143197e-05,
-      "loss": 0.2644,
+      "epoch": 2.016974808087361,
+      "grad_norm": 0.1775246411561966,
+      "learning_rate": 3.3852275888681076e-05,
+      "loss": 0.3975,
       "step": 55965
     },
     {
-      "epoch": 1.97,
-      "learning_rate": 3.457083610530987e-05,
-      "loss": 0.2862,
+      "epoch": 2.0171550077485856,
+      "grad_norm": 0.18977300822734833,
+      "learning_rate": 3.3849546757810536e-05,
+      "loss": 0.4093,
       "step": 55970
     },
     {
-      "epoch": 1.97,
-      "learning_rate": 3.4568204344916766e-05,
-      "loss": 0.2671,
+      "epoch": 2.01733520740981,
+      "grad_norm": 0.1697954684495926,
+      "learning_rate": 3.3846817506366625e-05,
+      "loss": 0.4306,
       "step": 55975
     },
     {
-      "epoch": 1.97,
-      "learning_rate": 3.4565572460286835e-05,
-      "loss": 0.2652,
+      "epoch": 2.0175154070710346,
+      "grad_norm": 0.20978444814682007,
+      "learning_rate": 3.384408813438655e-05,
+      "loss": 0.4178,
       "step": 55980
     },
     {
-      "epoch": 1.97,
-      "learning_rate": 3.456294045145425e-05,
-      "loss": 0.2871,
+      "epoch": 2.0176956067322593,
+      "grad_norm": 0.23438295722007751,
+      "learning_rate": 3.384135864190748e-05,
+      "loss": 0.4378,
       "step": 55985
     },
     {
-      "epoch": 1.97,
-      "learning_rate": 3.456030831845318e-05,
-      "loss": 0.2856,
+      "epoch": 2.017875806393484,
+      "grad_norm": 0.19239699840545654,
+      "learning_rate": 3.383862902896662e-05,
+      "loss": 0.3842,
       "step": 55990
     },
     {
-      "epoch": 1.97,
-      "learning_rate": 3.455767606131781e-05,
-      "loss": 0.3036,
+      "epoch": 2.018056006054709,
+      "grad_norm": 0.1987629383802414,
+      "learning_rate": 3.3835899295601146e-05,
+      "loss": 0.3805,
       "step": 55995
     },
     {
-      "epoch": 1.97,
-      "learning_rate": 3.455504368008232e-05,
-      "loss": 0.3041,
+      "epoch": 2.018236205715933,
+      "grad_norm": 0.19622477889060974,
+      "learning_rate": 3.383316944184825e-05,
+      "loss": 0.3615,
       "step": 56000
     },
     {
-      "epoch": 1.97,
-      "eval_loss": 0.27226722240448,
-      "eval_runtime": 10.5443,
-      "eval_samples_per_second": 9.484,
-      "eval_steps_per_second": 9.484,
+      "epoch": 2.018236205715933,
+      "eval_loss": 0.4372096359729767,
+      "eval_runtime": 3.5364,
+      "eval_samples_per_second": 28.278,
+      "eval_steps_per_second": 7.069,
       "step": 56000
     },
     {
-      "epoch": 1.97,
-      "learning_rate": 3.4552411174780886e-05,
-      "loss": 0.285,
+      "epoch": 2.018416405377158,
+      "grad_norm": 0.1874462217092514,
+      "learning_rate": 3.383043946774514e-05,
+      "loss": 0.4029,
       "step": 56005
     },
     {
-      "epoch": 1.97,
-      "learning_rate": 3.4549778545447685e-05,
-      "loss": 0.3094,
+      "epoch": 2.0185966050383826,
+      "grad_norm": 0.1793491393327713,
+      "learning_rate": 3.382770937332899e-05,
+      "loss": 0.4104,
       "step": 56010
     },
     {
-      "epoch": 1.97,
-      "learning_rate": 3.45471457921169e-05,
-      "loss": 0.2706,
+      "epoch": 2.0187768046996073,
+      "grad_norm": 0.15393778681755066,
+      "learning_rate": 3.382497915863702e-05,
+      "loss": 0.4142,
       "step": 56015
     },
     {
-      "epoch": 1.97,
-      "learning_rate": 3.454451291482272e-05,
-      "loss": 0.268,
+      "epoch": 2.0189570043608316,
+      "grad_norm": 0.17542561888694763,
+      "learning_rate": 3.3822248823706406e-05,
+      "loss": 0.3852,
       "step": 56020
     },
     {
-      "epoch": 1.97,
-      "learning_rate": 3.454187991359933e-05,
-      "loss": 0.2826,
+      "epoch": 2.0191372040220563,
+      "grad_norm": 0.20072583854198456,
+      "learning_rate": 3.3819518368574365e-05,
+      "loss": 0.4079,
       "step": 56025
     },
     {
-      "epoch": 1.97,
-      "learning_rate": 3.453924678848093e-05,
-      "loss": 0.2784,
+      "epoch": 2.019317403683281,
+      "grad_norm": 0.24251943826675415,
+      "learning_rate": 3.38167877932781e-05,
+      "loss": 0.4273,
       "step": 56030
     },
     {
-      "epoch": 1.97,
-      "learning_rate": 3.4536613539501684e-05,
-      "loss": 0.2817,
+      "epoch": 2.019497603344506,
+      "grad_norm": 0.19765913486480713,
+      "learning_rate": 3.3814057097854804e-05,
+      "loss": 0.3785,
       "step": 56035
     },
     {
-      "epoch": 1.97,
-      "learning_rate": 3.453398016669579e-05,
-      "loss": 0.2681,
+      "epoch": 2.0196778030057305,
+      "grad_norm": 0.1533811390399933,
+      "learning_rate": 3.381132628234168e-05,
+      "loss": 0.3913,
       "step": 56040
     },
     {
-      "epoch": 1.97,
-      "learning_rate": 3.453134667009745e-05,
-      "loss": 0.3129,
+      "epoch": 2.019858002666955,
+      "grad_norm": 0.15910130739212036,
+      "learning_rate": 3.380859534677595e-05,
+      "loss": 0.3877,
       "step": 56045
     },
     {
-      "epoch": 1.97,
-      "learning_rate": 3.452871304974086e-05,
-      "loss": 0.292,
+      "epoch": 2.0200382023281795,
+      "grad_norm": 0.298079252243042,
+      "learning_rate": 3.380586429119481e-05,
+      "loss": 0.418,
       "step": 56050
     },
     {
-      "epoch": 1.97,
-      "learning_rate": 3.452607930566022e-05,
-      "loss": 0.2589,
+      "epoch": 2.0202184019894043,
+      "grad_norm": 0.22058376669883728,
+      "learning_rate": 3.380313311563548e-05,
+      "loss": 0.4053,
       "step": 56055
     },
     {
-      "epoch": 1.97,
-      "learning_rate": 3.45234454378897e-05,
-      "loss": 0.2759,
+      "epoch": 2.020398601650629,
+      "grad_norm": 0.2122090458869934,
+      "learning_rate": 3.380040182013515e-05,
+      "loss": 0.4353,
       "step": 56060
     },
     {
-      "epoch": 1.97,
-      "learning_rate": 3.452081144646353e-05,
-      "loss": 0.3165,
+      "epoch": 2.0205788013118537,
+      "grad_norm": 0.15873844921588898,
+      "learning_rate": 3.3797670404731066e-05,
+      "loss": 0.4032,
       "step": 56065
     },
     {
-      "epoch": 1.97,
-      "learning_rate": 3.451817733141589e-05,
-      "loss": 0.2685,
+      "epoch": 2.020759000973078,
+      "grad_norm": 0.24122844636440277,
+      "learning_rate": 3.3794938869460416e-05,
+      "loss": 0.3695,
       "step": 56070
     },
     {
-      "epoch": 1.97,
-      "learning_rate": 3.4515543092781e-05,
-      "loss": 0.29,
+      "epoch": 2.0209392006343028,
+      "grad_norm": 0.15981855988502502,
+      "learning_rate": 3.3792207214360434e-05,
+      "loss": 0.3844,
       "step": 56075
     },
     {
-      "epoch": 1.97,
-      "learning_rate": 3.4512908730593044e-05,
-      "loss": 0.2866,
+      "epoch": 2.0211194002955275,
+      "grad_norm": 0.19484391808509827,
+      "learning_rate": 3.3789475439468324e-05,
+      "loss": 0.4208,
       "step": 56080
     },
     {
-      "epoch": 1.97,
-      "learning_rate": 3.451027424488624e-05,
-      "loss": 0.2695,
+      "epoch": 2.0212995999567522,
+      "grad_norm": 0.21466252207756042,
+      "learning_rate": 3.378674354482132e-05,
+      "loss": 0.4127,
       "step": 56085
     },
     {
-      "epoch": 1.97,
-      "learning_rate": 3.450763963569479e-05,
-      "loss": 0.2717,
+      "epoch": 2.0214797996179765,
+      "grad_norm": 0.20310963690280914,
+      "learning_rate": 3.378401153045663e-05,
+      "loss": 0.4158,
       "step": 56090
     },
     {
-      "epoch": 1.97,
-      "learning_rate": 3.4505004903052915e-05,
-      "loss": 0.2972,
+      "epoch": 2.0216599992792013,
+      "grad_norm": 0.21700692176818848,
+      "learning_rate": 3.3781279396411484e-05,
+      "loss": 0.4001,
       "step": 56095
     },
     {
-      "epoch": 1.97,
-      "learning_rate": 3.4502370046994806e-05,
-      "loss": 0.2862,
+      "epoch": 2.021840198940426,
+      "grad_norm": 0.17587092518806458,
+      "learning_rate": 3.3778547142723104e-05,
+      "loss": 0.3782,
       "step": 56100
     },
     {
-      "epoch": 1.97,
-      "learning_rate": 3.4499735067554686e-05,
-      "loss": 0.2698,
+      "epoch": 2.0220203986016507,
+      "grad_norm": 0.2222931832075119,
+      "learning_rate": 3.3775814769428725e-05,
+      "loss": 0.388,
       "step": 56105
     },
     {
-      "epoch": 1.97,
-      "learning_rate": 3.449709996476677e-05,
-      "loss": 0.2861,
+      "epoch": 2.0222005982628755,
+      "grad_norm": 0.2034565955400467,
+      "learning_rate": 3.377308227656557e-05,
+      "loss": 0.4348,
       "step": 56110
     },
     {
-      "epoch": 1.97,
-      "learning_rate": 3.449446473866527e-05,
-      "loss": 0.2772,
+      "epoch": 2.0223807979240997,
+      "grad_norm": 0.17816239595413208,
+      "learning_rate": 3.3770349664170865e-05,
+      "loss": 0.3529,
       "step": 56115
     },
     {
-      "epoch": 1.97,
-      "learning_rate": 3.4491829389284396e-05,
-      "loss": 0.2967,
+      "epoch": 2.0225609975853245,
+      "grad_norm": 0.18056336045265198,
+      "learning_rate": 3.376761693228184e-05,
+      "loss": 0.429,
       "step": 56120
     },
     {
-      "epoch": 1.97,
-      "learning_rate": 3.4489193916658374e-05,
-      "loss": 0.2691,
+      "epoch": 2.022741197246549,
+      "grad_norm": 0.21382421255111694,
+      "learning_rate": 3.3764884080935735e-05,
+      "loss": 0.3775,
       "step": 56125
     },
     {
-      "epoch": 1.97,
-      "learning_rate": 3.448655832082143e-05,
-      "loss": 0.2875,
+      "epoch": 2.022921396907774,
+      "grad_norm": 0.19596946239471436,
+      "learning_rate": 3.376215111016978e-05,
+      "loss": 0.4107,
       "step": 56130
     },
     {
-      "epoch": 1.97,
-      "learning_rate": 3.448392260180777e-05,
-      "loss": 0.3062,
+      "epoch": 2.0231015965689982,
+      "grad_norm": 0.1608893871307373,
+      "learning_rate": 3.375941802002122e-05,
+      "loss": 0.3723,
       "step": 56135
     },
     {
-      "epoch": 1.98,
-      "learning_rate": 3.448128675965163e-05,
-      "loss": 0.2866,
+      "epoch": 2.023281796230223,
+      "grad_norm": 0.22571824491024017,
+      "learning_rate": 3.3756684810527275e-05,
+      "loss": 0.4514,
       "step": 56140
     },
     {
-      "epoch": 1.98,
-      "learning_rate": 3.447865079438723e-05,
-      "loss": 0.295,
+      "epoch": 2.0234619958914477,
+      "grad_norm": 0.17446362972259521,
+      "learning_rate": 3.3753951481725196e-05,
+      "loss": 0.3783,
       "step": 56145
     },
     {
-      "epoch": 1.98,
-      "learning_rate": 3.447601470604879e-05,
-      "loss": 0.2987,
+      "epoch": 2.0236421955526724,
+      "grad_norm": 0.2468831092119217,
+      "learning_rate": 3.3751218033652225e-05,
+      "loss": 0.4116,
       "step": 56150
     },
     {
-      "epoch": 1.98,
-      "learning_rate": 3.447337849467055e-05,
-      "loss": 0.2784,
+      "epoch": 2.023822395213897,
+      "grad_norm": 0.22208808362483978,
+      "learning_rate": 3.374848446634561e-05,
+      "loss": 0.4228,
       "step": 56155
     },
     {
-      "epoch": 1.98,
-      "learning_rate": 3.447074216028673e-05,
-      "loss": 0.2711,
+      "epoch": 2.0240025948751215,
+      "grad_norm": 0.20974206924438477,
+      "learning_rate": 3.3745750779842575e-05,
+      "loss": 0.406,
       "step": 56160
     },
     {
-      "epoch": 1.98,
-      "learning_rate": 3.4468105702931565e-05,
-      "loss": 0.2909,
+      "epoch": 2.024182794536346,
+      "grad_norm": 0.14893017709255219,
+      "learning_rate": 3.374301697418039e-05,
+      "loss": 0.3712,
       "step": 56165
     },
     {
-      "epoch": 1.98,
-      "learning_rate": 3.446546912263929e-05,
-      "loss": 0.276,
+      "epoch": 2.024362994197571,
+      "grad_norm": 0.16727864742279053,
+      "learning_rate": 3.3740283049396285e-05,
+      "loss": 0.415,
       "step": 56170
     },
     {
-      "epoch": 1.98,
-      "learning_rate": 3.4462832419444135e-05,
-      "loss": 0.3056,
+      "epoch": 2.0245431938587957,
+      "grad_norm": 0.2555207908153534,
+      "learning_rate": 3.3737549005527514e-05,
+      "loss": 0.4133,
       "step": 56175
     },
     {
-      "epoch": 1.98,
-      "learning_rate": 3.446019559338034e-05,
-      "loss": 0.2793,
+      "epoch": 2.0247233935200204,
+      "grad_norm": 0.22973132133483887,
+      "learning_rate": 3.3734814842611326e-05,
+      "loss": 0.4124,
       "step": 56180
     },
     {
-      "epoch": 1.98,
-      "learning_rate": 3.445755864448213e-05,
-      "loss": 0.2806,
+      "epoch": 2.0249035931812447,
+      "grad_norm": 0.21145066618919373,
+      "learning_rate": 3.373208056068499e-05,
+      "loss": 0.3879,
       "step": 56185
     },
     {
-      "epoch": 1.98,
-      "learning_rate": 3.445492157278377e-05,
-      "loss": 0.288,
+      "epoch": 2.0250837928424694,
+      "grad_norm": 0.19409604370594025,
+      "learning_rate": 3.372934615978574e-05,
+      "loss": 0.3742,
       "step": 56190
     },
     {
-      "epoch": 1.98,
-      "learning_rate": 3.4452284378319465e-05,
-      "loss": 0.2845,
+      "epoch": 2.025263992503694,
+      "grad_norm": 0.21370406448841095,
+      "learning_rate": 3.372661163995084e-05,
+      "loss": 0.3909,
       "step": 56195
     },
     {
-      "epoch": 1.98,
-      "learning_rate": 3.444964706112349e-05,
-      "loss": 0.2675,
+      "epoch": 2.025444192164919,
+      "grad_norm": 0.22447529435157776,
+      "learning_rate": 3.372387700121754e-05,
+      "loss": 0.4121,
       "step": 56200
     },
     {
-      "epoch": 1.98,
-      "learning_rate": 3.444700962123007e-05,
-      "loss": 0.2838,
+      "epoch": 2.025624391826143,
+      "grad_norm": 0.22073820233345032,
+      "learning_rate": 3.372114224362311e-05,
+      "loss": 0.3848,
       "step": 56205
     },
     {
-      "epoch": 1.98,
-      "learning_rate": 3.444437205867346e-05,
-      "loss": 0.2862,
+      "epoch": 2.025804591487368,
+      "grad_norm": 0.18990257382392883,
+      "learning_rate": 3.3718407367204807e-05,
+      "loss": 0.3688,
       "step": 56210
     },
     {
-      "epoch": 1.98,
-      "learning_rate": 3.44417343734879e-05,
-      "loss": 0.2953,
+      "epoch": 2.0259847911485926,
+      "grad_norm": 0.172977477312088,
+      "learning_rate": 3.371567237199989e-05,
+      "loss": 0.3931,
       "step": 56215
     },
     {
-      "epoch": 1.98,
-      "learning_rate": 3.4439096565707647e-05,
-      "loss": 0.2968,
+      "epoch": 2.0261649908098174,
+      "grad_norm": 0.2128041833639145,
+      "learning_rate": 3.371293725804562e-05,
+      "loss": 0.4185,
       "step": 56220
     },
     {
-      "epoch": 1.98,
-      "learning_rate": 3.4436458635366935e-05,
-      "loss": 0.295,
+      "epoch": 2.026345190471042,
+      "grad_norm": 0.1763525903224945,
+      "learning_rate": 3.3710202025379276e-05,
+      "loss": 0.4008,
       "step": 56225
     },
     {
-      "epoch": 1.98,
-      "learning_rate": 3.4433820582500035e-05,
-      "loss": 0.2711,
+      "epoch": 2.0265253901322664,
+      "grad_norm": 0.19905276596546173,
+      "learning_rate": 3.370746667403811e-05,
+      "loss": 0.3789,
       "step": 56230
     },
     {
-      "epoch": 1.98,
-      "learning_rate": 3.4431182407141186e-05,
-      "loss": 0.2735,
+      "epoch": 2.026705589793491,
+      "grad_norm": 0.19667592644691467,
+      "learning_rate": 3.37047312040594e-05,
+      "loss": 0.3738,
       "step": 56235
     },
     {
-      "epoch": 1.98,
-      "learning_rate": 3.4428544109324656e-05,
-      "loss": 0.2823,
+      "epoch": 2.026885789454716,
+      "grad_norm": 0.19164909422397614,
+      "learning_rate": 3.37019956154804e-05,
+      "loss": 0.433,
       "step": 56240
     },
     {
-      "epoch": 1.98,
-      "learning_rate": 3.442590568908469e-05,
-      "loss": 0.2839,
+      "epoch": 2.0270659891159406,
+      "grad_norm": 0.1493864804506302,
+      "learning_rate": 3.369925990833841e-05,
+      "loss": 0.4096,
       "step": 56245
     },
     {
-      "epoch": 1.98,
-      "learning_rate": 3.442326714645555e-05,
-      "loss": 0.2897,
+      "epoch": 2.027246188777165,
+      "grad_norm": 0.18463338911533356,
+      "learning_rate": 3.3696524082670684e-05,
+      "loss": 0.3974,
       "step": 56250
     },
     {
-      "epoch": 1.98,
-      "learning_rate": 3.44206284814715e-05,
-      "loss": 0.2949,
+      "epoch": 2.0274263884383896,
+      "grad_norm": 0.22145789861679077,
+      "learning_rate": 3.369378813851449e-05,
+      "loss": 0.3888,
       "step": 56255
     },
     {
-      "epoch": 1.98,
-      "learning_rate": 3.441798969416679e-05,
-      "loss": 0.2782,
+      "epoch": 2.0276065880996144,
+      "grad_norm": 0.1729925125837326,
+      "learning_rate": 3.369105207590713e-05,
+      "loss": 0.4284,
       "step": 56260
     },
     {
-      "epoch": 1.98,
-      "learning_rate": 3.44153507845757e-05,
-      "loss": 0.2896,
+      "epoch": 2.027786787760839,
+      "grad_norm": 0.1994604617357254,
+      "learning_rate": 3.368831589488586e-05,
+      "loss": 0.4122,
       "step": 56265
     },
     {
-      "epoch": 1.98,
-      "learning_rate": 3.441271175273248e-05,
-      "loss": 0.2937,
+      "epoch": 2.027966987422064,
+      "grad_norm": 0.16139256954193115,
+      "learning_rate": 3.3685579595487985e-05,
+      "loss": 0.3947,
       "step": 56270
     },
     {
-      "epoch": 1.98,
-      "learning_rate": 3.44100725986714e-05,
-      "loss": 0.2904,
+      "epoch": 2.028147187083288,
+      "grad_norm": 0.1860417276620865,
+      "learning_rate": 3.3682843177750745e-05,
+      "loss": 0.4223,
       "step": 56275
     },
     {
-      "epoch": 1.98,
-      "learning_rate": 3.440743332242673e-05,
-      "loss": 0.3106,
+      "epoch": 2.028327386744513,
+      "grad_norm": 0.1829526275396347,
+      "learning_rate": 3.368010664171146e-05,
+      "loss": 0.3896,
       "step": 56280
     },
     {
-      "epoch": 1.98,
-      "learning_rate": 3.440479392403274e-05,
-      "loss": 0.2914,
+      "epoch": 2.0285075864057376,
+      "grad_norm": 0.23350392282009125,
+      "learning_rate": 3.36773699874074e-05,
+      "loss": 0.4205,
       "step": 56285
     },
     {
-      "epoch": 1.98,
-      "learning_rate": 3.440215440352369e-05,
-      "loss": 0.2849,
+      "epoch": 2.0286877860669623,
+      "grad_norm": 0.18550008535385132,
+      "learning_rate": 3.3674633214875855e-05,
+      "loss": 0.4167,
       "step": 56290
     },
     {
-      "epoch": 1.98,
-      "learning_rate": 3.439951476093387e-05,
-      "loss": 0.3019,
+      "epoch": 2.028867985728187,
+      "grad_norm": 0.2336433231830597,
+      "learning_rate": 3.367189632415411e-05,
+      "loss": 0.3576,
       "step": 56295
     },
     {
-      "epoch": 1.98,
-      "learning_rate": 3.4396874996297545e-05,
-      "loss": 0.2609,
+      "epoch": 2.0290481853894113,
+      "grad_norm": 0.18545769155025482,
+      "learning_rate": 3.366915931527946e-05,
+      "loss": 0.4111,
       "step": 56300
     },
     {
-      "epoch": 1.98,
-      "learning_rate": 3.4394235109649e-05,
-      "loss": 0.3112,
+      "epoch": 2.029228385050636,
+      "grad_norm": 0.1865958422422409,
+      "learning_rate": 3.3666969623134705e-05,
+      "loss": 0.4044,
       "step": 56305
     },
     {
-      "epoch": 1.98,
-      "learning_rate": 3.4391595101022486e-05,
-      "loss": 0.2749,
+      "epoch": 2.029408584711861,
+      "grad_norm": 0.202582448720932,
+      "learning_rate": 3.366423240167879e-05,
+      "loss": 0.3743,
       "step": 56310
     },
     {
-      "epoch": 1.98,
-      "learning_rate": 3.43889549704523e-05,
-      "loss": 0.272,
+      "epoch": 2.0295887843730855,
+      "grad_norm": 0.19475452601909637,
+      "learning_rate": 3.366149506217439e-05,
+      "loss": 0.419,
       "step": 56315
     },
     {
-      "epoch": 1.98,
-      "learning_rate": 3.438631471797273e-05,
-      "loss": 0.2844,
+      "epoch": 2.02976898403431,
+      "grad_norm": 0.19832675158977509,
+      "learning_rate": 3.36587576046588e-05,
+      "loss": 0.387,
       "step": 56320
     },
     {
-      "epoch": 1.98,
-      "learning_rate": 3.438367434361805e-05,
-      "loss": 0.2786,
+      "epoch": 2.0299491836955346,
+      "grad_norm": 0.19266216456890106,
+      "learning_rate": 3.3656020029169306e-05,
+      "loss": 0.4015,
       "step": 56325
     },
     {
-      "epoch": 1.98,
-      "learning_rate": 3.438103384742254e-05,
-      "loss": 0.2764,
+      "epoch": 2.0301293833567593,
+      "grad_norm": 0.19489936530590057,
+      "learning_rate": 3.365328233574322e-05,
+      "loss": 0.4126,
       "step": 56330
     },
     {
-      "epoch": 1.98,
-      "learning_rate": 3.437839322942049e-05,
-      "loss": 0.3176,
+      "epoch": 2.030309583017984,
+      "grad_norm": 0.1996261477470398,
+      "learning_rate": 3.365054452441783e-05,
+      "loss": 0.4134,
       "step": 56335
     },
     {
-      "epoch": 1.98,
-      "learning_rate": 3.437575248964618e-05,
-      "loss": 0.2639,
+      "epoch": 2.0304897826792088,
+      "grad_norm": 0.1988709717988968,
+      "learning_rate": 3.364780659523046e-05,
+      "loss": 0.4044,
       "step": 56340
     },
     {
-      "epoch": 1.98,
-      "learning_rate": 3.4373111628133906e-05,
-      "loss": 0.287,
+      "epoch": 2.030669982340433,
+      "grad_norm": 0.21509073674678802,
+      "learning_rate": 3.3645068548218385e-05,
+      "loss": 0.3803,
       "step": 56345
     },
     {
-      "epoch": 1.98,
-      "learning_rate": 3.437047064491795e-05,
-      "loss": 0.289,
+      "epoch": 2.030850182001658,
+      "grad_norm": 0.20546337962150574,
+      "learning_rate": 3.364233038341893e-05,
+      "loss": 0.3971,
       "step": 56350
     },
     {
-      "epoch": 1.98,
-      "learning_rate": 3.436782954003261e-05,
-      "loss": 0.2841,
+      "epoch": 2.0310303816628825,
+      "grad_norm": 0.20153018832206726,
+      "learning_rate": 3.36395921008694e-05,
+      "loss": 0.3793,
       "step": 56355
     },
     {
-      "epoch": 1.98,
-      "learning_rate": 3.436518831351218e-05,
-      "loss": 0.2736,
+      "epoch": 2.0312105813241073,
+      "grad_norm": 0.23244844377040863,
+      "learning_rate": 3.3636853700607095e-05,
+      "loss": 0.3696,
       "step": 56360
     },
     {
-      "epoch": 1.98,
-      "learning_rate": 3.436254696539095e-05,
-      "loss": 0.2837,
+      "epoch": 2.0313907809853315,
+      "grad_norm": 0.20720942318439484,
+      "learning_rate": 3.3634115182669335e-05,
+      "loss": 0.3894,
       "step": 56365
     },
     {
-      "epoch": 1.98,
-      "learning_rate": 3.435990549570321e-05,
-      "loss": 0.2807,
+      "epoch": 2.0315709806465563,
+      "grad_norm": 0.19618476927280426,
+      "learning_rate": 3.363137654709342e-05,
+      "loss": 0.4301,
       "step": 56370
     },
     {
-      "epoch": 1.98,
-      "learning_rate": 3.435726390448328e-05,
-      "loss": 0.2494,
+      "epoch": 2.031751180307781,
+      "grad_norm": 0.1612580120563507,
+      "learning_rate": 3.362863779391668e-05,
+      "loss": 0.3662,
       "step": 56375
     },
     {
-      "epoch": 1.98,
-      "learning_rate": 3.435462219176543e-05,
-      "loss": 0.2607,
+      "epoch": 2.0319313799690057,
+      "grad_norm": 0.14148502051830292,
+      "learning_rate": 3.362589892317641e-05,
+      "loss": 0.3881,
       "step": 56380
     },
     {
-      "epoch": 1.98,
-      "learning_rate": 3.435198035758399e-05,
-      "loss": 0.2763,
+      "epoch": 2.0321115796302305,
+      "grad_norm": 0.19896137714385986,
+      "learning_rate": 3.3623159934909954e-05,
+      "loss": 0.3847,
       "step": 56385
     },
     {
-      "epoch": 1.98,
-      "learning_rate": 3.434933840197324e-05,
-      "loss": 0.3139,
+      "epoch": 2.0322917792914548,
+      "grad_norm": 0.16365410387516022,
+      "learning_rate": 3.3620420829154605e-05,
+      "loss": 0.3822,
       "step": 56390
     },
     {
-      "epoch": 1.98,
-      "learning_rate": 3.4346696324967494e-05,
-      "loss": 0.2886,
+      "epoch": 2.0324719789526795,
+      "grad_norm": 0.1631069928407669,
+      "learning_rate": 3.361768160594768e-05,
+      "loss": 0.4095,
       "step": 56395
     },
     {
-      "epoch": 1.98,
-      "learning_rate": 3.434405412660106e-05,
-      "loss": 0.2772,
+      "epoch": 2.0326521786139042,
+      "grad_norm": 0.22321099042892456,
+      "learning_rate": 3.361494226532653e-05,
+      "loss": 0.3992,
       "step": 56400
     },
     {
-      "epoch": 1.98,
-      "learning_rate": 3.434141180690823e-05,
-      "loss": 0.2993,
+      "epoch": 2.032832378275129,
+      "grad_norm": 0.2479621022939682,
+      "learning_rate": 3.361220280732845e-05,
+      "loss": 0.3764,
       "step": 56405
     },
     {
-      "epoch": 1.98,
-      "learning_rate": 3.4338769365923335e-05,
-      "loss": 0.3046,
+      "epoch": 2.0330125779363533,
+      "grad_norm": 0.2009827196598053,
+      "learning_rate": 3.3609463231990774e-05,
+      "loss": 0.4009,
       "step": 56410
     },
     {
-      "epoch": 1.98,
-      "learning_rate": 3.4336126803680665e-05,
-      "loss": 0.2795,
+      "epoch": 2.033192777597578,
+      "grad_norm": 0.20739832520484924,
+      "learning_rate": 3.3606723539350825e-05,
+      "loss": 0.3841,
       "step": 56415
     },
     {
-      "epoch": 1.99,
-      "learning_rate": 3.433348412021456e-05,
-      "loss": 0.2812,
+      "epoch": 2.0333729772588027,
+      "grad_norm": 0.2028648406267166,
+      "learning_rate": 3.360398372944594e-05,
+      "loss": 0.4247,
       "step": 56420
     },
     {
-      "epoch": 1.99,
-      "learning_rate": 3.433084131555929e-05,
-      "loss": 0.2734,
+      "epoch": 2.0335531769200275,
+      "grad_norm": 0.19367647171020508,
+      "learning_rate": 3.360124380231344e-05,
+      "loss": 0.381,
       "step": 56425
     },
     {
-      "epoch": 1.99,
-      "learning_rate": 3.43281983897492e-05,
-      "loss": 0.2761,
+      "epoch": 2.033733376581252,
+      "grad_norm": 0.1919804960489273,
+      "learning_rate": 3.3598503757990664e-05,
+      "loss": 0.4649,
       "step": 56430
     },
     {
-      "epoch": 1.99,
-      "learning_rate": 3.4325555342818614e-05,
-      "loss": 0.2884,
+      "epoch": 2.0339135762424765,
+      "grad_norm": 0.17203310132026672,
+      "learning_rate": 3.359576359651493e-05,
+      "loss": 0.3987,
       "step": 56435
     },
     {
-      "epoch": 1.99,
-      "learning_rate": 3.432291217480183e-05,
-      "loss": 0.2656,
+      "epoch": 2.034093775903701,
+      "grad_norm": 0.23200438916683197,
+      "learning_rate": 3.359302331792358e-05,
+      "loss": 0.385,
       "step": 56440
     },
     {
-      "epoch": 1.99,
-      "learning_rate": 3.4320268885733174e-05,
-      "loss": 0.2916,
+      "epoch": 2.034273975564926,
+      "grad_norm": 0.20147477090358734,
+      "learning_rate": 3.3590282922253966e-05,
+      "loss": 0.3981,
       "step": 56445
     },
     {
-      "epoch": 1.99,
-      "learning_rate": 3.431762547564698e-05,
-      "loss": 0.2979,
+      "epoch": 2.0344541752261507,
+      "grad_norm": 0.20963551104068756,
+      "learning_rate": 3.3587542409543396e-05,
+      "loss": 0.4105,
       "step": 56450
     },
     {
-      "epoch": 1.99,
-      "learning_rate": 3.431498194457755e-05,
-      "loss": 0.2739,
+      "epoch": 2.0346343748873754,
+      "grad_norm": 0.20766286551952362,
+      "learning_rate": 3.358480177982923e-05,
+      "loss": 0.4136,
       "step": 56455
     },
     {
-      "epoch": 1.99,
-      "learning_rate": 3.4312338292559214e-05,
-      "loss": 0.2616,
+      "epoch": 2.0348145745485997,
+      "grad_norm": 0.22245804965496063,
+      "learning_rate": 3.3582061033148795e-05,
+      "loss": 0.4043,
       "step": 56460
     },
     {
-      "epoch": 1.99,
-      "learning_rate": 3.430969451962631e-05,
-      "loss": 0.2783,
+      "epoch": 2.0349947742098244,
+      "grad_norm": 0.18696880340576172,
+      "learning_rate": 3.3579320169539445e-05,
+      "loss": 0.3749,
       "step": 56465
     },
     {
-      "epoch": 1.99,
-      "learning_rate": 3.430705062581316e-05,
-      "loss": 0.2749,
+      "epoch": 2.035174973871049,
+      "grad_norm": 0.19157637655735016,
+      "learning_rate": 3.3576579189038525e-05,
+      "loss": 0.4225,
       "step": 56470
     },
     {
-      "epoch": 1.99,
-      "learning_rate": 3.4304406611154085e-05,
-      "loss": 0.2792,
+      "epoch": 2.035355173532274,
+      "grad_norm": 0.18588903546333313,
+      "learning_rate": 3.3573838091683366e-05,
+      "loss": 0.3961,
       "step": 56475
     },
     {
-      "epoch": 1.99,
-      "learning_rate": 3.430176247568342e-05,
-      "loss": 0.2631,
+      "epoch": 2.035535373193498,
+      "grad_norm": 0.196426123380661,
+      "learning_rate": 3.3571096877511324e-05,
+      "loss": 0.4047,
       "step": 56480
     },
     {
-      "epoch": 1.99,
-      "learning_rate": 3.42991182194355e-05,
-      "loss": 0.2848,
+      "epoch": 2.035715572854723,
+      "grad_norm": 0.1676120012998581,
+      "learning_rate": 3.356835554655975e-05,
+      "loss": 0.4061,
       "step": 56485
     },
     {
-      "epoch": 1.99,
-      "learning_rate": 3.4296473842444663e-05,
-      "loss": 0.2854,
+      "epoch": 2.0358957725159477,
+      "grad_norm": 0.22126492857933044,
+      "learning_rate": 3.3565614098865985e-05,
+      "loss": 0.4079,
       "step": 56490
     },
     {
-      "epoch": 1.99,
-      "learning_rate": 3.429382934474524e-05,
-      "loss": 0.2827,
+      "epoch": 2.0360759721771724,
+      "grad_norm": 0.1804613471031189,
+      "learning_rate": 3.356287253446738e-05,
+      "loss": 0.3552,
       "step": 56495
     },
     {
-      "epoch": 1.99,
-      "learning_rate": 3.429118472637157e-05,
-      "loss": 0.2898,
+      "epoch": 2.036256171838397,
+      "grad_norm": 0.16997934877872467,
+      "learning_rate": 3.3560130853401306e-05,
+      "loss": 0.3867,
       "step": 56500
     },
     {
-      "epoch": 1.99,
-      "eval_loss": 0.2718130946159363,
-      "eval_runtime": 10.5492,
-      "eval_samples_per_second": 9.479,
-      "eval_steps_per_second": 9.479,
+      "epoch": 2.036256171838397,
+      "eval_loss": 0.43729928135871887,
+      "eval_runtime": 3.5477,
+      "eval_samples_per_second": 28.187,
+      "eval_steps_per_second": 7.047,
       "step": 56500
     },
     {
-      "epoch": 1.99,
-      "learning_rate": 3.4288539987357984e-05,
-      "loss": 0.2907,
+      "epoch": 2.0364363714996214,
+      "grad_norm": 0.1884368509054184,
+      "learning_rate": 3.355738905570511e-05,
+      "loss": 0.3983,
       "step": 56505
     },
     {
-      "epoch": 1.99,
-      "learning_rate": 3.428589512773882e-05,
-      "loss": 0.2519,
+      "epoch": 2.036616571160846,
+      "grad_norm": 0.21548138558864594,
+      "learning_rate": 3.3554647141416136e-05,
+      "loss": 0.4085,
       "step": 56510
     },
     {
-      "epoch": 1.99,
-      "learning_rate": 3.428325014754844e-05,
-      "loss": 0.2855,
+      "epoch": 2.036796770822071,
+      "grad_norm": 0.23149925470352173,
+      "learning_rate": 3.355190511057175e-05,
+      "loss": 0.3907,
       "step": 56515
     },
     {
-      "epoch": 1.99,
-      "learning_rate": 3.4280605046821175e-05,
-      "loss": 0.2906,
+      "epoch": 2.0369769704832956,
+      "grad_norm": 0.18845608830451965,
+      "learning_rate": 3.354916296320931e-05,
+      "loss": 0.4065,
       "step": 56520
     },
     {
-      "epoch": 1.99,
-      "learning_rate": 3.4277959825591366e-05,
-      "loss": 0.2636,
+      "epoch": 2.03715717014452,
+      "grad_norm": 0.19569791853427887,
+      "learning_rate": 3.354642069936619e-05,
+      "loss": 0.3941,
       "step": 56525
     },
     {
-      "epoch": 1.99,
-      "learning_rate": 3.427531448389336e-05,
-      "loss": 0.2877,
+      "epoch": 2.0373373698057446,
+      "grad_norm": 0.24140161275863647,
+      "learning_rate": 3.354367831907974e-05,
+      "loss": 0.4019,
       "step": 56530
     },
     {
-      "epoch": 1.99,
-      "learning_rate": 3.427266902176151e-05,
-      "loss": 0.2604,
+      "epoch": 2.0375175694669694,
+      "grad_norm": 0.15203697979450226,
+      "learning_rate": 3.354093582238732e-05,
+      "loss": 0.3785,
       "step": 56535
     },
     {
-      "epoch": 1.99,
-      "learning_rate": 3.4270023439230166e-05,
-      "loss": 0.2789,
+      "epoch": 2.037697769128194,
+      "grad_norm": 0.17884066700935364,
+      "learning_rate": 3.353819320932632e-05,
+      "loss": 0.3748,
       "step": 56540
     },
     {
-      "epoch": 1.99,
-      "learning_rate": 3.4267377736333676e-05,
-      "loss": 0.3074,
+      "epoch": 2.037877968789419,
+      "grad_norm": 0.20854458212852478,
+      "learning_rate": 3.353545047993408e-05,
+      "loss": 0.3853,
       "step": 56545
     },
     {
-      "epoch": 1.99,
-      "learning_rate": 3.42647319131064e-05,
-      "loss": 0.2577,
+      "epoch": 2.038058168450643,
+      "grad_norm": 0.20811249315738678,
+      "learning_rate": 3.3532707634247975e-05,
+      "loss": 0.3865,
       "step": 56550
     },
     {
-      "epoch": 1.99,
-      "learning_rate": 3.4262085969582674e-05,
-      "loss": 0.2948,
+      "epoch": 2.038238368111868,
+      "grad_norm": 0.1722935438156128,
+      "learning_rate": 3.352996467230538e-05,
+      "loss": 0.3549,
       "step": 56555
     },
     {
-      "epoch": 1.99,
-      "learning_rate": 3.425943990579687e-05,
-      "loss": 0.2721,
+      "epoch": 2.0384185677730926,
+      "grad_norm": 0.2165447473526001,
+      "learning_rate": 3.3527221594143675e-05,
+      "loss": 0.3848,
       "step": 56560
     },
     {
-      "epoch": 1.99,
-      "learning_rate": 3.4256793721783345e-05,
-      "loss": 0.3173,
+      "epoch": 2.0385987674343173,
+      "grad_norm": 0.221002995967865,
+      "learning_rate": 3.352447839980022e-05,
+      "loss": 0.3889,
       "step": 56565
     },
     {
-      "epoch": 1.99,
-      "learning_rate": 3.4254147417576445e-05,
-      "loss": 0.2765,
+      "epoch": 2.038778967095542,
+      "grad_norm": 0.2625134587287903,
+      "learning_rate": 3.35217350893124e-05,
+      "loss": 0.3828,
       "step": 56570
     },
     {
-      "epoch": 1.99,
-      "learning_rate": 3.425150099321055e-05,
-      "loss": 0.2899,
+      "epoch": 2.0389591667567664,
+      "grad_norm": 0.21043117344379425,
+      "learning_rate": 3.35189916627176e-05,
+      "loss": 0.3867,
       "step": 56575
     },
     {
-      "epoch": 1.99,
-      "learning_rate": 3.4248854448720004e-05,
-      "loss": 0.2773,
+      "epoch": 2.039139366417991,
+      "grad_norm": 0.19859947264194489,
+      "learning_rate": 3.351624812005317e-05,
+      "loss": 0.3922,
       "step": 56580
     },
     {
-      "epoch": 1.99,
-      "learning_rate": 3.424620778413918e-05,
-      "loss": 0.2734,
+      "epoch": 2.039319566079216,
+      "grad_norm": 0.15382780134677887,
+      "learning_rate": 3.351350446135651e-05,
+      "loss": 0.3614,
       "step": 56585
     },
     {
-      "epoch": 1.99,
-      "learning_rate": 3.4243560999502445e-05,
-      "loss": 0.2944,
+      "epoch": 2.0394997657404406,
+      "grad_norm": 0.21880534291267395,
+      "learning_rate": 3.3510760686665014e-05,
+      "loss": 0.396,
       "step": 56590
     },
     {
-      "epoch": 1.99,
-      "learning_rate": 3.4240914094844154e-05,
-      "loss": 0.2746,
+      "epoch": 2.039679965401665,
+      "grad_norm": 0.21891264617443085,
+      "learning_rate": 3.3508016796016047e-05,
+      "loss": 0.392,
       "step": 56595
     },
     {
-      "epoch": 1.99,
-      "learning_rate": 3.4238267070198686e-05,
-      "loss": 0.2693,
+      "epoch": 2.0398601650628896,
+      "grad_norm": 0.20019477605819702,
+      "learning_rate": 3.350527278944699e-05,
+      "loss": 0.3729,
       "step": 56600
     },
     {
-      "epoch": 1.99,
-      "learning_rate": 3.4235619925600417e-05,
-      "loss": 0.2708,
+      "epoch": 2.0400403647241143,
+      "grad_norm": 0.18630538880825043,
+      "learning_rate": 3.350252866699525e-05,
+      "loss": 0.3704,
       "step": 56605
     },
     {
-      "epoch": 1.99,
-      "learning_rate": 3.42329726610837e-05,
-      "loss": 0.28,
+      "epoch": 2.040220564385339,
+      "grad_norm": 0.20929007232189178,
+      "learning_rate": 3.349978442869819e-05,
+      "loss": 0.4028,
       "step": 56610
     },
     {
-      "epoch": 1.99,
-      "learning_rate": 3.4230325276682924e-05,
-      "loss": 0.2826,
+      "epoch": 2.040400764046564,
+      "grad_norm": 0.18913328647613525,
+      "learning_rate": 3.3497040074593215e-05,
+      "loss": 0.3956,
       "step": 56615
     },
     {
-      "epoch": 1.99,
-      "learning_rate": 3.422767777243245e-05,
-      "loss": 0.294,
+      "epoch": 2.040580963707788,
+      "grad_norm": 0.1790267825126648,
+      "learning_rate": 3.349429560471772e-05,
+      "loss": 0.4009,
       "step": 56620
     },
     {
-      "epoch": 1.99,
-      "learning_rate": 3.4225030148366664e-05,
-      "loss": 0.2748,
+      "epoch": 2.040761163369013,
+      "grad_norm": 0.20545370876789093,
+      "learning_rate": 3.349155101910909e-05,
+      "loss": 0.3951,
       "step": 56625
     },
     {
-      "epoch": 1.99,
-      "learning_rate": 3.422238240451995e-05,
-      "loss": 0.2786,
+      "epoch": 2.0409413630302375,
+      "grad_norm": 0.21455180644989014,
+      "learning_rate": 3.348880631780472e-05,
+      "loss": 0.4135,
       "step": 56630
     },
     {
-      "epoch": 1.99,
-      "learning_rate": 3.4219734540926665e-05,
-      "loss": 0.2295,
+      "epoch": 2.0411215626914623,
+      "grad_norm": 0.19538956880569458,
+      "learning_rate": 3.3486061500842014e-05,
+      "loss": 0.4057,
       "step": 56635
     },
     {
-      "epoch": 1.99,
-      "learning_rate": 3.421708655762122e-05,
-      "loss": 0.2852,
+      "epoch": 2.0413017623526866,
+      "grad_norm": 0.19726307690143585,
+      "learning_rate": 3.348331656825835e-05,
+      "loss": 0.4318,
       "step": 56640
     },
     {
-      "epoch": 1.99,
-      "learning_rate": 3.421443845463797e-05,
-      "loss": 0.2934,
+      "epoch": 2.0414819620139113,
+      "grad_norm": 0.21314232051372528,
+      "learning_rate": 3.348057152009115e-05,
+      "loss": 0.4373,
       "step": 56645
     },
     {
-      "epoch": 1.99,
-      "learning_rate": 3.4211790232011305e-05,
-      "loss": 0.2506,
+      "epoch": 2.041662161675136,
+      "grad_norm": 0.19441983103752136,
+      "learning_rate": 3.347782635637781e-05,
+      "loss": 0.3647,
       "step": 56650
     },
     {
-      "epoch": 1.99,
-      "learning_rate": 3.4209141889775615e-05,
-      "loss": 0.265,
+      "epoch": 2.0418423613363608,
+      "grad_norm": 0.19044964015483856,
+      "learning_rate": 3.347508107715572e-05,
+      "loss": 0.3825,
       "step": 56655
     },
     {
-      "epoch": 1.99,
-      "learning_rate": 3.42064934279653e-05,
-      "loss": 0.3019,
+      "epoch": 2.0420225609975855,
+      "grad_norm": 0.19193421304225922,
+      "learning_rate": 3.347233568246228e-05,
+      "loss": 0.4275,
       "step": 56660
     },
     {
-      "epoch": 1.99,
-      "learning_rate": 3.4203844846614716e-05,
-      "loss": 0.2981,
+      "epoch": 2.04220276065881,
+      "grad_norm": 0.225186288356781,
+      "learning_rate": 3.3469590172334926e-05,
+      "loss": 0.3661,
       "step": 56665
     },
     {
-      "epoch": 1.99,
-      "learning_rate": 3.4201196145758286e-05,
-      "loss": 0.2878,
+      "epoch": 2.0423829603200345,
+      "grad_norm": 0.16008292138576508,
+      "learning_rate": 3.346684454681104e-05,
+      "loss": 0.355,
       "step": 56670
     },
     {
-      "epoch": 1.99,
-      "learning_rate": 3.419854732543038e-05,
-      "loss": 0.2791,
+      "epoch": 2.0425631599812593,
+      "grad_norm": 0.20448531210422516,
+      "learning_rate": 3.346409880592802e-05,
+      "loss": 0.3863,
       "step": 56675
     },
     {
-      "epoch": 1.99,
-      "learning_rate": 3.41958983856654e-05,
-      "loss": 0.2633,
+      "epoch": 2.042743359642484,
+      "grad_norm": 0.15298992395401,
+      "learning_rate": 3.346135294972331e-05,
+      "loss": 0.3827,
       "step": 56680
     },
     {
-      "epoch": 1.99,
-      "learning_rate": 3.4193249326497744e-05,
-      "loss": 0.2844,
+      "epoch": 2.0429235593037087,
+      "grad_norm": 0.1888943463563919,
+      "learning_rate": 3.34586069782343e-05,
+      "loss": 0.367,
       "step": 56685
     },
     {
-      "epoch": 1.99,
-      "learning_rate": 3.419060014796181e-05,
-      "loss": 0.3059,
+      "epoch": 2.043103758964933,
+      "grad_norm": 0.18571306765079498,
+      "learning_rate": 3.345586089149841e-05,
+      "loss": 0.3533,
       "step": 56690
     },
     {
-      "epoch": 1.99,
-      "learning_rate": 3.418795085009198e-05,
-      "loss": 0.2673,
+      "epoch": 2.0432839586261577,
+      "grad_norm": 0.2252894788980484,
+      "learning_rate": 3.345311468955305e-05,
+      "loss": 0.3834,
       "step": 56695
     },
     {
-      "epoch": 1.99,
-      "learning_rate": 3.418530143292267e-05,
-      "loss": 0.288,
+      "epoch": 2.0434641582873825,
+      "grad_norm": 0.20659992098808289,
+      "learning_rate": 3.3450368372435643e-05,
+      "loss": 0.4081,
       "step": 56700
     },
     {
-      "epoch": 2.0,
-      "learning_rate": 3.418265189648826e-05,
-      "loss": 0.2726,
+      "epoch": 2.043644357948607,
+      "grad_norm": 0.20119290053844452,
+      "learning_rate": 3.344762194018359e-05,
+      "loss": 0.3907,
       "step": 56705
     },
     {
-      "epoch": 2.0,
-      "learning_rate": 3.418000224082318e-05,
-      "loss": 0.2807,
+      "epoch": 2.0438245576098315,
+      "grad_norm": 0.17512202262878418,
+      "learning_rate": 3.344487539283434e-05,
+      "loss": 0.3676,
       "step": 56710
     },
     {
-      "epoch": 2.0,
-      "learning_rate": 3.417735246596182e-05,
-      "loss": 0.2939,
+      "epoch": 2.0440047572710562,
+      "grad_norm": 0.18519236147403717,
+      "learning_rate": 3.3442128730425295e-05,
+      "loss": 0.4245,
       "step": 56715
     },
     {
-      "epoch": 2.0,
-      "learning_rate": 3.417470257193859e-05,
-      "loss": 0.297,
+      "epoch": 2.044184956932281,
+      "grad_norm": 0.20480823516845703,
+      "learning_rate": 3.343938195299387e-05,
+      "loss": 0.4162,
       "step": 56720
     },
     {
-      "epoch": 2.0,
-      "learning_rate": 3.417205255878788e-05,
-      "loss": 0.2763,
+      "epoch": 2.0443651565935057,
+      "grad_norm": 0.2772522568702698,
+      "learning_rate": 3.3436635060577506e-05,
+      "loss": 0.3941,
       "step": 56725
     },
     {
-      "epoch": 2.0,
-      "learning_rate": 3.416940242654412e-05,
-      "loss": 0.2568,
+      "epoch": 2.0445453562547304,
+      "grad_norm": 0.19306902587413788,
+      "learning_rate": 3.3433888053213624e-05,
+      "loss": 0.4025,
       "step": 56730
     },
     {
-      "epoch": 2.0,
-      "learning_rate": 3.416675217524171e-05,
-      "loss": 0.2837,
+      "epoch": 2.0447255559159547,
+      "grad_norm": 0.22703427076339722,
+      "learning_rate": 3.3431140930939653e-05,
+      "loss": 0.3602,
       "step": 56735
     },
     {
-      "epoch": 2.0,
-      "learning_rate": 3.4164101804915064e-05,
-      "loss": 0.2868,
+      "epoch": 2.0449057555771795,
+      "grad_norm": 0.17691659927368164,
+      "learning_rate": 3.3428393693793006e-05,
+      "loss": 0.3831,
       "step": 56740
     },
     {
-      "epoch": 2.0,
-      "learning_rate": 3.41614513155986e-05,
-      "loss": 0.2832,
+      "epoch": 2.045085955238404,
+      "grad_norm": 0.2599165439605713,
+      "learning_rate": 3.342564634181114e-05,
+      "loss": 0.4334,
       "step": 56745
     },
     {
-      "epoch": 2.0,
-      "learning_rate": 3.4158800707326723e-05,
-      "loss": 0.2483,
+      "epoch": 2.045266154899629,
+      "grad_norm": 0.16780352592468262,
+      "learning_rate": 3.3422898875031475e-05,
+      "loss": 0.4114,
       "step": 56750
     },
     {
-      "epoch": 2.0,
-      "learning_rate": 3.4156680135084465e-05,
-      "loss": 0.2854,
+      "epoch": 2.045446354560853,
+      "grad_norm": 0.21082206070423126,
+      "learning_rate": 3.342015129349143e-05,
+      "loss": 0.3919,
       "step": 56755
     },
     {
-      "epoch": 2.0,
-      "learning_rate": 3.415402931277958e-05,
-      "loss": 0.2904,
+      "epoch": 2.045626554222078,
+      "grad_norm": 0.24321195483207703,
+      "learning_rate": 3.341740359722846e-05,
+      "loss": 0.4228,
       "step": 56760
     },
     {
-      "epoch": 2.0,
-      "learning_rate": 3.415137837161566e-05,
-      "loss": 0.2945,
+      "epoch": 2.0458067538833027,
+      "grad_norm": 0.189229354262352,
+      "learning_rate": 3.341465578628e-05,
+      "loss": 0.378,
       "step": 56765
     },
     {
-      "epoch": 2.0,
-      "learning_rate": 3.414872731162713e-05,
-      "loss": 0.2769,
+      "epoch": 2.0459869535445274,
+      "grad_norm": 0.2039814442396164,
+      "learning_rate": 3.341190786068349e-05,
+      "loss": 0.3869,
       "step": 56770
     },
     {
-      "epoch": 2.0,
-      "learning_rate": 3.4146076132848406e-05,
-      "loss": 0.2743,
+      "epoch": 2.046167153205752,
+      "grad_norm": 0.15500949323177338,
+      "learning_rate": 3.340915982047635e-05,
+      "loss": 0.3749,
       "step": 56775
     },
     {
-      "epoch": 2.0,
-      "learning_rate": 3.414342483531391e-05,
-      "loss": 0.2749,
+      "epoch": 2.0463473528669764,
+      "grad_norm": 0.21698758006095886,
+      "learning_rate": 3.340641166569604e-05,
+      "loss": 0.3944,
       "step": 56780
     },
     {
-      "epoch": 2.0,
-      "learning_rate": 3.414077341905806e-05,
-      "loss": 0.296,
+      "epoch": 2.046527552528201,
+      "grad_norm": 0.1889837086200714,
+      "learning_rate": 3.3403663396379997e-05,
+      "loss": 0.3515,
       "step": 56785
     },
     {
-      "epoch": 2.0,
-      "learning_rate": 3.41381218841153e-05,
-      "loss": 0.2932,
+      "epoch": 2.046707752189426,
+      "grad_norm": 0.1569356471300125,
+      "learning_rate": 3.340091501256567e-05,
+      "loss": 0.4297,
       "step": 56790
     },
     {
-      "epoch": 2.0,
-      "learning_rate": 3.413547023052004e-05,
-      "loss": 0.2935,
+      "epoch": 2.0468879518506506,
+      "grad_norm": 0.17683616280555725,
+      "learning_rate": 3.339816651429051e-05,
+      "loss": 0.4058,
       "step": 56795
     },
     {
-      "epoch": 2.0,
-      "learning_rate": 3.4132818458306735e-05,
-      "loss": 0.2933,
+      "epoch": 2.0470681515118754,
+      "grad_norm": 0.22128205001354218,
+      "learning_rate": 3.339541790159194e-05,
+      "loss": 0.429,
       "step": 56800
     },
     {
-      "epoch": 2.0,
-      "learning_rate": 3.4130166567509785e-05,
-      "loss": 0.284,
+      "epoch": 2.0472483511730997,
+      "grad_norm": 0.1818923056125641,
+      "learning_rate": 3.339266917450744e-05,
+      "loss": 0.4136,
       "step": 56805
     },
     {
-      "epoch": 2.0,
-      "learning_rate": 3.4127514558163656e-05,
-      "loss": 0.2524,
+      "epoch": 2.0474285508343244,
+      "grad_norm": 0.19361944496631622,
+      "learning_rate": 3.3389920333074434e-05,
+      "loss": 0.4062,
       "step": 56810
     },
     {
-      "epoch": 2.0,
-      "learning_rate": 3.412486243030275e-05,
-      "loss": 0.3027,
+      "epoch": 2.047608750495549,
+      "grad_norm": 0.1958514004945755,
+      "learning_rate": 3.33871713773304e-05,
+      "loss": 0.4092,
       "step": 56815
     },
     {
-      "epoch": 2.0,
-      "learning_rate": 3.412221018396152e-05,
-      "loss": 0.2836,
+      "epoch": 2.047788950156774,
+      "grad_norm": 0.2066863626241684,
+      "learning_rate": 3.338442230731278e-05,
+      "loss": 0.4023,
       "step": 56820
     },
     {
-      "epoch": 2.0,
-      "learning_rate": 3.4119557819174406e-05,
-      "loss": 0.2806,
+      "epoch": 2.047969149817998,
+      "grad_norm": 0.18415698409080505,
+      "learning_rate": 3.338167312305902e-05,
+      "loss": 0.3767,
       "step": 56825
     },
     {
-      "epoch": 2.0,
-      "learning_rate": 3.411690533597585e-05,
-      "loss": 0.2758,
+      "epoch": 2.048149349479223,
+      "grad_norm": 0.19590577483177185,
+      "learning_rate": 3.33789238246066e-05,
+      "loss": 0.3947,
       "step": 56830
     },
     {
-      "epoch": 2.0,
-      "learning_rate": 3.4114252734400274e-05,
-      "loss": 0.28,
+      "epoch": 2.0483295491404476,
+      "grad_norm": 0.17477436363697052,
+      "learning_rate": 3.337617441199295e-05,
+      "loss": 0.3764,
       "step": 56835
     },
     {
-      "epoch": 2.0,
-      "learning_rate": 3.411160001448214e-05,
-      "loss": 0.3095,
+      "epoch": 2.0485097488016724,
+      "grad_norm": 0.20771732926368713,
+      "learning_rate": 3.337342488525556e-05,
+      "loss": 0.437,
       "step": 56840
     },
     {
-      "epoch": 2.0,
-      "learning_rate": 3.4108947176255884e-05,
-      "loss": 0.3184,
+      "epoch": 2.048689948462897,
+      "grad_norm": 0.18982523679733276,
+      "learning_rate": 3.337067524443187e-05,
+      "loss": 0.4064,
       "step": 56845
     },
     {
-      "epoch": 2.0,
-      "learning_rate": 3.4106294219755955e-05,
-      "loss": 0.2491,
+      "epoch": 2.0488701481241214,
+      "grad_norm": 0.19457025825977325,
+      "learning_rate": 3.3367925489559346e-05,
+      "loss": 0.4211,
       "step": 56850
     },
     {
-      "epoch": 2.0,
-      "learning_rate": 3.410364114501679e-05,
-      "loss": 0.2714,
+      "epoch": 2.049050347785346,
+      "grad_norm": 0.17263978719711304,
+      "learning_rate": 3.336517562067546e-05,
+      "loss": 0.366,
       "step": 56855
     },
     {
-      "epoch": 2.0,
-      "learning_rate": 3.410098795207285e-05,
-      "loss": 0.2563,
+      "epoch": 2.049230547446571,
+      "grad_norm": 0.22003693878650665,
+      "learning_rate": 3.336242563781768e-05,
+      "loss": 0.4007,
       "step": 56860
     },
     {
-      "epoch": 2.0,
-      "learning_rate": 3.4098334640958584e-05,
-      "loss": 0.288,
+      "epoch": 2.0494107471077956,
+      "grad_norm": 0.24091579020023346,
+      "learning_rate": 3.3359675541023474e-05,
+      "loss": 0.4483,
       "step": 56865
     },
     {
-      "epoch": 2.0,
-      "learning_rate": 3.409568121170843e-05,
-      "loss": 0.2888,
+      "epoch": 2.04959094676902,
+      "grad_norm": 0.1734263151884079,
+      "learning_rate": 3.3356925330330294e-05,
+      "loss": 0.3737,
       "step": 56870
     },
     {
-      "epoch": 2.0,
-      "learning_rate": 3.409302766435685e-05,
-      "loss": 0.2673,
+      "epoch": 2.0497711464302446,
+      "grad_norm": 0.20855608582496643,
+      "learning_rate": 3.335417500577563e-05,
+      "loss": 0.4281,
       "step": 56875
     },
     {
-      "epoch": 2.0,
-      "learning_rate": 3.40903739989383e-05,
-      "loss": 0.2644,
+      "epoch": 2.0499513460914693,
+      "grad_norm": 0.2285163253545761,
+      "learning_rate": 3.335142456739695e-05,
+      "loss": 0.4231,
       "step": 56880
     },
     {
-      "epoch": 2.0,
-      "learning_rate": 3.408772021548724e-05,
-      "loss": 0.2681,
+      "epoch": 2.050131545752694,
+      "grad_norm": 0.17819838225841522,
+      "learning_rate": 3.334867401523173e-05,
+      "loss": 0.3831,
       "step": 56885
     },
     {
-      "epoch": 2.0,
-      "learning_rate": 3.408506631403811e-05,
-      "loss": 0.2682,
+      "epoch": 2.050311745413919,
+      "grad_norm": 0.2181788682937622,
+      "learning_rate": 3.3345923349317436e-05,
+      "loss": 0.4211,
       "step": 56890
     },
     {
-      "epoch": 2.0,
-      "learning_rate": 3.408241229462539e-05,
-      "loss": 0.2903,
+      "epoch": 2.050491945075143,
+      "grad_norm": 0.18571996688842773,
+      "learning_rate": 3.334317256969156e-05,
+      "loss": 0.4292,
       "step": 56895
     },
     {
-      "epoch": 2.0,
-      "learning_rate": 3.4079758157283526e-05,
-      "loss": 0.2452,
+      "epoch": 2.050672144736368,
+      "grad_norm": 0.20971627533435822,
+      "learning_rate": 3.334042167639157e-05,
+      "loss": 0.4386,
       "step": 56900
     },
     {
-      "epoch": 2.0,
-      "learning_rate": 3.407710390204699e-05,
-      "loss": 0.272,
+      "epoch": 2.0508523443975926,
+      "grad_norm": 0.17739515006542206,
+      "learning_rate": 3.3337670669454945e-05,
+      "loss": 0.3931,
       "step": 56905
     },
     {
-      "epoch": 2.0,
-      "learning_rate": 3.407444952895025e-05,
-      "loss": 0.2846,
+      "epoch": 2.0510325440588173,
+      "grad_norm": 0.24408304691314697,
+      "learning_rate": 3.3334919548919177e-05,
+      "loss": 0.3871,
       "step": 56910
     },
     {
-      "epoch": 2.0,
-      "learning_rate": 3.4071795038027755e-05,
-      "loss": 0.2791,
+      "epoch": 2.0512127437200416,
+      "grad_norm": 0.21893776953220367,
+      "learning_rate": 3.333216831482174e-05,
+      "loss": 0.381,
       "step": 56915
     },
     {
-      "epoch": 2.0,
-      "learning_rate": 3.406914042931398e-05,
-      "loss": 0.267,
+      "epoch": 2.0513929433812663,
+      "grad_norm": 0.19919031858444214,
+      "learning_rate": 3.332941696720012e-05,
+      "loss": 0.432,
       "step": 56920
     },
     {
-      "epoch": 2.0,
-      "learning_rate": 3.4066485702843385e-05,
-      "loss": 0.2816,
+      "epoch": 2.051573143042491,
+      "grad_norm": 0.18088489770889282,
+      "learning_rate": 3.3326665506091816e-05,
+      "loss": 0.3544,
       "step": 56925
     },
     {
-      "epoch": 2.0,
-      "learning_rate": 3.406383085865046e-05,
-      "loss": 0.2655,
+      "epoch": 2.051753342703716,
+      "grad_norm": 0.19796021282672882,
+      "learning_rate": 3.332391393153429e-05,
+      "loss": 0.3834,
       "step": 56930
     },
     {
-      "epoch": 2.0,
-      "learning_rate": 3.406117589676966e-05,
-      "loss": 0.2744,
+      "epoch": 2.0519335423649405,
+      "grad_norm": 0.15804626047611237,
+      "learning_rate": 3.3321162243565074e-05,
+      "loss": 0.351,
       "step": 56935
     },
     {
-      "epoch": 2.0,
-      "learning_rate": 3.405852081723547e-05,
-      "loss": 0.2504,
+      "epoch": 2.052113742026165,
+      "grad_norm": 0.18689633905887604,
+      "learning_rate": 3.3318410442221614e-05,
+      "loss": 0.4289,
       "step": 56940
     },
     {
-      "epoch": 2.0,
-      "learning_rate": 3.4055865620082346e-05,
-      "loss": 0.2855,
+      "epoch": 2.0522939416873895,
+      "grad_norm": 0.18491008877754211,
+      "learning_rate": 3.3315658527541425e-05,
+      "loss": 0.3782,
       "step": 56945
     },
     {
-      "epoch": 2.0,
-      "learning_rate": 3.405321030534479e-05,
-      "loss": 0.2821,
+      "epoch": 2.0524741413486143,
+      "grad_norm": 0.29741668701171875,
+      "learning_rate": 3.3312906499562e-05,
+      "loss": 0.4184,
       "step": 56950
     },
     {
-      "epoch": 2.0,
-      "learning_rate": 3.4050554873057245e-05,
-      "loss": 0.2763,
+      "epoch": 2.052654341009839,
+      "grad_norm": 0.20080535113811493,
+      "learning_rate": 3.331015435832084e-05,
+      "loss": 0.3833,
       "step": 56955
     },
     {
-      "epoch": 2.0,
-      "learning_rate": 3.404789932325422e-05,
-      "loss": 0.2822,
+      "epoch": 2.0528345406710637,
+      "grad_norm": 0.16579574346542358,
+      "learning_rate": 3.3307402103855426e-05,
+      "loss": 0.414,
       "step": 56960
     },
     {
-      "epoch": 2.0,
-      "learning_rate": 3.404524365597018e-05,
-      "loss": 0.2653,
+      "epoch": 2.053014740332288,
+      "grad_norm": 0.2213381975889206,
+      "learning_rate": 3.330464973620327e-05,
+      "loss": 0.3911,
       "step": 56965
     },
     {
-      "epoch": 2.0,
-      "learning_rate": 3.404258787123962e-05,
-      "loss": 0.2609,
+      "epoch": 2.0531949399935128,
+      "grad_norm": 0.1835223287343979,
+      "learning_rate": 3.330189725540187e-05,
+      "loss": 0.4325,
       "step": 56970
     },
     {
-      "epoch": 2.0,
-      "learning_rate": 3.4039931969097006e-05,
-      "loss": 0.2755,
+      "epoch": 2.0533751396547375,
+      "grad_norm": 0.18675781786441803,
+      "learning_rate": 3.329914466148872e-05,
+      "loss": 0.4228,
       "step": 56975
     },
     {
-      "epoch": 2.0,
-      "learning_rate": 3.403727594957684e-05,
-      "loss": 0.2875,
+      "epoch": 2.0535553393159622,
+      "grad_norm": 0.26438936591148376,
+      "learning_rate": 3.329639195450135e-05,
+      "loss": 0.4218,
       "step": 56980
     },
     {
-      "epoch": 2.0,
-      "learning_rate": 3.40346198127136e-05,
-      "loss": 0.2825,
+      "epoch": 2.0537355389771865,
+      "grad_norm": 0.1843392252922058,
+      "learning_rate": 3.329363913447723e-05,
+      "loss": 0.3871,
       "step": 56985
     },
     {
-      "epoch": 2.01,
-      "learning_rate": 3.403196355854178e-05,
-      "loss": 0.2678,
+      "epoch": 2.0539157386384113,
+      "grad_norm": 0.1797383427619934,
+      "learning_rate": 3.329088620145389e-05,
+      "loss": 0.3869,
       "step": 56990
     },
     {
-      "epoch": 2.01,
-      "learning_rate": 3.402930718709586e-05,
-      "loss": 0.2585,
+      "epoch": 2.054095938299636,
+      "grad_norm": 0.19929052889347076,
+      "learning_rate": 3.3288133155468826e-05,
+      "loss": 0.4024,
       "step": 56995
     },
     {
-      "epoch": 2.01,
-      "learning_rate": 3.402665069841034e-05,
-      "loss": 0.2683,
+      "epoch": 2.0542761379608607,
+      "grad_norm": 0.2309020608663559,
+      "learning_rate": 3.328537999655955e-05,
+      "loss": 0.394,
       "step": 57000
     },
     {
-      "epoch": 2.01,
-      "eval_loss": 0.271958589553833,
-      "eval_runtime": 10.5499,
-      "eval_samples_per_second": 9.479,
-      "eval_steps_per_second": 9.479,
+      "epoch": 2.0542761379608607,
+      "eval_loss": 0.43758705258369446,
+      "eval_runtime": 3.5257,
+      "eval_samples_per_second": 28.363,
+      "eval_steps_per_second": 7.091,
       "step": 57000
     },
     {
-      "epoch": 2.01,
-      "learning_rate": 3.402399409251971e-05,
-      "loss": 0.2793,
+      "epoch": 2.0544563376220855,
+      "grad_norm": 0.1692163497209549,
+      "learning_rate": 3.328262672476358e-05,
+      "loss": 0.3992,
       "step": 57005
     },
     {
-      "epoch": 2.01,
-      "learning_rate": 3.402133736945846e-05,
-      "loss": 0.2684,
+      "epoch": 2.0546365372833097,
+      "grad_norm": 0.2034154087305069,
+      "learning_rate": 3.327987334011842e-05,
+      "loss": 0.4069,
       "step": 57010
     },
     {
-      "epoch": 2.01,
-      "learning_rate": 3.401868052926108e-05,
-      "loss": 0.2791,
+      "epoch": 2.0548167369445345,
+      "grad_norm": 0.20764844119548798,
+      "learning_rate": 3.32771198426616e-05,
+      "loss": 0.3666,
       "step": 57015
     },
     {
-      "epoch": 2.01,
-      "learning_rate": 3.40160235719621e-05,
-      "loss": 0.2628,
+      "epoch": 2.054996936605759,
+      "grad_norm": 0.1821412444114685,
+      "learning_rate": 3.327436623243061e-05,
+      "loss": 0.4083,
       "step": 57020
     },
     {
-      "epoch": 2.01,
-      "learning_rate": 3.401336649759598e-05,
-      "loss": 0.2844,
+      "epoch": 2.055177136266984,
+      "grad_norm": 0.17971596121788025,
+      "learning_rate": 3.3271612509463e-05,
+      "loss": 0.4364,
       "step": 57025
     },
     {
-      "epoch": 2.01,
-      "learning_rate": 3.401070930619724e-05,
-      "loss": 0.2726,
+      "epoch": 2.0553573359282082,
+      "grad_norm": 0.17184744775295258,
+      "learning_rate": 3.326885867379625e-05,
+      "loss": 0.3749,
       "step": 57030
     },
     {
-      "epoch": 2.01,
-      "learning_rate": 3.400805199780037e-05,
-      "loss": 0.2613,
+      "epoch": 2.055537535589433,
+      "grad_norm": 0.20721890032291412,
+      "learning_rate": 3.326610472546792e-05,
+      "loss": 0.3955,
       "step": 57035
     },
     {
-      "epoch": 2.01,
-      "learning_rate": 3.400539457243989e-05,
-      "loss": 0.2656,
+      "epoch": 2.0557177352506577,
+      "grad_norm": 0.1466008871793747,
+      "learning_rate": 3.32633506645155e-05,
+      "loss": 0.4099,
       "step": 57040
     },
     {
-      "epoch": 2.01,
-      "learning_rate": 3.4002737030150294e-05,
-      "loss": 0.2687,
+      "epoch": 2.0558979349118824,
+      "grad_norm": 0.1782764494419098,
+      "learning_rate": 3.326059649097652e-05,
+      "loss": 0.4146,
       "step": 57045
     },
     {
-      "epoch": 2.01,
-      "learning_rate": 3.400007937096609e-05,
-      "loss": 0.2662,
+      "epoch": 2.056078134573107,
+      "grad_norm": 0.21149560809135437,
+      "learning_rate": 3.3257842204888536e-05,
+      "loss": 0.4045,
       "step": 57050
     },
     {
-      "epoch": 2.01,
-      "learning_rate": 3.3997421594921795e-05,
-      "loss": 0.2857,
+      "epoch": 2.0562583342343315,
+      "grad_norm": 0.1797577291727066,
+      "learning_rate": 3.325508780628903e-05,
+      "loss": 0.356,
       "step": 57055
     },
     {
-      "epoch": 2.01,
-      "learning_rate": 3.39947637020519e-05,
-      "loss": 0.2797,
+      "epoch": 2.056438533895556,
+      "grad_norm": 0.17433103919029236,
+      "learning_rate": 3.325233329521557e-05,
+      "loss": 0.4083,
       "step": 57060
     },
     {
-      "epoch": 2.01,
-      "learning_rate": 3.399210569239093e-05,
-      "loss": 0.2614,
+      "epoch": 2.056618733556781,
+      "grad_norm": 0.24138209223747253,
+      "learning_rate": 3.324957867170565e-05,
+      "loss": 0.3696,
       "step": 57065
     },
     {
-      "epoch": 2.01,
-      "learning_rate": 3.398944756597339e-05,
-      "loss": 0.282,
+      "epoch": 2.0567989332180057,
+      "grad_norm": 0.17933684587478638,
+      "learning_rate": 3.324682393579682e-05,
+      "loss": 0.4377,
       "step": 57070
     },
     {
-      "epoch": 2.01,
-      "learning_rate": 3.398678932283381e-05,
-      "loss": 0.2605,
+      "epoch": 2.0569791328792304,
+      "grad_norm": 0.21966612339019775,
+      "learning_rate": 3.324406908752662e-05,
+      "loss": 0.3704,
       "step": 57075
     },
     {
-      "epoch": 2.01,
-      "learning_rate": 3.398413096300668e-05,
-      "loss": 0.2592,
+      "epoch": 2.0571593325404547,
+      "grad_norm": 0.21100081503391266,
+      "learning_rate": 3.324131412693257e-05,
+      "loss": 0.382,
       "step": 57080
     },
     {
-      "epoch": 2.01,
-      "learning_rate": 3.398147248652653e-05,
-      "loss": 0.2464,
+      "epoch": 2.0573395322016794,
+      "grad_norm": 0.17946301400661469,
+      "learning_rate": 3.3238559054052206e-05,
+      "loss": 0.3886,
       "step": 57085
     },
     {
-      "epoch": 2.01,
-      "learning_rate": 3.3978813893427886e-05,
-      "loss": 0.2636,
+      "epoch": 2.057519731862904,
+      "grad_norm": 0.19525077939033508,
+      "learning_rate": 3.3235803868923077e-05,
+      "loss": 0.3696,
       "step": 57090
     },
     {
-      "epoch": 2.01,
-      "learning_rate": 3.397615518374526e-05,
-      "loss": 0.2526,
+      "epoch": 2.057699931524129,
+      "grad_norm": 0.22821834683418274,
+      "learning_rate": 3.3233048571582715e-05,
+      "loss": 0.4039,
       "step": 57095
     },
     {
-      "epoch": 2.01,
-      "learning_rate": 3.397349635751317e-05,
-      "loss": 0.2737,
+      "epoch": 2.057880131185353,
+      "grad_norm": 0.23264740407466888,
+      "learning_rate": 3.3230293162068655e-05,
+      "loss": 0.406,
       "step": 57100
     },
     {
-      "epoch": 2.01,
-      "learning_rate": 3.397083741476614e-05,
-      "loss": 0.2877,
+      "epoch": 2.058060330846578,
+      "grad_norm": 0.22399888932704926,
+      "learning_rate": 3.3227537640418445e-05,
+      "loss": 0.3937,
       "step": 57105
     },
     {
-      "epoch": 2.01,
-      "learning_rate": 3.396817835553871e-05,
-      "loss": 0.2706,
+      "epoch": 2.0582405305078026,
+      "grad_norm": 0.18442541360855103,
+      "learning_rate": 3.3224782006669624e-05,
+      "loss": 0.412,
       "step": 57110
     },
     {
-      "epoch": 2.01,
-      "learning_rate": 3.396551917986539e-05,
-      "loss": 0.2725,
+      "epoch": 2.0584207301690274,
+      "grad_norm": 0.23033219575881958,
+      "learning_rate": 3.3222026260859747e-05,
+      "loss": 0.3833,
       "step": 57115
     },
     {
-      "epoch": 2.01,
-      "learning_rate": 3.39628598877807e-05,
-      "loss": 0.2833,
+      "epoch": 2.058600929830252,
+      "grad_norm": 0.18063490092754364,
+      "learning_rate": 3.3219270403026354e-05,
+      "loss": 0.3814,
       "step": 57120
     },
     {
-      "epoch": 2.01,
-      "learning_rate": 3.39602004793192e-05,
-      "loss": 0.2687,
+      "epoch": 2.0587811294914764,
+      "grad_norm": 0.17391562461853027,
+      "learning_rate": 3.321651443320699e-05,
+      "loss": 0.3861,
       "step": 57125
     },
     {
-      "epoch": 2.01,
-      "learning_rate": 3.395754095451539e-05,
-      "loss": 0.2735,
+      "epoch": 2.058961329152701,
+      "grad_norm": 0.21318939328193665,
+      "learning_rate": 3.32137583514392e-05,
+      "loss": 0.4278,
       "step": 57130
     },
     {
-      "epoch": 2.01,
-      "learning_rate": 3.395488131340382e-05,
-      "loss": 0.2931,
+      "epoch": 2.059141528813926,
+      "grad_norm": 0.17975524067878723,
+      "learning_rate": 3.3211002157760554e-05,
+      "loss": 0.3949,
       "step": 57135
     },
     {
-      "epoch": 2.01,
-      "learning_rate": 3.395222155601902e-05,
-      "loss": 0.2854,
+      "epoch": 2.0593217284751506,
+      "grad_norm": 0.18724536895751953,
+      "learning_rate": 3.32082458522086e-05,
+      "loss": 0.3923,
       "step": 57140
     },
     {
-      "epoch": 2.01,
-      "learning_rate": 3.394956168239551e-05,
-      "loss": 0.2831,
+      "epoch": 2.059501928136375,
+      "grad_norm": 0.19911803305149078,
+      "learning_rate": 3.3205489434820876e-05,
+      "loss": 0.3708,
       "step": 57145
     },
     {
-      "epoch": 2.01,
-      "learning_rate": 3.3946901692567844e-05,
-      "loss": 0.2905,
+      "epoch": 2.0596821277975996,
+      "grad_norm": 0.21658672392368317,
+      "learning_rate": 3.3202732905634936e-05,
+      "loss": 0.4215,
       "step": 57150
     },
     {
-      "epoch": 2.01,
-      "learning_rate": 3.3944241586570565e-05,
-      "loss": 0.2555,
+      "epoch": 2.0598623274588244,
+      "grad_norm": 0.2002650946378708,
+      "learning_rate": 3.3199976264688365e-05,
+      "loss": 0.4014,
       "step": 57155
     },
     {
-      "epoch": 2.01,
-      "learning_rate": 3.394158136443819e-05,
-      "loss": 0.2624,
+      "epoch": 2.060042527120049,
+      "grad_norm": 0.20283368229866028,
+      "learning_rate": 3.3197219512018704e-05,
+      "loss": 0.4014,
       "step": 57160
     },
     {
-      "epoch": 2.01,
-      "learning_rate": 3.393892102620528e-05,
-      "loss": 0.2626,
+      "epoch": 2.060222726781274,
+      "grad_norm": 0.1828315407037735,
+      "learning_rate": 3.319446264766351e-05,
+      "loss": 0.4021,
       "step": 57165
     },
     {
-      "epoch": 2.01,
-      "learning_rate": 3.393626057190638e-05,
-      "loss": 0.28,
+      "epoch": 2.060402926442498,
+      "grad_norm": 0.1759759783744812,
+      "learning_rate": 3.319170567166034e-05,
+      "loss": 0.4176,
       "step": 57170
     },
     {
-      "epoch": 2.01,
-      "learning_rate": 3.393360000157601e-05,
-      "loss": 0.2743,
+      "epoch": 2.060583126103723,
+      "grad_norm": 0.19315093755722046,
+      "learning_rate": 3.318894858404678e-05,
+      "loss": 0.3938,
       "step": 57175
     },
     {
-      "epoch": 2.01,
-      "learning_rate": 3.393093931524875e-05,
-      "loss": 0.2711,
+      "epoch": 2.0607633257649476,
+      "grad_norm": 0.19143354892730713,
+      "learning_rate": 3.318619138486037e-05,
+      "loss": 0.3987,
       "step": 57180
     },
     {
-      "epoch": 2.01,
-      "learning_rate": 3.392827851295911e-05,
-      "loss": 0.2572,
+      "epoch": 2.0609435254261723,
+      "grad_norm": 0.21193043887615204,
+      "learning_rate": 3.31834340741387e-05,
+      "loss": 0.4053,
       "step": 57185
     },
     {
-      "epoch": 2.01,
-      "learning_rate": 3.392561759474167e-05,
-      "loss": 0.2988,
+      "epoch": 2.061123725087397,
+      "grad_norm": 0.17882151901721954,
+      "learning_rate": 3.318067665191932e-05,
+      "loss": 0.3777,
       "step": 57190
     },
     {
-      "epoch": 2.01,
-      "learning_rate": 3.392295656063096e-05,
-      "loss": 0.2747,
+      "epoch": 2.0613039247486213,
+      "grad_norm": 0.19089268147945404,
+      "learning_rate": 3.3177919118239794e-05,
+      "loss": 0.3783,
       "step": 57195
     },
     {
-      "epoch": 2.01,
-      "learning_rate": 3.392029541066154e-05,
-      "loss": 0.2537,
+      "epoch": 2.061484124409846,
+      "grad_norm": 0.1889115571975708,
+      "learning_rate": 3.317516147313771e-05,
+      "loss": 0.4081,
       "step": 57200
     },
     {
-      "epoch": 2.01,
-      "learning_rate": 3.3917634144867965e-05,
-      "loss": 0.2639,
+      "epoch": 2.061664324071071,
+      "grad_norm": 0.1909630447626114,
+      "learning_rate": 3.317240371665064e-05,
+      "loss": 0.3764,
       "step": 57205
     },
     {
-      "epoch": 2.01,
-      "learning_rate": 3.3914972763284794e-05,
-      "loss": 0.2387,
+      "epoch": 2.0618445237322955,
+      "grad_norm": 0.1816696971654892,
+      "learning_rate": 3.3169645848816146e-05,
+      "loss": 0.4064,
       "step": 57210
     },
     {
-      "epoch": 2.01,
-      "learning_rate": 3.391231126594657e-05,
-      "loss": 0.3002,
+      "epoch": 2.06202472339352,
+      "grad_norm": 0.15952745079994202,
+      "learning_rate": 3.316688786967181e-05,
+      "loss": 0.3982,
       "step": 57215
     },
     {
-      "epoch": 2.01,
-      "learning_rate": 3.390964965288786e-05,
-      "loss": 0.2646,
+      "epoch": 2.0622049230547446,
+      "grad_norm": 0.17730703949928284,
+      "learning_rate": 3.31641297792552e-05,
+      "loss": 0.421,
       "step": 57220
     },
     {
-      "epoch": 2.01,
-      "learning_rate": 3.390698792414322e-05,
-      "loss": 0.277,
+      "epoch": 2.0623851227159693,
+      "grad_norm": 0.19057762622833252,
+      "learning_rate": 3.316137157760392e-05,
+      "loss": 0.3998,
       "step": 57225
     },
     {
-      "epoch": 2.01,
-      "learning_rate": 3.390432607974721e-05,
-      "loss": 0.2545,
+      "epoch": 2.062565322377194,
+      "grad_norm": 0.20309016108512878,
+      "learning_rate": 3.315861326475552e-05,
+      "loss": 0.3808,
       "step": 57230
     },
     {
-      "epoch": 2.01,
-      "learning_rate": 3.390166411973439e-05,
-      "loss": 0.2576,
+      "epoch": 2.0627455220384188,
+      "grad_norm": 0.1632905900478363,
+      "learning_rate": 3.3155854840747596e-05,
+      "loss": 0.3994,
       "step": 57235
     },
     {
-      "epoch": 2.01,
-      "learning_rate": 3.389900204413934e-05,
-      "loss": 0.2717,
+      "epoch": 2.062925721699643,
+      "grad_norm": 0.20555098354816437,
+      "learning_rate": 3.315364802153166e-05,
+      "loss": 0.4086,
       "step": 57240
     },
     {
-      "epoch": 2.01,
-      "learning_rate": 3.3896339852996604e-05,
-      "loss": 0.2625,
+      "epoch": 2.063105921360868,
+      "grad_norm": 0.22128038108348846,
+      "learning_rate": 3.315088939753129e-05,
+      "loss": 0.4045,
       "step": 57245
     },
     {
-      "epoch": 2.01,
-      "learning_rate": 3.3893677546340765e-05,
-      "loss": 0.2338,
+      "epoch": 2.0632861210220925,
+      "grad_norm": 0.23114217817783356,
+      "learning_rate": 3.314813066247664e-05,
+      "loss": 0.3849,
       "step": 57250
     },
     {
-      "epoch": 2.01,
-      "learning_rate": 3.3891015124206376e-05,
-      "loss": 0.2643,
+      "epoch": 2.0634663206833173,
+      "grad_norm": 0.17400681972503662,
+      "learning_rate": 3.314537181640529e-05,
+      "loss": 0.3957,
       "step": 57255
     },
     {
-      "epoch": 2.01,
-      "learning_rate": 3.388835258662802e-05,
-      "loss": 0.2568,
+      "epoch": 2.0636465203445415,
+      "grad_norm": 0.18826141953468323,
+      "learning_rate": 3.314261285935483e-05,
+      "loss": 0.3964,
       "step": 57260
     },
     {
-      "epoch": 2.01,
-      "learning_rate": 3.388568993364026e-05,
-      "loss": 0.2942,
+      "epoch": 2.0638267200057663,
+      "grad_norm": 0.21883268654346466,
+      "learning_rate": 3.313985379136283e-05,
+      "loss": 0.4175,
       "step": 57265
     },
     {
-      "epoch": 2.01,
-      "learning_rate": 3.388302716527768e-05,
-      "loss": 0.2486,
+      "epoch": 2.064006919666991,
+      "grad_norm": 0.1698734164237976,
+      "learning_rate": 3.3137094612466914e-05,
+      "loss": 0.4036,
       "step": 57270
     },
     {
-      "epoch": 2.02,
-      "learning_rate": 3.3880364281574836e-05,
-      "loss": 0.2601,
+      "epoch": 2.0641871193282157,
+      "grad_norm": 0.17218467593193054,
+      "learning_rate": 3.3134335322704646e-05,
+      "loss": 0.3493,
       "step": 57275
     },
     {
-      "epoch": 2.02,
-      "learning_rate": 3.3877701282566314e-05,
-      "loss": 0.2685,
+      "epoch": 2.0643673189894405,
+      "grad_norm": 0.19277000427246094,
+      "learning_rate": 3.313157592211364e-05,
+      "loss": 0.4359,
       "step": 57280
     },
     {
-      "epoch": 2.02,
-      "learning_rate": 3.38750381682867e-05,
-      "loss": 0.255,
+      "epoch": 2.0645475186506648,
+      "grad_norm": 0.1942652016878128,
+      "learning_rate": 3.312881641073149e-05,
+      "loss": 0.4229,
       "step": 57285
     },
     {
-      "epoch": 2.02,
-      "learning_rate": 3.387237493877056e-05,
-      "loss": 0.2693,
+      "epoch": 2.0647277183118895,
+      "grad_norm": 0.23124489188194275,
+      "learning_rate": 3.3126056788595786e-05,
+      "loss": 0.38,
       "step": 57290
     },
     {
-      "epoch": 2.02,
-      "learning_rate": 3.386971159405248e-05,
-      "loss": 0.2861,
+      "epoch": 2.0649079179731142,
+      "grad_norm": 0.1828756481409073,
+      "learning_rate": 3.312329705574413e-05,
+      "loss": 0.4185,
       "step": 57295
     },
     {
-      "epoch": 2.02,
-      "learning_rate": 3.3867048134167035e-05,
-      "loss": 0.2617,
+      "epoch": 2.065088117634339,
+      "grad_norm": 0.18582990765571594,
+      "learning_rate": 3.3120537212214134e-05,
+      "loss": 0.376,
       "step": 57300
     },
     {
-      "epoch": 2.02,
-      "learning_rate": 3.386438455914882e-05,
-      "loss": 0.2772,
+      "epoch": 2.0652683172955637,
+      "grad_norm": 0.18808506429195404,
+      "learning_rate": 3.3117777258043384e-05,
+      "loss": 0.3729,
       "step": 57305
     },
     {
-      "epoch": 2.02,
-      "learning_rate": 3.386172086903241e-05,
-      "loss": 0.2497,
+      "epoch": 2.065448516956788,
+      "grad_norm": 0.2194932997226715,
+      "learning_rate": 3.311501719326948e-05,
+      "loss": 0.4381,
       "step": 57310
     },
     {
-      "epoch": 2.02,
-      "learning_rate": 3.38590570638524e-05,
-      "loss": 0.2822,
+      "epoch": 2.0656287166180127,
+      "grad_norm": 0.1534089893102646,
+      "learning_rate": 3.311225701793005e-05,
+      "loss": 0.3966,
       "step": 57315
     },
     {
-      "epoch": 2.02,
-      "learning_rate": 3.385639314364336e-05,
-      "loss": 0.2675,
+      "epoch": 2.0658089162792375,
+      "grad_norm": 0.19418910145759583,
+      "learning_rate": 3.3109496732062686e-05,
+      "loss": 0.4222,
       "step": 57320
     },
     {
-      "epoch": 2.02,
-      "learning_rate": 3.38537291084399e-05,
-      "loss": 0.2926,
+      "epoch": 2.065989115940462,
+      "grad_norm": 0.20485571026802063,
+      "learning_rate": 3.3106736335705e-05,
+      "loss": 0.3795,
       "step": 57325
     },
     {
-      "epoch": 2.02,
-      "learning_rate": 3.38510649582766e-05,
-      "loss": 0.2611,
+      "epoch": 2.0661693156016865,
+      "grad_norm": 0.198880136013031,
+      "learning_rate": 3.31039758288946e-05,
+      "loss": 0.4168,
       "step": 57330
     },
     {
-      "epoch": 2.02,
-      "learning_rate": 3.384840069318807e-05,
-      "loss": 0.2586,
+      "epoch": 2.066349515262911,
+      "grad_norm": 0.18784551322460175,
+      "learning_rate": 3.3101215211669104e-05,
+      "loss": 0.3876,
       "step": 57335
     },
     {
-      "epoch": 2.02,
-      "learning_rate": 3.384573631320886e-05,
-      "loss": 0.2682,
+      "epoch": 2.066529714924136,
+      "grad_norm": 0.1909671276807785,
+      "learning_rate": 3.309845448406612e-05,
+      "loss": 0.3914,
       "step": 57340
     },
     {
-      "epoch": 2.02,
-      "learning_rate": 3.3843071818373617e-05,
-      "loss": 0.2768,
+      "epoch": 2.0667099145853607,
+      "grad_norm": 0.18267397582530975,
+      "learning_rate": 3.309569364612325e-05,
+      "loss": 0.3791,
       "step": 57345
     },
     {
-      "epoch": 2.02,
-      "learning_rate": 3.38404072087169e-05,
-      "loss": 0.26,
+      "epoch": 2.0668901142465854,
+      "grad_norm": 0.16980095207691193,
+      "learning_rate": 3.3092932697878127e-05,
+      "loss": 0.3921,
       "step": 57350
     },
     {
-      "epoch": 2.02,
-      "learning_rate": 3.383774248427333e-05,
-      "loss": 0.2657,
+      "epoch": 2.0670703139078097,
+      "grad_norm": 0.2372709959745407,
+      "learning_rate": 3.3090171639368364e-05,
+      "loss": 0.4233,
       "step": 57355
     },
     {
-      "epoch": 2.02,
-      "learning_rate": 3.38350776450775e-05,
-      "loss": 0.2644,
+      "epoch": 2.0672505135690344,
+      "grad_norm": 0.2137833833694458,
+      "learning_rate": 3.3087410470631575e-05,
+      "loss": 0.4325,
       "step": 57360
     },
     {
-      "epoch": 2.02,
-      "learning_rate": 3.3832412691164e-05,
-      "loss": 0.2567,
+      "epoch": 2.067430713230259,
+      "grad_norm": 0.20736674964427948,
+      "learning_rate": 3.308464919170539e-05,
+      "loss": 0.4262,
       "step": 57365
     },
     {
-      "epoch": 2.02,
-      "learning_rate": 3.382974762256744e-05,
-      "loss": 0.2573,
+      "epoch": 2.067610912891484,
+      "grad_norm": 0.23613141477108002,
+      "learning_rate": 3.308188780262742e-05,
+      "loss": 0.3844,
       "step": 57370
     },
     {
-      "epoch": 2.02,
-      "learning_rate": 3.3827082439322445e-05,
-      "loss": 0.2531,
+      "epoch": 2.067791112552708,
+      "grad_norm": 0.23751363158226013,
+      "learning_rate": 3.30791263034353e-05,
+      "loss": 0.3884,
       "step": 57375
     },
     {
-      "epoch": 2.02,
-      "learning_rate": 3.382441714146358e-05,
-      "loss": 0.2777,
+      "epoch": 2.067971312213933,
+      "grad_norm": 0.22667773067951202,
+      "learning_rate": 3.307636469416664e-05,
+      "loss": 0.3871,
       "step": 57380
     },
     {
-      "epoch": 2.02,
-      "learning_rate": 3.382175172902549e-05,
-      "loss": 0.2603,
+      "epoch": 2.0681515118751577,
+      "grad_norm": 0.1814824640750885,
+      "learning_rate": 3.307360297485907e-05,
+      "loss": 0.3844,
       "step": 57385
     },
     {
-      "epoch": 2.02,
-      "learning_rate": 3.381908620204276e-05,
-      "loss": 0.2681,
+      "epoch": 2.0683317115363824,
+      "grad_norm": 0.19579993188381195,
+      "learning_rate": 3.307084114555023e-05,
+      "loss": 0.4153,
       "step": 57390
     },
     {
-      "epoch": 2.02,
-      "learning_rate": 3.3816420560550004e-05,
-      "loss": 0.2705,
+      "epoch": 2.068511911197607,
+      "grad_norm": 0.2431478351354599,
+      "learning_rate": 3.306807920627775e-05,
+      "loss": 0.3946,
       "step": 57395
     },
     {
-      "epoch": 2.02,
-      "learning_rate": 3.3813754804581844e-05,
-      "loss": 0.2606,
+      "epoch": 2.0686921108588314,
+      "grad_norm": 0.16477234661579132,
+      "learning_rate": 3.306531715707925e-05,
+      "loss": 0.3787,
       "step": 57400
     },
     {
-      "epoch": 2.02,
-      "learning_rate": 3.381108893417288e-05,
-      "loss": 0.2762,
+      "epoch": 2.068872310520056,
+      "grad_norm": 0.14775709807872772,
+      "learning_rate": 3.306255499799235e-05,
+      "loss": 0.3659,
       "step": 57405
     },
     {
-      "epoch": 2.02,
-      "learning_rate": 3.3808422949357746e-05,
-      "loss": 0.2848,
+      "epoch": 2.069052510181281,
+      "grad_norm": 0.2197529375553131,
+      "learning_rate": 3.305979272905472e-05,
+      "loss": 0.4249,
       "step": 57410
     },
     {
-      "epoch": 2.02,
-      "learning_rate": 3.3805756850171035e-05,
-      "loss": 0.2976,
+      "epoch": 2.0692327098425056,
+      "grad_norm": 0.1822102814912796,
+      "learning_rate": 3.3057030350303954e-05,
+      "loss": 0.4078,
       "step": 57415
     },
     {
-      "epoch": 2.02,
-      "learning_rate": 3.380309063664737e-05,
-      "loss": 0.2616,
+      "epoch": 2.06941290950373,
+      "grad_norm": 0.2318163812160492,
+      "learning_rate": 3.305426786177772e-05,
+      "loss": 0.4369,
       "step": 57420
     },
     {
-      "epoch": 2.02,
-      "learning_rate": 3.380042430882138e-05,
-      "loss": 0.2677,
+      "epoch": 2.0695931091649546,
+      "grad_norm": 0.1871294230222702,
+      "learning_rate": 3.305150526351365e-05,
+      "loss": 0.3798,
       "step": 57425
     },
     {
-      "epoch": 2.02,
-      "learning_rate": 3.379775786672768e-05,
-      "loss": 0.2842,
+      "epoch": 2.0697733088261794,
+      "grad_norm": 0.20110271871089935,
+      "learning_rate": 3.304874255554937e-05,
+      "loss": 0.3754,
       "step": 57430
     },
     {
-      "epoch": 2.02,
-      "learning_rate": 3.3795091310400895e-05,
-      "loss": 0.2763,
+      "epoch": 2.069953508487404,
+      "grad_norm": 0.17223583161830902,
+      "learning_rate": 3.304597973792254e-05,
+      "loss": 0.3773,
       "step": 57435
     },
     {
-      "epoch": 2.02,
-      "learning_rate": 3.379242463987564e-05,
-      "loss": 0.2363,
+      "epoch": 2.070133708148629,
+      "grad_norm": 0.19843506813049316,
+      "learning_rate": 3.304321681067079e-05,
+      "loss": 0.3627,
       "step": 57440
     },
     {
-      "epoch": 2.02,
-      "learning_rate": 3.378975785518655e-05,
-      "loss": 0.2622,
+      "epoch": 2.070313907809853,
+      "grad_norm": 0.21979078650474548,
+      "learning_rate": 3.304045377383177e-05,
+      "loss": 0.401,
       "step": 57445
     },
     {
-      "epoch": 2.02,
-      "learning_rate": 3.3787090956368244e-05,
-      "loss": 0.2869,
+      "epoch": 2.070494107471078,
+      "grad_norm": 0.18887178599834442,
+      "learning_rate": 3.303769062744312e-05,
+      "loss": 0.3996,
       "step": 57450
     },
     {
-      "epoch": 2.02,
-      "learning_rate": 3.3784423943455354e-05,
-      "loss": 0.2896,
+      "epoch": 2.0706743071323026,
+      "grad_norm": 0.19170895218849182,
+      "learning_rate": 3.303492737154249e-05,
+      "loss": 0.3618,
       "step": 57455
     },
     {
-      "epoch": 2.02,
-      "learning_rate": 3.378175681648251e-05,
-      "loss": 0.2588,
+      "epoch": 2.0708545067935273,
+      "grad_norm": 0.19946110248565674,
+      "learning_rate": 3.303216400616754e-05,
+      "loss": 0.3732,
       "step": 57460
     },
     {
-      "epoch": 2.02,
-      "learning_rate": 3.3779089575484335e-05,
-      "loss": 0.249,
+      "epoch": 2.071034706454752,
+      "grad_norm": 0.20270924270153046,
+      "learning_rate": 3.3029400531355897e-05,
+      "loss": 0.3963,
       "step": 57465
     },
     {
-      "epoch": 2.02,
-      "learning_rate": 3.377642222049548e-05,
-      "loss": 0.266,
+      "epoch": 2.0712149061159764,
+      "grad_norm": 0.24232977628707886,
+      "learning_rate": 3.302663694714524e-05,
+      "loss": 0.3853,
       "step": 57470
     },
     {
-      "epoch": 2.02,
-      "learning_rate": 3.3773754751550554e-05,
-      "loss": 0.279,
+      "epoch": 2.071395105777201,
+      "grad_norm": 0.17532899975776672,
+      "learning_rate": 3.30238732535732e-05,
+      "loss": 0.3874,
       "step": 57475
     },
     {
-      "epoch": 2.02,
-      "learning_rate": 3.377108716868421e-05,
-      "loss": 0.261,
+      "epoch": 2.071575305438426,
+      "grad_norm": 0.18288403749465942,
+      "learning_rate": 3.3021109450677445e-05,
+      "loss": 0.3988,
       "step": 57480
     },
     {
-      "epoch": 2.02,
-      "learning_rate": 3.376841947193108e-05,
-      "loss": 0.2634,
+      "epoch": 2.0717555050996506,
+      "grad_norm": 0.22599361836910248,
+      "learning_rate": 3.301834553849562e-05,
+      "loss": 0.4176,
       "step": 57485
     },
     {
-      "epoch": 2.02,
-      "learning_rate": 3.37657516613258e-05,
-      "loss": 0.279,
+      "epoch": 2.071935704760875,
+      "grad_norm": 0.20328468084335327,
+      "learning_rate": 3.30155815170654e-05,
+      "loss": 0.3994,
       "step": 57490
     },
     {
-      "epoch": 2.02,
-      "learning_rate": 3.376308373690301e-05,
-      "loss": 0.2691,
+      "epoch": 2.0721159044220996,
+      "grad_norm": 0.2231900542974472,
+      "learning_rate": 3.3012817386424416e-05,
+      "loss": 0.3832,
       "step": 57495
     },
     {
-      "epoch": 2.02,
-      "learning_rate": 3.376041569869735e-05,
-      "loss": 0.2587,
+      "epoch": 2.0722961040833243,
+      "grad_norm": 0.19576068222522736,
+      "learning_rate": 3.301005314661037e-05,
+      "loss": 0.3742,
       "step": 57500
     },
     {
-      "epoch": 2.02,
-      "eval_loss": 0.27126699686050415,
-      "eval_runtime": 10.5272,
-      "eval_samples_per_second": 9.499,
-      "eval_steps_per_second": 9.499,
+      "epoch": 2.0722961040833243,
+      "eval_loss": 0.43739715218544006,
+      "eval_runtime": 3.5417,
+      "eval_samples_per_second": 28.235,
+      "eval_steps_per_second": 7.059,
       "step": 57500
     },
     {
-      "epoch": 2.02,
-      "learning_rate": 3.375774754674347e-05,
-      "loss": 0.2689,
+      "epoch": 2.072476303744549,
+      "grad_norm": 0.18939457833766937,
+      "learning_rate": 3.3007288797660886e-05,
+      "loss": 0.3979,
       "step": 57505
     },
     {
-      "epoch": 2.02,
-      "learning_rate": 3.375507928107601e-05,
-      "loss": 0.2951,
+      "epoch": 2.072656503405774,
+      "grad_norm": 0.15962184965610504,
+      "learning_rate": 3.3004524339613636e-05,
+      "loss": 0.3678,
       "step": 57510
     },
     {
-      "epoch": 2.02,
-      "learning_rate": 3.375241090172961e-05,
-      "loss": 0.2765,
+      "epoch": 2.072836703066998,
+      "grad_norm": 0.20595142245292664,
+      "learning_rate": 3.30017597725063e-05,
+      "loss": 0.4048,
       "step": 57515
     },
     {
-      "epoch": 2.02,
-      "learning_rate": 3.3749742408738935e-05,
-      "loss": 0.2652,
+      "epoch": 2.073016902728223,
+      "grad_norm": 0.23062437772750854,
+      "learning_rate": 3.299899509637654e-05,
+      "loss": 0.3793,
       "step": 57520
     },
     {
-      "epoch": 2.02,
-      "learning_rate": 3.3747073802138616e-05,
-      "loss": 0.2837,
+      "epoch": 2.0731971023894475,
+      "grad_norm": 0.23280833661556244,
+      "learning_rate": 3.299623031126201e-05,
+      "loss": 0.4115,
       "step": 57525
     },
     {
-      "epoch": 2.02,
-      "learning_rate": 3.37444050819633e-05,
-      "loss": 0.2809,
+      "epoch": 2.0733773020506723,
+      "grad_norm": 0.18664391338825226,
+      "learning_rate": 3.2993465417200406e-05,
+      "loss": 0.4042,
       "step": 57530
     },
     {
-      "epoch": 2.02,
-      "learning_rate": 3.374173624824765e-05,
-      "loss": 0.2927,
+      "epoch": 2.0735575017118966,
+      "grad_norm": 0.19512483477592468,
+      "learning_rate": 3.299070041422937e-05,
+      "loss": 0.4043,
       "step": 57535
     },
     {
-      "epoch": 2.02,
-      "learning_rate": 3.373906730102632e-05,
-      "loss": 0.2504,
+      "epoch": 2.0737377013731213,
+      "grad_norm": 0.2118794023990631,
+      "learning_rate": 3.2987935302386594e-05,
+      "loss": 0.4164,
       "step": 57540
     },
     {
-      "epoch": 2.02,
-      "learning_rate": 3.373639824033396e-05,
-      "loss": 0.2715,
+      "epoch": 2.073917901034346,
+      "grad_norm": 0.20261244475841522,
+      "learning_rate": 3.298517008170974e-05,
+      "loss": 0.4024,
       "step": 57545
     },
     {
-      "epoch": 2.02,
-      "learning_rate": 3.3733729066205225e-05,
-      "loss": 0.2624,
+      "epoch": 2.0740981006955708,
+      "grad_norm": 0.18971537053585052,
+      "learning_rate": 3.29824047522365e-05,
+      "loss": 0.4133,
       "step": 57550
     },
     {
-      "epoch": 2.02,
-      "learning_rate": 3.373105977867478e-05,
-      "loss": 0.2941,
+      "epoch": 2.0742783003567955,
+      "grad_norm": 0.26411035656929016,
+      "learning_rate": 3.297963931400453e-05,
+      "loss": 0.397,
       "step": 57555
     },
     {
-      "epoch": 2.03,
-      "learning_rate": 3.3728390377777263e-05,
-      "loss": 0.2496,
+      "epoch": 2.07445850001802,
+      "grad_norm": 0.19595226645469666,
+      "learning_rate": 3.297687376705153e-05,
+      "loss": 0.4261,
       "step": 57560
     },
     {
-      "epoch": 2.03,
-      "learning_rate": 3.372572086354736e-05,
-      "loss": 0.2969,
+      "epoch": 2.0746386996792445,
+      "grad_norm": 0.18524418771266937,
+      "learning_rate": 3.297410811141516e-05,
+      "loss": 0.3958,
       "step": 57565
     },
     {
-      "epoch": 2.03,
-      "learning_rate": 3.372305123601972e-05,
-      "loss": 0.3003,
+      "epoch": 2.0748188993404693,
+      "grad_norm": 0.1883489340543747,
+      "learning_rate": 3.297134234713311e-05,
+      "loss": 0.3905,
       "step": 57570
     },
     {
-      "epoch": 2.03,
-      "learning_rate": 3.372038149522901e-05,
-      "loss": 0.2764,
+      "epoch": 2.074999099001694,
+      "grad_norm": 0.1916968673467636,
+      "learning_rate": 3.296857647424307e-05,
+      "loss": 0.3914,
       "step": 57575
     },
     {
-      "epoch": 2.03,
-      "learning_rate": 3.371771164120989e-05,
-      "loss": 0.2714,
+      "epoch": 2.0751792986629187,
+      "grad_norm": 0.19922398030757904,
+      "learning_rate": 3.296581049278272e-05,
+      "loss": 0.3591,
       "step": 57580
     },
     {
-      "epoch": 2.03,
-      "learning_rate": 3.3715041673997035e-05,
-      "loss": 0.2711,
+      "epoch": 2.075359498324143,
+      "grad_norm": 0.18488508462905884,
+      "learning_rate": 3.2963044402789736e-05,
+      "loss": 0.3581,
       "step": 57585
     },
     {
-      "epoch": 2.03,
-      "learning_rate": 3.37123715936251e-05,
-      "loss": 0.2781,
+      "epoch": 2.0755396979853677,
+      "grad_norm": 0.21221880614757538,
+      "learning_rate": 3.296027820430182e-05,
+      "loss": 0.4096,
       "step": 57590
     },
     {
-      "epoch": 2.03,
-      "learning_rate": 3.3709701400128765e-05,
-      "loss": 0.2579,
+      "epoch": 2.0757198976465925,
+      "grad_norm": 0.1988646239042282,
+      "learning_rate": 3.295751189735665e-05,
+      "loss": 0.3703,
       "step": 57595
     },
     {
-      "epoch": 2.03,
-      "learning_rate": 3.370703109354269e-05,
-      "loss": 0.2673,
+      "epoch": 2.075900097307817,
+      "grad_norm": 0.17219915986061096,
+      "learning_rate": 3.295474548199193e-05,
+      "loss": 0.3574,
       "step": 57600
     },
     {
-      "epoch": 2.03,
-      "learning_rate": 3.370436067390157e-05,
-      "loss": 0.2843,
+      "epoch": 2.0760802969690415,
+      "grad_norm": 0.21070365607738495,
+      "learning_rate": 3.2951978958245336e-05,
+      "loss": 0.4147,
       "step": 57605
     },
     {
-      "epoch": 2.03,
-      "learning_rate": 3.3701690141240056e-05,
-      "loss": 0.2998,
+      "epoch": 2.0762604966302662,
+      "grad_norm": 0.18376260995864868,
+      "learning_rate": 3.294921232615457e-05,
+      "loss": 0.4007,
       "step": 57610
     },
     {
-      "epoch": 2.03,
-      "learning_rate": 3.369901949559283e-05,
-      "loss": 0.2908,
+      "epoch": 2.076440696291491,
+      "grad_norm": 0.21397769451141357,
+      "learning_rate": 3.294644558575732e-05,
+      "loss": 0.3699,
       "step": 57615
     },
     {
-      "epoch": 2.03,
-      "learning_rate": 3.369634873699457e-05,
-      "loss": 0.276,
+      "epoch": 2.0766208959527157,
+      "grad_norm": 0.2026335448026657,
+      "learning_rate": 3.294367873709129e-05,
+      "loss": 0.4035,
       "step": 57620
     },
     {
-      "epoch": 2.03,
-      "learning_rate": 3.3693677865479945e-05,
-      "loss": 0.2981,
+      "epoch": 2.0768010956139404,
+      "grad_norm": 0.21820026636123657,
+      "learning_rate": 3.294091178019418e-05,
+      "loss": 0.3559,
       "step": 57625
     },
     {
-      "epoch": 2.03,
-      "learning_rate": 3.369100688108365e-05,
-      "loss": 0.2664,
+      "epoch": 2.0769812952751647,
+      "grad_norm": 0.1874457746744156,
+      "learning_rate": 3.293814471510368e-05,
+      "loss": 0.398,
       "step": 57630
     },
     {
-      "epoch": 2.03,
-      "learning_rate": 3.3688335783840355e-05,
-      "loss": 0.2672,
+      "epoch": 2.0771614949363895,
+      "grad_norm": 0.20042584836483002,
+      "learning_rate": 3.293537754185749e-05,
+      "loss": 0.366,
       "step": 57635
     },
     {
-      "epoch": 2.03,
-      "learning_rate": 3.368566457378474e-05,
-      "loss": 0.2837,
+      "epoch": 2.077341694597614,
+      "grad_norm": 0.24218867719173431,
+      "learning_rate": 3.293261026049333e-05,
+      "loss": 0.4413,
       "step": 57640
     },
     {
-      "epoch": 2.03,
-      "learning_rate": 3.36829932509515e-05,
-      "loss": 0.26,
+      "epoch": 2.077521894258839,
+      "grad_norm": 0.17626816034317017,
+      "learning_rate": 3.2929842871048885e-05,
+      "loss": 0.3598,
       "step": 57645
     },
     {
-      "epoch": 2.03,
-      "learning_rate": 3.368032181537532e-05,
-      "loss": 0.2865,
+      "epoch": 2.077702093920063,
+      "grad_norm": 0.2106868326663971,
+      "learning_rate": 3.292707537356186e-05,
+      "loss": 0.3975,
       "step": 57650
     },
     {
-      "epoch": 2.03,
-      "learning_rate": 3.3677650267090875e-05,
-      "loss": 0.3003,
+      "epoch": 2.077882293581288,
+      "grad_norm": 0.19412852823734283,
+      "learning_rate": 3.292430776806997e-05,
+      "loss": 0.3979,
       "step": 57655
     },
     {
-      "epoch": 2.03,
-      "learning_rate": 3.3674978606132866e-05,
-      "loss": 0.2764,
+      "epoch": 2.0780624932425127,
+      "grad_norm": 0.2266625612974167,
+      "learning_rate": 3.292154005461093e-05,
+      "loss": 0.4145,
       "step": 57660
     },
     {
-      "epoch": 2.03,
-      "learning_rate": 3.3672306832535974e-05,
-      "loss": 0.2719,
+      "epoch": 2.0782426929037374,
+      "grad_norm": 0.2035531848669052,
+      "learning_rate": 3.291877223322244e-05,
+      "loss": 0.3795,
       "step": 57665
     },
     {
-      "epoch": 2.03,
-      "learning_rate": 3.366963494633489e-05,
-      "loss": 0.269,
+      "epoch": 2.078422892564962,
+      "grad_norm": 0.22533543407917023,
+      "learning_rate": 3.29160043039422e-05,
+      "loss": 0.4332,
       "step": 57670
     },
     {
-      "epoch": 2.03,
-      "learning_rate": 3.366696294756431e-05,
-      "loss": 0.2546,
+      "epoch": 2.0786030922261864,
+      "grad_norm": 0.13781999051570892,
+      "learning_rate": 3.291323626680793e-05,
+      "loss": 0.3766,
       "step": 57675
     },
     {
-      "epoch": 2.03,
-      "learning_rate": 3.3664290836258934e-05,
-      "loss": 0.2789,
+      "epoch": 2.078783291887411,
+      "grad_norm": 0.20056338608264923,
+      "learning_rate": 3.291046812185736e-05,
+      "loss": 0.3835,
       "step": 57680
     },
     {
-      "epoch": 2.03,
-      "learning_rate": 3.3661618612453446e-05,
-      "loss": 0.2519,
+      "epoch": 2.078963491548636,
+      "grad_norm": 0.17711946368217468,
+      "learning_rate": 3.29076998691282e-05,
+      "loss": 0.4235,
       "step": 57685
     },
     {
-      "epoch": 2.03,
-      "learning_rate": 3.3658946276182556e-05,
-      "loss": 0.2899,
+      "epoch": 2.0791436912098606,
+      "grad_norm": 0.19499550759792328,
+      "learning_rate": 3.2904931508658154e-05,
+      "loss": 0.3613,
       "step": 57690
     },
     {
-      "epoch": 2.03,
-      "learning_rate": 3.3656273827480944e-05,
-      "loss": 0.2752,
+      "epoch": 2.079323890871085,
+      "grad_norm": 0.20534376800060272,
+      "learning_rate": 3.290216304048494e-05,
+      "loss": 0.3908,
       "step": 57695
     },
     {
-      "epoch": 2.03,
-      "learning_rate": 3.3653601266383326e-05,
-      "loss": 0.2817,
+      "epoch": 2.0795040905323097,
+      "grad_norm": 0.18097397685050964,
+      "learning_rate": 3.289939446464629e-05,
+      "loss": 0.409,
       "step": 57700
     },
     {
-      "epoch": 2.03,
-      "learning_rate": 3.3650928592924395e-05,
-      "loss": 0.2655,
+      "epoch": 2.0796842901935344,
+      "grad_norm": 0.18338914215564728,
+      "learning_rate": 3.289662578117992e-05,
+      "loss": 0.3821,
       "step": 57705
     },
     {
-      "epoch": 2.03,
-      "learning_rate": 3.3648255807138866e-05,
-      "loss": 0.2676,
+      "epoch": 2.079864489854759,
+      "grad_norm": 0.1914772242307663,
+      "learning_rate": 3.2893856990123544e-05,
+      "loss": 0.3561,
       "step": 57710
     },
     {
-      "epoch": 2.03,
-      "learning_rate": 3.364558290906143e-05,
-      "loss": 0.2963,
+      "epoch": 2.080044689515984,
+      "grad_norm": 0.25630638003349304,
+      "learning_rate": 3.2891088091514905e-05,
+      "loss": 0.3759,
       "step": 57715
     },
     {
-      "epoch": 2.03,
-      "learning_rate": 3.364290989872679e-05,
-      "loss": 0.2762,
+      "epoch": 2.080224889177208,
+      "grad_norm": 0.18817178905010223,
+      "learning_rate": 3.288831908539171e-05,
+      "loss": 0.3603,
       "step": 57720
     },
     {
-      "epoch": 2.03,
-      "learning_rate": 3.364023677616967e-05,
-      "loss": 0.2767,
+      "epoch": 2.080405088838433,
+      "grad_norm": 0.15790888667106628,
+      "learning_rate": 3.288554997179171e-05,
+      "loss": 0.3602,
       "step": 57725
     },
     {
-      "epoch": 2.03,
-      "learning_rate": 3.363756354142478e-05,
-      "loss": 0.2809,
+      "epoch": 2.0805852884996576,
+      "grad_norm": 0.17841728031635284,
+      "learning_rate": 3.2882780750752604e-05,
+      "loss": 0.3972,
       "step": 57730
     },
     {
-      "epoch": 2.03,
-      "learning_rate": 3.36348901945268e-05,
-      "loss": 0.2566,
+      "epoch": 2.0807654881608824,
+      "grad_norm": 0.20370912551879883,
+      "learning_rate": 3.288001142231214e-05,
+      "loss": 0.3692,
       "step": 57735
     },
     {
-      "epoch": 2.03,
-      "learning_rate": 3.363221673551047e-05,
-      "loss": 0.285,
+      "epoch": 2.080945687822107,
+      "grad_norm": 0.19954347610473633,
+      "learning_rate": 3.2877241986508045e-05,
+      "loss": 0.3974,
       "step": 57740
     },
     {
-      "epoch": 2.03,
-      "learning_rate": 3.3629543164410496e-05,
-      "loss": 0.247,
+      "epoch": 2.0811258874833314,
+      "grad_norm": 0.1677468866109848,
+      "learning_rate": 3.2874472443378056e-05,
+      "loss": 0.3678,
       "step": 57745
     },
     {
-      "epoch": 2.03,
-      "learning_rate": 3.3626869481261586e-05,
-      "loss": 0.2778,
+      "epoch": 2.081306087144556,
+      "grad_norm": 0.2071567326784134,
+      "learning_rate": 3.287170279295991e-05,
+      "loss": 0.3873,
       "step": 57750
     },
     {
-      "epoch": 2.03,
-      "learning_rate": 3.3624195686098465e-05,
-      "loss": 0.2722,
+      "epoch": 2.081486286805781,
+      "grad_norm": 0.1877603381872177,
+      "learning_rate": 3.286893303529132e-05,
+      "loss": 0.3882,
       "step": 57755
     },
     {
-      "epoch": 2.03,
-      "learning_rate": 3.362152177895584e-05,
-      "loss": 0.2861,
+      "epoch": 2.0816664864670056,
+      "grad_norm": 0.20741944015026093,
+      "learning_rate": 3.286616317041006e-05,
+      "loss": 0.4103,
       "step": 57760
     },
     {
-      "epoch": 2.03,
-      "learning_rate": 3.3618847759868445e-05,
-      "loss": 0.2812,
+      "epoch": 2.08184668612823,
+      "grad_norm": 0.18143180012702942,
+      "learning_rate": 3.286339319835384e-05,
+      "loss": 0.4236,
       "step": 57765
     },
     {
-      "epoch": 2.03,
-      "learning_rate": 3.361617362887099e-05,
-      "loss": 0.252,
+      "epoch": 2.0820268857894546,
+      "grad_norm": 0.19399969279766083,
+      "learning_rate": 3.286062311916041e-05,
+      "loss": 0.3963,
       "step": 57770
     },
     {
-      "epoch": 2.03,
-      "learning_rate": 3.3613499385998196e-05,
-      "loss": 0.2673,
+      "epoch": 2.0822070854506793,
+      "grad_norm": 0.19400350749492645,
+      "learning_rate": 3.285785293286751e-05,
+      "loss": 0.4274,
       "step": 57775
     },
     {
-      "epoch": 2.03,
-      "learning_rate": 3.3610825031284784e-05,
-      "loss": 0.2857,
+      "epoch": 2.082387285111904,
+      "grad_norm": 0.21012353897094727,
+      "learning_rate": 3.285508263951289e-05,
+      "loss": 0.4118,
       "step": 57780
     },
     {
-      "epoch": 2.03,
-      "learning_rate": 3.36081505647655e-05,
-      "loss": 0.2488,
+      "epoch": 2.082567484773129,
+      "grad_norm": 0.21954594552516937,
+      "learning_rate": 3.285231223913429e-05,
+      "loss": 0.4276,
       "step": 57785
     },
     {
-      "epoch": 2.03,
-      "learning_rate": 3.360547598647504e-05,
-      "loss": 0.2621,
+      "epoch": 2.082747684434353,
+      "grad_norm": 0.1681157350540161,
+      "learning_rate": 3.284954173176945e-05,
+      "loss": 0.3828,
       "step": 57790
     },
     {
-      "epoch": 2.03,
-      "learning_rate": 3.3602801296448154e-05,
-      "loss": 0.2739,
+      "epoch": 2.082927884095578,
+      "grad_norm": 0.19216038286685944,
+      "learning_rate": 3.2846771117456134e-05,
+      "loss": 0.3608,
       "step": 57795
     },
     {
-      "epoch": 2.03,
-      "learning_rate": 3.3600126494719555e-05,
-      "loss": 0.269,
+      "epoch": 2.0831080837568026,
+      "grad_norm": 0.20407375693321228,
+      "learning_rate": 3.2844000396232064e-05,
+      "loss": 0.3638,
       "step": 57800
     },
     {
-      "epoch": 2.03,
-      "learning_rate": 3.359745158132399e-05,
-      "loss": 0.2596,
+      "epoch": 2.0832882834180273,
+      "grad_norm": 0.186093270778656,
+      "learning_rate": 3.284122956813503e-05,
+      "loss": 0.4093,
       "step": 57805
     },
     {
-      "epoch": 2.03,
-      "learning_rate": 3.359477655629617e-05,
-      "loss": 0.3004,
+      "epoch": 2.083468483079252,
+      "grad_norm": 0.19043242931365967,
+      "learning_rate": 3.2838458633202745e-05,
+      "loss": 0.4267,
       "step": 57810
     },
     {
-      "epoch": 2.03,
-      "learning_rate": 3.359210141967085e-05,
-      "loss": 0.2488,
+      "epoch": 2.0836486827404763,
+      "grad_norm": 0.22312848269939423,
+      "learning_rate": 3.283568759147298e-05,
+      "loss": 0.3784,
       "step": 57815
     },
     {
-      "epoch": 2.03,
-      "learning_rate": 3.3589426171482754e-05,
-      "loss": 0.2864,
+      "epoch": 2.083828882401701,
+      "grad_norm": 0.207689568400383,
+      "learning_rate": 3.283291644298349e-05,
+      "loss": 0.3612,
       "step": 57820
     },
     {
-      "epoch": 2.03,
-      "learning_rate": 3.358675081176662e-05,
-      "loss": 0.2883,
+      "epoch": 2.084009082062926,
+      "grad_norm": 0.1909976601600647,
+      "learning_rate": 3.283014518777203e-05,
+      "loss": 0.412,
       "step": 57825
     },
     {
-      "epoch": 2.03,
-      "learning_rate": 3.358407534055718e-05,
-      "loss": 0.2591,
+      "epoch": 2.0841892817241505,
+      "grad_norm": 0.17157649993896484,
+      "learning_rate": 3.2827373825876364e-05,
+      "loss": 0.4281,
       "step": 57830
     },
     {
-      "epoch": 2.03,
-      "learning_rate": 3.358139975788918e-05,
-      "loss": 0.2733,
+      "epoch": 2.084369481385375,
+      "grad_norm": 0.20599566400051117,
+      "learning_rate": 3.282460235733424e-05,
+      "loss": 0.4165,
       "step": 57835
     },
     {
-      "epoch": 2.03,
-      "learning_rate": 3.357872406379736e-05,
-      "loss": 0.2702,
+      "epoch": 2.0845496810465995,
+      "grad_norm": 0.23705652356147766,
+      "learning_rate": 3.282183078218342e-05,
+      "loss": 0.4269,
       "step": 57840
     },
     {
-      "epoch": 2.04,
-      "learning_rate": 3.357604825831647e-05,
-      "loss": 0.2588,
+      "epoch": 2.0847298807078243,
+      "grad_norm": 0.17649759352207184,
+      "learning_rate": 3.281905910046167e-05,
+      "loss": 0.3876,
       "step": 57845
     },
     {
-      "epoch": 2.04,
-      "learning_rate": 3.357337234148124e-05,
-      "loss": 0.2755,
+      "epoch": 2.084910080369049,
+      "grad_norm": 0.16404461860656738,
+      "learning_rate": 3.281628731220676e-05,
+      "loss": 0.3754,
       "step": 57850
     },
     {
-      "epoch": 2.04,
-      "learning_rate": 3.357069631332642e-05,
-      "loss": 0.2558,
+      "epoch": 2.0850902800302737,
+      "grad_norm": 0.1914927363395691,
+      "learning_rate": 3.2813515417456456e-05,
+      "loss": 0.4207,
       "step": 57855
     },
     {
-      "epoch": 2.04,
-      "learning_rate": 3.356802017388675e-05,
-      "loss": 0.2694,
+      "epoch": 2.085270479691498,
+      "grad_norm": 0.19702866673469543,
+      "learning_rate": 3.2810743416248496e-05,
+      "loss": 0.4151,
       "step": 57860
     },
     {
-      "epoch": 2.04,
-      "learning_rate": 3.356534392319699e-05,
-      "loss": 0.2742,
+      "epoch": 2.0854506793527228,
+      "grad_norm": 0.19254851341247559,
+      "learning_rate": 3.2807971308620685e-05,
+      "loss": 0.389,
       "step": 57865
     },
     {
-      "epoch": 2.04,
-      "learning_rate": 3.356266756129189e-05,
-      "loss": 0.2663,
+      "epoch": 2.0856308790139475,
+      "grad_norm": 0.2003370225429535,
+      "learning_rate": 3.2805199094610774e-05,
+      "loss": 0.3934,
       "step": 57870
     },
     {
-      "epoch": 2.04,
-      "learning_rate": 3.355999108820619e-05,
-      "loss": 0.2779,
+      "epoch": 2.0858110786751722,
+      "grad_norm": 0.18016155064105988,
+      "learning_rate": 3.2802426774256534e-05,
+      "loss": 0.3857,
       "step": 57875
     },
     {
-      "epoch": 2.04,
-      "learning_rate": 3.3557314503974646e-05,
-      "loss": 0.3263,
+      "epoch": 2.0859912783363965,
+      "grad_norm": 0.16226695477962494,
+      "learning_rate": 3.2799654347595736e-05,
+      "loss": 0.4044,
       "step": 57880
     },
     {
-      "epoch": 2.04,
-      "learning_rate": 3.355463780863201e-05,
-      "loss": 0.2674,
+      "epoch": 2.0861714779976213,
+      "grad_norm": 0.1907840520143509,
+      "learning_rate": 3.2796881814666164e-05,
+      "loss": 0.4128,
       "step": 57885
     },
     {
-      "epoch": 2.04,
-      "learning_rate": 3.355196100221304e-05,
-      "loss": 0.2731,
+      "epoch": 2.086351677658846,
+      "grad_norm": 0.19597387313842773,
+      "learning_rate": 3.279410917550559e-05,
+      "loss": 0.4094,
       "step": 57890
     },
     {
-      "epoch": 2.04,
-      "learning_rate": 3.3549284084752494e-05,
-      "loss": 0.2736,
+      "epoch": 2.0865318773200707,
+      "grad_norm": 0.2085808515548706,
+      "learning_rate": 3.2791336430151775e-05,
+      "loss": 0.416,
       "step": 57895
     },
     {
-      "epoch": 2.04,
-      "learning_rate": 3.354660705628514e-05,
-      "loss": 0.2806,
+      "epoch": 2.0867120769812955,
+      "grad_norm": 0.22985954582691193,
+      "learning_rate": 3.278856357864252e-05,
+      "loss": 0.4266,
       "step": 57900
     },
     {
-      "epoch": 2.04,
-      "learning_rate": 3.354392991684571e-05,
-      "loss": 0.2514,
+      "epoch": 2.0868922766425198,
+      "grad_norm": 0.2437005490064621,
+      "learning_rate": 3.278579062101559e-05,
+      "loss": 0.4272,
       "step": 57905
     },
     {
-      "epoch": 2.04,
-      "learning_rate": 3.354125266646899e-05,
-      "loss": 0.2987,
+      "epoch": 2.0870724763037445,
+      "grad_norm": 0.1723114401102066,
+      "learning_rate": 3.278301755730878e-05,
+      "loss": 0.3741,
       "step": 57910
     },
     {
-      "epoch": 2.04,
-      "learning_rate": 3.353857530518972e-05,
-      "loss": 0.2642,
+      "epoch": 2.087252675964969,
+      "grad_norm": 0.1853064000606537,
+      "learning_rate": 3.2780244387559846e-05,
+      "loss": 0.4152,
       "step": 57915
     },
     {
-      "epoch": 2.04,
-      "learning_rate": 3.353589783304269e-05,
-      "loss": 0.258,
+      "epoch": 2.087432875626194,
+      "grad_norm": 0.2420477718114853,
+      "learning_rate": 3.27774711118066e-05,
+      "loss": 0.3977,
       "step": 57920
     },
     {
-      "epoch": 2.04,
-      "learning_rate": 3.3533220250062646e-05,
-      "loss": 0.2822,
+      "epoch": 2.0876130752874182,
+      "grad_norm": 0.23371827602386475,
+      "learning_rate": 3.277469773008681e-05,
+      "loss": 0.4223,
       "step": 57925
     },
     {
-      "epoch": 2.04,
-      "learning_rate": 3.3530542556284363e-05,
-      "loss": 0.2591,
+      "epoch": 2.087793274948643,
+      "grad_norm": 0.1876314878463745,
+      "learning_rate": 3.277192424243827e-05,
+      "loss": 0.4122,
       "step": 57930
     },
     {
-      "epoch": 2.04,
-      "learning_rate": 3.352786475174261e-05,
-      "loss": 0.2674,
+      "epoch": 2.0879734746098677,
+      "grad_norm": 0.1702897995710373,
+      "learning_rate": 3.276915064889877e-05,
+      "loss": 0.3766,
       "step": 57935
     },
     {
-      "epoch": 2.04,
-      "learning_rate": 3.3525186836472144e-05,
-      "loss": 0.2903,
+      "epoch": 2.0881536742710924,
+      "grad_norm": 0.22066879272460938,
+      "learning_rate": 3.2766376949506085e-05,
+      "loss": 0.4061,
       "step": 57940
     },
     {
-      "epoch": 2.04,
-      "learning_rate": 3.3522508810507746e-05,
-      "loss": 0.2626,
+      "epoch": 2.088333873932317,
+      "grad_norm": 0.19222991168498993,
+      "learning_rate": 3.2763603144298026e-05,
+      "loss": 0.372,
       "step": 57945
     },
     {
-      "epoch": 2.04,
-      "learning_rate": 3.35198306738842e-05,
-      "loss": 0.2595,
+      "epoch": 2.0885140735935415,
+      "grad_norm": 0.17189811170101166,
+      "learning_rate": 3.2760829233312385e-05,
+      "loss": 0.3949,
       "step": 57950
     },
     {
-      "epoch": 2.04,
-      "learning_rate": 3.351715242663626e-05,
-      "loss": 0.2622,
+      "epoch": 2.088694273254766,
+      "grad_norm": 0.19438187777996063,
+      "learning_rate": 3.275805521658694e-05,
+      "loss": 0.4451,
       "step": 57955
     },
     {
-      "epoch": 2.04,
-      "learning_rate": 3.351447406879871e-05,
-      "loss": 0.2506,
+      "epoch": 2.088874472915991,
+      "grad_norm": 0.17549294233322144,
+      "learning_rate": 3.27552810941595e-05,
+      "loss": 0.39,
       "step": 57960
     },
     {
-      "epoch": 2.04,
-      "learning_rate": 3.351179560040632e-05,
-      "loss": 0.2667,
+      "epoch": 2.0890546725772157,
+      "grad_norm": 0.18093203008174896,
+      "learning_rate": 3.275250686606784e-05,
+      "loss": 0.3649,
       "step": 57965
     },
     {
-      "epoch": 2.04,
-      "learning_rate": 3.350911702149388e-05,
-      "loss": 0.2623,
+      "epoch": 2.0892348722384404,
+      "grad_norm": 0.17545856535434723,
+      "learning_rate": 3.2749732532349796e-05,
+      "loss": 0.4106,
       "step": 57970
     },
     {
-      "epoch": 2.04,
-      "learning_rate": 3.350643833209616e-05,
-      "loss": 0.285,
+      "epoch": 2.0894150718996647,
+      "grad_norm": 0.21285952627658844,
+      "learning_rate": 3.2746958093043136e-05,
+      "loss": 0.3845,
       "step": 57975
     },
     {
-      "epoch": 2.04,
-      "learning_rate": 3.350375953224794e-05,
-      "loss": 0.2561,
+      "epoch": 2.0895952715608894,
+      "grad_norm": 0.2009621560573578,
+      "learning_rate": 3.2744183548185674e-05,
+      "loss": 0.3827,
       "step": 57980
     },
     {
-      "epoch": 2.04,
-      "learning_rate": 3.3501080621984013e-05,
-      "loss": 0.2873,
+      "epoch": 2.089775471222114,
+      "grad_norm": 0.21972990036010742,
+      "learning_rate": 3.274140889781521e-05,
+      "loss": 0.367,
       "step": 57985
     },
     {
-      "epoch": 2.04,
-      "learning_rate": 3.3498401601339156e-05,
-      "loss": 0.2427,
+      "epoch": 2.089955670883339,
+      "grad_norm": 0.1965899020433426,
+      "learning_rate": 3.273863414196955e-05,
+      "loss": 0.3854,
       "step": 57990
     },
     {
-      "epoch": 2.04,
-      "learning_rate": 3.349572247034816e-05,
-      "loss": 0.2907,
+      "epoch": 2.090135870544563,
+      "grad_norm": 0.17252902686595917,
+      "learning_rate": 3.27358592806865e-05,
+      "loss": 0.4063,
       "step": 57995
     },
     {
-      "epoch": 2.04,
-      "learning_rate": 3.34930432290458e-05,
-      "loss": 0.3012,
+      "epoch": 2.090316070205788,
+      "grad_norm": 0.1891898810863495,
+      "learning_rate": 3.273308431400386e-05,
+      "loss": 0.4243,
       "step": 58000
     },
     {
-      "epoch": 2.04,
-      "eval_loss": 0.2711896598339081,
-      "eval_runtime": 10.5364,
-      "eval_samples_per_second": 9.491,
-      "eval_steps_per_second": 9.491,
+      "epoch": 2.090316070205788,
+      "eval_loss": 0.4369398057460785,
+      "eval_runtime": 3.5377,
+      "eval_samples_per_second": 28.267,
+      "eval_steps_per_second": 7.067,
       "step": 58000
     },
     {
-      "epoch": 2.04,
-      "learning_rate": 3.349036387746687e-05,
-      "loss": 0.2664,
+      "epoch": 2.0904962698670126,
+      "grad_norm": 0.17718040943145752,
+      "learning_rate": 3.2730309241959446e-05,
+      "loss": 0.4142,
       "step": 58005
     },
     {
-      "epoch": 2.04,
-      "learning_rate": 3.348768441564616e-05,
-      "loss": 0.2721,
+      "epoch": 2.0906764695282374,
+      "grad_norm": 0.16075636446475983,
+      "learning_rate": 3.272753406459106e-05,
+      "loss": 0.4091,
       "step": 58010
     },
     {
-      "epoch": 2.04,
-      "learning_rate": 3.348500484361846e-05,
-      "loss": 0.2644,
+      "epoch": 2.090856669189462,
+      "grad_norm": 0.21720968186855316,
+      "learning_rate": 3.272475878193653e-05,
+      "loss": 0.404,
       "step": 58015
     },
     {
-      "epoch": 2.04,
-      "learning_rate": 3.348232516141857e-05,
-      "loss": 0.2656,
+      "epoch": 2.0910368688506864,
+      "grad_norm": 0.19920498132705688,
+      "learning_rate": 3.2721983394033645e-05,
+      "loss": 0.39,
       "step": 58020
     },
     {
-      "epoch": 2.04,
-      "learning_rate": 3.347964536908128e-05,
-      "loss": 0.2686,
+      "epoch": 2.091217068511911,
+      "grad_norm": 0.17115192115306854,
+      "learning_rate": 3.271920790092024e-05,
+      "loss": 0.398,
       "step": 58025
     },
     {
-      "epoch": 2.04,
-      "learning_rate": 3.347696546664137e-05,
-      "loss": 0.2754,
+      "epoch": 2.091397268173136,
+      "grad_norm": 0.15494772791862488,
+      "learning_rate": 3.2716432302634116e-05,
+      "loss": 0.4062,
       "step": 58030
     },
     {
-      "epoch": 2.04,
-      "learning_rate": 3.347428545413366e-05,
-      "loss": 0.2692,
+      "epoch": 2.0915774678343606,
+      "grad_norm": 0.1952543705701828,
+      "learning_rate": 3.2713656599213094e-05,
+      "loss": 0.4006,
       "step": 58035
     },
     {
-      "epoch": 2.04,
-      "learning_rate": 3.3471605331592935e-05,
-      "loss": 0.2756,
+      "epoch": 2.091757667495585,
+      "grad_norm": 0.2297617346048355,
+      "learning_rate": 3.2710880790695e-05,
+      "loss": 0.393,
       "step": 58040
     },
     {
-      "epoch": 2.04,
-      "learning_rate": 3.346892509905401e-05,
-      "loss": 0.2624,
+      "epoch": 2.0919378671568096,
+      "grad_norm": 0.18533745408058167,
+      "learning_rate": 3.2708104877117644e-05,
+      "loss": 0.375,
       "step": 58045
     },
     {
-      "epoch": 2.04,
-      "learning_rate": 3.3466244756551656e-05,
-      "loss": 0.2684,
+      "epoch": 2.0921180668180344,
+      "grad_norm": 0.19725355505943298,
+      "learning_rate": 3.270532885851886e-05,
+      "loss": 0.3976,
       "step": 58050
     },
     {
-      "epoch": 2.04,
-      "learning_rate": 3.34635643041207e-05,
-      "loss": 0.2734,
+      "epoch": 2.092298266479259,
+      "grad_norm": 0.2088615596294403,
+      "learning_rate": 3.2702552734936454e-05,
+      "loss": 0.423,
       "step": 58055
     },
     {
-      "epoch": 2.04,
-      "learning_rate": 3.346088374179593e-05,
-      "loss": 0.2706,
+      "epoch": 2.092478466140484,
+      "grad_norm": 0.1870979368686676,
+      "learning_rate": 3.2699776506408266e-05,
+      "loss": 0.3923,
       "step": 58060
     },
     {
-      "epoch": 2.04,
-      "learning_rate": 3.345820306961218e-05,
-      "loss": 0.2608,
+      "epoch": 2.092658665801708,
+      "grad_norm": 0.18204568326473236,
+      "learning_rate": 3.2697000172972106e-05,
+      "loss": 0.3888,
       "step": 58065
     },
     {
-      "epoch": 2.04,
-      "learning_rate": 3.345552228760422e-05,
-      "loss": 0.2507,
+      "epoch": 2.092838865462933,
+      "grad_norm": 0.16376551985740662,
+      "learning_rate": 3.2694223734665806e-05,
+      "loss": 0.3746,
       "step": 58070
     },
     {
-      "epoch": 2.04,
-      "learning_rate": 3.3452841395806887e-05,
-      "loss": 0.2956,
+      "epoch": 2.0930190651241576,
+      "grad_norm": 0.21093137562274933,
+      "learning_rate": 3.26914471915272e-05,
+      "loss": 0.4078,
       "step": 58075
     },
     {
-      "epoch": 2.04,
-      "learning_rate": 3.345016039425497e-05,
-      "loss": 0.2717,
+      "epoch": 2.0931992647853823,
+      "grad_norm": 0.199448361992836,
+      "learning_rate": 3.2688670543594116e-05,
+      "loss": 0.3742,
       "step": 58080
     },
     {
-      "epoch": 2.04,
-      "learning_rate": 3.34474792829833e-05,
-      "loss": 0.2607,
+      "epoch": 2.093379464446607,
+      "grad_norm": 0.21736840903759003,
+      "learning_rate": 3.268589379090439e-05,
+      "loss": 0.3834,
       "step": 58085
     },
     {
-      "epoch": 2.04,
-      "learning_rate": 3.3444798062026674e-05,
-      "loss": 0.2719,
+      "epoch": 2.0935596641078313,
+      "grad_norm": 0.22230195999145508,
+      "learning_rate": 3.2683116933495844e-05,
+      "loss": 0.4269,
       "step": 58090
     },
     {
-      "epoch": 2.04,
-      "learning_rate": 3.344211673141991e-05,
-      "loss": 0.2629,
+      "epoch": 2.093739863769056,
+      "grad_norm": 0.20408831536769867,
+      "learning_rate": 3.2680339971406325e-05,
+      "loss": 0.4019,
       "step": 58095
     },
     {
-      "epoch": 2.04,
-      "learning_rate": 3.3439435291197824e-05,
-      "loss": 0.2764,
+      "epoch": 2.093920063430281,
+      "grad_norm": 0.25087887048721313,
+      "learning_rate": 3.267756290467365e-05,
+      "loss": 0.4346,
       "step": 58100
     },
     {
-      "epoch": 2.04,
-      "learning_rate": 3.343675374139523e-05,
-      "loss": 0.2804,
+      "epoch": 2.0941002630915055,
+      "grad_norm": 0.2030712068080902,
+      "learning_rate": 3.267478573333567e-05,
+      "loss": 0.3983,
       "step": 58105
     },
     {
-      "epoch": 2.04,
-      "learning_rate": 3.343407208204695e-05,
-      "loss": 0.2788,
+      "epoch": 2.09428046275273,
+      "grad_norm": 0.24952729046344757,
+      "learning_rate": 3.267200845743022e-05,
+      "loss": 0.3967,
       "step": 58110
     },
     {
-      "epoch": 2.04,
-      "learning_rate": 3.34313903131878e-05,
-      "loss": 0.2597,
+      "epoch": 2.0944606624139546,
+      "grad_norm": 0.16245847940444946,
+      "learning_rate": 3.2669231076995146e-05,
+      "loss": 0.3667,
       "step": 58115
     },
     {
-      "epoch": 2.04,
-      "learning_rate": 3.342870843485261e-05,
-      "loss": 0.2737,
+      "epoch": 2.0946408620751793,
+      "grad_norm": 0.22543512284755707,
+      "learning_rate": 3.266645359206827e-05,
+      "loss": 0.3892,
       "step": 58120
     },
     {
-      "epoch": 2.04,
-      "learning_rate": 3.3426026447076186e-05,
-      "loss": 0.2646,
+      "epoch": 2.094821061736404,
+      "grad_norm": 0.1827520877122879,
+      "learning_rate": 3.266367600268746e-05,
+      "loss": 0.3929,
       "step": 58125
     },
     {
-      "epoch": 2.05,
-      "learning_rate": 3.342334434989337e-05,
-      "loss": 0.2729,
+      "epoch": 2.0950012613976288,
+      "grad_norm": 0.20801417529582977,
+      "learning_rate": 3.2660898308890546e-05,
+      "loss": 0.3763,
       "step": 58130
     },
     {
-      "epoch": 2.05,
-      "learning_rate": 3.3420662143338974e-05,
-      "loss": 0.2923,
+      "epoch": 2.095181461058853,
+      "grad_norm": 0.19347546994686127,
+      "learning_rate": 3.265812051071537e-05,
+      "loss": 0.3982,
       "step": 58135
     },
     {
-      "epoch": 2.05,
-      "learning_rate": 3.341797982744783e-05,
-      "loss": 0.2681,
+      "epoch": 2.095361660720078,
+      "grad_norm": 0.19839556515216827,
+      "learning_rate": 3.265534260819979e-05,
+      "loss": 0.4412,
       "step": 58140
     },
     {
-      "epoch": 2.05,
-      "learning_rate": 3.3415297402254764e-05,
-      "loss": 0.2529,
+      "epoch": 2.0955418603813025,
+      "grad_norm": 0.20302700996398926,
+      "learning_rate": 3.265256460138165e-05,
+      "loss": 0.4398,
       "step": 58145
     },
     {
-      "epoch": 2.05,
-      "learning_rate": 3.341261486779462e-05,
-      "loss": 0.2796,
+      "epoch": 2.0957220600425273,
+      "grad_norm": 0.19693662226200104,
+      "learning_rate": 3.2649786490298796e-05,
+      "loss": 0.4202,
       "step": 58150
     },
     {
-      "epoch": 2.05,
-      "learning_rate": 3.3409932224102205e-05,
-      "loss": 0.2909,
+      "epoch": 2.0959022597037515,
+      "grad_norm": 0.17689436674118042,
+      "learning_rate": 3.2647008274989087e-05,
+      "loss": 0.4202,
       "step": 58155
     },
     {
-      "epoch": 2.05,
-      "learning_rate": 3.340724947121236e-05,
-      "loss": 0.2871,
+      "epoch": 2.0960824593649763,
+      "grad_norm": 0.20521299540996552,
+      "learning_rate": 3.264422995549037e-05,
+      "loss": 0.3863,
       "step": 58160
     },
     {
-      "epoch": 2.05,
-      "learning_rate": 3.340456660915992e-05,
-      "loss": 0.2733,
+      "epoch": 2.096262659026201,
+      "grad_norm": 0.16490758955478668,
+      "learning_rate": 3.26414515318405e-05,
+      "loss": 0.372,
       "step": 58165
     },
     {
-      "epoch": 2.05,
-      "learning_rate": 3.340188363797973e-05,
-      "loss": 0.2731,
+      "epoch": 2.0964428586874257,
+      "grad_norm": 0.19816817343235016,
+      "learning_rate": 3.263867300407732e-05,
+      "loss": 0.4008,
       "step": 58170
     },
     {
-      "epoch": 2.05,
-      "learning_rate": 3.339920055770661e-05,
-      "loss": 0.2587,
+      "epoch": 2.0966230583486505,
+      "grad_norm": 0.19456301629543304,
+      "learning_rate": 3.263589437223871e-05,
+      "loss": 0.3828,
       "step": 58175
     },
     {
-      "epoch": 2.05,
-      "learning_rate": 3.339651736837542e-05,
-      "loss": 0.2766,
+      "epoch": 2.0968032580098748,
+      "grad_norm": 0.20074453949928284,
+      "learning_rate": 3.263311563636252e-05,
+      "loss": 0.3943,
       "step": 58180
     },
     {
-      "epoch": 2.05,
-      "learning_rate": 3.3393834070020966e-05,
-      "loss": 0.2629,
+      "epoch": 2.0969834576710995,
+      "grad_norm": 0.21630464494228363,
+      "learning_rate": 3.26303367964866e-05,
+      "loss": 0.3669,
       "step": 58185
     },
     {
-      "epoch": 2.05,
-      "learning_rate": 3.339115066267812e-05,
-      "loss": 0.2745,
+      "epoch": 2.0971636573323242,
+      "grad_norm": 0.17707021534442902,
+      "learning_rate": 3.262755785264882e-05,
+      "loss": 0.3598,
       "step": 58190
     },
     {
-      "epoch": 2.05,
-      "learning_rate": 3.338846714638171e-05,
-      "loss": 0.2617,
+      "epoch": 2.097343856993549,
+      "grad_norm": 0.1683293730020523,
+      "learning_rate": 3.262477880488705e-05,
+      "loss": 0.3759,
       "step": 58195
     },
     {
-      "epoch": 2.05,
-      "learning_rate": 3.338578352116659e-05,
-      "loss": 0.2845,
+      "epoch": 2.0975240566547733,
+      "grad_norm": 0.20540519058704376,
+      "learning_rate": 3.262199965323913e-05,
+      "loss": 0.4094,
       "step": 58200
     },
     {
-      "epoch": 2.05,
-      "learning_rate": 3.338309978706758e-05,
-      "loss": 0.2641,
+      "epoch": 2.097704256315998,
+      "grad_norm": 0.17812873423099518,
+      "learning_rate": 3.261922039774295e-05,
+      "loss": 0.401,
       "step": 58205
     },
     {
-      "epoch": 2.05,
-      "learning_rate": 3.338041594411956e-05,
-      "loss": 0.3062,
+      "epoch": 2.0978844559772227,
+      "grad_norm": 0.21470516920089722,
+      "learning_rate": 3.261644103843637e-05,
+      "loss": 0.3805,
       "step": 58210
     },
     {
-      "epoch": 2.05,
-      "learning_rate": 3.3377731992357355e-05,
-      "loss": 0.2742,
+      "epoch": 2.0980646556384475,
+      "grad_norm": 0.23254641890525818,
+      "learning_rate": 3.261366157535725e-05,
+      "loss": 0.4054,
       "step": 58215
     },
     {
-      "epoch": 2.05,
-      "learning_rate": 3.3375047931815816e-05,
-      "loss": 0.2501,
+      "epoch": 2.098244855299672,
+      "grad_norm": 0.2531212568283081,
+      "learning_rate": 3.2610882008543466e-05,
+      "loss": 0.3981,
       "step": 58220
     },
     {
-      "epoch": 2.05,
-      "learning_rate": 3.3372363762529815e-05,
-      "loss": 0.2379,
+      "epoch": 2.0984250549608965,
+      "grad_norm": 0.2381640523672104,
+      "learning_rate": 3.260810233803289e-05,
+      "loss": 0.4355,
       "step": 58225
     },
     {
-      "epoch": 2.05,
-      "learning_rate": 3.3369679484534175e-05,
-      "loss": 0.2645,
+      "epoch": 2.098605254622121,
+      "grad_norm": 0.1578020453453064,
+      "learning_rate": 3.260532256386338e-05,
+      "loss": 0.3769,
       "step": 58230
     },
     {
-      "epoch": 2.05,
-      "learning_rate": 3.336699509786377e-05,
-      "loss": 0.2771,
+      "epoch": 2.098785454283346,
+      "grad_norm": 0.23701877892017365,
+      "learning_rate": 3.2602542686072835e-05,
+      "loss": 0.4272,
       "step": 58235
     },
     {
-      "epoch": 2.05,
-      "learning_rate": 3.336431060255344e-05,
-      "loss": 0.2872,
+      "epoch": 2.0989656539445707,
+      "grad_norm": 0.2273239940404892,
+      "learning_rate": 3.259976270469912e-05,
+      "loss": 0.4038,
       "step": 58240
     },
     {
-      "epoch": 2.05,
-      "learning_rate": 3.336162599863805e-05,
-      "loss": 0.2668,
+      "epoch": 2.0991458536057954,
+      "grad_norm": 0.1812596321105957,
+      "learning_rate": 3.25969826197801e-05,
+      "loss": 0.3835,
       "step": 58245
     },
     {
-      "epoch": 2.05,
-      "learning_rate": 3.3358941286152465e-05,
-      "loss": 0.2605,
+      "epoch": 2.0993260532670197,
+      "grad_norm": 0.21768318116664886,
+      "learning_rate": 3.259420243135367e-05,
+      "loss": 0.4035,
       "step": 58250
     },
     {
-      "epoch": 2.05,
-      "learning_rate": 3.335625646513153e-05,
-      "loss": 0.2445,
+      "epoch": 2.0995062529282444,
+      "grad_norm": 0.16613122820854187,
+      "learning_rate": 3.25914221394577e-05,
+      "loss": 0.3626,
       "step": 58255
     },
     {
-      "epoch": 2.05,
-      "learning_rate": 3.335357153561011e-05,
-      "loss": 0.2861,
+      "epoch": 2.099686452589469,
+      "grad_norm": 0.21264012157917023,
+      "learning_rate": 3.258864174413008e-05,
+      "loss": 0.4092,
       "step": 58260
     },
     {
-      "epoch": 2.05,
-      "learning_rate": 3.335088649762308e-05,
-      "loss": 0.2569,
+      "epoch": 2.099866652250694,
+      "grad_norm": 0.22502678632736206,
+      "learning_rate": 3.2585861245408676e-05,
+      "loss": 0.4298,
       "step": 58265
     },
     {
-      "epoch": 2.05,
-      "learning_rate": 3.3348201351205274e-05,
-      "loss": 0.2772,
+      "epoch": 2.100046851911918,
+      "grad_norm": 0.17495203018188477,
+      "learning_rate": 3.2583080643331385e-05,
+      "loss": 0.3757,
       "step": 58270
     },
     {
-      "epoch": 2.05,
-      "learning_rate": 3.334551609639159e-05,
-      "loss": 0.2614,
+      "epoch": 2.100227051573143,
+      "grad_norm": 0.16976486146450043,
+      "learning_rate": 3.2580299937936084e-05,
+      "loss": 0.4213,
       "step": 58275
     },
     {
-      "epoch": 2.05,
-      "learning_rate": 3.334283073321687e-05,
-      "loss": 0.2743,
+      "epoch": 2.1004072512343677,
+      "grad_norm": 0.25477150082588196,
+      "learning_rate": 3.2577519129260666e-05,
+      "loss": 0.4203,
       "step": 58280
     },
     {
-      "epoch": 2.05,
-      "learning_rate": 3.3340145261716e-05,
-      "loss": 0.3164,
+      "epoch": 2.1005874508955924,
+      "grad_norm": 0.27590665221214294,
+      "learning_rate": 3.257473821734302e-05,
+      "loss": 0.4257,
       "step": 58285
     },
     {
-      "epoch": 2.05,
-      "learning_rate": 3.3337459681923835e-05,
-      "loss": 0.3013,
+      "epoch": 2.100767650556817,
+      "grad_norm": 0.18008361756801605,
+      "learning_rate": 3.257195720222103e-05,
+      "loss": 0.3971,
       "step": 58290
     },
     {
-      "epoch": 2.05,
-      "learning_rate": 3.3334773993875254e-05,
-      "loss": 0.2843,
+      "epoch": 2.1009478502180414,
+      "grad_norm": 0.22707924246788025,
+      "learning_rate": 3.256917608393259e-05,
+      "loss": 0.3946,
       "step": 58295
     },
     {
-      "epoch": 2.05,
-      "learning_rate": 3.3332088197605115e-05,
-      "loss": 0.2746,
+      "epoch": 2.101128049879266,
+      "grad_norm": 0.18620999157428741,
+      "learning_rate": 3.2566394862515596e-05,
+      "loss": 0.3527,
       "step": 58300
     },
     {
-      "epoch": 2.05,
-      "learning_rate": 3.332940229314832e-05,
-      "loss": 0.2719,
+      "epoch": 2.101308249540491,
+      "grad_norm": 0.20725572109222412,
+      "learning_rate": 3.256361353800793e-05,
+      "loss": 0.4064,
       "step": 58305
     },
     {
-      "epoch": 2.05,
-      "learning_rate": 3.33267162805397e-05,
-      "loss": 0.2672,
+      "epoch": 2.1014884492017156,
+      "grad_norm": 0.2142266035079956,
+      "learning_rate": 3.2560832110447495e-05,
+      "loss": 0.4219,
       "step": 58310
     },
     {
-      "epoch": 2.05,
-      "learning_rate": 3.332403015981418e-05,
-      "loss": 0.2518,
+      "epoch": 2.1016686488629404,
+      "grad_norm": 0.2480008453130722,
+      "learning_rate": 3.255805057987218e-05,
+      "loss": 0.4156,
       "step": 58315
     },
     {
-      "epoch": 2.05,
-      "learning_rate": 3.332134393100661e-05,
-      "loss": 0.2824,
+      "epoch": 2.1018488485241646,
+      "grad_norm": 0.22101950645446777,
+      "learning_rate": 3.255526894631991e-05,
+      "loss": 0.3769,
       "step": 58320
     },
     {
-      "epoch": 2.05,
-      "learning_rate": 3.331865759415186e-05,
-      "loss": 0.2766,
+      "epoch": 2.1020290481853894,
+      "grad_norm": 0.1711684614419937,
+      "learning_rate": 3.255248720982854e-05,
+      "loss": 0.4069,
       "step": 58325
     },
     {
-      "epoch": 2.05,
-      "learning_rate": 3.331597114928484e-05,
-      "loss": 0.263,
+      "epoch": 2.102209247846614,
+      "grad_norm": 0.2133340984582901,
+      "learning_rate": 3.254970537043601e-05,
+      "loss": 0.4075,
       "step": 58330
     },
     {
-      "epoch": 2.05,
-      "learning_rate": 3.33132845964404e-05,
-      "loss": 0.2864,
+      "epoch": 2.102389447507839,
+      "grad_norm": 0.27530649304389954,
+      "learning_rate": 3.2546923428180184e-05,
+      "loss": 0.4109,
       "step": 58335
     },
     {
-      "epoch": 2.05,
-      "learning_rate": 3.331059793565345e-05,
-      "loss": 0.3015,
+      "epoch": 2.102569647169063,
+      "grad_norm": 0.19913306832313538,
+      "learning_rate": 3.2544141383099014e-05,
+      "loss": 0.4136,
       "step": 58340
     },
     {
-      "epoch": 2.05,
-      "learning_rate": 3.330791116695886e-05,
-      "loss": 0.2648,
+      "epoch": 2.102749846830288,
+      "grad_norm": 0.2158804088830948,
+      "learning_rate": 3.254135923523037e-05,
+      "loss": 0.3929,
       "step": 58345
     },
     {
-      "epoch": 2.05,
-      "learning_rate": 3.330522429039152e-05,
-      "loss": 0.2498,
+      "epoch": 2.1029300464915126,
+      "grad_norm": 0.23708494007587433,
+      "learning_rate": 3.253857698461216e-05,
+      "loss": 0.4019,
       "step": 58350
     },
     {
-      "epoch": 2.05,
-      "learning_rate": 3.330253730598631e-05,
-      "loss": 0.2872,
+      "epoch": 2.1031102461527373,
+      "grad_norm": 0.20912696421146393,
+      "learning_rate": 3.25357946312823e-05,
+      "loss": 0.4115,
       "step": 58355
     },
     {
-      "epoch": 2.05,
-      "learning_rate": 3.3299850213778125e-05,
-      "loss": 0.281,
+      "epoch": 2.103290445813962,
+      "grad_norm": 0.1815493106842041,
+      "learning_rate": 3.25330121752787e-05,
+      "loss": 0.4154,
       "step": 58360
     },
     {
-      "epoch": 2.05,
-      "learning_rate": 3.329716301380185e-05,
-      "loss": 0.2943,
+      "epoch": 2.1034706454751864,
+      "grad_norm": 0.2595524489879608,
+      "learning_rate": 3.253022961663927e-05,
+      "loss": 0.4105,
       "step": 58365
     },
     {
-      "epoch": 2.05,
-      "learning_rate": 3.32944757060924e-05,
-      "loss": 0.2771,
+      "epoch": 2.103650845136411,
+      "grad_norm": 0.1517713963985443,
+      "learning_rate": 3.252744695540191e-05,
+      "loss": 0.4013,
       "step": 58370
     },
     {
-      "epoch": 2.05,
-      "learning_rate": 3.329178829068463e-05,
-      "loss": 0.277,
+      "epoch": 2.103831044797636,
+      "grad_norm": 0.1964835673570633,
+      "learning_rate": 3.252466419160455e-05,
+      "loss": 0.4256,
       "step": 58375
     },
     {
-      "epoch": 2.05,
-      "learning_rate": 3.3289100767613464e-05,
-      "loss": 0.2565,
+      "epoch": 2.1040112444588606,
+      "grad_norm": 0.18157154321670532,
+      "learning_rate": 3.25218813252851e-05,
+      "loss": 0.3714,
       "step": 58380
     },
     {
-      "epoch": 2.05,
-      "learning_rate": 3.328641313691378e-05,
-      "loss": 0.2783,
+      "epoch": 2.104191444120085,
+      "grad_norm": 0.18907707929611206,
+      "learning_rate": 3.251909835648147e-05,
+      "loss": 0.4031,
       "step": 58385
     },
     {
-      "epoch": 2.05,
-      "learning_rate": 3.328372539862049e-05,
-      "loss": 0.2651,
+      "epoch": 2.1043716437813096,
+      "grad_norm": 0.17940135300159454,
+      "learning_rate": 3.2516315285231576e-05,
+      "loss": 0.414,
       "step": 58390
     },
     {
-      "epoch": 2.05,
-      "learning_rate": 3.328103755276848e-05,
-      "loss": 0.2727,
+      "epoch": 2.1045518434425343,
+      "grad_norm": 0.18015572428703308,
+      "learning_rate": 3.251353211157334e-05,
+      "loss": 0.3899,
       "step": 58395
     },
     {
-      "epoch": 2.05,
-      "learning_rate": 3.3278349599392664e-05,
-      "loss": 0.2875,
+      "epoch": 2.104732043103759,
+      "grad_norm": 0.20908917486667633,
+      "learning_rate": 3.25107488355447e-05,
+      "loss": 0.4163,
       "step": 58400
     },
     {
-      "epoch": 2.05,
-      "learning_rate": 3.3275661538527926e-05,
-      "loss": 0.2729,
+      "epoch": 2.104912242764984,
+      "grad_norm": 0.1619907170534134,
+      "learning_rate": 3.2507965457183545e-05,
+      "loss": 0.4148,
       "step": 58405
     },
     {
-      "epoch": 2.06,
-      "learning_rate": 3.327297337020917e-05,
-      "loss": 0.2701,
+      "epoch": 2.105092442426208,
+      "grad_norm": 0.1768513172864914,
+      "learning_rate": 3.2505181976527815e-05,
+      "loss": 0.4,
       "step": 58410
     },
     {
-      "epoch": 2.06,
-      "learning_rate": 3.327028509447132e-05,
-      "loss": 0.2621,
+      "epoch": 2.105272642087433,
+      "grad_norm": 0.17166374623775482,
+      "learning_rate": 3.2502398393615435e-05,
+      "loss": 0.394,
       "step": 58415
     },
     {
-      "epoch": 2.06,
-      "learning_rate": 3.3267596711349256e-05,
-      "loss": 0.2715,
+      "epoch": 2.1054528417486575,
+      "grad_norm": 0.21615466475486755,
+      "learning_rate": 3.249961470848433e-05,
+      "loss": 0.3977,
       "step": 58420
     },
     {
-      "epoch": 2.06,
-      "learning_rate": 3.3264908220877904e-05,
-      "loss": 0.2651,
+      "epoch": 2.1056330414098823,
+      "grad_norm": 0.1769254207611084,
+      "learning_rate": 3.249683092117243e-05,
+      "loss": 0.3984,
       "step": 58425
     },
     {
-      "epoch": 2.06,
-      "learning_rate": 3.3262219623092174e-05,
-      "loss": 0.2773,
+      "epoch": 2.1058132410711066,
+      "grad_norm": 0.19884678721427917,
+      "learning_rate": 3.249404703171766e-05,
+      "loss": 0.3915,
       "step": 58430
     },
     {
-      "epoch": 2.06,
-      "learning_rate": 3.325953091802696e-05,
-      "loss": 0.2623,
+      "epoch": 2.1059934407323313,
+      "grad_norm": 0.186051145195961,
+      "learning_rate": 3.249126304015795e-05,
+      "loss": 0.388,
       "step": 58435
     },
     {
-      "epoch": 2.06,
-      "learning_rate": 3.325684210571717e-05,
-      "loss": 0.267,
+      "epoch": 2.106173640393556,
+      "grad_norm": 0.16118311882019043,
+      "learning_rate": 3.248847894653122e-05,
+      "loss": 0.3806,
       "step": 58440
     },
     {
-      "epoch": 2.06,
-      "learning_rate": 3.325415318619773e-05,
-      "loss": 0.2918,
+      "epoch": 2.1063538400547808,
+      "grad_norm": 0.24056154489517212,
+      "learning_rate": 3.248569475087544e-05,
+      "loss": 0.4179,
       "step": 58445
     },
     {
-      "epoch": 2.06,
-      "learning_rate": 3.325146415950356e-05,
-      "loss": 0.2827,
+      "epoch": 2.1065340397160055,
+      "grad_norm": 0.204376682639122,
+      "learning_rate": 3.24829104532285e-05,
+      "loss": 0.3641,
       "step": 58450
     },
     {
-      "epoch": 2.06,
-      "learning_rate": 3.324877502566957e-05,
-      "loss": 0.2751,
+      "epoch": 2.10671423937723,
+      "grad_norm": 0.2513475716114044,
+      "learning_rate": 3.248012605362836e-05,
+      "loss": 0.3944,
       "step": 58455
     },
     {
-      "epoch": 2.06,
-      "learning_rate": 3.324608578473066e-05,
-      "loss": 0.251,
+      "epoch": 2.1068944390384545,
+      "grad_norm": 0.24770568311214447,
+      "learning_rate": 3.247734155211294e-05,
+      "loss": 0.3903,
       "step": 58460
     },
     {
-      "epoch": 2.06,
-      "learning_rate": 3.324339643672176e-05,
-      "loss": 0.3062,
+      "epoch": 2.1070746386996793,
+      "grad_norm": 0.28164035081863403,
+      "learning_rate": 3.2474556948720197e-05,
+      "loss": 0.3925,
       "step": 58465
     },
     {
-      "epoch": 2.06,
-      "learning_rate": 3.324070698167779e-05,
-      "loss": 0.259,
+      "epoch": 2.107254838360904,
+      "grad_norm": 0.2520608603954315,
+      "learning_rate": 3.2471772243488064e-05,
+      "loss": 0.3713,
       "step": 58470
     },
     {
-      "epoch": 2.06,
-      "learning_rate": 3.3238017419633684e-05,
-      "loss": 0.2959,
+      "epoch": 2.1074350380221287,
+      "grad_norm": 0.18132475018501282,
+      "learning_rate": 3.2468987436454476e-05,
+      "loss": 0.4178,
       "step": 58475
     },
     {
-      "epoch": 2.06,
-      "learning_rate": 3.323532775062434e-05,
-      "loss": 0.2834,
+      "epoch": 2.107615237683353,
+      "grad_norm": 0.20063307881355286,
+      "learning_rate": 3.246620252765739e-05,
+      "loss": 0.417,
       "step": 58480
     },
     {
-      "epoch": 2.06,
-      "learning_rate": 3.32326379746847e-05,
-      "loss": 0.2654,
+      "epoch": 2.1077954373445777,
+      "grad_norm": 0.227056086063385,
+      "learning_rate": 3.2463417517134734e-05,
+      "loss": 0.4046,
       "step": 58485
     },
     {
-      "epoch": 2.06,
-      "learning_rate": 3.3229948091849674e-05,
-      "loss": 0.2705,
+      "epoch": 2.1079756370058025,
+      "grad_norm": 0.16620944440364838,
+      "learning_rate": 3.246063240492445e-05,
+      "loss": 0.3735,
       "step": 58490
     },
     {
-      "epoch": 2.06,
-      "learning_rate": 3.32272581021542e-05,
-      "loss": 0.2969,
+      "epoch": 2.108155836667027,
+      "grad_norm": 0.18251557648181915,
+      "learning_rate": 3.2457847191064505e-05,
+      "loss": 0.3888,
       "step": 58495
     },
     {
-      "epoch": 2.06,
-      "learning_rate": 3.3224568005633195e-05,
-      "loss": 0.2682,
+      "epoch": 2.1083360363282515,
+      "grad_norm": 0.20727889239788055,
+      "learning_rate": 3.245506187559283e-05,
+      "loss": 0.4018,
       "step": 58500
     },
     {
-      "epoch": 2.06,
-      "eval_loss": 0.2701320946216583,
-      "eval_runtime": 10.5303,
-      "eval_samples_per_second": 9.496,
-      "eval_steps_per_second": 9.496,
+      "epoch": 2.1083360363282515,
+      "eval_loss": 0.4373342990875244,
+      "eval_runtime": 3.5409,
+      "eval_samples_per_second": 28.242,
+      "eval_steps_per_second": 7.06,
       "step": 58500
     },
     {
-      "epoch": 2.06,
-      "learning_rate": 3.32218778023216e-05,
-      "loss": 0.2425,
+      "epoch": 2.1085162359894762,
+      "grad_norm": 0.2316305935382843,
+      "learning_rate": 3.245227645854739e-05,
+      "loss": 0.3647,
       "step": 58505
     },
     {
-      "epoch": 2.06,
-      "learning_rate": 3.321918749225434e-05,
-      "loss": 0.2702,
+      "epoch": 2.108696435650701,
+      "grad_norm": 0.22615645825862885,
+      "learning_rate": 3.244949093996612e-05,
+      "loss": 0.3822,
       "step": 58510
     },
     {
-      "epoch": 2.06,
-      "learning_rate": 3.321649707546635e-05,
-      "loss": 0.2658,
+      "epoch": 2.1088766353119257,
+      "grad_norm": 0.2182559221982956,
+      "learning_rate": 3.244670531988697e-05,
+      "loss": 0.3833,
       "step": 58515
     },
     {
-      "epoch": 2.06,
-      "learning_rate": 3.321380655199255e-05,
-      "loss": 0.2684,
+      "epoch": 2.1090568349731504,
+      "grad_norm": 0.19747740030288696,
+      "learning_rate": 3.244391959834791e-05,
+      "loss": 0.3731,
       "step": 58520
     },
     {
-      "epoch": 2.06,
-      "learning_rate": 3.3211115921867894e-05,
-      "loss": 0.2719,
+      "epoch": 2.1092370346343747,
+      "grad_norm": 0.165822371840477,
+      "learning_rate": 3.244113377538689e-05,
+      "loss": 0.3825,
       "step": 58525
     },
     {
-      "epoch": 2.06,
-      "learning_rate": 3.3208425185127305e-05,
-      "loss": 0.2789,
+      "epoch": 2.1094172342955995,
+      "grad_norm": 0.19013039767742157,
+      "learning_rate": 3.243834785104186e-05,
+      "loss": 0.3978,
       "step": 58530
     },
     {
-      "epoch": 2.06,
-      "learning_rate": 3.3205734341805725e-05,
-      "loss": 0.2787,
+      "epoch": 2.109597433956824,
+      "grad_norm": 0.16244149208068848,
+      "learning_rate": 3.243556182535077e-05,
+      "loss": 0.3966,
       "step": 58535
     },
     {
-      "epoch": 2.06,
-      "learning_rate": 3.3203043391938094e-05,
-      "loss": 0.2594,
+      "epoch": 2.109777633618049,
+      "grad_norm": 0.19332407414913177,
+      "learning_rate": 3.2432775698351605e-05,
+      "loss": 0.3793,
       "step": 58540
     },
     {
-      "epoch": 2.06,
-      "learning_rate": 3.320035233555935e-05,
-      "loss": 0.2918,
+      "epoch": 2.109957833279273,
+      "grad_norm": 0.18343211710453033,
+      "learning_rate": 3.242998947008231e-05,
+      "loss": 0.4015,
       "step": 58545
     },
     {
-      "epoch": 2.06,
-      "learning_rate": 3.319766117270444e-05,
-      "loss": 0.2422,
+      "epoch": 2.110138032940498,
+      "grad_norm": 0.17457517981529236,
+      "learning_rate": 3.242720314058084e-05,
+      "loss": 0.3965,
       "step": 58550
     },
     {
-      "epoch": 2.06,
-      "learning_rate": 3.3194969903408294e-05,
-      "loss": 0.281,
+      "epoch": 2.1103182326017227,
+      "grad_norm": 0.1578681319952011,
+      "learning_rate": 3.2424416709885165e-05,
+      "loss": 0.4025,
       "step": 58555
     },
     {
-      "epoch": 2.06,
-      "learning_rate": 3.319227852770587e-05,
-      "loss": 0.2878,
+      "epoch": 2.1104984322629474,
+      "grad_norm": 0.20855949819087982,
+      "learning_rate": 3.242163017803325e-05,
+      "loss": 0.4154,
       "step": 58560
     },
     {
-      "epoch": 2.06,
-      "learning_rate": 3.31895870456321e-05,
-      "loss": 0.2678,
+      "epoch": 2.110678631924172,
+      "grad_norm": 0.15797370672225952,
+      "learning_rate": 3.2418843545063065e-05,
+      "loss": 0.3684,
       "step": 58565
     },
     {
-      "epoch": 2.06,
-      "learning_rate": 3.318689545722195e-05,
-      "loss": 0.2968,
+      "epoch": 2.1108588315853964,
+      "grad_norm": 0.22277599573135376,
+      "learning_rate": 3.241605681101256e-05,
+      "loss": 0.401,
       "step": 58570
     },
     {
-      "epoch": 2.06,
-      "learning_rate": 3.318420376251035e-05,
-      "loss": 0.2631,
+      "epoch": 2.111039031246621,
+      "grad_norm": 0.19849325716495514,
+      "learning_rate": 3.2413269975919736e-05,
+      "loss": 0.4138,
       "step": 58575
     },
     {
-      "epoch": 2.06,
-      "learning_rate": 3.318151196153226e-05,
-      "loss": 0.2765,
+      "epoch": 2.111219230907846,
+      "grad_norm": 0.22037936747074127,
+      "learning_rate": 3.2410483039822527e-05,
+      "loss": 0.3959,
       "step": 58580
     },
     {
-      "epoch": 2.06,
-      "learning_rate": 3.317882005432263e-05,
-      "loss": 0.2809,
+      "epoch": 2.1113994305690706,
+      "grad_norm": 0.1882595717906952,
+      "learning_rate": 3.2407696002758936e-05,
+      "loss": 0.3915,
       "step": 58585
     },
     {
-      "epoch": 2.06,
-      "learning_rate": 3.317612804091641e-05,
-      "loss": 0.2923,
+      "epoch": 2.1115796302302954,
+      "grad_norm": 0.20610591769218445,
+      "learning_rate": 3.240490886476691e-05,
+      "loss": 0.4286,
       "step": 58590
     },
     {
-      "epoch": 2.06,
-      "learning_rate": 3.317343592134856e-05,
-      "loss": 0.2739,
+      "epoch": 2.1117598298915197,
+      "grad_norm": 0.22938185930252075,
+      "learning_rate": 3.240212162588444e-05,
+      "loss": 0.3866,
       "step": 58595
     },
     {
-      "epoch": 2.06,
-      "learning_rate": 3.3170743695654026e-05,
-      "loss": 0.2769,
+      "epoch": 2.1119400295527444,
+      "grad_norm": 0.20700225234031677,
+      "learning_rate": 3.2399334286149495e-05,
+      "loss": 0.3753,
       "step": 58600
     },
     {
-      "epoch": 2.06,
-      "learning_rate": 3.3168051363867774e-05,
-      "loss": 0.2654,
+      "epoch": 2.112120229213969,
+      "grad_norm": 0.19341452419757843,
+      "learning_rate": 3.239654684560005e-05,
+      "loss": 0.4153,
       "step": 58605
     },
     {
-      "epoch": 2.06,
-      "learning_rate": 3.316535892602476e-05,
-      "loss": 0.2636,
+      "epoch": 2.112300428875194,
+      "grad_norm": 0.20919573307037354,
+      "learning_rate": 3.23937593042741e-05,
+      "loss": 0.4088,
       "step": 58610
     },
     {
-      "epoch": 2.06,
-      "learning_rate": 3.3162666382159935e-05,
-      "loss": 0.264,
+      "epoch": 2.112480628536418,
+      "grad_norm": 0.20595782995224,
+      "learning_rate": 3.239097166220959e-05,
+      "loss": 0.3805,
       "step": 58615
     },
     {
-      "epoch": 2.06,
-      "learning_rate": 3.315997373230828e-05,
-      "loss": 0.3044,
+      "epoch": 2.112660828197643,
+      "grad_norm": 0.18179921805858612,
+      "learning_rate": 3.238818391944453e-05,
+      "loss": 0.4064,
       "step": 58620
     },
     {
-      "epoch": 2.06,
-      "learning_rate": 3.3157280976504726e-05,
-      "loss": 0.2707,
+      "epoch": 2.1128410278588676,
+      "grad_norm": 0.1617589145898819,
+      "learning_rate": 3.2385396076016896e-05,
+      "loss": 0.395,
       "step": 58625
     },
     {
-      "epoch": 2.06,
-      "learning_rate": 3.315458811478426e-05,
-      "loss": 0.315,
+      "epoch": 2.1130212275200924,
+      "grad_norm": 0.195877343416214,
+      "learning_rate": 3.2382608131964676e-05,
+      "loss": 0.4229,
       "step": 58630
     },
     {
-      "epoch": 2.06,
-      "learning_rate": 3.315189514718184e-05,
-      "loss": 0.2899,
+      "epoch": 2.113201427181317,
+      "grad_norm": 0.20379838347434998,
+      "learning_rate": 3.2379820087325844e-05,
+      "loss": 0.4083,
       "step": 58635
     },
     {
-      "epoch": 2.06,
-      "learning_rate": 3.3149202073732446e-05,
-      "loss": 0.2767,
+      "epoch": 2.1133816268425414,
+      "grad_norm": 0.2131894826889038,
+      "learning_rate": 3.237703194213839e-05,
+      "loss": 0.3798,
       "step": 58640
     },
     {
-      "epoch": 2.06,
-      "learning_rate": 3.314650889447102e-05,
-      "loss": 0.2539,
+      "epoch": 2.113561826503766,
+      "grad_norm": 0.23146046698093414,
+      "learning_rate": 3.2374243696440305e-05,
+      "loss": 0.4218,
       "step": 58645
     },
     {
-      "epoch": 2.06,
-      "learning_rate": 3.314381560943255e-05,
-      "loss": 0.2922,
+      "epoch": 2.113742026164991,
+      "grad_norm": 0.1999361366033554,
+      "learning_rate": 3.2371455350269574e-05,
+      "loss": 0.4121,
       "step": 58650
     },
     {
-      "epoch": 2.06,
-      "learning_rate": 3.3141122218652e-05,
-      "loss": 0.2856,
+      "epoch": 2.1139222258262156,
+      "grad_norm": 0.16494305431842804,
+      "learning_rate": 3.23686669036642e-05,
+      "loss": 0.392,
       "step": 58655
     },
     {
-      "epoch": 2.06,
-      "learning_rate": 3.313842872216434e-05,
-      "loss": 0.2456,
+      "epoch": 2.11410242548744,
+      "grad_norm": 0.24381648004055023,
+      "learning_rate": 3.236587835666216e-05,
+      "loss": 0.3991,
       "step": 58660
     },
     {
-      "epoch": 2.06,
-      "learning_rate": 3.313573512000455e-05,
-      "loss": 0.2928,
+      "epoch": 2.1142826251486646,
+      "grad_norm": 0.15115749835968018,
+      "learning_rate": 3.236308970930145e-05,
+      "loss": 0.3447,
       "step": 58665
     },
     {
-      "epoch": 2.06,
-      "learning_rate": 3.3133041412207596e-05,
-      "loss": 0.2632,
+      "epoch": 2.1144628248098893,
+      "grad_norm": 0.19442135095596313,
+      "learning_rate": 3.236030096162008e-05,
+      "loss": 0.3937,
       "step": 58670
     },
     {
-      "epoch": 2.06,
-      "learning_rate": 3.313034759880847e-05,
-      "loss": 0.2838,
+      "epoch": 2.114643024471114,
+      "grad_norm": 0.23755992949008942,
+      "learning_rate": 3.235751211365602e-05,
+      "loss": 0.381,
       "step": 58675
     },
     {
-      "epoch": 2.06,
-      "learning_rate": 3.312765367984213e-05,
-      "loss": 0.2953,
+      "epoch": 2.114823224132339,
+      "grad_norm": 0.19326242804527283,
+      "learning_rate": 3.23547231654473e-05,
+      "loss": 0.3761,
       "step": 58680
     },
     {
-      "epoch": 2.06,
-      "learning_rate": 3.3124959655343564e-05,
-      "loss": 0.264,
+      "epoch": 2.115003423793563,
+      "grad_norm": 0.2071000635623932,
+      "learning_rate": 3.2351934117031877e-05,
+      "loss": 0.4316,
       "step": 58685
     },
     {
-      "epoch": 2.06,
-      "learning_rate": 3.3122265525347754e-05,
-      "loss": 0.2863,
+      "epoch": 2.115183623454788,
+      "grad_norm": 0.17056724429130554,
+      "learning_rate": 3.23491449684478e-05,
+      "loss": 0.3951,
       "step": 58690
     },
     {
-      "epoch": 2.07,
-      "learning_rate": 3.3119571289889675e-05,
-      "loss": 0.2702,
+      "epoch": 2.1153638231160126,
+      "grad_norm": 0.172649085521698,
+      "learning_rate": 3.234635571973303e-05,
+      "loss": 0.3895,
       "step": 58695
     },
     {
-      "epoch": 2.07,
-      "learning_rate": 3.311687694900432e-05,
-      "loss": 0.303,
+      "epoch": 2.1155440227772373,
+      "grad_norm": 0.21372824907302856,
+      "learning_rate": 3.2343566370925594e-05,
+      "loss": 0.398,
       "step": 58700
     },
     {
-      "epoch": 2.07,
-      "learning_rate": 3.3114182502726664e-05,
-      "loss": 0.2653,
+      "epoch": 2.1157242224384616,
+      "grad_norm": 0.23039843142032623,
+      "learning_rate": 3.234077692206347e-05,
+      "loss": 0.4055,
       "step": 58705
     },
     {
-      "epoch": 2.07,
-      "learning_rate": 3.31114879510917e-05,
-      "loss": 0.2366,
+      "epoch": 2.1159044220996863,
+      "grad_norm": 0.22688081860542297,
+      "learning_rate": 3.2337987373184704e-05,
+      "loss": 0.3998,
       "step": 58710
     },
     {
-      "epoch": 2.07,
-      "learning_rate": 3.3108793294134405e-05,
-      "loss": 0.2524,
+      "epoch": 2.116084621760911,
+      "grad_norm": 0.15555809438228607,
+      "learning_rate": 3.233519772432727e-05,
+      "loss": 0.3709,
       "step": 58715
     },
     {
-      "epoch": 2.07,
-      "learning_rate": 3.310609853188977e-05,
-      "loss": 0.2888,
+      "epoch": 2.116264821422136,
+      "grad_norm": 0.22005966305732727,
+      "learning_rate": 3.233240797552919e-05,
+      "loss": 0.3701,
       "step": 58720
     },
     {
-      "epoch": 2.07,
-      "learning_rate": 3.3103403664392797e-05,
-      "loss": 0.2645,
+      "epoch": 2.1164450210833605,
+      "grad_norm": 0.21118348836898804,
+      "learning_rate": 3.232961812682847e-05,
+      "loss": 0.3888,
       "step": 58725
     },
     {
-      "epoch": 2.07,
-      "learning_rate": 3.310070869167846e-05,
-      "loss": 0.2613,
+      "epoch": 2.116625220744585,
+      "grad_norm": 0.18728987872600555,
+      "learning_rate": 3.2326828178263125e-05,
+      "loss": 0.4143,
       "step": 58730
     },
     {
-      "epoch": 2.07,
-      "learning_rate": 3.309801361378177e-05,
-      "loss": 0.2458,
+      "epoch": 2.1168054204058095,
+      "grad_norm": 0.19817517697811127,
+      "learning_rate": 3.2324038129871166e-05,
+      "loss": 0.4269,
       "step": 58735
     },
     {
-      "epoch": 2.07,
-      "learning_rate": 3.309531843073769e-05,
-      "loss": 0.2817,
+      "epoch": 2.1169856200670343,
+      "grad_norm": 0.19187265634536743,
+      "learning_rate": 3.232124798169059e-05,
+      "loss": 0.4259,
       "step": 58740
     },
     {
-      "epoch": 2.07,
-      "learning_rate": 3.309262314258125e-05,
-      "loss": 0.269,
+      "epoch": 2.117165819728259,
+      "grad_norm": 0.2186998724937439,
+      "learning_rate": 3.231845773375944e-05,
+      "loss": 0.4291,
       "step": 58745
     },
     {
-      "epoch": 2.07,
-      "learning_rate": 3.3089927749347426e-05,
-      "loss": 0.2837,
+      "epoch": 2.1173460193894837,
+      "grad_norm": 0.22712989151477814,
+      "learning_rate": 3.231566738611572e-05,
+      "loss": 0.3747,
       "step": 58750
     },
     {
-      "epoch": 2.07,
-      "learning_rate": 3.308723225107123e-05,
-      "loss": 0.2785,
+      "epoch": 2.117526219050708,
+      "grad_norm": 0.15697424113750458,
+      "learning_rate": 3.231287693879745e-05,
+      "loss": 0.3601,
       "step": 58755
     },
     {
-      "epoch": 2.07,
-      "learning_rate": 3.308453664778765e-05,
-      "loss": 0.2643,
+      "epoch": 2.1177064187119328,
+      "grad_norm": 0.23286950588226318,
+      "learning_rate": 3.231008639184265e-05,
+      "loss": 0.388,
       "step": 58760
     },
     {
-      "epoch": 2.07,
-      "learning_rate": 3.3081840939531684e-05,
-      "loss": 0.2622,
+      "epoch": 2.1178866183731575,
+      "grad_norm": 0.19460979104042053,
+      "learning_rate": 3.230729574528932e-05,
+      "loss": 0.4061,
       "step": 58765
     },
     {
-      "epoch": 2.07,
-      "learning_rate": 3.307914512633834e-05,
-      "loss": 0.2705,
+      "epoch": 2.1180668180343822,
+      "grad_norm": 0.1705007404088974,
+      "learning_rate": 3.230450499917552e-05,
+      "loss": 0.4024,
       "step": 58770
     },
     {
-      "epoch": 2.07,
-      "learning_rate": 3.3076449208242625e-05,
-      "loss": 0.2742,
+      "epoch": 2.1182470176956065,
+      "grad_norm": 0.20845447480678558,
+      "learning_rate": 3.2301714153539244e-05,
+      "loss": 0.4182,
       "step": 58775
     },
     {
-      "epoch": 2.07,
-      "learning_rate": 3.307375318527954e-05,
-      "loss": 0.2669,
+      "epoch": 2.1184272173568313,
+      "grad_norm": 0.18309171497821808,
+      "learning_rate": 3.2298923208418535e-05,
+      "loss": 0.3745,
       "step": 58780
     },
     {
-      "epoch": 2.07,
-      "learning_rate": 3.307105705748409e-05,
-      "loss": 0.2481,
+      "epoch": 2.118607417018056,
+      "grad_norm": 0.2173132747411728,
+      "learning_rate": 3.22961321638514e-05,
+      "loss": 0.397,
       "step": 58785
     },
     {
-      "epoch": 2.07,
-      "learning_rate": 3.306836082489128e-05,
-      "loss": 0.2822,
+      "epoch": 2.1187876166792807,
+      "grad_norm": 0.16794531047344208,
+      "learning_rate": 3.229334101987588e-05,
+      "loss": 0.3844,
       "step": 58790
     },
     {
-      "epoch": 2.07,
-      "learning_rate": 3.3065664487536124e-05,
-      "loss": 0.2952,
+      "epoch": 2.1189678163405055,
+      "grad_norm": 0.21813854575157166,
+      "learning_rate": 3.229054977653001e-05,
+      "loss": 0.42,
       "step": 58795
     },
     {
-      "epoch": 2.07,
-      "learning_rate": 3.306296804545362e-05,
-      "loss": 0.248,
+      "epoch": 2.1191480160017298,
+      "grad_norm": 0.21310067176818848,
+      "learning_rate": 3.22877584338518e-05,
+      "loss": 0.413,
       "step": 58800
     },
     {
-      "epoch": 2.07,
-      "learning_rate": 3.306027149867881e-05,
-      "loss": 0.2751,
+      "epoch": 2.1193282156629545,
+      "grad_norm": 0.20644977688789368,
+      "learning_rate": 3.2284966991879295e-05,
+      "loss": 0.3717,
       "step": 58805
     },
     {
-      "epoch": 2.07,
-      "learning_rate": 3.305757484724667e-05,
-      "loss": 0.2852,
+      "epoch": 2.119508415324179,
+      "grad_norm": 0.2088310867547989,
+      "learning_rate": 3.228217545065052e-05,
+      "loss": 0.4002,
       "step": 58810
     },
     {
-      "epoch": 2.07,
-      "learning_rate": 3.3054878091192245e-05,
-      "loss": 0.2712,
+      "epoch": 2.119688614985404,
+      "grad_norm": 0.23381415009498596,
+      "learning_rate": 3.227938381020353e-05,
+      "loss": 0.392,
       "step": 58815
     },
     {
-      "epoch": 2.07,
-      "learning_rate": 3.305218123055052e-05,
-      "loss": 0.2791,
+      "epoch": 2.1198688146466287,
+      "grad_norm": 0.19192180037498474,
+      "learning_rate": 3.227659207057633e-05,
+      "loss": 0.3862,
       "step": 58820
     },
     {
-      "epoch": 2.07,
-      "learning_rate": 3.3049484265356536e-05,
-      "loss": 0.2696,
+      "epoch": 2.120049014307853,
+      "grad_norm": 0.17391976714134216,
+      "learning_rate": 3.2273800231806974e-05,
+      "loss": 0.3916,
       "step": 58825
     },
     {
-      "epoch": 2.07,
-      "learning_rate": 3.3046787195645306e-05,
-      "loss": 0.2566,
+      "epoch": 2.1202292139690777,
+      "grad_norm": 0.236240416765213,
+      "learning_rate": 3.2271008293933496e-05,
+      "loss": 0.3979,
       "step": 58830
     },
     {
-      "epoch": 2.07,
-      "learning_rate": 3.3044090021451845e-05,
-      "loss": 0.2605,
+      "epoch": 2.1204094136303024,
+      "grad_norm": 0.15316404402256012,
+      "learning_rate": 3.2268216256993944e-05,
+      "loss": 0.3986,
       "step": 58835
     },
     {
-      "epoch": 2.07,
-      "learning_rate": 3.3041392742811184e-05,
-      "loss": 0.2699,
+      "epoch": 2.120589613291527,
+      "grad_norm": 0.20581558346748352,
+      "learning_rate": 3.2265424121026355e-05,
+      "loss": 0.3874,
       "step": 58840
     },
     {
-      "epoch": 2.07,
-      "learning_rate": 3.3038695359758324e-05,
-      "loss": 0.2714,
+      "epoch": 2.1207698129527515,
+      "grad_norm": 0.1844107061624527,
+      "learning_rate": 3.226263188606876e-05,
+      "loss": 0.4366,
       "step": 58845
     },
     {
-      "epoch": 2.07,
-      "learning_rate": 3.3035997872328304e-05,
-      "loss": 0.2648,
+      "epoch": 2.120950012613976,
+      "grad_norm": 0.17652803659439087,
+      "learning_rate": 3.225983955215922e-05,
+      "loss": 0.3898,
       "step": 58850
     },
     {
-      "epoch": 2.07,
-      "learning_rate": 3.303330028055616e-05,
-      "loss": 0.274,
+      "epoch": 2.121130212275201,
+      "grad_norm": 0.20647066831588745,
+      "learning_rate": 3.225704711933577e-05,
+      "loss": 0.3942,
       "step": 58855
     },
     {
-      "epoch": 2.07,
-      "learning_rate": 3.30306025844769e-05,
-      "loss": 0.2708,
+      "epoch": 2.1213104119364257,
+      "grad_norm": 0.24524955451488495,
+      "learning_rate": 3.2254254587636454e-05,
+      "loss": 0.4065,
       "step": 58860
     },
     {
-      "epoch": 2.07,
-      "learning_rate": 3.302790478412555e-05,
-      "loss": 0.2962,
+      "epoch": 2.1214906115976504,
+      "grad_norm": 0.23697435855865479,
+      "learning_rate": 3.225146195709933e-05,
+      "loss": 0.4129,
       "step": 58865
     },
     {
-      "epoch": 2.07,
-      "learning_rate": 3.3025206879537166e-05,
-      "loss": 0.2773,
+      "epoch": 2.1216708112588747,
+      "grad_norm": 0.2048749029636383,
+      "learning_rate": 3.2248669227762427e-05,
+      "loss": 0.39,
       "step": 58870
     },
     {
-      "epoch": 2.07,
-      "learning_rate": 3.302250887074674e-05,
-      "loss": 0.2597,
+      "epoch": 2.1218510109200994,
+      "grad_norm": 0.18117062747478485,
+      "learning_rate": 3.2245876399663826e-05,
+      "loss": 0.3636,
       "step": 58875
     },
     {
-      "epoch": 2.07,
-      "learning_rate": 3.3019810757789325e-05,
-      "loss": 0.2545,
+      "epoch": 2.122031210581324,
+      "grad_norm": 0.20858252048492432,
+      "learning_rate": 3.224308347284155e-05,
+      "loss": 0.3803,
       "step": 58880
     },
     {
-      "epoch": 2.07,
-      "learning_rate": 3.301711254069996e-05,
-      "loss": 0.2814,
+      "epoch": 2.122211410242549,
+      "grad_norm": 0.23120179772377014,
+      "learning_rate": 3.224029044733367e-05,
+      "loss": 0.4179,
       "step": 58885
     },
     {
-      "epoch": 2.07,
-      "learning_rate": 3.3014414219513676e-05,
-      "loss": 0.2782,
+      "epoch": 2.122391609903773,
+      "grad_norm": 0.20885401964187622,
+      "learning_rate": 3.2237497323178234e-05,
+      "loss": 0.3951,
       "step": 58890
     },
     {
-      "epoch": 2.07,
-      "learning_rate": 3.3011715794265494e-05,
-      "loss": 0.2859,
+      "epoch": 2.122571809564998,
+      "grad_norm": 0.22970037162303925,
+      "learning_rate": 3.223470410041329e-05,
+      "loss": 0.4036,
       "step": 58895
     },
     {
-      "epoch": 2.07,
-      "learning_rate": 3.300901726499046e-05,
-      "loss": 0.2637,
+      "epoch": 2.1227520092262226,
+      "grad_norm": 0.2186799943447113,
+      "learning_rate": 3.2231910779076916e-05,
+      "loss": 0.4109,
       "step": 58900
     },
     {
-      "epoch": 2.07,
-      "learning_rate": 3.300631863172363e-05,
-      "loss": 0.2925,
+      "epoch": 2.1229322088874474,
+      "grad_norm": 0.19462229311466217,
+      "learning_rate": 3.222911735920715e-05,
+      "loss": 0.4233,
       "step": 58905
     },
     {
-      "epoch": 2.07,
-      "learning_rate": 3.300361989450002e-05,
-      "loss": 0.2558,
+      "epoch": 2.123112408548672,
+      "grad_norm": 0.18175840377807617,
+      "learning_rate": 3.222632384084207e-05,
+      "loss": 0.3827,
       "step": 58910
     },
     {
-      "epoch": 2.07,
-      "learning_rate": 3.3000921053354675e-05,
-      "loss": 0.2598,
+      "epoch": 2.1232926082098964,
+      "grad_norm": 0.2110246866941452,
+      "learning_rate": 3.222353022401971e-05,
+      "loss": 0.3749,
       "step": 58915
     },
     {
-      "epoch": 2.07,
-      "learning_rate": 3.299822210832265e-05,
-      "loss": 0.2496,
+      "epoch": 2.123472807871121,
+      "grad_norm": 0.18468889594078064,
+      "learning_rate": 3.2220736508778166e-05,
+      "loss": 0.396,
       "step": 58920
     },
     {
-      "epoch": 2.07,
-      "learning_rate": 3.299552305943899e-05,
-      "loss": 0.2827,
+      "epoch": 2.123653007532346,
+      "grad_norm": 0.18783201277256012,
+      "learning_rate": 3.221794269515547e-05,
+      "loss": 0.3639,
       "step": 58925
     },
     {
-      "epoch": 2.07,
-      "learning_rate": 3.299282390673872e-05,
-      "loss": 0.2558,
+      "epoch": 2.1238332071935706,
+      "grad_norm": 0.21580557525157928,
+      "learning_rate": 3.2215148783189716e-05,
+      "loss": 0.4192,
       "step": 58930
     },
     {
-      "epoch": 2.07,
-      "learning_rate": 3.29901246502569e-05,
-      "loss": 0.2689,
+      "epoch": 2.124013406854795,
+      "grad_norm": 0.18120792508125305,
+      "learning_rate": 3.221235477291895e-05,
+      "loss": 0.3901,
       "step": 58935
     },
     {
-      "epoch": 2.07,
-      "learning_rate": 3.298742529002857e-05,
-      "loss": 0.282,
+      "epoch": 2.1241936065160196,
+      "grad_norm": 0.18788616359233856,
+      "learning_rate": 3.2209560664381244e-05,
+      "loss": 0.3894,
       "step": 58940
     },
     {
-      "epoch": 2.07,
-      "learning_rate": 3.298472582608881e-05,
-      "loss": 0.2818,
+      "epoch": 2.1243738061772444,
+      "grad_norm": 0.18515391647815704,
+      "learning_rate": 3.220676645761467e-05,
+      "loss": 0.4085,
       "step": 58945
     },
     {
-      "epoch": 2.07,
-      "learning_rate": 3.298202625847263e-05,
-      "loss": 0.2714,
+      "epoch": 2.124554005838469,
+      "grad_norm": 0.17624923586845398,
+      "learning_rate": 3.2203972152657294e-05,
+      "loss": 0.4079,
       "step": 58950
     },
     {
-      "epoch": 2.07,
-      "learning_rate": 3.297932658721511e-05,
-      "loss": 0.2718,
+      "epoch": 2.124734205499694,
+      "grad_norm": 0.2669857144355774,
+      "learning_rate": 3.22011777495472e-05,
+      "loss": 0.3783,
       "step": 58955
     },
     {
-      "epoch": 2.07,
-      "learning_rate": 3.2976626812351275e-05,
-      "loss": 0.2591,
+      "epoch": 2.124914405160918,
+      "grad_norm": 0.21977299451828003,
+      "learning_rate": 3.219838324832246e-05,
+      "loss": 0.3794,
       "step": 58960
     },
     {
-      "epoch": 2.07,
-      "learning_rate": 3.297392693391621e-05,
-      "loss": 0.277,
+      "epoch": 2.125094604822143,
+      "grad_norm": 0.18140974640846252,
+      "learning_rate": 3.219558864902113e-05,
+      "loss": 0.3855,
       "step": 58965
     },
     {
-      "epoch": 2.07,
-      "learning_rate": 3.2971226951944955e-05,
-      "loss": 0.26,
+      "epoch": 2.1252748044833676,
+      "grad_norm": 0.2620275020599365,
+      "learning_rate": 3.21927939516813e-05,
+      "loss": 0.3959,
       "step": 58970
     },
     {
-      "epoch": 2.07,
-      "learning_rate": 3.2968526866472576e-05,
-      "loss": 0.269,
+      "epoch": 2.1254550041445923,
+      "grad_norm": 0.2260800153017044,
+      "learning_rate": 3.218999915634105e-05,
+      "loss": 0.3754,
       "step": 58975
     },
     {
-      "epoch": 2.08,
-      "learning_rate": 3.2965826677534126e-05,
-      "loss": 0.2814,
+      "epoch": 2.125635203805817,
+      "grad_norm": 0.22423382103443146,
+      "learning_rate": 3.218720426303845e-05,
+      "loss": 0.3836,
       "step": 58980
     },
     {
-      "epoch": 2.08,
-      "learning_rate": 3.296312638516466e-05,
-      "loss": 0.3001,
+      "epoch": 2.1258154034670413,
+      "grad_norm": 0.21626518666744232,
+      "learning_rate": 3.2184409271811586e-05,
+      "loss": 0.4172,
       "step": 58985
     },
     {
-      "epoch": 2.08,
-      "learning_rate": 3.296042598939925e-05,
-      "loss": 0.2824,
+      "epoch": 2.125995603128266,
+      "grad_norm": 0.2051631510257721,
+      "learning_rate": 3.218161418269853e-05,
+      "loss": 0.384,
       "step": 58990
     },
     {
-      "epoch": 2.08,
-      "learning_rate": 3.295772549027296e-05,
-      "loss": 0.256,
+      "epoch": 2.126175802789491,
+      "grad_norm": 0.16707560420036316,
+      "learning_rate": 3.217881899573738e-05,
+      "loss": 0.4404,
       "step": 58995
     },
     {
-      "epoch": 2.08,
-      "learning_rate": 3.295502488782084e-05,
-      "loss": 0.2646,
+      "epoch": 2.1263560024507155,
+      "grad_norm": 0.20359165966510773,
+      "learning_rate": 3.21760237109662e-05,
+      "loss": 0.383,
       "step": 59000
     },
     {
-      "epoch": 2.08,
-      "eval_loss": 0.27035436034202576,
-      "eval_runtime": 10.5421,
-      "eval_samples_per_second": 9.486,
-      "eval_steps_per_second": 9.486,
+      "epoch": 2.1263560024507155,
+      "eval_loss": 0.43649742007255554,
+      "eval_runtime": 3.5299,
+      "eval_samples_per_second": 28.33,
+      "eval_steps_per_second": 7.082,
       "step": 59000
     },
     {
-      "epoch": 2.08,
-      "learning_rate": 3.295232418207798e-05,
-      "loss": 0.275,
+      "epoch": 2.12653620211194,
+      "grad_norm": 0.24366778135299683,
+      "learning_rate": 3.2173228328423095e-05,
+      "loss": 0.3883,
       "step": 59005
     },
     {
-      "epoch": 2.08,
-      "learning_rate": 3.2949623373079406e-05,
-      "loss": 0.2762,
+      "epoch": 2.1267164017731646,
+      "grad_norm": 0.18067045509815216,
+      "learning_rate": 3.217043284814614e-05,
+      "loss": 0.4133,
       "step": 59010
     },
     {
-      "epoch": 2.08,
-      "learning_rate": 3.294692246086023e-05,
-      "loss": 0.2561,
+      "epoch": 2.1268966014343893,
+      "grad_norm": 0.18387949466705322,
+      "learning_rate": 3.2167637270173425e-05,
+      "loss": 0.4048,
       "step": 59015
     },
     {
-      "epoch": 2.08,
-      "learning_rate": 3.2944221445455494e-05,
-      "loss": 0.2895,
+      "epoch": 2.127076801095614,
+      "grad_norm": 0.19910840690135956,
+      "learning_rate": 3.2164841594543044e-05,
+      "loss": 0.4296,
       "step": 59020
     },
     {
-      "epoch": 2.08,
-      "learning_rate": 3.294152032690029e-05,
-      "loss": 0.2872,
+      "epoch": 2.1272570007568388,
+      "grad_norm": 0.24461115896701813,
+      "learning_rate": 3.216204582129308e-05,
+      "loss": 0.3926,
       "step": 59025
     },
     {
-      "epoch": 2.08,
-      "learning_rate": 3.293881910522967e-05,
-      "loss": 0.2779,
+      "epoch": 2.127437200418063,
+      "grad_norm": 0.2087002843618393,
+      "learning_rate": 3.215924995046163e-05,
+      "loss": 0.4611,
       "step": 59030
     },
     {
-      "epoch": 2.08,
-      "learning_rate": 3.293611778047871e-05,
-      "loss": 0.2808,
+      "epoch": 2.127617400079288,
+      "grad_norm": 0.20317687094211578,
+      "learning_rate": 3.215645398208678e-05,
+      "loss": 0.4292,
       "step": 59035
     },
     {
-      "epoch": 2.08,
-      "learning_rate": 3.293341635268249e-05,
-      "loss": 0.2849,
+      "epoch": 2.1277975997405125,
+      "grad_norm": 0.17067931592464447,
+      "learning_rate": 3.215365791620664e-05,
+      "loss": 0.4072,
       "step": 59040
     },
     {
-      "epoch": 2.08,
-      "learning_rate": 3.2930714821876094e-05,
-      "loss": 0.242,
+      "epoch": 2.1279777994017373,
+      "grad_norm": 0.23524482548236847,
+      "learning_rate": 3.2150861752859286e-05,
+      "loss": 0.405,
       "step": 59045
     },
     {
-      "epoch": 2.08,
-      "learning_rate": 3.29280131880946e-05,
-      "loss": 0.2622,
+      "epoch": 2.128157999062962,
+      "grad_norm": 0.24678029119968414,
+      "learning_rate": 3.214806549208283e-05,
+      "loss": 0.4318,
       "step": 59050
     },
     {
-      "epoch": 2.08,
-      "learning_rate": 3.292531145137307e-05,
-      "loss": 0.2645,
+      "epoch": 2.1283381987241863,
+      "grad_norm": 0.22866828739643097,
+      "learning_rate": 3.214526913391536e-05,
+      "loss": 0.3886,
       "step": 59055
     },
     {
-      "epoch": 2.08,
-      "learning_rate": 3.2922609611746595e-05,
-      "loss": 0.2838,
+      "epoch": 2.128518398385411,
+      "grad_norm": 0.2560563385486603,
+      "learning_rate": 3.2142472678395e-05,
+      "loss": 0.422,
       "step": 59060
     },
     {
-      "epoch": 2.08,
-      "learning_rate": 3.291990766925026e-05,
-      "loss": 0.2518,
+      "epoch": 2.1286985980466357,
+      "grad_norm": 0.23088377714157104,
+      "learning_rate": 3.213967612555981e-05,
+      "loss": 0.4044,
       "step": 59065
     },
     {
-      "epoch": 2.08,
-      "learning_rate": 3.2917205623919134e-05,
-      "loss": 0.2829,
+      "epoch": 2.1288787977078605,
+      "grad_norm": 0.21493729948997498,
+      "learning_rate": 3.2136879475447924e-05,
+      "loss": 0.3938,
       "step": 59070
     },
     {
-      "epoch": 2.08,
-      "learning_rate": 3.291450347578832e-05,
-      "loss": 0.2837,
+      "epoch": 2.1290589973690848,
+      "grad_norm": 0.1736428588628769,
+      "learning_rate": 3.213408272809744e-05,
+      "loss": 0.4205,
       "step": 59075
     },
     {
-      "epoch": 2.08,
-      "learning_rate": 3.2911801224892894e-05,
-      "loss": 0.2655,
+      "epoch": 2.1292391970303095,
+      "grad_norm": 0.18080323934555054,
+      "learning_rate": 3.213128588354645e-05,
+      "loss": 0.363,
       "step": 59080
     },
     {
-      "epoch": 2.08,
-      "learning_rate": 3.2909098871267936e-05,
-      "loss": 0.2522,
+      "epoch": 2.1294193966915342,
+      "grad_norm": 0.19538411498069763,
+      "learning_rate": 3.212848894183308e-05,
+      "loss": 0.3855,
       "step": 59085
     },
     {
-      "epoch": 2.08,
-      "learning_rate": 3.290639641494855e-05,
-      "loss": 0.2826,
+      "epoch": 2.129599596352759,
+      "grad_norm": 0.2233283966779709,
+      "learning_rate": 3.212569190299542e-05,
+      "loss": 0.4047,
       "step": 59090
     },
     {
-      "epoch": 2.08,
-      "learning_rate": 3.290369385596981e-05,
-      "loss": 0.2664,
+      "epoch": 2.1297797960139837,
+      "grad_norm": 0.19378229975700378,
+      "learning_rate": 3.21228947670716e-05,
+      "loss": 0.3827,
       "step": 59095
     },
     {
-      "epoch": 2.08,
-      "learning_rate": 3.290099119436681e-05,
-      "loss": 0.2599,
+      "epoch": 2.129959995675208,
+      "grad_norm": 0.18152061104774475,
+      "learning_rate": 3.212009753409971e-05,
+      "loss": 0.4019,
       "step": 59100
     },
     {
-      "epoch": 2.08,
-      "learning_rate": 3.289828843017465e-05,
-      "loss": 0.2667,
+      "epoch": 2.1301401953364327,
+      "grad_norm": 0.2508603632450104,
+      "learning_rate": 3.2117300204117876e-05,
+      "loss": 0.399,
       "step": 59105
     },
     {
-      "epoch": 2.08,
-      "learning_rate": 3.289558556342843e-05,
-      "loss": 0.2931,
+      "epoch": 2.1303203949976575,
+      "grad_norm": 0.16633902490139008,
+      "learning_rate": 3.211450277716419e-05,
+      "loss": 0.3941,
       "step": 59110
     },
     {
-      "epoch": 2.08,
-      "learning_rate": 3.289288259416322e-05,
-      "loss": 0.2535,
+      "epoch": 2.130500594658882,
+      "grad_norm": 0.1936190277338028,
+      "learning_rate": 3.211170525327679e-05,
+      "loss": 0.4041,
       "step": 59115
     },
     {
-      "epoch": 2.08,
-      "learning_rate": 3.289017952241414e-05,
-      "loss": 0.265,
+      "epoch": 2.1306807943201065,
+      "grad_norm": 0.2269958257675171,
+      "learning_rate": 3.210890763249379e-05,
+      "loss": 0.4195,
       "step": 59120
     },
     {
-      "epoch": 2.08,
-      "learning_rate": 3.2887476348216275e-05,
-      "loss": 0.2608,
+      "epoch": 2.130860993981331,
+      "grad_norm": 0.18237541615962982,
+      "learning_rate": 3.210610991485329e-05,
+      "loss": 0.3866,
       "step": 59125
     },
     {
-      "epoch": 2.08,
-      "learning_rate": 3.288477307160474e-05,
-      "loss": 0.269,
+      "epoch": 2.131041193642556,
+      "grad_norm": 0.19430314004421234,
+      "learning_rate": 3.210331210039342e-05,
+      "loss": 0.4121,
       "step": 59130
     },
     {
-      "epoch": 2.08,
-      "learning_rate": 3.28820696926146e-05,
-      "loss": 0.2513,
+      "epoch": 2.1312213933037807,
+      "grad_norm": 0.18985673785209656,
+      "learning_rate": 3.2100514189152297e-05,
+      "loss": 0.4231,
       "step": 59135
     },
     {
-      "epoch": 2.08,
-      "learning_rate": 3.2879366211281e-05,
-      "loss": 0.2632,
+      "epoch": 2.1314015929650054,
+      "grad_norm": 0.19704566895961761,
+      "learning_rate": 3.209771618116805e-05,
+      "loss": 0.3891,
       "step": 59140
     },
     {
-      "epoch": 2.08,
-      "learning_rate": 3.287666262763901e-05,
-      "loss": 0.2539,
+      "epoch": 2.1315817926262297,
+      "grad_norm": 0.20834749937057495,
+      "learning_rate": 3.209491807647879e-05,
+      "loss": 0.3716,
       "step": 59145
     },
     {
-      "epoch": 2.08,
-      "learning_rate": 3.2873958941723745e-05,
-      "loss": 0.26,
+      "epoch": 2.1317619922874544,
+      "grad_norm": 0.15605294704437256,
+      "learning_rate": 3.2092119875122636e-05,
+      "loss": 0.3832,
       "step": 59150
     },
     {
-      "epoch": 2.08,
-      "learning_rate": 3.287125515357032e-05,
-      "loss": 0.2569,
+      "epoch": 2.131942191948679,
+      "grad_norm": 0.19391027092933655,
+      "learning_rate": 3.208932157713773e-05,
+      "loss": 0.4034,
       "step": 59155
     },
     {
-      "epoch": 2.08,
-      "learning_rate": 3.286855126321383e-05,
-      "loss": 0.2446,
+      "epoch": 2.132122391609904,
+      "grad_norm": 0.20829804241657257,
+      "learning_rate": 3.208652318256219e-05,
+      "loss": 0.3805,
       "step": 59160
     },
     {
-      "epoch": 2.08,
-      "learning_rate": 3.28658472706894e-05,
-      "loss": 0.2793,
+      "epoch": 2.132302591271128,
+      "grad_norm": 0.21649335324764252,
+      "learning_rate": 3.2083724691434145e-05,
+      "loss": 0.4418,
       "step": 59165
     },
     {
-      "epoch": 2.08,
-      "learning_rate": 3.286314317603211e-05,
-      "loss": 0.276,
+      "epoch": 2.132482790932353,
+      "grad_norm": 0.18842458724975586,
+      "learning_rate": 3.208092610379172e-05,
+      "loss": 0.396,
       "step": 59170
     },
     {
-      "epoch": 2.08,
-      "learning_rate": 3.2860438979277096e-05,
-      "loss": 0.294,
+      "epoch": 2.1326629905935777,
+      "grad_norm": 0.21697010099887848,
+      "learning_rate": 3.2078127419673046e-05,
+      "loss": 0.3576,
       "step": 59175
     },
     {
-      "epoch": 2.08,
-      "learning_rate": 3.2857734680459455e-05,
-      "loss": 0.2766,
+      "epoch": 2.1328431902548024,
+      "grad_norm": 0.20509681105613708,
+      "learning_rate": 3.207532863911627e-05,
+      "loss": 0.4059,
       "step": 59180
     },
     {
-      "epoch": 2.08,
-      "learning_rate": 3.2855030279614316e-05,
-      "loss": 0.2468,
+      "epoch": 2.133023389916027,
+      "grad_norm": 0.19861997663974762,
+      "learning_rate": 3.2072529762159495e-05,
+      "loss": 0.3848,
       "step": 59185
     },
     {
-      "epoch": 2.08,
-      "learning_rate": 3.285232577677679e-05,
-      "loss": 0.2683,
+      "epoch": 2.1332035895772514,
+      "grad_norm": 0.17639335989952087,
+      "learning_rate": 3.206973078884087e-05,
+      "loss": 0.4051,
       "step": 59190
     },
     {
-      "epoch": 2.08,
-      "learning_rate": 3.284962117198198e-05,
-      "loss": 0.266,
+      "epoch": 2.133383789238476,
+      "grad_norm": 0.2027795910835266,
+      "learning_rate": 3.206693171919854e-05,
+      "loss": 0.4244,
       "step": 59195
     },
     {
-      "epoch": 2.08,
-      "learning_rate": 3.2846916465265016e-05,
-      "loss": 0.2587,
+      "epoch": 2.133563988899701,
+      "grad_norm": 0.18750731647014618,
+      "learning_rate": 3.206413255327063e-05,
+      "loss": 0.4224,
       "step": 59200
     },
     {
-      "epoch": 2.08,
-      "learning_rate": 3.284421165666101e-05,
-      "loss": 0.2774,
+      "epoch": 2.1337441885609256,
+      "grad_norm": 0.21276669204235077,
+      "learning_rate": 3.206133329109529e-05,
+      "loss": 0.4186,
       "step": 59205
     },
     {
-      "epoch": 2.08,
-      "learning_rate": 3.284150674620508e-05,
-      "loss": 0.2789,
+      "epoch": 2.13392438822215,
+      "grad_norm": 0.17720721662044525,
+      "learning_rate": 3.205853393271064e-05,
+      "loss": 0.3877,
       "step": 59210
     },
     {
-      "epoch": 2.08,
-      "learning_rate": 3.283880173393236e-05,
-      "loss": 0.2867,
+      "epoch": 2.1341045878833746,
+      "grad_norm": 0.2149345576763153,
+      "learning_rate": 3.2055734478154834e-05,
+      "loss": 0.3937,
       "step": 59215
     },
     {
-      "epoch": 2.08,
-      "learning_rate": 3.283609661987797e-05,
-      "loss": 0.2893,
+      "epoch": 2.1342847875445994,
+      "grad_norm": 0.18279078602790833,
+      "learning_rate": 3.2052934927466016e-05,
+      "loss": 0.3948,
       "step": 59220
     },
     {
-      "epoch": 2.08,
-      "learning_rate": 3.2833391404077025e-05,
-      "loss": 0.3004,
+      "epoch": 2.134464987205824,
+      "grad_norm": 0.24067635834217072,
+      "learning_rate": 3.2050135280682327e-05,
+      "loss": 0.4092,
       "step": 59225
     },
     {
-      "epoch": 2.08,
-      "learning_rate": 3.283068608656465e-05,
-      "loss": 0.2544,
+      "epoch": 2.134645186867049,
+      "grad_norm": 0.22247518599033356,
+      "learning_rate": 3.20473355378419e-05,
+      "loss": 0.4003,
       "step": 59230
     },
     {
-      "epoch": 2.08,
-      "learning_rate": 3.282798066737598e-05,
-      "loss": 0.2583,
+      "epoch": 2.134825386528273,
+      "grad_norm": 0.16726024448871613,
+      "learning_rate": 3.2044535698982895e-05,
+      "loss": 0.3806,
       "step": 59235
     },
     {
-      "epoch": 2.08,
-      "learning_rate": 3.2825275146546136e-05,
-      "loss": 0.2954,
+      "epoch": 2.135005586189498,
+      "grad_norm": 0.18711987137794495,
+      "learning_rate": 3.204173576414345e-05,
+      "loss": 0.3994,
       "step": 59240
     },
     {
-      "epoch": 2.08,
-      "learning_rate": 3.282256952411026e-05,
-      "loss": 0.2731,
+      "epoch": 2.1351857858507226,
+      "grad_norm": 0.1729484349489212,
+      "learning_rate": 3.2038935733361734e-05,
+      "loss": 0.3954,
       "step": 59245
     },
     {
-      "epoch": 2.08,
-      "learning_rate": 3.2819863800103466e-05,
-      "loss": 0.2743,
+      "epoch": 2.1353659855119473,
+      "grad_norm": 0.16871348023414612,
+      "learning_rate": 3.203613560667587e-05,
+      "loss": 0.3775,
       "step": 59250
     },
     {
-      "epoch": 2.08,
-      "learning_rate": 3.2817157974560904e-05,
-      "loss": 0.2589,
+      "epoch": 2.135546185173172,
+      "grad_norm": 0.21796941757202148,
+      "learning_rate": 3.203333538412402e-05,
+      "loss": 0.4258,
       "step": 59255
     },
     {
-      "epoch": 2.08,
-      "learning_rate": 3.2814452047517685e-05,
-      "loss": 0.2905,
+      "epoch": 2.1357263848343964,
+      "grad_norm": 0.15358436107635498,
+      "learning_rate": 3.203053506574434e-05,
+      "loss": 0.3841,
       "step": 59260
     },
     {
-      "epoch": 2.09,
-      "learning_rate": 3.2811746019008966e-05,
-      "loss": 0.2651,
+      "epoch": 2.135906584495621,
+      "grad_norm": 0.2047918438911438,
+      "learning_rate": 3.202773465157498e-05,
+      "loss": 0.4093,
       "step": 59265
     },
     {
-      "epoch": 2.09,
-      "learning_rate": 3.280903988906986e-05,
-      "loss": 0.2655,
+      "epoch": 2.136086784156846,
+      "grad_norm": 0.22522784769535065,
+      "learning_rate": 3.20249341416541e-05,
+      "loss": 0.3782,
       "step": 59270
     },
     {
-      "epoch": 2.09,
-      "learning_rate": 3.280633365773553e-05,
-      "loss": 0.2618,
+      "epoch": 2.1362669838180706,
+      "grad_norm": 0.20467866957187653,
+      "learning_rate": 3.202213353601985e-05,
+      "loss": 0.3872,
       "step": 59275
     },
     {
-      "epoch": 2.09,
-      "learning_rate": 3.28036273250411e-05,
-      "loss": 0.2809,
+      "epoch": 2.136447183479295,
+      "grad_norm": 0.24087628722190857,
+      "learning_rate": 3.201933283471039e-05,
+      "loss": 0.4176,
       "step": 59280
     },
     {
-      "epoch": 2.09,
-      "learning_rate": 3.2800920891021714e-05,
-      "loss": 0.2727,
+      "epoch": 2.1366273831405196,
+      "grad_norm": 0.20225749909877777,
+      "learning_rate": 3.201653203776388e-05,
+      "loss": 0.3515,
       "step": 59285
     },
     {
-      "epoch": 2.09,
-      "learning_rate": 3.27982143557125e-05,
-      "loss": 0.2796,
+      "epoch": 2.1368075828017443,
+      "grad_norm": 0.24749909341335297,
+      "learning_rate": 3.201373114521847e-05,
+      "loss": 0.3979,
       "step": 59290
     },
     {
-      "epoch": 2.09,
-      "learning_rate": 3.2795507719148625e-05,
-      "loss": 0.2735,
+      "epoch": 2.136987782462969,
+      "grad_norm": 0.14851811528205872,
+      "learning_rate": 3.201093015711234e-05,
+      "loss": 0.3873,
       "step": 59295
     },
     {
-      "epoch": 2.09,
-      "learning_rate": 3.279280098136521e-05,
-      "loss": 0.246,
+      "epoch": 2.137167982124194,
+      "grad_norm": 0.2001311480998993,
+      "learning_rate": 3.200812907348364e-05,
+      "loss": 0.3825,
       "step": 59300
     },
     {
-      "epoch": 2.09,
-      "learning_rate": 3.279009414239742e-05,
-      "loss": 0.2973,
+      "epoch": 2.137348181785418,
+      "grad_norm": 0.1819891780614853,
+      "learning_rate": 3.200532789437055e-05,
+      "loss": 0.4069,
       "step": 59305
     },
     {
-      "epoch": 2.09,
-      "learning_rate": 3.278738720228039e-05,
-      "loss": 0.2841,
+      "epoch": 2.137528381446643,
+      "grad_norm": 0.20626020431518555,
+      "learning_rate": 3.200252661981121e-05,
+      "loss": 0.4005,
       "step": 59310
     },
     {
-      "epoch": 2.09,
-      "learning_rate": 3.278468016104926e-05,
-      "loss": 0.2841,
+      "epoch": 2.1377085811078675,
+      "grad_norm": 0.15189993381500244,
+      "learning_rate": 3.1999725249843806e-05,
+      "loss": 0.3802,
       "step": 59315
     },
     {
-      "epoch": 2.09,
-      "learning_rate": 3.278197301873919e-05,
-      "loss": 0.2635,
+      "epoch": 2.1378887807690923,
+      "grad_norm": 0.2139143943786621,
+      "learning_rate": 3.1996923784506494e-05,
+      "loss": 0.3658,
       "step": 59320
     },
     {
-      "epoch": 2.09,
-      "learning_rate": 3.2779265775385335e-05,
-      "loss": 0.2695,
+      "epoch": 2.138068980430317,
+      "grad_norm": 0.17770539224147797,
+      "learning_rate": 3.199412222383746e-05,
+      "loss": 0.3744,
       "step": 59325
     },
     {
-      "epoch": 2.09,
-      "learning_rate": 3.277655843102284e-05,
-      "loss": 0.2893,
+      "epoch": 2.1382491800915413,
+      "grad_norm": 0.18923962116241455,
+      "learning_rate": 3.1991320567874863e-05,
+      "loss": 0.399,
       "step": 59330
     },
     {
-      "epoch": 2.09,
-      "learning_rate": 3.277385098568685e-05,
-      "loss": 0.2398,
+      "epoch": 2.138429379752766,
+      "grad_norm": 0.2178812175989151,
+      "learning_rate": 3.198851881665687e-05,
+      "loss": 0.3778,
       "step": 59335
     },
     {
-      "epoch": 2.09,
-      "learning_rate": 3.277114343941254e-05,
-      "loss": 0.2809,
+      "epoch": 2.1386095794139908,
+      "grad_norm": 0.16876627504825592,
+      "learning_rate": 3.198571697022167e-05,
+      "loss": 0.4179,
       "step": 59340
     },
     {
-      "epoch": 2.09,
-      "learning_rate": 3.2768435792235044e-05,
-      "loss": 0.2598,
+      "epoch": 2.1387897790752155,
+      "grad_norm": 0.17547792196273804,
+      "learning_rate": 3.198291502860742e-05,
+      "loss": 0.3674,
       "step": 59345
     },
     {
-      "epoch": 2.09,
-      "learning_rate": 3.276572804418954e-05,
-      "loss": 0.2785,
+      "epoch": 2.13896997873644,
+      "grad_norm": 0.20084445178508759,
+      "learning_rate": 3.198011299185232e-05,
+      "loss": 0.3989,
       "step": 59350
     },
     {
-      "epoch": 2.09,
-      "learning_rate": 3.276302019531116e-05,
-      "loss": 0.2738,
+      "epoch": 2.1391501783976645,
+      "grad_norm": 0.1786831021308899,
+      "learning_rate": 3.197731085999451e-05,
+      "loss": 0.3793,
       "step": 59355
     },
     {
-      "epoch": 2.09,
-      "learning_rate": 3.2760312245635094e-05,
-      "loss": 0.293,
+      "epoch": 2.1393303780588893,
+      "grad_norm": 0.1777740865945816,
+      "learning_rate": 3.197450863307221e-05,
+      "loss": 0.3599,
       "step": 59360
     },
     {
-      "epoch": 2.09,
-      "learning_rate": 3.2757604195196477e-05,
-      "loss": 0.2736,
+      "epoch": 2.139510577720114,
+      "grad_norm": 0.19308814406394958,
+      "learning_rate": 3.1971706311123564e-05,
+      "loss": 0.422,
       "step": 59365
     },
     {
-      "epoch": 2.09,
-      "learning_rate": 3.275489604403048e-05,
-      "loss": 0.2656,
+      "epoch": 2.1396907773813387,
+      "grad_norm": 0.23540256917476654,
+      "learning_rate": 3.196890389418678e-05,
+      "loss": 0.4141,
       "step": 59370
     },
     {
-      "epoch": 2.09,
-      "learning_rate": 3.2752187792172276e-05,
-      "loss": 0.2904,
+      "epoch": 2.139870977042563,
+      "grad_norm": 0.2817511260509491,
+      "learning_rate": 3.196610138230003e-05,
+      "loss": 0.3921,
       "step": 59375
     },
     {
-      "epoch": 2.09,
-      "learning_rate": 3.274947943965703e-05,
-      "loss": 0.2758,
+      "epoch": 2.1400511767037877,
+      "grad_norm": 0.2218995839357376,
+      "learning_rate": 3.196329877550149e-05,
+      "loss": 0.3879,
       "step": 59380
     },
     {
-      "epoch": 2.09,
-      "learning_rate": 3.274677098651989e-05,
-      "loss": 0.2808,
+      "epoch": 2.1402313763650125,
+      "grad_norm": 0.1908760666847229,
+      "learning_rate": 3.196049607382936e-05,
+      "loss": 0.3864,
       "step": 59385
     },
     {
-      "epoch": 2.09,
-      "learning_rate": 3.274406243279603e-05,
-      "loss": 0.3196,
+      "epoch": 2.140411576026237,
+      "grad_norm": 0.22384685277938843,
+      "learning_rate": 3.195769327732181e-05,
+      "loss": 0.3835,
       "step": 59390
     },
     {
-      "epoch": 2.09,
-      "learning_rate": 3.2741353778520636e-05,
-      "loss": 0.2627,
+      "epoch": 2.1405917756874615,
+      "grad_norm": 0.20056529343128204,
+      "learning_rate": 3.195489038601704e-05,
+      "loss": 0.3914,
       "step": 59395
     },
     {
-      "epoch": 2.09,
-      "learning_rate": 3.273864502372885e-05,
-      "loss": 0.2666,
+      "epoch": 2.1407719753486862,
+      "grad_norm": 0.1963140219449997,
+      "learning_rate": 3.1952087399953236e-05,
+      "loss": 0.4141,
       "step": 59400
     },
     {
-      "epoch": 2.09,
-      "learning_rate": 3.2735936168455864e-05,
-      "loss": 0.2909,
+      "epoch": 2.140952175009911,
+      "grad_norm": 0.20005357265472412,
+      "learning_rate": 3.194928431916858e-05,
+      "loss": 0.4379,
       "step": 59405
     },
     {
-      "epoch": 2.09,
-      "learning_rate": 3.273322721273685e-05,
-      "loss": 0.2701,
+      "epoch": 2.1411323746711357,
+      "grad_norm": 0.1524960994720459,
+      "learning_rate": 3.194648114370129e-05,
+      "loss": 0.3916,
       "step": 59410
     },
     {
-      "epoch": 2.09,
-      "learning_rate": 3.273051815660698e-05,
-      "loss": 0.266,
+      "epoch": 2.1413125743323604,
+      "grad_norm": 0.20259173214435577,
+      "learning_rate": 3.1943677873589525e-05,
+      "loss": 0.4089,
       "step": 59415
     },
     {
-      "epoch": 2.09,
-      "learning_rate": 3.2727809000101413e-05,
-      "loss": 0.2632,
+      "epoch": 2.1414927739935847,
+      "grad_norm": 0.23878759145736694,
+      "learning_rate": 3.1940874508871496e-05,
+      "loss": 0.4275,
       "step": 59420
     },
     {
-      "epoch": 2.09,
-      "learning_rate": 3.2725099743255345e-05,
-      "loss": 0.2612,
+      "epoch": 2.1416729736548095,
+      "grad_norm": 0.15623390674591064,
+      "learning_rate": 3.193807104958539e-05,
+      "loss": 0.4258,
       "step": 59425
     },
     {
-      "epoch": 2.09,
-      "learning_rate": 3.2722390386103946e-05,
-      "loss": 0.2551,
+      "epoch": 2.141853173316034,
+      "grad_norm": 0.1934848129749298,
+      "learning_rate": 3.1935267495769416e-05,
+      "loss": 0.3741,
       "step": 59430
     },
     {
-      "epoch": 2.09,
-      "learning_rate": 3.271968092868241e-05,
-      "loss": 0.2569,
+      "epoch": 2.142033372977259,
+      "grad_norm": 0.2278786599636078,
+      "learning_rate": 3.193246384746176e-05,
+      "loss": 0.417,
       "step": 59435
     },
     {
-      "epoch": 2.09,
-      "learning_rate": 3.271697137102588e-05,
-      "loss": 0.2754,
+      "epoch": 2.142213572638483,
+      "grad_norm": 0.22830450534820557,
+      "learning_rate": 3.192966010470063e-05,
+      "loss": 0.3854,
       "step": 59440
     },
     {
-      "epoch": 2.09,
-      "learning_rate": 3.271426171316959e-05,
-      "loss": 0.2608,
+      "epoch": 2.142393772299708,
+      "grad_norm": 0.236038476228714,
+      "learning_rate": 3.192685626752422e-05,
+      "loss": 0.4147,
       "step": 59445
     },
     {
-      "epoch": 2.09,
-      "learning_rate": 3.271155195514868e-05,
-      "loss": 0.2745,
+      "epoch": 2.1425739719609327,
+      "grad_norm": 0.2309621423482895,
+      "learning_rate": 3.1924052335970736e-05,
+      "loss": 0.4035,
       "step": 59450
     },
     {
-      "epoch": 2.09,
-      "learning_rate": 3.2708842096998346e-05,
-      "loss": 0.2517,
+      "epoch": 2.1427541716221574,
+      "grad_norm": 0.2083134800195694,
+      "learning_rate": 3.1921248310078386e-05,
+      "loss": 0.4002,
       "step": 59455
     },
     {
-      "epoch": 2.09,
-      "learning_rate": 3.2706132138753784e-05,
-      "loss": 0.2879,
+      "epoch": 2.142934371283382,
+      "grad_norm": 0.22884568572044373,
+      "learning_rate": 3.1918444189885355e-05,
+      "loss": 0.4022,
       "step": 59460
     },
     {
-      "epoch": 2.09,
-      "learning_rate": 3.270342208045018e-05,
-      "loss": 0.2594,
+      "epoch": 2.1431145709446064,
+      "grad_norm": 0.18977230787277222,
+      "learning_rate": 3.191563997542987e-05,
+      "loss": 0.375,
       "step": 59465
     },
     {
-      "epoch": 2.09,
-      "learning_rate": 3.270071192212271e-05,
-      "loss": 0.2515,
+      "epoch": 2.143294770605831,
+      "grad_norm": 0.21738559007644653,
+      "learning_rate": 3.191283566675013e-05,
+      "loss": 0.3757,
       "step": 59470
     },
     {
-      "epoch": 2.09,
-      "learning_rate": 3.2698001663806575e-05,
-      "loss": 0.2823,
+      "epoch": 2.143474970267056,
+      "grad_norm": 0.19232676923274994,
+      "learning_rate": 3.1910031263884335e-05,
+      "loss": 0.4058,
       "step": 59475
     },
     {
-      "epoch": 2.09,
-      "learning_rate": 3.269529130553695e-05,
-      "loss": 0.2801,
+      "epoch": 2.1436551699282806,
+      "grad_norm": 0.21828657388687134,
+      "learning_rate": 3.1907226766870714e-05,
+      "loss": 0.4174,
       "step": 59480
     },
     {
-      "epoch": 2.09,
-      "learning_rate": 3.269258084734905e-05,
-      "loss": 0.2701,
+      "epoch": 2.1438353695895054,
+      "grad_norm": 0.19182120263576508,
+      "learning_rate": 3.190442217574745e-05,
+      "loss": 0.3884,
       "step": 59485
     },
     {
-      "epoch": 2.09,
-      "learning_rate": 3.268987028927806e-05,
-      "loss": 0.2805,
+      "epoch": 2.1440155692507297,
+      "grad_norm": 0.16395203769207,
+      "learning_rate": 3.190161749055279e-05,
+      "loss": 0.3748,
       "step": 59490
     },
     {
-      "epoch": 2.09,
-      "learning_rate": 3.2687159631359164e-05,
-      "loss": 0.2796,
+      "epoch": 2.1441957689119544,
+      "grad_norm": 0.18212929368019104,
+      "learning_rate": 3.189881271132491e-05,
+      "loss": 0.3492,
       "step": 59495
     },
     {
-      "epoch": 2.09,
-      "learning_rate": 3.268444887362757e-05,
-      "loss": 0.2773,
+      "epoch": 2.144375968573179,
+      "grad_norm": 0.19863685965538025,
+      "learning_rate": 3.189600783810205e-05,
+      "loss": 0.3785,
       "step": 59500
     },
     {
-      "epoch": 2.09,
-      "eval_loss": 0.26948508620262146,
-      "eval_runtime": 10.5443,
-      "eval_samples_per_second": 9.484,
-      "eval_steps_per_second": 9.484,
+      "epoch": 2.144375968573179,
+      "eval_loss": 0.43618133664131165,
+      "eval_runtime": 3.5383,
+      "eval_samples_per_second": 28.262,
+      "eval_steps_per_second": 7.065,
       "step": 59500
     },
     {
-      "epoch": 2.09,
-      "learning_rate": 3.268173801611847e-05,
-      "loss": 0.2598,
+      "epoch": 2.144556168234404,
+      "grad_norm": 0.17493978142738342,
+      "learning_rate": 3.1893202870922414e-05,
+      "loss": 0.3907,
       "step": 59505
     },
     {
-      "epoch": 2.09,
-      "learning_rate": 3.267902705886706e-05,
-      "loss": 0.2798,
+      "epoch": 2.144736367895628,
+      "grad_norm": 0.18247364461421967,
+      "learning_rate": 3.189039780982423e-05,
+      "loss": 0.4129,
       "step": 59510
     },
     {
-      "epoch": 2.09,
-      "learning_rate": 3.267631600190855e-05,
-      "loss": 0.2612,
+      "epoch": 2.144916567556853,
+      "grad_norm": 0.2383251041173935,
+      "learning_rate": 3.188759265484571e-05,
+      "loss": 0.3971,
       "step": 59515
     },
     {
-      "epoch": 2.09,
-      "learning_rate": 3.267360484527814e-05,
-      "loss": 0.2736,
+      "epoch": 2.1450967672180776,
+      "grad_norm": 0.1882295161485672,
+      "learning_rate": 3.188478740602506e-05,
+      "loss": 0.3815,
       "step": 59520
     },
     {
-      "epoch": 2.09,
-      "learning_rate": 3.2670893589011025e-05,
-      "loss": 0.2493,
+      "epoch": 2.1452769668793024,
+      "grad_norm": 0.19644425809383392,
+      "learning_rate": 3.1881982063400526e-05,
+      "loss": 0.3999,
       "step": 59525
     },
     {
-      "epoch": 2.09,
-      "learning_rate": 3.2668182233142414e-05,
-      "loss": 0.257,
+      "epoch": 2.145457166540527,
+      "grad_norm": 0.1579708456993103,
+      "learning_rate": 3.1879176627010324e-05,
+      "loss": 0.3654,
       "step": 59530
     },
     {
-      "epoch": 2.09,
-      "learning_rate": 3.266547077770751e-05,
-      "loss": 0.2727,
+      "epoch": 2.1456373662017514,
+      "grad_norm": 0.23646564781665802,
+      "learning_rate": 3.187637109689267e-05,
+      "loss": 0.3908,
       "step": 59535
     },
     {
-      "epoch": 2.09,
-      "learning_rate": 3.266275922274151e-05,
-      "loss": 0.2915,
+      "epoch": 2.145817565862976,
+      "grad_norm": 0.2409384548664093,
+      "learning_rate": 3.187356547308578e-05,
+      "loss": 0.4183,
       "step": 59540
     },
     {
-      "epoch": 2.09,
-      "learning_rate": 3.266004756827963e-05,
-      "loss": 0.2943,
+      "epoch": 2.145997765524201,
+      "grad_norm": 0.19728174805641174,
+      "learning_rate": 3.18707597556279e-05,
+      "loss": 0.4135,
       "step": 59545
     },
     {
-      "epoch": 2.1,
-      "learning_rate": 3.26573358143571e-05,
-      "loss": 0.2857,
+      "epoch": 2.1461779651854256,
+      "grad_norm": 0.22944805026054382,
+      "learning_rate": 3.186795394455725e-05,
+      "loss": 0.4424,
       "step": 59550
     },
     {
-      "epoch": 2.1,
-      "learning_rate": 3.26546239610091e-05,
-      "loss": 0.2579,
+      "epoch": 2.1463581648466503,
+      "grad_norm": 0.21976561844348907,
+      "learning_rate": 3.186514803991205e-05,
+      "loss": 0.4277,
       "step": 59555
     },
     {
-      "epoch": 2.1,
-      "learning_rate": 3.265191200827085e-05,
-      "loss": 0.2391,
+      "epoch": 2.1465383645078746,
+      "grad_norm": 0.2405269593000412,
+      "learning_rate": 3.1862342041730545e-05,
+      "loss": 0.3874,
       "step": 59560
     },
     {
-      "epoch": 2.1,
-      "learning_rate": 3.2649199956177564e-05,
-      "loss": 0.2757,
+      "epoch": 2.1467185641690993,
+      "grad_norm": 0.18481574952602386,
+      "learning_rate": 3.185953595005095e-05,
+      "loss": 0.3858,
       "step": 59565
     },
     {
-      "epoch": 2.1,
-      "learning_rate": 3.2646487804764456e-05,
-      "loss": 0.2885,
+      "epoch": 2.146898763830324,
+      "grad_norm": 0.1834956705570221,
+      "learning_rate": 3.1856729764911506e-05,
+      "loss": 0.4078,
       "step": 59570
     },
     {
-      "epoch": 2.1,
-      "learning_rate": 3.2643775554066754e-05,
-      "loss": 0.2694,
+      "epoch": 2.147078963491549,
+      "grad_norm": 0.2032243311405182,
+      "learning_rate": 3.1853923486350455e-05,
+      "loss": 0.4089,
       "step": 59575
     },
     {
-      "epoch": 2.1,
-      "learning_rate": 3.264106320411966e-05,
-      "loss": 0.2738,
+      "epoch": 2.147259163152773,
+      "grad_norm": 0.19479459524154663,
+      "learning_rate": 3.185111711440601e-05,
+      "loss": 0.4115,
       "step": 59580
     },
     {
-      "epoch": 2.1,
-      "learning_rate": 3.263835075495839e-05,
-      "loss": 0.2896,
+      "epoch": 2.147439362813998,
+      "grad_norm": 0.20436975359916687,
+      "learning_rate": 3.184831064911644e-05,
+      "loss": 0.3796,
       "step": 59585
     },
     {
-      "epoch": 2.1,
-      "learning_rate": 3.263563820661817e-05,
-      "loss": 0.2797,
+      "epoch": 2.1476195624752226,
+      "grad_norm": 0.2070939838886261,
+      "learning_rate": 3.1845504090519954e-05,
+      "loss": 0.3942,
       "step": 59590
     },
     {
-      "epoch": 2.1,
-      "learning_rate": 3.263292555913422e-05,
-      "loss": 0.2776,
+      "epoch": 2.1477997621364473,
+      "grad_norm": 0.18167893588542938,
+      "learning_rate": 3.18426974386548e-05,
+      "loss": 0.4023,
       "step": 59595
     },
     {
-      "epoch": 2.1,
-      "learning_rate": 3.263021281254177e-05,
-      "loss": 0.272,
+      "epoch": 2.147979961797672,
+      "grad_norm": 0.19285474717617035,
+      "learning_rate": 3.1839890693559216e-05,
+      "loss": 0.4162,
       "step": 59600
     },
     {
-      "epoch": 2.1,
-      "learning_rate": 3.262749996687603e-05,
-      "loss": 0.2816,
+      "epoch": 2.1481601614588963,
+      "grad_norm": 0.23616044223308563,
+      "learning_rate": 3.183708385527144e-05,
+      "loss": 0.3866,
       "step": 59605
     },
     {
-      "epoch": 2.1,
-      "learning_rate": 3.262478702217223e-05,
-      "loss": 0.2527,
+      "epoch": 2.148340361120121,
+      "grad_norm": 0.2100156992673874,
+      "learning_rate": 3.1834276923829734e-05,
+      "loss": 0.4258,
       "step": 59610
     },
     {
-      "epoch": 2.1,
-      "learning_rate": 3.262207397846559e-05,
-      "loss": 0.2852,
+      "epoch": 2.148520560781346,
+      "grad_norm": 0.23628145456314087,
+      "learning_rate": 3.183146989927232e-05,
+      "loss": 0.4279,
       "step": 59615
     },
     {
-      "epoch": 2.1,
-      "learning_rate": 3.261936083579135e-05,
-      "loss": 0.2905,
+      "epoch": 2.1487007604425705,
+      "grad_norm": 0.17379525303840637,
+      "learning_rate": 3.1828662781637455e-05,
+      "loss": 0.4271,
       "step": 59620
     },
     {
-      "epoch": 2.1,
-      "learning_rate": 3.2616647594184726e-05,
-      "loss": 0.2726,
+      "epoch": 2.148880960103795,
+      "grad_norm": 0.20716167986392975,
+      "learning_rate": 3.182585557096337e-05,
+      "loss": 0.4243,
       "step": 59625
     },
     {
-      "epoch": 2.1,
-      "learning_rate": 3.2613934253680956e-05,
-      "loss": 0.2932,
+      "epoch": 2.1490611597650195,
+      "grad_norm": 0.19413475692272186,
+      "learning_rate": 3.1823048267288336e-05,
+      "loss": 0.3949,
       "step": 59630
     },
     {
-      "epoch": 2.1,
-      "learning_rate": 3.261122081431527e-05,
-      "loss": 0.2569,
+      "epoch": 2.1492413594262443,
+      "grad_norm": 0.2092110812664032,
+      "learning_rate": 3.182024087065059e-05,
+      "loss": 0.3402,
       "step": 59635
     },
     {
-      "epoch": 2.1,
-      "learning_rate": 3.260850727612289e-05,
-      "loss": 0.2814,
+      "epoch": 2.149421559087469,
+      "grad_norm": 0.18373681604862213,
+      "learning_rate": 3.1817433381088385e-05,
+      "loss": 0.3934,
       "step": 59640
     },
     {
-      "epoch": 2.1,
-      "learning_rate": 3.260633637443745e-05,
-      "loss": 0.2669,
+      "epoch": 2.1496017587486937,
+      "grad_norm": 0.2630181908607483,
+      "learning_rate": 3.181462579863996e-05,
+      "loss": 0.415,
       "step": 59645
     },
     {
-      "epoch": 2.1,
-      "learning_rate": 3.260362265844583e-05,
-      "loss": 0.2827,
+      "epoch": 2.149781958409918,
+      "grad_norm": 0.20736628770828247,
+      "learning_rate": 3.1811818123343584e-05,
+      "loss": 0.4196,
       "step": 59650
     },
     {
-      "epoch": 2.1,
-      "learning_rate": 3.2600908843726175e-05,
-      "loss": 0.2866,
+      "epoch": 2.1499621580711428,
+      "grad_norm": 0.2118118554353714,
+      "learning_rate": 3.180901035523751e-05,
+      "loss": 0.408,
       "step": 59655
     },
     {
-      "epoch": 2.1,
-      "learning_rate": 3.2598194930313734e-05,
-      "loss": 0.2778,
+      "epoch": 2.1501423577323675,
+      "grad_norm": 0.218112975358963,
+      "learning_rate": 3.180620249435998e-05,
+      "loss": 0.3967,
       "step": 59660
     },
     {
-      "epoch": 2.1,
-      "learning_rate": 3.2595480918243734e-05,
-      "loss": 0.2808,
+      "epoch": 2.1503225573935922,
+      "grad_norm": 0.19679492712020874,
+      "learning_rate": 3.180339454074926e-05,
+      "loss": 0.3891,
       "step": 59665
     },
     {
-      "epoch": 2.1,
-      "learning_rate": 3.259276680755143e-05,
-      "loss": 0.2708,
+      "epoch": 2.1505027570548165,
+      "grad_norm": 0.19406099617481232,
+      "learning_rate": 3.18005864944436e-05,
+      "loss": 0.406,
       "step": 59670
     },
     {
-      "epoch": 2.1,
-      "learning_rate": 3.259005259827205e-05,
-      "loss": 0.2914,
+      "epoch": 2.1506829567160413,
+      "grad_norm": 0.21432217955589294,
+      "learning_rate": 3.1797778355481285e-05,
+      "loss": 0.4209,
       "step": 59675
     },
     {
-      "epoch": 2.1,
-      "learning_rate": 3.2587338290440824e-05,
-      "loss": 0.2822,
+      "epoch": 2.150863156377266,
+      "grad_norm": 0.20044691860675812,
+      "learning_rate": 3.179497012390054e-05,
+      "loss": 0.3638,
       "step": 59680
     },
     {
-      "epoch": 2.1,
-      "learning_rate": 3.2584623884093027e-05,
-      "loss": 0.2654,
+      "epoch": 2.1510433560384907,
+      "grad_norm": 0.2646994888782501,
+      "learning_rate": 3.179216179973964e-05,
+      "loss": 0.4082,
       "step": 59685
     },
     {
-      "epoch": 2.1,
-      "learning_rate": 3.258190937926387e-05,
-      "loss": 0.2899,
+      "epoch": 2.1512235556997155,
+      "grad_norm": 0.1870126724243164,
+      "learning_rate": 3.178935338303686e-05,
+      "loss": 0.3992,
       "step": 59690
     },
     {
-      "epoch": 2.1,
-      "learning_rate": 3.257919477598863e-05,
-      "loss": 0.2628,
+      "epoch": 2.1514037553609398,
+      "grad_norm": 0.19518372416496277,
+      "learning_rate": 3.178654487383045e-05,
+      "loss": 0.359,
       "step": 59695
     },
     {
-      "epoch": 2.1,
-      "learning_rate": 3.2576480074302544e-05,
-      "loss": 0.2657,
+      "epoch": 2.1515839550221645,
+      "grad_norm": 0.18647782504558563,
+      "learning_rate": 3.178373627215869e-05,
+      "loss": 0.4009,
       "step": 59700
     },
     {
-      "epoch": 2.1,
-      "learning_rate": 3.2573765274240856e-05,
-      "loss": 0.2717,
+      "epoch": 2.151764154683389,
+      "grad_norm": 0.16734398901462555,
+      "learning_rate": 3.178092757805982e-05,
+      "loss": 0.4091,
       "step": 59705
     },
     {
-      "epoch": 2.1,
-      "learning_rate": 3.2571050375838806e-05,
-      "loss": 0.2832,
+      "epoch": 2.151944354344614,
+      "grad_norm": 0.1782519370317459,
+      "learning_rate": 3.177811879157214e-05,
+      "loss": 0.384,
       "step": 59710
     },
     {
-      "epoch": 2.1,
-      "learning_rate": 3.256833537913166e-05,
-      "loss": 0.2799,
+      "epoch": 2.1521245540058382,
+      "grad_norm": 0.16030040383338928,
+      "learning_rate": 3.1775309912733897e-05,
+      "loss": 0.3843,
       "step": 59715
     },
     {
-      "epoch": 2.1,
-      "learning_rate": 3.256562028415467e-05,
-      "loss": 0.2696,
+      "epoch": 2.152304753667063,
+      "grad_norm": 0.19871199131011963,
+      "learning_rate": 3.177250094158336e-05,
+      "loss": 0.4164,
       "step": 59720
     },
     {
-      "epoch": 2.1,
-      "learning_rate": 3.2562905090943095e-05,
-      "loss": 0.2867,
+      "epoch": 2.1524849533282877,
+      "grad_norm": 0.2312350869178772,
+      "learning_rate": 3.1769691878158823e-05,
+      "loss": 0.3661,
       "step": 59725
     },
     {
-      "epoch": 2.1,
-      "learning_rate": 3.256018979953217e-05,
-      "loss": 0.2534,
+      "epoch": 2.1526651529895124,
+      "grad_norm": 0.20747533440589905,
+      "learning_rate": 3.176688272249854e-05,
+      "loss": 0.3765,
       "step": 59730
     },
     {
-      "epoch": 2.1,
-      "learning_rate": 3.255747440995717e-05,
-      "loss": 0.2905,
+      "epoch": 2.152845352650737,
+      "grad_norm": 0.19539090991020203,
+      "learning_rate": 3.176407347464079e-05,
+      "loss": 0.4251,
       "step": 59735
     },
     {
-      "epoch": 2.1,
-      "learning_rate": 3.2554758922253336e-05,
-      "loss": 0.2662,
+      "epoch": 2.1530255523119615,
+      "grad_norm": 0.21648086607456207,
+      "learning_rate": 3.1761264134623844e-05,
+      "loss": 0.3954,
       "step": 59740
     },
     {
-      "epoch": 2.1,
-      "learning_rate": 3.255204333645595e-05,
-      "loss": 0.2624,
+      "epoch": 2.153205751973186,
+      "grad_norm": 0.21644406020641327,
+      "learning_rate": 3.175845470248599e-05,
+      "loss": 0.3859,
       "step": 59745
     },
     {
-      "epoch": 2.1,
-      "learning_rate": 3.2549327652600245e-05,
-      "loss": 0.284,
+      "epoch": 2.153385951634411,
+      "grad_norm": 0.19673486053943634,
+      "learning_rate": 3.175564517826549e-05,
+      "loss": 0.3904,
       "step": 59750
     },
     {
-      "epoch": 2.1,
-      "learning_rate": 3.2546611870721506e-05,
-      "loss": 0.2606,
+      "epoch": 2.1535661512956357,
+      "grad_norm": 0.22667108476161957,
+      "learning_rate": 3.1752835562000636e-05,
+      "loss": 0.4018,
       "step": 59755
     },
     {
-      "epoch": 2.1,
-      "learning_rate": 3.254389599085498e-05,
-      "loss": 0.2445,
+      "epoch": 2.1537463509568604,
+      "grad_norm": 0.22213202714920044,
+      "learning_rate": 3.175002585372971e-05,
+      "loss": 0.3882,
       "step": 59760
     },
     {
-      "epoch": 2.1,
-      "learning_rate": 3.254118001303593e-05,
-      "loss": 0.2853,
+      "epoch": 2.1539265506180847,
+      "grad_norm": 0.1795310080051422,
+      "learning_rate": 3.174721605349099e-05,
+      "loss": 0.3719,
       "step": 59765
     },
     {
-      "epoch": 2.1,
-      "learning_rate": 3.2538463937299635e-05,
-      "loss": 0.2689,
+      "epoch": 2.1541067502793094,
+      "grad_norm": 0.20347318053245544,
+      "learning_rate": 3.174440616132275e-05,
+      "loss": 0.3895,
       "step": 59770
     },
     {
-      "epoch": 2.1,
-      "learning_rate": 3.253574776368135e-05,
-      "loss": 0.2578,
+      "epoch": 2.154286949940534,
+      "grad_norm": 0.2064104527235031,
+      "learning_rate": 3.174159617726329e-05,
+      "loss": 0.3691,
       "step": 59775
     },
     {
-      "epoch": 2.1,
-      "learning_rate": 3.253303149221635e-05,
-      "loss": 0.2666,
+      "epoch": 2.154467149601759,
+      "grad_norm": 0.22160401940345764,
+      "learning_rate": 3.1738786101350884e-05,
+      "loss": 0.394,
       "step": 59780
     },
     {
-      "epoch": 2.1,
-      "learning_rate": 3.25303151229399e-05,
-      "loss": 0.2814,
+      "epoch": 2.154647349262983,
+      "grad_norm": 0.18739666044712067,
+      "learning_rate": 3.173597593362381e-05,
+      "loss": 0.393,
       "step": 59785
     },
     {
-      "epoch": 2.1,
-      "learning_rate": 3.252759865588726e-05,
-      "loss": 0.2801,
+      "epoch": 2.154827548924208,
+      "grad_norm": 0.21694833040237427,
+      "learning_rate": 3.173316567412038e-05,
+      "loss": 0.3736,
       "step": 59790
     },
     {
-      "epoch": 2.1,
-      "learning_rate": 3.252488209109373e-05,
-      "loss": 0.2776,
+      "epoch": 2.1550077485854326,
+      "grad_norm": 0.2537897229194641,
+      "learning_rate": 3.173035532287887e-05,
+      "loss": 0.3989,
       "step": 59795
     },
     {
-      "epoch": 2.1,
-      "learning_rate": 3.252216542859456e-05,
-      "loss": 0.2657,
+      "epoch": 2.1551879482466574,
+      "grad_norm": 0.16658885776996613,
+      "learning_rate": 3.172754487993757e-05,
+      "loss": 0.4302,
       "step": 59800
     },
     {
-      "epoch": 2.1,
-      "learning_rate": 3.251944866842502e-05,
-      "loss": 0.2865,
+      "epoch": 2.155368147907882,
+      "grad_norm": 0.18496111035346985,
+      "learning_rate": 3.1724734345334775e-05,
+      "loss": 0.3784,
       "step": 59805
     },
     {
-      "epoch": 2.1,
-      "learning_rate": 3.25167318106204e-05,
-      "loss": 0.2521,
+      "epoch": 2.1555483475691064,
+      "grad_norm": 0.19085146486759186,
+      "learning_rate": 3.1721923719108775e-05,
+      "loss": 0.3818,
       "step": 59810
     },
     {
-      "epoch": 2.1,
-      "learning_rate": 3.251401485521598e-05,
-      "loss": 0.2784,
+      "epoch": 2.155728547230331,
+      "grad_norm": 0.21254359185695648,
+      "learning_rate": 3.1719113001297866e-05,
+      "loss": 0.4093,
       "step": 59815
     },
     {
-      "epoch": 2.1,
-      "learning_rate": 3.251129780224702e-05,
-      "loss": 0.2967,
+      "epoch": 2.155908746891556,
+      "grad_norm": 0.16708926856517792,
+      "learning_rate": 3.171630219194035e-05,
+      "loss": 0.4064,
       "step": 59820
     },
     {
-      "epoch": 2.1,
-      "learning_rate": 3.2508580651748804e-05,
-      "loss": 0.2709,
+      "epoch": 2.1560889465527806,
+      "grad_norm": 0.20576465129852295,
+      "learning_rate": 3.171349129107451e-05,
+      "loss": 0.4028,
       "step": 59825
     },
     {
-      "epoch": 2.1,
-      "learning_rate": 3.250586340375663e-05,
-      "loss": 0.2599,
+      "epoch": 2.1562691462140053,
+      "grad_norm": 0.16969583928585052,
+      "learning_rate": 3.171068029873865e-05,
+      "loss": 0.3922,
       "step": 59830
     },
     {
-      "epoch": 2.11,
-      "learning_rate": 3.250314605830575e-05,
-      "loss": 0.2538,
+      "epoch": 2.1564493458752296,
+      "grad_norm": 0.1903369426727295,
+      "learning_rate": 3.170786921497107e-05,
+      "loss": 0.3691,
       "step": 59835
     },
     {
-      "epoch": 2.11,
-      "learning_rate": 3.2500428615431474e-05,
-      "loss": 0.2799,
+      "epoch": 2.1566295455364544,
+      "grad_norm": 0.1657458245754242,
+      "learning_rate": 3.1705058039810075e-05,
+      "loss": 0.385,
       "step": 59840
     },
     {
-      "epoch": 2.11,
-      "learning_rate": 3.249771107516907e-05,
-      "loss": 0.2776,
+      "epoch": 2.156809745197679,
+      "grad_norm": 0.17674793303012848,
+      "learning_rate": 3.170224677329396e-05,
+      "loss": 0.3697,
       "step": 59845
     },
     {
-      "epoch": 2.11,
-      "learning_rate": 3.2494993437553836e-05,
-      "loss": 0.2709,
+      "epoch": 2.156989944858904,
+      "grad_norm": 0.1935698390007019,
+      "learning_rate": 3.1699435415461034e-05,
+      "loss": 0.3688,
       "step": 59850
     },
     {
-      "epoch": 2.11,
-      "learning_rate": 3.2492275702621045e-05,
-      "loss": 0.247,
+      "epoch": 2.157170144520128,
+      "grad_norm": 0.22389110922813416,
+      "learning_rate": 3.1696623966349586e-05,
+      "loss": 0.4113,
       "step": 59855
     },
     {
-      "epoch": 2.11,
-      "learning_rate": 3.2489557870406e-05,
-      "loss": 0.2678,
+      "epoch": 2.157350344181353,
+      "grad_norm": 0.2064765840768814,
+      "learning_rate": 3.1693812425997946e-05,
+      "loss": 0.3933,
       "step": 59860
     },
     {
-      "epoch": 2.11,
-      "learning_rate": 3.2486839940943967e-05,
-      "loss": 0.2833,
+      "epoch": 2.1575305438425776,
+      "grad_norm": 0.17828290164470673,
+      "learning_rate": 3.1691000794444404e-05,
+      "loss": 0.4174,
       "step": 59865
     },
     {
-      "epoch": 2.11,
-      "learning_rate": 3.248412191427027e-05,
-      "loss": 0.3118,
+      "epoch": 2.1577107435038023,
+      "grad_norm": 0.18194176256656647,
+      "learning_rate": 3.168818907172727e-05,
+      "loss": 0.421,
       "step": 59870
     },
     {
-      "epoch": 2.11,
-      "learning_rate": 3.2481403790420164e-05,
-      "loss": 0.2482,
+      "epoch": 2.157890943165027,
+      "grad_norm": 0.1584172546863556,
+      "learning_rate": 3.168537725788485e-05,
+      "loss": 0.4051,
       "step": 59875
     },
     {
-      "epoch": 2.11,
-      "learning_rate": 3.2478685569428975e-05,
-      "loss": 0.2719,
+      "epoch": 2.1580711428262513,
+      "grad_norm": 0.21431034803390503,
+      "learning_rate": 3.168256535295547e-05,
+      "loss": 0.4027,
       "step": 59880
     },
     {
-      "epoch": 2.11,
-      "learning_rate": 3.247596725133197e-05,
-      "loss": 0.2679,
+      "epoch": 2.158251342487476,
+      "grad_norm": 0.18599197268486023,
+      "learning_rate": 3.167975335697743e-05,
+      "loss": 0.4038,
       "step": 59885
     },
     {
-      "epoch": 2.11,
-      "learning_rate": 3.247324883616447e-05,
-      "loss": 0.3003,
+      "epoch": 2.158431542148701,
+      "grad_norm": 0.18698103725910187,
+      "learning_rate": 3.167694126998903e-05,
+      "loss": 0.4269,
       "step": 59890
     },
     {
-      "epoch": 2.11,
-      "learning_rate": 3.247053032396175e-05,
-      "loss": 0.268,
+      "epoch": 2.1586117418099255,
+      "grad_norm": 0.2108622044324875,
+      "learning_rate": 3.1674129092028607e-05,
+      "loss": 0.4332,
       "step": 59895
     },
     {
-      "epoch": 2.11,
-      "learning_rate": 3.2467811714759114e-05,
-      "loss": 0.2698,
+      "epoch": 2.15879194147115,
+      "grad_norm": 0.22758710384368896,
+      "learning_rate": 3.167131682313447e-05,
+      "loss": 0.4095,
       "step": 59900
     },
     {
-      "epoch": 2.11,
-      "learning_rate": 3.246509300859187e-05,
-      "loss": 0.2986,
+      "epoch": 2.1589721411323746,
+      "grad_norm": 0.17931969463825226,
+      "learning_rate": 3.1668504463344926e-05,
+      "loss": 0.4192,
       "step": 59905
     },
     {
-      "epoch": 2.11,
-      "learning_rate": 3.2462374205495314e-05,
-      "loss": 0.255,
+      "epoch": 2.1591523407935993,
+      "grad_norm": 0.1740187406539917,
+      "learning_rate": 3.166569201269831e-05,
+      "loss": 0.4038,
       "step": 59910
     },
     {
-      "epoch": 2.11,
-      "learning_rate": 3.245965530550475e-05,
-      "loss": 0.2731,
+      "epoch": 2.159332540454824,
+      "grad_norm": 0.199792742729187,
+      "learning_rate": 3.166287947123292e-05,
+      "loss": 0.4077,
       "step": 59915
     },
     {
-      "epoch": 2.11,
-      "learning_rate": 3.2456936308655475e-05,
-      "loss": 0.2569,
+      "epoch": 2.1595127401160488,
+      "grad_norm": 0.22314900159835815,
+      "learning_rate": 3.1660066838987095e-05,
+      "loss": 0.4176,
       "step": 59920
     },
     {
-      "epoch": 2.11,
-      "learning_rate": 3.245421721498279e-05,
-      "loss": 0.2728,
+      "epoch": 2.159692939777273,
+      "grad_norm": 0.2095964401960373,
+      "learning_rate": 3.165725411599914e-05,
+      "loss": 0.391,
       "step": 59925
     },
     {
-      "epoch": 2.11,
-      "learning_rate": 3.245149802452201e-05,
-      "loss": 0.2508,
+      "epoch": 2.159873139438498,
+      "grad_norm": 0.2291499823331833,
+      "learning_rate": 3.16544413023074e-05,
+      "loss": 0.4169,
       "step": 59930
     },
     {
-      "epoch": 2.11,
-      "learning_rate": 3.244877873730844e-05,
-      "loss": 0.2521,
+      "epoch": 2.1600533390997225,
+      "grad_norm": 0.17338679730892181,
+      "learning_rate": 3.165162839795017e-05,
+      "loss": 0.3882,
       "step": 59935
     },
     {
-      "epoch": 2.11,
-      "learning_rate": 3.244605935337739e-05,
-      "loss": 0.297,
+      "epoch": 2.1602335387609473,
+      "grad_norm": 0.16860181093215942,
+      "learning_rate": 3.16488154029658e-05,
+      "loss": 0.3689,
       "step": 59940
     },
     {
-      "epoch": 2.11,
-      "learning_rate": 3.244333987276417e-05,
-      "loss": 0.2943,
+      "epoch": 2.1604137384221715,
+      "grad_norm": 0.2292938083410263,
+      "learning_rate": 3.164600231739261e-05,
+      "loss": 0.3986,
       "step": 59945
     },
     {
-      "epoch": 2.11,
-      "learning_rate": 3.2440620295504076e-05,
-      "loss": 0.2698,
+      "epoch": 2.1605939380833963,
+      "grad_norm": 0.20005054771900177,
+      "learning_rate": 3.164318914126891e-05,
+      "loss": 0.4175,
       "step": 59950
     },
     {
-      "epoch": 2.11,
-      "learning_rate": 3.2437900621632435e-05,
-      "loss": 0.2422,
+      "epoch": 2.160774137744621,
+      "grad_norm": 0.18958815932273865,
+      "learning_rate": 3.164037587463306e-05,
+      "loss": 0.3758,
       "step": 59955
     },
     {
-      "epoch": 2.11,
-      "learning_rate": 3.243518085118456e-05,
-      "loss": 0.2802,
+      "epoch": 2.1609543374058457,
+      "grad_norm": 0.218599334359169,
+      "learning_rate": 3.1637562517523374e-05,
+      "loss": 0.3645,
       "step": 59960
     },
     {
-      "epoch": 2.11,
-      "learning_rate": 3.243246098419576e-05,
-      "loss": 0.2635,
+      "epoch": 2.1611345370670705,
+      "grad_norm": 0.19358861446380615,
+      "learning_rate": 3.163474906997818e-05,
+      "loss": 0.3981,
       "step": 59965
     },
     {
-      "epoch": 2.11,
-      "learning_rate": 3.242974102070134e-05,
-      "loss": 0.2706,
+      "epoch": 2.1613147367282948,
+      "grad_norm": 0.21078740060329437,
+      "learning_rate": 3.1631935532035814e-05,
+      "loss": 0.3875,
       "step": 59970
     },
     {
-      "epoch": 2.11,
-      "learning_rate": 3.2427020960736645e-05,
-      "loss": 0.2721,
+      "epoch": 2.1614949363895195,
+      "grad_norm": 0.14373530447483063,
+      "learning_rate": 3.16291219037346e-05,
+      "loss": 0.4011,
       "step": 59975
     },
     {
-      "epoch": 2.11,
-      "learning_rate": 3.242430080433698e-05,
-      "loss": 0.2893,
+      "epoch": 2.1616751360507442,
+      "grad_norm": 0.19193744659423828,
+      "learning_rate": 3.16263081851129e-05,
+      "loss": 0.3839,
       "step": 59980
     },
     {
-      "epoch": 2.11,
-      "learning_rate": 3.242158055153764e-05,
-      "loss": 0.2645,
+      "epoch": 2.161855335711969,
+      "grad_norm": 0.24244289100170135,
+      "learning_rate": 3.162349437620903e-05,
+      "loss": 0.3978,
       "step": 59985
     },
     {
-      "epoch": 2.11,
-      "learning_rate": 3.241886020237399e-05,
-      "loss": 0.2629,
+      "epoch": 2.1620355353731937,
+      "grad_norm": 0.24157002568244934,
+      "learning_rate": 3.162068047706133e-05,
+      "loss": 0.3703,
       "step": 59990
     },
     {
-      "epoch": 2.11,
-      "learning_rate": 3.241613975688132e-05,
-      "loss": 0.2837,
+      "epoch": 2.162215735034418,
+      "grad_norm": 0.20742827653884888,
+      "learning_rate": 3.161786648770813e-05,
+      "loss": 0.375,
       "step": 59995
     },
     {
-      "epoch": 2.11,
-      "learning_rate": 3.2413419215094966e-05,
-      "loss": 0.2808,
+      "epoch": 2.1623959346956427,
+      "grad_norm": 0.1863773912191391,
+      "learning_rate": 3.1615052408187797e-05,
+      "loss": 0.3623,
       "step": 60000
     },
     {
-      "epoch": 2.11,
-      "eval_loss": 0.2701481282711029,
-      "eval_runtime": 10.5458,
-      "eval_samples_per_second": 9.482,
-      "eval_steps_per_second": 9.482,
+      "epoch": 2.1623959346956427,
+      "eval_loss": 0.4364284873008728,
+      "eval_runtime": 3.5244,
+      "eval_samples_per_second": 28.373,
+      "eval_steps_per_second": 7.093,
       "step": 60000
     },
     {
-      "epoch": 2.11,
-      "learning_rate": 3.241069857705025e-05,
-      "loss": 0.2705,
+      "epoch": 2.1625761343568675,
+      "grad_norm": 0.21003571152687073,
+      "learning_rate": 3.1612238238538646e-05,
+      "loss": 0.3717,
       "step": 60005
     },
     {
-      "epoch": 2.11,
-      "learning_rate": 3.240797784278249e-05,
-      "loss": 0.2748,
+      "epoch": 2.162756334018092,
+      "grad_norm": 0.1988876461982727,
+      "learning_rate": 3.160942397879903e-05,
+      "loss": 0.3693,
       "step": 60010
     },
     {
-      "epoch": 2.11,
-      "learning_rate": 3.240525701232703e-05,
-      "loss": 0.2734,
+      "epoch": 2.1629365336793165,
+      "grad_norm": 0.22915050387382507,
+      "learning_rate": 3.160660962900729e-05,
+      "loss": 0.3922,
       "step": 60015
     },
     {
-      "epoch": 2.11,
-      "learning_rate": 3.240253608571918e-05,
-      "loss": 0.2739,
+      "epoch": 2.163116733340541,
+      "grad_norm": 0.18933100998401642,
+      "learning_rate": 3.160379518920177e-05,
+      "loss": 0.4227,
       "step": 60020
     },
     {
-      "epoch": 2.11,
-      "learning_rate": 3.239981506299428e-05,
-      "loss": 0.2664,
+      "epoch": 2.163296933001766,
+      "grad_norm": 0.19733920693397522,
+      "learning_rate": 3.160098065942084e-05,
+      "loss": 0.3909,
       "step": 60025
     },
     {
-      "epoch": 2.11,
-      "learning_rate": 3.2397093944187656e-05,
-      "loss": 0.2603,
+      "epoch": 2.1634771326629907,
+      "grad_norm": 0.21198596060276031,
+      "learning_rate": 3.1598166039702805e-05,
+      "loss": 0.3822,
       "step": 60030
     },
     {
-      "epoch": 2.11,
-      "learning_rate": 3.239437272933464e-05,
-      "loss": 0.2468,
+      "epoch": 2.1636573323242154,
+      "grad_norm": 0.20547007024288177,
+      "learning_rate": 3.159535133008604e-05,
+      "loss": 0.3746,
       "step": 60035
     },
     {
-      "epoch": 2.11,
-      "learning_rate": 3.239165141847057e-05,
-      "loss": 0.286,
+      "epoch": 2.1638375319854397,
+      "grad_norm": 0.21266385912895203,
+      "learning_rate": 3.1592536530608894e-05,
+      "loss": 0.372,
       "step": 60040
     },
     {
-      "epoch": 2.11,
-      "learning_rate": 3.238893001163078e-05,
-      "loss": 0.2758,
+      "epoch": 2.1640177316466644,
+      "grad_norm": 0.21298128366470337,
+      "learning_rate": 3.158972164130971e-05,
+      "loss": 0.3811,
       "step": 60045
     },
     {
-      "epoch": 2.11,
-      "learning_rate": 3.2386208508850605e-05,
-      "loss": 0.288,
+      "epoch": 2.164197931307889,
+      "grad_norm": 0.17953209578990936,
+      "learning_rate": 3.158690666222685e-05,
+      "loss": 0.3941,
       "step": 60050
     },
     {
-      "epoch": 2.11,
-      "learning_rate": 3.238348691016537e-05,
-      "loss": 0.2894,
+      "epoch": 2.164378130969114,
+      "grad_norm": 0.15556363761425018,
+      "learning_rate": 3.158409159339866e-05,
+      "loss": 0.3773,
       "step": 60055
     },
     {
-      "epoch": 2.11,
-      "learning_rate": 3.238076521561044e-05,
-      "loss": 0.2633,
+      "epoch": 2.1645583306303386,
+      "grad_norm": 0.1946418732404709,
+      "learning_rate": 3.158127643486349e-05,
+      "loss": 0.3913,
       "step": 60060
     },
     {
-      "epoch": 2.11,
-      "learning_rate": 3.2378043425221116e-05,
-      "loss": 0.2684,
+      "epoch": 2.164738530291563,
+      "grad_norm": 0.23327438533306122,
+      "learning_rate": 3.157846118665971e-05,
+      "loss": 0.4158,
       "step": 60065
     },
     {
-      "epoch": 2.11,
-      "learning_rate": 3.237532153903278e-05,
-      "loss": 0.3006,
+      "epoch": 2.1649187299527877,
+      "grad_norm": 0.1919216811656952,
+      "learning_rate": 3.1575645848825675e-05,
+      "loss": 0.3949,
       "step": 60070
     },
     {
-      "epoch": 2.11,
-      "learning_rate": 3.237259955708074e-05,
-      "loss": 0.2576,
+      "epoch": 2.1650989296140124,
+      "grad_norm": 0.18845520913600922,
+      "learning_rate": 3.1572830421399724e-05,
+      "loss": 0.4004,
       "step": 60075
     },
     {
-      "epoch": 2.11,
-      "learning_rate": 3.236987747940037e-05,
-      "loss": 0.2727,
+      "epoch": 2.165279129275237,
+      "grad_norm": 0.19371269643306732,
+      "learning_rate": 3.1570014904420245e-05,
+      "loss": 0.4108,
       "step": 60080
     },
     {
-      "epoch": 2.11,
-      "learning_rate": 3.236715530602698e-05,
-      "loss": 0.2846,
+      "epoch": 2.1654593289364614,
+      "grad_norm": 0.1754084974527359,
+      "learning_rate": 3.156719929792557e-05,
+      "loss": 0.4292,
       "step": 60085
     },
     {
-      "epoch": 2.11,
-      "learning_rate": 3.236443303699595e-05,
-      "loss": 0.2586,
+      "epoch": 2.165639528597686,
+      "grad_norm": 0.2297254204750061,
+      "learning_rate": 3.156438360195409e-05,
+      "loss": 0.3867,
       "step": 60090
     },
     {
-      "epoch": 2.11,
-      "learning_rate": 3.236171067234259e-05,
-      "loss": 0.2887,
+      "epoch": 2.165819728258911,
+      "grad_norm": 0.18021254241466522,
+      "learning_rate": 3.156156781654415e-05,
+      "loss": 0.4032,
       "step": 60095
     },
     {
-      "epoch": 2.11,
-      "learning_rate": 3.235898821210229e-05,
-      "loss": 0.2896,
+      "epoch": 2.1659999279201356,
+      "grad_norm": 0.20905311405658722,
+      "learning_rate": 3.155875194173411e-05,
+      "loss": 0.4037,
       "step": 60100
     },
     {
-      "epoch": 2.11,
-      "learning_rate": 3.2356265656310365e-05,
-      "loss": 0.2694,
+      "epoch": 2.1661801275813604,
+      "grad_norm": 0.1627390831708908,
+      "learning_rate": 3.1555935977562354e-05,
+      "loss": 0.3906,
       "step": 60105
     },
     {
-      "epoch": 2.11,
-      "learning_rate": 3.235354300500218e-05,
-      "loss": 0.25,
+      "epoch": 2.1663603272425846,
+      "grad_norm": 0.2780887484550476,
+      "learning_rate": 3.155311992406724e-05,
+      "loss": 0.4073,
       "step": 60110
     },
     {
-      "epoch": 2.12,
-      "learning_rate": 3.235082025821309e-05,
-      "loss": 0.2701,
+      "epoch": 2.1665405269038094,
+      "grad_norm": 0.17551511526107788,
+      "learning_rate": 3.155030378128713e-05,
+      "loss": 0.4206,
       "step": 60115
     },
     {
-      "epoch": 2.12,
-      "learning_rate": 3.234809741597844e-05,
-      "loss": 0.293,
+      "epoch": 2.166720726565034,
+      "grad_norm": 0.16955679655075073,
+      "learning_rate": 3.15474875492604e-05,
+      "loss": 0.402,
       "step": 60120
     },
     {
-      "epoch": 2.12,
-      "learning_rate": 3.234537447833358e-05,
-      "loss": 0.2771,
+      "epoch": 2.166900926226259,
+      "grad_norm": 0.16937977075576782,
+      "learning_rate": 3.154467122802543e-05,
+      "loss": 0.3718,
       "step": 60125
     },
     {
-      "epoch": 2.12,
-      "learning_rate": 3.234265144531389e-05,
-      "loss": 0.282,
+      "epoch": 2.167081125887483,
+      "grad_norm": 0.18345296382904053,
+      "learning_rate": 3.154185481762057e-05,
+      "loss": 0.4043,
       "step": 60130
     },
     {
-      "epoch": 2.12,
-      "learning_rate": 3.2339928316954696e-05,
-      "loss": 0.2647,
+      "epoch": 2.167261325548708,
+      "grad_norm": 0.19128628075122833,
+      "learning_rate": 3.153903831808421e-05,
+      "loss": 0.4128,
       "step": 60135
     },
     {
-      "epoch": 2.12,
-      "learning_rate": 3.2337205093291375e-05,
-      "loss": 0.282,
+      "epoch": 2.1674415252099326,
+      "grad_norm": 0.18635614216327667,
+      "learning_rate": 3.153622172945472e-05,
+      "loss": 0.3931,
       "step": 60140
     },
     {
-      "epoch": 2.12,
-      "learning_rate": 3.233448177435928e-05,
-      "loss": 0.2892,
+      "epoch": 2.1676217248711573,
+      "grad_norm": 0.22392097115516663,
+      "learning_rate": 3.153340505177047e-05,
+      "loss": 0.3779,
       "step": 60145
     },
     {
-      "epoch": 2.12,
-      "learning_rate": 3.2331758360193774e-05,
-      "loss": 0.2863,
+      "epoch": 2.167801924532382,
+      "grad_norm": 0.18912804126739502,
+      "learning_rate": 3.153058828506984e-05,
+      "loss": 0.387,
       "step": 60150
     },
     {
-      "epoch": 2.12,
-      "learning_rate": 3.232903485083022e-05,
-      "loss": 0.2757,
+      "epoch": 2.1679821241936064,
+      "grad_norm": 0.17587023973464966,
+      "learning_rate": 3.152777142939122e-05,
+      "loss": 0.4231,
       "step": 60155
     },
     {
-      "epoch": 2.12,
-      "learning_rate": 3.2326311246303967e-05,
-      "loss": 0.2936,
+      "epoch": 2.168162323854831,
+      "grad_norm": 0.19323548674583435,
+      "learning_rate": 3.152495448477296e-05,
+      "loss": 0.3857,
       "step": 60160
     },
     {
-      "epoch": 2.12,
-      "learning_rate": 3.2323587546650406e-05,
-      "loss": 0.254,
+      "epoch": 2.168342523516056,
+      "grad_norm": 0.2253098487854004,
+      "learning_rate": 3.152213745125348e-05,
+      "loss": 0.3647,
       "step": 60165
     },
     {
-      "epoch": 2.12,
-      "learning_rate": 3.2320863751904875e-05,
-      "loss": 0.2743,
+      "epoch": 2.1685227231772806,
+      "grad_norm": 0.26896703243255615,
+      "learning_rate": 3.151932032887112e-05,
+      "loss": 0.4163,
       "step": 60170
     },
     {
-      "epoch": 2.12,
-      "learning_rate": 3.231813986210276e-05,
-      "loss": 0.2725,
+      "epoch": 2.168702922838505,
+      "grad_norm": 0.16771872341632843,
+      "learning_rate": 3.151650311766429e-05,
+      "loss": 0.3935,
       "step": 60175
     },
     {
-      "epoch": 2.12,
-      "learning_rate": 3.231541587727942e-05,
-      "loss": 0.2518,
+      "epoch": 2.1688831224997296,
+      "grad_norm": 0.1727340817451477,
+      "learning_rate": 3.1513685817671365e-05,
+      "loss": 0.3809,
       "step": 60180
     },
     {
-      "epoch": 2.12,
-      "learning_rate": 3.231269179747023e-05,
-      "loss": 0.2639,
+      "epoch": 2.1690633221609543,
+      "grad_norm": 0.17910057306289673,
+      "learning_rate": 3.151086842893074e-05,
+      "loss": 0.4137,
       "step": 60185
     },
     {
-      "epoch": 2.12,
-      "learning_rate": 3.230996762271055e-05,
-      "loss": 0.2875,
+      "epoch": 2.169243521822179,
+      "grad_norm": 0.21039491891860962,
+      "learning_rate": 3.150805095148079e-05,
+      "loss": 0.3931,
       "step": 60190
     },
     {
-      "epoch": 2.12,
-      "learning_rate": 3.230724335303576e-05,
-      "loss": 0.2559,
+      "epoch": 2.169423721483404,
+      "grad_norm": 0.15780936181545258,
+      "learning_rate": 3.15052333853599e-05,
+      "loss": 0.3762,
       "step": 60195
     },
     {
-      "epoch": 2.12,
-      "learning_rate": 3.2304518988481234e-05,
-      "loss": 0.2703,
+      "epoch": 2.169603921144628,
+      "grad_norm": 0.2158997654914856,
+      "learning_rate": 3.150241573060647e-05,
+      "loss": 0.413,
       "step": 60200
     },
     {
-      "epoch": 2.12,
-      "learning_rate": 3.230179452908234e-05,
-      "loss": 0.272,
+      "epoch": 2.169784120805853,
+      "grad_norm": 0.22214800119400024,
+      "learning_rate": 3.1499597987258876e-05,
+      "loss": 0.406,
       "step": 60205
     },
     {
-      "epoch": 2.12,
-      "learning_rate": 3.229906997487445e-05,
-      "loss": 0.2638,
+      "epoch": 2.1699643204670775,
+      "grad_norm": 0.21257710456848145,
+      "learning_rate": 3.149678015535553e-05,
+      "loss": 0.4273,
       "step": 60210
     },
     {
-      "epoch": 2.12,
-      "learning_rate": 3.229634532589296e-05,
-      "loss": 0.2637,
+      "epoch": 2.1701445201283023,
+      "grad_norm": 0.18004845082759857,
+      "learning_rate": 3.14939622349348e-05,
+      "loss": 0.3658,
       "step": 60215
     },
     {
-      "epoch": 2.12,
-      "learning_rate": 3.2293620582173226e-05,
-      "loss": 0.2767,
+      "epoch": 2.1703247197895266,
+      "grad_norm": 0.20979435741901398,
+      "learning_rate": 3.149114422603511e-05,
+      "loss": 0.397,
       "step": 60220
     },
     {
-      "epoch": 2.12,
-      "learning_rate": 3.2290895743750634e-05,
-      "loss": 0.2669,
+      "epoch": 2.1705049194507513,
+      "grad_norm": 0.20553478598594666,
+      "learning_rate": 3.148832612869482e-05,
+      "loss": 0.3749,
       "step": 60225
     },
     {
-      "epoch": 2.12,
-      "learning_rate": 3.2288170810660565e-05,
-      "loss": 0.2701,
+      "epoch": 2.170685119111976,
+      "grad_norm": 0.15849637985229492,
+      "learning_rate": 3.148550794295235e-05,
+      "loss": 0.3821,
       "step": 60230
     },
     {
-      "epoch": 2.12,
-      "learning_rate": 3.228544578293841e-05,
-      "loss": 0.2598,
+      "epoch": 2.1708653187732008,
+      "grad_norm": 0.19608749449253082,
+      "learning_rate": 3.148268966884609e-05,
+      "loss": 0.3915,
       "step": 60235
     },
     {
-      "epoch": 2.12,
-      "learning_rate": 3.228272066061953e-05,
-      "loss": 0.2803,
+      "epoch": 2.1710455184344255,
+      "grad_norm": 0.16606873273849487,
+      "learning_rate": 3.147987130641443e-05,
+      "loss": 0.3923,
       "step": 60240
     },
     {
-      "epoch": 2.12,
-      "learning_rate": 3.227999544373933e-05,
-      "loss": 0.2956,
+      "epoch": 2.17122571809565,
+      "grad_norm": 0.17005008459091187,
+      "learning_rate": 3.147705285569579e-05,
+      "loss": 0.417,
       "step": 60245
     },
     {
-      "epoch": 2.12,
-      "learning_rate": 3.2277270132333185e-05,
-      "loss": 0.249,
+      "epoch": 2.1714059177568745,
+      "grad_norm": 0.1739397495985031,
+      "learning_rate": 3.1474234316728554e-05,
+      "loss": 0.4078,
       "step": 60250
     },
     {
-      "epoch": 2.12,
-      "learning_rate": 3.227454472643649e-05,
-      "loss": 0.259,
+      "epoch": 2.1715861174180993,
+      "grad_norm": 0.21615131199359894,
+      "learning_rate": 3.1471415689551124e-05,
+      "loss": 0.3828,
       "step": 60255
     },
     {
-      "epoch": 2.12,
-      "learning_rate": 3.227181922608462e-05,
-      "loss": 0.2769,
+      "epoch": 2.171766317079324,
+      "grad_norm": 0.18913507461547852,
+      "learning_rate": 3.1468596974201915e-05,
+      "loss": 0.4276,
       "step": 60260
     },
     {
-      "epoch": 2.12,
-      "learning_rate": 3.226909363131296e-05,
-      "loss": 0.2997,
+      "epoch": 2.1719465167405487,
+      "grad_norm": 0.17284362018108368,
+      "learning_rate": 3.1465778170719314e-05,
+      "loss": 0.4342,
       "step": 60265
     },
     {
-      "epoch": 2.12,
-      "learning_rate": 3.226636794215692e-05,
-      "loss": 0.2831,
+      "epoch": 2.172126716401773,
+      "grad_norm": 0.21640555560588837,
+      "learning_rate": 3.146295927914175e-05,
+      "loss": 0.4252,
       "step": 60270
     },
     {
-      "epoch": 2.12,
-      "learning_rate": 3.226364215865188e-05,
-      "loss": 0.2585,
+      "epoch": 2.1723069160629978,
+      "grad_norm": 0.1958705186843872,
+      "learning_rate": 3.1460140299507614e-05,
+      "loss": 0.4443,
       "step": 60275
     },
     {
-      "epoch": 2.12,
-      "learning_rate": 3.2260916280833244e-05,
-      "loss": 0.2614,
+      "epoch": 2.1724871157242225,
+      "grad_norm": 0.18334457278251648,
+      "learning_rate": 3.145732123185531e-05,
+      "loss": 0.3744,
       "step": 60280
     },
     {
-      "epoch": 2.12,
-      "learning_rate": 3.2258190308736375e-05,
-      "loss": 0.2785,
+      "epoch": 2.172667315385447,
+      "grad_norm": 0.2335662692785263,
+      "learning_rate": 3.1454502076223255e-05,
+      "loss": 0.4132,
       "step": 60285
     },
     {
-      "epoch": 2.12,
-      "learning_rate": 3.22554642423967e-05,
-      "loss": 0.2941,
+      "epoch": 2.1728475150466715,
+      "grad_norm": 0.1668962836265564,
+      "learning_rate": 3.145168283264987e-05,
+      "loss": 0.3935,
       "step": 60290
     },
     {
-      "epoch": 2.12,
-      "learning_rate": 3.22527380818496e-05,
-      "loss": 0.2816,
+      "epoch": 2.1730277147078962,
+      "grad_norm": 0.2275708168745041,
+      "learning_rate": 3.144886350117355e-05,
+      "loss": 0.3886,
       "step": 60295
     },
     {
-      "epoch": 2.12,
-      "learning_rate": 3.225001182713048e-05,
-      "loss": 0.2667,
+      "epoch": 2.173207914369121,
+      "grad_norm": 0.21035301685333252,
+      "learning_rate": 3.144604408183271e-05,
+      "loss": 0.4101,
       "step": 60300
     },
     {
-      "epoch": 2.12,
-      "learning_rate": 3.224728547827473e-05,
-      "loss": 0.2592,
+      "epoch": 2.1733881140303457,
+      "grad_norm": 0.20323359966278076,
+      "learning_rate": 3.144322457466577e-05,
+      "loss": 0.3783,
       "step": 60305
     },
     {
-      "epoch": 2.12,
-      "learning_rate": 3.224455903531775e-05,
-      "loss": 0.2858,
+      "epoch": 2.1735683136915704,
+      "grad_norm": 0.21606206893920898,
+      "learning_rate": 3.144040497971115e-05,
+      "loss": 0.4244,
       "step": 60310
     },
     {
-      "epoch": 2.12,
-      "learning_rate": 3.224183249829495e-05,
-      "loss": 0.2857,
+      "epoch": 2.1737485133527947,
+      "grad_norm": 0.1799047589302063,
+      "learning_rate": 3.143758529700724e-05,
+      "loss": 0.3813,
       "step": 60315
     },
     {
-      "epoch": 2.12,
-      "learning_rate": 3.223910586724173e-05,
-      "loss": 0.268,
+      "epoch": 2.1739287130140195,
+      "grad_norm": 0.17974518239498138,
+      "learning_rate": 3.143476552659249e-05,
+      "loss": 0.4338,
       "step": 60320
     },
     {
-      "epoch": 2.12,
-      "learning_rate": 3.223637914219349e-05,
-      "loss": 0.275,
+      "epoch": 2.174108912675244,
+      "grad_norm": 0.17822134494781494,
+      "learning_rate": 3.14319456685053e-05,
+      "loss": 0.3918,
       "step": 60325
     },
     {
-      "epoch": 2.12,
-      "learning_rate": 3.2233652323185636e-05,
-      "loss": 0.2762,
+      "epoch": 2.174289112336469,
+      "grad_norm": 0.21287919580936432,
+      "learning_rate": 3.1429125722784105e-05,
+      "loss": 0.4043,
       "step": 60330
     },
     {
-      "epoch": 2.12,
-      "learning_rate": 3.2230925410253576e-05,
-      "loss": 0.2881,
+      "epoch": 2.1744693119976937,
+      "grad_norm": 0.2515278160572052,
+      "learning_rate": 3.1426305689467304e-05,
+      "loss": 0.3708,
       "step": 60335
     },
     {
-      "epoch": 2.12,
-      "learning_rate": 3.2228198403432705e-05,
-      "loss": 0.2852,
+      "epoch": 2.174649511658918,
+      "grad_norm": 0.20817367732524872,
+      "learning_rate": 3.142348556859335e-05,
+      "loss": 0.4246,
       "step": 60340
     },
     {
-      "epoch": 2.12,
-      "learning_rate": 3.222547130275845e-05,
-      "loss": 0.2704,
+      "epoch": 2.1748297113201427,
+      "grad_norm": 0.20883257687091827,
+      "learning_rate": 3.142066536020063e-05,
+      "loss": 0.4121,
       "step": 60345
     },
     {
-      "epoch": 2.12,
-      "learning_rate": 3.222274410826621e-05,
-      "loss": 0.2745,
+      "epoch": 2.1750099109813674,
+      "grad_norm": 0.2272225171327591,
+      "learning_rate": 3.14178450643276e-05,
+      "loss": 0.4404,
       "step": 60350
     },
     {
-      "epoch": 2.12,
-      "learning_rate": 3.22200168199914e-05,
-      "loss": 0.2804,
+      "epoch": 2.175190110642592,
+      "grad_norm": 0.20350000262260437,
+      "learning_rate": 3.141502468101267e-05,
+      "loss": 0.4316,
       "step": 60355
     },
     {
-      "epoch": 2.12,
-      "learning_rate": 3.2217289437969435e-05,
-      "loss": 0.2782,
+      "epoch": 2.1753703103038164,
+      "grad_norm": 0.16587962210178375,
+      "learning_rate": 3.141220421029427e-05,
+      "loss": 0.367,
       "step": 60360
     },
     {
-      "epoch": 2.12,
-      "learning_rate": 3.22145619622357e-05,
-      "loss": 0.2681,
+      "epoch": 2.175550509965041,
+      "grad_norm": 0.16492369771003723,
+      "learning_rate": 3.140938365221082e-05,
+      "loss": 0.417,
       "step": 60365
     },
     {
-      "epoch": 2.12,
-      "learning_rate": 3.221183439282566e-05,
-      "loss": 0.2567,
+      "epoch": 2.175730709626266,
+      "grad_norm": 0.1600433737039566,
+      "learning_rate": 3.140656300680077e-05,
+      "loss": 0.3949,
       "step": 60370
     },
     {
-      "epoch": 2.12,
-      "learning_rate": 3.2209106729774686e-05,
-      "loss": 0.2655,
+      "epoch": 2.1759109092874906,
+      "grad_norm": 0.2284023016691208,
+      "learning_rate": 3.140374227410254e-05,
+      "loss": 0.433,
       "step": 60375
     },
     {
-      "epoch": 2.12,
-      "learning_rate": 3.220637897311822e-05,
-      "loss": 0.2645,
+      "epoch": 2.1760911089487154,
+      "grad_norm": 0.267976313829422,
+      "learning_rate": 3.1400921454154553e-05,
+      "loss": 0.3941,
       "step": 60380
     },
     {
-      "epoch": 2.12,
-      "learning_rate": 3.2203651122891664e-05,
-      "loss": 0.2739,
+      "epoch": 2.1762713086099397,
+      "grad_norm": 0.19785836338996887,
+      "learning_rate": 3.1398100546995256e-05,
+      "loss": 0.4047,
       "step": 60385
     },
     {
-      "epoch": 2.12,
-      "learning_rate": 3.2200923179130455e-05,
-      "loss": 0.282,
+      "epoch": 2.1764515082711644,
+      "grad_norm": 0.19465455412864685,
+      "learning_rate": 3.1395279552663075e-05,
+      "loss": 0.4211,
       "step": 60390
     },
     {
-      "epoch": 2.12,
-      "learning_rate": 3.2198195141869995e-05,
-      "loss": 0.2717,
+      "epoch": 2.176631707932389,
+      "grad_norm": 0.22500787675380707,
+      "learning_rate": 3.139245847119646e-05,
+      "loss": 0.4166,
       "step": 60395
     },
     {
-      "epoch": 2.13,
-      "learning_rate": 3.219546701114572e-05,
-      "loss": 0.251,
+      "epoch": 2.176811907593614,
+      "grad_norm": 0.19240856170654297,
+      "learning_rate": 3.1389637302633816e-05,
+      "loss": 0.3926,
       "step": 60400
     },
     {
-      "epoch": 2.13,
-      "learning_rate": 3.219273878699305e-05,
-      "loss": 0.2748,
+      "epoch": 2.176992107254838,
+      "grad_norm": 0.22199541330337524,
+      "learning_rate": 3.1386816047013615e-05,
+      "loss": 0.3848,
       "step": 60405
     },
     {
-      "epoch": 2.13,
-      "learning_rate": 3.219001046944741e-05,
-      "loss": 0.2541,
+      "epoch": 2.177172306916063,
+      "grad_norm": 0.21987438201904297,
+      "learning_rate": 3.1383994704374276e-05,
+      "loss": 0.3841,
       "step": 60410
     },
     {
-      "epoch": 2.13,
-      "learning_rate": 3.2187282058544214e-05,
-      "loss": 0.2729,
+      "epoch": 2.1773525065772876,
+      "grad_norm": 0.21120600402355194,
+      "learning_rate": 3.138117327475425e-05,
+      "loss": 0.3895,
       "step": 60415
     },
     {
-      "epoch": 2.13,
-      "learning_rate": 3.2184553554318894e-05,
-      "loss": 0.2877,
+      "epoch": 2.1775327062385124,
+      "grad_norm": 0.14966708421707153,
+      "learning_rate": 3.137835175819197e-05,
+      "loss": 0.3998,
       "step": 60420
     },
     {
-      "epoch": 2.13,
-      "learning_rate": 3.2181824956806896e-05,
-      "loss": 0.2689,
+      "epoch": 2.177712905899737,
+      "grad_norm": 0.18459533154964447,
+      "learning_rate": 3.137553015472588e-05,
+      "loss": 0.3855,
       "step": 60425
     },
     {
-      "epoch": 2.13,
-      "learning_rate": 3.217909626604362e-05,
-      "loss": 0.2677,
+      "epoch": 2.1778931055609614,
+      "grad_norm": 0.18099567294120789,
+      "learning_rate": 3.1372708464394427e-05,
+      "loss": 0.3645,
       "step": 60430
     },
     {
-      "epoch": 2.13,
-      "learning_rate": 3.217636748206452e-05,
-      "loss": 0.2678,
+      "epoch": 2.178073305222186,
+      "grad_norm": 0.20282356441020966,
+      "learning_rate": 3.136988668723606e-05,
+      "loss": 0.4062,
       "step": 60435
     },
     {
-      "epoch": 2.13,
-      "learning_rate": 3.217363860490501e-05,
-      "loss": 0.2515,
+      "epoch": 2.178253504883411,
+      "grad_norm": 0.22229717671871185,
+      "learning_rate": 3.136706482328922e-05,
+      "loss": 0.3824,
       "step": 60440
     },
     {
-      "epoch": 2.13,
-      "learning_rate": 3.217090963460053e-05,
-      "loss": 0.2667,
+      "epoch": 2.1784337045446356,
+      "grad_norm": 0.22084426879882812,
+      "learning_rate": 3.1364242872592345e-05,
+      "loss": 0.3796,
       "step": 60445
     },
     {
-      "epoch": 2.13,
-      "learning_rate": 3.216818057118652e-05,
-      "loss": 0.2902,
+      "epoch": 2.17861390420586,
+      "grad_norm": 0.24100591242313385,
+      "learning_rate": 3.1361420835183894e-05,
+      "loss": 0.4144,
       "step": 60450
     },
     {
-      "epoch": 2.13,
-      "learning_rate": 3.216545141469841e-05,
-      "loss": 0.2829,
+      "epoch": 2.1787941038670846,
+      "grad_norm": 0.18646858632564545,
+      "learning_rate": 3.1358598711102336e-05,
+      "loss": 0.3665,
       "step": 60455
     },
     {
-      "epoch": 2.13,
-      "learning_rate": 3.2162722165171624e-05,
-      "loss": 0.2775,
+      "epoch": 2.1789743035283093,
+      "grad_norm": 0.22386908531188965,
+      "learning_rate": 3.135577650038608e-05,
+      "loss": 0.3888,
       "step": 60460
     },
     {
-      "epoch": 2.13,
-      "learning_rate": 3.2159992822641626e-05,
-      "loss": 0.2978,
+      "epoch": 2.179154503189534,
+      "grad_norm": 0.1991269886493683,
+      "learning_rate": 3.135295420307361e-05,
+      "loss": 0.4148,
       "step": 60465
     },
     {
-      "epoch": 2.13,
-      "learning_rate": 3.215726338714383e-05,
-      "loss": 0.2594,
+      "epoch": 2.179334702850759,
+      "grad_norm": 0.18925313651561737,
+      "learning_rate": 3.135013181920336e-05,
+      "loss": 0.3654,
       "step": 60470
     },
     {
-      "epoch": 2.13,
-      "learning_rate": 3.215453385871369e-05,
-      "loss": 0.282,
+      "epoch": 2.179514902511983,
+      "grad_norm": 0.2189246565103531,
+      "learning_rate": 3.13473093488138e-05,
+      "loss": 0.3729,
       "step": 60475
     },
     {
-      "epoch": 2.13,
-      "learning_rate": 3.215180423738664e-05,
-      "loss": 0.2855,
+      "epoch": 2.179695102173208,
+      "grad_norm": 0.17129361629486084,
+      "learning_rate": 3.134448679194338e-05,
+      "loss": 0.4165,
       "step": 60480
     },
     {
-      "epoch": 2.13,
-      "learning_rate": 3.214907452319813e-05,
-      "loss": 0.268,
+      "epoch": 2.1798753018344326,
+      "grad_norm": 0.20182783901691437,
+      "learning_rate": 3.134166414863055e-05,
+      "loss": 0.3914,
       "step": 60485
     },
     {
-      "epoch": 2.13,
-      "learning_rate": 3.21463447161836e-05,
-      "loss": 0.2403,
+      "epoch": 2.1800555014956573,
+      "grad_norm": 0.18997080624103546,
+      "learning_rate": 3.1338841418913776e-05,
+      "loss": 0.3937,
       "step": 60490
     },
     {
-      "epoch": 2.13,
-      "learning_rate": 3.2143614816378486e-05,
-      "loss": 0.2631,
+      "epoch": 2.1802357011568816,
+      "grad_norm": 0.1698470115661621,
+      "learning_rate": 3.133601860283152e-05,
+      "loss": 0.3883,
       "step": 60495
     },
     {
-      "epoch": 2.13,
-      "learning_rate": 3.2140884823818246e-05,
-      "loss": 0.2759,
+      "epoch": 2.1804159008181063,
+      "grad_norm": 0.2348821759223938,
+      "learning_rate": 3.133319570042224e-05,
+      "loss": 0.4298,
       "step": 60500
     },
     {
-      "epoch": 2.13,
-      "eval_loss": 0.26962408423423767,
-      "eval_runtime": 10.5459,
-      "eval_samples_per_second": 9.482,
-      "eval_steps_per_second": 9.482,
+      "epoch": 2.1804159008181063,
+      "eval_loss": 0.4370137155056,
+      "eval_runtime": 3.5451,
+      "eval_samples_per_second": 28.208,
+      "eval_steps_per_second": 7.052,
       "step": 60500
     },
     {
-      "epoch": 2.13,
-      "learning_rate": 3.213815473853832e-05,
-      "loss": 0.2721,
+      "epoch": 2.180596100479331,
+      "grad_norm": 0.16450156271457672,
+      "learning_rate": 3.1330372711724385e-05,
+      "loss": 0.3918,
       "step": 60505
     },
     {
-      "epoch": 2.13,
-      "learning_rate": 3.213542456057416e-05,
-      "loss": 0.2745,
+      "epoch": 2.180776300140556,
+      "grad_norm": 0.23515671491622925,
+      "learning_rate": 3.132754963677643e-05,
+      "loss": 0.3826,
       "step": 60510
     },
     {
-      "epoch": 2.13,
-      "learning_rate": 3.213269428996121e-05,
-      "loss": 0.2793,
+      "epoch": 2.1809564998017805,
+      "grad_norm": 0.1602397859096527,
+      "learning_rate": 3.132472647561684e-05,
+      "loss": 0.3654,
       "step": 60515
     },
     {
-      "epoch": 2.13,
-      "learning_rate": 3.212996392673494e-05,
-      "loss": 0.2722,
+      "epoch": 2.181136699463005,
+      "grad_norm": 0.15597917139530182,
+      "learning_rate": 3.1321903228284064e-05,
+      "loss": 0.3848,
       "step": 60520
     },
     {
-      "epoch": 2.13,
-      "learning_rate": 3.212723347093077e-05,
-      "loss": 0.2784,
+      "epoch": 2.1813168991242295,
+      "grad_norm": 0.1703244298696518,
+      "learning_rate": 3.1319079894816595e-05,
+      "loss": 0.4033,
       "step": 60525
     },
     {
-      "epoch": 2.13,
-      "learning_rate": 3.212450292258418e-05,
-      "loss": 0.2722,
+      "epoch": 2.1814970987854543,
+      "grad_norm": 0.19135898351669312,
+      "learning_rate": 3.131625647525288e-05,
+      "loss": 0.389,
       "step": 60530
     },
     {
-      "epoch": 2.13,
-      "learning_rate": 3.2121772281730603e-05,
-      "loss": 0.2757,
+      "epoch": 2.181677298446679,
+      "grad_norm": 0.22922751307487488,
+      "learning_rate": 3.1313432969631397e-05,
+      "loss": 0.422,
       "step": 60535
     },
     {
-      "epoch": 2.13,
-      "learning_rate": 3.2119041548405526e-05,
-      "loss": 0.2651,
+      "epoch": 2.1818574981079037,
+      "grad_norm": 0.24647971987724304,
+      "learning_rate": 3.1310609377990603e-05,
+      "loss": 0.4223,
       "step": 60540
     },
     {
-      "epoch": 2.13,
-      "learning_rate": 3.2116310722644374e-05,
-      "loss": 0.2571,
+      "epoch": 2.182037697769128,
+      "grad_norm": 0.2550751268863678,
+      "learning_rate": 3.130778570036898e-05,
+      "loss": 0.4325,
       "step": 60545
     },
     {
-      "epoch": 2.13,
-      "learning_rate": 3.211357980448262e-05,
-      "loss": 0.2568,
+      "epoch": 2.1822178974303528,
+      "grad_norm": 0.20791274309158325,
+      "learning_rate": 3.130496193680501e-05,
+      "loss": 0.4028,
       "step": 60550
     },
     {
-      "epoch": 2.13,
-      "learning_rate": 3.211084879395572e-05,
-      "loss": 0.2969,
+      "epoch": 2.1823980970915775,
+      "grad_norm": 0.1871381402015686,
+      "learning_rate": 3.130213808733714e-05,
+      "loss": 0.3844,
       "step": 60555
     },
     {
-      "epoch": 2.13,
-      "learning_rate": 3.210811769109913e-05,
-      "loss": 0.2799,
+      "epoch": 2.1825782967528022,
+      "grad_norm": 0.2246485948562622,
+      "learning_rate": 3.129931415200387e-05,
+      "loss": 0.4274,
       "step": 60560
     },
     {
-      "epoch": 2.13,
-      "learning_rate": 3.210538649594832e-05,
-      "loss": 0.2843,
+      "epoch": 2.182758496414027,
+      "grad_norm": 0.22834356129169464,
+      "learning_rate": 3.129649013084365e-05,
+      "loss": 0.4065,
       "step": 60565
     },
     {
-      "epoch": 2.13,
-      "learning_rate": 3.210265520853875e-05,
-      "loss": 0.2762,
+      "epoch": 2.1829386960752513,
+      "grad_norm": 0.2202819585800171,
+      "learning_rate": 3.1293666023894984e-05,
+      "loss": 0.3958,
       "step": 60570
     },
     {
-      "epoch": 2.13,
-      "learning_rate": 3.2099923828905883e-05,
-      "loss": 0.2606,
+      "epoch": 2.183118895736476,
+      "grad_norm": 0.23861894011497498,
+      "learning_rate": 3.129084183119634e-05,
+      "loss": 0.3895,
       "step": 60575
     },
     {
-      "epoch": 2.13,
-      "learning_rate": 3.209719235708519e-05,
-      "loss": 0.2838,
+      "epoch": 2.1832990953977007,
+      "grad_norm": 0.20570367574691772,
+      "learning_rate": 3.128801755278618e-05,
+      "loss": 0.3865,
       "step": 60580
     },
     {
-      "epoch": 2.13,
-      "learning_rate": 3.209446079311211e-05,
-      "loss": 0.2753,
+      "epoch": 2.1834792950589255,
+      "grad_norm": 0.2041841596364975,
+      "learning_rate": 3.128519318870301e-05,
+      "loss": 0.417,
       "step": 60585
     },
     {
-      "epoch": 2.13,
-      "learning_rate": 3.209172913702215e-05,
-      "loss": 0.2683,
+      "epoch": 2.1836594947201498,
+      "grad_norm": 0.19558176398277283,
+      "learning_rate": 3.128236873898529e-05,
+      "loss": 0.3732,
       "step": 60590
     },
     {
-      "epoch": 2.13,
-      "learning_rate": 3.208899738885076e-05,
-      "loss": 0.2761,
+      "epoch": 2.1838396943813745,
+      "grad_norm": 0.2021927386522293,
+      "learning_rate": 3.1279544203671516e-05,
+      "loss": 0.4129,
       "step": 60595
     },
     {
-      "epoch": 2.13,
-      "learning_rate": 3.208626554863341e-05,
-      "loss": 0.2663,
+      "epoch": 2.184019894042599,
+      "grad_norm": 0.19117502868175507,
+      "learning_rate": 3.1276719582800176e-05,
+      "loss": 0.4208,
       "step": 60600
     },
     {
-      "epoch": 2.13,
-      "learning_rate": 3.208353361640557e-05,
-      "loss": 0.2608,
+      "epoch": 2.184200093703824,
+      "grad_norm": 0.18669839203357697,
+      "learning_rate": 3.127389487640974e-05,
+      "loss": 0.4171,
       "step": 60605
     },
     {
-      "epoch": 2.13,
-      "learning_rate": 3.2080801592202715e-05,
-      "loss": 0.2925,
+      "epoch": 2.1843802933650487,
+      "grad_norm": 0.18443822860717773,
+      "learning_rate": 3.1271070084538703e-05,
+      "loss": 0.4149,
       "step": 60610
     },
     {
-      "epoch": 2.13,
-      "learning_rate": 3.2078069476060315e-05,
-      "loss": 0.2767,
+      "epoch": 2.184560493026273,
+      "grad_norm": 0.20405270159244537,
+      "learning_rate": 3.126824520722554e-05,
+      "loss": 0.3927,
       "step": 60615
     },
     {
-      "epoch": 2.13,
-      "learning_rate": 3.207533726801385e-05,
-      "loss": 0.2809,
+      "epoch": 2.1847406926874977,
+      "grad_norm": 0.19594557583332062,
+      "learning_rate": 3.126542024450876e-05,
+      "loss": 0.3956,
       "step": 60620
     },
     {
-      "epoch": 2.13,
-      "learning_rate": 3.207260496809879e-05,
-      "loss": 0.2835,
+      "epoch": 2.1849208923487224,
+      "grad_norm": 0.21037110686302185,
+      "learning_rate": 3.126259519642684e-05,
+      "loss": 0.3782,
       "step": 60625
     },
     {
-      "epoch": 2.13,
-      "learning_rate": 3.206987257635062e-05,
-      "loss": 0.2855,
+      "epoch": 2.185101092009947,
+      "grad_norm": 0.21924084424972534,
+      "learning_rate": 3.125977006301828e-05,
+      "loss": 0.3774,
       "step": 60630
     },
     {
-      "epoch": 2.13,
-      "learning_rate": 3.206714009280481e-05,
-      "loss": 0.2797,
+      "epoch": 2.1852812916711715,
+      "grad_norm": 0.22989316284656525,
+      "learning_rate": 3.125694484432155e-05,
+      "loss": 0.4115,
       "step": 60635
     },
     {
-      "epoch": 2.13,
-      "learning_rate": 3.206440751749685e-05,
-      "loss": 0.2827,
+      "epoch": 2.185461491332396,
+      "grad_norm": 0.20585152506828308,
+      "learning_rate": 3.1254119540375173e-05,
+      "loss": 0.3983,
       "step": 60640
     },
     {
-      "epoch": 2.13,
-      "learning_rate": 3.206167485046221e-05,
-      "loss": 0.2775,
+      "epoch": 2.185641690993621,
+      "grad_norm": 0.20063400268554688,
+      "learning_rate": 3.1251294151217614e-05,
+      "loss": 0.3712,
       "step": 60645
     },
     {
-      "epoch": 2.13,
-      "learning_rate": 3.2058942091736366e-05,
-      "loss": 0.2738,
+      "epoch": 2.1858218906548457,
+      "grad_norm": 0.18336834013462067,
+      "learning_rate": 3.1248468676887396e-05,
+      "loss": 0.4123,
       "step": 60650
     },
     {
-      "epoch": 2.13,
-      "learning_rate": 3.2056209241354826e-05,
-      "loss": 0.2301,
+      "epoch": 2.1860020903160704,
+      "grad_norm": 0.2062855362892151,
+      "learning_rate": 3.124564311742299e-05,
+      "loss": 0.3857,
       "step": 60655
     },
     {
-      "epoch": 2.13,
-      "learning_rate": 3.205347629935305e-05,
-      "loss": 0.3128,
+      "epoch": 2.1861822899772947,
+      "grad_norm": 0.2320207953453064,
+      "learning_rate": 3.124281747286291e-05,
+      "loss": 0.4176,
       "step": 60660
     },
     {
-      "epoch": 2.13,
-      "learning_rate": 3.2050743265766534e-05,
-      "loss": 0.2745,
+      "epoch": 2.1863624896385194,
+      "grad_norm": 0.17851118743419647,
+      "learning_rate": 3.1239991743245656e-05,
+      "loss": 0.3797,
       "step": 60665
     },
     {
-      "epoch": 2.13,
-      "learning_rate": 3.204801014063077e-05,
-      "loss": 0.2572,
+      "epoch": 2.186542689299744,
+      "grad_norm": 0.17879897356033325,
+      "learning_rate": 3.123716592860971e-05,
+      "loss": 0.3958,
       "step": 60670
     },
     {
-      "epoch": 2.13,
-      "learning_rate": 3.204527692398124e-05,
-      "loss": 0.2781,
+      "epoch": 2.186722888960969,
+      "grad_norm": 0.24118387699127197,
+      "learning_rate": 3.12343400289936e-05,
+      "loss": 0.4083,
       "step": 60675
     },
     {
-      "epoch": 2.13,
-      "learning_rate": 3.204254361585342e-05,
-      "loss": 0.2621,
+      "epoch": 2.186903088622193,
+      "grad_norm": 0.17323794960975647,
+      "learning_rate": 3.123151404443581e-05,
+      "loss": 0.3856,
       "step": 60680
     },
     {
-      "epoch": 2.14,
-      "learning_rate": 3.203981021628282e-05,
-      "loss": 0.2497,
+      "epoch": 2.187083288283418,
+      "grad_norm": 0.19309526681900024,
+      "learning_rate": 3.122868797497485e-05,
+      "loss": 0.3939,
       "step": 60685
     },
     {
-      "epoch": 2.14,
-      "learning_rate": 3.203707672530493e-05,
-      "loss": 0.2819,
+      "epoch": 2.1872634879446426,
+      "grad_norm": 0.20537187159061432,
+      "learning_rate": 3.122586182064921e-05,
+      "loss": 0.3747,
       "step": 60690
     },
     {
-      "epoch": 2.14,
-      "learning_rate": 3.2034343142955225e-05,
-      "loss": 0.249,
+      "epoch": 2.1874436876058674,
+      "grad_norm": 0.22961921989917755,
+      "learning_rate": 3.122303558149742e-05,
+      "loss": 0.4318,
       "step": 60695
     },
     {
-      "epoch": 2.14,
-      "learning_rate": 3.203160946926921e-05,
-      "loss": 0.2888,
+      "epoch": 2.187623887267092,
+      "grad_norm": 0.2008984237909317,
+      "learning_rate": 3.122020925755797e-05,
+      "loss": 0.4176,
       "step": 60700
     },
     {
-      "epoch": 2.14,
-      "learning_rate": 3.202887570428239e-05,
-      "loss": 0.2805,
+      "epoch": 2.1878040869283164,
+      "grad_norm": 0.2251536101102829,
+      "learning_rate": 3.121738284886938e-05,
+      "loss": 0.3891,
       "step": 60705
     },
     {
-      "epoch": 2.14,
-      "learning_rate": 3.2026141848030244e-05,
-      "loss": 0.2742,
+      "epoch": 2.187984286589541,
+      "grad_norm": 0.16935603320598602,
+      "learning_rate": 3.121455635547014e-05,
+      "loss": 0.3819,
       "step": 60710
     },
     {
-      "epoch": 2.14,
-      "learning_rate": 3.202340790054828e-05,
-      "loss": 0.2432,
+      "epoch": 2.188164486250766,
+      "grad_norm": 0.20971350371837616,
+      "learning_rate": 3.121172977739878e-05,
+      "loss": 0.4172,
       "step": 60715
     },
     {
-      "epoch": 2.14,
-      "learning_rate": 3.202067386187199e-05,
-      "loss": 0.2833,
+      "epoch": 2.1883446859119906,
+      "grad_norm": 0.25330233573913574,
+      "learning_rate": 3.1208903114693804e-05,
+      "loss": 0.403,
       "step": 60720
     },
     {
-      "epoch": 2.14,
-      "learning_rate": 3.201793973203688e-05,
-      "loss": 0.267,
+      "epoch": 2.188524885573215,
+      "grad_norm": 0.17658570408821106,
+      "learning_rate": 3.1206076367393724e-05,
+      "loss": 0.3783,
       "step": 60725
     },
     {
-      "epoch": 2.14,
-      "learning_rate": 3.201520551107844e-05,
-      "loss": 0.2622,
+      "epoch": 2.1887050852344396,
+      "grad_norm": 0.18725471198558807,
+      "learning_rate": 3.1203249535537056e-05,
+      "loss": 0.3887,
       "step": 60730
     },
     {
-      "epoch": 2.14,
-      "learning_rate": 3.201247119903219e-05,
-      "loss": 0.2818,
+      "epoch": 2.1888852848956644,
+      "grad_norm": 0.17408056557178497,
+      "learning_rate": 3.1200422619162315e-05,
+      "loss": 0.3688,
       "step": 60735
     },
     {
-      "epoch": 2.14,
-      "learning_rate": 3.2009736795933616e-05,
-      "loss": 0.2603,
+      "epoch": 2.189065484556889,
+      "grad_norm": 0.21276924014091492,
+      "learning_rate": 3.119759561830802e-05,
+      "loss": 0.4114,
       "step": 60740
     },
     {
-      "epoch": 2.14,
-      "learning_rate": 3.200700230181823e-05,
-      "loss": 0.2846,
+      "epoch": 2.189245684218114,
+      "grad_norm": 0.2514096200466156,
+      "learning_rate": 3.119476853301268e-05,
+      "loss": 0.4205,
       "step": 60745
     },
     {
-      "epoch": 2.14,
-      "learning_rate": 3.2004267716721534e-05,
-      "loss": 0.2779,
+      "epoch": 2.189425883879338,
+      "grad_norm": 0.1897374540567398,
+      "learning_rate": 3.1191941363314814e-05,
+      "loss": 0.383,
       "step": 60750
     },
     {
-      "epoch": 2.14,
-      "learning_rate": 3.200153304067904e-05,
-      "loss": 0.2703,
+      "epoch": 2.189606083540563,
+      "grad_norm": 0.23337703943252563,
+      "learning_rate": 3.1189114109252946e-05,
+      "loss": 0.4357,
       "step": 60755
     },
     {
-      "epoch": 2.14,
-      "learning_rate": 3.199879827372626e-05,
-      "loss": 0.2681,
+      "epoch": 2.1897862832017876,
+      "grad_norm": 0.2705073654651642,
+      "learning_rate": 3.118628677086561e-05,
+      "loss": 0.4014,
       "step": 60760
     },
     {
-      "epoch": 2.14,
-      "learning_rate": 3.199606341589869e-05,
-      "loss": 0.276,
+      "epoch": 2.1899664828630123,
+      "grad_norm": 0.20422200858592987,
+      "learning_rate": 3.1183459348191296e-05,
+      "loss": 0.4171,
       "step": 60765
     },
     {
-      "epoch": 2.14,
-      "learning_rate": 3.199332846723185e-05,
-      "loss": 0.2633,
+      "epoch": 2.190146682524237,
+      "grad_norm": 0.20398341119289398,
+      "learning_rate": 3.118063184126856e-05,
+      "loss": 0.3475,
       "step": 60770
     },
     {
-      "epoch": 2.14,
-      "learning_rate": 3.199059342776125e-05,
-      "loss": 0.2834,
+      "epoch": 2.1903268821854613,
+      "grad_norm": 0.17860305309295654,
+      "learning_rate": 3.11778042501359e-05,
+      "loss": 0.3701,
       "step": 60775
     },
     {
-      "epoch": 2.14,
-      "learning_rate": 3.19878582975224e-05,
-      "loss": 0.2831,
+      "epoch": 2.190507081846686,
+      "grad_norm": 0.19039638340473175,
+      "learning_rate": 3.117497657483187e-05,
+      "loss": 0.4299,
       "step": 60780
     },
     {
-      "epoch": 2.14,
-      "learning_rate": 3.198512307655081e-05,
-      "loss": 0.2631,
+      "epoch": 2.190687281507911,
+      "grad_norm": 0.2850659489631653,
+      "learning_rate": 3.117214881539496e-05,
+      "loss": 0.4103,
       "step": 60785
     },
     {
-      "epoch": 2.14,
-      "learning_rate": 3.1982387764882005e-05,
-      "loss": 0.279,
+      "epoch": 2.1908674811691355,
+      "grad_norm": 0.24027515947818756,
+      "learning_rate": 3.116932097186373e-05,
+      "loss": 0.422,
       "step": 60790
     },
     {
-      "epoch": 2.14,
-      "learning_rate": 3.19796523625515e-05,
-      "loss": 0.2773,
+      "epoch": 2.19104768083036,
+      "grad_norm": 0.18420986831188202,
+      "learning_rate": 3.116649304427669e-05,
+      "loss": 0.3831,
       "step": 60795
     },
     {
-      "epoch": 2.14,
-      "learning_rate": 3.197691686959481e-05,
-      "loss": 0.2691,
+      "epoch": 2.1912278804915846,
+      "grad_norm": 0.23363029956817627,
+      "learning_rate": 3.116366503267238e-05,
+      "loss": 0.4279,
       "step": 60800
     },
     {
-      "epoch": 2.14,
-      "learning_rate": 3.1974181286047435e-05,
-      "loss": 0.2634,
+      "epoch": 2.1914080801528093,
+      "grad_norm": 0.18705283105373383,
+      "learning_rate": 3.116083693708933e-05,
+      "loss": 0.377,
       "step": 60805
     },
     {
-      "epoch": 2.14,
-      "learning_rate": 3.197144561194493e-05,
-      "loss": 0.2702,
+      "epoch": 2.191588279814034,
+      "grad_norm": 0.22469288110733032,
+      "learning_rate": 3.115800875756606e-05,
+      "loss": 0.3951,
       "step": 60810
     },
     {
-      "epoch": 2.14,
-      "learning_rate": 3.196870984732279e-05,
-      "loss": 0.2642,
+      "epoch": 2.1917684794752588,
+      "grad_norm": 0.17964445054531097,
+      "learning_rate": 3.115518049414112e-05,
+      "loss": 0.3898,
       "step": 60815
     },
     {
-      "epoch": 2.14,
-      "learning_rate": 3.1965973992216554e-05,
-      "loss": 0.2564,
+      "epoch": 2.191948679136483,
+      "grad_norm": 0.22461974620819092,
+      "learning_rate": 3.115235214685303e-05,
+      "loss": 0.4117,
       "step": 60820
     },
     {
-      "epoch": 2.14,
-      "learning_rate": 3.196323804666173e-05,
-      "loss": 0.2799,
+      "epoch": 2.192128878797708,
+      "grad_norm": 0.1743023842573166,
+      "learning_rate": 3.114952371574035e-05,
+      "loss": 0.4163,
       "step": 60825
     },
     {
-      "epoch": 2.14,
-      "learning_rate": 3.1960502010693844e-05,
-      "loss": 0.2689,
+      "epoch": 2.1923090784589325,
+      "grad_norm": 0.22755619883537292,
+      "learning_rate": 3.114669520084158e-05,
+      "loss": 0.4297,
       "step": 60830
     },
     {
-      "epoch": 2.14,
-      "learning_rate": 3.1957765884348427e-05,
-      "loss": 0.271,
+      "epoch": 2.1924892781201573,
+      "grad_norm": 0.21366189420223236,
+      "learning_rate": 3.114386660219528e-05,
+      "loss": 0.3677,
       "step": 60835
     },
     {
-      "epoch": 2.14,
-      "learning_rate": 3.1955029667661016e-05,
-      "loss": 0.2715,
+      "epoch": 2.192669477781382,
+      "grad_norm": 0.19875569641590118,
+      "learning_rate": 3.1141037919839996e-05,
+      "loss": 0.377,
       "step": 60840
     },
     {
-      "epoch": 2.14,
-      "learning_rate": 3.195229336066712e-05,
-      "loss": 0.277,
+      "epoch": 2.1928496774426063,
+      "grad_norm": 0.18235017359256744,
+      "learning_rate": 3.113820915381426e-05,
+      "loss": 0.4023,
       "step": 60845
     },
     {
-      "epoch": 2.14,
-      "learning_rate": 3.194955696340228e-05,
-      "loss": 0.2632,
+      "epoch": 2.193029877103831,
+      "grad_norm": 0.22413848340511322,
+      "learning_rate": 3.1135380304156614e-05,
+      "loss": 0.4025,
       "step": 60850
     },
     {
-      "epoch": 2.14,
-      "learning_rate": 3.1946820475902024e-05,
-      "loss": 0.2803,
+      "epoch": 2.1932100767650557,
+      "grad_norm": 0.226275235414505,
+      "learning_rate": 3.11325513709056e-05,
+      "loss": 0.4236,
       "step": 60855
     },
     {
-      "epoch": 2.14,
-      "learning_rate": 3.1944083898201874e-05,
-      "loss": 0.2787,
+      "epoch": 2.1933902764262805,
+      "grad_norm": 0.23043504357337952,
+      "learning_rate": 3.1129722354099746e-05,
+      "loss": 0.3911,
       "step": 60860
     },
     {
-      "epoch": 2.14,
-      "learning_rate": 3.194134723033738e-05,
-      "loss": 0.2935,
+      "epoch": 2.1935704760875048,
+      "grad_norm": 0.22926190495491028,
+      "learning_rate": 3.1126893253777626e-05,
+      "loss": 0.3957,
       "step": 60865
     },
     {
-      "epoch": 2.14,
-      "learning_rate": 3.193861047234406e-05,
-      "loss": 0.279,
+      "epoch": 2.1937506757487295,
+      "grad_norm": 0.24694472551345825,
+      "learning_rate": 3.1124064069977766e-05,
+      "loss": 0.3715,
       "step": 60870
     },
     {
-      "epoch": 2.14,
-      "learning_rate": 3.193587362425746e-05,
-      "loss": 0.2632,
+      "epoch": 2.1939308754099542,
+      "grad_norm": 0.16308090090751648,
+      "learning_rate": 3.112123480273872e-05,
+      "loss": 0.3668,
       "step": 60875
     },
     {
-      "epoch": 2.14,
-      "learning_rate": 3.193313668611311e-05,
-      "loss": 0.2805,
+      "epoch": 2.194111075071179,
+      "grad_norm": 0.17201615869998932,
+      "learning_rate": 3.111840545209903e-05,
+      "loss": 0.3633,
       "step": 60880
     },
     {
-      "epoch": 2.14,
-      "learning_rate": 3.193039965794655e-05,
-      "loss": 0.2754,
+      "epoch": 2.1942912747324037,
+      "grad_norm": 0.20584902167320251,
+      "learning_rate": 3.1115576018097264e-05,
+      "loss": 0.4049,
       "step": 60885
     },
     {
-      "epoch": 2.14,
-      "learning_rate": 3.192766253979332e-05,
-      "loss": 0.259,
+      "epoch": 2.194471474393628,
+      "grad_norm": 0.16585664451122284,
+      "learning_rate": 3.111274650077195e-05,
+      "loss": 0.3564,
       "step": 60890
     },
     {
-      "epoch": 2.14,
-      "learning_rate": 3.192492533168896e-05,
-      "loss": 0.2614,
+      "epoch": 2.1946516740548527,
+      "grad_norm": 0.18358787894248962,
+      "learning_rate": 3.110991690016165e-05,
+      "loss": 0.4176,
       "step": 60895
     },
     {
-      "epoch": 2.14,
-      "learning_rate": 3.1922188033668995e-05,
-      "loss": 0.2569,
+      "epoch": 2.1948318737160775,
+      "grad_norm": 0.28625836968421936,
+      "learning_rate": 3.11070872163049e-05,
+      "loss": 0.4292,
       "step": 60900
     },
     {
-      "epoch": 2.14,
-      "learning_rate": 3.1919450645769e-05,
-      "loss": 0.2703,
+      "epoch": 2.195012073377302,
+      "grad_norm": 0.2175116091966629,
+      "learning_rate": 3.110425744924029e-05,
+      "loss": 0.4274,
       "step": 60905
     },
     {
-      "epoch": 2.14,
-      "learning_rate": 3.1916713168024484e-05,
-      "loss": 0.2464,
+      "epoch": 2.1951922730385265,
+      "grad_norm": 0.18882659077644348,
+      "learning_rate": 3.1101427599006346e-05,
+      "loss": 0.3686,
       "step": 60910
     },
     {
-      "epoch": 2.14,
-      "learning_rate": 3.1913975600471006e-05,
-      "loss": 0.2377,
+      "epoch": 2.195372472699751,
+      "grad_norm": 0.2079334855079651,
+      "learning_rate": 3.109859766564163e-05,
+      "loss": 0.4185,
       "step": 60915
     },
     {
-      "epoch": 2.14,
-      "learning_rate": 3.191123794314411e-05,
-      "loss": 0.2752,
+      "epoch": 2.195552672360976,
+      "grad_norm": 0.18339209258556366,
+      "learning_rate": 3.1095767649184704e-05,
+      "loss": 0.399,
       "step": 60920
     },
     {
-      "epoch": 2.14,
-      "learning_rate": 3.190850019607935e-05,
-      "loss": 0.2745,
+      "epoch": 2.1957328720222007,
+      "grad_norm": 0.17973054945468903,
+      "learning_rate": 3.1092937549674126e-05,
+      "loss": 0.3799,
       "step": 60925
     },
     {
-      "epoch": 2.14,
-      "learning_rate": 3.190576235931227e-05,
-      "loss": 0.2692,
+      "epoch": 2.1959130716834254,
+      "grad_norm": 0.24256204068660736,
+      "learning_rate": 3.109010736714845e-05,
+      "loss": 0.3867,
       "step": 60930
     },
     {
-      "epoch": 2.14,
-      "learning_rate": 3.190302443287841e-05,
-      "loss": 0.2722,
+      "epoch": 2.1960932713446497,
+      "grad_norm": 0.21241244673728943,
+      "learning_rate": 3.1087277101646244e-05,
+      "loss": 0.4304,
       "step": 60935
     },
     {
-      "epoch": 2.14,
-      "learning_rate": 3.190028641681334e-05,
-      "loss": 0.28,
+      "epoch": 2.1962734710058744,
+      "grad_norm": 0.1754169911146164,
+      "learning_rate": 3.108444675320607e-05,
+      "loss": 0.4074,
       "step": 60940
     },
     {
-      "epoch": 2.14,
-      "learning_rate": 3.189754831115258e-05,
-      "loss": 0.2689,
+      "epoch": 2.196453670667099,
+      "grad_norm": 0.20329289138317108,
+      "learning_rate": 3.108161632186648e-05,
+      "loss": 0.4086,
       "step": 60945
     },
     {
-      "epoch": 2.14,
-      "learning_rate": 3.189481011593171e-05,
-      "loss": 0.2435,
+      "epoch": 2.196633870328324,
+      "grad_norm": 0.2189503312110901,
+      "learning_rate": 3.107878580766604e-05,
+      "loss": 0.4192,
       "step": 60950
     },
     {
-      "epoch": 2.14,
-      "learning_rate": 3.1892071831186275e-05,
-      "loss": 0.2531,
+      "epoch": 2.196814069989548,
+      "grad_norm": 0.1868715137243271,
+      "learning_rate": 3.107595521064333e-05,
+      "loss": 0.3519,
       "step": 60955
     },
     {
-      "epoch": 2.14,
-      "learning_rate": 3.188933345695183e-05,
-      "loss": 0.2654,
+      "epoch": 2.196994269650773,
+      "grad_norm": 0.25199243426322937,
+      "learning_rate": 3.1073124530836894e-05,
+      "loss": 0.3997,
       "step": 60960
     },
     {
-      "epoch": 2.14,
-      "learning_rate": 3.188659499326393e-05,
-      "loss": 0.2704,
+      "epoch": 2.1971744693119977,
+      "grad_norm": 0.21293796598911285,
+      "learning_rate": 3.107029376828533e-05,
+      "loss": 0.3592,
       "step": 60965
     },
     {
-      "epoch": 2.15,
-      "learning_rate": 3.1883856440158124e-05,
-      "loss": 0.3151,
+      "epoch": 2.1973546689732224,
+      "grad_norm": 0.21643023192882538,
+      "learning_rate": 3.1067462923027174e-05,
+      "loss": 0.418,
       "step": 60970
     },
     {
-      "epoch": 2.15,
-      "learning_rate": 3.188111779766998e-05,
-      "loss": 0.2528,
+      "epoch": 2.197534868634447,
+      "grad_norm": 0.20323172211647034,
+      "learning_rate": 3.106463199510102e-05,
+      "loss": 0.4273,
       "step": 60975
     },
     {
-      "epoch": 2.15,
-      "learning_rate": 3.187837906583507e-05,
-      "loss": 0.2974,
+      "epoch": 2.1977150682956714,
+      "grad_norm": 0.20973147451877594,
+      "learning_rate": 3.106180098454542e-05,
+      "loss": 0.4134,
       "step": 60980
     },
     {
-      "epoch": 2.15,
-      "learning_rate": 3.187564024468893e-05,
-      "loss": 0.2805,
+      "epoch": 2.197895267956896,
+      "grad_norm": 0.19453154504299164,
+      "learning_rate": 3.105896989139896e-05,
+      "loss": 0.3847,
       "step": 60985
     },
     {
-      "epoch": 2.15,
-      "learning_rate": 3.187290133426714e-05,
-      "loss": 0.2844,
+      "epoch": 2.198075467618121,
+      "grad_norm": 0.2046874463558197,
+      "learning_rate": 3.105613871570021e-05,
+      "loss": 0.3834,
       "step": 60990
     },
     {
-      "epoch": 2.15,
-      "learning_rate": 3.187016233460525e-05,
-      "loss": 0.2758,
+      "epoch": 2.1982556672793456,
+      "grad_norm": 0.2236749827861786,
+      "learning_rate": 3.105330745748774e-05,
+      "loss": 0.4089,
       "step": 60995
     },
     {
-      "epoch": 2.15,
-      "learning_rate": 3.186742324573883e-05,
-      "loss": 0.2762,
+      "epoch": 2.19843586694057,
+      "grad_norm": 0.2126908153295517,
+      "learning_rate": 3.105047611680013e-05,
+      "loss": 0.4004,
       "step": 61000
     },
     {
-      "epoch": 2.15,
-      "eval_loss": 0.2689875662326813,
-      "eval_runtime": 10.5446,
-      "eval_samples_per_second": 9.483,
-      "eval_steps_per_second": 9.483,
+      "epoch": 2.19843586694057,
+      "eval_loss": 0.43686825037002563,
+      "eval_runtime": 3.5139,
+      "eval_samples_per_second": 28.458,
+      "eval_steps_per_second": 7.115,
       "step": 61000
     },
     {
-      "epoch": 2.15,
-      "learning_rate": 3.186468406770344e-05,
-      "loss": 0.2781,
+      "epoch": 2.1986160666017946,
+      "grad_norm": 0.19092722237110138,
+      "learning_rate": 3.1047644693675956e-05,
+      "loss": 0.3829,
       "step": 61005
     },
     {
-      "epoch": 2.15,
-      "learning_rate": 3.186194480053467e-05,
-      "loss": 0.2567,
+      "epoch": 2.1987962662630194,
+      "grad_norm": 0.24691295623779297,
+      "learning_rate": 3.104481318815379e-05,
+      "loss": 0.4313,
       "step": 61010
     },
     {
-      "epoch": 2.15,
-      "learning_rate": 3.185920544426805e-05,
-      "loss": 0.2522,
+      "epoch": 2.198976465924244,
+      "grad_norm": 0.2079208791255951,
+      "learning_rate": 3.104198160027222e-05,
+      "loss": 0.425,
       "step": 61015
     },
     {
-      "epoch": 2.15,
-      "learning_rate": 3.185646599893918e-05,
-      "loss": 0.2552,
+      "epoch": 2.199156665585469,
+      "grad_norm": 0.19778048992156982,
+      "learning_rate": 3.103914993006981e-05,
+      "loss": 0.3969,
       "step": 61020
     },
     {
-      "epoch": 2.15,
-      "learning_rate": 3.18537264645836e-05,
-      "loss": 0.2444,
+      "epoch": 2.199336865246693,
+      "grad_norm": 0.16332995891571045,
+      "learning_rate": 3.1036318177585156e-05,
+      "loss": 0.4092,
       "step": 61025
     },
     {
-      "epoch": 2.15,
-      "learning_rate": 3.185098684123691e-05,
-      "loss": 0.2853,
+      "epoch": 2.199517064907918,
+      "grad_norm": 0.1742788553237915,
+      "learning_rate": 3.103348634285684e-05,
+      "loss": 0.4048,
       "step": 61030
     },
     {
-      "epoch": 2.15,
-      "learning_rate": 3.184824712893467e-05,
-      "loss": 0.3003,
+      "epoch": 2.1996972645691426,
+      "grad_norm": 0.14753888547420502,
+      "learning_rate": 3.103065442592344e-05,
+      "loss": 0.3834,
       "step": 61035
     },
     {
-      "epoch": 2.15,
-      "learning_rate": 3.184550732771246e-05,
-      "loss": 0.2714,
+      "epoch": 2.1998774642303673,
+      "grad_norm": 0.22011804580688477,
+      "learning_rate": 3.1027822426823536e-05,
+      "loss": 0.4417,
       "step": 61040
     },
     {
-      "epoch": 2.15,
-      "learning_rate": 3.184276743760584e-05,
-      "loss": 0.2848,
+      "epoch": 2.200057663891592,
+      "grad_norm": 0.20552678406238556,
+      "learning_rate": 3.1024990345595725e-05,
+      "loss": 0.4126,
       "step": 61045
     },
     {
-      "epoch": 2.15,
-      "learning_rate": 3.18400274586504e-05,
-      "loss": 0.2724,
+      "epoch": 2.2002378635528164,
+      "grad_norm": 0.18421411514282227,
+      "learning_rate": 3.1022158182278584e-05,
+      "loss": 0.4111,
       "step": 61050
     },
     {
-      "epoch": 2.15,
-      "learning_rate": 3.183728739088171e-05,
-      "loss": 0.2825,
+      "epoch": 2.200418063214041,
+      "grad_norm": 0.21071533858776093,
+      "learning_rate": 3.1019325936910696e-05,
+      "loss": 0.4239,
       "step": 61055
     },
     {
-      "epoch": 2.15,
-      "learning_rate": 3.183454723433535e-05,
-      "loss": 0.2541,
+      "epoch": 2.200598262875266,
+      "grad_norm": 0.19925324618816376,
+      "learning_rate": 3.1016493609530666e-05,
+      "loss": 0.3629,
       "step": 61060
     },
     {
-      "epoch": 2.15,
-      "learning_rate": 3.183180698904689e-05,
-      "loss": 0.2586,
+      "epoch": 2.2007784625364906,
+      "grad_norm": 0.19704802334308624,
+      "learning_rate": 3.101366120017707e-05,
+      "loss": 0.3847,
       "step": 61065
     },
     {
-      "epoch": 2.15,
-      "learning_rate": 3.182906665505193e-05,
-      "loss": 0.2811,
+      "epoch": 2.2009586621977153,
+      "grad_norm": 0.17782747745513916,
+      "learning_rate": 3.1010828708888516e-05,
+      "loss": 0.3889,
       "step": 61070
     },
     {
-      "epoch": 2.15,
-      "learning_rate": 3.182632623238604e-05,
-      "loss": 0.2755,
+      "epoch": 2.2011388618589396,
+      "grad_norm": 0.1703190803527832,
+      "learning_rate": 3.1007996135703576e-05,
+      "loss": 0.3616,
       "step": 61075
     },
     {
-      "epoch": 2.15,
-      "learning_rate": 3.182358572108479e-05,
-      "loss": 0.2842,
+      "epoch": 2.2013190615201643,
+      "grad_norm": 0.22071696817874908,
+      "learning_rate": 3.100516348066085e-05,
+      "loss": 0.4017,
       "step": 61080
     },
     {
-      "epoch": 2.15,
-      "learning_rate": 3.1820845121183786e-05,
-      "loss": 0.2524,
+      "epoch": 2.201499261181389,
+      "grad_norm": 0.20108157396316528,
+      "learning_rate": 3.100233074379894e-05,
+      "loss": 0.3965,
       "step": 61085
     },
     {
-      "epoch": 2.15,
-      "learning_rate": 3.1818104432718596e-05,
-      "loss": 0.2618,
+      "epoch": 2.201679460842614,
+      "grad_norm": 0.16138023138046265,
+      "learning_rate": 3.099949792515643e-05,
+      "loss": 0.4049,
       "step": 61090
     },
     {
-      "epoch": 2.15,
-      "learning_rate": 3.181536365572483e-05,
-      "loss": 0.2663,
+      "epoch": 2.201859660503838,
+      "grad_norm": 0.1845240294933319,
+      "learning_rate": 3.0996665024771924e-05,
+      "loss": 0.4032,
       "step": 61095
     },
     {
-      "epoch": 2.15,
-      "learning_rate": 3.181262279023804e-05,
-      "loss": 0.2534,
+      "epoch": 2.202039860165063,
+      "grad_norm": 0.17944936454296112,
+      "learning_rate": 3.099383204268402e-05,
+      "loss": 0.3838,
       "step": 61100
     },
     {
-      "epoch": 2.15,
-      "learning_rate": 3.180988183629384e-05,
-      "loss": 0.259,
+      "epoch": 2.2022200598262875,
+      "grad_norm": 0.17411671578884125,
+      "learning_rate": 3.0990998978931315e-05,
+      "loss": 0.4018,
       "step": 61105
     },
     {
-      "epoch": 2.15,
-      "learning_rate": 3.180714079392781e-05,
-      "loss": 0.2751,
+      "epoch": 2.2024002594875123,
+      "grad_norm": 0.1903943568468094,
+      "learning_rate": 3.0988165833552404e-05,
+      "loss": 0.3924,
       "step": 61110
     },
     {
-      "epoch": 2.15,
-      "learning_rate": 3.180439966317555e-05,
-      "loss": 0.2816,
+      "epoch": 2.202580459148737,
+      "grad_norm": 0.1921924352645874,
+      "learning_rate": 3.0985332606585905e-05,
+      "loss": 0.3686,
       "step": 61115
     },
     {
-      "epoch": 2.15,
-      "learning_rate": 3.180165844407264e-05,
-      "loss": 0.2554,
+      "epoch": 2.2027606588099613,
+      "grad_norm": 0.21751470863819122,
+      "learning_rate": 3.0982499298070394e-05,
+      "loss": 0.387,
       "step": 61120
     },
     {
-      "epoch": 2.15,
-      "learning_rate": 3.1798917136654676e-05,
-      "loss": 0.2584,
+      "epoch": 2.202940858471186,
+      "grad_norm": 0.18325336277484894,
+      "learning_rate": 3.0979665908044495e-05,
+      "loss": 0.3853,
       "step": 61125
     },
     {
-      "epoch": 2.15,
-      "learning_rate": 3.179617574095725e-05,
-      "loss": 0.2674,
+      "epoch": 2.2031210581324108,
+      "grad_norm": 0.21470201015472412,
+      "learning_rate": 3.097683243654681e-05,
+      "loss": 0.4171,
       "step": 61130
     },
     {
-      "epoch": 2.15,
-      "learning_rate": 3.1793434257015965e-05,
-      "loss": 0.272,
+      "epoch": 2.2033012577936355,
+      "grad_norm": 0.2272443175315857,
+      "learning_rate": 3.097399888361593e-05,
+      "loss": 0.3425,
       "step": 61135
     },
     {
-      "epoch": 2.15,
-      "learning_rate": 3.1790692684866414e-05,
-      "loss": 0.2681,
+      "epoch": 2.20348145745486,
+      "grad_norm": 0.22367946803569794,
+      "learning_rate": 3.097116524929048e-05,
+      "loss": 0.3819,
       "step": 61140
     },
     {
-      "epoch": 2.15,
-      "learning_rate": 3.1787951024544196e-05,
-      "loss": 0.2902,
+      "epoch": 2.2036616571160845,
+      "grad_norm": 0.16753867268562317,
+      "learning_rate": 3.096833153360905e-05,
+      "loss": 0.3832,
       "step": 61145
     },
     {
-      "epoch": 2.15,
-      "learning_rate": 3.178520927608491e-05,
-      "loss": 0.2537,
+      "epoch": 2.2038418567773093,
+      "grad_norm": 0.18004858493804932,
+      "learning_rate": 3.096549773661027e-05,
+      "loss": 0.4408,
       "step": 61150
     },
     {
-      "epoch": 2.15,
-      "learning_rate": 3.178246743952414e-05,
-      "loss": 0.289,
+      "epoch": 2.204022056438534,
+      "grad_norm": 0.20204921066761017,
+      "learning_rate": 3.096266385833273e-05,
+      "loss": 0.3311,
       "step": 61155
     },
     {
-      "epoch": 2.15,
-      "learning_rate": 3.177972551489751e-05,
-      "loss": 0.2592,
+      "epoch": 2.2042022560997587,
+      "grad_norm": 0.19747331738471985,
+      "learning_rate": 3.0959829898815053e-05,
+      "loss": 0.4104,
       "step": 61160
     },
     {
-      "epoch": 2.15,
-      "learning_rate": 3.177698350224061e-05,
-      "loss": 0.2744,
+      "epoch": 2.204382455760983,
+      "grad_norm": 0.20761699974536896,
+      "learning_rate": 3.095699585809584e-05,
+      "loss": 0.3564,
       "step": 61165
     },
     {
-      "epoch": 2.15,
-      "learning_rate": 3.1774241401589044e-05,
-      "loss": 0.2709,
+      "epoch": 2.2045626554222078,
+      "grad_norm": 0.18775121867656708,
+      "learning_rate": 3.0954161736213725e-05,
+      "loss": 0.3941,
       "step": 61170
     },
     {
-      "epoch": 2.15,
-      "learning_rate": 3.177149921297842e-05,
-      "loss": 0.2811,
+      "epoch": 2.2047428550834325,
+      "grad_norm": 0.1855042576789856,
+      "learning_rate": 3.0951327533207305e-05,
+      "loss": 0.398,
       "step": 61175
     },
     {
-      "epoch": 2.15,
-      "learning_rate": 3.176875693644434e-05,
-      "loss": 0.2822,
+      "epoch": 2.204923054744657,
+      "grad_norm": 0.15789079666137695,
+      "learning_rate": 3.094849324911519e-05,
+      "loss": 0.3827,
       "step": 61180
     },
     {
-      "epoch": 2.15,
-      "learning_rate": 3.1766014572022415e-05,
-      "loss": 0.278,
+      "epoch": 2.2051032544058815,
+      "grad_norm": 0.20551757514476776,
+      "learning_rate": 3.0945658883976014e-05,
+      "loss": 0.4138,
       "step": 61185
     },
     {
-      "epoch": 2.15,
-      "learning_rate": 3.1763272119748244e-05,
-      "loss": 0.2664,
+      "epoch": 2.2052834540671062,
+      "grad_norm": 0.20195716619491577,
+      "learning_rate": 3.0942824437828386e-05,
+      "loss": 0.4424,
       "step": 61190
     },
     {
-      "epoch": 2.15,
-      "learning_rate": 3.176052957965745e-05,
-      "loss": 0.2495,
+      "epoch": 2.205463653728331,
+      "grad_norm": 0.18200279772281647,
+      "learning_rate": 3.093998991071093e-05,
+      "loss": 0.4024,
       "step": 61195
     },
     {
-      "epoch": 2.15,
-      "learning_rate": 3.175778695178563e-05,
-      "loss": 0.2562,
+      "epoch": 2.2056438533895557,
+      "grad_norm": 0.21861431002616882,
+      "learning_rate": 3.0937155302662256e-05,
+      "loss": 0.4019,
       "step": 61200
     },
     {
-      "epoch": 2.15,
-      "learning_rate": 3.17550442361684e-05,
-      "loss": 0.2734,
+      "epoch": 2.2058240530507804,
+      "grad_norm": 0.18421269953250885,
+      "learning_rate": 3.093432061372098e-05,
+      "loss": 0.3865,
       "step": 61205
     },
     {
-      "epoch": 2.15,
-      "learning_rate": 3.175230143284137e-05,
-      "loss": 0.2879,
+      "epoch": 2.2060042527120047,
+      "grad_norm": 0.2282218486070633,
+      "learning_rate": 3.093148584392576e-05,
+      "loss": 0.3667,
       "step": 61210
     },
     {
-      "epoch": 2.15,
-      "learning_rate": 3.174955854184016e-05,
-      "loss": 0.264,
+      "epoch": 2.2061844523732295,
+      "grad_norm": 0.17307542264461517,
+      "learning_rate": 3.0928650993315175e-05,
+      "loss": 0.4014,
       "step": 61215
     },
     {
-      "epoch": 2.15,
-      "learning_rate": 3.174681556320038e-05,
-      "loss": 0.2764,
+      "epoch": 2.206364652034454,
+      "grad_norm": 0.23555909097194672,
+      "learning_rate": 3.092581606192788e-05,
+      "loss": 0.4212,
       "step": 61220
     },
     {
-      "epoch": 2.15,
-      "learning_rate": 3.174407249695764e-05,
-      "loss": 0.2588,
+      "epoch": 2.206544851695679,
+      "grad_norm": 0.2506897747516632,
+      "learning_rate": 3.092298104980247e-05,
+      "loss": 0.3723,
       "step": 61225
     },
     {
-      "epoch": 2.15,
-      "learning_rate": 3.1741329343147575e-05,
-      "loss": 0.2544,
+      "epoch": 2.2067250513569032,
+      "grad_norm": 0.2201319932937622,
+      "learning_rate": 3.0920145956977606e-05,
+      "loss": 0.3987,
       "step": 61230
     },
     {
-      "epoch": 2.15,
-      "learning_rate": 3.173858610180578e-05,
-      "loss": 0.2865,
+      "epoch": 2.206905251018128,
+      "grad_norm": 0.17496994137763977,
+      "learning_rate": 3.09173107834919e-05,
+      "loss": 0.3906,
       "step": 61235
     },
     {
-      "epoch": 2.15,
-      "learning_rate": 3.1735842772967895e-05,
-      "loss": 0.2596,
+      "epoch": 2.2070854506793527,
+      "grad_norm": 0.20881304144859314,
+      "learning_rate": 3.0914475529383966e-05,
+      "loss": 0.3911,
       "step": 61240
     },
     {
-      "epoch": 2.15,
-      "learning_rate": 3.1733099356669514e-05,
-      "loss": 0.2852,
+      "epoch": 2.2072656503405774,
+      "grad_norm": 0.21811749041080475,
+      "learning_rate": 3.091164019469246e-05,
+      "loss": 0.3785,
       "step": 61245
     },
     {
-      "epoch": 2.15,
-      "learning_rate": 3.1730355852946295e-05,
-      "loss": 0.2792,
+      "epoch": 2.207445850001802,
+      "grad_norm": 0.18549318611621857,
+      "learning_rate": 3.0908804779456e-05,
+      "loss": 0.3929,
       "step": 61250
     },
     {
-      "epoch": 2.16,
-      "learning_rate": 3.1727612261833826e-05,
-      "loss": 0.2798,
+      "epoch": 2.2076260496630264,
+      "grad_norm": 0.21845227479934692,
+      "learning_rate": 3.090596928371322e-05,
+      "loss": 0.3835,
       "step": 61255
     },
     {
-      "epoch": 2.16,
-      "learning_rate": 3.1724868583367756e-05,
-      "loss": 0.277,
+      "epoch": 2.207806249324251,
+      "grad_norm": 0.2296115756034851,
+      "learning_rate": 3.0903133707502744e-05,
+      "loss": 0.4349,
       "step": 61260
     },
     {
-      "epoch": 2.16,
-      "learning_rate": 3.172212481758369e-05,
-      "loss": 0.2579,
+      "epoch": 2.207986448985476,
+      "grad_norm": 0.19739088416099548,
+      "learning_rate": 3.090029805086322e-05,
+      "loss": 0.3676,
       "step": 61265
     },
     {
-      "epoch": 2.16,
-      "learning_rate": 3.171938096451727e-05,
-      "loss": 0.256,
+      "epoch": 2.2081666486467006,
+      "grad_norm": 0.23218046128749847,
+      "learning_rate": 3.089746231383327e-05,
+      "loss": 0.3863,
       "step": 61270
     },
     {
-      "epoch": 2.16,
-      "learning_rate": 3.171663702420412e-05,
-      "loss": 0.2693,
+      "epoch": 2.2083468483079254,
+      "grad_norm": 0.17996297776699066,
+      "learning_rate": 3.089462649645155e-05,
+      "loss": 0.3793,
       "step": 61275
     },
     {
-      "epoch": 2.16,
-      "learning_rate": 3.171389299667986e-05,
-      "loss": 0.2638,
+      "epoch": 2.2085270479691497,
+      "grad_norm": 0.22149422764778137,
+      "learning_rate": 3.089179059875668e-05,
+      "loss": 0.4028,
       "step": 61280
     },
     {
-      "epoch": 2.16,
-      "learning_rate": 3.171114888198012e-05,
-      "loss": 0.2672,
+      "epoch": 2.2087072476303744,
+      "grad_norm": 0.18242286145687103,
+      "learning_rate": 3.0888954620787306e-05,
+      "loss": 0.3801,
       "step": 61285
     },
     {
-      "epoch": 2.16,
-      "learning_rate": 3.170840468014054e-05,
-      "loss": 0.2712,
+      "epoch": 2.208887447291599,
+      "grad_norm": 0.18819750845432281,
+      "learning_rate": 3.0886118562582056e-05,
+      "loss": 0.3964,
       "step": 61290
     },
     {
-      "epoch": 2.16,
-      "learning_rate": 3.170566039119675e-05,
-      "loss": 0.277,
+      "epoch": 2.209067646952824,
+      "grad_norm": 0.19808700680732727,
+      "learning_rate": 3.0883282424179586e-05,
+      "loss": 0.4066,
       "step": 61295
     },
     {
-      "epoch": 2.16,
-      "learning_rate": 3.170291601518436e-05,
-      "loss": 0.2836,
+      "epoch": 2.209247846614048,
+      "grad_norm": 0.1916135549545288,
+      "learning_rate": 3.088044620561854e-05,
+      "loss": 0.4026,
       "step": 61300
     },
     {
-      "epoch": 2.16,
-      "learning_rate": 3.1700171552139036e-05,
-      "loss": 0.2654,
+      "epoch": 2.209428046275273,
+      "grad_norm": 0.2114742547273636,
+      "learning_rate": 3.0877609906937545e-05,
+      "loss": 0.3913,
       "step": 61305
     },
     {
-      "epoch": 2.16,
-      "learning_rate": 3.16974270020964e-05,
-      "loss": 0.2694,
+      "epoch": 2.2096082459364976,
+      "grad_norm": 0.2341557890176773,
+      "learning_rate": 3.0874773528175245e-05,
+      "loss": 0.3725,
       "step": 61310
     },
     {
-      "epoch": 2.16,
-      "learning_rate": 3.1694682365092096e-05,
-      "loss": 0.249,
+      "epoch": 2.2097884455977224,
+      "grad_norm": 0.20667120814323425,
+      "learning_rate": 3.087193706937031e-05,
+      "loss": 0.4142,
       "step": 61315
     },
     {
-      "epoch": 2.16,
-      "learning_rate": 3.169193764116175e-05,
-      "loss": 0.2703,
+      "epoch": 2.209968645258947,
+      "grad_norm": 0.23338265717029572,
+      "learning_rate": 3.086910053056136e-05,
+      "loss": 0.406,
       "step": 61320
     },
     {
-      "epoch": 2.16,
-      "learning_rate": 3.1689192830340984e-05,
-      "loss": 0.2533,
+      "epoch": 2.2101488449201714,
+      "grad_norm": 0.22917935252189636,
+      "learning_rate": 3.086626391178705e-05,
+      "loss": 0.3956,
       "step": 61325
     },
     {
-      "epoch": 2.16,
-      "learning_rate": 3.1686447932665475e-05,
-      "loss": 0.2801,
+      "epoch": 2.210329044581396,
+      "grad_norm": 0.2156306803226471,
+      "learning_rate": 3.0863427213086026e-05,
+      "loss": 0.4162,
       "step": 61330
     },
     {
-      "epoch": 2.16,
-      "learning_rate": 3.1683702948170845e-05,
-      "loss": 0.2747,
+      "epoch": 2.210509244242621,
+      "grad_norm": 0.20012244582176208,
+      "learning_rate": 3.086059043449695e-05,
+      "loss": 0.4189,
       "step": 61335
     },
     {
-      "epoch": 2.16,
-      "learning_rate": 3.1680957876892735e-05,
-      "loss": 0.2789,
+      "epoch": 2.2106894439038456,
+      "grad_norm": 0.18535156548023224,
+      "learning_rate": 3.085775357605847e-05,
+      "loss": 0.4042,
       "step": 61340
     },
     {
-      "epoch": 2.16,
-      "learning_rate": 3.16782127188668e-05,
-      "loss": 0.2675,
+      "epoch": 2.2108696435650703,
+      "grad_norm": 0.18791675567626953,
+      "learning_rate": 3.0854916637809215e-05,
+      "loss": 0.3618,
       "step": 61345
     },
     {
-      "epoch": 2.16,
-      "learning_rate": 3.167546747412866e-05,
-      "loss": 0.2792,
+      "epoch": 2.2110498432262946,
+      "grad_norm": 0.2098093181848526,
+      "learning_rate": 3.085207961978786e-05,
+      "loss": 0.4419,
       "step": 61350
     },
     {
-      "epoch": 2.16,
-      "learning_rate": 3.167272214271398e-05,
-      "loss": 0.2758,
+      "epoch": 2.2112300428875193,
+      "grad_norm": 0.22902563214302063,
+      "learning_rate": 3.0849242522033064e-05,
+      "loss": 0.4385,
       "step": 61355
     },
     {
-      "epoch": 2.16,
-      "learning_rate": 3.16699767246584e-05,
-      "loss": 0.3046,
+      "epoch": 2.211410242548744,
+      "grad_norm": 0.20538847148418427,
+      "learning_rate": 3.084640534458346e-05,
+      "loss": 0.385,
       "step": 61360
     },
     {
-      "epoch": 2.16,
-      "learning_rate": 3.166723121999757e-05,
-      "loss": 0.285,
+      "epoch": 2.211590442209969,
+      "grad_norm": 0.24018481373786926,
+      "learning_rate": 3.084356808747772e-05,
+      "loss": 0.4011,
       "step": 61365
     },
     {
-      "epoch": 2.16,
-      "learning_rate": 3.1664485628767126e-05,
-      "loss": 0.2992,
+      "epoch": 2.211770641871193,
+      "grad_norm": 0.17867416143417358,
+      "learning_rate": 3.084073075075449e-05,
+      "loss": 0.3826,
       "step": 61370
     },
     {
-      "epoch": 2.16,
-      "learning_rate": 3.166173995100275e-05,
-      "loss": 0.3017,
+      "epoch": 2.211950841532418,
+      "grad_norm": 0.22849269211292267,
+      "learning_rate": 3.083789333445244e-05,
+      "loss": 0.4173,
       "step": 61375
     },
     {
-      "epoch": 2.16,
-      "learning_rate": 3.165899418674005e-05,
-      "loss": 0.3009,
+      "epoch": 2.2121310411936426,
+      "grad_norm": 0.20987308025360107,
+      "learning_rate": 3.0835055838610224e-05,
+      "loss": 0.414,
       "step": 61380
     },
     {
-      "epoch": 2.16,
-      "learning_rate": 3.1656248336014706e-05,
-      "loss": 0.2688,
+      "epoch": 2.2123112408548673,
+      "grad_norm": 0.19698165357112885,
+      "learning_rate": 3.08322182632665e-05,
+      "loss": 0.3898,
       "step": 61385
     },
     {
-      "epoch": 2.16,
-      "learning_rate": 3.1653502398862356e-05,
-      "loss": 0.297,
+      "epoch": 2.212491440516092,
+      "grad_norm": 0.2028915286064148,
+      "learning_rate": 3.082938060845993e-05,
+      "loss": 0.4183,
       "step": 61390
     },
     {
-      "epoch": 2.16,
-      "learning_rate": 3.165075637531867e-05,
-      "loss": 0.2663,
+      "epoch": 2.2126716401773163,
+      "grad_norm": 0.19117394089698792,
+      "learning_rate": 3.082654287422918e-05,
+      "loss": 0.4264,
       "step": 61395
     },
     {
-      "epoch": 2.16,
-      "learning_rate": 3.16480102654193e-05,
-      "loss": 0.2722,
+      "epoch": 2.212851839838541,
+      "grad_norm": 0.23229442536830902,
+      "learning_rate": 3.082370506061291e-05,
+      "loss": 0.3847,
       "step": 61400
     },
     {
-      "epoch": 2.16,
-      "learning_rate": 3.1645264069199896e-05,
-      "loss": 0.2626,
+      "epoch": 2.213032039499766,
+      "grad_norm": 0.1863051801919937,
+      "learning_rate": 3.0820867167649794e-05,
+      "loss": 0.3803,
       "step": 61405
     },
     {
-      "epoch": 2.16,
-      "learning_rate": 3.164251778669611e-05,
-      "loss": 0.2683,
+      "epoch": 2.2132122391609905,
+      "grad_norm": 0.17941780388355255,
+      "learning_rate": 3.081802919537847e-05,
+      "loss": 0.3799,
       "step": 61410
     },
     {
-      "epoch": 2.16,
-      "learning_rate": 3.1639771417943616e-05,
-      "loss": 0.2736,
+      "epoch": 2.213392438822215,
+      "grad_norm": 0.20453837513923645,
+      "learning_rate": 3.081519114383764e-05,
+      "loss": 0.4076,
       "step": 61415
     },
     {
-      "epoch": 2.16,
-      "learning_rate": 3.1637024962978064e-05,
-      "loss": 0.272,
+      "epoch": 2.2135726384834395,
+      "grad_norm": 0.1980455219745636,
+      "learning_rate": 3.081235301306596e-05,
+      "loss": 0.4094,
       "step": 61420
     },
     {
-      "epoch": 2.16,
-      "learning_rate": 3.1634278421835116e-05,
-      "loss": 0.2703,
+      "epoch": 2.2137528381446643,
+      "grad_norm": 0.23042725026607513,
+      "learning_rate": 3.080951480310209e-05,
+      "loss": 0.4069,
       "step": 61425
     },
     {
-      "epoch": 2.16,
-      "learning_rate": 3.1631531794550435e-05,
-      "loss": 0.2385,
+      "epoch": 2.213933037805889,
+      "grad_norm": 0.197109192609787,
+      "learning_rate": 3.08066765139847e-05,
+      "loss": 0.3933,
       "step": 61430
     },
     {
-      "epoch": 2.16,
-      "learning_rate": 3.162878508115968e-05,
-      "loss": 0.2926,
+      "epoch": 2.2141132374671137,
+      "grad_norm": 0.19078285992145538,
+      "learning_rate": 3.0803838145752465e-05,
+      "loss": 0.4077,
       "step": 61435
     },
     {
-      "epoch": 2.16,
-      "learning_rate": 3.162603828169852e-05,
-      "loss": 0.2872,
+      "epoch": 2.214293437128338,
+      "grad_norm": 0.20791934430599213,
+      "learning_rate": 3.0800999698444065e-05,
+      "loss": 0.3886,
       "step": 61440
     },
     {
-      "epoch": 2.16,
-      "learning_rate": 3.162329139620262e-05,
-      "loss": 0.2672,
+      "epoch": 2.2144736367895628,
+      "grad_norm": 0.24572139978408813,
+      "learning_rate": 3.0798161172098175e-05,
+      "loss": 0.4195,
       "step": 61445
     },
     {
-      "epoch": 2.16,
-      "learning_rate": 3.162054442470765e-05,
-      "loss": 0.2541,
+      "epoch": 2.2146538364507875,
+      "grad_norm": 0.2402421236038208,
+      "learning_rate": 3.079532256675345e-05,
+      "loss": 0.3967,
       "step": 61450
     },
     {
-      "epoch": 2.16,
-      "learning_rate": 3.1617797367249274e-05,
-      "loss": 0.2762,
+      "epoch": 2.2148340361120122,
+      "grad_norm": 0.20814205706119537,
+      "learning_rate": 3.079248388244858e-05,
+      "loss": 0.4013,
       "step": 61455
     },
     {
-      "epoch": 2.16,
-      "learning_rate": 3.161505022386316e-05,
-      "loss": 0.2837,
+      "epoch": 2.2150142357732365,
+      "grad_norm": 0.17785018682479858,
+      "learning_rate": 3.078964511922224e-05,
+      "loss": 0.4192,
       "step": 61460
     },
     {
-      "epoch": 2.16,
-      "learning_rate": 3.1612302994584976e-05,
-      "loss": 0.2656,
+      "epoch": 2.2151944354344613,
+      "grad_norm": 0.2151138335466385,
+      "learning_rate": 3.0786806277113115e-05,
+      "loss": 0.3994,
       "step": 61465
     },
     {
-      "epoch": 2.16,
-      "learning_rate": 3.16095556794504e-05,
-      "loss": 0.2893,
+      "epoch": 2.215374635095686,
+      "grad_norm": 0.18419857323169708,
+      "learning_rate": 3.0783967356159856e-05,
+      "loss": 0.4021,
       "step": 61470
     },
     {
-      "epoch": 2.16,
-      "learning_rate": 3.1606808278495094e-05,
-      "loss": 0.2888,
+      "epoch": 2.2155548347569107,
+      "grad_norm": 0.18258464336395264,
+      "learning_rate": 3.078112835640117e-05,
+      "loss": 0.3915,
       "step": 61475
     },
     {
-      "epoch": 2.16,
-      "learning_rate": 3.160406079175474e-05,
-      "loss": 0.2957,
+      "epoch": 2.2157350344181355,
+      "grad_norm": 0.2090165615081787,
+      "learning_rate": 3.077828927787574e-05,
+      "loss": 0.3759,
       "step": 61480
     },
     {
-      "epoch": 2.16,
-      "learning_rate": 3.1601313219265e-05,
-      "loss": 0.2729,
+      "epoch": 2.2159152340793598,
+      "grad_norm": 0.21204036474227905,
+      "learning_rate": 3.0775450120622215e-05,
+      "loss": 0.4124,
       "step": 61485
     },
     {
-      "epoch": 2.16,
-      "learning_rate": 3.159856556106157e-05,
-      "loss": 0.2943,
+      "epoch": 2.2160954337405845,
+      "grad_norm": 0.1751544028520584,
+      "learning_rate": 3.077261088467932e-05,
+      "loss": 0.3981,
       "step": 61490
     },
     {
-      "epoch": 2.16,
-      "learning_rate": 3.1595817817180104e-05,
-      "loss": 0.2943,
+      "epoch": 2.216275633401809,
+      "grad_norm": 0.23898641765117645,
+      "learning_rate": 3.076977157008571e-05,
+      "loss": 0.395,
       "step": 61495
     },
     {
-      "epoch": 2.16,
-      "learning_rate": 3.15930699876563e-05,
-      "loss": 0.2764,
+      "epoch": 2.216455833063034,
+      "grad_norm": 0.17369025945663452,
+      "learning_rate": 3.076693217688009e-05,
+      "loss": 0.4193,
       "step": 61500
     },
     {
-      "epoch": 2.16,
-      "eval_loss": 0.26903456449508667,
-      "eval_runtime": 10.5402,
-      "eval_samples_per_second": 9.487,
-      "eval_steps_per_second": 9.487,
+      "epoch": 2.216455833063034,
+      "eval_loss": 0.4364326000213623,
+      "eval_runtime": 3.523,
+      "eval_samples_per_second": 28.385,
+      "eval_steps_per_second": 7.096,
       "step": 61500
     },
     {
-      "epoch": 2.16,
-      "learning_rate": 3.159032207252582e-05,
-      "loss": 0.2633,
+      "epoch": 2.2166360327242582,
+      "grad_norm": 0.22703151404857635,
+      "learning_rate": 3.076409270510112e-05,
+      "loss": 0.3983,
       "step": 61505
     },
     {
-      "epoch": 2.16,
-      "learning_rate": 3.1587574071824367e-05,
-      "loss": 0.2699,
+      "epoch": 2.216816232385483,
+      "grad_norm": 0.21053709089756012,
+      "learning_rate": 3.076125315478752e-05,
+      "loss": 0.3839,
       "step": 61510
     },
     {
-      "epoch": 2.16,
-      "learning_rate": 3.158482598558759e-05,
-      "loss": 0.265,
+      "epoch": 2.2169964320467077,
+      "grad_norm": 0.17118942737579346,
+      "learning_rate": 3.075841352597794e-05,
+      "loss": 0.3901,
       "step": 61515
     },
     {
-      "epoch": 2.16,
-      "learning_rate": 3.158207781385119e-05,
-      "loss": 0.2584,
+      "epoch": 2.2171766317079324,
+      "grad_norm": 0.21972143650054932,
+      "learning_rate": 3.07555738187111e-05,
+      "loss": 0.4219,
       "step": 61520
     },
     {
-      "epoch": 2.16,
-      "learning_rate": 3.157932955665085e-05,
-      "loss": 0.2815,
+      "epoch": 2.217356831369157,
+      "grad_norm": 0.1972840577363968,
+      "learning_rate": 3.0752734033025694e-05,
+      "loss": 0.3747,
       "step": 61525
     },
     {
-      "epoch": 2.16,
-      "learning_rate": 3.157658121402225e-05,
-      "loss": 0.2744,
+      "epoch": 2.2175370310303815,
+      "grad_norm": 0.20041783154010773,
+      "learning_rate": 3.074989416896038e-05,
+      "loss": 0.3963,
       "step": 61530
     },
     {
-      "epoch": 2.16,
-      "learning_rate": 3.157383278600108e-05,
-      "loss": 0.2673,
+      "epoch": 2.217717230691606,
+      "grad_norm": 0.23045070469379425,
+      "learning_rate": 3.0747054226553886e-05,
+      "loss": 0.3746,
       "step": 61535
     },
     {
-      "epoch": 2.17,
-      "learning_rate": 3.157108427262302e-05,
-      "loss": 0.2736,
+      "epoch": 2.217897430352831,
+      "grad_norm": 0.17599715292453766,
+      "learning_rate": 3.0744214205844884e-05,
+      "loss": 0.393,
       "step": 61540
     },
     {
-      "epoch": 2.17,
-      "learning_rate": 3.1568335673923756e-05,
-      "loss": 0.2542,
+      "epoch": 2.2180776300140557,
+      "grad_norm": 0.17511135339736938,
+      "learning_rate": 3.0741374106872084e-05,
+      "loss": 0.4001,
       "step": 61545
     },
     {
-      "epoch": 2.17,
-      "learning_rate": 3.1565586989938996e-05,
-      "loss": 0.2861,
+      "epoch": 2.2182578296752804,
+      "grad_norm": 0.18678990006446838,
+      "learning_rate": 3.0738533929674165e-05,
+      "loss": 0.4134,
       "step": 61550
     },
     {
-      "epoch": 2.17,
-      "learning_rate": 3.15628382207044e-05,
-      "loss": 0.2817,
+      "epoch": 2.2184380293365047,
+      "grad_norm": 0.21317550539970398,
+      "learning_rate": 3.0735693674289826e-05,
+      "loss": 0.4136,
       "step": 61555
     },
     {
-      "epoch": 2.17,
-      "learning_rate": 3.1560089366255684e-05,
-      "loss": 0.2694,
+      "epoch": 2.2186182289977294,
+      "grad_norm": 0.24046799540519714,
+      "learning_rate": 3.073285334075778e-05,
+      "loss": 0.3942,
       "step": 61560
     },
     {
-      "epoch": 2.17,
-      "learning_rate": 3.155734042662853e-05,
-      "loss": 0.2562,
+      "epoch": 2.218798428658954,
+      "grad_norm": 0.19992442429065704,
+      "learning_rate": 3.073001292911672e-05,
+      "loss": 0.3546,
       "step": 61565
     },
     {
-      "epoch": 2.17,
-      "learning_rate": 3.1554591401858634e-05,
-      "loss": 0.2769,
+      "epoch": 2.218978628320179,
+      "grad_norm": 0.2136695832014084,
+      "learning_rate": 3.072717243940534e-05,
+      "loss": 0.407,
       "step": 61570
     },
     {
-      "epoch": 2.17,
-      "learning_rate": 3.155184229198168e-05,
-      "loss": 0.2685,
+      "epoch": 2.2191588279814036,
+      "grad_norm": 0.2068341076374054,
+      "learning_rate": 3.072433187166234e-05,
+      "loss": 0.3849,
       "step": 61575
     },
     {
-      "epoch": 2.17,
-      "learning_rate": 3.154909309703338e-05,
-      "loss": 0.2839,
+      "epoch": 2.219339027642628,
+      "grad_norm": 0.2191094160079956,
+      "learning_rate": 3.072149122592643e-05,
+      "loss": 0.4173,
       "step": 61580
     },
     {
-      "epoch": 2.17,
-      "learning_rate": 3.154634381704942e-05,
-      "loss": 0.2619,
+      "epoch": 2.2195192273038526,
+      "grad_norm": 0.20730826258659363,
+      "learning_rate": 3.071865050223631e-05,
+      "loss": 0.3846,
       "step": 61585
     },
     {
-      "epoch": 2.17,
-      "learning_rate": 3.15435944520655e-05,
-      "loss": 0.2865,
+      "epoch": 2.2196994269650774,
+      "grad_norm": 0.20395426452159882,
+      "learning_rate": 3.071580970063068e-05,
+      "loss": 0.3963,
       "step": 61590
     },
     {
-      "epoch": 2.17,
-      "learning_rate": 3.154084500211732e-05,
-      "loss": 0.2774,
+      "epoch": 2.219879626626302,
+      "grad_norm": 0.19093595445156097,
+      "learning_rate": 3.071296882114826e-05,
+      "loss": 0.39,
       "step": 61595
     },
     {
-      "epoch": 2.17,
-      "learning_rate": 3.1538095467240574e-05,
-      "loss": 0.2769,
+      "epoch": 2.2200598262875264,
+      "grad_norm": 0.2352960854768753,
+      "learning_rate": 3.071012786382774e-05,
+      "loss": 0.4021,
       "step": 61600
     },
     {
-      "epoch": 2.17,
-      "learning_rate": 3.153534584747097e-05,
-      "loss": 0.2741,
+      "epoch": 2.220240025948751,
+      "grad_norm": 0.21295884251594543,
+      "learning_rate": 3.0707286828707835e-05,
+      "loss": 0.3812,
       "step": 61605
     },
     {
-      "epoch": 2.17,
-      "learning_rate": 3.1532596142844205e-05,
-      "loss": 0.2868,
+      "epoch": 2.220420225609976,
+      "grad_norm": 0.19988127052783966,
+      "learning_rate": 3.0704445715827246e-05,
+      "loss": 0.4401,
       "step": 61610
     },
     {
-      "epoch": 2.17,
-      "learning_rate": 3.1529846353395986e-05,
-      "loss": 0.2889,
+      "epoch": 2.2206004252712006,
+      "grad_norm": 0.19108517467975616,
+      "learning_rate": 3.0701604525224695e-05,
+      "loss": 0.3947,
       "step": 61615
     },
     {
-      "epoch": 2.17,
-      "learning_rate": 3.152709647916202e-05,
-      "loss": 0.2689,
+      "epoch": 2.2207806249324253,
+      "grad_norm": 0.21383409202098846,
+      "learning_rate": 3.0698763256938876e-05,
+      "loss": 0.3947,
       "step": 61620
     },
     {
-      "epoch": 2.17,
-      "learning_rate": 3.152434652017801e-05,
-      "loss": 0.2735,
+      "epoch": 2.2209608245936496,
+      "grad_norm": 0.201728954911232,
+      "learning_rate": 3.069592191100852e-05,
+      "loss": 0.3971,
       "step": 61625
     },
     {
-      "epoch": 2.17,
-      "learning_rate": 3.152159647647964e-05,
-      "loss": 0.2756,
+      "epoch": 2.2211410242548744,
+      "grad_norm": 0.20201417803764343,
+      "learning_rate": 3.069308048747233e-05,
+      "loss": 0.3725,
       "step": 61630
     },
     {
-      "epoch": 2.17,
-      "learning_rate": 3.151884634810266e-05,
-      "loss": 0.2433,
+      "epoch": 2.221321223916099,
+      "grad_norm": 0.19340789318084717,
+      "learning_rate": 3.069023898636902e-05,
+      "loss": 0.3866,
       "step": 61635
     },
     {
-      "epoch": 2.17,
-      "learning_rate": 3.151609613508274e-05,
-      "loss": 0.2631,
+      "epoch": 2.221501423577324,
+      "grad_norm": 0.22743280231952667,
+      "learning_rate": 3.0687397407737306e-05,
+      "loss": 0.3558,
       "step": 61640
     },
     {
-      "epoch": 2.17,
-      "learning_rate": 3.151334583745562e-05,
-      "loss": 0.2759,
+      "epoch": 2.221681623238548,
+      "grad_norm": 0.18240687251091003,
+      "learning_rate": 3.06845557516159e-05,
+      "loss": 0.4093,
       "step": 61645
     },
     {
-      "epoch": 2.17,
-      "learning_rate": 3.151059545525698e-05,
-      "loss": 0.2369,
+      "epoch": 2.221861822899773,
+      "grad_norm": 0.22113995254039764,
+      "learning_rate": 3.068171401804353e-05,
+      "loss": 0.3851,
       "step": 61650
     },
     {
-      "epoch": 2.17,
-      "learning_rate": 3.150784498852256e-05,
-      "loss": 0.2596,
+      "epoch": 2.2220420225609976,
+      "grad_norm": 0.22573597729206085,
+      "learning_rate": 3.0678872207058903e-05,
+      "loss": 0.4202,
       "step": 61655
     },
     {
-      "epoch": 2.17,
-      "learning_rate": 3.150509443728806e-05,
-      "loss": 0.286,
+      "epoch": 2.2222222222222223,
+      "grad_norm": 0.1990000456571579,
+      "learning_rate": 3.067603031870074e-05,
+      "loss": 0.4169,
       "step": 61660
     },
     {
-      "epoch": 2.17,
-      "learning_rate": 3.15023438015892e-05,
-      "loss": 0.28,
+      "epoch": 2.222402421883447,
+      "grad_norm": 0.18506182730197906,
+      "learning_rate": 3.067318835300777e-05,
+      "loss": 0.3866,
       "step": 61665
     },
     {
-      "epoch": 2.17,
-      "learning_rate": 3.149959308146168e-05,
-      "loss": 0.2702,
+      "epoch": 2.2225826215446713,
+      "grad_norm": 0.16143712401390076,
+      "learning_rate": 3.067034631001869e-05,
+      "loss": 0.3977,
       "step": 61670
     },
     {
-      "epoch": 2.17,
-      "learning_rate": 3.149684227694123e-05,
-      "loss": 0.2794,
+      "epoch": 2.222762821205896,
+      "grad_norm": 0.2105492502450943,
+      "learning_rate": 3.066750418977225e-05,
+      "loss": 0.3876,
       "step": 61675
     },
     {
-      "epoch": 2.17,
-      "learning_rate": 3.1494091388063573e-05,
-      "loss": 0.278,
+      "epoch": 2.222943020867121,
+      "grad_norm": 0.2138456106185913,
+      "learning_rate": 3.066466199230716e-05,
+      "loss": 0.4023,
       "step": 61680
     },
     {
-      "epoch": 2.17,
-      "learning_rate": 3.149134041486441e-05,
-      "loss": 0.2791,
+      "epoch": 2.2231232205283455,
+      "grad_norm": 0.2341010868549347,
+      "learning_rate": 3.066181971766215e-05,
+      "loss": 0.394,
       "step": 61685
     },
     {
-      "epoch": 2.17,
-      "learning_rate": 3.1488589357379476e-05,
-      "loss": 0.2698,
+      "epoch": 2.22330342018957,
+      "grad_norm": 0.2601710557937622,
+      "learning_rate": 3.065897736587594e-05,
+      "loss": 0.36,
       "step": 61690
     },
     {
-      "epoch": 2.17,
-      "learning_rate": 3.148583821564448e-05,
-      "loss": 0.2683,
+      "epoch": 2.2234836198507946,
+      "grad_norm": 0.20886510610580444,
+      "learning_rate": 3.065613493698726e-05,
+      "loss": 0.4128,
       "step": 61695
     },
     {
-      "epoch": 2.17,
-      "learning_rate": 3.148308698969516e-05,
-      "loss": 0.2948,
+      "epoch": 2.2236638195120193,
+      "grad_norm": 0.21699224412441254,
+      "learning_rate": 3.0653292431034845e-05,
+      "loss": 0.385,
       "step": 61700
     },
     {
-      "epoch": 2.17,
-      "learning_rate": 3.148033567956722e-05,
-      "loss": 0.2722,
+      "epoch": 2.223844019173244,
+      "grad_norm": 0.22566765546798706,
+      "learning_rate": 3.0650449848057404e-05,
+      "loss": 0.3951,
       "step": 61705
     },
     {
-      "epoch": 2.17,
-      "learning_rate": 3.1477584285296385e-05,
-      "loss": 0.2688,
+      "epoch": 2.2240242188344688,
+      "grad_norm": 0.16451287269592285,
+      "learning_rate": 3.064760718809368e-05,
+      "loss": 0.3828,
       "step": 61710
     },
     {
-      "epoch": 2.17,
-      "learning_rate": 3.1474832806918395e-05,
-      "loss": 0.2561,
+      "epoch": 2.224204418495693,
+      "grad_norm": 0.2075955718755722,
+      "learning_rate": 3.06447644511824e-05,
+      "loss": 0.3986,
       "step": 61715
     },
     {
-      "epoch": 2.17,
-      "learning_rate": 3.147208124446897e-05,
-      "loss": 0.257,
+      "epoch": 2.224384618156918,
+      "grad_norm": 0.1905994862318039,
+      "learning_rate": 3.06419216373623e-05,
+      "loss": 0.4061,
       "step": 61720
     },
     {
-      "epoch": 2.17,
-      "learning_rate": 3.1469329597983834e-05,
-      "loss": 0.249,
+      "epoch": 2.2245648178181425,
+      "grad_norm": 0.2250581681728363,
+      "learning_rate": 3.063907874667211e-05,
+      "loss": 0.4116,
       "step": 61725
     },
     {
-      "epoch": 2.17,
-      "learning_rate": 3.146657786749871e-05,
-      "loss": 0.256,
+      "epoch": 2.2247450174793673,
+      "grad_norm": 0.20644237101078033,
+      "learning_rate": 3.063623577915056e-05,
+      "loss": 0.3753,
       "step": 61730
     },
     {
-      "epoch": 2.17,
-      "learning_rate": 3.146382605304935e-05,
-      "loss": 0.2751,
+      "epoch": 2.2249252171405915,
+      "grad_norm": 0.2077893614768982,
+      "learning_rate": 3.0633392734836395e-05,
+      "loss": 0.3853,
       "step": 61735
     },
     {
-      "epoch": 2.17,
-      "learning_rate": 3.146107415467145e-05,
-      "loss": 0.2642,
+      "epoch": 2.2251054168018163,
+      "grad_norm": 0.18702587485313416,
+      "learning_rate": 3.063054961376834e-05,
+      "loss": 0.412,
       "step": 61740
     },
     {
-      "epoch": 2.17,
-      "learning_rate": 3.1458322172400766e-05,
-      "loss": 0.2778,
+      "epoch": 2.225285616463041,
+      "grad_norm": 0.20379160344600677,
+      "learning_rate": 3.0627706415985146e-05,
+      "loss": 0.4043,
       "step": 61745
     },
     {
-      "epoch": 2.17,
-      "learning_rate": 3.145557010627303e-05,
-      "loss": 0.2913,
+      "epoch": 2.2254658161242657,
+      "grad_norm": 0.17889107763767242,
+      "learning_rate": 3.062486314152553e-05,
+      "loss": 0.4084,
       "step": 61750
     },
     {
-      "epoch": 2.17,
-      "learning_rate": 3.1452817956323967e-05,
-      "loss": 0.2724,
+      "epoch": 2.2256460157854905,
+      "grad_norm": 0.23296573758125305,
+      "learning_rate": 3.0622019790428254e-05,
+      "loss": 0.4243,
       "step": 61755
     },
     {
-      "epoch": 2.17,
-      "learning_rate": 3.1450065722589316e-05,
-      "loss": 0.2712,
+      "epoch": 2.2258262154467148,
+      "grad_norm": 0.22724519670009613,
+      "learning_rate": 3.0619176362732035e-05,
+      "loss": 0.4073,
       "step": 61760
     },
     {
-      "epoch": 2.17,
-      "learning_rate": 3.1447313405104805e-05,
-      "loss": 0.2703,
+      "epoch": 2.2260064151079395,
+      "grad_norm": 0.27276527881622314,
+      "learning_rate": 3.0616332858475625e-05,
+      "loss": 0.4086,
       "step": 61765
     },
     {
-      "epoch": 2.17,
-      "learning_rate": 3.144456100390619e-05,
-      "loss": 0.2594,
+      "epoch": 2.2261866147691642,
+      "grad_norm": 0.21575415134429932,
+      "learning_rate": 3.0613489277697784e-05,
+      "loss": 0.4389,
       "step": 61770
     },
     {
-      "epoch": 2.17,
-      "learning_rate": 3.144180851902919e-05,
-      "loss": 0.2719,
+      "epoch": 2.226366814430389,
+      "grad_norm": 0.19603478908538818,
+      "learning_rate": 3.061064562043722e-05,
+      "loss": 0.4233,
       "step": 61775
     },
     {
-      "epoch": 2.17,
-      "learning_rate": 3.143905595050956e-05,
-      "loss": 0.2843,
+      "epoch": 2.2265470140916137,
+      "grad_norm": 0.2517644464969635,
+      "learning_rate": 3.06078018867327e-05,
+      "loss": 0.4004,
       "step": 61780
     },
     {
-      "epoch": 2.17,
-      "learning_rate": 3.1436303298383027e-05,
-      "loss": 0.2774,
+      "epoch": 2.226727213752838,
+      "grad_norm": 0.2613866627216339,
+      "learning_rate": 3.060495807662297e-05,
+      "loss": 0.3955,
       "step": 61785
     },
     {
-      "epoch": 2.17,
-      "learning_rate": 3.1433550562685335e-05,
-      "loss": 0.2669,
+      "epoch": 2.2269074134140627,
+      "grad_norm": 0.18380023539066315,
+      "learning_rate": 3.0602114190146775e-05,
+      "loss": 0.3754,
       "step": 61790
     },
     {
-      "epoch": 2.17,
-      "learning_rate": 3.143079774345223e-05,
-      "loss": 0.2658,
+      "epoch": 2.2270876130752875,
+      "grad_norm": 0.1809166818857193,
+      "learning_rate": 3.0599270227342845e-05,
+      "loss": 0.3961,
       "step": 61795
     },
     {
-      "epoch": 2.17,
-      "learning_rate": 3.142804484071945e-05,
-      "loss": 0.2773,
+      "epoch": 2.227267812736512,
+      "grad_norm": 0.1867736577987671,
+      "learning_rate": 3.059642618824996e-05,
+      "loss": 0.3826,
       "step": 61800
     },
     {
-      "epoch": 2.17,
-      "learning_rate": 3.142529185452275e-05,
-      "loss": 0.2765,
+      "epoch": 2.2274480123977365,
+      "grad_norm": 0.24860471487045288,
+      "learning_rate": 3.059358207290683e-05,
+      "loss": 0.4294,
       "step": 61805
     },
     {
-      "epoch": 2.17,
-      "learning_rate": 3.142253878489787e-05,
-      "loss": 0.2758,
+      "epoch": 2.227628212058961,
+      "grad_norm": 0.2238049954175949,
+      "learning_rate": 3.059073788135224e-05,
+      "loss": 0.44,
       "step": 61810
     },
     {
-      "epoch": 2.17,
-      "learning_rate": 3.141978563188056e-05,
-      "loss": 0.2643,
+      "epoch": 2.227808411720186,
+      "grad_norm": 0.1938982456922531,
+      "learning_rate": 3.0587893613624934e-05,
+      "loss": 0.3959,
       "step": 61815
     },
     {
-      "epoch": 2.17,
-      "learning_rate": 3.141703239550656e-05,
-      "loss": 0.2776,
+      "epoch": 2.2279886113814107,
+      "grad_norm": 0.1805550903081894,
+      "learning_rate": 3.058504926976365e-05,
+      "loss": 0.3997,
       "step": 61820
     },
     {
-      "epoch": 2.18,
-      "learning_rate": 3.141427907581162e-05,
-      "loss": 0.2638,
+      "epoch": 2.2281688110426354,
+      "grad_norm": 0.2023649662733078,
+      "learning_rate": 3.0582204849807154e-05,
+      "loss": 0.4009,
       "step": 61825
     },
     {
-      "epoch": 2.18,
-      "learning_rate": 3.141152567283149e-05,
-      "loss": 0.2818,
+      "epoch": 2.2283490107038597,
+      "grad_norm": 0.18612384796142578,
+      "learning_rate": 3.05793603537942e-05,
+      "loss": 0.4259,
       "step": 61830
     },
     {
-      "epoch": 2.18,
-      "learning_rate": 3.140877218660194e-05,
-      "loss": 0.2655,
+      "epoch": 2.2285292103650844,
+      "grad_norm": 0.22503447532653809,
+      "learning_rate": 3.0576515781763545e-05,
+      "loss": 0.4017,
       "step": 61835
     },
     {
-      "epoch": 2.18,
-      "learning_rate": 3.1406018617158695e-05,
-      "loss": 0.2551,
+      "epoch": 2.228709410026309,
+      "grad_norm": 0.2309921830892563,
+      "learning_rate": 3.057367113375393e-05,
+      "loss": 0.3924,
       "step": 61840
     },
     {
-      "epoch": 2.18,
-      "learning_rate": 3.1403264964537525e-05,
-      "loss": 0.2567,
+      "epoch": 2.228889609687534,
+      "grad_norm": 0.18473617732524872,
+      "learning_rate": 3.0570826409804135e-05,
+      "loss": 0.3755,
       "step": 61845
     },
     {
-      "epoch": 2.18,
-      "learning_rate": 3.140051122877417e-05,
-      "loss": 0.2728,
+      "epoch": 2.2290698093487586,
+      "grad_norm": 0.19264580309391022,
+      "learning_rate": 3.056798160995291e-05,
+      "loss": 0.4029,
       "step": 61850
     },
     {
-      "epoch": 2.18,
-      "learning_rate": 3.1397757409904405e-05,
-      "loss": 0.2919,
+      "epoch": 2.229250009009983,
+      "grad_norm": 0.21103538572788239,
+      "learning_rate": 3.0565136734239006e-05,
+      "loss": 0.4169,
       "step": 61855
     },
     {
-      "epoch": 2.18,
-      "learning_rate": 3.139500350796397e-05,
-      "loss": 0.2607,
+      "epoch": 2.2294302086712077,
+      "grad_norm": 0.21922504901885986,
+      "learning_rate": 3.0562291782701194e-05,
+      "loss": 0.3899,
       "step": 61860
     },
     {
-      "epoch": 2.18,
-      "learning_rate": 3.1392249522988636e-05,
-      "loss": 0.2678,
+      "epoch": 2.2296104083324324,
+      "grad_norm": 0.19922494888305664,
+      "learning_rate": 3.0559446755378226e-05,
+      "loss": 0.3969,
       "step": 61865
     },
     {
-      "epoch": 2.18,
-      "learning_rate": 3.1389495455014154e-05,
-      "loss": 0.2838,
+      "epoch": 2.229790607993657,
+      "grad_norm": 0.2152629941701889,
+      "learning_rate": 3.055660165230888e-05,
+      "loss": 0.4128,
       "step": 61870
     },
     {
-      "epoch": 2.18,
-      "learning_rate": 3.138674130407628e-05,
-      "loss": 0.265,
+      "epoch": 2.2299708076548814,
+      "grad_norm": 0.19293451309204102,
+      "learning_rate": 3.055375647353191e-05,
+      "loss": 0.3871,
       "step": 61875
     },
     {
-      "epoch": 2.18,
-      "learning_rate": 3.138398707021077e-05,
-      "loss": 0.2934,
+      "epoch": 2.230151007316106,
+      "grad_norm": 0.2509141266345978,
+      "learning_rate": 3.0550911219086084e-05,
+      "loss": 0.4476,
       "step": 61880
     },
     {
-      "epoch": 2.18,
-      "learning_rate": 3.1381232753453416e-05,
-      "loss": 0.2583,
+      "epoch": 2.230331206977331,
+      "grad_norm": 0.24453610181808472,
+      "learning_rate": 3.0548065889010176e-05,
+      "loss": 0.3912,
       "step": 61885
     },
     {
-      "epoch": 2.18,
-      "learning_rate": 3.137847835383995e-05,
-      "loss": 0.2492,
+      "epoch": 2.2305114066385556,
+      "grad_norm": 0.24755702912807465,
+      "learning_rate": 3.054522048334294e-05,
+      "loss": 0.3986,
       "step": 61890
     },
     {
-      "epoch": 2.18,
-      "learning_rate": 3.1375723871406146e-05,
-      "loss": 0.2666,
+      "epoch": 2.2306916062997804,
+      "grad_norm": 0.1772449016571045,
+      "learning_rate": 3.0542375002123145e-05,
+      "loss": 0.3765,
       "step": 61895
     },
     {
-      "epoch": 2.18,
-      "learning_rate": 3.137296930618777e-05,
-      "loss": 0.2915,
+      "epoch": 2.2308718059610047,
+      "grad_norm": 0.1797766089439392,
+      "learning_rate": 3.053952944538956e-05,
+      "loss": 0.3711,
       "step": 61900
     },
     {
-      "epoch": 2.18,
-      "learning_rate": 3.1370214658220574e-05,
-      "loss": 0.2795,
+      "epoch": 2.2310520056222294,
+      "grad_norm": 0.21006938815116882,
+      "learning_rate": 3.0536683813180966e-05,
+      "loss": 0.4086,
       "step": 61905
     },
     {
-      "epoch": 2.18,
-      "learning_rate": 3.1367459927540354e-05,
-      "loss": 0.2567,
+      "epoch": 2.231232205283454,
+      "grad_norm": 0.18526072800159454,
+      "learning_rate": 3.053383810553613e-05,
+      "loss": 0.3676,
       "step": 61910
     },
     {
-      "epoch": 2.18,
-      "learning_rate": 3.1364705114182855e-05,
-      "loss": 0.2675,
+      "epoch": 2.231412404944679,
+      "grad_norm": 0.20477519929409027,
+      "learning_rate": 3.053099232249381e-05,
+      "loss": 0.4048,
       "step": 61915
     },
     {
-      "epoch": 2.18,
-      "learning_rate": 3.136195021818387e-05,
-      "loss": 0.2815,
+      "epoch": 2.231592604605903,
+      "grad_norm": 0.17895658314228058,
+      "learning_rate": 3.05281464640928e-05,
+      "loss": 0.4049,
       "step": 61920
     },
     {
-      "epoch": 2.18,
-      "learning_rate": 3.1359195239579145e-05,
-      "loss": 0.2679,
+      "epoch": 2.231772804267128,
+      "grad_norm": 0.20034639537334442,
+      "learning_rate": 3.0525300530371846e-05,
+      "loss": 0.4089,
       "step": 61925
     },
     {
-      "epoch": 2.18,
-      "learning_rate": 3.135644017840445e-05,
-      "loss": 0.2546,
+      "epoch": 2.2319530039283526,
+      "grad_norm": 0.24000723659992218,
+      "learning_rate": 3.052245452136977e-05,
+      "loss": 0.3823,
       "step": 61930
     },
     {
-      "epoch": 2.18,
-      "learning_rate": 3.135368503469558e-05,
-      "loss": 0.2836,
+      "epoch": 2.2321332035895773,
+      "grad_norm": 0.20230047404766083,
+      "learning_rate": 3.0519608437125305e-05,
+      "loss": 0.4144,
       "step": 61935
     },
     {
-      "epoch": 2.18,
-      "learning_rate": 3.13509298084883e-05,
-      "loss": 0.2724,
+      "epoch": 2.232313403250802,
+      "grad_norm": 0.16788794100284576,
+      "learning_rate": 3.0516762277677243e-05,
+      "loss": 0.3911,
       "step": 61940
     },
     {
-      "epoch": 2.18,
-      "learning_rate": 3.1348174499818375e-05,
-      "loss": 0.2578,
+      "epoch": 2.2324936029120264,
+      "grad_norm": 0.1820405274629593,
+      "learning_rate": 3.051391604306436e-05,
+      "loss": 0.3892,
       "step": 61945
     },
     {
-      "epoch": 2.18,
-      "learning_rate": 3.134541910872159e-05,
-      "loss": 0.2742,
+      "epoch": 2.232673802573251,
+      "grad_norm": 0.24724507331848145,
+      "learning_rate": 3.0511069733325447e-05,
+      "loss": 0.376,
       "step": 61950
     },
     {
-      "epoch": 2.18,
-      "learning_rate": 3.1342663635233713e-05,
-      "loss": 0.267,
+      "epoch": 2.232854002234476,
+      "grad_norm": 0.21663494408130646,
+      "learning_rate": 3.0508223348499275e-05,
+      "loss": 0.393,
       "step": 61955
     },
     {
-      "epoch": 2.18,
-      "learning_rate": 3.133990807939053e-05,
-      "loss": 0.2694,
+      "epoch": 2.2330342018957006,
+      "grad_norm": 0.17999057471752167,
+      "learning_rate": 3.0505376888624622e-05,
+      "loss": 0.4099,
       "step": 61960
     },
     {
-      "epoch": 2.18,
-      "learning_rate": 3.133715244122783e-05,
-      "loss": 0.2745,
+      "epoch": 2.233214401556925,
+      "grad_norm": 0.15646670758724213,
+      "learning_rate": 3.0502530353740278e-05,
+      "loss": 0.3565,
       "step": 61965
     },
     {
-      "epoch": 2.18,
-      "learning_rate": 3.133439672078138e-05,
-      "loss": 0.2624,
+      "epoch": 2.2333946012181496,
+      "grad_norm": 0.1919204592704773,
+      "learning_rate": 3.049968374388502e-05,
+      "loss": 0.4127,
       "step": 61970
     },
     {
-      "epoch": 2.18,
-      "learning_rate": 3.1331640918086956e-05,
-      "loss": 0.2828,
+      "epoch": 2.2335748008793743,
+      "grad_norm": 0.1685272455215454,
+      "learning_rate": 3.0496837059097645e-05,
+      "loss": 0.4332,
       "step": 61975
     },
     {
-      "epoch": 2.18,
-      "learning_rate": 3.132888503318036e-05,
-      "loss": 0.2696,
+      "epoch": 2.233755000540599,
+      "grad_norm": 0.1549319177865982,
+      "learning_rate": 3.0493990299416913e-05,
+      "loss": 0.4114,
       "step": 61980
     },
     {
-      "epoch": 2.18,
-      "learning_rate": 3.132612906609734e-05,
-      "loss": 0.2563,
+      "epoch": 2.233935200201824,
+      "grad_norm": 0.22232964634895325,
+      "learning_rate": 3.0491143464881634e-05,
+      "loss": 0.3857,
       "step": 61985
     },
     {
-      "epoch": 2.18,
-      "learning_rate": 3.1323373016873715e-05,
-      "loss": 0.2596,
+      "epoch": 2.234115399863048,
+      "grad_norm": 0.14861559867858887,
+      "learning_rate": 3.0488296555530594e-05,
+      "loss": 0.3871,
       "step": 61990
     },
     {
-      "epoch": 2.18,
-      "learning_rate": 3.132061688554526e-05,
-      "loss": 0.2638,
+      "epoch": 2.234295599524273,
+      "grad_norm": 0.20374061167240143,
+      "learning_rate": 3.0485449571402573e-05,
+      "loss": 0.3995,
       "step": 61995
     },
     {
-      "epoch": 2.18,
-      "learning_rate": 3.131786067214777e-05,
-      "loss": 0.29,
+      "epoch": 2.2344757991854975,
+      "grad_norm": 0.21584060788154602,
+      "learning_rate": 3.048260251253636e-05,
+      "loss": 0.4383,
       "step": 62000
     },
     {
-      "epoch": 2.18,
-      "eval_loss": 0.2678808569908142,
-      "eval_runtime": 10.5389,
-      "eval_samples_per_second": 9.489,
-      "eval_steps_per_second": 9.489,
+      "epoch": 2.2344757991854975,
+      "eval_loss": 0.43634411692619324,
+      "eval_runtime": 3.5361,
+      "eval_samples_per_second": 28.28,
+      "eval_steps_per_second": 7.07,
       "step": 62000
     },
     {
-      "epoch": 2.18,
-      "learning_rate": 3.1315104376717e-05,
-      "loss": 0.3023,
+      "epoch": 2.2346559988467223,
+      "grad_norm": 0.18200920522212982,
+      "learning_rate": 3.0479755378970753e-05,
+      "loss": 0.4092,
       "step": 62005
     },
     {
-      "epoch": 2.18,
-      "learning_rate": 3.1312347999288777e-05,
-      "loss": 0.2886,
+      "epoch": 2.2348361985079466,
+      "grad_norm": 0.18603946268558502,
+      "learning_rate": 3.047690817074454e-05,
+      "loss": 0.3547,
       "step": 62010
     },
     {
-      "epoch": 2.18,
-      "learning_rate": 3.1309591539898866e-05,
-      "loss": 0.2419,
+      "epoch": 2.2350163981691713,
+      "grad_norm": 0.17530210316181183,
+      "learning_rate": 3.0474060887896518e-05,
+      "loss": 0.3952,
       "step": 62015
     },
     {
-      "epoch": 2.18,
-      "learning_rate": 3.130683499858308e-05,
-      "loss": 0.2869,
+      "epoch": 2.235196597830396,
+      "grad_norm": 0.1511596292257309,
+      "learning_rate": 3.0471213530465465e-05,
+      "loss": 0.3703,
       "step": 62020
     },
     {
-      "epoch": 2.18,
-      "learning_rate": 3.130407837537719e-05,
-      "loss": 0.2772,
+      "epoch": 2.2353767974916208,
+      "grad_norm": 0.22197118401527405,
+      "learning_rate": 3.046836609849019e-05,
+      "loss": 0.4399,
       "step": 62025
     },
     {
-      "epoch": 2.18,
-      "learning_rate": 3.1301321670316994e-05,
-      "loss": 0.2674,
+      "epoch": 2.2355569971528455,
+      "grad_norm": 0.2044990509748459,
+      "learning_rate": 3.046551859200949e-05,
+      "loss": 0.3614,
       "step": 62030
     },
     {
-      "epoch": 2.18,
-      "learning_rate": 3.1298564883438295e-05,
-      "loss": 0.2838,
+      "epoch": 2.23573719681407,
+      "grad_norm": 0.2116348147392273,
+      "learning_rate": 3.0462671011062162e-05,
+      "loss": 0.3879,
       "step": 62035
     },
     {
-      "epoch": 2.18,
-      "learning_rate": 3.1295808014776884e-05,
-      "loss": 0.2683,
+      "epoch": 2.2359173964752945,
+      "grad_norm": 0.18771792948246002,
+      "learning_rate": 3.045982335568699e-05,
+      "loss": 0.4121,
       "step": 62040
     },
     {
-      "epoch": 2.18,
-      "learning_rate": 3.129305106436855e-05,
-      "loss": 0.2963,
+      "epoch": 2.2360975961365193,
+      "grad_norm": 0.2089640349149704,
+      "learning_rate": 3.0456975625922784e-05,
+      "loss": 0.3884,
       "step": 62045
     },
     {
-      "epoch": 2.18,
-      "learning_rate": 3.129029403224909e-05,
-      "loss": 0.2604,
+      "epoch": 2.236277795797744,
+      "grad_norm": 0.1947159469127655,
+      "learning_rate": 3.0454127821808337e-05,
+      "loss": 0.3953,
       "step": 62050
     },
     {
-      "epoch": 2.18,
-      "learning_rate": 3.128753691845432e-05,
-      "loss": 0.2684,
+      "epoch": 2.2364579954589687,
+      "grad_norm": 0.23277081549167633,
+      "learning_rate": 3.045127994338246e-05,
+      "loss": 0.397,
       "step": 62055
     },
     {
-      "epoch": 2.18,
-      "learning_rate": 3.128477972302001e-05,
-      "loss": 0.2561,
+      "epoch": 2.236638195120193,
+      "grad_norm": 0.17750267684459686,
+      "learning_rate": 3.0448431990683958e-05,
+      "loss": 0.3889,
       "step": 62060
     },
     {
-      "epoch": 2.18,
-      "learning_rate": 3.128202244598199e-05,
-      "loss": 0.2732,
+      "epoch": 2.2368183947814178,
+      "grad_norm": 0.16454583406448364,
+      "learning_rate": 3.0445583963751606e-05,
+      "loss": 0.3797,
       "step": 62065
     },
     {
-      "epoch": 2.18,
-      "learning_rate": 3.1279265087376044e-05,
-      "loss": 0.2673,
+      "epoch": 2.2369985944426425,
+      "grad_norm": 0.20463277399539948,
+      "learning_rate": 3.044273586262424e-05,
+      "loss": 0.3917,
       "step": 62070
     },
     {
-      "epoch": 2.18,
-      "learning_rate": 3.1276507647237984e-05,
-      "loss": 0.2787,
+      "epoch": 2.237178794103867,
+      "grad_norm": 0.20936442911624908,
+      "learning_rate": 3.043988768734065e-05,
+      "loss": 0.4067,
       "step": 62075
     },
     {
-      "epoch": 2.18,
-      "learning_rate": 3.1273750125603604e-05,
-      "loss": 0.2969,
+      "epoch": 2.237358993765092,
+      "grad_norm": 0.20374515652656555,
+      "learning_rate": 3.043703943793964e-05,
+      "loss": 0.392,
       "step": 62080
     },
     {
-      "epoch": 2.18,
-      "learning_rate": 3.127099252250872e-05,
-      "loss": 0.2773,
+      "epoch": 2.2375391934263162,
+      "grad_norm": 0.20407363772392273,
+      "learning_rate": 3.043419111446002e-05,
+      "loss": 0.3983,
       "step": 62085
     },
     {
-      "epoch": 2.18,
-      "learning_rate": 3.126823483798912e-05,
-      "loss": 0.251,
+      "epoch": 2.237719393087541,
+      "grad_norm": 0.19166499376296997,
+      "learning_rate": 3.0431342716940602e-05,
+      "loss": 0.396,
       "step": 62090
     },
     {
-      "epoch": 2.18,
-      "learning_rate": 3.126547707208063e-05,
-      "loss": 0.27,
+      "epoch": 2.2378995927487657,
+      "grad_norm": 0.22127671539783478,
+      "learning_rate": 3.0428494245420192e-05,
+      "loss": 0.3952,
       "step": 62095
     },
     {
-      "epoch": 2.18,
-      "learning_rate": 3.1262719224819044e-05,
-      "loss": 0.2743,
+      "epoch": 2.2380797924099904,
+      "grad_norm": 0.19556500017642975,
+      "learning_rate": 3.0425645699937594e-05,
+      "loss": 0.4074,
       "step": 62100
     },
     {
-      "epoch": 2.19,
-      "learning_rate": 3.125996129624018e-05,
-      "loss": 0.2802,
+      "epoch": 2.2382599920712147,
+      "grad_norm": 0.18051710724830627,
+      "learning_rate": 3.0422797080531628e-05,
+      "loss": 0.3663,
       "step": 62105
     },
     {
-      "epoch": 2.19,
-      "learning_rate": 3.1257203286379845e-05,
-      "loss": 0.2797,
+      "epoch": 2.2384401917324395,
+      "grad_norm": 0.1607939898967743,
+      "learning_rate": 3.0419948387241093e-05,
+      "loss": 0.3728,
       "step": 62110
     },
     {
-      "epoch": 2.19,
-      "learning_rate": 3.1254445195273844e-05,
-      "loss": 0.2787,
+      "epoch": 2.238620391393664,
+      "grad_norm": 0.21384070813655853,
+      "learning_rate": 3.0417099620104826e-05,
+      "loss": 0.4021,
       "step": 62115
     },
     {
-      "epoch": 2.19,
-      "learning_rate": 3.1251687022958e-05,
-      "loss": 0.266,
+      "epoch": 2.238800591054889,
+      "grad_norm": 0.16077342629432678,
+      "learning_rate": 3.041425077916161e-05,
+      "loss": 0.3476,
       "step": 62120
     },
     {
-      "epoch": 2.19,
-      "learning_rate": 3.124892876946811e-05,
-      "loss": 0.2657,
+      "epoch": 2.2389807907161137,
+      "grad_norm": 0.20714950561523438,
+      "learning_rate": 3.0411401864450283e-05,
+      "loss": 0.402,
       "step": 62125
     },
     {
-      "epoch": 2.19,
-      "learning_rate": 3.124617043484001e-05,
-      "loss": 0.2578,
+      "epoch": 2.239160990377338,
+      "grad_norm": 0.2111266702413559,
+      "learning_rate": 3.040855287600965e-05,
+      "loss": 0.4138,
       "step": 62130
     },
     {
-      "epoch": 2.19,
-      "learning_rate": 3.1243412019109496e-05,
-      "loss": 0.2822,
+      "epoch": 2.2393411900385627,
+      "grad_norm": 0.21749116480350494,
+      "learning_rate": 3.0405703813878534e-05,
+      "loss": 0.3802,
       "step": 62135
     },
     {
-      "epoch": 2.19,
-      "learning_rate": 3.124065352231239e-05,
-      "loss": 0.2852,
+      "epoch": 2.2395213896997874,
+      "grad_norm": 0.2501288056373596,
+      "learning_rate": 3.0402854678095754e-05,
+      "loss": 0.4045,
       "step": 62140
     },
     {
-      "epoch": 2.19,
-      "learning_rate": 3.123789494448451e-05,
-      "loss": 0.2712,
+      "epoch": 2.239701589361012,
+      "grad_norm": 0.18549959361553192,
+      "learning_rate": 3.0400005468700115e-05,
+      "loss": 0.4179,
       "step": 62145
     },
     {
-      "epoch": 2.19,
-      "learning_rate": 3.123513628566168e-05,
-      "loss": 0.273,
+      "epoch": 2.2398817890222364,
+      "grad_norm": 0.19732961058616638,
+      "learning_rate": 3.0397156185730453e-05,
+      "loss": 0.3776,
       "step": 62150
     },
     {
-      "epoch": 2.19,
-      "learning_rate": 3.123237754587971e-05,
-      "loss": 0.2482,
+      "epoch": 2.240061988683461,
+      "grad_norm": 0.2629784345626831,
+      "learning_rate": 3.039430682922558e-05,
+      "loss": 0.3572,
       "step": 62155
     },
     {
-      "epoch": 2.19,
-      "learning_rate": 3.1229618725174436e-05,
-      "loss": 0.2561,
+      "epoch": 2.240242188344686,
+      "grad_norm": 0.1689624935388565,
+      "learning_rate": 3.0391457399224325e-05,
+      "loss": 0.3499,
       "step": 62160
     },
     {
-      "epoch": 2.19,
-      "learning_rate": 3.122685982358165e-05,
-      "loss": 0.2584,
+      "epoch": 2.2404223880059106,
+      "grad_norm": 0.20784690976142883,
+      "learning_rate": 3.038860789576551e-05,
+      "loss": 0.4062,
       "step": 62165
     },
     {
-      "epoch": 2.19,
-      "learning_rate": 3.1224100841137217e-05,
-      "loss": 0.2677,
+      "epoch": 2.2406025876671354,
+      "grad_norm": 0.17468421161174774,
+      "learning_rate": 3.038575831888794e-05,
+      "loss": 0.4152,
       "step": 62170
     },
     {
-      "epoch": 2.19,
-      "learning_rate": 3.122134177787692e-05,
-      "loss": 0.2588,
+      "epoch": 2.2407827873283597,
+      "grad_norm": 0.21627172827720642,
+      "learning_rate": 3.0382908668630477e-05,
+      "loss": 0.4106,
       "step": 62175
     },
     {
-      "epoch": 2.19,
-      "learning_rate": 3.12185826338366e-05,
-      "loss": 0.2661,
+      "epoch": 2.2409629869895844,
+      "grad_norm": 0.19633768498897552,
+      "learning_rate": 3.0380058945031907e-05,
+      "loss": 0.3889,
       "step": 62180
     },
     {
-      "epoch": 2.19,
-      "learning_rate": 3.121582340905209e-05,
-      "loss": 0.2692,
+      "epoch": 2.241143186650809,
+      "grad_norm": 0.18723514676094055,
+      "learning_rate": 3.0377209148131086e-05,
+      "loss": 0.4013,
       "step": 62185
     },
     {
-      "epoch": 2.19,
-      "learning_rate": 3.1213064103559206e-05,
-      "loss": 0.2429,
+      "epoch": 2.241323386312034,
+      "grad_norm": 0.17972959578037262,
+      "learning_rate": 3.037435927796683e-05,
+      "loss": 0.3765,
       "step": 62190
     },
     {
-      "epoch": 2.19,
-      "learning_rate": 3.1210304717393775e-05,
-      "loss": 0.2833,
+      "epoch": 2.241503585973258,
+      "grad_norm": 0.21043039858341217,
+      "learning_rate": 3.037150933457797e-05,
+      "loss": 0.3866,
       "step": 62195
     },
     {
-      "epoch": 2.19,
-      "learning_rate": 3.120754525059163e-05,
-      "loss": 0.2604,
+      "epoch": 2.241683785634483,
+      "grad_norm": 0.20892179012298584,
+      "learning_rate": 3.0368659318003335e-05,
+      "loss": 0.4267,
       "step": 62200
     },
     {
-      "epoch": 2.19,
-      "learning_rate": 3.12047857031886e-05,
-      "loss": 0.2753,
+      "epoch": 2.2418639852957076,
+      "grad_norm": 0.22525620460510254,
+      "learning_rate": 3.036580922828175e-05,
+      "loss": 0.3884,
       "step": 62205
     },
     {
-      "epoch": 2.19,
-      "learning_rate": 3.1202026075220525e-05,
-      "loss": 0.2935,
+      "epoch": 2.2420441849569324,
+      "grad_norm": 0.23556070029735565,
+      "learning_rate": 3.0362959065452057e-05,
+      "loss": 0.4409,
       "step": 62210
     },
     {
-      "epoch": 2.19,
-      "learning_rate": 3.1199266366723224e-05,
-      "loss": 0.257,
+      "epoch": 2.242224384618157,
+      "grad_norm": 0.20461376011371613,
+      "learning_rate": 3.036010882955308e-05,
+      "loss": 0.4232,
       "step": 62215
     },
     {
-      "epoch": 2.19,
-      "learning_rate": 3.1196506577732545e-05,
-      "loss": 0.2837,
+      "epoch": 2.2424045842793814,
+      "grad_norm": 0.19230584800243378,
+      "learning_rate": 3.035725852062367e-05,
+      "loss": 0.391,
       "step": 62220
     },
     {
-      "epoch": 2.19,
-      "learning_rate": 3.1193746708284294e-05,
-      "loss": 0.2723,
+      "epoch": 2.242584783940606,
+      "grad_norm": 0.17546550929546356,
+      "learning_rate": 3.0354408138702645e-05,
+      "loss": 0.3665,
       "step": 62225
     },
     {
-      "epoch": 2.19,
-      "learning_rate": 3.119098675841433e-05,
-      "loss": 0.2625,
+      "epoch": 2.242764983601831,
+      "grad_norm": 0.24618138372898102,
+      "learning_rate": 3.0351557683828846e-05,
+      "loss": 0.4114,
       "step": 62230
     },
     {
-      "epoch": 2.19,
-      "learning_rate": 3.1188226728158485e-05,
-      "loss": 0.2771,
+      "epoch": 2.2429451832630556,
+      "grad_norm": 0.2337779998779297,
+      "learning_rate": 3.03487071560411e-05,
+      "loss": 0.3832,
       "step": 62235
     },
     {
-      "epoch": 2.19,
-      "learning_rate": 3.1185466617552594e-05,
-      "loss": 0.2816,
+      "epoch": 2.24312538292428,
+      "grad_norm": 0.19449256360530853,
+      "learning_rate": 3.0345856555378267e-05,
+      "loss": 0.4175,
       "step": 62240
     },
     {
-      "epoch": 2.19,
-      "learning_rate": 3.1182706426632496e-05,
-      "loss": 0.2625,
+      "epoch": 2.2433055825855046,
+      "grad_norm": 0.19468238949775696,
+      "learning_rate": 3.0343005881879172e-05,
+      "loss": 0.3843,
       "step": 62245
     },
     {
-      "epoch": 2.19,
-      "learning_rate": 3.117994615543403e-05,
-      "loss": 0.2642,
+      "epoch": 2.2434857822467293,
+      "grad_norm": 0.24619287252426147,
+      "learning_rate": 3.0340155135582648e-05,
+      "loss": 0.403,
       "step": 62250
     },
     {
-      "epoch": 2.19,
-      "learning_rate": 3.117718580399304e-05,
-      "loss": 0.2604,
+      "epoch": 2.243665981907954,
+      "grad_norm": 0.15675576031208038,
+      "learning_rate": 3.0337304316527548e-05,
+      "loss": 0.3894,
       "step": 62255
     },
     {
-      "epoch": 2.19,
-      "learning_rate": 3.117442537234535e-05,
-      "loss": 0.2451,
+      "epoch": 2.243846181569179,
+      "grad_norm": 0.18250007927417755,
+      "learning_rate": 3.0334453424752712e-05,
+      "loss": 0.4095,
       "step": 62260
     },
     {
-      "epoch": 2.19,
-      "learning_rate": 3.1171664860526826e-05,
-      "loss": 0.2944,
+      "epoch": 2.244026381230403,
+      "grad_norm": 0.17939797043800354,
+      "learning_rate": 3.033160246029698e-05,
+      "loss": 0.3893,
       "step": 62265
     },
     {
-      "epoch": 2.19,
-      "learning_rate": 3.1168904268573294e-05,
-      "loss": 0.2968,
+      "epoch": 2.244206580891628,
+      "grad_norm": 0.15464815497398376,
+      "learning_rate": 3.0328751423199198e-05,
+      "loss": 0.3849,
       "step": 62270
     },
     {
-      "epoch": 2.19,
-      "learning_rate": 3.116614359652061e-05,
-      "loss": 0.2615,
+      "epoch": 2.2443867805528526,
+      "grad_norm": 0.1741357445716858,
+      "learning_rate": 3.03259003134982e-05,
+      "loss": 0.4079,
       "step": 62275
     },
     {
-      "epoch": 2.19,
-      "learning_rate": 3.116338284440462e-05,
-      "loss": 0.2496,
+      "epoch": 2.2445669802140773,
+      "grad_norm": 0.1852835714817047,
+      "learning_rate": 3.0323049131232863e-05,
+      "loss": 0.388,
       "step": 62280
     },
     {
-      "epoch": 2.19,
-      "learning_rate": 3.116062201226116e-05,
-      "loss": 0.2591,
+      "epoch": 2.244747179875302,
+      "grad_norm": 0.17500340938568115,
+      "learning_rate": 3.0320197876441992e-05,
+      "loss": 0.4356,
       "step": 62285
     },
     {
-      "epoch": 2.19,
-      "learning_rate": 3.115786110012608e-05,
-      "loss": 0.265,
+      "epoch": 2.2449273795365263,
+      "grad_norm": 0.20107300579547882,
+      "learning_rate": 3.0317346549164465e-05,
+      "loss": 0.3873,
       "step": 62290
     },
     {
-      "epoch": 2.19,
-      "learning_rate": 3.115510010803524e-05,
-      "loss": 0.2671,
+      "epoch": 2.245107579197751,
+      "grad_norm": 0.19302281737327576,
+      "learning_rate": 3.0314495149439115e-05,
+      "loss": 0.3943,
       "step": 62295
     },
     {
-      "epoch": 2.19,
-      "learning_rate": 3.115233903602448e-05,
-      "loss": 0.2622,
+      "epoch": 2.245287778858976,
+      "grad_norm": 0.20569072663784027,
+      "learning_rate": 3.03116436773048e-05,
+      "loss": 0.4247,
       "step": 62300
     },
     {
-      "epoch": 2.19,
-      "learning_rate": 3.114957788412965e-05,
-      "loss": 0.266,
+      "epoch": 2.2454679785202005,
+      "grad_norm": 0.1855345070362091,
+      "learning_rate": 3.0308792132800368e-05,
+      "loss": 0.3976,
       "step": 62305
     },
     {
-      "epoch": 2.19,
-      "learning_rate": 3.11468166523866e-05,
-      "loss": 0.3061,
+      "epoch": 2.245648178181425,
+      "grad_norm": 0.1881013661623001,
+      "learning_rate": 3.0305940515964672e-05,
+      "loss": 0.4136,
       "step": 62310
     },
     {
-      "epoch": 2.19,
-      "learning_rate": 3.114405534083119e-05,
-      "loss": 0.2594,
+      "epoch": 2.2458283778426495,
+      "grad_norm": 0.20543043315410614,
+      "learning_rate": 3.0303088826836563e-05,
+      "loss": 0.3816,
       "step": 62315
     },
     {
-      "epoch": 2.19,
-      "learning_rate": 3.114129394949927e-05,
-      "loss": 0.2714,
+      "epoch": 2.2460085775038743,
+      "grad_norm": 0.28218919038772583,
+      "learning_rate": 3.0300237065454894e-05,
+      "loss": 0.4104,
       "step": 62320
     },
     {
-      "epoch": 2.19,
-      "learning_rate": 3.113853247842669e-05,
-      "loss": 0.285,
+      "epoch": 2.246188777165099,
+      "grad_norm": 0.21315787732601166,
+      "learning_rate": 3.0297385231858526e-05,
+      "loss": 0.4041,
       "step": 62325
     },
     {
-      "epoch": 2.19,
-      "learning_rate": 3.113577092764932e-05,
-      "loss": 0.2718,
+      "epoch": 2.2463689768263237,
+      "grad_norm": 0.19718264043331146,
+      "learning_rate": 3.0294533326086305e-05,
+      "loss": 0.4086,
       "step": 62330
     },
     {
-      "epoch": 2.19,
-      "learning_rate": 3.113300929720302e-05,
-      "loss": 0.2715,
+      "epoch": 2.246549176487548,
+      "grad_norm": 0.1942020058631897,
+      "learning_rate": 3.0291681348177092e-05,
+      "loss": 0.3855,
       "step": 62335
     },
     {
-      "epoch": 2.19,
-      "learning_rate": 3.113024758712361e-05,
-      "loss": 0.2728,
+      "epoch": 2.2467293761487728,
+      "grad_norm": 0.1929059475660324,
+      "learning_rate": 3.0288829298169753e-05,
+      "loss": 0.3879,
       "step": 62340
     },
     {
-      "epoch": 2.19,
-      "learning_rate": 3.112748579744699e-05,
-      "loss": 0.264,
+      "epoch": 2.2469095758099975,
+      "grad_norm": 0.2083449363708496,
+      "learning_rate": 3.0285977176103132e-05,
+      "loss": 0.3884,
       "step": 62345
     },
     {
-      "epoch": 2.19,
-      "learning_rate": 3.1124723928209e-05,
-      "loss": 0.2649,
+      "epoch": 2.2470897754712222,
+      "grad_norm": 0.22706344723701477,
+      "learning_rate": 3.028312498201609e-05,
+      "loss": 0.4189,
       "step": 62350
     },
     {
-      "epoch": 2.19,
-      "learning_rate": 3.1121961979445506e-05,
-      "loss": 0.2562,
+      "epoch": 2.247269975132447,
+      "grad_norm": 0.2006312906742096,
+      "learning_rate": 3.02802727159475e-05,
+      "loss": 0.408,
       "step": 62355
     },
     {
-      "epoch": 2.19,
-      "learning_rate": 3.111919995119237e-05,
-      "loss": 0.2866,
+      "epoch": 2.2474501747936713,
+      "grad_norm": 0.2271738350391388,
+      "learning_rate": 3.0277420377936222e-05,
+      "loss": 0.4563,
       "step": 62360
     },
     {
-      "epoch": 2.19,
-      "learning_rate": 3.1116437843485465e-05,
-      "loss": 0.2639,
+      "epoch": 2.247630374454896,
+      "grad_norm": 0.19731751084327698,
+      "learning_rate": 3.0274567968021107e-05,
+      "loss": 0.4227,
       "step": 62365
     },
     {
-      "epoch": 2.19,
-      "learning_rate": 3.111367565636063e-05,
-      "loss": 0.2655,
+      "epoch": 2.2478105741161207,
+      "grad_norm": 0.19308514893054962,
+      "learning_rate": 3.0271715486241026e-05,
+      "loss": 0.4012,
       "step": 62370
     },
     {
-      "epoch": 2.19,
-      "learning_rate": 3.111091338985376e-05,
-      "loss": 0.2735,
+      "epoch": 2.2479907737773455,
+      "grad_norm": 0.19526423513889313,
+      "learning_rate": 3.0268862932634832e-05,
+      "loss": 0.3879,
       "step": 62375
     },
     {
-      "epoch": 2.19,
-      "learning_rate": 3.110815104400069e-05,
-      "loss": 0.2723,
+      "epoch": 2.2481709734385698,
+      "grad_norm": 0.20854395627975464,
+      "learning_rate": 3.026601030724141e-05,
+      "loss": 0.3913,
       "step": 62380
     },
     {
-      "epoch": 2.19,
-      "learning_rate": 3.1105388618837314e-05,
-      "loss": 0.3042,
+      "epoch": 2.2483511730997945,
+      "grad_norm": 0.1848597228527069,
+      "learning_rate": 3.0263157610099625e-05,
+      "loss": 0.3934,
       "step": 62385
     },
     {
-      "epoch": 2.2,
-      "learning_rate": 3.11026261143995e-05,
-      "loss": 0.2662,
+      "epoch": 2.248531372761019,
+      "grad_norm": 0.22897033393383026,
+      "learning_rate": 3.0260304841248322e-05,
+      "loss": 0.3878,
       "step": 62390
     },
     {
-      "epoch": 2.2,
-      "learning_rate": 3.109986353072309e-05,
-      "loss": 0.2803,
+      "epoch": 2.248711572422244,
+      "grad_norm": 0.1810678094625473,
+      "learning_rate": 3.025745200072639e-05,
+      "loss": 0.4209,
       "step": 62395
     },
     {
-      "epoch": 2.2,
-      "learning_rate": 3.109710086784398e-05,
-      "loss": 0.28,
+      "epoch": 2.2488917720834687,
+      "grad_norm": 0.19008022546768188,
+      "learning_rate": 3.0254599088572688e-05,
+      "loss": 0.3718,
       "step": 62400
     },
     {
-      "epoch": 2.2,
-      "learning_rate": 3.1094338125798035e-05,
-      "loss": 0.2713,
+      "epoch": 2.249071971744693,
+      "grad_norm": 0.18570220470428467,
+      "learning_rate": 3.02517461048261e-05,
+      "loss": 0.3892,
       "step": 62405
     },
     {
-      "epoch": 2.2,
-      "learning_rate": 3.1091575304621126e-05,
-      "loss": 0.2577,
+      "epoch": 2.2492521714059177,
+      "grad_norm": 0.23810695111751556,
+      "learning_rate": 3.0248893049525483e-05,
+      "loss": 0.42,
       "step": 62410
     },
     {
-      "epoch": 2.2,
-      "learning_rate": 3.108881240434912e-05,
-      "loss": 0.2785,
+      "epoch": 2.2494323710671424,
+      "grad_norm": 0.23687207698822021,
+      "learning_rate": 3.024603992270971e-05,
+      "loss": 0.4312,
       "step": 62415
     },
     {
-      "epoch": 2.2,
-      "learning_rate": 3.108604942501791e-05,
-      "loss": 0.2456,
+      "epoch": 2.249612570728367,
+      "grad_norm": 0.1891470104455948,
+      "learning_rate": 3.0243186724417664e-05,
+      "loss": 0.3563,
       "step": 62420
     },
     {
-      "epoch": 2.2,
-      "learning_rate": 3.108328636666335e-05,
-      "loss": 0.2876,
+      "epoch": 2.2497927703895915,
+      "grad_norm": 0.18046557903289795,
+      "learning_rate": 3.0240333454688214e-05,
+      "loss": 0.3756,
       "step": 62425
     },
     {
-      "epoch": 2.2,
-      "learning_rate": 3.108052322932132e-05,
-      "loss": 0.2638,
+      "epoch": 2.249972970050816,
+      "grad_norm": 0.19262628257274628,
+      "learning_rate": 3.0237480113560235e-05,
+      "loss": 0.3925,
       "step": 62430
     },
     {
-      "epoch": 2.2,
-      "learning_rate": 3.107776001302771e-05,
-      "loss": 0.2657,
+      "epoch": 2.250153169712041,
+      "grad_norm": 0.22571761906147003,
+      "learning_rate": 3.02346267010726e-05,
+      "loss": 0.3438,
       "step": 62435
     },
     {
-      "epoch": 2.2,
-      "learning_rate": 3.10749967178184e-05,
-      "loss": 0.2973,
+      "epoch": 2.2503333693732657,
+      "grad_norm": 0.19609612226486206,
+      "learning_rate": 3.02317732172642e-05,
+      "loss": 0.3906,
       "step": 62440
     },
     {
-      "epoch": 2.2,
-      "learning_rate": 3.107223334372925e-05,
-      "loss": 0.2855,
+      "epoch": 2.2505135690344904,
+      "grad_norm": 0.2036208212375641,
+      "learning_rate": 3.0228919662173895e-05,
+      "loss": 0.4103,
       "step": 62445
     },
     {
-      "epoch": 2.2,
-      "learning_rate": 3.106946989079616e-05,
-      "loss": 0.2855,
+      "epoch": 2.2506937686957147,
+      "grad_norm": 0.20205596089363098,
+      "learning_rate": 3.0226066035840568e-05,
+      "loss": 0.4045,
       "step": 62450
     },
     {
-      "epoch": 2.2,
-      "learning_rate": 3.1066706359054984e-05,
-      "loss": 0.2701,
+      "epoch": 2.2508739683569394,
+      "grad_norm": 0.18515092134475708,
+      "learning_rate": 3.0223212338303108e-05,
+      "loss": 0.4139,
       "step": 62455
     },
     {
-      "epoch": 2.2,
-      "learning_rate": 3.1063942748541644e-05,
-      "loss": 0.2738,
+      "epoch": 2.251054168018164,
+      "grad_norm": 0.19203773140907288,
+      "learning_rate": 3.0220358569600388e-05,
+      "loss": 0.3927,
       "step": 62460
     },
     {
-      "epoch": 2.2,
-      "learning_rate": 3.1061179059292e-05,
-      "loss": 0.2961,
+      "epoch": 2.251234367679389,
+      "grad_norm": 0.2266424596309662,
+      "learning_rate": 3.02175047297713e-05,
+      "loss": 0.4254,
       "step": 62465
     },
     {
-      "epoch": 2.2,
-      "learning_rate": 3.1058415291341944e-05,
-      "loss": 0.2745,
+      "epoch": 2.251414567340613,
+      "grad_norm": 0.23780296742916107,
+      "learning_rate": 3.0214650818854713e-05,
+      "loss": 0.3954,
       "step": 62470
     },
     {
-      "epoch": 2.2,
-      "learning_rate": 3.1055651444727345e-05,
-      "loss": 0.2487,
+      "epoch": 2.251594767001838,
+      "grad_norm": 0.2071855366230011,
+      "learning_rate": 3.021179683688952e-05,
+      "loss": 0.3667,
       "step": 62475
     },
     {
-      "epoch": 2.2,
-      "learning_rate": 3.105288751948411e-05,
-      "loss": 0.2876,
+      "epoch": 2.2517749666630626,
+      "grad_norm": 0.23853279650211334,
+      "learning_rate": 3.0208942783914606e-05,
+      "loss": 0.4328,
       "step": 62480
     },
     {
-      "epoch": 2.2,
-      "learning_rate": 3.105012351564812e-05,
-      "loss": 0.2709,
+      "epoch": 2.2519551663242874,
+      "grad_norm": 0.23091068863868713,
+      "learning_rate": 3.0206088659968857e-05,
+      "loss": 0.4059,
       "step": 62485
     },
     {
-      "epoch": 2.2,
-      "learning_rate": 3.1047359433255263e-05,
-      "loss": 0.2986,
+      "epoch": 2.252135365985512,
+      "grad_norm": 0.23153997957706451,
+      "learning_rate": 3.0203234465091157e-05,
+      "loss": 0.3665,
       "step": 62490
     },
     {
-      "epoch": 2.2,
-      "learning_rate": 3.1044595272341425e-05,
-      "loss": 0.2757,
+      "epoch": 2.2523155656467364,
+      "grad_norm": 0.18207654356956482,
+      "learning_rate": 3.020038019932039e-05,
+      "loss": 0.3626,
       "step": 62495
     },
     {
-      "epoch": 2.2,
-      "learning_rate": 3.104183103294251e-05,
-      "loss": 0.2839,
+      "epoch": 2.252495765307961,
+      "grad_norm": 0.19852490723133087,
+      "learning_rate": 3.0197525862695452e-05,
+      "loss": 0.4133,
       "step": 62500
     },
     {
-      "epoch": 2.2,
-      "eval_loss": 0.2684471011161804,
-      "eval_runtime": 10.5326,
-      "eval_samples_per_second": 9.494,
-      "eval_steps_per_second": 9.494,
+      "epoch": 2.252495765307961,
+      "eval_loss": 0.4360164999961853,
+      "eval_runtime": 3.5278,
+      "eval_samples_per_second": 28.347,
+      "eval_steps_per_second": 7.087,
       "step": 62500
     },
     {
-      "epoch": 2.2,
-      "learning_rate": 3.103906671509439e-05,
-      "loss": 0.2537,
+      "epoch": 2.252675964969186,
+      "grad_norm": 0.23163332045078278,
+      "learning_rate": 3.019467145525523e-05,
+      "loss": 0.3957,
       "step": 62505
     },
     {
-      "epoch": 2.2,
-      "learning_rate": 3.103630231883297e-05,
-      "loss": 0.2632,
+      "epoch": 2.2528561646304106,
+      "grad_norm": 0.21983997523784637,
+      "learning_rate": 3.0191816977038622e-05,
+      "loss": 0.3898,
       "step": 62510
     },
     {
-      "epoch": 2.2,
-      "learning_rate": 3.1033537844194155e-05,
-      "loss": 0.2969,
+      "epoch": 2.253036364291635,
+      "grad_norm": 0.1701703816652298,
+      "learning_rate": 3.01889624280845e-05,
+      "loss": 0.3719,
       "step": 62515
     },
     {
-      "epoch": 2.2,
-      "learning_rate": 3.103077329121382e-05,
-      "loss": 0.2819,
+      "epoch": 2.2532165639528596,
+      "grad_norm": 0.17912349104881287,
+      "learning_rate": 3.018610780843178e-05,
+      "loss": 0.3743,
       "step": 62520
     },
     {
-      "epoch": 2.2,
-      "learning_rate": 3.1028008659927866e-05,
-      "loss": 0.237,
+      "epoch": 2.2533967636140844,
+      "grad_norm": 0.20103944838047028,
+      "learning_rate": 3.0183253118119342e-05,
+      "loss": 0.4066,
       "step": 62525
     },
     {
-      "epoch": 2.2,
-      "learning_rate": 3.102524395037219e-05,
-      "loss": 0.2558,
+      "epoch": 2.253576963275309,
+      "grad_norm": 0.20848768949508667,
+      "learning_rate": 3.018039835718608e-05,
+      "loss": 0.4139,
       "step": 62530
     },
     {
-      "epoch": 2.2,
-      "learning_rate": 3.1022479162582695e-05,
-      "loss": 0.2682,
+      "epoch": 2.253757162936534,
+      "grad_norm": 0.1789763867855072,
+      "learning_rate": 3.0177543525670898e-05,
+      "loss": 0.404,
       "step": 62535
     },
     {
-      "epoch": 2.2,
-      "learning_rate": 3.101971429659528e-05,
-      "loss": 0.249,
+      "epoch": 2.2539373625977586,
+      "grad_norm": 0.1968272179365158,
+      "learning_rate": 3.0174688623612684e-05,
+      "loss": 0.39,
       "step": 62540
     },
     {
-      "epoch": 2.2,
-      "learning_rate": 3.101694935244584e-05,
-      "loss": 0.2778,
+      "epoch": 2.254117562258983,
+      "grad_norm": 0.2528758645057678,
+      "learning_rate": 3.0171833651050352e-05,
+      "loss": 0.4127,
       "step": 62545
     },
     {
-      "epoch": 2.2,
-      "learning_rate": 3.101418433017027e-05,
-      "loss": 0.2924,
+      "epoch": 2.2542977619202076,
+      "grad_norm": 0.23864948749542236,
+      "learning_rate": 3.016897860802277e-05,
+      "loss": 0.3495,
       "step": 62550
     },
     {
-      "epoch": 2.2,
-      "learning_rate": 3.10114192298045e-05,
-      "loss": 0.2784,
+      "epoch": 2.2544779615814323,
+      "grad_norm": 0.21808521449565887,
+      "learning_rate": 3.0166123494568865e-05,
+      "loss": 0.3913,
       "step": 62555
     },
     {
-      "epoch": 2.2,
-      "learning_rate": 3.10086540513844e-05,
-      "loss": 0.2703,
+      "epoch": 2.254658161242657,
+      "grad_norm": 0.17956504225730896,
+      "learning_rate": 3.016326831072752e-05,
+      "loss": 0.3843,
       "step": 62560
     },
     {
-      "epoch": 2.2,
-      "learning_rate": 3.1005888794945884e-05,
-      "loss": 0.285,
+      "epoch": 2.2548383609038813,
+      "grad_norm": 0.1821889877319336,
+      "learning_rate": 3.0160413056537656e-05,
+      "loss": 0.376,
       "step": 62565
     },
     {
-      "epoch": 2.2,
-      "learning_rate": 3.1003123460524865e-05,
-      "loss": 0.2675,
+      "epoch": 2.255018560565106,
+      "grad_norm": 0.1842850148677826,
+      "learning_rate": 3.0157557732038155e-05,
+      "loss": 0.3938,
       "step": 62570
     },
     {
-      "epoch": 2.2,
-      "learning_rate": 3.100035804815724e-05,
-      "loss": 0.2688,
+      "epoch": 2.255198760226331,
+      "grad_norm": 0.17056186497211456,
+      "learning_rate": 3.0154702337267926e-05,
+      "loss": 0.3657,
       "step": 62575
     },
     {
-      "epoch": 2.2,
-      "learning_rate": 3.0997592557878926e-05,
-      "loss": 0.2583,
+      "epoch": 2.2553789598875555,
+      "grad_norm": 0.23995059728622437,
+      "learning_rate": 3.0151846872265883e-05,
+      "loss": 0.4019,
       "step": 62580
     },
     {
-      "epoch": 2.2,
-      "learning_rate": 3.099482698972582e-05,
-      "loss": 0.2913,
+      "epoch": 2.2555591595487803,
+      "grad_norm": 0.25965413451194763,
+      "learning_rate": 3.014899133707092e-05,
+      "loss": 0.4349,
       "step": 62585
     },
     {
-      "epoch": 2.2,
-      "learning_rate": 3.0992061343733835e-05,
-      "loss": 0.308,
+      "epoch": 2.2557393592100046,
+      "grad_norm": 0.16279491782188416,
+      "learning_rate": 3.0146135731721946e-05,
+      "loss": 0.3995,
       "step": 62590
     },
     {
-      "epoch": 2.2,
-      "learning_rate": 3.098929561993889e-05,
-      "loss": 0.274,
+      "epoch": 2.2559195588712293,
+      "grad_norm": 0.2433793842792511,
+      "learning_rate": 3.0143280056257874e-05,
+      "loss": 0.3998,
       "step": 62595
     },
     {
-      "epoch": 2.2,
-      "learning_rate": 3.098652981837688e-05,
-      "loss": 0.2655,
+      "epoch": 2.256099758532454,
+      "grad_norm": 0.17817862331867218,
+      "learning_rate": 3.0140424310717597e-05,
+      "loss": 0.3904,
       "step": 62600
     },
     {
-      "epoch": 2.2,
-      "learning_rate": 3.098376393908373e-05,
-      "loss": 0.2737,
+      "epoch": 2.2562799581936788,
+      "grad_norm": 0.18165291845798492,
+      "learning_rate": 3.0137568495140045e-05,
+      "loss": 0.3814,
       "step": 62605
     },
     {
-      "epoch": 2.2,
-      "learning_rate": 3.098099798209535e-05,
-      "loss": 0.2755,
+      "epoch": 2.256460157854903,
+      "grad_norm": 0.17060628533363342,
+      "learning_rate": 3.0134712609564114e-05,
+      "loss": 0.4033,
       "step": 62610
     },
     {
-      "epoch": 2.2,
-      "learning_rate": 3.097823194744765e-05,
-      "loss": 0.2636,
+      "epoch": 2.256640357516128,
+      "grad_norm": 0.19508397579193115,
+      "learning_rate": 3.0131856654028716e-05,
+      "loss": 0.3854,
       "step": 62615
     },
     {
-      "epoch": 2.2,
-      "learning_rate": 3.0975465835176545e-05,
-      "loss": 0.2543,
+      "epoch": 2.2568205571773525,
+      "grad_norm": 0.21551352739334106,
+      "learning_rate": 3.012900062857276e-05,
+      "loss": 0.4103,
       "step": 62620
     },
     {
-      "epoch": 2.2,
-      "learning_rate": 3.097269964531796e-05,
-      "loss": 0.2555,
+      "epoch": 2.2570007568385773,
+      "grad_norm": 0.20918194949626923,
+      "learning_rate": 3.0126144533235172e-05,
+      "loss": 0.3812,
       "step": 62625
     },
     {
-      "epoch": 2.2,
-      "learning_rate": 3.096993337790781e-05,
-      "loss": 0.2465,
+      "epoch": 2.257180956499802,
+      "grad_norm": 0.2055574357509613,
+      "learning_rate": 3.0123288368054857e-05,
+      "loss": 0.3614,
       "step": 62630
     },
     {
-      "epoch": 2.2,
-      "learning_rate": 3.0967167032982e-05,
-      "loss": 0.2686,
+      "epoch": 2.2573611561610263,
+      "grad_norm": 0.201625257730484,
+      "learning_rate": 3.012043213307072e-05,
+      "loss": 0.3726,
       "step": 62635
     },
     {
-      "epoch": 2.2,
-      "learning_rate": 3.096440061057647e-05,
-      "loss": 0.2759,
+      "epoch": 2.257541355822251,
+      "grad_norm": 0.2094610631465912,
+      "learning_rate": 3.0117575828321697e-05,
+      "loss": 0.3753,
       "step": 62640
     },
     {
-      "epoch": 2.2,
-      "learning_rate": 3.096163411072711e-05,
-      "loss": 0.2765,
+      "epoch": 2.2577215554834758,
+      "grad_norm": 0.19469302892684937,
+      "learning_rate": 3.0114719453846684e-05,
+      "loss": 0.3766,
       "step": 62645
     },
     {
-      "epoch": 2.2,
-      "learning_rate": 3.095886753346988e-05,
-      "loss": 0.2789,
+      "epoch": 2.2579017551447005,
+      "grad_norm": 0.26172351837158203,
+      "learning_rate": 3.0111863009684627e-05,
+      "loss": 0.4036,
       "step": 62650
     },
     {
-      "epoch": 2.2,
-      "learning_rate": 3.095610087884067e-05,
-      "loss": 0.2703,
+      "epoch": 2.2580819548059248,
+      "grad_norm": 0.16828452050685883,
+      "learning_rate": 3.0109006495874408e-05,
+      "loss": 0.3963,
       "step": 62655
     },
     {
-      "epoch": 2.2,
-      "learning_rate": 3.095333414687542e-05,
-      "loss": 0.2798,
+      "epoch": 2.2582621544671495,
+      "grad_norm": 0.17929872870445251,
+      "learning_rate": 3.0106149912454974e-05,
+      "loss": 0.3774,
       "step": 62660
     },
     {
-      "epoch": 2.2,
-      "learning_rate": 3.095056733761005e-05,
-      "loss": 0.274,
+      "epoch": 2.2584423541283742,
+      "grad_norm": 0.19596625864505768,
+      "learning_rate": 3.0103293259465227e-05,
+      "loss": 0.3821,
       "step": 62665
     },
     {
-      "epoch": 2.2,
-      "learning_rate": 3.094780045108048e-05,
-      "loss": 0.2811,
+      "epoch": 2.258622553789599,
+      "grad_norm": 0.17597244679927826,
+      "learning_rate": 3.0100436536944105e-05,
+      "loss": 0.4049,
       "step": 62670
     },
     {
-      "epoch": 2.21,
-      "learning_rate": 3.094503348732264e-05,
-      "loss": 0.2785,
+      "epoch": 2.2588027534508237,
+      "grad_norm": 0.22724290192127228,
+      "learning_rate": 3.0097579744930522e-05,
+      "loss": 0.4355,
       "step": 62675
     },
     {
-      "epoch": 2.21,
-      "learning_rate": 3.094226644637246e-05,
-      "loss": 0.2748,
+      "epoch": 2.258982953112048,
+      "grad_norm": 0.24098409712314606,
+      "learning_rate": 3.0094722883463396e-05,
+      "loss": 0.4125,
       "step": 62680
     },
     {
-      "epoch": 2.21,
-      "learning_rate": 3.093949932826587e-05,
-      "loss": 0.2676,
+      "epoch": 2.2591631527732727,
+      "grad_norm": 0.22858020663261414,
+      "learning_rate": 3.0091865952581665e-05,
+      "loss": 0.4162,
       "step": 62685
     },
     {
-      "epoch": 2.21,
-      "learning_rate": 3.0936732133038795e-05,
-      "loss": 0.2652,
+      "epoch": 2.2593433524344975,
+      "grad_norm": 0.16358384490013123,
+      "learning_rate": 3.0089008952324243e-05,
+      "loss": 0.3844,
       "step": 62690
     },
     {
-      "epoch": 2.21,
-      "learning_rate": 3.0933964860727167e-05,
-      "loss": 0.2779,
+      "epoch": 2.259523552095722,
+      "grad_norm": 0.2142842710018158,
+      "learning_rate": 3.0086151882730063e-05,
+      "loss": 0.3829,
       "step": 62695
     },
     {
-      "epoch": 2.21,
-      "learning_rate": 3.0931197511366914e-05,
-      "loss": 0.2713,
+      "epoch": 2.2597037517569465,
+      "grad_norm": 0.26690196990966797,
+      "learning_rate": 3.0083294743838036e-05,
+      "loss": 0.4132,
       "step": 62700
     },
     {
-      "epoch": 2.21,
-      "learning_rate": 3.0928430084993965e-05,
-      "loss": 0.2618,
+      "epoch": 2.259883951418171,
+      "grad_norm": 0.16452406346797943,
+      "learning_rate": 3.008043753568711e-05,
+      "loss": 0.3578,
       "step": 62705
     },
     {
-      "epoch": 2.21,
-      "learning_rate": 3.092566258164427e-05,
-      "loss": 0.2765,
+      "epoch": 2.260064151079396,
+      "grad_norm": 0.21890616416931152,
+      "learning_rate": 3.0077580258316213e-05,
+      "loss": 0.3875,
       "step": 62710
     },
     {
-      "epoch": 2.21,
-      "learning_rate": 3.0922895001353736e-05,
-      "loss": 0.2803,
+      "epoch": 2.2602443507406207,
+      "grad_norm": 0.22171573340892792,
+      "learning_rate": 3.0074722911764258e-05,
+      "loss": 0.3939,
       "step": 62715
     },
     {
-      "epoch": 2.21,
-      "learning_rate": 3.0920127344158326e-05,
-      "loss": 0.2767,
+      "epoch": 2.2604245504018454,
+      "grad_norm": 0.2403387874364853,
+      "learning_rate": 3.007186549607019e-05,
+      "loss": 0.4446,
       "step": 62720
     },
     {
-      "epoch": 2.21,
-      "learning_rate": 3.091735961009396e-05,
-      "loss": 0.2781,
+      "epoch": 2.2606047500630697,
+      "grad_norm": 0.1787687987089157,
+      "learning_rate": 3.0069008011272936e-05,
+      "loss": 0.3989,
       "step": 62725
     },
     {
-      "epoch": 2.21,
-      "learning_rate": 3.091459179919657e-05,
-      "loss": 0.2569,
+      "epoch": 2.2607849497242944,
+      "grad_norm": 0.17443282902240753,
+      "learning_rate": 3.006615045741143e-05,
+      "loss": 0.4032,
       "step": 62730
     },
     {
-      "epoch": 2.21,
-      "learning_rate": 3.0911823911502116e-05,
-      "loss": 0.2774,
+      "epoch": 2.260965149385519,
+      "grad_norm": 0.18043413758277893,
+      "learning_rate": 3.0063292834524604e-05,
+      "loss": 0.415,
       "step": 62735
     },
     {
-      "epoch": 2.21,
-      "learning_rate": 3.090905594704652e-05,
-      "loss": 0.2852,
+      "epoch": 2.261145349046744,
+      "grad_norm": 0.22148096561431885,
+      "learning_rate": 3.0060435142651387e-05,
+      "loss": 0.423,
       "step": 62740
     },
     {
-      "epoch": 2.21,
-      "learning_rate": 3.090628790586573e-05,
-      "loss": 0.2657,
+      "epoch": 2.261325548707968,
+      "grad_norm": 0.18598996102809906,
+      "learning_rate": 3.0057577381830732e-05,
+      "loss": 0.3616,
       "step": 62745
     },
     {
-      "epoch": 2.21,
-      "learning_rate": 3.0903519787995684e-05,
-      "loss": 0.2806,
+      "epoch": 2.261505748369193,
+      "grad_norm": 0.17888624966144562,
+      "learning_rate": 3.005471955210156e-05,
+      "loss": 0.3796,
       "step": 62750
     },
     {
-      "epoch": 2.21,
-      "learning_rate": 3.090075159347232e-05,
-      "loss": 0.2882,
+      "epoch": 2.2616859480304177,
+      "grad_norm": 0.18961434066295624,
+      "learning_rate": 3.0051861653502815e-05,
+      "loss": 0.3933,
       "step": 62755
     },
     {
-      "epoch": 2.21,
-      "learning_rate": 3.0897983322331586e-05,
-      "loss": 0.2767,
+      "epoch": 2.2618661476916424,
+      "grad_norm": 0.23519083857536316,
+      "learning_rate": 3.004900368607343e-05,
+      "loss": 0.3984,
       "step": 62760
     },
     {
-      "epoch": 2.21,
-      "learning_rate": 3.0895214974609425e-05,
-      "loss": 0.2908,
+      "epoch": 2.262046347352867,
+      "grad_norm": 0.24493834376335144,
+      "learning_rate": 3.0046145649852347e-05,
+      "loss": 0.3918,
       "step": 62765
     },
     {
-      "epoch": 2.21,
-      "learning_rate": 3.0892446550341783e-05,
-      "loss": 0.2871,
+      "epoch": 2.2622265470140914,
+      "grad_norm": 0.16111089289188385,
+      "learning_rate": 3.0043287544878513e-05,
+      "loss": 0.4249,
       "step": 62770
     },
     {
-      "epoch": 2.21,
-      "learning_rate": 3.088967804956461e-05,
-      "loss": 0.2766,
+      "epoch": 2.262406746675316,
+      "grad_norm": 0.24691814184188843,
+      "learning_rate": 3.0040429371190855e-05,
+      "loss": 0.3989,
       "step": 62775
     },
     {
-      "epoch": 2.21,
-      "learning_rate": 3.088690947231385e-05,
-      "loss": 0.2849,
+      "epoch": 2.262586946336541,
+      "grad_norm": 0.22405491769313812,
+      "learning_rate": 3.0037571128828323e-05,
+      "loss": 0.3911,
       "step": 62780
     },
     {
-      "epoch": 2.21,
-      "learning_rate": 3.088414081862544e-05,
-      "loss": 0.2766,
+      "epoch": 2.2627671459977656,
+      "grad_norm": 0.16816742718219757,
+      "learning_rate": 3.003471281782986e-05,
+      "loss": 0.3844,
       "step": 62785
     },
     {
-      "epoch": 2.21,
-      "learning_rate": 3.088137208853535e-05,
-      "loss": 0.252,
+      "epoch": 2.26294734565899,
+      "grad_norm": 0.21033096313476562,
+      "learning_rate": 3.0031854438234413e-05,
+      "loss": 0.4017,
       "step": 62790
     },
     {
-      "epoch": 2.21,
-      "learning_rate": 3.0878603282079516e-05,
-      "loss": 0.2417,
+      "epoch": 2.2631275453202147,
+      "grad_norm": 0.16821134090423584,
+      "learning_rate": 3.002899599008092e-05,
+      "loss": 0.3908,
       "step": 62795
     },
     {
-      "epoch": 2.21,
-      "learning_rate": 3.087583439929389e-05,
-      "loss": 0.2696,
+      "epoch": 2.2633077449814394,
+      "grad_norm": 0.23085516691207886,
+      "learning_rate": 3.0026137473408332e-05,
+      "loss": 0.4195,
       "step": 62800
     },
     {
-      "epoch": 2.21,
-      "learning_rate": 3.087306544021443e-05,
-      "loss": 0.2914,
+      "epoch": 2.263487944642664,
+      "grad_norm": 0.22191166877746582,
+      "learning_rate": 3.0023278888255595e-05,
+      "loss": 0.4187,
       "step": 62805
     },
     {
-      "epoch": 2.21,
-      "learning_rate": 3.087029640487708e-05,
-      "loss": 0.2581,
+      "epoch": 2.263668144303889,
+      "grad_norm": 0.173319473862648,
+      "learning_rate": 3.0020420234661655e-05,
+      "loss": 0.3752,
       "step": 62810
     },
     {
-      "epoch": 2.21,
-      "learning_rate": 3.08675272933178e-05,
-      "loss": 0.2741,
+      "epoch": 2.2638483439651136,
+      "grad_norm": 0.20444796979427338,
+      "learning_rate": 3.001756151266546e-05,
+      "loss": 0.4048,
       "step": 62815
     },
     {
-      "epoch": 2.21,
-      "learning_rate": 3.086475810557255e-05,
-      "loss": 0.2921,
+      "epoch": 2.264028543626338,
+      "grad_norm": 0.18994398415088654,
+      "learning_rate": 3.0014702722305958e-05,
+      "loss": 0.4,
       "step": 62820
     },
     {
-      "epoch": 2.21,
-      "learning_rate": 3.086198884167728e-05,
-      "loss": 0.2578,
+      "epoch": 2.2642087432875626,
+      "grad_norm": 0.1826678216457367,
+      "learning_rate": 3.0011843863622112e-05,
+      "loss": 0.4135,
       "step": 62825
     },
     {
-      "epoch": 2.21,
-      "learning_rate": 3.0859219501667944e-05,
-      "loss": 0.2567,
+      "epoch": 2.2643889429487873,
+      "grad_norm": 0.19503554701805115,
+      "learning_rate": 3.000898493665285e-05,
+      "loss": 0.4056,
       "step": 62830
     },
     {
-      "epoch": 2.21,
-      "learning_rate": 3.085645008558051e-05,
-      "loss": 0.2587,
+      "epoch": 2.264569142610012,
+      "grad_norm": 0.22407963871955872,
+      "learning_rate": 3.0006125941437157e-05,
+      "loss": 0.4232,
       "step": 62835
     },
     {
-      "epoch": 2.21,
-      "learning_rate": 3.085368059345093e-05,
-      "loss": 0.2745,
+      "epoch": 2.2647493422712364,
+      "grad_norm": 0.1850503832101822,
+      "learning_rate": 3.000326687801395e-05,
+      "loss": 0.4157,
       "step": 62840
     },
     {
-      "epoch": 2.21,
-      "learning_rate": 3.085091102531516e-05,
-      "loss": 0.2569,
+      "epoch": 2.264929541932461,
+      "grad_norm": 0.17998819053173065,
+      "learning_rate": 3.000040774642221e-05,
+      "loss": 0.3862,
       "step": 62845
     },
     {
-      "epoch": 2.21,
-      "learning_rate": 3.084814138120917e-05,
-      "loss": 0.279,
+      "epoch": 2.265109741593686,
+      "grad_norm": 0.27859246730804443,
+      "learning_rate": 2.999754854670087e-05,
+      "loss": 0.4156,
       "step": 62850
     },
     {
-      "epoch": 2.21,
-      "learning_rate": 3.084537166116892e-05,
-      "loss": 0.2621,
+      "epoch": 2.2652899412549106,
+      "grad_norm": 0.21499688923358917,
+      "learning_rate": 2.9994689278888914e-05,
+      "loss": 0.3995,
       "step": 62855
     },
     {
-      "epoch": 2.21,
-      "learning_rate": 3.084260186523038e-05,
-      "loss": 0.2597,
+      "epoch": 2.2654701409161353,
+      "grad_norm": 0.1794780045747757,
+      "learning_rate": 2.999182994302528e-05,
+      "loss": 0.3926,
       "step": 62860
     },
     {
-      "epoch": 2.21,
-      "learning_rate": 3.083983199342949e-05,
-      "loss": 0.2743,
+      "epoch": 2.2656503405773596,
+      "grad_norm": 0.20833353698253632,
+      "learning_rate": 2.998897053914892e-05,
+      "loss": 0.4182,
       "step": 62865
     },
     {
-      "epoch": 2.21,
-      "learning_rate": 3.0837062045802235e-05,
-      "loss": 0.2986,
+      "epoch": 2.2658305402385843,
+      "grad_norm": 0.2357635349035263,
+      "learning_rate": 2.998611106729881e-05,
+      "loss": 0.4268,
       "step": 62870
     },
     {
-      "epoch": 2.21,
-      "learning_rate": 3.083429202238457e-05,
-      "loss": 0.2834,
+      "epoch": 2.266010739899809,
+      "grad_norm": 0.18633919954299927,
+      "learning_rate": 2.9983251527513906e-05,
+      "loss": 0.3511,
       "step": 62875
     },
     {
-      "epoch": 2.21,
-      "learning_rate": 3.0831521923212484e-05,
-      "loss": 0.2812,
+      "epoch": 2.266190939561034,
+      "grad_norm": 0.2481304407119751,
+      "learning_rate": 2.9980391919833156e-05,
+      "loss": 0.3642,
       "step": 62880
     },
     {
-      "epoch": 2.21,
-      "learning_rate": 3.0828751748321914e-05,
-      "loss": 0.2694,
+      "epoch": 2.266371139222258,
+      "grad_norm": 0.24019500613212585,
+      "learning_rate": 2.9977532244295537e-05,
+      "loss": 0.391,
       "step": 62885
     },
     {
-      "epoch": 2.21,
-      "learning_rate": 3.082598149774884e-05,
-      "loss": 0.2614,
+      "epoch": 2.266551338883483,
+      "grad_norm": 0.21231551468372345,
+      "learning_rate": 2.9974672500939994e-05,
+      "loss": 0.3907,
       "step": 62890
     },
     {
-      "epoch": 2.21,
-      "learning_rate": 3.082321117152925e-05,
-      "loss": 0.2899,
+      "epoch": 2.2667315385447075,
+      "grad_norm": 0.19400948286056519,
+      "learning_rate": 2.997181268980552e-05,
+      "loss": 0.3995,
       "step": 62895
     },
     {
-      "epoch": 2.21,
-      "learning_rate": 3.08204407696991e-05,
-      "loss": 0.2703,
+      "epoch": 2.2669117382059323,
+      "grad_norm": 0.18428564071655273,
+      "learning_rate": 2.9968952810931044e-05,
+      "loss": 0.4017,
       "step": 62900
     },
     {
-      "epoch": 2.21,
-      "learning_rate": 3.081767029229435e-05,
-      "loss": 0.2616,
+      "epoch": 2.267091937867157,
+      "grad_norm": 0.20284946262836456,
+      "learning_rate": 2.9966092864355556e-05,
+      "loss": 0.3902,
       "step": 62905
     },
     {
-      "epoch": 2.21,
-      "learning_rate": 3.0814899739350996e-05,
-      "loss": 0.2762,
+      "epoch": 2.2672721375283813,
+      "grad_norm": 0.21810269355773926,
+      "learning_rate": 2.9963232850118006e-05,
+      "loss": 0.4044,
       "step": 62910
     },
     {
-      "epoch": 2.21,
-      "learning_rate": 3.0812129110905e-05,
-      "loss": 0.2624,
+      "epoch": 2.267452337189606,
+      "grad_norm": 0.18810847401618958,
+      "learning_rate": 2.9960372768257378e-05,
+      "loss": 0.4189,
       "step": 62915
     },
     {
-      "epoch": 2.21,
-      "learning_rate": 3.080935840699234e-05,
-      "loss": 0.2833,
+      "epoch": 2.2676325368508308,
+      "grad_norm": 0.2173568606376648,
+      "learning_rate": 2.9957512618812634e-05,
+      "loss": 0.3888,
       "step": 62920
     },
     {
-      "epoch": 2.21,
-      "learning_rate": 3.080658762764898e-05,
-      "loss": 0.2786,
+      "epoch": 2.2678127365120555,
+      "grad_norm": 0.2114708125591278,
+      "learning_rate": 2.9954652401822732e-05,
+      "loss": 0.4262,
       "step": 62925
     },
     {
-      "epoch": 2.21,
-      "learning_rate": 3.080381677291091e-05,
-      "loss": 0.2778,
+      "epoch": 2.26799293617328,
+      "grad_norm": 0.20578879117965698,
+      "learning_rate": 2.9951792117326648e-05,
+      "loss": 0.3686,
       "step": 62930
     },
     {
-      "epoch": 2.21,
-      "learning_rate": 3.080104584281411e-05,
-      "loss": 0.27,
+      "epoch": 2.2681731358345045,
+      "grad_norm": 0.18849804997444153,
+      "learning_rate": 2.9948931765363364e-05,
+      "loss": 0.4066,
       "step": 62935
     },
     {
-      "epoch": 2.21,
-      "learning_rate": 3.0798274837394554e-05,
-      "loss": 0.2539,
+      "epoch": 2.2683533354957293,
+      "grad_norm": 0.2172732949256897,
+      "learning_rate": 2.9946071345971842e-05,
+      "loss": 0.4361,
       "step": 62940
     },
     {
-      "epoch": 2.21,
-      "learning_rate": 3.079550375668821e-05,
-      "loss": 0.2617,
+      "epoch": 2.268533535156954,
+      "grad_norm": 0.19015686213970184,
+      "learning_rate": 2.994321085919105e-05,
+      "loss": 0.4048,
       "step": 62945
     },
     {
-      "epoch": 2.21,
-      "learning_rate": 3.0792732600731075e-05,
-      "loss": 0.2757,
+      "epoch": 2.2687137348181787,
+      "grad_norm": 0.18530982732772827,
+      "learning_rate": 2.9940350305059972e-05,
+      "loss": 0.3978,
       "step": 62950
     },
     {
-      "epoch": 2.21,
-      "learning_rate": 3.078996136955912e-05,
-      "loss": 0.2698,
+      "epoch": 2.268893934479403,
+      "grad_norm": 0.2398093044757843,
+      "learning_rate": 2.9937489683617577e-05,
+      "loss": 0.4006,
       "step": 62955
     },
     {
-      "epoch": 2.22,
-      "learning_rate": 3.078719006320834e-05,
-      "loss": 0.2779,
+      "epoch": 2.2690741341406278,
+      "grad_norm": 0.18880394101142883,
+      "learning_rate": 2.9934628994902836e-05,
+      "loss": 0.3719,
       "step": 62960
     },
     {
-      "epoch": 2.22,
-      "learning_rate": 3.078441868171471e-05,
-      "loss": 0.2606,
+      "epoch": 2.2692543338018525,
+      "grad_norm": 0.21370282769203186,
+      "learning_rate": 2.993176823895474e-05,
+      "loss": 0.4061,
       "step": 62965
     },
     {
-      "epoch": 2.22,
-      "learning_rate": 3.078164722511421e-05,
-      "loss": 0.2839,
+      "epoch": 2.269434533463077,
+      "grad_norm": 0.20947441458702087,
+      "learning_rate": 2.992890741581224e-05,
+      "loss": 0.4062,
       "step": 62970
     },
     {
-      "epoch": 2.22,
-      "learning_rate": 3.077887569344283e-05,
-      "loss": 0.2678,
+      "epoch": 2.2696147331243015,
+      "grad_norm": 0.18674728274345398,
+      "learning_rate": 2.9926046525514345e-05,
+      "loss": 0.3753,
       "step": 62975
     },
     {
-      "epoch": 2.22,
-      "learning_rate": 3.0776104086736556e-05,
-      "loss": 0.2597,
+      "epoch": 2.2697949327855262,
+      "grad_norm": 0.19669872522354126,
+      "learning_rate": 2.9923185568100014e-05,
+      "loss": 0.3642,
       "step": 62980
     },
     {
-      "epoch": 2.22,
-      "learning_rate": 3.077333240503139e-05,
-      "loss": 0.2734,
+      "epoch": 2.269975132446751,
+      "grad_norm": 0.18584516644477844,
+      "learning_rate": 2.992032454360824e-05,
+      "loss": 0.373,
       "step": 62985
     },
     {
-      "epoch": 2.22,
-      "learning_rate": 3.07705606483633e-05,
-      "loss": 0.2606,
+      "epoch": 2.2701553321079757,
+      "grad_norm": 0.19654756784439087,
+      "learning_rate": 2.9917463452077986e-05,
+      "loss": 0.3898,
       "step": 62990
     },
     {
-      "epoch": 2.22,
-      "learning_rate": 3.0767788816768285e-05,
-      "loss": 0.2728,
+      "epoch": 2.2703355317692004,
+      "grad_norm": 0.18995268642902374,
+      "learning_rate": 2.991460229354825e-05,
+      "loss": 0.4269,
       "step": 62995
     },
     {
-      "epoch": 2.22,
-      "learning_rate": 3.076501691028233e-05,
-      "loss": 0.2611,
+      "epoch": 2.2705157314304247,
+      "grad_norm": 0.2119891196489334,
+      "learning_rate": 2.9911741068058012e-05,
+      "loss": 0.3993,
       "step": 63000
     },
     {
-      "epoch": 2.22,
-      "eval_loss": 0.2684405446052551,
-      "eval_runtime": 10.5301,
-      "eval_samples_per_second": 9.497,
-      "eval_steps_per_second": 9.497,
+      "epoch": 2.2705157314304247,
+      "eval_loss": 0.4354766011238098,
+      "eval_runtime": 3.5274,
+      "eval_samples_per_second": 28.349,
+      "eval_steps_per_second": 7.087,
       "step": 63000
     },
     {
-      "epoch": 2.22,
-      "learning_rate": 3.076224492894143e-05,
-      "loss": 0.2829,
+      "epoch": 2.2706959310916495,
+      "grad_norm": 0.22078485786914825,
+      "learning_rate": 2.9908879775646247e-05,
+      "loss": 0.3753,
       "step": 63005
     },
     {
-      "epoch": 2.22,
-      "learning_rate": 3.0759472872781576e-05,
-      "loss": 0.2912,
+      "epoch": 2.270876130752874,
+      "grad_norm": 0.20009629428386688,
+      "learning_rate": 2.990601841635195e-05,
+      "loss": 0.4031,
       "step": 63010
     },
     {
-      "epoch": 2.22,
-      "learning_rate": 3.075670074183877e-05,
-      "loss": 0.2739,
+      "epoch": 2.271056330414099,
+      "grad_norm": 0.17347082495689392,
+      "learning_rate": 2.9903156990214097e-05,
+      "loss": 0.3555,
       "step": 63015
     },
     {
-      "epoch": 2.22,
-      "learning_rate": 3.0753928536148986e-05,
-      "loss": 0.2833,
+      "epoch": 2.2712365300753232,
+      "grad_norm": 0.261214941740036,
+      "learning_rate": 2.9900295497271687e-05,
+      "loss": 0.3902,
       "step": 63020
     },
     {
-      "epoch": 2.22,
-      "learning_rate": 3.075115625574824e-05,
-      "loss": 0.2689,
+      "epoch": 2.271416729736548,
+      "grad_norm": 0.19732660055160522,
+      "learning_rate": 2.9897433937563696e-05,
+      "loss": 0.4313,
       "step": 63025
     },
     {
-      "epoch": 2.22,
-      "learning_rate": 3.074838390067252e-05,
-      "loss": 0.2743,
+      "epoch": 2.2715969293977727,
+      "grad_norm": 0.22915062308311462,
+      "learning_rate": 2.989457231112911e-05,
+      "loss": 0.4075,
       "step": 63030
     },
     {
-      "epoch": 2.22,
-      "learning_rate": 3.074561147095782e-05,
-      "loss": 0.2613,
+      "epoch": 2.2717771290589974,
+      "grad_norm": 0.20028504729270935,
+      "learning_rate": 2.989171061800693e-05,
+      "loss": 0.4022,
       "step": 63035
     },
     {
-      "epoch": 2.22,
-      "learning_rate": 3.074283896664014e-05,
-      "loss": 0.2802,
+      "epoch": 2.271957328720222,
+      "grad_norm": 0.23854713141918182,
+      "learning_rate": 2.988884885823614e-05,
+      "loss": 0.3984,
       "step": 63040
     },
     {
-      "epoch": 2.22,
-      "learning_rate": 3.0740066387755496e-05,
-      "loss": 0.2777,
+      "epoch": 2.272137528381447,
+      "grad_norm": 0.2059304416179657,
+      "learning_rate": 2.9885987031855733e-05,
+      "loss": 0.4019,
       "step": 63045
     },
     {
-      "epoch": 2.22,
-      "learning_rate": 3.073729373433986e-05,
-      "loss": 0.2518,
+      "epoch": 2.272317728042671,
+      "grad_norm": 0.16523705422878265,
+      "learning_rate": 2.9883125138904693e-05,
+      "loss": 0.3699,
       "step": 63050
     },
     {
-      "epoch": 2.22,
-      "learning_rate": 3.073452100642924e-05,
-      "loss": 0.2616,
+      "epoch": 2.272497927703896,
+      "grad_norm": 0.20402154326438904,
+      "learning_rate": 2.988026317942202e-05,
+      "loss": 0.3868,
       "step": 63055
     },
     {
-      "epoch": 2.22,
-      "learning_rate": 3.073174820405965e-05,
-      "loss": 0.277,
+      "epoch": 2.2726781273651206,
+      "grad_norm": 0.22401557862758636,
+      "learning_rate": 2.987740115344671e-05,
+      "loss": 0.3785,
       "step": 63060
     },
     {
-      "epoch": 2.22,
-      "learning_rate": 3.07289753272671e-05,
-      "loss": 0.3092,
+      "epoch": 2.2728583270263454,
+      "grad_norm": 0.21607942879199982,
+      "learning_rate": 2.9874539061017746e-05,
+      "loss": 0.3896,
       "step": 63065
     },
     {
-      "epoch": 2.22,
-      "learning_rate": 3.072620237608756e-05,
-      "loss": 0.2681,
+      "epoch": 2.2730385266875697,
+      "grad_norm": 0.2318323701620102,
+      "learning_rate": 2.987167690217414e-05,
+      "loss": 0.4374,
       "step": 63070
     },
     {
-      "epoch": 2.22,
-      "learning_rate": 3.072342935055707e-05,
-      "loss": 0.2587,
+      "epoch": 2.2732187263487944,
+      "grad_norm": 0.20749138295650482,
+      "learning_rate": 2.986881467695487e-05,
+      "loss": 0.4096,
       "step": 63075
     },
     {
-      "epoch": 2.22,
-      "learning_rate": 3.0720656250711614e-05,
-      "loss": 0.2689,
+      "epoch": 2.273398926010019,
+      "grad_norm": 0.20667487382888794,
+      "learning_rate": 2.9865952385398955e-05,
+      "loss": 0.4047,
       "step": 63080
     },
     {
-      "epoch": 2.22,
-      "learning_rate": 3.07178830765872e-05,
-      "loss": 0.2684,
+      "epoch": 2.273579125671244,
+      "grad_norm": 0.23267313838005066,
+      "learning_rate": 2.9863090027545364e-05,
+      "loss": 0.4173,
       "step": 63085
     },
     {
-      "epoch": 2.22,
-      "learning_rate": 3.0715109828219856e-05,
-      "loss": 0.2542,
+      "epoch": 2.2737593253324686,
+      "grad_norm": 0.19276480376720428,
+      "learning_rate": 2.986022760343313e-05,
+      "loss": 0.3679,
       "step": 63090
     },
     {
-      "epoch": 2.22,
-      "learning_rate": 3.0712336505645573e-05,
-      "loss": 0.2734,
+      "epoch": 2.273939524993693,
+      "grad_norm": 0.20015855133533478,
+      "learning_rate": 2.985736511310122e-05,
+      "loss": 0.38,
       "step": 63095
     },
     {
-      "epoch": 2.22,
-      "learning_rate": 3.070956310890037e-05,
-      "loss": 0.2789,
+      "epoch": 2.2741197246549176,
+      "grad_norm": 0.2319357693195343,
+      "learning_rate": 2.985450255658866e-05,
+      "loss": 0.3908,
       "step": 63100
     },
     {
-      "epoch": 2.22,
-      "learning_rate": 3.070678963802025e-05,
-      "loss": 0.2828,
+      "epoch": 2.2742999243161424,
+      "grad_norm": 0.2397158294916153,
+      "learning_rate": 2.9851639933934446e-05,
+      "loss": 0.3892,
       "step": 63105
     },
     {
-      "epoch": 2.22,
-      "learning_rate": 3.070401609304122e-05,
-      "loss": 0.3018,
+      "epoch": 2.274480123977367,
+      "grad_norm": 0.162200927734375,
+      "learning_rate": 2.9848777245177568e-05,
+      "loss": 0.4061,
       "step": 63110
     },
     {
-      "epoch": 2.22,
-      "learning_rate": 3.0701242473999304e-05,
-      "loss": 0.2615,
+      "epoch": 2.2746603236385914,
+      "grad_norm": 0.2283955216407776,
+      "learning_rate": 2.9845914490357046e-05,
+      "loss": 0.4059,
       "step": 63115
     },
     {
-      "epoch": 2.22,
-      "learning_rate": 3.069846878093052e-05,
-      "loss": 0.2628,
+      "epoch": 2.274840523299816,
+      "grad_norm": 0.20629531145095825,
+      "learning_rate": 2.9843051669511872e-05,
+      "loss": 0.3769,
       "step": 63120
     },
     {
-      "epoch": 2.22,
-      "learning_rate": 3.069569501387086e-05,
-      "loss": 0.2609,
+      "epoch": 2.275020722961041,
+      "grad_norm": 0.16891133785247803,
+      "learning_rate": 2.9840188782681062e-05,
+      "loss": 0.3967,
       "step": 63125
     },
     {
-      "epoch": 2.22,
-      "learning_rate": 3.069292117285637e-05,
-      "loss": 0.2709,
+      "epoch": 2.2752009226222656,
+      "grad_norm": 0.19010712206363678,
+      "learning_rate": 2.983732582990361e-05,
+      "loss": 0.4143,
       "step": 63130
     },
     {
-      "epoch": 2.22,
-      "learning_rate": 3.069014725792304e-05,
-      "loss": 0.294,
+      "epoch": 2.2753811222834903,
+      "grad_norm": 0.2108030766248703,
+      "learning_rate": 2.9834462811218534e-05,
+      "loss": 0.3775,
       "step": 63135
     },
     {
-      "epoch": 2.22,
-      "learning_rate": 3.0687373269106894e-05,
-      "loss": 0.2662,
+      "epoch": 2.2755613219447146,
+      "grad_norm": 0.21494841575622559,
+      "learning_rate": 2.9831599726664844e-05,
+      "loss": 0.3723,
       "step": 63140
     },
     {
-      "epoch": 2.22,
-      "learning_rate": 3.068459920644396e-05,
-      "loss": 0.289,
+      "epoch": 2.2757415216059393,
+      "grad_norm": 0.19775624573230743,
+      "learning_rate": 2.9828736576281535e-05,
+      "loss": 0.4012,
       "step": 63145
     },
     {
-      "epoch": 2.22,
-      "learning_rate": 3.068182506997026e-05,
-      "loss": 0.2689,
+      "epoch": 2.275921721267164,
+      "grad_norm": 0.20917865633964539,
+      "learning_rate": 2.9825873360107626e-05,
+      "loss": 0.3865,
       "step": 63150
     },
     {
-      "epoch": 2.22,
-      "learning_rate": 3.067905085972179e-05,
-      "loss": 0.2564,
+      "epoch": 2.276101920928389,
+      "grad_norm": 0.1742408573627472,
+      "learning_rate": 2.982301007818213e-05,
+      "loss": 0.3552,
       "step": 63155
     },
     {
-      "epoch": 2.22,
-      "learning_rate": 3.06762765757346e-05,
-      "loss": 0.2651,
+      "epoch": 2.276282120589613,
+      "grad_norm": 0.20653748512268066,
+      "learning_rate": 2.9820146730544052e-05,
+      "loss": 0.3645,
       "step": 63160
     },
     {
-      "epoch": 2.22,
-      "learning_rate": 3.0673502218044694e-05,
-      "loss": 0.2759,
+      "epoch": 2.276462320250838,
+      "grad_norm": 0.20753364264965057,
+      "learning_rate": 2.9817283317232413e-05,
+      "loss": 0.3959,
       "step": 63165
     },
     {
-      "epoch": 2.22,
-      "learning_rate": 3.067072778668809e-05,
-      "loss": 0.2634,
+      "epoch": 2.2766425199120626,
+      "grad_norm": 0.20545120537281036,
+      "learning_rate": 2.9814419838286213e-05,
+      "loss": 0.4067,
       "step": 63170
     },
     {
-      "epoch": 2.22,
-      "learning_rate": 3.0667953281700834e-05,
-      "loss": 0.2703,
+      "epoch": 2.2768227195732873,
+      "grad_norm": 0.1817047894001007,
+      "learning_rate": 2.981155629374448e-05,
+      "loss": 0.4138,
       "step": 63175
     },
     {
-      "epoch": 2.22,
-      "learning_rate": 3.0665178703118945e-05,
-      "loss": 0.261,
+      "epoch": 2.277002919234512,
+      "grad_norm": 0.1619819700717926,
+      "learning_rate": 2.980869268364622e-05,
+      "loss": 0.4093,
       "step": 63180
     },
     {
-      "epoch": 2.22,
-      "learning_rate": 3.066240405097844e-05,
-      "loss": 0.257,
+      "epoch": 2.2771831188957363,
+      "grad_norm": 0.18216536939144135,
+      "learning_rate": 2.9805829008030466e-05,
+      "loss": 0.388,
       "step": 63185
     },
     {
-      "epoch": 2.22,
-      "learning_rate": 3.065962932531534e-05,
-      "loss": 0.2566,
+      "epoch": 2.277363318556961,
+      "grad_norm": 0.23608942329883575,
+      "learning_rate": 2.9802965266936213e-05,
+      "loss": 0.3851,
       "step": 63190
     },
     {
-      "epoch": 2.22,
-      "learning_rate": 3.0656854526165695e-05,
-      "loss": 0.2697,
+      "epoch": 2.277543518218186,
+      "grad_norm": 0.22420279681682587,
+      "learning_rate": 2.9800101460402484e-05,
+      "loss": 0.3876,
       "step": 63195
     },
     {
-      "epoch": 2.22,
-      "learning_rate": 3.065407965356552e-05,
-      "loss": 0.2739,
+      "epoch": 2.2777237178794105,
+      "grad_norm": 0.20337367057800293,
+      "learning_rate": 2.9797237588468308e-05,
+      "loss": 0.398,
       "step": 63200
     },
     {
-      "epoch": 2.22,
-      "learning_rate": 3.0651304707550846e-05,
-      "loss": 0.2766,
+      "epoch": 2.277903917540635,
+      "grad_norm": 0.20395952463150024,
+      "learning_rate": 2.9794373651172696e-05,
+      "loss": 0.4199,
       "step": 63205
     },
     {
-      "epoch": 2.22,
-      "learning_rate": 3.0648529688157704e-05,
-      "loss": 0.2808,
+      "epoch": 2.2780841172018595,
+      "grad_norm": 0.19904713332653046,
+      "learning_rate": 2.9791509648554678e-05,
+      "loss": 0.3991,
       "step": 63210
     },
     {
-      "epoch": 2.22,
-      "learning_rate": 3.064575459542213e-05,
-      "loss": 0.2638,
+      "epoch": 2.2782643168630843,
+      "grad_norm": 0.17146888375282288,
+      "learning_rate": 2.9788645580653257e-05,
+      "loss": 0.3952,
       "step": 63215
     },
     {
-      "epoch": 2.22,
-      "learning_rate": 3.0642979429380144e-05,
-      "loss": 0.2747,
+      "epoch": 2.278444516524309,
+      "grad_norm": 0.2011823207139969,
+      "learning_rate": 2.9785781447507476e-05,
+      "loss": 0.3915,
       "step": 63220
     },
     {
-      "epoch": 2.22,
-      "learning_rate": 3.064020419006779e-05,
-      "loss": 0.2562,
+      "epoch": 2.2786247161855337,
+      "grad_norm": 0.2076421082019806,
+      "learning_rate": 2.9782917249156346e-05,
+      "loss": 0.386,
       "step": 63225
     },
     {
-      "epoch": 2.22,
-      "learning_rate": 3.0637428877521115e-05,
-      "loss": 0.2634,
+      "epoch": 2.278804915846758,
+      "grad_norm": 0.18694275617599487,
+      "learning_rate": 2.9780052985638896e-05,
+      "loss": 0.3882,
       "step": 63230
     },
     {
-      "epoch": 2.22,
-      "learning_rate": 3.063465349177613e-05,
-      "loss": 0.2749,
+      "epoch": 2.2789851155079828,
+      "grad_norm": 0.2005622684955597,
+      "learning_rate": 2.9777188656994147e-05,
+      "loss": 0.3676,
       "step": 63235
     },
     {
-      "epoch": 2.22,
-      "learning_rate": 3.063187803286889e-05,
-      "loss": 0.2611,
+      "epoch": 2.2791653151692075,
+      "grad_norm": 0.17751555144786835,
+      "learning_rate": 2.9774324263261126e-05,
+      "loss": 0.3941,
       "step": 63240
     },
     {
-      "epoch": 2.23,
-      "learning_rate": 3.062910250083542e-05,
-      "loss": 0.274,
+      "epoch": 2.2793455148304322,
+      "grad_norm": 0.20505878329277039,
+      "learning_rate": 2.9771459804478868e-05,
+      "loss": 0.3837,
       "step": 63245
     },
     {
-      "epoch": 2.23,
-      "learning_rate": 3.062632689571176e-05,
-      "loss": 0.2638,
+      "epoch": 2.2795257144916565,
+      "grad_norm": 0.20655593276023865,
+      "learning_rate": 2.976859528068639e-05,
+      "loss": 0.4162,
       "step": 63250
     },
     {
-      "epoch": 2.23,
-      "learning_rate": 3.062355121753396e-05,
-      "loss": 0.2689,
+      "epoch": 2.2797059141528813,
+      "grad_norm": 0.21149560809135437,
+      "learning_rate": 2.9765730691922723e-05,
+      "loss": 0.3893,
       "step": 63255
     },
     {
-      "epoch": 2.23,
-      "learning_rate": 3.062077546633805e-05,
-      "loss": 0.2442,
+      "epoch": 2.279886113814106,
+      "grad_norm": 0.2167680561542511,
+      "learning_rate": 2.97628660382269e-05,
+      "loss": 0.4075,
       "step": 63260
     },
     {
-      "epoch": 2.23,
-      "learning_rate": 3.061799964216008e-05,
-      "loss": 0.278,
+      "epoch": 2.2800663134753307,
+      "grad_norm": 0.23179109394550323,
+      "learning_rate": 2.9760001319637955e-05,
+      "loss": 0.3927,
       "step": 63265
     },
     {
-      "epoch": 2.23,
-      "learning_rate": 3.061522374503608e-05,
-      "loss": 0.2693,
+      "epoch": 2.2802465131365555,
+      "grad_norm": 0.19186845421791077,
+      "learning_rate": 2.97571365361949e-05,
+      "loss": 0.3835,
       "step": 63270
     },
     {
-      "epoch": 2.23,
-      "learning_rate": 3.06124477750021e-05,
-      "loss": 0.2743,
+      "epoch": 2.2804267127977798,
+      "grad_norm": 0.21390306949615479,
+      "learning_rate": 2.975427168793679e-05,
+      "loss": 0.4458,
       "step": 63275
     },
     {
-      "epoch": 2.23,
-      "learning_rate": 3.060967173209419e-05,
-      "loss": 0.2704,
+      "epoch": 2.2806069124590045,
+      "grad_norm": 0.2087387591600418,
+      "learning_rate": 2.9751406774902645e-05,
+      "loss": 0.3978,
       "step": 63280
     },
     {
-      "epoch": 2.23,
-      "learning_rate": 3.060689561634838e-05,
-      "loss": 0.25,
+      "epoch": 2.280787112120229,
+      "grad_norm": 0.2157135307788849,
+      "learning_rate": 2.9748541797131506e-05,
+      "loss": 0.4081,
       "step": 63285
     },
     {
-      "epoch": 2.23,
-      "learning_rate": 3.060411942780074e-05,
-      "loss": 0.2897,
+      "epoch": 2.280967311781454,
+      "grad_norm": 0.18895265460014343,
+      "learning_rate": 2.9745676754662405e-05,
+      "loss": 0.3776,
       "step": 63290
     },
     {
-      "epoch": 2.23,
-      "learning_rate": 3.06013431664873e-05,
-      "loss": 0.2894,
+      "epoch": 2.2811475114426782,
+      "grad_norm": 0.2101006656885147,
+      "learning_rate": 2.9742811647534373e-05,
+      "loss": 0.3925,
       "step": 63295
     },
     {
-      "epoch": 2.23,
-      "learning_rate": 3.05985668324441e-05,
-      "loss": 0.2574,
+      "epoch": 2.281327711103903,
+      "grad_norm": 0.21534328162670135,
+      "learning_rate": 2.9739946475786452e-05,
+      "loss": 0.4033,
       "step": 63300
     },
     {
-      "epoch": 2.23,
-      "learning_rate": 3.059579042570721e-05,
-      "loss": 0.2719,
+      "epoch": 2.2815079107651277,
+      "grad_norm": 0.17622357606887817,
+      "learning_rate": 2.9737081239457683e-05,
+      "loss": 0.41,
       "step": 63305
     },
     {
-      "epoch": 2.23,
-      "learning_rate": 3.059301394631266e-05,
-      "loss": 0.2507,
+      "epoch": 2.2816881104263524,
+      "grad_norm": 0.2728975713253021,
+      "learning_rate": 2.973421593858709e-05,
+      "loss": 0.4151,
       "step": 63310
     },
     {
-      "epoch": 2.23,
-      "learning_rate": 3.059023739429651e-05,
-      "loss": 0.2601,
+      "epoch": 2.281868310087577,
+      "grad_norm": 0.18991267681121826,
+      "learning_rate": 2.973135057321373e-05,
+      "loss": 0.426,
       "step": 63315
     },
     {
-      "epoch": 2.23,
-      "learning_rate": 3.058746076969482e-05,
-      "loss": 0.2822,
+      "epoch": 2.282048509748802,
+      "grad_norm": 0.16258344054222107,
+      "learning_rate": 2.972848514337662e-05,
+      "loss": 0.3736,
       "step": 63320
     },
     {
-      "epoch": 2.23,
-      "learning_rate": 3.058468407254363e-05,
-      "loss": 0.2661,
+      "epoch": 2.282228709410026,
+      "grad_norm": 0.19868828356266022,
+      "learning_rate": 2.972561964911484e-05,
+      "loss": 0.4063,
       "step": 63325
     },
     {
-      "epoch": 2.23,
-      "learning_rate": 3.058190730287899e-05,
-      "loss": 0.2667,
+      "epoch": 2.282408909071251,
+      "grad_norm": 0.21427805721759796,
+      "learning_rate": 2.9722754090467385e-05,
+      "loss": 0.3628,
       "step": 63330
     },
     {
-      "epoch": 2.23,
-      "learning_rate": 3.0579130460736974e-05,
-      "loss": 0.2643,
+      "epoch": 2.2825891087324757,
+      "grad_norm": 0.16630151867866516,
+      "learning_rate": 2.9719888467473333e-05,
+      "loss": 0.3724,
       "step": 63335
     },
     {
-      "epoch": 2.23,
-      "learning_rate": 3.057635354615362e-05,
-      "loss": 0.2804,
+      "epoch": 2.2827693083937004,
+      "grad_norm": 0.20577584207057953,
+      "learning_rate": 2.9717022780171704e-05,
+      "loss": 0.4154,
       "step": 63340
     },
     {
-      "epoch": 2.23,
-      "learning_rate": 3.057357655916499e-05,
-      "loss": 0.2649,
+      "epoch": 2.2829495080549247,
+      "grad_norm": 0.24059414863586426,
+      "learning_rate": 2.9714157028601558e-05,
+      "loss": 0.3972,
       "step": 63345
     },
     {
-      "epoch": 2.23,
-      "learning_rate": 3.0570799499807145e-05,
-      "loss": 0.2675,
+      "epoch": 2.2831297077161494,
+      "grad_norm": 0.1928698569536209,
+      "learning_rate": 2.971129121280194e-05,
+      "loss": 0.3794,
       "step": 63350
     },
     {
-      "epoch": 2.23,
-      "learning_rate": 3.056802236811614e-05,
-      "loss": 0.2487,
+      "epoch": 2.283309907377374,
+      "grad_norm": 0.20987224578857422,
+      "learning_rate": 2.9708425332811883e-05,
+      "loss": 0.3816,
       "step": 63355
     },
     {
-      "epoch": 2.23,
-      "learning_rate": 3.056524516412803e-05,
-      "loss": 0.2895,
+      "epoch": 2.283490107038599,
+      "grad_norm": 0.23037847876548767,
+      "learning_rate": 2.9705559388670446e-05,
+      "loss": 0.4389,
       "step": 63360
     },
     {
-      "epoch": 2.23,
-      "learning_rate": 3.0562467887878876e-05,
-      "loss": 0.2858,
+      "epoch": 2.2836703066998236,
+      "grad_norm": 0.19015249609947205,
+      "learning_rate": 2.970269338041668e-05,
+      "loss": 0.3849,
       "step": 63365
     },
     {
-      "epoch": 2.23,
-      "learning_rate": 3.055969053940475e-05,
-      "loss": 0.2805,
+      "epoch": 2.283850506361048,
+      "grad_norm": 0.21437907218933105,
+      "learning_rate": 2.9699827308089617e-05,
+      "loss": 0.4241,
       "step": 63370
     },
     {
-      "epoch": 2.23,
-      "learning_rate": 3.05569131187417e-05,
-      "loss": 0.2549,
+      "epoch": 2.2840307060222726,
+      "grad_norm": 0.2129935771226883,
+      "learning_rate": 2.969696117172832e-05,
+      "loss": 0.402,
       "step": 63375
     },
     {
-      "epoch": 2.23,
-      "learning_rate": 3.0554135625925806e-05,
-      "loss": 0.2794,
+      "epoch": 2.2842109056834974,
+      "grad_norm": 0.2034277319908142,
+      "learning_rate": 2.969409497137184e-05,
+      "loss": 0.4077,
       "step": 63380
     },
     {
-      "epoch": 2.23,
-      "learning_rate": 3.055135806099311e-05,
-      "loss": 0.2655,
+      "epoch": 2.284391105344722,
+      "grad_norm": 0.21894314885139465,
+      "learning_rate": 2.9691228707059216e-05,
+      "loss": 0.4069,
       "step": 63385
     },
     {
-      "epoch": 2.23,
-      "learning_rate": 3.054858042397969e-05,
-      "loss": 0.303,
+      "epoch": 2.2845713050059464,
+      "grad_norm": 0.20878823101520538,
+      "learning_rate": 2.9688362378829514e-05,
+      "loss": 0.3912,
       "step": 63390
     },
     {
-      "epoch": 2.23,
-      "learning_rate": 3.054580271492161e-05,
-      "loss": 0.2728,
+      "epoch": 2.284751504667171,
+      "grad_norm": 0.21701130270957947,
+      "learning_rate": 2.968549598672179e-05,
+      "loss": 0.3831,
       "step": 63395
     },
     {
-      "epoch": 2.23,
-      "learning_rate": 3.0543024933854944e-05,
-      "loss": 0.2585,
+      "epoch": 2.284931704328396,
+      "grad_norm": 0.20921416580677032,
+      "learning_rate": 2.9682629530775075e-05,
+      "loss": 0.3879,
       "step": 63400
     },
     {
-      "epoch": 2.23,
-      "learning_rate": 3.054024708081575e-05,
-      "loss": 0.2828,
+      "epoch": 2.2851119039896206,
+      "grad_norm": 0.22157233953475952,
+      "learning_rate": 2.967976301102845e-05,
+      "loss": 0.4179,
       "step": 63405
     },
     {
-      "epoch": 2.23,
-      "learning_rate": 3.0537469155840095e-05,
-      "loss": 0.2673,
+      "epoch": 2.2852921036508453,
+      "grad_norm": 0.2359781712293625,
+      "learning_rate": 2.967689642752096e-05,
+      "loss": 0.3867,
       "step": 63410
     },
     {
-      "epoch": 2.23,
-      "learning_rate": 3.053469115896405e-05,
-      "loss": 0.2639,
+      "epoch": 2.2854723033120696,
+      "grad_norm": 0.18056300282478333,
+      "learning_rate": 2.967402978029166e-05,
+      "loss": 0.3854,
       "step": 63415
     },
     {
-      "epoch": 2.23,
-      "learning_rate": 3.0531913090223693e-05,
-      "loss": 0.2824,
+      "epoch": 2.2856525029732944,
+      "grad_norm": 0.1796717792749405,
+      "learning_rate": 2.967116306937961e-05,
+      "loss": 0.3722,
       "step": 63420
     },
     {
-      "epoch": 2.23,
-      "learning_rate": 3.052913494965509e-05,
-      "loss": 0.2922,
+      "epoch": 2.285832702634519,
+      "grad_norm": 0.21330176293849945,
+      "learning_rate": 2.9668296294823862e-05,
+      "loss": 0.4205,
       "step": 63425
     },
     {
-      "epoch": 2.23,
-      "learning_rate": 3.0526356737294306e-05,
-      "loss": 0.2759,
+      "epoch": 2.286012902295744,
+      "grad_norm": 0.16301804780960083,
+      "learning_rate": 2.966542945666349e-05,
+      "loss": 0.4064,
       "step": 63430
     },
     {
-      "epoch": 2.23,
-      "learning_rate": 3.052357845317743e-05,
-      "loss": 0.2859,
+      "epoch": 2.286193101956968,
+      "grad_norm": 0.2423224300146103,
+      "learning_rate": 2.9662562554937534e-05,
+      "loss": 0.4057,
       "step": 63435
     },
     {
-      "epoch": 2.23,
-      "learning_rate": 3.0520800097340526e-05,
-      "loss": 0.2588,
+      "epoch": 2.286373301618193,
+      "grad_norm": 0.18112295866012573,
+      "learning_rate": 2.9659695589685076e-05,
+      "loss": 0.3972,
       "step": 63440
     },
     {
-      "epoch": 2.23,
-      "learning_rate": 3.051802166981967e-05,
-      "loss": 0.2786,
+      "epoch": 2.2865535012794176,
+      "grad_norm": 0.1735352873802185,
+      "learning_rate": 2.965682856094516e-05,
+      "loss": 0.391,
       "step": 63445
     },
     {
-      "epoch": 2.23,
-      "learning_rate": 3.051524317065093e-05,
-      "loss": 0.2616,
+      "epoch": 2.2867337009406423,
+      "grad_norm": 0.208638533949852,
+      "learning_rate": 2.9653961468756863e-05,
+      "loss": 0.3966,
       "step": 63450
     },
     {
-      "epoch": 2.23,
-      "learning_rate": 3.051246459987041e-05,
-      "loss": 0.2702,
+      "epoch": 2.286913900601867,
+      "grad_norm": 0.19702517986297607,
+      "learning_rate": 2.965109431315924e-05,
+      "loss": 0.412,
       "step": 63455
     },
     {
-      "epoch": 2.23,
-      "learning_rate": 3.050968595751416e-05,
-      "loss": 0.2691,
+      "epoch": 2.2870941002630913,
+      "grad_norm": 0.21106953918933868,
+      "learning_rate": 2.964822709419135e-05,
+      "loss": 0.4016,
       "step": 63460
     },
     {
-      "epoch": 2.23,
-      "learning_rate": 3.050690724361827e-05,
-      "loss": 0.2864,
+      "epoch": 2.287274299924316,
+      "grad_norm": 0.18838860094547272,
+      "learning_rate": 2.9645359811892275e-05,
+      "loss": 0.3739,
       "step": 63465
     },
     {
-      "epoch": 2.23,
-      "learning_rate": 3.0504128458218815e-05,
-      "loss": 0.2618,
+      "epoch": 2.287454499585541,
+      "grad_norm": 0.22702044248580933,
+      "learning_rate": 2.964249246630107e-05,
+      "loss": 0.3855,
       "step": 63470
     },
     {
-      "epoch": 2.23,
-      "learning_rate": 3.0501349601351887e-05,
-      "loss": 0.2828,
+      "epoch": 2.2876346992467655,
+      "grad_norm": 0.18174223601818085,
+      "learning_rate": 2.9639625057456805e-05,
+      "loss": 0.3885,
       "step": 63475
     },
     {
-      "epoch": 2.23,
-      "learning_rate": 3.0498570673053557e-05,
-      "loss": 0.2864,
+      "epoch": 2.28781489890799,
+      "grad_norm": 0.19665606319904327,
+      "learning_rate": 2.9636757585398544e-05,
+      "loss": 0.4166,
       "step": 63480
     },
     {
-      "epoch": 2.23,
-      "learning_rate": 3.049579167335991e-05,
-      "loss": 0.2753,
+      "epoch": 2.2879950985692146,
+      "grad_norm": 0.1689072996377945,
+      "learning_rate": 2.9633890050165357e-05,
+      "loss": 0.3923,
       "step": 63485
     },
     {
-      "epoch": 2.23,
-      "learning_rate": 3.049301260230703e-05,
-      "loss": 0.2759,
+      "epoch": 2.2881752982304393,
+      "grad_norm": 0.20292653143405914,
+      "learning_rate": 2.963102245179632e-05,
+      "loss": 0.3948,
       "step": 63490
     },
     {
-      "epoch": 2.23,
-      "learning_rate": 3.0490233459931e-05,
-      "loss": 0.2643,
+      "epoch": 2.288355497891664,
+      "grad_norm": 0.1763313263654709,
+      "learning_rate": 2.9628154790330498e-05,
+      "loss": 0.3818,
       "step": 63495
     },
     {
-      "epoch": 2.23,
-      "learning_rate": 3.0487454246267904e-05,
-      "loss": 0.2826,
+      "epoch": 2.2885356975528888,
+      "grad_norm": 0.2226899415254593,
+      "learning_rate": 2.9625287065806962e-05,
+      "loss": 0.393,
       "step": 63500
     },
     {
-      "epoch": 2.23,
-      "eval_loss": 0.26817554235458374,
-      "eval_runtime": 10.5266,
-      "eval_samples_per_second": 9.5,
-      "eval_steps_per_second": 9.5,
+      "epoch": 2.2885356975528888,
+      "eval_loss": 0.4357220530509949,
+      "eval_runtime": 3.5317,
+      "eval_samples_per_second": 28.315,
+      "eval_steps_per_second": 7.079,
       "step": 63500
     },
     {
-      "epoch": 2.23,
-      "learning_rate": 3.048467496135384e-05,
-      "loss": 0.2642,
+      "epoch": 2.288715897214113,
+      "grad_norm": 0.17390502989292145,
+      "learning_rate": 2.962241927826478e-05,
+      "loss": 0.4268,
       "step": 63505
     },
     {
-      "epoch": 2.23,
-      "learning_rate": 3.048189560522488e-05,
-      "loss": 0.2709,
+      "epoch": 2.288896096875338,
+      "grad_norm": 0.250418484210968,
+      "learning_rate": 2.9619551427743042e-05,
+      "loss": 0.4187,
       "step": 63510
     },
     {
-      "epoch": 2.23,
-      "learning_rate": 3.047911617791712e-05,
-      "loss": 0.2787,
+      "epoch": 2.2890762965365625,
+      "grad_norm": 0.25360366702079773,
+      "learning_rate": 2.9616683514280798e-05,
+      "loss": 0.3917,
       "step": 63515
     },
     {
-      "epoch": 2.23,
-      "learning_rate": 3.0476336679466645e-05,
-      "loss": 0.2723,
+      "epoch": 2.2892564961977873,
+      "grad_norm": 0.18742060661315918,
+      "learning_rate": 2.9613815537917145e-05,
+      "loss": 0.3793,
       "step": 63520
     },
     {
-      "epoch": 2.23,
-      "learning_rate": 3.047355710990955e-05,
-      "loss": 0.2784,
+      "epoch": 2.2894366958590116,
+      "grad_norm": 0.17973770201206207,
+      "learning_rate": 2.961094749869114e-05,
+      "loss": 0.3881,
       "step": 63525
     },
     {
-      "epoch": 2.24,
-      "learning_rate": 3.0470777469281918e-05,
-      "loss": 0.2844,
+      "epoch": 2.2896168955202363,
+      "grad_norm": 0.229409322142601,
+      "learning_rate": 2.9608079396641868e-05,
+      "loss": 0.4357,
       "step": 63530
     },
     {
-      "epoch": 2.24,
-      "learning_rate": 3.0467997757619853e-05,
-      "loss": 0.2933,
+      "epoch": 2.289797095181461,
+      "grad_norm": 0.2348993420600891,
+      "learning_rate": 2.9605211231808417e-05,
+      "loss": 0.4039,
       "step": 63535
     },
     {
-      "epoch": 2.24,
-      "learning_rate": 3.0465217974959436e-05,
-      "loss": 0.2765,
+      "epoch": 2.2899772948426858,
+      "grad_norm": 0.21660517156124115,
+      "learning_rate": 2.9602343004229842e-05,
+      "loss": 0.4293,
       "step": 63540
     },
     {
-      "epoch": 2.24,
-      "learning_rate": 3.0462438121336767e-05,
-      "loss": 0.291,
+      "epoch": 2.2901574945039105,
+      "grad_norm": 0.1841536909341812,
+      "learning_rate": 2.9599474713945242e-05,
+      "loss": 0.394,
       "step": 63545
     },
     {
-      "epoch": 2.24,
-      "learning_rate": 3.0459658196787937e-05,
-      "loss": 0.2678,
+      "epoch": 2.290337694165135,
+      "grad_norm": 0.2169848531484604,
+      "learning_rate": 2.959660636099369e-05,
+      "loss": 0.3922,
       "step": 63550
     },
     {
-      "epoch": 2.24,
-      "learning_rate": 3.045687820134904e-05,
-      "loss": 0.2615,
+      "epoch": 2.2905178938263595,
+      "grad_norm": 0.22825659811496735,
+      "learning_rate": 2.9593737945414264e-05,
+      "loss": 0.3977,
       "step": 63555
     },
     {
-      "epoch": 2.24,
-      "learning_rate": 3.0454098135056186e-05,
-      "loss": 0.2461,
+      "epoch": 2.2906980934875842,
+      "grad_norm": 0.21935486793518066,
+      "learning_rate": 2.9590869467246047e-05,
+      "loss": 0.4143,
       "step": 63560
     },
     {
-      "epoch": 2.24,
-      "learning_rate": 3.0451317997945454e-05,
-      "loss": 0.2817,
+      "epoch": 2.290878293148809,
+      "grad_norm": 0.18911194801330566,
+      "learning_rate": 2.9588000926528126e-05,
+      "loss": 0.3835,
       "step": 63565
     },
     {
-      "epoch": 2.24,
-      "learning_rate": 3.0448537790052962e-05,
-      "loss": 0.264,
+      "epoch": 2.2910584928100337,
+      "grad_norm": 0.16770300269126892,
+      "learning_rate": 2.958513232329957e-05,
+      "loss": 0.4211,
       "step": 63570
     },
     {
-      "epoch": 2.24,
-      "learning_rate": 3.0445757511414785e-05,
-      "loss": 0.2779,
+      "epoch": 2.291238692471258,
+      "grad_norm": 0.20675276219844818,
+      "learning_rate": 2.9582263657599485e-05,
+      "loss": 0.384,
       "step": 63575
     },
     {
-      "epoch": 2.24,
-      "learning_rate": 3.044297716206704e-05,
-      "loss": 0.2688,
+      "epoch": 2.2914188921324827,
+      "grad_norm": 0.26239025592803955,
+      "learning_rate": 2.9579394929466943e-05,
+      "loss": 0.394,
       "step": 63580
     },
     {
-      "epoch": 2.24,
-      "learning_rate": 3.044019674204582e-05,
-      "loss": 0.2604,
+      "epoch": 2.2915990917937075,
+      "grad_norm": 0.2336747944355011,
+      "learning_rate": 2.9576526138941025e-05,
+      "loss": 0.4154,
       "step": 63585
     },
     {
-      "epoch": 2.24,
-      "learning_rate": 3.0437416251387234e-05,
-      "loss": 0.2821,
+      "epoch": 2.291779291454932,
+      "grad_norm": 0.18274347484111786,
+      "learning_rate": 2.957365728606083e-05,
+      "loss": 0.3395,
       "step": 63590
     },
     {
-      "epoch": 2.24,
-      "learning_rate": 3.0434635690127382e-05,
-      "loss": 0.2489,
+      "epoch": 2.291959491116157,
+      "grad_norm": 0.17019642889499664,
+      "learning_rate": 2.9570788370865443e-05,
+      "loss": 0.3737,
       "step": 63595
     },
     {
-      "epoch": 2.24,
-      "learning_rate": 3.043185505830237e-05,
-      "loss": 0.2696,
+      "epoch": 2.2921396907773812,
+      "grad_norm": 0.2089766263961792,
+      "learning_rate": 2.956791939339394e-05,
+      "loss": 0.3783,
       "step": 63600
     },
     {
-      "epoch": 2.24,
-      "learning_rate": 3.0429074355948295e-05,
-      "loss": 0.2913,
+      "epoch": 2.292319890438606,
+      "grad_norm": 0.2029276192188263,
+      "learning_rate": 2.956505035368543e-05,
+      "loss": 0.4223,
       "step": 63605
     },
     {
-      "epoch": 2.24,
-      "learning_rate": 3.0426293583101267e-05,
-      "loss": 0.2934,
+      "epoch": 2.2925000900998307,
+      "grad_norm": 0.1698850840330124,
+      "learning_rate": 2.9562181251778986e-05,
+      "loss": 0.4032,
       "step": 63610
     },
     {
-      "epoch": 2.24,
-      "learning_rate": 3.0423512739797387e-05,
-      "loss": 0.2659,
+      "epoch": 2.2926802897610554,
+      "grad_norm": 0.17825381457805634,
+      "learning_rate": 2.9559312087713714e-05,
+      "loss": 0.3808,
       "step": 63615
     },
     {
-      "epoch": 2.24,
-      "learning_rate": 3.0420731826072785e-05,
-      "loss": 0.2681,
+      "epoch": 2.2928604894222797,
+      "grad_norm": 0.21084287762641907,
+      "learning_rate": 2.9556442861528688e-05,
+      "loss": 0.4042,
       "step": 63620
     },
     {
-      "epoch": 2.24,
-      "learning_rate": 3.041795084196354e-05,
-      "loss": 0.2589,
+      "epoch": 2.2930406890835044,
+      "grad_norm": 0.21861131489276886,
+      "learning_rate": 2.9553573573263016e-05,
+      "loss": 0.4222,
       "step": 63625
     },
     {
-      "epoch": 2.24,
-      "learning_rate": 3.0415169787505777e-05,
-      "loss": 0.2642,
+      "epoch": 2.293220888744729,
+      "grad_norm": 0.1877928525209427,
+      "learning_rate": 2.9550704222955778e-05,
+      "loss": 0.378,
       "step": 63630
     },
     {
-      "epoch": 2.24,
-      "learning_rate": 3.04123886627356e-05,
-      "loss": 0.2544,
+      "epoch": 2.293401088405954,
+      "grad_norm": 0.23301534354686737,
+      "learning_rate": 2.954783481064608e-05,
+      "loss": 0.407,
       "step": 63635
     },
     {
-      "epoch": 2.24,
-      "learning_rate": 3.0409607467689127e-05,
-      "loss": 0.2651,
+      "epoch": 2.2935812880671786,
+      "grad_norm": 0.2327142059803009,
+      "learning_rate": 2.954496533637302e-05,
+      "loss": 0.396,
       "step": 63640
     },
     {
-      "epoch": 2.24,
-      "learning_rate": 3.0406826202402466e-05,
-      "loss": 0.2479,
+      "epoch": 2.293761487728403,
+      "grad_norm": 0.22101731598377228,
+      "learning_rate": 2.954209580017568e-05,
+      "loss": 0.378,
       "step": 63645
     },
     {
-      "epoch": 2.24,
-      "learning_rate": 3.0404044866911736e-05,
-      "loss": 0.2829,
+      "epoch": 2.2939416873896277,
+      "grad_norm": 0.22895368933677673,
+      "learning_rate": 2.9539226202093162e-05,
+      "loss": 0.4179,
       "step": 63650
     },
     {
-      "epoch": 2.24,
-      "learning_rate": 3.0401263461253033e-05,
-      "loss": 0.2997,
+      "epoch": 2.2941218870508524,
+      "grad_norm": 0.19855442643165588,
+      "learning_rate": 2.9536356542164573e-05,
+      "loss": 0.3768,
       "step": 63655
     },
     {
-      "epoch": 2.24,
-      "learning_rate": 3.039848198546249e-05,
-      "loss": 0.2897,
+      "epoch": 2.294302086712077,
+      "grad_norm": 0.20044022798538208,
+      "learning_rate": 2.9533486820429e-05,
+      "loss": 0.3758,
       "step": 63660
     },
     {
-      "epoch": 2.24,
-      "learning_rate": 3.039570043957621e-05,
-      "loss": 0.278,
+      "epoch": 2.2944822863733014,
+      "grad_norm": 0.23310525715351105,
+      "learning_rate": 2.9530617036925545e-05,
+      "loss": 0.4063,
       "step": 63665
     },
     {
-      "epoch": 2.24,
-      "learning_rate": 3.039291882363032e-05,
-      "loss": 0.2656,
+      "epoch": 2.294662486034526,
+      "grad_norm": 0.2389223724603653,
+      "learning_rate": 2.9527747191693318e-05,
+      "loss": 0.3821,
       "step": 63670
     },
     {
-      "epoch": 2.24,
-      "learning_rate": 3.0390137137660933e-05,
-      "loss": 0.2906,
+      "epoch": 2.294842685695751,
+      "grad_norm": 0.21088966727256775,
+      "learning_rate": 2.9524877284771406e-05,
+      "loss": 0.4099,
       "step": 63675
     },
     {
-      "epoch": 2.24,
-      "learning_rate": 3.038735538170417e-05,
-      "loss": 0.28,
+      "epoch": 2.2950228853569756,
+      "grad_norm": 0.19195705652236938,
+      "learning_rate": 2.952200731619892e-05,
+      "loss": 0.3833,
       "step": 63680
     },
     {
-      "epoch": 2.24,
-      "learning_rate": 3.038457355579615e-05,
-      "loss": 0.2634,
+      "epoch": 2.2952030850182004,
+      "grad_norm": 0.1707751750946045,
+      "learning_rate": 2.9519137286014957e-05,
+      "loss": 0.3551,
       "step": 63685
     },
     {
-      "epoch": 2.24,
-      "learning_rate": 3.038179165997298e-05,
-      "loss": 0.2563,
+      "epoch": 2.2953832846794247,
+      "grad_norm": 0.23566772043704987,
+      "learning_rate": 2.9516267194258618e-05,
+      "loss": 0.3843,
       "step": 63690
     },
     {
-      "epoch": 2.24,
-      "learning_rate": 3.03790096942708e-05,
-      "loss": 0.2743,
+      "epoch": 2.2955634843406494,
+      "grad_norm": 0.19582681357860565,
+      "learning_rate": 2.9513397040969025e-05,
+      "loss": 0.4201,
       "step": 63695
     },
     {
-      "epoch": 2.24,
-      "learning_rate": 3.037622765872572e-05,
-      "loss": 0.2518,
+      "epoch": 2.295743684001874,
+      "grad_norm": 0.2070237249135971,
+      "learning_rate": 2.9510526826185263e-05,
+      "loss": 0.3976,
       "step": 63700
     },
     {
-      "epoch": 2.24,
-      "learning_rate": 3.0373445553373874e-05,
-      "loss": 0.2593,
+      "epoch": 2.295923883663099,
+      "grad_norm": 0.19391392171382904,
+      "learning_rate": 2.950765654994645e-05,
+      "loss": 0.4019,
       "step": 63705
     },
     {
-      "epoch": 2.24,
-      "learning_rate": 3.0370663378251374e-05,
-      "loss": 0.2665,
+      "epoch": 2.296104083324323,
+      "grad_norm": 0.2209123820066452,
+      "learning_rate": 2.9504786212291685e-05,
+      "loss": 0.3918,
       "step": 63710
     },
     {
-      "epoch": 2.24,
-      "learning_rate": 3.036788113339435e-05,
-      "loss": 0.2536,
+      "epoch": 2.296284282985548,
+      "grad_norm": 0.18248969316482544,
+      "learning_rate": 2.9501915813260084e-05,
+      "loss": 0.3817,
       "step": 63715
     },
     {
-      "epoch": 2.24,
-      "learning_rate": 3.0365098818838923e-05,
-      "loss": 0.2749,
+      "epoch": 2.2964644826467726,
+      "grad_norm": 0.17542307078838348,
+      "learning_rate": 2.9499045352890754e-05,
+      "loss": 0.3842,
       "step": 63720
     },
     {
-      "epoch": 2.24,
-      "learning_rate": 3.036231643462123e-05,
-      "loss": 0.2811,
+      "epoch": 2.2966446823079973,
+      "grad_norm": 0.19635504484176636,
+      "learning_rate": 2.9496174831222796e-05,
+      "loss": 0.3686,
       "step": 63725
     },
     {
-      "epoch": 2.24,
-      "learning_rate": 3.035953398077739e-05,
-      "loss": 0.2634,
+      "epoch": 2.296824881969222,
+      "grad_norm": 0.1903241127729416,
+      "learning_rate": 2.9493304248295327e-05,
+      "loss": 0.3946,
       "step": 63730
     },
     {
-      "epoch": 2.24,
-      "learning_rate": 3.0356751457343534e-05,
-      "loss": 0.2931,
+      "epoch": 2.2970050816304464,
+      "grad_norm": 0.19925326108932495,
+      "learning_rate": 2.949043360414746e-05,
+      "loss": 0.3765,
       "step": 63735
     },
     {
-      "epoch": 2.24,
-      "learning_rate": 3.035396886435579e-05,
-      "loss": 0.2714,
+      "epoch": 2.297185281291671,
+      "grad_norm": 0.19926919043064117,
+      "learning_rate": 2.9487562898818304e-05,
+      "loss": 0.4365,
       "step": 63740
     },
     {
-      "epoch": 2.24,
-      "learning_rate": 3.0351186201850286e-05,
-      "loss": 0.2481,
+      "epoch": 2.297365480952896,
+      "grad_norm": 0.23359358310699463,
+      "learning_rate": 2.9484692132346974e-05,
+      "loss": 0.3822,
       "step": 63745
     },
     {
-      "epoch": 2.24,
-      "learning_rate": 3.0348403469863156e-05,
-      "loss": 0.2698,
+      "epoch": 2.2975456806141206,
+      "grad_norm": 0.2015082836151123,
+      "learning_rate": 2.9481821304772572e-05,
+      "loss": 0.404,
       "step": 63750
     },
     {
-      "epoch": 2.24,
-      "learning_rate": 3.0345620668430535e-05,
-      "loss": 0.2541,
+      "epoch": 2.297725880275345,
+      "grad_norm": 0.2047175019979477,
+      "learning_rate": 2.9478950416134243e-05,
+      "loss": 0.4028,
       "step": 63755
     },
     {
-      "epoch": 2.24,
-      "learning_rate": 3.0342837797588552e-05,
-      "loss": 0.2757,
+      "epoch": 2.2979060799365696,
+      "grad_norm": 0.16767185926437378,
+      "learning_rate": 2.9476079466471063e-05,
+      "loss": 0.412,
       "step": 63760
     },
     {
-      "epoch": 2.24,
-      "learning_rate": 3.0340054857373344e-05,
-      "loss": 0.255,
+      "epoch": 2.2980862795977943,
+      "grad_norm": 0.21474462747573853,
+      "learning_rate": 2.9473208455822178e-05,
+      "loss": 0.386,
       "step": 63765
     },
     {
-      "epoch": 2.24,
-      "learning_rate": 3.033727184782103e-05,
-      "loss": 0.2832,
+      "epoch": 2.298266479259019,
+      "grad_norm": 0.22542273998260498,
+      "learning_rate": 2.947033738422668e-05,
+      "loss": 0.3809,
       "step": 63770
     },
     {
-      "epoch": 2.24,
-      "learning_rate": 3.033448876896777e-05,
-      "loss": 0.2753,
+      "epoch": 2.298446678920244,
+      "grad_norm": 0.18820738792419434,
+      "learning_rate": 2.9467466251723713e-05,
+      "loss": 0.4045,
       "step": 63775
     },
     {
-      "epoch": 2.24,
-      "learning_rate": 3.033170562084969e-05,
-      "loss": 0.2628,
+      "epoch": 2.298626878581468,
+      "grad_norm": 0.23645508289337158,
+      "learning_rate": 2.9464595058352383e-05,
+      "loss": 0.3902,
       "step": 63780
     },
     {
-      "epoch": 2.24,
-      "learning_rate": 3.032892240350292e-05,
-      "loss": 0.277,
+      "epoch": 2.298807078242693,
+      "grad_norm": 0.2576570510864258,
+      "learning_rate": 2.9461723804151802e-05,
+      "loss": 0.4025,
       "step": 63785
     },
     {
-      "epoch": 2.24,
-      "learning_rate": 3.032613911696361e-05,
-      "loss": 0.3037,
+      "epoch": 2.2989872779039175,
+      "grad_norm": 0.2036072313785553,
+      "learning_rate": 2.9458852489161102e-05,
+      "loss": 0.4015,
       "step": 63790
     },
     {
-      "epoch": 2.24,
-      "learning_rate": 3.0323355761267892e-05,
-      "loss": 0.2693,
+      "epoch": 2.2991674775651423,
+      "grad_norm": 0.1797807663679123,
+      "learning_rate": 2.94559811134194e-05,
+      "loss": 0.3861,
       "step": 63795
     },
     {
-      "epoch": 2.24,
-      "learning_rate": 3.03205723364519e-05,
-      "loss": 0.2608,
+      "epoch": 2.2993476772263666,
+      "grad_norm": 0.2656286954879761,
+      "learning_rate": 2.945310967696583e-05,
+      "loss": 0.3931,
       "step": 63800
     },
     {
-      "epoch": 2.24,
-      "learning_rate": 3.031778884255179e-05,
-      "loss": 0.2782,
+      "epoch": 2.2995278768875913,
+      "grad_norm": 0.2161501944065094,
+      "learning_rate": 2.9450238179839483e-05,
+      "loss": 0.4066,
       "step": 63805
     },
     {
-      "epoch": 2.25,
-      "learning_rate": 3.0315005279603696e-05,
-      "loss": 0.2889,
+      "epoch": 2.299708076548816,
+      "grad_norm": 0.24081888794898987,
+      "learning_rate": 2.9447366622079515e-05,
+      "loss": 0.4195,
       "step": 63810
     },
     {
-      "epoch": 2.25,
-      "learning_rate": 3.031222164764376e-05,
-      "loss": 0.2652,
+      "epoch": 2.2998882762100408,
+      "grad_norm": 0.18698875606060028,
+      "learning_rate": 2.9444495003725033e-05,
+      "loss": 0.3576,
       "step": 63815
     },
     {
-      "epoch": 2.25,
-      "learning_rate": 3.030943794670813e-05,
-      "loss": 0.2428,
+      "epoch": 2.3000684758712655,
+      "grad_norm": 0.22651271522045135,
+      "learning_rate": 2.9441623324815166e-05,
+      "loss": 0.4172,
       "step": 63820
     },
     {
-      "epoch": 2.25,
-      "learning_rate": 3.030665417683295e-05,
-      "loss": 0.2678,
+      "epoch": 2.3002486755324902,
+      "grad_norm": 0.23613341152668,
+      "learning_rate": 2.9438751585389047e-05,
+      "loss": 0.4038,
       "step": 63825
     },
     {
-      "epoch": 2.25,
-      "learning_rate": 3.0303870338054354e-05,
-      "loss": 0.2648,
+      "epoch": 2.3004288751937145,
+      "grad_norm": 0.19608378410339355,
+      "learning_rate": 2.9435879785485788e-05,
+      "loss": 0.3828,
       "step": 63830
     },
     {
-      "epoch": 2.25,
-      "learning_rate": 3.0301086430408498e-05,
-      "loss": 0.2691,
+      "epoch": 2.3006090748549393,
+      "grad_norm": 0.19310517609119415,
+      "learning_rate": 2.943300792514453e-05,
+      "loss": 0.3665,
       "step": 63835
     },
     {
-      "epoch": 2.25,
-      "learning_rate": 3.0298302453931533e-05,
-      "loss": 0.2628,
+      "epoch": 2.300789274516164,
+      "grad_norm": 0.23690401017665863,
+      "learning_rate": 2.9430136004404402e-05,
+      "loss": 0.4279,
       "step": 63840
     },
     {
-      "epoch": 2.25,
-      "learning_rate": 3.02955184086596e-05,
-      "loss": 0.2702,
+      "epoch": 2.3009694741773887,
+      "grad_norm": 0.201737180352211,
+      "learning_rate": 2.9427264023304523e-05,
+      "loss": 0.3869,
       "step": 63845
     },
     {
-      "epoch": 2.25,
-      "learning_rate": 3.029273429462885e-05,
-      "loss": 0.2707,
+      "epoch": 2.301149673838613,
+      "grad_norm": 0.192640483379364,
+      "learning_rate": 2.942439198188403e-05,
+      "loss": 0.4047,
       "step": 63850
     },
     {
-      "epoch": 2.25,
-      "learning_rate": 3.028995011187543e-05,
-      "loss": 0.29,
+      "epoch": 2.3013298734998378,
+      "grad_norm": 0.17696553468704224,
+      "learning_rate": 2.9421519880182047e-05,
+      "loss": 0.3942,
       "step": 63855
     },
     {
-      "epoch": 2.25,
-      "learning_rate": 3.02871658604355e-05,
-      "loss": 0.2597,
+      "epoch": 2.3015100731610625,
+      "grad_norm": 0.19767604768276215,
+      "learning_rate": 2.9418647718237724e-05,
+      "loss": 0.4076,
       "step": 63860
     },
     {
-      "epoch": 2.25,
-      "learning_rate": 3.0284381540345203e-05,
-      "loss": 0.2668,
+      "epoch": 2.301690272822287,
+      "grad_norm": 0.25544825196266174,
+      "learning_rate": 2.9415775496090174e-05,
+      "loss": 0.3982,
       "step": 63865
     },
     {
-      "epoch": 2.25,
-      "learning_rate": 3.02815971516407e-05,
-      "loss": 0.2517,
+      "epoch": 2.301870472483512,
+      "grad_norm": 0.20909026265144348,
+      "learning_rate": 2.941290321377854e-05,
+      "loss": 0.4101,
       "step": 63870
     },
     {
-      "epoch": 2.25,
-      "learning_rate": 3.0278812694358127e-05,
-      "loss": 0.2788,
+      "epoch": 2.3020506721447362,
+      "grad_norm": 0.16474294662475586,
+      "learning_rate": 2.941003087134195e-05,
+      "loss": 0.3868,
       "step": 63875
     },
     {
-      "epoch": 2.25,
-      "learning_rate": 3.027602816853366e-05,
-      "loss": 0.2658,
+      "epoch": 2.302230871805961,
+      "grad_norm": 0.18018871545791626,
+      "learning_rate": 2.940715846881955e-05,
+      "loss": 0.3999,
       "step": 63880
     },
     {
-      "epoch": 2.25,
-      "learning_rate": 3.0273243574203436e-05,
-      "loss": 0.2614,
+      "epoch": 2.3024110714671857,
+      "grad_norm": 0.20449526607990265,
+      "learning_rate": 2.9404286006250464e-05,
+      "loss": 0.4119,
       "step": 63885
     },
     {
-      "epoch": 2.25,
-      "learning_rate": 3.0270458911403625e-05,
-      "loss": 0.2683,
+      "epoch": 2.3025912711284104,
+      "grad_norm": 0.2095811516046524,
+      "learning_rate": 2.9401413483673833e-05,
+      "loss": 0.3833,
       "step": 63890
     },
     {
-      "epoch": 2.25,
-      "learning_rate": 3.026767418017038e-05,
-      "loss": 0.2708,
+      "epoch": 2.3027714707896347,
+      "grad_norm": 0.19565238058567047,
+      "learning_rate": 2.93985409011288e-05,
+      "loss": 0.3665,
       "step": 63895
     },
     {
-      "epoch": 2.25,
-      "learning_rate": 3.026488938053985e-05,
-      "loss": 0.272,
+      "epoch": 2.3029516704508595,
+      "grad_norm": 0.18948253989219666,
+      "learning_rate": 2.9395668258654497e-05,
+      "loss": 0.3988,
       "step": 63900
     },
     {
-      "epoch": 2.25,
-      "learning_rate": 3.0262104512548212e-05,
-      "loss": 0.2595,
+      "epoch": 2.303131870112084,
+      "grad_norm": 0.19964943826198578,
+      "learning_rate": 2.9392795556290064e-05,
+      "loss": 0.4107,
       "step": 63905
     },
     {
-      "epoch": 2.25,
-      "learning_rate": 3.0259319576231603e-05,
-      "loss": 0.2552,
+      "epoch": 2.303312069773309,
+      "grad_norm": 0.2215508222579956,
+      "learning_rate": 2.9389922794074643e-05,
+      "loss": 0.379,
       "step": 63910
     },
     {
-      "epoch": 2.25,
-      "learning_rate": 3.0256534571626206e-05,
-      "loss": 0.2814,
+      "epoch": 2.3034922694345337,
+      "grad_norm": 0.20707103610038757,
+      "learning_rate": 2.9387049972047376e-05,
+      "loss": 0.3797,
       "step": 63915
     },
     {
-      "epoch": 2.25,
-      "learning_rate": 3.0253749498768163e-05,
-      "loss": 0.2735,
+      "epoch": 2.303672469095758,
+      "grad_norm": 0.24852395057678223,
+      "learning_rate": 2.93841770902474e-05,
+      "loss": 0.4041,
       "step": 63920
     },
     {
-      "epoch": 2.25,
-      "learning_rate": 3.0250964357693652e-05,
-      "loss": 0.2509,
+      "epoch": 2.3038526687569827,
+      "grad_norm": 0.2447209507226944,
+      "learning_rate": 2.9381304148713856e-05,
+      "loss": 0.4051,
       "step": 63925
     },
     {
-      "epoch": 2.25,
-      "learning_rate": 3.0248179148438826e-05,
-      "loss": 0.2908,
+      "epoch": 2.3040328684182074,
+      "grad_norm": 0.22265037894248962,
+      "learning_rate": 2.9378431147485896e-05,
+      "loss": 0.396,
       "step": 63930
     },
     {
-      "epoch": 2.25,
-      "learning_rate": 3.024539387103985e-05,
-      "loss": 0.2475,
+      "epoch": 2.304213068079432,
+      "grad_norm": 0.18129128217697144,
+      "learning_rate": 2.9375558086602657e-05,
+      "loss": 0.3808,
       "step": 63935
     },
     {
-      "epoch": 2.25,
-      "learning_rate": 3.024260852553289e-05,
-      "loss": 0.2848,
+      "epoch": 2.3043932677406564,
+      "grad_norm": 0.25198519229888916,
+      "learning_rate": 2.9372684966103293e-05,
+      "loss": 0.4207,
       "step": 63940
     },
     {
-      "epoch": 2.25,
-      "learning_rate": 3.0239823111954124e-05,
-      "loss": 0.258,
+      "epoch": 2.304573467401881,
+      "grad_norm": 0.2193840593099594,
+      "learning_rate": 2.9369811786026935e-05,
+      "loss": 0.4081,
       "step": 63945
     },
     {
-      "epoch": 2.25,
-      "learning_rate": 3.02370376303397e-05,
-      "loss": 0.2805,
+      "epoch": 2.304753667063106,
+      "grad_norm": 0.25426429510116577,
+      "learning_rate": 2.9366938546412744e-05,
+      "loss": 0.3851,
       "step": 63950
     },
     {
-      "epoch": 2.25,
-      "learning_rate": 3.0234252080725796e-05,
-      "loss": 0.2801,
+      "epoch": 2.3049338667243306,
+      "grad_norm": 0.21279068291187286,
+      "learning_rate": 2.9364065247299848e-05,
+      "loss": 0.4081,
       "step": 63955
     },
     {
-      "epoch": 2.25,
-      "learning_rate": 3.0231466463148577e-05,
-      "loss": 0.2646,
+      "epoch": 2.3051140663855554,
+      "grad_norm": 0.1979341357946396,
+      "learning_rate": 2.9361191888727417e-05,
+      "loss": 0.3702,
       "step": 63960
     },
     {
-      "epoch": 2.25,
-      "learning_rate": 3.0228680777644214e-05,
-      "loss": 0.2717,
+      "epoch": 2.3052942660467797,
+      "grad_norm": 0.1835552603006363,
+      "learning_rate": 2.93583184707346e-05,
+      "loss": 0.3754,
       "step": 63965
     },
     {
-      "epoch": 2.25,
-      "learning_rate": 3.0225895024248874e-05,
-      "loss": 0.2827,
+      "epoch": 2.3054744657080044,
+      "grad_norm": 0.16004249453544617,
+      "learning_rate": 2.935544499336052e-05,
+      "loss": 0.4186,
       "step": 63970
     },
     {
-      "epoch": 2.25,
-      "learning_rate": 3.0223109202998734e-05,
-      "loss": 0.2582,
+      "epoch": 2.305654665369229,
+      "grad_norm": 0.22096280753612518,
+      "learning_rate": 2.9352571456644357e-05,
+      "loss": 0.3709,
       "step": 63975
     },
     {
-      "epoch": 2.25,
-      "learning_rate": 3.0220323313929965e-05,
-      "loss": 0.2712,
+      "epoch": 2.305834865030454,
+      "grad_norm": 0.19364888966083527,
+      "learning_rate": 2.934969786062524e-05,
+      "loss": 0.3871,
       "step": 63980
     },
     {
-      "epoch": 2.25,
-      "learning_rate": 3.021753735707874e-05,
-      "loss": 0.2704,
+      "epoch": 2.306015064691678,
+      "grad_norm": 0.17691142857074738,
+      "learning_rate": 2.934682420534235e-05,
+      "loss": 0.4034,
       "step": 63985
     },
     {
-      "epoch": 2.25,
-      "learning_rate": 3.0214751332481218e-05,
-      "loss": 0.2845,
+      "epoch": 2.306195264352903,
+      "grad_norm": 0.239870086312294,
+      "learning_rate": 2.9343950490834806e-05,
+      "loss": 0.4087,
       "step": 63990
     },
     {
-      "epoch": 2.25,
-      "learning_rate": 3.02119652401736e-05,
-      "loss": 0.2768,
+      "epoch": 2.3063754640141276,
+      "grad_norm": 0.2206905335187912,
+      "learning_rate": 2.934107671714178e-05,
+      "loss": 0.3875,
       "step": 63995
     },
     {
-      "epoch": 2.25,
-      "learning_rate": 3.020917908019204e-05,
-      "loss": 0.2669,
+      "epoch": 2.3065556636753524,
+      "grad_norm": 0.18567106127738953,
+      "learning_rate": 2.9338202884302423e-05,
+      "loss": 0.4075,
       "step": 64000
     },
     {
-      "epoch": 2.25,
-      "eval_loss": 0.26856276392936707,
-      "eval_runtime": 10.5405,
-      "eval_samples_per_second": 9.487,
-      "eval_steps_per_second": 9.487,
+      "epoch": 2.3065556636753524,
+      "eval_loss": 0.43620339035987854,
+      "eval_runtime": 3.5308,
+      "eval_samples_per_second": 28.322,
+      "eval_steps_per_second": 7.081,
       "step": 64000
     },
     {
-      "epoch": 2.25,
-      "learning_rate": 3.0206392852572733e-05,
-      "loss": 0.2684,
+      "epoch": 2.306735863336577,
+      "grad_norm": 0.1770346313714981,
+      "learning_rate": 2.93353289923559e-05,
+      "loss": 0.395,
       "step": 64005
     },
     {
-      "epoch": 2.25,
-      "learning_rate": 3.0203606557351836e-05,
-      "loss": 0.2997,
+      "epoch": 2.3069160629978014,
+      "grad_norm": 0.20676358044147491,
+      "learning_rate": 2.9332455041341355e-05,
+      "loss": 0.3684,
       "step": 64010
     },
     {
-      "epoch": 2.25,
-      "learning_rate": 3.0200820194565537e-05,
-      "loss": 0.297,
+      "epoch": 2.307096262659026,
+      "grad_norm": 0.21041785180568695,
+      "learning_rate": 2.932958103129794e-05,
+      "loss": 0.3907,
       "step": 64015
     },
     {
-      "epoch": 2.25,
-      "learning_rate": 3.0198033764250023e-05,
-      "loss": 0.2825,
+      "epoch": 2.307276462320251,
+      "grad_norm": 0.19155463576316833,
+      "learning_rate": 2.9326706962264832e-05,
+      "loss": 0.3741,
       "step": 64020
     },
     {
-      "epoch": 2.25,
-      "learning_rate": 3.0195247266441458e-05,
-      "loss": 0.2744,
+      "epoch": 2.3074566619814756,
+      "grad_norm": 0.1932283639907837,
+      "learning_rate": 2.9323832834281177e-05,
+      "loss": 0.3953,
       "step": 64025
     },
     {
-      "epoch": 2.25,
-      "learning_rate": 3.019246070117604e-05,
-      "loss": 0.2528,
+      "epoch": 2.3076368616427,
+      "grad_norm": 0.19772790372371674,
+      "learning_rate": 2.932095864738613e-05,
+      "loss": 0.4258,
       "step": 64030
     },
     {
-      "epoch": 2.25,
-      "learning_rate": 3.018967406848993e-05,
-      "loss": 0.2631,
+      "epoch": 2.3078170613039246,
+      "grad_norm": 0.18976783752441406,
+      "learning_rate": 2.9318084401618872e-05,
+      "loss": 0.403,
       "step": 64035
     },
     {
-      "epoch": 2.25,
-      "learning_rate": 3.0186887368419337e-05,
-      "loss": 0.2773,
+      "epoch": 2.3079972609651493,
+      "grad_norm": 0.19546537101268768,
+      "learning_rate": 2.9315210097018535e-05,
+      "loss": 0.3811,
       "step": 64040
     },
     {
-      "epoch": 2.25,
-      "learning_rate": 3.018410060100042e-05,
-      "loss": 0.2583,
+      "epoch": 2.308177460626374,
+      "grad_norm": 0.17727042734622955,
+      "learning_rate": 2.9312335733624312e-05,
+      "loss": 0.3558,
       "step": 64045
     },
     {
-      "epoch": 2.25,
-      "learning_rate": 3.0181313766269376e-05,
-      "loss": 0.2694,
+      "epoch": 2.308357660287599,
+      "grad_norm": 0.17102521657943726,
+      "learning_rate": 2.930946131147534e-05,
+      "loss": 0.3756,
       "step": 64050
     },
     {
-      "epoch": 2.25,
-      "learning_rate": 3.017852686426238e-05,
-      "loss": 0.2807,
+      "epoch": 2.3085378599488235,
+      "grad_norm": 0.21491068601608276,
+      "learning_rate": 2.9306586830610794e-05,
+      "loss": 0.4153,
       "step": 64055
     },
     {
-      "epoch": 2.25,
-      "learning_rate": 3.0175739895015636e-05,
-      "loss": 0.2918,
+      "epoch": 2.308718059610048,
+      "grad_norm": 0.18855595588684082,
+      "learning_rate": 2.930371229106983e-05,
+      "loss": 0.3981,
       "step": 64060
     },
     {
-      "epoch": 2.25,
-      "learning_rate": 3.0172952858565317e-05,
-      "loss": 0.3113,
+      "epoch": 2.3088982592712726,
+      "grad_norm": 0.19503138959407806,
+      "learning_rate": 2.9300837692891627e-05,
+      "loss": 0.4133,
       "step": 64065
     },
     {
-      "epoch": 2.25,
-      "learning_rate": 3.017016575494761e-05,
-      "loss": 0.2758,
+      "epoch": 2.3090784589324973,
+      "grad_norm": 0.18976575136184692,
+      "learning_rate": 2.929796303611534e-05,
+      "loss": 0.3941,
       "step": 64070
     },
     {
-      "epoch": 2.25,
-      "learning_rate": 3.0167378584198714e-05,
-      "loss": 0.2863,
+      "epoch": 2.309258658593722,
+      "grad_norm": 0.1784740537405014,
+      "learning_rate": 2.9295088320780133e-05,
+      "loss": 0.408,
       "step": 64075
     },
     {
-      "epoch": 2.25,
-      "learning_rate": 3.016459134635481e-05,
-      "loss": 0.2638,
+      "epoch": 2.3094388582549463,
+      "grad_norm": 0.21722306311130524,
+      "learning_rate": 2.929221354692519e-05,
+      "loss": 0.4086,
       "step": 64080
     },
     {
-      "epoch": 2.25,
-      "learning_rate": 3.0161804041452092e-05,
-      "loss": 0.2724,
+      "epoch": 2.309619057916171,
+      "grad_norm": 0.2144099622964859,
+      "learning_rate": 2.9289338714589664e-05,
+      "loss": 0.3952,
       "step": 64085
     },
     {
-      "epoch": 2.25,
-      "learning_rate": 3.0159016669526752e-05,
-      "loss": 0.2684,
+      "epoch": 2.309799257577396,
+      "grad_norm": 0.1988193541765213,
+      "learning_rate": 2.9286463823812732e-05,
+      "loss": 0.3712,
       "step": 64090
     },
     {
-      "epoch": 2.26,
-      "learning_rate": 3.0156229230614973e-05,
-      "loss": 0.2513,
+      "epoch": 2.3099794572386205,
+      "grad_norm": 0.18340793251991272,
+      "learning_rate": 2.928358887463355e-05,
+      "loss": 0.4021,
       "step": 64095
     },
     {
-      "epoch": 2.26,
-      "learning_rate": 3.015344172475296e-05,
-      "loss": 0.2973,
+      "epoch": 2.3101596568998453,
+      "grad_norm": 0.24040386080741882,
+      "learning_rate": 2.928071386709131e-05,
+      "loss": 0.4141,
       "step": 64100
     },
     {
-      "epoch": 2.26,
-      "learning_rate": 3.0150654151976905e-05,
-      "loss": 0.2869,
+      "epoch": 2.3103398565610695,
+      "grad_norm": 0.19517391920089722,
+      "learning_rate": 2.9277838801225165e-05,
+      "loss": 0.368,
       "step": 64105
     },
     {
-      "epoch": 2.26,
-      "learning_rate": 3.0147866512322996e-05,
-      "loss": 0.2659,
+      "epoch": 2.3105200562222943,
+      "grad_norm": 0.1790241003036499,
+      "learning_rate": 2.9274963677074298e-05,
+      "loss": 0.3828,
       "step": 64110
     },
     {
-      "epoch": 2.26,
-      "learning_rate": 3.0145078805827437e-05,
-      "loss": 0.3028,
+      "epoch": 2.310700255883519,
+      "grad_norm": 0.15677717328071594,
+      "learning_rate": 2.927208849467788e-05,
+      "loss": 0.4316,
       "step": 64115
     },
     {
-      "epoch": 2.26,
-      "learning_rate": 3.014229103252642e-05,
-      "loss": 0.2946,
+      "epoch": 2.3108804555447438,
+      "grad_norm": 0.21783356368541718,
+      "learning_rate": 2.9269213254075074e-05,
+      "loss": 0.4188,
       "step": 64120
     },
     {
-      "epoch": 2.26,
-      "learning_rate": 3.013950319245614e-05,
-      "loss": 0.276,
+      "epoch": 2.311060655205968,
+      "grad_norm": 0.23642495274543762,
+      "learning_rate": 2.926633795530508e-05,
+      "loss": 0.3957,
       "step": 64125
     },
     {
-      "epoch": 2.26,
-      "learning_rate": 3.0136715285652795e-05,
-      "loss": 0.2682,
+      "epoch": 2.3112408548671928,
+      "grad_norm": 0.18061873316764832,
+      "learning_rate": 2.9263462598407044e-05,
+      "loss": 0.3771,
       "step": 64130
     },
     {
-      "epoch": 2.26,
-      "learning_rate": 3.0133927312152593e-05,
-      "loss": 0.2826,
+      "epoch": 2.3114210545284175,
+      "grad_norm": 0.1819910854101181,
+      "learning_rate": 2.9260587183420164e-05,
+      "loss": 0.3468,
       "step": 64135
     },
     {
-      "epoch": 2.26,
-      "learning_rate": 3.0131139271991726e-05,
-      "loss": 0.2862,
+      "epoch": 2.3116012541896422,
+      "grad_norm": 0.2120652049779892,
+      "learning_rate": 2.9257711710383595e-05,
+      "loss": 0.4319,
       "step": 64140
     },
     {
-      "epoch": 2.26,
-      "learning_rate": 3.0128351165206404e-05,
-      "loss": 0.2731,
+      "epoch": 2.311781453850867,
+      "grad_norm": 0.20193518698215485,
+      "learning_rate": 2.925483617933654e-05,
+      "loss": 0.3839,
       "step": 64145
     },
     {
-      "epoch": 2.26,
-      "learning_rate": 3.012556299183281e-05,
-      "loss": 0.2835,
+      "epoch": 2.3119616535120913,
+      "grad_norm": 0.19727912545204163,
+      "learning_rate": 2.9251960590318167e-05,
+      "loss": 0.4062,
       "step": 64150
     },
     {
-      "epoch": 2.26,
-      "learning_rate": 3.012277475190716e-05,
-      "loss": 0.2742,
+      "epoch": 2.312141853173316,
+      "grad_norm": 0.17611241340637207,
+      "learning_rate": 2.9249084943367648e-05,
+      "loss": 0.3645,
       "step": 64155
     },
     {
-      "epoch": 2.26,
-      "learning_rate": 3.0119986445465663e-05,
-      "loss": 0.268,
+      "epoch": 2.3123220528345407,
+      "grad_norm": 0.17656007409095764,
+      "learning_rate": 2.9246209238524176e-05,
+      "loss": 0.3498,
       "step": 64160
     },
     {
-      "epoch": 2.26,
-      "learning_rate": 3.0117198072544518e-05,
-      "loss": 0.2489,
+      "epoch": 2.3125022524957655,
+      "grad_norm": 0.1993638426065445,
+      "learning_rate": 2.9243333475826916e-05,
+      "loss": 0.381,
       "step": 64165
     },
     {
-      "epoch": 2.26,
-      "learning_rate": 3.011440963317992e-05,
-      "loss": 0.2575,
+      "epoch": 2.3126824521569898,
+      "grad_norm": 0.23477478325366974,
+      "learning_rate": 2.924045765531506e-05,
+      "loss": 0.361,
       "step": 64170
     },
     {
-      "epoch": 2.26,
-      "learning_rate": 3.011162112740809e-05,
-      "loss": 0.268,
+      "epoch": 2.3128626518182145,
+      "grad_norm": 0.23515748977661133,
+      "learning_rate": 2.92375817770278e-05,
+      "loss": 0.4028,
       "step": 64175
     },
     {
-      "epoch": 2.26,
-      "learning_rate": 3.0108832555265225e-05,
-      "loss": 0.272,
+      "epoch": 2.313042851479439,
+      "grad_norm": 0.17302124202251434,
+      "learning_rate": 2.9234705841004295e-05,
+      "loss": 0.4029,
       "step": 64180
     },
     {
-      "epoch": 2.26,
-      "learning_rate": 3.0106043916787534e-05,
-      "loss": 0.2692,
+      "epoch": 2.313223051140664,
+      "grad_norm": 0.2219003438949585,
+      "learning_rate": 2.9231829847283753e-05,
+      "loss": 0.3592,
       "step": 64185
     },
     {
-      "epoch": 2.26,
-      "learning_rate": 3.010325521201123e-05,
-      "loss": 0.2692,
+      "epoch": 2.3134032508018887,
+      "grad_norm": 0.23449034988880157,
+      "learning_rate": 2.922895379590535e-05,
+      "loss": 0.408,
       "step": 64190
     },
     {
-      "epoch": 2.26,
-      "learning_rate": 3.0100466440972523e-05,
-      "loss": 0.2626,
+      "epoch": 2.313583450463113,
+      "grad_norm": 0.21402695775032043,
+      "learning_rate": 2.9226077686908265e-05,
+      "loss": 0.4229,
       "step": 64195
     },
     {
-      "epoch": 2.26,
-      "learning_rate": 3.0097677603707614e-05,
-      "loss": 0.2806,
+      "epoch": 2.3137636501243377,
+      "grad_norm": 0.2237950712442398,
+      "learning_rate": 2.922320152033169e-05,
+      "loss": 0.4359,
       "step": 64200
     },
     {
-      "epoch": 2.26,
-      "learning_rate": 3.009488870025273e-05,
-      "loss": 0.2461,
+      "epoch": 2.3139438497855624,
+      "grad_norm": 0.20469123125076294,
+      "learning_rate": 2.922032529621481e-05,
+      "loss": 0.3704,
       "step": 64205
     },
     {
-      "epoch": 2.26,
-      "learning_rate": 3.0092099730644064e-05,
-      "loss": 0.2769,
+      "epoch": 2.314124049446787,
+      "grad_norm": 0.17067334055900574,
+      "learning_rate": 2.921744901459682e-05,
+      "loss": 0.3781,
       "step": 64210
     },
     {
-      "epoch": 2.26,
-      "learning_rate": 3.0089310694917845e-05,
-      "loss": 0.2555,
+      "epoch": 2.3143042491080115,
+      "grad_norm": 0.2015264630317688,
+      "learning_rate": 2.92145726755169e-05,
+      "loss": 0.4129,
       "step": 64215
     },
     {
-      "epoch": 2.26,
-      "learning_rate": 3.008652159311028e-05,
-      "loss": 0.279,
+      "epoch": 2.314484448769236,
+      "grad_norm": 0.19434256851673126,
+      "learning_rate": 2.9211696279014244e-05,
+      "loss": 0.3847,
       "step": 64220
     },
     {
-      "epoch": 2.26,
-      "learning_rate": 3.0083732425257583e-05,
-      "loss": 0.2507,
+      "epoch": 2.314664648430461,
+      "grad_norm": 0.18357601761817932,
+      "learning_rate": 2.920881982512804e-05,
+      "loss": 0.4103,
       "step": 64225
     },
     {
-      "epoch": 2.26,
-      "learning_rate": 3.008094319139597e-05,
-      "loss": 0.2701,
+      "epoch": 2.3148448480916857,
+      "grad_norm": 0.1933751106262207,
+      "learning_rate": 2.9205943313897487e-05,
+      "loss": 0.3704,
       "step": 64230
     },
     {
-      "epoch": 2.26,
-      "learning_rate": 3.007815389156166e-05,
-      "loss": 0.2702,
+      "epoch": 2.3150250477529104,
+      "grad_norm": 0.20393973588943481,
+      "learning_rate": 2.920306674536177e-05,
+      "loss": 0.3816,
       "step": 64235
     },
     {
-      "epoch": 2.26,
-      "learning_rate": 3.0075364525790865e-05,
-      "loss": 0.2656,
+      "epoch": 2.3152052474141347,
+      "grad_norm": 0.23503683507442474,
+      "learning_rate": 2.9200190119560077e-05,
+      "loss": 0.3898,
       "step": 64240
     },
     {
-      "epoch": 2.26,
-      "learning_rate": 3.0072575094119813e-05,
-      "loss": 0.281,
+      "epoch": 2.3153854470753594,
+      "grad_norm": 0.18053038418293,
+      "learning_rate": 2.9197313436531604e-05,
+      "loss": 0.393,
       "step": 64245
     },
     {
-      "epoch": 2.26,
-      "learning_rate": 3.006978559658471e-05,
-      "loss": 0.2743,
+      "epoch": 2.315565646736584,
+      "grad_norm": 0.21024245023727417,
+      "learning_rate": 2.919443669631555e-05,
+      "loss": 0.4292,
       "step": 64250
     },
     {
-      "epoch": 2.26,
-      "learning_rate": 3.0066996033221785e-05,
-      "loss": 0.2509,
+      "epoch": 2.315745846397809,
+      "grad_norm": 0.1868125945329666,
+      "learning_rate": 2.9191559898951115e-05,
+      "loss": 0.4292,
       "step": 64255
     },
     {
-      "epoch": 2.26,
-      "learning_rate": 3.0064206404067258e-05,
-      "loss": 0.2651,
+      "epoch": 2.315926046059033,
+      "grad_norm": 0.19413365423679352,
+      "learning_rate": 2.9188683044477484e-05,
+      "loss": 0.4133,
       "step": 64260
     },
     {
-      "epoch": 2.26,
-      "learning_rate": 3.0061416709157343e-05,
-      "loss": 0.2743,
+      "epoch": 2.316106245720258,
+      "grad_norm": 0.25181156396865845,
+      "learning_rate": 2.9185806132933856e-05,
+      "loss": 0.4184,
       "step": 64265
     },
     {
-      "epoch": 2.26,
-      "learning_rate": 3.005862694852827e-05,
-      "loss": 0.262,
+      "epoch": 2.3162864453814827,
+      "grad_norm": 0.20910336077213287,
+      "learning_rate": 2.9182929164359436e-05,
+      "loss": 0.375,
       "step": 64270
     },
     {
-      "epoch": 2.26,
-      "learning_rate": 3.005583712221626e-05,
-      "loss": 0.2638,
+      "epoch": 2.3164666450427074,
+      "grad_norm": 0.1750982105731964,
+      "learning_rate": 2.9180052138793413e-05,
+      "loss": 0.3782,
       "step": 64275
     },
     {
-      "epoch": 2.26,
-      "learning_rate": 3.0053047230257535e-05,
-      "loss": 0.2714,
+      "epoch": 2.316646844703932,
+      "grad_norm": 0.18748337030410767,
+      "learning_rate": 2.917717505627498e-05,
+      "loss": 0.3698,
       "step": 64280
     },
     {
-      "epoch": 2.26,
-      "learning_rate": 3.0050257272688327e-05,
-      "loss": 0.2561,
+      "epoch": 2.3168270443651564,
+      "grad_norm": 0.2219618856906891,
+      "learning_rate": 2.917429791684335e-05,
+      "loss": 0.4027,
       "step": 64285
     },
     {
-      "epoch": 2.26,
-      "learning_rate": 3.0047467249544854e-05,
-      "loss": 0.2601,
+      "epoch": 2.317007244026381,
+      "grad_norm": 0.19255468249320984,
+      "learning_rate": 2.917142072053773e-05,
+      "loss": 0.373,
       "step": 64290
     },
     {
-      "epoch": 2.26,
-      "learning_rate": 3.0044677160863338e-05,
-      "loss": 0.2726,
+      "epoch": 2.317187443687606,
+      "grad_norm": 0.19871120154857635,
+      "learning_rate": 2.9168543467397298e-05,
+      "loss": 0.3987,
       "step": 64295
     },
     {
-      "epoch": 2.26,
-      "learning_rate": 3.0041887006680026e-05,
-      "loss": 0.2747,
+      "epoch": 2.3173676433488306,
+      "grad_norm": 0.2183811515569687,
+      "learning_rate": 2.9165666157461268e-05,
+      "loss": 0.3834,
       "step": 64300
     },
     {
-      "epoch": 2.26,
-      "learning_rate": 3.0039096787031123e-05,
-      "loss": 0.2926,
+      "epoch": 2.317547843010055,
+      "grad_norm": 0.19588343799114227,
+      "learning_rate": 2.9162788790768847e-05,
+      "loss": 0.4093,
       "step": 64305
     },
     {
-      "epoch": 2.26,
-      "learning_rate": 3.0036306501952882e-05,
-      "loss": 0.2767,
+      "epoch": 2.3177280426712796,
+      "grad_norm": 0.1578301340341568,
+      "learning_rate": 2.9159911367359238e-05,
+      "loss": 0.3794,
       "step": 64310
     },
     {
-      "epoch": 2.26,
-      "learning_rate": 3.0033516151481512e-05,
-      "loss": 0.291,
+      "epoch": 2.3179082423325044,
+      "grad_norm": 0.21837523579597473,
+      "learning_rate": 2.915703388727164e-05,
+      "loss": 0.4073,
       "step": 64315
     },
     {
-      "epoch": 2.26,
-      "learning_rate": 3.0030725735653246e-05,
-      "loss": 0.2831,
+      "epoch": 2.318088441993729,
+      "grad_norm": 0.2273416817188263,
+      "learning_rate": 2.9154156350545252e-05,
+      "loss": 0.3673,
       "step": 64320
     },
     {
-      "epoch": 2.26,
-      "learning_rate": 3.002793525450433e-05,
-      "loss": 0.248,
+      "epoch": 2.318268641654954,
+      "grad_norm": 0.2129017859697342,
+      "learning_rate": 2.9151278757219296e-05,
+      "loss": 0.3983,
       "step": 64325
     },
     {
-      "epoch": 2.26,
-      "learning_rate": 3.002514470807099e-05,
-      "loss": 0.2657,
+      "epoch": 2.3184488413161786,
+      "grad_norm": 0.17560027539730072,
+      "learning_rate": 2.9148401107332972e-05,
+      "loss": 0.3978,
       "step": 64330
     },
     {
-      "epoch": 2.26,
-      "learning_rate": 3.002235409638945e-05,
-      "loss": 0.2686,
+      "epoch": 2.318629040977403,
+      "grad_norm": 0.20387177169322968,
+      "learning_rate": 2.914552340092549e-05,
+      "loss": 0.4097,
       "step": 64335
     },
     {
-      "epoch": 2.26,
-      "learning_rate": 3.001956341949596e-05,
-      "loss": 0.2589,
+      "epoch": 2.3188092406386276,
+      "grad_norm": 0.23885492980480194,
+      "learning_rate": 2.9142645638036042e-05,
+      "loss": 0.3813,
       "step": 64340
     },
     {
-      "epoch": 2.26,
-      "learning_rate": 3.001677267742674e-05,
-      "loss": 0.2523,
+      "epoch": 2.3189894402998523,
+      "grad_norm": 0.20437073707580566,
+      "learning_rate": 2.9139767818703857e-05,
+      "loss": 0.3662,
       "step": 64345
     },
     {
-      "epoch": 2.26,
-      "learning_rate": 3.0013981870218034e-05,
-      "loss": 0.2848,
+      "epoch": 2.319169639961077,
+      "grad_norm": 0.20243240892887115,
+      "learning_rate": 2.9136889942968138e-05,
+      "loss": 0.3974,
       "step": 64350
     },
     {
-      "epoch": 2.26,
-      "learning_rate": 3.001119099790608e-05,
-      "loss": 0.2721,
+      "epoch": 2.3193498396223013,
+      "grad_norm": 0.19991594552993774,
+      "learning_rate": 2.9134012010868082e-05,
+      "loss": 0.4184,
       "step": 64355
     },
     {
-      "epoch": 2.26,
-      "learning_rate": 3.0008400060527113e-05,
-      "loss": 0.2901,
+      "epoch": 2.319530039283526,
+      "grad_norm": 0.1909300833940506,
+      "learning_rate": 2.913113402244293e-05,
+      "loss": 0.3714,
       "step": 64360
     },
     {
-      "epoch": 2.26,
-      "learning_rate": 3.0005609058117374e-05,
-      "loss": 0.2814,
+      "epoch": 2.319710238944751,
+      "grad_norm": 0.21020810306072235,
+      "learning_rate": 2.912825597773186e-05,
+      "loss": 0.4092,
       "step": 64365
     },
     {
-      "epoch": 2.26,
-      "learning_rate": 3.0002817990713096e-05,
-      "loss": 0.256,
+      "epoch": 2.3198904386059755,
+      "grad_norm": 0.29738765954971313,
+      "learning_rate": 2.9125377876774118e-05,
+      "loss": 0.4022,
       "step": 64370
     },
     {
-      "epoch": 2.26,
-      "learning_rate": 3.0000026858350522e-05,
-      "loss": 0.2928,
+      "epoch": 2.3200706382672003,
+      "grad_norm": 0.21587322652339935,
+      "learning_rate": 2.9122499719608894e-05,
+      "loss": 0.4157,
       "step": 64375
     },
     {
-      "epoch": 2.27,
-      "learning_rate": 2.99972356610659e-05,
-      "loss": 0.2758,
+      "epoch": 2.3202508379284246,
+      "grad_norm": 0.19749046862125397,
+      "learning_rate": 2.9119621506275407e-05,
+      "loss": 0.3893,
       "step": 64380
     },
     {
-      "epoch": 2.27,
-      "learning_rate": 2.9994444398895464e-05,
-      "loss": 0.2717,
+      "epoch": 2.3204310375896493,
+      "grad_norm": 0.24363020062446594,
+      "learning_rate": 2.9116743236812878e-05,
+      "loss": 0.4346,
       "step": 64385
     },
     {
-      "epoch": 2.27,
-      "learning_rate": 2.999165307187546e-05,
-      "loss": 0.2795,
+      "epoch": 2.320611237250874,
+      "grad_norm": 0.18595632910728455,
+      "learning_rate": 2.911386491126052e-05,
+      "loss": 0.3947,
       "step": 64390
     },
     {
-      "epoch": 2.27,
-      "learning_rate": 2.998886168004213e-05,
-      "loss": 0.263,
+      "epoch": 2.3207914369120988,
+      "grad_norm": 0.2038854956626892,
+      "learning_rate": 2.9110986529657547e-05,
+      "loss": 0.4065,
       "step": 64395
     },
     {
-      "epoch": 2.27,
-      "learning_rate": 2.998607022343172e-05,
-      "loss": 0.2792,
+      "epoch": 2.320971636573323,
+      "grad_norm": 0.26512575149536133,
+      "learning_rate": 2.9108108092043173e-05,
+      "loss": 0.3958,
       "step": 64400
     },
     {
-      "epoch": 2.27,
-      "learning_rate": 2.9983278702080468e-05,
-      "loss": 0.2497,
+      "epoch": 2.321151836234548,
+      "grad_norm": 0.2215341031551361,
+      "learning_rate": 2.910522959845663e-05,
+      "loss": 0.3719,
       "step": 64405
     },
     {
-      "epoch": 2.27,
-      "learning_rate": 2.998048711602463e-05,
-      "loss": 0.2719,
+      "epoch": 2.3213320358957725,
+      "grad_norm": 0.20210741460323334,
+      "learning_rate": 2.9102351048937122e-05,
+      "loss": 0.4205,
       "step": 64410
     },
     {
-      "epoch": 2.27,
-      "learning_rate": 2.9977695465300455e-05,
-      "loss": 0.2633,
+      "epoch": 2.3215122355569973,
+      "grad_norm": 0.21246850490570068,
+      "learning_rate": 2.9099472443523885e-05,
+      "loss": 0.4175,
       "step": 64415
     },
     {
-      "epoch": 2.27,
-      "learning_rate": 2.997490374994418e-05,
-      "loss": 0.2752,
+      "epoch": 2.321692435218222,
+      "grad_norm": 0.23316890001296997,
+      "learning_rate": 2.9096593782256122e-05,
+      "loss": 0.4295,
       "step": 64420
     },
     {
-      "epoch": 2.27,
-      "learning_rate": 2.9972111969992055e-05,
-      "loss": 0.3034,
+      "epoch": 2.3218726348794463,
+      "grad_norm": 0.17650622129440308,
+      "learning_rate": 2.9093715065173066e-05,
+      "loss": 0.4047,
       "step": 64425
     },
     {
-      "epoch": 2.27,
-      "learning_rate": 2.996932012548033e-05,
-      "loss": 0.2559,
+      "epoch": 2.322052834540671,
+      "grad_norm": 0.20012375712394714,
+      "learning_rate": 2.9090836292313928e-05,
+      "loss": 0.4085,
       "step": 64430
     },
     {
-      "epoch": 2.27,
-      "learning_rate": 2.9966528216445265e-05,
-      "loss": 0.2829,
+      "epoch": 2.3222330342018958,
+      "grad_norm": 0.2151358723640442,
+      "learning_rate": 2.908795746371794e-05,
+      "loss": 0.3799,
       "step": 64435
     },
     {
-      "epoch": 2.27,
-      "learning_rate": 2.9963736242923107e-05,
-      "loss": 0.2869,
+      "epoch": 2.3224132338631205,
+      "grad_norm": 0.20195366442203522,
+      "learning_rate": 2.908507857942433e-05,
+      "loss": 0.3912,
       "step": 64440
     },
     {
-      "epoch": 2.27,
-      "learning_rate": 2.9960944204950105e-05,
-      "loss": 0.2867,
+      "epoch": 2.3225934335243448,
+      "grad_norm": 0.24572651088237762,
+      "learning_rate": 2.9082199639472306e-05,
+      "loss": 0.3691,
       "step": 64445
     },
     {
-      "epoch": 2.27,
-      "learning_rate": 2.9958152102562503e-05,
-      "loss": 0.2968,
+      "epoch": 2.3227736331855695,
+      "grad_norm": 0.20580525696277618,
+      "learning_rate": 2.907932064390111e-05,
+      "loss": 0.3815,
       "step": 64450
     },
     {
-      "epoch": 2.27,
-      "learning_rate": 2.9955359935796568e-05,
-      "loss": 0.2643,
+      "epoch": 2.3229538328467942,
+      "grad_norm": 0.26529133319854736,
+      "learning_rate": 2.9076441592749954e-05,
+      "loss": 0.3954,
       "step": 64455
     },
     {
-      "epoch": 2.27,
-      "learning_rate": 2.995256770468855e-05,
-      "loss": 0.2917,
+      "epoch": 2.323134032508019,
+      "grad_norm": 0.19922053813934326,
+      "learning_rate": 2.907356248605807e-05,
+      "loss": 0.3686,
       "step": 64460
     },
     {
-      "epoch": 2.27,
-      "learning_rate": 2.9949775409274706e-05,
-      "loss": 0.2657,
+      "epoch": 2.3233142321692437,
+      "grad_norm": 0.22310146689414978,
+      "learning_rate": 2.9071259160741603e-05,
+      "loss": 0.4364,
       "step": 64465
     },
     {
-      "epoch": 2.27,
-      "learning_rate": 2.994698304959129e-05,
-      "loss": 0.2644,
+      "epoch": 2.323494431830468,
+      "grad_norm": 0.19857893884181976,
+      "learning_rate": 2.9068379954175267e-05,
+      "loss": 0.4276,
       "step": 64470
     },
     {
-      "epoch": 2.27,
-      "learning_rate": 2.994419062567456e-05,
-      "loss": 0.2726,
+      "epoch": 2.3236746314916927,
+      "grad_norm": 0.20761960744857788,
+      "learning_rate": 2.906550069217804e-05,
+      "loss": 0.3837,
       "step": 64475
     },
     {
-      "epoch": 2.27,
-      "learning_rate": 2.9941398137560772e-05,
-      "loss": 0.272,
+      "epoch": 2.3238548311529175,
+      "grad_norm": 0.22300156950950623,
+      "learning_rate": 2.906262137478915e-05,
+      "loss": 0.421,
       "step": 64480
     },
     {
-      "epoch": 2.27,
-      "learning_rate": 2.9938605585286182e-05,
-      "loss": 0.2778,
+      "epoch": 2.324035030814142,
+      "grad_norm": 0.19981767237186432,
+      "learning_rate": 2.9059742002047835e-05,
+      "loss": 0.3987,
       "step": 64485
     },
     {
-      "epoch": 2.27,
-      "learning_rate": 2.9935812968887055e-05,
-      "loss": 0.2645,
+      "epoch": 2.3242152304753665,
+      "grad_norm": 0.21243509650230408,
+      "learning_rate": 2.9056862573993322e-05,
+      "loss": 0.4114,
       "step": 64490
     },
     {
-      "epoch": 2.27,
-      "learning_rate": 2.993302028839965e-05,
-      "loss": 0.2834,
+      "epoch": 2.3243954301365912,
+      "grad_norm": 0.22393706440925598,
+      "learning_rate": 2.9053983090664838e-05,
+      "loss": 0.3966,
       "step": 64495
     },
     {
-      "epoch": 2.27,
-      "learning_rate": 2.993022754386023e-05,
-      "loss": 0.2824,
+      "epoch": 2.324575629797816,
+      "grad_norm": 0.19257164001464844,
+      "learning_rate": 2.9051103552101623e-05,
+      "loss": 0.389,
       "step": 64500
     },
     {
-      "epoch": 2.27,
-      "eval_loss": 0.26770463585853577,
-      "eval_runtime": 10.5455,
-      "eval_samples_per_second": 9.483,
-      "eval_steps_per_second": 9.483,
+      "epoch": 2.324575629797816,
+      "eval_loss": 0.4352463483810425,
+      "eval_runtime": 3.5326,
+      "eval_samples_per_second": 28.308,
+      "eval_steps_per_second": 7.077,
       "step": 64500
     },
     {
-      "epoch": 2.27,
-      "learning_rate": 2.992743473530505e-05,
-      "loss": 0.2888,
+      "epoch": 2.3247558294590407,
+      "grad_norm": 0.18349726498126984,
+      "learning_rate": 2.9048223958342906e-05,
+      "loss": 0.407,
       "step": 64505
     },
     {
-      "epoch": 2.27,
-      "learning_rate": 2.9924641862770376e-05,
-      "loss": 0.2819,
+      "epoch": 2.3249360291202654,
+      "grad_norm": 0.2214827537536621,
+      "learning_rate": 2.9045344309427924e-05,
+      "loss": 0.3542,
       "step": 64510
     },
     {
-      "epoch": 2.27,
-      "learning_rate": 2.992184892629248e-05,
-      "loss": 0.2553,
+      "epoch": 2.3251162287814897,
+      "grad_norm": 0.19785992801189423,
+      "learning_rate": 2.9042464605395898e-05,
+      "loss": 0.4088,
       "step": 64515
     },
     {
-      "epoch": 2.27,
-      "learning_rate": 2.991905592590762e-05,
-      "loss": 0.2743,
+      "epoch": 2.3252964284427144,
+      "grad_norm": 0.24504734575748444,
+      "learning_rate": 2.9039584846286083e-05,
+      "loss": 0.3953,
       "step": 64520
     },
     {
-      "epoch": 2.27,
-      "learning_rate": 2.991626286165205e-05,
-      "loss": 0.2775,
+      "epoch": 2.325476628103939,
+      "grad_norm": 0.22237750887870789,
+      "learning_rate": 2.9036705032137707e-05,
+      "loss": 0.3808,
       "step": 64525
     },
     {
-      "epoch": 2.27,
-      "learning_rate": 2.991346973356206e-05,
-      "loss": 0.2794,
+      "epoch": 2.325656827765164,
+      "grad_norm": 0.23211441934108734,
+      "learning_rate": 2.903382516299e-05,
+      "loss": 0.4144,
       "step": 64530
     },
     {
-      "epoch": 2.27,
-      "learning_rate": 2.9910676541673893e-05,
-      "loss": 0.2578,
+      "epoch": 2.325837027426388,
+      "grad_norm": 0.21520470082759857,
+      "learning_rate": 2.9030945238882214e-05,
+      "loss": 0.4161,
       "step": 64535
     },
     {
-      "epoch": 2.27,
-      "learning_rate": 2.990788328602383e-05,
-      "loss": 0.2791,
+      "epoch": 2.326017227087613,
+      "grad_norm": 0.21746765077114105,
+      "learning_rate": 2.9028065259853572e-05,
+      "loss": 0.4375,
       "step": 64540
     },
     {
-      "epoch": 2.27,
-      "learning_rate": 2.9905089966648137e-05,
-      "loss": 0.2512,
+      "epoch": 2.3261974267488377,
+      "grad_norm": 0.22863152623176575,
+      "learning_rate": 2.902518522594333e-05,
+      "loss": 0.409,
       "step": 64545
     },
     {
-      "epoch": 2.27,
-      "learning_rate": 2.990229658358309e-05,
-      "loss": 0.2719,
+      "epoch": 2.3263776264100624,
+      "grad_norm": 0.1936892420053482,
+      "learning_rate": 2.902230513719071e-05,
+      "loss": 0.4021,
       "step": 64550
     },
     {
-      "epoch": 2.27,
-      "learning_rate": 2.989950313686494e-05,
-      "loss": 0.2566,
+      "epoch": 2.326557826071287,
+      "grad_norm": 0.22578535974025726,
+      "learning_rate": 2.9019424993634968e-05,
+      "loss": 0.402,
       "step": 64555
     },
     {
-      "epoch": 2.27,
-      "learning_rate": 2.9896709626529986e-05,
-      "loss": 0.2486,
+      "epoch": 2.326738025732512,
+      "grad_norm": 0.23165376484394073,
+      "learning_rate": 2.901654479531533e-05,
+      "loss": 0.4211,
       "step": 64560
     },
     {
-      "epoch": 2.27,
-      "learning_rate": 2.9893916052614474e-05,
-      "loss": 0.2557,
+      "epoch": 2.326918225393736,
+      "grad_norm": 0.1904148906469345,
+      "learning_rate": 2.9013664542271057e-05,
+      "loss": 0.4131,
       "step": 64565
     },
     {
-      "epoch": 2.27,
-      "learning_rate": 2.989112241515469e-05,
-      "loss": 0.2585,
+      "epoch": 2.327098425054961,
+      "grad_norm": 0.16796857118606567,
+      "learning_rate": 2.9010784234541376e-05,
+      "loss": 0.4064,
       "step": 64570
     },
     {
-      "epoch": 2.27,
-      "learning_rate": 2.9888328714186898e-05,
-      "loss": 0.294,
+      "epoch": 2.3272786247161856,
+      "grad_norm": 0.1999785602092743,
+      "learning_rate": 2.900790387216553e-05,
+      "loss": 0.3945,
       "step": 64575
     },
     {
-      "epoch": 2.27,
-      "learning_rate": 2.988553494974739e-05,
-      "loss": 0.2827,
+      "epoch": 2.3274588243774104,
+      "grad_norm": 0.19684360921382904,
+      "learning_rate": 2.9005023455182784e-05,
+      "loss": 0.3999,
       "step": 64580
     },
     {
-      "epoch": 2.27,
-      "learning_rate": 2.9882741121872425e-05,
-      "loss": 0.2649,
+      "epoch": 2.3276390240386347,
+      "grad_norm": 0.20736943185329437,
+      "learning_rate": 2.9002142983632365e-05,
+      "loss": 0.3924,
       "step": 64585
     },
     {
-      "epoch": 2.27,
-      "learning_rate": 2.9879947230598283e-05,
-      "loss": 0.2765,
+      "epoch": 2.3278192236998594,
+      "grad_norm": 0.21448162198066711,
+      "learning_rate": 2.8999262457553518e-05,
+      "loss": 0.3711,
       "step": 64590
     },
     {
-      "epoch": 2.27,
-      "learning_rate": 2.9877153275961244e-05,
-      "loss": 0.2723,
+      "epoch": 2.327999423361084,
+      "grad_norm": 0.17774830758571625,
+      "learning_rate": 2.899638187698549e-05,
+      "loss": 0.3944,
       "step": 64595
     },
     {
-      "epoch": 2.27,
-      "learning_rate": 2.987435925799758e-05,
-      "loss": 0.2668,
+      "epoch": 2.328179623022309,
+      "grad_norm": 0.19229644536972046,
+      "learning_rate": 2.899350124196754e-05,
+      "loss": 0.3717,
       "step": 64600
     },
     {
-      "epoch": 2.27,
-      "learning_rate": 2.9871565176743583e-05,
-      "loss": 0.2791,
+      "epoch": 2.3283598226835336,
+      "grad_norm": 0.22813409566879272,
+      "learning_rate": 2.8990620552538905e-05,
+      "loss": 0.4118,
       "step": 64605
     },
     {
-      "epoch": 2.27,
-      "learning_rate": 2.9868771032235514e-05,
-      "loss": 0.2842,
+      "epoch": 2.328540022344758,
+      "grad_norm": 0.16227515041828156,
+      "learning_rate": 2.898773980873884e-05,
+      "loss": 0.4017,
       "step": 64610
     },
     {
-      "epoch": 2.27,
-      "learning_rate": 2.9865976824509673e-05,
-      "loss": 0.2497,
+      "epoch": 2.3287202220059826,
+      "grad_norm": 0.20297592878341675,
+      "learning_rate": 2.898485901060659e-05,
+      "loss": 0.4339,
       "step": 64615
     },
     {
-      "epoch": 2.27,
-      "learning_rate": 2.986318255360232e-05,
-      "loss": 0.2682,
+      "epoch": 2.3289004216672073,
+      "grad_norm": 0.2151886522769928,
+      "learning_rate": 2.8981978158181412e-05,
+      "loss": 0.4276,
       "step": 64620
     },
     {
-      "epoch": 2.27,
-      "learning_rate": 2.9860388219549744e-05,
-      "loss": 0.2814,
+      "epoch": 2.329080621328432,
+      "grad_norm": 0.21326357126235962,
+      "learning_rate": 2.8979097251502548e-05,
+      "loss": 0.36,
       "step": 64625
     },
     {
-      "epoch": 2.27,
-      "learning_rate": 2.9857593822388237e-05,
-      "loss": 0.286,
+      "epoch": 2.3292608209896564,
+      "grad_norm": 0.20007814466953278,
+      "learning_rate": 2.8976216290609258e-05,
+      "loss": 0.4148,
       "step": 64630
     },
     {
-      "epoch": 2.27,
-      "learning_rate": 2.9854799362154074e-05,
-      "loss": 0.2544,
+      "epoch": 2.329441020650881,
+      "grad_norm": 0.20268851518630981,
+      "learning_rate": 2.8973335275540785e-05,
+      "loss": 0.407,
       "step": 64635
     },
     {
-      "epoch": 2.27,
-      "learning_rate": 2.985200483888354e-05,
-      "loss": 0.2455,
+      "epoch": 2.329621220312106,
+      "grad_norm": 0.2361319661140442,
+      "learning_rate": 2.8970454206336393e-05,
+      "loss": 0.4504,
       "step": 64640
     },
     {
-      "epoch": 2.27,
-      "learning_rate": 2.9849210252612926e-05,
-      "loss": 0.2825,
+      "epoch": 2.3298014199733306,
+      "grad_norm": 0.17205223441123962,
+      "learning_rate": 2.8967573083035327e-05,
+      "loss": 0.3959,
       "step": 64645
     },
     {
-      "epoch": 2.27,
-      "learning_rate": 2.9846415603378496e-05,
-      "loss": 0.2637,
+      "epoch": 2.3299816196345553,
+      "grad_norm": 0.19118523597717285,
+      "learning_rate": 2.8964691905676856e-05,
+      "loss": 0.4025,
       "step": 64650
     },
     {
-      "epoch": 2.27,
-      "learning_rate": 2.9843620891216566e-05,
-      "loss": 0.2724,
+      "epoch": 2.3301618192957796,
+      "grad_norm": 0.20011954009532928,
+      "learning_rate": 2.8961810674300217e-05,
+      "loss": 0.4142,
       "step": 64655
     },
     {
-      "epoch": 2.27,
-      "learning_rate": 2.984082611616341e-05,
-      "loss": 0.2574,
+      "epoch": 2.3303420189570043,
+      "grad_norm": 0.21023406088352203,
+      "learning_rate": 2.895892938894468e-05,
+      "loss": 0.4246,
       "step": 64660
     },
     {
-      "epoch": 2.28,
-      "learning_rate": 2.983803127825532e-05,
-      "loss": 0.2862,
+      "epoch": 2.330522218618229,
+      "grad_norm": 0.19074782729148865,
+      "learning_rate": 2.8956048049649487e-05,
+      "loss": 0.4024,
       "step": 64665
     },
     {
-      "epoch": 2.28,
-      "learning_rate": 2.9835236377528575e-05,
-      "loss": 0.2622,
+      "epoch": 2.330702418279454,
+      "grad_norm": 0.24010269343852997,
+      "learning_rate": 2.895316665645391e-05,
+      "loss": 0.3951,
       "step": 64670
     },
     {
-      "epoch": 2.28,
-      "learning_rate": 2.9832441414019474e-05,
-      "loss": 0.2929,
+      "epoch": 2.330882617940678,
+      "grad_norm": 0.23700617253780365,
+      "learning_rate": 2.8950285209397206e-05,
+      "loss": 0.3924,
       "step": 64675
     },
     {
-      "epoch": 2.28,
-      "learning_rate": 2.982964638776431e-05,
-      "loss": 0.2792,
+      "epoch": 2.331062817601903,
+      "grad_norm": 0.23753370344638824,
+      "learning_rate": 2.8947403708518623e-05,
+      "loss": 0.4263,
       "step": 64680
     },
     {
-      "epoch": 2.28,
-      "learning_rate": 2.982685129879937e-05,
-      "loss": 0.2724,
+      "epoch": 2.3312430172631275,
+      "grad_norm": 0.17111685872077942,
+      "learning_rate": 2.8944522153857433e-05,
+      "loss": 0.4019,
       "step": 64685
     },
     {
-      "epoch": 2.28,
-      "learning_rate": 2.9824056147160944e-05,
-      "loss": 0.2722,
+      "epoch": 2.3314232169243523,
+      "grad_norm": 0.18538156151771545,
+      "learning_rate": 2.8941640545452898e-05,
+      "loss": 0.4268,
       "step": 64690
     },
     {
-      "epoch": 2.28,
-      "learning_rate": 2.982126093288533e-05,
-      "loss": 0.2795,
+      "epoch": 2.331603416585577,
+      "grad_norm": 0.2004711627960205,
+      "learning_rate": 2.893875888334427e-05,
+      "loss": 0.3993,
       "step": 64695
     },
     {
-      "epoch": 2.28,
-      "learning_rate": 2.981846565600882e-05,
-      "loss": 0.2508,
+      "epoch": 2.3317836162468013,
+      "grad_norm": 0.18873223662376404,
+      "learning_rate": 2.8935877167570814e-05,
+      "loss": 0.3939,
       "step": 64700
     },
     {
-      "epoch": 2.28,
-      "learning_rate": 2.9815670316567706e-05,
-      "loss": 0.2602,
+      "epoch": 2.331963815908026,
+      "grad_norm": 0.1974310278892517,
+      "learning_rate": 2.8932995398171793e-05,
+      "loss": 0.3892,
       "step": 64705
     },
     {
-      "epoch": 2.28,
-      "learning_rate": 2.9812874914598288e-05,
-      "loss": 0.2481,
+      "epoch": 2.3321440155692508,
+      "grad_norm": 0.2253679633140564,
+      "learning_rate": 2.8930113575186473e-05,
+      "loss": 0.4034,
       "step": 64710
     },
     {
-      "epoch": 2.28,
-      "learning_rate": 2.9810079450136863e-05,
-      "loss": 0.2904,
+      "epoch": 2.3323242152304755,
+      "grad_norm": 0.2432539016008377,
+      "learning_rate": 2.892723169865411e-05,
+      "loss": 0.4074,
       "step": 64715
     },
     {
-      "epoch": 2.28,
-      "learning_rate": 2.9807283923219724e-05,
-      "loss": 0.2735,
+      "epoch": 2.3325044148917,
+      "grad_norm": 0.22504882514476776,
+      "learning_rate": 2.8924349768613984e-05,
+      "loss": 0.4394,
       "step": 64720
     },
     {
-      "epoch": 2.28,
-      "learning_rate": 2.9804488333883172e-05,
-      "loss": 0.2911,
+      "epoch": 2.3326846145529245,
+      "grad_norm": 0.24001994729042053,
+      "learning_rate": 2.8921467785105342e-05,
+      "loss": 0.3778,
       "step": 64725
     },
     {
-      "epoch": 2.28,
-      "learning_rate": 2.9801692682163508e-05,
-      "loss": 0.2698,
+      "epoch": 2.3328648142141493,
+      "grad_norm": 0.19272451102733612,
+      "learning_rate": 2.8918585748167475e-05,
+      "loss": 0.4748,
       "step": 64730
     },
     {
-      "epoch": 2.28,
-      "learning_rate": 2.9798896968097025e-05,
-      "loss": 0.2831,
+      "epoch": 2.333045013875374,
+      "grad_norm": 0.20509444177150726,
+      "learning_rate": 2.8915703657839627e-05,
+      "loss": 0.4025,
       "step": 64735
     },
     {
-      "epoch": 2.28,
-      "learning_rate": 2.9796101191720033e-05,
-      "loss": 0.2828,
+      "epoch": 2.3332252135365987,
+      "grad_norm": 0.1780424416065216,
+      "learning_rate": 2.8912821514161077e-05,
+      "loss": 0.3849,
       "step": 64740
     },
     {
-      "epoch": 2.28,
-      "learning_rate": 2.9793305353068816e-05,
-      "loss": 0.2739,
+      "epoch": 2.333405413197823,
+      "grad_norm": 0.20018140971660614,
+      "learning_rate": 2.8909939317171086e-05,
+      "loss": 0.4019,
       "step": 64745
     },
     {
-      "epoch": 2.28,
-      "learning_rate": 2.9790509452179706e-05,
-      "loss": 0.2611,
+      "epoch": 2.3335856128590478,
+      "grad_norm": 0.22508153319358826,
+      "learning_rate": 2.8907057066908934e-05,
+      "loss": 0.3987,
       "step": 64750
     },
     {
-      "epoch": 2.28,
-      "learning_rate": 2.9787713489088975e-05,
-      "loss": 0.2712,
+      "epoch": 2.3337658125202725,
+      "grad_norm": 0.22255435585975647,
+      "learning_rate": 2.8904174763413888e-05,
+      "loss": 0.3795,
       "step": 64755
     },
     {
-      "epoch": 2.28,
-      "learning_rate": 2.978491746383294e-05,
-      "loss": 0.2625,
+      "epoch": 2.333946012181497,
+      "grad_norm": 0.17312286794185638,
+      "learning_rate": 2.890129240672521e-05,
+      "loss": 0.4022,
       "step": 64760
     },
     {
-      "epoch": 2.28,
-      "learning_rate": 2.97821213764479e-05,
-      "loss": 0.3069,
+      "epoch": 2.3341262118427215,
+      "grad_norm": 0.18471607565879822,
+      "learning_rate": 2.8898409996882182e-05,
+      "loss": 0.3776,
       "step": 64765
     },
     {
-      "epoch": 2.28,
-      "learning_rate": 2.9779325226970183e-05,
-      "loss": 0.2856,
+      "epoch": 2.3343064115039462,
+      "grad_norm": 0.19598588347434998,
+      "learning_rate": 2.8895527533924076e-05,
+      "loss": 0.3784,
       "step": 64770
     },
     {
-      "epoch": 2.28,
-      "learning_rate": 2.9776529015436065e-05,
-      "loss": 0.2727,
+      "epoch": 2.334486611165171,
+      "grad_norm": 0.20340387523174286,
+      "learning_rate": 2.889264501789015e-05,
+      "loss": 0.3934,
       "step": 64775
     },
     {
-      "epoch": 2.28,
-      "learning_rate": 2.977373274188187e-05,
-      "loss": 0.2896,
+      "epoch": 2.3346668108263957,
+      "grad_norm": 0.2492077350616455,
+      "learning_rate": 2.88897624488197e-05,
+      "loss": 0.4267,
       "step": 64780
     },
     {
-      "epoch": 2.28,
-      "learning_rate": 2.9770936406343895e-05,
-      "loss": 0.2778,
+      "epoch": 2.3348470104876204,
+      "grad_norm": 0.21682408452033997,
+      "learning_rate": 2.8886879826751982e-05,
+      "loss": 0.4055,
       "step": 64785
     },
     {
-      "epoch": 2.28,
-      "learning_rate": 2.9768140008858464e-05,
-      "loss": 0.2569,
+      "epoch": 2.3350272101488447,
+      "grad_norm": 0.2086905837059021,
+      "learning_rate": 2.8883997151726293e-05,
+      "loss": 0.353,
       "step": 64790
     },
     {
-      "epoch": 2.28,
-      "learning_rate": 2.9765343549461866e-05,
-      "loss": 0.2996,
+      "epoch": 2.3352074098100695,
+      "grad_norm": 0.18733155727386475,
+      "learning_rate": 2.8881114423781885e-05,
+      "loss": 0.372,
       "step": 64795
     },
     {
-      "epoch": 2.28,
-      "learning_rate": 2.9762547028190436e-05,
-      "loss": 0.2896,
+      "epoch": 2.335387609471294,
+      "grad_norm": 0.18295986950397491,
+      "learning_rate": 2.8878231642958044e-05,
+      "loss": 0.3722,
       "step": 64800
     },
     {
-      "epoch": 2.28,
-      "learning_rate": 2.9759750445080465e-05,
-      "loss": 0.2651,
+      "epoch": 2.335567809132519,
+      "grad_norm": 0.21052880585193634,
+      "learning_rate": 2.8875348809294044e-05,
+      "loss": 0.3814,
       "step": 64805
     },
     {
-      "epoch": 2.28,
-      "learning_rate": 2.9756953800168262e-05,
-      "loss": 0.2642,
+      "epoch": 2.3357480087937432,
+      "grad_norm": 0.20123307406902313,
+      "learning_rate": 2.887246592282917e-05,
+      "loss": 0.3888,
       "step": 64810
     },
     {
-      "epoch": 2.28,
-      "learning_rate": 2.975415709349016e-05,
-      "loss": 0.2653,
+      "epoch": 2.335928208454968,
+      "grad_norm": 0.23528067767620087,
+      "learning_rate": 2.8869582983602705e-05,
+      "loss": 0.3984,
       "step": 64815
     },
     {
-      "epoch": 2.28,
-      "learning_rate": 2.975136032508246e-05,
-      "loss": 0.2717,
+      "epoch": 2.3361084081161927,
+      "grad_norm": 0.22291633486747742,
+      "learning_rate": 2.8866699991653913e-05,
+      "loss": 0.4111,
       "step": 64820
     },
     {
-      "epoch": 2.28,
-      "learning_rate": 2.974856349498148e-05,
-      "loss": 0.2564,
+      "epoch": 2.3362886077774174,
+      "grad_norm": 0.16556419432163239,
+      "learning_rate": 2.8863816947022087e-05,
+      "loss": 0.3962,
       "step": 64825
     },
     {
-      "epoch": 2.28,
-      "learning_rate": 2.974576660322353e-05,
-      "loss": 0.2669,
+      "epoch": 2.336468807438642,
+      "grad_norm": 0.2400919497013092,
+      "learning_rate": 2.8860933849746496e-05,
+      "loss": 0.3776,
       "step": 64830
     },
     {
-      "epoch": 2.28,
-      "learning_rate": 2.9742969649844926e-05,
-      "loss": 0.2428,
+      "epoch": 2.336649007099867,
+      "grad_norm": 0.2065679430961609,
+      "learning_rate": 2.8858050699866442e-05,
+      "loss": 0.4298,
       "step": 64835
     },
     {
-      "epoch": 2.28,
-      "learning_rate": 2.9740172634881986e-05,
-      "loss": 0.291,
+      "epoch": 2.336829206761091,
+      "grad_norm": 0.19219578802585602,
+      "learning_rate": 2.885516749742118e-05,
+      "loss": 0.3546,
       "step": 64840
     },
     {
-      "epoch": 2.28,
-      "learning_rate": 2.9737375558371026e-05,
-      "loss": 0.2643,
+      "epoch": 2.337009406422316,
+      "grad_norm": 0.2160220742225647,
+      "learning_rate": 2.8852284242450013e-05,
+      "loss": 0.4224,
       "step": 64845
     },
     {
-      "epoch": 2.28,
-      "learning_rate": 2.973457842034837e-05,
-      "loss": 0.2804,
+      "epoch": 2.3371896060835406,
+      "grad_norm": 0.17602595686912537,
+      "learning_rate": 2.8849400934992215e-05,
+      "loss": 0.4115,
       "step": 64850
     },
     {
-      "epoch": 2.28,
-      "learning_rate": 2.973178122085034e-05,
-      "loss": 0.2886,
+      "epoch": 2.3373698057447654,
+      "grad_norm": 0.21248827874660492,
+      "learning_rate": 2.8846517575087086e-05,
+      "loss": 0.4115,
       "step": 64855
     },
     {
-      "epoch": 2.28,
-      "learning_rate": 2.9728983959913243e-05,
-      "loss": 0.2789,
+      "epoch": 2.3375500054059897,
+      "grad_norm": 0.1953934133052826,
+      "learning_rate": 2.884363416277389e-05,
+      "loss": 0.388,
       "step": 64860
     },
     {
-      "epoch": 2.28,
-      "learning_rate": 2.9726186637573404e-05,
-      "loss": 0.2683,
+      "epoch": 2.3377302050672144,
+      "grad_norm": 0.19074764847755432,
+      "learning_rate": 2.8840750698091924e-05,
+      "loss": 0.3821,
       "step": 64865
     },
     {
-      "epoch": 2.28,
-      "learning_rate": 2.9723389253867146e-05,
-      "loss": 0.2639,
+      "epoch": 2.337910404728439,
+      "grad_norm": 0.20078332722187042,
+      "learning_rate": 2.883786718108047e-05,
+      "loss": 0.4138,
       "step": 64870
     },
     {
-      "epoch": 2.28,
-      "learning_rate": 2.9720591808830796e-05,
-      "loss": 0.2614,
+      "epoch": 2.338090604389664,
+      "grad_norm": 0.20209373533725739,
+      "learning_rate": 2.8834983611778827e-05,
+      "loss": 0.4012,
       "step": 64875
     },
     {
-      "epoch": 2.28,
-      "learning_rate": 2.971779430250067e-05,
-      "loss": 0.2755,
+      "epoch": 2.3382708040508886,
+      "grad_norm": 0.21952852606773376,
+      "learning_rate": 2.8832099990226268e-05,
+      "loss": 0.3731,
       "step": 64880
     },
     {
-      "epoch": 2.28,
-      "learning_rate": 2.97149967349131e-05,
-      "loss": 0.3162,
+      "epoch": 2.338451003712113,
+      "grad_norm": 0.18298974633216858,
+      "learning_rate": 2.8829216316462092e-05,
+      "loss": 0.3778,
       "step": 64885
     },
     {
-      "epoch": 2.28,
-      "learning_rate": 2.9712199106104392e-05,
-      "loss": 0.2874,
+      "epoch": 2.3386312033733376,
+      "grad_norm": 0.2111566960811615,
+      "learning_rate": 2.8826332590525583e-05,
+      "loss": 0.4025,
       "step": 64890
     },
     {
-      "epoch": 2.28,
-      "learning_rate": 2.9709401416110895e-05,
-      "loss": 0.2674,
+      "epoch": 2.3388114030345624,
+      "grad_norm": 0.21694448590278625,
+      "learning_rate": 2.882344881245604e-05,
+      "loss": 0.414,
       "step": 64895
     },
     {
-      "epoch": 2.28,
-      "learning_rate": 2.9706603664968913e-05,
-      "loss": 0.26,
+      "epoch": 2.338991602695787,
+      "grad_norm": 0.188668355345726,
+      "learning_rate": 2.8820564982292736e-05,
+      "loss": 0.4043,
       "step": 64900
     },
     {
-      "epoch": 2.28,
-      "learning_rate": 2.9703805852714796e-05,
-      "loss": 0.2534,
+      "epoch": 2.3391718023570114,
+      "grad_norm": 0.18398909270763397,
+      "learning_rate": 2.881768110007498e-05,
+      "loss": 0.3847,
       "step": 64905
     },
     {
-      "epoch": 2.28,
-      "learning_rate": 2.9701007979384855e-05,
-      "loss": 0.294,
+      "epoch": 2.339352002018236,
+      "grad_norm": 0.19492432475090027,
+      "learning_rate": 2.8814797165842057e-05,
+      "loss": 0.38,
       "step": 64910
     },
     {
-      "epoch": 2.28,
-      "learning_rate": 2.9698210045015424e-05,
-      "loss": 0.2643,
+      "epoch": 2.339532201679461,
+      "grad_norm": 0.16589581966400146,
+      "learning_rate": 2.881191317963326e-05,
+      "loss": 0.3701,
       "step": 64915
     },
     {
-      "epoch": 2.28,
-      "learning_rate": 2.9695971653595856e-05,
-      "loss": 0.2674,
+      "epoch": 2.3397124013406856,
+      "grad_norm": 0.23120105266571045,
+      "learning_rate": 2.8809029141487886e-05,
+      "loss": 0.4066,
       "step": 64920
     },
     {
-      "epoch": 2.28,
-      "learning_rate": 2.9693173609446895e-05,
-      "loss": 0.2858,
+      "epoch": 2.3398926010019103,
+      "grad_norm": 0.19977350533008575,
+      "learning_rate": 2.8806145051445225e-05,
+      "loss": 0.3919,
       "step": 64925
     },
     {
-      "epoch": 2.28,
-      "learning_rate": 2.9690375504360162e-05,
-      "loss": 0.2574,
+      "epoch": 2.3400728006631346,
+      "grad_norm": 0.2010623812675476,
+      "learning_rate": 2.8803260909544578e-05,
+      "loss": 0.4305,
       "step": 64930
     },
     {
-      "epoch": 2.28,
-      "learning_rate": 2.9687577338371997e-05,
-      "loss": 0.2608,
+      "epoch": 2.3402530003243593,
+      "grad_norm": 0.17667075991630554,
+      "learning_rate": 2.880037671582524e-05,
+      "loss": 0.3747,
       "step": 64935
     },
     {
-      "epoch": 2.28,
-      "learning_rate": 2.9684779111518728e-05,
-      "loss": 0.2728,
+      "epoch": 2.340433199985584,
+      "grad_norm": 0.15319837629795074,
+      "learning_rate": 2.8797492470326497e-05,
+      "loss": 0.3922,
       "step": 64940
     },
     {
-      "epoch": 2.28,
-      "learning_rate": 2.9681980823836696e-05,
-      "loss": 0.2791,
+      "epoch": 2.340613399646809,
+      "grad_norm": 0.20747162401676178,
+      "learning_rate": 2.8794608173087655e-05,
+      "loss": 0.3958,
       "step": 64945
     },
     {
-      "epoch": 2.29,
-      "learning_rate": 2.967918247536222e-05,
-      "loss": 0.2715,
+      "epoch": 2.340793599308033,
+      "grad_norm": 0.2174191176891327,
+      "learning_rate": 2.879172382414802e-05,
+      "loss": 0.3928,
       "step": 64950
     },
     {
-      "epoch": 2.29,
-      "learning_rate": 2.9676384066131646e-05,
-      "loss": 0.2619,
+      "epoch": 2.340973798969258,
+      "grad_norm": 0.21791304647922516,
+      "learning_rate": 2.8788839423546877e-05,
+      "loss": 0.4441,
       "step": 64955
     },
     {
-      "epoch": 2.29,
-      "learning_rate": 2.9673585596181303e-05,
-      "loss": 0.2621,
+      "epoch": 2.3411539986304826,
+      "grad_norm": 0.22638213634490967,
+      "learning_rate": 2.8785954971323526e-05,
+      "loss": 0.402,
       "step": 64960
     },
     {
-      "epoch": 2.29,
-      "learning_rate": 2.9670787065547538e-05,
-      "loss": 0.2665,
+      "epoch": 2.3413341982917073,
+      "grad_norm": 0.20641210675239563,
+      "learning_rate": 2.8783070467517277e-05,
+      "loss": 0.4119,
       "step": 64965
     },
     {
-      "epoch": 2.29,
-      "learning_rate": 2.9667988474266678e-05,
-      "loss": 0.2658,
+      "epoch": 2.341514397952932,
+      "grad_norm": 0.2296299785375595,
+      "learning_rate": 2.8780185912167424e-05,
+      "loss": 0.4335,
       "step": 64970
     },
     {
-      "epoch": 2.29,
-      "learning_rate": 2.966518982237507e-05,
-      "loss": 0.2694,
+      "epoch": 2.3416945976141563,
+      "grad_norm": 0.18867331743240356,
+      "learning_rate": 2.8777301305313276e-05,
+      "loss": 0.3613,
       "step": 64975
     },
     {
-      "epoch": 2.29,
-      "learning_rate": 2.9662391109909038e-05,
-      "loss": 0.2698,
+      "epoch": 2.341874797275381,
+      "grad_norm": 0.27165400981903076,
+      "learning_rate": 2.8774416646994117e-05,
+      "loss": 0.4134,
       "step": 64980
     },
     {
-      "epoch": 2.29,
-      "learning_rate": 2.9659592336904933e-05,
-      "loss": 0.279,
+      "epoch": 2.342054996936606,
+      "grad_norm": 0.23474054038524628,
+      "learning_rate": 2.877153193724927e-05,
+      "loss": 0.3988,
       "step": 64985
     },
     {
-      "epoch": 2.29,
-      "learning_rate": 2.9656793503399095e-05,
-      "loss": 0.2622,
+      "epoch": 2.3422351965978305,
+      "grad_norm": 0.2028888314962387,
+      "learning_rate": 2.8768647176118024e-05,
+      "loss": 0.3996,
       "step": 64990
     },
     {
-      "epoch": 2.29,
-      "learning_rate": 2.965399460942786e-05,
-      "loss": 0.2711,
+      "epoch": 2.342415396259055,
+      "grad_norm": 0.29794272780418396,
+      "learning_rate": 2.8765762363639692e-05,
+      "loss": 0.4289,
       "step": 64995
     },
     {
-      "epoch": 2.29,
-      "learning_rate": 2.965119565502757e-05,
-      "loss": 0.2372,
+      "epoch": 2.3425955959202795,
+      "grad_norm": 0.2336840033531189,
+      "learning_rate": 2.8762877499853586e-05,
+      "loss": 0.3917,
       "step": 65000
     },
     {
-      "epoch": 2.29,
-      "eval_loss": 0.2671353220939636,
-      "eval_runtime": 10.5464,
-      "eval_samples_per_second": 9.482,
-      "eval_steps_per_second": 9.482,
+      "epoch": 2.3425955959202795,
+      "eval_loss": 0.4352496862411499,
+      "eval_runtime": 3.5177,
+      "eval_samples_per_second": 28.428,
+      "eval_steps_per_second": 7.107,
       "step": 65000
     },
     {
-      "epoch": 2.29,
-      "learning_rate": 2.9648396640234578e-05,
-      "loss": 0.2522,
+      "epoch": 2.3427757955815043,
+      "grad_norm": 0.21938785910606384,
+      "learning_rate": 2.8759992584798988e-05,
+      "loss": 0.4125,
       "step": 65005
     },
     {
-      "epoch": 2.29,
-      "learning_rate": 2.9645597565085208e-05,
-      "loss": 0.2762,
+      "epoch": 2.342955995242729,
+      "grad_norm": 0.1729237288236618,
+      "learning_rate": 2.8757107618515227e-05,
+      "loss": 0.405,
       "step": 65010
     },
     {
-      "epoch": 2.29,
-      "learning_rate": 2.964279842961582e-05,
-      "loss": 0.2681,
+      "epoch": 2.3431361949039538,
+      "grad_norm": 0.19324518740177155,
+      "learning_rate": 2.8754222601041603e-05,
+      "loss": 0.3719,
       "step": 65015
     },
     {
-      "epoch": 2.29,
-      "learning_rate": 2.9639999233862752e-05,
-      "loss": 0.2478,
+      "epoch": 2.343316394565178,
+      "grad_norm": 0.2103443443775177,
+      "learning_rate": 2.8751337532417428e-05,
+      "loss": 0.4041,
       "step": 65020
     },
     {
-      "epoch": 2.29,
-      "learning_rate": 2.963719997786235e-05,
-      "loss": 0.2876,
+      "epoch": 2.3434965942264028,
+      "grad_norm": 0.2286413609981537,
+      "learning_rate": 2.8748452412681994e-05,
+      "loss": 0.4066,
       "step": 65025
     },
     {
-      "epoch": 2.29,
-      "learning_rate": 2.9634400661650973e-05,
-      "loss": 0.2809,
+      "epoch": 2.3436767938876275,
+      "grad_norm": 0.2286667674779892,
+      "learning_rate": 2.8745567241874627e-05,
+      "loss": 0.3864,
       "step": 65030
     },
     {
-      "epoch": 2.29,
-      "learning_rate": 2.9631601285264947e-05,
-      "loss": 0.2557,
+      "epoch": 2.3438569935488522,
+      "grad_norm": 0.25851571559906006,
+      "learning_rate": 2.874268202003463e-05,
+      "loss": 0.4042,
       "step": 65035
     },
     {
-      "epoch": 2.29,
-      "learning_rate": 2.9628801848740633e-05,
-      "loss": 0.2618,
+      "epoch": 2.3440371932100765,
+      "grad_norm": 0.17988860607147217,
+      "learning_rate": 2.8739796747201313e-05,
+      "loss": 0.412,
       "step": 65040
     },
     {
-      "epoch": 2.29,
-      "learning_rate": 2.962600235211438e-05,
-      "loss": 0.2638,
+      "epoch": 2.3442173928713013,
+      "grad_norm": 0.21041817963123322,
+      "learning_rate": 2.8736911423414e-05,
+      "loss": 0.4372,
       "step": 65045
     },
     {
-      "epoch": 2.29,
-      "learning_rate": 2.9623202795422532e-05,
-      "loss": 0.2836,
+      "epoch": 2.344397592532526,
+      "grad_norm": 0.25124192237854004,
+      "learning_rate": 2.8734026048711975e-05,
+      "loss": 0.3904,
       "step": 65050
     },
     {
-      "epoch": 2.29,
-      "learning_rate": 2.9620403178701445e-05,
-      "loss": 0.261,
+      "epoch": 2.3445777921937507,
+      "grad_norm": 0.18272101879119873,
+      "learning_rate": 2.8731140623134583e-05,
+      "loss": 0.3723,
       "step": 65055
     },
     {
-      "epoch": 2.29,
-      "learning_rate": 2.9617603501987473e-05,
-      "loss": 0.2787,
+      "epoch": 2.3447579918549755,
+      "grad_norm": 0.2411685287952423,
+      "learning_rate": 2.8728255146721117e-05,
+      "loss": 0.3785,
       "step": 65060
     },
     {
-      "epoch": 2.29,
-      "learning_rate": 2.9614803765316956e-05,
-      "loss": 0.2619,
+      "epoch": 2.3449381915162,
+      "grad_norm": 0.2054976224899292,
+      "learning_rate": 2.8725369619510895e-05,
+      "loss": 0.4264,
       "step": 65065
     },
     {
-      "epoch": 2.29,
-      "learning_rate": 2.9612003968726254e-05,
-      "loss": 0.2937,
+      "epoch": 2.3451183911774245,
+      "grad_norm": 0.3253336250782013,
+      "learning_rate": 2.8722484041543233e-05,
+      "loss": 0.3911,
       "step": 65070
     },
     {
-      "epoch": 2.29,
-      "learning_rate": 2.9609204112251715e-05,
-      "loss": 0.2591,
+      "epoch": 2.345298590838649,
+      "grad_norm": 0.23166240751743317,
+      "learning_rate": 2.8719598412857445e-05,
+      "loss": 0.3862,
       "step": 65075
     },
     {
-      "epoch": 2.29,
-      "learning_rate": 2.960640419592971e-05,
-      "loss": 0.2509,
+      "epoch": 2.345478790499874,
+      "grad_norm": 0.17862895131111145,
+      "learning_rate": 2.871671273349286e-05,
+      "loss": 0.3967,
       "step": 65080
     },
     {
-      "epoch": 2.29,
-      "learning_rate": 2.9603604219796573e-05,
-      "loss": 0.2657,
+      "epoch": 2.3456589901610987,
+      "grad_norm": 0.2060977816581726,
+      "learning_rate": 2.8713827003488776e-05,
+      "loss": 0.3993,
       "step": 65085
     },
     {
-      "epoch": 2.29,
-      "learning_rate": 2.9600804183888675e-05,
-      "loss": 0.2633,
+      "epoch": 2.345839189822323,
+      "grad_norm": 0.1897481232881546,
+      "learning_rate": 2.8710941222884515e-05,
+      "loss": 0.3673,
       "step": 65090
     },
     {
-      "epoch": 2.29,
-      "learning_rate": 2.9598004088242355e-05,
-      "loss": 0.2567,
+      "epoch": 2.3460193894835477,
+      "grad_norm": 0.2062787413597107,
+      "learning_rate": 2.8708055391719396e-05,
+      "loss": 0.429,
       "step": 65095
     },
     {
-      "epoch": 2.29,
-      "learning_rate": 2.9595203932893996e-05,
-      "loss": 0.2692,
+      "epoch": 2.3461995891447724,
+      "grad_norm": 0.21658870577812195,
+      "learning_rate": 2.8705169510032747e-05,
+      "loss": 0.4001,
       "step": 65100
     },
     {
-      "epoch": 2.29,
-      "learning_rate": 2.9592403717879935e-05,
-      "loss": 0.2629,
+      "epoch": 2.346379788805997,
+      "grad_norm": 0.2051638811826706,
+      "learning_rate": 2.8702283577863883e-05,
+      "loss": 0.3861,
       "step": 65105
     },
     {
-      "epoch": 2.29,
-      "learning_rate": 2.958960344323654e-05,
-      "loss": 0.2529,
+      "epoch": 2.346559988467222,
+      "grad_norm": 0.22257257997989655,
+      "learning_rate": 2.8699397595252116e-05,
+      "loss": 0.3834,
       "step": 65110
     },
     {
-      "epoch": 2.29,
-      "learning_rate": 2.958680310900017e-05,
-      "loss": 0.278,
+      "epoch": 2.346740188128446,
+      "grad_norm": 0.1782115399837494,
+      "learning_rate": 2.8696511562236776e-05,
+      "loss": 0.4001,
       "step": 65115
     },
     {
-      "epoch": 2.29,
-      "learning_rate": 2.9584002715207182e-05,
-      "loss": 0.2807,
+      "epoch": 2.346920387789671,
+      "grad_norm": 0.2436537742614746,
+      "learning_rate": 2.8693625478857182e-05,
+      "loss": 0.4371,
       "step": 65120
     },
     {
-      "epoch": 2.29,
-      "learning_rate": 2.9581202261893937e-05,
-      "loss": 0.2419,
+      "epoch": 2.3471005874508957,
+      "grad_norm": 0.22284133732318878,
+      "learning_rate": 2.8690739345152658e-05,
+      "loss": 0.4084,
       "step": 65125
     },
     {
-      "epoch": 2.29,
-      "learning_rate": 2.9578401749096806e-05,
-      "loss": 0.2967,
+      "epoch": 2.3472807871121204,
+      "grad_norm": 0.19572477042675018,
+      "learning_rate": 2.8687853161162514e-05,
+      "loss": 0.3609,
       "step": 65130
     },
     {
-      "epoch": 2.29,
-      "learning_rate": 2.9575601176852146e-05,
-      "loss": 0.2738,
+      "epoch": 2.3474609867733447,
+      "grad_norm": 0.21674886345863342,
+      "learning_rate": 2.8684966926926092e-05,
+      "loss": 0.3779,
       "step": 65135
     },
     {
-      "epoch": 2.29,
-      "learning_rate": 2.9572800545196317e-05,
-      "loss": 0.2837,
+      "epoch": 2.3476411864345694,
+      "grad_norm": 0.2136157900094986,
+      "learning_rate": 2.8682080642482717e-05,
+      "loss": 0.4004,
       "step": 65140
     },
     {
-      "epoch": 2.29,
-      "learning_rate": 2.9569999854165687e-05,
-      "loss": 0.2811,
+      "epoch": 2.347821386095794,
+      "grad_norm": 0.22586745023727417,
+      "learning_rate": 2.8679194307871694e-05,
+      "loss": 0.4183,
       "step": 65145
     },
     {
-      "epoch": 2.29,
-      "learning_rate": 2.956719910379662e-05,
-      "loss": 0.2877,
+      "epoch": 2.348001585757019,
+      "grad_norm": 0.20550450682640076,
+      "learning_rate": 2.8676307923132367e-05,
+      "loss": 0.3483,
       "step": 65150
     },
     {
-      "epoch": 2.29,
-      "learning_rate": 2.9564398294125485e-05,
-      "loss": 0.2804,
+      "epoch": 2.3481817854182436,
+      "grad_norm": 0.18559207022190094,
+      "learning_rate": 2.8673421488304048e-05,
+      "loss": 0.3986,
       "step": 65155
     },
     {
-      "epoch": 2.29,
-      "learning_rate": 2.9561597425188642e-05,
-      "loss": 0.2756,
+      "epoch": 2.348361985079468,
+      "grad_norm": 0.23518766462802887,
+      "learning_rate": 2.867053500342609e-05,
+      "loss": 0.3831,
       "step": 65160
     },
     {
-      "epoch": 2.29,
-      "learning_rate": 2.955879649702247e-05,
-      "loss": 0.2892,
+      "epoch": 2.3485421847406927,
+      "grad_norm": 0.193081796169281,
+      "learning_rate": 2.866764846853779e-05,
+      "loss": 0.4086,
       "step": 65165
     },
     {
-      "epoch": 2.29,
-      "learning_rate": 2.955599550966332e-05,
-      "loss": 0.2629,
+      "epoch": 2.3487223844019174,
+      "grad_norm": 0.24136221408843994,
+      "learning_rate": 2.8664761883678493e-05,
+      "loss": 0.391,
       "step": 65170
     },
     {
-      "epoch": 2.29,
-      "learning_rate": 2.955319446314757e-05,
-      "loss": 0.2763,
+      "epoch": 2.348902584063142,
+      "grad_norm": 0.1842050552368164,
+      "learning_rate": 2.8661875248887515e-05,
+      "loss": 0.3772,
       "step": 65175
     },
     {
-      "epoch": 2.29,
-      "learning_rate": 2.9550393357511596e-05,
-      "loss": 0.2644,
+      "epoch": 2.3490827837243664,
+      "grad_norm": 0.16665658354759216,
+      "learning_rate": 2.865898856420421e-05,
+      "loss": 0.3921,
       "step": 65180
     },
     {
-      "epoch": 2.29,
-      "learning_rate": 2.9547592192791766e-05,
-      "loss": 0.2685,
+      "epoch": 2.349262983385591,
+      "grad_norm": 0.24213731288909912,
+      "learning_rate": 2.8656101829667892e-05,
+      "loss": 0.4142,
       "step": 65185
     },
     {
-      "epoch": 2.29,
-      "learning_rate": 2.954479096902445e-05,
-      "loss": 0.2634,
+      "epoch": 2.349443183046816,
+      "grad_norm": 0.16028551757335663,
+      "learning_rate": 2.8653215045317882e-05,
+      "loss": 0.4045,
       "step": 65190
     },
     {
-      "epoch": 2.29,
-      "learning_rate": 2.954198968624602e-05,
-      "loss": 0.304,
+      "epoch": 2.3496233827080406,
+      "grad_norm": 0.21428442001342773,
+      "learning_rate": 2.865032821119354e-05,
+      "loss": 0.3671,
       "step": 65195
     },
     {
-      "epoch": 2.29,
-      "learning_rate": 2.9539188344492837e-05,
-      "loss": 0.2807,
+      "epoch": 2.3498035823692653,
+      "grad_norm": 0.2171047478914261,
+      "learning_rate": 2.8647441327334166e-05,
+      "loss": 0.3788,
       "step": 65200
     },
     {
-      "epoch": 2.29,
-      "learning_rate": 2.9536386943801287e-05,
-      "loss": 0.3002,
+      "epoch": 2.3499837820304896,
+      "grad_norm": 0.17178626358509064,
+      "learning_rate": 2.8644554393779127e-05,
+      "loss": 0.404,
       "step": 65205
     },
     {
-      "epoch": 2.29,
-      "learning_rate": 2.9533585484207743e-05,
-      "loss": 0.253,
+      "epoch": 2.3501639816917144,
+      "grad_norm": 0.19908472895622253,
+      "learning_rate": 2.8641667410567736e-05,
+      "loss": 0.415,
       "step": 65210
     },
     {
-      "epoch": 2.29,
-      "learning_rate": 2.953078396574858e-05,
-      "loss": 0.295,
+      "epoch": 2.350344181352939,
+      "grad_norm": 0.1920705884695053,
+      "learning_rate": 2.8638780377739323e-05,
+      "loss": 0.4006,
       "step": 65215
     },
     {
-      "epoch": 2.29,
-      "learning_rate": 2.9527982388460175e-05,
-      "loss": 0.2554,
+      "epoch": 2.350524381014164,
+      "grad_norm": 0.22549034655094147,
+      "learning_rate": 2.863589329533324e-05,
+      "loss": 0.4142,
       "step": 65220
     },
     {
-      "epoch": 2.29,
-      "learning_rate": 2.9525180752378904e-05,
-      "loss": 0.274,
+      "epoch": 2.350704580675388,
+      "grad_norm": 0.19695831835269928,
+      "learning_rate": 2.8633006163388815e-05,
+      "loss": 0.3613,
       "step": 65225
     },
     {
-      "epoch": 2.29,
-      "learning_rate": 2.952237905754114e-05,
-      "loss": 0.275,
+      "epoch": 2.350884780336613,
+      "grad_norm": 0.2582644522190094,
+      "learning_rate": 2.8630118981945385e-05,
+      "loss": 0.3895,
       "step": 65230
     },
     {
-      "epoch": 2.3,
-      "learning_rate": 2.9519577303983264e-05,
-      "loss": 0.2911,
+      "epoch": 2.3510649799978376,
+      "grad_norm": 0.2204061895608902,
+      "learning_rate": 2.862723175104228e-05,
+      "loss": 0.4464,
       "step": 65235
     },
     {
-      "epoch": 2.3,
-      "learning_rate": 2.9516775491741654e-05,
-      "loss": 0.2803,
+      "epoch": 2.3512451796590623,
+      "grad_norm": 0.2712995707988739,
+      "learning_rate": 2.8624344470718852e-05,
+      "loss": 0.4043,
       "step": 65240
     },
     {
-      "epoch": 2.3,
-      "learning_rate": 2.95139736208527e-05,
-      "loss": 0.2559,
+      "epoch": 2.351425379320287,
+      "grad_norm": 0.17687354981899261,
+      "learning_rate": 2.862145714101443e-05,
+      "loss": 0.3772,
       "step": 65245
     },
     {
-      "epoch": 2.3,
-      "learning_rate": 2.9511171691352764e-05,
-      "loss": 0.2758,
+      "epoch": 2.3516055789815113,
+      "grad_norm": 0.22399690747261047,
+      "learning_rate": 2.861856976196835e-05,
+      "loss": 0.4168,
       "step": 65250
     },
     {
-      "epoch": 2.3,
-      "learning_rate": 2.950836970327823e-05,
-      "loss": 0.276,
+      "epoch": 2.351785778642736,
+      "grad_norm": 0.22232884168624878,
+      "learning_rate": 2.8615682333619957e-05,
+      "loss": 0.3784,
       "step": 65255
     },
     {
-      "epoch": 2.3,
-      "learning_rate": 2.9505567656665495e-05,
-      "loss": 0.2704,
+      "epoch": 2.351965978303961,
+      "grad_norm": 0.23076751828193665,
+      "learning_rate": 2.8612794856008595e-05,
+      "loss": 0.3769,
       "step": 65260
     },
     {
-      "epoch": 2.3,
-      "learning_rate": 2.950276555155093e-05,
-      "loss": 0.2763,
+      "epoch": 2.3521461779651855,
+      "grad_norm": 0.21458113193511963,
+      "learning_rate": 2.8609907329173613e-05,
+      "loss": 0.4084,
       "step": 65265
     },
     {
-      "epoch": 2.3,
-      "learning_rate": 2.949996338797093e-05,
-      "loss": 0.2608,
+      "epoch": 2.35232637762641,
+      "grad_norm": 0.22074739634990692,
+      "learning_rate": 2.860701975315433e-05,
+      "loss": 0.399,
       "step": 65270
     },
     {
-      "epoch": 2.3,
-      "learning_rate": 2.949716116596187e-05,
-      "loss": 0.2825,
+      "epoch": 2.3525065772876346,
+      "grad_norm": 0.20703339576721191,
+      "learning_rate": 2.8604132127990107e-05,
+      "loss": 0.3878,
       "step": 65275
     },
     {
-      "epoch": 2.3,
-      "learning_rate": 2.9494358885560138e-05,
-      "loss": 0.2638,
+      "epoch": 2.3526867769488593,
+      "grad_norm": 0.25511276721954346,
+      "learning_rate": 2.8601244453720276e-05,
+      "loss": 0.3976,
       "step": 65280
     },
     {
-      "epoch": 2.3,
-      "learning_rate": 2.9491556546802106e-05,
-      "loss": 0.2725,
+      "epoch": 2.352866976610084,
+      "grad_norm": 0.21042507886886597,
+      "learning_rate": 2.859835673038419e-05,
+      "loss": 0.4136,
       "step": 65285
     },
     {
-      "epoch": 2.3,
-      "learning_rate": 2.9488754149724175e-05,
-      "loss": 0.2593,
+      "epoch": 2.3530471762713088,
+      "grad_norm": 0.24099446833133698,
+      "learning_rate": 2.85954689580212e-05,
+      "loss": 0.4005,
       "step": 65290
     },
     {
-      "epoch": 2.3,
-      "learning_rate": 2.9485951694362734e-05,
-      "loss": 0.2668,
+      "epoch": 2.353227375932533,
+      "grad_norm": 0.18244995176792145,
+      "learning_rate": 2.8592581136670625e-05,
+      "loss": 0.3733,
       "step": 65295
     },
     {
-      "epoch": 2.3,
-      "learning_rate": 2.9483149180754168e-05,
-      "loss": 0.2634,
+      "epoch": 2.353407575593758,
+      "grad_norm": 0.25763794779777527,
+      "learning_rate": 2.8589693266371837e-05,
+      "loss": 0.394,
       "step": 65300
     },
     {
-      "epoch": 2.3,
-      "learning_rate": 2.948034660893486e-05,
-      "loss": 0.2597,
+      "epoch": 2.3535877752549825,
+      "grad_norm": 0.19655391573905945,
+      "learning_rate": 2.858680534716417e-05,
+      "loss": 0.4198,
       "step": 65305
     },
     {
-      "epoch": 2.3,
-      "learning_rate": 2.9477543978941213e-05,
-      "loss": 0.2492,
+      "epoch": 2.3537679749162073,
+      "grad_norm": 0.16229073703289032,
+      "learning_rate": 2.858391737908698e-05,
+      "loss": 0.3837,
       "step": 65310
     },
     {
-      "epoch": 2.3,
-      "learning_rate": 2.9474741290809592e-05,
-      "loss": 0.2786,
+      "epoch": 2.3539481745774316,
+      "grad_norm": 0.2285899817943573,
+      "learning_rate": 2.85810293621796e-05,
+      "loss": 0.4059,
       "step": 65315
     },
     {
-      "epoch": 2.3,
-      "learning_rate": 2.9471938544576417e-05,
-      "loss": 0.2849,
+      "epoch": 2.3541283742386563,
+      "grad_norm": 0.20628534257411957,
+      "learning_rate": 2.85781412964814e-05,
+      "loss": 0.4214,
       "step": 65320
     },
     {
-      "epoch": 2.3,
-      "learning_rate": 2.9469135740278066e-05,
-      "loss": 0.279,
+      "epoch": 2.354308573899881,
+      "grad_norm": 0.1957215815782547,
+      "learning_rate": 2.857525318203171e-05,
+      "loss": 0.391,
       "step": 65325
     },
     {
-      "epoch": 2.3,
-      "learning_rate": 2.946633287795093e-05,
-      "loss": 0.268,
+      "epoch": 2.3544887735611058,
+      "grad_norm": 0.1700507253408432,
+      "learning_rate": 2.8572365018869884e-05,
+      "loss": 0.3891,
       "step": 65330
     },
     {
-      "epoch": 2.3,
-      "learning_rate": 2.9463529957631404e-05,
-      "loss": 0.2755,
+      "epoch": 2.3546689732223305,
+      "grad_norm": 0.21826016902923584,
+      "learning_rate": 2.8569476807035278e-05,
+      "loss": 0.3958,
       "step": 65335
     },
     {
-      "epoch": 2.3,
-      "learning_rate": 2.946072697935588e-05,
-      "loss": 0.2858,
+      "epoch": 2.354849172883555,
+      "grad_norm": 0.1989716738462448,
+      "learning_rate": 2.8566588546567243e-05,
+      "loss": 0.4297,
       "step": 65340
     },
     {
-      "epoch": 2.3,
-      "learning_rate": 2.945792394316076e-05,
-      "loss": 0.2783,
+      "epoch": 2.3550293725447795,
+      "grad_norm": 0.20529745519161224,
+      "learning_rate": 2.856370023750513e-05,
+      "loss": 0.3922,
       "step": 65345
     },
     {
-      "epoch": 2.3,
-      "learning_rate": 2.9455120849082442e-05,
-      "loss": 0.2666,
+      "epoch": 2.3552095722060042,
+      "grad_norm": 0.1638742834329605,
+      "learning_rate": 2.8560811879888294e-05,
+      "loss": 0.3881,
       "step": 65350
     },
     {
-      "epoch": 2.3,
-      "learning_rate": 2.9452317697157307e-05,
-      "loss": 0.2716,
+      "epoch": 2.355389771867229,
+      "grad_norm": 0.16526326537132263,
+      "learning_rate": 2.855792347375608e-05,
+      "loss": 0.3958,
       "step": 65355
     },
     {
-      "epoch": 2.3,
-      "learning_rate": 2.9449514487421764e-05,
-      "loss": 0.2684,
+      "epoch": 2.3555699715284537,
+      "grad_norm": 0.17560982704162598,
+      "learning_rate": 2.8555035019147857e-05,
+      "loss": 0.3716,
       "step": 65360
     },
     {
-      "epoch": 2.3,
-      "learning_rate": 2.9446711219912203e-05,
-      "loss": 0.2683,
+      "epoch": 2.355750171189678,
+      "grad_norm": 0.21271197497844696,
+      "learning_rate": 2.855214651610296e-05,
+      "loss": 0.3795,
       "step": 65365
     },
     {
-      "epoch": 2.3,
-      "learning_rate": 2.9443907894665036e-05,
-      "loss": 0.2721,
+      "epoch": 2.3559303708509027,
+      "grad_norm": 0.18829388916492462,
+      "learning_rate": 2.8549257964660765e-05,
+      "loss": 0.4005,
       "step": 65370
     },
     {
-      "epoch": 2.3,
-      "learning_rate": 2.9441104511716645e-05,
-      "loss": 0.2548,
+      "epoch": 2.3561105705121275,
+      "grad_norm": 0.19685223698616028,
+      "learning_rate": 2.8546369364860608e-05,
+      "loss": 0.3688,
       "step": 65375
     },
     {
-      "epoch": 2.3,
-      "learning_rate": 2.943830107110344e-05,
-      "loss": 0.2763,
+      "epoch": 2.356290770173352,
+      "grad_norm": 0.2065257430076599,
+      "learning_rate": 2.854348071674186e-05,
+      "loss": 0.3742,
       "step": 65380
     },
     {
-      "epoch": 2.3,
-      "learning_rate": 2.9435497572861824e-05,
-      "loss": 0.2818,
+      "epoch": 2.356470969834577,
+      "grad_norm": 0.22780166566371918,
+      "learning_rate": 2.8540592020343872e-05,
+      "loss": 0.394,
       "step": 65385
     },
     {
-      "epoch": 2.3,
-      "learning_rate": 2.9432694017028195e-05,
-      "loss": 0.2857,
+      "epoch": 2.3566511694958012,
+      "grad_norm": 0.1743927299976349,
+      "learning_rate": 2.8537703275705997e-05,
+      "loss": 0.4064,
       "step": 65390
     },
     {
-      "epoch": 2.3,
-      "learning_rate": 2.9429890403638955e-05,
-      "loss": 0.2841,
+      "epoch": 2.356831369157026,
+      "grad_norm": 0.15711626410484314,
+      "learning_rate": 2.853481448286761e-05,
+      "loss": 0.3607,
       "step": 65395
     },
     {
-      "epoch": 2.3,
-      "learning_rate": 2.9427086732730506e-05,
-      "loss": 0.279,
+      "epoch": 2.3570115688182507,
+      "grad_norm": 0.2312869429588318,
+      "learning_rate": 2.853192564186805e-05,
+      "loss": 0.3852,
       "step": 65400
     },
     {
-      "epoch": 2.3,
-      "learning_rate": 2.942428300433926e-05,
-      "loss": 0.2693,
+      "epoch": 2.3571917684794754,
+      "grad_norm": 0.16384528577327728,
+      "learning_rate": 2.8529036752746697e-05,
+      "loss": 0.4077,
       "step": 65405
     },
     {
-      "epoch": 2.3,
-      "learning_rate": 2.9421479218501607e-05,
-      "loss": 0.2909,
+      "epoch": 2.3573719681406997,
+      "grad_norm": 0.24304990470409393,
+      "learning_rate": 2.8526147815542898e-05,
+      "loss": 0.4212,
       "step": 65410
     },
     {
-      "epoch": 2.3,
-      "learning_rate": 2.9418675375253968e-05,
-      "loss": 0.2845,
+      "epoch": 2.3575521678019244,
+      "grad_norm": 0.21454188227653503,
+      "learning_rate": 2.8523258830296017e-05,
+      "loss": 0.4003,
       "step": 65415
     },
     {
-      "epoch": 2.3,
-      "learning_rate": 2.9415871474632735e-05,
-      "loss": 0.2763,
+      "epoch": 2.357732367463149,
+      "grad_norm": 0.23705005645751953,
+      "learning_rate": 2.852036979704541e-05,
+      "loss": 0.3912,
       "step": 65420
     },
     {
-      "epoch": 2.3,
-      "learning_rate": 2.941306751667432e-05,
-      "loss": 0.2548,
+      "epoch": 2.357912567124374,
+      "grad_norm": 0.2235097885131836,
+      "learning_rate": 2.851748071583046e-05,
+      "loss": 0.4037,
       "step": 65425
     },
     {
-      "epoch": 2.3,
-      "learning_rate": 2.941026350141513e-05,
-      "loss": 0.2905,
+      "epoch": 2.3580927667855986,
+      "grad_norm": 0.2439454346895218,
+      "learning_rate": 2.851459158669051e-05,
+      "loss": 0.4144,
       "step": 65430
     },
     {
-      "epoch": 2.3,
-      "learning_rate": 2.9407459428891583e-05,
-      "loss": 0.2622,
+      "epoch": 2.358272966446823,
+      "grad_norm": 0.23461012542247772,
+      "learning_rate": 2.851170240966492e-05,
+      "loss": 0.4088,
       "step": 65435
     },
     {
-      "epoch": 2.3,
-      "learning_rate": 2.940465529914007e-05,
-      "loss": 0.2889,
+      "epoch": 2.3584531661080477,
+      "grad_norm": 0.21233990788459778,
+      "learning_rate": 2.8508813184793074e-05,
+      "loss": 0.3875,
       "step": 65440
     },
     {
-      "epoch": 2.3,
-      "learning_rate": 2.9401851112197022e-05,
-      "loss": 0.2797,
+      "epoch": 2.3586333657692724,
+      "grad_norm": 0.21089306473731995,
+      "learning_rate": 2.8505923912114324e-05,
+      "loss": 0.3812,
       "step": 65445
     },
     {
-      "epoch": 2.3,
-      "learning_rate": 2.9399046868098833e-05,
-      "loss": 0.2415,
+      "epoch": 2.358813565430497,
+      "grad_norm": 0.23103797435760498,
+      "learning_rate": 2.850303459166805e-05,
+      "loss": 0.4117,
       "step": 65450
     },
     {
-      "epoch": 2.3,
-      "learning_rate": 2.9396242566881916e-05,
-      "loss": 0.2487,
+      "epoch": 2.3589937650917214,
+      "grad_norm": 0.21833232045173645,
+      "learning_rate": 2.8500145223493595e-05,
+      "loss": 0.4149,
       "step": 65455
     },
     {
-      "epoch": 2.3,
-      "learning_rate": 2.939343820858269e-05,
-      "loss": 0.2523,
+      "epoch": 2.359173964752946,
+      "grad_norm": 0.20648404955863953,
+      "learning_rate": 2.8497255807630346e-05,
+      "loss": 0.4214,
       "step": 65460
     },
     {
-      "epoch": 2.3,
-      "learning_rate": 2.9390633793237566e-05,
-      "loss": 0.2733,
+      "epoch": 2.359354164414171,
+      "grad_norm": 0.1955990344285965,
+      "learning_rate": 2.849436634411766e-05,
+      "loss": 0.3793,
       "step": 65465
     },
     {
-      "epoch": 2.3,
-      "learning_rate": 2.9387829320882954e-05,
-      "loss": 0.2728,
+      "epoch": 2.3595343640753956,
+      "grad_norm": 0.18235990405082703,
+      "learning_rate": 2.8491476832994912e-05,
+      "loss": 0.3988,
       "step": 65470
     },
     {
-      "epoch": 2.3,
-      "learning_rate": 2.9385024791555267e-05,
-      "loss": 0.2709,
+      "epoch": 2.3597145637366204,
+      "grad_norm": 0.19686979055404663,
+      "learning_rate": 2.848858727430147e-05,
+      "loss": 0.4026,
       "step": 65475
     },
     {
-      "epoch": 2.3,
-      "learning_rate": 2.9382220205290923e-05,
-      "loss": 0.2774,
+      "epoch": 2.3598947633978447,
+      "grad_norm": 0.16368722915649414,
+      "learning_rate": 2.8485697668076694e-05,
+      "loss": 0.4115,
       "step": 65480
     },
     {
-      "epoch": 2.3,
-      "learning_rate": 2.9379415562126344e-05,
-      "loss": 0.2885,
+      "epoch": 2.3600749630590694,
+      "grad_norm": 0.21965758502483368,
+      "learning_rate": 2.848280801435997e-05,
+      "loss": 0.3961,
       "step": 65485
     },
     {
-      "epoch": 2.3,
-      "learning_rate": 2.9376610862097936e-05,
-      "loss": 0.2733,
+      "epoch": 2.360255162720294,
+      "grad_norm": 0.1920892298221588,
+      "learning_rate": 2.8479918313190657e-05,
+      "loss": 0.4003,
       "step": 65490
     },
     {
-      "epoch": 2.3,
-      "learning_rate": 2.937380610524212e-05,
-      "loss": 0.2483,
+      "epoch": 2.360435362381519,
+      "grad_norm": 0.18833613395690918,
+      "learning_rate": 2.8477028564608126e-05,
+      "loss": 0.3545,
       "step": 65495
     },
     {
-      "epoch": 2.3,
-      "learning_rate": 2.9371001291595314e-05,
-      "loss": 0.2939,
+      "epoch": 2.360615562042743,
+      "grad_norm": 0.17606264352798462,
+      "learning_rate": 2.8474138768651764e-05,
+      "loss": 0.3993,
       "step": 65500
     },
     {
-      "epoch": 2.3,
-      "eval_loss": 0.26703858375549316,
-      "eval_runtime": 10.541,
-      "eval_samples_per_second": 9.487,
-      "eval_steps_per_second": 9.487,
+      "epoch": 2.360615562042743,
+      "eval_loss": 0.43418970704078674,
+      "eval_runtime": 3.5195,
+      "eval_samples_per_second": 28.413,
+      "eval_steps_per_second": 7.103,
       "step": 65500
     },
     {
-      "epoch": 2.3,
-      "learning_rate": 2.936819642119394e-05,
-      "loss": 0.287,
+      "epoch": 2.360795761703968,
+      "grad_norm": 0.2237166315317154,
+      "learning_rate": 2.8471248925360927e-05,
+      "loss": 0.4224,
       "step": 65505
     },
     {
-      "epoch": 2.3,
-      "learning_rate": 2.9365391494074407e-05,
-      "loss": 0.2491,
+      "epoch": 2.3609759613651926,
+      "grad_norm": 0.2391594648361206,
+      "learning_rate": 2.8468359034775007e-05,
+      "loss": 0.3891,
       "step": 65510
     },
     {
-      "epoch": 2.3,
-      "learning_rate": 2.9362586510273143e-05,
-      "loss": 0.2859,
+      "epoch": 2.3611561610264173,
+      "grad_norm": 0.22656749188899994,
+      "learning_rate": 2.8465469096933352e-05,
+      "loss": 0.3941,
       "step": 65515
     },
     {
-      "epoch": 2.31,
-      "learning_rate": 2.935978146982658e-05,
-      "loss": 0.263,
+      "epoch": 2.361336360687642,
+      "grad_norm": 0.24217578768730164,
+      "learning_rate": 2.846257911187536e-05,
+      "loss": 0.3866,
       "step": 65520
     },
     {
-      "epoch": 2.31,
-      "learning_rate": 2.9356976372771117e-05,
-      "loss": 0.2789,
+      "epoch": 2.3615165603488664,
+      "grad_norm": 0.21993565559387207,
+      "learning_rate": 2.845968907964039e-05,
+      "loss": 0.4155,
       "step": 65525
     },
     {
-      "epoch": 2.31,
-      "learning_rate": 2.9354171219143193e-05,
-      "loss": 0.2859,
+      "epoch": 2.361696760010091,
+      "grad_norm": 0.2261502742767334,
+      "learning_rate": 2.845679900026783e-05,
+      "loss": 0.3222,
       "step": 65530
     },
     {
-      "epoch": 2.31,
-      "learning_rate": 2.9351366008979215e-05,
-      "loss": 0.2607,
+      "epoch": 2.361876959671316,
+      "grad_norm": 0.22237887978553772,
+      "learning_rate": 2.8453908873797058e-05,
+      "loss": 0.4065,
       "step": 65535
     },
     {
-      "epoch": 2.31,
-      "learning_rate": 2.934856074231563e-05,
-      "loss": 0.2644,
+      "epoch": 2.3620571593325406,
+      "grad_norm": 0.22978180646896362,
+      "learning_rate": 2.845101870026744e-05,
+      "loss": 0.3804,
       "step": 65540
     },
     {
-      "epoch": 2.31,
-      "learning_rate": 2.9345755419188843e-05,
-      "loss": 0.289,
+      "epoch": 2.362237358993765,
+      "grad_norm": 0.2067323923110962,
+      "learning_rate": 2.8448128479718363e-05,
+      "loss": 0.4058,
       "step": 65545
     },
     {
-      "epoch": 2.31,
-      "learning_rate": 2.9342950039635287e-05,
-      "loss": 0.287,
+      "epoch": 2.3624175586549896,
+      "grad_norm": 0.18871329724788666,
+      "learning_rate": 2.8445238212189208e-05,
+      "loss": 0.3929,
       "step": 65550
     },
     {
-      "epoch": 2.31,
-      "learning_rate": 2.9340144603691387e-05,
-      "loss": 0.2628,
+      "epoch": 2.3625977583162143,
+      "grad_norm": 0.20891360938549042,
+      "learning_rate": 2.8442347897719347e-05,
+      "loss": 0.3813,
       "step": 65555
     },
     {
-      "epoch": 2.31,
-      "learning_rate": 2.933733911139357e-05,
-      "loss": 0.2871,
+      "epoch": 2.362777957977439,
+      "grad_norm": 0.20510460436344147,
+      "learning_rate": 2.843945753634815e-05,
+      "loss": 0.3578,
       "step": 65560
     },
     {
-      "epoch": 2.31,
-      "learning_rate": 2.933453356277826e-05,
-      "loss": 0.2694,
+      "epoch": 2.362958157638664,
+      "grad_norm": 0.20665214955806732,
+      "learning_rate": 2.8436567128115022e-05,
+      "loss": 0.406,
       "step": 65565
     },
     {
-      "epoch": 2.31,
-      "learning_rate": 2.9331727957881892e-05,
-      "loss": 0.2994,
+      "epoch": 2.3631383572998885,
+      "grad_norm": 0.22137491405010223,
+      "learning_rate": 2.8433676673059335e-05,
+      "loss": 0.4062,
       "step": 65570
     },
     {
-      "epoch": 2.31,
-      "learning_rate": 2.932892229674089e-05,
-      "loss": 0.2825,
+      "epoch": 2.363318556961113,
+      "grad_norm": 0.17183031141757965,
+      "learning_rate": 2.8430786171220457e-05,
+      "loss": 0.3879,
       "step": 65575
     },
     {
-      "epoch": 2.31,
-      "learning_rate": 2.932611657939169e-05,
-      "loss": 0.2706,
+      "epoch": 2.3634987566223375,
+      "grad_norm": 0.2274576723575592,
+      "learning_rate": 2.842789562263779e-05,
+      "loss": 0.3876,
       "step": 65580
     },
     {
-      "epoch": 2.31,
-      "learning_rate": 2.932331080587071e-05,
-      "loss": 0.2541,
+      "epoch": 2.3636789562835623,
+      "grad_norm": 0.23537667095661163,
+      "learning_rate": 2.8425005027350704e-05,
+      "loss": 0.4282,
       "step": 65585
     },
     {
-      "epoch": 2.31,
-      "learning_rate": 2.9320504976214385e-05,
-      "loss": 0.2803,
+      "epoch": 2.363859155944787,
+      "grad_norm": 0.2088640332221985,
+      "learning_rate": 2.8422114385398597e-05,
+      "loss": 0.3923,
       "step": 65590
     },
     {
-      "epoch": 2.31,
-      "learning_rate": 2.9317699090459154e-05,
-      "loss": 0.2711,
+      "epoch": 2.3640393556060113,
+      "grad_norm": 0.19276586174964905,
+      "learning_rate": 2.841922369682083e-05,
+      "loss": 0.3885,
       "step": 65595
     },
     {
-      "epoch": 2.31,
-      "learning_rate": 2.9314893148641455e-05,
-      "loss": 0.2808,
+      "epoch": 2.364219555267236,
+      "grad_norm": 0.17863479256629944,
+      "learning_rate": 2.8416332961656812e-05,
+      "loss": 0.3814,
       "step": 65600
     },
     {
-      "epoch": 2.31,
-      "learning_rate": 2.9312087150797697e-05,
-      "loss": 0.2655,
+      "epoch": 2.3643997549284608,
+      "grad_norm": 0.21695052087306976,
+      "learning_rate": 2.8413442179945915e-05,
+      "loss": 0.395,
       "step": 65605
     },
     {
-      "epoch": 2.31,
-      "learning_rate": 2.9309281096964336e-05,
-      "loss": 0.2863,
+      "epoch": 2.3645799545896855,
+      "grad_norm": 0.20921245217323303,
+      "learning_rate": 2.8410551351727532e-05,
+      "loss": 0.3627,
       "step": 65610
     },
     {
-      "epoch": 2.31,
-      "learning_rate": 2.93064749871778e-05,
-      "loss": 0.2395,
+      "epoch": 2.3647601542509102,
+      "grad_norm": 0.18168587982654572,
+      "learning_rate": 2.8407660477041044e-05,
+      "loss": 0.3791,
       "step": 65615
     },
     {
-      "epoch": 2.31,
-      "learning_rate": 2.9303668821474522e-05,
-      "loss": 0.2728,
+      "epoch": 2.3649403539121345,
+      "grad_norm": 0.21238493919372559,
+      "learning_rate": 2.840476955592584e-05,
+      "loss": 0.4178,
       "step": 65620
     },
     {
-      "epoch": 2.31,
-      "learning_rate": 2.9300862599890944e-05,
-      "loss": 0.2665,
+      "epoch": 2.3651205535733593,
+      "grad_norm": 0.19079379737377167,
+      "learning_rate": 2.8401878588421316e-05,
+      "loss": 0.3875,
       "step": 65625
     },
     {
-      "epoch": 2.31,
-      "learning_rate": 2.92980563224635e-05,
-      "loss": 0.2773,
+      "epoch": 2.365300753234584,
+      "grad_norm": 0.17326843738555908,
+      "learning_rate": 2.8398987574566848e-05,
+      "loss": 0.3903,
       "step": 65630
     },
     {
-      "epoch": 2.31,
-      "learning_rate": 2.9295249989228628e-05,
-      "loss": 0.2719,
+      "epoch": 2.3654809528958087,
+      "grad_norm": 0.2072458267211914,
+      "learning_rate": 2.8396096514401838e-05,
+      "loss": 0.4034,
       "step": 65635
     },
     {
-      "epoch": 2.31,
-      "learning_rate": 2.9292443600222764e-05,
-      "loss": 0.2502,
+      "epoch": 2.365661152557033,
+      "grad_norm": 0.22681084275245667,
+      "learning_rate": 2.839320540796567e-05,
+      "loss": 0.3799,
       "step": 65640
     },
     {
-      "epoch": 2.31,
-      "learning_rate": 2.9289637155482346e-05,
-      "loss": 0.2657,
+      "epoch": 2.3658413522182578,
+      "grad_norm": 0.20382151007652283,
+      "learning_rate": 2.839031425529773e-05,
+      "loss": 0.4067,
       "step": 65645
     },
     {
-      "epoch": 2.31,
-      "learning_rate": 2.9286830655043817e-05,
-      "loss": 0.2493,
+      "epoch": 2.3660215518794825,
+      "grad_norm": 0.19017788767814636,
+      "learning_rate": 2.8387423056437417e-05,
+      "loss": 0.4157,
       "step": 65650
     },
     {
-      "epoch": 2.31,
-      "learning_rate": 2.9284024098943618e-05,
-      "loss": 0.2523,
+      "epoch": 2.366201751540707,
+      "grad_norm": 0.24520789086818695,
+      "learning_rate": 2.8384531811424127e-05,
+      "loss": 0.415,
       "step": 65655
     },
     {
-      "epoch": 2.31,
-      "learning_rate": 2.9281217487218193e-05,
-      "loss": 0.2934,
+      "epoch": 2.366381951201932,
+      "grad_norm": 0.23502486944198608,
+      "learning_rate": 2.8381640520297244e-05,
+      "loss": 0.4009,
       "step": 65660
     },
     {
-      "epoch": 2.31,
-      "learning_rate": 2.9278410819903978e-05,
-      "loss": 0.2807,
+      "epoch": 2.3665621508631562,
+      "grad_norm": 0.2212107628583908,
+      "learning_rate": 2.8378749183096154e-05,
+      "loss": 0.3814,
       "step": 65665
     },
     {
-      "epoch": 2.31,
-      "learning_rate": 2.9275604097037418e-05,
-      "loss": 0.3029,
+      "epoch": 2.366742350524381,
+      "grad_norm": 0.16753430664539337,
+      "learning_rate": 2.8375857799860272e-05,
+      "loss": 0.3745,
       "step": 65670
     },
     {
-      "epoch": 2.31,
-      "learning_rate": 2.9272797318654954e-05,
-      "loss": 0.2729,
+      "epoch": 2.3669225501856057,
+      "grad_norm": 0.1829628348350525,
+      "learning_rate": 2.8372966370628978e-05,
+      "loss": 0.4017,
       "step": 65675
     },
     {
-      "epoch": 2.31,
-      "learning_rate": 2.926999048479303e-05,
-      "loss": 0.284,
+      "epoch": 2.3671027498468304,
+      "grad_norm": 0.210256427526474,
+      "learning_rate": 2.8370074895441666e-05,
+      "loss": 0.3829,
       "step": 65680
     },
     {
-      "epoch": 2.31,
-      "learning_rate": 2.9267183595488106e-05,
-      "loss": 0.2547,
+      "epoch": 2.3672829495080547,
+      "grad_norm": 0.21395932137966156,
+      "learning_rate": 2.8367183374337743e-05,
+      "loss": 0.3984,
       "step": 65685
     },
     {
-      "epoch": 2.31,
-      "learning_rate": 2.9264376650776604e-05,
-      "loss": 0.2649,
+      "epoch": 2.3674631491692795,
+      "grad_norm": 0.22579269111156464,
+      "learning_rate": 2.8364291807356587e-05,
+      "loss": 0.4096,
       "step": 65690
     },
     {
-      "epoch": 2.31,
-      "learning_rate": 2.9261569650694986e-05,
-      "loss": 0.2765,
+      "epoch": 2.367643348830504,
+      "grad_norm": 0.22070454061031342,
+      "learning_rate": 2.8361400194537623e-05,
+      "loss": 0.4038,
       "step": 65695
     },
     {
-      "epoch": 2.31,
-      "learning_rate": 2.9258762595279693e-05,
-      "loss": 0.2556,
+      "epoch": 2.367823548491729,
+      "grad_norm": 0.18013955652713776,
+      "learning_rate": 2.835850853592022e-05,
+      "loss": 0.3868,
       "step": 65700
     },
     {
-      "epoch": 2.31,
-      "learning_rate": 2.9255955484567176e-05,
-      "loss": 0.3017,
+      "epoch": 2.3680037481529537,
+      "grad_norm": 0.1805589497089386,
+      "learning_rate": 2.835561683154379e-05,
+      "loss": 0.3842,
       "step": 65705
     },
     {
-      "epoch": 2.31,
-      "learning_rate": 2.925314831859388e-05,
-      "loss": 0.2805,
+      "epoch": 2.368183947814178,
+      "grad_norm": 0.17691951990127563,
+      "learning_rate": 2.835272508144773e-05,
+      "loss": 0.3759,
       "step": 65710
     },
     {
-      "epoch": 2.31,
-      "learning_rate": 2.9250341097396262e-05,
-      "loss": 0.2693,
+      "epoch": 2.3683641474754027,
+      "grad_norm": 0.18184374272823334,
+      "learning_rate": 2.8349833285671444e-05,
+      "loss": 0.3885,
       "step": 65715
     },
     {
-      "epoch": 2.31,
-      "learning_rate": 2.9247533821010758e-05,
-      "loss": 0.2638,
+      "epoch": 2.3685443471366274,
+      "grad_norm": 0.18779203295707703,
+      "learning_rate": 2.8346941444254327e-05,
+      "loss": 0.3972,
       "step": 65720
     },
     {
-      "epoch": 2.31,
-      "learning_rate": 2.9244726489473827e-05,
-      "loss": 0.2623,
+      "epoch": 2.368724546797852,
+      "grad_norm": 0.20252971351146698,
+      "learning_rate": 2.8344049557235775e-05,
+      "loss": 0.3733,
       "step": 65725
     },
     {
-      "epoch": 2.31,
-      "learning_rate": 2.924191910282192e-05,
-      "loss": 0.2724,
+      "epoch": 2.3689047464590764,
+      "grad_norm": 0.17999647557735443,
+      "learning_rate": 2.8341157624655202e-05,
+      "loss": 0.3842,
       "step": 65730
     },
     {
-      "epoch": 2.31,
-      "learning_rate": 2.9239111661091488e-05,
-      "loss": 0.2673,
+      "epoch": 2.369084946120301,
+      "grad_norm": 0.2151997685432434,
+      "learning_rate": 2.8338265646552002e-05,
+      "loss": 0.4194,
       "step": 65735
     },
     {
-      "epoch": 2.31,
-      "learning_rate": 2.9236304164318995e-05,
-      "loss": 0.2617,
+      "epoch": 2.369265145781526,
+      "grad_norm": 0.22379030287265778,
+      "learning_rate": 2.8335373622965576e-05,
+      "loss": 0.4103,
       "step": 65740
     },
     {
-      "epoch": 2.31,
-      "learning_rate": 2.9233496612540874e-05,
-      "loss": 0.2669,
+      "epoch": 2.3694453454427507,
+      "grad_norm": 0.20626239478588104,
+      "learning_rate": 2.833248155393533e-05,
+      "loss": 0.4013,
       "step": 65745
     },
     {
-      "epoch": 2.31,
-      "learning_rate": 2.9230689005793598e-05,
-      "loss": 0.2578,
+      "epoch": 2.3696255451039754,
+      "grad_norm": 0.2125103622674942,
+      "learning_rate": 2.8329589439500677e-05,
+      "loss": 0.4139,
       "step": 65750
     },
     {
-      "epoch": 2.31,
-      "learning_rate": 2.92278813441136e-05,
-      "loss": 0.2796,
+      "epoch": 2.3698057447651997,
+      "grad_norm": 0.20649980008602142,
+      "learning_rate": 2.8326697279701002e-05,
+      "loss": 0.4195,
       "step": 65755
     },
     {
-      "epoch": 2.31,
-      "learning_rate": 2.9225073627537358e-05,
-      "loss": 0.2737,
+      "epoch": 2.3699859444264244,
+      "grad_norm": 0.20015624165534973,
+      "learning_rate": 2.832380507457572e-05,
+      "loss": 0.3944,
       "step": 65760
     },
     {
-      "epoch": 2.31,
-      "learning_rate": 2.9222265856101317e-05,
-      "loss": 0.2745,
+      "epoch": 2.370166144087649,
+      "grad_norm": 0.20391733944416046,
+      "learning_rate": 2.8320912824164248e-05,
+      "loss": 0.4466,
       "step": 65765
     },
     {
-      "epoch": 2.31,
-      "learning_rate": 2.9219458029841945e-05,
-      "loss": 0.2825,
+      "epoch": 2.370346343748874,
+      "grad_norm": 0.21429723501205444,
+      "learning_rate": 2.8318020528505967e-05,
+      "loss": 0.4048,
       "step": 65770
     },
     {
-      "epoch": 2.31,
-      "learning_rate": 2.9216650148795683e-05,
-      "loss": 0.2576,
+      "epoch": 2.370526543410098,
+      "grad_norm": 0.2534154951572418,
+      "learning_rate": 2.831512818764031e-05,
+      "loss": 0.3978,
       "step": 65775
     },
     {
-      "epoch": 2.31,
-      "learning_rate": 2.9213842212998994e-05,
-      "loss": 0.2439,
+      "epoch": 2.370706743071323,
+      "grad_norm": 0.20116816461086273,
+      "learning_rate": 2.8312235801606674e-05,
+      "loss": 0.3887,
       "step": 65780
     },
     {
-      "epoch": 2.31,
-      "learning_rate": 2.9211034222488343e-05,
-      "loss": 0.2586,
+      "epoch": 2.3708869427325476,
+      "grad_norm": 0.199702188372612,
+      "learning_rate": 2.8309343370444457e-05,
+      "loss": 0.3907,
       "step": 65785
     },
     {
-      "epoch": 2.31,
-      "learning_rate": 2.9208226177300195e-05,
-      "loss": 0.2714,
+      "epoch": 2.3710671423937724,
+      "grad_norm": 0.21394966542720795,
+      "learning_rate": 2.8306450894193086e-05,
+      "loss": 0.3903,
       "step": 65790
     },
     {
-      "epoch": 2.31,
-      "learning_rate": 2.9205418077470998e-05,
-      "loss": 0.2634,
+      "epoch": 2.371247342054997,
+      "grad_norm": 0.21755504608154297,
+      "learning_rate": 2.8303558372891953e-05,
+      "loss": 0.3855,
       "step": 65795
     },
     {
-      "epoch": 2.32,
-      "learning_rate": 2.9202609923037222e-05,
-      "loss": 0.2927,
+      "epoch": 2.3714275417162214,
+      "grad_norm": 0.23619835078716278,
+      "learning_rate": 2.830066580658049e-05,
+      "loss": 0.4036,
       "step": 65800
     },
     {
-      "epoch": 2.32,
-      "learning_rate": 2.9199801714035324e-05,
-      "loss": 0.27,
+      "epoch": 2.371607741377446,
+      "grad_norm": 0.16918082535266876,
+      "learning_rate": 2.8297773195298084e-05,
+      "loss": 0.39,
       "step": 65805
     },
     {
-      "epoch": 2.32,
-      "learning_rate": 2.9196993450501764e-05,
-      "loss": 0.2663,
+      "epoch": 2.371787941038671,
+      "grad_norm": 0.20656408369541168,
+      "learning_rate": 2.8294880539084163e-05,
+      "loss": 0.3941,
       "step": 65810
     },
     {
-      "epoch": 2.32,
-      "learning_rate": 2.9194185132473013e-05,
-      "loss": 0.2596,
+      "epoch": 2.3719681406998956,
+      "grad_norm": 0.18240268528461456,
+      "learning_rate": 2.8291987837978125e-05,
+      "loss": 0.4155,
       "step": 65815
     },
     {
-      "epoch": 2.32,
-      "learning_rate": 2.9191376759985534e-05,
-      "loss": 0.268,
+      "epoch": 2.37214834036112,
+      "grad_norm": 0.19765466451644897,
+      "learning_rate": 2.8289095092019396e-05,
+      "loss": 0.3989,
       "step": 65820
     },
     {
-      "epoch": 2.32,
-      "learning_rate": 2.9188568333075787e-05,
-      "loss": 0.2997,
+      "epoch": 2.3723285400223446,
+      "grad_norm": 0.2510075271129608,
+      "learning_rate": 2.8286202301247382e-05,
+      "loss": 0.401,
       "step": 65825
     },
     {
-      "epoch": 2.32,
-      "learning_rate": 2.9185759851780247e-05,
-      "loss": 0.2841,
+      "epoch": 2.3725087396835693,
+      "grad_norm": 0.23590603470802307,
+      "learning_rate": 2.828330946570149e-05,
+      "loss": 0.4273,
       "step": 65830
     },
     {
-      "epoch": 2.32,
-      "learning_rate": 2.9182951316135367e-05,
-      "loss": 0.2822,
+      "epoch": 2.372688939344794,
+      "grad_norm": 0.20524172484874725,
+      "learning_rate": 2.8280416585421155e-05,
+      "loss": 0.3926,
       "step": 65835
     },
     {
-      "epoch": 2.32,
-      "learning_rate": 2.9180142726177616e-05,
-      "loss": 0.2858,
+      "epoch": 2.372869139006019,
+      "grad_norm": 0.21674352884292603,
+      "learning_rate": 2.8277523660445776e-05,
+      "loss": 0.3717,
       "step": 65840
     },
     {
-      "epoch": 2.32,
-      "learning_rate": 2.917733408194348e-05,
-      "loss": 0.2727,
+      "epoch": 2.3730493386672435,
+      "grad_norm": 0.20333261787891388,
+      "learning_rate": 2.827463069081477e-05,
+      "loss": 0.4245,
       "step": 65845
     },
     {
-      "epoch": 2.32,
-      "learning_rate": 2.9174525383469408e-05,
-      "loss": 0.2686,
+      "epoch": 2.373229538328468,
+      "grad_norm": 0.16487130522727966,
+      "learning_rate": 2.827173767656755e-05,
+      "loss": 0.3902,
       "step": 65850
     },
     {
-      "epoch": 2.32,
-      "learning_rate": 2.9171716630791878e-05,
-      "loss": 0.2368,
+      "epoch": 2.3734097379896926,
+      "grad_norm": 0.1875753402709961,
+      "learning_rate": 2.8268844617743544e-05,
+      "loss": 0.4189,
       "step": 65855
     },
     {
-      "epoch": 2.32,
-      "learning_rate": 2.916890782394735e-05,
-      "loss": 0.2818,
+      "epoch": 2.3735899376509173,
+      "grad_norm": 0.20067238807678223,
+      "learning_rate": 2.826595151438216e-05,
+      "loss": 0.3785,
       "step": 65860
     },
     {
-      "epoch": 2.32,
-      "learning_rate": 2.9166098962972306e-05,
-      "loss": 0.256,
+      "epoch": 2.373770137312142,
+      "grad_norm": 0.23048771917819977,
+      "learning_rate": 2.826305836652282e-05,
+      "loss": 0.4162,
       "step": 65865
     },
     {
-      "epoch": 2.32,
-      "learning_rate": 2.9163290047903208e-05,
-      "loss": 0.2594,
+      "epoch": 2.3739503369733663,
+      "grad_norm": 0.19079884886741638,
+      "learning_rate": 2.8260165174204938e-05,
+      "loss": 0.3493,
       "step": 65870
     },
     {
-      "epoch": 2.32,
-      "learning_rate": 2.9160481078776535e-05,
-      "loss": 0.259,
+      "epoch": 2.374130536634591,
+      "grad_norm": 0.1478530913591385,
+      "learning_rate": 2.825727193746794e-05,
+      "loss": 0.3911,
       "step": 65875
     },
     {
-      "epoch": 2.32,
-      "learning_rate": 2.915767205562876e-05,
-      "loss": 0.278,
+      "epoch": 2.374310736295816,
+      "grad_norm": 0.2297230362892151,
+      "learning_rate": 2.825437865635125e-05,
+      "loss": 0.4319,
       "step": 65880
     },
     {
-      "epoch": 2.32,
-      "learning_rate": 2.9154862978496354e-05,
-      "loss": 0.2681,
+      "epoch": 2.3744909359570405,
+      "grad_norm": 0.22544041275978088,
+      "learning_rate": 2.8251485330894266e-05,
+      "loss": 0.3452,
       "step": 65885
     },
     {
-      "epoch": 2.32,
-      "learning_rate": 2.9152053847415782e-05,
-      "loss": 0.2755,
+      "epoch": 2.3746711356182653,
+      "grad_norm": 0.19545534253120422,
+      "learning_rate": 2.8248591961136435e-05,
+      "loss": 0.4169,
       "step": 65890
     },
     {
-      "epoch": 2.32,
-      "learning_rate": 2.9149244662423535e-05,
-      "loss": 0.2759,
+      "epoch": 2.3748513352794896,
+      "grad_norm": 0.18258170783519745,
+      "learning_rate": 2.8245698547117162e-05,
+      "loss": 0.3857,
       "step": 65895
     },
     {
-      "epoch": 2.32,
-      "learning_rate": 2.914643542355608e-05,
-      "loss": 0.2689,
+      "epoch": 2.3750315349407143,
+      "grad_norm": 0.19753779470920563,
+      "learning_rate": 2.8242805088875874e-05,
+      "loss": 0.3941,
       "step": 65900
     },
     {
-      "epoch": 2.32,
-      "learning_rate": 2.9143626130849895e-05,
-      "loss": 0.2655,
+      "epoch": 2.375211734601939,
+      "grad_norm": 0.21284906566143036,
+      "learning_rate": 2.8239911586452e-05,
+      "loss": 0.3825,
       "step": 65905
     },
     {
-      "epoch": 2.32,
-      "learning_rate": 2.9140816784341445e-05,
-      "loss": 0.2749,
+      "epoch": 2.3753919342631638,
+      "grad_norm": 0.18012243509292603,
+      "learning_rate": 2.823701803988495e-05,
+      "loss": 0.3937,
       "step": 65910
     },
     {
-      "epoch": 2.32,
-      "learning_rate": 2.913800738406723e-05,
-      "loss": 0.277,
+      "epoch": 2.375572133924388,
+      "grad_norm": 0.2071056365966797,
+      "learning_rate": 2.8234124449214163e-05,
+      "loss": 0.3982,
       "step": 65915
     },
     {
-      "epoch": 2.32,
-      "learning_rate": 2.9135197930063708e-05,
-      "loss": 0.2265,
+      "epoch": 2.3757523335856128,
+      "grad_norm": 0.21273143589496613,
+      "learning_rate": 2.8231230814479052e-05,
+      "loss": 0.3928,
       "step": 65920
     },
     {
-      "epoch": 2.32,
-      "learning_rate": 2.9132388422367375e-05,
-      "loss": 0.2685,
+      "epoch": 2.3759325332468375,
+      "grad_norm": 0.21067939698696136,
+      "learning_rate": 2.8228337135719046e-05,
+      "loss": 0.3612,
       "step": 65925
     },
     {
-      "epoch": 2.32,
-      "learning_rate": 2.912957886101469e-05,
-      "loss": 0.2863,
+      "epoch": 2.3761127329080622,
+      "grad_norm": 0.22029119729995728,
+      "learning_rate": 2.822544341297358e-05,
+      "loss": 0.3829,
       "step": 65930
     },
     {
-      "epoch": 2.32,
-      "learning_rate": 2.912676924604216e-05,
-      "loss": 0.2608,
+      "epoch": 2.376292932569287,
+      "grad_norm": 0.2073868066072464,
+      "learning_rate": 2.822254964628206e-05,
+      "loss": 0.4605,
       "step": 65935
     },
     {
-      "epoch": 2.32,
-      "learning_rate": 2.912395957748624e-05,
-      "loss": 0.2767,
+      "epoch": 2.3764731322305113,
+      "grad_norm": 0.1779036968946457,
+      "learning_rate": 2.821965583568394e-05,
+      "loss": 0.3912,
       "step": 65940
     },
     {
-      "epoch": 2.32,
-      "learning_rate": 2.9121149855383417e-05,
-      "loss": 0.2627,
+      "epoch": 2.376653331891736,
+      "grad_norm": 0.1958044022321701,
+      "learning_rate": 2.821676198121862e-05,
+      "loss": 0.4263,
       "step": 65945
     },
     {
-      "epoch": 2.32,
-      "learning_rate": 2.911834007977019e-05,
-      "loss": 0.2609,
+      "epoch": 2.3768335315529607,
+      "grad_norm": 0.1888829916715622,
+      "learning_rate": 2.8213868082925542e-05,
+      "loss": 0.3649,
       "step": 65950
     },
     {
-      "epoch": 2.32,
-      "learning_rate": 2.911553025068302e-05,
-      "loss": 0.3032,
+      "epoch": 2.3770137312141855,
+      "grad_norm": 0.2218841016292572,
+      "learning_rate": 2.8210974140844137e-05,
+      "loss": 0.3925,
       "step": 65955
     },
     {
-      "epoch": 2.32,
-      "learning_rate": 2.911272036815842e-05,
-      "loss": 0.303,
+      "epoch": 2.3771939308754098,
+      "grad_norm": 0.19127243757247925,
+      "learning_rate": 2.8208080155013826e-05,
+      "loss": 0.4105,
       "step": 65960
     },
     {
-      "epoch": 2.32,
-      "learning_rate": 2.910991043223284e-05,
-      "loss": 0.2768,
+      "epoch": 2.3773741305366345,
+      "grad_norm": 0.19066543877124786,
+      "learning_rate": 2.8205186125474054e-05,
+      "loss": 0.3651,
       "step": 65965
     },
     {
-      "epoch": 2.32,
-      "learning_rate": 2.9107100442942793e-05,
-      "loss": 0.247,
+      "epoch": 2.3775543301978592,
+      "grad_norm": 0.1706237643957138,
+      "learning_rate": 2.8202292052264234e-05,
+      "loss": 0.4082,
       "step": 65970
     },
     {
-      "epoch": 2.32,
-      "learning_rate": 2.910429040032474e-05,
-      "loss": 0.2768,
+      "epoch": 2.377734529859084,
+      "grad_norm": 0.21013911068439484,
+      "learning_rate": 2.8199397935423805e-05,
+      "loss": 0.424,
       "step": 65975
     },
     {
-      "epoch": 2.32,
-      "learning_rate": 2.9101480304415196e-05,
-      "loss": 0.2787,
+      "epoch": 2.3779147295203087,
+      "grad_norm": 0.22082720696926117,
+      "learning_rate": 2.81965037749922e-05,
+      "loss": 0.3968,
       "step": 65980
     },
     {
-      "epoch": 2.32,
-      "learning_rate": 2.9098670155250624e-05,
-      "loss": 0.2531,
+      "epoch": 2.378094929181533,
+      "grad_norm": 0.21986991167068481,
+      "learning_rate": 2.8193609571008856e-05,
+      "loss": 0.4173,
       "step": 65985
     },
     {
-      "epoch": 2.32,
-      "learning_rate": 2.9095859952867527e-05,
-      "loss": 0.253,
+      "epoch": 2.3782751288427577,
+      "grad_norm": 0.22359991073608398,
+      "learning_rate": 2.819071532351319e-05,
+      "loss": 0.4279,
       "step": 65990
     },
     {
-      "epoch": 2.32,
-      "learning_rate": 2.9093049697302382e-05,
-      "loss": 0.2818,
+      "epoch": 2.3784553285039824,
+      "grad_norm": 0.18317356705665588,
+      "learning_rate": 2.8187821032544648e-05,
+      "loss": 0.382,
       "step": 65995
     },
     {
-      "epoch": 2.32,
-      "learning_rate": 2.9090239388591684e-05,
-      "loss": 0.2416,
+      "epoch": 2.378635528165207,
+      "grad_norm": 0.21548272669315338,
+      "learning_rate": 2.818492669814266e-05,
+      "loss": 0.3584,
       "step": 66000
     },
     {
-      "epoch": 2.32,
-      "eval_loss": 0.2667825520038605,
-      "eval_runtime": 10.5519,
-      "eval_samples_per_second": 9.477,
-      "eval_steps_per_second": 9.477,
+      "epoch": 2.378635528165207,
+      "eval_loss": 0.4347288906574249,
+      "eval_runtime": 3.5302,
+      "eval_samples_per_second": 28.327,
+      "eval_steps_per_second": 7.082,
       "step": 66000
     },
     {
-      "epoch": 2.32,
-      "learning_rate": 2.9087429026771924e-05,
-      "loss": 0.2977,
+      "epoch": 2.3788157278264315,
+      "grad_norm": 0.2195282280445099,
+      "learning_rate": 2.8182032320346668e-05,
+      "loss": 0.3977,
       "step": 66005
     },
     {
-      "epoch": 2.32,
-      "learning_rate": 2.9084618611879598e-05,
-      "loss": 0.2883,
+      "epoch": 2.378995927487656,
+      "grad_norm": 0.19582293927669525,
+      "learning_rate": 2.81791378991961e-05,
+      "loss": 0.3921,
       "step": 66010
     },
     {
-      "epoch": 2.32,
-      "learning_rate": 2.9081808143951185e-05,
-      "loss": 0.2677,
+      "epoch": 2.379176127148881,
+      "grad_norm": 0.1854901909828186,
+      "learning_rate": 2.8176243434730388e-05,
+      "loss": 0.3933,
       "step": 66015
     },
     {
-      "epoch": 2.32,
-      "learning_rate": 2.907899762302319e-05,
-      "loss": 0.2862,
+      "epoch": 2.3793563268101057,
+      "grad_norm": 0.2092180699110031,
+      "learning_rate": 2.817334892698898e-05,
+      "loss": 0.4335,
       "step": 66020
     },
     {
-      "epoch": 2.32,
-      "learning_rate": 2.907618704913209e-05,
-      "loss": 0.2749,
+      "epoch": 2.3795365264713304,
+      "grad_norm": 0.1754889339208603,
+      "learning_rate": 2.8170454376011307e-05,
+      "loss": 0.4288,
       "step": 66025
     },
     {
-      "epoch": 2.32,
-      "learning_rate": 2.907337642231439e-05,
-      "loss": 0.2824,
+      "epoch": 2.3797167261325547,
+      "grad_norm": 0.21382570266723633,
+      "learning_rate": 2.816755978183681e-05,
+      "loss": 0.4057,
       "step": 66030
     },
     {
-      "epoch": 2.32,
-      "learning_rate": 2.9070565742606582e-05,
-      "loss": 0.2703,
+      "epoch": 2.3798969257937794,
+      "grad_norm": 0.183826744556427,
+      "learning_rate": 2.8164665144504914e-05,
+      "loss": 0.3857,
       "step": 66035
     },
     {
-      "epoch": 2.32,
-      "learning_rate": 2.9067755010045167e-05,
-      "loss": 0.2743,
+      "epoch": 2.380077125455004,
+      "grad_norm": 0.17409420013427734,
+      "learning_rate": 2.8161770464055077e-05,
+      "loss": 0.394,
       "step": 66040
     },
     {
-      "epoch": 2.32,
-      "learning_rate": 2.9064944224666623e-05,
-      "loss": 0.2623,
+      "epoch": 2.380257325116229,
+      "grad_norm": 0.19534289836883545,
+      "learning_rate": 2.815887574052673e-05,
+      "loss": 0.4162,
       "step": 66045
     },
     {
-      "epoch": 2.32,
-      "learning_rate": 2.9062133386507467e-05,
-      "loss": 0.246,
+      "epoch": 2.380437524777453,
+      "grad_norm": 0.2002120018005371,
+      "learning_rate": 2.8155980973959308e-05,
+      "loss": 0.3953,
       "step": 66050
     },
     {
-      "epoch": 2.32,
-      "learning_rate": 2.9059322495604173e-05,
-      "loss": 0.2824,
+      "epoch": 2.380617724438678,
+      "grad_norm": 0.23090575635433197,
+      "learning_rate": 2.8153086164392258e-05,
+      "loss": 0.4104,
       "step": 66055
     },
     {
-      "epoch": 2.32,
-      "learning_rate": 2.9056511551993264e-05,
-      "loss": 0.2938,
+      "epoch": 2.3807979240999027,
+      "grad_norm": 0.22412335872650146,
+      "learning_rate": 2.8150191311865014e-05,
+      "loss": 0.4029,
       "step": 66060
     },
     {
-      "epoch": 2.32,
-      "learning_rate": 2.905370055571122e-05,
-      "loss": 0.2722,
+      "epoch": 2.3809781237611274,
+      "grad_norm": 0.1874934881925583,
+      "learning_rate": 2.814729641641703e-05,
+      "loss": 0.4051,
       "step": 66065
     },
     {
-      "epoch": 2.32,
-      "learning_rate": 2.9050889506794547e-05,
-      "loss": 0.26,
+      "epoch": 2.381158323422352,
+      "grad_norm": 0.23350152373313904,
+      "learning_rate": 2.8144401478087744e-05,
+      "loss": 0.4113,
       "step": 66070
     },
     {
-      "epoch": 2.32,
-      "learning_rate": 2.904807840527975e-05,
-      "loss": 0.2683,
+      "epoch": 2.381338523083577,
+      "grad_norm": 0.21121875941753387,
+      "learning_rate": 2.8141506496916586e-05,
+      "loss": 0.3871,
       "step": 66075
     },
     {
-      "epoch": 2.32,
-      "learning_rate": 2.904526725120332e-05,
-      "loss": 0.2762,
+      "epoch": 2.381518722744801,
+      "grad_norm": 0.1994001567363739,
+      "learning_rate": 2.8138611472943023e-05,
+      "loss": 0.4012,
       "step": 66080
     },
     {
-      "epoch": 2.33,
-      "learning_rate": 2.904245604460175e-05,
-      "loss": 0.2937,
+      "epoch": 2.381698922406026,
+      "grad_norm": 0.18203619122505188,
+      "learning_rate": 2.813571640620648e-05,
+      "loss": 0.3756,
       "step": 66085
     },
     {
-      "epoch": 2.33,
-      "learning_rate": 2.903964478551156e-05,
-      "loss": 0.2587,
+      "epoch": 2.3818791220672506,
+      "grad_norm": 0.18496555089950562,
+      "learning_rate": 2.813282129674641e-05,
+      "loss": 0.4213,
       "step": 66090
     },
     {
-      "epoch": 2.33,
-      "learning_rate": 2.903683347396925e-05,
-      "loss": 0.2781,
+      "epoch": 2.3820593217284753,
+      "grad_norm": 0.22467875480651855,
+      "learning_rate": 2.8129926144602247e-05,
+      "loss": 0.4402,
       "step": 66095
     },
     {
-      "epoch": 2.33,
-      "learning_rate": 2.9034022110011305e-05,
-      "loss": 0.2881,
+      "epoch": 2.3822395213896996,
+      "grad_norm": 0.1796996146440506,
+      "learning_rate": 2.812703094981346e-05,
+      "loss": 0.4387,
       "step": 66100
     },
     {
-      "epoch": 2.33,
-      "learning_rate": 2.9031210693674256e-05,
-      "loss": 0.2518,
+      "epoch": 2.3824197210509244,
+      "grad_norm": 0.20980200171470642,
+      "learning_rate": 2.8124135712419476e-05,
+      "loss": 0.4291,
       "step": 66105
     },
     {
-      "epoch": 2.33,
-      "learning_rate": 2.9028399224994586e-05,
-      "loss": 0.2767,
+      "epoch": 2.382599920712149,
+      "grad_norm": 0.1975623220205307,
+      "learning_rate": 2.8121240432459746e-05,
+      "loss": 0.3491,
       "step": 66110
     },
     {
-      "epoch": 2.33,
-      "learning_rate": 2.90255877040088e-05,
-      "loss": 0.2892,
+      "epoch": 2.382780120373374,
+      "grad_norm": 0.2462427318096161,
+      "learning_rate": 2.811834510997372e-05,
+      "loss": 0.3775,
       "step": 66115
     },
     {
-      "epoch": 2.33,
-      "learning_rate": 2.9022776130753414e-05,
-      "loss": 0.2792,
+      "epoch": 2.3829603200345986,
+      "grad_norm": 0.201690673828125,
+      "learning_rate": 2.8115449745000845e-05,
+      "loss": 0.4109,
       "step": 66120
     },
     {
-      "epoch": 2.33,
-      "learning_rate": 2.901996450526494e-05,
-      "loss": 0.2875,
+      "epoch": 2.383140519695823,
+      "grad_norm": 0.18545566499233246,
+      "learning_rate": 2.8112554337580577e-05,
+      "loss": 0.3889,
       "step": 66125
     },
     {
-      "epoch": 2.33,
-      "learning_rate": 2.901715282757987e-05,
-      "loss": 0.2502,
+      "epoch": 2.3833207193570476,
+      "grad_norm": 0.19686880707740784,
+      "learning_rate": 2.810965888775235e-05,
+      "loss": 0.3594,
       "step": 66130
     },
     {
-      "epoch": 2.33,
-      "learning_rate": 2.9014341097734717e-05,
-      "loss": 0.2566,
+      "epoch": 2.3835009190182723,
+      "grad_norm": 0.24028317630290985,
+      "learning_rate": 2.810676339555563e-05,
+      "loss": 0.4291,
       "step": 66135
     },
     {
-      "epoch": 2.33,
-      "learning_rate": 2.901152931576599e-05,
-      "loss": 0.2717,
+      "epoch": 2.383681118679497,
+      "grad_norm": 0.2691000998020172,
+      "learning_rate": 2.8103867861029852e-05,
+      "loss": 0.3862,
       "step": 66140
     },
     {
-      "epoch": 2.33,
-      "learning_rate": 2.90087174817102e-05,
-      "loss": 0.2481,
+      "epoch": 2.3838613183407213,
+      "grad_norm": 0.18645727634429932,
+      "learning_rate": 2.8100972284214476e-05,
+      "loss": 0.3964,
       "step": 66145
     },
     {
-      "epoch": 2.33,
-      "learning_rate": 2.9005905595603854e-05,
-      "loss": 0.2758,
+      "epoch": 2.384041518001946,
+      "grad_norm": 0.21478243172168732,
+      "learning_rate": 2.8098076665148965e-05,
+      "loss": 0.4038,
       "step": 66150
     },
     {
-      "epoch": 2.33,
-      "learning_rate": 2.900309365748347e-05,
-      "loss": 0.2538,
+      "epoch": 2.384221717663171,
+      "grad_norm": 0.22936570644378662,
+      "learning_rate": 2.8095181003872746e-05,
+      "loss": 0.3795,
       "step": 66155
     },
     {
-      "epoch": 2.33,
-      "learning_rate": 2.900028166738555e-05,
-      "loss": 0.2659,
+      "epoch": 2.3844019173243955,
+      "grad_norm": 0.22326894104480743,
+      "learning_rate": 2.809228530042529e-05,
+      "loss": 0.4041,
       "step": 66160
     },
     {
-      "epoch": 2.33,
-      "learning_rate": 2.89974696253466e-05,
-      "loss": 0.2753,
+      "epoch": 2.3845821169856203,
+      "grad_norm": 0.22681626677513123,
+      "learning_rate": 2.8089389554846048e-05,
+      "loss": 0.4157,
       "step": 66165
     },
     {
-      "epoch": 2.33,
-      "learning_rate": 2.8994657531403153e-05,
-      "loss": 0.2585,
+      "epoch": 2.3847623166468446,
+      "grad_norm": 0.19719672203063965,
+      "learning_rate": 2.808649376717447e-05,
+      "loss": 0.37,
       "step": 66170
     },
     {
-      "epoch": 2.33,
-      "learning_rate": 2.899184538559171e-05,
-      "loss": 0.2701,
+      "epoch": 2.3849425163080693,
+      "grad_norm": 0.22357481718063354,
+      "learning_rate": 2.8083597937450006e-05,
+      "loss": 0.385,
       "step": 66175
     },
     {
-      "epoch": 2.33,
-      "learning_rate": 2.8989033187948778e-05,
-      "loss": 0.2448,
+      "epoch": 2.385122715969294,
+      "grad_norm": 0.22523802518844604,
+      "learning_rate": 2.8080702065712127e-05,
+      "loss": 0.3971,
       "step": 66180
     },
     {
-      "epoch": 2.33,
-      "learning_rate": 2.8986220938510884e-05,
-      "loss": 0.2692,
+      "epoch": 2.3853029156305188,
+      "grad_norm": 0.2830953001976013,
+      "learning_rate": 2.8077806152000273e-05,
+      "loss": 0.3931,
       "step": 66185
     },
     {
-      "epoch": 2.33,
-      "learning_rate": 2.8983408637314535e-05,
-      "loss": 0.2859,
+      "epoch": 2.385483115291743,
+      "grad_norm": 0.21233409643173218,
+      "learning_rate": 2.8074910196353904e-05,
+      "loss": 0.3689,
       "step": 66190
     },
     {
-      "epoch": 2.33,
-      "learning_rate": 2.8980596284396246e-05,
-      "loss": 0.2792,
+      "epoch": 2.385663314952968,
+      "grad_norm": 0.20997081696987152,
+      "learning_rate": 2.8072014198812487e-05,
+      "loss": 0.3721,
       "step": 66195
     },
     {
-      "epoch": 2.33,
-      "learning_rate": 2.897778387979255e-05,
-      "loss": 0.2653,
+      "epoch": 2.3858435146141925,
+      "grad_norm": 0.19072303175926208,
+      "learning_rate": 2.8069118159415458e-05,
+      "loss": 0.3859,
       "step": 66200
     },
     {
-      "epoch": 2.33,
-      "learning_rate": 2.8974971423539937e-05,
-      "loss": 0.2734,
+      "epoch": 2.3860237142754173,
+      "grad_norm": 0.2057289481163025,
+      "learning_rate": 2.8066222078202303e-05,
+      "loss": 0.3966,
       "step": 66205
     },
     {
-      "epoch": 2.33,
-      "learning_rate": 2.897215891567495e-05,
-      "loss": 0.2732,
+      "epoch": 2.386203913936642,
+      "grad_norm": 0.1966446340084076,
+      "learning_rate": 2.806332595521246e-05,
+      "loss": 0.3794,
       "step": 66210
     },
     {
-      "epoch": 2.33,
-      "learning_rate": 2.8969346356234095e-05,
-      "loss": 0.2985,
+      "epoch": 2.3863841135978663,
+      "grad_norm": 0.2165762335062027,
+      "learning_rate": 2.8060429790485386e-05,
+      "loss": 0.4308,
       "step": 66215
     },
     {
-      "epoch": 2.33,
-      "learning_rate": 2.8966533745253893e-05,
-      "loss": 0.2597,
+      "epoch": 2.386564313259091,
+      "grad_norm": 0.1911364644765854,
+      "learning_rate": 2.8057533584060558e-05,
+      "loss": 0.3758,
       "step": 66220
     },
     {
-      "epoch": 2.33,
-      "learning_rate": 2.8963721082770863e-05,
-      "loss": 0.2916,
+      "epoch": 2.3867445129203158,
+      "grad_norm": 0.2317955642938614,
+      "learning_rate": 2.8054637335977423e-05,
+      "loss": 0.4062,
       "step": 66225
     },
     {
-      "epoch": 2.33,
-      "learning_rate": 2.896090836882153e-05,
-      "loss": 0.2687,
+      "epoch": 2.3869247125815405,
+      "grad_norm": 0.21293920278549194,
+      "learning_rate": 2.8051741046275453e-05,
+      "loss": 0.4013,
       "step": 66230
     },
     {
-      "epoch": 2.33,
-      "learning_rate": 2.895809560344241e-05,
-      "loss": 0.2851,
+      "epoch": 2.3871049122427648,
+      "grad_norm": 0.17797954380512238,
+      "learning_rate": 2.8048844714994092e-05,
+      "loss": 0.3899,
       "step": 66235
     },
     {
-      "epoch": 2.33,
-      "learning_rate": 2.895528278667003e-05,
-      "loss": 0.249,
+      "epoch": 2.3872851119039895,
+      "grad_norm": 0.21565930545330048,
+      "learning_rate": 2.8045948342172823e-05,
+      "loss": 0.3822,
       "step": 66240
     },
     {
-      "epoch": 2.33,
-      "learning_rate": 2.8952469918540908e-05,
-      "loss": 0.2783,
+      "epoch": 2.3874653115652142,
+      "grad_norm": 0.19301265478134155,
+      "learning_rate": 2.8043051927851083e-05,
+      "loss": 0.4029,
       "step": 66245
     },
     {
-      "epoch": 2.33,
-      "learning_rate": 2.8949656999091568e-05,
-      "loss": 0.2905,
+      "epoch": 2.387645511226439,
+      "grad_norm": 0.20318666100502014,
+      "learning_rate": 2.8040155472068365e-05,
+      "loss": 0.3968,
       "step": 66250
     },
     {
-      "epoch": 2.33,
-      "learning_rate": 2.8946844028358533e-05,
-      "loss": 0.255,
+      "epoch": 2.3878257108876637,
+      "grad_norm": 0.18554876744747162,
+      "learning_rate": 2.8037258974864105e-05,
+      "loss": 0.3907,
       "step": 66255
     },
     {
-      "epoch": 2.33,
-      "learning_rate": 2.8944031006378332e-05,
-      "loss": 0.253,
+      "epoch": 2.388005910548888,
+      "grad_norm": 0.2470291256904602,
+      "learning_rate": 2.8034362436277784e-05,
+      "loss": 0.4145,
       "step": 66260
     },
     {
-      "epoch": 2.33,
-      "learning_rate": 2.8941217933187487e-05,
-      "loss": 0.2725,
+      "epoch": 2.3881861102101127,
+      "grad_norm": 0.20082451403141022,
+      "learning_rate": 2.8031465856348865e-05,
+      "loss": 0.4123,
       "step": 66265
     },
     {
-      "epoch": 2.33,
-      "learning_rate": 2.8938404808822527e-05,
-      "loss": 0.2718,
+      "epoch": 2.3883663098713375,
+      "grad_norm": 0.19475050270557404,
+      "learning_rate": 2.802856923511681e-05,
+      "loss": 0.4013,
       "step": 66270
     },
     {
-      "epoch": 2.33,
-      "learning_rate": 2.893559163331997e-05,
-      "loss": 0.2692,
+      "epoch": 2.388546509532562,
+      "grad_norm": 0.22373121976852417,
+      "learning_rate": 2.8025672572621088e-05,
+      "loss": 0.3921,
       "step": 66275
     },
     {
-      "epoch": 2.33,
-      "learning_rate": 2.8932778406716353e-05,
-      "loss": 0.2846,
+      "epoch": 2.3887267091937865,
+      "grad_norm": 0.25062522292137146,
+      "learning_rate": 2.8022775868901153e-05,
+      "loss": 0.4029,
       "step": 66280
     },
     {
-      "epoch": 2.33,
-      "learning_rate": 2.8929965129048197e-05,
-      "loss": 0.2979,
+      "epoch": 2.3889069088550112,
+      "grad_norm": 0.19427599012851715,
+      "learning_rate": 2.801987912399649e-05,
+      "loss": 0.3677,
       "step": 66285
     },
     {
-      "epoch": 2.33,
-      "learning_rate": 2.8927151800352042e-05,
-      "loss": 0.2936,
+      "epoch": 2.389087108516236,
+      "grad_norm": 0.1876882165670395,
+      "learning_rate": 2.8016982337946556e-05,
+      "loss": 0.3781,
       "step": 66290
     },
     {
-      "epoch": 2.33,
-      "learning_rate": 2.8924338420664403e-05,
-      "loss": 0.2795,
+      "epoch": 2.3892673081774607,
+      "grad_norm": 0.18821240961551666,
+      "learning_rate": 2.801408551079082e-05,
+      "loss": 0.3557,
       "step": 66295
     },
     {
-      "epoch": 2.33,
-      "learning_rate": 2.8921524990021816e-05,
-      "loss": 0.2871,
+      "epoch": 2.3894475078386854,
+      "grad_norm": 0.21924707293510437,
+      "learning_rate": 2.801118864256876e-05,
+      "loss": 0.4088,
       "step": 66300
     },
     {
-      "epoch": 2.33,
-      "learning_rate": 2.8918711508460805e-05,
-      "loss": 0.2471,
+      "epoch": 2.3896277074999097,
+      "grad_norm": 0.19032910466194153,
+      "learning_rate": 2.8008291733319824e-05,
+      "loss": 0.3856,
       "step": 66305
     },
     {
-      "epoch": 2.33,
-      "learning_rate": 2.8915897976017915e-05,
-      "loss": 0.2777,
+      "epoch": 2.3898079071611344,
+      "grad_norm": 0.1727897822856903,
+      "learning_rate": 2.800539478308351e-05,
+      "loss": 0.4037,
       "step": 66310
     },
     {
-      "epoch": 2.33,
-      "learning_rate": 2.891308439272967e-05,
-      "loss": 0.2781,
+      "epoch": 2.389988106822359,
+      "grad_norm": 0.21995660662651062,
+      "learning_rate": 2.8002497791899268e-05,
+      "loss": 0.4113,
       "step": 66315
     },
     {
-      "epoch": 2.33,
-      "learning_rate": 2.8910270758632606e-05,
-      "loss": 0.2709,
+      "epoch": 2.390168306483584,
+      "grad_norm": 0.20189730823040009,
+      "learning_rate": 2.7999600759806577e-05,
+      "loss": 0.3665,
       "step": 66320
     },
     {
-      "epoch": 2.33,
-      "learning_rate": 2.8907457073763254e-05,
-      "loss": 0.2599,
+      "epoch": 2.390348506144808,
+      "grad_norm": 0.21424639225006104,
+      "learning_rate": 2.79967036868449e-05,
+      "loss": 0.4236,
       "step": 66325
     },
     {
-      "epoch": 2.33,
-      "learning_rate": 2.890464333815814e-05,
-      "loss": 0.291,
+      "epoch": 2.390528705806033,
+      "grad_norm": 0.2475886195898056,
+      "learning_rate": 2.7993806573053723e-05,
+      "loss": 0.4123,
       "step": 66330
     },
     {
-      "epoch": 2.33,
-      "learning_rate": 2.890182955185381e-05,
-      "loss": 0.2766,
+      "epoch": 2.3907089054672577,
+      "grad_norm": 0.17715711891651154,
+      "learning_rate": 2.7990909418472505e-05,
+      "loss": 0.4205,
       "step": 66335
     },
     {
-      "epoch": 2.33,
-      "learning_rate": 2.889901571488679e-05,
-      "loss": 0.2551,
+      "epoch": 2.3908891051284824,
+      "grad_norm": 0.20529335737228394,
+      "learning_rate": 2.7988012223140726e-05,
+      "loss": 0.4083,
       "step": 66340
     },
     {
-      "epoch": 2.33,
-      "learning_rate": 2.889620182729363e-05,
-      "loss": 0.2641,
+      "epoch": 2.391069304789707,
+      "grad_norm": 0.18749858438968658,
+      "learning_rate": 2.798511498709786e-05,
+      "loss": 0.4141,
       "step": 66345
     },
     {
-      "epoch": 2.33,
-      "learning_rate": 2.8893387889110845e-05,
-      "loss": 0.2837,
+      "epoch": 2.391249504450932,
+      "grad_norm": 0.18986235558986664,
+      "learning_rate": 2.7982217710383386e-05,
+      "loss": 0.4154,
       "step": 66350
     },
     {
-      "epoch": 2.33,
-      "learning_rate": 2.889057390037499e-05,
-      "loss": 0.2869,
+      "epoch": 2.391429704112156,
+      "grad_norm": 0.19709980487823486,
+      "learning_rate": 2.7979320393036762e-05,
+      "loss": 0.3889,
       "step": 66355
     },
     {
-      "epoch": 2.33,
-      "learning_rate": 2.8887759861122594e-05,
-      "loss": 0.2369,
+      "epoch": 2.391609903773381,
+      "grad_norm": 0.17932918667793274,
+      "learning_rate": 2.7976423035097486e-05,
+      "loss": 0.3893,
       "step": 66360
     },
     {
-      "epoch": 2.33,
-      "learning_rate": 2.8884945771390203e-05,
-      "loss": 0.2778,
+      "epoch": 2.3917901034346056,
+      "grad_norm": 0.20083005726337433,
+      "learning_rate": 2.7973525636605014e-05,
+      "loss": 0.4004,
       "step": 66365
     },
     {
-      "epoch": 2.34,
-      "learning_rate": 2.888213163121435e-05,
-      "loss": 0.277,
+      "epoch": 2.3919703030958304,
+      "grad_norm": 0.21819722652435303,
+      "learning_rate": 2.797062819759884e-05,
+      "loss": 0.414,
       "step": 66370
     },
     {
-      "epoch": 2.34,
-      "learning_rate": 2.887931744063158e-05,
-      "loss": 0.2471,
+      "epoch": 2.3921505027570547,
+      "grad_norm": 0.22273634374141693,
+      "learning_rate": 2.7967730718118424e-05,
+      "loss": 0.4096,
       "step": 66375
     },
     {
-      "epoch": 2.34,
-      "learning_rate": 2.8876503199678424e-05,
-      "loss": 0.2734,
+      "epoch": 2.3923307024182794,
+      "grad_norm": 0.18526360392570496,
+      "learning_rate": 2.7964833198203254e-05,
+      "loss": 0.3985,
       "step": 66380
     },
     {
-      "epoch": 2.34,
-      "learning_rate": 2.8873688908391426e-05,
-      "loss": 0.2756,
+      "epoch": 2.392510902079504,
+      "grad_norm": 0.2259715348482132,
+      "learning_rate": 2.79619356378928e-05,
+      "loss": 0.3778,
       "step": 66385
     },
     {
-      "epoch": 2.34,
-      "learning_rate": 2.8870874566807134e-05,
-      "loss": 0.2497,
+      "epoch": 2.392691101740729,
+      "grad_norm": 0.17341482639312744,
+      "learning_rate": 2.7959038037226554e-05,
+      "loss": 0.3816,
       "step": 66390
     },
     {
-      "epoch": 2.34,
-      "learning_rate": 2.8868060174962092e-05,
-      "loss": 0.2631,
+      "epoch": 2.3928713014019536,
+      "grad_norm": 0.22242747247219086,
+      "learning_rate": 2.7956140396243986e-05,
+      "loss": 0.378,
       "step": 66395
     },
     {
-      "epoch": 2.34,
-      "learning_rate": 2.886524573289283e-05,
-      "loss": 0.2787,
+      "epoch": 2.393051501063178,
+      "grad_norm": 0.19790546596050262,
+      "learning_rate": 2.7953242714984573e-05,
+      "loss": 0.3698,
       "step": 66400
     },
     {
-      "epoch": 2.34,
-      "learning_rate": 2.8862431240635906e-05,
-      "loss": 0.2505,
+      "epoch": 2.3932317007244026,
+      "grad_norm": 0.2102694809436798,
+      "learning_rate": 2.7950344993487803e-05,
+      "loss": 0.4262,
       "step": 66405
     },
     {
-      "epoch": 2.34,
-      "learning_rate": 2.8859616698227854e-05,
-      "loss": 0.2918,
+      "epoch": 2.3934119003856273,
+      "grad_norm": 0.19927994906902313,
+      "learning_rate": 2.7947447231793156e-05,
+      "loss": 0.4135,
       "step": 66410
     },
     {
-      "epoch": 2.34,
-      "learning_rate": 2.8856802105705216e-05,
-      "loss": 0.2516,
+      "epoch": 2.393592100046852,
+      "grad_norm": 0.20652970671653748,
+      "learning_rate": 2.7944549429940115e-05,
+      "loss": 0.3911,
       "step": 66415
     },
     {
-      "epoch": 2.34,
-      "learning_rate": 2.8853987463104554e-05,
-      "loss": 0.2788,
+      "epoch": 2.3937722997080764,
+      "grad_norm": 0.17121002078056335,
+      "learning_rate": 2.7941651587968147e-05,
+      "loss": 0.394,
       "step": 66420
     },
     {
-      "epoch": 2.34,
-      "learning_rate": 2.8851172770462398e-05,
-      "loss": 0.2767,
+      "epoch": 2.393952499369301,
+      "grad_norm": 0.20938289165496826,
+      "learning_rate": 2.7938753705916752e-05,
+      "loss": 0.3841,
       "step": 66425
     },
     {
-      "epoch": 2.34,
-      "learning_rate": 2.884835802781531e-05,
-      "loss": 0.2687,
+      "epoch": 2.394132699030526,
+      "grad_norm": 0.2229117602109909,
+      "learning_rate": 2.7935855783825406e-05,
+      "loss": 0.4056,
       "step": 66430
     },
     {
-      "epoch": 2.34,
-      "learning_rate": 2.8845543235199824e-05,
-      "loss": 0.2584,
+      "epoch": 2.3943128986917506,
+      "grad_norm": 0.2586416006088257,
+      "learning_rate": 2.7932957821733592e-05,
+      "loss": 0.4423,
       "step": 66435
     },
     {
-      "epoch": 2.34,
-      "learning_rate": 2.884272839265249e-05,
-      "loss": 0.2691,
+      "epoch": 2.3944930983529753,
+      "grad_norm": 0.22627079486846924,
+      "learning_rate": 2.79300598196808e-05,
+      "loss": 0.3877,
       "step": 66440
     },
     {
-      "epoch": 2.34,
-      "learning_rate": 2.8839913500209863e-05,
-      "loss": 0.269,
+      "epoch": 2.3946732980141996,
+      "grad_norm": 0.20481657981872559,
+      "learning_rate": 2.79271617777065e-05,
+      "loss": 0.3804,
       "step": 66445
     },
     {
-      "epoch": 2.34,
-      "learning_rate": 2.8837098557908493e-05,
-      "loss": 0.2869,
+      "epoch": 2.3948534976754243,
+      "grad_norm": 0.18298093974590302,
+      "learning_rate": 2.7924263695850196e-05,
+      "loss": 0.4435,
       "step": 66450
     },
     {
-      "epoch": 2.34,
-      "learning_rate": 2.8834283565784927e-05,
-      "loss": 0.266,
+      "epoch": 2.395033697336649,
+      "grad_norm": 0.2255946695804596,
+      "learning_rate": 2.7921365574151364e-05,
+      "loss": 0.3889,
       "step": 66455
     },
     {
-      "epoch": 2.34,
-      "learning_rate": 2.8831468523875715e-05,
-      "loss": 0.2755,
+      "epoch": 2.395213896997874,
+      "grad_norm": 0.15484662353992462,
+      "learning_rate": 2.7918467412649495e-05,
+      "loss": 0.3927,
       "step": 66460
     },
     {
-      "epoch": 2.34,
-      "learning_rate": 2.8828653432217405e-05,
-      "loss": 0.2646,
+      "epoch": 2.395394096659098,
+      "grad_norm": 0.23011015355587006,
+      "learning_rate": 2.7915569211384064e-05,
+      "loss": 0.3822,
       "step": 66465
     },
     {
-      "epoch": 2.34,
-      "learning_rate": 2.882583829084656e-05,
-      "loss": 0.2931,
+      "epoch": 2.395574296320323,
+      "grad_norm": 0.21641449630260468,
+      "learning_rate": 2.791267097039457e-05,
+      "loss": 0.3903,
       "step": 66470
     },
     {
-      "epoch": 2.34,
-      "learning_rate": 2.8823023099799717e-05,
-      "loss": 0.2787,
+      "epoch": 2.3957544959815475,
+      "grad_norm": 0.19246269762516022,
+      "learning_rate": 2.7909772689720503e-05,
+      "loss": 0.3717,
       "step": 66475
     },
     {
-      "epoch": 2.34,
-      "learning_rate": 2.882020785911345e-05,
-      "loss": 0.26,
+      "epoch": 2.3959346956427723,
+      "grad_norm": 0.1879129558801651,
+      "learning_rate": 2.7906874369401342e-05,
+      "loss": 0.4046,
       "step": 66480
     },
     {
-      "epoch": 2.34,
-      "learning_rate": 2.8817392568824292e-05,
-      "loss": 0.2788,
+      "epoch": 2.396114895303997,
+      "grad_norm": 0.20422397553920746,
+      "learning_rate": 2.7903976009476584e-05,
+      "loss": 0.353,
       "step": 66485
     },
     {
-      "epoch": 2.34,
-      "learning_rate": 2.8814577228968815e-05,
-      "loss": 0.2496,
+      "epoch": 2.3962950949652213,
+      "grad_norm": 0.19214318692684174,
+      "learning_rate": 2.7901077609985708e-05,
+      "loss": 0.3937,
       "step": 66490
     },
     {
-      "epoch": 2.34,
-      "learning_rate": 2.8811761839583558e-05,
-      "loss": 0.2569,
+      "epoch": 2.396475294626446,
+      "grad_norm": 0.20487315952777863,
+      "learning_rate": 2.789817917096822e-05,
+      "loss": 0.3955,
       "step": 66495
     },
     {
-      "epoch": 2.34,
-      "learning_rate": 2.8808946400705094e-05,
-      "loss": 0.3004,
+      "epoch": 2.3966554942876708,
+      "grad_norm": 0.19864313304424286,
+      "learning_rate": 2.78952806924636e-05,
+      "loss": 0.3657,
       "step": 66500
     },
     {
-      "epoch": 2.34,
-      "eval_loss": 0.2670895457267761,
-      "eval_runtime": 10.5428,
-      "eval_samples_per_second": 9.485,
-      "eval_steps_per_second": 9.485,
+      "epoch": 2.3966554942876708,
+      "eval_loss": 0.43440091609954834,
+      "eval_runtime": 3.5323,
+      "eval_samples_per_second": 28.31,
+      "eval_steps_per_second": 7.077,
       "step": 66500
     },
     {
-      "epoch": 2.34,
-      "learning_rate": 2.8806130912369966e-05,
-      "loss": 0.2656,
+      "epoch": 2.3968356939488955,
+      "grad_norm": 0.19604401290416718,
+      "learning_rate": 2.7892382174511333e-05,
+      "loss": 0.3751,
       "step": 66505
     },
     {
-      "epoch": 2.34,
-      "learning_rate": 2.8803315374614738e-05,
-      "loss": 0.2754,
+      "epoch": 2.39701589361012,
+      "grad_norm": 0.27748361229896545,
+      "learning_rate": 2.7889483617150924e-05,
+      "loss": 0.3847,
       "step": 66510
     },
     {
-      "epoch": 2.34,
-      "learning_rate": 2.8800499787475964e-05,
-      "loss": 0.273,
+      "epoch": 2.3971960932713445,
+      "grad_norm": 0.20608752965927124,
+      "learning_rate": 2.7886585020421863e-05,
+      "loss": 0.364,
       "step": 66515
     },
     {
-      "epoch": 2.34,
-      "learning_rate": 2.8797684150990206e-05,
-      "loss": 0.2715,
+      "epoch": 2.3973762929325693,
+      "grad_norm": 0.20215480029582977,
+      "learning_rate": 2.788368638436364e-05,
+      "loss": 0.3916,
       "step": 66520
     },
     {
-      "epoch": 2.34,
-      "learning_rate": 2.8794868465194025e-05,
-      "loss": 0.2471,
+      "epoch": 2.397556492593794,
+      "grad_norm": 0.25804924964904785,
+      "learning_rate": 2.7880787709015742e-05,
+      "loss": 0.4008,
       "step": 66525
     },
     {
-      "epoch": 2.34,
-      "learning_rate": 2.8792052730123976e-05,
-      "loss": 0.2625,
+      "epoch": 2.3977366922550187,
+      "grad_norm": 0.2014016956090927,
+      "learning_rate": 2.787788899441768e-05,
+      "loss": 0.3776,
       "step": 66530
     },
     {
-      "epoch": 2.34,
-      "learning_rate": 2.8789236945816624e-05,
-      "loss": 0.283,
+      "epoch": 2.397916891916243,
+      "grad_norm": 0.15427666902542114,
+      "learning_rate": 2.7874990240608927e-05,
+      "loss": 0.3746,
       "step": 66535
     },
     {
-      "epoch": 2.34,
-      "learning_rate": 2.878642111230852e-05,
-      "loss": 0.273,
+      "epoch": 2.3980970915774678,
+      "grad_norm": 0.15082351863384247,
+      "learning_rate": 2.787209144762899e-05,
+      "loss": 0.3814,
       "step": 66540
     },
     {
-      "epoch": 2.34,
-      "learning_rate": 2.878360522963624e-05,
-      "loss": 0.2841,
+      "epoch": 2.3982772912386925,
+      "grad_norm": 0.1636931151151657,
+      "learning_rate": 2.786919261551737e-05,
+      "loss": 0.3839,
       "step": 66545
     },
     {
-      "epoch": 2.34,
-      "learning_rate": 2.8780789297836334e-05,
-      "loss": 0.2855,
+      "epoch": 2.398457490899917,
+      "grad_norm": 0.17769426107406616,
+      "learning_rate": 2.786629374431355e-05,
+      "loss": 0.3759,
       "step": 66550
     },
     {
-      "epoch": 2.34,
-      "learning_rate": 2.8777973316945374e-05,
-      "loss": 0.2768,
+      "epoch": 2.3986376905611415,
+      "grad_norm": 0.2125890552997589,
+      "learning_rate": 2.7863394834057038e-05,
+      "loss": 0.4056,
       "step": 66555
     },
     {
-      "epoch": 2.34,
-      "learning_rate": 2.877515728699992e-05,
-      "loss": 0.2538,
+      "epoch": 2.3988178902223662,
+      "grad_norm": 0.18982288241386414,
+      "learning_rate": 2.7860495884787318e-05,
+      "loss": 0.4115,
       "step": 66560
     },
     {
-      "epoch": 2.34,
-      "learning_rate": 2.877234120803654e-05,
-      "loss": 0.2692,
+      "epoch": 2.398998089883591,
+      "grad_norm": 0.18557365238666534,
+      "learning_rate": 2.78575968965439e-05,
+      "loss": 0.3919,
       "step": 66565
     },
     {
-      "epoch": 2.34,
-      "learning_rate": 2.876952508009179e-05,
-      "loss": 0.2605,
+      "epoch": 2.3991782895448157,
+      "grad_norm": 0.17741166055202484,
+      "learning_rate": 2.7854697869366277e-05,
+      "loss": 0.4147,
       "step": 66570
     },
     {
-      "epoch": 2.34,
-      "learning_rate": 2.8766708903202245e-05,
-      "loss": 0.2634,
+      "epoch": 2.3993584892060404,
+      "grad_norm": 0.20264872908592224,
+      "learning_rate": 2.785179880329395e-05,
+      "loss": 0.3773,
       "step": 66575
     },
     {
-      "epoch": 2.34,
-      "learning_rate": 2.8763892677404465e-05,
-      "loss": 0.2392,
+      "epoch": 2.399538688867265,
+      "grad_norm": 0.20139038562774658,
+      "learning_rate": 2.7848899698366414e-05,
+      "loss": 0.3799,
       "step": 66580
     },
     {
-      "epoch": 2.34,
-      "learning_rate": 2.876107640273502e-05,
-      "loss": 0.2734,
+      "epoch": 2.3997188885284895,
+      "grad_norm": 0.21390214562416077,
+      "learning_rate": 2.7846000554623168e-05,
+      "loss": 0.4303,
       "step": 66585
     },
     {
-      "epoch": 2.34,
-      "learning_rate": 2.8758260079230475e-05,
-      "loss": 0.2385,
+      "epoch": 2.399899088189714,
+      "grad_norm": 0.20651130378246307,
+      "learning_rate": 2.7843101372103726e-05,
+      "loss": 0.4191,
       "step": 66590
     },
     {
-      "epoch": 2.34,
-      "learning_rate": 2.8755443706927403e-05,
-      "loss": 0.2735,
+      "epoch": 2.400079287850939,
+      "grad_norm": 0.20792073011398315,
+      "learning_rate": 2.7840202150847566e-05,
+      "loss": 0.3953,
       "step": 66595
     },
     {
-      "epoch": 2.34,
-      "learning_rate": 2.8752627285862365e-05,
-      "loss": 0.2632,
+      "epoch": 2.4002594875121637,
+      "grad_norm": 0.17147089540958405,
+      "learning_rate": 2.783730289089421e-05,
+      "loss": 0.3817,
       "step": 66600
     },
     {
-      "epoch": 2.34,
-      "learning_rate": 2.8749810816071936e-05,
-      "loss": 0.2805,
+      "epoch": 2.400439687173388,
+      "grad_norm": 0.1750619262456894,
+      "learning_rate": 2.7834403592283147e-05,
+      "loss": 0.4007,
       "step": 66605
     },
     {
-      "epoch": 2.34,
-      "learning_rate": 2.8746994297592688e-05,
-      "loss": 0.289,
+      "epoch": 2.4006198868346127,
+      "grad_norm": 0.20801058411598206,
+      "learning_rate": 2.7831504255053886e-05,
+      "loss": 0.3996,
       "step": 66610
     },
     {
-      "epoch": 2.34,
-      "learning_rate": 2.8744177730461185e-05,
-      "loss": 0.2455,
+      "epoch": 2.4008000864958374,
+      "grad_norm": 0.2012956142425537,
+      "learning_rate": 2.7828604879245923e-05,
+      "loss": 0.383,
       "step": 66615
     },
     {
-      "epoch": 2.34,
-      "learning_rate": 2.8741361114714006e-05,
-      "loss": 0.263,
+      "epoch": 2.400980286157062,
+      "grad_norm": 0.18420088291168213,
+      "learning_rate": 2.7825705464898777e-05,
+      "loss": 0.3725,
       "step": 66620
     },
     {
-      "epoch": 2.34,
-      "learning_rate": 2.8738544450387716e-05,
-      "loss": 0.2629,
+      "epoch": 2.401160485818287,
+      "grad_norm": 0.2484653741121292,
+      "learning_rate": 2.7822806012051934e-05,
+      "loss": 0.4174,
       "step": 66625
     },
     {
-      "epoch": 2.34,
-      "learning_rate": 2.8735727737518893e-05,
-      "loss": 0.2732,
+      "epoch": 2.401340685479511,
+      "grad_norm": 0.19384010136127472,
+      "learning_rate": 2.7819906520744903e-05,
+      "loss": 0.3899,
       "step": 66630
     },
     {
-      "epoch": 2.34,
-      "learning_rate": 2.87329109761441e-05,
-      "loss": 0.2799,
+      "epoch": 2.401520885140736,
+      "grad_norm": 0.19840480387210846,
+      "learning_rate": 2.7817006991017196e-05,
+      "loss": 0.3819,
       "step": 66635
     },
     {
-      "epoch": 2.34,
-      "learning_rate": 2.873009416629992e-05,
-      "loss": 0.2525,
+      "epoch": 2.4017010848019607,
+      "grad_norm": 0.20642586052417755,
+      "learning_rate": 2.7814107422908315e-05,
+      "loss": 0.3706,
       "step": 66640
     },
     {
-      "epoch": 2.34,
-      "learning_rate": 2.872727730802293e-05,
-      "loss": 0.2636,
+      "epoch": 2.4018812844631854,
+      "grad_norm": 0.20059265196323395,
+      "learning_rate": 2.7811207816457756e-05,
+      "loss": 0.3905,
       "step": 66645
     },
     {
-      "epoch": 2.34,
-      "learning_rate": 2.8724460401349696e-05,
-      "loss": 0.268,
+      "epoch": 2.4020614841244097,
+      "grad_norm": 0.19518163800239563,
+      "learning_rate": 2.7808308171705045e-05,
+      "loss": 0.3732,
       "step": 66650
     },
     {
-      "epoch": 2.35,
-      "learning_rate": 2.87216434463168e-05,
-      "loss": 0.2448,
+      "epoch": 2.4022416837856344,
+      "grad_norm": 0.2086179107427597,
+      "learning_rate": 2.780540848868967e-05,
+      "loss": 0.3932,
       "step": 66655
     },
     {
-      "epoch": 2.35,
-      "learning_rate": 2.871882644296081e-05,
-      "loss": 0.2645,
+      "epoch": 2.402421883446859,
+      "grad_norm": 0.18908485770225525,
+      "learning_rate": 2.780250876745116e-05,
+      "loss": 0.3602,
       "step": 66660
     },
     {
-      "epoch": 2.35,
-      "learning_rate": 2.8716009391318317e-05,
-      "loss": 0.2918,
+      "epoch": 2.402602083108084,
+      "grad_norm": 0.19856788218021393,
+      "learning_rate": 2.7799609008029004e-05,
+      "loss": 0.3977,
       "step": 66665
     },
     {
-      "epoch": 2.35,
-      "learning_rate": 2.8713192291425893e-05,
-      "loss": 0.2543,
+      "epoch": 2.4027822827693086,
+      "grad_norm": 0.1787065863609314,
+      "learning_rate": 2.779670921046272e-05,
+      "loss": 0.3638,
       "step": 66670
     },
     {
-      "epoch": 2.35,
-      "learning_rate": 2.8710375143320113e-05,
-      "loss": 0.2864,
+      "epoch": 2.402962482430533,
+      "grad_norm": 0.225900337100029,
+      "learning_rate": 2.7793809374791807e-05,
+      "loss": 0.3711,
       "step": 66675
     },
     {
-      "epoch": 2.35,
-      "learning_rate": 2.8707557947037555e-05,
-      "loss": 0.2812,
+      "epoch": 2.4031426820917576,
+      "grad_norm": 0.20942628383636475,
+      "learning_rate": 2.779090950105579e-05,
+      "loss": 0.3831,
       "step": 66680
     },
     {
-      "epoch": 2.35,
-      "learning_rate": 2.87047407026148e-05,
-      "loss": 0.2605,
+      "epoch": 2.4033228817529824,
+      "grad_norm": 0.1625223010778427,
+      "learning_rate": 2.7788009589294167e-05,
+      "loss": 0.368,
       "step": 66685
     },
     {
-      "epoch": 2.35,
-      "learning_rate": 2.8701923410088426e-05,
-      "loss": 0.2609,
+      "epoch": 2.403503081414207,
+      "grad_norm": 0.17064473032951355,
+      "learning_rate": 2.7785109639546454e-05,
+      "loss": 0.4337,
       "step": 66690
     },
     {
-      "epoch": 2.35,
-      "learning_rate": 2.869910606949502e-05,
-      "loss": 0.2737,
+      "epoch": 2.4036832810754314,
+      "grad_norm": 0.1989968866109848,
+      "learning_rate": 2.778220965185216e-05,
+      "loss": 0.3975,
       "step": 66695
     },
     {
-      "epoch": 2.35,
-      "learning_rate": 2.8696288680871164e-05,
-      "loss": 0.2603,
+      "epoch": 2.403863480736656,
+      "grad_norm": 0.1930345594882965,
+      "learning_rate": 2.77793096262508e-05,
+      "loss": 0.3987,
       "step": 66700
     },
     {
-      "epoch": 2.35,
-      "learning_rate": 2.869347124425343e-05,
-      "loss": 0.2734,
+      "epoch": 2.404043680397881,
+      "grad_norm": 0.22240734100341797,
+      "learning_rate": 2.7776409562781885e-05,
+      "loss": 0.4242,
       "step": 66705
     },
     {
-      "epoch": 2.35,
-      "learning_rate": 2.869065375967841e-05,
-      "loss": 0.2654,
+      "epoch": 2.4042238800591056,
+      "grad_norm": 0.22761568427085876,
+      "learning_rate": 2.7773509461484924e-05,
+      "loss": 0.3846,
       "step": 66710
     },
     {
-      "epoch": 2.35,
-      "learning_rate": 2.8687836227182668e-05,
-      "loss": 0.2671,
+      "epoch": 2.4044040797203303,
+      "grad_norm": 0.18993456661701202,
+      "learning_rate": 2.7770609322399438e-05,
+      "loss": 0.3828,
       "step": 66715
     },
     {
-      "epoch": 2.35,
-      "learning_rate": 2.868501864680282e-05,
-      "loss": 0.2596,
+      "epoch": 2.4045842793815546,
+      "grad_norm": 0.253111869096756,
+      "learning_rate": 2.7767709145564936e-05,
+      "loss": 0.3801,
       "step": 66720
     },
     {
-      "epoch": 2.35,
-      "learning_rate": 2.8682201018575426e-05,
-      "loss": 0.2768,
+      "epoch": 2.4047644790427793,
+      "grad_norm": 0.2091599851846695,
+      "learning_rate": 2.7764808931020923e-05,
+      "loss": 0.3715,
       "step": 66725
     },
     {
-      "epoch": 2.35,
-      "learning_rate": 2.8679383342537087e-05,
-      "loss": 0.2812,
+      "epoch": 2.404944678704004,
+      "grad_norm": 0.19851459562778473,
+      "learning_rate": 2.7761908678806937e-05,
+      "loss": 0.4124,
       "step": 66730
     },
     {
-      "epoch": 2.35,
-      "learning_rate": 2.867656561872437e-05,
-      "loss": 0.2778,
+      "epoch": 2.405124878365229,
+      "grad_norm": 0.19328032433986664,
+      "learning_rate": 2.775900838896247e-05,
+      "loss": 0.3716,
       "step": 66735
     },
     {
-      "epoch": 2.35,
-      "learning_rate": 2.867374784717387e-05,
-      "loss": 0.2537,
+      "epoch": 2.405305078026453,
+      "grad_norm": 0.18993335962295532,
+      "learning_rate": 2.7756108061527062e-05,
+      "loss": 0.3552,
       "step": 66740
     },
     {
-      "epoch": 2.35,
-      "learning_rate": 2.8670930027922183e-05,
-      "loss": 0.3001,
+      "epoch": 2.405485277687678,
+      "grad_norm": 0.19817394018173218,
+      "learning_rate": 2.77532076965402e-05,
+      "loss": 0.4001,
       "step": 66745
     },
     {
-      "epoch": 2.35,
-      "learning_rate": 2.866811216100589e-05,
-      "loss": 0.2672,
+      "epoch": 2.4056654773489026,
+      "grad_norm": 0.19719624519348145,
+      "learning_rate": 2.7750307294041423e-05,
+      "loss": 0.3777,
       "step": 66750
     },
     {
-      "epoch": 2.35,
-      "learning_rate": 2.866529424646157e-05,
-      "loss": 0.2582,
+      "epoch": 2.4058456770101273,
+      "grad_norm": 0.20810894668102264,
+      "learning_rate": 2.774740685407024e-05,
+      "loss": 0.3946,
       "step": 66755
     },
     {
-      "epoch": 2.35,
-      "learning_rate": 2.8662476284325823e-05,
-      "loss": 0.2792,
+      "epoch": 2.406025876671352,
+      "grad_norm": 0.1935310959815979,
+      "learning_rate": 2.7744506376666175e-05,
+      "loss": 0.3903,
       "step": 66760
     },
     {
-      "epoch": 2.35,
-      "learning_rate": 2.865965827463524e-05,
-      "loss": 0.2657,
+      "epoch": 2.4062060763325763,
+      "grad_norm": 0.24938994646072388,
+      "learning_rate": 2.7741605861868735e-05,
+      "loss": 0.3914,
       "step": 66765
     },
     {
-      "epoch": 2.35,
-      "learning_rate": 2.8656840217426394e-05,
-      "loss": 0.2705,
+      "epoch": 2.406386275993801,
+      "grad_norm": 0.22419728338718414,
+      "learning_rate": 2.7738705309717444e-05,
+      "loss": 0.4173,
       "step": 66770
     },
     {
-      "epoch": 2.35,
-      "learning_rate": 2.8654022112735895e-05,
-      "loss": 0.2326,
+      "epoch": 2.406566475655026,
+      "grad_norm": 0.17850545048713684,
+      "learning_rate": 2.7735804720251835e-05,
+      "loss": 0.4045,
       "step": 66775
     },
     {
-      "epoch": 2.35,
-      "learning_rate": 2.8651203960600325e-05,
-      "loss": 0.2847,
+      "epoch": 2.4067466753162505,
+      "grad_norm": 0.23113694787025452,
+      "learning_rate": 2.773290409351141e-05,
+      "loss": 0.3859,
       "step": 66780
     },
     {
-      "epoch": 2.35,
-      "learning_rate": 2.8648385761056285e-05,
-      "loss": 0.2878,
+      "epoch": 2.406926874977475,
+      "grad_norm": 0.24465039372444153,
+      "learning_rate": 2.7730003429535688e-05,
+      "loss": 0.3736,
       "step": 66785
     },
     {
-      "epoch": 2.35,
-      "learning_rate": 2.8645567514140355e-05,
-      "loss": 0.2512,
+      "epoch": 2.4071070746386996,
+      "grad_norm": 0.22059732675552368,
+      "learning_rate": 2.7727102728364207e-05,
+      "loss": 0.3946,
       "step": 66790
     },
     {
-      "epoch": 2.35,
-      "learning_rate": 2.864274921988913e-05,
-      "loss": 0.2916,
+      "epoch": 2.4072872742999243,
+      "grad_norm": 0.20629587769508362,
+      "learning_rate": 2.7724201990036474e-05,
+      "loss": 0.3921,
       "step": 66795
     },
     {
-      "epoch": 2.35,
-      "learning_rate": 2.863993087833921e-05,
-      "loss": 0.2613,
+      "epoch": 2.407467473961149,
+      "grad_norm": 0.2906322479248047,
+      "learning_rate": 2.772130121459202e-05,
+      "loss": 0.3928,
       "step": 66800
     },
     {
-      "epoch": 2.35,
-      "learning_rate": 2.8637112489527195e-05,
-      "loss": 0.2886,
+      "epoch": 2.4076476736223738,
+      "grad_norm": 0.1948506087064743,
+      "learning_rate": 2.771840040207037e-05,
+      "loss": 0.3751,
       "step": 66805
     },
     {
-      "epoch": 2.35,
-      "learning_rate": 2.8634294053489658e-05,
-      "loss": 0.2658,
+      "epoch": 2.407827873283598,
+      "grad_norm": 0.1990462988615036,
+      "learning_rate": 2.7715499552511036e-05,
+      "loss": 0.4181,
       "step": 66810
     },
     {
-      "epoch": 2.35,
-      "learning_rate": 2.8631475570263218e-05,
-      "loss": 0.2933,
+      "epoch": 2.4080080729448228,
+      "grad_norm": 0.23790279030799866,
+      "learning_rate": 2.7712598665953543e-05,
+      "loss": 0.3669,
       "step": 66815
     },
     {
-      "epoch": 2.35,
-      "learning_rate": 2.8628657039884455e-05,
-      "loss": 0.2801,
+      "epoch": 2.4081882726060475,
+      "grad_norm": 0.18465937674045563,
+      "learning_rate": 2.7709697742437418e-05,
+      "loss": 0.4092,
       "step": 66820
     },
     {
-      "epoch": 2.35,
-      "learning_rate": 2.862583846238997e-05,
-      "loss": 0.2542,
+      "epoch": 2.4083684722672722,
+      "grad_norm": 0.16770188510417938,
+      "learning_rate": 2.7706796782002192e-05,
+      "loss": 0.37,
       "step": 66825
     },
     {
-      "epoch": 2.35,
-      "learning_rate": 2.8623019837816367e-05,
-      "loss": 0.2774,
+      "epoch": 2.4085486719284965,
+      "grad_norm": 0.2589314877986908,
+      "learning_rate": 2.770389578468738e-05,
+      "loss": 0.4184,
       "step": 66830
     },
     {
-      "epoch": 2.35,
-      "learning_rate": 2.8620201166200238e-05,
-      "loss": 0.2683,
+      "epoch": 2.4087288715897213,
+      "grad_norm": 0.18053090572357178,
+      "learning_rate": 2.770099475053251e-05,
+      "loss": 0.3487,
       "step": 66835
     },
     {
-      "epoch": 2.35,
-      "learning_rate": 2.8617382447578178e-05,
-      "loss": 0.2716,
+      "epoch": 2.408909071250946,
+      "grad_norm": 0.2627122700214386,
+      "learning_rate": 2.7698093679577108e-05,
+      "loss": 0.4316,
       "step": 66840
     },
     {
-      "epoch": 2.35,
-      "learning_rate": 2.86145636819868e-05,
-      "loss": 0.2712,
+      "epoch": 2.4090892709121707,
+      "grad_norm": 0.18675784766674042,
+      "learning_rate": 2.769519257186071e-05,
+      "loss": 0.3685,
       "step": 66845
     },
     {
-      "epoch": 2.35,
-      "learning_rate": 2.8611744869462688e-05,
-      "loss": 0.2498,
+      "epoch": 2.4092694705733955,
+      "grad_norm": 0.27119845151901245,
+      "learning_rate": 2.7692291427422828e-05,
+      "loss": 0.4013,
       "step": 66850
     },
     {
-      "epoch": 2.35,
-      "learning_rate": 2.8608926010042447e-05,
-      "loss": 0.2782,
+      "epoch": 2.40944967023462,
+      "grad_norm": 0.19368845224380493,
+      "learning_rate": 2.7689390246302997e-05,
+      "loss": 0.3814,
       "step": 66855
     },
     {
-      "epoch": 2.35,
-      "learning_rate": 2.8606107103762688e-05,
-      "loss": 0.2876,
+      "epoch": 2.4096298698958445,
+      "grad_norm": 0.19200463593006134,
+      "learning_rate": 2.7686489028540748e-05,
+      "loss": 0.3739,
       "step": 66860
     },
     {
-      "epoch": 2.35,
-      "learning_rate": 2.860328815066e-05,
-      "loss": 0.2708,
+      "epoch": 2.4098100695570692,
+      "grad_norm": 0.22479012608528137,
+      "learning_rate": 2.76835877741756e-05,
+      "loss": 0.4098,
       "step": 66865
     },
     {
-      "epoch": 2.35,
-      "learning_rate": 2.860046915077099e-05,
-      "loss": 0.2864,
+      "epoch": 2.409990269218294,
+      "grad_norm": 0.23108981549739838,
+      "learning_rate": 2.7680686483247098e-05,
+      "loss": 0.3996,
       "step": 66870
     },
     {
-      "epoch": 2.35,
-      "learning_rate": 2.8597650104132257e-05,
-      "loss": 0.2729,
+      "epoch": 2.4101704688795187,
+      "grad_norm": 0.20057092607021332,
+      "learning_rate": 2.767778515579475e-05,
+      "loss": 0.4096,
       "step": 66875
     },
     {
-      "epoch": 2.35,
-      "learning_rate": 2.859483101078041e-05,
-      "loss": 0.2683,
+      "epoch": 2.410350668540743,
+      "grad_norm": 0.17036274075508118,
+      "learning_rate": 2.7674883791858107e-05,
+      "loss": 0.4209,
       "step": 66880
     },
     {
-      "epoch": 2.35,
-      "learning_rate": 2.8592011870752062e-05,
-      "loss": 0.2692,
+      "epoch": 2.4105308682019677,
+      "grad_norm": 0.19558821618556976,
+      "learning_rate": 2.7671982391476686e-05,
+      "loss": 0.4133,
       "step": 66885
     },
     {
-      "epoch": 2.35,
-      "learning_rate": 2.8589192684083794e-05,
-      "loss": 0.2591,
+      "epoch": 2.4107110678631924,
+      "grad_norm": 0.2363506406545639,
+      "learning_rate": 2.7669080954690023e-05,
+      "loss": 0.4149,
       "step": 66890
     },
     {
-      "epoch": 2.35,
-      "learning_rate": 2.8586373450812233e-05,
-      "loss": 0.2617,
+      "epoch": 2.410891267524417,
+      "grad_norm": 0.23225028812885284,
+      "learning_rate": 2.7666179481537646e-05,
+      "loss": 0.4357,
       "step": 66895
     },
     {
-      "epoch": 2.35,
-      "learning_rate": 2.858355417097397e-05,
-      "loss": 0.2512,
+      "epoch": 2.411071467185642,
+      "grad_norm": 0.20364603400230408,
+      "learning_rate": 2.76632779720591e-05,
+      "loss": 0.3863,
       "step": 66900
     },
     {
-      "epoch": 2.35,
-      "learning_rate": 2.8580734844605624e-05,
-      "loss": 0.2595,
+      "epoch": 2.411251666846866,
+      "grad_norm": 0.14624075591564178,
+      "learning_rate": 2.76603764262939e-05,
+      "loss": 0.3693,
       "step": 66905
     },
     {
-      "epoch": 2.35,
-      "learning_rate": 2.857791547174379e-05,
-      "loss": 0.2547,
+      "epoch": 2.411431866508091,
+      "grad_norm": 0.20455005764961243,
+      "learning_rate": 2.7657474844281577e-05,
+      "loss": 0.3922,
       "step": 66910
     },
     {
-      "epoch": 2.35,
-      "learning_rate": 2.857509605242508e-05,
-      "loss": 0.2712,
+      "epoch": 2.4116120661693157,
+      "grad_norm": 0.19567883014678955,
+      "learning_rate": 2.7654573226061686e-05,
+      "loss": 0.4302,
       "step": 66915
     },
     {
-      "epoch": 2.35,
-      "learning_rate": 2.8572276586686107e-05,
-      "loss": 0.2826,
+      "epoch": 2.4117922658305404,
+      "grad_norm": 0.2601669430732727,
+      "learning_rate": 2.7651671571673743e-05,
+      "loss": 0.3964,
       "step": 66920
     },
     {
-      "epoch": 2.35,
-      "learning_rate": 2.8569457074563477e-05,
-      "loss": 0.278,
+      "epoch": 2.4119724654917647,
+      "grad_norm": 0.22823162376880646,
+      "learning_rate": 2.76487698811573e-05,
+      "loss": 0.3906,
       "step": 66925
     },
     {
-      "epoch": 2.35,
-      "learning_rate": 2.8566637516093807e-05,
-      "loss": 0.266,
+      "epoch": 2.4121526651529894,
+      "grad_norm": 0.2383589893579483,
+      "learning_rate": 2.764586815455187e-05,
+      "loss": 0.4254,
       "step": 66930
     },
     {
-      "epoch": 2.35,
-      "learning_rate": 2.8563817911313684e-05,
-      "loss": 0.2973,
+      "epoch": 2.412332864814214,
+      "grad_norm": 0.18745693564414978,
+      "learning_rate": 2.764296639189699e-05,
+      "loss": 0.3825,
       "step": 66935
     },
     {
-      "epoch": 2.36,
-      "learning_rate": 2.8560998260259748e-05,
-      "loss": 0.2799,
+      "epoch": 2.412513064475439,
+      "grad_norm": 0.2072102427482605,
+      "learning_rate": 2.7640064593232218e-05,
+      "loss": 0.3881,
       "step": 66940
     },
     {
-      "epoch": 2.36,
-      "learning_rate": 2.8558178562968583e-05,
-      "loss": 0.281,
+      "epoch": 2.4126932641366636,
+      "grad_norm": 0.1848241686820984,
+      "learning_rate": 2.7637162758597073e-05,
+      "loss": 0.3747,
       "step": 66945
     },
     {
-      "epoch": 2.36,
-      "learning_rate": 2.855535881947683e-05,
-      "loss": 0.2711,
+      "epoch": 2.412873463797888,
+      "grad_norm": 0.1639891117811203,
+      "learning_rate": 2.7634260888031098e-05,
+      "loss": 0.3715,
       "step": 66950
     },
     {
-      "epoch": 2.36,
-      "learning_rate": 2.8552539029821074e-05,
-      "loss": 0.274,
+      "epoch": 2.4130536634591127,
+      "grad_norm": 0.20181863009929657,
+      "learning_rate": 2.7631358981573824e-05,
+      "loss": 0.4048,
       "step": 66955
     },
     {
-      "epoch": 2.36,
-      "learning_rate": 2.8549719194037942e-05,
-      "loss": 0.2637,
+      "epoch": 2.4132338631203374,
+      "grad_norm": 0.1724659949541092,
+      "learning_rate": 2.76284570392648e-05,
+      "loss": 0.4168,
       "step": 66960
     },
     {
-      "epoch": 2.36,
-      "learning_rate": 2.8546899312164045e-05,
-      "loss": 0.2716,
+      "epoch": 2.413414062781562,
+      "grad_norm": 0.23063142597675323,
+      "learning_rate": 2.7625555061143553e-05,
+      "loss": 0.4061,
       "step": 66965
     },
     {
-      "epoch": 2.36,
-      "learning_rate": 2.8544079384236004e-05,
-      "loss": 0.2458,
+      "epoch": 2.4135942624427864,
+      "grad_norm": 0.23275382816791534,
+      "learning_rate": 2.7622653047249626e-05,
+      "loss": 0.4226,
       "step": 66970
     },
     {
-      "epoch": 2.36,
-      "learning_rate": 2.8541259410290422e-05,
-      "loss": 0.2872,
+      "epoch": 2.413774462104011,
+      "grad_norm": 0.17771199345588684,
+      "learning_rate": 2.7619750997622564e-05,
+      "loss": 0.4079,
       "step": 66975
     },
     {
-      "epoch": 2.36,
-      "learning_rate": 2.8538439390363924e-05,
-      "loss": 0.2763,
+      "epoch": 2.413954661765236,
+      "grad_norm": 0.20910033583641052,
+      "learning_rate": 2.7616848912301892e-05,
+      "loss": 0.4074,
       "step": 66980
     },
     {
-      "epoch": 2.36,
-      "learning_rate": 2.8535619324493125e-05,
-      "loss": 0.2574,
+      "epoch": 2.4141348614264606,
+      "grad_norm": 0.2285621017217636,
+      "learning_rate": 2.7613946791327167e-05,
+      "loss": 0.4133,
       "step": 66985
     },
     {
-      "epoch": 2.36,
-      "learning_rate": 2.8532799212714628e-05,
-      "loss": 0.2851,
+      "epoch": 2.4143150610876853,
+      "grad_norm": 0.22376610338687897,
+      "learning_rate": 2.7611044634737926e-05,
+      "loss": 0.3761,
       "step": 66990
     },
     {
-      "epoch": 2.36,
-      "learning_rate": 2.8529979055065072e-05,
-      "loss": 0.2583,
+      "epoch": 2.4144952607489096,
+      "grad_norm": 0.20609983801841736,
+      "learning_rate": 2.7608142442573704e-05,
+      "loss": 0.376,
       "step": 66995
     },
     {
-      "epoch": 2.36,
-      "learning_rate": 2.852715885158106e-05,
-      "loss": 0.2546,
+      "epoch": 2.4146754604101344,
+      "grad_norm": 0.20516866445541382,
+      "learning_rate": 2.7605240214874038e-05,
+      "loss": 0.3812,
       "step": 67000
     },
     {
-      "epoch": 2.36,
-      "eval_loss": 0.26735541224479675,
-      "eval_runtime": 10.5619,
-      "eval_samples_per_second": 9.468,
-      "eval_steps_per_second": 9.468,
+      "epoch": 2.4146754604101344,
+      "eval_loss": 0.4341346025466919,
+      "eval_runtime": 3.524,
+      "eval_samples_per_second": 28.377,
+      "eval_steps_per_second": 7.094,
       "step": 67000
     },
     {
-      "epoch": 2.36,
-      "learning_rate": 2.852433860229921e-05,
-      "loss": 0.2645,
+      "epoch": 2.414855660071359,
+      "grad_norm": 0.19318906962871552,
+      "learning_rate": 2.760233795167849e-05,
+      "loss": 0.376,
       "step": 67005
     },
     {
-      "epoch": 2.36,
-      "learning_rate": 2.852151830725615e-05,
-      "loss": 0.2531,
+      "epoch": 2.415035859732584,
+      "grad_norm": 0.19978414475917816,
+      "learning_rate": 2.759943565302659e-05,
+      "loss": 0.3926,
       "step": 67010
     },
     {
-      "epoch": 2.36,
-      "learning_rate": 2.8518697966488493e-05,
-      "loss": 0.2846,
+      "epoch": 2.415216059393808,
+      "grad_norm": 0.19029457867145538,
+      "learning_rate": 2.759653331895788e-05,
+      "loss": 0.3724,
       "step": 67015
     },
     {
-      "epoch": 2.36,
-      "learning_rate": 2.8515877580032868e-05,
-      "loss": 0.2697,
+      "epoch": 2.415396259055033,
+      "grad_norm": 0.1965416818857193,
+      "learning_rate": 2.7593630949511908e-05,
+      "loss": 0.4342,
       "step": 67020
     },
     {
-      "epoch": 2.36,
-      "learning_rate": 2.8513057147925894e-05,
-      "loss": 0.2515,
+      "epoch": 2.4155764587162576,
+      "grad_norm": 0.19365379214286804,
+      "learning_rate": 2.7590728544728213e-05,
+      "loss": 0.3787,
       "step": 67025
     },
     {
-      "epoch": 2.36,
-      "learning_rate": 2.8510236670204176e-05,
-      "loss": 0.2567,
+      "epoch": 2.4157566583774823,
+      "grad_norm": 0.24059845507144928,
+      "learning_rate": 2.7587826104646348e-05,
+      "loss": 0.3632,
       "step": 67030
     },
     {
-      "epoch": 2.36,
-      "learning_rate": 2.8507416146904358e-05,
-      "loss": 0.2844,
+      "epoch": 2.415936858038707,
+      "grad_norm": 0.2164888083934784,
+      "learning_rate": 2.758492362930585e-05,
+      "loss": 0.4078,
       "step": 67035
     },
     {
-      "epoch": 2.36,
-      "learning_rate": 2.8504595578063047e-05,
-      "loss": 0.2846,
+      "epoch": 2.4161170576999313,
+      "grad_norm": 0.21107304096221924,
+      "learning_rate": 2.7582021118746266e-05,
+      "loss": 0.3813,
       "step": 67040
     },
     {
-      "epoch": 2.36,
-      "learning_rate": 2.850177496371687e-05,
-      "loss": 0.2816,
+      "epoch": 2.416297257361156,
+      "grad_norm": 0.23840226233005524,
+      "learning_rate": 2.7579118573007144e-05,
+      "loss": 0.3851,
       "step": 67045
     },
     {
-      "epoch": 2.36,
-      "learning_rate": 2.8498954303902452e-05,
-      "loss": 0.2844,
+      "epoch": 2.416477457022381,
+      "grad_norm": 0.23817472159862518,
+      "learning_rate": 2.7576215992128034e-05,
+      "loss": 0.3961,
       "step": 67050
     },
     {
-      "epoch": 2.36,
-      "learning_rate": 2.8496133598656426e-05,
-      "loss": 0.2526,
+      "epoch": 2.4166576566836055,
+      "grad_norm": 0.23038597404956818,
+      "learning_rate": 2.7573313376148478e-05,
+      "loss": 0.4176,
       "step": 67055
     },
     {
-      "epoch": 2.36,
-      "learning_rate": 2.8493312848015407e-05,
-      "loss": 0.2688,
+      "epoch": 2.41683785634483,
+      "grad_norm": 0.18033789098262787,
+      "learning_rate": 2.7570410725108027e-05,
+      "loss": 0.4053,
       "step": 67060
     },
     {
-      "epoch": 2.36,
-      "learning_rate": 2.8490492052016028e-05,
-      "loss": 0.2706,
+      "epoch": 2.4170180560060546,
+      "grad_norm": 0.24066656827926636,
+      "learning_rate": 2.7567508039046232e-05,
+      "loss": 0.4378,
       "step": 67065
     },
     {
-      "epoch": 2.36,
-      "learning_rate": 2.84876712106949e-05,
-      "loss": 0.2574,
+      "epoch": 2.4171982556672793,
+      "grad_norm": 0.2486283779144287,
+      "learning_rate": 2.7564605318002627e-05,
+      "loss": 0.4038,
       "step": 67070
     },
     {
-      "epoch": 2.36,
-      "learning_rate": 2.8484850324088662e-05,
-      "loss": 0.2623,
+      "epoch": 2.417378455328504,
+      "grad_norm": 0.2280510812997818,
+      "learning_rate": 2.756170256201678e-05,
+      "loss": 0.3974,
       "step": 67075
     },
     {
-      "epoch": 2.36,
-      "learning_rate": 2.848202939223394e-05,
-      "loss": 0.2618,
+      "epoch": 2.4175586549897288,
+      "grad_norm": 0.19952905178070068,
+      "learning_rate": 2.7558799771128223e-05,
+      "loss": 0.408,
       "step": 67080
     },
     {
-      "epoch": 2.36,
-      "learning_rate": 2.8479208415167363e-05,
-      "loss": 0.2743,
+      "epoch": 2.417738854650953,
+      "grad_norm": 0.19479811191558838,
+      "learning_rate": 2.755589694537652e-05,
+      "loss": 0.3922,
       "step": 67085
     },
     {
-      "epoch": 2.36,
-      "learning_rate": 2.8476387392925552e-05,
-      "loss": 0.2593,
+      "epoch": 2.417919054312178,
+      "grad_norm": 0.2107398509979248,
+      "learning_rate": 2.755299408480122e-05,
+      "loss": 0.4211,
       "step": 67090
     },
     {
-      "epoch": 2.36,
-      "learning_rate": 2.8473566325545148e-05,
-      "loss": 0.2704,
+      "epoch": 2.4180992539734025,
+      "grad_norm": 0.21813732385635376,
+      "learning_rate": 2.755009118944186e-05,
+      "loss": 0.3841,
       "step": 67095
     },
     {
-      "epoch": 2.36,
-      "learning_rate": 2.8470745213062767e-05,
-      "loss": 0.2534,
+      "epoch": 2.4182794536346273,
+      "grad_norm": 0.18200333416461945,
+      "learning_rate": 2.754718825933801e-05,
+      "loss": 0.4034,
       "step": 67100
     },
     {
-      "epoch": 2.36,
-      "learning_rate": 2.846792405551506e-05,
-      "loss": 0.2744,
+      "epoch": 2.418459653295852,
+      "grad_norm": 0.18498298525810242,
+      "learning_rate": 2.7544285294529204e-05,
+      "loss": 0.3802,
       "step": 67105
     },
     {
-      "epoch": 2.36,
-      "learning_rate": 2.846510285293863e-05,
-      "loss": 0.2651,
+      "epoch": 2.4186398529570763,
+      "grad_norm": 0.21120643615722656,
+      "learning_rate": 2.7541382295055014e-05,
+      "loss": 0.4255,
       "step": 67110
     },
     {
-      "epoch": 2.36,
-      "learning_rate": 2.8462281605370133e-05,
-      "loss": 0.2744,
+      "epoch": 2.418820052618301,
+      "grad_norm": 0.2860146164894104,
+      "learning_rate": 2.753847926095498e-05,
+      "loss": 0.402,
       "step": 67115
     },
     {
-      "epoch": 2.36,
-      "learning_rate": 2.8459460312846186e-05,
-      "loss": 0.2822,
+      "epoch": 2.4190002522795258,
+      "grad_norm": 0.21103060245513916,
+      "learning_rate": 2.7535576192268648e-05,
+      "loss": 0.3978,
       "step": 67120
     },
     {
-      "epoch": 2.36,
-      "learning_rate": 2.8456638975403428e-05,
-      "loss": 0.2749,
+      "epoch": 2.4191804519407505,
+      "grad_norm": 0.1569247990846634,
+      "learning_rate": 2.7532673089035588e-05,
+      "loss": 0.3973,
       "step": 67125
     },
     {
-      "epoch": 2.36,
-      "learning_rate": 2.8453817593078492e-05,
-      "loss": 0.2597,
+      "epoch": 2.419360651601975,
+      "grad_norm": 0.23274268209934235,
+      "learning_rate": 2.7529769951295355e-05,
+      "loss": 0.3947,
       "step": 67130
     },
     {
-      "epoch": 2.36,
-      "learning_rate": 2.8450996165908006e-05,
-      "loss": 0.2588,
+      "epoch": 2.4195408512631995,
+      "grad_norm": 0.1998492181301117,
+      "learning_rate": 2.752686677908749e-05,
+      "loss": 0.4082,
       "step": 67135
     },
     {
-      "epoch": 2.36,
-      "learning_rate": 2.844817469392862e-05,
-      "loss": 0.276,
+      "epoch": 2.4197210509244242,
+      "grad_norm": 0.19233012199401855,
+      "learning_rate": 2.7523963572451545e-05,
+      "loss": 0.4093,
       "step": 67140
     },
     {
-      "epoch": 2.36,
-      "learning_rate": 2.8445353177176946e-05,
-      "loss": 0.2702,
+      "epoch": 2.419901250585649,
+      "grad_norm": 0.22969095408916473,
+      "learning_rate": 2.7521060331427096e-05,
+      "loss": 0.4131,
       "step": 67145
     },
     {
-      "epoch": 2.36,
-      "learning_rate": 2.8442531615689644e-05,
-      "loss": 0.2784,
+      "epoch": 2.4200814502468737,
+      "grad_norm": 0.2119283676147461,
+      "learning_rate": 2.751815705605369e-05,
+      "loss": 0.4162,
       "step": 67150
     },
     {
-      "epoch": 2.36,
-      "learning_rate": 2.8439710009503322e-05,
-      "loss": 0.2722,
+      "epoch": 2.420261649908098,
+      "grad_norm": 0.18394353985786438,
+      "learning_rate": 2.7515253746370873e-05,
+      "loss": 0.3837,
       "step": 67155
     },
     {
-      "epoch": 2.36,
-      "learning_rate": 2.8436888358654644e-05,
-      "loss": 0.2814,
+      "epoch": 2.4204418495693227,
+      "grad_norm": 0.19230255484580994,
+      "learning_rate": 2.7512350402418212e-05,
+      "loss": 0.3603,
       "step": 67160
     },
     {
-      "epoch": 2.36,
-      "learning_rate": 2.843406666318023e-05,
-      "loss": 0.2579,
+      "epoch": 2.4206220492305475,
+      "grad_norm": 0.2101142406463623,
+      "learning_rate": 2.7509447024235262e-05,
+      "loss": 0.3784,
       "step": 67165
     },
     {
-      "epoch": 2.36,
-      "learning_rate": 2.843124492311673e-05,
-      "loss": 0.2651,
+      "epoch": 2.420802248891772,
+      "grad_norm": 0.1899431347846985,
+      "learning_rate": 2.7506543611861598e-05,
+      "loss": 0.3301,
       "step": 67170
     },
     {
-      "epoch": 2.36,
-      "learning_rate": 2.842842313850077e-05,
-      "loss": 0.2666,
+      "epoch": 2.420982448552997,
+      "grad_norm": 0.2313234508037567,
+      "learning_rate": 2.750364016533674e-05,
+      "loss": 0.3929,
       "step": 67175
     },
     {
-      "epoch": 2.36,
-      "learning_rate": 2.8425601309369e-05,
-      "loss": 0.2696,
+      "epoch": 2.4211626482142212,
+      "grad_norm": 0.20797547698020935,
+      "learning_rate": 2.7500736684700283e-05,
+      "loss": 0.3923,
       "step": 67180
     },
     {
-      "epoch": 2.36,
-      "learning_rate": 2.8422779435758045e-05,
-      "loss": 0.2789,
+      "epoch": 2.421342847875446,
+      "grad_norm": 0.2217523455619812,
+      "learning_rate": 2.7497833169991767e-05,
+      "loss": 0.4297,
       "step": 67185
     },
     {
-      "epoch": 2.36,
-      "learning_rate": 2.8419957517704567e-05,
-      "loss": 0.2796,
+      "epoch": 2.4215230475366707,
+      "grad_norm": 0.19359783828258514,
+      "learning_rate": 2.7494929621250765e-05,
+      "loss": 0.3811,
       "step": 67190
     },
     {
-      "epoch": 2.36,
-      "learning_rate": 2.8417135555245183e-05,
-      "loss": 0.2679,
+      "epoch": 2.4217032471978954,
+      "grad_norm": 0.21590293943881989,
+      "learning_rate": 2.7492026038516826e-05,
+      "loss": 0.3841,
       "step": 67195
     },
     {
-      "epoch": 2.36,
-      "learning_rate": 2.841431354841656e-05,
-      "loss": 0.2773,
+      "epoch": 2.4218834468591197,
+      "grad_norm": 0.17126692831516266,
+      "learning_rate": 2.748912242182951e-05,
+      "loss": 0.3761,
       "step": 67200
     },
     {
-      "epoch": 2.36,
-      "learning_rate": 2.8411491497255316e-05,
-      "loss": 0.2451,
+      "epoch": 2.4220636465203444,
+      "grad_norm": 0.24167068302631378,
+      "learning_rate": 2.7486218771228388e-05,
+      "loss": 0.3784,
       "step": 67205
     },
     {
-      "epoch": 2.36,
-      "learning_rate": 2.8408669401798098e-05,
-      "loss": 0.2846,
+      "epoch": 2.422243846181569,
+      "grad_norm": 0.21758897602558136,
+      "learning_rate": 2.7483315086753015e-05,
+      "loss": 0.4132,
       "step": 67210
     },
     {
-      "epoch": 2.36,
-      "learning_rate": 2.840584726208156e-05,
-      "loss": 0.2712,
+      "epoch": 2.422424045842794,
+      "grad_norm": 0.20795123279094696,
+      "learning_rate": 2.7480411368442956e-05,
+      "loss": 0.3909,
       "step": 67215
     },
     {
-      "epoch": 2.36,
-      "learning_rate": 2.840302507814234e-05,
-      "loss": 0.2558,
+      "epoch": 2.4226042455040186,
+      "grad_norm": 0.1906597912311554,
+      "learning_rate": 2.7477507616337773e-05,
+      "loss": 0.3742,
       "step": 67220
     },
     {
-      "epoch": 2.37,
-      "learning_rate": 2.8400202850017087e-05,
-      "loss": 0.2631,
+      "epoch": 2.422784445165243,
+      "grad_norm": 0.22008202970027924,
+      "learning_rate": 2.747460383047702e-05,
+      "loss": 0.4114,
       "step": 67225
     },
     {
-      "epoch": 2.37,
-      "learning_rate": 2.839738057774244e-05,
-      "loss": 0.2752,
+      "epoch": 2.4229646448264677,
+      "grad_norm": 0.2209016978740692,
+      "learning_rate": 2.747170001090028e-05,
+      "loss": 0.3567,
       "step": 67230
     },
     {
-      "epoch": 2.37,
-      "learning_rate": 2.839455826135503e-05,
-      "loss": 0.2842,
+      "epoch": 2.4231448444876924,
+      "grad_norm": 0.2042957842350006,
+      "learning_rate": 2.74687961576471e-05,
+      "loss": 0.3943,
       "step": 67235
     },
     {
-      "epoch": 2.37,
-      "learning_rate": 2.839173590089153e-05,
-      "loss": 0.3026,
+      "epoch": 2.423325044148917,
+      "grad_norm": 0.2651365399360657,
+      "learning_rate": 2.7465892270757055e-05,
+      "loss": 0.4139,
       "step": 67240
     },
     {
-      "epoch": 2.37,
-      "learning_rate": 2.838891349638858e-05,
-      "loss": 0.2557,
+      "epoch": 2.4235052438101414,
+      "grad_norm": 0.2116497904062271,
+      "learning_rate": 2.7462988350269698e-05,
+      "loss": 0.4203,
       "step": 67245
     },
     {
-      "epoch": 2.37,
-      "learning_rate": 2.838609104788281e-05,
-      "loss": 0.2689,
+      "epoch": 2.423685443471366,
+      "grad_norm": 0.24418489634990692,
+      "learning_rate": 2.7460084396224606e-05,
+      "loss": 0.3874,
       "step": 67250
     },
     {
-      "epoch": 2.37,
-      "learning_rate": 2.838326855541089e-05,
-      "loss": 0.2744,
+      "epoch": 2.423865643132591,
+      "grad_norm": 0.18240700662136078,
+      "learning_rate": 2.7457180408661343e-05,
+      "loss": 0.3952,
       "step": 67255
     },
     {
-      "epoch": 2.37,
-      "learning_rate": 2.8380446019009448e-05,
-      "loss": 0.2816,
+      "epoch": 2.4240458427938156,
+      "grad_norm": 0.20729464292526245,
+      "learning_rate": 2.7454276387619465e-05,
+      "loss": 0.378,
       "step": 67260
     },
     {
-      "epoch": 2.37,
-      "learning_rate": 2.8377623438715152e-05,
-      "loss": 0.2801,
+      "epoch": 2.4242260424550404,
+      "grad_norm": 0.22618576884269714,
+      "learning_rate": 2.745137233313856e-05,
+      "loss": 0.408,
       "step": 67265
     },
     {
-      "epoch": 2.37,
-      "learning_rate": 2.8374800814564627e-05,
-      "loss": 0.289,
+      "epoch": 2.4244062421162647,
+      "grad_norm": 0.1970742791891098,
+      "learning_rate": 2.7448468245258162e-05,
+      "loss": 0.4034,
       "step": 67270
     },
     {
-      "epoch": 2.37,
-      "learning_rate": 2.837197814659455e-05,
-      "loss": 0.276,
+      "epoch": 2.4245864417774894,
+      "grad_norm": 0.18698358535766602,
+      "learning_rate": 2.744556412401788e-05,
+      "loss": 0.3862,
       "step": 67275
     },
     {
-      "epoch": 2.37,
-      "learning_rate": 2.8369155434841554e-05,
-      "loss": 0.2614,
+      "epoch": 2.424766641438714,
+      "grad_norm": 0.18384528160095215,
+      "learning_rate": 2.7442659969457247e-05,
+      "loss": 0.3994,
       "step": 67280
     },
     {
-      "epoch": 2.37,
-      "learning_rate": 2.8366332679342305e-05,
-      "loss": 0.2795,
+      "epoch": 2.424946841099939,
+      "grad_norm": 0.199675053358078,
+      "learning_rate": 2.7439755781615845e-05,
+      "loss": 0.3781,
       "step": 67285
     },
     {
-      "epoch": 2.37,
-      "learning_rate": 2.836350988013343e-05,
-      "loss": 0.2556,
+      "epoch": 2.425127040761163,
+      "grad_norm": 0.2059258222579956,
+      "learning_rate": 2.7436851560533246e-05,
+      "loss": 0.4037,
       "step": 67290
     },
     {
-      "epoch": 2.37,
-      "learning_rate": 2.8360687037251603e-05,
-      "loss": 0.2745,
+      "epoch": 2.425307240422388,
+      "grad_norm": 0.21822243928909302,
+      "learning_rate": 2.7433947306249014e-05,
+      "loss": 0.3944,
       "step": 67295
     },
     {
-      "epoch": 2.37,
-      "learning_rate": 2.835786415073347e-05,
-      "loss": 0.2803,
+      "epoch": 2.4254874400836126,
+      "grad_norm": 0.19655758142471313,
+      "learning_rate": 2.7431043018802725e-05,
+      "loss": 0.3946,
       "step": 67300
     },
     {
-      "epoch": 2.37,
-      "learning_rate": 2.8355041220615685e-05,
-      "loss": 0.2531,
+      "epoch": 2.4256676397448373,
+      "grad_norm": 0.2280445247888565,
+      "learning_rate": 2.742813869823394e-05,
+      "loss": 0.3906,
       "step": 67305
     },
     {
-      "epoch": 2.37,
-      "learning_rate": 2.8352218246934897e-05,
-      "loss": 0.2607,
+      "epoch": 2.425847839406062,
+      "grad_norm": 0.16742421686649323,
+      "learning_rate": 2.742523434458224e-05,
+      "loss": 0.3632,
       "step": 67310
     },
     {
-      "epoch": 2.37,
-      "learning_rate": 2.8349395229727766e-05,
-      "loss": 0.2564,
+      "epoch": 2.4260280390672864,
+      "grad_norm": 0.19926291704177856,
+      "learning_rate": 2.7422329957887188e-05,
+      "loss": 0.4446,
       "step": 67315
     },
     {
-      "epoch": 2.37,
-      "learning_rate": 2.8346572169030938e-05,
-      "loss": 0.2854,
+      "epoch": 2.426208238728511,
+      "grad_norm": 0.17569011449813843,
+      "learning_rate": 2.7419425538188365e-05,
+      "loss": 0.3709,
       "step": 67320
     },
     {
-      "epoch": 2.37,
-      "learning_rate": 2.8344313689185543e-05,
-      "loss": 0.2996,
+      "epoch": 2.426388438389736,
+      "grad_norm": 0.26853156089782715,
+      "learning_rate": 2.741652108552533e-05,
+      "loss": 0.4286,
       "step": 67325
     },
     {
-      "epoch": 2.37,
-      "learning_rate": 2.834149055029965e-05,
-      "loss": 0.2884,
+      "epoch": 2.4265686380509606,
+      "grad_norm": 0.20490770041942596,
+      "learning_rate": 2.741361659993767e-05,
+      "loss": 0.3812,
       "step": 67330
     },
     {
-      "epoch": 2.37,
-      "learning_rate": 2.833866736802671e-05,
-      "loss": 0.2499,
+      "epoch": 2.426748837712185,
+      "grad_norm": 0.24885760247707367,
+      "learning_rate": 2.7410712081464947e-05,
+      "loss": 0.4166,
       "step": 67335
     },
     {
-      "epoch": 2.37,
-      "learning_rate": 2.8335844142403366e-05,
-      "loss": 0.245,
+      "epoch": 2.4269290373734096,
+      "grad_norm": 0.24645163118839264,
+      "learning_rate": 2.7407807530146735e-05,
+      "loss": 0.3814,
       "step": 67340
     },
     {
-      "epoch": 2.37,
-      "learning_rate": 2.8333020873466297e-05,
-      "loss": 0.2702,
+      "epoch": 2.4271092370346343,
+      "grad_norm": 0.24521587789058685,
+      "learning_rate": 2.740490294602262e-05,
+      "loss": 0.3759,
       "step": 67345
     },
     {
-      "epoch": 2.37,
-      "learning_rate": 2.8330197561252142e-05,
-      "loss": 0.2931,
+      "epoch": 2.427289436695859,
+      "grad_norm": 0.20544666051864624,
+      "learning_rate": 2.7401998329132162e-05,
+      "loss": 0.404,
       "step": 67350
     },
     {
-      "epoch": 2.37,
-      "learning_rate": 2.8327374205797562e-05,
-      "loss": 0.2842,
+      "epoch": 2.427469636357084,
+      "grad_norm": 0.17196045815944672,
+      "learning_rate": 2.7399093679514947e-05,
+      "loss": 0.3892,
       "step": 67355
     },
     {
-      "epoch": 2.37,
-      "learning_rate": 2.8324550807139223e-05,
-      "loss": 0.2538,
+      "epoch": 2.4276498360183085,
+      "grad_norm": 0.18687868118286133,
+      "learning_rate": 2.7396188997210538e-05,
+      "loss": 0.3991,
       "step": 67360
     },
     {
-      "epoch": 2.37,
-      "learning_rate": 2.8321727365313788e-05,
-      "loss": 0.2573,
+      "epoch": 2.427830035679533,
+      "grad_norm": 0.23785381019115448,
+      "learning_rate": 2.7393284282258517e-05,
+      "loss": 0.4111,
       "step": 67365
     },
     {
-      "epoch": 2.37,
-      "learning_rate": 2.8318903880357906e-05,
-      "loss": 0.266,
+      "epoch": 2.4280102353407575,
+      "grad_norm": 0.23783884942531586,
+      "learning_rate": 2.739037953469847e-05,
+      "loss": 0.3777,
       "step": 67370
     },
     {
-      "epoch": 2.37,
-      "learning_rate": 2.831608035230825e-05,
-      "loss": 0.2564,
+      "epoch": 2.4281904350019823,
+      "grad_norm": 0.16781599819660187,
+      "learning_rate": 2.7387474754569964e-05,
+      "loss": 0.3757,
       "step": 67375
     },
     {
-      "epoch": 2.37,
-      "learning_rate": 2.831325678120147e-05,
-      "loss": 0.2514,
+      "epoch": 2.428370634663207,
+      "grad_norm": 0.15680092573165894,
+      "learning_rate": 2.7384569941912574e-05,
+      "loss": 0.4104,
       "step": 67380
     },
     {
-      "epoch": 2.37,
-      "learning_rate": 2.831043316707424e-05,
-      "loss": 0.2565,
+      "epoch": 2.4285508343244313,
+      "grad_norm": 0.23837818205356598,
+      "learning_rate": 2.7381665096765874e-05,
+      "loss": 0.3977,
       "step": 67385
     },
     {
-      "epoch": 2.37,
-      "learning_rate": 2.830760950996321e-05,
-      "loss": 0.2854,
+      "epoch": 2.428731033985656,
+      "grad_norm": 0.2140997052192688,
+      "learning_rate": 2.7378760219169453e-05,
+      "loss": 0.4186,
       "step": 67390
     },
     {
-      "epoch": 2.37,
-      "learning_rate": 2.830478580990506e-05,
-      "loss": 0.2671,
+      "epoch": 2.4289112336468808,
+      "grad_norm": 0.25618883967399597,
+      "learning_rate": 2.7375855309162886e-05,
+      "loss": 0.3869,
       "step": 67395
     },
     {
-      "epoch": 2.37,
-      "learning_rate": 2.8301962066936437e-05,
-      "loss": 0.2981,
+      "epoch": 2.4290914333081055,
+      "grad_norm": 0.26297783851623535,
+      "learning_rate": 2.7372950366785748e-05,
+      "loss": 0.411,
       "step": 67400
     },
     {
-      "epoch": 2.37,
-      "learning_rate": 2.829913828109401e-05,
-      "loss": 0.2826,
+      "epoch": 2.4292716329693302,
+      "grad_norm": 0.22148928046226501,
+      "learning_rate": 2.7370045392077626e-05,
+      "loss": 0.3997,
       "step": 67405
     },
     {
-      "epoch": 2.37,
-      "learning_rate": 2.8296314452414453e-05,
-      "loss": 0.2819,
+      "epoch": 2.4294518326305545,
+      "grad_norm": 0.21536856889724731,
+      "learning_rate": 2.736714038507808e-05,
+      "loss": 0.4008,
       "step": 67410
     },
     {
-      "epoch": 2.37,
-      "learning_rate": 2.8293490580934422e-05,
-      "loss": 0.2803,
+      "epoch": 2.4296320322917793,
+      "grad_norm": 0.18646790087223053,
+      "learning_rate": 2.736423534582672e-05,
+      "loss": 0.3861,
       "step": 67415
     },
     {
-      "epoch": 2.37,
-      "learning_rate": 2.829066666669059e-05,
-      "loss": 0.2906,
+      "epoch": 2.429812231953004,
+      "grad_norm": 0.17240692675113678,
+      "learning_rate": 2.7361330274363113e-05,
+      "loss": 0.3718,
       "step": 67420
     },
     {
-      "epoch": 2.37,
-      "learning_rate": 2.8287842709719618e-05,
-      "loss": 0.27,
+      "epoch": 2.4299924316142287,
+      "grad_norm": 0.17794083058834076,
+      "learning_rate": 2.735842517072683e-05,
+      "loss": 0.4088,
       "step": 67425
     },
     {
-      "epoch": 2.37,
-      "learning_rate": 2.8285018710058176e-05,
-      "loss": 0.2333,
+      "epoch": 2.430172631275453,
+      "grad_norm": 0.17455384135246277,
+      "learning_rate": 2.735552003495746e-05,
+      "loss": 0.4008,
       "step": 67430
     },
     {
-      "epoch": 2.37,
-      "learning_rate": 2.828219466774293e-05,
-      "loss": 0.2658,
+      "epoch": 2.4303528309366778,
+      "grad_norm": 0.22316579520702362,
+      "learning_rate": 2.735261486709459e-05,
+      "loss": 0.4001,
       "step": 67435
     },
     {
-      "epoch": 2.37,
-      "learning_rate": 2.8279370582810544e-05,
-      "loss": 0.2773,
+      "epoch": 2.4305330305979025,
+      "grad_norm": 0.18151316046714783,
+      "learning_rate": 2.734970966717779e-05,
+      "loss": 0.4002,
       "step": 67440
     },
     {
-      "epoch": 2.37,
-      "learning_rate": 2.8276546455297694e-05,
-      "loss": 0.2568,
+      "epoch": 2.4307132302591272,
+      "grad_norm": 0.27039554715156555,
+      "learning_rate": 2.7346804435246658e-05,
+      "loss": 0.4181,
       "step": 67445
     },
     {
-      "epoch": 2.37,
-      "learning_rate": 2.8273722285241056e-05,
-      "loss": 0.2769,
+      "epoch": 2.430893429920352,
+      "grad_norm": 0.24719054996967316,
+      "learning_rate": 2.734389917134077e-05,
+      "loss": 0.417,
       "step": 67450
     },
     {
-      "epoch": 2.37,
-      "learning_rate": 2.8270898072677282e-05,
-      "loss": 0.2584,
+      "epoch": 2.4310736295815762,
+      "grad_norm": 0.2097766101360321,
+      "learning_rate": 2.7340993875499704e-05,
+      "loss": 0.387,
       "step": 67455
     },
     {
-      "epoch": 2.37,
-      "learning_rate": 2.8268073817643053e-05,
-      "loss": 0.2837,
+      "epoch": 2.431253829242801,
+      "grad_norm": 0.23086480796337128,
+      "learning_rate": 2.7338088547763057e-05,
+      "loss": 0.4455,
       "step": 67460
     },
     {
-      "epoch": 2.37,
-      "learning_rate": 2.826524952017504e-05,
-      "loss": 0.2787,
+      "epoch": 2.4314340289040257,
+      "grad_norm": 0.21231964230537415,
+      "learning_rate": 2.7335183188170398e-05,
+      "loss": 0.3746,
       "step": 67465
     },
     {
-      "epoch": 2.37,
-      "learning_rate": 2.8262425180309927e-05,
-      "loss": 0.2792,
+      "epoch": 2.4316142285652504,
+      "grad_norm": 0.1851814240217209,
+      "learning_rate": 2.733227779676133e-05,
+      "loss": 0.4,
       "step": 67470
     },
     {
-      "epoch": 2.37,
-      "learning_rate": 2.8259600798084356e-05,
-      "loss": 0.2781,
+      "epoch": 2.4317944282264747,
+      "grad_norm": 0.1971953958272934,
+      "learning_rate": 2.7329372373575418e-05,
+      "loss": 0.398,
       "step": 67475
     },
     {
-      "epoch": 2.37,
-      "learning_rate": 2.8256776373535026e-05,
-      "loss": 0.2648,
+      "epoch": 2.4319746278876995,
+      "grad_norm": 0.24778111279010773,
+      "learning_rate": 2.7326466918652267e-05,
+      "loss": 0.4247,
       "step": 67480
     },
     {
-      "epoch": 2.37,
-      "learning_rate": 2.82539519066986e-05,
-      "loss": 0.2737,
+      "epoch": 2.432154827548924,
+      "grad_norm": 0.20682400465011597,
+      "learning_rate": 2.7323561432031452e-05,
+      "loss": 0.383,
       "step": 67485
     },
     {
-      "epoch": 2.37,
-      "learning_rate": 2.8251127397611748e-05,
-      "loss": 0.256,
+      "epoch": 2.432335027210149,
+      "grad_norm": 0.20953048765659332,
+      "learning_rate": 2.7320655913752558e-05,
+      "loss": 0.4237,
       "step": 67490
     },
     {
-      "epoch": 2.37,
-      "learning_rate": 2.8248302846311154e-05,
-      "loss": 0.2713,
+      "epoch": 2.4325152268713737,
+      "grad_norm": 0.20623236894607544,
+      "learning_rate": 2.7317750363855184e-05,
+      "loss": 0.3984,
       "step": 67495
     },
     {
-      "epoch": 2.37,
-      "learning_rate": 2.824547825283349e-05,
-      "loss": 0.2877,
+      "epoch": 2.432695426532598,
+      "grad_norm": 0.19964554905891418,
+      "learning_rate": 2.7314844782378907e-05,
+      "loss": 0.3993,
       "step": 67500
     },
     {
-      "epoch": 2.37,
-      "eval_loss": 0.2662014961242676,
-      "eval_runtime": 10.538,
-      "eval_samples_per_second": 9.489,
-      "eval_steps_per_second": 9.489,
+      "epoch": 2.432695426532598,
+      "eval_loss": 0.43458136916160583,
+      "eval_runtime": 3.5307,
+      "eval_samples_per_second": 28.323,
+      "eval_steps_per_second": 7.081,
       "step": 67500
     },
     {
-      "epoch": 2.38,
-      "learning_rate": 2.8242653617215424e-05,
-      "loss": 0.2656,
+      "epoch": 2.4328756261938227,
+      "grad_norm": 0.2280506193637848,
+      "learning_rate": 2.731193916936331e-05,
+      "loss": 0.3918,
       "step": 67505
     },
     {
-      "epoch": 2.38,
-      "learning_rate": 2.8239828939493644e-05,
-      "loss": 0.2672,
+      "epoch": 2.4330558258550474,
+      "grad_norm": 0.2009294182062149,
+      "learning_rate": 2.7309033524848e-05,
+      "loss": 0.3965,
       "step": 67510
     },
     {
-      "epoch": 2.38,
-      "learning_rate": 2.8237004219704817e-05,
-      "loss": 0.2692,
+      "epoch": 2.433236025516272,
+      "grad_norm": 0.18741028010845184,
+      "learning_rate": 2.7306127848872547e-05,
+      "loss": 0.3826,
       "step": 67515
     },
     {
-      "epoch": 2.38,
-      "learning_rate": 2.823417945788562e-05,
-      "loss": 0.277,
+      "epoch": 2.4334162251774965,
+      "grad_norm": 0.2168714553117752,
+      "learning_rate": 2.7303222141476566e-05,
+      "loss": 0.3582,
       "step": 67520
     },
     {
-      "epoch": 2.38,
-      "learning_rate": 2.8231354654072733e-05,
-      "loss": 0.2665,
+      "epoch": 2.433596424838721,
+      "grad_norm": 0.2223115712404251,
+      "learning_rate": 2.730031640269961e-05,
+      "loss": 0.3941,
       "step": 67525
     },
     {
-      "epoch": 2.38,
-      "learning_rate": 2.8228529808302845e-05,
-      "loss": 0.2728,
+      "epoch": 2.433776624499946,
+      "grad_norm": 0.24095144867897034,
+      "learning_rate": 2.7297410632581293e-05,
+      "loss": 0.4198,
       "step": 67530
     },
     {
-      "epoch": 2.38,
-      "learning_rate": 2.8225704920612616e-05,
-      "loss": 0.2625,
+      "epoch": 2.4339568241611707,
+      "grad_norm": 0.162398099899292,
+      "learning_rate": 2.72945048311612e-05,
+      "loss": 0.3884,
       "step": 67535
     },
     {
-      "epoch": 2.38,
-      "learning_rate": 2.8222879991038736e-05,
-      "loss": 0.2947,
+      "epoch": 2.4341370238223954,
+      "grad_norm": 0.2290612757205963,
+      "learning_rate": 2.7291598998478927e-05,
+      "loss": 0.3902,
       "step": 67540
     },
     {
-      "epoch": 2.38,
-      "learning_rate": 2.8220055019617887e-05,
-      "loss": 0.2678,
+      "epoch": 2.4343172234836197,
+      "grad_norm": 0.19016504287719727,
+      "learning_rate": 2.7288693134574063e-05,
+      "loss": 0.3899,
       "step": 67545
     },
     {
-      "epoch": 2.38,
-      "learning_rate": 2.821723000638674e-05,
-      "loss": 0.2809,
+      "epoch": 2.4344974231448444,
+      "grad_norm": 0.19795559346675873,
+      "learning_rate": 2.7285787239486192e-05,
+      "loss": 0.3749,
       "step": 67550
     },
     {
-      "epoch": 2.38,
-      "learning_rate": 2.821440495138199e-05,
-      "loss": 0.2604,
+      "epoch": 2.434677622806069,
+      "grad_norm": 0.18482926487922668,
+      "learning_rate": 2.7282881313254916e-05,
+      "loss": 0.3787,
       "step": 67555
     },
     {
-      "epoch": 2.38,
-      "learning_rate": 2.82115798546403e-05,
-      "loss": 0.2859,
+      "epoch": 2.434857822467294,
+      "grad_norm": 0.23885639011859894,
+      "learning_rate": 2.727997535591983e-05,
+      "loss": 0.4029,
       "step": 67560
     },
     {
-      "epoch": 2.38,
-      "learning_rate": 2.8208754716198376e-05,
-      "loss": 0.2527,
+      "epoch": 2.435038022128518,
+      "grad_norm": 0.17567263543605804,
+      "learning_rate": 2.7277069367520513e-05,
+      "loss": 0.3902,
       "step": 67565
     },
     {
-      "epoch": 2.38,
-      "learning_rate": 2.8205929536092878e-05,
-      "loss": 0.2474,
+      "epoch": 2.435218221789743,
+      "grad_norm": 0.195444256067276,
+      "learning_rate": 2.7274163348096564e-05,
+      "loss": 0.4088,
       "step": 67570
     },
     {
-      "epoch": 2.38,
-      "learning_rate": 2.8203104314360495e-05,
-      "loss": 0.2517,
+      "epoch": 2.4353984214509676,
+      "grad_norm": 0.20873726904392242,
+      "learning_rate": 2.727125729768759e-05,
+      "loss": 0.4304,
       "step": 67575
     },
     {
-      "epoch": 2.38,
-      "learning_rate": 2.8200279051037914e-05,
-      "loss": 0.305,
+      "epoch": 2.4355786211121924,
+      "grad_norm": 0.22101178765296936,
+      "learning_rate": 2.7268351216333164e-05,
+      "loss": 0.4032,
       "step": 67580
     },
     {
-      "epoch": 2.38,
-      "learning_rate": 2.8197453746161827e-05,
-      "loss": 0.2641,
+      "epoch": 2.435758820773417,
+      "grad_norm": 0.23092830181121826,
+      "learning_rate": 2.72654451040729e-05,
+      "loss": 0.4214,
       "step": 67585
     },
     {
-      "epoch": 2.38,
-      "learning_rate": 2.81946283997689e-05,
-      "loss": 0.2854,
+      "epoch": 2.4359390204346414,
+      "grad_norm": 0.21683692932128906,
+      "learning_rate": 2.7262538960946382e-05,
+      "loss": 0.4102,
       "step": 67590
     },
     {
-      "epoch": 2.38,
-      "learning_rate": 2.819180301189584e-05,
-      "loss": 0.2652,
+      "epoch": 2.436119220095866,
+      "grad_norm": 0.21636758744716644,
+      "learning_rate": 2.7259632786993205e-05,
+      "loss": 0.3927,
       "step": 67595
     },
     {
-      "epoch": 2.38,
-      "learning_rate": 2.8188977582579308e-05,
-      "loss": 0.2688,
+      "epoch": 2.436299419757091,
+      "grad_norm": 0.22299599647521973,
+      "learning_rate": 2.725672658225298e-05,
+      "loss": 0.3997,
       "step": 67600
     },
     {
-      "epoch": 2.38,
-      "learning_rate": 2.8186152111856013e-05,
-      "loss": 0.273,
+      "epoch": 2.4364796194183156,
+      "grad_norm": 0.1862391084432602,
+      "learning_rate": 2.7253820346765285e-05,
+      "loss": 0.4002,
       "step": 67605
     },
     {
-      "epoch": 2.38,
-      "learning_rate": 2.8183326599762627e-05,
-      "loss": 0.2844,
+      "epoch": 2.4366598190795403,
+      "grad_norm": 0.18761450052261353,
+      "learning_rate": 2.725091408056972e-05,
+      "loss": 0.3688,
       "step": 67610
     },
     {
-      "epoch": 2.38,
-      "learning_rate": 2.818050104633585e-05,
-      "loss": 0.2694,
+      "epoch": 2.4368400187407646,
+      "grad_norm": 0.19558748602867126,
+      "learning_rate": 2.724800778370589e-05,
+      "loss": 0.4267,
       "step": 67615
     },
     {
-      "epoch": 2.38,
-      "learning_rate": 2.8177675451612358e-05,
-      "loss": 0.2896,
+      "epoch": 2.4370202184019893,
+      "grad_norm": 0.154653400182724,
+      "learning_rate": 2.724510145621339e-05,
+      "loss": 0.3866,
       "step": 67620
     },
     {
-      "epoch": 2.38,
-      "learning_rate": 2.8174849815628846e-05,
-      "loss": 0.2841,
+      "epoch": 2.437200418063214,
+      "grad_norm": 0.18840929865837097,
+      "learning_rate": 2.7242195098131824e-05,
+      "loss": 0.3955,
       "step": 67625
     },
     {
-      "epoch": 2.38,
-      "learning_rate": 2.8172024138422e-05,
-      "loss": 0.2586,
+      "epoch": 2.437380617724439,
+      "grad_norm": 0.2518814206123352,
+      "learning_rate": 2.7239288709500772e-05,
+      "loss": 0.4223,
       "step": 67630
     },
     {
-      "epoch": 2.38,
-      "learning_rate": 2.816919842002851e-05,
-      "loss": 0.2634,
+      "epoch": 2.4375608173856635,
+      "grad_norm": 0.2483137845993042,
+      "learning_rate": 2.723638229035985e-05,
+      "loss": 0.3919,
       "step": 67635
     },
     {
-      "epoch": 2.38,
-      "learning_rate": 2.816637266048507e-05,
-      "loss": 0.2691,
+      "epoch": 2.437741017046888,
+      "grad_norm": 0.18424572050571442,
+      "learning_rate": 2.7233475840748645e-05,
+      "loss": 0.3912,
       "step": 67640
     },
     {
-      "epoch": 2.38,
-      "learning_rate": 2.816354685982837e-05,
-      "loss": 0.2768,
+      "epoch": 2.4379212167081126,
+      "grad_norm": 0.18739376962184906,
+      "learning_rate": 2.7230569360706777e-05,
+      "loss": 0.4361,
       "step": 67645
     },
     {
-      "epoch": 2.38,
-      "learning_rate": 2.8160721018095094e-05,
-      "loss": 0.2891,
+      "epoch": 2.4381014163693373,
+      "grad_norm": 0.16617445647716522,
+      "learning_rate": 2.7227662850273827e-05,
+      "loss": 0.3941,
       "step": 67650
     },
     {
-      "epoch": 2.38,
-      "learning_rate": 2.815789513532194e-05,
-      "loss": 0.2806,
+      "epoch": 2.438281616030562,
+      "grad_norm": 0.23537923395633698,
+      "learning_rate": 2.7224756309489403e-05,
+      "loss": 0.3962,
       "step": 67655
     },
     {
-      "epoch": 2.38,
-      "learning_rate": 2.8155069211545594e-05,
-      "loss": 0.272,
+      "epoch": 2.4384618156917863,
+      "grad_norm": 0.21504931151866913,
+      "learning_rate": 2.7221849738393103e-05,
+      "loss": 0.4205,
       "step": 67660
     },
     {
-      "epoch": 2.38,
-      "learning_rate": 2.8152243246802757e-05,
-      "loss": 0.2733,
+      "epoch": 2.438642015353011,
+      "grad_norm": 0.18787220120429993,
+      "learning_rate": 2.7218943137024532e-05,
+      "loss": 0.3643,
       "step": 67665
     },
     {
-      "epoch": 2.38,
-      "learning_rate": 2.8149417241130122e-05,
-      "loss": 0.266,
+      "epoch": 2.438822215014236,
+      "grad_norm": 0.24440403282642365,
+      "learning_rate": 2.7216036505423293e-05,
+      "loss": 0.3797,
       "step": 67670
     },
     {
-      "epoch": 2.38,
-      "learning_rate": 2.8146591194564377e-05,
-      "loss": 0.2856,
+      "epoch": 2.4390024146754605,
+      "grad_norm": 0.2223052680492401,
+      "learning_rate": 2.721312984362898e-05,
+      "loss": 0.4005,
       "step": 67675
     },
     {
-      "epoch": 2.38,
-      "learning_rate": 2.8143765107142218e-05,
-      "loss": 0.2567,
+      "epoch": 2.4391826143366853,
+      "grad_norm": 0.18885168433189392,
+      "learning_rate": 2.7210223151681207e-05,
+      "loss": 0.3727,
       "step": 67680
     },
     {
-      "epoch": 2.38,
-      "learning_rate": 2.8140938978900345e-05,
-      "loss": 0.284,
+      "epoch": 2.4393628139979096,
+      "grad_norm": 0.1814367175102234,
+      "learning_rate": 2.7207316429619566e-05,
+      "loss": 0.3837,
       "step": 67685
     },
     {
-      "epoch": 2.38,
-      "learning_rate": 2.813811280987545e-05,
-      "loss": 0.2534,
+      "epoch": 2.4395430136591343,
+      "grad_norm": 0.18942394852638245,
+      "learning_rate": 2.7204409677483668e-05,
+      "loss": 0.4423,
       "step": 67690
     },
     {
-      "epoch": 2.38,
-      "learning_rate": 2.8135286600104222e-05,
-      "loss": 0.2557,
+      "epoch": 2.439723213320359,
+      "grad_norm": 0.16898776590824127,
+      "learning_rate": 2.720150289531312e-05,
+      "loss": 0.3894,
       "step": 67695
     },
     {
-      "epoch": 2.38,
-      "learning_rate": 2.813246034962337e-05,
-      "loss": 0.252,
+      "epoch": 2.4399034129815838,
+      "grad_norm": 0.1935228556394577,
+      "learning_rate": 2.719859608314751e-05,
+      "loss": 0.4158,
       "step": 67700
     },
     {
-      "epoch": 2.38,
-      "learning_rate": 2.812963405846958e-05,
-      "loss": 0.2668,
+      "epoch": 2.440083612642808,
+      "grad_norm": 0.22555245459079742,
+      "learning_rate": 2.719568924102647e-05,
+      "loss": 0.3861,
       "step": 67705
     },
     {
-      "epoch": 2.38,
-      "learning_rate": 2.8126807726679556e-05,
-      "loss": 0.2756,
+      "epoch": 2.4402638123040328,
+      "grad_norm": 0.22015783190727234,
+      "learning_rate": 2.7192782368989577e-05,
+      "loss": 0.3656,
       "step": 67710
     },
     {
-      "epoch": 2.38,
-      "learning_rate": 2.812398135428999e-05,
-      "loss": 0.2436,
+      "epoch": 2.4404440119652575,
+      "grad_norm": 0.20583826303482056,
+      "learning_rate": 2.7189875467076454e-05,
+      "loss": 0.4075,
       "step": 67715
     },
     {
-      "epoch": 2.38,
-      "learning_rate": 2.812115494133759e-05,
-      "loss": 0.2669,
+      "epoch": 2.4406242116264822,
+      "grad_norm": 0.2194293737411499,
+      "learning_rate": 2.7186968535326694e-05,
+      "loss": 0.4094,
       "step": 67720
     },
     {
-      "epoch": 2.38,
-      "learning_rate": 2.8118328487859053e-05,
-      "loss": 0.2457,
+      "epoch": 2.440804411287707,
+      "grad_norm": 0.15374071896076202,
+      "learning_rate": 2.7184061573779918e-05,
+      "loss": 0.3868,
       "step": 67725
     },
     {
-      "epoch": 2.38,
-      "learning_rate": 2.8115501993891074e-05,
-      "loss": 0.3102,
+      "epoch": 2.4409846109489313,
+      "grad_norm": 0.2262052297592163,
+      "learning_rate": 2.7181154582475726e-05,
+      "loss": 0.4121,
       "step": 67730
     },
     {
-      "epoch": 2.38,
-      "learning_rate": 2.811267545947035e-05,
-      "loss": 0.2651,
+      "epoch": 2.441164810610156,
+      "grad_norm": 0.21393641829490662,
+      "learning_rate": 2.7178247561453724e-05,
+      "loss": 0.3887,
       "step": 67735
     },
     {
-      "epoch": 2.38,
-      "learning_rate": 2.8109848884633588e-05,
-      "loss": 0.2928,
+      "epoch": 2.4413450102713807,
+      "grad_norm": 0.23438535630702972,
+      "learning_rate": 2.717534051075352e-05,
+      "loss": 0.4055,
       "step": 67740
     },
     {
-      "epoch": 2.38,
-      "learning_rate": 2.8107022269417488e-05,
-      "loss": 0.2644,
+      "epoch": 2.4415252099326055,
+      "grad_norm": 0.1914755254983902,
+      "learning_rate": 2.7172433430414725e-05,
+      "loss": 0.4044,
       "step": 67745
     },
     {
-      "epoch": 2.38,
-      "learning_rate": 2.8104195613858754e-05,
-      "loss": 0.2744,
+      "epoch": 2.4417054095938298,
+      "grad_norm": 0.22475279867649078,
+      "learning_rate": 2.7169526320476944e-05,
+      "loss": 0.3659,
       "step": 67750
     },
     {
-      "epoch": 2.38,
-      "learning_rate": 2.8101368917994082e-05,
-      "loss": 0.2925,
+      "epoch": 2.4418856092550545,
+      "grad_norm": 0.2241634875535965,
+      "learning_rate": 2.7166619180979784e-05,
+      "loss": 0.4037,
       "step": 67755
     },
     {
-      "epoch": 2.38,
-      "learning_rate": 2.8098542181860183e-05,
-      "loss": 0.2604,
+      "epoch": 2.4420658089162792,
+      "grad_norm": 0.23309601843357086,
+      "learning_rate": 2.716371201196286e-05,
+      "loss": 0.3942,
       "step": 67760
     },
     {
-      "epoch": 2.38,
-      "learning_rate": 2.809571540549375e-05,
-      "loss": 0.2893,
+      "epoch": 2.442246008577504,
+      "grad_norm": 0.20057342946529388,
+      "learning_rate": 2.716080481346578e-05,
+      "loss": 0.3939,
       "step": 67765
     },
     {
-      "epoch": 2.38,
-      "learning_rate": 2.80928885889315e-05,
-      "loss": 0.2932,
+      "epoch": 2.4424262082387287,
+      "grad_norm": 0.19580478966236115,
+      "learning_rate": 2.715789758552815e-05,
+      "loss": 0.3698,
       "step": 67770
     },
     {
-      "epoch": 2.38,
-      "learning_rate": 2.809006173221012e-05,
-      "loss": 0.2549,
+      "epoch": 2.442606407899953,
+      "grad_norm": 0.21272897720336914,
+      "learning_rate": 2.7154990328189583e-05,
+      "loss": 0.4033,
       "step": 67775
     },
     {
-      "epoch": 2.38,
-      "learning_rate": 2.8087234835366333e-05,
-      "loss": 0.2662,
+      "epoch": 2.4427866075611777,
+      "grad_norm": 0.22313345968723297,
+      "learning_rate": 2.715208304148969e-05,
+      "loss": 0.3641,
       "step": 67780
     },
     {
-      "epoch": 2.38,
-      "learning_rate": 2.8084407898436837e-05,
-      "loss": 0.2651,
+      "epoch": 2.4429668072224024,
+      "grad_norm": 0.20916157960891724,
+      "learning_rate": 2.7149175725468086e-05,
+      "loss": 0.3875,
       "step": 67785
     },
     {
-      "epoch": 2.39,
-      "learning_rate": 2.8081580921458335e-05,
-      "loss": 0.2785,
+      "epoch": 2.443147006883627,
+      "grad_norm": 0.20672792196273804,
+      "learning_rate": 2.7146268380164376e-05,
+      "loss": 0.4256,
       "step": 67790
     },
     {
-      "epoch": 2.39,
-      "learning_rate": 2.8078753904467532e-05,
-      "loss": 0.2767,
+      "epoch": 2.4433272065448515,
+      "grad_norm": 0.19972656667232513,
+      "learning_rate": 2.7143361005618168e-05,
+      "loss": 0.4102,
       "step": 67795
     },
     {
-      "epoch": 2.39,
-      "learning_rate": 2.8075926847501137e-05,
-      "loss": 0.2838,
+      "epoch": 2.443507406206076,
+      "grad_norm": 0.18391171097755432,
+      "learning_rate": 2.7140453601869093e-05,
+      "loss": 0.4145,
       "step": 67800
     },
     {
-      "epoch": 2.39,
-      "learning_rate": 2.8073099750595866e-05,
-      "loss": 0.2798,
+      "epoch": 2.443687605867301,
+      "grad_norm": 0.17627693712711334,
+      "learning_rate": 2.713754616895674e-05,
+      "loss": 0.3597,
       "step": 67805
     },
     {
-      "epoch": 2.39,
-      "learning_rate": 2.8070272613788417e-05,
-      "loss": 0.276,
+      "epoch": 2.4438678055285257,
+      "grad_norm": 0.25177159905433655,
+      "learning_rate": 2.7134638706920752e-05,
+      "loss": 0.4215,
       "step": 67810
     },
     {
-      "epoch": 2.39,
-      "learning_rate": 2.8067445437115504e-05,
-      "loss": 0.2661,
+      "epoch": 2.4440480051897504,
+      "grad_norm": 0.2532104253768921,
+      "learning_rate": 2.713173121580071e-05,
+      "loss": 0.3984,
       "step": 67815
     },
     {
-      "epoch": 2.39,
-      "learning_rate": 2.8064618220613826e-05,
-      "loss": 0.2742,
+      "epoch": 2.4442282048509747,
+      "grad_norm": 0.19765953719615936,
+      "learning_rate": 2.7128823695636253e-05,
+      "loss": 0.4179,
       "step": 67820
     },
     {
-      "epoch": 2.39,
-      "learning_rate": 2.8061790964320108e-05,
-      "loss": 0.2534,
+      "epoch": 2.4444084045121994,
+      "grad_norm": 0.2110573947429657,
+      "learning_rate": 2.7125916146466973e-05,
+      "loss": 0.3967,
       "step": 67825
     },
     {
-      "epoch": 2.39,
-      "learning_rate": 2.8058963668271047e-05,
-      "loss": 0.2752,
+      "epoch": 2.444588604173424,
+      "grad_norm": 0.22106719017028809,
+      "learning_rate": 2.7123008568332504e-05,
+      "loss": 0.4195,
       "step": 67830
     },
     {
-      "epoch": 2.39,
-      "learning_rate": 2.8056136332503362e-05,
-      "loss": 0.2537,
+      "epoch": 2.444768803834649,
+      "grad_norm": 0.21176251769065857,
+      "learning_rate": 2.7120100961272455e-05,
+      "loss": 0.3917,
       "step": 67835
     },
     {
-      "epoch": 2.39,
-      "learning_rate": 2.805330895705376e-05,
-      "loss": 0.2581,
+      "epoch": 2.444949003495873,
+      "grad_norm": 0.2130795568227768,
+      "learning_rate": 2.711719332532644e-05,
+      "loss": 0.3974,
       "step": 67840
     },
     {
-      "epoch": 2.39,
-      "learning_rate": 2.805048154195895e-05,
-      "loss": 0.2517,
+      "epoch": 2.445129203157098,
+      "grad_norm": 0.20041079819202423,
+      "learning_rate": 2.7114285660534078e-05,
+      "loss": 0.415,
       "step": 67845
     },
     {
-      "epoch": 2.39,
-      "learning_rate": 2.804765408725565e-05,
-      "loss": 0.2553,
+      "epoch": 2.4453094028183227,
+      "grad_norm": 0.2205459475517273,
+      "learning_rate": 2.711137796693498e-05,
+      "loss": 0.4196,
       "step": 67850
     },
     {
-      "epoch": 2.39,
-      "learning_rate": 2.8044826592980577e-05,
-      "loss": 0.2766,
+      "epoch": 2.4454896024795474,
+      "grad_norm": 0.2311592847108841,
+      "learning_rate": 2.7108470244568767e-05,
+      "loss": 0.4111,
       "step": 67855
     },
     {
-      "epoch": 2.39,
-      "learning_rate": 2.8041999059170426e-05,
-      "loss": 0.2772,
+      "epoch": 2.445669802140772,
+      "grad_norm": 0.2623636722564697,
+      "learning_rate": 2.7105562493475052e-05,
+      "loss": 0.4283,
       "step": 67860
     },
     {
-      "epoch": 2.39,
-      "learning_rate": 2.8039171485861932e-05,
-      "loss": 0.2644,
+      "epoch": 2.445850001801997,
+      "grad_norm": 0.21312451362609863,
+      "learning_rate": 2.710265471369346e-05,
+      "loss": 0.3945,
       "step": 67865
     },
     {
-      "epoch": 2.39,
-      "learning_rate": 2.8036343873091793e-05,
-      "loss": 0.2892,
+      "epoch": 2.446030201463221,
+      "grad_norm": 0.1985563039779663,
+      "learning_rate": 2.7099746905263606e-05,
+      "loss": 0.356,
       "step": 67870
     },
     {
-      "epoch": 2.39,
-      "learning_rate": 2.803351622089673e-05,
-      "loss": 0.2932,
+      "epoch": 2.446210401124446,
+      "grad_norm": 0.20180106163024902,
+      "learning_rate": 2.70968390682251e-05,
+      "loss": 0.3865,
       "step": 67875
     },
     {
-      "epoch": 2.39,
-      "learning_rate": 2.803068852931346e-05,
-      "loss": 0.2749,
+      "epoch": 2.4463906007856706,
+      "grad_norm": 0.1858830749988556,
+      "learning_rate": 2.7093931202617572e-05,
+      "loss": 0.417,
       "step": 67880
     },
     {
-      "epoch": 2.39,
-      "learning_rate": 2.8027860798378697e-05,
-      "loss": 0.2781,
+      "epoch": 2.4465708004468953,
+      "grad_norm": 0.2230648547410965,
+      "learning_rate": 2.7091023308480632e-05,
+      "loss": 0.4184,
       "step": 67885
     },
     {
-      "epoch": 2.39,
-      "learning_rate": 2.8025033028129155e-05,
-      "loss": 0.2682,
+      "epoch": 2.4467510001081196,
+      "grad_norm": 0.24601806700229645,
+      "learning_rate": 2.708811538585392e-05,
+      "loss": 0.4014,
       "step": 67890
     },
     {
-      "epoch": 2.39,
-      "learning_rate": 2.802220521860155e-05,
-      "loss": 0.2645,
+      "epoch": 2.4469311997693444,
+      "grad_norm": 0.1883206069469452,
+      "learning_rate": 2.7085207434777026e-05,
+      "loss": 0.3919,
       "step": 67895
     },
     {
-      "epoch": 2.39,
-      "learning_rate": 2.8019377369832605e-05,
-      "loss": 0.2776,
+      "epoch": 2.447111399430569,
+      "grad_norm": 0.22426816821098328,
+      "learning_rate": 2.7082299455289588e-05,
+      "loss": 0.4058,
       "step": 67900
     },
     {
-      "epoch": 2.39,
-      "learning_rate": 2.8016549481859035e-05,
-      "loss": 0.2684,
+      "epoch": 2.447291599091794,
+      "grad_norm": 0.2288169413805008,
+      "learning_rate": 2.7079391447431218e-05,
+      "loss": 0.3892,
       "step": 67905
     },
     {
-      "epoch": 2.39,
-      "learning_rate": 2.801372155471756e-05,
-      "loss": 0.2765,
+      "epoch": 2.4474717987530186,
+      "grad_norm": 0.2036815732717514,
+      "learning_rate": 2.7076483411241547e-05,
+      "loss": 0.3782,
       "step": 67910
     },
     {
-      "epoch": 2.39,
-      "learning_rate": 2.8010893588444892e-05,
-      "loss": 0.2764,
+      "epoch": 2.447651998414243,
+      "grad_norm": 0.1831190586090088,
+      "learning_rate": 2.7073575346760194e-05,
+      "loss": 0.3705,
       "step": 67915
     },
     {
-      "epoch": 2.39,
-      "learning_rate": 2.8008065583077763e-05,
-      "loss": 0.2738,
+      "epoch": 2.4478321980754676,
+      "grad_norm": 0.1885637491941452,
+      "learning_rate": 2.707066725402677e-05,
+      "loss": 0.4212,
       "step": 67920
     },
     {
-      "epoch": 2.39,
-      "learning_rate": 2.8005237538652878e-05,
-      "loss": 0.2562,
+      "epoch": 2.4480123977366923,
+      "grad_norm": 0.202529639005661,
+      "learning_rate": 2.7067759133080917e-05,
+      "loss": 0.4009,
       "step": 67925
     },
     {
-      "epoch": 2.39,
-      "learning_rate": 2.8002409455206964e-05,
-      "loss": 0.2766,
+      "epoch": 2.448192597397917,
+      "grad_norm": 0.22661788761615753,
+      "learning_rate": 2.7064850983962243e-05,
+      "loss": 0.4497,
       "step": 67930
     },
     {
-      "epoch": 2.39,
-      "learning_rate": 2.799958133277674e-05,
-      "loss": 0.272,
+      "epoch": 2.4483727970591413,
+      "grad_norm": 0.2267916351556778,
+      "learning_rate": 2.7061942806710368e-05,
+      "loss": 0.3874,
       "step": 67935
     },
     {
-      "epoch": 2.39,
-      "learning_rate": 2.7996753171398937e-05,
-      "loss": 0.2884,
+      "epoch": 2.448552996720366,
+      "grad_norm": 0.22525878250598907,
+      "learning_rate": 2.7059034601364925e-05,
+      "loss": 0.3741,
       "step": 67940
     },
     {
-      "epoch": 2.39,
-      "learning_rate": 2.7993924971110262e-05,
-      "loss": 0.2617,
+      "epoch": 2.448733196381591,
+      "grad_norm": 0.18289059400558472,
+      "learning_rate": 2.705612636796553e-05,
+      "loss": 0.3917,
       "step": 67945
     },
     {
-      "epoch": 2.39,
-      "learning_rate": 2.7991096731947454e-05,
-      "loss": 0.2923,
+      "epoch": 2.4489133960428155,
+      "grad_norm": 0.24663716554641724,
+      "learning_rate": 2.7053218106551825e-05,
+      "loss": 0.3758,
       "step": 67950
     },
     {
-      "epoch": 2.39,
-      "learning_rate": 2.798826845394722e-05,
-      "loss": 0.2731,
+      "epoch": 2.4490935957040403,
+      "grad_norm": 0.25891220569610596,
+      "learning_rate": 2.7050309817163407e-05,
+      "loss": 0.4179,
       "step": 67955
     },
     {
-      "epoch": 2.39,
-      "learning_rate": 2.7985440137146283e-05,
-      "loss": 0.2692,
+      "epoch": 2.4492737953652646,
+      "grad_norm": 0.20873096585273743,
+      "learning_rate": 2.704740149983992e-05,
+      "loss": 0.4246,
       "step": 67960
     },
     {
-      "epoch": 2.39,
-      "learning_rate": 2.7982611781581382e-05,
-      "loss": 0.2848,
+      "epoch": 2.4494539950264893,
+      "grad_norm": 0.2050144225358963,
+      "learning_rate": 2.7044493154620975e-05,
+      "loss": 0.4267,
       "step": 67965
     },
     {
-      "epoch": 2.39,
-      "learning_rate": 2.7979783387289233e-05,
-      "loss": 0.2871,
+      "epoch": 2.449634194687714,
+      "grad_norm": 0.189696803689003,
+      "learning_rate": 2.704158478154622e-05,
+      "loss": 0.3697,
       "step": 67970
     },
     {
-      "epoch": 2.39,
-      "learning_rate": 2.7976954954306557e-05,
-      "loss": 0.2772,
+      "epoch": 2.4498143943489388,
+      "grad_norm": 0.2150600403547287,
+      "learning_rate": 2.7038676380655265e-05,
+      "loss": 0.3711,
       "step": 67975
     },
     {
-      "epoch": 2.39,
-      "learning_rate": 2.7974126482670078e-05,
-      "loss": 0.2792,
+      "epoch": 2.449994594010163,
+      "grad_norm": 0.2863604724407196,
+      "learning_rate": 2.7035767951987735e-05,
+      "loss": 0.4377,
       "step": 67980
     },
     {
-      "epoch": 2.39,
-      "learning_rate": 2.7971297972416527e-05,
-      "loss": 0.2637,
+      "epoch": 2.450174793671388,
+      "grad_norm": 0.21974705159664154,
+      "learning_rate": 2.7032859495583258e-05,
+      "loss": 0.4083,
       "step": 67985
     },
     {
-      "epoch": 2.39,
-      "learning_rate": 2.7968469423582637e-05,
-      "loss": 0.264,
+      "epoch": 2.4503549933326125,
+      "grad_norm": 0.22870804369449615,
+      "learning_rate": 2.7029951011481464e-05,
+      "loss": 0.3969,
       "step": 67990
     },
     {
-      "epoch": 2.39,
-      "learning_rate": 2.7965640836205127e-05,
-      "loss": 0.2702,
+      "epoch": 2.4505351929938373,
+      "grad_norm": 0.19309473037719727,
+      "learning_rate": 2.7027042499721994e-05,
+      "loss": 0.3831,
       "step": 67995
     },
     {
-      "epoch": 2.39,
-      "learning_rate": 2.7962812210320726e-05,
-      "loss": 0.2813,
+      "epoch": 2.450715392655062,
+      "grad_norm": 0.22232699394226074,
+      "learning_rate": 2.702413396034445e-05,
+      "loss": 0.3634,
       "step": 68000
     },
     {
-      "epoch": 2.39,
-      "eval_loss": 0.2646848261356354,
-      "eval_runtime": 10.557,
-      "eval_samples_per_second": 9.472,
-      "eval_steps_per_second": 9.472,
+      "epoch": 2.450715392655062,
+      "eval_loss": 0.43447232246398926,
+      "eval_runtime": 3.5294,
+      "eval_samples_per_second": 28.333,
+      "eval_steps_per_second": 7.083,
       "step": 68000
     },
     {
-      "epoch": 2.39,
-      "learning_rate": 2.7959983545966155e-05,
-      "loss": 0.2767,
+      "epoch": 2.4508955923162863,
+      "grad_norm": 0.16693948209285736,
+      "learning_rate": 2.7021225393388477e-05,
+      "loss": 0.3795,
       "step": 68005
     },
     {
-      "epoch": 2.39,
-      "learning_rate": 2.7957154843178153e-05,
-      "loss": 0.2649,
+      "epoch": 2.451075791977511,
+      "grad_norm": 0.19517579674720764,
+      "learning_rate": 2.7018316798893695e-05,
+      "loss": 0.396,
       "step": 68010
     },
     {
-      "epoch": 2.39,
-      "learning_rate": 2.7954326101993445e-05,
-      "loss": 0.2803,
+      "epoch": 2.4512559916387358,
+      "grad_norm": 0.18335868418216705,
+      "learning_rate": 2.701540817689974e-05,
+      "loss": 0.3768,
       "step": 68015
     },
     {
-      "epoch": 2.39,
-      "learning_rate": 2.7951497322448757e-05,
-      "loss": 0.2935,
+      "epoch": 2.4514361912999605,
+      "grad_norm": 0.24918295443058014,
+      "learning_rate": 2.701249952744624e-05,
+      "loss": 0.3984,
       "step": 68020
     },
     {
-      "epoch": 2.39,
-      "learning_rate": 2.794866850458082e-05,
-      "loss": 0.2573,
+      "epoch": 2.4516163909611848,
+      "grad_norm": 0.1807776242494583,
+      "learning_rate": 2.700959085057282e-05,
+      "loss": 0.41,
       "step": 68025
     },
     {
-      "epoch": 2.39,
-      "learning_rate": 2.794583964842637e-05,
-      "loss": 0.2796,
+      "epoch": 2.4517965906224095,
+      "grad_norm": 0.18439331650733948,
+      "learning_rate": 2.700668214631912e-05,
+      "loss": 0.3853,
       "step": 68030
     },
     {
-      "epoch": 2.39,
-      "learning_rate": 2.7943010754022137e-05,
-      "loss": 0.2686,
+      "epoch": 2.4519767902836342,
+      "grad_norm": 0.19809627532958984,
+      "learning_rate": 2.700377341472476e-05,
+      "loss": 0.3885,
       "step": 68035
     },
     {
-      "epoch": 2.39,
-      "learning_rate": 2.7940181821404832e-05,
-      "loss": 0.263,
+      "epoch": 2.452156989944859,
+      "grad_norm": 0.22589033842086792,
+      "learning_rate": 2.7000864655829377e-05,
+      "loss": 0.4058,
       "step": 68040
     },
     {
-      "epoch": 2.39,
-      "learning_rate": 2.793735285061122e-05,
-      "loss": 0.2584,
+      "epoch": 2.4523371896060837,
+      "grad_norm": 0.2336047887802124,
+      "learning_rate": 2.699795586967259e-05,
+      "loss": 0.435,
       "step": 68045
     },
     {
-      "epoch": 2.39,
-      "learning_rate": 2.7934523841678013e-05,
-      "loss": 0.2866,
+      "epoch": 2.452517389267308,
+      "grad_norm": 0.20116937160491943,
+      "learning_rate": 2.6995047056294054e-05,
+      "loss": 0.3978,
       "step": 68050
     },
     {
-      "epoch": 2.39,
-      "learning_rate": 2.793169479464195e-05,
-      "loss": 0.2616,
+      "epoch": 2.4526975889285327,
+      "grad_norm": 0.2507151961326599,
+      "learning_rate": 2.6992138215733385e-05,
+      "loss": 0.4326,
       "step": 68055
     },
     {
-      "epoch": 2.39,
-      "learning_rate": 2.7928865709539754e-05,
-      "loss": 0.253,
+      "epoch": 2.4528777885897575,
+      "grad_norm": 0.21996821463108063,
+      "learning_rate": 2.6989229348030214e-05,
+      "loss": 0.3858,
       "step": 68060
     },
     {
-      "epoch": 2.39,
-      "learning_rate": 2.7926036586408172e-05,
-      "loss": 0.2722,
+      "epoch": 2.453057988250982,
+      "grad_norm": 0.2189149707555771,
+      "learning_rate": 2.698632045322419e-05,
+      "loss": 0.434,
       "step": 68065
     },
     {
-      "epoch": 2.39,
-      "learning_rate": 2.7923207425283933e-05,
-      "loss": 0.2763,
+      "epoch": 2.4532381879122065,
+      "grad_norm": 0.17697672545909882,
+      "learning_rate": 2.6983411531354918e-05,
+      "loss": 0.3774,
       "step": 68070
     },
     {
-      "epoch": 2.4,
-      "learning_rate": 2.7920378226203775e-05,
-      "loss": 0.2651,
+      "epoch": 2.4534183875734312,
+      "grad_norm": 0.17975100874900818,
+      "learning_rate": 2.6980502582462063e-05,
+      "loss": 0.4279,
       "step": 68075
     },
     {
-      "epoch": 2.4,
-      "learning_rate": 2.7917548989204422e-05,
-      "loss": 0.2692,
+      "epoch": 2.453598587234656,
+      "grad_norm": 0.2552946209907532,
+      "learning_rate": 2.6977593606585232e-05,
+      "loss": 0.3939,
       "step": 68080
     },
     {
-      "epoch": 2.4,
-      "learning_rate": 2.7914719714322628e-05,
-      "loss": 0.2624,
+      "epoch": 2.4537787868958807,
+      "grad_norm": 0.23870886862277985,
+      "learning_rate": 2.697468460376408e-05,
+      "loss": 0.3806,
       "step": 68085
     },
     {
-      "epoch": 2.4,
-      "learning_rate": 2.7911890401595114e-05,
-      "loss": 0.2823,
+      "epoch": 2.4539589865571054,
+      "grad_norm": 0.27226266264915466,
+      "learning_rate": 2.6971775574038226e-05,
+      "loss": 0.4057,
       "step": 68090
     },
     {
-      "epoch": 2.4,
-      "learning_rate": 2.790906105105862e-05,
-      "loss": 0.2674,
+      "epoch": 2.4541391862183297,
+      "grad_norm": 0.1971409022808075,
+      "learning_rate": 2.6968866517447317e-05,
+      "loss": 0.3958,
       "step": 68095
     },
     {
-      "epoch": 2.4,
-      "learning_rate": 2.7906231662749886e-05,
-      "loss": 0.2812,
+      "epoch": 2.4543193858795544,
+      "grad_norm": 0.21888650953769684,
+      "learning_rate": 2.696595743403098e-05,
+      "loss": 0.3885,
       "step": 68100
     },
     {
-      "epoch": 2.4,
-      "learning_rate": 2.790340223670566e-05,
-      "loss": 0.262,
+      "epoch": 2.454499585540779,
+      "grad_norm": 0.2371811419725418,
+      "learning_rate": 2.6963048323828848e-05,
+      "loss": 0.3722,
       "step": 68105
     },
     {
-      "epoch": 2.4,
-      "learning_rate": 2.790057277296266e-05,
-      "loss": 0.2699,
+      "epoch": 2.454679785202004,
+      "grad_norm": 0.23788423836231232,
+      "learning_rate": 2.6960139186880574e-05,
+      "loss": 0.3851,
       "step": 68110
     },
     {
-      "epoch": 2.4,
-      "learning_rate": 2.7897743271557634e-05,
-      "loss": 0.2597,
+      "epoch": 2.4548599848632287,
+      "grad_norm": 0.23677048087120056,
+      "learning_rate": 2.6957230023225778e-05,
+      "loss": 0.4163,
       "step": 68115
     },
     {
-      "epoch": 2.4,
-      "learning_rate": 2.789491373252732e-05,
-      "loss": 0.2699,
+      "epoch": 2.455040184524453,
+      "grad_norm": 0.17837996780872345,
+      "learning_rate": 2.6954320832904094e-05,
+      "loss": 0.388,
       "step": 68120
     },
     {
-      "epoch": 2.4,
-      "learning_rate": 2.7892084155908465e-05,
-      "loss": 0.2745,
+      "epoch": 2.4552203841856777,
+      "grad_norm": 0.19991345703601837,
+      "learning_rate": 2.695141161595518e-05,
+      "loss": 0.381,
       "step": 68125
     },
     {
-      "epoch": 2.4,
-      "learning_rate": 2.7889254541737795e-05,
-      "loss": 0.2705,
+      "epoch": 2.4554005838469024,
+      "grad_norm": 0.22142352163791656,
+      "learning_rate": 2.694850237241865e-05,
+      "loss": 0.357,
       "step": 68130
     },
     {
-      "epoch": 2.4,
-      "learning_rate": 2.7886424890052067e-05,
-      "loss": 0.2734,
+      "epoch": 2.455580783508127,
+      "grad_norm": 0.2008516937494278,
+      "learning_rate": 2.6945593102334166e-05,
+      "loss": 0.3499,
       "step": 68135
     },
     {
-      "epoch": 2.4,
-      "learning_rate": 2.7883595200888014e-05,
-      "loss": 0.2759,
+      "epoch": 2.455760983169352,
+      "grad_norm": 0.2108553797006607,
+      "learning_rate": 2.6942683805741337e-05,
+      "loss": 0.4014,
       "step": 68140
     },
     {
-      "epoch": 2.4,
-      "learning_rate": 2.7880765474282374e-05,
-      "loss": 0.2572,
+      "epoch": 2.455941182830576,
+      "grad_norm": 0.21089521050453186,
+      "learning_rate": 2.693977448267983e-05,
+      "loss": 0.4221,
       "step": 68145
     },
     {
-      "epoch": 2.4,
-      "learning_rate": 2.7877935710271884e-05,
-      "loss": 0.266,
+      "epoch": 2.456121382491801,
+      "grad_norm": 0.22154204547405243,
+      "learning_rate": 2.693686513318926e-05,
+      "loss": 0.3958,
       "step": 68150
     },
     {
-      "epoch": 2.4,
-      "learning_rate": 2.7875105908893305e-05,
-      "loss": 0.2584,
+      "epoch": 2.4563015821530256,
+      "grad_norm": 0.17124603688716888,
+      "learning_rate": 2.693395575730928e-05,
+      "loss": 0.4072,
       "step": 68155
     },
     {
-      "epoch": 2.4,
-      "learning_rate": 2.7872276070183373e-05,
-      "loss": 0.2862,
+      "epoch": 2.4564817818142504,
+      "grad_norm": 0.2264595329761505,
+      "learning_rate": 2.6931046355079538e-05,
+      "loss": 0.3937,
       "step": 68160
     },
     {
-      "epoch": 2.4,
-      "learning_rate": 2.7869446194178827e-05,
-      "loss": 0.2482,
+      "epoch": 2.4566619814754747,
+      "grad_norm": 0.22285620868206024,
+      "learning_rate": 2.692813692653965e-05,
+      "loss": 0.3988,
       "step": 68165
     },
     {
-      "epoch": 2.4,
-      "learning_rate": 2.7866616280916412e-05,
-      "loss": 0.2581,
+      "epoch": 2.4568421811366994,
+      "grad_norm": 0.20908579230308533,
+      "learning_rate": 2.6925227471729275e-05,
+      "loss": 0.3614,
       "step": 68170
     },
     {
-      "epoch": 2.4,
-      "learning_rate": 2.7863786330432873e-05,
-      "loss": 0.2524,
+      "epoch": 2.457022380797924,
+      "grad_norm": 0.18623116612434387,
+      "learning_rate": 2.692231799068805e-05,
+      "loss": 0.4202,
       "step": 68175
     },
     {
-      "epoch": 2.4,
-      "learning_rate": 2.786095634276496e-05,
-      "loss": 0.25,
+      "epoch": 2.457202580459149,
+      "grad_norm": 0.216128870844841,
+      "learning_rate": 2.6919408483455615e-05,
+      "loss": 0.3951,
       "step": 68180
     },
     {
-      "epoch": 2.4,
-      "learning_rate": 2.7858126317949407e-05,
-      "loss": 0.2789,
+      "epoch": 2.4573827801203736,
+      "grad_norm": 0.18577782809734344,
+      "learning_rate": 2.6916498950071605e-05,
+      "loss": 0.4192,
       "step": 68185
     },
     {
-      "epoch": 2.4,
-      "learning_rate": 2.785529625602298e-05,
-      "loss": 0.2712,
+      "epoch": 2.457562979781598,
+      "grad_norm": 0.18905296921730042,
+      "learning_rate": 2.6913589390575677e-05,
+      "loss": 0.3804,
       "step": 68190
     },
     {
-      "epoch": 2.4,
-      "learning_rate": 2.78524661570224e-05,
-      "loss": 0.2741,
+      "epoch": 2.4577431794428226,
+      "grad_norm": 0.23588091135025024,
+      "learning_rate": 2.6910679805007454e-05,
+      "loss": 0.4285,
       "step": 68195
     },
     {
-      "epoch": 2.4,
-      "learning_rate": 2.7849636020984425e-05,
-      "loss": 0.2724,
+      "epoch": 2.4579233791040473,
+      "grad_norm": 0.23398438096046448,
+      "learning_rate": 2.6907770193406595e-05,
+      "loss": 0.3918,
       "step": 68200
     },
     {
-      "epoch": 2.4,
-      "learning_rate": 2.784680584794581e-05,
-      "loss": 0.2737,
+      "epoch": 2.458103578765272,
+      "grad_norm": 0.22657303512096405,
+      "learning_rate": 2.6904860555812738e-05,
+      "loss": 0.3779,
       "step": 68205
     },
     {
-      "epoch": 2.4,
-      "learning_rate": 2.7843975637943303e-05,
-      "loss": 0.2678,
+      "epoch": 2.4582837784264964,
+      "grad_norm": 0.17407497763633728,
+      "learning_rate": 2.6901950892265514e-05,
+      "loss": 0.4045,
       "step": 68210
     },
     {
-      "epoch": 2.4,
-      "learning_rate": 2.784114539101364e-05,
-      "loss": 0.269,
+      "epoch": 2.458463978087721,
+      "grad_norm": 0.19006119668483734,
+      "learning_rate": 2.6899041202804587e-05,
+      "loss": 0.3903,
       "step": 68215
     },
     {
-      "epoch": 2.4,
-      "learning_rate": 2.783831510719358e-05,
-      "loss": 0.2992,
+      "epoch": 2.458644177748946,
+      "grad_norm": 0.22295033931732178,
+      "learning_rate": 2.6896131487469595e-05,
+      "loss": 0.3926,
       "step": 68220
     },
     {
-      "epoch": 2.4,
-      "learning_rate": 2.7835484786519868e-05,
-      "loss": 0.302,
+      "epoch": 2.4588243774101706,
+      "grad_norm": 0.21144932508468628,
+      "learning_rate": 2.6893221746300167e-05,
+      "loss": 0.4296,
       "step": 68225
     },
     {
-      "epoch": 2.4,
-      "learning_rate": 2.7832654429029248e-05,
-      "loss": 0.2639,
+      "epoch": 2.4590045770713953,
+      "grad_norm": 0.2753607928752899,
+      "learning_rate": 2.689031197933597e-05,
+      "loss": 0.4244,
       "step": 68230
     },
     {
-      "epoch": 2.4,
-      "learning_rate": 2.7829824034758483e-05,
-      "loss": 0.2903,
+      "epoch": 2.4591847767326196,
+      "grad_norm": 0.18635913729667664,
+      "learning_rate": 2.6887402186616627e-05,
+      "loss": 0.4052,
       "step": 68235
     },
     {
-      "epoch": 2.4,
-      "learning_rate": 2.7826993603744323e-05,
-      "loss": 0.2771,
+      "epoch": 2.4593649763938443,
+      "grad_norm": 0.1916455328464508,
+      "learning_rate": 2.6884492368181808e-05,
+      "loss": 0.4064,
       "step": 68240
     },
     {
-      "epoch": 2.4,
-      "learning_rate": 2.7824163136023506e-05,
-      "loss": 0.2762,
+      "epoch": 2.459545176055069,
+      "grad_norm": 0.21135953068733215,
+      "learning_rate": 2.6881582524071137e-05,
+      "loss": 0.4178,
       "step": 68245
     },
     {
-      "epoch": 2.4,
-      "learning_rate": 2.782133263163279e-05,
-      "loss": 0.2731,
+      "epoch": 2.459725375716294,
+      "grad_norm": 0.20750294625759125,
+      "learning_rate": 2.6878672654324272e-05,
+      "loss": 0.4014,
       "step": 68250
     },
     {
-      "epoch": 2.4,
-      "learning_rate": 2.7818502090608937e-05,
-      "loss": 0.294,
+      "epoch": 2.459905575377518,
+      "grad_norm": 0.18974362313747406,
+      "learning_rate": 2.687576275898085e-05,
+      "loss": 0.3691,
       "step": 68255
     },
     {
-      "epoch": 2.4,
-      "learning_rate": 2.7815671512988685e-05,
-      "loss": 0.2546,
+      "epoch": 2.460085775038743,
+      "grad_norm": 0.19954413175582886,
+      "learning_rate": 2.6872852838080524e-05,
+      "loss": 0.3698,
       "step": 68260
     },
     {
-      "epoch": 2.4,
-      "learning_rate": 2.7812840898808797e-05,
-      "loss": 0.2527,
+      "epoch": 2.4602659746999676,
+      "grad_norm": 0.20801982283592224,
+      "learning_rate": 2.6869942891662947e-05,
+      "loss": 0.3933,
       "step": 68265
     },
     {
-      "epoch": 2.4,
-      "learning_rate": 2.7810010248106024e-05,
-      "loss": 0.2724,
+      "epoch": 2.4604461743611923,
+      "grad_norm": 0.2402622401714325,
+      "learning_rate": 2.6867032919767754e-05,
+      "loss": 0.4223,
       "step": 68270
     },
     {
-      "epoch": 2.4,
-      "learning_rate": 2.7807179560917123e-05,
-      "loss": 0.2793,
+      "epoch": 2.460626374022417,
+      "grad_norm": 0.2088380753993988,
+      "learning_rate": 2.6864122922434603e-05,
+      "loss": 0.4071,
       "step": 68275
     },
     {
-      "epoch": 2.4,
-      "learning_rate": 2.7804348837278836e-05,
-      "loss": 0.263,
+      "epoch": 2.4608065736836413,
+      "grad_norm": 0.2807213366031647,
+      "learning_rate": 2.6861212899703142e-05,
+      "loss": 0.4163,
       "step": 68280
     },
     {
-      "epoch": 2.4,
-      "learning_rate": 2.7801518077227935e-05,
-      "loss": 0.2702,
+      "epoch": 2.460986773344866,
+      "grad_norm": 0.1840617060661316,
+      "learning_rate": 2.685830285161301e-05,
+      "loss": 0.39,
       "step": 68285
     },
     {
-      "epoch": 2.4,
-      "learning_rate": 2.7798687280801166e-05,
-      "loss": 0.2803,
+      "epoch": 2.4611669730060908,
+      "grad_norm": 0.21558882296085358,
+      "learning_rate": 2.6855392778203858e-05,
+      "loss": 0.3968,
       "step": 68290
     },
     {
-      "epoch": 2.4,
-      "learning_rate": 2.779585644803529e-05,
-      "loss": 0.287,
+      "epoch": 2.4613471726673155,
+      "grad_norm": 0.2157306671142578,
+      "learning_rate": 2.6852482679515345e-05,
+      "loss": 0.4016,
       "step": 68295
     },
     {
-      "epoch": 2.4,
-      "learning_rate": 2.779302557896706e-05,
-      "loss": 0.262,
+      "epoch": 2.46152737232854,
+      "grad_norm": 0.18490919470787048,
+      "learning_rate": 2.6849572555587116e-05,
+      "loss": 0.4155,
       "step": 68300
     },
     {
-      "epoch": 2.4,
-      "learning_rate": 2.7790194673633228e-05,
-      "loss": 0.2756,
+      "epoch": 2.4617075719897645,
+      "grad_norm": 0.2350340336561203,
+      "learning_rate": 2.684666240645881e-05,
+      "loss": 0.4039,
       "step": 68305
     },
     {
-      "epoch": 2.4,
-      "learning_rate": 2.778736373207056e-05,
-      "loss": 0.2656,
+      "epoch": 2.4618877716509893,
+      "grad_norm": 0.1901131123304367,
+      "learning_rate": 2.6843752232170095e-05,
+      "loss": 0.3888,
       "step": 68310
     },
     {
-      "epoch": 2.4,
-      "learning_rate": 2.7784532754315812e-05,
-      "loss": 0.2581,
+      "epoch": 2.462067971312214,
+      "grad_norm": 0.1914856880903244,
+      "learning_rate": 2.68408420327606e-05,
+      "loss": 0.3938,
       "step": 68315
     },
     {
-      "epoch": 2.4,
-      "learning_rate": 2.778170174040574e-05,
-      "loss": 0.2646,
+      "epoch": 2.4622481709734387,
+      "grad_norm": 0.23871468007564545,
+      "learning_rate": 2.6837931808270006e-05,
+      "loss": 0.3918,
       "step": 68320
     },
     {
-      "epoch": 2.4,
-      "learning_rate": 2.7778870690377108e-05,
-      "loss": 0.2627,
+      "epoch": 2.462428370634663,
+      "grad_norm": 0.1966165006160736,
+      "learning_rate": 2.6835021558737932e-05,
+      "loss": 0.4099,
       "step": 68325
     },
     {
-      "epoch": 2.4,
-      "learning_rate": 2.7776039604266667e-05,
-      "loss": 0.265,
+      "epoch": 2.4626085702958878,
+      "grad_norm": 0.224965900182724,
+      "learning_rate": 2.6832111284204054e-05,
+      "loss": 0.4068,
       "step": 68330
     },
     {
-      "epoch": 2.4,
-      "learning_rate": 2.7773208482111186e-05,
-      "loss": 0.2562,
+      "epoch": 2.4627887699571125,
+      "grad_norm": 0.20366400480270386,
+      "learning_rate": 2.682920098470801e-05,
+      "loss": 0.3795,
       "step": 68335
     },
     {
-      "epoch": 2.4,
-      "learning_rate": 2.777037732394741e-05,
-      "loss": 0.2778,
+      "epoch": 2.4629689696183372,
+      "grad_norm": 0.20222342014312744,
+      "learning_rate": 2.682629066028946e-05,
+      "loss": 0.3814,
       "step": 68340
     },
     {
-      "epoch": 2.4,
-      "learning_rate": 2.7767546129812123e-05,
-      "loss": 0.2595,
+      "epoch": 2.4631491692795615,
+      "grad_norm": 0.21054887771606445,
+      "learning_rate": 2.682338031098805e-05,
+      "loss": 0.4116,
       "step": 68345
     },
     {
-      "epoch": 2.4,
-      "learning_rate": 2.7764714899742063e-05,
-      "loss": 0.2597,
+      "epoch": 2.4633293689407862,
+      "grad_norm": 0.23802423477172852,
+      "learning_rate": 2.6820469936843434e-05,
+      "loss": 0.4045,
       "step": 68350
     },
     {
-      "epoch": 2.4,
-      "learning_rate": 2.776188363377401e-05,
-      "loss": 0.2698,
+      "epoch": 2.463509568602011,
+      "grad_norm": 0.2170737087726593,
+      "learning_rate": 2.6817559537895276e-05,
+      "loss": 0.3764,
       "step": 68355
     },
     {
-      "epoch": 2.41,
-      "learning_rate": 2.7759052331944713e-05,
-      "loss": 0.3009,
+      "epoch": 2.4636897682632357,
+      "grad_norm": 0.2287246584892273,
+      "learning_rate": 2.6814649114183216e-05,
+      "loss": 0.4225,
       "step": 68360
     },
     {
-      "epoch": 2.41,
-      "learning_rate": 2.7756220994290938e-05,
-      "loss": 0.2374,
+      "epoch": 2.4638699679244604,
+      "grad_norm": 0.17304402589797974,
+      "learning_rate": 2.681173866574691e-05,
+      "loss": 0.3751,
       "step": 68365
     },
     {
-      "epoch": 2.41,
-      "learning_rate": 2.7753389620849452e-05,
-      "loss": 0.2609,
+      "epoch": 2.464050167585685,
+      "grad_norm": 0.22758109867572784,
+      "learning_rate": 2.6808828192626022e-05,
+      "loss": 0.4037,
       "step": 68370
     },
     {
-      "epoch": 2.41,
-      "learning_rate": 2.775055821165702e-05,
-      "loss": 0.2568,
+      "epoch": 2.4642303672469095,
+      "grad_norm": 0.22057542204856873,
+      "learning_rate": 2.6805917694860195e-05,
+      "loss": 0.4412,
       "step": 68375
     },
     {
-      "epoch": 2.41,
-      "learning_rate": 2.7747726766750405e-05,
-      "loss": 0.2988,
+      "epoch": 2.464410566908134,
+      "grad_norm": 0.22053518891334534,
+      "learning_rate": 2.680300717248909e-05,
+      "loss": 0.3812,
       "step": 68380
     },
     {
-      "epoch": 2.41,
-      "learning_rate": 2.7744895286166357e-05,
-      "loss": 0.2684,
+      "epoch": 2.464590766569359,
+      "grad_norm": 0.21665138006210327,
+      "learning_rate": 2.6800096625552368e-05,
+      "loss": 0.3685,
       "step": 68385
     },
     {
-      "epoch": 2.41,
-      "learning_rate": 2.7742063769941662e-05,
-      "loss": 0.2592,
+      "epoch": 2.4647709662305837,
+      "grad_norm": 0.1874067336320877,
+      "learning_rate": 2.6797186054089674e-05,
+      "loss": 0.4096,
       "step": 68390
     },
     {
-      "epoch": 2.41,
-      "learning_rate": 2.7739232218113076e-05,
-      "loss": 0.2616,
+      "epoch": 2.464951165891808,
+      "grad_norm": 0.2060876190662384,
+      "learning_rate": 2.679427545814066e-05,
+      "loss": 0.3738,
       "step": 68395
     },
     {
-      "epoch": 2.41,
-      "learning_rate": 2.773640063071736e-05,
-      "loss": 0.2722,
+      "epoch": 2.4651313655530327,
+      "grad_norm": 0.2187156081199646,
+      "learning_rate": 2.6791364837744998e-05,
+      "loss": 0.4185,
       "step": 68400
     },
     {
-      "epoch": 2.41,
-      "learning_rate": 2.7733569007791284e-05,
-      "loss": 0.262,
+      "epoch": 2.4653115652142574,
+      "grad_norm": 0.214223712682724,
+      "learning_rate": 2.6788454192942342e-05,
+      "loss": 0.4253,
       "step": 68405
     },
     {
-      "epoch": 2.41,
-      "learning_rate": 2.773073734937162e-05,
-      "loss": 0.276,
+      "epoch": 2.465491764875482,
+      "grad_norm": 0.1790245622396469,
+      "learning_rate": 2.6785543523772334e-05,
+      "loss": 0.4309,
       "step": 68410
     },
     {
-      "epoch": 2.41,
-      "learning_rate": 2.7727905655495135e-05,
-      "loss": 0.2585,
+      "epoch": 2.465671964536707,
+      "grad_norm": 0.20714068412780762,
+      "learning_rate": 2.6782632830274645e-05,
+      "loss": 0.4007,
       "step": 68415
     },
     {
-      "epoch": 2.41,
-      "learning_rate": 2.7725073926198587e-05,
-      "loss": 0.2751,
+      "epoch": 2.465852164197931,
+      "grad_norm": 0.21486568450927734,
+      "learning_rate": 2.6779722112488924e-05,
+      "loss": 0.3859,
       "step": 68420
     },
     {
-      "epoch": 2.41,
-      "learning_rate": 2.7722242161518746e-05,
-      "loss": 0.2607,
+      "epoch": 2.466032363859156,
+      "grad_norm": 0.24496997892856598,
+      "learning_rate": 2.6776811370454848e-05,
+      "loss": 0.3808,
       "step": 68425
     },
     {
-      "epoch": 2.41,
-      "learning_rate": 2.7719410361492393e-05,
-      "loss": 0.2654,
+      "epoch": 2.4662125635203807,
+      "grad_norm": 0.24687924981117249,
+      "learning_rate": 2.677390060421204e-05,
+      "loss": 0.3642,
       "step": 68430
     },
     {
-      "epoch": 2.41,
-      "learning_rate": 2.7716578526156285e-05,
-      "loss": 0.2693,
+      "epoch": 2.4663927631816054,
+      "grad_norm": 0.17569184303283691,
+      "learning_rate": 2.6770989813800197e-05,
+      "loss": 0.4155,
       "step": 68435
     },
     {
-      "epoch": 2.41,
-      "learning_rate": 2.7713746655547202e-05,
-      "loss": 0.2475,
+      "epoch": 2.4665729628428297,
+      "grad_norm": 0.2444978952407837,
+      "learning_rate": 2.676807899925895e-05,
+      "loss": 0.4116,
       "step": 68440
     },
     {
-      "epoch": 2.41,
-      "learning_rate": 2.7710914749701893e-05,
-      "loss": 0.2863,
+      "epoch": 2.4667531625040544,
+      "grad_norm": 0.16109678149223328,
+      "learning_rate": 2.676516816062798e-05,
+      "loss": 0.3992,
       "step": 68445
     },
     {
-      "epoch": 2.41,
-      "learning_rate": 2.7708082808657148e-05,
-      "loss": 0.2811,
+      "epoch": 2.466933362165279,
+      "grad_norm": 0.20728158950805664,
+      "learning_rate": 2.676225729794693e-05,
+      "loss": 0.3597,
       "step": 68450
     },
     {
-      "epoch": 2.41,
-      "learning_rate": 2.7705250832449737e-05,
-      "loss": 0.2761,
+      "epoch": 2.467113561826504,
+      "grad_norm": 0.2165517657995224,
+      "learning_rate": 2.675934641125546e-05,
+      "loss": 0.3792,
       "step": 68455
     },
     {
-      "epoch": 2.41,
-      "learning_rate": 2.7702418821116426e-05,
-      "loss": 0.27,
+      "epoch": 2.4672937614877286,
+      "grad_norm": 0.274471253156662,
+      "learning_rate": 2.6756435500593242e-05,
+      "loss": 0.4132,
       "step": 68460
     },
     {
-      "epoch": 2.41,
-      "learning_rate": 2.7699586774693986e-05,
-      "loss": 0.2699,
+      "epoch": 2.467473961148953,
+      "grad_norm": 0.23226816952228546,
+      "learning_rate": 2.675352456599993e-05,
+      "loss": 0.4197,
       "step": 68465
     },
     {
-      "epoch": 2.41,
-      "learning_rate": 2.7696754693219197e-05,
-      "loss": 0.2681,
+      "epoch": 2.4676541608101776,
+      "grad_norm": 0.24921727180480957,
+      "learning_rate": 2.6750613607515184e-05,
+      "loss": 0.4038,
       "step": 68470
     },
     {
-      "epoch": 2.41,
-      "learning_rate": 2.7693922576728816e-05,
-      "loss": 0.2782,
+      "epoch": 2.4678343604714024,
+      "grad_norm": 0.17365936934947968,
+      "learning_rate": 2.6747702625178667e-05,
+      "loss": 0.4263,
       "step": 68475
     },
     {
-      "epoch": 2.41,
-      "learning_rate": 2.769109042525962e-05,
-      "loss": 0.2844,
+      "epoch": 2.468014560132627,
+      "grad_norm": 0.24769355356693268,
+      "learning_rate": 2.6744791619030042e-05,
+      "loss": 0.3998,
       "step": 68480
     },
     {
-      "epoch": 2.41,
-      "learning_rate": 2.7688258238848403e-05,
-      "loss": 0.2863,
+      "epoch": 2.4681947597938514,
+      "grad_norm": 0.24163182079792023,
+      "learning_rate": 2.674188058910897e-05,
+      "loss": 0.3979,
       "step": 68485
     },
     {
-      "epoch": 2.41,
-      "learning_rate": 2.7685426017531917e-05,
-      "loss": 0.2722,
+      "epoch": 2.468374959455076,
+      "grad_norm": 0.21235689520835876,
+      "learning_rate": 2.6738969535455104e-05,
+      "loss": 0.3893,
       "step": 68490
     },
     {
-      "epoch": 2.41,
-      "learning_rate": 2.7682593761346948e-05,
-      "loss": 0.2669,
+      "epoch": 2.468555159116301,
+      "grad_norm": 0.23081441223621368,
+      "learning_rate": 2.673605845810812e-05,
+      "loss": 0.3736,
       "step": 68495
     },
     {
-      "epoch": 2.41,
-      "learning_rate": 2.7679761470330262e-05,
-      "loss": 0.2725,
+      "epoch": 2.4687353587775256,
+      "grad_norm": 0.2072082906961441,
+      "learning_rate": 2.6733147357107673e-05,
+      "loss": 0.4183,
       "step": 68500
     },
     {
-      "epoch": 2.41,
-      "eval_loss": 0.26537927985191345,
-      "eval_runtime": 10.5494,
-      "eval_samples_per_second": 9.479,
-      "eval_steps_per_second": 9.479,
+      "epoch": 2.4687353587775256,
+      "eval_loss": 0.43393146991729736,
+      "eval_runtime": 3.5293,
+      "eval_samples_per_second": 28.334,
+      "eval_steps_per_second": 7.084,
       "step": 68500
     },
     {
-      "epoch": 2.41,
-      "learning_rate": 2.7676929144518642e-05,
-      "loss": 0.2678,
+      "epoch": 2.4689155584387503,
+      "grad_norm": 0.245635524392128,
+      "learning_rate": 2.673023623249344e-05,
+      "loss": 0.3877,
       "step": 68505
     },
     {
-      "epoch": 2.41,
-      "learning_rate": 2.767409678394886e-05,
-      "loss": 0.2775,
+      "epoch": 2.4690957580999746,
+      "grad_norm": 0.21268950402736664,
+      "learning_rate": 2.672732508430506e-05,
+      "loss": 0.3474,
       "step": 68510
     },
     {
-      "epoch": 2.41,
-      "learning_rate": 2.76712643886577e-05,
-      "loss": 0.2876,
+      "epoch": 2.4692759577611993,
+      "grad_norm": 0.22742435336112976,
+      "learning_rate": 2.6724413912582208e-05,
+      "loss": 0.3821,
       "step": 68515
     },
     {
-      "epoch": 2.41,
-      "learning_rate": 2.7668431958681924e-05,
-      "loss": 0.2665,
+      "epoch": 2.469456157422424,
+      "grad_norm": 0.26614660024642944,
+      "learning_rate": 2.672150271736456e-05,
+      "loss": 0.3755,
       "step": 68520
     },
     {
-      "epoch": 2.41,
-      "learning_rate": 2.7665599494058325e-05,
-      "loss": 0.2805,
+      "epoch": 2.469636357083649,
+      "grad_norm": 0.20712517201900482,
+      "learning_rate": 2.671859149869177e-05,
+      "loss": 0.4151,
       "step": 68525
     },
     {
-      "epoch": 2.41,
-      "learning_rate": 2.7662766994823665e-05,
-      "loss": 0.2676,
+      "epoch": 2.469816556744873,
+      "grad_norm": 0.16104906797409058,
+      "learning_rate": 2.6715680256603503e-05,
+      "loss": 0.3874,
       "step": 68530
     },
     {
-      "epoch": 2.41,
-      "learning_rate": 2.765993446101473e-05,
-      "loss": 0.2759,
+      "epoch": 2.469996756406098,
+      "grad_norm": 0.20770768821239471,
+      "learning_rate": 2.671276899113942e-05,
+      "loss": 0.4159,
       "step": 68535
     },
     {
-      "epoch": 2.41,
-      "learning_rate": 2.7657101892668303e-05,
-      "loss": 0.2572,
+      "epoch": 2.4701769560673226,
+      "grad_norm": 0.16630055010318756,
+      "learning_rate": 2.670985770233919e-05,
+      "loss": 0.4349,
       "step": 68540
     },
     {
-      "epoch": 2.41,
-      "learning_rate": 2.7654269289821165e-05,
-      "loss": 0.2697,
+      "epoch": 2.4703571557285473,
+      "grad_norm": 0.18788853287696838,
+      "learning_rate": 2.6706946390242488e-05,
+      "loss": 0.3773,
       "step": 68545
     },
     {
-      "epoch": 2.41,
-      "learning_rate": 2.7651436652510082e-05,
-      "loss": 0.2701,
+      "epoch": 2.470537355389772,
+      "grad_norm": 0.24081262946128845,
+      "learning_rate": 2.6704035054888964e-05,
+      "loss": 0.3926,
       "step": 68550
     },
     {
-      "epoch": 2.41,
-      "learning_rate": 2.7648603980771846e-05,
-      "loss": 0.2857,
+      "epoch": 2.4707175550509963,
+      "grad_norm": 0.21271935105323792,
+      "learning_rate": 2.67011236963183e-05,
+      "loss": 0.3867,
       "step": 68555
     },
     {
-      "epoch": 2.41,
-      "learning_rate": 2.7645771274643223e-05,
-      "loss": 0.2857,
+      "epoch": 2.470897754712221,
+      "grad_norm": 0.19515545666217804,
+      "learning_rate": 2.6698212314570148e-05,
+      "loss": 0.3917,
       "step": 68560
     },
     {
-      "epoch": 2.41,
-      "learning_rate": 2.7642938534161016e-05,
-      "loss": 0.2729,
+      "epoch": 2.471077954373446,
+      "grad_norm": 0.18669338524341583,
+      "learning_rate": 2.669530090968419e-05,
+      "loss": 0.3656,
       "step": 68565
     },
     {
-      "epoch": 2.41,
-      "learning_rate": 2.7640105759361985e-05,
-      "loss": 0.279,
+      "epoch": 2.4712581540346705,
+      "grad_norm": 0.19376404583454132,
+      "learning_rate": 2.669238948170008e-05,
+      "loss": 0.3925,
       "step": 68570
     },
     {
-      "epoch": 2.41,
-      "learning_rate": 2.7637272950282927e-05,
-      "loss": 0.2692,
+      "epoch": 2.471438353695895,
+      "grad_norm": 0.22779560089111328,
+      "learning_rate": 2.6689478030657496e-05,
+      "loss": 0.4522,
       "step": 68575
     },
     {
-      "epoch": 2.41,
-      "learning_rate": 2.7634440106960614e-05,
-      "loss": 0.2535,
+      "epoch": 2.4716185533571196,
+      "grad_norm": 0.19855950772762299,
+      "learning_rate": 2.6686566556596087e-05,
+      "loss": 0.3975,
       "step": 68580
     },
     {
-      "epoch": 2.41,
-      "learning_rate": 2.7631607229431834e-05,
-      "loss": 0.2822,
+      "epoch": 2.4717987530183443,
+      "grad_norm": 0.2209222912788391,
+      "learning_rate": 2.6683655059555544e-05,
+      "loss": 0.3871,
       "step": 68585
     },
     {
-      "epoch": 2.41,
-      "learning_rate": 2.7628774317733362e-05,
-      "loss": 0.2818,
+      "epoch": 2.471978952679569,
+      "grad_norm": 0.1904340535402298,
+      "learning_rate": 2.668074353957553e-05,
+      "loss": 0.3682,
       "step": 68590
     },
     {
-      "epoch": 2.41,
-      "learning_rate": 2.7625941371901997e-05,
-      "loss": 0.2645,
+      "epoch": 2.4721591523407938,
+      "grad_norm": 0.2192251831293106,
+      "learning_rate": 2.6677831996695706e-05,
+      "loss": 0.3725,
       "step": 68595
     },
     {
-      "epoch": 2.41,
-      "learning_rate": 2.762310839197451e-05,
-      "loss": 0.2604,
+      "epoch": 2.472339352002018,
+      "grad_norm": 0.20693185925483704,
+      "learning_rate": 2.6674920430955753e-05,
+      "loss": 0.3858,
       "step": 68600
     },
     {
-      "epoch": 2.41,
-      "learning_rate": 2.7620275377987686e-05,
-      "loss": 0.2736,
+      "epoch": 2.4725195516632428,
+      "grad_norm": 0.25332531332969666,
+      "learning_rate": 2.6672008842395323e-05,
+      "loss": 0.3742,
       "step": 68605
     },
     {
-      "epoch": 2.41,
-      "learning_rate": 2.7617442329978323e-05,
-      "loss": 0.2814,
+      "epoch": 2.4726997513244675,
+      "grad_norm": 0.2141808420419693,
+      "learning_rate": 2.666909723105411e-05,
+      "loss": 0.4053,
       "step": 68610
     },
     {
-      "epoch": 2.41,
-      "learning_rate": 2.7614609247983193e-05,
-      "loss": 0.2776,
+      "epoch": 2.4728799509856922,
+      "grad_norm": 0.20601968467235565,
+      "learning_rate": 2.6666185596971753e-05,
+      "loss": 0.3986,
       "step": 68615
     },
     {
-      "epoch": 2.41,
-      "learning_rate": 2.7611776132039074e-05,
-      "loss": 0.2643,
+      "epoch": 2.473060150646917,
+      "grad_norm": 0.21490851044654846,
+      "learning_rate": 2.6663273940187956e-05,
+      "loss": 0.3877,
       "step": 68620
     },
     {
-      "epoch": 2.41,
-      "learning_rate": 2.7608942982182774e-05,
-      "loss": 0.2721,
+      "epoch": 2.4732403503081413,
+      "grad_norm": 0.1867503970861435,
+      "learning_rate": 2.666036226074236e-05,
+      "loss": 0.3738,
       "step": 68625
     },
     {
-      "epoch": 2.41,
-      "learning_rate": 2.760610979845107e-05,
-      "loss": 0.2682,
+      "epoch": 2.473420549969366,
+      "grad_norm": 0.21378225088119507,
+      "learning_rate": 2.6657450558674658e-05,
+      "loss": 0.3711,
       "step": 68630
     },
     {
-      "epoch": 2.41,
-      "learning_rate": 2.760327658088074e-05,
-      "loss": 0.2631,
+      "epoch": 2.4736007496305907,
+      "grad_norm": 0.20687870681285858,
+      "learning_rate": 2.665453883402451e-05,
+      "loss": 0.4117,
       "step": 68635
     },
     {
-      "epoch": 2.41,
-      "learning_rate": 2.7600443329508578e-05,
-      "loss": 0.2766,
+      "epoch": 2.4737809492918155,
+      "grad_norm": 0.17772191762924194,
+      "learning_rate": 2.665162708683159e-05,
+      "loss": 0.3953,
       "step": 68640
     },
     {
-      "epoch": 2.42,
-      "learning_rate": 2.759761004437138e-05,
-      "loss": 0.2757,
+      "epoch": 2.47396114895304,
+      "grad_norm": 0.19218339025974274,
+      "learning_rate": 2.6648715317135575e-05,
+      "loss": 0.3936,
       "step": 68645
     },
     {
-      "epoch": 2.42,
-      "learning_rate": 2.759477672550593e-05,
-      "loss": 0.2499,
+      "epoch": 2.4741413486142645,
+      "grad_norm": 0.2298266589641571,
+      "learning_rate": 2.6645803524976133e-05,
+      "loss": 0.3804,
       "step": 68650
     },
     {
-      "epoch": 2.42,
-      "learning_rate": 2.759194337294901e-05,
-      "loss": 0.2766,
+      "epoch": 2.4743215482754892,
+      "grad_norm": 0.1857258528470993,
+      "learning_rate": 2.664289171039293e-05,
+      "loss": 0.3942,
       "step": 68655
     },
     {
-      "epoch": 2.42,
-      "learning_rate": 2.7589109986737416e-05,
-      "loss": 0.2666,
+      "epoch": 2.474501747936714,
+      "grad_norm": 0.1851481795310974,
+      "learning_rate": 2.6639979873425652e-05,
+      "loss": 0.4023,
       "step": 68660
     },
     {
-      "epoch": 2.42,
-      "learning_rate": 2.7586276566907925e-05,
-      "loss": 0.2593,
+      "epoch": 2.4746819475979387,
+      "grad_norm": 0.19881199300289154,
+      "learning_rate": 2.663706801411396e-05,
+      "loss": 0.4297,
       "step": 68665
     },
     {
-      "epoch": 2.42,
-      "learning_rate": 2.7583443113497348e-05,
-      "loss": 0.264,
+      "epoch": 2.474862147259163,
+      "grad_norm": 0.22843848168849945,
+      "learning_rate": 2.6634156132497538e-05,
+      "loss": 0.3793,
       "step": 68670
     },
     {
-      "epoch": 2.42,
-      "learning_rate": 2.7580609626542457e-05,
-      "loss": 0.284,
+      "epoch": 2.4750423469203877,
+      "grad_norm": 0.20685456693172455,
+      "learning_rate": 2.6631244228616053e-05,
+      "loss": 0.3826,
       "step": 68675
     },
     {
-      "epoch": 2.42,
-      "learning_rate": 2.757777610608006e-05,
-      "loss": 0.2797,
+      "epoch": 2.4752225465816124,
+      "grad_norm": 0.24268147349357605,
+      "learning_rate": 2.6628332302509186e-05,
+      "loss": 0.3882,
       "step": 68680
     },
     {
-      "epoch": 2.42,
-      "learning_rate": 2.757494255214693e-05,
-      "loss": 0.2877,
+      "epoch": 2.475402746242837,
+      "grad_norm": 0.21436165273189545,
+      "learning_rate": 2.6625420354216597e-05,
+      "loss": 0.4387,
       "step": 68685
     },
     {
-      "epoch": 2.42,
-      "learning_rate": 2.7572108964779873e-05,
-      "loss": 0.2632,
+      "epoch": 2.475582945904062,
+      "grad_norm": 0.20121796429157257,
+      "learning_rate": 2.6622508383777977e-05,
+      "loss": 0.4059,
       "step": 68690
     },
     {
-      "epoch": 2.42,
-      "learning_rate": 2.7569275344015678e-05,
-      "loss": 0.2612,
+      "epoch": 2.475763145565286,
+      "grad_norm": 0.1770712435245514,
+      "learning_rate": 2.6619596391233e-05,
+      "loss": 0.3906,
       "step": 68695
     },
     {
-      "epoch": 2.42,
-      "learning_rate": 2.756644168989113e-05,
-      "loss": 0.2647,
+      "epoch": 2.475943345226511,
+      "grad_norm": 0.2208649069070816,
+      "learning_rate": 2.661668437662132e-05,
+      "loss": 0.4162,
       "step": 68700
     },
     {
-      "epoch": 2.42,
-      "learning_rate": 2.756360800244303e-05,
-      "loss": 0.2814,
+      "epoch": 2.4761235448877357,
+      "grad_norm": 0.23539263010025024,
+      "learning_rate": 2.661377233998264e-05,
+      "loss": 0.4129,
       "step": 68705
     },
     {
-      "epoch": 2.42,
-      "learning_rate": 2.7560774281708174e-05,
-      "loss": 0.2855,
+      "epoch": 2.4763037445489604,
+      "grad_norm": 0.18581153452396393,
+      "learning_rate": 2.6610860281356627e-05,
+      "loss": 0.4237,
       "step": 68710
     },
     {
-      "epoch": 2.42,
-      "learning_rate": 2.7557940527723348e-05,
-      "loss": 0.2755,
+      "epoch": 2.4764839442101847,
+      "grad_norm": 0.20653104782104492,
+      "learning_rate": 2.6607948200782944e-05,
+      "loss": 0.3858,
       "step": 68715
     },
     {
-      "epoch": 2.42,
-      "learning_rate": 2.755510674052535e-05,
-      "loss": 0.2472,
+      "epoch": 2.4766641438714094,
+      "grad_norm": 0.18141742050647736,
+      "learning_rate": 2.6605036098301283e-05,
+      "loss": 0.3642,
       "step": 68720
     },
     {
-      "epoch": 2.42,
-      "learning_rate": 2.7552272920150966e-05,
-      "loss": 0.2665,
+      "epoch": 2.476844343532634,
+      "grad_norm": 0.22775548696517944,
+      "learning_rate": 2.6602123973951314e-05,
+      "loss": 0.4237,
       "step": 68725
     },
     {
-      "epoch": 2.42,
-      "learning_rate": 2.754943906663701e-05,
-      "loss": 0.2592,
+      "epoch": 2.477024543193859,
+      "grad_norm": 0.1954440176486969,
+      "learning_rate": 2.6599211827772724e-05,
+      "loss": 0.403,
       "step": 68730
     },
     {
-      "epoch": 2.42,
-      "learning_rate": 2.7546605180020273e-05,
-      "loss": 0.2856,
+      "epoch": 2.4772047428550836,
+      "grad_norm": 0.18037401139736176,
+      "learning_rate": 2.6596299659805173e-05,
+      "loss": 0.3839,
       "step": 68735
     },
     {
-      "epoch": 2.42,
-      "learning_rate": 2.7543771260337538e-05,
-      "loss": 0.263,
+      "epoch": 2.477384942516308,
+      "grad_norm": 0.2156188189983368,
+      "learning_rate": 2.6593387470088354e-05,
+      "loss": 0.4249,
       "step": 68740
     },
     {
-      "epoch": 2.42,
-      "learning_rate": 2.7540937307625613e-05,
-      "loss": 0.2834,
+      "epoch": 2.4775651421775327,
+      "grad_norm": 0.2691105008125305,
+      "learning_rate": 2.6590475258661935e-05,
+      "loss": 0.4494,
       "step": 68745
     },
     {
-      "epoch": 2.42,
-      "learning_rate": 2.7538103321921287e-05,
-      "loss": 0.2778,
+      "epoch": 2.4777453418387574,
+      "grad_norm": 0.1580956131219864,
+      "learning_rate": 2.6587563025565604e-05,
+      "loss": 0.3839,
       "step": 68750
     },
     {
-      "epoch": 2.42,
-      "learning_rate": 2.753526930326137e-05,
-      "loss": 0.2541,
+      "epoch": 2.477925541499982,
+      "grad_norm": 0.17688532173633575,
+      "learning_rate": 2.6584650770839026e-05,
+      "loss": 0.4133,
       "step": 68755
     },
     {
-      "epoch": 2.42,
-      "learning_rate": 2.7532435251682643e-05,
-      "loss": 0.2764,
+      "epoch": 2.4781057411612064,
+      "grad_norm": 0.19186891615390778,
+      "learning_rate": 2.6581738494521898e-05,
+      "loss": 0.3996,
       "step": 68760
     },
     {
-      "epoch": 2.42,
-      "learning_rate": 2.752960116722192e-05,
-      "loss": 0.2536,
+      "epoch": 2.478285940822431,
+      "grad_norm": 0.23146666586399078,
+      "learning_rate": 2.657882619665388e-05,
+      "loss": 0.3699,
       "step": 68765
     },
     {
-      "epoch": 2.42,
-      "learning_rate": 2.7526767049915992e-05,
-      "loss": 0.2796,
+      "epoch": 2.478466140483656,
+      "grad_norm": 0.2393793761730194,
+      "learning_rate": 2.6575913877274666e-05,
+      "loss": 0.3812,
       "step": 68770
     },
     {
-      "epoch": 2.42,
-      "learning_rate": 2.7523932899801657e-05,
-      "loss": 0.2645,
+      "epoch": 2.4786463401448806,
+      "grad_norm": 0.18306025862693787,
+      "learning_rate": 2.657300153642393e-05,
+      "loss": 0.4051,
       "step": 68775
     },
     {
-      "epoch": 2.42,
-      "learning_rate": 2.752109871691571e-05,
-      "loss": 0.2534,
+      "epoch": 2.4788265398061053,
+      "grad_norm": 0.22833152115345,
+      "learning_rate": 2.657008917414135e-05,
+      "loss": 0.3908,
       "step": 68780
     },
     {
-      "epoch": 2.42,
-      "learning_rate": 2.7518264501294964e-05,
-      "loss": 0.2548,
+      "epoch": 2.4790067394673296,
+      "grad_norm": 0.19252647459506989,
+      "learning_rate": 2.6567176790466613e-05,
+      "loss": 0.3735,
       "step": 68785
     },
     {
-      "epoch": 2.42,
-      "learning_rate": 2.7515430252976215e-05,
-      "loss": 0.2627,
+      "epoch": 2.4791869391285544,
+      "grad_norm": 0.202143132686615,
+      "learning_rate": 2.6564264385439385e-05,
+      "loss": 0.411,
       "step": 68790
     },
     {
-      "epoch": 2.42,
-      "learning_rate": 2.7512595971996265e-05,
-      "loss": 0.2931,
+      "epoch": 2.479367138789779,
+      "grad_norm": 0.1779770851135254,
+      "learning_rate": 2.656135195909938e-05,
+      "loss": 0.3841,
       "step": 68795
     },
     {
-      "epoch": 2.42,
-      "learning_rate": 2.750976165839191e-05,
-      "loss": 0.2687,
+      "epoch": 2.479547338451004,
+      "grad_norm": 0.17505523562431335,
+      "learning_rate": 2.6558439511486232e-05,
+      "loss": 0.3515,
       "step": 68800
     },
     {
-      "epoch": 2.42,
-      "learning_rate": 2.750692731219995e-05,
-      "loss": 0.2772,
+      "epoch": 2.479727538112228,
+      "grad_norm": 0.2236313372850418,
+      "learning_rate": 2.655552704263966e-05,
+      "loss": 0.4296,
       "step": 68805
     },
     {
-      "epoch": 2.42,
-      "learning_rate": 2.7504092933457187e-05,
-      "loss": 0.2637,
+      "epoch": 2.479907737773453,
+      "grad_norm": 0.21467982232570648,
+      "learning_rate": 2.655261455259933e-05,
+      "loss": 0.3938,
       "step": 68810
     },
     {
-      "epoch": 2.42,
-      "learning_rate": 2.7501258522200436e-05,
-      "loss": 0.2547,
+      "epoch": 2.4800879374346776,
+      "grad_norm": 0.1988895684480667,
+      "learning_rate": 2.6549702041404932e-05,
+      "loss": 0.4324,
       "step": 68815
     },
     {
-      "epoch": 2.42,
-      "learning_rate": 2.749842407846649e-05,
-      "loss": 0.2584,
+      "epoch": 2.4802681370959023,
+      "grad_norm": 0.17672614753246307,
+      "learning_rate": 2.6546789509096144e-05,
+      "loss": 0.4032,
       "step": 68820
     },
     {
-      "epoch": 2.42,
-      "learning_rate": 2.7495589602292154e-05,
-      "loss": 0.2928,
+      "epoch": 2.480448336757127,
+      "grad_norm": 0.22681747376918793,
+      "learning_rate": 2.6543876955712637e-05,
+      "loss": 0.4158,
       "step": 68825
     },
     {
-      "epoch": 2.42,
-      "learning_rate": 2.749275509371423e-05,
-      "loss": 0.2746,
+      "epoch": 2.4806285364183513,
+      "grad_norm": 0.16268977522850037,
+      "learning_rate": 2.654096438129412e-05,
+      "loss": 0.3774,
       "step": 68830
     },
     {
-      "epoch": 2.42,
-      "learning_rate": 2.748992055276952e-05,
-      "loss": 0.2673,
+      "epoch": 2.480808736079576,
+      "grad_norm": 0.18678313493728638,
+      "learning_rate": 2.6538051785880254e-05,
+      "loss": 0.3649,
       "step": 68835
     },
     {
-      "epoch": 2.42,
-      "learning_rate": 2.7487085979494838e-05,
-      "loss": 0.2751,
+      "epoch": 2.480988935740801,
+      "grad_norm": 0.2044302076101303,
+      "learning_rate": 2.6535139169510727e-05,
+      "loss": 0.3747,
       "step": 68840
     },
     {
-      "epoch": 2.42,
-      "learning_rate": 2.7484251373926977e-05,
-      "loss": 0.2683,
+      "epoch": 2.4811691354020255,
+      "grad_norm": 0.2173137664794922,
+      "learning_rate": 2.6532226532225235e-05,
+      "loss": 0.3782,
       "step": 68845
     },
     {
-      "epoch": 2.42,
-      "learning_rate": 2.748141673610276e-05,
-      "loss": 0.2614,
+      "epoch": 2.48134933506325,
+      "grad_norm": 0.22002023458480835,
+      "learning_rate": 2.6529313874063445e-05,
+      "loss": 0.3879,
       "step": 68850
     },
     {
-      "epoch": 2.42,
-      "learning_rate": 2.7478582066058974e-05,
-      "loss": 0.2732,
+      "epoch": 2.4815295347244746,
+      "grad_norm": 0.19581471383571625,
+      "learning_rate": 2.652640119506506e-05,
+      "loss": 0.3761,
       "step": 68855
     },
     {
-      "epoch": 2.42,
-      "learning_rate": 2.7475747363832432e-05,
-      "loss": 0.2437,
+      "epoch": 2.4817097343856993,
+      "grad_norm": 0.18339914083480835,
+      "learning_rate": 2.6523488495269744e-05,
+      "loss": 0.4203,
       "step": 68860
     },
     {
-      "epoch": 2.42,
-      "learning_rate": 2.7472912629459946e-05,
-      "loss": 0.265,
+      "epoch": 2.481889934046924,
+      "grad_norm": 0.19573746621608734,
+      "learning_rate": 2.6520575774717194e-05,
+      "loss": 0.4061,
       "step": 68865
     },
     {
-      "epoch": 2.42,
-      "learning_rate": 2.7470077862978322e-05,
-      "loss": 0.2429,
+      "epoch": 2.4820701337081488,
+      "grad_norm": 0.24082912504673004,
+      "learning_rate": 2.6517663033447092e-05,
+      "loss": 0.4076,
       "step": 68870
     },
     {
-      "epoch": 2.42,
-      "learning_rate": 2.7467243064424358e-05,
-      "loss": 0.2671,
+      "epoch": 2.4822503333693735,
+      "grad_norm": 0.1869727075099945,
+      "learning_rate": 2.6514750271499127e-05,
+      "loss": 0.4011,
       "step": 68875
     },
     {
-      "epoch": 2.42,
-      "learning_rate": 2.7464408233834876e-05,
-      "loss": 0.2554,
+      "epoch": 2.482430533030598,
+      "grad_norm": 0.19371429085731506,
+      "learning_rate": 2.6511837488912988e-05,
+      "loss": 0.3798,
       "step": 68880
     },
     {
-      "epoch": 2.42,
-      "learning_rate": 2.7461573371246674e-05,
-      "loss": 0.271,
+      "epoch": 2.4826107326918225,
+      "grad_norm": 0.20784878730773926,
+      "learning_rate": 2.650892468572835e-05,
+      "loss": 0.4097,
       "step": 68885
     },
     {
-      "epoch": 2.42,
-      "learning_rate": 2.745873847669656e-05,
-      "loss": 0.2693,
+      "epoch": 2.4827909323530473,
+      "grad_norm": 0.17571403086185455,
+      "learning_rate": 2.65060118619849e-05,
+      "loss": 0.3819,
       "step": 68890
     },
     {
-      "epoch": 2.42,
-      "learning_rate": 2.745590355022135e-05,
-      "loss": 0.2692,
+      "epoch": 2.482971132014272,
+      "grad_norm": 0.17903995513916016,
+      "learning_rate": 2.6503099017722343e-05,
+      "loss": 0.3743,
       "step": 68895
     },
     {
-      "epoch": 2.42,
-      "learning_rate": 2.7453068591857857e-05,
-      "loss": 0.2953,
+      "epoch": 2.4831513316754963,
+      "grad_norm": 0.23476238548755646,
+      "learning_rate": 2.650018615298035e-05,
+      "loss": 0.3742,
       "step": 68900
     },
     {
-      "epoch": 2.42,
-      "learning_rate": 2.7450233601642878e-05,
-      "loss": 0.2494,
+      "epoch": 2.483331531336721,
+      "grad_norm": 0.21859110891819,
+      "learning_rate": 2.6497273267798605e-05,
+      "loss": 0.3606,
       "step": 68905
     },
     {
-      "epoch": 2.42,
-      "learning_rate": 2.7447398579613233e-05,
-      "loss": 0.2435,
+      "epoch": 2.4835117309979458,
+      "grad_norm": 0.2518290579319,
+      "learning_rate": 2.6494360362216803e-05,
+      "loss": 0.4162,
       "step": 68910
     },
     {
-      "epoch": 2.42,
-      "learning_rate": 2.7444563525805728e-05,
-      "loss": 0.284,
+      "epoch": 2.4836919306591705,
+      "grad_norm": 0.1847134232521057,
+      "learning_rate": 2.6491447436274637e-05,
+      "loss": 0.3972,
       "step": 68915
     },
     {
-      "epoch": 2.42,
-      "learning_rate": 2.7441728440257174e-05,
-      "loss": 0.2486,
+      "epoch": 2.483872130320395,
+      "grad_norm": 0.18543338775634766,
+      "learning_rate": 2.648853449001178e-05,
+      "loss": 0.4026,
       "step": 68920
     },
     {
-      "epoch": 2.42,
-      "learning_rate": 2.743889332300439e-05,
-      "loss": 0.255,
+      "epoch": 2.4840523299816195,
+      "grad_norm": 0.20667828619480133,
+      "learning_rate": 2.648562152346793e-05,
+      "loss": 0.4239,
       "step": 68925
     },
     {
-      "epoch": 2.43,
-      "learning_rate": 2.7436058174084188e-05,
-      "loss": 0.2625,
+      "epoch": 2.4842325296428442,
+      "grad_norm": 0.21871261298656464,
+      "learning_rate": 2.6482708536682777e-05,
+      "loss": 0.4119,
       "step": 68930
     },
     {
-      "epoch": 2.43,
-      "learning_rate": 2.743322299353337e-05,
-      "loss": 0.276,
+      "epoch": 2.484412729304069,
+      "grad_norm": 0.18584778904914856,
+      "learning_rate": 2.647979552969601e-05,
+      "loss": 0.3987,
       "step": 68935
     },
     {
-      "epoch": 2.43,
-      "learning_rate": 2.7430387781388756e-05,
-      "loss": 0.2862,
+      "epoch": 2.4845929289652937,
+      "grad_norm": 0.1998424530029297,
+      "learning_rate": 2.6476882502547305e-05,
+      "loss": 0.3779,
       "step": 68940
     },
     {
-      "epoch": 2.43,
-      "learning_rate": 2.742755253768715e-05,
-      "loss": 0.2511,
+      "epoch": 2.484773128626518,
+      "grad_norm": 0.18726879358291626,
+      "learning_rate": 2.647396945527637e-05,
+      "loss": 0.4084,
       "step": 68945
     },
     {
-      "epoch": 2.43,
-      "learning_rate": 2.742471726246538e-05,
-      "loss": 0.2575,
+      "epoch": 2.4849533282877427,
+      "grad_norm": 0.19069020450115204,
+      "learning_rate": 2.6471056387922886e-05,
+      "loss": 0.3962,
       "step": 68950
     },
     {
-      "epoch": 2.43,
-      "learning_rate": 2.7421881955760265e-05,
-      "loss": 0.2964,
+      "epoch": 2.4851335279489675,
+      "grad_norm": 0.231498122215271,
+      "learning_rate": 2.6468143300526543e-05,
+      "loss": 0.3819,
       "step": 68955
     },
     {
-      "epoch": 2.43,
-      "learning_rate": 2.7419046617608597e-05,
-      "loss": 0.2514,
+      "epoch": 2.485313727610192,
+      "grad_norm": 0.19024336338043213,
+      "learning_rate": 2.6465230193127033e-05,
+      "loss": 0.3738,
       "step": 68960
     },
     {
-      "epoch": 2.43,
-      "learning_rate": 2.7416211248047212e-05,
-      "loss": 0.2529,
+      "epoch": 2.485493927271417,
+      "grad_norm": 0.23656608164310455,
+      "learning_rate": 2.6462317065764043e-05,
+      "loss": 0.4132,
       "step": 68965
     },
     {
-      "epoch": 2.43,
-      "learning_rate": 2.741337584711291e-05,
-      "loss": 0.2616,
+      "epoch": 2.4856741269326412,
+      "grad_norm": 0.2034633606672287,
+      "learning_rate": 2.645940391847727e-05,
+      "loss": 0.4149,
       "step": 68970
     },
     {
-      "epoch": 2.43,
-      "learning_rate": 2.741054041484251e-05,
-      "loss": 0.2552,
+      "epoch": 2.485854326593866,
+      "grad_norm": 0.20113669335842133,
+      "learning_rate": 2.6456490751306395e-05,
+      "loss": 0.3942,
       "step": 68975
     },
     {
-      "epoch": 2.43,
-      "learning_rate": 2.7407704951272833e-05,
-      "loss": 0.2736,
+      "epoch": 2.4860345262550907,
+      "grad_norm": 0.18040354549884796,
+      "learning_rate": 2.645357756429112e-05,
+      "loss": 0.403,
       "step": 68980
     },
     {
-      "epoch": 2.43,
-      "learning_rate": 2.7404869456440703e-05,
-      "loss": 0.2452,
+      "epoch": 2.4862147259163154,
+      "grad_norm": 0.2093953937292099,
+      "learning_rate": 2.645066435747113e-05,
+      "loss": 0.3853,
       "step": 68985
     },
     {
-      "epoch": 2.43,
-      "learning_rate": 2.740203393038292e-05,
-      "loss": 0.2571,
+      "epoch": 2.4863949255775397,
+      "grad_norm": 0.18795111775398254,
+      "learning_rate": 2.6447751130886117e-05,
+      "loss": 0.4241,
       "step": 68990
     },
     {
-      "epoch": 2.43,
-      "learning_rate": 2.739919837313631e-05,
-      "loss": 0.2656,
+      "epoch": 2.4865751252387644,
+      "grad_norm": 0.24407252669334412,
+      "learning_rate": 2.644483788457578e-05,
+      "loss": 0.3794,
       "step": 68995
     },
     {
-      "epoch": 2.43,
-      "learning_rate": 2.7396362784737682e-05,
-      "loss": 0.2927,
+      "epoch": 2.486755324899989,
+      "grad_norm": 0.18721212446689606,
+      "learning_rate": 2.6441924618579807e-05,
+      "loss": 0.3853,
       "step": 69000
     },
     {
-      "epoch": 2.43,
-      "eval_loss": 0.2651735842227936,
-      "eval_runtime": 10.5525,
-      "eval_samples_per_second": 9.476,
-      "eval_steps_per_second": 9.476,
+      "epoch": 2.486755324899989,
+      "eval_loss": 0.43400391936302185,
+      "eval_runtime": 3.5358,
+      "eval_samples_per_second": 28.282,
+      "eval_steps_per_second": 7.071,
       "step": 69000
     },
     {
-      "epoch": 2.43,
-      "learning_rate": 2.7393527165223875e-05,
-      "loss": 0.2494,
+      "epoch": 2.486935524561214,
+      "grad_norm": 0.21904613077640533,
+      "learning_rate": 2.643901133293789e-05,
+      "loss": 0.4267,
       "step": 69005
     },
     {
-      "epoch": 2.43,
-      "learning_rate": 2.7390691514631693e-05,
-      "loss": 0.2765,
+      "epoch": 2.4871157242224387,
+      "grad_norm": 0.18053406476974487,
+      "learning_rate": 2.6436098027689714e-05,
+      "loss": 0.3885,
       "step": 69010
     },
     {
-      "epoch": 2.43,
-      "learning_rate": 2.738785583299796e-05,
-      "loss": 0.2723,
+      "epoch": 2.487295923883663,
+      "grad_norm": 0.18730615079402924,
+      "learning_rate": 2.6433184702874993e-05,
+      "loss": 0.371,
       "step": 69015
     },
     {
-      "epoch": 2.43,
-      "learning_rate": 2.7385020120359483e-05,
-      "loss": 0.2716,
+      "epoch": 2.4874761235448877,
+      "grad_norm": 0.21903234720230103,
+      "learning_rate": 2.6430271358533397e-05,
+      "loss": 0.4299,
       "step": 69020
     },
     {
-      "epoch": 2.43,
-      "learning_rate": 2.7382184376753096e-05,
-      "loss": 0.2638,
+      "epoch": 2.4876563232061124,
+      "grad_norm": 0.21092742681503296,
+      "learning_rate": 2.6427357994704633e-05,
+      "loss": 0.4174,
       "step": 69025
     },
     {
-      "epoch": 2.43,
-      "learning_rate": 2.7379348602215616e-05,
-      "loss": 0.2854,
+      "epoch": 2.487836522867337,
+      "grad_norm": 0.20386511087417603,
+      "learning_rate": 2.6424444611428396e-05,
+      "loss": 0.4047,
       "step": 69030
     },
     {
-      "epoch": 2.43,
-      "learning_rate": 2.7376512796783867e-05,
-      "loss": 0.263,
+      "epoch": 2.4880167225285614,
+      "grad_norm": 0.2634836733341217,
+      "learning_rate": 2.642153120874437e-05,
+      "loss": 0.424,
       "step": 69035
     },
     {
-      "epoch": 2.43,
-      "learning_rate": 2.737367696049466e-05,
-      "loss": 0.2827,
+      "epoch": 2.488196922189786,
+      "grad_norm": 0.22535258531570435,
+      "learning_rate": 2.6418617786692273e-05,
+      "loss": 0.4279,
       "step": 69040
     },
     {
-      "epoch": 2.43,
-      "learning_rate": 2.7370841093384824e-05,
-      "loss": 0.295,
+      "epoch": 2.488377121851011,
+      "grad_norm": 0.19794537127017975,
+      "learning_rate": 2.6415704345311764e-05,
+      "loss": 0.3665,
       "step": 69045
     },
     {
-      "epoch": 2.43,
-      "learning_rate": 2.7368005195491176e-05,
-      "loss": 0.2831,
+      "epoch": 2.4885573215122356,
+      "grad_norm": 0.20642319321632385,
+      "learning_rate": 2.641279088464257e-05,
+      "loss": 0.3978,
       "step": 69050
     },
     {
-      "epoch": 2.43,
-      "learning_rate": 2.736516926685054e-05,
-      "loss": 0.2635,
+      "epoch": 2.4887375211734604,
+      "grad_norm": 0.22682100534439087,
+      "learning_rate": 2.6409877404724363e-05,
+      "loss": 0.4194,
       "step": 69055
     },
     {
-      "epoch": 2.43,
-      "learning_rate": 2.736290050182495e-05,
-      "loss": 0.2612,
+      "epoch": 2.4889177208346847,
+      "grad_norm": 0.24420538544654846,
+      "learning_rate": 2.640696390559686e-05,
+      "loss": 0.3943,
       "step": 69060
     },
     {
-      "epoch": 2.43,
-      "learning_rate": 2.7360064517932533e-05,
-      "loss": 0.2668,
+      "epoch": 2.4890979204959094,
+      "grad_norm": 0.16333971917629242,
+      "learning_rate": 2.6404050387299744e-05,
+      "loss": 0.3718,
       "step": 69065
     },
     {
-      "epoch": 2.43,
-      "learning_rate": 2.735722850339623e-05,
-      "loss": 0.2729,
+      "epoch": 2.489278120157134,
+      "grad_norm": 0.19951343536376953,
+      "learning_rate": 2.640113684987271e-05,
+      "loss": 0.3919,
       "step": 69070
     },
     {
-      "epoch": 2.43,
-      "learning_rate": 2.7354392458252876e-05,
-      "loss": 0.2574,
+      "epoch": 2.489458319818359,
+      "grad_norm": 0.18329237401485443,
+      "learning_rate": 2.6398223293355455e-05,
+      "loss": 0.408,
       "step": 69075
     },
     {
-      "epoch": 2.43,
-      "learning_rate": 2.7351556382539283e-05,
-      "loss": 0.2612,
+      "epoch": 2.489638519479583,
+      "grad_norm": 0.2317817211151123,
+      "learning_rate": 2.6395309717787686e-05,
+      "loss": 0.384,
       "step": 69080
     },
     {
-      "epoch": 2.43,
-      "learning_rate": 2.7348720276292284e-05,
-      "loss": 0.2877,
+      "epoch": 2.489818719140808,
+      "grad_norm": 0.20693840086460114,
+      "learning_rate": 2.6392396123209085e-05,
+      "loss": 0.3992,
       "step": 69085
     },
     {
-      "epoch": 2.43,
-      "learning_rate": 2.734588413954869e-05,
-      "loss": 0.2746,
+      "epoch": 2.4899989188020326,
+      "grad_norm": 0.18415316939353943,
+      "learning_rate": 2.6389482509659365e-05,
+      "loss": 0.3955,
       "step": 69090
     },
     {
-      "epoch": 2.43,
-      "learning_rate": 2.7343047972345347e-05,
-      "loss": 0.2714,
+      "epoch": 2.4901791184632573,
+      "grad_norm": 0.18562282621860504,
+      "learning_rate": 2.6386568877178204e-05,
+      "loss": 0.3756,
       "step": 69095
     },
     {
-      "epoch": 2.43,
-      "learning_rate": 2.7340211774719064e-05,
-      "loss": 0.2651,
+      "epoch": 2.490359318124482,
+      "grad_norm": 0.1960134655237198,
+      "learning_rate": 2.6383655225805326e-05,
+      "loss": 0.3731,
       "step": 69100
     },
     {
-      "epoch": 2.43,
-      "learning_rate": 2.7337375546706677e-05,
-      "loss": 0.2713,
+      "epoch": 2.4905395177857064,
+      "grad_norm": 0.22196051478385925,
+      "learning_rate": 2.6380741555580398e-05,
+      "loss": 0.3898,
       "step": 69105
     },
     {
-      "epoch": 2.43,
-      "learning_rate": 2.7334539288345012e-05,
-      "loss": 0.2508,
+      "epoch": 2.490719717446931,
+      "grad_norm": 0.19862577319145203,
+      "learning_rate": 2.6377827866543142e-05,
+      "loss": 0.371,
       "step": 69110
     },
     {
-      "epoch": 2.43,
-      "learning_rate": 2.7331702999670883e-05,
-      "loss": 0.2733,
+      "epoch": 2.490899917108156,
+      "grad_norm": 0.1888255774974823,
+      "learning_rate": 2.6374914158733238e-05,
+      "loss": 0.3924,
       "step": 69115
     },
     {
-      "epoch": 2.43,
-      "learning_rate": 2.7328866680721134e-05,
-      "loss": 0.2548,
+      "epoch": 2.4910801167693806,
+      "grad_norm": 0.18701927363872528,
+      "learning_rate": 2.6372000432190407e-05,
+      "loss": 0.369,
       "step": 69120
     },
     {
-      "epoch": 2.43,
-      "learning_rate": 2.7326030331532586e-05,
-      "loss": 0.2789,
+      "epoch": 2.4912603164306053,
+      "grad_norm": 0.25110724568367004,
+      "learning_rate": 2.636908668695433e-05,
+      "loss": 0.3578,
       "step": 69125
     },
     {
-      "epoch": 2.43,
-      "learning_rate": 2.732319395214207e-05,
-      "loss": 0.2904,
+      "epoch": 2.4914405160918296,
+      "grad_norm": 0.21141697466373444,
+      "learning_rate": 2.6366172923064714e-05,
+      "loss": 0.4043,
       "step": 69130
     },
     {
-      "epoch": 2.43,
-      "learning_rate": 2.7320357542586405e-05,
-      "loss": 0.2713,
+      "epoch": 2.4916207157530543,
+      "grad_norm": 0.23328730463981628,
+      "learning_rate": 2.6363259140561252e-05,
+      "loss": 0.4044,
       "step": 69135
     },
     {
-      "epoch": 2.43,
-      "learning_rate": 2.7317521102902433e-05,
-      "loss": 0.2806,
+      "epoch": 2.491800915414279,
+      "grad_norm": 0.17210394144058228,
+      "learning_rate": 2.6360345339483655e-05,
+      "loss": 0.3999,
       "step": 69140
     },
     {
-      "epoch": 2.43,
-      "learning_rate": 2.7314684633126964e-05,
-      "loss": 0.2691,
+      "epoch": 2.491981115075504,
+      "grad_norm": 0.19488251209259033,
+      "learning_rate": 2.6357431519871612e-05,
+      "loss": 0.3879,
       "step": 69145
     },
     {
-      "epoch": 2.43,
-      "learning_rate": 2.7311848133296853e-05,
-      "loss": 0.2929,
+      "epoch": 2.4921613147367285,
+      "grad_norm": 0.23553623259067535,
+      "learning_rate": 2.635451768176483e-05,
+      "loss": 0.412,
       "step": 69150
     },
     {
-      "epoch": 2.43,
-      "learning_rate": 2.7309011603448914e-05,
-      "loss": 0.2708,
+      "epoch": 2.492341514397953,
+      "grad_norm": 0.21684016287326813,
+      "learning_rate": 2.6351603825203003e-05,
+      "loss": 0.425,
       "step": 69155
     },
     {
-      "epoch": 2.43,
-      "learning_rate": 2.730617504361998e-05,
-      "loss": 0.2812,
+      "epoch": 2.4925217140591776,
+      "grad_norm": 0.22947993874549866,
+      "learning_rate": 2.634868995022584e-05,
+      "loss": 0.3921,
       "step": 69160
     },
     {
-      "epoch": 2.43,
-      "learning_rate": 2.7303338453846884e-05,
-      "loss": 0.2603,
+      "epoch": 2.4927019137204023,
+      "grad_norm": 0.1686110943555832,
+      "learning_rate": 2.634577605687303e-05,
+      "loss": 0.3758,
       "step": 69165
     },
     {
-      "epoch": 2.43,
-      "learning_rate": 2.7300501834166452e-05,
-      "loss": 0.2747,
+      "epoch": 2.492882113381627,
+      "grad_norm": 0.2178329974412918,
+      "learning_rate": 2.6342862145184287e-05,
+      "loss": 0.3766,
       "step": 69170
     },
     {
-      "epoch": 2.43,
-      "learning_rate": 2.7297665184615522e-05,
-      "loss": 0.2567,
+      "epoch": 2.4930623130428513,
+      "grad_norm": 0.16872194409370422,
+      "learning_rate": 2.6339948215199304e-05,
+      "loss": 0.3641,
       "step": 69175
     },
     {
-      "epoch": 2.43,
-      "learning_rate": 2.7294828505230936e-05,
-      "loss": 0.2734,
+      "epoch": 2.493242512704076,
+      "grad_norm": 0.2486531138420105,
+      "learning_rate": 2.633703426695779e-05,
+      "loss": 0.4011,
       "step": 69180
     },
     {
-      "epoch": 2.43,
-      "learning_rate": 2.72919917960495e-05,
-      "loss": 0.2757,
+      "epoch": 2.4934227123653008,
+      "grad_norm": 0.1963934600353241,
+      "learning_rate": 2.6334120300499443e-05,
+      "loss": 0.3931,
       "step": 69185
     },
     {
-      "epoch": 2.43,
-      "learning_rate": 2.728915505710807e-05,
-      "loss": 0.2778,
+      "epoch": 2.4936029120265255,
+      "grad_norm": 0.20075222849845886,
+      "learning_rate": 2.6331206315863966e-05,
+      "loss": 0.3863,
       "step": 69190
     },
     {
-      "epoch": 2.43,
-      "learning_rate": 2.728631828844346e-05,
-      "loss": 0.2741,
+      "epoch": 2.4937831116877502,
+      "grad_norm": 0.19349929690361023,
+      "learning_rate": 2.6328292313091056e-05,
+      "loss": 0.3631,
       "step": 69195
     },
     {
-      "epoch": 2.43,
-      "learning_rate": 2.7283481490092516e-05,
-      "loss": 0.2572,
+      "epoch": 2.4939633113489745,
+      "grad_norm": 0.14709793031215668,
+      "learning_rate": 2.6325378292220428e-05,
+      "loss": 0.3709,
       "step": 69200
     },
     {
-      "epoch": 2.43,
-      "learning_rate": 2.728064466209207e-05,
-      "loss": 0.2774,
+      "epoch": 2.4941435110101993,
+      "grad_norm": 0.21333067119121552,
+      "learning_rate": 2.6322464253291775e-05,
+      "loss": 0.3988,
       "step": 69205
     },
     {
-      "epoch": 2.43,
-      "learning_rate": 2.7277807804478965e-05,
-      "loss": 0.2795,
+      "epoch": 2.494323710671424,
+      "grad_norm": 0.19537271559238434,
+      "learning_rate": 2.6319550196344793e-05,
+      "loss": 0.3719,
       "step": 69210
     },
     {
-      "epoch": 2.44,
-      "learning_rate": 2.7274970917290017e-05,
-      "loss": 0.2495,
+      "epoch": 2.4945039103326487,
+      "grad_norm": 0.22581617534160614,
+      "learning_rate": 2.63166361214192e-05,
+      "loss": 0.4023,
       "step": 69215
     },
     {
-      "epoch": 2.44,
-      "learning_rate": 2.7272134000562076e-05,
-      "loss": 0.2688,
+      "epoch": 2.494684109993873,
+      "grad_norm": 0.2060622274875641,
+      "learning_rate": 2.6313722028554692e-05,
+      "loss": 0.3808,
       "step": 69220
     },
     {
-      "epoch": 2.44,
-      "learning_rate": 2.7269297054331965e-05,
-      "loss": 0.2808,
+      "epoch": 2.4948643096550978,
+      "grad_norm": 0.19554631412029266,
+      "learning_rate": 2.631080791779099e-05,
+      "loss": 0.4116,
       "step": 69225
     },
     {
-      "epoch": 2.44,
-      "learning_rate": 2.7266460078636534e-05,
-      "loss": 0.2556,
+      "epoch": 2.4950445093163225,
+      "grad_norm": 0.1858394294977188,
+      "learning_rate": 2.630789378916777e-05,
+      "loss": 0.3857,
       "step": 69230
     },
     {
-      "epoch": 2.44,
-      "learning_rate": 2.726362307351261e-05,
-      "loss": 0.2753,
+      "epoch": 2.4952247089775472,
+      "grad_norm": 0.23908282816410065,
+      "learning_rate": 2.6304979642724754e-05,
+      "loss": 0.3515,
       "step": 69235
     },
     {
-      "epoch": 2.44,
-      "learning_rate": 2.7260786038997045e-05,
-      "loss": 0.286,
+      "epoch": 2.495404908638772,
+      "grad_norm": 0.1817556619644165,
+      "learning_rate": 2.630206547850165e-05,
+      "loss": 0.4181,
       "step": 69240
     },
     {
-      "epoch": 2.44,
-      "learning_rate": 2.7257948975126647e-05,
-      "loss": 0.2793,
+      "epoch": 2.4955851082999962,
+      "grad_norm": 0.19453182816505432,
+      "learning_rate": 2.629915129653815e-05,
+      "loss": 0.3915,
       "step": 69245
     },
     {
-      "epoch": 2.44,
-      "learning_rate": 2.7255111881938273e-05,
-      "loss": 0.2538,
+      "epoch": 2.495765307961221,
+      "grad_norm": 0.18795980513095856,
+      "learning_rate": 2.6296237096873964e-05,
+      "loss": 0.3749,
       "step": 69250
     },
     {
-      "epoch": 2.44,
-      "learning_rate": 2.7252274759468754e-05,
-      "loss": 0.2577,
+      "epoch": 2.4959455076224457,
+      "grad_norm": 0.16127502918243408,
+      "learning_rate": 2.6293322879548792e-05,
+      "loss": 0.3953,
       "step": 69255
     },
     {
-      "epoch": 2.44,
-      "learning_rate": 2.7249437607754937e-05,
-      "loss": 0.2796,
+      "epoch": 2.4961257072836704,
+      "grad_norm": 0.20099008083343506,
+      "learning_rate": 2.629040864460236e-05,
+      "loss": 0.3947,
       "step": 69260
     },
     {
-      "epoch": 2.44,
-      "learning_rate": 2.7246600426833656e-05,
-      "loss": 0.272,
+      "epoch": 2.4963059069448947,
+      "grad_norm": 0.18468354642391205,
+      "learning_rate": 2.6287494392074352e-05,
+      "loss": 0.3996,
       "step": 69265
     },
     {
-      "epoch": 2.44,
-      "learning_rate": 2.7243763216741745e-05,
-      "loss": 0.2648,
+      "epoch": 2.4964861066061195,
+      "grad_norm": 0.190368190407753,
+      "learning_rate": 2.6284580122004482e-05,
+      "loss": 0.3393,
       "step": 69270
     },
     {
-      "epoch": 2.44,
-      "learning_rate": 2.724092597751606e-05,
-      "loss": 0.2778,
+      "epoch": 2.496666306267344,
+      "grad_norm": 0.2398798167705536,
+      "learning_rate": 2.6281665834432462e-05,
+      "loss": 0.396,
       "step": 69275
     },
     {
-      "epoch": 2.44,
-      "learning_rate": 2.723808870919341e-05,
-      "loss": 0.2485,
+      "epoch": 2.496846505928569,
+      "grad_norm": 0.21134978532791138,
+      "learning_rate": 2.6278751529397983e-05,
+      "loss": 0.3844,
       "step": 69280
     },
     {
-      "epoch": 2.44,
-      "learning_rate": 2.7235251411810657e-05,
-      "loss": 0.2654,
+      "epoch": 2.4970267055897937,
+      "grad_norm": 0.221641406416893,
+      "learning_rate": 2.6275837206940772e-05,
+      "loss": 0.3982,
       "step": 69285
     },
     {
-      "epoch": 2.44,
-      "learning_rate": 2.7232414085404643e-05,
-      "loss": 0.2599,
+      "epoch": 2.497206905251018,
+      "grad_norm": 0.2198064923286438,
+      "learning_rate": 2.6272922867100524e-05,
+      "loss": 0.4137,
       "step": 69290
     },
     {
-      "epoch": 2.44,
-      "learning_rate": 2.7229576730012207e-05,
-      "loss": 0.232,
+      "epoch": 2.4973871049122427,
+      "grad_norm": 0.26445555686950684,
+      "learning_rate": 2.627000850991695e-05,
+      "loss": 0.4162,
       "step": 69295
     },
     {
-      "epoch": 2.44,
-      "learning_rate": 2.7226739345670176e-05,
-      "loss": 0.2623,
+      "epoch": 2.4975673045734674,
+      "grad_norm": 0.2406071126461029,
+      "learning_rate": 2.6267094135429748e-05,
+      "loss": 0.3863,
       "step": 69300
     },
     {
-      "epoch": 2.44,
-      "learning_rate": 2.7223901932415415e-05,
-      "loss": 0.2579,
+      "epoch": 2.497747504234692,
+      "grad_norm": 0.21584929525852203,
+      "learning_rate": 2.6264179743678642e-05,
+      "loss": 0.3893,
       "step": 69305
     },
     {
-      "epoch": 2.44,
-      "learning_rate": 2.7221064490284738e-05,
-      "loss": 0.2594,
+      "epoch": 2.4979277038959165,
+      "grad_norm": 0.19201698899269104,
+      "learning_rate": 2.6261265334703327e-05,
+      "loss": 0.4101,
       "step": 69310
     },
     {
-      "epoch": 2.44,
-      "learning_rate": 2.721822701931501e-05,
-      "loss": 0.2513,
+      "epoch": 2.498107903557141,
+      "grad_norm": 0.22103743255138397,
+      "learning_rate": 2.625835090854351e-05,
+      "loss": 0.4085,
       "step": 69315
     },
     {
-      "epoch": 2.44,
-      "learning_rate": 2.7215389519543068e-05,
-      "loss": 0.3011,
+      "epoch": 2.498288103218366,
+      "grad_norm": 0.24371977150440216,
+      "learning_rate": 2.625543646523892e-05,
+      "loss": 0.4247,
       "step": 69320
     },
     {
-      "epoch": 2.44,
-      "learning_rate": 2.7212551991005753e-05,
-      "loss": 0.256,
+      "epoch": 2.4984683028795907,
+      "grad_norm": 0.21865001320838928,
+      "learning_rate": 2.6252522004829243e-05,
+      "loss": 0.3971,
       "step": 69325
     },
     {
-      "epoch": 2.44,
-      "learning_rate": 2.7209714433739903e-05,
-      "loss": 0.2623,
+      "epoch": 2.4986485025408154,
+      "grad_norm": 0.18151873350143433,
+      "learning_rate": 2.6249607527354198e-05,
+      "loss": 0.4078,
       "step": 69330
     },
     {
-      "epoch": 2.44,
-      "learning_rate": 2.720687684778237e-05,
-      "loss": 0.2762,
+      "epoch": 2.4988287022020397,
+      "grad_norm": 0.19175831973552704,
+      "learning_rate": 2.6246693032853486e-05,
+      "loss": 0.4046,
       "step": 69335
     },
     {
-      "epoch": 2.44,
-      "learning_rate": 2.7204039233169992e-05,
-      "loss": 0.2626,
+      "epoch": 2.4990089018632644,
+      "grad_norm": 0.18904618918895721,
+      "learning_rate": 2.6243778521366835e-05,
+      "loss": 0.3663,
       "step": 69340
     },
     {
-      "epoch": 2.44,
-      "learning_rate": 2.720120158993963e-05,
-      "loss": 0.281,
+      "epoch": 2.499189101524489,
+      "grad_norm": 0.21790605783462524,
+      "learning_rate": 2.6240863992933936e-05,
+      "loss": 0.4033,
       "step": 69345
     },
     {
-      "epoch": 2.44,
-      "learning_rate": 2.71983639181281e-05,
-      "loss": 0.2703,
+      "epoch": 2.499369301185714,
+      "grad_norm": 0.20878206193447113,
+      "learning_rate": 2.6237949447594502e-05,
+      "loss": 0.4134,
       "step": 69350
     },
     {
-      "epoch": 2.44,
-      "learning_rate": 2.7195526217772278e-05,
-      "loss": 0.2765,
+      "epoch": 2.499549500846938,
+      "grad_norm": 0.2429923415184021,
+      "learning_rate": 2.6235034885388247e-05,
+      "loss": 0.3973,
       "step": 69355
     },
     {
-      "epoch": 2.44,
-      "learning_rate": 2.7192688488908984e-05,
-      "loss": 0.282,
+      "epoch": 2.499729700508163,
+      "grad_norm": 0.19370168447494507,
+      "learning_rate": 2.6232120306354884e-05,
+      "loss": 0.3918,
       "step": 69360
     },
     {
-      "epoch": 2.44,
-      "learning_rate": 2.7189850731575077e-05,
-      "loss": 0.2712,
+      "epoch": 2.4999099001693876,
+      "grad_norm": 0.22662904858589172,
+      "learning_rate": 2.6229205710534126e-05,
+      "loss": 0.37,
       "step": 69365
     },
     {
-      "epoch": 2.44,
-      "learning_rate": 2.71870129458074e-05,
-      "loss": 0.2931,
+      "epoch": 2.5000900998306124,
+      "grad_norm": 0.18225222826004028,
+      "learning_rate": 2.6226291097965668e-05,
+      "loss": 0.4206,
       "step": 69370
     },
     {
-      "epoch": 2.44,
-      "learning_rate": 2.7184175131642803e-05,
-      "loss": 0.2832,
+      "epoch": 2.500270299491837,
+      "grad_norm": 0.2310526967048645,
+      "learning_rate": 2.622337646868923e-05,
+      "loss": 0.4082,
       "step": 69375
     },
     {
-      "epoch": 2.44,
-      "learning_rate": 2.7181337289118137e-05,
-      "loss": 0.2733,
+      "epoch": 2.500450499153062,
+      "grad_norm": 0.20121249556541443,
+      "learning_rate": 2.6220461822744536e-05,
+      "loss": 0.407,
       "step": 69380
     },
     {
-      "epoch": 2.44,
-      "learning_rate": 2.7178499418270238e-05,
-      "loss": 0.2858,
+      "epoch": 2.500630698814286,
+      "grad_norm": 0.24337324500083923,
+      "learning_rate": 2.6217547160171274e-05,
+      "loss": 0.4463,
       "step": 69385
     },
     {
-      "epoch": 2.44,
-      "learning_rate": 2.7175661519135958e-05,
-      "loss": 0.2905,
+      "epoch": 2.500810898475511,
+      "grad_norm": 0.21460457146167755,
+      "learning_rate": 2.6214632481009176e-05,
+      "loss": 0.3511,
       "step": 69390
     },
     {
-      "epoch": 2.44,
-      "learning_rate": 2.717282359175215e-05,
-      "loss": 0.2579,
+      "epoch": 2.5009910981367356,
+      "grad_norm": 0.19160525500774384,
+      "learning_rate": 2.621171778529794e-05,
+      "loss": 0.3843,
       "step": 69395
     },
     {
-      "epoch": 2.44,
-      "learning_rate": 2.7169985636155663e-05,
-      "loss": 0.2739,
+      "epoch": 2.50117129779796,
+      "grad_norm": 0.20841006934642792,
+      "learning_rate": 2.6208803073077294e-05,
+      "loss": 0.4167,
       "step": 69400
     },
     {
-      "epoch": 2.44,
-      "learning_rate": 2.7167147652383334e-05,
-      "loss": 0.2675,
+      "epoch": 2.5013514974591846,
+      "grad_norm": 0.21460412442684174,
+      "learning_rate": 2.6205888344386927e-05,
+      "loss": 0.3881,
       "step": 69405
     },
     {
-      "epoch": 2.44,
-      "learning_rate": 2.716430964047203e-05,
-      "loss": 0.2827,
+      "epoch": 2.5015316971204093,
+      "grad_norm": 0.20934461057186127,
+      "learning_rate": 2.6202973599266573e-05,
+      "loss": 0.371,
       "step": 69410
     },
     {
-      "epoch": 2.44,
-      "learning_rate": 2.7161471600458588e-05,
-      "loss": 0.2544,
+      "epoch": 2.501711896781634,
+      "grad_norm": 0.23594777286052704,
+      "learning_rate": 2.620005883775593e-05,
+      "loss": 0.3876,
       "step": 69415
     },
     {
-      "epoch": 2.44,
-      "learning_rate": 2.7158633532379863e-05,
-      "loss": 0.301,
+      "epoch": 2.501892096442859,
+      "grad_norm": 0.21076056361198425,
+      "learning_rate": 2.6197144059894724e-05,
+      "loss": 0.4082,
       "step": 69420
     },
     {
-      "epoch": 2.44,
-      "learning_rate": 2.7155795436272696e-05,
-      "loss": 0.2817,
+      "epoch": 2.5020722961040835,
+      "grad_norm": 0.21848683059215546,
+      "learning_rate": 2.619422926572266e-05,
+      "loss": 0.3747,
       "step": 69425
     },
     {
-      "epoch": 2.44,
-      "learning_rate": 2.715295731217396e-05,
-      "loss": 0.2906,
+      "epoch": 2.502252495765308,
+      "grad_norm": 0.20957939326763153,
+      "learning_rate": 2.6191314455279453e-05,
+      "loss": 0.4058,
       "step": 69430
     },
     {
-      "epoch": 2.44,
-      "learning_rate": 2.715011916012048e-05,
-      "loss": 0.2861,
+      "epoch": 2.5024326954265326,
+      "grad_norm": 0.23492544889450073,
+      "learning_rate": 2.618839962860482e-05,
+      "loss": 0.383,
       "step": 69435
     },
     {
-      "epoch": 2.44,
-      "learning_rate": 2.714728098014913e-05,
-      "loss": 0.2622,
+      "epoch": 2.5026128950877573,
+      "grad_norm": 0.16732336580753326,
+      "learning_rate": 2.6185484785738467e-05,
+      "loss": 0.391,
       "step": 69440
     },
     {
-      "epoch": 2.44,
-      "learning_rate": 2.7144442772296742e-05,
-      "loss": 0.2833,
+      "epoch": 2.5027930947489816,
+      "grad_norm": 0.18726083636283875,
+      "learning_rate": 2.6182569926720117e-05,
+      "loss": 0.381,
       "step": 69445
     },
     {
-      "epoch": 2.44,
-      "learning_rate": 2.7141604536600185e-05,
-      "loss": 0.246,
+      "epoch": 2.5029732944102063,
+      "grad_norm": 0.201826810836792,
+      "learning_rate": 2.617965505158948e-05,
+      "loss": 0.3941,
       "step": 69450
     },
     {
-      "epoch": 2.44,
-      "learning_rate": 2.7138766273096296e-05,
-      "loss": 0.2596,
+      "epoch": 2.503153494071431,
+      "grad_norm": 0.20765550434589386,
+      "learning_rate": 2.617674016038627e-05,
+      "loss": 0.4062,
       "step": 69455
     },
     {
-      "epoch": 2.44,
-      "learning_rate": 2.7135927981821947e-05,
-      "loss": 0.2619,
+      "epoch": 2.503333693732656,
+      "grad_norm": 0.23174385726451874,
+      "learning_rate": 2.61738252531502e-05,
+      "loss": 0.3968,
       "step": 69460
     },
     {
-      "epoch": 2.44,
-      "learning_rate": 2.7133089662813977e-05,
-      "loss": 0.2508,
+      "epoch": 2.5035138933938805,
+      "grad_norm": 0.21468636393547058,
+      "learning_rate": 2.6170910329920994e-05,
+      "loss": 0.3938,
       "step": 69465
     },
     {
-      "epoch": 2.44,
-      "learning_rate": 2.7130251316109233e-05,
-      "loss": 0.2653,
+      "epoch": 2.5036940930551053,
+      "grad_norm": 0.23031465709209442,
+      "learning_rate": 2.6167995390738366e-05,
+      "loss": 0.4022,
       "step": 69470
     },
     {
-      "epoch": 2.44,
-      "learning_rate": 2.712741294174459e-05,
-      "loss": 0.2602,
+      "epoch": 2.5038742927163296,
+      "grad_norm": 0.19280219078063965,
+      "learning_rate": 2.6165080435642014e-05,
+      "loss": 0.3947,
       "step": 69475
     },
     {
-      "epoch": 2.44,
-      "learning_rate": 2.71245745397569e-05,
-      "loss": 0.2717,
+      "epoch": 2.5040544923775543,
+      "grad_norm": 0.20321400463581085,
+      "learning_rate": 2.616216546467168e-05,
+      "loss": 0.3962,
       "step": 69480
     },
     {
-      "epoch": 2.44,
-      "learning_rate": 2.7121736110182994e-05,
-      "loss": 0.2808,
+      "epoch": 2.504234692038779,
+      "grad_norm": 0.16147299110889435,
+      "learning_rate": 2.6159250477867053e-05,
+      "loss": 0.4046,
       "step": 69485
     },
     {
-      "epoch": 2.44,
-      "learning_rate": 2.711889765305975e-05,
-      "loss": 0.2695,
+      "epoch": 2.5044148917000038,
+      "grad_norm": 0.21102051436901093,
+      "learning_rate": 2.6156335475267874e-05,
+      "loss": 0.4235,
       "step": 69490
     },
     {
-      "epoch": 2.45,
-      "learning_rate": 2.711605916842402e-05,
-      "loss": 0.281,
+      "epoch": 2.504595091361228,
+      "grad_norm": 0.22560648620128632,
+      "learning_rate": 2.615342045691384e-05,
+      "loss": 0.4053,
       "step": 69495
     },
     {
-      "epoch": 2.45,
-      "learning_rate": 2.711322065631265e-05,
-      "loss": 0.2699,
+      "epoch": 2.5047752910224528,
+      "grad_norm": 0.17885218560695648,
+      "learning_rate": 2.615050542284468e-05,
+      "loss": 0.35,
       "step": 69500
     },
     {
-      "epoch": 2.45,
-      "eval_loss": 0.2653324604034424,
-      "eval_runtime": 10.5362,
-      "eval_samples_per_second": 9.491,
-      "eval_steps_per_second": 9.491,
+      "epoch": 2.5047752910224528,
+      "eval_loss": 0.4340173304080963,
+      "eval_runtime": 3.5362,
+      "eval_samples_per_second": 28.279,
+      "eval_steps_per_second": 7.07,
       "step": 69500
     },
     {
-      "epoch": 2.45,
-      "learning_rate": 2.71103821167625e-05,
-      "loss": 0.2805,
+      "epoch": 2.5049554906836775,
+      "grad_norm": 0.23639249801635742,
+      "learning_rate": 2.6147590373100106e-05,
+      "loss": 0.4151,
       "step": 69505
     },
     {
-      "epoch": 2.45,
-      "learning_rate": 2.7107543549810433e-05,
-      "loss": 0.2772,
+      "epoch": 2.5051356903449022,
+      "grad_norm": 0.260985404253006,
+      "learning_rate": 2.6144675307719835e-05,
+      "loss": 0.4056,
       "step": 69510
     },
     {
-      "epoch": 2.45,
-      "learning_rate": 2.7104704955493303e-05,
-      "loss": 0.2752,
+      "epoch": 2.505315890006127,
+      "grad_norm": 0.21319666504859924,
+      "learning_rate": 2.614176022674359e-05,
+      "loss": 0.3559,
       "step": 69515
     },
     {
-      "epoch": 2.45,
-      "learning_rate": 2.7101866333847962e-05,
-      "loss": 0.2571,
+      "epoch": 2.5054960896673513,
+      "grad_norm": 0.20121723413467407,
+      "learning_rate": 2.613884513021107e-05,
+      "loss": 0.4013,
       "step": 69520
     },
     {
-      "epoch": 2.45,
-      "learning_rate": 2.709902768491128e-05,
-      "loss": 0.2747,
+      "epoch": 2.505676289328576,
+      "grad_norm": 0.1814257949590683,
+      "learning_rate": 2.613593001816201e-05,
+      "loss": 0.3781,
       "step": 69525
     },
     {
-      "epoch": 2.45,
-      "learning_rate": 2.7096189008720096e-05,
-      "loss": 0.262,
+      "epoch": 2.5058564889898007,
+      "grad_norm": 0.20692424476146698,
+      "learning_rate": 2.613301489063613e-05,
+      "loss": 0.3848,
       "step": 69530
     },
     {
-      "epoch": 2.45,
-      "learning_rate": 2.7093350305311288e-05,
-      "loss": 0.2717,
+      "epoch": 2.5060366886510255,
+      "grad_norm": 0.21745166182518005,
+      "learning_rate": 2.6130099747673136e-05,
+      "loss": 0.3875,
       "step": 69535
     },
     {
-      "epoch": 2.45,
-      "learning_rate": 2.7090511574721705e-05,
-      "loss": 0.2975,
+      "epoch": 2.5062168883122498,
+      "grad_norm": 0.19533394277095795,
+      "learning_rate": 2.612718458931276e-05,
+      "loss": 0.4273,
       "step": 69540
     },
     {
-      "epoch": 2.45,
-      "learning_rate": 2.708767281698821e-05,
-      "loss": 0.2871,
+      "epoch": 2.5063970879734745,
+      "grad_norm": 0.18785609304904938,
+      "learning_rate": 2.6124269415594698e-05,
+      "loss": 0.4146,
       "step": 69545
     },
     {
-      "epoch": 2.45,
-      "learning_rate": 2.708483403214765e-05,
-      "loss": 0.2548,
+      "epoch": 2.5065772876346992,
+      "grad_norm": 0.21127115190029144,
+      "learning_rate": 2.6121354226558692e-05,
+      "loss": 0.4068,
       "step": 69550
     },
     {
-      "epoch": 2.45,
-      "learning_rate": 2.7081995220236893e-05,
-      "loss": 0.269,
+      "epoch": 2.506757487295924,
+      "grad_norm": 0.2023439109325409,
+      "learning_rate": 2.611843902224445e-05,
+      "loss": 0.4257,
       "step": 69555
     },
     {
-      "epoch": 2.45,
-      "learning_rate": 2.707915638129281e-05,
-      "loss": 0.2799,
+      "epoch": 2.5069376869571487,
+      "grad_norm": 0.1987898051738739,
+      "learning_rate": 2.6115523802691695e-05,
+      "loss": 0.4159,
       "step": 69560
     },
     {
-      "epoch": 2.45,
-      "learning_rate": 2.7076317515352247e-05,
-      "loss": 0.2619,
+      "epoch": 2.5071178866183734,
+      "grad_norm": 0.2078784704208374,
+      "learning_rate": 2.6112608567940138e-05,
+      "loss": 0.3997,
       "step": 69565
     },
     {
-      "epoch": 2.45,
-      "learning_rate": 2.707347862245207e-05,
-      "loss": 0.2628,
+      "epoch": 2.5072980862795977,
+      "grad_norm": 0.27035263180732727,
+      "learning_rate": 2.610969331802951e-05,
+      "loss": 0.4235,
       "step": 69570
     },
     {
-      "epoch": 2.45,
-      "learning_rate": 2.7070639702629142e-05,
-      "loss": 0.2729,
+      "epoch": 2.5074782859408224,
+      "grad_norm": 0.22846317291259766,
+      "learning_rate": 2.610677805299953e-05,
+      "loss": 0.3812,
       "step": 69575
     },
     {
-      "epoch": 2.45,
-      "learning_rate": 2.706780075592032e-05,
-      "loss": 0.2924,
+      "epoch": 2.507658485602047,
+      "grad_norm": 0.20302127301692963,
+      "learning_rate": 2.6103862772889902e-05,
+      "loss": 0.3991,
       "step": 69580
     },
     {
-      "epoch": 2.45,
-      "learning_rate": 2.7064961782362462e-05,
-      "loss": 0.2756,
+      "epoch": 2.5078386852632715,
+      "grad_norm": 0.20338794589042664,
+      "learning_rate": 2.6100947477740367e-05,
+      "loss": 0.3879,
       "step": 69585
     },
     {
-      "epoch": 2.45,
-      "learning_rate": 2.7062122781992438e-05,
-      "loss": 0.2735,
+      "epoch": 2.508018884924496,
+      "grad_norm": 0.1619461178779602,
+      "learning_rate": 2.609803216759063e-05,
+      "loss": 0.4004,
       "step": 69590
     },
     {
-      "epoch": 2.45,
-      "learning_rate": 2.705928375484712e-05,
-      "loss": 0.2717,
+      "epoch": 2.508199084585721,
+      "grad_norm": 0.2222161740064621,
+      "learning_rate": 2.6095116842480417e-05,
+      "loss": 0.3937,
       "step": 69595
     },
     {
-      "epoch": 2.45,
-      "learning_rate": 2.7056444700963346e-05,
-      "loss": 0.258,
+      "epoch": 2.5083792842469457,
+      "grad_norm": 0.18493251502513885,
+      "learning_rate": 2.6092201502449455e-05,
+      "loss": 0.3873,
       "step": 69600
     },
     {
-      "epoch": 2.45,
-      "learning_rate": 2.7053605620377998e-05,
-      "loss": 0.2592,
+      "epoch": 2.5085594839081704,
+      "grad_norm": 0.21007417142391205,
+      "learning_rate": 2.6089286147537452e-05,
+      "loss": 0.3758,
       "step": 69605
     },
     {
-      "epoch": 2.45,
-      "learning_rate": 2.7050766513127934e-05,
-      "loss": 0.2745,
+      "epoch": 2.508739683569395,
+      "grad_norm": 0.2484482377767563,
+      "learning_rate": 2.608637077778414e-05,
+      "loss": 0.4411,
       "step": 69610
     },
     {
-      "epoch": 2.45,
-      "learning_rate": 2.7047927379250016e-05,
-      "loss": 0.274,
+      "epoch": 2.5089198832306194,
+      "grad_norm": 0.2558552920818329,
+      "learning_rate": 2.608345539322924e-05,
+      "loss": 0.4278,
       "step": 69615
     },
     {
-      "epoch": 2.45,
-      "learning_rate": 2.7045088218781116e-05,
-      "loss": 0.2726,
+      "epoch": 2.509100082891844,
+      "grad_norm": 0.2762976586818695,
+      "learning_rate": 2.6080539993912467e-05,
+      "loss": 0.4425,
       "step": 69620
     },
     {
-      "epoch": 2.45,
-      "learning_rate": 2.7042249031758088e-05,
-      "loss": 0.2749,
+      "epoch": 2.509280282553069,
+      "grad_norm": 0.2021670788526535,
+      "learning_rate": 2.607762457987354e-05,
+      "loss": 0.3999,
       "step": 69625
     },
     {
-      "epoch": 2.45,
-      "learning_rate": 2.703940981821781e-05,
-      "loss": 0.2606,
+      "epoch": 2.509460482214293,
+      "grad_norm": 0.22645121812820435,
+      "learning_rate": 2.6074709151152193e-05,
+      "loss": 0.4255,
       "step": 69630
     },
     {
-      "epoch": 2.45,
-      "learning_rate": 2.7036570578197134e-05,
-      "loss": 0.2564,
+      "epoch": 2.509640681875518,
+      "grad_norm": 0.23237501084804535,
+      "learning_rate": 2.6071793707788138e-05,
+      "loss": 0.431,
       "step": 69635
     },
     {
-      "epoch": 2.45,
-      "learning_rate": 2.703373131173293e-05,
-      "loss": 0.2579,
+      "epoch": 2.5098208815367427,
+      "grad_norm": 0.22652441263198853,
+      "learning_rate": 2.6068878249821106e-05,
+      "loss": 0.3742,
       "step": 69640
     },
     {
-      "epoch": 2.45,
-      "learning_rate": 2.7030892018862066e-05,
-      "loss": 0.2936,
+      "epoch": 2.5100010811979674,
+      "grad_norm": 0.17889447510242462,
+      "learning_rate": 2.606596277729081e-05,
+      "loss": 0.3447,
       "step": 69645
     },
     {
-      "epoch": 2.45,
-      "learning_rate": 2.702805269962141e-05,
-      "loss": 0.2619,
+      "epoch": 2.510181280859192,
+      "grad_norm": 0.21000555157661438,
+      "learning_rate": 2.6063047290236974e-05,
+      "loss": 0.3839,
       "step": 69650
     },
     {
-      "epoch": 2.45,
-      "learning_rate": 2.7025213354047824e-05,
-      "loss": 0.2891,
+      "epoch": 2.510361480520417,
+      "grad_norm": 0.18811647593975067,
+      "learning_rate": 2.6060131788699343e-05,
+      "loss": 0.4068,
       "step": 69655
     },
     {
-      "epoch": 2.45,
-      "learning_rate": 2.7022373982178182e-05,
-      "loss": 0.2682,
+      "epoch": 2.510541680181641,
+      "grad_norm": 0.21040557324886322,
+      "learning_rate": 2.6057216272717605e-05,
+      "loss": 0.3839,
       "step": 69660
     },
     {
-      "epoch": 2.45,
-      "learning_rate": 2.7019534584049333e-05,
-      "loss": 0.2676,
+      "epoch": 2.510721879842866,
+      "grad_norm": 0.23716652393341064,
+      "learning_rate": 2.6054300742331498e-05,
+      "loss": 0.4053,
       "step": 69665
     },
     {
-      "epoch": 2.45,
-      "learning_rate": 2.7016695159698174e-05,
-      "loss": 0.2531,
+      "epoch": 2.5109020795040906,
+      "grad_norm": 0.2132147252559662,
+      "learning_rate": 2.6051385197580757e-05,
+      "loss": 0.3949,
       "step": 69670
     },
     {
-      "epoch": 2.45,
-      "learning_rate": 2.701385570916155e-05,
-      "loss": 0.271,
+      "epoch": 2.511082279165315,
+      "grad_norm": 0.20189203321933746,
+      "learning_rate": 2.6048469638505092e-05,
+      "loss": 0.4238,
       "step": 69675
     },
     {
-      "epoch": 2.45,
-      "learning_rate": 2.7011016232476344e-05,
-      "loss": 0.2571,
+      "epoch": 2.5112624788265396,
+      "grad_norm": 0.2348521649837494,
+      "learning_rate": 2.6045554065144234e-05,
+      "loss": 0.3931,
       "step": 69680
     },
     {
-      "epoch": 2.45,
-      "learning_rate": 2.7008176729679407e-05,
-      "loss": 0.2842,
+      "epoch": 2.5114426784877644,
+      "grad_norm": 0.1996840536594391,
+      "learning_rate": 2.60426384775379e-05,
+      "loss": 0.4236,
       "step": 69685
     },
     {
-      "epoch": 2.45,
-      "learning_rate": 2.7005337200807624e-05,
-      "loss": 0.2799,
+      "epoch": 2.511622878148989,
+      "grad_norm": 0.20005884766578674,
+      "learning_rate": 2.603972287572582e-05,
+      "loss": 0.4177,
       "step": 69690
     },
     {
-      "epoch": 2.45,
-      "learning_rate": 2.7002497645897855e-05,
-      "loss": 0.266,
+      "epoch": 2.511803077810214,
+      "grad_norm": 0.19069145619869232,
+      "learning_rate": 2.603680725974772e-05,
+      "loss": 0.4015,
       "step": 69695
     },
     {
-      "epoch": 2.45,
-      "learning_rate": 2.6999658064986983e-05,
-      "loss": 0.2721,
+      "epoch": 2.5119832774714386,
+      "grad_norm": 0.211561918258667,
+      "learning_rate": 2.6033891629643314e-05,
+      "loss": 0.3864,
       "step": 69700
     },
     {
-      "epoch": 2.45,
-      "learning_rate": 2.6996818458111862e-05,
-      "loss": 0.2712,
+      "epoch": 2.512163477132663,
+      "grad_norm": 0.2280474752187729,
+      "learning_rate": 2.6030975985452344e-05,
+      "loss": 0.3813,
       "step": 69705
     },
     {
-      "epoch": 2.45,
-      "learning_rate": 2.6993978825309374e-05,
-      "loss": 0.2511,
+      "epoch": 2.5123436767938876,
+      "grad_norm": 0.1532917618751526,
+      "learning_rate": 2.602806032721452e-05,
+      "loss": 0.3638,
       "step": 69710
     },
     {
-      "epoch": 2.45,
-      "learning_rate": 2.6991139166616375e-05,
-      "loss": 0.2838,
+      "epoch": 2.5125238764551123,
+      "grad_norm": 0.22681502997875214,
+      "learning_rate": 2.602514465496958e-05,
+      "loss": 0.4054,
       "step": 69715
     },
     {
-      "epoch": 2.45,
-      "learning_rate": 2.698829948206975e-05,
-      "loss": 0.2664,
+      "epoch": 2.512704076116337,
+      "grad_norm": 0.18224921822547913,
+      "learning_rate": 2.6022228968757233e-05,
+      "loss": 0.36,
       "step": 69720
     },
     {
-      "epoch": 2.45,
-      "learning_rate": 2.6985459771706367e-05,
-      "loss": 0.2854,
+      "epoch": 2.5128842757775613,
+      "grad_norm": 0.18726272881031036,
+      "learning_rate": 2.6019313268617223e-05,
+      "loss": 0.3718,
       "step": 69725
     },
     {
-      "epoch": 2.45,
-      "learning_rate": 2.69826200355631e-05,
-      "loss": 0.2769,
+      "epoch": 2.513064475438786,
+      "grad_norm": 0.21877476572990417,
+      "learning_rate": 2.601639755458926e-05,
+      "loss": 0.398,
       "step": 69730
     },
     {
-      "epoch": 2.45,
-      "learning_rate": 2.697978027367682e-05,
-      "loss": 0.2712,
+      "epoch": 2.513244675100011,
+      "grad_norm": 0.2230885922908783,
+      "learning_rate": 2.6013481826713083e-05,
+      "loss": 0.3982,
       "step": 69735
     },
     {
-      "epoch": 2.45,
-      "learning_rate": 2.6976940486084394e-05,
-      "loss": 0.2626,
+      "epoch": 2.5134248747612356,
+      "grad_norm": 0.22673091292381287,
+      "learning_rate": 2.6010566085028408e-05,
+      "loss": 0.4168,
       "step": 69740
     },
     {
-      "epoch": 2.45,
-      "learning_rate": 2.6974100672822705e-05,
-      "loss": 0.2676,
+      "epoch": 2.5136050744224603,
+      "grad_norm": 0.19574376940727234,
+      "learning_rate": 2.6007650329574968e-05,
+      "loss": 0.3864,
       "step": 69745
     },
     {
-      "epoch": 2.45,
-      "learning_rate": 2.6971260833928607e-05,
-      "loss": 0.2453,
+      "epoch": 2.5137852740836846,
+      "grad_norm": 0.1650337278842926,
+      "learning_rate": 2.6004734560392487e-05,
+      "loss": 0.4042,
       "step": 69750
     },
     {
-      "epoch": 2.45,
-      "learning_rate": 2.6968420969439002e-05,
-      "loss": 0.2894,
+      "epoch": 2.5139654737449093,
+      "grad_norm": 0.19529931247234344,
+      "learning_rate": 2.6001818777520692e-05,
+      "loss": 0.3919,
       "step": 69755
     },
     {
-      "epoch": 2.45,
-      "learning_rate": 2.6965581079390734e-05,
-      "loss": 0.2795,
+      "epoch": 2.514145673406134,
+      "grad_norm": 0.21473079919815063,
+      "learning_rate": 2.5998902980999314e-05,
+      "loss": 0.3822,
       "step": 69760
     },
     {
-      "epoch": 2.45,
-      "learning_rate": 2.6962741163820703e-05,
-      "loss": 0.2667,
+      "epoch": 2.5143258730673588,
+      "grad_norm": 0.20333951711654663,
+      "learning_rate": 2.5995987170868068e-05,
+      "loss": 0.3926,
       "step": 69765
     },
     {
-      "epoch": 2.45,
-      "learning_rate": 2.6959901222765764e-05,
-      "loss": 0.2664,
+      "epoch": 2.514506072728583,
+      "grad_norm": 0.2425466775894165,
+      "learning_rate": 2.5993071347166693e-05,
+      "loss": 0.3759,
       "step": 69770
     },
     {
-      "epoch": 2.45,
-      "learning_rate": 2.69570612562628e-05,
-      "loss": 0.2816,
+      "epoch": 2.514686272389808,
+      "grad_norm": 0.20622824132442474,
+      "learning_rate": 2.599015550993491e-05,
+      "loss": 0.3813,
       "step": 69775
     },
     {
-      "epoch": 2.46,
-      "learning_rate": 2.6954221264348684e-05,
-      "loss": 0.2869,
+      "epoch": 2.5148664720510325,
+      "grad_norm": 0.18303707242012024,
+      "learning_rate": 2.598723965921246e-05,
+      "loss": 0.4201,
       "step": 69780
     },
     {
-      "epoch": 2.46,
-      "learning_rate": 2.69513812470603e-05,
-      "loss": 0.2762,
+      "epoch": 2.5150466717122573,
+      "grad_norm": 0.1895645707845688,
+      "learning_rate": 2.5984323795039057e-05,
+      "loss": 0.362,
       "step": 69785
     },
     {
-      "epoch": 2.46,
-      "learning_rate": 2.694854120443451e-05,
-      "loss": 0.278,
+      "epoch": 2.515226871373482,
+      "grad_norm": 0.20545552670955658,
+      "learning_rate": 2.5981407917454427e-05,
+      "loss": 0.384,
       "step": 69790
     },
     {
-      "epoch": 2.46,
-      "learning_rate": 2.69457011365082e-05,
-      "loss": 0.2699,
+      "epoch": 2.5154070710347067,
+      "grad_norm": 0.20499677956104279,
+      "learning_rate": 2.5978492026498308e-05,
+      "loss": 0.3918,
       "step": 69795
     },
     {
-      "epoch": 2.46,
-      "learning_rate": 2.694286104331824e-05,
-      "loss": 0.2747,
+      "epoch": 2.515587270695931,
+      "grad_norm": 0.22519873082637787,
+      "learning_rate": 2.5975576122210422e-05,
+      "loss": 0.4266,
       "step": 69800
     },
     {
-      "epoch": 2.46,
-      "learning_rate": 2.694002092490151e-05,
-      "loss": 0.2802,
+      "epoch": 2.5157674703571558,
+      "grad_norm": 0.21015673875808716,
+      "learning_rate": 2.59726602046305e-05,
+      "loss": 0.3873,
       "step": 69805
     },
     {
-      "epoch": 2.46,
-      "learning_rate": 2.6937180781294885e-05,
-      "loss": 0.288,
+      "epoch": 2.5159476700183805,
+      "grad_norm": 0.19208891689777374,
+      "learning_rate": 2.5969744273798274e-05,
+      "loss": 0.397,
       "step": 69810
     },
     {
-      "epoch": 2.46,
-      "learning_rate": 2.6934340612535247e-05,
-      "loss": 0.2781,
+      "epoch": 2.516127869679605,
+      "grad_norm": 0.23260699212551117,
+      "learning_rate": 2.5966828329753467e-05,
+      "loss": 0.4168,
       "step": 69815
     },
     {
-      "epoch": 2.46,
-      "learning_rate": 2.6931500418659473e-05,
-      "loss": 0.284,
+      "epoch": 2.5163080693408295,
+      "grad_norm": 0.2245369702577591,
+      "learning_rate": 2.596391237253582e-05,
+      "loss": 0.4131,
       "step": 69820
     },
     {
-      "epoch": 2.46,
-      "learning_rate": 2.6928660199704435e-05,
-      "loss": 0.2531,
+      "epoch": 2.5164882690020542,
+      "grad_norm": 0.17223899066448212,
+      "learning_rate": 2.5960996402185043e-05,
+      "loss": 0.373,
       "step": 69825
     },
     {
-      "epoch": 2.46,
-      "learning_rate": 2.6925819955707014e-05,
-      "loss": 0.2725,
+      "epoch": 2.516668468663279,
+      "grad_norm": 0.23170410096645355,
+      "learning_rate": 2.595808041874088e-05,
+      "loss": 0.4081,
       "step": 69830
     },
     {
-      "epoch": 2.46,
-      "learning_rate": 2.6922979686704097e-05,
-      "loss": 0.2799,
+      "epoch": 2.5168486683245037,
+      "grad_norm": 0.21683074533939362,
+      "learning_rate": 2.5955164422243054e-05,
+      "loss": 0.4064,
       "step": 69835
     },
     {
-      "epoch": 2.46,
-      "learning_rate": 2.6920139392732552e-05,
-      "loss": 0.2975,
+      "epoch": 2.5170288679857284,
+      "grad_norm": 0.16481082141399384,
+      "learning_rate": 2.5952248412731305e-05,
+      "loss": 0.3402,
       "step": 69840
     },
     {
-      "epoch": 2.46,
-      "learning_rate": 2.691729907382926e-05,
-      "loss": 0.2886,
+      "epoch": 2.5172090676469527,
+      "grad_norm": 0.15683116018772125,
+      "learning_rate": 2.594933239024535e-05,
+      "loss": 0.4006,
       "step": 69845
     },
     {
-      "epoch": 2.46,
-      "learning_rate": 2.691445873003111e-05,
-      "loss": 0.2714,
+      "epoch": 2.5173892673081775,
+      "grad_norm": 0.18406714498996735,
+      "learning_rate": 2.5946416354824927e-05,
+      "loss": 0.3865,
       "step": 69850
     },
     {
-      "epoch": 2.46,
-      "learning_rate": 2.691161836137497e-05,
-      "loss": 0.2786,
+      "epoch": 2.517569466969402,
+      "grad_norm": 0.19561605155467987,
+      "learning_rate": 2.5943500306509765e-05,
+      "loss": 0.4131,
       "step": 69855
     },
     {
-      "epoch": 2.46,
-      "learning_rate": 2.6908777967897718e-05,
-      "loss": 0.2714,
+      "epoch": 2.5177496666306265,
+      "grad_norm": 0.19253993034362793,
+      "learning_rate": 2.5940584245339593e-05,
+      "loss": 0.3926,
       "step": 69860
     },
     {
-      "epoch": 2.46,
-      "learning_rate": 2.690593754963625e-05,
-      "loss": 0.2534,
+      "epoch": 2.5179298662918512,
+      "grad_norm": 0.24067455530166626,
+      "learning_rate": 2.5937668171354145e-05,
+      "loss": 0.3809,
       "step": 69865
     },
     {
-      "epoch": 2.46,
-      "learning_rate": 2.6903097106627447e-05,
-      "loss": 0.2817,
+      "epoch": 2.518110065953076,
+      "grad_norm": 0.17989800870418549,
+      "learning_rate": 2.5934752084593143e-05,
+      "loss": 0.3891,
       "step": 69870
     },
     {
-      "epoch": 2.46,
-      "learning_rate": 2.6900256638908174e-05,
-      "loss": 0.2838,
+      "epoch": 2.5182902656143007,
+      "grad_norm": 0.21168993413448334,
+      "learning_rate": 2.593183598509633e-05,
+      "loss": 0.3663,
       "step": 69875
     },
     {
-      "epoch": 2.46,
-      "learning_rate": 2.689741614651532e-05,
-      "loss": 0.2638,
+      "epoch": 2.5184704652755254,
+      "grad_norm": 0.19849541783332825,
+      "learning_rate": 2.5928919872903434e-05,
+      "loss": 0.4195,
       "step": 69880
     },
     {
-      "epoch": 2.46,
-      "learning_rate": 2.6894575629485762e-05,
-      "loss": 0.2688,
+      "epoch": 2.51865066493675,
+      "grad_norm": 0.22839267551898956,
+      "learning_rate": 2.592600374805418e-05,
+      "loss": 0.4412,
       "step": 69885
     },
     {
-      "epoch": 2.46,
-      "learning_rate": 2.68917350878564e-05,
-      "loss": 0.2442,
+      "epoch": 2.5188308645979745,
+      "grad_norm": 0.2030458301305771,
+      "learning_rate": 2.5923087610588305e-05,
+      "loss": 0.3982,
       "step": 69890
     },
     {
-      "epoch": 2.46,
-      "learning_rate": 2.68888945216641e-05,
-      "loss": 0.2397,
+      "epoch": 2.519011064259199,
+      "grad_norm": 0.161631777882576,
+      "learning_rate": 2.592017146054554e-05,
+      "loss": 0.3629,
       "step": 69895
     },
     {
-      "epoch": 2.46,
-      "learning_rate": 2.6886053930945755e-05,
-      "loss": 0.2594,
+      "epoch": 2.519191263920424,
+      "grad_norm": 0.21375270187854767,
+      "learning_rate": 2.5917255297965625e-05,
+      "loss": 0.3821,
       "step": 69900
     },
     {
-      "epoch": 2.46,
-      "learning_rate": 2.6883213315738232e-05,
-      "loss": 0.257,
+      "epoch": 2.519371463581648,
+      "grad_norm": 0.19490370154380798,
+      "learning_rate": 2.5914339122888272e-05,
+      "loss": 0.411,
       "step": 69905
     },
     {
-      "epoch": 2.46,
-      "learning_rate": 2.6880372676078423e-05,
-      "loss": 0.2749,
+      "epoch": 2.519551663242873,
+      "grad_norm": 0.2110588401556015,
+      "learning_rate": 2.591142293535323e-05,
+      "loss": 0.3623,
       "step": 69910
     },
     {
-      "epoch": 2.46,
-      "learning_rate": 2.6877532012003225e-05,
-      "loss": 0.2849,
+      "epoch": 2.5197318629040977,
+      "grad_norm": 0.21977630257606506,
+      "learning_rate": 2.5908506735400223e-05,
+      "loss": 0.4114,
       "step": 69915
     },
     {
-      "epoch": 2.46,
-      "learning_rate": 2.687469132354951e-05,
-      "loss": 0.2624,
+      "epoch": 2.5199120625653224,
+      "grad_norm": 0.20621120929718018,
+      "learning_rate": 2.5905590523068995e-05,
+      "loss": 0.3517,
       "step": 69920
     },
     {
-      "epoch": 2.46,
-      "learning_rate": 2.6871850610754156e-05,
-      "loss": 0.2645,
+      "epoch": 2.520092262226547,
+      "grad_norm": 0.20464351773262024,
+      "learning_rate": 2.5902674298399273e-05,
+      "loss": 0.3803,
       "step": 69925
     },
     {
-      "epoch": 2.46,
-      "learning_rate": 2.6869009873654066e-05,
-      "loss": 0.2821,
+      "epoch": 2.520272461887772,
+      "grad_norm": 0.20108066499233246,
+      "learning_rate": 2.5899758061430777e-05,
+      "loss": 0.4085,
       "step": 69930
     },
     {
-      "epoch": 2.46,
-      "learning_rate": 2.6866169112286105e-05,
-      "loss": 0.2655,
+      "epoch": 2.520452661548996,
+      "grad_norm": 0.2749893367290497,
+      "learning_rate": 2.5896841812203265e-05,
+      "loss": 0.3936,
       "step": 69935
     },
     {
-      "epoch": 2.46,
-      "learning_rate": 2.686332832668717e-05,
-      "loss": 0.2726,
+      "epoch": 2.520632861210221,
+      "grad_norm": 0.21574847400188446,
+      "learning_rate": 2.589392555075645e-05,
+      "loss": 0.3946,
       "step": 69940
     },
     {
-      "epoch": 2.46,
-      "learning_rate": 2.6860487516894144e-05,
-      "loss": 0.2955,
+      "epoch": 2.5208130608714456,
+      "grad_norm": 0.17388522624969482,
+      "learning_rate": 2.5891009277130073e-05,
+      "loss": 0.4072,
       "step": 69945
     },
     {
-      "epoch": 2.46,
-      "learning_rate": 2.685764668294392e-05,
-      "loss": 0.2693,
+      "epoch": 2.52099326053267,
+      "grad_norm": 0.21994316577911377,
+      "learning_rate": 2.5888092991363867e-05,
+      "loss": 0.4174,
       "step": 69950
     },
     {
-      "epoch": 2.46,
-      "learning_rate": 2.6854805824873376e-05,
-      "loss": 0.2529,
+      "epoch": 2.5211734601938947,
+      "grad_norm": 0.17600910365581512,
+      "learning_rate": 2.5885176693497558e-05,
+      "loss": 0.3881,
       "step": 69955
     },
     {
-      "epoch": 2.46,
-      "learning_rate": 2.6851964942719394e-05,
-      "loss": 0.2596,
+      "epoch": 2.5213536598551194,
+      "grad_norm": 0.19850102066993713,
+      "learning_rate": 2.5882260383570906e-05,
+      "loss": 0.3902,
       "step": 69960
     },
     {
-      "epoch": 2.46,
-      "learning_rate": 2.684912403651888e-05,
-      "loss": 0.2836,
+      "epoch": 2.521533859516344,
+      "grad_norm": 0.20933623611927032,
+      "learning_rate": 2.587934406162361e-05,
+      "loss": 0.3793,
       "step": 69965
     },
     {
-      "epoch": 2.46,
-      "learning_rate": 2.684628310630869e-05,
-      "loss": 0.2747,
+      "epoch": 2.521714059177569,
+      "grad_norm": 0.2151513695716858,
+      "learning_rate": 2.5876427727695433e-05,
+      "loss": 0.3761,
       "step": 69970
     },
     {
-      "epoch": 2.46,
-      "learning_rate": 2.6843442152125748e-05,
-      "loss": 0.2686,
+      "epoch": 2.5218942588387936,
+      "grad_norm": 0.19325150549411774,
+      "learning_rate": 2.5873511381826087e-05,
+      "loss": 0.3629,
       "step": 69975
     },
     {
-      "epoch": 2.46,
-      "learning_rate": 2.6840601174006918e-05,
-      "loss": 0.2522,
+      "epoch": 2.522074458500018,
+      "grad_norm": 0.1916101574897766,
+      "learning_rate": 2.5870595024055328e-05,
+      "loss": 0.3827,
       "step": 69980
     },
     {
-      "epoch": 2.46,
-      "learning_rate": 2.6837760171989094e-05,
-      "loss": 0.2775,
+      "epoch": 2.5222546581612426,
+      "grad_norm": 0.17457756400108337,
+      "learning_rate": 2.586767865442288e-05,
+      "loss": 0.4084,
       "step": 69985
     },
     {
-      "epoch": 2.46,
-      "learning_rate": 2.6834919146109167e-05,
-      "loss": 0.2693,
+      "epoch": 2.5224348578224673,
+      "grad_norm": 0.19067886471748352,
+      "learning_rate": 2.586476227296847e-05,
+      "loss": 0.3825,
       "step": 69990
     },
     {
-      "epoch": 2.46,
-      "learning_rate": 2.683207809640402e-05,
-      "loss": 0.2827,
+      "epoch": 2.522615057483692,
+      "grad_norm": 0.18429118394851685,
+      "learning_rate": 2.586184587973185e-05,
+      "loss": 0.3885,
       "step": 69995
     },
     {
-      "epoch": 2.46,
-      "learning_rate": 2.6829237022910552e-05,
-      "loss": 0.238,
+      "epoch": 2.5227952571449164,
+      "grad_norm": 0.18464790284633636,
+      "learning_rate": 2.5858929474752734e-05,
+      "loss": 0.4124,
       "step": 70000
     },
     {
-      "epoch": 2.46,
-      "eval_loss": 0.26445090770721436,
-      "eval_runtime": 10.5622,
-      "eval_samples_per_second": 9.468,
-      "eval_steps_per_second": 9.468,
+      "epoch": 2.5227952571449164,
+      "eval_loss": 0.43379154801368713,
+      "eval_runtime": 3.5327,
+      "eval_samples_per_second": 28.307,
+      "eval_steps_per_second": 7.077,
       "step": 70000
     },
     {
-      "epoch": 2.46,
-      "learning_rate": 2.682639592566565e-05,
-      "loss": 0.2826,
+      "epoch": 2.522975456806141,
+      "grad_norm": 0.20850814878940582,
+      "learning_rate": 2.5856013058070888e-05,
+      "loss": 0.3697,
       "step": 70005
     },
     {
-      "epoch": 2.46,
-      "learning_rate": 2.6823554804706192e-05,
-      "loss": 0.2633,
+      "epoch": 2.523155656467366,
+      "grad_norm": 0.23533137142658234,
+      "learning_rate": 2.5853096629726022e-05,
+      "loss": 0.4432,
       "step": 70010
     },
     {
-      "epoch": 2.46,
-      "learning_rate": 2.6820713660069087e-05,
-      "loss": 0.2642,
+      "epoch": 2.5233358561285906,
+      "grad_norm": 0.22087322175502777,
+      "learning_rate": 2.5850180189757878e-05,
+      "loss": 0.3875,
       "step": 70015
     },
     {
-      "epoch": 2.46,
-      "learning_rate": 2.68178724917912e-05,
-      "loss": 0.2651,
+      "epoch": 2.5235160557898153,
+      "grad_norm": 0.23013746738433838,
+      "learning_rate": 2.584726373820619e-05,
+      "loss": 0.3855,
       "step": 70020
     },
     {
-      "epoch": 2.46,
-      "learning_rate": 2.681503129990945e-05,
-      "loss": 0.2732,
+      "epoch": 2.5236962554510396,
+      "grad_norm": 0.17639127373695374,
+      "learning_rate": 2.5844347275110702e-05,
+      "loss": 0.3714,
       "step": 70025
     },
     {
-      "epoch": 2.46,
-      "learning_rate": 2.6812190084460704e-05,
-      "loss": 0.2791,
+      "epoch": 2.5238764551122643,
+      "grad_norm": 0.18888616561889648,
+      "learning_rate": 2.5841430800511145e-05,
+      "loss": 0.4043,
       "step": 70030
     },
     {
-      "epoch": 2.46,
-      "learning_rate": 2.6809348845481873e-05,
-      "loss": 0.2664,
+      "epoch": 2.524056654773489,
+      "grad_norm": 0.2257877141237259,
+      "learning_rate": 2.5838514314447255e-05,
+      "loss": 0.3873,
       "step": 70035
     },
     {
-      "epoch": 2.46,
-      "learning_rate": 2.6806507583009837e-05,
-      "loss": 0.2579,
+      "epoch": 2.524236854434714,
+      "grad_norm": 0.20586729049682617,
+      "learning_rate": 2.5835597816958774e-05,
+      "loss": 0.406,
       "step": 70040
     },
     {
-      "epoch": 2.46,
-      "learning_rate": 2.680366629708149e-05,
-      "loss": 0.2769,
+      "epoch": 2.524417054095938,
+      "grad_norm": 0.2077612429857254,
+      "learning_rate": 2.583268130808543e-05,
+      "loss": 0.3753,
       "step": 70045
     },
     {
-      "epoch": 2.46,
-      "learning_rate": 2.680082498773372e-05,
-      "loss": 0.2742,
+      "epoch": 2.524597253757163,
+      "grad_norm": 0.17256224155426025,
+      "learning_rate": 2.5829764787866974e-05,
+      "loss": 0.35,
       "step": 70050
     },
     {
-      "epoch": 2.46,
-      "learning_rate": 2.6797983655003433e-05,
-      "loss": 0.2812,
+      "epoch": 2.5247774534183876,
+      "grad_norm": 0.28003284335136414,
+      "learning_rate": 2.5826848256343116e-05,
+      "loss": 0.4164,
       "step": 70055
     },
     {
-      "epoch": 2.46,
-      "learning_rate": 2.6795142298927507e-05,
-      "loss": 0.2794,
+      "epoch": 2.5249576530796123,
+      "grad_norm": 0.234117329120636,
+      "learning_rate": 2.582393171355363e-05,
+      "loss": 0.3949,
       "step": 70060
     },
     {
-      "epoch": 2.47,
-      "learning_rate": 2.6792300919542845e-05,
-      "loss": 0.2661,
+      "epoch": 2.525137852740837,
+      "grad_norm": 0.21253302693367004,
+      "learning_rate": 2.582101515953822e-05,
+      "loss": 0.3964,
       "step": 70065
     },
     {
-      "epoch": 2.47,
-      "learning_rate": 2.6789459516886334e-05,
-      "loss": 0.2733,
+      "epoch": 2.5253180524020618,
+      "grad_norm": 0.2189071625471115,
+      "learning_rate": 2.5818098594336636e-05,
+      "loss": 0.3859,
       "step": 70070
     },
     {
-      "epoch": 2.47,
-      "learning_rate": 2.678661809099487e-05,
-      "loss": 0.2739,
+      "epoch": 2.525498252063286,
+      "grad_norm": 0.2182706594467163,
+      "learning_rate": 2.5815182017988626e-05,
+      "loss": 0.3934,
       "step": 70075
     },
     {
-      "epoch": 2.47,
-      "learning_rate": 2.6783776641905346e-05,
-      "loss": 0.2775,
+      "epoch": 2.5256784517245108,
+      "grad_norm": 0.1897723227739334,
+      "learning_rate": 2.5812265430533916e-05,
+      "loss": 0.4036,
       "step": 70080
     },
     {
-      "epoch": 2.47,
-      "learning_rate": 2.6780935169654657e-05,
-      "loss": 0.2773,
+      "epoch": 2.5258586513857355,
+      "grad_norm": 0.20511159300804138,
+      "learning_rate": 2.5809348832012253e-05,
+      "loss": 0.4152,
       "step": 70085
     },
     {
-      "epoch": 2.47,
-      "learning_rate": 2.6778093674279708e-05,
-      "loss": 0.2615,
+      "epoch": 2.52603885104696,
+      "grad_norm": 0.21636447310447693,
+      "learning_rate": 2.5806432222463356e-05,
+      "loss": 0.379,
       "step": 70090
     },
     {
-      "epoch": 2.47,
-      "learning_rate": 2.6775252155817375e-05,
-      "loss": 0.3045,
+      "epoch": 2.5262190507081845,
+      "grad_norm": 0.2109370082616806,
+      "learning_rate": 2.580351560192698e-05,
+      "loss": 0.386,
       "step": 70095
     },
     {
-      "epoch": 2.47,
-      "learning_rate": 2.6772410614304572e-05,
-      "loss": 0.2739,
+      "epoch": 2.5263992503694093,
+      "grad_norm": 0.20028775930404663,
+      "learning_rate": 2.5800598970442862e-05,
+      "loss": 0.4101,
       "step": 70100
     },
     {
-      "epoch": 2.47,
-      "learning_rate": 2.6769569049778175e-05,
-      "loss": 0.2818,
+      "epoch": 2.526579450030634,
+      "grad_norm": 0.2018347680568695,
+      "learning_rate": 2.5797682328050744e-05,
+      "loss": 0.3857,
       "step": 70105
     },
     {
-      "epoch": 2.47,
-      "learning_rate": 2.6766727462275098e-05,
-      "loss": 0.2558,
+      "epoch": 2.5267596496918587,
+      "grad_norm": 0.2127305567264557,
+      "learning_rate": 2.579476567479035e-05,
+      "loss": 0.3927,
       "step": 70110
     },
     {
-      "epoch": 2.47,
-      "learning_rate": 2.6763885851832228e-05,
-      "loss": 0.2704,
+      "epoch": 2.5269398493530835,
+      "grad_norm": 0.19696669280529022,
+      "learning_rate": 2.579184901070143e-05,
+      "loss": 0.4074,
       "step": 70115
     },
     {
-      "epoch": 2.47,
-      "learning_rate": 2.6761044218486457e-05,
-      "loss": 0.2763,
+      "epoch": 2.5271200490143078,
+      "grad_norm": 0.2948433458805084,
+      "learning_rate": 2.578893233582372e-05,
+      "loss": 0.3763,
       "step": 70120
     },
     {
-      "epoch": 2.47,
-      "learning_rate": 2.6758202562274693e-05,
-      "loss": 0.2631,
+      "epoch": 2.5273002486755325,
+      "grad_norm": 0.1763259321451187,
+      "learning_rate": 2.578601565019696e-05,
+      "loss": 0.3824,
       "step": 70125
     },
     {
-      "epoch": 2.47,
-      "learning_rate": 2.6755360883233826e-05,
-      "loss": 0.2597,
+      "epoch": 2.5274804483367572,
+      "grad_norm": 0.22030363976955414,
+      "learning_rate": 2.5783098953860883e-05,
+      "loss": 0.424,
       "step": 70130
     },
     {
-      "epoch": 2.47,
-      "learning_rate": 2.675251918140075e-05,
-      "loss": 0.2679,
+      "epoch": 2.5276606479979815,
+      "grad_norm": 0.2222214639186859,
+      "learning_rate": 2.5780182246855245e-05,
+      "loss": 0.421,
       "step": 70135
     },
     {
-      "epoch": 2.47,
-      "learning_rate": 2.6749677456812376e-05,
-      "loss": 0.2464,
+      "epoch": 2.5278408476592062,
+      "grad_norm": 0.2385231852531433,
+      "learning_rate": 2.5777265529219767e-05,
+      "loss": 0.3931,
       "step": 70140
     },
     {
-      "epoch": 2.47,
-      "learning_rate": 2.6746835709505586e-05,
-      "loss": 0.2559,
+      "epoch": 2.528021047320431,
+      "grad_norm": 0.1919146180152893,
+      "learning_rate": 2.577434880099421e-05,
+      "loss": 0.3618,
       "step": 70145
     },
     {
-      "epoch": 2.47,
-      "learning_rate": 2.674399393951729e-05,
-      "loss": 0.2623,
+      "epoch": 2.5282012469816557,
+      "grad_norm": 0.21808387339115143,
+      "learning_rate": 2.5771432062218286e-05,
+      "loss": 0.4243,
       "step": 70150
     },
     {
-      "epoch": 2.47,
-      "learning_rate": 2.6741152146884386e-05,
-      "loss": 0.2735,
+      "epoch": 2.5283814466428804,
+      "grad_norm": 0.1588166505098343,
+      "learning_rate": 2.576851531293176e-05,
+      "loss": 0.3825,
       "step": 70155
     },
     {
-      "epoch": 2.47,
-      "learning_rate": 2.6738310331643755e-05,
-      "loss": 0.2737,
+      "epoch": 2.528561646304105,
+      "grad_norm": 0.20755480229854584,
+      "learning_rate": 2.576559855317435e-05,
+      "loss": 0.4111,
       "step": 70160
     },
     {
-      "epoch": 2.47,
-      "learning_rate": 2.6735468493832322e-05,
-      "loss": 0.2854,
+      "epoch": 2.5287418459653295,
+      "grad_norm": 0.22904981672763824,
+      "learning_rate": 2.5762681782985816e-05,
+      "loss": 0.3904,
       "step": 70165
     },
     {
-      "epoch": 2.47,
-      "learning_rate": 2.6732626633486973e-05,
-      "loss": 0.2754,
+      "epoch": 2.528922045626554,
+      "grad_norm": 0.18714340031147003,
+      "learning_rate": 2.575976500240589e-05,
+      "loss": 0.3692,
       "step": 70170
     },
     {
-      "epoch": 2.47,
-      "learning_rate": 2.6729784750644605e-05,
-      "loss": 0.2867,
+      "epoch": 2.529102245287779,
+      "grad_norm": 0.2257447987794876,
+      "learning_rate": 2.5756848211474306e-05,
+      "loss": 0.398,
       "step": 70175
     },
     {
-      "epoch": 2.47,
-      "learning_rate": 2.672694284534213e-05,
-      "loss": 0.2767,
+      "epoch": 2.5292824449490032,
+      "grad_norm": 0.21989066898822784,
+      "learning_rate": 2.575393141023082e-05,
+      "loss": 0.3885,
       "step": 70180
     },
     {
-      "epoch": 2.47,
-      "learning_rate": 2.672410091761643e-05,
-      "loss": 0.2667,
+      "epoch": 2.529462644610228,
+      "grad_norm": 0.21285291016101837,
+      "learning_rate": 2.5751014598715155e-05,
+      "loss": 0.3877,
       "step": 70185
     },
     {
-      "epoch": 2.47,
-      "learning_rate": 2.672125896750441e-05,
-      "loss": 0.2848,
+      "epoch": 2.5296428442714527,
+      "grad_norm": 0.20206022262573242,
+      "learning_rate": 2.5748097776967073e-05,
+      "loss": 0.4259,
       "step": 70190
     },
     {
-      "epoch": 2.47,
-      "learning_rate": 2.671841699504299e-05,
-      "loss": 0.2816,
+      "epoch": 2.5298230439326774,
+      "grad_norm": 0.18147850036621094,
+      "learning_rate": 2.5745180945026298e-05,
+      "loss": 0.3914,
       "step": 70195
     },
     {
-      "epoch": 2.47,
-      "learning_rate": 2.6715575000269055e-05,
-      "loss": 0.2844,
+      "epoch": 2.530003243593902,
+      "grad_norm": 0.18609163165092468,
+      "learning_rate": 2.574226410293258e-05,
+      "loss": 0.4179,
       "step": 70200
     },
     {
-      "epoch": 2.47,
-      "learning_rate": 2.6712732983219514e-05,
-      "loss": 0.2748,
+      "epoch": 2.530183443255127,
+      "grad_norm": 0.1647646427154541,
+      "learning_rate": 2.573934725072565e-05,
+      "loss": 0.3955,
       "step": 70205
     },
     {
-      "epoch": 2.47,
-      "learning_rate": 2.6709890943931252e-05,
-      "loss": 0.2603,
+      "epoch": 2.530363642916351,
+      "grad_norm": 0.21977296471595764,
+      "learning_rate": 2.573643038844526e-05,
+      "loss": 0.3912,
       "step": 70210
     },
     {
-      "epoch": 2.47,
-      "learning_rate": 2.670704888244119e-05,
-      "loss": 0.2777,
+      "epoch": 2.530543842577576,
+      "grad_norm": 0.18793781101703644,
+      "learning_rate": 2.5733513516131153e-05,
+      "loss": 0.3975,
       "step": 70215
     },
     {
-      "epoch": 2.47,
-      "learning_rate": 2.670420679878622e-05,
-      "loss": 0.2673,
+      "epoch": 2.5307240422388007,
+      "grad_norm": 0.18933290243148804,
+      "learning_rate": 2.573059663382306e-05,
+      "loss": 0.4017,
       "step": 70220
     },
     {
-      "epoch": 2.47,
-      "learning_rate": 2.6701364693003255e-05,
-      "loss": 0.2878,
+      "epoch": 2.5309042419000254,
+      "grad_norm": 0.17556174099445343,
+      "learning_rate": 2.5727679741560734e-05,
+      "loss": 0.4003,
       "step": 70225
     },
     {
-      "epoch": 2.47,
-      "learning_rate": 2.669852256512918e-05,
-      "loss": 0.2912,
+      "epoch": 2.5310844415612497,
+      "grad_norm": 0.18023037910461426,
+      "learning_rate": 2.5724762839383915e-05,
+      "loss": 0.4288,
       "step": 70230
     },
     {
-      "epoch": 2.47,
-      "learning_rate": 2.669568041520092e-05,
-      "loss": 0.2746,
+      "epoch": 2.5312646412224744,
+      "grad_norm": 0.1804884523153305,
+      "learning_rate": 2.5721845927332333e-05,
+      "loss": 0.3854,
       "step": 70235
     },
     {
-      "epoch": 2.47,
-      "learning_rate": 2.6692838243255364e-05,
-      "loss": 0.2432,
+      "epoch": 2.531444840883699,
+      "grad_norm": 0.26398801803588867,
+      "learning_rate": 2.5718929005445746e-05,
+      "loss": 0.4287,
       "step": 70240
     },
     {
-      "epoch": 2.47,
-      "learning_rate": 2.6689996049329413e-05,
-      "loss": 0.2665,
+      "epoch": 2.531625040544924,
+      "grad_norm": 0.2065039724111557,
+      "learning_rate": 2.5716012073763883e-05,
+      "loss": 0.397,
       "step": 70245
     },
     {
-      "epoch": 2.47,
-      "learning_rate": 2.6687153833459982e-05,
-      "loss": 0.2876,
+      "epoch": 2.5318052402061486,
+      "grad_norm": 0.1797478199005127,
+      "learning_rate": 2.5713095132326515e-05,
+      "loss": 0.3692,
       "step": 70250
     },
     {
-      "epoch": 2.47,
-      "learning_rate": 2.668431159568397e-05,
-      "loss": 0.3013,
+      "epoch": 2.531985439867373,
+      "grad_norm": 0.19305941462516785,
+      "learning_rate": 2.5710178181173344e-05,
+      "loss": 0.3386,
       "step": 70255
     },
     {
-      "epoch": 2.47,
-      "learning_rate": 2.668146933603828e-05,
-      "loss": 0.2851,
+      "epoch": 2.5321656395285976,
+      "grad_norm": 0.2119571417570114,
+      "learning_rate": 2.5707261220344143e-05,
+      "loss": 0.4244,
       "step": 70260
     },
     {
-      "epoch": 2.47,
-      "learning_rate": 2.6678627054559824e-05,
-      "loss": 0.2648,
+      "epoch": 2.5323458391898224,
+      "grad_norm": 0.21070972084999084,
+      "learning_rate": 2.5704344249878637e-05,
+      "loss": 0.3988,
       "step": 70265
     },
     {
-      "epoch": 2.47,
-      "learning_rate": 2.6675784751285494e-05,
-      "loss": 0.2773,
+      "epoch": 2.532526038851047,
+      "grad_norm": 0.2026742547750473,
+      "learning_rate": 2.570142726981658e-05,
+      "loss": 0.4084,
       "step": 70270
     },
     {
-      "epoch": 2.47,
-      "learning_rate": 2.6672942426252212e-05,
-      "loss": 0.2921,
+      "epoch": 2.5327062385122714,
+      "grad_norm": 0.16865421831607819,
+      "learning_rate": 2.569851028019771e-05,
+      "loss": 0.3662,
       "step": 70275
     },
     {
-      "epoch": 2.47,
-      "learning_rate": 2.667010007949687e-05,
-      "loss": 0.2451,
+      "epoch": 2.532886438173496,
+      "grad_norm": 0.23617038130760193,
+      "learning_rate": 2.5695593281061774e-05,
+      "loss": 0.3986,
       "step": 70280
     },
     {
-      "epoch": 2.47,
-      "learning_rate": 2.666725771105638e-05,
-      "loss": 0.2775,
+      "epoch": 2.533066637834721,
+      "grad_norm": 0.2046816498041153,
+      "learning_rate": 2.5692676272448517e-05,
+      "loss": 0.3803,
       "step": 70285
     },
     {
-      "epoch": 2.47,
-      "learning_rate": 2.6664415320967644e-05,
-      "loss": 0.2334,
+      "epoch": 2.5332468374959456,
+      "grad_norm": 0.18710558116436005,
+      "learning_rate": 2.5689759254397683e-05,
+      "loss": 0.3968,
       "step": 70290
     },
     {
-      "epoch": 2.47,
-      "learning_rate": 2.6661572909267574e-05,
-      "loss": 0.2681,
+      "epoch": 2.5334270371571703,
+      "grad_norm": 0.21419739723205566,
+      "learning_rate": 2.5686842226949008e-05,
+      "loss": 0.4224,
       "step": 70295
     },
     {
-      "epoch": 2.47,
-      "learning_rate": 2.6658730475993077e-05,
-      "loss": 0.2791,
+      "epoch": 2.533607236818395,
+      "grad_norm": 0.18646739423274994,
+      "learning_rate": 2.568392519014224e-05,
+      "loss": 0.4011,
       "step": 70300
     },
     {
-      "epoch": 2.47,
-      "learning_rate": 2.6655888021181052e-05,
-      "loss": 0.2672,
+      "epoch": 2.5337874364796193,
+      "grad_norm": 0.19873277842998505,
+      "learning_rate": 2.5681008144017128e-05,
+      "loss": 0.3635,
       "step": 70305
     },
     {
-      "epoch": 2.47,
-      "learning_rate": 2.665304554486842e-05,
-      "loss": 0.2755,
+      "epoch": 2.533967636140844,
+      "grad_norm": 0.18021991848945618,
+      "learning_rate": 2.5678091088613408e-05,
+      "loss": 0.3867,
       "step": 70310
     },
     {
-      "epoch": 2.47,
-      "learning_rate": 2.665020304709208e-05,
-      "loss": 0.2966,
+      "epoch": 2.534147835802069,
+      "grad_norm": 0.17604483664035797,
+      "learning_rate": 2.5675174023970826e-05,
+      "loss": 0.3857,
       "step": 70315
     },
     {
-      "epoch": 2.47,
-      "learning_rate": 2.6647360527888943e-05,
-      "loss": 0.2693,
+      "epoch": 2.534328035463293,
+      "grad_norm": 0.2240837961435318,
+      "learning_rate": 2.567225695012913e-05,
+      "loss": 0.3711,
       "step": 70320
     },
     {
-      "epoch": 2.47,
-      "learning_rate": 2.6644517987295903e-05,
-      "loss": 0.2733,
+      "epoch": 2.534508235124518,
+      "grad_norm": 0.1983606368303299,
+      "learning_rate": 2.566933986712806e-05,
+      "loss": 0.3828,
       "step": 70325
     },
     {
-      "epoch": 2.47,
-      "learning_rate": 2.6641675425349894e-05,
-      "loss": 0.263,
+      "epoch": 2.5346884347857426,
+      "grad_norm": 0.22788797318935394,
+      "learning_rate": 2.566642277500738e-05,
+      "loss": 0.3859,
       "step": 70330
     },
     {
-      "epoch": 2.47,
-      "learning_rate": 2.6638832842087807e-05,
-      "loss": 0.2887,
+      "epoch": 2.5348686344469673,
+      "grad_norm": 0.2181435376405716,
+      "learning_rate": 2.5663505673806805e-05,
+      "loss": 0.4207,
       "step": 70335
     },
     {
-      "epoch": 2.47,
-      "learning_rate": 2.663599023754656e-05,
-      "loss": 0.2671,
+      "epoch": 2.535048834108192,
+      "grad_norm": 0.19292283058166504,
+      "learning_rate": 2.5660588563566097e-05,
+      "loss": 0.4102,
       "step": 70340
     },
     {
-      "epoch": 2.47,
-      "learning_rate": 2.6633147611763047e-05,
-      "loss": 0.2613,
+      "epoch": 2.5352290337694168,
+      "grad_norm": 0.18651723861694336,
+      "learning_rate": 2.5657671444324994e-05,
+      "loss": 0.4028,
       "step": 70345
     },
     {
-      "epoch": 2.48,
-      "learning_rate": 2.6630304964774195e-05,
-      "loss": 0.2824,
+      "epoch": 2.535409233430641,
+      "grad_norm": 0.20241175591945648,
+      "learning_rate": 2.5654754316123248e-05,
+      "loss": 0.4,
       "step": 70350
     },
     {
-      "epoch": 2.48,
-      "learning_rate": 2.6627462296616906e-05,
-      "loss": 0.2821,
+      "epoch": 2.535589433091866,
+      "grad_norm": 0.19530391693115234,
+      "learning_rate": 2.5651837179000605e-05,
+      "loss": 0.3573,
       "step": 70355
     },
     {
-      "epoch": 2.48,
-      "learning_rate": 2.6624619607328093e-05,
-      "loss": 0.2637,
+      "epoch": 2.5357696327530905,
+      "grad_norm": 0.1831565499305725,
+      "learning_rate": 2.5648920032996794e-05,
+      "loss": 0.3696,
       "step": 70360
     },
     {
-      "epoch": 2.48,
-      "learning_rate": 2.6621776896944666e-05,
-      "loss": 0.283,
+      "epoch": 2.535949832414315,
+      "grad_norm": 0.22338707745075226,
+      "learning_rate": 2.5646002878151586e-05,
+      "loss": 0.3642,
       "step": 70365
     },
     {
-      "epoch": 2.48,
-      "learning_rate": 2.661893416550354e-05,
-      "loss": 0.2689,
+      "epoch": 2.5361300320755396,
+      "grad_norm": 0.23138903081417084,
+      "learning_rate": 2.5643669147936305e-05,
+      "loss": 0.3546,
       "step": 70370
     },
     {
-      "epoch": 2.48,
-      "learning_rate": 2.661609141304161e-05,
-      "loss": 0.2667,
+      "epoch": 2.5363102317367643,
+      "grad_norm": 0.18024463951587677,
+      "learning_rate": 2.5640751977276717e-05,
+      "loss": 0.377,
       "step": 70375
     },
     {
-      "epoch": 2.48,
-      "learning_rate": 2.66132486395958e-05,
-      "loss": 0.2465,
+      "epoch": 2.536490431397989,
+      "grad_norm": 0.19997356832027435,
+      "learning_rate": 2.5637834797887005e-05,
+      "loss": 0.3983,
       "step": 70380
     },
     {
-      "epoch": 2.48,
-      "learning_rate": 2.661040584520302e-05,
-      "loss": 0.2643,
+      "epoch": 2.5366706310592138,
+      "grad_norm": 0.20950496196746826,
+      "learning_rate": 2.5634917609806917e-05,
+      "loss": 0.3991,
       "step": 70385
     },
     {
-      "epoch": 2.48,
-      "learning_rate": 2.660756302990019e-05,
-      "loss": 0.2812,
+      "epoch": 2.5368508307204385,
+      "grad_norm": 0.20799300074577332,
+      "learning_rate": 2.56320004130762e-05,
+      "loss": 0.3923,
       "step": 70390
     },
     {
-      "epoch": 2.48,
-      "learning_rate": 2.66047201937242e-05,
-      "loss": 0.2553,
+      "epoch": 2.5370310303816628,
+      "grad_norm": 0.23899616301059723,
+      "learning_rate": 2.5629083207734595e-05,
+      "loss": 0.4264,
       "step": 70395
     },
     {
-      "epoch": 2.48,
-      "learning_rate": 2.660187733671199e-05,
-      "loss": 0.2678,
+      "epoch": 2.5372112300428875,
+      "grad_norm": 0.18154381215572357,
+      "learning_rate": 2.5626165993821866e-05,
+      "loss": 0.3814,
       "step": 70400
     },
     {
-      "epoch": 2.48,
-      "learning_rate": 2.6599034458900452e-05,
-      "loss": 0.2585,
+      "epoch": 2.5373914297041122,
+      "grad_norm": 0.1529223769903183,
+      "learning_rate": 2.5623248771377733e-05,
+      "loss": 0.3888,
       "step": 70405
     },
     {
-      "epoch": 2.48,
-      "learning_rate": 2.65961915603265e-05,
-      "loss": 0.2741,
+      "epoch": 2.5375716293653365,
+      "grad_norm": 0.18600383400917053,
+      "learning_rate": 2.5620331540441956e-05,
+      "loss": 0.3829,
       "step": 70410
     },
     {
-      "epoch": 2.48,
-      "learning_rate": 2.6593348641027056e-05,
-      "loss": 0.2839,
+      "epoch": 2.5377518290265613,
+      "grad_norm": 0.17343242466449738,
+      "learning_rate": 2.5617414301054288e-05,
+      "loss": 0.3795,
       "step": 70415
     },
     {
-      "epoch": 2.48,
-      "learning_rate": 2.6590505701039032e-05,
-      "loss": 0.2588,
+      "epoch": 2.537932028687786,
+      "grad_norm": 0.33770492672920227,
+      "learning_rate": 2.5614497053254464e-05,
+      "loss": 0.3925,
       "step": 70420
     },
     {
-      "epoch": 2.48,
-      "learning_rate": 2.6587662740399338e-05,
-      "loss": 0.2524,
+      "epoch": 2.5381122283490107,
+      "grad_norm": 0.20434433221817017,
+      "learning_rate": 2.5611579797082252e-05,
+      "loss": 0.3664,
       "step": 70425
     },
     {
-      "epoch": 2.48,
-      "learning_rate": 2.6584819759144892e-05,
-      "loss": 0.2595,
+      "epoch": 2.5382924280102355,
+      "grad_norm": 0.2171066850423813,
+      "learning_rate": 2.5608662532577367e-05,
+      "loss": 0.403,
       "step": 70430
     },
     {
-      "epoch": 2.48,
-      "learning_rate": 2.65819767573126e-05,
-      "loss": 0.2641,
+      "epoch": 2.53847262767146,
+      "grad_norm": 0.21163609623908997,
+      "learning_rate": 2.5605745259779578e-05,
+      "loss": 0.4073,
       "step": 70435
     },
     {
-      "epoch": 2.48,
-      "learning_rate": 2.6579133734939383e-05,
-      "loss": 0.2606,
+      "epoch": 2.5386528273326845,
+      "grad_norm": 0.19679218530654907,
+      "learning_rate": 2.5602827978728626e-05,
+      "loss": 0.3859,
       "step": 70440
     },
     {
-      "epoch": 2.48,
-      "learning_rate": 2.6576290692062162e-05,
-      "loss": 0.2499,
+      "epoch": 2.5388330269939092,
+      "grad_norm": 0.19141094386577606,
+      "learning_rate": 2.5599910689464263e-05,
+      "loss": 0.3611,
       "step": 70445
     },
     {
-      "epoch": 2.48,
-      "learning_rate": 2.657344762871784e-05,
-      "loss": 0.2609,
+      "epoch": 2.539013226655134,
+      "grad_norm": 0.18265005946159363,
+      "learning_rate": 2.559699339202623e-05,
+      "loss": 0.4121,
       "step": 70450
     },
     {
-      "epoch": 2.48,
-      "learning_rate": 2.6570604544943346e-05,
-      "loss": 0.2638,
+      "epoch": 2.5391934263163582,
+      "grad_norm": 0.1961100548505783,
+      "learning_rate": 2.559407608645428e-05,
+      "loss": 0.387,
       "step": 70455
     },
     {
-      "epoch": 2.48,
-      "learning_rate": 2.656776144077558e-05,
-      "loss": 0.2473,
+      "epoch": 2.539373625977583,
+      "grad_norm": 0.1709757298231125,
+      "learning_rate": 2.559115877278816e-05,
+      "loss": 0.3952,
       "step": 70460
     },
     {
-      "epoch": 2.48,
-      "learning_rate": 2.6564918316251463e-05,
-      "loss": 0.2855,
+      "epoch": 2.5395538256388077,
+      "grad_norm": 0.193019837141037,
+      "learning_rate": 2.558824145106761e-05,
+      "loss": 0.4121,
       "step": 70465
     },
     {
-      "epoch": 2.48,
-      "learning_rate": 2.6562075171407914e-05,
-      "loss": 0.2692,
+      "epoch": 2.5397340253000324,
+      "grad_norm": 0.23720510303974152,
+      "learning_rate": 2.5585324121332394e-05,
+      "loss": 0.3731,
       "step": 70470
     },
     {
-      "epoch": 2.48,
-      "learning_rate": 2.6559232006281853e-05,
-      "loss": 0.2746,
+      "epoch": 2.539914224961257,
+      "grad_norm": 0.21066009998321533,
+      "learning_rate": 2.558240678362224e-05,
+      "loss": 0.4199,
       "step": 70475
     },
     {
-      "epoch": 2.48,
-      "learning_rate": 2.6556388820910183e-05,
-      "loss": 0.2747,
+      "epoch": 2.540094424622482,
+      "grad_norm": 0.23059865832328796,
+      "learning_rate": 2.557948943797691e-05,
+      "loss": 0.3951,
       "step": 70480
     },
     {
-      "epoch": 2.48,
-      "learning_rate": 2.6553545615329838e-05,
-      "loss": 0.2568,
+      "epoch": 2.540274624283706,
+      "grad_norm": 0.19941085577011108,
+      "learning_rate": 2.5576572084436153e-05,
+      "loss": 0.4155,
       "step": 70485
     },
     {
-      "epoch": 2.48,
-      "learning_rate": 2.655070238957772e-05,
-      "loss": 0.2693,
+      "epoch": 2.540454823944931,
+      "grad_norm": 0.20750625431537628,
+      "learning_rate": 2.5573654723039704e-05,
+      "loss": 0.3913,
       "step": 70490
     },
     {
-      "epoch": 2.48,
-      "learning_rate": 2.6547859143690762e-05,
-      "loss": 0.265,
+      "epoch": 2.5406350236061557,
+      "grad_norm": 0.1879345029592514,
+      "learning_rate": 2.5570737353827323e-05,
+      "loss": 0.4201,
       "step": 70495
     },
     {
-      "epoch": 2.48,
-      "learning_rate": 2.6545015877705865e-05,
-      "loss": 0.2696,
+      "epoch": 2.5408152232673804,
+      "grad_norm": 0.217643141746521,
+      "learning_rate": 2.5567819976838752e-05,
+      "loss": 0.3992,
       "step": 70500
     },
     {
-      "epoch": 2.48,
-      "eval_loss": 0.2644537389278412,
-      "eval_runtime": 10.5368,
-      "eval_samples_per_second": 9.491,
-      "eval_steps_per_second": 9.491,
+      "epoch": 2.5408152232673804,
+      "eval_loss": 0.4338361620903015,
+      "eval_runtime": 3.5352,
+      "eval_samples_per_second": 28.287,
+      "eval_steps_per_second": 7.072,
       "step": 70500
     },
     {
-      "epoch": 2.48,
-      "learning_rate": 2.6542172591659962e-05,
-      "loss": 0.2577,
+      "epoch": 2.5409954229286047,
+      "grad_norm": 0.24064627289772034,
+      "learning_rate": 2.556490259211376e-05,
+      "loss": 0.3785,
       "step": 70505
     },
     {
-      "epoch": 2.48,
-      "learning_rate": 2.6539329285589963e-05,
-      "loss": 0.2593,
+      "epoch": 2.5411756225898294,
+      "grad_norm": 0.18428386747837067,
+      "learning_rate": 2.5561985199692062e-05,
+      "loss": 0.4018,
       "step": 70510
     },
     {
-      "epoch": 2.48,
-      "learning_rate": 2.653648595953278e-05,
-      "loss": 0.2707,
+      "epoch": 2.541355822251054,
+      "grad_norm": 0.20441985130310059,
+      "learning_rate": 2.5559067799613434e-05,
+      "loss": 0.3607,
       "step": 70515
     },
     {
-      "epoch": 2.48,
-      "learning_rate": 2.6533642613525345e-05,
-      "loss": 0.2658,
+      "epoch": 2.541536021912279,
+      "grad_norm": 0.20763367414474487,
+      "learning_rate": 2.555615039191761e-05,
+      "loss": 0.3537,
       "step": 70520
     },
     {
-      "epoch": 2.48,
-      "learning_rate": 2.6530799247604577e-05,
-      "loss": 0.2745,
+      "epoch": 2.5417162215735036,
+      "grad_norm": 0.2506287693977356,
+      "learning_rate": 2.5553232976644352e-05,
+      "loss": 0.382,
       "step": 70525
     },
     {
-      "epoch": 2.48,
-      "learning_rate": 2.6527955861807385e-05,
-      "loss": 0.2485,
+      "epoch": 2.541896421234728,
+      "grad_norm": 0.2257665991783142,
+      "learning_rate": 2.55503155538334e-05,
+      "loss": 0.4206,
       "step": 70530
     },
     {
-      "epoch": 2.48,
-      "learning_rate": 2.6525112456170692e-05,
-      "loss": 0.2742,
+      "epoch": 2.5420766208959527,
+      "grad_norm": 0.2143712192773819,
+      "learning_rate": 2.5547398123524495e-05,
+      "loss": 0.3677,
       "step": 70535
     },
     {
-      "epoch": 2.48,
-      "learning_rate": 2.6522269030731416e-05,
-      "loss": 0.2785,
+      "epoch": 2.5422568205571774,
+      "grad_norm": 0.21773605048656464,
+      "learning_rate": 2.554448068575741e-05,
+      "loss": 0.4063,
       "step": 70540
     },
     {
-      "epoch": 2.48,
-      "learning_rate": 2.651942558552648e-05,
-      "loss": 0.2837,
+      "epoch": 2.542437020218402,
+      "grad_norm": 0.19206026196479797,
+      "learning_rate": 2.5541563240571877e-05,
+      "loss": 0.4034,
       "step": 70545
     },
     {
-      "epoch": 2.48,
-      "learning_rate": 2.651658212059281e-05,
-      "loss": 0.2881,
+      "epoch": 2.5426172198796264,
+      "grad_norm": 0.19149447977542877,
+      "learning_rate": 2.553864578800765e-05,
+      "loss": 0.377,
       "step": 70550
     },
     {
-      "epoch": 2.48,
-      "learning_rate": 2.6513738635967316e-05,
-      "loss": 0.2577,
+      "epoch": 2.542797419540851,
+      "grad_norm": 0.18062478303909302,
+      "learning_rate": 2.553572832810447e-05,
+      "loss": 0.3932,
       "step": 70555
     },
     {
-      "epoch": 2.48,
-      "learning_rate": 2.6510895131686932e-05,
-      "loss": 0.2778,
+      "epoch": 2.542977619202076,
+      "grad_norm": 0.1797667294740677,
+      "learning_rate": 2.55328108609021e-05,
+      "loss": 0.3771,
       "step": 70560
     },
     {
-      "epoch": 2.48,
-      "learning_rate": 2.650805160778856e-05,
-      "loss": 0.2704,
+      "epoch": 2.5431578188633006,
+      "grad_norm": 0.25208544731140137,
+      "learning_rate": 2.5529893386440295e-05,
+      "loss": 0.4192,
       "step": 70565
     },
     {
-      "epoch": 2.48,
-      "learning_rate": 2.6505208064309134e-05,
-      "loss": 0.2713,
+      "epoch": 2.5433380185245253,
+      "grad_norm": 0.2537004351615906,
+      "learning_rate": 2.552697590475878e-05,
+      "loss": 0.3734,
       "step": 70570
     },
     {
-      "epoch": 2.48,
-      "learning_rate": 2.650236450128557e-05,
-      "loss": 0.2803,
+      "epoch": 2.54351821818575,
+      "grad_norm": 0.27715036273002625,
+      "learning_rate": 2.5524058415897328e-05,
+      "loss": 0.3997,
       "step": 70575
     },
     {
-      "epoch": 2.48,
-      "learning_rate": 2.6499520918754805e-05,
-      "loss": 0.2946,
+      "epoch": 2.5436984178469744,
+      "grad_norm": 0.19086778163909912,
+      "learning_rate": 2.552114091989568e-05,
+      "loss": 0.3765,
       "step": 70580
     },
     {
-      "epoch": 2.48,
-      "learning_rate": 2.6496677316753738e-05,
-      "loss": 0.2721,
+      "epoch": 2.543878617508199,
+      "grad_norm": 0.22903601825237274,
+      "learning_rate": 2.5518223416793592e-05,
+      "loss": 0.4433,
       "step": 70585
     },
     {
-      "epoch": 2.48,
-      "learning_rate": 2.6493833695319308e-05,
-      "loss": 0.2775,
+      "epoch": 2.544058817169424,
+      "grad_norm": 0.21384993195533752,
+      "learning_rate": 2.5515305906630805e-05,
+      "loss": 0.4386,
       "step": 70590
     },
     {
-      "epoch": 2.48,
-      "learning_rate": 2.649099005448843e-05,
-      "loss": 0.2534,
+      "epoch": 2.544239016830648,
+      "grad_norm": 0.2099718600511551,
+      "learning_rate": 2.5512388389447074e-05,
+      "loss": 0.4044,
       "step": 70595
     },
     {
-      "epoch": 2.48,
-      "learning_rate": 2.648814639429803e-05,
-      "loss": 0.2831,
+      "epoch": 2.544419216491873,
+      "grad_norm": 0.20232221484184265,
+      "learning_rate": 2.5509470865282155e-05,
+      "loss": 0.4041,
       "step": 70600
     },
     {
-      "epoch": 2.48,
-      "learning_rate": 2.6485302714785027e-05,
-      "loss": 0.2819,
+      "epoch": 2.5445994161530976,
+      "grad_norm": 0.19650457799434662,
+      "learning_rate": 2.550655333417579e-05,
+      "loss": 0.3656,
       "step": 70605
     },
     {
-      "epoch": 2.48,
-      "learning_rate": 2.6482459015986348e-05,
-      "loss": 0.256,
+      "epoch": 2.5447796158143223,
+      "grad_norm": 0.25365570187568665,
+      "learning_rate": 2.550363579616774e-05,
+      "loss": 0.3818,
       "step": 70610
     },
     {
-      "epoch": 2.48,
-      "learning_rate": 2.6479615297938914e-05,
-      "loss": 0.2702,
+      "epoch": 2.544959815475547,
+      "grad_norm": 0.19008086621761322,
+      "learning_rate": 2.5500718251297746e-05,
+      "loss": 0.4,
       "step": 70615
     },
     {
-      "epoch": 2.48,
-      "learning_rate": 2.6476771560679657e-05,
-      "loss": 0.2729,
+      "epoch": 2.545140015136772,
+      "grad_norm": 0.17244011163711548,
+      "learning_rate": 2.5497800699605563e-05,
+      "loss": 0.4169,
       "step": 70620
     },
     {
-      "epoch": 2.48,
-      "learning_rate": 2.647392780424549e-05,
-      "loss": 0.2365,
+      "epoch": 2.545320214797996,
+      "grad_norm": 0.21005220711231232,
+      "learning_rate": 2.5494883141130938e-05,
+      "loss": 0.3926,
       "step": 70625
     },
     {
-      "epoch": 2.48,
-      "learning_rate": 2.6471084028673344e-05,
-      "loss": 0.2691,
+      "epoch": 2.545500414459221,
+      "grad_norm": 0.15076076984405518,
+      "learning_rate": 2.5491965575913635e-05,
+      "loss": 0.3612,
       "step": 70630
     },
     {
-      "epoch": 2.49,
-      "learning_rate": 2.6468240234000134e-05,
-      "loss": 0.2795,
+      "epoch": 2.5456806141204456,
+      "grad_norm": 0.2208922952413559,
+      "learning_rate": 2.548904800399339e-05,
+      "loss": 0.4058,
       "step": 70635
     },
     {
-      "epoch": 2.49,
-      "learning_rate": 2.64653964202628e-05,
-      "loss": 0.2887,
+      "epoch": 2.54586081378167,
+      "grad_norm": 0.2186022400856018,
+      "learning_rate": 2.548613042540996e-05,
+      "loss": 0.4301,
       "step": 70640
     },
     {
-      "epoch": 2.49,
-      "learning_rate": 2.6462552587498256e-05,
-      "loss": 0.2906,
+      "epoch": 2.5460410134428946,
+      "grad_norm": 0.2868672311306,
+      "learning_rate": 2.5483212840203097e-05,
+      "loss": 0.3947,
       "step": 70645
     },
     {
-      "epoch": 2.49,
-      "learning_rate": 2.645970873574343e-05,
-      "loss": 0.2476,
+      "epoch": 2.5462212131041193,
+      "grad_norm": 0.22895467281341553,
+      "learning_rate": 2.5480295248412555e-05,
+      "loss": 0.4076,
       "step": 70650
     },
     {
-      "epoch": 2.49,
-      "learning_rate": 2.6456864865035246e-05,
-      "loss": 0.2553,
+      "epoch": 2.546401412765344,
+      "grad_norm": 0.23754490911960602,
+      "learning_rate": 2.5477377650078087e-05,
+      "loss": 0.3852,
       "step": 70655
     },
     {
-      "epoch": 2.49,
-      "learning_rate": 2.6454020975410636e-05,
-      "loss": 0.2771,
+      "epoch": 2.5465816124265688,
+      "grad_norm": 0.17783097922801971,
+      "learning_rate": 2.5474460045239435e-05,
+      "loss": 0.408,
       "step": 70660
     },
     {
-      "epoch": 2.49,
-      "learning_rate": 2.6451177066906525e-05,
-      "loss": 0.2717,
+      "epoch": 2.5467618120877935,
+      "grad_norm": 0.21512363851070404,
+      "learning_rate": 2.5471542433936358e-05,
+      "loss": 0.4015,
       "step": 70665
     },
     {
-      "epoch": 2.49,
-      "learning_rate": 2.6448333139559834e-05,
-      "loss": 0.2877,
+      "epoch": 2.546942011749018,
+      "grad_norm": 0.2190975397825241,
+      "learning_rate": 2.546862481620861e-05,
+      "loss": 0.3845,
       "step": 70670
     },
     {
-      "epoch": 2.49,
-      "learning_rate": 2.644548919340749e-05,
-      "loss": 0.2463,
+      "epoch": 2.5471222114102425,
+      "grad_norm": 0.19571487605571747,
+      "learning_rate": 2.5465707192095927e-05,
+      "loss": 0.4478,
       "step": 70675
     },
     {
-      "epoch": 2.49,
-      "learning_rate": 2.644264522848642e-05,
-      "loss": 0.2666,
+      "epoch": 2.5473024110714673,
+      "grad_norm": 0.2442859262228012,
+      "learning_rate": 2.546278956163809e-05,
+      "loss": 0.3756,
       "step": 70680
     },
     {
-      "epoch": 2.49,
-      "learning_rate": 2.6439801244833552e-05,
-      "loss": 0.2872,
+      "epoch": 2.5474826107326916,
+      "grad_norm": 0.20132243633270264,
+      "learning_rate": 2.5459871924874822e-05,
+      "loss": 0.3883,
       "step": 70685
     },
     {
-      "epoch": 2.49,
-      "learning_rate": 2.6436957242485817e-05,
-      "loss": 0.255,
+      "epoch": 2.5476628103939163,
+      "grad_norm": 0.22069872915744781,
+      "learning_rate": 2.54569542818459e-05,
+      "loss": 0.4054,
       "step": 70690
     },
     {
-      "epoch": 2.49,
-      "learning_rate": 2.643411322148014e-05,
-      "loss": 0.2743,
+      "epoch": 2.547843010055141,
+      "grad_norm": 0.1703534722328186,
+      "learning_rate": 2.5454036632591055e-05,
+      "loss": 0.4059,
       "step": 70695
     },
     {
-      "epoch": 2.49,
-      "learning_rate": 2.643126918185344e-05,
-      "loss": 0.2633,
+      "epoch": 2.5480232097163658,
+      "grad_norm": 0.1875164955854416,
+      "learning_rate": 2.5451118977150053e-05,
+      "loss": 0.3853,
       "step": 70700
     },
     {
-      "epoch": 2.49,
-      "learning_rate": 2.6428425123642664e-05,
-      "loss": 0.2505,
+      "epoch": 2.5482034093775905,
+      "grad_norm": 0.22562925517559052,
+      "learning_rate": 2.544820131556264e-05,
+      "loss": 0.4122,
       "step": 70705
     },
     {
-      "epoch": 2.49,
-      "learning_rate": 2.6425581046884718e-05,
-      "loss": 0.2791,
+      "epoch": 2.5483836090388152,
+      "grad_norm": 0.1860308051109314,
+      "learning_rate": 2.5445283647868574e-05,
+      "loss": 0.3849,
       "step": 70710
     },
     {
-      "epoch": 2.49,
-      "learning_rate": 2.642273695161655e-05,
-      "loss": 0.2804,
+      "epoch": 2.5485638087000395,
+      "grad_norm": 0.21737992763519287,
+      "learning_rate": 2.54423659741076e-05,
+      "loss": 0.4249,
       "step": 70715
     },
     {
-      "epoch": 2.49,
-      "learning_rate": 2.6419892837875077e-05,
-      "loss": 0.2931,
+      "epoch": 2.5487440083612642,
+      "grad_norm": 0.2317013442516327,
+      "learning_rate": 2.543944829431948e-05,
+      "loss": 0.3782,
       "step": 70720
     },
     {
-      "epoch": 2.49,
-      "learning_rate": 2.6417048705697232e-05,
-      "loss": 0.276,
+      "epoch": 2.548924208022489,
+      "grad_norm": 0.183481365442276,
+      "learning_rate": 2.543653060854396e-05,
+      "loss": 0.4066,
       "step": 70725
     },
     {
-      "epoch": 2.49,
-      "learning_rate": 2.641420455511994e-05,
-      "loss": 0.2728,
+      "epoch": 2.5491044076837137,
+      "grad_norm": 0.2388121783733368,
+      "learning_rate": 2.5433612916820798e-05,
+      "loss": 0.3918,
       "step": 70730
     },
     {
-      "epoch": 2.49,
-      "learning_rate": 2.641136038618014e-05,
-      "loss": 0.2851,
+      "epoch": 2.549284607344938,
+      "grad_norm": 0.22414498031139374,
+      "learning_rate": 2.5430695219189738e-05,
+      "loss": 0.4132,
       "step": 70735
     },
     {
-      "epoch": 2.49,
-      "learning_rate": 2.640851619891475e-05,
-      "loss": 0.2897,
+      "epoch": 2.5494648070061627,
+      "grad_norm": 0.18682889640331268,
+      "learning_rate": 2.5427777515690543e-05,
+      "loss": 0.4106,
       "step": 70740
     },
     {
-      "epoch": 2.49,
-      "learning_rate": 2.6405671993360716e-05,
-      "loss": 0.2527,
+      "epoch": 2.5496450066673875,
+      "grad_norm": 0.18605414032936096,
+      "learning_rate": 2.5424859806362954e-05,
+      "loss": 0.3922,
       "step": 70745
     },
     {
-      "epoch": 2.49,
-      "learning_rate": 2.6402827769554943e-05,
-      "loss": 0.2513,
+      "epoch": 2.549825206328612,
+      "grad_norm": 0.19950343668460846,
+      "learning_rate": 2.542194209124675e-05,
+      "loss": 0.3529,
       "step": 70750
     },
     {
-      "epoch": 2.49,
-      "learning_rate": 2.6399983527534388e-05,
-      "loss": 0.2765,
+      "epoch": 2.550005405989837,
+      "grad_norm": 0.22607798874378204,
+      "learning_rate": 2.541902437038165e-05,
+      "loss": 0.3833,
       "step": 70755
     },
     {
-      "epoch": 2.49,
-      "learning_rate": 2.6397139267335963e-05,
-      "loss": 0.2954,
+      "epoch": 2.5501856056510612,
+      "grad_norm": 0.24269573390483856,
+      "learning_rate": 2.5416106643807434e-05,
+      "loss": 0.4602,
       "step": 70760
     },
     {
-      "epoch": 2.49,
-      "learning_rate": 2.6394294988996604e-05,
-      "loss": 0.2749,
+      "epoch": 2.550365805312286,
+      "grad_norm": 0.22042149305343628,
+      "learning_rate": 2.541318891156384e-05,
+      "loss": 0.385,
       "step": 70765
     },
     {
-      "epoch": 2.49,
-      "learning_rate": 2.6391450692553243e-05,
-      "loss": 0.2695,
+      "epoch": 2.5505460049735107,
+      "grad_norm": 0.21216388046741486,
+      "learning_rate": 2.541027117369063e-05,
+      "loss": 0.4038,
       "step": 70770
     },
     {
-      "epoch": 2.49,
-      "learning_rate": 2.6388606378042812e-05,
-      "loss": 0.2888,
+      "epoch": 2.5507262046347354,
+      "grad_norm": 0.21207298338413239,
+      "learning_rate": 2.5407353430227554e-05,
+      "loss": 0.4195,
       "step": 70775
     },
     {
-      "epoch": 2.49,
-      "learning_rate": 2.638576204550225e-05,
-      "loss": 0.2716,
+      "epoch": 2.5509064042959597,
+      "grad_norm": 0.219013512134552,
+      "learning_rate": 2.540443568121436e-05,
+      "loss": 0.4065,
       "step": 70780
     },
     {
-      "epoch": 2.49,
-      "learning_rate": 2.6382917694968474e-05,
-      "loss": 0.2771,
+      "epoch": 2.5510866039571845,
+      "grad_norm": 0.23736967146396637,
+      "learning_rate": 2.5401517926690816e-05,
+      "loss": 0.3827,
       "step": 70785
     },
     {
-      "epoch": 2.49,
-      "learning_rate": 2.638007332647843e-05,
-      "loss": 0.2758,
+      "epoch": 2.551266803618409,
+      "grad_norm": 0.20479167997837067,
+      "learning_rate": 2.5398600166696668e-05,
+      "loss": 0.4108,
       "step": 70790
     },
     {
-      "epoch": 2.49,
-      "learning_rate": 2.6377228940069032e-05,
-      "loss": 0.2839,
+      "epoch": 2.551447003279634,
+      "grad_norm": 0.24934591352939606,
+      "learning_rate": 2.539568240127167e-05,
+      "loss": 0.3904,
       "step": 70795
     },
     {
-      "epoch": 2.49,
-      "learning_rate": 2.6374384535777235e-05,
-      "loss": 0.2812,
+      "epoch": 2.5516272029408587,
+      "grad_norm": 0.20699314773082733,
+      "learning_rate": 2.5392764630455574e-05,
+      "loss": 0.4098,
       "step": 70800
     },
     {
-      "epoch": 2.49,
-      "learning_rate": 2.6371540113639952e-05,
-      "loss": 0.2791,
+      "epoch": 2.5518074026020834,
+      "grad_norm": 0.22364868223667145,
+      "learning_rate": 2.5389846854288135e-05,
+      "loss": 0.4066,
       "step": 70805
     },
     {
-      "epoch": 2.49,
-      "learning_rate": 2.6368695673694133e-05,
-      "loss": 0.2697,
+      "epoch": 2.5519876022633077,
+      "grad_norm": 0.23530325293540955,
+      "learning_rate": 2.5386929072809107e-05,
+      "loss": 0.4167,
       "step": 70810
     },
     {
-      "epoch": 2.49,
-      "learning_rate": 2.6365851215976695e-05,
-      "loss": 0.3063,
+      "epoch": 2.5521678019245324,
+      "grad_norm": 0.2264833003282547,
+      "learning_rate": 2.5384011286058245e-05,
+      "loss": 0.3679,
       "step": 70815
     },
     {
-      "epoch": 2.49,
-      "learning_rate": 2.6363006740524583e-05,
-      "loss": 0.2718,
+      "epoch": 2.552348001585757,
+      "grad_norm": 0.20546197891235352,
+      "learning_rate": 2.5381093494075308e-05,
+      "loss": 0.3812,
       "step": 70820
     },
     {
-      "epoch": 2.49,
-      "learning_rate": 2.636016224737472e-05,
-      "loss": 0.2651,
+      "epoch": 2.5525282012469814,
+      "grad_norm": 0.1657361537218094,
+      "learning_rate": 2.5378175696900042e-05,
+      "loss": 0.3598,
       "step": 70825
     },
     {
-      "epoch": 2.49,
-      "learning_rate": 2.6357317736564058e-05,
-      "loss": 0.2578,
+      "epoch": 2.552708400908206,
+      "grad_norm": 0.24184440076351166,
+      "learning_rate": 2.5375257894572207e-05,
+      "loss": 0.3985,
       "step": 70830
     },
     {
-      "epoch": 2.49,
-      "learning_rate": 2.6354473208129516e-05,
-      "loss": 0.2959,
+      "epoch": 2.552888600569431,
+      "grad_norm": 0.22822557389736176,
+      "learning_rate": 2.537234008713156e-05,
+      "loss": 0.3811,
       "step": 70835
     },
     {
-      "epoch": 2.49,
-      "learning_rate": 2.635162866210803e-05,
-      "loss": 0.2696,
+      "epoch": 2.5530688002306556,
+      "grad_norm": 0.2141774296760559,
+      "learning_rate": 2.5369422274617844e-05,
+      "loss": 0.4142,
       "step": 70840
     },
     {
-      "epoch": 2.49,
-      "learning_rate": 2.634878409853654e-05,
-      "loss": 0.2828,
+      "epoch": 2.5532489998918804,
+      "grad_norm": 0.16231632232666016,
+      "learning_rate": 2.5366504457070826e-05,
+      "loss": 0.3693,
       "step": 70845
     },
     {
-      "epoch": 2.49,
-      "learning_rate": 2.6345939517451968e-05,
-      "loss": 0.2892,
+      "epoch": 2.553429199553105,
+      "grad_norm": 0.1983339488506317,
+      "learning_rate": 2.536358663453025e-05,
+      "loss": 0.4081,
       "step": 70850
     },
     {
-      "epoch": 2.49,
-      "learning_rate": 2.6343094918891264e-05,
-      "loss": 0.268,
+      "epoch": 2.5536093992143294,
+      "grad_norm": 0.21507583558559418,
+      "learning_rate": 2.5360668807035887e-05,
+      "loss": 0.3717,
       "step": 70855
     },
     {
-      "epoch": 2.49,
-      "learning_rate": 2.634025030289136e-05,
-      "loss": 0.28,
+      "epoch": 2.553789598875554,
+      "grad_norm": 0.2187241017818451,
+      "learning_rate": 2.535775097462747e-05,
+      "loss": 0.3672,
       "step": 70860
     },
     {
-      "epoch": 2.49,
-      "learning_rate": 2.6337405669489183e-05,
-      "loss": 0.301,
+      "epoch": 2.553969798536779,
+      "grad_norm": 0.20706336200237274,
+      "learning_rate": 2.5354833137344773e-05,
+      "loss": 0.407,
       "step": 70865
     },
     {
-      "epoch": 2.49,
-      "learning_rate": 2.633456101872168e-05,
-      "loss": 0.2413,
+      "epoch": 2.554149998198003,
+      "grad_norm": 0.21873357892036438,
+      "learning_rate": 2.5351915295227535e-05,
+      "loss": 0.4204,
       "step": 70870
     },
     {
-      "epoch": 2.49,
-      "learning_rate": 2.633171635062578e-05,
-      "loss": 0.2937,
+      "epoch": 2.554330197859228,
+      "grad_norm": 0.1820431351661682,
+      "learning_rate": 2.534899744831553e-05,
+      "loss": 0.3873,
       "step": 70875
     },
     {
-      "epoch": 2.49,
-      "learning_rate": 2.6328871665238425e-05,
-      "loss": 0.2646,
+      "epoch": 2.5545103975204526,
+      "grad_norm": 0.17599505186080933,
+      "learning_rate": 2.5346079596648497e-05,
+      "loss": 0.3794,
       "step": 70880
     },
     {
-      "epoch": 2.49,
-      "learning_rate": 2.632602696259654e-05,
-      "loss": 0.2688,
+      "epoch": 2.5546905971816773,
+      "grad_norm": 0.2262696623802185,
+      "learning_rate": 2.534316174026619e-05,
+      "loss": 0.4126,
       "step": 70885
     },
     {
-      "epoch": 2.49,
-      "learning_rate": 2.632318224273707e-05,
-      "loss": 0.2705,
+      "epoch": 2.554870796842902,
+      "grad_norm": 0.23906999826431274,
+      "learning_rate": 2.534024387920838e-05,
+      "loss": 0.3999,
       "step": 70890
     },
     {
-      "epoch": 2.49,
-      "learning_rate": 2.6320337505696956e-05,
-      "loss": 0.2516,
+      "epoch": 2.555050996504127,
+      "grad_norm": 0.16976365447044373,
+      "learning_rate": 2.5337326013514817e-05,
+      "loss": 0.3864,
       "step": 70895
     },
     {
-      "epoch": 2.49,
-      "learning_rate": 2.6317492751513128e-05,
-      "loss": 0.2815,
+      "epoch": 2.555231196165351,
+      "grad_norm": 0.2087818831205368,
+      "learning_rate": 2.5334408143225246e-05,
+      "loss": 0.4226,
       "step": 70900
     },
     {
-      "epoch": 2.49,
-      "learning_rate": 2.6314647980222524e-05,
-      "loss": 0.2841,
+      "epoch": 2.555411395826576,
+      "grad_norm": 0.21390338242053986,
+      "learning_rate": 2.533149026837942e-05,
+      "loss": 0.3674,
       "step": 70905
     },
     {
-      "epoch": 2.49,
-      "learning_rate": 2.631180319186208e-05,
-      "loss": 0.2748,
+      "epoch": 2.5555915954878006,
+      "grad_norm": 0.22753164172172546,
+      "learning_rate": 2.532857238901712e-05,
+      "loss": 0.4407,
       "step": 70910
     },
     {
-      "epoch": 2.49,
-      "learning_rate": 2.630895838646874e-05,
-      "loss": 0.2425,
+      "epoch": 2.555771795149025,
+      "grad_norm": 0.179393008351326,
+      "learning_rate": 2.5325654505178074e-05,
+      "loss": 0.4156,
       "step": 70915
     },
     {
-      "epoch": 2.5,
-      "learning_rate": 2.630611356407943e-05,
-      "loss": 0.2694,
+      "epoch": 2.5559519948102496,
+      "grad_norm": 0.22798627614974976,
+      "learning_rate": 2.53233201949103e-05,
+      "loss": 0.4018,
       "step": 70920
     },
     {
-      "epoch": 2.5,
-      "learning_rate": 2.6303268724731112e-05,
-      "loss": 0.2638,
+      "epoch": 2.5561321944714743,
+      "grad_norm": 0.203562393784523,
+      "learning_rate": 2.5320402303113327e-05,
+      "loss": 0.4037,
       "step": 70925
     },
     {
-      "epoch": 2.5,
-      "learning_rate": 2.6300423868460688e-05,
-      "loss": 0.2543,
+      "epoch": 2.556312394132699,
+      "grad_norm": 0.22589832544326782,
+      "learning_rate": 2.531748440695092e-05,
+      "loss": 0.3967,
       "step": 70930
     },
     {
-      "epoch": 2.5,
-      "learning_rate": 2.6297578995305133e-05,
-      "loss": 0.2726,
+      "epoch": 2.556492593793924,
+      "grad_norm": 0.23674431443214417,
+      "learning_rate": 2.5314566506462856e-05,
+      "loss": 0.3992,
       "step": 70935
     },
     {
-      "epoch": 2.5,
-      "learning_rate": 2.629473410530136e-05,
-      "loss": 0.2665,
+      "epoch": 2.5566727934551485,
+      "grad_norm": 0.20412121713161469,
+      "learning_rate": 2.5311648601688876e-05,
+      "loss": 0.4023,
       "step": 70940
     },
     {
-      "epoch": 2.5,
-      "learning_rate": 2.629188919848633e-05,
-      "loss": 0.2788,
+      "epoch": 2.556852993116373,
+      "grad_norm": 0.17974796891212463,
+      "learning_rate": 2.5308730692668754e-05,
+      "loss": 0.3787,
       "step": 70945
     },
     {
-      "epoch": 2.5,
-      "learning_rate": 2.6289044274896963e-05,
-      "loss": 0.2554,
+      "epoch": 2.5570331927775976,
+      "grad_norm": 0.16718092560768127,
+      "learning_rate": 2.5305812779442232e-05,
+      "loss": 0.407,
       "step": 70950
     },
     {
-      "epoch": 2.5,
-      "learning_rate": 2.62861993345702e-05,
-      "loss": 0.2959,
+      "epoch": 2.5572133924388223,
+      "grad_norm": 0.16526854038238525,
+      "learning_rate": 2.530289486204907e-05,
+      "loss": 0.3996,
       "step": 70955
     },
     {
-      "epoch": 2.5,
-      "learning_rate": 2.628335437754299e-05,
-      "loss": 0.2694,
+      "epoch": 2.5573935921000466,
+      "grad_norm": 0.2542569935321808,
+      "learning_rate": 2.529997694052903e-05,
+      "loss": 0.3735,
       "step": 70960
     },
     {
-      "epoch": 2.5,
-      "learning_rate": 2.6280509403852277e-05,
-      "loss": 0.2644,
+      "epoch": 2.5575737917612713,
+      "grad_norm": 0.21731750667095184,
+      "learning_rate": 2.5297059014921857e-05,
+      "loss": 0.381,
       "step": 70965
     },
     {
-      "epoch": 2.5,
-      "learning_rate": 2.627766441353498e-05,
-      "loss": 0.2655,
+      "epoch": 2.557753991422496,
+      "grad_norm": 0.18733660876750946,
+      "learning_rate": 2.5294141085267315e-05,
+      "loss": 0.3858,
       "step": 70970
     },
     {
-      "epoch": 2.5,
-      "learning_rate": 2.6274819406628066e-05,
-      "loss": 0.2675,
+      "epoch": 2.5579341910837208,
+      "grad_norm": 0.23402203619480133,
+      "learning_rate": 2.5291223151605147e-05,
+      "loss": 0.4118,
       "step": 70975
     },
     {
-      "epoch": 2.5,
-      "learning_rate": 2.6271974383168453e-05,
-      "loss": 0.2728,
+      "epoch": 2.5581143907449455,
+      "grad_norm": 0.19820962846279144,
+      "learning_rate": 2.5288305213975132e-05,
+      "loss": 0.3861,
       "step": 70980
     },
     {
-      "epoch": 2.5,
-      "learning_rate": 2.6269129343193093e-05,
-      "loss": 0.246,
+      "epoch": 2.5582945904061702,
+      "grad_norm": 0.2706112861633301,
+      "learning_rate": 2.5285387272417015e-05,
+      "loss": 0.4072,
       "step": 70985
     },
     {
-      "epoch": 2.5,
-      "learning_rate": 2.6266284286738924e-05,
-      "loss": 0.2662,
+      "epoch": 2.5584747900673945,
+      "grad_norm": 0.21151851117610931,
+      "learning_rate": 2.5282469326970543e-05,
+      "loss": 0.3774,
       "step": 70990
     },
     {
-      "epoch": 2.5,
-      "learning_rate": 2.6263439213842893e-05,
-      "loss": 0.2847,
+      "epoch": 2.5586549897286193,
+      "grad_norm": 0.1977061629295349,
+      "learning_rate": 2.5279551377675488e-05,
+      "loss": 0.4204,
       "step": 70995
     },
     {
-      "epoch": 2.5,
-      "learning_rate": 2.6260594124541932e-05,
-      "loss": 0.2974,
+      "epoch": 2.558835189389844,
+      "grad_norm": 0.19445201754570007,
+      "learning_rate": 2.5276633424571594e-05,
+      "loss": 0.3868,
       "step": 71000
     },
     {
-      "epoch": 2.5,
-      "eval_loss": 0.26469552516937256,
-      "eval_runtime": 10.5298,
-      "eval_samples_per_second": 9.497,
-      "eval_steps_per_second": 9.497,
+      "epoch": 2.558835189389844,
+      "eval_loss": 0.43283751606941223,
+      "eval_runtime": 3.5328,
+      "eval_samples_per_second": 28.306,
+      "eval_steps_per_second": 7.077,
       "step": 71000
     },
     {
-      "epoch": 2.5,
-      "learning_rate": 2.6257749018872983e-05,
-      "loss": 0.2703,
+      "epoch": 2.5590153890510687,
+      "grad_norm": 0.17734093964099884,
+      "learning_rate": 2.5273715467698633e-05,
+      "loss": 0.3773,
       "step": 71005
     },
     {
-      "epoch": 2.5,
-      "learning_rate": 2.6254903896872996e-05,
-      "loss": 0.2498,
+      "epoch": 2.559195588712293,
+      "grad_norm": 0.23937411606311798,
+      "learning_rate": 2.527079750709634e-05,
+      "loss": 0.4013,
       "step": 71010
     },
     {
-      "epoch": 2.5,
-      "learning_rate": 2.6252058758578907e-05,
-      "loss": 0.2616,
+      "epoch": 2.5593757883735178,
+      "grad_norm": 0.1706862598657608,
+      "learning_rate": 2.5267879542804484e-05,
+      "loss": 0.3689,
       "step": 71015
     },
     {
-      "epoch": 2.5,
-      "learning_rate": 2.6249213604027667e-05,
-      "loss": 0.2757,
+      "epoch": 2.5595559880347425,
+      "grad_norm": 0.20813287794589996,
+      "learning_rate": 2.526496157486283e-05,
+      "loss": 0.3894,
       "step": 71020
     },
     {
-      "epoch": 2.5,
-      "learning_rate": 2.6246368433256206e-05,
-      "loss": 0.2803,
+      "epoch": 2.5597361876959672,
+      "grad_norm": 0.24396944046020508,
+      "learning_rate": 2.526204360331112e-05,
+      "loss": 0.3836,
       "step": 71025
     },
     {
-      "epoch": 2.5,
-      "learning_rate": 2.6243523246301477e-05,
-      "loss": 0.2634,
+      "epoch": 2.559916387357192,
+      "grad_norm": 0.18609154224395752,
+      "learning_rate": 2.525912562818912e-05,
+      "loss": 0.3683,
       "step": 71030
     },
     {
-      "epoch": 2.5,
-      "learning_rate": 2.6240678043200416e-05,
-      "loss": 0.2533,
+      "epoch": 2.5600965870184162,
+      "grad_norm": 0.24493709206581116,
+      "learning_rate": 2.525620764953658e-05,
+      "loss": 0.3732,
       "step": 71035
     },
     {
-      "epoch": 2.5,
-      "learning_rate": 2.6237832823989967e-05,
-      "loss": 0.2823,
+      "epoch": 2.560276786679641,
+      "grad_norm": 0.19621089100837708,
+      "learning_rate": 2.525328966739326e-05,
+      "loss": 0.3917,
       "step": 71040
     },
     {
-      "epoch": 2.5,
-      "learning_rate": 2.623498758870707e-05,
-      "loss": 0.2766,
+      "epoch": 2.5604569863408657,
+      "grad_norm": 0.20630283653736115,
+      "learning_rate": 2.525037168179892e-05,
+      "loss": 0.4063,
       "step": 71045
     },
     {
-      "epoch": 2.5,
-      "learning_rate": 2.6232142337388684e-05,
-      "loss": 0.3033,
+      "epoch": 2.5606371860020904,
+      "grad_norm": 0.2241489142179489,
+      "learning_rate": 2.524745369279331e-05,
+      "loss": 0.4282,
       "step": 71050
     },
     {
-      "epoch": 2.5,
-      "learning_rate": 2.6229297070071734e-05,
-      "loss": 0.2837,
+      "epoch": 2.5608173856633147,
+      "grad_norm": 0.20686471462249756,
+      "learning_rate": 2.52445357004162e-05,
+      "loss": 0.4066,
       "step": 71055
     },
     {
-      "epoch": 2.5,
-      "learning_rate": 2.6226451786793177e-05,
-      "loss": 0.2547,
+      "epoch": 2.5609975853245395,
+      "grad_norm": 0.2341819703578949,
+      "learning_rate": 2.5241617704707328e-05,
+      "loss": 0.4269,
       "step": 71060
     },
     {
-      "epoch": 2.5,
-      "learning_rate": 2.6223606487589946e-05,
-      "loss": 0.2853,
+      "epoch": 2.561177784985764,
+      "grad_norm": 0.17863284051418304,
+      "learning_rate": 2.523869970570647e-05,
+      "loss": 0.4245,
       "step": 71065
     },
     {
-      "epoch": 2.5,
-      "learning_rate": 2.6220761172498993e-05,
-      "loss": 0.2503,
+      "epoch": 2.561357984646989,
+      "grad_norm": 0.31622689962387085,
+      "learning_rate": 2.523578170345337e-05,
+      "loss": 0.437,
       "step": 71070
     },
     {
-      "epoch": 2.5,
-      "learning_rate": 2.6217915841557255e-05,
-      "loss": 0.2546,
+      "epoch": 2.5615381843082137,
+      "grad_norm": 0.24306558072566986,
+      "learning_rate": 2.5232863697987796e-05,
+      "loss": 0.4115,
       "step": 71075
     },
     {
-      "epoch": 2.5,
-      "learning_rate": 2.6215070494801693e-05,
-      "loss": 0.2663,
+      "epoch": 2.5617183839694384,
+      "grad_norm": 0.23051370680332184,
+      "learning_rate": 2.5229945689349487e-05,
+      "loss": 0.4229,
       "step": 71080
     },
     {
-      "epoch": 2.5,
-      "learning_rate": 2.6212225132269234e-05,
-      "loss": 0.2875,
+      "epoch": 2.5618985836306627,
+      "grad_norm": 0.22686095535755157,
+      "learning_rate": 2.5227027677578224e-05,
+      "loss": 0.3949,
       "step": 71085
     },
     {
-      "epoch": 2.5,
-      "learning_rate": 2.6209379753996833e-05,
-      "loss": 0.2916,
+      "epoch": 2.5620787832918874,
+      "grad_norm": 0.24122454226016998,
+      "learning_rate": 2.5224109662713752e-05,
+      "loss": 0.382,
       "step": 71090
     },
     {
-      "epoch": 2.5,
-      "learning_rate": 2.6206534360021427e-05,
-      "loss": 0.286,
+      "epoch": 2.562258982953112,
+      "grad_norm": 0.19796620309352875,
+      "learning_rate": 2.5221191644795822e-05,
+      "loss": 0.3849,
       "step": 71095
     },
     {
-      "epoch": 2.5,
-      "learning_rate": 2.6203688950379978e-05,
-      "loss": 0.2548,
+      "epoch": 2.5624391826143365,
+      "grad_norm": 0.20853497087955475,
+      "learning_rate": 2.5218273623864202e-05,
+      "loss": 0.3628,
       "step": 71100
     },
     {
-      "epoch": 2.5,
-      "learning_rate": 2.6200843525109418e-05,
-      "loss": 0.2594,
+      "epoch": 2.562619382275561,
+      "grad_norm": 0.24205127358436584,
+      "learning_rate": 2.5215355599958647e-05,
+      "loss": 0.4302,
       "step": 71105
     },
     {
-      "epoch": 2.5,
-      "learning_rate": 2.61979980842467e-05,
-      "loss": 0.2408,
+      "epoch": 2.562799581936786,
+      "grad_norm": 0.18552717566490173,
+      "learning_rate": 2.521243757311892e-05,
+      "loss": 0.3786,
       "step": 71110
     },
     {
-      "epoch": 2.5,
-      "learning_rate": 2.619515262782876e-05,
-      "loss": 0.2577,
+      "epoch": 2.5629797815980107,
+      "grad_norm": 0.15225747227668762,
+      "learning_rate": 2.5209519543384763e-05,
+      "loss": 0.3853,
       "step": 71115
     },
     {
-      "epoch": 2.5,
-      "learning_rate": 2.6192307155892547e-05,
-      "loss": 0.2733,
+      "epoch": 2.5631599812592354,
+      "grad_norm": 0.2108543962240219,
+      "learning_rate": 2.5206601510795948e-05,
+      "loss": 0.4087,
       "step": 71120
     },
     {
-      "epoch": 2.5,
-      "learning_rate": 2.6189461668475018e-05,
-      "loss": 0.2338,
+      "epoch": 2.56334018092046,
+      "grad_norm": 0.26248225569725037,
+      "learning_rate": 2.520368347539222e-05,
+      "loss": 0.3905,
       "step": 71125
     },
     {
-      "epoch": 2.5,
-      "learning_rate": 2.618661616561311e-05,
-      "loss": 0.2518,
+      "epoch": 2.5635203805816844,
+      "grad_norm": 0.18176917731761932,
+      "learning_rate": 2.5200765437213347e-05,
+      "loss": 0.3922,
       "step": 71130
     },
     {
-      "epoch": 2.5,
-      "learning_rate": 2.618377064734378e-05,
-      "loss": 0.2524,
+      "epoch": 2.563700580242909,
+      "grad_norm": 0.1985873579978943,
+      "learning_rate": 2.519784739629909e-05,
+      "loss": 0.3832,
       "step": 71135
     },
     {
-      "epoch": 2.5,
-      "learning_rate": 2.618092511370396e-05,
-      "loss": 0.244,
+      "epoch": 2.563880779904134,
+      "grad_norm": 0.217198446393013,
+      "learning_rate": 2.519492935268919e-05,
+      "loss": 0.4287,
       "step": 71140
     },
     {
-      "epoch": 2.5,
-      "learning_rate": 2.6178079564730613e-05,
-      "loss": 0.2627,
+      "epoch": 2.564060979565358,
+      "grad_norm": 0.18571309745311737,
+      "learning_rate": 2.5192011306423424e-05,
+      "loss": 0.3855,
       "step": 71145
     },
     {
-      "epoch": 2.5,
-      "learning_rate": 2.617523400046067e-05,
-      "loss": 0.2476,
+      "epoch": 2.564241179226583,
+      "grad_norm": 0.22361566126346588,
+      "learning_rate": 2.518909325754154e-05,
+      "loss": 0.419,
       "step": 71150
     },
     {
-      "epoch": 2.5,
-      "learning_rate": 2.6172388420931093e-05,
-      "loss": 0.2441,
+      "epoch": 2.5644213788878076,
+      "grad_norm": 0.16923829913139343,
+      "learning_rate": 2.51861752060833e-05,
+      "loss": 0.3755,
       "step": 71155
     },
     {
-      "epoch": 2.5,
-      "learning_rate": 2.6169542826178822e-05,
-      "loss": 0.2498,
+      "epoch": 2.5646015785490324,
+      "grad_norm": 0.19118502736091614,
+      "learning_rate": 2.518325715208845e-05,
+      "loss": 0.3829,
       "step": 71160
     },
     {
-      "epoch": 2.5,
-      "learning_rate": 2.6166697216240816e-05,
-      "loss": 0.2706,
+      "epoch": 2.564781778210257,
+      "grad_norm": 0.2314773052930832,
+      "learning_rate": 2.5180339095596755e-05,
+      "loss": 0.3592,
       "step": 71165
     },
     {
-      "epoch": 2.5,
-      "learning_rate": 2.616385159115401e-05,
-      "loss": 0.275,
+      "epoch": 2.564961977871482,
+      "grad_norm": 0.2128879874944687,
+      "learning_rate": 2.517742103664799e-05,
+      "loss": 0.3965,
       "step": 71170
     },
     {
-      "epoch": 2.5,
-      "learning_rate": 2.6161005950955358e-05,
-      "loss": 0.2578,
+      "epoch": 2.565142177532706,
+      "grad_norm": 0.20926570892333984,
+      "learning_rate": 2.5174502975281887e-05,
+      "loss": 0.3829,
       "step": 71175
     },
     {
-      "epoch": 2.5,
-      "learning_rate": 2.6158160295681804e-05,
-      "loss": 0.2837,
+      "epoch": 2.565322377193931,
+      "grad_norm": 0.2084009349346161,
+      "learning_rate": 2.517158491153822e-05,
+      "loss": 0.3937,
       "step": 71180
     },
     {
-      "epoch": 2.5,
-      "learning_rate": 2.6155314625370315e-05,
-      "loss": 0.2616,
+      "epoch": 2.5655025768551556,
+      "grad_norm": 0.1534343957901001,
+      "learning_rate": 2.5168666845456733e-05,
+      "loss": 0.3473,
       "step": 71185
     },
     {
-      "epoch": 2.5,
-      "learning_rate": 2.615246894005782e-05,
-      "loss": 0.2713,
+      "epoch": 2.56568277651638,
+      "grad_norm": 0.16337619721889496,
+      "learning_rate": 2.5165748777077197e-05,
+      "loss": 0.3902,
       "step": 71190
     },
     {
-      "epoch": 2.5,
-      "learning_rate": 2.6149623239781272e-05,
-      "loss": 0.2844,
+      "epoch": 2.5658629761776046,
+      "grad_norm": 0.23396508395671844,
+      "learning_rate": 2.516283070643937e-05,
+      "loss": 0.4145,
       "step": 71195
     },
     {
-      "epoch": 2.51,
-      "learning_rate": 2.6146777524577627e-05,
-      "loss": 0.283,
+      "epoch": 2.5660431758388293,
+      "grad_norm": 0.22363629937171936,
+      "learning_rate": 2.5159912633582998e-05,
+      "loss": 0.3752,
       "step": 71200
     },
     {
-      "epoch": 2.51,
-      "learning_rate": 2.6143931794483827e-05,
-      "loss": 0.2695,
+      "epoch": 2.566223375500054,
+      "grad_norm": 0.22132565081119537,
+      "learning_rate": 2.5156994558547857e-05,
+      "loss": 0.3862,
       "step": 71205
     },
     {
-      "epoch": 2.51,
-      "learning_rate": 2.6141086049536833e-05,
-      "loss": 0.2553,
+      "epoch": 2.566403575161279,
+      "grad_norm": 0.16280272603034973,
+      "learning_rate": 2.5154076481373694e-05,
+      "loss": 0.3692,
       "step": 71210
     },
     {
-      "epoch": 2.51,
-      "learning_rate": 2.6138240289773592e-05,
-      "loss": 0.2613,
+      "epoch": 2.5665837748225035,
+      "grad_norm": 0.17420467734336853,
+      "learning_rate": 2.5151158402100268e-05,
+      "loss": 0.3694,
       "step": 71215
     },
     {
-      "epoch": 2.51,
-      "learning_rate": 2.6135394515231037e-05,
-      "loss": 0.2718,
+      "epoch": 2.566763974483728,
+      "grad_norm": 0.22650916874408722,
+      "learning_rate": 2.514824032076733e-05,
+      "loss": 0.4037,
       "step": 71220
     },
     {
-      "epoch": 2.51,
-      "learning_rate": 2.6132548725946147e-05,
-      "loss": 0.2702,
+      "epoch": 2.5669441741449526,
+      "grad_norm": 0.22971905767917633,
+      "learning_rate": 2.514532223741466e-05,
+      "loss": 0.4182,
       "step": 71225
     },
     {
-      "epoch": 2.51,
-      "learning_rate": 2.6129702921955844e-05,
-      "loss": 0.2659,
+      "epoch": 2.5671243738061773,
+      "grad_norm": 0.19210931658744812,
+      "learning_rate": 2.5142404152081993e-05,
+      "loss": 0.3954,
       "step": 71230
     },
     {
-      "epoch": 2.51,
-      "learning_rate": 2.612685710329711e-05,
-      "loss": 0.2463,
+      "epoch": 2.567304573467402,
+      "grad_norm": 0.22867514193058014,
+      "learning_rate": 2.5139486064809097e-05,
+      "loss": 0.3992,
       "step": 71235
     },
     {
-      "epoch": 2.51,
-      "learning_rate": 2.6124011270006866e-05,
-      "loss": 0.2611,
+      "epoch": 2.5674847731286263,
+      "grad_norm": 0.19859769940376282,
+      "learning_rate": 2.5136567975635733e-05,
+      "loss": 0.4205,
       "step": 71240
     },
     {
-      "epoch": 2.51,
-      "learning_rate": 2.612116542212208e-05,
-      "loss": 0.2812,
+      "epoch": 2.567664972789851,
+      "grad_norm": 0.20891262590885162,
+      "learning_rate": 2.513364988460165e-05,
+      "loss": 0.4269,
       "step": 71245
     },
     {
-      "epoch": 2.51,
-      "learning_rate": 2.61183195596797e-05,
-      "loss": 0.2586,
+      "epoch": 2.567845172451076,
+      "grad_norm": 0.20854581892490387,
+      "learning_rate": 2.5130731791746627e-05,
+      "loss": 0.3829,
       "step": 71250
     },
     {
-      "epoch": 2.51,
-      "learning_rate": 2.611547368271668e-05,
-      "loss": 0.2627,
+      "epoch": 2.5680253721123005,
+      "grad_norm": 0.23713940382003784,
+      "learning_rate": 2.5127813697110398e-05,
+      "loss": 0.3889,
       "step": 71255
     },
     {
-      "epoch": 2.51,
-      "learning_rate": 2.611262779126996e-05,
-      "loss": 0.2599,
+      "epoch": 2.5682055717735253,
+      "grad_norm": 0.17710144817829132,
+      "learning_rate": 2.512489560073274e-05,
+      "loss": 0.3795,
       "step": 71260
     },
     {
-      "epoch": 2.51,
-      "learning_rate": 2.610978188537651e-05,
-      "loss": 0.2799,
+      "epoch": 2.5683857714347496,
+      "grad_norm": 0.16682571172714233,
+      "learning_rate": 2.5121977502653395e-05,
+      "loss": 0.3858,
       "step": 71265
     },
     {
-      "epoch": 2.51,
-      "learning_rate": 2.6106935965073275e-05,
-      "loss": 0.261,
+      "epoch": 2.5685659710959743,
+      "grad_norm": 0.2379252016544342,
+      "learning_rate": 2.511905940291214e-05,
+      "loss": 0.4112,
       "step": 71270
     },
     {
-      "epoch": 2.51,
-      "learning_rate": 2.6104090030397206e-05,
-      "loss": 0.2442,
+      "epoch": 2.568746170757199,
+      "grad_norm": 0.19735756516456604,
+      "learning_rate": 2.5116141301548714e-05,
+      "loss": 0.3893,
       "step": 71275
     },
     {
-      "epoch": 2.51,
-      "learning_rate": 2.6101244081385257e-05,
-      "loss": 0.2656,
+      "epoch": 2.5689263704184238,
+      "grad_norm": 0.2579612135887146,
+      "learning_rate": 2.5113223198602885e-05,
+      "loss": 0.3624,
       "step": 71280
     },
     {
-      "epoch": 2.51,
-      "learning_rate": 2.609839811807437e-05,
-      "loss": 0.263,
+      "epoch": 2.569106570079648,
+      "grad_norm": 0.2095087617635727,
+      "learning_rate": 2.5110305094114416e-05,
+      "loss": 0.377,
       "step": 71285
     },
     {
-      "epoch": 2.51,
-      "learning_rate": 2.6095552140501512e-05,
-      "loss": 0.251,
+      "epoch": 2.5692867697408728,
+      "grad_norm": 0.19634920358657837,
+      "learning_rate": 2.510738698812306e-05,
+      "loss": 0.3583,
       "step": 71290
     },
     {
-      "epoch": 2.51,
-      "learning_rate": 2.609270614870363e-05,
-      "loss": 0.2839,
+      "epoch": 2.5694669694020975,
+      "grad_norm": 0.2694518268108368,
+      "learning_rate": 2.5104468880668587e-05,
+      "loss": 0.4229,
       "step": 71295
     },
     {
-      "epoch": 2.51,
-      "learning_rate": 2.6089860142717682e-05,
-      "loss": 0.2422,
+      "epoch": 2.5696471690633222,
+      "grad_norm": 0.1848263442516327,
+      "learning_rate": 2.5101550771790733e-05,
+      "loss": 0.4172,
       "step": 71300
     },
     {
-      "epoch": 2.51,
-      "learning_rate": 2.6087014122580617e-05,
-      "loss": 0.2632,
+      "epoch": 2.569827368724547,
+      "grad_norm": 0.16099122166633606,
+      "learning_rate": 2.509863266152927e-05,
+      "loss": 0.3887,
       "step": 71305
     },
     {
-      "epoch": 2.51,
-      "learning_rate": 2.6084168088329384e-05,
-      "loss": 0.293,
+      "epoch": 2.5700075683857713,
+      "grad_norm": 0.2326761782169342,
+      "learning_rate": 2.509571454992396e-05,
+      "loss": 0.4158,
       "step": 71310
     },
     {
-      "epoch": 2.51,
-      "learning_rate": 2.608132204000095e-05,
-      "loss": 0.2771,
+      "epoch": 2.570187768046996,
+      "grad_norm": 0.23651954531669617,
+      "learning_rate": 2.509279643701456e-05,
+      "loss": 0.3646,
       "step": 71315
     },
     {
-      "epoch": 2.51,
-      "learning_rate": 2.6078475977632268e-05,
-      "loss": 0.2842,
+      "epoch": 2.5703679677082207,
+      "grad_norm": 0.21195195615291595,
+      "learning_rate": 2.5089878322840826e-05,
+      "loss": 0.3927,
       "step": 71320
     },
     {
-      "epoch": 2.51,
-      "learning_rate": 2.6075629901260272e-05,
-      "loss": 0.2775,
+      "epoch": 2.5705481673694455,
+      "grad_norm": 0.23943771421909332,
+      "learning_rate": 2.5086960207442512e-05,
+      "loss": 0.3641,
       "step": 71325
     },
     {
-      "epoch": 2.51,
-      "learning_rate": 2.6072783810921946e-05,
-      "loss": 0.2772,
+      "epoch": 2.5707283670306698,
+      "grad_norm": 0.2217801809310913,
+      "learning_rate": 2.5084042090859382e-05,
+      "loss": 0.3776,
       "step": 71330
     },
     {
-      "epoch": 2.51,
-      "learning_rate": 2.6069937706654218e-05,
-      "loss": 0.2903,
+      "epoch": 2.5709085666918945,
+      "grad_norm": 0.17442776262760162,
+      "learning_rate": 2.5081123973131205e-05,
+      "loss": 0.3989,
       "step": 71335
     },
     {
-      "epoch": 2.51,
-      "learning_rate": 2.606709158849405e-05,
-      "loss": 0.274,
+      "epoch": 2.5710887663531192,
+      "grad_norm": 0.21039848029613495,
+      "learning_rate": 2.5078205854297715e-05,
+      "loss": 0.4116,
       "step": 71340
     },
     {
-      "epoch": 2.51,
-      "learning_rate": 2.606424545647841e-05,
-      "loss": 0.2756,
+      "epoch": 2.571268966014344,
+      "grad_norm": 0.1870678812265396,
+      "learning_rate": 2.5075287734398695e-05,
+      "loss": 0.3692,
       "step": 71345
     },
     {
-      "epoch": 2.51,
-      "learning_rate": 2.6061399310644246e-05,
-      "loss": 0.2881,
+      "epoch": 2.5714491656755687,
+      "grad_norm": 0.22971993684768677,
+      "learning_rate": 2.507236961347389e-05,
+      "loss": 0.4334,
       "step": 71350
     },
     {
-      "epoch": 2.51,
-      "learning_rate": 2.605855315102851e-05,
-      "loss": 0.2835,
+      "epoch": 2.5716293653367934,
+      "grad_norm": 0.1942625790834427,
+      "learning_rate": 2.5069451491563073e-05,
+      "loss": 0.3755,
       "step": 71355
     },
     {
-      "epoch": 2.51,
-      "learning_rate": 2.6055706977668147e-05,
-      "loss": 0.2657,
+      "epoch": 2.5718095649980177,
+      "grad_norm": 0.20081962645053864,
+      "learning_rate": 2.506653336870598e-05,
+      "loss": 0.381,
       "step": 71360
     },
     {
-      "epoch": 2.51,
-      "learning_rate": 2.6052860790600143e-05,
-      "loss": 0.2821,
+      "epoch": 2.5719897646592425,
+      "grad_norm": 0.23155467212200165,
+      "learning_rate": 2.506361524494239e-05,
+      "loss": 0.3836,
       "step": 71365
     },
     {
-      "epoch": 2.51,
-      "learning_rate": 2.6050014589861416e-05,
-      "loss": 0.2426,
+      "epoch": 2.572169964320467,
+      "grad_norm": 0.21292930841445923,
+      "learning_rate": 2.506069712031205e-05,
+      "loss": 0.3763,
       "step": 71370
     },
     {
-      "epoch": 2.51,
-      "learning_rate": 2.6047168375488956e-05,
-      "loss": 0.2741,
+      "epoch": 2.5723501639816915,
+      "grad_norm": 0.1886298656463623,
+      "learning_rate": 2.5057778994854724e-05,
+      "loss": 0.3779,
       "step": 71375
     },
     {
-      "epoch": 2.51,
-      "learning_rate": 2.60443221475197e-05,
-      "loss": 0.2739,
+      "epoch": 2.572530363642916,
+      "grad_norm": 0.211809903383255,
+      "learning_rate": 2.5054860868610173e-05,
+      "loss": 0.3735,
       "step": 71380
     },
     {
-      "epoch": 2.51,
-      "learning_rate": 2.6041475905990614e-05,
-      "loss": 0.2688,
+      "epoch": 2.572710563304141,
+      "grad_norm": 0.18282023072242737,
+      "learning_rate": 2.505194274161815e-05,
+      "loss": 0.3909,
       "step": 71385
     },
     {
-      "epoch": 2.51,
-      "learning_rate": 2.6038629650938644e-05,
-      "loss": 0.2555,
+      "epoch": 2.5728907629653657,
+      "grad_norm": 0.22584885358810425,
+      "learning_rate": 2.504902461391842e-05,
+      "loss": 0.4002,
       "step": 71390
     },
     {
-      "epoch": 2.51,
-      "learning_rate": 2.603578338240075e-05,
-      "loss": 0.2414,
+      "epoch": 2.5730709626265904,
+      "grad_norm": 0.18001751601696014,
+      "learning_rate": 2.504610648555074e-05,
+      "loss": 0.4059,
       "step": 71395
     },
     {
-      "epoch": 2.51,
-      "learning_rate": 2.6032937100413896e-05,
-      "loss": 0.2605,
+      "epoch": 2.573251162287815,
+      "grad_norm": 0.22520719468593597,
+      "learning_rate": 2.504318835655487e-05,
+      "loss": 0.4089,
       "step": 71400
     },
     {
-      "epoch": 2.51,
-      "learning_rate": 2.603009080501504e-05,
-      "loss": 0.2634,
+      "epoch": 2.5734313619490394,
+      "grad_norm": 0.17423519492149353,
+      "learning_rate": 2.5040270226970558e-05,
+      "loss": 0.3849,
       "step": 71405
     },
     {
-      "epoch": 2.51,
-      "learning_rate": 2.602724449624112e-05,
-      "loss": 0.2889,
+      "epoch": 2.573611561610264,
+      "grad_norm": 0.2317894846200943,
+      "learning_rate": 2.503735209683758e-05,
+      "loss": 0.3856,
       "step": 71410
     },
     {
-      "epoch": 2.51,
-      "learning_rate": 2.6024398174129122e-05,
-      "loss": 0.2832,
+      "epoch": 2.573791761271489,
+      "grad_norm": 0.2146165519952774,
+      "learning_rate": 2.5034433966195686e-05,
+      "loss": 0.3974,
       "step": 71415
     },
     {
-      "epoch": 2.51,
-      "learning_rate": 2.602155183871598e-05,
-      "loss": 0.2571,
+      "epoch": 2.573971960932713,
+      "grad_norm": 0.23812200129032135,
+      "learning_rate": 2.5031515835084636e-05,
+      "loss": 0.4132,
       "step": 71420
     },
     {
-      "epoch": 2.51,
-      "learning_rate": 2.601870549003866e-05,
-      "loss": 0.2695,
+      "epoch": 2.574152160593938,
+      "grad_norm": 0.229581817984581,
+      "learning_rate": 2.502859770354419e-05,
+      "loss": 0.3816,
       "step": 71425
     },
     {
-      "epoch": 2.51,
-      "learning_rate": 2.6015859128134124e-05,
-      "loss": 0.28,
+      "epoch": 2.5743323602551627,
+      "grad_norm": 0.2108045518398285,
+      "learning_rate": 2.5025679571614095e-05,
+      "loss": 0.3487,
       "step": 71430
     },
     {
-      "epoch": 2.51,
-      "learning_rate": 2.601301275303933e-05,
-      "loss": 0.2626,
+      "epoch": 2.5745125599163874,
+      "grad_norm": 0.21576263010501862,
+      "learning_rate": 2.502276143933413e-05,
+      "loss": 0.3742,
       "step": 71435
     },
     {
-      "epoch": 2.51,
-      "learning_rate": 2.6010166364791223e-05,
-      "loss": 0.2621,
+      "epoch": 2.574692759577612,
+      "grad_norm": 0.21896515786647797,
+      "learning_rate": 2.5019843306744045e-05,
+      "loss": 0.386,
       "step": 71440
     },
     {
-      "epoch": 2.51,
-      "learning_rate": 2.600731996342678e-05,
-      "loss": 0.2734,
+      "epoch": 2.574872959238837,
+      "grad_norm": 0.25227516889572144,
+      "learning_rate": 2.50169251738836e-05,
+      "loss": 0.3987,
       "step": 71445
     },
     {
-      "epoch": 2.51,
-      "learning_rate": 2.6004473548982945e-05,
-      "loss": 0.2753,
+      "epoch": 2.575053158900061,
+      "grad_norm": 0.24805094301700592,
+      "learning_rate": 2.5014007040792548e-05,
+      "loss": 0.3862,
       "step": 71450
     },
     {
-      "epoch": 2.51,
-      "learning_rate": 2.6001627121496687e-05,
-      "loss": 0.2534,
+      "epoch": 2.575233358561286,
+      "grad_norm": 0.1973733752965927,
+      "learning_rate": 2.5011088907510648e-05,
+      "loss": 0.3989,
       "step": 71455
     },
     {
-      "epoch": 2.51,
-      "learning_rate": 2.599878068100496e-05,
-      "loss": 0.289,
+      "epoch": 2.5754135582225106,
+      "grad_norm": 0.20170655846595764,
+      "learning_rate": 2.500817077407768e-05,
+      "loss": 0.3698,
       "step": 71460
     },
     {
-      "epoch": 2.51,
-      "learning_rate": 2.5995934227544728e-05,
-      "loss": 0.2631,
+      "epoch": 2.575593757883735,
+      "grad_norm": 0.22908920049667358,
+      "learning_rate": 2.5005252640533374e-05,
+      "loss": 0.3889,
       "step": 71465
     },
     {
-      "epoch": 2.51,
-      "learning_rate": 2.599308776115294e-05,
-      "loss": 0.2656,
+      "epoch": 2.5757739575449596,
+      "grad_norm": 0.17910641431808472,
+      "learning_rate": 2.500233450691751e-05,
+      "loss": 0.361,
       "step": 71470
     },
     {
-      "epoch": 2.51,
-      "learning_rate": 2.599024128186656e-05,
-      "loss": 0.263,
+      "epoch": 2.5759541572061844,
+      "grad_norm": 0.22365345060825348,
+      "learning_rate": 2.4999416373269836e-05,
+      "loss": 0.4144,
       "step": 71475
     },
     {
-      "epoch": 2.51,
-      "learning_rate": 2.5987394789722554e-05,
-      "loss": 0.2728,
+      "epoch": 2.576134356867409,
+      "grad_norm": 0.20202086865901947,
+      "learning_rate": 2.4996498239630105e-05,
+      "loss": 0.3748,
       "step": 71480
     },
     {
-      "epoch": 2.52,
-      "learning_rate": 2.5984548284757877e-05,
-      "loss": 0.2724,
+      "epoch": 2.576314556528634,
+      "grad_norm": 0.17802977561950684,
+      "learning_rate": 2.4993580106038096e-05,
+      "loss": 0.4047,
       "step": 71485
     },
     {
-      "epoch": 2.52,
-      "learning_rate": 2.598170176700949e-05,
-      "loss": 0.2847,
+      "epoch": 2.5764947561898586,
+      "grad_norm": 0.22806064784526825,
+      "learning_rate": 2.499066197253355e-05,
+      "loss": 0.388,
       "step": 71490
     },
     {
-      "epoch": 2.52,
-      "learning_rate": 2.597885523651435e-05,
-      "loss": 0.2776,
+      "epoch": 2.576674955851083,
+      "grad_norm": 0.1772589087486267,
+      "learning_rate": 2.4987743839156234e-05,
+      "loss": 0.3769,
       "step": 71495
     },
     {
-      "epoch": 2.52,
-      "learning_rate": 2.597600869330943e-05,
-      "loss": 0.2942,
+      "epoch": 2.5768551555123076,
+      "grad_norm": 0.2450006902217865,
+      "learning_rate": 2.4984825705945912e-05,
+      "loss": 0.3806,
       "step": 71500
     },
     {
-      "epoch": 2.52,
-      "eval_loss": 0.2644185721874237,
-      "eval_runtime": 10.558,
-      "eval_samples_per_second": 9.471,
-      "eval_steps_per_second": 9.471,
+      "epoch": 2.5768551555123076,
+      "eval_loss": 0.4331854581832886,
+      "eval_runtime": 3.5274,
+      "eval_samples_per_second": 28.349,
+      "eval_steps_per_second": 7.087,
       "step": 71500
     },
     {
-      "epoch": 2.52,
-      "learning_rate": 2.5973162137431668e-05,
-      "loss": 0.2832,
+      "epoch": 2.5770353551735323,
+      "grad_norm": 0.1859433799982071,
+      "learning_rate": 2.4981907572942326e-05,
+      "loss": 0.3643,
       "step": 71505
     },
     {
-      "epoch": 2.52,
-      "learning_rate": 2.597031556891804e-05,
-      "loss": 0.2688,
+      "epoch": 2.577215554834757,
+      "grad_norm": 0.21410761773586273,
+      "learning_rate": 2.4978989440185254e-05,
+      "loss": 0.3997,
       "step": 71510
     },
     {
-      "epoch": 2.52,
-      "learning_rate": 2.5967468987805504e-05,
-      "loss": 0.2701,
+      "epoch": 2.5773957544959814,
+      "grad_norm": 0.19530677795410156,
+      "learning_rate": 2.4976071307714446e-05,
+      "loss": 0.403,
       "step": 71515
     },
     {
-      "epoch": 2.52,
-      "learning_rate": 2.5964622394131023e-05,
-      "loss": 0.2681,
+      "epoch": 2.577575954157206,
+      "grad_norm": 0.19589672982692719,
+      "learning_rate": 2.4973153175569657e-05,
+      "loss": 0.3772,
       "step": 71520
     },
     {
-      "epoch": 2.52,
-      "learning_rate": 2.5961775787931553e-05,
-      "loss": 0.2971,
+      "epoch": 2.577756153818431,
+      "grad_norm": 0.16859115660190582,
+      "learning_rate": 2.4970235043790657e-05,
+      "loss": 0.3551,
       "step": 71525
     },
     {
-      "epoch": 2.52,
-      "learning_rate": 2.5958929169244063e-05,
-      "loss": 0.312,
+      "epoch": 2.5779363534796556,
+      "grad_norm": 0.2540879547595978,
+      "learning_rate": 2.4967316912417204e-05,
+      "loss": 0.3946,
       "step": 71530
     },
     {
-      "epoch": 2.52,
-      "learning_rate": 2.5956082538105504e-05,
-      "loss": 0.2934,
+      "epoch": 2.5781165531408803,
+      "grad_norm": 0.1967131346464157,
+      "learning_rate": 2.4964398781489035e-05,
+      "loss": 0.4273,
       "step": 71535
     },
     {
-      "epoch": 2.52,
-      "learning_rate": 2.5953235894552847e-05,
-      "loss": 0.271,
+      "epoch": 2.5782967528021046,
+      "grad_norm": 0.2016594558954239,
+      "learning_rate": 2.496148065104594e-05,
+      "loss": 0.4031,
       "step": 71540
     },
     {
-      "epoch": 2.52,
-      "learning_rate": 2.5950389238623047e-05,
-      "loss": 0.2588,
+      "epoch": 2.5784769524633293,
+      "grad_norm": 0.22788941860198975,
+      "learning_rate": 2.495856252112765e-05,
+      "loss": 0.3956,
       "step": 71545
     },
     {
-      "epoch": 2.52,
-      "learning_rate": 2.5947542570353077e-05,
-      "loss": 0.2516,
+      "epoch": 2.578657152124554,
+      "grad_norm": 0.18338678777217865,
+      "learning_rate": 2.4955644391773954e-05,
+      "loss": 0.3795,
       "step": 71550
     },
     {
-      "epoch": 2.52,
-      "learning_rate": 2.594469588977988e-05,
-      "loss": 0.2648,
+      "epoch": 2.5788373517857788,
+      "grad_norm": 0.16157367825508118,
+      "learning_rate": 2.4952726263024588e-05,
+      "loss": 0.3509,
       "step": 71555
     },
     {
-      "epoch": 2.52,
-      "learning_rate": 2.594184919694044e-05,
-      "loss": 0.2894,
+      "epoch": 2.579017551447003,
+      "grad_norm": 0.1809634417295456,
+      "learning_rate": 2.4949808134919312e-05,
+      "loss": 0.3658,
       "step": 71560
     },
     {
-      "epoch": 2.52,
-      "learning_rate": 2.59390024918717e-05,
-      "loss": 0.2683,
+      "epoch": 2.579197751108228,
+      "grad_norm": 0.2137366533279419,
+      "learning_rate": 2.4946890007497898e-05,
+      "loss": 0.3747,
       "step": 71565
     },
     {
-      "epoch": 2.52,
-      "learning_rate": 2.593615577461064e-05,
-      "loss": 0.2838,
+      "epoch": 2.5793779507694525,
+      "grad_norm": 0.1947045475244522,
+      "learning_rate": 2.4943971880800093e-05,
+      "loss": 0.4341,
       "step": 71570
     },
     {
-      "epoch": 2.52,
-      "learning_rate": 2.5933309045194204e-05,
-      "loss": 0.2564,
+      "epoch": 2.5795581504306773,
+      "grad_norm": 0.2671882212162018,
+      "learning_rate": 2.4941053754865658e-05,
+      "loss": 0.4122,
       "step": 71575
     },
     {
-      "epoch": 2.52,
-      "learning_rate": 2.5930462303659375e-05,
-      "loss": 0.2731,
+      "epoch": 2.579738350091902,
+      "grad_norm": 0.17521758377552032,
+      "learning_rate": 2.4938135629734356e-05,
+      "loss": 0.3967,
       "step": 71580
     },
     {
-      "epoch": 2.52,
-      "learning_rate": 2.59276155500431e-05,
-      "loss": 0.2804,
+      "epoch": 2.5799185497531267,
+      "grad_norm": 0.1912911832332611,
+      "learning_rate": 2.4935217505445947e-05,
+      "loss": 0.384,
       "step": 71585
     },
     {
-      "epoch": 2.52,
-      "learning_rate": 2.592476878438234e-05,
-      "loss": 0.267,
+      "epoch": 2.580098749414351,
+      "grad_norm": 0.19323128461837769,
+      "learning_rate": 2.4932299382040183e-05,
+      "loss": 0.4119,
       "step": 71590
     },
     {
-      "epoch": 2.52,
-      "learning_rate": 2.5921922006714085e-05,
-      "loss": 0.2696,
+      "epoch": 2.5802789490755758,
+      "grad_norm": 0.22955751419067383,
+      "learning_rate": 2.4929381259556835e-05,
+      "loss": 0.4337,
       "step": 71595
     },
     {
-      "epoch": 2.52,
-      "learning_rate": 2.5919075217075268e-05,
-      "loss": 0.2778,
+      "epoch": 2.5804591487368005,
+      "grad_norm": 0.21088866889476776,
+      "learning_rate": 2.492646313803564e-05,
+      "loss": 0.3519,
       "step": 71600
     },
     {
-      "epoch": 2.52,
-      "learning_rate": 2.5916228415502875e-05,
-      "loss": 0.2754,
+      "epoch": 2.580639348398025,
+      "grad_norm": 0.23416496813297272,
+      "learning_rate": 2.492354501751638e-05,
+      "loss": 0.4205,
       "step": 71605
     },
     {
-      "epoch": 2.52,
-      "learning_rate": 2.5913381602033847e-05,
-      "loss": 0.2525,
+      "epoch": 2.5808195480592495,
+      "grad_norm": 0.22170627117156982,
+      "learning_rate": 2.4920626898038806e-05,
+      "loss": 0.4031,
       "step": 71610
     },
     {
-      "epoch": 2.52,
-      "learning_rate": 2.591053477670517e-05,
-      "loss": 0.2565,
+      "epoch": 2.5809997477204742,
+      "grad_norm": 0.18647152185440063,
+      "learning_rate": 2.491770877964267e-05,
+      "loss": 0.3669,
       "step": 71615
     },
     {
-      "epoch": 2.52,
-      "learning_rate": 2.5907687939553794e-05,
-      "loss": 0.2581,
+      "epoch": 2.581179947381699,
+      "grad_norm": 0.24669143557548523,
+      "learning_rate": 2.4914790662367737e-05,
+      "loss": 0.4038,
       "step": 71620
     },
     {
-      "epoch": 2.52,
-      "learning_rate": 2.5904841090616695e-05,
-      "loss": 0.2581,
+      "epoch": 2.5813601470429237,
+      "grad_norm": 0.2305600941181183,
+      "learning_rate": 2.491187254625376e-05,
+      "loss": 0.3989,
       "step": 71625
     },
     {
-      "epoch": 2.52,
-      "learning_rate": 2.5901994229930824e-05,
-      "loss": 0.2522,
+      "epoch": 2.5815403467041484,
+      "grad_norm": 0.17700450122356415,
+      "learning_rate": 2.4908954431340513e-05,
+      "loss": 0.3687,
       "step": 71630
     },
     {
-      "epoch": 2.52,
-      "learning_rate": 2.589914735753316e-05,
-      "loss": 0.2773,
+      "epoch": 2.5817205463653727,
+      "grad_norm": 0.20750737190246582,
+      "learning_rate": 2.4906036317667744e-05,
+      "loss": 0.3824,
       "step": 71635
     },
     {
-      "epoch": 2.52,
-      "learning_rate": 2.5896300473460655e-05,
-      "loss": 0.2666,
+      "epoch": 2.5819007460265975,
+      "grad_norm": 0.2014245092868805,
+      "learning_rate": 2.49031182052752e-05,
+      "loss": 0.3933,
       "step": 71640
     },
     {
-      "epoch": 2.52,
-      "learning_rate": 2.5893453577750277e-05,
-      "loss": 0.2703,
+      "epoch": 2.582080945687822,
+      "grad_norm": 0.21627309918403625,
+      "learning_rate": 2.4900200094202663e-05,
+      "loss": 0.4101,
       "step": 71645
     },
     {
-      "epoch": 2.52,
-      "learning_rate": 2.5890606670438994e-05,
-      "loss": 0.2686,
+      "epoch": 2.5822611453490465,
+      "grad_norm": 0.24448803067207336,
+      "learning_rate": 2.4897281984489868e-05,
+      "loss": 0.3986,
       "step": 71650
     },
     {
-      "epoch": 2.52,
-      "learning_rate": 2.5887759751563777e-05,
-      "loss": 0.2582,
+      "epoch": 2.5824413450102712,
+      "grad_norm": 0.20513613522052765,
+      "learning_rate": 2.4894363876176602e-05,
+      "loss": 0.3929,
       "step": 71655
     },
     {
-      "epoch": 2.52,
-      "learning_rate": 2.588491282116158e-05,
-      "loss": 0.2581,
+      "epoch": 2.582621544671496,
+      "grad_norm": 0.16580136120319366,
+      "learning_rate": 2.48914457693026e-05,
+      "loss": 0.3893,
       "step": 71660
     },
     {
-      "epoch": 2.52,
-      "learning_rate": 2.5882065879269373e-05,
-      "loss": 0.259,
+      "epoch": 2.5828017443327207,
+      "grad_norm": 0.17390620708465576,
+      "learning_rate": 2.4888527663907627e-05,
+      "loss": 0.3854,
       "step": 71665
     },
     {
-      "epoch": 2.52,
-      "learning_rate": 2.5879218925924116e-05,
-      "loss": 0.2782,
+      "epoch": 2.5829819439939454,
+      "grad_norm": 0.21692463755607605,
+      "learning_rate": 2.4885609560031445e-05,
+      "loss": 0.3975,
       "step": 71670
     },
     {
-      "epoch": 2.52,
-      "learning_rate": 2.587637196116279e-05,
-      "loss": 0.2828,
+      "epoch": 2.58316214365517,
+      "grad_norm": 0.20750007033348083,
+      "learning_rate": 2.4882691457713813e-05,
+      "loss": 0.4188,
       "step": 71675
     },
     {
-      "epoch": 2.52,
-      "learning_rate": 2.5873524985022347e-05,
-      "loss": 0.2558,
+      "epoch": 2.5833423433163945,
+      "grad_norm": 0.1784081906080246,
+      "learning_rate": 2.4879773356994478e-05,
+      "loss": 0.3624,
       "step": 71680
     },
     {
-      "epoch": 2.52,
-      "learning_rate": 2.587067799753976e-05,
-      "loss": 0.2871,
+      "epoch": 2.583522542977619,
+      "grad_norm": 0.2542138993740082,
+      "learning_rate": 2.4876855257913217e-05,
+      "loss": 0.3848,
       "step": 71685
     },
     {
-      "epoch": 2.52,
-      "learning_rate": 2.586783099875199e-05,
-      "loss": 0.2673,
+      "epoch": 2.583702742638844,
+      "grad_norm": 0.19215676188468933,
+      "learning_rate": 2.4873937160509772e-05,
+      "loss": 0.3796,
       "step": 71690
     },
     {
-      "epoch": 2.52,
-      "learning_rate": 2.5864983988696007e-05,
-      "loss": 0.2785,
+      "epoch": 2.583882942300068,
+      "grad_norm": 0.23548319935798645,
+      "learning_rate": 2.4871019064823918e-05,
+      "loss": 0.4226,
       "step": 71695
     },
     {
-      "epoch": 2.52,
-      "learning_rate": 2.5862136967408773e-05,
-      "loss": 0.2656,
+      "epoch": 2.584063141961293,
+      "grad_norm": 0.23626303672790527,
+      "learning_rate": 2.48681009708954e-05,
+      "loss": 0.4119,
       "step": 71700
     },
     {
-      "epoch": 2.52,
-      "learning_rate": 2.585928993492727e-05,
-      "loss": 0.2634,
+      "epoch": 2.5842433416225177,
+      "grad_norm": 0.20605653524398804,
+      "learning_rate": 2.4865182878763975e-05,
+      "loss": 0.3605,
       "step": 71705
     },
     {
-      "epoch": 2.52,
-      "learning_rate": 2.5856442891288447e-05,
-      "loss": 0.2504,
+      "epoch": 2.5844235412837424,
+      "grad_norm": 0.23323751986026764,
+      "learning_rate": 2.4862264788469414e-05,
+      "loss": 0.4294,
       "step": 71710
     },
     {
-      "epoch": 2.52,
-      "learning_rate": 2.5853595836529277e-05,
-      "loss": 0.2628,
+      "epoch": 2.584603740944967,
+      "grad_norm": 0.19780725240707397,
+      "learning_rate": 2.4859346700051474e-05,
+      "loss": 0.3779,
       "step": 71715
     },
     {
-      "epoch": 2.52,
-      "learning_rate": 2.5850748770686728e-05,
-      "loss": 0.255,
+      "epoch": 2.584783940606192,
+      "grad_norm": 0.2039288431406021,
+      "learning_rate": 2.4856428613549892e-05,
+      "loss": 0.3828,
       "step": 71720
     },
     {
-      "epoch": 2.52,
-      "learning_rate": 2.5847901693797765e-05,
-      "loss": 0.2527,
+      "epoch": 2.584964140267416,
+      "grad_norm": 0.1562710702419281,
+      "learning_rate": 2.485351052900446e-05,
+      "loss": 0.3812,
       "step": 71725
     },
     {
-      "epoch": 2.52,
-      "learning_rate": 2.584505460589936e-05,
-      "loss": 0.2515,
+      "epoch": 2.585144339928641,
+      "grad_norm": 0.18016739189624786,
+      "learning_rate": 2.48505924464549e-05,
+      "loss": 0.3886,
       "step": 71730
     },
     {
-      "epoch": 2.52,
-      "learning_rate": 2.5842207507028478e-05,
-      "loss": 0.297,
+      "epoch": 2.5853245395898656,
+      "grad_norm": 0.1871825009584427,
+      "learning_rate": 2.4847674365941e-05,
+      "loss": 0.3554,
       "step": 71735
     },
     {
-      "epoch": 2.52,
-      "learning_rate": 2.5839360397222083e-05,
-      "loss": 0.2261,
+      "epoch": 2.5855047392510904,
+      "grad_norm": 0.2108829915523529,
+      "learning_rate": 2.4844756287502514e-05,
+      "loss": 0.37,
       "step": 71740
     },
     {
-      "epoch": 2.52,
-      "learning_rate": 2.583651327651715e-05,
-      "loss": 0.2677,
+      "epoch": 2.5856849389123147,
+      "grad_norm": 0.2470017671585083,
+      "learning_rate": 2.4841838211179175e-05,
+      "loss": 0.3934,
       "step": 71745
     },
     {
-      "epoch": 2.52,
-      "learning_rate": 2.583366614495064e-05,
-      "loss": 0.2555,
+      "epoch": 2.5858651385735394,
+      "grad_norm": 0.22116664052009583,
+      "learning_rate": 2.4838920137010776e-05,
+      "loss": 0.4159,
       "step": 71750
     },
     {
-      "epoch": 2.52,
-      "learning_rate": 2.5830819002559526e-05,
-      "loss": 0.2722,
+      "epoch": 2.586045338234764,
+      "grad_norm": 0.1974770575761795,
+      "learning_rate": 2.4836002065037056e-05,
+      "loss": 0.4125,
       "step": 71755
     },
     {
-      "epoch": 2.52,
-      "learning_rate": 2.582797184938078e-05,
-      "loss": 0.256,
+      "epoch": 2.586225537895989,
+      "grad_norm": 0.2147388607263565,
+      "learning_rate": 2.4833083995297772e-05,
+      "loss": 0.4346,
       "step": 71760
     },
     {
-      "epoch": 2.52,
-      "learning_rate": 2.5825124685451363e-05,
-      "loss": 0.2765,
+      "epoch": 2.5864057375572136,
+      "grad_norm": 0.21751351654529572,
+      "learning_rate": 2.483016592783269e-05,
+      "loss": 0.354,
       "step": 71765
     },
     {
-      "epoch": 2.53,
-      "learning_rate": 2.5822277510808246e-05,
-      "loss": 0.2664,
+      "epoch": 2.586585937218438,
+      "grad_norm": 0.1897956132888794,
+      "learning_rate": 2.4827247862681556e-05,
+      "loss": 0.3771,
       "step": 71770
     },
     {
-      "epoch": 2.53,
-      "learning_rate": 2.5819430325488393e-05,
-      "loss": 0.2628,
+      "epoch": 2.5867661368796626,
+      "grad_norm": 0.2261020839214325,
+      "learning_rate": 2.4824329799884144e-05,
+      "loss": 0.3636,
       "step": 71775
     },
     {
-      "epoch": 2.53,
-      "learning_rate": 2.5816583129528776e-05,
-      "loss": 0.2742,
+      "epoch": 2.5869463365408873,
+      "grad_norm": 0.20903125405311584,
+      "learning_rate": 2.4821411739480206e-05,
+      "loss": 0.4079,
       "step": 71780
     },
     {
-      "epoch": 2.53,
-      "learning_rate": 2.581373592296637e-05,
-      "loss": 0.2894,
+      "epoch": 2.587126536202112,
+      "grad_norm": 0.21350517868995667,
+      "learning_rate": 2.481849368150949e-05,
+      "loss": 0.3354,
       "step": 71785
     },
     {
-      "epoch": 2.53,
-      "learning_rate": 2.5810888705838136e-05,
-      "loss": 0.2772,
+      "epoch": 2.5873067358633364,
+      "grad_norm": 0.18678854405879974,
+      "learning_rate": 2.481557562601177e-05,
+      "loss": 0.4028,
       "step": 71790
     },
     {
-      "epoch": 2.53,
-      "learning_rate": 2.580804147818105e-05,
-      "loss": 0.2661,
+      "epoch": 2.587486935524561,
+      "grad_norm": 0.20988419651985168,
+      "learning_rate": 2.4812657573026797e-05,
+      "loss": 0.3798,
       "step": 71795
     },
     {
-      "epoch": 2.53,
-      "learning_rate": 2.5805194240032083e-05,
-      "loss": 0.2905,
+      "epoch": 2.587667135185786,
+      "grad_norm": 0.21831747889518738,
+      "learning_rate": 2.4809739522594318e-05,
+      "loss": 0.3999,
       "step": 71800
     },
     {
-      "epoch": 2.53,
-      "learning_rate": 2.580234699142819e-05,
-      "loss": 0.2824,
+      "epoch": 2.5878473348470106,
+      "grad_norm": 0.18378116190433502,
+      "learning_rate": 2.4806821474754112e-05,
+      "loss": 0.41,
       "step": 71805
     },
     {
-      "epoch": 2.53,
-      "learning_rate": 2.579949973240635e-05,
-      "loss": 0.2648,
+      "epoch": 2.5880275345082353,
+      "grad_norm": 0.22383366525173187,
+      "learning_rate": 2.4803903429545918e-05,
+      "loss": 0.3942,
       "step": 71810
     },
     {
-      "epoch": 2.53,
-      "learning_rate": 2.5796652463003535e-05,
-      "loss": 0.2654,
+      "epoch": 2.5882077341694596,
+      "grad_norm": 0.15554063022136688,
+      "learning_rate": 2.480098538700951e-05,
+      "loss": 0.3572,
       "step": 71815
     },
     {
-      "epoch": 2.53,
-      "learning_rate": 2.579380518325672e-05,
-      "loss": 0.2823,
+      "epoch": 2.5883879338306843,
+      "grad_norm": 0.20337338745594025,
+      "learning_rate": 2.4798067347184638e-05,
+      "loss": 0.4069,
       "step": 71820
     },
     {
-      "epoch": 2.53,
-      "learning_rate": 2.5790957893202856e-05,
-      "loss": 0.2606,
+      "epoch": 2.588568133491909,
+      "grad_norm": 0.25683242082595825,
+      "learning_rate": 2.4795149310111047e-05,
+      "loss": 0.4472,
       "step": 71825
     },
     {
-      "epoch": 2.53,
-      "learning_rate": 2.5788110592878928e-05,
-      "loss": 0.2698,
+      "epoch": 2.588748333153134,
+      "grad_norm": 0.16255348920822144,
+      "learning_rate": 2.479223127582852e-05,
+      "loss": 0.3799,
       "step": 71830
     },
     {
-      "epoch": 2.53,
-      "learning_rate": 2.578526328232191e-05,
-      "loss": 0.273,
+      "epoch": 2.588928532814358,
+      "grad_norm": 0.23991474509239197,
+      "learning_rate": 2.478931324437679e-05,
+      "loss": 0.3806,
       "step": 71835
     },
     {
-      "epoch": 2.53,
-      "learning_rate": 2.5782415961568756e-05,
-      "loss": 0.2733,
+      "epoch": 2.589108732475583,
+      "grad_norm": 0.17780093848705292,
+      "learning_rate": 2.478639521579564e-05,
+      "loss": 0.371,
       "step": 71840
     },
     {
-      "epoch": 2.53,
-      "learning_rate": 2.5779568630656452e-05,
-      "loss": 0.2512,
+      "epoch": 2.5892889321368076,
+      "grad_norm": 0.21101722121238708,
+      "learning_rate": 2.478347719012481e-05,
+      "loss": 0.3911,
       "step": 71845
     },
     {
-      "epoch": 2.53,
-      "learning_rate": 2.5776721289621965e-05,
-      "loss": 0.264,
+      "epoch": 2.5894691317980323,
+      "grad_norm": 0.1497720628976822,
+      "learning_rate": 2.4780559167404054e-05,
+      "loss": 0.3663,
       "step": 71850
     },
     {
-      "epoch": 2.53,
-      "learning_rate": 2.5773873938502263e-05,
-      "loss": 0.2789,
+      "epoch": 2.589649331459257,
+      "grad_norm": 0.14714904129505157,
+      "learning_rate": 2.4777641147673144e-05,
+      "loss": 0.3627,
       "step": 71855
     },
     {
-      "epoch": 2.53,
-      "learning_rate": 2.577102657733432e-05,
-      "loss": 0.2529,
+      "epoch": 2.5898295311204818,
+      "grad_norm": 0.1989908665418625,
+      "learning_rate": 2.477472313097183e-05,
+      "loss": 0.3685,
       "step": 71860
     },
     {
-      "epoch": 2.53,
-      "learning_rate": 2.5768179206155098e-05,
-      "loss": 0.2685,
+      "epoch": 2.590009730781706,
+      "grad_norm": 0.2075403332710266,
+      "learning_rate": 2.4771805117339863e-05,
+      "loss": 0.3824,
       "step": 71865
     },
     {
-      "epoch": 2.53,
-      "learning_rate": 2.5765331825001577e-05,
-      "loss": 0.2743,
+      "epoch": 2.5901899304429308,
+      "grad_norm": 0.17004312574863434,
+      "learning_rate": 2.476888710681701e-05,
+      "loss": 0.3466,
       "step": 71870
     },
     {
-      "epoch": 2.53,
-      "learning_rate": 2.5762484433910732e-05,
-      "loss": 0.2568,
+      "epoch": 2.5903701301041555,
+      "grad_norm": 0.20493172109127045,
+      "learning_rate": 2.4765969099443025e-05,
+      "loss": 0.4133,
       "step": 71875
     },
     {
-      "epoch": 2.53,
-      "learning_rate": 2.5759637032919527e-05,
-      "loss": 0.2392,
+      "epoch": 2.59055032976538,
+      "grad_norm": 0.2063346803188324,
+      "learning_rate": 2.476305109525767e-05,
+      "loss": 0.4252,
       "step": 71880
     },
     {
-      "epoch": 2.53,
-      "learning_rate": 2.5756789622064937e-05,
-      "loss": 0.2603,
+      "epoch": 2.5907305294266045,
+      "grad_norm": 0.19341862201690674,
+      "learning_rate": 2.4760133094300697e-05,
+      "loss": 0.3719,
       "step": 71885
     },
     {
-      "epoch": 2.53,
-      "learning_rate": 2.5753942201383925e-05,
-      "loss": 0.2906,
+      "epoch": 2.5909107290878293,
+      "grad_norm": 0.17939716577529907,
+      "learning_rate": 2.475721509661186e-05,
+      "loss": 0.4111,
       "step": 71890
     },
     {
-      "epoch": 2.53,
-      "learning_rate": 2.5751094770913476e-05,
-      "loss": 0.2739,
+      "epoch": 2.591090928749054,
+      "grad_norm": 0.19604428112506866,
+      "learning_rate": 2.4754297102230923e-05,
+      "loss": 0.3756,
       "step": 71895
     },
     {
-      "epoch": 2.53,
-      "learning_rate": 2.5748247330690557e-05,
-      "loss": 0.2532,
+      "epoch": 2.5912711284102787,
+      "grad_norm": 0.20184142887592316,
+      "learning_rate": 2.4751379111197642e-05,
+      "loss": 0.3813,
       "step": 71900
     },
     {
-      "epoch": 2.53,
-      "learning_rate": 2.5745399880752143e-05,
-      "loss": 0.2563,
+      "epoch": 2.5914513280715035,
+      "grad_norm": 0.21380701661109924,
+      "learning_rate": 2.4748461123551768e-05,
+      "loss": 0.3822,
       "step": 71905
     },
     {
-      "epoch": 2.53,
-      "learning_rate": 2.5742552421135195e-05,
-      "loss": 0.2803,
+      "epoch": 2.5916315277327278,
+      "grad_norm": 0.22639767825603485,
+      "learning_rate": 2.4745543139333067e-05,
+      "loss": 0.4095,
       "step": 71910
     },
     {
-      "epoch": 2.53,
-      "learning_rate": 2.573970495187669e-05,
-      "loss": 0.2673,
+      "epoch": 2.5918117273939525,
+      "grad_norm": 0.21239355206489563,
+      "learning_rate": 2.4742625158581286e-05,
+      "loss": 0.4324,
       "step": 71915
     },
     {
-      "epoch": 2.53,
-      "learning_rate": 2.5736857473013616e-05,
-      "loss": 0.2969,
+      "epoch": 2.5919919270551772,
+      "grad_norm": 0.19603148102760315,
+      "learning_rate": 2.4739707181336193e-05,
+      "loss": 0.3851,
       "step": 71920
     },
     {
-      "epoch": 2.53,
-      "learning_rate": 2.573400998458293e-05,
-      "loss": 0.2601,
+      "epoch": 2.5921721267164015,
+      "grad_norm": 0.21165809035301208,
+      "learning_rate": 2.4736789207637543e-05,
+      "loss": 0.3982,
       "step": 71925
     },
     {
-      "epoch": 2.53,
-      "learning_rate": 2.5731162486621602e-05,
-      "loss": 0.2824,
+      "epoch": 2.5923523263776262,
+      "grad_norm": 0.22291170060634613,
+      "learning_rate": 2.473387123752508e-05,
+      "loss": 0.3934,
       "step": 71930
     },
     {
-      "epoch": 2.53,
-      "learning_rate": 2.5728314979166618e-05,
-      "loss": 0.2398,
+      "epoch": 2.592532526038851,
+      "grad_norm": 0.1911042034626007,
+      "learning_rate": 2.4730953271038575e-05,
+      "loss": 0.3858,
       "step": 71935
     },
     {
-      "epoch": 2.53,
-      "learning_rate": 2.572546746225494e-05,
-      "loss": 0.253,
+      "epoch": 2.5927127257000757,
+      "grad_norm": 0.2135748714208603,
+      "learning_rate": 2.472803530821778e-05,
+      "loss": 0.3912,
       "step": 71940
     },
     {
-      "epoch": 2.53,
-      "learning_rate": 2.572261993592354e-05,
-      "loss": 0.266,
+      "epoch": 2.5928929253613004,
+      "grad_norm": 0.20679445564746857,
+      "learning_rate": 2.472511734910245e-05,
+      "loss": 0.369,
       "step": 71945
     },
     {
-      "epoch": 2.53,
-      "learning_rate": 2.57197724002094e-05,
-      "loss": 0.2691,
+      "epoch": 2.593073125022525,
+      "grad_norm": 0.22252236306667328,
+      "learning_rate": 2.472219939373234e-05,
+      "loss": 0.4035,
       "step": 71950
     },
     {
-      "epoch": 2.53,
-      "learning_rate": 2.5716924855149495e-05,
-      "loss": 0.2848,
+      "epoch": 2.5932533246837495,
+      "grad_norm": 0.20789240300655365,
+      "learning_rate": 2.471928144214721e-05,
+      "loss": 0.3496,
       "step": 71955
     },
     {
-      "epoch": 2.53,
-      "learning_rate": 2.5714077300780792e-05,
-      "loss": 0.2687,
+      "epoch": 2.593433524344974,
+      "grad_norm": 0.24072077870368958,
+      "learning_rate": 2.4716363494386817e-05,
+      "loss": 0.3821,
       "step": 71960
     },
     {
-      "epoch": 2.53,
-      "learning_rate": 2.5711229737140263e-05,
-      "loss": 0.2578,
+      "epoch": 2.593613724006199,
+      "grad_norm": 0.21749582886695862,
+      "learning_rate": 2.471344555049092e-05,
+      "loss": 0.3826,
       "step": 71965
     },
     {
-      "epoch": 2.53,
-      "learning_rate": 2.5708382164264883e-05,
-      "loss": 0.247,
+      "epoch": 2.5937939236674232,
+      "grad_norm": 0.17097270488739014,
+      "learning_rate": 2.4710527610499265e-05,
+      "loss": 0.352,
       "step": 71970
     },
     {
-      "epoch": 2.53,
-      "learning_rate": 2.5705534582191626e-05,
-      "loss": 0.2634,
+      "epoch": 2.593974123328648,
+      "grad_norm": 0.20925599336624146,
+      "learning_rate": 2.470760967445162e-05,
+      "loss": 0.3946,
       "step": 71975
     },
     {
-      "epoch": 2.53,
-      "learning_rate": 2.5702686990957474e-05,
-      "loss": 0.2676,
+      "epoch": 2.5941543229898727,
+      "grad_norm": 0.16980622708797455,
+      "learning_rate": 2.470469174238774e-05,
+      "loss": 0.3889,
       "step": 71980
     },
     {
-      "epoch": 2.53,
-      "learning_rate": 2.569983939059939e-05,
-      "loss": 0.2804,
+      "epoch": 2.5943345226510974,
+      "grad_norm": 0.1989830583333969,
+      "learning_rate": 2.4701773814347366e-05,
+      "loss": 0.3903,
       "step": 71985
     },
     {
-      "epoch": 2.53,
-      "learning_rate": 2.569699178115436e-05,
-      "loss": 0.2901,
+      "epoch": 2.594514722312322,
+      "grad_norm": 0.22819073498249054,
+      "learning_rate": 2.469885589037028e-05,
+      "loss": 0.3905,
       "step": 71990
     },
     {
-      "epoch": 2.53,
-      "learning_rate": 2.5694144162659344e-05,
-      "loss": 0.2714,
+      "epoch": 2.594694921973547,
+      "grad_norm": 0.23240187764167786,
+      "learning_rate": 2.4695937970496212e-05,
+      "loss": 0.3952,
       "step": 71995
     },
     {
-      "epoch": 2.53,
-      "learning_rate": 2.569129653515132e-05,
-      "loss": 0.2763,
+      "epoch": 2.594875121634771,
+      "grad_norm": 0.20458155870437622,
+      "learning_rate": 2.4693020054764936e-05,
+      "loss": 0.3712,
       "step": 72000
     },
     {
-      "epoch": 2.53,
-      "eval_loss": 0.26415836811065674,
-      "eval_runtime": 10.5519,
-      "eval_samples_per_second": 9.477,
-      "eval_steps_per_second": 9.477,
+      "epoch": 2.594875121634771,
+      "eval_loss": 0.43242183327674866,
+      "eval_runtime": 3.5303,
+      "eval_samples_per_second": 28.327,
+      "eval_steps_per_second": 7.082,
       "step": 72000
     },
     {
-      "epoch": 2.53,
-      "learning_rate": 2.5688448898667274e-05,
-      "loss": 0.2591,
+      "epoch": 2.595055321295996,
+      "grad_norm": 0.2061101496219635,
+      "learning_rate": 2.4690102143216214e-05,
+      "loss": 0.3734,
       "step": 72005
     },
     {
-      "epoch": 2.53,
-      "learning_rate": 2.5685601253244172e-05,
-      "loss": 0.2776,
+      "epoch": 2.5952355209572207,
+      "grad_norm": 0.2840164303779602,
+      "learning_rate": 2.4687184235889768e-05,
+      "loss": 0.3939,
       "step": 72010
     },
     {
-      "epoch": 2.53,
-      "learning_rate": 2.5682753598918985e-05,
-      "loss": 0.2975,
+      "epoch": 2.5954157206184454,
+      "grad_norm": 0.3389698266983032,
+      "learning_rate": 2.468426633282539e-05,
+      "loss": 0.3925,
       "step": 72015
     },
     {
-      "epoch": 2.53,
-      "learning_rate": 2.56799059357287e-05,
-      "loss": 0.2368,
+      "epoch": 2.5955959202796697,
+      "grad_norm": 0.22343645989894867,
+      "learning_rate": 2.4681348434062825e-05,
+      "loss": 0.4075,
       "step": 72020
     },
     {
-      "epoch": 2.53,
-      "learning_rate": 2.5677058263710274e-05,
-      "loss": 0.2852,
+      "epoch": 2.5957761199408944,
+      "grad_norm": 0.23399618268013,
+      "learning_rate": 2.467843053964181e-05,
+      "loss": 0.3981,
       "step": 72025
     },
     {
-      "epoch": 2.53,
-      "learning_rate": 2.5674210582900698e-05,
-      "loss": 0.3003,
+      "epoch": 2.595956319602119,
+      "grad_norm": 0.25679734349250793,
+      "learning_rate": 2.4675512649602134e-05,
+      "loss": 0.3848,
       "step": 72030
     },
     {
-      "epoch": 2.53,
-      "learning_rate": 2.567136289333694e-05,
-      "loss": 0.2511,
+      "epoch": 2.596136519263344,
+      "grad_norm": 0.19120201468467712,
+      "learning_rate": 2.467259476398352e-05,
+      "loss": 0.3813,
       "step": 72035
     },
     {
-      "epoch": 2.53,
-      "learning_rate": 2.5668515195055988e-05,
-      "loss": 0.2747,
+      "epoch": 2.5963167189245686,
+      "grad_norm": 0.24785567820072174,
+      "learning_rate": 2.4669676882825754e-05,
+      "loss": 0.3926,
       "step": 72040
     },
     {
-      "epoch": 2.53,
-      "learning_rate": 2.566566748809479e-05,
-      "loss": 0.2807,
+      "epoch": 2.596496918585793,
+      "grad_norm": 0.21031877398490906,
+      "learning_rate": 2.4666759006168572e-05,
+      "loss": 0.3782,
       "step": 72045
     },
     {
-      "epoch": 2.53,
-      "learning_rate": 2.5662819772490348e-05,
-      "loss": 0.272,
+      "epoch": 2.5966771182470176,
+      "grad_norm": 0.22502188384532928,
+      "learning_rate": 2.4663841134051727e-05,
+      "loss": 0.389,
       "step": 72050
     },
     {
-      "epoch": 2.54,
-      "learning_rate": 2.5659972048279623e-05,
-      "loss": 0.2606,
+      "epoch": 2.5968573179082424,
+      "grad_norm": 0.20378191769123077,
+      "learning_rate": 2.4660923266514986e-05,
+      "loss": 0.4028,
       "step": 72055
     },
     {
-      "epoch": 2.54,
-      "learning_rate": 2.5657124315499598e-05,
-      "loss": 0.2775,
+      "epoch": 2.597037517569467,
+      "grad_norm": 0.19002820551395416,
+      "learning_rate": 2.4658005403598098e-05,
+      "loss": 0.3727,
       "step": 72060
     },
     {
-      "epoch": 2.54,
-      "learning_rate": 2.5654276574187254e-05,
-      "loss": 0.2746,
+      "epoch": 2.5972177172306914,
+      "grad_norm": 0.19522058963775635,
+      "learning_rate": 2.4655087545340823e-05,
+      "loss": 0.3779,
       "step": 72065
     },
     {
-      "epoch": 2.54,
-      "learning_rate": 2.565142882437955e-05,
-      "loss": 0.2642,
+      "epoch": 2.597397916891916,
+      "grad_norm": 0.23693622648715973,
+      "learning_rate": 2.4652169691782914e-05,
+      "loss": 0.3623,
       "step": 72070
     },
     {
-      "epoch": 2.54,
-      "learning_rate": 2.564858106611348e-05,
-      "loss": 0.2627,
+      "epoch": 2.597578116553141,
+      "grad_norm": 0.18634121119976044,
+      "learning_rate": 2.464925184296412e-05,
+      "loss": 0.3996,
       "step": 72075
     },
     {
-      "epoch": 2.54,
-      "learning_rate": 2.5645733299426006e-05,
-      "loss": 0.2554,
+      "epoch": 2.5977583162143656,
+      "grad_norm": 0.18742458522319794,
+      "learning_rate": 2.4646333998924205e-05,
+      "loss": 0.3633,
       "step": 72080
     },
     {
-      "epoch": 2.54,
-      "learning_rate": 2.564288552435411e-05,
-      "loss": 0.2609,
+      "epoch": 2.5979385158755903,
+      "grad_norm": 0.21051351726055145,
+      "learning_rate": 2.4643416159702925e-05,
+      "loss": 0.4086,
       "step": 72085
     },
     {
-      "epoch": 2.54,
-      "learning_rate": 2.564003774093477e-05,
-      "loss": 0.2533,
+      "epoch": 2.598118715536815,
+      "grad_norm": 0.1712328940629959,
+      "learning_rate": 2.4640498325340022e-05,
+      "loss": 0.3798,
       "step": 72090
     },
     {
-      "epoch": 2.54,
-      "learning_rate": 2.563718994920497e-05,
-      "loss": 0.2824,
+      "epoch": 2.5982989151980393,
+      "grad_norm": 0.19666485488414764,
+      "learning_rate": 2.4637580495875267e-05,
+      "loss": 0.3856,
       "step": 72095
     },
     {
-      "epoch": 2.54,
-      "learning_rate": 2.5634342149201667e-05,
-      "loss": 0.2591,
+      "epoch": 2.598479114859264,
+      "grad_norm": 0.15711526572704315,
+      "learning_rate": 2.4634662671348403e-05,
+      "loss": 0.3996,
       "step": 72100
     },
     {
-      "epoch": 2.54,
-      "learning_rate": 2.5631494340961858e-05,
-      "loss": 0.2786,
+      "epoch": 2.598659314520489,
+      "grad_norm": 0.21674515306949615,
+      "learning_rate": 2.4631744851799192e-05,
+      "loss": 0.3851,
       "step": 72105
     },
     {
-      "epoch": 2.54,
-      "learning_rate": 2.5628646524522503e-05,
-      "loss": 0.2509,
+      "epoch": 2.598839514181713,
+      "grad_norm": 0.1570635885000229,
+      "learning_rate": 2.4628827037267397e-05,
+      "loss": 0.3768,
       "step": 72110
     },
     {
-      "epoch": 2.54,
-      "learning_rate": 2.5625798699920596e-05,
-      "loss": 0.2841,
+      "epoch": 2.599019713842938,
+      "grad_norm": 0.188796266913414,
+      "learning_rate": 2.462590922779274e-05,
+      "loss": 0.4152,
       "step": 72115
     },
     {
-      "epoch": 2.54,
-      "learning_rate": 2.5622950867193095e-05,
-      "loss": 0.2978,
+      "epoch": 2.5991999135041626,
+      "grad_norm": 0.22848844528198242,
+      "learning_rate": 2.4622991423415016e-05,
+      "loss": 0.3614,
       "step": 72120
     },
     {
-      "epoch": 2.54,
-      "learning_rate": 2.5620103026377003e-05,
-      "loss": 0.2747,
+      "epoch": 2.5993801131653873,
+      "grad_norm": 0.23881231248378754,
+      "learning_rate": 2.4620073624173952e-05,
+      "loss": 0.3662,
       "step": 72125
     },
     {
-      "epoch": 2.54,
-      "learning_rate": 2.561725517750927e-05,
-      "loss": 0.2687,
+      "epoch": 2.599560312826612,
+      "grad_norm": 0.1681346446275711,
+      "learning_rate": 2.461715583010931e-05,
+      "loss": 0.4039,
       "step": 72130
     },
     {
-      "epoch": 2.54,
-      "learning_rate": 2.5614407320626886e-05,
-      "loss": 0.2913,
+      "epoch": 2.5997405124878368,
+      "grad_norm": 0.19078193604946136,
+      "learning_rate": 2.461423804126085e-05,
+      "loss": 0.3721,
       "step": 72135
     },
     {
-      "epoch": 2.54,
-      "learning_rate": 2.5611559455766825e-05,
-      "loss": 0.2905,
+      "epoch": 2.599920712149061,
+      "grad_norm": 0.19204656779766083,
+      "learning_rate": 2.4611320257668318e-05,
+      "loss": 0.3735,
       "step": 72140
     },
     {
-      "epoch": 2.54,
-      "learning_rate": 2.5608711582966077e-05,
-      "loss": 0.2821,
+      "epoch": 2.600100911810286,
+      "grad_norm": 0.2286890149116516,
+      "learning_rate": 2.4608402479371475e-05,
+      "loss": 0.3929,
       "step": 72145
     },
     {
-      "epoch": 2.54,
-      "learning_rate": 2.5605863702261607e-05,
-      "loss": 0.2703,
+      "epoch": 2.6002811114715105,
+      "grad_norm": 0.18372632563114166,
+      "learning_rate": 2.4605484706410072e-05,
+      "loss": 0.3658,
       "step": 72150
     },
     {
-      "epoch": 2.54,
-      "learning_rate": 2.5603015813690396e-05,
-      "loss": 0.2875,
+      "epoch": 2.600461311132735,
+      "grad_norm": 0.1332695037126541,
+      "learning_rate": 2.460256693882386e-05,
+      "loss": 0.3649,
       "step": 72155
     },
     {
-      "epoch": 2.54,
-      "learning_rate": 2.560016791728942e-05,
-      "loss": 0.2575,
+      "epoch": 2.6006415107939596,
+      "grad_norm": 0.22151631116867065,
+      "learning_rate": 2.4599649176652602e-05,
+      "loss": 0.425,
       "step": 72160
     },
     {
-      "epoch": 2.54,
-      "learning_rate": 2.5597320013095656e-05,
-      "loss": 0.2838,
+      "epoch": 2.6008217104551843,
+      "grad_norm": 0.25563111901283264,
+      "learning_rate": 2.4596731419936044e-05,
+      "loss": 0.4232,
       "step": 72165
     },
     {
-      "epoch": 2.54,
-      "learning_rate": 2.559447210114609e-05,
-      "loss": 0.2722,
+      "epoch": 2.601001910116409,
+      "grad_norm": 0.19068589806556702,
+      "learning_rate": 2.4593813668713942e-05,
+      "loss": 0.3822,
       "step": 72170
     },
     {
-      "epoch": 2.54,
-      "learning_rate": 2.5591624181477693e-05,
-      "loss": 0.2918,
+      "epoch": 2.6011821097776338,
+      "grad_norm": 0.2579037845134735,
+      "learning_rate": 2.459089592302605e-05,
+      "loss": 0.3904,
       "step": 72175
     },
     {
-      "epoch": 2.54,
-      "learning_rate": 2.558877625412745e-05,
-      "loss": 0.2727,
+      "epoch": 2.6013623094388585,
+      "grad_norm": 0.1858731061220169,
+      "learning_rate": 2.458797818291212e-05,
+      "loss": 0.3894,
       "step": 72180
     },
     {
-      "epoch": 2.54,
-      "learning_rate": 2.558592831913233e-05,
-      "loss": 0.2597,
+      "epoch": 2.601542509100083,
+      "grad_norm": 0.2007628083229065,
+      "learning_rate": 2.458506044841191e-05,
+      "loss": 0.399,
       "step": 72185
     },
     {
-      "epoch": 2.54,
-      "learning_rate": 2.5583080376529318e-05,
-      "loss": 0.2616,
+      "epoch": 2.6017227087613075,
+      "grad_norm": 0.22041386365890503,
+      "learning_rate": 2.4582142719565173e-05,
+      "loss": 0.4187,
       "step": 72190
     },
     {
-      "epoch": 2.54,
-      "learning_rate": 2.558023242635539e-05,
-      "loss": 0.2793,
+      "epoch": 2.6019029084225322,
+      "grad_norm": 0.22453515231609344,
+      "learning_rate": 2.4579224996411655e-05,
+      "loss": 0.4086,
       "step": 72195
     },
     {
-      "epoch": 2.54,
-      "learning_rate": 2.5577384468647537e-05,
-      "loss": 0.2957,
+      "epoch": 2.6020831080837565,
+      "grad_norm": 0.2213824838399887,
+      "learning_rate": 2.457630727899112e-05,
+      "loss": 0.4141,
       "step": 72200
     },
     {
-      "epoch": 2.54,
-      "learning_rate": 2.5574536503442715e-05,
-      "loss": 0.2607,
+      "epoch": 2.6022633077449813,
+      "grad_norm": 0.20723369717597961,
+      "learning_rate": 2.4573389567343323e-05,
+      "loss": 0.3806,
       "step": 72205
     },
     {
-      "epoch": 2.54,
-      "learning_rate": 2.5571688530777928e-05,
-      "loss": 0.2442,
+      "epoch": 2.602443507406206,
+      "grad_norm": 0.2115570306777954,
+      "learning_rate": 2.4570471861507994e-05,
+      "loss": 0.38,
       "step": 72210
     },
     {
-      "epoch": 2.54,
-      "learning_rate": 2.5568840550690127e-05,
-      "loss": 0.2868,
+      "epoch": 2.6026237070674307,
+      "grad_norm": 0.21702751517295837,
+      "learning_rate": 2.4567554161524917e-05,
+      "loss": 0.3766,
       "step": 72215
     },
     {
-      "epoch": 2.54,
-      "learning_rate": 2.5565992563216313e-05,
-      "loss": 0.2637,
+      "epoch": 2.6028039067286555,
+      "grad_norm": 0.22824609279632568,
+      "learning_rate": 2.4564636467433814e-05,
+      "loss": 0.3602,
       "step": 72220
     },
     {
-      "epoch": 2.54,
-      "learning_rate": 2.5563144568393456e-05,
-      "loss": 0.2579,
+      "epoch": 2.60298410638988,
+      "grad_norm": 0.2511116564273834,
+      "learning_rate": 2.4561718779274474e-05,
+      "loss": 0.3884,
       "step": 72225
     },
     {
-      "epoch": 2.54,
-      "learning_rate": 2.5560296566258546e-05,
-      "loss": 0.2757,
+      "epoch": 2.6031643060511045,
+      "grad_norm": 0.19371597468852997,
+      "learning_rate": 2.4558801097086627e-05,
+      "loss": 0.3869,
       "step": 72230
     },
     {
-      "epoch": 2.54,
-      "learning_rate": 2.5557448556848545e-05,
-      "loss": 0.2823,
+      "epoch": 2.6033445057123292,
+      "grad_norm": 0.27392515540122986,
+      "learning_rate": 2.455588342091002e-05,
+      "loss": 0.4111,
       "step": 72235
     },
     {
-      "epoch": 2.54,
-      "learning_rate": 2.555460054020045e-05,
-      "loss": 0.3012,
+      "epoch": 2.603524705373554,
+      "grad_norm": 0.20092566311359406,
+      "learning_rate": 2.4552965750784422e-05,
+      "loss": 0.4155,
       "step": 72240
     },
     {
-      "epoch": 2.54,
-      "learning_rate": 2.555175251635123e-05,
-      "loss": 0.2845,
+      "epoch": 2.6037049050347787,
+      "grad_norm": 0.18741832673549652,
+      "learning_rate": 2.4550048086749572e-05,
+      "loss": 0.4058,
       "step": 72245
     },
     {
-      "epoch": 2.54,
-      "learning_rate": 2.5548904485337864e-05,
-      "loss": 0.2942,
+      "epoch": 2.603885104696003,
+      "grad_norm": 0.20643602311611176,
+      "learning_rate": 2.454713042884524e-05,
+      "loss": 0.4002,
       "step": 72250
     },
     {
-      "epoch": 2.54,
-      "learning_rate": 2.5546056447197335e-05,
-      "loss": 0.2608,
+      "epoch": 2.6040653043572277,
+      "grad_norm": 0.1865357607603073,
+      "learning_rate": 2.4544212777111164e-05,
+      "loss": 0.3739,
       "step": 72255
     },
     {
-      "epoch": 2.54,
-      "learning_rate": 2.554320840196663e-05,
-      "loss": 0.2834,
+      "epoch": 2.6042455040184525,
+      "grad_norm": 0.165154829621315,
+      "learning_rate": 2.4541295131587098e-05,
+      "loss": 0.4122,
       "step": 72260
     },
     {
-      "epoch": 2.54,
-      "learning_rate": 2.5540360349682713e-05,
-      "loss": 0.2809,
+      "epoch": 2.604425703679677,
+      "grad_norm": 0.19582092761993408,
+      "learning_rate": 2.4538377492312797e-05,
+      "loss": 0.4369,
       "step": 72265
     },
     {
-      "epoch": 2.54,
-      "learning_rate": 2.5537512290382576e-05,
-      "loss": 0.2622,
+      "epoch": 2.604605903340902,
+      "grad_norm": 0.18715718388557434,
+      "learning_rate": 2.453545985932802e-05,
+      "loss": 0.3845,
       "step": 72270
     },
     {
-      "epoch": 2.54,
-      "learning_rate": 2.5534664224103193e-05,
-      "loss": 0.2804,
+      "epoch": 2.604786103002126,
+      "grad_norm": 0.22059041261672974,
+      "learning_rate": 2.4532542232672504e-05,
+      "loss": 0.3972,
       "step": 72275
     },
     {
-      "epoch": 2.54,
-      "learning_rate": 2.5531816150881554e-05,
-      "loss": 0.2637,
+      "epoch": 2.604966302663351,
+      "grad_norm": 0.22253043949604034,
+      "learning_rate": 2.4529624612386015e-05,
+      "loss": 0.4599,
       "step": 72280
     },
     {
-      "epoch": 2.54,
-      "learning_rate": 2.552896807075463e-05,
-      "loss": 0.2903,
+      "epoch": 2.6051465023245757,
+      "grad_norm": 0.22209565341472626,
+      "learning_rate": 2.4526706998508296e-05,
+      "loss": 0.39,
       "step": 72285
     },
     {
-      "epoch": 2.54,
-      "learning_rate": 2.5526119983759404e-05,
-      "loss": 0.2552,
+      "epoch": 2.6053267019858004,
+      "grad_norm": 0.2502153515815735,
+      "learning_rate": 2.4523789391079103e-05,
+      "loss": 0.3973,
       "step": 72290
     },
     {
-      "epoch": 2.54,
-      "learning_rate": 2.552327188993286e-05,
-      "loss": 0.2634,
+      "epoch": 2.6055069016470247,
+      "grad_norm": 0.2121538519859314,
+      "learning_rate": 2.4520871790138196e-05,
+      "loss": 0.362,
       "step": 72295
     },
     {
-      "epoch": 2.54,
-      "learning_rate": 2.5520423789311972e-05,
-      "loss": 0.2673,
+      "epoch": 2.6056871013082494,
+      "grad_norm": 0.22222217917442322,
+      "learning_rate": 2.4517954195725305e-05,
+      "loss": 0.4204,
       "step": 72300
     },
     {
-      "epoch": 2.54,
-      "learning_rate": 2.551757568193372e-05,
-      "loss": 0.2947,
+      "epoch": 2.605867300969474,
+      "grad_norm": 0.1976117193698883,
+      "learning_rate": 2.4515036607880208e-05,
+      "loss": 0.3909,
       "step": 72305
     },
     {
-      "epoch": 2.54,
-      "learning_rate": 2.551472756783509e-05,
-      "loss": 0.258,
+      "epoch": 2.606047500630699,
+      "grad_norm": 0.16748464107513428,
+      "learning_rate": 2.451211902664264e-05,
+      "loss": 0.3622,
       "step": 72310
     },
     {
-      "epoch": 2.54,
-      "learning_rate": 2.5511879447053068e-05,
-      "loss": 0.2688,
+      "epoch": 2.6062277002919236,
+      "grad_norm": 0.15831992030143738,
+      "learning_rate": 2.4509201452052338e-05,
+      "loss": 0.3806,
       "step": 72315
     },
     {
-      "epoch": 2.54,
-      "learning_rate": 2.5509031319624627e-05,
-      "loss": 0.2813,
+      "epoch": 2.606407899953148,
+      "grad_norm": 0.18840594589710236,
+      "learning_rate": 2.4506283884149094e-05,
+      "loss": 0.4006,
       "step": 72320
     },
     {
-      "epoch": 2.54,
-      "learning_rate": 2.5506183185586752e-05,
-      "loss": 0.2619,
+      "epoch": 2.6065880996143727,
+      "grad_norm": 0.28430864214897156,
+      "learning_rate": 2.450336632297262e-05,
+      "loss": 0.4045,
       "step": 72325
     },
     {
-      "epoch": 2.54,
-      "learning_rate": 2.5503335044976405e-05,
-      "loss": 0.2896,
+      "epoch": 2.6067682992755974,
+      "grad_norm": 0.24897444248199463,
+      "learning_rate": 2.45004487685627e-05,
+      "loss": 0.4207,
       "step": 72330
     },
     {
-      "epoch": 2.54,
-      "learning_rate": 2.5500486897830607e-05,
-      "loss": 0.2547,
+      "epoch": 2.606948498936822,
+      "grad_norm": 0.2031540423631668,
+      "learning_rate": 2.4497531220959056e-05,
+      "loss": 0.3999,
       "step": 72335
     },
     {
-      "epoch": 2.55,
-      "learning_rate": 2.5497638744186304e-05,
-      "loss": 0.2922,
+      "epoch": 2.6071286985980464,
+      "grad_norm": 0.26375287771224976,
+      "learning_rate": 2.4494613680201456e-05,
+      "loss": 0.4072,
       "step": 72340
     },
     {
-      "epoch": 2.55,
-      "learning_rate": 2.5494790584080498e-05,
-      "loss": 0.2762,
+      "epoch": 2.607308898259271,
+      "grad_norm": 0.2009999305009842,
+      "learning_rate": 2.4491696146329648e-05,
+      "loss": 0.366,
       "step": 72345
     },
     {
-      "epoch": 2.55,
-      "learning_rate": 2.5491942417550154e-05,
-      "loss": 0.2727,
+      "epoch": 2.607489097920496,
+      "grad_norm": 0.24075284600257874,
+      "learning_rate": 2.448877861938338e-05,
+      "loss": 0.4361,
       "step": 72350
     },
     {
-      "epoch": 2.55,
-      "learning_rate": 2.548909424463226e-05,
-      "loss": 0.2848,
+      "epoch": 2.6076692975817206,
+      "grad_norm": 0.22679272294044495,
+      "learning_rate": 2.4485861099402402e-05,
+      "loss": 0.4212,
       "step": 72355
     },
     {
-      "epoch": 2.55,
-      "learning_rate": 2.5486246065363807e-05,
-      "loss": 0.2811,
+      "epoch": 2.6078494972429453,
+      "grad_norm": 0.18782539665699005,
+      "learning_rate": 2.4482943586426473e-05,
+      "loss": 0.3904,
       "step": 72360
     },
     {
-      "epoch": 2.55,
-      "learning_rate": 2.5483397879781768e-05,
-      "loss": 0.2931,
+      "epoch": 2.60802969690417,
+      "grad_norm": 0.18936201930046082,
+      "learning_rate": 2.448002608049533e-05,
+      "loss": 0.3494,
       "step": 72365
     },
     {
-      "epoch": 2.55,
-      "learning_rate": 2.5480549687923123e-05,
-      "loss": 0.2764,
+      "epoch": 2.6082098965653944,
+      "grad_norm": 0.22821679711341858,
+      "learning_rate": 2.4477108581648734e-05,
+      "loss": 0.4238,
       "step": 72370
     },
     {
-      "epoch": 2.55,
-      "learning_rate": 2.5477701489824863e-05,
-      "loss": 0.2898,
+      "epoch": 2.608390096226619,
+      "grad_norm": 0.19849362969398499,
+      "learning_rate": 2.4474191089926438e-05,
+      "loss": 0.3797,
       "step": 72375
     },
     {
-      "epoch": 2.55,
-      "learning_rate": 2.5474853285523958e-05,
-      "loss": 0.2849,
+      "epoch": 2.608570295887844,
+      "grad_norm": 0.21136881411075592,
+      "learning_rate": 2.447127360536818e-05,
+      "loss": 0.3807,
       "step": 72380
     },
     {
-      "epoch": 2.55,
-      "learning_rate": 2.54720050750574e-05,
-      "loss": 0.267,
+      "epoch": 2.608750495549068,
+      "grad_norm": 0.23229160904884338,
+      "learning_rate": 2.446835612801372e-05,
+      "loss": 0.4021,
       "step": 72385
     },
     {
-      "epoch": 2.55,
-      "learning_rate": 2.5469156858462156e-05,
-      "loss": 0.2518,
+      "epoch": 2.608930695210293,
+      "grad_norm": 0.2442099153995514,
+      "learning_rate": 2.446543865790281e-05,
+      "loss": 0.4003,
       "step": 72390
     },
     {
-      "epoch": 2.55,
-      "learning_rate": 2.5466308635775238e-05,
-      "loss": 0.2578,
+      "epoch": 2.6091108948715176,
+      "grad_norm": 0.2331792116165161,
+      "learning_rate": 2.446252119507518e-05,
+      "loss": 0.4425,
       "step": 72395
     },
     {
-      "epoch": 2.55,
-      "learning_rate": 2.5463460407033597e-05,
-      "loss": 0.2703,
+      "epoch": 2.6092910945327423,
+      "grad_norm": 0.16892333328723907,
+      "learning_rate": 2.445960373957061e-05,
+      "loss": 0.3777,
       "step": 72400
     },
     {
-      "epoch": 2.55,
-      "learning_rate": 2.5460612172274234e-05,
-      "loss": 0.2511,
+      "epoch": 2.609471294193967,
+      "grad_norm": 0.18657229840755463,
+      "learning_rate": 2.445668629142882e-05,
+      "loss": 0.3742,
       "step": 72405
     },
     {
-      "epoch": 2.55,
-      "learning_rate": 2.5457763931534117e-05,
-      "loss": 0.2541,
+      "epoch": 2.609651493855192,
+      "grad_norm": 0.2221583127975464,
+      "learning_rate": 2.4453768850689587e-05,
+      "loss": 0.4194,
       "step": 72410
     },
     {
-      "epoch": 2.55,
-      "learning_rate": 2.5454915684850238e-05,
-      "loss": 0.2788,
+      "epoch": 2.609831693516416,
+      "grad_norm": 0.1793508529663086,
+      "learning_rate": 2.4450851417392645e-05,
+      "loss": 0.397,
       "step": 72415
     },
     {
-      "epoch": 2.55,
-      "learning_rate": 2.5452067432259585e-05,
-      "loss": 0.2648,
+      "epoch": 2.610011893177641,
+      "grad_norm": 0.1674244999885559,
+      "learning_rate": 2.444793399157774e-05,
+      "loss": 0.3605,
       "step": 72420
     },
     {
-      "epoch": 2.55,
-      "learning_rate": 2.5449219173799132e-05,
-      "loss": 0.2611,
+      "epoch": 2.6101920928388656,
+      "grad_norm": 0.18199168145656586,
+      "learning_rate": 2.4445016573284632e-05,
+      "loss": 0.3821,
       "step": 72425
     },
     {
-      "epoch": 2.55,
-      "learning_rate": 2.5446370909505868e-05,
-      "loss": 0.2631,
+      "epoch": 2.61037229250009,
+      "grad_norm": 0.20687104761600494,
+      "learning_rate": 2.444209916255306e-05,
+      "loss": 0.3999,
       "step": 72430
     },
     {
-      "epoch": 2.55,
-      "learning_rate": 2.5443522639416767e-05,
-      "loss": 0.2589,
+      "epoch": 2.6105524921613146,
+      "grad_norm": 0.21444140374660492,
+      "learning_rate": 2.4439181759422787e-05,
+      "loss": 0.3575,
       "step": 72435
     },
     {
-      "epoch": 2.55,
-      "learning_rate": 2.544067436356882e-05,
-      "loss": 0.2697,
+      "epoch": 2.6107326918225393,
+      "grad_norm": 0.18158109486103058,
+      "learning_rate": 2.4436264363933554e-05,
+      "loss": 0.3997,
       "step": 72440
     },
     {
-      "epoch": 2.55,
-      "learning_rate": 2.5437826081999e-05,
-      "loss": 0.2398,
+      "epoch": 2.610912891483764,
+      "grad_norm": 0.23257912695407867,
+      "learning_rate": 2.4433346976125103e-05,
+      "loss": 0.4247,
       "step": 72445
     },
     {
-      "epoch": 2.55,
-      "learning_rate": 2.543497779474431e-05,
-      "loss": 0.2626,
+      "epoch": 2.6110930911449888,
+      "grad_norm": 0.24175713956356049,
+      "learning_rate": 2.4430429596037195e-05,
+      "loss": 0.407,
       "step": 72450
     },
     {
-      "epoch": 2.55,
-      "learning_rate": 2.5432129501841704e-05,
-      "loss": 0.2503,
+      "epoch": 2.6112732908062135,
+      "grad_norm": 0.24281500279903412,
+      "learning_rate": 2.442751222370957e-05,
+      "loss": 0.4246,
       "step": 72455
     },
     {
-      "epoch": 2.55,
-      "learning_rate": 2.5429281203328198e-05,
-      "loss": 0.2749,
+      "epoch": 2.611453490467438,
+      "grad_norm": 0.21452686190605164,
+      "learning_rate": 2.4424594859181978e-05,
+      "loss": 0.3739,
       "step": 72460
     },
     {
-      "epoch": 2.55,
-      "learning_rate": 2.542643289924075e-05,
-      "loss": 0.2486,
+      "epoch": 2.6116336901286625,
+      "grad_norm": 0.23282234370708466,
+      "learning_rate": 2.4421677502494175e-05,
+      "loss": 0.3948,
       "step": 72465
     },
     {
-      "epoch": 2.55,
-      "learning_rate": 2.5423584589616346e-05,
-      "loss": 0.2838,
+      "epoch": 2.6118138897898873,
+      "grad_norm": 0.24855384230613708,
+      "learning_rate": 2.44187601536859e-05,
+      "loss": 0.4206,
       "step": 72470
     },
     {
-      "epoch": 2.55,
-      "learning_rate": 2.542073627449198e-05,
-      "loss": 0.2734,
+      "epoch": 2.6119940894511116,
+      "grad_norm": 0.22032064199447632,
+      "learning_rate": 2.441584281279691e-05,
+      "loss": 0.3899,
       "step": 72475
     },
     {
-      "epoch": 2.55,
-      "learning_rate": 2.5417887953904633e-05,
-      "loss": 0.295,
+      "epoch": 2.6121742891123363,
+      "grad_norm": 0.28760719299316406,
+      "learning_rate": 2.4412925479866946e-05,
+      "loss": 0.3476,
       "step": 72480
     },
     {
-      "epoch": 2.55,
-      "learning_rate": 2.5415039627891287e-05,
-      "loss": 0.2504,
+      "epoch": 2.612354488773561,
+      "grad_norm": 0.22636237740516663,
+      "learning_rate": 2.4410008154935757e-05,
+      "loss": 0.3523,
       "step": 72485
     },
     {
-      "epoch": 2.55,
-      "learning_rate": 2.541219129648892e-05,
-      "loss": 0.2823,
+      "epoch": 2.6125346884347858,
+      "grad_norm": 0.19048437476158142,
+      "learning_rate": 2.4407090838043097e-05,
+      "loss": 0.3876,
       "step": 72490
     },
     {
-      "epoch": 2.55,
-      "learning_rate": 2.5409342959734522e-05,
-      "loss": 0.2523,
+      "epoch": 2.6127148880960105,
+      "grad_norm": 0.20822523534297943,
+      "learning_rate": 2.440417352922871e-05,
+      "loss": 0.4092,
       "step": 72495
     },
     {
-      "epoch": 2.55,
-      "learning_rate": 2.5406494617665084e-05,
-      "loss": 0.2634,
+      "epoch": 2.6128950877572352,
+      "grad_norm": 0.20144416391849518,
+      "learning_rate": 2.4401256228532334e-05,
+      "loss": 0.3772,
       "step": 72500
     },
     {
-      "epoch": 2.55,
-      "eval_loss": 0.26341068744659424,
-      "eval_runtime": 10.5346,
-      "eval_samples_per_second": 9.493,
-      "eval_steps_per_second": 9.493,
+      "epoch": 2.6128950877572352,
+      "eval_loss": 0.432780385017395,
+      "eval_runtime": 3.5337,
+      "eval_samples_per_second": 28.299,
+      "eval_steps_per_second": 7.075,
       "step": 72500
     },
     {
-      "epoch": 2.55,
-      "learning_rate": 2.5403646270317573e-05,
-      "loss": 0.2763,
+      "epoch": 2.6130752874184595,
+      "grad_norm": 0.20547565817832947,
+      "learning_rate": 2.4398338935993742e-05,
+      "loss": 0.4123,
       "step": 72505
     },
     {
-      "epoch": 2.55,
-      "learning_rate": 2.540079791772898e-05,
-      "loss": 0.2665,
+      "epoch": 2.6132554870796842,
+      "grad_norm": 0.2339104562997818,
+      "learning_rate": 2.4395421651652646e-05,
+      "loss": 0.4156,
       "step": 72510
     },
     {
-      "epoch": 2.55,
-      "learning_rate": 2.5397949559936297e-05,
-      "loss": 0.2826,
+      "epoch": 2.613435686740909,
+      "grad_norm": 0.209024116396904,
+      "learning_rate": 2.439250437554883e-05,
+      "loss": 0.3592,
       "step": 72515
     },
     {
-      "epoch": 2.55,
-      "learning_rate": 2.5395101196976488e-05,
-      "loss": 0.2844,
+      "epoch": 2.6136158864021337,
+      "grad_norm": 0.20917484164237976,
+      "learning_rate": 2.438958710772202e-05,
+      "loss": 0.3925,
       "step": 72520
     },
     {
-      "epoch": 2.55,
-      "learning_rate": 2.539225282888656e-05,
-      "loss": 0.2834,
+      "epoch": 2.613796086063358,
+      "grad_norm": 0.23410077393054962,
+      "learning_rate": 2.438666984821196e-05,
+      "loss": 0.3747,
       "step": 72525
     },
     {
-      "epoch": 2.55,
-      "learning_rate": 2.5389404455703487e-05,
-      "loss": 0.2522,
+      "epoch": 2.6139762857245827,
+      "grad_norm": 0.18121454119682312,
+      "learning_rate": 2.4383752597058414e-05,
+      "loss": 0.3926,
       "step": 72530
     },
     {
-      "epoch": 2.55,
-      "learning_rate": 2.5386556077464258e-05,
-      "loss": 0.2763,
+      "epoch": 2.6141564853858075,
+      "grad_norm": 0.1945260465145111,
+      "learning_rate": 2.4380835354301117e-05,
+      "loss": 0.4095,
       "step": 72535
     },
     {
-      "epoch": 2.55,
-      "learning_rate": 2.5383707694205843e-05,
-      "loss": 0.2537,
+      "epoch": 2.614336685047032,
+      "grad_norm": 0.21491825580596924,
+      "learning_rate": 2.437791811997981e-05,
+      "loss": 0.4031,
       "step": 72540
     },
     {
-      "epoch": 2.55,
-      "learning_rate": 2.538085930596525e-05,
-      "loss": 0.2708,
+      "epoch": 2.614516884708257,
+      "grad_norm": 0.23276425898075104,
+      "learning_rate": 2.4375000894134257e-05,
+      "loss": 0.3753,
       "step": 72545
     },
     {
-      "epoch": 2.55,
-      "learning_rate": 2.5378010912779432e-05,
-      "loss": 0.2655,
+      "epoch": 2.6146970843694812,
+      "grad_norm": 0.20412658154964447,
+      "learning_rate": 2.4372083676804187e-05,
+      "loss": 0.4058,
       "step": 72550
     },
     {
-      "epoch": 2.55,
-      "learning_rate": 2.5375162514685408e-05,
-      "loss": 0.309,
+      "epoch": 2.614877284030706,
+      "grad_norm": 0.2030048817396164,
+      "learning_rate": 2.4369166468029367e-05,
+      "loss": 0.371,
       "step": 72555
     },
     {
-      "epoch": 2.55,
-      "learning_rate": 2.5372314111720135e-05,
-      "loss": 0.2705,
+      "epoch": 2.6150574836919307,
+      "grad_norm": 0.22809742391109467,
+      "learning_rate": 2.436624926784953e-05,
+      "loss": 0.3964,
       "step": 72560
     },
     {
-      "epoch": 2.55,
-      "learning_rate": 2.536946570392062e-05,
-      "loss": 0.2456,
+      "epoch": 2.6152376833531554,
+      "grad_norm": 0.20488815009593964,
+      "learning_rate": 2.4363332076304413e-05,
+      "loss": 0.3581,
       "step": 72565
     },
     {
-      "epoch": 2.55,
-      "learning_rate": 2.5366617291323825e-05,
-      "loss": 0.2688,
+      "epoch": 2.6154178830143797,
+      "grad_norm": 0.2278917282819748,
+      "learning_rate": 2.4360414893433784e-05,
+      "loss": 0.3737,
       "step": 72570
     },
     {
-      "epoch": 2.55,
-      "learning_rate": 2.5363768873966747e-05,
-      "loss": 0.2542,
+      "epoch": 2.6155980826756045,
+      "grad_norm": 0.2145715057849884,
+      "learning_rate": 2.435749771927738e-05,
+      "loss": 0.3794,
       "step": 72575
     },
     {
-      "epoch": 2.55,
-      "learning_rate": 2.5360920451886373e-05,
-      "loss": 0.2535,
+      "epoch": 2.615778282336829,
+      "grad_norm": 0.1873069554567337,
+      "learning_rate": 2.435458055387493e-05,
+      "loss": 0.3869,
       "step": 72580
     },
     {
-      "epoch": 2.55,
-      "learning_rate": 2.5358072025119687e-05,
-      "loss": 0.2642,
+      "epoch": 2.615958481998054,
+      "grad_norm": 0.16591453552246094,
+      "learning_rate": 2.4351663397266213e-05,
+      "loss": 0.3601,
       "step": 72585
     },
     {
-      "epoch": 2.55,
-      "learning_rate": 2.535522359370367e-05,
-      "loss": 0.272,
+      "epoch": 2.6161386816592787,
+      "grad_norm": 0.21023093163967133,
+      "learning_rate": 2.434874624949094e-05,
+      "loss": 0.41,
       "step": 72590
     },
     {
-      "epoch": 2.55,
-      "learning_rate": 2.535237515767531e-05,
-      "loss": 0.2651,
+      "epoch": 2.6163188813205034,
+      "grad_norm": 0.17836223542690277,
+      "learning_rate": 2.4345829110588892e-05,
+      "loss": 0.3976,
       "step": 72595
     },
     {
-      "epoch": 2.55,
-      "learning_rate": 2.534952671707159e-05,
-      "loss": 0.2561,
+      "epoch": 2.6164990809817277,
+      "grad_norm": 0.2159523069858551,
+      "learning_rate": 2.4342911980599788e-05,
+      "loss": 0.4025,
       "step": 72600
     },
     {
-      "epoch": 2.55,
-      "learning_rate": 2.534667827192949e-05,
-      "loss": 0.2688,
+      "epoch": 2.6166792806429524,
+      "grad_norm": 0.23821072280406952,
+      "learning_rate": 2.4339994859563368e-05,
+      "loss": 0.4003,
       "step": 72605
     },
     {
-      "epoch": 2.55,
-      "learning_rate": 2.5343829822286008e-05,
-      "loss": 0.2615,
+      "epoch": 2.616859480304177,
+      "grad_norm": 0.2293204665184021,
+      "learning_rate": 2.433707774751941e-05,
+      "loss": 0.3821,
       "step": 72610
     },
     {
-      "epoch": 2.55,
-      "learning_rate": 2.5340981368178125e-05,
-      "loss": 0.2712,
+      "epoch": 2.6170396799654014,
+      "grad_norm": 0.2255268394947052,
+      "learning_rate": 2.433416064450763e-05,
+      "loss": 0.3848,
       "step": 72615
     },
     {
-      "epoch": 2.55,
-      "learning_rate": 2.533813290964282e-05,
-      "loss": 0.3022,
+      "epoch": 2.617219879626626,
+      "grad_norm": 0.18856073915958405,
+      "learning_rate": 2.433124355056778e-05,
+      "loss": 0.4097,
       "step": 72620
     },
     {
-      "epoch": 2.56,
-      "learning_rate": 2.5335284446717083e-05,
-      "loss": 0.2799,
+      "epoch": 2.617400079287851,
+      "grad_norm": 0.221920907497406,
+      "learning_rate": 2.432832646573961e-05,
+      "loss": 0.3756,
       "step": 72625
     },
     {
-      "epoch": 2.56,
-      "learning_rate": 2.533243597943789e-05,
-      "loss": 0.2795,
+      "epoch": 2.6175802789490756,
+      "grad_norm": 0.23313437402248383,
+      "learning_rate": 2.432540939006286e-05,
+      "loss": 0.4022,
       "step": 72630
     },
     {
-      "epoch": 2.56,
-      "learning_rate": 2.5329587507842246e-05,
-      "loss": 0.2702,
+      "epoch": 2.6177604786103004,
+      "grad_norm": 0.17918843030929565,
+      "learning_rate": 2.432249232357728e-05,
+      "loss": 0.394,
       "step": 72635
     },
     {
-      "epoch": 2.56,
-      "learning_rate": 2.5326739031967118e-05,
-      "loss": 0.2808,
+      "epoch": 2.617940678271525,
+      "grad_norm": 0.1900051385164261,
+      "learning_rate": 2.4319575266322607e-05,
+      "loss": 0.4176,
       "step": 72640
     },
     {
-      "epoch": 2.56,
-      "learning_rate": 2.53238905518495e-05,
-      "loss": 0.2706,
+      "epoch": 2.6181208779327494,
+      "grad_norm": 0.21642592549324036,
+      "learning_rate": 2.431665821833859e-05,
+      "loss": 0.4371,
       "step": 72645
     },
     {
-      "epoch": 2.56,
-      "learning_rate": 2.5321042067526383e-05,
-      "loss": 0.2738,
+      "epoch": 2.618301077593974,
+      "grad_norm": 0.21715706586837769,
+      "learning_rate": 2.4313741179664974e-05,
+      "loss": 0.3757,
       "step": 72650
     },
     {
-      "epoch": 2.56,
-      "learning_rate": 2.5318193579034737e-05,
-      "loss": 0.27,
+      "epoch": 2.618481277255199,
+      "grad_norm": 0.21024832129478455,
+      "learning_rate": 2.4310824150341497e-05,
+      "loss": 0.3762,
       "step": 72655
     },
     {
-      "epoch": 2.56,
-      "learning_rate": 2.531534508641156e-05,
-      "loss": 0.2633,
+      "epoch": 2.618661476916423,
+      "grad_norm": 0.2558997869491577,
+      "learning_rate": 2.4307907130407916e-05,
+      "loss": 0.4358,
       "step": 72660
     },
     {
-      "epoch": 2.56,
-      "learning_rate": 2.531249658969384e-05,
-      "loss": 0.2696,
+      "epoch": 2.618841676577648,
+      "grad_norm": 0.1785580962896347,
+      "learning_rate": 2.4304990119903966e-05,
+      "loss": 0.4231,
       "step": 72665
     },
     {
-      "epoch": 2.56,
-      "learning_rate": 2.5309648088918558e-05,
-      "loss": 0.2552,
+      "epoch": 2.6190218762388726,
+      "grad_norm": 0.18506887555122375,
+      "learning_rate": 2.430207311886938e-05,
+      "loss": 0.3657,
       "step": 72670
     },
     {
-      "epoch": 2.56,
-      "learning_rate": 2.5306799584122693e-05,
-      "loss": 0.2602,
+      "epoch": 2.6192020759000973,
+      "grad_norm": 0.1763153374195099,
+      "learning_rate": 2.429915612734392e-05,
+      "loss": 0.3529,
       "step": 72675
     },
     {
-      "epoch": 2.56,
-      "learning_rate": 2.5303951075343245e-05,
-      "loss": 0.2569,
+      "epoch": 2.619382275561322,
+      "grad_norm": 0.21150760352611542,
+      "learning_rate": 2.4296239145367333e-05,
+      "loss": 0.3704,
       "step": 72680
     },
     {
-      "epoch": 2.56,
-      "learning_rate": 2.5301102562617186e-05,
-      "loss": 0.2762,
+      "epoch": 2.619562475222547,
+      "grad_norm": 0.2270565927028656,
+      "learning_rate": 2.429332217297933e-05,
+      "loss": 0.3814,
       "step": 72685
     },
     {
-      "epoch": 2.56,
-      "learning_rate": 2.529825404598151e-05,
-      "loss": 0.2722,
+      "epoch": 2.619742674883771,
+      "grad_norm": 0.20172244310379028,
+      "learning_rate": 2.4290405210219694e-05,
+      "loss": 0.3692,
       "step": 72690
     },
     {
-      "epoch": 2.56,
-      "learning_rate": 2.5295405525473205e-05,
-      "loss": 0.2748,
+      "epoch": 2.619922874544996,
+      "grad_norm": 0.20030339062213898,
+      "learning_rate": 2.428748825712813e-05,
+      "loss": 0.3969,
       "step": 72695
     },
     {
-      "epoch": 2.56,
-      "learning_rate": 2.5292557001129252e-05,
-      "loss": 0.2611,
+      "epoch": 2.6201030742062206,
+      "grad_norm": 0.23000763356685638,
+      "learning_rate": 2.4284571313744424e-05,
+      "loss": 0.4319,
       "step": 72700
     },
     {
-      "epoch": 2.56,
-      "learning_rate": 2.5289708472986638e-05,
-      "loss": 0.2969,
+      "epoch": 2.620283273867445,
+      "grad_norm": 0.18827667832374573,
+      "learning_rate": 2.428165438010828e-05,
+      "loss": 0.3748,
       "step": 72705
     },
     {
-      "epoch": 2.56,
-      "learning_rate": 2.5286859941082347e-05,
-      "loss": 0.2757,
+      "epoch": 2.6204634735286696,
+      "grad_norm": 0.23635447025299072,
+      "learning_rate": 2.427873745625946e-05,
+      "loss": 0.4177,
       "step": 72710
     },
     {
-      "epoch": 2.56,
-      "learning_rate": 2.528401140545337e-05,
-      "loss": 0.259,
+      "epoch": 2.6206436731898943,
+      "grad_norm": 0.18466027081012726,
+      "learning_rate": 2.4275820542237703e-05,
+      "loss": 0.3716,
       "step": 72715
     },
     {
-      "epoch": 2.56,
-      "learning_rate": 2.5281162866136705e-05,
-      "loss": 0.2553,
+      "epoch": 2.620823872851119,
+      "grad_norm": 0.21398019790649414,
+      "learning_rate": 2.427290363808275e-05,
+      "loss": 0.3842,
       "step": 72720
     },
     {
-      "epoch": 2.56,
-      "learning_rate": 2.527831432316931e-05,
-      "loss": 0.252,
+      "epoch": 2.621004072512344,
+      "grad_norm": 0.24221466481685638,
+      "learning_rate": 2.426998674383434e-05,
+      "loss": 0.4314,
       "step": 72725
     },
     {
-      "epoch": 2.56,
-      "learning_rate": 2.52754657765882e-05,
-      "loss": 0.2745,
+      "epoch": 2.6211842721735685,
+      "grad_norm": 0.18980202078819275,
+      "learning_rate": 2.4267069859532228e-05,
+      "loss": 0.3927,
       "step": 72730
     },
     {
-      "epoch": 2.56,
-      "learning_rate": 2.527261722643034e-05,
-      "loss": 0.2559,
+      "epoch": 2.621364471834793,
+      "grad_norm": 0.2422516644001007,
+      "learning_rate": 2.4264152985216136e-05,
+      "loss": 0.398,
       "step": 72735
     },
     {
-      "epoch": 2.56,
-      "learning_rate": 2.5269768672732726e-05,
-      "loss": 0.2649,
+      "epoch": 2.6215446714960176,
+      "grad_norm": 0.18269525468349457,
+      "learning_rate": 2.4261236120925828e-05,
+      "loss": 0.4142,
       "step": 72740
     },
     {
-      "epoch": 2.56,
-      "learning_rate": 2.5266920115532338e-05,
-      "loss": 0.262,
+      "epoch": 2.6217248711572423,
+      "grad_norm": 0.17914442718029022,
+      "learning_rate": 2.4258319266701032e-05,
+      "loss": 0.3988,
       "step": 72745
     },
     {
-      "epoch": 2.56,
-      "learning_rate": 2.5264071554866176e-05,
-      "loss": 0.2882,
+      "epoch": 2.621905070818467,
+      "grad_norm": 0.22300316393375397,
+      "learning_rate": 2.4255402422581485e-05,
+      "loss": 0.3862,
       "step": 72750
     },
     {
-      "epoch": 2.56,
-      "learning_rate": 2.5261222990771218e-05,
-      "loss": 0.2757,
+      "epoch": 2.6220852704796913,
+      "grad_norm": 0.21765939891338348,
+      "learning_rate": 2.4252485588606947e-05,
+      "loss": 0.3836,
       "step": 72755
     },
     {
-      "epoch": 2.56,
-      "learning_rate": 2.525837442328445e-05,
-      "loss": 0.2497,
+      "epoch": 2.622265470140916,
+      "grad_norm": 0.2101963460445404,
+      "learning_rate": 2.424956876481715e-05,
+      "loss": 0.4349,
       "step": 72760
     },
     {
-      "epoch": 2.56,
-      "learning_rate": 2.5255525852442863e-05,
-      "loss": 0.2631,
+      "epoch": 2.6224456698021408,
+      "grad_norm": 0.20395536720752716,
+      "learning_rate": 2.4246651951251815e-05,
+      "loss": 0.3813,
       "step": 72765
     },
     {
-      "epoch": 2.56,
-      "learning_rate": 2.5252677278283432e-05,
-      "loss": 0.2461,
+      "epoch": 2.6226258694633655,
+      "grad_norm": 0.208245649933815,
+      "learning_rate": 2.4243735147950715e-05,
+      "loss": 0.3975,
       "step": 72770
     },
     {
-      "epoch": 2.56,
-      "learning_rate": 2.5249828700843165e-05,
-      "loss": 0.2635,
+      "epoch": 2.6228060691245902,
+      "grad_norm": 0.20807571709156036,
+      "learning_rate": 2.424081835495357e-05,
+      "loss": 0.3674,
       "step": 72775
     },
     {
-      "epoch": 2.56,
-      "learning_rate": 2.5246980120159026e-05,
-      "loss": 0.2524,
+      "epoch": 2.6229862687858145,
+      "grad_norm": 0.2246702015399933,
+      "learning_rate": 2.423790157230014e-05,
+      "loss": 0.3766,
       "step": 72780
     },
     {
-      "epoch": 2.56,
-      "learning_rate": 2.5244131536268024e-05,
-      "loss": 0.2744,
+      "epoch": 2.6231664684470393,
+      "grad_norm": 0.16798308491706848,
+      "learning_rate": 2.423498480003015e-05,
+      "loss": 0.3983,
       "step": 72785
     },
     {
-      "epoch": 2.56,
-      "learning_rate": 2.524128294920713e-05,
-      "loss": 0.2592,
+      "epoch": 2.623346668108264,
+      "grad_norm": 0.18455500900745392,
+      "learning_rate": 2.423206803818333e-05,
+      "loss": 0.3787,
       "step": 72790
     },
     {
-      "epoch": 2.56,
-      "learning_rate": 2.523843435901333e-05,
-      "loss": 0.2661,
+      "epoch": 2.6235268677694887,
+      "grad_norm": 0.22922283411026,
+      "learning_rate": 2.422915128679945e-05,
+      "loss": 0.4127,
       "step": 72795
     },
     {
-      "epoch": 2.56,
-      "learning_rate": 2.5235585765723624e-05,
-      "loss": 0.2602,
+      "epoch": 2.623707067430713,
+      "grad_norm": 0.19777072966098785,
+      "learning_rate": 2.422623454591823e-05,
+      "loss": 0.379,
       "step": 72800
     },
     {
-      "epoch": 2.56,
-      "learning_rate": 2.5232737169374993e-05,
-      "loss": 0.2876,
+      "epoch": 2.6238872670919378,
+      "grad_norm": 0.2073335349559784,
+      "learning_rate": 2.4223317815579414e-05,
+      "loss": 0.3929,
       "step": 72805
     },
     {
-      "epoch": 2.56,
-      "learning_rate": 2.5229888570004416e-05,
-      "loss": 0.2709,
+      "epoch": 2.6240674667531625,
+      "grad_norm": 0.19318436086177826,
+      "learning_rate": 2.4220401095822742e-05,
+      "loss": 0.3986,
       "step": 72810
     },
     {
-      "epoch": 2.56,
-      "learning_rate": 2.52270399676489e-05,
-      "loss": 0.2644,
+      "epoch": 2.6242476664143872,
+      "grad_norm": 0.20948757231235504,
+      "learning_rate": 2.421748438668795e-05,
+      "loss": 0.432,
       "step": 72815
     },
     {
-      "epoch": 2.56,
-      "learning_rate": 2.522419136234541e-05,
-      "loss": 0.2693,
+      "epoch": 2.624427866075612,
+      "grad_norm": 0.24110877513885498,
+      "learning_rate": 2.4214567688214788e-05,
+      "loss": 0.4188,
       "step": 72820
     },
     {
-      "epoch": 2.56,
-      "learning_rate": 2.5221342754130943e-05,
-      "loss": 0.2959,
+      "epoch": 2.6246080657368362,
+      "grad_norm": 0.24411988258361816,
+      "learning_rate": 2.4211651000442988e-05,
+      "loss": 0.3898,
       "step": 72825
     },
     {
-      "epoch": 2.56,
-      "learning_rate": 2.521849414304249e-05,
-      "loss": 0.2704,
+      "epoch": 2.624788265398061,
+      "grad_norm": 0.21839167177677155,
+      "learning_rate": 2.4208734323412284e-05,
+      "loss": 0.4199,
       "step": 72830
     },
     {
-      "epoch": 2.56,
-      "learning_rate": 2.5215645529117033e-05,
-      "loss": 0.2734,
+      "epoch": 2.6249684650592857,
+      "grad_norm": 0.2261667400598526,
+      "learning_rate": 2.420581765716243e-05,
+      "loss": 0.4064,
       "step": 72835
     },
     {
-      "epoch": 2.56,
-      "learning_rate": 2.521279691239156e-05,
-      "loss": 0.2744,
+      "epoch": 2.6251486647205104,
+      "grad_norm": 0.20573599636554718,
+      "learning_rate": 2.420290100173315e-05,
+      "loss": 0.3813,
       "step": 72840
     },
     {
-      "epoch": 2.56,
-      "learning_rate": 2.5209948292903063e-05,
-      "loss": 0.254,
+      "epoch": 2.6253288643817347,
+      "grad_norm": 0.17270725965499878,
+      "learning_rate": 2.4199984357164197e-05,
+      "loss": 0.3894,
       "step": 72845
     },
     {
-      "epoch": 2.56,
-      "learning_rate": 2.5207099670688516e-05,
-      "loss": 0.2688,
+      "epoch": 2.6255090640429595,
+      "grad_norm": 0.2366410195827484,
+      "learning_rate": 2.4197067723495298e-05,
+      "loss": 0.3717,
       "step": 72850
     },
     {
-      "epoch": 2.56,
-      "learning_rate": 2.5204251045784927e-05,
-      "loss": 0.2712,
+      "epoch": 2.625689263704184,
+      "grad_norm": 0.2237681895494461,
+      "learning_rate": 2.4194151100766193e-05,
+      "loss": 0.3762,
       "step": 72855
     },
     {
-      "epoch": 2.56,
-      "learning_rate": 2.5201402418229265e-05,
-      "loss": 0.2834,
+      "epoch": 2.625869463365409,
+      "grad_norm": 0.20603452622890472,
+      "learning_rate": 2.419123448901663e-05,
+      "loss": 0.4403,
       "step": 72860
     },
     {
-      "epoch": 2.56,
-      "learning_rate": 2.5198553788058533e-05,
-      "loss": 0.2676,
+      "epoch": 2.6260496630266337,
+      "grad_norm": 0.21571598947048187,
+      "learning_rate": 2.4188317888286345e-05,
+      "loss": 0.3842,
       "step": 72865
     },
     {
-      "epoch": 2.56,
-      "learning_rate": 2.5195705155309705e-05,
-      "loss": 0.2897,
+      "epoch": 2.6262298626878584,
+      "grad_norm": 0.20403259992599487,
+      "learning_rate": 2.4185401298615052e-05,
+      "loss": 0.3822,
       "step": 72870
     },
     {
-      "epoch": 2.56,
-      "learning_rate": 2.5192856520019774e-05,
-      "loss": 0.2422,
+      "epoch": 2.6264100623490827,
+      "grad_norm": 0.20591306686401367,
+      "learning_rate": 2.418248472004253e-05,
+      "loss": 0.4068,
       "step": 72875
     },
     {
-      "epoch": 2.56,
-      "learning_rate": 2.519000788222573e-05,
-      "loss": 0.2494,
+      "epoch": 2.6265902620103074,
+      "grad_norm": 0.233648881316185,
+      "learning_rate": 2.4179568152608476e-05,
+      "loss": 0.3983,
       "step": 72880
     },
     {
-      "epoch": 2.56,
-      "learning_rate": 2.518715924196456e-05,
-      "loss": 0.2354,
+      "epoch": 2.626770461671532,
+      "grad_norm": 0.21279099583625793,
+      "learning_rate": 2.4176651596352657e-05,
+      "loss": 0.3876,
       "step": 72885
     },
     {
-      "epoch": 2.56,
-      "learning_rate": 2.518431059927325e-05,
-      "loss": 0.2622,
+      "epoch": 2.6269506613327565,
+      "grad_norm": 0.24593648314476013,
+      "learning_rate": 2.417373505131481e-05,
+      "loss": 0.4046,
       "step": 72890
     },
     {
-      "epoch": 2.56,
-      "learning_rate": 2.518146195418879e-05,
-      "loss": 0.2665,
+      "epoch": 2.627130860993981,
+      "grad_norm": 0.18338236212730408,
+      "learning_rate": 2.4170818517534642e-05,
+      "loss": 0.3814,
       "step": 72895
     },
     {
-      "epoch": 2.56,
-      "learning_rate": 2.5178613306748166e-05,
-      "loss": 0.2761,
+      "epoch": 2.627311060655206,
+      "grad_norm": 0.19615016877651215,
+      "learning_rate": 2.416790199505193e-05,
+      "loss": 0.3829,
       "step": 72900
     },
     {
-      "epoch": 2.57,
-      "learning_rate": 2.5175764656988365e-05,
-      "loss": 0.2379,
+      "epoch": 2.6274912603164307,
+      "grad_norm": 0.17791742086410522,
+      "learning_rate": 2.4164985483906384e-05,
+      "loss": 0.3914,
       "step": 72905
     },
     {
-      "epoch": 2.57,
-      "learning_rate": 2.5172916004946372e-05,
-      "loss": 0.2582,
+      "epoch": 2.6276714599776554,
+      "grad_norm": 0.22163866460323334,
+      "learning_rate": 2.416206898413775e-05,
+      "loss": 0.4144,
       "step": 72910
     },
     {
-      "epoch": 2.57,
-      "learning_rate": 2.5170067350659183e-05,
-      "loss": 0.2882,
+      "epoch": 2.62785165963888,
+      "grad_norm": 0.18475517630577087,
+      "learning_rate": 2.4159152495785765e-05,
+      "loss": 0.3815,
       "step": 72915
     },
     {
-      "epoch": 2.57,
-      "learning_rate": 2.5167218694163786e-05,
-      "loss": 0.28,
+      "epoch": 2.6280318593001044,
+      "grad_norm": 0.22653162479400635,
+      "learning_rate": 2.415623601889016e-05,
+      "loss": 0.403,
       "step": 72920
     },
     {
-      "epoch": 2.57,
-      "learning_rate": 2.5164370035497152e-05,
-      "loss": 0.2516,
+      "epoch": 2.628212058961329,
+      "grad_norm": 0.1823846399784088,
+      "learning_rate": 2.4153319553490677e-05,
+      "loss": 0.3878,
       "step": 72925
     },
     {
-      "epoch": 2.57,
-      "learning_rate": 2.516152137469629e-05,
-      "loss": 0.2771,
+      "epoch": 2.628392258622554,
+      "grad_norm": 0.21723927557468414,
+      "learning_rate": 2.4150403099627056e-05,
+      "loss": 0.4188,
       "step": 72930
     },
     {
-      "epoch": 2.57,
-      "learning_rate": 2.5158672711798174e-05,
-      "loss": 0.2699,
+      "epoch": 2.628572458283778,
+      "grad_norm": 0.2315875142812729,
+      "learning_rate": 2.414748665733902e-05,
+      "loss": 0.3892,
       "step": 72935
     },
     {
-      "epoch": 2.57,
-      "learning_rate": 2.5155824046839803e-05,
-      "loss": 0.2599,
+      "epoch": 2.628752657945003,
+      "grad_norm": 0.24299949407577515,
+      "learning_rate": 2.4144570226666325e-05,
+      "loss": 0.4066,
       "step": 72940
     },
     {
-      "epoch": 2.57,
-      "learning_rate": 2.5152975379858155e-05,
-      "loss": 0.2754,
+      "epoch": 2.6289328576062276,
+      "grad_norm": 0.19684675335884094,
+      "learning_rate": 2.414165380764869e-05,
+      "loss": 0.4139,
       "step": 72945
     },
     {
-      "epoch": 2.57,
-      "learning_rate": 2.5150126710890232e-05,
-      "loss": 0.2775,
+      "epoch": 2.6291130572674524,
+      "grad_norm": 0.232985258102417,
+      "learning_rate": 2.413873740032585e-05,
+      "loss": 0.3786,
       "step": 72950
     },
     {
-      "epoch": 2.57,
-      "learning_rate": 2.5147278039973005e-05,
-      "loss": 0.2599,
+      "epoch": 2.629293256928677,
+      "grad_norm": 0.20983463525772095,
+      "learning_rate": 2.413582100473755e-05,
+      "loss": 0.3665,
       "step": 72955
     },
     {
-      "epoch": 2.57,
-      "learning_rate": 2.514442936714347e-05,
-      "loss": 0.2474,
+      "epoch": 2.629473456589902,
+      "grad_norm": 0.1940142810344696,
+      "learning_rate": 2.4132904620923518e-05,
+      "loss": 0.3979,
       "step": 72960
     },
     {
-      "epoch": 2.57,
-      "learning_rate": 2.514158069243861e-05,
-      "loss": 0.2508,
+      "epoch": 2.629653656251126,
+      "grad_norm": 0.20459094643592834,
+      "learning_rate": 2.41299882489235e-05,
+      "loss": 0.3806,
       "step": 72965
     },
     {
-      "epoch": 2.57,
-      "learning_rate": 2.513873201589543e-05,
-      "loss": 0.2771,
+      "epoch": 2.629833855912351,
+      "grad_norm": 0.22843003273010254,
+      "learning_rate": 2.4127071888777227e-05,
+      "loss": 0.3792,
       "step": 72970
     },
     {
-      "epoch": 2.57,
-      "learning_rate": 2.5135883337550898e-05,
-      "loss": 0.2758,
+      "epoch": 2.6300140555735756,
+      "grad_norm": 0.25140973925590515,
+      "learning_rate": 2.4124155540524414e-05,
+      "loss": 0.4148,
       "step": 72975
     },
     {
-      "epoch": 2.57,
-      "learning_rate": 2.513303465744201e-05,
-      "loss": 0.2695,
+      "epoch": 2.6301942552348,
+      "grad_norm": 0.2235340029001236,
+      "learning_rate": 2.4121239204204833e-05,
+      "loss": 0.3699,
       "step": 72980
     },
     {
-      "epoch": 2.57,
-      "learning_rate": 2.5130185975605756e-05,
-      "loss": 0.2611,
+      "epoch": 2.6303744548960246,
+      "grad_norm": 0.2354481816291809,
+      "learning_rate": 2.4118322879858187e-05,
+      "loss": 0.4014,
       "step": 72985
     },
     {
-      "epoch": 2.57,
-      "learning_rate": 2.5127337292079122e-05,
-      "loss": 0.2607,
+      "epoch": 2.6305546545572493,
+      "grad_norm": 0.17458322644233704,
+      "learning_rate": 2.4115406567524217e-05,
+      "loss": 0.3573,
       "step": 72990
     },
     {
-      "epoch": 2.57,
-      "learning_rate": 2.51244886068991e-05,
-      "loss": 0.2707,
+      "epoch": 2.630734854218474,
+      "grad_norm": 0.1971106380224228,
+      "learning_rate": 2.4112490267242665e-05,
+      "loss": 0.3703,
       "step": 72995
     },
     {
-      "epoch": 2.57,
-      "learning_rate": 2.5121639920102675e-05,
-      "loss": 0.2887,
+      "epoch": 2.630915053879699,
+      "grad_norm": 0.2097444236278534,
+      "learning_rate": 2.410957397905326e-05,
+      "loss": 0.4161,
       "step": 73000
     },
     {
-      "epoch": 2.57,
-      "eval_loss": 0.2631213068962097,
-      "eval_runtime": 10.5493,
-      "eval_samples_per_second": 9.479,
-      "eval_steps_per_second": 9.479,
+      "epoch": 2.630915053879699,
+      "eval_loss": 0.43243083357810974,
+      "eval_runtime": 3.5273,
+      "eval_samples_per_second": 28.35,
+      "eval_steps_per_second": 7.087,
       "step": 73000
     },
     {
-      "epoch": 2.57,
-      "learning_rate": 2.5118791231726835e-05,
-      "loss": 0.2693,
+      "epoch": 2.6310952535409236,
+      "grad_norm": 0.2296389639377594,
+      "learning_rate": 2.410665770299574e-05,
+      "loss": 0.386,
       "step": 73005
     },
     {
-      "epoch": 2.57,
-      "learning_rate": 2.5115942541808568e-05,
-      "loss": 0.2585,
+      "epoch": 2.631275453202148,
+      "grad_norm": 0.2250434011220932,
+      "learning_rate": 2.4103741439109835e-05,
+      "loss": 0.3863,
       "step": 73010
     },
     {
-      "epoch": 2.57,
-      "learning_rate": 2.5113093850384863e-05,
-      "loss": 0.2537,
+      "epoch": 2.6314556528633726,
+      "grad_norm": 0.20868182182312012,
+      "learning_rate": 2.410082518743528e-05,
+      "loss": 0.366,
       "step": 73015
     },
     {
-      "epoch": 2.57,
-      "learning_rate": 2.5110245157492707e-05,
-      "loss": 0.2874,
+      "epoch": 2.6316358525245973,
+      "grad_norm": 0.2829134166240692,
+      "learning_rate": 2.4097908948011804e-05,
+      "loss": 0.3913,
       "step": 73020
     },
     {
-      "epoch": 2.57,
-      "learning_rate": 2.5107396463169097e-05,
-      "loss": 0.2721,
+      "epoch": 2.631816052185822,
+      "grad_norm": 0.21935230493545532,
+      "learning_rate": 2.4094992720879144e-05,
+      "loss": 0.3815,
       "step": 73025
     },
     {
-      "epoch": 2.57,
-      "learning_rate": 2.510454776745101e-05,
-      "loss": 0.2768,
+      "epoch": 2.6319962518470463,
+      "grad_norm": 0.1936255842447281,
+      "learning_rate": 2.4092076506077036e-05,
+      "loss": 0.3789,
       "step": 73030
     },
     {
-      "epoch": 2.57,
-      "learning_rate": 2.5101699070375446e-05,
-      "loss": 0.2624,
+      "epoch": 2.632176451508271,
+      "grad_norm": 0.18805243074893951,
+      "learning_rate": 2.408916030364521e-05,
+      "loss": 0.3998,
       "step": 73035
     },
     {
-      "epoch": 2.57,
-      "learning_rate": 2.5098850371979375e-05,
-      "loss": 0.2713,
+      "epoch": 2.632356651169496,
+      "grad_norm": 0.19446292519569397,
+      "learning_rate": 2.4086244113623395e-05,
+      "loss": 0.4257,
       "step": 73040
     },
     {
-      "epoch": 2.57,
-      "learning_rate": 2.5096001672299806e-05,
-      "loss": 0.2588,
+      "epoch": 2.6325368508307205,
+      "grad_norm": 0.2560482621192932,
+      "learning_rate": 2.408332793605133e-05,
+      "loss": 0.387,
       "step": 73045
     },
     {
-      "epoch": 2.57,
-      "learning_rate": 2.5093152971373712e-05,
-      "loss": 0.2834,
+      "epoch": 2.6327170504919453,
+      "grad_norm": 0.1912938952445984,
+      "learning_rate": 2.4080411770968746e-05,
+      "loss": 0.4343,
       "step": 73050
     },
     {
-      "epoch": 2.57,
-      "learning_rate": 2.5090304269238096e-05,
-      "loss": 0.25,
+      "epoch": 2.6328972501531696,
+      "grad_norm": 0.2362576127052307,
+      "learning_rate": 2.4077495618415367e-05,
+      "loss": 0.3869,
       "step": 73055
     },
     {
-      "epoch": 2.57,
-      "learning_rate": 2.5087455565929936e-05,
-      "loss": 0.2558,
+      "epoch": 2.6330774498143943,
+      "grad_norm": 0.19982530176639557,
+      "learning_rate": 2.4074579478430942e-05,
+      "loss": 0.3916,
       "step": 73060
     },
     {
-      "epoch": 2.57,
-      "learning_rate": 2.508460686148622e-05,
-      "loss": 0.2671,
+      "epoch": 2.633257649475619,
+      "grad_norm": 0.2339782565832138,
+      "learning_rate": 2.407166335105518e-05,
+      "loss": 0.3988,
       "step": 73065
     },
     {
-      "epoch": 2.57,
-      "learning_rate": 2.5081758155943935e-05,
-      "loss": 0.2846,
+      "epoch": 2.6334378491368438,
+      "grad_norm": 0.2075488567352295,
+      "learning_rate": 2.4068747236327838e-05,
+      "loss": 0.4047,
       "step": 73070
     },
     {
-      "epoch": 2.57,
-      "learning_rate": 2.507890944934009e-05,
-      "loss": 0.2752,
+      "epoch": 2.633618048798068,
+      "grad_norm": 0.20350724458694458,
+      "learning_rate": 2.4065831134288635e-05,
+      "loss": 0.3793,
       "step": 73075
     },
     {
-      "epoch": 2.57,
-      "learning_rate": 2.5076060741711643e-05,
-      "loss": 0.2523,
+      "epoch": 2.633798248459293,
+      "grad_norm": 0.20746393501758575,
+      "learning_rate": 2.4062915044977284e-05,
+      "loss": 0.401,
       "step": 73080
     },
     {
-      "epoch": 2.57,
-      "learning_rate": 2.5073212033095605e-05,
-      "loss": 0.2534,
+      "epoch": 2.6339784481205175,
+      "grad_norm": 0.21754801273345947,
+      "learning_rate": 2.4059998968433553e-05,
+      "loss": 0.3765,
       "step": 73085
     },
     {
-      "epoch": 2.57,
-      "learning_rate": 2.5070363323528956e-05,
-      "loss": 0.2498,
+      "epoch": 2.6341586477817422,
+      "grad_norm": 0.16761621832847595,
+      "learning_rate": 2.4057082904697152e-05,
+      "loss": 0.393,
       "step": 73090
     },
     {
-      "epoch": 2.57,
-      "learning_rate": 2.5067514613048683e-05,
-      "loss": 0.271,
+      "epoch": 2.634338847442967,
+      "grad_norm": 0.23849821090698242,
+      "learning_rate": 2.4054166853807803e-05,
+      "loss": 0.4177,
       "step": 73095
     },
     {
-      "epoch": 2.57,
-      "learning_rate": 2.5064665901691775e-05,
-      "loss": 0.2751,
+      "epoch": 2.6345190471041917,
+      "grad_norm": 0.2702219784259796,
+      "learning_rate": 2.4051250815805253e-05,
+      "loss": 0.4136,
       "step": 73100
     },
     {
-      "epoch": 2.57,
-      "learning_rate": 2.506181718949523e-05,
-      "loss": 0.2606,
+      "epoch": 2.634699246765416,
+      "grad_norm": 0.22448520362377167,
+      "learning_rate": 2.4048334790729225e-05,
+      "loss": 0.423,
       "step": 73105
     },
     {
-      "epoch": 2.57,
-      "learning_rate": 2.5058968476496024e-05,
-      "loss": 0.2337,
+      "epoch": 2.6348794464266407,
+      "grad_norm": 0.24827606976032257,
+      "learning_rate": 2.4045418778619456e-05,
+      "loss": 0.4137,
       "step": 73110
     },
     {
-      "epoch": 2.57,
-      "learning_rate": 2.505611976273115e-05,
-      "loss": 0.2835,
+      "epoch": 2.6350596460878655,
+      "grad_norm": 0.2294238954782486,
+      "learning_rate": 2.4042502779515668e-05,
+      "loss": 0.3948,
       "step": 73115
     },
     {
-      "epoch": 2.57,
-      "learning_rate": 2.50532710482376e-05,
-      "loss": 0.2665,
+      "epoch": 2.6352398457490898,
+      "grad_norm": 0.2006445825099945,
+      "learning_rate": 2.4039586793457593e-05,
+      "loss": 0.3847,
       "step": 73120
     },
     {
-      "epoch": 2.57,
-      "learning_rate": 2.5050422333052366e-05,
-      "loss": 0.2758,
+      "epoch": 2.6354200454103145,
+      "grad_norm": 0.2071392983198166,
+      "learning_rate": 2.4036670820484964e-05,
+      "loss": 0.3991,
       "step": 73125
     },
     {
-      "epoch": 2.57,
-      "learning_rate": 2.504757361721242e-05,
-      "loss": 0.2706,
+      "epoch": 2.6356002450715392,
+      "grad_norm": 0.23408040404319763,
+      "learning_rate": 2.403375486063751e-05,
+      "loss": 0.4106,
       "step": 73130
     },
     {
-      "epoch": 2.57,
-      "learning_rate": 2.5044724900754775e-05,
-      "loss": 0.2624,
+      "epoch": 2.635780444732764,
+      "grad_norm": 0.2418873906135559,
+      "learning_rate": 2.4030838913954955e-05,
+      "loss": 0.4135,
       "step": 73135
     },
     {
-      "epoch": 2.57,
-      "learning_rate": 2.50418761837164e-05,
-      "loss": 0.2817,
+      "epoch": 2.6359606443939887,
+      "grad_norm": 0.18823467195034027,
+      "learning_rate": 2.4027922980477036e-05,
+      "loss": 0.3632,
       "step": 73140
     },
     {
-      "epoch": 2.57,
-      "learning_rate": 2.5039027466134292e-05,
-      "loss": 0.2912,
+      "epoch": 2.6361408440552134,
+      "grad_norm": 0.22256679832935333,
+      "learning_rate": 2.402500706024347e-05,
+      "loss": 0.3916,
       "step": 73145
     },
     {
-      "epoch": 2.57,
-      "learning_rate": 2.503617874804543e-05,
-      "loss": 0.2722,
+      "epoch": 2.6363210437164377,
+      "grad_norm": 0.2467431128025055,
+      "learning_rate": 2.4022091153294004e-05,
+      "loss": 0.4101,
       "step": 73150
     },
     {
-      "epoch": 2.57,
-      "learning_rate": 2.5033330029486818e-05,
-      "loss": 0.3058,
+      "epoch": 2.6365012433776625,
+      "grad_norm": 0.20016492903232574,
+      "learning_rate": 2.4019175259668362e-05,
+      "loss": 0.3946,
       "step": 73155
     },
     {
-      "epoch": 2.57,
-      "learning_rate": 2.503048131049544e-05,
-      "loss": 0.2577,
+      "epoch": 2.636681443038887,
+      "grad_norm": 0.2600384056568146,
+      "learning_rate": 2.4016259379406247e-05,
+      "loss": 0.4024,
       "step": 73160
     },
     {
-      "epoch": 2.57,
-      "learning_rate": 2.502763259110828e-05,
-      "loss": 0.2728,
+      "epoch": 2.6368616427001115,
+      "grad_norm": 0.2585957646369934,
+      "learning_rate": 2.401334351254743e-05,
+      "loss": 0.3854,
       "step": 73165
     },
     {
-      "epoch": 2.57,
-      "learning_rate": 2.5024783871362327e-05,
-      "loss": 0.2937,
+      "epoch": 2.637041842361336,
+      "grad_norm": 0.1720488965511322,
+      "learning_rate": 2.4010427659131604e-05,
+      "loss": 0.3705,
       "step": 73170
     },
     {
-      "epoch": 2.57,
-      "learning_rate": 2.5021935151294573e-05,
-      "loss": 0.2686,
+      "epoch": 2.637222042022561,
+      "grad_norm": 0.23503340780735016,
+      "learning_rate": 2.4007511819198503e-05,
+      "loss": 0.4219,
       "step": 73175
     },
     {
-      "epoch": 2.57,
-      "learning_rate": 2.5019086430942006e-05,
-      "loss": 0.2744,
+      "epoch": 2.6374022416837857,
+      "grad_norm": 0.21345233917236328,
+      "learning_rate": 2.4004595992787877e-05,
+      "loss": 0.4177,
       "step": 73180
     },
     {
-      "epoch": 2.57,
-      "learning_rate": 2.501623771034161e-05,
-      "loss": 0.2831,
+      "epoch": 2.6375824413450104,
+      "grad_norm": 0.2077561467885971,
+      "learning_rate": 2.400168017993942e-05,
+      "loss": 0.4072,
       "step": 73185
     },
     {
-      "epoch": 2.58,
-      "learning_rate": 2.5013388989530388e-05,
-      "loss": 0.2692,
+      "epoch": 2.637762641006235,
+      "grad_norm": 0.2376868724822998,
+      "learning_rate": 2.3998764380692896e-05,
+      "loss": 0.4042,
       "step": 73190
     },
     {
-      "epoch": 2.58,
-      "learning_rate": 2.5010540268545314e-05,
-      "loss": 0.2712,
+      "epoch": 2.6379428406674594,
+      "grad_norm": 0.24904000759124756,
+      "learning_rate": 2.3995848595088008e-05,
+      "loss": 0.423,
       "step": 73195
     },
     {
-      "epoch": 2.58,
-      "learning_rate": 2.5007691547423377e-05,
-      "loss": 0.2667,
+      "epoch": 2.638123040328684,
+      "grad_norm": 0.20672129094600677,
+      "learning_rate": 2.3992932823164483e-05,
+      "loss": 0.397,
       "step": 73200
     },
     {
-      "epoch": 2.58,
-      "learning_rate": 2.5004842826201574e-05,
-      "loss": 0.2671,
+      "epoch": 2.638303239989909,
+      "grad_norm": 0.17566558718681335,
+      "learning_rate": 2.3990017064962056e-05,
+      "loss": 0.4138,
       "step": 73205
     },
     {
-      "epoch": 2.58,
-      "learning_rate": 2.5001994104916888e-05,
-      "loss": 0.2742,
+      "epoch": 2.638483439651133,
+      "grad_norm": 0.19264771044254303,
+      "learning_rate": 2.398710132052045e-05,
+      "loss": 0.3775,
       "step": 73210
     },
     {
-      "epoch": 2.58,
-      "learning_rate": 2.4999145383606313e-05,
-      "loss": 0.2682,
+      "epoch": 2.638663639312358,
+      "grad_norm": 0.17429447174072266,
+      "learning_rate": 2.3984185589879395e-05,
+      "loss": 0.3963,
       "step": 73215
     },
     {
-      "epoch": 2.58,
-      "learning_rate": 2.4996296662306834e-05,
-      "loss": 0.2402,
+      "epoch": 2.6388438389735827,
+      "grad_norm": 0.19570399820804596,
+      "learning_rate": 2.3981269873078613e-05,
+      "loss": 0.4127,
       "step": 73220
     },
     {
-      "epoch": 2.58,
-      "learning_rate": 2.4993447941055446e-05,
-      "loss": 0.2721,
+      "epoch": 2.6390240386348074,
+      "grad_norm": 0.23415052890777588,
+      "learning_rate": 2.3978354170157828e-05,
+      "loss": 0.3912,
       "step": 73225
     },
     {
-      "epoch": 2.58,
-      "learning_rate": 2.4990599219889114e-05,
-      "loss": 0.2458,
+      "epoch": 2.639204238296032,
+      "grad_norm": 0.22329415380954742,
+      "learning_rate": 2.3975438481156772e-05,
+      "loss": 0.4042,
       "step": 73230
     },
     {
-      "epoch": 2.58,
-      "learning_rate": 2.498775049884486e-05,
-      "loss": 0.2727,
+      "epoch": 2.639384437957257,
+      "grad_norm": 0.17532110214233398,
+      "learning_rate": 2.3972522806115176e-05,
+      "loss": 0.4019,
       "step": 73235
     },
     {
-      "epoch": 2.58,
-      "learning_rate": 2.4984901777959655e-05,
-      "loss": 0.2571,
+      "epoch": 2.639564637618481,
+      "grad_norm": 0.2220272421836853,
+      "learning_rate": 2.3969607145072747e-05,
+      "loss": 0.4198,
       "step": 73240
     },
     {
-      "epoch": 2.58,
-      "learning_rate": 2.4982053057270483e-05,
-      "loss": 0.2896,
+      "epoch": 2.639744837279706,
+      "grad_norm": 0.2223370224237442,
+      "learning_rate": 2.3966691498069228e-05,
+      "loss": 0.4365,
       "step": 73245
     },
     {
-      "epoch": 2.58,
-      "learning_rate": 2.497920433681435e-05,
-      "loss": 0.276,
+      "epoch": 2.6399250369409306,
+      "grad_norm": 0.2924058437347412,
+      "learning_rate": 2.396377586514433e-05,
+      "loss": 0.3981,
       "step": 73250
     },
     {
-      "epoch": 2.58,
-      "learning_rate": 2.4976355616628236e-05,
-      "loss": 0.266,
+      "epoch": 2.6401052366021553,
+      "grad_norm": 0.22655771672725677,
+      "learning_rate": 2.396086024633779e-05,
+      "loss": 0.4427,
       "step": 73255
     },
     {
-      "epoch": 2.58,
-      "learning_rate": 2.4973506896749123e-05,
-      "loss": 0.291,
+      "epoch": 2.6402854362633796,
+      "grad_norm": 0.17349043488502502,
+      "learning_rate": 2.3957944641689335e-05,
+      "loss": 0.3575,
       "step": 73260
     },
     {
-      "epoch": 2.58,
-      "learning_rate": 2.4970658177214007e-05,
-      "loss": 0.2903,
+      "epoch": 2.6404656359246044,
+      "grad_norm": 0.18977303802967072,
+      "learning_rate": 2.3955029051238666e-05,
+      "loss": 0.3798,
       "step": 73265
     },
     {
-      "epoch": 2.58,
-      "learning_rate": 2.4967809458059873e-05,
-      "loss": 0.2851,
+      "epoch": 2.640645835585829,
+      "grad_norm": 0.2203160524368286,
+      "learning_rate": 2.3952113475025543e-05,
+      "loss": 0.3796,
       "step": 73270
     },
     {
-      "epoch": 2.58,
-      "learning_rate": 2.496496073932372e-05,
-      "loss": 0.2526,
+      "epoch": 2.640826035247054,
+      "grad_norm": 0.16892023384571075,
+      "learning_rate": 2.394919791308966e-05,
+      "loss": 0.4015,
       "step": 73275
     },
     {
-      "epoch": 2.58,
-      "learning_rate": 2.4962112021042523e-05,
-      "loss": 0.2813,
+      "epoch": 2.6410062349082786,
+      "grad_norm": 0.22856315970420837,
+      "learning_rate": 2.3946282365470755e-05,
+      "loss": 0.4011,
       "step": 73280
     },
     {
-      "epoch": 2.58,
-      "learning_rate": 2.495926330325327e-05,
-      "loss": 0.2818,
+      "epoch": 2.641186434569503,
+      "grad_norm": 0.21979033946990967,
+      "learning_rate": 2.3943366832208548e-05,
+      "loss": 0.4229,
       "step": 73285
     },
     {
-      "epoch": 2.58,
-      "learning_rate": 2.4956414585992963e-05,
-      "loss": 0.2748,
+      "epoch": 2.6413666342307276,
+      "grad_norm": 0.180423304438591,
+      "learning_rate": 2.3940451313342757e-05,
+      "loss": 0.3786,
       "step": 73290
     },
     {
-      "epoch": 2.58,
-      "learning_rate": 2.4953565869298586e-05,
-      "loss": 0.262,
+      "epoch": 2.6415468338919523,
+      "grad_norm": 0.2113245725631714,
+      "learning_rate": 2.393753580891312e-05,
+      "loss": 0.3818,
       "step": 73295
     },
     {
-      "epoch": 2.58,
-      "learning_rate": 2.495071715320713e-05,
-      "loss": 0.2837,
+      "epoch": 2.641727033553177,
+      "grad_norm": 0.21729253232479095,
+      "learning_rate": 2.393462031895935e-05,
+      "loss": 0.4212,
       "step": 73300
     },
     {
-      "epoch": 2.58,
-      "learning_rate": 2.4947868437755563e-05,
-      "loss": 0.2752,
+      "epoch": 2.6419072332144014,
+      "grad_norm": 0.20087195932865143,
+      "learning_rate": 2.393170484352117e-05,
+      "loss": 0.4411,
       "step": 73305
     },
     {
-      "epoch": 2.58,
-      "learning_rate": 2.4945019722980904e-05,
-      "loss": 0.2715,
+      "epoch": 2.642087432875626,
+      "grad_norm": 0.21028025448322296,
+      "learning_rate": 2.3928789382638305e-05,
+      "loss": 0.424,
       "step": 73310
     },
     {
-      "epoch": 2.58,
-      "learning_rate": 2.494217100892012e-05,
-      "loss": 0.273,
+      "epoch": 2.642267632536851,
+      "grad_norm": 0.18576066195964813,
+      "learning_rate": 2.392587393635048e-05,
+      "loss": 0.3946,
       "step": 73315
     },
     {
-      "epoch": 2.58,
-      "learning_rate": 2.4939322295610216e-05,
-      "loss": 0.2473,
+      "epoch": 2.6424478321980756,
+      "grad_norm": 0.18550072610378265,
+      "learning_rate": 2.392295850469741e-05,
+      "loss": 0.4047,
       "step": 73320
     },
     {
-      "epoch": 2.58,
-      "learning_rate": 2.493647358308816e-05,
-      "loss": 0.2795,
+      "epoch": 2.6426280318593003,
+      "grad_norm": 0.17761926352977753,
+      "learning_rate": 2.3920043087718826e-05,
+      "loss": 0.3817,
       "step": 73325
     },
     {
-      "epoch": 2.58,
-      "learning_rate": 2.4933624871390962e-05,
-      "loss": 0.2874,
+      "epoch": 2.6428082315205246,
+      "grad_norm": 0.20507977902889252,
+      "learning_rate": 2.3917127685454442e-05,
+      "loss": 0.409,
       "step": 73330
     },
     {
-      "epoch": 2.58,
-      "learning_rate": 2.4930776160555603e-05,
-      "loss": 0.2605,
+      "epoch": 2.6429884311817493,
+      "grad_norm": 0.24363189935684204,
+      "learning_rate": 2.391421229794399e-05,
+      "loss": 0.4085,
       "step": 73335
     },
     {
-      "epoch": 2.58,
-      "learning_rate": 2.492792745061906e-05,
-      "loss": 0.2756,
+      "epoch": 2.643168630842974,
+      "grad_norm": 0.17502768337726593,
+      "learning_rate": 2.3911296925227182e-05,
+      "loss": 0.4124,
       "step": 73340
     },
     {
-      "epoch": 2.58,
-      "learning_rate": 2.4925078741618338e-05,
-      "loss": 0.2642,
+      "epoch": 2.6433488305041988,
+      "grad_norm": 0.21647201478481293,
+      "learning_rate": 2.3908381567343736e-05,
+      "loss": 0.3596,
       "step": 73345
     },
     {
-      "epoch": 2.58,
-      "learning_rate": 2.4922230033590415e-05,
-      "loss": 0.2525,
+      "epoch": 2.643529030165423,
+      "grad_norm": 0.24254228174686432,
+      "learning_rate": 2.390546622433339e-05,
+      "loss": 0.4148,
       "step": 73350
     },
     {
-      "epoch": 2.58,
-      "learning_rate": 2.491938132657229e-05,
-      "loss": 0.2883,
+      "epoch": 2.643709229826648,
+      "grad_norm": 0.22495651245117188,
+      "learning_rate": 2.3902550896235855e-05,
+      "loss": 0.3867,
       "step": 73355
     },
     {
-      "epoch": 2.58,
-      "learning_rate": 2.4916532620600937e-05,
-      "loss": 0.2925,
+      "epoch": 2.6438894294878725,
+      "grad_norm": 0.23512765765190125,
+      "learning_rate": 2.3899635583090837e-05,
+      "loss": 0.4377,
       "step": 73360
     },
     {
-      "epoch": 2.58,
-      "learning_rate": 2.4913683915713362e-05,
-      "loss": 0.2917,
+      "epoch": 2.6440696291490973,
+      "grad_norm": 0.18619103729724884,
+      "learning_rate": 2.389672028493809e-05,
+      "loss": 0.4124,
       "step": 73365
     },
     {
-      "epoch": 2.58,
-      "learning_rate": 2.491083521194654e-05,
-      "loss": 0.2734,
+      "epoch": 2.644249828810322,
+      "grad_norm": 0.2616622745990753,
+      "learning_rate": 2.3893805001817298e-05,
+      "loss": 0.3899,
       "step": 73370
     },
     {
-      "epoch": 2.58,
-      "learning_rate": 2.4907986509337467e-05,
-      "loss": 0.2758,
+      "epoch": 2.6444300284715467,
+      "grad_norm": 0.23114174604415894,
+      "learning_rate": 2.3890889733768215e-05,
+      "loss": 0.4002,
       "step": 73375
     },
     {
-      "epoch": 2.58,
-      "learning_rate": 2.4905137807923116e-05,
-      "loss": 0.2821,
+      "epoch": 2.644610228132771,
+      "grad_norm": 0.19029228389263153,
+      "learning_rate": 2.388797448083054e-05,
+      "loss": 0.3984,
       "step": 73380
     },
     {
-      "epoch": 2.58,
-      "learning_rate": 2.4902289107740508e-05,
-      "loss": 0.2439,
+      "epoch": 2.6447904277939958,
+      "grad_norm": 0.2512267529964447,
+      "learning_rate": 2.388505924304399e-05,
+      "loss": 0.4141,
       "step": 73385
     },
     {
-      "epoch": 2.58,
-      "learning_rate": 2.4899440408826602e-05,
-      "loss": 0.2789,
+      "epoch": 2.6449706274552205,
+      "grad_norm": 0.21497133374214172,
+      "learning_rate": 2.3882144020448297e-05,
+      "loss": 0.4142,
       "step": 73390
     },
     {
-      "epoch": 2.58,
-      "learning_rate": 2.4896591711218394e-05,
-      "loss": 0.2544,
+      "epoch": 2.645150827116445,
+      "grad_norm": 0.20400524139404297,
+      "learning_rate": 2.3879228813083175e-05,
+      "loss": 0.4061,
       "step": 73395
     },
     {
-      "epoch": 2.58,
-      "learning_rate": 2.4893743014952875e-05,
-      "loss": 0.2764,
+      "epoch": 2.6453310267776695,
+      "grad_norm": 0.1923159807920456,
+      "learning_rate": 2.387631362098834e-05,
+      "loss": 0.4065,
       "step": 73400
     },
     {
-      "epoch": 2.58,
-      "learning_rate": 2.489089432006703e-05,
-      "loss": 0.2583,
+      "epoch": 2.6455112264388942,
+      "grad_norm": 0.2456676959991455,
+      "learning_rate": 2.387339844420352e-05,
+      "loss": 0.3786,
       "step": 73405
     },
     {
-      "epoch": 2.58,
-      "learning_rate": 2.488804562659786e-05,
-      "loss": 0.2913,
+      "epoch": 2.645691426100119,
+      "grad_norm": 0.20096588134765625,
+      "learning_rate": 2.3870483282768422e-05,
+      "loss": 0.4352,
       "step": 73410
     },
     {
-      "epoch": 2.58,
-      "learning_rate": 2.4885196934582343e-05,
-      "loss": 0.2417,
+      "epoch": 2.6458716257613437,
+      "grad_norm": 0.19349773228168488,
+      "learning_rate": 2.3867568136722777e-05,
+      "loss": 0.4179,
       "step": 73415
     },
     {
-      "epoch": 2.58,
-      "learning_rate": 2.488234824405746e-05,
-      "loss": 0.28,
+      "epoch": 2.6460518254225684,
+      "grad_norm": 0.24120834469795227,
+      "learning_rate": 2.3864653006106298e-05,
+      "loss": 0.3935,
       "step": 73420
     },
     {
-      "epoch": 2.58,
-      "learning_rate": 2.487949955506021e-05,
-      "loss": 0.2668,
+      "epoch": 2.6462320250837927,
+      "grad_norm": 0.1798119693994522,
+      "learning_rate": 2.386173789095869e-05,
+      "loss": 0.3727,
       "step": 73425
     },
     {
-      "epoch": 2.58,
-      "learning_rate": 2.487665086762758e-05,
-      "loss": 0.2542,
+      "epoch": 2.6464122247450175,
+      "grad_norm": 0.2536999583244324,
+      "learning_rate": 2.3858822791319693e-05,
+      "loss": 0.3995,
       "step": 73430
     },
     {
-      "epoch": 2.58,
-      "learning_rate": 2.487380218179656e-05,
-      "loss": 0.2687,
+      "epoch": 2.646592424406242,
+      "grad_norm": 0.24796532094478607,
+      "learning_rate": 2.385590770722901e-05,
+      "loss": 0.4144,
       "step": 73435
     },
     {
-      "epoch": 2.58,
-      "learning_rate": 2.4870953497604126e-05,
-      "loss": 0.2709,
+      "epoch": 2.6467726240674665,
+      "grad_norm": 0.17443668842315674,
+      "learning_rate": 2.3852992638726368e-05,
+      "loss": 0.3669,
       "step": 73440
     },
     {
-      "epoch": 2.58,
-      "learning_rate": 2.4868104815087284e-05,
-      "loss": 0.2685,
+      "epoch": 2.6469528237286912,
+      "grad_norm": 0.2030688226222992,
+      "learning_rate": 2.385007758585148e-05,
+      "loss": 0.3755,
       "step": 73445
     },
     {
-      "epoch": 2.58,
-      "learning_rate": 2.486525613428301e-05,
-      "loss": 0.2622,
+      "epoch": 2.647133023389916,
+      "grad_norm": 0.1865500956773758,
+      "learning_rate": 2.3847162548644054e-05,
+      "loss": 0.3985,
       "step": 73450
     },
     {
-      "epoch": 2.58,
-      "learning_rate": 2.4862407455228303e-05,
-      "loss": 0.291,
+      "epoch": 2.6473132230511407,
+      "grad_norm": 0.16397204995155334,
+      "learning_rate": 2.3844247527143826e-05,
+      "loss": 0.3599,
       "step": 73455
     },
     {
-      "epoch": 2.58,
-      "learning_rate": 2.4859558777960135e-05,
-      "loss": 0.265,
+      "epoch": 2.6474934227123654,
+      "grad_norm": 0.16662755608558655,
+      "learning_rate": 2.3841332521390496e-05,
+      "loss": 0.3815,
       "step": 73460
     },
     {
-      "epoch": 2.58,
-      "learning_rate": 2.4856710102515513e-05,
-      "loss": 0.2648,
+      "epoch": 2.64767362237359,
+      "grad_norm": 0.2906810939311981,
+      "learning_rate": 2.383841753142378e-05,
+      "loss": 0.3871,
       "step": 73465
     },
     {
-      "epoch": 2.58,
-      "learning_rate": 2.4853861428931412e-05,
-      "loss": 0.3027,
+      "epoch": 2.6478538220348145,
+      "grad_norm": 0.21434706449508667,
+      "learning_rate": 2.383550255728341e-05,
+      "loss": 0.3876,
       "step": 73470
     },
     {
-      "epoch": 2.59,
-      "learning_rate": 2.4851012757244818e-05,
-      "loss": 0.3025,
+      "epoch": 2.648034021696039,
+      "grad_norm": 0.212555930018425,
+      "learning_rate": 2.3832587599009083e-05,
+      "loss": 0.4124,
       "step": 73475
     },
     {
-      "epoch": 2.59,
-      "learning_rate": 2.4848164087492735e-05,
-      "loss": 0.2561,
+      "epoch": 2.648214221357264,
+      "grad_norm": 0.26868897676467896,
+      "learning_rate": 2.3829672656640534e-05,
+      "loss": 0.4283,
       "step": 73480
     },
     {
-      "epoch": 2.59,
-      "learning_rate": 2.4845315419712135e-05,
-      "loss": 0.2634,
+      "epoch": 2.648394421018488,
+      "grad_norm": 0.20437636971473694,
+      "learning_rate": 2.3826757730217467e-05,
+      "loss": 0.3905,
       "step": 73485
     },
     {
-      "epoch": 2.59,
-      "learning_rate": 2.484246675394002e-05,
-      "loss": 0.2709,
+      "epoch": 2.648574620679713,
+      "grad_norm": 0.2129250317811966,
+      "learning_rate": 2.382384281977959e-05,
+      "loss": 0.4188,
       "step": 73490
     },
     {
-      "epoch": 2.59,
-      "learning_rate": 2.4839618090213357e-05,
-      "loss": 0.2787,
+      "epoch": 2.6487548203409377,
+      "grad_norm": 0.2285563051700592,
+      "learning_rate": 2.3820927925366634e-05,
+      "loss": 0.3996,
       "step": 73495
     },
     {
-      "epoch": 2.59,
-      "learning_rate": 2.4836769428569158e-05,
-      "loss": 0.2796,
+      "epoch": 2.6489350200021624,
+      "grad_norm": 0.20043610036373138,
+      "learning_rate": 2.3818013047018304e-05,
+      "loss": 0.403,
       "step": 73500
     },
     {
-      "epoch": 2.59,
-      "eval_loss": 0.2626335024833679,
-      "eval_runtime": 10.5465,
-      "eval_samples_per_second": 9.482,
-      "eval_steps_per_second": 9.482,
+      "epoch": 2.6489350200021624,
+      "eval_loss": 0.4324740469455719,
+      "eval_runtime": 3.5359,
+      "eval_samples_per_second": 28.282,
+      "eval_steps_per_second": 7.07,
       "step": 73500
     },
     {
-      "epoch": 2.59,
-      "learning_rate": 2.4833920769044402e-05,
-      "loss": 0.2616,
+      "epoch": 2.649115219663387,
+      "grad_norm": 0.21913573145866394,
+      "learning_rate": 2.3815098184774318e-05,
+      "loss": 0.4078,
       "step": 73505
     },
     {
-      "epoch": 2.59,
-      "learning_rate": 2.483107211167607e-05,
-      "loss": 0.2657,
+      "epoch": 2.649295419324612,
+      "grad_norm": 0.1827915608882904,
+      "learning_rate": 2.3812183338674393e-05,
+      "loss": 0.4111,
       "step": 73510
     },
     {
-      "epoch": 2.59,
-      "learning_rate": 2.482822345650115e-05,
-      "loss": 0.245,
+      "epoch": 2.649475618985836,
+      "grad_norm": 0.2369619905948639,
+      "learning_rate": 2.3809268508758232e-05,
+      "loss": 0.3877,
       "step": 73515
     },
     {
-      "epoch": 2.59,
-      "learning_rate": 2.4825374803556643e-05,
-      "loss": 0.2641,
+      "epoch": 2.649655818647061,
+      "grad_norm": 0.19070202112197876,
+      "learning_rate": 2.3806353695065564e-05,
+      "loss": 0.4125,
       "step": 73520
     },
     {
-      "epoch": 2.59,
-      "learning_rate": 2.4822526152879533e-05,
-      "loss": 0.2623,
+      "epoch": 2.6498360183082856,
+      "grad_norm": 0.2124054878950119,
+      "learning_rate": 2.3803438897636095e-05,
+      "loss": 0.3944,
       "step": 73525
     },
     {
-      "epoch": 2.59,
-      "learning_rate": 2.4819677504506795e-05,
-      "loss": 0.2704,
+      "epoch": 2.6500162179695104,
+      "grad_norm": 0.22031430900096893,
+      "learning_rate": 2.3800524116509537e-05,
+      "loss": 0.3687,
       "step": 73530
     },
     {
-      "epoch": 2.59,
-      "learning_rate": 2.481682885847543e-05,
-      "loss": 0.2854,
+      "epoch": 2.6501964176307347,
+      "grad_norm": 0.20872965455055237,
+      "learning_rate": 2.379760935172561e-05,
+      "loss": 0.376,
       "step": 73535
     },
     {
-      "epoch": 2.59,
-      "learning_rate": 2.4813980214822418e-05,
-      "loss": 0.2653,
+      "epoch": 2.6503766172919594,
+      "grad_norm": 0.2545884847640991,
+      "learning_rate": 2.3794694603324026e-05,
+      "loss": 0.3869,
       "step": 73540
     },
     {
-      "epoch": 2.59,
-      "learning_rate": 2.4811131573584757e-05,
-      "loss": 0.2757,
+      "epoch": 2.650556816953184,
+      "grad_norm": 0.23649878799915314,
+      "learning_rate": 2.379177987134448e-05,
+      "loss": 0.3706,
       "step": 73545
     },
     {
-      "epoch": 2.59,
-      "learning_rate": 2.480828293479943e-05,
-      "loss": 0.2732,
+      "epoch": 2.650737016614409,
+      "grad_norm": 0.20820178091526031,
+      "learning_rate": 2.3788865155826716e-05,
+      "loss": 0.4016,
       "step": 73550
     },
     {
-      "epoch": 2.59,
-      "learning_rate": 2.4805434298503407e-05,
-      "loss": 0.2669,
+      "epoch": 2.6509172162756336,
+      "grad_norm": 0.2441217005252838,
+      "learning_rate": 2.378595045681041e-05,
+      "loss": 0.3979,
       "step": 73555
     },
     {
-      "epoch": 2.59,
-      "learning_rate": 2.4802585664733708e-05,
-      "loss": 0.2607,
+      "epoch": 2.651097415936858,
+      "grad_norm": 0.2145996391773224,
+      "learning_rate": 2.3783035774335313e-05,
+      "loss": 0.4015,
       "step": 73560
     },
     {
-      "epoch": 2.59,
-      "learning_rate": 2.4799737033527296e-05,
-      "loss": 0.2629,
+      "epoch": 2.6512776155980826,
+      "grad_norm": 0.2503126263618469,
+      "learning_rate": 2.3780121108441116e-05,
+      "loss": 0.4389,
       "step": 73565
     },
     {
-      "epoch": 2.59,
-      "learning_rate": 2.4797458130432597e-05,
-      "loss": 0.2736,
+      "epoch": 2.6514578152593073,
+      "grad_norm": 0.19581691920757294,
+      "learning_rate": 2.377720645916752e-05,
+      "loss": 0.3767,
       "step": 73570
     },
     {
-      "epoch": 2.59,
-      "learning_rate": 2.4794609503933324e-05,
-      "loss": 0.2633,
+      "epoch": 2.651638014920532,
+      "grad_norm": 0.19556576013565063,
+      "learning_rate": 2.377429182655426e-05,
+      "loss": 0.389,
       "step": 73575
     },
     {
-      "epoch": 2.59,
-      "learning_rate": 2.4791760880100916e-05,
-      "loss": 0.2567,
+      "epoch": 2.6518182145817564,
+      "grad_norm": 0.2089264839887619,
+      "learning_rate": 2.3771377210641035e-05,
+      "loss": 0.3602,
       "step": 73580
     },
     {
-      "epoch": 2.59,
-      "learning_rate": 2.4788912258972347e-05,
-      "loss": 0.2552,
+      "epoch": 2.651998414242981,
+      "grad_norm": 0.21499642729759216,
+      "learning_rate": 2.3768462611467552e-05,
+      "loss": 0.4022,
       "step": 73585
     },
     {
-      "epoch": 2.59,
-      "learning_rate": 2.4786063640584623e-05,
-      "loss": 0.2663,
+      "epoch": 2.652178613904206,
+      "grad_norm": 0.27903100848197937,
+      "learning_rate": 2.3765548029073535e-05,
+      "loss": 0.417,
       "step": 73590
     },
     {
-      "epoch": 2.59,
-      "learning_rate": 2.4783215024974717e-05,
-      "loss": 0.2864,
+      "epoch": 2.6523588135654306,
+      "grad_norm": 0.2061304748058319,
+      "learning_rate": 2.3762633463498677e-05,
+      "loss": 0.3962,
       "step": 73595
     },
     {
-      "epoch": 2.59,
-      "learning_rate": 2.4780366412179627e-05,
-      "loss": 0.2785,
+      "epoch": 2.6525390132266553,
+      "grad_norm": 0.23685584962368011,
+      "learning_rate": 2.375971891478271e-05,
+      "loss": 0.3888,
       "step": 73600
     },
     {
-      "epoch": 2.59,
-      "learning_rate": 2.4777517802236323e-05,
-      "loss": 0.2649,
+      "epoch": 2.65271921288788,
+      "grad_norm": 0.22389626502990723,
+      "learning_rate": 2.3756804382965324e-05,
+      "loss": 0.4113,
       "step": 73605
     },
     {
-      "epoch": 2.59,
-      "learning_rate": 2.4774669195181813e-05,
-      "loss": 0.2524,
+      "epoch": 2.6528994125491043,
+      "grad_norm": 0.1898953765630722,
+      "learning_rate": 2.375388986808624e-05,
+      "loss": 0.3737,
       "step": 73610
     },
     {
-      "epoch": 2.59,
-      "learning_rate": 2.4771820591053075e-05,
-      "loss": 0.2715,
+      "epoch": 2.653079612210329,
+      "grad_norm": 0.22866353392601013,
+      "learning_rate": 2.375097537018517e-05,
+      "loss": 0.3694,
       "step": 73615
     },
     {
-      "epoch": 2.59,
-      "learning_rate": 2.476897198988709e-05,
-      "loss": 0.2782,
+      "epoch": 2.653259811871554,
+      "grad_norm": 0.22122244536876678,
+      "learning_rate": 2.374806088930181e-05,
+      "loss": 0.4329,
       "step": 73620
     },
     {
-      "epoch": 2.59,
-      "learning_rate": 2.4766123391720855e-05,
-      "loss": 0.2799,
+      "epoch": 2.653440011532778,
+      "grad_norm": 0.17833849787712097,
+      "learning_rate": 2.3745146425475884e-05,
+      "loss": 0.4094,
       "step": 73625
     },
     {
-      "epoch": 2.59,
-      "learning_rate": 2.4763274796591352e-05,
-      "loss": 0.2584,
+      "epoch": 2.653620211194003,
+      "grad_norm": 0.2136964350938797,
+      "learning_rate": 2.37422319787471e-05,
+      "loss": 0.4217,
       "step": 73630
     },
     {
-      "epoch": 2.59,
-      "learning_rate": 2.4760426204535575e-05,
-      "loss": 0.2706,
+      "epoch": 2.6538004108552276,
+      "grad_norm": 0.18438690900802612,
+      "learning_rate": 2.3739317549155148e-05,
+      "loss": 0.4136,
       "step": 73635
     },
     {
-      "epoch": 2.59,
-      "learning_rate": 2.4757577615590507e-05,
-      "loss": 0.2862,
+      "epoch": 2.6539806105164523,
+      "grad_norm": 0.23062624037265778,
+      "learning_rate": 2.373640313673976e-05,
+      "loss": 0.3857,
       "step": 73640
     },
     {
-      "epoch": 2.59,
-      "learning_rate": 2.475472902979312e-05,
-      "loss": 0.2709,
+      "epoch": 2.654160810177677,
+      "grad_norm": 0.26239144802093506,
+      "learning_rate": 2.373348874154064e-05,
+      "loss": 0.4017,
       "step": 73645
     },
     {
-      "epoch": 2.59,
-      "learning_rate": 2.4751880447180427e-05,
-      "loss": 0.2287,
+      "epoch": 2.6543410098389018,
+      "grad_norm": 0.22294588387012482,
+      "learning_rate": 2.3730574363597476e-05,
+      "loss": 0.4033,
       "step": 73650
     },
     {
-      "epoch": 2.59,
-      "learning_rate": 2.47490318677894e-05,
-      "loss": 0.2704,
+      "epoch": 2.654521209500126,
+      "grad_norm": 0.17945276200771332,
+      "learning_rate": 2.3727660002950006e-05,
+      "loss": 0.3721,
       "step": 73655
     },
     {
-      "epoch": 2.59,
-      "learning_rate": 2.4746183291657022e-05,
-      "loss": 0.2665,
+      "epoch": 2.654701409161351,
+      "grad_norm": 0.22688071429729462,
+      "learning_rate": 2.3724745659637902e-05,
+      "loss": 0.3883,
       "step": 73660
     },
     {
-      "epoch": 2.59,
-      "learning_rate": 2.4743334718820298e-05,
-      "loss": 0.2821,
+      "epoch": 2.6548816088225755,
+      "grad_norm": 0.21472230553627014,
+      "learning_rate": 2.3721831333700913e-05,
+      "loss": 0.4231,
       "step": 73665
     },
     {
-      "epoch": 2.59,
-      "learning_rate": 2.4740486149316204e-05,
-      "loss": 0.2545,
+      "epoch": 2.6550618084838,
+      "grad_norm": 0.23527327179908752,
+      "learning_rate": 2.371891702517872e-05,
+      "loss": 0.3958,
       "step": 73670
     },
     {
-      "epoch": 2.59,
-      "learning_rate": 2.4737637583181716e-05,
-      "loss": 0.2813,
+      "epoch": 2.6552420081450245,
+      "grad_norm": 0.19936297833919525,
+      "learning_rate": 2.3716002734111024e-05,
+      "loss": 0.3937,
       "step": 73675
     },
     {
-      "epoch": 2.59,
-      "learning_rate": 2.4734789020453835e-05,
-      "loss": 0.2617,
+      "epoch": 2.6554222078062493,
+      "grad_norm": 0.19073469936847687,
+      "learning_rate": 2.371308846053755e-05,
+      "loss": 0.3815,
       "step": 73680
     },
     {
-      "epoch": 2.59,
-      "learning_rate": 2.4731940461169543e-05,
-      "loss": 0.2424,
+      "epoch": 2.655602407467474,
+      "grad_norm": 0.19665153324604034,
+      "learning_rate": 2.3710174204497997e-05,
+      "loss": 0.4036,
       "step": 73685
     },
     {
-      "epoch": 2.59,
-      "learning_rate": 2.4729091905365835e-05,
-      "loss": 0.277,
+      "epoch": 2.6557826071286987,
+      "grad_norm": 0.2125919610261917,
+      "learning_rate": 2.3707259966032064e-05,
+      "loss": 0.4332,
       "step": 73690
     },
     {
-      "epoch": 2.59,
-      "learning_rate": 2.4726243353079685e-05,
-      "loss": 0.2643,
+      "epoch": 2.6559628067899235,
+      "grad_norm": 0.16980135440826416,
+      "learning_rate": 2.370434574517947e-05,
+      "loss": 0.4054,
       "step": 73695
     },
     {
-      "epoch": 2.59,
-      "learning_rate": 2.472339480434808e-05,
-      "loss": 0.273,
+      "epoch": 2.6561430064511478,
+      "grad_norm": 0.16812558472156525,
+      "learning_rate": 2.370143154197991e-05,
+      "loss": 0.3967,
       "step": 73700
     },
     {
-      "epoch": 2.59,
-      "learning_rate": 2.472054625920802e-05,
-      "loss": 0.2446,
+      "epoch": 2.6563232061123725,
+      "grad_norm": 0.23627859354019165,
+      "learning_rate": 2.3698517356473098e-05,
+      "loss": 0.3897,
       "step": 73705
     },
     {
-      "epoch": 2.59,
-      "learning_rate": 2.4717697717696475e-05,
-      "loss": 0.2484,
+      "epoch": 2.6565034057735972,
+      "grad_norm": 0.2503635287284851,
+      "learning_rate": 2.3695603188698733e-05,
+      "loss": 0.3958,
       "step": 73710
     },
     {
-      "epoch": 2.59,
-      "learning_rate": 2.4714849179850448e-05,
-      "loss": 0.2785,
+      "epoch": 2.6566836054348215,
+      "grad_norm": 0.2157202512025833,
+      "learning_rate": 2.3692689038696518e-05,
+      "loss": 0.4087,
       "step": 73715
     },
     {
-      "epoch": 2.59,
-      "learning_rate": 2.47120006457069e-05,
-      "loss": 0.2813,
+      "epoch": 2.6568638050960462,
+      "grad_norm": 0.1902592033147812,
+      "learning_rate": 2.368977490650617e-05,
+      "loss": 0.389,
       "step": 73720
     },
     {
-      "epoch": 2.59,
-      "learning_rate": 2.470915211530285e-05,
-      "loss": 0.2686,
+      "epoch": 2.657044004757271,
+      "grad_norm": 0.188002347946167,
+      "learning_rate": 2.368686079216739e-05,
+      "loss": 0.3733,
       "step": 73725
     },
     {
-      "epoch": 2.59,
-      "learning_rate": 2.4706303588675267e-05,
-      "loss": 0.2724,
+      "epoch": 2.6572242044184957,
+      "grad_norm": 0.22462768852710724,
+      "learning_rate": 2.3683946695719857e-05,
+      "loss": 0.3922,
       "step": 73730
     },
     {
-      "epoch": 2.59,
-      "learning_rate": 2.4703455065861133e-05,
-      "loss": 0.268,
+      "epoch": 2.6574044040797205,
+      "grad_norm": 0.22756509482860565,
+      "learning_rate": 2.3681032617203317e-05,
+      "loss": 0.4083,
       "step": 73735
     },
     {
-      "epoch": 2.59,
-      "learning_rate": 2.4700606546897432e-05,
-      "loss": 0.2808,
+      "epoch": 2.657584603740945,
+      "grad_norm": 0.2529270052909851,
+      "learning_rate": 2.367811855665743e-05,
+      "loss": 0.3848,
       "step": 73740
     },
     {
-      "epoch": 2.59,
-      "learning_rate": 2.469775803182117e-05,
-      "loss": 0.2872,
+      "epoch": 2.6577648034021695,
+      "grad_norm": 0.2393999844789505,
+      "learning_rate": 2.3675204514121942e-05,
+      "loss": 0.3802,
       "step": 73745
     },
     {
-      "epoch": 2.59,
-      "learning_rate": 2.469490952066932e-05,
-      "loss": 0.2504,
+      "epoch": 2.657945003063394,
+      "grad_norm": 0.2106008678674698,
+      "learning_rate": 2.3672290489636533e-05,
+      "loss": 0.4216,
       "step": 73750
     },
     {
-      "epoch": 2.59,
-      "learning_rate": 2.469206101347887e-05,
-      "loss": 0.2543,
+      "epoch": 2.658125202724619,
+      "grad_norm": 0.19306325912475586,
+      "learning_rate": 2.3669376483240894e-05,
+      "loss": 0.3937,
       "step": 73755
     },
     {
-      "epoch": 2.6,
-      "learning_rate": 2.4689212510286793e-05,
-      "loss": 0.2908,
+      "epoch": 2.6583054023858437,
+      "grad_norm": 0.2211955040693283,
+      "learning_rate": 2.366646249497476e-05,
+      "loss": 0.4201,
       "step": 73760
     },
     {
-      "epoch": 2.6,
-      "learning_rate": 2.4686364011130094e-05,
-      "loss": 0.2758,
+      "epoch": 2.658485602047068,
+      "grad_norm": 0.20790930092334747,
+      "learning_rate": 2.366354852487781e-05,
+      "loss": 0.368,
       "step": 73765
     },
     {
-      "epoch": 2.6,
-      "learning_rate": 2.4683515516045756e-05,
-      "loss": 0.2654,
+      "epoch": 2.6586658017082927,
+      "grad_norm": 0.2273816168308258,
+      "learning_rate": 2.3660634572989747e-05,
+      "loss": 0.3965,
       "step": 73770
     },
     {
-      "epoch": 2.6,
-      "learning_rate": 2.468066702507075e-05,
-      "loss": 0.2751,
+      "epoch": 2.6588460013695174,
+      "grad_norm": 0.22291794419288635,
+      "learning_rate": 2.3657720639350288e-05,
+      "loss": 0.4491,
       "step": 73775
     },
     {
-      "epoch": 2.6,
-      "learning_rate": 2.467781853824208e-05,
-      "loss": 0.2788,
+      "epoch": 2.659026201030742,
+      "grad_norm": 0.21958285570144653,
+      "learning_rate": 2.3654806723999117e-05,
+      "loss": 0.3876,
       "step": 73780
     },
     {
-      "epoch": 2.6,
-      "learning_rate": 2.4674970055596727e-05,
-      "loss": 0.2844,
+      "epoch": 2.659206400691967,
+      "grad_norm": 0.23919081687927246,
+      "learning_rate": 2.365189282697595e-05,
+      "loss": 0.3955,
       "step": 73785
     },
     {
-      "epoch": 2.6,
-      "learning_rate": 2.4672121577171667e-05,
-      "loss": 0.2669,
+      "epoch": 2.659386600353191,
+      "grad_norm": 0.2069309651851654,
+      "learning_rate": 2.3648978948320483e-05,
+      "loss": 0.4031,
       "step": 73790
     },
     {
-      "epoch": 2.6,
-      "learning_rate": 2.4669273103003886e-05,
-      "loss": 0.2827,
+      "epoch": 2.659566800014416,
+      "grad_norm": 0.22566114366054535,
+      "learning_rate": 2.3646065088072407e-05,
+      "loss": 0.3822,
       "step": 73795
     },
     {
-      "epoch": 2.6,
-      "learning_rate": 2.4666424633130393e-05,
-      "loss": 0.2853,
+      "epoch": 2.6597469996756407,
+      "grad_norm": 0.193036288022995,
+      "learning_rate": 2.364315124627144e-05,
+      "loss": 0.3603,
       "step": 73800
     },
     {
-      "epoch": 2.6,
-      "learning_rate": 2.466357616758815e-05,
-      "loss": 0.2893,
+      "epoch": 2.6599271993368654,
+      "grad_norm": 0.19962070882320404,
+      "learning_rate": 2.3640237422957275e-05,
+      "loss": 0.3869,
       "step": 73805
     },
     {
-      "epoch": 2.6,
-      "learning_rate": 2.4660727706414148e-05,
-      "loss": 0.2785,
+      "epoch": 2.6601073989980897,
+      "grad_norm": 0.20868006348609924,
+      "learning_rate": 2.3637323618169606e-05,
+      "loss": 0.3864,
       "step": 73810
     },
     {
-      "epoch": 2.6,
-      "learning_rate": 2.4657879249645366e-05,
-      "loss": 0.2621,
+      "epoch": 2.6602875986593144,
+      "grad_norm": 0.22715534269809723,
+      "learning_rate": 2.3634409831948144e-05,
+      "loss": 0.3757,
       "step": 73815
     },
     {
-      "epoch": 2.6,
-      "learning_rate": 2.4655030797318797e-05,
-      "loss": 0.2654,
+      "epoch": 2.660467798320539,
+      "grad_norm": 0.22947651147842407,
+      "learning_rate": 2.363149606433258e-05,
+      "loss": 0.3707,
       "step": 73820
     },
     {
-      "epoch": 2.6,
-      "learning_rate": 2.4652182349471437e-05,
-      "loss": 0.2726,
+      "epoch": 2.660647997981764,
+      "grad_norm": 0.17915059626102448,
+      "learning_rate": 2.3628582315362625e-05,
+      "loss": 0.4104,
       "step": 73825
     },
     {
-      "epoch": 2.6,
-      "learning_rate": 2.4649333906140256e-05,
-      "loss": 0.2621,
+      "epoch": 2.6608281976429886,
+      "grad_norm": 0.2261868566274643,
+      "learning_rate": 2.362566858507797e-05,
+      "loss": 0.3774,
       "step": 73830
     },
     {
-      "epoch": 2.6,
-      "learning_rate": 2.4646485467362234e-05,
-      "loss": 0.287,
+      "epoch": 2.661008397304213,
+      "grad_norm": 0.23218120634555817,
+      "learning_rate": 2.3622754873518303e-05,
+      "loss": 0.4169,
       "step": 73835
     },
     {
-      "epoch": 2.6,
-      "learning_rate": 2.4643637033174378e-05,
-      "loss": 0.2776,
+      "epoch": 2.6611885969654376,
+      "grad_norm": 0.2152547836303711,
+      "learning_rate": 2.3619841180723345e-05,
+      "loss": 0.4173,
       "step": 73840
     },
     {
-      "epoch": 2.6,
-      "learning_rate": 2.4640788603613652e-05,
-      "loss": 0.2683,
+      "epoch": 2.6613687966266624,
+      "grad_norm": 0.22858308255672455,
+      "learning_rate": 2.3616927506732773e-05,
+      "loss": 0.3866,
       "step": 73845
     },
     {
-      "epoch": 2.6,
-      "learning_rate": 2.4637940178717053e-05,
-      "loss": 0.2783,
+      "epoch": 2.661548996287887,
+      "grad_norm": 0.2074364870786667,
+      "learning_rate": 2.361401385158631e-05,
+      "loss": 0.3856,
       "step": 73850
     },
     {
-      "epoch": 2.6,
-      "learning_rate": 2.4635091758521554e-05,
-      "loss": 0.2617,
+      "epoch": 2.6617291959491114,
+      "grad_norm": 0.1993214190006256,
+      "learning_rate": 2.361110021532363e-05,
+      "loss": 0.4015,
       "step": 73855
     },
     {
-      "epoch": 2.6,
-      "learning_rate": 2.4632243343064158e-05,
-      "loss": 0.2614,
+      "epoch": 2.661909395610336,
+      "grad_norm": 0.22513341903686523,
+      "learning_rate": 2.360818659798444e-05,
+      "loss": 0.3876,
       "step": 73860
     },
     {
-      "epoch": 2.6,
-      "learning_rate": 2.4629394932381842e-05,
-      "loss": 0.2846,
+      "epoch": 2.662089595271561,
+      "grad_norm": 0.20720557868480682,
+      "learning_rate": 2.3605272999608442e-05,
+      "loss": 0.4208,
       "step": 73865
     },
     {
-      "epoch": 2.6,
-      "learning_rate": 2.462654652651158e-05,
-      "loss": 0.2668,
+      "epoch": 2.6622697949327856,
+      "grad_norm": 0.19978566467761993,
+      "learning_rate": 2.3602359420235333e-05,
+      "loss": 0.4022,
       "step": 73870
     },
     {
-      "epoch": 2.6,
-      "learning_rate": 2.4623698125490364e-05,
-      "loss": 0.2688,
+      "epoch": 2.6624499945940103,
+      "grad_norm": 0.23757880926132202,
+      "learning_rate": 2.3599445859904798e-05,
+      "loss": 0.3755,
       "step": 73875
     },
     {
-      "epoch": 2.6,
-      "learning_rate": 2.4620849729355186e-05,
-      "loss": 0.267,
+      "epoch": 2.662630194255235,
+      "grad_norm": 0.17947709560394287,
+      "learning_rate": 2.3596532318656547e-05,
+      "loss": 0.4252,
       "step": 73880
     },
     {
-      "epoch": 2.6,
-      "learning_rate": 2.4618001338143027e-05,
-      "loss": 0.2587,
+      "epoch": 2.6628103939164594,
+      "grad_norm": 0.22774136066436768,
+      "learning_rate": 2.3593618796530268e-05,
+      "loss": 0.3899,
       "step": 73885
     },
     {
-      "epoch": 2.6,
-      "learning_rate": 2.461515295189086e-05,
-      "loss": 0.2811,
+      "epoch": 2.662990593577684,
+      "grad_norm": 0.23940308392047882,
+      "learning_rate": 2.3590705293565663e-05,
+      "loss": 0.3636,
       "step": 73890
     },
     {
-      "epoch": 2.6,
-      "learning_rate": 2.4612304570635684e-05,
-      "loss": 0.2919,
+      "epoch": 2.663170793238909,
+      "grad_norm": 0.1792476326227188,
+      "learning_rate": 2.3587791809802427e-05,
+      "loss": 0.3771,
       "step": 73895
     },
     {
-      "epoch": 2.6,
-      "learning_rate": 2.4609456194414473e-05,
-      "loss": 0.2534,
+      "epoch": 2.663350992900133,
+      "grad_norm": 0.15978147089481354,
+      "learning_rate": 2.358487834528025e-05,
+      "loss": 0.4011,
       "step": 73900
     },
     {
-      "epoch": 2.6,
-      "learning_rate": 2.4606607823264225e-05,
-      "loss": 0.2737,
+      "epoch": 2.663531192561358,
+      "grad_norm": 0.20006360113620758,
+      "learning_rate": 2.3581964900038836e-05,
+      "loss": 0.3993,
       "step": 73905
     },
     {
-      "epoch": 2.6,
-      "learning_rate": 2.4603759457221904e-05,
-      "loss": 0.2857,
+      "epoch": 2.6637113922225826,
+      "grad_norm": 0.1868210732936859,
+      "learning_rate": 2.357905147411788e-05,
+      "loss": 0.3967,
       "step": 73910
     },
     {
-      "epoch": 2.6,
-      "learning_rate": 2.4600911096324513e-05,
-      "loss": 0.2629,
+      "epoch": 2.6638915918838073,
+      "grad_norm": 0.21746650338172913,
+      "learning_rate": 2.3576138067557058e-05,
+      "loss": 0.3769,
       "step": 73915
     },
     {
-      "epoch": 2.6,
-      "learning_rate": 2.4598062740609028e-05,
-      "loss": 0.2631,
+      "epoch": 2.664071791545032,
+      "grad_norm": 0.21393951773643494,
+      "learning_rate": 2.357322468039609e-05,
+      "loss": 0.4063,
       "step": 73920
     },
     {
-      "epoch": 2.6,
-      "learning_rate": 2.459521439011243e-05,
-      "loss": 0.2634,
+      "epoch": 2.6642519912062568,
+      "grad_norm": 0.21100027859210968,
+      "learning_rate": 2.3570311312674654e-05,
+      "loss": 0.4181,
       "step": 73925
     },
     {
-      "epoch": 2.6,
-      "learning_rate": 2.4592366044871708e-05,
-      "loss": 0.2748,
+      "epoch": 2.664432190867481,
+      "grad_norm": 0.20578476786613464,
+      "learning_rate": 2.356739796443245e-05,
+      "loss": 0.396,
       "step": 73930
     },
     {
-      "epoch": 2.6,
-      "learning_rate": 2.4589517704923843e-05,
-      "loss": 0.2676,
+      "epoch": 2.664612390528706,
+      "grad_norm": 0.2200337052345276,
+      "learning_rate": 2.356448463570918e-05,
+      "loss": 0.3953,
       "step": 73935
     },
     {
-      "epoch": 2.6,
-      "learning_rate": 2.4586669370305825e-05,
-      "loss": 0.2539,
+      "epoch": 2.6647925901899305,
+      "grad_norm": 0.21242652833461761,
+      "learning_rate": 2.356157132654452e-05,
+      "loss": 0.4087,
       "step": 73940
     },
     {
-      "epoch": 2.6,
-      "learning_rate": 2.4583821041054634e-05,
-      "loss": 0.2718,
+      "epoch": 2.664972789851155,
+      "grad_norm": 0.2014140486717224,
+      "learning_rate": 2.3558658036978183e-05,
+      "loss": 0.3968,
       "step": 73945
     },
     {
-      "epoch": 2.6,
-      "learning_rate": 2.458097271720724e-05,
-      "loss": 0.2683,
+      "epoch": 2.6651529895123796,
+      "grad_norm": 0.2067977786064148,
+      "learning_rate": 2.355574476704984e-05,
+      "loss": 0.3766,
       "step": 73950
     },
     {
-      "epoch": 2.6,
-      "learning_rate": 2.4578124398800645e-05,
-      "loss": 0.2919,
+      "epoch": 2.6653331891736043,
+      "grad_norm": 0.18431727588176727,
+      "learning_rate": 2.35528315167992e-05,
+      "loss": 0.3723,
       "step": 73955
     },
     {
-      "epoch": 2.6,
-      "learning_rate": 2.457527608587183e-05,
-      "loss": 0.282,
+      "epoch": 2.665513388834829,
+      "grad_norm": 0.2466651052236557,
+      "learning_rate": 2.3549918286265947e-05,
+      "loss": 0.3828,
       "step": 73960
     },
     {
-      "epoch": 2.6,
-      "learning_rate": 2.457242777845778e-05,
-      "loss": 0.2822,
+      "epoch": 2.6656935884960538,
+      "grad_norm": 0.22020891308784485,
+      "learning_rate": 2.3547005075489776e-05,
+      "loss": 0.3619,
       "step": 73965
     },
     {
-      "epoch": 2.6,
-      "learning_rate": 2.456957947659546e-05,
-      "loss": 0.2758,
+      "epoch": 2.6658737881572785,
+      "grad_norm": 0.173415869474411,
+      "learning_rate": 2.3544091884510383e-05,
+      "loss": 0.3719,
       "step": 73970
     },
     {
-      "epoch": 2.6,
-      "learning_rate": 2.4566731180321877e-05,
-      "loss": 0.2454,
+      "epoch": 2.666053987818503,
+      "grad_norm": 0.17796094715595245,
+      "learning_rate": 2.3541178713367456e-05,
+      "loss": 0.4016,
       "step": 73975
     },
     {
-      "epoch": 2.6,
-      "learning_rate": 2.4563882889674e-05,
-      "loss": 0.2699,
+      "epoch": 2.6662341874797275,
+      "grad_norm": 0.235768124461174,
+      "learning_rate": 2.353826556210068e-05,
+      "loss": 0.3881,
       "step": 73980
     },
     {
-      "epoch": 2.6,
-      "learning_rate": 2.4561034604688822e-05,
-      "loss": 0.2696,
+      "epoch": 2.6664143871409522,
+      "grad_norm": 0.253508597612381,
+      "learning_rate": 2.3535352430749763e-05,
+      "loss": 0.4,
       "step": 73985
     },
     {
-      "epoch": 2.6,
-      "learning_rate": 2.4558186325403305e-05,
-      "loss": 0.276,
+      "epoch": 2.6665945868021765,
+      "grad_norm": 0.20636709034442902,
+      "learning_rate": 2.353243931935438e-05,
+      "loss": 0.3943,
       "step": 73990
     },
     {
-      "epoch": 2.6,
-      "learning_rate": 2.4555338051854465e-05,
-      "loss": 0.2672,
+      "epoch": 2.6667747864634013,
+      "grad_norm": 0.20948311686515808,
+      "learning_rate": 2.3529526227954225e-05,
+      "loss": 0.3992,
       "step": 73995
     },
     {
-      "epoch": 2.6,
-      "learning_rate": 2.4552489784079262e-05,
-      "loss": 0.286,
+      "epoch": 2.666954986124626,
+      "grad_norm": 0.19002006947994232,
+      "learning_rate": 2.3526613156588997e-05,
+      "loss": 0.3784,
       "step": 74000
     },
     {
-      "epoch": 2.6,
-      "eval_loss": 0.26184189319610596,
-      "eval_runtime": 10.5446,
-      "eval_samples_per_second": 9.484,
-      "eval_steps_per_second": 9.484,
+      "epoch": 2.666954986124626,
+      "eval_loss": 0.4320438802242279,
+      "eval_runtime": 3.5274,
+      "eval_samples_per_second": 28.35,
+      "eval_steps_per_second": 7.087,
       "step": 74000
     },
     {
-      "epoch": 2.6,
-      "learning_rate": 2.4549641522114678e-05,
-      "loss": 0.2697,
+      "epoch": 2.6671351857858507,
+      "grad_norm": 0.2230537235736847,
+      "learning_rate": 2.3523700105298372e-05,
+      "loss": 0.3806,
       "step": 74005
     },
     {
-      "epoch": 2.6,
-      "learning_rate": 2.4546793265997708e-05,
-      "loss": 0.2528,
+      "epoch": 2.6673153854470755,
+      "grad_norm": 0.20656515657901764,
+      "learning_rate": 2.3520787074122052e-05,
+      "loss": 0.3979,
       "step": 74010
     },
     {
-      "epoch": 2.6,
-      "learning_rate": 2.4543945015765327e-05,
-      "loss": 0.2671,
+      "epoch": 2.6674955851083,
+      "grad_norm": 0.2289830446243286,
+      "learning_rate": 2.351787406309973e-05,
+      "loss": 0.4243,
       "step": 74015
     },
     {
-      "epoch": 2.6,
-      "learning_rate": 2.4541096771454525e-05,
-      "loss": 0.2972,
+      "epoch": 2.6676757847695245,
+      "grad_norm": 0.22350530326366425,
+      "learning_rate": 2.3514961072271068e-05,
+      "loss": 0.407,
       "step": 74020
     },
     {
-      "epoch": 2.6,
-      "learning_rate": 2.4538248533102265e-05,
-      "loss": 0.2878,
+      "epoch": 2.6678559844307492,
+      "grad_norm": 0.224583700299263,
+      "learning_rate": 2.351204810167579e-05,
+      "loss": 0.3582,
       "step": 74025
     },
     {
-      "epoch": 2.6,
-      "learning_rate": 2.4535400300745555e-05,
-      "loss": 0.282,
+      "epoch": 2.668036184091974,
+      "grad_norm": 0.24546033143997192,
+      "learning_rate": 2.3509135151353553e-05,
+      "loss": 0.4033,
       "step": 74030
     },
     {
-      "epoch": 2.6,
-      "learning_rate": 2.4532552074421363e-05,
-      "loss": 0.2478,
+      "epoch": 2.6682163837531987,
+      "grad_norm": 0.19116534292697906,
+      "learning_rate": 2.350622222134408e-05,
+      "loss": 0.4039,
       "step": 74035
     },
     {
-      "epoch": 2.6,
-      "learning_rate": 2.452970385416668e-05,
-      "loss": 0.2523,
+      "epoch": 2.668396583414423,
+      "grad_norm": 0.1550498604774475,
+      "learning_rate": 2.350330931168703e-05,
+      "loss": 0.3879,
       "step": 74040
     },
     {
-      "epoch": 2.61,
-      "learning_rate": 2.452685564001847e-05,
-      "loss": 0.2701,
+      "epoch": 2.6685767830756477,
+      "grad_norm": 0.2186582088470459,
+      "learning_rate": 2.3500396422422092e-05,
+      "loss": 0.3957,
       "step": 74045
     },
     {
-      "epoch": 2.61,
-      "learning_rate": 2.452400743201374e-05,
-      "loss": 0.2813,
+      "epoch": 2.6687569827368725,
+      "grad_norm": 0.21175815165042877,
+      "learning_rate": 2.3497483553588977e-05,
+      "loss": 0.3772,
       "step": 74050
     },
     {
-      "epoch": 2.61,
-      "learning_rate": 2.4521159230189456e-05,
-      "loss": 0.2408,
+      "epoch": 2.668937182398097,
+      "grad_norm": 0.25158798694610596,
+      "learning_rate": 2.3494570705227355e-05,
+      "loss": 0.3367,
       "step": 74055
     },
     {
-      "epoch": 2.61,
-      "learning_rate": 2.45183110345826e-05,
-      "loss": 0.2699,
+      "epoch": 2.669117382059322,
+      "grad_norm": 0.21257662773132324,
+      "learning_rate": 2.349165787737691e-05,
+      "loss": 0.3836,
       "step": 74060
     },
     {
-      "epoch": 2.61,
-      "learning_rate": 2.4515462845230158e-05,
-      "loss": 0.2767,
+      "epoch": 2.669297581720546,
+      "grad_norm": 0.22067958116531372,
+      "learning_rate": 2.348874507007734e-05,
+      "loss": 0.4208,
       "step": 74065
     },
     {
-      "epoch": 2.61,
-      "learning_rate": 2.4512614662169116e-05,
-      "loss": 0.2737,
+      "epoch": 2.669477781381771,
+      "grad_norm": 0.18371091783046722,
+      "learning_rate": 2.348583228336832e-05,
+      "loss": 0.3965,
       "step": 74070
     },
     {
-      "epoch": 2.61,
-      "learning_rate": 2.450976648543645e-05,
-      "loss": 0.2806,
+      "epoch": 2.6696579810429957,
+      "grad_norm": 0.2284574806690216,
+      "learning_rate": 2.3482919517289543e-05,
+      "loss": 0.3892,
       "step": 74075
     },
     {
-      "epoch": 2.61,
-      "learning_rate": 2.4506918315069148e-05,
-      "loss": 0.2699,
+      "epoch": 2.6698381807042204,
+      "grad_norm": 0.20297889411449432,
+      "learning_rate": 2.34800067718807e-05,
+      "loss": 0.4246,
       "step": 74080
     },
     {
-      "epoch": 2.61,
-      "learning_rate": 2.4504070151104172e-05,
-      "loss": 0.262,
+      "epoch": 2.6700183803654447,
+      "grad_norm": 0.23391233384609222,
+      "learning_rate": 2.3477094047181463e-05,
+      "loss": 0.4102,
       "step": 74085
     },
     {
-      "epoch": 2.61,
-      "learning_rate": 2.4501221993578533e-05,
-      "loss": 0.2613,
+      "epoch": 2.6701985800266694,
+      "grad_norm": 0.1801353245973587,
+      "learning_rate": 2.347418134323153e-05,
+      "loss": 0.3542,
       "step": 74090
     },
     {
-      "epoch": 2.61,
-      "learning_rate": 2.4498373842529187e-05,
-      "loss": 0.2739,
+      "epoch": 2.670378779687894,
+      "grad_norm": 0.20619447529315948,
+      "learning_rate": 2.347126866007058e-05,
+      "loss": 0.3497,
       "step": 74095
     },
     {
-      "epoch": 2.61,
-      "learning_rate": 2.4495525697993134e-05,
-      "loss": 0.2618,
+      "epoch": 2.670558979349119,
+      "grad_norm": 0.15153087675571442,
+      "learning_rate": 2.3468355997738293e-05,
+      "loss": 0.4128,
       "step": 74100
     },
     {
-      "epoch": 2.61,
-      "learning_rate": 2.449267756000734e-05,
-      "loss": 0.261,
+      "epoch": 2.6707391790103436,
+      "grad_norm": 0.26451894640922546,
+      "learning_rate": 2.3465443356274365e-05,
+      "loss": 0.4019,
       "step": 74105
     },
     {
-      "epoch": 2.61,
-      "learning_rate": 2.4489829428608802e-05,
-      "loss": 0.2796,
+      "epoch": 2.6709193786715684,
+      "grad_norm": 0.24107202887535095,
+      "learning_rate": 2.3462530735718472e-05,
+      "loss": 0.4251,
       "step": 74110
     },
     {
-      "epoch": 2.61,
-      "learning_rate": 2.4486981303834485e-05,
-      "loss": 0.2656,
+      "epoch": 2.6710995783327927,
+      "grad_norm": 0.2847641110420227,
+      "learning_rate": 2.34596181361103e-05,
+      "loss": 0.4012,
       "step": 74115
     },
     {
-      "epoch": 2.61,
-      "learning_rate": 2.448413318572138e-05,
-      "loss": 0.2802,
+      "epoch": 2.6712797779940174,
+      "grad_norm": 0.25060930848121643,
+      "learning_rate": 2.3456705557489543e-05,
+      "loss": 0.4052,
       "step": 74120
     },
     {
-      "epoch": 2.61,
-      "learning_rate": 2.4481285074306468e-05,
-      "loss": 0.2774,
+      "epoch": 2.671459977655242,
+      "grad_norm": 0.2532309591770172,
+      "learning_rate": 2.3453792999895855e-05,
+      "loss": 0.3964,
       "step": 74125
     },
     {
-      "epoch": 2.61,
-      "learning_rate": 2.4478436969626727e-05,
-      "loss": 0.2624,
+      "epoch": 2.6716401773164664,
+      "grad_norm": 0.1897568553686142,
+      "learning_rate": 2.3450880463368955e-05,
+      "loss": 0.3838,
       "step": 74130
     },
     {
-      "epoch": 2.61,
-      "learning_rate": 2.4475588871719144e-05,
-      "loss": 0.2983,
+      "epoch": 2.671820376977691,
+      "grad_norm": 0.2430182695388794,
+      "learning_rate": 2.3447967947948503e-05,
+      "loss": 0.4141,
       "step": 74135
     },
     {
-      "epoch": 2.61,
-      "learning_rate": 2.4472740780620678e-05,
-      "loss": 0.2542,
+      "epoch": 2.672000576638916,
+      "grad_norm": 0.22214165329933167,
+      "learning_rate": 2.344505545367418e-05,
+      "loss": 0.4151,
       "step": 74140
     },
     {
-      "epoch": 2.61,
-      "learning_rate": 2.446989269636834e-05,
-      "loss": 0.2763,
+      "epoch": 2.6721807763001406,
+      "grad_norm": 0.19647006690502167,
+      "learning_rate": 2.344214298058568e-05,
+      "loss": 0.4073,
       "step": 74145
     },
     {
-      "epoch": 2.61,
-      "learning_rate": 2.446704461899909e-05,
-      "loss": 0.2834,
+      "epoch": 2.6723609759613653,
+      "grad_norm": 0.18476322293281555,
+      "learning_rate": 2.343923052872268e-05,
+      "loss": 0.3706,
       "step": 74150
     },
     {
-      "epoch": 2.61,
-      "learning_rate": 2.446419654854992e-05,
-      "loss": 0.2549,
+      "epoch": 2.67254117562259,
+      "grad_norm": 0.18611617386341095,
+      "learning_rate": 2.3436318098124864e-05,
+      "loss": 0.3995,
       "step": 74155
     },
     {
-      "epoch": 2.61,
-      "learning_rate": 2.4461348485057792e-05,
-      "loss": 0.2625,
+      "epoch": 2.6727213752838144,
+      "grad_norm": 0.2218163013458252,
+      "learning_rate": 2.3433405688831915e-05,
+      "loss": 0.3928,
       "step": 74160
     },
     {
-      "epoch": 2.61,
-      "learning_rate": 2.445850042855971e-05,
-      "loss": 0.2741,
+      "epoch": 2.672901574945039,
+      "grad_norm": 0.16507494449615479,
+      "learning_rate": 2.34304933008835e-05,
+      "loss": 0.3753,
       "step": 74165
     },
     {
-      "epoch": 2.61,
-      "learning_rate": 2.4455652379092644e-05,
-      "loss": 0.2667,
+      "epoch": 2.673081774606264,
+      "grad_norm": 0.19643135368824005,
+      "learning_rate": 2.3427580934319314e-05,
+      "loss": 0.4089,
       "step": 74170
     },
     {
-      "epoch": 2.61,
-      "learning_rate": 2.445280433669357e-05,
-      "loss": 0.2613,
+      "epoch": 2.673261974267488,
+      "grad_norm": 0.22430889308452606,
+      "learning_rate": 2.3424668589179037e-05,
+      "loss": 0.3765,
       "step": 74175
     },
     {
-      "epoch": 2.61,
-      "learning_rate": 2.4449956301399464e-05,
-      "loss": 0.2699,
+      "epoch": 2.673442173928713,
+      "grad_norm": 0.2428254932165146,
+      "learning_rate": 2.342175626550234e-05,
+      "loss": 0.3831,
       "step": 74180
     },
     {
-      "epoch": 2.61,
-      "learning_rate": 2.444710827324732e-05,
-      "loss": 0.2531,
+      "epoch": 2.6736223735899376,
+      "grad_norm": 0.2370181381702423,
+      "learning_rate": 2.3418843963328912e-05,
+      "loss": 0.3586,
       "step": 74185
     },
     {
-      "epoch": 2.61,
-      "learning_rate": 2.4444260252274116e-05,
-      "loss": 0.2748,
+      "epoch": 2.6738025732511623,
+      "grad_norm": 0.19188377261161804,
+      "learning_rate": 2.3415931682698427e-05,
+      "loss": 0.3616,
       "step": 74190
     },
     {
-      "epoch": 2.61,
-      "learning_rate": 2.4441412238516823e-05,
-      "loss": 0.2512,
+      "epoch": 2.673982772912387,
+      "grad_norm": 0.22248739004135132,
+      "learning_rate": 2.341301942365057e-05,
+      "loss": 0.4284,
       "step": 74195
     },
     {
-      "epoch": 2.61,
-      "learning_rate": 2.4438564232012412e-05,
-      "loss": 0.2867,
+      "epoch": 2.674162972573612,
+      "grad_norm": 0.2631896138191223,
+      "learning_rate": 2.3410107186225015e-05,
+      "loss": 0.4025,
       "step": 74200
     },
     {
-      "epoch": 2.61,
-      "learning_rate": 2.4435716232797882e-05,
-      "loss": 0.2848,
+      "epoch": 2.674343172234836,
+      "grad_norm": 0.19557088613510132,
+      "learning_rate": 2.3407194970461435e-05,
+      "loss": 0.4003,
       "step": 74205
     },
     {
-      "epoch": 2.61,
-      "learning_rate": 2.443286824091021e-05,
-      "loss": 0.2665,
+      "epoch": 2.674523371896061,
+      "grad_norm": 0.20168626308441162,
+      "learning_rate": 2.3404282776399523e-05,
+      "loss": 0.3501,
       "step": 74210
     },
     {
-      "epoch": 2.61,
-      "learning_rate": 2.443002025638637e-05,
-      "loss": 0.2785,
+      "epoch": 2.6747035715572856,
+      "grad_norm": 0.23849454522132874,
+      "learning_rate": 2.340137060407894e-05,
+      "loss": 0.3978,
       "step": 74215
     },
     {
-      "epoch": 2.61,
-      "learning_rate": 2.4427172279263333e-05,
-      "loss": 0.2872,
+      "epoch": 2.67488377121851,
+      "grad_norm": 0.2284935563802719,
+      "learning_rate": 2.3398458453539385e-05,
+      "loss": 0.3623,
       "step": 74220
     },
     {
-      "epoch": 2.61,
-      "learning_rate": 2.4424324309578096e-05,
-      "loss": 0.2626,
+      "epoch": 2.6750639708797346,
+      "grad_norm": 0.2417210191488266,
+      "learning_rate": 2.3395546324820526e-05,
+      "loss": 0.4028,
       "step": 74225
     },
     {
-      "epoch": 2.61,
-      "learning_rate": 2.4421476347367623e-05,
-      "loss": 0.2686,
+      "epoch": 2.6752441705409593,
+      "grad_norm": 0.18951061367988586,
+      "learning_rate": 2.3392634217962018e-05,
+      "loss": 0.3889,
       "step": 74230
     },
     {
-      "epoch": 2.61,
-      "learning_rate": 2.441862839266889e-05,
-      "loss": 0.2656,
+      "epoch": 2.675424370202184,
+      "grad_norm": 0.24638362228870392,
+      "learning_rate": 2.3389722133003577e-05,
+      "loss": 0.3558,
       "step": 74235
     },
     {
-      "epoch": 2.61,
-      "learning_rate": 2.44157804455189e-05,
-      "loss": 0.2722,
+      "epoch": 2.6756045698634088,
+      "grad_norm": 0.18396976590156555,
+      "learning_rate": 2.3386810069984856e-05,
+      "loss": 0.3805,
       "step": 74240
     },
     {
-      "epoch": 2.61,
-      "learning_rate": 2.441293250595461e-05,
-      "loss": 0.266,
+      "epoch": 2.6757847695246335,
+      "grad_norm": 0.20471645891666412,
+      "learning_rate": 2.3383898028945528e-05,
+      "loss": 0.4248,
       "step": 74245
     },
     {
-      "epoch": 2.61,
-      "learning_rate": 2.441008457401301e-05,
-      "loss": 0.2711,
+      "epoch": 2.675964969185858,
+      "grad_norm": 0.1948273777961731,
+      "learning_rate": 2.338098600992528e-05,
+      "loss": 0.3789,
       "step": 74250
     },
     {
-      "epoch": 2.61,
-      "learning_rate": 2.4407236649731056e-05,
-      "loss": 0.2852,
+      "epoch": 2.6761451688470825,
+      "grad_norm": 0.22328773140907288,
+      "learning_rate": 2.3378074012963783e-05,
+      "loss": 0.4053,
       "step": 74255
     },
     {
-      "epoch": 2.61,
-      "learning_rate": 2.4404388733145756e-05,
-      "loss": 0.2881,
+      "epoch": 2.6763253685083073,
+      "grad_norm": 0.25258857011795044,
+      "learning_rate": 2.3375162038100716e-05,
+      "loss": 0.4313,
       "step": 74260
     },
     {
-      "epoch": 2.61,
-      "learning_rate": 2.440154082429408e-05,
-      "loss": 0.2974,
+      "epoch": 2.676505568169532,
+      "grad_norm": 0.1960865557193756,
+      "learning_rate": 2.337225008537575e-05,
+      "loss": 0.3928,
       "step": 74265
     },
     {
-      "epoch": 2.61,
-      "learning_rate": 2.4398692923213e-05,
-      "loss": 0.294,
+      "epoch": 2.6766857678307563,
+      "grad_norm": 0.190299391746521,
+      "learning_rate": 2.3369338154828564e-05,
+      "loss": 0.3747,
       "step": 74270
     },
     {
-      "epoch": 2.61,
-      "learning_rate": 2.439584502993948e-05,
-      "loss": 0.2635,
+      "epoch": 2.676865967491981,
+      "grad_norm": 0.2319035530090332,
+      "learning_rate": 2.336642624649883e-05,
+      "loss": 0.3932,
       "step": 74275
     },
     {
-      "epoch": 2.61,
-      "learning_rate": 2.4392997144510533e-05,
-      "loss": 0.2973,
+      "epoch": 2.6770461671532058,
+      "grad_norm": 0.17021389305591583,
+      "learning_rate": 2.336351436042622e-05,
+      "loss": 0.362,
       "step": 74280
     },
     {
-      "epoch": 2.61,
-      "learning_rate": 2.439014926696311e-05,
-      "loss": 0.2617,
+      "epoch": 2.6772263668144305,
+      "grad_norm": 0.1885906457901001,
+      "learning_rate": 2.3360602496650406e-05,
+      "loss": 0.4038,
       "step": 74285
     },
     {
-      "epoch": 2.61,
-      "learning_rate": 2.4387301397334203e-05,
-      "loss": 0.2876,
+      "epoch": 2.6774065664756552,
+      "grad_norm": 0.2009781450033188,
+      "learning_rate": 2.3357690655211072e-05,
+      "loss": 0.4274,
       "step": 74290
     },
     {
-      "epoch": 2.61,
-      "learning_rate": 2.438445353566077e-05,
-      "loss": 0.247,
+      "epoch": 2.6775867661368795,
+      "grad_norm": 0.2083226591348648,
+      "learning_rate": 2.335477883614788e-05,
+      "loss": 0.4083,
       "step": 74295
     },
     {
-      "epoch": 2.61,
-      "learning_rate": 2.4381605681979817e-05,
-      "loss": 0.2682,
+      "epoch": 2.6777669657981042,
+      "grad_norm": 0.18861836194992065,
+      "learning_rate": 2.3351867039500513e-05,
+      "loss": 0.3834,
       "step": 74300
     },
     {
-      "epoch": 2.61,
-      "learning_rate": 2.4378757836328304e-05,
-      "loss": 0.2722,
+      "epoch": 2.677947165459329,
+      "grad_norm": 0.2695733904838562,
+      "learning_rate": 2.3348955265308642e-05,
+      "loss": 0.4182,
       "step": 74305
     },
     {
-      "epoch": 2.61,
-      "learning_rate": 2.4375909998743205e-05,
-      "loss": 0.259,
+      "epoch": 2.6781273651205537,
+      "grad_norm": 0.17389146983623505,
+      "learning_rate": 2.334604351361192e-05,
+      "loss": 0.4131,
       "step": 74310
     },
     {
-      "epoch": 2.61,
-      "learning_rate": 2.4373062169261498e-05,
-      "loss": 0.2706,
+      "epoch": 2.678307564781778,
+      "grad_norm": 0.18311482667922974,
+      "learning_rate": 2.3343131784450055e-05,
+      "loss": 0.4198,
       "step": 74315
     },
     {
-      "epoch": 2.61,
-      "learning_rate": 2.437021434792018e-05,
-      "loss": 0.2927,
+      "epoch": 2.6784877644430027,
+      "grad_norm": 0.2671164572238922,
+      "learning_rate": 2.334022007786269e-05,
+      "loss": 0.3622,
       "step": 74320
     },
     {
-      "epoch": 2.61,
-      "learning_rate": 2.4367366534756213e-05,
-      "loss": 0.282,
+      "epoch": 2.6786679641042275,
+      "grad_norm": 0.2175154834985733,
+      "learning_rate": 2.3337308393889493e-05,
+      "loss": 0.4046,
       "step": 74325
     },
     {
-      "epoch": 2.62,
-      "learning_rate": 2.4364518729806572e-05,
-      "loss": 0.2638,
+      "epoch": 2.678848163765452,
+      "grad_norm": 0.22409701347351074,
+      "learning_rate": 2.3334396732570167e-05,
+      "loss": 0.3906,
       "step": 74330
     },
     {
-      "epoch": 2.62,
-      "learning_rate": 2.4361670933108224e-05,
-      "loss": 0.2674,
+      "epoch": 2.679028363426677,
+      "grad_norm": 0.203602135181427,
+      "learning_rate": 2.3331485093944344e-05,
+      "loss": 0.4466,
       "step": 74335
     },
     {
-      "epoch": 2.62,
-      "learning_rate": 2.435882314469817e-05,
-      "loss": 0.2617,
+      "epoch": 2.6792085630879012,
+      "grad_norm": 0.19964513182640076,
+      "learning_rate": 2.332857347805173e-05,
+      "loss": 0.4134,
       "step": 74340
     },
     {
-      "epoch": 2.62,
-      "learning_rate": 2.4355975364613372e-05,
-      "loss": 0.2523,
+      "epoch": 2.679388762749126,
+      "grad_norm": 0.1884617805480957,
+      "learning_rate": 2.3325661884931972e-05,
+      "loss": 0.3872,
       "step": 74345
     },
     {
-      "epoch": 2.62,
-      "learning_rate": 2.4353127592890804e-05,
-      "loss": 0.2723,
+      "epoch": 2.6795689624103507,
+      "grad_norm": 0.2019912749528885,
+      "learning_rate": 2.3322750314624747e-05,
+      "loss": 0.4068,
       "step": 74350
     },
     {
-      "epoch": 2.62,
-      "learning_rate": 2.435027982956746e-05,
-      "loss": 0.2797,
+      "epoch": 2.6797491620715754,
+      "grad_norm": 0.2299565076828003,
+      "learning_rate": 2.3319838767169725e-05,
+      "loss": 0.4014,
       "step": 74355
     },
     {
-      "epoch": 2.62,
-      "learning_rate": 2.43474320746803e-05,
-      "loss": 0.2813,
+      "epoch": 2.6799293617327997,
+      "grad_norm": 0.24402816593647003,
+      "learning_rate": 2.3316927242606575e-05,
+      "loss": 0.3874,
       "step": 74360
     },
     {
-      "epoch": 2.62,
-      "learning_rate": 2.43445843282663e-05,
-      "loss": 0.2605,
+      "epoch": 2.6801095613940245,
+      "grad_norm": 0.25074100494384766,
+      "learning_rate": 2.331401574097496e-05,
+      "loss": 0.3988,
       "step": 74365
     },
     {
-      "epoch": 2.62,
-      "learning_rate": 2.4341736590362445e-05,
-      "loss": 0.2604,
+      "epoch": 2.680289761055249,
+      "grad_norm": 0.19405324757099152,
+      "learning_rate": 2.3311104262314563e-05,
+      "loss": 0.3689,
       "step": 74370
     },
     {
-      "epoch": 2.62,
-      "learning_rate": 2.43388888610057e-05,
-      "loss": 0.2687,
+      "epoch": 2.680469960716474,
+      "grad_norm": 0.19997498393058777,
+      "learning_rate": 2.3308192806665034e-05,
+      "loss": 0.3886,
       "step": 74375
     },
     {
-      "epoch": 2.62,
-      "learning_rate": 2.4336041140233058e-05,
-      "loss": 0.2694,
+      "epoch": 2.6806501603776987,
+      "grad_norm": 0.22032709419727325,
+      "learning_rate": 2.3305281374066057e-05,
+      "loss": 0.4064,
       "step": 74380
     },
     {
-      "epoch": 2.62,
-      "learning_rate": 2.4333193428081483e-05,
-      "loss": 0.2787,
+      "epoch": 2.6808303600389234,
+      "grad_norm": 0.25573626160621643,
+      "learning_rate": 2.3302369964557292e-05,
+      "loss": 0.3774,
       "step": 74385
     },
     {
-      "epoch": 2.62,
-      "learning_rate": 2.4330345724587937e-05,
-      "loss": 0.2734,
+      "epoch": 2.6810105597001477,
+      "grad_norm": 0.23341487348079681,
+      "learning_rate": 2.3299458578178403e-05,
+      "loss": 0.3918,
       "step": 74390
     },
     {
-      "epoch": 2.62,
-      "learning_rate": 2.4327498029789422e-05,
-      "loss": 0.2889,
+      "epoch": 2.6811907593613724,
+      "grad_norm": 0.19004227221012115,
+      "learning_rate": 2.3296547214969066e-05,
+      "loss": 0.4126,
       "step": 74395
     },
     {
-      "epoch": 2.62,
-      "learning_rate": 2.4324650343722905e-05,
-      "loss": 0.2638,
+      "epoch": 2.681370959022597,
+      "grad_norm": 0.18891394138336182,
+      "learning_rate": 2.3293635874968954e-05,
+      "loss": 0.3896,
       "step": 74400
     },
     {
-      "epoch": 2.62,
-      "learning_rate": 2.4321802666425357e-05,
-      "loss": 0.2766,
+      "epoch": 2.6815511586838214,
+      "grad_norm": 0.20582923293113708,
+      "learning_rate": 2.32907245582177e-05,
+      "loss": 0.433,
       "step": 74405
     },
     {
-      "epoch": 2.62,
-      "learning_rate": 2.4318954997933743e-05,
-      "loss": 0.2545,
+      "epoch": 2.681731358345046,
+      "grad_norm": 0.22365623712539673,
+      "learning_rate": 2.328781326475501e-05,
+      "loss": 0.3804,
       "step": 74410
     },
     {
-      "epoch": 2.62,
-      "learning_rate": 2.4316107338285062e-05,
-      "loss": 0.2671,
+      "epoch": 2.681911558006271,
+      "grad_norm": 0.2613740563392639,
+      "learning_rate": 2.328490199462052e-05,
+      "loss": 0.3879,
       "step": 74415
     },
     {
-      "epoch": 2.62,
-      "learning_rate": 2.431325968751627e-05,
-      "loss": 0.2798,
+      "epoch": 2.6820917576674956,
+      "grad_norm": 0.16647237539291382,
+      "learning_rate": 2.3281990747853925e-05,
+      "loss": 0.3906,
       "step": 74420
     },
     {
-      "epoch": 2.62,
-      "learning_rate": 2.431041204566436e-05,
-      "loss": 0.2563,
+      "epoch": 2.6822719573287204,
+      "grad_norm": 0.21178102493286133,
+      "learning_rate": 2.3279079524494864e-05,
+      "loss": 0.4273,
       "step": 74425
     },
     {
-      "epoch": 2.62,
-      "learning_rate": 2.4307564412766276e-05,
-      "loss": 0.2845,
+      "epoch": 2.682452156989945,
+      "grad_norm": 0.22928111255168915,
+      "learning_rate": 2.327616832458301e-05,
+      "loss": 0.3926,
       "step": 74430
     },
     {
-      "epoch": 2.62,
-      "learning_rate": 2.4304716788859024e-05,
-      "loss": 0.2695,
+      "epoch": 2.6826323566511694,
+      "grad_norm": 0.29340752959251404,
+      "learning_rate": 2.327325714815803e-05,
+      "loss": 0.3821,
       "step": 74435
     },
     {
-      "epoch": 2.62,
-      "learning_rate": 2.4301869173979568e-05,
-      "loss": 0.2853,
+      "epoch": 2.682812556312394,
+      "grad_norm": 0.22144730389118195,
+      "learning_rate": 2.3270345995259586e-05,
+      "loss": 0.4207,
       "step": 74440
     },
     {
-      "epoch": 2.62,
-      "learning_rate": 2.4299021568164876e-05,
-      "loss": 0.2705,
+      "epoch": 2.682992755973619,
+      "grad_norm": 0.2856822907924652,
+      "learning_rate": 2.3267434865927345e-05,
+      "loss": 0.3741,
       "step": 74445
     },
     {
-      "epoch": 2.62,
-      "learning_rate": 2.429617397145193e-05,
-      "loss": 0.2638,
+      "epoch": 2.683172955634843,
+      "grad_norm": 0.16166545450687408,
+      "learning_rate": 2.326452376020097e-05,
+      "loss": 0.3838,
       "step": 74450
     },
     {
-      "epoch": 2.62,
-      "learning_rate": 2.4293326383877694e-05,
-      "loss": 0.2754,
+      "epoch": 2.683353155296068,
+      "grad_norm": 0.21917329728603363,
+      "learning_rate": 2.3261612678120118e-05,
+      "loss": 0.4113,
       "step": 74455
     },
     {
-      "epoch": 2.62,
-      "learning_rate": 2.4290478805479158e-05,
-      "loss": 0.2483,
+      "epoch": 2.6835333549572926,
+      "grad_norm": 0.25646519660949707,
+      "learning_rate": 2.325870161972446e-05,
+      "loss": 0.4198,
       "step": 74460
     },
     {
-      "epoch": 2.62,
-      "learning_rate": 2.4287631236293287e-05,
-      "loss": 0.25,
+      "epoch": 2.6837135546185173,
+      "grad_norm": 0.18003955483436584,
+      "learning_rate": 2.3255790585053654e-05,
+      "loss": 0.4351,
       "step": 74465
     },
     {
-      "epoch": 2.62,
-      "learning_rate": 2.4284783676357044e-05,
-      "loss": 0.2723,
+      "epoch": 2.683893754279742,
+      "grad_norm": 0.2252659797668457,
+      "learning_rate": 2.3252879574147363e-05,
+      "loss": 0.3775,
       "step": 74470
     },
     {
-      "epoch": 2.62,
-      "learning_rate": 2.4281936125707418e-05,
-      "loss": 0.2838,
+      "epoch": 2.684073953940967,
+      "grad_norm": 0.21669524908065796,
+      "learning_rate": 2.3249968587045253e-05,
+      "loss": 0.4371,
       "step": 74475
     },
     {
-      "epoch": 2.62,
-      "learning_rate": 2.4279088584381386e-05,
-      "loss": 0.2648,
+      "epoch": 2.684254153602191,
+      "grad_norm": 0.2279970496892929,
+      "learning_rate": 2.3247057623786974e-05,
+      "loss": 0.4017,
       "step": 74480
     },
     {
-      "epoch": 2.62,
-      "learning_rate": 2.4276241052415898e-05,
-      "loss": 0.2948,
+      "epoch": 2.684434353263416,
+      "grad_norm": 0.1842222362756729,
+      "learning_rate": 2.3244146684412205e-05,
+      "loss": 0.3811,
       "step": 74485
     },
     {
-      "epoch": 2.62,
-      "learning_rate": 2.427339352984796e-05,
-      "loss": 0.2717,
+      "epoch": 2.6846145529246406,
+      "grad_norm": 0.24382725358009338,
+      "learning_rate": 2.3241235768960595e-05,
+      "loss": 0.423,
       "step": 74490
     },
     {
-      "epoch": 2.62,
-      "learning_rate": 2.4270546016714525e-05,
-      "loss": 0.2738,
+      "epoch": 2.684794752585865,
+      "grad_norm": 0.24431829154491425,
+      "learning_rate": 2.3238324877471804e-05,
+      "loss": 0.3774,
       "step": 74495
     },
     {
-      "epoch": 2.62,
-      "learning_rate": 2.4267698513052565e-05,
-      "loss": 0.2835,
+      "epoch": 2.6849749522470896,
+      "grad_norm": 0.22015175223350525,
+      "learning_rate": 2.3235414009985498e-05,
+      "loss": 0.4201,
       "step": 74500
     },
     {
-      "epoch": 2.62,
-      "eval_loss": 0.2619902491569519,
-      "eval_runtime": 10.556,
-      "eval_samples_per_second": 9.473,
-      "eval_steps_per_second": 9.473,
+      "epoch": 2.6849749522470896,
+      "eval_loss": 0.4314359724521637,
+      "eval_runtime": 3.5281,
+      "eval_samples_per_second": 28.344,
+      "eval_steps_per_second": 7.086,
       "step": 74500
     },
     {
-      "epoch": 2.62,
-      "learning_rate": 2.4264851018899058e-05,
-      "loss": 0.2681,
+      "epoch": 2.6851551519083143,
+      "grad_norm": 0.19529901444911957,
+      "learning_rate": 2.323250316654134e-05,
+      "loss": 0.4097,
       "step": 74505
     },
     {
-      "epoch": 2.62,
-      "learning_rate": 2.4262003534290972e-05,
-      "loss": 0.254,
+      "epoch": 2.685335351569539,
+      "grad_norm": 0.1773121953010559,
+      "learning_rate": 2.322959234717897e-05,
+      "loss": 0.4016,
       "step": 74510
     },
     {
-      "epoch": 2.62,
-      "learning_rate": 2.4259156059265292e-05,
-      "loss": 0.2792,
+      "epoch": 2.685515551230764,
+      "grad_norm": 0.24559174478054047,
+      "learning_rate": 2.322668155193808e-05,
+      "loss": 0.3864,
       "step": 74515
     },
     {
-      "epoch": 2.62,
-      "learning_rate": 2.4256308593858982e-05,
-      "loss": 0.2778,
+      "epoch": 2.6856957508919885,
+      "grad_norm": 0.214651420712471,
+      "learning_rate": 2.322377078085829e-05,
+      "loss": 0.4201,
       "step": 74520
     },
     {
-      "epoch": 2.62,
-      "learning_rate": 2.4253461138109004e-05,
-      "loss": 0.2737,
+      "epoch": 2.685875950553213,
+      "grad_norm": 0.19618816673755646,
+      "learning_rate": 2.3220860033979296e-05,
+      "loss": 0.4003,
       "step": 74525
     },
     {
-      "epoch": 2.62,
-      "learning_rate": 2.4250613692052352e-05,
-      "loss": 0.2618,
+      "epoch": 2.6860561502144376,
+      "grad_norm": 0.22414274513721466,
+      "learning_rate": 2.3217949311340733e-05,
+      "loss": 0.3972,
       "step": 74530
     },
     {
-      "epoch": 2.62,
-      "learning_rate": 2.424776625572598e-05,
-      "loss": 0.2901,
+      "epoch": 2.6862363498756623,
+      "grad_norm": 0.1811375916004181,
+      "learning_rate": 2.3215038612982265e-05,
+      "loss": 0.4129,
       "step": 74535
     },
     {
-      "epoch": 2.62,
-      "learning_rate": 2.4244918829166876e-05,
-      "loss": 0.275,
+      "epoch": 2.686416549536887,
+      "grad_norm": 0.21368905901908875,
+      "learning_rate": 2.3212127938943552e-05,
+      "loss": 0.3923,
       "step": 74540
     },
     {
-      "epoch": 2.62,
-      "learning_rate": 2.4242071412411994e-05,
-      "loss": 0.2533,
+      "epoch": 2.6865967491981113,
+      "grad_norm": 0.2091556340456009,
+      "learning_rate": 2.320921728926425e-05,
+      "loss": 0.3873,
       "step": 74545
     },
     {
-      "epoch": 2.62,
-      "learning_rate": 2.423922400549832e-05,
-      "loss": 0.2717,
+      "epoch": 2.686776948859336,
+      "grad_norm": 0.246855229139328,
+      "learning_rate": 2.3206306663984013e-05,
+      "loss": 0.4176,
       "step": 74550
     },
     {
-      "epoch": 2.62,
-      "learning_rate": 2.423637660846282e-05,
-      "loss": 0.2673,
+      "epoch": 2.686957148520561,
+      "grad_norm": 0.20908492803573608,
+      "learning_rate": 2.3203396063142503e-05,
+      "loss": 0.3964,
       "step": 74555
     },
     {
-      "epoch": 2.62,
-      "learning_rate": 2.4233529221342473e-05,
-      "loss": 0.2636,
+      "epoch": 2.6871373481817855,
+      "grad_norm": 0.21159091591835022,
+      "learning_rate": 2.3200485486779367e-05,
+      "loss": 0.3901,
       "step": 74560
     },
     {
-      "epoch": 2.62,
-      "learning_rate": 2.4230681844174228e-05,
-      "loss": 0.2511,
+      "epoch": 2.6873175478430102,
+      "grad_norm": 0.20965726673603058,
+      "learning_rate": 2.3197574934934274e-05,
+      "loss": 0.3859,
       "step": 74565
     },
     {
-      "epoch": 2.62,
-      "learning_rate": 2.4227834476995086e-05,
-      "loss": 0.2544,
+      "epoch": 2.6874977475042345,
+      "grad_norm": 0.2037397176027298,
+      "learning_rate": 2.3194664407646876e-05,
+      "loss": 0.3793,
       "step": 74570
     },
     {
-      "epoch": 2.62,
-      "learning_rate": 2.4224987119842006e-05,
-      "loss": 0.276,
+      "epoch": 2.6876779471654593,
+      "grad_norm": 0.21732820570468903,
+      "learning_rate": 2.319175390495682e-05,
+      "loss": 0.4485,
       "step": 74575
     },
     {
-      "epoch": 2.62,
-      "learning_rate": 2.4222139772751952e-05,
-      "loss": 0.2681,
+      "epoch": 2.687858146826684,
+      "grad_norm": 0.19209305942058563,
+      "learning_rate": 2.3188843426903774e-05,
+      "loss": 0.3827,
       "step": 74580
     },
     {
-      "epoch": 2.62,
-      "learning_rate": 2.4219292435761895e-05,
-      "loss": 0.2883,
+      "epoch": 2.6880383464879087,
+      "grad_norm": 0.2674555480480194,
+      "learning_rate": 2.3185932973527386e-05,
+      "loss": 0.4225,
       "step": 74585
     },
     {
-      "epoch": 2.62,
-      "learning_rate": 2.4216445108908817e-05,
-      "loss": 0.2474,
+      "epoch": 2.688218546149133,
+      "grad_norm": 0.1785704642534256,
+      "learning_rate": 2.3183022544867298e-05,
+      "loss": 0.4047,
       "step": 74590
     },
     {
-      "epoch": 2.62,
-      "learning_rate": 2.4213597792229685e-05,
-      "loss": 0.2678,
+      "epoch": 2.6883987458103578,
+      "grad_norm": 0.18283165991306305,
+      "learning_rate": 2.318011214096319e-05,
+      "loss": 0.3662,
       "step": 74595
     },
     {
-      "epoch": 2.62,
-      "learning_rate": 2.4210750485761457e-05,
-      "loss": 0.271,
+      "epoch": 2.6885789454715825,
+      "grad_norm": 0.22287532687187195,
+      "learning_rate": 2.3177201761854686e-05,
+      "loss": 0.4041,
       "step": 74600
     },
     {
-      "epoch": 2.62,
-      "learning_rate": 2.4207903189541126e-05,
-      "loss": 0.2604,
+      "epoch": 2.6887591451328072,
+      "grad_norm": 0.2034824639558792,
+      "learning_rate": 2.317429140758147e-05,
+      "loss": 0.3986,
       "step": 74605
     },
     {
-      "epoch": 2.62,
-      "learning_rate": 2.420505590360565e-05,
-      "loss": 0.2676,
+      "epoch": 2.688939344794032,
+      "grad_norm": 0.23879247903823853,
+      "learning_rate": 2.317138107818318e-05,
+      "loss": 0.3647,
       "step": 74610
     },
     {
-      "epoch": 2.63,
-      "learning_rate": 2.4202208627991994e-05,
-      "loss": 0.2623,
+      "epoch": 2.6891195444552567,
+      "grad_norm": 0.2760174572467804,
+      "learning_rate": 2.3168470773699452e-05,
+      "loss": 0.3891,
       "step": 74615
     },
     {
-      "epoch": 2.63,
-      "learning_rate": 2.4199361362737127e-05,
-      "loss": 0.2597,
+      "epoch": 2.689299744116481,
+      "grad_norm": 0.20346973836421967,
+      "learning_rate": 2.3165560494169973e-05,
+      "loss": 0.4036,
       "step": 74620
     },
     {
-      "epoch": 2.63,
-      "learning_rate": 2.4196514107878038e-05,
-      "loss": 0.254,
+      "epoch": 2.6894799437777057,
+      "grad_norm": 0.27176687121391296,
+      "learning_rate": 2.3162650239634363e-05,
+      "loss": 0.4122,
       "step": 74625
     },
     {
-      "epoch": 2.63,
-      "learning_rate": 2.419366686345168e-05,
-      "loss": 0.2658,
+      "epoch": 2.6896601434389305,
+      "grad_norm": 0.232636496424675,
+      "learning_rate": 2.3159740010132304e-05,
+      "loss": 0.4164,
       "step": 74630
     },
     {
-      "epoch": 2.63,
-      "learning_rate": 2.419081962949503e-05,
-      "loss": 0.2685,
+      "epoch": 2.6898403431001547,
+      "grad_norm": 0.1852894425392151,
+      "learning_rate": 2.3156829805703422e-05,
+      "loss": 0.4268,
       "step": 74635
     },
     {
-      "epoch": 2.63,
-      "learning_rate": 2.418797240604504e-05,
-      "loss": 0.2688,
+      "epoch": 2.6900205427613795,
+      "grad_norm": 0.1811005026102066,
+      "learning_rate": 2.3153919626387374e-05,
+      "loss": 0.4163,
       "step": 74640
     },
     {
-      "epoch": 2.63,
-      "learning_rate": 2.41851251931387e-05,
-      "loss": 0.2555,
+      "epoch": 2.690200742422604,
+      "grad_norm": 0.22391211986541748,
+      "learning_rate": 2.315100947222382e-05,
+      "loss": 0.3755,
       "step": 74645
     },
     {
-      "epoch": 2.63,
-      "learning_rate": 2.4182277990812976e-05,
-      "loss": 0.2566,
+      "epoch": 2.690380942083829,
+      "grad_norm": 0.19471041858196259,
+      "learning_rate": 2.31480993432524e-05,
+      "loss": 0.3779,
       "step": 74650
     },
     {
-      "epoch": 2.63,
-      "learning_rate": 2.4179430799104834e-05,
-      "loss": 0.2696,
+      "epoch": 2.6905611417450537,
+      "grad_norm": 0.21801535785198212,
+      "learning_rate": 2.3145189239512765e-05,
+      "loss": 0.3664,
       "step": 74655
     },
     {
-      "epoch": 2.63,
-      "learning_rate": 2.4176583618051236e-05,
-      "loss": 0.2613,
+      "epoch": 2.6907413414062784,
+      "grad_norm": 0.21162468194961548,
+      "learning_rate": 2.3142279161044575e-05,
+      "loss": 0.3548,
       "step": 74660
     },
     {
-      "epoch": 2.63,
-      "learning_rate": 2.4173736447689165e-05,
-      "loss": 0.2532,
+      "epoch": 2.6909215410675027,
+      "grad_norm": 0.20754042267799377,
+      "learning_rate": 2.3139369107887467e-05,
+      "loss": 0.3827,
       "step": 74665
     },
     {
-      "epoch": 2.63,
-      "learning_rate": 2.4170889288055572e-05,
-      "loss": 0.2818,
+      "epoch": 2.6911017407287274,
+      "grad_norm": 0.175797700881958,
+      "learning_rate": 2.3136459080081096e-05,
+      "loss": 0.3674,
       "step": 74670
     },
     {
-      "epoch": 2.63,
-      "learning_rate": 2.4168042139187447e-05,
-      "loss": 0.2913,
+      "epoch": 2.691281940389952,
+      "grad_norm": 0.19099979102611542,
+      "learning_rate": 2.3133549077665114e-05,
+      "loss": 0.411,
       "step": 74675
     },
     {
-      "epoch": 2.63,
-      "learning_rate": 2.4165195001121732e-05,
-      "loss": 0.2811,
+      "epoch": 2.6914621400511765,
+      "grad_norm": 0.20772726833820343,
+      "learning_rate": 2.313063910067915e-05,
+      "loss": 0.3753,
       "step": 74680
     },
     {
-      "epoch": 2.63,
-      "learning_rate": 2.416234787389542e-05,
-      "loss": 0.2603,
+      "epoch": 2.691642339712401,
+      "grad_norm": 0.24013733863830566,
+      "learning_rate": 2.312772914916288e-05,
+      "loss": 0.3797,
       "step": 74685
     },
     {
-      "epoch": 2.63,
-      "learning_rate": 2.4159500757545473e-05,
-      "loss": 0.2629,
+      "epoch": 2.691822539373626,
+      "grad_norm": 0.191719189286232,
+      "learning_rate": 2.312481922315594e-05,
+      "loss": 0.3634,
       "step": 74690
     },
     {
-      "epoch": 2.63,
-      "learning_rate": 2.4156653652108847e-05,
-      "loss": 0.2686,
+      "epoch": 2.6920027390348507,
+      "grad_norm": 0.1952929049730301,
+      "learning_rate": 2.312190932269796e-05,
+      "loss": 0.374,
       "step": 74695
     },
     {
-      "epoch": 2.63,
-      "learning_rate": 2.415380655762251e-05,
-      "loss": 0.2839,
+      "epoch": 2.6921829386960754,
+      "grad_norm": 0.19985808432102203,
+      "learning_rate": 2.3118999447828617e-05,
+      "loss": 0.3761,
       "step": 74700
     },
     {
-      "epoch": 2.63,
-      "learning_rate": 2.4150959474123454e-05,
-      "loss": 0.2787,
+      "epoch": 2.6923631383573,
+      "grad_norm": 0.22037459909915924,
+      "learning_rate": 2.311608959858753e-05,
+      "loss": 0.3537,
       "step": 74705
     },
     {
-      "epoch": 2.63,
-      "learning_rate": 2.4148112401648624e-05,
-      "loss": 0.2747,
+      "epoch": 2.6925433380185244,
+      "grad_norm": 0.25286975502967834,
+      "learning_rate": 2.311317977501437e-05,
+      "loss": 0.3598,
       "step": 74710
     },
     {
-      "epoch": 2.63,
-      "learning_rate": 2.4145265340234983e-05,
-      "loss": 0.2692,
+      "epoch": 2.692723537679749,
+      "grad_norm": 0.21672144532203674,
+      "learning_rate": 2.3110269977148765e-05,
+      "loss": 0.3825,
       "step": 74715
     },
     {
-      "epoch": 2.63,
-      "learning_rate": 2.414241828991952e-05,
-      "loss": 0.2783,
+      "epoch": 2.692903737340974,
+      "grad_norm": 0.2222987562417984,
+      "learning_rate": 2.310736020503036e-05,
+      "loss": 0.3665,
       "step": 74720
     },
     {
-      "epoch": 2.63,
-      "learning_rate": 2.4139571250739183e-05,
-      "loss": 0.2737,
+      "epoch": 2.693083937002198,
+      "grad_norm": 0.1981443464756012,
+      "learning_rate": 2.3104450458698816e-05,
+      "loss": 0.4095,
       "step": 74725
     },
     {
-      "epoch": 2.63,
-      "learning_rate": 2.4136724222730957e-05,
-      "loss": 0.2638,
+      "epoch": 2.693264136663423,
+      "grad_norm": 0.19497603178024292,
+      "learning_rate": 2.3101540738193762e-05,
+      "loss": 0.4022,
       "step": 74730
     },
     {
-      "epoch": 2.63,
-      "learning_rate": 2.413387720593178e-05,
-      "loss": 0.2488,
+      "epoch": 2.6934443363246476,
+      "grad_norm": 0.20269820094108582,
+      "learning_rate": 2.3098631043554845e-05,
+      "loss": 0.3861,
       "step": 74735
     },
     {
-      "epoch": 2.63,
-      "learning_rate": 2.4131030200378653e-05,
-      "loss": 0.2688,
+      "epoch": 2.6936245359858724,
+      "grad_norm": 0.20779010653495789,
+      "learning_rate": 2.3095721374821716e-05,
+      "loss": 0.3832,
       "step": 74740
     },
     {
-      "epoch": 2.63,
-      "learning_rate": 2.4128183206108524e-05,
-      "loss": 0.2412,
+      "epoch": 2.693804735647097,
+      "grad_norm": 0.18731550872325897,
+      "learning_rate": 2.309281173203401e-05,
+      "loss": 0.3904,
       "step": 74745
     },
     {
-      "epoch": 2.63,
-      "learning_rate": 2.4125336223158358e-05,
-      "loss": 0.2759,
+      "epoch": 2.693984935308322,
+      "grad_norm": 0.2547980844974518,
+      "learning_rate": 2.3089902115231378e-05,
+      "loss": 0.3837,
       "step": 74750
     },
     {
-      "epoch": 2.63,
-      "learning_rate": 2.4122489251565117e-05,
-      "loss": 0.302,
+      "epoch": 2.694165134969546,
+      "grad_norm": 0.20064884424209595,
+      "learning_rate": 2.3086992524453462e-05,
+      "loss": 0.3908,
       "step": 74755
     },
     {
-      "epoch": 2.63,
-      "learning_rate": 2.4119642291365785e-05,
-      "loss": 0.2656,
+      "epoch": 2.694345334630771,
+      "grad_norm": 0.1967686116695404,
+      "learning_rate": 2.30840829597399e-05,
+      "loss": 0.423,
       "step": 74760
     },
     {
-      "epoch": 2.63,
-      "learning_rate": 2.411679534259732e-05,
-      "loss": 0.2767,
+      "epoch": 2.6945255342919956,
+      "grad_norm": 0.21911115944385529,
+      "learning_rate": 2.3081173421130336e-05,
+      "loss": 0.4434,
       "step": 74765
     },
     {
-      "epoch": 2.63,
-      "learning_rate": 2.4113948405296684e-05,
-      "loss": 0.2715,
+      "epoch": 2.6947057339532203,
+      "grad_norm": 0.20522983372211456,
+      "learning_rate": 2.307826390866442e-05,
+      "loss": 0.403,
       "step": 74770
     },
     {
-      "epoch": 2.63,
-      "learning_rate": 2.4111101479500835e-05,
-      "loss": 0.2581,
+      "epoch": 2.6948859336144446,
+      "grad_norm": 0.19616533815860748,
+      "learning_rate": 2.307535442238177e-05,
+      "loss": 0.3686,
       "step": 74775
     },
     {
-      "epoch": 2.63,
-      "learning_rate": 2.410825456524675e-05,
-      "loss": 0.2648,
+      "epoch": 2.6950661332756694,
+      "grad_norm": 0.2152736634016037,
+      "learning_rate": 2.3072444962322056e-05,
+      "loss": 0.4022,
       "step": 74780
     },
     {
-      "epoch": 2.63,
-      "learning_rate": 2.41054076625714e-05,
-      "loss": 0.2657,
+      "epoch": 2.695246332936894,
+      "grad_norm": 0.17979800701141357,
+      "learning_rate": 2.3069535528524902e-05,
+      "loss": 0.3845,
       "step": 74785
     },
     {
-      "epoch": 2.63,
-      "learning_rate": 2.410256077151174e-05,
-      "loss": 0.2561,
+      "epoch": 2.695426532598119,
+      "grad_norm": 0.2075699120759964,
+      "learning_rate": 2.3066626121029954e-05,
+      "loss": 0.4106,
       "step": 74790
     },
     {
-      "epoch": 2.63,
-      "learning_rate": 2.4099713892104727e-05,
-      "loss": 0.2544,
+      "epoch": 2.6956067322593436,
+      "grad_norm": 0.1615006923675537,
+      "learning_rate": 2.306371673987686e-05,
+      "loss": 0.3906,
       "step": 74795
     },
     {
-      "epoch": 2.63,
-      "learning_rate": 2.4096867024387345e-05,
-      "loss": 0.2728,
+      "epoch": 2.695786931920568,
+      "grad_norm": 0.2201845496892929,
+      "learning_rate": 2.306080738510523e-05,
+      "loss": 0.3822,
       "step": 74800
     },
     {
-      "epoch": 2.63,
-      "learning_rate": 2.4094020168396548e-05,
-      "loss": 0.2933,
+      "epoch": 2.6959671315817926,
+      "grad_norm": 0.17273396253585815,
+      "learning_rate": 2.3057898056754744e-05,
+      "loss": 0.401,
       "step": 74805
     },
     {
-      "epoch": 2.63,
-      "learning_rate": 2.4091173324169304e-05,
-      "loss": 0.2865,
+      "epoch": 2.6961473312430173,
+      "grad_norm": 0.1817549616098404,
+      "learning_rate": 2.3054988754865015e-05,
+      "loss": 0.3613,
       "step": 74810
     },
     {
-      "epoch": 2.63,
-      "learning_rate": 2.4088326491742567e-05,
-      "loss": 0.2505,
+      "epoch": 2.696327530904242,
+      "grad_norm": 0.18268698453903198,
+      "learning_rate": 2.3052079479475683e-05,
+      "loss": 0.4106,
       "step": 74815
     },
     {
-      "epoch": 2.63,
-      "learning_rate": 2.408547967115332e-05,
-      "loss": 0.2515,
+      "epoch": 2.6965077305654663,
+      "grad_norm": 0.22823894023895264,
+      "learning_rate": 2.3049170230626395e-05,
+      "loss": 0.3843,
       "step": 74820
     },
     {
-      "epoch": 2.63,
-      "learning_rate": 2.4082632862438514e-05,
-      "loss": 0.2667,
+      "epoch": 2.696687930226691,
+      "grad_norm": 0.22048458456993103,
+      "learning_rate": 2.304626100835678e-05,
+      "loss": 0.4166,
       "step": 74825
     },
     {
-      "epoch": 2.63,
-      "learning_rate": 2.4079786065635107e-05,
-      "loss": 0.2822,
+      "epoch": 2.696868129887916,
+      "grad_norm": 0.22666330635547638,
+      "learning_rate": 2.3043351812706486e-05,
+      "loss": 0.3873,
       "step": 74830
     },
     {
-      "epoch": 2.63,
-      "learning_rate": 2.4076939280780075e-05,
-      "loss": 0.261,
+      "epoch": 2.6970483295491405,
+      "grad_norm": 0.22461573779582977,
+      "learning_rate": 2.3040442643715142e-05,
+      "loss": 0.4052,
       "step": 74835
     },
     {
-      "epoch": 2.63,
-      "learning_rate": 2.4074092507910386e-05,
-      "loss": 0.2563,
+      "epoch": 2.6972285292103653,
+      "grad_norm": 0.2268482744693756,
+      "learning_rate": 2.3037533501422384e-05,
+      "loss": 0.4011,
       "step": 74840
     },
     {
-      "epoch": 2.63,
-      "learning_rate": 2.4071245747062996e-05,
-      "loss": 0.2818,
+      "epoch": 2.6974087288715896,
+      "grad_norm": 0.2219397872686386,
+      "learning_rate": 2.303462438586786e-05,
+      "loss": 0.387,
       "step": 74845
     },
     {
-      "epoch": 2.63,
-      "learning_rate": 2.4068398998274854e-05,
-      "loss": 0.2844,
+      "epoch": 2.6975889285328143,
+      "grad_norm": 0.21480488777160645,
+      "learning_rate": 2.3031715297091188e-05,
+      "loss": 0.4225,
       "step": 74850
     },
     {
-      "epoch": 2.63,
-      "learning_rate": 2.4065552261582952e-05,
-      "loss": 0.2628,
+      "epoch": 2.697769128194039,
+      "grad_norm": 0.1800355166196823,
+      "learning_rate": 2.302880623513202e-05,
+      "loss": 0.3738,
       "step": 74855
     },
     {
-      "epoch": 2.63,
-      "learning_rate": 2.406270553702423e-05,
-      "loss": 0.2968,
+      "epoch": 2.6979493278552638,
+      "grad_norm": 0.2048688679933548,
+      "learning_rate": 2.302589720002999e-05,
+      "loss": 0.3953,
       "step": 74860
     },
     {
-      "epoch": 2.63,
-      "learning_rate": 2.4059858824635663e-05,
-      "loss": 0.263,
+      "epoch": 2.698129527516488,
+      "grad_norm": 0.19619779288768768,
+      "learning_rate": 2.302298819182472e-05,
+      "loss": 0.3967,
       "step": 74865
     },
     {
-      "epoch": 2.63,
-      "learning_rate": 2.4057012124454205e-05,
-      "loss": 0.2644,
+      "epoch": 2.698309727177713,
+      "grad_norm": 0.22370363771915436,
+      "learning_rate": 2.302007921055586e-05,
+      "loss": 0.4012,
       "step": 74870
     },
     {
-      "epoch": 2.63,
-      "learning_rate": 2.4054165436516825e-05,
-      "loss": 0.256,
+      "epoch": 2.6984899268389375,
+      "grad_norm": 0.2369144707918167,
+      "learning_rate": 2.301717025626304e-05,
+      "loss": 0.3758,
       "step": 74875
     },
     {
-      "epoch": 2.63,
-      "learning_rate": 2.4051318760860488e-05,
-      "loss": 0.2898,
+      "epoch": 2.6986701265001622,
+      "grad_norm": 0.18146106600761414,
+      "learning_rate": 2.301426132898588e-05,
+      "loss": 0.4064,
       "step": 74880
     },
     {
-      "epoch": 2.63,
-      "learning_rate": 2.4048472097522148e-05,
-      "loss": 0.2598,
+      "epoch": 2.698850326161387,
+      "grad_norm": 0.2013833373785019,
+      "learning_rate": 2.301135242876404e-05,
+      "loss": 0.3753,
       "step": 74885
     },
     {
-      "epoch": 2.63,
-      "learning_rate": 2.404562544653877e-05,
-      "loss": 0.2734,
+      "epoch": 2.6990305258226117,
+      "grad_norm": 0.18798790872097015,
+      "learning_rate": 2.3008443555637116e-05,
+      "loss": 0.4047,
       "step": 74890
     },
     {
-      "epoch": 2.64,
-      "learning_rate": 2.4042778807947313e-05,
-      "loss": 0.2798,
+      "epoch": 2.699210725483836,
+      "grad_norm": 0.23676003515720367,
+      "learning_rate": 2.3005534709644784e-05,
+      "loss": 0.3704,
       "step": 74895
     },
     {
-      "epoch": 2.64,
-      "learning_rate": 2.4039932181784754e-05,
-      "loss": 0.272,
+      "epoch": 2.6993909251450607,
+      "grad_norm": 0.20541468262672424,
+      "learning_rate": 2.300262589082665e-05,
+      "loss": 0.3796,
       "step": 74900
     },
     {
-      "epoch": 2.64,
-      "learning_rate": 2.403708556808804e-05,
-      "loss": 0.2886,
+      "epoch": 2.6995711248062855,
+      "grad_norm": 0.2497328221797943,
+      "learning_rate": 2.299971709922234e-05,
+      "loss": 0.3954,
       "step": 74905
     },
     {
-      "epoch": 2.64,
-      "learning_rate": 2.403423896689412e-05,
-      "loss": 0.2677,
+      "epoch": 2.6997513244675098,
+      "grad_norm": 0.2306727170944214,
+      "learning_rate": 2.2996808334871513e-05,
+      "loss": 0.4071,
       "step": 74910
     },
     {
-      "epoch": 2.64,
-      "learning_rate": 2.403139237823998e-05,
-      "loss": 0.2875,
+      "epoch": 2.6999315241287345,
+      "grad_norm": 0.2239367663860321,
+      "learning_rate": 2.2993899597813778e-05,
+      "loss": 0.3672,
       "step": 74915
     },
     {
-      "epoch": 2.64,
-      "learning_rate": 2.4028545802162578e-05,
-      "loss": 0.2836,
+      "epoch": 2.7001117237899592,
+      "grad_norm": 0.19781452417373657,
+      "learning_rate": 2.299099088808877e-05,
+      "loss": 0.3691,
       "step": 74920
     },
     {
-      "epoch": 2.64,
-      "learning_rate": 2.402569923869887e-05,
-      "loss": 0.2768,
+      "epoch": 2.700291923451184,
+      "grad_norm": 0.21767933666706085,
+      "learning_rate": 2.298808220573613e-05,
+      "loss": 0.4065,
       "step": 74925
     },
     {
-      "epoch": 2.64,
-      "learning_rate": 2.40228526878858e-05,
-      "loss": 0.3029,
+      "epoch": 2.7004721231124087,
+      "grad_norm": 0.18852004408836365,
+      "learning_rate": 2.298517355079547e-05,
+      "loss": 0.3676,
       "step": 74930
     },
     {
-      "epoch": 2.64,
-      "learning_rate": 2.4020006149760355e-05,
-      "loss": 0.26,
+      "epoch": 2.7006523227736334,
+      "grad_norm": 0.21407105028629303,
+      "learning_rate": 2.2982264923306435e-05,
+      "loss": 0.4,
       "step": 74935
     },
     {
-      "epoch": 2.64,
-      "learning_rate": 2.401715962435948e-05,
-      "loss": 0.2837,
+      "epoch": 2.7008325224348577,
+      "grad_norm": 0.1654965728521347,
+      "learning_rate": 2.2979356323308653e-05,
+      "loss": 0.4009,
       "step": 74940
     },
     {
-      "epoch": 2.64,
-      "learning_rate": 2.401431311172014e-05,
-      "loss": 0.2707,
+      "epoch": 2.7010127220960825,
+      "grad_norm": 0.1940935105085373,
+      "learning_rate": 2.2976447750841742e-05,
+      "loss": 0.379,
       "step": 74945
     },
     {
-      "epoch": 2.64,
-      "learning_rate": 2.4011466611879297e-05,
-      "loss": 0.2681,
+      "epoch": 2.701192921757307,
+      "grad_norm": 0.21042941510677338,
+      "learning_rate": 2.2973539205945347e-05,
+      "loss": 0.437,
       "step": 74950
     },
     {
-      "epoch": 2.64,
-      "learning_rate": 2.400862012487391e-05,
-      "loss": 0.2783,
+      "epoch": 2.7013731214185315,
+      "grad_norm": 0.26631516218185425,
+      "learning_rate": 2.2970630688659086e-05,
+      "loss": 0.4402,
       "step": 74955
     },
     {
-      "epoch": 2.64,
-      "learning_rate": 2.400577365074094e-05,
-      "loss": 0.2826,
+      "epoch": 2.701553321079756,
+      "grad_norm": 0.23533588647842407,
+      "learning_rate": 2.2967722199022585e-05,
+      "loss": 0.3937,
       "step": 74960
     },
     {
-      "epoch": 2.64,
-      "learning_rate": 2.400292718951733e-05,
-      "loss": 0.2472,
+      "epoch": 2.701733520740981,
+      "grad_norm": 0.22266995906829834,
+      "learning_rate": 2.296481373707548e-05,
+      "loss": 0.3798,
       "step": 74965
     },
     {
-      "epoch": 2.64,
-      "learning_rate": 2.4000080741240066e-05,
-      "loss": 0.2662,
+      "epoch": 2.7019137204022057,
+      "grad_norm": 0.22777952253818512,
+      "learning_rate": 2.2961905302857387e-05,
+      "loss": 0.4028,
       "step": 74970
     },
     {
-      "epoch": 2.64,
-      "learning_rate": 2.399723430594609e-05,
-      "loss": 0.2496,
+      "epoch": 2.7020939200634304,
+      "grad_norm": 0.2449411302804947,
+      "learning_rate": 2.295899689640795e-05,
+      "loss": 0.3865,
       "step": 74975
     },
     {
-      "epoch": 2.64,
-      "learning_rate": 2.399438788367237e-05,
-      "loss": 0.2745,
+      "epoch": 2.702274119724655,
+      "grad_norm": 0.188294917345047,
+      "learning_rate": 2.2956088517766784e-05,
+      "loss": 0.4135,
       "step": 74980
     },
     {
-      "epoch": 2.64,
-      "learning_rate": 2.399154147445585e-05,
-      "loss": 0.2454,
+      "epoch": 2.7024543193858794,
+      "grad_norm": 0.19857686758041382,
+      "learning_rate": 2.2953180166973505e-05,
+      "loss": 0.3972,
       "step": 74985
     },
     {
-      "epoch": 2.64,
-      "learning_rate": 2.398869507833351e-05,
-      "loss": 0.2886,
+      "epoch": 2.702634519047104,
+      "grad_norm": 0.2035861313343048,
+      "learning_rate": 2.295027184406776e-05,
+      "loss": 0.3869,
       "step": 74990
     },
     {
-      "epoch": 2.64,
-      "learning_rate": 2.3985848695342294e-05,
-      "loss": 0.2567,
+      "epoch": 2.702814718708329,
+      "grad_norm": 0.20147527754306793,
+      "learning_rate": 2.2947363549089164e-05,
+      "loss": 0.3817,
       "step": 74995
     },
     {
-      "epoch": 2.64,
-      "learning_rate": 2.398300232551917e-05,
-      "loss": 0.2774,
+      "epoch": 2.702994918369553,
+      "grad_norm": 0.18223348259925842,
+      "learning_rate": 2.2944455282077337e-05,
+      "loss": 0.4237,
       "step": 75000
     },
     {
-      "epoch": 2.64,
-      "eval_loss": 0.2608758807182312,
-      "eval_runtime": 10.5423,
-      "eval_samples_per_second": 9.486,
-      "eval_steps_per_second": 9.486,
+      "epoch": 2.702994918369553,
+      "eval_loss": 0.43115875124931335,
+      "eval_runtime": 3.5382,
+      "eval_samples_per_second": 28.263,
+      "eval_steps_per_second": 7.066,
       "step": 75000
     },
     {
-      "epoch": 2.64,
-      "learning_rate": 2.3980155968901078e-05,
-      "loss": 0.2513,
+      "epoch": 2.703175118030778,
+      "grad_norm": 0.1930094212293625,
+      "learning_rate": 2.2941547043071916e-05,
+      "loss": 0.3965,
       "step": 75005
     },
     {
-      "epoch": 2.64,
-      "learning_rate": 2.3977309625524998e-05,
-      "loss": 0.2618,
+      "epoch": 2.7033553176920027,
+      "grad_norm": 0.1886611133813858,
+      "learning_rate": 2.2938638832112507e-05,
+      "loss": 0.4005,
       "step": 75010
     },
     {
-      "epoch": 2.64,
-      "learning_rate": 2.397446329542788e-05,
-      "loss": 0.2612,
+      "epoch": 2.7035355173532274,
+      "grad_norm": 0.2501041293144226,
+      "learning_rate": 2.2935730649238753e-05,
+      "loss": 0.4165,
       "step": 75015
     },
     {
-      "epoch": 2.64,
-      "learning_rate": 2.3971616978646678e-05,
-      "loss": 0.2762,
+      "epoch": 2.703715717014452,
+      "grad_norm": 0.2163308709859848,
+      "learning_rate": 2.293282249449027e-05,
+      "loss": 0.4063,
       "step": 75020
     },
     {
-      "epoch": 2.64,
-      "learning_rate": 2.396877067521835e-05,
-      "loss": 0.2579,
+      "epoch": 2.703895916675677,
+      "grad_norm": 0.20869512856006622,
+      "learning_rate": 2.2929914367906674e-05,
+      "loss": 0.4237,
       "step": 75025
     },
     {
-      "epoch": 2.64,
-      "learning_rate": 2.3965924385179853e-05,
-      "loss": 0.2517,
+      "epoch": 2.704076116336901,
+      "grad_norm": 0.22830967605113983,
+      "learning_rate": 2.2927006269527597e-05,
+      "loss": 0.3773,
       "step": 75030
     },
     {
-      "epoch": 2.64,
-      "learning_rate": 2.396307810856815e-05,
-      "loss": 0.2847,
+      "epoch": 2.704256315998126,
+      "grad_norm": 0.2236100733280182,
+      "learning_rate": 2.2924098199392658e-05,
+      "loss": 0.4216,
       "step": 75035
     },
     {
-      "epoch": 2.64,
-      "learning_rate": 2.39602318454202e-05,
-      "loss": 0.2795,
+      "epoch": 2.7044365156593506,
+      "grad_norm": 0.20641185343265533,
+      "learning_rate": 2.292119015754148e-05,
+      "loss": 0.3819,
       "step": 75040
     },
     {
-      "epoch": 2.64,
-      "learning_rate": 2.395738559577294e-05,
-      "loss": 0.2711,
+      "epoch": 2.7046167153205753,
+      "grad_norm": 0.2077999860048294,
+      "learning_rate": 2.2918282144013685e-05,
+      "loss": 0.4211,
       "step": 75045
     },
     {
-      "epoch": 2.64,
-      "learning_rate": 2.395453935966335e-05,
-      "loss": 0.2744,
+      "epoch": 2.7047969149817996,
+      "grad_norm": 0.21933238208293915,
+      "learning_rate": 2.291537415884889e-05,
+      "loss": 0.3865,
       "step": 75050
     },
     {
-      "epoch": 2.64,
-      "learning_rate": 2.3951693137128377e-05,
-      "loss": 0.2652,
+      "epoch": 2.7049771146430244,
+      "grad_norm": 0.1914396733045578,
+      "learning_rate": 2.291246620208672e-05,
+      "loss": 0.3647,
       "step": 75055
     },
     {
-      "epoch": 2.64,
-      "learning_rate": 2.394884692820497e-05,
-      "loss": 0.2746,
+      "epoch": 2.705157314304249,
+      "grad_norm": 0.24134762585163116,
+      "learning_rate": 2.2909558273766802e-05,
+      "loss": 0.3848,
       "step": 75060
     },
     {
-      "epoch": 2.64,
-      "learning_rate": 2.3946000732930105e-05,
-      "loss": 0.2683,
+      "epoch": 2.705337513965474,
+      "grad_norm": 0.22134195268154144,
+      "learning_rate": 2.290665037392873e-05,
+      "loss": 0.3949,
       "step": 75065
     },
     {
-      "epoch": 2.64,
-      "learning_rate": 2.3943154551340724e-05,
-      "loss": 0.2474,
+      "epoch": 2.7055177136266986,
+      "grad_norm": 0.17406292259693146,
+      "learning_rate": 2.2903742502612153e-05,
+      "loss": 0.3943,
       "step": 75070
     },
     {
-      "epoch": 2.64,
-      "learning_rate": 2.3940308383473788e-05,
-      "loss": 0.2929,
+      "epoch": 2.705697913287923,
+      "grad_norm": 0.20366698503494263,
+      "learning_rate": 2.2900834659856673e-05,
+      "loss": 0.3911,
       "step": 75075
     },
     {
-      "epoch": 2.64,
-      "learning_rate": 2.3937462229366234e-05,
-      "loss": 0.2724,
+      "epoch": 2.7058781129491476,
+      "grad_norm": 0.1711958944797516,
+      "learning_rate": 2.289792684570192e-05,
+      "loss": 0.3821,
       "step": 75080
     },
     {
-      "epoch": 2.64,
-      "learning_rate": 2.3934616089055038e-05,
-      "loss": 0.2569,
+      "epoch": 2.7060583126103723,
+      "grad_norm": 0.21541287004947662,
+      "learning_rate": 2.289501906018751e-05,
+      "loss": 0.4022,
       "step": 75085
     },
     {
-      "epoch": 2.64,
-      "learning_rate": 2.393176996257716e-05,
-      "loss": 0.2704,
+      "epoch": 2.706238512271597,
+      "grad_norm": 0.2087329775094986,
+      "learning_rate": 2.2892111303353046e-05,
+      "loss": 0.3882,
       "step": 75090
     },
     {
-      "epoch": 2.64,
-      "learning_rate": 2.3928923849969543e-05,
-      "loss": 0.2602,
+      "epoch": 2.7064187119328214,
+      "grad_norm": 0.2614741325378418,
+      "learning_rate": 2.288920357523817e-05,
+      "loss": 0.3778,
       "step": 75095
     },
     {
-      "epoch": 2.64,
-      "learning_rate": 2.392607775126913e-05,
-      "loss": 0.2738,
+      "epoch": 2.706598911594046,
+      "grad_norm": 0.286384642124176,
+      "learning_rate": 2.2886295875882484e-05,
+      "loss": 0.4149,
       "step": 75100
     },
     {
-      "epoch": 2.64,
-      "learning_rate": 2.3923231666512906e-05,
-      "loss": 0.2869,
+      "epoch": 2.706779111255271,
+      "grad_norm": 0.22412830591201782,
+      "learning_rate": 2.28833882053256e-05,
+      "loss": 0.4088,
       "step": 75105
     },
     {
-      "epoch": 2.64,
-      "learning_rate": 2.39203855957378e-05,
-      "loss": 0.2515,
+      "epoch": 2.7069593109164956,
+      "grad_norm": 0.1880389153957367,
+      "learning_rate": 2.2880480563607145e-05,
+      "loss": 0.4153,
       "step": 75110
     },
     {
-      "epoch": 2.64,
-      "learning_rate": 2.391753953898078e-05,
-      "loss": 0.2623,
+      "epoch": 2.7071395105777203,
+      "grad_norm": 0.19842107594013214,
+      "learning_rate": 2.287757295076673e-05,
+      "loss": 0.3912,
       "step": 75115
     },
     {
-      "epoch": 2.64,
-      "learning_rate": 2.3914693496278787e-05,
-      "loss": 0.2683,
+      "epoch": 2.707319710238945,
+      "grad_norm": 0.20793603360652924,
+      "learning_rate": 2.2874665366843977e-05,
+      "loss": 0.4447,
       "step": 75120
     },
     {
-      "epoch": 2.64,
-      "learning_rate": 2.3911847467668793e-05,
-      "loss": 0.2926,
+      "epoch": 2.7074999099001693,
+      "grad_norm": 0.18394573032855988,
+      "learning_rate": 2.2871757811878497e-05,
+      "loss": 0.3807,
       "step": 75125
     },
     {
-      "epoch": 2.64,
-      "learning_rate": 2.3909001453187747e-05,
-      "loss": 0.2708,
+      "epoch": 2.707680109561394,
+      "grad_norm": 0.2318859100341797,
+      "learning_rate": 2.2868850285909897e-05,
+      "loss": 0.407,
       "step": 75130
     },
     {
-      "epoch": 2.64,
-      "learning_rate": 2.3906155452872588e-05,
-      "loss": 0.2827,
+      "epoch": 2.7078603092226188,
+      "grad_norm": 0.19001048803329468,
+      "learning_rate": 2.286594278897781e-05,
+      "loss": 0.3766,
       "step": 75135
     },
     {
-      "epoch": 2.64,
-      "learning_rate": 2.3903309466760274e-05,
-      "loss": 0.3036,
+      "epoch": 2.708040508883843,
+      "grad_norm": 0.2057991772890091,
+      "learning_rate": 2.2863035321121836e-05,
+      "loss": 0.3665,
       "step": 75140
     },
     {
-      "epoch": 2.64,
-      "learning_rate": 2.3900463494887776e-05,
-      "loss": 0.2739,
+      "epoch": 2.708220708545068,
+      "grad_norm": 0.23362144827842712,
+      "learning_rate": 2.286012788238159e-05,
+      "loss": 0.3889,
       "step": 75145
     },
     {
-      "epoch": 2.64,
-      "learning_rate": 2.3897617537292036e-05,
-      "loss": 0.2748,
+      "epoch": 2.7084009082062925,
+      "grad_norm": 0.1938500553369522,
+      "learning_rate": 2.2857220472796688e-05,
+      "loss": 0.4234,
       "step": 75150
     },
     {
-      "epoch": 2.64,
-      "learning_rate": 2.3894771594010004e-05,
-      "loss": 0.2815,
+      "epoch": 2.7085811078675173,
+      "grad_norm": 0.17589519917964935,
+      "learning_rate": 2.285431309240674e-05,
+      "loss": 0.3794,
       "step": 75155
     },
     {
-      "epoch": 2.64,
-      "learning_rate": 2.3891925665078624e-05,
-      "loss": 0.2412,
+      "epoch": 2.708761307528742,
+      "grad_norm": 0.2224626988172531,
+      "learning_rate": 2.285140574125136e-05,
+      "loss": 0.3915,
       "step": 75160
     },
     {
-      "epoch": 2.64,
-      "learning_rate": 2.3889079750534865e-05,
-      "loss": 0.2747,
+      "epoch": 2.7089415071899667,
+      "grad_norm": 0.17219646275043488,
+      "learning_rate": 2.2848498419370174e-05,
+      "loss": 0.4047,
       "step": 75165
     },
     {
-      "epoch": 2.64,
-      "learning_rate": 2.388623385041568e-05,
-      "loss": 0.2702,
+      "epoch": 2.709121706851191,
+      "grad_norm": 0.26903530955314636,
+      "learning_rate": 2.284559112680276e-05,
+      "loss": 0.3869,
       "step": 75170
     },
     {
-      "epoch": 2.64,
-      "learning_rate": 2.3883387964758e-05,
-      "loss": 0.2612,
+      "epoch": 2.7093019065124158,
+      "grad_norm": 0.18370231986045837,
+      "learning_rate": 2.2842683863588766e-05,
+      "loss": 0.3703,
       "step": 75175
     },
     {
-      "epoch": 2.65,
-      "learning_rate": 2.3880542093598805e-05,
-      "loss": 0.2415,
+      "epoch": 2.7094821061736405,
+      "grad_norm": 0.20962488651275635,
+      "learning_rate": 2.2839776629767785e-05,
+      "loss": 0.3904,
       "step": 75180
     },
     {
-      "epoch": 2.65,
-      "learning_rate": 2.3877696236975034e-05,
-      "loss": 0.268,
+      "epoch": 2.709662305834865,
+      "grad_norm": 0.2684512734413147,
+      "learning_rate": 2.283686942537942e-05,
+      "loss": 0.3828,
       "step": 75185
     },
     {
-      "epoch": 2.65,
-      "learning_rate": 2.387485039492363e-05,
-      "loss": 0.2636,
+      "epoch": 2.7098425054960895,
+      "grad_norm": 0.20460106432437897,
+      "learning_rate": 2.2833962250463293e-05,
+      "loss": 0.4142,
       "step": 75190
     },
     {
-      "epoch": 2.65,
-      "learning_rate": 2.387200456748155e-05,
-      "loss": 0.2544,
+      "epoch": 2.7100227051573142,
+      "grad_norm": 0.21397659182548523,
+      "learning_rate": 2.2831055105059007e-05,
+      "loss": 0.3773,
       "step": 75195
     },
     {
-      "epoch": 2.65,
-      "learning_rate": 2.3869158754685753e-05,
-      "loss": 0.2749,
+      "epoch": 2.710202904818539,
+      "grad_norm": 0.20380306243896484,
+      "learning_rate": 2.282814798920619e-05,
+      "loss": 0.3858,
       "step": 75200
     },
     {
-      "epoch": 2.65,
-      "learning_rate": 2.3866312956573188e-05,
-      "loss": 0.2571,
+      "epoch": 2.7103831044797637,
+      "grad_norm": 0.2668623626232147,
+      "learning_rate": 2.282524090294443e-05,
+      "loss": 0.3665,
       "step": 75205
     },
     {
-      "epoch": 2.65,
-      "learning_rate": 2.38634671731808e-05,
-      "loss": 0.2724,
+      "epoch": 2.7105633041409884,
+      "grad_norm": 0.1861308068037033,
+      "learning_rate": 2.2822333846313332e-05,
+      "loss": 0.3891,
       "step": 75210
     },
     {
-      "epoch": 2.65,
-      "learning_rate": 2.386062140454553e-05,
-      "loss": 0.2596,
+      "epoch": 2.7107435038022127,
+      "grad_norm": 0.25733551383018494,
+      "learning_rate": 2.2819426819352525e-05,
+      "loss": 0.3773,
       "step": 75215
     },
     {
-      "epoch": 2.65,
-      "learning_rate": 2.385777565070435e-05,
-      "loss": 0.2694,
+      "epoch": 2.7109237034634375,
+      "grad_norm": 0.19054418802261353,
+      "learning_rate": 2.2816519822101596e-05,
+      "loss": 0.3619,
       "step": 75220
     },
     {
-      "epoch": 2.65,
-      "learning_rate": 2.3854929911694203e-05,
-      "loss": 0.2482,
+      "epoch": 2.711103903124662,
+      "grad_norm": 0.2656010687351227,
+      "learning_rate": 2.281361285460017e-05,
+      "loss": 0.3848,
       "step": 75225
     },
     {
-      "epoch": 2.65,
-      "learning_rate": 2.385208418755204e-05,
-      "loss": 0.2841,
+      "epoch": 2.7112841027858865,
+      "grad_norm": 0.21391278505325317,
+      "learning_rate": 2.281070591688784e-05,
+      "loss": 0.3942,
       "step": 75230
     },
     {
-      "epoch": 2.65,
-      "learning_rate": 2.384923847831479e-05,
-      "loss": 0.2664,
+      "epoch": 2.7114643024471112,
+      "grad_norm": 0.20552653074264526,
+      "learning_rate": 2.280779900900422e-05,
+      "loss": 0.3606,
       "step": 75235
     },
     {
-      "epoch": 2.65,
-      "learning_rate": 2.3846392784019436e-05,
-      "loss": 0.2859,
+      "epoch": 2.711644502108336,
+      "grad_norm": 0.2136029154062271,
+      "learning_rate": 2.2804892130988916e-05,
+      "loss": 0.4099,
       "step": 75240
     },
     {
-      "epoch": 2.65,
-      "learning_rate": 2.3843547104702903e-05,
-      "loss": 0.2856,
+      "epoch": 2.7118247017695607,
+      "grad_norm": 0.20137493312358856,
+      "learning_rate": 2.2801985282881532e-05,
+      "loss": 0.3833,
       "step": 75245
     },
     {
-      "epoch": 2.65,
-      "learning_rate": 2.3840701440402156e-05,
-      "loss": 0.2696,
+      "epoch": 2.7120049014307854,
+      "grad_norm": 0.17652560770511627,
+      "learning_rate": 2.279907846472167e-05,
+      "loss": 0.3546,
       "step": 75250
     },
     {
-      "epoch": 2.65,
-      "learning_rate": 2.3837855791154127e-05,
-      "loss": 0.2782,
+      "epoch": 2.71218510109201,
+      "grad_norm": 0.2107645869255066,
+      "learning_rate": 2.279617167654894e-05,
+      "loss": 0.3693,
       "step": 75255
     },
     {
-      "epoch": 2.65,
-      "learning_rate": 2.3835010156995783e-05,
-      "loss": 0.2685,
+      "epoch": 2.7123653007532345,
+      "grad_norm": 0.17879261076450348,
+      "learning_rate": 2.279326491840294e-05,
+      "loss": 0.4006,
       "step": 75260
     },
     {
-      "epoch": 2.65,
-      "learning_rate": 2.3832164537964066e-05,
-      "loss": 0.2724,
+      "epoch": 2.712545500414459,
+      "grad_norm": 0.1885097324848175,
+      "learning_rate": 2.279035819032328e-05,
+      "loss": 0.4,
       "step": 75265
     },
     {
-      "epoch": 2.65,
-      "learning_rate": 2.3829318934095918e-05,
-      "loss": 0.2717,
+      "epoch": 2.712725700075684,
+      "grad_norm": 0.19376368820667267,
+      "learning_rate": 2.2787451492349574e-05,
+      "loss": 0.3943,
       "step": 75270
     },
     {
-      "epoch": 2.65,
-      "learning_rate": 2.3826473345428285e-05,
-      "loss": 0.2621,
+      "epoch": 2.7129058997369087,
+      "grad_norm": 0.22324879467487335,
+      "learning_rate": 2.2784544824521392e-05,
+      "loss": 0.4092,
       "step": 75275
     },
     {
-      "epoch": 2.65,
-      "learning_rate": 2.3823627771998134e-05,
-      "loss": 0.2779,
+      "epoch": 2.713086099398133,
+      "grad_norm": 0.18715398013591766,
+      "learning_rate": 2.2781638186878375e-05,
+      "loss": 0.4188,
       "step": 75280
     },
     {
-      "epoch": 2.65,
-      "learning_rate": 2.38207822138424e-05,
-      "loss": 0.2559,
+      "epoch": 2.7132662990593577,
+      "grad_norm": 0.24565525352954865,
+      "learning_rate": 2.2778731579460105e-05,
+      "loss": 0.4029,
       "step": 75285
     },
     {
-      "epoch": 2.65,
-      "learning_rate": 2.381793667099802e-05,
-      "loss": 0.2871,
+      "epoch": 2.7134464987205824,
+      "grad_norm": 0.28024908900260925,
+      "learning_rate": 2.277582500230618e-05,
+      "loss": 0.3778,
       "step": 75290
     },
     {
-      "epoch": 2.65,
-      "learning_rate": 2.3815091143501968e-05,
-      "loss": 0.2464,
+      "epoch": 2.713626698381807,
+      "grad_norm": 0.1823398321866989,
+      "learning_rate": 2.2772918455456215e-05,
+      "loss": 0.3794,
       "step": 75295
     },
     {
-      "epoch": 2.65,
-      "learning_rate": 2.3812245631391168e-05,
-      "loss": 0.266,
+      "epoch": 2.713806898043032,
+      "grad_norm": 0.22716470062732697,
+      "learning_rate": 2.27700119389498e-05,
+      "loss": 0.3845,
       "step": 75300
     },
     {
-      "epoch": 2.65,
-      "learning_rate": 2.3809400134702582e-05,
-      "loss": 0.2678,
+      "epoch": 2.713987097704256,
+      "grad_norm": 0.21729101240634918,
+      "learning_rate": 2.2767105452826542e-05,
+      "loss": 0.3898,
       "step": 75305
     },
     {
-      "epoch": 2.65,
-      "learning_rate": 2.380655465347314e-05,
-      "loss": 0.244,
+      "epoch": 2.714167297365481,
+      "grad_norm": 0.22123655676841736,
+      "learning_rate": 2.2764198997126043e-05,
+      "loss": 0.3791,
       "step": 75310
     },
     {
-      "epoch": 2.65,
-      "learning_rate": 2.380370918773981e-05,
-      "loss": 0.2402,
+      "epoch": 2.7143474970267056,
+      "grad_norm": 0.222514346241951,
+      "learning_rate": 2.2761292571887894e-05,
+      "loss": 0.3756,
       "step": 75315
     },
     {
-      "epoch": 2.65,
-      "learning_rate": 2.380086373753953e-05,
-      "loss": 0.2624,
+      "epoch": 2.7145276966879304,
+      "grad_norm": 0.23015157878398895,
+      "learning_rate": 2.2758386177151707e-05,
+      "loss": 0.3883,
       "step": 75320
     },
     {
-      "epoch": 2.65,
-      "learning_rate": 2.3798018302909235e-05,
-      "loss": 0.265,
+      "epoch": 2.7147078963491547,
+      "grad_norm": 0.23218706250190735,
+      "learning_rate": 2.2755479812957074e-05,
+      "loss": 0.3997,
       "step": 75325
     },
     {
-      "epoch": 2.65,
-      "learning_rate": 2.379517288388588e-05,
-      "loss": 0.2735,
+      "epoch": 2.7148880960103794,
+      "grad_norm": 0.22764332592487335,
+      "learning_rate": 2.2752573479343588e-05,
+      "loss": 0.3802,
       "step": 75330
     },
     {
-      "epoch": 2.65,
-      "learning_rate": 2.3792327480506412e-05,
-      "loss": 0.2528,
+      "epoch": 2.715068295671604,
+      "grad_norm": 0.21155443787574768,
+      "learning_rate": 2.2749667176350857e-05,
+      "loss": 0.4231,
       "step": 75335
     },
     {
-      "epoch": 2.65,
-      "learning_rate": 2.3789482092807784e-05,
-      "loss": 0.2571,
+      "epoch": 2.715248495332829,
+      "grad_norm": 0.19925571978092194,
+      "learning_rate": 2.2746760904018472e-05,
+      "loss": 0.3896,
       "step": 75340
     },
     {
-      "epoch": 2.65,
-      "learning_rate": 2.3786636720826934e-05,
-      "loss": 0.2701,
+      "epoch": 2.7154286949940536,
+      "grad_norm": 0.22160978615283966,
+      "learning_rate": 2.2743854662386035e-05,
+      "loss": 0.3815,
       "step": 75345
     },
     {
-      "epoch": 2.65,
-      "learning_rate": 2.378379136460079e-05,
-      "loss": 0.2734,
+      "epoch": 2.715608894655278,
+      "grad_norm": 0.22966960072517395,
+      "learning_rate": 2.2740948451493148e-05,
+      "loss": 0.3712,
       "step": 75350
     },
     {
-      "epoch": 2.65,
-      "learning_rate": 2.378094602416632e-05,
-      "loss": 0.2922,
+      "epoch": 2.7157890943165026,
+      "grad_norm": 0.1914483606815338,
+      "learning_rate": 2.2738042271379388e-05,
+      "loss": 0.3998,
       "step": 75355
     },
     {
-      "epoch": 2.65,
-      "learning_rate": 2.3778100699560474e-05,
-      "loss": 0.2715,
+      "epoch": 2.7159692939777274,
+      "grad_norm": 0.22551338374614716,
+      "learning_rate": 2.2735136122084374e-05,
+      "loss": 0.3848,
       "step": 75360
     },
     {
-      "epoch": 2.65,
-      "learning_rate": 2.377525539082018e-05,
-      "loss": 0.2577,
+      "epoch": 2.716149493638952,
+      "grad_norm": 0.2782670259475708,
+      "learning_rate": 2.2732230003647698e-05,
+      "loss": 0.4082,
       "step": 75365
     },
     {
-      "epoch": 2.65,
-      "learning_rate": 2.377241009798238e-05,
-      "loss": 0.2751,
+      "epoch": 2.7163296933001764,
+      "grad_norm": 0.21808646619319916,
+      "learning_rate": 2.272932391610893e-05,
+      "loss": 0.3989,
       "step": 75370
     },
     {
-      "epoch": 2.65,
-      "learning_rate": 2.376956482108403e-05,
-      "loss": 0.2759,
+      "epoch": 2.716509892961401,
+      "grad_norm": 0.22317920625209808,
+      "learning_rate": 2.2726417859507703e-05,
+      "loss": 0.4206,
       "step": 75375
     },
     {
-      "epoch": 2.65,
-      "learning_rate": 2.376671956016207e-05,
-      "loss": 0.2581,
+      "epoch": 2.716690092622626,
+      "grad_norm": 0.19897177815437317,
+      "learning_rate": 2.2723511833883574e-05,
+      "loss": 0.3832,
       "step": 75380
     },
     {
-      "epoch": 2.65,
-      "learning_rate": 2.376387431525345e-05,
-      "loss": 0.2553,
+      "epoch": 2.7168702922838506,
+      "grad_norm": 0.20371364057064056,
+      "learning_rate": 2.2720605839276173e-05,
+      "loss": 0.4227,
       "step": 75385
     },
     {
-      "epoch": 2.65,
-      "learning_rate": 2.3761029086395095e-05,
-      "loss": 0.2709,
+      "epoch": 2.7170504919450753,
+      "grad_norm": 0.2331681102514267,
+      "learning_rate": 2.2717699875725072e-05,
+      "loss": 0.3592,
       "step": 75390
     },
     {
-      "epoch": 2.65,
-      "learning_rate": 2.375818387362397e-05,
-      "loss": 0.2801,
+      "epoch": 2.7172306916063,
+      "grad_norm": 0.2340395599603653,
+      "learning_rate": 2.271479394326986e-05,
+      "loss": 0.4067,
       "step": 75395
     },
     {
-      "epoch": 2.65,
-      "learning_rate": 2.3755338676977012e-05,
-      "loss": 0.2996,
+      "epoch": 2.7174108912675243,
+      "grad_norm": 0.2065713256597519,
+      "learning_rate": 2.2711888041950143e-05,
+      "loss": 0.3794,
       "step": 75400
     },
     {
-      "epoch": 2.65,
-      "learning_rate": 2.3752493496491148e-05,
-      "loss": 0.2742,
+      "epoch": 2.717591090928749,
+      "grad_norm": 0.19182537496089935,
+      "learning_rate": 2.2708982171805512e-05,
+      "loss": 0.388,
       "step": 75405
     },
     {
-      "epoch": 2.65,
-      "learning_rate": 2.3749648332203346e-05,
-      "loss": 0.257,
+      "epoch": 2.717771290589974,
+      "grad_norm": 0.2015332728624344,
+      "learning_rate": 2.2706076332875546e-05,
+      "loss": 0.3948,
       "step": 75410
     },
     {
-      "epoch": 2.65,
-      "learning_rate": 2.374680318415053e-05,
-      "loss": 0.278,
+      "epoch": 2.717951490251198,
+      "grad_norm": 0.2081226408481598,
+      "learning_rate": 2.2703170525199856e-05,
+      "loss": 0.3549,
       "step": 75415
     },
     {
-      "epoch": 2.65,
-      "learning_rate": 2.3743958052369652e-05,
-      "loss": 0.2875,
+      "epoch": 2.718131689912423,
+      "grad_norm": 0.23630069196224213,
+      "learning_rate": 2.2700264748818015e-05,
+      "loss": 0.4129,
       "step": 75420
     },
     {
-      "epoch": 2.65,
-      "learning_rate": 2.3741112936897646e-05,
-      "loss": 0.2619,
+      "epoch": 2.7183118895736476,
+      "grad_norm": 0.20923583209514618,
+      "learning_rate": 2.2697359003769627e-05,
+      "loss": 0.4091,
       "step": 75425
     },
     {
-      "epoch": 2.65,
-      "learning_rate": 2.3738267837771464e-05,
-      "loss": 0.2756,
+      "epoch": 2.7184920892348723,
+      "grad_norm": 0.2418445646762848,
+      "learning_rate": 2.2694453290094276e-05,
+      "loss": 0.4168,
       "step": 75430
     },
     {
-      "epoch": 2.65,
-      "learning_rate": 2.373542275502804e-05,
-      "loss": 0.2495,
+      "epoch": 2.718672288896097,
+      "grad_norm": 0.20278342068195343,
+      "learning_rate": 2.2691547607831547e-05,
+      "loss": 0.4111,
       "step": 75435
     },
     {
-      "epoch": 2.65,
-      "learning_rate": 2.3732577688704323e-05,
-      "loss": 0.2631,
+      "epoch": 2.7188524885573218,
+      "grad_norm": 0.2284698486328125,
+      "learning_rate": 2.2688641957021043e-05,
+      "loss": 0.3705,
       "step": 75440
     },
     {
-      "epoch": 2.65,
-      "learning_rate": 2.3729732638837238e-05,
-      "loss": 0.2677,
+      "epoch": 2.719032688218546,
+      "grad_norm": 0.21057669818401337,
+      "learning_rate": 2.2685736337702336e-05,
+      "loss": 0.3889,
       "step": 75445
     },
     {
-      "epoch": 2.65,
-      "learning_rate": 2.3726887605463748e-05,
-      "loss": 0.2637,
+      "epoch": 2.719212887879771,
+      "grad_norm": 0.23061637580394745,
+      "learning_rate": 2.268283074991503e-05,
+      "loss": 0.412,
       "step": 75450
     },
     {
-      "epoch": 2.65,
-      "learning_rate": 2.3724042588620786e-05,
-      "loss": 0.2702,
+      "epoch": 2.7193930875409955,
+      "grad_norm": 0.1986396461725235,
+      "learning_rate": 2.2679925193698713e-05,
+      "loss": 0.3599,
       "step": 75455
     },
     {
-      "epoch": 2.65,
-      "learning_rate": 2.3721197588345285e-05,
-      "loss": 0.2665,
+      "epoch": 2.71957328720222,
+      "grad_norm": 0.20687183737754822,
+      "learning_rate": 2.267701966909295e-05,
+      "loss": 0.3999,
       "step": 75460
     },
     {
-      "epoch": 2.66,
-      "learning_rate": 2.3718352604674187e-05,
-      "loss": 0.2869,
+      "epoch": 2.7197534868634445,
+      "grad_norm": 0.22581927478313446,
+      "learning_rate": 2.267411417613736e-05,
+      "loss": 0.3888,
       "step": 75465
     },
     {
-      "epoch": 2.66,
-      "learning_rate": 2.3715507637644437e-05,
-      "loss": 0.2535,
+      "epoch": 2.7199336865246693,
+      "grad_norm": 0.1779545694589615,
+      "learning_rate": 2.2671208714871507e-05,
+      "loss": 0.3675,
       "step": 75470
     },
     {
-      "epoch": 2.66,
-      "learning_rate": 2.371266268729298e-05,
-      "loss": 0.2708,
+      "epoch": 2.720113886185894,
+      "grad_norm": 0.19196881353855133,
+      "learning_rate": 2.2668303285334974e-05,
+      "loss": 0.3927,
       "step": 75475
     },
     {
-      "epoch": 2.66,
-      "learning_rate": 2.3709817753656748e-05,
-      "loss": 0.27,
+      "epoch": 2.7202940858471187,
+      "grad_norm": 0.21891996264457703,
+      "learning_rate": 2.2665397887567374e-05,
+      "loss": 0.3826,
       "step": 75480
     },
     {
-      "epoch": 2.66,
-      "learning_rate": 2.3706972836772673e-05,
-      "loss": 0.2495,
+      "epoch": 2.7204742855083435,
+      "grad_norm": 0.2428896129131317,
+      "learning_rate": 2.2662492521608263e-05,
+      "loss": 0.3847,
       "step": 75485
     },
     {
-      "epoch": 2.66,
-      "learning_rate": 2.3704127936677713e-05,
-      "loss": 0.2591,
+      "epoch": 2.7206544851695678,
+      "grad_norm": 0.21766485273838043,
+      "learning_rate": 2.2659587187497248e-05,
+      "loss": 0.408,
       "step": 75490
     },
     {
-      "epoch": 2.66,
-      "learning_rate": 2.3701283053408795e-05,
-      "loss": 0.2735,
+      "epoch": 2.7208346848307925,
+      "grad_norm": 0.26046082377433777,
+      "learning_rate": 2.2656681885273907e-05,
+      "loss": 0.3949,
       "step": 75495
     },
     {
-      "epoch": 2.66,
-      "learning_rate": 2.3698438187002863e-05,
-      "loss": 0.2667,
+      "epoch": 2.7210148844920172,
+      "grad_norm": 0.2007281333208084,
+      "learning_rate": 2.2653776614977813e-05,
+      "loss": 0.3955,
       "step": 75500
     },
     {
-      "epoch": 2.66,
-      "eval_loss": 0.2609610855579376,
-      "eval_runtime": 10.5588,
-      "eval_samples_per_second": 9.471,
-      "eval_steps_per_second": 9.471,
+      "epoch": 2.7210148844920172,
+      "eval_loss": 0.4320552349090576,
+      "eval_runtime": 3.5362,
+      "eval_samples_per_second": 28.279,
+      "eval_steps_per_second": 7.07,
       "step": 75500
     },
     {
-      "epoch": 2.66,
-      "learning_rate": 2.3695593337496847e-05,
-      "loss": 0.2506,
+      "epoch": 2.7211950841532415,
+      "grad_norm": 0.29919010400772095,
+      "learning_rate": 2.265087137664856e-05,
+      "loss": 0.3914,
       "step": 75505
     },
     {
-      "epoch": 2.66,
-      "learning_rate": 2.36927485049277e-05,
-      "loss": 0.2631,
+      "epoch": 2.7213752838144663,
+      "grad_norm": 0.24532043933868408,
+      "learning_rate": 2.2647966170325733e-05,
+      "loss": 0.4182,
       "step": 75510
     },
     {
-      "epoch": 2.66,
-      "learning_rate": 2.3689903689332353e-05,
-      "loss": 0.2895,
+      "epoch": 2.721555483475691,
+      "grad_norm": 0.2970276176929474,
+      "learning_rate": 2.2645060996048904e-05,
+      "loss": 0.3703,
       "step": 75515
     },
     {
-      "epoch": 2.66,
-      "learning_rate": 2.368705889074773e-05,
-      "loss": 0.266,
+      "epoch": 2.7217356831369157,
+      "grad_norm": 0.19525529444217682,
+      "learning_rate": 2.2642155853857673e-05,
+      "loss": 0.4027,
       "step": 75520
     },
     {
-      "epoch": 2.66,
-      "learning_rate": 2.3684214109210788e-05,
-      "loss": 0.2753,
+      "epoch": 2.7219158827981405,
+      "grad_norm": 0.21934610605239868,
+      "learning_rate": 2.26392507437916e-05,
+      "loss": 0.4019,
       "step": 75525
     },
     {
-      "epoch": 2.66,
-      "learning_rate": 2.3681369344758464e-05,
-      "loss": 0.2822,
+      "epoch": 2.722096082459365,
+      "grad_norm": 0.20460151135921478,
+      "learning_rate": 2.263634566589029e-05,
+      "loss": 0.3934,
       "step": 75530
     },
     {
-      "epoch": 2.66,
-      "learning_rate": 2.367852459742769e-05,
-      "loss": 0.2874,
+      "epoch": 2.7222762821205895,
+      "grad_norm": 0.18550460040569305,
+      "learning_rate": 2.263344062019331e-05,
+      "loss": 0.4085,
       "step": 75535
     },
     {
-      "epoch": 2.66,
-      "learning_rate": 2.367567986725539e-05,
-      "loss": 0.2457,
+      "epoch": 2.722456481781814,
+      "grad_norm": 0.23762153089046478,
+      "learning_rate": 2.263053560674024e-05,
+      "loss": 0.4393,
       "step": 75540
     },
     {
-      "epoch": 2.66,
-      "learning_rate": 2.3672835154278534e-05,
-      "loss": 0.2878,
+      "epoch": 2.722636681443039,
+      "grad_norm": 0.20546315610408783,
+      "learning_rate": 2.2627630625570666e-05,
+      "loss": 0.3945,
       "step": 75545
     },
     {
-      "epoch": 2.66,
-      "learning_rate": 2.3669990458534023e-05,
-      "loss": 0.2828,
+      "epoch": 2.7228168811042637,
+      "grad_norm": 0.22104191780090332,
+      "learning_rate": 2.2624725676724175e-05,
+      "loss": 0.3628,
       "step": 75550
     },
     {
-      "epoch": 2.66,
-      "learning_rate": 2.3667145780058823e-05,
-      "loss": 0.263,
+      "epoch": 2.722997080765488,
+      "grad_norm": 0.21103298664093018,
+      "learning_rate": 2.2621820760240316e-05,
+      "loss": 0.3632,
       "step": 75555
     },
     {
-      "epoch": 2.66,
-      "learning_rate": 2.3664301118889842e-05,
-      "loss": 0.2772,
+      "epoch": 2.7231772804267127,
+      "grad_norm": 0.20641304552555084,
+      "learning_rate": 2.261891587615871e-05,
+      "loss": 0.4007,
       "step": 75560
     },
     {
-      "epoch": 2.66,
-      "learning_rate": 2.3661456475064043e-05,
-      "loss": 0.2477,
+      "epoch": 2.7233574800879374,
+      "grad_norm": 0.19765135645866394,
+      "learning_rate": 2.261601102451889e-05,
+      "loss": 0.3494,
       "step": 75565
     },
     {
-      "epoch": 2.66,
-      "learning_rate": 2.3658611848618346e-05,
-      "loss": 0.2818,
+      "epoch": 2.723537679749162,
+      "grad_norm": 0.22192902863025665,
+      "learning_rate": 2.261310620536048e-05,
+      "loss": 0.3931,
       "step": 75570
     },
     {
-      "epoch": 2.66,
-      "learning_rate": 2.3655767239589687e-05,
-      "loss": 0.259,
+      "epoch": 2.723717879410387,
+      "grad_norm": 0.23835474252700806,
+      "learning_rate": 2.261020141872303e-05,
+      "loss": 0.3943,
       "step": 75575
     },
     {
-      "epoch": 2.66,
-      "learning_rate": 2.3652922648015e-05,
-      "loss": 0.2561,
+      "epoch": 2.723898079071611,
+      "grad_norm": 0.19383686780929565,
+      "learning_rate": 2.260729666464612e-05,
+      "loss": 0.3632,
       "step": 75580
     },
     {
-      "epoch": 2.66,
-      "learning_rate": 2.3650078073931237e-05,
-      "loss": 0.2811,
+      "epoch": 2.724078278732836,
+      "grad_norm": 0.18293000757694244,
+      "learning_rate": 2.260439194316933e-05,
+      "loss": 0.3706,
       "step": 75585
     },
     {
-      "epoch": 2.66,
-      "learning_rate": 2.3647233517375317e-05,
-      "loss": 0.2783,
+      "epoch": 2.7242584783940607,
+      "grad_norm": 0.27218204736709595,
+      "learning_rate": 2.260148725433224e-05,
+      "loss": 0.4146,
       "step": 75590
     },
     {
-      "epoch": 2.66,
-      "learning_rate": 2.364438897838418e-05,
-      "loss": 0.2699,
+      "epoch": 2.7244386780552854,
+      "grad_norm": 0.22335471212863922,
+      "learning_rate": 2.259858259817441e-05,
+      "loss": 0.3762,
       "step": 75595
     },
     {
-      "epoch": 2.66,
-      "learning_rate": 2.3641544456994747e-05,
-      "loss": 0.2716,
+      "epoch": 2.7246188777165097,
+      "grad_norm": 0.19925439357757568,
+      "learning_rate": 2.259567797473544e-05,
+      "loss": 0.3917,
       "step": 75600
     },
     {
-      "epoch": 2.66,
-      "learning_rate": 2.363869995324397e-05,
-      "loss": 0.2684,
+      "epoch": 2.7247990773777344,
+      "grad_norm": 0.18860496580600739,
+      "learning_rate": 2.2592773384054883e-05,
+      "loss": 0.3926,
       "step": 75605
     },
     {
-      "epoch": 2.66,
-      "learning_rate": 2.3635855467168783e-05,
-      "loss": 0.292,
+      "epoch": 2.724979277038959,
+      "grad_norm": 0.23276196420192719,
+      "learning_rate": 2.258986882617233e-05,
+      "loss": 0.4083,
       "step": 75610
     },
     {
-      "epoch": 2.66,
-      "learning_rate": 2.3633010998806113e-05,
-      "loss": 0.2756,
+      "epoch": 2.725159476700184,
+      "grad_norm": 0.21294569969177246,
+      "learning_rate": 2.258696430112735e-05,
+      "loss": 0.392,
       "step": 75615
     },
     {
-      "epoch": 2.66,
-      "learning_rate": 2.363016654819288e-05,
-      "loss": 0.2436,
+      "epoch": 2.7253396763614086,
+      "grad_norm": 0.21455052495002747,
+      "learning_rate": 2.25840598089595e-05,
+      "loss": 0.4136,
       "step": 75620
     },
     {
-      "epoch": 2.66,
-      "learning_rate": 2.362732211536605e-05,
-      "loss": 0.2806,
+      "epoch": 2.7255198760226333,
+      "grad_norm": 0.2718030512332916,
+      "learning_rate": 2.2581736238923366e-05,
+      "loss": 0.4269,
       "step": 75625
     },
     {
-      "epoch": 2.66,
-      "learning_rate": 2.3624477700362526e-05,
-      "loss": 0.271,
+      "epoch": 2.7257000756838576,
+      "grad_norm": 0.21406759321689606,
+      "learning_rate": 2.2578831806034116e-05,
+      "loss": 0.3959,
       "step": 75630
     },
     {
-      "epoch": 2.66,
-      "learning_rate": 2.362163330321925e-05,
-      "loss": 0.2794,
+      "epoch": 2.7258802753450824,
+      "grad_norm": 0.21770311892032623,
+      "learning_rate": 2.2575927406132795e-05,
+      "loss": 0.3844,
       "step": 75635
     },
     {
-      "epoch": 2.66,
-      "learning_rate": 2.3618788923973167e-05,
-      "loss": 0.2581,
+      "epoch": 2.726060475006307,
+      "grad_norm": 0.22125771641731262,
+      "learning_rate": 2.2573023039259013e-05,
+      "loss": 0.3554,
       "step": 75640
     },
     {
-      "epoch": 2.66,
-      "learning_rate": 2.36159445626612e-05,
-      "loss": 0.2853,
+      "epoch": 2.7262406746675314,
+      "grad_norm": 0.2546500861644745,
+      "learning_rate": 2.2570118705452317e-05,
+      "loss": 0.4081,
       "step": 75645
     },
     {
-      "epoch": 2.66,
-      "learning_rate": 2.3613100219320282e-05,
-      "loss": 0.2975,
+      "epoch": 2.726420874328756,
+      "grad_norm": 0.19551582634449005,
+      "learning_rate": 2.2567214404752273e-05,
+      "loss": 0.4039,
       "step": 75650
     },
     {
-      "epoch": 2.66,
-      "learning_rate": 2.3610255893987328e-05,
-      "loss": 0.2612,
+      "epoch": 2.726601073989981,
+      "grad_norm": 0.21097378432750702,
+      "learning_rate": 2.2564310137198474e-05,
+      "loss": 0.4068,
       "step": 75655
     },
     {
-      "epoch": 2.66,
-      "learning_rate": 2.3607411586699296e-05,
-      "loss": 0.2624,
+      "epoch": 2.7267812736512056,
+      "grad_norm": 0.20981697738170624,
+      "learning_rate": 2.2561405902830464e-05,
+      "loss": 0.3682,
       "step": 75660
     },
     {
-      "epoch": 2.66,
-      "learning_rate": 2.360456729749311e-05,
-      "loss": 0.2454,
+      "epoch": 2.7269614733124303,
+      "grad_norm": 0.22551362216472626,
+      "learning_rate": 2.2558501701687847e-05,
+      "loss": 0.3908,
       "step": 75665
     },
     {
-      "epoch": 2.66,
-      "learning_rate": 2.36017230264057e-05,
-      "loss": 0.2618,
+      "epoch": 2.727141672973655,
+      "grad_norm": 0.23293738067150116,
+      "learning_rate": 2.255559753381016e-05,
+      "loss": 0.3983,
       "step": 75670
     },
     {
-      "epoch": 2.66,
-      "learning_rate": 2.3598878773473982e-05,
-      "loss": 0.2853,
+      "epoch": 2.7273218726348794,
+      "grad_norm": 0.19451351463794708,
+      "learning_rate": 2.2552693399236978e-05,
+      "loss": 0.3788,
       "step": 75675
     },
     {
-      "epoch": 2.66,
-      "learning_rate": 2.3596034538734908e-05,
-      "loss": 0.2617,
+      "epoch": 2.727502072296104,
+      "grad_norm": 0.22846585512161255,
+      "learning_rate": 2.2549789298007884e-05,
+      "loss": 0.3828,
       "step": 75680
     },
     {
-      "epoch": 2.66,
-      "learning_rate": 2.35931903222254e-05,
-      "loss": 0.2671,
+      "epoch": 2.727682271957329,
+      "grad_norm": 0.20074519515037537,
+      "learning_rate": 2.2546885230162435e-05,
+      "loss": 0.3768,
       "step": 75685
     },
     {
-      "epoch": 2.66,
-      "learning_rate": 2.3590346123982394e-05,
-      "loss": 0.2818,
+      "epoch": 2.727862471618553,
+      "grad_norm": 0.23105072975158691,
+      "learning_rate": 2.2543981195740194e-05,
+      "loss": 0.4379,
       "step": 75690
     },
     {
-      "epoch": 2.66,
-      "learning_rate": 2.35875019440428e-05,
-      "loss": 0.2639,
+      "epoch": 2.728042671279778,
+      "grad_norm": 0.20044337213039398,
+      "learning_rate": 2.254107719478074e-05,
+      "loss": 0.3861,
       "step": 75695
     },
     {
-      "epoch": 2.66,
-      "learning_rate": 2.3584657782443575e-05,
-      "loss": 0.2507,
+      "epoch": 2.7282228709410026,
+      "grad_norm": 0.1897222101688385,
+      "learning_rate": 2.2538173227323626e-05,
+      "loss": 0.3739,
       "step": 75700
     },
     {
-      "epoch": 2.66,
-      "learning_rate": 2.3581813639221637e-05,
-      "loss": 0.2559,
+      "epoch": 2.7284030706022273,
+      "grad_norm": 0.19628849625587463,
+      "learning_rate": 2.253526929340843e-05,
+      "loss": 0.3837,
       "step": 75705
     },
     {
-      "epoch": 2.66,
-      "learning_rate": 2.3578969514413908e-05,
-      "loss": 0.2618,
+      "epoch": 2.728583270263452,
+      "grad_norm": 0.21223433315753937,
+      "learning_rate": 2.2532365393074715e-05,
+      "loss": 0.4052,
       "step": 75710
     },
     {
-      "epoch": 2.66,
-      "learning_rate": 2.3576125408057317e-05,
-      "loss": 0.2597,
+      "epoch": 2.7287634699246768,
+      "grad_norm": 0.19551092386245728,
+      "learning_rate": 2.2529461526362037e-05,
+      "loss": 0.4148,
       "step": 75715
     },
     {
-      "epoch": 2.66,
-      "learning_rate": 2.3573281320188813e-05,
-      "loss": 0.2762,
+      "epoch": 2.728943669585901,
+      "grad_norm": 0.1927630603313446,
+      "learning_rate": 2.2526557693309974e-05,
+      "loss": 0.39,
       "step": 75720
     },
     {
-      "epoch": 2.66,
-      "learning_rate": 2.3570437250845306e-05,
-      "loss": 0.2716,
+      "epoch": 2.729123869247126,
+      "grad_norm": 0.21203365921974182,
+      "learning_rate": 2.252365389395809e-05,
+      "loss": 0.4016,
       "step": 75725
     },
     {
-      "epoch": 2.66,
-      "learning_rate": 2.3567593200063733e-05,
-      "loss": 0.2634,
+      "epoch": 2.7293040689083505,
+      "grad_norm": 0.18629692494869232,
+      "learning_rate": 2.252075012834592e-05,
+      "loss": 0.4215,
       "step": 75730
     },
     {
-      "epoch": 2.66,
-      "learning_rate": 2.3564749167881007e-05,
-      "loss": 0.2626,
+      "epoch": 2.729484268569575,
+      "grad_norm": 0.24014006555080414,
+      "learning_rate": 2.251784639651307e-05,
+      "loss": 0.4002,
       "step": 75735
     },
     {
-      "epoch": 2.66,
-      "learning_rate": 2.356190515433407e-05,
-      "loss": 0.2668,
+      "epoch": 2.7296644682307996,
+      "grad_norm": 0.17442671954631805,
+      "learning_rate": 2.2514942698499067e-05,
+      "loss": 0.4086,
       "step": 75740
     },
     {
-      "epoch": 2.66,
-      "learning_rate": 2.3559061159459854e-05,
-      "loss": 0.2505,
+      "epoch": 2.7298446678920243,
+      "grad_norm": 0.23989979922771454,
+      "learning_rate": 2.2512039034343504e-05,
+      "loss": 0.3988,
       "step": 75745
     },
     {
-      "epoch": 2.67,
-      "learning_rate": 2.355621718329528e-05,
-      "loss": 0.2781,
+      "epoch": 2.730024867553249,
+      "grad_norm": 0.23518751561641693,
+      "learning_rate": 2.250913540408592e-05,
+      "loss": 0.3873,
       "step": 75750
     },
     {
-      "epoch": 2.67,
-      "learning_rate": 2.3553373225877263e-05,
-      "loss": 0.2879,
+      "epoch": 2.7302050672144738,
+      "grad_norm": 0.23601964116096497,
+      "learning_rate": 2.2506231807765882e-05,
+      "loss": 0.3869,
       "step": 75755
     },
     {
-      "epoch": 2.67,
-      "learning_rate": 2.355052928724275e-05,
-      "loss": 0.2771,
+      "epoch": 2.7303852668756985,
+      "grad_norm": 0.269535630941391,
+      "learning_rate": 2.2503328245422957e-05,
+      "loss": 0.3967,
       "step": 75760
     },
     {
-      "epoch": 2.67,
-      "learning_rate": 2.3547685367428656e-05,
-      "loss": 0.2718,
+      "epoch": 2.730565466536923,
+      "grad_norm": 0.20272400975227356,
+      "learning_rate": 2.2500424717096702e-05,
+      "loss": 0.3563,
       "step": 75765
     },
     {
-      "epoch": 2.67,
-      "learning_rate": 2.354484146647191e-05,
-      "loss": 0.2845,
+      "epoch": 2.7307456661981475,
+      "grad_norm": 0.19777542352676392,
+      "learning_rate": 2.2497521222826667e-05,
+      "loss": 0.4028,
       "step": 75770
     },
     {
-      "epoch": 2.67,
-      "learning_rate": 2.3541997584409438e-05,
-      "loss": 0.2798,
+      "epoch": 2.7309258658593722,
+      "grad_norm": 0.18318623304367065,
+      "learning_rate": 2.2494617762652433e-05,
+      "loss": 0.3846,
       "step": 75775
     },
     {
-      "epoch": 2.67,
-      "learning_rate": 2.353915372127817e-05,
-      "loss": 0.2676,
+      "epoch": 2.731106065520597,
+      "grad_norm": 0.17845581471920013,
+      "learning_rate": 2.2491714336613534e-05,
+      "loss": 0.4036,
       "step": 75780
     },
     {
-      "epoch": 2.67,
-      "learning_rate": 2.3536309877115033e-05,
-      "loss": 0.2759,
+      "epoch": 2.7312862651818213,
+      "grad_norm": 0.20689518749713898,
+      "learning_rate": 2.248881094474955e-05,
+      "loss": 0.3367,
       "step": 75785
     },
     {
-      "epoch": 2.67,
-      "learning_rate": 2.3533466051956938e-05,
-      "loss": 0.2623,
+      "epoch": 2.731466464843046,
+      "grad_norm": 0.22084331512451172,
+      "learning_rate": 2.2485907587100034e-05,
+      "loss": 0.3819,
       "step": 75790
     },
     {
-      "epoch": 2.67,
-      "learning_rate": 2.3530622245840818e-05,
-      "loss": 0.2667,
+      "epoch": 2.7316466645042707,
+      "grad_norm": 0.2153303027153015,
+      "learning_rate": 2.248300426370453e-05,
+      "loss": 0.3989,
       "step": 75795
     },
     {
-      "epoch": 2.67,
-      "learning_rate": 2.352777845880361e-05,
-      "loss": 0.2681,
+      "epoch": 2.7318268641654955,
+      "grad_norm": 0.23561865091323853,
+      "learning_rate": 2.2480100974602613e-05,
+      "loss": 0.3838,
       "step": 75800
     },
     {
-      "epoch": 2.67,
-      "learning_rate": 2.352493469088223e-05,
-      "loss": 0.2697,
+      "epoch": 2.73200706382672,
+      "grad_norm": 0.24491235613822937,
+      "learning_rate": 2.247719771983384e-05,
+      "loss": 0.3881,
       "step": 75805
     },
     {
-      "epoch": 2.67,
-      "learning_rate": 2.3522659690333325e-05,
-      "loss": 0.2463,
+      "epoch": 2.7321872634879445,
+      "grad_norm": 0.1986062377691269,
+      "learning_rate": 2.247429449943774e-05,
+      "loss": 0.3913,
       "step": 75810
     },
     {
-      "epoch": 2.67,
-      "learning_rate": 2.3519815956913475e-05,
-      "loss": 0.2921,
+      "epoch": 2.7323674631491692,
+      "grad_norm": 0.20403479039669037,
+      "learning_rate": 2.24713913134539e-05,
+      "loss": 0.3769,
       "step": 75815
     },
     {
-      "epoch": 2.67,
-      "learning_rate": 2.3516972242712848e-05,
-      "loss": 0.2793,
+      "epoch": 2.732547662810394,
+      "grad_norm": 0.232151597738266,
+      "learning_rate": 2.2468488161921858e-05,
+      "loss": 0.4194,
       "step": 75820
     },
     {
-      "epoch": 2.67,
-      "learning_rate": 2.3514128547768342e-05,
-      "loss": 0.2738,
+      "epoch": 2.7327278624716187,
+      "grad_norm": 0.19631794095039368,
+      "learning_rate": 2.2465585044881182e-05,
+      "loss": 0.4294,
       "step": 75825
     },
     {
-      "epoch": 2.67,
-      "learning_rate": 2.3511284872116906e-05,
-      "loss": 0.2991,
+      "epoch": 2.732908062132843,
+      "grad_norm": 0.20293517410755157,
+      "learning_rate": 2.2462681962371424e-05,
+      "loss": 0.3644,
       "step": 75830
     },
     {
-      "epoch": 2.67,
-      "learning_rate": 2.350844121579546e-05,
-      "loss": 0.2796,
+      "epoch": 2.7330882617940677,
+      "grad_norm": 0.19923901557922363,
+      "learning_rate": 2.2459778914432116e-05,
+      "loss": 0.3929,
       "step": 75835
     },
     {
-      "epoch": 2.67,
-      "learning_rate": 2.350559757884091e-05,
-      "loss": 0.2683,
+      "epoch": 2.7332684614552925,
+      "grad_norm": 0.24110738933086395,
+      "learning_rate": 2.2456875901102845e-05,
+      "loss": 0.3879,
       "step": 75840
     },
     {
-      "epoch": 2.67,
-      "learning_rate": 2.3502753961290204e-05,
-      "loss": 0.2594,
+      "epoch": 2.733448661116517,
+      "grad_norm": 0.21238499879837036,
+      "learning_rate": 2.245397292242313e-05,
+      "loss": 0.3906,
       "step": 75845
     },
     {
-      "epoch": 2.67,
-      "learning_rate": 2.3499910363180246e-05,
-      "loss": 0.2886,
+      "epoch": 2.733628860777742,
+      "grad_norm": 0.23390400409698486,
+      "learning_rate": 2.245106997843256e-05,
+      "loss": 0.4043,
       "step": 75850
     },
     {
-      "epoch": 2.67,
-      "learning_rate": 2.349706678454796e-05,
-      "loss": 0.2805,
+      "epoch": 2.733809060438966,
+      "grad_norm": 0.17990170419216156,
+      "learning_rate": 2.244816706917066e-05,
+      "loss": 0.4187,
       "step": 75855
     },
     {
-      "epoch": 2.67,
-      "learning_rate": 2.3494223225430266e-05,
-      "loss": 0.2628,
+      "epoch": 2.733989260100191,
+      "grad_norm": 0.20971287786960602,
+      "learning_rate": 2.244526419467699e-05,
+      "loss": 0.3834,
       "step": 75860
     },
     {
-      "epoch": 2.67,
-      "learning_rate": 2.3491379685864105e-05,
-      "loss": 0.272,
+      "epoch": 2.7341694597614157,
+      "grad_norm": 0.21100729703903198,
+      "learning_rate": 2.24423613549911e-05,
+      "loss": 0.3752,
       "step": 75865
     },
     {
-      "epoch": 2.67,
-      "learning_rate": 2.3488536165886384e-05,
-      "loss": 0.2463,
+      "epoch": 2.7343496594226404,
+      "grad_norm": 0.19430191814899445,
+      "learning_rate": 2.2439458550152544e-05,
+      "loss": 0.4059,
       "step": 75870
     },
     {
-      "epoch": 2.67,
-      "learning_rate": 2.348569266553402e-05,
-      "loss": 0.2638,
+      "epoch": 2.7345298590838647,
+      "grad_norm": 0.1902761608362198,
+      "learning_rate": 2.243655578020086e-05,
+      "loss": 0.4116,
       "step": 75875
     },
     {
-      "epoch": 2.67,
-      "learning_rate": 2.348284918484393e-05,
-      "loss": 0.2661,
+      "epoch": 2.7347100587450894,
+      "grad_norm": 0.20415811240673065,
+      "learning_rate": 2.2433653045175614e-05,
+      "loss": 0.3747,
       "step": 75880
     },
     {
-      "epoch": 2.67,
-      "learning_rate": 2.348000572385305e-05,
-      "loss": 0.2953,
+      "epoch": 2.734890258406314,
+      "grad_norm": 0.21149209141731262,
+      "learning_rate": 2.2430750345116346e-05,
+      "loss": 0.3851,
       "step": 75885
     },
     {
-      "epoch": 2.67,
-      "learning_rate": 2.34771622825983e-05,
-      "loss": 0.2886,
+      "epoch": 2.735070458067539,
+      "grad_norm": 0.2101927250623703,
+      "learning_rate": 2.2427847680062612e-05,
+      "loss": 0.4146,
       "step": 75890
     },
     {
-      "epoch": 2.67,
-      "learning_rate": 2.3474318861116594e-05,
-      "loss": 0.2792,
+      "epoch": 2.7352506577287636,
+      "grad_norm": 0.18154318630695343,
+      "learning_rate": 2.2424945050053954e-05,
+      "loss": 0.3912,
       "step": 75895
     },
     {
-      "epoch": 2.67,
-      "learning_rate": 2.347147545944484e-05,
-      "loss": 0.2792,
+      "epoch": 2.7354308573899884,
+      "grad_norm": 0.2463860660791397,
+      "learning_rate": 2.2422042455129916e-05,
+      "loss": 0.3867,
       "step": 75900
     },
     {
-      "epoch": 2.67,
-      "learning_rate": 2.346863207761998e-05,
-      "loss": 0.2595,
+      "epoch": 2.7356110570512127,
+      "grad_norm": 0.23593199253082275,
+      "learning_rate": 2.2419139895330058e-05,
+      "loss": 0.4111,
       "step": 75905
     },
     {
-      "epoch": 2.67,
-      "learning_rate": 2.3465788715678922e-05,
-      "loss": 0.263,
+      "epoch": 2.7357912567124374,
+      "grad_norm": 0.24448728561401367,
+      "learning_rate": 2.2416237370693922e-05,
+      "loss": 0.4308,
       "step": 75910
     },
     {
-      "epoch": 2.67,
-      "learning_rate": 2.3462945373658588e-05,
-      "loss": 0.2568,
+      "epoch": 2.735971456373662,
+      "grad_norm": 0.21843284368515015,
+      "learning_rate": 2.2413334881261038e-05,
+      "loss": 0.3489,
       "step": 75915
     },
     {
-      "epoch": 2.67,
-      "learning_rate": 2.3460102051595885e-05,
-      "loss": 0.2462,
+      "epoch": 2.7361516560348864,
+      "grad_norm": 0.20660178363323212,
+      "learning_rate": 2.2410432427070975e-05,
+      "loss": 0.3702,
       "step": 75920
     },
     {
-      "epoch": 2.67,
-      "learning_rate": 2.345725874952775e-05,
-      "loss": 0.2756,
+      "epoch": 2.736331855696111,
+      "grad_norm": 0.19898898899555206,
+      "learning_rate": 2.240753000816326e-05,
+      "loss": 0.3856,
       "step": 75925
     },
     {
-      "epoch": 2.67,
-      "learning_rate": 2.3454415467491102e-05,
-      "loss": 0.2813,
+      "epoch": 2.736512055357336,
+      "grad_norm": 0.2137184739112854,
+      "learning_rate": 2.240462762457746e-05,
+      "loss": 0.3925,
       "step": 75930
     },
     {
-      "epoch": 2.67,
-      "learning_rate": 2.345157220552284e-05,
-      "loss": 0.259,
+      "epoch": 2.7366922550185606,
+      "grad_norm": 0.21520856022834778,
+      "learning_rate": 2.2401725276353103e-05,
+      "loss": 0.4215,
       "step": 75935
     },
     {
-      "epoch": 2.67,
-      "learning_rate": 2.3448728963659885e-05,
-      "loss": 0.2697,
+      "epoch": 2.7368724546797853,
+      "grad_norm": 0.22829484939575195,
+      "learning_rate": 2.2398822963529722e-05,
+      "loss": 0.3858,
       "step": 75940
     },
     {
-      "epoch": 2.67,
-      "learning_rate": 2.3445885741939175e-05,
-      "loss": 0.2627,
+      "epoch": 2.73705265434101,
+      "grad_norm": 0.19753465056419373,
+      "learning_rate": 2.2395920686146894e-05,
+      "loss": 0.418,
       "step": 75945
     },
     {
-      "epoch": 2.67,
-      "learning_rate": 2.3443042540397615e-05,
-      "loss": 0.2602,
+      "epoch": 2.7372328540022344,
+      "grad_norm": 0.19238927960395813,
+      "learning_rate": 2.2393018444244132e-05,
+      "loss": 0.4194,
       "step": 75950
     },
     {
-      "epoch": 2.67,
-      "learning_rate": 2.344019935907211e-05,
-      "loss": 0.2678,
+      "epoch": 2.737413053663459,
+      "grad_norm": 0.2262929379940033,
+      "learning_rate": 2.2390116237860985e-05,
+      "loss": 0.373,
       "step": 75955
     },
     {
-      "epoch": 2.67,
-      "learning_rate": 2.34373561979996e-05,
-      "loss": 0.2482,
+      "epoch": 2.737593253324684,
+      "grad_norm": 0.1645156592130661,
+      "learning_rate": 2.238721406703701e-05,
+      "loss": 0.3634,
       "step": 75960
     },
     {
-      "epoch": 2.67,
-      "learning_rate": 2.3434513057216984e-05,
-      "loss": 0.2637,
+      "epoch": 2.737773452985908,
+      "grad_norm": 0.20983080565929413,
+      "learning_rate": 2.2384311931811728e-05,
+      "loss": 0.3576,
       "step": 75965
     },
     {
-      "epoch": 2.67,
-      "learning_rate": 2.343166993676119e-05,
-      "loss": 0.2795,
+      "epoch": 2.737953652647133,
+      "grad_norm": 0.21075214445590973,
+      "learning_rate": 2.238199024928914e-05,
+      "loss": 0.4086,
       "step": 75970
     },
     {
-      "epoch": 2.67,
-      "learning_rate": 2.3428826836669116e-05,
-      "loss": 0.277,
+      "epoch": 2.7381338523083576,
+      "grad_norm": 0.2087734490633011,
+      "learning_rate": 2.237908817824117e-05,
+      "loss": 0.428,
       "step": 75975
     },
     {
-      "epoch": 2.67,
-      "learning_rate": 2.34259837569777e-05,
-      "loss": 0.246,
+      "epoch": 2.7383140519695823,
+      "grad_norm": 0.22745364904403687,
+      "learning_rate": 2.2376186142902626e-05,
+      "loss": 0.4003,
       "step": 75980
     },
     {
-      "epoch": 2.67,
-      "learning_rate": 2.342314069772385e-05,
-      "loss": 0.2809,
+      "epoch": 2.738494251630807,
+      "grad_norm": 0.214681476354599,
+      "learning_rate": 2.2373284143313015e-05,
+      "loss": 0.3532,
       "step": 75985
     },
     {
-      "epoch": 2.67,
-      "learning_rate": 2.3420297658944472e-05,
-      "loss": 0.267,
+      "epoch": 2.738674451292032,
+      "grad_norm": 0.25850626826286316,
+      "learning_rate": 2.2370382179511915e-05,
+      "loss": 0.3884,
       "step": 75990
     },
     {
-      "epoch": 2.67,
-      "learning_rate": 2.341745464067649e-05,
-      "loss": 0.274,
+      "epoch": 2.738854650953256,
+      "grad_norm": 0.2193749099969864,
+      "learning_rate": 2.2367480251538842e-05,
+      "loss": 0.3821,
       "step": 75995
     },
     {
-      "epoch": 2.67,
-      "learning_rate": 2.3414611642956816e-05,
-      "loss": 0.2866,
+      "epoch": 2.739034850614481,
+      "grad_norm": 0.24161560833454132,
+      "learning_rate": 2.2364578359433345e-05,
+      "loss": 0.3848,
       "step": 76000
     },
     {
-      "epoch": 2.67,
-      "eval_loss": 0.2612769603729248,
-      "eval_runtime": 10.5493,
-      "eval_samples_per_second": 9.479,
-      "eval_steps_per_second": 9.479,
+      "epoch": 2.739034850614481,
+      "eval_loss": 0.43177616596221924,
+      "eval_runtime": 3.531,
+      "eval_samples_per_second": 28.321,
+      "eval_steps_per_second": 7.08,
       "step": 76000
     },
     {
-      "epoch": 2.67,
-      "learning_rate": 2.341176866582237e-05,
-      "loss": 0.2699,
+      "epoch": 2.7392150502757056,
+      "grad_norm": 0.21785877645015717,
+      "learning_rate": 2.2361676503234963e-05,
+      "loss": 0.3629,
       "step": 76005
     },
     {
-      "epoch": 2.67,
-      "learning_rate": 2.3408925709310063e-05,
-      "loss": 0.2581,
+      "epoch": 2.73939524993693,
+      "grad_norm": 0.24404390156269073,
+      "learning_rate": 2.2358774682983213e-05,
+      "loss": 0.4146,
       "step": 76010
     },
     {
-      "epoch": 2.67,
-      "learning_rate": 2.3406082773456792e-05,
-      "loss": 0.2552,
+      "epoch": 2.7395754495981546,
+      "grad_norm": 0.20576035976409912,
+      "learning_rate": 2.235587289871766e-05,
+      "loss": 0.4005,
       "step": 76015
     },
     {
-      "epoch": 2.67,
-      "learning_rate": 2.340323985829949e-05,
-      "loss": 0.259,
+      "epoch": 2.7397556492593793,
+      "grad_norm": 0.1887839436531067,
+      "learning_rate": 2.2352971150477824e-05,
+      "loss": 0.3787,
       "step": 76020
     },
     {
-      "epoch": 2.67,
-      "learning_rate": 2.340039696387508e-05,
-      "loss": 0.2766,
+      "epoch": 2.739935848920604,
+      "grad_norm": 0.20635947585105896,
+      "learning_rate": 2.235006943830324e-05,
+      "loss": 0.3778,
       "step": 76025
     },
     {
-      "epoch": 2.67,
-      "learning_rate": 2.3397554090220456e-05,
-      "loss": 0.2715,
+      "epoch": 2.740116048581829,
+      "grad_norm": 0.21008598804473877,
+      "learning_rate": 2.234716776223345e-05,
+      "loss": 0.3746,
       "step": 76030
     },
     {
-      "epoch": 2.68,
-      "learning_rate": 2.3394711237372525e-05,
-      "loss": 0.2591,
+      "epoch": 2.7402962482430535,
+      "grad_norm": 0.2424369752407074,
+      "learning_rate": 2.234426612230798e-05,
+      "loss": 0.4026,
       "step": 76035
     },
     {
-      "epoch": 2.68,
-      "learning_rate": 2.3391868405368218e-05,
-      "loss": 0.2483,
+      "epoch": 2.740476447904278,
+      "grad_norm": 0.2240283191204071,
+      "learning_rate": 2.2341364518566378e-05,
+      "loss": 0.4191,
       "step": 76040
     },
     {
-      "epoch": 2.68,
-      "learning_rate": 2.338902559424444e-05,
-      "loss": 0.2635,
+      "epoch": 2.7406566475655025,
+      "grad_norm": 0.22698961198329926,
+      "learning_rate": 2.2338462951048165e-05,
+      "loss": 0.4108,
       "step": 76045
     },
     {
-      "epoch": 2.68,
-      "learning_rate": 2.3386182804038108e-05,
-      "loss": 0.2694,
+      "epoch": 2.7408368472267273,
+      "grad_norm": 0.1691308468580246,
+      "learning_rate": 2.2335561419792876e-05,
+      "loss": 0.3827,
       "step": 76050
     },
     {
-      "epoch": 2.68,
-      "learning_rate": 2.3383340034786118e-05,
-      "loss": 0.2895,
+      "epoch": 2.741017046887952,
+      "grad_norm": 0.18437933921813965,
+      "learning_rate": 2.233265992484005e-05,
+      "loss": 0.3761,
       "step": 76055
     },
     {
-      "epoch": 2.68,
-      "learning_rate": 2.33804972865254e-05,
-      "loss": 0.2687,
+      "epoch": 2.7411972465491763,
+      "grad_norm": 0.23850589990615845,
+      "learning_rate": 2.232975846622922e-05,
+      "loss": 0.4086,
       "step": 76060
     },
     {
-      "epoch": 2.68,
-      "learning_rate": 2.3377654559292862e-05,
-      "loss": 0.279,
+      "epoch": 2.741377446210401,
+      "grad_norm": 0.20492541790008545,
+      "learning_rate": 2.2326857043999906e-05,
+      "loss": 0.3932,
       "step": 76065
     },
     {
-      "epoch": 2.68,
-      "learning_rate": 2.3374811853125393e-05,
-      "loss": 0.275,
+      "epoch": 2.7415576458716258,
+      "grad_norm": 0.219580739736557,
+      "learning_rate": 2.2323955658191653e-05,
+      "loss": 0.3816,
       "step": 76070
     },
     {
-      "epoch": 2.68,
-      "learning_rate": 2.3371969168059937e-05,
-      "loss": 0.2605,
+      "epoch": 2.7417378455328505,
+      "grad_norm": 0.23200367391109467,
+      "learning_rate": 2.232105430884398e-05,
+      "loss": 0.3838,
       "step": 76075
     },
     {
-      "epoch": 2.68,
-      "learning_rate": 2.336912650413338e-05,
-      "loss": 0.2477,
+      "epoch": 2.7419180451940752,
+      "grad_norm": 0.24806389212608337,
+      "learning_rate": 2.231815299599643e-05,
+      "loss": 0.4068,
       "step": 76080
     },
     {
-      "epoch": 2.68,
-      "learning_rate": 2.3366283861382648e-05,
-      "loss": 0.2602,
+      "epoch": 2.7420982448552995,
+      "grad_norm": 0.17643065750598907,
+      "learning_rate": 2.2315251719688536e-05,
+      "loss": 0.3776,
       "step": 76085
     },
     {
-      "epoch": 2.68,
-      "learning_rate": 2.3363441239844632e-05,
-      "loss": 0.2465,
+      "epoch": 2.7422784445165242,
+      "grad_norm": 0.2253817468881607,
+      "learning_rate": 2.231235047995979e-05,
+      "loss": 0.3983,
       "step": 76090
     },
     {
-      "epoch": 2.68,
-      "learning_rate": 2.3360598639556268e-05,
-      "loss": 0.2853,
+      "epoch": 2.742458644177749,
+      "grad_norm": 0.24909770488739014,
+      "learning_rate": 2.2309449276849775e-05,
+      "loss": 0.4178,
       "step": 76095
     },
     {
-      "epoch": 2.68,
-      "learning_rate": 2.3357756060554445e-05,
-      "loss": 0.2683,
+      "epoch": 2.7426388438389737,
+      "grad_norm": 0.21218940615653992,
+      "learning_rate": 2.2306548110397973e-05,
+      "loss": 0.3915,
       "step": 76100
     },
     {
-      "epoch": 2.68,
-      "learning_rate": 2.3354913502876072e-05,
-      "loss": 0.282,
+      "epoch": 2.742819043500198,
+      "grad_norm": 0.19259501993656158,
+      "learning_rate": 2.2303646980643948e-05,
+      "loss": 0.3977,
       "step": 76105
     },
     {
-      "epoch": 2.68,
-      "learning_rate": 2.335207096655806e-05,
-      "loss": 0.2867,
+      "epoch": 2.7429992431614227,
+      "grad_norm": 0.22517803311347961,
+      "learning_rate": 2.2300745887627206e-05,
+      "loss": 0.383,
       "step": 76110
     },
     {
-      "epoch": 2.68,
-      "learning_rate": 2.3349228451637327e-05,
-      "loss": 0.2511,
+      "epoch": 2.7431794428226475,
+      "grad_norm": 0.2479083389043808,
+      "learning_rate": 2.2297844831387265e-05,
+      "loss": 0.392,
       "step": 76115
     },
     {
-      "epoch": 2.68,
-      "learning_rate": 2.3346385958150778e-05,
-      "loss": 0.2747,
+      "epoch": 2.743359642483872,
+      "grad_norm": 0.2069377303123474,
+      "learning_rate": 2.2294943811963682e-05,
+      "loss": 0.4116,
       "step": 76120
     },
     {
-      "epoch": 2.68,
-      "learning_rate": 2.3343543486135314e-05,
-      "loss": 0.2776,
+      "epoch": 2.743539842145097,
+      "grad_norm": 0.21161840856075287,
+      "learning_rate": 2.2292042829395964e-05,
+      "loss": 0.4078,
       "step": 76125
     },
     {
-      "epoch": 2.68,
-      "learning_rate": 2.3340701035627843e-05,
-      "loss": 0.2435,
+      "epoch": 2.7437200418063217,
+      "grad_norm": 0.24700872600078583,
+      "learning_rate": 2.228914188372363e-05,
+      "loss": 0.4206,
       "step": 76130
     },
     {
-      "epoch": 2.68,
-      "learning_rate": 2.333785860666528e-05,
-      "loss": 0.2765,
+      "epoch": 2.743900241467546,
+      "grad_norm": 0.19092321395874023,
+      "learning_rate": 2.2286240974986218e-05,
+      "loss": 0.4151,
       "step": 76135
     },
     {
-      "epoch": 2.68,
-      "learning_rate": 2.3335016199284528e-05,
-      "loss": 0.2888,
+      "epoch": 2.7440804411287707,
+      "grad_norm": 0.20470459759235382,
+      "learning_rate": 2.2283340103223243e-05,
+      "loss": 0.4057,
       "step": 76140
     },
     {
-      "epoch": 2.68,
-      "learning_rate": 2.3332173813522497e-05,
-      "loss": 0.2604,
+      "epoch": 2.7442606407899954,
+      "grad_norm": 0.18111029267311096,
+      "learning_rate": 2.2280439268474236e-05,
+      "loss": 0.3995,
       "step": 76145
     },
     {
-      "epoch": 2.68,
-      "learning_rate": 2.3329331449416077e-05,
-      "loss": 0.2646,
+      "epoch": 2.7444408404512197,
+      "grad_norm": 0.20974625647068024,
+      "learning_rate": 2.2277538470778722e-05,
+      "loss": 0.3997,
       "step": 76150
     },
     {
-      "epoch": 2.68,
-      "learning_rate": 2.3326489107002194e-05,
-      "loss": 0.2574,
+      "epoch": 2.7446210401124445,
+      "grad_norm": 0.24757510423660278,
+      "learning_rate": 2.227463771017621e-05,
+      "loss": 0.4023,
       "step": 76155
     },
     {
-      "epoch": 2.68,
-      "learning_rate": 2.3323646786317748e-05,
-      "loss": 0.2908,
+      "epoch": 2.744801239773669,
+      "grad_norm": 0.2829555869102478,
+      "learning_rate": 2.227173698670624e-05,
+      "loss": 0.3991,
       "step": 76160
     },
     {
-      "epoch": 2.68,
-      "learning_rate": 2.3320804487399647e-05,
-      "loss": 0.2683,
+      "epoch": 2.744981439434894,
+      "grad_norm": 0.1985725462436676,
+      "learning_rate": 2.2268836300408323e-05,
+      "loss": 0.4086,
       "step": 76165
     },
     {
-      "epoch": 2.68,
-      "learning_rate": 2.3317962210284784e-05,
-      "loss": 0.2552,
+      "epoch": 2.7451616390961187,
+      "grad_norm": 0.23659586906433105,
+      "learning_rate": 2.226593565132198e-05,
+      "loss": 0.3794,
       "step": 76170
     },
     {
-      "epoch": 2.68,
-      "learning_rate": 2.331511995501008e-05,
-      "loss": 0.2676,
+      "epoch": 2.7453418387573434,
+      "grad_norm": 0.21936027705669403,
+      "learning_rate": 2.2263035039486734e-05,
+      "loss": 0.3971,
       "step": 76175
     },
     {
-      "epoch": 2.68,
-      "learning_rate": 2.3312277721612434e-05,
-      "loss": 0.2623,
+      "epoch": 2.7455220384185677,
+      "grad_norm": 0.21939417719841003,
+      "learning_rate": 2.2260134464942108e-05,
+      "loss": 0.4245,
       "step": 76180
     },
     {
-      "epoch": 2.68,
-      "learning_rate": 2.330943551012874e-05,
-      "loss": 0.2738,
+      "epoch": 2.7457022380797924,
+      "grad_norm": 0.2699432969093323,
+      "learning_rate": 2.225723392772762e-05,
+      "loss": 0.4004,
       "step": 76185
     },
     {
-      "epoch": 2.68,
-      "learning_rate": 2.3306593320595918e-05,
-      "loss": 0.2851,
+      "epoch": 2.745882437741017,
+      "grad_norm": 0.20219436287879944,
+      "learning_rate": 2.2254333427882795e-05,
+      "loss": 0.4058,
       "step": 76190
     },
     {
-      "epoch": 2.68,
-      "learning_rate": 2.330375115305087e-05,
-      "loss": 0.2665,
+      "epoch": 2.7460626374022414,
+      "grad_norm": 0.2475823611021042,
+      "learning_rate": 2.2251432965447126e-05,
+      "loss": 0.4338,
       "step": 76195
     },
     {
-      "epoch": 2.68,
-      "learning_rate": 2.3300909007530493e-05,
-      "loss": 0.2861,
+      "epoch": 2.746242837063466,
+      "grad_norm": 0.18280702829360962,
+      "learning_rate": 2.2248532540460173e-05,
+      "loss": 0.3821,
       "step": 76200
     },
     {
-      "epoch": 2.68,
-      "learning_rate": 2.3298066884071686e-05,
-      "loss": 0.2728,
+      "epoch": 2.746423036724691,
+      "grad_norm": 0.2569371461868286,
+      "learning_rate": 2.2245632152961425e-05,
+      "loss": 0.4106,
       "step": 76205
     },
     {
-      "epoch": 2.68,
-      "learning_rate": 2.329522478271137e-05,
-      "loss": 0.2538,
+      "epoch": 2.7466032363859156,
+      "grad_norm": 0.22354593873023987,
+      "learning_rate": 2.2242731802990396e-05,
+      "loss": 0.3885,
       "step": 76210
     },
     {
-      "epoch": 2.68,
-      "learning_rate": 2.3292382703486428e-05,
-      "loss": 0.2773,
+      "epoch": 2.7467834360471404,
+      "grad_norm": 0.17847710847854614,
+      "learning_rate": 2.223983149058662e-05,
+      "loss": 0.4149,
       "step": 76215
     },
     {
-      "epoch": 2.68,
-      "learning_rate": 2.3289540646433777e-05,
-      "loss": 0.2909,
+      "epoch": 2.746963635708365,
+      "grad_norm": 0.22239890694618225,
+      "learning_rate": 2.2236931215789604e-05,
+      "loss": 0.3831,
       "step": 76220
     },
     {
-      "epoch": 2.68,
-      "learning_rate": 2.3286698611590304e-05,
-      "loss": 0.2853,
+      "epoch": 2.7471438353695894,
+      "grad_norm": 0.22584442794322968,
+      "learning_rate": 2.2234030978638865e-05,
+      "loss": 0.3784,
       "step": 76225
     },
     {
-      "epoch": 2.68,
-      "learning_rate": 2.328385659899293e-05,
-      "loss": 0.2606,
+      "epoch": 2.747324035030814,
+      "grad_norm": 0.2300465852022171,
+      "learning_rate": 2.223113077917392e-05,
+      "loss": 0.3863,
       "step": 76230
     },
     {
-      "epoch": 2.68,
-      "learning_rate": 2.328101460867855e-05,
-      "loss": 0.2718,
+      "epoch": 2.747504234692039,
+      "grad_norm": 0.21806028485298157,
+      "learning_rate": 2.2228230617434276e-05,
+      "loss": 0.408,
       "step": 76235
     },
     {
-      "epoch": 2.68,
-      "learning_rate": 2.3278172640684056e-05,
-      "loss": 0.2591,
+      "epoch": 2.747684434353263,
+      "grad_norm": 0.26781561970710754,
+      "learning_rate": 2.222533049345946e-05,
+      "loss": 0.4062,
       "step": 76240
     },
     {
-      "epoch": 2.68,
-      "learning_rate": 2.327533069504635e-05,
-      "loss": 0.264,
+      "epoch": 2.747864634014488,
+      "grad_norm": 0.23583123087882996,
+      "learning_rate": 2.222243040728898e-05,
+      "loss": 0.3975,
       "step": 76245
     },
     {
-      "epoch": 2.68,
-      "learning_rate": 2.327248877180235e-05,
-      "loss": 0.2514,
+      "epoch": 2.7480448336757126,
+      "grad_norm": 0.1862766593694687,
+      "learning_rate": 2.221953035896234e-05,
+      "loss": 0.4025,
       "step": 76250
     },
     {
-      "epoch": 2.68,
-      "learning_rate": 2.3269646870988944e-05,
-      "loss": 0.2671,
+      "epoch": 2.7482250333369374,
+      "grad_norm": 0.21025675535202026,
+      "learning_rate": 2.2216630348519067e-05,
+      "loss": 0.4092,
       "step": 76255
     },
     {
-      "epoch": 2.68,
-      "learning_rate": 2.326680499264303e-05,
-      "loss": 0.2581,
+      "epoch": 2.748405232998162,
+      "grad_norm": 0.1862116903066635,
+      "learning_rate": 2.2213730375998663e-05,
+      "loss": 0.398,
       "step": 76260
     },
     {
-      "epoch": 2.68,
-      "learning_rate": 2.3263963136801505e-05,
-      "loss": 0.281,
+      "epoch": 2.748585432659387,
+      "grad_norm": 0.18690145015716553,
+      "learning_rate": 2.2210830441440646e-05,
+      "loss": 0.3963,
       "step": 76265
     },
     {
-      "epoch": 2.68,
-      "learning_rate": 2.3261121303501274e-05,
-      "loss": 0.2668,
+      "epoch": 2.748765632320611,
+      "grad_norm": 0.20919784903526306,
+      "learning_rate": 2.220793054488453e-05,
+      "loss": 0.4236,
       "step": 76270
     },
     {
-      "epoch": 2.68,
-      "learning_rate": 2.325827949277925e-05,
-      "loss": 0.2715,
+      "epoch": 2.748945831981836,
+      "grad_norm": 0.24846714735031128,
+      "learning_rate": 2.2205030686369805e-05,
+      "loss": 0.3915,
       "step": 76275
     },
     {
-      "epoch": 2.68,
-      "learning_rate": 2.3255437704672312e-05,
-      "loss": 0.2676,
+      "epoch": 2.7491260316430606,
+      "grad_norm": 0.19307979941368103,
+      "learning_rate": 2.2202130865936006e-05,
+      "loss": 0.3978,
       "step": 76280
     },
     {
-      "epoch": 2.68,
-      "learning_rate": 2.3252595939217357e-05,
-      "loss": 0.3085,
+      "epoch": 2.7493062313042853,
+      "grad_norm": 0.2004070281982422,
+      "learning_rate": 2.2199231083622627e-05,
+      "loss": 0.4069,
       "step": 76285
     },
     {
-      "epoch": 2.68,
-      "learning_rate": 2.3249754196451304e-05,
-      "loss": 0.2605,
+      "epoch": 2.7494864309655096,
+      "grad_norm": 0.20964334905147552,
+      "learning_rate": 2.2196331339469187e-05,
+      "loss": 0.3888,
       "step": 76290
     },
     {
-      "epoch": 2.68,
-      "learning_rate": 2.324691247641103e-05,
-      "loss": 0.2552,
+      "epoch": 2.7496666306267343,
+      "grad_norm": 0.1950366199016571,
+      "learning_rate": 2.2193431633515194e-05,
+      "loss": 0.3997,
       "step": 76295
     },
     {
-      "epoch": 2.68,
-      "learning_rate": 2.3244070779133442e-05,
-      "loss": 0.2588,
+      "epoch": 2.749846830287959,
+      "grad_norm": 0.19967371225357056,
+      "learning_rate": 2.2190531965800138e-05,
+      "loss": 0.4061,
       "step": 76300
     },
     {
-      "epoch": 2.68,
-      "learning_rate": 2.324122910465545e-05,
-      "loss": 0.2745,
+      "epoch": 2.750027029949184,
+      "grad_norm": 0.22163501381874084,
+      "learning_rate": 2.2187632336363555e-05,
+      "loss": 0.3802,
       "step": 76305
     },
     {
-      "epoch": 2.68,
-      "learning_rate": 2.3238387453013934e-05,
-      "loss": 0.2781,
+      "epoch": 2.7502072296104085,
+      "grad_norm": 0.21866455674171448,
+      "learning_rate": 2.218473274524493e-05,
+      "loss": 0.4057,
       "step": 76310
     },
     {
-      "epoch": 2.68,
-      "learning_rate": 2.3235545824245797e-05,
-      "loss": 0.2418,
+      "epoch": 2.750387429271633,
+      "grad_norm": 0.20910955965518951,
+      "learning_rate": 2.2181833192483774e-05,
+      "loss": 0.4417,
       "step": 76315
     },
     {
-      "epoch": 2.69,
-      "learning_rate": 2.3232704218387923e-05,
-      "loss": 0.2643,
+      "epoch": 2.7505676289328576,
+      "grad_norm": 0.21740221977233887,
+      "learning_rate": 2.2178933678119598e-05,
+      "loss": 0.3984,
       "step": 76320
     },
     {
-      "epoch": 2.69,
-      "learning_rate": 2.3229862635477228e-05,
-      "loss": 0.2813,
+      "epoch": 2.7507478285940823,
+      "grad_norm": 0.28222358226776123,
+      "learning_rate": 2.21760342021919e-05,
+      "loss": 0.3966,
       "step": 76325
     },
     {
-      "epoch": 2.69,
-      "learning_rate": 2.3227021075550606e-05,
-      "loss": 0.2506,
+      "epoch": 2.750928028255307,
+      "grad_norm": 0.16867958009243011,
+      "learning_rate": 2.2173134764740196e-05,
+      "loss": 0.4051,
       "step": 76330
     },
     {
-      "epoch": 2.69,
-      "learning_rate": 2.3224179538644944e-05,
-      "loss": 0.2613,
+      "epoch": 2.7511082279165313,
+      "grad_norm": 0.26838427782058716,
+      "learning_rate": 2.217023536580398e-05,
+      "loss": 0.4368,
       "step": 76335
     },
     {
-      "epoch": 2.69,
-      "learning_rate": 2.322133802479713e-05,
-      "loss": 0.2741,
+      "epoch": 2.751288427577756,
+      "grad_norm": 0.21444910764694214,
+      "learning_rate": 2.2167336005422758e-05,
+      "loss": 0.3754,
       "step": 76340
     },
     {
-      "epoch": 2.69,
-      "learning_rate": 2.3218496534044082e-05,
-      "loss": 0.2694,
+      "epoch": 2.751468627238981,
+      "grad_norm": 0.23337894678115845,
+      "learning_rate": 2.2164436683636035e-05,
+      "loss": 0.4093,
       "step": 76345
     },
     {
-      "epoch": 2.69,
-      "learning_rate": 2.321565506642268e-05,
-      "loss": 0.2591,
+      "epoch": 2.7516488269002055,
+      "grad_norm": 0.2542593479156494,
+      "learning_rate": 2.216153740048332e-05,
+      "loss": 0.4148,
       "step": 76350
     },
     {
-      "epoch": 2.69,
-      "learning_rate": 2.321281362196982e-05,
-      "loss": 0.2826,
+      "epoch": 2.7518290265614302,
+      "grad_norm": 0.24587145447731018,
+      "learning_rate": 2.2158638156004098e-05,
+      "loss": 0.3793,
       "step": 76355
     },
     {
-      "epoch": 2.69,
-      "learning_rate": 2.320997220072239e-05,
-      "loss": 0.2798,
+      "epoch": 2.7520092262226545,
+      "grad_norm": 0.19757163524627686,
+      "learning_rate": 2.2155738950237887e-05,
+      "loss": 0.4132,
       "step": 76360
     },
     {
-      "epoch": 2.69,
-      "learning_rate": 2.3207130802717304e-05,
-      "loss": 0.278,
+      "epoch": 2.7521894258838793,
+      "grad_norm": 0.21361534297466278,
+      "learning_rate": 2.215283978322418e-05,
+      "loss": 0.3932,
       "step": 76365
     },
     {
-      "epoch": 2.69,
-      "learning_rate": 2.3204289427991437e-05,
-      "loss": 0.2806,
+      "epoch": 2.752369625545104,
+      "grad_norm": 0.17598572373390198,
+      "learning_rate": 2.214994065500248e-05,
+      "loss": 0.3777,
       "step": 76370
     },
     {
-      "epoch": 2.69,
-      "learning_rate": 2.3201448076581688e-05,
-      "loss": 0.2719,
+      "epoch": 2.7525498252063287,
+      "grad_norm": 0.17586344480514526,
+      "learning_rate": 2.2147041565612294e-05,
+      "loss": 0.39,
       "step": 76375
     },
     {
-      "epoch": 2.69,
-      "learning_rate": 2.319860674852494e-05,
-      "loss": 0.2532,
+      "epoch": 2.752730024867553,
+      "grad_norm": 0.17030631005764008,
+      "learning_rate": 2.2144142515093097e-05,
+      "loss": 0.3821,
       "step": 76380
     },
     {
-      "epoch": 2.69,
-      "learning_rate": 2.3195765443858116e-05,
-      "loss": 0.2692,
+      "epoch": 2.7529102245287778,
+      "grad_norm": 0.22826333343982697,
+      "learning_rate": 2.2141243503484426e-05,
+      "loss": 0.4265,
       "step": 76385
     },
     {
-      "epoch": 2.69,
-      "learning_rate": 2.319292416261808e-05,
-      "loss": 0.2773,
+      "epoch": 2.7530904241900025,
+      "grad_norm": 0.2388213574886322,
+      "learning_rate": 2.213834453082575e-05,
+      "loss": 0.3792,
       "step": 76390
     },
     {
-      "epoch": 2.69,
-      "learning_rate": 2.3190082904841737e-05,
-      "loss": 0.272,
+      "epoch": 2.7532706238512272,
+      "grad_norm": 0.19387775659561157,
+      "learning_rate": 2.213544559715657e-05,
+      "loss": 0.4057,
       "step": 76395
     },
     {
-      "epoch": 2.69,
-      "learning_rate": 2.318724167056596e-05,
-      "loss": 0.2646,
+      "epoch": 2.753450823512452,
+      "grad_norm": 0.22617575526237488,
+      "learning_rate": 2.2132546702516395e-05,
+      "loss": 0.4345,
       "step": 76400
     },
     {
-      "epoch": 2.69,
-      "learning_rate": 2.3184400459827666e-05,
-      "loss": 0.2476,
+      "epoch": 2.7536310231736767,
+      "grad_norm": 0.19259920716285706,
+      "learning_rate": 2.2129647846944708e-05,
+      "loss": 0.367,
       "step": 76405
     },
     {
-      "epoch": 2.69,
-      "learning_rate": 2.3181559272663737e-05,
-      "loss": 0.2573,
+      "epoch": 2.753811222834901,
+      "grad_norm": 0.20966337621212006,
+      "learning_rate": 2.2126749030481026e-05,
+      "loss": 0.3878,
       "step": 76410
     },
     {
-      "epoch": 2.69,
-      "learning_rate": 2.3178718109111052e-05,
-      "loss": 0.248,
+      "epoch": 2.7539914224961257,
+      "grad_norm": 0.22299322485923767,
+      "learning_rate": 2.2123850253164826e-05,
+      "loss": 0.4122,
       "step": 76415
     },
     {
-      "epoch": 2.69,
-      "learning_rate": 2.3175876969206524e-05,
-      "loss": 0.2775,
+      "epoch": 2.7541716221573505,
+      "grad_norm": 0.1778857707977295,
+      "learning_rate": 2.2120951515035605e-05,
+      "loss": 0.4066,
       "step": 76420
     },
     {
-      "epoch": 2.69,
-      "learning_rate": 2.3173035852987036e-05,
-      "loss": 0.2733,
+      "epoch": 2.7543518218185747,
+      "grad_norm": 0.22316712141036987,
+      "learning_rate": 2.2118052816132873e-05,
+      "loss": 0.4025,
       "step": 76425
     },
     {
-      "epoch": 2.69,
-      "learning_rate": 2.317019476048946e-05,
-      "loss": 0.2875,
+      "epoch": 2.7545320214797995,
+      "grad_norm": 0.19544607400894165,
+      "learning_rate": 2.2115154156496105e-05,
+      "loss": 0.4107,
       "step": 76430
     },
     {
-      "epoch": 2.69,
-      "learning_rate": 2.3167353691750703e-05,
-      "loss": 0.2397,
+      "epoch": 2.754712221141024,
+      "grad_norm": 0.25387707352638245,
+      "learning_rate": 2.2112255536164807e-05,
+      "loss": 0.3657,
       "step": 76435
     },
     {
-      "epoch": 2.69,
-      "learning_rate": 2.316451264680765e-05,
-      "loss": 0.2967,
+      "epoch": 2.754892420802249,
+      "grad_norm": 0.21383005380630493,
+      "learning_rate": 2.2109356955178463e-05,
+      "loss": 0.3826,
       "step": 76440
     },
     {
-      "epoch": 2.69,
-      "learning_rate": 2.3161671625697197e-05,
-      "loss": 0.2505,
+      "epoch": 2.7550726204634737,
+      "grad_norm": 0.19232189655303955,
+      "learning_rate": 2.2106458413576573e-05,
+      "loss": 0.403,
       "step": 76445
     },
     {
-      "epoch": 2.69,
-      "learning_rate": 2.3158830628456225e-05,
-      "loss": 0.2745,
+      "epoch": 2.7552528201246984,
+      "grad_norm": 0.23429124057292938,
+      "learning_rate": 2.210355991139863e-05,
+      "loss": 0.3711,
       "step": 76450
     },
     {
-      "epoch": 2.69,
-      "learning_rate": 2.3155989655121614e-05,
-      "loss": 0.2636,
+      "epoch": 2.7554330197859227,
+      "grad_norm": 0.2584487795829773,
+      "learning_rate": 2.2100661448684123e-05,
+      "loss": 0.389,
       "step": 76455
     },
     {
-      "epoch": 2.69,
-      "learning_rate": 2.315314870573027e-05,
-      "loss": 0.2789,
+      "epoch": 2.7556132194471474,
+      "grad_norm": 0.2131696194410324,
+      "learning_rate": 2.2097763025472536e-05,
+      "loss": 0.3589,
       "step": 76460
     },
     {
-      "epoch": 2.69,
-      "learning_rate": 2.315030778031907e-05,
-      "loss": 0.2802,
+      "epoch": 2.755793419108372,
+      "grad_norm": 0.21315625309944153,
+      "learning_rate": 2.2094864641803372e-05,
+      "loss": 0.3698,
       "step": 76465
     },
     {
-      "epoch": 2.69,
-      "learning_rate": 2.3147466878924915e-05,
-      "loss": 0.2532,
+      "epoch": 2.7559736187695965,
+      "grad_norm": 0.2009851336479187,
+      "learning_rate": 2.209196629771612e-05,
+      "loss": 0.3787,
       "step": 76470
     },
     {
-      "epoch": 2.69,
-      "learning_rate": 2.3144626001584663e-05,
-      "loss": 0.2909,
+      "epoch": 2.756153818430821,
+      "grad_norm": 0.24308665096759796,
+      "learning_rate": 2.208906799325025e-05,
+      "loss": 0.3962,
       "step": 76475
     },
     {
-      "epoch": 2.69,
-      "learning_rate": 2.3141785148335234e-05,
-      "loss": 0.256,
+      "epoch": 2.756334018092046,
+      "grad_norm": 0.2490856796503067,
+      "learning_rate": 2.2086169728445276e-05,
+      "loss": 0.4196,
       "step": 76480
     },
     {
-      "epoch": 2.69,
-      "learning_rate": 2.3138944319213497e-05,
-      "loss": 0.2513,
+      "epoch": 2.7565142177532707,
+      "grad_norm": 0.20798978209495544,
+      "learning_rate": 2.2083271503340662e-05,
+      "loss": 0.3762,
       "step": 76485
     },
     {
-      "epoch": 2.69,
-      "learning_rate": 2.313610351425634e-05,
-      "loss": 0.2494,
+      "epoch": 2.7566944174144954,
+      "grad_norm": 0.21434953808784485,
+      "learning_rate": 2.2080373317975927e-05,
+      "loss": 0.4108,
       "step": 76490
     },
     {
-      "epoch": 2.69,
-      "learning_rate": 2.3133262733500646e-05,
-      "loss": 0.2747,
+      "epoch": 2.75687461707572,
+      "grad_norm": 0.19935797154903412,
+      "learning_rate": 2.2077475172390536e-05,
+      "loss": 0.4004,
       "step": 76495
     },
     {
-      "epoch": 2.69,
-      "learning_rate": 2.3130421976983314e-05,
-      "loss": 0.2887,
+      "epoch": 2.7570548167369444,
+      "grad_norm": 0.19653324782848358,
+      "learning_rate": 2.2074577066623974e-05,
+      "loss": 0.3803,
       "step": 76500
     },
     {
-      "epoch": 2.69,
-      "eval_loss": 0.2611083686351776,
-      "eval_runtime": 10.5437,
-      "eval_samples_per_second": 9.484,
-      "eval_steps_per_second": 9.484,
+      "epoch": 2.7570548167369444,
+      "eval_loss": 0.4315706491470337,
+      "eval_runtime": 3.524,
+      "eval_samples_per_second": 28.377,
+      "eval_steps_per_second": 7.094,
       "step": 76500
     },
     {
-      "epoch": 2.69,
-      "learning_rate": 2.312758124474122e-05,
-      "loss": 0.2711,
+      "epoch": 2.757235016398169,
+      "grad_norm": 0.2254483997821808,
+      "learning_rate": 2.2071679000715733e-05,
+      "loss": 0.4047,
       "step": 76505
     },
     {
-      "epoch": 2.69,
-      "learning_rate": 2.3124740536811245e-05,
-      "loss": 0.2671,
+      "epoch": 2.757415216059394,
+      "grad_norm": 0.20184098184108734,
+      "learning_rate": 2.2068780974705298e-05,
+      "loss": 0.386,
       "step": 76510
     },
     {
-      "epoch": 2.69,
-      "learning_rate": 2.3121899853230278e-05,
-      "loss": 0.2561,
+      "epoch": 2.757595415720618,
+      "grad_norm": 0.20205454528331757,
+      "learning_rate": 2.206588298863216e-05,
+      "loss": 0.3598,
       "step": 76515
     },
     {
-      "epoch": 2.69,
-      "learning_rate": 2.3119059194035202e-05,
-      "loss": 0.2709,
+      "epoch": 2.757775615381843,
+      "grad_norm": 0.18086735904216766,
+      "learning_rate": 2.2062985042535797e-05,
+      "loss": 0.3592,
       "step": 76520
     },
     {
-      "epoch": 2.69,
-      "learning_rate": 2.3116218559262912e-05,
-      "loss": 0.2708,
+      "epoch": 2.7579558150430676,
+      "grad_norm": 0.18240606784820557,
+      "learning_rate": 2.2060087136455687e-05,
+      "loss": 0.3759,
       "step": 76525
     },
     {
-      "epoch": 2.69,
-      "learning_rate": 2.3113377948950267e-05,
-      "loss": 0.2657,
+      "epoch": 2.7581360147042924,
+      "grad_norm": 0.19021601974964142,
+      "learning_rate": 2.205718927043133e-05,
+      "loss": 0.4057,
       "step": 76530
     },
     {
-      "epoch": 2.69,
-      "learning_rate": 2.311053736313418e-05,
-      "loss": 0.2494,
+      "epoch": 2.758316214365517,
+      "grad_norm": 0.19175371527671814,
+      "learning_rate": 2.2054291444502198e-05,
+      "loss": 0.3937,
       "step": 76535
     },
     {
-      "epoch": 2.69,
-      "learning_rate": 2.3107696801851513e-05,
-      "loss": 0.264,
+      "epoch": 2.758496414026742,
+      "grad_norm": 0.2313561886548996,
+      "learning_rate": 2.2051393658707766e-05,
+      "loss": 0.3928,
       "step": 76540
     },
     {
-      "epoch": 2.69,
-      "learning_rate": 2.3104856265139164e-05,
-      "loss": 0.2555,
+      "epoch": 2.758676613687966,
+      "grad_norm": 0.19368553161621094,
+      "learning_rate": 2.2048495913087535e-05,
+      "loss": 0.3741,
       "step": 76545
     },
     {
-      "epoch": 2.69,
-      "learning_rate": 2.3102015753033996e-05,
-      "loss": 0.2725,
+      "epoch": 2.758856813349191,
+      "grad_norm": 0.20152942836284637,
+      "learning_rate": 2.2045598207680968e-05,
+      "loss": 0.3813,
       "step": 76550
     },
     {
-      "epoch": 2.69,
-      "learning_rate": 2.3099175265572912e-05,
-      "loss": 0.2636,
+      "epoch": 2.7590370130104156,
+      "grad_norm": 0.2724568843841553,
+      "learning_rate": 2.204270054252756e-05,
+      "loss": 0.386,
       "step": 76555
     },
     {
-      "epoch": 2.69,
-      "learning_rate": 2.3096334802792787e-05,
-      "loss": 0.2668,
+      "epoch": 2.7592172126716403,
+      "grad_norm": 0.2629745602607727,
+      "learning_rate": 2.203980291766679e-05,
+      "loss": 0.4035,
       "step": 76560
     },
     {
-      "epoch": 2.69,
-      "learning_rate": 2.309349436473049e-05,
-      "loss": 0.2711,
+      "epoch": 2.7593974123328646,
+      "grad_norm": 0.19395357370376587,
+      "learning_rate": 2.2036905333138115e-05,
+      "loss": 0.3902,
       "step": 76565
     },
     {
-      "epoch": 2.69,
-      "learning_rate": 2.3090653951422914e-05,
-      "loss": 0.2713,
+      "epoch": 2.7595776119940894,
+      "grad_norm": 0.19974055886268616,
+      "learning_rate": 2.203400778898104e-05,
+      "loss": 0.3891,
       "step": 76570
     },
     {
-      "epoch": 2.69,
-      "learning_rate": 2.308781356290694e-05,
-      "loss": 0.2534,
+      "epoch": 2.759757811655314,
+      "grad_norm": 0.21690985560417175,
+      "learning_rate": 2.203111028523504e-05,
+      "loss": 0.3544,
       "step": 76575
     },
     {
-      "epoch": 2.69,
-      "learning_rate": 2.3084973199219452e-05,
-      "loss": 0.2498,
+      "epoch": 2.759938011316539,
+      "grad_norm": 0.16456058621406555,
+      "learning_rate": 2.2028212821939576e-05,
+      "loss": 0.378,
       "step": 76580
     },
     {
-      "epoch": 2.69,
-      "learning_rate": 2.308213286039732e-05,
-      "loss": 0.271,
+      "epoch": 2.7601182109777636,
+      "grad_norm": 0.17220036685466766,
+      "learning_rate": 2.202531539913415e-05,
+      "loss": 0.3982,
       "step": 76585
     },
     {
-      "epoch": 2.69,
-      "learning_rate": 2.3079292546477423e-05,
-      "loss": 0.2704,
+      "epoch": 2.760298410638988,
+      "grad_norm": 0.2109745293855667,
+      "learning_rate": 2.202241801685821e-05,
+      "loss": 0.3956,
       "step": 76590
     },
     {
-      "epoch": 2.69,
-      "learning_rate": 2.3076452257496658e-05,
-      "loss": 0.2688,
+      "epoch": 2.7604786103002126,
+      "grad_norm": 0.18770918250083923,
+      "learning_rate": 2.2019520675151263e-05,
+      "loss": 0.3672,
       "step": 76595
     },
     {
-      "epoch": 2.7,
-      "learning_rate": 2.3073611993491885e-05,
-      "loss": 0.2491,
+      "epoch": 2.7606588099614373,
+      "grad_norm": 0.22950011491775513,
+      "learning_rate": 2.2016623374052766e-05,
+      "loss": 0.3903,
       "step": 76600
     },
     {
-      "epoch": 2.7,
-      "learning_rate": 2.30707717545e-05,
-      "loss": 0.2701,
+      "epoch": 2.760839009622662,
+      "grad_norm": 0.19520814716815948,
+      "learning_rate": 2.2013726113602192e-05,
+      "loss": 0.3848,
       "step": 76605
     },
     {
-      "epoch": 2.7,
-      "learning_rate": 2.3067931540557855e-05,
-      "loss": 0.27,
+      "epoch": 2.7610192092838863,
+      "grad_norm": 0.20134703814983368,
+      "learning_rate": 2.201082889383903e-05,
+      "loss": 0.4014,
       "step": 76610
     },
     {
-      "epoch": 2.7,
-      "learning_rate": 2.306509135170236e-05,
-      "loss": 0.2353,
+      "epoch": 2.761199408945111,
+      "grad_norm": 0.19556838274002075,
+      "learning_rate": 2.200793171480274e-05,
+      "loss": 0.4127,
       "step": 76615
     },
     {
-      "epoch": 2.7,
-      "learning_rate": 2.306225118797038e-05,
-      "loss": 0.2765,
+      "epoch": 2.761379608606336,
+      "grad_norm": 0.20522183179855347,
+      "learning_rate": 2.20050345765328e-05,
+      "loss": 0.4185,
       "step": 76620
     },
     {
-      "epoch": 2.7,
-      "learning_rate": 2.3059411049398784e-05,
-      "loss": 0.261,
+      "epoch": 2.7615598082675605,
+      "grad_norm": 0.20291735231876373,
+      "learning_rate": 2.2002137479068684e-05,
+      "loss": 0.3848,
       "step": 76625
     },
     {
-      "epoch": 2.7,
-      "learning_rate": 2.305657093602445e-05,
-      "loss": 0.2716,
+      "epoch": 2.7617400079287853,
+      "grad_norm": 0.18468672037124634,
+      "learning_rate": 2.1999240422449863e-05,
+      "loss": 0.4037,
       "step": 76630
     },
     {
-      "epoch": 2.7,
-      "learning_rate": 2.3053730847884274e-05,
-      "loss": 0.2438,
+      "epoch": 2.76192020759001,
+      "grad_norm": 0.19746728241443634,
+      "learning_rate": 2.1996343406715815e-05,
+      "loss": 0.3784,
       "step": 76635
     },
     {
-      "epoch": 2.7,
-      "learning_rate": 2.3050890785015122e-05,
-      "loss": 0.2526,
+      "epoch": 2.7621004072512343,
+      "grad_norm": 0.17763979732990265,
+      "learning_rate": 2.1993446431906007e-05,
+      "loss": 0.4116,
       "step": 76640
     },
     {
-      "epoch": 2.7,
-      "learning_rate": 2.3048050747453857e-05,
-      "loss": 0.2388,
+      "epoch": 2.762280606912459,
+      "grad_norm": 0.2670569121837616,
+      "learning_rate": 2.1990549498059906e-05,
+      "loss": 0.3857,
       "step": 76645
     },
     {
-      "epoch": 2.7,
-      "learning_rate": 2.3045210735237373e-05,
-      "loss": 0.2811,
+      "epoch": 2.7624608065736838,
+      "grad_norm": 0.21598556637763977,
+      "learning_rate": 2.198765260521699e-05,
+      "loss": 0.3965,
       "step": 76650
     },
     {
-      "epoch": 2.7,
-      "learning_rate": 2.304237074840254e-05,
-      "loss": 0.2666,
+      "epoch": 2.762641006234908,
+      "grad_norm": 0.17488926649093628,
+      "learning_rate": 2.1984755753416728e-05,
+      "loss": 0.3939,
       "step": 76655
     },
     {
-      "epoch": 2.7,
-      "learning_rate": 2.3039530786986237e-05,
-      "loss": 0.2566,
+      "epoch": 2.762821205896133,
+      "grad_norm": 0.2201049029827118,
+      "learning_rate": 2.1981858942698568e-05,
+      "loss": 0.4167,
       "step": 76660
     },
     {
-      "epoch": 2.7,
-      "learning_rate": 2.3036690851025324e-05,
-      "loss": 0.2593,
+      "epoch": 2.7630014055573575,
+      "grad_norm": 0.21279020607471466,
+      "learning_rate": 2.1978962173102015e-05,
+      "loss": 0.3794,
       "step": 76665
     },
     {
-      "epoch": 2.7,
-      "learning_rate": 2.303385094055669e-05,
-      "loss": 0.2807,
+      "epoch": 2.7631816052185822,
+      "grad_norm": 0.1922799050807953,
+      "learning_rate": 2.1976065444666495e-05,
+      "loss": 0.4126,
       "step": 76670
     },
     {
-      "epoch": 2.7,
-      "learning_rate": 2.3031011055617217e-05,
-      "loss": 0.2642,
+      "epoch": 2.763361804879807,
+      "grad_norm": 0.19106921553611755,
+      "learning_rate": 2.197316875743152e-05,
+      "loss": 0.3927,
       "step": 76675
     },
     {
-      "epoch": 2.7,
-      "learning_rate": 2.3028171196243756e-05,
-      "loss": 0.2517,
+      "epoch": 2.7635420045410317,
+      "grad_norm": 0.25177741050720215,
+      "learning_rate": 2.1970272111436527e-05,
+      "loss": 0.3636,
       "step": 76680
     },
     {
-      "epoch": 2.7,
-      "learning_rate": 2.3025331362473187e-05,
-      "loss": 0.2673,
+      "epoch": 2.763722204202256,
+      "grad_norm": 0.21334508061408997,
+      "learning_rate": 2.196737550672098e-05,
+      "loss": 0.3626,
       "step": 76685
     },
     {
-      "epoch": 2.7,
-      "learning_rate": 2.30224915543424e-05,
-      "loss": 0.2477,
+      "epoch": 2.7639024038634807,
+      "grad_norm": 0.21705521643161774,
+      "learning_rate": 2.196447894332437e-05,
+      "loss": 0.377,
       "step": 76690
     },
     {
-      "epoch": 2.7,
-      "learning_rate": 2.301965177188826e-05,
-      "loss": 0.2783,
+      "epoch": 2.7640826035247055,
+      "grad_norm": 0.20139169692993164,
+      "learning_rate": 2.196158242128613e-05,
+      "loss": 0.4,
       "step": 76695
     },
     {
-      "epoch": 2.7,
-      "learning_rate": 2.3016812015147635e-05,
-      "loss": 0.2897,
+      "epoch": 2.7642628031859298,
+      "grad_norm": 0.17421554028987885,
+      "learning_rate": 2.195868594064576e-05,
+      "loss": 0.3998,
       "step": 76700
     },
     {
-      "epoch": 2.7,
-      "learning_rate": 2.3013972284157386e-05,
-      "loss": 0.2525,
+      "epoch": 2.7644430028471545,
+      "grad_norm": 0.1658887267112732,
+      "learning_rate": 2.1955789501442696e-05,
+      "loss": 0.3811,
       "step": 76705
     },
     {
-      "epoch": 2.7,
-      "learning_rate": 2.3011132578954407e-05,
-      "loss": 0.2556,
+      "epoch": 2.7646232025083792,
+      "grad_norm": 0.1670427918434143,
+      "learning_rate": 2.1952893103716408e-05,
+      "loss": 0.3959,
       "step": 76710
     },
     {
-      "epoch": 2.7,
-      "learning_rate": 2.3008292899575562e-05,
-      "loss": 0.2586,
+      "epoch": 2.764803402169604,
+      "grad_norm": 0.21571218967437744,
+      "learning_rate": 2.194999674750637e-05,
+      "loss": 0.4167,
       "step": 76715
     },
     {
-      "epoch": 2.7,
-      "learning_rate": 2.3005453246057725e-05,
-      "loss": 0.2679,
+      "epoch": 2.7649836018308287,
+      "grad_norm": 0.25938844680786133,
+      "learning_rate": 2.194710043285203e-05,
+      "loss": 0.3906,
       "step": 76720
     },
     {
-      "epoch": 2.7,
-      "learning_rate": 2.300261361843775e-05,
-      "loss": 0.2646,
+      "epoch": 2.7651638014920534,
+      "grad_norm": 0.1947462111711502,
+      "learning_rate": 2.1944204159792854e-05,
+      "loss": 0.3942,
       "step": 76725
     },
     {
-      "epoch": 2.7,
-      "learning_rate": 2.299977401675253e-05,
-      "loss": 0.259,
+      "epoch": 2.7653440011532777,
+      "grad_norm": 0.19629132747650146,
+      "learning_rate": 2.1941307928368305e-05,
+      "loss": 0.4012,
       "step": 76730
     },
     {
-      "epoch": 2.7,
-      "learning_rate": 2.2996934441038925e-05,
-      "loss": 0.2711,
+      "epoch": 2.7655242008145025,
+      "grad_norm": 0.20872166752815247,
+      "learning_rate": 2.1938411738617843e-05,
+      "loss": 0.4147,
       "step": 76735
     },
     {
-      "epoch": 2.7,
-      "learning_rate": 2.299409489133381e-05,
-      "loss": 0.2804,
+      "epoch": 2.765704400475727,
+      "grad_norm": 0.2229888141155243,
+      "learning_rate": 2.1935515590580934e-05,
+      "loss": 0.372,
       "step": 76740
     },
     {
-      "epoch": 2.7,
-      "learning_rate": 2.2991255367674043e-05,
-      "loss": 0.2622,
+      "epoch": 2.7658846001369515,
+      "grad_norm": 0.23434416949748993,
+      "learning_rate": 2.193261948429703e-05,
+      "loss": 0.404,
       "step": 76745
     },
     {
-      "epoch": 2.7,
-      "learning_rate": 2.298841587009651e-05,
-      "loss": 0.2763,
+      "epoch": 2.766064799798176,
+      "grad_norm": 0.2255534827709198,
+      "learning_rate": 2.1929723419805582e-05,
+      "loss": 0.4228,
       "step": 76750
     },
     {
-      "epoch": 2.7,
-      "learning_rate": 2.298557639863807e-05,
-      "loss": 0.2788,
+      "epoch": 2.766244999459401,
+      "grad_norm": 0.21123521029949188,
+      "learning_rate": 2.192682739714607e-05,
+      "loss": 0.371,
       "step": 76755
     },
     {
-      "epoch": 2.7,
-      "learning_rate": 2.298273695333558e-05,
-      "loss": 0.275,
+      "epoch": 2.7664251991206257,
+      "grad_norm": 0.18391209840774536,
+      "learning_rate": 2.1923931416357944e-05,
+      "loss": 0.3993,
       "step": 76760
     },
     {
-      "epoch": 2.7,
-      "learning_rate": 2.2979897534225928e-05,
-      "loss": 0.2933,
+      "epoch": 2.7666053987818504,
+      "grad_norm": 0.19269075989723206,
+      "learning_rate": 2.1921035477480636e-05,
+      "loss": 0.3941,
       "step": 76765
     },
     {
-      "epoch": 2.7,
-      "learning_rate": 2.297705814134598e-05,
-      "loss": 0.2498,
+      "epoch": 2.766785598443075,
+      "grad_norm": 0.18515953421592712,
+      "learning_rate": 2.1918139580553644e-05,
+      "loss": 0.4063,
       "step": 76770
     },
     {
-      "epoch": 2.7,
-      "learning_rate": 2.2974218774732604e-05,
-      "loss": 0.2771,
+      "epoch": 2.7669657981042994,
+      "grad_norm": 0.21497590839862823,
+      "learning_rate": 2.1915243725616386e-05,
+      "loss": 0.4265,
       "step": 76775
     },
     {
-      "epoch": 2.7,
-      "learning_rate": 2.2971379434422647e-05,
-      "loss": 0.2682,
+      "epoch": 2.767145997765524,
+      "grad_norm": 0.19425266981124878,
+      "learning_rate": 2.1912347912708354e-05,
+      "loss": 0.405,
       "step": 76780
     },
     {
-      "epoch": 2.7,
-      "learning_rate": 2.2968540120453003e-05,
-      "loss": 0.2858,
+      "epoch": 2.767326197426749,
+      "grad_norm": 0.24012631177902222,
+      "learning_rate": 2.1909452141868975e-05,
+      "loss": 0.4209,
       "step": 76785
     },
     {
-      "epoch": 2.7,
-      "learning_rate": 2.296570083286052e-05,
-      "loss": 0.2574,
+      "epoch": 2.7675063970879736,
+      "grad_norm": 0.2343771904706955,
+      "learning_rate": 2.190655641313771e-05,
+      "loss": 0.3743,
       "step": 76790
     },
     {
-      "epoch": 2.7,
-      "learning_rate": 2.2962861571682075e-05,
-      "loss": 0.2707,
+      "epoch": 2.767686596749198,
+      "grad_norm": 0.23053786158561707,
+      "learning_rate": 2.1903660726554016e-05,
+      "loss": 0.4217,
       "step": 76795
     },
     {
-      "epoch": 2.7,
-      "learning_rate": 2.296002233695452e-05,
-      "loss": 0.2781,
+      "epoch": 2.7678667964104227,
+      "grad_norm": 0.19305402040481567,
+      "learning_rate": 2.1900765082157347e-05,
+      "loss": 0.3926,
       "step": 76800
     },
     {
-      "epoch": 2.7,
-      "learning_rate": 2.295718312871475e-05,
-      "loss": 0.2666,
+      "epoch": 2.7680469960716474,
+      "grad_norm": 0.18887373805046082,
+      "learning_rate": 2.1897869479987148e-05,
+      "loss": 0.3979,
       "step": 76805
     },
     {
-      "epoch": 2.7,
-      "learning_rate": 2.2954343946999602e-05,
-      "loss": 0.2449,
+      "epoch": 2.768227195732872,
+      "grad_norm": 0.2448185682296753,
+      "learning_rate": 2.1894973920082884e-05,
+      "loss": 0.4011,
       "step": 76810
     },
     {
-      "epoch": 2.7,
-      "learning_rate": 2.2951504791845945e-05,
-      "loss": 0.2578,
+      "epoch": 2.768407395394097,
+      "grad_norm": 0.2089599370956421,
+      "learning_rate": 2.189207840248399e-05,
+      "loss": 0.383,
       "step": 76815
     },
     {
-      "epoch": 2.7,
-      "learning_rate": 2.2948665663290644e-05,
-      "loss": 0.2534,
+      "epoch": 2.768587595055321,
+      "grad_norm": 0.24234536290168762,
+      "learning_rate": 2.1889182927229936e-05,
+      "loss": 0.3862,
       "step": 76820
     },
     {
-      "epoch": 2.7,
-      "learning_rate": 2.294582656137058e-05,
-      "loss": 0.2635,
+      "epoch": 2.768767794716546,
+      "grad_norm": 0.28005969524383545,
+      "learning_rate": 2.188628749436016e-05,
+      "loss": 0.405,
       "step": 76825
     },
     {
-      "epoch": 2.7,
-      "learning_rate": 2.29429874861226e-05,
-      "loss": 0.2841,
+      "epoch": 2.7689479943777706,
+      "grad_norm": 0.20914676785469055,
+      "learning_rate": 2.188339210391411e-05,
+      "loss": 0.3951,
       "step": 76830
     },
     {
-      "epoch": 2.7,
-      "learning_rate": 2.2940148437583575e-05,
-      "loss": 0.2808,
+      "epoch": 2.7691281940389953,
+      "grad_norm": 0.20570850372314453,
+      "learning_rate": 2.1880496755931244e-05,
+      "loss": 0.4507,
       "step": 76835
     },
     {
-      "epoch": 2.7,
-      "learning_rate": 2.2937309415790353e-05,
-      "loss": 0.2633,
+      "epoch": 2.7693083937002196,
+      "grad_norm": 0.20999915897846222,
+      "learning_rate": 2.187760145045101e-05,
+      "loss": 0.4017,
       "step": 76840
     },
     {
-      "epoch": 2.7,
-      "learning_rate": 2.2934470420779815e-05,
-      "loss": 0.2678,
+      "epoch": 2.7694885933614444,
+      "grad_norm": 0.16798564791679382,
+      "learning_rate": 2.1874706187512836e-05,
+      "loss": 0.3827,
       "step": 76845
     },
     {
-      "epoch": 2.7,
-      "learning_rate": 2.2931631452588827e-05,
-      "loss": 0.2716,
+      "epoch": 2.769668793022669,
+      "grad_norm": 0.20106874406337738,
+      "learning_rate": 2.18718109671562e-05,
+      "loss": 0.4012,
       "step": 76850
     },
     {
-      "epoch": 2.7,
-      "learning_rate": 2.2928792511254235e-05,
-      "loss": 0.2717,
+      "epoch": 2.769848992683894,
+      "grad_norm": 0.20237289369106293,
+      "learning_rate": 2.1868915789420522e-05,
+      "loss": 0.3978,
       "step": 76855
     },
     {
-      "epoch": 2.7,
-      "learning_rate": 2.2925953596812898e-05,
-      "loss": 0.2703,
+      "epoch": 2.7700291923451186,
+      "grad_norm": 0.2165011614561081,
+      "learning_rate": 2.186602065434527e-05,
+      "loss": 0.4484,
       "step": 76860
     },
     {
-      "epoch": 2.7,
-      "learning_rate": 2.29231147093017e-05,
-      "loss": 0.2869,
+      "epoch": 2.770209392006343,
+      "grad_norm": 0.19253584742546082,
+      "learning_rate": 2.1863125561969883e-05,
+      "loss": 0.3802,
       "step": 76865
     },
     {
-      "epoch": 2.7,
-      "learning_rate": 2.292027584875748e-05,
-      "loss": 0.2913,
+      "epoch": 2.7703895916675676,
+      "grad_norm": 0.19273778796195984,
+      "learning_rate": 2.1860230512333784e-05,
+      "loss": 0.3859,
       "step": 76870
     },
     {
-      "epoch": 2.7,
-      "learning_rate": 2.291743701521711e-05,
-      "loss": 0.2903,
+      "epoch": 2.7705697913287923,
+      "grad_norm": 0.1797461360692978,
+      "learning_rate": 2.185733550547646e-05,
+      "loss": 0.3808,
       "step": 76875
     },
     {
-      "epoch": 2.7,
-      "learning_rate": 2.2914598208717446e-05,
-      "loss": 0.2753,
+      "epoch": 2.770749990990017,
+      "grad_norm": 0.2192516326904297,
+      "learning_rate": 2.185444054143731e-05,
+      "loss": 0.4253,
       "step": 76880
     },
     {
-      "epoch": 2.71,
-      "learning_rate": 2.2911759429295357e-05,
-      "loss": 0.274,
+      "epoch": 2.7709301906512414,
+      "grad_norm": 0.17714661359786987,
+      "learning_rate": 2.1851545620255816e-05,
+      "loss": 0.4122,
       "step": 76885
     },
     {
-      "epoch": 2.71,
-      "learning_rate": 2.2908920676987693e-05,
-      "loss": 0.2704,
+      "epoch": 2.771110390312466,
+      "grad_norm": 0.26212644577026367,
+      "learning_rate": 2.1848650741971395e-05,
+      "loss": 0.4195,
       "step": 76890
     },
     {
-      "epoch": 2.71,
-      "learning_rate": 2.2906081951831308e-05,
-      "loss": 0.2785,
+      "epoch": 2.771290589973691,
+      "grad_norm": 0.1727093756198883,
+      "learning_rate": 2.1845755906623498e-05,
+      "loss": 0.4085,
       "step": 76895
     },
     {
-      "epoch": 2.71,
-      "learning_rate": 2.2903243253863075e-05,
-      "loss": 0.2702,
+      "epoch": 2.7714707896349156,
+      "grad_norm": 0.18318147957324982,
+      "learning_rate": 2.184286111425157e-05,
+      "loss": 0.349,
       "step": 76900
     },
     {
-      "epoch": 2.71,
-      "learning_rate": 2.290040458311985e-05,
-      "loss": 0.2588,
+      "epoch": 2.7716509892961403,
+      "grad_norm": 0.23283013701438904,
+      "learning_rate": 2.1839966364895042e-05,
+      "loss": 0.3714,
       "step": 76905
     },
     {
-      "epoch": 2.71,
-      "learning_rate": 2.2897565939638487e-05,
-      "loss": 0.2606,
+      "epoch": 2.771831188957365,
+      "grad_norm": 0.20274753868579865,
+      "learning_rate": 2.183707165859336e-05,
+      "loss": 0.3953,
       "step": 76910
     },
     {
-      "epoch": 2.71,
-      "learning_rate": 2.2894727323455835e-05,
-      "loss": 0.2743,
+      "epoch": 2.7720113886185893,
+      "grad_norm": 0.25053125619888306,
+      "learning_rate": 2.183417699538597e-05,
+      "loss": 0.4063,
       "step": 76915
     },
     {
-      "epoch": 2.71,
-      "learning_rate": 2.2891888734608775e-05,
-      "loss": 0.249,
+      "epoch": 2.772191588279814,
+      "grad_norm": 0.17475423216819763,
+      "learning_rate": 2.18312823753123e-05,
+      "loss": 0.386,
       "step": 76920
     },
     {
-      "epoch": 2.71,
-      "learning_rate": 2.288905017313414e-05,
-      "loss": 0.2775,
+      "epoch": 2.772371787941039,
+      "grad_norm": 0.19454878568649292,
+      "learning_rate": 2.18283877984118e-05,
+      "loss": 0.382,
       "step": 76925
     },
     {
-      "epoch": 2.71,
-      "learning_rate": 2.2886211639068806e-05,
-      "loss": 0.2406,
+      "epoch": 2.772551987602263,
+      "grad_norm": 0.2621554434299469,
+      "learning_rate": 2.1825493264723902e-05,
+      "loss": 0.4036,
       "step": 76930
     },
     {
-      "epoch": 2.71,
-      "learning_rate": 2.2883373132449608e-05,
-      "loss": 0.2914,
+      "epoch": 2.772732187263488,
+      "grad_norm": 0.211991548538208,
+      "learning_rate": 2.1822598774288034e-05,
+      "loss": 0.3894,
       "step": 76935
     },
     {
-      "epoch": 2.71,
-      "learning_rate": 2.2880534653313428e-05,
-      "loss": 0.2439,
+      "epoch": 2.7729123869247125,
+      "grad_norm": 0.20913714170455933,
+      "learning_rate": 2.181970432714365e-05,
+      "loss": 0.3712,
       "step": 76940
     },
     {
-      "epoch": 2.71,
-      "learning_rate": 2.2877696201697107e-05,
-      "loss": 0.2857,
+      "epoch": 2.7730925865859373,
+      "grad_norm": 0.22146856784820557,
+      "learning_rate": 2.1816809923330188e-05,
+      "loss": 0.3821,
       "step": 76945
     },
     {
-      "epoch": 2.71,
-      "learning_rate": 2.28748577776375e-05,
-      "loss": 0.2752,
+      "epoch": 2.773272786247162,
+      "grad_norm": 0.16859367489814758,
+      "learning_rate": 2.1813915562887054e-05,
+      "loss": 0.4161,
       "step": 76950
     },
     {
-      "epoch": 2.71,
-      "learning_rate": 2.287201938117146e-05,
-      "loss": 0.2735,
+      "epoch": 2.7734529859083867,
+      "grad_norm": 0.2146807461977005,
+      "learning_rate": 2.1811021245853724e-05,
+      "loss": 0.4039,
       "step": 76955
     },
     {
-      "epoch": 2.71,
-      "learning_rate": 2.2869181012335848e-05,
-      "loss": 0.2812,
+      "epoch": 2.773633185569611,
+      "grad_norm": 0.23717908561229706,
+      "learning_rate": 2.1808126972269594e-05,
+      "loss": 0.403,
       "step": 76960
     },
     {
-      "epoch": 2.71,
-      "learning_rate": 2.286634267116752e-05,
-      "loss": 0.254,
+      "epoch": 2.7738133852308358,
+      "grad_norm": 0.23532505333423615,
+      "learning_rate": 2.180523274217413e-05,
+      "loss": 0.3809,
       "step": 76965
     },
     {
-      "epoch": 2.71,
-      "learning_rate": 2.2863504357703328e-05,
-      "loss": 0.2942,
+      "epoch": 2.7739935848920605,
+      "grad_norm": 0.21510958671569824,
+      "learning_rate": 2.180233855560675e-05,
+      "loss": 0.3985,
       "step": 76970
     },
     {
-      "epoch": 2.71,
-      "learning_rate": 2.286066607198011e-05,
-      "loss": 0.2571,
+      "epoch": 2.774173784553285,
+      "grad_norm": 0.21763652563095093,
+      "learning_rate": 2.1799444412606873e-05,
+      "loss": 0.3586,
       "step": 76975
     },
     {
-      "epoch": 2.71,
-      "learning_rate": 2.2857827814034743e-05,
-      "loss": 0.2972,
+      "epoch": 2.7743539842145095,
+      "grad_norm": 0.23686784505844116,
+      "learning_rate": 2.1796550313213963e-05,
+      "loss": 0.4179,
       "step": 76980
     },
     {
-      "epoch": 2.71,
-      "learning_rate": 2.2854989583904074e-05,
-      "loss": 0.2408,
+      "epoch": 2.7745341838757343,
+      "grad_norm": 0.1995023488998413,
+      "learning_rate": 2.1793656257467432e-05,
+      "loss": 0.4106,
       "step": 76985
     },
     {
-      "epoch": 2.71,
-      "learning_rate": 2.2852151381624952e-05,
-      "loss": 0.261,
+      "epoch": 2.774714383536959,
+      "grad_norm": 0.20156660676002502,
+      "learning_rate": 2.179076224540671e-05,
+      "loss": 0.3956,
       "step": 76990
     },
     {
-      "epoch": 2.71,
-      "learning_rate": 2.2849313207234217e-05,
-      "loss": 0.2507,
+      "epoch": 2.7748945831981837,
+      "grad_norm": 0.19944174587726593,
+      "learning_rate": 2.1787868277071234e-05,
+      "loss": 0.3759,
       "step": 76995
     },
     {
-      "epoch": 2.71,
-      "learning_rate": 2.2846475060768743e-05,
-      "loss": 0.2842,
+      "epoch": 2.7750747828594085,
+      "grad_norm": 0.22275997698307037,
+      "learning_rate": 2.1784974352500423e-05,
+      "loss": 0.4041,
       "step": 77000
     },
     {
-      "epoch": 2.71,
-      "eval_loss": 0.2610674202442169,
-      "eval_runtime": 10.5451,
-      "eval_samples_per_second": 9.483,
-      "eval_steps_per_second": 9.483,
+      "epoch": 2.7750747828594085,
+      "eval_loss": 0.4315272569656372,
+      "eval_runtime": 3.5286,
+      "eval_samples_per_second": 28.34,
+      "eval_steps_per_second": 7.085,
       "step": 77000
     },
     {
-      "epoch": 2.71,
-      "learning_rate": 2.2843636942265367e-05,
-      "loss": 0.2581,
+      "epoch": 2.7752549825206327,
+      "grad_norm": 0.19483999907970428,
+      "learning_rate": 2.178208047173372e-05,
+      "loss": 0.411,
       "step": 77005
     },
     {
-      "epoch": 2.71,
-      "learning_rate": 2.284079885176094e-05,
-      "loss": 0.2934,
+      "epoch": 2.7754351821818575,
+      "grad_norm": 0.18567782640457153,
+      "learning_rate": 2.177918663481055e-05,
+      "loss": 0.3731,
       "step": 77010
     },
     {
-      "epoch": 2.71,
-      "learning_rate": 2.2837960789292322e-05,
-      "loss": 0.2504,
+      "epoch": 2.775615381843082,
+      "grad_norm": 0.22382299602031708,
+      "learning_rate": 2.1776292841770333e-05,
+      "loss": 0.4029,
       "step": 77015
     },
     {
-      "epoch": 2.71,
-      "learning_rate": 2.2835122754896367e-05,
-      "loss": 0.2727,
+      "epoch": 2.7757955815043065,
+      "grad_norm": 0.20525696873664856,
+      "learning_rate": 2.1773399092652507e-05,
+      "loss": 0.3781,
       "step": 77020
     },
     {
-      "epoch": 2.71,
-      "learning_rate": 2.283228474860991e-05,
-      "loss": 0.2551,
+      "epoch": 2.7759757811655312,
+      "grad_norm": 0.21928441524505615,
+      "learning_rate": 2.1770505387496494e-05,
+      "loss": 0.3852,
       "step": 77025
     },
     {
-      "epoch": 2.71,
-      "learning_rate": 2.2829446770469796e-05,
-      "loss": 0.2866,
+      "epoch": 2.776155980826756,
+      "grad_norm": 0.25740548968315125,
+      "learning_rate": 2.1767611726341714e-05,
+      "loss": 0.4268,
       "step": 77030
     },
     {
-      "epoch": 2.71,
-      "learning_rate": 2.28266088205129e-05,
-      "loss": 0.2693,
+      "epoch": 2.7763361804879807,
+      "grad_norm": 0.21519418060779572,
+      "learning_rate": 2.17647181092276e-05,
+      "loss": 0.4104,
       "step": 77035
     },
     {
-      "epoch": 2.71,
-      "learning_rate": 2.2823770898776047e-05,
-      "loss": 0.2849,
+      "epoch": 2.7765163801492054,
+      "grad_norm": 0.19074714183807373,
+      "learning_rate": 2.1761824536193575e-05,
+      "loss": 0.4052,
       "step": 77040
     },
     {
-      "epoch": 2.71,
-      "learning_rate": 2.28209330052961e-05,
-      "loss": 0.2797,
+      "epoch": 2.77669657981043,
+      "grad_norm": 0.20910242199897766,
+      "learning_rate": 2.1758931007279067e-05,
+      "loss": 0.3869,
       "step": 77045
     },
     {
-      "epoch": 2.71,
-      "learning_rate": 2.2818095140109892e-05,
-      "loss": 0.2796,
+      "epoch": 2.7768767794716545,
+      "grad_norm": 0.22091153264045715,
+      "learning_rate": 2.1756037522523503e-05,
+      "loss": 0.4058,
       "step": 77050
     },
     {
-      "epoch": 2.71,
-      "learning_rate": 2.2815257303254296e-05,
-      "loss": 0.2764,
+      "epoch": 2.777056979132879,
+      "grad_norm": 0.18852129578590393,
+      "learning_rate": 2.175314408196628e-05,
+      "loss": 0.4099,
       "step": 77055
     },
     {
-      "epoch": 2.71,
-      "learning_rate": 2.281241949476614e-05,
-      "loss": 0.2723,
+      "epoch": 2.777237178794104,
+      "grad_norm": 0.23839035630226135,
+      "learning_rate": 2.1750250685646863e-05,
+      "loss": 0.3927,
       "step": 77060
     },
     {
-      "epoch": 2.71,
-      "learning_rate": 2.2809581714682268e-05,
-      "loss": 0.2533,
+      "epoch": 2.7774173784553287,
+      "grad_norm": 0.20940519869327545,
+      "learning_rate": 2.174735733360464e-05,
+      "loss": 0.3632,
       "step": 77065
     },
     {
-      "epoch": 2.71,
-      "learning_rate": 2.280674396303953e-05,
-      "loss": 0.2869,
+      "epoch": 2.777597578116553,
+      "grad_norm": 0.24172858893871307,
+      "learning_rate": 2.174446402587904e-05,
+      "loss": 0.3759,
       "step": 77070
     },
     {
-      "epoch": 2.71,
-      "learning_rate": 2.280390623987479e-05,
-      "loss": 0.2614,
+      "epoch": 2.7777777777777777,
+      "grad_norm": 0.2583145201206207,
+      "learning_rate": 2.1741570762509495e-05,
+      "loss": 0.3658,
       "step": 77075
     },
     {
-      "epoch": 2.71,
-      "learning_rate": 2.2801068545224876e-05,
-      "loss": 0.2631,
+      "epoch": 2.7779579774390024,
+      "grad_norm": 0.21428237855434418,
+      "learning_rate": 2.173867754353541e-05,
+      "loss": 0.4038,
       "step": 77080
     },
     {
-      "epoch": 2.71,
-      "learning_rate": 2.2798230879126636e-05,
-      "loss": 0.2867,
+      "epoch": 2.778138177100227,
+      "grad_norm": 0.1586180180311203,
+      "learning_rate": 2.173578436899622e-05,
+      "loss": 0.3844,
       "step": 77085
     },
     {
-      "epoch": 2.71,
-      "learning_rate": 2.2795393241616913e-05,
-      "loss": 0.2628,
+      "epoch": 2.778318376761452,
+      "grad_norm": 0.22997400164604187,
+      "learning_rate": 2.1732891238931334e-05,
+      "loss": 0.3767,
       "step": 77090
     },
     {
-      "epoch": 2.71,
-      "learning_rate": 2.279255563273256e-05,
-      "loss": 0.2489,
+      "epoch": 2.778498576422676,
+      "grad_norm": 0.22127170860767365,
+      "learning_rate": 2.1729998153380165e-05,
+      "loss": 0.3797,
       "step": 77095
     },
     {
-      "epoch": 2.71,
-      "learning_rate": 2.2789718052510424e-05,
-      "loss": 0.2667,
+      "epoch": 2.778678776083901,
+      "grad_norm": 0.25849655270576477,
+      "learning_rate": 2.1727105112382147e-05,
+      "loss": 0.4181,
       "step": 77100
     },
     {
-      "epoch": 2.71,
-      "learning_rate": 2.2786880500987336e-05,
-      "loss": 0.2879,
+      "epoch": 2.7788589757451256,
+      "grad_norm": 0.18317921459674835,
+      "learning_rate": 2.172421211597668e-05,
+      "loss": 0.3667,
       "step": 77105
     },
     {
-      "epoch": 2.71,
-      "learning_rate": 2.2784042978200142e-05,
-      "loss": 0.2719,
+      "epoch": 2.7790391754063504,
+      "grad_norm": 0.21405282616615295,
+      "learning_rate": 2.1721319164203195e-05,
+      "loss": 0.4073,
       "step": 77110
     },
     {
-      "epoch": 2.71,
-      "learning_rate": 2.2781205484185695e-05,
-      "loss": 0.2512,
+      "epoch": 2.7792193750675747,
+      "grad_norm": 0.18047797679901123,
+      "learning_rate": 2.1718426257101103e-05,
+      "loss": 0.4006,
       "step": 77115
     },
     {
-      "epoch": 2.71,
-      "learning_rate": 2.2778368018980827e-05,
-      "loss": 0.2795,
+      "epoch": 2.7793995747287994,
+      "grad_norm": 0.1952604204416275,
+      "learning_rate": 2.1715533394709807e-05,
+      "loss": 0.3855,
       "step": 77120
     },
     {
-      "epoch": 2.71,
-      "learning_rate": 2.2775530582622385e-05,
-      "loss": 0.2752,
+      "epoch": 2.779579774390024,
+      "grad_norm": 0.18268941342830658,
+      "learning_rate": 2.1712640577068743e-05,
+      "loss": 0.435,
       "step": 77125
     },
     {
-      "epoch": 2.71,
-      "learning_rate": 2.277269317514722e-05,
-      "loss": 0.2545,
+      "epoch": 2.779759974051249,
+      "grad_norm": 0.23162731528282166,
+      "learning_rate": 2.1709747804217324e-05,
+      "loss": 0.3652,
       "step": 77130
     },
     {
-      "epoch": 2.71,
-      "learning_rate": 2.276985579659217e-05,
-      "loss": 0.2642,
+      "epoch": 2.7799401737124736,
+      "grad_norm": 0.22689977288246155,
+      "learning_rate": 2.170685507619493e-05,
+      "loss": 0.3839,
       "step": 77135
     },
     {
-      "epoch": 2.71,
-      "learning_rate": 2.276701844699407e-05,
-      "loss": 0.2255,
+      "epoch": 2.7801203733736983,
+      "grad_norm": 0.2621477544307709,
+      "learning_rate": 2.1703962393041015e-05,
+      "loss": 0.4038,
       "step": 77140
     },
     {
-      "epoch": 2.71,
-      "learning_rate": 2.2764181126389753e-05,
-      "loss": 0.2544,
+      "epoch": 2.7803005730349226,
+      "grad_norm": 0.22769121825695038,
+      "learning_rate": 2.1701069754794966e-05,
+      "loss": 0.4086,
       "step": 77145
     },
     {
-      "epoch": 2.71,
-      "learning_rate": 2.2761343834816072e-05,
-      "loss": 0.2894,
+      "epoch": 2.7804807726961474,
+      "grad_norm": 0.21255040168762207,
+      "learning_rate": 2.1698177161496205e-05,
+      "loss": 0.4162,
       "step": 77150
     },
     {
-      "epoch": 2.71,
-      "learning_rate": 2.2758506572309878e-05,
-      "loss": 0.2732,
+      "epoch": 2.780660972357372,
+      "grad_norm": 0.1877208799123764,
+      "learning_rate": 2.1695284613184154e-05,
+      "loss": 0.3944,
       "step": 77155
     },
     {
-      "epoch": 2.71,
-      "learning_rate": 2.2755669338907997e-05,
-      "loss": 0.2722,
+      "epoch": 2.7808411720185964,
+      "grad_norm": 0.24076056480407715,
+      "learning_rate": 2.1692392109898185e-05,
+      "loss": 0.4121,
       "step": 77160
     },
     {
-      "epoch": 2.71,
-      "learning_rate": 2.275283213464726e-05,
-      "loss": 0.2762,
+      "epoch": 2.781021371679821,
+      "grad_norm": 0.22234618663787842,
+      "learning_rate": 2.1689499651677754e-05,
+      "loss": 0.3533,
       "step": 77165
     },
     {
-      "epoch": 2.72,
-      "learning_rate": 2.2749994959564528e-05,
-      "loss": 0.2824,
+      "epoch": 2.781201571341046,
+      "grad_norm": 0.21065817773342133,
+      "learning_rate": 2.1686607238562245e-05,
+      "loss": 0.3904,
       "step": 77170
     },
     {
-      "epoch": 2.72,
-      "learning_rate": 2.274715781369662e-05,
-      "loss": 0.2778,
+      "epoch": 2.7813817710022706,
+      "grad_norm": 0.24153132736682892,
+      "learning_rate": 2.1683714870591067e-05,
+      "loss": 0.4196,
       "step": 77175
     },
     {
-      "epoch": 2.72,
-      "learning_rate": 2.2744320697080394e-05,
-      "loss": 0.2657,
+      "epoch": 2.7815619706634953,
+      "grad_norm": 0.25936198234558105,
+      "learning_rate": 2.1680822547803635e-05,
+      "loss": 0.4077,
       "step": 77180
     },
     {
-      "epoch": 2.72,
-      "learning_rate": 2.274148360975266e-05,
-      "loss": 0.2686,
+      "epoch": 2.78174217032472,
+      "grad_norm": 0.257414847612381,
+      "learning_rate": 2.1677930270239343e-05,
+      "loss": 0.4039,
       "step": 77185
     },
     {
-      "epoch": 2.72,
-      "learning_rate": 2.2738646551750288e-05,
-      "loss": 0.2729,
+      "epoch": 2.7819223699859443,
+      "grad_norm": 0.17230790853500366,
+      "learning_rate": 2.167503803793762e-05,
+      "loss": 0.3765,
       "step": 77190
     },
     {
-      "epoch": 2.72,
-      "learning_rate": 2.27358095231101e-05,
-      "loss": 0.2537,
+      "epoch": 2.782102569647169,
+      "grad_norm": 0.21409516036510468,
+      "learning_rate": 2.167214585093786e-05,
+      "loss": 0.3749,
       "step": 77195
     },
     {
-      "epoch": 2.72,
-      "learning_rate": 2.2732972523868926e-05,
-      "loss": 0.2587,
+      "epoch": 2.782282769308394,
+      "grad_norm": 0.17981880903244019,
+      "learning_rate": 2.1669253709279458e-05,
+      "loss": 0.4021,
       "step": 77200
     },
     {
-      "epoch": 2.72,
-      "learning_rate": 2.2730135554063607e-05,
-      "loss": 0.2661,
+      "epoch": 2.782462968969618,
+      "grad_norm": 0.1959773600101471,
+      "learning_rate": 2.166636161300184e-05,
+      "loss": 0.4146,
       "step": 77205
     },
     {
-      "epoch": 2.72,
-      "learning_rate": 2.272729861373099e-05,
-      "loss": 0.2629,
+      "epoch": 2.782643168630843,
+      "grad_norm": 0.30680885910987854,
+      "learning_rate": 2.1663469562144395e-05,
+      "loss": 0.3919,
       "step": 77210
     },
     {
-      "epoch": 2.72,
-      "learning_rate": 2.2724461702907905e-05,
-      "loss": 0.2707,
+      "epoch": 2.7828233682920676,
+      "grad_norm": 0.2044559270143509,
+      "learning_rate": 2.1660577556746527e-05,
+      "loss": 0.391,
       "step": 77215
     },
     {
-      "epoch": 2.72,
-      "learning_rate": 2.2721624821631183e-05,
-      "loss": 0.2674,
+      "epoch": 2.7830035679532923,
+      "grad_norm": 0.2195388227701187,
+      "learning_rate": 2.165768559684765e-05,
+      "loss": 0.395,
       "step": 77220
     },
     {
-      "epoch": 2.72,
-      "learning_rate": 2.271878796993765e-05,
-      "loss": 0.2789,
+      "epoch": 2.783183767614517,
+      "grad_norm": 0.2089926302433014,
+      "learning_rate": 2.1654793682487157e-05,
+      "loss": 0.3878,
       "step": 77225
     },
     {
-      "epoch": 2.72,
-      "learning_rate": 2.271595114786416e-05,
-      "loss": 0.2739,
+      "epoch": 2.7833639672757418,
+      "grad_norm": 0.21790191531181335,
+      "learning_rate": 2.1651901813704452e-05,
+      "loss": 0.4098,
       "step": 77230
     },
     {
-      "epoch": 2.72,
-      "learning_rate": 2.271311435544754e-05,
-      "loss": 0.2704,
+      "epoch": 2.783544166936966,
+      "grad_norm": 0.22714762389659882,
+      "learning_rate": 2.1649009990538947e-05,
+      "loss": 0.3506,
       "step": 77235
     },
     {
-      "epoch": 2.72,
-      "learning_rate": 2.2710277592724616e-05,
-      "loss": 0.2693,
+      "epoch": 2.783724366598191,
+      "grad_norm": 0.18509992957115173,
+      "learning_rate": 2.1646118213030015e-05,
+      "loss": 0.3868,
       "step": 77240
     },
     {
-      "epoch": 2.72,
-      "learning_rate": 2.270744085973224e-05,
-      "loss": 0.2677,
+      "epoch": 2.7839045662594155,
+      "grad_norm": 0.1982908993959427,
+      "learning_rate": 2.1643226481217084e-05,
+      "loss": 0.4092,
       "step": 77245
     },
     {
-      "epoch": 2.72,
-      "learning_rate": 2.270460415650723e-05,
-      "loss": 0.2619,
+      "epoch": 2.78408476592064,
+      "grad_norm": 0.1664113849401474,
+      "learning_rate": 2.1640334795139545e-05,
+      "loss": 0.3934,
       "step": 77250
     },
     {
-      "epoch": 2.72,
-      "learning_rate": 2.270176748308642e-05,
-      "loss": 0.267,
+      "epoch": 2.7842649655818645,
+      "grad_norm": 0.22491081058979034,
+      "learning_rate": 2.163744315483678e-05,
+      "loss": 0.4087,
       "step": 77255
     },
     {
-      "epoch": 2.72,
-      "learning_rate": 2.269893083950664e-05,
-      "loss": 0.2798,
+      "epoch": 2.7844451652430893,
+      "grad_norm": 0.22320492565631866,
+      "learning_rate": 2.1634551560348213e-05,
+      "loss": 0.4185,
       "step": 77260
     },
     {
-      "epoch": 2.72,
-      "learning_rate": 2.2696094225804735e-05,
-      "loss": 0.2499,
+      "epoch": 2.784625364904314,
+      "grad_norm": 0.1935398429632187,
+      "learning_rate": 2.1631660011713218e-05,
+      "loss": 0.3915,
       "step": 77265
     },
     {
-      "epoch": 2.72,
-      "learning_rate": 2.269325764201753e-05,
-      "loss": 0.2957,
+      "epoch": 2.7848055645655387,
+      "grad_norm": 0.205779567360878,
+      "learning_rate": 2.1628768508971213e-05,
+      "loss": 0.4085,
       "step": 77270
     },
     {
-      "epoch": 2.72,
-      "learning_rate": 2.2690421088181852e-05,
-      "loss": 0.2505,
+      "epoch": 2.7849857642267635,
+      "grad_norm": 0.19906838238239288,
+      "learning_rate": 2.1625877052161586e-05,
+      "loss": 0.4113,
       "step": 77275
     },
     {
-      "epoch": 2.72,
-      "learning_rate": 2.2687584564334524e-05,
-      "loss": 0.2756,
+      "epoch": 2.7851659638879878,
+      "grad_norm": 0.23320987820625305,
+      "learning_rate": 2.1622985641323724e-05,
+      "loss": 0.3884,
       "step": 77280
     },
     {
-      "epoch": 2.72,
-      "learning_rate": 2.268474807051239e-05,
-      "loss": 0.2733,
+      "epoch": 2.7853461635492125,
+      "grad_norm": 0.22347044944763184,
+      "learning_rate": 2.1620094276497033e-05,
+      "loss": 0.3888,
       "step": 77285
     },
     {
-      "epoch": 2.72,
-      "learning_rate": 2.2681911606752286e-05,
-      "loss": 0.2585,
+      "epoch": 2.7855263632104372,
+      "grad_norm": 0.27736568450927734,
+      "learning_rate": 2.16172029577209e-05,
+      "loss": 0.4216,
       "step": 77290
     },
     {
-      "epoch": 2.72,
-      "learning_rate": 2.267907517309103e-05,
-      "loss": 0.279,
+      "epoch": 2.785706562871662,
+      "grad_norm": 0.2748807668685913,
+      "learning_rate": 2.1614311685034726e-05,
+      "loss": 0.4152,
       "step": 77295
     },
     {
-      "epoch": 2.72,
-      "learning_rate": 2.267623876956544e-05,
-      "loss": 0.2715,
+      "epoch": 2.7858867625328863,
+      "grad_norm": 0.2487899661064148,
+      "learning_rate": 2.16114204584779e-05,
+      "loss": 0.4088,
       "step": 77300
     },
     {
-      "epoch": 2.72,
-      "learning_rate": 2.2673402396212374e-05,
-      "loss": 0.2629,
+      "epoch": 2.786066962194111,
+      "grad_norm": 0.21085642278194427,
+      "learning_rate": 2.1608529278089808e-05,
+      "loss": 0.3912,
       "step": 77305
     },
     {
-      "epoch": 2.72,
-      "learning_rate": 2.2670566053068637e-05,
-      "loss": 0.2589,
+      "epoch": 2.7862471618553357,
+      "grad_norm": 0.19442741572856903,
+      "learning_rate": 2.160563814390985e-05,
+      "loss": 0.3965,
       "step": 77310
     },
     {
-      "epoch": 2.72,
-      "learning_rate": 2.266772974017107e-05,
-      "loss": 0.2688,
+      "epoch": 2.7864273615165605,
+      "grad_norm": 0.19160184264183044,
+      "learning_rate": 2.1602747055977417e-05,
+      "loss": 0.37,
       "step": 77315
     },
     {
-      "epoch": 2.72,
-      "learning_rate": 2.2664893457556486e-05,
-      "loss": 0.2875,
+      "epoch": 2.786607561177785,
+      "grad_norm": 0.2595144212245941,
+      "learning_rate": 2.1599856014331895e-05,
+      "loss": 0.4168,
       "step": 77320
     },
     {
-      "epoch": 2.72,
-      "learning_rate": 2.266205720526173e-05,
-      "loss": 0.2793,
+      "epoch": 2.7867877608390095,
+      "grad_norm": 0.23235271871089935,
+      "learning_rate": 2.1596965019012682e-05,
+      "loss": 0.4176,
       "step": 77325
     },
     {
-      "epoch": 2.72,
-      "learning_rate": 2.265922098332362e-05,
-      "loss": 0.2463,
+      "epoch": 2.786967960500234,
+      "grad_norm": 0.1904749721288681,
+      "learning_rate": 2.1594074070059155e-05,
+      "loss": 0.3688,
       "step": 77330
     },
     {
-      "epoch": 2.72,
-      "learning_rate": 2.2656384791778977e-05,
-      "loss": 0.2618,
+      "epoch": 2.787148160161459,
+      "grad_norm": 0.22936037182807922,
+      "learning_rate": 2.1591183167510714e-05,
+      "loss": 0.3988,
       "step": 77335
     },
     {
-      "epoch": 2.72,
-      "learning_rate": 2.2653548630664632e-05,
-      "loss": 0.2857,
+      "epoch": 2.7873283598226837,
+      "grad_norm": 0.21783271431922913,
+      "learning_rate": 2.158829231140675e-05,
+      "loss": 0.4026,
       "step": 77340
     },
     {
-      "epoch": 2.72,
-      "learning_rate": 2.265071250001742e-05,
-      "loss": 0.2459,
+      "epoch": 2.787508559483908,
+      "grad_norm": 0.193325474858284,
+      "learning_rate": 2.1585401501786622e-05,
+      "loss": 0.4321,
       "step": 77345
     },
     {
-      "epoch": 2.72,
-      "learning_rate": 2.2647876399874156e-05,
-      "loss": 0.2834,
+      "epoch": 2.7876887591451327,
+      "grad_norm": 0.24753624200820923,
+      "learning_rate": 2.158251073868976e-05,
+      "loss": 0.4161,
       "step": 77350
     },
     {
-      "epoch": 2.72,
-      "learning_rate": 2.2645040330271658e-05,
-      "loss": 0.2687,
+      "epoch": 2.7878689588063574,
+      "grad_norm": 0.18600989878177643,
+      "learning_rate": 2.1579620022155516e-05,
+      "loss": 0.3775,
       "step": 77355
     },
     {
-      "epoch": 2.72,
-      "learning_rate": 2.2642204291246772e-05,
-      "loss": 0.271,
+      "epoch": 2.788049158467582,
+      "grad_norm": 0.19241438806056976,
+      "learning_rate": 2.1576729352223285e-05,
+      "loss": 0.395,
       "step": 77360
     },
     {
-      "epoch": 2.72,
-      "learning_rate": 2.26393682828363e-05,
-      "loss": 0.2495,
+      "epoch": 2.788229358128807,
+      "grad_norm": 0.2080182284116745,
+      "learning_rate": 2.1573838728932462e-05,
+      "loss": 0.3651,
       "step": 77365
     },
     {
-      "epoch": 2.72,
-      "learning_rate": 2.2636532305077083e-05,
-      "loss": 0.2624,
+      "epoch": 2.788409557790031,
+      "grad_norm": 0.23465201258659363,
+      "learning_rate": 2.157094815232241e-05,
+      "loss": 0.3976,
       "step": 77370
     },
     {
-      "epoch": 2.72,
-      "learning_rate": 2.2633696358005928e-05,
-      "loss": 0.2692,
+      "epoch": 2.788589757451256,
+      "grad_norm": 0.21152657270431519,
+      "learning_rate": 2.1568057622432536e-05,
+      "loss": 0.4053,
       "step": 77375
     },
     {
-      "epoch": 2.72,
-      "learning_rate": 2.263086044165967e-05,
-      "loss": 0.2721,
+      "epoch": 2.7887699571124807,
+      "grad_norm": 0.20607246458530426,
+      "learning_rate": 2.1565167139302213e-05,
+      "loss": 0.3984,
       "step": 77380
     },
     {
-      "epoch": 2.72,
-      "learning_rate": 2.2628024556075134e-05,
-      "loss": 0.2617,
+      "epoch": 2.7889501567737054,
+      "grad_norm": 0.20786413550376892,
+      "learning_rate": 2.1562276702970816e-05,
+      "loss": 0.3772,
       "step": 77385
     },
     {
-      "epoch": 2.72,
-      "learning_rate": 2.262518870128913e-05,
-      "loss": 0.246,
+      "epoch": 2.7891303564349297,
+      "grad_norm": 0.21368098258972168,
+      "learning_rate": 2.155938631347774e-05,
+      "loss": 0.3946,
       "step": 77390
     },
     {
-      "epoch": 2.72,
-      "learning_rate": 2.2622352877338482e-05,
-      "loss": 0.274,
+      "epoch": 2.7893105560961544,
+      "grad_norm": 0.22205038368701935,
+      "learning_rate": 2.1556495970862356e-05,
+      "loss": 0.3638,
       "step": 77395
     },
     {
-      "epoch": 2.72,
-      "learning_rate": 2.2619517084260018e-05,
-      "loss": 0.2585,
+      "epoch": 2.789490755757379,
+      "grad_norm": 0.22499975562095642,
+      "learning_rate": 2.1553605675164047e-05,
+      "loss": 0.4032,
       "step": 77400
     },
     {
-      "epoch": 2.72,
-      "learning_rate": 2.261668132209056e-05,
-      "loss": 0.2608,
+      "epoch": 2.789670955418604,
+      "grad_norm": 0.2418082356452942,
+      "learning_rate": 2.1550715426422194e-05,
+      "loss": 0.3953,
       "step": 77405
     },
     {
-      "epoch": 2.72,
-      "learning_rate": 2.2613845590866924e-05,
-      "loss": 0.2521,
+      "epoch": 2.7898511550798286,
+      "grad_norm": 0.19211691617965698,
+      "learning_rate": 2.1547825224676173e-05,
+      "loss": 0.3994,
       "step": 77410
     },
     {
-      "epoch": 2.72,
-      "learning_rate": 2.2611009890625924e-05,
-      "loss": 0.2686,
+      "epoch": 2.7900313547410533,
+      "grad_norm": 0.2310890257358551,
+      "learning_rate": 2.154493506996537e-05,
+      "loss": 0.3642,
       "step": 77415
     },
     {
-      "epoch": 2.72,
-      "learning_rate": 2.2608174221404384e-05,
-      "loss": 0.2489,
+      "epoch": 2.7902115544022776,
+      "grad_norm": 0.24673838913440704,
+      "learning_rate": 2.1542044962329163e-05,
+      "loss": 0.4185,
       "step": 77420
     },
     {
-      "epoch": 2.72,
-      "learning_rate": 2.2605338583239138e-05,
-      "loss": 0.2762,
+      "epoch": 2.7903917540635024,
+      "grad_norm": 0.2144293338060379,
+      "learning_rate": 2.153915490180691e-05,
+      "loss": 0.4018,
       "step": 77425
     },
     {
-      "epoch": 2.72,
-      "learning_rate": 2.260250297616699e-05,
-      "loss": 0.249,
+      "epoch": 2.790571953724727,
+      "grad_norm": 0.1878860741853714,
+      "learning_rate": 2.1536264888438005e-05,
+      "loss": 0.3942,
       "step": 77430
     },
     {
-      "epoch": 2.72,
-      "learning_rate": 2.2599667400224748e-05,
-      "loss": 0.2607,
+      "epoch": 2.7907521533859514,
+      "grad_norm": 0.2370370328426361,
+      "learning_rate": 2.1533374922261835e-05,
+      "loss": 0.3942,
       "step": 77435
     },
     {
-      "epoch": 2.72,
-      "learning_rate": 2.2596831855449255e-05,
-      "loss": 0.2503,
+      "epoch": 2.790932353047176,
+      "grad_norm": 0.20887020230293274,
+      "learning_rate": 2.153048500331774e-05,
+      "loss": 0.4095,
       "step": 77440
     },
     {
-      "epoch": 2.72,
-      "learning_rate": 2.259399634187731e-05,
-      "loss": 0.2347,
+      "epoch": 2.791112552708401,
+      "grad_norm": 0.21834127604961395,
+      "learning_rate": 2.152759513164513e-05,
+      "loss": 0.379,
       "step": 77445
     },
     {
-      "epoch": 2.72,
-      "learning_rate": 2.2591160859545747e-05,
-      "loss": 0.2706,
+      "epoch": 2.7912927523696256,
+      "grad_norm": 0.16670918464660645,
+      "learning_rate": 2.1524705307283348e-05,
+      "loss": 0.3774,
       "step": 77450
     },
     {
-      "epoch": 2.73,
-      "learning_rate": 2.2588325408491356e-05,
-      "loss": 0.2669,
+      "epoch": 2.7914729520308503,
+      "grad_norm": 0.20966099202632904,
+      "learning_rate": 2.15218155302718e-05,
+      "loss": 0.3779,
       "step": 77455
     },
     {
-      "epoch": 2.73,
-      "learning_rate": 2.2585489988750984e-05,
-      "loss": 0.2725,
+      "epoch": 2.791653151692075,
+      "grad_norm": 0.1819358766078949,
+      "learning_rate": 2.1518925800649836e-05,
+      "loss": 0.4119,
       "step": 77460
     },
     {
-      "epoch": 2.73,
-      "learning_rate": 2.2582654600361433e-05,
-      "loss": 0.2924,
+      "epoch": 2.7918333513532994,
+      "grad_norm": 0.22907419502735138,
+      "learning_rate": 2.151603611845683e-05,
+      "loss": 0.3981,
       "step": 77465
     },
     {
-      "epoch": 2.73,
-      "learning_rate": 2.2579819243359507e-05,
-      "loss": 0.2495,
+      "epoch": 2.792013551014524,
+      "grad_norm": 0.2390015870332718,
+      "learning_rate": 2.1513146483732163e-05,
+      "loss": 0.3907,
       "step": 77470
     },
     {
-      "epoch": 2.73,
-      "learning_rate": 2.2576983917782046e-05,
-      "loss": 0.267,
+      "epoch": 2.792193750675749,
+      "grad_norm": 0.20397347211837769,
+      "learning_rate": 2.1510256896515195e-05,
+      "loss": 0.3836,
       "step": 77475
     },
     {
-      "epoch": 2.73,
-      "learning_rate": 2.2574148623665845e-05,
-      "loss": 0.2731,
+      "epoch": 2.792373950336973,
+      "grad_norm": 0.2034919261932373,
+      "learning_rate": 2.1507367356845304e-05,
+      "loss": 0.398,
       "step": 77480
     },
     {
-      "epoch": 2.73,
-      "learning_rate": 2.2571313361047734e-05,
-      "loss": 0.2742,
+      "epoch": 2.792554149998198,
+      "grad_norm": 0.1811286211013794,
+      "learning_rate": 2.150447786476186e-05,
+      "loss": 0.3666,
       "step": 77485
     },
     {
-      "epoch": 2.73,
-      "learning_rate": 2.2568478129964503e-05,
-      "loss": 0.2896,
+      "epoch": 2.7927343496594226,
+      "grad_norm": 0.23901815712451935,
+      "learning_rate": 2.1501588420304218e-05,
+      "loss": 0.4266,
       "step": 77490
     },
     {
-      "epoch": 2.73,
-      "learning_rate": 2.256564293045299e-05,
-      "loss": 0.2554,
+      "epoch": 2.7929145493206473,
+      "grad_norm": 0.23680728673934937,
+      "learning_rate": 2.1498699023511766e-05,
+      "loss": 0.419,
       "step": 77495
     },
     {
-      "epoch": 2.73,
-      "learning_rate": 2.2562807762550006e-05,
-      "loss": 0.2943,
+      "epoch": 2.793094748981872,
+      "grad_norm": 0.22944191098213196,
+      "learning_rate": 2.1495809674423865e-05,
+      "loss": 0.3911,
       "step": 77500
     },
     {
-      "epoch": 2.73,
-      "eval_loss": 0.2611740827560425,
-      "eval_runtime": 10.5438,
-      "eval_samples_per_second": 9.484,
-      "eval_steps_per_second": 9.484,
+      "epoch": 2.793094748981872,
+      "eval_loss": 0.4314689338207245,
+      "eval_runtime": 3.521,
+      "eval_samples_per_second": 28.401,
+      "eval_steps_per_second": 7.1,
       "step": 77500
     },
     {
-      "epoch": 2.73,
-      "learning_rate": 2.255997262629235e-05,
-      "loss": 0.2799,
+      "epoch": 2.7932749486430968,
+      "grad_norm": 0.22808948159217834,
+      "learning_rate": 2.1492920373079867e-05,
+      "loss": 0.3854,
       "step": 77505
     },
     {
-      "epoch": 2.73,
-      "learning_rate": 2.2557137521716834e-05,
-      "loss": 0.2513,
+      "epoch": 2.793455148304321,
+      "grad_norm": 0.17345234751701355,
+      "learning_rate": 2.149003111951916e-05,
+      "loss": 0.3919,
       "step": 77510
     },
     {
-      "epoch": 2.73,
-      "learning_rate": 2.255430244886029e-05,
-      "loss": 0.2622,
+      "epoch": 2.793635347965546,
+      "grad_norm": 0.18954497575759888,
+      "learning_rate": 2.148714191378109e-05,
+      "loss": 0.3631,
       "step": 77515
     },
     {
-      "epoch": 2.73,
-      "learning_rate": 2.2551467407759515e-05,
-      "loss": 0.2908,
+      "epoch": 2.7938155476267705,
+      "grad_norm": 0.28941574692726135,
+      "learning_rate": 2.148425275590504e-05,
+      "loss": 0.396,
       "step": 77520
     },
     {
-      "epoch": 2.73,
-      "learning_rate": 2.2548632398451316e-05,
-      "loss": 0.2826,
+      "epoch": 2.793995747287995,
+      "grad_norm": 0.2047412395477295,
+      "learning_rate": 2.1481363645930367e-05,
+      "loss": 0.3768,
       "step": 77525
     },
     {
-      "epoch": 2.73,
-      "learning_rate": 2.2545797420972514e-05,
-      "loss": 0.2632,
+      "epoch": 2.7941759469492196,
+      "grad_norm": 0.17408131062984467,
+      "learning_rate": 2.147847458389642e-05,
+      "loss": 0.409,
       "step": 77530
     },
     {
-      "epoch": 2.73,
-      "learning_rate": 2.254296247535991e-05,
-      "loss": 0.2727,
+      "epoch": 2.7943561466104443,
+      "grad_norm": 0.23604151606559753,
+      "learning_rate": 2.147558556984259e-05,
+      "loss": 0.3933,
       "step": 77535
     },
     {
-      "epoch": 2.73,
-      "learning_rate": 2.2540127561650327e-05,
-      "loss": 0.2617,
+      "epoch": 2.794536346271669,
+      "grad_norm": 0.22431814670562744,
+      "learning_rate": 2.147269660380822e-05,
+      "loss": 0.4178,
       "step": 77540
     },
     {
-      "epoch": 2.73,
-      "learning_rate": 2.2537292679880567e-05,
-      "loss": 0.2768,
+      "epoch": 2.7947165459328938,
+      "grad_norm": 0.2082727700471878,
+      "learning_rate": 2.1469807685832672e-05,
+      "loss": 0.3803,
       "step": 77545
     },
     {
-      "epoch": 2.73,
-      "learning_rate": 2.2534457830087428e-05,
-      "loss": 0.2644,
+      "epoch": 2.7948967455941185,
+      "grad_norm": 0.2045706808567047,
+      "learning_rate": 2.146691881595531e-05,
+      "loss": 0.401,
       "step": 77550
     },
     {
-      "epoch": 2.73,
-      "learning_rate": 2.253162301230774e-05,
-      "loss": 0.2605,
+      "epoch": 2.795076945255343,
+      "grad_norm": 0.17748849093914032,
+      "learning_rate": 2.146402999421549e-05,
+      "loss": 0.3753,
       "step": 77555
     },
     {
-      "epoch": 2.73,
-      "learning_rate": 2.2528788226578292e-05,
-      "loss": 0.2768,
+      "epoch": 2.7952571449165675,
+      "grad_norm": 0.19480100274085999,
+      "learning_rate": 2.1461141220652593e-05,
+      "loss": 0.407,
       "step": 77560
     },
     {
-      "epoch": 2.73,
-      "learning_rate": 2.252595347293591e-05,
-      "loss": 0.2968,
+      "epoch": 2.7954373445777922,
+      "grad_norm": 0.2130713313817978,
+      "learning_rate": 2.1458252495305954e-05,
+      "loss": 0.368,
       "step": 77565
     },
     {
-      "epoch": 2.73,
-      "learning_rate": 2.252311875141738e-05,
-      "loss": 0.2626,
+      "epoch": 2.795617544239017,
+      "grad_norm": 0.22572968900203705,
+      "learning_rate": 2.145536381821493e-05,
+      "loss": 0.396,
       "step": 77570
     },
     {
-      "epoch": 2.73,
-      "learning_rate": 2.2520284062059527e-05,
-      "loss": 0.2552,
+      "epoch": 2.7957977439002413,
+      "grad_norm": 0.21448364853858948,
+      "learning_rate": 2.14524751894189e-05,
+      "loss": 0.3612,
       "step": 77575
     },
     {
-      "epoch": 2.73,
-      "learning_rate": 2.2517449404899153e-05,
-      "loss": 0.2794,
+      "epoch": 2.795977943561466,
+      "grad_norm": 0.2889059782028198,
+      "learning_rate": 2.1449586608957207e-05,
+      "loss": 0.4111,
       "step": 77580
     },
     {
-      "epoch": 2.73,
-      "learning_rate": 2.2514614779973055e-05,
-      "loss": 0.2895,
+      "epoch": 2.7961581432226907,
+      "grad_norm": 0.23263846337795258,
+      "learning_rate": 2.1446698076869203e-05,
+      "loss": 0.4039,
       "step": 77585
     },
     {
-      "epoch": 2.73,
-      "learning_rate": 2.251178018731805e-05,
-      "loss": 0.2437,
+      "epoch": 2.7963383428839155,
+      "grad_norm": 0.22178532183170319,
+      "learning_rate": 2.1443809593194253e-05,
+      "loss": 0.4065,
       "step": 77590
     },
     {
-      "epoch": 2.73,
-      "learning_rate": 2.2508945626970945e-05,
-      "loss": 0.2697,
+      "epoch": 2.79651854254514,
+      "grad_norm": 0.19402441382408142,
+      "learning_rate": 2.1440921157971706e-05,
+      "loss": 0.3759,
       "step": 77595
     },
     {
-      "epoch": 2.73,
-      "learning_rate": 2.2506111098968537e-05,
-      "loss": 0.259,
+      "epoch": 2.7966987422063645,
+      "grad_norm": 0.22122874855995178,
+      "learning_rate": 2.1438032771240925e-05,
+      "loss": 0.4054,
       "step": 77600
     },
     {
-      "epoch": 2.73,
-      "learning_rate": 2.2503276603347626e-05,
-      "loss": 0.26,
+      "epoch": 2.7968789418675892,
+      "grad_norm": 0.2299935668706894,
+      "learning_rate": 2.143514443304126e-05,
+      "loss": 0.3833,
       "step": 77605
     },
     {
-      "epoch": 2.73,
-      "learning_rate": 2.250044214014503e-05,
-      "loss": 0.261,
+      "epoch": 2.797059141528814,
+      "grad_norm": 0.23036040365695953,
+      "learning_rate": 2.143225614341205e-05,
+      "loss": 0.4046,
       "step": 77610
     },
     {
-      "epoch": 2.73,
-      "learning_rate": 2.2497607709397543e-05,
-      "loss": 0.2581,
+      "epoch": 2.7972393411900387,
+      "grad_norm": 0.21337628364562988,
+      "learning_rate": 2.1429367902392668e-05,
+      "loss": 0.3539,
       "step": 77615
     },
     {
-      "epoch": 2.73,
-      "learning_rate": 2.2494773311141974e-05,
-      "loss": 0.2515,
+      "epoch": 2.797419540851263,
+      "grad_norm": 0.2017783671617508,
+      "learning_rate": 2.1426479710022463e-05,
+      "loss": 0.3555,
       "step": 77620
     },
     {
-      "epoch": 2.73,
-      "learning_rate": 2.2491938945415114e-05,
-      "loss": 0.2459,
+      "epoch": 2.7975997405124877,
+      "grad_norm": 0.2635432183742523,
+      "learning_rate": 2.142359156634076e-05,
+      "loss": 0.4249,
       "step": 77625
     },
     {
-      "epoch": 2.73,
-      "learning_rate": 2.2489104612253784e-05,
-      "loss": 0.254,
+      "epoch": 2.7977799401737125,
+      "grad_norm": 0.1856364905834198,
+      "learning_rate": 2.1420703471386952e-05,
+      "loss": 0.41,
       "step": 77630
     },
     {
-      "epoch": 2.73,
-      "learning_rate": 2.2486270311694775e-05,
-      "loss": 0.2882,
+      "epoch": 2.797960139834937,
+      "grad_norm": 0.1753203123807907,
+      "learning_rate": 2.1417815425200346e-05,
+      "loss": 0.3796,
       "step": 77635
     },
     {
-      "epoch": 2.73,
-      "learning_rate": 2.2483436043774884e-05,
-      "loss": 0.274,
+      "epoch": 2.798140339496162,
+      "grad_norm": 0.23167075216770172,
+      "learning_rate": 2.141492742782033e-05,
+      "loss": 0.3789,
       "step": 77640
     },
     {
-      "epoch": 2.73,
-      "learning_rate": 2.2480601808530914e-05,
-      "loss": 0.2703,
+      "epoch": 2.7983205391573867,
+      "grad_norm": 0.17792221903800964,
+      "learning_rate": 2.141203947928623e-05,
+      "loss": 0.3429,
       "step": 77645
     },
     {
-      "epoch": 2.73,
-      "learning_rate": 2.2477767605999676e-05,
-      "loss": 0.2312,
+      "epoch": 2.798500738818611,
+      "grad_norm": 0.23163169622421265,
+      "learning_rate": 2.1409151579637386e-05,
+      "loss": 0.4213,
       "step": 77650
     },
     {
-      "epoch": 2.73,
-      "learning_rate": 2.2474933436217964e-05,
-      "loss": 0.2711,
+      "epoch": 2.7986809384798357,
+      "grad_norm": 0.21783442795276642,
+      "learning_rate": 2.1406263728913163e-05,
+      "loss": 0.3908,
       "step": 77655
     },
     {
-      "epoch": 2.73,
-      "learning_rate": 2.2472099299222583e-05,
-      "loss": 0.2628,
+      "epoch": 2.7988611381410604,
+      "grad_norm": 0.2356899380683899,
+      "learning_rate": 2.1403375927152903e-05,
+      "loss": 0.3945,
       "step": 77660
     },
     {
-      "epoch": 2.73,
-      "learning_rate": 2.2469265195050308e-05,
-      "loss": 0.2846,
+      "epoch": 2.7990413378022847,
+      "grad_norm": 0.23855198919773102,
+      "learning_rate": 2.1400488174395946e-05,
+      "loss": 0.4182,
       "step": 77665
     },
     {
-      "epoch": 2.73,
-      "learning_rate": 2.2466431123737968e-05,
-      "loss": 0.2754,
+      "epoch": 2.7992215374635094,
+      "grad_norm": 0.20778150856494904,
+      "learning_rate": 2.1397600470681643e-05,
+      "loss": 0.3909,
       "step": 77670
     },
     {
-      "epoch": 2.73,
-      "learning_rate": 2.2463597085322352e-05,
-      "loss": 0.2636,
+      "epoch": 2.799401737124734,
+      "grad_norm": 0.19535931944847107,
+      "learning_rate": 2.1394712816049332e-05,
+      "loss": 0.3873,
       "step": 77675
     },
     {
-      "epoch": 2.73,
-      "learning_rate": 2.2460763079840257e-05,
-      "loss": 0.2799,
+      "epoch": 2.799581936785959,
+      "grad_norm": 0.21138401329517365,
+      "learning_rate": 2.1391825210538366e-05,
+      "loss": 0.4423,
       "step": 77680
     },
     {
-      "epoch": 2.73,
-      "learning_rate": 2.2457929107328473e-05,
-      "loss": 0.2776,
+      "epoch": 2.7997621364471836,
+      "grad_norm": 0.25198328495025635,
+      "learning_rate": 2.138893765418808e-05,
+      "loss": 0.417,
       "step": 77685
     },
     {
-      "epoch": 2.73,
-      "learning_rate": 2.245509516782381e-05,
-      "loss": 0.2927,
+      "epoch": 2.7999423361084084,
+      "grad_norm": 0.1913152039051056,
+      "learning_rate": 2.138605014703782e-05,
+      "loss": 0.3844,
       "step": 77690
     },
     {
-      "epoch": 2.73,
-      "learning_rate": 2.2452261261363056e-05,
-      "loss": 0.2628,
+      "epoch": 2.8001225357696327,
+      "grad_norm": 0.19458575546741486,
+      "learning_rate": 2.138316268912693e-05,
+      "loss": 0.3818,
       "step": 77695
     },
     {
-      "epoch": 2.73,
-      "learning_rate": 2.2449427387983004e-05,
-      "loss": 0.2494,
+      "epoch": 2.8003027354308574,
+      "grad_norm": 0.25212234258651733,
+      "learning_rate": 2.138027528049474e-05,
+      "loss": 0.3994,
       "step": 77700
     },
     {
-      "epoch": 2.73,
-      "learning_rate": 2.244659354772047e-05,
-      "loss": 0.2591,
+      "epoch": 2.800482935092082,
+      "grad_norm": 0.19638745486736298,
+      "learning_rate": 2.137738792118061e-05,
+      "loss": 0.3779,
       "step": 77705
     },
     {
-      "epoch": 2.73,
-      "learning_rate": 2.2443759740612234e-05,
-      "loss": 0.2602,
+      "epoch": 2.8006631347533064,
+      "grad_norm": 0.2299245297908783,
+      "learning_rate": 2.1374500611223867e-05,
+      "loss": 0.4019,
       "step": 77710
     },
     {
-      "epoch": 2.73,
-      "learning_rate": 2.2440925966695093e-05,
-      "loss": 0.2782,
+      "epoch": 2.800843334414531,
+      "grad_norm": 0.2781182527542114,
+      "learning_rate": 2.1371613350663837e-05,
+      "loss": 0.3782,
       "step": 77715
     },
     {
-      "epoch": 2.73,
-      "learning_rate": 2.2438092226005832e-05,
-      "loss": 0.267,
+      "epoch": 2.801023534075756,
+      "grad_norm": 0.2580137848854065,
+      "learning_rate": 2.1368726139539885e-05,
+      "loss": 0.4054,
       "step": 77720
     },
     {
-      "epoch": 2.73,
-      "learning_rate": 2.243525851858126e-05,
-      "loss": 0.2685,
+      "epoch": 2.8012037337369806,
+      "grad_norm": 0.2500860095024109,
+      "learning_rate": 2.1365838977891344e-05,
+      "loss": 0.4145,
       "step": 77725
     },
     {
-      "epoch": 2.73,
-      "learning_rate": 2.2432424844458168e-05,
-      "loss": 0.2624,
+      "epoch": 2.8013839333982054,
+      "grad_norm": 0.26811492443084717,
+      "learning_rate": 2.1362951865757523e-05,
+      "loss": 0.3767,
       "step": 77730
     },
     {
-      "epoch": 2.73,
-      "learning_rate": 2.2429591203673353e-05,
-      "loss": 0.2586,
+      "epoch": 2.80156413305943,
+      "grad_norm": 0.22250352799892426,
+      "learning_rate": 2.1360064803177793e-05,
+      "loss": 0.3912,
       "step": 77735
     },
     {
-      "epoch": 2.74,
-      "learning_rate": 2.242675759626359e-05,
-      "loss": 0.276,
+      "epoch": 2.8017443327206544,
+      "grad_norm": 0.16757109761238098,
+      "learning_rate": 2.1357177790191463e-05,
+      "loss": 0.3749,
       "step": 77740
     },
     {
-      "epoch": 2.74,
-      "learning_rate": 2.242392402226569e-05,
-      "loss": 0.2892,
+      "epoch": 2.801924532381879,
+      "grad_norm": 0.1849580705165863,
+      "learning_rate": 2.13542908268379e-05,
+      "loss": 0.3972,
       "step": 77745
     },
     {
-      "epoch": 2.74,
-      "learning_rate": 2.242109048171644e-05,
-      "loss": 0.2714,
+      "epoch": 2.802104732043104,
+      "grad_norm": 0.2086080014705658,
+      "learning_rate": 2.1351403913156403e-05,
+      "loss": 0.4013,
       "step": 77750
     },
     {
-      "epoch": 2.74,
-      "learning_rate": 2.241825697465263e-05,
-      "loss": 0.2524,
+      "epoch": 2.802284931704328,
+      "grad_norm": 0.22372741997241974,
+      "learning_rate": 2.1348517049186323e-05,
+      "loss": 0.4117,
       "step": 77755
     },
     {
-      "epoch": 2.74,
-      "learning_rate": 2.2415423501111046e-05,
-      "loss": 0.2848,
+      "epoch": 2.802465131365553,
+      "grad_norm": 0.18635831773281097,
+      "learning_rate": 2.1345630234966997e-05,
+      "loss": 0.3779,
       "step": 77760
     },
     {
-      "epoch": 2.74,
-      "learning_rate": 2.241259006112849e-05,
-      "loss": 0.2867,
+      "epoch": 2.8026453310267776,
+      "grad_norm": 0.19005261361598969,
+      "learning_rate": 2.134274347053775e-05,
+      "loss": 0.4248,
       "step": 77765
     },
     {
-      "epoch": 2.74,
-      "learning_rate": 2.2409756654741753e-05,
-      "loss": 0.2579,
+      "epoch": 2.8028255306880023,
+      "grad_norm": 0.19459028542041779,
+      "learning_rate": 2.133985675593791e-05,
+      "loss": 0.4361,
       "step": 77770
     },
     {
-      "epoch": 2.74,
-      "learning_rate": 2.240692328198761e-05,
-      "loss": 0.2494,
+      "epoch": 2.803005730349227,
+      "grad_norm": 0.19851277768611908,
+      "learning_rate": 2.1336970091206814e-05,
+      "loss": 0.3814,
       "step": 77775
     },
     {
-      "epoch": 2.74,
-      "learning_rate": 2.2404089942902855e-05,
-      "loss": 0.2742,
+      "epoch": 2.803185930010452,
+      "grad_norm": 0.1843992918729782,
+      "learning_rate": 2.1334083476383794e-05,
+      "loss": 0.4244,
       "step": 77780
     },
     {
-      "epoch": 2.74,
-      "learning_rate": 2.240125663752429e-05,
-      "loss": 0.2523,
+      "epoch": 2.803366129671676,
+      "grad_norm": 0.24044600129127502,
+      "learning_rate": 2.133119691150818e-05,
+      "loss": 0.3871,
       "step": 77785
     },
     {
-      "epoch": 2.74,
-      "learning_rate": 2.2398423365888697e-05,
-      "loss": 0.2723,
+      "epoch": 2.803546329332901,
+      "grad_norm": 0.1807718724012375,
+      "learning_rate": 2.1328310396619296e-05,
+      "loss": 0.4099,
       "step": 77790
     },
     {
-      "epoch": 2.74,
-      "learning_rate": 2.239559012803286e-05,
-      "loss": 0.2681,
+      "epoch": 2.8037265289941256,
+      "grad_norm": 0.22799517214298248,
+      "learning_rate": 2.1325423931756463e-05,
+      "loss": 0.3899,
       "step": 77795
     },
     {
-      "epoch": 2.74,
-      "learning_rate": 2.239275692399356e-05,
-      "loss": 0.2579,
+      "epoch": 2.8039067286553503,
+      "grad_norm": 0.2450421303510666,
+      "learning_rate": 2.1322537516959026e-05,
+      "loss": 0.3654,
       "step": 77800
     },
     {
-      "epoch": 2.74,
-      "learning_rate": 2.23899237538076e-05,
-      "loss": 0.2741,
+      "epoch": 2.8040869283165746,
+      "grad_norm": 0.21578456461429596,
+      "learning_rate": 2.1319651152266313e-05,
+      "loss": 0.3766,
       "step": 77805
     },
     {
-      "epoch": 2.74,
-      "learning_rate": 2.2387090617511763e-05,
-      "loss": 0.2572,
+      "epoch": 2.8042671279777993,
+      "grad_norm": 0.22823040187358856,
+      "learning_rate": 2.1316764837717618e-05,
+      "loss": 0.4372,
       "step": 77810
     },
     {
-      "epoch": 2.74,
-      "learning_rate": 2.238425751514282e-05,
-      "loss": 0.2713,
+      "epoch": 2.804447327639024,
+      "grad_norm": 0.2035275250673294,
+      "learning_rate": 2.131387857335231e-05,
+      "loss": 0.3643,
       "step": 77815
     },
     {
-      "epoch": 2.74,
-      "learning_rate": 2.2381424446737585e-05,
-      "loss": 0.2708,
+      "epoch": 2.804627527300249,
+      "grad_norm": 0.23069792985916138,
+      "learning_rate": 2.1310992359209673e-05,
+      "loss": 0.3928,
       "step": 77820
     },
     {
-      "epoch": 2.74,
-      "learning_rate": 2.237859141233282e-05,
-      "loss": 0.2514,
+      "epoch": 2.8048077269614735,
+      "grad_norm": 0.18597543239593506,
+      "learning_rate": 2.130810619532907e-05,
+      "loss": 0.3802,
       "step": 77825
     },
     {
-      "epoch": 2.74,
-      "learning_rate": 2.237575841196532e-05,
-      "loss": 0.2613,
+      "epoch": 2.804987926622698,
+      "grad_norm": 0.14916902780532837,
+      "learning_rate": 2.1305220081749798e-05,
+      "loss": 0.3602,
       "step": 77830
     },
     {
-      "epoch": 2.74,
-      "learning_rate": 2.2372925445671866e-05,
-      "loss": 0.2606,
+      "epoch": 2.8051681262839225,
+      "grad_norm": 0.20437051355838776,
+      "learning_rate": 2.130233401851118e-05,
+      "loss": 0.3659,
       "step": 77835
     },
     {
-      "epoch": 2.74,
-      "learning_rate": 2.237009251348924e-05,
-      "loss": 0.2792,
+      "epoch": 2.8053483259451473,
+      "grad_norm": 0.25175103545188904,
+      "learning_rate": 2.1299448005652552e-05,
+      "loss": 0.3947,
       "step": 77840
     },
     {
-      "epoch": 2.74,
-      "learning_rate": 2.236725961545424e-05,
-      "loss": 0.2749,
+      "epoch": 2.805528525606372,
+      "grad_norm": 0.18190394341945648,
+      "learning_rate": 2.129656204321323e-05,
+      "loss": 0.3906,
       "step": 77845
     },
     {
-      "epoch": 2.74,
-      "learning_rate": 2.2364426751603637e-05,
-      "loss": 0.2729,
+      "epoch": 2.8057087252675963,
+      "grad_norm": 0.16877366602420807,
+      "learning_rate": 2.1293676131232526e-05,
+      "loss": 0.3896,
       "step": 77850
     },
     {
-      "epoch": 2.74,
-      "learning_rate": 2.23615939219742e-05,
-      "loss": 0.2763,
+      "epoch": 2.805888924928821,
+      "grad_norm": 0.19917549192905426,
+      "learning_rate": 2.1290790269749774e-05,
+      "loss": 0.3843,
       "step": 77855
     },
     {
-      "epoch": 2.74,
-      "learning_rate": 2.235876112660274e-05,
-      "loss": 0.2885,
+      "epoch": 2.8060691245900458,
+      "grad_norm": 0.23629820346832275,
+      "learning_rate": 2.128790445880428e-05,
+      "loss": 0.3683,
       "step": 77860
     },
     {
-      "epoch": 2.74,
-      "learning_rate": 2.235592836552603e-05,
-      "loss": 0.2678,
+      "epoch": 2.8062493242512705,
+      "grad_norm": 0.2526760697364807,
+      "learning_rate": 2.128501869843538e-05,
+      "loss": 0.3883,
       "step": 77865
     },
     {
-      "epoch": 2.74,
-      "learning_rate": 2.2353095638780845e-05,
-      "loss": 0.288,
+      "epoch": 2.8064295239124952,
+      "grad_norm": 0.22685180604457855,
+      "learning_rate": 2.1282132988682374e-05,
+      "loss": 0.3851,
       "step": 77870
     },
     {
-      "epoch": 2.74,
-      "learning_rate": 2.235026294640396e-05,
-      "loss": 0.2635,
+      "epoch": 2.8066097235737195,
+      "grad_norm": 0.24433378875255585,
+      "learning_rate": 2.1279247329584582e-05,
+      "loss": 0.4229,
       "step": 77875
     },
     {
-      "epoch": 2.74,
-      "learning_rate": 2.2347430288432172e-05,
-      "loss": 0.2696,
+      "epoch": 2.8067899232349443,
+      "grad_norm": 0.2599703371524811,
+      "learning_rate": 2.1276361721181332e-05,
+      "loss": 0.4136,
       "step": 77880
     },
     {
-      "epoch": 2.74,
-      "learning_rate": 2.2344597664902254e-05,
-      "loss": 0.2782,
+      "epoch": 2.806970122896169,
+      "grad_norm": 0.22800195217132568,
+      "learning_rate": 2.1273476163511924e-05,
+      "loss": 0.3937,
       "step": 77885
     },
     {
-      "epoch": 2.74,
-      "learning_rate": 2.2341765075850985e-05,
-      "loss": 0.2874,
+      "epoch": 2.8071503225573937,
+      "grad_norm": 0.26085367798805237,
+      "learning_rate": 2.127059065661569e-05,
+      "loss": 0.4329,
       "step": 77890
     },
     {
-      "epoch": 2.74,
-      "learning_rate": 2.2338932521315136e-05,
-      "loss": 0.2768,
+      "epoch": 2.807330522218618,
+      "grad_norm": 0.1912272721529007,
+      "learning_rate": 2.1267705200531933e-05,
+      "loss": 0.4017,
       "step": 77895
     },
     {
-      "epoch": 2.74,
-      "learning_rate": 2.2336100001331506e-05,
-      "loss": 0.2596,
+      "epoch": 2.8075107218798427,
+      "grad_norm": 0.2735271751880646,
+      "learning_rate": 2.126481979529997e-05,
+      "loss": 0.4036,
       "step": 77900
     },
     {
-      "epoch": 2.74,
-      "learning_rate": 2.2333267515936863e-05,
-      "loss": 0.2539,
+      "epoch": 2.8076909215410675,
+      "grad_norm": 0.2603655755519867,
+      "learning_rate": 2.1261934440959115e-05,
+      "loss": 0.4244,
       "step": 77905
     },
     {
-      "epoch": 2.74,
-      "learning_rate": 2.2330435065167975e-05,
-      "loss": 0.2689,
+      "epoch": 2.807871121202292,
+      "grad_norm": 0.21912328898906708,
+      "learning_rate": 2.1259049137548686e-05,
+      "loss": 0.4253,
       "step": 77910
     },
     {
-      "epoch": 2.74,
-      "learning_rate": 2.232760264906163e-05,
-      "loss": 0.256,
+      "epoch": 2.808051320863517,
+      "grad_norm": 0.19126076996326447,
+      "learning_rate": 2.1256163885107973e-05,
+      "loss": 0.4077,
       "step": 77915
     },
     {
-      "epoch": 2.74,
-      "learning_rate": 2.2324770267654603e-05,
-      "loss": 0.2695,
+      "epoch": 2.8082315205247417,
+      "grad_norm": 0.19802503287792206,
+      "learning_rate": 2.125327868367632e-05,
+      "loss": 0.3619,
       "step": 77920
     },
     {
-      "epoch": 2.74,
-      "learning_rate": 2.2321937920983675e-05,
-      "loss": 0.2547,
+      "epoch": 2.808411720185966,
+      "grad_norm": 0.2620820105075836,
+      "learning_rate": 2.1250393533293e-05,
+      "loss": 0.4068,
       "step": 77925
     },
     {
-      "epoch": 2.74,
-      "learning_rate": 2.2319105609085612e-05,
-      "loss": 0.2681,
+      "epoch": 2.8085919198471907,
+      "grad_norm": 0.269182413816452,
+      "learning_rate": 2.124750843399736e-05,
+      "loss": 0.4109,
       "step": 77930
     },
     {
-      "epoch": 2.74,
-      "learning_rate": 2.2316273331997204e-05,
-      "loss": 0.2799,
+      "epoch": 2.8087721195084154,
+      "grad_norm": 0.2248949408531189,
+      "learning_rate": 2.1244623385828687e-05,
+      "loss": 0.4336,
       "step": 77935
     },
     {
-      "epoch": 2.74,
-      "learning_rate": 2.2313441089755217e-05,
-      "loss": 0.2527,
+      "epoch": 2.8089523191696397,
+      "grad_norm": 0.2180756777524948,
+      "learning_rate": 2.1241738388826288e-05,
+      "loss": 0.3815,
       "step": 77940
     },
     {
-      "epoch": 2.74,
-      "learning_rate": 2.2310608882396425e-05,
-      "loss": 0.2718,
+      "epoch": 2.8091325188308645,
+      "grad_norm": 0.22202380001544952,
+      "learning_rate": 2.1238853443029476e-05,
+      "loss": 0.3617,
       "step": 77945
     },
     {
-      "epoch": 2.74,
-      "learning_rate": 2.2307776709957596e-05,
-      "loss": 0.2551,
+      "epoch": 2.809312718492089,
+      "grad_norm": 0.24584850668907166,
+      "learning_rate": 2.1235968548477564e-05,
+      "loss": 0.3859,
       "step": 77950
     },
     {
-      "epoch": 2.74,
-      "learning_rate": 2.2304944572475523e-05,
-      "loss": 0.2872,
+      "epoch": 2.809492918153314,
+      "grad_norm": 0.18127258121967316,
+      "learning_rate": 2.1233083705209845e-05,
+      "loss": 0.3916,
       "step": 77955
     },
     {
-      "epoch": 2.74,
-      "learning_rate": 2.230211246998697e-05,
-      "loss": 0.2838,
+      "epoch": 2.8096731178145387,
+      "grad_norm": 0.2721177637577057,
+      "learning_rate": 2.1230198913265635e-05,
+      "loss": 0.3872,
       "step": 77960
     },
     {
-      "epoch": 2.74,
-      "learning_rate": 2.2299280402528702e-05,
-      "loss": 0.275,
+      "epoch": 2.8098533174757634,
+      "grad_norm": 0.23304437100887299,
+      "learning_rate": 2.122731417268423e-05,
+      "loss": 0.4107,
       "step": 77965
     },
     {
-      "epoch": 2.74,
-      "learning_rate": 2.2296448370137498e-05,
-      "loss": 0.26,
+      "epoch": 2.8100335171369877,
+      "grad_norm": 0.21200516819953918,
+      "learning_rate": 2.122442948350494e-05,
+      "loss": 0.4215,
       "step": 77970
     },
     {
-      "epoch": 2.74,
-      "learning_rate": 2.229361637285013e-05,
-      "loss": 0.2651,
+      "epoch": 2.8102137167982124,
+      "grad_norm": 0.1965387612581253,
+      "learning_rate": 2.1221544845767074e-05,
+      "loss": 0.4209,
       "step": 77975
     },
     {
-      "epoch": 2.74,
-      "learning_rate": 2.229078441070338e-05,
-      "loss": 0.2758,
+      "epoch": 2.810393916459437,
+      "grad_norm": 0.19175685942173004,
+      "learning_rate": 2.121866025950992e-05,
+      "loss": 0.383,
       "step": 77980
     },
     {
-      "epoch": 2.74,
-      "learning_rate": 2.2287952483734004e-05,
-      "loss": 0.277,
+      "epoch": 2.8105741161206614,
+      "grad_norm": 0.2140149623155594,
+      "learning_rate": 2.1215775724772794e-05,
+      "loss": 0.4087,
       "step": 77985
     },
     {
-      "epoch": 2.74,
-      "learning_rate": 2.2285120591978766e-05,
-      "loss": 0.2619,
+      "epoch": 2.810754315781886,
+      "grad_norm": 0.239614799618721,
+      "learning_rate": 2.1212891241594996e-05,
+      "loss": 0.3841,
       "step": 77990
     },
     {
-      "epoch": 2.74,
-      "learning_rate": 2.228228873547446e-05,
-      "loss": 0.2605,
+      "epoch": 2.810934515443111,
+      "grad_norm": 0.24127493798732758,
+      "learning_rate": 2.1210006810015803e-05,
+      "loss": 0.375,
       "step": 77995
     },
     {
-      "epoch": 2.74,
-      "learning_rate": 2.227945691425784e-05,
-      "loss": 0.2698,
+      "epoch": 2.8111147151043356,
+      "grad_norm": 0.20837131142616272,
+      "learning_rate": 2.120712243007455e-05,
+      "loss": 0.4066,
       "step": 78000
     },
     {
-      "epoch": 2.74,
-      "eval_loss": 0.26091867685317993,
-      "eval_runtime": 10.5455,
-      "eval_samples_per_second": 9.483,
-      "eval_steps_per_second": 9.483,
+      "epoch": 2.8111147151043356,
+      "eval_loss": 0.4307841956615448,
+      "eval_runtime": 3.5376,
+      "eval_samples_per_second": 28.268,
+      "eval_steps_per_second": 7.067,
       "step": 78000
     },
     {
-      "epoch": 2.74,
-      "learning_rate": 2.2276625128365683e-05,
-      "loss": 0.2536,
+      "epoch": 2.8112949147655604,
+      "grad_norm": 0.21309497952461243,
+      "learning_rate": 2.1204238101810507e-05,
+      "loss": 0.3937,
       "step": 78005
     },
     {
-      "epoch": 2.74,
-      "learning_rate": 2.2273793377834744e-05,
-      "loss": 0.2584,
+      "epoch": 2.811475114426785,
+      "grad_norm": 0.23278546333312988,
+      "learning_rate": 2.1201353825262996e-05,
+      "loss": 0.4088,
       "step": 78010
     },
     {
-      "epoch": 2.74,
-      "learning_rate": 2.227096166270181e-05,
-      "loss": 0.251,
+      "epoch": 2.8116553140880094,
+      "grad_norm": 0.1590275913476944,
+      "learning_rate": 2.1198469600471308e-05,
+      "loss": 0.3719,
       "step": 78015
     },
     {
-      "epoch": 2.74,
-      "learning_rate": 2.2268129983003643e-05,
-      "loss": 0.2848,
+      "epoch": 2.811835513749234,
+      "grad_norm": 0.19892996549606323,
+      "learning_rate": 2.119558542747472e-05,
+      "loss": 0.3786,
       "step": 78020
     },
     {
-      "epoch": 2.75,
-      "learning_rate": 2.2265298338777003e-05,
-      "loss": 0.2954,
+      "epoch": 2.812015713410459,
+      "grad_norm": 0.22030261158943176,
+      "learning_rate": 2.1192701306312556e-05,
+      "loss": 0.3856,
       "step": 78025
     },
     {
-      "epoch": 2.75,
-      "learning_rate": 2.2262466730058652e-05,
-      "loss": 0.2684,
+      "epoch": 2.812195913071683,
+      "grad_norm": 0.2299405336380005,
+      "learning_rate": 2.11898172370241e-05,
+      "loss": 0.4055,
       "step": 78030
     },
     {
-      "epoch": 2.75,
-      "learning_rate": 2.2259635156885384e-05,
-      "loss": 0.2402,
+      "epoch": 2.812376112732908,
+      "grad_norm": 0.22635617852210999,
+      "learning_rate": 2.1186933219648636e-05,
+      "loss": 0.3831,
       "step": 78035
     },
     {
-      "epoch": 2.75,
-      "learning_rate": 2.225680361929394e-05,
-      "loss": 0.2574,
+      "epoch": 2.8125563123941326,
+      "grad_norm": 0.1920858919620514,
+      "learning_rate": 2.1184049254225477e-05,
+      "loss": 0.3861,
       "step": 78040
     },
     {
-      "epoch": 2.75,
-      "learning_rate": 2.2253972117321088e-05,
-      "loss": 0.2703,
+      "epoch": 2.8127365120553574,
+      "grad_norm": 0.256536066532135,
+      "learning_rate": 2.1181165340793902e-05,
+      "loss": 0.4304,
       "step": 78045
     },
     {
-      "epoch": 2.75,
-      "learning_rate": 2.2251140651003608e-05,
-      "loss": 0.2725,
+      "epoch": 2.812916711716582,
+      "grad_norm": 0.1779921054840088,
+      "learning_rate": 2.1178281479393213e-05,
+      "loss": 0.3844,
       "step": 78050
     },
     {
-      "epoch": 2.75,
-      "learning_rate": 2.224830922037825e-05,
-      "loss": 0.2775,
+      "epoch": 2.813096911377807,
+      "grad_norm": 0.20656806230545044,
+      "learning_rate": 2.11753976700627e-05,
+      "loss": 0.3907,
       "step": 78055
     },
     {
-      "epoch": 2.75,
-      "learning_rate": 2.2245477825481786e-05,
-      "loss": 0.2936,
+      "epoch": 2.813277111039031,
+      "grad_norm": 0.2381805181503296,
+      "learning_rate": 2.117251391284165e-05,
+      "loss": 0.4102,
       "step": 78060
     },
     {
-      "epoch": 2.75,
-      "learning_rate": 2.2242646466350966e-05,
-      "loss": 0.2614,
+      "epoch": 2.813457310700256,
+      "grad_norm": 0.2606278657913208,
+      "learning_rate": 2.1169630207769366e-05,
+      "loss": 0.3904,
       "step": 78065
     },
     {
-      "epoch": 2.75,
-      "learning_rate": 2.223981514302258e-05,
-      "loss": 0.2573,
+      "epoch": 2.8136375103614806,
+      "grad_norm": 0.24265384674072266,
+      "learning_rate": 2.116674655488512e-05,
+      "loss": 0.3964,
       "step": 78070
     },
     {
-      "epoch": 2.75,
-      "learning_rate": 2.223698385553337e-05,
-      "loss": 0.2834,
+      "epoch": 2.8138177100227053,
+      "grad_norm": 0.20753905177116394,
+      "learning_rate": 2.116386295422821e-05,
+      "loss": 0.395,
       "step": 78075
     },
     {
-      "epoch": 2.75,
-      "learning_rate": 2.2234152603920102e-05,
-      "loss": 0.266,
+      "epoch": 2.8139979096839296,
+      "grad_norm": 0.19484953582286835,
+      "learning_rate": 2.116097940583793e-05,
+      "loss": 0.4158,
       "step": 78080
     },
     {
-      "epoch": 2.75,
-      "learning_rate": 2.2231321388219532e-05,
-      "loss": 0.2543,
+      "epoch": 2.8141781093451543,
+      "grad_norm": 0.17793631553649902,
+      "learning_rate": 2.1158095909753555e-05,
+      "loss": 0.404,
       "step": 78085
     },
     {
-      "epoch": 2.75,
-      "learning_rate": 2.222849020846844e-05,
-      "loss": 0.2598,
+      "epoch": 2.814358309006379,
+      "grad_norm": 0.226637065410614,
+      "learning_rate": 2.1155212466014383e-05,
+      "loss": 0.3846,
       "step": 78090
     },
     {
-      "epoch": 2.75,
-      "learning_rate": 2.222565906470358e-05,
-      "loss": 0.2721,
+      "epoch": 2.814538508667604,
+      "grad_norm": 0.24022680521011353,
+      "learning_rate": 2.1152329074659707e-05,
+      "loss": 0.3918,
       "step": 78095
     },
     {
-      "epoch": 2.75,
-      "learning_rate": 2.2222827956961707e-05,
-      "loss": 0.2614,
+      "epoch": 2.8147187083288285,
+      "grad_norm": 0.23119503259658813,
+      "learning_rate": 2.1149445735728784e-05,
+      "loss": 0.3771,
       "step": 78100
     },
     {
-      "epoch": 2.75,
-      "learning_rate": 2.221999688527957e-05,
-      "loss": 0.2784,
+      "epoch": 2.814898907990053,
+      "grad_norm": 0.1578037291765213,
+      "learning_rate": 2.1146562449260933e-05,
+      "loss": 0.3745,
       "step": 78105
     },
     {
-      "epoch": 2.75,
-      "learning_rate": 2.2217165849693948e-05,
-      "loss": 0.2608,
+      "epoch": 2.8150791076512776,
+      "grad_norm": 0.22053155303001404,
+      "learning_rate": 2.1143679215295407e-05,
+      "loss": 0.4288,
       "step": 78110
     },
     {
-      "epoch": 2.75,
-      "learning_rate": 2.2214334850241598e-05,
-      "loss": 0.2884,
+      "epoch": 2.8152593073125023,
+      "grad_norm": 0.21788404881954193,
+      "learning_rate": 2.114079603387152e-05,
+      "loss": 0.4265,
       "step": 78115
     },
     {
-      "epoch": 2.75,
-      "learning_rate": 2.2211503886959272e-05,
-      "loss": 0.2564,
+      "epoch": 2.815439506973727,
+      "grad_norm": 0.21442186832427979,
+      "learning_rate": 2.1137912905028537e-05,
+      "loss": 0.362,
       "step": 78120
     },
     {
-      "epoch": 2.75,
-      "learning_rate": 2.2208672959883726e-05,
-      "loss": 0.2679,
+      "epoch": 2.8156197066349513,
+      "grad_norm": 0.2007633000612259,
+      "learning_rate": 2.113502982880573e-05,
+      "loss": 0.3879,
       "step": 78125
     },
     {
-      "epoch": 2.75,
-      "learning_rate": 2.2205842069051727e-05,
-      "loss": 0.2764,
+      "epoch": 2.815799906296176,
+      "grad_norm": 0.23273354768753052,
+      "learning_rate": 2.113214680524241e-05,
+      "loss": 0.386,
       "step": 78130
     },
     {
-      "epoch": 2.75,
-      "learning_rate": 2.2203011214500026e-05,
-      "loss": 0.2784,
+      "epoch": 2.815980105957401,
+      "grad_norm": 0.23062081634998322,
+      "learning_rate": 2.1129263834377838e-05,
+      "loss": 0.3886,
       "step": 78135
     },
     {
-      "epoch": 2.75,
-      "learning_rate": 2.2200180396265384e-05,
-      "loss": 0.2653,
+      "epoch": 2.8161603056186255,
+      "grad_norm": 0.20802180469036102,
+      "learning_rate": 2.1126380916251287e-05,
+      "loss": 0.4092,
       "step": 78140
     },
     {
-      "epoch": 2.75,
-      "learning_rate": 2.2197349614384543e-05,
-      "loss": 0.2874,
+      "epoch": 2.8163405052798502,
+      "grad_norm": 0.27845528721809387,
+      "learning_rate": 2.1123498050902055e-05,
+      "loss": 0.4517,
       "step": 78145
     },
     {
-      "epoch": 2.75,
-      "learning_rate": 2.2194518868894283e-05,
-      "loss": 0.2918,
+      "epoch": 2.816520704941075,
+      "grad_norm": 0.20084762573242188,
+      "learning_rate": 2.1120615238369407e-05,
+      "loss": 0.4151,
       "step": 78150
     },
     {
-      "epoch": 2.75,
-      "learning_rate": 2.2191688159831346e-05,
-      "loss": 0.2566,
+      "epoch": 2.8167009046022993,
+      "grad_norm": 0.2194603532552719,
+      "learning_rate": 2.1117732478692627e-05,
+      "loss": 0.4013,
       "step": 78155
     },
     {
-      "epoch": 2.75,
-      "learning_rate": 2.2188857487232477e-05,
-      "loss": 0.2439,
+      "epoch": 2.816881104263524,
+      "grad_norm": 0.18530698120594025,
+      "learning_rate": 2.111484977191099e-05,
+      "loss": 0.37,
       "step": 78160
     },
     {
-      "epoch": 2.75,
-      "learning_rate": 2.2186026851134445e-05,
-      "loss": 0.2777,
+      "epoch": 2.8170613039247487,
+      "grad_norm": 0.21759265661239624,
+      "learning_rate": 2.1111967118063772e-05,
+      "loss": 0.3906,
       "step": 78165
     },
     {
-      "epoch": 2.75,
-      "learning_rate": 2.2183196251574003e-05,
-      "loss": 0.2715,
+      "epoch": 2.817241503585973,
+      "grad_norm": 0.18584828078746796,
+      "learning_rate": 2.110908451719025e-05,
+      "loss": 0.3837,
       "step": 78170
     },
     {
-      "epoch": 2.75,
-      "learning_rate": 2.2180365688587906e-05,
-      "loss": 0.2752,
+      "epoch": 2.8174217032471978,
+      "grad_norm": 0.23392297327518463,
+      "learning_rate": 2.11062019693297e-05,
+      "loss": 0.4138,
       "step": 78175
     },
     {
-      "epoch": 2.75,
-      "learning_rate": 2.217753516221289e-05,
-      "loss": 0.2789,
+      "epoch": 2.8176019029084225,
+      "grad_norm": 0.21533243358135223,
+      "learning_rate": 2.110331947452139e-05,
+      "loss": 0.4038,
       "step": 78180
     },
     {
-      "epoch": 2.75,
-      "learning_rate": 2.217470467248573e-05,
-      "loss": 0.2584,
+      "epoch": 2.8177821025696472,
+      "grad_norm": 0.18144142627716064,
+      "learning_rate": 2.11004370328046e-05,
+      "loss": 0.4086,
       "step": 78185
     },
     {
-      "epoch": 2.75,
-      "learning_rate": 2.2171874219443163e-05,
-      "loss": 0.2661,
+      "epoch": 2.817962302230872,
+      "grad_norm": 0.1902519166469574,
+      "learning_rate": 2.10975546442186e-05,
+      "loss": 0.366,
       "step": 78190
     },
     {
-      "epoch": 2.75,
-      "learning_rate": 2.216904380312195e-05,
-      "loss": 0.2743,
+      "epoch": 2.8181425018920967,
+      "grad_norm": 0.259445458650589,
+      "learning_rate": 2.1094672308802666e-05,
+      "loss": 0.448,
       "step": 78195
     },
     {
-      "epoch": 2.75,
-      "learning_rate": 2.216621342355883e-05,
-      "loss": 0.2647,
+      "epoch": 2.818322701553321,
+      "grad_norm": 0.2690201997756958,
+      "learning_rate": 2.109179002659607e-05,
+      "loss": 0.393,
       "step": 78200
     },
     {
-      "epoch": 2.75,
-      "learning_rate": 2.2163383080790567e-05,
-      "loss": 0.2367,
+      "epoch": 2.8185029012145457,
+      "grad_norm": 0.17392581701278687,
+      "learning_rate": 2.108890779763806e-05,
+      "loss": 0.4098,
       "step": 78205
     },
     {
-      "epoch": 2.75,
-      "learning_rate": 2.216055277485391e-05,
-      "loss": 0.2617,
+      "epoch": 2.8186831008757705,
+      "grad_norm": 0.22515340149402618,
+      "learning_rate": 2.108602562196794e-05,
+      "loss": 0.392,
       "step": 78210
     },
     {
-      "epoch": 2.75,
-      "learning_rate": 2.2157722505785595e-05,
-      "loss": 0.2747,
+      "epoch": 2.8188633005369947,
+      "grad_norm": 0.27040407061576843,
+      "learning_rate": 2.1083143499624965e-05,
+      "loss": 0.4095,
       "step": 78215
     },
     {
-      "epoch": 2.75,
-      "learning_rate": 2.2154892273622377e-05,
-      "loss": 0.2815,
+      "epoch": 2.8190435001982195,
+      "grad_norm": 0.2060219943523407,
+      "learning_rate": 2.108026143064839e-05,
+      "loss": 0.4017,
       "step": 78220
     },
     {
-      "epoch": 2.75,
-      "learning_rate": 2.2152062078401018e-05,
-      "loss": 0.2643,
+      "epoch": 2.819223699859444,
+      "grad_norm": 0.2325987070798874,
+      "learning_rate": 2.10773794150775e-05,
+      "loss": 0.3988,
       "step": 78225
     },
     {
-      "epoch": 2.75,
-      "learning_rate": 2.2149231920158256e-05,
-      "loss": 0.2734,
+      "epoch": 2.819403899520669,
+      "grad_norm": 0.2505181133747101,
+      "learning_rate": 2.107449745295155e-05,
+      "loss": 0.3817,
       "step": 78230
     },
     {
-      "epoch": 2.75,
-      "learning_rate": 2.214640179893084e-05,
-      "loss": 0.2741,
+      "epoch": 2.8195840991818937,
+      "grad_norm": 0.1897851973772049,
+      "learning_rate": 2.107161554430982e-05,
+      "loss": 0.4017,
       "step": 78235
     },
     {
-      "epoch": 2.75,
-      "learning_rate": 2.21435717147555e-05,
-      "loss": 0.29,
+      "epoch": 2.8197642988431184,
+      "grad_norm": 0.19039909541606903,
+      "learning_rate": 2.1068733689191567e-05,
+      "loss": 0.399,
       "step": 78240
     },
     {
-      "epoch": 2.75,
-      "learning_rate": 2.214074166766901e-05,
-      "loss": 0.2804,
+      "epoch": 2.8199444985043427,
+      "grad_norm": 0.20617642998695374,
+      "learning_rate": 2.106585188763605e-05,
+      "loss": 0.4053,
       "step": 78245
     },
     {
-      "epoch": 2.75,
-      "learning_rate": 2.2137911657708107e-05,
-      "loss": 0.3018,
+      "epoch": 2.8201246981655674,
+      "grad_norm": 0.26967623829841614,
+      "learning_rate": 2.1062970139682546e-05,
+      "loss": 0.3708,
       "step": 78250
     },
     {
-      "epoch": 2.75,
-      "learning_rate": 2.2135081684909536e-05,
-      "loss": 0.2525,
+      "epoch": 2.820304897826792,
+      "grad_norm": 0.1973973661661148,
+      "learning_rate": 2.106008844537031e-05,
+      "loss": 0.4058,
       "step": 78255
     },
     {
-      "epoch": 2.75,
-      "learning_rate": 2.2132251749310032e-05,
-      "loss": 0.2621,
+      "epoch": 2.8204850974880165,
+      "grad_norm": 0.2256660908460617,
+      "learning_rate": 2.1057206804738602e-05,
+      "loss": 0.3945,
       "step": 78260
     },
     {
-      "epoch": 2.75,
-      "learning_rate": 2.2129421850946357e-05,
-      "loss": 0.2863,
+      "epoch": 2.820665297149241,
+      "grad_norm": 0.2466830313205719,
+      "learning_rate": 2.1054325217826694e-05,
+      "loss": 0.4094,
       "step": 78265
     },
     {
-      "epoch": 2.75,
-      "learning_rate": 2.2126591989855246e-05,
-      "loss": 0.2516,
+      "epoch": 2.820845496810466,
+      "grad_norm": 0.22766727209091187,
+      "learning_rate": 2.1051443684673832e-05,
+      "loss": 0.4063,
       "step": 78270
     },
     {
-      "epoch": 2.75,
-      "learning_rate": 2.2123762166073446e-05,
-      "loss": 0.2547,
+      "epoch": 2.8210256964716907,
+      "grad_norm": 0.19182877242565155,
+      "learning_rate": 2.1048562205319295e-05,
+      "loss": 0.3953,
       "step": 78275
     },
     {
-      "epoch": 2.75,
-      "learning_rate": 2.212093237963769e-05,
-      "loss": 0.2712,
+      "epoch": 2.8212058961329154,
+      "grad_norm": 0.2543238699436188,
+      "learning_rate": 2.1045680779802336e-05,
+      "loss": 0.3709,
       "step": 78280
     },
     {
-      "epoch": 2.75,
-      "learning_rate": 2.2118102630584743e-05,
-      "loss": 0.2611,
+      "epoch": 2.82138609579414,
+      "grad_norm": 0.16309164464473724,
+      "learning_rate": 2.1042799408162194e-05,
+      "loss": 0.3363,
       "step": 78285
     },
     {
-      "epoch": 2.75,
-      "learning_rate": 2.2115272918951328e-05,
-      "loss": 0.2859,
+      "epoch": 2.8215662954553644,
+      "grad_norm": 0.1783873587846756,
+      "learning_rate": 2.1039918090438156e-05,
+      "loss": 0.3677,
       "step": 78290
     },
     {
-      "epoch": 2.75,
-      "learning_rate": 2.2112443244774188e-05,
-      "loss": 0.2507,
+      "epoch": 2.821746495116589,
+      "grad_norm": 0.2764756679534912,
+      "learning_rate": 2.1037036826669463e-05,
+      "loss": 0.4055,
       "step": 78295
     },
     {
-      "epoch": 2.75,
-      "learning_rate": 2.210961360809007e-05,
-      "loss": 0.2666,
+      "epoch": 2.821926694777814,
+      "grad_norm": 0.23792728781700134,
+      "learning_rate": 2.103415561689538e-05,
+      "loss": 0.4224,
       "step": 78300
     },
     {
-      "epoch": 2.75,
-      "learning_rate": 2.2106784008935723e-05,
-      "loss": 0.2525,
+      "epoch": 2.8221068944390386,
+      "grad_norm": 0.18397817015647888,
+      "learning_rate": 2.1031274461155164e-05,
+      "loss": 0.3565,
       "step": 78305
     },
     {
-      "epoch": 2.76,
-      "learning_rate": 2.210395444734788e-05,
-      "loss": 0.2667,
+      "epoch": 2.822287094100263,
+      "grad_norm": 0.16802199184894562,
+      "learning_rate": 2.1028393359488048e-05,
+      "loss": 0.3974,
       "step": 78310
     },
     {
-      "epoch": 2.76,
-      "learning_rate": 2.2101124923363267e-05,
-      "loss": 0.2904,
+      "epoch": 2.8224672937614876,
+      "grad_norm": 0.20538485050201416,
+      "learning_rate": 2.1025512311933324e-05,
+      "loss": 0.3369,
       "step": 78315
     },
     {
-      "epoch": 2.76,
-      "learning_rate": 2.2098295437018648e-05,
-      "loss": 0.2897,
+      "epoch": 2.8226474934227124,
+      "grad_norm": 0.2252337634563446,
+      "learning_rate": 2.102263131853022e-05,
+      "loss": 0.3967,
       "step": 78320
     },
     {
-      "epoch": 2.76,
-      "learning_rate": 2.2095465988350745e-05,
-      "loss": 0.2825,
+      "epoch": 2.822827693083937,
+      "grad_norm": 0.2148447036743164,
+      "learning_rate": 2.101975037931798e-05,
+      "loss": 0.4041,
       "step": 78325
     },
     {
-      "epoch": 2.76,
-      "learning_rate": 2.2092636577396306e-05,
-      "loss": 0.29,
+      "epoch": 2.823007892745162,
+      "grad_norm": 0.24140605330467224,
+      "learning_rate": 2.1016869494335882e-05,
+      "loss": 0.365,
       "step": 78330
     },
     {
-      "epoch": 2.76,
-      "learning_rate": 2.2089807204192056e-05,
-      "loss": 0.2703,
+      "epoch": 2.823188092406386,
+      "grad_norm": 0.21468763053417206,
+      "learning_rate": 2.1013988663623165e-05,
+      "loss": 0.3998,
       "step": 78335
     },
     {
-      "epoch": 2.76,
-      "learning_rate": 2.208697786877475e-05,
-      "loss": 0.2609,
+      "epoch": 2.823368292067611,
+      "grad_norm": 0.18902160227298737,
+      "learning_rate": 2.101110788721908e-05,
+      "loss": 0.3891,
       "step": 78340
     },
     {
-      "epoch": 2.76,
-      "learning_rate": 2.208414857118112e-05,
-      "loss": 0.2636,
+      "epoch": 2.8235484917288356,
+      "grad_norm": 0.18640626966953278,
+      "learning_rate": 2.1008227165162877e-05,
+      "loss": 0.3851,
       "step": 78345
     },
     {
-      "epoch": 2.76,
-      "learning_rate": 2.208131931144789e-05,
-      "loss": 0.289,
+      "epoch": 2.8237286913900603,
+      "grad_norm": 0.21738696098327637,
+      "learning_rate": 2.1005346497493807e-05,
+      "loss": 0.4259,
       "step": 78350
     },
     {
-      "epoch": 2.76,
-      "learning_rate": 2.2078490089611808e-05,
-      "loss": 0.2759,
+      "epoch": 2.8239088910512846,
+      "grad_norm": 0.18691019713878632,
+      "learning_rate": 2.1002465884251116e-05,
+      "loss": 0.3784,
       "step": 78355
     },
     {
-      "epoch": 2.76,
-      "learning_rate": 2.2075660905709606e-05,
-      "loss": 0.2688,
+      "epoch": 2.8240890907125094,
+      "grad_norm": 0.17924901843070984,
+      "learning_rate": 2.0999585325474057e-05,
+      "loss": 0.3571,
       "step": 78360
     },
     {
-      "epoch": 2.76,
-      "learning_rate": 2.2072831759778025e-05,
-      "loss": 0.2762,
+      "epoch": 2.824269290373734,
+      "grad_norm": 0.20223569869995117,
+      "learning_rate": 2.0996704821201867e-05,
+      "loss": 0.3781,
       "step": 78365
     },
     {
-      "epoch": 2.76,
-      "learning_rate": 2.2070002651853797e-05,
-      "loss": 0.2652,
+      "epoch": 2.824449490034959,
+      "grad_norm": 0.2247609794139862,
+      "learning_rate": 2.09938243714738e-05,
+      "loss": 0.3776,
       "step": 78370
     },
     {
-      "epoch": 2.76,
-      "learning_rate": 2.206717358197364e-05,
-      "loss": 0.2634,
+      "epoch": 2.8246296896961836,
+      "grad_norm": 0.2532680034637451,
+      "learning_rate": 2.09909439763291e-05,
+      "loss": 0.4274,
       "step": 78375
     },
     {
-      "epoch": 2.76,
-      "learning_rate": 2.206434455017431e-05,
-      "loss": 0.2625,
+      "epoch": 2.824809889357408,
+      "grad_norm": 0.24293555319309235,
+      "learning_rate": 2.0988063635807022e-05,
+      "loss": 0.3768,
       "step": 78380
     },
     {
-      "epoch": 2.76,
-      "learning_rate": 2.206151555649253e-05,
-      "loss": 0.2498,
+      "epoch": 2.8249900890186326,
+      "grad_norm": 0.17587457597255707,
+      "learning_rate": 2.09851833499468e-05,
+      "loss": 0.3666,
       "step": 78385
     },
     {
-      "epoch": 2.76,
-      "learning_rate": 2.205868660096504e-05,
-      "loss": 0.2632,
+      "epoch": 2.8251702886798573,
+      "grad_norm": 0.20782296359539032,
+      "learning_rate": 2.0982303118787662e-05,
+      "loss": 0.4201,
       "step": 78390
     },
     {
-      "epoch": 2.76,
-      "learning_rate": 2.2055857683628554e-05,
-      "loss": 0.263,
+      "epoch": 2.825350488341082,
+      "grad_norm": 0.200978621840477,
+      "learning_rate": 2.0979422942368882e-05,
+      "loss": 0.3939,
       "step": 78395
     },
     {
-      "epoch": 2.76,
-      "learning_rate": 2.2053028804519825e-05,
-      "loss": 0.2532,
+      "epoch": 2.8255306880023063,
+      "grad_norm": 0.18631309270858765,
+      "learning_rate": 2.097654282072968e-05,
+      "loss": 0.4075,
       "step": 78400
     },
     {
-      "epoch": 2.76,
-      "learning_rate": 2.2050199963675573e-05,
-      "loss": 0.2502,
+      "epoch": 2.825710887663531,
+      "grad_norm": 0.22163410484790802,
+      "learning_rate": 2.09736627539093e-05,
+      "loss": 0.3522,
       "step": 78405
     },
     {
-      "epoch": 2.76,
-      "learning_rate": 2.2047371161132523e-05,
-      "loss": 0.2625,
+      "epoch": 2.825891087324756,
+      "grad_norm": 0.18656250834465027,
+      "learning_rate": 2.0970782741946987e-05,
+      "loss": 0.4021,
       "step": 78410
     },
     {
-      "epoch": 2.76,
-      "learning_rate": 2.204454239692742e-05,
-      "loss": 0.2671,
+      "epoch": 2.8260712869859805,
+      "grad_norm": 0.23782800137996674,
+      "learning_rate": 2.096790278488197e-05,
+      "loss": 0.3833,
       "step": 78415
     },
     {
-      "epoch": 2.76,
-      "learning_rate": 2.204171367109699e-05,
-      "loss": 0.258,
+      "epoch": 2.8262514866472053,
+      "grad_norm": 0.21709121763706207,
+      "learning_rate": 2.0965022882753516e-05,
+      "loss": 0.3792,
       "step": 78420
     },
     {
-      "epoch": 2.76,
-      "learning_rate": 2.203888498367796e-05,
-      "loss": 0.2777,
+      "epoch": 2.82643168630843,
+      "grad_norm": 0.22497032582759857,
+      "learning_rate": 2.0962143035600833e-05,
+      "loss": 0.4084,
       "step": 78425
     },
     {
-      "epoch": 2.76,
-      "learning_rate": 2.2036056334707038e-05,
-      "loss": 0.269,
+      "epoch": 2.8266118859696543,
+      "grad_norm": 0.2216866910457611,
+      "learning_rate": 2.095926324346317e-05,
+      "loss": 0.406,
       "step": 78430
     },
     {
-      "epoch": 2.76,
-      "learning_rate": 2.2033227724220988e-05,
-      "loss": 0.2586,
+      "epoch": 2.826792085630879,
+      "grad_norm": 0.2360861450433731,
+      "learning_rate": 2.0956383506379764e-05,
+      "loss": 0.4031,
       "step": 78435
     },
     {
-      "epoch": 2.76,
-      "learning_rate": 2.2030399152256516e-05,
-      "loss": 0.2547,
+      "epoch": 2.8269722852921038,
+      "grad_norm": 0.2680504620075226,
+      "learning_rate": 2.0953503824389853e-05,
+      "loss": 0.369,
       "step": 78440
     },
     {
-      "epoch": 2.76,
-      "learning_rate": 2.202757061885036e-05,
-      "loss": 0.2568,
+      "epoch": 2.827152484953328,
+      "grad_norm": 0.21421213448047638,
+      "learning_rate": 2.0950624197532662e-05,
+      "loss": 0.3846,
       "step": 78445
     },
     {
-      "epoch": 2.76,
-      "learning_rate": 2.2024742124039228e-05,
-      "loss": 0.2524,
+      "epoch": 2.827332684614553,
+      "grad_norm": 0.1895037740468979,
+      "learning_rate": 2.0947744625847437e-05,
+      "loss": 0.4286,
       "step": 78450
     },
     {
-      "epoch": 2.76,
-      "learning_rate": 2.202191366785987e-05,
-      "loss": 0.2505,
+      "epoch": 2.8275128842757775,
+      "grad_norm": 0.20965850353240967,
+      "learning_rate": 2.0944865109373405e-05,
+      "loss": 0.408,
       "step": 78455
     },
     {
-      "epoch": 2.76,
-      "learning_rate": 2.2019085250349e-05,
-      "loss": 0.2746,
+      "epoch": 2.8276930839370022,
+      "grad_norm": 0.21087834239006042,
+      "learning_rate": 2.0941985648149804e-05,
+      "loss": 0.3806,
       "step": 78460
     },
     {
-      "epoch": 2.76,
-      "learning_rate": 2.201625687154334e-05,
-      "loss": 0.2663,
+      "epoch": 2.827873283598227,
+      "grad_norm": 0.19764740765094757,
+      "learning_rate": 2.093910624221586e-05,
+      "loss": 0.3897,
       "step": 78465
     },
     {
-      "epoch": 2.76,
-      "learning_rate": 2.201342853147961e-05,
-      "loss": 0.2835,
+      "epoch": 2.8280534832594517,
+      "grad_norm": 0.25473564863204956,
+      "learning_rate": 2.0936226891610806e-05,
+      "loss": 0.3687,
       "step": 78470
     },
     {
-      "epoch": 2.76,
-      "learning_rate": 2.2010600230194555e-05,
-      "loss": 0.2738,
+      "epoch": 2.828233682920676,
+      "grad_norm": 0.2085646688938141,
+      "learning_rate": 2.0933347596373876e-05,
+      "loss": 0.4079,
       "step": 78475
     },
     {
-      "epoch": 2.76,
-      "learning_rate": 2.2007771967724887e-05,
-      "loss": 0.2533,
+      "epoch": 2.8284138825819007,
+      "grad_norm": 0.22956323623657227,
+      "learning_rate": 2.0930468356544294e-05,
+      "loss": 0.3709,
       "step": 78480
     },
     {
-      "epoch": 2.76,
-      "learning_rate": 2.200494374410732e-05,
-      "loss": 0.257,
+      "epoch": 2.8285940822431255,
+      "grad_norm": 0.1897376924753189,
+      "learning_rate": 2.09275891721613e-05,
+      "loss": 0.3705,
       "step": 78485
     },
     {
-      "epoch": 2.76,
-      "learning_rate": 2.2002115559378585e-05,
-      "loss": 0.2756,
+      "epoch": 2.8287742819043498,
+      "grad_norm": 0.25354906916618347,
+      "learning_rate": 2.0924710043264116e-05,
+      "loss": 0.4236,
       "step": 78490
     },
     {
-      "epoch": 2.76,
-      "learning_rate": 2.1999287413575403e-05,
-      "loss": 0.2491,
+      "epoch": 2.8289544815655745,
+      "grad_norm": 0.20143790543079376,
+      "learning_rate": 2.0921830969891955e-05,
+      "loss": 0.3469,
       "step": 78495
     },
     {
-      "epoch": 2.76,
-      "learning_rate": 2.1996459306734503e-05,
-      "loss": 0.2764,
+      "epoch": 2.8291346812267992,
+      "grad_norm": 0.22372838854789734,
+      "learning_rate": 2.0918951952084077e-05,
+      "loss": 0.3999,
       "step": 78500
     },
     {
-      "epoch": 2.76,
-      "eval_loss": 0.26042640209198,
-      "eval_runtime": 10.5472,
-      "eval_samples_per_second": 9.481,
-      "eval_steps_per_second": 9.481,
+      "epoch": 2.8291346812267992,
+      "eval_loss": 0.43108808994293213,
+      "eval_runtime": 3.5275,
+      "eval_samples_per_second": 28.348,
+      "eval_steps_per_second": 7.087,
       "step": 78500
     },
     {
-      "epoch": 2.76,
-      "learning_rate": 2.1993631238892602e-05,
-      "loss": 0.2786,
+      "epoch": 2.829314880888024,
+      "grad_norm": 0.21421097218990326,
+      "learning_rate": 2.0916072989879678e-05,
+      "loss": 0.3838,
       "step": 78505
     },
     {
-      "epoch": 2.76,
-      "learning_rate": 2.19908032100864e-05,
-      "loss": 0.2912,
+      "epoch": 2.8294950805492487,
+      "grad_norm": 0.23731063306331635,
+      "learning_rate": 2.091319408331799e-05,
+      "loss": 0.3827,
       "step": 78510
     },
     {
-      "epoch": 2.76,
-      "learning_rate": 2.1987975220352655e-05,
-      "loss": 0.2569,
+      "epoch": 2.8296752802104734,
+      "grad_norm": 0.1949160099029541,
+      "learning_rate": 2.0910315232438248e-05,
+      "loss": 0.3793,
       "step": 78515
     },
     {
-      "epoch": 2.76,
-      "learning_rate": 2.1985147269728056e-05,
-      "loss": 0.2628,
+      "epoch": 2.8298554798716977,
+      "grad_norm": 0.27850422263145447,
+      "learning_rate": 2.0907436437279662e-05,
+      "loss": 0.3917,
       "step": 78520
     },
     {
-      "epoch": 2.76,
-      "learning_rate": 2.198231935824933e-05,
-      "loss": 0.2807,
+      "epoch": 2.8300356795329225,
+      "grad_norm": 0.19794079661369324,
+      "learning_rate": 2.090455769788147e-05,
+      "loss": 0.4164,
       "step": 78525
     },
     {
-      "epoch": 2.76,
-      "learning_rate": 2.1979491485953207e-05,
-      "loss": 0.2718,
+      "epoch": 2.830215879194147,
+      "grad_norm": 0.1963154375553131,
+      "learning_rate": 2.0901679014282882e-05,
+      "loss": 0.3783,
       "step": 78530
     },
     {
-      "epoch": 2.76,
-      "learning_rate": 2.19766636528764e-05,
-      "loss": 0.2648,
+      "epoch": 2.8303960788553715,
+      "grad_norm": 0.24002398550510406,
+      "learning_rate": 2.0898800386523123e-05,
+      "loss": 0.3993,
       "step": 78535
     },
     {
-      "epoch": 2.76,
-      "learning_rate": 2.1973835859055616e-05,
-      "loss": 0.2811,
+      "epoch": 2.830576278516596,
+      "grad_norm": 0.20864902436733246,
+      "learning_rate": 2.0895921814641416e-05,
+      "loss": 0.3889,
       "step": 78540
     },
     {
-      "epoch": 2.76,
-      "learning_rate": 2.1971008104527574e-05,
-      "loss": 0.2712,
+      "epoch": 2.830756478177821,
+      "grad_norm": 0.2638551890850067,
+      "learning_rate": 2.0893043298676986e-05,
+      "loss": 0.3931,
       "step": 78545
     },
     {
-      "epoch": 2.76,
-      "learning_rate": 2.1968180389329e-05,
-      "loss": 0.2597,
+      "epoch": 2.8309366778390457,
+      "grad_norm": 0.20083938539028168,
+      "learning_rate": 2.0890164838669036e-05,
+      "loss": 0.3713,
       "step": 78550
     },
     {
-      "epoch": 2.76,
-      "learning_rate": 2.196535271349661e-05,
-      "loss": 0.2871,
+      "epoch": 2.8311168775002704,
+      "grad_norm": 0.19976069033145905,
+      "learning_rate": 2.08872864346568e-05,
+      "loss": 0.4218,
       "step": 78555
     },
     {
-      "epoch": 2.76,
-      "learning_rate": 2.1962525077067117e-05,
-      "loss": 0.2908,
+      "epoch": 2.831297077161495,
+      "grad_norm": 0.19148698449134827,
+      "learning_rate": 2.0884408086679485e-05,
+      "loss": 0.3993,
       "step": 78560
     },
     {
-      "epoch": 2.76,
-      "learning_rate": 2.195969748007722e-05,
-      "loss": 0.2651,
+      "epoch": 2.8314772768227194,
+      "grad_norm": 0.20389747619628906,
+      "learning_rate": 2.088152979477632e-05,
+      "loss": 0.3613,
       "step": 78565
     },
     {
-      "epoch": 2.76,
-      "learning_rate": 2.195686992256366e-05,
-      "loss": 0.2564,
+      "epoch": 2.831657476483944,
+      "grad_norm": 0.16990236937999725,
+      "learning_rate": 2.087865155898652e-05,
+      "loss": 0.3767,
       "step": 78570
     },
     {
-      "epoch": 2.76,
-      "learning_rate": 2.1954042404563134e-05,
-      "loss": 0.2597,
+      "epoch": 2.831837676145169,
+      "grad_norm": 0.195389986038208,
+      "learning_rate": 2.0875773379349273e-05,
+      "loss": 0.3924,
       "step": 78575
     },
     {
-      "epoch": 2.76,
-      "learning_rate": 2.1951214926112366e-05,
-      "loss": 0.2854,
+      "epoch": 2.8320178758063936,
+      "grad_norm": 0.19705142080783844,
+      "learning_rate": 2.087289525590383e-05,
+      "loss": 0.4067,
       "step": 78580
     },
     {
-      "epoch": 2.76,
-      "learning_rate": 2.1948387487248052e-05,
-      "loss": 0.2616,
+      "epoch": 2.832198075467618,
+      "grad_norm": 0.18426579236984253,
+      "learning_rate": 2.0870017188689394e-05,
+      "loss": 0.4144,
       "step": 78585
     },
     {
-      "epoch": 2.77,
-      "learning_rate": 2.1945560088006926e-05,
-      "loss": 0.2823,
+      "epoch": 2.8323782751288427,
+      "grad_norm": 0.23681855201721191,
+      "learning_rate": 2.086713917774516e-05,
+      "loss": 0.4152,
       "step": 78590
     },
     {
-      "epoch": 2.77,
-      "learning_rate": 2.1942732728425685e-05,
-      "loss": 0.261,
+      "epoch": 2.8325584747900674,
+      "grad_norm": 0.18488551676273346,
+      "learning_rate": 2.0864261223110372e-05,
+      "loss": 0.3904,
       "step": 78595
     },
     {
-      "epoch": 2.77,
-      "learning_rate": 2.1939905408541046e-05,
-      "loss": 0.2808,
+      "epoch": 2.832738674451292,
+      "grad_norm": 0.21363814175128937,
+      "learning_rate": 2.0861383324824206e-05,
+      "loss": 0.4118,
       "step": 78600
     },
     {
-      "epoch": 2.77,
-      "learning_rate": 2.193707812838971e-05,
-      "loss": 0.2677,
+      "epoch": 2.832918874112517,
+      "grad_norm": 0.24424757063388824,
+      "learning_rate": 2.0858505482925905e-05,
+      "loss": 0.3585,
       "step": 78605
     },
     {
-      "epoch": 2.77,
-      "learning_rate": 2.1934250888008408e-05,
-      "loss": 0.2687,
+      "epoch": 2.833099073773741,
+      "grad_norm": 0.20323701202869415,
+      "learning_rate": 2.0855627697454662e-05,
+      "loss": 0.3676,
       "step": 78610
     },
     {
-      "epoch": 2.77,
-      "learning_rate": 2.1931423687433835e-05,
-      "loss": 0.2635,
+      "epoch": 2.833279273434966,
+      "grad_norm": 0.1850632280111313,
+      "learning_rate": 2.0852749968449684e-05,
+      "loss": 0.3832,
       "step": 78615
     },
     {
-      "epoch": 2.77,
-      "learning_rate": 2.1928596526702706e-05,
-      "loss": 0.2494,
+      "epoch": 2.8334594730961906,
+      "grad_norm": 0.19925178587436676,
+      "learning_rate": 2.084987229595019e-05,
+      "loss": 0.3802,
       "step": 78620
     },
     {
-      "epoch": 2.77,
-      "learning_rate": 2.1925769405851715e-05,
-      "loss": 0.2584,
+      "epoch": 2.8336396727574154,
+      "grad_norm": 0.21943655610084534,
+      "learning_rate": 2.084699467999538e-05,
+      "loss": 0.3908,
       "step": 78625
     },
     {
-      "epoch": 2.77,
-      "learning_rate": 2.1922942324917588e-05,
-      "loss": 0.2525,
+      "epoch": 2.8338198724186396,
+      "grad_norm": 0.2018982470035553,
+      "learning_rate": 2.0844117120624463e-05,
+      "loss": 0.3912,
       "step": 78630
     },
     {
-      "epoch": 2.77,
-      "learning_rate": 2.1920115283937033e-05,
-      "loss": 0.2502,
+      "epoch": 2.8340000720798644,
+      "grad_norm": 0.20863167941570282,
+      "learning_rate": 2.0841239617876647e-05,
+      "loss": 0.4028,
       "step": 78635
     },
     {
-      "epoch": 2.77,
-      "learning_rate": 2.191728828294674e-05,
-      "loss": 0.2859,
+      "epoch": 2.834180271741089,
+      "grad_norm": 0.19766663014888763,
+      "learning_rate": 2.0838362171791133e-05,
+      "loss": 0.3733,
       "step": 78640
     },
     {
-      "epoch": 2.77,
-      "learning_rate": 2.1914461321983436e-05,
-      "loss": 0.2472,
+      "epoch": 2.834360471402314,
+      "grad_norm": 0.24761301279067993,
+      "learning_rate": 2.083548478240713e-05,
+      "loss": 0.3987,
       "step": 78645
     },
     {
-      "epoch": 2.77,
-      "learning_rate": 2.191163440108382e-05,
-      "loss": 0.2693,
+      "epoch": 2.8345406710635386,
+      "grad_norm": 0.22107259929180145,
+      "learning_rate": 2.0832607449763843e-05,
+      "loss": 0.4007,
       "step": 78650
     },
     {
-      "epoch": 2.77,
-      "learning_rate": 2.190880752028459e-05,
-      "loss": 0.2925,
+      "epoch": 2.8347208707247633,
+      "grad_norm": 0.2298266589641571,
+      "learning_rate": 2.082973017390047e-05,
+      "loss": 0.3774,
       "step": 78655
     },
     {
-      "epoch": 2.77,
-      "learning_rate": 2.190598067962245e-05,
-      "loss": 0.2807,
+      "epoch": 2.8349010703859876,
+      "grad_norm": 0.2100599855184555,
+      "learning_rate": 2.0826852954856217e-05,
+      "loss": 0.3824,
       "step": 78660
     },
     {
-      "epoch": 2.77,
-      "learning_rate": 2.1903153879134126e-05,
-      "loss": 0.277,
+      "epoch": 2.8350812700472123,
+      "grad_norm": 0.19169162213802338,
+      "learning_rate": 2.0823975792670292e-05,
+      "loss": 0.3735,
       "step": 78665
     },
     {
-      "epoch": 2.77,
-      "learning_rate": 2.1900327118856305e-05,
-      "loss": 0.2835,
+      "epoch": 2.835261469708437,
+      "grad_norm": 0.2121647298336029,
+      "learning_rate": 2.0821098687381874e-05,
+      "loss": 0.4001,
       "step": 78670
     },
     {
-      "epoch": 2.77,
-      "learning_rate": 2.1897500398825695e-05,
-      "loss": 0.2835,
+      "epoch": 2.8354416693696614,
+      "grad_norm": 0.2263890504837036,
+      "learning_rate": 2.081822163903019e-05,
+      "loss": 0.4074,
       "step": 78675
     },
     {
-      "epoch": 2.77,
-      "learning_rate": 2.189467371907899e-05,
-      "loss": 0.2598,
+      "epoch": 2.835621869030886,
+      "grad_norm": 0.2616461515426636,
+      "learning_rate": 2.0815344647654413e-05,
+      "loss": 0.4573,
       "step": 78680
     },
     {
-      "epoch": 2.77,
-      "learning_rate": 2.1891847079652903e-05,
-      "loss": 0.2599,
+      "epoch": 2.835802068692111,
+      "grad_norm": 0.2435300052165985,
+      "learning_rate": 2.081246771329377e-05,
+      "loss": 0.3558,
       "step": 78685
     },
     {
-      "epoch": 2.77,
-      "learning_rate": 2.1889020480584137e-05,
-      "loss": 0.2563,
+      "epoch": 2.8359822683533356,
+      "grad_norm": 0.20610463619232178,
+      "learning_rate": 2.080959083598744e-05,
+      "loss": 0.4027,
       "step": 78690
     },
     {
-      "epoch": 2.77,
-      "learning_rate": 2.1886193921909394e-05,
-      "loss": 0.2537,
+      "epoch": 2.8361624680145603,
+      "grad_norm": 0.23141418397426605,
+      "learning_rate": 2.0806714015774613e-05,
+      "loss": 0.3734,
       "step": 78695
     },
     {
-      "epoch": 2.77,
-      "learning_rate": 2.1883367403665356e-05,
-      "loss": 0.2813,
+      "epoch": 2.836342667675785,
+      "grad_norm": 0.1831231713294983,
+      "learning_rate": 2.080383725269451e-05,
+      "loss": 0.4006,
       "step": 78700
     },
     {
-      "epoch": 2.77,
-      "learning_rate": 2.188054092588875e-05,
-      "loss": 0.2766,
+      "epoch": 2.8365228673370093,
+      "grad_norm": 0.23543448746204376,
+      "learning_rate": 2.0800960546786293e-05,
+      "loss": 0.4156,
       "step": 78705
     },
     {
-      "epoch": 2.77,
-      "learning_rate": 2.1877714488616262e-05,
-      "loss": 0.2852,
+      "epoch": 2.836703066998234,
+      "grad_norm": 0.1677248328924179,
+      "learning_rate": 2.0798083898089193e-05,
+      "loss": 0.3878,
       "step": 78710
     },
     {
-      "epoch": 2.77,
-      "learning_rate": 2.1874888091884593e-05,
-      "loss": 0.2666,
+      "epoch": 2.836883266659459,
+      "grad_norm": 0.20878414809703827,
+      "learning_rate": 2.0795207306642383e-05,
+      "loss": 0.3831,
       "step": 78715
     },
     {
-      "epoch": 2.77,
-      "learning_rate": 2.1872061735730436e-05,
-      "loss": 0.2682,
+      "epoch": 2.837063466320683,
+      "grad_norm": 0.24195106327533722,
+      "learning_rate": 2.0792330772485055e-05,
+      "loss": 0.4067,
       "step": 78720
     },
     {
-      "epoch": 2.77,
-      "learning_rate": 2.186923542019051e-05,
-      "loss": 0.2746,
+      "epoch": 2.837243665981908,
+      "grad_norm": 0.19112065434455872,
+      "learning_rate": 2.078945429565641e-05,
+      "loss": 0.4271,
       "step": 78725
     },
     {
-      "epoch": 2.77,
-      "learning_rate": 2.186640914530149e-05,
-      "loss": 0.2646,
+      "epoch": 2.8374238656431325,
+      "grad_norm": 0.2498737871646881,
+      "learning_rate": 2.0786577876195633e-05,
+      "loss": 0.3881,
       "step": 78730
     },
     {
-      "epoch": 2.77,
-      "learning_rate": 2.1863582911100083e-05,
-      "loss": 0.2447,
+      "epoch": 2.8376040653043573,
+      "grad_norm": 0.1864451766014099,
+      "learning_rate": 2.0783701514141916e-05,
+      "loss": 0.3809,
       "step": 78735
     },
     {
-      "epoch": 2.77,
-      "learning_rate": 2.1860756717622978e-05,
-      "loss": 0.2704,
+      "epoch": 2.837784264965582,
+      "grad_norm": 0.24682430922985077,
+      "learning_rate": 2.0780825209534448e-05,
+      "loss": 0.3981,
       "step": 78740
     },
     {
-      "epoch": 2.77,
-      "learning_rate": 2.185793056490689e-05,
-      "loss": 0.27,
+      "epoch": 2.8379644646268067,
+      "grad_norm": 0.2143888771533966,
+      "learning_rate": 2.077794896241242e-05,
+      "loss": 0.3755,
       "step": 78745
     },
     {
-      "epoch": 2.77,
-      "learning_rate": 2.1855104452988505e-05,
-      "loss": 0.2776,
+      "epoch": 2.838144664288031,
+      "grad_norm": 0.2417534738779068,
+      "learning_rate": 2.0775072772815023e-05,
+      "loss": 0.4129,
       "step": 78750
     },
     {
-      "epoch": 2.77,
-      "learning_rate": 2.18522783819045e-05,
-      "loss": 0.2765,
+      "epoch": 2.8383248639492558,
+      "grad_norm": 0.2023160457611084,
+      "learning_rate": 2.0772196640781444e-05,
+      "loss": 0.3503,
       "step": 78755
     },
     {
-      "epoch": 2.77,
-      "learning_rate": 2.18494523516916e-05,
-      "loss": 0.2562,
+      "epoch": 2.8385050636104805,
+      "grad_norm": 0.26836487650871277,
+      "learning_rate": 2.076932056635086e-05,
+      "loss": 0.433,
       "step": 78760
     },
     {
-      "epoch": 2.77,
-      "learning_rate": 2.1846626362386478e-05,
-      "loss": 0.3022,
+      "epoch": 2.838685263271705,
+      "grad_norm": 0.2009926289319992,
+      "learning_rate": 2.076644454956247e-05,
+      "loss": 0.452,
       "step": 78765
     },
     {
-      "epoch": 2.77,
-      "learning_rate": 2.184380041402584e-05,
-      "loss": 0.2736,
+      "epoch": 2.8388654629329295,
+      "grad_norm": 0.20589284598827362,
+      "learning_rate": 2.0763568590455458e-05,
+      "loss": 0.3995,
       "step": 78770
     },
     {
-      "epoch": 2.77,
-      "learning_rate": 2.1840974506646363e-05,
-      "loss": 0.2568,
+      "epoch": 2.8390456625941543,
+      "grad_norm": 0.23348768055438995,
+      "learning_rate": 2.0760692689068988e-05,
+      "loss": 0.3783,
       "step": 78775
     },
     {
-      "epoch": 2.77,
-      "learning_rate": 2.1838148640284763e-05,
-      "loss": 0.2483,
+      "epoch": 2.839225862255379,
+      "grad_norm": 0.2545948028564453,
+      "learning_rate": 2.0757816845442274e-05,
+      "loss": 0.423,
       "step": 78780
     },
     {
-      "epoch": 2.77,
-      "learning_rate": 2.1835322814977718e-05,
-      "loss": 0.2659,
+      "epoch": 2.8394060619166037,
+      "grad_norm": 0.19799551367759705,
+      "learning_rate": 2.075494105961447e-05,
+      "loss": 0.417,
       "step": 78785
     },
     {
-      "epoch": 2.77,
-      "learning_rate": 2.1832497030761916e-05,
-      "loss": 0.2842,
+      "epoch": 2.8395862615778285,
+      "grad_norm": 0.22839927673339844,
+      "learning_rate": 2.0752065331624788e-05,
+      "loss": 0.3828,
       "step": 78790
     },
     {
-      "epoch": 2.77,
-      "learning_rate": 2.182967128767405e-05,
-      "loss": 0.2506,
+      "epoch": 2.8397664612390527,
+      "grad_norm": 0.1921931505203247,
+      "learning_rate": 2.0749189661512387e-05,
+      "loss": 0.4218,
       "step": 78795
     },
     {
-      "epoch": 2.77,
-      "learning_rate": 2.1826845585750817e-05,
-      "loss": 0.2736,
+      "epoch": 2.8399466609002775,
+      "grad_norm": 0.20182907581329346,
+      "learning_rate": 2.074631404931645e-05,
+      "loss": 0.4186,
       "step": 78800
     },
     {
-      "epoch": 2.77,
-      "learning_rate": 2.1824019925028905e-05,
-      "loss": 0.2783,
+      "epoch": 2.840126860561502,
+      "grad_norm": 0.20706386864185333,
+      "learning_rate": 2.0743438495076164e-05,
+      "loss": 0.394,
       "step": 78805
     },
     {
-      "epoch": 2.77,
-      "learning_rate": 2.1821194305545002e-05,
-      "loss": 0.2795,
+      "epoch": 2.840307060222727,
+      "grad_norm": 0.23786523938179016,
+      "learning_rate": 2.0740562998830706e-05,
+      "loss": 0.3553,
       "step": 78810
     },
     {
-      "epoch": 2.77,
-      "learning_rate": 2.1818368727335784e-05,
-      "loss": 0.2575,
+      "epoch": 2.8404872598839512,
+      "grad_norm": 0.1973177045583725,
+      "learning_rate": 2.073768756061925e-05,
+      "loss": 0.3815,
       "step": 78815
     },
     {
-      "epoch": 2.77,
-      "learning_rate": 2.1815543190437963e-05,
-      "loss": 0.2811,
+      "epoch": 2.840667459545176,
+      "grad_norm": 0.21319210529327393,
+      "learning_rate": 2.0734812180480976e-05,
+      "loss": 0.3736,
       "step": 78820
     },
     {
-      "epoch": 2.77,
-      "learning_rate": 2.1812717694888207e-05,
-      "loss": 0.2559,
+      "epoch": 2.8408476592064007,
+      "grad_norm": 0.21359285712242126,
+      "learning_rate": 2.0731936858455057e-05,
+      "loss": 0.358,
       "step": 78825
     },
     {
-      "epoch": 2.77,
-      "learning_rate": 2.1809892240723224e-05,
-      "loss": 0.2639,
+      "epoch": 2.8410278588676254,
+      "grad_norm": 0.24287675321102142,
+      "learning_rate": 2.0729061594580677e-05,
+      "loss": 0.4399,
       "step": 78830
     },
     {
-      "epoch": 2.77,
-      "learning_rate": 2.180706682797967e-05,
-      "loss": 0.2542,
+      "epoch": 2.84120805852885,
+      "grad_norm": 0.2060524970293045,
+      "learning_rate": 2.0726186388897007e-05,
+      "loss": 0.3973,
       "step": 78835
     },
     {
-      "epoch": 2.77,
-      "learning_rate": 2.1804241456694262e-05,
-      "loss": 0.2483,
+      "epoch": 2.8413882581900745,
+      "grad_norm": 0.21056078374385834,
+      "learning_rate": 2.072331124144321e-05,
+      "loss": 0.3884,
       "step": 78840
     },
     {
-      "epoch": 2.77,
-      "learning_rate": 2.1801416126903667e-05,
-      "loss": 0.2842,
+      "epoch": 2.841568457851299,
+      "grad_norm": 0.22964583337306976,
+      "learning_rate": 2.0720436152258483e-05,
+      "loss": 0.3679,
       "step": 78845
     },
     {
-      "epoch": 2.77,
-      "learning_rate": 2.1798590838644584e-05,
-      "loss": 0.2807,
+      "epoch": 2.841748657512524,
+      "grad_norm": 0.1646958887577057,
+      "learning_rate": 2.0717561121381983e-05,
+      "loss": 0.3748,
       "step": 78850
     },
     {
-      "epoch": 2.77,
-      "learning_rate": 2.1795765591953678e-05,
-      "loss": 0.2798,
+      "epoch": 2.8419288571737487,
+      "grad_norm": 0.2664015591144562,
+      "learning_rate": 2.0714686148852873e-05,
+      "loss": 0.4059,
       "step": 78855
     },
     {
-      "epoch": 2.77,
-      "learning_rate": 2.179294038686765e-05,
-      "loss": 0.2552,
+      "epoch": 2.842109056834973,
+      "grad_norm": 0.22343656420707703,
+      "learning_rate": 2.0711811234710347e-05,
+      "loss": 0.4021,
       "step": 78860
     },
     {
-      "epoch": 2.77,
-      "learning_rate": 2.179011522342318e-05,
-      "loss": 0.268,
+      "epoch": 2.8422892564961977,
+      "grad_norm": 0.2661823332309723,
+      "learning_rate": 2.0708936378993545e-05,
+      "loss": 0.3696,
       "step": 78865
     },
     {
-      "epoch": 2.77,
-      "learning_rate": 2.1787290101656937e-05,
-      "loss": 0.2761,
+      "epoch": 2.8424694561574224,
+      "grad_norm": 0.25233519077301025,
+      "learning_rate": 2.0706061581741667e-05,
+      "loss": 0.4061,
       "step": 78870
     },
     {
-      "epoch": 2.78,
-      "learning_rate": 2.1784465021605627e-05,
-      "loss": 0.276,
+      "epoch": 2.842649655818647,
+      "grad_norm": 0.19629564881324768,
+      "learning_rate": 2.0703186842993878e-05,
+      "loss": 0.4031,
       "step": 78875
     },
     {
-      "epoch": 2.78,
-      "learning_rate": 2.1781639983305917e-05,
-      "loss": 0.2701,
+      "epoch": 2.842829855479872,
+      "grad_norm": 0.17642425000667572,
+      "learning_rate": 2.0700312162789316e-05,
+      "loss": 0.4057,
       "step": 78880
     },
     {
-      "epoch": 2.78,
-      "learning_rate": 2.1778814986794492e-05,
-      "loss": 0.2942,
+      "epoch": 2.843010055141096,
+      "grad_norm": 0.16491912305355072,
+      "learning_rate": 2.0697437541167182e-05,
+      "loss": 0.3817,
       "step": 78885
     },
     {
-      "epoch": 2.78,
-      "learning_rate": 2.177599003210802e-05,
-      "loss": 0.275,
+      "epoch": 2.843190254802321,
+      "grad_norm": 0.18968559801578522,
+      "learning_rate": 2.0694562978166617e-05,
+      "loss": 0.4262,
       "step": 78890
     },
     {
-      "epoch": 2.78,
-      "learning_rate": 2.1773165119283206e-05,
-      "loss": 0.2897,
+      "epoch": 2.8433704544635456,
+      "grad_norm": 0.23654945194721222,
+      "learning_rate": 2.0691688473826813e-05,
+      "loss": 0.4081,
       "step": 78895
     },
     {
-      "epoch": 2.78,
-      "learning_rate": 2.1770340248356714e-05,
-      "loss": 0.285,
+      "epoch": 2.8435506541247704,
+      "grad_norm": 0.21302512288093567,
+      "learning_rate": 2.068881402818691e-05,
+      "loss": 0.3762,
       "step": 78900
     },
     {
-      "epoch": 2.78,
-      "learning_rate": 2.176751541936522e-05,
-      "loss": 0.2668,
+      "epoch": 2.8437308537859947,
+      "grad_norm": 0.19468043744564056,
+      "learning_rate": 2.068593964128608e-05,
+      "loss": 0.4304,
       "step": 78905
     },
     {
-      "epoch": 2.78,
-      "learning_rate": 2.1764690632345407e-05,
-      "loss": 0.2558,
+      "epoch": 2.8439110534472194,
+      "grad_norm": 0.20956778526306152,
+      "learning_rate": 2.0683065313163493e-05,
+      "loss": 0.3958,
       "step": 78910
     },
     {
-      "epoch": 2.78,
-      "learning_rate": 2.1761865887333962e-05,
-      "loss": 0.2576,
+      "epoch": 2.844091253108444,
+      "grad_norm": 0.2271938920021057,
+      "learning_rate": 2.0680191043858303e-05,
+      "loss": 0.4014,
       "step": 78915
     },
     {
-      "epoch": 2.78,
-      "learning_rate": 2.1759041184367558e-05,
-      "loss": 0.29,
+      "epoch": 2.844271452769669,
+      "grad_norm": 0.2164512425661087,
+      "learning_rate": 2.0677316833409672e-05,
+      "loss": 0.4368,
       "step": 78920
     },
     {
-      "epoch": 2.78,
-      "learning_rate": 2.175621652348286e-05,
-      "loss": 0.2788,
+      "epoch": 2.8444516524308936,
+      "grad_norm": 0.2096407115459442,
+      "learning_rate": 2.0674442681856764e-05,
+      "loss": 0.3887,
       "step": 78925
     },
     {
-      "epoch": 2.78,
-      "learning_rate": 2.175339190471655e-05,
-      "loss": 0.2555,
+      "epoch": 2.8446318520921183,
+      "grad_norm": 0.199036106467247,
+      "learning_rate": 2.0671568589238734e-05,
+      "loss": 0.389,
       "step": 78930
     },
     {
-      "epoch": 2.78,
-      "learning_rate": 2.1750567328105308e-05,
-      "loss": 0.2706,
+      "epoch": 2.8448120517533426,
+      "grad_norm": 0.20308394730091095,
+      "learning_rate": 2.0668694555594746e-05,
+      "loss": 0.3894,
       "step": 78935
     },
     {
-      "epoch": 2.78,
-      "learning_rate": 2.1747742793685816e-05,
-      "loss": 0.2622,
+      "epoch": 2.8449922514145674,
+      "grad_norm": 0.2252362221479416,
+      "learning_rate": 2.0665820580963957e-05,
+      "loss": 0.3844,
       "step": 78940
     },
     {
-      "epoch": 2.78,
-      "learning_rate": 2.174491830149474e-05,
-      "loss": 0.2678,
+      "epoch": 2.845172451075792,
+      "grad_norm": 0.18750686943531036,
+      "learning_rate": 2.066294666538552e-05,
+      "loss": 0.4232,
       "step": 78945
     },
     {
-      "epoch": 2.78,
-      "learning_rate": 2.174209385156874e-05,
-      "loss": 0.2656,
+      "epoch": 2.8453526507370164,
+      "grad_norm": 0.20952045917510986,
+      "learning_rate": 2.06600728088986e-05,
+      "loss": 0.4092,
       "step": 78950
     },
     {
-      "epoch": 2.78,
-      "learning_rate": 2.1739269443944515e-05,
-      "loss": 0.2589,
+      "epoch": 2.845532850398241,
+      "grad_norm": 0.24399419128894806,
+      "learning_rate": 2.0657199011542352e-05,
+      "loss": 0.4103,
       "step": 78955
     },
     {
-      "epoch": 2.78,
-      "learning_rate": 2.1736445078658726e-05,
-      "loss": 0.2621,
+      "epoch": 2.845713050059466,
+      "grad_norm": 0.20559486746788025,
+      "learning_rate": 2.065432527335591e-05,
+      "loss": 0.3815,
       "step": 78960
     },
     {
-      "epoch": 2.78,
-      "learning_rate": 2.1733620755748054e-05,
-      "loss": 0.2753,
+      "epoch": 2.8458932497206906,
+      "grad_norm": 0.24112515151500702,
+      "learning_rate": 2.0651451594378462e-05,
+      "loss": 0.4345,
       "step": 78965
     },
     {
-      "epoch": 2.78,
-      "learning_rate": 2.1730796475249148e-05,
-      "loss": 0.2798,
+      "epoch": 2.8460734493819153,
+      "grad_norm": 0.24008318781852722,
+      "learning_rate": 2.064857797464913e-05,
+      "loss": 0.4203,
       "step": 78970
     },
     {
-      "epoch": 2.78,
-      "learning_rate": 2.1727972237198708e-05,
-      "loss": 0.274,
+      "epoch": 2.84625364904314,
+      "grad_norm": 0.1976160705089569,
+      "learning_rate": 2.0645704414207096e-05,
+      "loss": 0.3979,
       "step": 78975
     },
     {
-      "epoch": 2.78,
-      "learning_rate": 2.1725148041633388e-05,
-      "loss": 0.2559,
+      "epoch": 2.8464338487043643,
+      "grad_norm": 0.21798163652420044,
+      "learning_rate": 2.064283091309149e-05,
+      "loss": 0.3961,
       "step": 78980
     },
     {
-      "epoch": 2.78,
-      "learning_rate": 2.1722323888589857e-05,
-      "loss": 0.279,
+      "epoch": 2.846614048365589,
+      "grad_norm": 0.30236032605171204,
+      "learning_rate": 2.0639957471341463e-05,
+      "loss": 0.3842,
       "step": 78985
     },
     {
-      "epoch": 2.78,
-      "learning_rate": 2.1719499778104796e-05,
-      "loss": 0.2681,
+      "epoch": 2.846794248026814,
+      "grad_norm": 0.2003178596496582,
+      "learning_rate": 2.0637084088996175e-05,
+      "loss": 0.4055,
       "step": 78990
     },
     {
-      "epoch": 2.78,
-      "learning_rate": 2.171667571021487e-05,
-      "loss": 0.2853,
+      "epoch": 2.846974447688038,
+      "grad_norm": 0.18531939387321472,
+      "learning_rate": 2.0634210766094775e-05,
+      "loss": 0.3818,
       "step": 78995
     },
     {
-      "epoch": 2.78,
-      "learning_rate": 2.171385168495675e-05,
-      "loss": 0.258,
+      "epoch": 2.847154647349263,
+      "grad_norm": 0.16672350466251373,
+      "learning_rate": 2.06313375026764e-05,
+      "loss": 0.3738,
       "step": 79000
     },
     {
-      "epoch": 2.78,
-      "eval_loss": 0.260810911655426,
-      "eval_runtime": 10.5622,
-      "eval_samples_per_second": 9.468,
-      "eval_steps_per_second": 9.468,
+      "epoch": 2.847154647349263,
+      "eval_loss": 0.4311943054199219,
+      "eval_runtime": 3.5347,
+      "eval_samples_per_second": 28.291,
+      "eval_steps_per_second": 7.073,
       "step": 79000
     },
     {
-      "epoch": 2.78,
-      "learning_rate": 2.1711027702367086e-05,
-      "loss": 0.2571,
+      "epoch": 2.8473348470104876,
+      "grad_norm": 0.1916167140007019,
+      "learning_rate": 2.0628464298780215e-05,
+      "loss": 0.39,
       "step": 79005
     },
     {
-      "epoch": 2.78,
-      "learning_rate": 2.1708203762482574e-05,
-      "loss": 0.2718,
+      "epoch": 2.8475150466717123,
+      "grad_norm": 0.4534687399864197,
+      "learning_rate": 2.0625591154445348e-05,
+      "loss": 0.3945,
       "step": 79010
     },
     {
-      "epoch": 2.78,
-      "learning_rate": 2.170537986533986e-05,
-      "loss": 0.2733,
+      "epoch": 2.847695246332937,
+      "grad_norm": 0.19475337862968445,
+      "learning_rate": 2.062271806971096e-05,
+      "loss": 0.4248,
       "step": 79015
     },
     {
-      "epoch": 2.78,
-      "learning_rate": 2.1702556010975623e-05,
-      "loss": 0.2939,
+      "epoch": 2.8478754459941618,
+      "grad_norm": 0.20240464806556702,
+      "learning_rate": 2.0619845044616195e-05,
+      "loss": 0.4073,
       "step": 79020
     },
     {
-      "epoch": 2.78,
-      "learning_rate": 2.169973219942651e-05,
-      "loss": 0.2649,
+      "epoch": 2.848055645655386,
+      "grad_norm": 0.21959447860717773,
+      "learning_rate": 2.0616972079200185e-05,
+      "loss": 0.4052,
       "step": 79025
     },
     {
-      "epoch": 2.78,
-      "learning_rate": 2.1696908430729214e-05,
-      "loss": 0.2827,
+      "epoch": 2.848235845316611,
+      "grad_norm": 0.21866481006145477,
+      "learning_rate": 2.061409917350209e-05,
+      "loss": 0.4036,
       "step": 79030
     },
     {
-      "epoch": 2.78,
-      "learning_rate": 2.169408470492038e-05,
-      "loss": 0.286,
+      "epoch": 2.8484160449778355,
+      "grad_norm": 0.19479642808437347,
+      "learning_rate": 2.0611226327561042e-05,
+      "loss": 0.3468,
       "step": 79035
     },
     {
-      "epoch": 2.78,
-      "learning_rate": 2.1691261022036674e-05,
-      "loss": 0.2421,
+      "epoch": 2.84859624463906,
+      "grad_norm": 0.24999023973941803,
+      "learning_rate": 2.060835354141618e-05,
+      "loss": 0.3691,
       "step": 79040
     },
     {
-      "epoch": 2.78,
-      "learning_rate": 2.168843738211476e-05,
-      "loss": 0.2687,
+      "epoch": 2.8487764443002845,
+      "grad_norm": 0.23853568732738495,
+      "learning_rate": 2.0605480815106656e-05,
+      "loss": 0.4224,
       "step": 79045
     },
     {
-      "epoch": 2.78,
-      "learning_rate": 2.1685613785191312e-05,
-      "loss": 0.2901,
+      "epoch": 2.8489566439615093,
+      "grad_norm": 0.2416432797908783,
+      "learning_rate": 2.0602608148671602e-05,
+      "loss": 0.3812,
       "step": 79050
     },
     {
-      "epoch": 2.78,
-      "learning_rate": 2.1682790231302988e-05,
-      "loss": 0.2639,
+      "epoch": 2.849136843622734,
+      "grad_norm": 0.18531100451946259,
+      "learning_rate": 2.0599735542150164e-05,
+      "loss": 0.4103,
       "step": 79055
     },
     {
-      "epoch": 2.78,
-      "learning_rate": 2.1679966720486445e-05,
-      "loss": 0.2533,
+      "epoch": 2.8493170432839587,
+      "grad_norm": 0.1804402768611908,
+      "learning_rate": 2.0596862995581485e-05,
+      "loss": 0.3648,
       "step": 79060
     },
     {
-      "epoch": 2.78,
-      "learning_rate": 2.1677143252778335e-05,
-      "loss": 0.2734,
+      "epoch": 2.8494972429451835,
+      "grad_norm": 0.21126359701156616,
+      "learning_rate": 2.0593990509004675e-05,
+      "loss": 0.3624,
       "step": 79065
     },
     {
-      "epoch": 2.78,
-      "learning_rate": 2.1674319828215337e-05,
-      "loss": 0.2918,
+      "epoch": 2.8496774426064078,
+      "grad_norm": 0.23747804760932922,
+      "learning_rate": 2.059111808245891e-05,
+      "loss": 0.4037,
       "step": 79070
     },
     {
-      "epoch": 2.78,
-      "learning_rate": 2.1671496446834108e-05,
-      "loss": 0.2567,
+      "epoch": 2.8498576422676325,
+      "grad_norm": 0.2273935079574585,
+      "learning_rate": 2.058824571598329e-05,
+      "loss": 0.4083,
       "step": 79075
     },
     {
-      "epoch": 2.78,
-      "learning_rate": 2.1668673108671306e-05,
-      "loss": 0.2541,
+      "epoch": 2.8500378419288572,
+      "grad_norm": 0.22935499250888824,
+      "learning_rate": 2.0585373409616985e-05,
+      "loss": 0.3983,
       "step": 79080
     },
     {
-      "epoch": 2.78,
-      "learning_rate": 2.1665849813763576e-05,
-      "loss": 0.2684,
+      "epoch": 2.850218041590082,
+      "grad_norm": 0.28681254386901855,
+      "learning_rate": 2.058250116339911e-05,
+      "loss": 0.39,
       "step": 79085
     },
     {
-      "epoch": 2.78,
-      "learning_rate": 2.16630265621476e-05,
-      "loss": 0.2823,
+      "epoch": 2.8503982412513063,
+      "grad_norm": 0.18276835978031158,
+      "learning_rate": 2.0579628977368792e-05,
+      "loss": 0.3573,
       "step": 79090
     },
     {
-      "epoch": 2.78,
-      "learning_rate": 2.166020335386002e-05,
-      "loss": 0.2759,
+      "epoch": 2.850578440912531,
+      "grad_norm": 0.19169297814369202,
+      "learning_rate": 2.0576756851565182e-05,
+      "loss": 0.4129,
       "step": 79095
     },
     {
-      "epoch": 2.78,
-      "learning_rate": 2.1657380188937494e-05,
-      "loss": 0.284,
+      "epoch": 2.8507586405737557,
+      "grad_norm": 0.2185550034046173,
+      "learning_rate": 2.05738847860274e-05,
+      "loss": 0.4006,
       "step": 79100
     },
     {
-      "epoch": 2.78,
-      "learning_rate": 2.165455706741669e-05,
-      "loss": 0.2545,
+      "epoch": 2.8509388402349805,
+      "grad_norm": 0.24736441671848297,
+      "learning_rate": 2.0571012780794577e-05,
+      "loss": 0.3866,
       "step": 79105
     },
     {
-      "epoch": 2.78,
-      "learning_rate": 2.165173398933426e-05,
-      "loss": 0.2777,
+      "epoch": 2.851119039896205,
+      "grad_norm": 0.19851014018058777,
+      "learning_rate": 2.056814083590585e-05,
+      "loss": 0.4227,
       "step": 79110
     },
     {
-      "epoch": 2.78,
-      "learning_rate": 2.164891095472686e-05,
-      "loss": 0.2842,
+      "epoch": 2.8512992395574295,
+      "grad_norm": 0.19932295382022858,
+      "learning_rate": 2.0565268951400346e-05,
+      "loss": 0.3941,
       "step": 79115
     },
     {
-      "epoch": 2.78,
-      "learning_rate": 2.164608796363113e-05,
-      "loss": 0.2896,
+      "epoch": 2.851479439218654,
+      "grad_norm": 0.21272258460521698,
+      "learning_rate": 2.0562397127317197e-05,
+      "loss": 0.37,
       "step": 79120
     },
     {
-      "epoch": 2.78,
-      "learning_rate": 2.164326501608374e-05,
-      "loss": 0.2724,
+      "epoch": 2.851659638879879,
+      "grad_norm": 0.2207152098417282,
+      "learning_rate": 2.055952536369553e-05,
+      "loss": 0.3985,
       "step": 79125
     },
     {
-      "epoch": 2.78,
-      "learning_rate": 2.164044211212135e-05,
-      "loss": 0.2505,
+      "epoch": 2.8518398385411037,
+      "grad_norm": 0.2772730588912964,
+      "learning_rate": 2.0556653660574464e-05,
+      "loss": 0.385,
       "step": 79130
     },
     {
-      "epoch": 2.78,
-      "learning_rate": 2.1637619251780604e-05,
-      "loss": 0.2728,
+      "epoch": 2.852020038202328,
+      "grad_norm": 0.2133176475763321,
+      "learning_rate": 2.0553782017993135e-05,
+      "loss": 0.3741,
       "step": 79135
     },
     {
-      "epoch": 2.78,
-      "learning_rate": 2.163479643509815e-05,
-      "loss": 0.2859,
+      "epoch": 2.8522002378635527,
+      "grad_norm": 0.266476571559906,
+      "learning_rate": 2.055091043599067e-05,
+      "loss": 0.4071,
       "step": 79140
     },
     {
-      "epoch": 2.78,
-      "learning_rate": 2.163197366211065e-05,
-      "loss": 0.2734,
+      "epoch": 2.8523804375247774,
+      "grad_norm": 0.20669589936733246,
+      "learning_rate": 2.0548038914606174e-05,
+      "loss": 0.417,
       "step": 79145
     },
     {
-      "epoch": 2.78,
-      "learning_rate": 2.1629150932854752e-05,
-      "loss": 0.2739,
+      "epoch": 2.852560637186002,
+      "grad_norm": 0.21725420653820038,
+      "learning_rate": 2.0545167453878804e-05,
+      "loss": 0.3852,
       "step": 79150
     },
     {
-      "epoch": 2.78,
-      "learning_rate": 2.1626328247367112e-05,
-      "loss": 0.2747,
+      "epoch": 2.852740836847227,
+      "grad_norm": 0.23247067630290985,
+      "learning_rate": 2.0542296053847647e-05,
+      "loss": 0.399,
       "step": 79155
     },
     {
-      "epoch": 2.79,
-      "learning_rate": 2.1623505605684364e-05,
-      "loss": 0.2626,
+      "epoch": 2.8529210365084516,
+      "grad_norm": 0.2411661595106125,
+      "learning_rate": 2.0539424714551852e-05,
+      "loss": 0.4155,
       "step": 79160
     },
     {
-      "epoch": 2.79,
-      "learning_rate": 2.1620683007843184e-05,
-      "loss": 0.2693,
+      "epoch": 2.853101236169676,
+      "grad_norm": 0.2001282423734665,
+      "learning_rate": 2.053655343603054e-05,
+      "loss": 0.4257,
       "step": 79165
     },
     {
-      "epoch": 2.79,
-      "learning_rate": 2.1617860453880205e-05,
-      "loss": 0.2699,
+      "epoch": 2.8532814358309007,
+      "grad_norm": 0.21642963588237762,
+      "learning_rate": 2.0533682218322807e-05,
+      "loss": 0.4346,
       "step": 79170
     },
     {
-      "epoch": 2.79,
-      "learning_rate": 2.1615037943832076e-05,
-      "loss": 0.2775,
+      "epoch": 2.8534616354921254,
+      "grad_norm": 0.24576754868030548,
+      "learning_rate": 2.0530811061467802e-05,
+      "loss": 0.4105,
       "step": 79175
     },
     {
-      "epoch": 2.79,
-      "learning_rate": 2.1612215477735443e-05,
-      "loss": 0.2562,
+      "epoch": 2.8536418351533497,
+      "grad_norm": 0.16596747934818268,
+      "learning_rate": 2.052793996550463e-05,
+      "loss": 0.3769,
       "step": 79180
     },
     {
-      "epoch": 2.79,
-      "learning_rate": 2.160939305562697e-05,
-      "loss": 0.277,
+      "epoch": 2.8538220348145744,
+      "grad_norm": 0.20932526886463165,
+      "learning_rate": 2.05250689304724e-05,
+      "loss": 0.4079,
       "step": 79185
     },
     {
-      "epoch": 2.79,
-      "learning_rate": 2.1606570677543293e-05,
-      "loss": 0.2747,
+      "epoch": 2.854002234475799,
+      "grad_norm": 0.23779064416885376,
+      "learning_rate": 2.052219795641025e-05,
+      "loss": 0.4114,
       "step": 79190
     },
     {
-      "epoch": 2.79,
-      "learning_rate": 2.1603748343521062e-05,
-      "loss": 0.2458,
+      "epoch": 2.854182434137024,
+      "grad_norm": 0.17086093127727509,
+      "learning_rate": 2.0519327043357278e-05,
+      "loss": 0.3968,
       "step": 79195
     },
     {
-      "epoch": 2.79,
-      "learning_rate": 2.1600926053596904e-05,
-      "loss": 0.2618,
+      "epoch": 2.8543626337982486,
+      "grad_norm": 0.22263990342617035,
+      "learning_rate": 2.0516456191352612e-05,
+      "loss": 0.3955,
       "step": 79200
     },
     {
-      "epoch": 2.79,
-      "learning_rate": 2.159810380780749e-05,
-      "loss": 0.2707,
+      "epoch": 2.8545428334594733,
+      "grad_norm": 0.24070154130458832,
+      "learning_rate": 2.0513585400435363e-05,
+      "loss": 0.3586,
       "step": 79205
     },
     {
-      "epoch": 2.79,
-      "learning_rate": 2.1595281606189464e-05,
-      "loss": 0.2677,
+      "epoch": 2.8547230331206976,
+      "grad_norm": 0.18672329187393188,
+      "learning_rate": 2.0510714670644643e-05,
+      "loss": 0.3952,
       "step": 79210
     },
     {
-      "epoch": 2.79,
-      "learning_rate": 2.1592459448779446e-05,
-      "loss": 0.2631,
+      "epoch": 2.8549032327819224,
+      "grad_norm": 0.20222359895706177,
+      "learning_rate": 2.0507844002019564e-05,
+      "loss": 0.3989,
       "step": 79215
     },
     {
-      "epoch": 2.79,
-      "learning_rate": 2.1589637335614114e-05,
-      "loss": 0.2719,
+      "epoch": 2.855083432443147,
+      "grad_norm": 0.19654959440231323,
+      "learning_rate": 2.0504973394599247e-05,
+      "loss": 0.3831,
       "step": 79220
     },
     {
-      "epoch": 2.79,
-      "learning_rate": 2.1586815266730088e-05,
-      "loss": 0.2509,
+      "epoch": 2.8552636321043714,
+      "grad_norm": 0.24008645117282867,
+      "learning_rate": 2.050210284842279e-05,
+      "loss": 0.39,
       "step": 79225
     },
     {
-      "epoch": 2.79,
-      "learning_rate": 2.1583993242164012e-05,
-      "loss": 0.2838,
+      "epoch": 2.855443831765596,
+      "grad_norm": 0.20987041294574738,
+      "learning_rate": 2.0499232363529315e-05,
+      "loss": 0.3742,
       "step": 79230
     },
     {
-      "epoch": 2.79,
-      "learning_rate": 2.1581171261952535e-05,
-      "loss": 0.2725,
+      "epoch": 2.855624031426821,
+      "grad_norm": 0.22069920599460602,
+      "learning_rate": 2.0496361939957926e-05,
+      "loss": 0.4068,
       "step": 79235
     },
     {
-      "epoch": 2.79,
-      "learning_rate": 2.1578349326132293e-05,
-      "loss": 0.2576,
+      "epoch": 2.8558042310880456,
+      "grad_norm": 0.21462880074977875,
+      "learning_rate": 2.0493491577747738e-05,
+      "loss": 0.4358,
       "step": 79240
     },
     {
-      "epoch": 2.79,
-      "learning_rate": 2.157552743473994e-05,
-      "loss": 0.2681,
+      "epoch": 2.8559844307492703,
+      "grad_norm": 0.22585731744766235,
+      "learning_rate": 2.0490621276937853e-05,
+      "loss": 0.3979,
       "step": 79245
     },
     {
-      "epoch": 2.79,
-      "learning_rate": 2.15727055878121e-05,
-      "loss": 0.2652,
+      "epoch": 2.856164630410495,
+      "grad_norm": 0.20403482019901276,
+      "learning_rate": 2.048775103756737e-05,
+      "loss": 0.386,
       "step": 79250
     },
     {
-      "epoch": 2.79,
-      "learning_rate": 2.1569883785385413e-05,
-      "loss": 0.2611,
+      "epoch": 2.8563448300717194,
+      "grad_norm": 0.1699545532464981,
+      "learning_rate": 2.0484880859675422e-05,
+      "loss": 0.3828,
       "step": 79255
     },
     {
-      "epoch": 2.79,
-      "learning_rate": 2.1567062027496538e-05,
-      "loss": 0.2867,
+      "epoch": 2.856525029732944,
+      "grad_norm": 0.15183869004249573,
+      "learning_rate": 2.0482010743301093e-05,
+      "loss": 0.3445,
       "step": 79260
     },
     {
-      "epoch": 2.79,
-      "learning_rate": 2.156424031418209e-05,
-      "loss": 0.2822,
+      "epoch": 2.856705229394169,
+      "grad_norm": 0.25381314754486084,
+      "learning_rate": 2.0479140688483485e-05,
+      "loss": 0.4077,
       "step": 79265
     },
     {
-      "epoch": 2.79,
-      "learning_rate": 2.1561418645478722e-05,
-      "loss": 0.2569,
+      "epoch": 2.856885429055393,
+      "grad_norm": 0.18916507065296173,
+      "learning_rate": 2.0476270695261716e-05,
+      "loss": 0.3801,
       "step": 79270
     },
     {
-      "epoch": 2.79,
-      "learning_rate": 2.155859702142306e-05,
-      "loss": 0.2762,
+      "epoch": 2.857065628716618,
+      "grad_norm": 0.18349602818489075,
+      "learning_rate": 2.0473400763674876e-05,
+      "loss": 0.4151,
       "step": 79275
     },
     {
-      "epoch": 2.79,
-      "learning_rate": 2.1555775442051756e-05,
-      "loss": 0.2956,
+      "epoch": 2.8572458283778426,
+      "grad_norm": 0.1902550309896469,
+      "learning_rate": 2.0470530893762087e-05,
+      "loss": 0.4191,
       "step": 79280
     },
     {
-      "epoch": 2.79,
-      "learning_rate": 2.1552953907401433e-05,
-      "loss": 0.2615,
+      "epoch": 2.8574260280390673,
+      "grad_norm": 0.23861198127269745,
+      "learning_rate": 2.046766108556243e-05,
+      "loss": 0.3875,
       "step": 79285
     },
     {
-      "epoch": 2.79,
-      "learning_rate": 2.1550132417508736e-05,
-      "loss": 0.2734,
+      "epoch": 2.857606227700292,
+      "grad_norm": 0.19769497215747833,
+      "learning_rate": 2.0464791339115014e-05,
+      "loss": 0.3807,
       "step": 79290
     },
     {
-      "epoch": 2.79,
-      "learning_rate": 2.1547310972410287e-05,
-      "loss": 0.2429,
+      "epoch": 2.857786427361517,
+      "grad_norm": 0.20650269091129303,
+      "learning_rate": 2.0461921654458938e-05,
+      "loss": 0.3896,
       "step": 79295
     },
     {
-      "epoch": 2.79,
-      "learning_rate": 2.154448957214274e-05,
-      "loss": 0.2657,
+      "epoch": 2.857966627022741,
+      "grad_norm": 0.19639898836612701,
+      "learning_rate": 2.0459052031633297e-05,
+      "loss": 0.3815,
       "step": 79300
     },
     {
-      "epoch": 2.79,
-      "learning_rate": 2.154166821674272e-05,
-      "loss": 0.2652,
+      "epoch": 2.858146826683966,
+      "grad_norm": 0.23616774380207062,
+      "learning_rate": 2.0456182470677198e-05,
+      "loss": 0.4199,
       "step": 79305
     },
     {
-      "epoch": 2.79,
-      "learning_rate": 2.1538846906246853e-05,
-      "loss": 0.2836,
+      "epoch": 2.8583270263451905,
+      "grad_norm": 0.17947378754615784,
+      "learning_rate": 2.0453312971629734e-05,
+      "loss": 0.3959,
       "step": 79310
     },
     {
-      "epoch": 2.79,
-      "learning_rate": 2.1536025640691774e-05,
-      "loss": 0.2665,
+      "epoch": 2.8585072260064153,
+      "grad_norm": 0.18461818993091583,
+      "learning_rate": 2.0450443534529995e-05,
+      "loss": 0.3596,
       "step": 79315
     },
     {
-      "epoch": 2.79,
-      "learning_rate": 2.1533204420114122e-05,
-      "loss": 0.2522,
+      "epoch": 2.8586874256676396,
+      "grad_norm": 0.19616423547267914,
+      "learning_rate": 2.044757415941709e-05,
+      "loss": 0.3676,
       "step": 79320
     },
     {
-      "epoch": 2.79,
-      "learning_rate": 2.1530383244550534e-05,
-      "loss": 0.2672,
+      "epoch": 2.8588676253288643,
+      "grad_norm": 0.18989914655685425,
+      "learning_rate": 2.0444704846330098e-05,
+      "loss": 0.4095,
       "step": 79325
     },
     {
-      "epoch": 2.79,
-      "learning_rate": 2.152756211403762e-05,
-      "loss": 0.2577,
+      "epoch": 2.859047824990089,
+      "grad_norm": 0.20153529942035675,
+      "learning_rate": 2.0441835595308122e-05,
+      "loss": 0.4075,
       "step": 79330
     },
     {
-      "epoch": 2.79,
-      "learning_rate": 2.1524741028612034e-05,
-      "loss": 0.2838,
+      "epoch": 2.8592280246513138,
+      "grad_norm": 0.1888132393360138,
+      "learning_rate": 2.0438966406390256e-05,
+      "loss": 0.4005,
       "step": 79335
     },
     {
-      "epoch": 2.79,
-      "learning_rate": 2.1521919988310392e-05,
-      "loss": 0.287,
+      "epoch": 2.8594082243125385,
+      "grad_norm": 0.21628032624721527,
+      "learning_rate": 2.0436097279615585e-05,
+      "loss": 0.4074,
       "step": 79340
     },
     {
-      "epoch": 2.79,
-      "learning_rate": 2.151909899316933e-05,
-      "loss": 0.2482,
+      "epoch": 2.859588423973763,
+      "grad_norm": 0.18922530114650726,
+      "learning_rate": 2.0433228215023213e-05,
+      "loss": 0.3821,
       "step": 79345
     },
     {
-      "epoch": 2.79,
-      "learning_rate": 2.151627804322546e-05,
-      "loss": 0.2592,
+      "epoch": 2.8597686236349875,
+      "grad_norm": 0.24143609404563904,
+      "learning_rate": 2.0430359212652224e-05,
+      "loss": 0.3769,
       "step": 79350
     },
     {
-      "epoch": 2.79,
-      "learning_rate": 2.151345713851544e-05,
-      "loss": 0.2645,
+      "epoch": 2.8599488232962123,
+      "grad_norm": 0.22710378468036652,
+      "learning_rate": 2.042749027254169e-05,
+      "loss": 0.4179,
       "step": 79355
     },
     {
-      "epoch": 2.79,
-      "learning_rate": 2.1510636279075878e-05,
-      "loss": 0.2633,
+      "epoch": 2.860129022957437,
+      "grad_norm": 0.23377621173858643,
+      "learning_rate": 2.0424621394730735e-05,
+      "loss": 0.3965,
       "step": 79360
     },
     {
-      "epoch": 2.79,
-      "learning_rate": 2.1507815464943397e-05,
-      "loss": 0.2484,
+      "epoch": 2.8603092226186613,
+      "grad_norm": 0.19734351336956024,
+      "learning_rate": 2.042175257925842e-05,
+      "loss": 0.3964,
       "step": 79365
     },
     {
-      "epoch": 2.79,
-      "learning_rate": 2.1504994696154633e-05,
-      "loss": 0.2675,
+      "epoch": 2.860489422279886,
+      "grad_norm": 0.21778082847595215,
+      "learning_rate": 2.0418883826163833e-05,
+      "loss": 0.4076,
       "step": 79370
     },
     {
-      "epoch": 2.79,
-      "learning_rate": 2.1502173972746215e-05,
-      "loss": 0.2762,
+      "epoch": 2.8606696219411107,
+      "grad_norm": 0.1940212845802307,
+      "learning_rate": 2.0416015135486074e-05,
+      "loss": 0.3652,
       "step": 79375
     },
     {
-      "epoch": 2.79,
-      "learning_rate": 2.149935329475476e-05,
-      "loss": 0.2842,
+      "epoch": 2.8608498216023355,
+      "grad_norm": 0.21918119490146637,
+      "learning_rate": 2.0413146507264216e-05,
+      "loss": 0.411,
       "step": 79380
     },
     {
-      "epoch": 2.79,
-      "learning_rate": 2.14965326622169e-05,
-      "loss": 0.2683,
+      "epoch": 2.86103002126356,
+      "grad_norm": 0.21737195551395416,
+      "learning_rate": 2.0410277941537352e-05,
+      "loss": 0.3977,
       "step": 79385
     },
     {
-      "epoch": 2.79,
-      "learning_rate": 2.149371207516924e-05,
-      "loss": 0.2686,
+      "epoch": 2.8612102209247845,
+      "grad_norm": 0.1796942949295044,
+      "learning_rate": 2.0407409438344566e-05,
+      "loss": 0.3913,
       "step": 79390
     },
     {
-      "epoch": 2.79,
-      "learning_rate": 2.1490891533648436e-05,
-      "loss": 0.2529,
+      "epoch": 2.8613904205860092,
+      "grad_norm": 0.20798861980438232,
+      "learning_rate": 2.040454099772493e-05,
+      "loss": 0.4107,
       "step": 79395
     },
     {
-      "epoch": 2.79,
-      "learning_rate": 2.148807103769108e-05,
-      "loss": 0.268,
+      "epoch": 2.861570620247234,
+      "grad_norm": 0.2642776370048523,
+      "learning_rate": 2.040167261971754e-05,
+      "loss": 0.3942,
       "step": 79400
     },
     {
-      "epoch": 2.79,
-      "learning_rate": 2.1485250587333817e-05,
-      "loss": 0.2525,
+      "epoch": 2.8617508199084587,
+      "grad_norm": 0.2090744972229004,
+      "learning_rate": 2.039880430436147e-05,
+      "loss": 0.3992,
       "step": 79405
     },
     {
-      "epoch": 2.79,
-      "learning_rate": 2.1482430182613248e-05,
-      "loss": 0.278,
+      "epoch": 2.861931019569683,
+      "grad_norm": 0.208735391497612,
+      "learning_rate": 2.0395936051695794e-05,
+      "loss": 0.37,
       "step": 79410
     },
     {
-      "epoch": 2.79,
-      "learning_rate": 2.1479609823566013e-05,
-      "loss": 0.2466,
+      "epoch": 2.8621112192309077,
+      "grad_norm": 0.2147226631641388,
+      "learning_rate": 2.0393067861759604e-05,
+      "loss": 0.4038,
       "step": 79415
     },
     {
-      "epoch": 2.79,
-      "learning_rate": 2.1476789510228725e-05,
-      "loss": 0.2907,
+      "epoch": 2.8622914188921325,
+      "grad_norm": 0.21452739834785461,
+      "learning_rate": 2.0390199734591967e-05,
+      "loss": 0.4,
       "step": 79420
     },
     {
-      "epoch": 2.79,
-      "learning_rate": 2.1473969242638e-05,
-      "loss": 0.2831,
+      "epoch": 2.862471618553357,
+      "grad_norm": 0.21164734661579132,
+      "learning_rate": 2.0387331670231972e-05,
+      "loss": 0.3923,
       "step": 79425
     },
     {
-      "epoch": 2.79,
-      "learning_rate": 2.1471149020830454e-05,
-      "loss": 0.2742,
+      "epoch": 2.862651818214582,
+      "grad_norm": 0.25900834798812866,
+      "learning_rate": 2.0384463668718695e-05,
+      "loss": 0.4046,
       "step": 79430
     },
     {
-      "epoch": 2.79,
-      "learning_rate": 2.1468328844842722e-05,
-      "loss": 0.2506,
+      "epoch": 2.8628320178758067,
+      "grad_norm": 0.2033701092004776,
+      "learning_rate": 2.0381595730091187e-05,
+      "loss": 0.4171,
       "step": 79435
     },
     {
-      "epoch": 2.79,
-      "learning_rate": 2.1465508714711413e-05,
-      "loss": 0.2554,
+      "epoch": 2.863012217537031,
+      "grad_norm": 0.23494963347911835,
+      "learning_rate": 2.0378727854388557e-05,
+      "loss": 0.4213,
       "step": 79440
     },
     {
-      "epoch": 2.8,
-      "learning_rate": 2.146268863047314e-05,
-      "loss": 0.2887,
+      "epoch": 2.8631924171982557,
+      "grad_norm": 0.21937774121761322,
+      "learning_rate": 2.0375860041649874e-05,
+      "loss": 0.3904,
       "step": 79445
     },
     {
-      "epoch": 2.8,
-      "learning_rate": 2.145986859216452e-05,
-      "loss": 0.264,
+      "epoch": 2.8633726168594804,
+      "grad_norm": 0.18949083983898163,
+      "learning_rate": 2.0372992291914182e-05,
+      "loss": 0.3691,
       "step": 79450
     },
     {
-      "epoch": 2.8,
-      "learning_rate": 2.145704859982218e-05,
-      "loss": 0.2591,
+      "epoch": 2.8635528165207047,
+      "grad_norm": 0.23285529017448425,
+      "learning_rate": 2.037012460522059e-05,
+      "loss": 0.4134,
       "step": 79455
     },
     {
-      "epoch": 2.8,
-      "learning_rate": 2.145422865348273e-05,
-      "loss": 0.2317,
+      "epoch": 2.8637330161819294,
+      "grad_norm": 0.2178962379693985,
+      "learning_rate": 2.036725698160814e-05,
+      "loss": 0.3796,
       "step": 79460
     },
     {
-      "epoch": 2.8,
-      "learning_rate": 2.1451408753182772e-05,
-      "loss": 0.2664,
+      "epoch": 2.863913215843154,
+      "grad_norm": 0.20775160193443298,
+      "learning_rate": 2.0364389421115936e-05,
+      "loss": 0.3863,
       "step": 79465
     },
     {
-      "epoch": 2.8,
-      "learning_rate": 2.1448588898958947e-05,
-      "loss": 0.2619,
+      "epoch": 2.864093415504379,
+      "grad_norm": 0.20087379217147827,
+      "learning_rate": 2.0361521923783018e-05,
+      "loss": 0.3904,
       "step": 79470
     },
     {
-      "epoch": 2.8,
-      "learning_rate": 2.1445769090847854e-05,
-      "loss": 0.2431,
+      "epoch": 2.8642736151656036,
+      "grad_norm": 0.2299887239933014,
+      "learning_rate": 2.035865448964846e-05,
+      "loss": 0.4115,
       "step": 79475
     },
     {
-      "epoch": 2.8,
-      "learning_rate": 2.14429493288861e-05,
-      "loss": 0.249,
+      "epoch": 2.8644538148268284,
+      "grad_norm": 0.22085849940776825,
+      "learning_rate": 2.0355787118751346e-05,
+      "loss": 0.3658,
       "step": 79480
     },
     {
-      "epoch": 2.8,
-      "learning_rate": 2.1440129613110302e-05,
-      "loss": 0.2691,
+      "epoch": 2.8646340144880527,
+      "grad_norm": 0.22253592312335968,
+      "learning_rate": 2.0352919811130726e-05,
+      "loss": 0.4168,
       "step": 79485
     },
     {
-      "epoch": 2.8,
-      "learning_rate": 2.143730994355708e-05,
-      "loss": 0.2644,
+      "epoch": 2.8648142141492774,
+      "grad_norm": 0.20002683997154236,
+      "learning_rate": 2.035005256682568e-05,
+      "loss": 0.4205,
       "step": 79490
     },
     {
-      "epoch": 2.8,
-      "learning_rate": 2.1434490320263048e-05,
-      "loss": 0.2756,
+      "epoch": 2.864994413810502,
+      "grad_norm": 0.22108720242977142,
+      "learning_rate": 2.034718538587526e-05,
+      "loss": 0.4173,
       "step": 79495
     },
     {
-      "epoch": 2.8,
-      "learning_rate": 2.1431670743264806e-05,
-      "loss": 0.2753,
+      "epoch": 2.8651746134717264,
+      "grad_norm": 0.20260627567768097,
+      "learning_rate": 2.034431826831854e-05,
+      "loss": 0.4003,
       "step": 79500
     },
     {
-      "epoch": 2.8,
-      "eval_loss": 0.2603822350502014,
-      "eval_runtime": 10.5518,
-      "eval_samples_per_second": 9.477,
-      "eval_steps_per_second": 9.477,
+      "epoch": 2.8651746134717264,
+      "eval_loss": 0.43116477131843567,
+      "eval_runtime": 3.5328,
+      "eval_samples_per_second": 28.306,
+      "eval_steps_per_second": 7.077,
       "step": 79500
     },
     {
-      "epoch": 2.8,
-      "learning_rate": 2.1428851212598958e-05,
-      "loss": 0.2806,
+      "epoch": 2.865354813132951,
+      "grad_norm": 0.20049650967121124,
+      "learning_rate": 2.0341451214194586e-05,
+      "loss": 0.3975,
       "step": 79505
     },
     {
-      "epoch": 2.8,
-      "learning_rate": 2.1426031728302126e-05,
-      "loss": 0.2657,
+      "epoch": 2.865535012794176,
+      "grad_norm": 0.21907275915145874,
+      "learning_rate": 2.0338584223542462e-05,
+      "loss": 0.4186,
       "step": 79510
     },
     {
-      "epoch": 2.8,
-      "learning_rate": 2.1423212290410922e-05,
-      "loss": 0.2625,
+      "epoch": 2.8657152124554006,
+      "grad_norm": 0.2359967678785324,
+      "learning_rate": 2.033571729640121e-05,
+      "loss": 0.3657,
       "step": 79515
     },
     {
-      "epoch": 2.8,
-      "learning_rate": 2.1420392898961956e-05,
-      "loss": 0.2688,
+      "epoch": 2.8658954121166254,
+      "grad_norm": 0.24726665019989014,
+      "learning_rate": 2.0332850432809922e-05,
+      "loss": 0.3807,
       "step": 79520
     },
     {
-      "epoch": 2.8,
-      "learning_rate": 2.141757355399181e-05,
-      "loss": 0.2591,
+      "epoch": 2.86607561177785,
+      "grad_norm": 0.24529293179512024,
+      "learning_rate": 2.0329983632807632e-05,
+      "loss": 0.4069,
       "step": 79525
     },
     {
-      "epoch": 2.8,
-      "learning_rate": 2.1414754255537126e-05,
-      "loss": 0.2542,
+      "epoch": 2.8662558114390744,
+      "grad_norm": 0.19809693098068237,
+      "learning_rate": 2.032711689643342e-05,
+      "loss": 0.3777,
       "step": 79530
     },
     {
-      "epoch": 2.8,
-      "learning_rate": 2.141193500363449e-05,
-      "loss": 0.2542,
+      "epoch": 2.866436011100299,
+      "grad_norm": 0.1797298640012741,
+      "learning_rate": 2.032425022372634e-05,
+      "loss": 0.3542,
       "step": 79535
     },
     {
-      "epoch": 2.8,
-      "learning_rate": 2.140911579832052e-05,
-      "loss": 0.2587,
+      "epoch": 2.866616210761524,
+      "grad_norm": 0.1798580288887024,
+      "learning_rate": 2.032138361472543e-05,
+      "loss": 0.3837,
       "step": 79540
     },
     {
-      "epoch": 2.8,
-      "learning_rate": 2.1406296639631798e-05,
-      "loss": 0.2797,
+      "epoch": 2.866796410422748,
+      "grad_norm": 0.2182854264974594,
+      "learning_rate": 2.0318517069469775e-05,
+      "loss": 0.406,
       "step": 79545
     },
     {
-      "epoch": 2.8,
-      "learning_rate": 2.140347752760496e-05,
-      "loss": 0.2718,
+      "epoch": 2.866976610083973,
+      "grad_norm": 0.19084390997886658,
+      "learning_rate": 2.0315650587998416e-05,
+      "loss": 0.3815,
       "step": 79550
     },
     {
-      "epoch": 2.8,
-      "learning_rate": 2.1400658462276595e-05,
-      "loss": 0.278,
+      "epoch": 2.8671568097451976,
+      "grad_norm": 0.20089589059352875,
+      "learning_rate": 2.0312784170350404e-05,
+      "loss": 0.3958,
       "step": 79555
     },
     {
-      "epoch": 2.8,
-      "learning_rate": 2.1397839443683305e-05,
-      "loss": 0.2467,
+      "epoch": 2.8673370094064223,
+      "grad_norm": 0.26835712790489197,
+      "learning_rate": 2.030991781656481e-05,
+      "loss": 0.3866,
       "step": 79560
     },
     {
-      "epoch": 2.8,
-      "learning_rate": 2.1395020471861688e-05,
-      "loss": 0.2569,
+      "epoch": 2.867517209067647,
+      "grad_norm": 0.1739826798439026,
+      "learning_rate": 2.030705152668066e-05,
+      "loss": 0.3778,
       "step": 79565
     },
     {
-      "epoch": 2.8,
-      "learning_rate": 2.1392201546848368e-05,
-      "loss": 0.2563,
+      "epoch": 2.867697408728872,
+      "grad_norm": 0.23499102890491486,
+      "learning_rate": 2.0304185300737046e-05,
+      "loss": 0.3978,
       "step": 79570
     },
     {
-      "epoch": 2.8,
-      "learning_rate": 2.138938266867993e-05,
-      "loss": 0.2755,
+      "epoch": 2.867877608390096,
+      "grad_norm": 0.19982177019119263,
+      "learning_rate": 2.030131913877299e-05,
+      "loss": 0.4005,
       "step": 79575
     },
     {
-      "epoch": 2.8,
-      "learning_rate": 2.138656383739297e-05,
-      "loss": 0.2489,
+      "epoch": 2.868057808051321,
+      "grad_norm": 0.18712134659290314,
+      "learning_rate": 2.0298453040827544e-05,
+      "loss": 0.4348,
       "step": 79580
     },
     {
-      "epoch": 2.8,
-      "learning_rate": 2.1383745053024106e-05,
-      "loss": 0.2614,
+      "epoch": 2.8682380077125456,
+      "grad_norm": 0.26873958110809326,
+      "learning_rate": 2.0295587006939772e-05,
+      "loss": 0.3693,
       "step": 79585
     },
     {
-      "epoch": 2.8,
-      "learning_rate": 2.138092631560993e-05,
-      "loss": 0.2598,
+      "epoch": 2.8684182073737703,
+      "grad_norm": 0.1947907954454422,
+      "learning_rate": 2.0292721037148717e-05,
+      "loss": 0.3872,
       "step": 79590
     },
     {
-      "epoch": 2.8,
-      "learning_rate": 2.137810762518704e-05,
-      "loss": 0.2725,
+      "epoch": 2.8685984070349946,
+      "grad_norm": 0.16691918671131134,
+      "learning_rate": 2.0289855131493422e-05,
+      "loss": 0.3799,
       "step": 79595
     },
     {
-      "epoch": 2.8,
-      "learning_rate": 2.1375288981792026e-05,
-      "loss": 0.2691,
+      "epoch": 2.8687786066962193,
+      "grad_norm": 0.15523312985897064,
+      "learning_rate": 2.0286989290012945e-05,
+      "loss": 0.3837,
       "step": 79600
     },
     {
-      "epoch": 2.8,
-      "learning_rate": 2.1372470385461506e-05,
-      "loss": 0.2584,
+      "epoch": 2.868958806357444,
+      "grad_norm": 0.22626052796840668,
+      "learning_rate": 2.0284123512746317e-05,
+      "loss": 0.3807,
       "step": 79605
     },
     {
-      "epoch": 2.8,
-      "learning_rate": 2.136965183623207e-05,
-      "loss": 0.2645,
+      "epoch": 2.869139006018669,
+      "grad_norm": 0.25360921025276184,
+      "learning_rate": 2.0281257799732602e-05,
+      "loss": 0.401,
       "step": 79610
     },
     {
-      "epoch": 2.8,
-      "learning_rate": 2.1366833334140302e-05,
-      "loss": 0.2702,
+      "epoch": 2.8693192056798935,
+      "grad_norm": 0.21159762144088745,
+      "learning_rate": 2.0278392151010837e-05,
+      "loss": 0.4108,
       "step": 79615
     },
     {
-      "epoch": 2.8,
-      "learning_rate": 2.1364014879222807e-05,
-      "loss": 0.2578,
+      "epoch": 2.869499405341118,
+      "grad_norm": 0.19640955328941345,
+      "learning_rate": 2.0275526566620058e-05,
+      "loss": 0.3899,
       "step": 79620
     },
     {
-      "epoch": 2.8,
-      "learning_rate": 2.136119647151619e-05,
-      "loss": 0.2612,
+      "epoch": 2.8696796050023425,
+      "grad_norm": 0.20014135539531708,
+      "learning_rate": 2.0272661046599318e-05,
+      "loss": 0.3942,
       "step": 79625
     },
     {
-      "epoch": 2.8,
-      "learning_rate": 2.1358378111057046e-05,
-      "loss": 0.2815,
+      "epoch": 2.8698598046635673,
+      "grad_norm": 0.20158907771110535,
+      "learning_rate": 2.026979559098766e-05,
+      "loss": 0.367,
       "step": 79630
     },
     {
-      "epoch": 2.8,
-      "learning_rate": 2.1355559797881957e-05,
-      "loss": 0.2783,
+      "epoch": 2.870040004324792,
+      "grad_norm": 0.18245342373847961,
+      "learning_rate": 2.0266930199824108e-05,
+      "loss": 0.3979,
       "step": 79635
     },
     {
-      "epoch": 2.8,
-      "learning_rate": 2.1353305181410992e-05,
-      "loss": 0.2755,
+      "epoch": 2.8702202039860163,
+      "grad_norm": 0.2652152478694916,
+      "learning_rate": 2.0264064873147735e-05,
+      "loss": 0.3925,
       "step": 79640
     },
     {
-      "epoch": 2.8,
-      "learning_rate": 2.1350486953439432e-05,
-      "loss": 0.2867,
+      "epoch": 2.870400403647241,
+      "grad_norm": 0.22294139862060547,
+      "learning_rate": 2.026119961099754e-05,
+      "loss": 0.4227,
       "step": 79645
     },
     {
-      "epoch": 2.8,
-      "learning_rate": 2.134766877285438e-05,
-      "loss": 0.2578,
+      "epoch": 2.8705806033084658,
+      "grad_norm": 0.20929698646068573,
+      "learning_rate": 2.02583344134126e-05,
+      "loss": 0.3903,
       "step": 79650
     },
     {
-      "epoch": 2.8,
-      "learning_rate": 2.1344850639692447e-05,
-      "loss": 0.2673,
+      "epoch": 2.8707608029696905,
+      "grad_norm": 0.2380889505147934,
+      "learning_rate": 2.0255469280431932e-05,
+      "loss": 0.3933,
       "step": 79655
     },
     {
-      "epoch": 2.8,
-      "learning_rate": 2.1342032553990228e-05,
-      "loss": 0.2736,
+      "epoch": 2.8709410026309152,
+      "grad_norm": 0.21243393421173096,
+      "learning_rate": 2.025260421209457e-05,
+      "loss": 0.4065,
       "step": 79660
     },
     {
-      "epoch": 2.8,
-      "learning_rate": 2.13392145157843e-05,
-      "loss": 0.2522,
+      "epoch": 2.87112120229214,
+      "grad_norm": 0.2032131552696228,
+      "learning_rate": 2.0249739208439562e-05,
+      "loss": 0.373,
       "step": 79665
     },
     {
-      "epoch": 2.8,
-      "learning_rate": 2.1336396525111252e-05,
-      "loss": 0.266,
+      "epoch": 2.8713014019533643,
+      "grad_norm": 0.2194889783859253,
+      "learning_rate": 2.0246874269505934e-05,
+      "loss": 0.4264,
       "step": 79670
     },
     {
-      "epoch": 2.8,
-      "learning_rate": 2.133357858200769e-05,
-      "loss": 0.2527,
+      "epoch": 2.871481601614589,
+      "grad_norm": 0.20919016003608704,
+      "learning_rate": 2.0244009395332725e-05,
+      "loss": 0.4062,
       "step": 79675
     },
     {
-      "epoch": 2.8,
-      "learning_rate": 2.1330760686510188e-05,
-      "loss": 0.2538,
+      "epoch": 2.8716618012758137,
+      "grad_norm": 0.17519910633563995,
+      "learning_rate": 2.024114458595897e-05,
+      "loss": 0.3697,
       "step": 79680
     },
     {
-      "epoch": 2.8,
-      "learning_rate": 2.132794283865535e-05,
-      "loss": 0.2634,
+      "epoch": 2.871842000937038,
+      "grad_norm": 0.2320595383644104,
+      "learning_rate": 2.0238279841423693e-05,
+      "loss": 0.3617,
       "step": 79685
     },
     {
-      "epoch": 2.8,
-      "learning_rate": 2.1325125038479736e-05,
-      "loss": 0.2837,
+      "epoch": 2.8720222005982627,
+      "grad_norm": 0.20876125991344452,
+      "learning_rate": 2.0235415161765936e-05,
+      "loss": 0.4013,
       "step": 79690
     },
     {
-      "epoch": 2.8,
-      "learning_rate": 2.132230728601997e-05,
-      "loss": 0.2694,
+      "epoch": 2.8722024002594875,
+      "grad_norm": 0.2067393809556961,
+      "learning_rate": 2.0232550547024726e-05,
+      "loss": 0.4045,
       "step": 79695
     },
     {
-      "epoch": 2.8,
-      "learning_rate": 2.1319489581312614e-05,
-      "loss": 0.2669,
+      "epoch": 2.872382599920712,
+      "grad_norm": 0.2197972983121872,
+      "learning_rate": 2.0229685997239088e-05,
+      "loss": 0.3972,
       "step": 79700
     },
     {
-      "epoch": 2.8,
-      "learning_rate": 2.131667192439426e-05,
-      "loss": 0.2794,
+      "epoch": 2.872562799581937,
+      "grad_norm": 0.23758886754512787,
+      "learning_rate": 2.0226821512448057e-05,
+      "loss": 0.3835,
       "step": 79705
     },
     {
-      "epoch": 2.8,
-      "learning_rate": 2.1313854315301486e-05,
-      "loss": 0.2697,
+      "epoch": 2.8727429992431617,
+      "grad_norm": 0.20239755511283875,
+      "learning_rate": 2.0223957092690653e-05,
+      "loss": 0.3908,
       "step": 79710
     },
     {
-      "epoch": 2.8,
-      "learning_rate": 2.1311036754070897e-05,
-      "loss": 0.2478,
+      "epoch": 2.872923198904386,
+      "grad_norm": 0.23541153967380524,
+      "learning_rate": 2.0221092738005916e-05,
+      "loss": 0.3804,
       "step": 79715
     },
     {
-      "epoch": 2.8,
-      "learning_rate": 2.130821924073906e-05,
-      "loss": 0.2492,
+      "epoch": 2.8731033985656107,
+      "grad_norm": 0.19039641320705414,
+      "learning_rate": 2.021822844843287e-05,
+      "loss": 0.3887,
       "step": 79720
     },
     {
-      "epoch": 2.8,
-      "learning_rate": 2.1305401775342566e-05,
-      "loss": 0.2565,
+      "epoch": 2.8732835982268354,
+      "grad_norm": 0.20644930005073547,
+      "learning_rate": 2.0215364224010522e-05,
+      "loss": 0.3808,
       "step": 79725
     },
     {
-      "epoch": 2.81,
-      "learning_rate": 2.1302584357917985e-05,
-      "loss": 0.2698,
+      "epoch": 2.8734637978880597,
+      "grad_norm": 0.2427452951669693,
+      "learning_rate": 2.021250006477792e-05,
+      "loss": 0.4126,
       "step": 79730
     },
     {
-      "epoch": 2.81,
-      "learning_rate": 2.1299766988501913e-05,
-      "loss": 0.2696,
+      "epoch": 2.8736439975492845,
+      "grad_norm": 0.19568517804145813,
+      "learning_rate": 2.020963597077408e-05,
+      "loss": 0.4055,
       "step": 79735
     },
     {
-      "epoch": 2.81,
-      "learning_rate": 2.1296949667130935e-05,
-      "loss": 0.278,
+      "epoch": 2.873824197210509,
+      "grad_norm": 0.21443352103233337,
+      "learning_rate": 2.0206771942038008e-05,
+      "loss": 0.3971,
       "step": 79740
     },
     {
-      "epoch": 2.81,
-      "learning_rate": 2.1294132393841627e-05,
-      "loss": 0.2425,
+      "epoch": 2.874004396871734,
+      "grad_norm": 0.19338439404964447,
+      "learning_rate": 2.020390797860876e-05,
+      "loss": 0.3834,
       "step": 79745
     },
     {
-      "epoch": 2.81,
-      "learning_rate": 2.1291315168670553e-05,
-      "loss": 0.2866,
+      "epoch": 2.8741845965329587,
+      "grad_norm": 0.1852528303861618,
+      "learning_rate": 2.0201044080525314e-05,
+      "loss": 0.3668,
       "step": 79750
     },
     {
-      "epoch": 2.81,
-      "learning_rate": 2.1288497991654322e-05,
-      "loss": 0.2734,
+      "epoch": 2.8743647961941834,
+      "grad_norm": 0.2564043402671814,
+      "learning_rate": 2.0198180247826734e-05,
+      "loss": 0.4105,
       "step": 79755
     },
     {
-      "epoch": 2.81,
-      "learning_rate": 2.128568086282949e-05,
-      "loss": 0.2391,
+      "epoch": 2.8745449958554077,
+      "grad_norm": 0.227681502699852,
+      "learning_rate": 2.0195316480552013e-05,
+      "loss": 0.3855,
       "step": 79760
     },
     {
-      "epoch": 2.81,
-      "learning_rate": 2.128286378223264e-05,
-      "loss": 0.2667,
+      "epoch": 2.8747251955166324,
+      "grad_norm": 0.21400003135204315,
+      "learning_rate": 2.0192452778740166e-05,
+      "loss": 0.384,
       "step": 79765
     },
     {
-      "epoch": 2.81,
-      "learning_rate": 2.1280046749900364e-05,
-      "loss": 0.2573,
+      "epoch": 2.874905395177857,
+      "grad_norm": 0.17560800909996033,
+      "learning_rate": 2.018958914243023e-05,
+      "loss": 0.3919,
       "step": 79770
     },
     {
-      "epoch": 2.81,
-      "learning_rate": 2.127722976586923e-05,
-      "loss": 0.2598,
+      "epoch": 2.8750855948390814,
+      "grad_norm": 0.2656613886356354,
+      "learning_rate": 2.0186725571661207e-05,
+      "loss": 0.4083,
       "step": 79775
     },
     {
-      "epoch": 2.81,
-      "learning_rate": 2.1274412830175813e-05,
-      "loss": 0.264,
+      "epoch": 2.875265794500306,
+      "grad_norm": 0.23810866475105286,
+      "learning_rate": 2.018386206647211e-05,
+      "loss": 0.3885,
       "step": 79780
     },
     {
-      "epoch": 2.81,
-      "learning_rate": 2.127159594285668e-05,
-      "loss": 0.2813,
+      "epoch": 2.875445994161531,
+      "grad_norm": 0.23813889920711517,
+      "learning_rate": 2.0180998626901966e-05,
+      "loss": 0.3846,
       "step": 79785
     },
     {
-      "epoch": 2.81,
-      "learning_rate": 2.1268779103948423e-05,
-      "loss": 0.2484,
+      "epoch": 2.8756261938227556,
+      "grad_norm": 0.19781778752803802,
+      "learning_rate": 2.017813525298978e-05,
+      "loss": 0.385,
       "step": 79790
     },
     {
-      "epoch": 2.81,
-      "learning_rate": 2.1265962313487618e-05,
-      "loss": 0.2816,
+      "epoch": 2.8758063934839804,
+      "grad_norm": 0.23655954003334045,
+      "learning_rate": 2.017527194477457e-05,
+      "loss": 0.406,
       "step": 79795
     },
     {
-      "epoch": 2.81,
-      "learning_rate": 2.126314557151083e-05,
-      "loss": 0.2595,
+      "epoch": 2.875986593145205,
+      "grad_norm": 0.2517338991165161,
+      "learning_rate": 2.0172408702295347e-05,
+      "loss": 0.3935,
       "step": 79800
     },
     {
-      "epoch": 2.81,
-      "learning_rate": 2.1260328878054618e-05,
-      "loss": 0.2617,
+      "epoch": 2.8761667928064294,
+      "grad_norm": 0.19464854896068573,
+      "learning_rate": 2.0169545525591117e-05,
+      "loss": 0.41,
       "step": 79805
     },
     {
-      "epoch": 2.81,
-      "learning_rate": 2.1257512233155587e-05,
-      "loss": 0.2855,
+      "epoch": 2.876346992467654,
+      "grad_norm": 0.2122073471546173,
+      "learning_rate": 2.0166682414700896e-05,
+      "loss": 0.3577,
       "step": 79810
     },
     {
-      "epoch": 2.81,
-      "learning_rate": 2.1254695636850288e-05,
-      "loss": 0.2838,
+      "epoch": 2.876527192128879,
+      "grad_norm": 0.22305898368358612,
+      "learning_rate": 2.01638193696637e-05,
+      "loss": 0.4146,
       "step": 79815
     },
     {
-      "epoch": 2.81,
-      "learning_rate": 2.12518790891753e-05,
-      "loss": 0.2773,
+      "epoch": 2.8767073917901036,
+      "grad_norm": 0.2052910327911377,
+      "learning_rate": 2.0160956390518508e-05,
+      "loss": 0.3865,
       "step": 79820
     },
     {
-      "epoch": 2.81,
-      "learning_rate": 2.1249062590167186e-05,
-      "loss": 0.2413,
+      "epoch": 2.876887591451328,
+      "grad_norm": 0.23175741732120514,
+      "learning_rate": 2.0158093477304367e-05,
+      "loss": 0.3845,
       "step": 79825
     },
     {
-      "epoch": 2.81,
-      "learning_rate": 2.124624613986253e-05,
-      "loss": 0.2865,
+      "epoch": 2.8770677911125526,
+      "grad_norm": 0.14851652085781097,
+      "learning_rate": 2.0155230630060252e-05,
+      "loss": 0.3451,
       "step": 79830
     },
     {
-      "epoch": 2.81,
-      "learning_rate": 2.1243429738297895e-05,
-      "loss": 0.2846,
+      "epoch": 2.8772479907737774,
+      "grad_norm": 0.21722625195980072,
+      "learning_rate": 2.0152367848825195e-05,
+      "loss": 0.4043,
       "step": 79835
     },
     {
-      "epoch": 2.81,
-      "learning_rate": 2.124061338550984e-05,
-      "loss": 0.2647,
+      "epoch": 2.877428190435002,
+      "grad_norm": 0.17405612766742706,
+      "learning_rate": 2.014950513363818e-05,
+      "loss": 0.3895,
       "step": 79840
     },
     {
-      "epoch": 2.81,
-      "learning_rate": 2.1237797081534948e-05,
-      "loss": 0.2382,
+      "epoch": 2.877608390096227,
+      "grad_norm": 0.23024611175060272,
+      "learning_rate": 2.0146642484538217e-05,
+      "loss": 0.4022,
       "step": 79845
     },
     {
-      "epoch": 2.81,
-      "learning_rate": 2.123498082640978e-05,
-      "loss": 0.2995,
+      "epoch": 2.877788589757451,
+      "grad_norm": 0.23710598051548004,
+      "learning_rate": 2.0143779901564313e-05,
+      "loss": 0.3921,
       "step": 79850
     },
     {
-      "epoch": 2.81,
-      "learning_rate": 2.1232164620170908e-05,
-      "loss": 0.2746,
+      "epoch": 2.877968789418676,
+      "grad_norm": 0.19851794838905334,
+      "learning_rate": 2.014091738475547e-05,
+      "loss": 0.4004,
       "step": 79855
     },
     {
-      "epoch": 2.81,
-      "learning_rate": 2.1229348462854898e-05,
-      "loss": 0.2655,
+      "epoch": 2.8781489890799006,
+      "grad_norm": 0.19404460489749908,
+      "learning_rate": 2.013805493415068e-05,
+      "loss": 0.3714,
       "step": 79860
     },
     {
-      "epoch": 2.81,
-      "learning_rate": 2.1226532354498303e-05,
-      "loss": 0.2675,
+      "epoch": 2.8783291887411253,
+      "grad_norm": 0.1905554085969925,
+      "learning_rate": 2.0135192549788955e-05,
+      "loss": 0.3373,
       "step": 79865
     },
     {
-      "epoch": 2.81,
-      "learning_rate": 2.12237162951377e-05,
-      "loss": 0.2652,
+      "epoch": 2.8785093884023496,
+      "grad_norm": 0.20806780457496643,
+      "learning_rate": 2.0132330231709287e-05,
+      "loss": 0.387,
       "step": 79870
     },
     {
-      "epoch": 2.81,
-      "learning_rate": 2.1220900284809657e-05,
-      "loss": 0.2666,
+      "epoch": 2.8786895880635743,
+      "grad_norm": 0.2478901594877243,
+      "learning_rate": 2.012946797995068e-05,
+      "loss": 0.4016,
       "step": 79875
     },
     {
-      "epoch": 2.81,
-      "learning_rate": 2.1218084323550728e-05,
-      "loss": 0.2566,
+      "epoch": 2.878869787724799,
+      "grad_norm": 0.17782168090343475,
+      "learning_rate": 2.012660579455213e-05,
+      "loss": 0.4067,
       "step": 79880
     },
     {
-      "epoch": 2.81,
-      "learning_rate": 2.121526841139749e-05,
-      "loss": 0.2406,
+      "epoch": 2.879049987386024,
+      "grad_norm": 0.2104375660419464,
+      "learning_rate": 2.0123743675552624e-05,
+      "loss": 0.3679,
       "step": 79885
     },
     {
-      "epoch": 2.81,
-      "learning_rate": 2.1212452548386497e-05,
-      "loss": 0.2773,
+      "epoch": 2.8792301870472485,
+      "grad_norm": 0.16701386868953705,
+      "learning_rate": 2.0120881622991178e-05,
+      "loss": 0.3655,
       "step": 79890
     },
     {
-      "epoch": 2.81,
-      "learning_rate": 2.1209636734554307e-05,
-      "loss": 0.2602,
+      "epoch": 2.879410386708473,
+      "grad_norm": 0.24730798602104187,
+      "learning_rate": 2.0118019636906765e-05,
+      "loss": 0.3892,
       "step": 79895
     },
     {
-      "epoch": 2.81,
-      "learning_rate": 2.1206820969937485e-05,
-      "loss": 0.2835,
+      "epoch": 2.8795905863696976,
+      "grad_norm": 0.19160984456539154,
+      "learning_rate": 2.0115157717338396e-05,
+      "loss": 0.4107,
       "step": 79900
     },
     {
-      "epoch": 2.81,
-      "learning_rate": 2.1204005254572594e-05,
-      "loss": 0.2683,
+      "epoch": 2.8797707860309223,
+      "grad_norm": 0.22043399512767792,
+      "learning_rate": 2.0112295864325057e-05,
+      "loss": 0.3838,
       "step": 79905
     },
     {
-      "epoch": 2.81,
-      "learning_rate": 2.1201189588496197e-05,
-      "loss": 0.2697,
+      "epoch": 2.879950985692147,
+      "grad_norm": 0.1867411732673645,
+      "learning_rate": 2.0109434077905737e-05,
+      "loss": 0.4019,
       "step": 79910
     },
     {
-      "epoch": 2.81,
-      "learning_rate": 2.119837397174485e-05,
-      "loss": 0.2709,
+      "epoch": 2.8801311853533713,
+      "grad_norm": 0.23410554230213165,
+      "learning_rate": 2.0106572358119433e-05,
+      "loss": 0.398,
       "step": 79915
     },
     {
-      "epoch": 2.81,
-      "learning_rate": 2.1195558404355104e-05,
-      "loss": 0.2803,
+      "epoch": 2.880311385014596,
+      "grad_norm": 0.2368057370185852,
+      "learning_rate": 2.0103710705005142e-05,
+      "loss": 0.3796,
       "step": 79920
     },
     {
-      "epoch": 2.81,
-      "learning_rate": 2.1192742886363532e-05,
-      "loss": 0.2833,
+      "epoch": 2.880491584675821,
+      "grad_norm": 0.14740437269210815,
+      "learning_rate": 2.0100849118601824e-05,
+      "loss": 0.3645,
       "step": 79925
     },
     {
-      "epoch": 2.81,
-      "learning_rate": 2.1189927417806678e-05,
-      "loss": 0.2837,
+      "epoch": 2.8806717843370455,
+      "grad_norm": 0.18294650316238403,
+      "learning_rate": 2.0097987598948507e-05,
+      "loss": 0.3756,
       "step": 79930
     },
     {
-      "epoch": 2.81,
-      "learning_rate": 2.1187111998721114e-05,
-      "loss": 0.2748,
+      "epoch": 2.8808519839982702,
+      "grad_norm": 0.18560317158699036,
+      "learning_rate": 2.0095126146084145e-05,
+      "loss": 0.3951,
       "step": 79935
     },
     {
-      "epoch": 2.81,
-      "learning_rate": 2.1184296629143377e-05,
-      "loss": 0.2607,
+      "epoch": 2.881032183659495,
+      "grad_norm": 0.2644931375980377,
+      "learning_rate": 2.0092264760047758e-05,
+      "loss": 0.3868,
       "step": 79940
     },
     {
-      "epoch": 2.81,
-      "learning_rate": 2.1181481309110043e-05,
-      "loss": 0.2764,
+      "epoch": 2.8812123833207193,
+      "grad_norm": 0.197056844830513,
+      "learning_rate": 2.00894034408783e-05,
+      "loss": 0.3664,
       "step": 79945
     },
     {
-      "epoch": 2.81,
-      "learning_rate": 2.1178666038657653e-05,
-      "loss": 0.2747,
+      "epoch": 2.881392582981944,
+      "grad_norm": 0.2113136351108551,
+      "learning_rate": 2.0086542188614772e-05,
+      "loss": 0.3959,
       "step": 79950
     },
     {
-      "epoch": 2.81,
-      "learning_rate": 2.1175850817822774e-05,
-      "loss": 0.3014,
+      "epoch": 2.8815727826431687,
+      "grad_norm": 0.21340829133987427,
+      "learning_rate": 2.0083681003296158e-05,
+      "loss": 0.3783,
       "step": 79955
     },
     {
-      "epoch": 2.81,
-      "learning_rate": 2.1173035646641938e-05,
-      "loss": 0.2642,
+      "epoch": 2.881752982304393,
+      "grad_norm": 0.2382301241159439,
+      "learning_rate": 2.0080819884961437e-05,
+      "loss": 0.3865,
       "step": 79960
     },
     {
-      "epoch": 2.81,
-      "learning_rate": 2.1170220525151724e-05,
-      "loss": 0.2579,
+      "epoch": 2.8819331819656178,
+      "grad_norm": 0.2526952028274536,
+      "learning_rate": 2.007795883364959e-05,
+      "loss": 0.3977,
       "step": 79965
     },
     {
-      "epoch": 2.81,
-      "learning_rate": 2.1167405453388673e-05,
-      "loss": 0.2858,
+      "epoch": 2.8821133816268425,
+      "grad_norm": 0.2137371152639389,
+      "learning_rate": 2.0075097849399603e-05,
+      "loss": 0.3798,
       "step": 79970
     },
     {
-      "epoch": 2.81,
-      "learning_rate": 2.1164590431389326e-05,
-      "loss": 0.268,
+      "epoch": 2.8822935812880672,
+      "grad_norm": 0.2008858621120453,
+      "learning_rate": 2.007223693225045e-05,
+      "loss": 0.4151,
       "step": 79975
     },
     {
-      "epoch": 2.81,
-      "learning_rate": 2.116177545919025e-05,
-      "loss": 0.2737,
+      "epoch": 2.882473780949292,
+      "grad_norm": 0.20356670022010803,
+      "learning_rate": 2.006937608224112e-05,
+      "loss": 0.4056,
       "step": 79980
     },
     {
-      "epoch": 2.81,
-      "learning_rate": 2.115896053682799e-05,
-      "loss": 0.2739,
+      "epoch": 2.8826539806105167,
+      "grad_norm": 0.23385298252105713,
+      "learning_rate": 2.006651529941059e-05,
+      "loss": 0.4071,
       "step": 79985
     },
     {
-      "epoch": 2.81,
-      "learning_rate": 2.11561456643391e-05,
-      "loss": 0.2873,
+      "epoch": 2.882834180271741,
+      "grad_norm": 0.24331295490264893,
+      "learning_rate": 2.0063654583797825e-05,
+      "loss": 0.3944,
       "step": 79990
     },
     {
-      "epoch": 2.81,
-      "learning_rate": 2.1153330841760115e-05,
-      "loss": 0.2966,
+      "epoch": 2.8830143799329657,
+      "grad_norm": 0.22254779934883118,
+      "learning_rate": 2.0060793935441818e-05,
+      "loss": 0.4021,
       "step": 79995
     },
     {
-      "epoch": 2.81,
-      "learning_rate": 2.1150516069127603e-05,
-      "loss": 0.2477,
+      "epoch": 2.8831945795941905,
+      "grad_norm": 0.24361281096935272,
+      "learning_rate": 2.0057933354381543e-05,
+      "loss": 0.4402,
       "step": 80000
     },
     {
-      "epoch": 2.81,
-      "eval_loss": 0.26057764887809753,
-      "eval_runtime": 10.5527,
-      "eval_samples_per_second": 9.476,
-      "eval_steps_per_second": 9.476,
+      "epoch": 2.8831945795941905,
+      "eval_loss": 0.43133535981178284,
+      "eval_runtime": 3.5314,
+      "eval_samples_per_second": 28.317,
+      "eval_steps_per_second": 7.079,
       "step": 80000
     },
     {
-      "epoch": 2.81,
-      "learning_rate": 2.114770134647811e-05,
-      "loss": 0.261,
+      "epoch": 2.8833747792554147,
+      "grad_norm": 0.19643712043762207,
+      "learning_rate": 2.0055072840655952e-05,
+      "loss": 0.3874,
       "step": 80005
     },
     {
-      "epoch": 2.81,
-      "learning_rate": 2.1144886673848164e-05,
-      "loss": 0.2834,
+      "epoch": 2.8835549789166395,
+      "grad_norm": 0.23182834684848785,
+      "learning_rate": 2.005221239430405e-05,
+      "loss": 0.3392,
       "step": 80010
     },
     {
-      "epoch": 2.82,
-      "learning_rate": 2.114207205127432e-05,
-      "loss": 0.2665,
+      "epoch": 2.883735178577864,
+      "grad_norm": 0.19430822134017944,
+      "learning_rate": 2.004935201536478e-05,
+      "loss": 0.4039,
       "step": 80015
     },
     {
-      "epoch": 2.82,
-      "learning_rate": 2.1139257478793137e-05,
-      "loss": 0.2911,
+      "epoch": 2.883915378239089,
+      "grad_norm": 0.2064819037914276,
+      "learning_rate": 2.0046491703877143e-05,
+      "loss": 0.3976,
       "step": 80020
     },
     {
-      "epoch": 2.82,
-      "learning_rate": 2.1136442956441155e-05,
-      "loss": 0.2903,
+      "epoch": 2.8840955779003137,
+      "grad_norm": 0.22902020812034607,
+      "learning_rate": 2.0043631459880103e-05,
+      "loss": 0.3979,
       "step": 80025
     },
     {
-      "epoch": 2.82,
-      "learning_rate": 2.1133628484254906e-05,
-      "loss": 0.2478,
+      "epoch": 2.8842757775615384,
+      "grad_norm": 0.26856574416160583,
+      "learning_rate": 2.004077128341261e-05,
+      "loss": 0.3935,
       "step": 80030
     },
     {
-      "epoch": 2.82,
-      "learning_rate": 2.1130814062270944e-05,
-      "loss": 0.2896,
+      "epoch": 2.8844559772227627,
+      "grad_norm": 0.2205887734889984,
+      "learning_rate": 2.0037911174513663e-05,
+      "loss": 0.3823,
       "step": 80035
     },
     {
-      "epoch": 2.82,
-      "learning_rate": 2.112799969052581e-05,
-      "loss": 0.2863,
+      "epoch": 2.8846361768839874,
+      "grad_norm": 0.17064271867275238,
+      "learning_rate": 2.003505113322221e-05,
+      "loss": 0.348,
       "step": 80040
     },
     {
-      "epoch": 2.82,
-      "learning_rate": 2.112518536905606e-05,
-      "loss": 0.2933,
+      "epoch": 2.884816376545212,
+      "grad_norm": 0.2369140088558197,
+      "learning_rate": 2.003219115957722e-05,
+      "loss": 0.3831,
       "step": 80045
     },
     {
-      "epoch": 2.82,
-      "learning_rate": 2.112237109789822e-05,
-      "loss": 0.2573,
+      "epoch": 2.8849965762064365,
+      "grad_norm": 0.19578613340854645,
+      "learning_rate": 2.0029331253617666e-05,
+      "loss": 0.3514,
       "step": 80050
     },
     {
-      "epoch": 2.82,
-      "learning_rate": 2.1119556877088824e-05,
-      "loss": 0.2869,
+      "epoch": 2.885176775867661,
+      "grad_norm": 0.22336263954639435,
+      "learning_rate": 2.0026471415382507e-05,
+      "loss": 0.4095,
       "step": 80055
     },
     {
-      "epoch": 2.82,
-      "learning_rate": 2.111674270666444e-05,
-      "loss": 0.2613,
+      "epoch": 2.885356975528886,
+      "grad_norm": 0.20325559377670288,
+      "learning_rate": 2.0023611644910716e-05,
+      "loss": 0.412,
       "step": 80060
     },
     {
-      "epoch": 2.82,
-      "learning_rate": 2.1113928586661583e-05,
-      "loss": 0.2761,
+      "epoch": 2.8855371751901107,
+      "grad_norm": 0.21424368023872375,
+      "learning_rate": 2.0020751942241255e-05,
+      "loss": 0.3845,
       "step": 80065
     },
     {
-      "epoch": 2.82,
-      "learning_rate": 2.111111451711681e-05,
-      "loss": 0.2667,
+      "epoch": 2.8857173748513354,
+      "grad_norm": 0.2809605300426483,
+      "learning_rate": 2.001789230741308e-05,
+      "loss": 0.3961,
       "step": 80070
     },
     {
-      "epoch": 2.82,
-      "learning_rate": 2.1108300498066638e-05,
-      "loss": 0.267,
+      "epoch": 2.88589757451256,
+      "grad_norm": 0.18413975834846497,
+      "learning_rate": 2.0015032740465165e-05,
+      "loss": 0.4107,
       "step": 80075
     },
     {
-      "epoch": 2.82,
-      "learning_rate": 2.1105486529547633e-05,
-      "loss": 0.2622,
+      "epoch": 2.8860777741737844,
+      "grad_norm": 0.28514033555984497,
+      "learning_rate": 2.0012173241436456e-05,
+      "loss": 0.3854,
       "step": 80080
     },
     {
-      "epoch": 2.82,
-      "learning_rate": 2.1102672611596318e-05,
-      "loss": 0.2929,
+      "epoch": 2.886257973835009,
+      "grad_norm": 0.27392441034317017,
+      "learning_rate": 2.0009313810365925e-05,
+      "loss": 0.3835,
       "step": 80085
     },
     {
-      "epoch": 2.82,
-      "learning_rate": 2.1099858744249226e-05,
-      "loss": 0.252,
+      "epoch": 2.886438173496234,
+      "grad_norm": 0.1907772570848465,
+      "learning_rate": 2.000645444729253e-05,
+      "loss": 0.3915,
       "step": 80090
     },
     {
-      "epoch": 2.82,
-      "learning_rate": 2.1097044927542892e-05,
-      "loss": 0.2905,
+      "epoch": 2.8866183731574586,
+      "grad_norm": 0.2062581181526184,
+      "learning_rate": 2.0003595152255218e-05,
+      "loss": 0.3623,
       "step": 80095
     },
     {
-      "epoch": 2.82,
-      "learning_rate": 2.109423116151387e-05,
-      "loss": 0.2739,
+      "epoch": 2.886798572818683,
+      "grad_norm": 0.20222362875938416,
+      "learning_rate": 2.000073592529296e-05,
+      "loss": 0.369,
       "step": 80100
     },
     {
-      "epoch": 2.82,
-      "learning_rate": 2.109141744619868e-05,
-      "loss": 0.2775,
+      "epoch": 2.8869787724799076,
+      "grad_norm": 0.21699479222297668,
+      "learning_rate": 1.9997876766444716e-05,
+      "loss": 0.3965,
       "step": 80105
     },
     {
-      "epoch": 2.82,
-      "learning_rate": 2.1088603781633847e-05,
-      "loss": 0.2554,
+      "epoch": 2.8871589721411324,
+      "grad_norm": 0.24619553983211517,
+      "learning_rate": 1.999501767574941e-05,
+      "loss": 0.4102,
       "step": 80110
     },
     {
-      "epoch": 2.82,
-      "learning_rate": 2.108579016785593e-05,
-      "loss": 0.2898,
+      "epoch": 2.887339171802357,
+      "grad_norm": 0.21754342317581177,
+      "learning_rate": 1.9992158653246042e-05,
+      "loss": 0.3494,
       "step": 80115
     },
     {
-      "epoch": 2.82,
-      "learning_rate": 2.108297660490144e-05,
-      "loss": 0.2458,
+      "epoch": 2.887519371463582,
+      "grad_norm": 0.22781917452812195,
+      "learning_rate": 1.9989299698973525e-05,
+      "loss": 0.3932,
       "step": 80120
     },
     {
-      "epoch": 2.82,
-      "learning_rate": 2.1080163092806927e-05,
-      "loss": 0.2766,
+      "epoch": 2.887699571124806,
+      "grad_norm": 0.17550964653491974,
+      "learning_rate": 1.9986440812970844e-05,
+      "loss": 0.3504,
       "step": 80125
     },
     {
-      "epoch": 2.82,
-      "learning_rate": 2.1077349631608894e-05,
-      "loss": 0.2797,
+      "epoch": 2.887879770786031,
+      "grad_norm": 0.20450954139232635,
+      "learning_rate": 1.998358199527693e-05,
+      "loss": 0.4168,
       "step": 80130
     },
     {
-      "epoch": 2.82,
-      "learning_rate": 2.107453622134391e-05,
-      "loss": 0.2439,
+      "epoch": 2.8880599704472556,
+      "grad_norm": 0.20474529266357422,
+      "learning_rate": 1.9980723245930737e-05,
+      "loss": 0.3699,
       "step": 80135
     },
     {
-      "epoch": 2.82,
-      "learning_rate": 2.107172286204848e-05,
-      "loss": 0.2781,
+      "epoch": 2.8882401701084803,
+      "grad_norm": 0.22211550176143646,
+      "learning_rate": 1.9977864564971225e-05,
+      "loss": 0.3711,
       "step": 80140
     },
     {
-      "epoch": 2.82,
-      "learning_rate": 2.1068909553759138e-05,
-      "loss": 0.2773,
+      "epoch": 2.8884203697697046,
+      "grad_norm": 0.2340330183506012,
+      "learning_rate": 1.9975005952437336e-05,
+      "loss": 0.3975,
       "step": 80145
     },
     {
-      "epoch": 2.82,
-      "learning_rate": 2.106609629651241e-05,
-      "loss": 0.2765,
+      "epoch": 2.8886005694309294,
+      "grad_norm": 0.211472749710083,
+      "learning_rate": 1.9972147408368008e-05,
+      "loss": 0.3676,
       "step": 80150
     },
     {
-      "epoch": 2.82,
-      "learning_rate": 2.106328309034484e-05,
-      "loss": 0.2547,
+      "epoch": 2.888780769092154,
+      "grad_norm": 0.17939423024654388,
+      "learning_rate": 1.9969288932802205e-05,
+      "loss": 0.3778,
       "step": 80155
     },
     {
-      "epoch": 2.82,
-      "learning_rate": 2.106046993529294e-05,
-      "loss": 0.2779,
+      "epoch": 2.888960968753379,
+      "grad_norm": 0.21015068888664246,
+      "learning_rate": 1.996643052577886e-05,
+      "loss": 0.3659,
       "step": 80160
     },
     {
-      "epoch": 2.82,
-      "learning_rate": 2.1057656831393246e-05,
-      "loss": 0.2747,
+      "epoch": 2.8891411684146036,
+      "grad_norm": 0.20937329530715942,
+      "learning_rate": 1.9963572187336935e-05,
+      "loss": 0.3988,
       "step": 80165
     },
     {
-      "epoch": 2.82,
-      "learning_rate": 2.1054843778682272e-05,
-      "loss": 0.2883,
+      "epoch": 2.8893213680758283,
+      "grad_norm": 0.17730148136615753,
+      "learning_rate": 1.996071391751536e-05,
+      "loss": 0.3907,
       "step": 80170
     },
     {
-      "epoch": 2.82,
-      "learning_rate": 2.105203077719655e-05,
-      "loss": 0.2727,
+      "epoch": 2.8895015677370526,
+      "grad_norm": 0.18675543367862701,
+      "learning_rate": 1.995785571635308e-05,
+      "loss": 0.4226,
       "step": 80175
     },
     {
-      "epoch": 2.82,
-      "learning_rate": 2.1049217826972617e-05,
-      "loss": 0.2562,
+      "epoch": 2.8896817673982773,
+      "grad_norm": 0.20549610257148743,
+      "learning_rate": 1.995499758388904e-05,
+      "loss": 0.3824,
       "step": 80180
     },
     {
-      "epoch": 2.82,
-      "learning_rate": 2.1046404928046983e-05,
-      "loss": 0.2653,
+      "epoch": 2.889861967059502,
+      "grad_norm": 0.21189923584461212,
+      "learning_rate": 1.9952139520162186e-05,
+      "loss": 0.402,
       "step": 80185
     },
     {
-      "epoch": 2.82,
-      "learning_rate": 2.1043592080456168e-05,
-      "loss": 0.2759,
+      "epoch": 2.8900421667207263,
+      "grad_norm": 0.24374066293239594,
+      "learning_rate": 1.9949281525211446e-05,
+      "loss": 0.4361,
       "step": 80190
     },
     {
-      "epoch": 2.82,
-      "learning_rate": 2.104077928423671e-05,
-      "loss": 0.2759,
+      "epoch": 2.890222366381951,
+      "grad_norm": 0.2334618866443634,
+      "learning_rate": 1.9946423599075774e-05,
+      "loss": 0.3988,
       "step": 80195
     },
     {
-      "epoch": 2.82,
-      "learning_rate": 2.103796653942512e-05,
-      "loss": 0.2461,
+      "epoch": 2.890402566043176,
+      "grad_norm": 0.25280308723449707,
+      "learning_rate": 1.9943565741794095e-05,
+      "loss": 0.4341,
       "step": 80200
     },
     {
-      "epoch": 2.82,
-      "learning_rate": 2.103515384605793e-05,
-      "loss": 0.262,
+      "epoch": 2.8905827657044005,
+      "grad_norm": 0.21369872987270355,
+      "learning_rate": 1.994070795340536e-05,
+      "loss": 0.365,
       "step": 80205
     },
     {
-      "epoch": 2.82,
-      "learning_rate": 2.1032341204171645e-05,
-      "loss": 0.2633,
+      "epoch": 2.8907629653656253,
+      "grad_norm": 0.23660530149936676,
+      "learning_rate": 1.9937850233948505e-05,
+      "loss": 0.3767,
       "step": 80210
     },
     {
-      "epoch": 2.82,
-      "learning_rate": 2.10295286138028e-05,
-      "loss": 0.275,
+      "epoch": 2.89094316502685,
+      "grad_norm": 0.2095620483160019,
+      "learning_rate": 1.9934992583462443e-05,
+      "loss": 0.3988,
       "step": 80215
     },
     {
-      "epoch": 2.82,
-      "learning_rate": 2.1026716074987912e-05,
-      "loss": 0.271,
+      "epoch": 2.8911233646880743,
+      "grad_norm": 0.22344286739826202,
+      "learning_rate": 1.9932135001986144e-05,
+      "loss": 0.4129,
       "step": 80220
     },
     {
-      "epoch": 2.82,
-      "learning_rate": 2.102390358776349e-05,
-      "loss": 0.2527,
+      "epoch": 2.891303564349299,
+      "grad_norm": 0.18746274709701538,
+      "learning_rate": 1.9929277489558517e-05,
+      "loss": 0.41,
       "step": 80225
     },
     {
-      "epoch": 2.82,
-      "learning_rate": 2.102109115216606e-05,
-      "loss": 0.2725,
+      "epoch": 2.8914837640105238,
+      "grad_norm": 0.17481876909732819,
+      "learning_rate": 1.9926420046218503e-05,
+      "loss": 0.3326,
       "step": 80230
     },
     {
-      "epoch": 2.82,
-      "learning_rate": 2.1018278768232146e-05,
-      "loss": 0.2702,
+      "epoch": 2.891663963671748,
+      "grad_norm": 0.185255765914917,
+      "learning_rate": 1.9923562672005033e-05,
+      "loss": 0.4074,
       "step": 80235
     },
     {
-      "epoch": 2.82,
-      "learning_rate": 2.101546643599826e-05,
-      "loss": 0.2668,
+      "epoch": 2.891844163332973,
+      "grad_norm": 0.16853371262550354,
+      "learning_rate": 1.9920705366957035e-05,
+      "loss": 0.3553,
       "step": 80240
     },
     {
-      "epoch": 2.82,
-      "learning_rate": 2.1012654155500904e-05,
-      "loss": 0.2583,
+      "epoch": 2.8920243629941975,
+      "grad_norm": 0.25939464569091797,
+      "learning_rate": 1.991784813111345e-05,
+      "loss": 0.3817,
       "step": 80245
     },
     {
-      "epoch": 2.82,
-      "learning_rate": 2.1009841926776616e-05,
-      "loss": 0.2684,
+      "epoch": 2.8922045626554223,
+      "grad_norm": 0.20100079476833344,
+      "learning_rate": 1.9914990964513196e-05,
+      "loss": 0.3873,
       "step": 80250
     },
     {
-      "epoch": 2.82,
-      "learning_rate": 2.1007029749861897e-05,
-      "loss": 0.2692,
+      "epoch": 2.892384762316647,
+      "grad_norm": 0.16807907819747925,
+      "learning_rate": 1.9912133867195203e-05,
+      "loss": 0.387,
       "step": 80255
     },
     {
-      "epoch": 2.82,
-      "learning_rate": 2.1004217624793272e-05,
-      "loss": 0.2727,
+      "epoch": 2.8925649619778717,
+      "grad_norm": 0.2342177778482437,
+      "learning_rate": 1.990927683919841e-05,
+      "loss": 0.3891,
       "step": 80260
     },
     {
-      "epoch": 2.82,
-      "learning_rate": 2.1001405551607234e-05,
-      "loss": 0.2685,
+      "epoch": 2.892745161639096,
+      "grad_norm": 0.27248871326446533,
+      "learning_rate": 1.9906419880561725e-05,
+      "loss": 0.4153,
       "step": 80265
     },
     {
-      "epoch": 2.82,
-      "learning_rate": 2.0998593530340322e-05,
-      "loss": 0.2716,
+      "epoch": 2.8929253613003207,
+      "grad_norm": 0.33067587018013,
+      "learning_rate": 1.9903562991324088e-05,
+      "loss": 0.4321,
       "step": 80270
     },
     {
-      "epoch": 2.82,
-      "learning_rate": 2.0995781561029035e-05,
-      "loss": 0.2605,
+      "epoch": 2.8931055609615455,
+      "grad_norm": 0.22115959227085114,
+      "learning_rate": 1.990070617152442e-05,
+      "loss": 0.3908,
       "step": 80275
     },
     {
-      "epoch": 2.82,
-      "learning_rate": 2.099296964370988e-05,
-      "loss": 0.2623,
+      "epoch": 2.8932857606227698,
+      "grad_norm": 0.23916247487068176,
+      "learning_rate": 1.9897849421201636e-05,
+      "loss": 0.4196,
       "step": 80280
     },
     {
-      "epoch": 2.82,
-      "learning_rate": 2.0990157778419375e-05,
-      "loss": 0.2637,
+      "epoch": 2.8934659602839945,
+      "grad_norm": 0.2263944149017334,
+      "learning_rate": 1.9894992740394674e-05,
+      "loss": 0.3994,
       "step": 80285
     },
     {
-      "epoch": 2.82,
-      "learning_rate": 2.0987345965194028e-05,
-      "loss": 0.2739,
+      "epoch": 2.8936461599452192,
+      "grad_norm": 0.2417251467704773,
+      "learning_rate": 1.9892136129142452e-05,
+      "loss": 0.3894,
       "step": 80290
     },
     {
-      "epoch": 2.83,
-      "learning_rate": 2.0984534204070354e-05,
-      "loss": 0.2722,
+      "epoch": 2.893826359606444,
+      "grad_norm": 0.18903349339962006,
+      "learning_rate": 1.988927958748387e-05,
+      "loss": 0.3851,
       "step": 80295
     },
     {
-      "epoch": 2.83,
-      "learning_rate": 2.098172249508486e-05,
-      "loss": 0.2858,
+      "epoch": 2.8940065592676687,
+      "grad_norm": 0.23545511066913605,
+      "learning_rate": 1.988642311545788e-05,
+      "loss": 0.3862,
       "step": 80300
     },
     {
-      "epoch": 2.83,
-      "learning_rate": 2.0978910838274038e-05,
-      "loss": 0.262,
+      "epoch": 2.8941867589288934,
+      "grad_norm": 0.27978673577308655,
+      "learning_rate": 1.9883566713103368e-05,
+      "loss": 0.4131,
       "step": 80305
     },
     {
-      "epoch": 2.83,
-      "learning_rate": 2.0976099233674413e-05,
-      "loss": 0.2545,
+      "epoch": 2.8943669585901177,
+      "grad_norm": 0.171412855386734,
+      "learning_rate": 1.988071038045928e-05,
+      "loss": 0.3753,
       "step": 80310
     },
     {
-      "epoch": 2.83,
-      "learning_rate": 2.0973287681322497e-05,
-      "loss": 0.2886,
+      "epoch": 2.8945471582513425,
+      "grad_norm": 0.18040184676647186,
+      "learning_rate": 1.987785411756453e-05,
+      "loss": 0.3975,
       "step": 80315
     },
     {
-      "epoch": 2.83,
-      "learning_rate": 2.0970476181254784e-05,
-      "loss": 0.2829,
+      "epoch": 2.894727357912567,
+      "grad_norm": 0.2326379269361496,
+      "learning_rate": 1.9874997924458007e-05,
+      "loss": 0.4205,
       "step": 80320
     },
     {
-      "epoch": 2.83,
-      "learning_rate": 2.0967664733507768e-05,
-      "loss": 0.2694,
+      "epoch": 2.894907557573792,
+      "grad_norm": 0.24374447762966156,
+      "learning_rate": 1.987214180117866e-05,
+      "loss": 0.4147,
       "step": 80325
     },
     {
-      "epoch": 2.83,
-      "learning_rate": 2.0964853338117983e-05,
-      "loss": 0.2816,
+      "epoch": 2.895087757235016,
+      "grad_norm": 0.20812088251113892,
+      "learning_rate": 1.9869285747765387e-05,
+      "loss": 0.3989,
       "step": 80330
     },
     {
-      "epoch": 2.83,
-      "learning_rate": 2.0962041995121912e-05,
-      "loss": 0.28,
+      "epoch": 2.895267956896241,
+      "grad_norm": 0.22509217262268066,
+      "learning_rate": 1.9866429764257092e-05,
+      "loss": 0.3967,
       "step": 80335
     },
     {
-      "epoch": 2.83,
-      "learning_rate": 2.0959230704556066e-05,
-      "loss": 0.2756,
+      "epoch": 2.8954481565574657,
+      "grad_norm": 0.244933620095253,
+      "learning_rate": 1.9863573850692706e-05,
+      "loss": 0.3987,
       "step": 80340
     },
     {
-      "epoch": 2.83,
-      "learning_rate": 2.0956419466456945e-05,
-      "loss": 0.2736,
+      "epoch": 2.8956283562186904,
+      "grad_norm": 0.17772509157657623,
+      "learning_rate": 1.986071800711113e-05,
+      "loss": 0.4069,
       "step": 80345
     },
     {
-      "epoch": 2.83,
-      "learning_rate": 2.095360828086106e-05,
-      "loss": 0.2538,
+      "epoch": 2.895808555879915,
+      "grad_norm": 0.21300829946994781,
+      "learning_rate": 1.9857862233551274e-05,
+      "loss": 0.405,
       "step": 80350
     },
     {
-      "epoch": 2.83,
-      "learning_rate": 2.0950797147804903e-05,
-      "loss": 0.2854,
+      "epoch": 2.8959887555411394,
+      "grad_norm": 0.20396198332309723,
+      "learning_rate": 1.9855006530052055e-05,
+      "loss": 0.3692,
       "step": 80355
     },
     {
-      "epoch": 2.83,
-      "learning_rate": 2.0947986067324962e-05,
-      "loss": 0.2723,
+      "epoch": 2.896168955202364,
+      "grad_norm": 0.2265167236328125,
+      "learning_rate": 1.985215089665237e-05,
+      "loss": 0.387,
       "step": 80360
     },
     {
-      "epoch": 2.83,
-      "learning_rate": 2.0945175039457765e-05,
-      "loss": 0.2605,
+      "epoch": 2.896349154863589,
+      "grad_norm": 0.2407812774181366,
+      "learning_rate": 1.9849295333391134e-05,
+      "loss": 0.3945,
       "step": 80365
     },
     {
-      "epoch": 2.83,
-      "learning_rate": 2.0942364064239795e-05,
-      "loss": 0.2816,
+      "epoch": 2.8965293545248136,
+      "grad_norm": 0.19000792503356934,
+      "learning_rate": 1.984643984030726e-05,
+      "loss": 0.3897,
       "step": 80370
     },
     {
-      "epoch": 2.83,
-      "learning_rate": 2.093955314170756e-05,
-      "loss": 0.27,
+      "epoch": 2.896709554186038,
+      "grad_norm": 0.21029935777187347,
+      "learning_rate": 1.9843584417439633e-05,
+      "loss": 0.4148,
       "step": 80375
     },
     {
-      "epoch": 2.83,
-      "learning_rate": 2.0936742271897536e-05,
-      "loss": 0.2619,
+      "epoch": 2.8968897538472627,
+      "grad_norm": 0.21043239533901215,
+      "learning_rate": 1.9840729064827173e-05,
+      "loss": 0.3786,
       "step": 80380
     },
     {
-      "epoch": 2.83,
-      "learning_rate": 2.0933931454846245e-05,
-      "loss": 0.2715,
+      "epoch": 2.8970699535084874,
+      "grad_norm": 0.24211741983890533,
+      "learning_rate": 1.983787378250878e-05,
+      "loss": 0.4162,
       "step": 80385
     },
     {
-      "epoch": 2.83,
-      "learning_rate": 2.093112069059017e-05,
-      "loss": 0.2664,
+      "epoch": 2.897250153169712,
+      "grad_norm": 0.20542925596237183,
+      "learning_rate": 1.9835018570523363e-05,
+      "loss": 0.4269,
       "step": 80390
     },
     {
-      "epoch": 2.83,
-      "learning_rate": 2.0928309979165824e-05,
-      "loss": 0.2681,
+      "epoch": 2.897430352830937,
+      "grad_norm": 0.22718103229999542,
+      "learning_rate": 1.983216342890982e-05,
+      "loss": 0.3968,
       "step": 80395
     },
     {
-      "epoch": 2.83,
-      "learning_rate": 2.0925499320609674e-05,
-      "loss": 0.2559,
+      "epoch": 2.897610552492161,
+      "grad_norm": 0.20570969581604004,
+      "learning_rate": 1.9829308357707037e-05,
+      "loss": 0.4045,
       "step": 80400
     },
     {
-      "epoch": 2.83,
-      "learning_rate": 2.092268871495824e-05,
-      "loss": 0.2453,
+      "epoch": 2.897790752153386,
+      "grad_norm": 0.26240652799606323,
+      "learning_rate": 1.982645335695394e-05,
+      "loss": 0.3815,
       "step": 80405
     },
     {
-      "epoch": 2.83,
-      "learning_rate": 2.0919878162248008e-05,
-      "loss": 0.2544,
+      "epoch": 2.8979709518146106,
+      "grad_norm": 0.23519177734851837,
+      "learning_rate": 1.982359842668941e-05,
+      "loss": 0.4191,
       "step": 80410
     },
     {
-      "epoch": 2.83,
-      "learning_rate": 2.091706766251546e-05,
-      "loss": 0.2612,
+      "epoch": 2.8981511514758354,
+      "grad_norm": 0.23879577219486237,
+      "learning_rate": 1.9820743566952348e-05,
+      "loss": 0.4106,
       "step": 80415
     },
     {
-      "epoch": 2.83,
-      "learning_rate": 2.09142572157971e-05,
-      "loss": 0.2587,
+      "epoch": 2.8983313511370596,
+      "grad_norm": 0.22001482546329498,
+      "learning_rate": 1.981788877778165e-05,
+      "loss": 0.3793,
       "step": 80420
     },
     {
-      "epoch": 2.83,
-      "learning_rate": 2.091144682212942e-05,
-      "loss": 0.2638,
+      "epoch": 2.8985115507982844,
+      "grad_norm": 0.1924242228269577,
+      "learning_rate": 1.9815034059216214e-05,
+      "loss": 0.4246,
       "step": 80425
     },
     {
-      "epoch": 2.83,
-      "learning_rate": 2.090863648154891e-05,
-      "loss": 0.2679,
+      "epoch": 2.898691750459509,
+      "grad_norm": 0.2032802700996399,
+      "learning_rate": 1.981217941129494e-05,
+      "loss": 0.3554,
       "step": 80430
     },
     {
-      "epoch": 2.83,
-      "learning_rate": 2.0905826194092064e-05,
-      "loss": 0.2748,
+      "epoch": 2.898871950120734,
+      "grad_norm": 0.2488231360912323,
+      "learning_rate": 1.9809324834056713e-05,
+      "loss": 0.4157,
       "step": 80435
     },
     {
-      "epoch": 2.83,
-      "learning_rate": 2.090301595979535e-05,
-      "loss": 0.2592,
+      "epoch": 2.8990521497819586,
+      "grad_norm": 0.20548661053180695,
+      "learning_rate": 1.980647032754043e-05,
+      "loss": 0.3964,
       "step": 80440
     },
     {
-      "epoch": 2.83,
-      "learning_rate": 2.0900205778695285e-05,
-      "loss": 0.2609,
+      "epoch": 2.8992323494431833,
+      "grad_norm": 0.20527927577495575,
+      "learning_rate": 1.9803615891784987e-05,
+      "loss": 0.3843,
       "step": 80445
     },
     {
-      "epoch": 2.83,
-      "learning_rate": 2.0897395650828343e-05,
-      "loss": 0.2677,
+      "epoch": 2.8994125491044076,
+      "grad_norm": 0.20890048146247864,
+      "learning_rate": 1.980076152682927e-05,
+      "loss": 0.418,
       "step": 80450
     },
     {
-      "epoch": 2.83,
-      "learning_rate": 2.0894585576231012e-05,
-      "loss": 0.2443,
+      "epoch": 2.8995927487656323,
+      "grad_norm": 0.23384051024913788,
+      "learning_rate": 1.9797907232712166e-05,
+      "loss": 0.368,
       "step": 80455
     },
     {
-      "epoch": 2.83,
-      "learning_rate": 2.0891775554939792e-05,
-      "loss": 0.2822,
+      "epoch": 2.899772948426857,
+      "grad_norm": 0.20361943542957306,
+      "learning_rate": 1.9795053009472574e-05,
+      "loss": 0.4239,
       "step": 80460
     },
     {
-      "epoch": 2.83,
-      "learning_rate": 2.0888965586991156e-05,
-      "loss": 0.2614,
+      "epoch": 2.8999531480880814,
+      "grad_norm": 0.21155259013175964,
+      "learning_rate": 1.9792198857149375e-05,
+      "loss": 0.4109,
       "step": 80465
     },
     {
-      "epoch": 2.83,
-      "learning_rate": 2.0886155672421588e-05,
-      "loss": 0.2685,
+      "epoch": 2.900133347749306,
+      "grad_norm": 0.24116067588329315,
+      "learning_rate": 1.978934477578146e-05,
+      "loss": 0.4045,
       "step": 80470
     },
     {
-      "epoch": 2.83,
-      "learning_rate": 2.0883345811267574e-05,
-      "loss": 0.2603,
+      "epoch": 2.900313547410531,
+      "grad_norm": 0.20510059595108032,
+      "learning_rate": 1.9786490765407713e-05,
+      "loss": 0.4006,
       "step": 80475
     },
     {
-      "epoch": 2.83,
-      "learning_rate": 2.0880536003565605e-05,
-      "loss": 0.2708,
+      "epoch": 2.9004937470717556,
+      "grad_norm": 0.2083190232515335,
+      "learning_rate": 1.9783636826067015e-05,
+      "loss": 0.3923,
       "step": 80480
     },
     {
-      "epoch": 2.83,
-      "learning_rate": 2.087772624935217e-05,
-      "loss": 0.2762,
+      "epoch": 2.9006739467329803,
+      "grad_norm": 0.22528444230556488,
+      "learning_rate": 1.9780782957798263e-05,
+      "loss": 0.3764,
       "step": 80485
     },
     {
-      "epoch": 2.83,
-      "learning_rate": 2.0874916548663738e-05,
-      "loss": 0.2803,
+      "epoch": 2.900854146394205,
+      "grad_norm": 0.22572170197963715,
+      "learning_rate": 1.9777929160640325e-05,
+      "loss": 0.3875,
       "step": 80490
     },
     {
-      "epoch": 2.83,
-      "learning_rate": 2.0872106901536786e-05,
-      "loss": 0.2708,
+      "epoch": 2.9010343460554293,
+      "grad_norm": 0.1767890602350235,
+      "learning_rate": 1.97750754346321e-05,
+      "loss": 0.3535,
       "step": 80495
     },
     {
-      "epoch": 2.83,
-      "learning_rate": 2.086929730800782e-05,
-      "loss": 0.2658,
+      "epoch": 2.901214545716654,
+      "grad_norm": 0.24823251366615295,
+      "learning_rate": 1.977222177981246e-05,
+      "loss": 0.4085,
       "step": 80500
     },
     {
-      "epoch": 2.83,
-      "eval_loss": 0.26011475920677185,
-      "eval_runtime": 10.5627,
-      "eval_samples_per_second": 9.467,
-      "eval_steps_per_second": 9.467,
+      "epoch": 2.901214545716654,
+      "eval_loss": 0.43115851283073425,
+      "eval_runtime": 3.533,
+      "eval_samples_per_second": 28.304,
+      "eval_steps_per_second": 7.076,
       "step": 80500
     },
     {
-      "epoch": 2.83,
-      "learning_rate": 2.0866487768113297e-05,
-      "loss": 0.2571,
+      "epoch": 2.901394745377879,
+      "grad_norm": 0.284284383058548,
+      "learning_rate": 1.9769368196220275e-05,
+      "loss": 0.4208,
       "step": 80505
     },
     {
-      "epoch": 2.83,
-      "learning_rate": 2.0863678281889713e-05,
-      "loss": 0.2468,
+      "epoch": 2.901574945039103,
+      "grad_norm": 0.22723108530044556,
+      "learning_rate": 1.976651468389445e-05,
+      "loss": 0.415,
       "step": 80510
     },
     {
-      "epoch": 2.83,
-      "learning_rate": 2.0860868849373528e-05,
-      "loss": 0.2418,
+      "epoch": 2.901755144700328,
+      "grad_norm": 0.2633325159549713,
+      "learning_rate": 1.9763661242873845e-05,
+      "loss": 0.4283,
       "step": 80515
     },
     {
-      "epoch": 2.83,
-      "learning_rate": 2.0858059470601243e-05,
-      "loss": 0.242,
+      "epoch": 2.9019353443615525,
+      "grad_norm": 0.2002272605895996,
+      "learning_rate": 1.9760807873197336e-05,
+      "loss": 0.3805,
       "step": 80520
     },
     {
-      "epoch": 2.83,
-      "learning_rate": 2.085525014560933e-05,
-      "loss": 0.2763,
+      "epoch": 2.9021155440227773,
+      "grad_norm": 0.1985533982515335,
+      "learning_rate": 1.975795457490381e-05,
+      "loss": 0.3822,
       "step": 80525
     },
     {
-      "epoch": 2.83,
-      "learning_rate": 2.0852440874434256e-05,
-      "loss": 0.2448,
+      "epoch": 2.902295743684002,
+      "grad_norm": 0.18595746159553528,
+      "learning_rate": 1.9755101348032136e-05,
+      "loss": 0.3732,
       "step": 80530
     },
     {
-      "epoch": 2.83,
-      "learning_rate": 2.0849631657112496e-05,
-      "loss": 0.2852,
+      "epoch": 2.9024759433452267,
+      "grad_norm": 0.22532545030117035,
+      "learning_rate": 1.975224819262119e-05,
+      "loss": 0.3716,
       "step": 80535
     },
     {
-      "epoch": 2.83,
-      "learning_rate": 2.0846822493680548e-05,
-      "loss": 0.2857,
+      "epoch": 2.902656143006451,
+      "grad_norm": 0.20030663907527924,
+      "learning_rate": 1.974939510870985e-05,
+      "loss": 0.3823,
       "step": 80540
     },
     {
-      "epoch": 2.83,
-      "learning_rate": 2.084401338417487e-05,
-      "loss": 0.3082,
+      "epoch": 2.9028363426676758,
+      "grad_norm": 0.2480957806110382,
+      "learning_rate": 1.974654209633698e-05,
+      "loss": 0.3838,
       "step": 80545
     },
     {
-      "epoch": 2.83,
-      "learning_rate": 2.0841204328631942e-05,
-      "loss": 0.2949,
+      "epoch": 2.9030165423289005,
+      "grad_norm": 0.23294273018836975,
+      "learning_rate": 1.9743689155541458e-05,
+      "loss": 0.412,
       "step": 80550
     },
     {
-      "epoch": 2.83,
-      "learning_rate": 2.0838395327088223e-05,
-      "loss": 0.2487,
+      "epoch": 2.903196741990125,
+      "grad_norm": 0.2072293609380722,
+      "learning_rate": 1.974083628636216e-05,
+      "loss": 0.4121,
       "step": 80555
     },
     {
-      "epoch": 2.83,
-      "learning_rate": 2.0835586379580203e-05,
-      "loss": 0.2753,
+      "epoch": 2.9033769416513495,
+      "grad_norm": 0.20918940007686615,
+      "learning_rate": 1.973798348883794e-05,
+      "loss": 0.3944,
       "step": 80560
     },
     {
-      "epoch": 2.83,
-      "learning_rate": 2.0832777486144357e-05,
-      "loss": 0.275,
+      "epoch": 2.9035571413125743,
+      "grad_norm": 0.24100269377231598,
+      "learning_rate": 1.973513076300768e-05,
+      "loss": 0.416,
       "step": 80565
     },
     {
-      "epoch": 2.83,
-      "learning_rate": 2.0829968646817135e-05,
-      "loss": 0.277,
+      "epoch": 2.903737340973799,
+      "grad_norm": 0.1925951987504959,
+      "learning_rate": 1.9732278108910243e-05,
+      "loss": 0.3716,
       "step": 80570
     },
     {
-      "epoch": 2.83,
-      "learning_rate": 2.0827159861635034e-05,
-      "loss": 0.2595,
+      "epoch": 2.9039175406350237,
+      "grad_norm": 0.21889275312423706,
+      "learning_rate": 1.97294255265845e-05,
+      "loss": 0.4215,
       "step": 80575
     },
     {
-      "epoch": 2.84,
-      "learning_rate": 2.0824351130634513e-05,
-      "loss": 0.2627,
+      "epoch": 2.9040977402962485,
+      "grad_norm": 0.20681817829608917,
+      "learning_rate": 1.972657301606932e-05,
+      "loss": 0.4319,
       "step": 80580
     },
     {
-      "epoch": 2.84,
-      "learning_rate": 2.0821542453852032e-05,
-      "loss": 0.2589,
+      "epoch": 2.9042779399574727,
+      "grad_norm": 0.23390960693359375,
+      "learning_rate": 1.9723720577403546e-05,
+      "loss": 0.4255,
       "step": 80585
     },
     {
-      "epoch": 2.84,
-      "learning_rate": 2.081873383132407e-05,
-      "loss": 0.2754,
+      "epoch": 2.9044581396186975,
+      "grad_norm": 0.21536414325237274,
+      "learning_rate": 1.9720868210626067e-05,
+      "loss": 0.3939,
       "step": 80590
     },
     {
-      "epoch": 2.84,
-      "learning_rate": 2.08159252630871e-05,
-      "loss": 0.2601,
+      "epoch": 2.904638339279922,
+      "grad_norm": 0.245896577835083,
+      "learning_rate": 1.9718015915775743e-05,
+      "loss": 0.3782,
       "step": 80595
     },
     {
-      "epoch": 2.84,
-      "learning_rate": 2.081311674917758e-05,
-      "loss": 0.2757,
+      "epoch": 2.904818538941147,
+      "grad_norm": 0.20185402035713196,
+      "learning_rate": 1.9715163692891416e-05,
+      "loss": 0.4153,
       "step": 80600
     },
     {
-      "epoch": 2.84,
-      "learning_rate": 2.0810308289631984e-05,
-      "loss": 0.2631,
+      "epoch": 2.9049987386023712,
+      "grad_norm": 0.19402766227722168,
+      "learning_rate": 1.971231154201197e-05,
+      "loss": 0.4159,
       "step": 80605
     },
     {
-      "epoch": 2.84,
-      "learning_rate": 2.0807499884486763e-05,
-      "loss": 0.2876,
+      "epoch": 2.905178938263596,
+      "grad_norm": 0.23426468670368195,
+      "learning_rate": 1.9709459463176243e-05,
+      "loss": 0.3907,
       "step": 80610
     },
     {
-      "epoch": 2.84,
-      "learning_rate": 2.0804691533778398e-05,
-      "loss": 0.2869,
+      "epoch": 2.9053591379248207,
+      "grad_norm": 0.2931303381919861,
+      "learning_rate": 1.9706607456423122e-05,
+      "loss": 0.432,
       "step": 80615
     },
     {
-      "epoch": 2.84,
-      "learning_rate": 2.0801883237543355e-05,
-      "loss": 0.2653,
+      "epoch": 2.9055393375860454,
+      "grad_norm": 0.25184187293052673,
+      "learning_rate": 1.9703755521791445e-05,
+      "loss": 0.3872,
       "step": 80620
     },
     {
-      "epoch": 2.84,
-      "learning_rate": 2.0799074995818087e-05,
-      "loss": 0.2704,
+      "epoch": 2.90571953724727,
+      "grad_norm": 0.2202758938074112,
+      "learning_rate": 1.970090365932007e-05,
+      "loss": 0.359,
       "step": 80625
     },
     {
-      "epoch": 2.84,
-      "learning_rate": 2.0796266808639058e-05,
-      "loss": 0.275,
+      "epoch": 2.9058997369084945,
+      "grad_norm": 0.2390233725309372,
+      "learning_rate": 1.969805186904786e-05,
+      "loss": 0.3898,
       "step": 80630
     },
     {
-      "epoch": 2.84,
-      "learning_rate": 2.0793458676042737e-05,
-      "loss": 0.252,
+      "epoch": 2.906079936569719,
+      "grad_norm": 0.1913604885339737,
+      "learning_rate": 1.9695200151013666e-05,
+      "loss": 0.3645,
       "step": 80635
     },
     {
-      "epoch": 2.84,
-      "learning_rate": 2.079065059806558e-05,
-      "loss": 0.2659,
+      "epoch": 2.906260136230944,
+      "grad_norm": 0.22707746922969818,
+      "learning_rate": 1.9692348505256335e-05,
+      "loss": 0.3945,
       "step": 80640
     },
     {
-      "epoch": 2.84,
-      "learning_rate": 2.0787842574744058e-05,
-      "loss": 0.265,
+      "epoch": 2.9064403358921687,
+      "grad_norm": 0.2409844696521759,
+      "learning_rate": 1.968949693181474e-05,
+      "loss": 0.3892,
       "step": 80645
     },
     {
-      "epoch": 2.84,
-      "learning_rate": 2.0785034606114613e-05,
-      "loss": 0.2435,
+      "epoch": 2.906620535553393,
+      "grad_norm": 0.21036100387573242,
+      "learning_rate": 1.968664543072771e-05,
+      "loss": 0.3957,
       "step": 80650
     },
     {
-      "epoch": 2.84,
-      "learning_rate": 2.078222669221373e-05,
-      "loss": 0.2861,
+      "epoch": 2.9068007352146177,
+      "grad_norm": 0.2081831395626068,
+      "learning_rate": 1.9683794002034115e-05,
+      "loss": 0.3975,
       "step": 80655
     },
     {
-      "epoch": 2.84,
-      "learning_rate": 2.0779418833077848e-05,
-      "loss": 0.2312,
+      "epoch": 2.9069809348758424,
+      "grad_norm": 0.17619097232818604,
+      "learning_rate": 1.96809426457728e-05,
+      "loss": 0.3864,
       "step": 80660
     },
     {
-      "epoch": 2.84,
-      "learning_rate": 2.077661102874343e-05,
-      "loss": 0.2818,
+      "epoch": 2.907161134537067,
+      "grad_norm": 0.25311923027038574,
+      "learning_rate": 1.9678091361982602e-05,
+      "loss": 0.3982,
       "step": 80665
     },
     {
-      "epoch": 2.84,
-      "learning_rate": 2.0773803279246924e-05,
-      "loss": 0.2593,
+      "epoch": 2.907341334198292,
+      "grad_norm": 0.25830528140068054,
+      "learning_rate": 1.967524015070239e-05,
+      "loss": 0.4033,
       "step": 80670
     },
     {
-      "epoch": 2.84,
-      "learning_rate": 2.0770995584624815e-05,
-      "loss": 0.259,
+      "epoch": 2.9075215338595166,
+      "grad_norm": 0.20592638850212097,
+      "learning_rate": 1.967238901197099e-05,
+      "loss": 0.3828,
       "step": 80675
     },
     {
-      "epoch": 2.84,
-      "learning_rate": 2.0768187944913537e-05,
-      "loss": 0.2681,
+      "epoch": 2.907701733520741,
+      "grad_norm": 0.19659774005413055,
+      "learning_rate": 1.9669537945827265e-05,
+      "loss": 0.3768,
       "step": 80680
     },
     {
-      "epoch": 2.84,
-      "learning_rate": 2.076538036014955e-05,
-      "loss": 0.2568,
+      "epoch": 2.9078819331819656,
+      "grad_norm": 0.24102912843227386,
+      "learning_rate": 1.9666686952310057e-05,
+      "loss": 0.4266,
       "step": 80685
     },
     {
-      "epoch": 2.84,
-      "learning_rate": 2.07625728303693e-05,
-      "loss": 0.2833,
+      "epoch": 2.9080621328431904,
+      "grad_norm": 0.18135298788547516,
+      "learning_rate": 1.9663836031458195e-05,
+      "loss": 0.4413,
       "step": 80690
     },
     {
-      "epoch": 2.84,
-      "learning_rate": 2.075976535560925e-05,
-      "loss": 0.2708,
+      "epoch": 2.9082423325044147,
+      "grad_norm": 0.2321069985628128,
+      "learning_rate": 1.9660985183310543e-05,
+      "loss": 0.4058,
       "step": 80695
     },
     {
-      "epoch": 2.84,
-      "learning_rate": 2.075695793590586e-05,
-      "loss": 0.2695,
+      "epoch": 2.9084225321656394,
+      "grad_norm": 0.21384575963020325,
+      "learning_rate": 1.9658134407905935e-05,
+      "loss": 0.382,
       "step": 80700
     },
     {
-      "epoch": 2.84,
-      "learning_rate": 2.0754150571295563e-05,
-      "loss": 0.2673,
+      "epoch": 2.908602731826864,
+      "grad_norm": 0.2350618541240692,
+      "learning_rate": 1.9655283705283205e-05,
+      "loss": 0.4001,
       "step": 80705
     },
     {
-      "epoch": 2.84,
-      "learning_rate": 2.0751343261814838e-05,
-      "loss": 0.2859,
+      "epoch": 2.908782931488089,
+      "grad_norm": 0.19745108485221863,
+      "learning_rate": 1.9652433075481202e-05,
+      "loss": 0.425,
       "step": 80710
     },
     {
-      "epoch": 2.84,
-      "learning_rate": 2.0748536007500116e-05,
-      "loss": 0.2642,
+      "epoch": 2.9089631311493136,
+      "grad_norm": 0.21076776087284088,
+      "learning_rate": 1.9649582518538757e-05,
+      "loss": 0.4067,
       "step": 80715
     },
     {
-      "epoch": 2.84,
-      "learning_rate": 2.074572880838785e-05,
-      "loss": 0.2722,
+      "epoch": 2.9091433308105383,
+      "grad_norm": 0.20827296376228333,
+      "learning_rate": 1.9646732034494726e-05,
+      "loss": 0.3839,
       "step": 80720
     },
     {
-      "epoch": 2.84,
-      "learning_rate": 2.0742921664514486e-05,
-      "loss": 0.2735,
+      "epoch": 2.9093235304717626,
+      "grad_norm": 0.18359017372131348,
+      "learning_rate": 1.964388162338793e-05,
+      "loss": 0.3977,
       "step": 80725
     },
     {
-      "epoch": 2.84,
-      "learning_rate": 2.0740114575916484e-05,
-      "loss": 0.2475,
+      "epoch": 2.9095037301329874,
+      "grad_norm": 0.23850063979625702,
+      "learning_rate": 1.9641031285257205e-05,
+      "loss": 0.4165,
       "step": 80730
     },
     {
-      "epoch": 2.84,
-      "learning_rate": 2.073730754263029e-05,
-      "loss": 0.2844,
+      "epoch": 2.909683929794212,
+      "grad_norm": 0.2121899276971817,
+      "learning_rate": 1.96381810201414e-05,
+      "loss": 0.3906,
       "step": 80735
     },
     {
-      "epoch": 2.84,
-      "learning_rate": 2.073450056469235e-05,
-      "loss": 0.2879,
+      "epoch": 2.9098641294554364,
+      "grad_norm": 0.26047274470329285,
+      "learning_rate": 1.9635330828079335e-05,
+      "loss": 0.3992,
       "step": 80740
     },
     {
-      "epoch": 2.84,
-      "learning_rate": 2.07316936421391e-05,
-      "loss": 0.2833,
+      "epoch": 2.910044329116661,
+      "grad_norm": 0.16215106844902039,
+      "learning_rate": 1.9632480709109845e-05,
+      "loss": 0.4113,
       "step": 80745
     },
     {
-      "epoch": 2.84,
-      "learning_rate": 2.0728886775006997e-05,
-      "loss": 0.2413,
+      "epoch": 2.910224528777886,
+      "grad_norm": 0.2457340657711029,
+      "learning_rate": 1.9629630663271776e-05,
+      "loss": 0.4222,
       "step": 80750
     },
     {
-      "epoch": 2.84,
-      "learning_rate": 2.072607996333249e-05,
-      "loss": 0.2701,
+      "epoch": 2.9104047284391106,
+      "grad_norm": 0.24519392848014832,
+      "learning_rate": 1.962678069060394e-05,
+      "loss": 0.4321,
       "step": 80755
     },
     {
-      "epoch": 2.84,
-      "learning_rate": 2.072327320715202e-05,
-      "loss": 0.2796,
+      "epoch": 2.9105849281003353,
+      "grad_norm": 0.169399693608284,
+      "learning_rate": 1.9623930791145184e-05,
+      "loss": 0.3873,
       "step": 80760
     },
     {
-      "epoch": 2.84,
-      "learning_rate": 2.072046650650202e-05,
-      "loss": 0.249,
+      "epoch": 2.91076512776156,
+      "grad_norm": 0.20469214022159576,
+      "learning_rate": 1.9621080964934326e-05,
+      "loss": 0.3688,
       "step": 80765
     },
     {
-      "epoch": 2.84,
-      "learning_rate": 2.071765986141895e-05,
-      "loss": 0.2712,
+      "epoch": 2.9109453274227843,
+      "grad_norm": 0.21101103723049164,
+      "learning_rate": 1.9618231212010195e-05,
+      "loss": 0.4403,
       "step": 80770
     },
     {
-      "epoch": 2.84,
-      "learning_rate": 2.0714853271939235e-05,
-      "loss": 0.2685,
+      "epoch": 2.911125527084009,
+      "grad_norm": 0.193876713514328,
+      "learning_rate": 1.9615381532411632e-05,
+      "loss": 0.3991,
       "step": 80775
     },
     {
-      "epoch": 2.84,
-      "learning_rate": 2.071204673809934e-05,
-      "loss": 0.2792,
+      "epoch": 2.911305726745234,
+      "grad_norm": 0.1711498647928238,
+      "learning_rate": 1.9612531926177453e-05,
+      "loss": 0.3954,
       "step": 80780
     },
     {
-      "epoch": 2.84,
-      "learning_rate": 2.0709240259935676e-05,
-      "loss": 0.2435,
+      "epoch": 2.911485926406458,
+      "grad_norm": 0.2698494791984558,
+      "learning_rate": 1.960968239334647e-05,
+      "loss": 0.4084,
       "step": 80785
     },
     {
-      "epoch": 2.84,
-      "learning_rate": 2.070643383748471e-05,
-      "loss": 0.2449,
+      "epoch": 2.911666126067683,
+      "grad_norm": 0.22966143488883972,
+      "learning_rate": 1.9606832933957536e-05,
+      "loss": 0.3625,
       "step": 80790
     },
     {
-      "epoch": 2.84,
-      "learning_rate": 2.070362747078287e-05,
-      "loss": 0.2518,
+      "epoch": 2.9118463257289076,
+      "grad_norm": 0.22466062009334564,
+      "learning_rate": 1.9603983548049444e-05,
+      "loss": 0.3633,
       "step": 80795
     },
     {
-      "epoch": 2.84,
-      "learning_rate": 2.0700821159866597e-05,
-      "loss": 0.2533,
+      "epoch": 2.9120265253901323,
+      "grad_norm": 0.23133142292499542,
+      "learning_rate": 1.9601134235661047e-05,
+      "loss": 0.3741,
       "step": 80800
     },
     {
-      "epoch": 2.84,
-      "learning_rate": 2.0698014904772323e-05,
-      "loss": 0.2424,
+      "epoch": 2.912206725051357,
+      "grad_norm": 0.22116337716579437,
+      "learning_rate": 1.9598284996831145e-05,
+      "loss": 0.3696,
       "step": 80805
     },
     {
-      "epoch": 2.84,
-      "learning_rate": 2.0695208705536493e-05,
-      "loss": 0.2789,
+      "epoch": 2.9123869247125818,
+      "grad_norm": 0.22359761595726013,
+      "learning_rate": 1.9595435831598558e-05,
+      "loss": 0.4236,
       "step": 80810
     },
     {
-      "epoch": 2.84,
-      "learning_rate": 2.0692402562195544e-05,
-      "loss": 0.2933,
+      "epoch": 2.912567124373806,
+      "grad_norm": 0.24041809141635895,
+      "learning_rate": 1.9592586740002116e-05,
+      "loss": 0.3961,
       "step": 80815
     },
     {
-      "epoch": 2.84,
-      "learning_rate": 2.06895964747859e-05,
-      "loss": 0.2417,
+      "epoch": 2.912747324035031,
+      "grad_norm": 0.23116619884967804,
+      "learning_rate": 1.9589737722080637e-05,
+      "loss": 0.3997,
       "step": 80820
     },
     {
-      "epoch": 2.84,
-      "learning_rate": 2.0686790443344013e-05,
-      "loss": 0.271,
+      "epoch": 2.9129275236962555,
+      "grad_norm": 0.23118247091770172,
+      "learning_rate": 1.9586888777872925e-05,
+      "loss": 0.389,
       "step": 80825
     },
     {
-      "epoch": 2.84,
-      "learning_rate": 2.0683984467906306e-05,
-      "loss": 0.2566,
+      "epoch": 2.9131077233574802,
+      "grad_norm": 0.2595424950122833,
+      "learning_rate": 1.9584039907417812e-05,
+      "loss": 0.4273,
       "step": 80830
     },
     {
-      "epoch": 2.84,
-      "learning_rate": 2.0681178548509223e-05,
-      "loss": 0.2505,
+      "epoch": 2.9132879230187045,
+      "grad_norm": 0.1928250938653946,
+      "learning_rate": 1.95811911107541e-05,
+      "loss": 0.3858,
       "step": 80835
     },
     {
-      "epoch": 2.84,
-      "learning_rate": 2.0678372685189178e-05,
-      "loss": 0.2572,
+      "epoch": 2.9134681226799293,
+      "grad_norm": 0.22600649297237396,
+      "learning_rate": 1.957834238792062e-05,
+      "loss": 0.402,
       "step": 80840
     },
     {
-      "epoch": 2.84,
-      "learning_rate": 2.0675566877982626e-05,
-      "loss": 0.2604,
+      "epoch": 2.913648322341154,
+      "grad_norm": 0.1833333820104599,
+      "learning_rate": 1.9575493738956168e-05,
+      "loss": 0.4063,
       "step": 80845
     },
     {
-      "epoch": 2.84,
-      "learning_rate": 2.067276112692599e-05,
-      "loss": 0.2816,
+      "epoch": 2.9138285220023787,
+      "grad_norm": 0.21622666716575623,
+      "learning_rate": 1.9572645163899563e-05,
+      "loss": 0.4199,
       "step": 80850
     },
     {
-      "epoch": 2.84,
-      "learning_rate": 2.0669955432055695e-05,
-      "loss": 0.2688,
+      "epoch": 2.9140087216636035,
+      "grad_norm": 0.21689878404140472,
+      "learning_rate": 1.9569796662789623e-05,
+      "loss": 0.4067,
       "step": 80855
     },
     {
-      "epoch": 2.84,
-      "learning_rate": 2.066714979340817e-05,
-      "loss": 0.2857,
+      "epoch": 2.9141889213248278,
+      "grad_norm": 0.21402768790721893,
+      "learning_rate": 1.9566948235665144e-05,
+      "loss": 0.3951,
       "step": 80860
     },
     {
-      "epoch": 2.85,
-      "learning_rate": 2.066434421101985e-05,
-      "loss": 0.2708,
+      "epoch": 2.9143691209860525,
+      "grad_norm": 0.24621036648750305,
+      "learning_rate": 1.956409988256495e-05,
+      "loss": 0.3993,
       "step": 80865
     },
     {
-      "epoch": 2.85,
-      "learning_rate": 2.0661538684927175e-05,
-      "loss": 0.2721,
+      "epoch": 2.9145493206472772,
+      "grad_norm": 0.21831317245960236,
+      "learning_rate": 1.9561251603527846e-05,
+      "loss": 0.3955,
       "step": 80870
     },
     {
-      "epoch": 2.85,
-      "learning_rate": 2.0658733215166555e-05,
-      "loss": 0.2632,
+      "epoch": 2.914729520308502,
+      "grad_norm": 0.20578089356422424,
+      "learning_rate": 1.9558403398592625e-05,
+      "loss": 0.4011,
       "step": 80875
     },
     {
-      "epoch": 2.85,
-      "learning_rate": 2.0655927801774412e-05,
-      "loss": 0.2533,
+      "epoch": 2.9149097199697263,
+      "grad_norm": 0.18618744611740112,
+      "learning_rate": 1.955555526779811e-05,
+      "loss": 0.4194,
       "step": 80880
     },
     {
-      "epoch": 2.85,
-      "learning_rate": 2.06531224447872e-05,
-      "loss": 0.2479,
+      "epoch": 2.915089919630951,
+      "grad_norm": 0.21536029875278473,
+      "learning_rate": 1.9552707211183107e-05,
+      "loss": 0.3765,
       "step": 80885
     },
     {
-      "epoch": 2.85,
-      "learning_rate": 2.065031714424132e-05,
-      "loss": 0.2733,
+      "epoch": 2.9152701192921757,
+      "grad_norm": 0.20636074244976044,
+      "learning_rate": 1.95498592287864e-05,
+      "loss": 0.3903,
       "step": 80890
     },
     {
-      "epoch": 2.85,
-      "learning_rate": 2.0647511900173208e-05,
-      "loss": 0.2774,
+      "epoch": 2.9154503189534005,
+      "grad_norm": 0.201945498585701,
+      "learning_rate": 1.9547011320646817e-05,
+      "loss": 0.3891,
       "step": 80895
     },
     {
-      "epoch": 2.85,
-      "learning_rate": 2.0644706712619276e-05,
-      "loss": 0.2727,
+      "epoch": 2.915630518614625,
+      "grad_norm": 0.16355274617671967,
+      "learning_rate": 1.9544163486803134e-05,
+      "loss": 0.4209,
       "step": 80900
     },
     {
-      "epoch": 2.85,
-      "learning_rate": 2.064190158161597e-05,
-      "loss": 0.2752,
+      "epoch": 2.9158107182758495,
+      "grad_norm": 0.23773586750030518,
+      "learning_rate": 1.9541315727294188e-05,
+      "loss": 0.4007,
       "step": 80905
     },
     {
-      "epoch": 2.85,
-      "learning_rate": 2.0639096507199693e-05,
-      "loss": 0.2747,
+      "epoch": 2.915990917937074,
+      "grad_norm": 0.1974399834871292,
+      "learning_rate": 1.953846804215875e-05,
+      "loss": 0.3999,
       "step": 80910
     },
     {
-      "epoch": 2.85,
-      "learning_rate": 2.0636291489406878e-05,
-      "loss": 0.2662,
+      "epoch": 2.916171117598299,
+      "grad_norm": 0.24811916053295135,
+      "learning_rate": 1.9535620431435623e-05,
+      "loss": 0.3971,
       "step": 80915
     },
     {
-      "epoch": 2.85,
-      "learning_rate": 2.0633486528273934e-05,
-      "loss": 0.2834,
+      "epoch": 2.9163513172595237,
+      "grad_norm": 0.20471271872520447,
+      "learning_rate": 1.9532772895163616e-05,
+      "loss": 0.4142,
       "step": 80920
     },
     {
-      "epoch": 2.85,
-      "learning_rate": 2.06306816238373e-05,
-      "loss": 0.2694,
+      "epoch": 2.916531516920748,
+      "grad_norm": 0.19927844405174255,
+      "learning_rate": 1.9529925433381517e-05,
+      "loss": 0.354,
       "step": 80925
     },
     {
-      "epoch": 2.85,
-      "learning_rate": 2.0627876776133386e-05,
-      "loss": 0.261,
+      "epoch": 2.9167117165819727,
+      "grad_norm": 0.18068204820156097,
+      "learning_rate": 1.9527078046128118e-05,
+      "loss": 0.3941,
       "step": 80930
     },
     {
-      "epoch": 2.85,
-      "learning_rate": 2.0625071985198597e-05,
-      "loss": 0.2763,
+      "epoch": 2.9168919162431974,
+      "grad_norm": 0.21484220027923584,
+      "learning_rate": 1.952423073344223e-05,
+      "loss": 0.3866,
       "step": 80935
     },
     {
-      "epoch": 2.85,
-      "learning_rate": 2.062226725106938e-05,
-      "loss": 0.2674,
+      "epoch": 2.917072115904422,
+      "grad_norm": 0.21851830184459686,
+      "learning_rate": 1.9521383495362634e-05,
+      "loss": 0.3903,
       "step": 80940
     },
     {
-      "epoch": 2.85,
-      "learning_rate": 2.0619462573782132e-05,
-      "loss": 0.2511,
+      "epoch": 2.917252315565647,
+      "grad_norm": 0.2002904862165451,
+      "learning_rate": 1.9518536331928127e-05,
+      "loss": 0.3791,
       "step": 80945
     },
     {
-      "epoch": 2.85,
-      "learning_rate": 2.061665795337328e-05,
-      "loss": 0.2441,
+      "epoch": 2.9174325152268716,
+      "grad_norm": 0.23152127861976624,
+      "learning_rate": 1.9515689243177508e-05,
+      "loss": 0.4406,
       "step": 80950
     },
     {
-      "epoch": 2.85,
-      "learning_rate": 2.0613853389879224e-05,
-      "loss": 0.2774,
+      "epoch": 2.917612714888096,
+      "grad_norm": 0.1651298850774765,
+      "learning_rate": 1.9512842229149553e-05,
+      "loss": 0.3888,
       "step": 80955
     },
     {
-      "epoch": 2.85,
-      "learning_rate": 2.06110488833364e-05,
-      "loss": 0.2761,
+      "epoch": 2.9177929145493207,
+      "grad_norm": 0.17865732312202454,
+      "learning_rate": 1.950999528988306e-05,
+      "loss": 0.3686,
       "step": 80960
     },
     {
-      "epoch": 2.85,
-      "learning_rate": 2.0608244433781216e-05,
-      "loss": 0.2804,
+      "epoch": 2.9179731142105454,
+      "grad_norm": 0.2451142817735672,
+      "learning_rate": 1.9507148425416832e-05,
+      "loss": 0.423,
       "step": 80965
     },
     {
-      "epoch": 2.85,
-      "learning_rate": 2.0605440041250078e-05,
-      "loss": 0.2761,
+      "epoch": 2.9181533138717697,
+      "grad_norm": 0.16584070026874542,
+      "learning_rate": 1.9504301635789623e-05,
+      "loss": 0.3668,
       "step": 80970
     },
     {
-      "epoch": 2.85,
-      "learning_rate": 2.0602635705779403e-05,
-      "loss": 0.2573,
+      "epoch": 2.9183335135329944,
+      "grad_norm": 0.22449831664562225,
+      "learning_rate": 1.9501454921040256e-05,
+      "loss": 0.3977,
       "step": 80975
     },
     {
-      "epoch": 2.85,
-      "learning_rate": 2.0599831427405612e-05,
-      "loss": 0.2798,
+      "epoch": 2.918513713194219,
+      "grad_norm": 0.21804749965667725,
+      "learning_rate": 1.949860828120749e-05,
+      "loss": 0.3792,
       "step": 80980
     },
     {
-      "epoch": 2.85,
-      "learning_rate": 2.059702720616511e-05,
-      "loss": 0.2702,
+      "epoch": 2.918693912855444,
+      "grad_norm": 0.20953933894634247,
+      "learning_rate": 1.9495761716330133e-05,
+      "loss": 0.4228,
       "step": 80985
     },
     {
-      "epoch": 2.85,
-      "learning_rate": 2.0594223042094308e-05,
-      "loss": 0.2639,
+      "epoch": 2.9188741125166686,
+      "grad_norm": 0.20794197916984558,
+      "learning_rate": 1.949291522644695e-05,
+      "loss": 0.4034,
       "step": 80990
     },
     {
-      "epoch": 2.85,
-      "learning_rate": 2.059141893522961e-05,
-      "loss": 0.2854,
+      "epoch": 2.9190543121778934,
+      "grad_norm": 0.2192896008491516,
+      "learning_rate": 1.949006881159673e-05,
+      "loss": 0.4113,
       "step": 80995
     },
     {
-      "epoch": 2.85,
-      "learning_rate": 2.0588614885607434e-05,
-      "loss": 0.252,
+      "epoch": 2.9192345118391176,
+      "grad_norm": 0.18000654876232147,
+      "learning_rate": 1.9487222471818255e-05,
+      "loss": 0.4156,
       "step": 81000
     },
     {
-      "epoch": 2.85,
-      "eval_loss": 0.26019665598869324,
-      "eval_runtime": 10.5615,
-      "eval_samples_per_second": 9.468,
-      "eval_steps_per_second": 9.468,
+      "epoch": 2.9192345118391176,
+      "eval_loss": 0.4301445782184601,
+      "eval_runtime": 3.5355,
+      "eval_samples_per_second": 28.285,
+      "eval_steps_per_second": 7.071,
       "step": 81000
     },
     {
-      "epoch": 2.85,
-      "learning_rate": 2.058581089326419e-05,
-      "loss": 0.284,
+      "epoch": 2.9194147115003424,
+      "grad_norm": 0.292233943939209,
+      "learning_rate": 1.9484376207150314e-05,
+      "loss": 0.3844,
       "step": 81005
     },
     {
-      "epoch": 2.85,
-      "learning_rate": 2.0583006958236283e-05,
-      "loss": 0.2698,
+      "epoch": 2.919594911161567,
+      "grad_norm": 0.21577784419059753,
+      "learning_rate": 1.948153001763167e-05,
+      "loss": 0.3826,
       "step": 81010
     },
     {
-      "epoch": 2.85,
-      "learning_rate": 2.0580203080560112e-05,
-      "loss": 0.2365,
+      "epoch": 2.9197751108227914,
+      "grad_norm": 0.2999797463417053,
+      "learning_rate": 1.9478683903301116e-05,
+      "loss": 0.4494,
       "step": 81015
     },
     {
-      "epoch": 2.85,
-      "learning_rate": 2.05773992602721e-05,
-      "loss": 0.2765,
+      "epoch": 2.919955310484016,
+      "grad_norm": 0.24274799227714539,
+      "learning_rate": 1.9475837864197418e-05,
+      "loss": 0.3721,
       "step": 81020
     },
     {
-      "epoch": 2.85,
-      "learning_rate": 2.0574595497408637e-05,
-      "loss": 0.2547,
+      "epoch": 2.920135510145241,
+      "grad_norm": 0.2308402806520462,
+      "learning_rate": 1.947299190035937e-05,
+      "loss": 0.4185,
       "step": 81025
     },
     {
-      "epoch": 2.85,
-      "learning_rate": 2.0571791792006147e-05,
-      "loss": 0.2475,
+      "epoch": 2.9203157098064656,
+      "grad_norm": 0.2198919653892517,
+      "learning_rate": 1.9470146011825733e-05,
+      "loss": 0.3646,
       "step": 81030
     },
     {
-      "epoch": 2.85,
-      "learning_rate": 2.0568988144101005e-05,
-      "loss": 0.2706,
+      "epoch": 2.9204959094676903,
+      "grad_norm": 0.21217882633209229,
+      "learning_rate": 1.946730019863528e-05,
+      "loss": 0.3997,
       "step": 81035
     },
     {
-      "epoch": 2.85,
-      "learning_rate": 2.0566184553729645e-05,
-      "loss": 0.276,
+      "epoch": 2.920676109128915,
+      "grad_norm": 0.17091058194637299,
+      "learning_rate": 1.94644544608268e-05,
+      "loss": 0.3784,
       "step": 81040
     },
     {
-      "epoch": 2.85,
-      "learning_rate": 2.0563381020928458e-05,
-      "loss": 0.2681,
+      "epoch": 2.9208563087901394,
+      "grad_norm": 0.23678135871887207,
+      "learning_rate": 1.9461608798439055e-05,
+      "loss": 0.3841,
       "step": 81045
     },
     {
-      "epoch": 2.85,
-      "learning_rate": 2.056057754573383e-05,
-      "loss": 0.2537,
+      "epoch": 2.921036508451364,
+      "grad_norm": 0.2015039473772049,
+      "learning_rate": 1.945876321151081e-05,
+      "loss": 0.4026,
       "step": 81050
     },
     {
-      "epoch": 2.85,
-      "learning_rate": 2.0557774128182184e-05,
-      "loss": 0.2597,
+      "epoch": 2.921216708112589,
+      "grad_norm": 0.2211407572031021,
+      "learning_rate": 1.945591770008085e-05,
+      "loss": 0.3967,
       "step": 81055
     },
     {
-      "epoch": 2.85,
-      "learning_rate": 2.055497076830992e-05,
-      "loss": 0.2749,
+      "epoch": 2.921396907773813,
+      "grad_norm": 0.21739064157009125,
+      "learning_rate": 1.945307226418793e-05,
+      "loss": 0.3956,
       "step": 81060
     },
     {
-      "epoch": 2.85,
-      "learning_rate": 2.0552167466153428e-05,
-      "loss": 0.2832,
+      "epoch": 2.921577107435038,
+      "grad_norm": 0.2019902616739273,
+      "learning_rate": 1.945022690387084e-05,
+      "loss": 0.4075,
       "step": 81065
     },
     {
-      "epoch": 2.85,
-      "learning_rate": 2.05493642217491e-05,
-      "loss": 0.2556,
+      "epoch": 2.9217573070962626,
+      "grad_norm": 0.1831779032945633,
+      "learning_rate": 1.9447381619168332e-05,
+      "loss": 0.3816,
       "step": 81070
     },
     {
-      "epoch": 2.85,
-      "learning_rate": 2.054656103513336e-05,
-      "loss": 0.2678,
+      "epoch": 2.9219375067574873,
+      "grad_norm": 0.21984632313251495,
+      "learning_rate": 1.944453641011916e-05,
+      "loss": 0.4448,
       "step": 81075
     },
     {
-      "epoch": 2.85,
-      "learning_rate": 2.054375790634258e-05,
-      "loss": 0.2551,
+      "epoch": 2.922117706418712,
+      "grad_norm": 0.19661669433116913,
+      "learning_rate": 1.9441691276762123e-05,
+      "loss": 0.41,
       "step": 81080
     },
     {
-      "epoch": 2.85,
-      "learning_rate": 2.0540954835413176e-05,
-      "loss": 0.2738,
+      "epoch": 2.922297906079937,
+      "grad_norm": 0.2180146425962448,
+      "learning_rate": 1.9438846219135948e-05,
+      "loss": 0.3727,
       "step": 81085
     },
     {
-      "epoch": 2.85,
-      "learning_rate": 2.0538151822381522e-05,
-      "loss": 0.2672,
+      "epoch": 2.922478105741161,
+      "grad_norm": 0.2767757475376129,
+      "learning_rate": 1.9436001237279432e-05,
+      "loss": 0.4416,
       "step": 81090
     },
     {
-      "epoch": 2.85,
-      "learning_rate": 2.0535348867284036e-05,
-      "loss": 0.247,
+      "epoch": 2.922658305402386,
+      "grad_norm": 0.20873478055000305,
+      "learning_rate": 1.9433156331231314e-05,
+      "loss": 0.3978,
       "step": 81095
     },
     {
-      "epoch": 2.85,
-      "learning_rate": 2.05325459701571e-05,
-      "loss": 0.2663,
+      "epoch": 2.9228385050636105,
+      "grad_norm": 0.26107144355773926,
+      "learning_rate": 1.9430311501030362e-05,
+      "loss": 0.3829,
       "step": 81100
     },
     {
-      "epoch": 2.85,
-      "learning_rate": 2.052974313103711e-05,
-      "loss": 0.2895,
+      "epoch": 2.9230187047248353,
+      "grad_norm": 0.267535924911499,
+      "learning_rate": 1.9427466746715342e-05,
+      "loss": 0.4117,
       "step": 81105
     },
     {
-      "epoch": 2.85,
-      "learning_rate": 2.0526940349960455e-05,
-      "loss": 0.3062,
+      "epoch": 2.9231989043860596,
+      "grad_norm": 0.21252234280109406,
+      "learning_rate": 1.942462206832501e-05,
+      "loss": 0.379,
       "step": 81110
     },
     {
-      "epoch": 2.85,
-      "learning_rate": 2.052413762696354e-05,
-      "loss": 0.2912,
+      "epoch": 2.9233791040472843,
+      "grad_norm": 0.1903425008058548,
+      "learning_rate": 1.9421777465898114e-05,
+      "loss": 0.352,
       "step": 81115
     },
     {
-      "epoch": 2.85,
-      "learning_rate": 2.0521334962082748e-05,
-      "loss": 0.2731,
+      "epoch": 2.923559303708509,
+      "grad_norm": 0.1975475549697876,
+      "learning_rate": 1.9418932939473426e-05,
+      "loss": 0.3898,
       "step": 81120
     },
     {
-      "epoch": 2.85,
-      "learning_rate": 2.051853235535447e-05,
-      "loss": 0.2838,
+      "epoch": 2.9237395033697338,
+      "grad_norm": 0.22255592048168182,
+      "learning_rate": 1.941608848908969e-05,
+      "loss": 0.3482,
       "step": 81125
     },
     {
-      "epoch": 2.85,
-      "learning_rate": 2.0515729806815084e-05,
-      "loss": 0.2635,
+      "epoch": 2.9239197030309585,
+      "grad_norm": 0.24407097697257996,
+      "learning_rate": 1.941324411478568e-05,
+      "loss": 0.3865,
       "step": 81130
     },
     {
-      "epoch": 2.85,
-      "learning_rate": 2.0512927316500995e-05,
-      "loss": 0.2871,
+      "epoch": 2.924099902692183,
+      "grad_norm": 0.18102699518203735,
+      "learning_rate": 1.9410399816600128e-05,
+      "loss": 0.3786,
       "step": 81135
     },
     {
-      "epoch": 2.85,
-      "learning_rate": 2.05101248844486e-05,
-      "loss": 0.2514,
+      "epoch": 2.9242801023534075,
+      "grad_norm": 0.23479129374027252,
+      "learning_rate": 1.9407555594571796e-05,
+      "loss": 0.3847,
       "step": 81140
     },
     {
-      "epoch": 2.85,
-      "learning_rate": 2.050732251069427e-05,
-      "loss": 0.2808,
+      "epoch": 2.9244603020146323,
+      "grad_norm": 0.20247110724449158,
+      "learning_rate": 1.9404711448739442e-05,
+      "loss": 0.3794,
       "step": 81145
     },
     {
-      "epoch": 2.86,
-      "learning_rate": 2.0504520195274384e-05,
-      "loss": 0.2765,
+      "epoch": 2.924640501675857,
+      "grad_norm": 0.2692834436893463,
+      "learning_rate": 1.9401867379141815e-05,
+      "loss": 0.4258,
       "step": 81150
     },
     {
-      "epoch": 2.86,
-      "learning_rate": 2.0501717938225355e-05,
-      "loss": 0.2493,
+      "epoch": 2.9248207013370813,
+      "grad_norm": 0.20718295872211456,
+      "learning_rate": 1.9399023385817644e-05,
+      "loss": 0.4251,
       "step": 81155
     },
     {
-      "epoch": 2.86,
-      "learning_rate": 2.0498915739583547e-05,
-      "loss": 0.2634,
+      "epoch": 2.925000900998306,
+      "grad_norm": 0.18877676129341125,
+      "learning_rate": 1.939617946880571e-05,
+      "loss": 0.3919,
       "step": 81160
     },
     {
-      "epoch": 2.86,
-      "learning_rate": 2.0496113599385353e-05,
-      "loss": 0.2576,
+      "epoch": 2.9251811006595307,
+      "grad_norm": 0.2178078591823578,
+      "learning_rate": 1.939333562814473e-05,
+      "loss": 0.4156,
       "step": 81165
     },
     {
-      "epoch": 2.86,
-      "learning_rate": 2.0493311517667153e-05,
-      "loss": 0.2958,
+      "epoch": 2.9253613003207555,
+      "grad_norm": 0.17405816912651062,
+      "learning_rate": 1.9390491863873482e-05,
+      "loss": 0.371,
       "step": 81170
     },
     {
-      "epoch": 2.86,
-      "learning_rate": 2.049050949446534e-05,
-      "loss": 0.2549,
+      "epoch": 2.92554149998198,
+      "grad_norm": 0.20669405162334442,
+      "learning_rate": 1.9387648176030697e-05,
+      "loss": 0.389,
       "step": 81175
     },
     {
-      "epoch": 2.86,
-      "learning_rate": 2.048770752981629e-05,
-      "loss": 0.2545,
+      "epoch": 2.925721699643205,
+      "grad_norm": 0.20877471566200256,
+      "learning_rate": 1.9384804564655106e-05,
+      "loss": 0.3857,
       "step": 81180
     },
     {
-      "epoch": 2.86,
-      "learning_rate": 2.048490562375637e-05,
-      "loss": 0.2702,
+      "epoch": 2.9259018993044292,
+      "grad_norm": 0.2562907934188843,
+      "learning_rate": 1.9381961029785485e-05,
+      "loss": 0.396,
       "step": 81185
     },
     {
-      "epoch": 2.86,
-      "learning_rate": 2.0482103776321983e-05,
-      "loss": 0.291,
+      "epoch": 2.926082098965654,
+      "grad_norm": 0.2052457183599472,
+      "learning_rate": 1.9379117571460545e-05,
+      "loss": 0.3769,
       "step": 81190
     },
     {
-      "epoch": 2.86,
-      "learning_rate": 2.047930198754951e-05,
-      "loss": 0.2739,
+      "epoch": 2.9262622986268787,
+      "grad_norm": 0.26629289984703064,
+      "learning_rate": 1.9376274189719034e-05,
+      "loss": 0.4164,
       "step": 81195
     },
     {
-      "epoch": 2.86,
-      "learning_rate": 2.047650025747532e-05,
-      "loss": 0.2484,
+      "epoch": 2.926442498288103,
+      "grad_norm": 0.23448923230171204,
+      "learning_rate": 1.9373430884599707e-05,
+      "loss": 0.3989,
       "step": 81200
     },
     {
-      "epoch": 2.86,
-      "learning_rate": 2.0473698586135776e-05,
-      "loss": 0.286,
+      "epoch": 2.9266226979493277,
+      "grad_norm": 0.2015734761953354,
+      "learning_rate": 1.937058765614129e-05,
+      "loss": 0.4129,
       "step": 81205
     },
     {
-      "epoch": 2.86,
-      "learning_rate": 2.047089697356729e-05,
-      "loss": 0.2554,
+      "epoch": 2.9268028976105525,
+      "grad_norm": 0.3036739230155945,
+      "learning_rate": 1.936774450438253e-05,
+      "loss": 0.3919,
       "step": 81210
     },
     {
-      "epoch": 2.86,
-      "learning_rate": 2.046809541980621e-05,
-      "loss": 0.303,
+      "epoch": 2.926983097271777,
+      "grad_norm": 0.19649726152420044,
+      "learning_rate": 1.9364901429362164e-05,
+      "loss": 0.4088,
       "step": 81215
     },
     {
-      "epoch": 2.86,
-      "learning_rate": 2.0465293924888936e-05,
-      "loss": 0.2859,
+      "epoch": 2.927163296933002,
+      "grad_norm": 0.2458265870809555,
+      "learning_rate": 1.9362058431118917e-05,
+      "loss": 0.4255,
       "step": 81220
     },
     {
-      "epoch": 2.86,
-      "learning_rate": 2.046249248885182e-05,
-      "loss": 0.2765,
+      "epoch": 2.9273434965942267,
+      "grad_norm": 0.22339041531085968,
+      "learning_rate": 1.935921550969154e-05,
+      "loss": 0.3925,
       "step": 81225
     },
     {
-      "epoch": 2.86,
-      "learning_rate": 2.0459691111731258e-05,
-      "loss": 0.2501,
+      "epoch": 2.927523696255451,
+      "grad_norm": 0.18713003396987915,
+      "learning_rate": 1.9356372665118754e-05,
+      "loss": 0.3988,
       "step": 81230
     },
     {
-      "epoch": 2.86,
-      "learning_rate": 2.0456889793563617e-05,
-      "loss": 0.2619,
+      "epoch": 2.9277038959166757,
+      "grad_norm": 0.2202412635087967,
+      "learning_rate": 1.9353529897439298e-05,
+      "loss": 0.3597,
       "step": 81235
     },
     {
-      "epoch": 2.86,
-      "learning_rate": 2.045408853438526e-05,
-      "loss": 0.2544,
+      "epoch": 2.9278840955779004,
+      "grad_norm": 0.21688272058963776,
+      "learning_rate": 1.9350687206691904e-05,
+      "loss": 0.4228,
       "step": 81240
     },
     {
-      "epoch": 2.86,
-      "learning_rate": 2.0451287334232567e-05,
-      "loss": 0.2714,
+      "epoch": 2.9280642952391247,
+      "grad_norm": 0.22464942932128906,
+      "learning_rate": 1.9347844592915303e-05,
+      "loss": 0.4061,
       "step": 81245
     },
     {
-      "epoch": 2.86,
-      "learning_rate": 2.0448486193141914e-05,
-      "loss": 0.2519,
+      "epoch": 2.9282444949003494,
+      "grad_norm": 0.18823909759521484,
+      "learning_rate": 1.934500205614823e-05,
+      "loss": 0.4211,
       "step": 81250
     },
     {
-      "epoch": 2.86,
-      "learning_rate": 2.044568511114967e-05,
-      "loss": 0.2771,
+      "epoch": 2.928424694561574,
+      "grad_norm": 0.1837102770805359,
+      "learning_rate": 1.9342159596429414e-05,
+      "loss": 0.3719,
       "step": 81255
     },
     {
-      "epoch": 2.86,
-      "learning_rate": 2.0442884088292204e-05,
-      "loss": 0.2701,
+      "epoch": 2.928604894222799,
+      "grad_norm": 0.19186070561408997,
+      "learning_rate": 1.933931721379756e-05,
+      "loss": 0.3568,
       "step": 81260
     },
     {
-      "epoch": 2.86,
-      "learning_rate": 2.0440083124605873e-05,
-      "loss": 0.2639,
+      "epoch": 2.9287850938840236,
+      "grad_norm": 0.2203979641199112,
+      "learning_rate": 1.933647490829143e-05,
+      "loss": 0.3601,
       "step": 81265
     },
     {
-      "epoch": 2.86,
-      "learning_rate": 2.0437282220127067e-05,
-      "loss": 0.2588,
+      "epoch": 2.9289652935452484,
+      "grad_norm": 0.1754385232925415,
+      "learning_rate": 1.9333632679949713e-05,
+      "loss": 0.3799,
       "step": 81270
     },
     {
-      "epoch": 2.86,
-      "learning_rate": 2.0434481374892148e-05,
-      "loss": 0.2643,
+      "epoch": 2.9291454932064727,
+      "grad_norm": 0.24749146401882172,
+      "learning_rate": 1.9330790528811177e-05,
+      "loss": 0.3833,
       "step": 81275
     },
     {
-      "epoch": 2.86,
-      "learning_rate": 2.043168058893747e-05,
-      "loss": 0.2891,
+      "epoch": 2.9293256928676974,
+      "grad_norm": 0.19745169579982758,
+      "learning_rate": 1.9327948454914514e-05,
+      "loss": 0.3694,
       "step": 81280
     },
     {
-      "epoch": 2.86,
-      "learning_rate": 2.0428879862299414e-05,
-      "loss": 0.2688,
+      "epoch": 2.929505892528922,
+      "grad_norm": 0.24451091885566711,
+      "learning_rate": 1.932510645829845e-05,
+      "loss": 0.3911,
       "step": 81285
     },
     {
-      "epoch": 2.86,
-      "learning_rate": 2.0426079195014347e-05,
-      "loss": 0.2552,
+      "epoch": 2.9296860921901464,
+      "grad_norm": 0.22178997099399567,
+      "learning_rate": 1.9322264539001716e-05,
+      "loss": 0.3695,
       "step": 81290
     },
     {
-      "epoch": 2.86,
-      "learning_rate": 2.042327858711862e-05,
-      "loss": 0.2722,
+      "epoch": 2.929866291851371,
+      "grad_norm": 0.2586722671985626,
+      "learning_rate": 1.9319422697063027e-05,
+      "loss": 0.402,
       "step": 81295
     },
     {
-      "epoch": 2.86,
-      "learning_rate": 2.0420478038648604e-05,
-      "loss": 0.2542,
+      "epoch": 2.930046491512596,
+      "grad_norm": 0.21260590851306915,
+      "learning_rate": 1.9316580932521102e-05,
+      "loss": 0.4434,
       "step": 81300
     },
     {
-      "epoch": 2.86,
-      "learning_rate": 2.0417677549640662e-05,
-      "loss": 0.2934,
+      "epoch": 2.9302266911738206,
+      "grad_norm": 0.21419310569763184,
+      "learning_rate": 1.9313739245414665e-05,
+      "loss": 0.3775,
       "step": 81305
     },
     {
-      "epoch": 2.86,
-      "learning_rate": 2.0414877120131163e-05,
-      "loss": 0.2756,
+      "epoch": 2.9304068908350454,
+      "grad_norm": 0.18965986371040344,
+      "learning_rate": 1.9310897635782426e-05,
+      "loss": 0.4149,
       "step": 81310
     },
     {
-      "epoch": 2.86,
-      "learning_rate": 2.0412076750156463e-05,
-      "loss": 0.276,
+      "epoch": 2.93058709049627,
+      "grad_norm": 0.2388392835855484,
+      "learning_rate": 1.930805610366311e-05,
+      "loss": 0.4024,
       "step": 81315
     },
     {
-      "epoch": 2.86,
-      "learning_rate": 2.040927643975291e-05,
-      "loss": 0.2806,
+      "epoch": 2.9307672901574944,
+      "grad_norm": 0.21854186058044434,
+      "learning_rate": 1.9305214649095426e-05,
+      "loss": 0.3543,
       "step": 81320
     },
     {
-      "epoch": 2.86,
-      "learning_rate": 2.0406476188956892e-05,
-      "loss": 0.2896,
+      "epoch": 2.930947489818719,
+      "grad_norm": 0.18977004289627075,
+      "learning_rate": 1.930237327211809e-05,
+      "loss": 0.3614,
       "step": 81325
     },
     {
-      "epoch": 2.86,
-      "learning_rate": 2.0403675997804748e-05,
-      "loss": 0.2642,
+      "epoch": 2.931127689479944,
+      "grad_norm": 0.20082025229930878,
+      "learning_rate": 1.9299531972769813e-05,
+      "loss": 0.3993,
       "step": 81330
     },
     {
-      "epoch": 2.86,
-      "learning_rate": 2.0400875866332845e-05,
-      "loss": 0.2652,
+      "epoch": 2.9313078891411686,
+      "grad_norm": 0.20026427507400513,
+      "learning_rate": 1.9296690751089312e-05,
+      "loss": 0.3545,
       "step": 81335
     },
     {
-      "epoch": 2.86,
-      "learning_rate": 2.0398075794577527e-05,
-      "loss": 0.2599,
+      "epoch": 2.931488088802393,
+      "grad_norm": 0.22359101474285126,
+      "learning_rate": 1.9293849607115292e-05,
+      "loss": 0.3945,
       "step": 81340
     },
     {
-      "epoch": 2.86,
-      "learning_rate": 2.0395275782575172e-05,
-      "loss": 0.2457,
+      "epoch": 2.9316682884636176,
+      "grad_norm": 0.2114863246679306,
+      "learning_rate": 1.929100854088647e-05,
+      "loss": 0.3842,
       "step": 81345
     },
     {
-      "epoch": 2.86,
-      "learning_rate": 2.0392475830362122e-05,
-      "loss": 0.274,
+      "epoch": 2.9318484881248423,
+      "grad_norm": 0.22665932774543762,
+      "learning_rate": 1.928816755244155e-05,
+      "loss": 0.3776,
       "step": 81350
     },
     {
-      "epoch": 2.86,
-      "learning_rate": 2.0389675937974745e-05,
-      "loss": 0.2751,
+      "epoch": 2.932028687786067,
+      "grad_norm": 0.27405086159706116,
+      "learning_rate": 1.928532664181924e-05,
+      "loss": 0.4077,
       "step": 81355
     },
     {
-      "epoch": 2.86,
-      "learning_rate": 2.0386876105449372e-05,
-      "loss": 0.2649,
+      "epoch": 2.932208887447292,
+      "grad_norm": 0.16725726425647736,
+      "learning_rate": 1.9282485809058254e-05,
+      "loss": 0.3516,
       "step": 81360
     },
     {
-      "epoch": 2.86,
-      "learning_rate": 2.0384076332822387e-05,
-      "loss": 0.2685,
+      "epoch": 2.932389087108516,
+      "grad_norm": 0.22895494103431702,
+      "learning_rate": 1.927964505419728e-05,
+      "loss": 0.3507,
       "step": 81365
     },
     {
-      "epoch": 2.86,
-      "learning_rate": 2.0381276620130126e-05,
-      "loss": 0.2657,
+      "epoch": 2.932569286769741,
+      "grad_norm": 0.2224702686071396,
+      "learning_rate": 1.9276804377275047e-05,
+      "loss": 0.3727,
       "step": 81370
     },
     {
-      "epoch": 2.86,
-      "learning_rate": 2.0378476967408942e-05,
-      "loss": 0.282,
+      "epoch": 2.9327494864309656,
+      "grad_norm": 0.2331514060497284,
+      "learning_rate": 1.927396377833024e-05,
+      "loss": 0.4005,
       "step": 81375
     },
     {
-      "epoch": 2.86,
-      "learning_rate": 2.0375677374695188e-05,
-      "loss": 0.283,
+      "epoch": 2.9329296860921903,
+      "grad_norm": 0.18143446743488312,
+      "learning_rate": 1.9271123257401568e-05,
+      "loss": 0.3625,
       "step": 81380
     },
     {
-      "epoch": 2.86,
-      "learning_rate": 2.0372877842025214e-05,
-      "loss": 0.273,
+      "epoch": 2.9331098857534146,
+      "grad_norm": 0.20633216202259064,
+      "learning_rate": 1.9268282814527737e-05,
+      "loss": 0.3695,
       "step": 81385
     },
     {
-      "epoch": 2.86,
-      "learning_rate": 2.0370078369435378e-05,
-      "loss": 0.2746,
+      "epoch": 2.9332900854146393,
+      "grad_norm": 0.2796785235404968,
+      "learning_rate": 1.9265442449747432e-05,
+      "loss": 0.4057,
       "step": 81390
     },
     {
-      "epoch": 2.86,
-      "learning_rate": 2.0367278956962012e-05,
-      "loss": 0.2806,
+      "epoch": 2.933470285075864,
+      "grad_norm": 0.23631520569324493,
+      "learning_rate": 1.9262602163099375e-05,
+      "loss": 0.3928,
       "step": 81395
     },
     {
-      "epoch": 2.86,
-      "learning_rate": 2.036447960464149e-05,
-      "loss": 0.2817,
+      "epoch": 2.933650484737089,
+      "grad_norm": 0.2131415605545044,
+      "learning_rate": 1.925976195462225e-05,
+      "loss": 0.3849,
       "step": 81400
     },
     {
-      "epoch": 2.86,
-      "learning_rate": 2.036168031251014e-05,
-      "loss": 0.2665,
+      "epoch": 2.9338306843983135,
+      "grad_norm": 0.2613097131252289,
+      "learning_rate": 1.925692182435475e-05,
+      "loss": 0.418,
       "step": 81405
     },
     {
-      "epoch": 2.86,
-      "learning_rate": 2.0358881080604313e-05,
-      "loss": 0.2534,
+      "epoch": 2.934010884059538,
+      "grad_norm": 0.24537603557109833,
+      "learning_rate": 1.9254081772335586e-05,
+      "loss": 0.4091,
       "step": 81410
     },
     {
-      "epoch": 2.86,
-      "learning_rate": 2.035608190896035e-05,
-      "loss": 0.2562,
+      "epoch": 2.9341910837207625,
+      "grad_norm": 0.19642150402069092,
+      "learning_rate": 1.9251241798603447e-05,
+      "loss": 0.4022,
       "step": 81415
     },
     {
-      "epoch": 2.86,
-      "learning_rate": 2.035328279761461e-05,
-      "loss": 0.2575,
+      "epoch": 2.9343712833819873,
+      "grad_norm": 0.1668914407491684,
+      "learning_rate": 1.9248401903197015e-05,
+      "loss": 0.3614,
       "step": 81420
     },
     {
-      "epoch": 2.86,
-      "learning_rate": 2.0350483746603437e-05,
-      "loss": 0.292,
+      "epoch": 2.934551483043212,
+      "grad_norm": 0.20156939327716827,
+      "learning_rate": 1.9245562086155004e-05,
+      "loss": 0.4329,
       "step": 81425
     },
     {
-      "epoch": 2.86,
-      "learning_rate": 2.0347684755963164e-05,
-      "loss": 0.2702,
+      "epoch": 2.9347316827044363,
+      "grad_norm": 0.22902165353298187,
+      "learning_rate": 1.9242722347516088e-05,
+      "loss": 0.3377,
       "step": 81430
     },
     {
-      "epoch": 2.87,
-      "learning_rate": 2.034488582573013e-05,
-      "loss": 0.2691,
+      "epoch": 2.934911882365661,
+      "grad_norm": 0.1990087330341339,
+      "learning_rate": 1.9239882687318972e-05,
+      "loss": 0.4288,
       "step": 81435
     },
     {
-      "epoch": 2.87,
-      "learning_rate": 2.034208695594069e-05,
-      "loss": 0.2523,
+      "epoch": 2.9350920820268858,
+      "grad_norm": 0.1762835681438446,
+      "learning_rate": 1.9237043105602342e-05,
+      "loss": 0.3982,
       "step": 81440
     },
     {
-      "epoch": 2.87,
-      "learning_rate": 2.0339288146631185e-05,
-      "loss": 0.3037,
+      "epoch": 2.9352722816881105,
+      "grad_norm": 0.23856672644615173,
+      "learning_rate": 1.923420360240487e-05,
+      "loss": 0.4047,
       "step": 81445
     },
     {
-      "epoch": 2.87,
-      "learning_rate": 2.0336489397837956e-05,
-      "loss": 0.264,
+      "epoch": 2.9354524813493352,
+      "grad_norm": 0.19141538441181183,
+      "learning_rate": 1.923136417776527e-05,
+      "loss": 0.4154,
       "step": 81450
     },
     {
-      "epoch": 2.87,
-      "learning_rate": 2.0333690709597326e-05,
-      "loss": 0.2505,
+      "epoch": 2.93563268101056,
+      "grad_norm": 0.21707025170326233,
+      "learning_rate": 1.9228524831722206e-05,
+      "loss": 0.3691,
       "step": 81455
     },
     {
-      "epoch": 2.87,
-      "learning_rate": 2.0330892081945657e-05,
-      "loss": 0.2608,
+      "epoch": 2.9358128806717843,
+      "grad_norm": 0.2077445536851883,
+      "learning_rate": 1.922568556431438e-05,
+      "loss": 0.3898,
       "step": 81460
     },
     {
-      "epoch": 2.87,
-      "learning_rate": 2.032809351491927e-05,
-      "loss": 0.2412,
+      "epoch": 2.935993080333009,
+      "grad_norm": 0.17210866510868073,
+      "learning_rate": 1.9222846375580478e-05,
+      "loss": 0.3616,
       "step": 81465
     },
     {
-      "epoch": 2.87,
-      "learning_rate": 2.0325295008554517e-05,
-      "loss": 0.2593,
+      "epoch": 2.9361732799942337,
+      "grad_norm": 0.1976180225610733,
+      "learning_rate": 1.922000726555916e-05,
+      "loss": 0.3905,
       "step": 81470
     },
     {
-      "epoch": 2.87,
-      "learning_rate": 2.032249656288772e-05,
-      "loss": 0.2561,
+      "epoch": 2.936353479655458,
+      "grad_norm": 0.2516196072101593,
+      "learning_rate": 1.9217168234289136e-05,
+      "loss": 0.4011,
       "step": 81475
     },
     {
-      "epoch": 2.87,
-      "learning_rate": 2.0319698177955228e-05,
-      "loss": 0.2668,
+      "epoch": 2.9365336793166827,
+      "grad_norm": 0.2320529818534851,
+      "learning_rate": 1.9214329281809074e-05,
+      "loss": 0.3951,
       "step": 81480
     },
     {
-      "epoch": 2.87,
-      "learning_rate": 2.0316899853793374e-05,
-      "loss": 0.2754,
+      "epoch": 2.9367138789779075,
+      "grad_norm": 0.19517144560813904,
+      "learning_rate": 1.921149040815765e-05,
+      "loss": 0.3724,
       "step": 81485
     },
     {
-      "epoch": 2.87,
-      "learning_rate": 2.0314101590438484e-05,
-      "loss": 0.2649,
+      "epoch": 2.936894078639132,
+      "grad_norm": 0.21140725910663605,
+      "learning_rate": 1.9208651613373553e-05,
+      "loss": 0.3686,
       "step": 81490
     },
     {
-      "epoch": 2.87,
-      "learning_rate": 2.0311303387926888e-05,
-      "loss": 0.2693,
+      "epoch": 2.937074278300357,
+      "grad_norm": 0.19208674132823944,
+      "learning_rate": 1.9205812897495448e-05,
+      "loss": 0.3837,
       "step": 81495
     },
     {
-      "epoch": 2.87,
-      "learning_rate": 2.0308505246294942e-05,
-      "loss": 0.2814,
+      "epoch": 2.9372544779615817,
+      "grad_norm": 0.19858016073703766,
+      "learning_rate": 1.920297426056203e-05,
+      "loss": 0.3975,
       "step": 81500
     },
     {
-      "epoch": 2.87,
-      "eval_loss": 0.2588658034801483,
-      "eval_runtime": 10.5545,
-      "eval_samples_per_second": 9.475,
-      "eval_steps_per_second": 9.475,
+      "epoch": 2.9372544779615817,
+      "eval_loss": 0.43074142932891846,
+      "eval_runtime": 3.5337,
+      "eval_samples_per_second": 28.299,
+      "eval_steps_per_second": 7.075,
       "step": 81500
     },
     {
-      "epoch": 2.87,
-      "learning_rate": 2.0305707165578963e-05,
-      "loss": 0.2613,
+      "epoch": 2.937434677622806,
+      "grad_norm": 0.20274978876113892,
+      "learning_rate": 1.920013570261196e-05,
+      "loss": 0.3892,
       "step": 81505
     },
     {
-      "epoch": 2.87,
-      "learning_rate": 2.030290914581527e-05,
-      "loss": 0.2867,
+      "epoch": 2.9376148772840307,
+      "grad_norm": 0.2162855863571167,
+      "learning_rate": 1.919729722368392e-05,
+      "loss": 0.4381,
       "step": 81510
     },
     {
-      "epoch": 2.87,
-      "learning_rate": 2.0300111187040222e-05,
-      "loss": 0.2575,
+      "epoch": 2.9377950769452554,
+      "grad_norm": 0.212267205119133,
+      "learning_rate": 1.9194458823816583e-05,
+      "loss": 0.3685,
       "step": 81515
     },
     {
-      "epoch": 2.87,
-      "learning_rate": 2.0297313289290122e-05,
-      "loss": 0.2701,
+      "epoch": 2.9379752766064797,
+      "grad_norm": 0.18356037139892578,
+      "learning_rate": 1.919162050304862e-05,
+      "loss": 0.3908,
       "step": 81520
     },
     {
-      "epoch": 2.87,
-      "learning_rate": 2.0294515452601316e-05,
-      "loss": 0.2571,
+      "epoch": 2.9381554762677045,
+      "grad_norm": 0.23290087282657623,
+      "learning_rate": 1.91887822614187e-05,
+      "loss": 0.4314,
       "step": 81525
     },
     {
-      "epoch": 2.87,
-      "learning_rate": 2.0291717677010115e-05,
-      "loss": 0.2659,
+      "epoch": 2.938335675928929,
+      "grad_norm": 0.2154729664325714,
+      "learning_rate": 1.9185944098965502e-05,
+      "loss": 0.3872,
       "step": 81530
     },
     {
-      "epoch": 2.87,
-      "learning_rate": 2.0288919962552873e-05,
-      "loss": 0.2722,
+      "epoch": 2.938515875590154,
+      "grad_norm": 0.19503213465213776,
+      "learning_rate": 1.918310601572768e-05,
+      "loss": 0.3756,
       "step": 81535
     },
     {
-      "epoch": 2.87,
-      "learning_rate": 2.0286122309265894e-05,
-      "loss": 0.2492,
+      "epoch": 2.9386960752513787,
+      "grad_norm": 0.22458268702030182,
+      "learning_rate": 1.9180268011743925e-05,
+      "loss": 0.3672,
       "step": 81540
     },
     {
-      "epoch": 2.87,
-      "learning_rate": 2.0283324717185505e-05,
-      "loss": 0.2694,
+      "epoch": 2.9388762749126034,
+      "grad_norm": 0.20593036711215973,
+      "learning_rate": 1.9177430087052892e-05,
+      "loss": 0.4,
       "step": 81545
     },
     {
-      "epoch": 2.87,
-      "learning_rate": 2.0280527186348035e-05,
-      "loss": 0.2685,
+      "epoch": 2.9390564745738277,
+      "grad_norm": 0.21392624080181122,
+      "learning_rate": 1.917459224169323e-05,
+      "loss": 0.3966,
       "step": 81550
     },
     {
-      "epoch": 2.87,
-      "learning_rate": 2.0277729716789817e-05,
-      "loss": 0.2735,
+      "epoch": 2.9392366742350524,
+      "grad_norm": 0.19831807911396027,
+      "learning_rate": 1.917175447570364e-05,
+      "loss": 0.3983,
       "step": 81555
     },
     {
-      "epoch": 2.87,
-      "learning_rate": 2.027493230854717e-05,
-      "loss": 0.2833,
+      "epoch": 2.939416873896277,
+      "grad_norm": 0.20181818306446075,
+      "learning_rate": 1.916891678912276e-05,
+      "loss": 0.3699,
       "step": 81560
     },
     {
-      "epoch": 2.87,
-      "learning_rate": 2.027213496165641e-05,
-      "loss": 0.3045,
+      "epoch": 2.9395970735575014,
+      "grad_norm": 0.19284601509571075,
+      "learning_rate": 1.916607918198926e-05,
+      "loss": 0.3628,
       "step": 81565
     },
     {
-      "epoch": 2.87,
-      "learning_rate": 2.0269337676153852e-05,
-      "loss": 0.2866,
+      "epoch": 2.939777273218726,
+      "grad_norm": 0.20783284306526184,
+      "learning_rate": 1.91632416543418e-05,
+      "loss": 0.4283,
       "step": 81570
     },
     {
-      "epoch": 2.87,
-      "learning_rate": 2.026654045207583e-05,
-      "loss": 0.2906,
+      "epoch": 2.939957472879951,
+      "grad_norm": 0.20689019560813904,
+      "learning_rate": 1.916040420621904e-05,
+      "loss": 0.3616,
       "step": 81575
     },
     {
-      "epoch": 2.87,
-      "learning_rate": 2.026374328945867e-05,
-      "loss": 0.2755,
+      "epoch": 2.9401376725411756,
+      "grad_norm": 0.3156082332134247,
+      "learning_rate": 1.9157566837659645e-05,
+      "loss": 0.4052,
       "step": 81580
     },
     {
-      "epoch": 2.87,
-      "learning_rate": 2.0260946188338677e-05,
-      "loss": 0.271,
+      "epoch": 2.9403178722024004,
+      "grad_norm": 0.23160193860530853,
+      "learning_rate": 1.9154729548702272e-05,
+      "loss": 0.3514,
       "step": 81585
     },
     {
-      "epoch": 2.87,
-      "learning_rate": 2.0258149148752168e-05,
-      "loss": 0.2736,
+      "epoch": 2.940498071863625,
+      "grad_norm": 0.18180207908153534,
+      "learning_rate": 1.9151892339385573e-05,
+      "loss": 0.4118,
       "step": 81590
     },
     {
-      "epoch": 2.87,
-      "learning_rate": 2.0255352170735477e-05,
-      "loss": 0.2743,
+      "epoch": 2.9406782715248494,
+      "grad_norm": 0.17559707164764404,
+      "learning_rate": 1.9149055209748214e-05,
+      "loss": 0.3459,
       "step": 81595
     },
     {
-      "epoch": 2.87,
-      "learning_rate": 2.0252555254324905e-05,
-      "loss": 0.2799,
+      "epoch": 2.940858471186074,
+      "grad_norm": 0.2852272689342499,
+      "learning_rate": 1.9146218159828845e-05,
+      "loss": 0.4089,
       "step": 81600
     },
     {
-      "epoch": 2.87,
-      "learning_rate": 2.024975839955678e-05,
-      "loss": 0.2507,
+      "epoch": 2.941038670847299,
+      "grad_norm": 0.19265016913414001,
+      "learning_rate": 1.914338118966611e-05,
+      "loss": 0.4113,
       "step": 81605
     },
     {
-      "epoch": 2.87,
-      "learning_rate": 2.0246961606467402e-05,
-      "loss": 0.2396,
+      "epoch": 2.9412188705085236,
+      "grad_norm": 0.27521640062332153,
+      "learning_rate": 1.9140544299298687e-05,
+      "loss": 0.4314,
       "step": 81610
     },
     {
-      "epoch": 2.87,
-      "learning_rate": 2.0244164875093106e-05,
-      "loss": 0.2552,
+      "epoch": 2.941399070169748,
+      "grad_norm": 0.22525955736637115,
+      "learning_rate": 1.91377074887652e-05,
+      "loss": 0.4097,
       "step": 81615
     },
     {
-      "epoch": 2.87,
-      "learning_rate": 2.0241368205470194e-05,
-      "loss": 0.2729,
+      "epoch": 2.9415792698309726,
+      "grad_norm": 0.1709366887807846,
+      "learning_rate": 1.9134870758104324e-05,
+      "loss": 0.3707,
       "step": 81620
     },
     {
-      "epoch": 2.87,
-      "learning_rate": 2.023857159763497e-05,
-      "loss": 0.2491,
+      "epoch": 2.9417594694921974,
+      "grad_norm": 0.20338067412376404,
+      "learning_rate": 1.9132034107354693e-05,
+      "loss": 0.4064,
       "step": 81625
     },
     {
-      "epoch": 2.87,
-      "learning_rate": 2.0235775051623756e-05,
-      "loss": 0.2673,
+      "epoch": 2.941939669153422,
+      "grad_norm": 0.21200573444366455,
+      "learning_rate": 1.912919753655496e-05,
+      "loss": 0.3752,
       "step": 81630
     },
     {
-      "epoch": 2.87,
-      "learning_rate": 2.0232978567472875e-05,
-      "loss": 0.2717,
+      "epoch": 2.942119868814647,
+      "grad_norm": 0.19498129189014435,
+      "learning_rate": 1.9126361045743778e-05,
+      "loss": 0.3777,
       "step": 81635
     },
     {
-      "epoch": 2.87,
-      "learning_rate": 2.0230182145218624e-05,
-      "loss": 0.2596,
+      "epoch": 2.942300068475871,
+      "grad_norm": 0.24946102499961853,
+      "learning_rate": 1.9123524634959794e-05,
+      "loss": 0.4022,
       "step": 81640
     },
     {
-      "epoch": 2.87,
-      "learning_rate": 2.02273857848973e-05,
-      "loss": 0.2645,
+      "epoch": 2.942480268137096,
+      "grad_norm": 0.25150594115257263,
+      "learning_rate": 1.9120688304241636e-05,
+      "loss": 0.3575,
       "step": 81645
     },
     {
-      "epoch": 2.87,
-      "learning_rate": 2.022458948654524e-05,
-      "loss": 0.2759,
+      "epoch": 2.9426604677983206,
+      "grad_norm": 0.23567576706409454,
+      "learning_rate": 1.9117852053627978e-05,
+      "loss": 0.4141,
       "step": 81650
     },
     {
-      "epoch": 2.87,
-      "learning_rate": 2.0221793250198738e-05,
-      "loss": 0.2445,
+      "epoch": 2.9428406674595453,
+      "grad_norm": 0.218300923705101,
+      "learning_rate": 1.911501588315743e-05,
+      "loss": 0.3508,
       "step": 81655
     },
     {
-      "epoch": 2.87,
-      "learning_rate": 2.02189970758941e-05,
-      "loss": 0.2694,
+      "epoch": 2.9430208671207696,
+      "grad_norm": 0.24794436991214752,
+      "learning_rate": 1.9112179792868662e-05,
+      "loss": 0.3964,
       "step": 81660
     },
     {
-      "epoch": 2.87,
-      "learning_rate": 2.0216200963667624e-05,
-      "loss": 0.2725,
+      "epoch": 2.9432010667819943,
+      "grad_norm": 0.1559266299009323,
+      "learning_rate": 1.9109343782800305e-05,
+      "loss": 0.4062,
       "step": 81665
     },
     {
-      "epoch": 2.87,
-      "learning_rate": 2.021340491355564e-05,
-      "loss": 0.266,
+      "epoch": 2.943381266443219,
+      "grad_norm": 0.25583502650260925,
+      "learning_rate": 1.910650785299099e-05,
+      "loss": 0.3751,
       "step": 81670
     },
     {
-      "epoch": 2.87,
-      "learning_rate": 2.0210608925594438e-05,
-      "loss": 0.249,
+      "epoch": 2.943561466104444,
+      "grad_norm": 0.2357344627380371,
+      "learning_rate": 1.9103672003479372e-05,
+      "loss": 0.3932,
       "step": 81675
     },
     {
-      "epoch": 2.87,
-      "learning_rate": 2.0207812999820316e-05,
-      "loss": 0.2599,
+      "epoch": 2.9437416657656685,
+      "grad_norm": 0.2820563018321991,
+      "learning_rate": 1.9100836234304076e-05,
+      "loss": 0.3736,
       "step": 81680
     },
     {
-      "epoch": 2.87,
-      "learning_rate": 2.0205017136269587e-05,
-      "loss": 0.2606,
+      "epoch": 2.9439218654268933,
+      "grad_norm": 0.227774977684021,
+      "learning_rate": 1.9098000545503745e-05,
+      "loss": 0.4056,
       "step": 81685
     },
     {
-      "epoch": 2.87,
-      "learning_rate": 2.0202221334978548e-05,
-      "loss": 0.2802,
+      "epoch": 2.9441020650881176,
+      "grad_norm": 0.23130254447460175,
+      "learning_rate": 1.909516493711702e-05,
+      "loss": 0.3757,
       "step": 81690
     },
     {
-      "epoch": 2.87,
-      "learning_rate": 2.0199425595983505e-05,
-      "loss": 0.247,
+      "epoch": 2.9442822647493423,
+      "grad_norm": 0.24160918593406677,
+      "learning_rate": 1.909232940918252e-05,
+      "loss": 0.4112,
       "step": 81695
     },
     {
-      "epoch": 2.87,
-      "learning_rate": 2.019662991932076e-05,
-      "loss": 0.2425,
+      "epoch": 2.944462464410567,
+      "grad_norm": 0.19820939004421234,
+      "learning_rate": 1.9089493961738896e-05,
+      "loss": 0.3657,
       "step": 81700
     },
     {
-      "epoch": 2.87,
-      "learning_rate": 2.0193834305026603e-05,
-      "loss": 0.2506,
+      "epoch": 2.9446426640717913,
+      "grad_norm": 0.21389785408973694,
+      "learning_rate": 1.9086658594824774e-05,
+      "loss": 0.3894,
       "step": 81705
     },
     {
-      "epoch": 2.87,
-      "learning_rate": 2.0191038753137344e-05,
-      "loss": 0.2518,
+      "epoch": 2.944822863733016,
+      "grad_norm": 0.24125300347805023,
+      "learning_rate": 1.9083823308478776e-05,
+      "loss": 0.4037,
       "step": 81710
     },
     {
-      "epoch": 2.87,
-      "learning_rate": 2.0188243263689276e-05,
-      "loss": 0.2504,
+      "epoch": 2.945003063394241,
+      "grad_norm": 0.23977577686309814,
+      "learning_rate": 1.9080988102739543e-05,
+      "loss": 0.4222,
       "step": 81715
     },
     {
-      "epoch": 2.88,
-      "learning_rate": 2.018544783671871e-05,
-      "loss": 0.2751,
+      "epoch": 2.9451832630554655,
+      "grad_norm": 0.2184176743030548,
+      "learning_rate": 1.90781529776457e-05,
+      "loss": 0.3853,
       "step": 81720
     },
     {
-      "epoch": 2.88,
-      "learning_rate": 2.018265247226191e-05,
-      "loss": 0.2701,
+      "epoch": 2.9453634627166903,
+      "grad_norm": 0.21365883946418762,
+      "learning_rate": 1.9075884935661265e-05,
+      "loss": 0.3671,
       "step": 81725
     },
     {
-      "epoch": 2.88,
-      "learning_rate": 2.0179857170355207e-05,
-      "loss": 0.2809,
+      "epoch": 2.945543662377915,
+      "grad_norm": 0.2217915952205658,
+      "learning_rate": 1.9073049955826478e-05,
+      "loss": 0.3974,
       "step": 81730
     },
     {
-      "epoch": 2.88,
-      "learning_rate": 2.0177061931034875e-05,
-      "loss": 0.2608,
+      "epoch": 2.9457238620391393,
+      "grad_norm": 0.2563817799091339,
+      "learning_rate": 1.9070215056745234e-05,
+      "loss": 0.4093,
       "step": 81735
     },
     {
-      "epoch": 2.88,
-      "learning_rate": 2.0174266754337217e-05,
-      "loss": 0.286,
+      "epoch": 2.945904061700364,
+      "grad_norm": 0.21083958446979523,
+      "learning_rate": 1.906738023845615e-05,
+      "loss": 0.3654,
       "step": 81740
     },
     {
-      "epoch": 2.88,
-      "learning_rate": 2.017147164029853e-05,
-      "loss": 0.2641,
+      "epoch": 2.9460842613615887,
+      "grad_norm": 0.2113693654537201,
+      "learning_rate": 1.906454550099787e-05,
+      "loss": 0.3854,
       "step": 81745
     },
     {
-      "epoch": 2.88,
-      "learning_rate": 2.01686765889551e-05,
-      "loss": 0.2808,
+      "epoch": 2.946264461022813,
+      "grad_norm": 0.2557770311832428,
+      "learning_rate": 1.9061710844409007e-05,
+      "loss": 0.3904,
       "step": 81750
     },
     {
-      "epoch": 2.88,
-      "learning_rate": 2.016588160034322e-05,
-      "loss": 0.2879,
+      "epoch": 2.9464446606840378,
+      "grad_norm": 0.23025763034820557,
+      "learning_rate": 1.9058876268728188e-05,
+      "loss": 0.3876,
       "step": 81755
     },
     {
-      "epoch": 2.88,
-      "learning_rate": 2.0163086674499175e-05,
-      "loss": 0.2807,
+      "epoch": 2.9466248603452625,
+      "grad_norm": 0.28059622645378113,
+      "learning_rate": 1.9056041773994025e-05,
+      "loss": 0.4097,
       "step": 81760
     },
     {
-      "epoch": 2.88,
-      "learning_rate": 2.016029181145927e-05,
-      "loss": 0.2345,
+      "epoch": 2.9468050600064872,
+      "grad_norm": 0.19796496629714966,
+      "learning_rate": 1.905320736024514e-05,
+      "loss": 0.4118,
       "step": 81765
     },
     {
-      "epoch": 2.88,
-      "learning_rate": 2.0157497011259785e-05,
-      "loss": 0.2527,
+      "epoch": 2.946985259667712,
+      "grad_norm": 0.19060267508029938,
+      "learning_rate": 1.9050373027520157e-05,
+      "loss": 0.3784,
       "step": 81770
     },
     {
-      "epoch": 2.88,
-      "learning_rate": 2.015470227393701e-05,
-      "loss": 0.2662,
+      "epoch": 2.9471654593289367,
+      "grad_norm": 0.1780025213956833,
+      "learning_rate": 1.9047538775857694e-05,
+      "loss": 0.384,
       "step": 81775
     },
     {
-      "epoch": 2.88,
-      "learning_rate": 2.0151907599527225e-05,
-      "loss": 0.2849,
+      "epoch": 2.947345658990161,
+      "grad_norm": 0.20874053239822388,
+      "learning_rate": 1.9044704605296353e-05,
+      "loss": 0.3901,
       "step": 81780
     },
     {
-      "epoch": 2.88,
-      "learning_rate": 2.0149112988066734e-05,
-      "loss": 0.2453,
+      "epoch": 2.9475258586513857,
+      "grad_norm": 0.18418976664543152,
+      "learning_rate": 1.9041870515874767e-05,
+      "loss": 0.365,
       "step": 81785
     },
     {
-      "epoch": 2.88,
-      "learning_rate": 2.014631843959181e-05,
-      "loss": 0.2547,
+      "epoch": 2.9477060583126105,
+      "grad_norm": 0.2519192695617676,
+      "learning_rate": 1.903903650763153e-05,
+      "loss": 0.3867,
       "step": 81790
     },
     {
-      "epoch": 2.88,
-      "learning_rate": 2.014352395413875e-05,
-      "loss": 0.2393,
+      "epoch": 2.9478862579738347,
+      "grad_norm": 0.20696429908275604,
+      "learning_rate": 1.903620258060528e-05,
+      "loss": 0.3652,
       "step": 81795
     },
     {
-      "epoch": 2.88,
-      "learning_rate": 2.0140729531743816e-05,
-      "loss": 0.265,
+      "epoch": 2.9480664576350595,
+      "grad_norm": 0.25634706020355225,
+      "learning_rate": 1.903336873483461e-05,
+      "loss": 0.366,
       "step": 81800
     },
     {
-      "epoch": 2.88,
-      "learning_rate": 2.013793517244332e-05,
-      "loss": 0.2536,
+      "epoch": 2.948246657296284,
+      "grad_norm": 0.17828556895256042,
+      "learning_rate": 1.9030534970358135e-05,
+      "loss": 0.438,
       "step": 81805
     },
     {
-      "epoch": 2.88,
-      "learning_rate": 2.013514087627353e-05,
-      "loss": 0.2445,
+      "epoch": 2.948426856957509,
+      "grad_norm": 0.21666783094406128,
+      "learning_rate": 1.9027701287214466e-05,
+      "loss": 0.4145,
       "step": 81810
     },
     {
-      "epoch": 2.88,
-      "learning_rate": 2.0132346643270725e-05,
-      "loss": 0.3005,
+      "epoch": 2.9486070566187337,
+      "grad_norm": 0.22079980373382568,
+      "learning_rate": 1.902486768544222e-05,
+      "loss": 0.372,
       "step": 81815
     },
     {
-      "epoch": 2.88,
-      "learning_rate": 2.012955247347119e-05,
-      "loss": 0.2731,
+      "epoch": 2.9487872562799584,
+      "grad_norm": 0.22830483317375183,
+      "learning_rate": 1.9022034165079977e-05,
+      "loss": 0.3996,
       "step": 81820
     },
     {
-      "epoch": 2.88,
-      "learning_rate": 2.0126758366911207e-05,
-      "loss": 0.2605,
+      "epoch": 2.9489674559411827,
+      "grad_norm": 0.20484139025211334,
+      "learning_rate": 1.901920072616638e-05,
+      "loss": 0.3791,
       "step": 81825
     },
     {
-      "epoch": 2.88,
-      "learning_rate": 2.012396432362706e-05,
-      "loss": 0.2733,
+      "epoch": 2.9491476556024074,
+      "grad_norm": 0.2294071912765503,
+      "learning_rate": 1.9016367368739994e-05,
+      "loss": 0.4122,
       "step": 81830
     },
     {
-      "epoch": 2.88,
-      "learning_rate": 2.0121170343655024e-05,
-      "loss": 0.2584,
+      "epoch": 2.949327855263632,
+      "grad_norm": 0.20010051131248474,
+      "learning_rate": 1.9013534092839463e-05,
+      "loss": 0.3743,
       "step": 81835
     },
     {
-      "epoch": 2.88,
-      "learning_rate": 2.0118376427031367e-05,
-      "loss": 0.2725,
+      "epoch": 2.949508054924857,
+      "grad_norm": 0.2721427381038666,
+      "learning_rate": 1.9010700898503362e-05,
+      "loss": 0.4054,
       "step": 81840
     },
     {
-      "epoch": 2.88,
-      "learning_rate": 2.0115582573792382e-05,
-      "loss": 0.2574,
+      "epoch": 2.949688254586081,
+      "grad_norm": 0.21804888546466827,
+      "learning_rate": 1.90078677857703e-05,
+      "loss": 0.4004,
       "step": 81845
     },
     {
-      "epoch": 2.88,
-      "learning_rate": 2.0112788783974335e-05,
-      "loss": 0.2502,
+      "epoch": 2.949868454247306,
+      "grad_norm": 0.20582912862300873,
+      "learning_rate": 1.9005034754678887e-05,
+      "loss": 0.3752,
       "step": 81850
     },
     {
-      "epoch": 2.88,
-      "learning_rate": 2.01099950576135e-05,
-      "loss": 0.2637,
+      "epoch": 2.9500486539085307,
+      "grad_norm": 0.22233439981937408,
+      "learning_rate": 1.900220180526771e-05,
+      "loss": 0.3631,
       "step": 81855
     },
     {
-      "epoch": 2.88,
-      "learning_rate": 2.0107201394746165e-05,
-      "loss": 0.2621,
+      "epoch": 2.9502288535697554,
+      "grad_norm": 0.20850253105163574,
+      "learning_rate": 1.8999368937575367e-05,
+      "loss": 0.3808,
       "step": 81860
     },
     {
-      "epoch": 2.88,
-      "learning_rate": 2.0104407795408596e-05,
-      "loss": 0.2777,
+      "epoch": 2.95040905323098,
+      "grad_norm": 0.23026061058044434,
+      "learning_rate": 1.8996536151640472e-05,
+      "loss": 0.4305,
       "step": 81865
     },
     {
-      "epoch": 2.88,
-      "learning_rate": 2.0101614259637065e-05,
-      "loss": 0.2842,
+      "epoch": 2.9505892528922044,
+      "grad_norm": 0.2322373390197754,
+      "learning_rate": 1.89937034475016e-05,
+      "loss": 0.4094,
       "step": 81870
     },
     {
-      "epoch": 2.88,
-      "learning_rate": 2.0098820787467835e-05,
-      "loss": 0.2921,
+      "epoch": 2.950769452553429,
+      "grad_norm": 0.2333720475435257,
+      "learning_rate": 1.899087082519736e-05,
+      "loss": 0.4051,
       "step": 81875
     },
     {
-      "epoch": 2.88,
-      "learning_rate": 2.0096027378937193e-05,
-      "loss": 0.2932,
+      "epoch": 2.950949652214654,
+      "grad_norm": 0.2258276641368866,
+      "learning_rate": 1.898803828476634e-05,
+      "loss": 0.4222,
       "step": 81880
     },
     {
-      "epoch": 2.88,
-      "learning_rate": 2.0093234034081406e-05,
-      "loss": 0.2608,
+      "epoch": 2.9511298518758786,
+      "grad_norm": 0.21611566841602325,
+      "learning_rate": 1.898520582624713e-05,
+      "loss": 0.4085,
       "step": 81885
     },
     {
-      "epoch": 2.88,
-      "learning_rate": 2.0090440752936744e-05,
-      "loss": 0.2765,
+      "epoch": 2.951310051537103,
+      "grad_norm": 0.21131113171577454,
+      "learning_rate": 1.8982373449678336e-05,
+      "loss": 0.413,
       "step": 81890
     },
     {
-      "epoch": 2.88,
-      "learning_rate": 2.008764753553946e-05,
-      "loss": 0.2695,
+      "epoch": 2.9514902511983276,
+      "grad_norm": 0.21428248286247253,
+      "learning_rate": 1.8979541155098528e-05,
+      "loss": 0.3984,
       "step": 81895
     },
     {
-      "epoch": 2.88,
-      "learning_rate": 2.008485438192585e-05,
-      "loss": 0.2474,
+      "epoch": 2.9516704508595524,
+      "grad_norm": 0.1810792088508606,
+      "learning_rate": 1.8976708942546316e-05,
+      "loss": 0.4067,
       "step": 81900
     },
     {
-      "epoch": 2.88,
-      "learning_rate": 2.008206129213216e-05,
-      "loss": 0.2591,
+      "epoch": 2.951850650520777,
+      "grad_norm": 0.20499011874198914,
+      "learning_rate": 1.897387681206028e-05,
+      "loss": 0.3654,
       "step": 81905
     },
     {
-      "epoch": 2.88,
-      "learning_rate": 2.0079268266194668e-05,
-      "loss": 0.279,
+      "epoch": 2.952030850182002,
+      "grad_norm": 0.2227361649274826,
+      "learning_rate": 1.8971044763678995e-05,
+      "loss": 0.3786,
       "step": 81910
     },
     {
-      "epoch": 2.88,
-      "learning_rate": 2.0076475304149624e-05,
-      "loss": 0.263,
+      "epoch": 2.952211049843226,
+      "grad_norm": 0.19713594019412994,
+      "learning_rate": 1.8968212797441064e-05,
+      "loss": 0.3725,
       "step": 81915
     },
     {
-      "epoch": 2.88,
-      "learning_rate": 2.0073682406033315e-05,
-      "loss": 0.2706,
+      "epoch": 2.952391249504451,
+      "grad_norm": 0.250509113073349,
+      "learning_rate": 1.896538091338507e-05,
+      "loss": 0.3395,
       "step": 81920
     },
     {
-      "epoch": 2.88,
-      "learning_rate": 2.0070889571881993e-05,
-      "loss": 0.2994,
+      "epoch": 2.9525714491656756,
+      "grad_norm": 0.1757110208272934,
+      "learning_rate": 1.8962549111549578e-05,
+      "loss": 0.4056,
       "step": 81925
     },
     {
-      "epoch": 2.88,
-      "learning_rate": 2.0068096801731915e-05,
-      "loss": 0.2746,
+      "epoch": 2.9527516488269003,
+      "grad_norm": 0.23518085479736328,
+      "learning_rate": 1.8959717391973204e-05,
+      "loss": 0.381,
       "step": 81930
     },
     {
-      "epoch": 2.88,
-      "learning_rate": 2.0065304095619344e-05,
-      "loss": 0.2926,
+      "epoch": 2.9529318484881246,
+      "grad_norm": 0.2347138673067093,
+      "learning_rate": 1.8956885754694495e-05,
+      "loss": 0.4076,
       "step": 81935
     },
     {
-      "epoch": 2.88,
-      "learning_rate": 2.006251145358056e-05,
-      "loss": 0.2821,
+      "epoch": 2.9531120481493494,
+      "grad_norm": 0.19113247096538544,
+      "learning_rate": 1.8954054199752063e-05,
+      "loss": 0.3994,
       "step": 81940
     },
     {
-      "epoch": 2.88,
-      "learning_rate": 2.0059718875651812e-05,
-      "loss": 0.2838,
+      "epoch": 2.953292247810574,
+      "grad_norm": 0.1934749335050583,
+      "learning_rate": 1.8951222727184466e-05,
+      "loss": 0.381,
       "step": 81945
     },
     {
-      "epoch": 2.88,
-      "learning_rate": 2.0056926361869354e-05,
-      "loss": 0.2792,
+      "epoch": 2.953472447471799,
+      "grad_norm": 0.21083234250545502,
+      "learning_rate": 1.894839133703028e-05,
+      "loss": 0.3682,
       "step": 81950
     },
     {
-      "epoch": 2.88,
-      "learning_rate": 2.0054692397053082e-05,
-      "loss": 0.2797,
+      "epoch": 2.9536526471330236,
+      "grad_norm": 0.22241094708442688,
+      "learning_rate": 1.89455600293281e-05,
+      "loss": 0.3661,
       "step": 81955
     },
     {
-      "epoch": 2.88,
-      "learning_rate": 2.0051899998825325e-05,
-      "loss": 0.2649,
+      "epoch": 2.9538328467942483,
+      "grad_norm": 0.17992718517780304,
+      "learning_rate": 1.894272880411649e-05,
+      "loss": 0.4289,
       "step": 81960
     },
     {
-      "epoch": 2.88,
-      "learning_rate": 2.0049107664845383e-05,
-      "loss": 0.2662,
+      "epoch": 2.9540130464554726,
+      "grad_norm": 0.20856983959674835,
+      "learning_rate": 1.8939897661434022e-05,
+      "loss": 0.412,
       "step": 81965
     },
     {
-      "epoch": 2.88,
-      "learning_rate": 2.0046315395149516e-05,
-      "loss": 0.2687,
+      "epoch": 2.9541932461166973,
+      "grad_norm": 0.21314571797847748,
+      "learning_rate": 1.893706660131928e-05,
+      "loss": 0.3842,
       "step": 81970
     },
     {
-      "epoch": 2.88,
-      "learning_rate": 2.0043523189773983e-05,
-      "loss": 0.2927,
+      "epoch": 2.954373445777922,
+      "grad_norm": 0.19296015799045563,
+      "learning_rate": 1.8934235623810833e-05,
+      "loss": 0.3784,
       "step": 81975
     },
     {
-      "epoch": 2.88,
-      "learning_rate": 2.0040731048755038e-05,
-      "loss": 0.2563,
+      "epoch": 2.9545536454391463,
+      "grad_norm": 0.24017472565174103,
+      "learning_rate": 1.8931404728947248e-05,
+      "loss": 0.4024,
       "step": 81980
     },
     {
-      "epoch": 2.88,
-      "learning_rate": 2.003793897212892e-05,
-      "loss": 0.265,
+      "epoch": 2.954733845100371,
+      "grad_norm": 0.27945849299430847,
+      "learning_rate": 1.89285739167671e-05,
+      "loss": 0.4157,
       "step": 81985
     },
     {
-      "epoch": 2.88,
-      "learning_rate": 2.0035146959931905e-05,
-      "loss": 0.2949,
+      "epoch": 2.954914044761596,
+      "grad_norm": 0.23929205536842346,
+      "learning_rate": 1.8925743187308954e-05,
+      "loss": 0.385,
       "step": 81990
     },
     {
-      "epoch": 2.88,
-      "learning_rate": 2.003235501220023e-05,
-      "loss": 0.2663,
+      "epoch": 2.9550942444228205,
+      "grad_norm": 0.20242777466773987,
+      "learning_rate": 1.8922912540611387e-05,
+      "loss": 0.4123,
       "step": 81995
     },
     {
-      "epoch": 2.88,
-      "learning_rate": 2.0029563128970156e-05,
-      "loss": 0.2588,
+      "epoch": 2.9552744440840453,
+      "grad_norm": 0.2415585070848465,
+      "learning_rate": 1.8920081976712963e-05,
+      "loss": 0.3732,
       "step": 82000
     },
     {
-      "epoch": 2.88,
-      "eval_loss": 0.2587546408176422,
-      "eval_runtime": 10.5627,
-      "eval_samples_per_second": 9.467,
-      "eval_steps_per_second": 9.467,
+      "epoch": 2.9552744440840453,
+      "eval_loss": 0.4305243194103241,
+      "eval_runtime": 3.5265,
+      "eval_samples_per_second": 28.356,
+      "eval_steps_per_second": 7.089,
       "step": 82000
     },
     {
-      "epoch": 2.89,
-      "learning_rate": 2.0026771310277916e-05,
-      "loss": 0.2722,
+      "epoch": 2.95545464374527,
+      "grad_norm": 0.16947853565216064,
+      "learning_rate": 1.8917251495652234e-05,
+      "loss": 0.38,
       "step": 82005
     },
     {
-      "epoch": 2.89,
-      "learning_rate": 2.002397955615979e-05,
-      "loss": 0.2583,
+      "epoch": 2.9556348434064943,
+      "grad_norm": 0.24770832061767578,
+      "learning_rate": 1.8914421097467787e-05,
+      "loss": 0.4288,
       "step": 82010
     },
     {
-      "epoch": 2.89,
-      "learning_rate": 2.0021187866651997e-05,
-      "loss": 0.2724,
+      "epoch": 2.955815043067719,
+      "grad_norm": 0.21834175288677216,
+      "learning_rate": 1.891159078219816e-05,
+      "loss": 0.3732,
       "step": 82015
     },
     {
-      "epoch": 2.89,
-      "learning_rate": 2.001839624179081e-05,
-      "loss": 0.246,
+      "epoch": 2.9559952427289438,
+      "grad_norm": 0.1806267648935318,
+      "learning_rate": 1.8908760549881943e-05,
+      "loss": 0.3953,
       "step": 82020
     },
     {
-      "epoch": 2.89,
-      "learning_rate": 2.001560468161245e-05,
-      "loss": 0.3009,
+      "epoch": 2.956175442390168,
+      "grad_norm": 0.26217013597488403,
+      "learning_rate": 1.890593040055768e-05,
+      "loss": 0.4025,
       "step": 82025
     },
     {
-      "epoch": 2.89,
-      "learning_rate": 2.001281318615319e-05,
-      "loss": 0.2928,
+      "epoch": 2.956355642051393,
+      "grad_norm": 0.15443582832813263,
+      "learning_rate": 1.8903100334263933e-05,
+      "loss": 0.3855,
       "step": 82030
     },
     {
-      "epoch": 2.89,
-      "learning_rate": 2.001002175544927e-05,
-      "loss": 0.2486,
+      "epoch": 2.9565358417126175,
+      "grad_norm": 0.21527697145938873,
+      "learning_rate": 1.8900270351039263e-05,
+      "loss": 0.3965,
       "step": 82035
     },
     {
-      "epoch": 2.89,
-      "learning_rate": 2.0007230389536916e-05,
-      "loss": 0.2804,
+      "epoch": 2.9567160413738423,
+      "grad_norm": 0.2215004861354828,
+      "learning_rate": 1.8897440450922234e-05,
+      "loss": 0.3823,
       "step": 82040
     },
     {
-      "epoch": 2.89,
-      "learning_rate": 2.0004439088452386e-05,
-      "loss": 0.2471,
+      "epoch": 2.956896241035067,
+      "grad_norm": 0.2365996241569519,
+      "learning_rate": 1.8894610633951392e-05,
+      "loss": 0.3848,
       "step": 82045
     },
     {
-      "epoch": 2.89,
-      "learning_rate": 2.0001647852231924e-05,
-      "loss": 0.2752,
+      "epoch": 2.9570764406962917,
+      "grad_norm": 0.2466432750225067,
+      "learning_rate": 1.88917809001653e-05,
+      "loss": 0.4159,
       "step": 82050
     },
     {
-      "epoch": 2.89,
-      "learning_rate": 1.9998856680911777e-05,
-      "loss": 0.2607,
+      "epoch": 2.957256640357516,
+      "grad_norm": 0.20089565217494965,
+      "learning_rate": 1.888895124960251e-05,
+      "loss": 0.3936,
       "step": 82055
     },
     {
-      "epoch": 2.89,
-      "learning_rate": 1.9996065574528168e-05,
-      "loss": 0.2797,
+      "epoch": 2.9574368400187407,
+      "grad_norm": 0.23963125050067902,
+      "learning_rate": 1.8886121682301576e-05,
+      "loss": 0.3979,
       "step": 82060
     },
     {
-      "epoch": 2.89,
-      "learning_rate": 1.9993274533117363e-05,
-      "loss": 0.2755,
+      "epoch": 2.9576170396799655,
+      "grad_norm": 0.2524862289428711,
+      "learning_rate": 1.8883292198301056e-05,
+      "loss": 0.3805,
       "step": 82065
     },
     {
-      "epoch": 2.89,
-      "learning_rate": 1.9990483556715587e-05,
-      "loss": 0.2648,
+      "epoch": 2.9577972393411898,
+      "grad_norm": 0.18411079049110413,
+      "learning_rate": 1.8880462797639487e-05,
+      "loss": 0.3945,
       "step": 82070
     },
     {
-      "epoch": 2.89,
-      "learning_rate": 1.9987692645359075e-05,
-      "loss": 0.2684,
+      "epoch": 2.9579774390024145,
+      "grad_norm": 0.20663435757160187,
+      "learning_rate": 1.8877633480355434e-05,
+      "loss": 0.3686,
       "step": 82075
     },
     {
-      "epoch": 2.89,
-      "learning_rate": 1.9984901799084064e-05,
-      "loss": 0.2473,
+      "epoch": 2.9581576386636392,
+      "grad_norm": 0.20612770318984985,
+      "learning_rate": 1.8874804246487437e-05,
+      "loss": 0.4012,
       "step": 82080
     },
     {
-      "epoch": 2.89,
-      "learning_rate": 1.9982111017926812e-05,
-      "loss": 0.2687,
+      "epoch": 2.958337838324864,
+      "grad_norm": 0.18578004837036133,
+      "learning_rate": 1.887197509607404e-05,
+      "loss": 0.3866,
       "step": 82085
     },
     {
-      "epoch": 2.89,
-      "learning_rate": 1.9979320301923537e-05,
-      "loss": 0.2566,
+      "epoch": 2.9585180379860887,
+      "grad_norm": 0.19520944356918335,
+      "learning_rate": 1.8869146029153805e-05,
+      "loss": 0.4054,
       "step": 82090
     },
     {
-      "epoch": 2.89,
-      "learning_rate": 1.9976529651110483e-05,
-      "loss": 0.2715,
+      "epoch": 2.9586982376473134,
+      "grad_norm": 0.2129872888326645,
+      "learning_rate": 1.8866317045765264e-05,
+      "loss": 0.3716,
       "step": 82095
     },
     {
-      "epoch": 2.89,
-      "learning_rate": 1.9973739065523865e-05,
-      "loss": 0.2655,
+      "epoch": 2.9588784373085377,
+      "grad_norm": 0.2196713089942932,
+      "learning_rate": 1.8863488145946965e-05,
+      "loss": 0.3829,
       "step": 82100
     },
     {
-      "epoch": 2.89,
-      "learning_rate": 1.9970948545199943e-05,
-      "loss": 0.2656,
+      "epoch": 2.9590586369697625,
+      "grad_norm": 0.21013066172599792,
+      "learning_rate": 1.8860659329737467e-05,
+      "loss": 0.403,
       "step": 82105
     },
     {
-      "epoch": 2.89,
-      "learning_rate": 1.9968158090174946e-05,
-      "loss": 0.2702,
+      "epoch": 2.959238836630987,
+      "grad_norm": 0.19054462015628815,
+      "learning_rate": 1.8857830597175273e-05,
+      "loss": 0.3886,
       "step": 82110
     },
     {
-      "epoch": 2.89,
-      "learning_rate": 1.99653677004851e-05,
-      "loss": 0.2788,
+      "epoch": 2.959419036292212,
+      "grad_norm": 0.2173452377319336,
+      "learning_rate": 1.8855001948298966e-05,
+      "loss": 0.3946,
       "step": 82115
     },
     {
-      "epoch": 2.89,
-      "learning_rate": 1.9962577376166622e-05,
-      "loss": 0.2872,
+      "epoch": 2.959599235953436,
+      "grad_norm": 0.22703658044338226,
+      "learning_rate": 1.885217338314705e-05,
+      "loss": 0.3912,
       "step": 82120
     },
     {
-      "epoch": 2.89,
-      "learning_rate": 1.995978711725577e-05,
-      "loss": 0.2743,
+      "epoch": 2.959779435614661,
+      "grad_norm": 0.1947726011276245,
+      "learning_rate": 1.8849344901758102e-05,
+      "loss": 0.3932,
       "step": 82125
     },
     {
-      "epoch": 2.89,
-      "learning_rate": 1.9956996923788757e-05,
-      "loss": 0.2629,
+      "epoch": 2.9599596352758857,
+      "grad_norm": 0.23506419360637665,
+      "learning_rate": 1.884651650417063e-05,
+      "loss": 0.3539,
       "step": 82130
     },
     {
-      "epoch": 2.89,
-      "learning_rate": 1.9954206795801815e-05,
-      "loss": 0.2858,
+      "epoch": 2.9601398349371104,
+      "grad_norm": 0.23253336548805237,
+      "learning_rate": 1.8843688190423175e-05,
+      "loss": 0.3832,
       "step": 82135
     },
     {
-      "epoch": 2.89,
-      "learning_rate": 1.995141673333117e-05,
-      "loss": 0.2714,
+      "epoch": 2.960320034598335,
+      "grad_norm": 0.19187670946121216,
+      "learning_rate": 1.884085996055428e-05,
+      "loss": 0.3612,
       "step": 82140
     },
     {
-      "epoch": 2.89,
-      "learning_rate": 1.9948626736413055e-05,
-      "loss": 0.2657,
+      "epoch": 2.9605002342595594,
+      "grad_norm": 0.2828364670276642,
+      "learning_rate": 1.883803181460248e-05,
+      "loss": 0.3848,
       "step": 82145
     },
     {
-      "epoch": 2.89,
-      "learning_rate": 1.9945836805083695e-05,
-      "loss": 0.2455,
+      "epoch": 2.960680433920784,
+      "grad_norm": 0.19735077023506165,
+      "learning_rate": 1.8835203752606294e-05,
+      "loss": 0.3615,
       "step": 82150
     },
     {
-      "epoch": 2.89,
-      "learning_rate": 1.994304693937931e-05,
-      "loss": 0.2824,
+      "epoch": 2.960860633582009,
+      "grad_norm": 0.2174484133720398,
+      "learning_rate": 1.8832375774604272e-05,
+      "loss": 0.3571,
       "step": 82155
     },
     {
-      "epoch": 2.89,
-      "learning_rate": 1.9940257139336122e-05,
-      "loss": 0.2578,
+      "epoch": 2.9610408332432336,
+      "grad_norm": 0.1863182634115219,
+      "learning_rate": 1.882954788063493e-05,
+      "loss": 0.3636,
       "step": 82160
     },
     {
-      "epoch": 2.89,
-      "learning_rate": 1.993746740499037e-05,
-      "loss": 0.2316,
+      "epoch": 2.961221032904458,
+      "grad_norm": 0.1913149356842041,
+      "learning_rate": 1.8826720070736804e-05,
+      "loss": 0.3881,
       "step": 82165
     },
     {
-      "epoch": 2.89,
-      "learning_rate": 1.9934677736378265e-05,
-      "loss": 0.2439,
+      "epoch": 2.9614012325656827,
+      "grad_norm": 0.21392560005187988,
+      "learning_rate": 1.8823892344948428e-05,
+      "loss": 0.3875,
       "step": 82170
     },
     {
-      "epoch": 2.89,
-      "learning_rate": 1.9931888133536027e-05,
-      "loss": 0.2705,
+      "epoch": 2.9615814322269074,
+      "grad_norm": 0.18154218792915344,
+      "learning_rate": 1.8821064703308315e-05,
+      "loss": 0.4115,
       "step": 82175
     },
     {
-      "epoch": 2.89,
-      "learning_rate": 1.9929098596499887e-05,
-      "loss": 0.2591,
+      "epoch": 2.961761631888132,
+      "grad_norm": 0.2097567617893219,
+      "learning_rate": 1.8818237145855004e-05,
+      "loss": 0.3887,
       "step": 82180
     },
     {
-      "epoch": 2.89,
-      "learning_rate": 1.9926309125306057e-05,
-      "loss": 0.2901,
+      "epoch": 2.961941831549357,
+      "grad_norm": 0.218689426779747,
+      "learning_rate": 1.881540967262702e-05,
+      "loss": 0.3997,
       "step": 82185
     },
     {
-      "epoch": 2.89,
-      "learning_rate": 1.9923519719990763e-05,
-      "loss": 0.2306,
+      "epoch": 2.9621220312105816,
+      "grad_norm": 0.19388610124588013,
+      "learning_rate": 1.8812582283662865e-05,
+      "loss": 0.3715,
       "step": 82190
     },
     {
-      "epoch": 2.89,
-      "learning_rate": 1.9920730380590212e-05,
-      "loss": 0.2629,
+      "epoch": 2.962302230871806,
+      "grad_norm": 0.19577103853225708,
+      "learning_rate": 1.88097549790011e-05,
+      "loss": 0.3694,
       "step": 82195
     },
     {
-      "epoch": 2.89,
-      "learning_rate": 1.9917941107140635e-05,
-      "loss": 0.2528,
+      "epoch": 2.9624824305330306,
+      "grad_norm": 0.21979494392871857,
+      "learning_rate": 1.8806927758680203e-05,
+      "loss": 0.4049,
       "step": 82200
     },
     {
-      "epoch": 2.89,
-      "learning_rate": 1.991515189967825e-05,
-      "loss": 0.2852,
+      "epoch": 2.9626626301942554,
+      "grad_norm": 0.19626237452030182,
+      "learning_rate": 1.880410062273873e-05,
+      "loss": 0.3738,
       "step": 82205
     },
     {
-      "epoch": 2.89,
-      "learning_rate": 1.9912362758239258e-05,
-      "loss": 0.2471,
+      "epoch": 2.9628428298554796,
+      "grad_norm": 0.21775542199611664,
+      "learning_rate": 1.8801273571215193e-05,
+      "loss": 0.3772,
       "step": 82210
     },
     {
-      "epoch": 2.89,
-      "learning_rate": 1.990957368285988e-05,
-      "loss": 0.2607,
+      "epoch": 2.9630230295167044,
+      "grad_norm": 0.16890302300453186,
+      "learning_rate": 1.8798446604148085e-05,
+      "loss": 0.3957,
       "step": 82215
     },
     {
-      "epoch": 2.89,
-      "learning_rate": 1.990678467357634e-05,
-      "loss": 0.2535,
+      "epoch": 2.963203229177929,
+      "grad_norm": 0.2454291582107544,
+      "learning_rate": 1.879561972157596e-05,
+      "loss": 0.367,
       "step": 82220
     },
     {
-      "epoch": 2.89,
-      "learning_rate": 1.9903995730424846e-05,
-      "loss": 0.2794,
+      "epoch": 2.963383428839154,
+      "grad_norm": 0.21003590524196625,
+      "learning_rate": 1.8792792923537304e-05,
+      "loss": 0.3988,
       "step": 82225
     },
     {
-      "epoch": 2.89,
-      "learning_rate": 1.9901206853441607e-05,
-      "loss": 0.2707,
+      "epoch": 2.9635636285003786,
+      "grad_norm": 0.21214927732944489,
+      "learning_rate": 1.878996621007064e-05,
+      "loss": 0.3633,
       "step": 82230
     },
     {
-      "epoch": 2.89,
-      "learning_rate": 1.989841804266283e-05,
-      "loss": 0.2502,
+      "epoch": 2.9637438281616033,
+      "grad_norm": 0.24998506903648376,
+      "learning_rate": 1.878713958121449e-05,
+      "loss": 0.3908,
       "step": 82235
     },
     {
-      "epoch": 2.89,
-      "learning_rate": 1.9895629298124734e-05,
-      "loss": 0.2555,
+      "epoch": 2.9639240278228276,
+      "grad_norm": 0.21911552548408508,
+      "learning_rate": 1.8784313037007355e-05,
+      "loss": 0.3708,
       "step": 82240
     },
     {
-      "epoch": 2.89,
-      "learning_rate": 1.9892840619863535e-05,
-      "loss": 0.2574,
+      "epoch": 2.9641042274840523,
+      "grad_norm": 0.24015755951404572,
+      "learning_rate": 1.8781486577487763e-05,
+      "loss": 0.43,
       "step": 82245
     },
     {
-      "epoch": 2.89,
-      "learning_rate": 1.9890052007915435e-05,
-      "loss": 0.2736,
+      "epoch": 2.964284427145277,
+      "grad_norm": 0.17950226366519928,
+      "learning_rate": 1.8778660202694203e-05,
+      "loss": 0.3546,
       "step": 82250
     },
     {
-      "epoch": 2.89,
-      "learning_rate": 1.988726346231663e-05,
-      "loss": 0.2708,
+      "epoch": 2.9644646268065014,
+      "grad_norm": 0.14863833785057068,
+      "learning_rate": 1.877583391266519e-05,
+      "loss": 0.3852,
       "step": 82255
     },
     {
-      "epoch": 2.89,
-      "learning_rate": 1.988447498310335e-05,
-      "loss": 0.2763,
+      "epoch": 2.964644826467726,
+      "grad_norm": 0.207986518740654,
+      "learning_rate": 1.8773007707439245e-05,
+      "loss": 0.3883,
       "step": 82260
     },
     {
-      "epoch": 2.89,
-      "learning_rate": 1.9881686570311788e-05,
-      "loss": 0.2744,
+      "epoch": 2.964825026128951,
+      "grad_norm": 0.1966986507177353,
+      "learning_rate": 1.877018158705486e-05,
+      "loss": 0.3868,
       "step": 82265
     },
     {
-      "epoch": 2.89,
-      "learning_rate": 1.987889822397816e-05,
-      "loss": 0.2648,
+      "epoch": 2.9650052257901756,
+      "grad_norm": 0.19852174818515778,
+      "learning_rate": 1.876735555155054e-05,
+      "loss": 0.429,
       "step": 82270
     },
     {
-      "epoch": 2.89,
-      "learning_rate": 1.987610994413865e-05,
-      "loss": 0.2605,
+      "epoch": 2.9651854254514003,
+      "grad_norm": 0.2330932915210724,
+      "learning_rate": 1.87645296009648e-05,
+      "loss": 0.4144,
       "step": 82275
     },
     {
-      "epoch": 2.89,
-      "learning_rate": 1.9873321730829486e-05,
-      "loss": 0.264,
+      "epoch": 2.965365625112625,
+      "grad_norm": 0.2199668288230896,
+      "learning_rate": 1.8761703735336134e-05,
+      "loss": 0.4,
       "step": 82280
     },
     {
-      "epoch": 2.9,
-      "learning_rate": 1.987053358408686e-05,
-      "loss": 0.2632,
+      "epoch": 2.9655458247738493,
+      "grad_norm": 0.24273623526096344,
+      "learning_rate": 1.8758877954703047e-05,
+      "loss": 0.3756,
       "step": 82285
     },
     {
-      "epoch": 2.9,
-      "learning_rate": 1.9867745503946968e-05,
-      "loss": 0.2622,
+      "epoch": 2.965726024435074,
+      "grad_norm": 0.2339378446340561,
+      "learning_rate": 1.8756052259104048e-05,
+      "loss": 0.4029,
       "step": 82290
     },
     {
-      "epoch": 2.9,
-      "learning_rate": 1.9864957490446018e-05,
-      "loss": 0.2775,
+      "epoch": 2.965906224096299,
+      "grad_norm": 0.1760188341140747,
+      "learning_rate": 1.8753226648577613e-05,
+      "loss": 0.3748,
       "step": 82295
     },
     {
-      "epoch": 2.9,
-      "learning_rate": 1.9862169543620222e-05,
-      "loss": 0.2779,
+      "epoch": 2.966086423757523,
+      "grad_norm": 0.24585866928100586,
+      "learning_rate": 1.875040112316227e-05,
+      "loss": 0.4141,
       "step": 82300
     },
     {
-      "epoch": 2.9,
-      "learning_rate": 1.9859381663505763e-05,
-      "loss": 0.2398,
+      "epoch": 2.966266623418748,
+      "grad_norm": 0.2293357104063034,
+      "learning_rate": 1.8747575682896483e-05,
+      "loss": 0.3367,
       "step": 82305
     },
     {
-      "epoch": 2.9,
-      "learning_rate": 1.9856593850138837e-05,
-      "loss": 0.2947,
+      "epoch": 2.9664468230799725,
+      "grad_norm": 0.19390980899333954,
+      "learning_rate": 1.874475032781879e-05,
+      "loss": 0.3877,
       "step": 82310
     },
     {
-      "epoch": 2.9,
-      "learning_rate": 1.9853806103555662e-05,
-      "loss": 0.2575,
+      "epoch": 2.9666270227411973,
+      "grad_norm": 0.20741982758045197,
+      "learning_rate": 1.874192505796765e-05,
+      "loss": 0.4016,
       "step": 82315
     },
     {
-      "epoch": 2.9,
-      "learning_rate": 1.9851018423792416e-05,
-      "loss": 0.2662,
+      "epoch": 2.966807222402422,
+      "grad_norm": 0.21530795097351074,
+      "learning_rate": 1.873909987338157e-05,
+      "loss": 0.3675,
       "step": 82320
     },
     {
-      "epoch": 2.9,
-      "learning_rate": 1.984823081088531e-05,
-      "loss": 0.2516,
+      "epoch": 2.9669874220636467,
+      "grad_norm": 0.2282307744026184,
+      "learning_rate": 1.8736274774099044e-05,
+      "loss": 0.3606,
       "step": 82325
     },
     {
-      "epoch": 2.9,
-      "learning_rate": 1.9845443264870522e-05,
-      "loss": 0.2661,
+      "epoch": 2.967167621724871,
+      "grad_norm": 0.22964274883270264,
+      "learning_rate": 1.873344976015856e-05,
+      "loss": 0.3933,
       "step": 82330
     },
     {
-      "epoch": 2.9,
-      "learning_rate": 1.9842655785784265e-05,
-      "loss": 0.2357,
+      "epoch": 2.9673478213860958,
+      "grad_norm": 0.2597173750400543,
+      "learning_rate": 1.873062483159861e-05,
+      "loss": 0.3803,
       "step": 82335
     },
     {
-      "epoch": 2.9,
-      "learning_rate": 1.9839868373662722e-05,
-      "loss": 0.2718,
+      "epoch": 2.9675280210473205,
+      "grad_norm": 0.19982276856899261,
+      "learning_rate": 1.8727799988457682e-05,
+      "loss": 0.4066,
       "step": 82340
     },
     {
-      "epoch": 2.9,
-      "learning_rate": 1.9837081028542088e-05,
-      "loss": 0.271,
+      "epoch": 2.9677082207085452,
+      "grad_norm": 0.19702062010765076,
+      "learning_rate": 1.8724975230774265e-05,
+      "loss": 0.3729,
       "step": 82345
     },
     {
-      "epoch": 2.9,
-      "learning_rate": 1.983429375045855e-05,
-      "loss": 0.2681,
+      "epoch": 2.9678884203697695,
+      "grad_norm": 0.19733615219593048,
+      "learning_rate": 1.872215055858685e-05,
+      "loss": 0.4182,
       "step": 82350
     },
     {
-      "epoch": 2.9,
-      "learning_rate": 1.9831506539448302e-05,
-      "loss": 0.2809,
+      "epoch": 2.9680686200309943,
+      "grad_norm": 0.21504078805446625,
+      "learning_rate": 1.8719325971933912e-05,
+      "loss": 0.3605,
       "step": 82355
     },
     {
-      "epoch": 2.9,
-      "learning_rate": 1.9828719395547545e-05,
-      "loss": 0.2644,
+      "epoch": 2.968248819692219,
+      "grad_norm": 0.19036151468753815,
+      "learning_rate": 1.8716501470853942e-05,
+      "loss": 0.4204,
       "step": 82360
     },
     {
-      "epoch": 2.9,
-      "learning_rate": 1.9825932318792454e-05,
-      "loss": 0.2748,
+      "epoch": 2.9684290193534437,
+      "grad_norm": 0.2703343331813812,
+      "learning_rate": 1.8713677055385425e-05,
+      "loss": 0.3874,
       "step": 82365
     },
     {
-      "epoch": 2.9,
-      "learning_rate": 1.9823145309219213e-05,
-      "loss": 0.2707,
+      "epoch": 2.9686092190146685,
+      "grad_norm": 0.20502404868602753,
+      "learning_rate": 1.8710852725566847e-05,
+      "loss": 0.3794,
       "step": 82370
     },
     {
-      "epoch": 2.9,
-      "learning_rate": 1.9820358366864026e-05,
-      "loss": 0.2676,
+      "epoch": 2.9687894186758927,
+      "grad_norm": 0.2101486623287201,
+      "learning_rate": 1.8708028481436672e-05,
+      "loss": 0.4003,
       "step": 82375
     },
     {
-      "epoch": 2.9,
-      "learning_rate": 1.9817571491763076e-05,
-      "loss": 0.249,
+      "epoch": 2.9689696183371175,
+      "grad_norm": 0.19613702595233917,
+      "learning_rate": 1.8705204323033398e-05,
+      "loss": 0.3774,
       "step": 82380
     },
     {
-      "epoch": 2.9,
-      "learning_rate": 1.9814784683952544e-05,
-      "loss": 0.2791,
+      "epoch": 2.969149817998342,
+      "grad_norm": 0.22090743482112885,
+      "learning_rate": 1.870238025039549e-05,
+      "loss": 0.4098,
       "step": 82385
     },
     {
-      "epoch": 2.9,
-      "learning_rate": 1.9811997943468605e-05,
-      "loss": 0.2651,
+      "epoch": 2.969330017659567,
+      "grad_norm": 0.2624460756778717,
+      "learning_rate": 1.869955626356144e-05,
+      "loss": 0.3689,
       "step": 82390
     },
     {
-      "epoch": 2.9,
-      "learning_rate": 1.9809211270347465e-05,
-      "loss": 0.2612,
+      "epoch": 2.9695102173207912,
+      "grad_norm": 0.2443309724330902,
+      "learning_rate": 1.869673236256972e-05,
+      "loss": 0.4143,
       "step": 82395
     },
     {
-      "epoch": 2.9,
-      "learning_rate": 1.980642466462529e-05,
-      "loss": 0.27,
+      "epoch": 2.969690416982016,
+      "grad_norm": 0.2217528223991394,
+      "learning_rate": 1.8693908547458782e-05,
+      "loss": 0.386,
       "step": 82400
     },
     {
-      "epoch": 2.9,
-      "learning_rate": 1.980363812633827e-05,
-      "loss": 0.2555,
+      "epoch": 2.9698706166432407,
+      "grad_norm": 0.19909748435020447,
+      "learning_rate": 1.869108481826714e-05,
+      "loss": 0.4024,
       "step": 82405
     },
     {
-      "epoch": 2.9,
-      "learning_rate": 1.980085165552258e-05,
-      "loss": 0.2816,
+      "epoch": 2.9700508163044654,
+      "grad_norm": 0.23337341845035553,
+      "learning_rate": 1.8688261175033238e-05,
+      "loss": 0.4011,
       "step": 82410
     },
     {
-      "epoch": 2.9,
-      "learning_rate": 1.9798065252214415e-05,
-      "loss": 0.2741,
+      "epoch": 2.97023101596569,
+      "grad_norm": 0.20584703981876373,
+      "learning_rate": 1.868543761779555e-05,
+      "loss": 0.4166,
       "step": 82415
     },
     {
-      "epoch": 2.9,
-      "learning_rate": 1.979527891644994e-05,
-      "loss": 0.2858,
+      "epoch": 2.9704112156269145,
+      "grad_norm": 0.17910712957382202,
+      "learning_rate": 1.8682614146592558e-05,
+      "loss": 0.3962,
       "step": 82420
     },
     {
-      "epoch": 2.9,
-      "learning_rate": 1.979249264826533e-05,
-      "loss": 0.263,
+      "epoch": 2.970591415288139,
+      "grad_norm": 0.16632866859436035,
+      "learning_rate": 1.8679790761462717e-05,
+      "loss": 0.388,
       "step": 82425
     },
     {
-      "epoch": 2.9,
-      "learning_rate": 1.9789706447696785e-05,
-      "loss": 0.2587,
+      "epoch": 2.970771614949364,
+      "grad_norm": 0.24808615446090698,
+      "learning_rate": 1.8676967462444513e-05,
+      "loss": 0.3801,
       "step": 82430
     },
     {
-      "epoch": 2.9,
-      "learning_rate": 1.9786920314780462e-05,
-      "loss": 0.2738,
+      "epoch": 2.9709518146105887,
+      "grad_norm": 0.2281518578529358,
+      "learning_rate": 1.8674144249576403e-05,
+      "loss": 0.363,
       "step": 82435
     },
     {
-      "epoch": 2.9,
-      "learning_rate": 1.9784134249552544e-05,
-      "loss": 0.2599,
+      "epoch": 2.971132014271813,
+      "grad_norm": 0.14790934324264526,
+      "learning_rate": 1.8671321122896847e-05,
+      "loss": 0.3396,
       "step": 82440
     },
     {
-      "epoch": 2.9,
-      "learning_rate": 1.9781348252049196e-05,
-      "loss": 0.2548,
+      "epoch": 2.9713122139330377,
+      "grad_norm": 0.2927576005458832,
+      "learning_rate": 1.866849808244432e-05,
+      "loss": 0.3867,
       "step": 82445
     },
     {
-      "epoch": 2.9,
-      "learning_rate": 1.977856232230661e-05,
-      "loss": 0.2498,
+      "epoch": 2.9714924135942624,
+      "grad_norm": 0.24208620190620422,
+      "learning_rate": 1.866567512825728e-05,
+      "loss": 0.4101,
       "step": 82450
     },
     {
-      "epoch": 2.9,
-      "learning_rate": 1.9775776460360952e-05,
-      "loss": 0.2721,
+      "epoch": 2.971672613255487,
+      "grad_norm": 0.19803206622600555,
+      "learning_rate": 1.8662852260374183e-05,
+      "loss": 0.3986,
       "step": 82455
     },
     {
-      "epoch": 2.9,
-      "learning_rate": 1.97729906662484e-05,
-      "loss": 0.2808,
+      "epoch": 2.971852812916712,
+      "grad_norm": 0.22639377415180206,
+      "learning_rate": 1.866002947883351e-05,
+      "loss": 0.4211,
       "step": 82460
     },
     {
-      "epoch": 2.9,
-      "learning_rate": 1.9770204940005103e-05,
-      "loss": 0.2835,
+      "epoch": 2.9720330125779366,
+      "grad_norm": 0.20173002779483795,
+      "learning_rate": 1.865720678367369e-05,
+      "loss": 0.3984,
       "step": 82465
     },
     {
-      "epoch": 2.9,
-      "learning_rate": 1.976741928166726e-05,
-      "loss": 0.2708,
+      "epoch": 2.972213212239161,
+      "grad_norm": 0.28004732728004456,
+      "learning_rate": 1.8654384174933214e-05,
+      "loss": 0.4313,
       "step": 82470
     },
     {
-      "epoch": 2.9,
-      "learning_rate": 1.976463369127103e-05,
-      "loss": 0.2586,
+      "epoch": 2.9723934119003856,
+      "grad_norm": 0.19623763859272003,
+      "learning_rate": 1.865156165265053e-05,
+      "loss": 0.408,
       "step": 82475
     },
     {
-      "epoch": 2.9,
-      "learning_rate": 1.9761848168852575e-05,
-      "loss": 0.2569,
+      "epoch": 2.9725736115616104,
+      "grad_norm": 0.17043015360832214,
+      "learning_rate": 1.864873921686407e-05,
+      "loss": 0.3979,
       "step": 82480
     },
     {
-      "epoch": 2.9,
-      "learning_rate": 1.975906271444807e-05,
-      "loss": 0.2628,
+      "epoch": 2.9727538112228347,
+      "grad_norm": 0.2070167362689972,
+      "learning_rate": 1.864591686761233e-05,
+      "loss": 0.3852,
       "step": 82485
     },
     {
-      "epoch": 2.9,
-      "learning_rate": 1.9756277328093685e-05,
-      "loss": 0.3001,
+      "epoch": 2.9729340108840594,
+      "grad_norm": 0.21379338204860687,
+      "learning_rate": 1.864309460493372e-05,
+      "loss": 0.4143,
       "step": 82490
     },
     {
-      "epoch": 2.9,
-      "learning_rate": 1.9753492009825587e-05,
-      "loss": 0.2645,
+      "epoch": 2.973114210545284,
+      "grad_norm": 0.2483922243118286,
+      "learning_rate": 1.864027242886673e-05,
+      "loss": 0.4423,
       "step": 82495
     },
     {
-      "epoch": 2.9,
-      "learning_rate": 1.9750706759679934e-05,
-      "loss": 0.2419,
+      "epoch": 2.973294410206509,
+      "grad_norm": 0.2155703455209732,
+      "learning_rate": 1.86374503394498e-05,
+      "loss": 0.4044,
       "step": 82500
     },
     {
-      "epoch": 2.9,
-      "eval_loss": 0.25884824991226196,
-      "eval_runtime": 10.5461,
-      "eval_samples_per_second": 9.482,
-      "eval_steps_per_second": 9.482,
+      "epoch": 2.973294410206509,
+      "eval_loss": 0.4298676550388336,
+      "eval_runtime": 3.5326,
+      "eval_samples_per_second": 28.308,
+      "eval_steps_per_second": 7.077,
       "step": 82500
     },
     {
-      "epoch": 2.9,
-      "learning_rate": 1.9747921577692884e-05,
-      "loss": 0.266,
+      "epoch": 2.9734746098677336,
+      "grad_norm": 0.24112021923065186,
+      "learning_rate": 1.863462833672136e-05,
+      "loss": 0.3763,
       "step": 82505
     },
     {
-      "epoch": 2.9,
-      "learning_rate": 1.9745136463900626e-05,
-      "loss": 0.2801,
+      "epoch": 2.9736548095289583,
+      "grad_norm": 0.19381298124790192,
+      "learning_rate": 1.8631806420719896e-05,
+      "loss": 0.3785,
       "step": 82510
     },
     {
-      "epoch": 2.9,
-      "learning_rate": 1.97423514183393e-05,
-      "loss": 0.2743,
+      "epoch": 2.9738350091901826,
+      "grad_norm": 0.21174710988998413,
+      "learning_rate": 1.862898459148383e-05,
+      "loss": 0.3688,
       "step": 82515
     },
     {
-      "epoch": 2.9,
-      "learning_rate": 1.9739566441045072e-05,
-      "loss": 0.2633,
+      "epoch": 2.9740152088514074,
+      "grad_norm": 0.18652644753456116,
+      "learning_rate": 1.862616284905161e-05,
+      "loss": 0.3974,
       "step": 82520
     },
     {
-      "epoch": 2.9,
-      "learning_rate": 1.9736781532054115e-05,
-      "loss": 0.2905,
+      "epoch": 2.974195408512632,
+      "grad_norm": 0.24465037882328033,
+      "learning_rate": 1.8623341193461696e-05,
+      "loss": 0.4212,
       "step": 82525
     },
     {
-      "epoch": 2.9,
-      "learning_rate": 1.9733996691402583e-05,
-      "loss": 0.2409,
+      "epoch": 2.9743756081738564,
+      "grad_norm": 0.19063496589660645,
+      "learning_rate": 1.8620519624752512e-05,
+      "loss": 0.3724,
       "step": 82530
     },
     {
-      "epoch": 2.9,
-      "learning_rate": 1.9731211919126634e-05,
-      "loss": 0.275,
+      "epoch": 2.974555807835081,
+      "grad_norm": 0.21079134941101074,
+      "learning_rate": 1.8617698142962523e-05,
+      "loss": 0.401,
       "step": 82535
     },
     {
-      "epoch": 2.9,
-      "learning_rate": 1.9728427215262415e-05,
-      "loss": 0.2609,
+      "epoch": 2.974736007496306,
+      "grad_norm": 0.22930555045604706,
+      "learning_rate": 1.861487674813016e-05,
+      "loss": 0.3471,
       "step": 82540
     },
     {
-      "epoch": 2.9,
-      "learning_rate": 1.97256425798461e-05,
-      "loss": 0.272,
+      "epoch": 2.9749162071575306,
+      "grad_norm": 0.22107750177383423,
+      "learning_rate": 1.8612055440293856e-05,
+      "loss": 0.3809,
       "step": 82545
     },
     {
-      "epoch": 2.9,
-      "learning_rate": 1.9722858012913843e-05,
-      "loss": 0.2581,
+      "epoch": 2.9750964068187553,
+      "grad_norm": 0.21329925954341888,
+      "learning_rate": 1.8609234219492072e-05,
+      "loss": 0.3947,
       "step": 82550
     },
     {
-      "epoch": 2.9,
-      "learning_rate": 1.9720073514501802e-05,
-      "loss": 0.2868,
+      "epoch": 2.97527660647998,
+      "grad_norm": 0.1962546557188034,
+      "learning_rate": 1.860641308576323e-05,
+      "loss": 0.3516,
       "step": 82555
     },
     {
-      "epoch": 2.9,
-      "learning_rate": 1.971728908464611e-05,
-      "loss": 0.2457,
+      "epoch": 2.9754568061412043,
+      "grad_norm": 0.19978949427604675,
+      "learning_rate": 1.8603592039145767e-05,
+      "loss": 0.3896,
       "step": 82560
     },
     {
-      "epoch": 2.9,
-      "learning_rate": 1.9714504723382952e-05,
-      "loss": 0.2821,
+      "epoch": 2.975637005802429,
+      "grad_norm": 0.23347966372966766,
+      "learning_rate": 1.860077107967813e-05,
+      "loss": 0.3658,
       "step": 82565
     },
     {
-      "epoch": 2.91,
-      "learning_rate": 1.9711720430748466e-05,
-      "loss": 0.2635,
+      "epoch": 2.975817205463654,
+      "grad_norm": 0.19242475926876068,
+      "learning_rate": 1.8597950207398745e-05,
+      "loss": 0.4014,
       "step": 82570
     },
     {
-      "epoch": 2.91,
-      "learning_rate": 1.9708936206778803e-05,
-      "loss": 0.2541,
+      "epoch": 2.975997405124878,
+      "grad_norm": 0.20746643841266632,
+      "learning_rate": 1.8595129422346054e-05,
+      "loss": 0.3827,
       "step": 82575
     },
     {
-      "epoch": 2.91,
-      "learning_rate": 1.9706152051510106e-05,
-      "loss": 0.2585,
+      "epoch": 2.976177604786103,
+      "grad_norm": 0.20090548694133759,
+      "learning_rate": 1.8592308724558485e-05,
+      "loss": 0.3629,
       "step": 82580
     },
     {
-      "epoch": 2.91,
-      "learning_rate": 1.970336796497855e-05,
-      "loss": 0.2627,
+      "epoch": 2.9763578044473276,
+      "grad_norm": 0.17902597784996033,
+      "learning_rate": 1.8589488114074456e-05,
+      "loss": 0.3842,
       "step": 82585
     },
     {
-      "epoch": 2.91,
-      "learning_rate": 1.9700583947220268e-05,
-      "loss": 0.2771,
+      "epoch": 2.9765380041085523,
+      "grad_norm": 0.2532462477684021,
+      "learning_rate": 1.8586667590932426e-05,
+      "loss": 0.4095,
       "step": 82590
     },
     {
-      "epoch": 2.91,
-      "learning_rate": 1.969779999827141e-05,
-      "loss": 0.2966,
+      "epoch": 2.976718203769777,
+      "grad_norm": 0.26254743337631226,
+      "learning_rate": 1.8583847155170803e-05,
+      "loss": 0.3585,
       "step": 82595
     },
     {
-      "epoch": 2.91,
-      "learning_rate": 1.9695016118168117e-05,
-      "loss": 0.2555,
+      "epoch": 2.9768984034310018,
+      "grad_norm": 0.21005457639694214,
+      "learning_rate": 1.858102680682802e-05,
+      "loss": 0.3766,
       "step": 82600
     },
     {
-      "epoch": 2.91,
-      "learning_rate": 1.969223230694655e-05,
-      "loss": 0.2502,
+      "epoch": 2.977078603092226,
+      "grad_norm": 0.23283547163009644,
+      "learning_rate": 1.85782065459425e-05,
+      "loss": 0.3884,
       "step": 82605
     },
     {
-      "epoch": 2.91,
-      "learning_rate": 1.9689448564642855e-05,
-      "loss": 0.238,
+      "epoch": 2.977258802753451,
+      "grad_norm": 0.21179614961147308,
+      "learning_rate": 1.8575386372552673e-05,
+      "loss": 0.3647,
       "step": 82610
     },
     {
-      "epoch": 2.91,
-      "learning_rate": 1.9686664891293167e-05,
-      "loss": 0.275,
+      "epoch": 2.9774390024146755,
+      "grad_norm": 0.1830110400915146,
+      "learning_rate": 1.857256628669697e-05,
+      "loss": 0.3991,
       "step": 82615
     },
     {
-      "epoch": 2.91,
-      "learning_rate": 1.9683881286933625e-05,
-      "loss": 0.2512,
+      "epoch": 2.9776192020759003,
+      "grad_norm": 0.24650508165359497,
+      "learning_rate": 1.8569746288413802e-05,
+      "loss": 0.363,
       "step": 82620
     },
     {
-      "epoch": 2.91,
-      "learning_rate": 1.9681097751600387e-05,
-      "loss": 0.2571,
+      "epoch": 2.9777994017371245,
+      "grad_norm": 0.23117853701114655,
+      "learning_rate": 1.856692637774159e-05,
+      "loss": 0.3747,
       "step": 82625
     },
     {
-      "epoch": 2.91,
-      "learning_rate": 1.967831428532959e-05,
-      "loss": 0.2597,
+      "epoch": 2.9779796013983493,
+      "grad_norm": 0.18374405801296234,
+      "learning_rate": 1.8564106554718767e-05,
+      "loss": 0.3854,
       "step": 82630
     },
     {
-      "epoch": 2.91,
-      "learning_rate": 1.9675530888157366e-05,
-      "loss": 0.2538,
+      "epoch": 2.978159801059574,
+      "grad_norm": 0.1933712363243103,
+      "learning_rate": 1.856128681938375e-05,
+      "loss": 0.3719,
       "step": 82635
     },
     {
-      "epoch": 2.91,
-      "learning_rate": 1.9672747560119877e-05,
-      "loss": 0.2731,
+      "epoch": 2.9783400007207987,
+      "grad_norm": 0.24124352633953094,
+      "learning_rate": 1.8558467171774946e-05,
+      "loss": 0.4093,
       "step": 82640
     },
     {
-      "epoch": 2.91,
-      "learning_rate": 1.9669964301253252e-05,
-      "loss": 0.2618,
+      "epoch": 2.9785202003820235,
+      "grad_norm": 0.22895336151123047,
+      "learning_rate": 1.8555647611930788e-05,
+      "loss": 0.3979,
       "step": 82645
     },
     {
-      "epoch": 2.91,
-      "learning_rate": 1.9667181111593618e-05,
-      "loss": 0.2696,
+      "epoch": 2.9787004000432478,
+      "grad_norm": 0.240083709359169,
+      "learning_rate": 1.8552828139889673e-05,
+      "loss": 0.403,
       "step": 82650
     },
     {
-      "epoch": 2.91,
-      "learning_rate": 1.966439799117712e-05,
-      "loss": 0.2296,
+      "epoch": 2.9788805997044725,
+      "grad_norm": 0.18165172636508942,
+      "learning_rate": 1.8550008755690034e-05,
+      "loss": 0.3757,
       "step": 82655
     },
     {
-      "epoch": 2.91,
-      "learning_rate": 1.966161494003991e-05,
-      "loss": 0.263,
+      "epoch": 2.9790607993656972,
+      "grad_norm": 0.2198115587234497,
+      "learning_rate": 1.8547189459370275e-05,
+      "loss": 0.374,
       "step": 82660
     },
     {
-      "epoch": 2.91,
-      "learning_rate": 1.9658831958218107e-05,
-      "loss": 0.2833,
+      "epoch": 2.979240999026922,
+      "grad_norm": 0.19602136313915253,
+      "learning_rate": 1.8544370250968808e-05,
+      "loss": 0.3852,
       "step": 82665
     },
     {
-      "epoch": 2.91,
-      "learning_rate": 1.9656049045747853e-05,
-      "loss": 0.2817,
+      "epoch": 2.9794211986881463,
+      "grad_norm": 0.24065439403057098,
+      "learning_rate": 1.854155113052405e-05,
+      "loss": 0.3971,
       "step": 82670
     },
     {
-      "epoch": 2.91,
-      "learning_rate": 1.965326620266527e-05,
-      "loss": 0.2714,
+      "epoch": 2.979601398349371,
+      "grad_norm": 0.20084550976753235,
+      "learning_rate": 1.853873209807441e-05,
+      "loss": 0.3652,
       "step": 82675
     },
     {
-      "epoch": 2.91,
-      "learning_rate": 1.9650483429006512e-05,
-      "loss": 0.2619,
+      "epoch": 2.9797815980105957,
+      "grad_norm": 0.18546749651432037,
+      "learning_rate": 1.8535913153658284e-05,
+      "loss": 0.3908,
       "step": 82680
     },
     {
-      "epoch": 2.91,
-      "learning_rate": 1.96477007248077e-05,
-      "loss": 0.2632,
+      "epoch": 2.9799617976718205,
+      "grad_norm": 0.19629612565040588,
+      "learning_rate": 1.8533094297314108e-05,
+      "loss": 0.3711,
       "step": 82685
     },
     {
-      "epoch": 2.91,
-      "learning_rate": 1.9644918090104968e-05,
-      "loss": 0.2504,
+      "epoch": 2.980141997333045,
+      "grad_norm": 0.19778327643871307,
+      "learning_rate": 1.8530275529080248e-05,
+      "loss": 0.4318,
       "step": 82690
     },
     {
-      "epoch": 2.91,
-      "learning_rate": 1.9642135524934434e-05,
-      "loss": 0.2637,
+      "epoch": 2.98032219699427,
+      "grad_norm": 0.19688297808170319,
+      "learning_rate": 1.8527456848995148e-05,
+      "loss": 0.3831,
       "step": 82695
     },
     {
-      "epoch": 2.91,
-      "learning_rate": 1.9639353029332255e-05,
-      "loss": 0.2801,
+      "epoch": 2.980502396655494,
+      "grad_norm": 0.22742833197116852,
+      "learning_rate": 1.852463825709719e-05,
+      "loss": 0.4061,
       "step": 82700
     },
     {
-      "epoch": 2.91,
-      "learning_rate": 1.9636570603334537e-05,
-      "loss": 0.2835,
+      "epoch": 2.980682596316719,
+      "grad_norm": 0.20782898366451263,
+      "learning_rate": 1.852181975342478e-05,
+      "loss": 0.3998,
       "step": 82705
     },
     {
-      "epoch": 2.91,
-      "learning_rate": 1.9633788246977415e-05,
-      "loss": 0.2811,
+      "epoch": 2.9808627959779437,
+      "grad_norm": 0.20285415649414062,
+      "learning_rate": 1.8519001338016324e-05,
+      "loss": 0.3738,
       "step": 82710
     },
     {
-      "epoch": 2.91,
-      "learning_rate": 1.963100596029701e-05,
-      "loss": 0.2644,
+      "epoch": 2.981042995639168,
+      "grad_norm": 0.26227110624313354,
+      "learning_rate": 1.8516183010910216e-05,
+      "loss": 0.3849,
       "step": 82715
     },
     {
-      "epoch": 2.91,
-      "learning_rate": 1.9628223743329466e-05,
-      "loss": 0.2563,
+      "epoch": 2.9812231953003927,
+      "grad_norm": 0.20241287350654602,
+      "learning_rate": 1.8513364772144863e-05,
+      "loss": 0.3869,
       "step": 82720
     },
     {
-      "epoch": 2.91,
-      "learning_rate": 1.9625441596110894e-05,
-      "loss": 0.2486,
+      "epoch": 2.9814033949616174,
+      "grad_norm": 0.25886163115501404,
+      "learning_rate": 1.851054662175866e-05,
+      "loss": 0.43,
       "step": 82725
     },
     {
-      "epoch": 2.91,
-      "learning_rate": 1.9622659518677416e-05,
-      "loss": 0.2612,
+      "epoch": 2.981583594622842,
+      "grad_norm": 0.20351019501686096,
+      "learning_rate": 1.8507728559789996e-05,
+      "loss": 0.3872,
       "step": 82730
     },
     {
-      "epoch": 2.91,
-      "learning_rate": 1.9619877511065157e-05,
-      "loss": 0.2785,
+      "epoch": 2.981763794284067,
+      "grad_norm": 0.27192798256874084,
+      "learning_rate": 1.850491058627728e-05,
+      "loss": 0.3813,
       "step": 82735
     },
     {
-      "epoch": 2.91,
-      "learning_rate": 1.961709557331025e-05,
-      "loss": 0.2814,
+      "epoch": 2.9819439939452916,
+      "grad_norm": 0.19435738027095795,
+      "learning_rate": 1.85020927012589e-05,
+      "loss": 0.3965,
       "step": 82740
     },
     {
-      "epoch": 2.91,
-      "learning_rate": 1.9614313705448814e-05,
-      "loss": 0.2843,
+      "epoch": 2.982124193606516,
+      "grad_norm": 0.20769937336444855,
+      "learning_rate": 1.8499274904773245e-05,
+      "loss": 0.3737,
       "step": 82745
     },
     {
-      "epoch": 2.91,
-      "learning_rate": 1.961153190751695e-05,
-      "loss": 0.2801,
+      "epoch": 2.9823043932677407,
+      "grad_norm": 0.22676123678684235,
+      "learning_rate": 1.8496457196858714e-05,
+      "loss": 0.3807,
       "step": 82750
     },
     {
-      "epoch": 2.91,
-      "learning_rate": 1.9608750179550806e-05,
-      "loss": 0.2575,
+      "epoch": 2.9824845929289654,
+      "grad_norm": 0.18023423850536346,
+      "learning_rate": 1.849363957755369e-05,
+      "loss": 0.4336,
       "step": 82755
     },
     {
-      "epoch": 2.91,
-      "learning_rate": 1.960596852158648e-05,
-      "loss": 0.272,
+      "epoch": 2.9826647925901897,
+      "grad_norm": 0.2146742194890976,
+      "learning_rate": 1.849082204689657e-05,
+      "loss": 0.3597,
       "step": 82760
     },
     {
-      "epoch": 2.91,
-      "learning_rate": 1.960318693366011e-05,
-      "loss": 0.2707,
+      "epoch": 2.9828449922514144,
+      "grad_norm": 0.24750038981437683,
+      "learning_rate": 1.848800460492575e-05,
+      "loss": 0.3814,
       "step": 82765
     },
     {
-      "epoch": 2.91,
-      "learning_rate": 1.9600405415807786e-05,
-      "loss": 0.2527,
+      "epoch": 2.983025191912639,
+      "grad_norm": 0.20825496315956116,
+      "learning_rate": 1.8485187251679586e-05,
+      "loss": 0.4108,
       "step": 82770
     },
     {
-      "epoch": 2.91,
-      "learning_rate": 1.9597623968065647e-05,
-      "loss": 0.2772,
+      "epoch": 2.983205391573864,
+      "grad_norm": 0.23861125111579895,
+      "learning_rate": 1.8482369987196503e-05,
+      "loss": 0.4002,
       "step": 82775
     },
     {
-      "epoch": 2.91,
-      "learning_rate": 1.9594842590469807e-05,
-      "loss": 0.2358,
+      "epoch": 2.9833855912350886,
+      "grad_norm": 0.2297501564025879,
+      "learning_rate": 1.847955281151486e-05,
+      "loss": 0.4099,
       "step": 82780
     },
     {
-      "epoch": 2.91,
-      "learning_rate": 1.9592061283056363e-05,
-      "loss": 0.2604,
+      "epoch": 2.9835657908963134,
+      "grad_norm": 0.2385878562927246,
+      "learning_rate": 1.847673572467304e-05,
+      "loss": 0.4512,
       "step": 82785
     },
     {
-      "epoch": 2.91,
-      "learning_rate": 1.958928004586144e-05,
-      "loss": 0.2802,
+      "epoch": 2.9837459905575376,
+      "grad_norm": 0.21177738904953003,
+      "learning_rate": 1.8473918726709442e-05,
+      "loss": 0.394,
       "step": 82790
     },
     {
-      "epoch": 2.91,
-      "learning_rate": 1.9586498878921153e-05,
-      "loss": 0.2797,
+      "epoch": 2.9839261902187624,
+      "grad_norm": 0.17175810039043427,
+      "learning_rate": 1.8471101817662428e-05,
+      "loss": 0.4103,
       "step": 82795
     },
     {
-      "epoch": 2.91,
-      "learning_rate": 1.9583717782271614e-05,
-      "loss": 0.2408,
+      "epoch": 2.984106389879987,
+      "grad_norm": 0.22006794810295105,
+      "learning_rate": 1.8468284997570404e-05,
+      "loss": 0.3797,
       "step": 82800
     },
     {
-      "epoch": 2.91,
-      "learning_rate": 1.9580936755948932e-05,
-      "loss": 0.2661,
+      "epoch": 2.9842865895412114,
+      "grad_norm": 0.20512345433235168,
+      "learning_rate": 1.8465468266471724e-05,
+      "loss": 0.4287,
       "step": 82805
     },
     {
-      "epoch": 2.91,
-      "learning_rate": 1.95781557999892e-05,
-      "loss": 0.2735,
+      "epoch": 2.984466789202436,
+      "grad_norm": 0.24116021394729614,
+      "learning_rate": 1.846265162440477e-05,
+      "loss": 0.3939,
       "step": 82810
     },
     {
-      "epoch": 2.91,
-      "learning_rate": 1.9575374914428548e-05,
-      "loss": 0.2756,
+      "epoch": 2.984646988863661,
+      "grad_norm": 0.21826866269111633,
+      "learning_rate": 1.8459835071407928e-05,
+      "loss": 0.3739,
       "step": 82815
     },
     {
-      "epoch": 2.91,
-      "learning_rate": 1.9572594099303086e-05,
-      "loss": 0.2808,
+      "epoch": 2.9848271885248856,
+      "grad_norm": 0.20197410881519318,
+      "learning_rate": 1.8457018607519567e-05,
+      "loss": 0.4008,
       "step": 82820
     },
     {
-      "epoch": 2.91,
-      "learning_rate": 1.956981335464891e-05,
-      "loss": 0.2507,
+      "epoch": 2.9850073881861103,
+      "grad_norm": 0.21542875468730927,
+      "learning_rate": 1.845420223277805e-05,
+      "loss": 0.3851,
       "step": 82825
     },
     {
-      "epoch": 2.91,
-      "learning_rate": 1.956703268050212e-05,
-      "loss": 0.2484,
+      "epoch": 2.985187587847335,
+      "grad_norm": 0.19995024800300598,
+      "learning_rate": 1.8451385947221773e-05,
+      "loss": 0.3678,
       "step": 82830
     },
     {
-      "epoch": 2.91,
-      "learning_rate": 1.9564252076898842e-05,
-      "loss": 0.2662,
+      "epoch": 2.9853677875085594,
+      "grad_norm": 0.20630434155464172,
+      "learning_rate": 1.8448569750889083e-05,
+      "loss": 0.3832,
       "step": 82835
     },
     {
-      "epoch": 2.91,
-      "learning_rate": 1.956147154387516e-05,
-      "loss": 0.2683,
+      "epoch": 2.985547987169784,
+      "grad_norm": 0.20057667791843414,
+      "learning_rate": 1.844575364381837e-05,
+      "loss": 0.394,
       "step": 82840
     },
     {
-      "epoch": 2.91,
-      "learning_rate": 1.9558691081467193e-05,
-      "loss": 0.2644,
+      "epoch": 2.985728186831009,
+      "grad_norm": 0.2470291405916214,
+      "learning_rate": 1.8442937626047995e-05,
+      "loss": 0.3905,
       "step": 82845
     },
     {
-      "epoch": 2.91,
-      "learning_rate": 1.9555910689711025e-05,
-      "loss": 0.256,
+      "epoch": 2.9859083864922336,
+      "grad_norm": 0.20389066636562347,
+      "learning_rate": 1.844012169761632e-05,
+      "loss": 0.4083,
       "step": 82850
     },
     {
-      "epoch": 2.92,
-      "learning_rate": 1.955313036864278e-05,
-      "loss": 0.2545,
+      "epoch": 2.986088586153458,
+      "grad_norm": 0.26393190026283264,
+      "learning_rate": 1.843730585856172e-05,
+      "loss": 0.4205,
       "step": 82855
     },
     {
-      "epoch": 2.92,
-      "learning_rate": 1.955035011829855e-05,
-      "loss": 0.2803,
+      "epoch": 2.9862687858146826,
+      "grad_norm": 0.2350768744945526,
+      "learning_rate": 1.8434490108922566e-05,
+      "loss": 0.3914,
       "step": 82860
     },
     {
-      "epoch": 2.92,
-      "learning_rate": 1.9547569938714415e-05,
-      "loss": 0.2853,
+      "epoch": 2.9864489854759073,
+      "grad_norm": 0.20153537392616272,
+      "learning_rate": 1.8431674448737195e-05,
+      "loss": 0.3954,
       "step": 82865
     },
     {
-      "epoch": 2.92,
-      "learning_rate": 1.9544789829926507e-05,
-      "loss": 0.2946,
+      "epoch": 2.986629185137132,
+      "grad_norm": 0.22236128151416779,
+      "learning_rate": 1.8428858878044007e-05,
+      "loss": 0.4122,
       "step": 82870
     },
     {
-      "epoch": 2.92,
-      "learning_rate": 1.95420097919709e-05,
-      "loss": 0.2517,
+      "epoch": 2.986809384798357,
+      "grad_norm": 0.18074600398540497,
+      "learning_rate": 1.8426043396881326e-05,
+      "loss": 0.3938,
       "step": 82875
     },
     {
-      "epoch": 2.92,
-      "learning_rate": 1.953922982488371e-05,
-      "loss": 0.2771,
+      "epoch": 2.986989584459581,
+      "grad_norm": 0.21660077571868896,
+      "learning_rate": 1.8423228005287546e-05,
+      "loss": 0.4038,
       "step": 82880
     },
     {
-      "epoch": 2.92,
-      "learning_rate": 1.9536449928701007e-05,
-      "loss": 0.2789,
+      "epoch": 2.987169784120806,
+      "grad_norm": 0.23685570061206818,
+      "learning_rate": 1.8420412703301004e-05,
+      "loss": 0.3751,
       "step": 82885
     },
     {
-      "epoch": 2.92,
-      "learning_rate": 1.9533670103458912e-05,
-      "loss": 0.2807,
+      "epoch": 2.9873499837820305,
+      "grad_norm": 0.20489926636219025,
+      "learning_rate": 1.8417597490960066e-05,
+      "loss": 0.4195,
       "step": 82890
     },
     {
-      "epoch": 2.92,
-      "learning_rate": 1.9530890349193502e-05,
-      "loss": 0.2674,
+      "epoch": 2.9875301834432553,
+      "grad_norm": 0.23835612833499908,
+      "learning_rate": 1.8414782368303088e-05,
+      "loss": 0.3734,
       "step": 82895
     },
     {
-      "epoch": 2.92,
-      "learning_rate": 1.9528110665940882e-05,
-      "loss": 0.2566,
+      "epoch": 2.9877103831044796,
+      "grad_norm": 0.24969185888767242,
+      "learning_rate": 1.841196733536842e-05,
+      "loss": 0.3929,
       "step": 82900
     },
     {
-      "epoch": 2.92,
-      "learning_rate": 1.9525331053737128e-05,
-      "loss": 0.2781,
+      "epoch": 2.9878905827657043,
+      "grad_norm": 0.22482214868068695,
+      "learning_rate": 1.8409152392194427e-05,
+      "loss": 0.4001,
       "step": 82905
     },
     {
-      "epoch": 2.92,
-      "learning_rate": 1.9522551512618354e-05,
-      "loss": 0.2592,
+      "epoch": 2.988070782426929,
+      "grad_norm": 0.19898438453674316,
+      "learning_rate": 1.8406337538819462e-05,
+      "loss": 0.3767,
       "step": 82910
     },
     {
-      "epoch": 2.92,
-      "learning_rate": 1.9519772042620637e-05,
-      "loss": 0.2896,
+      "epoch": 2.9882509820881538,
+      "grad_norm": 0.2599036991596222,
+      "learning_rate": 1.840352277528186e-05,
+      "loss": 0.4169,
       "step": 82915
     },
     {
-      "epoch": 2.92,
-      "learning_rate": 1.9516992643780068e-05,
-      "loss": 0.2845,
+      "epoch": 2.9884311817493785,
+      "grad_norm": 0.1831098049879074,
+      "learning_rate": 1.840070810161999e-05,
+      "loss": 0.3763,
       "step": 82920
     },
     {
-      "epoch": 2.92,
-      "learning_rate": 1.9514213316132732e-05,
-      "loss": 0.256,
+      "epoch": 2.988611381410603,
+      "grad_norm": 0.26548436284065247,
+      "learning_rate": 1.8397893517872196e-05,
+      "loss": 0.3815,
       "step": 82925
     },
     {
-      "epoch": 2.92,
-      "learning_rate": 1.951143405971472e-05,
-      "loss": 0.2472,
+      "epoch": 2.9887915810718275,
+      "grad_norm": 0.24817430973052979,
+      "learning_rate": 1.839507902407682e-05,
+      "loss": 0.4,
       "step": 82930
     },
     {
-      "epoch": 2.92,
-      "learning_rate": 1.9508654874562127e-05,
-      "loss": 0.2716,
+      "epoch": 2.9889717807330523,
+      "grad_norm": NaN,
+      "learning_rate": 1.839282749383203e-05,
+      "loss": 0.4199,
       "step": 82935
     },
     {
-      "epoch": 2.92,
-      "learning_rate": 1.950587576071103e-05,
-      "loss": 0.2728,
+      "epoch": 2.989151980394277,
+      "grad_norm": 0.2185213267803192,
+      "learning_rate": 1.839001316204766e-05,
+      "loss": 0.3663,
       "step": 82940
     },
     {
-      "epoch": 2.92,
-      "learning_rate": 1.9503096718197507e-05,
-      "loss": 0.279,
+      "epoch": 2.9893321800555013,
+      "grad_norm": 0.24561047554016113,
+      "learning_rate": 1.8387198920323063e-05,
+      "loss": 0.4042,
       "step": 82945
     },
     {
-      "epoch": 2.92,
-      "learning_rate": 1.9500317747057662e-05,
-      "loss": 0.2568,
+      "epoch": 2.989512379716726,
+      "grad_norm": 0.1906471997499466,
+      "learning_rate": 1.8384384768696615e-05,
+      "loss": 0.3998,
       "step": 82950
     },
     {
-      "epoch": 2.92,
-      "learning_rate": 1.9497538847327557e-05,
-      "loss": 0.2763,
+      "epoch": 2.9896925793779507,
+      "grad_norm": 0.2308565229177475,
+      "learning_rate": 1.8381570707206637e-05,
+      "loss": 0.3625,
       "step": 82955
     },
     {
-      "epoch": 2.92,
-      "learning_rate": 1.9494760019043292e-05,
-      "loss": 0.2828,
+      "epoch": 2.9898727790391755,
+      "grad_norm": 0.2136189490556717,
+      "learning_rate": 1.837875673589147e-05,
+      "loss": 0.3868,
       "step": 82960
     },
     {
-      "epoch": 2.92,
-      "learning_rate": 1.949198126224093e-05,
-      "loss": 0.3088,
+      "epoch": 2.9900529787004,
+      "grad_norm": 0.21687231957912445,
+      "learning_rate": 1.8375942854789475e-05,
+      "loss": 0.3923,
       "step": 82965
     },
     {
-      "epoch": 2.92,
-      "learning_rate": 1.948920257695657e-05,
-      "loss": 0.253,
+      "epoch": 2.990233178361625,
+      "grad_norm": 0.24081559479236603,
+      "learning_rate": 1.837312906393896e-05,
+      "loss": 0.3877,
       "step": 82970
     },
     {
-      "epoch": 2.92,
-      "learning_rate": 1.9486423963226284e-05,
-      "loss": 0.2667,
+      "epoch": 2.9904133780228492,
+      "grad_norm": 0.19234254956245422,
+      "learning_rate": 1.8370315363378293e-05,
+      "loss": 0.4074,
       "step": 82975
     },
     {
-      "epoch": 2.92,
-      "learning_rate": 1.9483645421086137e-05,
-      "loss": 0.2419,
+      "epoch": 2.990593577684074,
+      "grad_norm": 0.22776223719120026,
+      "learning_rate": 1.8367501753145792e-05,
+      "loss": 0.4108,
       "step": 82980
     },
     {
-      "epoch": 2.92,
-      "learning_rate": 1.9480866950572227e-05,
-      "loss": 0.2721,
+      "epoch": 2.9907737773452987,
+      "grad_norm": 0.23385420441627502,
+      "learning_rate": 1.836468823327979e-05,
+      "loss": 0.3966,
       "step": 82985
     },
     {
-      "epoch": 2.92,
-      "learning_rate": 1.9478088551720626e-05,
-      "loss": 0.2733,
+      "epoch": 2.990953977006523,
+      "grad_norm": 0.25053873658180237,
+      "learning_rate": 1.836187480381863e-05,
+      "loss": 0.4033,
       "step": 82990
     },
     {
-      "epoch": 2.92,
-      "learning_rate": 1.9475310224567405e-05,
-      "loss": 0.2441,
+      "epoch": 2.9911341766677477,
+      "grad_norm": 0.22640636563301086,
+      "learning_rate": 1.835906146480064e-05,
+      "loss": 0.3687,
       "step": 82995
     },
     {
-      "epoch": 2.92,
-      "learning_rate": 1.9472531969148632e-05,
-      "loss": 0.2591,
+      "epoch": 2.9913143763289725,
+      "grad_norm": 0.2172614485025406,
+      "learning_rate": 1.835624821626415e-05,
+      "loss": 0.3921,
       "step": 83000
     },
     {
-      "epoch": 2.92,
-      "eval_loss": 0.25857260823249817,
-      "eval_runtime": 10.5923,
-      "eval_samples_per_second": 9.441,
-      "eval_steps_per_second": 9.441,
+      "epoch": 2.9913143763289725,
+      "eval_loss": 0.4294629693031311,
+      "eval_runtime": 3.5292,
+      "eval_samples_per_second": 28.335,
+      "eval_steps_per_second": 7.084,
       "step": 83000
     },
     {
-      "epoch": 2.92,
-      "learning_rate": 1.9469753785500397e-05,
-      "loss": 0.2793,
+      "epoch": 2.991494575990197,
+      "grad_norm": 0.17745208740234375,
+      "learning_rate": 1.8353435058247496e-05,
+      "loss": 0.3833,
       "step": 83005
     },
     {
-      "epoch": 2.92,
-      "learning_rate": 1.946697567365876e-05,
-      "loss": 0.2883,
+      "epoch": 2.991674775651422,
+      "grad_norm": 0.25056445598602295,
+      "learning_rate": 1.8350621990788998e-05,
+      "loss": 0.4054,
       "step": 83010
     },
     {
-      "epoch": 2.92,
-      "learning_rate": 1.9464197633659802e-05,
-      "loss": 0.2625,
+      "epoch": 2.9918549753126467,
+      "grad_norm": 0.20053359866142273,
+      "learning_rate": 1.834780901392699e-05,
+      "loss": 0.4106,
       "step": 83015
     },
     {
-      "epoch": 2.92,
-      "learning_rate": 1.9461419665539582e-05,
-      "loss": 0.2913,
+      "epoch": 2.992035174973871,
+      "grad_norm": 0.18675009906291962,
+      "learning_rate": 1.83449961276998e-05,
+      "loss": 0.3792,
       "step": 83020
     },
     {
-      "epoch": 2.92,
-      "learning_rate": 1.9458641769334184e-05,
-      "loss": 0.2658,
+      "epoch": 2.9922153746350957,
+      "grad_norm": 0.22878988087177277,
+      "learning_rate": 1.834218333214574e-05,
+      "loss": 0.4279,
       "step": 83025
     },
     {
-      "epoch": 2.92,
-      "learning_rate": 1.945586394507967e-05,
-      "loss": 0.2692,
+      "epoch": 2.9923955742963204,
+      "grad_norm": 0.20670390129089355,
+      "learning_rate": 1.833937062730315e-05,
+      "loss": 0.3876,
       "step": 83030
     },
     {
-      "epoch": 2.92,
-      "learning_rate": 1.945308619281211e-05,
-      "loss": 0.2514,
+      "epoch": 2.9925757739575447,
+      "grad_norm": 0.18799588084220886,
+      "learning_rate": 1.8336558013210355e-05,
+      "loss": 0.3937,
       "step": 83035
     },
     {
-      "epoch": 2.92,
-      "learning_rate": 1.9450308512567562e-05,
-      "loss": 0.2617,
+      "epoch": 2.9927559736187694,
+      "grad_norm": 0.19332025945186615,
+      "learning_rate": 1.8333745489905646e-05,
+      "loss": 0.4095,
       "step": 83040
     },
     {
-      "epoch": 2.92,
-      "learning_rate": 1.9447530904382108e-05,
-      "loss": 0.2585,
+      "epoch": 2.992936173279994,
+      "grad_norm": 0.22102457284927368,
+      "learning_rate": 1.8330933057427378e-05,
+      "loss": 0.3956,
       "step": 83045
     },
     {
-      "epoch": 2.92,
-      "learning_rate": 1.9444753368291808e-05,
-      "loss": 0.2485,
+      "epoch": 2.993116372941219,
+      "grad_norm": 0.2692544758319855,
+      "learning_rate": 1.8328120715813845e-05,
+      "loss": 0.403,
       "step": 83050
     },
     {
-      "epoch": 2.92,
-      "learning_rate": 1.9441975904332727e-05,
-      "loss": 0.2593,
+      "epoch": 2.9932965726024436,
+      "grad_norm": 0.20256653428077698,
+      "learning_rate": 1.832530846510339e-05,
+      "loss": 0.3552,
       "step": 83055
     },
     {
-      "epoch": 2.92,
-      "learning_rate": 1.9439198512540913e-05,
-      "loss": 0.2562,
+      "epoch": 2.9934767722636684,
+      "grad_norm": 0.2464403659105301,
+      "learning_rate": 1.8322496305334312e-05,
+      "loss": 0.3758,
       "step": 83060
     },
     {
-      "epoch": 2.92,
-      "learning_rate": 1.9436421192952445e-05,
-      "loss": 0.2747,
+      "epoch": 2.9936569719248927,
+      "grad_norm": 0.20932310819625854,
+      "learning_rate": 1.831968423654492e-05,
+      "loss": 0.3809,
       "step": 83065
     },
     {
-      "epoch": 2.92,
-      "learning_rate": 1.943364394560339e-05,
-      "loss": 0.2603,
+      "epoch": 2.9938371715861174,
+      "grad_norm": 0.19015516340732574,
+      "learning_rate": 1.8316872258773543e-05,
+      "loss": 0.4073,
       "step": 83070
     },
     {
-      "epoch": 2.92,
-      "learning_rate": 1.94308667705298e-05,
-      "loss": 0.2622,
+      "epoch": 2.994017371247342,
+      "grad_norm": 0.2418580800294876,
+      "learning_rate": 1.831406037205849e-05,
+      "loss": 0.432,
       "step": 83075
     },
     {
-      "epoch": 2.92,
-      "learning_rate": 1.9428089667767722e-05,
-      "loss": 0.2696,
+      "epoch": 2.9941975709085664,
+      "grad_norm": 0.23939929902553558,
+      "learning_rate": 1.831124857643806e-05,
+      "loss": 0.4024,
       "step": 83080
     },
     {
-      "epoch": 2.92,
-      "learning_rate": 1.942531263735324e-05,
-      "loss": 0.2567,
+      "epoch": 2.994377770569791,
+      "grad_norm": 0.259090781211853,
+      "learning_rate": 1.8308436871950584e-05,
+      "loss": 0.3824,
       "step": 83085
     },
     {
-      "epoch": 2.92,
-      "learning_rate": 1.942253567932239e-05,
-      "loss": 0.2608,
+      "epoch": 2.994557970231016,
+      "grad_norm": 0.2107851803302765,
+      "learning_rate": 1.8305625258634353e-05,
+      "loss": 0.4087,
       "step": 83090
     },
     {
-      "epoch": 2.92,
-      "learning_rate": 1.941975879371124e-05,
-      "loss": 0.2979,
+      "epoch": 2.9947381698922406,
+      "grad_norm": 0.22065068781375885,
+      "learning_rate": 1.8302813736527686e-05,
+      "loss": 0.4059,
       "step": 83095
     },
     {
-      "epoch": 2.92,
-      "learning_rate": 1.941698198055585e-05,
-      "loss": 0.2658,
+      "epoch": 2.9949183695534654,
+      "grad_norm": 0.21330714225769043,
+      "learning_rate": 1.8300002305668884e-05,
+      "loss": 0.3922,
       "step": 83100
     },
     {
-      "epoch": 2.92,
-      "learning_rate": 1.941420523989227e-05,
-      "loss": 0.2573,
+      "epoch": 2.99509856921469,
+      "grad_norm": 0.19063615798950195,
+      "learning_rate": 1.829719096609625e-05,
+      "loss": 0.3862,
       "step": 83105
     },
     {
-      "epoch": 2.92,
-      "learning_rate": 1.941142857175656e-05,
-      "loss": 0.2538,
+      "epoch": 2.9952787688759144,
+      "grad_norm": 0.2311249077320099,
+      "learning_rate": 1.8294379717848095e-05,
+      "loss": 0.3986,
       "step": 83110
     },
     {
-      "epoch": 2.92,
-      "learning_rate": 1.940865197618475e-05,
-      "loss": 0.2743,
+      "epoch": 2.995458968537139,
+      "grad_norm": 0.18495197594165802,
+      "learning_rate": 1.8291568560962723e-05,
+      "loss": 0.4199,
       "step": 83115
     },
     {
-      "epoch": 2.92,
-      "learning_rate": 1.940587545321291e-05,
-      "loss": 0.2432,
+      "epoch": 2.995639168198364,
+      "grad_norm": 0.1809265911579132,
+      "learning_rate": 1.828875749547842e-05,
+      "loss": 0.4077,
       "step": 83120
     },
     {
-      "epoch": 2.92,
-      "learning_rate": 1.9403099002877108e-05,
-      "loss": 0.2777,
+      "epoch": 2.9958193678595886,
+      "grad_norm": 0.2279941588640213,
+      "learning_rate": 1.82859465214335e-05,
+      "loss": 0.4244,
       "step": 83125
     },
     {
-      "epoch": 2.92,
-      "learning_rate": 1.940032262521337e-05,
-      "loss": 0.2655,
+      "epoch": 2.995999567520813,
+      "grad_norm": 0.23656921088695526,
+      "learning_rate": 1.8283135638866263e-05,
+      "loss": 0.3996,
       "step": 83130
     },
     {
-      "epoch": 2.92,
-      "learning_rate": 1.939754632025774e-05,
-      "loss": 0.2827,
+      "epoch": 2.9961797671820376,
+      "grad_norm": 0.2122529149055481,
+      "learning_rate": 1.8280324847815e-05,
+      "loss": 0.4041,
       "step": 83135
     },
     {
-      "epoch": 2.93,
-      "learning_rate": 1.9394770088046294e-05,
-      "loss": 0.265,
+      "epoch": 2.9963599668432623,
+      "grad_norm": 0.21249407529830933,
+      "learning_rate": 1.8277514148318014e-05,
+      "loss": 0.3736,
       "step": 83140
     },
     {
-      "epoch": 2.93,
-      "learning_rate": 1.939199392861506e-05,
-      "loss": 0.2621,
+      "epoch": 2.996540166504487,
+      "grad_norm": 0.2389555722475052,
+      "learning_rate": 1.8274703540413584e-05,
+      "loss": 0.3884,
       "step": 83145
     },
     {
-      "epoch": 2.93,
-      "learning_rate": 1.9389217842000088e-05,
-      "loss": 0.2781,
+      "epoch": 2.996720366165712,
+      "grad_norm": 0.24687263369560242,
+      "learning_rate": 1.8271893024140034e-05,
+      "loss": 0.4034,
       "step": 83150
     },
     {
-      "epoch": 2.93,
-      "learning_rate": 1.938644182823742e-05,
-      "loss": 0.2644,
+      "epoch": 2.996900565826936,
+      "grad_norm": 0.19258186221122742,
+      "learning_rate": 1.826908259953562e-05,
+      "loss": 0.3823,
       "step": 83155
     },
     {
-      "epoch": 2.93,
-      "learning_rate": 1.9383665887363113e-05,
-      "loss": 0.2719,
+      "epoch": 2.997080765488161,
+      "grad_norm": 0.21487314999103546,
+      "learning_rate": 1.826627226663867e-05,
+      "loss": 0.4198,
       "step": 83160
     },
     {
-      "epoch": 2.93,
-      "learning_rate": 1.93808900194132e-05,
-      "loss": 0.2746,
+      "epoch": 2.9972609651493856,
+      "grad_norm": 0.17476046085357666,
+      "learning_rate": 1.826346202548745e-05,
+      "loss": 0.3885,
       "step": 83165
     },
     {
-      "epoch": 2.93,
-      "learning_rate": 1.9378114224423726e-05,
-      "loss": 0.2669,
+      "epoch": 2.9974411648106103,
+      "grad_norm": 0.22289486229419708,
+      "learning_rate": 1.826065187612025e-05,
+      "loss": 0.3922,
       "step": 83170
     },
     {
-      "epoch": 2.93,
-      "learning_rate": 1.937533850243073e-05,
-      "loss": 0.2738,
+      "epoch": 2.9976213644718346,
+      "grad_norm": 0.21170969307422638,
+      "learning_rate": 1.825784181857537e-05,
+      "loss": 0.3817,
       "step": 83175
     },
     {
-      "epoch": 2.93,
-      "learning_rate": 1.9372562853470263e-05,
-      "loss": 0.2719,
+      "epoch": 2.9978015641330593,
+      "grad_norm": 0.17165972292423248,
+      "learning_rate": 1.825503185289109e-05,
+      "loss": 0.3939,
       "step": 83180
     },
     {
-      "epoch": 2.93,
-      "learning_rate": 1.9369787277578365e-05,
-      "loss": 0.2708,
+      "epoch": 2.997981763794284,
+      "grad_norm": 0.2255578637123108,
+      "learning_rate": 1.8252221979105687e-05,
+      "loss": 0.3985,
       "step": 83185
     },
     {
-      "epoch": 2.93,
-      "learning_rate": 1.9367011774791062e-05,
-      "loss": 0.2549,
+      "epoch": 2.998161963455509,
+      "grad_norm": 0.23678001761436462,
+      "learning_rate": 1.824941219725746e-05,
+      "loss": 0.4077,
       "step": 83190
     },
     {
-      "epoch": 2.93,
-      "learning_rate": 1.9364236345144393e-05,
-      "loss": 0.2744,
+      "epoch": 2.9983421631167335,
+      "grad_norm": 0.32480308413505554,
+      "learning_rate": 1.824660250738468e-05,
+      "loss": 0.4023,
       "step": 83195
     },
     {
-      "epoch": 2.93,
-      "learning_rate": 1.93614609886744e-05,
-      "loss": 0.3034,
+      "epoch": 2.9985223627779583,
+      "grad_norm": 0.18831786513328552,
+      "learning_rate": 1.8243792909525643e-05,
+      "loss": 0.3703,
       "step": 83200
     },
     {
-      "epoch": 2.93,
-      "learning_rate": 1.9358685705417133e-05,
-      "loss": 0.2712,
+      "epoch": 2.9987025624391825,
+      "grad_norm": 0.22749020159244537,
+      "learning_rate": 1.8240983403718614e-05,
+      "loss": 0.3943,
       "step": 83205
     },
     {
-      "epoch": 2.93,
-      "learning_rate": 1.935591049540861e-05,
-      "loss": 0.2648,
+      "epoch": 2.9988827621004073,
+      "grad_norm": 0.2319597750902176,
+      "learning_rate": 1.8238173990001874e-05,
+      "loss": 0.3893,
       "step": 83210
     },
     {
-      "epoch": 2.93,
-      "learning_rate": 1.935313535868486e-05,
-      "loss": 0.2597,
+      "epoch": 2.999062961761632,
+      "grad_norm": 0.2220054417848587,
+      "learning_rate": 1.8235364668413705e-05,
+      "loss": 0.3845,
       "step": 83215
     },
     {
-      "epoch": 2.93,
-      "learning_rate": 1.9350360295281937e-05,
-      "loss": 0.2538,
+      "epoch": 2.9992431614228563,
+      "grad_norm": 0.29346784949302673,
+      "learning_rate": 1.8232555438992395e-05,
+      "loss": 0.4334,
       "step": 83220
     },
     {
-      "epoch": 2.93,
-      "learning_rate": 1.9347585305235858e-05,
-      "loss": 0.2593,
+      "epoch": 2.999423361084081,
+      "grad_norm": 0.19180335104465485,
+      "learning_rate": 1.8229746301776187e-05,
+      "loss": 0.3889,
       "step": 83225
     },
     {
-      "epoch": 2.93,
-      "learning_rate": 1.934481038858266e-05,
-      "loss": 0.2686,
+      "epoch": 2.9996035607453058,
+      "grad_norm": 0.2530876100063324,
+      "learning_rate": 1.8226937256803396e-05,
+      "loss": 0.3928,
       "step": 83230
     },
     {
-      "epoch": 2.93,
-      "learning_rate": 1.9342035545358366e-05,
-      "loss": 0.2657,
+      "epoch": 2.9997837604065305,
+      "grad_norm": 0.1911705732345581,
+      "learning_rate": 1.8224128304112252e-05,
+      "loss": 0.4052,
       "step": 83235
     },
     {
-      "epoch": 2.93,
-      "learning_rate": 1.933926077559902e-05,
-      "loss": 0.2594,
+      "epoch": 2.9999639600677552,
+      "grad_norm": 0.21257303655147552,
+      "learning_rate": 1.822131944374106e-05,
+      "loss": 0.3966,
       "step": 83240
     },
     {
-      "epoch": 2.93,
-      "learning_rate": 1.9336486079340643e-05,
-      "loss": 0.2823,
+      "epoch": 3.0001441597289795,
+      "grad_norm": 0.24478384852409363,
+      "learning_rate": 1.8218510675728085e-05,
+      "loss": 0.3972,
       "step": 83245
     },
     {
-      "epoch": 2.93,
-      "learning_rate": 1.933371145661925e-05,
-      "loss": 0.2676,
+      "epoch": 3.0003243593902043,
+      "grad_norm": 0.21207980811595917,
+      "learning_rate": 1.821570200011157e-05,
+      "loss": 0.3956,
       "step": 83250
     },
     {
-      "epoch": 2.93,
-      "learning_rate": 1.9330936907470886e-05,
-      "loss": 0.2641,
+      "epoch": 3.000504559051429,
+      "grad_norm": 0.19810383021831512,
+      "learning_rate": 1.8212893416929817e-05,
+      "loss": 0.3571,
       "step": 83255
     },
     {
-      "epoch": 2.93,
-      "learning_rate": 1.9328162431931575e-05,
-      "loss": 0.2608,
+      "epoch": 3.0006847587126537,
+      "grad_norm": 0.21242433786392212,
+      "learning_rate": 1.8210084926221078e-05,
+      "loss": 0.3709,
       "step": 83260
     },
     {
-      "epoch": 2.93,
-      "learning_rate": 1.932538803003734e-05,
-      "loss": 0.2673,
+      "epoch": 3.0008649583738785,
+      "grad_norm": 0.21151970326900482,
+      "learning_rate": 1.820727652802361e-05,
+      "loss": 0.3948,
       "step": 83265
     },
     {
-      "epoch": 2.93,
-      "learning_rate": 1.9322613701824188e-05,
-      "loss": 0.2695,
+      "epoch": 3.0010451580351027,
+      "grad_norm": 0.27678802609443665,
+      "learning_rate": 1.8204468222375687e-05,
+      "loss": 0.3665,
       "step": 83270
     },
     {
-      "epoch": 2.93,
-      "learning_rate": 1.9319839447328167e-05,
-      "loss": 0.2779,
+      "epoch": 3.0012253576963275,
+      "grad_norm": 0.2054269164800644,
+      "learning_rate": 1.8201660009315563e-05,
+      "loss": 0.3515,
       "step": 83275
     },
     {
-      "epoch": 2.93,
-      "learning_rate": 1.9317065266585283e-05,
-      "loss": 0.2721,
+      "epoch": 3.001405557357552,
+      "grad_norm": 0.27250099182128906,
+      "learning_rate": 1.8198851888881513e-05,
+      "loss": 0.3725,
       "step": 83280
     },
     {
-      "epoch": 2.93,
-      "learning_rate": 1.9314291159631566e-05,
-      "loss": 0.2656,
+      "epoch": 3.001585757018777,
+      "grad_norm": 0.22905905544757843,
+      "learning_rate": 1.8196043861111788e-05,
+      "loss": 0.3983,
       "step": 83285
     },
     {
-      "epoch": 2.93,
-      "learning_rate": 1.931151712650302e-05,
-      "loss": 0.2559,
+      "epoch": 3.0017659566800012,
+      "grad_norm": 0.19048994779586792,
+      "learning_rate": 1.819323592604464e-05,
+      "loss": 0.3538,
       "step": 83290
     },
     {
-      "epoch": 2.93,
-      "learning_rate": 1.9308743167235683e-05,
-      "loss": 0.2672,
+      "epoch": 3.001946156341226,
+      "grad_norm": 0.2080589234828949,
+      "learning_rate": 1.8190428083718346e-05,
+      "loss": 0.3769,
       "step": 83295
     },
     {
-      "epoch": 2.93,
-      "learning_rate": 1.9305969281865568e-05,
-      "loss": 0.2542,
+      "epoch": 3.0021263560024507,
+      "grad_norm": 0.20519669353961945,
+      "learning_rate": 1.8187620334171147e-05,
+      "loss": 0.38,
       "step": 83300
     },
     {
-      "epoch": 2.93,
-      "learning_rate": 1.9303195470428687e-05,
-      "loss": 0.2666,
+      "epoch": 3.0023065556636754,
+      "grad_norm": 0.20568309724330902,
+      "learning_rate": 1.81848126774413e-05,
+      "loss": 0.4157,
       "step": 83305
     },
     {
-      "epoch": 2.93,
-      "learning_rate": 1.9300421732961055e-05,
-      "loss": 0.2709,
+      "epoch": 3.0024867553249,
+      "grad_norm": 0.1868577003479004,
+      "learning_rate": 1.8182005113567064e-05,
+      "loss": 0.3971,
       "step": 83310
     },
     {
-      "epoch": 2.93,
-      "learning_rate": 1.9297648069498692e-05,
-      "loss": 0.2806,
+      "epoch": 3.0026669549861245,
+      "grad_norm": 0.21570152044296265,
+      "learning_rate": 1.817919764258668e-05,
+      "loss": 0.3962,
       "step": 83315
     },
     {
-      "epoch": 2.93,
-      "learning_rate": 1.9294874480077613e-05,
-      "loss": 0.2487,
+      "epoch": 3.002847154647349,
+      "grad_norm": 0.2413526475429535,
+      "learning_rate": 1.817639026453842e-05,
+      "loss": 0.3816,
       "step": 83320
     },
     {
-      "epoch": 2.93,
-      "learning_rate": 1.9292100964733835e-05,
-      "loss": 0.2662,
+      "epoch": 3.003027354308574,
+      "grad_norm": 0.2983512282371521,
+      "learning_rate": 1.8173582979460517e-05,
+      "loss": 0.4202,
       "step": 83325
     },
     {
-      "epoch": 2.93,
-      "learning_rate": 1.928932752350335e-05,
-      "loss": 0.2892,
+      "epoch": 3.0032075539697987,
+      "grad_norm": 0.18553189933300018,
+      "learning_rate": 1.8170775787391213e-05,
+      "loss": 0.3656,
       "step": 83330
     },
     {
-      "epoch": 2.93,
-      "learning_rate": 1.9286554156422188e-05,
-      "loss": 0.2582,
+      "epoch": 3.0033877536310234,
+      "grad_norm": 0.2300480157136917,
+      "learning_rate": 1.8167968688368787e-05,
+      "loss": 0.3999,
       "step": 83335
     },
     {
-      "epoch": 2.93,
-      "learning_rate": 1.928378086352636e-05,
-      "loss": 0.2557,
+      "epoch": 3.0035679532922477,
+      "grad_norm": 0.18656012415885925,
+      "learning_rate": 1.8165161682431444e-05,
+      "loss": 0.3956,
       "step": 83340
     },
     {
-      "epoch": 2.93,
-      "learning_rate": 1.928100764485186e-05,
-      "loss": 0.269,
+      "epoch": 3.0037481529534724,
+      "grad_norm": 0.18259282410144806,
+      "learning_rate": 1.8162354769617468e-05,
+      "loss": 0.3691,
       "step": 83345
     },
     {
-      "epoch": 2.93,
-      "learning_rate": 1.927823450043472e-05,
-      "loss": 0.2733,
+      "epoch": 3.003928352614697,
+      "grad_norm": 0.21580132842063904,
+      "learning_rate": 1.815954794996508e-05,
+      "loss": 0.3893,
       "step": 83350
     },
     {
-      "epoch": 2.93,
-      "learning_rate": 1.927546143031093e-05,
-      "loss": 0.267,
+      "epoch": 3.004108552275922,
+      "grad_norm": 0.20566125214099884,
+      "learning_rate": 1.815674122351252e-05,
+      "loss": 0.3643,
       "step": 83355
     },
     {
-      "epoch": 2.93,
-      "learning_rate": 1.92726884345165e-05,
-      "loss": 0.2704,
+      "epoch": 3.004288751937146,
+      "grad_norm": 0.2558181583881378,
+      "learning_rate": 1.815393459029804e-05,
+      "loss": 0.3879,
       "step": 83360
     },
     {
-      "epoch": 2.93,
-      "learning_rate": 1.9269915513087437e-05,
-      "loss": 0.277,
+      "epoch": 3.004468951598371,
+      "grad_norm": 0.1984751671552658,
+      "learning_rate": 1.815112805035988e-05,
+      "loss": 0.4131,
       "step": 83365
     },
     {
-      "epoch": 2.93,
-      "learning_rate": 1.926714266605974e-05,
-      "loss": 0.2227,
+      "epoch": 3.0046491512595956,
+      "grad_norm": 0.19502590596675873,
+      "learning_rate": 1.8148321603736263e-05,
+      "loss": 0.3646,
       "step": 83370
     },
     {
-      "epoch": 2.93,
-      "learning_rate": 1.9264369893469424e-05,
-      "loss": 0.2612,
+      "epoch": 3.0048293509208204,
+      "grad_norm": 0.22936676442623138,
+      "learning_rate": 1.8145515250465446e-05,
+      "loss": 0.3653,
       "step": 83375
     },
     {
-      "epoch": 2.93,
-      "learning_rate": 1.9261597195352486e-05,
-      "loss": 0.2579,
+      "epoch": 3.005009550582045,
+      "grad_norm": 0.2669048309326172,
+      "learning_rate": 1.814270899058565e-05,
+      "loss": 0.4142,
       "step": 83380
     },
     {
-      "epoch": 2.93,
-      "learning_rate": 1.9258824571744917e-05,
-      "loss": 0.2713,
+      "epoch": 3.0051897502432694,
+      "grad_norm": 0.21119844913482666,
+      "learning_rate": 1.8139902824135124e-05,
+      "loss": 0.4132,
       "step": 83385
     },
     {
-      "epoch": 2.93,
-      "learning_rate": 1.9256052022682738e-05,
-      "loss": 0.2766,
+      "epoch": 3.005369949904494,
+      "grad_norm": 0.20271877944469452,
+      "learning_rate": 1.8137096751152093e-05,
+      "loss": 0.3699,
       "step": 83390
     },
     {
-      "epoch": 2.93,
-      "learning_rate": 1.925327954820193e-05,
-      "loss": 0.27,
+      "epoch": 3.005550149565719,
+      "grad_norm": 0.21155229210853577,
+      "learning_rate": 1.8134290771674784e-05,
+      "loss": 0.3365,
       "step": 83395
     },
     {
-      "epoch": 2.93,
-      "learning_rate": 1.925050714833851e-05,
-      "loss": 0.2737,
+      "epoch": 3.0057303492269436,
+      "grad_norm": 0.21775977313518524,
+      "learning_rate": 1.813148488574144e-05,
+      "loss": 0.3988,
       "step": 83400
     },
     {
-      "epoch": 2.93,
-      "learning_rate": 1.924773482312845e-05,
-      "loss": 0.2652,
+      "epoch": 3.005910548888168,
+      "grad_norm": 0.2257860004901886,
+      "learning_rate": 1.8128679093390282e-05,
+      "loss": 0.3699,
       "step": 83405
     },
     {
-      "epoch": 2.93,
-      "learning_rate": 1.9244962572607776e-05,
-      "loss": 0.2849,
+      "epoch": 3.0060907485493926,
+      "grad_norm": 0.1963503360748291,
+      "learning_rate": 1.8125873394659543e-05,
+      "loss": 0.358,
       "step": 83410
     },
     {
-      "epoch": 2.93,
-      "learning_rate": 1.9242190396812472e-05,
-      "loss": 0.2707,
+      "epoch": 3.0062709482106174,
+      "grad_norm": 0.17363089323043823,
+      "learning_rate": 1.8123067789587443e-05,
+      "loss": 0.3867,
       "step": 83415
     },
     {
-      "epoch": 2.93,
-      "learning_rate": 1.9239418295778518e-05,
-      "loss": 0.2855,
+      "epoch": 3.006451147871842,
+      "grad_norm": 0.22728854417800903,
+      "learning_rate": 1.8120262278212216e-05,
+      "loss": 0.3821,
       "step": 83420
     },
     {
-      "epoch": 2.94,
-      "learning_rate": 1.923664626954192e-05,
-      "loss": 0.2633,
+      "epoch": 3.006631347533067,
+      "grad_norm": 0.2198934406042099,
+      "learning_rate": 1.8117456860572085e-05,
+      "loss": 0.3799,
       "step": 83425
     },
     {
-      "epoch": 2.94,
-      "learning_rate": 1.9233874318138688e-05,
-      "loss": 0.2519,
+      "epoch": 3.006811547194291,
+      "grad_norm": 0.27746328711509705,
+      "learning_rate": 1.811465153670528e-05,
+      "loss": 0.3917,
       "step": 83430
     },
     {
-      "epoch": 2.94,
-      "learning_rate": 1.9231102441604793e-05,
-      "loss": 0.256,
+      "epoch": 3.006991746855516,
+      "grad_norm": 0.25935912132263184,
+      "learning_rate": 1.811184630665e-05,
+      "loss": 0.3629,
       "step": 83435
     },
     {
-      "epoch": 2.94,
-      "learning_rate": 1.9228330639976223e-05,
-      "loss": 0.2413,
+      "epoch": 3.0071719465167406,
+      "grad_norm": 0.2055579125881195,
+      "learning_rate": 1.810904117044449e-05,
+      "loss": 0.361,
       "step": 83440
     },
     {
-      "epoch": 2.94,
-      "learning_rate": 1.9225558913288982e-05,
-      "loss": 0.2396,
+      "epoch": 3.0073521461779653,
+      "grad_norm": 0.2863004207611084,
+      "learning_rate": 1.810623612812696e-05,
+      "loss": 0.3877,
       "step": 83445
     },
     {
-      "epoch": 2.94,
-      "learning_rate": 1.922278726157905e-05,
-      "loss": 0.2843,
+      "epoch": 3.0075323458391896,
+      "grad_norm": 0.21766206622123718,
+      "learning_rate": 1.810343117973562e-05,
+      "loss": 0.3755,
       "step": 83450
     },
     {
-      "epoch": 2.94,
-      "learning_rate": 1.922001568488242e-05,
-      "loss": 0.256,
+      "epoch": 3.0077125455004143,
+      "grad_norm": 0.21764706075191498,
+      "learning_rate": 1.81006263253087e-05,
+      "loss": 0.4227,
       "step": 83455
     },
     {
-      "epoch": 2.94,
-      "learning_rate": 1.9217244183235073e-05,
-      "loss": 0.2692,
+      "epoch": 3.007892745161639,
+      "grad_norm": 0.25549066066741943,
+      "learning_rate": 1.8097821564884408e-05,
+      "loss": 0.4138,
       "step": 83460
     },
     {
-      "epoch": 2.94,
-      "learning_rate": 1.9214472756673008e-05,
-      "loss": 0.2827,
+      "epoch": 3.008072944822864,
+      "grad_norm": 0.22530651092529297,
+      "learning_rate": 1.8095016898500962e-05,
+      "loss": 0.4006,
       "step": 83465
     },
     {
-      "epoch": 2.94,
-      "learning_rate": 1.92117014052322e-05,
-      "loss": 0.2706,
+      "epoch": 3.0082531444840885,
+      "grad_norm": 0.20702116191387177,
+      "learning_rate": 1.8092212326196578e-05,
+      "loss": 0.3831,
       "step": 83470
     },
     {
-      "epoch": 2.94,
-      "learning_rate": 1.920893012894863e-05,
-      "loss": 0.2947,
+      "epoch": 3.008433344145313,
+      "grad_norm": 0.21814846992492676,
+      "learning_rate": 1.8089407848009457e-05,
+      "loss": 0.3723,
       "step": 83475
     },
     {
-      "epoch": 2.94,
-      "learning_rate": 1.920615892785828e-05,
-      "loss": 0.278,
+      "epoch": 3.0086135438065376,
+      "grad_norm": 0.19958257675170898,
+      "learning_rate": 1.808660346397782e-05,
+      "loss": 0.3971,
       "step": 83480
     },
     {
-      "epoch": 2.94,
-      "learning_rate": 1.9203387801997148e-05,
-      "loss": 0.2691,
+      "epoch": 3.0087937434677623,
+      "grad_norm": 0.2572198510169983,
+      "learning_rate": 1.8083799174139876e-05,
+      "loss": 0.3909,
       "step": 83485
     },
     {
-      "epoch": 2.94,
-      "learning_rate": 1.9200616751401207e-05,
-      "loss": 0.2807,
+      "epoch": 3.008973943128987,
+      "grad_norm": 0.1767335683107376,
+      "learning_rate": 1.8080994978533822e-05,
+      "loss": 0.3866,
       "step": 83490
     },
     {
-      "epoch": 2.94,
-      "learning_rate": 1.919784577610643e-05,
-      "loss": 0.2716,
+      "epoch": 3.0091541427902118,
+      "grad_norm": 0.22340673208236694,
+      "learning_rate": 1.8078190877197876e-05,
+      "loss": 0.3917,
       "step": 83495
     },
     {
-      "epoch": 2.94,
-      "learning_rate": 1.91950748761488e-05,
-      "loss": 0.2561,
+      "epoch": 3.009334342451436,
+      "grad_norm": 0.21768926084041595,
+      "learning_rate": 1.8075386870170233e-05,
+      "loss": 0.3831,
       "step": 83500
     },
     {
-      "epoch": 2.94,
-      "eval_loss": 0.2586362063884735,
-      "eval_runtime": 10.5563,
-      "eval_samples_per_second": 9.473,
-      "eval_steps_per_second": 9.473,
+      "epoch": 3.009334342451436,
+      "eval_loss": 0.4312828481197357,
+      "eval_runtime": 3.5249,
+      "eval_samples_per_second": 28.369,
+      "eval_steps_per_second": 7.092,
       "step": 83500
     },
     {
-      "epoch": 2.94,
-      "learning_rate": 1.9192304051564296e-05,
-      "loss": 0.2662,
+      "epoch": 3.009514542112661,
+      "grad_norm": 0.23713305592536926,
+      "learning_rate": 1.8072582957489108e-05,
+      "loss": 0.3742,
       "step": 83505
     },
     {
-      "epoch": 2.94,
-      "learning_rate": 1.9189533302388906e-05,
-      "loss": 0.2656,
+      "epoch": 3.0096947417738855,
+      "grad_norm": 0.18255163729190826,
+      "learning_rate": 1.806977913919271e-05,
+      "loss": 0.407,
       "step": 83510
     },
     {
-      "epoch": 2.94,
-      "learning_rate": 1.9186762628658594e-05,
-      "loss": 0.2826,
+      "epoch": 3.0098749414351103,
+      "grad_norm": 0.20699362456798553,
+      "learning_rate": 1.806697541531921e-05,
+      "loss": 0.348,
       "step": 83515
     },
     {
-      "epoch": 2.94,
-      "learning_rate": 1.9183992030409333e-05,
-      "loss": 0.2547,
+      "epoch": 3.0100551410963345,
+      "grad_norm": 0.22750404477119446,
+      "learning_rate": 1.806417178590684e-05,
+      "loss": 0.3584,
       "step": 83520
     },
     {
-      "epoch": 2.94,
-      "learning_rate": 1.9181221507677104e-05,
-      "loss": 0.2424,
+      "epoch": 3.0102353407575593,
+      "grad_norm": 0.2072783261537552,
+      "learning_rate": 1.8061368250993777e-05,
+      "loss": 0.42,
       "step": 83525
     },
     {
-      "epoch": 2.94,
-      "learning_rate": 1.917845106049788e-05,
-      "loss": 0.2857,
+      "epoch": 3.010415540418784,
+      "grad_norm": 0.18794316053390503,
+      "learning_rate": 1.8058564810618233e-05,
+      "loss": 0.3754,
       "step": 83530
     },
     {
-      "epoch": 2.94,
-      "learning_rate": 1.917568068890764e-05,
-      "loss": 0.2393,
+      "epoch": 3.0105957400800087,
+      "grad_norm": 0.15637893974781036,
+      "learning_rate": 1.805576146481841e-05,
+      "loss": 0.3791,
       "step": 83535
     },
     {
-      "epoch": 2.94,
-      "learning_rate": 1.917291039294233e-05,
-      "loss": 0.2903,
+      "epoch": 3.0107759397412335,
+      "grad_norm": 0.23354141414165497,
+      "learning_rate": 1.8052958213632477e-05,
+      "loss": 0.3636,
       "step": 83540
     },
     {
-      "epoch": 2.94,
-      "learning_rate": 1.9170140172637953e-05,
-      "loss": 0.2738,
+      "epoch": 3.0109561394024578,
+      "grad_norm": 0.20593926310539246,
+      "learning_rate": 1.8050155057098656e-05,
+      "loss": 0.35,
       "step": 83545
     },
     {
-      "epoch": 2.94,
-      "learning_rate": 1.916737002803046e-05,
-      "loss": 0.2613,
+      "epoch": 3.0111363390636825,
+      "grad_norm": 0.26360243558883667,
+      "learning_rate": 1.8047351995255123e-05,
+      "loss": 0.3946,
       "step": 83550
     },
     {
-      "epoch": 2.94,
-      "learning_rate": 1.9164599959155823e-05,
-      "loss": 0.2761,
+      "epoch": 3.0113165387249072,
+      "grad_norm": 0.20182831585407257,
+      "learning_rate": 1.8044549028140068e-05,
+      "loss": 0.3674,
       "step": 83555
     },
     {
-      "epoch": 2.94,
-      "learning_rate": 1.9161829966050003e-05,
-      "loss": 0.27,
+      "epoch": 3.011496738386132,
+      "grad_norm": 0.1877097487449646,
+      "learning_rate": 1.804174615579169e-05,
+      "loss": 0.3906,
       "step": 83560
     },
     {
-      "epoch": 2.94,
-      "learning_rate": 1.9159060048748984e-05,
-      "loss": 0.2818,
+      "epoch": 3.0116769380473563,
+      "grad_norm": 0.23039346933364868,
+      "learning_rate": 1.8038943378248165e-05,
+      "loss": 0.3893,
       "step": 83565
     },
     {
-      "epoch": 2.94,
-      "learning_rate": 1.9156290207288723e-05,
-      "loss": 0.2836,
+      "epoch": 3.011857137708581,
+      "grad_norm": 0.25139108300209045,
+      "learning_rate": 1.80361406955477e-05,
+      "loss": 0.3811,
       "step": 83570
     },
     {
-      "epoch": 2.94,
-      "learning_rate": 1.9153520441705168e-05,
-      "loss": 0.2686,
+      "epoch": 3.0120373373698057,
+      "grad_norm": 0.2260846644639969,
+      "learning_rate": 1.8033338107728465e-05,
+      "loss": 0.3805,
       "step": 83575
     },
     {
-      "epoch": 2.94,
-      "learning_rate": 1.915075075203431e-05,
-      "loss": 0.2763,
+      "epoch": 3.0122175370310305,
+      "grad_norm": 0.21495521068572998,
+      "learning_rate": 1.803053561482865e-05,
+      "loss": 0.3935,
       "step": 83580
     },
     {
-      "epoch": 2.94,
-      "learning_rate": 1.9147981138312093e-05,
-      "loss": 0.2754,
+      "epoch": 3.012397736692255,
+      "grad_norm": 0.24546730518341064,
+      "learning_rate": 1.802773321688644e-05,
+      "loss": 0.3861,
       "step": 83585
     },
     {
-      "epoch": 2.94,
-      "learning_rate": 1.914521160057449e-05,
-      "loss": 0.2634,
+      "epoch": 3.0125779363534795,
+      "grad_norm": 0.2514590322971344,
+      "learning_rate": 1.8024930913940015e-05,
+      "loss": 0.3718,
       "step": 83590
     },
     {
-      "epoch": 2.94,
-      "learning_rate": 1.914244213885744e-05,
-      "loss": 0.2856,
+      "epoch": 3.012758136014704,
+      "grad_norm": 0.23116329312324524,
+      "learning_rate": 1.8022128706027547e-05,
+      "loss": 0.354,
       "step": 83595
     },
     {
-      "epoch": 2.94,
-      "learning_rate": 1.9139672753196936e-05,
-      "loss": 0.2785,
+      "epoch": 3.012938335675929,
+      "grad_norm": 0.17475327849388123,
+      "learning_rate": 1.8019326593187232e-05,
+      "loss": 0.3732,
       "step": 83600
     },
     {
-      "epoch": 2.94,
-      "learning_rate": 1.9136903443628916e-05,
-      "loss": 0.255,
+      "epoch": 3.0131185353371537,
+      "grad_norm": 0.2183002233505249,
+      "learning_rate": 1.801652457545724e-05,
+      "loss": 0.3881,
       "step": 83605
     },
     {
-      "epoch": 2.94,
-      "learning_rate": 1.913413421018934e-05,
-      "loss": 0.2541,
+      "epoch": 3.0132987349983784,
+      "grad_norm": 0.23634588718414307,
+      "learning_rate": 1.8013722652875747e-05,
+      "loss": 0.4003,
       "step": 83610
     },
     {
-      "epoch": 2.94,
-      "learning_rate": 1.9131365052914154e-05,
-      "loss": 0.2687,
+      "epoch": 3.0134789346596027,
+      "grad_norm": 0.22611938416957855,
+      "learning_rate": 1.801092082548094e-05,
+      "loss": 0.3966,
       "step": 83615
     },
     {
-      "epoch": 2.94,
-      "learning_rate": 1.912859597183934e-05,
-      "loss": 0.2711,
+      "epoch": 3.0136591343208274,
+      "grad_norm": 0.23099973797798157,
+      "learning_rate": 1.8008119093310965e-05,
+      "loss": 0.3739,
       "step": 83620
     },
     {
-      "epoch": 2.94,
-      "learning_rate": 1.9125826967000836e-05,
-      "loss": 0.2686,
+      "epoch": 3.013839333982052,
+      "grad_norm": 0.2718352675437927,
+      "learning_rate": 1.8005317456404034e-05,
+      "loss": 0.3796,
       "step": 83625
     },
     {
-      "epoch": 2.94,
-      "learning_rate": 1.91230580384346e-05,
-      "loss": 0.2434,
+      "epoch": 3.014019533643277,
+      "grad_norm": 0.18968239426612854,
+      "learning_rate": 1.800251591479829e-05,
+      "loss": 0.3575,
       "step": 83630
     },
     {
-      "epoch": 2.94,
-      "learning_rate": 1.912028918617657e-05,
-      "loss": 0.3043,
+      "epoch": 3.014199733304501,
+      "grad_norm": 0.1754741370677948,
+      "learning_rate": 1.7999714468531906e-05,
+      "loss": 0.374,
       "step": 83635
     },
     {
-      "epoch": 2.94,
-      "learning_rate": 1.911752041026271e-05,
-      "loss": 0.2653,
+      "epoch": 3.014379932965726,
+      "grad_norm": 0.25998741388320923,
+      "learning_rate": 1.7996913117643064e-05,
+      "loss": 0.3882,
       "step": 83640
     },
     {
-      "epoch": 2.94,
-      "learning_rate": 1.9114751710728983e-05,
-      "loss": 0.2592,
+      "epoch": 3.0145601326269507,
+      "grad_norm": 0.25412383675575256,
+      "learning_rate": 1.799411186216992e-05,
+      "loss": 0.3873,
       "step": 83645
     },
     {
-      "epoch": 2.94,
-      "learning_rate": 1.9111983087611328e-05,
-      "loss": 0.2862,
+      "epoch": 3.0147403322881754,
+      "grad_norm": 0.23785263299942017,
+      "learning_rate": 1.7991310702150647e-05,
+      "loss": 0.3314,
       "step": 83650
     },
     {
-      "epoch": 2.94,
-      "learning_rate": 1.9109214540945677e-05,
-      "loss": 0.2584,
+      "epoch": 3.0149205319494,
+      "grad_norm": 0.21944360435009003,
+      "learning_rate": 1.7988509637623414e-05,
+      "loss": 0.3885,
       "step": 83655
     },
     {
-      "epoch": 2.94,
-      "learning_rate": 1.9106446070768007e-05,
-      "loss": 0.2845,
+      "epoch": 3.0151007316106244,
+      "grad_norm": 0.20874114334583282,
+      "learning_rate": 1.7985708668626373e-05,
+      "loss": 0.3503,
       "step": 83660
     },
     {
-      "epoch": 2.94,
-      "learning_rate": 1.9103677677114244e-05,
-      "loss": 0.264,
+      "epoch": 3.015280931271849,
+      "grad_norm": 0.19543851912021637,
+      "learning_rate": 1.7982907795197695e-05,
+      "loss": 0.3552,
       "step": 83665
     },
     {
-      "epoch": 2.94,
-      "learning_rate": 1.9100909360020343e-05,
-      "loss": 0.2732,
+      "epoch": 3.015461130933074,
+      "grad_norm": 0.17719921469688416,
+      "learning_rate": 1.7980107017375543e-05,
+      "loss": 0.3681,
       "step": 83670
     },
     {
-      "epoch": 2.94,
-      "learning_rate": 1.909814111952224e-05,
-      "loss": 0.2679,
+      "epoch": 3.0156413305942986,
+      "grad_norm": 0.20741024613380432,
+      "learning_rate": 1.7977306335198067e-05,
+      "loss": 0.3569,
       "step": 83675
     },
     {
-      "epoch": 2.94,
-      "learning_rate": 1.9095372955655894e-05,
-      "loss": 0.267,
+      "epoch": 3.015821530255523,
+      "grad_norm": 0.2016208916902542,
+      "learning_rate": 1.797450574870344e-05,
+      "loss": 0.3684,
       "step": 83680
     },
     {
-      "epoch": 2.94,
-      "learning_rate": 1.9092604868457238e-05,
-      "loss": 0.2704,
+      "epoch": 3.0160017299167476,
+      "grad_norm": 0.23120005428791046,
+      "learning_rate": 1.7971705257929806e-05,
+      "loss": 0.3627,
       "step": 83685
     },
     {
-      "epoch": 2.94,
-      "learning_rate": 1.9089836857962203e-05,
-      "loss": 0.2713,
+      "epoch": 3.0161819295779724,
+      "grad_norm": 0.22338317334651947,
+      "learning_rate": 1.796890486291533e-05,
+      "loss": 0.3849,
       "step": 83690
     },
     {
-      "epoch": 2.94,
-      "learning_rate": 1.908706892420675e-05,
-      "loss": 0.2662,
+      "epoch": 3.016362129239197,
+      "grad_norm": 0.2729624807834625,
+      "learning_rate": 1.7966104563698165e-05,
+      "loss": 0.3771,
       "step": 83695
     },
     {
-      "epoch": 2.94,
-      "learning_rate": 1.9084301067226815e-05,
-      "loss": 0.2738,
+      "epoch": 3.016542328900422,
+      "grad_norm": 0.24947573244571686,
+      "learning_rate": 1.796330436031646e-05,
+      "loss": 0.3744,
       "step": 83700
     },
     {
-      "epoch": 2.94,
-      "learning_rate": 1.9081533287058328e-05,
-      "loss": 0.2564,
+      "epoch": 3.016722528561646,
+      "grad_norm": 0.22686618566513062,
+      "learning_rate": 1.7960504252808374e-05,
+      "loss": 0.3786,
       "step": 83705
     },
     {
-      "epoch": 2.95,
-      "learning_rate": 1.9078765583737223e-05,
-      "loss": 0.2507,
+      "epoch": 3.016902728222871,
+      "grad_norm": 0.2134973108768463,
+      "learning_rate": 1.7957704241212063e-05,
+      "loss": 0.4183,
       "step": 83710
     },
     {
-      "epoch": 2.95,
-      "learning_rate": 1.9075997957299456e-05,
-      "loss": 0.2706,
+      "epoch": 3.0170829278840956,
+      "grad_norm": 0.17952316999435425,
+      "learning_rate": 1.795490432556565e-05,
+      "loss": 0.3963,
       "step": 83715
     },
     {
-      "epoch": 2.95,
-      "learning_rate": 1.9073230407780947e-05,
-      "loss": 0.2618,
+      "epoch": 3.0172631275453203,
+      "grad_norm": 0.22514955699443817,
+      "learning_rate": 1.795210450590732e-05,
+      "loss": 0.3789,
       "step": 83720
     },
     {
-      "epoch": 2.95,
-      "learning_rate": 1.9070462935217638e-05,
-      "loss": 0.277,
+      "epoch": 3.017443327206545,
+      "grad_norm": 0.19981583952903748,
+      "learning_rate": 1.794930478227518e-05,
+      "loss": 0.3763,
       "step": 83725
     },
     {
-      "epoch": 2.95,
-      "learning_rate": 1.9067695539645453e-05,
-      "loss": 0.2748,
+      "epoch": 3.0176235268677694,
+      "grad_norm": 0.2752089500427246,
+      "learning_rate": 1.7946505154707416e-05,
+      "loss": 0.3775,
       "step": 83730
     },
     {
-      "epoch": 2.95,
-      "learning_rate": 1.906492822110034e-05,
-      "loss": 0.2872,
+      "epoch": 3.017803726528994,
+      "grad_norm": 0.21153953671455383,
+      "learning_rate": 1.7943705623242153e-05,
+      "loss": 0.3939,
       "step": 83735
     },
     {
-      "epoch": 2.95,
-      "learning_rate": 1.9062160979618224e-05,
-      "loss": 0.256,
+      "epoch": 3.017983926190219,
+      "grad_norm": 0.22683672606945038,
+      "learning_rate": 1.794090618791752e-05,
+      "loss": 0.3739,
       "step": 83740
     },
     {
-      "epoch": 2.95,
-      "learning_rate": 1.9059393815235026e-05,
-      "loss": 0.2739,
+      "epoch": 3.0181641258514436,
+      "grad_norm": 0.2554776668548584,
+      "learning_rate": 1.793810684877169e-05,
+      "loss": 0.4071,
       "step": 83745
     },
     {
-      "epoch": 2.95,
-      "learning_rate": 1.9056626727986686e-05,
-      "loss": 0.2743,
+      "epoch": 3.018344325512668,
+      "grad_norm": 0.2107311338186264,
+      "learning_rate": 1.7935307605842775e-05,
+      "loss": 0.3646,
       "step": 83750
     },
     {
-      "epoch": 2.95,
-      "learning_rate": 1.9053859717909132e-05,
-      "loss": 0.2674,
+      "epoch": 3.0185245251738926,
+      "grad_norm": 0.22267073392868042,
+      "learning_rate": 1.793250845916893e-05,
+      "loss": 0.4121,
       "step": 83755
     },
     {
-      "epoch": 2.95,
-      "learning_rate": 1.9051092785038297e-05,
-      "loss": 0.2634,
+      "epoch": 3.0187047248351173,
+      "grad_norm": 0.2603355646133423,
+      "learning_rate": 1.7929709408788292e-05,
+      "loss": 0.371,
       "step": 83760
     },
     {
-      "epoch": 2.95,
-      "learning_rate": 1.9048325929410103e-05,
-      "loss": 0.2678,
+      "epoch": 3.018884924496342,
+      "grad_norm": 0.22211819887161255,
+      "learning_rate": 1.792691045473899e-05,
+      "loss": 0.353,
       "step": 83765
     },
     {
-      "epoch": 2.95,
-      "learning_rate": 1.904555915106046e-05,
-      "loss": 0.2741,
+      "epoch": 3.019065124157567,
+      "grad_norm": 0.1939542144536972,
+      "learning_rate": 1.792411159705917e-05,
+      "loss": 0.3708,
       "step": 83770
     },
     {
-      "epoch": 2.95,
-      "learning_rate": 1.9042792450025313e-05,
-      "loss": 0.2709,
+      "epoch": 3.019245323818791,
+      "grad_norm": 0.23032665252685547,
+      "learning_rate": 1.7921312835786952e-05,
+      "loss": 0.3856,
       "step": 83775
     },
     {
-      "epoch": 2.95,
-      "learning_rate": 1.904002582634059e-05,
-      "loss": 0.2552,
+      "epoch": 3.019425523480016,
+      "grad_norm": 0.22775132954120636,
+      "learning_rate": 1.7918514170960478e-05,
+      "loss": 0.3811,
       "step": 83780
     },
     {
-      "epoch": 2.95,
-      "learning_rate": 1.9037259280042197e-05,
-      "loss": 0.2854,
+      "epoch": 3.0196057231412405,
+      "grad_norm": 0.2607553005218506,
+      "learning_rate": 1.791571560261788e-05,
+      "loss": 0.3733,
       "step": 83785
     },
     {
-      "epoch": 2.95,
-      "learning_rate": 1.9034492811166056e-05,
-      "loss": 0.2707,
+      "epoch": 3.0197859228024653,
+      "grad_norm": 0.22313840687274933,
+      "learning_rate": 1.791291713079728e-05,
+      "loss": 0.4006,
       "step": 83790
     },
     {
-      "epoch": 2.95,
-      "learning_rate": 1.9031726419748103e-05,
-      "loss": 0.2613,
+      "epoch": 3.0199661224636896,
+      "grad_norm": 0.22880607843399048,
+      "learning_rate": 1.791011875553682e-05,
+      "loss": 0.4368,
       "step": 83795
     },
     {
-      "epoch": 2.95,
-      "learning_rate": 1.9028960105824245e-05,
-      "loss": 0.27,
+      "epoch": 3.0201463221249143,
+      "grad_norm": 0.2502082288265228,
+      "learning_rate": 1.790732047687462e-05,
+      "loss": 0.3866,
       "step": 83800
     },
     {
-      "epoch": 2.95,
-      "learning_rate": 1.9026193869430404e-05,
-      "loss": 0.2898,
+      "epoch": 3.020326521786139,
+      "grad_norm": 0.23791562020778656,
+      "learning_rate": 1.7904522294848792e-05,
+      "loss": 0.3749,
       "step": 83805
     },
     {
-      "epoch": 2.95,
-      "learning_rate": 1.9023427710602504e-05,
-      "loss": 0.261,
+      "epoch": 3.0205067214473638,
+      "grad_norm": 0.22234022617340088,
+      "learning_rate": 1.790172420949749e-05,
+      "loss": 0.3662,
       "step": 83810
     },
     {
-      "epoch": 2.95,
-      "learning_rate": 1.9020661629376456e-05,
-      "loss": 0.2876,
+      "epoch": 3.0206869211085885,
+      "grad_norm": 0.169847309589386,
+      "learning_rate": 1.7898926220858817e-05,
+      "loss": 0.361,
       "step": 83815
     },
     {
-      "epoch": 2.95,
-      "learning_rate": 1.9017895625788178e-05,
-      "loss": 0.2652,
+      "epoch": 3.020867120769813,
+      "grad_norm": 0.1907045543193817,
+      "learning_rate": 1.7896128328970886e-05,
+      "loss": 0.3964,
       "step": 83820
     },
     {
-      "epoch": 2.95,
-      "learning_rate": 1.9015129699873573e-05,
-      "loss": 0.2626,
+      "epoch": 3.0210473204310375,
+      "grad_norm": 0.21488279104232788,
+      "learning_rate": 1.789333053387185e-05,
+      "loss": 0.3986,
       "step": 83825
     },
     {
-      "epoch": 2.95,
-      "learning_rate": 1.9012363851668576e-05,
-      "loss": 0.2728,
+      "epoch": 3.0212275200922623,
+      "grad_norm": 0.193211168050766,
+      "learning_rate": 1.7890532835599793e-05,
+      "loss": 0.3955,
       "step": 83830
     },
     {
-      "epoch": 2.95,
-      "learning_rate": 1.900959808120908e-05,
-      "loss": 0.2711,
+      "epoch": 3.021407719753487,
+      "grad_norm": 0.2113611400127411,
+      "learning_rate": 1.7887735234192864e-05,
+      "loss": 0.3767,
       "step": 83835
     },
     {
-      "epoch": 2.95,
-      "learning_rate": 1.900683238853101e-05,
-      "loss": 0.2751,
+      "epoch": 3.0215879194147117,
+      "grad_norm": 0.2665266990661621,
+      "learning_rate": 1.788493772968916e-05,
+      "loss": 0.3705,
       "step": 83840
     },
     {
-      "epoch": 2.95,
-      "learning_rate": 1.9004066773670265e-05,
-      "loss": 0.2988,
+      "epoch": 3.021768119075936,
+      "grad_norm": 0.24903304874897003,
+      "learning_rate": 1.78821403221268e-05,
+      "loss": 0.4004,
       "step": 83845
     },
     {
-      "epoch": 2.95,
-      "learning_rate": 1.900130123666277e-05,
-      "loss": 0.2605,
+      "epoch": 3.0219483187371607,
+      "grad_norm": 0.2464405745267868,
+      "learning_rate": 1.7879343011543905e-05,
+      "loss": 0.3908,
       "step": 83850
     },
     {
-      "epoch": 2.95,
-      "learning_rate": 1.8998535777544424e-05,
-      "loss": 0.2607,
+      "epoch": 3.0221285183983855,
+      "grad_norm": 0.2581247389316559,
+      "learning_rate": 1.7876545797978576e-05,
+      "loss": 0.3983,
       "step": 83855
     },
     {
-      "epoch": 2.95,
-      "learning_rate": 1.8995770396351134e-05,
-      "loss": 0.2659,
+      "epoch": 3.02230871805961,
+      "grad_norm": 0.3195907473564148,
+      "learning_rate": 1.7873748681468932e-05,
+      "loss": 0.3965,
       "step": 83860
     },
     {
-      "epoch": 2.95,
-      "learning_rate": 1.89930050931188e-05,
-      "loss": 0.2775,
+      "epoch": 3.0224889177208345,
+      "grad_norm": 0.22984401881694794,
+      "learning_rate": 1.787095166205309e-05,
+      "loss": 0.3762,
       "step": 83865
     },
     {
-      "epoch": 2.95,
-      "learning_rate": 1.8990239867883348e-05,
-      "loss": 0.2859,
+      "epoch": 3.0226691173820592,
+      "grad_norm": 0.22616103291511536,
+      "learning_rate": 1.7868154739769144e-05,
+      "loss": 0.3776,
       "step": 83870
     },
     {
-      "epoch": 2.95,
-      "learning_rate": 1.898747472068067e-05,
-      "loss": 0.2673,
+      "epoch": 3.022849317043284,
+      "grad_norm": 0.17865821719169617,
+      "learning_rate": 1.7865357914655212e-05,
+      "loss": 0.3676,
       "step": 83875
     },
     {
-      "epoch": 2.95,
-      "learning_rate": 1.8984709651546667e-05,
-      "loss": 0.2706,
+      "epoch": 3.0230295167045087,
+      "grad_norm": 0.21580396592617035,
+      "learning_rate": 1.78625611867494e-05,
+      "loss": 0.3549,
       "step": 83880
     },
     {
-      "epoch": 2.95,
-      "learning_rate": 1.8981944660517243e-05,
-      "loss": 0.27,
+      "epoch": 3.0232097163657334,
+      "grad_norm": 0.22355122864246368,
+      "learning_rate": 1.7859764556089804e-05,
+      "loss": 0.4141,
       "step": 83885
     },
     {
-      "epoch": 2.95,
-      "learning_rate": 1.89791797476283e-05,
-      "loss": 0.2736,
+      "epoch": 3.0233899160269577,
+      "grad_norm": 0.2217842936515808,
+      "learning_rate": 1.7856968022714535e-05,
+      "loss": 0.387,
       "step": 83890
     },
     {
-      "epoch": 2.95,
-      "learning_rate": 1.8976414912915748e-05,
-      "loss": 0.2677,
+      "epoch": 3.0235701156881825,
+      "grad_norm": 0.29023489356040955,
+      "learning_rate": 1.78541715866617e-05,
+      "loss": 0.3755,
       "step": 83895
     },
     {
-      "epoch": 2.95,
-      "learning_rate": 1.897365015641548e-05,
-      "loss": 0.2639,
+      "epoch": 3.023750315349407,
+      "grad_norm": 0.19948424398899078,
+      "learning_rate": 1.785137524796938e-05,
+      "loss": 0.3434,
       "step": 83900
     },
     {
-      "epoch": 2.95,
-      "learning_rate": 1.897088547816338e-05,
-      "loss": 0.2631,
+      "epoch": 3.023930515010632,
+      "grad_norm": 0.1968270093202591,
+      "learning_rate": 1.7848579006675707e-05,
+      "loss": 0.3828,
       "step": 83905
     },
     {
-      "epoch": 2.95,
-      "learning_rate": 1.896812087819537e-05,
-      "loss": 0.2578,
+      "epoch": 3.024110714671856,
+      "grad_norm": 0.20622436702251434,
+      "learning_rate": 1.784578286281874e-05,
+      "loss": 0.3828,
       "step": 83910
     },
     {
-      "epoch": 2.95,
-      "learning_rate": 1.896535635654733e-05,
-      "loss": 0.2558,
+      "epoch": 3.024290914333081,
+      "grad_norm": 0.25507232546806335,
+      "learning_rate": 1.7842986816436613e-05,
+      "loss": 0.3875,
       "step": 83915
     },
     {
-      "epoch": 2.95,
-      "learning_rate": 1.8962591913255158e-05,
-      "loss": 0.2719,
+      "epoch": 3.0244711139943057,
+      "grad_norm": 0.18650664389133453,
+      "learning_rate": 1.7840190867567403e-05,
+      "loss": 0.3826,
       "step": 83920
     },
     {
-      "epoch": 2.95,
-      "learning_rate": 1.895982754835476e-05,
-      "loss": 0.2821,
+      "epoch": 3.0246513136555304,
+      "grad_norm": 0.23238393664360046,
+      "learning_rate": 1.7837395016249198e-05,
+      "loss": 0.3914,
       "step": 83925
     },
     {
-      "epoch": 2.95,
-      "learning_rate": 1.895706326188202e-05,
-      "loss": 0.2708,
+      "epoch": 3.024831513316755,
+      "grad_norm": 0.23915240168571472,
+      "learning_rate": 1.7834599262520102e-05,
+      "loss": 0.3834,
       "step": 83930
     },
     {
-      "epoch": 2.95,
-      "learning_rate": 1.895429905387283e-05,
-      "loss": 0.282,
+      "epoch": 3.0250117129779794,
+      "grad_norm": 0.2193489819765091,
+      "learning_rate": 1.78318036064182e-05,
+      "loss": 0.4019,
       "step": 83935
     },
     {
-      "epoch": 2.95,
-      "learning_rate": 1.895153492436307e-05,
-      "loss": 0.2764,
+      "epoch": 3.025191912639204,
+      "grad_norm": 0.2331564873456955,
+      "learning_rate": 1.7829008047981594e-05,
+      "loss": 0.3714,
       "step": 83940
     },
     {
-      "epoch": 2.95,
-      "learning_rate": 1.8948770873388654e-05,
-      "loss": 0.2581,
+      "epoch": 3.025372112300429,
+      "grad_norm": 0.22607645392417908,
+      "learning_rate": 1.7826212587248364e-05,
+      "loss": 0.4046,
       "step": 83945
     },
     {
-      "epoch": 2.95,
-      "learning_rate": 1.894600690098546e-05,
-      "loss": 0.2626,
+      "epoch": 3.0255523119616536,
+      "grad_norm": 0.235866516828537,
+      "learning_rate": 1.7823417224256594e-05,
+      "loss": 0.3759,
       "step": 83950
     },
     {
-      "epoch": 2.95,
-      "learning_rate": 1.894324300718938e-05,
-      "loss": 0.2693,
+      "epoch": 3.025732511622878,
+      "grad_norm": 0.23294222354888916,
+      "learning_rate": 1.782062195904438e-05,
+      "loss": 0.3864,
       "step": 83955
     },
     {
-      "epoch": 2.95,
-      "learning_rate": 1.8940479192036286e-05,
-      "loss": 0.2833,
+      "epoch": 3.0259127112841027,
+      "grad_norm": 0.22986343502998352,
+      "learning_rate": 1.78178267916498e-05,
+      "loss": 0.3758,
       "step": 83960
     },
     {
-      "epoch": 2.95,
-      "learning_rate": 1.893771545556209e-05,
-      "loss": 0.2659,
+      "epoch": 3.0260929109453274,
+      "grad_norm": 0.17476344108581543,
+      "learning_rate": 1.7815031722110935e-05,
+      "loss": 0.3921,
       "step": 83965
     },
     {
-      "epoch": 2.95,
-      "learning_rate": 1.8934951797802654e-05,
-      "loss": 0.2795,
+      "epoch": 3.026273110606552,
+      "grad_norm": 0.23283426463603973,
+      "learning_rate": 1.7812236750465876e-05,
+      "loss": 0.4051,
       "step": 83970
     },
     {
-      "epoch": 2.95,
-      "learning_rate": 1.893218821879388e-05,
-      "loss": 0.2528,
+      "epoch": 3.026453310267777,
+      "grad_norm": 0.19704559445381165,
+      "learning_rate": 1.7809441876752695e-05,
+      "loss": 0.3977,
       "step": 83975
     },
     {
-      "epoch": 2.95,
-      "learning_rate": 1.892942471857163e-05,
-      "loss": 0.2465,
+      "epoch": 3.026633509929001,
+      "grad_norm": 0.1991409957408905,
+      "learning_rate": 1.7806647101009484e-05,
+      "loss": 0.3977,
       "step": 83980
     },
     {
-      "epoch": 2.95,
-      "learning_rate": 1.8926661297171816e-05,
-      "loss": 0.2652,
+      "epoch": 3.026813709590226,
+      "grad_norm": 0.2306070774793625,
+      "learning_rate": 1.780385242327431e-05,
+      "loss": 0.3909,
       "step": 83985
     },
     {
-      "epoch": 2.96,
-      "learning_rate": 1.8923897954630298e-05,
-      "loss": 0.2615,
+      "epoch": 3.0269939092514506,
+      "grad_norm": 0.25500649213790894,
+      "learning_rate": 1.780105784358525e-05,
+      "loss": 0.4118,
       "step": 83990
     },
     {
-      "epoch": 2.96,
-      "learning_rate": 1.8921134690982957e-05,
-      "loss": 0.2867,
+      "epoch": 3.0271741089126754,
+      "grad_norm": 0.19943642616271973,
+      "learning_rate": 1.7798263361980388e-05,
+      "loss": 0.366,
       "step": 83995
     },
     {
-      "epoch": 2.96,
-      "learning_rate": 1.8918371506265666e-05,
-      "loss": 0.2851,
+      "epoch": 3.0273543085739,
+      "grad_norm": 0.22723056375980377,
+      "learning_rate": 1.77954689784978e-05,
+      "loss": 0.4048,
       "step": 84000
     },
     {
-      "epoch": 2.96,
-      "eval_loss": 0.2586404085159302,
-      "eval_runtime": 10.5437,
-      "eval_samples_per_second": 9.484,
-      "eval_steps_per_second": 9.484,
+      "epoch": 3.0273543085739,
+      "eval_loss": 0.4311674237251282,
+      "eval_runtime": 3.5287,
+      "eval_samples_per_second": 28.339,
+      "eval_steps_per_second": 7.085,
       "step": 84000
     },
     {
-      "epoch": 2.96,
-      "learning_rate": 1.891560840051433e-05,
-      "loss": 0.2446,
+      "epoch": 3.0275345082351244,
+      "grad_norm": 0.20490515232086182,
+      "learning_rate": 1.7792674693175534e-05,
+      "loss": 0.3958,
       "step": 84005
     },
     {
-      "epoch": 2.96,
-      "learning_rate": 1.8912845373764804e-05,
-      "loss": 0.2668,
+      "epoch": 3.027714707896349,
+      "grad_norm": 0.24055179953575134,
+      "learning_rate": 1.77898805060517e-05,
+      "loss": 0.3762,
       "step": 84010
     },
     {
-      "epoch": 2.96,
-      "learning_rate": 1.8910082426052973e-05,
-      "loss": 0.2509,
+      "epoch": 3.027894907557574,
+      "grad_norm": 0.25771138072013855,
+      "learning_rate": 1.778708641716433e-05,
+      "loss": 0.4009,
       "step": 84015
     },
     {
-      "epoch": 2.96,
-      "learning_rate": 1.8907319557414696e-05,
-      "loss": 0.2543,
+      "epoch": 3.0280751072187986,
+      "grad_norm": 0.19675886631011963,
+      "learning_rate": 1.7784292426551525e-05,
+      "loss": 0.3787,
       "step": 84020
     },
     {
-      "epoch": 2.96,
-      "learning_rate": 1.8904556767885865e-05,
-      "loss": 0.2615,
+      "epoch": 3.028255306880023,
+      "grad_norm": 0.2063799649477005,
+      "learning_rate": 1.7781498534251334e-05,
+      "loss": 0.428,
       "step": 84025
     },
     {
-      "epoch": 2.96,
-      "learning_rate": 1.890179405750235e-05,
-      "loss": 0.2538,
+      "epoch": 3.0284355065412476,
+      "grad_norm": 0.23454803228378296,
+      "learning_rate": 1.7778704740301823e-05,
+      "loss": 0.3933,
       "step": 84030
     },
     {
-      "epoch": 2.96,
-      "learning_rate": 1.889903142630001e-05,
-      "loss": 0.2779,
+      "epoch": 3.0286157062024723,
+      "grad_norm": 0.20957069098949432,
+      "learning_rate": 1.7775911044741067e-05,
+      "loss": 0.3725,
       "step": 84035
     },
     {
-      "epoch": 2.96,
-      "learning_rate": 1.8896268874314743e-05,
-      "loss": 0.2719,
+      "epoch": 3.028795905863697,
+      "grad_norm": 0.24163176119327545,
+      "learning_rate": 1.7773117447607128e-05,
+      "loss": 0.3617,
       "step": 84040
     },
     {
-      "epoch": 2.96,
-      "learning_rate": 1.8893506401582397e-05,
-      "loss": 0.261,
+      "epoch": 3.028976105524922,
+      "grad_norm": 0.19776524603366852,
+      "learning_rate": 1.7770323948938056e-05,
+      "loss": 0.3595,
       "step": 84045
     },
     {
-      "epoch": 2.96,
-      "learning_rate": 1.889074400813884e-05,
-      "loss": 0.2607,
+      "epoch": 3.029156305186146,
+      "grad_norm": 0.21786345541477203,
+      "learning_rate": 1.7767530548771926e-05,
+      "loss": 0.4257,
       "step": 84050
     },
     {
-      "epoch": 2.96,
-      "learning_rate": 1.8887981694019946e-05,
-      "loss": 0.2538,
+      "epoch": 3.029336504847371,
+      "grad_norm": 0.2272718995809555,
+      "learning_rate": 1.7764737247146783e-05,
+      "loss": 0.3771,
       "step": 84055
     },
     {
-      "epoch": 2.96,
-      "learning_rate": 1.8885219459261588e-05,
-      "loss": 0.2449,
+      "epoch": 3.0295167045085956,
+      "grad_norm": 0.2050115466117859,
+      "learning_rate": 1.7761944044100707e-05,
+      "loss": 0.3718,
       "step": 84060
     },
     {
-      "epoch": 2.96,
-      "learning_rate": 1.8882457303899626e-05,
-      "loss": 0.2915,
+      "epoch": 3.0296969041698203,
+      "grad_norm": 0.26552021503448486,
+      "learning_rate": 1.7759150939671735e-05,
+      "loss": 0.3983,
       "step": 84065
     },
     {
-      "epoch": 2.96,
-      "learning_rate": 1.8879695227969923e-05,
-      "loss": 0.2925,
+      "epoch": 3.0298771038310446,
+      "grad_norm": 0.22904445230960846,
+      "learning_rate": 1.775635793389793e-05,
+      "loss": 0.377,
       "step": 84070
     },
     {
-      "epoch": 2.96,
-      "learning_rate": 1.8876933231508337e-05,
-      "loss": 0.2451,
+      "epoch": 3.0300573034922693,
+      "grad_norm": 0.1924734264612198,
+      "learning_rate": 1.7753565026817348e-05,
+      "loss": 0.3702,
       "step": 84075
     },
     {
-      "epoch": 2.96,
-      "learning_rate": 1.887417131455074e-05,
-      "loss": 0.272,
+      "epoch": 3.030237503153494,
+      "grad_norm": 0.2412857860326767,
+      "learning_rate": 1.775077221846805e-05,
+      "loss": 0.3691,
       "step": 84080
     },
     {
-      "epoch": 2.96,
-      "learning_rate": 1.8871409477132997e-05,
-      "loss": 0.265,
+      "epoch": 3.030417702814719,
+      "grad_norm": 0.2012851983308792,
+      "learning_rate": 1.7747979508888053e-05,
+      "loss": 0.3935,
       "step": 84085
     },
     {
-      "epoch": 2.96,
-      "learning_rate": 1.8868647719290966e-05,
-      "loss": 0.2835,
+      "epoch": 3.0305979024759435,
+      "grad_norm": 0.21327096223831177,
+      "learning_rate": 1.774518689811545e-05,
+      "loss": 0.3731,
       "step": 84090
     },
     {
-      "epoch": 2.96,
-      "learning_rate": 1.886588604106049e-05,
-      "loss": 0.2811,
+      "epoch": 3.030778102137168,
+      "grad_norm": 0.19363613426685333,
+      "learning_rate": 1.7742394386188255e-05,
+      "loss": 0.3753,
       "step": 84095
     },
     {
-      "epoch": 2.96,
-      "learning_rate": 1.8863124442477453e-05,
-      "loss": 0.2649,
+      "epoch": 3.0309583017983925,
+      "grad_norm": 0.19046692550182343,
+      "learning_rate": 1.7739601973144547e-05,
+      "loss": 0.3963,
       "step": 84100
     },
     {
-      "epoch": 2.96,
-      "learning_rate": 1.8860362923577694e-05,
-      "loss": 0.2909,
+      "epoch": 3.0311385014596173,
+      "grad_norm": 0.2194364368915558,
+      "learning_rate": 1.7736809659022354e-05,
+      "loss": 0.374,
       "step": 84105
     },
     {
-      "epoch": 2.96,
-      "learning_rate": 1.8857601484397082e-05,
-      "loss": 0.2592,
+      "epoch": 3.031318701120842,
+      "grad_norm": 0.1966681331396103,
+      "learning_rate": 1.7734017443859715e-05,
+      "loss": 0.3419,
       "step": 84110
     },
     {
-      "epoch": 2.96,
-      "learning_rate": 1.885484012497146e-05,
-      "loss": 0.2774,
+      "epoch": 3.0314989007820667,
+      "grad_norm": 0.2259567826986313,
+      "learning_rate": 1.7731225327694688e-05,
+      "loss": 0.3931,
       "step": 84115
     },
     {
-      "epoch": 2.96,
-      "learning_rate": 1.8852078845336694e-05,
-      "loss": 0.2713,
+      "epoch": 3.031679100443291,
+      "grad_norm": 0.21452829241752625,
+      "learning_rate": 1.7728433310565302e-05,
+      "loss": 0.3811,
       "step": 84120
     },
     {
-      "epoch": 2.96,
-      "learning_rate": 1.8849317645528636e-05,
-      "loss": 0.2646,
+      "epoch": 3.0318593001045158,
+      "grad_norm": 0.20977729558944702,
+      "learning_rate": 1.7725641392509614e-05,
+      "loss": 0.3805,
       "step": 84125
     },
     {
-      "epoch": 2.96,
-      "learning_rate": 1.8846556525583127e-05,
-      "loss": 0.2542,
+      "epoch": 3.0320394997657405,
+      "grad_norm": 0.2387772798538208,
+      "learning_rate": 1.772284957356565e-05,
+      "loss": 0.3516,
       "step": 84130
     },
     {
-      "epoch": 2.96,
-      "learning_rate": 1.8843795485536027e-05,
-      "loss": 0.2764,
+      "epoch": 3.0322196994269652,
+      "grad_norm": 0.21212005615234375,
+      "learning_rate": 1.772005785377145e-05,
+      "loss": 0.3751,
       "step": 84135
     },
     {
-      "epoch": 2.96,
-      "learning_rate": 1.884103452542319e-05,
-      "loss": 0.2772,
+      "epoch": 3.0323998990881895,
+      "grad_norm": 0.22513705492019653,
+      "learning_rate": 1.7717266233165054e-05,
+      "loss": 0.3696,
       "step": 84140
     },
     {
-      "epoch": 2.96,
-      "learning_rate": 1.883827364528046e-05,
-      "loss": 0.281,
+      "epoch": 3.0325800987494143,
+      "grad_norm": 0.25777748227119446,
+      "learning_rate": 1.7714474711784496e-05,
+      "loss": 0.3954,
       "step": 84145
     },
     {
-      "epoch": 2.96,
-      "learning_rate": 1.883551284514368e-05,
-      "loss": 0.2541,
+      "epoch": 3.032760298410639,
+      "grad_norm": 0.20755411684513092,
+      "learning_rate": 1.77116832896678e-05,
+      "loss": 0.3365,
       "step": 84150
     },
     {
-      "epoch": 2.96,
-      "learning_rate": 1.8832752125048713e-05,
-      "loss": 0.2678,
+      "epoch": 3.0329404980718637,
+      "grad_norm": 0.18738418817520142,
+      "learning_rate": 1.770889196685302e-05,
+      "loss": 0.3949,
       "step": 84155
     },
     {
-      "epoch": 2.96,
-      "learning_rate": 1.8829991485031387e-05,
-      "loss": 0.2587,
+      "epoch": 3.0331206977330885,
+      "grad_norm": 0.21156027913093567,
+      "learning_rate": 1.7706100743378168e-05,
+      "loss": 0.3566,
       "step": 84160
     },
     {
-      "epoch": 2.96,
-      "learning_rate": 1.8827230925127566e-05,
-      "loss": 0.2556,
+      "epoch": 3.0333008973943127,
+      "grad_norm": 0.28246161341667175,
+      "learning_rate": 1.770330961928129e-05,
+      "loss": 0.3625,
       "step": 84165
     },
     {
-      "epoch": 2.96,
-      "learning_rate": 1.8824470445373065e-05,
-      "loss": 0.2801,
+      "epoch": 3.0334810970555375,
+      "grad_norm": 0.22130875289440155,
+      "learning_rate": 1.77005185946004e-05,
+      "loss": 0.3402,
       "step": 84170
     },
     {
-      "epoch": 2.96,
-      "learning_rate": 1.882171004580376e-05,
-      "loss": 0.2753,
+      "epoch": 3.033661296716762,
+      "grad_norm": 0.22155120968818665,
+      "learning_rate": 1.7697727669373525e-05,
+      "loss": 0.375,
       "step": 84175
     },
     {
-      "epoch": 2.96,
-      "learning_rate": 1.881894972645548e-05,
-      "loss": 0.2691,
+      "epoch": 3.033841496377987,
+      "grad_norm": 0.22286799550056458,
+      "learning_rate": 1.7694936843638707e-05,
+      "loss": 0.3858,
       "step": 84180
     },
     {
-      "epoch": 2.96,
-      "learning_rate": 1.8816189487364056e-05,
-      "loss": 0.2618,
+      "epoch": 3.0340216960392112,
+      "grad_norm": 0.1948540061712265,
+      "learning_rate": 1.7692146117433965e-05,
+      "loss": 0.3905,
       "step": 84185
     },
     {
-      "epoch": 2.96,
-      "learning_rate": 1.8813429328565342e-05,
-      "loss": 0.244,
+      "epoch": 3.034201895700436,
+      "grad_norm": 0.228069469332695,
+      "learning_rate": 1.76893554907973e-05,
+      "loss": 0.41,
       "step": 84190
     },
     {
-      "epoch": 2.96,
-      "learning_rate": 1.8810669250095163e-05,
-      "loss": 0.2633,
+      "epoch": 3.0343820953616607,
+      "grad_norm": 0.21206477284431458,
+      "learning_rate": 1.7686564963766765e-05,
+      "loss": 0.3649,
       "step": 84195
     },
     {
-      "epoch": 2.96,
-      "learning_rate": 1.8807909251989376e-05,
-      "loss": 0.2525,
+      "epoch": 3.0345622950228854,
+      "grad_norm": 0.2614726722240448,
+      "learning_rate": 1.768377453638035e-05,
+      "loss": 0.377,
       "step": 84200
     },
     {
-      "epoch": 2.96,
-      "learning_rate": 1.8805149334283807e-05,
-      "loss": 0.2616,
+      "epoch": 3.03474249468411,
+      "grad_norm": 0.21512839198112488,
+      "learning_rate": 1.768098420867611e-05,
+      "loss": 0.3931,
       "step": 84205
     },
     {
-      "epoch": 2.96,
-      "learning_rate": 1.880238949701428e-05,
-      "loss": 0.2629,
+      "epoch": 3.0349226943453345,
+      "grad_norm": 0.22823360562324524,
+      "learning_rate": 1.7678193980692036e-05,
+      "loss": 0.368,
       "step": 84210
     },
     {
-      "epoch": 2.96,
-      "learning_rate": 1.8799629740216643e-05,
-      "loss": 0.2652,
+      "epoch": 3.035102894006559,
+      "grad_norm": 0.2092510610818863,
+      "learning_rate": 1.767540385246615e-05,
+      "loss": 0.3904,
       "step": 84215
     },
     {
-      "epoch": 2.96,
-      "learning_rate": 1.8796870063926733e-05,
-      "loss": 0.2691,
+      "epoch": 3.035283093667784,
+      "grad_norm": 0.22960545122623444,
+      "learning_rate": 1.767261382403647e-05,
+      "loss": 0.4228,
       "step": 84220
     },
     {
-      "epoch": 2.96,
-      "learning_rate": 1.8794110468180383e-05,
-      "loss": 0.2754,
+      "epoch": 3.0354632933290087,
+      "grad_norm": 0.2698182761669159,
+      "learning_rate": 1.7669823895441007e-05,
+      "loss": 0.3652,
       "step": 84225
     },
     {
-      "epoch": 2.96,
-      "learning_rate": 1.8791350953013402e-05,
-      "loss": 0.281,
+      "epoch": 3.0356434929902334,
+      "grad_norm": 0.27109652757644653,
+      "learning_rate": 1.7667034066717768e-05,
+      "loss": 0.3765,
       "step": 84230
     },
     {
-      "epoch": 2.96,
-      "learning_rate": 1.8788591518461647e-05,
-      "loss": 0.2688,
+      "epoch": 3.0358236926514577,
+      "grad_norm": 0.178133025765419,
+      "learning_rate": 1.766424433790478e-05,
+      "loss": 0.3916,
       "step": 84235
     },
     {
-      "epoch": 2.96,
-      "learning_rate": 1.8785832164560937e-05,
-      "loss": 0.2692,
+      "epoch": 3.0360038923126824,
+      "grad_norm": 0.26019546389579773,
+      "learning_rate": 1.7661454709040036e-05,
+      "loss": 0.4084,
       "step": 84240
     },
     {
-      "epoch": 2.96,
-      "learning_rate": 1.8783072891347105e-05,
-      "loss": 0.2761,
+      "epoch": 3.036184091973907,
+      "grad_norm": 0.21639469265937805,
+      "learning_rate": 1.7658665180161555e-05,
+      "loss": 0.3909,
       "step": 84245
     },
     {
-      "epoch": 2.96,
-      "learning_rate": 1.8780313698855964e-05,
-      "loss": 0.2574,
+      "epoch": 3.036364291635132,
+      "grad_norm": 0.2332758754491806,
+      "learning_rate": 1.7655875751307338e-05,
+      "loss": 0.3712,
       "step": 84250
     },
     {
-      "epoch": 2.96,
-      "learning_rate": 1.877755458712336e-05,
-      "loss": 0.2711,
+      "epoch": 3.036544491296356,
+      "grad_norm": 0.19095510244369507,
+      "learning_rate": 1.765308642251539e-05,
+      "loss": 0.3552,
       "step": 84255
     },
     {
-      "epoch": 2.96,
-      "learning_rate": 1.877479555618511e-05,
-      "loss": 0.2504,
+      "epoch": 3.036724690957581,
+      "grad_norm": 0.18508897721767426,
+      "learning_rate": 1.765029719382372e-05,
+      "loss": 0.4014,
       "step": 84260
     },
     {
-      "epoch": 2.96,
-      "learning_rate": 1.8772036606077025e-05,
-      "loss": 0.2722,
+      "epoch": 3.0369048906188056,
+      "grad_norm": 0.22159694135189056,
+      "learning_rate": 1.7647508065270335e-05,
+      "loss": 0.3882,
       "step": 84265
     },
     {
-      "epoch": 2.96,
-      "learning_rate": 1.8769277736834954e-05,
-      "loss": 0.2722,
+      "epoch": 3.0370850902800304,
+      "grad_norm": 0.22813962399959564,
+      "learning_rate": 1.764471903689321e-05,
+      "loss": 0.3698,
       "step": 84270
     },
     {
-      "epoch": 2.97,
-      "learning_rate": 1.87665189484947e-05,
-      "loss": 0.2855,
+      "epoch": 3.037265289941255,
+      "grad_norm": 0.21553412079811096,
+      "learning_rate": 1.7641930108730377e-05,
+      "loss": 0.3682,
       "step": 84275
     },
     {
-      "epoch": 2.97,
-      "learning_rate": 1.876376024109209e-05,
-      "loss": 0.2642,
+      "epoch": 3.0374454896024794,
+      "grad_norm": 0.2294461727142334,
+      "learning_rate": 1.7639141280819815e-05,
+      "loss": 0.3852,
       "step": 84280
     },
     {
-      "epoch": 2.97,
-      "learning_rate": 1.8761001614662938e-05,
-      "loss": 0.2801,
+      "epoch": 3.037625689263704,
+      "grad_norm": 0.240443617105484,
+      "learning_rate": 1.7636352553199537e-05,
+      "loss": 0.3859,
       "step": 84285
     },
     {
-      "epoch": 2.97,
-      "learning_rate": 1.8758243069243077e-05,
-      "loss": 0.2736,
+      "epoch": 3.037805888924929,
+      "grad_norm": 0.21389326453208923,
+      "learning_rate": 1.7633563925907533e-05,
+      "loss": 0.36,
       "step": 84290
     },
     {
-      "epoch": 2.97,
-      "learning_rate": 1.8755484604868315e-05,
-      "loss": 0.2919,
+      "epoch": 3.0379860885861536,
+      "grad_norm": 0.21817508339881897,
+      "learning_rate": 1.7630775398981777e-05,
+      "loss": 0.4133,
       "step": 84295
     },
     {
-      "epoch": 2.97,
-      "learning_rate": 1.8752726221574466e-05,
-      "loss": 0.2704,
+      "epoch": 3.038166288247378,
+      "grad_norm": 0.2270621359348297,
+      "learning_rate": 1.7627986972460298e-05,
+      "loss": 0.4165,
       "step": 84300
     },
     {
-      "epoch": 2.97,
-      "learning_rate": 1.8749967919397342e-05,
-      "loss": 0.2514,
+      "epoch": 3.0383464879086026,
+      "grad_norm": 0.24196819961071014,
+      "learning_rate": 1.762519864638106e-05,
+      "loss": 0.393,
       "step": 84305
     },
     {
-      "epoch": 2.97,
-      "learning_rate": 1.874720969837278e-05,
-      "loss": 0.2576,
+      "epoch": 3.0385266875698274,
+      "grad_norm": 0.2399672567844391,
+      "learning_rate": 1.7622410420782058e-05,
+      "loss": 0.4056,
       "step": 84310
     },
     {
-      "epoch": 2.97,
-      "learning_rate": 1.874445155853658e-05,
-      "loss": 0.2626,
+      "epoch": 3.038706887231052,
+      "grad_norm": 0.23226651549339294,
+      "learning_rate": 1.7619622295701296e-05,
+      "loss": 0.3552,
       "step": 84315
     },
     {
-      "epoch": 2.97,
-      "learning_rate": 1.8741693499924546e-05,
-      "loss": 0.269,
+      "epoch": 3.038887086892277,
+      "grad_norm": 0.2692866027355194,
+      "learning_rate": 1.761683427117674e-05,
+      "loss": 0.3849,
       "step": 84320
     },
     {
-      "epoch": 2.97,
-      "learning_rate": 1.8738935522572498e-05,
-      "loss": 0.2554,
+      "epoch": 3.039067286553501,
+      "grad_norm": 0.2093464732170105,
+      "learning_rate": 1.76140463472464e-05,
+      "loss": 0.3731,
       "step": 84325
     },
     {
-      "epoch": 2.97,
-      "learning_rate": 1.8736177626516246e-05,
-      "loss": 0.266,
+      "epoch": 3.039247486214726,
+      "grad_norm": 0.227401465177536,
+      "learning_rate": 1.7611258523948242e-05,
+      "loss": 0.3783,
       "step": 84330
     },
     {
-      "epoch": 2.97,
-      "learning_rate": 1.8733419811791604e-05,
-      "loss": 0.2702,
+      "epoch": 3.0394276858759506,
+      "grad_norm": 0.212887704372406,
+      "learning_rate": 1.7608470801320253e-05,
+      "loss": 0.354,
       "step": 84335
     },
     {
-      "epoch": 2.97,
-      "learning_rate": 1.873066207843438e-05,
-      "loss": 0.2851,
+      "epoch": 3.0396078855371753,
+      "grad_norm": 0.2043050229549408,
+      "learning_rate": 1.7605683179400423e-05,
+      "loss": 0.3788,
       "step": 84340
     },
     {
-      "epoch": 2.97,
-      "learning_rate": 1.8727904426480362e-05,
-      "loss": 0.2812,
+      "epoch": 3.0397880851984,
+      "grad_norm": 0.24479785561561584,
+      "learning_rate": 1.7602895658226725e-05,
+      "loss": 0.3753,
       "step": 84345
     },
     {
-      "epoch": 2.97,
-      "learning_rate": 1.8725146855965385e-05,
-      "loss": 0.2573,
+      "epoch": 3.0399682848596243,
+      "grad_norm": 0.2218136191368103,
+      "learning_rate": 1.7600108237837144e-05,
+      "loss": 0.4084,
       "step": 84350
     },
     {
-      "epoch": 2.97,
-      "learning_rate": 1.8722389366925232e-05,
-      "loss": 0.2504,
+      "epoch": 3.040148484520849,
+      "grad_norm": 0.23123931884765625,
+      "learning_rate": 1.7597320918269658e-05,
+      "loss": 0.4059,
       "step": 84355
     },
     {
-      "epoch": 2.97,
-      "learning_rate": 1.871963195939572e-05,
-      "loss": 0.2567,
+      "epoch": 3.040328684182074,
+      "grad_norm": 0.2367209494113922,
+      "learning_rate": 1.7594533699562234e-05,
+      "loss": 0.3901,
       "step": 84360
     },
     {
-      "epoch": 2.97,
-      "learning_rate": 1.8716874633412647e-05,
-      "loss": 0.2856,
+      "epoch": 3.0405088838432985,
+      "grad_norm": 0.22968047857284546,
+      "learning_rate": 1.759174658175286e-05,
+      "loss": 0.3823,
       "step": 84365
     },
     {
-      "epoch": 2.97,
-      "learning_rate": 1.871411738901182e-05,
-      "loss": 0.2759,
+      "epoch": 3.040689083504523,
+      "grad_norm": 0.23370803892612457,
+      "learning_rate": 1.758895956487951e-05,
+      "loss": 0.384,
       "step": 84370
     },
     {
-      "epoch": 2.97,
-      "learning_rate": 1.8711360226229035e-05,
-      "loss": 0.2461,
+      "epoch": 3.0408692831657476,
+      "grad_norm": 0.19679541885852814,
+      "learning_rate": 1.758617264898014e-05,
+      "loss": 0.3785,
       "step": 84375
     },
     {
-      "epoch": 2.97,
-      "learning_rate": 1.8708603145100085e-05,
-      "loss": 0.2644,
+      "epoch": 3.0410494828269723,
+      "grad_norm": 0.2017953097820282,
+      "learning_rate": 1.758338583409274e-05,
+      "loss": 0.3986,
       "step": 84380
     },
     {
-      "epoch": 2.97,
-      "learning_rate": 1.8705846145660782e-05,
-      "loss": 0.2525,
+      "epoch": 3.041229682488197,
+      "grad_norm": 0.1906876564025879,
+      "learning_rate": 1.7580599120255263e-05,
+      "loss": 0.3619,
       "step": 84385
     },
     {
-      "epoch": 2.97,
-      "learning_rate": 1.8703089227946924e-05,
-      "loss": 0.274,
+      "epoch": 3.0414098821494218,
+      "grad_norm": 0.1913730502128601,
+      "learning_rate": 1.75778125075057e-05,
+      "loss": 0.3926,
       "step": 84390
     },
     {
-      "epoch": 2.97,
-      "learning_rate": 1.87003323919943e-05,
-      "loss": 0.2611,
+      "epoch": 3.041590081810646,
+      "grad_norm": 0.24644358456134796,
+      "learning_rate": 1.7575025995881998e-05,
+      "loss": 0.3757,
       "step": 84395
     },
     {
-      "epoch": 2.97,
-      "learning_rate": 1.8697575637838703e-05,
-      "loss": 0.2785,
+      "epoch": 3.041770281471871,
+      "grad_norm": 0.2385040670633316,
+      "learning_rate": 1.7572239585422116e-05,
+      "loss": 0.4107,
       "step": 84400
     },
     {
-      "epoch": 2.97,
-      "learning_rate": 1.8694818965515938e-05,
-      "loss": 0.2823,
+      "epoch": 3.0419504811330955,
+      "grad_norm": 0.18101677298545837,
+      "learning_rate": 1.7569453276164054e-05,
+      "loss": 0.387,
       "step": 84405
     },
     {
-      "epoch": 2.97,
-      "learning_rate": 1.869206237506179e-05,
-      "loss": 0.2583,
+      "epoch": 3.0421306807943203,
+      "grad_norm": 0.21769458055496216,
+      "learning_rate": 1.756666706814574e-05,
+      "loss": 0.4045,
       "step": 84410
     },
     {
-      "epoch": 2.97,
-      "learning_rate": 1.8689305866512063e-05,
-      "loss": 0.2773,
+      "epoch": 3.0423108804555445,
+      "grad_norm": 0.2499932199716568,
+      "learning_rate": 1.7563880961405148e-05,
+      "loss": 0.3898,
       "step": 84415
     },
     {
-      "epoch": 2.97,
-      "learning_rate": 1.8686549439902528e-05,
-      "loss": 0.2528,
+      "epoch": 3.0424910801167693,
+      "grad_norm": 0.24783965945243835,
+      "learning_rate": 1.7561094955980247e-05,
+      "loss": 0.3607,
       "step": 84420
     },
     {
-      "epoch": 2.97,
-      "learning_rate": 1.8683793095268997e-05,
-      "loss": 0.2625,
+      "epoch": 3.042671279777994,
+      "grad_norm": 0.21402627229690552,
+      "learning_rate": 1.7558309051908976e-05,
+      "loss": 0.3521,
       "step": 84425
     },
     {
-      "epoch": 2.97,
-      "learning_rate": 1.8681036832647253e-05,
-      "loss": 0.274,
+      "epoch": 3.0428514794392187,
+      "grad_norm": 0.21035562455654144,
+      "learning_rate": 1.7555523249229312e-05,
+      "loss": 0.4272,
       "step": 84430
     },
     {
-      "epoch": 2.97,
-      "learning_rate": 1.8678280652073074e-05,
-      "loss": 0.2494,
+      "epoch": 3.0430316791004435,
+      "grad_norm": 0.2314184308052063,
+      "learning_rate": 1.7552737547979197e-05,
+      "loss": 0.3946,
       "step": 84435
     },
     {
-      "epoch": 2.97,
-      "learning_rate": 1.867552455358225e-05,
-      "loss": 0.272,
+      "epoch": 3.0432118787616678,
+      "grad_norm": 0.18554693460464478,
+      "learning_rate": 1.754995194819659e-05,
+      "loss": 0.3658,
       "step": 84440
     },
     {
-      "epoch": 2.97,
-      "learning_rate": 1.8672768537210583e-05,
-      "loss": 0.2845,
+      "epoch": 3.0433920784228925,
+      "grad_norm": 0.20421580970287323,
+      "learning_rate": 1.754716644991945e-05,
+      "loss": 0.3572,
       "step": 84445
     },
     {
-      "epoch": 2.97,
-      "learning_rate": 1.8670012602993843e-05,
-      "loss": 0.2485,
+      "epoch": 3.0435722780841172,
+      "grad_norm": 0.29508501291275024,
+      "learning_rate": 1.7544381053185723e-05,
+      "loss": 0.4097,
       "step": 84450
     },
     {
-      "epoch": 2.97,
-      "learning_rate": 1.866725675096782e-05,
-      "loss": 0.267,
+      "epoch": 3.043752477745342,
+      "grad_norm": 0.24256087839603424,
+      "learning_rate": 1.7541595758033357e-05,
+      "loss": 0.3888,
       "step": 84455
     },
     {
-      "epoch": 2.97,
-      "learning_rate": 1.8664500981168286e-05,
-      "loss": 0.3005,
+      "epoch": 3.0439326774065663,
+      "grad_norm": 0.17925018072128296,
+      "learning_rate": 1.753881056450031e-05,
+      "loss": 0.3381,
       "step": 84460
     },
     {
-      "epoch": 2.97,
-      "learning_rate": 1.8661745293631036e-05,
-      "loss": 0.2624,
+      "epoch": 3.044112877067791,
+      "grad_norm": 0.2275160253047943,
+      "learning_rate": 1.7536025472624518e-05,
+      "loss": 0.4123,
       "step": 84465
     },
     {
-      "epoch": 2.97,
-      "learning_rate": 1.865898968839185e-05,
-      "loss": 0.255,
+      "epoch": 3.0442930767290157,
+      "grad_norm": 0.2509187161922455,
+      "learning_rate": 1.7533240482443942e-05,
+      "loss": 0.3866,
       "step": 84470
     },
     {
-      "epoch": 2.97,
-      "learning_rate": 1.8656234165486506e-05,
-      "loss": 0.27,
+      "epoch": 3.0444732763902405,
+      "grad_norm": 0.231712207198143,
+      "learning_rate": 1.753045559399652e-05,
+      "loss": 0.4006,
       "step": 84475
     },
     {
-      "epoch": 2.97,
-      "learning_rate": 1.865347872495077e-05,
-      "loss": 0.2593,
+      "epoch": 3.044653476051465,
+      "grad_norm": 0.23315437138080597,
+      "learning_rate": 1.7527670807320183e-05,
+      "loss": 0.4009,
       "step": 84480
     },
     {
-      "epoch": 2.97,
-      "learning_rate": 1.865072336682044e-05,
-      "loss": 0.2904,
+      "epoch": 3.0448336757126895,
+      "grad_norm": 0.24659566581249237,
+      "learning_rate": 1.7524886122452897e-05,
+      "loss": 0.4056,
       "step": 84485
     },
     {
-      "epoch": 2.97,
-      "learning_rate": 1.8647968091131273e-05,
-      "loss": 0.2726,
+      "epoch": 3.045013875373914,
+      "grad_norm": 0.22627463936805725,
+      "learning_rate": 1.7522101539432582e-05,
+      "loss": 0.3803,
       "step": 84490
     },
     {
-      "epoch": 2.97,
-      "learning_rate": 1.8645212897919067e-05,
-      "loss": 0.261,
+      "epoch": 3.045194075035139,
+      "grad_norm": 0.2701631486415863,
+      "learning_rate": 1.7519317058297188e-05,
+      "loss": 0.3936,
       "step": 84495
     },
     {
-      "epoch": 2.97,
-      "learning_rate": 1.864245778721957e-05,
-      "loss": 0.2694,
+      "epoch": 3.0453742746963637,
+      "grad_norm": 0.23009376227855682,
+      "learning_rate": 1.7516532679084652e-05,
+      "loss": 0.3687,
       "step": 84500
     },
     {
-      "epoch": 2.97,
-      "eval_loss": 0.2578907310962677,
-      "eval_runtime": 10.553,
-      "eval_samples_per_second": 9.476,
-      "eval_steps_per_second": 9.476,
+      "epoch": 3.0453742746963637,
+      "eval_loss": 0.43199047446250916,
+      "eval_runtime": 3.5354,
+      "eval_samples_per_second": 28.286,
+      "eval_steps_per_second": 7.071,
       "step": 84500
     },
     {
-      "epoch": 2.97,
-      "learning_rate": 1.8639702759068578e-05,
-      "loss": 0.2505,
+      "epoch": 3.0455544743575884,
+      "grad_norm": 0.24204762279987335,
+      "learning_rate": 1.7513748401832904e-05,
+      "loss": 0.3531,
       "step": 84505
     },
     {
-      "epoch": 2.97,
-      "learning_rate": 1.8636947813501853e-05,
-      "loss": 0.2453,
+      "epoch": 3.0457346740188127,
+      "grad_norm": 0.27944275736808777,
+      "learning_rate": 1.751096422657989e-05,
+      "loss": 0.3562,
       "step": 84510
     },
     {
-      "epoch": 2.97,
-      "learning_rate": 1.8634192950555163e-05,
-      "loss": 0.2729,
+      "epoch": 3.0459148736800374,
+      "grad_norm": 0.2434234917163849,
+      "learning_rate": 1.750818015336354e-05,
+      "loss": 0.406,
       "step": 84515
     },
     {
-      "epoch": 2.97,
-      "learning_rate": 1.863143817026428e-05,
-      "loss": 0.2606,
+      "epoch": 3.046095073341262,
+      "grad_norm": 0.18763190507888794,
+      "learning_rate": 1.7505396182221777e-05,
+      "loss": 0.3698,
       "step": 84520
     },
     {
-      "epoch": 2.97,
-      "learning_rate": 1.8628683472664988e-05,
-      "loss": 0.2258,
+      "epoch": 3.046275273002487,
+      "grad_norm": 0.21916723251342773,
+      "learning_rate": 1.750261231319255e-05,
+      "loss": 0.4321,
       "step": 84525
     },
     {
-      "epoch": 2.97,
-      "learning_rate": 1.8625928857793034e-05,
-      "loss": 0.2844,
+      "epoch": 3.046455472663711,
+      "grad_norm": 0.22559642791748047,
+      "learning_rate": 1.749982854631377e-05,
+      "loss": 0.3608,
       "step": 84530
     },
     {
-      "epoch": 2.97,
-      "learning_rate": 1.8623725225483194e-05,
-      "loss": 0.2798,
+      "epoch": 3.046635672324936,
+      "grad_norm": 0.20946092903614044,
+      "learning_rate": 1.749704488162338e-05,
+      "loss": 0.3531,
       "step": 84535
     },
     {
-      "epoch": 2.97,
-      "learning_rate": 1.8620970759610597e-05,
-      "loss": 0.2855,
+      "epoch": 3.0468158719861607,
+      "grad_norm": 0.2142314463853836,
+      "learning_rate": 1.74942613191593e-05,
+      "loss": 0.3761,
       "step": 84540
     },
     {
-      "epoch": 2.97,
-      "learning_rate": 1.861821637656549e-05,
-      "loss": 0.2655,
+      "epoch": 3.0469960716473854,
+      "grad_norm": 0.18263986706733704,
+      "learning_rate": 1.7491477858959453e-05,
+      "loss": 0.3924,
       "step": 84545
     },
     {
-      "epoch": 2.97,
-      "learning_rate": 1.8615462076383633e-05,
-      "loss": 0.279,
+      "epoch": 3.04717627130861,
+      "grad_norm": 0.19309696555137634,
+      "learning_rate": 1.748869450106177e-05,
+      "loss": 0.3453,
       "step": 84550
     },
     {
-      "epoch": 2.97,
-      "learning_rate": 1.8612707859100796e-05,
-      "loss": 0.2723,
+      "epoch": 3.0473564709698344,
+      "grad_norm": 0.2552222013473511,
+      "learning_rate": 1.7485911245504173e-05,
+      "loss": 0.3855,
       "step": 84555
     },
     {
-      "epoch": 2.98,
-      "learning_rate": 1.8609953724752745e-05,
-      "loss": 0.2618,
+      "epoch": 3.047536670631059,
+      "grad_norm": 0.2853567600250244,
+      "learning_rate": 1.7483128092324567e-05,
+      "loss": 0.4424,
       "step": 84560
     },
     {
-      "epoch": 2.98,
-      "learning_rate": 1.860719967337523e-05,
-      "loss": 0.2769,
+      "epoch": 3.047716870292284,
+      "grad_norm": 0.2120617926120758,
+      "learning_rate": 1.74803450415609e-05,
+      "loss": 0.3929,
       "step": 84565
     },
     {
-      "epoch": 2.98,
-      "learning_rate": 1.860444570500401e-05,
-      "loss": 0.276,
+      "epoch": 3.0478970699535086,
+      "grad_norm": 0.2228347659111023,
+      "learning_rate": 1.7477562093251066e-05,
+      "loss": 0.3744,
       "step": 84570
     },
     {
-      "epoch": 2.98,
-      "learning_rate": 1.8601691819674847e-05,
-      "loss": 0.2715,
+      "epoch": 3.048077269614733,
+      "grad_norm": 0.2142227590084076,
+      "learning_rate": 1.7474779247432998e-05,
+      "loss": 0.4072,
       "step": 84575
     },
     {
-      "epoch": 2.98,
-      "learning_rate": 1.8598938017423505e-05,
-      "loss": 0.2972,
+      "epoch": 3.0482574692759576,
+      "grad_norm": 0.23196277022361755,
+      "learning_rate": 1.7471996504144612e-05,
+      "loss": 0.3609,
       "step": 84580
     },
     {
-      "epoch": 2.98,
-      "learning_rate": 1.8596184298285725e-05,
-      "loss": 0.2717,
+      "epoch": 3.0484376689371824,
+      "grad_norm": 0.20943579077720642,
+      "learning_rate": 1.74692138634238e-05,
+      "loss": 0.3989,
       "step": 84585
     },
     {
-      "epoch": 2.98,
-      "learning_rate": 1.859343066229728e-05,
-      "loss": 0.2735,
+      "epoch": 3.048617868598407,
+      "grad_norm": 0.28706827759742737,
+      "learning_rate": 1.7466431325308507e-05,
+      "loss": 0.3669,
       "step": 84590
     },
     {
-      "epoch": 2.98,
-      "learning_rate": 1.8590677109493918e-05,
-      "loss": 0.2694,
+      "epoch": 3.048798068259632,
+      "grad_norm": 0.20854423940181732,
+      "learning_rate": 1.7463648889836627e-05,
+      "loss": 0.3551,
       "step": 84595
     },
     {
-      "epoch": 2.98,
-      "learning_rate": 1.858792363991139e-05,
-      "loss": 0.2491,
+      "epoch": 3.048978267920856,
+      "grad_norm": 0.20572051405906677,
+      "learning_rate": 1.746086655704606e-05,
+      "loss": 0.3609,
       "step": 84600
     },
     {
-      "epoch": 2.98,
-      "learning_rate": 1.8585170253585436e-05,
-      "loss": 0.2665,
+      "epoch": 3.049158467582081,
+      "grad_norm": 0.1955312341451645,
+      "learning_rate": 1.7458084326974732e-05,
+      "loss": 0.3764,
       "step": 84605
     },
     {
-      "epoch": 2.98,
-      "learning_rate": 1.8582416950551823e-05,
-      "loss": 0.2561,
+      "epoch": 3.0493386672433056,
+      "grad_norm": 0.1977349817752838,
+      "learning_rate": 1.7455302199660544e-05,
+      "loss": 0.3836,
       "step": 84610
     },
     {
-      "epoch": 2.98,
-      "learning_rate": 1.8579663730846305e-05,
-      "loss": 0.2626,
+      "epoch": 3.0495188669045303,
+      "grad_norm": 0.2958627939224243,
+      "learning_rate": 1.7452520175141406e-05,
+      "loss": 0.4121,
       "step": 84615
     },
     {
-      "epoch": 2.98,
-      "learning_rate": 1.857691059450462e-05,
-      "loss": 0.2642,
+      "epoch": 3.049699066565755,
+      "grad_norm": 0.20005503296852112,
+      "learning_rate": 1.7449738253455223e-05,
+      "loss": 0.3998,
       "step": 84620
     },
     {
-      "epoch": 2.98,
-      "learning_rate": 1.857415754156251e-05,
-      "loss": 0.2656,
+      "epoch": 3.0498792662269794,
+      "grad_norm": 0.2881629168987274,
+      "learning_rate": 1.7446956434639884e-05,
+      "loss": 0.4127,
       "step": 84625
     },
     {
-      "epoch": 2.98,
-      "learning_rate": 1.857140457205574e-05,
-      "loss": 0.2666,
+      "epoch": 3.050059465888204,
+      "grad_norm": 0.2254873812198639,
+      "learning_rate": 1.7444174718733308e-05,
+      "loss": 0.4088,
       "step": 84630
     },
     {
-      "epoch": 2.98,
-      "learning_rate": 1.8568651686020037e-05,
-      "loss": 0.288,
+      "epoch": 3.050239665549429,
+      "grad_norm": 0.21899674832820892,
+      "learning_rate": 1.744139310577339e-05,
+      "loss": 0.3857,
       "step": 84635
     },
     {
-      "epoch": 2.98,
-      "learning_rate": 1.856589888349116e-05,
-      "loss": 0.2713,
+      "epoch": 3.0504198652106536,
+      "grad_norm": 0.23545971512794495,
+      "learning_rate": 1.743861159579802e-05,
+      "loss": 0.3672,
       "step": 84640
     },
     {
-      "epoch": 2.98,
-      "learning_rate": 1.856314616450483e-05,
-      "loss": 0.2631,
+      "epoch": 3.050600064871878,
+      "grad_norm": 0.23571446537971497,
+      "learning_rate": 1.743583018884511e-05,
+      "loss": 0.3931,
       "step": 84645
     },
     {
-      "epoch": 2.98,
-      "learning_rate": 1.8560393529096827e-05,
-      "loss": 0.2697,
+      "epoch": 3.0507802645331026,
+      "grad_norm": 0.23098576068878174,
+      "learning_rate": 1.7433048884952548e-05,
+      "loss": 0.3968,
       "step": 84650
     },
     {
-      "epoch": 2.98,
-      "learning_rate": 1.855764097730286e-05,
-      "loss": 0.2666,
+      "epoch": 3.0509604641943273,
+      "grad_norm": 0.2752792537212372,
+      "learning_rate": 1.7430267684158226e-05,
+      "loss": 0.3622,
       "step": 84655
     },
     {
-      "epoch": 2.98,
-      "learning_rate": 1.855488850915868e-05,
-      "loss": 0.2862,
+      "epoch": 3.051140663855552,
+      "grad_norm": 0.2383970022201538,
+      "learning_rate": 1.7427486586500053e-05,
+      "loss": 0.4099,
       "step": 84660
     },
     {
-      "epoch": 2.98,
-      "learning_rate": 1.8552136124700014e-05,
-      "loss": 0.264,
+      "epoch": 3.051320863516777,
+      "grad_norm": 0.22088317573070526,
+      "learning_rate": 1.742470559201589e-05,
+      "loss": 0.382,
       "step": 84665
     },
     {
-      "epoch": 2.98,
-      "learning_rate": 1.8549383823962625e-05,
-      "loss": 0.2594,
+      "epoch": 3.051501063178001,
+      "grad_norm": 0.2405635267496109,
+      "learning_rate": 1.7421924700743668e-05,
+      "loss": 0.3674,
       "step": 84670
     },
     {
-      "epoch": 2.98,
-      "learning_rate": 1.854663160698224e-05,
-      "loss": 0.2513,
+      "epoch": 3.051681262839226,
+      "grad_norm": 0.2541629672050476,
+      "learning_rate": 1.741914391272124e-05,
+      "loss": 0.3796,
       "step": 84675
     },
     {
-      "epoch": 2.98,
-      "learning_rate": 1.8543879473794584e-05,
-      "loss": 0.2686,
+      "epoch": 3.0518614625004505,
+      "grad_norm": 0.18502090871334076,
+      "learning_rate": 1.7416363227986507e-05,
+      "loss": 0.4146,
       "step": 84680
     },
     {
-      "epoch": 2.98,
-      "learning_rate": 1.854112742443539e-05,
-      "loss": 0.2723,
+      "epoch": 3.0520416621616753,
+      "grad_norm": 0.20814743638038635,
+      "learning_rate": 1.741358264657737e-05,
+      "loss": 0.3825,
       "step": 84685
     },
     {
-      "epoch": 2.98,
-      "learning_rate": 1.8538375458940408e-05,
-      "loss": 0.2528,
+      "epoch": 3.0522218618228996,
+      "grad_norm": 0.22809316217899323,
+      "learning_rate": 1.7410802168531684e-05,
+      "loss": 0.4106,
       "step": 84690
     },
     {
-      "epoch": 2.98,
-      "learning_rate": 1.8535623577345364e-05,
-      "loss": 0.285,
+      "epoch": 3.0524020614841243,
+      "grad_norm": 0.2442580610513687,
+      "learning_rate": 1.7408021793887363e-05,
+      "loss": 0.3786,
       "step": 84695
     },
     {
-      "epoch": 2.98,
-      "learning_rate": 1.853287177968598e-05,
-      "loss": 0.2968,
+      "epoch": 3.052582261145349,
+      "grad_norm": 0.2511539161205292,
+      "learning_rate": 1.7405241522682276e-05,
+      "loss": 0.4042,
       "step": 84700
     },
     {
-      "epoch": 2.98,
-      "learning_rate": 1.8530120065998004e-05,
-      "loss": 0.2695,
+      "epoch": 3.0527624608065738,
+      "grad_norm": 0.20690898597240448,
+      "learning_rate": 1.740246135495429e-05,
+      "loss": 0.3782,
       "step": 84705
     },
     {
-      "epoch": 2.98,
-      "learning_rate": 1.852736843631715e-05,
-      "loss": 0.2702,
+      "epoch": 3.0529426604677985,
+      "grad_norm": 0.20105348527431488,
+      "learning_rate": 1.7399681290741308e-05,
+      "loss": 0.3946,
       "step": 84710
     },
     {
-      "epoch": 2.98,
-      "learning_rate": 1.852461689067915e-05,
-      "loss": 0.2808,
+      "epoch": 3.053122860129023,
+      "grad_norm": 0.2233029156923294,
+      "learning_rate": 1.739690133008119e-05,
+      "loss": 0.3783,
       "step": 84715
     },
     {
-      "epoch": 2.98,
-      "learning_rate": 1.852186542911973e-05,
-      "loss": 0.2427,
+      "epoch": 3.0533030597902475,
+      "grad_norm": 0.2557271122932434,
+      "learning_rate": 1.7394121473011825e-05,
+      "loss": 0.3672,
       "step": 84720
     },
     {
-      "epoch": 2.98,
-      "learning_rate": 1.8519114051674626e-05,
-      "loss": 0.2538,
+      "epoch": 3.0534832594514723,
+      "grad_norm": 0.2504826784133911,
+      "learning_rate": 1.739134171957108e-05,
+      "loss": 0.3653,
       "step": 84725
     },
     {
-      "epoch": 2.98,
-      "learning_rate": 1.8516362758379552e-05,
-      "loss": 0.2875,
+      "epoch": 3.053663459112697,
+      "grad_norm": 0.19489239156246185,
+      "learning_rate": 1.7388562069796827e-05,
+      "loss": 0.3771,
       "step": 84730
     },
     {
-      "epoch": 2.98,
-      "learning_rate": 1.851361154927024e-05,
-      "loss": 0.2583,
+      "epoch": 3.0538436587739217,
+      "grad_norm": 0.18892985582351685,
+      "learning_rate": 1.738578252372695e-05,
+      "loss": 0.3507,
       "step": 84735
     },
     {
-      "epoch": 2.98,
-      "learning_rate": 1.8510860424382392e-05,
-      "loss": 0.2762,
+      "epoch": 3.054023858435146,
+      "grad_norm": 0.2819010317325592,
+      "learning_rate": 1.7383003081399308e-05,
+      "loss": 0.3892,
       "step": 84740
     },
     {
-      "epoch": 2.98,
-      "learning_rate": 1.8508109383751754e-05,
-      "loss": 0.2963,
+      "epoch": 3.0542040580963707,
+      "grad_norm": 0.22074325382709503,
+      "learning_rate": 1.738022374285177e-05,
+      "loss": 0.3679,
       "step": 84745
     },
     {
-      "epoch": 2.98,
-      "learning_rate": 1.8505358427414037e-05,
-      "loss": 0.258,
+      "epoch": 3.0543842577575955,
+      "grad_norm": 0.23581378161907196,
+      "learning_rate": 1.7377444508122215e-05,
+      "loss": 0.3676,
       "step": 84750
     },
     {
-      "epoch": 2.98,
-      "learning_rate": 1.8502607555404966e-05,
-      "loss": 0.2605,
+      "epoch": 3.05456445741882,
+      "grad_norm": 0.23171649873256683,
+      "learning_rate": 1.73746653772485e-05,
+      "loss": 0.3905,
       "step": 84755
     },
     {
-      "epoch": 2.98,
-      "learning_rate": 1.8499856767760243e-05,
-      "loss": 0.2667,
+      "epoch": 3.0547446570800445,
+      "grad_norm": 0.22440069913864136,
+      "learning_rate": 1.7371886350268494e-05,
+      "loss": 0.3745,
       "step": 84760
     },
     {
-      "epoch": 2.98,
-      "learning_rate": 1.8497106064515605e-05,
-      "loss": 0.2566,
+      "epoch": 3.0549248567412692,
+      "grad_norm": 0.17826688289642334,
+      "learning_rate": 1.7369107427220066e-05,
+      "loss": 0.3795,
       "step": 84765
     },
     {
-      "epoch": 2.98,
-      "learning_rate": 1.8494355445706755e-05,
-      "loss": 0.2565,
+      "epoch": 3.055105056402494,
+      "grad_norm": 0.2648460268974304,
+      "learning_rate": 1.7366328608141057e-05,
+      "loss": 0.3735,
       "step": 84770
     },
     {
-      "epoch": 2.98,
-      "learning_rate": 1.8491604911369418e-05,
-      "loss": 0.2795,
+      "epoch": 3.0552852560637187,
+      "grad_norm": 0.18997584283351898,
+      "learning_rate": 1.7363549893069355e-05,
+      "loss": 0.3955,
       "step": 84775
     },
     {
-      "epoch": 2.98,
-      "learning_rate": 1.848885446153929e-05,
-      "loss": 0.2754,
+      "epoch": 3.0554654557249434,
+      "grad_norm": 0.18987253308296204,
+      "learning_rate": 1.7360771282042807e-05,
+      "loss": 0.383,
       "step": 84780
     },
     {
-      "epoch": 2.98,
-      "learning_rate": 1.8486104096252104e-05,
-      "loss": 0.2819,
+      "epoch": 3.0556456553861677,
+      "grad_norm": 0.23954762518405914,
+      "learning_rate": 1.7357992775099264e-05,
+      "loss": 0.407,
       "step": 84785
     },
     {
-      "epoch": 2.98,
-      "learning_rate": 1.848335381554357e-05,
-      "loss": 0.255,
+      "epoch": 3.0558258550473925,
+      "grad_norm": 0.19741961359977722,
+      "learning_rate": 1.7355214372276596e-05,
+      "loss": 0.3922,
       "step": 84790
     },
     {
-      "epoch": 2.98,
-      "learning_rate": 1.8480603619449383e-05,
-      "loss": 0.2634,
+      "epoch": 3.056006054708617,
+      "grad_norm": 0.20754849910736084,
+      "learning_rate": 1.7352436073612644e-05,
+      "loss": 0.3883,
       "step": 84795
     },
     {
-      "epoch": 2.98,
-      "learning_rate": 1.8477853508005255e-05,
-      "loss": 0.2802,
+      "epoch": 3.056186254369842,
+      "grad_norm": 0.2479812055826187,
+      "learning_rate": 1.7349657879145274e-05,
+      "loss": 0.3505,
       "step": 84800
     },
     {
-      "epoch": 2.98,
-      "learning_rate": 1.8475103481246912e-05,
-      "loss": 0.277,
+      "epoch": 3.056366454031066,
+      "grad_norm": 0.26793172955513,
+      "learning_rate": 1.734687978891234e-05,
+      "loss": 0.4124,
       "step": 84805
     },
     {
-      "epoch": 2.98,
-      "learning_rate": 1.8472353539210052e-05,
-      "loss": 0.2462,
+      "epoch": 3.056546653692291,
+      "grad_norm": 0.2121279090642929,
+      "learning_rate": 1.734410180295168e-05,
+      "loss": 0.3928,
       "step": 84810
     },
     {
-      "epoch": 2.98,
-      "learning_rate": 1.846960368193037e-05,
-      "loss": 0.2716,
+      "epoch": 3.0567268533535157,
+      "grad_norm": 0.2421426624059677,
+      "learning_rate": 1.7341323921301154e-05,
+      "loss": 0.3936,
       "step": 84815
     },
     {
-      "epoch": 2.98,
-      "learning_rate": 1.8466853909443588e-05,
-      "loss": 0.2653,
+      "epoch": 3.0569070530147404,
+      "grad_norm": 0.21488474309444427,
+      "learning_rate": 1.733854614399861e-05,
+      "loss": 0.3913,
       "step": 84820
     },
     {
-      "epoch": 2.98,
-      "learning_rate": 1.84641042217854e-05,
-      "loss": 0.281,
+      "epoch": 3.057087252675965,
+      "grad_norm": 0.18251733481884003,
+      "learning_rate": 1.733576847108188e-05,
+      "loss": 0.4056,
       "step": 84825
     },
     {
-      "epoch": 2.98,
-      "learning_rate": 1.846135461899151e-05,
-      "loss": 0.2615,
+      "epoch": 3.0572674523371894,
+      "grad_norm": 0.25495290756225586,
+      "learning_rate": 1.733299090258883e-05,
+      "loss": 0.3676,
       "step": 84830
     },
     {
-      "epoch": 2.98,
-      "learning_rate": 1.8458605101097618e-05,
-      "loss": 0.2775,
+      "epoch": 3.057447651998414,
+      "grad_norm": 0.23162417113780975,
+      "learning_rate": 1.733021343855729e-05,
+      "loss": 0.3768,
       "step": 84835
     },
     {
-      "epoch": 2.98,
-      "learning_rate": 1.8455855668139434e-05,
-      "loss": 0.2839,
+      "epoch": 3.057627851659639,
+      "grad_norm": 0.20512209832668304,
+      "learning_rate": 1.7327436079025112e-05,
+      "loss": 0.3293,
       "step": 84840
     },
     {
-      "epoch": 2.99,
-      "learning_rate": 1.8453106320152654e-05,
-      "loss": 0.2763,
+      "epoch": 3.0578080513208636,
+      "grad_norm": 0.197612926363945,
+      "learning_rate": 1.7324658824030133e-05,
+      "loss": 0.3904,
       "step": 84845
     },
     {
-      "epoch": 2.99,
-      "learning_rate": 1.8450357057172967e-05,
-      "loss": 0.2809,
+      "epoch": 3.0579882509820884,
+      "grad_norm": 0.29313725233078003,
+      "learning_rate": 1.7321881673610184e-05,
+      "loss": 0.396,
       "step": 84850
     },
     {
-      "epoch": 2.99,
-      "learning_rate": 1.8447607879236078e-05,
-      "loss": 0.2501,
+      "epoch": 3.0581684506433127,
+      "grad_norm": 0.2089136838912964,
+      "learning_rate": 1.7319104627803117e-05,
+      "loss": 0.3205,
       "step": 84855
     },
     {
-      "epoch": 2.99,
-      "learning_rate": 1.8444858786377682e-05,
-      "loss": 0.2753,
+      "epoch": 3.0583486503045374,
+      "grad_norm": 0.25626543164253235,
+      "learning_rate": 1.7316327686646767e-05,
+      "loss": 0.3811,
       "step": 84860
     },
     {
-      "epoch": 2.99,
-      "learning_rate": 1.844210977863348e-05,
-      "loss": 0.2645,
+      "epoch": 3.058528849965762,
+      "grad_norm": 0.2188102900981903,
+      "learning_rate": 1.731355085017895e-05,
+      "loss": 0.3782,
       "step": 84865
     },
     {
-      "epoch": 2.99,
-      "learning_rate": 1.8439360856039163e-05,
-      "loss": 0.3029,
+      "epoch": 3.058709049626987,
+      "grad_norm": 0.2118126004934311,
+      "learning_rate": 1.731077411843753e-05,
+      "loss": 0.3928,
       "step": 84870
     },
     {
-      "epoch": 2.99,
-      "learning_rate": 1.843661201863041e-05,
-      "loss": 0.2619,
+      "epoch": 3.058889249288211,
+      "grad_norm": 0.25317180156707764,
+      "learning_rate": 1.7307997491460306e-05,
+      "loss": 0.3693,
       "step": 84875
     },
     {
-      "epoch": 2.99,
-      "learning_rate": 1.8433863266442937e-05,
-      "loss": 0.2809,
+      "epoch": 3.059069448949436,
+      "grad_norm": 0.2602880299091339,
+      "learning_rate": 1.7305220969285148e-05,
+      "loss": 0.3928,
       "step": 84880
     },
     {
-      "epoch": 2.99,
-      "learning_rate": 1.8431114599512412e-05,
-      "loss": 0.2983,
+      "epoch": 3.0592496486106606,
+      "grad_norm": 0.23756670951843262,
+      "learning_rate": 1.7302444551949853e-05,
+      "loss": 0.3839,
       "step": 84885
     },
     {
-      "epoch": 2.99,
-      "learning_rate": 1.842836601787454e-05,
-      "loss": 0.2685,
+      "epoch": 3.0594298482718854,
+      "grad_norm": 0.2189938724040985,
+      "learning_rate": 1.729966823949226e-05,
+      "loss": 0.3825,
       "step": 84890
     },
     {
-      "epoch": 2.99,
-      "learning_rate": 1.8425617521564998e-05,
-      "loss": 0.2727,
+      "epoch": 3.05961004793311,
+      "grad_norm": 0.24415577948093414,
+      "learning_rate": 1.72968920319502e-05,
+      "loss": 0.3731,
       "step": 84895
     },
     {
-      "epoch": 2.99,
-      "learning_rate": 1.842286911061949e-05,
-      "loss": 0.2634,
+      "epoch": 3.0597902475943344,
+      "grad_norm": 0.1736556589603424,
+      "learning_rate": 1.7294115929361492e-05,
+      "loss": 0.3822,
       "step": 84900
     },
     {
-      "epoch": 2.99,
-      "learning_rate": 1.8420120785073685e-05,
-      "loss": 0.2866,
+      "epoch": 3.059970447255559,
+      "grad_norm": 0.2100590169429779,
+      "learning_rate": 1.7291339931763962e-05,
+      "loss": 0.3883,
       "step": 84905
     },
     {
-      "epoch": 2.99,
-      "learning_rate": 1.841737254496328e-05,
-      "loss": 0.2641,
+      "epoch": 3.060150646916784,
+      "grad_norm": 0.2625679671764374,
+      "learning_rate": 1.7288564039195434e-05,
+      "loss": 0.3733,
       "step": 84910
     },
     {
-      "epoch": 2.99,
-      "learning_rate": 1.8414624390323946e-05,
-      "loss": 0.2716,
+      "epoch": 3.0603308465780086,
+      "grad_norm": 0.2331903725862503,
+      "learning_rate": 1.7285788251693723e-05,
+      "loss": 0.3834,
       "step": 84915
     },
     {
-      "epoch": 2.99,
-      "learning_rate": 1.8411876321191384e-05,
-      "loss": 0.2535,
+      "epoch": 3.060511046239233,
+      "grad_norm": 0.21731112897396088,
+      "learning_rate": 1.7283012569296665e-05,
+      "loss": 0.3721,
       "step": 84920
     },
     {
-      "epoch": 2.99,
-      "learning_rate": 1.8409128337601267e-05,
-      "loss": 0.2594,
+      "epoch": 3.0606912459004576,
+      "grad_norm": 0.22888799011707306,
+      "learning_rate": 1.728023699204206e-05,
+      "loss": 0.3532,
       "step": 84925
     },
     {
-      "epoch": 2.99,
-      "learning_rate": 1.8406380439589262e-05,
-      "loss": 0.2455,
+      "epoch": 3.0608714455616823,
+      "grad_norm": 0.2064668983221054,
+      "learning_rate": 1.727746151996773e-05,
+      "loss": 0.3789,
       "step": 84930
     },
     {
-      "epoch": 2.99,
-      "learning_rate": 1.8403632627191074e-05,
-      "loss": 0.2776,
+      "epoch": 3.061051645222907,
+      "grad_norm": 0.1944301277399063,
+      "learning_rate": 1.7274686153111496e-05,
+      "loss": 0.3762,
       "step": 84935
     },
     {
-      "epoch": 2.99,
-      "learning_rate": 1.840088490044236e-05,
-      "loss": 0.2514,
+      "epoch": 3.061231844884132,
+      "grad_norm": 0.2432466596364975,
+      "learning_rate": 1.7271910891511163e-05,
+      "loss": 0.3953,
       "step": 84940
     },
     {
-      "epoch": 2.99,
-      "learning_rate": 1.8398137259378816e-05,
-      "loss": 0.2845,
+      "epoch": 3.061412044545356,
+      "grad_norm": 0.1888858526945114,
+      "learning_rate": 1.726913573520455e-05,
+      "loss": 0.3832,
       "step": 84945
     },
     {
-      "epoch": 2.99,
-      "learning_rate": 1.8395389704036096e-05,
-      "loss": 0.2742,
+      "epoch": 3.061592244206581,
+      "grad_norm": 0.27773165702819824,
+      "learning_rate": 1.7266360684229473e-05,
+      "loss": 0.3922,
       "step": 84950
     },
     {
-      "epoch": 2.99,
-      "learning_rate": 1.8392642234449903e-05,
-      "loss": 0.2583,
+      "epoch": 3.0617724438678056,
+      "grad_norm": 0.1935848444700241,
+      "learning_rate": 1.7263585738623715e-05,
+      "loss": 0.3716,
       "step": 84955
     },
     {
-      "epoch": 2.99,
-      "learning_rate": 1.8389894850655895e-05,
-      "loss": 0.2654,
+      "epoch": 3.0619526435290303,
+      "grad_norm": 0.2068513184785843,
+      "learning_rate": 1.7260810898425126e-05,
+      "loss": 0.3646,
       "step": 84960
     },
     {
-      "epoch": 2.99,
-      "learning_rate": 1.8387147552689737e-05,
-      "loss": 0.2815,
+      "epoch": 3.0621328431902546,
+      "grad_norm": 0.22322344779968262,
+      "learning_rate": 1.725803616367148e-05,
+      "loss": 0.3975,
       "step": 84965
     },
     {
-      "epoch": 2.99,
-      "learning_rate": 1.838440034058711e-05,
-      "loss": 0.2674,
+      "epoch": 3.0623130428514793,
+      "grad_norm": 0.19500023126602173,
+      "learning_rate": 1.7255261534400585e-05,
+      "loss": 0.3771,
       "step": 84970
     },
     {
-      "epoch": 2.99,
-      "learning_rate": 1.8381653214383692e-05,
-      "loss": 0.2691,
+      "epoch": 3.062493242512704,
+      "grad_norm": 0.20665855705738068,
+      "learning_rate": 1.7252487010650266e-05,
+      "loss": 0.3912,
       "step": 84975
     },
     {
-      "epoch": 2.99,
-      "learning_rate": 1.8378906174115145e-05,
-      "loss": 0.2662,
+      "epoch": 3.062673442173929,
+      "grad_norm": 0.24489200115203857,
+      "learning_rate": 1.7249712592458294e-05,
+      "loss": 0.3861,
       "step": 84980
     },
     {
-      "epoch": 2.99,
-      "learning_rate": 1.8376159219817135e-05,
-      "loss": 0.2625,
+      "epoch": 3.0628536418351535,
+      "grad_norm": 0.22476324439048767,
+      "learning_rate": 1.7246938279862508e-05,
+      "loss": 0.35,
       "step": 84985
     },
     {
-      "epoch": 2.99,
-      "learning_rate": 1.837341235152533e-05,
-      "loss": 0.2713,
+      "epoch": 3.063033841496378,
+      "grad_norm": 0.2135080248117447,
+      "learning_rate": 1.7244164072900678e-05,
+      "loss": 0.3628,
       "step": 84990
     },
     {
-      "epoch": 2.99,
-      "learning_rate": 1.83706655692754e-05,
-      "loss": 0.2785,
+      "epoch": 3.0632140411576025,
+      "grad_norm": 0.2153581976890564,
+      "learning_rate": 1.7241389971610607e-05,
+      "loss": 0.3808,
       "step": 84995
     },
     {
-      "epoch": 2.99,
-      "learning_rate": 1.8367918873103012e-05,
-      "loss": 0.2897,
+      "epoch": 3.0633942408188273,
+      "grad_norm": 0.27249398827552795,
+      "learning_rate": 1.72386159760301e-05,
+      "loss": 0.3824,
       "step": 85000
     },
     {
-      "epoch": 2.99,
-      "eval_loss": 0.2575870752334595,
-      "eval_runtime": 10.5655,
-      "eval_samples_per_second": 9.465,
-      "eval_steps_per_second": 9.465,
+      "epoch": 3.0633942408188273,
+      "eval_loss": 0.43113651871681213,
+      "eval_runtime": 3.5309,
+      "eval_samples_per_second": 28.322,
+      "eval_steps_per_second": 7.08,
       "step": 85000
     },
     {
-      "epoch": 2.99,
-      "learning_rate": 1.8365172263043827e-05,
-      "loss": 0.2663,
+      "epoch": 3.063574440480052,
+      "grad_norm": 0.23285353183746338,
+      "learning_rate": 1.723584208619695e-05,
+      "loss": 0.3787,
       "step": 85005
     },
     {
-      "epoch": 2.99,
-      "learning_rate": 1.83624257391335e-05,
-      "loss": 0.2902,
+      "epoch": 3.0637546401412767,
+      "grad_norm": 0.21958500146865845,
+      "learning_rate": 1.7233068302148943e-05,
+      "loss": 0.4006,
       "step": 85010
     },
     {
-      "epoch": 2.99,
-      "learning_rate": 1.8359679301407705e-05,
-      "loss": 0.2739,
+      "epoch": 3.063934839802501,
+      "grad_norm": 0.26001232862472534,
+      "learning_rate": 1.7230294623923876e-05,
+      "loss": 0.3732,
       "step": 85015
     },
     {
-      "epoch": 2.99,
-      "learning_rate": 1.83569329499021e-05,
-      "loss": 0.2879,
+      "epoch": 3.0641150394637258,
+      "grad_norm": 0.20157134532928467,
+      "learning_rate": 1.722752105155954e-05,
+      "loss": 0.3474,
       "step": 85020
     },
     {
-      "epoch": 2.99,
-      "learning_rate": 1.8354186684652343e-05,
-      "loss": 0.2705,
+      "epoch": 3.0642952391249505,
+      "grad_norm": 0.2562786638736725,
+      "learning_rate": 1.722474758509373e-05,
+      "loss": 0.4045,
       "step": 85025
     },
     {
-      "epoch": 2.99,
-      "learning_rate": 1.8351440505694083e-05,
-      "loss": 0.2706,
+      "epoch": 3.0644754387861752,
+      "grad_norm": 0.22175267338752747,
+      "learning_rate": 1.722197422456423e-05,
+      "loss": 0.3797,
       "step": 85030
     },
     {
-      "epoch": 2.99,
-      "learning_rate": 1.8348694413063e-05,
-      "loss": 0.2817,
+      "epoch": 3.0646556384473995,
+      "grad_norm": 0.19869141280651093,
+      "learning_rate": 1.721920097000882e-05,
+      "loss": 0.3716,
       "step": 85035
     },
     {
-      "epoch": 2.99,
-      "learning_rate": 1.8345948406794733e-05,
-      "loss": 0.2729,
+      "epoch": 3.0648358381086243,
+      "grad_norm": 0.2057720124721527,
+      "learning_rate": 1.7216427821465292e-05,
+      "loss": 0.3932,
       "step": 85040
     },
     {
-      "epoch": 2.99,
-      "learning_rate": 1.8343202486924933e-05,
-      "loss": 0.2621,
+      "epoch": 3.065016037769849,
+      "grad_norm": 0.24935896694660187,
+      "learning_rate": 1.7213654778971436e-05,
+      "loss": 0.3798,
       "step": 85045
     },
     {
-      "epoch": 2.99,
-      "learning_rate": 1.8340456653489264e-05,
-      "loss": 0.2582,
+      "epoch": 3.0651962374310737,
+      "grad_norm": 0.2382826954126358,
+      "learning_rate": 1.7210881842565007e-05,
+      "loss": 0.3826,
       "step": 85050
     },
     {
-      "epoch": 2.99,
-      "learning_rate": 1.8337710906523384e-05,
-      "loss": 0.2626,
+      "epoch": 3.0653764370922985,
+      "grad_norm": 0.2057560682296753,
+      "learning_rate": 1.7208109012283824e-05,
+      "loss": 0.3825,
       "step": 85055
     },
     {
-      "epoch": 2.99,
-      "learning_rate": 1.8334965246062936e-05,
-      "loss": 0.2544,
+      "epoch": 3.0655566367535227,
+      "grad_norm": 0.20751681923866272,
+      "learning_rate": 1.720533628816563e-05,
+      "loss": 0.3988,
       "step": 85060
     },
     {
-      "epoch": 2.99,
-      "learning_rate": 1.833221967214356e-05,
-      "loss": 0.2565,
+      "epoch": 3.0657368364147475,
+      "grad_norm": 0.22527416050434113,
+      "learning_rate": 1.720256367024824e-05,
+      "loss": 0.3742,
       "step": 85065
     },
     {
-      "epoch": 2.99,
-      "learning_rate": 1.8329474184800932e-05,
-      "loss": 0.2458,
+      "epoch": 3.065917036075972,
+      "grad_norm": 0.24858717620372772,
+      "learning_rate": 1.71997911585694e-05,
+      "loss": 0.385,
       "step": 85070
     },
     {
-      "epoch": 2.99,
-      "learning_rate": 1.8326728784070676e-05,
-      "loss": 0.2674,
+      "epoch": 3.066097235737197,
+      "grad_norm": 0.2016977220773697,
+      "learning_rate": 1.7197018753166895e-05,
+      "loss": 0.3803,
       "step": 85075
     },
     {
-      "epoch": 2.99,
-      "learning_rate": 1.8323983469988458e-05,
-      "loss": 0.2548,
+      "epoch": 3.0662774353984212,
+      "grad_norm": 0.23507557809352875,
+      "learning_rate": 1.71942464540785e-05,
+      "loss": 0.3956,
       "step": 85080
     },
     {
-      "epoch": 2.99,
-      "learning_rate": 1.83212382425899e-05,
-      "loss": 0.279,
+      "epoch": 3.066457635059646,
+      "grad_norm": 0.18204432725906372,
+      "learning_rate": 1.7191474261341982e-05,
+      "loss": 0.4038,
       "step": 85085
     },
     {
-      "epoch": 2.99,
-      "learning_rate": 1.8318493101910676e-05,
-      "loss": 0.2629,
+      "epoch": 3.0666378347208707,
+      "grad_norm": 0.19573719799518585,
+      "learning_rate": 1.7188702174995115e-05,
+      "loss": 0.3751,
       "step": 85090
     },
     {
-      "epoch": 2.99,
-      "learning_rate": 1.8315748047986415e-05,
-      "loss": 0.2786,
+      "epoch": 3.0668180343820954,
+      "grad_norm": 0.2764735221862793,
+      "learning_rate": 1.7185930195075675e-05,
+      "loss": 0.3545,
       "step": 85095
     },
     {
-      "epoch": 2.99,
-      "learning_rate": 1.8313003080852755e-05,
-      "loss": 0.2567,
+      "epoch": 3.06699823404332,
+      "grad_norm": 0.2690604031085968,
+      "learning_rate": 1.718315832162142e-05,
+      "loss": 0.3845,
       "step": 85100
     },
     {
-      "epoch": 2.99,
-      "learning_rate": 1.8310258200545338e-05,
-      "loss": 0.2676,
+      "epoch": 3.0671784337045445,
+      "grad_norm": 0.22735117375850677,
+      "learning_rate": 1.718038655467012e-05,
+      "loss": 0.4114,
       "step": 85105
     },
     {
-      "epoch": 2.99,
-      "learning_rate": 1.830751340709982e-05,
-      "loss": 0.2571,
+      "epoch": 3.067358633365769,
+      "grad_norm": 0.25227391719818115,
+      "learning_rate": 1.7177614894259538e-05,
+      "loss": 0.4067,
       "step": 85110
     },
     {
-      "epoch": 2.99,
-      "learning_rate": 1.830476870055183e-05,
-      "loss": 0.2583,
+      "epoch": 3.067538833026994,
+      "grad_norm": 0.22190016508102417,
+      "learning_rate": 1.717484334042744e-05,
+      "loss": 0.4165,
       "step": 85115
     },
     {
-      "epoch": 2.99,
-      "learning_rate": 1.8302024080937008e-05,
-      "loss": 0.2619,
+      "epoch": 3.0677190326882187,
+      "grad_norm": 0.2443481832742691,
+      "learning_rate": 1.7172071893211583e-05,
+      "loss": 0.4004,
       "step": 85120
     },
     {
-      "epoch": 2.99,
-      "learning_rate": 1.8299279548290975e-05,
-      "loss": 0.2721,
+      "epoch": 3.0678992323494434,
+      "grad_norm": 0.19374127686023712,
+      "learning_rate": 1.716930055264973e-05,
+      "loss": 0.3809,
       "step": 85125
     },
     {
-      "epoch": 3.0,
-      "learning_rate": 1.8296535102649388e-05,
-      "loss": 0.2721,
+      "epoch": 3.0680794320106677,
+      "grad_norm": 0.21551166474819183,
+      "learning_rate": 1.716652931877965e-05,
+      "loss": 0.3959,
       "step": 85130
     },
     {
-      "epoch": 3.0,
-      "learning_rate": 1.829379074404788e-05,
-      "loss": 0.2777,
+      "epoch": 3.0682596316718924,
+      "grad_norm": 0.2144090086221695,
+      "learning_rate": 1.7163758191639085e-05,
+      "loss": 0.4067,
       "step": 85135
     },
     {
-      "epoch": 3.0,
-      "learning_rate": 1.829104647252208e-05,
-      "loss": 0.2647,
+      "epoch": 3.068439831333117,
+      "grad_norm": 0.25093093514442444,
+      "learning_rate": 1.7160987171265798e-05,
+      "loss": 0.3913,
       "step": 85140
     },
     {
-      "epoch": 3.0,
-      "learning_rate": 1.8288302288107613e-05,
-      "loss": 0.2806,
+      "epoch": 3.068620030994342,
+      "grad_norm": 0.24009042978286743,
+      "learning_rate": 1.7158216257697545e-05,
+      "loss": 0.3445,
       "step": 85145
     },
     {
-      "epoch": 3.0,
-      "learning_rate": 1.8285558190840123e-05,
-      "loss": 0.2671,
+      "epoch": 3.068800230655566,
+      "grad_norm": 0.21893219649791718,
+      "learning_rate": 1.715544545097208e-05,
+      "loss": 0.3871,
       "step": 85150
     },
     {
-      "epoch": 3.0,
-      "learning_rate": 1.828281418075523e-05,
-      "loss": 0.2809,
+      "epoch": 3.068980430316791,
+      "grad_norm": 0.21242910623550415,
+      "learning_rate": 1.715267475112714e-05,
+      "loss": 0.3822,
       "step": 85155
     },
     {
-      "epoch": 3.0,
-      "learning_rate": 1.8280070257888565e-05,
-      "loss": 0.2558,
+      "epoch": 3.0691606299780156,
+      "grad_norm": 0.209553524851799,
+      "learning_rate": 1.7149904158200504e-05,
+      "loss": 0.379,
       "step": 85160
     },
     {
-      "epoch": 3.0,
-      "learning_rate": 1.8277326422275767e-05,
-      "loss": 0.2798,
+      "epoch": 3.0693408296392404,
+      "grad_norm": 0.2270086407661438,
+      "learning_rate": 1.7147133672229885e-05,
+      "loss": 0.3807,
       "step": 85165
     },
     {
-      "epoch": 3.0,
-      "learning_rate": 1.8274582673952458e-05,
-      "loss": 0.2723,
+      "epoch": 3.069521029300465,
+      "grad_norm": 0.1918162852525711,
+      "learning_rate": 1.7144363293253066e-05,
+      "loss": 0.3994,
       "step": 85170
     },
     {
-      "epoch": 3.0,
-      "learning_rate": 1.8271839012954258e-05,
-      "loss": 0.2397,
+      "epoch": 3.0697012289616894,
+      "grad_norm": 0.18857698142528534,
+      "learning_rate": 1.7141593021307774e-05,
+      "loss": 0.412,
       "step": 85175
     },
     {
-      "epoch": 3.0,
-      "learning_rate": 1.8269095439316786e-05,
-      "loss": 0.259,
+      "epoch": 3.069881428622914,
+      "grad_norm": 0.27960577607154846,
+      "learning_rate": 1.7138822856431746e-05,
+      "loss": 0.3907,
       "step": 85180
     },
     {
-      "epoch": 3.0,
-      "learning_rate": 1.8266351953075676e-05,
-      "loss": 0.2452,
+      "epoch": 3.070061628284139,
+      "grad_norm": 0.21300694346427917,
+      "learning_rate": 1.713605279866274e-05,
+      "loss": 0.3445,
       "step": 85185
     },
     {
-      "epoch": 3.0,
-      "learning_rate": 1.826360855426656e-05,
-      "loss": 0.2612,
+      "epoch": 3.0702418279453636,
+      "grad_norm": 0.2177317589521408,
+      "learning_rate": 1.7133282848038495e-05,
+      "loss": 0.3532,
       "step": 85190
     },
     {
-      "epoch": 3.0,
-      "learning_rate": 1.8260865242925042e-05,
-      "loss": 0.2567,
+      "epoch": 3.070422027606588,
+      "grad_norm": 0.2585134208202362,
+      "learning_rate": 1.713051300459674e-05,
+      "loss": 0.4105,
       "step": 85195
     },
     {
-      "epoch": 3.0,
-      "learning_rate": 1.8258122019086736e-05,
-      "loss": 0.2513,
+      "epoch": 3.0706022272678126,
+      "grad_norm": 0.2013564258813858,
+      "learning_rate": 1.712774326837523e-05,
+      "loss": 0.3809,
       "step": 85200
     },
     {
-      "epoch": 3.0,
-      "learning_rate": 1.825537888278729e-05,
-      "loss": 0.2751,
+      "epoch": 3.0707824269290374,
+      "grad_norm": 0.1960568130016327,
+      "learning_rate": 1.7124973639411686e-05,
+      "loss": 0.3597,
       "step": 85205
     },
     {
-      "epoch": 3.0,
-      "learning_rate": 1.8252635834062292e-05,
-      "loss": 0.2635,
+      "epoch": 3.070962626590262,
+      "grad_norm": 0.2271984964609146,
+      "learning_rate": 1.712220411774386e-05,
+      "loss": 0.3732,
       "step": 85210
     },
     {
-      "epoch": 3.0,
-      "learning_rate": 1.824989287294738e-05,
-      "loss": 0.2698,
+      "epoch": 3.071142826251487,
+      "grad_norm": 0.2209077626466751,
+      "learning_rate": 1.7119434703409475e-05,
+      "loss": 0.3632,
       "step": 85215
     },
     {
-      "epoch": 3.0,
-      "learning_rate": 1.824714999947815e-05,
-      "loss": 0.2512,
+      "epoch": 3.071323025912711,
+      "grad_norm": 0.24356262385845184,
+      "learning_rate": 1.7116665396446262e-05,
+      "loss": 0.3656,
       "step": 85220
     },
     {
-      "epoch": 3.0,
-      "learning_rate": 1.8244407213690234e-05,
-      "loss": 0.287,
+      "epoch": 3.071503225573936,
+      "grad_norm": 0.2605443298816681,
+      "learning_rate": 1.7113896196891963e-05,
+      "loss": 0.3577,
       "step": 85225
     },
     {
-      "epoch": 3.0,
-      "learning_rate": 1.8241664515619245e-05,
-      "loss": 0.2691,
+      "epoch": 3.0716834252351606,
+      "grad_norm": 0.23770549893379211,
+      "learning_rate": 1.7111127104784305e-05,
+      "loss": 0.3835,
       "step": 85230
     },
     {
-      "epoch": 3.0,
-      "learning_rate": 1.8238921905300777e-05,
-      "loss": 0.2627,
+      "epoch": 3.0718636248963853,
+      "grad_norm": 0.225930318236351,
+      "learning_rate": 1.7108358120160997e-05,
+      "loss": 0.4269,
       "step": 85235
     },
     {
-      "epoch": 3.0,
-      "learning_rate": 1.8236179382770448e-05,
-      "loss": 0.2863,
+      "epoch": 3.0720438245576096,
+      "grad_norm": 0.24659954011440277,
+      "learning_rate": 1.7105589243059798e-05,
+      "loss": 0.3578,
       "step": 85240
     },
     {
-      "epoch": 3.0,
-      "learning_rate": 1.8233436948063885e-05,
-      "loss": 0.2653,
+      "epoch": 3.0722240242188343,
+      "grad_norm": 0.17880700528621674,
+      "learning_rate": 1.7102820473518404e-05,
+      "loss": 0.3679,
       "step": 85245
     },
     {
-      "epoch": 3.0,
-      "learning_rate": 1.823069460121668e-05,
-      "loss": 0.2855,
+      "epoch": 3.072404223880059,
+      "grad_norm": 0.21226419508457184,
+      "learning_rate": 1.7100051811574564e-05,
+      "loss": 0.3876,
       "step": 85250
     },
     {
-      "epoch": 3.0,
-      "learning_rate": 1.822795234226445e-05,
-      "loss": 0.289,
+      "epoch": 3.072584423541284,
+      "grad_norm": 0.23851273953914642,
+      "learning_rate": 1.7097283257265983e-05,
+      "loss": 0.3621,
       "step": 85255
     },
     {
-      "epoch": 3.0,
-      "learning_rate": 1.8225210171242778e-05,
-      "loss": 0.2794,
+      "epoch": 3.0727646232025085,
+      "grad_norm": 0.20797961950302124,
+      "learning_rate": 1.709451481063038e-05,
+      "loss": 0.3759,
       "step": 85260
     },
     {
-      "epoch": 3.0,
-      "learning_rate": 1.82224680881873e-05,
-      "loss": 0.2918,
+      "epoch": 3.072944822863733,
+      "grad_norm": 0.22600434720516205,
+      "learning_rate": 1.70917464717055e-05,
+      "loss": 0.3779,
       "step": 85265
     },
     {
-      "epoch": 3.0,
-      "learning_rate": 1.821972609313361e-05,
-      "loss": 0.2634,
+      "epoch": 3.0731250225249576,
+      "grad_norm": 0.2033650130033493,
+      "learning_rate": 1.7088978240529034e-05,
+      "loss": 0.3632,
       "step": 85270
     },
     {
-      "epoch": 3.0,
-      "learning_rate": 1.8216984186117297e-05,
-      "loss": 0.249,
+      "epoch": 3.0733052221861823,
+      "grad_norm": 0.22226615250110626,
+      "learning_rate": 1.70862101171387e-05,
+      "loss": 0.3842,
       "step": 85275
     },
     {
-      "epoch": 3.0,
-      "learning_rate": 1.8214242367173984e-05,
-      "loss": 0.2576,
+      "epoch": 3.073485421847407,
+      "grad_norm": 0.22112597525119781,
+      "learning_rate": 1.7083442101572235e-05,
+      "loss": 0.3838,
       "step": 85280
     },
     {
-      "epoch": 3.0,
-      "learning_rate": 1.821150063633926e-05,
-      "loss": 0.2529,
+      "epoch": 3.0736656215086318,
+      "grad_norm": 0.2128104269504547,
+      "learning_rate": 1.7080674193867325e-05,
+      "loss": 0.3375,
       "step": 85285
     },
     {
-      "epoch": 3.0,
-      "learning_rate": 1.820875899364873e-05,
-      "loss": 0.2693,
+      "epoch": 3.073845821169856,
+      "grad_norm": 0.2777286469936371,
+      "learning_rate": 1.7077906394061703e-05,
+      "loss": 0.3895,
       "step": 85290
     },
     {
-      "epoch": 3.0,
-      "learning_rate": 1.820601743913798e-05,
-      "loss": 0.2434,
+      "epoch": 3.074026020831081,
+      "grad_norm": 0.22353100776672363,
+      "learning_rate": 1.7075138702193074e-05,
+      "loss": 0.3957,
       "step": 85295
     },
     {
-      "epoch": 3.0,
-      "learning_rate": 1.820327597284262e-05,
-      "loss": 0.2495,
+      "epoch": 3.0742062204923055,
+      "grad_norm": 0.2158191204071045,
+      "learning_rate": 1.7072371118299142e-05,
+      "loss": 0.4044,
       "step": 85300
     },
     {
-      "epoch": 3.0,
-      "learning_rate": 1.8200534594798244e-05,
-      "loss": 0.247,
+      "epoch": 3.0743864201535303,
+      "grad_norm": 0.2143658846616745,
+      "learning_rate": 1.7069603642417622e-05,
+      "loss": 0.3425,
       "step": 85305
     },
     {
-      "epoch": 3.0,
-      "learning_rate": 1.819779330504045e-05,
-      "loss": 0.2358,
+      "epoch": 3.0745666198147545,
+      "grad_norm": 0.21374057233333588,
+      "learning_rate": 1.7066836274586214e-05,
+      "loss": 0.3924,
       "step": 85310
     },
     {
-      "epoch": 3.0,
-      "learning_rate": 1.819505210360481e-05,
-      "loss": 0.2602,
+      "epoch": 3.0747468194759793,
+      "grad_norm": 0.21760691702365875,
+      "learning_rate": 1.7064069014842626e-05,
+      "loss": 0.4043,
       "step": 85315
     },
     {
-      "epoch": 3.0,
-      "learning_rate": 1.819231099052695e-05,
-      "loss": 0.2637,
+      "epoch": 3.074927019137204,
+      "grad_norm": 0.25695380568504333,
+      "learning_rate": 1.7061301863224566e-05,
+      "loss": 0.3793,
       "step": 85320
     },
     {
-      "epoch": 3.0,
-      "learning_rate": 1.818956996584244e-05,
-      "loss": 0.2622,
+      "epoch": 3.0751072187984287,
+      "grad_norm": 0.22985416650772095,
+      "learning_rate": 1.7058534819769724e-05,
+      "loss": 0.3659,
       "step": 85325
     },
     {
-      "epoch": 3.0,
-      "learning_rate": 1.818682902958688e-05,
-      "loss": 0.2495,
+      "epoch": 3.0752874184596535,
+      "grad_norm": 0.2104073315858841,
+      "learning_rate": 1.7055767884515815e-05,
+      "loss": 0.3891,
       "step": 85330
     },
     {
-      "epoch": 3.0,
-      "learning_rate": 1.8184088181795844e-05,
-      "loss": 0.2399,
+      "epoch": 3.0754676181208778,
+      "grad_norm": 0.22502779960632324,
+      "learning_rate": 1.7053001057500534e-05,
+      "loss": 0.3922,
       "step": 85335
     },
     {
-      "epoch": 3.0,
-      "learning_rate": 1.818134742250494e-05,
-      "loss": 0.2366,
+      "epoch": 3.0756478177821025,
+      "grad_norm": 0.26956361532211304,
+      "learning_rate": 1.705023433876156e-05,
+      "loss": 0.3993,
       "step": 85340
     },
     {
-      "epoch": 3.0,
-      "learning_rate": 1.8178606751749742e-05,
-      "loss": 0.2724,
+      "epoch": 3.0758280174433272,
+      "grad_norm": 0.22516946494579315,
+      "learning_rate": 1.704746772833662e-05,
+      "loss": 0.3681,
       "step": 85345
     },
     {
-      "epoch": 3.0,
-      "learning_rate": 1.817586616956585e-05,
-      "loss": 0.2627,
+      "epoch": 3.076008217104552,
+      "grad_norm": 0.2101748287677765,
+      "learning_rate": 1.7044701226263374e-05,
+      "loss": 0.3783,
       "step": 85350
     },
     {
-      "epoch": 3.0,
-      "learning_rate": 1.8173125675988822e-05,
-      "loss": 0.2537,
+      "epoch": 3.0761884167657767,
+      "grad_norm": 0.2022121697664261,
+      "learning_rate": 1.704193483257955e-05,
+      "loss": 0.4104,
       "step": 85355
     },
     {
-      "epoch": 3.0,
-      "learning_rate": 1.817038527105427e-05,
-      "loss": 0.2662,
+      "epoch": 3.076368616427001,
+      "grad_norm": 0.2616555690765381,
+      "learning_rate": 1.703916854732282e-05,
+      "loss": 0.395,
       "step": 85360
     },
     {
-      "epoch": 3.0,
-      "learning_rate": 1.8167644954797763e-05,
-      "loss": 0.2653,
+      "epoch": 3.0765488160882257,
+      "grad_norm": 0.2326449602842331,
+      "learning_rate": 1.7036402370530867e-05,
+      "loss": 0.3604,
       "step": 85365
     },
     {
-      "epoch": 3.0,
-      "learning_rate": 1.8164904727254883e-05,
-      "loss": 0.2675,
+      "epoch": 3.0767290157494505,
+      "grad_norm": 0.25773555040359497,
+      "learning_rate": 1.70336363022414e-05,
+      "loss": 0.3656,
       "step": 85370
     },
     {
-      "epoch": 3.0,
-      "learning_rate": 1.8162164588461207e-05,
-      "loss": 0.2528,
+      "epoch": 3.076909215410675,
+      "grad_norm": 0.228920578956604,
+      "learning_rate": 1.7030870342492098e-05,
+      "loss": 0.3941,
       "step": 85375
     },
     {
-      "epoch": 3.0,
-      "learning_rate": 1.815942453845232e-05,
-      "loss": 0.265,
+      "epoch": 3.0770894150718995,
+      "grad_norm": 0.21083427965641022,
+      "learning_rate": 1.7028104491320636e-05,
+      "loss": 0.4001,
       "step": 85380
     },
     {
-      "epoch": 3.0,
-      "learning_rate": 1.81566845772638e-05,
-      "loss": 0.2637,
+      "epoch": 3.077269614733124,
+      "grad_norm": 0.20418711006641388,
+      "learning_rate": 1.7025338748764713e-05,
+      "loss": 0.37,
       "step": 85385
     },
     {
-      "epoch": 3.0,
-      "learning_rate": 1.815394470493121e-05,
-      "loss": 0.2915,
+      "epoch": 3.077449814394349,
+      "grad_norm": 0.2817372977733612,
+      "learning_rate": 1.7022573114862e-05,
+      "loss": 0.3804,
       "step": 85390
     },
     {
-      "epoch": 3.0,
-      "learning_rate": 1.8151204921490146e-05,
-      "loss": 0.2486,
+      "epoch": 3.0776300140555737,
+      "grad_norm": 0.2094438374042511,
+      "learning_rate": 1.7019807589650187e-05,
+      "loss": 0.3896,
       "step": 85395
     },
     {
-      "epoch": 3.0,
-      "learning_rate": 1.8148465226976175e-05,
-      "loss": 0.2617,
+      "epoch": 3.0778102137167984,
+      "grad_norm": 0.2494175136089325,
+      "learning_rate": 1.701704217316696e-05,
+      "loss": 0.3862,
       "step": 85400
     },
     {
-      "epoch": 3.0,
-      "learning_rate": 1.814572562142486e-05,
-      "loss": 0.2724,
+      "epoch": 3.0779904133780227,
+      "grad_norm": 0.19059158861637115,
+      "learning_rate": 1.7014276865449973e-05,
+      "loss": 0.3787,
       "step": 85405
     },
     {
-      "epoch": 3.0,
-      "learning_rate": 1.814298610487178e-05,
-      "loss": 0.2499,
+      "epoch": 3.0781706130392474,
+      "grad_norm": 0.21895138919353485,
+      "learning_rate": 1.7011511666536923e-05,
+      "loss": 0.3882,
       "step": 85410
     },
     {
-      "epoch": 3.01,
-      "learning_rate": 1.8140246677352514e-05,
-      "loss": 0.2662,
+      "epoch": 3.078350812700472,
+      "grad_norm": 0.20317794382572174,
+      "learning_rate": 1.700874657646549e-05,
+      "loss": 0.3827,
       "step": 85415
     },
     {
-      "epoch": 3.01,
-      "learning_rate": 1.8137507338902625e-05,
-      "loss": 0.2476,
+      "epoch": 3.078531012361697,
+      "grad_norm": 0.23056404292583466,
+      "learning_rate": 1.700598159527332e-05,
+      "loss": 0.3973,
       "step": 85420
     },
     {
-      "epoch": 3.01,
-      "learning_rate": 1.8134768089557675e-05,
-      "loss": 0.2447,
+      "epoch": 3.078711212022921,
+      "grad_norm": 0.22910591959953308,
+      "learning_rate": 1.7003216722998112e-05,
+      "loss": 0.3841,
       "step": 85425
     },
     {
-      "epoch": 3.01,
-      "learning_rate": 1.8132028929353235e-05,
-      "loss": 0.2333,
+      "epoch": 3.078891411684146,
+      "grad_norm": 0.22995778918266296,
+      "learning_rate": 1.700045195967752e-05,
+      "loss": 0.4064,
       "step": 85430
     },
     {
-      "epoch": 3.01,
-      "learning_rate": 1.8129289858324875e-05,
-      "loss": 0.271,
+      "epoch": 3.0790716113453707,
+      "grad_norm": 0.24501994252204895,
+      "learning_rate": 1.6997687305349234e-05,
+      "loss": 0.3691,
       "step": 85435
     },
     {
-      "epoch": 3.01,
-      "learning_rate": 1.8126550876508163e-05,
-      "loss": 0.2516,
+      "epoch": 3.0792518110065954,
+      "grad_norm": 0.19479656219482422,
+      "learning_rate": 1.699492276005091e-05,
+      "loss": 0.3536,
       "step": 85440
     },
     {
-      "epoch": 3.01,
-      "learning_rate": 1.812381198393866e-05,
-      "loss": 0.2651,
+      "epoch": 3.07943201066782,
+      "grad_norm": 0.24624988436698914,
+      "learning_rate": 1.69921583238202e-05,
+      "loss": 0.3855,
       "step": 85445
     },
     {
-      "epoch": 3.01,
-      "learning_rate": 1.8121073180651915e-05,
-      "loss": 0.2639,
+      "epoch": 3.0796122103290444,
+      "grad_norm": 0.23105396330356598,
+      "learning_rate": 1.6989393996694797e-05,
+      "loss": 0.3399,
       "step": 85450
     },
     {
-      "epoch": 3.01,
-      "learning_rate": 1.811833446668351e-05,
-      "loss": 0.2613,
+      "epoch": 3.079792409990269,
+      "grad_norm": 0.23639746010303497,
+      "learning_rate": 1.6986629778712344e-05,
+      "loss": 0.4176,
       "step": 85455
     },
     {
-      "epoch": 3.01,
-      "learning_rate": 1.8115595842068995e-05,
-      "loss": 0.2641,
+      "epoch": 3.079972609651494,
+      "grad_norm": 0.27155548334121704,
+      "learning_rate": 1.6983865669910504e-05,
+      "loss": 0.3482,
       "step": 85460
     },
     {
-      "epoch": 3.01,
-      "learning_rate": 1.8112857306843933e-05,
-      "loss": 0.2608,
+      "epoch": 3.0801528093127186,
+      "grad_norm": 0.2704147398471832,
+      "learning_rate": 1.6981101670326945e-05,
+      "loss": 0.3908,
       "step": 85465
     },
     {
-      "epoch": 3.01,
-      "learning_rate": 1.811011886104387e-05,
-      "loss": 0.2452,
+      "epoch": 3.080333008973943,
+      "grad_norm": 0.21449273824691772,
+      "learning_rate": 1.697833777999932e-05,
+      "loss": 0.3481,
       "step": 85470
     },
     {
-      "epoch": 3.01,
-      "learning_rate": 1.8107380504704386e-05,
-      "loss": 0.2699,
+      "epoch": 3.0805132086351676,
+      "grad_norm": 0.2628542482852936,
+      "learning_rate": 1.6975573998965298e-05,
+      "loss": 0.3956,
       "step": 85475
     },
     {
-      "epoch": 3.01,
-      "learning_rate": 1.810464223786102e-05,
-      "loss": 0.2613,
+      "epoch": 3.0806934082963924,
+      "grad_norm": 0.20272031426429749,
+      "learning_rate": 1.697281032726252e-05,
+      "loss": 0.3353,
       "step": 85480
     },
     {
-      "epoch": 3.01,
-      "learning_rate": 1.8101904060549328e-05,
-      "loss": 0.271,
+      "epoch": 3.080873607957617,
+      "grad_norm": 0.20484726130962372,
+      "learning_rate": 1.697004676492865e-05,
+      "loss": 0.3335,
       "step": 85485
     },
     {
-      "epoch": 3.01,
-      "learning_rate": 1.809916597280486e-05,
-      "loss": 0.2667,
+      "epoch": 3.081053807618842,
+      "grad_norm": 0.21416746079921722,
+      "learning_rate": 1.696728331200134e-05,
+      "loss": 0.3687,
       "step": 85490
     },
     {
-      "epoch": 3.01,
-      "learning_rate": 1.8096427974663183e-05,
-      "loss": 0.2599,
+      "epoch": 3.081234007280066,
+      "grad_norm": 0.2598028779029846,
+      "learning_rate": 1.696451996851824e-05,
+      "loss": 0.3842,
       "step": 85495
     },
     {
-      "epoch": 3.01,
-      "learning_rate": 1.809369006615984e-05,
-      "loss": 0.2713,
+      "epoch": 3.081414206941291,
+      "grad_norm": 0.23536814749240875,
+      "learning_rate": 1.6961756734516994e-05,
+      "loss": 0.3904,
       "step": 85500
     },
     {
-      "epoch": 3.01,
-      "eval_loss": 0.2571566104888916,
-      "eval_runtime": 10.5571,
-      "eval_samples_per_second": 9.472,
-      "eval_steps_per_second": 9.472,
+      "epoch": 3.081414206941291,
+      "eval_loss": 0.4317238926887512,
+      "eval_runtime": 3.5362,
+      "eval_samples_per_second": 28.279,
+      "eval_steps_per_second": 7.07,
       "step": 85500
     },
     {
-      "epoch": 3.01,
-      "learning_rate": 1.8090952247330374e-05,
-      "loss": 0.2468,
+      "epoch": 3.0815944066025156,
+      "grad_norm": 0.17988234758377075,
+      "learning_rate": 1.695899361003526e-05,
+      "loss": 0.3702,
       "step": 85505
     },
     {
-      "epoch": 3.01,
-      "learning_rate": 1.8088214518210345e-05,
-      "loss": 0.2664,
+      "epoch": 3.0817746062637403,
+      "grad_norm": 0.22040648758411407,
+      "learning_rate": 1.695623059511068e-05,
+      "loss": 0.3716,
       "step": 85510
     },
     {
-      "epoch": 3.01,
-      "learning_rate": 1.808547687883529e-05,
-      "loss": 0.2476,
+      "epoch": 3.081954805924965,
+      "grad_norm": 0.23150260746479034,
+      "learning_rate": 1.69534676897809e-05,
+      "loss": 0.3891,
       "step": 85515
     },
     {
-      "epoch": 3.01,
-      "learning_rate": 1.808273932924077e-05,
-      "loss": 0.2681,
+      "epoch": 3.0821350055861894,
+      "grad_norm": 0.21168793737888336,
+      "learning_rate": 1.6950704894083575e-05,
+      "loss": 0.363,
       "step": 85520
     },
     {
-      "epoch": 3.01,
-      "learning_rate": 1.8080001869462308e-05,
-      "loss": 0.2499,
+      "epoch": 3.082315205247414,
+      "grad_norm": 0.21894006431102753,
+      "learning_rate": 1.6947942208056316e-05,
+      "loss": 0.3604,
       "step": 85525
     },
     {
-      "epoch": 3.01,
-      "learning_rate": 1.807726449953547e-05,
-      "loss": 0.2518,
+      "epoch": 3.082495404908639,
+      "grad_norm": 0.23681262135505676,
+      "learning_rate": 1.6945179631736807e-05,
+      "loss": 0.3973,
       "step": 85530
     },
     {
-      "epoch": 3.01,
-      "learning_rate": 1.8074527219495793e-05,
-      "loss": 0.2667,
+      "epoch": 3.0826756045698636,
+      "grad_norm": 0.2256009578704834,
+      "learning_rate": 1.6942417165162648e-05,
+      "loss": 0.3661,
       "step": 85535
     },
     {
-      "epoch": 3.01,
-      "learning_rate": 1.807179002937881e-05,
-      "loss": 0.268,
+      "epoch": 3.082855804231088,
+      "grad_norm": 0.21623054146766663,
+      "learning_rate": 1.6939654808371515e-05,
+      "loss": 0.3936,
       "step": 85540
     },
     {
-      "epoch": 3.01,
-      "learning_rate": 1.8069052929220064e-05,
-      "loss": 0.2868,
+      "epoch": 3.0830360038923126,
+      "grad_norm": 0.2143433690071106,
+      "learning_rate": 1.693689256140101e-05,
+      "loss": 0.3677,
       "step": 85545
     },
     {
-      "epoch": 3.01,
-      "learning_rate": 1.8066315919055106e-05,
-      "loss": 0.2774,
+      "epoch": 3.0832162035535373,
+      "grad_norm": 0.2282250076532364,
+      "learning_rate": 1.693413042428878e-05,
+      "loss": 0.3712,
       "step": 85550
     },
     {
-      "epoch": 3.01,
-      "learning_rate": 1.8063578998919467e-05,
-      "loss": 0.2597,
+      "epoch": 3.083396403214762,
+      "grad_norm": 0.19568027555942535,
+      "learning_rate": 1.693136839707248e-05,
+      "loss": 0.3908,
       "step": 85555
     },
     {
-      "epoch": 3.01,
-      "learning_rate": 1.8060842168848686e-05,
-      "loss": 0.269,
+      "epoch": 3.083576602875987,
+      "grad_norm": 0.20706535875797272,
+      "learning_rate": 1.692860647978971e-05,
+      "loss": 0.388,
       "step": 85560
     },
     {
-      "epoch": 3.01,
-      "learning_rate": 1.805810542887828e-05,
-      "loss": 0.2514,
+      "epoch": 3.083756802537211,
+      "grad_norm": 0.23546600341796875,
+      "learning_rate": 1.6925844672478115e-05,
+      "loss": 0.3772,
       "step": 85565
     },
     {
-      "epoch": 3.01,
-      "learning_rate": 1.805536877904381e-05,
-      "loss": 0.2565,
+      "epoch": 3.083937002198436,
+      "grad_norm": 0.2380140721797943,
+      "learning_rate": 1.6923082975175325e-05,
+      "loss": 0.3468,
       "step": 85570
     },
     {
-      "epoch": 3.01,
-      "learning_rate": 1.8052632219380806e-05,
-      "loss": 0.2527,
+      "epoch": 3.0841172018596605,
+      "grad_norm": 0.23840586841106415,
+      "learning_rate": 1.6920321387918957e-05,
+      "loss": 0.3679,
       "step": 85575
     },
     {
-      "epoch": 3.01,
-      "learning_rate": 1.804989574992479e-05,
-      "loss": 0.2334,
+      "epoch": 3.0842974015208853,
+      "grad_norm": 0.2252638339996338,
+      "learning_rate": 1.6917559910746657e-05,
+      "loss": 0.4021,
       "step": 85580
     },
     {
-      "epoch": 3.01,
-      "learning_rate": 1.804715937071129e-05,
-      "loss": 0.2542,
+      "epoch": 3.0844776011821096,
+      "grad_norm": 0.20066937804222107,
+      "learning_rate": 1.6914798543696036e-05,
+      "loss": 0.4079,
       "step": 85585
     },
     {
-      "epoch": 3.01,
-      "learning_rate": 1.8044423081775853e-05,
-      "loss": 0.2429,
+      "epoch": 3.0846578008433343,
+      "grad_norm": 0.22038015723228455,
+      "learning_rate": 1.6912037286804717e-05,
+      "loss": 0.3824,
       "step": 85590
     },
     {
-      "epoch": 3.01,
-      "learning_rate": 1.8041686883153992e-05,
-      "loss": 0.2496,
+      "epoch": 3.084838000504559,
+      "grad_norm": 0.21279798448085785,
+      "learning_rate": 1.6909276140110324e-05,
+      "loss": 0.3655,
       "step": 85595
     },
     {
-      "epoch": 3.01,
-      "learning_rate": 1.8038950774881245e-05,
-      "loss": 0.2554,
+      "epoch": 3.0850182001657838,
+      "grad_norm": 0.24839600920677185,
+      "learning_rate": 1.6906515103650482e-05,
+      "loss": 0.3607,
       "step": 85600
     },
     {
-      "epoch": 3.01,
-      "learning_rate": 1.803621475699313e-05,
-      "loss": 0.239,
+      "epoch": 3.0851983998270085,
+      "grad_norm": 0.23541690409183502,
+      "learning_rate": 1.69037541774628e-05,
+      "loss": 0.3618,
       "step": 85605
     },
     {
-      "epoch": 3.01,
-      "learning_rate": 1.8033478829525185e-05,
-      "loss": 0.2629,
+      "epoch": 3.085378599488233,
+      "grad_norm": 0.24776875972747803,
+      "learning_rate": 1.69009933615849e-05,
+      "loss": 0.3814,
       "step": 85610
     },
     {
-      "epoch": 3.01,
-      "learning_rate": 1.8030742992512924e-05,
-      "loss": 0.258,
+      "epoch": 3.0855587991494575,
+      "grad_norm": 0.19632692635059357,
+      "learning_rate": 1.6898232656054398e-05,
+      "loss": 0.413,
       "step": 85615
     },
     {
-      "epoch": 3.01,
-      "learning_rate": 1.8028007245991863e-05,
-      "loss": 0.2504,
+      "epoch": 3.0857389988106823,
+      "grad_norm": 0.24694256484508514,
+      "learning_rate": 1.6895472060908907e-05,
+      "loss": 0.387,
       "step": 85620
     },
     {
-      "epoch": 3.01,
-      "learning_rate": 1.802527158999754e-05,
-      "loss": 0.2538,
+      "epoch": 3.085919198471907,
+      "grad_norm": 0.21106529235839844,
+      "learning_rate": 1.689271157618605e-05,
+      "loss": 0.3476,
       "step": 85625
     },
     {
-      "epoch": 3.01,
-      "learning_rate": 1.8022536024565474e-05,
-      "loss": 0.251,
+      "epoch": 3.0860993981331317,
+      "grad_norm": 0.24675370752811432,
+      "learning_rate": 1.688995120192341e-05,
+      "loss": 0.396,
       "step": 85630
     },
     {
-      "epoch": 3.01,
-      "learning_rate": 1.801980054973118e-05,
-      "loss": 0.2519,
+      "epoch": 3.086279597794356,
+      "grad_norm": 0.20427581667900085,
+      "learning_rate": 1.688719093815863e-05,
+      "loss": 0.3869,
       "step": 85635
     },
     {
-      "epoch": 3.01,
-      "learning_rate": 1.8017065165530162e-05,
-      "loss": 0.2593,
+      "epoch": 3.0864597974555807,
+      "grad_norm": 0.2514338493347168,
+      "learning_rate": 1.68844307849293e-05,
+      "loss": 0.3824,
       "step": 85640
     },
     {
-      "epoch": 3.01,
-      "learning_rate": 1.8014329871997963e-05,
-      "loss": 0.3128,
+      "epoch": 3.0866399971168055,
+      "grad_norm": 0.25234454870224,
+      "learning_rate": 1.6881670742273023e-05,
+      "loss": 0.3886,
       "step": 85645
     },
     {
-      "epoch": 3.01,
-      "learning_rate": 1.8011594669170078e-05,
-      "loss": 0.2533,
+      "epoch": 3.08682019677803,
+      "grad_norm": 0.26698270440101624,
+      "learning_rate": 1.6878910810227417e-05,
+      "loss": 0.3554,
       "step": 85650
     },
     {
-      "epoch": 3.01,
-      "learning_rate": 1.800885955708204e-05,
-      "loss": 0.2519,
+      "epoch": 3.0870003964392545,
+      "grad_norm": 0.21970665454864502,
+      "learning_rate": 1.687615098883007e-05,
+      "loss": 0.3698,
       "step": 85655
     },
     {
-      "epoch": 3.01,
-      "learning_rate": 1.8006124535769345e-05,
-      "loss": 0.2527,
+      "epoch": 3.0871805961004792,
+      "grad_norm": 0.2174971103668213,
+      "learning_rate": 1.68733912781186e-05,
+      "loss": 0.3598,
       "step": 85660
     },
     {
-      "epoch": 3.01,
-      "learning_rate": 1.800338960526752e-05,
-      "loss": 0.2399,
+      "epoch": 3.087360795761704,
+      "grad_norm": 0.26610609889030457,
+      "learning_rate": 1.68706316781306e-05,
+      "loss": 0.3654,
       "step": 85665
     },
     {
-      "epoch": 3.01,
-      "learning_rate": 1.8000654765612074e-05,
-      "loss": 0.2729,
+      "epoch": 3.0875409954229287,
+      "grad_norm": 0.22846719622612,
+      "learning_rate": 1.6867872188903667e-05,
+      "loss": 0.3823,
       "step": 85670
     },
     {
-      "epoch": 3.01,
-      "learning_rate": 1.7997920016838504e-05,
-      "loss": 0.2545,
+      "epoch": 3.0877211950841534,
+      "grad_norm": 0.23386985063552856,
+      "learning_rate": 1.6865112810475403e-05,
+      "loss": 0.3609,
       "step": 85675
     },
     {
-      "epoch": 3.01,
-      "learning_rate": 1.7995185358982323e-05,
-      "loss": 0.2372,
+      "epoch": 3.0879013947453777,
+      "grad_norm": 0.19239924848079681,
+      "learning_rate": 1.6862353542883404e-05,
+      "loss": 0.3711,
       "step": 85680
     },
     {
-      "epoch": 3.01,
-      "learning_rate": 1.7992450792079056e-05,
-      "loss": 0.2608,
+      "epoch": 3.0880815944066025,
+      "grad_norm": 0.18704915046691895,
+      "learning_rate": 1.6859594386165255e-05,
+      "loss": 0.3522,
       "step": 85685
     },
     {
-      "epoch": 3.01,
-      "learning_rate": 1.7989716316164198e-05,
-      "loss": 0.2648,
+      "epoch": 3.088261794067827,
+      "grad_norm": 0.2616778016090393,
+      "learning_rate": 1.6856835340358563e-05,
+      "loss": 0.3941,
       "step": 85690
     },
     {
-      "epoch": 3.01,
-      "learning_rate": 1.7986981931273254e-05,
-      "loss": 0.2433,
+      "epoch": 3.088441993729052,
+      "grad_norm": 0.21047228574752808,
+      "learning_rate": 1.685407640550091e-05,
+      "loss": 0.3471,
       "step": 85695
     },
     {
-      "epoch": 3.02,
-      "learning_rate": 1.7984247637441714e-05,
-      "loss": 0.2697,
+      "epoch": 3.088622193390276,
+      "grad_norm": 0.23516976833343506,
+      "learning_rate": 1.685131758162989e-05,
+      "loss": 0.3674,
       "step": 85700
     },
     {
-      "epoch": 3.02,
-      "learning_rate": 1.79815134347051e-05,
-      "loss": 0.2455,
+      "epoch": 3.088802393051501,
+      "grad_norm": 0.23034684360027313,
+      "learning_rate": 1.6848558868783098e-05,
+      "loss": 0.394,
       "step": 85705
     },
     {
-      "epoch": 3.02,
-      "learning_rate": 1.797877932309892e-05,
-      "loss": 0.2586,
+      "epoch": 3.0889825927127257,
+      "grad_norm": 0.2383045256137848,
+      "learning_rate": 1.6845800266998098e-05,
+      "loss": 0.3637,
       "step": 85710
     },
     {
-      "epoch": 3.02,
-      "learning_rate": 1.7976045302658657e-05,
-      "loss": 0.282,
+      "epoch": 3.0891627923739504,
+      "grad_norm": 0.21178589761257172,
+      "learning_rate": 1.6843041776312503e-05,
+      "loss": 0.364,
       "step": 85715
     },
     {
-      "epoch": 3.02,
-      "learning_rate": 1.797331137341981e-05,
-      "loss": 0.2418,
+      "epoch": 3.089342992035175,
+      "grad_norm": 0.2509348392486572,
+      "learning_rate": 1.6840283396763872e-05,
+      "loss": 0.3614,
       "step": 85720
     },
     {
-      "epoch": 3.02,
-      "learning_rate": 1.7970577535417892e-05,
-      "loss": 0.2651,
+      "epoch": 3.0895231916963994,
+      "grad_norm": 0.24692773818969727,
+      "learning_rate": 1.683752512838981e-05,
+      "loss": 0.3893,
       "step": 85725
     },
     {
-      "epoch": 3.02,
-      "learning_rate": 1.7967843788688388e-05,
-      "loss": 0.2649,
+      "epoch": 3.089703391357624,
+      "grad_norm": 0.19897426664829254,
+      "learning_rate": 1.6834766971227893e-05,
+      "loss": 0.3856,
       "step": 85730
     },
     {
-      "epoch": 3.02,
-      "learning_rate": 1.7965110133266806e-05,
-      "loss": 0.2598,
+      "epoch": 3.089883591018849,
+      "grad_norm": 0.21682973206043243,
+      "learning_rate": 1.6832008925315677e-05,
+      "loss": 0.3699,
       "step": 85735
     },
     {
-      "epoch": 3.02,
-      "learning_rate": 1.7962376569188617e-05,
-      "loss": 0.2503,
+      "epoch": 3.0900637906800736,
+      "grad_norm": 0.27642059326171875,
+      "learning_rate": 1.682925099069078e-05,
+      "loss": 0.3887,
       "step": 85740
     },
     {
-      "epoch": 3.02,
-      "learning_rate": 1.7959643096489342e-05,
-      "loss": 0.2473,
+      "epoch": 3.090243990341298,
+      "grad_norm": 0.23973903059959412,
+      "learning_rate": 1.6826493167390746e-05,
+      "loss": 0.3442,
       "step": 85745
     },
     {
-      "epoch": 3.02,
-      "learning_rate": 1.795690971520447e-05,
-      "loss": 0.2483,
+      "epoch": 3.0904241900025227,
+      "grad_norm": 0.21967796981334686,
+      "learning_rate": 1.6823735455453155e-05,
+      "loss": 0.4142,
       "step": 85750
     },
     {
-      "epoch": 3.02,
-      "learning_rate": 1.7954176425369463e-05,
-      "loss": 0.2506,
+      "epoch": 3.0906043896637474,
+      "grad_norm": 0.1877850890159607,
+      "learning_rate": 1.6820977854915593e-05,
+      "loss": 0.3623,
       "step": 85755
     },
     {
-      "epoch": 3.02,
-      "learning_rate": 1.7951443227019843e-05,
-      "loss": 0.2643,
+      "epoch": 3.090784589324972,
+      "grad_norm": 0.22713260352611542,
+      "learning_rate": 1.681822036581562e-05,
+      "loss": 0.3917,
       "step": 85760
     },
     {
-      "epoch": 3.02,
-      "learning_rate": 1.7948710120191086e-05,
-      "loss": 0.257,
+      "epoch": 3.090964788986197,
+      "grad_norm": 0.2456943690776825,
+      "learning_rate": 1.681546298819081e-05,
+      "loss": 0.3744,
       "step": 85765
     },
     {
-      "epoch": 3.02,
-      "learning_rate": 1.794597710491869e-05,
-      "loss": 0.2715,
+      "epoch": 3.091144988647421,
+      "grad_norm": 0.2482386976480484,
+      "learning_rate": 1.6812705722078738e-05,
+      "loss": 0.3686,
       "step": 85770
     },
     {
-      "epoch": 3.02,
-      "learning_rate": 1.794324418123811e-05,
-      "loss": 0.2561,
+      "epoch": 3.091325188308646,
+      "grad_norm": 0.28733116388320923,
+      "learning_rate": 1.6809948567516955e-05,
+      "loss": 0.3872,
       "step": 85775
     },
     {
-      "epoch": 3.02,
-      "learning_rate": 1.7940511349184873e-05,
-      "loss": 0.2645,
+      "epoch": 3.0915053879698706,
+      "grad_norm": 0.266248881816864,
+      "learning_rate": 1.6807191524543045e-05,
+      "loss": 0.3957,
       "step": 85780
     },
     {
-      "epoch": 3.02,
-      "learning_rate": 1.7937778608794437e-05,
-      "loss": 0.2468,
+      "epoch": 3.0916855876310954,
+      "grad_norm": 0.21898359060287476,
+      "learning_rate": 1.6804434593194565e-05,
+      "loss": 0.3749,
       "step": 85785
     },
     {
-      "epoch": 3.02,
-      "learning_rate": 1.7935045960102298e-05,
-      "loss": 0.2588,
+      "epoch": 3.09186578729232,
+      "grad_norm": 0.22204652428627014,
+      "learning_rate": 1.6801677773509074e-05,
+      "loss": 0.3562,
       "step": 85790
     },
     {
-      "epoch": 3.02,
-      "learning_rate": 1.793231340314392e-05,
-      "loss": 0.2506,
+      "epoch": 3.0920459869535444,
+      "grad_norm": 0.228468656539917,
+      "learning_rate": 1.6798921065524138e-05,
+      "loss": 0.3875,
       "step": 85795
     },
     {
-      "epoch": 3.02,
-      "learning_rate": 1.7929580937954803e-05,
-      "loss": 0.2624,
+      "epoch": 3.092226186614769,
+      "grad_norm": 0.2888953685760498,
+      "learning_rate": 1.679616446927731e-05,
+      "loss": 0.3727,
       "step": 85800
     },
     {
-      "epoch": 3.02,
-      "learning_rate": 1.792684856457042e-05,
-      "loss": 0.2708,
+      "epoch": 3.092406386275994,
+      "grad_norm": 0.2353111207485199,
+      "learning_rate": 1.6793407984806153e-05,
+      "loss": 0.4015,
       "step": 85805
     },
     {
-      "epoch": 3.02,
-      "learning_rate": 1.7924116283026244e-05,
-      "loss": 0.2593,
+      "epoch": 3.0925865859372186,
+      "grad_norm": 0.21720938384532928,
+      "learning_rate": 1.6790651612148235e-05,
+      "loss": 0.3622,
       "step": 85810
     },
     {
-      "epoch": 3.02,
-      "learning_rate": 1.7921384093357753e-05,
-      "loss": 0.2432,
+      "epoch": 3.092766785598443,
+      "grad_norm": 0.25723299384117126,
+      "learning_rate": 1.6787895351341083e-05,
+      "loss": 0.4135,
       "step": 85815
     },
     {
-      "epoch": 3.02,
-      "learning_rate": 1.7918651995600423e-05,
-      "loss": 0.2573,
+      "epoch": 3.0929469852596676,
+      "grad_norm": 0.2206987887620926,
+      "learning_rate": 1.678513920242228e-05,
+      "loss": 0.4013,
       "step": 85820
     },
     {
-      "epoch": 3.02,
-      "learning_rate": 1.7915919989789733e-05,
-      "loss": 0.2652,
+      "epoch": 3.0931271849208923,
+      "grad_norm": 0.23632338643074036,
+      "learning_rate": 1.6782383165429364e-05,
+      "loss": 0.3843,
       "step": 85825
     },
     {
-      "epoch": 3.02,
-      "learning_rate": 1.7913188075961156e-05,
-      "loss": 0.2361,
+      "epoch": 3.093307384582117,
+      "grad_norm": 0.19661295413970947,
+      "learning_rate": 1.677962724039987e-05,
+      "loss": 0.3777,
       "step": 85830
     },
     {
-      "epoch": 3.02,
-      "learning_rate": 1.7910456254150152e-05,
-      "loss": 0.242,
+      "epoch": 3.093487584243342,
+      "grad_norm": 0.2105666697025299,
+      "learning_rate": 1.6776871427371384e-05,
+      "loss": 0.3652,
       "step": 85835
     },
     {
-      "epoch": 3.02,
-      "learning_rate": 1.790772452439221e-05,
-      "loss": 0.2871,
+      "epoch": 3.093667783904566,
+      "grad_norm": 0.21203243732452393,
+      "learning_rate": 1.677411572638142e-05,
+      "loss": 0.3687,
       "step": 85840
     },
     {
-      "epoch": 3.02,
-      "learning_rate": 1.7904992886722787e-05,
-      "loss": 0.2457,
+      "epoch": 3.093847983565791,
+      "grad_norm": 0.20673683285713196,
+      "learning_rate": 1.6771360137467546e-05,
+      "loss": 0.3816,
       "step": 85845
     },
     {
-      "epoch": 3.02,
-      "learning_rate": 1.7902261341177363e-05,
-      "loss": 0.2623,
+      "epoch": 3.0940281832270156,
+      "grad_norm": 0.23549970984458923,
+      "learning_rate": 1.6768604660667293e-05,
+      "loss": 0.3927,
       "step": 85850
     },
     {
-      "epoch": 3.02,
-      "learning_rate": 1.7899529887791386e-05,
-      "loss": 0.2535,
+      "epoch": 3.0942083828882403,
+      "grad_norm": 0.18549276888370514,
+      "learning_rate": 1.6765849296018203e-05,
+      "loss": 0.3757,
       "step": 85855
     },
     {
-      "epoch": 3.02,
-      "learning_rate": 1.7896798526600345e-05,
-      "loss": 0.2648,
+      "epoch": 3.094388582549465,
+      "grad_norm": 0.2178041934967041,
+      "learning_rate": 1.6763094043557825e-05,
+      "loss": 0.3638,
       "step": 85860
     },
     {
-      "epoch": 3.02,
-      "learning_rate": 1.7894067257639686e-05,
-      "loss": 0.2569,
+      "epoch": 3.0945687822106893,
+      "grad_norm": 0.25327563285827637,
+      "learning_rate": 1.67603389033237e-05,
+      "loss": 0.3797,
       "step": 85865
     },
     {
-      "epoch": 3.02,
-      "learning_rate": 1.7891336080944882e-05,
-      "loss": 0.2785,
+      "epoch": 3.094748981871914,
+      "grad_norm": 0.24706877768039703,
+      "learning_rate": 1.6757583875353355e-05,
+      "loss": 0.3832,
       "step": 85870
     },
     {
-      "epoch": 3.02,
-      "learning_rate": 1.7888604996551394e-05,
-      "loss": 0.2462,
+      "epoch": 3.094929181533139,
+      "grad_norm": 0.20161230862140656,
+      "learning_rate": 1.675482895968434e-05,
+      "loss": 0.3857,
       "step": 85875
     },
     {
-      "epoch": 3.02,
-      "learning_rate": 1.7885874004494692e-05,
-      "loss": 0.265,
+      "epoch": 3.0951093811943635,
+      "grad_norm": 0.23628760874271393,
+      "learning_rate": 1.675207415635418e-05,
+      "loss": 0.3709,
       "step": 85880
     },
     {
-      "epoch": 3.02,
-      "learning_rate": 1.7883143104810223e-05,
-      "loss": 0.2593,
+      "epoch": 3.095289580855588,
+      "grad_norm": 0.2771933376789093,
+      "learning_rate": 1.6749319465400414e-05,
+      "loss": 0.3934,
       "step": 85885
     },
     {
-      "epoch": 3.02,
-      "learning_rate": 1.7880412297533444e-05,
-      "loss": 0.2675,
+      "epoch": 3.0954697805168125,
+      "grad_norm": 0.2511608898639679,
+      "learning_rate": 1.6746564886860577e-05,
+      "loss": 0.4084,
       "step": 85890
     },
     {
-      "epoch": 3.02,
-      "learning_rate": 1.7877681582699827e-05,
-      "loss": 0.2416,
+      "epoch": 3.0956499801780373,
+      "grad_norm": 0.2165483683347702,
+      "learning_rate": 1.6743810420772186e-05,
+      "loss": 0.3714,
       "step": 85895
     },
     {
-      "epoch": 3.02,
-      "learning_rate": 1.7874950960344818e-05,
-      "loss": 0.2507,
+      "epoch": 3.095830179839262,
+      "grad_norm": 0.22518706321716309,
+      "learning_rate": 1.674105606717279e-05,
+      "loss": 0.375,
       "step": 85900
     },
     {
-      "epoch": 3.02,
-      "learning_rate": 1.7872220430503878e-05,
-      "loss": 0.2235,
+      "epoch": 3.0960103795004867,
+      "grad_norm": 0.21256521344184875,
+      "learning_rate": 1.67383018260999e-05,
+      "loss": 0.3688,
       "step": 85905
     },
     {
-      "epoch": 3.02,
-      "learning_rate": 1.7869489993212445e-05,
-      "loss": 0.2617,
+      "epoch": 3.096190579161711,
+      "grad_norm": 0.2093464583158493,
+      "learning_rate": 1.673554769759105e-05,
+      "loss": 0.3451,
       "step": 85910
     },
     {
-      "epoch": 3.02,
-      "learning_rate": 1.7866759648506e-05,
-      "loss": 0.2712,
+      "epoch": 3.0963707788229358,
+      "grad_norm": 0.16468511521816254,
+      "learning_rate": 1.673279368168377e-05,
+      "loss": 0.4124,
       "step": 85915
     },
     {
-      "epoch": 3.02,
-      "learning_rate": 1.786402939641998e-05,
-      "loss": 0.2656,
+      "epoch": 3.0965509784841605,
+      "grad_norm": 0.21614669263362885,
+      "learning_rate": 1.673003977841556e-05,
+      "loss": 0.3886,
       "step": 85920
     },
     {
-      "epoch": 3.02,
-      "learning_rate": 1.7861299236989832e-05,
-      "loss": 0.2583,
+      "epoch": 3.0967311781453852,
+      "grad_norm": 0.25867360830307007,
+      "learning_rate": 1.672728598782397e-05,
+      "loss": 0.388,
       "step": 85925
     },
     {
-      "epoch": 3.02,
-      "learning_rate": 1.7858569170251e-05,
-      "loss": 0.2673,
+      "epoch": 3.0969113778066095,
+      "grad_norm": 0.25877588987350464,
+      "learning_rate": 1.6724532309946506e-05,
+      "loss": 0.4002,
       "step": 85930
     },
     {
-      "epoch": 3.02,
-      "learning_rate": 1.7855839196238956e-05,
-      "loss": 0.2758,
+      "epoch": 3.0970915774678343,
+      "grad_norm": 0.22398416697978973,
+      "learning_rate": 1.6721778744820678e-05,
+      "loss": 0.4284,
       "step": 85935
     },
     {
-      "epoch": 3.02,
-      "learning_rate": 1.7853109314989125e-05,
-      "loss": 0.2585,
+      "epoch": 3.097271777129059,
+      "grad_norm": 0.2191021740436554,
+      "learning_rate": 1.671902529248402e-05,
+      "loss": 0.4128,
       "step": 85940
     },
     {
-      "epoch": 3.02,
-      "learning_rate": 1.7850379526536958e-05,
-      "loss": 0.2618,
+      "epoch": 3.0974519767902837,
+      "grad_norm": 0.23091648519039154,
+      "learning_rate": 1.6716271952974033e-05,
+      "loss": 0.3747,
       "step": 85945
     },
     {
-      "epoch": 3.02,
-      "learning_rate": 1.78476498309179e-05,
-      "loss": 0.2605,
+      "epoch": 3.0976321764515085,
+      "grad_norm": 0.2016901671886444,
+      "learning_rate": 1.671351872632824e-05,
+      "loss": 0.3443,
       "step": 85950
     },
     {
-      "epoch": 3.02,
-      "learning_rate": 1.7844920228167397e-05,
-      "loss": 0.2557,
+      "epoch": 3.0978123761127327,
+      "grad_norm": 0.2278120517730713,
+      "learning_rate": 1.671076561258415e-05,
+      "loss": 0.4025,
       "step": 85955
     },
     {
-      "epoch": 3.02,
-      "learning_rate": 1.7842190718320896e-05,
-      "loss": 0.2569,
+      "epoch": 3.0979925757739575,
+      "grad_norm": 0.18883484601974487,
+      "learning_rate": 1.6708012611779273e-05,
+      "loss": 0.3807,
       "step": 85960
     },
     {
-      "epoch": 3.02,
-      "learning_rate": 1.783946130141383e-05,
-      "loss": 0.2488,
+      "epoch": 3.098172775435182,
+      "grad_norm": 0.2371702492237091,
+      "learning_rate": 1.670525972395112e-05,
+      "loss": 0.3861,
       "step": 85965
     },
     {
-      "epoch": 3.02,
-      "learning_rate": 1.7836731977481628e-05,
-      "loss": 0.2511,
+      "epoch": 3.098352975096407,
+      "grad_norm": 0.2216133028268814,
+      "learning_rate": 1.6702506949137197e-05,
+      "loss": 0.3951,
       "step": 85970
     },
     {
-      "epoch": 3.02,
-      "learning_rate": 1.7834002746559754e-05,
-      "loss": 0.2645,
+      "epoch": 3.0985331747576312,
+      "grad_norm": 0.25640398263931274,
+      "learning_rate": 1.669975428737501e-05,
+      "loss": 0.3929,
       "step": 85975
     },
     {
-      "epoch": 3.03,
-      "learning_rate": 1.783127360868362e-05,
-      "loss": 0.2593,
+      "epoch": 3.098713374418856,
+      "grad_norm": 0.22617505490779877,
+      "learning_rate": 1.669700173870206e-05,
+      "loss": 0.4057,
       "step": 85980
     },
     {
-      "epoch": 3.03,
-      "learning_rate": 1.7828544563888675e-05,
-      "loss": 0.2706,
+      "epoch": 3.0988935740800807,
+      "grad_norm": 0.25993672013282776,
+      "learning_rate": 1.6694249303155857e-05,
+      "loss": 0.3954,
       "step": 85985
     },
     {
-      "epoch": 3.03,
-      "learning_rate": 1.782581561221036e-05,
-      "loss": 0.2762,
+      "epoch": 3.0990737737413054,
+      "grad_norm": 0.23373205959796906,
+      "learning_rate": 1.6691496980773903e-05,
+      "loss": 0.4111,
       "step": 85990
     },
     {
-      "epoch": 3.03,
-      "learning_rate": 1.7823086753684103e-05,
-      "loss": 0.2745,
+      "epoch": 3.09925397340253,
+      "grad_norm": 0.19884000718593597,
+      "learning_rate": 1.668874477159369e-05,
+      "loss": 0.4071,
       "step": 85995
     },
     {
-      "epoch": 3.03,
-      "learning_rate": 1.7820357988345336e-05,
-      "loss": 0.2397,
+      "epoch": 3.0994341730637545,
+      "grad_norm": 0.23313197493553162,
+      "learning_rate": 1.6685992675652717e-05,
+      "loss": 0.3718,
       "step": 86000
     },
     {
-      "epoch": 3.03,
-      "eval_loss": 0.2575805187225342,
-      "eval_runtime": 10.5481,
-      "eval_samples_per_second": 9.48,
-      "eval_steps_per_second": 9.48,
+      "epoch": 3.0994341730637545,
+      "eval_loss": 0.4313110411167145,
+      "eval_runtime": 3.5318,
+      "eval_samples_per_second": 28.314,
+      "eval_steps_per_second": 7.079,
       "step": 86000
     },
     {
-      "epoch": 3.03,
-      "learning_rate": 1.7817629316229474e-05,
-      "loss": 0.2688,
+      "epoch": 3.099614372724979,
+      "grad_norm": 0.22738061845302582,
+      "learning_rate": 1.6683240692988492e-05,
+      "loss": 0.3886,
       "step": 86005
     },
     {
-      "epoch": 3.03,
-      "learning_rate": 1.781490073737197e-05,
-      "loss": 0.2693,
+      "epoch": 3.099794572386204,
+      "grad_norm": 0.21752703189849854,
+      "learning_rate": 1.6680488823638508e-05,
+      "loss": 0.4185,
       "step": 86010
     },
     {
-      "epoch": 3.03,
-      "learning_rate": 1.781217225180825e-05,
-      "loss": 0.2262,
+      "epoch": 3.0999747720474287,
+      "grad_norm": 0.22272847592830658,
+      "learning_rate": 1.6677737067640232e-05,
+      "loss": 0.3903,
       "step": 86015
     },
     {
-      "epoch": 3.03,
-      "learning_rate": 1.7809443859573738e-05,
-      "loss": 0.2576,
+      "epoch": 3.1001549717086534,
+      "grad_norm": 0.2768046259880066,
+      "learning_rate": 1.6674985425031197e-05,
+      "loss": 0.3713,
       "step": 86020
     },
     {
-      "epoch": 3.03,
-      "learning_rate": 1.7806715560703845e-05,
-      "loss": 0.2587,
+      "epoch": 3.1003351713698777,
+      "grad_norm": 0.23231996595859528,
+      "learning_rate": 1.667223389584886e-05,
+      "loss": 0.3758,
       "step": 86025
     },
     {
-      "epoch": 3.03,
-      "learning_rate": 1.7803987355234023e-05,
-      "loss": 0.2747,
+      "epoch": 3.1005153710311024,
+      "grad_norm": 0.21626095473766327,
+      "learning_rate": 1.6669482480130734e-05,
+      "loss": 0.3965,
       "step": 86030
     },
     {
-      "epoch": 3.03,
-      "learning_rate": 1.7801259243199676e-05,
-      "loss": 0.2475,
+      "epoch": 3.100695570692327,
+      "grad_norm": 0.2107696682214737,
+      "learning_rate": 1.6666731177914292e-05,
+      "loss": 0.3717,
       "step": 86035
     },
     {
-      "epoch": 3.03,
-      "learning_rate": 1.779853122463624e-05,
-      "loss": 0.2707,
+      "epoch": 3.100875770353552,
+      "grad_norm": 0.24785815179347992,
+      "learning_rate": 1.666397998923702e-05,
+      "loss": 0.3698,
       "step": 86040
     },
     {
-      "epoch": 3.03,
-      "learning_rate": 1.7795803299579116e-05,
-      "loss": 0.2519,
+      "epoch": 3.101055970014776,
+      "grad_norm": 0.21663641929626465,
+      "learning_rate": 1.6661228914136414e-05,
+      "loss": 0.3634,
       "step": 86045
     },
     {
-      "epoch": 3.03,
-      "learning_rate": 1.7793075468063755e-05,
-      "loss": 0.2443,
+      "epoch": 3.101236169676001,
+      "grad_norm": 0.23445236682891846,
+      "learning_rate": 1.6658477952649946e-05,
+      "loss": 0.387,
       "step": 86050
     },
     {
-      "epoch": 3.03,
-      "learning_rate": 1.7790347730125555e-05,
-      "loss": 0.2526,
+      "epoch": 3.1014163693372256,
+      "grad_norm": 0.20155306160449982,
+      "learning_rate": 1.6655727104815104e-05,
+      "loss": 0.3826,
       "step": 86055
     },
     {
-      "epoch": 3.03,
-      "learning_rate": 1.7787620085799935e-05,
-      "loss": 0.2438,
+      "epoch": 3.1015965689984504,
+      "grad_norm": 0.20625048875808716,
+      "learning_rate": 1.6652976370669362e-05,
+      "loss": 0.388,
       "step": 86060
     },
     {
-      "epoch": 3.03,
-      "learning_rate": 1.7784892535122312e-05,
-      "loss": 0.2435,
+      "epoch": 3.101776768659675,
+      "grad_norm": 0.2746394872665405,
+      "learning_rate": 1.6650225750250197e-05,
+      "loss": 0.3988,
       "step": 86065
     },
     {
-      "epoch": 3.03,
-      "learning_rate": 1.778216507812811e-05,
-      "loss": 0.2692,
+      "epoch": 3.1019569683208994,
+      "grad_norm": 0.21470344066619873,
+      "learning_rate": 1.6647475243595096e-05,
+      "loss": 0.3931,
       "step": 86070
     },
     {
-      "epoch": 3.03,
-      "learning_rate": 1.777943771485274e-05,
-      "loss": 0.2492,
+      "epoch": 3.102137167982124,
+      "grad_norm": 0.22546519339084625,
+      "learning_rate": 1.6644724850741528e-05,
+      "loss": 0.3552,
       "step": 86075
     },
     {
-      "epoch": 3.03,
-      "learning_rate": 1.7776710445331613e-05,
-      "loss": 0.2531,
+      "epoch": 3.102317367643349,
+      "grad_norm": 0.24429123103618622,
+      "learning_rate": 1.6641974571726958e-05,
+      "loss": 0.3673,
       "step": 86080
     },
     {
-      "epoch": 3.03,
-      "learning_rate": 1.777398326960013e-05,
-      "loss": 0.248,
+      "epoch": 3.1024975673045736,
+      "grad_norm": 0.17867647111415863,
+      "learning_rate": 1.6639224406588876e-05,
+      "loss": 0.3625,
       "step": 86085
     },
     {
-      "epoch": 3.03,
-      "learning_rate": 1.777125618769372e-05,
-      "loss": 0.2593,
+      "epoch": 3.102677766965798,
+      "grad_norm": 0.23398911952972412,
+      "learning_rate": 1.6636474355364746e-05,
+      "loss": 0.3595,
       "step": 86090
     },
     {
-      "epoch": 3.03,
-      "learning_rate": 1.7768529199647784e-05,
-      "loss": 0.2566,
+      "epoch": 3.1028579666270226,
+      "grad_norm": 0.2413238137960434,
+      "learning_rate": 1.6633724418092016e-05,
+      "loss": 0.3635,
       "step": 86095
     },
     {
-      "epoch": 3.03,
-      "learning_rate": 1.7765802305497724e-05,
-      "loss": 0.2718,
+      "epoch": 3.1030381662882474,
+      "grad_norm": 0.23708127439022064,
+      "learning_rate": 1.6630974594808192e-05,
+      "loss": 0.3597,
       "step": 86100
     },
     {
-      "epoch": 3.03,
-      "learning_rate": 1.7763075505278962e-05,
-      "loss": 0.2584,
+      "epoch": 3.103218365949472,
+      "grad_norm": 0.20175603032112122,
+      "learning_rate": 1.6628224885550697e-05,
+      "loss": 0.4123,
       "step": 86105
     },
     {
-      "epoch": 3.03,
-      "learning_rate": 1.7760348799026896e-05,
-      "loss": 0.267,
+      "epoch": 3.103398565610697,
+      "grad_norm": 0.2378152459859848,
+      "learning_rate": 1.6625475290357037e-05,
+      "loss": 0.3938,
       "step": 86110
     },
     {
-      "epoch": 3.03,
-      "learning_rate": 1.7757622186776922e-05,
-      "loss": 0.2841,
+      "epoch": 3.103578765271921,
+      "grad_norm": 0.19391091167926788,
+      "learning_rate": 1.662272580926465e-05,
+      "loss": 0.3861,
       "step": 86115
     },
     {
-      "epoch": 3.03,
-      "learning_rate": 1.7754895668564452e-05,
-      "loss": 0.2695,
+      "epoch": 3.103758964933146,
+      "grad_norm": 0.20529715716838837,
+      "learning_rate": 1.6619976442310987e-05,
+      "loss": 0.3877,
       "step": 86120
     },
     {
-      "epoch": 3.03,
-      "learning_rate": 1.7752169244424894e-05,
-      "loss": 0.2629,
+      "epoch": 3.1039391645943706,
+      "grad_norm": 0.21286916732788086,
+      "learning_rate": 1.661722718953354e-05,
+      "loss": 0.3867,
       "step": 86125
     },
     {
-      "epoch": 3.03,
-      "learning_rate": 1.7749442914393644e-05,
-      "loss": 0.2655,
+      "epoch": 3.1041193642555953,
+      "grad_norm": 0.2500961422920227,
+      "learning_rate": 1.6614478050969738e-05,
+      "loss": 0.3611,
       "step": 86130
     },
     {
-      "epoch": 3.03,
-      "learning_rate": 1.7746716678506094e-05,
-      "loss": 0.2585,
+      "epoch": 3.10429956391682,
+      "grad_norm": 0.2693064212799072,
+      "learning_rate": 1.661172902665706e-05,
+      "loss": 0.3954,
       "step": 86135
     },
     {
-      "epoch": 3.03,
-      "learning_rate": 1.774399053679764e-05,
-      "loss": 0.2798,
+      "epoch": 3.1044797635780443,
+      "grad_norm": 0.22724072635173798,
+      "learning_rate": 1.6608980116632945e-05,
+      "loss": 0.3488,
       "step": 86140
     },
     {
-      "epoch": 3.03,
-      "learning_rate": 1.7741264489303694e-05,
-      "loss": 0.2878,
+      "epoch": 3.104659963239269,
+      "grad_norm": 0.24723055958747864,
+      "learning_rate": 1.660623132093485e-05,
+      "loss": 0.3801,
       "step": 86145
     },
     {
-      "epoch": 3.03,
-      "learning_rate": 1.773853853605965e-05,
-      "loss": 0.2479,
+      "epoch": 3.104840162900494,
+      "grad_norm": 0.26212194561958313,
+      "learning_rate": 1.6603482639600232e-05,
+      "loss": 0.3625,
       "step": 86150
     },
     {
-      "epoch": 3.03,
-      "learning_rate": 1.7735812677100896e-05,
-      "loss": 0.2513,
+      "epoch": 3.1050203625617185,
+      "grad_norm": 0.19167543947696686,
+      "learning_rate": 1.6600734072666535e-05,
+      "loss": 0.3962,
       "step": 86155
     },
     {
-      "epoch": 3.03,
-      "learning_rate": 1.7733086912462817e-05,
-      "loss": 0.2343,
+      "epoch": 3.105200562222943,
+      "grad_norm": 0.2559322118759155,
+      "learning_rate": 1.6597985620171207e-05,
+      "loss": 0.3864,
       "step": 86160
     },
     {
-      "epoch": 3.03,
-      "learning_rate": 1.7730361242180827e-05,
-      "loss": 0.2601,
+      "epoch": 3.1053807618841676,
+      "grad_norm": 0.17590028047561646,
+      "learning_rate": 1.65952372821517e-05,
+      "loss": 0.368,
       "step": 86165
     },
     {
-      "epoch": 3.03,
-      "learning_rate": 1.7727635666290298e-05,
-      "loss": 0.2608,
+      "epoch": 3.1055609615453923,
+      "grad_norm": 0.23424407839775085,
+      "learning_rate": 1.6592489058645455e-05,
+      "loss": 0.3734,
       "step": 86170
     },
     {
-      "epoch": 3.03,
-      "learning_rate": 1.772491018482664e-05,
-      "loss": 0.2794,
+      "epoch": 3.105741161206617,
+      "grad_norm": 0.215243399143219,
+      "learning_rate": 1.6589740949689926e-05,
+      "loss": 0.3799,
       "step": 86175
     },
     {
-      "epoch": 3.03,
-      "learning_rate": 1.772218479782521e-05,
-      "loss": 0.2484,
+      "epoch": 3.1059213608678418,
+      "grad_norm": 0.22049564123153687,
+      "learning_rate": 1.6586992955322546e-05,
+      "loss": 0.3542,
       "step": 86180
     },
     {
-      "epoch": 3.03,
-      "learning_rate": 1.7719459505321425e-05,
-      "loss": 0.2363,
+      "epoch": 3.106101560529066,
+      "grad_norm": 0.24883852899074554,
+      "learning_rate": 1.6584245075580753e-05,
+      "loss": 0.3818,
       "step": 86185
     },
     {
-      "epoch": 3.03,
-      "learning_rate": 1.7716734307350664e-05,
-      "loss": 0.2372,
+      "epoch": 3.106281760190291,
+      "grad_norm": 0.25981536507606506,
+      "learning_rate": 1.6581497310502e-05,
+      "loss": 0.3753,
       "step": 86190
     },
     {
-      "epoch": 3.03,
-      "learning_rate": 1.77140092039483e-05,
-      "loss": 0.2774,
+      "epoch": 3.1064619598515155,
+      "grad_norm": 0.23729638755321503,
+      "learning_rate": 1.6578749660123715e-05,
+      "loss": 0.3649,
       "step": 86195
     },
     {
-      "epoch": 3.03,
-      "learning_rate": 1.771128419514972e-05,
-      "loss": 0.2725,
+      "epoch": 3.1066421595127403,
+      "grad_norm": 0.1951550394296646,
+      "learning_rate": 1.6576002124483324e-05,
+      "loss": 0.3514,
       "step": 86200
     },
     {
-      "epoch": 3.03,
-      "learning_rate": 1.7708559280990322e-05,
-      "loss": 0.2564,
+      "epoch": 3.1068223591739645,
+      "grad_norm": 0.23986147344112396,
+      "learning_rate": 1.657325470361829e-05,
+      "loss": 0.3732,
       "step": 86205
     },
     {
-      "epoch": 3.03,
-      "learning_rate": 1.7705834461505477e-05,
-      "loss": 0.2536,
+      "epoch": 3.1070025588351893,
+      "grad_norm": 0.27552303671836853,
+      "learning_rate": 1.657050739756601e-05,
+      "loss": 0.3641,
       "step": 86210
     },
     {
-      "epoch": 3.03,
-      "learning_rate": 1.7703109736730556e-05,
-      "loss": 0.2554,
+      "epoch": 3.107182758496414,
+      "grad_norm": 0.2763289511203766,
+      "learning_rate": 1.6567760206363953e-05,
+      "loss": 0.3584,
       "step": 86215
     },
     {
-      "epoch": 3.03,
-      "learning_rate": 1.770038510670095e-05,
-      "loss": 0.2723,
+      "epoch": 3.1073629581576387,
+      "grad_norm": 0.21052367985248566,
+      "learning_rate": 1.6565013130049526e-05,
+      "loss": 0.3756,
       "step": 86220
     },
     {
-      "epoch": 3.03,
-      "learning_rate": 1.7697660571452034e-05,
-      "loss": 0.2475,
+      "epoch": 3.1075431578188635,
+      "grad_norm": 0.2306980937719345,
+      "learning_rate": 1.6562266168660153e-05,
+      "loss": 0.3861,
       "step": 86225
     },
     {
-      "epoch": 3.03,
-      "learning_rate": 1.7694936131019186e-05,
-      "loss": 0.2406,
+      "epoch": 3.1077233574800878,
+      "grad_norm": 0.2257169485092163,
+      "learning_rate": 1.655951932223328e-05,
+      "loss": 0.4018,
       "step": 86230
     },
     {
-      "epoch": 3.03,
-      "learning_rate": 1.7692211785437768e-05,
-      "loss": 0.259,
+      "epoch": 3.1079035571413125,
+      "grad_norm": 0.20288598537445068,
+      "learning_rate": 1.6556772590806318e-05,
+      "loss": 0.3691,
       "step": 86235
     },
     {
-      "epoch": 3.03,
-      "learning_rate": 1.7689487534743177e-05,
-      "loss": 0.2834,
+      "epoch": 3.1080837568025372,
+      "grad_norm": 0.23873363435268402,
+      "learning_rate": 1.6554025974416693e-05,
+      "loss": 0.3547,
       "step": 86240
     },
     {
-      "epoch": 3.03,
-      "learning_rate": 1.7686763378970767e-05,
-      "loss": 0.289,
+      "epoch": 3.108263956463762,
+      "grad_norm": 0.2184680998325348,
+      "learning_rate": 1.6551279473101834e-05,
+      "loss": 0.3309,
       "step": 86245
     },
     {
-      "epoch": 3.03,
-      "learning_rate": 1.7684039318155914e-05,
-      "loss": 0.2726,
+      "epoch": 3.1084441561249863,
+      "grad_norm": 0.26064011454582214,
+      "learning_rate": 1.654853308689915e-05,
+      "loss": 0.3793,
       "step": 86250
     },
     {
-      "epoch": 3.03,
-      "learning_rate": 1.7681315352333987e-05,
-      "loss": 0.2704,
+      "epoch": 3.108624355786211,
+      "grad_norm": 0.22625210881233215,
+      "learning_rate": 1.6545786815846067e-05,
+      "loss": 0.401,
       "step": 86255
     },
     {
-      "epoch": 3.03,
-      "learning_rate": 1.7678591481540357e-05,
-      "loss": 0.2619,
+      "epoch": 3.1088045554474357,
+      "grad_norm": 0.21573030948638916,
+      "learning_rate": 1.654304065998001e-05,
+      "loss": 0.3825,
       "step": 86260
     },
     {
-      "epoch": 3.04,
-      "learning_rate": 1.76758677058104e-05,
-      "loss": 0.2641,
+      "epoch": 3.1089847551086605,
+      "grad_norm": 0.21487176418304443,
+      "learning_rate": 1.6540294619338377e-05,
+      "loss": 0.3793,
       "step": 86265
     },
     {
-      "epoch": 3.04,
-      "learning_rate": 1.7673144025179473e-05,
-      "loss": 0.2632,
+      "epoch": 3.109164954769885,
+      "grad_norm": 0.19863709807395935,
+      "learning_rate": 1.6537548693958593e-05,
+      "loss": 0.3945,
       "step": 86270
     },
     {
-      "epoch": 3.04,
-      "learning_rate": 1.767042043968293e-05,
-      "loss": 0.26,
+      "epoch": 3.1093451544311095,
+      "grad_norm": 0.2572442889213562,
+      "learning_rate": 1.6534802883878083e-05,
+      "loss": 0.3819,
       "step": 86275
     },
     {
-      "epoch": 3.04,
-      "learning_rate": 1.7667696949356156e-05,
-      "loss": 0.2451,
+      "epoch": 3.109525354092334,
+      "grad_norm": 0.26214733719825745,
+      "learning_rate": 1.6532057189134227e-05,
+      "loss": 0.4062,
       "step": 86280
     },
     {
-      "epoch": 3.04,
-      "learning_rate": 1.7664973554234505e-05,
-      "loss": 0.2431,
+      "epoch": 3.109705553753559,
+      "grad_norm": 0.21623064577579498,
+      "learning_rate": 1.6529311609764458e-05,
+      "loss": 0.3626,
       "step": 86285
     },
     {
-      "epoch": 3.04,
-      "learning_rate": 1.7662250254353337e-05,
-      "loss": 0.2845,
+      "epoch": 3.1098857534147837,
+      "grad_norm": 0.24655792117118835,
+      "learning_rate": 1.6526566145806177e-05,
+      "loss": 0.4102,
       "step": 86290
     },
     {
-      "epoch": 3.04,
-      "learning_rate": 1.765952704974801e-05,
-      "loss": 0.2732,
+      "epoch": 3.1100659530760084,
+      "grad_norm": 0.1786569356918335,
+      "learning_rate": 1.6523820797296796e-05,
+      "loss": 0.3942,
       "step": 86295
     },
     {
-      "epoch": 3.04,
-      "learning_rate": 1.7656803940453893e-05,
-      "loss": 0.227,
+      "epoch": 3.1102461527372327,
+      "grad_norm": 0.23387044668197632,
+      "learning_rate": 1.652107556427372e-05,
+      "loss": 0.3798,
       "step": 86300
     },
     {
-      "epoch": 3.04,
-      "learning_rate": 1.7654080926506335e-05,
-      "loss": 0.2505,
+      "epoch": 3.1104263523984574,
+      "grad_norm": 0.21776267886161804,
+      "learning_rate": 1.6518330446774334e-05,
+      "loss": 0.3593,
       "step": 86305
     },
     {
-      "epoch": 3.04,
-      "learning_rate": 1.7651358007940694e-05,
-      "loss": 0.2697,
+      "epoch": 3.110606552059682,
+      "grad_norm": 0.21297864615917206,
+      "learning_rate": 1.651558544483607e-05,
+      "loss": 0.3782,
       "step": 86310
     },
     {
-      "epoch": 3.04,
-      "learning_rate": 1.764863518479232e-05,
-      "loss": 0.2797,
+      "epoch": 3.110786751720907,
+      "grad_norm": 0.2200326770544052,
+      "learning_rate": 1.6512840558496295e-05,
+      "loss": 0.3814,
       "step": 86315
     },
     {
-      "epoch": 3.04,
-      "learning_rate": 1.7645912457096586e-05,
-      "loss": 0.2775,
+      "epoch": 3.110966951382131,
+      "grad_norm": 0.2481064796447754,
+      "learning_rate": 1.651009578779244e-05,
+      "loss": 0.3499,
       "step": 86320
     },
     {
-      "epoch": 3.04,
-      "learning_rate": 1.7643189824888827e-05,
-      "loss": 0.2449,
+      "epoch": 3.111147151043356,
+      "grad_norm": 0.20087507367134094,
+      "learning_rate": 1.650735113276188e-05,
+      "loss": 0.3656,
       "step": 86325
     },
     {
-      "epoch": 3.04,
-      "learning_rate": 1.764046728820439e-05,
-      "loss": 0.2385,
+      "epoch": 3.1113273507045807,
+      "grad_norm": 0.225091814994812,
+      "learning_rate": 1.6504606593442014e-05,
+      "loss": 0.3703,
       "step": 86330
     },
     {
-      "epoch": 3.04,
-      "learning_rate": 1.7637744847078652e-05,
-      "loss": 0.2674,
+      "epoch": 3.1115075503658054,
+      "grad_norm": 0.20192007720470428,
+      "learning_rate": 1.650186216987024e-05,
+      "loss": 0.4075,
       "step": 86335
     },
     {
-      "epoch": 3.04,
-      "learning_rate": 1.7635022501546938e-05,
-      "loss": 0.2579,
+      "epoch": 3.11168775002703,
+      "grad_norm": 0.2589082717895508,
+      "learning_rate": 1.6499117862083953e-05,
+      "loss": 0.405,
       "step": 86340
     },
     {
-      "epoch": 3.04,
-      "learning_rate": 1.7632300251644606e-05,
-      "loss": 0.2655,
+      "epoch": 3.1118679496882544,
+      "grad_norm": 0.2047402262687683,
+      "learning_rate": 1.649637367012053e-05,
+      "loss": 0.3555,
       "step": 86345
     },
     {
-      "epoch": 3.04,
-      "learning_rate": 1.762957809740699e-05,
-      "loss": 0.2744,
+      "epoch": 3.112048149349479,
+      "grad_norm": 0.2570793032646179,
+      "learning_rate": 1.6493629594017376e-05,
+      "loss": 0.3869,
       "step": 86350
     },
     {
-      "epoch": 3.04,
-      "learning_rate": 1.7626856038869456e-05,
-      "loss": 0.247,
+      "epoch": 3.112228349010704,
+      "grad_norm": 0.23956820368766785,
+      "learning_rate": 1.6490885633811868e-05,
+      "loss": 0.3472,
       "step": 86355
     },
     {
-      "epoch": 3.04,
-      "learning_rate": 1.7624134076067336e-05,
-      "loss": 0.268,
+      "epoch": 3.1124085486719286,
+      "grad_norm": 0.24650585651397705,
+      "learning_rate": 1.64881417895414e-05,
+      "loss": 0.3631,
       "step": 86360
     },
     {
-      "epoch": 3.04,
-      "learning_rate": 1.762141220903597e-05,
-      "loss": 0.2387,
+      "epoch": 3.1125887483331534,
+      "grad_norm": 0.22965595126152039,
+      "learning_rate": 1.6485398061243353e-05,
+      "loss": 0.3844,
       "step": 86365
     },
     {
-      "epoch": 3.04,
-      "learning_rate": 1.7618690437810697e-05,
-      "loss": 0.2421,
+      "epoch": 3.1127689479943776,
+      "grad_norm": 0.22221916913986206,
+      "learning_rate": 1.6482654448955097e-05,
+      "loss": 0.3907,
       "step": 86370
     },
     {
-      "epoch": 3.04,
-      "learning_rate": 1.7615968762426876e-05,
-      "loss": 0.2817,
+      "epoch": 3.1129491476556024,
+      "grad_norm": 0.22202599048614502,
+      "learning_rate": 1.6479910952714038e-05,
+      "loss": 0.4044,
       "step": 86375
     },
     {
-      "epoch": 3.04,
-      "learning_rate": 1.7613247182919836e-05,
-      "loss": 0.2491,
+      "epoch": 3.113129347316827,
+      "grad_norm": 0.21764975786209106,
+      "learning_rate": 1.6477167572557547e-05,
+      "loss": 0.4054,
       "step": 86380
     },
     {
-      "epoch": 3.04,
-      "learning_rate": 1.7610525699324908e-05,
-      "loss": 0.2615,
+      "epoch": 3.113309546978052,
+      "grad_norm": 0.29039353132247925,
+      "learning_rate": 1.647442430852298e-05,
+      "loss": 0.3839,
       "step": 86385
     },
     {
-      "epoch": 3.04,
-      "learning_rate": 1.7607804311677427e-05,
-      "loss": 0.2524,
+      "epoch": 3.113489746639276,
+      "grad_norm": 0.1968279629945755,
+      "learning_rate": 1.6471681160647752e-05,
+      "loss": 0.3557,
       "step": 86390
     },
     {
-      "epoch": 3.04,
-      "learning_rate": 1.7605083020012744e-05,
-      "loss": 0.2461,
+      "epoch": 3.113669946300501,
+      "grad_norm": 0.22088350355625153,
+      "learning_rate": 1.6468938128969194e-05,
+      "loss": 0.3762,
       "step": 86395
     },
     {
-      "epoch": 3.04,
-      "learning_rate": 1.7602361824366186e-05,
-      "loss": 0.2614,
+      "epoch": 3.1138501459617256,
+      "grad_norm": 0.30237895250320435,
+      "learning_rate": 1.6466195213524722e-05,
+      "loss": 0.4007,
       "step": 86400
     },
     {
-      "epoch": 3.04,
-      "learning_rate": 1.7599640724773085e-05,
-      "loss": 0.2538,
+      "epoch": 3.1140303456229503,
+      "grad_norm": 0.21387836337089539,
+      "learning_rate": 1.6463452414351683e-05,
+      "loss": 0.3915,
       "step": 86405
     },
     {
-      "epoch": 3.04,
-      "learning_rate": 1.759691972126876e-05,
-      "loss": 0.2505,
+      "epoch": 3.114210545284175,
+      "grad_norm": 0.22359785437583923,
+      "learning_rate": 1.646070973148744e-05,
+      "loss": 0.3664,
       "step": 86410
     },
     {
-      "epoch": 3.04,
-      "learning_rate": 1.7594198813888568e-05,
-      "loss": 0.2782,
+      "epoch": 3.1143907449453994,
+      "grad_norm": 0.2460165023803711,
+      "learning_rate": 1.645796716496939e-05,
+      "loss": 0.4142,
       "step": 86415
     },
     {
-      "epoch": 3.04,
-      "learning_rate": 1.7591478002667817e-05,
-      "loss": 0.2593,
+      "epoch": 3.114570944606624,
+      "grad_norm": 0.20692665874958038,
+      "learning_rate": 1.6455224714834876e-05,
+      "loss": 0.3838,
       "step": 86420
     },
     {
-      "epoch": 3.04,
-      "learning_rate": 1.7588757287641844e-05,
-      "loss": 0.2685,
+      "epoch": 3.114751144267849,
+      "grad_norm": 0.18075329065322876,
+      "learning_rate": 1.645248238112127e-05,
+      "loss": 0.3807,
       "step": 86425
     },
     {
-      "epoch": 3.04,
-      "learning_rate": 1.7586036668845966e-05,
-      "loss": 0.282,
+      "epoch": 3.1149313439290736,
+      "grad_norm": 0.21312923729419708,
+      "learning_rate": 1.644974016386594e-05,
+      "loss": 0.3996,
       "step": 86430
     },
     {
-      "epoch": 3.04,
-      "learning_rate": 1.7583316146315525e-05,
-      "loss": 0.2636,
+      "epoch": 3.115111543590298,
+      "grad_norm": 0.1856447458267212,
+      "learning_rate": 1.644699806310624e-05,
+      "loss": 0.3807,
       "step": 86435
     },
     {
-      "epoch": 3.04,
-      "learning_rate": 1.7580595720085835e-05,
-      "loss": 0.2661,
+      "epoch": 3.1152917432515226,
+      "grad_norm": 0.23847022652626038,
+      "learning_rate": 1.6444256078879537e-05,
+      "loss": 0.3593,
       "step": 86440
     },
     {
-      "epoch": 3.04,
-      "learning_rate": 1.757787539019221e-05,
-      "loss": 0.2676,
+      "epoch": 3.1154719429127473,
+      "grad_norm": 0.2461748719215393,
+      "learning_rate": 1.6441514211223197e-05,
+      "loss": 0.3845,
       "step": 86445
     },
     {
-      "epoch": 3.04,
-      "learning_rate": 1.7575155156669993e-05,
-      "loss": 0.257,
+      "epoch": 3.115652142573972,
+      "grad_norm": 0.17290343344211578,
+      "learning_rate": 1.6438772460174558e-05,
+      "loss": 0.385,
       "step": 86450
     },
     {
-      "epoch": 3.04,
-      "learning_rate": 1.757243501955449e-05,
-      "loss": 0.2919,
+      "epoch": 3.115832342235197,
+      "grad_norm": 0.2519152760505676,
+      "learning_rate": 1.6436030825770992e-05,
+      "loss": 0.3341,
       "step": 86455
     },
     {
-      "epoch": 3.04,
-      "learning_rate": 1.756971497888103e-05,
-      "loss": 0.2632,
+      "epoch": 3.116012541896421,
+      "grad_norm": 0.22315552830696106,
+      "learning_rate": 1.6433289308049847e-05,
+      "loss": 0.3806,
       "step": 86460
     },
     {
-      "epoch": 3.04,
-      "learning_rate": 1.7566995034684908e-05,
-      "loss": 0.2783,
+      "epoch": 3.116192741557646,
+      "grad_norm": 0.20392777025699615,
+      "learning_rate": 1.6430547907048474e-05,
+      "loss": 0.3915,
       "step": 86465
     },
     {
-      "epoch": 3.04,
-      "learning_rate": 1.7564275187001472e-05,
-      "loss": 0.2767,
+      "epoch": 3.1163729412188705,
+      "grad_norm": 0.2624046206474304,
+      "learning_rate": 1.6427806622804233e-05,
+      "loss": 0.4327,
       "step": 86470
     },
     {
-      "epoch": 3.04,
-      "learning_rate": 1.7561555435866012e-05,
-      "loss": 0.2671,
+      "epoch": 3.1165531408800953,
+      "grad_norm": 0.1981981098651886,
+      "learning_rate": 1.6425065455354457e-05,
+      "loss": 0.3494,
       "step": 86475
     },
     {
-      "epoch": 3.04,
-      "learning_rate": 1.755883578131386e-05,
-      "loss": 0.2403,
+      "epoch": 3.1167333405413196,
+      "grad_norm": 0.2108498364686966,
+      "learning_rate": 1.6422324404736512e-05,
+      "loss": 0.3365,
       "step": 86480
     },
     {
-      "epoch": 3.04,
-      "learning_rate": 1.7556116223380315e-05,
-      "loss": 0.2492,
+      "epoch": 3.1169135402025443,
+      "grad_norm": 0.30326491594314575,
+      "learning_rate": 1.6419583470987742e-05,
+      "loss": 0.3961,
       "step": 86485
     },
     {
-      "epoch": 3.04,
-      "learning_rate": 1.7553396762100702e-05,
-      "loss": 0.257,
+      "epoch": 3.117093739863769,
+      "grad_norm": 0.22434242069721222,
+      "learning_rate": 1.641684265414547e-05,
+      "loss": 0.3698,
       "step": 86490
     },
     {
-      "epoch": 3.04,
-      "learning_rate": 1.7550677397510324e-05,
-      "loss": 0.2787,
+      "epoch": 3.1172739395249938,
+      "grad_norm": 0.2361210584640503,
+      "learning_rate": 1.641410195424707e-05,
+      "loss": 0.3758,
       "step": 86495
     },
     {
-      "epoch": 3.04,
-      "learning_rate": 1.754795812964449e-05,
-      "loss": 0.2483,
+      "epoch": 3.1174541391862185,
+      "grad_norm": 0.23360228538513184,
+      "learning_rate": 1.6411361371329852e-05,
+      "loss": 0.3541,
       "step": 86500
     },
     {
-      "epoch": 3.04,
-      "eval_loss": 0.2569468021392822,
-      "eval_runtime": 10.5367,
-      "eval_samples_per_second": 9.491,
-      "eval_steps_per_second": 9.491,
+      "epoch": 3.1174541391862185,
+      "eval_loss": 0.43187737464904785,
+      "eval_runtime": 3.5321,
+      "eval_samples_per_second": 28.311,
+      "eval_steps_per_second": 7.078,
       "step": 86500
     },
     {
-      "epoch": 3.04,
-      "learning_rate": 1.7545238958538496e-05,
-      "loss": 0.2703,
+      "epoch": 3.117634338847443,
+      "grad_norm": 0.26229509711265564,
+      "learning_rate": 1.6408620905431192e-05,
+      "loss": 0.3866,
       "step": 86505
     },
     {
-      "epoch": 3.04,
-      "learning_rate": 1.7542519884227677e-05,
-      "loss": 0.2612,
+      "epoch": 3.1178145385086675,
+      "grad_norm": 0.22185193002223969,
+      "learning_rate": 1.64058805565884e-05,
+      "loss": 0.3794,
       "step": 86510
     },
     {
-      "epoch": 3.04,
-      "learning_rate": 1.753980090674732e-05,
-      "loss": 0.2541,
+      "epoch": 3.1179947381698923,
+      "grad_norm": 0.26995113492012024,
+      "learning_rate": 1.6403140324838817e-05,
+      "loss": 0.3945,
       "step": 86515
     },
     {
-      "epoch": 3.04,
-      "learning_rate": 1.7537082026132734e-05,
-      "loss": 0.2626,
+      "epoch": 3.118174937831117,
+      "grad_norm": 0.30093368887901306,
+      "learning_rate": 1.640040021021979e-05,
+      "loss": 0.3729,
       "step": 86520
     },
     {
-      "epoch": 3.04,
-      "learning_rate": 1.7534363242419208e-05,
-      "loss": 0.2525,
+      "epoch": 3.1183551374923417,
+      "grad_norm": 0.24983744323253632,
+      "learning_rate": 1.6397660212768643e-05,
+      "loss": 0.369,
       "step": 86525
     },
     {
-      "epoch": 3.04,
-      "learning_rate": 1.753164455564206e-05,
-      "loss": 0.2516,
+      "epoch": 3.118535337153566,
+      "grad_norm": 0.2591390609741211,
+      "learning_rate": 1.6394920332522705e-05,
+      "loss": 0.398,
       "step": 86530
     },
     {
-      "epoch": 3.04,
-      "learning_rate": 1.752892596583659e-05,
-      "loss": 0.2643,
+      "epoch": 3.1187155368147907,
+      "grad_norm": 0.23025575280189514,
+      "learning_rate": 1.6392180569519316e-05,
+      "loss": 0.38,
       "step": 86535
     },
     {
-      "epoch": 3.04,
-      "learning_rate": 1.7526207473038096e-05,
-      "loss": 0.2558,
+      "epoch": 3.1188957364760155,
+      "grad_norm": 0.1807858794927597,
+      "learning_rate": 1.6389440923795795e-05,
+      "loss": 0.3726,
       "step": 86540
     },
     {
-      "epoch": 3.04,
-      "learning_rate": 1.752348907728186e-05,
-      "loss": 0.2617,
+      "epoch": 3.11907593613724,
+      "grad_norm": 0.24324113130569458,
+      "learning_rate": 1.6386701395389484e-05,
+      "loss": 0.3855,
       "step": 86545
     },
     {
-      "epoch": 3.05,
-      "learning_rate": 1.7520770778603203e-05,
-      "loss": 0.2511,
+      "epoch": 3.1192561357984645,
+      "grad_norm": 0.17595435678958893,
+      "learning_rate": 1.6383961984337694e-05,
+      "loss": 0.3635,
       "step": 86550
     },
     {
-      "epoch": 3.05,
-      "learning_rate": 1.7518052577037404e-05,
-      "loss": 0.271,
+      "epoch": 3.1194363354596892,
+      "grad_norm": 0.2983537018299103,
+      "learning_rate": 1.638122269067775e-05,
+      "loss": 0.393,
       "step": 86555
     },
     {
-      "epoch": 3.05,
-      "learning_rate": 1.751533447261975e-05,
-      "loss": 0.2645,
+      "epoch": 3.119616535120914,
+      "grad_norm": 0.25049325823783875,
+      "learning_rate": 1.6378483514446984e-05,
+      "loss": 0.3885,
       "step": 86560
     },
     {
-      "epoch": 3.05,
-      "learning_rate": 1.7512616465385562e-05,
-      "loss": 0.2534,
+      "epoch": 3.1197967347821387,
+      "grad_norm": 0.26442599296569824,
+      "learning_rate": 1.637574445568272e-05,
+      "loss": 0.3671,
       "step": 86565
     },
     {
-      "epoch": 3.05,
-      "learning_rate": 1.7509898555370113e-05,
-      "loss": 0.2519,
+      "epoch": 3.1199769344433634,
+      "grad_norm": 0.2337508350610733,
+      "learning_rate": 1.6373005514422246e-05,
+      "loss": 0.3758,
       "step": 86570
     },
     {
-      "epoch": 3.05,
-      "learning_rate": 1.7507180742608695e-05,
-      "loss": 0.2461,
+      "epoch": 3.1201571341045877,
+      "grad_norm": 0.2032548189163208,
+      "learning_rate": 1.6370266690702912e-05,
+      "loss": 0.3562,
       "step": 86575
     },
     {
-      "epoch": 3.05,
-      "learning_rate": 1.7504463027136585e-05,
-      "loss": 0.249,
+      "epoch": 3.1203373337658125,
+      "grad_norm": 0.24068522453308105,
+      "learning_rate": 1.6367527984562024e-05,
+      "loss": 0.36,
       "step": 86580
     },
     {
-      "epoch": 3.05,
-      "learning_rate": 1.750174540898909e-05,
-      "loss": 0.2673,
+      "epoch": 3.120517533427037,
+      "grad_norm": 0.26474300026893616,
+      "learning_rate": 1.6364789396036895e-05,
+      "loss": 0.4097,
       "step": 86585
     },
     {
-      "epoch": 3.05,
-      "learning_rate": 1.7499027888201496e-05,
-      "loss": 0.2567,
+      "epoch": 3.120697733088262,
+      "grad_norm": 0.17933666706085205,
+      "learning_rate": 1.6362050925164846e-05,
+      "loss": 0.3599,
       "step": 86590
     },
     {
-      "epoch": 3.05,
-      "learning_rate": 1.749631046480908e-05,
-      "loss": 0.2565,
+      "epoch": 3.120877932749486,
+      "grad_norm": 0.23188619315624237,
+      "learning_rate": 1.635931257198317e-05,
+      "loss": 0.3753,
       "step": 86595
     },
     {
-      "epoch": 3.05,
-      "learning_rate": 1.7493593138847113e-05,
-      "loss": 0.2932,
+      "epoch": 3.121058132410711,
+      "grad_norm": 0.19330739974975586,
+      "learning_rate": 1.6356574336529196e-05,
+      "loss": 0.3503,
       "step": 86600
     },
     {
-      "epoch": 3.05,
-      "learning_rate": 1.749087591035091e-05,
-      "loss": 0.2615,
+      "epoch": 3.1212383320719357,
+      "grad_norm": 0.21732577681541443,
+      "learning_rate": 1.6353836218840223e-05,
+      "loss": 0.4139,
       "step": 86605
     },
     {
-      "epoch": 3.05,
-      "learning_rate": 1.7488158779355724e-05,
-      "loss": 0.2538,
+      "epoch": 3.1214185317331604,
+      "grad_norm": 0.2698850929737091,
+      "learning_rate": 1.635109821895355e-05,
+      "loss": 0.3719,
       "step": 86610
     },
     {
-      "epoch": 3.05,
-      "learning_rate": 1.7485441745896853e-05,
-      "loss": 0.2511,
+      "epoch": 3.121598731394385,
+      "grad_norm": 0.21727710962295532,
+      "learning_rate": 1.6348360336906492e-05,
+      "loss": 0.367,
       "step": 86615
     },
     {
-      "epoch": 3.05,
-      "learning_rate": 1.748272481000956e-05,
-      "loss": 0.2427,
+      "epoch": 3.1217789310556094,
+      "grad_norm": 0.23365622758865356,
+      "learning_rate": 1.6345622572736345e-05,
+      "loss": 0.4036,
       "step": 86620
     },
     {
-      "epoch": 3.05,
-      "learning_rate": 1.748000797172914e-05,
-      "loss": 0.2435,
+      "epoch": 3.121959130716834,
+      "grad_norm": 0.23919281363487244,
+      "learning_rate": 1.6342884926480425e-05,
+      "loss": 0.3967,
       "step": 86625
     },
     {
-      "epoch": 3.05,
-      "learning_rate": 1.7477291231090857e-05,
-      "loss": 0.242,
+      "epoch": 3.122139330378059,
+      "grad_norm": 0.1928279846906662,
+      "learning_rate": 1.634014739817602e-05,
+      "loss": 0.4055,
       "step": 86630
     },
     {
-      "epoch": 3.05,
-      "learning_rate": 1.7474574588129987e-05,
-      "loss": 0.2656,
+      "epoch": 3.1223195300392836,
+      "grad_norm": 0.21561673283576965,
+      "learning_rate": 1.6337409987860423e-05,
+      "loss": 0.3846,
       "step": 86635
     },
     {
-      "epoch": 3.05,
-      "learning_rate": 1.7471858042881805e-05,
-      "loss": 0.2405,
+      "epoch": 3.1224997297005084,
+      "grad_norm": 0.2624969780445099,
+      "learning_rate": 1.6334672695570942e-05,
+      "loss": 0.4019,
       "step": 86640
     },
     {
-      "epoch": 3.05,
-      "learning_rate": 1.7469141595381594e-05,
-      "loss": 0.2518,
+      "epoch": 3.1226799293617327,
+      "grad_norm": 0.18644605576992035,
+      "learning_rate": 1.633193552134487e-05,
+      "loss": 0.3881,
       "step": 86645
     },
     {
-      "epoch": 3.05,
-      "learning_rate": 1.7466425245664614e-05,
-      "loss": 0.2752,
+      "epoch": 3.1228601290229574,
+      "grad_norm": 0.20565631985664368,
+      "learning_rate": 1.632919846521949e-05,
+      "loss": 0.3493,
       "step": 86650
     },
     {
-      "epoch": 3.05,
-      "learning_rate": 1.7463708993766137e-05,
-      "loss": 0.2634,
+      "epoch": 3.123040328684182,
+      "grad_norm": 0.2321755588054657,
+      "learning_rate": 1.632646152723211e-05,
+      "loss": 0.3675,
       "step": 86655
     },
     {
-      "epoch": 3.05,
-      "learning_rate": 1.7460992839721422e-05,
-      "loss": 0.2472,
+      "epoch": 3.123220528345407,
+      "grad_norm": 0.20580174028873444,
+      "learning_rate": 1.6323724707420013e-05,
+      "loss": 0.3705,
       "step": 86660
     },
     {
-      "epoch": 3.05,
-      "learning_rate": 1.7458276783565754e-05,
-      "loss": 0.2352,
+      "epoch": 3.123400728006631,
+      "grad_norm": 0.19330060482025146,
+      "learning_rate": 1.6320988005820484e-05,
+      "loss": 0.3382,
       "step": 86665
     },
     {
-      "epoch": 3.05,
-      "learning_rate": 1.74555608253344e-05,
-      "loss": 0.2356,
+      "epoch": 3.123580927667856,
+      "grad_norm": 0.20522934198379517,
+      "learning_rate": 1.631825142247082e-05,
+      "loss": 0.3555,
       "step": 86670
     },
     {
-      "epoch": 3.05,
-      "learning_rate": 1.7452844965062598e-05,
-      "loss": 0.2326,
+      "epoch": 3.1237611273290806,
+      "grad_norm": 0.23507462441921234,
+      "learning_rate": 1.6315514957408284e-05,
+      "loss": 0.3862,
       "step": 86675
     },
     {
-      "epoch": 3.05,
-      "learning_rate": 1.7450129202785647e-05,
-      "loss": 0.2642,
+      "epoch": 3.1239413269903054,
+      "grad_norm": 0.2328176349401474,
+      "learning_rate": 1.6312778610670193e-05,
+      "loss": 0.3903,
       "step": 86680
     },
     {
-      "epoch": 3.05,
-      "learning_rate": 1.744741353853879e-05,
-      "loss": 0.2588,
+      "epoch": 3.12412152665153,
+      "grad_norm": 0.220947727560997,
+      "learning_rate": 1.6310042382293806e-05,
+      "loss": 0.3965,
       "step": 86685
     },
     {
-      "epoch": 3.05,
-      "learning_rate": 1.7444697972357287e-05,
-      "loss": 0.2801,
+      "epoch": 3.1243017263127544,
+      "grad_norm": 0.2658975124359131,
+      "learning_rate": 1.6307306272316402e-05,
+      "loss": 0.3763,
       "step": 86690
     },
     {
-      "epoch": 3.05,
-      "learning_rate": 1.74419825042764e-05,
-      "loss": 0.2571,
+      "epoch": 3.124481925973979,
+      "grad_norm": 0.2326037585735321,
+      "learning_rate": 1.6304570280775275e-05,
+      "loss": 0.4033,
       "step": 86695
     },
     {
-      "epoch": 3.05,
-      "learning_rate": 1.7439267134331393e-05,
-      "loss": 0.2463,
+      "epoch": 3.124662125635204,
+      "grad_norm": 0.22777698934078217,
+      "learning_rate": 1.6301834407707684e-05,
+      "loss": 0.4019,
       "step": 86700
     },
     {
-      "epoch": 3.05,
-      "learning_rate": 1.743655186255752e-05,
-      "loss": 0.2573,
+      "epoch": 3.1248423252964286,
+      "grad_norm": 0.2394534796476364,
+      "learning_rate": 1.6299098653150926e-05,
+      "loss": 0.3871,
       "step": 86705
     },
     {
-      "epoch": 3.05,
-      "learning_rate": 1.743383668899004e-05,
-      "loss": 0.2593,
+      "epoch": 3.125022524957653,
+      "grad_norm": 0.18719437718391418,
+      "learning_rate": 1.6296363017142264e-05,
+      "loss": 0.3503,
       "step": 86710
     },
     {
-      "epoch": 3.05,
-      "learning_rate": 1.7431121613664193e-05,
-      "loss": 0.257,
+      "epoch": 3.1252027246188776,
+      "grad_norm": 0.24379460513591766,
+      "learning_rate": 1.6293627499718962e-05,
+      "loss": 0.3867,
       "step": 86715
     },
     {
-      "epoch": 3.05,
-      "learning_rate": 1.7428406636615252e-05,
-      "loss": 0.2623,
+      "epoch": 3.1253829242801023,
+      "grad_norm": 0.20296038687229156,
+      "learning_rate": 1.629089210091831e-05,
+      "loss": 0.3609,
       "step": 86720
     },
     {
-      "epoch": 3.05,
-      "learning_rate": 1.742569175787846e-05,
-      "loss": 0.2597,
+      "epoch": 3.125563123941327,
+      "grad_norm": 0.21820303797721863,
+      "learning_rate": 1.6288156820777556e-05,
+      "loss": 0.3685,
       "step": 86725
     },
     {
-      "epoch": 3.05,
-      "learning_rate": 1.742297697748907e-05,
-      "loss": 0.2608,
+      "epoch": 3.125743323602552,
+      "grad_norm": 0.25657927989959717,
+      "learning_rate": 1.628542165933399e-05,
+      "loss": 0.4238,
       "step": 86730
     },
     {
-      "epoch": 3.05,
-      "learning_rate": 1.742026229548232e-05,
-      "loss": 0.2469,
+      "epoch": 3.125923523263776,
+      "grad_norm": 0.26692578196525574,
+      "learning_rate": 1.628268661662486e-05,
+      "loss": 0.3551,
       "step": 86735
     },
     {
-      "epoch": 3.05,
-      "learning_rate": 1.7417547711893485e-05,
-      "loss": 0.2675,
+      "epoch": 3.126103722925001,
+      "grad_norm": 0.23433983325958252,
+      "learning_rate": 1.6279951692687433e-05,
+      "loss": 0.3719,
       "step": 86740
     },
     {
-      "epoch": 3.05,
-      "learning_rate": 1.7414833226757786e-05,
-      "loss": 0.2315,
+      "epoch": 3.1262839225862256,
+      "grad_norm": 0.19793805480003357,
+      "learning_rate": 1.6277216887558982e-05,
+      "loss": 0.3766,
       "step": 86745
     },
     {
-      "epoch": 3.05,
-      "learning_rate": 1.7412118840110484e-05,
-      "loss": 0.2394,
+      "epoch": 3.1264641222474503,
+      "grad_norm": 0.2359396368265152,
+      "learning_rate": 1.6274482201276765e-05,
+      "loss": 0.3728,
       "step": 86750
     },
     {
-      "epoch": 3.05,
-      "learning_rate": 1.740940455198681e-05,
-      "loss": 0.267,
+      "epoch": 3.1266443219086746,
+      "grad_norm": 0.24847714602947235,
+      "learning_rate": 1.6271747633878032e-05,
+      "loss": 0.3969,
       "step": 86755
     },
     {
-      "epoch": 3.05,
-      "learning_rate": 1.7406690362422023e-05,
-      "loss": 0.2702,
+      "epoch": 3.1268245215698993,
+      "grad_norm": 0.26364535093307495,
+      "learning_rate": 1.6269013185400054e-05,
+      "loss": 0.3949,
       "step": 86760
     },
     {
-      "epoch": 3.05,
-      "learning_rate": 1.7403976271451358e-05,
-      "loss": 0.2547,
+      "epoch": 3.127004721231124,
+      "grad_norm": 0.22220592200756073,
+      "learning_rate": 1.6266278855880075e-05,
+      "loss": 0.3691,
       "step": 86765
     },
     {
-      "epoch": 3.05,
-      "learning_rate": 1.7401805069686268e-05,
-      "loss": 0.2618,
+      "epoch": 3.127184920892349,
+      "grad_norm": 0.2486381232738495,
+      "learning_rate": 1.626354464535536e-05,
+      "loss": 0.3708,
       "step": 86770
     },
     {
-      "epoch": 3.05,
-      "learning_rate": 1.7399091156273828e-05,
-      "loss": 0.2611,
+      "epoch": 3.1273651205535735,
+      "grad_norm": 0.23804841935634613,
+      "learning_rate": 1.626081055386316e-05,
+      "loss": 0.3858,
       "step": 86775
     },
     {
-      "epoch": 3.05,
-      "learning_rate": 1.7396377341554176e-05,
-      "loss": 0.2334,
+      "epoch": 3.127545320214798,
+      "grad_norm": 0.19268541038036346,
+      "learning_rate": 1.6258076581440712e-05,
+      "loss": 0.3594,
       "step": 86780
     },
     {
-      "epoch": 3.05,
-      "learning_rate": 1.7393663625562544e-05,
-      "loss": 0.2733,
+      "epoch": 3.1277255198760225,
+      "grad_norm": 0.2013455331325531,
+      "learning_rate": 1.6255342728125295e-05,
+      "loss": 0.3667,
       "step": 86785
     },
     {
-      "epoch": 3.05,
-      "learning_rate": 1.7390950008334194e-05,
-      "loss": 0.2609,
+      "epoch": 3.1279057195372473,
+      "grad_norm": 0.26369357109069824,
+      "learning_rate": 1.6252608993954132e-05,
+      "loss": 0.4069,
       "step": 86790
     },
     {
-      "epoch": 3.05,
-      "learning_rate": 1.738823648990434e-05,
-      "loss": 0.2555,
+      "epoch": 3.128085919198472,
+      "grad_norm": 0.23781251907348633,
+      "learning_rate": 1.6249875378964475e-05,
+      "loss": 0.3672,
       "step": 86795
     },
     {
-      "epoch": 3.05,
-      "learning_rate": 1.7385523070308218e-05,
-      "loss": 0.2658,
+      "epoch": 3.1282661188596967,
+      "grad_norm": 0.1853938102722168,
+      "learning_rate": 1.6247141883193578e-05,
+      "loss": 0.3969,
       "step": 86800
     },
     {
-      "epoch": 3.05,
-      "learning_rate": 1.738280974958105e-05,
-      "loss": 0.2902,
+      "epoch": 3.128446318520921,
+      "grad_norm": 0.1940661072731018,
+      "learning_rate": 1.624440850667867e-05,
+      "loss": 0.3897,
       "step": 86805
     },
     {
-      "epoch": 3.05,
-      "learning_rate": 1.738009652775809e-05,
-      "loss": 0.2581,
+      "epoch": 3.1286265181821458,
+      "grad_norm": 0.21411295235157013,
+      "learning_rate": 1.624167524945701e-05,
+      "loss": 0.386,
       "step": 86810
     },
     {
-      "epoch": 3.05,
-      "learning_rate": 1.737738340487455e-05,
-      "loss": 0.2536,
+      "epoch": 3.1288067178433705,
+      "grad_norm": 0.19931474328041077,
+      "learning_rate": 1.6238942111565826e-05,
+      "loss": 0.4114,
       "step": 86815
     },
     {
-      "epoch": 3.05,
-      "learning_rate": 1.7374670380965668e-05,
-      "loss": 0.2576,
+      "epoch": 3.1289869175045952,
+      "grad_norm": 0.19459813833236694,
+      "learning_rate": 1.6236209093042355e-05,
+      "loss": 0.3829,
       "step": 86820
     },
     {
-      "epoch": 3.05,
-      "learning_rate": 1.7371957456066657e-05,
-      "loss": 0.2596,
+      "epoch": 3.1291671171658195,
+      "grad_norm": 0.18339698016643524,
+      "learning_rate": 1.623347619392384e-05,
+      "loss": 0.3981,
       "step": 86825
     },
     {
-      "epoch": 3.05,
-      "learning_rate": 1.7369244630212757e-05,
-      "loss": 0.2531,
+      "epoch": 3.1293473168270443,
+      "grad_norm": 0.21319806575775146,
+      "learning_rate": 1.6230743414247522e-05,
+      "loss": 0.3896,
       "step": 86830
     },
     {
-      "epoch": 3.06,
-      "learning_rate": 1.7366531903439184e-05,
-      "loss": 0.2439,
+      "epoch": 3.129527516488269,
+      "grad_norm": 0.2697208523750305,
+      "learning_rate": 1.6228010754050615e-05,
+      "loss": 0.4042,
       "step": 86835
     },
     {
-      "epoch": 3.06,
-      "learning_rate": 1.7363819275781168e-05,
-      "loss": 0.2591,
+      "epoch": 3.1297077161494937,
+      "grad_norm": 0.2193807065486908,
+      "learning_rate": 1.6225278213370373e-05,
+      "loss": 0.4035,
       "step": 86840
     },
     {
-      "epoch": 3.06,
-      "learning_rate": 1.7361106747273918e-05,
-      "loss": 0.2785,
+      "epoch": 3.1298879158107185,
+      "grad_norm": 0.195016011595726,
+      "learning_rate": 1.622254579224401e-05,
+      "loss": 0.3613,
       "step": 86845
     },
     {
-      "epoch": 3.06,
-      "learning_rate": 1.7358394317952668e-05,
-      "loss": 0.2839,
+      "epoch": 3.1300681154719427,
+      "grad_norm": 0.19086046516895294,
+      "learning_rate": 1.6219813490708767e-05,
+      "loss": 0.4081,
       "step": 86850
     },
     {
-      "epoch": 3.06,
-      "learning_rate": 1.735568198785264e-05,
-      "loss": 0.2684,
+      "epoch": 3.1302483151331675,
+      "grad_norm": 0.1884526163339615,
+      "learning_rate": 1.621708130880187e-05,
+      "loss": 0.3415,
       "step": 86855
     },
     {
-      "epoch": 3.06,
-      "learning_rate": 1.7352969757009025e-05,
-      "loss": 0.2733,
+      "epoch": 3.130428514794392,
+      "grad_norm": 0.2053939402103424,
+      "learning_rate": 1.6214349246560522e-05,
+      "loss": 0.38,
       "step": 86860
     },
     {
-      "epoch": 3.06,
-      "learning_rate": 1.735025762545707e-05,
-      "loss": 0.2549,
+      "epoch": 3.130608714455617,
+      "grad_norm": 0.18308919668197632,
+      "learning_rate": 1.6211617304021977e-05,
+      "loss": 0.3578,
       "step": 86865
     },
     {
-      "epoch": 3.06,
-      "learning_rate": 1.734754559323198e-05,
-      "loss": 0.2776,
+      "epoch": 3.1307889141168417,
+      "grad_norm": 0.21582022309303284,
+      "learning_rate": 1.620888548122345e-05,
+      "loss": 0.3648,
       "step": 86870
     },
     {
-      "epoch": 3.06,
-      "learning_rate": 1.7344833660368967e-05,
-      "loss": 0.2678,
+      "epoch": 3.130969113778066,
+      "grad_norm": 0.2522687315940857,
+      "learning_rate": 1.6206153778202144e-05,
+      "loss": 0.3881,
       "step": 86875
     },
     {
-      "epoch": 3.06,
-      "learning_rate": 1.7342121826903235e-05,
-      "loss": 0.2734,
+      "epoch": 3.1311493134392907,
+      "grad_norm": 0.2382495403289795,
+      "learning_rate": 1.62034221949953e-05,
+      "loss": 0.3718,
       "step": 86880
     },
     {
-      "epoch": 3.06,
-      "learning_rate": 1.7339410092870013e-05,
-      "loss": 0.2364,
+      "epoch": 3.1313295131005154,
+      "grad_norm": 0.2659468650817871,
+      "learning_rate": 1.6200690731640112e-05,
+      "loss": 0.3654,
       "step": 86885
     },
     {
-      "epoch": 3.06,
-      "learning_rate": 1.7336698458304497e-05,
-      "loss": 0.2606,
+      "epoch": 3.13150971276174,
+      "grad_norm": 0.24328552186489105,
+      "learning_rate": 1.6197959388173825e-05,
+      "loss": 0.3785,
       "step": 86890
     },
     {
-      "epoch": 3.06,
-      "learning_rate": 1.7333986923241906e-05,
-      "loss": 0.2835,
+      "epoch": 3.1316899124229645,
+      "grad_norm": 0.22692778706550598,
+      "learning_rate": 1.6195228164633634e-05,
+      "loss": 0.363,
       "step": 86895
     },
     {
-      "epoch": 3.06,
-      "learning_rate": 1.7331275487717437e-05,
-      "loss": 0.239,
+      "epoch": 3.131870112084189,
+      "grad_norm": 0.24278061091899872,
+      "learning_rate": 1.6192497061056747e-05,
+      "loss": 0.3806,
       "step": 86900
     },
     {
-      "epoch": 3.06,
-      "learning_rate": 1.732856415176631e-05,
-      "loss": 0.2545,
+      "epoch": 3.132050311745414,
+      "grad_norm": 0.23407572507858276,
+      "learning_rate": 1.6189766077480386e-05,
+      "loss": 0.3659,
       "step": 86905
     },
     {
-      "epoch": 3.06,
-      "learning_rate": 1.732585291542372e-05,
-      "loss": 0.2363,
+      "epoch": 3.1322305114066387,
+      "grad_norm": 0.24145354330539703,
+      "learning_rate": 1.6187035213941754e-05,
+      "loss": 0.3704,
       "step": 86910
     },
     {
-      "epoch": 3.06,
-      "learning_rate": 1.732314177872487e-05,
-      "loss": 0.2535,
+      "epoch": 3.1324107110678634,
+      "grad_norm": 0.18080276250839233,
+      "learning_rate": 1.618430447047807e-05,
+      "loss": 0.3888,
       "step": 86915
     },
     {
-      "epoch": 3.06,
-      "learning_rate": 1.7320430741704962e-05,
-      "loss": 0.266,
+      "epoch": 3.1325909107290877,
+      "grad_norm": 0.2304733842611313,
+      "learning_rate": 1.6181573847126525e-05,
+      "loss": 0.3904,
       "step": 86920
     },
     {
-      "epoch": 3.06,
-      "learning_rate": 1.73177198043992e-05,
-      "loss": 0.2702,
+      "epoch": 3.1327711103903124,
+      "grad_norm": 0.261940598487854,
+      "learning_rate": 1.617884334392432e-05,
+      "loss": 0.3794,
       "step": 86925
     },
     {
-      "epoch": 3.06,
-      "learning_rate": 1.7315008966842788e-05,
-      "loss": 0.2557,
+      "epoch": 3.132951310051537,
+      "grad_norm": 0.21325023472309113,
+      "learning_rate": 1.617611296090868e-05,
+      "loss": 0.3466,
       "step": 86930
     },
     {
-      "epoch": 3.06,
-      "learning_rate": 1.731229822907092e-05,
-      "loss": 0.2587,
+      "epoch": 3.133131509712762,
+      "grad_norm": 0.21175888180732727,
+      "learning_rate": 1.617338269811679e-05,
+      "loss": 0.3665,
       "step": 86935
     },
     {
-      "epoch": 3.06,
-      "learning_rate": 1.7309587591118787e-05,
-      "loss": 0.2599,
+      "epoch": 3.133311709373986,
+      "grad_norm": 0.2089497148990631,
+      "learning_rate": 1.6170652555585847e-05,
+      "loss": 0.3469,
       "step": 86940
     },
     {
-      "epoch": 3.06,
-      "learning_rate": 1.7306877053021602e-05,
-      "loss": 0.2537,
+      "epoch": 3.133491909035211,
+      "grad_norm": 0.2231149524450302,
+      "learning_rate": 1.6167922533353057e-05,
+      "loss": 0.3621,
       "step": 86945
     },
     {
-      "epoch": 3.06,
-      "learning_rate": 1.7304166614814538e-05,
-      "loss": 0.2791,
+      "epoch": 3.1336721086964356,
+      "grad_norm": 0.19638942182064056,
+      "learning_rate": 1.6165192631455605e-05,
+      "loss": 0.3358,
       "step": 86950
     },
     {
-      "epoch": 3.06,
-      "learning_rate": 1.7301456276532808e-05,
-      "loss": 0.2666,
+      "epoch": 3.1338523083576604,
+      "grad_norm": 0.22249744832515717,
+      "learning_rate": 1.6162462849930704e-05,
+      "loss": 0.3728,
       "step": 86955
     },
     {
-      "epoch": 3.06,
-      "learning_rate": 1.729874603821158e-05,
-      "loss": 0.2499,
+      "epoch": 3.134032508018885,
+      "grad_norm": 0.27574974298477173,
+      "learning_rate": 1.6159733188815537e-05,
+      "loss": 0.3883,
       "step": 86960
     },
     {
-      "epoch": 3.06,
-      "learning_rate": 1.729603589988607e-05,
-      "loss": 0.2402,
+      "epoch": 3.1342127076801094,
+      "grad_norm": 0.23349010944366455,
+      "learning_rate": 1.615700364814728e-05,
+      "loss": 0.3996,
       "step": 86965
     },
     {
-      "epoch": 3.06,
-      "learning_rate": 1.7293325861591456e-05,
-      "loss": 0.2629,
+      "epoch": 3.134392907341334,
+      "grad_norm": 0.2402789145708084,
+      "learning_rate": 1.615427422796315e-05,
+      "loss": 0.4117,
       "step": 86970
     },
     {
-      "epoch": 3.06,
-      "learning_rate": 1.729061592336293e-05,
-      "loss": 0.2554,
+      "epoch": 3.134573107002559,
+      "grad_norm": 0.2527726888656616,
+      "learning_rate": 1.6151544928300317e-05,
+      "loss": 0.3607,
       "step": 86975
     },
     {
-      "epoch": 3.06,
-      "learning_rate": 1.7287906085235662e-05,
-      "loss": 0.2523,
+      "epoch": 3.1347533066637836,
+      "grad_norm": 0.2573539614677429,
+      "learning_rate": 1.614881574919596e-05,
+      "loss": 0.3634,
       "step": 86980
     },
     {
-      "epoch": 3.06,
-      "learning_rate": 1.7285196347244868e-05,
-      "loss": 0.2578,
+      "epoch": 3.134933506325008,
+      "grad_norm": 0.2581014037132263,
+      "learning_rate": 1.6146086690687294e-05,
+      "loss": 0.3679,
       "step": 86985
     },
     {
-      "epoch": 3.06,
-      "learning_rate": 1.728248670942571e-05,
-      "loss": 0.2509,
+      "epoch": 3.1351137059862326,
+      "grad_norm": 0.2501007616519928,
+      "learning_rate": 1.6143357752811463e-05,
+      "loss": 0.3852,
       "step": 86990
     },
     {
-      "epoch": 3.06,
-      "learning_rate": 1.7279777171813366e-05,
-      "loss": 0.2498,
+      "epoch": 3.1352939056474574,
+      "grad_norm": 0.2757335603237152,
+      "learning_rate": 1.6140628935605684e-05,
+      "loss": 0.3801,
       "step": 86995
     },
     {
-      "epoch": 3.06,
-      "learning_rate": 1.7277067734443038e-05,
-      "loss": 0.2589,
+      "epoch": 3.135474105308682,
+      "grad_norm": 0.22444647550582886,
+      "learning_rate": 1.6137900239107118e-05,
+      "loss": 0.4285,
       "step": 87000
     },
     {
-      "epoch": 3.06,
-      "eval_loss": 0.25700682401657104,
-      "eval_runtime": 10.5666,
-      "eval_samples_per_second": 9.464,
-      "eval_steps_per_second": 9.464,
+      "epoch": 3.135474105308682,
+      "eval_loss": 0.43149107694625854,
+      "eval_runtime": 3.5463,
+      "eval_samples_per_second": 28.199,
+      "eval_steps_per_second": 7.05,
       "step": 87000
     },
     {
-      "epoch": 3.06,
-      "learning_rate": 1.7274358397349895e-05,
-      "loss": 0.2571,
+      "epoch": 3.135654304969907,
+      "grad_norm": 0.20283716917037964,
+      "learning_rate": 1.613517166335294e-05,
+      "loss": 0.3473,
       "step": 87005
     },
     {
-      "epoch": 3.06,
-      "learning_rate": 1.727164916056912e-05,
-      "loss": 0.2686,
+      "epoch": 3.135834504631131,
+      "grad_norm": 0.2266281247138977,
+      "learning_rate": 1.6132443208380333e-05,
+      "loss": 0.3585,
       "step": 87010
     },
     {
-      "epoch": 3.06,
-      "learning_rate": 1.726894002413587e-05,
-      "loss": 0.2568,
+      "epoch": 3.136014704292356,
+      "grad_norm": 0.21769681572914124,
+      "learning_rate": 1.6129714874226473e-05,
+      "loss": 0.3786,
       "step": 87015
     },
     {
-      "epoch": 3.06,
-      "learning_rate": 1.726623098808536e-05,
-      "loss": 0.2401,
+      "epoch": 3.1361949039535806,
+      "grad_norm": 0.19556161761283875,
+      "learning_rate": 1.6126986660928525e-05,
+      "loss": 0.3848,
       "step": 87020
     },
     {
-      "epoch": 3.06,
-      "learning_rate": 1.7263522052452734e-05,
-      "loss": 0.2597,
+      "epoch": 3.1363751036148053,
+      "grad_norm": 0.2081514149904251,
+      "learning_rate": 1.6124258568523672e-05,
+      "loss": 0.4028,
       "step": 87025
     },
     {
-      "epoch": 3.06,
-      "learning_rate": 1.7260813217273178e-05,
-      "loss": 0.2436,
+      "epoch": 3.13655530327603,
+      "grad_norm": 0.2071877270936966,
+      "learning_rate": 1.6121530597049072e-05,
+      "loss": 0.4077,
       "step": 87030
     },
     {
-      "epoch": 3.06,
-      "learning_rate": 1.7258104482581854e-05,
-      "loss": 0.2491,
+      "epoch": 3.1367355029372543,
+      "grad_norm": 0.1915891468524933,
+      "learning_rate": 1.61188027465419e-05,
+      "loss": 0.3885,
       "step": 87035
     },
     {
-      "epoch": 3.06,
-      "learning_rate": 1.7255395848413954e-05,
-      "loss": 0.25,
+      "epoch": 3.136915702598479,
+      "grad_norm": 0.2584543228149414,
+      "learning_rate": 1.611607501703933e-05,
+      "loss": 0.3638,
       "step": 87040
     },
     {
-      "epoch": 3.06,
-      "learning_rate": 1.725268731480463e-05,
-      "loss": 0.2528,
+      "epoch": 3.137095902259704,
+      "grad_norm": 0.20328520238399506,
+      "learning_rate": 1.6113347408578504e-05,
+      "loss": 0.3868,
       "step": 87045
     },
     {
-      "epoch": 3.06,
-      "learning_rate": 1.7249978881789053e-05,
-      "loss": 0.2766,
+      "epoch": 3.1372761019209285,
+      "grad_norm": 0.2620421350002289,
+      "learning_rate": 1.6110619921196607e-05,
+      "loss": 0.4069,
       "step": 87050
     },
     {
-      "epoch": 3.06,
-      "learning_rate": 1.7247270549402393e-05,
-      "loss": 0.2714,
+      "epoch": 3.137456301582153,
+      "grad_norm": 0.20489437878131866,
+      "learning_rate": 1.61078925549308e-05,
+      "loss": 0.352,
       "step": 87055
     },
     {
-      "epoch": 3.06,
-      "learning_rate": 1.7244562317679816e-05,
-      "loss": 0.2443,
+      "epoch": 3.1376365012433776,
+      "grad_norm": 0.2264631986618042,
+      "learning_rate": 1.610516530981822e-05,
+      "loss": 0.3975,
       "step": 87060
     },
     {
-      "epoch": 3.06,
-      "learning_rate": 1.7241854186656493e-05,
-      "loss": 0.2568,
+      "epoch": 3.1378167009046023,
+      "grad_norm": 0.21898336708545685,
+      "learning_rate": 1.6102438185896052e-05,
+      "loss": 0.3688,
       "step": 87065
     },
     {
-      "epoch": 3.06,
-      "learning_rate": 1.723914615636758e-05,
-      "loss": 0.2646,
+      "epoch": 3.137996900565827,
+      "grad_norm": 0.21995946764945984,
+      "learning_rate": 1.6099711183201427e-05,
+      "loss": 0.388,
       "step": 87070
     },
     {
-      "epoch": 3.06,
-      "learning_rate": 1.7236438226848224e-05,
-      "loss": 0.2628,
+      "epoch": 3.1381771002270518,
+      "grad_norm": 0.2543283998966217,
+      "learning_rate": 1.6096984301771535e-05,
+      "loss": 0.3847,
       "step": 87075
     },
     {
-      "epoch": 3.06,
-      "learning_rate": 1.7233730398133616e-05,
-      "loss": 0.2443,
+      "epoch": 3.138357299888276,
+      "grad_norm": 0.210366889834404,
+      "learning_rate": 1.6094257541643494e-05,
+      "loss": 0.4029,
       "step": 87080
     },
     {
-      "epoch": 3.06,
-      "learning_rate": 1.7231022670258894e-05,
-      "loss": 0.2564,
+      "epoch": 3.138537499549501,
+      "grad_norm": 0.22760994732379913,
+      "learning_rate": 1.609153090285447e-05,
+      "loss": 0.4536,
       "step": 87085
     },
     {
-      "epoch": 3.06,
-      "learning_rate": 1.722831504325923e-05,
-      "loss": 0.2679,
+      "epoch": 3.1387176992107255,
+      "grad_norm": 0.23996824026107788,
+      "learning_rate": 1.6088804385441616e-05,
+      "loss": 0.3483,
       "step": 87090
     },
     {
-      "epoch": 3.06,
-      "learning_rate": 1.722560751716976e-05,
-      "loss": 0.2628,
+      "epoch": 3.1388978988719503,
+      "grad_norm": 0.2677284777164459,
+      "learning_rate": 1.608607798944207e-05,
+      "loss": 0.3906,
       "step": 87095
     },
     {
-      "epoch": 3.06,
-      "learning_rate": 1.7222900092025668e-05,
-      "loss": 0.2385,
+      "epoch": 3.139078098533175,
+      "grad_norm": 0.28425002098083496,
+      "learning_rate": 1.6083351714892997e-05,
+      "loss": 0.3969,
       "step": 87100
     },
     {
-      "epoch": 3.06,
-      "learning_rate": 1.7220192767862085e-05,
-      "loss": 0.2463,
+      "epoch": 3.1392582981943993,
+      "grad_norm": 0.2697744369506836,
+      "learning_rate": 1.6080625561831528e-05,
+      "loss": 0.3645,
       "step": 87105
     },
     {
-      "epoch": 3.06,
-      "learning_rate": 1.7217485544714167e-05,
-      "loss": 0.2827,
+      "epoch": 3.139438497855624,
+      "grad_norm": 0.1847570687532425,
+      "learning_rate": 1.60778995302948e-05,
+      "loss": 0.3889,
       "step": 87110
     },
     {
-      "epoch": 3.06,
-      "learning_rate": 1.7214778422617077e-05,
-      "loss": 0.2391,
+      "epoch": 3.1396186975168487,
+      "grad_norm": 0.2160649597644806,
+      "learning_rate": 1.6075173620319972e-05,
+      "loss": 0.3851,
       "step": 87115
     },
     {
-      "epoch": 3.07,
-      "learning_rate": 1.7212071401605962e-05,
-      "loss": 0.2367,
+      "epoch": 3.1397988971780735,
+      "grad_norm": 0.22959807515144348,
+      "learning_rate": 1.6072447831944177e-05,
+      "loss": 0.36,
       "step": 87120
     },
     {
-      "epoch": 3.07,
-      "learning_rate": 1.7209364481715968e-05,
-      "loss": 0.2658,
+      "epoch": 3.1399790968392978,
+      "grad_norm": 0.2232397347688675,
+      "learning_rate": 1.6069722165204544e-05,
+      "loss": 0.3793,
       "step": 87125
     },
     {
-      "epoch": 3.07,
-      "learning_rate": 1.7206657662982227e-05,
-      "loss": 0.245,
+      "epoch": 3.1401592965005225,
+      "grad_norm": 0.22282375395298004,
+      "learning_rate": 1.6066996620138224e-05,
+      "loss": 0.3569,
       "step": 87130
     },
     {
-      "epoch": 3.07,
-      "learning_rate": 1.7203950945439917e-05,
-      "loss": 0.261,
+      "epoch": 3.1403394961617472,
+      "grad_norm": 0.25350263714790344,
+      "learning_rate": 1.6064271196782337e-05,
+      "loss": 0.4028,
       "step": 87135
     },
     {
-      "epoch": 3.07,
-      "learning_rate": 1.720124432912416e-05,
-      "loss": 0.2542,
+      "epoch": 3.140519695822972,
+      "grad_norm": 0.19245287775993347,
+      "learning_rate": 1.6061545895174036e-05,
+      "loss": 0.3635,
       "step": 87140
     },
     {
-      "epoch": 3.07,
-      "learning_rate": 1.719853781407011e-05,
-      "loss": 0.2666,
+      "epoch": 3.1406998954841967,
+      "grad_norm": 0.18547998368740082,
+      "learning_rate": 1.6058820715350438e-05,
+      "loss": 0.3974,
       "step": 87145
     },
     {
-      "epoch": 3.07,
-      "learning_rate": 1.719583140031289e-05,
-      "loss": 0.2677,
+      "epoch": 3.140880095145421,
+      "grad_norm": 0.20061656832695007,
+      "learning_rate": 1.605609565734867e-05,
+      "loss": 0.3442,
       "step": 87150
     },
     {
-      "epoch": 3.07,
-      "learning_rate": 1.7193125087887673e-05,
-      "loss": 0.27,
+      "epoch": 3.1410602948066457,
+      "grad_norm": 0.2687702178955078,
+      "learning_rate": 1.6053370721205877e-05,
+      "loss": 0.3709,
       "step": 87155
     },
     {
-      "epoch": 3.07,
-      "learning_rate": 1.7190418876829577e-05,
-      "loss": 0.2322,
+      "epoch": 3.1412404944678705,
+      "grad_norm": 0.22766318917274475,
+      "learning_rate": 1.6050645906959178e-05,
+      "loss": 0.3863,
       "step": 87160
     },
     {
-      "epoch": 3.07,
-      "learning_rate": 1.7187712767173747e-05,
-      "loss": 0.2537,
+      "epoch": 3.141420694129095,
+      "grad_norm": 0.21267448365688324,
+      "learning_rate": 1.6047921214645677e-05,
+      "loss": 0.3719,
       "step": 87165
     },
     {
-      "epoch": 3.07,
-      "learning_rate": 1.718500675895531e-05,
-      "loss": 0.2604,
+      "epoch": 3.1416008937903195,
+      "grad_norm": 0.19360697269439697,
+      "learning_rate": 1.604519664430254e-05,
+      "loss": 0.39,
       "step": 87170
     },
     {
-      "epoch": 3.07,
-      "learning_rate": 1.7182300852209416e-05,
-      "loss": 0.2583,
+      "epoch": 3.141781093451544,
+      "grad_norm": 0.1883769929409027,
+      "learning_rate": 1.6042472195966843e-05,
+      "loss": 0.3667,
       "step": 87175
     },
     {
-      "epoch": 3.07,
-      "learning_rate": 1.7179595046971196e-05,
-      "loss": 0.2391,
+      "epoch": 3.141961293112769,
+      "grad_norm": 0.22920463979244232,
+      "learning_rate": 1.6039747869675747e-05,
+      "loss": 0.3861,
       "step": 87180
     },
     {
-      "epoch": 3.07,
-      "learning_rate": 1.717688934327578e-05,
-      "loss": 0.2704,
+      "epoch": 3.1421414927739937,
+      "grad_norm": 0.23282212018966675,
+      "learning_rate": 1.6037023665466345e-05,
+      "loss": 0.3846,
       "step": 87185
     },
     {
-      "epoch": 3.07,
-      "learning_rate": 1.717418374115829e-05,
-      "loss": 0.2694,
+      "epoch": 3.1423216924352184,
+      "grad_norm": 0.23070281744003296,
+      "learning_rate": 1.6034299583375752e-05,
+      "loss": 0.408,
       "step": 87190
     },
     {
-      "epoch": 3.07,
-      "learning_rate": 1.717147824065387e-05,
-      "loss": 0.2692,
+      "epoch": 3.1425018920964427,
+      "grad_norm": 0.22182480990886688,
+      "learning_rate": 1.60315756234411e-05,
+      "loss": 0.3623,
       "step": 87195
     },
     {
-      "epoch": 3.07,
-      "learning_rate": 1.7168772841797658e-05,
-      "loss": 0.2604,
+      "epoch": 3.1426820917576674,
+      "grad_norm": 0.23669365048408508,
+      "learning_rate": 1.6028851785699496e-05,
+      "loss": 0.3811,
       "step": 87200
     },
     {
-      "epoch": 3.07,
-      "learning_rate": 1.7166067544624765e-05,
-      "loss": 0.2707,
+      "epoch": 3.142862291418892,
+      "grad_norm": 0.2434239238500595,
+      "learning_rate": 1.602612807018804e-05,
+      "loss": 0.4078,
       "step": 87205
     },
     {
-      "epoch": 3.07,
-      "learning_rate": 1.716336234917031e-05,
-      "loss": 0.2606,
+      "epoch": 3.143042491080117,
+      "grad_norm": 0.190851092338562,
+      "learning_rate": 1.602340447694386e-05,
+      "loss": 0.363,
       "step": 87210
     },
     {
-      "epoch": 3.07,
-      "learning_rate": 1.7160657255469442e-05,
-      "loss": 0.236,
+      "epoch": 3.143222690741341,
+      "grad_norm": 0.26526564359664917,
+      "learning_rate": 1.6020681006004046e-05,
+      "loss": 0.4028,
       "step": 87215
     },
     {
-      "epoch": 3.07,
-      "learning_rate": 1.715795226355727e-05,
-      "loss": 0.2382,
+      "epoch": 3.143402890402566,
+      "grad_norm": 0.2453896403312683,
+      "learning_rate": 1.6017957657405724e-05,
+      "loss": 0.3361,
       "step": 87220
     },
     {
-      "epoch": 3.07,
-      "learning_rate": 1.7155247373468912e-05,
-      "loss": 0.2898,
+      "epoch": 3.1435830900637907,
+      "grad_norm": 0.28423580527305603,
+      "learning_rate": 1.601523443118599e-05,
+      "loss": 0.3749,
       "step": 87225
     },
     {
-      "epoch": 3.07,
-      "learning_rate": 1.715254258523951e-05,
-      "loss": 0.2667,
+      "epoch": 3.1437632897250154,
+      "grad_norm": 0.24882547557353973,
+      "learning_rate": 1.601251132738194e-05,
+      "loss": 0.4034,
       "step": 87230
     },
     {
-      "epoch": 3.07,
-      "learning_rate": 1.714983789890417e-05,
-      "loss": 0.2648,
+      "epoch": 3.14394348938624,
+      "grad_norm": 0.2349334955215454,
+      "learning_rate": 1.600978834603069e-05,
+      "loss": 0.3975,
       "step": 87235
     },
     {
-      "epoch": 3.07,
-      "learning_rate": 1.7147133314498008e-05,
-      "loss": 0.2587,
+      "epoch": 3.1441236890474644,
+      "grad_norm": 0.2091837227344513,
+      "learning_rate": 1.6007065487169338e-05,
+      "loss": 0.3942,
       "step": 87240
     },
     {
-      "epoch": 3.07,
-      "learning_rate": 1.714442883205614e-05,
-      "loss": 0.2615,
+      "epoch": 3.144303888708689,
+      "grad_norm": 0.19795401394367218,
+      "learning_rate": 1.6004342750834965e-05,
+      "loss": 0.3833,
       "step": 87245
     },
     {
-      "epoch": 3.07,
-      "learning_rate": 1.714172445161369e-05,
-      "loss": 0.2606,
+      "epoch": 3.144484088369914,
+      "grad_norm": 0.21962173283100128,
+      "learning_rate": 1.6001620137064694e-05,
+      "loss": 0.3709,
       "step": 87250
     },
     {
-      "epoch": 3.07,
-      "learning_rate": 1.713902017320577e-05,
-      "loss": 0.2385,
+      "epoch": 3.1446642880311386,
+      "grad_norm": 0.2462879866361618,
+      "learning_rate": 1.5998897645895587e-05,
+      "loss": 0.3942,
       "step": 87255
     },
     {
-      "epoch": 3.07,
-      "learning_rate": 1.7136315996867497e-05,
-      "loss": 0.243,
+      "epoch": 3.144844487692363,
+      "grad_norm": 0.21551023423671722,
+      "learning_rate": 1.5996175277364777e-05,
+      "loss": 0.3778,
       "step": 87260
     },
     {
-      "epoch": 3.07,
-      "learning_rate": 1.713361192263397e-05,
-      "loss": 0.2508,
+      "epoch": 3.1450246873535876,
+      "grad_norm": 0.23634564876556396,
+      "learning_rate": 1.599345303150933e-05,
+      "loss": 0.3598,
       "step": 87265
     },
     {
-      "epoch": 3.07,
-      "learning_rate": 1.713090795054031e-05,
-      "loss": 0.2614,
+      "epoch": 3.1452048870148124,
+      "grad_norm": 0.18265362083911896,
+      "learning_rate": 1.599073090836633e-05,
+      "loss": 0.3772,
       "step": 87270
     },
     {
-      "epoch": 3.07,
-      "learning_rate": 1.7128204080621627e-05,
-      "loss": 0.2454,
+      "epoch": 3.145385086676037,
+      "grad_norm": 0.1682208925485611,
+      "learning_rate": 1.5988008907972894e-05,
+      "loss": 0.336,
       "step": 87275
     },
     {
-      "epoch": 3.07,
-      "learning_rate": 1.712550031291303e-05,
-      "loss": 0.2599,
+      "epoch": 3.145565286337262,
+      "grad_norm": 0.28196775913238525,
+      "learning_rate": 1.5985287030366086e-05,
+      "loss": 0.4021,
       "step": 87280
     },
     {
-      "epoch": 3.07,
-      "learning_rate": 1.712279664744961e-05,
-      "loss": 0.2705,
+      "epoch": 3.145745485998486,
+      "grad_norm": 0.23696090281009674,
+      "learning_rate": 1.598256527558299e-05,
+      "loss": 0.4015,
       "step": 87285
     },
     {
-      "epoch": 3.07,
-      "learning_rate": 1.7120093084266493e-05,
-      "loss": 0.2621,
+      "epoch": 3.145925685659711,
+      "grad_norm": 0.24363209307193756,
+      "learning_rate": 1.5979843643660703e-05,
+      "loss": 0.3606,
       "step": 87290
     },
     {
-      "epoch": 3.07,
-      "learning_rate": 1.7117389623398776e-05,
-      "loss": 0.242,
+      "epoch": 3.1461058853209356,
+      "grad_norm": 0.24763990938663483,
+      "learning_rate": 1.5977122134636297e-05,
+      "loss": 0.3774,
       "step": 87295
     },
     {
-      "epoch": 3.07,
-      "learning_rate": 1.7114686264881557e-05,
-      "loss": 0.2857,
+      "epoch": 3.1462860849821603,
+      "grad_norm": 0.23061999678611755,
+      "learning_rate": 1.597440074854686e-05,
+      "loss": 0.3816,
       "step": 87300
     },
     {
-      "epoch": 3.07,
-      "learning_rate": 1.7111983008749937e-05,
-      "loss": 0.2446,
+      "epoch": 3.146466284643385,
+      "grad_norm": 0.22484789788722992,
+      "learning_rate": 1.5971679485429457e-05,
+      "loss": 0.3834,
       "step": 87305
     },
     {
-      "epoch": 3.07,
-      "learning_rate": 1.7109279855039023e-05,
-      "loss": 0.2541,
+      "epoch": 3.1466464843046094,
+      "grad_norm": 0.24081788957118988,
+      "learning_rate": 1.5968958345321178e-05,
+      "loss": 0.3798,
       "step": 87310
     },
     {
-      "epoch": 3.07,
-      "learning_rate": 1.7106576803783913e-05,
-      "loss": 0.2635,
+      "epoch": 3.146826683965834,
+      "grad_norm": 0.21690495312213898,
+      "learning_rate": 1.5966237328259092e-05,
+      "loss": 0.3613,
       "step": 87315
     },
     {
-      "epoch": 3.07,
-      "learning_rate": 1.71038738550197e-05,
-      "loss": 0.2522,
+      "epoch": 3.147006883627059,
+      "grad_norm": 0.1942356824874878,
+      "learning_rate": 1.5963516434280275e-05,
+      "loss": 0.3946,
       "step": 87320
     },
     {
-      "epoch": 3.07,
-      "learning_rate": 1.710117100878147e-05,
-      "loss": 0.2472,
+      "epoch": 3.1471870832882836,
+      "grad_norm": 0.23814404010772705,
+      "learning_rate": 1.5960795663421796e-05,
+      "loss": 0.4,
       "step": 87325
     },
     {
-      "epoch": 3.07,
-      "learning_rate": 1.7098468265104336e-05,
-      "loss": 0.2591,
+      "epoch": 3.147367282949508,
+      "grad_norm": 0.20741520822048187,
+      "learning_rate": 1.595807501572073e-05,
+      "loss": 0.3594,
       "step": 87330
     },
     {
-      "epoch": 3.07,
-      "learning_rate": 1.709576562402339e-05,
-      "loss": 0.2769,
+      "epoch": 3.1475474826107326,
+      "grad_norm": 0.24436409771442413,
+      "learning_rate": 1.595535449121413e-05,
+      "loss": 0.383,
       "step": 87335
     },
     {
-      "epoch": 3.07,
-      "learning_rate": 1.709306308557371e-05,
-      "loss": 0.285,
+      "epoch": 3.1477276822719573,
+      "grad_norm": 0.24962151050567627,
+      "learning_rate": 1.5952634089939083e-05,
+      "loss": 0.4004,
       "step": 87340
     },
     {
-      "epoch": 3.07,
-      "learning_rate": 1.7090360649790398e-05,
-      "loss": 0.2444,
+      "epoch": 3.147907881933182,
+      "grad_norm": 0.2126917541027069,
+      "learning_rate": 1.5949913811932654e-05,
+      "loss": 0.4005,
       "step": 87345
     },
     {
-      "epoch": 3.07,
-      "learning_rate": 1.708765831670855e-05,
-      "loss": 0.2723,
+      "epoch": 3.148088081594407,
+      "grad_norm": 0.19167962670326233,
+      "learning_rate": 1.5947193657231878e-05,
+      "loss": 0.3429,
       "step": 87350
     },
     {
-      "epoch": 3.07,
-      "learning_rate": 1.7084956086363228e-05,
-      "loss": 0.2508,
+      "epoch": 3.148268281255631,
+      "grad_norm": 0.22857992351055145,
+      "learning_rate": 1.594447362587385e-05,
+      "loss": 0.368,
       "step": 87355
     },
     {
-      "epoch": 3.07,
-      "learning_rate": 1.7082253958789545e-05,
-      "loss": 0.2719,
+      "epoch": 3.148448480916856,
+      "grad_norm": 0.25584468245506287,
+      "learning_rate": 1.5941753717895604e-05,
+      "loss": 0.4065,
       "step": 87360
     },
     {
-      "epoch": 3.07,
-      "learning_rate": 1.707955193402257e-05,
-      "loss": 0.2668,
+      "epoch": 3.1486286805780805,
+      "grad_norm": 0.25212666392326355,
+      "learning_rate": 1.5939033933334222e-05,
+      "loss": 0.3659,
       "step": 87365
     },
     {
-      "epoch": 3.07,
-      "learning_rate": 1.7076850012097405e-05,
-      "loss": 0.2523,
+      "epoch": 3.1488088802393053,
+      "grad_norm": 0.21378327906131744,
+      "learning_rate": 1.5936314272226743e-05,
+      "loss": 0.3726,
       "step": 87370
     },
     {
-      "epoch": 3.07,
-      "learning_rate": 1.7074148193049116e-05,
-      "loss": 0.2612,
+      "epoch": 3.14898907990053,
+      "grad_norm": 0.2281913012266159,
+      "learning_rate": 1.5933594734610225e-05,
+      "loss": 0.3816,
       "step": 87375
     },
     {
-      "epoch": 3.07,
-      "learning_rate": 1.7071446476912776e-05,
-      "loss": 0.2518,
+      "epoch": 3.1491692795617543,
+      "grad_norm": 0.21938398480415344,
+      "learning_rate": 1.5930875320521728e-05,
+      "loss": 0.3704,
       "step": 87380
     },
     {
-      "epoch": 3.07,
-      "learning_rate": 1.7068744863723496e-05,
-      "loss": 0.2548,
+      "epoch": 3.149349479222979,
+      "grad_norm": 0.23328323662281036,
+      "learning_rate": 1.5928156029998296e-05,
+      "loss": 0.3732,
       "step": 87385
     },
     {
-      "epoch": 3.07,
-      "learning_rate": 1.706604335351633e-05,
-      "loss": 0.2838,
+      "epoch": 3.1495296788842038,
+      "grad_norm": 0.26327353715896606,
+      "learning_rate": 1.5925436863076975e-05,
+      "loss": 0.3539,
       "step": 87390
     },
     {
-      "epoch": 3.07,
-      "learning_rate": 1.7063341946326366e-05,
-      "loss": 0.2316,
+      "epoch": 3.1497098785454285,
+      "grad_norm": 0.2839113175868988,
+      "learning_rate": 1.592271781979483e-05,
+      "loss": 0.3575,
       "step": 87395
     },
     {
-      "epoch": 3.07,
-      "learning_rate": 1.7060640642188664e-05,
-      "loss": 0.2461,
+      "epoch": 3.149890078206653,
+      "grad_norm": 0.23041868209838867,
+      "learning_rate": 1.5919998900188894e-05,
+      "loss": 0.3847,
       "step": 87400
     },
     {
-      "epoch": 3.08,
-      "learning_rate": 1.7057939441138328e-05,
-      "loss": 0.2472,
+      "epoch": 3.1500702778678775,
+      "grad_norm": 0.20966331660747528,
+      "learning_rate": 1.5917280104296213e-05,
+      "loss": 0.3631,
       "step": 87405
     },
     {
-      "epoch": 3.08,
-      "learning_rate": 1.70552383432104e-05,
-      "loss": 0.2563,
+      "epoch": 3.1502504775291023,
+      "grad_norm": 0.25130027532577515,
+      "learning_rate": 1.5914561432153836e-05,
+      "loss": 0.4388,
       "step": 87410
     },
     {
-      "epoch": 3.08,
-      "learning_rate": 1.7052537348439977e-05,
-      "loss": 0.2647,
+      "epoch": 3.150430677190327,
+      "grad_norm": 0.20757076144218445,
+      "learning_rate": 1.5911842883798796e-05,
+      "loss": 0.3988,
       "step": 87415
     },
     {
-      "epoch": 3.08,
-      "learning_rate": 1.704983645686211e-05,
-      "loss": 0.2284,
+      "epoch": 3.1506108768515517,
+      "grad_norm": 0.19644039869308472,
+      "learning_rate": 1.590912445926814e-05,
+      "loss": 0.386,
       "step": 87420
     },
     {
-      "epoch": 3.08,
-      "learning_rate": 1.704713566851188e-05,
-      "loss": 0.2526,
+      "epoch": 3.150791076512776,
+      "grad_norm": 0.23030658066272736,
+      "learning_rate": 1.590640615859891e-05,
+      "loss": 0.3882,
       "step": 87425
     },
     {
-      "epoch": 3.08,
-      "learning_rate": 1.7044434983424363e-05,
-      "loss": 0.2665,
+      "epoch": 3.1509712761740007,
+      "grad_norm": 0.24640274047851562,
+      "learning_rate": 1.590368798182812e-05,
+      "loss": 0.3869,
       "step": 87430
     },
     {
-      "epoch": 3.08,
-      "learning_rate": 1.70417344016346e-05,
-      "loss": 0.2559,
+      "epoch": 3.1511514758352255,
+      "grad_norm": 0.24621576070785522,
+      "learning_rate": 1.5900969928992832e-05,
+      "loss": 0.3922,
       "step": 87435
     },
     {
-      "epoch": 3.08,
-      "learning_rate": 1.7039033923177676e-05,
-      "loss": 0.255,
+      "epoch": 3.15133167549645,
+      "grad_norm": 0.2239450216293335,
+      "learning_rate": 1.589825200013006e-05,
+      "loss": 0.3877,
       "step": 87440
     },
     {
-      "epoch": 3.08,
-      "learning_rate": 1.7036333548088652e-05,
-      "loss": 0.2554,
+      "epoch": 3.1515118751576745,
+      "grad_norm": 0.22715716063976288,
+      "learning_rate": 1.5895534195276846e-05,
+      "loss": 0.3739,
       "step": 87445
     },
     {
-      "epoch": 3.08,
-      "learning_rate": 1.703363327640259e-05,
-      "loss": 0.2613,
+      "epoch": 3.1516920748188992,
+      "grad_norm": 0.22491981089115143,
+      "learning_rate": 1.5892816514470223e-05,
+      "loss": 0.363,
       "step": 87450
     },
     {
-      "epoch": 3.08,
-      "learning_rate": 1.7030933108154544e-05,
-      "loss": 0.2755,
+      "epoch": 3.151872274480124,
+      "grad_norm": 0.23676250874996185,
+      "learning_rate": 1.5890098957747197e-05,
+      "loss": 0.3657,
       "step": 87455
     },
     {
-      "epoch": 3.08,
-      "learning_rate": 1.7028233043379584e-05,
-      "loss": 0.2605,
+      "epoch": 3.1520524741413487,
+      "grad_norm": 0.22871702909469604,
+      "learning_rate": 1.5887381525144823e-05,
+      "loss": 0.4015,
       "step": 87460
     },
     {
-      "epoch": 3.08,
-      "learning_rate": 1.7025533082112768e-05,
-      "loss": 0.2295,
+      "epoch": 3.1522326738025734,
+      "grad_norm": 0.2012801617383957,
+      "learning_rate": 1.5884664216700107e-05,
+      "loss": 0.3338,
       "step": 87465
     },
     {
-      "epoch": 3.08,
-      "learning_rate": 1.7022833224389144e-05,
-      "loss": 0.2576,
+      "epoch": 3.1524128734637977,
+      "grad_norm": 0.2810083031654358,
+      "learning_rate": 1.588194703245007e-05,
+      "loss": 0.3743,
       "step": 87470
     },
     {
-      "epoch": 3.08,
-      "learning_rate": 1.7020133470243768e-05,
-      "loss": 0.27,
+      "epoch": 3.1525930731250225,
+      "grad_norm": 0.21618230640888214,
+      "learning_rate": 1.5879229972431748e-05,
+      "loss": 0.3838,
       "step": 87475
     },
     {
-      "epoch": 3.08,
-      "learning_rate": 1.7017433819711707e-05,
-      "loss": 0.2392,
+      "epoch": 3.152773272786247,
+      "grad_norm": 0.23298971354961395,
+      "learning_rate": 1.5876513036682143e-05,
+      "loss": 0.359,
       "step": 87480
     },
     {
-      "epoch": 3.08,
-      "learning_rate": 1.7014734272828016e-05,
-      "loss": 0.2587,
+      "epoch": 3.152953472447472,
+      "grad_norm": 0.23836317658424377,
+      "learning_rate": 1.5873796225238284e-05,
+      "loss": 0.3654,
       "step": 87485
     },
     {
-      "epoch": 3.08,
-      "learning_rate": 1.7012034829627727e-05,
-      "loss": 0.252,
+      "epoch": 3.153133672108696,
+      "grad_norm": 0.20612043142318726,
+      "learning_rate": 1.587107953813719e-05,
+      "loss": 0.3643,
       "step": 87490
     },
     {
-      "epoch": 3.08,
-      "learning_rate": 1.70093354901459e-05,
-      "loss": 0.2711,
+      "epoch": 3.153313871769921,
+      "grad_norm": 0.20500215888023376,
+      "learning_rate": 1.5868362975415862e-05,
+      "loss": 0.3322,
       "step": 87495
     },
     {
-      "epoch": 3.08,
-      "learning_rate": 1.7006636254417583e-05,
-      "loss": 0.275,
+      "epoch": 3.1534940714311457,
+      "grad_norm": 0.22878599166870117,
+      "learning_rate": 1.586564653711133e-05,
+      "loss": 0.347,
       "step": 87500
     },
     {
-      "epoch": 3.08,
-      "eval_loss": 0.25675198435783386,
-      "eval_runtime": 10.5938,
-      "eval_samples_per_second": 9.439,
-      "eval_steps_per_second": 9.439,
+      "epoch": 3.1534940714311457,
+      "eval_loss": 0.4312251806259155,
+      "eval_runtime": 3.5283,
+      "eval_samples_per_second": 28.342,
+      "eval_steps_per_second": 7.086,
       "step": 87500
     },
     {
-      "epoch": 3.08,
-      "learning_rate": 1.700393712247784e-05,
-      "loss": 0.2602,
+      "epoch": 3.1536742710923704,
+      "grad_norm": 0.23596695065498352,
+      "learning_rate": 1.5862930223260587e-05,
+      "loss": 0.4027,
       "step": 87505
     },
     {
-      "epoch": 3.08,
-      "learning_rate": 1.70012380943617e-05,
-      "loss": 0.2764,
+      "epoch": 3.153854470753595,
+      "grad_norm": 0.2374979853630066,
+      "learning_rate": 1.5860214033900657e-05,
+      "loss": 0.3555,
       "step": 87510
     },
     {
-      "epoch": 3.08,
-      "learning_rate": 1.69985391701042e-05,
-      "loss": 0.2733,
+      "epoch": 3.1540346704148194,
+      "grad_norm": 0.2398587167263031,
+      "learning_rate": 1.5857497969068535e-05,
+      "loss": 0.3814,
       "step": 87515
     },
     {
-      "epoch": 3.08,
-      "learning_rate": 1.699584034974041e-05,
-      "loss": 0.2469,
+      "epoch": 3.154214870076044,
+      "grad_norm": 0.1979122906923294,
+      "learning_rate": 1.5854782028801235e-05,
+      "loss": 0.3692,
       "step": 87520
     },
     {
-      "epoch": 3.08,
-      "learning_rate": 1.699314163330535e-05,
-      "loss": 0.2683,
+      "epoch": 3.154395069737269,
+      "grad_norm": 0.23171058297157288,
+      "learning_rate": 1.5852066213135764e-05,
+      "loss": 0.3734,
       "step": 87525
     },
     {
-      "epoch": 3.08,
-      "learning_rate": 1.6990443020834073e-05,
-      "loss": 0.2443,
+      "epoch": 3.1545752693984936,
+      "grad_norm": 0.221579447388649,
+      "learning_rate": 1.5849350522109124e-05,
+      "loss": 0.3978,
       "step": 87530
     },
     {
-      "epoch": 3.08,
-      "learning_rate": 1.6987744512361607e-05,
-      "loss": 0.2555,
+      "epoch": 3.1547554690597184,
+      "grad_norm": 0.23226810991764069,
+      "learning_rate": 1.5846634955758294e-05,
+      "loss": 0.3808,
       "step": 87535
     },
     {
-      "epoch": 3.08,
-      "learning_rate": 1.698504610792301e-05,
-      "loss": 0.2549,
+      "epoch": 3.1549356687209427,
+      "grad_norm": 0.25539645552635193,
+      "learning_rate": 1.584391951412031e-05,
+      "loss": 0.3677,
       "step": 87540
     },
     {
-      "epoch": 3.08,
-      "learning_rate": 1.69823478075533e-05,
-      "loss": 0.268,
+      "epoch": 3.1551158683821674,
+      "grad_norm": 0.23339438438415527,
+      "learning_rate": 1.5841204197232128e-05,
+      "loss": 0.4242,
       "step": 87545
     },
     {
-      "epoch": 3.08,
-      "learning_rate": 1.697964961128752e-05,
-      "loss": 0.2675,
+      "epoch": 3.155296068043392,
+      "grad_norm": 0.20781894028186798,
+      "learning_rate": 1.5838489005130784e-05,
+      "loss": 0.3916,
       "step": 87550
     },
     {
-      "epoch": 3.08,
-      "learning_rate": 1.69769515191607e-05,
-      "loss": 0.2432,
+      "epoch": 3.155476267704617,
+      "grad_norm": 0.2721303701400757,
+      "learning_rate": 1.5835773937853254e-05,
+      "loss": 0.3906,
       "step": 87555
     },
     {
-      "epoch": 3.08,
-      "learning_rate": 1.6974253531207886e-05,
-      "loss": 0.2515,
+      "epoch": 3.155656467365841,
+      "grad_norm": 0.20113010704517365,
+      "learning_rate": 1.5833058995436512e-05,
+      "loss": 0.3924,
       "step": 87560
     },
     {
-      "epoch": 3.08,
-      "learning_rate": 1.6971555647464103e-05,
-      "loss": 0.2636,
+      "epoch": 3.155836667027066,
+      "grad_norm": 0.22510038316249847,
+      "learning_rate": 1.5830344177917587e-05,
+      "loss": 0.4007,
       "step": 87565
     },
     {
-      "epoch": 3.08,
-      "learning_rate": 1.6968857867964367e-05,
-      "loss": 0.2557,
+      "epoch": 3.1560168666882906,
+      "grad_norm": 0.2468447983264923,
+      "learning_rate": 1.582762948533344e-05,
+      "loss": 0.3602,
       "step": 87570
     },
     {
-      "epoch": 3.08,
-      "learning_rate": 1.6966160192743727e-05,
-      "loss": 0.2648,
+      "epoch": 3.1561970663495154,
+      "grad_norm": 0.22670796513557434,
+      "learning_rate": 1.582491491772106e-05,
+      "loss": 0.3703,
       "step": 87575
     },
     {
-      "epoch": 3.08,
-      "learning_rate": 1.69634626218372e-05,
-      "loss": 0.2296,
+      "epoch": 3.15637726601074,
+      "grad_norm": 0.2570918798446655,
+      "learning_rate": 1.5822200475117442e-05,
+      "loss": 0.3967,
       "step": 87580
     },
     {
-      "epoch": 3.08,
-      "learning_rate": 1.6960765155279818e-05,
-      "loss": 0.2627,
+      "epoch": 3.1565574656719644,
+      "grad_norm": 0.23322150111198425,
+      "learning_rate": 1.581948615755956e-05,
+      "loss": 0.416,
       "step": 87585
     },
     {
-      "epoch": 3.08,
-      "learning_rate": 1.695806779310659e-05,
-      "loss": 0.2599,
+      "epoch": 3.156737665333189,
+      "grad_norm": 0.20641043782234192,
+      "learning_rate": 1.581677196508441e-05,
+      "loss": 0.3588,
       "step": 87590
     },
     {
-      "epoch": 3.08,
-      "learning_rate": 1.6955370535352566e-05,
-      "loss": 0.2587,
+      "epoch": 3.156917864994414,
+      "grad_norm": 0.21482810378074646,
+      "learning_rate": 1.5814057897728964e-05,
+      "loss": 0.3548,
       "step": 87595
     },
     {
-      "epoch": 3.08,
-      "learning_rate": 1.695267338205275e-05,
-      "loss": 0.2623,
+      "epoch": 3.1570980646556386,
+      "grad_norm": 0.2732873260974884,
+      "learning_rate": 1.5811343955530194e-05,
+      "loss": 0.3961,
       "step": 87600
     },
     {
-      "epoch": 3.08,
-      "learning_rate": 1.694997633324217e-05,
-      "loss": 0.2587,
+      "epoch": 3.1572782643168633,
+      "grad_norm": 0.23953673243522644,
+      "learning_rate": 1.5808630138525088e-05,
+      "loss": 0.4481,
       "step": 87605
     },
     {
-      "epoch": 3.08,
-      "learning_rate": 1.694727938895583e-05,
-      "loss": 0.2652,
+      "epoch": 3.1574584639780876,
+      "grad_norm": 0.23516350984573364,
+      "learning_rate": 1.580591644675062e-05,
+      "loss": 0.3804,
       "step": 87610
     },
     {
-      "epoch": 3.08,
-      "learning_rate": 1.694458254922877e-05,
-      "loss": 0.2439,
+      "epoch": 3.1576386636393123,
+      "grad_norm": 0.21418847143650055,
+      "learning_rate": 1.5803202880243754e-05,
+      "loss": 0.3417,
       "step": 87615
     },
     {
-      "epoch": 3.08,
-      "learning_rate": 1.6941885814096004e-05,
-      "loss": 0.245,
+      "epoch": 3.157818863300537,
+      "grad_norm": 0.21651922166347504,
+      "learning_rate": 1.580048943904147e-05,
+      "loss": 0.3643,
       "step": 87620
     },
     {
-      "epoch": 3.08,
-      "learning_rate": 1.6939189183592537e-05,
-      "loss": 0.2677,
+      "epoch": 3.157999062961762,
+      "grad_norm": 0.309662789106369,
+      "learning_rate": 1.5797776123180736e-05,
+      "loss": 0.4107,
       "step": 87625
     },
     {
-      "epoch": 3.08,
-      "learning_rate": 1.6936492657753373e-05,
-      "loss": 0.266,
+      "epoch": 3.158179262622986,
+      "grad_norm": 0.2053297758102417,
+      "learning_rate": 1.5795062932698522e-05,
+      "loss": 0.3574,
       "step": 87630
     },
     {
-      "epoch": 3.08,
-      "learning_rate": 1.693379623661354e-05,
-      "loss": 0.253,
+      "epoch": 3.158359462284211,
+      "grad_norm": 0.2058108001947403,
+      "learning_rate": 1.57923498676318e-05,
+      "loss": 0.3742,
       "step": 87635
     },
     {
-      "epoch": 3.08,
-      "learning_rate": 1.6931099920208058e-05,
-      "loss": 0.2574,
+      "epoch": 3.1585396619454356,
+      "grad_norm": 0.22333082556724548,
+      "learning_rate": 1.5789636928017514e-05,
+      "loss": 0.3823,
       "step": 87640
     },
     {
-      "epoch": 3.08,
-      "learning_rate": 1.6928403708571924e-05,
-      "loss": 0.2643,
+      "epoch": 3.1587198616066603,
+      "grad_norm": 0.2727280259132385,
+      "learning_rate": 1.5786924113892655e-05,
+      "loss": 0.3995,
       "step": 87645
     },
     {
-      "epoch": 3.08,
-      "learning_rate": 1.6925707601740136e-05,
-      "loss": 0.2571,
+      "epoch": 3.158900061267885,
+      "grad_norm": 0.2285049855709076,
+      "learning_rate": 1.578421142529417e-05,
+      "loss": 0.3648,
       "step": 87650
     },
     {
-      "epoch": 3.08,
-      "learning_rate": 1.692301159974773e-05,
-      "loss": 0.2617,
+      "epoch": 3.1590802609291093,
+      "grad_norm": 0.24408283829689026,
+      "learning_rate": 1.5781498862259007e-05,
+      "loss": 0.4071,
       "step": 87655
     },
     {
-      "epoch": 3.08,
-      "learning_rate": 1.6920315702629685e-05,
-      "loss": 0.2567,
+      "epoch": 3.159260460590334,
+      "grad_norm": 0.23655414581298828,
+      "learning_rate": 1.577878642482415e-05,
+      "loss": 0.3995,
       "step": 87660
     },
     {
-      "epoch": 3.08,
-      "learning_rate": 1.6917619910421025e-05,
-      "loss": 0.2675,
+      "epoch": 3.159440660251559,
+      "grad_norm": 0.21964088082313538,
+      "learning_rate": 1.5776074113026533e-05,
+      "loss": 0.3779,
       "step": 87665
     },
     {
-      "epoch": 3.08,
-      "learning_rate": 1.6914924223156727e-05,
-      "loss": 0.2791,
+      "epoch": 3.1596208599127835,
+      "grad_norm": 0.18904650211334229,
+      "learning_rate": 1.577336192690313e-05,
+      "loss": 0.404,
       "step": 87670
     },
     {
-      "epoch": 3.08,
-      "learning_rate": 1.6912228640871823e-05,
-      "loss": 0.26,
+      "epoch": 3.159801059574008,
+      "grad_norm": 0.2197270691394806,
+      "learning_rate": 1.577064986649088e-05,
+      "loss": 0.3763,
       "step": 87675
     },
     {
-      "epoch": 3.08,
-      "learning_rate": 1.6909533163601305e-05,
-      "loss": 0.2729,
+      "epoch": 3.1599812592352325,
+      "grad_norm": 0.2542881667613983,
+      "learning_rate": 1.576793793182674e-05,
+      "loss": 0.4108,
       "step": 87680
     },
     {
-      "epoch": 3.09,
-      "learning_rate": 1.6906837791380152e-05,
-      "loss": 0.2604,
+      "epoch": 3.1601614588964573,
+      "grad_norm": 0.2189742773771286,
+      "learning_rate": 1.5765768474658893e-05,
+      "loss": 0.4039,
       "step": 87685
     },
     {
-      "epoch": 3.09,
-      "learning_rate": 1.6904142524243383e-05,
-      "loss": 0.2677,
+      "epoch": 3.160341658557682,
+      "grad_norm": 0.2676211893558502,
+      "learning_rate": 1.576305676643446e-05,
+      "loss": 0.3957,
       "step": 87690
     },
     {
-      "epoch": 3.09,
-      "learning_rate": 1.6901447362225992e-05,
-      "loss": 0.2759,
+      "epoch": 3.1605218582189067,
+      "grad_norm": 0.2037624716758728,
+      "learning_rate": 1.5760345184061586e-05,
+      "loss": 0.3458,
       "step": 87695
     },
     {
-      "epoch": 3.09,
-      "learning_rate": 1.689875230536297e-05,
-      "loss": 0.2578,
+      "epoch": 3.160702057880131,
+      "grad_norm": 0.21000508964061737,
+      "learning_rate": 1.575763372757723e-05,
+      "loss": 0.3576,
       "step": 87700
     },
     {
-      "epoch": 3.09,
-      "learning_rate": 1.6896057353689305e-05,
-      "loss": 0.2391,
+      "epoch": 3.1608822575413558,
+      "grad_norm": 0.27110737562179565,
+      "learning_rate": 1.5754922397018323e-05,
+      "loss": 0.35,
       "step": 87705
     },
     {
-      "epoch": 3.09,
-      "learning_rate": 1.6893362507239998e-05,
-      "loss": 0.252,
+      "epoch": 3.1610624572025805,
+      "grad_norm": 0.24699506163597107,
+      "learning_rate": 1.5752211192421794e-05,
+      "loss": 0.3493,
       "step": 87710
     },
     {
-      "epoch": 3.09,
-      "learning_rate": 1.6890667766050032e-05,
-      "loss": 0.2659,
+      "epoch": 3.1612426568638052,
+      "grad_norm": 0.25574126839637756,
+      "learning_rate": 1.574950011382462e-05,
+      "loss": 0.3795,
       "step": 87715
     },
     {
-      "epoch": 3.09,
-      "learning_rate": 1.688797313015441e-05,
-      "loss": 0.2584,
+      "epoch": 3.1614228565250295,
+      "grad_norm": 0.24138563871383667,
+      "learning_rate": 1.5746789161263702e-05,
+      "loss": 0.3382,
       "step": 87720
     },
     {
-      "epoch": 3.09,
-      "learning_rate": 1.68852785995881e-05,
-      "loss": 0.2694,
+      "epoch": 3.1616030561862543,
+      "grad_norm": 0.2855709493160248,
+      "learning_rate": 1.574407833477599e-05,
+      "loss": 0.3956,
       "step": 87725
     },
     {
-      "epoch": 3.09,
-      "learning_rate": 1.6882584174386106e-05,
-      "loss": 0.2415,
+      "epoch": 3.161783255847479,
+      "grad_norm": 0.1987280547618866,
+      "learning_rate": 1.574136763439842e-05,
+      "loss": 0.3667,
       "step": 87730
     },
     {
-      "epoch": 3.09,
-      "learning_rate": 1.687988985458341e-05,
-      "loss": 0.2396,
+      "epoch": 3.1619634555087037,
+      "grad_norm": 0.19212450087070465,
+      "learning_rate": 1.5738657060167916e-05,
+      "loss": 0.3855,
       "step": 87735
     },
     {
-      "epoch": 3.09,
-      "learning_rate": 1.6877195640214987e-05,
-      "loss": 0.2591,
+      "epoch": 3.1621436551699285,
+      "grad_norm": 0.20062510669231415,
+      "learning_rate": 1.573594661212143e-05,
+      "loss": 0.3914,
       "step": 87740
     },
     {
-      "epoch": 3.09,
-      "learning_rate": 1.6874501531315828e-05,
-      "loss": 0.2512,
+      "epoch": 3.1623238548311527,
+      "grad_norm": 0.24137015640735626,
+      "learning_rate": 1.5733236290295873e-05,
+      "loss": 0.3458,
       "step": 87745
     },
     {
-      "epoch": 3.09,
-      "learning_rate": 1.6871807527920907e-05,
-      "loss": 0.2601,
+      "epoch": 3.1625040544923775,
+      "grad_norm": 0.1870691478252411,
+      "learning_rate": 1.5730526094728173e-05,
+      "loss": 0.3664,
       "step": 87750
     },
     {
-      "epoch": 3.09,
-      "learning_rate": 1.6869113630065217e-05,
-      "loss": 0.2553,
+      "epoch": 3.162684254153602,
+      "grad_norm": 0.21274974942207336,
+      "learning_rate": 1.5727816025455267e-05,
+      "loss": 0.3506,
       "step": 87755
     },
     {
-      "epoch": 3.09,
-      "learning_rate": 1.6866419837783727e-05,
-      "loss": 0.2651,
+      "epoch": 3.162864453814827,
+      "grad_norm": 0.21418865025043488,
+      "learning_rate": 1.5725106082514066e-05,
+      "loss": 0.4101,
       "step": 87760
     },
     {
-      "epoch": 3.09,
-      "learning_rate": 1.6863726151111403e-05,
-      "loss": 0.244,
+      "epoch": 3.1630446534760512,
+      "grad_norm": 0.2074471414089203,
+      "learning_rate": 1.5722396265941504e-05,
+      "loss": 0.3705,
       "step": 87765
     },
     {
-      "epoch": 3.09,
-      "learning_rate": 1.686103257008324e-05,
-      "loss": 0.2777,
+      "epoch": 3.163224853137276,
+      "grad_norm": 0.26662349700927734,
+      "learning_rate": 1.5719686575774494e-05,
+      "loss": 0.3715,
       "step": 87770
     },
     {
-      "epoch": 3.09,
-      "learning_rate": 1.6858339094734212e-05,
-      "loss": 0.2576,
+      "epoch": 3.1634050527985007,
+      "grad_norm": 0.22255268692970276,
+      "learning_rate": 1.5716977012049958e-05,
+      "loss": 0.3973,
       "step": 87775
     },
     {
-      "epoch": 3.09,
-      "learning_rate": 1.6855645725099283e-05,
-      "loss": 0.2737,
+      "epoch": 3.1635852524597254,
+      "grad_norm": 0.2301638126373291,
+      "learning_rate": 1.5714267574804814e-05,
+      "loss": 0.3646,
       "step": 87780
     },
     {
-      "epoch": 3.09,
-      "learning_rate": 1.6852952461213416e-05,
-      "loss": 0.2445,
+      "epoch": 3.16376545212095,
+      "grad_norm": 0.2992319166660309,
+      "learning_rate": 1.5711558264075978e-05,
+      "loss": 0.3886,
       "step": 87785
     },
     {
-      "epoch": 3.09,
-      "learning_rate": 1.6850259303111606e-05,
-      "loss": 0.2697,
+      "epoch": 3.1639456517821745,
+      "grad_norm": 0.2510296106338501,
+      "learning_rate": 1.5708849079900355e-05,
+      "loss": 0.3778,
       "step": 87790
     },
     {
-      "epoch": 3.09,
-      "learning_rate": 1.68475662508288e-05,
-      "loss": 0.247,
+      "epoch": 3.164125851443399,
+      "grad_norm": 0.23558221757411957,
+      "learning_rate": 1.5706140022314874e-05,
+      "loss": 0.3634,
       "step": 87795
     },
     {
-      "epoch": 3.09,
-      "learning_rate": 1.6844873304399977e-05,
-      "loss": 0.2542,
+      "epoch": 3.164306051104624,
+      "grad_norm": 0.1713656485080719,
+      "learning_rate": 1.570343109135643e-05,
+      "loss": 0.3438,
       "step": 87800
     },
     {
-      "epoch": 3.09,
-      "learning_rate": 1.6842180463860095e-05,
-      "loss": 0.2503,
+      "epoch": 3.1644862507658487,
+      "grad_norm": 0.26629698276519775,
+      "learning_rate": 1.570072228706194e-05,
+      "loss": 0.3836,
       "step": 87805
     },
     {
-      "epoch": 3.09,
-      "learning_rate": 1.683948772924413e-05,
-      "loss": 0.2418,
+      "epoch": 3.1646664504270734,
+      "grad_norm": 0.2552080750465393,
+      "learning_rate": 1.5698013609468314e-05,
+      "loss": 0.3914,
       "step": 87810
     },
     {
-      "epoch": 3.09,
-      "learning_rate": 1.683679510058704e-05,
-      "loss": 0.2551,
+      "epoch": 3.1648466500882977,
+      "grad_norm": 0.2017066776752472,
+      "learning_rate": 1.5695305058612436e-05,
+      "loss": 0.3464,
       "step": 87815
     },
     {
-      "epoch": 3.09,
-      "learning_rate": 1.6834102577923777e-05,
-      "loss": 0.275,
+      "epoch": 3.1650268497495224,
+      "grad_norm": 0.18393248319625854,
+      "learning_rate": 1.569259663453124e-05,
+      "loss": 0.4058,
       "step": 87820
     },
     {
-      "epoch": 3.09,
-      "learning_rate": 1.6831410161289317e-05,
-      "loss": 0.2568,
+      "epoch": 3.165207049410747,
+      "grad_norm": 0.2161942571401596,
+      "learning_rate": 1.568988833726161e-05,
+      "loss": 0.3795,
       "step": 87825
     },
     {
-      "epoch": 3.09,
-      "learning_rate": 1.682871785071861e-05,
-      "loss": 0.2469,
+      "epoch": 3.165387249071972,
+      "grad_norm": 0.21547988057136536,
+      "learning_rate": 1.5687180166840442e-05,
+      "loss": 0.3861,
       "step": 87830
     },
     {
-      "epoch": 3.09,
-      "learning_rate": 1.6826025646246623e-05,
-      "loss": 0.2618,
+      "epoch": 3.165567448733196,
+      "grad_norm": 0.27021363377571106,
+      "learning_rate": 1.5684472123304648e-05,
+      "loss": 0.3855,
       "step": 87835
     },
     {
-      "epoch": 3.09,
-      "learning_rate": 1.6823333547908292e-05,
-      "loss": 0.2836,
+      "epoch": 3.165747648394421,
+      "grad_norm": 0.2427399456501007,
+      "learning_rate": 1.5681764206691114e-05,
+      "loss": 0.3669,
       "step": 87840
     },
     {
-      "epoch": 3.09,
-      "learning_rate": 1.68206415557386e-05,
-      "loss": 0.2508,
+      "epoch": 3.1659278480556456,
+      "grad_norm": 0.2702856957912445,
+      "learning_rate": 1.5679056417036743e-05,
+      "loss": 0.403,
       "step": 87845
     },
     {
-      "epoch": 3.09,
-      "learning_rate": 1.6817949669772483e-05,
-      "loss": 0.246,
+      "epoch": 3.1661080477168704,
+      "grad_norm": 0.24403263628482819,
+      "learning_rate": 1.567634875437842e-05,
+      "loss": 0.3985,
       "step": 87850
     },
     {
-      "epoch": 3.09,
-      "learning_rate": 1.6815257890044898e-05,
-      "loss": 0.2478,
+      "epoch": 3.166288247378095,
+      "grad_norm": 0.24820050597190857,
+      "learning_rate": 1.567364121875304e-05,
+      "loss": 0.3987,
       "step": 87855
     },
     {
-      "epoch": 3.09,
-      "learning_rate": 1.6812566216590786e-05,
-      "loss": 0.259,
+      "epoch": 3.1664684470393194,
+      "grad_norm": 0.20138351619243622,
+      "learning_rate": 1.567093381019749e-05,
+      "loss": 0.3929,
       "step": 87860
     },
     {
-      "epoch": 3.09,
-      "learning_rate": 1.6809874649445116e-05,
-      "loss": 0.2453,
+      "epoch": 3.166648646700544,
+      "grad_norm": 0.22523026168346405,
+      "learning_rate": 1.566822652874867e-05,
+      "loss": 0.3857,
       "step": 87865
     },
     {
-      "epoch": 3.09,
-      "learning_rate": 1.6807183188642827e-05,
-      "loss": 0.249,
+      "epoch": 3.166828846361769,
+      "grad_norm": 0.20283663272857666,
+      "learning_rate": 1.5665519374443447e-05,
+      "loss": 0.3864,
       "step": 87870
     },
     {
-      "epoch": 3.09,
-      "learning_rate": 1.6804491834218864e-05,
-      "loss": 0.2558,
+      "epoch": 3.1670090460229936,
+      "grad_norm": 0.22071103751659393,
+      "learning_rate": 1.566281234731872e-05,
+      "loss": 0.3571,
       "step": 87875
     },
     {
-      "epoch": 3.09,
-      "learning_rate": 1.6801800586208168e-05,
-      "loss": 0.2504,
+      "epoch": 3.1671892456842183,
+      "grad_norm": 0.262087345123291,
+      "learning_rate": 1.5660105447411364e-05,
+      "loss": 0.365,
       "step": 87880
     },
     {
-      "epoch": 3.09,
-      "learning_rate": 1.679910944464569e-05,
-      "loss": 0.2629,
+      "epoch": 3.1673694453454426,
+      "grad_norm": 0.26408714056015015,
+      "learning_rate": 1.565739867475827e-05,
+      "loss": 0.4295,
       "step": 87885
     },
     {
-      "epoch": 3.09,
-      "learning_rate": 1.679641840956638e-05,
-      "loss": 0.2806,
+      "epoch": 3.1675496450066674,
+      "grad_norm": 0.30089375376701355,
+      "learning_rate": 1.5654692029396308e-05,
+      "loss": 0.4188,
       "step": 87890
     },
     {
-      "epoch": 3.09,
-      "learning_rate": 1.6793727481005168e-05,
-      "loss": 0.2629,
+      "epoch": 3.167729844667892,
+      "grad_norm": 0.24746596813201904,
+      "learning_rate": 1.565198551136235e-05,
+      "loss": 0.3927,
       "step": 87895
     },
     {
-      "epoch": 3.09,
-      "learning_rate": 1.6791036658996984e-05,
-      "loss": 0.2511,
+      "epoch": 3.167910044329117,
+      "grad_norm": 0.2371826320886612,
+      "learning_rate": 1.5649279120693283e-05,
+      "loss": 0.3807,
       "step": 87900
     },
     {
-      "epoch": 3.09,
-      "learning_rate": 1.678834594357679e-05,
-      "loss": 0.2563,
+      "epoch": 3.168090243990341,
+      "grad_norm": 0.2455732524394989,
+      "learning_rate": 1.564657285742599e-05,
+      "loss": 0.3812,
       "step": 87905
     },
     {
-      "epoch": 3.09,
-      "learning_rate": 1.678565533477951e-05,
-      "loss": 0.2638,
+      "epoch": 3.168270443651566,
+      "grad_norm": 0.22298307716846466,
+      "learning_rate": 1.564386672159731e-05,
+      "loss": 0.4142,
       "step": 87910
     },
     {
-      "epoch": 3.09,
-      "learning_rate": 1.6782964832640075e-05,
-      "loss": 0.2509,
+      "epoch": 3.1684506433127906,
+      "grad_norm": 0.24019892513751984,
+      "learning_rate": 1.5641160713244153e-05,
+      "loss": 0.3813,
       "step": 87915
     },
     {
-      "epoch": 3.09,
-      "learning_rate": 1.6780274437193435e-05,
-      "loss": 0.2488,
+      "epoch": 3.1686308429740153,
+      "grad_norm": 0.22240878641605377,
+      "learning_rate": 1.563845483240335e-05,
+      "loss": 0.4095,
       "step": 87920
     },
     {
-      "epoch": 3.09,
-      "learning_rate": 1.6777584148474518e-05,
-      "loss": 0.2512,
+      "epoch": 3.16881104263524,
+      "grad_norm": 0.22564803063869476,
+      "learning_rate": 1.5635749079111807e-05,
+      "loss": 0.3787,
       "step": 87925
     },
     {
-      "epoch": 3.09,
-      "learning_rate": 1.6774893966518242e-05,
-      "loss": 0.2355,
+      "epoch": 3.1689912422964643,
+      "grad_norm": 0.23460455238819122,
+      "learning_rate": 1.5633043453406364e-05,
+      "loss": 0.3668,
       "step": 87930
     },
     {
-      "epoch": 3.09,
-      "learning_rate": 1.677220389135955e-05,
-      "loss": 0.2639,
+      "epoch": 3.169171441957689,
+      "grad_norm": 0.2320113182067871,
+      "learning_rate": 1.5630337955323883e-05,
+      "loss": 0.401,
       "step": 87935
     },
     {
-      "epoch": 3.09,
-      "learning_rate": 1.676951392303337e-05,
-      "loss": 0.2628,
+      "epoch": 3.169351641618914,
+      "grad_norm": 0.21181990206241608,
+      "learning_rate": 1.562763258490124e-05,
+      "loss": 0.379,
       "step": 87940
     },
     {
-      "epoch": 3.09,
-      "learning_rate": 1.6766824061574632e-05,
-      "loss": 0.2853,
+      "epoch": 3.1695318412801385,
+      "grad_norm": 0.23171097040176392,
+      "learning_rate": 1.5624927342175287e-05,
+      "loss": 0.3618,
       "step": 87945
     },
     {
-      "epoch": 3.09,
-      "learning_rate": 1.676413430701826e-05,
-      "loss": 0.2306,
+      "epoch": 3.169712040941363,
+      "grad_norm": 0.2735525965690613,
+      "learning_rate": 1.5622222227182884e-05,
+      "loss": 0.4178,
       "step": 87950
     },
     {
-      "epoch": 3.09,
-      "learning_rate": 1.6761444659399163e-05,
-      "loss": 0.2469,
+      "epoch": 3.1698922406025876,
+      "grad_norm": 0.23134300112724304,
+      "learning_rate": 1.561951723996089e-05,
+      "loss": 0.3827,
       "step": 87955
     },
     {
-      "epoch": 3.09,
-      "learning_rate": 1.6758755118752284e-05,
-      "loss": 0.2654,
+      "epoch": 3.1700724402638123,
+      "grad_norm": 0.21240036189556122,
+      "learning_rate": 1.561681238054615e-05,
+      "loss": 0.382,
       "step": 87960
     },
     {
-      "epoch": 3.09,
-      "learning_rate": 1.6756065685112538e-05,
-      "loss": 0.2666,
+      "epoch": 3.170252639925037,
+      "grad_norm": 0.26621800661087036,
+      "learning_rate": 1.5614107648975532e-05,
+      "loss": 0.3772,
       "step": 87965
     },
     {
-      "epoch": 3.1,
-      "learning_rate": 1.6753376358514852e-05,
-      "loss": 0.2589,
+      "epoch": 3.1704328395862618,
+      "grad_norm": 0.22227172553539276,
+      "learning_rate": 1.561140304528588e-05,
+      "loss": 0.3594,
       "step": 87970
     },
     {
-      "epoch": 3.1,
-      "learning_rate": 1.6750687138994125e-05,
-      "loss": 0.2552,
+      "epoch": 3.170613039247486,
+      "grad_norm": 0.23364649713039398,
+      "learning_rate": 1.5608698569514037e-05,
+      "loss": 0.372,
       "step": 87975
     },
     {
-      "epoch": 3.1,
-      "learning_rate": 1.6747998026585306e-05,
-      "loss": 0.2464,
+      "epoch": 3.170793238908711,
+      "grad_norm": 0.23514457046985626,
+      "learning_rate": 1.5605994221696864e-05,
+      "loss": 0.4162,
       "step": 87980
     },
     {
-      "epoch": 3.1,
-      "learning_rate": 1.6745309021323287e-05,
-      "loss": 0.2492,
+      "epoch": 3.1709734385699355,
+      "grad_norm": 0.25838416814804077,
+      "learning_rate": 1.5603290001871197e-05,
+      "loss": 0.4076,
       "step": 87985
     },
     {
-      "epoch": 3.1,
-      "learning_rate": 1.674262012324299e-05,
-      "loss": 0.2546,
+      "epoch": 3.1711536382311603,
+      "grad_norm": 0.2554473280906677,
+      "learning_rate": 1.560058591007389e-05,
+      "loss": 0.422,
       "step": 87990
     },
     {
-      "epoch": 3.1,
-      "learning_rate": 1.673993133237932e-05,
-      "loss": 0.2565,
+      "epoch": 3.1713338378923845,
+      "grad_norm": 0.23790916800498962,
+      "learning_rate": 1.559788194634178e-05,
+      "loss": 0.3746,
       "step": 87995
     },
     {
-      "epoch": 3.1,
-      "learning_rate": 1.6737242648767215e-05,
-      "loss": 0.2637,
+      "epoch": 3.1715140375536093,
+      "grad_norm": 0.21130967140197754,
+      "learning_rate": 1.55951781107117e-05,
+      "loss": 0.3621,
       "step": 88000
     },
     {
-      "epoch": 3.1,
-      "eval_loss": 0.25597599148750305,
-      "eval_runtime": 10.5513,
-      "eval_samples_per_second": 9.477,
-      "eval_steps_per_second": 9.477,
+      "epoch": 3.1715140375536093,
+      "eval_loss": 0.4313719570636749,
+      "eval_runtime": 3.532,
+      "eval_samples_per_second": 28.313,
+      "eval_steps_per_second": 7.078,
       "step": 88000
     },
     {
-      "epoch": 3.1,
-      "learning_rate": 1.6734554072441563e-05,
-      "loss": 0.2452,
+      "epoch": 3.171694237214834,
+      "grad_norm": 0.2613160312175751,
+      "learning_rate": 1.559247440322051e-05,
+      "loss": 0.3657,
       "step": 88005
     },
     {
-      "epoch": 3.1,
-      "learning_rate": 1.673186560343728e-05,
-      "loss": 0.2522,
+      "epoch": 3.1718744368760587,
+      "grad_norm": 0.17358632385730743,
+      "learning_rate": 1.5589770823905027e-05,
+      "loss": 0.36,
       "step": 88010
     },
     {
-      "epoch": 3.1,
-      "learning_rate": 1.672917724178926e-05,
-      "loss": 0.2383,
+      "epoch": 3.1720546365372835,
+      "grad_norm": 0.2220027893781662,
+      "learning_rate": 1.5587067372802092e-05,
+      "loss": 0.3705,
       "step": 88015
     },
     {
-      "epoch": 3.1,
-      "learning_rate": 1.6726488987532434e-05,
-      "loss": 0.2508,
+      "epoch": 3.1722348361985078,
+      "grad_norm": 0.20974379777908325,
+      "learning_rate": 1.5584364049948548e-05,
+      "loss": 0.3666,
       "step": 88020
     },
     {
-      "epoch": 3.1,
-      "learning_rate": 1.6723800840701702e-05,
-      "loss": 0.266,
+      "epoch": 3.1724150358597325,
+      "grad_norm": 0.20734605193138123,
+      "learning_rate": 1.5581660855381207e-05,
+      "loss": 0.3625,
       "step": 88025
     },
     {
-      "epoch": 3.1,
-      "learning_rate": 1.6721112801331946e-05,
-      "loss": 0.2572,
+      "epoch": 3.1725952355209572,
+      "grad_norm": 0.24206040799617767,
+      "learning_rate": 1.557895778913693e-05,
+      "loss": 0.3759,
       "step": 88030
     },
     {
-      "epoch": 3.1,
-      "learning_rate": 1.67184248694581e-05,
-      "loss": 0.2678,
+      "epoch": 3.172775435182182,
+      "grad_norm": 0.20642173290252686,
+      "learning_rate": 1.5576254851252526e-05,
+      "loss": 0.3891,
       "step": 88035
     },
     {
-      "epoch": 3.1,
-      "learning_rate": 1.6715737045115043e-05,
-      "loss": 0.2583,
+      "epoch": 3.1729556348434063,
+      "grad_norm": 0.23246215283870697,
+      "learning_rate": 1.5573552041764818e-05,
+      "loss": 0.4263,
       "step": 88040
     },
     {
-      "epoch": 3.1,
-      "learning_rate": 1.6713049328337683e-05,
-      "loss": 0.2586,
+      "epoch": 3.173135834504631,
+      "grad_norm": 0.2090509980916977,
+      "learning_rate": 1.557084936071064e-05,
+      "loss": 0.3807,
       "step": 88045
     },
     {
-      "epoch": 3.1,
-      "learning_rate": 1.6710361719160905e-05,
-      "loss": 0.2745,
+      "epoch": 3.1733160341658557,
+      "grad_norm": 0.22444801032543182,
+      "learning_rate": 1.5568146808126823e-05,
+      "loss": 0.3697,
       "step": 88050
     },
     {
-      "epoch": 3.1,
-      "learning_rate": 1.670767421761963e-05,
-      "loss": 0.2549,
+      "epoch": 3.1734962338270805,
+      "grad_norm": 0.24101091921329498,
+      "learning_rate": 1.556544438405017e-05,
+      "loss": 0.3802,
       "step": 88055
     },
     {
-      "epoch": 3.1,
-      "learning_rate": 1.670498682374874e-05,
-      "loss": 0.2587,
+      "epoch": 3.173676433488305,
+      "grad_norm": 0.2750411927700043,
+      "learning_rate": 1.556274208851752e-05,
+      "loss": 0.391,
       "step": 88060
     },
     {
-      "epoch": 3.1,
-      "learning_rate": 1.6702299537583126e-05,
-      "loss": 0.2475,
+      "epoch": 3.1738566331495295,
+      "grad_norm": 0.21167200803756714,
+      "learning_rate": 1.5560039921565667e-05,
+      "loss": 0.3587,
       "step": 88065
     },
     {
-      "epoch": 3.1,
-      "learning_rate": 1.669961235915768e-05,
-      "loss": 0.2702,
+      "epoch": 3.174036832810754,
+      "grad_norm": 0.28249591588974,
+      "learning_rate": 1.5557337883231455e-05,
+      "loss": 0.3977,
       "step": 88070
     },
     {
-      "epoch": 3.1,
-      "learning_rate": 1.6696925288507297e-05,
-      "loss": 0.2693,
+      "epoch": 3.174217032471979,
+      "grad_norm": 0.23094874620437622,
+      "learning_rate": 1.5554635973551684e-05,
+      "loss": 0.3555,
       "step": 88075
     },
     {
-      "epoch": 3.1,
-      "learning_rate": 1.669423832566688e-05,
-      "loss": 0.2599,
+      "epoch": 3.1743972321332037,
+      "grad_norm": 0.24824340641498566,
+      "learning_rate": 1.5551934192563165e-05,
+      "loss": 0.3596,
       "step": 88080
     },
     {
-      "epoch": 3.1,
-      "learning_rate": 1.6691551470671296e-05,
-      "loss": 0.2832,
+      "epoch": 3.1745774317944284,
+      "grad_norm": 0.23412884771823883,
+      "learning_rate": 1.5549232540302718e-05,
+      "loss": 0.4005,
       "step": 88085
     },
     {
-      "epoch": 3.1,
-      "learning_rate": 1.6688864723555436e-05,
-      "loss": 0.255,
+      "epoch": 3.1747576314556527,
+      "grad_norm": 0.2383238971233368,
+      "learning_rate": 1.5546531016807152e-05,
+      "loss": 0.4006,
       "step": 88090
     },
     {
-      "epoch": 3.1,
-      "learning_rate": 1.6686178084354198e-05,
-      "loss": 0.2701,
+      "epoch": 3.1749378311168774,
+      "grad_norm": 0.2111070156097412,
+      "learning_rate": 1.5543829622113253e-05,
+      "loss": 0.3623,
       "step": 88095
     },
     {
-      "epoch": 3.1,
-      "learning_rate": 1.6683491553102454e-05,
-      "loss": 0.2571,
+      "epoch": 3.175118030778102,
+      "grad_norm": 0.2079268842935562,
+      "learning_rate": 1.5541128356257866e-05,
+      "loss": 0.3812,
       "step": 88100
     },
     {
-      "epoch": 3.1,
-      "learning_rate": 1.66808051298351e-05,
-      "loss": 0.2642,
+      "epoch": 3.175298230439327,
+      "grad_norm": 0.2111787348985672,
+      "learning_rate": 1.5538427219277756e-05,
+      "loss": 0.4066,
       "step": 88105
     },
     {
-      "epoch": 3.1,
-      "learning_rate": 1.667811881458699e-05,
-      "loss": 0.2762,
+      "epoch": 3.1754784301005516,
+      "grad_norm": 0.22174887359142303,
+      "learning_rate": 1.5535726211209758e-05,
+      "loss": 0.3579,
       "step": 88110
     },
     {
-      "epoch": 3.1,
-      "learning_rate": 1.667543260739304e-05,
-      "loss": 0.2425,
+      "epoch": 3.175658629761776,
+      "grad_norm": 0.2652639150619507,
+      "learning_rate": 1.5533025332090655e-05,
+      "loss": 0.4119,
       "step": 88115
     },
     {
-      "epoch": 3.1,
-      "learning_rate": 1.667274650828811e-05,
-      "loss": 0.2539,
+      "epoch": 3.1758388294230007,
+      "grad_norm": 0.22351694107055664,
+      "learning_rate": 1.553032458195724e-05,
+      "loss": 0.3969,
       "step": 88120
     },
     {
-      "epoch": 3.1,
-      "learning_rate": 1.6670060517307072e-05,
-      "loss": 0.2692,
+      "epoch": 3.1760190290842254,
+      "grad_norm": 0.2620088756084442,
+      "learning_rate": 1.5527623960846328e-05,
+      "loss": 0.3849,
       "step": 88125
     },
     {
-      "epoch": 3.1,
-      "learning_rate": 1.6667374634484806e-05,
-      "loss": 0.2809,
+      "epoch": 3.17619922874545,
+      "grad_norm": 0.24307698011398315,
+      "learning_rate": 1.55249234687947e-05,
+      "loss": 0.38,
       "step": 88130
     },
     {
-      "epoch": 3.1,
-      "learning_rate": 1.6664688859856196e-05,
-      "loss": 0.2754,
+      "epoch": 3.1763794284066744,
+      "grad_norm": 0.2753578722476959,
+      "learning_rate": 1.5522223105839157e-05,
+      "loss": 0.369,
       "step": 88135
     },
     {
-      "epoch": 3.1,
-      "learning_rate": 1.666200319345611e-05,
-      "loss": 0.2544,
+      "epoch": 3.176559628067899,
+      "grad_norm": 0.21059057116508484,
+      "learning_rate": 1.551952287201649e-05,
+      "loss": 0.4148,
       "step": 88140
     },
     {
-      "epoch": 3.1,
-      "learning_rate": 1.665931763531942e-05,
-      "loss": 0.2284,
+      "epoch": 3.176739827729124,
+      "grad_norm": 0.21120493113994598,
+      "learning_rate": 1.551682276736349e-05,
+      "loss": 0.3609,
       "step": 88145
     },
     {
-      "epoch": 3.1,
-      "learning_rate": 1.665663218548098e-05,
-      "loss": 0.2623,
+      "epoch": 3.1769200273903486,
+      "grad_norm": 0.22285541892051697,
+      "learning_rate": 1.551412279191694e-05,
+      "loss": 0.3983,
       "step": 88150
     },
     {
-      "epoch": 3.1,
-      "learning_rate": 1.665394684397568e-05,
-      "loss": 0.246,
+      "epoch": 3.1771002270515734,
+      "grad_norm": 0.22482122480869293,
+      "learning_rate": 1.5511422945713633e-05,
+      "loss": 0.3582,
       "step": 88155
     },
     {
-      "epoch": 3.1,
-      "learning_rate": 1.6651261610838383e-05,
-      "loss": 0.2496,
+      "epoch": 3.1772804267127976,
+      "grad_norm": 0.21789222955703735,
+      "learning_rate": 1.550872322879035e-05,
+      "loss": 0.3925,
       "step": 88160
     },
     {
-      "epoch": 3.1,
-      "learning_rate": 1.664857648610394e-05,
-      "loss": 0.2721,
+      "epoch": 3.1774606263740224,
+      "grad_norm": 0.21532319486141205,
+      "learning_rate": 1.5506023641183876e-05,
+      "loss": 0.3784,
       "step": 88165
     },
     {
-      "epoch": 3.1,
-      "learning_rate": 1.6645891469807242e-05,
-      "loss": 0.2511,
+      "epoch": 3.177640826035247,
+      "grad_norm": 0.23733597993850708,
+      "learning_rate": 1.550332418293099e-05,
+      "loss": 0.3538,
       "step": 88170
     },
     {
-      "epoch": 3.1,
-      "learning_rate": 1.6643206561983138e-05,
-      "loss": 0.2559,
+      "epoch": 3.177821025696472,
+      "grad_norm": 0.18654675781726837,
+      "learning_rate": 1.550062485406848e-05,
+      "loss": 0.4173,
       "step": 88175
     },
     {
-      "epoch": 3.1,
-      "learning_rate": 1.6640521762666482e-05,
-      "loss": 0.2766,
+      "epoch": 3.178001225357696,
+      "grad_norm": 0.2497045248746872,
+      "learning_rate": 1.5497925654633118e-05,
+      "loss": 0.41,
       "step": 88180
     },
     {
-      "epoch": 3.1,
-      "learning_rate": 1.663783707189214e-05,
-      "loss": 0.2513,
+      "epoch": 3.178181425018921,
+      "grad_norm": 0.27806994318962097,
+      "learning_rate": 1.5495226584661664e-05,
+      "loss": 0.428,
       "step": 88185
     },
     {
-      "epoch": 3.1,
-      "learning_rate": 1.6635152489694976e-05,
-      "loss": 0.2542,
+      "epoch": 3.1783616246801456,
+      "grad_norm": 0.20720724761486053,
+      "learning_rate": 1.549252764419092e-05,
+      "loss": 0.3697,
       "step": 88190
     },
     {
-      "epoch": 3.1,
-      "learning_rate": 1.6632468016109848e-05,
-      "loss": 0.26,
+      "epoch": 3.1785418243413703,
+      "grad_norm": 0.21904359757900238,
+      "learning_rate": 1.548982883325765e-05,
+      "loss": 0.3729,
       "step": 88195
     },
     {
-      "epoch": 3.1,
-      "learning_rate": 1.662978365117161e-05,
-      "loss": 0.2627,
+      "epoch": 3.178722024002595,
+      "grad_norm": 0.2742276191711426,
+      "learning_rate": 1.5487130151898603e-05,
+      "loss": 0.3845,
       "step": 88200
     },
     {
-      "epoch": 3.1,
-      "learning_rate": 1.66270993949151e-05,
-      "loss": 0.275,
+      "epoch": 3.1789022236638194,
+      "grad_norm": 0.2257348746061325,
+      "learning_rate": 1.5484431600150587e-05,
+      "loss": 0.3508,
       "step": 88205
     },
     {
-      "epoch": 3.1,
-      "learning_rate": 1.66244152473752e-05,
-      "loss": 0.2463,
+      "epoch": 3.179082423325044,
+      "grad_norm": 0.23579144477844238,
+      "learning_rate": 1.548173317805033e-05,
+      "loss": 0.3784,
       "step": 88210
     },
     {
-      "epoch": 3.1,
-      "learning_rate": 1.662173120858675e-05,
-      "loss": 0.2575,
+      "epoch": 3.179262622986269,
+      "grad_norm": 0.2549505829811096,
+      "learning_rate": 1.547903488563463e-05,
+      "loss": 0.3866,
       "step": 88215
     },
     {
-      "epoch": 3.1,
-      "learning_rate": 1.6619047278584596e-05,
-      "loss": 0.2504,
+      "epoch": 3.1794428226474936,
+      "grad_norm": 0.2171338051557541,
+      "learning_rate": 1.5476336722940233e-05,
+      "loss": 0.3705,
       "step": 88220
     },
     {
-      "epoch": 3.1,
-      "learning_rate": 1.661636345740358e-05,
-      "loss": 0.2562,
+      "epoch": 3.179623022308718,
+      "grad_norm": 0.21523796021938324,
+      "learning_rate": 1.54736386900039e-05,
+      "loss": 0.3732,
       "step": 88225
     },
     {
-      "epoch": 3.1,
-      "learning_rate": 1.661367974507857e-05,
-      "loss": 0.2457,
+      "epoch": 3.1798032219699426,
+      "grad_norm": 0.22669003903865814,
+      "learning_rate": 1.5470940786862397e-05,
+      "loss": 0.3728,
       "step": 88230
     },
     {
-      "epoch": 3.1,
-      "learning_rate": 1.6610996141644397e-05,
-      "loss": 0.2446,
+      "epoch": 3.1799834216311673,
+      "grad_norm": 0.20555952191352844,
+      "learning_rate": 1.5468243013552486e-05,
+      "loss": 0.3688,
       "step": 88235
     },
     {
-      "epoch": 3.1,
-      "learning_rate": 1.6608312647135917e-05,
-      "loss": 0.2849,
+      "epoch": 3.180163621292392,
+      "grad_norm": 0.24928292632102966,
+      "learning_rate": 1.546554537011091e-05,
+      "loss": 0.4122,
       "step": 88240
     },
     {
-      "epoch": 3.1,
-      "learning_rate": 1.6605629261587957e-05,
-      "loss": 0.2788,
+      "epoch": 3.180343820953617,
+      "grad_norm": 0.2341477870941162,
+      "learning_rate": 1.5462847856574437e-05,
+      "loss": 0.396,
       "step": 88245
     },
     {
-      "epoch": 3.1,
-      "learning_rate": 1.660294598503538e-05,
-      "loss": 0.2552,
+      "epoch": 3.180524020614841,
+      "grad_norm": 0.21584312617778778,
+      "learning_rate": 1.546015047297981e-05,
+      "loss": 0.3566,
       "step": 88250
     },
     {
-      "epoch": 3.11,
-      "learning_rate": 1.660026281751301e-05,
-      "loss": 0.2568,
+      "epoch": 3.180704220276066,
+      "grad_norm": 0.25343307852745056,
+      "learning_rate": 1.5457453219363794e-05,
+      "loss": 0.3825,
       "step": 88255
     },
     {
-      "epoch": 3.11,
-      "learning_rate": 1.6597579759055688e-05,
-      "loss": 0.2444,
+      "epoch": 3.1808844199372905,
+      "grad_norm": 0.26470595598220825,
+      "learning_rate": 1.545475609576313e-05,
+      "loss": 0.3954,
       "step": 88260
     },
     {
-      "epoch": 3.11,
-      "learning_rate": 1.659489680969826e-05,
-      "loss": 0.2542,
+      "epoch": 3.1810646195985153,
+      "grad_norm": 0.19923968613147736,
+      "learning_rate": 1.5452059102214558e-05,
+      "loss": 0.3574,
       "step": 88265
     },
     {
-      "epoch": 3.11,
-      "learning_rate": 1.6592213969475557e-05,
-      "loss": 0.2692,
+      "epoch": 3.1812448192597396,
+      "grad_norm": 0.20713098347187042,
+      "learning_rate": 1.5449362238754838e-05,
+      "loss": 0.3896,
       "step": 88270
     },
     {
-      "epoch": 3.11,
-      "learning_rate": 1.6589531238422416e-05,
-      "loss": 0.2584,
+      "epoch": 3.1814250189209643,
+      "grad_norm": 0.24674662947654724,
+      "learning_rate": 1.5446665505420714e-05,
+      "loss": 0.3901,
       "step": 88275
     },
     {
-      "epoch": 3.11,
-      "learning_rate": 1.658684861657366e-05,
-      "loss": 0.2641,
+      "epoch": 3.181605218582189,
+      "grad_norm": 0.22390885651111603,
+      "learning_rate": 1.5443968902248913e-05,
+      "loss": 0.375,
       "step": 88280
     },
     {
-      "epoch": 3.11,
-      "learning_rate": 1.6584166103964143e-05,
-      "loss": 0.2619,
+      "epoch": 3.1817854182434138,
+      "grad_norm": 0.252699077129364,
+      "learning_rate": 1.5441272429276193e-05,
+      "loss": 0.3771,
       "step": 88285
     },
     {
-      "epoch": 3.11,
-      "learning_rate": 1.6581483700628676e-05,
-      "loss": 0.2526,
+      "epoch": 3.1819656179046385,
+      "grad_norm": 0.25395792722702026,
+      "learning_rate": 1.5438576086539275e-05,
+      "loss": 0.389,
       "step": 88290
     },
     {
-      "epoch": 3.11,
-      "learning_rate": 1.65788014066021e-05,
-      "loss": 0.2417,
+      "epoch": 3.182145817565863,
+      "grad_norm": 0.27715232968330383,
+      "learning_rate": 1.543587987407492e-05,
+      "loss": 0.3793,
       "step": 88295
     },
     {
-      "epoch": 3.11,
-      "learning_rate": 1.6576119221919228e-05,
-      "loss": 0.2669,
+      "epoch": 3.1823260172270875,
+      "grad_norm": 0.27703672647476196,
+      "learning_rate": 1.5433183791919844e-05,
+      "loss": 0.3654,
       "step": 88300
     },
     {
-      "epoch": 3.11,
-      "learning_rate": 1.6573437146614906e-05,
-      "loss": 0.2453,
+      "epoch": 3.1825062168883123,
+      "grad_norm": 0.21371644735336304,
+      "learning_rate": 1.543048784011078e-05,
+      "loss": 0.3736,
       "step": 88305
     },
     {
-      "epoch": 3.11,
-      "learning_rate": 1.657075518072395e-05,
-      "loss": 0.2621,
+      "epoch": 3.182686416549537,
+      "grad_norm": 0.18627633154392242,
+      "learning_rate": 1.542779201868448e-05,
+      "loss": 0.332,
       "step": 88310
     },
     {
-      "epoch": 3.11,
-      "learning_rate": 1.6568073324281178e-05,
-      "loss": 0.2774,
+      "epoch": 3.1828666162107617,
+      "grad_norm": 0.24603888392448425,
+      "learning_rate": 1.542509632767765e-05,
+      "loss": 0.3981,
       "step": 88315
     },
     {
-      "epoch": 3.11,
-      "learning_rate": 1.6565391577321415e-05,
-      "loss": 0.2645,
+      "epoch": 3.183046815871986,
+      "grad_norm": 0.26360347867012024,
+      "learning_rate": 1.5422400767127034e-05,
+      "loss": 0.3647,
       "step": 88320
     },
     {
-      "epoch": 3.11,
-      "learning_rate": 1.6562709939879486e-05,
-      "loss": 0.245,
+      "epoch": 3.1832270155332107,
+      "grad_norm": 0.24063238501548767,
+      "learning_rate": 1.5419705337069353e-05,
+      "loss": 0.3559,
       "step": 88325
     },
     {
-      "epoch": 3.11,
-      "learning_rate": 1.6560028411990213e-05,
-      "loss": 0.2613,
+      "epoch": 3.1834072151944355,
+      "grad_norm": 0.20714794099330902,
+      "learning_rate": 1.5417010037541322e-05,
+      "loss": 0.3922,
       "step": 88330
     },
     {
-      "epoch": 3.11,
-      "learning_rate": 1.6557346993688412e-05,
-      "loss": 0.2542,
+      "epoch": 3.18358741485566,
+      "grad_norm": 0.24574214220046997,
+      "learning_rate": 1.5414314868579687e-05,
+      "loss": 0.384,
       "step": 88335
     },
     {
-      "epoch": 3.11,
-      "learning_rate": 1.655466568500888e-05,
-      "loss": 0.2499,
+      "epoch": 3.1837676145168845,
+      "grad_norm": 0.2243025153875351,
+      "learning_rate": 1.541161983022115e-05,
+      "loss": 0.3988,
       "step": 88340
     },
     {
-      "epoch": 3.11,
-      "learning_rate": 1.655198448598646e-05,
-      "loss": 0.2386,
+      "epoch": 3.1839478141781092,
+      "grad_norm": 0.22772471606731415,
+      "learning_rate": 1.5408924922502437e-05,
+      "loss": 0.4002,
       "step": 88345
     },
     {
-      "epoch": 3.11,
-      "learning_rate": 1.6549303396655953e-05,
-      "loss": 0.2671,
+      "epoch": 3.184128013839334,
+      "grad_norm": 0.27433350682258606,
+      "learning_rate": 1.5406230145460264e-05,
+      "loss": 0.3967,
       "step": 88350
     },
     {
-      "epoch": 3.11,
-      "learning_rate": 1.6546622417052174e-05,
-      "loss": 0.2524,
+      "epoch": 3.1843082135005587,
+      "grad_norm": 0.24873776733875275,
+      "learning_rate": 1.540353549913135e-05,
+      "loss": 0.3847,
       "step": 88355
     },
     {
-      "epoch": 3.11,
-      "learning_rate": 1.6543941547209923e-05,
-      "loss": 0.2415,
+      "epoch": 3.1844884131617834,
+      "grad_norm": 0.21438409388065338,
+      "learning_rate": 1.5400840983552405e-05,
+      "loss": 0.3762,
       "step": 88360
     },
     {
-      "epoch": 3.11,
-      "learning_rate": 1.6541260787164026e-05,
-      "loss": 0.2601,
+      "epoch": 3.1846686128230077,
+      "grad_norm": 0.20840908586978912,
+      "learning_rate": 1.539814659876015e-05,
+      "loss": 0.4169,
       "step": 88365
     },
     {
-      "epoch": 3.11,
-      "learning_rate": 1.653858013694928e-05,
-      "loss": 0.2599,
+      "epoch": 3.1848488124842325,
+      "grad_norm": 0.30924826860427856,
+      "learning_rate": 1.5395452344791277e-05,
+      "loss": 0.3751,
       "step": 88370
     },
     {
-      "epoch": 3.11,
-      "learning_rate": 1.65358995966005e-05,
-      "loss": 0.2621,
+      "epoch": 3.185029012145457,
+      "grad_norm": 0.25707584619522095,
+      "learning_rate": 1.5392758221682516e-05,
+      "loss": 0.4209,
       "step": 88375
     },
     {
-      "epoch": 3.11,
-      "learning_rate": 1.6533219166152473e-05,
-      "loss": 0.2534,
+      "epoch": 3.185209211806682,
+      "grad_norm": 0.26053330302238464,
+      "learning_rate": 1.5390064229470568e-05,
+      "loss": 0.3743,
       "step": 88380
     },
     {
-      "epoch": 3.11,
-      "learning_rate": 1.653053884564003e-05,
-      "loss": 0.25,
+      "epoch": 3.1853894114679067,
+      "grad_norm": 0.24853472411632538,
+      "learning_rate": 1.5387370368192116e-05,
+      "loss": 0.3825,
       "step": 88385
     },
     {
-      "epoch": 3.11,
-      "learning_rate": 1.6527858635097955e-05,
-      "loss": 0.2317,
+      "epoch": 3.185569611129131,
+      "grad_norm": 0.20806756615638733,
+      "learning_rate": 1.5384676637883904e-05,
+      "loss": 0.3562,
       "step": 88390
     },
     {
-      "epoch": 3.11,
-      "learning_rate": 1.652517853456104e-05,
-      "loss": 0.2635,
+      "epoch": 3.1857498107903557,
+      "grad_norm": 0.2653723359107971,
+      "learning_rate": 1.5381983038582588e-05,
+      "loss": 0.374,
       "step": 88395
     },
     {
-      "epoch": 3.11,
-      "learning_rate": 1.6522498544064113e-05,
-      "loss": 0.2467,
+      "epoch": 3.1859300104515804,
+      "grad_norm": 0.25022828578948975,
+      "learning_rate": 1.5379289570324912e-05,
+      "loss": 0.3863,
       "step": 88400
     },
     {
-      "epoch": 3.11,
-      "learning_rate": 1.651981866364195e-05,
-      "loss": 0.2529,
+      "epoch": 3.186110210112805,
+      "grad_norm": 0.2051696926355362,
+      "learning_rate": 1.5376596233147543e-05,
+      "loss": 0.4092,
       "step": 88405
     },
     {
-      "epoch": 3.11,
-      "learning_rate": 1.6517138893329355e-05,
-      "loss": 0.2823,
+      "epoch": 3.1862904097740294,
+      "grad_norm": 0.1869521141052246,
+      "learning_rate": 1.5373903027087185e-05,
+      "loss": 0.3755,
       "step": 88410
     },
     {
-      "epoch": 3.11,
-      "learning_rate": 1.651445923316111e-05,
-      "loss": 0.2919,
+      "epoch": 3.186470609435254,
+      "grad_norm": 0.17560672760009766,
+      "learning_rate": 1.537120995218054e-05,
+      "loss": 0.3518,
       "step": 88415
     },
     {
-      "epoch": 3.11,
-      "learning_rate": 1.651177968317203e-05,
-      "loss": 0.2554,
+      "epoch": 3.186650809096479,
+      "grad_norm": 0.24432684481143951,
+      "learning_rate": 1.5368517008464295e-05,
+      "loss": 0.3625,
       "step": 88420
     },
     {
-      "epoch": 3.11,
-      "learning_rate": 1.65091002433969e-05,
-      "loss": 0.233,
+      "epoch": 3.1868310087577036,
+      "grad_norm": 0.2695041298866272,
+      "learning_rate": 1.5365824195975137e-05,
+      "loss": 0.3801,
       "step": 88425
     },
     {
-      "epoch": 3.11,
-      "learning_rate": 1.65064209138705e-05,
-      "loss": 0.2487,
+      "epoch": 3.1870112084189284,
+      "grad_norm": 0.2649804651737213,
+      "learning_rate": 1.5363131514749762e-05,
+      "loss": 0.3637,
       "step": 88430
     },
     {
-      "epoch": 3.11,
-      "learning_rate": 1.650374169462762e-05,
-      "loss": 0.2457,
+      "epoch": 3.1871914080801527,
+      "grad_norm": 0.2614179849624634,
+      "learning_rate": 1.5360438964824846e-05,
+      "loss": 0.3891,
       "step": 88435
     },
     {
-      "epoch": 3.11,
-      "learning_rate": 1.650106258570307e-05,
-      "loss": 0.254,
+      "epoch": 3.1873716077413774,
+      "grad_norm": 0.2579090893268585,
+      "learning_rate": 1.5357746546237093e-05,
+      "loss": 0.3654,
       "step": 88440
     },
     {
-      "epoch": 3.11,
-      "learning_rate": 1.649838358713162e-05,
-      "loss": 0.2363,
+      "epoch": 3.187551807402602,
+      "grad_norm": 0.2515721619129181,
+      "learning_rate": 1.5355054259023176e-05,
+      "loss": 0.414,
       "step": 88445
     },
     {
-      "epoch": 3.11,
-      "learning_rate": 1.6495704698948057e-05,
-      "loss": 0.2572,
+      "epoch": 3.187732007063827,
+      "grad_norm": 0.19676761329174042,
+      "learning_rate": 1.5352362103219772e-05,
+      "loss": 0.4121,
       "step": 88450
     },
     {
-      "epoch": 3.11,
-      "learning_rate": 1.649302592118715e-05,
-      "loss": 0.259,
+      "epoch": 3.187912206725051,
+      "grad_norm": 0.3330390751361847,
+      "learning_rate": 1.534967007886357e-05,
+      "loss": 0.4198,
       "step": 88455
     },
     {
-      "epoch": 3.11,
-      "learning_rate": 1.64903472538837e-05,
-      "loss": 0.2359,
+      "epoch": 3.188092406386276,
+      "grad_norm": 0.1731852889060974,
+      "learning_rate": 1.5346978185991254e-05,
+      "loss": 0.3721,
       "step": 88460
     },
     {
-      "epoch": 3.11,
-      "learning_rate": 1.6487668697072493e-05,
-      "loss": 0.2649,
+      "epoch": 3.1882726060475006,
+      "grad_norm": 0.21727561950683594,
+      "learning_rate": 1.5344286424639473e-05,
+      "loss": 0.3973,
       "step": 88465
     },
     {
-      "epoch": 3.11,
-      "learning_rate": 1.6484990250788293e-05,
-      "loss": 0.2425,
+      "epoch": 3.1884528057087254,
+      "grad_norm": 0.21226783096790314,
+      "learning_rate": 1.534159479484493e-05,
+      "loss": 0.3798,
       "step": 88470
     },
     {
-      "epoch": 3.11,
-      "learning_rate": 1.6482311915065874e-05,
-      "loss": 0.2795,
+      "epoch": 3.18863300536995,
+      "grad_norm": 0.21172350645065308,
+      "learning_rate": 1.5338903296644282e-05,
+      "loss": 0.3675,
       "step": 88475
     },
     {
-      "epoch": 3.11,
-      "learning_rate": 1.6479633689940032e-05,
-      "loss": 0.2461,
+      "epoch": 3.1888132050311744,
+      "grad_norm": 0.23403914272785187,
+      "learning_rate": 1.533621193007421e-05,
+      "loss": 0.3956,
       "step": 88480
     },
     {
-      "epoch": 3.11,
-      "learning_rate": 1.647695557544553e-05,
-      "loss": 0.2561,
+      "epoch": 3.188993404692399,
+      "grad_norm": 0.24958845973014832,
+      "learning_rate": 1.533352069517139e-05,
+      "loss": 0.3753,
       "step": 88485
     },
     {
-      "epoch": 3.11,
-      "learning_rate": 1.6474277571617145e-05,
-      "loss": 0.2542,
+      "epoch": 3.189173604353624,
+      "grad_norm": 0.29716435074806213,
+      "learning_rate": 1.5330829591972463e-05,
+      "loss": 0.3674,
       "step": 88490
     },
     {
-      "epoch": 3.11,
-      "learning_rate": 1.6471599678489634e-05,
-      "loss": 0.2431,
+      "epoch": 3.1893538040148486,
+      "grad_norm": 0.23734360933303833,
+      "learning_rate": 1.5328138620514125e-05,
+      "loss": 0.3635,
       "step": 88495
     },
     {
-      "epoch": 3.11,
-      "learning_rate": 1.6468921896097793e-05,
-      "loss": 0.2672,
+      "epoch": 3.189534003676073,
+      "grad_norm": 0.22716392576694489,
+      "learning_rate": 1.532544778083302e-05,
+      "loss": 0.3993,
       "step": 88500
     },
     {
-      "epoch": 3.11,
-      "eval_loss": 0.25621533393859863,
-      "eval_runtime": 10.5456,
-      "eval_samples_per_second": 9.483,
-      "eval_steps_per_second": 9.483,
+      "epoch": 3.189534003676073,
+      "eval_loss": 0.4313284754753113,
+      "eval_runtime": 3.5346,
+      "eval_samples_per_second": 28.292,
+      "eval_steps_per_second": 7.073,
       "step": 88500
     },
     {
-      "epoch": 3.11,
-      "learning_rate": 1.646624422447638e-05,
-      "loss": 0.2621,
+      "epoch": 3.1897142033372976,
+      "grad_norm": 0.24912869930267334,
+      "learning_rate": 1.532275707296581e-05,
+      "loss": 0.3939,
       "step": 88505
     },
     {
-      "epoch": 3.11,
-      "learning_rate": 1.6463566663660153e-05,
-      "loss": 0.2563,
+      "epoch": 3.1898944029985223,
+      "grad_norm": 0.20654475688934326,
+      "learning_rate": 1.5320066496949164e-05,
+      "loss": 0.3825,
       "step": 88510
     },
     {
-      "epoch": 3.11,
-      "learning_rate": 1.6460889213683892e-05,
-      "loss": 0.2306,
+      "epoch": 3.190074602659747,
+      "grad_norm": 0.23317840695381165,
+      "learning_rate": 1.531737605281974e-05,
+      "loss": 0.3779,
       "step": 88515
     },
     {
-      "epoch": 3.11,
-      "learning_rate": 1.645821187458236e-05,
-      "loss": 0.2811,
+      "epoch": 3.190254802320972,
+      "grad_norm": 0.2050599455833435,
+      "learning_rate": 1.531468574061419e-05,
+      "loss": 0.3616,
       "step": 88520
     },
     {
-      "epoch": 3.11,
-      "learning_rate": 1.645553464639032e-05,
-      "loss": 0.2564,
+      "epoch": 3.190435001982196,
+      "grad_norm": 0.24573275446891785,
+      "learning_rate": 1.5311995560369176e-05,
+      "loss": 0.3505,
       "step": 88525
     },
     {
-      "epoch": 3.11,
-      "learning_rate": 1.6452857529142518e-05,
-      "loss": 0.2592,
+      "epoch": 3.190615201643421,
+      "grad_norm": 0.22783836722373962,
+      "learning_rate": 1.530930551212134e-05,
+      "loss": 0.3509,
       "step": 88530
     },
     {
-      "epoch": 3.11,
-      "learning_rate": 1.645018052287374e-05,
-      "loss": 0.2565,
+      "epoch": 3.1907954013046456,
+      "grad_norm": 0.22740013897418976,
+      "learning_rate": 1.5306615595907348e-05,
+      "loss": 0.35,
       "step": 88535
     },
     {
-      "epoch": 3.12,
-      "learning_rate": 1.6447503627618728e-05,
-      "loss": 0.2429,
+      "epoch": 3.1909756009658703,
+      "grad_norm": 0.23683089017868042,
+      "learning_rate": 1.5303925811763833e-05,
+      "loss": 0.3592,
       "step": 88540
     },
     {
-      "epoch": 3.12,
-      "learning_rate": 1.6444826843412254e-05,
-      "loss": 0.2409,
+      "epoch": 3.1911558006270946,
+      "grad_norm": 0.2351250797510147,
+      "learning_rate": 1.530123615972746e-05,
+      "loss": 0.4032,
       "step": 88545
     },
     {
-      "epoch": 3.12,
-      "learning_rate": 1.644215017028905e-05,
-      "loss": 0.2629,
+      "epoch": 3.1913360002883193,
+      "grad_norm": 0.23128874599933624,
+      "learning_rate": 1.5298546639834867e-05,
+      "loss": 0.3866,
       "step": 88550
     },
     {
-      "epoch": 3.12,
-      "learning_rate": 1.6439473608283894e-05,
-      "loss": 0.2676,
+      "epoch": 3.191516199949544,
+      "grad_norm": 0.24116584658622742,
+      "learning_rate": 1.529585725212269e-05,
+      "loss": 0.3896,
       "step": 88555
     },
     {
-      "epoch": 3.12,
-      "learning_rate": 1.643679715743154e-05,
-      "loss": 0.2427,
+      "epoch": 3.191696399610769,
+      "grad_norm": 0.19982759654521942,
+      "learning_rate": 1.5293167996627583e-05,
+      "loss": 0.3896,
       "step": 88560
     },
     {
-      "epoch": 3.12,
-      "learning_rate": 1.6434120817766723e-05,
-      "loss": 0.2589,
+      "epoch": 3.1918765992719935,
+      "grad_norm": 0.19958491623401642,
+      "learning_rate": 1.529047887338619e-05,
+      "loss": 0.3574,
       "step": 88565
     },
     {
-      "epoch": 3.12,
-      "learning_rate": 1.6431444589324193e-05,
-      "loss": 0.2767,
+      "epoch": 3.192056798933218,
+      "grad_norm": 0.22124148905277252,
+      "learning_rate": 1.5287789882435126e-05,
+      "loss": 0.3592,
       "step": 88570
     },
     {
-      "epoch": 3.12,
-      "learning_rate": 1.6428768472138723e-05,
-      "loss": 0.2622,
+      "epoch": 3.1922369985944425,
+      "grad_norm": 0.19804465770721436,
+      "learning_rate": 1.528510102381106e-05,
+      "loss": 0.3993,
       "step": 88575
     },
     {
-      "epoch": 3.12,
-      "learning_rate": 1.6426092466245044e-05,
-      "loss": 0.2695,
+      "epoch": 3.1924171982556673,
+      "grad_norm": 0.18390753865242004,
+      "learning_rate": 1.5282412297550603e-05,
+      "loss": 0.4034,
       "step": 88580
     },
     {
-      "epoch": 3.12,
-      "learning_rate": 1.6423416571677904e-05,
-      "loss": 0.2686,
+      "epoch": 3.192597397916892,
+      "grad_norm": 0.2427562028169632,
+      "learning_rate": 1.5279723703690404e-05,
+      "loss": 0.3776,
       "step": 88585
     },
     {
-      "epoch": 3.12,
-      "learning_rate": 1.642074078847204e-05,
-      "loss": 0.2474,
+      "epoch": 3.1927775975781167,
+      "grad_norm": 0.2625272572040558,
+      "learning_rate": 1.5277035242267085e-05,
+      "loss": 0.4091,
       "step": 88590
     },
     {
-      "epoch": 3.12,
-      "learning_rate": 1.6418065116662202e-05,
-      "loss": 0.2619,
+      "epoch": 3.192957797239341,
+      "grad_norm": 0.20139677822589874,
+      "learning_rate": 1.527434691331727e-05,
+      "loss": 0.4056,
       "step": 88595
     },
     {
-      "epoch": 3.12,
-      "learning_rate": 1.6415389556283144e-05,
-      "loss": 0.2584,
+      "epoch": 3.1931379969005658,
+      "grad_norm": 0.22303380072116852,
+      "learning_rate": 1.5271658716877607e-05,
+      "loss": 0.3861,
       "step": 88600
     },
     {
-      "epoch": 3.12,
-      "learning_rate": 1.6412714107369593e-05,
-      "loss": 0.2847,
+      "epoch": 3.1933181965617905,
+      "grad_norm": 0.24220505356788635,
+      "learning_rate": 1.526897065298471e-05,
+      "loss": 0.3655,
       "step": 88605
     },
     {
-      "epoch": 3.12,
-      "learning_rate": 1.6410038769956277e-05,
-      "loss": 0.2579,
+      "epoch": 3.1934983962230152,
+      "grad_norm": 0.2237461358308792,
+      "learning_rate": 1.5266282721675196e-05,
+      "loss": 0.3396,
       "step": 88610
     },
     {
-      "epoch": 3.12,
-      "learning_rate": 1.640736354407796e-05,
-      "loss": 0.2679,
+      "epoch": 3.19367859588424,
+      "grad_norm": 0.22719275951385498,
+      "learning_rate": 1.5263594922985698e-05,
+      "loss": 0.389,
       "step": 88615
     },
     {
-      "epoch": 3.12,
-      "learning_rate": 1.6404688429769362e-05,
-      "loss": 0.2467,
+      "epoch": 3.1938587955454643,
+      "grad_norm": 0.20860248804092407,
+      "learning_rate": 1.5260907256952832e-05,
+      "loss": 0.376,
       "step": 88620
     },
     {
-      "epoch": 3.12,
-      "learning_rate": 1.6402013427065215e-05,
-      "loss": 0.2191,
+      "epoch": 3.194038995206689,
+      "grad_norm": 0.25672057271003723,
+      "learning_rate": 1.5258219723613226e-05,
+      "loss": 0.3926,
       "step": 88625
     },
     {
-      "epoch": 3.12,
-      "learning_rate": 1.6399338536000258e-05,
-      "loss": 0.2667,
+      "epoch": 3.1942191948679137,
+      "grad_norm": 0.24725201725959778,
+      "learning_rate": 1.5255532323003488e-05,
+      "loss": 0.3647,
       "step": 88630
     },
     {
-      "epoch": 3.12,
-      "learning_rate": 1.639666375660923e-05,
-      "loss": 0.2494,
+      "epoch": 3.1943993945291385,
+      "grad_norm": 0.21268433332443237,
+      "learning_rate": 1.5252845055160231e-05,
+      "loss": 0.4103,
       "step": 88635
     },
     {
-      "epoch": 3.12,
-      "learning_rate": 1.639398908892685e-05,
-      "loss": 0.2712,
+      "epoch": 3.1945795941903627,
+      "grad_norm": 0.2236996740102768,
+      "learning_rate": 1.5250157920120079e-05,
+      "loss": 0.3829,
       "step": 88640
     },
     {
-      "epoch": 3.12,
-      "learning_rate": 1.6391314532987844e-05,
-      "loss": 0.2481,
+      "epoch": 3.1947597938515875,
+      "grad_norm": 0.2377428263425827,
+      "learning_rate": 1.5247470917919634e-05,
+      "loss": 0.3897,
       "step": 88645
     },
     {
-      "epoch": 3.12,
-      "learning_rate": 1.6388640088826946e-05,
-      "loss": 0.2413,
+      "epoch": 3.194939993512812,
+      "grad_norm": 0.1965668499469757,
+      "learning_rate": 1.5244784048595506e-05,
+      "loss": 0.4063,
       "step": 88650
     },
     {
-      "epoch": 3.12,
-      "learning_rate": 1.6385965756478887e-05,
-      "loss": 0.2479,
+      "epoch": 3.195120193174037,
+      "grad_norm": 0.21718864142894745,
+      "learning_rate": 1.5242097312184314e-05,
+      "loss": 0.3925,
       "step": 88655
     },
     {
-      "epoch": 3.12,
-      "learning_rate": 1.638329153597839e-05,
-      "loss": 0.2568,
+      "epoch": 3.1953003928352617,
+      "grad_norm": 0.25005796551704407,
+      "learning_rate": 1.5239410708722651e-05,
+      "loss": 0.4104,
       "step": 88660
     },
     {
-      "epoch": 3.12,
-      "learning_rate": 1.6380617427360164e-05,
-      "loss": 0.2788,
+      "epoch": 3.195480592496486,
+      "grad_norm": 0.18256092071533203,
+      "learning_rate": 1.5236724238247132e-05,
+      "loss": 0.3653,
       "step": 88665
     },
     {
-      "epoch": 3.12,
-      "learning_rate": 1.6377943430658942e-05,
-      "loss": 0.2436,
+      "epoch": 3.1956607921577107,
+      "grad_norm": 0.25699880719184875,
+      "learning_rate": 1.5234037900794362e-05,
+      "loss": 0.3736,
       "step": 88670
     },
     {
-      "epoch": 3.12,
-      "learning_rate": 1.6375269545909445e-05,
-      "loss": 0.2451,
+      "epoch": 3.1958409918189354,
+      "grad_norm": 0.20626196265220642,
+      "learning_rate": 1.5231351696400915e-05,
+      "loss": 0.3671,
       "step": 88675
     },
     {
-      "epoch": 3.12,
-      "learning_rate": 1.637259577314639e-05,
-      "loss": 0.2385,
+      "epoch": 3.19602119148016,
+      "grad_norm": 0.2111903876066208,
+      "learning_rate": 1.5228665625103428e-05,
+      "loss": 0.3878,
       "step": 88680
     },
     {
-      "epoch": 3.12,
-      "learning_rate": 1.6369922112404488e-05,
-      "loss": 0.2284,
+      "epoch": 3.1962013911413845,
+      "grad_norm": 0.22062833607196808,
+      "learning_rate": 1.5225979686938474e-05,
+      "loss": 0.3727,
       "step": 88685
     },
     {
-      "epoch": 3.12,
-      "learning_rate": 1.6367248563718468e-05,
-      "loss": 0.2568,
+      "epoch": 3.196381590802609,
+      "grad_norm": 0.2414456009864807,
+      "learning_rate": 1.5223293881942653e-05,
+      "loss": 0.3594,
       "step": 88690
     },
     {
-      "epoch": 3.12,
-      "learning_rate": 1.6364575127123034e-05,
-      "loss": 0.2702,
+      "epoch": 3.196561790463834,
+      "grad_norm": 0.21987886726856232,
+      "learning_rate": 1.5220608210152559e-05,
+      "loss": 0.3376,
       "step": 88695
     },
     {
-      "epoch": 3.12,
-      "learning_rate": 1.6361901802652896e-05,
-      "loss": 0.2529,
+      "epoch": 3.1967419901250587,
+      "grad_norm": 0.20280367136001587,
+      "learning_rate": 1.5217922671604784e-05,
+      "loss": 0.3523,
       "step": 88700
     },
     {
-      "epoch": 3.12,
-      "learning_rate": 1.6359228590342774e-05,
-      "loss": 0.2551,
+      "epoch": 3.1969221897862834,
+      "grad_norm": 0.259671151638031,
+      "learning_rate": 1.5215237266335925e-05,
+      "loss": 0.3875,
       "step": 88705
     },
     {
-      "epoch": 3.12,
-      "learning_rate": 1.635655549022737e-05,
-      "loss": 0.2418,
+      "epoch": 3.1971023894475077,
+      "grad_norm": 0.26262468099594116,
+      "learning_rate": 1.5212551994382562e-05,
+      "loss": 0.3624,
       "step": 88710
     },
     {
-      "epoch": 3.12,
-      "learning_rate": 1.6353882502341404e-05,
-      "loss": 0.2548,
+      "epoch": 3.1972825891087324,
+      "grad_norm": 0.2648344933986664,
+      "learning_rate": 1.5209866855781277e-05,
+      "loss": 0.3959,
       "step": 88715
     },
     {
-      "epoch": 3.12,
-      "learning_rate": 1.6351209626719578e-05,
-      "loss": 0.2475,
+      "epoch": 3.197462788769957,
+      "grad_norm": 0.18271714448928833,
+      "learning_rate": 1.5207181850568666e-05,
+      "loss": 0.3252,
       "step": 88720
     },
     {
-      "epoch": 3.12,
-      "learning_rate": 1.634853686339658e-05,
-      "loss": 0.2506,
+      "epoch": 3.197642988431182,
+      "grad_norm": 0.23713482916355133,
+      "learning_rate": 1.5204496978781301e-05,
+      "loss": 0.3915,
       "step": 88725
     },
     {
-      "epoch": 3.12,
-      "learning_rate": 1.6345864212407136e-05,
-      "loss": 0.2439,
+      "epoch": 3.197823188092406,
+      "grad_norm": 0.2027212232351303,
+      "learning_rate": 1.5201812240455776e-05,
+      "loss": 0.368,
       "step": 88730
     },
     {
-      "epoch": 3.12,
-      "learning_rate": 1.6343191673785953e-05,
-      "loss": 0.2548,
+      "epoch": 3.198003387753631,
+      "grad_norm": 0.26085880398750305,
+      "learning_rate": 1.519912763562866e-05,
+      "loss": 0.3783,
       "step": 88735
     },
     {
-      "epoch": 3.12,
-      "learning_rate": 1.6340519247567702e-05,
-      "loss": 0.2592,
+      "epoch": 3.1981835874148556,
+      "grad_norm": 0.23765386641025543,
+      "learning_rate": 1.5196443164336526e-05,
+      "loss": 0.3829,
       "step": 88740
     },
     {
-      "epoch": 3.12,
-      "learning_rate": 1.6337846933787116e-05,
-      "loss": 0.2338,
+      "epoch": 3.1983637870760804,
+      "grad_norm": 0.2554900348186493,
+      "learning_rate": 1.5193758826615964e-05,
+      "loss": 0.3891,
       "step": 88745
     },
     {
-      "epoch": 3.12,
-      "learning_rate": 1.633517473247888e-05,
-      "loss": 0.2454,
+      "epoch": 3.198543986737305,
+      "grad_norm": 0.21230515837669373,
+      "learning_rate": 1.5191074622503542e-05,
+      "loss": 0.4042,
       "step": 88750
     },
     {
-      "epoch": 3.12,
-      "learning_rate": 1.6332502643677683e-05,
-      "loss": 0.2575,
+      "epoch": 3.1987241863985294,
+      "grad_norm": 0.31726622581481934,
+      "learning_rate": 1.5188390552035813e-05,
+      "loss": 0.3832,
       "step": 88755
     },
     {
-      "epoch": 3.12,
-      "learning_rate": 1.6329830667418228e-05,
-      "loss": 0.2732,
+      "epoch": 3.198904386059754,
+      "grad_norm": 0.21085330843925476,
+      "learning_rate": 1.5185706615249378e-05,
+      "loss": 0.375,
       "step": 88760
     },
     {
-      "epoch": 3.12,
-      "learning_rate": 1.632715880373521e-05,
-      "loss": 0.2445,
+      "epoch": 3.199084585720979,
+      "grad_norm": 0.27633991837501526,
+      "learning_rate": 1.5183022812180783e-05,
+      "loss": 0.4169,
       "step": 88765
     },
     {
-      "epoch": 3.12,
-      "learning_rate": 1.632448705266332e-05,
-      "loss": 0.2622,
+      "epoch": 3.1992647853822036,
+      "grad_norm": 0.21804660558700562,
+      "learning_rate": 1.5180339142866607e-05,
+      "loss": 0.3719,
       "step": 88770
     },
     {
-      "epoch": 3.12,
-      "learning_rate": 1.632181541423725e-05,
-      "loss": 0.2481,
+      "epoch": 3.199444985043428,
+      "grad_norm": 0.24333439767360687,
+      "learning_rate": 1.517765560734341e-05,
+      "loss": 0.381,
       "step": 88775
     },
     {
-      "epoch": 3.12,
-      "learning_rate": 1.6319143888491675e-05,
-      "loss": 0.2467,
+      "epoch": 3.1996251847046526,
+      "grad_norm": 0.23350374400615692,
+      "learning_rate": 1.5174972205647744e-05,
+      "loss": 0.3529,
       "step": 88780
     },
     {
-      "epoch": 3.12,
-      "learning_rate": 1.631647247546131e-05,
-      "loss": 0.2521,
+      "epoch": 3.1998053843658774,
+      "grad_norm": 0.23045288026332855,
+      "learning_rate": 1.5172288937816193e-05,
+      "loss": 0.3478,
       "step": 88785
     },
     {
-      "epoch": 3.12,
-      "learning_rate": 1.6313801175180822e-05,
-      "loss": 0.2684,
+      "epoch": 3.199985584027102,
+      "grad_norm": 0.19861705601215363,
+      "learning_rate": 1.5169605803885296e-05,
+      "loss": 0.3737,
       "step": 88790
     },
     {
-      "epoch": 3.12,
-      "learning_rate": 1.6311129987684902e-05,
-      "loss": 0.2602,
+      "epoch": 3.200165783688327,
+      "grad_norm": 0.3103479743003845,
+      "learning_rate": 1.5166922803891615e-05,
+      "loss": 0.3906,
       "step": 88795
     },
     {
-      "epoch": 3.12,
-      "learning_rate": 1.6308458913008228e-05,
-      "loss": 0.2543,
+      "epoch": 3.200345983349551,
+      "grad_norm": 0.2640889585018158,
+      "learning_rate": 1.5164239937871708e-05,
+      "loss": 0.3636,
       "step": 88800
     },
     {
-      "epoch": 3.12,
-      "learning_rate": 1.6305787951185494e-05,
-      "loss": 0.2528,
+      "epoch": 3.200526183010776,
+      "grad_norm": 0.22964945435523987,
+      "learning_rate": 1.5161557205862127e-05,
+      "loss": 0.3578,
       "step": 88805
     },
     {
-      "epoch": 3.12,
-      "learning_rate": 1.630311710225137e-05,
-      "loss": 0.2517,
+      "epoch": 3.2007063826720006,
+      "grad_norm": 0.20563103258609772,
+      "learning_rate": 1.5158874607899423e-05,
+      "loss": 0.3629,
       "step": 88810
     },
     {
-      "epoch": 3.12,
-      "learning_rate": 1.6300446366240543e-05,
-      "loss": 0.2709,
+      "epoch": 3.2008865823332253,
+      "grad_norm": 0.22984810173511505,
+      "learning_rate": 1.515619214402015e-05,
+      "loss": 0.4054,
       "step": 88815
     },
     {
-      "epoch": 3.12,
-      "learning_rate": 1.6297775743187677e-05,
-      "loss": 0.2588,
+      "epoch": 3.20106678199445,
+      "grad_norm": 0.23353759944438934,
+      "learning_rate": 1.5153509814260849e-05,
+      "loss": 0.3607,
       "step": 88820
     },
     {
-      "epoch": 3.13,
-      "learning_rate": 1.6295105233127468e-05,
-      "loss": 0.2637,
+      "epoch": 3.2012469816556743,
+      "grad_norm": 0.2417394518852234,
+      "learning_rate": 1.5150827618658075e-05,
+      "loss": 0.3985,
       "step": 88825
     },
     {
-      "epoch": 3.13,
-      "learning_rate": 1.6292434836094584e-05,
-      "loss": 0.2487,
+      "epoch": 3.201427181316899,
+      "grad_norm": 0.2204267978668213,
+      "learning_rate": 1.514814555724837e-05,
+      "loss": 0.3738,
       "step": 88830
     },
     {
-      "epoch": 3.13,
-      "learning_rate": 1.6289764552123688e-05,
-      "loss": 0.2458,
+      "epoch": 3.201607380978124,
+      "grad_norm": 0.2334311157464981,
+      "learning_rate": 1.5145463630068268e-05,
+      "loss": 0.3758,
       "step": 88835
     },
     {
-      "epoch": 3.13,
-      "learning_rate": 1.6287094381249457e-05,
-      "loss": 0.2593,
+      "epoch": 3.2017875806393485,
+      "grad_norm": 0.21753154695034027,
+      "learning_rate": 1.5142781837154319e-05,
+      "loss": 0.3782,
       "step": 88840
     },
     {
-      "epoch": 3.13,
-      "learning_rate": 1.6284424323506564e-05,
-      "loss": 0.237,
+      "epoch": 3.201967780300573,
+      "grad_norm": 0.27690520882606506,
+      "learning_rate": 1.5140100178543057e-05,
+      "loss": 0.377,
       "step": 88845
     },
     {
-      "epoch": 3.13,
-      "learning_rate": 1.6281754378929684e-05,
-      "loss": 0.2576,
+      "epoch": 3.2021479799617976,
+      "grad_norm": 0.21353106200695038,
+      "learning_rate": 1.5137418654271025e-05,
+      "loss": 0.3773,
       "step": 88850
     },
     {
-      "epoch": 3.13,
-      "learning_rate": 1.6279084547553465e-05,
-      "loss": 0.2742,
+      "epoch": 3.2023281796230223,
+      "grad_norm": 0.19500304758548737,
+      "learning_rate": 1.513473726437476e-05,
+      "loss": 0.351,
       "step": 88855
     },
     {
-      "epoch": 3.13,
-      "learning_rate": 1.62764148294126e-05,
-      "loss": 0.2643,
+      "epoch": 3.202508379284247,
+      "grad_norm": 0.21235820651054382,
+      "learning_rate": 1.5132056008890771e-05,
+      "loss": 0.3599,
       "step": 88860
     },
     {
-      "epoch": 3.13,
-      "learning_rate": 1.6273745224541735e-05,
-      "loss": 0.2551,
+      "epoch": 3.2026885789454718,
+      "grad_norm": 0.2564550042152405,
+      "learning_rate": 1.5129374887855629e-05,
+      "loss": 0.3966,
       "step": 88865
     },
     {
-      "epoch": 3.13,
-      "learning_rate": 1.627107573297554e-05,
-      "loss": 0.2677,
+      "epoch": 3.202868778606696,
+      "grad_norm": 0.24082764983177185,
+      "learning_rate": 1.5126693901305836e-05,
+      "loss": 0.3829,
       "step": 88870
     },
     {
-      "epoch": 3.13,
-      "learning_rate": 1.6268406354748662e-05,
-      "loss": 0.2393,
+      "epoch": 3.203048978267921,
+      "grad_norm": 0.22911947965621948,
+      "learning_rate": 1.512401304927792e-05,
+      "loss": 0.3408,
       "step": 88875
     },
     {
-      "epoch": 3.13,
-      "learning_rate": 1.6265737089895783e-05,
-      "loss": 0.2798,
+      "epoch": 3.2032291779291455,
+      "grad_norm": 0.26279905438423157,
+      "learning_rate": 1.5121332331808421e-05,
+      "loss": 0.4124,
       "step": 88880
     },
     {
-      "epoch": 3.13,
-      "learning_rate": 1.6263067938451553e-05,
-      "loss": 0.2618,
+      "epoch": 3.2034093775903703,
+      "grad_norm": 0.2682557702064514,
+      "learning_rate": 1.511865174893385e-05,
+      "loss": 0.3681,
       "step": 88885
     },
     {
-      "epoch": 3.13,
-      "learning_rate": 1.626039890045063e-05,
-      "loss": 0.2606,
+      "epoch": 3.203589577251595,
+      "grad_norm": 0.25883546471595764,
+      "learning_rate": 1.5115971300690746e-05,
+      "loss": 0.4007,
       "step": 88890
     },
     {
-      "epoch": 3.13,
-      "learning_rate": 1.6257729975927656e-05,
-      "loss": 0.2465,
+      "epoch": 3.2037697769128193,
+      "grad_norm": 0.21053394675254822,
+      "learning_rate": 1.5113290987115614e-05,
+      "loss": 0.3837,
       "step": 88895
     },
     {
-      "epoch": 3.13,
-      "learning_rate": 1.62550611649173e-05,
-      "loss": 0.2683,
+      "epoch": 3.203949976574044,
+      "grad_norm": 0.22569601237773895,
+      "learning_rate": 1.5110610808244974e-05,
+      "loss": 0.3969,
       "step": 88900
     },
     {
-      "epoch": 3.13,
-      "learning_rate": 1.625239246745422e-05,
-      "loss": 0.2552,
+      "epoch": 3.2041301762352687,
+      "grad_norm": 0.24073798954486847,
+      "learning_rate": 1.5107930764115352e-05,
+      "loss": 0.3974,
       "step": 88905
     },
     {
-      "epoch": 3.13,
-      "learning_rate": 1.624972388357306e-05,
-      "loss": 0.2637,
+      "epoch": 3.2043103758964935,
+      "grad_norm": 0.24545975029468536,
+      "learning_rate": 1.5105250854763259e-05,
+      "loss": 0.3976,
       "step": 88910
     },
     {
-      "epoch": 3.13,
-      "learning_rate": 1.6247055413308455e-05,
-      "loss": 0.2644,
+      "epoch": 3.2044905755577178,
+      "grad_norm": 0.18453896045684814,
+      "learning_rate": 1.5102571080225203e-05,
+      "loss": 0.3786,
       "step": 88915
     },
     {
-      "epoch": 3.13,
-      "learning_rate": 1.624438705669508e-05,
-      "loss": 0.2455,
+      "epoch": 3.2046707752189425,
+      "grad_norm": 0.2695164084434509,
+      "learning_rate": 1.5099891440537705e-05,
+      "loss": 0.3783,
       "step": 88920
     },
     {
-      "epoch": 3.13,
-      "learning_rate": 1.624171881376756e-05,
-      "loss": 0.2699,
+      "epoch": 3.2048509748801672,
+      "grad_norm": 0.22112925350666046,
+      "learning_rate": 1.5097211935737263e-05,
+      "loss": 0.4013,
       "step": 88925
     },
     {
-      "epoch": 3.13,
-      "learning_rate": 1.6239050684560557e-05,
-      "loss": 0.2428,
+      "epoch": 3.205031174541392,
+      "grad_norm": 0.2127871960401535,
+      "learning_rate": 1.5094532565860398e-05,
+      "loss": 0.3885,
       "step": 88930
     },
     {
-      "epoch": 3.13,
-      "learning_rate": 1.6236382669108696e-05,
-      "loss": 0.2529,
+      "epoch": 3.2052113742026167,
+      "grad_norm": 0.22440172731876373,
+      "learning_rate": 1.5091853330943606e-05,
+      "loss": 0.3864,
       "step": 88935
     },
     {
-      "epoch": 3.13,
-      "learning_rate": 1.6233714767446645e-05,
-      "loss": 0.2753,
+      "epoch": 3.205391573863841,
+      "grad_norm": 0.2635382115840912,
+      "learning_rate": 1.508917423102339e-05,
+      "loss": 0.3542,
       "step": 88940
     },
     {
-      "epoch": 3.13,
-      "learning_rate": 1.6231046979609026e-05,
-      "loss": 0.2709,
+      "epoch": 3.2055717735250657,
+      "grad_norm": 0.20323844254016876,
+      "learning_rate": 1.5086495266136263e-05,
+      "loss": 0.3546,
       "step": 88945
     },
     {
-      "epoch": 3.13,
-      "learning_rate": 1.622837930563048e-05,
-      "loss": 0.2455,
+      "epoch": 3.2057519731862905,
+      "grad_norm": 0.21139764785766602,
+      "learning_rate": 1.5083816436318716e-05,
+      "loss": 0.3542,
       "step": 88950
     },
     {
-      "epoch": 3.13,
-      "learning_rate": 1.622571174554564e-05,
-      "loss": 0.2726,
+      "epoch": 3.205932172847515,
+      "grad_norm": 0.20650440454483032,
+      "learning_rate": 1.5081137741607249e-05,
+      "loss": 0.3755,
       "step": 88955
     },
     {
-      "epoch": 3.13,
-      "learning_rate": 1.6223044299389163e-05,
-      "loss": 0.2659,
+      "epoch": 3.2061123725087395,
+      "grad_norm": 0.22922226786613464,
+      "learning_rate": 1.5078459182038367e-05,
+      "loss": 0.3651,
       "step": 88960
     },
     {
-      "epoch": 3.13,
-      "learning_rate": 1.6220376967195673e-05,
-      "loss": 0.2429,
+      "epoch": 3.206292572169964,
+      "grad_norm": 0.2639060914516449,
+      "learning_rate": 1.5075780757648544e-05,
+      "loss": 0.4042,
       "step": 88965
     },
     {
-      "epoch": 3.13,
-      "learning_rate": 1.6217709748999787e-05,
-      "loss": 0.2912,
+      "epoch": 3.206472771831189,
+      "grad_norm": 0.20227624475955963,
+      "learning_rate": 1.5073102468474304e-05,
+      "loss": 0.3607,
       "step": 88970
     },
     {
-      "epoch": 3.13,
-      "learning_rate": 1.621504264483617e-05,
-      "loss": 0.2627,
+      "epoch": 3.2066529714924137,
+      "grad_norm": 0.23656994104385376,
+      "learning_rate": 1.5070424314552112e-05,
+      "loss": 0.4143,
       "step": 88975
     },
     {
-      "epoch": 3.13,
-      "learning_rate": 1.621237565473942e-05,
-      "loss": 0.2782,
+      "epoch": 3.2068331711536384,
+      "grad_norm": 0.21116869151592255,
+      "learning_rate": 1.5067746295918462e-05,
+      "loss": 0.3836,
       "step": 88980
     },
     {
-      "epoch": 3.13,
-      "learning_rate": 1.6210242144813453e-05,
-      "loss": 0.27,
+      "epoch": 3.2070133708148627,
+      "grad_norm": 0.20764394104480743,
+      "learning_rate": 1.506506841260985e-05,
+      "loss": 0.3969,
       "step": 88985
     },
     {
-      "epoch": 3.13,
-      "learning_rate": 1.620757536012436e-05,
-      "loss": 0.2398,
+      "epoch": 3.2071935704760874,
+      "grad_norm": 0.2260044366121292,
+      "learning_rate": 1.5062390664662757e-05,
+      "loss": 0.3741,
       "step": 88990
     },
     {
-      "epoch": 3.13,
-      "learning_rate": 1.6204908689599114e-05,
-      "loss": 0.2631,
+      "epoch": 3.207373770137312,
+      "grad_norm": 0.2571514844894409,
+      "learning_rate": 1.5059713052113666e-05,
+      "loss": 0.3532,
       "step": 88995
     },
     {
-      "epoch": 3.13,
-      "learning_rate": 1.6202242133272327e-05,
-      "loss": 0.2362,
+      "epoch": 3.207553969798537,
+      "grad_norm": 0.27601492404937744,
+      "learning_rate": 1.5057035574999067e-05,
+      "loss": 0.3968,
       "step": 89000
     },
     {
-      "epoch": 3.13,
-      "eval_loss": 0.25561562180519104,
-      "eval_runtime": 10.5484,
-      "eval_samples_per_second": 9.48,
-      "eval_steps_per_second": 9.48,
+      "epoch": 3.207553969798537,
+      "eval_loss": 0.4308409094810486,
+      "eval_runtime": 3.5383,
+      "eval_samples_per_second": 28.262,
+      "eval_steps_per_second": 7.066,
       "step": 89000
     },
     {
-      "epoch": 3.13,
-      "learning_rate": 1.619957569117862e-05,
-      "loss": 0.2543,
+      "epoch": 3.207734169459761,
+      "grad_norm": 0.22598758339881897,
+      "learning_rate": 1.5054358233355423e-05,
+      "loss": 0.373,
       "step": 89005
     },
     {
-      "epoch": 3.13,
-      "learning_rate": 1.6196909363352637e-05,
-      "loss": 0.2474,
+      "epoch": 3.207914369120986,
+      "grad_norm": 0.23747682571411133,
+      "learning_rate": 1.5051681027219228e-05,
+      "loss": 0.366,
       "step": 89010
     },
     {
-      "epoch": 3.13,
-      "learning_rate": 1.6194243149828974e-05,
-      "loss": 0.2468,
+      "epoch": 3.2080945687822107,
+      "grad_norm": 0.1627287119626999,
+      "learning_rate": 1.5049003956626957e-05,
+      "loss": 0.3545,
       "step": 89015
     },
     {
-      "epoch": 3.13,
-      "learning_rate": 1.6191577050642267e-05,
-      "loss": 0.2666,
+      "epoch": 3.2082747684434354,
+      "grad_norm": 0.28767460584640503,
+      "learning_rate": 1.504632702161507e-05,
+      "loss": 0.3935,
       "step": 89020
     },
     {
-      "epoch": 3.13,
-      "learning_rate": 1.6188911065827118e-05,
-      "loss": 0.2764,
+      "epoch": 3.20845496810466,
+      "grad_norm": 0.28209447860717773,
+      "learning_rate": 1.5043650222220057e-05,
+      "loss": 0.3836,
       "step": 89025
     },
     {
-      "epoch": 3.13,
-      "learning_rate": 1.6186245195418162e-05,
-      "loss": 0.2577,
+      "epoch": 3.2086351677658844,
+      "grad_norm": 0.2737613022327423,
+      "learning_rate": 1.504097355847838e-05,
+      "loss": 0.3978,
       "step": 89030
     },
     {
-      "epoch": 3.13,
-      "learning_rate": 1.6183579439449998e-05,
-      "loss": 0.2554,
+      "epoch": 3.208815367427109,
+      "grad_norm": 0.2262285202741623,
+      "learning_rate": 1.5038297030426513e-05,
+      "loss": 0.3948,
       "step": 89035
     },
     {
-      "epoch": 3.13,
-      "learning_rate": 1.6180913797957246e-05,
-      "loss": 0.2567,
+      "epoch": 3.208995567088334,
+      "grad_norm": 0.24471984803676605,
+      "learning_rate": 1.5035620638100926e-05,
+      "loss": 0.4265,
       "step": 89040
     },
     {
-      "epoch": 3.13,
-      "learning_rate": 1.6178248270974515e-05,
-      "loss": 0.273,
+      "epoch": 3.2091757667495586,
+      "grad_norm": 0.2190433293581009,
+      "learning_rate": 1.5032944381538061e-05,
+      "loss": 0.3845,
       "step": 89045
     },
     {
-      "epoch": 3.13,
-      "learning_rate": 1.617558285853642e-05,
-      "loss": 0.2565,
+      "epoch": 3.209355966410783,
+      "grad_norm": 0.22950243949890137,
+      "learning_rate": 1.503026826077441e-05,
+      "loss": 0.4014,
       "step": 89050
     },
     {
-      "epoch": 3.13,
-      "learning_rate": 1.617291756067757e-05,
-      "loss": 0.2484,
+      "epoch": 3.2095361660720076,
+      "grad_norm": 0.23515675961971283,
+      "learning_rate": 1.502759227584643e-05,
+      "loss": 0.3942,
       "step": 89055
     },
     {
-      "epoch": 3.13,
-      "learning_rate": 1.6170252377432554e-05,
-      "loss": 0.2745,
+      "epoch": 3.2097163657332324,
+      "grad_norm": 0.24395707249641418,
+      "learning_rate": 1.502491642679056e-05,
+      "loss": 0.3746,
       "step": 89060
     },
     {
-      "epoch": 3.13,
-      "learning_rate": 1.6167587308836007e-05,
-      "loss": 0.2313,
+      "epoch": 3.209896565394457,
+      "grad_norm": 0.28980615735054016,
+      "learning_rate": 1.5022240713643287e-05,
+      "loss": 0.401,
       "step": 89065
     },
     {
-      "epoch": 3.13,
-      "learning_rate": 1.6164922354922507e-05,
-      "loss": 0.2444,
+      "epoch": 3.210076765055682,
+      "grad_norm": 0.1918734461069107,
+      "learning_rate": 1.5019565136441038e-05,
+      "loss": 0.3971,
       "step": 89070
     },
     {
-      "epoch": 3.13,
-      "learning_rate": 1.6162257515726678e-05,
-      "loss": 0.2472,
+      "epoch": 3.210256964716906,
+      "grad_norm": 0.22146931290626526,
+      "learning_rate": 1.50168896952203e-05,
+      "loss": 0.3756,
       "step": 89075
     },
     {
-      "epoch": 3.13,
-      "learning_rate": 1.6159592791283097e-05,
-      "loss": 0.2551,
+      "epoch": 3.210437164378131,
+      "grad_norm": 0.20011606812477112,
+      "learning_rate": 1.5014214390017496e-05,
+      "loss": 0.3584,
       "step": 89080
     },
     {
-      "epoch": 3.13,
-      "learning_rate": 1.615692818162639e-05,
-      "loss": 0.2704,
+      "epoch": 3.2106173640393556,
+      "grad_norm": 0.2056010365486145,
+      "learning_rate": 1.5011539220869084e-05,
+      "loss": 0.3509,
       "step": 89085
     },
     {
-      "epoch": 3.13,
-      "learning_rate": 1.6154263686791143e-05,
-      "loss": 0.266,
+      "epoch": 3.2107975637005803,
+      "grad_norm": 0.2631101608276367,
+      "learning_rate": 1.5008864187811522e-05,
+      "loss": 0.3855,
       "step": 89090
     },
     {
-      "epoch": 3.13,
-      "learning_rate": 1.6151599306811946e-05,
-      "loss": 0.2748,
+      "epoch": 3.210977763361805,
+      "grad_norm": 0.2047474980354309,
+      "learning_rate": 1.500618929088125e-05,
+      "loss": 0.3843,
       "step": 89095
     },
     {
-      "epoch": 3.13,
-      "learning_rate": 1.6148935041723396e-05,
-      "loss": 0.2489,
+      "epoch": 3.2111579630230294,
+      "grad_norm": 0.2543022632598877,
+      "learning_rate": 1.5003514530114712e-05,
+      "loss": 0.3901,
       "step": 89100
     },
     {
-      "epoch": 3.13,
-      "learning_rate": 1.61462708915601e-05,
-      "loss": 0.24,
+      "epoch": 3.211338162684254,
+      "grad_norm": 0.22199147939682007,
+      "learning_rate": 1.5000839905548359e-05,
+      "loss": 0.372,
       "step": 89105
     },
     {
-      "epoch": 3.14,
-      "learning_rate": 1.6143606856356644e-05,
-      "loss": 0.2735,
+      "epoch": 3.211518362345479,
+      "grad_norm": 0.19281704723834991,
+      "learning_rate": 1.4998165417218618e-05,
+      "loss": 0.341,
       "step": 89110
     },
     {
-      "epoch": 3.14,
-      "learning_rate": 1.6140942936147615e-05,
-      "loss": 0.2626,
+      "epoch": 3.2116985620067036,
+      "grad_norm": 0.2250460386276245,
+      "learning_rate": 1.499549106516194e-05,
+      "loss": 0.3559,
       "step": 89115
     },
     {
-      "epoch": 3.14,
-      "learning_rate": 1.6138279130967594e-05,
-      "loss": 0.2559,
+      "epoch": 3.2118787616679283,
+      "grad_norm": 0.2156555950641632,
+      "learning_rate": 1.4992816849414765e-05,
+      "loss": 0.3516,
       "step": 89120
     },
     {
-      "epoch": 3.14,
-      "learning_rate": 1.6135615440851185e-05,
-      "loss": 0.2752,
+      "epoch": 3.2120589613291526,
+      "grad_norm": 0.2305130809545517,
+      "learning_rate": 1.4990142770013512e-05,
+      "loss": 0.3911,
       "step": 89125
     },
     {
-      "epoch": 3.14,
-      "learning_rate": 1.6132951865832968e-05,
-      "loss": 0.2571,
+      "epoch": 3.2122391609903773,
+      "grad_norm": 0.23964077234268188,
+      "learning_rate": 1.4987468826994632e-05,
+      "loss": 0.4231,
       "step": 89130
     },
     {
-      "epoch": 3.14,
-      "learning_rate": 1.6130288405947534e-05,
-      "loss": 0.2491,
+      "epoch": 3.212419360651602,
+      "grad_norm": 0.2621869742870331,
+      "learning_rate": 1.4984795020394548e-05,
+      "loss": 0.378,
       "step": 89135
     },
     {
-      "epoch": 3.14,
-      "learning_rate": 1.6127625061229447e-05,
-      "loss": 0.2669,
+      "epoch": 3.212599560312827,
+      "grad_norm": 0.18443959951400757,
+      "learning_rate": 1.49821213502497e-05,
+      "loss": 0.3616,
       "step": 89140
     },
     {
-      "epoch": 3.14,
-      "learning_rate": 1.612496183171331e-05,
-      "loss": 0.2549,
+      "epoch": 3.212779759974051,
+      "grad_norm": 0.20639346539974213,
+      "learning_rate": 1.4979447816596508e-05,
+      "loss": 0.3615,
       "step": 89145
     },
     {
-      "epoch": 3.14,
-      "learning_rate": 1.612229871743369e-05,
-      "loss": 0.27,
+      "epoch": 3.212959959635276,
+      "grad_norm": 0.24619300663471222,
+      "learning_rate": 1.497677441947139e-05,
+      "loss": 0.4164,
       "step": 89150
     },
     {
-      "epoch": 3.14,
-      "learning_rate": 1.6119635718425163e-05,
-      "loss": 0.2503,
+      "epoch": 3.2131401592965005,
+      "grad_norm": 0.3176526427268982,
+      "learning_rate": 1.4974101158910791e-05,
+      "loss": 0.3682,
       "step": 89155
     },
     {
-      "epoch": 3.14,
-      "learning_rate": 1.6116972834722328e-05,
-      "loss": 0.2405,
+      "epoch": 3.2133203589577253,
+      "grad_norm": 0.2009301781654358,
+      "learning_rate": 1.497142803495112e-05,
+      "loss": 0.3765,
       "step": 89160
     },
     {
-      "epoch": 3.14,
-      "learning_rate": 1.6114310066359746e-05,
-      "loss": 0.2495,
+      "epoch": 3.21350055861895,
+      "grad_norm": 0.2197410613298416,
+      "learning_rate": 1.4968755047628796e-05,
+      "loss": 0.3795,
       "step": 89165
     },
     {
-      "epoch": 3.14,
-      "learning_rate": 1.6111647413371992e-05,
-      "loss": 0.2459,
+      "epoch": 3.2136807582801743,
+      "grad_norm": 0.23432065546512604,
+      "learning_rate": 1.4966082196980247e-05,
+      "loss": 0.3779,
       "step": 89170
     },
     {
-      "epoch": 3.14,
-      "learning_rate": 1.6108984875793626e-05,
-      "loss": 0.2613,
+      "epoch": 3.213860957941399,
+      "grad_norm": 0.29888421297073364,
+      "learning_rate": 1.4963409483041874e-05,
+      "loss": 0.3599,
       "step": 89175
     },
     {
-      "epoch": 3.14,
-      "learning_rate": 1.610632245365924e-05,
-      "loss": 0.2814,
+      "epoch": 3.2140411576026238,
+      "grad_norm": 0.19982536137104034,
+      "learning_rate": 1.496073690585012e-05,
+      "loss": 0.3989,
       "step": 89180
     },
     {
-      "epoch": 3.14,
-      "learning_rate": 1.6103660147003402e-05,
-      "loss": 0.2423,
+      "epoch": 3.2142213572638485,
+      "grad_norm": 0.22147499024868011,
+      "learning_rate": 1.4958064465441374e-05,
+      "loss": 0.4284,
       "step": 89185
     },
     {
-      "epoch": 3.14,
-      "learning_rate": 1.610099795586067e-05,
-      "loss": 0.2627,
+      "epoch": 3.214401556925073,
+      "grad_norm": 0.23238062858581543,
+      "learning_rate": 1.4955392161852056e-05,
+      "loss": 0.368,
       "step": 89190
     },
     {
-      "epoch": 3.14,
-      "learning_rate": 1.6098335880265608e-05,
-      "loss": 0.2506,
+      "epoch": 3.2145817565862975,
+      "grad_norm": 0.251499205827713,
+      "learning_rate": 1.4952719995118574e-05,
+      "loss": 0.4231,
       "step": 89195
     },
     {
-      "epoch": 3.14,
-      "learning_rate": 1.6095673920252795e-05,
-      "loss": 0.2539,
+      "epoch": 3.2147619562475223,
+      "grad_norm": 0.24169008433818817,
+      "learning_rate": 1.4950047965277342e-05,
+      "loss": 0.3682,
       "step": 89200
     },
     {
-      "epoch": 3.14,
-      "learning_rate": 1.6093012075856787e-05,
-      "loss": 0.2608,
+      "epoch": 3.214942155908747,
+      "grad_norm": 0.21621721982955933,
+      "learning_rate": 1.4947376072364752e-05,
+      "loss": 0.3801,
       "step": 89205
     },
     {
-      "epoch": 3.14,
-      "learning_rate": 1.6090350347112148e-05,
-      "loss": 0.267,
+      "epoch": 3.2151223555699717,
+      "grad_norm": 0.2918843626976013,
+      "learning_rate": 1.4944704316417223e-05,
+      "loss": 0.3745,
       "step": 89210
     },
     {
-      "epoch": 3.14,
-      "learning_rate": 1.608768873405343e-05,
-      "loss": 0.2623,
+      "epoch": 3.215302555231196,
+      "grad_norm": 0.1639867126941681,
+      "learning_rate": 1.4942032697471148e-05,
+      "loss": 0.3755,
       "step": 89215
     },
     {
-      "epoch": 3.14,
-      "learning_rate": 1.6085027236715212e-05,
-      "loss": 0.2672,
+      "epoch": 3.2154827548924207,
+      "grad_norm": 0.2433352768421173,
+      "learning_rate": 1.493936121556293e-05,
+      "loss": 0.3648,
       "step": 89220
     },
     {
-      "epoch": 3.14,
-      "learning_rate": 1.608236585513204e-05,
-      "loss": 0.2301,
+      "epoch": 3.2156629545536455,
+      "grad_norm": 0.20584319531917572,
+      "learning_rate": 1.4936689870728974e-05,
+      "loss": 0.3858,
       "step": 89225
     },
     {
-      "epoch": 3.14,
-      "learning_rate": 1.6079704589338463e-05,
-      "loss": 0.2536,
+      "epoch": 3.21584315421487,
+      "grad_norm": 0.2562110424041748,
+      "learning_rate": 1.4934018663005662e-05,
+      "loss": 0.3917,
       "step": 89230
     },
     {
-      "epoch": 3.14,
-      "learning_rate": 1.6077043439369037e-05,
-      "loss": 0.2361,
+      "epoch": 3.2160233538760945,
+      "grad_norm": 0.19980406761169434,
+      "learning_rate": 1.4931347592429401e-05,
+      "loss": 0.3751,
       "step": 89235
     },
     {
-      "epoch": 3.14,
-      "learning_rate": 1.6074382405258337e-05,
-      "loss": 0.273,
+      "epoch": 3.2162035535373192,
+      "grad_norm": 0.24331307411193848,
+      "learning_rate": 1.4928676659036586e-05,
+      "loss": 0.3779,
       "step": 89240
     },
     {
-      "epoch": 3.14,
-      "learning_rate": 1.6071721487040897e-05,
-      "loss": 0.2442,
+      "epoch": 3.216383753198544,
+      "grad_norm": 0.20756937563419342,
+      "learning_rate": 1.4926005862863584e-05,
+      "loss": 0.38,
       "step": 89245
     },
     {
-      "epoch": 3.14,
-      "learning_rate": 1.606906068475127e-05,
-      "loss": 0.2665,
+      "epoch": 3.2165639528597687,
+      "grad_norm": 0.21686789393424988,
+      "learning_rate": 1.492333520394682e-05,
+      "loss": 0.3911,
       "step": 89250
     },
     {
-      "epoch": 3.14,
-      "learning_rate": 1.606639999842399e-05,
-      "loss": 0.2583,
+      "epoch": 3.2167441525209934,
+      "grad_norm": 0.1949186772108078,
+      "learning_rate": 1.492066468232265e-05,
+      "loss": 0.3791,
       "step": 89255
     },
     {
-      "epoch": 3.14,
-      "learning_rate": 1.6063739428093626e-05,
-      "loss": 0.2467,
+      "epoch": 3.2169243521822177,
+      "grad_norm": 0.19429568946361542,
+      "learning_rate": 1.4917994298027482e-05,
+      "loss": 0.3407,
       "step": 89260
     },
     {
-      "epoch": 3.14,
-      "learning_rate": 1.6061078973794725e-05,
-      "loss": 0.2623,
+      "epoch": 3.2171045518434425,
+      "grad_norm": 0.25820302963256836,
+      "learning_rate": 1.4915324051097688e-05,
+      "loss": 0.3799,
       "step": 89265
     },
     {
-      "epoch": 3.14,
-      "learning_rate": 1.6058418635561805e-05,
-      "loss": 0.2491,
+      "epoch": 3.217284751504667,
+      "grad_norm": 0.2778339684009552,
+      "learning_rate": 1.4912653941569642e-05,
+      "loss": 0.3973,
       "step": 89270
     },
     {
-      "epoch": 3.14,
-      "learning_rate": 1.605575841342944e-05,
-      "loss": 0.2509,
+      "epoch": 3.217464951165892,
+      "grad_norm": 0.2582213878631592,
+      "learning_rate": 1.4909983969479747e-05,
+      "loss": 0.3764,
       "step": 89275
     },
     {
-      "epoch": 3.14,
-      "learning_rate": 1.6053098307432162e-05,
-      "loss": 0.2567,
+      "epoch": 3.217645150827116,
+      "grad_norm": 0.1809670329093933,
+      "learning_rate": 1.490731413486436e-05,
+      "loss": 0.3945,
       "step": 89280
     },
     {
-      "epoch": 3.14,
-      "learning_rate": 1.6050438317604496e-05,
-      "loss": 0.2441,
+      "epoch": 3.217825350488341,
+      "grad_norm": 0.2597140371799469,
+      "learning_rate": 1.490464443775986e-05,
+      "loss": 0.4122,
       "step": 89285
     },
     {
-      "epoch": 3.14,
-      "learning_rate": 1.6047778443980992e-05,
-      "loss": 0.2644,
+      "epoch": 3.2180055501495657,
+      "grad_norm": 0.23340429365634918,
+      "learning_rate": 1.490197487820263e-05,
+      "loss": 0.3487,
       "step": 89290
     },
     {
-      "epoch": 3.14,
-      "learning_rate": 1.6045118686596182e-05,
-      "loss": 0.2314,
+      "epoch": 3.2181857498107904,
+      "grad_norm": 0.2497396469116211,
+      "learning_rate": 1.4899305456229032e-05,
+      "loss": 0.3891,
       "step": 89295
     },
     {
-      "epoch": 3.14,
-      "learning_rate": 1.6042459045484614e-05,
-      "loss": 0.2554,
+      "epoch": 3.218365949472015,
+      "grad_norm": 0.28371942043304443,
+      "learning_rate": 1.4896636171875447e-05,
+      "loss": 0.4068,
       "step": 89300
     },
     {
-      "epoch": 3.14,
-      "learning_rate": 1.603979952068081e-05,
-      "loss": 0.2759,
+      "epoch": 3.2185461491332394,
+      "grad_norm": 0.21911130845546722,
+      "learning_rate": 1.489396702517824e-05,
+      "loss": 0.4053,
       "step": 89305
     },
     {
-      "epoch": 3.14,
-      "learning_rate": 1.6037140112219294e-05,
-      "loss": 0.2714,
+      "epoch": 3.218726348794464,
+      "grad_norm": 0.2320789396762848,
+      "learning_rate": 1.489129801617377e-05,
+      "loss": 0.3911,
       "step": 89310
     },
     {
-      "epoch": 3.14,
-      "learning_rate": 1.6034480820134615e-05,
-      "loss": 0.2457,
+      "epoch": 3.218906548455689,
+      "grad_norm": 0.21250326931476593,
+      "learning_rate": 1.4888629144898409e-05,
+      "loss": 0.3706,
       "step": 89315
     },
     {
-      "epoch": 3.14,
-      "learning_rate": 1.6031821644461296e-05,
-      "loss": 0.2447,
+      "epoch": 3.2190867481169136,
+      "grad_norm": 0.26723039150238037,
+      "learning_rate": 1.4885960411388517e-05,
+      "loss": 0.3946,
       "step": 89320
     },
     {
-      "epoch": 3.14,
-      "learning_rate": 1.6029162585233866e-05,
-      "loss": 0.2676,
+      "epoch": 3.2192669477781384,
+      "grad_norm": 0.20231136679649353,
+      "learning_rate": 1.4883291815680459e-05,
+      "loss": 0.3582,
       "step": 89325
     },
     {
-      "epoch": 3.14,
-      "learning_rate": 1.6026503642486835e-05,
-      "loss": 0.2476,
+      "epoch": 3.2194471474393627,
+      "grad_norm": 0.2787347435951233,
+      "learning_rate": 1.4880623357810597e-05,
+      "loss": 0.3847,
       "step": 89330
     },
     {
-      "epoch": 3.14,
-      "learning_rate": 1.602384481625475e-05,
-      "loss": 0.2467,
+      "epoch": 3.2196273471005874,
+      "grad_norm": 0.2298644334077835,
+      "learning_rate": 1.4877955037815267e-05,
+      "loss": 0.3643,
       "step": 89335
     },
     {
-      "epoch": 3.14,
-      "learning_rate": 1.602118610657212e-05,
-      "loss": 0.2636,
+      "epoch": 3.219807546761812,
+      "grad_norm": 0.23724813759326935,
+      "learning_rate": 1.4875286855730852e-05,
+      "loss": 0.4065,
       "step": 89340
     },
     {
-      "epoch": 3.14,
-      "learning_rate": 1.6018527513473476e-05,
-      "loss": 0.2591,
+      "epoch": 3.219987746423037,
+      "grad_norm": 0.1757110357284546,
+      "learning_rate": 1.4872618811593697e-05,
+      "loss": 0.3819,
       "step": 89345
     },
     {
-      "epoch": 3.14,
-      "learning_rate": 1.6015869036993326e-05,
-      "loss": 0.2523,
+      "epoch": 3.220167946084261,
+      "grad_norm": 0.20195148885250092,
+      "learning_rate": 1.4869950905440136e-05,
+      "loss": 0.3823,
       "step": 89350
     },
     {
-      "epoch": 3.14,
-      "learning_rate": 1.60132106771662e-05,
-      "loss": 0.2732,
+      "epoch": 3.220348145745486,
+      "grad_norm": 0.1882009357213974,
+      "learning_rate": 1.4867283137306547e-05,
+      "loss": 0.339,
       "step": 89355
     },
     {
-      "epoch": 3.14,
-      "learning_rate": 1.6010552434026615e-05,
-      "loss": 0.2625,
+      "epoch": 3.2205283454067106,
+      "grad_norm": 0.24396227300167084,
+      "learning_rate": 1.4864615507229246e-05,
+      "loss": 0.3963,
       "step": 89360
     },
     {
-      "epoch": 3.14,
-      "learning_rate": 1.6007894307609078e-05,
-      "loss": 0.2702,
+      "epoch": 3.2207085450679354,
+      "grad_norm": 0.2632978558540344,
+      "learning_rate": 1.4861948015244615e-05,
+      "loss": 0.4177,
       "step": 89365
     },
     {
-      "epoch": 3.14,
-      "learning_rate": 1.6005236297948107e-05,
-      "loss": 0.2705,
+      "epoch": 3.22088874472916,
+      "grad_norm": 0.20694774389266968,
+      "learning_rate": 1.4859280661388974e-05,
+      "loss": 0.367,
       "step": 89370
     },
     {
-      "epoch": 3.14,
-      "learning_rate": 1.600257840507821e-05,
-      "loss": 0.2774,
+      "epoch": 3.2210689443903844,
+      "grad_norm": 0.21386392414569855,
+      "learning_rate": 1.4856613445698664e-05,
+      "loss": 0.359,
       "step": 89375
     },
     {
-      "epoch": 3.14,
-      "learning_rate": 1.5999920629033914e-05,
-      "loss": 0.2673,
+      "epoch": 3.221249144051609,
+      "grad_norm": 0.1855984479188919,
+      "learning_rate": 1.4853946368210036e-05,
+      "loss": 0.3242,
       "step": 89380
     },
     {
-      "epoch": 3.14,
-      "learning_rate": 1.5997262969849712e-05,
-      "loss": 0.25,
+      "epoch": 3.221429343712834,
+      "grad_norm": 0.23867137730121613,
+      "learning_rate": 1.485127942895943e-05,
+      "loss": 0.363,
       "step": 89385
     },
     {
-      "epoch": 3.14,
-      "learning_rate": 1.5994605427560112e-05,
-      "loss": 0.2511,
+      "epoch": 3.2216095433740586,
+      "grad_norm": 0.19533032178878784,
+      "learning_rate": 1.4848612627983166e-05,
+      "loss": 0.4011,
       "step": 89390
     },
     {
-      "epoch": 3.15,
-      "learning_rate": 1.599194800219963e-05,
-      "loss": 0.2466,
+      "epoch": 3.2217897430352833,
+      "grad_norm": 0.22574682533740997,
+      "learning_rate": 1.4845945965317598e-05,
+      "loss": 0.3869,
       "step": 89395
     },
     {
-      "epoch": 3.15,
-      "learning_rate": 1.598929069380277e-05,
-      "loss": 0.2701,
+      "epoch": 3.2219699426965076,
+      "grad_norm": 0.268163800239563,
+      "learning_rate": 1.4843279440999044e-05,
+      "loss": 0.3802,
       "step": 89400
     },
     {
-      "epoch": 3.15,
-      "learning_rate": 1.5986633502404024e-05,
-      "loss": 0.2589,
+      "epoch": 3.2221501423577323,
+      "grad_norm": 0.23362210392951965,
+      "learning_rate": 1.4840613055063845e-05,
+      "loss": 0.4094,
       "step": 89405
     },
     {
-      "epoch": 3.15,
-      "learning_rate": 1.598397642803791e-05,
-      "loss": 0.2441,
+      "epoch": 3.222330342018957,
+      "grad_norm": 0.2540440559387207,
+      "learning_rate": 1.483794680754833e-05,
+      "loss": 0.366,
       "step": 89410
     },
     {
-      "epoch": 3.15,
-      "learning_rate": 1.598131947073892e-05,
-      "loss": 0.2767,
+      "epoch": 3.222510541680182,
+      "grad_norm": 0.2499903291463852,
+      "learning_rate": 1.4835280698488812e-05,
+      "loss": 0.3656,
       "step": 89415
     },
     {
-      "epoch": 3.15,
-      "learning_rate": 1.597866263054155e-05,
-      "loss": 0.2721,
+      "epoch": 3.222690741341406,
+      "grad_norm": 0.21211077272891998,
+      "learning_rate": 1.4832614727921634e-05,
+      "loss": 0.3654,
       "step": 89420
     },
     {
-      "epoch": 3.15,
-      "learning_rate": 1.5976005907480303e-05,
-      "loss": 0.2415,
+      "epoch": 3.222870941002631,
+      "grad_norm": 0.20574845373630524,
+      "learning_rate": 1.482994889588312e-05,
+      "loss": 0.39,
       "step": 89425
     },
     {
-      "epoch": 3.15,
-      "learning_rate": 1.5973349301589668e-05,
-      "loss": 0.276,
+      "epoch": 3.2230511406638556,
+      "grad_norm": 0.21333856880664825,
+      "learning_rate": 1.4827283202409564e-05,
+      "loss": 0.4014,
       "step": 89430
     },
     {
-      "epoch": 3.15,
-      "learning_rate": 1.5970692812904148e-05,
-      "loss": 0.264,
+      "epoch": 3.2232313403250803,
+      "grad_norm": 0.1983533501625061,
+      "learning_rate": 1.4824617647537323e-05,
+      "loss": 0.3733,
       "step": 89435
     },
     {
-      "epoch": 3.15,
-      "learning_rate": 1.5968036441458233e-05,
-      "loss": 0.244,
+      "epoch": 3.223411539986305,
+      "grad_norm": 0.1866043210029602,
+      "learning_rate": 1.482195223130268e-05,
+      "loss": 0.3471,
       "step": 89440
     },
     {
-      "epoch": 3.15,
-      "learning_rate": 1.59653801872864e-05,
-      "loss": 0.2484,
+      "epoch": 3.2235917396475293,
+      "grad_norm": 0.22153297066688538,
+      "learning_rate": 1.481928695374198e-05,
+      "loss": 0.3847,
       "step": 89445
     },
     {
-      "epoch": 3.15,
-      "learning_rate": 1.5962724050423168e-05,
-      "loss": 0.2579,
+      "epoch": 3.223771939308754,
+      "grad_norm": 0.26498228311538696,
+      "learning_rate": 1.4816621814891524e-05,
+      "loss": 0.3581,
       "step": 89450
     },
     {
-      "epoch": 3.15,
-      "learning_rate": 1.5960068030902997e-05,
-      "loss": 0.2605,
+      "epoch": 3.223952138969979,
+      "grad_norm": 0.24101939797401428,
+      "learning_rate": 1.4813956814787616e-05,
+      "loss": 0.3513,
       "step": 89455
     },
     {
-      "epoch": 3.15,
-      "learning_rate": 1.595741212876039e-05,
-      "loss": 0.254,
+      "epoch": 3.2241323386312035,
+      "grad_norm": 0.23949559032917023,
+      "learning_rate": 1.481129195346658e-05,
+      "loss": 0.375,
       "step": 89460
     },
     {
-      "epoch": 3.15,
-      "learning_rate": 1.595475634402982e-05,
-      "loss": 0.2555,
+      "epoch": 3.224312538292428,
+      "grad_norm": 0.20637662708759308,
+      "learning_rate": 1.4808627230964717e-05,
+      "loss": 0.3764,
       "step": 89465
     },
     {
-      "epoch": 3.15,
-      "learning_rate": 1.5952100676745792e-05,
-      "loss": 0.2482,
+      "epoch": 3.2244927379536525,
+      "grad_norm": 0.25291964411735535,
+      "learning_rate": 1.4805962647318333e-05,
+      "loss": 0.3823,
       "step": 89470
     },
     {
-      "epoch": 3.15,
-      "learning_rate": 1.594944512694276e-05,
-      "loss": 0.2741,
+      "epoch": 3.2246729376148773,
+      "grad_norm": 0.2150401920080185,
+      "learning_rate": 1.4803298202563737e-05,
+      "loss": 0.3555,
       "step": 89475
     },
     {
-      "epoch": 3.15,
-      "learning_rate": 1.5946789694655228e-05,
-      "loss": 0.2761,
+      "epoch": 3.224853137276102,
+      "grad_norm": 0.25628653168678284,
+      "learning_rate": 1.4800633896737226e-05,
+      "loss": 0.3762,
       "step": 89480
     },
     {
-      "epoch": 3.15,
-      "learning_rate": 1.5944134379917657e-05,
-      "loss": 0.2633,
+      "epoch": 3.2250333369373267,
+      "grad_norm": 0.2144256830215454,
+      "learning_rate": 1.4797969729875106e-05,
+      "loss": 0.3503,
       "step": 89485
     },
     {
-      "epoch": 3.15,
-      "learning_rate": 1.5941479182764542e-05,
-      "loss": 0.236,
+      "epoch": 3.225213536598551,
+      "grad_norm": 0.2198164016008377,
+      "learning_rate": 1.4795305702013674e-05,
+      "loss": 0.3467,
       "step": 89490
     },
     {
-      "epoch": 3.15,
-      "learning_rate": 1.5938824103230344e-05,
-      "loss": 0.2626,
+      "epoch": 3.2253937362597758,
+      "grad_norm": 0.2541256844997406,
+      "learning_rate": 1.4792641813189222e-05,
+      "loss": 0.3787,
       "step": 89495
     },
     {
-      "epoch": 3.15,
-      "learning_rate": 1.5936169141349548e-05,
-      "loss": 0.2246,
+      "epoch": 3.2255739359210005,
+      "grad_norm": 0.21869847178459167,
+      "learning_rate": 1.4789978063438054e-05,
+      "loss": 0.3738,
       "step": 89500
     },
     {
-      "epoch": 3.15,
-      "eval_loss": 0.25540173053741455,
-      "eval_runtime": 10.5366,
-      "eval_samples_per_second": 9.491,
-      "eval_steps_per_second": 9.491,
+      "epoch": 3.2255739359210005,
+      "eval_loss": 0.43070846796035767,
+      "eval_runtime": 3.5348,
+      "eval_samples_per_second": 28.29,
+      "eval_steps_per_second": 7.073,
       "step": 89500
     },
     {
-      "epoch": 3.15,
-      "learning_rate": 1.5933514297156614e-05,
-      "loss": 0.2681,
+      "epoch": 3.2257541355822252,
+      "grad_norm": 0.3209418058395386,
+      "learning_rate": 1.4787314452796458e-05,
+      "loss": 0.3811,
       "step": 89505
     },
     {
-      "epoch": 3.15,
-      "learning_rate": 1.593085957068603e-05,
-      "loss": 0.2602,
+      "epoch": 3.2259343352434495,
+      "grad_norm": 0.2772040069103241,
+      "learning_rate": 1.4784650981300718e-05,
+      "loss": 0.3696,
       "step": 89510
     },
     {
-      "epoch": 3.15,
-      "learning_rate": 1.5928204961972257e-05,
-      "loss": 0.2618,
+      "epoch": 3.2261145349046743,
+      "grad_norm": 0.2352096289396286,
+      "learning_rate": 1.4781987648987138e-05,
+      "loss": 0.3697,
       "step": 89515
     },
     {
-      "epoch": 3.15,
-      "learning_rate": 1.5925550471049756e-05,
-      "loss": 0.2692,
+      "epoch": 3.226294734565899,
+      "grad_norm": 0.19842156767845154,
+      "learning_rate": 1.4779324455891994e-05,
+      "loss": 0.3671,
       "step": 89520
     },
     {
-      "epoch": 3.15,
-      "learning_rate": 1.592289609795301e-05,
-      "loss": 0.2481,
+      "epoch": 3.2264749342271237,
+      "grad_norm": 0.2521081864833832,
+      "learning_rate": 1.4776661402051578e-05,
+      "loss": 0.3927,
       "step": 89525
     },
     {
-      "epoch": 3.15,
-      "learning_rate": 1.5920241842716477e-05,
-      "loss": 0.2424,
+      "epoch": 3.2266551338883485,
+      "grad_norm": 0.23003843426704407,
+      "learning_rate": 1.4773998487502178e-05,
+      "loss": 0.3673,
       "step": 89530
     },
     {
-      "epoch": 3.15,
-      "learning_rate": 1.5917587705374614e-05,
-      "loss": 0.2458,
+      "epoch": 3.2268353335495727,
+      "grad_norm": 0.20491987466812134,
+      "learning_rate": 1.4771335712280054e-05,
+      "loss": 0.37,
       "step": 89535
     },
     {
-      "epoch": 3.15,
-      "learning_rate": 1.5914933685961886e-05,
-      "loss": 0.266,
+      "epoch": 3.2270155332107975,
+      "grad_norm": 0.21708033978939056,
+      "learning_rate": 1.4768673076421512e-05,
+      "loss": 0.3611,
       "step": 89540
     },
     {
-      "epoch": 3.15,
-      "learning_rate": 1.5912279784512767e-05,
-      "loss": 0.2782,
+      "epoch": 3.227195732872022,
+      "grad_norm": 0.21462254226207733,
+      "learning_rate": 1.4766010579962808e-05,
+      "loss": 0.392,
       "step": 89545
     },
     {
-      "epoch": 3.15,
-      "learning_rate": 1.5909626001061707e-05,
-      "loss": 0.2647,
+      "epoch": 3.227375932533247,
+      "grad_norm": 0.20276306569576263,
+      "learning_rate": 1.4763348222940243e-05,
+      "loss": 0.3416,
       "step": 89550
     },
     {
-      "epoch": 3.15,
-      "learning_rate": 1.590697233564316e-05,
-      "loss": 0.2916,
+      "epoch": 3.2275561321944712,
+      "grad_norm": 0.2577986419200897,
+      "learning_rate": 1.4760686005390068e-05,
+      "loss": 0.3846,
       "step": 89555
     },
     {
-      "epoch": 3.15,
-      "learning_rate": 1.5904318788291577e-05,
-      "loss": 0.2814,
+      "epoch": 3.227736331855696,
+      "grad_norm": 0.22566761076450348,
+      "learning_rate": 1.4758023927348563e-05,
+      "loss": 0.416,
       "step": 89560
     },
     {
-      "epoch": 3.15,
-      "learning_rate": 1.5901665359041425e-05,
-      "loss": 0.2549,
+      "epoch": 3.2279165315169207,
+      "grad_norm": 0.2527579665184021,
+      "learning_rate": 1.4755361988852002e-05,
+      "loss": 0.4009,
       "step": 89565
     },
     {
-      "epoch": 3.15,
-      "learning_rate": 1.5899012047927155e-05,
-      "loss": 0.2596,
+      "epoch": 3.2280967311781454,
+      "grad_norm": 0.23741693794727325,
+      "learning_rate": 1.4752700189936649e-05,
+      "loss": 0.3604,
       "step": 89570
     },
     {
-      "epoch": 3.15,
-      "learning_rate": 1.5896358854983216e-05,
-      "loss": 0.2437,
+      "epoch": 3.22827693083937,
+      "grad_norm": 0.2663268446922302,
+      "learning_rate": 1.4750038530638769e-05,
+      "loss": 0.3795,
       "step": 89575
     },
     {
-      "epoch": 3.15,
-      "learning_rate": 1.589370578024405e-05,
-      "loss": 0.2533,
+      "epoch": 3.2284571305005945,
+      "grad_norm": 0.2541162371635437,
+      "learning_rate": 1.4747377010994634e-05,
+      "loss": 0.367,
       "step": 89580
     },
     {
-      "epoch": 3.15,
-      "learning_rate": 1.589105282374412e-05,
-      "loss": 0.2313,
+      "epoch": 3.228637330161819,
+      "grad_norm": 0.21635161340236664,
+      "learning_rate": 1.4744715631040495e-05,
+      "loss": 0.3491,
       "step": 89585
     },
     {
-      "epoch": 3.15,
-      "learning_rate": 1.5888399985517863e-05,
-      "loss": 0.2523,
+      "epoch": 3.228817529823044,
+      "grad_norm": 0.2601666748523712,
+      "learning_rate": 1.4742054390812631e-05,
+      "loss": 0.4016,
       "step": 89590
     },
     {
-      "epoch": 3.15,
-      "learning_rate": 1.588574726559973e-05,
-      "loss": 0.2657,
+      "epoch": 3.2289977294842687,
+      "grad_norm": 0.2198321521282196,
+      "learning_rate": 1.4739393290347286e-05,
+      "loss": 0.3752,
       "step": 89595
     },
     {
-      "epoch": 3.15,
-      "learning_rate": 1.5883094664024155e-05,
-      "loss": 0.2559,
+      "epoch": 3.2291779291454934,
+      "grad_norm": 0.22527672350406647,
+      "learning_rate": 1.4736732329680714e-05,
+      "loss": 0.3589,
       "step": 89600
     },
     {
-      "epoch": 3.15,
-      "learning_rate": 1.58804421808256e-05,
-      "loss": 0.2689,
+      "epoch": 3.2293581288067177,
+      "grad_norm": 0.28329557180404663,
+      "learning_rate": 1.4734071508849184e-05,
+      "loss": 0.3664,
       "step": 89605
     },
     {
-      "epoch": 3.15,
-      "learning_rate": 1.587778981603849e-05,
-      "loss": 0.25,
+      "epoch": 3.2295383284679424,
+      "grad_norm": 0.2408505380153656,
+      "learning_rate": 1.4731410827888947e-05,
+      "loss": 0.4055,
       "step": 89610
     },
     {
-      "epoch": 3.15,
-      "learning_rate": 1.587513756969726e-05,
-      "loss": 0.2579,
+      "epoch": 3.229718528129167,
+      "grad_norm": 0.2271721065044403,
+      "learning_rate": 1.4728750286836236e-05,
+      "loss": 0.3716,
       "step": 89615
     },
     {
-      "epoch": 3.15,
-      "learning_rate": 1.5872485441836353e-05,
-      "loss": 0.2435,
+      "epoch": 3.229898727790392,
+      "grad_norm": 0.20692966878414154,
+      "learning_rate": 1.472608988572733e-05,
+      "loss": 0.3708,
       "step": 89620
     },
     {
-      "epoch": 3.15,
-      "learning_rate": 1.586983343249022e-05,
-      "loss": 0.2647,
+      "epoch": 3.2300789274516166,
+      "grad_norm": 0.19603079557418823,
+      "learning_rate": 1.472342962459844e-05,
+      "loss": 0.3608,
       "step": 89625
     },
     {
-      "epoch": 3.15,
-      "learning_rate": 1.5867181541693278e-05,
-      "loss": 0.2701,
+      "epoch": 3.230259127112841,
+      "grad_norm": 0.21330444514751434,
+      "learning_rate": 1.4720769503485845e-05,
+      "loss": 0.3808,
       "step": 89630
     },
     {
-      "epoch": 3.15,
-      "learning_rate": 1.586452976947996e-05,
-      "loss": 0.2476,
+      "epoch": 3.2304393267740656,
+      "grad_norm": 0.22534000873565674,
+      "learning_rate": 1.471810952242578e-05,
+      "loss": 0.4035,
       "step": 89635
     },
     {
-      "epoch": 3.15,
-      "learning_rate": 1.586187811588471e-05,
-      "loss": 0.2859,
+      "epoch": 3.2306195264352904,
+      "grad_norm": 0.22428487241268158,
+      "learning_rate": 1.4715449681454468e-05,
+      "loss": 0.3776,
       "step": 89640
     },
     {
-      "epoch": 3.15,
-      "learning_rate": 1.5859226580941945e-05,
-      "loss": 0.2638,
+      "epoch": 3.230799726096515,
+      "grad_norm": 0.2968350052833557,
+      "learning_rate": 1.4712789980608177e-05,
+      "loss": 0.3924,
       "step": 89645
     },
     {
-      "epoch": 3.15,
-      "learning_rate": 1.5856575164686104e-05,
-      "loss": 0.2614,
+      "epoch": 3.2309799257577394,
+      "grad_norm": 0.2184687852859497,
+      "learning_rate": 1.4710130419923123e-05,
+      "loss": 0.3787,
       "step": 89650
     },
     {
-      "epoch": 3.15,
-      "learning_rate": 1.5853923867151596e-05,
-      "loss": 0.243,
+      "epoch": 3.231160125418964,
+      "grad_norm": 0.23696725070476532,
+      "learning_rate": 1.470747099943555e-05,
+      "loss": 0.3938,
       "step": 89655
     },
     {
-      "epoch": 3.15,
-      "learning_rate": 1.585127268837287e-05,
-      "loss": 0.2492,
+      "epoch": 3.231340325080189,
+      "grad_norm": 0.24137650430202484,
+      "learning_rate": 1.4704811719181693e-05,
+      "loss": 0.4011,
       "step": 89660
     },
     {
-      "epoch": 3.15,
-      "learning_rate": 1.584862162838434e-05,
-      "loss": 0.2499,
+      "epoch": 3.2315205247414136,
+      "grad_norm": 0.2342267483472824,
+      "learning_rate": 1.4702152579197782e-05,
+      "loss": 0.3619,
       "step": 89665
     },
     {
-      "epoch": 3.15,
-      "learning_rate": 1.5845970687220424e-05,
-      "loss": 0.2493,
+      "epoch": 3.2317007244026383,
+      "grad_norm": 0.2602199912071228,
+      "learning_rate": 1.4699493579520052e-05,
+      "loss": 0.4182,
       "step": 89670
     },
     {
-      "epoch": 3.16,
-      "learning_rate": 1.584331986491554e-05,
-      "loss": 0.2587,
+      "epoch": 3.2318809240638626,
+      "grad_norm": 0.2467493712902069,
+      "learning_rate": 1.4696834720184725e-05,
+      "loss": 0.3843,
       "step": 89675
     },
     {
-      "epoch": 3.16,
-      "learning_rate": 1.5840669161504125e-05,
-      "loss": 0.2524,
+      "epoch": 3.2320611237250874,
+      "grad_norm": 0.23336918652057648,
+      "learning_rate": 1.4694176001228027e-05,
+      "loss": 0.3789,
       "step": 89680
     },
     {
-      "epoch": 3.16,
-      "learning_rate": 1.583801857702058e-05,
-      "loss": 0.2829,
+      "epoch": 3.232241323386312,
+      "grad_norm": 0.22967779636383057,
+      "learning_rate": 1.4691517422686191e-05,
+      "loss": 0.3782,
       "step": 89685
     },
     {
-      "epoch": 3.16,
-      "learning_rate": 1.5835368111499328e-05,
-      "loss": 0.2596,
+      "epoch": 3.232421523047537,
+      "grad_norm": 0.19825366139411926,
+      "learning_rate": 1.4688858984595432e-05,
+      "loss": 0.3776,
       "step": 89690
     },
     {
-      "epoch": 3.16,
-      "learning_rate": 1.5832717764974772e-05,
-      "loss": 0.2582,
+      "epoch": 3.232601722708761,
+      "grad_norm": 0.18190155923366547,
+      "learning_rate": 1.4686200686991972e-05,
+      "loss": 0.3398,
       "step": 89695
     },
     {
-      "epoch": 3.16,
-      "learning_rate": 1.5830067537481337e-05,
-      "loss": 0.2688,
+      "epoch": 3.232781922369986,
+      "grad_norm": 0.21826298534870148,
+      "learning_rate": 1.468354252991203e-05,
+      "loss": 0.371,
       "step": 89700
     },
     {
-      "epoch": 3.16,
-      "learning_rate": 1.582741742905344e-05,
-      "loss": 0.243,
+      "epoch": 3.2329621220312106,
+      "grad_norm": 0.2672412097454071,
+      "learning_rate": 1.4680884513391824e-05,
+      "loss": 0.3691,
       "step": 89705
     },
     {
-      "epoch": 3.16,
-      "learning_rate": 1.582476743972548e-05,
-      "loss": 0.2272,
+      "epoch": 3.2331423216924353,
+      "grad_norm": 0.25871455669403076,
+      "learning_rate": 1.4678226637467574e-05,
+      "loss": 0.3965,
       "step": 89710
     },
     {
-      "epoch": 3.16,
-      "learning_rate": 1.582211756953186e-05,
-      "loss": 0.2525,
+      "epoch": 3.23332252135366,
+      "grad_norm": 0.24801315367221832,
+      "learning_rate": 1.467556890217549e-05,
+      "loss": 0.4047,
       "step": 89715
     },
     {
-      "epoch": 3.16,
-      "learning_rate": 1.5819467818507005e-05,
-      "loss": 0.2644,
+      "epoch": 3.2335027210148843,
+      "grad_norm": 0.20735400915145874,
+      "learning_rate": 1.4672911307551768e-05,
+      "loss": 0.3561,
       "step": 89720
     },
     {
-      "epoch": 3.16,
-      "learning_rate": 1.5816818186685306e-05,
-      "loss": 0.2635,
+      "epoch": 3.233682920676109,
+      "grad_norm": 0.24072608351707458,
+      "learning_rate": 1.4670253853632642e-05,
+      "loss": 0.3899,
       "step": 89725
     },
     {
-      "epoch": 3.16,
-      "learning_rate": 1.5814168674101174e-05,
-      "loss": 0.2387,
+      "epoch": 3.233863120337334,
+      "grad_norm": 0.24866987764835358,
+      "learning_rate": 1.4667596540454296e-05,
+      "loss": 0.3894,
       "step": 89730
     },
     {
-      "epoch": 3.16,
-      "learning_rate": 1.5811519280789e-05,
-      "loss": 0.2879,
+      "epoch": 3.2340433199985585,
+      "grad_norm": 0.2773805856704712,
+      "learning_rate": 1.466493936805296e-05,
+      "loss": 0.4052,
       "step": 89735
     },
     {
-      "epoch": 3.16,
-      "learning_rate": 1.58088700067832e-05,
-      "loss": 0.2594,
+      "epoch": 3.234223519659783,
+      "grad_norm": 0.2559952735900879,
+      "learning_rate": 1.4662282336464817e-05,
+      "loss": 0.3995,
       "step": 89740
     },
     {
-      "epoch": 3.16,
-      "learning_rate": 1.5806220852118166e-05,
-      "loss": 0.2604,
+      "epoch": 3.2344037193210076,
+      "grad_norm": 0.20046544075012207,
+      "learning_rate": 1.4659625445726068e-05,
+      "loss": 0.388,
       "step": 89745
     },
     {
-      "epoch": 3.16,
-      "learning_rate": 1.5803571816828287e-05,
-      "loss": 0.2514,
+      "epoch": 3.2345839189822323,
+      "grad_norm": 0.24034267663955688,
+      "learning_rate": 1.4656968695872936e-05,
+      "loss": 0.3999,
       "step": 89750
     },
     {
-      "epoch": 3.16,
-      "learning_rate": 1.5800922900947973e-05,
-      "loss": 0.2494,
+      "epoch": 3.234764118643457,
+      "grad_norm": 0.20936627686023712,
+      "learning_rate": 1.4654312086941597e-05,
+      "loss": 0.3932,
       "step": 89755
     },
     {
-      "epoch": 3.16,
-      "learning_rate": 1.5798274104511614e-05,
-      "loss": 0.2555,
+      "epoch": 3.2349443183046818,
+      "grad_norm": 0.25857529044151306,
+      "learning_rate": 1.4651655618968246e-05,
+      "loss": 0.3954,
       "step": 89760
     },
     {
-      "epoch": 3.16,
-      "learning_rate": 1.5795625427553606e-05,
-      "loss": 0.2639,
+      "epoch": 3.235124517965906,
+      "grad_norm": 0.20144201815128326,
+      "learning_rate": 1.4648999291989093e-05,
+      "loss": 0.3765,
       "step": 89765
     },
     {
-      "epoch": 3.16,
-      "learning_rate": 1.5792976870108324e-05,
-      "loss": 0.2454,
+      "epoch": 3.235304717627131,
+      "grad_norm": 0.25030654668807983,
+      "learning_rate": 1.4646343106040314e-05,
+      "loss": 0.3872,
       "step": 89770
     },
     {
-      "epoch": 3.16,
-      "learning_rate": 1.5790328432210176e-05,
-      "loss": 0.253,
+      "epoch": 3.2354849172883555,
+      "grad_norm": 0.18118837475776672,
+      "learning_rate": 1.464368706115811e-05,
+      "loss": 0.3797,
       "step": 89775
     },
     {
-      "epoch": 3.16,
-      "learning_rate": 1.5787680113893544e-05,
-      "loss": 0.2621,
+      "epoch": 3.2356651169495803,
+      "grad_norm": 0.2649378180503845,
+      "learning_rate": 1.4641031157378665e-05,
+      "loss": 0.3662,
       "step": 89780
     },
     {
-      "epoch": 3.16,
-      "learning_rate": 1.5785031915192818e-05,
-      "loss": 0.2478,
+      "epoch": 3.2358453166108045,
+      "grad_norm": 0.22662107646465302,
+      "learning_rate": 1.4638375394738157e-05,
+      "loss": 0.3622,
       "step": 89785
     },
     {
-      "epoch": 3.16,
-      "learning_rate": 1.578238383614237e-05,
-      "loss": 0.2364,
+      "epoch": 3.2360255162720293,
+      "grad_norm": 0.2865629494190216,
+      "learning_rate": 1.4635719773272788e-05,
+      "loss": 0.4174,
       "step": 89790
     },
     {
-      "epoch": 3.16,
-      "learning_rate": 1.5779735876776604e-05,
-      "loss": 0.2658,
+      "epoch": 3.236205715933254,
+      "grad_norm": 0.2678220868110657,
+      "learning_rate": 1.4633064293018728e-05,
+      "loss": 0.3557,
       "step": 89795
     },
     {
-      "epoch": 3.16,
-      "learning_rate": 1.5777088037129888e-05,
-      "loss": 0.2406,
+      "epoch": 3.2363859155944787,
+      "grad_norm": 0.2645963728427887,
+      "learning_rate": 1.4630408954012153e-05,
+      "loss": 0.3925,
       "step": 89800
     },
     {
-      "epoch": 3.16,
-      "learning_rate": 1.5774440317236606e-05,
-      "loss": 0.2387,
+      "epoch": 3.2365661152557035,
+      "grad_norm": 0.21265912055969238,
+      "learning_rate": 1.4627753756289258e-05,
+      "loss": 0.349,
       "step": 89805
     },
     {
-      "epoch": 3.16,
-      "learning_rate": 1.5771792717131133e-05,
-      "loss": 0.2783,
+      "epoch": 3.2367463149169278,
+      "grad_norm": 0.23955461382865906,
+      "learning_rate": 1.4625098699886204e-05,
+      "loss": 0.42,
       "step": 89810
     },
     {
-      "epoch": 3.16,
-      "learning_rate": 1.576914523684785e-05,
-      "loss": 0.2633,
+      "epoch": 3.2369265145781525,
+      "grad_norm": 0.28037208318710327,
+      "learning_rate": 1.4622443784839174e-05,
+      "loss": 0.3545,
       "step": 89815
     },
     {
-      "epoch": 3.16,
-      "learning_rate": 1.576649787642114e-05,
-      "loss": 0.2534,
+      "epoch": 3.2371067142393772,
+      "grad_norm": 0.20563095808029175,
+      "learning_rate": 1.4619789011184343e-05,
+      "loss": 0.3833,
       "step": 89820
     },
     {
-      "epoch": 3.16,
-      "learning_rate": 1.5763850635885367e-05,
-      "loss": 0.2456,
+      "epoch": 3.237286913900602,
+      "grad_norm": 0.2627568542957306,
+      "learning_rate": 1.4617134378957864e-05,
+      "loss": 0.3868,
       "step": 89825
     },
     {
-      "epoch": 3.16,
-      "learning_rate": 1.5761203515274898e-05,
-      "loss": 0.2555,
+      "epoch": 3.2374671135618267,
+      "grad_norm": 0.23573757708072662,
+      "learning_rate": 1.4614479888195931e-05,
+      "loss": 0.3716,
       "step": 89830
     },
     {
-      "epoch": 3.16,
-      "learning_rate": 1.5758556514624118e-05,
-      "loss": 0.2793,
+      "epoch": 3.237647313223051,
+      "grad_norm": 0.22677142918109894,
+      "learning_rate": 1.4611825538934699e-05,
+      "loss": 0.39,
       "step": 89835
     },
     {
-      "epoch": 3.16,
-      "learning_rate": 1.575590963396739e-05,
-      "loss": 0.259,
+      "epoch": 3.2378275128842757,
+      "grad_norm": 0.23571011424064636,
+      "learning_rate": 1.4609171331210324e-05,
+      "loss": 0.383,
       "step": 89840
     },
     {
-      "epoch": 3.16,
-      "learning_rate": 1.5753262873339093e-05,
-      "loss": 0.2493,
+      "epoch": 3.2380077125455005,
+      "grad_norm": 0.2776787281036377,
+      "learning_rate": 1.4606517265058982e-05,
+      "loss": 0.3464,
       "step": 89845
     },
     {
-      "epoch": 3.16,
-      "learning_rate": 1.575061623277357e-05,
-      "loss": 0.2773,
+      "epoch": 3.238187912206725,
+      "grad_norm": 0.21779267489910126,
+      "learning_rate": 1.4603863340516826e-05,
+      "loss": 0.3814,
       "step": 89850
     },
     {
-      "epoch": 3.16,
-      "learning_rate": 1.574796971230521e-05,
-      "loss": 0.2326,
+      "epoch": 3.2383681118679495,
+      "grad_norm": 0.2640026807785034,
+      "learning_rate": 1.4601209557620027e-05,
+      "loss": 0.3906,
       "step": 89855
     },
     {
-      "epoch": 3.16,
-      "learning_rate": 1.574532331196836e-05,
-      "loss": 0.2391,
+      "epoch": 3.238548311529174,
+      "grad_norm": 0.2678093910217285,
+      "learning_rate": 1.4598555916404732e-05,
+      "loss": 0.3533,
       "step": 89860
     },
     {
-      "epoch": 3.16,
-      "learning_rate": 1.5742677031797388e-05,
-      "loss": 0.2515,
+      "epoch": 3.238728511190399,
+      "grad_norm": 0.24538059532642365,
+      "learning_rate": 1.4595902416907092e-05,
+      "loss": 0.3712,
       "step": 89865
     },
     {
-      "epoch": 3.16,
-      "learning_rate": 1.5740030871826656e-05,
-      "loss": 0.2605,
+      "epoch": 3.2389087108516237,
+      "grad_norm": 0.27062126994132996,
+      "learning_rate": 1.4593249059163283e-05,
+      "loss": 0.3648,
       "step": 89870
     },
     {
-      "epoch": 3.16,
-      "learning_rate": 1.5737384832090523e-05,
-      "loss": 0.2575,
+      "epoch": 3.2390889105128484,
+      "grad_norm": 0.24212132394313812,
+      "learning_rate": 1.4590595843209432e-05,
+      "loss": 0.3855,
       "step": 89875
     },
     {
-      "epoch": 3.16,
-      "learning_rate": 1.573473891262334e-05,
-      "loss": 0.2595,
+      "epoch": 3.2392691101740727,
+      "grad_norm": 0.2248118668794632,
+      "learning_rate": 1.4587942769081689e-05,
+      "loss": 0.347,
       "step": 89880
     },
     {
-      "epoch": 3.16,
-      "learning_rate": 1.5732093113459466e-05,
-      "loss": 0.242,
+      "epoch": 3.2394493098352974,
+      "grad_norm": 0.21070916950702667,
+      "learning_rate": 1.4585289836816218e-05,
+      "loss": 0.3758,
       "step": 89885
     },
     {
-      "epoch": 3.16,
-      "learning_rate": 1.572944743463326e-05,
-      "loss": 0.2755,
+      "epoch": 3.239629509496522,
+      "grad_norm": 0.19322839379310608,
+      "learning_rate": 1.4582637046449148e-05,
+      "loss": 0.3913,
       "step": 89890
     },
     {
-      "epoch": 3.16,
-      "learning_rate": 1.5726801876179066e-05,
-      "loss": 0.2453,
+      "epoch": 3.239809709157747,
+      "grad_norm": 0.2393057942390442,
+      "learning_rate": 1.4579984398016649e-05,
+      "loss": 0.3938,
       "step": 89895
     },
     {
-      "epoch": 3.16,
-      "learning_rate": 1.572415643813124e-05,
-      "loss": 0.2467,
+      "epoch": 3.2399899088189716,
+      "grad_norm": 0.24970926344394684,
+      "learning_rate": 1.4577331891554839e-05,
+      "loss": 0.3932,
       "step": 89900
     },
     {
-      "epoch": 3.16,
-      "learning_rate": 1.5721511120524117e-05,
-      "loss": 0.2782,
+      "epoch": 3.240170108480196,
+      "grad_norm": 0.23957009613513947,
+      "learning_rate": 1.4574679527099852e-05,
+      "loss": 0.3849,
       "step": 89905
     },
     {
-      "epoch": 3.16,
-      "learning_rate": 1.5718865923392073e-05,
-      "loss": 0.2259,
+      "epoch": 3.2403503081414207,
+      "grad_norm": 0.22815951704978943,
+      "learning_rate": 1.4572027304687847e-05,
+      "loss": 0.3937,
       "step": 89910
     },
     {
-      "epoch": 3.16,
-      "learning_rate": 1.5716220846769434e-05,
-      "loss": 0.2757,
+      "epoch": 3.2405305078026454,
+      "grad_norm": 0.20002481341362,
+      "learning_rate": 1.4569375224354953e-05,
+      "loss": 0.3943,
       "step": 89915
     },
     {
-      "epoch": 3.16,
-      "learning_rate": 1.5713575890690553e-05,
-      "loss": 0.2559,
+      "epoch": 3.24071070746387,
+      "grad_norm": 0.2669081687927246,
+      "learning_rate": 1.4566723286137301e-05,
+      "loss": 0.4006,
       "step": 89920
     },
     {
-      "epoch": 3.16,
-      "learning_rate": 1.571093105518976e-05,
-      "loss": 0.2471,
+      "epoch": 3.2408909071250944,
+      "grad_norm": 0.2839129567146301,
+      "learning_rate": 1.4564071490071024e-05,
+      "loss": 0.3667,
       "step": 89925
     },
     {
-      "epoch": 3.16,
-      "learning_rate": 1.570828634030141e-05,
-      "loss": 0.2502,
+      "epoch": 3.241071106786319,
+      "grad_norm": 0.1999591588973999,
+      "learning_rate": 1.4561419836192241e-05,
+      "loss": 0.3618,
       "step": 89930
     },
     {
-      "epoch": 3.16,
-      "learning_rate": 1.5705641746059843e-05,
-      "loss": 0.2415,
+      "epoch": 3.241251306447544,
+      "grad_norm": 0.21648964285850525,
+      "learning_rate": 1.4558768324537105e-05,
+      "loss": 0.3854,
       "step": 89935
     },
     {
-      "epoch": 3.16,
-      "learning_rate": 1.5702997272499387e-05,
-      "loss": 0.2449,
+      "epoch": 3.2414315061087686,
+      "grad_norm": 0.17775964736938477,
+      "learning_rate": 1.4556116955141729e-05,
+      "loss": 0.3296,
       "step": 89940
     },
     {
-      "epoch": 3.16,
-      "learning_rate": 1.5700352919654386e-05,
-      "loss": 0.2644,
+      "epoch": 3.2416117057699934,
+      "grad_norm": 0.24124689400196075,
+      "learning_rate": 1.4553465728042236e-05,
+      "loss": 0.3892,
       "step": 89945
     },
     {
-      "epoch": 3.16,
-      "learning_rate": 1.569770868755917e-05,
-      "loss": 0.2435,
+      "epoch": 3.2417919054312176,
+      "grad_norm": 0.19329720735549927,
+      "learning_rate": 1.4550814643274752e-05,
+      "loss": 0.3676,
       "step": 89950
     },
     {
-      "epoch": 3.16,
-      "learning_rate": 1.5695064576248088e-05,
-      "loss": 0.2489,
+      "epoch": 3.2419721050924424,
+      "grad_norm": 0.1987384855747223,
+      "learning_rate": 1.4548163700875384e-05,
+      "loss": 0.3755,
       "step": 89955
     },
     {
-      "epoch": 3.17,
-      "learning_rate": 1.569242058575545e-05,
-      "loss": 0.2684,
+      "epoch": 3.242152304753667,
+      "grad_norm": 0.2366037666797638,
+      "learning_rate": 1.4545512900880271e-05,
+      "loss": 0.3561,
       "step": 89960
     },
     {
-      "epoch": 3.17,
-      "learning_rate": 1.5689776716115594e-05,
-      "loss": 0.2492,
+      "epoch": 3.242332504414892,
+      "grad_norm": 0.2112271934747696,
+      "learning_rate": 1.454286224332553e-05,
+      "loss": 0.381,
       "step": 89965
     },
     {
-      "epoch": 3.17,
-      "learning_rate": 1.568713296736286e-05,
-      "loss": 0.266,
+      "epoch": 3.242512704076116,
+      "grad_norm": 0.2448207587003708,
+      "learning_rate": 1.454021172824725e-05,
+      "loss": 0.4151,
       "step": 89970
     },
     {
-      "epoch": 3.17,
-      "learning_rate": 1.5684489339531563e-05,
-      "loss": 0.2666,
+      "epoch": 3.242692903737341,
+      "grad_norm": 0.24623465538024902,
+      "learning_rate": 1.4537561355681567e-05,
+      "loss": 0.369,
       "step": 89975
     },
     {
-      "epoch": 3.17,
-      "learning_rate": 1.568184583265603e-05,
-      "loss": 0.2632,
+      "epoch": 3.2428731033985656,
+      "grad_norm": 0.2527907192707062,
+      "learning_rate": 1.4534911125664585e-05,
+      "loss": 0.4135,
       "step": 89980
     },
     {
-      "epoch": 3.17,
-      "learning_rate": 1.567920244677059e-05,
-      "loss": 0.2733,
+      "epoch": 3.2430533030597903,
+      "grad_norm": 0.21829581260681152,
+      "learning_rate": 1.4532261038232414e-05,
+      "loss": 0.3546,
       "step": 89985
     },
     {
-      "epoch": 3.17,
-      "learning_rate": 1.567655918190957e-05,
-      "loss": 0.2579,
+      "epoch": 3.243233502721015,
+      "grad_norm": 0.196324422955513,
+      "learning_rate": 1.4529611093421158e-05,
+      "loss": 0.4074,
       "step": 89990
     },
     {
-      "epoch": 3.17,
-      "learning_rate": 1.5673916038107283e-05,
-      "loss": 0.2433,
+      "epoch": 3.2434137023822394,
+      "grad_norm": 0.22843465209007263,
+      "learning_rate": 1.4526961291266916e-05,
+      "loss": 0.3604,
       "step": 89995
     },
     {
-      "epoch": 3.17,
-      "learning_rate": 1.5671273015398042e-05,
-      "loss": 0.2469,
+      "epoch": 3.243593902043464,
+      "grad_norm": 0.19749246537685394,
+      "learning_rate": 1.452431163180581e-05,
+      "loss": 0.3636,
       "step": 90000
     },
     {
-      "epoch": 3.17,
-      "eval_loss": 0.2552323043346405,
-      "eval_runtime": 10.5438,
-      "eval_samples_per_second": 9.484,
-      "eval_steps_per_second": 9.484,
+      "epoch": 3.243593902043464,
+      "eval_loss": 0.4301738440990448,
+      "eval_runtime": 3.5334,
+      "eval_samples_per_second": 28.301,
+      "eval_steps_per_second": 7.075,
       "step": 90000
     },
     {
-      "epoch": 3.17,
-      "learning_rate": 1.5668630113816173e-05,
-      "loss": 0.2563,
+      "epoch": 3.243774101704689,
+      "grad_norm": 0.2458840310573578,
+      "learning_rate": 1.4521662115073925e-05,
+      "loss": 0.3655,
       "step": 90005
     },
     {
-      "epoch": 3.17,
-      "learning_rate": 1.5665987333396003e-05,
-      "loss": 0.256,
+      "epoch": 3.2439543013659136,
+      "grad_norm": 0.23517417907714844,
+      "learning_rate": 1.4519012741107368e-05,
+      "loss": 0.3811,
       "step": 90010
     },
     {
-      "epoch": 3.17,
-      "learning_rate": 1.5663344674171832e-05,
-      "loss": 0.2445,
+      "epoch": 3.244134501027138,
+      "grad_norm": 0.21552741527557373,
+      "learning_rate": 1.4516363509942233e-05,
+      "loss": 0.3699,
       "step": 90015
     },
     {
-      "epoch": 3.17,
-      "learning_rate": 1.5660702136177968e-05,
-      "loss": 0.253,
+      "epoch": 3.2443147006883626,
+      "grad_norm": 0.2056073546409607,
+      "learning_rate": 1.4513714421614617e-05,
+      "loss": 0.3778,
       "step": 90020
     },
     {
-      "epoch": 3.17,
-      "learning_rate": 1.5658059719448745e-05,
-      "loss": 0.2588,
+      "epoch": 3.2444949003495873,
+      "grad_norm": 0.219630628824234,
+      "learning_rate": 1.4511065476160598e-05,
+      "loss": 0.3692,
       "step": 90025
     },
     {
-      "epoch": 3.17,
-      "learning_rate": 1.5655417424018455e-05,
-      "loss": 0.2653,
+      "epoch": 3.244675100010812,
+      "grad_norm": 0.2458181381225586,
+      "learning_rate": 1.4508416673616292e-05,
+      "loss": 0.3989,
       "step": 90030
     },
     {
-      "epoch": 3.17,
-      "learning_rate": 1.5652775249921413e-05,
-      "loss": 0.2451,
+      "epoch": 3.244855299672037,
+      "grad_norm": 0.23041844367980957,
+      "learning_rate": 1.4505768014017782e-05,
+      "loss": 0.3763,
       "step": 90035
     },
     {
-      "epoch": 3.17,
-      "learning_rate": 1.5650133197191914e-05,
-      "loss": 0.2793,
+      "epoch": 3.245035499333261,
+      "grad_norm": 0.2688606083393097,
+      "learning_rate": 1.4503119497401147e-05,
+      "loss": 0.3814,
       "step": 90040
     },
     {
-      "epoch": 3.17,
-      "learning_rate": 1.5647491265864287e-05,
-      "loss": 0.2702,
+      "epoch": 3.245215698994486,
+      "grad_norm": 0.20905734598636627,
+      "learning_rate": 1.4500471123802478e-05,
+      "loss": 0.3603,
       "step": 90045
     },
     {
-      "epoch": 3.17,
-      "learning_rate": 1.564484945597282e-05,
-      "loss": 0.2383,
+      "epoch": 3.2453958986557105,
+      "grad_norm": 0.2165406346321106,
+      "learning_rate": 1.4497822893257845e-05,
+      "loss": 0.3778,
       "step": 90050
     },
     {
-      "epoch": 3.17,
-      "learning_rate": 1.5642207767551814e-05,
-      "loss": 0.2614,
+      "epoch": 3.2455760983169353,
+      "grad_norm": 0.28338685631752014,
+      "learning_rate": 1.4495174805803358e-05,
+      "loss": 0.3849,
       "step": 90055
     },
     {
-      "epoch": 3.17,
-      "learning_rate": 1.5639566200635565e-05,
-      "loss": 0.2682,
+      "epoch": 3.2457562979781596,
+      "grad_norm": 0.24347181618213654,
+      "learning_rate": 1.4492526861475084e-05,
+      "loss": 0.3794,
       "step": 90060
     },
     {
-      "epoch": 3.17,
-      "learning_rate": 1.5636924755258396e-05,
-      "loss": 0.2693,
+      "epoch": 3.2459364976393843,
+      "grad_norm": 0.17866498231887817,
+      "learning_rate": 1.448987906030908e-05,
+      "loss": 0.3434,
       "step": 90065
     },
     {
-      "epoch": 3.17,
-      "learning_rate": 1.5634283431454582e-05,
-      "loss": 0.2385,
+      "epoch": 3.246116697300609,
+      "grad_norm": 0.23330989480018616,
+      "learning_rate": 1.4487231402341447e-05,
+      "loss": 0.377,
       "step": 90070
     },
     {
-      "epoch": 3.17,
-      "learning_rate": 1.563164222925843e-05,
-      "loss": 0.2838,
+      "epoch": 3.2462968969618338,
+      "grad_norm": 0.23260824382305145,
+      "learning_rate": 1.4484583887608243e-05,
+      "loss": 0.3774,
       "step": 90075
     },
     {
-      "epoch": 3.17,
-      "learning_rate": 1.5629001148704218e-05,
-      "loss": 0.263,
+      "epoch": 3.2464770966230585,
+      "grad_norm": 0.20770181715488434,
+      "learning_rate": 1.4481936516145567e-05,
+      "loss": 0.3724,
       "step": 90080
     },
     {
-      "epoch": 3.17,
-      "learning_rate": 1.5626360189826257e-05,
-      "loss": 0.2606,
+      "epoch": 3.246657296284283,
+      "grad_norm": 0.20958863198757172,
+      "learning_rate": 1.447928928798946e-05,
+      "loss": 0.3703,
       "step": 90085
     },
     {
-      "epoch": 3.17,
-      "learning_rate": 1.5623719352658833e-05,
-      "loss": 0.2521,
+      "epoch": 3.2468374959455075,
+      "grad_norm": 0.2624190151691437,
+      "learning_rate": 1.4476642203175988e-05,
+      "loss": 0.3906,
       "step": 90090
     },
     {
-      "epoch": 3.17,
-      "learning_rate": 1.5621078637236224e-05,
-      "loss": 0.239,
+      "epoch": 3.2470176956067323,
+      "grad_norm": 0.27643921971321106,
+      "learning_rate": 1.4473995261741238e-05,
+      "loss": 0.36,
       "step": 90095
     },
     {
-      "epoch": 3.17,
-      "learning_rate": 1.5618438043592744e-05,
-      "loss": 0.2759,
+      "epoch": 3.247197895267957,
+      "grad_norm": 0.24671484529972076,
+      "learning_rate": 1.4471348463721268e-05,
+      "loss": 0.3819,
       "step": 90100
     },
     {
-      "epoch": 3.17,
-      "learning_rate": 1.5615797571762658e-05,
-      "loss": 0.2607,
+      "epoch": 3.2473780949291817,
+      "grad_norm": 0.22529368102550507,
+      "learning_rate": 1.4468701809152135e-05,
+      "loss": 0.352,
       "step": 90105
     },
     {
-      "epoch": 3.17,
-      "learning_rate": 1.561315722178025e-05,
-      "loss": 0.2598,
+      "epoch": 3.247558294590406,
+      "grad_norm": 0.22821401059627533,
+      "learning_rate": 1.44660552980699e-05,
+      "loss": 0.3884,
       "step": 90110
     },
     {
-      "epoch": 3.17,
-      "learning_rate": 1.561051699367981e-05,
-      "loss": 0.2588,
+      "epoch": 3.2477384942516307,
+      "grad_norm": 0.22324234247207642,
+      "learning_rate": 1.4463408930510614e-05,
+      "loss": 0.3781,
       "step": 90115
     },
     {
-      "epoch": 3.17,
-      "learning_rate": 1.5607876887495628e-05,
-      "loss": 0.2403,
+      "epoch": 3.2479186939128555,
+      "grad_norm": 0.25382617115974426,
+      "learning_rate": 1.446076270651035e-05,
+      "loss": 0.3944,
       "step": 90120
     },
     {
-      "epoch": 3.17,
-      "learning_rate": 1.560523690326197e-05,
-      "loss": 0.232,
+      "epoch": 3.24809889357408,
+      "grad_norm": 0.2712723910808563,
+      "learning_rate": 1.4458116626105154e-05,
+      "loss": 0.3762,
       "step": 90125
     },
     {
-      "epoch": 3.17,
-      "learning_rate": 1.5602597041013122e-05,
-      "loss": 0.2458,
+      "epoch": 3.248279093235305,
+      "grad_norm": 0.2119145393371582,
+      "learning_rate": 1.4455470689331078e-05,
+      "loss": 0.3578,
       "step": 90130
     },
     {
-      "epoch": 3.17,
-      "learning_rate": 1.5599957300783346e-05,
-      "loss": 0.2349,
+      "epoch": 3.2484592928965292,
+      "grad_norm": 0.2648928165435791,
+      "learning_rate": 1.4452824896224171e-05,
+      "loss": 0.3913,
       "step": 90135
     },
     {
-      "epoch": 3.17,
-      "learning_rate": 1.559731768260694e-05,
-      "loss": 0.2434,
+      "epoch": 3.248639492557754,
+      "grad_norm": 0.21984988451004028,
+      "learning_rate": 1.4450179246820475e-05,
+      "loss": 0.3591,
       "step": 90140
     },
     {
-      "epoch": 3.17,
-      "learning_rate": 1.5594678186518168e-05,
-      "loss": 0.2512,
+      "epoch": 3.2488196922189787,
+      "grad_norm": 0.25489935278892517,
+      "learning_rate": 1.4447533741156055e-05,
+      "loss": 0.3913,
       "step": 90145
     },
     {
-      "epoch": 3.17,
-      "learning_rate": 1.5592038812551303e-05,
-      "loss": 0.2682,
+      "epoch": 3.2489998918802034,
+      "grad_norm": 0.19354936480522156,
+      "learning_rate": 1.4444888379266952e-05,
+      "loss": 0.3815,
       "step": 90150
     },
     {
-      "epoch": 3.17,
-      "learning_rate": 1.5589399560740603e-05,
-      "loss": 0.2573,
+      "epoch": 3.2491800915414277,
+      "grad_norm": 0.2649206519126892,
+      "learning_rate": 1.4442243161189184e-05,
+      "loss": 0.3611,
       "step": 90155
     },
     {
-      "epoch": 3.17,
-      "learning_rate": 1.5586760431120358e-05,
-      "loss": 0.2615,
+      "epoch": 3.2493602912026525,
+      "grad_norm": 0.23419325053691864,
+      "learning_rate": 1.4439598086958816e-05,
+      "loss": 0.3869,
       "step": 90160
     },
     {
-      "epoch": 3.17,
-      "learning_rate": 1.558412142372482e-05,
-      "loss": 0.2723,
+      "epoch": 3.249540490863877,
+      "grad_norm": 0.2443428486585617,
+      "learning_rate": 1.4436953156611881e-05,
+      "loss": 0.4055,
       "step": 90165
     },
     {
-      "epoch": 3.17,
-      "learning_rate": 1.5581482538588265e-05,
-      "loss": 0.2662,
+      "epoch": 3.249720690525102,
+      "grad_norm": 0.18674413859844208,
+      "learning_rate": 1.4434308370184413e-05,
+      "loss": 0.3629,
       "step": 90170
     },
     {
-      "epoch": 3.17,
-      "learning_rate": 1.5578843775744945e-05,
-      "loss": 0.2276,
+      "epoch": 3.2499008901863267,
+      "grad_norm": 0.2267744094133377,
+      "learning_rate": 1.443166372771245e-05,
+      "loss": 0.3401,
       "step": 90175
     },
     {
-      "epoch": 3.17,
-      "learning_rate": 1.5576205135229137e-05,
-      "loss": 0.2404,
+      "epoch": 3.250081089847551,
+      "grad_norm": 0.1851803958415985,
+      "learning_rate": 1.442901922923201e-05,
+      "loss": 0.3667,
       "step": 90180
     },
     {
-      "epoch": 3.17,
-      "learning_rate": 1.5573566617075096e-05,
-      "loss": 0.2428,
+      "epoch": 3.2502612895087757,
+      "grad_norm": 0.2092495560646057,
+      "learning_rate": 1.4426374874779146e-05,
+      "loss": 0.4123,
       "step": 90185
     },
     {
-      "epoch": 3.17,
-      "learning_rate": 1.5570928221317075e-05,
-      "loss": 0.2637,
+      "epoch": 3.2504414891700004,
+      "grad_norm": 0.20582100749015808,
+      "learning_rate": 1.4423730664389883e-05,
+      "loss": 0.4061,
       "step": 90190
     },
     {
-      "epoch": 3.17,
-      "learning_rate": 1.556828994798933e-05,
-      "loss": 0.276,
+      "epoch": 3.250621688831225,
+      "grad_norm": 0.23701712489128113,
+      "learning_rate": 1.4421086598100237e-05,
+      "loss": 0.3766,
       "step": 90195
     },
     {
-      "epoch": 3.17,
-      "learning_rate": 1.5565651797126137e-05,
-      "loss": 0.2541,
+      "epoch": 3.2508018884924494,
+      "grad_norm": 0.1818612813949585,
+      "learning_rate": 1.441844267594624e-05,
+      "loss": 0.3682,
       "step": 90200
     },
     {
-      "epoch": 3.17,
-      "learning_rate": 1.5563013768761735e-05,
-      "loss": 0.2626,
+      "epoch": 3.250982088153674,
+      "grad_norm": 0.23264910280704498,
+      "learning_rate": 1.4415798897963911e-05,
+      "loss": 0.3912,
       "step": 90205
     },
     {
-      "epoch": 3.17,
-      "learning_rate": 1.5560375862930376e-05,
-      "loss": 0.2614,
+      "epoch": 3.251162287814899,
+      "grad_norm": 0.23024223744869232,
+      "learning_rate": 1.4413155264189266e-05,
+      "loss": 0.3769,
       "step": 90210
     },
     {
-      "epoch": 3.17,
-      "learning_rate": 1.555773807966632e-05,
-      "loss": 0.2798,
+      "epoch": 3.2513424874761236,
+      "grad_norm": 0.1898718625307083,
+      "learning_rate": 1.4410511774658336e-05,
+      "loss": 0.3937,
       "step": 90215
     },
     {
-      "epoch": 3.17,
-      "learning_rate": 1.555510041900381e-05,
-      "loss": 0.2506,
+      "epoch": 3.2515226871373484,
+      "grad_norm": 0.2725496292114258,
+      "learning_rate": 1.4407868429407138e-05,
+      "loss": 0.4097,
       "step": 90220
     },
     {
-      "epoch": 3.17,
-      "learning_rate": 1.5552462880977098e-05,
-      "loss": 0.2327,
+      "epoch": 3.2517028867985727,
+      "grad_norm": 0.2061624974012375,
+      "learning_rate": 1.4405225228471678e-05,
+      "loss": 0.3782,
       "step": 90225
     },
     {
-      "epoch": 3.17,
-      "learning_rate": 1.554982546562042e-05,
-      "loss": 0.2532,
+      "epoch": 3.2518830864597974,
+      "grad_norm": 0.240001380443573,
+      "learning_rate": 1.4402582171887973e-05,
+      "loss": 0.3728,
       "step": 90230
     },
     {
-      "epoch": 3.17,
-      "learning_rate": 1.5547188172968048e-05,
-      "loss": 0.2477,
+      "epoch": 3.252063286121022,
+      "grad_norm": 0.22183412313461304,
+      "learning_rate": 1.4399939259692027e-05,
+      "loss": 0.3642,
       "step": 90235
     },
     {
-      "epoch": 3.17,
-      "learning_rate": 1.55445510030542e-05,
-      "loss": 0.2468,
+      "epoch": 3.252243485782247,
+      "grad_norm": 0.2705990672111511,
+      "learning_rate": 1.4397296491919865e-05,
+      "loss": 0.3807,
       "step": 90240
     },
     {
-      "epoch": 3.18,
-      "learning_rate": 1.5541913955913124e-05,
-      "loss": 0.2551,
+      "epoch": 3.252423685443471,
+      "grad_norm": 0.2311464250087738,
+      "learning_rate": 1.4394653868607497e-05,
+      "loss": 0.3711,
       "step": 90245
     },
     {
-      "epoch": 3.18,
-      "learning_rate": 1.5539277031579058e-05,
-      "loss": 0.263,
+      "epoch": 3.252603885104696,
+      "grad_norm": 0.226415753364563,
+      "learning_rate": 1.4392011389790893e-05,
+      "loss": 0.4024,
       "step": 90250
     },
     {
-      "epoch": 3.18,
-      "learning_rate": 1.553664023008625e-05,
-      "loss": 0.2437,
+      "epoch": 3.2527840847659206,
+      "grad_norm": 0.24060435593128204,
+      "learning_rate": 1.4389369055506094e-05,
+      "loss": 0.4227,
       "step": 90255
     },
     {
-      "epoch": 3.18,
-      "learning_rate": 1.5534003551468934e-05,
-      "loss": 0.2818,
+      "epoch": 3.2529642844271454,
+      "grad_norm": 0.2353263795375824,
+      "learning_rate": 1.4386726865789077e-05,
+      "loss": 0.3729,
       "step": 90260
     },
     {
-      "epoch": 3.18,
-      "learning_rate": 1.5531366995761347e-05,
-      "loss": 0.2727,
+      "epoch": 3.25314448408837,
+      "grad_norm": 0.23941341042518616,
+      "learning_rate": 1.438408482067587e-05,
+      "loss": 0.3689,
       "step": 90265
     },
     {
-      "epoch": 3.18,
-      "learning_rate": 1.5528730562997705e-05,
-      "loss": 0.2619,
+      "epoch": 3.2533246837495944,
+      "grad_norm": 0.24477094411849976,
+      "learning_rate": 1.4381442920202442e-05,
+      "loss": 0.4205,
       "step": 90270
     },
     {
-      "epoch": 3.18,
-      "learning_rate": 1.5526094253212264e-05,
-      "loss": 0.2655,
+      "epoch": 3.253504883410819,
+      "grad_norm": 0.21615129709243774,
+      "learning_rate": 1.437880116440479e-05,
+      "loss": 0.3556,
       "step": 90275
     },
     {
-      "epoch": 3.18,
-      "learning_rate": 1.5523458066439248e-05,
-      "loss": 0.2708,
+      "epoch": 3.253685083072044,
+      "grad_norm": 0.2210584431886673,
+      "learning_rate": 1.4376159553318924e-05,
+      "loss": 0.3909,
       "step": 90280
     },
     {
-      "epoch": 3.18,
-      "learning_rate": 1.5520822002712883e-05,
-      "loss": 0.2733,
+      "epoch": 3.2538652827332686,
+      "grad_norm": 0.26296374201774597,
+      "learning_rate": 1.4373518086980827e-05,
+      "loss": 0.3978,
       "step": 90285
     },
     {
-      "epoch": 3.18,
-      "learning_rate": 1.551818606206739e-05,
-      "loss": 0.2635,
+      "epoch": 3.254045482394493,
+      "grad_norm": 0.26196447014808655,
+      "learning_rate": 1.437087676542649e-05,
+      "loss": 0.3911,
       "step": 90290
     },
     {
-      "epoch": 3.18,
-      "learning_rate": 1.551555024453701e-05,
-      "loss": 0.2432,
+      "epoch": 3.2542256820557176,
+      "grad_norm": 0.2408955991268158,
+      "learning_rate": 1.43682355886919e-05,
+      "loss": 0.38,
       "step": 90295
     },
     {
-      "epoch": 3.18,
-      "learning_rate": 1.5512914550155956e-05,
-      "loss": 0.2506,
+      "epoch": 3.2544058817169423,
+      "grad_norm": 0.21061177551746368,
+      "learning_rate": 1.436559455681303e-05,
+      "loss": 0.3891,
       "step": 90300
     },
     {
-      "epoch": 3.18,
-      "learning_rate": 1.551027897895846e-05,
-      "loss": 0.2604,
+      "epoch": 3.254586081378167,
+      "grad_norm": 0.31329545378685,
+      "learning_rate": 1.4362953669825885e-05,
+      "loss": 0.3841,
       "step": 90305
     },
     {
-      "epoch": 3.18,
-      "learning_rate": 1.550764353097873e-05,
-      "loss": 0.2649,
+      "epoch": 3.254766281039392,
+      "grad_norm": 0.22759318351745605,
+      "learning_rate": 1.4360312927766439e-05,
+      "loss": 0.3843,
       "step": 90310
     },
     {
-      "epoch": 3.18,
-      "learning_rate": 1.5505008206250995e-05,
-      "loss": 0.2424,
+      "epoch": 3.254946480700616,
+      "grad_norm": 0.22500398755073547,
+      "learning_rate": 1.4357672330670668e-05,
+      "loss": 0.389,
       "step": 90315
     },
     {
-      "epoch": 3.18,
-      "learning_rate": 1.5502373004809475e-05,
-      "loss": 0.2583,
+      "epoch": 3.255126680361841,
+      "grad_norm": 0.2273656278848648,
+      "learning_rate": 1.435503187857455e-05,
+      "loss": 0.4046,
       "step": 90320
     },
     {
-      "epoch": 3.18,
-      "learning_rate": 1.5499737926688374e-05,
-      "loss": 0.2623,
+      "epoch": 3.2553068800230656,
+      "grad_norm": 0.22215020656585693,
+      "learning_rate": 1.4352391571514053e-05,
+      "loss": 0.4307,
       "step": 90325
     },
     {
-      "epoch": 3.18,
-      "learning_rate": 1.5497102971921924e-05,
-      "loss": 0.2689,
+      "epoch": 3.2554870796842903,
+      "grad_norm": 0.2710254490375519,
+      "learning_rate": 1.434975140952517e-05,
+      "loss": 0.3778,
       "step": 90330
     },
     {
-      "epoch": 3.18,
-      "learning_rate": 1.5494468140544323e-05,
-      "loss": 0.2514,
+      "epoch": 3.2556672793455146,
+      "grad_norm": 0.2155771702528,
+      "learning_rate": 1.4347111392643873e-05,
+      "loss": 0.3671,
       "step": 90335
     },
     {
-      "epoch": 3.18,
-      "learning_rate": 1.5491833432589796e-05,
-      "loss": 0.2469,
+      "epoch": 3.2558474790067393,
+      "grad_norm": 0.2274942249059677,
+      "learning_rate": 1.43444715209061e-05,
+      "loss": 0.4002,
       "step": 90340
     },
     {
-      "epoch": 3.18,
-      "learning_rate": 1.5489198848092533e-05,
-      "loss": 0.24,
+      "epoch": 3.256027678667964,
+      "grad_norm": 0.22359316051006317,
+      "learning_rate": 1.4341831794347849e-05,
+      "loss": 0.3652,
       "step": 90345
     },
     {
-      "epoch": 3.18,
-      "learning_rate": 1.5486564387086766e-05,
-      "loss": 0.2573,
+      "epoch": 3.256207878329189,
+      "grad_norm": 0.239881694316864,
+      "learning_rate": 1.4339192213005076e-05,
+      "loss": 0.3881,
       "step": 90350
     },
     {
-      "epoch": 3.18,
-      "learning_rate": 1.5483930049606686e-05,
-      "loss": 0.2465,
+      "epoch": 3.2563880779904135,
+      "grad_norm": 0.22023212909698486,
+      "learning_rate": 1.4336552776913733e-05,
+      "loss": 0.3707,
       "step": 90355
     },
     {
-      "epoch": 3.18,
-      "learning_rate": 1.548129583568651e-05,
-      "loss": 0.2557,
+      "epoch": 3.2565682776516383,
+      "grad_norm": 0.2799815535545349,
+      "learning_rate": 1.4333913486109817e-05,
+      "loss": 0.3659,
       "step": 90360
     },
     {
-      "epoch": 3.18,
-      "learning_rate": 1.5478661745360423e-05,
-      "loss": 0.258,
+      "epoch": 3.2567484773128625,
+      "grad_norm": 0.19893276691436768,
+      "learning_rate": 1.4331274340629245e-05,
+      "loss": 0.4441,
       "step": 90365
     },
     {
-      "epoch": 3.18,
-      "learning_rate": 1.547602777866265e-05,
-      "loss": 0.2579,
+      "epoch": 3.2569286769740873,
+      "grad_norm": 0.2362791895866394,
+      "learning_rate": 1.4328635340508001e-05,
+      "loss": 0.3203,
       "step": 90370
     },
     {
-      "epoch": 3.18,
-      "learning_rate": 1.5473393935627377e-05,
-      "loss": 0.2479,
+      "epoch": 3.257108876635312,
+      "grad_norm": 0.29334524273872375,
+      "learning_rate": 1.4325996485782039e-05,
+      "loss": 0.3978,
       "step": 90375
     },
     {
-      "epoch": 3.18,
-      "learning_rate": 1.5470760216288804e-05,
-      "loss": 0.2549,
+      "epoch": 3.2572890762965367,
+      "grad_norm": 0.20384852588176727,
+      "learning_rate": 1.4323357776487307e-05,
+      "loss": 0.3841,
       "step": 90380
     },
     {
-      "epoch": 3.18,
-      "learning_rate": 1.5468126620681125e-05,
-      "loss": 0.2813,
+      "epoch": 3.257469275957761,
+      "grad_norm": 0.21409177780151367,
+      "learning_rate": 1.4320719212659756e-05,
+      "loss": 0.3804,
       "step": 90385
     },
     {
-      "epoch": 3.18,
-      "learning_rate": 1.5465493148838546e-05,
-      "loss": 0.2543,
+      "epoch": 3.2576494756189858,
+      "grad_norm": 0.26546648144721985,
+      "learning_rate": 1.431808079433534e-05,
+      "loss": 0.3746,
       "step": 90390
     },
     {
-      "epoch": 3.18,
-      "learning_rate": 1.5462859800795253e-05,
-      "loss": 0.2616,
+      "epoch": 3.2578296752802105,
+      "grad_norm": 0.24117311835289001,
+      "learning_rate": 1.4315442521549993e-05,
+      "loss": 0.3813,
       "step": 90395
     },
     {
-      "epoch": 3.18,
-      "learning_rate": 1.5460226576585447e-05,
-      "loss": 0.2329,
+      "epoch": 3.2580098749414352,
+      "grad_norm": 0.2505226731300354,
+      "learning_rate": 1.4312804394339686e-05,
+      "loss": 0.3831,
       "step": 90400
     },
     {
-      "epoch": 3.18,
-      "learning_rate": 1.5457593476243295e-05,
-      "loss": 0.245,
+      "epoch": 3.25819007460266,
+      "grad_norm": 0.24122489988803864,
+      "learning_rate": 1.431016641274035e-05,
+      "loss": 0.3852,
       "step": 90405
     },
     {
-      "epoch": 3.18,
-      "learning_rate": 1.5454960499803016e-05,
-      "loss": 0.2666,
+      "epoch": 3.2583702742638843,
+      "grad_norm": 0.2857030928134918,
+      "learning_rate": 1.4307528576787927e-05,
+      "loss": 0.3887,
       "step": 90410
     },
     {
-      "epoch": 3.18,
-      "learning_rate": 1.5452327647298782e-05,
-      "loss": 0.2771,
+      "epoch": 3.258550473925109,
+      "grad_norm": 0.22176389396190643,
+      "learning_rate": 1.4304890886518352e-05,
+      "loss": 0.3954,
       "step": 90415
     },
     {
-      "epoch": 3.18,
-      "learning_rate": 1.5449694918764785e-05,
-      "loss": 0.2595,
+      "epoch": 3.2587306735863337,
+      "grad_norm": 0.2822429835796356,
+      "learning_rate": 1.4302253341967564e-05,
+      "loss": 0.3836,
       "step": 90420
     },
     {
-      "epoch": 3.18,
-      "learning_rate": 1.5447062314235196e-05,
-      "loss": 0.2372,
+      "epoch": 3.2589108732475585,
+      "grad_norm": 0.23664531111717224,
+      "learning_rate": 1.4299615943171509e-05,
+      "loss": 0.3671,
       "step": 90425
     },
     {
-      "epoch": 3.18,
-      "learning_rate": 1.5444429833744213e-05,
-      "loss": 0.2799,
+      "epoch": 3.2590910729087827,
+      "grad_norm": 0.24214255809783936,
+      "learning_rate": 1.429697869016613e-05,
+      "loss": 0.404,
       "step": 90430
     },
     {
-      "epoch": 3.18,
-      "learning_rate": 1.5441797477326017e-05,
-      "loss": 0.2616,
+      "epoch": 3.2592712725700075,
+      "grad_norm": 0.21557815372943878,
+      "learning_rate": 1.429434158298732e-05,
+      "loss": 0.3875,
       "step": 90435
     },
     {
-      "epoch": 3.18,
-      "learning_rate": 1.543916524501477e-05,
-      "loss": 0.2696,
+      "epoch": 3.259451472231232,
+      "grad_norm": 0.2581924498081207,
+      "learning_rate": 1.4291704621671043e-05,
+      "loss": 0.3645,
       "step": 90440
     },
     {
-      "epoch": 3.18,
-      "learning_rate": 1.5436533136844665e-05,
-      "loss": 0.2675,
+      "epoch": 3.259631671892457,
+      "grad_norm": 0.23095859587192535,
+      "learning_rate": 1.4289067806253208e-05,
+      "loss": 0.3912,
       "step": 90445
     },
     {
-      "epoch": 3.18,
-      "learning_rate": 1.543390115284988e-05,
-      "loss": 0.264,
+      "epoch": 3.2598118715536817,
+      "grad_norm": 0.25443336367607117,
+      "learning_rate": 1.4286431136769774e-05,
+      "loss": 0.3588,
       "step": 90450
     },
     {
-      "epoch": 3.18,
-      "learning_rate": 1.5431269293064585e-05,
-      "loss": 0.2748,
+      "epoch": 3.259992071214906,
+      "grad_norm": 0.3058500289916992,
+      "learning_rate": 1.428379461325663e-05,
+      "loss": 0.4221,
       "step": 90455
     },
     {
-      "epoch": 3.18,
-      "learning_rate": 1.5428637557522942e-05,
-      "loss": 0.2724,
+      "epoch": 3.2601722708761307,
+      "grad_norm": 0.2576063573360443,
+      "learning_rate": 1.4281158235749698e-05,
+      "loss": 0.3781,
       "step": 90460
     },
     {
-      "epoch": 3.18,
-      "learning_rate": 1.542600594625914e-05,
-      "loss": 0.2642,
+      "epoch": 3.2603524705373554,
+      "grad_norm": 0.25434017181396484,
+      "learning_rate": 1.4278522004284919e-05,
+      "loss": 0.3894,
       "step": 90465
     },
     {
-      "epoch": 3.18,
-      "learning_rate": 1.542337445930734e-05,
-      "loss": 0.2555,
+      "epoch": 3.26053267019858,
+      "grad_norm": 0.2770051956176758,
+      "learning_rate": 1.4275885918898201e-05,
+      "loss": 0.4269,
       "step": 90470
     },
     {
-      "epoch": 3.18,
-      "learning_rate": 1.5420743096701717e-05,
-      "loss": 0.2335,
+      "epoch": 3.2607128698598045,
+      "grad_norm": 0.24559049308300018,
+      "learning_rate": 1.4273249979625464e-05,
+      "loss": 0.3713,
       "step": 90475
     },
     {
-      "epoch": 3.18,
-      "learning_rate": 1.541811185847642e-05,
-      "loss": 0.2708,
+      "epoch": 3.260893069521029,
+      "grad_norm": 0.24192962050437927,
+      "learning_rate": 1.427061418650262e-05,
+      "loss": 0.3708,
       "step": 90480
     },
     {
-      "epoch": 3.18,
-      "learning_rate": 1.5415480744665635e-05,
-      "loss": 0.2621,
+      "epoch": 3.261073269182254,
+      "grad_norm": 0.23208360373973846,
+      "learning_rate": 1.4268505657256397e-05,
+      "loss": 0.3847,
       "step": 90485
     },
     {
-      "epoch": 3.18,
-      "learning_rate": 1.5412849755303514e-05,
-      "loss": 0.2507,
+      "epoch": 3.2612534688434787,
+      "grad_norm": 0.26153233647346497,
+      "learning_rate": 1.4265870127293846e-05,
+      "loss": 0.425,
       "step": 90490
     },
     {
-      "epoch": 3.18,
-      "learning_rate": 1.5410218890424222e-05,
-      "loss": 0.2686,
+      "epoch": 3.2614336685047034,
+      "grad_norm": 0.22417205572128296,
+      "learning_rate": 1.4263234743581744e-05,
+      "loss": 0.396,
       "step": 90495
     },
     {
-      "epoch": 3.18,
-      "learning_rate": 1.5407588150061907e-05,
-      "loss": 0.2645,
+      "epoch": 3.2616138681659277,
+      "grad_norm": 0.3092048764228821,
+      "learning_rate": 1.4260599506156002e-05,
+      "loss": 0.392,
       "step": 90500
     },
     {
-      "epoch": 3.18,
-      "eval_loss": 0.2550298273563385,
-      "eval_runtime": 10.5455,
-      "eval_samples_per_second": 9.483,
-      "eval_steps_per_second": 9.483,
+      "epoch": 3.2616138681659277,
+      "eval_loss": 0.4304320514202118,
+      "eval_runtime": 3.5275,
+      "eval_samples_per_second": 28.349,
+      "eval_steps_per_second": 7.087,
       "step": 90500
     },
     {
-      "epoch": 3.18,
-      "learning_rate": 1.5404957534250752e-05,
-      "loss": 0.2426,
+      "epoch": 3.2617940678271524,
+      "grad_norm": 0.19845503568649292,
+      "learning_rate": 1.4257964415052491e-05,
+      "loss": 0.3698,
       "step": 90505
     },
     {
-      "epoch": 3.18,
-      "learning_rate": 1.54023270430249e-05,
-      "loss": 0.2579,
+      "epoch": 3.261974267488377,
+      "grad_norm": 0.29068851470947266,
+      "learning_rate": 1.4255329470307146e-05,
+      "loss": 0.393,
       "step": 90510
     },
     {
-      "epoch": 3.18,
-      "learning_rate": 1.5399696676418506e-05,
-      "loss": 0.2536,
+      "epoch": 3.262154467149602,
+      "grad_norm": 0.22075437009334564,
+      "learning_rate": 1.4252694671955847e-05,
+      "loss": 0.3644,
       "step": 90515
     },
     {
-      "epoch": 3.18,
-      "learning_rate": 1.5397066434465712e-05,
-      "loss": 0.245,
+      "epoch": 3.262334666810826,
+      "grad_norm": 0.2166365683078766,
+      "learning_rate": 1.4250060020034511e-05,
+      "loss": 0.362,
       "step": 90520
     },
     {
-      "epoch": 3.18,
-      "learning_rate": 1.5394436317200687e-05,
-      "loss": 0.2526,
+      "epoch": 3.262514866472051,
+      "grad_norm": 0.23188582062721252,
+      "learning_rate": 1.4247425514579038e-05,
+      "loss": 0.3859,
       "step": 90525
     },
     {
-      "epoch": 3.19,
-      "learning_rate": 1.539180632465758e-05,
-      "loss": 0.2504,
+      "epoch": 3.2626950661332756,
+      "grad_norm": 0.1937790811061859,
+      "learning_rate": 1.4244791155625289e-05,
+      "loss": 0.3916,
       "step": 90530
     },
     {
-      "epoch": 3.19,
-      "learning_rate": 1.538917645687054e-05,
-      "loss": 0.266,
+      "epoch": 3.2628752657945004,
+      "grad_norm": 0.20566551387310028,
+      "learning_rate": 1.424215694320919e-05,
+      "loss": 0.3681,
       "step": 90535
     },
     {
-      "epoch": 3.19,
-      "learning_rate": 1.5386546713873695e-05,
-      "loss": 0.2471,
+      "epoch": 3.263055465455725,
+      "grad_norm": 0.21733631193637848,
+      "learning_rate": 1.4239522877366618e-05,
+      "loss": 0.3715,
       "step": 90540
     },
     {
-      "epoch": 3.19,
-      "learning_rate": 1.5383917095701218e-05,
-      "loss": 0.2568,
+      "epoch": 3.2632356651169494,
+      "grad_norm": 0.23890431225299835,
+      "learning_rate": 1.4236888958133464e-05,
+      "loss": 0.3793,
       "step": 90545
     },
     {
-      "epoch": 3.19,
-      "learning_rate": 1.5381287602387236e-05,
-      "loss": 0.2708,
+      "epoch": 3.263415864778174,
+      "grad_norm": 0.19045878946781158,
+      "learning_rate": 1.4234255185545614e-05,
+      "loss": 0.3516,
       "step": 90550
     },
     {
-      "epoch": 3.19,
-      "learning_rate": 1.5378658233965893e-05,
-      "loss": 0.2876,
+      "epoch": 3.263596064439399,
+      "grad_norm": 0.27482134103775024,
+      "learning_rate": 1.4231621559638941e-05,
+      "loss": 0.3766,
       "step": 90555
     },
     {
-      "epoch": 3.19,
-      "learning_rate": 1.5376028990471336e-05,
-      "loss": 0.2632,
+      "epoch": 3.2637762641006236,
+      "grad_norm": 0.20117837190628052,
+      "learning_rate": 1.422898808044935e-05,
+      "loss": 0.3557,
       "step": 90560
     },
     {
-      "epoch": 3.19,
-      "learning_rate": 1.5373399871937706e-05,
-      "loss": 0.2386,
+      "epoch": 3.263956463761848,
+      "grad_norm": 0.30366307497024536,
+      "learning_rate": 1.4226354748012716e-05,
+      "loss": 0.4117,
       "step": 90565
     },
     {
-      "epoch": 3.19,
-      "learning_rate": 1.537077087839913e-05,
-      "loss": 0.2629,
+      "epoch": 3.2641366634230726,
+      "grad_norm": 0.23265069723129272,
+      "learning_rate": 1.4223721562364908e-05,
+      "loss": 0.3705,
       "step": 90570
     },
     {
-      "epoch": 3.19,
-      "learning_rate": 1.5368142009889737e-05,
-      "loss": 0.2412,
+      "epoch": 3.2643168630842974,
+      "grad_norm": 0.2623595893383026,
+      "learning_rate": 1.4221088523541809e-05,
+      "loss": 0.3727,
       "step": 90575
     },
     {
-      "epoch": 3.19,
-      "learning_rate": 1.5365513266443683e-05,
-      "loss": 0.2609,
+      "epoch": 3.264497062745522,
+      "grad_norm": 0.22619140148162842,
+      "learning_rate": 1.421845563157928e-05,
+      "loss": 0.4174,
       "step": 90580
     },
     {
-      "epoch": 3.19,
-      "learning_rate": 1.536288464809509e-05,
-      "loss": 0.2865,
+      "epoch": 3.264677262406747,
+      "grad_norm": 0.25267648696899414,
+      "learning_rate": 1.421582288651322e-05,
+      "loss": 0.395,
       "step": 90585
     },
     {
-      "epoch": 3.19,
-      "learning_rate": 1.5360256154878093e-05,
-      "loss": 0.2689,
+      "epoch": 3.2648574620679716,
+      "grad_norm": 0.30874866247177124,
+      "learning_rate": 1.421319028837948e-05,
+      "loss": 0.4297,
       "step": 90590
     },
     {
-      "epoch": 3.19,
-      "learning_rate": 1.53576277868268e-05,
-      "loss": 0.2651,
+      "epoch": 3.265037661729196,
+      "grad_norm": 0.2350592315196991,
+      "learning_rate": 1.421055783721394e-05,
+      "loss": 0.3709,
       "step": 90595
     },
     {
-      "epoch": 3.19,
-      "learning_rate": 1.5354999543975372e-05,
-      "loss": 0.251,
+      "epoch": 3.2652178613904206,
+      "grad_norm": 0.22834807634353638,
+      "learning_rate": 1.4207925533052455e-05,
+      "loss": 0.3666,
       "step": 90600
     },
     {
-      "epoch": 3.19,
-      "learning_rate": 1.5352371426357913e-05,
-      "loss": 0.2549,
+      "epoch": 3.2653980610516453,
+      "grad_norm": 0.21915757656097412,
+      "learning_rate": 1.4205293375930898e-05,
+      "loss": 0.3658,
       "step": 90605
     },
     {
-      "epoch": 3.19,
-      "learning_rate": 1.5349743434008556e-05,
-      "loss": 0.245,
+      "epoch": 3.26557826071287,
+      "grad_norm": 0.2512879967689514,
+      "learning_rate": 1.4202661365885116e-05,
+      "loss": 0.3719,
       "step": 90610
     },
     {
-      "epoch": 3.19,
-      "learning_rate": 1.534711556696141e-05,
-      "loss": 0.2429,
+      "epoch": 3.2657584603740943,
+      "grad_norm": 0.19401970505714417,
+      "learning_rate": 1.4200029502951007e-05,
+      "loss": 0.3696,
       "step": 90615
     },
     {
-      "epoch": 3.19,
-      "learning_rate": 1.5344487825250624e-05,
-      "loss": 0.2646,
+      "epoch": 3.265938660035319,
+      "grad_norm": 0.2430254966020584,
+      "learning_rate": 1.419739778716438e-05,
+      "loss": 0.3845,
       "step": 90620
     },
     {
-      "epoch": 3.19,
-      "learning_rate": 1.5341860208910292e-05,
-      "loss": 0.2681,
+      "epoch": 3.266118859696544,
+      "grad_norm": 0.24256938695907593,
+      "learning_rate": 1.4194766218561128e-05,
+      "loss": 0.3866,
       "step": 90625
     },
     {
-      "epoch": 3.19,
-      "learning_rate": 1.533923271797454e-05,
-      "loss": 0.2423,
+      "epoch": 3.2662990593577685,
+      "grad_norm": 0.2600161135196686,
+      "learning_rate": 1.4192134797177098e-05,
+      "loss": 0.3928,
       "step": 90630
     },
     {
-      "epoch": 3.19,
-      "learning_rate": 1.5336605352477476e-05,
-      "loss": 0.2569,
+      "epoch": 3.2664792590189933,
+      "grad_norm": 0.21937114000320435,
+      "learning_rate": 1.4189503523048137e-05,
+      "loss": 0.4063,
       "step": 90635
     },
     {
-      "epoch": 3.19,
-      "learning_rate": 1.5333978112453235e-05,
-      "loss": 0.2557,
+      "epoch": 3.2666594586802176,
+      "grad_norm": 0.21056115627288818,
+      "learning_rate": 1.4186872396210096e-05,
+      "loss": 0.3868,
       "step": 90640
     },
     {
-      "epoch": 3.19,
-      "learning_rate": 1.5331350997935916e-05,
-      "loss": 0.2587,
+      "epoch": 3.2668396583414423,
+      "grad_norm": 0.2408798485994339,
+      "learning_rate": 1.4184241416698827e-05,
+      "loss": 0.3905,
       "step": 90645
     },
     {
-      "epoch": 3.19,
-      "learning_rate": 1.5328724008959637e-05,
-      "loss": 0.2696,
+      "epoch": 3.267019858002667,
+      "grad_norm": 0.2301928699016571,
+      "learning_rate": 1.4181610584550165e-05,
+      "loss": 0.4008,
       "step": 90650
     },
     {
-      "epoch": 3.19,
-      "learning_rate": 1.532609714555849e-05,
-      "loss": 0.2512,
+      "epoch": 3.2672000576638918,
+      "grad_norm": 0.21639235317707062,
+      "learning_rate": 1.4178979899799971e-05,
+      "loss": 0.3618,
       "step": 90655
     },
     {
-      "epoch": 3.19,
-      "learning_rate": 1.5323470407766602e-05,
-      "loss": 0.2534,
+      "epoch": 3.267380257325116,
+      "grad_norm": 0.20622292160987854,
+      "learning_rate": 1.4176349362484085e-05,
+      "loss": 0.3987,
       "step": 90660
     },
     {
-      "epoch": 3.19,
-      "learning_rate": 1.532084379561808e-05,
-      "loss": 0.2649,
+      "epoch": 3.267560456986341,
+      "grad_norm": 0.24098491668701172,
+      "learning_rate": 1.4173718972638339e-05,
+      "loss": 0.4114,
       "step": 90665
     },
     {
-      "epoch": 3.19,
-      "learning_rate": 1.531821730914702e-05,
-      "loss": 0.26,
+      "epoch": 3.2677406566475655,
+      "grad_norm": 0.2708558142185211,
+      "learning_rate": 1.417108873029858e-05,
+      "loss": 0.388,
       "step": 90670
     },
     {
-      "epoch": 3.19,
-      "learning_rate": 1.5315590948387516e-05,
-      "loss": 0.2505,
+      "epoch": 3.2679208563087903,
+      "grad_norm": 0.24035510420799255,
+      "learning_rate": 1.4168458635500628e-05,
+      "loss": 0.3668,
       "step": 90675
     },
     {
-      "epoch": 3.19,
-      "learning_rate": 1.5312964713373695e-05,
-      "loss": 0.2635,
+      "epoch": 3.268101055970015,
+      "grad_norm": 0.22504648566246033,
+      "learning_rate": 1.416582868828034e-05,
+      "loss": 0.4099,
       "step": 90680
     },
     {
-      "epoch": 3.19,
-      "learning_rate": 1.5310338604139638e-05,
-      "loss": 0.2791,
+      "epoch": 3.2682812556312393,
+      "grad_norm": 0.21226103603839874,
+      "learning_rate": 1.416319888867354e-05,
+      "loss": 0.3926,
       "step": 90685
     },
     {
-      "epoch": 3.19,
-      "learning_rate": 1.5307712620719444e-05,
-      "loss": 0.2511,
+      "epoch": 3.268461455292464,
+      "grad_norm": 0.2621418535709381,
+      "learning_rate": 1.4160569236716059e-05,
+      "loss": 0.4155,
       "step": 90690
     },
     {
-      "epoch": 3.19,
-      "learning_rate": 1.530508676314722e-05,
-      "loss": 0.2455,
+      "epoch": 3.2686416549536887,
+      "grad_norm": 0.2505376636981964,
+      "learning_rate": 1.415793973244372e-05,
+      "loss": 0.3616,
       "step": 90695
     },
     {
-      "epoch": 3.19,
-      "learning_rate": 1.5302461031457057e-05,
-      "loss": 0.2518,
+      "epoch": 3.2688218546149135,
+      "grad_norm": 0.24424095451831818,
+      "learning_rate": 1.4155310375892345e-05,
+      "loss": 0.39,
       "step": 90700
     },
     {
-      "epoch": 3.19,
-      "learning_rate": 1.529983542568305e-05,
-      "loss": 0.2622,
+      "epoch": 3.2690020542761378,
+      "grad_norm": 0.2086549997329712,
+      "learning_rate": 1.4152681167097775e-05,
+      "loss": 0.3737,
       "step": 90705
     },
     {
-      "epoch": 3.19,
-      "learning_rate": 1.5297209945859274e-05,
-      "loss": 0.2675,
+      "epoch": 3.2691822539373625,
+      "grad_norm": 0.25878942012786865,
+      "learning_rate": 1.4150052106095834e-05,
+      "loss": 0.4075,
       "step": 90710
     },
     {
-      "epoch": 3.19,
-      "learning_rate": 1.5294584592019835e-05,
-      "loss": 0.2519,
+      "epoch": 3.2693624535985872,
+      "grad_norm": 0.21511436998844147,
+      "learning_rate": 1.4147423192922314e-05,
+      "loss": 0.3487,
       "step": 90715
     },
     {
-      "epoch": 3.19,
-      "learning_rate": 1.529195936419883e-05,
-      "loss": 0.2756,
+      "epoch": 3.269542653259812,
+      "grad_norm": 0.27404505014419556,
+      "learning_rate": 1.4144794427613061e-05,
+      "loss": 0.3914,
       "step": 90720
     },
     {
-      "epoch": 3.19,
-      "learning_rate": 1.5289334262430332e-05,
-      "loss": 0.2521,
+      "epoch": 3.2697228529210367,
+      "grad_norm": 0.24359211325645447,
+      "learning_rate": 1.4142165810203881e-05,
+      "loss": 0.3774,
       "step": 90725
     },
     {
-      "epoch": 3.19,
-      "learning_rate": 1.5286709286748417e-05,
-      "loss": 0.2765,
+      "epoch": 3.269903052582261,
+      "grad_norm": 0.19888465106487274,
+      "learning_rate": 1.413953734073059e-05,
+      "loss": 0.3661,
       "step": 90730
     },
     {
-      "epoch": 3.19,
-      "learning_rate": 1.5284084437187194e-05,
-      "loss": 0.2716,
+      "epoch": 3.2700832522434857,
+      "grad_norm": 0.250573992729187,
+      "learning_rate": 1.4136909019229003e-05,
+      "loss": 0.3367,
       "step": 90735
     },
     {
-      "epoch": 3.19,
-      "learning_rate": 1.5281459713780726e-05,
-      "loss": 0.2615,
+      "epoch": 3.2702634519047105,
+      "grad_norm": 0.2471703588962555,
+      "learning_rate": 1.4134280845734915e-05,
+      "loss": 0.3536,
       "step": 90740
     },
     {
-      "epoch": 3.19,
-      "learning_rate": 1.52788351165631e-05,
-      "loss": 0.2552,
+      "epoch": 3.270443651565935,
+      "grad_norm": 0.2185695767402649,
+      "learning_rate": 1.4131652820284158e-05,
+      "loss": 0.39,
       "step": 90745
     },
     {
-      "epoch": 3.19,
-      "learning_rate": 1.527621064556839e-05,
-      "loss": 0.2574,
+      "epoch": 3.2706238512271595,
+      "grad_norm": 0.22518201172351837,
+      "learning_rate": 1.4129024942912525e-05,
+      "loss": 0.3999,
       "step": 90750
     },
     {
-      "epoch": 3.19,
-      "learning_rate": 1.5273586300830684e-05,
-      "loss": 0.2539,
+      "epoch": 3.270804050888384,
+      "grad_norm": 0.23426784574985504,
+      "learning_rate": 1.4126397213655824e-05,
+      "loss": 0.3719,
       "step": 90755
     },
     {
-      "epoch": 3.19,
-      "learning_rate": 1.5270962082384047e-05,
-      "loss": 0.2696,
+      "epoch": 3.270984250549609,
+      "grad_norm": 0.28812581300735474,
+      "learning_rate": 1.4123769632549855e-05,
+      "loss": 0.3956,
       "step": 90760
     },
     {
-      "epoch": 3.19,
-      "learning_rate": 1.5268337990262555e-05,
-      "loss": 0.2712,
+      "epoch": 3.2711644502108337,
+      "grad_norm": 0.2103385478258133,
+      "learning_rate": 1.412114219963041e-05,
+      "loss": 0.4033,
       "step": 90765
     },
     {
-      "epoch": 3.19,
-      "learning_rate": 1.5265714024500276e-05,
-      "loss": 0.279,
+      "epoch": 3.2713446498720584,
+      "grad_norm": 0.23068909347057343,
+      "learning_rate": 1.4118514914933306e-05,
+      "loss": 0.3719,
       "step": 90770
     },
     {
-      "epoch": 3.19,
-      "learning_rate": 1.5263090185131287e-05,
-      "loss": 0.2613,
+      "epoch": 3.2715248495332827,
+      "grad_norm": 0.219300776720047,
+      "learning_rate": 1.411588777849433e-05,
+      "loss": 0.3917,
       "step": 90775
     },
     {
-      "epoch": 3.19,
-      "learning_rate": 1.526046647218966e-05,
-      "loss": 0.2419,
+      "epoch": 3.2717050491945074,
+      "grad_norm": 0.18937250971794128,
+      "learning_rate": 1.4113260790349275e-05,
+      "loss": 0.3667,
       "step": 90780
     },
     {
-      "epoch": 3.19,
-      "learning_rate": 1.525784288570946e-05,
-      "loss": 0.2895,
+      "epoch": 3.271885248855732,
+      "grad_norm": 0.24838852882385254,
+      "learning_rate": 1.4110633950533936e-05,
+      "loss": 0.3858,
       "step": 90785
     },
     {
-      "epoch": 3.19,
-      "learning_rate": 1.5255219425724737e-05,
-      "loss": 0.2659,
+      "epoch": 3.272065448516957,
+      "grad_norm": 0.23256389796733856,
+      "learning_rate": 1.4108007259084099e-05,
+      "loss": 0.3784,
       "step": 90790
     },
     {
-      "epoch": 3.19,
-      "learning_rate": 1.5252596092269575e-05,
-      "loss": 0.2426,
+      "epoch": 3.272245648178181,
+      "grad_norm": 0.2195148915052414,
+      "learning_rate": 1.4105380716035544e-05,
+      "loss": 0.3441,
       "step": 90795
     },
     {
-      "epoch": 3.19,
-      "learning_rate": 1.5249972885378034e-05,
-      "loss": 0.2553,
+      "epoch": 3.272425847839406,
+      "grad_norm": 0.23615360260009766,
+      "learning_rate": 1.4102754321424088e-05,
+      "loss": 0.4046,
       "step": 90800
     },
     {
-      "epoch": 3.19,
-      "learning_rate": 1.5247349805084155e-05,
-      "loss": 0.2577,
+      "epoch": 3.2726060475006307,
+      "grad_norm": 0.25502121448516846,
+      "learning_rate": 1.4100128075285473e-05,
+      "loss": 0.3555,
       "step": 90805
     },
     {
-      "epoch": 3.19,
-      "learning_rate": 1.5244726851422026e-05,
-      "loss": 0.2744,
+      "epoch": 3.2727862471618554,
+      "grad_norm": 0.23027649521827698,
+      "learning_rate": 1.4097501977655517e-05,
+      "loss": 0.3715,
       "step": 90810
     },
     {
-      "epoch": 3.2,
-      "learning_rate": 1.5242104024425691e-05,
-      "loss": 0.2499,
+      "epoch": 3.27296644682308,
+      "grad_norm": 0.25443360209465027,
+      "learning_rate": 1.4094876028569983e-05,
+      "loss": 0.3668,
       "step": 90815
     },
     {
-      "epoch": 3.2,
-      "learning_rate": 1.5239481324129195e-05,
-      "loss": 0.2641,
+      "epoch": 3.2731466464843044,
+      "grad_norm": 0.28707796335220337,
+      "learning_rate": 1.4092250228064643e-05,
+      "loss": 0.4048,
       "step": 90820
     },
     {
-      "epoch": 3.2,
-      "learning_rate": 1.5236858750566604e-05,
-      "loss": 0.2535,
+      "epoch": 3.273326846145529,
+      "grad_norm": 0.2626727521419525,
+      "learning_rate": 1.4089624576175301e-05,
+      "loss": 0.4063,
       "step": 90825
     },
     {
-      "epoch": 3.2,
-      "learning_rate": 1.5234236303771967e-05,
-      "loss": 0.2481,
+      "epoch": 3.273507045806754,
+      "grad_norm": 0.2237686812877655,
+      "learning_rate": 1.4086999072937702e-05,
+      "loss": 0.3797,
       "step": 90830
     },
     {
-      "epoch": 3.2,
-      "learning_rate": 1.5231613983779341e-05,
-      "loss": 0.2672,
+      "epoch": 3.2736872454679786,
+      "grad_norm": 0.20242911577224731,
+      "learning_rate": 1.4084373718387617e-05,
+      "loss": 0.3658,
       "step": 90835
     },
     {
-      "epoch": 3.2,
-      "learning_rate": 1.522899179062277e-05,
-      "loss": 0.2875,
+      "epoch": 3.273867445129203,
+      "grad_norm": 0.18086695671081543,
+      "learning_rate": 1.4081748512560838e-05,
+      "loss": 0.3622,
       "step": 90840
     },
     {
-      "epoch": 3.2,
-      "learning_rate": 1.5226369724336292e-05,
-      "loss": 0.2434,
+      "epoch": 3.2740476447904276,
+      "grad_norm": 0.269579142332077,
+      "learning_rate": 1.4079123455493121e-05,
+      "loss": 0.4093,
       "step": 90845
     },
     {
-      "epoch": 3.2,
-      "learning_rate": 1.5223747784953974e-05,
-      "loss": 0.2503,
+      "epoch": 3.2742278444516524,
+      "grad_norm": 0.20566615462303162,
+      "learning_rate": 1.4076498547220234e-05,
+      "loss": 0.3837,
       "step": 90850
     },
     {
-      "epoch": 3.2,
-      "learning_rate": 1.5221125972509839e-05,
-      "loss": 0.2584,
+      "epoch": 3.274408044112877,
+      "grad_norm": 0.22256791591644287,
+      "learning_rate": 1.4073873787777936e-05,
+      "loss": 0.3359,
       "step": 90855
     },
     {
-      "epoch": 3.2,
-      "learning_rate": 1.5218504287037947e-05,
-      "loss": 0.2519,
+      "epoch": 3.274588243774102,
+      "grad_norm": 0.2186269313097,
+      "learning_rate": 1.4071249177201984e-05,
+      "loss": 0.372,
       "step": 90860
     },
     {
-      "epoch": 3.2,
-      "learning_rate": 1.5215882728572317e-05,
-      "loss": 0.2567,
+      "epoch": 3.2747684434353266,
+      "grad_norm": 0.21954451501369476,
+      "learning_rate": 1.4068624715528158e-05,
+      "loss": 0.3913,
       "step": 90865
     },
     {
-      "epoch": 3.2,
-      "learning_rate": 1.5213261297147014e-05,
-      "loss": 0.2547,
+      "epoch": 3.274948643096551,
+      "grad_norm": 0.2596687972545624,
+      "learning_rate": 1.40660004027922e-05,
+      "loss": 0.371,
       "step": 90870
     },
     {
-      "epoch": 3.2,
-      "learning_rate": 1.5210639992796063e-05,
-      "loss": 0.2447,
+      "epoch": 3.2751288427577756,
+      "grad_norm": 0.2262258678674698,
+      "learning_rate": 1.4063376239029875e-05,
+      "loss": 0.4175,
       "step": 90875
     },
     {
-      "epoch": 3.2,
-      "learning_rate": 1.5208018815553496e-05,
-      "loss": 0.2541,
+      "epoch": 3.2753090424190003,
+      "grad_norm": 0.26203519105911255,
+      "learning_rate": 1.406075222427693e-05,
+      "loss": 0.3837,
       "step": 90880
     },
     {
-      "epoch": 3.2,
-      "learning_rate": 1.5205397765453348e-05,
-      "loss": 0.2599,
+      "epoch": 3.275489242080225,
+      "grad_norm": 0.20488256216049194,
+      "learning_rate": 1.4058128358569106e-05,
+      "loss": 0.3964,
       "step": 90885
     },
     {
-      "epoch": 3.2,
-      "learning_rate": 1.5202776842529662e-05,
-      "loss": 0.2661,
+      "epoch": 3.2756694417414494,
+      "grad_norm": 0.23428840935230255,
+      "learning_rate": 1.4055504641942175e-05,
+      "loss": 0.3849,
       "step": 90890
     },
     {
-      "epoch": 3.2,
-      "learning_rate": 1.5200156046816463e-05,
-      "loss": 0.2734,
+      "epoch": 3.275849641402674,
+      "grad_norm": 0.23843036592006683,
+      "learning_rate": 1.4052881074431884e-05,
+      "loss": 0.3872,
       "step": 90895
     },
     {
-      "epoch": 3.2,
-      "learning_rate": 1.5197535378347775e-05,
-      "loss": 0.2833,
+      "epoch": 3.276029841063899,
+      "grad_norm": 0.22145730257034302,
+      "learning_rate": 1.4050257656073947e-05,
+      "loss": 0.4031,
       "step": 90900
     },
     {
-      "epoch": 3.2,
-      "learning_rate": 1.5194914837157631e-05,
-      "loss": 0.2635,
+      "epoch": 3.2762100407251236,
+      "grad_norm": 0.2753254175186157,
+      "learning_rate": 1.4047634386904142e-05,
+      "loss": 0.3998,
       "step": 90905
     },
     {
-      "epoch": 3.2,
-      "learning_rate": 1.5192294423280052e-05,
-      "loss": 0.2675,
+      "epoch": 3.2763902403863483,
+      "grad_norm": 0.18128567934036255,
+      "learning_rate": 1.4045011266958197e-05,
+      "loss": 0.4172,
       "step": 90910
     },
     {
-      "epoch": 3.2,
-      "learning_rate": 1.5189674136749077e-05,
-      "loss": 0.2532,
+      "epoch": 3.2765704400475726,
+      "grad_norm": 0.2414032369852066,
+      "learning_rate": 1.4042388296271856e-05,
+      "loss": 0.3816,
       "step": 90915
     },
     {
-      "epoch": 3.2,
-      "learning_rate": 1.5187053977598702e-05,
-      "loss": 0.2469,
+      "epoch": 3.2767506397087973,
+      "grad_norm": 0.21249370276927948,
+      "learning_rate": 1.4039765474880851e-05,
+      "loss": 0.3528,
       "step": 90920
     },
     {
-      "epoch": 3.2,
-      "learning_rate": 1.518443394586298e-05,
-      "loss": 0.2738,
+      "epoch": 3.276930839370022,
+      "grad_norm": 0.238897442817688,
+      "learning_rate": 1.4037142802820907e-05,
+      "loss": 0.4205,
       "step": 90925
     },
     {
-      "epoch": 3.2,
-      "learning_rate": 1.5181814041575908e-05,
-      "loss": 0.2736,
+      "epoch": 3.277111039031247,
+      "grad_norm": 0.2670360803604126,
+      "learning_rate": 1.4034520280127782e-05,
+      "loss": 0.3664,
       "step": 90930
     },
     {
-      "epoch": 3.2,
-      "learning_rate": 1.5179194264771512e-05,
-      "loss": 0.2674,
+      "epoch": 3.277291238692471,
+      "grad_norm": 0.20420727133750916,
+      "learning_rate": 1.4031897906837194e-05,
+      "loss": 0.3831,
       "step": 90935
     },
     {
-      "epoch": 3.2,
-      "learning_rate": 1.5176574615483796e-05,
-      "loss": 0.2715,
+      "epoch": 3.277471438353696,
+      "grad_norm": 0.2537064850330353,
+      "learning_rate": 1.4029275682984878e-05,
+      "loss": 0.3979,
       "step": 90940
     },
     {
-      "epoch": 3.2,
-      "learning_rate": 1.51739550937468e-05,
-      "loss": 0.2848,
+      "epoch": 3.2776516380149205,
+      "grad_norm": 0.3141496181488037,
+      "learning_rate": 1.402665360860655e-05,
+      "loss": 0.3572,
       "step": 90945
     },
     {
-      "epoch": 3.2,
-      "learning_rate": 1.5171335699594519e-05,
-      "loss": 0.2449,
+      "epoch": 3.2778318376761453,
+      "grad_norm": 0.18175874650478363,
+      "learning_rate": 1.4024031683737937e-05,
+      "loss": 0.387,
       "step": 90950
     },
     {
-      "epoch": 3.2,
-      "learning_rate": 1.5168716433060964e-05,
-      "loss": 0.2451,
+      "epoch": 3.27801203733737,
+      "grad_norm": 0.2895739674568176,
+      "learning_rate": 1.4021409908414773e-05,
+      "loss": 0.367,
       "step": 90955
     },
     {
-      "epoch": 3.2,
-      "learning_rate": 1.516609729418014e-05,
-      "loss": 0.2459,
+      "epoch": 3.2781922369985943,
+      "grad_norm": 0.21518366038799286,
+      "learning_rate": 1.401878828267278e-05,
+      "loss": 0.3816,
       "step": 90960
     },
     {
-      "epoch": 3.2,
-      "learning_rate": 1.5163478282986065e-05,
-      "loss": 0.2462,
+      "epoch": 3.278372436659819,
+      "grad_norm": 0.25440242886543274,
+      "learning_rate": 1.4016166806547664e-05,
+      "loss": 0.3703,
       "step": 90965
     },
     {
-      "epoch": 3.2,
-      "learning_rate": 1.516085939951275e-05,
-      "loss": 0.2721,
+      "epoch": 3.2785526363210438,
+      "grad_norm": 0.271392822265625,
+      "learning_rate": 1.4013545480075153e-05,
+      "loss": 0.4075,
       "step": 90970
     },
     {
-      "epoch": 3.2,
-      "learning_rate": 1.515824064379419e-05,
-      "loss": 0.275,
+      "epoch": 3.2787328359822685,
+      "grad_norm": 0.22662685811519623,
+      "learning_rate": 1.4010924303290957e-05,
+      "loss": 0.3931,
       "step": 90975
     },
     {
-      "epoch": 3.2,
-      "learning_rate": 1.515562201586438e-05,
-      "loss": 0.2739,
+      "epoch": 3.278913035643493,
+      "grad_norm": 0.24230170249938965,
+      "learning_rate": 1.4008303276230777e-05,
+      "loss": 0.378,
       "step": 90980
     },
     {
-      "epoch": 3.2,
-      "learning_rate": 1.5153003515757335e-05,
-      "loss": 0.2541,
+      "epoch": 3.2790932353047175,
+      "grad_norm": 0.2077885866165161,
+      "learning_rate": 1.4005682398930347e-05,
+      "loss": 0.3544,
       "step": 90985
     },
     {
-      "epoch": 3.2,
-      "learning_rate": 1.5150385143507052e-05,
-      "loss": 0.2435,
+      "epoch": 3.2792734349659423,
+      "grad_norm": 0.22285114228725433,
+      "learning_rate": 1.4003061671425368e-05,
+      "loss": 0.3823,
       "step": 90990
     },
     {
-      "epoch": 3.2,
-      "learning_rate": 1.5147766899147531e-05,
-      "loss": 0.2621,
+      "epoch": 3.279453634627167,
+      "grad_norm": 0.22684328258037567,
+      "learning_rate": 1.4000441093751546e-05,
+      "loss": 0.3657,
       "step": 90995
     },
     {
-      "epoch": 3.2,
-      "learning_rate": 1.5145148782712748e-05,
-      "loss": 0.2644,
+      "epoch": 3.2796338342883917,
+      "grad_norm": 0.2049197107553482,
+      "learning_rate": 1.3997820665944584e-05,
+      "loss": 0.4181,
       "step": 91000
     },
     {
-      "epoch": 3.2,
-      "eval_loss": 0.25480425357818604,
-      "eval_runtime": 10.5461,
-      "eval_samples_per_second": 9.482,
-      "eval_steps_per_second": 9.482,
+      "epoch": 3.2796338342883917,
+      "eval_loss": 0.4301852881908417,
+      "eval_runtime": 3.5299,
+      "eval_samples_per_second": 28.329,
+      "eval_steps_per_second": 7.082,
       "step": 91000
     },
     {
-      "epoch": 3.2,
-      "learning_rate": 1.5142530794236731e-05,
-      "loss": 0.2754,
+      "epoch": 3.279814033949616,
+      "grad_norm": 0.22128911316394806,
+      "learning_rate": 1.3995200388040172e-05,
+      "loss": 0.3903,
       "step": 91005
     },
     {
-      "epoch": 3.2,
-      "learning_rate": 1.5139912933753451e-05,
-      "loss": 0.2911,
+      "epoch": 3.2799942336108407,
+      "grad_norm": 0.3121122717857361,
+      "learning_rate": 1.3992580260074046e-05,
+      "loss": 0.4014,
       "step": 91010
     },
     {
-      "epoch": 3.2,
-      "learning_rate": 1.5137295201296897e-05,
-      "loss": 0.2636,
+      "epoch": 3.2801744332720655,
+      "grad_norm": 0.25039294362068176,
+      "learning_rate": 1.3989960282081874e-05,
+      "loss": 0.3716,
       "step": 91015
     },
     {
-      "epoch": 3.2,
-      "learning_rate": 1.5134677596901065e-05,
-      "loss": 0.2511,
+      "epoch": 3.28035463293329,
+      "grad_norm": 0.20780634880065918,
+      "learning_rate": 1.398734045409935e-05,
+      "loss": 0.3885,
       "step": 91020
     },
     {
-      "epoch": 3.2,
-      "learning_rate": 1.513206012059995e-05,
-      "loss": 0.281,
+      "epoch": 3.2805348325945145,
+      "grad_norm": 0.204863503575325,
+      "learning_rate": 1.3984720776162192e-05,
+      "loss": 0.3911,
       "step": 91025
     },
     {
-      "epoch": 3.2,
-      "learning_rate": 1.5129442772427531e-05,
-      "loss": 0.2354,
+      "epoch": 3.2807150322557392,
+      "grad_norm": 0.24138136208057404,
+      "learning_rate": 1.398210124830608e-05,
+      "loss": 0.3948,
       "step": 91030
     },
     {
-      "epoch": 3.2,
-      "learning_rate": 1.5126825552417784e-05,
-      "loss": 0.2607,
+      "epoch": 3.280895231916964,
+      "grad_norm": 0.21834784746170044,
+      "learning_rate": 1.3979481870566703e-05,
+      "loss": 0.3692,
       "step": 91035
     },
     {
-      "epoch": 3.2,
-      "learning_rate": 1.5124208460604708e-05,
-      "loss": 0.2657,
+      "epoch": 3.2810754315781887,
+      "grad_norm": 0.20662029087543488,
+      "learning_rate": 1.3976862642979755e-05,
+      "loss": 0.4021,
       "step": 91040
     },
     {
-      "epoch": 3.2,
-      "learning_rate": 1.5121591497022275e-05,
-      "loss": 0.2594,
+      "epoch": 3.2812556312394134,
+      "grad_norm": 0.2747882604598999,
+      "learning_rate": 1.3974243565580907e-05,
+      "loss": 0.355,
       "step": 91045
     },
     {
-      "epoch": 3.2,
-      "learning_rate": 1.5118974661704465e-05,
-      "loss": 0.2513,
+      "epoch": 3.2814358309006377,
+      "grad_norm": 0.2811215817928314,
+      "learning_rate": 1.3971624638405867e-05,
+      "loss": 0.3675,
       "step": 91050
     },
     {
-      "epoch": 3.2,
-      "learning_rate": 1.5116357954685257e-05,
-      "loss": 0.2565,
+      "epoch": 3.2816160305618625,
+      "grad_norm": 0.3407902121543884,
+      "learning_rate": 1.3969005861490305e-05,
+      "loss": 0.429,
       "step": 91055
     },
     {
-      "epoch": 3.2,
-      "learning_rate": 1.511374137599863e-05,
-      "loss": 0.2841,
+      "epoch": 3.281796230223087,
+      "grad_norm": 0.1697690635919571,
+      "learning_rate": 1.3966387234869904e-05,
+      "loss": 0.3446,
       "step": 91060
     },
     {
-      "epoch": 3.2,
-      "learning_rate": 1.5111124925678558e-05,
-      "loss": 0.2329,
+      "epoch": 3.281976429884312,
+      "grad_norm": 0.20448936522006989,
+      "learning_rate": 1.3963768758580342e-05,
+      "loss": 0.3826,
       "step": 91065
     },
     {
-      "epoch": 3.2,
-      "learning_rate": 1.5108508603759008e-05,
-      "loss": 0.2446,
+      "epoch": 3.282156629545536,
+      "grad_norm": 0.2799994647502899,
+      "learning_rate": 1.3961150432657283e-05,
+      "loss": 0.3621,
       "step": 91070
     },
     {
-      "epoch": 3.2,
-      "learning_rate": 1.5105892410273947e-05,
-      "loss": 0.2464,
+      "epoch": 3.282336829206761,
+      "grad_norm": 0.21096967160701752,
+      "learning_rate": 1.395853225713642e-05,
+      "loss": 0.3655,
       "step": 91075
     },
     {
-      "epoch": 3.2,
-      "learning_rate": 1.5103276345257368e-05,
-      "loss": 0.2742,
+      "epoch": 3.2825170288679857,
+      "grad_norm": 0.23864832520484924,
+      "learning_rate": 1.395591423205343e-05,
+      "loss": 0.3979,
       "step": 91080
     },
     {
-      "epoch": 3.2,
-      "learning_rate": 1.5100660408743218e-05,
-      "loss": 0.2279,
+      "epoch": 3.2826972285292104,
+      "grad_norm": 0.2114400416612625,
+      "learning_rate": 1.3953296357443946e-05,
+      "loss": 0.4144,
       "step": 91085
     },
     {
-      "epoch": 3.2,
-      "learning_rate": 1.509804460076547e-05,
-      "loss": 0.252,
+      "epoch": 3.282877428190435,
+      "grad_norm": 0.2655703127384186,
+      "learning_rate": 1.3950678633343675e-05,
+      "loss": 0.3678,
       "step": 91090
     },
     {
-      "epoch": 3.2,
-      "learning_rate": 1.5095428921358079e-05,
-      "loss": 0.2564,
+      "epoch": 3.28305762785166,
+      "grad_norm": 0.24133309721946716,
+      "learning_rate": 1.3948061059788267e-05,
+      "loss": 0.3951,
       "step": 91095
     },
     {
-      "epoch": 3.21,
-      "learning_rate": 1.509281337055502e-05,
-      "loss": 0.2621,
+      "epoch": 3.283237827512884,
+      "grad_norm": 0.22033952176570892,
+      "learning_rate": 1.3945443636813376e-05,
+      "loss": 0.3699,
       "step": 91100
     },
     {
-      "epoch": 3.21,
-      "learning_rate": 1.5090197948390255e-05,
-      "loss": 0.2564,
+      "epoch": 3.283418027174109,
+      "grad_norm": 0.23784691095352173,
+      "learning_rate": 1.3942826364454697e-05,
+      "loss": 0.4171,
       "step": 91105
     },
     {
-      "epoch": 3.21,
-      "learning_rate": 1.5087582654897738e-05,
-      "loss": 0.2645,
+      "epoch": 3.2835982268353336,
+      "grad_norm": 0.17935135960578918,
+      "learning_rate": 1.3940209242747847e-05,
+      "loss": 0.3836,
       "step": 91110
     },
     {
-      "epoch": 3.21,
-      "learning_rate": 1.5084967490111419e-05,
-      "loss": 0.265,
+      "epoch": 3.2837784264965584,
+      "grad_norm": 0.2037777453660965,
+      "learning_rate": 1.3937592271728517e-05,
+      "loss": 0.3883,
       "step": 91115
     },
     {
-      "epoch": 3.21,
-      "learning_rate": 1.5082352454065274e-05,
-      "loss": 0.2581,
+      "epoch": 3.2839586261577827,
+      "grad_norm": 0.24445310235023499,
+      "learning_rate": 1.3934975451432346e-05,
+      "loss": 0.3805,
       "step": 91120
     },
     {
-      "epoch": 3.21,
-      "learning_rate": 1.5079737546793237e-05,
-      "loss": 0.2452,
+      "epoch": 3.2841388258190074,
+      "grad_norm": 0.24390850961208344,
+      "learning_rate": 1.3932358781894996e-05,
+      "loss": 0.3939,
       "step": 91125
     },
     {
-      "epoch": 3.21,
-      "learning_rate": 1.5077122768329282e-05,
-      "loss": 0.2548,
+      "epoch": 3.284319025480232,
+      "grad_norm": 0.19269582629203796,
+      "learning_rate": 1.3929742263152115e-05,
+      "loss": 0.3994,
       "step": 91130
     },
     {
-      "epoch": 3.21,
-      "learning_rate": 1.5074508118707334e-05,
-      "loss": 0.2566,
+      "epoch": 3.284499225141457,
+      "grad_norm": 0.2648945748806,
+      "learning_rate": 1.3927125895239352e-05,
+      "loss": 0.3578,
       "step": 91135
     },
     {
-      "epoch": 3.21,
-      "learning_rate": 1.5071893597961367e-05,
-      "loss": 0.2787,
+      "epoch": 3.2846794248026816,
+      "grad_norm": 0.18639974296092987,
+      "learning_rate": 1.3924509678192343e-05,
+      "loss": 0.3366,
       "step": 91140
     },
     {
-      "epoch": 3.21,
-      "learning_rate": 1.506927920612532e-05,
-      "loss": 0.2527,
+      "epoch": 3.284859624463906,
+      "grad_norm": 0.1827782690525055,
+      "learning_rate": 1.3921893612046757e-05,
+      "loss": 0.3588,
       "step": 91145
     },
     {
-      "epoch": 3.21,
-      "learning_rate": 1.5066664943233128e-05,
-      "loss": 0.2547,
+      "epoch": 3.2850398241251306,
+      "grad_norm": 0.20156405866146088,
+      "learning_rate": 1.3919277696838227e-05,
+      "loss": 0.3694,
       "step": 91150
     },
     {
-      "epoch": 3.21,
-      "learning_rate": 1.506405080931875e-05,
-      "loss": 0.236,
+      "epoch": 3.2852200237863554,
+      "grad_norm": 0.1734452098608017,
+      "learning_rate": 1.391666193260239e-05,
+      "loss": 0.3947,
       "step": 91155
     },
     {
-      "epoch": 3.21,
-      "learning_rate": 1.5061436804416129e-05,
-      "loss": 0.2815,
+      "epoch": 3.28540022344758,
+      "grad_norm": 0.2353818714618683,
+      "learning_rate": 1.3914046319374891e-05,
+      "loss": 0.3996,
       "step": 91160
     },
     {
-      "epoch": 3.21,
-      "learning_rate": 1.5058822928559202e-05,
-      "loss": 0.2583,
+      "epoch": 3.2855804231088044,
+      "grad_norm": 0.23739945888519287,
+      "learning_rate": 1.3911430857191351e-05,
+      "loss": 0.3798,
       "step": 91165
     },
     {
-      "epoch": 3.21,
-      "learning_rate": 1.5056209181781896e-05,
-      "loss": 0.2591,
+      "epoch": 3.285760622770029,
+      "grad_norm": 0.20489168167114258,
+      "learning_rate": 1.3908815546087434e-05,
+      "loss": 0.3993,
       "step": 91170
     },
     {
-      "epoch": 3.21,
-      "learning_rate": 1.5053595564118172e-05,
-      "loss": 0.2473,
+      "epoch": 3.285940822431254,
+      "grad_norm": 0.22435033321380615,
+      "learning_rate": 1.3906200386098753e-05,
+      "loss": 0.3806,
       "step": 91175
     },
     {
-      "epoch": 3.21,
-      "learning_rate": 1.5050982075601949e-05,
-      "loss": 0.2496,
+      "epoch": 3.2861210220924786,
+      "grad_norm": 0.227759450674057,
+      "learning_rate": 1.3903585377260947e-05,
+      "loss": 0.3604,
       "step": 91180
     },
     {
-      "epoch": 3.21,
-      "learning_rate": 1.5048368716267175e-05,
-      "loss": 0.2478,
+      "epoch": 3.2863012217537033,
+      "grad_norm": 0.23122897744178772,
+      "learning_rate": 1.390097051960964e-05,
+      "loss": 0.3973,
       "step": 91185
     },
     {
-      "epoch": 3.21,
-      "learning_rate": 1.5045755486147761e-05,
-      "loss": 0.2934,
+      "epoch": 3.2864814214149276,
+      "grad_norm": 0.2522181570529938,
+      "learning_rate": 1.3898355813180453e-05,
+      "loss": 0.3383,
       "step": 91190
     },
     {
-      "epoch": 3.21,
-      "learning_rate": 1.5043142385277664e-05,
-      "loss": 0.2542,
+      "epoch": 3.2866616210761523,
+      "grad_norm": 0.3048706650733948,
+      "learning_rate": 1.3895741258009038e-05,
+      "loss": 0.3729,
       "step": 91195
     },
     {
-      "epoch": 3.21,
-      "learning_rate": 1.5040529413690802e-05,
-      "loss": 0.2574,
+      "epoch": 3.286841820737377,
+      "grad_norm": 0.20635949075222015,
+      "learning_rate": 1.3893126854130985e-05,
+      "loss": 0.3899,
       "step": 91200
     },
     {
-      "epoch": 3.21,
-      "learning_rate": 1.5037916571421094e-05,
-      "loss": 0.2496,
+      "epoch": 3.287022020398602,
+      "grad_norm": 0.2326965183019638,
+      "learning_rate": 1.3890512601581923e-05,
+      "loss": 0.3528,
       "step": 91205
     },
     {
-      "epoch": 3.21,
-      "learning_rate": 1.5035303858502474e-05,
-      "loss": 0.2459,
+      "epoch": 3.287202220059826,
+      "grad_norm": 0.22335182130336761,
+      "learning_rate": 1.3887898500397484e-05,
+      "loss": 0.3865,
       "step": 91210
     },
     {
-      "epoch": 3.21,
-      "learning_rate": 1.5032691274968868e-05,
-      "loss": 0.2913,
+      "epoch": 3.287382419721051,
+      "grad_norm": 0.2084750384092331,
+      "learning_rate": 1.3885284550613264e-05,
+      "loss": 0.3724,
       "step": 91215
     },
     {
-      "epoch": 3.21,
-      "learning_rate": 1.5030078820854205e-05,
-      "loss": 0.28,
+      "epoch": 3.2875626193822756,
+      "grad_norm": 0.18253043293952942,
+      "learning_rate": 1.3882670752264915e-05,
+      "loss": 0.3941,
       "step": 91220
     },
     {
-      "epoch": 3.21,
-      "learning_rate": 1.5027466496192393e-05,
-      "loss": 0.2536,
+      "epoch": 3.2877428190435003,
+      "grad_norm": 0.20223382115364075,
+      "learning_rate": 1.388005710538801e-05,
+      "loss": 0.3569,
       "step": 91225
     },
     {
-      "epoch": 3.21,
-      "learning_rate": 1.5024854301017349e-05,
-      "loss": 0.2663,
+      "epoch": 3.287923018704725,
+      "grad_norm": 0.20182040333747864,
+      "learning_rate": 1.3877443610018168e-05,
+      "loss": 0.3787,
       "step": 91230
     },
     {
-      "epoch": 3.21,
-      "learning_rate": 1.5022242235363004e-05,
-      "loss": 0.2726,
+      "epoch": 3.2881032183659493,
+      "grad_norm": 0.2429923266172409,
+      "learning_rate": 1.3874830266191014e-05,
+      "loss": 0.3608,
       "step": 91235
     },
     {
-      "epoch": 3.21,
-      "learning_rate": 1.501963029926327e-05,
-      "loss": 0.2587,
+      "epoch": 3.288283418027174,
+      "grad_norm": 0.23385964334011078,
+      "learning_rate": 1.387221707394214e-05,
+      "loss": 0.3892,
       "step": 91240
     },
     {
-      "epoch": 3.21,
-      "learning_rate": 1.501701849275206e-05,
-      "loss": 0.2533,
+      "epoch": 3.288463617688399,
+      "grad_norm": 0.20496410131454468,
+      "learning_rate": 1.3869604033307154e-05,
+      "loss": 0.3716,
       "step": 91245
     },
     {
-      "epoch": 3.21,
-      "learning_rate": 1.501440681586328e-05,
-      "loss": 0.2719,
+      "epoch": 3.2886438173496235,
+      "grad_norm": 0.20968787372112274,
+      "learning_rate": 1.3866991144321661e-05,
+      "loss": 0.3497,
       "step": 91250
     },
     {
-      "epoch": 3.21,
-      "learning_rate": 1.5011795268630852e-05,
-      "loss": 0.2614,
+      "epoch": 3.288824017010848,
+      "grad_norm": 0.2523500323295593,
+      "learning_rate": 1.3864378407021244e-05,
+      "loss": 0.3817,
       "step": 91255
     },
     {
-      "epoch": 3.21,
-      "learning_rate": 1.5009183851088677e-05,
-      "loss": 0.2336,
+      "epoch": 3.2890042166720725,
+      "grad_norm": 0.1904435008764267,
+      "learning_rate": 1.386176582144153e-05,
+      "loss": 0.3555,
       "step": 91260
     },
     {
-      "epoch": 3.21,
-      "learning_rate": 1.5006572563270665e-05,
-      "loss": 0.26,
+      "epoch": 3.2891844163332973,
+      "grad_norm": 0.2404657006263733,
+      "learning_rate": 1.3859153387618098e-05,
+      "loss": 0.3649,
       "step": 91265
     },
     {
-      "epoch": 3.21,
-      "learning_rate": 1.5003961405210723e-05,
-      "loss": 0.2684,
+      "epoch": 3.289364615994522,
+      "grad_norm": 0.23486708104610443,
+      "learning_rate": 1.3856541105586545e-05,
+      "loss": 0.3715,
       "step": 91270
     },
     {
-      "epoch": 3.21,
-      "learning_rate": 1.5001350376942758e-05,
-      "loss": 0.2504,
+      "epoch": 3.2895448156557467,
+      "grad_norm": 0.20871885120868683,
+      "learning_rate": 1.3853928975382464e-05,
+      "loss": 0.3904,
       "step": 91275
     },
     {
-      "epoch": 3.21,
-      "learning_rate": 1.499873947850067e-05,
-      "loss": 0.2491,
+      "epoch": 3.289725015316971,
+      "grad_norm": 0.23163028061389923,
+      "learning_rate": 1.3851316997041438e-05,
+      "loss": 0.3893,
       "step": 91280
     },
     {
-      "epoch": 3.21,
-      "learning_rate": 1.499612870991835e-05,
-      "loss": 0.2461,
+      "epoch": 3.2899052149781958,
+      "grad_norm": 0.32008373737335205,
+      "learning_rate": 1.3848705170599053e-05,
+      "loss": 0.3878,
       "step": 91285
     },
     {
-      "epoch": 3.21,
-      "learning_rate": 1.4993518071229712e-05,
-      "loss": 0.2483,
+      "epoch": 3.2900854146394205,
+      "grad_norm": 0.26068592071533203,
+      "learning_rate": 1.3846093496090918e-05,
+      "loss": 0.3768,
       "step": 91290
     },
     {
-      "epoch": 3.21,
-      "learning_rate": 1.4990907562468647e-05,
-      "loss": 0.2515,
+      "epoch": 3.2902656143006452,
+      "grad_norm": 0.2309846431016922,
+      "learning_rate": 1.384348197355258e-05,
+      "loss": 0.4094,
       "step": 91295
     },
     {
-      "epoch": 3.21,
-      "learning_rate": 1.4988297183669054e-05,
-      "loss": 0.2458,
+      "epoch": 3.2904458139618695,
+      "grad_norm": 0.250918447971344,
+      "learning_rate": 1.3840870603019655e-05,
+      "loss": 0.3681,
       "step": 91300
     },
     {
-      "epoch": 3.21,
-      "learning_rate": 1.4985686934864812e-05,
-      "loss": 0.2697,
+      "epoch": 3.2906260136230943,
+      "grad_norm": 0.22328318655490875,
+      "learning_rate": 1.3838259384527702e-05,
+      "loss": 0.3665,
       "step": 91305
     },
     {
-      "epoch": 3.21,
-      "learning_rate": 1.498307681608984e-05,
-      "loss": 0.2606,
+      "epoch": 3.290806213284319,
+      "grad_norm": 0.25075024366378784,
+      "learning_rate": 1.3835648318112307e-05,
+      "loss": 0.3973,
       "step": 91310
     },
     {
-      "epoch": 3.21,
-      "learning_rate": 1.4980466827378009e-05,
-      "loss": 0.2456,
+      "epoch": 3.2909864129455437,
+      "grad_norm": 0.20350109040737152,
+      "learning_rate": 1.383303740380904e-05,
+      "loss": 0.3423,
       "step": 91315
     },
     {
-      "epoch": 3.21,
-      "learning_rate": 1.4977856968763212e-05,
-      "loss": 0.2617,
+      "epoch": 3.2911666126067685,
+      "grad_norm": 0.2606266438961029,
+      "learning_rate": 1.383042664165348e-05,
+      "loss": 0.3894,
       "step": 91320
     },
     {
-      "epoch": 3.21,
-      "learning_rate": 1.4975247240279332e-05,
-      "loss": 0.2436,
+      "epoch": 3.2913468122679927,
+      "grad_norm": 0.20949092507362366,
+      "learning_rate": 1.3827816031681182e-05,
+      "loss": 0.4185,
       "step": 91325
     },
     {
-      "epoch": 3.21,
-      "learning_rate": 1.497263764196027e-05,
-      "loss": 0.2471,
+      "epoch": 3.2915270119292175,
+      "grad_norm": 0.23206953704357147,
+      "learning_rate": 1.3825205573927736e-05,
+      "loss": 0.3802,
       "step": 91330
     },
     {
-      "epoch": 3.21,
-      "learning_rate": 1.4970028173839901e-05,
-      "loss": 0.2406,
+      "epoch": 3.291707211590442,
+      "grad_norm": 0.21940544247627258,
+      "learning_rate": 1.3822595268428703e-05,
+      "loss": 0.3655,
       "step": 91335
     },
     {
-      "epoch": 3.21,
-      "learning_rate": 1.4967418835952095e-05,
-      "loss": 0.2429,
+      "epoch": 3.291887411251667,
+      "grad_norm": 0.1804596483707428,
+      "learning_rate": 1.3819985115219644e-05,
+      "loss": 0.3701,
       "step": 91340
     },
     {
-      "epoch": 3.21,
-      "learning_rate": 1.496480962833075e-05,
-      "loss": 0.2761,
+      "epoch": 3.2920676109128912,
+      "grad_norm": 0.22188700735569,
+      "learning_rate": 1.3817375114336123e-05,
+      "loss": 0.3908,
       "step": 91345
     },
     {
-      "epoch": 3.21,
-      "learning_rate": 1.4962200551009735e-05,
-      "loss": 0.2409,
+      "epoch": 3.292247810574116,
+      "grad_norm": 0.2676689326763153,
+      "learning_rate": 1.3814765265813695e-05,
+      "loss": 0.4065,
       "step": 91350
     },
     {
-      "epoch": 3.21,
-      "learning_rate": 1.4959591604022937e-05,
-      "loss": 0.2541,
+      "epoch": 3.2924280102353407,
+      "grad_norm": 0.20868746936321259,
+      "learning_rate": 1.381215556968793e-05,
+      "loss": 0.3471,
       "step": 91355
     },
     {
-      "epoch": 3.21,
-      "learning_rate": 1.4956982787404228e-05,
-      "loss": 0.2431,
+      "epoch": 3.2926082098965654,
+      "grad_norm": 0.21724528074264526,
+      "learning_rate": 1.380954602599438e-05,
+      "loss": 0.382,
       "step": 91360
     },
     {
-      "epoch": 3.21,
-      "learning_rate": 1.4954374101187465e-05,
-      "loss": 0.241,
+      "epoch": 3.29278840955779,
+      "grad_norm": 0.21302174031734467,
+      "learning_rate": 1.3806936634768603e-05,
+      "loss": 0.37,
       "step": 91365
     },
     {
-      "epoch": 3.21,
-      "learning_rate": 1.4951765545406547e-05,
-      "loss": 0.2682,
+      "epoch": 3.292968609219015,
+      "grad_norm": 0.21098196506500244,
+      "learning_rate": 1.3804327396046143e-05,
+      "loss": 0.3938,
       "step": 91370
     },
     {
-      "epoch": 3.21,
-      "learning_rate": 1.4949157120095324e-05,
-      "loss": 0.2601,
+      "epoch": 3.293148808880239,
+      "grad_norm": 0.20903551578521729,
+      "learning_rate": 1.3801718309862546e-05,
+      "loss": 0.3812,
       "step": 91375
     },
     {
-      "epoch": 3.22,
-      "learning_rate": 1.4946548825287668e-05,
-      "loss": 0.2532,
+      "epoch": 3.293329008541464,
+      "grad_norm": 0.2137085646390915,
+      "learning_rate": 1.3799109376253378e-05,
+      "loss": 0.3976,
       "step": 91380
     },
     {
-      "epoch": 3.22,
-      "learning_rate": 1.4943940661017458e-05,
-      "loss": 0.2727,
+      "epoch": 3.2935092082026887,
+      "grad_norm": 0.2167220115661621,
+      "learning_rate": 1.3796500595254187e-05,
+      "loss": 0.3906,
       "step": 91385
     },
     {
-      "epoch": 3.22,
-      "learning_rate": 1.4941332627318555e-05,
-      "loss": 0.2476,
+      "epoch": 3.2936894078639134,
+      "grad_norm": 0.19560924172401428,
+      "learning_rate": 1.3793891966900488e-05,
+      "loss": 0.3904,
       "step": 91390
     },
     {
-      "epoch": 3.22,
-      "learning_rate": 1.4938724724224817e-05,
-      "loss": 0.256,
+      "epoch": 3.2938696075251377,
+      "grad_norm": 0.24281176924705505,
+      "learning_rate": 1.379128349122785e-05,
+      "loss": 0.3647,
       "step": 91395
     },
     {
-      "epoch": 3.22,
-      "learning_rate": 1.4936116951770095e-05,
-      "loss": 0.2726,
+      "epoch": 3.2940498071863624,
+      "grad_norm": 0.2524743676185608,
+      "learning_rate": 1.3788675168271791e-05,
+      "loss": 0.3615,
       "step": 91400
     },
     {
-      "epoch": 3.22,
-      "learning_rate": 1.4933509309988271e-05,
-      "loss": 0.2599,
+      "epoch": 3.294230006847587,
+      "grad_norm": 0.2531737983226776,
+      "learning_rate": 1.3786066998067887e-05,
+      "loss": 0.3995,
       "step": 91405
     },
     {
-      "epoch": 3.22,
-      "learning_rate": 1.49309017989132e-05,
-      "loss": 0.2461,
+      "epoch": 3.294410206508812,
+      "grad_norm": 0.2063448578119278,
+      "learning_rate": 1.3783458980651637e-05,
+      "loss": 0.3758,
       "step": 91410
     },
     {
-      "epoch": 3.22,
-      "learning_rate": 1.4928294418578732e-05,
-      "loss": 0.2535,
+      "epoch": 3.2945904061700366,
+      "grad_norm": 0.23044107854366302,
+      "learning_rate": 1.3780851116058579e-05,
+      "loss": 0.3673,
       "step": 91415
     },
     {
-      "epoch": 3.22,
-      "learning_rate": 1.4925687169018712e-05,
-      "loss": 0.2698,
+      "epoch": 3.294770605831261,
+      "grad_norm": 0.23507073521614075,
+      "learning_rate": 1.377824340432426e-05,
+      "loss": 0.3559,
       "step": 91420
     },
     {
-      "epoch": 3.22,
-      "learning_rate": 1.492308005026702e-05,
-      "loss": 0.2688,
+      "epoch": 3.2949508054924856,
+      "grad_norm": 0.22942350804805756,
+      "learning_rate": 1.3775635845484203e-05,
+      "loss": 0.3899,
       "step": 91425
     },
     {
-      "epoch": 3.22,
-      "learning_rate": 1.4920473062357481e-05,
-      "loss": 0.2572,
+      "epoch": 3.2951310051537104,
+      "grad_norm": 0.282632976770401,
+      "learning_rate": 1.3773028439573935e-05,
+      "loss": 0.3621,
       "step": 91430
     },
     {
-      "epoch": 3.22,
-      "learning_rate": 1.4917866205323966e-05,
-      "loss": 0.233,
+      "epoch": 3.295311204814935,
+      "grad_norm": 0.2871980369091034,
+      "learning_rate": 1.3770421186628979e-05,
+      "loss": 0.3672,
       "step": 91435
     },
     {
-      "epoch": 3.22,
-      "learning_rate": 1.49152594792003e-05,
-      "loss": 0.2626,
+      "epoch": 3.2954914044761594,
+      "grad_norm": 0.27828335762023926,
+      "learning_rate": 1.3767814086684853e-05,
+      "loss": 0.4216,
       "step": 91440
     },
     {
-      "epoch": 3.22,
-      "learning_rate": 1.4912652884020356e-05,
-      "loss": 0.2667,
+      "epoch": 3.295671604137384,
+      "grad_norm": 0.268162339925766,
+      "learning_rate": 1.3765207139777091e-05,
+      "loss": 0.4116,
       "step": 91445
     },
     {
-      "epoch": 3.22,
-      "learning_rate": 1.4910046419817966e-05,
-      "loss": 0.2822,
+      "epoch": 3.295851803798609,
+      "grad_norm": 0.20781943202018738,
+      "learning_rate": 1.3762600345941213e-05,
+      "loss": 0.3331,
       "step": 91450
     },
     {
-      "epoch": 3.22,
-      "learning_rate": 1.490744008662697e-05,
-      "loss": 0.2621,
+      "epoch": 3.2960320034598336,
+      "grad_norm": 0.3062985837459564,
+      "learning_rate": 1.3759993705212726e-05,
+      "loss": 0.392,
       "step": 91455
     },
     {
-      "epoch": 3.22,
-      "learning_rate": 1.4904833884481206e-05,
-      "loss": 0.2532,
+      "epoch": 3.2962122031210583,
+      "grad_norm": 0.1981145143508911,
+      "learning_rate": 1.3757387217627146e-05,
+      "loss": 0.3496,
       "step": 91460
     },
     {
-      "epoch": 3.22,
-      "learning_rate": 1.4902227813414531e-05,
-      "loss": 0.2704,
+      "epoch": 3.2963924027822826,
+      "grad_norm": 0.2891653776168823,
+      "learning_rate": 1.3754780883219992e-05,
+      "loss": 0.4288,
       "step": 91465
     },
     {
-      "epoch": 3.22,
-      "learning_rate": 1.4899621873460778e-05,
-      "loss": 0.2505,
+      "epoch": 3.2965726024435074,
+      "grad_norm": 0.23202994465827942,
+      "learning_rate": 1.375217470202676e-05,
+      "loss": 0.4005,
       "step": 91470
     },
     {
-      "epoch": 3.22,
-      "learning_rate": 1.4897016064653774e-05,
-      "loss": 0.2427,
+      "epoch": 3.296752802104732,
+      "grad_norm": 0.26071614027023315,
+      "learning_rate": 1.374956867408299e-05,
+      "loss": 0.4132,
       "step": 91475
     },
     {
-      "epoch": 3.22,
-      "learning_rate": 1.489441038702735e-05,
-      "loss": 0.2428,
+      "epoch": 3.296933001765957,
+      "grad_norm": 0.22965455055236816,
+      "learning_rate": 1.3746962799424151e-05,
+      "loss": 0.3659,
       "step": 91480
     },
     {
-      "epoch": 3.22,
-      "learning_rate": 1.4891804840615353e-05,
-      "loss": 0.2644,
+      "epoch": 3.297113201427181,
+      "grad_norm": 0.19163905084133148,
+      "learning_rate": 1.3744357078085773e-05,
+      "loss": 0.3467,
       "step": 91485
     },
     {
-      "epoch": 3.22,
-      "learning_rate": 1.488919942545161e-05,
-      "loss": 0.2517,
+      "epoch": 3.297293401088406,
+      "grad_norm": 0.21488019824028015,
+      "learning_rate": 1.3741751510103352e-05,
+      "loss": 0.3741,
       "step": 91490
     },
     {
-      "epoch": 3.22,
-      "learning_rate": 1.4886594141569944e-05,
-      "loss": 0.2629,
+      "epoch": 3.2974736007496306,
+      "grad_norm": 0.19963303208351135,
+      "learning_rate": 1.3739146095512383e-05,
+      "loss": 0.3876,
       "step": 91495
     },
     {
-      "epoch": 3.22,
-      "learning_rate": 1.4883988989004194e-05,
-      "loss": 0.2442,
+      "epoch": 3.2976538004108553,
+      "grad_norm": 0.18200761079788208,
+      "learning_rate": 1.3736540834348372e-05,
+      "loss": 0.3782,
       "step": 91500
     },
     {
-      "epoch": 3.22,
-      "eval_loss": 0.25463610887527466,
-      "eval_runtime": 10.5511,
-      "eval_samples_per_second": 9.478,
-      "eval_steps_per_second": 9.478,
+      "epoch": 3.2976538004108553,
+      "eval_loss": 0.430542528629303,
+      "eval_runtime": 3.5303,
+      "eval_samples_per_second": 28.326,
+      "eval_steps_per_second": 7.081,
       "step": 91500
     },
     {
-      "epoch": 3.22,
-      "learning_rate": 1.4881383967788181e-05,
-      "loss": 0.2399,
+      "epoch": 3.29783400007208,
+      "grad_norm": 0.2648855149745941,
+      "learning_rate": 1.3733935726646812e-05,
+      "loss": 0.3551,
       "step": 91505
     },
     {
-      "epoch": 3.22,
-      "learning_rate": 1.4878779077955723e-05,
-      "loss": 0.2549,
+      "epoch": 3.2980141997333043,
+      "grad_norm": 0.26277956366539,
+      "learning_rate": 1.3731330772443186e-05,
+      "loss": 0.4238,
       "step": 91510
     },
     {
-      "epoch": 3.22,
-      "learning_rate": 1.4876174319540647e-05,
-      "loss": 0.2428,
+      "epoch": 3.298194399394529,
+      "grad_norm": 0.2899339199066162,
+      "learning_rate": 1.3728725971773004e-05,
+      "loss": 0.3986,
       "step": 91515
     },
     {
-      "epoch": 3.22,
-      "learning_rate": 1.4873569692576784e-05,
-      "loss": 0.2622,
+      "epoch": 3.298374599055754,
+      "grad_norm": 0.23275549709796906,
+      "learning_rate": 1.372612132467175e-05,
+      "loss": 0.3944,
       "step": 91520
     },
     {
-      "epoch": 3.22,
-      "learning_rate": 1.4870965197097944e-05,
-      "loss": 0.2583,
+      "epoch": 3.2985547987169785,
+      "grad_norm": 0.3062422573566437,
+      "learning_rate": 1.3723516831174912e-05,
+      "loss": 0.4204,
       "step": 91525
     },
     {
-      "epoch": 3.22,
-      "learning_rate": 1.4868360833137945e-05,
-      "loss": 0.2568,
+      "epoch": 3.298734998378203,
+      "grad_norm": 0.17250539362430573,
+      "learning_rate": 1.3720912491317972e-05,
+      "loss": 0.3552,
       "step": 91530
     },
     {
-      "epoch": 3.22,
-      "learning_rate": 1.4865756600730597e-05,
-      "loss": 0.2574,
+      "epoch": 3.2989151980394276,
+      "grad_norm": 0.17668567597866058,
+      "learning_rate": 1.3718308305136407e-05,
+      "loss": 0.3811,
       "step": 91535
     },
     {
-      "epoch": 3.22,
-      "learning_rate": 1.4863152499909722e-05,
-      "loss": 0.2483,
+      "epoch": 3.2990953977006523,
+      "grad_norm": 0.267254501581192,
+      "learning_rate": 1.3715704272665717e-05,
+      "loss": 0.3756,
       "step": 91540
     },
     {
-      "epoch": 3.22,
-      "learning_rate": 1.4861069314018015e-05,
-      "loss": 0.2347,
+      "epoch": 3.299275597361877,
+      "grad_norm": 0.23662297427654266,
+      "learning_rate": 1.3713100393941369e-05,
+      "loss": 0.366,
       "step": 91545
     },
     {
-      "epoch": 3.22,
-      "learning_rate": 1.4858465450137993e-05,
-      "loss": 0.2609,
+      "epoch": 3.2994557970231018,
+      "grad_norm": 0.21058867871761322,
+      "learning_rate": 1.3710496668998845e-05,
+      "loss": 0.3798,
       "step": 91550
     },
     {
-      "epoch": 3.22,
-      "learning_rate": 1.485586171793912e-05,
-      "loss": 0.2563,
+      "epoch": 3.299635996684326,
+      "grad_norm": 0.22281815111637115,
+      "learning_rate": 1.3707893097873619e-05,
+      "loss": 0.3874,
       "step": 91555
     },
     {
-      "epoch": 3.22,
-      "learning_rate": 1.4853258117455202e-05,
-      "loss": 0.2392,
+      "epoch": 3.299816196345551,
+      "grad_norm": 0.29249441623687744,
+      "learning_rate": 1.3705289680601152e-05,
+      "loss": 0.3828,
       "step": 91560
     },
     {
-      "epoch": 3.22,
-      "learning_rate": 1.4850654648720031e-05,
-      "loss": 0.2914,
+      "epoch": 3.2999963960067755,
+      "grad_norm": 0.2529570460319519,
+      "learning_rate": 1.3702686417216937e-05,
+      "loss": 0.3894,
       "step": 91565
     },
     {
-      "epoch": 3.22,
-      "learning_rate": 1.4848051311767424e-05,
-      "loss": 0.2669,
+      "epoch": 3.3001765956680003,
+      "grad_norm": 0.23126383125782013,
+      "learning_rate": 1.3700083307756443e-05,
+      "loss": 0.3985,
       "step": 91570
     },
     {
-      "epoch": 3.22,
-      "learning_rate": 1.4845448106631182e-05,
-      "loss": 0.2672,
+      "epoch": 3.3003567953292245,
+      "grad_norm": 0.21352465450763702,
+      "learning_rate": 1.3697480352255105e-05,
+      "loss": 0.3434,
       "step": 91575
     },
     {
-      "epoch": 3.22,
-      "learning_rate": 1.4842845033345106e-05,
-      "loss": 0.2848,
+      "epoch": 3.3005369949904493,
+      "grad_norm": 0.2859989404678345,
+      "learning_rate": 1.3694877550748419e-05,
+      "loss": 0.3767,
       "step": 91580
     },
     {
-      "epoch": 3.22,
-      "learning_rate": 1.4840242091942986e-05,
-      "loss": 0.2635,
+      "epoch": 3.300717194651674,
+      "grad_norm": 0.2089170217514038,
+      "learning_rate": 1.3692274903271828e-05,
+      "loss": 0.3966,
       "step": 91585
     },
     {
-      "epoch": 3.22,
-      "learning_rate": 1.4837639282458635e-05,
-      "loss": 0.2705,
+      "epoch": 3.3008973943128987,
+      "grad_norm": 0.24411603808403015,
+      "learning_rate": 1.368967240986082e-05,
+      "loss": 0.3803,
       "step": 91590
     },
     {
-      "epoch": 3.22,
-      "learning_rate": 1.483503660492584e-05,
-      "loss": 0.2306,
+      "epoch": 3.3010775939741235,
+      "grad_norm": 0.2391398698091507,
+      "learning_rate": 1.3687070070550829e-05,
+      "loss": 0.361,
       "step": 91595
     },
     {
-      "epoch": 3.22,
-      "learning_rate": 1.4832434059378392e-05,
-      "loss": 0.2859,
+      "epoch": 3.301257793635348,
+      "grad_norm": 0.23561078310012817,
+      "learning_rate": 1.3684467885377306e-05,
+      "loss": 0.373,
       "step": 91600
     },
     {
-      "epoch": 3.22,
-      "learning_rate": 1.4829831645850084e-05,
-      "loss": 0.2609,
+      "epoch": 3.3014379932965725,
+      "grad_norm": 0.20145799219608307,
+      "learning_rate": 1.3681865854375728e-05,
+      "loss": 0.3838,
       "step": 91605
     },
     {
-      "epoch": 3.22,
-      "learning_rate": 1.4827229364374718e-05,
-      "loss": 0.2488,
+      "epoch": 3.3016181929577972,
+      "grad_norm": 0.26969730854034424,
+      "learning_rate": 1.3679263977581536e-05,
+      "loss": 0.388,
       "step": 91610
     },
     {
-      "epoch": 3.22,
-      "learning_rate": 1.4824627214986072e-05,
-      "loss": 0.2563,
+      "epoch": 3.301798392619022,
+      "grad_norm": 0.28870394825935364,
+      "learning_rate": 1.3676662255030182e-05,
+      "loss": 0.4321,
       "step": 91615
     },
     {
-      "epoch": 3.22,
-      "learning_rate": 1.4822025197717937e-05,
-      "loss": 0.2649,
+      "epoch": 3.3019785922802467,
+      "grad_norm": 0.2489212453365326,
+      "learning_rate": 1.3674060686757113e-05,
+      "loss": 0.3873,
       "step": 91620
     },
     {
-      "epoch": 3.22,
-      "learning_rate": 1.4819423312604091e-05,
-      "loss": 0.2611,
+      "epoch": 3.302158791941471,
+      "grad_norm": 0.2643071115016937,
+      "learning_rate": 1.367145927279776e-05,
+      "loss": 0.3973,
       "step": 91625
     },
     {
-      "epoch": 3.22,
-      "learning_rate": 1.4816821559678324e-05,
-      "loss": 0.2774,
+      "epoch": 3.3023389916026957,
+      "grad_norm": 0.25098204612731934,
+      "learning_rate": 1.3668858013187597e-05,
+      "loss": 0.3781,
       "step": 91630
     },
     {
-      "epoch": 3.22,
-      "learning_rate": 1.4814219938974428e-05,
-      "loss": 0.258,
+      "epoch": 3.3025191912639205,
+      "grad_norm": 0.20655451714992523,
+      "learning_rate": 1.3666256907962044e-05,
+      "loss": 0.3913,
       "step": 91635
     },
     {
-      "epoch": 3.22,
-      "learning_rate": 1.481161845052617e-05,
-      "loss": 0.2518,
+      "epoch": 3.302699390925145,
+      "grad_norm": 0.20496463775634766,
+      "learning_rate": 1.366365595715655e-05,
+      "loss": 0.3778,
       "step": 91640
     },
     {
-      "epoch": 3.22,
-      "learning_rate": 1.4809017094367322e-05,
-      "loss": 0.2379,
+      "epoch": 3.30287959058637,
+      "grad_norm": 0.2476882040500641,
+      "learning_rate": 1.3661055160806546e-05,
+      "loss": 0.398,
       "step": 91645
     },
     {
-      "epoch": 3.22,
-      "learning_rate": 1.4806415870531681e-05,
-      "loss": 0.2669,
+      "epoch": 3.303059790247594,
+      "grad_norm": 0.20680633187294006,
+      "learning_rate": 1.3658454518947475e-05,
+      "loss": 0.353,
       "step": 91650
     },
     {
-      "epoch": 3.22,
-      "learning_rate": 1.480381477905301e-05,
-      "loss": 0.2635,
+      "epoch": 3.303239989908819,
+      "grad_norm": 0.21282508969306946,
+      "learning_rate": 1.3655854031614751e-05,
+      "loss": 0.3972,
       "step": 91655
     },
     {
-      "epoch": 3.22,
-      "learning_rate": 1.4801213819965087e-05,
-      "loss": 0.2717,
+      "epoch": 3.3034201895700437,
+      "grad_norm": 0.2705020010471344,
+      "learning_rate": 1.3653253698843844e-05,
+      "loss": 0.3935,
       "step": 91660
     },
     {
-      "epoch": 3.23,
-      "learning_rate": 1.479861299330167e-05,
-      "loss": 0.2847,
+      "epoch": 3.3036003892312684,
+      "grad_norm": 0.32260480523109436,
+      "learning_rate": 1.3650653520670134e-05,
+      "loss": 0.3799,
       "step": 91665
     },
     {
-      "epoch": 3.23,
-      "learning_rate": 1.4796012299096551e-05,
-      "loss": 0.2471,
+      "epoch": 3.3037805888924927,
+      "grad_norm": 0.2574411928653717,
+      "learning_rate": 1.3648053497129082e-05,
+      "loss": 0.4028,
       "step": 91670
     },
     {
-      "epoch": 3.23,
-      "learning_rate": 1.4793411737383488e-05,
-      "loss": 0.2323,
+      "epoch": 3.3039607885537174,
+      "grad_norm": 0.22409643232822418,
+      "learning_rate": 1.3645453628256105e-05,
+      "loss": 0.3615,
       "step": 91675
     },
     {
-      "epoch": 3.23,
-      "learning_rate": 1.479081130819624e-05,
-      "loss": 0.2785,
+      "epoch": 3.304140988214942,
+      "grad_norm": 0.2090173065662384,
+      "learning_rate": 1.3642853914086617e-05,
+      "loss": 0.373,
       "step": 91680
     },
     {
-      "epoch": 3.23,
-      "learning_rate": 1.4788211011568581e-05,
-      "loss": 0.2688,
+      "epoch": 3.304321187876167,
+      "grad_norm": 0.1917799711227417,
+      "learning_rate": 1.3640254354656062e-05,
+      "loss": 0.3696,
       "step": 91685
     },
     {
-      "epoch": 3.23,
-      "learning_rate": 1.478561084753428e-05,
-      "loss": 0.2751,
+      "epoch": 3.3045013875373916,
+      "grad_norm": 0.2533906400203705,
+      "learning_rate": 1.3637654949999833e-05,
+      "loss": 0.4081,
       "step": 91690
     },
     {
-      "epoch": 3.23,
-      "learning_rate": 1.4783010816127088e-05,
-      "loss": 0.2484,
+      "epoch": 3.304681587198616,
+      "grad_norm": 0.19297879934310913,
+      "learning_rate": 1.3635055700153346e-05,
+      "loss": 0.3785,
       "step": 91695
     },
     {
-      "epoch": 3.23,
-      "learning_rate": 1.4780410917380757e-05,
-      "loss": 0.2553,
+      "epoch": 3.3048617868598407,
+      "grad_norm": 0.20561909675598145,
+      "learning_rate": 1.3632456605152033e-05,
+      "loss": 0.3966,
       "step": 91700
     },
     {
-      "epoch": 3.23,
-      "learning_rate": 1.4777811151329068e-05,
-      "loss": 0.2591,
+      "epoch": 3.3050419865210654,
+      "grad_norm": 0.22225363552570343,
+      "learning_rate": 1.36298576650313e-05,
+      "loss": 0.4088,
       "step": 91705
     },
     {
-      "epoch": 3.23,
-      "learning_rate": 1.4775211518005756e-05,
-      "loss": 0.2395,
+      "epoch": 3.30522218618229,
+      "grad_norm": 0.23833619058132172,
+      "learning_rate": 1.3627258879826554e-05,
+      "loss": 0.3973,
       "step": 91710
     },
     {
-      "epoch": 3.23,
-      "learning_rate": 1.4772612017444592e-05,
-      "loss": 0.2454,
+      "epoch": 3.3054023858435144,
+      "grad_norm": 0.23555156588554382,
+      "learning_rate": 1.3624660249573207e-05,
+      "loss": 0.3813,
       "step": 91715
     },
     {
-      "epoch": 3.23,
-      "learning_rate": 1.477001264967931e-05,
-      "loss": 0.2699,
+      "epoch": 3.305582585504739,
+      "grad_norm": 0.21567942202091217,
+      "learning_rate": 1.3622061774306647e-05,
+      "loss": 0.381,
       "step": 91720
     },
     {
-      "epoch": 3.23,
-      "learning_rate": 1.476741341474368e-05,
-      "loss": 0.269,
+      "epoch": 3.305762785165964,
+      "grad_norm": 0.2811535894870758,
+      "learning_rate": 1.3619463454062304e-05,
+      "loss": 0.3864,
       "step": 91725
     },
     {
-      "epoch": 3.23,
-      "learning_rate": 1.4764814312671446e-05,
-      "loss": 0.2733,
+      "epoch": 3.3059429848271886,
+      "grad_norm": 0.23214657604694366,
+      "learning_rate": 1.3616865288875568e-05,
+      "loss": 0.3726,
       "step": 91730
     },
     {
-      "epoch": 3.23,
-      "learning_rate": 1.4762215343496346e-05,
-      "loss": 0.2514,
+      "epoch": 3.3061231844884134,
+      "grad_norm": 0.2762070596218109,
+      "learning_rate": 1.3614267278781839e-05,
+      "loss": 0.3761,
       "step": 91735
     },
     {
-      "epoch": 3.23,
-      "learning_rate": 1.4759616507252127e-05,
-      "loss": 0.2639,
+      "epoch": 3.3063033841496376,
+      "grad_norm": 0.21005326509475708,
+      "learning_rate": 1.3611669423816514e-05,
+      "loss": 0.3974,
       "step": 91740
     },
     {
-      "epoch": 3.23,
-      "learning_rate": 1.4757017803972551e-05,
-      "loss": 0.2544,
+      "epoch": 3.3064835838108624,
+      "grad_norm": 0.2921294569969177,
+      "learning_rate": 1.360907172401498e-05,
+      "loss": 0.3958,
       "step": 91745
     },
     {
-      "epoch": 3.23,
-      "learning_rate": 1.475441923369135e-05,
-      "loss": 0.2346,
+      "epoch": 3.306663783472087,
+      "grad_norm": 0.24011696875095367,
+      "learning_rate": 1.3606474179412645e-05,
+      "loss": 0.3846,
       "step": 91750
     },
     {
-      "epoch": 3.23,
-      "learning_rate": 1.4751820796442261e-05,
-      "loss": 0.2813,
+      "epoch": 3.306843983133312,
+      "grad_norm": 0.2656587064266205,
+      "learning_rate": 1.3603876790044906e-05,
+      "loss": 0.3611,
       "step": 91755
     },
     {
-      "epoch": 3.23,
-      "learning_rate": 1.4749222492259013e-05,
-      "loss": 0.2688,
+      "epoch": 3.307024182794536,
+      "grad_norm": 0.2405286282300949,
+      "learning_rate": 1.360127955594712e-05,
+      "loss": 0.3816,
       "step": 91760
     },
     {
-      "epoch": 3.23,
-      "learning_rate": 1.4746624321175365e-05,
-      "loss": 0.2435,
+      "epoch": 3.307204382455761,
+      "grad_norm": 0.23016515374183655,
+      "learning_rate": 1.3598682477154701e-05,
+      "loss": 0.3776,
       "step": 91765
     },
     {
-      "epoch": 3.23,
-      "learning_rate": 1.4744026283225045e-05,
-      "loss": 0.265,
+      "epoch": 3.3073845821169856,
+      "grad_norm": 0.24398526549339294,
+      "learning_rate": 1.3596085553703014e-05,
+      "loss": 0.3715,
       "step": 91770
     },
     {
-      "epoch": 3.23,
-      "learning_rate": 1.4741428378441786e-05,
-      "loss": 0.2716,
+      "epoch": 3.3075647817782103,
+      "grad_norm": 0.25272274017333984,
+      "learning_rate": 1.3593488785627478e-05,
+      "loss": 0.3561,
       "step": 91775
     },
     {
-      "epoch": 3.23,
-      "learning_rate": 1.4738830606859309e-05,
-      "loss": 0.2371,
+      "epoch": 3.307744981439435,
+      "grad_norm": 0.22344271838665009,
+      "learning_rate": 1.3590892172963437e-05,
+      "loss": 0.3864,
       "step": 91780
     },
     {
-      "epoch": 3.23,
-      "learning_rate": 1.4736232968511362e-05,
-      "loss": 0.2622,
+      "epoch": 3.3079251811006594,
+      "grad_norm": 0.2128763496875763,
+      "learning_rate": 1.3588295715746272e-05,
+      "loss": 0.3617,
       "step": 91785
     },
     {
-      "epoch": 3.23,
-      "learning_rate": 1.4733635463431661e-05,
-      "loss": 0.2683,
+      "epoch": 3.308105380761884,
+      "grad_norm": 0.24568325281143188,
+      "learning_rate": 1.3585699414011376e-05,
+      "loss": 0.3945,
       "step": 91790
     },
     {
-      "epoch": 3.23,
-      "learning_rate": 1.4731038091653943e-05,
-      "loss": 0.249,
+      "epoch": 3.308285580423109,
+      "grad_norm": 0.230985626578331,
+      "learning_rate": 1.3583103267794117e-05,
+      "loss": 0.3645,
       "step": 91795
     },
     {
-      "epoch": 3.23,
-      "learning_rate": 1.4728440853211923e-05,
-      "loss": 0.2776,
+      "epoch": 3.3084657800843336,
+      "grad_norm": 0.25532665848731995,
+      "learning_rate": 1.3580507277129867e-05,
+      "loss": 0.3828,
       "step": 91800
     },
     {
-      "epoch": 3.23,
-      "learning_rate": 1.4725843748139335e-05,
-      "loss": 0.2587,
+      "epoch": 3.308645979745558,
+      "grad_norm": 0.26051023602485657,
+      "learning_rate": 1.3577911442053992e-05,
+      "loss": 0.377,
       "step": 91805
     },
     {
-      "epoch": 3.23,
-      "learning_rate": 1.4723246776469896e-05,
-      "loss": 0.2436,
+      "epoch": 3.3088261794067826,
+      "grad_norm": 0.22641240060329437,
+      "learning_rate": 1.3575315762601853e-05,
+      "loss": 0.3672,
       "step": 91810
     },
     {
-      "epoch": 3.23,
-      "learning_rate": 1.4720649938237313e-05,
-      "loss": 0.267,
+      "epoch": 3.3090063790680073,
+      "grad_norm": 0.29088813066482544,
+      "learning_rate": 1.357272023880884e-05,
+      "loss": 0.3662,
       "step": 91815
     },
     {
-      "epoch": 3.23,
-      "learning_rate": 1.471805323347532e-05,
-      "loss": 0.2601,
+      "epoch": 3.309186578729232,
+      "grad_norm": 0.24642913043498993,
+      "learning_rate": 1.3570124870710293e-05,
+      "loss": 0.3872,
       "step": 91820
     },
     {
-      "epoch": 3.23,
-      "learning_rate": 1.4715456662217639e-05,
-      "loss": 0.2803,
+      "epoch": 3.309366778390457,
+      "grad_norm": 0.20131155848503113,
+      "learning_rate": 1.3567529658341587e-05,
+      "loss": 0.3784,
       "step": 91825
     },
     {
-      "epoch": 3.23,
-      "learning_rate": 1.4712860224497973e-05,
-      "loss": 0.2677,
+      "epoch": 3.309546978051681,
+      "grad_norm": 0.2340930700302124,
+      "learning_rate": 1.3564934601738075e-05,
+      "loss": 0.409,
       "step": 91830
     },
     {
-      "epoch": 3.23,
-      "learning_rate": 1.4710263920350028e-05,
-      "loss": 0.2623,
+      "epoch": 3.309727177712906,
+      "grad_norm": 0.22841165959835052,
+      "learning_rate": 1.3562339700935114e-05,
+      "loss": 0.3586,
       "step": 91835
     },
     {
-      "epoch": 3.23,
-      "learning_rate": 1.4707667749807535e-05,
-      "loss": 0.2359,
+      "epoch": 3.3099073773741305,
+      "grad_norm": 0.2462332546710968,
+      "learning_rate": 1.355974495596805e-05,
+      "loss": 0.4212,
       "step": 91840
     },
     {
-      "epoch": 3.23,
-      "learning_rate": 1.4705071712904189e-05,
-      "loss": 0.2593,
+      "epoch": 3.3100875770353553,
+      "grad_norm": 0.2860056459903717,
+      "learning_rate": 1.355715036687226e-05,
+      "loss": 0.3739,
       "step": 91845
     },
     {
-      "epoch": 3.23,
-      "learning_rate": 1.4702475809673707e-05,
-      "loss": 0.2581,
+      "epoch": 3.3102677766965796,
+      "grad_norm": 0.24574606120586395,
+      "learning_rate": 1.3554555933683077e-05,
+      "loss": 0.3489,
       "step": 91850
     },
     {
-      "epoch": 3.23,
-      "learning_rate": 1.469988004014978e-05,
-      "loss": 0.2673,
+      "epoch": 3.3104479763578043,
+      "grad_norm": 0.20963186025619507,
+      "learning_rate": 1.3551961656435852e-05,
+      "loss": 0.3463,
       "step": 91855
     },
     {
-      "epoch": 3.23,
-      "learning_rate": 1.4697284404366133e-05,
-      "loss": 0.2592,
+      "epoch": 3.310628176019029,
+      "grad_norm": 0.2524944841861725,
+      "learning_rate": 1.3549367535165935e-05,
+      "loss": 0.3816,
       "step": 91860
     },
     {
-      "epoch": 3.23,
-      "learning_rate": 1.4694688902356458e-05,
-      "loss": 0.2567,
+      "epoch": 3.3108083756802538,
+      "grad_norm": 0.22844867408275604,
+      "learning_rate": 1.3546773569908661e-05,
+      "loss": 0.3623,
       "step": 91865
     },
     {
-      "epoch": 3.23,
-      "learning_rate": 1.469209353415445e-05,
-      "loss": 0.2353,
+      "epoch": 3.3109885753414785,
+      "grad_norm": 0.20473162829875946,
+      "learning_rate": 1.35441797606994e-05,
+      "loss": 0.4076,
       "step": 91870
     },
     {
-      "epoch": 3.23,
-      "learning_rate": 1.4689498299793813e-05,
-      "loss": 0.258,
+      "epoch": 3.3111687750027032,
+      "grad_norm": 0.2806645631790161,
+      "learning_rate": 1.3541586107573456e-05,
+      "loss": 0.3582,
       "step": 91875
     },
     {
-      "epoch": 3.23,
-      "learning_rate": 1.468690319930825e-05,
-      "loss": 0.2777,
+      "epoch": 3.3113489746639275,
+      "grad_norm": 0.2380189746618271,
+      "learning_rate": 1.3538992610566175e-05,
+      "loss": 0.3897,
       "step": 91880
     },
     {
-      "epoch": 3.23,
-      "learning_rate": 1.4684308232731457e-05,
-      "loss": 0.2432,
+      "epoch": 3.3115291743251523,
+      "grad_norm": 0.20420801639556885,
+      "learning_rate": 1.3536399269712912e-05,
+      "loss": 0.3863,
       "step": 91885
     },
     {
-      "epoch": 3.23,
-      "learning_rate": 1.468171340009712e-05,
-      "loss": 0.2557,
+      "epoch": 3.311709373986377,
+      "grad_norm": 0.27330854535102844,
+      "learning_rate": 1.3533806085048991e-05,
+      "loss": 0.3967,
       "step": 91890
     },
     {
-      "epoch": 3.23,
-      "learning_rate": 1.4679118701438926e-05,
-      "loss": 0.2605,
+      "epoch": 3.3118895736476017,
+      "grad_norm": 0.26601719856262207,
+      "learning_rate": 1.3531213056609744e-05,
+      "loss": 0.3883,
       "step": 91895
     },
     {
-      "epoch": 3.23,
-      "learning_rate": 1.4676524136790584e-05,
-      "loss": 0.251,
+      "epoch": 3.312069773308826,
+      "grad_norm": 0.22995233535766602,
+      "learning_rate": 1.352862018443049e-05,
+      "loss": 0.3689,
       "step": 91900
     },
     {
-      "epoch": 3.23,
-      "learning_rate": 1.4673929706185769e-05,
-      "loss": 0.2479,
+      "epoch": 3.3122499729700507,
+      "grad_norm": 0.26952049136161804,
+      "learning_rate": 1.3526027468546562e-05,
+      "loss": 0.3816,
       "step": 91905
     },
     {
-      "epoch": 3.23,
-      "learning_rate": 1.4671335409658175e-05,
-      "loss": 0.2499,
+      "epoch": 3.3124301726312755,
+      "grad_norm": 0.23863959312438965,
+      "learning_rate": 1.3523434908993299e-05,
+      "loss": 0.3797,
       "step": 91910
     },
     {
-      "epoch": 3.23,
-      "learning_rate": 1.4668741247241475e-05,
-      "loss": 0.2418,
+      "epoch": 3.3126103722925,
+      "grad_norm": 0.2401474565267563,
+      "learning_rate": 1.3520842505806008e-05,
+      "loss": 0.3767,
       "step": 91915
     },
     {
-      "epoch": 3.23,
-      "learning_rate": 1.4666147218969367e-05,
-      "loss": 0.2717,
+      "epoch": 3.312790571953725,
+      "grad_norm": 0.24980275332927704,
+      "learning_rate": 1.3518250259020021e-05,
+      "loss": 0.3933,
       "step": 91920
     },
     {
-      "epoch": 3.23,
-      "learning_rate": 1.4663553324875529e-05,
-      "loss": 0.2779,
+      "epoch": 3.3129707716149492,
+      "grad_norm": 0.2250710129737854,
+      "learning_rate": 1.3515658168670647e-05,
+      "loss": 0.4035,
       "step": 91925
     },
     {
-      "epoch": 3.23,
-      "learning_rate": 1.466095956499363e-05,
-      "loss": 0.2812,
+      "epoch": 3.313150971276174,
+      "grad_norm": 0.23497453331947327,
+      "learning_rate": 1.3513066234793198e-05,
+      "loss": 0.3815,
       "step": 91930
     },
     {
-      "epoch": 3.23,
-      "learning_rate": 1.4658365939357366e-05,
-      "loss": 0.2558,
+      "epoch": 3.3133311709373987,
+      "grad_norm": Infinity,
+      "learning_rate": 1.3510992800374772e-05,
+      "loss": 0.3933,
       "step": 91935
     },
     {
-      "epoch": 3.23,
-      "learning_rate": 1.4655772448000408e-05,
-      "loss": 0.2468,
+      "epoch": 3.3135113705986234,
+      "grad_norm": 0.20314912497997284,
+      "learning_rate": 1.3508401148235816e-05,
+      "loss": 0.4087,
       "step": 91940
     },
     {
-      "epoch": 3.23,
-      "learning_rate": 1.4653179090956425e-05,
-      "loss": 0.2882,
+      "epoch": 3.3136915702598477,
+      "grad_norm": 0.25075605511665344,
+      "learning_rate": 1.3505809652667658e-05,
+      "loss": 0.3749,
       "step": 91945
     },
     {
-      "epoch": 3.24,
-      "learning_rate": 1.4650585868259087e-05,
-      "loss": 0.2731,
+      "epoch": 3.3138717699210725,
+      "grad_norm": 0.2547110617160797,
+      "learning_rate": 1.3503218313705612e-05,
+      "loss": 0.4021,
       "step": 91950
     },
     {
-      "epoch": 3.24,
-      "learning_rate": 1.4647992779942077e-05,
-      "loss": 0.2524,
+      "epoch": 3.314051969582297,
+      "grad_norm": 0.28053396940231323,
+      "learning_rate": 1.3500627131384996e-05,
+      "loss": 0.399,
       "step": 91955
     },
     {
-      "epoch": 3.24,
-      "learning_rate": 1.4645399826039052e-05,
-      "loss": 0.264,
+      "epoch": 3.314232169243522,
+      "grad_norm": 0.2546713054180145,
+      "learning_rate": 1.3498036105741113e-05,
+      "loss": 0.394,
       "step": 91960
     },
     {
-      "epoch": 3.24,
-      "learning_rate": 1.4642807006583698e-05,
-      "loss": 0.2514,
+      "epoch": 3.3144123689047467,
+      "grad_norm": 0.24178683757781982,
+      "learning_rate": 1.3495445236809263e-05,
+      "loss": 0.3621,
       "step": 91965
     },
     {
-      "epoch": 3.24,
-      "learning_rate": 1.4640214321609653e-05,
-      "loss": 0.2804,
+      "epoch": 3.314592568565971,
+      "grad_norm": 0.2593505084514618,
+      "learning_rate": 1.3492854524624737e-05,
+      "loss": 0.3819,
       "step": 91970
     },
     {
-      "epoch": 3.24,
-      "learning_rate": 1.463762177115061e-05,
-      "loss": 0.2562,
+      "epoch": 3.3147727682271957,
+      "grad_norm": 0.25846415758132935,
+      "learning_rate": 1.3490263969222838e-05,
+      "loss": 0.3856,
       "step": 91975
     },
     {
-      "epoch": 3.24,
-      "learning_rate": 1.4635029355240221e-05,
-      "loss": 0.2612,
+      "epoch": 3.3149529678884204,
+      "grad_norm": 0.20164726674556732,
+      "learning_rate": 1.348767357063887e-05,
+      "loss": 0.3679,
       "step": 91980
     },
     {
-      "epoch": 3.24,
-      "learning_rate": 1.4632437073912136e-05,
-      "loss": 0.2569,
+      "epoch": 3.315133167549645,
+      "grad_norm": 0.22583597898483276,
+      "learning_rate": 1.348508332890812e-05,
+      "loss": 0.3902,
       "step": 91985
     },
     {
-      "epoch": 3.24,
-      "learning_rate": 1.462984492720002e-05,
-      "loss": 0.2519,
+      "epoch": 3.3153133672108694,
+      "grad_norm": 0.223631352186203,
+      "learning_rate": 1.348249324406588e-05,
+      "loss": 0.3685,
       "step": 91990
     },
     {
-      "epoch": 3.24,
-      "learning_rate": 1.4627252915137546e-05,
-      "loss": 0.2588,
+      "epoch": 3.315493566872094,
+      "grad_norm": 0.2747786343097687,
+      "learning_rate": 1.347990331614744e-05,
+      "loss": 0.4107,
       "step": 91995
     },
     {
-      "epoch": 3.24,
-      "learning_rate": 1.4624661037758356e-05,
-      "loss": 0.2677,
+      "epoch": 3.315673766533319,
+      "grad_norm": 0.22893761098384857,
+      "learning_rate": 1.347731354518808e-05,
+      "loss": 0.3957,
       "step": 92000
     },
     {
-      "epoch": 3.24,
-      "eval_loss": 0.2548165023326874,
-      "eval_runtime": 10.5617,
-      "eval_samples_per_second": 9.468,
-      "eval_steps_per_second": 9.468,
+      "epoch": 3.315673766533319,
+      "eval_loss": 0.4305948317050934,
+      "eval_runtime": 3.5291,
+      "eval_samples_per_second": 28.336,
+      "eval_steps_per_second": 7.084,
       "step": 92000
     },
     {
-      "epoch": 3.24,
-      "learning_rate": 1.4622069295096096e-05,
-      "loss": 0.2453,
+      "epoch": 3.3158539661945436,
+      "grad_norm": 0.269552081823349,
+      "learning_rate": 1.3474723931223102e-05,
+      "loss": 0.3697,
       "step": 92005
     },
     {
-      "epoch": 3.24,
-      "learning_rate": 1.461947768718443e-05,
-      "loss": 0.2439,
+      "epoch": 3.3160341658557684,
+      "grad_norm": 0.21320883929729462,
+      "learning_rate": 1.3472134474287777e-05,
+      "loss": 0.3593,
       "step": 92010
     },
     {
-      "epoch": 3.24,
-      "learning_rate": 1.4616886214057007e-05,
-      "loss": 0.2609,
+      "epoch": 3.3162143655169927,
+      "grad_norm": 0.25191807746887207,
+      "learning_rate": 1.3469545174417386e-05,
+      "loss": 0.4161,
       "step": 92015
     },
     {
-      "epoch": 3.24,
-      "learning_rate": 1.4614294875747475e-05,
-      "loss": 0.2477,
+      "epoch": 3.3163945651782174,
+      "grad_norm": 0.3065241575241089,
+      "learning_rate": 1.3466956031647212e-05,
+      "loss": 0.3889,
       "step": 92020
     },
     {
-      "epoch": 3.24,
-      "learning_rate": 1.4611703672289484e-05,
-      "loss": 0.2486,
+      "epoch": 3.316574764839442,
+      "grad_norm": 0.22870443761348724,
+      "learning_rate": 1.3464367046012521e-05,
+      "loss": 0.3896,
       "step": 92025
     },
     {
-      "epoch": 3.24,
-      "learning_rate": 1.460911260371666e-05,
-      "loss": 0.2642,
+      "epoch": 3.316754964500667,
+      "grad_norm": 0.2070244699716568,
+      "learning_rate": 1.3461778217548603e-05,
+      "loss": 0.4069,
       "step": 92030
     },
     {
-      "epoch": 3.24,
-      "learning_rate": 1.4606521670062673e-05,
-      "loss": 0.2674,
+      "epoch": 3.316935164161891,
+      "grad_norm": 0.20733942091464996,
+      "learning_rate": 1.3459189546290735e-05,
+      "loss": 0.3833,
       "step": 92035
     },
     {
-      "epoch": 3.24,
-      "learning_rate": 1.460393087136115e-05,
-      "loss": 0.2648,
+      "epoch": 3.317115363823116,
+      "grad_norm": 0.22954271733760834,
+      "learning_rate": 1.3456601032274153e-05,
+      "loss": 0.4083,
       "step": 92040
     },
     {
-      "epoch": 3.24,
-      "learning_rate": 1.4601340207645726e-05,
-      "loss": 0.2583,
+      "epoch": 3.3172955634843406,
+      "grad_norm": 0.23598207533359528,
+      "learning_rate": 1.3454012675534156e-05,
+      "loss": 0.3831,
       "step": 92045
     },
     {
-      "epoch": 3.24,
-      "learning_rate": 1.4598749678950052e-05,
-      "loss": 0.2811,
+      "epoch": 3.3174757631455654,
+      "grad_norm": 0.2571483850479126,
+      "learning_rate": 1.3451424476106004e-05,
+      "loss": 0.3722,
       "step": 92050
     },
     {
-      "epoch": 3.24,
-      "learning_rate": 1.4596159285307758e-05,
-      "loss": 0.2561,
+      "epoch": 3.31765596280679,
+      "grad_norm": 0.22039219737052917,
+      "learning_rate": 1.3448836434024955e-05,
+      "loss": 0.3507,
       "step": 92055
     },
     {
-      "epoch": 3.24,
-      "learning_rate": 1.4593569026752484e-05,
-      "loss": 0.2414,
+      "epoch": 3.3178361624680144,
+      "grad_norm": 0.2569846510887146,
+      "learning_rate": 1.3446248549326274e-05,
+      "loss": 0.3714,
       "step": 92060
     },
     {
-      "epoch": 3.24,
-      "learning_rate": 1.459097890331785e-05,
-      "loss": 0.2629,
+      "epoch": 3.318016362129239,
+      "grad_norm": 0.24034719169139862,
+      "learning_rate": 1.3443660822045211e-05,
+      "loss": 0.3962,
       "step": 92065
     },
     {
-      "epoch": 3.24,
-      "learning_rate": 1.4588388915037504e-05,
-      "loss": 0.2519,
+      "epoch": 3.318196561790464,
+      "grad_norm": 0.27031049132347107,
+      "learning_rate": 1.344107325221704e-05,
+      "loss": 0.3683,
       "step": 92070
     },
     {
-      "epoch": 3.24,
-      "learning_rate": 1.4585799061945065e-05,
-      "loss": 0.2487,
+      "epoch": 3.3183767614516886,
+      "grad_norm": 0.31037384271621704,
+      "learning_rate": 1.343848583987701e-05,
+      "loss": 0.4435,
       "step": 92075
     },
     {
-      "epoch": 3.24,
-      "learning_rate": 1.4583209344074164e-05,
-      "loss": 0.2793,
+      "epoch": 3.318556961112913,
+      "grad_norm": 0.21680322289466858,
+      "learning_rate": 1.3435898585060372e-05,
+      "loss": 0.3767,
       "step": 92080
     },
     {
-      "epoch": 3.24,
-      "learning_rate": 1.4580619761458414e-05,
-      "loss": 0.2651,
+      "epoch": 3.3187371607741376,
+      "grad_norm": 0.22905753552913666,
+      "learning_rate": 1.3433311487802375e-05,
+      "loss": 0.3935,
       "step": 92085
     },
     {
-      "epoch": 3.24,
-      "learning_rate": 1.4578030314131457e-05,
-      "loss": 0.2645,
+      "epoch": 3.3189173604353623,
+      "grad_norm": 0.27699750661849976,
+      "learning_rate": 1.343072454813827e-05,
+      "loss": 0.4165,
       "step": 92090
     },
     {
-      "epoch": 3.24,
-      "learning_rate": 1.4575441002126913e-05,
-      "loss": 0.2459,
+      "epoch": 3.319097560096587,
+      "grad_norm": 0.19221347570419312,
+      "learning_rate": 1.3428137766103293e-05,
+      "loss": 0.3773,
       "step": 92095
     },
     {
-      "epoch": 3.24,
-      "learning_rate": 1.4572851825478395e-05,
-      "loss": 0.2574,
+      "epoch": 3.319277759757812,
+      "grad_norm": 0.2585064172744751,
+      "learning_rate": 1.3425551141732712e-05,
+      "loss": 0.3909,
       "step": 92100
     },
     {
-      "epoch": 3.24,
-      "learning_rate": 1.4570262784219513e-05,
-      "loss": 0.2627,
+      "epoch": 3.3194579594190365,
+      "grad_norm": 0.21457339823246002,
+      "learning_rate": 1.3422964675061752e-05,
+      "loss": 0.3761,
       "step": 92105
     },
     {
-      "epoch": 3.24,
-      "learning_rate": 1.4567673878383897e-05,
-      "loss": 0.2483,
+      "epoch": 3.319638159080261,
+      "grad_norm": 0.22903072834014893,
+      "learning_rate": 1.3420378366125657e-05,
+      "loss": 0.3808,
       "step": 92110
     },
     {
-      "epoch": 3.24,
-      "learning_rate": 1.4565085108005172e-05,
-      "loss": 0.2788,
+      "epoch": 3.3198183587414856,
+      "grad_norm": 0.24865801632404327,
+      "learning_rate": 1.3417792214959668e-05,
+      "loss": 0.3829,
       "step": 92115
     },
     {
-      "epoch": 3.24,
-      "learning_rate": 1.4562496473116938e-05,
-      "loss": 0.2663,
+      "epoch": 3.3199985584027103,
+      "grad_norm": 0.22578608989715576,
+      "learning_rate": 1.3415206221599006e-05,
+      "loss": 0.3397,
       "step": 92120
     },
     {
-      "epoch": 3.24,
-      "learning_rate": 1.4559907973752798e-05,
-      "loss": 0.2607,
+      "epoch": 3.320178758063935,
+      "grad_norm": 0.21695365011692047,
+      "learning_rate": 1.3412620386078933e-05,
+      "loss": 0.3552,
       "step": 92125
     },
     {
-      "epoch": 3.24,
-      "learning_rate": 1.4557319609946385e-05,
-      "loss": 0.254,
+      "epoch": 3.3203589577251593,
+      "grad_norm": 0.19830626249313354,
+      "learning_rate": 1.341003470843466e-05,
+      "loss": 0.3794,
       "step": 92130
     },
     {
-      "epoch": 3.24,
-      "learning_rate": 1.4554731381731291e-05,
-      "loss": 0.2415,
+      "epoch": 3.320539157386384,
+      "grad_norm": 0.25142696499824524,
+      "learning_rate": 1.3407449188701408e-05,
+      "loss": 0.387,
       "step": 92135
     },
     {
-      "epoch": 3.24,
-      "learning_rate": 1.4552143289141129e-05,
-      "loss": 0.2497,
+      "epoch": 3.320719357047609,
+      "grad_norm": 0.25763440132141113,
+      "learning_rate": 1.3404863826914424e-05,
+      "loss": 0.4225,
       "step": 92140
     },
     {
-      "epoch": 3.24,
-      "learning_rate": 1.454955533220949e-05,
-      "loss": 0.2357,
+      "epoch": 3.3208995567088335,
+      "grad_norm": 0.2544826567173004,
+      "learning_rate": 1.3402278623108915e-05,
+      "loss": 0.3771,
       "step": 92145
     },
     {
-      "epoch": 3.24,
-      "learning_rate": 1.454696751097e-05,
-      "loss": 0.2616,
+      "epoch": 3.3210797563700583,
+      "grad_norm": 0.24022381007671356,
+      "learning_rate": 1.3399693577320137e-05,
+      "loss": 0.3724,
       "step": 92150
     },
     {
-      "epoch": 3.24,
-      "learning_rate": 1.4544379825456247e-05,
-      "loss": 0.2607,
+      "epoch": 3.3212599560312825,
+      "grad_norm": 0.24866530299186707,
+      "learning_rate": 1.3397108689583276e-05,
+      "loss": 0.3865,
       "step": 92155
     },
     {
-      "epoch": 3.24,
-      "learning_rate": 1.4541792275701821e-05,
-      "loss": 0.2507,
+      "epoch": 3.3214401556925073,
+      "grad_norm": 0.2202133685350418,
+      "learning_rate": 1.3394523959933559e-05,
+      "loss": 0.3777,
       "step": 92160
     },
     {
-      "epoch": 3.24,
-      "learning_rate": 1.453920486174033e-05,
-      "loss": 0.2505,
+      "epoch": 3.321620355353732,
+      "grad_norm": 0.1919008046388626,
+      "learning_rate": 1.339193938840621e-05,
+      "loss": 0.3427,
       "step": 92165
     },
     {
-      "epoch": 3.24,
-      "learning_rate": 1.4536617583605381e-05,
-      "loss": 0.2799,
+      "epoch": 3.3218005550149567,
+      "grad_norm": 0.21456719934940338,
+      "learning_rate": 1.338935497503644e-05,
+      "loss": 0.406,
       "step": 92170
     },
     {
-      "epoch": 3.24,
-      "learning_rate": 1.4534030441330556e-05,
-      "loss": 0.2376,
+      "epoch": 3.321980754676181,
+      "grad_norm": 0.17022041976451874,
+      "learning_rate": 1.338677071985946e-05,
+      "loss": 0.3656,
       "step": 92175
     },
     {
-      "epoch": 3.24,
-      "learning_rate": 1.4531443434949439e-05,
-      "loss": 0.2609,
+      "epoch": 3.3221609543374058,
+      "grad_norm": 0.2558014392852783,
+      "learning_rate": 1.3384186622910482e-05,
+      "loss": 0.3812,
       "step": 92180
     },
     {
-      "epoch": 3.24,
-      "learning_rate": 1.4528856564495642e-05,
-      "loss": 0.2734,
+      "epoch": 3.3223411539986305,
+      "grad_norm": 0.24474850296974182,
+      "learning_rate": 1.3381602684224703e-05,
+      "loss": 0.372,
       "step": 92185
     },
     {
-      "epoch": 3.24,
-      "learning_rate": 1.4526269830002737e-05,
-      "loss": 0.2529,
+      "epoch": 3.3225213536598552,
+      "grad_norm": 0.2648925483226776,
+      "learning_rate": 1.3379018903837348e-05,
+      "loss": 0.3879,
       "step": 92190
     },
     {
-      "epoch": 3.24,
-      "learning_rate": 1.4523683231504318e-05,
-      "loss": 0.263,
+      "epoch": 3.32270155332108,
+      "grad_norm": 0.2510066330432892,
+      "learning_rate": 1.3376435281783608e-05,
+      "loss": 0.3624,
       "step": 92195
     },
     {
-      "epoch": 3.24,
-      "learning_rate": 1.4521096769033959e-05,
-      "loss": 0.2631,
+      "epoch": 3.3228817529823043,
+      "grad_norm": 0.20649553835391998,
+      "learning_rate": 1.337385181809869e-05,
+      "loss": 0.3811,
       "step": 92200
     },
     {
-      "epoch": 3.24,
-      "learning_rate": 1.451851044262526e-05,
-      "loss": 0.2638,
+      "epoch": 3.323061952643529,
+      "grad_norm": 0.24206553399562836,
+      "learning_rate": 1.337126851281779e-05,
+      "loss": 0.3888,
       "step": 92205
     },
     {
-      "epoch": 3.24,
-      "learning_rate": 1.4515924252311796e-05,
-      "loss": 0.2381,
+      "epoch": 3.3232421523047537,
+      "grad_norm": 0.21972547471523285,
+      "learning_rate": 1.3368685365976091e-05,
+      "loss": 0.3822,
       "step": 92210
     },
     {
-      "epoch": 3.24,
-      "learning_rate": 1.4513338198127142e-05,
-      "loss": 0.2614,
+      "epoch": 3.3234223519659785,
+      "grad_norm": 0.22937068343162537,
+      "learning_rate": 1.3366102377608817e-05,
+      "loss": 0.3684,
       "step": 92215
     },
     {
-      "epoch": 3.24,
-      "learning_rate": 1.4510752280104872e-05,
-      "loss": 0.2711,
+      "epoch": 3.3236025516272028,
+      "grad_norm": 0.23047223687171936,
+      "learning_rate": 1.3363519547751152e-05,
+      "loss": 0.3663,
       "step": 92220
     },
     {
-      "epoch": 3.24,
-      "learning_rate": 1.4508166498278586e-05,
-      "loss": 0.2606,
+      "epoch": 3.3237827512884275,
+      "grad_norm": 0.19826874136924744,
+      "learning_rate": 1.3360936876438263e-05,
+      "loss": 0.3945,
       "step": 92225
     },
     {
-      "epoch": 3.24,
-      "learning_rate": 1.4505580852681843e-05,
-      "loss": 0.2653,
+      "epoch": 3.323962950949652,
+      "grad_norm": 0.2385808378458023,
+      "learning_rate": 1.3358354363705365e-05,
+      "loss": 0.4018,
       "step": 92230
     },
     {
-      "epoch": 3.25,
-      "learning_rate": 1.4502995343348218e-05,
-      "loss": 0.2511,
+      "epoch": 3.324143150610877,
+      "grad_norm": 0.24572807550430298,
+      "learning_rate": 1.3355772009587634e-05,
+      "loss": 0.405,
       "step": 92235
     },
     {
-      "epoch": 3.25,
-      "learning_rate": 1.4500409970311268e-05,
-      "loss": 0.2755,
+      "epoch": 3.3243233502721017,
+      "grad_norm": 0.237590491771698,
+      "learning_rate": 1.3353189814120253e-05,
+      "loss": 0.3893,
       "step": 92240
     },
     {
-      "epoch": 3.25,
-      "learning_rate": 1.4497824733604588e-05,
-      "loss": 0.2681,
+      "epoch": 3.324503549933326,
+      "grad_norm": 0.20121124386787415,
+      "learning_rate": 1.3350607777338409e-05,
+      "loss": 0.3707,
       "step": 92245
     },
     {
-      "epoch": 3.25,
-      "learning_rate": 1.4495239633261728e-05,
-      "loss": 0.2792,
+      "epoch": 3.3246837495945507,
+      "grad_norm": 0.260689377784729,
+      "learning_rate": 1.334802589927727e-05,
+      "loss": 0.3887,
       "step": 92250
     },
     {
-      "epoch": 3.25,
-      "learning_rate": 1.4492654669316263e-05,
-      "loss": 0.2421,
+      "epoch": 3.3248639492557754,
+      "grad_norm": 0.31178975105285645,
+      "learning_rate": 1.3345444179972027e-05,
+      "loss": 0.4013,
       "step": 92255
     },
     {
-      "epoch": 3.25,
-      "learning_rate": 1.449006984180174e-05,
-      "loss": 0.2473,
+      "epoch": 3.325044148917,
+      "grad_norm": 0.24525390565395355,
+      "learning_rate": 1.3342862619457858e-05,
+      "loss": 0.3908,
       "step": 92260
     },
     {
-      "epoch": 3.25,
-      "learning_rate": 1.4487485150751745e-05,
-      "loss": 0.2751,
+      "epoch": 3.3252243485782245,
+      "grad_norm": 0.20837341248989105,
+      "learning_rate": 1.3340281217769922e-05,
+      "loss": 0.3367,
       "step": 92265
     },
     {
-      "epoch": 3.25,
-      "learning_rate": 1.4484900596199816e-05,
-      "loss": 0.2476,
+      "epoch": 3.325404548239449,
+      "grad_norm": 0.24146051704883575,
+      "learning_rate": 1.3337699974943401e-05,
+      "loss": 0.3905,
       "step": 92270
     },
     {
-      "epoch": 3.25,
-      "learning_rate": 1.4482316178179534e-05,
-      "loss": 0.261,
+      "epoch": 3.325584747900674,
+      "grad_norm": 0.2158283293247223,
+      "learning_rate": 1.3335118891013459e-05,
+      "loss": 0.3847,
       "step": 92275
     },
     {
-      "epoch": 3.25,
-      "learning_rate": 1.4479731896724436e-05,
-      "loss": 0.2639,
+      "epoch": 3.3257649475618987,
+      "grad_norm": 0.2629423141479492,
+      "learning_rate": 1.3332537966015251e-05,
+      "loss": 0.4141,
       "step": 92280
     },
     {
-      "epoch": 3.25,
-      "learning_rate": 1.4477147751868092e-05,
-      "loss": 0.2642,
+      "epoch": 3.3259451472231234,
+      "grad_norm": 0.22925181686878204,
+      "learning_rate": 1.3329957199983967e-05,
+      "loss": 0.3676,
       "step": 92285
     },
     {
-      "epoch": 3.25,
-      "learning_rate": 1.4474563743644054e-05,
-      "loss": 0.2661,
+      "epoch": 3.3261253468843477,
+      "grad_norm": 0.2732671797275543,
+      "learning_rate": 1.3327376592954757e-05,
+      "loss": 0.3693,
       "step": 92290
     },
     {
-      "epoch": 3.25,
-      "learning_rate": 1.4471979872085856e-05,
-      "loss": 0.2732,
+      "epoch": 3.3263055465455724,
+      "grad_norm": 0.21771977841854095,
+      "learning_rate": 1.3324796144962781e-05,
+      "loss": 0.3854,
       "step": 92295
     },
     {
-      "epoch": 3.25,
-      "learning_rate": 1.4469396137227076e-05,
-      "loss": 0.2934,
+      "epoch": 3.326485746206797,
+      "grad_norm": 0.19893218576908112,
+      "learning_rate": 1.3322215856043196e-05,
+      "loss": 0.3719,
       "step": 92300
     },
     {
-      "epoch": 3.25,
-      "learning_rate": 1.4466812539101242e-05,
-      "loss": 0.2657,
+      "epoch": 3.326665945868022,
+      "grad_norm": 0.253020703792572,
+      "learning_rate": 1.331963572623115e-05,
+      "loss": 0.3934,
       "step": 92305
     },
     {
-      "epoch": 3.25,
-      "learning_rate": 1.446422907774191e-05,
-      "loss": 0.2484,
+      "epoch": 3.326846145529246,
+      "grad_norm": 0.272172749042511,
+      "learning_rate": 1.3317055755561817e-05,
+      "loss": 0.4027,
       "step": 92310
     },
     {
-      "epoch": 3.25,
-      "learning_rate": 1.4461645753182607e-05,
-      "loss": 0.2537,
+      "epoch": 3.327026345190471,
+      "grad_norm": 0.22048524022102356,
+      "learning_rate": 1.3314475944070345e-05,
+      "loss": 0.383,
       "step": 92315
     },
     {
-      "epoch": 3.25,
-      "learning_rate": 1.44590625654569e-05,
-      "loss": 0.2774,
+      "epoch": 3.3272065448516956,
+      "grad_norm": 0.188111811876297,
+      "learning_rate": 1.3311896291791859e-05,
+      "loss": 0.3955,
       "step": 92320
     },
     {
-      "epoch": 3.25,
-      "learning_rate": 1.445647951459831e-05,
-      "loss": 0.2674,
+      "epoch": 3.3273867445129204,
+      "grad_norm": 0.21891506016254425,
+      "learning_rate": 1.3309316798761529e-05,
+      "loss": 0.3913,
       "step": 92325
     },
     {
-      "epoch": 3.25,
-      "learning_rate": 1.445389660064039e-05,
-      "loss": 0.2582,
+      "epoch": 3.327566944174145,
+      "grad_norm": 0.24252718687057495,
+      "learning_rate": 1.3306737465014485e-05,
+      "loss": 0.3619,
       "step": 92330
     },
     {
-      "epoch": 3.25,
-      "learning_rate": 1.445131382361667e-05,
-      "loss": 0.2502,
+      "epoch": 3.3277471438353694,
+      "grad_norm": 0.23667490482330322,
+      "learning_rate": 1.33041582905859e-05,
+      "loss": 0.3844,
       "step": 92335
     },
     {
-      "epoch": 3.25,
-      "learning_rate": 1.4448731183560694e-05,
-      "loss": 0.2593,
+      "epoch": 3.327927343496594,
+      "grad_norm": 0.21754367649555206,
+      "learning_rate": 1.3301579275510887e-05,
+      "loss": 0.3438,
       "step": 92340
     },
     {
-      "epoch": 3.25,
-      "learning_rate": 1.4446148680505989e-05,
-      "loss": 0.2414,
+      "epoch": 3.328107543157819,
+      "grad_norm": 0.2473175972700119,
+      "learning_rate": 1.3299000419824578e-05,
+      "loss": 0.4039,
       "step": 92345
     },
     {
-      "epoch": 3.25,
-      "learning_rate": 1.444356631448609e-05,
-      "loss": 0.2283,
+      "epoch": 3.3282877428190436,
+      "grad_norm": 0.2296813279390335,
+      "learning_rate": 1.3296421723562133e-05,
+      "loss": 0.3446,
       "step": 92350
     },
     {
-      "epoch": 3.25,
-      "learning_rate": 1.4440984085534514e-05,
-      "loss": 0.272,
+      "epoch": 3.328467942480268,
+      "grad_norm": 0.24328921735286713,
+      "learning_rate": 1.3293843186758675e-05,
+      "loss": 0.3843,
       "step": 92355
     },
     {
-      "epoch": 3.25,
-      "learning_rate": 1.4438401993684811e-05,
-      "loss": 0.2576,
+      "epoch": 3.3286481421414926,
+      "grad_norm": 0.23065999150276184,
+      "learning_rate": 1.329126480944934e-05,
+      "loss": 0.3811,
       "step": 92360
     },
     {
-      "epoch": 3.25,
-      "learning_rate": 1.4435820038970498e-05,
-      "loss": 0.2794,
+      "epoch": 3.3288283418027174,
+      "grad_norm": 0.2626056373119354,
+      "learning_rate": 1.3288686591669255e-05,
+      "loss": 0.3911,
       "step": 92365
     },
     {
-      "epoch": 3.25,
-      "learning_rate": 1.44332382214251e-05,
-      "loss": 0.2481,
+      "epoch": 3.329008541463942,
+      "grad_norm": 0.2469651997089386,
+      "learning_rate": 1.3286108533453537e-05,
+      "loss": 0.3647,
       "step": 92370
     },
     {
-      "epoch": 3.25,
-      "learning_rate": 1.4430656541082133e-05,
-      "loss": 0.2631,
+      "epoch": 3.329188741125167,
+      "grad_norm": 0.23114638030529022,
+      "learning_rate": 1.3283530634837333e-05,
+      "loss": 0.374,
       "step": 92375
     },
     {
-      "epoch": 3.25,
-      "learning_rate": 1.4428074997975132e-05,
-      "loss": 0.2682,
+      "epoch": 3.3293689407863916,
+      "grad_norm": 0.209586039185524,
+      "learning_rate": 1.3280952895855753e-05,
+      "loss": 0.3747,
       "step": 92380
     },
     {
-      "epoch": 3.25,
-      "learning_rate": 1.4425493592137601e-05,
-      "loss": 0.2555,
+      "epoch": 3.329549140447616,
+      "grad_norm": 0.25766298174858093,
+      "learning_rate": 1.3278375316543924e-05,
+      "loss": 0.3928,
       "step": 92385
     },
     {
-      "epoch": 3.25,
-      "learning_rate": 1.4422912323603066e-05,
-      "loss": 0.2567,
+      "epoch": 3.3297293401088406,
+      "grad_norm": 0.20684154331684113,
+      "learning_rate": 1.3275797896936958e-05,
+      "loss": 0.3196,
       "step": 92390
     },
     {
-      "epoch": 3.25,
-      "learning_rate": 1.4420331192405053e-05,
-      "loss": 0.2441,
+      "epoch": 3.3299095397700653,
+      "grad_norm": 0.27292490005493164,
+      "learning_rate": 1.3273220637069971e-05,
+      "loss": 0.3889,
       "step": 92395
     },
     {
-      "epoch": 3.25,
-      "learning_rate": 1.4417750198577069e-05,
-      "loss": 0.2532,
+      "epoch": 3.33008973943129,
+      "grad_norm": 0.2662750780582428,
+      "learning_rate": 1.327064353697809e-05,
+      "loss": 0.385,
       "step": 92400
     },
     {
-      "epoch": 3.25,
-      "learning_rate": 1.4415169342152624e-05,
-      "loss": 0.2683,
+      "epoch": 3.3302699390925143,
+      "grad_norm": 0.23216348886489868,
+      "learning_rate": 1.326806659669643e-05,
+      "loss": 0.3605,
       "step": 92405
     },
     {
-      "epoch": 3.25,
-      "learning_rate": 1.4412588623165218e-05,
-      "loss": 0.2695,
+      "epoch": 3.330450138753739,
+      "grad_norm": 0.2415631115436554,
+      "learning_rate": 1.3265489816260071e-05,
+      "loss": 0.4063,
       "step": 92410
     },
     {
-      "epoch": 3.25,
-      "learning_rate": 1.4410008041648388e-05,
-      "loss": 0.257,
+      "epoch": 3.330630338414964,
+      "grad_norm": 0.2386670559644699,
+      "learning_rate": 1.3262913195704152e-05,
+      "loss": 0.405,
       "step": 92415
     },
     {
-      "epoch": 3.25,
-      "learning_rate": 1.4407427597635616e-05,
-      "loss": 0.2384,
+      "epoch": 3.3308105380761885,
+      "grad_norm": 0.29892489314079285,
+      "learning_rate": 1.3260336735063766e-05,
+      "loss": 0.3817,
       "step": 92420
     },
     {
-      "epoch": 3.25,
-      "learning_rate": 1.4404847291160423e-05,
-      "loss": 0.2613,
+      "epoch": 3.3309907377374133,
+      "grad_norm": 0.2595389485359192,
+      "learning_rate": 1.325776043437401e-05,
+      "loss": 0.4289,
       "step": 92425
     },
     {
-      "epoch": 3.25,
-      "learning_rate": 1.4402267122256291e-05,
-      "loss": 0.2498,
+      "epoch": 3.3311709373986376,
+      "grad_norm": 0.23359209299087524,
+      "learning_rate": 1.325518429367002e-05,
+      "loss": 0.3704,
       "step": 92430
     },
     {
-      "epoch": 3.25,
-      "learning_rate": 1.4399687090956746e-05,
-      "loss": 0.2649,
+      "epoch": 3.3313511370598623,
+      "grad_norm": 0.22734783589839935,
+      "learning_rate": 1.3252608312986844e-05,
+      "loss": 0.3439,
       "step": 92435
     },
     {
-      "epoch": 3.25,
-      "learning_rate": 1.4397107197295268e-05,
-      "loss": 0.2418,
+      "epoch": 3.331531336721087,
+      "grad_norm": 0.21022412180900574,
+      "learning_rate": 1.3250032492359618e-05,
+      "loss": 0.3662,
       "step": 92440
     },
     {
-      "epoch": 3.25,
-      "learning_rate": 1.4394527441305378e-05,
-      "loss": 0.2674,
+      "epoch": 3.3317115363823118,
+      "grad_norm": 0.2911469042301178,
+      "learning_rate": 1.3247456831823423e-05,
+      "loss": 0.3742,
       "step": 92445
     },
     {
-      "epoch": 3.25,
-      "learning_rate": 1.439194782302055e-05,
-      "loss": 0.2477,
+      "epoch": 3.331891736043536,
+      "grad_norm": 0.24090684950351715,
+      "learning_rate": 1.3244881331413356e-05,
+      "loss": 0.371,
       "step": 92450
     },
     {
-      "epoch": 3.25,
-      "learning_rate": 1.4389368342474296e-05,
-      "loss": 0.2359,
+      "epoch": 3.332071935704761,
+      "grad_norm": 0.27764245867729187,
+      "learning_rate": 1.32423059911645e-05,
+      "loss": 0.3933,
       "step": 92455
     },
     {
-      "epoch": 3.25,
-      "learning_rate": 1.4386788999700102e-05,
-      "loss": 0.2487,
+      "epoch": 3.3322521353659855,
+      "grad_norm": 0.23873978853225708,
+      "learning_rate": 1.3239730811111952e-05,
+      "loss": 0.3861,
       "step": 92460
     },
     {
-      "epoch": 3.25,
-      "learning_rate": 1.4384209794731463e-05,
-      "loss": 0.2506,
+      "epoch": 3.3324323350272103,
+      "grad_norm": 0.19886447489261627,
+      "learning_rate": 1.3237155791290784e-05,
+      "loss": 0.3883,
       "step": 92465
     },
     {
-      "epoch": 3.25,
-      "learning_rate": 1.4381630727601847e-05,
-      "loss": 0.255,
+      "epoch": 3.332612534688435,
+      "grad_norm": 0.24735817313194275,
+      "learning_rate": 1.3234580931736102e-05,
+      "loss": 0.366,
       "step": 92470
     },
     {
-      "epoch": 3.25,
-      "learning_rate": 1.4379051798344772e-05,
-      "loss": 0.2906,
+      "epoch": 3.3327927343496593,
+      "grad_norm": 0.21200679242610931,
+      "learning_rate": 1.3232006232482978e-05,
+      "loss": 0.3777,
       "step": 92475
     },
     {
-      "epoch": 3.25,
-      "learning_rate": 1.437647300699371e-05,
-      "loss": 0.2423,
+      "epoch": 3.332972934010884,
+      "grad_norm": 0.22038812935352325,
+      "learning_rate": 1.3229431693566488e-05,
+      "loss": 0.3966,
       "step": 92480
     },
     {
-      "epoch": 3.25,
-      "learning_rate": 1.4373894353582135e-05,
-      "loss": 0.2642,
+      "epoch": 3.3331531336721087,
+      "grad_norm": 0.21706204116344452,
+      "learning_rate": 1.3226857315021712e-05,
+      "loss": 0.3784,
       "step": 92485
     },
     {
-      "epoch": 3.25,
-      "learning_rate": 1.4371315838143545e-05,
-      "loss": 0.2657,
+      "epoch": 3.3333333333333335,
+      "grad_norm": 0.26988691091537476,
+      "learning_rate": 1.3224283096883717e-05,
+      "loss": 0.3762,
       "step": 92490
     },
     {
-      "epoch": 3.25,
-      "learning_rate": 1.4368737460711406e-05,
-      "loss": 0.2582,
+      "epoch": 3.3335135329945578,
+      "grad_norm": 0.2401093691587448,
+      "learning_rate": 1.3221709039187597e-05,
+      "loss": 0.4154,
       "step": 92495
     },
     {
-      "epoch": 3.25,
-      "learning_rate": 1.436615922131922e-05,
-      "loss": 0.2564,
+      "epoch": 3.3336937326557825,
+      "grad_norm": 0.18479788303375244,
+      "learning_rate": 1.321913514196842e-05,
+      "loss": 0.3788,
       "step": 92500
     },
     {
-      "epoch": 3.25,
-      "eval_loss": 0.25434496998786926,
-      "eval_runtime": 10.5432,
-      "eval_samples_per_second": 9.485,
-      "eval_steps_per_second": 9.485,
+      "epoch": 3.3336937326557825,
+      "eval_loss": 0.42996159195899963,
+      "eval_runtime": 3.5289,
+      "eval_samples_per_second": 28.337,
+      "eval_steps_per_second": 7.084,
       "step": 92500
     },
     {
-      "epoch": 3.25,
-      "learning_rate": 1.4363581120000432e-05,
-      "loss": 0.2536,
+      "epoch": 3.3338739323170072,
+      "grad_norm": 0.28573232889175415,
+      "learning_rate": 1.3216561405261224e-05,
+      "loss": 0.4017,
       "step": 92505
     },
     {
-      "epoch": 3.25,
-      "learning_rate": 1.436100315678855e-05,
-      "loss": 0.2728,
+      "epoch": 3.334054131978232,
+      "grad_norm": 0.2534043490886688,
+      "learning_rate": 1.3213987829101108e-05,
+      "loss": 0.3497,
       "step": 92510
     },
     {
-      "epoch": 3.25,
-      "learning_rate": 1.4358425331717023e-05,
-      "loss": 0.2573,
+      "epoch": 3.3342343316394567,
+      "grad_norm": 0.218218594789505,
+      "learning_rate": 1.3211414413523115e-05,
+      "loss": 0.4098,
       "step": 92515
     },
     {
-      "epoch": 3.26,
-      "learning_rate": 1.4355847644819336e-05,
-      "loss": 0.2449,
+      "epoch": 3.334414531300681,
+      "grad_norm": 0.21875832974910736,
+      "learning_rate": 1.320884115856234e-05,
+      "loss": 0.3693,
       "step": 92520
     },
     {
-      "epoch": 3.26,
-      "learning_rate": 1.435327009612894e-05,
-      "loss": 0.245,
+      "epoch": 3.3345947309619057,
+      "grad_norm": 0.2261316180229187,
+      "learning_rate": 1.320626806425381e-05,
+      "loss": 0.3835,
       "step": 92525
     },
     {
-      "epoch": 3.26,
-      "learning_rate": 1.4350692685679323e-05,
-      "loss": 0.2713,
+      "epoch": 3.3347749306231305,
+      "grad_norm": 0.21817557513713837,
+      "learning_rate": 1.3203695130632587e-05,
+      "loss": 0.3939,
       "step": 92530
     },
     {
-      "epoch": 3.26,
-      "learning_rate": 1.4348115413503948e-05,
-      "loss": 0.2275,
+      "epoch": 3.334955130284355,
+      "grad_norm": 0.21988247334957123,
+      "learning_rate": 1.3201122357733742e-05,
+      "loss": 0.4075,
       "step": 92535
     },
     {
-      "epoch": 3.26,
-      "learning_rate": 1.4345538279636273e-05,
-      "loss": 0.2387,
+      "epoch": 3.3351353299455795,
+      "grad_norm": 0.26080963015556335,
+      "learning_rate": 1.319854974559232e-05,
+      "loss": 0.4094,
       "step": 92540
     },
     {
-      "epoch": 3.26,
-      "learning_rate": 1.4342961284109751e-05,
-      "loss": 0.2323,
+      "epoch": 3.335315529606804,
+      "grad_norm": 0.2803710103034973,
+      "learning_rate": 1.3195977294243372e-05,
+      "loss": 0.431,
       "step": 92545
     },
     {
-      "epoch": 3.26,
-      "learning_rate": 1.4340384426957854e-05,
-      "loss": 0.254,
+      "epoch": 3.335495729268029,
+      "grad_norm": 0.21274006366729736,
+      "learning_rate": 1.3193405003721951e-05,
+      "loss": 0.3808,
       "step": 92550
     },
     {
-      "epoch": 3.26,
-      "learning_rate": 1.4337807708214052e-05,
-      "loss": 0.2489,
+      "epoch": 3.3356759289292537,
+      "grad_norm": 0.23850922286510468,
+      "learning_rate": 1.3190832874063092e-05,
+      "loss": 0.406,
       "step": 92555
     },
     {
-      "epoch": 3.26,
-      "learning_rate": 1.4335231127911786e-05,
-      "loss": 0.2568,
+      "epoch": 3.3358561285904784,
+      "grad_norm": 0.22884266078472137,
+      "learning_rate": 1.3188260905301852e-05,
+      "loss": 0.373,
       "step": 92560
     },
     {
-      "epoch": 3.26,
-      "learning_rate": 1.4332654686084507e-05,
-      "loss": 0.2389,
+      "epoch": 3.3360363282517027,
+      "grad_norm": 0.20494328439235687,
+      "learning_rate": 1.3185689097473278e-05,
+      "loss": 0.3745,
       "step": 92565
     },
     {
-      "epoch": 3.26,
-      "learning_rate": 1.4330078382765688e-05,
-      "loss": 0.2694,
+      "epoch": 3.3362165279129274,
+      "grad_norm": 0.24264514446258545,
+      "learning_rate": 1.3183117450612398e-05,
+      "loss": 0.3718,
       "step": 92570
     },
     {
-      "epoch": 3.26,
-      "learning_rate": 1.4327502217988769e-05,
-      "loss": 0.2535,
+      "epoch": 3.336396727574152,
+      "grad_norm": 0.1675427258014679,
+      "learning_rate": 1.3180545964754257e-05,
+      "loss": 0.3765,
       "step": 92575
     },
     {
-      "epoch": 3.26,
-      "learning_rate": 1.4324926191787202e-05,
-      "loss": 0.236,
+      "epoch": 3.336576927235377,
+      "grad_norm": 0.25368401408195496,
+      "learning_rate": 1.3177974639933877e-05,
+      "loss": 0.3799,
       "step": 92580
     },
     {
-      "epoch": 3.26,
-      "learning_rate": 1.432235030419442e-05,
-      "loss": 0.2577,
+      "epoch": 3.336757126896601,
+      "grad_norm": 0.30531957745552063,
+      "learning_rate": 1.3175403476186315e-05,
+      "loss": 0.4008,
       "step": 92585
     },
     {
-      "epoch": 3.26,
-      "learning_rate": 1.4319774555243895e-05,
-      "loss": 0.2554,
+      "epoch": 3.336937326557826,
+      "grad_norm": 0.2088593989610672,
+      "learning_rate": 1.3172832473546593e-05,
+      "loss": 0.3525,
       "step": 92590
     },
     {
-      "epoch": 3.26,
-      "learning_rate": 1.4317198944969057e-05,
-      "loss": 0.2546,
+      "epoch": 3.3371175262190507,
+      "grad_norm": 0.28222382068634033,
+      "learning_rate": 1.3170261632049736e-05,
+      "loss": 0.3578,
       "step": 92595
     },
     {
-      "epoch": 3.26,
-      "learning_rate": 1.4314623473403343e-05,
-      "loss": 0.2744,
+      "epoch": 3.3372977258802754,
+      "grad_norm": 0.18309058248996735,
+      "learning_rate": 1.3167690951730777e-05,
+      "loss": 0.3696,
       "step": 92600
     },
     {
-      "epoch": 3.26,
-      "learning_rate": 1.43120481405802e-05,
-      "loss": 0.2617,
+      "epoch": 3.3374779255415,
+      "grad_norm": 0.22721010446548462,
+      "learning_rate": 1.3165120432624734e-05,
+      "loss": 0.3963,
       "step": 92605
     },
     {
-      "epoch": 3.26,
-      "learning_rate": 1.4309472946533075e-05,
-      "loss": 0.2554,
+      "epoch": 3.337658125202725,
+      "grad_norm": 0.30004847049713135,
+      "learning_rate": 1.3162550074766627e-05,
+      "loss": 0.4081,
       "step": 92610
     },
     {
-      "epoch": 3.26,
-      "learning_rate": 1.4306897891295403e-05,
-      "loss": 0.2573,
+      "epoch": 3.337838324863949,
+      "grad_norm": 0.29648730158805847,
+      "learning_rate": 1.3159979878191503e-05,
+      "loss": 0.3961,
       "step": 92615
     },
     {
-      "epoch": 3.26,
-      "learning_rate": 1.4304322974900605e-05,
-      "loss": 0.2565,
+      "epoch": 3.338018524525174,
+      "grad_norm": 0.19478410482406616,
+      "learning_rate": 1.3157409842934338e-05,
+      "loss": 0.3629,
       "step": 92620
     },
     {
-      "epoch": 3.26,
-      "learning_rate": 1.4301748197382131e-05,
-      "loss": 0.2631,
+      "epoch": 3.3381987241863986,
+      "grad_norm": 0.2053857296705246,
+      "learning_rate": 1.3154839969030186e-05,
+      "loss": 0.3789,
       "step": 92625
     },
     {
-      "epoch": 3.26,
-      "learning_rate": 1.429917355877341e-05,
-      "loss": 0.269,
+      "epoch": 3.3383789238476234,
+      "grad_norm": 0.19927579164505005,
+      "learning_rate": 1.3152270256514044e-05,
+      "loss": 0.3949,
       "step": 92630
     },
     {
-      "epoch": 3.26,
-      "learning_rate": 1.4296599059107865e-05,
-      "loss": 0.2527,
+      "epoch": 3.3385591235088476,
+      "grad_norm": 0.2706872224807739,
+      "learning_rate": 1.3149700705420923e-05,
+      "loss": 0.3705,
       "step": 92635
     },
     {
-      "epoch": 3.26,
-      "learning_rate": 1.4294024698418917e-05,
-      "loss": 0.2742,
+      "epoch": 3.3387393231700724,
+      "grad_norm": 0.2197032868862152,
+      "learning_rate": 1.3147131315785837e-05,
+      "loss": 0.3917,
       "step": 92640
     },
     {
-      "epoch": 3.26,
-      "learning_rate": 1.4291450476740018e-05,
-      "loss": 0.2426,
+      "epoch": 3.338919522831297,
+      "grad_norm": 0.2595154643058777,
+      "learning_rate": 1.3144562087643794e-05,
+      "loss": 0.3981,
       "step": 92645
     },
     {
-      "epoch": 3.26,
-      "learning_rate": 1.4288876394104572e-05,
-      "loss": 0.2568,
+      "epoch": 3.339099722492522,
+      "grad_norm": 0.20744380354881287,
+      "learning_rate": 1.3141993021029786e-05,
+      "loss": 0.3941,
       "step": 92650
     },
     {
-      "epoch": 3.26,
-      "learning_rate": 1.4286302450546003e-05,
-      "loss": 0.2631,
+      "epoch": 3.3392799221537466,
+      "grad_norm": 0.2614065408706665,
+      "learning_rate": 1.3139424115978838e-05,
+      "loss": 0.3667,
       "step": 92655
     },
     {
-      "epoch": 3.26,
-      "learning_rate": 1.4283728646097727e-05,
-      "loss": 0.2552,
+      "epoch": 3.339460121814971,
+      "grad_norm": 0.26152557134628296,
+      "learning_rate": 1.3136855372525939e-05,
+      "loss": 0.3953,
       "step": 92660
     },
     {
-      "epoch": 3.26,
-      "learning_rate": 1.4281154980793187e-05,
-      "loss": 0.2428,
+      "epoch": 3.3396403214761956,
+      "grad_norm": 0.2180803269147873,
+      "learning_rate": 1.313428679070609e-05,
+      "loss": 0.3865,
       "step": 92665
     },
     {
-      "epoch": 3.26,
-      "learning_rate": 1.4278581454665787e-05,
-      "loss": 0.2668,
+      "epoch": 3.3398205211374203,
+      "grad_norm": 0.21196460723876953,
+      "learning_rate": 1.3131718370554286e-05,
+      "loss": 0.3598,
       "step": 92670
     },
     {
-      "epoch": 3.26,
-      "learning_rate": 1.4276008067748941e-05,
-      "loss": 0.2527,
+      "epoch": 3.340000720798645,
+      "grad_norm": 0.23230072855949402,
+      "learning_rate": 1.312915011210551e-05,
+      "loss": 0.3827,
       "step": 92675
     },
     {
-      "epoch": 3.26,
-      "learning_rate": 1.4273434820076053e-05,
-      "loss": 0.2457,
+      "epoch": 3.3401809204598694,
+      "grad_norm": 0.24028299748897552,
+      "learning_rate": 1.3126582015394778e-05,
+      "loss": 0.41,
       "step": 92680
     },
     {
-      "epoch": 3.26,
-      "learning_rate": 1.4270861711680553e-05,
-      "loss": 0.252,
+      "epoch": 3.340361120121094,
+      "grad_norm": 0.29149261116981506,
+      "learning_rate": 1.3124014080457076e-05,
+      "loss": 0.3888,
       "step": 92685
     },
     {
-      "epoch": 3.26,
-      "learning_rate": 1.4268288742595842e-05,
-      "loss": 0.2754,
+      "epoch": 3.340541319782319,
+      "grad_norm": 0.2105678766965866,
+      "learning_rate": 1.3121446307327362e-05,
+      "loss": 0.3682,
       "step": 92690
     },
     {
-      "epoch": 3.26,
-      "learning_rate": 1.4265715912855331e-05,
-      "loss": 0.2431,
+      "epoch": 3.3407215194435436,
+      "grad_norm": 0.29433146119117737,
+      "learning_rate": 1.3118878696040655e-05,
+      "loss": 0.3464,
       "step": 92695
     },
     {
-      "epoch": 3.26,
-      "learning_rate": 1.4263143222492414e-05,
-      "loss": 0.2656,
+      "epoch": 3.3409017191047683,
+      "grad_norm": 0.19459545612335205,
+      "learning_rate": 1.311631124663191e-05,
+      "loss": 0.372,
       "step": 92700
     },
     {
-      "epoch": 3.26,
-      "learning_rate": 1.426057067154052e-05,
-      "loss": 0.2387,
+      "epoch": 3.3410819187659926,
+      "grad_norm": 0.22271791100502014,
+      "learning_rate": 1.3113743959136137e-05,
+      "loss": 0.3698,
       "step": 92705
     },
     {
-      "epoch": 3.26,
-      "learning_rate": 1.4257998260033024e-05,
-      "loss": 0.2546,
+      "epoch": 3.3412621184272173,
+      "grad_norm": 0.24294120073318481,
+      "learning_rate": 1.3111176833588312e-05,
+      "loss": 0.3947,
       "step": 92710
     },
     {
-      "epoch": 3.26,
-      "learning_rate": 1.425542598800335e-05,
-      "loss": 0.2652,
+      "epoch": 3.341442318088442,
+      "grad_norm": 0.2713625729084015,
+      "learning_rate": 1.3108609870023381e-05,
+      "loss": 0.3898,
       "step": 92715
     },
     {
-      "epoch": 3.26,
-      "learning_rate": 1.4252853855484883e-05,
-      "loss": 0.2448,
+      "epoch": 3.341622517749667,
+      "grad_norm": 0.24030660092830658,
+      "learning_rate": 1.3106043068476348e-05,
+      "loss": 0.3998,
       "step": 92720
     },
     {
-      "epoch": 3.26,
-      "learning_rate": 1.4250281862511034e-05,
-      "loss": 0.2666,
+      "epoch": 3.341802717410891,
+      "grad_norm": 0.23403751850128174,
+      "learning_rate": 1.3103476428982172e-05,
+      "loss": 0.4276,
       "step": 92725
     },
     {
-      "epoch": 3.26,
-      "learning_rate": 1.4247710009115192e-05,
-      "loss": 0.2428,
+      "epoch": 3.341982917072116,
+      "grad_norm": 0.23179000616073608,
+      "learning_rate": 1.3100909951575829e-05,
+      "loss": 0.3936,
       "step": 92730
     },
     {
-      "epoch": 3.26,
-      "learning_rate": 1.424513829533074e-05,
-      "loss": 0.2671,
+      "epoch": 3.3421631167333405,
+      "grad_norm": 0.2009132206439972,
+      "learning_rate": 1.3098343636292287e-05,
+      "loss": 0.3873,
       "step": 92735
     },
     {
-      "epoch": 3.26,
-      "learning_rate": 1.4242566721191087e-05,
-      "loss": 0.2536,
+      "epoch": 3.3423433163945653,
+      "grad_norm": 0.22999610006809235,
+      "learning_rate": 1.3095777483166493e-05,
+      "loss": 0.3685,
       "step": 92740
     },
     {
-      "epoch": 3.26,
-      "learning_rate": 1.4239995286729615e-05,
-      "loss": 0.2517,
+      "epoch": 3.34252351605579,
+      "grad_norm": 0.19379504024982452,
+      "learning_rate": 1.3093211492233443e-05,
+      "loss": 0.3747,
       "step": 92745
     },
     {
-      "epoch": 3.26,
-      "learning_rate": 1.4237423991979715e-05,
-      "loss": 0.2407,
+      "epoch": 3.3427037157170143,
+      "grad_norm": 0.21033187210559845,
+      "learning_rate": 1.3090645663528075e-05,
+      "loss": 0.3403,
       "step": 92750
     },
     {
-      "epoch": 3.26,
-      "learning_rate": 1.423485283697476e-05,
-      "loss": 0.244,
+      "epoch": 3.342883915378239,
+      "grad_norm": 0.19492049515247345,
+      "learning_rate": 1.3088079997085356e-05,
+      "loss": 0.3598,
       "step": 92755
     },
     {
-      "epoch": 3.26,
-      "learning_rate": 1.423228182174816e-05,
-      "loss": 0.2608,
+      "epoch": 3.3430641150394638,
+      "grad_norm": 0.21218262612819672,
+      "learning_rate": 1.3085514492940243e-05,
+      "loss": 0.3883,
       "step": 92760
     },
     {
-      "epoch": 3.26,
-      "learning_rate": 1.4229710946333272e-05,
-      "loss": 0.2484,
+      "epoch": 3.3432443147006885,
+      "grad_norm": 0.21746402978897095,
+      "learning_rate": 1.3082949151127688e-05,
+      "loss": 0.3835,
       "step": 92765
     },
     {
-      "epoch": 3.26,
-      "learning_rate": 1.42271402107635e-05,
-      "loss": 0.2848,
+      "epoch": 3.343424514361913,
+      "grad_norm": 0.22117747366428375,
+      "learning_rate": 1.308038397168263e-05,
+      "loss": 0.3615,
       "step": 92770
     },
     {
-      "epoch": 3.26,
-      "learning_rate": 1.4224569615072203e-05,
-      "loss": 0.2537,
+      "epoch": 3.3436047140231375,
+      "grad_norm": 0.19114363193511963,
+      "learning_rate": 1.3077818954640048e-05,
+      "loss": 0.3707,
       "step": 92775
     },
     {
-      "epoch": 3.26,
-      "learning_rate": 1.4221999159292776e-05,
-      "loss": 0.2552,
+      "epoch": 3.3437849136843623,
+      "grad_norm": 0.19133096933364868,
+      "learning_rate": 1.307525410003487e-05,
+      "loss": 0.3634,
       "step": 92780
     },
     {
-      "epoch": 3.26,
-      "learning_rate": 1.4219428843458588e-05,
-      "loss": 0.2513,
+      "epoch": 3.343965113345587,
+      "grad_norm": 0.21959801018238068,
+      "learning_rate": 1.3072689407902047e-05,
+      "loss": 0.3866,
       "step": 92785
     },
     {
-      "epoch": 3.26,
-      "learning_rate": 1.4216858667603015e-05,
-      "loss": 0.2473,
+      "epoch": 3.3441453130068117,
+      "grad_norm": 0.26673710346221924,
+      "learning_rate": 1.3070124878276524e-05,
+      "loss": 0.3999,
       "step": 92790
     },
     {
-      "epoch": 3.26,
-      "learning_rate": 1.4214288631759413e-05,
-      "loss": 0.2535,
+      "epoch": 3.344325512668036,
+      "grad_norm": 0.27120256423950195,
+      "learning_rate": 1.3067560511193227e-05,
+      "loss": 0.3931,
       "step": 92795
     },
     {
-      "epoch": 3.26,
-      "learning_rate": 1.4211718735961174e-05,
-      "loss": 0.2803,
+      "epoch": 3.3445057123292607,
+      "grad_norm": 0.22337371110916138,
+      "learning_rate": 1.3064996306687127e-05,
+      "loss": 0.3907,
       "step": 92800
     },
     {
-      "epoch": 3.27,
-      "learning_rate": 1.420914898024166e-05,
-      "loss": 0.2466,
+      "epoch": 3.3446859119904855,
+      "grad_norm": 0.22518573701381683,
+      "learning_rate": 1.3062432264793122e-05,
+      "loss": 0.3561,
       "step": 92805
     },
     {
-      "epoch": 3.27,
-      "learning_rate": 1.4206579364634231e-05,
-      "loss": 0.2617,
+      "epoch": 3.34486611165171,
+      "grad_norm": 0.23381398618221283,
+      "learning_rate": 1.3059868385546181e-05,
+      "loss": 0.3455,
       "step": 92810
     },
     {
-      "epoch": 3.27,
-      "learning_rate": 1.4204009889172248e-05,
-      "loss": 0.2561,
+      "epoch": 3.3450463113129345,
+      "grad_norm": 0.20269441604614258,
+      "learning_rate": 1.3057304668981218e-05,
+      "loss": 0.3855,
       "step": 92815
     },
     {
-      "epoch": 3.27,
-      "learning_rate": 1.4201440553889095e-05,
-      "loss": 0.2463,
+      "epoch": 3.3452265109741592,
+      "grad_norm": 0.20035311579704285,
+      "learning_rate": 1.3054741115133168e-05,
+      "loss": 0.3773,
       "step": 92820
     },
     {
-      "epoch": 3.27,
-      "learning_rate": 1.4198871358818106e-05,
-      "loss": 0.2531,
+      "epoch": 3.345406710635384,
+      "grad_norm": 0.1864159256219864,
+      "learning_rate": 1.3052177724036957e-05,
+      "loss": 0.3679,
       "step": 92825
     },
     {
-      "epoch": 3.27,
-      "learning_rate": 1.4196302303992668e-05,
-      "loss": 0.2655,
+      "epoch": 3.3455869102966087,
+      "grad_norm": 0.22926171123981476,
+      "learning_rate": 1.3049614495727507e-05,
+      "loss": 0.348,
       "step": 92830
     },
     {
-      "epoch": 3.27,
-      "learning_rate": 1.4193733389446113e-05,
-      "loss": 0.2617,
+      "epoch": 3.3457671099578334,
+      "grad_norm": 0.27731531858444214,
+      "learning_rate": 1.3047051430239744e-05,
+      "loss": 0.3804,
       "step": 92835
     },
     {
-      "epoch": 3.27,
-      "learning_rate": 1.4191164615211816e-05,
-      "loss": 0.2699,
+      "epoch": 3.3459473096190577,
+      "grad_norm": 0.2885668873786926,
+      "learning_rate": 1.3044488527608595e-05,
+      "loss": 0.397,
       "step": 92840
     },
     {
-      "epoch": 3.27,
-      "learning_rate": 1.4188595981323124e-05,
-      "loss": 0.2499,
+      "epoch": 3.3461275092802825,
+      "grad_norm": 0.23300670087337494,
+      "learning_rate": 1.3041925787868978e-05,
+      "loss": 0.3717,
       "step": 92845
     },
     {
-      "epoch": 3.27,
-      "learning_rate": 1.418602748781338e-05,
-      "loss": 0.2351,
+      "epoch": 3.346307708941507,
+      "grad_norm": 0.2299395352602005,
+      "learning_rate": 1.3039363211055806e-05,
+      "loss": 0.3944,
       "step": 92850
     },
     {
-      "epoch": 3.27,
-      "learning_rate": 1.4183459134715952e-05,
-      "loss": 0.2457,
+      "epoch": 3.346487908602732,
+      "grad_norm": 0.23562583327293396,
+      "learning_rate": 1.3036800797203997e-05,
+      "loss": 0.3838,
       "step": 92855
     },
     {
-      "epoch": 3.27,
-      "learning_rate": 1.4180890922064177e-05,
-      "loss": 0.2566,
+      "epoch": 3.346668108263956,
+      "grad_norm": 0.18811137974262238,
+      "learning_rate": 1.3034238546348449e-05,
+      "loss": 0.4058,
       "step": 92860
     },
     {
-      "epoch": 3.27,
-      "learning_rate": 1.4178322849891404e-05,
-      "loss": 0.2558,
+      "epoch": 3.346848307925181,
+      "grad_norm": 0.3136909306049347,
+      "learning_rate": 1.3031676458524092e-05,
+      "loss": 0.3802,
       "step": 92865
     },
     {
-      "epoch": 3.27,
-      "learning_rate": 1.4175754918230966e-05,
-      "loss": 0.256,
+      "epoch": 3.3470285075864057,
+      "grad_norm": 0.2185976505279541,
+      "learning_rate": 1.302911453376583e-05,
+      "loss": 0.3782,
       "step": 92870
     },
     {
-      "epoch": 3.27,
-      "learning_rate": 1.417318712711623e-05,
-      "loss": 0.2528,
+      "epoch": 3.3472087072476304,
+      "grad_norm": 0.2267122119665146,
+      "learning_rate": 1.302655277210856e-05,
+      "loss": 0.3729,
       "step": 92875
     },
     {
-      "epoch": 3.27,
-      "learning_rate": 1.4170619476580513e-05,
-      "loss": 0.2599,
+      "epoch": 3.347388906908855,
+      "grad_norm": 0.20721164345741272,
+      "learning_rate": 1.3023991173587193e-05,
+      "loss": 0.3868,
       "step": 92880
     },
     {
-      "epoch": 3.27,
-      "learning_rate": 1.4168051966657176e-05,
-      "loss": 0.2291,
+      "epoch": 3.34756910657008,
+      "grad_norm": 0.21538670361042023,
+      "learning_rate": 1.3021429738236618e-05,
+      "loss": 0.3675,
       "step": 92885
     },
     {
-      "epoch": 3.27,
-      "learning_rate": 1.4165484597379534e-05,
-      "loss": 0.2616,
+      "epoch": 3.347749306231304,
+      "grad_norm": 0.24019305408000946,
+      "learning_rate": 1.3018868466091755e-05,
+      "loss": 0.3555,
       "step": 92890
     },
     {
-      "epoch": 3.27,
-      "learning_rate": 1.4162917368780943e-05,
-      "loss": 0.2558,
+      "epoch": 3.347929505892529,
+      "grad_norm": 0.25520575046539307,
+      "learning_rate": 1.3016307357187496e-05,
+      "loss": 0.4072,
       "step": 92895
     },
     {
-      "epoch": 3.27,
-      "learning_rate": 1.416035028089473e-05,
-      "loss": 0.2478,
+      "epoch": 3.3481097055537536,
+      "grad_norm": 0.21573813259601593,
+      "learning_rate": 1.3013746411558714e-05,
+      "loss": 0.3914,
       "step": 92900
     },
     {
-      "epoch": 3.27,
-      "learning_rate": 1.4157783333754226e-05,
-      "loss": 0.2743,
+      "epoch": 3.3482899052149784,
+      "grad_norm": 0.22230984270572662,
+      "learning_rate": 1.3011185629240321e-05,
+      "loss": 0.3971,
       "step": 92905
     },
     {
-      "epoch": 3.27,
-      "learning_rate": 1.4155216527392747e-05,
-      "loss": 0.2864,
+      "epoch": 3.3484701048762027,
+      "grad_norm": 0.21699829399585724,
+      "learning_rate": 1.3008625010267206e-05,
+      "loss": 0.373,
       "step": 92910
     },
     {
-      "epoch": 3.27,
-      "learning_rate": 1.4152649861843648e-05,
-      "loss": 0.2724,
+      "epoch": 3.3486503045374274,
+      "grad_norm": 0.27921581268310547,
+      "learning_rate": 1.3006064554674252e-05,
+      "loss": 0.3984,
       "step": 92915
     },
     {
-      "epoch": 3.27,
-      "learning_rate": 1.415008333714024e-05,
-      "loss": 0.2614,
+      "epoch": 3.348830504198652,
+      "grad_norm": 0.20376195013523102,
+      "learning_rate": 1.3003504262496347e-05,
+      "loss": 0.3565,
       "step": 92920
     },
     {
-      "epoch": 3.27,
-      "learning_rate": 1.4147516953315836e-05,
-      "loss": 0.2682,
+      "epoch": 3.349010703859877,
+      "grad_norm": 0.26665177941322327,
+      "learning_rate": 1.3000944133768368e-05,
+      "loss": 0.3716,
       "step": 92925
     },
     {
-      "epoch": 3.27,
-      "learning_rate": 1.4144950710403788e-05,
-      "loss": 0.2617,
+      "epoch": 3.3491909035211016,
+      "grad_norm": 0.22333024442195892,
+      "learning_rate": 1.2998384168525208e-05,
+      "loss": 0.3586,
       "step": 92930
     },
     {
-      "epoch": 3.27,
-      "learning_rate": 1.414238460843739e-05,
-      "loss": 0.2711,
+      "epoch": 3.349371103182326,
+      "grad_norm": 0.26054447889328003,
+      "learning_rate": 1.299582436680174e-05,
+      "loss": 0.4145,
       "step": 92935
     },
     {
-      "epoch": 3.27,
-      "learning_rate": 1.4139818647449981e-05,
-      "loss": 0.2508,
+      "epoch": 3.3495513028435506,
+      "grad_norm": 0.25818753242492676,
+      "learning_rate": 1.299326472863284e-05,
+      "loss": 0.3409,
       "step": 92940
     },
     {
-      "epoch": 3.27,
-      "learning_rate": 1.413725282747487e-05,
-      "loss": 0.2638,
+      "epoch": 3.3497315025047754,
+      "grad_norm": 0.17389363050460815,
+      "learning_rate": 1.2990705254053385e-05,
+      "loss": 0.3626,
       "step": 92945
     },
     {
-      "epoch": 3.27,
-      "learning_rate": 1.4134687148545362e-05,
-      "loss": 0.2476,
+      "epoch": 3.349911702166,
+      "grad_norm": 0.24069735407829285,
+      "learning_rate": 1.2988145943098245e-05,
+      "loss": 0.3704,
       "step": 92950
     },
     {
-      "epoch": 3.27,
-      "learning_rate": 1.4132121610694792e-05,
-      "loss": 0.2468,
+      "epoch": 3.3500919018272244,
+      "grad_norm": 0.2535139322280884,
+      "learning_rate": 1.2985586795802279e-05,
+      "loss": 0.3725,
       "step": 92955
     },
     {
-      "epoch": 3.27,
-      "learning_rate": 1.412955621395646e-05,
-      "loss": 0.2461,
+      "epoch": 3.350272101488449,
+      "grad_norm": 0.2423035055398941,
+      "learning_rate": 1.2983027812200382e-05,
+      "loss": 0.3808,
       "step": 92960
     },
     {
-      "epoch": 3.27,
-      "learning_rate": 1.4126990958363668e-05,
-      "loss": 0.2656,
+      "epoch": 3.350452301149674,
+      "grad_norm": 0.2165379524230957,
+      "learning_rate": 1.2980468992327402e-05,
+      "loss": 0.3942,
       "step": 92965
     },
     {
-      "epoch": 3.27,
-      "learning_rate": 1.412442584394974e-05,
-      "loss": 0.2755,
+      "epoch": 3.3506325008108986,
+      "grad_norm": 0.27460017800331116,
+      "learning_rate": 1.2977910336218204e-05,
+      "loss": 0.3583,
       "step": 92970
     },
     {
-      "epoch": 3.27,
-      "learning_rate": 1.4121860870747978e-05,
-      "loss": 0.2617,
+      "epoch": 3.3508127004721233,
+      "grad_norm": 0.2811183035373688,
+      "learning_rate": 1.2975351843907652e-05,
+      "loss": 0.369,
       "step": 92975
     },
     {
-      "epoch": 3.27,
-      "learning_rate": 1.4119296038791683e-05,
-      "loss": 0.2502,
+      "epoch": 3.3509929001333476,
+      "grad_norm": 0.22959589958190918,
+      "learning_rate": 1.2972793515430592e-05,
+      "loss": 0.3604,
       "step": 92980
     },
     {
-      "epoch": 3.27,
-      "learning_rate": 1.4116731348114148e-05,
-      "loss": 0.2519,
+      "epoch": 3.3511730997945723,
+      "grad_norm": 0.2398064136505127,
+      "learning_rate": 1.2970235350821911e-05,
+      "loss": 0.3713,
       "step": 92985
     },
     {
-      "epoch": 3.27,
-      "learning_rate": 1.4114166798748684e-05,
-      "loss": 0.2423,
+      "epoch": 3.351353299455797,
+      "grad_norm": 0.23232856392860413,
+      "learning_rate": 1.2967677350116425e-05,
+      "loss": 0.3658,
       "step": 92990
     },
     {
-      "epoch": 3.27,
-      "learning_rate": 1.4111602390728602e-05,
-      "loss": 0.2438,
+      "epoch": 3.351533499117022,
+      "grad_norm": 0.2607458829879761,
+      "learning_rate": 1.2965119513349016e-05,
+      "loss": 0.3745,
       "step": 92995
     },
     {
-      "epoch": 3.27,
-      "learning_rate": 1.4109038124087181e-05,
-      "loss": 0.2675,
+      "epoch": 3.351713698778246,
+      "grad_norm": 0.17985029518604279,
+      "learning_rate": 1.2962561840554524e-05,
+      "loss": 0.3852,
       "step": 93000
     },
     {
-      "epoch": 3.27,
-      "eval_loss": 0.2542540431022644,
-      "eval_runtime": 10.5543,
-      "eval_samples_per_second": 9.475,
-      "eval_steps_per_second": 9.475,
+      "epoch": 3.351713698778246,
+      "eval_loss": 0.4302482306957245,
+      "eval_runtime": 3.5321,
+      "eval_samples_per_second": 28.312,
+      "eval_steps_per_second": 7.078,
       "step": 93000
     },
     {
-      "epoch": 3.27,
-      "learning_rate": 1.4106473998857717e-05,
-      "loss": 0.2569,
+      "epoch": 3.351893898439471,
+      "grad_norm": 0.22021250426769257,
+      "learning_rate": 1.2960004331767786e-05,
+      "loss": 0.36,
       "step": 93005
     },
     {
-      "epoch": 3.27,
-      "learning_rate": 1.4103910015073518e-05,
-      "loss": 0.2748,
+      "epoch": 3.3520740981006956,
+      "grad_norm": 0.24476410448551178,
+      "learning_rate": 1.2957446987023678e-05,
+      "loss": 0.4055,
       "step": 93010
     },
     {
-      "epoch": 3.27,
-      "learning_rate": 1.4101346172767871e-05,
-      "loss": 0.2652,
+      "epoch": 3.3522542977619203,
+      "grad_norm": 0.20762693881988525,
+      "learning_rate": 1.2954889806357012e-05,
+      "loss": 0.3946,
       "step": 93015
     },
     {
-      "epoch": 3.27,
-      "learning_rate": 1.4098782471974057e-05,
-      "loss": 0.2259,
+      "epoch": 3.352434497423145,
+      "grad_norm": 0.30094394087791443,
+      "learning_rate": 1.2952332789802631e-05,
+      "loss": 0.3695,
       "step": 93020
     },
     {
-      "epoch": 3.27,
-      "learning_rate": 1.4096218912725356e-05,
-      "loss": 0.2508,
+      "epoch": 3.3526146970843693,
+      "grad_norm": 0.27970486879348755,
+      "learning_rate": 1.2949775937395392e-05,
+      "loss": 0.3885,
       "step": 93025
     },
     {
-      "epoch": 3.27,
-      "learning_rate": 1.4093655495055081e-05,
-      "loss": 0.2728,
+      "epoch": 3.352794896745594,
+      "grad_norm": 0.2970288395881653,
+      "learning_rate": 1.2947219249170128e-05,
+      "loss": 0.3899,
       "step": 93030
     },
     {
-      "epoch": 3.27,
-      "learning_rate": 1.4091092218996502e-05,
-      "loss": 0.2435,
+      "epoch": 3.352975096406819,
+      "grad_norm": 0.18664143979549408,
+      "learning_rate": 1.2944662725161663e-05,
+      "loss": 0.3715,
       "step": 93035
     },
     {
-      "epoch": 3.27,
-      "learning_rate": 1.408852908458289e-05,
-      "loss": 0.2542,
+      "epoch": 3.3531552960680435,
+      "grad_norm": 0.2144254446029663,
+      "learning_rate": 1.2942106365404839e-05,
+      "loss": 0.3757,
       "step": 93040
     },
     {
-      "epoch": 3.27,
-      "learning_rate": 1.4085966091847535e-05,
-      "loss": 0.2551,
+      "epoch": 3.353335495729268,
+      "grad_norm": 0.24365659058094025,
+      "learning_rate": 1.2939550169934466e-05,
+      "loss": 0.3741,
       "step": 93045
     },
     {
-      "epoch": 3.27,
-      "learning_rate": 1.4083403240823732e-05,
-      "loss": 0.2636,
+      "epoch": 3.3535156953904925,
+      "grad_norm": 0.2775290012359619,
+      "learning_rate": 1.29369941387854e-05,
+      "loss": 0.3854,
       "step": 93050
     },
     {
-      "epoch": 3.27,
-      "learning_rate": 1.408084053154474e-05,
-      "loss": 0.2521,
+      "epoch": 3.3536958950517173,
+      "grad_norm": 0.20349471271038055,
+      "learning_rate": 1.2934438271992452e-05,
+      "loss": 0.3809,
       "step": 93055
     },
     {
-      "epoch": 3.27,
-      "learning_rate": 1.407827796404384e-05,
-      "loss": 0.2496,
+      "epoch": 3.353876094712942,
+      "grad_norm": 0.27783820033073425,
+      "learning_rate": 1.2931882569590448e-05,
+      "loss": 0.353,
       "step": 93060
     },
     {
-      "epoch": 3.27,
-      "learning_rate": 1.407571553835429e-05,
-      "loss": 0.2704,
+      "epoch": 3.3540562943741667,
+      "grad_norm": 0.26392999291419983,
+      "learning_rate": 1.2929327031614204e-05,
+      "loss": 0.4006,
       "step": 93065
     },
     {
-      "epoch": 3.27,
-      "learning_rate": 1.4073153254509386e-05,
-      "loss": 0.2535,
+      "epoch": 3.354236494035391,
+      "grad_norm": 0.2420482635498047,
+      "learning_rate": 1.2926771658098535e-05,
+      "loss": 0.3974,
       "step": 93070
     },
     {
-      "epoch": 3.27,
-      "learning_rate": 1.4070591112542387e-05,
-      "loss": 0.2552,
+      "epoch": 3.3544166936966158,
+      "grad_norm": 0.2563861012458801,
+      "learning_rate": 1.2924216449078274e-05,
+      "loss": 0.3876,
       "step": 93075
     },
     {
-      "epoch": 3.27,
-      "learning_rate": 1.4068029112486547e-05,
-      "loss": 0.2795,
+      "epoch": 3.3545968933578405,
+      "grad_norm": 0.23027673363685608,
+      "learning_rate": 1.292166140458824e-05,
+      "loss": 0.3777,
       "step": 93080
     },
     {
-      "epoch": 3.27,
-      "learning_rate": 1.4065467254375159e-05,
-      "loss": 0.2735,
+      "epoch": 3.3547770930190652,
+      "grad_norm": 0.2890183925628662,
+      "learning_rate": 1.2919106524663208e-05,
+      "loss": 0.3627,
       "step": 93085
     },
     {
-      "epoch": 3.28,
-      "learning_rate": 1.406290553824147e-05,
-      "loss": 0.2424,
+      "epoch": 3.3549572926802895,
+      "grad_norm": 0.2011183500289917,
+      "learning_rate": 1.2916551809338018e-05,
+      "loss": 0.3793,
       "step": 93090
     },
     {
-      "epoch": 3.28,
-      "learning_rate": 1.4060343964118733e-05,
-      "loss": 0.273,
+      "epoch": 3.3551374923415143,
+      "grad_norm": 0.23442140221595764,
+      "learning_rate": 1.2913997258647476e-05,
+      "loss": 0.3532,
       "step": 93095
     },
     {
-      "epoch": 3.28,
-      "learning_rate": 1.405778253204022e-05,
-      "loss": 0.2706,
+      "epoch": 3.355317692002739,
+      "grad_norm": 0.18820983171463013,
+      "learning_rate": 1.2911442872626376e-05,
+      "loss": 0.3741,
       "step": 93100
     },
     {
-      "epoch": 3.28,
-      "learning_rate": 1.4055221242039202e-05,
-      "loss": 0.2269,
+      "epoch": 3.3554978916639637,
+      "grad_norm": 0.17393651604652405,
+      "learning_rate": 1.2908888651309528e-05,
+      "loss": 0.3554,
       "step": 93105
     },
     {
-      "epoch": 3.28,
-      "learning_rate": 1.4052660094148922e-05,
-      "loss": 0.248,
+      "epoch": 3.3556780913251885,
+      "grad_norm": 0.2209923267364502,
+      "learning_rate": 1.2906334594731722e-05,
+      "loss": 0.4142,
       "step": 93110
     },
     {
-      "epoch": 3.28,
-      "learning_rate": 1.405009908840264e-05,
-      "loss": 0.254,
+      "epoch": 3.355858290986413,
+      "grad_norm": 0.2804467976093292,
+      "learning_rate": 1.2903780702927776e-05,
+      "loss": 0.3901,
       "step": 93115
     },
     {
-      "epoch": 3.28,
-      "learning_rate": 1.404753822483359e-05,
-      "loss": 0.2408,
+      "epoch": 3.3560384906476375,
+      "grad_norm": 0.2660714089870453,
+      "learning_rate": 1.2901226975932474e-05,
+      "loss": 0.3976,
       "step": 93120
     },
     {
-      "epoch": 3.28,
-      "learning_rate": 1.4044977503475051e-05,
-      "loss": 0.266,
+      "epoch": 3.356218690308862,
+      "grad_norm": 0.24791589379310608,
+      "learning_rate": 1.2898673413780616e-05,
+      "loss": 0.4206,
       "step": 93125
     },
     {
-      "epoch": 3.28,
-      "learning_rate": 1.4042416924360257e-05,
-      "loss": 0.2522,
+      "epoch": 3.356398889970087,
+      "grad_norm": 0.25095242261886597,
+      "learning_rate": 1.2896120016506985e-05,
+      "loss": 0.373,
       "step": 93130
     },
     {
-      "epoch": 3.28,
-      "learning_rate": 1.4039856487522463e-05,
-      "loss": 0.2745,
+      "epoch": 3.3565790896313117,
+      "grad_norm": 0.31840965151786804,
+      "learning_rate": 1.2893566784146377e-05,
+      "loss": 0.3655,
       "step": 93135
     },
     {
-      "epoch": 3.28,
-      "learning_rate": 1.4037296192994893e-05,
-      "loss": 0.2434,
+      "epoch": 3.356759289292536,
+      "grad_norm": 0.20555394887924194,
+      "learning_rate": 1.2891013716733569e-05,
+      "loss": 0.3744,
       "step": 93140
     },
     {
-      "epoch": 3.28,
-      "learning_rate": 1.403473604081082e-05,
-      "loss": 0.2449,
+      "epoch": 3.3569394889537607,
+      "grad_norm": 0.24305355548858643,
+      "learning_rate": 1.2888460814303363e-05,
+      "loss": 0.3667,
       "step": 93145
     },
     {
-      "epoch": 3.28,
-      "learning_rate": 1.4032176031003463e-05,
-      "loss": 0.2645,
+      "epoch": 3.3571196886149854,
+      "grad_norm": 0.22731393575668335,
+      "learning_rate": 1.2885908076890536e-05,
+      "loss": 0.3577,
       "step": 93150
     },
     {
-      "epoch": 3.28,
-      "learning_rate": 1.4029616163606085e-05,
-      "loss": 0.2864,
+      "epoch": 3.35729988827621,
+      "grad_norm": 0.22640764713287354,
+      "learning_rate": 1.2883355504529865e-05,
+      "loss": 0.3424,
       "step": 93155
     },
     {
-      "epoch": 3.28,
-      "learning_rate": 1.4027056438651898e-05,
-      "loss": 0.2473,
+      "epoch": 3.357480087937435,
+      "grad_norm": 0.21849682927131653,
+      "learning_rate": 1.2880803097256128e-05,
+      "loss": 0.3742,
       "step": 93160
     },
     {
-      "epoch": 3.28,
-      "learning_rate": 1.4024496856174162e-05,
-      "loss": 0.2521,
+      "epoch": 3.357660287598659,
+      "grad_norm": 0.20635724067687988,
+      "learning_rate": 1.2878250855104094e-05,
+      "loss": 0.4043,
       "step": 93165
     },
     {
-      "epoch": 3.28,
-      "learning_rate": 1.4021937416206101e-05,
-      "loss": 0.2574,
+      "epoch": 3.357840487259884,
+      "grad_norm": 0.17877210676670074,
+      "learning_rate": 1.2875698778108552e-05,
+      "loss": 0.3761,
       "step": 93170
     },
     {
-      "epoch": 3.28,
-      "learning_rate": 1.4019378118780951e-05,
-      "loss": 0.2324,
+      "epoch": 3.3580206869211087,
+      "grad_norm": 0.19038045406341553,
+      "learning_rate": 1.287314686630427e-05,
+      "loss": 0.3983,
       "step": 93175
     },
     {
-      "epoch": 3.28,
-      "learning_rate": 1.4016818963931927e-05,
-      "loss": 0.2426,
+      "epoch": 3.3582008865823334,
+      "grad_norm": 0.25942298769950867,
+      "learning_rate": 1.2870595119726015e-05,
+      "loss": 0.42,
       "step": 93180
     },
     {
-      "epoch": 3.28,
-      "learning_rate": 1.4014259951692279e-05,
-      "loss": 0.244,
+      "epoch": 3.3583810862435577,
+      "grad_norm": 0.2063370645046234,
+      "learning_rate": 1.2868043538408553e-05,
+      "loss": 0.4054,
       "step": 93185
     },
     {
-      "epoch": 3.28,
-      "learning_rate": 1.401170108209523e-05,
-      "loss": 0.2634,
+      "epoch": 3.3585612859047824,
+      "grad_norm": 0.22962912917137146,
+      "learning_rate": 1.2865492122386636e-05,
+      "loss": 0.3537,
       "step": 93190
     },
     {
-      "epoch": 3.28,
-      "learning_rate": 1.4009142355173996e-05,
-      "loss": 0.2655,
+      "epoch": 3.358741485566007,
+      "grad_norm": 0.21510834991931915,
+      "learning_rate": 1.2862940871695062e-05,
+      "loss": 0.3888,
       "step": 93195
     },
     {
-      "epoch": 3.28,
-      "learning_rate": 1.4006583770961798e-05,
-      "loss": 0.249,
+      "epoch": 3.358921685227232,
+      "grad_norm": 0.22031091153621674,
+      "learning_rate": 1.2860389786368559e-05,
+      "loss": 0.3752,
       "step": 93200
     },
     {
-      "epoch": 3.28,
-      "learning_rate": 1.4004025329491866e-05,
-      "loss": 0.2634,
+      "epoch": 3.3591018848884566,
+      "grad_norm": 0.2270020842552185,
+      "learning_rate": 1.2857838866441885e-05,
+      "loss": 0.3664,
       "step": 93205
     },
     {
-      "epoch": 3.28,
-      "learning_rate": 1.4001467030797428e-05,
-      "loss": 0.2556,
+      "epoch": 3.359282084549681,
+      "grad_norm": 0.22233691811561584,
+      "learning_rate": 1.2855288111949811e-05,
+      "loss": 0.3757,
       "step": 93210
     },
     {
-      "epoch": 3.28,
-      "learning_rate": 1.3998908874911679e-05,
-      "loss": 0.2586,
+      "epoch": 3.3594622842109056,
+      "grad_norm": 0.25379857420921326,
+      "learning_rate": 1.2852737522927089e-05,
+      "loss": 0.3823,
       "step": 93215
     },
     {
-      "epoch": 3.28,
-      "learning_rate": 1.399635086186786e-05,
-      "loss": 0.2485,
+      "epoch": 3.3596424838721304,
+      "grad_norm": 0.19441834092140198,
+      "learning_rate": 1.2850187099408467e-05,
+      "loss": 0.3814,
       "step": 93220
     },
     {
-      "epoch": 3.28,
-      "learning_rate": 1.3993792991699173e-05,
-      "loss": 0.2416,
+      "epoch": 3.359822683533355,
+      "grad_norm": 0.21130891144275665,
+      "learning_rate": 1.2847636841428695e-05,
+      "loss": 0.3634,
       "step": 93225
     },
     {
-      "epoch": 3.28,
-      "learning_rate": 1.3991235264438833e-05,
-      "loss": 0.2645,
+      "epoch": 3.3600028831945794,
+      "grad_norm": 0.22752270102500916,
+      "learning_rate": 1.2845086749022506e-05,
+      "loss": 0.3821,
       "step": 93230
     },
     {
-      "epoch": 3.28,
-      "learning_rate": 1.3988677680120036e-05,
-      "loss": 0.2557,
+      "epoch": 3.360183082855804,
+      "grad_norm": 0.2163986712694168,
+      "learning_rate": 1.2842536822224666e-05,
+      "loss": 0.3591,
       "step": 93235
     },
     {
-      "epoch": 3.28,
-      "learning_rate": 1.3986120238776012e-05,
-      "loss": 0.2617,
+      "epoch": 3.360363282517029,
+      "grad_norm": 0.2803928852081299,
+      "learning_rate": 1.2839987061069908e-05,
+      "loss": 0.356,
       "step": 93240
     },
     {
-      "epoch": 3.28,
-      "learning_rate": 1.398356294043996e-05,
-      "loss": 0.2646,
+      "epoch": 3.3605434821782536,
+      "grad_norm": 0.26921606063842773,
+      "learning_rate": 1.2837437465592972e-05,
+      "loss": 0.4204,
       "step": 93245
     },
     {
-      "epoch": 3.28,
-      "learning_rate": 1.3981005785145085e-05,
-      "loss": 0.2721,
+      "epoch": 3.3607236818394783,
+      "grad_norm": 0.2192206233739853,
+      "learning_rate": 1.2834888035828596e-05,
+      "loss": 0.3889,
       "step": 93250
     },
     {
-      "epoch": 3.28,
-      "learning_rate": 1.3978448772924574e-05,
-      "loss": 0.2445,
+      "epoch": 3.3609038815007026,
+      "grad_norm": 0.26289498805999756,
+      "learning_rate": 1.2832338771811508e-05,
+      "loss": 0.3843,
       "step": 93255
     },
     {
-      "epoch": 3.28,
-      "learning_rate": 1.3975891903811653e-05,
-      "loss": 0.2669,
+      "epoch": 3.3610840811619274,
+      "grad_norm": 0.2620001435279846,
+      "learning_rate": 1.2829789673576456e-05,
+      "loss": 0.3524,
       "step": 93260
     },
     {
-      "epoch": 3.28,
-      "learning_rate": 1.3973335177839503e-05,
-      "loss": 0.256,
+      "epoch": 3.361264280823152,
+      "grad_norm": 0.2221224457025528,
+      "learning_rate": 1.2827240741158175e-05,
+      "loss": 0.3894,
       "step": 93265
     },
     {
-      "epoch": 3.28,
-      "learning_rate": 1.3970778595041333e-05,
-      "loss": 0.2373,
+      "epoch": 3.361444480484377,
+      "grad_norm": 0.21585452556610107,
+      "learning_rate": 1.282469197459136e-05,
+      "loss": 0.3812,
       "step": 93270
     },
     {
-      "epoch": 3.28,
-      "learning_rate": 1.396822215545033e-05,
-      "loss": 0.2583,
+      "epoch": 3.361624680145601,
+      "grad_norm": 0.21924617886543274,
+      "learning_rate": 1.2822143373910771e-05,
+      "loss": 0.3903,
       "step": 93275
     },
     {
-      "epoch": 3.28,
-      "learning_rate": 1.3965665859099702e-05,
-      "loss": 0.2802,
+      "epoch": 3.361804879806826,
+      "grad_norm": 0.1916201114654541,
+      "learning_rate": 1.281959493915112e-05,
+      "loss": 0.386,
       "step": 93280
     },
     {
-      "epoch": 3.28,
-      "learning_rate": 1.396310970602263e-05,
-      "loss": 0.2618,
+      "epoch": 3.3619850794680506,
+      "grad_norm": 0.27522167563438416,
+      "learning_rate": 1.2817046670347121e-05,
+      "loss": 0.3835,
       "step": 93285
     },
     {
-      "epoch": 3.28,
-      "learning_rate": 1.3960553696252304e-05,
-      "loss": 0.2519,
+      "epoch": 3.3621652791292753,
+      "grad_norm": 0.25006136298179626,
+      "learning_rate": 1.2814498567533523e-05,
+      "loss": 0.3698,
       "step": 93290
     },
     {
-      "epoch": 3.28,
-      "learning_rate": 1.3957997829821903e-05,
-      "loss": 0.2962,
+      "epoch": 3.3623454787905,
+      "grad_norm": 0.25098854303359985,
+      "learning_rate": 1.2811950630745e-05,
+      "loss": 0.3341,
       "step": 93295
     },
     {
-      "epoch": 3.28,
-      "learning_rate": 1.3955442106764632e-05,
-      "loss": 0.2457,
+      "epoch": 3.3625256784517243,
+      "grad_norm": 0.25163203477859497,
+      "learning_rate": 1.28094028600163e-05,
+      "loss": 0.4086,
       "step": 93300
     },
     {
-      "epoch": 3.28,
-      "learning_rate": 1.395288652711367e-05,
-      "loss": 0.2763,
+      "epoch": 3.362705878112949,
+      "grad_norm": 0.321817547082901,
+      "learning_rate": 1.2806855255382127e-05,
+      "loss": 0.3856,
       "step": 93305
     },
     {
-      "epoch": 3.28,
-      "learning_rate": 1.3950331090902186e-05,
-      "loss": 0.2768,
+      "epoch": 3.362886077774174,
+      "grad_norm": 0.22713923454284668,
+      "learning_rate": 1.2804307816877193e-05,
+      "loss": 0.3978,
       "step": 93310
     },
     {
-      "epoch": 3.28,
-      "learning_rate": 1.3947775798163377e-05,
-      "loss": 0.291,
+      "epoch": 3.3630662774353985,
+      "grad_norm": 0.24719440937042236,
+      "learning_rate": 1.2801760544536202e-05,
+      "loss": 0.3796,
       "step": 93315
     },
     {
-      "epoch": 3.28,
-      "learning_rate": 1.3945220648930408e-05,
-      "loss": 0.2642,
+      "epoch": 3.363246477096623,
+      "grad_norm": 0.24784217774868011,
+      "learning_rate": 1.2799213438393858e-05,
+      "loss": 0.3683,
       "step": 93320
     },
     {
-      "epoch": 3.28,
-      "learning_rate": 1.3942665643236475e-05,
-      "loss": 0.2497,
+      "epoch": 3.3634266767578476,
+      "grad_norm": 0.24816350638866425,
+      "learning_rate": 1.2796666498484865e-05,
+      "loss": 0.3849,
       "step": 93325
     },
     {
-      "epoch": 3.28,
-      "learning_rate": 1.3940110781114729e-05,
-      "loss": 0.2789,
+      "epoch": 3.3636068764190723,
+      "grad_norm": 0.2373788207769394,
+      "learning_rate": 1.2794119724843934e-05,
+      "loss": 0.388,
       "step": 93330
     },
     {
-      "epoch": 3.28,
-      "learning_rate": 1.3937556062598367e-05,
-      "loss": 0.2427,
+      "epoch": 3.363787076080297,
+      "grad_norm": 0.2272450029850006,
+      "learning_rate": 1.2791573117505761e-05,
+      "loss": 0.3771,
       "step": 93335
     },
     {
-      "epoch": 3.28,
-      "learning_rate": 1.3935001487720547e-05,
-      "loss": 0.262,
+      "epoch": 3.3639672757415218,
+      "grad_norm": 0.2307075560092926,
+      "learning_rate": 1.2789026676505039e-05,
+      "loss": 0.362,
       "step": 93340
     },
     {
-      "epoch": 3.28,
-      "learning_rate": 1.3932447056514442e-05,
-      "loss": 0.2798,
+      "epoch": 3.364147475402746,
+      "grad_norm": 0.22494341433048248,
+      "learning_rate": 1.2786480401876466e-05,
+      "loss": 0.3934,
       "step": 93345
     },
     {
-      "epoch": 3.28,
-      "learning_rate": 1.3929892769013208e-05,
-      "loss": 0.2792,
+      "epoch": 3.364327675063971,
+      "grad_norm": 0.2284454107284546,
+      "learning_rate": 1.2783934293654726e-05,
+      "loss": 0.4008,
       "step": 93350
     },
     {
-      "epoch": 3.28,
-      "learning_rate": 1.3927338625250033e-05,
-      "loss": 0.241,
+      "epoch": 3.3645078747251955,
+      "grad_norm": 0.2578500807285309,
+      "learning_rate": 1.2781388351874518e-05,
+      "loss": 0.3791,
       "step": 93355
     },
     {
-      "epoch": 3.28,
-      "learning_rate": 1.3924784625258069e-05,
-      "loss": 0.2577,
+      "epoch": 3.3646880743864203,
+      "grad_norm": 0.23964065313339233,
+      "learning_rate": 1.2778842576570543e-05,
+      "loss": 0.3465,
       "step": 93360
     },
     {
-      "epoch": 3.28,
-      "learning_rate": 1.3922230769070465e-05,
-      "loss": 0.2489,
+      "epoch": 3.3648682740476445,
+      "grad_norm": 0.23482295870780945,
+      "learning_rate": 1.277629696777745e-05,
+      "loss": 0.3947,
       "step": 93365
     },
     {
-      "epoch": 3.29,
-      "learning_rate": 1.3919677056720405e-05,
-      "loss": 0.2431,
+      "epoch": 3.3650484737088693,
+      "grad_norm": 0.27758488059043884,
+      "learning_rate": 1.277375152552996e-05,
+      "loss": 0.3551,
       "step": 93370
     },
     {
-      "epoch": 3.29,
-      "learning_rate": 1.3917123488241025e-05,
-      "loss": 0.2847,
+      "epoch": 3.365228673370094,
+      "grad_norm": 0.26252129673957825,
+      "learning_rate": 1.2771206249862722e-05,
+      "loss": 0.3843,
       "step": 93375
     },
     {
-      "epoch": 3.29,
-      "learning_rate": 1.3914570063665505e-05,
-      "loss": 0.2718,
+      "epoch": 3.3654088730313187,
+      "grad_norm": 0.23185403645038605,
+      "learning_rate": 1.2768661140810454e-05,
+      "loss": 0.4336,
       "step": 93380
     },
     {
-      "epoch": 3.29,
-      "learning_rate": 1.3912016783026984e-05,
-      "loss": 0.2637,
+      "epoch": 3.3655890726925435,
+      "grad_norm": 0.32735198736190796,
+      "learning_rate": 1.2766116198407798e-05,
+      "loss": 0.3843,
       "step": 93385
     },
     {
-      "epoch": 3.29,
-      "learning_rate": 1.3909463646358611e-05,
-      "loss": 0.2626,
+      "epoch": 3.365769272353768,
+      "grad_norm": 0.2612536549568176,
+      "learning_rate": 1.2763571422689432e-05,
+      "loss": 0.3721,
       "step": 93390
     },
     {
-      "epoch": 3.29,
-      "learning_rate": 1.390691065369355e-05,
-      "loss": 0.2238,
+      "epoch": 3.3659494720149925,
+      "grad_norm": 0.22687096893787384,
+      "learning_rate": 1.2761026813690047e-05,
+      "loss": 0.3867,
       "step": 93395
     },
     {
-      "epoch": 3.29,
-      "learning_rate": 1.390435780506495e-05,
-      "loss": 0.2608,
+      "epoch": 3.3661296716762172,
+      "grad_norm": 0.2720807194709778,
+      "learning_rate": 1.2758482371444299e-05,
+      "loss": 0.3975,
       "step": 93400
     },
     {
-      "epoch": 3.29,
-      "learning_rate": 1.3901805100505943e-05,
-      "loss": 0.2597,
+      "epoch": 3.366309871337442,
+      "grad_norm": 0.2100588083267212,
+      "learning_rate": 1.2755938095986861e-05,
+      "loss": 0.3865,
       "step": 93405
     },
     {
-      "epoch": 3.29,
-      "learning_rate": 1.3899252540049679e-05,
-      "loss": 0.2819,
+      "epoch": 3.3664900709986667,
+      "grad_norm": 0.23490869998931885,
+      "learning_rate": 1.2753393987352396e-05,
+      "loss": 0.4074,
       "step": 93410
     },
     {
-      "epoch": 3.29,
-      "learning_rate": 1.3896700123729315e-05,
-      "loss": 0.2807,
+      "epoch": 3.366670270659891,
+      "grad_norm": 0.25652775168418884,
+      "learning_rate": 1.2750850045575558e-05,
+      "loss": 0.3773,
       "step": 93415
     },
     {
-      "epoch": 3.29,
-      "learning_rate": 1.3894147851577979e-05,
-      "loss": 0.2479,
+      "epoch": 3.3668504703211157,
+      "grad_norm": 0.25951749086380005,
+      "learning_rate": 1.2748306270691024e-05,
+      "loss": 0.3564,
       "step": 93420
     },
     {
-      "epoch": 3.29,
-      "learning_rate": 1.389159572362881e-05,
-      "loss": 0.2543,
+      "epoch": 3.3670306699823405,
+      "grad_norm": 0.2929254174232483,
+      "learning_rate": 1.2745762662733445e-05,
+      "loss": 0.3865,
       "step": 93425
     },
     {
-      "epoch": 3.29,
-      "learning_rate": 1.3889043739914943e-05,
-      "loss": 0.2714,
+      "epoch": 3.367210869643565,
+      "grad_norm": 0.24950219690799713,
+      "learning_rate": 1.274321922173748e-05,
+      "loss": 0.3821,
       "step": 93430
     },
     {
-      "epoch": 3.29,
-      "learning_rate": 1.3886491900469534e-05,
-      "loss": 0.258,
+      "epoch": 3.36739106930479,
+      "grad_norm": 0.2351795881986618,
+      "learning_rate": 1.2740675947737777e-05,
+      "loss": 0.4021,
       "step": 93435
     },
     {
-      "epoch": 3.29,
-      "learning_rate": 1.3883940205325701e-05,
-      "loss": 0.2697,
+      "epoch": 3.367571268966014,
+      "grad_norm": 0.267691969871521,
+      "learning_rate": 1.273813284076898e-05,
+      "loss": 0.3983,
       "step": 93440
     },
     {
-      "epoch": 3.29,
-      "learning_rate": 1.3881388654516573e-05,
-      "loss": 0.2575,
+      "epoch": 3.367751468627239,
+      "grad_norm": 0.18753398954868317,
+      "learning_rate": 1.2735589900865762e-05,
+      "loss": 0.3566,
       "step": 93445
     },
     {
-      "epoch": 3.29,
-      "learning_rate": 1.3878837248075294e-05,
-      "loss": 0.2603,
+      "epoch": 3.3679316682884637,
+      "grad_norm": 0.243241548538208,
+      "learning_rate": 1.2733047128062756e-05,
+      "loss": 0.3818,
       "step": 93450
     },
     {
-      "epoch": 3.29,
-      "learning_rate": 1.3876285986034981e-05,
-      "loss": 0.2569,
+      "epoch": 3.3681118679496884,
+      "grad_norm": 0.28631412982940674,
+      "learning_rate": 1.273050452239461e-05,
+      "loss": 0.3818,
       "step": 93455
     },
     {
-      "epoch": 3.29,
-      "learning_rate": 1.3873734868428768e-05,
-      "loss": 0.232,
+      "epoch": 3.3682920676109127,
+      "grad_norm": 0.22822198271751404,
+      "learning_rate": 1.272796208389596e-05,
+      "loss": 0.3824,
       "step": 93460
     },
     {
-      "epoch": 3.29,
-      "learning_rate": 1.3871183895289763e-05,
-      "loss": 0.2485,
+      "epoch": 3.3684722672721374,
+      "grad_norm": 0.262082040309906,
+      "learning_rate": 1.2725419812601453e-05,
+      "loss": 0.377,
       "step": 93465
     },
     {
-      "epoch": 3.29,
-      "learning_rate": 1.3868633066651113e-05,
-      "loss": 0.2465,
+      "epoch": 3.368652466933362,
+      "grad_norm": 0.21219515800476074,
+      "learning_rate": 1.2722877708545717e-05,
+      "loss": 0.3848,
       "step": 93470
     },
     {
-      "epoch": 3.29,
-      "learning_rate": 1.3866082382545925e-05,
-      "loss": 0.2844,
+      "epoch": 3.368832666594587,
+      "grad_norm": 0.2659531533718109,
+      "learning_rate": 1.272033577176341e-05,
+      "loss": 0.3382,
       "step": 93475
     },
     {
-      "epoch": 3.29,
-      "learning_rate": 1.386353184300731e-05,
-      "loss": 0.2281,
+      "epoch": 3.3690128662558116,
+      "grad_norm": 0.2504463791847229,
+      "learning_rate": 1.2717794002289134e-05,
+      "loss": 0.3823,
       "step": 93480
     },
     {
-      "epoch": 3.29,
-      "learning_rate": 1.3860981448068394e-05,
-      "loss": 0.2563,
+      "epoch": 3.369193065917036,
+      "grad_norm": 0.27462002635002136,
+      "learning_rate": 1.2715252400157548e-05,
+      "loss": 0.4201,
       "step": 93485
     },
     {
-      "epoch": 3.29,
-      "learning_rate": 1.3858431197762306e-05,
-      "loss": 0.2635,
+      "epoch": 3.3693732655782607,
+      "grad_norm": 0.2492084950208664,
+      "learning_rate": 1.2712710965403269e-05,
+      "loss": 0.3777,
       "step": 93490
     },
     {
-      "epoch": 3.29,
-      "learning_rate": 1.3855881092122142e-05,
-      "loss": 0.2785,
+      "epoch": 3.3695534652394854,
+      "grad_norm": 0.2124946117401123,
+      "learning_rate": 1.2710169698060922e-05,
+      "loss": 0.3578,
       "step": 93495
     },
     {
-      "epoch": 3.29,
-      "learning_rate": 1.3853331131181022e-05,
-      "loss": 0.2459,
+      "epoch": 3.36973366490071,
+      "grad_norm": 0.24669291079044342,
+      "learning_rate": 1.2707628598165137e-05,
+      "loss": 0.4265,
       "step": 93500
     },
     {
-      "epoch": 3.29,
-      "eval_loss": 0.2541661560535431,
-      "eval_runtime": 10.5324,
-      "eval_samples_per_second": 9.495,
-      "eval_steps_per_second": 9.495,
+      "epoch": 3.36973366490071,
+      "eval_loss": 0.4303239583969116,
+      "eval_runtime": 3.5298,
+      "eval_samples_per_second": 28.331,
+      "eval_steps_per_second": 7.083,
       "step": 93500
     },
     {
-      "epoch": 3.29,
-      "learning_rate": 1.3850781314972038e-05,
-      "loss": 0.251,
+      "epoch": 3.3699138645619344,
+      "grad_norm": 0.22059577703475952,
+      "learning_rate": 1.2705087665750531e-05,
+      "loss": 0.3647,
       "step": 93505
     },
     {
-      "epoch": 3.29,
-      "learning_rate": 1.3848231643528325e-05,
-      "loss": 0.2656,
+      "epoch": 3.370094064223159,
+      "grad_norm": 0.20390857756137848,
+      "learning_rate": 1.2702546900851715e-05,
+      "loss": 0.3714,
       "step": 93510
     },
     {
-      "epoch": 3.29,
-      "learning_rate": 1.3845682116882972e-05,
-      "loss": 0.2656,
+      "epoch": 3.370274263884384,
+      "grad_norm": 0.22258242964744568,
+      "learning_rate": 1.2700006303503325e-05,
+      "loss": 0.3766,
       "step": 93515
     },
     {
-      "epoch": 3.29,
-      "learning_rate": 1.3843132735069087e-05,
-      "loss": 0.2524,
+      "epoch": 3.3704544635456086,
+      "grad_norm": 0.26116088032722473,
+      "learning_rate": 1.2697465873739966e-05,
+      "loss": 0.403,
       "step": 93520
     },
     {
-      "epoch": 3.29,
-      "learning_rate": 1.3840583498119759e-05,
-      "loss": 0.2857,
+      "epoch": 3.3706346632068334,
+      "grad_norm": 0.2494215965270996,
+      "learning_rate": 1.269492561159626e-05,
+      "loss": 0.3681,
       "step": 93525
     },
     {
-      "epoch": 3.29,
-      "learning_rate": 1.383803440606811e-05,
-      "loss": 0.2584,
+      "epoch": 3.3708148628680576,
+      "grad_norm": 0.24651680886745453,
+      "learning_rate": 1.2692385517106802e-05,
+      "loss": 0.3774,
       "step": 93530
     },
     {
-      "epoch": 3.29,
-      "learning_rate": 1.3835485458947217e-05,
-      "loss": 0.2466,
+      "epoch": 3.3709950625292824,
+      "grad_norm": 0.22229456901550293,
+      "learning_rate": 1.2689845590306204e-05,
+      "loss": 0.3997,
       "step": 93535
     },
     {
-      "epoch": 3.29,
-      "learning_rate": 1.3832936656790202e-05,
-      "loss": 0.2683,
+      "epoch": 3.371175262190507,
+      "grad_norm": 0.2545780539512634,
+      "learning_rate": 1.2687305831229084e-05,
+      "loss": 0.3661,
       "step": 93540
     },
     {
-      "epoch": 3.29,
-      "learning_rate": 1.3830387999630132e-05,
-      "loss": 0.2416,
+      "epoch": 3.371355461851732,
+      "grad_norm": 0.19883494079113007,
+      "learning_rate": 1.268476623991005e-05,
+      "loss": 0.3843,
       "step": 93545
     },
     {
-      "epoch": 3.29,
-      "learning_rate": 1.3827839487500121e-05,
-      "loss": 0.2635,
+      "epoch": 3.371535661512956,
+      "grad_norm": 0.19383443892002106,
+      "learning_rate": 1.2682226816383668e-05,
+      "loss": 0.3768,
       "step": 93550
     },
     {
-      "epoch": 3.29,
-      "learning_rate": 1.3825291120433257e-05,
-      "loss": 0.2477,
+      "epoch": 3.371715861174181,
+      "grad_norm": 0.2644541263580322,
+      "learning_rate": 1.2679687560684578e-05,
+      "loss": 0.3745,
       "step": 93555
     },
     {
-      "epoch": 3.29,
-      "learning_rate": 1.382274289846261e-05,
-      "loss": 0.2627,
+      "epoch": 3.3718960608354056,
+      "grad_norm": 0.24296367168426514,
+      "learning_rate": 1.2677148472847344e-05,
+      "loss": 0.4342,
       "step": 93560
     },
     {
-      "epoch": 3.29,
-      "learning_rate": 1.382019482162129e-05,
-      "loss": 0.2665,
+      "epoch": 3.3720762604966303,
+      "grad_norm": 0.23606744408607483,
+      "learning_rate": 1.2674609552906586e-05,
+      "loss": 0.3628,
       "step": 93565
     },
     {
-      "epoch": 3.29,
-      "learning_rate": 1.3817646889942373e-05,
-      "loss": 0.2546,
+      "epoch": 3.372256460157855,
+      "grad_norm": 0.20120225846767426,
+      "learning_rate": 1.2672070800896899e-05,
+      "loss": 0.3939,
       "step": 93570
     },
     {
-      "epoch": 3.29,
-      "learning_rate": 1.3815099103458942e-05,
-      "loss": 0.2686,
+      "epoch": 3.3724366598190794,
+      "grad_norm": 0.20473100244998932,
+      "learning_rate": 1.2669532216852842e-05,
+      "loss": 0.3502,
       "step": 93575
     },
     {
-      "epoch": 3.29,
-      "learning_rate": 1.3812551462204071e-05,
-      "loss": 0.2588,
+      "epoch": 3.372616859480304,
+      "grad_norm": 0.2535666823387146,
+      "learning_rate": 1.2666993800809026e-05,
+      "loss": 0.3759,
       "step": 93580
     },
     {
-      "epoch": 3.29,
-      "learning_rate": 1.3810003966210849e-05,
-      "loss": 0.2995,
+      "epoch": 3.372797059141529,
+      "grad_norm": 0.2333008199930191,
+      "learning_rate": 1.2664455552800024e-05,
+      "loss": 0.3961,
       "step": 93585
     },
     {
-      "epoch": 3.29,
-      "learning_rate": 1.3807456615512348e-05,
-      "loss": 0.2539,
+      "epoch": 3.3729772588027536,
+      "grad_norm": 0.1824401170015335,
+      "learning_rate": 1.2661917472860449e-05,
+      "loss": 0.3831,
       "step": 93590
     },
     {
-      "epoch": 3.29,
-      "learning_rate": 1.3804909410141652e-05,
-      "loss": 0.2459,
+      "epoch": 3.373157458463978,
+      "grad_norm": 0.26639214158058167,
+      "learning_rate": 1.2659379561024848e-05,
+      "loss": 0.3991,
       "step": 93595
     },
     {
-      "epoch": 3.29,
-      "learning_rate": 1.3802362350131826e-05,
-      "loss": 0.2339,
+      "epoch": 3.3733376581252026,
+      "grad_norm": 0.2088332623243332,
+      "learning_rate": 1.2656841817327803e-05,
+      "loss": 0.3711,
       "step": 93600
     },
     {
-      "epoch": 3.29,
-      "learning_rate": 1.379981543551595e-05,
-      "loss": 0.2507,
+      "epoch": 3.3735178577864273,
+      "grad_norm": 0.24687723815441132,
+      "learning_rate": 1.2654304241803904e-05,
+      "loss": 0.4086,
       "step": 93605
     },
     {
-      "epoch": 3.29,
-      "learning_rate": 1.379726866632709e-05,
-      "loss": 0.2696,
+      "epoch": 3.373698057447652,
+      "grad_norm": 0.2668485641479492,
+      "learning_rate": 1.2651766834487724e-05,
+      "loss": 0.3932,
       "step": 93610
     },
     {
-      "epoch": 3.29,
-      "learning_rate": 1.3794722042598312e-05,
-      "loss": 0.248,
+      "epoch": 3.373878257108877,
+      "grad_norm": 0.26824715733528137,
+      "learning_rate": 1.2649229595413831e-05,
+      "loss": 0.4194,
       "step": 93615
     },
     {
-      "epoch": 3.29,
-      "learning_rate": 1.3792175564362676e-05,
-      "loss": 0.2534,
+      "epoch": 3.3740584567701015,
+      "grad_norm": 0.27278998494148254,
+      "learning_rate": 1.2646692524616788e-05,
+      "loss": 0.3771,
       "step": 93620
     },
     {
-      "epoch": 3.29,
-      "learning_rate": 1.3789629231653264e-05,
-      "loss": 0.2547,
+      "epoch": 3.374238656431326,
+      "grad_norm": 0.2048804759979248,
+      "learning_rate": 1.2644155622131163e-05,
+      "loss": 0.3519,
       "step": 93625
     },
     {
-      "epoch": 3.29,
-      "learning_rate": 1.378708304450313e-05,
-      "loss": 0.2439,
+      "epoch": 3.3744188560925505,
+      "grad_norm": 0.2662324011325836,
+      "learning_rate": 1.2641618887991532e-05,
+      "loss": 0.3949,
       "step": 93630
     },
     {
-      "epoch": 3.29,
-      "learning_rate": 1.378453700294533e-05,
-      "loss": 0.2218,
+      "epoch": 3.3745990557537753,
+      "grad_norm": 0.24462181329727173,
+      "learning_rate": 1.2639082322232455e-05,
+      "loss": 0.3855,
       "step": 93635
     },
     {
-      "epoch": 3.29,
-      "learning_rate": 1.3781991107012914e-05,
-      "loss": 0.2338,
+      "epoch": 3.374779255415,
+      "grad_norm": 0.2650309205055237,
+      "learning_rate": 1.2636545924888485e-05,
+      "loss": 0.3813,
       "step": 93640
     },
     {
-      "epoch": 3.29,
-      "learning_rate": 1.3779445356738956e-05,
-      "loss": 0.2689,
+      "epoch": 3.3749594550762243,
+      "grad_norm": 0.26425033807754517,
+      "learning_rate": 1.2634009695994182e-05,
+      "loss": 0.3943,
       "step": 93645
     },
     {
-      "epoch": 3.29,
-      "learning_rate": 1.3776899752156513e-05,
-      "loss": 0.2705,
+      "epoch": 3.375139654737449,
+      "grad_norm": 0.21526920795440674,
+      "learning_rate": 1.2631473635584107e-05,
+      "loss": 0.389,
       "step": 93650
     },
     {
-      "epoch": 3.3,
-      "learning_rate": 1.377435429329863e-05,
-      "loss": 0.2428,
+      "epoch": 3.3753198543986738,
+      "grad_norm": 0.2526695430278778,
+      "learning_rate": 1.2628937743692795e-05,
+      "loss": 0.383,
       "step": 93655
     },
     {
-      "epoch": 3.3,
-      "learning_rate": 1.3771808980198347e-05,
-      "loss": 0.2571,
+      "epoch": 3.3755000540598985,
+      "grad_norm": 0.26634764671325684,
+      "learning_rate": 1.2626402020354832e-05,
+      "loss": 0.3667,
       "step": 93660
     },
     {
-      "epoch": 3.3,
-      "learning_rate": 1.3769263812888738e-05,
-      "loss": 0.2609,
+      "epoch": 3.3756802537211232,
+      "grad_norm": 0.21848170459270477,
+      "learning_rate": 1.2623866465604727e-05,
+      "loss": 0.3813,
       "step": 93665
     },
     {
-      "epoch": 3.3,
-      "learning_rate": 1.3766718791402836e-05,
-      "loss": 0.2526,
+      "epoch": 3.3758604533823475,
+      "grad_norm": 0.25666651129722595,
+      "learning_rate": 1.2621331079477056e-05,
+      "loss": 0.3716,
       "step": 93670
     },
     {
-      "epoch": 3.3,
-      "learning_rate": 1.3764173915773676e-05,
-      "loss": 0.2505,
+      "epoch": 3.3760406530435723,
+      "grad_norm": 0.2850119471549988,
+      "learning_rate": 1.261879586200635e-05,
+      "loss": 0.3825,
       "step": 93675
     },
     {
-      "epoch": 3.3,
-      "learning_rate": 1.3761629186034325e-05,
-      "loss": 0.2853,
+      "epoch": 3.376220852704797,
+      "grad_norm": 0.22654448449611664,
+      "learning_rate": 1.2616260813227155e-05,
+      "loss": 0.3777,
       "step": 93680
     },
     {
-      "epoch": 3.3,
-      "learning_rate": 1.3759084602217814e-05,
-      "loss": 0.2664,
+      "epoch": 3.3764010523660217,
+      "grad_norm": 0.2590138912200928,
+      "learning_rate": 1.2613725933174009e-05,
+      "loss": 0.3641,
       "step": 93685
     },
     {
-      "epoch": 3.3,
-      "learning_rate": 1.375654016435718e-05,
-      "loss": 0.2423,
+      "epoch": 3.376581252027246,
+      "grad_norm": 0.20931139588356018,
+      "learning_rate": 1.2611191221881449e-05,
+      "loss": 0.3663,
       "step": 93690
     },
     {
-      "epoch": 3.3,
-      "learning_rate": 1.3753995872485448e-05,
-      "loss": 0.2628,
+      "epoch": 3.3767614516884707,
+      "grad_norm": 0.2435794174671173,
+      "learning_rate": 1.2608656679384002e-05,
+      "loss": 0.3863,
       "step": 93695
     },
     {
-      "epoch": 3.3,
-      "learning_rate": 1.3751451726635684e-05,
-      "loss": 0.2537,
+      "epoch": 3.3769416513496955,
+      "grad_norm": 0.26772207021713257,
+      "learning_rate": 1.2606122305716215e-05,
+      "loss": 0.3818,
       "step": 93700
     },
     {
-      "epoch": 3.3,
-      "learning_rate": 1.3748907726840892e-05,
-      "loss": 0.2458,
+      "epoch": 3.37712185101092,
+      "grad_norm": 0.21198633313179016,
+      "learning_rate": 1.2603588100912611e-05,
+      "loss": 0.3792,
       "step": 93705
     },
     {
-      "epoch": 3.3,
-      "learning_rate": 1.3746363873134132e-05,
-      "loss": 0.2434,
+      "epoch": 3.377302050672145,
+      "grad_norm": 0.18529768288135529,
+      "learning_rate": 1.2601054065007722e-05,
+      "loss": 0.3644,
       "step": 93710
     },
     {
-      "epoch": 3.3,
-      "learning_rate": 1.374382016554841e-05,
-      "loss": 0.2754,
+      "epoch": 3.3774822503333692,
+      "grad_norm": 0.24963034689426422,
+      "learning_rate": 1.2598520198036074e-05,
+      "loss": 0.3925,
       "step": 93715
     },
     {
-      "epoch": 3.3,
-      "learning_rate": 1.374127660411677e-05,
-      "loss": 0.273,
+      "epoch": 3.377662449994594,
+      "grad_norm": 0.27459418773651123,
+      "learning_rate": 1.259598650003217e-05,
+      "loss": 0.3957,
       "step": 93720
     },
     {
-      "epoch": 3.3,
-      "learning_rate": 1.3738733188872238e-05,
-      "loss": 0.2431,
+      "epoch": 3.3778426496558187,
+      "grad_norm": 0.23690108954906464,
+      "learning_rate": 1.2593452971030564e-05,
+      "loss": 0.3768,
       "step": 93725
     },
     {
-      "epoch": 3.3,
-      "learning_rate": 1.3736189919847835e-05,
-      "loss": 0.2741,
+      "epoch": 3.3780228493170434,
+      "grad_norm": 0.25995904207229614,
+      "learning_rate": 1.2590919611065757e-05,
+      "loss": 0.3882,
       "step": 93730
     },
     {
-      "epoch": 3.3,
-      "learning_rate": 1.3733646797076571e-05,
-      "loss": 0.2543,
+      "epoch": 3.3782030489782677,
+      "grad_norm": 0.2160549908876419,
+      "learning_rate": 1.2588386420172269e-05,
+      "loss": 0.3617,
       "step": 93735
     },
     {
-      "epoch": 3.3,
-      "learning_rate": 1.373110382059149e-05,
-      "loss": 0.2623,
+      "epoch": 3.3783832486394925,
+      "grad_norm": 0.20487286150455475,
+      "learning_rate": 1.2585853398384612e-05,
+      "loss": 0.3748,
       "step": 93740
     },
     {
-      "epoch": 3.3,
-      "learning_rate": 1.3728560990425598e-05,
-      "loss": 0.2502,
+      "epoch": 3.378563448300717,
+      "grad_norm": 0.23326444625854492,
+      "learning_rate": 1.258332054573729e-05,
+      "loss": 0.3864,
       "step": 93745
     },
     {
-      "epoch": 3.3,
-      "learning_rate": 1.3726018306611905e-05,
-      "loss": 0.2691,
+      "epoch": 3.378743647961942,
+      "grad_norm": 0.23716379702091217,
+      "learning_rate": 1.2580787862264832e-05,
+      "loss": 0.3846,
       "step": 93750
     },
     {
-      "epoch": 3.3,
-      "learning_rate": 1.3723475769183442e-05,
-      "loss": 0.2653,
+      "epoch": 3.3789238476231667,
+      "grad_norm": 0.23601892590522766,
+      "learning_rate": 1.2578255348001741e-05,
+      "loss": 0.371,
       "step": 93755
     },
     {
-      "epoch": 3.3,
-      "learning_rate": 1.3720933378173207e-05,
-      "loss": 0.275,
+      "epoch": 3.379104047284391,
+      "grad_norm": 0.2128582000732422,
+      "learning_rate": 1.2575723002982498e-05,
+      "loss": 0.3677,
       "step": 93760
     },
     {
-      "epoch": 3.3,
-      "learning_rate": 1.3718391133614227e-05,
-      "loss": 0.2911,
+      "epoch": 3.3792842469456157,
+      "grad_norm": 0.24130862951278687,
+      "learning_rate": 1.2573190827241637e-05,
+      "loss": 0.3728,
       "step": 93765
     },
     {
-      "epoch": 3.3,
-      "learning_rate": 1.3715849035539503e-05,
-      "loss": 0.252,
+      "epoch": 3.3794644466068404,
+      "grad_norm": 0.24159899353981018,
+      "learning_rate": 1.2570658820813633e-05,
+      "loss": 0.3939,
       "step": 93770
     },
     {
-      "epoch": 3.3,
-      "learning_rate": 1.3713307083982035e-05,
-      "loss": 0.2553,
+      "epoch": 3.379644646268065,
+      "grad_norm": 0.2206299901008606,
+      "learning_rate": 1.2568126983733015e-05,
+      "loss": 0.3902,
       "step": 93775
     },
     {
-      "epoch": 3.3,
-      "learning_rate": 1.3710765278974847e-05,
-      "loss": 0.2501,
+      "epoch": 3.3798248459292894,
+      "grad_norm": 0.20361188054084778,
+      "learning_rate": 1.256559531603425e-05,
+      "loss": 0.3666,
       "step": 93780
     },
     {
-      "epoch": 3.3,
-      "learning_rate": 1.3708223620550928e-05,
-      "loss": 0.2694,
+      "epoch": 3.380005045590514,
+      "grad_norm": 0.22459423542022705,
+      "learning_rate": 1.2563063817751834e-05,
+      "loss": 0.3898,
       "step": 93785
     },
     {
-      "epoch": 3.3,
-      "learning_rate": 1.3705682108743276e-05,
-      "loss": 0.26,
+      "epoch": 3.380185245251739,
+      "grad_norm": 0.3262191414833069,
+      "learning_rate": 1.2560532488920274e-05,
+      "loss": 0.3945,
       "step": 93790
     },
     {
-      "epoch": 3.3,
-      "learning_rate": 1.3703140743584909e-05,
-      "loss": 0.2596,
+      "epoch": 3.3803654449129636,
+      "grad_norm": 0.23222355544567108,
+      "learning_rate": 1.2558001329574049e-05,
+      "loss": 0.3799,
       "step": 93795
     },
     {
-      "epoch": 3.3,
-      "learning_rate": 1.3700599525108812e-05,
-      "loss": 0.2495,
+      "epoch": 3.3805456445741884,
+      "grad_norm": 0.24489018321037292,
+      "learning_rate": 1.255547033974765e-05,
+      "loss": 0.4148,
       "step": 93800
     },
     {
-      "epoch": 3.3,
-      "learning_rate": 1.3698058453347975e-05,
-      "loss": 0.2526,
+      "epoch": 3.3807258442354127,
+      "grad_norm": 0.246272012591362,
+      "learning_rate": 1.2552939519475554e-05,
+      "loss": 0.3642,
       "step": 93805
     },
     {
-      "epoch": 3.3,
-      "learning_rate": 1.3695517528335411e-05,
-      "loss": 0.2751,
+      "epoch": 3.3809060438966374,
+      "grad_norm": 0.2430204451084137,
+      "learning_rate": 1.255040886879224e-05,
+      "loss": 0.3893,
       "step": 93810
     },
     {
-      "epoch": 3.3,
-      "learning_rate": 1.3692976750104092e-05,
-      "loss": 0.2642,
+      "epoch": 3.381086243557862,
+      "grad_norm": 0.28567835688591003,
+      "learning_rate": 1.2547878387732203e-05,
+      "loss": 0.4011,
       "step": 93815
     },
     {
-      "epoch": 3.3,
-      "learning_rate": 1.3690436118687031e-05,
-      "loss": 0.2701,
+      "epoch": 3.381266443219087,
+      "grad_norm": 0.28451603651046753,
+      "learning_rate": 1.2545348076329916e-05,
+      "loss": 0.3964,
       "step": 93820
     },
     {
-      "epoch": 3.3,
-      "learning_rate": 1.3687895634117199e-05,
-      "loss": 0.2818,
+      "epoch": 3.381446642880311,
+      "grad_norm": 0.20606903731822968,
+      "learning_rate": 1.254281793461985e-05,
+      "loss": 0.3684,
       "step": 93825
     },
     {
-      "epoch": 3.3,
-      "learning_rate": 1.3685355296427582e-05,
-      "loss": 0.2594,
+      "epoch": 3.381626842541536,
+      "grad_norm": 0.2716697156429291,
+      "learning_rate": 1.2540287962636473e-05,
+      "loss": 0.3818,
       "step": 93830
     },
     {
-      "epoch": 3.3,
-      "learning_rate": 1.3682815105651175e-05,
-      "loss": 0.2336,
+      "epoch": 3.3818070422027606,
+      "grad_norm": 0.2406664490699768,
+      "learning_rate": 1.2537758160414265e-05,
+      "loss": 0.3719,
       "step": 93835
     },
     {
-      "epoch": 3.3,
-      "learning_rate": 1.3680275061820957e-05,
-      "loss": 0.2715,
+      "epoch": 3.3819872418639854,
+      "grad_norm": 0.23287048935890198,
+      "learning_rate": 1.2535228527987674e-05,
+      "loss": 0.373,
       "step": 93840
     },
     {
-      "epoch": 3.3,
-      "learning_rate": 1.3677735164969909e-05,
-      "loss": 0.264,
+      "epoch": 3.38216744152521,
+      "grad_norm": 0.20795047283172607,
+      "learning_rate": 1.2532699065391206e-05,
+      "loss": 0.3667,
       "step": 93845
     },
     {
-      "epoch": 3.3,
-      "learning_rate": 1.3675195415130999e-05,
-      "loss": 0.2618,
+      "epoch": 3.3823476411864344,
+      "grad_norm": 0.23873348534107208,
+      "learning_rate": 1.2530169772659278e-05,
+      "loss": 0.3937,
       "step": 93850
     },
     {
-      "epoch": 3.3,
-      "learning_rate": 1.3672655812337221e-05,
-      "loss": 0.2503,
+      "epoch": 3.382527840847659,
+      "grad_norm": 0.3395892083644867,
+      "learning_rate": 1.2527640649826384e-05,
+      "loss": 0.408,
       "step": 93855
     },
     {
-      "epoch": 3.3,
-      "learning_rate": 1.3670116356621544e-05,
-      "loss": 0.2506,
+      "epoch": 3.382708040508884,
+      "grad_norm": 0.2514219582080841,
+      "learning_rate": 1.2525111696926967e-05,
+      "loss": 0.401,
       "step": 93860
     },
     {
-      "epoch": 3.3,
-      "learning_rate": 1.3667577048016927e-05,
-      "loss": 0.2564,
+      "epoch": 3.3828882401701086,
+      "grad_norm": 0.23013897240161896,
+      "learning_rate": 1.2522582913995484e-05,
+      "loss": 0.3651,
       "step": 93865
     },
     {
-      "epoch": 3.3,
-      "learning_rate": 1.3665037886556353e-05,
-      "loss": 0.2567,
+      "epoch": 3.383068439831333,
+      "grad_norm": 0.2501782476902008,
+      "learning_rate": 1.2520054301066409e-05,
+      "loss": 0.3569,
       "step": 93870
     },
     {
-      "epoch": 3.3,
-      "learning_rate": 1.3662498872272806e-05,
-      "loss": 0.244,
+      "epoch": 3.3832486394925576,
+      "grad_norm": 0.2740953266620636,
+      "learning_rate": 1.2517525858174166e-05,
+      "loss": 0.3837,
       "step": 93875
     },
     {
-      "epoch": 3.3,
-      "learning_rate": 1.3659960005199238e-05,
-      "loss": 0.2331,
+      "epoch": 3.3834288391537823,
+      "grad_norm": 0.2235153615474701,
+      "learning_rate": 1.251499758535321e-05,
+      "loss": 0.344,
       "step": 93880
     },
     {
-      "epoch": 3.3,
-      "learning_rate": 1.3657421285368615e-05,
-      "loss": 0.2539,
+      "epoch": 3.383609038815007,
+      "grad_norm": 0.21870172023773193,
+      "learning_rate": 1.2512469482638006e-05,
+      "loss": 0.3887,
       "step": 93885
     },
     {
-      "epoch": 3.3,
-      "learning_rate": 1.3654882712813893e-05,
-      "loss": 0.2625,
+      "epoch": 3.383789238476232,
+      "grad_norm": 0.21588364243507385,
+      "learning_rate": 1.2509941550062987e-05,
+      "loss": 0.3651,
       "step": 93890
     },
     {
-      "epoch": 3.3,
-      "learning_rate": 1.365234428756805e-05,
-      "loss": 0.2449,
+      "epoch": 3.3839694381374565,
+      "grad_norm": 0.24432948231697083,
+      "learning_rate": 1.2507413787662592e-05,
+      "loss": 0.3971,
       "step": 93895
     },
     {
-      "epoch": 3.3,
-      "learning_rate": 1.364980600966404e-05,
-      "loss": 0.2693,
+      "epoch": 3.384149637798681,
+      "grad_norm": 0.23872052133083344,
+      "learning_rate": 1.2504886195471272e-05,
+      "loss": 0.3919,
       "step": 93900
     },
     {
-      "epoch": 3.3,
-      "learning_rate": 1.3647267879134807e-05,
-      "loss": 0.2634,
+      "epoch": 3.3843298374599056,
+      "grad_norm": 0.25494420528411865,
+      "learning_rate": 1.2502358773523443e-05,
+      "loss": 0.3789,
       "step": 93905
     },
     {
-      "epoch": 3.3,
-      "learning_rate": 1.3644729896013331e-05,
-      "loss": 0.2605,
+      "epoch": 3.3845100371211303,
+      "grad_norm": 0.2569286525249481,
+      "learning_rate": 1.2499831521853567e-05,
+      "loss": 0.3544,
       "step": 93910
     },
     {
-      "epoch": 3.3,
-      "learning_rate": 1.3642192060332554e-05,
-      "loss": 0.2513,
+      "epoch": 3.384690236782355,
+      "grad_norm": 0.2678029537200928,
+      "learning_rate": 1.2497304440496068e-05,
+      "loss": 0.4016,
       "step": 93915
     },
     {
-      "epoch": 3.3,
-      "learning_rate": 1.3639654372125416e-05,
-      "loss": 0.2845,
+      "epoch": 3.3848704364435793,
+      "grad_norm": 0.21244113147258759,
+      "learning_rate": 1.2494777529485374e-05,
+      "loss": 0.3636,
       "step": 93920
     },
     {
-      "epoch": 3.3,
-      "learning_rate": 1.3637116831424879e-05,
-      "loss": 0.2483,
+      "epoch": 3.385050636104804,
+      "grad_norm": 0.21646173298358917,
+      "learning_rate": 1.2492250788855916e-05,
+      "loss": 0.3687,
       "step": 93925
     },
     {
-      "epoch": 3.3,
-      "learning_rate": 1.3634579438263902e-05,
-      "loss": 0.2557,
+      "epoch": 3.385230835766029,
+      "grad_norm": 0.214036226272583,
+      "learning_rate": 1.248972421864211e-05,
+      "loss": 0.3638,
       "step": 93930
     },
     {
-      "epoch": 3.3,
-      "learning_rate": 1.3632042192675424e-05,
-      "loss": 0.2434,
+      "epoch": 3.3854110354272535,
+      "grad_norm": 0.19108587503433228,
+      "learning_rate": 1.2487197818878399e-05,
+      "loss": 0.3568,
       "step": 93935
     },
     {
-      "epoch": 3.31,
-      "learning_rate": 1.3629505094692382e-05,
-      "loss": 0.2606,
+      "epoch": 3.3855912350884783,
+      "grad_norm": 0.2635771632194519,
+      "learning_rate": 1.2484671589599204e-05,
+      "loss": 0.3701,
       "step": 93940
     },
     {
-      "epoch": 3.31,
-      "learning_rate": 1.3626968144347715e-05,
-      "loss": 0.2532,
+      "epoch": 3.3857714347497025,
+      "grad_norm": 0.27943238615989685,
+      "learning_rate": 1.2482145530838918e-05,
+      "loss": 0.3563,
       "step": 93945
     },
     {
-      "epoch": 3.31,
-      "learning_rate": 1.3624431341674383e-05,
-      "loss": 0.2282,
+      "epoch": 3.3859516344109273,
+      "grad_norm": 0.2727324366569519,
+      "learning_rate": 1.2479619642631985e-05,
+      "loss": 0.3783,
       "step": 93950
     },
     {
-      "epoch": 3.31,
-      "learning_rate": 1.362189468670531e-05,
-      "loss": 0.2614,
+      "epoch": 3.386131834072152,
+      "grad_norm": 0.28211259841918945,
+      "learning_rate": 1.2477093925012808e-05,
+      "loss": 0.4027,
       "step": 93955
     },
     {
-      "epoch": 3.31,
-      "learning_rate": 1.3619358179473438e-05,
-      "loss": 0.2579,
+      "epoch": 3.3863120337333767,
+      "grad_norm": 0.24526701867580414,
+      "learning_rate": 1.2474568378015802e-05,
+      "loss": 0.3788,
       "step": 93960
     },
     {
-      "epoch": 3.31,
-      "learning_rate": 1.3616821820011689e-05,
-      "loss": 0.2629,
+      "epoch": 3.386492233394601,
+      "grad_norm": 0.2575979232788086,
+      "learning_rate": 1.247204300167538e-05,
+      "loss": 0.358,
       "step": 93965
     },
     {
-      "epoch": 3.31,
-      "learning_rate": 1.3614285608353017e-05,
-      "loss": 0.2686,
+      "epoch": 3.3866724330558258,
+      "grad_norm": 0.261115700006485,
+      "learning_rate": 1.2469517796025934e-05,
+      "loss": 0.3946,
       "step": 93970
     },
     {
-      "epoch": 3.31,
-      "learning_rate": 1.3611749544530334e-05,
-      "loss": 0.2524,
+      "epoch": 3.3868526327170505,
+      "grad_norm": 0.2626388370990753,
+      "learning_rate": 1.2466992761101893e-05,
+      "loss": 0.3957,
       "step": 93975
     },
     {
-      "epoch": 3.31,
-      "learning_rate": 1.360921362857659e-05,
-      "loss": 0.2663,
+      "epoch": 3.3870328323782752,
+      "grad_norm": 0.2778167724609375,
+      "learning_rate": 1.2464467896937649e-05,
+      "loss": 0.3642,
       "step": 93980
     },
     {
-      "epoch": 3.31,
-      "learning_rate": 1.3606677860524686e-05,
-      "loss": 0.2573,
+      "epoch": 3.3872130320395,
+      "grad_norm": 0.2699746787548065,
+      "learning_rate": 1.2461943203567602e-05,
+      "loss": 0.3913,
       "step": 93985
     },
     {
-      "epoch": 3.31,
-      "learning_rate": 1.3604142240407575e-05,
-      "loss": 0.2445,
+      "epoch": 3.3873932317007243,
+      "grad_norm": 0.2563049793243408,
+      "learning_rate": 1.2459418681026152e-05,
+      "loss": 0.3911,
       "step": 93990
     },
     {
-      "epoch": 3.31,
-      "learning_rate": 1.3601606768258163e-05,
-      "loss": 0.2402,
+      "epoch": 3.387573431361949,
+      "grad_norm": 0.27618205547332764,
+      "learning_rate": 1.2456894329347685e-05,
+      "loss": 0.4132,
       "step": 93995
     },
     {
-      "epoch": 3.31,
-      "learning_rate": 1.3599071444109378e-05,
-      "loss": 0.2714,
+      "epoch": 3.3877536310231737,
+      "grad_norm": 0.19964580237865448,
+      "learning_rate": 1.2454370148566613e-05,
+      "loss": 0.3732,
       "step": 94000
     },
     {
-      "epoch": 3.31,
-      "eval_loss": 0.2543759047985077,
-      "eval_runtime": 10.548,
-      "eval_samples_per_second": 9.48,
-      "eval_steps_per_second": 9.48,
+      "epoch": 3.3877536310231737,
+      "eval_loss": 0.43040722608566284,
+      "eval_runtime": 3.5294,
+      "eval_samples_per_second": 28.334,
+      "eval_steps_per_second": 7.083,
       "step": 94000
     },
     {
-      "epoch": 3.31,
-      "learning_rate": 1.3596536267994126e-05,
-      "loss": 0.2596,
+      "epoch": 3.3879338306843985,
+      "grad_norm": 0.2523888647556305,
+      "learning_rate": 1.2451846138717321e-05,
+      "loss": 0.3964,
       "step": 94005
     },
     {
-      "epoch": 3.31,
-      "learning_rate": 1.3594001239945345e-05,
-      "loss": 0.2738,
+      "epoch": 3.3881140303456228,
+      "grad_norm": 0.2638121247291565,
+      "learning_rate": 1.2449322299834196e-05,
+      "loss": 0.4108,
       "step": 94010
     },
     {
-      "epoch": 3.31,
-      "learning_rate": 1.3591466359995942e-05,
-      "loss": 0.2628,
+      "epoch": 3.3882942300068475,
+      "grad_norm": 0.2701154053211212,
+      "learning_rate": 1.2446798631951623e-05,
+      "loss": 0.3477,
       "step": 94015
     },
     {
-      "epoch": 3.31,
-      "learning_rate": 1.358893162817882e-05,
-      "loss": 0.2545,
+      "epoch": 3.388474429668072,
+      "grad_norm": 0.2638227641582489,
+      "learning_rate": 1.2444275135103988e-05,
+      "loss": 0.4001,
       "step": 94020
     },
     {
-      "epoch": 3.31,
-      "learning_rate": 1.358639704452691e-05,
-      "loss": 0.2568,
+      "epoch": 3.388654629329297,
+      "grad_norm": 0.20842356979846954,
+      "learning_rate": 1.2441751809325666e-05,
+      "loss": 0.3629,
       "step": 94025
     },
     {
-      "epoch": 3.31,
-      "learning_rate": 1.35838626090731e-05,
-      "loss": 0.2592,
+      "epoch": 3.3888348289905217,
+      "grad_norm": 0.24552541971206665,
+      "learning_rate": 1.2439228654651053e-05,
+      "loss": 0.381,
       "step": 94030
     },
     {
-      "epoch": 3.31,
-      "learning_rate": 1.3581328321850322e-05,
-      "loss": 0.2703,
+      "epoch": 3.389015028651746,
+      "grad_norm": 0.22910676896572113,
+      "learning_rate": 1.2436705671114515e-05,
+      "loss": 0.3913,
       "step": 94035
     },
     {
-      "epoch": 3.31,
-      "learning_rate": 1.3578794182891464e-05,
-      "loss": 0.2574,
+      "epoch": 3.3891952283129707,
+      "grad_norm": 0.1990845948457718,
+      "learning_rate": 1.2434182858750431e-05,
+      "loss": 0.4065,
       "step": 94040
     },
     {
-      "epoch": 3.31,
-      "learning_rate": 1.3576260192229445e-05,
-      "loss": 0.2624,
+      "epoch": 3.3893754279741954,
+      "grad_norm": 0.24163967370986938,
+      "learning_rate": 1.2431660217593175e-05,
+      "loss": 0.4221,
       "step": 94045
     },
     {
-      "epoch": 3.31,
-      "learning_rate": 1.3573726349897159e-05,
-      "loss": 0.2587,
+      "epoch": 3.38955562763542,
+      "grad_norm": 0.20850235223770142,
+      "learning_rate": 1.2429137747677105e-05,
+      "loss": 0.3812,
       "step": 94050
     },
     {
-      "epoch": 3.31,
-      "learning_rate": 1.3571192655927503e-05,
-      "loss": 0.2631,
+      "epoch": 3.3897358272966445,
+      "grad_norm": 0.22851799428462982,
+      "learning_rate": 1.2426615449036619e-05,
+      "loss": 0.4011,
       "step": 94055
     },
     {
-      "epoch": 3.31,
-      "learning_rate": 1.356865911035337e-05,
-      "loss": 0.2539,
+      "epoch": 3.389916026957869,
+      "grad_norm": 0.24421176314353943,
+      "learning_rate": 1.2424093321706052e-05,
+      "loss": 0.432,
       "step": 94060
     },
     {
-      "epoch": 3.31,
-      "learning_rate": 1.3566125713207678e-05,
-      "loss": 0.25,
+      "epoch": 3.390096226619094,
+      "grad_norm": 0.20470160245895386,
+      "learning_rate": 1.242157136571977e-05,
+      "loss": 0.4326,
       "step": 94065
     },
     {
-      "epoch": 3.31,
-      "learning_rate": 1.3563592464523305e-05,
-      "loss": 0.2691,
+      "epoch": 3.3902764262803187,
+      "grad_norm": 0.2451028823852539,
+      "learning_rate": 1.2419049581112152e-05,
+      "loss": 0.329,
       "step": 94070
     },
     {
-      "epoch": 3.31,
-      "learning_rate": 1.3561059364333147e-05,
-      "loss": 0.2674,
+      "epoch": 3.3904566259415434,
+      "grad_norm": 0.20993046462535858,
+      "learning_rate": 1.2416527967917548e-05,
+      "loss": 0.3511,
       "step": 94075
     },
     {
-      "epoch": 3.31,
-      "learning_rate": 1.3558526412670086e-05,
-      "loss": 0.2457,
+      "epoch": 3.3906368256027677,
+      "grad_norm": 0.2813032567501068,
+      "learning_rate": 1.2414006526170312e-05,
+      "loss": 0.3764,
       "step": 94080
     },
     {
-      "epoch": 3.31,
-      "learning_rate": 1.355599360956702e-05,
-      "loss": 0.262,
+      "epoch": 3.3908170252639924,
+      "grad_norm": 0.22048485279083252,
+      "learning_rate": 1.2411485255904807e-05,
+      "loss": 0.3629,
       "step": 94085
     },
     {
-      "epoch": 3.31,
-      "learning_rate": 1.3553460955056845e-05,
-      "loss": 0.2832,
+      "epoch": 3.390997224925217,
+      "grad_norm": 0.2146688997745514,
+      "learning_rate": 1.2408964157155362e-05,
+      "loss": 0.3739,
       "step": 94090
     },
     {
-      "epoch": 3.31,
-      "learning_rate": 1.3550928449172433e-05,
-      "loss": 0.2359,
+      "epoch": 3.391177424586442,
+      "grad_norm": 0.26328352093696594,
+      "learning_rate": 1.2406443229956355e-05,
+      "loss": 0.4042,
       "step": 94095
     },
     {
-      "epoch": 3.31,
-      "learning_rate": 1.3548396091946663e-05,
-      "loss": 0.2592,
+      "epoch": 3.391357624247666,
+      "grad_norm": 0.22812364995479584,
+      "learning_rate": 1.240392247434212e-05,
+      "loss": 0.3294,
       "step": 94100
     },
     {
-      "epoch": 3.31,
-      "learning_rate": 1.3545863883412432e-05,
-      "loss": 0.2594,
+      "epoch": 3.391537823908891,
+      "grad_norm": 0.19867856800556183,
+      "learning_rate": 1.2401401890347003e-05,
+      "loss": 0.4121,
       "step": 94105
     },
     {
-      "epoch": 3.31,
-      "learning_rate": 1.3543331823602613e-05,
-      "loss": 0.2325,
+      "epoch": 3.3917180235701156,
+      "grad_norm": 0.24459275603294373,
+      "learning_rate": 1.2398881478005345e-05,
+      "loss": 0.3386,
       "step": 94110
     },
     {
-      "epoch": 3.31,
-      "learning_rate": 1.3540799912550073e-05,
-      "loss": 0.2471,
+      "epoch": 3.3918982232313404,
+      "grad_norm": 0.2276408076286316,
+      "learning_rate": 1.2396361237351478e-05,
+      "loss": 0.3889,
       "step": 94115
     },
     {
-      "epoch": 3.31,
-      "learning_rate": 1.3538268150287688e-05,
-      "loss": 0.2797,
+      "epoch": 3.392078422892565,
+      "grad_norm": 0.1976245641708374,
+      "learning_rate": 1.2393841168419759e-05,
+      "loss": 0.4016,
       "step": 94120
     },
     {
-      "epoch": 3.31,
-      "learning_rate": 1.3535736536848349e-05,
-      "loss": 0.2508,
+      "epoch": 3.39225862255379,
+      "grad_norm": 0.2941203713417053,
+      "learning_rate": 1.2391321271244524e-05,
+      "loss": 0.3932,
       "step": 94125
     },
     {
-      "epoch": 3.31,
-      "learning_rate": 1.3533205072264916e-05,
-      "loss": 0.2521,
+      "epoch": 3.392438822215014,
+      "grad_norm": 0.3133787214756012,
+      "learning_rate": 1.2388801545860076e-05,
+      "loss": 0.3873,
       "step": 94130
     },
     {
-      "epoch": 3.31,
-      "learning_rate": 1.3530673756570245e-05,
-      "loss": 0.2397,
+      "epoch": 3.392619021876239,
+      "grad_norm": 0.2651798725128174,
+      "learning_rate": 1.2386281992300775e-05,
+      "loss": 0.4417,
       "step": 94135
     },
     {
-      "epoch": 3.31,
-      "learning_rate": 1.3528142589797227e-05,
-      "loss": 0.2531,
+      "epoch": 3.3927992215374636,
+      "grad_norm": 0.23838862776756287,
+      "learning_rate": 1.2383762610600938e-05,
+      "loss": 0.3781,
       "step": 94140
     },
     {
-      "epoch": 3.31,
-      "learning_rate": 1.3525611571978708e-05,
-      "loss": 0.2361,
+      "epoch": 3.3929794211986883,
+      "grad_norm": 0.2366458624601364,
+      "learning_rate": 1.2381243400794885e-05,
+      "loss": 0.3972,
       "step": 94145
     },
     {
-      "epoch": 3.31,
-      "learning_rate": 1.3523080703147572e-05,
-      "loss": 0.2691,
+      "epoch": 3.3931596208599126,
+      "grad_norm": 0.2220826894044876,
+      "learning_rate": 1.2378724362916966e-05,
+      "loss": 0.4127,
       "step": 94150
     },
     {
-      "epoch": 3.31,
-      "learning_rate": 1.3520549983336658e-05,
-      "loss": 0.2774,
+      "epoch": 3.3933398205211374,
+      "grad_norm": 0.20894181728363037,
+      "learning_rate": 1.2376205497001465e-05,
+      "loss": 0.3961,
       "step": 94155
     },
     {
-      "epoch": 3.31,
-      "learning_rate": 1.3518019412578848e-05,
-      "loss": 0.2514,
+      "epoch": 3.393520020182362,
+      "grad_norm": 0.24846115708351135,
+      "learning_rate": 1.2373686803082728e-05,
+      "loss": 0.3631,
       "step": 94160
     },
     {
-      "epoch": 3.31,
-      "learning_rate": 1.3515488990906989e-05,
-      "loss": 0.266,
+      "epoch": 3.393700219843587,
+      "grad_norm": 0.20306502282619476,
+      "learning_rate": 1.2371168281195067e-05,
+      "loss": 0.3865,
       "step": 94165
     },
     {
-      "epoch": 3.31,
-      "learning_rate": 1.3512958718353937e-05,
-      "loss": 0.2292,
+      "epoch": 3.3938804195048116,
+      "grad_norm": 0.222940593957901,
+      "learning_rate": 1.2368649931372791e-05,
+      "loss": 0.3733,
       "step": 94170
     },
     {
-      "epoch": 3.31,
-      "learning_rate": 1.3510428594952537e-05,
-      "loss": 0.2441,
+      "epoch": 3.394060619166036,
+      "grad_norm": 0.2514864206314087,
+      "learning_rate": 1.2366131753650214e-05,
+      "loss": 0.4155,
       "step": 94175
     },
     {
-      "epoch": 3.31,
-      "learning_rate": 1.3507898620735659e-05,
-      "loss": 0.2651,
+      "epoch": 3.3942408188272606,
+      "grad_norm": 0.2028576135635376,
+      "learning_rate": 1.2363613748061639e-05,
+      "loss": 0.3471,
       "step": 94180
     },
     {
-      "epoch": 3.31,
-      "learning_rate": 1.3505368795736143e-05,
-      "loss": 0.2452,
+      "epoch": 3.3944210184884853,
+      "grad_norm": 0.21311317384243011,
+      "learning_rate": 1.2361095914641388e-05,
+      "loss": 0.3696,
       "step": 94185
     },
     {
-      "epoch": 3.31,
-      "learning_rate": 1.3502839119986827e-05,
-      "loss": 0.2532,
+      "epoch": 3.39460121814971,
+      "grad_norm": 0.21538734436035156,
+      "learning_rate": 1.2358578253423757e-05,
+      "loss": 0.3949,
       "step": 94190
     },
     {
-      "epoch": 3.31,
-      "learning_rate": 1.350030959352058e-05,
-      "loss": 0.2476,
+      "epoch": 3.3947814178109343,
+      "grad_norm": 0.25031110644340515,
+      "learning_rate": 1.235606076444305e-05,
+      "loss": 0.3706,
       "step": 94195
     },
     {
-      "epoch": 3.31,
-      "learning_rate": 1.3497780216370226e-05,
-      "loss": 0.2817,
+      "epoch": 3.394961617472159,
+      "grad_norm": 0.1601521223783493,
+      "learning_rate": 1.235354344773357e-05,
+      "loss": 0.3664,
       "step": 94200
     },
     {
-      "epoch": 3.31,
-      "learning_rate": 1.349525098856862e-05,
-      "loss": 0.2676,
+      "epoch": 3.395141817133384,
+      "grad_norm": 0.24917049705982208,
+      "learning_rate": 1.235102630332961e-05,
+      "loss": 0.37,
       "step": 94205
     },
     {
-      "epoch": 3.31,
-      "learning_rate": 1.3492721910148604e-05,
-      "loss": 0.246,
+      "epoch": 3.3953220167946085,
+      "grad_norm": 0.24421679973602295,
+      "learning_rate": 1.234850933126546e-05,
+      "loss": 0.3709,
       "step": 94210
     },
     {
-      "epoch": 3.31,
-      "learning_rate": 1.3490192981142996e-05,
-      "loss": 0.2373,
+      "epoch": 3.3955022164558333,
+      "grad_norm": 0.20778076350688934,
+      "learning_rate": 1.2345992531575426e-05,
+      "loss": 0.4043,
       "step": 94215
     },
     {
-      "epoch": 3.31,
-      "learning_rate": 1.3487664201584659e-05,
-      "loss": 0.2551,
+      "epoch": 3.3956824161170576,
+      "grad_norm": 0.22638878226280212,
+      "learning_rate": 1.2343475904293797e-05,
+      "loss": 0.3479,
       "step": 94220
     },
     {
-      "epoch": 3.32,
-      "learning_rate": 1.3485135571506416e-05,
-      "loss": 0.2407,
+      "epoch": 3.3958626157782823,
+      "grad_norm": 0.2236730009317398,
+      "learning_rate": 1.2340959449454859e-05,
+      "loss": 0.3658,
       "step": 94225
     },
     {
-      "epoch": 3.32,
-      "learning_rate": 1.3482607090941096e-05,
-      "loss": 0.2685,
+      "epoch": 3.396042815439507,
+      "grad_norm": 0.22104784846305847,
+      "learning_rate": 1.2338443167092898e-05,
+      "loss": 0.3816,
       "step": 94230
     },
     {
-      "epoch": 3.32,
-      "learning_rate": 1.3480078759921527e-05,
-      "loss": 0.2463,
+      "epoch": 3.3962230151007318,
+      "grad_norm": 0.26684364676475525,
+      "learning_rate": 1.2335927057242185e-05,
+      "loss": 0.3682,
       "step": 94235
     },
     {
-      "epoch": 3.32,
-      "learning_rate": 1.3477550578480553e-05,
-      "loss": 0.2558,
+      "epoch": 3.396403214761956,
+      "grad_norm": 0.22945067286491394,
+      "learning_rate": 1.2333411119937036e-05,
+      "loss": 0.3779,
       "step": 94240
     },
     {
-      "epoch": 3.32,
-      "learning_rate": 1.3475022546650979e-05,
-      "loss": 0.2484,
+      "epoch": 3.396583414423181,
+      "grad_norm": 0.18376512825489044,
+      "learning_rate": 1.2330895355211697e-05,
+      "loss": 0.3449,
       "step": 94245
     },
     {
-      "epoch": 3.32,
-      "learning_rate": 1.3472494664465652e-05,
-      "loss": 0.2492,
+      "epoch": 3.3967636140844055,
+      "grad_norm": 0.24991095066070557,
+      "learning_rate": 1.2328379763100445e-05,
+      "loss": 0.4067,
       "step": 94250
     },
     {
-      "epoch": 3.32,
-      "learning_rate": 1.3469966931957374e-05,
-      "loss": 0.2552,
+      "epoch": 3.3969438137456303,
+      "grad_norm": 0.2116982489824295,
+      "learning_rate": 1.2325864343637577e-05,
+      "loss": 0.3399,
       "step": 94255
     },
     {
-      "epoch": 3.32,
-      "learning_rate": 1.346743934915899e-05,
-      "loss": 0.2591,
+      "epoch": 3.397124013406855,
+      "grad_norm": 0.23185938596725464,
+      "learning_rate": 1.2323349096857354e-05,
+      "loss": 0.378,
       "step": 94260
     },
     {
-      "epoch": 3.32,
-      "learning_rate": 1.34649119161033e-05,
-      "loss": 0.2093,
+      "epoch": 3.3973042130680793,
+      "grad_norm": 0.21932083368301392,
+      "learning_rate": 1.2320834022794045e-05,
+      "loss": 0.3541,
       "step": 94265
     },
     {
-      "epoch": 3.32,
-      "learning_rate": 1.3462384632823122e-05,
-      "loss": 0.2402,
+      "epoch": 3.397484412729304,
+      "grad_norm": 0.20772342383861542,
+      "learning_rate": 1.2318319121481917e-05,
+      "loss": 0.3874,
       "step": 94270
     },
     {
-      "epoch": 3.32,
-      "learning_rate": 1.3459857499351286e-05,
-      "loss": 0.2607,
+      "epoch": 3.3976646123905287,
+      "grad_norm": 0.2540047764778137,
+      "learning_rate": 1.2315804392955228e-05,
+      "loss": 0.4273,
       "step": 94275
     },
     {
-      "epoch": 3.32,
-      "learning_rate": 1.3457330515720592e-05,
-      "loss": 0.2511,
+      "epoch": 3.3978448120517535,
+      "grad_norm": 0.25412896275520325,
+      "learning_rate": 1.2313289837248254e-05,
+      "loss": 0.3906,
       "step": 94280
     },
     {
-      "epoch": 3.32,
-      "learning_rate": 1.3454803681963856e-05,
-      "loss": 0.253,
+      "epoch": 3.3980250117129778,
+      "grad_norm": 0.2159910500049591,
+      "learning_rate": 1.2310775454395252e-05,
+      "loss": 0.3793,
       "step": 94285
     },
     {
-      "epoch": 3.32,
-      "learning_rate": 1.3452276998113878e-05,
-      "loss": 0.2711,
+      "epoch": 3.3982052113742025,
+      "grad_norm": 0.23486000299453735,
+      "learning_rate": 1.2308261244430477e-05,
+      "loss": 0.3846,
       "step": 94290
     },
     {
-      "epoch": 3.32,
-      "learning_rate": 1.344975046420348e-05,
-      "loss": 0.2385,
+      "epoch": 3.3983854110354272,
+      "grad_norm": 0.2207733690738678,
+      "learning_rate": 1.2305747207388187e-05,
+      "loss": 0.3535,
       "step": 94295
     },
     {
-      "epoch": 3.32,
-      "learning_rate": 1.3447224080265462e-05,
-      "loss": 0.2699,
+      "epoch": 3.398565610696652,
+      "grad_norm": 0.22273029386997223,
+      "learning_rate": 1.2303233343302623e-05,
+      "loss": 0.3769,
       "step": 94300
     },
     {
-      "epoch": 3.32,
-      "learning_rate": 1.3444697846332612e-05,
-      "loss": 0.2332,
+      "epoch": 3.3987458103578767,
+      "grad_norm": 0.2068636119365692,
+      "learning_rate": 1.2300719652208057e-05,
+      "loss": 0.355,
       "step": 94305
     },
     {
-      "epoch": 3.32,
-      "learning_rate": 1.344217176243775e-05,
-      "loss": 0.2502,
+      "epoch": 3.398926010019101,
+      "grad_norm": 0.2813783884048462,
+      "learning_rate": 1.2298206134138726e-05,
+      "loss": 0.3902,
       "step": 94310
     },
     {
-      "epoch": 3.32,
-      "learning_rate": 1.3439645828613678e-05,
-      "loss": 0.2498,
+      "epoch": 3.3991062096803257,
+      "grad_norm": 0.2261374592781067,
+      "learning_rate": 1.2295692789128877e-05,
+      "loss": 0.3624,
       "step": 94315
     },
     {
-      "epoch": 3.32,
-      "learning_rate": 1.3437120044893187e-05,
-      "loss": 0.2674,
+      "epoch": 3.3992864093415505,
+      "grad_norm": 0.21216978132724762,
+      "learning_rate": 1.2293179617212755e-05,
+      "loss": 0.3872,
       "step": 94320
     },
     {
-      "epoch": 3.32,
-      "learning_rate": 1.343459441130907e-05,
-      "loss": 0.2475,
+      "epoch": 3.399466609002775,
+      "grad_norm": 0.22973038256168365,
+      "learning_rate": 1.22906666184246e-05,
+      "loss": 0.386,
       "step": 94325
     },
     {
-      "epoch": 3.32,
-      "learning_rate": 1.3432068927894118e-05,
-      "loss": 0.2423,
+      "epoch": 3.3996468086639995,
+      "grad_norm": 0.18079350888729095,
+      "learning_rate": 1.2288153792798642e-05,
+      "loss": 0.3672,
       "step": 94330
     },
     {
-      "epoch": 3.32,
-      "learning_rate": 1.3429543594681133e-05,
-      "loss": 0.267,
+      "epoch": 3.399827008325224,
+      "grad_norm": 0.2188836932182312,
+      "learning_rate": 1.2285641140369147e-05,
+      "loss": 0.4015,
       "step": 94335
     },
     {
-      "epoch": 3.32,
-      "learning_rate": 1.3427018411702901e-05,
-      "loss": 0.2778,
+      "epoch": 3.400007207986449,
+      "grad_norm": 0.2155511975288391,
+      "learning_rate": 1.2283128661170313e-05,
+      "loss": 0.358,
       "step": 94340
     },
     {
-      "epoch": 3.32,
-      "learning_rate": 1.342449337899221e-05,
-      "loss": 0.2546,
+      "epoch": 3.4001874076476737,
+      "grad_norm": 0.2768744230270386,
+      "learning_rate": 1.2280616355236397e-05,
+      "loss": 0.4151,
       "step": 94345
     },
     {
-      "epoch": 3.32,
-      "learning_rate": 1.3421968496581827e-05,
-      "loss": 0.2568,
+      "epoch": 3.4003676073088984,
+      "grad_norm": 0.20749439299106598,
+      "learning_rate": 1.2278104222601618e-05,
+      "loss": 0.3595,
       "step": 94350
     },
     {
-      "epoch": 3.32,
-      "learning_rate": 1.3419443764504569e-05,
-      "loss": 0.2568,
+      "epoch": 3.4005478069701227,
+      "grad_norm": 0.28811633586883545,
+      "learning_rate": 1.227559226330021e-05,
+      "loss": 0.4218,
       "step": 94355
     },
     {
-      "epoch": 3.32,
-      "learning_rate": 1.3416919182793192e-05,
-      "loss": 0.2476,
+      "epoch": 3.4007280066313474,
+      "grad_norm": 0.24326983094215393,
+      "learning_rate": 1.2273080477366391e-05,
+      "loss": 0.3668,
       "step": 94360
     },
     {
-      "epoch": 3.32,
-      "learning_rate": 1.3414394751480492e-05,
-      "loss": 0.2513,
+      "epoch": 3.400908206292572,
+      "grad_norm": 0.20822405815124512,
+      "learning_rate": 1.2270568864834378e-05,
+      "loss": 0.3486,
       "step": 94365
     },
     {
-      "epoch": 3.32,
-      "learning_rate": 1.3411870470599233e-05,
-      "loss": 0.2527,
+      "epoch": 3.401088405953797,
+      "grad_norm": 0.21231502294540405,
+      "learning_rate": 1.2268057425738408e-05,
+      "loss": 0.3724,
       "step": 94370
     },
     {
-      "epoch": 3.32,
-      "learning_rate": 1.340934634018221e-05,
-      "loss": 0.2936,
+      "epoch": 3.401268605615021,
+      "grad_norm": 0.21015721559524536,
+      "learning_rate": 1.2265546160112692e-05,
+      "loss": 0.3906,
       "step": 94375
     },
     {
-      "epoch": 3.32,
-      "learning_rate": 1.3406822360262183e-05,
-      "loss": 0.2555,
+      "epoch": 3.401448805276246,
+      "grad_norm": 0.20890061557292938,
+      "learning_rate": 1.2263035067991443e-05,
+      "loss": 0.3918,
       "step": 94380
     },
     {
-      "epoch": 3.32,
-      "learning_rate": 1.3404298530871922e-05,
-      "loss": 0.2488,
+      "epoch": 3.4016290049374707,
+      "grad_norm": 0.23352204263210297,
+      "learning_rate": 1.2260524149408875e-05,
+      "loss": 0.3573,
       "step": 94385
     },
     {
-      "epoch": 3.32,
-      "learning_rate": 1.3401774852044211e-05,
-      "loss": 0.2521,
+      "epoch": 3.4018092045986954,
+      "grad_norm": 0.24678070843219757,
+      "learning_rate": 1.2258013404399202e-05,
+      "loss": 0.3975,
       "step": 94390
     },
     {
-      "epoch": 3.32,
-      "learning_rate": 1.3399251323811807e-05,
-      "loss": 0.2486,
+      "epoch": 3.40198940425992,
+      "grad_norm": 0.27147865295410156,
+      "learning_rate": 1.2255502832996619e-05,
+      "loss": 0.3848,
       "step": 94395
     },
     {
-      "epoch": 3.32,
-      "learning_rate": 1.3396727946207482e-05,
-      "loss": 0.2644,
+      "epoch": 3.402169603921145,
+      "grad_norm": 0.2716798484325409,
+      "learning_rate": 1.225299243523535e-05,
+      "loss": 0.3492,
       "step": 94400
     },
     {
-      "epoch": 3.32,
-      "learning_rate": 1.3394204719263989e-05,
-      "loss": 0.2582,
+      "epoch": 3.402349803582369,
+      "grad_norm": 0.2307744324207306,
+      "learning_rate": 1.2250482211149591e-05,
+      "loss": 0.3821,
       "step": 94405
     },
     {
-      "epoch": 3.32,
-      "learning_rate": 1.3391681643014109e-05,
-      "loss": 0.2385,
+      "epoch": 3.402530003243594,
+      "grad_norm": 0.2626960575580597,
+      "learning_rate": 1.2247972160773544e-05,
+      "loss": 0.3784,
       "step": 94410
     },
     {
-      "epoch": 3.32,
-      "learning_rate": 1.338915871749058e-05,
-      "loss": 0.2645,
+      "epoch": 3.4027102029048186,
+      "grad_norm": 0.29123780131340027,
+      "learning_rate": 1.224546228414141e-05,
+      "loss": 0.3983,
       "step": 94415
     },
     {
-      "epoch": 3.32,
-      "learning_rate": 1.3386635942726183e-05,
-      "loss": 0.2564,
+      "epoch": 3.4028904025660434,
+      "grad_norm": 0.23781827092170715,
+      "learning_rate": 1.224295258128737e-05,
+      "loss": 0.4109,
       "step": 94420
     },
     {
-      "epoch": 3.32,
-      "learning_rate": 1.338411331875366e-05,
-      "loss": 0.255,
+      "epoch": 3.4030706022272676,
+      "grad_norm": 0.2665836811065674,
+      "learning_rate": 1.2240443052245651e-05,
+      "loss": 0.399,
       "step": 94425
     },
     {
-      "epoch": 3.32,
-      "learning_rate": 1.3381590845605776e-05,
-      "loss": 0.2679,
+      "epoch": 3.4032508018884924,
+      "grad_norm": 0.27731844782829285,
+      "learning_rate": 1.2237933697050416e-05,
+      "loss": 0.4108,
       "step": 94430
     },
     {
-      "epoch": 3.32,
-      "learning_rate": 1.337906852331528e-05,
-      "loss": 0.2662,
+      "epoch": 3.403431001549717,
+      "grad_norm": 0.20404113829135895,
+      "learning_rate": 1.2235424515735855e-05,
+      "loss": 0.3812,
       "step": 94435
     },
     {
-      "epoch": 3.32,
-      "learning_rate": 1.337654635191492e-05,
-      "loss": 0.2493,
+      "epoch": 3.403611201210942,
+      "grad_norm": 0.19817714393138885,
+      "learning_rate": 1.2232915508336173e-05,
+      "loss": 0.3763,
       "step": 94440
     },
     {
-      "epoch": 3.32,
-      "learning_rate": 1.3374024331437435e-05,
-      "loss": 0.2521,
+      "epoch": 3.4037914008721666,
+      "grad_norm": 0.2466011792421341,
+      "learning_rate": 1.2230406674885536e-05,
+      "loss": 0.3551,
       "step": 94445
     },
     {
-      "epoch": 3.32,
-      "learning_rate": 1.3371502461915592e-05,
-      "loss": 0.2722,
+      "epoch": 3.403971600533391,
+      "grad_norm": 0.2783677577972412,
+      "learning_rate": 1.2227898015418154e-05,
+      "loss": 0.3643,
       "step": 94450
     },
     {
-      "epoch": 3.32,
-      "learning_rate": 1.3368980743382126e-05,
-      "loss": 0.2694,
+      "epoch": 3.4041518001946156,
+      "grad_norm": 0.2562623620033264,
+      "learning_rate": 1.222538952996818e-05,
+      "loss": 0.4023,
       "step": 94455
     },
     {
-      "epoch": 3.32,
-      "learning_rate": 1.336645917586978e-05,
-      "loss": 0.2432,
+      "epoch": 3.4043319998558403,
+      "grad_norm": 0.2551863193511963,
+      "learning_rate": 1.222288121856979e-05,
+      "loss": 0.3695,
       "step": 94460
     },
     {
-      "epoch": 3.32,
-      "learning_rate": 1.3363937759411287e-05,
-      "loss": 0.2721,
+      "epoch": 3.404512199517065,
+      "grad_norm": 0.27139002084732056,
+      "learning_rate": 1.222037308125718e-05,
+      "loss": 0.3763,
       "step": 94465
     },
     {
-      "epoch": 3.32,
-      "learning_rate": 1.336141649403939e-05,
-      "loss": 0.2625,
+      "epoch": 3.4046923991782894,
+      "grad_norm": 0.21469560265541077,
+      "learning_rate": 1.2217865118064512e-05,
+      "loss": 0.3879,
       "step": 94470
     },
     {
-      "epoch": 3.32,
-      "learning_rate": 1.3358895379786842e-05,
-      "loss": 0.2666,
+      "epoch": 3.404872598839514,
+      "grad_norm": 0.20434612035751343,
+      "learning_rate": 1.2215357329025956e-05,
+      "loss": 0.3651,
       "step": 94475
     },
     {
-      "epoch": 3.32,
-      "learning_rate": 1.3356374416686368e-05,
-      "loss": 0.2839,
+      "epoch": 3.405052798500739,
+      "grad_norm": 0.26074668765068054,
+      "learning_rate": 1.2212849714175684e-05,
+      "loss": 0.3782,
       "step": 94480
     },
     {
-      "epoch": 3.32,
-      "learning_rate": 1.3353853604770683e-05,
-      "loss": 0.2386,
+      "epoch": 3.4052329981619636,
+      "grad_norm": 0.22299236059188843,
+      "learning_rate": 1.2210342273547848e-05,
+      "loss": 0.3674,
       "step": 94485
     },
     {
-      "epoch": 3.32,
-      "learning_rate": 1.3351332944072548e-05,
-      "loss": 0.2651,
+      "epoch": 3.4054131978231883,
+      "grad_norm": 0.2286439836025238,
+      "learning_rate": 1.2207835007176632e-05,
+      "loss": 0.3969,
       "step": 94490
     },
     {
-      "epoch": 3.32,
-      "learning_rate": 1.3348812434624677e-05,
-      "loss": 0.2861,
+      "epoch": 3.4055933974844126,
+      "grad_norm": 0.21935893595218658,
+      "learning_rate": 1.2205327915096187e-05,
+      "loss": 0.3621,
       "step": 94495
     },
     {
-      "epoch": 3.32,
-      "learning_rate": 1.3346292076459788e-05,
-      "loss": 0.261,
+      "epoch": 3.4057735971456373,
+      "grad_norm": 0.2769947946071625,
+      "learning_rate": 1.2202820997340673e-05,
+      "loss": 0.4064,
       "step": 94500
     },
     {
-      "epoch": 3.32,
-      "eval_loss": 0.25450006127357483,
-      "eval_runtime": 10.5436,
-      "eval_samples_per_second": 9.484,
-      "eval_steps_per_second": 9.484,
+      "epoch": 3.4057735971456373,
+      "eval_loss": 0.4305441975593567,
+      "eval_runtime": 3.5328,
+      "eval_samples_per_second": 28.307,
+      "eval_steps_per_second": 7.077,
       "step": 94500
     },
     {
-      "epoch": 3.32,
-      "learning_rate": 1.3343771869610627e-05,
-      "loss": 0.2545,
+      "epoch": 3.405953796806862,
+      "grad_norm": 0.22937120497226715,
+      "learning_rate": 1.2200314253944243e-05,
+      "loss": 0.4212,
       "step": 94505
     },
     {
-      "epoch": 3.33,
-      "learning_rate": 1.3341251814109906e-05,
-      "loss": 0.2607,
+      "epoch": 3.406133996468087,
+      "grad_norm": 0.2228795886039734,
+      "learning_rate": 1.2197807684941054e-05,
+      "loss": 0.3915,
       "step": 94510
     },
     {
-      "epoch": 3.33,
-      "learning_rate": 1.3338731909990348e-05,
-      "loss": 0.2558,
+      "epoch": 3.406314196129311,
+      "grad_norm": 0.2143542468547821,
+      "learning_rate": 1.2195301290365248e-05,
+      "loss": 0.38,
       "step": 94515
     },
     {
-      "epoch": 3.33,
-      "learning_rate": 1.3336212157284658e-05,
-      "loss": 0.2538,
+      "epoch": 3.406494395790536,
+      "grad_norm": 0.22707048058509827,
+      "learning_rate": 1.2192795070251001e-05,
+      "loss": 0.3598,
       "step": 94520
     },
     {
-      "epoch": 3.33,
-      "learning_rate": 1.3333692556025568e-05,
-      "loss": 0.2324,
+      "epoch": 3.4066745954517605,
+      "grad_norm": 0.24929481744766235,
+      "learning_rate": 1.2190289024632424e-05,
+      "loss": 0.389,
       "step": 94525
     },
     {
-      "epoch": 3.33,
-      "learning_rate": 1.3331173106245804e-05,
-      "loss": 0.2737,
+      "epoch": 3.4068547951129853,
+      "grad_norm": 0.2553023397922516,
+      "learning_rate": 1.2187783153543689e-05,
+      "loss": 0.3685,
       "step": 94530
     },
     {
-      "epoch": 3.33,
-      "learning_rate": 1.3328653807978064e-05,
-      "loss": 0.2455,
+      "epoch": 3.40703499477421,
+      "grad_norm": 0.2211211770772934,
+      "learning_rate": 1.2185277457018926e-05,
+      "loss": 0.3369,
       "step": 94535
     },
     {
-      "epoch": 3.33,
-      "learning_rate": 1.3326134661255055e-05,
-      "loss": 0.2488,
+      "epoch": 3.4072151944354343,
+      "grad_norm": 0.27645251154899597,
+      "learning_rate": 1.2182771935092277e-05,
+      "loss": 0.3792,
       "step": 94540
     },
     {
-      "epoch": 3.33,
-      "learning_rate": 1.3323615666109502e-05,
-      "loss": 0.2556,
+      "epoch": 3.407395394096659,
+      "grad_norm": 0.23426538705825806,
+      "learning_rate": 1.2180266587797881e-05,
+      "loss": 0.3623,
       "step": 94545
     },
     {
-      "epoch": 3.33,
-      "learning_rate": 1.3321096822574106e-05,
-      "loss": 0.2651,
+      "epoch": 3.4075755937578838,
+      "grad_norm": 0.204464390873909,
+      "learning_rate": 1.2177761415169869e-05,
+      "loss": 0.3999,
       "step": 94550
     },
     {
-      "epoch": 3.33,
-      "learning_rate": 1.331857813068157e-05,
-      "loss": 0.2757,
+      "epoch": 3.4077557934191085,
+      "grad_norm": 0.22453001141548157,
+      "learning_rate": 1.2175256417242365e-05,
+      "loss": 0.3882,
       "step": 94555
     },
     {
-      "epoch": 3.33,
-      "learning_rate": 1.3316059590464591e-05,
-      "loss": 0.2661,
+      "epoch": 3.407935993080333,
+      "grad_norm": 0.22461944818496704,
+      "learning_rate": 1.217275159404952e-05,
+      "loss": 0.4007,
       "step": 94560
     },
     {
-      "epoch": 3.33,
-      "learning_rate": 1.3313541201955885e-05,
-      "loss": 0.2849,
+      "epoch": 3.4081161927415575,
+      "grad_norm": 0.2279561460018158,
+      "learning_rate": 1.2170246945625451e-05,
+      "loss": 0.3697,
       "step": 94565
     },
     {
-      "epoch": 3.33,
-      "learning_rate": 1.3311022965188147e-05,
-      "loss": 0.2361,
+      "epoch": 3.4082963924027823,
+      "grad_norm": 0.20913352072238922,
+      "learning_rate": 1.2167742472004284e-05,
+      "loss": 0.3608,
       "step": 94570
     },
     {
-      "epoch": 3.33,
-      "learning_rate": 1.330850488019406e-05,
-      "loss": 0.2496,
+      "epoch": 3.408476592064007,
+      "grad_norm": 0.2152101993560791,
+      "learning_rate": 1.2165238173220139e-05,
+      "loss": 0.3632,
       "step": 94575
     },
     {
-      "epoch": 3.33,
-      "learning_rate": 1.3305986947006343e-05,
-      "loss": 0.2545,
+      "epoch": 3.4086567917252317,
+      "grad_norm": 0.21769098937511444,
+      "learning_rate": 1.216273404930713e-05,
+      "loss": 0.3927,
       "step": 94580
     },
     {
-      "epoch": 3.33,
-      "learning_rate": 1.3303469165657673e-05,
-      "loss": 0.2494,
+      "epoch": 3.408836991386456,
+      "grad_norm": 0.22281882166862488,
+      "learning_rate": 1.2160230100299397e-05,
+      "loss": 0.3571,
       "step": 94585
     },
     {
-      "epoch": 3.33,
-      "learning_rate": 1.330095153618075e-05,
-      "loss": 0.2668,
+      "epoch": 3.4090171910476808,
+      "grad_norm": 0.20765550434589386,
+      "learning_rate": 1.2157726326231041e-05,
+      "loss": 0.3766,
       "step": 94590
     },
     {
-      "epoch": 3.33,
-      "learning_rate": 1.3298434058608267e-05,
-      "loss": 0.2508,
+      "epoch": 3.4091973907089055,
+      "grad_norm": 0.23559367656707764,
+      "learning_rate": 1.215522272713618e-05,
+      "loss": 0.3658,
       "step": 94595
     },
     {
-      "epoch": 3.33,
-      "learning_rate": 1.3295916732972896e-05,
-      "loss": 0.2632,
+      "epoch": 3.40937759037013,
+      "grad_norm": 0.23205064237117767,
+      "learning_rate": 1.2152719303048919e-05,
+      "loss": 0.4183,
       "step": 94600
     },
     {
-      "epoch": 3.33,
-      "learning_rate": 1.329339955930734e-05,
-      "loss": 0.2781,
+      "epoch": 3.4095577900313545,
+      "grad_norm": 0.2513408958911896,
+      "learning_rate": 1.2150216054003361e-05,
+      "loss": 0.3678,
       "step": 94605
     },
     {
-      "epoch": 3.33,
-      "learning_rate": 1.3290882537644278e-05,
-      "loss": 0.2595,
+      "epoch": 3.4097379896925792,
+      "grad_norm": 0.24810358881950378,
+      "learning_rate": 1.2147712980033629e-05,
+      "loss": 0.384,
       "step": 94610
     },
     {
-      "epoch": 3.33,
-      "learning_rate": 1.3288365668016378e-05,
-      "loss": 0.264,
+      "epoch": 3.409918189353804,
+      "grad_norm": Infinity,
+      "learning_rate": 1.2145710646935363e-05,
+      "loss": 0.3946,
       "step": 94615
     },
     {
-      "epoch": 3.33,
-      "learning_rate": 1.3285848950456341e-05,
-      "loss": 0.2602,
+      "epoch": 3.4100983890150287,
+      "grad_norm": 0.2287181168794632,
+      "learning_rate": 1.2143207888188035e-05,
+      "loss": 0.3305,
       "step": 94620
     },
     {
-      "epoch": 3.33,
-      "learning_rate": 1.3283332384996838e-05,
-      "loss": 0.2481,
+      "epoch": 3.4102785886762534,
+      "grad_norm": 0.28771016001701355,
+      "learning_rate": 1.2140705304612024e-05,
+      "loss": 0.4085,
       "step": 94625
     },
     {
-      "epoch": 3.33,
-      "learning_rate": 1.3280815971670535e-05,
-      "loss": 0.2513,
+      "epoch": 3.4104587883374777,
+      "grad_norm": 0.23076020181179047,
+      "learning_rate": 1.2138202896241413e-05,
+      "loss": 0.3834,
       "step": 94630
     },
     {
-      "epoch": 3.33,
-      "learning_rate": 1.3278299710510117e-05,
-      "loss": 0.2562,
+      "epoch": 3.4106389879987025,
+      "grad_norm": 0.25198495388031006,
+      "learning_rate": 1.2135700663110295e-05,
+      "loss": 0.3626,
       "step": 94635
     },
     {
-      "epoch": 3.33,
-      "learning_rate": 1.3275783601548248e-05,
-      "loss": 0.2786,
+      "epoch": 3.410819187659927,
+      "grad_norm": 0.19510357081890106,
+      "learning_rate": 1.2133198605252767e-05,
+      "loss": 0.4004,
       "step": 94640
     },
     {
-      "epoch": 3.33,
-      "learning_rate": 1.3273267644817607e-05,
-      "loss": 0.2617,
+      "epoch": 3.410999387321152,
+      "grad_norm": 0.21469490230083466,
+      "learning_rate": 1.2130696722702917e-05,
+      "loss": 0.3654,
       "step": 94645
     },
     {
-      "epoch": 3.33,
-      "learning_rate": 1.3270751840350861e-05,
-      "loss": 0.2527,
+      "epoch": 3.4111795869823767,
+      "grad_norm": 0.22288212180137634,
+      "learning_rate": 1.212819501549482e-05,
+      "loss": 0.3701,
       "step": 94650
     },
     {
-      "epoch": 3.33,
-      "learning_rate": 1.3268236188180661e-05,
-      "loss": 0.2464,
+      "epoch": 3.411359786643601,
+      "grad_norm": 0.25711116194725037,
+      "learning_rate": 1.2125693483662586e-05,
+      "loss": 0.3539,
       "step": 94655
     },
     {
-      "epoch": 3.33,
-      "learning_rate": 1.3265720688339695e-05,
-      "loss": 0.2613,
+      "epoch": 3.4115399863048257,
+      "grad_norm": 0.30416256189346313,
+      "learning_rate": 1.2123192127240286e-05,
+      "loss": 0.3982,
       "step": 94660
     },
     {
-      "epoch": 3.33,
-      "learning_rate": 1.326320534086061e-05,
-      "loss": 0.2414,
+      "epoch": 3.4117201859660504,
+      "grad_norm": 0.23342035710811615,
+      "learning_rate": 1.2120690946262e-05,
+      "loss": 0.3734,
       "step": 94665
     },
     {
-      "epoch": 3.33,
-      "learning_rate": 1.326069014577607e-05,
-      "loss": 0.2516,
+      "epoch": 3.411900385627275,
+      "grad_norm": 0.3036738932132721,
+      "learning_rate": 1.2118189940761807e-05,
+      "loss": 0.3875,
       "step": 94670
     },
     {
-      "epoch": 3.33,
-      "learning_rate": 1.3258175103118719e-05,
-      "loss": 0.2528,
+      "epoch": 3.4120805852885,
+      "grad_norm": 0.20019620656967163,
+      "learning_rate": 1.2115689110773771e-05,
+      "loss": 0.3689,
       "step": 94675
     },
     {
-      "epoch": 3.33,
-      "learning_rate": 1.325566021292124e-05,
-      "loss": 0.2653,
+      "epoch": 3.412260784949724,
+      "grad_norm": 0.23406051099300385,
+      "learning_rate": 1.2113188456331987e-05,
+      "loss": 0.3826,
       "step": 94680
     },
     {
-      "epoch": 3.33,
-      "learning_rate": 1.3253145475216261e-05,
-      "loss": 0.2526,
+      "epoch": 3.412440984610949,
+      "grad_norm": 0.1972576528787613,
+      "learning_rate": 1.2110687977470522e-05,
+      "loss": 0.3868,
       "step": 94685
     },
     {
-      "epoch": 3.33,
-      "learning_rate": 1.3250630890036458e-05,
-      "loss": 0.2568,
+      "epoch": 3.4126211842721736,
+      "grad_norm": 0.21116188168525696,
+      "learning_rate": 1.210818767422342e-05,
+      "loss": 0.3706,
       "step": 94690
     },
     {
-      "epoch": 3.33,
-      "learning_rate": 1.3248116457414461e-05,
-      "loss": 0.289,
+      "epoch": 3.4128013839333984,
+      "grad_norm": 0.23035746812820435,
+      "learning_rate": 1.2105687546624777e-05,
+      "loss": 0.3828,
       "step": 94695
     },
     {
-      "epoch": 3.33,
-      "learning_rate": 1.3245602177382936e-05,
-      "loss": 0.2483,
+      "epoch": 3.4129815835946227,
+      "grad_norm": 0.25209763646125793,
+      "learning_rate": 1.2103187594708635e-05,
+      "loss": 0.4037,
       "step": 94700
     },
     {
-      "epoch": 3.33,
-      "learning_rate": 1.3243088049974516e-05,
-      "loss": 0.2605,
+      "epoch": 3.4131617832558474,
+      "grad_norm": 0.23511983454227448,
+      "learning_rate": 1.2100687818509083e-05,
+      "loss": 0.3718,
       "step": 94705
     },
     {
-      "epoch": 3.33,
-      "learning_rate": 1.3240574075221856e-05,
-      "loss": 0.2474,
+      "epoch": 3.413341982917072,
+      "grad_norm": 0.2819865047931671,
+      "learning_rate": 1.2098188218060153e-05,
+      "loss": 0.4107,
       "step": 94710
     },
     {
-      "epoch": 3.33,
-      "learning_rate": 1.323806025315758e-05,
-      "loss": 0.2783,
+      "epoch": 3.413522182578297,
+      "grad_norm": 0.21473956108093262,
+      "learning_rate": 1.2095688793395898e-05,
+      "loss": 0.4104,
       "step": 94715
     },
     {
-      "epoch": 3.33,
-      "learning_rate": 1.3235546583814348e-05,
-      "loss": 0.2654,
+      "epoch": 3.4137023822395216,
+      "grad_norm": 0.24196425080299377,
+      "learning_rate": 1.2093189544550398e-05,
+      "loss": 0.4131,
       "step": 94720
     },
     {
-      "epoch": 3.33,
-      "learning_rate": 1.3233033067224792e-05,
-      "loss": 0.2342,
+      "epoch": 3.413882581900746,
+      "grad_norm": 0.2057386040687561,
+      "learning_rate": 1.2090690471557689e-05,
+      "loss": 0.3766,
       "step": 94725
     },
     {
-      "epoch": 3.33,
-      "learning_rate": 1.3230519703421537e-05,
-      "loss": 0.2719,
+      "epoch": 3.4140627815619706,
+      "grad_norm": 0.268764466047287,
+      "learning_rate": 1.2088191574451827e-05,
+      "loss": 0.375,
       "step": 94730
     },
     {
-      "epoch": 3.33,
-      "learning_rate": 1.3228006492437236e-05,
-      "loss": 0.2502,
+      "epoch": 3.4142429812231954,
+      "grad_norm": 0.27153804898262024,
+      "learning_rate": 1.2085692853266852e-05,
+      "loss": 0.4177,
       "step": 94735
     },
     {
-      "epoch": 3.33,
-      "learning_rate": 1.3225493434304514e-05,
-      "loss": 0.2529,
+      "epoch": 3.41442318088442,
+      "grad_norm": 0.23466871678829193,
+      "learning_rate": 1.2083194308036803e-05,
+      "loss": 0.4096,
       "step": 94740
     },
     {
-      "epoch": 3.33,
-      "learning_rate": 1.322298052905599e-05,
-      "loss": 0.2508,
+      "epoch": 3.4146033805456444,
+      "grad_norm": 0.22995726764202118,
+      "learning_rate": 1.2080695938795739e-05,
+      "loss": 0.359,
       "step": 94745
     },
     {
-      "epoch": 3.33,
-      "learning_rate": 1.3220467776724301e-05,
-      "loss": 0.252,
+      "epoch": 3.414783580206869,
+      "grad_norm": 0.2112264633178711,
+      "learning_rate": 1.2078197745577693e-05,
+      "loss": 0.3687,
       "step": 94750
     },
     {
-      "epoch": 3.33,
-      "learning_rate": 1.3217955177342087e-05,
-      "loss": 0.2732,
+      "epoch": 3.414963779868094,
+      "grad_norm": 0.19448639452457428,
+      "learning_rate": 1.20756997284167e-05,
+      "loss": 0.3947,
       "step": 94755
     },
     {
-      "epoch": 3.33,
-      "learning_rate": 1.321544273094196e-05,
-      "loss": 0.264,
+      "epoch": 3.4151439795293186,
+      "grad_norm": 0.20815762877464294,
+      "learning_rate": 1.2073201887346797e-05,
+      "loss": 0.4007,
       "step": 94760
     },
     {
-      "epoch": 3.33,
-      "learning_rate": 1.3212930437556542e-05,
-      "loss": 0.2485,
+      "epoch": 3.4153241791905433,
+      "grad_norm": 0.21343150734901428,
+      "learning_rate": 1.2070704222402016e-05,
+      "loss": 0.3743,
       "step": 94765
     },
     {
-      "epoch": 3.33,
-      "learning_rate": 1.3210418297218446e-05,
-      "loss": 0.2582,
+      "epoch": 3.4155043788517676,
+      "grad_norm": 0.2795332372188568,
+      "learning_rate": 1.2068206733616375e-05,
+      "loss": 0.3962,
       "step": 94770
     },
     {
-      "epoch": 3.33,
-      "learning_rate": 1.3207906309960305e-05,
-      "loss": 0.2761,
+      "epoch": 3.4156845785129923,
+      "grad_norm": 0.26403021812438965,
+      "learning_rate": 1.2065709421023923e-05,
+      "loss": 0.3737,
       "step": 94775
     },
     {
-      "epoch": 3.33,
-      "learning_rate": 1.3205394475814732e-05,
-      "loss": 0.2472,
+      "epoch": 3.415864778174217,
+      "grad_norm": 0.21119317412376404,
+      "learning_rate": 1.2063212284658679e-05,
+      "loss": 0.363,
       "step": 94780
     },
     {
-      "epoch": 3.33,
-      "learning_rate": 1.320288279481434e-05,
-      "loss": 0.2437,
+      "epoch": 3.416044977835442,
+      "grad_norm": 0.23738868534564972,
+      "learning_rate": 1.206071532455466e-05,
+      "loss": 0.3517,
       "step": 94785
     },
     {
-      "epoch": 3.33,
-      "learning_rate": 1.320037126699173e-05,
-      "loss": 0.2663,
+      "epoch": 3.416225177496666,
+      "grad_norm": 0.22581474483013153,
+      "learning_rate": 1.2058218540745891e-05,
+      "loss": 0.3721,
       "step": 94790
     },
     {
-      "epoch": 3.34,
-      "learning_rate": 1.319785989237953e-05,
-      "loss": 0.2613,
+      "epoch": 3.416405377157891,
+      "grad_norm": 0.20880122482776642,
+      "learning_rate": 1.2055721933266381e-05,
+      "loss": 0.3756,
       "step": 94795
     },
     {
-      "epoch": 3.34,
-      "learning_rate": 1.3195348671010333e-05,
-      "loss": 0.2638,
+      "epoch": 3.4165855768191156,
+      "grad_norm": 0.19523420929908752,
+      "learning_rate": 1.205322550215017e-05,
+      "loss": 0.3655,
       "step": 94800
     },
     {
-      "epoch": 3.34,
-      "learning_rate": 1.3192837602916764e-05,
-      "loss": 0.2416,
+      "epoch": 3.4167657764803403,
+      "grad_norm": 0.26016107201576233,
+      "learning_rate": 1.2050729247431238e-05,
+      "loss": 0.3844,
       "step": 94805
     },
     {
-      "epoch": 3.34,
-      "learning_rate": 1.3190326688131407e-05,
-      "loss": 0.2669,
+      "epoch": 3.416945976141565,
+      "grad_norm": 0.21247242391109467,
+      "learning_rate": 1.2048233169143623e-05,
+      "loss": 0.409,
       "step": 94810
     },
     {
-      "epoch": 3.34,
-      "learning_rate": 1.3187815926686886e-05,
-      "loss": 0.2434,
+      "epoch": 3.4171261758027893,
+      "grad_norm": 0.27394580841064453,
+      "learning_rate": 1.204573726732132e-05,
+      "loss": 0.438,
       "step": 94815
     },
     {
-      "epoch": 3.34,
-      "learning_rate": 1.318530531861579e-05,
-      "loss": 0.2341,
+      "epoch": 3.417306375464014,
+      "grad_norm": 0.22982816398143768,
+      "learning_rate": 1.204324154199834e-05,
+      "loss": 0.3659,
       "step": 94820
     },
     {
-      "epoch": 3.34,
-      "learning_rate": 1.318279486395072e-05,
-      "loss": 0.2696,
+      "epoch": 3.417486575125239,
+      "grad_norm": 0.2429938167333603,
+      "learning_rate": 1.2040745993208685e-05,
+      "loss": 0.374,
       "step": 94825
     },
     {
-      "epoch": 3.34,
-      "learning_rate": 1.318028456272426e-05,
-      "loss": 0.2584,
+      "epoch": 3.4176667747864635,
+      "grad_norm": 0.2131299376487732,
+      "learning_rate": 1.2038250620986358e-05,
+      "loss": 0.4089,
       "step": 94830
     },
     {
-      "epoch": 3.34,
-      "learning_rate": 1.3177774414969023e-05,
-      "loss": 0.2407,
+      "epoch": 3.417846974447688,
+      "grad_norm": 0.2698366641998291,
+      "learning_rate": 1.2035755425365348e-05,
+      "loss": 0.3917,
       "step": 94835
     },
     {
-      "epoch": 3.34,
-      "learning_rate": 1.3175264420717594e-05,
-      "loss": 0.2605,
+      "epoch": 3.4180271741089125,
+      "grad_norm": 0.22712182998657227,
+      "learning_rate": 1.2033260406379668e-05,
+      "loss": 0.3908,
       "step": 94840
     },
     {
-      "epoch": 3.34,
-      "learning_rate": 1.3172754580002556e-05,
-      "loss": 0.2672,
+      "epoch": 3.4182073737701373,
+      "grad_norm": 0.25237366557121277,
+      "learning_rate": 1.2030765564063306e-05,
+      "loss": 0.4028,
       "step": 94845
     },
     {
-      "epoch": 3.34,
-      "learning_rate": 1.3170746817998638e-05,
-      "loss": 0.2632,
+      "epoch": 3.418387573431362,
+      "grad_norm": 0.22257663309574127,
+      "learning_rate": 1.2028270898450254e-05,
+      "loss": 0.3829,
       "step": 94850
     },
     {
-      "epoch": 3.34,
-      "learning_rate": 1.316823725373125e-05,
-      "loss": 0.2434,
+      "epoch": 3.4185677730925867,
+      "grad_norm": 0.22527682781219482,
+      "learning_rate": 1.2025776409574496e-05,
+      "loss": 0.3725,
       "step": 94855
     },
     {
-      "epoch": 3.34,
-      "learning_rate": 1.3165727843091493e-05,
-      "loss": 0.2421,
+      "epoch": 3.418747972753811,
+      "grad_norm": 0.24640600383281708,
+      "learning_rate": 1.2023282097470016e-05,
+      "loss": 0.4155,
       "step": 94860
     },
     {
-      "epoch": 3.34,
-      "learning_rate": 1.3163218586111972e-05,
-      "loss": 0.2645,
+      "epoch": 3.4189281724150358,
+      "grad_norm": 0.2200070023536682,
+      "learning_rate": 1.202078796217081e-05,
+      "loss": 0.3464,
       "step": 94865
     },
     {
-      "epoch": 3.34,
-      "learning_rate": 1.3160709482825257e-05,
-      "loss": 0.2445,
+      "epoch": 3.4191083720762605,
+      "grad_norm": 0.19475413858890533,
+      "learning_rate": 1.2018294003710867e-05,
+      "loss": 0.3841,
       "step": 94870
     },
     {
-      "epoch": 3.34,
-      "learning_rate": 1.3158200533263926e-05,
-      "loss": 0.2647,
+      "epoch": 3.4192885717374852,
+      "grad_norm": 0.2146916389465332,
+      "learning_rate": 1.2015800222124137e-05,
+      "loss": 0.3938,
       "step": 94875
     },
     {
-      "epoch": 3.34,
-      "learning_rate": 1.3155691737460545e-05,
-      "loss": 0.2784,
+      "epoch": 3.4194687713987095,
+      "grad_norm": 0.2292756289243698,
+      "learning_rate": 1.2013306617444625e-05,
+      "loss": 0.3847,
       "step": 94880
     },
     {
-      "epoch": 3.34,
-      "learning_rate": 1.3153183095447713e-05,
-      "loss": 0.25,
+      "epoch": 3.4196489710599343,
+      "grad_norm": 0.26337769627571106,
+      "learning_rate": 1.2010813189706285e-05,
+      "loss": 0.4117,
       "step": 94885
     },
     {
-      "epoch": 3.34,
-      "learning_rate": 1.3150674607257983e-05,
-      "loss": 0.2598,
+      "epoch": 3.419829170721159,
+      "grad_norm": 0.2760777175426483,
+      "learning_rate": 1.200831993894311e-05,
+      "loss": 0.3819,
       "step": 94890
     },
     {
-      "epoch": 3.34,
-      "learning_rate": 1.314816627292394e-05,
-      "loss": 0.2393,
+      "epoch": 3.4200093703823837,
+      "grad_norm": 0.22346773743629456,
+      "learning_rate": 1.200582686518907e-05,
+      "loss": 0.3848,
       "step": 94895
     },
     {
-      "epoch": 3.34,
-      "learning_rate": 1.3145658092478141e-05,
-      "loss": 0.2418,
+      "epoch": 3.4201895700436085,
+      "grad_norm": 0.2624737322330475,
+      "learning_rate": 1.2003333968478106e-05,
+      "loss": 0.3789,
       "step": 94900
     },
     {
-      "epoch": 3.34,
-      "learning_rate": 1.314315006595317e-05,
-      "loss": 0.2527,
+      "epoch": 3.420369769704833,
+      "grad_norm": 0.33500173687934875,
+      "learning_rate": 1.200084124884421e-05,
+      "loss": 0.3858,
       "step": 94905
     },
     {
-      "epoch": 3.34,
-      "learning_rate": 1.3140642193381578e-05,
-      "loss": 0.2348,
+      "epoch": 3.4205499693660575,
+      "grad_norm": 0.19514542818069458,
+      "learning_rate": 1.1998348706321333e-05,
+      "loss": 0.3719,
       "step": 94910
     },
     {
-      "epoch": 3.34,
-      "learning_rate": 1.3138134474795921e-05,
-      "loss": 0.2663,
+      "epoch": 3.420730169027282,
+      "grad_norm": 0.25635966658592224,
+      "learning_rate": 1.1995856340943439e-05,
+      "loss": 0.3643,
       "step": 94915
     },
     {
-      "epoch": 3.34,
-      "learning_rate": 1.3135626910228784e-05,
-      "loss": 0.2632,
+      "epoch": 3.420910368688507,
+      "grad_norm": 0.1672072410583496,
+      "learning_rate": 1.1993364152744485e-05,
+      "loss": 0.3837,
       "step": 94920
     },
     {
-      "epoch": 3.34,
-      "learning_rate": 1.3133119499712712e-05,
-      "loss": 0.2536,
+      "epoch": 3.4210905683497317,
+      "grad_norm": 0.23457878828048706,
+      "learning_rate": 1.1990872141758417e-05,
+      "loss": 0.3667,
       "step": 94925
     },
     {
-      "epoch": 3.34,
-      "learning_rate": 1.313061224328026e-05,
-      "loss": 0.2499,
+      "epoch": 3.421270768010956,
+      "grad_norm": 0.23055920004844666,
+      "learning_rate": 1.1988380308019207e-05,
+      "loss": 0.3776,
       "step": 94930
     },
     {
-      "epoch": 3.34,
-      "learning_rate": 1.3128105140963981e-05,
-      "loss": 0.2533,
+      "epoch": 3.4214509676721807,
+      "grad_norm": 0.24517036974430084,
+      "learning_rate": 1.1985888651560795e-05,
+      "loss": 0.3673,
       "step": 94935
     },
     {
-      "epoch": 3.34,
-      "learning_rate": 1.312559819279644e-05,
-      "loss": 0.2737,
+      "epoch": 3.4216311673334054,
+      "grad_norm": 0.22914618253707886,
+      "learning_rate": 1.1983397172417129e-05,
+      "loss": 0.3627,
       "step": 94940
     },
     {
-      "epoch": 3.34,
-      "learning_rate": 1.312309139881017e-05,
-      "loss": 0.2586,
+      "epoch": 3.42181136699463,
+      "grad_norm": 0.2351149469614029,
+      "learning_rate": 1.1980905870622159e-05,
+      "loss": 0.3745,
       "step": 94945
     },
     {
-      "epoch": 3.34,
-      "learning_rate": 1.3120584759037741e-05,
-      "loss": 0.2556,
+      "epoch": 3.421991566655855,
+      "grad_norm": 0.20726028084754944,
+      "learning_rate": 1.1978414746209826e-05,
+      "loss": 0.3554,
       "step": 94950
     },
     {
-      "epoch": 3.34,
-      "learning_rate": 1.311807827351168e-05,
-      "loss": 0.2735,
+      "epoch": 3.422171766317079,
+      "grad_norm": 0.2245931476354599,
+      "learning_rate": 1.1975923799214061e-05,
+      "loss": 0.4048,
       "step": 94955
     },
     {
-      "epoch": 3.34,
-      "learning_rate": 1.3115571942264554e-05,
-      "loss": 0.2577,
+      "epoch": 3.422351965978304,
+      "grad_norm": 0.2029707431793213,
+      "learning_rate": 1.1973433029668821e-05,
+      "loss": 0.3594,
       "step": 94960
     },
     {
-      "epoch": 3.34,
-      "learning_rate": 1.3113065765328892e-05,
-      "loss": 0.2492,
+      "epoch": 3.4225321656395287,
+      "grad_norm": 0.22745849192142487,
+      "learning_rate": 1.1970942437608035e-05,
+      "loss": 0.4059,
       "step": 94965
     },
     {
-      "epoch": 3.34,
-      "learning_rate": 1.3110559742737239e-05,
-      "loss": 0.2564,
+      "epoch": 3.4227123653007534,
+      "grad_norm": 0.283438116312027,
+      "learning_rate": 1.1968452023065635e-05,
+      "loss": 0.4164,
       "step": 94970
     },
     {
-      "epoch": 3.34,
-      "learning_rate": 1.310805387452212e-05,
-      "loss": 0.2664,
+      "epoch": 3.4228925649619777,
+      "grad_norm": 0.181619331240654,
+      "learning_rate": 1.1965961786075556e-05,
+      "loss": 0.3374,
       "step": 94975
     },
     {
-      "epoch": 3.34,
-      "learning_rate": 1.3105548160716097e-05,
-      "loss": 0.2677,
+      "epoch": 3.4230727646232024,
+      "grad_norm": 0.1780719757080078,
+      "learning_rate": 1.196347172667171e-05,
+      "loss": 0.3894,
       "step": 94980
     },
     {
-      "epoch": 3.34,
-      "learning_rate": 1.310304260135169e-05,
-      "loss": 0.2575,
+      "epoch": 3.423252964284427,
+      "grad_norm": 0.24398981034755707,
+      "learning_rate": 1.196098184488806e-05,
+      "loss": 0.3389,
       "step": 94985
     },
     {
-      "epoch": 3.34,
-      "learning_rate": 1.3100537196461427e-05,
-      "loss": 0.2438,
+      "epoch": 3.423433163945652,
+      "grad_norm": 0.21179978549480438,
+      "learning_rate": 1.1958492140758497e-05,
+      "loss": 0.3897,
       "step": 94990
     },
     {
-      "epoch": 3.34,
-      "learning_rate": 1.3098031946077854e-05,
-      "loss": 0.2433,
+      "epoch": 3.4236133636068766,
+      "grad_norm": 0.2665993273258209,
+      "learning_rate": 1.1956002614316946e-05,
+      "loss": 0.3834,
       "step": 94995
     },
     {
-      "epoch": 3.34,
-      "learning_rate": 1.3095526850233486e-05,
-      "loss": 0.2506,
+      "epoch": 3.423793563268101,
+      "grad_norm": 0.24159680306911469,
+      "learning_rate": 1.1953513265597344e-05,
+      "loss": 0.3963,
       "step": 95000
     },
     {
-      "epoch": 3.34,
-      "eval_loss": 0.25423306226730347,
-      "eval_runtime": 10.5407,
-      "eval_samples_per_second": 9.487,
-      "eval_steps_per_second": 9.487,
+      "epoch": 3.423793563268101,
+      "eval_loss": 0.4299665093421936,
+      "eval_runtime": 3.5329,
+      "eval_samples_per_second": 28.305,
+      "eval_steps_per_second": 7.076,
       "step": 95000
     },
     {
-      "epoch": 3.34,
-      "learning_rate": 1.3093021908960862e-05,
-      "loss": 0.2546,
+      "epoch": 3.4239737629293256,
+      "grad_norm": 0.2492501586675644,
+      "learning_rate": 1.1951024094633594e-05,
+      "loss": 0.3966,
       "step": 95005
     },
     {
-      "epoch": 3.34,
-      "learning_rate": 1.3090517122292505e-05,
-      "loss": 0.234,
+      "epoch": 3.4241539625905504,
+      "grad_norm": 0.2158864438533783,
+      "learning_rate": 1.1948535101459619e-05,
+      "loss": 0.3756,
       "step": 95010
     },
     {
-      "epoch": 3.34,
-      "learning_rate": 1.3088012490260924e-05,
-      "loss": 0.2825,
+      "epoch": 3.424334162251775,
+      "grad_norm": 0.20892734825611115,
+      "learning_rate": 1.1946046286109321e-05,
+      "loss": 0.3704,
       "step": 95015
     },
     {
-      "epoch": 3.34,
-      "learning_rate": 1.3085508012898659e-05,
-      "loss": 0.2709,
+      "epoch": 3.4245143619129994,
+      "grad_norm": 0.2723451852798462,
+      "learning_rate": 1.194355764861661e-05,
+      "loss": 0.3775,
       "step": 95020
     },
     {
-      "epoch": 3.34,
-      "learning_rate": 1.308300369023822e-05,
-      "loss": 0.2723,
+      "epoch": 3.424694561574224,
+      "grad_norm": 0.24041932821273804,
+      "learning_rate": 1.1941069189015406e-05,
+      "loss": 0.3849,
       "step": 95025
     },
     {
-      "epoch": 3.34,
-      "learning_rate": 1.3080499522312118e-05,
-      "loss": 0.2856,
+      "epoch": 3.424874761235449,
+      "grad_norm": 0.23057153820991516,
+      "learning_rate": 1.1938580907339606e-05,
+      "loss": 0.3625,
       "step": 95030
     },
     {
-      "epoch": 3.34,
-      "learning_rate": 1.3077995509152882e-05,
-      "loss": 0.2701,
+      "epoch": 3.4250549608966736,
+      "grad_norm": 0.23812946677207947,
+      "learning_rate": 1.1936092803623114e-05,
+      "loss": 0.3765,
       "step": 95035
     },
     {
-      "epoch": 3.34,
-      "learning_rate": 1.3075491650793017e-05,
-      "loss": 0.2622,
+      "epoch": 3.4252351605578983,
+      "grad_norm": 0.18412302434444427,
+      "learning_rate": 1.1933604877899825e-05,
+      "loss": 0.3616,
       "step": 95040
     },
     {
-      "epoch": 3.34,
-      "learning_rate": 1.3072987947265037e-05,
-      "loss": 0.2584,
+      "epoch": 3.4254153602191226,
+      "grad_norm": 0.20551741123199463,
+      "learning_rate": 1.1931117130203634e-05,
+      "loss": 0.3636,
       "step": 95045
     },
     {
-      "epoch": 3.34,
-      "learning_rate": 1.3070484398601434e-05,
-      "loss": 0.2455,
+      "epoch": 3.4255955598803474,
+      "grad_norm": 0.2384248822927475,
+      "learning_rate": 1.192862956056845e-05,
+      "loss": 0.399,
       "step": 95050
     },
     {
-      "epoch": 3.34,
-      "learning_rate": 1.3067981004834734e-05,
-      "loss": 0.2491,
+      "epoch": 3.425775759541572,
+      "grad_norm": 0.27354708313941956,
+      "learning_rate": 1.1926142169028156e-05,
+      "loss": 0.3748,
       "step": 95055
     },
     {
-      "epoch": 3.34,
-      "learning_rate": 1.3065477765997442e-05,
-      "loss": 0.2469,
+      "epoch": 3.425955959202797,
+      "grad_norm": 0.23982170224189758,
+      "learning_rate": 1.1923654955616645e-05,
+      "loss": 0.3492,
       "step": 95060
     },
     {
-      "epoch": 3.34,
-      "learning_rate": 1.3062974682122059e-05,
-      "loss": 0.2695,
+      "epoch": 3.426136158864021,
+      "grad_norm": 0.20187516510486603,
+      "learning_rate": 1.1921167920367805e-05,
+      "loss": 0.3737,
       "step": 95065
     },
     {
-      "epoch": 3.34,
-      "learning_rate": 1.3060471753241074e-05,
-      "loss": 0.2539,
+      "epoch": 3.426316358525246,
+      "grad_norm": 0.2167653739452362,
+      "learning_rate": 1.1918681063315507e-05,
+      "loss": 0.3887,
       "step": 95070
     },
     {
-      "epoch": 3.35,
-      "learning_rate": 1.3057968979387004e-05,
-      "loss": 0.2544,
+      "epoch": 3.4264965581864706,
+      "grad_norm": 0.2231149524450302,
+      "learning_rate": 1.1916194384493659e-05,
+      "loss": 0.3895,
       "step": 95075
     },
     {
-      "epoch": 3.35,
-      "learning_rate": 1.3055466360592337e-05,
-      "loss": 0.2343,
+      "epoch": 3.4266767578476953,
+      "grad_norm": 0.2766813337802887,
+      "learning_rate": 1.1913707883936136e-05,
+      "loss": 0.3737,
       "step": 95080
     },
     {
-      "epoch": 3.35,
-      "learning_rate": 1.3052963896889565e-05,
-      "loss": 0.2827,
+      "epoch": 3.42685695750892,
+      "grad_norm": 0.2061295360326767,
+      "learning_rate": 1.1911221561676795e-05,
+      "loss": 0.3716,
       "step": 95085
     },
     {
-      "epoch": 3.35,
-      "learning_rate": 1.3050461588311175e-05,
-      "loss": 0.2629,
+      "epoch": 3.4270371571701443,
+      "grad_norm": 0.21205422282218933,
+      "learning_rate": 1.1908735417749534e-05,
+      "loss": 0.3941,
       "step": 95090
     },
     {
-      "epoch": 3.35,
-      "learning_rate": 1.3047959434889679e-05,
-      "loss": 0.2746,
+      "epoch": 3.427217356831369,
+      "grad_norm": 0.2193949669599533,
+      "learning_rate": 1.1906249452188217e-05,
+      "loss": 0.3787,
       "step": 95095
     },
     {
-      "epoch": 3.35,
-      "learning_rate": 1.304545743665755e-05,
-      "loss": 0.2742,
+      "epoch": 3.427397556492594,
+      "grad_norm": 0.1897018849849701,
+      "learning_rate": 1.1903763665026719e-05,
+      "loss": 0.3418,
       "step": 95100
     },
     {
-      "epoch": 3.35,
-      "learning_rate": 1.3042955593647267e-05,
-      "loss": 0.2586,
+      "epoch": 3.4275777561538185,
+      "grad_norm": 0.2177032083272934,
+      "learning_rate": 1.1901278056298901e-05,
+      "loss": 0.3671,
       "step": 95105
     },
     {
-      "epoch": 3.35,
-      "learning_rate": 1.3040453905891328e-05,
-      "loss": 0.2966,
+      "epoch": 3.427757955815043,
+      "grad_norm": 0.29652655124664307,
+      "learning_rate": 1.1898792626038626e-05,
+      "loss": 0.366,
       "step": 95110
     },
     {
-      "epoch": 3.35,
-      "learning_rate": 1.3037952373422225e-05,
-      "loss": 0.2524,
+      "epoch": 3.4279381554762676,
+      "grad_norm": 0.20908688008785248,
+      "learning_rate": 1.1896307374279777e-05,
+      "loss": 0.3806,
       "step": 95115
     },
     {
-      "epoch": 3.35,
-      "learning_rate": 1.3035450996272422e-05,
-      "loss": 0.2349,
+      "epoch": 3.4281183551374923,
+      "grad_norm": 0.2735453248023987,
+      "learning_rate": 1.18938223010562e-05,
+      "loss": 0.3945,
       "step": 95120
     },
     {
-      "epoch": 3.35,
-      "learning_rate": 1.3032949774474407e-05,
-      "loss": 0.2827,
+      "epoch": 3.428298554798717,
+      "grad_norm": 0.23692834377288818,
+      "learning_rate": 1.1891337406401758e-05,
+      "loss": 0.3743,
       "step": 95125
     },
     {
-      "epoch": 3.35,
-      "learning_rate": 1.3030448708060639e-05,
-      "loss": 0.2449,
+      "epoch": 3.4284787544599418,
+      "grad_norm": 0.23678283393383026,
+      "learning_rate": 1.1888852690350305e-05,
+      "loss": 0.3876,
       "step": 95130
     },
     {
-      "epoch": 3.35,
-      "learning_rate": 1.3027947797063619e-05,
-      "loss": 0.2624,
+      "epoch": 3.428658954121166,
+      "grad_norm": 0.30739691853523254,
+      "learning_rate": 1.1886368152935698e-05,
+      "loss": 0.3843,
       "step": 95135
     },
     {
-      "epoch": 3.35,
-      "learning_rate": 1.3025447041515806e-05,
-      "loss": 0.2681,
+      "epoch": 3.428839153782391,
+      "grad_norm": 0.23699170351028442,
+      "learning_rate": 1.1883883794191774e-05,
+      "loss": 0.363,
       "step": 95140
     },
     {
-      "epoch": 3.35,
-      "learning_rate": 1.3022946441449663e-05,
-      "loss": 0.2653,
+      "epoch": 3.4290193534436155,
+      "grad_norm": 0.2635442912578583,
+      "learning_rate": 1.1881399614152406e-05,
+      "loss": 0.3764,
       "step": 95145
     },
     {
-      "epoch": 3.35,
-      "learning_rate": 1.3020445996897673e-05,
-      "loss": 0.2469,
+      "epoch": 3.4291995531048403,
+      "grad_norm": 0.23307177424430847,
+      "learning_rate": 1.1878915612851428e-05,
+      "loss": 0.3758,
       "step": 95150
     },
     {
-      "epoch": 3.35,
-      "learning_rate": 1.3017945707892305e-05,
-      "loss": 0.2505,
+      "epoch": 3.429379752766065,
+      "grad_norm": 0.208553284406662,
+      "learning_rate": 1.1876431790322682e-05,
+      "loss": 0.3888,
       "step": 95155
     },
     {
-      "epoch": 3.35,
-      "learning_rate": 1.3015445574465999e-05,
-      "loss": 0.249,
+      "epoch": 3.4295599524272893,
+      "grad_norm": 0.30267196893692017,
+      "learning_rate": 1.1873948146600012e-05,
+      "loss": 0.4137,
       "step": 95160
     },
     {
-      "epoch": 3.35,
-      "learning_rate": 1.3012945596651243e-05,
-      "loss": 0.2543,
+      "epoch": 3.429740152088514,
+      "grad_norm": 0.21039345860481262,
+      "learning_rate": 1.1871464681717249e-05,
+      "loss": 0.382,
       "step": 95165
     },
     {
-      "epoch": 3.35,
-      "learning_rate": 1.3010445774480484e-05,
-      "loss": 0.2588,
+      "epoch": 3.4299203517497387,
+      "grad_norm": 0.21296541392803192,
+      "learning_rate": 1.1868981395708256e-05,
+      "loss": 0.3729,
       "step": 95170
     },
     {
-      "epoch": 3.35,
-      "learning_rate": 1.3007946107986191e-05,
-      "loss": 0.2457,
+      "epoch": 3.4301005514109635,
+      "grad_norm": 0.2397422194480896,
+      "learning_rate": 1.1866498288606843e-05,
+      "loss": 0.3815,
       "step": 95175
     },
     {
-      "epoch": 3.35,
-      "learning_rate": 1.3005446597200815e-05,
-      "loss": 0.2692,
+      "epoch": 3.430280751072188,
+      "grad_norm": 0.259847491979599,
+      "learning_rate": 1.1864015360446834e-05,
+      "loss": 0.3923,
       "step": 95180
     },
     {
-      "epoch": 3.35,
-      "learning_rate": 1.3002947242156805e-05,
-      "loss": 0.2647,
+      "epoch": 3.4304609507334125,
+      "grad_norm": 0.23978331685066223,
+      "learning_rate": 1.1861532611262086e-05,
+      "loss": 0.3823,
       "step": 95185
     },
     {
-      "epoch": 3.35,
-      "learning_rate": 1.3000448042886626e-05,
-      "loss": 0.2713,
+      "epoch": 3.4306411503946372,
+      "grad_norm": 0.20742113888263702,
+      "learning_rate": 1.18590500410864e-05,
+      "loss": 0.3816,
       "step": 95190
     },
     {
-      "epoch": 3.35,
-      "learning_rate": 1.2997948999422719e-05,
-      "loss": 0.2444,
+      "epoch": 3.430821350055862,
+      "grad_norm": 0.24058598279953003,
+      "learning_rate": 1.1856567649953635e-05,
+      "loss": 0.3628,
       "step": 95195
     },
     {
-      "epoch": 3.35,
-      "learning_rate": 1.2995450111797535e-05,
-      "loss": 0.2554,
+      "epoch": 3.4310015497170867,
+      "grad_norm": 0.1972092241048813,
+      "learning_rate": 1.1854085437897578e-05,
+      "loss": 0.3614,
       "step": 95200
     },
     {
-      "epoch": 3.35,
-      "learning_rate": 1.299295138004351e-05,
-      "loss": 0.2439,
+      "epoch": 3.431181749378311,
+      "grad_norm": 0.22613808512687683,
+      "learning_rate": 1.1851603404952056e-05,
+      "loss": 0.3998,
       "step": 95205
     },
     {
-      "epoch": 3.35,
-      "learning_rate": 1.299045280419311e-05,
-      "loss": 0.2501,
+      "epoch": 3.4313619490395357,
+      "grad_norm": 0.24530638754367828,
+      "learning_rate": 1.1849121551150899e-05,
+      "loss": 0.3455,
       "step": 95210
     },
     {
-      "epoch": 3.35,
-      "learning_rate": 1.2987954384278756e-05,
-      "loss": 0.2727,
+      "epoch": 3.4315421487007605,
+      "grad_norm": 0.22673511505126953,
+      "learning_rate": 1.1846639876527915e-05,
+      "loss": 0.3772,
       "step": 95215
     },
     {
-      "epoch": 3.35,
-      "learning_rate": 1.2985456120332906e-05,
-      "loss": 0.2381,
+      "epoch": 3.431722348361985,
+      "grad_norm": 0.24466091394424438,
+      "learning_rate": 1.1844158381116916e-05,
+      "loss": 0.3992,
       "step": 95220
     },
     {
-      "epoch": 3.35,
-      "learning_rate": 1.2982958012387985e-05,
-      "loss": 0.2882,
+      "epoch": 3.43190254802321,
+      "grad_norm": 0.24469353258609772,
+      "learning_rate": 1.1841677064951714e-05,
+      "loss": 0.3863,
       "step": 95225
     },
     {
-      "epoch": 3.35,
-      "learning_rate": 1.298046006047644e-05,
-      "loss": 0.247,
+      "epoch": 3.432082747684434,
+      "grad_norm": 0.23059558868408203,
+      "learning_rate": 1.1839195928066102e-05,
+      "loss": 0.4255,
       "step": 95230
     },
     {
-      "epoch": 3.35,
-      "learning_rate": 1.29779622646307e-05,
-      "loss": 0.2704,
+      "epoch": 3.432262947345659,
+      "grad_norm": 0.18520836532115936,
+      "learning_rate": 1.183671497049391e-05,
+      "loss": 0.3682,
       "step": 95235
     },
     {
-      "epoch": 3.35,
-      "learning_rate": 1.29754646248832e-05,
-      "loss": 0.2382,
+      "epoch": 3.4324431470068837,
+      "grad_norm": 0.2409636229276657,
+      "learning_rate": 1.1834234192268926e-05,
+      "loss": 0.3583,
       "step": 95240
     },
     {
-      "epoch": 3.35,
-      "learning_rate": 1.2972967141266357e-05,
-      "loss": 0.2845,
+      "epoch": 3.4326233466681084,
+      "grad_norm": 0.1873437464237213,
+      "learning_rate": 1.1831753593424951e-05,
+      "loss": 0.3832,
       "step": 95245
     },
     {
-      "epoch": 3.35,
-      "learning_rate": 1.2970469813812619e-05,
-      "loss": 0.2614,
+      "epoch": 3.4328035463293327,
+      "grad_norm": 0.27376478910446167,
+      "learning_rate": 1.1829273173995787e-05,
+      "loss": 0.3519,
       "step": 95250
     },
     {
-      "epoch": 3.35,
-      "learning_rate": 1.29679726425544e-05,
-      "loss": 0.2622,
+      "epoch": 3.4329837459905574,
+      "grad_norm": 0.24432578682899475,
+      "learning_rate": 1.1826792934015216e-05,
+      "loss": 0.4051,
       "step": 95255
     },
     {
-      "epoch": 3.35,
-      "learning_rate": 1.2965475627524122e-05,
-      "loss": 0.2555,
+      "epoch": 3.433163945651782,
+      "grad_norm": 0.3113139867782593,
+      "learning_rate": 1.182431287351705e-05,
+      "loss": 0.41,
       "step": 95260
     },
     {
-      "epoch": 3.35,
-      "learning_rate": 1.2962978768754217e-05,
-      "loss": 0.2572,
+      "epoch": 3.433344145313007,
+      "grad_norm": 0.24303504824638367,
+      "learning_rate": 1.1821832992535078e-05,
+      "loss": 0.3869,
       "step": 95265
     },
     {
-      "epoch": 3.35,
-      "learning_rate": 1.296048206627709e-05,
-      "loss": 0.2654,
+      "epoch": 3.4335243449742316,
+      "grad_norm": 0.19319789111614227,
+      "learning_rate": 1.1819353291103064e-05,
+      "loss": 0.3631,
       "step": 95270
     },
     {
-      "epoch": 3.35,
-      "learning_rate": 1.2957985520125182e-05,
-      "loss": 0.2568,
+      "epoch": 3.433704544635456,
+      "grad_norm": 0.24024613201618195,
+      "learning_rate": 1.1816873769254816e-05,
+      "loss": 0.3829,
       "step": 95275
     },
     {
-      "epoch": 3.35,
-      "learning_rate": 1.2955489130330884e-05,
-      "loss": 0.2593,
+      "epoch": 3.4338847442966807,
+      "grad_norm": 0.23290759325027466,
+      "learning_rate": 1.1814394427024112e-05,
+      "loss": 0.3731,
       "step": 95280
     },
     {
-      "epoch": 3.35,
-      "learning_rate": 1.2952992896926636e-05,
-      "loss": 0.2639,
+      "epoch": 3.4340649439579054,
+      "grad_norm": 0.21962322294712067,
+      "learning_rate": 1.1811915264444732e-05,
+      "loss": 0.3682,
       "step": 95285
     },
     {
-      "epoch": 3.35,
-      "learning_rate": 1.295049681994483e-05,
-      "loss": 0.2456,
+      "epoch": 3.43424514361913,
+      "grad_norm": 0.2735930383205414,
+      "learning_rate": 1.1809436281550451e-05,
+      "loss": 0.4133,
       "step": 95290
     },
     {
-      "epoch": 3.35,
-      "learning_rate": 1.2948000899417887e-05,
-      "loss": 0.2499,
+      "epoch": 3.4344253432803544,
+      "grad_norm": 0.2372862547636032,
+      "learning_rate": 1.1806957478375038e-05,
+      "loss": 0.3902,
       "step": 95295
     },
     {
-      "epoch": 3.35,
-      "learning_rate": 1.2945505135378196e-05,
-      "loss": 0.2455,
+      "epoch": 3.434605542941579,
+      "grad_norm": 0.24612626433372498,
+      "learning_rate": 1.1804478854952287e-05,
+      "loss": 0.4095,
       "step": 95300
     },
     {
-      "epoch": 3.35,
-      "learning_rate": 1.2943009527858187e-05,
-      "loss": 0.2623,
+      "epoch": 3.434785742602804,
+      "grad_norm": 0.21510684490203857,
+      "learning_rate": 1.1802000411315953e-05,
+      "loss": 0.3714,
       "step": 95305
     },
     {
-      "epoch": 3.35,
-      "learning_rate": 1.2940514076890257e-05,
-      "loss": 0.2515,
+      "epoch": 3.4349659422640286,
+      "grad_norm": 0.23222926259040833,
+      "learning_rate": 1.1799522147499812e-05,
+      "loss": 0.3823,
       "step": 95310
     },
     {
-      "epoch": 3.35,
-      "learning_rate": 1.29380187825068e-05,
-      "loss": 0.2423,
+      "epoch": 3.4351461419252534,
+      "grad_norm": 0.303889662027359,
+      "learning_rate": 1.1797044063537625e-05,
+      "loss": 0.374,
       "step": 95315
     },
     {
-      "epoch": 3.35,
-      "learning_rate": 1.2935523644740213e-05,
-      "loss": 0.256,
+      "epoch": 3.4353263415864776,
+      "grad_norm": 0.22583414614200592,
+      "learning_rate": 1.1794566159463155e-05,
+      "loss": 0.3554,
       "step": 95320
     },
     {
-      "epoch": 3.35,
-      "learning_rate": 1.2933028663622909e-05,
-      "loss": 0.2667,
+      "epoch": 3.4355065412477024,
+      "grad_norm": 0.21528246998786926,
+      "learning_rate": 1.1792088435310156e-05,
+      "loss": 0.4096,
       "step": 95325
     },
     {
-      "epoch": 3.35,
-      "learning_rate": 1.2930533839187265e-05,
-      "loss": 0.271,
+      "epoch": 3.435686740908927,
+      "grad_norm": 0.237309068441391,
+      "learning_rate": 1.1789610891112402e-05,
+      "loss": 0.3965,
       "step": 95330
     },
     {
-      "epoch": 3.35,
-      "learning_rate": 1.2928039171465697e-05,
-      "loss": 0.2523,
+      "epoch": 3.435866940570152,
+      "grad_norm": 0.22394055128097534,
+      "learning_rate": 1.1787133526903646e-05,
+      "loss": 0.3985,
       "step": 95335
     },
     {
-      "epoch": 3.35,
-      "learning_rate": 1.2925544660490573e-05,
-      "loss": 0.2573,
+      "epoch": 3.436047140231376,
+      "grad_norm": 0.23511168360710144,
+      "learning_rate": 1.1784656342717634e-05,
+      "loss": 0.4072,
       "step": 95340
     },
     {
-      "epoch": 3.35,
-      "learning_rate": 1.2923050306294305e-05,
-      "loss": 0.262,
+      "epoch": 3.436227339892601,
+      "grad_norm": 0.2214852273464203,
+      "learning_rate": 1.1782179338588124e-05,
+      "loss": 0.3582,
       "step": 95345
     },
     {
-      "epoch": 3.35,
-      "learning_rate": 1.2920556108909271e-05,
-      "loss": 0.2591,
+      "epoch": 3.4364075395538256,
+      "grad_norm": 0.2489766925573349,
+      "learning_rate": 1.1779702514548848e-05,
+      "loss": 0.361,
       "step": 95350
     },
     {
-      "epoch": 3.35,
-      "learning_rate": 1.2918062068367853e-05,
-      "loss": 0.2597,
+      "epoch": 3.4365877392150503,
+      "grad_norm": 0.23350292444229126,
+      "learning_rate": 1.1777225870633579e-05,
+      "loss": 0.3675,
       "step": 95355
     },
     {
-      "epoch": 3.36,
-      "learning_rate": 1.2915568184702425e-05,
-      "loss": 0.2418,
+      "epoch": 3.436767938876275,
+      "grad_norm": 0.23286855220794678,
+      "learning_rate": 1.1774749406876057e-05,
+      "loss": 0.4119,
       "step": 95360
     },
     {
-      "epoch": 3.36,
-      "learning_rate": 1.2913074457945396e-05,
-      "loss": 0.2486,
+      "epoch": 3.4369481385374994,
+      "grad_norm": 0.2144162505865097,
+      "learning_rate": 1.1772273123309993e-05,
+      "loss": 0.3873,
       "step": 95365
     },
     {
-      "epoch": 3.36,
-      "learning_rate": 1.2910580888129126e-05,
-      "loss": 0.2508,
+      "epoch": 3.437128338198724,
+      "grad_norm": 0.24179521203041077,
+      "learning_rate": 1.1769797019969156e-05,
+      "loss": 0.3731,
       "step": 95370
     },
     {
-      "epoch": 3.36,
-      "learning_rate": 1.2908087475285984e-05,
-      "loss": 0.2783,
+      "epoch": 3.437308537859949,
+      "grad_norm": 0.22237004339694977,
+      "learning_rate": 1.1767321096887265e-05,
+      "loss": 0.3584,
       "step": 95375
     },
     {
-      "epoch": 3.36,
-      "learning_rate": 1.2905594219448368e-05,
-      "loss": 0.2565,
+      "epoch": 3.4374887375211736,
+      "grad_norm": 0.23531267046928406,
+      "learning_rate": 1.1764845354098081e-05,
+      "loss": 0.4297,
       "step": 95380
     },
     {
-      "epoch": 3.36,
-      "learning_rate": 1.2903101120648637e-05,
-      "loss": 0.2734,
+      "epoch": 3.437668937182398,
+      "grad_norm": 0.2168639749288559,
+      "learning_rate": 1.1762369791635305e-05,
+      "loss": 0.4128,
       "step": 95385
     },
     {
-      "epoch": 3.36,
-      "learning_rate": 1.290060817891917e-05,
-      "loss": 0.2653,
+      "epoch": 3.4378491368436226,
+      "grad_norm": 0.24747677147388458,
+      "learning_rate": 1.175989440953267e-05,
+      "loss": 0.3953,
       "step": 95390
     },
     {
-      "epoch": 3.36,
-      "learning_rate": 1.2898115394292321e-05,
-      "loss": 0.2656,
+      "epoch": 3.4380293365048473,
+      "grad_norm": 0.2616941034793854,
+      "learning_rate": 1.1757419207823918e-05,
+      "loss": 0.3724,
       "step": 95395
     },
     {
-      "epoch": 3.36,
-      "learning_rate": 1.289562276680048e-05,
-      "loss": 0.2528,
+      "epoch": 3.438209536166072,
+      "grad_norm": 0.21548521518707275,
+      "learning_rate": 1.1754944186542768e-05,
+      "loss": 0.3945,
       "step": 95400
     },
     {
-      "epoch": 3.36,
-      "learning_rate": 1.2893130296476e-05,
-      "loss": 0.2752,
+      "epoch": 3.438389735827297,
+      "grad_norm": 0.20728285610675812,
+      "learning_rate": 1.1752469345722938e-05,
+      "loss": 0.4152,
       "step": 95405
     },
     {
-      "epoch": 3.36,
-      "learning_rate": 1.2890637983351244e-05,
-      "loss": 0.2442,
+      "epoch": 3.4385699354885215,
+      "grad_norm": 0.24086813628673553,
+      "learning_rate": 1.1749994685398149e-05,
+      "loss": 0.3667,
       "step": 95410
     },
     {
-      "epoch": 3.36,
-      "learning_rate": 1.2888145827458564e-05,
-      "loss": 0.2665,
+      "epoch": 3.438750135149746,
+      "grad_norm": 0.21392937004566193,
+      "learning_rate": 1.1747520205602108e-05,
+      "loss": 0.3966,
       "step": 95415
     },
     {
-      "epoch": 3.36,
-      "learning_rate": 1.288565382883034e-05,
-      "loss": 0.2613,
+      "epoch": 3.4389303348109705,
+      "grad_norm": 0.21268320083618164,
+      "learning_rate": 1.1745045906368549e-05,
+      "loss": 0.3754,
       "step": 95420
     },
     {
-      "epoch": 3.36,
-      "learning_rate": 1.2883161987498915e-05,
-      "loss": 0.2647,
+      "epoch": 3.4391105344721953,
+      "grad_norm": 0.2621963918209076,
+      "learning_rate": 1.1742571787731172e-05,
+      "loss": 0.4324,
       "step": 95425
     },
     {
-      "epoch": 3.36,
-      "learning_rate": 1.288067030349664e-05,
-      "loss": 0.2593,
+      "epoch": 3.43929073413342,
+      "grad_norm": 0.28907299041748047,
+      "learning_rate": 1.1740097849723686e-05,
+      "loss": 0.3717,
       "step": 95430
     },
     {
-      "epoch": 3.36,
-      "learning_rate": 1.287817877685588e-05,
-      "loss": 0.2676,
+      "epoch": 3.4394709337946443,
+      "grad_norm": 0.24965962767601013,
+      "learning_rate": 1.17376240923798e-05,
+      "loss": 0.3629,
       "step": 95435
     },
     {
-      "epoch": 3.36,
-      "learning_rate": 1.2875687407608972e-05,
-      "loss": 0.2488,
+      "epoch": 3.439651133455869,
+      "grad_norm": 0.27768054604530334,
+      "learning_rate": 1.1735150515733209e-05,
+      "loss": 0.3604,
       "step": 95440
     },
     {
-      "epoch": 3.36,
-      "learning_rate": 1.2873196195788279e-05,
-      "loss": 0.263,
+      "epoch": 3.4398313331170938,
+      "grad_norm": 0.25251927971839905,
+      "learning_rate": 1.1732677119817634e-05,
+      "loss": 0.3643,
       "step": 95445
     },
     {
-      "epoch": 3.36,
-      "learning_rate": 1.2870705141426146e-05,
-      "loss": 0.2592,
+      "epoch": 3.4400115327783185,
+      "grad_norm": 0.2550899386405945,
+      "learning_rate": 1.1730203904666773e-05,
+      "loss": 0.3879,
       "step": 95450
     },
     {
-      "epoch": 3.36,
-      "learning_rate": 1.2868214244554902e-05,
-      "loss": 0.2432,
+      "epoch": 3.4401917324395432,
+      "grad_norm": 0.2622126638889313,
+      "learning_rate": 1.1727730870314298e-05,
+      "loss": 0.3498,
       "step": 95455
     },
     {
-      "epoch": 3.36,
-      "learning_rate": 1.2865723505206912e-05,
-      "loss": 0.2637,
+      "epoch": 3.4403719321007675,
+      "grad_norm": 0.2463812530040741,
+      "learning_rate": 1.1725258016793927e-05,
+      "loss": 0.413,
       "step": 95460
     },
     {
-      "epoch": 3.36,
-      "learning_rate": 1.2863232923414504e-05,
-      "loss": 0.2573,
+      "epoch": 3.4405521317619923,
+      "grad_norm": 0.2597421109676361,
+      "learning_rate": 1.1722785344139347e-05,
+      "loss": 0.3833,
       "step": 95465
     },
     {
-      "epoch": 3.36,
-      "learning_rate": 1.2860742499210016e-05,
-      "loss": 0.2492,
+      "epoch": 3.440732331423217,
+      "grad_norm": 0.20518064498901367,
+      "learning_rate": 1.1720312852384239e-05,
+      "loss": 0.3924,
       "step": 95470
     },
     {
-      "epoch": 3.36,
-      "learning_rate": 1.2858252232625778e-05,
-      "loss": 0.2399,
+      "epoch": 3.4409125310844417,
+      "grad_norm": 0.20788519084453583,
+      "learning_rate": 1.1717840541562312e-05,
+      "loss": 0.3612,
       "step": 95475
     },
     {
-      "epoch": 3.36,
-      "learning_rate": 1.2855762123694145e-05,
-      "loss": 0.2297,
+      "epoch": 3.441092730745666,
+      "grad_norm": 0.21892179548740387,
+      "learning_rate": 1.171536841170722e-05,
+      "loss": 0.3725,
       "step": 95480
     },
     {
-      "epoch": 3.36,
-      "learning_rate": 1.2853272172447437e-05,
-      "loss": 0.2794,
+      "epoch": 3.4412729304068908,
+      "grad_norm": 0.23169714212417603,
+      "learning_rate": 1.1712896462852668e-05,
+      "loss": 0.3949,
       "step": 95485
     },
     {
-      "epoch": 3.36,
-      "learning_rate": 1.2850782378917975e-05,
-      "loss": 0.2753,
+      "epoch": 3.4414531300681155,
+      "grad_norm": 0.24189196527004242,
+      "learning_rate": 1.1710424695032332e-05,
+      "loss": 0.368,
       "step": 95490
     },
     {
-      "epoch": 3.36,
-      "learning_rate": 1.2848292743138097e-05,
-      "loss": 0.2583,
+      "epoch": 3.44163332972934,
+      "grad_norm": 0.21008506417274475,
+      "learning_rate": 1.1707953108279888e-05,
+      "loss": 0.4096,
       "step": 95495
     },
     {
-      "epoch": 3.36,
-      "learning_rate": 1.2845803265140138e-05,
-      "loss": 0.2693,
+      "epoch": 3.441813529390565,
+      "grad_norm": 0.23108026385307312,
+      "learning_rate": 1.1705481702629007e-05,
+      "loss": 0.3929,
       "step": 95500
     },
     {
-      "epoch": 3.36,
-      "eval_loss": 0.2542358934879303,
-      "eval_runtime": 10.5386,
-      "eval_samples_per_second": 9.489,
-      "eval_steps_per_second": 9.489,
+      "epoch": 3.441813529390565,
+      "eval_loss": 0.4303448796272278,
+      "eval_runtime": 3.5356,
+      "eval_samples_per_second": 28.284,
+      "eval_steps_per_second": 7.071,
       "step": 95500
     },
     {
-      "epoch": 3.36,
-      "learning_rate": 1.2843313944956414e-05,
-      "loss": 0.2593,
+      "epoch": 3.4419937290517892,
+      "grad_norm": 0.1974146068096161,
+      "learning_rate": 1.1703010478113363e-05,
+      "loss": 0.3409,
       "step": 95505
     },
     {
-      "epoch": 3.36,
-      "learning_rate": 1.2840824782619237e-05,
-      "loss": 0.2635,
+      "epoch": 3.442173928713014,
+      "grad_norm": 0.2331085056066513,
+      "learning_rate": 1.1700539434766619e-05,
+      "loss": 0.3589,
       "step": 95510
     },
     {
-      "epoch": 3.36,
-      "learning_rate": 1.2838335778160948e-05,
-      "loss": 0.2435,
+      "epoch": 3.4423541283742387,
+      "grad_norm": 0.23634552955627441,
+      "learning_rate": 1.1698068572622458e-05,
+      "loss": 0.3905,
       "step": 95515
     },
     {
-      "epoch": 3.36,
-      "learning_rate": 1.2835846931613853e-05,
-      "loss": 0.2347,
+      "epoch": 3.4425343280354634,
+      "grad_norm": 0.3103722929954529,
+      "learning_rate": 1.1695597891714539e-05,
+      "loss": 0.3924,
       "step": 95520
     },
     {
-      "epoch": 3.36,
-      "learning_rate": 1.2833358243010268e-05,
-      "loss": 0.2579,
+      "epoch": 3.4427145276966877,
+      "grad_norm": 0.2085963785648346,
+      "learning_rate": 1.1693127392076522e-05,
+      "loss": 0.3817,
       "step": 95525
     },
     {
-      "epoch": 3.36,
-      "learning_rate": 1.2830869712382504e-05,
-      "loss": 0.2742,
+      "epoch": 3.4428947273579125,
+      "grad_norm": 0.2746676504611969,
+      "learning_rate": 1.169065707374207e-05,
+      "loss": 0.4077,
       "step": 95530
     },
     {
-      "epoch": 3.36,
-      "learning_rate": 1.282838133976288e-05,
-      "loss": 0.261,
+      "epoch": 3.443074927019137,
+      "grad_norm": 0.2430097460746765,
+      "learning_rate": 1.1688186936744827e-05,
+      "loss": 0.3698,
       "step": 95535
     },
     {
-      "epoch": 3.36,
-      "learning_rate": 1.2825893125183709e-05,
-      "loss": 0.2383,
+      "epoch": 3.443255126680362,
+      "grad_norm": 0.2820988595485687,
+      "learning_rate": 1.1685716981118469e-05,
+      "loss": 0.3811,
       "step": 95540
     },
     {
-      "epoch": 3.36,
-      "learning_rate": 1.2823405068677277e-05,
-      "loss": 0.2771,
+      "epoch": 3.4434353263415867,
+      "grad_norm": 0.2863827347755432,
+      "learning_rate": 1.168324720689665e-05,
+      "loss": 0.4197,
       "step": 95545
     },
     {
-      "epoch": 3.36,
-      "learning_rate": 1.2820917170275909e-05,
-      "loss": 0.2735,
+      "epoch": 3.443615526002811,
+      "grad_norm": 0.2102956622838974,
+      "learning_rate": 1.1680777614112992e-05,
+      "loss": 0.3898,
       "step": 95550
     },
     {
-      "epoch": 3.36,
-      "learning_rate": 1.2818429430011914e-05,
-      "loss": 0.2574,
+      "epoch": 3.4437957256640357,
+      "grad_norm": 0.20017428696155548,
+      "learning_rate": 1.1678308202801166e-05,
+      "loss": 0.3685,
       "step": 95555
     },
     {
-      "epoch": 3.36,
-      "learning_rate": 1.2815941847917584e-05,
-      "loss": 0.2508,
+      "epoch": 3.4439759253252604,
+      "grad_norm": 0.2171812206506729,
+      "learning_rate": 1.1675838972994805e-05,
+      "loss": 0.4052,
       "step": 95560
     },
     {
-      "epoch": 3.36,
-      "learning_rate": 1.281345442402522e-05,
-      "loss": 0.2654,
+      "epoch": 3.444156124986485,
+      "grad_norm": 0.20312188565731049,
+      "learning_rate": 1.1673369924727576e-05,
+      "loss": 0.3949,
       "step": 95565
     },
     {
-      "epoch": 3.36,
-      "learning_rate": 1.281096715836711e-05,
-      "loss": 0.254,
+      "epoch": 3.4443363246477094,
+      "grad_norm": 0.2849917709827423,
+      "learning_rate": 1.1670901058033093e-05,
+      "loss": 0.3912,
       "step": 95570
     },
     {
-      "epoch": 3.36,
-      "learning_rate": 1.2808480050975569e-05,
-      "loss": 0.262,
+      "epoch": 3.444516524308934,
+      "grad_norm": 0.2215384542942047,
+      "learning_rate": 1.1668432372944992e-05,
+      "loss": 0.3403,
       "step": 95575
     },
     {
-      "epoch": 3.36,
-      "learning_rate": 1.280599310188288e-05,
-      "loss": 0.2479,
+      "epoch": 3.444696723970159,
+      "grad_norm": 0.26748472452163696,
+      "learning_rate": 1.1665963869496932e-05,
+      "loss": 0.3769,
       "step": 95580
     },
     {
-      "epoch": 3.36,
-      "learning_rate": 1.2803506311121332e-05,
-      "loss": 0.2974,
+      "epoch": 3.4448769236313836,
+      "grad_norm": 0.20163197815418243,
+      "learning_rate": 1.1663495547722528e-05,
+      "loss": 0.3837,
       "step": 95585
     },
     {
-      "epoch": 3.36,
-      "learning_rate": 1.2801019678723208e-05,
-      "loss": 0.2636,
+      "epoch": 3.4450571232926084,
+      "grad_norm": 0.23869015276432037,
+      "learning_rate": 1.1661027407655417e-05,
+      "loss": 0.3995,
       "step": 95590
     },
     {
-      "epoch": 3.36,
-      "learning_rate": 1.2798533204720814e-05,
-      "loss": 0.2263,
+      "epoch": 3.4452373229538327,
+      "grad_norm": 0.25916892290115356,
+      "learning_rate": 1.1658559449329227e-05,
+      "loss": 0.3736,
       "step": 95595
     },
     {
-      "epoch": 3.36,
-      "learning_rate": 1.2796046889146417e-05,
-      "loss": 0.2639,
+      "epoch": 3.4454175226150574,
+      "grad_norm": 0.2509896755218506,
+      "learning_rate": 1.1656091672777569e-05,
+      "loss": 0.3497,
       "step": 95600
     },
     {
-      "epoch": 3.36,
-      "learning_rate": 1.2793560732032317e-05,
-      "loss": 0.254,
+      "epoch": 3.445597722276282,
+      "grad_norm": 0.2731640636920929,
+      "learning_rate": 1.165362407803409e-05,
+      "loss": 0.3667,
       "step": 95605
     },
     {
-      "epoch": 3.36,
-      "learning_rate": 1.2791074733410776e-05,
-      "loss": 0.2232,
+      "epoch": 3.445777921937507,
+      "grad_norm": 0.23170186579227448,
+      "learning_rate": 1.1651156665132396e-05,
+      "loss": 0.3817,
       "step": 95610
     },
     {
-      "epoch": 3.36,
-      "learning_rate": 1.2788588893314096e-05,
-      "loss": 0.2799,
+      "epoch": 3.445958121598731,
+      "grad_norm": 0.21757963299751282,
+      "learning_rate": 1.164868943410611e-05,
+      "loss": 0.3697,
       "step": 95615
     },
     {
-      "epoch": 3.36,
-      "learning_rate": 1.2786103211774538e-05,
-      "loss": 0.2552,
+      "epoch": 3.446138321259956,
+      "grad_norm": 0.2630380690097809,
+      "learning_rate": 1.1646222384988847e-05,
+      "loss": 0.3495,
       "step": 95620
     },
     {
-      "epoch": 3.36,
-      "learning_rate": 1.2783617688824373e-05,
-      "loss": 0.2556,
+      "epoch": 3.4463185209211806,
+      "grad_norm": 0.23066528141498566,
+      "learning_rate": 1.164375551781421e-05,
+      "loss": 0.3468,
       "step": 95625
     },
     {
-      "epoch": 3.36,
-      "learning_rate": 1.2781132324495892e-05,
-      "loss": 0.2572,
+      "epoch": 3.4464987205824054,
+      "grad_norm": 0.27254176139831543,
+      "learning_rate": 1.1641288832615823e-05,
+      "loss": 0.4197,
       "step": 95630
     },
     {
-      "epoch": 3.36,
-      "learning_rate": 1.2778647118821352e-05,
-      "loss": 0.2541,
+      "epoch": 3.44667892024363,
+      "grad_norm": 0.2096617966890335,
+      "learning_rate": 1.1638822329427293e-05,
+      "loss": 0.3646,
       "step": 95635
     },
     {
-      "epoch": 3.36,
-      "learning_rate": 1.2776162071833028e-05,
-      "loss": 0.2767,
+      "epoch": 3.4468591199048544,
+      "grad_norm": 0.20088768005371094,
+      "learning_rate": 1.1636356008282221e-05,
+      "loss": 0.3663,
       "step": 95640
     },
     {
-      "epoch": 3.37,
-      "learning_rate": 1.2773677183563171e-05,
-      "loss": 0.2583,
+      "epoch": 3.447039319566079,
+      "grad_norm": 0.2403293251991272,
+      "learning_rate": 1.163388986921421e-05,
+      "loss": 0.3651,
       "step": 95645
     },
     {
-      "epoch": 3.37,
-      "learning_rate": 1.2771192454044073e-05,
-      "loss": 0.2749,
+      "epoch": 3.447219519227304,
+      "grad_norm": 0.23152384161949158,
+      "learning_rate": 1.163142391225686e-05,
+      "loss": 0.3991,
       "step": 95650
     },
     {
-      "epoch": 3.37,
-      "learning_rate": 1.2768707883307968e-05,
-      "loss": 0.26,
+      "epoch": 3.4473997188885286,
+      "grad_norm": 0.29287877678871155,
+      "learning_rate": 1.1628958137443766e-05,
+      "loss": 0.4082,
       "step": 95655
     },
     {
-      "epoch": 3.37,
-      "learning_rate": 1.276622347138714e-05,
-      "loss": 0.2622,
+      "epoch": 3.4475799185497533,
+      "grad_norm": 0.2720036208629608,
+      "learning_rate": 1.1626492544808545e-05,
+      "loss": 0.3914,
       "step": 95660
     },
     {
-      "epoch": 3.37,
-      "learning_rate": 1.276373921831383e-05,
-      "loss": 0.2501,
+      "epoch": 3.4477601182109776,
+      "grad_norm": 0.24344083666801453,
+      "learning_rate": 1.1624027134384757e-05,
+      "loss": 0.3794,
       "step": 95665
     },
     {
-      "epoch": 3.37,
-      "learning_rate": 1.2761255124120313e-05,
-      "loss": 0.2671,
+      "epoch": 3.4479403178722023,
+      "grad_norm": 0.22803980112075806,
+      "learning_rate": 1.1621561906206019e-05,
+      "loss": 0.3875,
       "step": 95670
     },
     {
-      "epoch": 3.37,
-      "learning_rate": 1.2758771188838831e-05,
-      "loss": 0.2439,
+      "epoch": 3.448120517533427,
+      "grad_norm": 0.23738615214824677,
+      "learning_rate": 1.1619096860305905e-05,
+      "loss": 0.4092,
       "step": 95675
     },
     {
-      "epoch": 3.37,
-      "learning_rate": 1.2756287412501638e-05,
-      "loss": 0.2651,
+      "epoch": 3.448300717194652,
+      "grad_norm": 0.20066243410110474,
+      "learning_rate": 1.161663199671801e-05,
+      "loss": 0.3773,
       "step": 95680
     },
     {
-      "epoch": 3.37,
-      "learning_rate": 1.2753803795140973e-05,
-      "loss": 0.2558,
+      "epoch": 3.4484809168558765,
+      "grad_norm": 0.3553462326526642,
+      "learning_rate": 1.1614167315475913e-05,
+      "loss": 0.3927,
       "step": 95685
     },
     {
-      "epoch": 3.37,
-      "learning_rate": 1.2751320336789108e-05,
-      "loss": 0.2531,
+      "epoch": 3.448661116517101,
+      "grad_norm": 0.23177509009838104,
+      "learning_rate": 1.1611702816613195e-05,
+      "loss": 0.4163,
       "step": 95690
     },
     {
-      "epoch": 3.37,
-      "learning_rate": 1.2748837037478272e-05,
-      "loss": 0.2478,
+      "epoch": 3.4488413161783256,
+      "grad_norm": 0.21542638540267944,
+      "learning_rate": 1.1609238500163424e-05,
+      "loss": 0.4101,
       "step": 95695
     },
     {
-      "epoch": 3.37,
-      "learning_rate": 1.2746353897240715e-05,
-      "loss": 0.2585,
+      "epoch": 3.4490215158395503,
+      "grad_norm": 0.2028144747018814,
+      "learning_rate": 1.1606774366160194e-05,
+      "loss": 0.3794,
       "step": 95700
     },
     {
-      "epoch": 3.37,
-      "learning_rate": 1.2743870916108668e-05,
-      "loss": 0.2654,
+      "epoch": 3.449201715500775,
+      "grad_norm": 0.23155230283737183,
+      "learning_rate": 1.160431041463707e-05,
+      "loss": 0.3431,
       "step": 95705
     },
     {
-      "epoch": 3.37,
-      "learning_rate": 1.2741388094114382e-05,
-      "loss": 0.2564,
+      "epoch": 3.4493819151619993,
+      "grad_norm": 0.2762317359447479,
+      "learning_rate": 1.1601846645627625e-05,
+      "loss": 0.3731,
       "step": 95710
     },
     {
-      "epoch": 3.37,
-      "learning_rate": 1.27389054312901e-05,
-      "loss": 0.2517,
+      "epoch": 3.449562114823224,
+      "grad_norm": 0.2540960907936096,
+      "learning_rate": 1.1599383059165425e-05,
+      "loss": 0.3865,
       "step": 95715
     },
     {
-      "epoch": 3.37,
-      "learning_rate": 1.2736422927668052e-05,
-      "loss": 0.2956,
+      "epoch": 3.449742314484449,
+      "grad_norm": 0.19944801926612854,
+      "learning_rate": 1.1596919655284027e-05,
+      "loss": 0.3779,
       "step": 95720
     },
     {
-      "epoch": 3.37,
-      "learning_rate": 1.2733940583280457e-05,
-      "loss": 0.2528,
+      "epoch": 3.4499225141456735,
+      "grad_norm": 0.23415467143058777,
+      "learning_rate": 1.1594456434017013e-05,
+      "loss": 0.3739,
       "step": 95725
     },
     {
-      "epoch": 3.37,
-      "learning_rate": 1.2731458398159573e-05,
-      "loss": 0.2563,
+      "epoch": 3.4501027138068983,
+      "grad_norm": 0.21086935698986053,
+      "learning_rate": 1.1591993395397944e-05,
+      "loss": 0.3912,
       "step": 95730
     },
     {
-      "epoch": 3.37,
-      "learning_rate": 1.2728976372337617e-05,
-      "loss": 0.2618,
+      "epoch": 3.4502829134681225,
+      "grad_norm": 0.23152147233486176,
+      "learning_rate": 1.158953053946035e-05,
+      "loss": 0.4073,
       "step": 95735
     },
     {
-      "epoch": 3.37,
-      "learning_rate": 1.2726494505846803e-05,
-      "loss": 0.2516,
+      "epoch": 3.4504631131293473,
+      "grad_norm": 0.29169178009033203,
+      "learning_rate": 1.1587067866237816e-05,
+      "loss": 0.41,
       "step": 95740
     },
     {
-      "epoch": 3.37,
-      "learning_rate": 1.272401279871938e-05,
-      "loss": 0.2465,
+      "epoch": 3.450643312790572,
+      "grad_norm": 0.24295637011528015,
+      "learning_rate": 1.1584605375763876e-05,
+      "loss": 0.3491,
       "step": 95745
     },
     {
-      "epoch": 3.37,
-      "learning_rate": 1.2721531250987565e-05,
-      "loss": 0.2483,
+      "epoch": 3.4508235124517967,
+      "grad_norm": 0.19027990102767944,
+      "learning_rate": 1.158214306807211e-05,
+      "loss": 0.385,
       "step": 95750
     },
     {
-      "epoch": 3.37,
-      "learning_rate": 1.2719049862683572e-05,
-      "loss": 0.2505,
+      "epoch": 3.451003712113021,
+      "grad_norm": 0.21917863190174103,
+      "learning_rate": 1.1579680943196036e-05,
+      "loss": 0.3643,
       "step": 95755
     },
     {
-      "epoch": 3.37,
-      "learning_rate": 1.2716568633839612e-05,
-      "loss": 0.2483,
+      "epoch": 3.4511839117742458,
+      "grad_norm": 0.22133587300777435,
+      "learning_rate": 1.1577219001169204e-05,
+      "loss": 0.3636,
       "step": 95760
     },
     {
-      "epoch": 3.37,
-      "learning_rate": 1.2714087564487925e-05,
-      "loss": 0.2572,
+      "epoch": 3.4513641114354705,
+      "grad_norm": 0.22902527451515198,
+      "learning_rate": 1.1574757242025169e-05,
+      "loss": 0.3693,
       "step": 95765
     },
     {
-      "epoch": 3.37,
-      "learning_rate": 1.2711606654660701e-05,
-      "loss": 0.2597,
+      "epoch": 3.4515443110966952,
+      "grad_norm": 0.21824954450130463,
+      "learning_rate": 1.1572295665797472e-05,
+      "loss": 0.3672,
       "step": 95770
     },
     {
-      "epoch": 3.37,
-      "learning_rate": 1.270912590439018e-05,
-      "loss": 0.2515,
+      "epoch": 3.45172451075792,
+      "grad_norm": 0.29460370540618896,
+      "learning_rate": 1.1569834272519644e-05,
+      "loss": 0.413,
       "step": 95775
     },
     {
-      "epoch": 3.37,
-      "learning_rate": 1.2706645313708548e-05,
-      "loss": 0.2443,
+      "epoch": 3.4519047104191443,
+      "grad_norm": 0.259569376707077,
+      "learning_rate": 1.1567373062225226e-05,
+      "loss": 0.3763,
       "step": 95780
     },
     {
-      "epoch": 3.37,
-      "learning_rate": 1.2704164882648034e-05,
-      "loss": 0.2494,
+      "epoch": 3.452084910080369,
+      "grad_norm": 0.28914976119995117,
+      "learning_rate": 1.1564912034947739e-05,
+      "loss": 0.3791,
       "step": 95785
     },
     {
-      "epoch": 3.37,
-      "learning_rate": 1.2701684611240835e-05,
-      "loss": 0.2548,
+      "epoch": 3.4522651097415937,
+      "grad_norm": 0.2676786482334137,
+      "learning_rate": 1.1562451190720733e-05,
+      "loss": 0.3611,
       "step": 95790
     },
     {
-      "epoch": 3.37,
-      "learning_rate": 1.2699204499519157e-05,
-      "loss": 0.2632,
+      "epoch": 3.4524453094028185,
+      "grad_norm": 0.2162621021270752,
+      "learning_rate": 1.155999052957773e-05,
+      "loss": 0.3395,
       "step": 95795
     },
     {
-      "epoch": 3.37,
-      "learning_rate": 1.2696724547515193e-05,
-      "loss": 0.2684,
+      "epoch": 3.4526255090640428,
+      "grad_norm": 0.26020073890686035,
+      "learning_rate": 1.155753005155225e-05,
+      "loss": 0.3507,
       "step": 95800
     },
     {
-      "epoch": 3.37,
-      "learning_rate": 1.2694244755261164e-05,
-      "loss": 0.2437,
+      "epoch": 3.4528057087252675,
+      "grad_norm": 0.2733500003814697,
+      "learning_rate": 1.1555069756677824e-05,
+      "loss": 0.377,
       "step": 95805
     },
     {
-      "epoch": 3.37,
-      "learning_rate": 1.2691765122789256e-05,
-      "loss": 0.2587,
+      "epoch": 3.452985908386492,
+      "grad_norm": 0.25482839345932007,
+      "learning_rate": 1.155260964498796e-05,
+      "loss": 0.3767,
       "step": 95810
     },
     {
-      "epoch": 3.37,
-      "learning_rate": 1.2689285650131654e-05,
-      "loss": 0.2623,
+      "epoch": 3.453166108047717,
+      "grad_norm": 0.2589060068130493,
+      "learning_rate": 1.1550149716516196e-05,
+      "loss": 0.3995,
       "step": 95815
     },
     {
-      "epoch": 3.37,
-      "learning_rate": 1.2686806337320578e-05,
-      "loss": 0.2859,
+      "epoch": 3.4533463077089417,
+      "grad_norm": 0.26133349537849426,
+      "learning_rate": 1.1547689971296038e-05,
+      "loss": 0.3989,
       "step": 95820
     },
     {
-      "epoch": 3.37,
-      "learning_rate": 1.2684327184388195e-05,
-      "loss": 0.2551,
+      "epoch": 3.453526507370166,
+      "grad_norm": 0.2330733835697174,
+      "learning_rate": 1.1545230409360996e-05,
+      "loss": 0.3813,
       "step": 95825
     },
     {
-      "epoch": 3.37,
-      "learning_rate": 1.2681848191366719e-05,
-      "loss": 0.2557,
+      "epoch": 3.4537067070313907,
+      "grad_norm": 0.2710002660751343,
+      "learning_rate": 1.154277103074459e-05,
+      "loss": 0.3821,
       "step": 95830
     },
     {
-      "epoch": 3.37,
-      "learning_rate": 1.2679369358288323e-05,
-      "loss": 0.2455,
+      "epoch": 3.4538869066926154,
+      "grad_norm": 0.24312447011470795,
+      "learning_rate": 1.154031183548032e-05,
+      "loss": 0.3863,
       "step": 95835
     },
     {
-      "epoch": 3.37,
-      "learning_rate": 1.2676890685185188e-05,
-      "loss": 0.2614,
+      "epoch": 3.45406710635384,
+      "grad_norm": 0.21377715468406677,
+      "learning_rate": 1.1537852823601685e-05,
+      "loss": 0.3886,
       "step": 95840
     },
     {
-      "epoch": 3.37,
-      "learning_rate": 1.2674412172089517e-05,
-      "loss": 0.2673,
+      "epoch": 3.4542473060150645,
+      "grad_norm": 0.2441663295030594,
+      "learning_rate": 1.1535393995142222e-05,
+      "loss": 0.3983,
       "step": 95845
     },
     {
-      "epoch": 3.37,
-      "learning_rate": 1.2671933819033482e-05,
-      "loss": 0.255,
+      "epoch": 3.454427505676289,
+      "grad_norm": 0.21451100707054138,
+      "learning_rate": 1.1532935350135387e-05,
+      "loss": 0.3596,
       "step": 95850
     },
     {
-      "epoch": 3.37,
-      "learning_rate": 1.2669455626049251e-05,
-      "loss": 0.2618,
+      "epoch": 3.454607705337514,
+      "grad_norm": 0.22826121747493744,
+      "learning_rate": 1.1530476888614711e-05,
+      "loss": 0.3795,
       "step": 95855
     },
     {
-      "epoch": 3.37,
-      "learning_rate": 1.266697759316902e-05,
-      "loss": 0.247,
+      "epoch": 3.4547879049987387,
+      "grad_norm": 0.2341785728931427,
+      "learning_rate": 1.1528018610613678e-05,
+      "loss": 0.3178,
       "step": 95860
     },
     {
-      "epoch": 3.37,
-      "learning_rate": 1.266449972042496e-05,
-      "loss": 0.2691,
+      "epoch": 3.4549681046599634,
+      "grad_norm": 0.25270435214042664,
+      "learning_rate": 1.1525560516165784e-05,
+      "loss": 0.3635,
       "step": 95865
     },
     {
-      "epoch": 3.37,
-      "learning_rate": 1.2662022007849234e-05,
-      "loss": 0.2587,
+      "epoch": 3.4551483043211877,
+      "grad_norm": 0.2552870512008667,
+      "learning_rate": 1.152310260530452e-05,
+      "loss": 0.3629,
       "step": 95870
     },
     {
-      "epoch": 3.37,
-      "learning_rate": 1.2659544455474026e-05,
-      "loss": 0.2492,
+      "epoch": 3.4553285039824124,
+      "grad_norm": 0.24541820585727692,
+      "learning_rate": 1.1520644878063374e-05,
+      "loss": 0.3886,
       "step": 95875
     },
     {
-      "epoch": 3.37,
-      "learning_rate": 1.2657067063331492e-05,
-      "loss": 0.2823,
+      "epoch": 3.455508703643637,
+      "grad_norm": 0.22713278234004974,
+      "learning_rate": 1.1518187334475822e-05,
+      "loss": 0.3915,
       "step": 95880
     },
     {
-      "epoch": 3.37,
-      "learning_rate": 1.265458983145382e-05,
-      "loss": 0.2437,
+      "epoch": 3.455688903304862,
+      "grad_norm": 0.2381686270236969,
+      "learning_rate": 1.1515729974575369e-05,
+      "loss": 0.4036,
       "step": 95885
     },
     {
-      "epoch": 3.37,
-      "learning_rate": 1.265211275987316e-05,
-      "loss": 0.2672,
+      "epoch": 3.455869102966086,
+      "grad_norm": 0.2205180674791336,
+      "learning_rate": 1.151327279839548e-05,
+      "loss": 0.4164,
       "step": 95890
     },
     {
-      "epoch": 3.37,
-      "learning_rate": 1.2649635848621672e-05,
-      "loss": 0.2517,
+      "epoch": 3.456049302627311,
+      "grad_norm": 0.2498241364955902,
+      "learning_rate": 1.1510815805969641e-05,
+      "loss": 0.3929,
       "step": 95895
     },
     {
-      "epoch": 3.37,
-      "learning_rate": 1.2647159097731532e-05,
-      "loss": 0.2606,
+      "epoch": 3.4562295022885356,
+      "grad_norm": 0.20146778225898743,
+      "learning_rate": 1.1508358997331322e-05,
+      "loss": 0.378,
       "step": 95900
     },
     {
-      "epoch": 3.37,
-      "learning_rate": 1.264468250723489e-05,
-      "loss": 0.2482,
+      "epoch": 3.4564097019497604,
+      "grad_norm": 0.22101639211177826,
+      "learning_rate": 1.1505902372513994e-05,
+      "loss": 0.3808,
       "step": 95905
     },
     {
-      "epoch": 3.37,
-      "learning_rate": 1.26422060771639e-05,
-      "loss": 0.2635,
+      "epoch": 3.456589901610985,
+      "grad_norm": 0.19309233129024506,
+      "learning_rate": 1.150344593155114e-05,
+      "loss": 0.4071,
       "step": 95910
     },
     {
-      "epoch": 3.37,
-      "learning_rate": 1.2639729807550715e-05,
-      "loss": 0.2477,
+      "epoch": 3.45677010127221,
+      "grad_norm": 0.2546674907207489,
+      "learning_rate": 1.1500989674476234e-05,
+      "loss": 0.3693,
       "step": 95915
     },
     {
-      "epoch": 3.37,
-      "learning_rate": 1.2637253698427498e-05,
-      "loss": 0.2564,
+      "epoch": 3.456950300933434,
+      "grad_norm": 0.2687970697879791,
+      "learning_rate": 1.1498533601322711e-05,
+      "loss": 0.3548,
       "step": 95920
     },
     {
-      "epoch": 3.37,
-      "learning_rate": 1.26347777498264e-05,
-      "loss": 0.2578,
+      "epoch": 3.457130500594659,
+      "grad_norm": 0.22413934767246246,
+      "learning_rate": 1.1496077712124062e-05,
+      "loss": 0.3728,
       "step": 95925
     },
     {
-      "epoch": 3.38,
-      "learning_rate": 1.2632301961779547e-05,
-      "loss": 0.2635,
+      "epoch": 3.4573107002558836,
+      "grad_norm": 0.20262411236763,
+      "learning_rate": 1.1493622006913728e-05,
+      "loss": 0.3563,
       "step": 95930
     },
     {
-      "epoch": 3.38,
-      "learning_rate": 1.2629826334319108e-05,
-      "loss": 0.2843,
+      "epoch": 3.4574908999171083,
+      "grad_norm": 0.18897296488285065,
+      "learning_rate": 1.1491166485725194e-05,
+      "loss": 0.3703,
       "step": 95935
     },
     {
-      "epoch": 3.38,
-      "learning_rate": 1.2627350867477231e-05,
-      "loss": 0.2645,
+      "epoch": 3.4576710995783326,
+      "grad_norm": 0.2627841830253601,
+      "learning_rate": 1.1488711148591907e-05,
+      "loss": 0.4015,
       "step": 95940
     },
     {
-      "epoch": 3.38,
-      "learning_rate": 1.2624875561286049e-05,
-      "loss": 0.262,
+      "epoch": 3.4578512992395574,
+      "grad_norm": 0.19592086970806122,
+      "learning_rate": 1.1486255995547299e-05,
+      "loss": 0.3791,
       "step": 95945
     },
     {
-      "epoch": 3.38,
-      "learning_rate": 1.2622400415777699e-05,
-      "loss": 0.2535,
+      "epoch": 3.458031498900782,
+      "grad_norm": 0.27027082443237305,
+      "learning_rate": 1.1483801026624844e-05,
+      "loss": 0.3186,
       "step": 95950
     },
     {
-      "epoch": 3.38,
-      "learning_rate": 1.2619925430984313e-05,
-      "loss": 0.2517,
+      "epoch": 3.458211698562007,
+      "grad_norm": 0.2537202537059784,
+      "learning_rate": 1.1481346241857982e-05,
+      "loss": 0.3907,
       "step": 95955
     },
     {
-      "epoch": 3.38,
-      "learning_rate": 1.2617450606938052e-05,
-      "loss": 0.2612,
+      "epoch": 3.4583918982232316,
+      "grad_norm": 0.2511982023715973,
+      "learning_rate": 1.1478891641280162e-05,
+      "loss": 0.3686,
       "step": 95960
     },
     {
-      "epoch": 3.38,
-      "learning_rate": 1.2614975943671031e-05,
-      "loss": 0.2554,
+      "epoch": 3.458572097884456,
+      "grad_norm": 0.2458512932062149,
+      "learning_rate": 1.1476437224924824e-05,
+      "loss": 0.3945,
       "step": 95965
     },
     {
-      "epoch": 3.38,
-      "learning_rate": 1.2612501441215374e-05,
-      "loss": 0.2715,
+      "epoch": 3.4587522975456806,
+      "grad_norm": 0.2354070544242859,
+      "learning_rate": 1.1473982992825403e-05,
+      "loss": 0.3661,
       "step": 95970
     },
     {
-      "epoch": 3.38,
-      "learning_rate": 1.2610027099603233e-05,
-      "loss": 0.2533,
+      "epoch": 3.4589324972069053,
+      "grad_norm": 0.22366875410079956,
+      "learning_rate": 1.1471528945015352e-05,
+      "loss": 0.4073,
       "step": 95975
     },
     {
-      "epoch": 3.38,
-      "learning_rate": 1.2607552918866725e-05,
-      "loss": 0.2504,
+      "epoch": 3.45911269686813,
+      "grad_norm": 0.3030519485473633,
+      "learning_rate": 1.1469075081528102e-05,
+      "loss": 0.387,
       "step": 95980
     },
     {
-      "epoch": 3.38,
-      "learning_rate": 1.2605078899037964e-05,
-      "loss": 0.2624,
+      "epoch": 3.4592928965293543,
+      "grad_norm": 0.1697065532207489,
+      "learning_rate": 1.1466621402397085e-05,
+      "loss": 0.383,
       "step": 95985
     },
     {
-      "epoch": 3.38,
-      "learning_rate": 1.260260504014909e-05,
-      "loss": 0.2599,
+      "epoch": 3.459473096190579,
+      "grad_norm": 0.2466737926006317,
+      "learning_rate": 1.1464167907655731e-05,
+      "loss": 0.3841,
       "step": 95990
     },
     {
-      "epoch": 3.38,
-      "learning_rate": 1.2600131342232224e-05,
-      "loss": 0.2644,
+      "epoch": 3.459653295851804,
+      "grad_norm": 0.2090192586183548,
+      "learning_rate": 1.1461714597337459e-05,
+      "loss": 0.3883,
       "step": 95995
     },
     {
-      "epoch": 3.38,
-      "learning_rate": 1.2597657805319485e-05,
-      "loss": 0.2574,
+      "epoch": 3.4598334955130285,
+      "grad_norm": 0.20262013375759125,
+      "learning_rate": 1.1459261471475713e-05,
+      "loss": 0.3893,
       "step": 96000
     },
     {
-      "epoch": 3.38,
-      "eval_loss": 0.2541818916797638,
-      "eval_runtime": 10.5477,
-      "eval_samples_per_second": 9.481,
-      "eval_steps_per_second": 9.481,
+      "epoch": 3.4598334955130285,
+      "eval_loss": 0.4300161600112915,
+      "eval_runtime": 3.5329,
+      "eval_samples_per_second": 28.305,
+      "eval_steps_per_second": 7.076,
       "step": 96000
     },
     {
-      "epoch": 3.38,
-      "learning_rate": 1.259518442944298e-05,
-      "loss": 0.2428,
+      "epoch": 3.4600136951742533,
+      "grad_norm": 0.23101657629013062,
+      "learning_rate": 1.1456808530103907e-05,
+      "loss": 0.3993,
       "step": 96005
     },
     {
-      "epoch": 3.38,
-      "learning_rate": 1.2592711214634826e-05,
-      "loss": 0.251,
+      "epoch": 3.4601938948354776,
+      "grad_norm": 0.2159048616886139,
+      "learning_rate": 1.1454355773255465e-05,
+      "loss": 0.3827,
       "step": 96010
     },
     {
-      "epoch": 3.38,
-      "learning_rate": 1.2590238160927148e-05,
-      "loss": 0.2491,
+      "epoch": 3.4603740944967023,
+      "grad_norm": 0.26098620891571045,
+      "learning_rate": 1.14519032009638e-05,
+      "loss": 0.3952,
       "step": 96015
     },
     {
-      "epoch": 3.38,
-      "learning_rate": 1.2587765268352048e-05,
-      "loss": 0.2583,
+      "epoch": 3.460554294157927,
+      "grad_norm": 0.22338443994522095,
+      "learning_rate": 1.1449450813262336e-05,
+      "loss": 0.366,
       "step": 96020
     },
     {
-      "epoch": 3.38,
-      "learning_rate": 1.2585292536941637e-05,
-      "loss": 0.2616,
+      "epoch": 3.4607344938191518,
+      "grad_norm": 0.218563973903656,
+      "learning_rate": 1.1446998610184467e-05,
+      "loss": 0.3993,
       "step": 96025
     },
     {
-      "epoch": 3.38,
-      "learning_rate": 1.258281996672801e-05,
-      "loss": 0.2655,
+      "epoch": 3.460914693480376,
+      "grad_norm": 0.2379787415266037,
+      "learning_rate": 1.1444546591763639e-05,
+      "loss": 0.3897,
       "step": 96030
     },
     {
-      "epoch": 3.38,
-      "learning_rate": 1.2580347557743294e-05,
-      "loss": 0.2491,
+      "epoch": 3.461094893141601,
+      "grad_norm": 0.2325468510389328,
+      "learning_rate": 1.144209475803322e-05,
+      "loss": 0.3757,
       "step": 96035
     },
     {
-      "epoch": 3.38,
-      "learning_rate": 1.2577875310019571e-05,
-      "loss": 0.2575,
+      "epoch": 3.4612750928028255,
+      "grad_norm": 0.237319678068161,
+      "learning_rate": 1.1439643109026643e-05,
+      "loss": 0.3858,
       "step": 96040
     },
     {
-      "epoch": 3.38,
-      "learning_rate": 1.2575403223588956e-05,
-      "loss": 0.2572,
+      "epoch": 3.4614552924640503,
+      "grad_norm": 0.19926400482654572,
+      "learning_rate": 1.1437191644777307e-05,
+      "loss": 0.3948,
       "step": 96045
     },
     {
-      "epoch": 3.38,
-      "learning_rate": 1.2572931298483537e-05,
-      "loss": 0.2645,
+      "epoch": 3.461635492125275,
+      "grad_norm": 0.20922797918319702,
+      "learning_rate": 1.1434740365318597e-05,
+      "loss": 0.352,
       "step": 96050
     },
     {
-      "epoch": 3.38,
-      "learning_rate": 1.2570459534735424e-05,
-      "loss": 0.2598,
+      "epoch": 3.4618156917864993,
+      "grad_norm": 0.2545210123062134,
+      "learning_rate": 1.1432289270683941e-05,
+      "loss": 0.4189,
       "step": 96055
     },
     {
-      "epoch": 3.38,
-      "learning_rate": 1.25679879323767e-05,
-      "loss": 0.2502,
+      "epoch": 3.461995891447724,
+      "grad_norm": 0.20487937331199646,
+      "learning_rate": 1.1429838360906708e-05,
+      "loss": 0.398,
       "step": 96060
     },
     {
-      "epoch": 3.38,
-      "learning_rate": 1.256551649143946e-05,
-      "loss": 0.2473,
+      "epoch": 3.4621760911089488,
+      "grad_norm": 0.27374470233917236,
+      "learning_rate": 1.1427387636020292e-05,
+      "loss": 0.3865,
       "step": 96065
     },
     {
-      "epoch": 3.38,
-      "learning_rate": 1.2563045211955787e-05,
-      "loss": 0.2557,
+      "epoch": 3.4623562907701735,
+      "grad_norm": 0.24212025105953217,
+      "learning_rate": 1.14249370960581e-05,
+      "loss": 0.3911,
       "step": 96070
     },
     {
-      "epoch": 3.38,
-      "learning_rate": 1.2560574093957784e-05,
-      "loss": 0.2662,
+      "epoch": 3.4625364904313978,
+      "grad_norm": 0.24381519854068756,
+      "learning_rate": 1.1422486741053513e-05,
+      "loss": 0.3624,
       "step": 96075
     },
     {
-      "epoch": 3.38,
-      "learning_rate": 1.2558103137477527e-05,
-      "loss": 0.2786,
+      "epoch": 3.4627166900926225,
+      "grad_norm": 0.23481976985931396,
+      "learning_rate": 1.1420036571039915e-05,
+      "loss": 0.3895,
       "step": 96080
     },
     {
-      "epoch": 3.38,
-      "learning_rate": 1.2555632342547096e-05,
-      "loss": 0.2703,
+      "epoch": 3.4628968897538472,
+      "grad_norm": 0.27169761061668396,
+      "learning_rate": 1.141758658605069e-05,
+      "loss": 0.39,
       "step": 96085
     },
     {
-      "epoch": 3.38,
-      "learning_rate": 1.2553161709198583e-05,
-      "loss": 0.2625,
+      "epoch": 3.463077089415072,
+      "grad_norm": 0.24660931527614594,
+      "learning_rate": 1.1415136786119207e-05,
+      "loss": 0.3867,
       "step": 96090
     },
     {
-      "epoch": 3.38,
-      "learning_rate": 1.2550691237464054e-05,
-      "loss": 0.2605,
+      "epoch": 3.4632572890762967,
+      "grad_norm": 0.23132185637950897,
+      "learning_rate": 1.1412687171278865e-05,
+      "loss": 0.4086,
       "step": 96095
     },
     {
-      "epoch": 3.38,
-      "learning_rate": 1.2548220927375604e-05,
-      "loss": 0.228,
+      "epoch": 3.463437488737521,
+      "grad_norm": 0.266564279794693,
+      "learning_rate": 1.1410237741563031e-05,
+      "loss": 0.3929,
       "step": 96100
     },
     {
-      "epoch": 3.38,
-      "learning_rate": 1.2545750778965292e-05,
-      "loss": 0.2524,
+      "epoch": 3.4636176883987457,
+      "grad_norm": 0.19676515460014343,
+      "learning_rate": 1.1407788497005076e-05,
+      "loss": 0.3475,
       "step": 96105
     },
     {
-      "epoch": 3.38,
-      "learning_rate": 1.2543280792265206e-05,
-      "loss": 0.2701,
+      "epoch": 3.4637978880599705,
+      "grad_norm": 0.2622312605381012,
+      "learning_rate": 1.140533943763837e-05,
+      "loss": 0.3491,
       "step": 96110
     },
     {
-      "epoch": 3.38,
-      "learning_rate": 1.2540810967307409e-05,
-      "loss": 0.2554,
+      "epoch": 3.463978087721195,
+      "grad_norm": 0.27703362703323364,
+      "learning_rate": 1.1402890563496276e-05,
+      "loss": 0.3801,
       "step": 96115
     },
     {
-      "epoch": 3.38,
-      "learning_rate": 1.2538341304123973e-05,
-      "loss": 0.2725,
+      "epoch": 3.4641582873824195,
+      "grad_norm": 0.24706853926181793,
+      "learning_rate": 1.1400441874612173e-05,
+      "loss": 0.3824,
       "step": 96120
     },
     {
-      "epoch": 3.38,
-      "learning_rate": 1.2535871802746951e-05,
-      "loss": 0.2588,
+      "epoch": 3.464338487043644,
+      "grad_norm": 0.2970876693725586,
+      "learning_rate": 1.1397993371019427e-05,
+      "loss": 0.3695,
       "step": 96125
     },
     {
-      "epoch": 3.38,
-      "learning_rate": 1.2533402463208432e-05,
-      "loss": 0.2784,
+      "epoch": 3.464518686704869,
+      "grad_norm": 0.24940580129623413,
+      "learning_rate": 1.1395545052751366e-05,
+      "loss": 0.3726,
       "step": 96130
     },
     {
-      "epoch": 3.38,
-      "learning_rate": 1.2530933285540464e-05,
-      "loss": 0.2701,
+      "epoch": 3.4646988863660937,
+      "grad_norm": 0.21381749212741852,
+      "learning_rate": 1.1393096919841384e-05,
+      "loss": 0.3976,
       "step": 96135
     },
     {
-      "epoch": 3.38,
-      "learning_rate": 1.2528464269775108e-05,
-      "loss": 0.2697,
+      "epoch": 3.4648790860273184,
+      "grad_norm": 0.2594471275806427,
+      "learning_rate": 1.1390648972322817e-05,
+      "loss": 0.3572,
       "step": 96140
     },
     {
-      "epoch": 3.38,
-      "learning_rate": 1.2525995415944419e-05,
-      "loss": 0.2534,
+      "epoch": 3.4650592856885427,
+      "grad_norm": 0.22999241948127747,
+      "learning_rate": 1.1388201210229027e-05,
+      "loss": 0.3397,
       "step": 96145
     },
     {
-      "epoch": 3.38,
-      "learning_rate": 1.2523526724080456e-05,
-      "loss": 0.272,
+      "epoch": 3.4652394853497674,
+      "grad_norm": 0.28377705812454224,
+      "learning_rate": 1.1385753633593357e-05,
+      "loss": 0.3886,
       "step": 96150
     },
     {
-      "epoch": 3.38,
-      "learning_rate": 1.2521058194215284e-05,
-      "loss": 0.2502,
+      "epoch": 3.465419685010992,
+      "grad_norm": 0.21517440676689148,
+      "learning_rate": 1.1383306242449152e-05,
+      "loss": 0.4021,
       "step": 96155
     },
     {
-      "epoch": 3.38,
-      "learning_rate": 1.251858982638095e-05,
-      "loss": 0.2615,
+      "epoch": 3.465599884672217,
+      "grad_norm": 0.26591652631759644,
+      "learning_rate": 1.138085903682977e-05,
+      "loss": 0.3855,
       "step": 96160
     },
     {
-      "epoch": 3.38,
-      "learning_rate": 1.2516121620609495e-05,
-      "loss": 0.2699,
+      "epoch": 3.4657800843334416,
+      "grad_norm": 0.2830233871936798,
+      "learning_rate": 1.1378412016768547e-05,
+      "loss": 0.3945,
       "step": 96165
     },
     {
-      "epoch": 3.38,
-      "learning_rate": 1.2513653576932977e-05,
-      "loss": 0.2711,
+      "epoch": 3.465960283994666,
+      "grad_norm": 0.2300427407026291,
+      "learning_rate": 1.1375965182298822e-05,
+      "loss": 0.3664,
       "step": 96170
     },
     {
-      "epoch": 3.38,
-      "learning_rate": 1.2511185695383443e-05,
-      "loss": 0.2523,
+      "epoch": 3.4661404836558907,
+      "grad_norm": 0.23842735588550568,
+      "learning_rate": 1.1373518533453937e-05,
+      "loss": 0.3729,
       "step": 96175
     },
     {
-      "epoch": 3.38,
-      "learning_rate": 1.2508717975992932e-05,
-      "loss": 0.2591,
+      "epoch": 3.4663206833171154,
+      "grad_norm": 0.2205519825220108,
+      "learning_rate": 1.1371072070267221e-05,
+      "loss": 0.3791,
       "step": 96180
     },
     {
-      "epoch": 3.38,
-      "learning_rate": 1.2506250418793475e-05,
-      "loss": 0.2689,
+      "epoch": 3.46650088297834,
+      "grad_norm": 0.22283774614334106,
+      "learning_rate": 1.1368625792772e-05,
+      "loss": 0.3938,
       "step": 96185
     },
     {
-      "epoch": 3.38,
-      "learning_rate": 1.250378302381713e-05,
-      "loss": 0.2366,
+      "epoch": 3.466681082639565,
+      "grad_norm": 0.23954735696315765,
+      "learning_rate": 1.1366179701001625e-05,
+      "loss": 0.4172,
       "step": 96190
     },
     {
-      "epoch": 3.38,
-      "learning_rate": 1.250131579109593e-05,
-      "loss": 0.2754,
+      "epoch": 3.466861282300789,
+      "grad_norm": 0.20571085810661316,
+      "learning_rate": 1.136373379498941e-05,
+      "loss": 0.3765,
       "step": 96195
     },
     {
-      "epoch": 3.38,
-      "learning_rate": 1.2498848720661899e-05,
-      "loss": 0.2485,
+      "epoch": 3.467041481962014,
+      "grad_norm": 0.28632012009620667,
+      "learning_rate": 1.1361288074768684e-05,
+      "loss": 0.3678,
       "step": 96200
     },
     {
-      "epoch": 3.38,
-      "learning_rate": 1.2496381812547086e-05,
-      "loss": 0.2825,
+      "epoch": 3.4672216816232386,
+      "grad_norm": 0.2133149951696396,
+      "learning_rate": 1.1358842540372767e-05,
+      "loss": 0.4013,
       "step": 96205
     },
     {
-      "epoch": 3.38,
-      "learning_rate": 1.2493915066783506e-05,
-      "loss": 0.2758,
+      "epoch": 3.4674018812844634,
+      "grad_norm": 0.17300593852996826,
+      "learning_rate": 1.1356397191834967e-05,
+      "loss": 0.3512,
       "step": 96210
     },
     {
-      "epoch": 3.39,
-      "learning_rate": 1.2491448483403206e-05,
-      "loss": 0.2512,
+      "epoch": 3.4675820809456877,
+      "grad_norm": 0.23370462656021118,
+      "learning_rate": 1.1353952029188625e-05,
+      "loss": 0.4099,
       "step": 96215
     },
     {
-      "epoch": 3.39,
-      "learning_rate": 1.2488982062438196e-05,
-      "loss": 0.2478,
+      "epoch": 3.4677622806069124,
+      "grad_norm": 0.2420395463705063,
+      "learning_rate": 1.1351507052467044e-05,
+      "loss": 0.3867,
       "step": 96220
     },
     {
-      "epoch": 3.39,
-      "learning_rate": 1.2486515803920518e-05,
-      "loss": 0.2593,
+      "epoch": 3.467942480268137,
+      "grad_norm": 0.2655438780784607,
+      "learning_rate": 1.1349062261703538e-05,
+      "loss": 0.3969,
       "step": 96225
     },
     {
-      "epoch": 3.39,
-      "learning_rate": 1.2484049707882187e-05,
-      "loss": 0.2603,
+      "epoch": 3.468122679929362,
+      "grad_norm": 0.24291475117206573,
+      "learning_rate": 1.1346617656931416e-05,
+      "loss": 0.4008,
       "step": 96230
     },
     {
-      "epoch": 3.39,
-      "learning_rate": 1.2481583774355218e-05,
-      "loss": 0.2691,
+      "epoch": 3.4683028795905866,
+      "grad_norm": 0.23957978188991547,
+      "learning_rate": 1.1344173238183974e-05,
+      "loss": 0.3572,
       "step": 96235
     },
     {
-      "epoch": 3.39,
-      "learning_rate": 1.2479118003371628e-05,
-      "loss": 0.2614,
+      "epoch": 3.468483079251811,
+      "grad_norm": 0.23568493127822876,
+      "learning_rate": 1.134172900549455e-05,
+      "loss": 0.3883,
       "step": 96240
     },
     {
-      "epoch": 3.39,
-      "learning_rate": 1.2476652394963448e-05,
-      "loss": 0.2582,
+      "epoch": 3.4686632789130356,
+      "grad_norm": 0.2132604718208313,
+      "learning_rate": 1.1339284958896412e-05,
+      "loss": 0.3945,
       "step": 96245
     },
     {
-      "epoch": 3.39,
-      "learning_rate": 1.2474186949162681e-05,
-      "loss": 0.2583,
+      "epoch": 3.4688434785742603,
+      "grad_norm": 0.25484153628349304,
+      "learning_rate": 1.1336841098422862e-05,
+      "loss": 0.3929,
       "step": 96250
     },
     {
-      "epoch": 3.39,
-      "learning_rate": 1.2471721666001334e-05,
-      "loss": 0.2465,
+      "epoch": 3.469023678235485,
+      "grad_norm": 0.3113678991794586,
+      "learning_rate": 1.1334397424107218e-05,
+      "loss": 0.4022,
       "step": 96255
     },
     {
-      "epoch": 3.39,
-      "learning_rate": 1.2469256545511435e-05,
-      "loss": 0.2863,
+      "epoch": 3.4692038778967094,
+      "grad_norm": 0.1955181211233139,
+      "learning_rate": 1.1331953935982761e-05,
+      "loss": 0.3452,
       "step": 96260
     },
     {
-      "epoch": 3.39,
-      "learning_rate": 1.2466791587724972e-05,
-      "loss": 0.2601,
+      "epoch": 3.469384077557934,
+      "grad_norm": 0.2540067732334137,
+      "learning_rate": 1.1329510634082787e-05,
+      "loss": 0.4174,
       "step": 96265
     },
     {
-      "epoch": 3.39,
-      "learning_rate": 1.2464326792673969e-05,
-      "loss": 0.2444,
+      "epoch": 3.469564277219159,
+      "grad_norm": 0.2300526648759842,
+      "learning_rate": 1.1327067518440585e-05,
+      "loss": 0.4029,
       "step": 96270
     },
     {
-      "epoch": 3.39,
-      "learning_rate": 1.2461862160390423e-05,
-      "loss": 0.2784,
+      "epoch": 3.4697444768803836,
+      "grad_norm": 0.2536962330341339,
+      "learning_rate": 1.132462458908943e-05,
+      "loss": 0.3517,
       "step": 96275
     },
     {
-      "epoch": 3.39,
-      "learning_rate": 1.2459397690906321e-05,
-      "loss": 0.247,
+      "epoch": 3.4699246765416083,
+      "grad_norm": 0.24605917930603027,
+      "learning_rate": 1.1322181846062626e-05,
+      "loss": 0.3896,
       "step": 96280
     },
     {
-      "epoch": 3.39,
-      "learning_rate": 1.2456933384253689e-05,
-      "loss": 0.2483,
+      "epoch": 3.4701048762028326,
+      "grad_norm": 0.21459710597991943,
+      "learning_rate": 1.1319739289393448e-05,
+      "loss": 0.4341,
       "step": 96285
     },
     {
-      "epoch": 3.39,
-      "learning_rate": 1.2454469240464506e-05,
-      "loss": 0.255,
+      "epoch": 3.4702850758640573,
+      "grad_norm": 0.2235480546951294,
+      "learning_rate": 1.1317296919115171e-05,
+      "loss": 0.3664,
       "step": 96290
     },
     {
-      "epoch": 3.39,
-      "learning_rate": 1.2452005259570776e-05,
-      "loss": 0.2449,
+      "epoch": 3.470465275525282,
+      "grad_norm": 0.23965181410312653,
+      "learning_rate": 1.1314854735261076e-05,
+      "loss": 0.3913,
       "step": 96295
     },
     {
-      "epoch": 3.39,
-      "learning_rate": 1.2449541441604478e-05,
-      "loss": 0.2525,
+      "epoch": 3.470645475186507,
+      "grad_norm": 0.29223349690437317,
+      "learning_rate": 1.1312412737864429e-05,
+      "loss": 0.4089,
       "step": 96300
     },
     {
-      "epoch": 3.39,
-      "learning_rate": 1.2447077786597621e-05,
-      "loss": 0.264,
+      "epoch": 3.470825674847731,
+      "grad_norm": 0.2220560908317566,
+      "learning_rate": 1.1309970926958513e-05,
+      "loss": 0.3817,
       "step": 96305
     },
     {
-      "epoch": 3.39,
-      "learning_rate": 1.244461429458218e-05,
-      "loss": 0.2626,
+      "epoch": 3.471005874508956,
+      "grad_norm": 0.26034578680992126,
+      "learning_rate": 1.1307529302576605e-05,
+      "loss": 0.3736,
       "step": 96310
     },
     {
-      "epoch": 3.39,
-      "learning_rate": 1.2442150965590154e-05,
-      "loss": 0.2405,
+      "epoch": 3.4711860741701805,
+      "grad_norm": 0.23067830502986908,
+      "learning_rate": 1.130508786475194e-05,
+      "loss": 0.3885,
       "step": 96315
     },
     {
-      "epoch": 3.39,
-      "learning_rate": 1.2439687799653516e-05,
-      "loss": 0.2791,
+      "epoch": 3.4713662738314053,
+      "grad_norm": 0.2320612072944641,
+      "learning_rate": 1.1302646613517812e-05,
+      "loss": 0.3833,
       "step": 96320
     },
     {
-      "epoch": 3.39,
-      "learning_rate": 1.2437224796804262e-05,
-      "loss": 0.2476,
+      "epoch": 3.47154647349263,
+      "grad_norm": 0.22921355068683624,
+      "learning_rate": 1.130020554890747e-05,
+      "loss": 0.3706,
       "step": 96325
     },
     {
-      "epoch": 3.39,
-      "learning_rate": 1.2434761957074365e-05,
-      "loss": 0.2756,
+      "epoch": 3.4717266731538543,
+      "grad_norm": 0.24659670889377594,
+      "learning_rate": 1.1297764670954166e-05,
+      "loss": 0.3904,
       "step": 96330
     },
     {
-      "epoch": 3.39,
-      "learning_rate": 1.2432299280495794e-05,
-      "loss": 0.2364,
+      "epoch": 3.471906872815079,
+      "grad_norm": 0.2158777266740799,
+      "learning_rate": 1.1295323979691183e-05,
+      "loss": 0.389,
       "step": 96335
     },
     {
-      "epoch": 3.39,
-      "learning_rate": 1.2429836767100541e-05,
-      "loss": 0.2747,
+      "epoch": 3.4720870724763038,
+      "grad_norm": 0.21258202195167542,
+      "learning_rate": 1.1292883475151741e-05,
+      "loss": 0.3774,
       "step": 96340
     },
     {
-      "epoch": 3.39,
-      "learning_rate": 1.2427374416920575e-05,
-      "loss": 0.2762,
+      "epoch": 3.4722672721375285,
+      "grad_norm": 0.20848548412322998,
+      "learning_rate": 1.1290443157369115e-05,
+      "loss": 0.373,
       "step": 96345
     },
     {
-      "epoch": 3.39,
-      "learning_rate": 1.242491222998787e-05,
-      "loss": 0.2775,
+      "epoch": 3.472447471798753,
+      "grad_norm": 0.268587201833725,
+      "learning_rate": 1.1288003026376551e-05,
+      "loss": 0.3654,
       "step": 96350
     },
     {
-      "epoch": 3.39,
-      "learning_rate": 1.242245020633438e-05,
-      "loss": 0.241,
+      "epoch": 3.4726276714599775,
+      "grad_norm": 0.25624382495880127,
+      "learning_rate": 1.1285563082207286e-05,
+      "loss": 0.3601,
       "step": 96355
     },
     {
-      "epoch": 3.39,
-      "learning_rate": 1.2419988345992095e-05,
-      "loss": 0.2515,
+      "epoch": 3.4728078711212023,
+      "grad_norm": 0.1996915340423584,
+      "learning_rate": 1.1283123324894573e-05,
+      "loss": 0.3577,
       "step": 96360
     },
     {
-      "epoch": 3.39,
-      "learning_rate": 1.2417526648992972e-05,
-      "loss": 0.2556,
+      "epoch": 3.472988070782427,
+      "grad_norm": 0.32410600781440735,
+      "learning_rate": 1.1280683754471647e-05,
+      "loss": 0.3696,
       "step": 96365
     },
     {
-      "epoch": 3.39,
-      "learning_rate": 1.2415065115368962e-05,
-      "loss": 0.2698,
+      "epoch": 3.4731682704436517,
+      "grad_norm": 0.20257310569286346,
+      "learning_rate": 1.1278244370971739e-05,
+      "loss": 0.3503,
       "step": 96370
     },
     {
-      "epoch": 3.39,
-      "learning_rate": 1.2412603745152038e-05,
-      "loss": 0.2618,
+      "epoch": 3.473348470104876,
+      "grad_norm": 0.2658899426460266,
+      "learning_rate": 1.1275805174428102e-05,
+      "loss": 0.4159,
       "step": 96375
     },
     {
-      "epoch": 3.39,
-      "learning_rate": 1.2410142538374167e-05,
-      "loss": 0.2633,
+      "epoch": 3.4735286697661008,
+      "grad_norm": 0.22585046291351318,
+      "learning_rate": 1.1273366164873967e-05,
+      "loss": 0.3779,
       "step": 96380
     },
     {
-      "epoch": 3.39,
-      "learning_rate": 1.2407681495067303e-05,
-      "loss": 0.2549,
+      "epoch": 3.4737088694273255,
+      "grad_norm": 0.24556881189346313,
+      "learning_rate": 1.1270927342342558e-05,
+      "loss": 0.4055,
       "step": 96385
     },
     {
-      "epoch": 3.39,
-      "learning_rate": 1.240522061526339e-05,
-      "loss": 0.2463,
+      "epoch": 3.47388906908855,
+      "grad_norm": 0.20033571124076843,
+      "learning_rate": 1.1268488706867105e-05,
+      "loss": 0.3803,
       "step": 96390
     },
     {
-      "epoch": 3.39,
-      "learning_rate": 1.2402759898994382e-05,
-      "loss": 0.2453,
+      "epoch": 3.4740692687497745,
+      "grad_norm": 0.2153259962797165,
+      "learning_rate": 1.1266050258480829e-05,
+      "loss": 0.4001,
       "step": 96395
     },
     {
-      "epoch": 3.39,
-      "learning_rate": 1.2400299346292244e-05,
-      "loss": 0.2395,
+      "epoch": 3.4742494684109992,
+      "grad_norm": 0.24744583666324615,
+      "learning_rate": 1.1263611997216966e-05,
+      "loss": 0.3978,
       "step": 96400
     },
     {
-      "epoch": 3.39,
-      "learning_rate": 1.2397838957188912e-05,
-      "loss": 0.2588,
+      "epoch": 3.474429668072224,
+      "grad_norm": 0.2664938271045685,
+      "learning_rate": 1.1261173923108731e-05,
+      "loss": 0.3655,
       "step": 96405
     },
     {
-      "epoch": 3.39,
-      "learning_rate": 1.239537873171634e-05,
-      "loss": 0.2691,
+      "epoch": 3.4746098677334487,
+      "grad_norm": 0.23613092303276062,
+      "learning_rate": 1.125873603618934e-05,
+      "loss": 0.407,
       "step": 96410
     },
     {
-      "epoch": 3.39,
-      "learning_rate": 1.2392918669906456e-05,
-      "loss": 0.2514,
+      "epoch": 3.4747900673946734,
+      "grad_norm": 0.2801758348941803,
+      "learning_rate": 1.1256298336492013e-05,
+      "loss": 0.3698,
       "step": 96415
     },
     {
-      "epoch": 3.39,
-      "learning_rate": 1.2390458771791227e-05,
-      "loss": 0.2651,
+      "epoch": 3.474970267055898,
+      "grad_norm": 0.24601031839847565,
+      "learning_rate": 1.1253860824049948e-05,
+      "loss": 0.4104,
       "step": 96420
     },
     {
-      "epoch": 3.39,
-      "learning_rate": 1.238799903740257e-05,
-      "loss": 0.2565,
+      "epoch": 3.4751504667171225,
+      "grad_norm": 0.2248278707265854,
+      "learning_rate": 1.1251423498896389e-05,
+      "loss": 0.3917,
       "step": 96425
     },
     {
-      "epoch": 3.39,
-      "learning_rate": 1.2385539466772434e-05,
-      "loss": 0.2823,
+      "epoch": 3.475330666378347,
+      "grad_norm": 0.2194819152355194,
+      "learning_rate": 1.1248986361064511e-05,
+      "loss": 0.3757,
       "step": 96430
     },
     {
-      "epoch": 3.39,
-      "learning_rate": 1.2383080059932764e-05,
-      "loss": 0.2646,
+      "epoch": 3.475510866039572,
+      "grad_norm": 0.22775894403457642,
+      "learning_rate": 1.1246549410587522e-05,
+      "loss": 0.3658,
       "step": 96435
     },
     {
-      "epoch": 3.39,
-      "learning_rate": 1.2380620816915484e-05,
-      "loss": 0.2515,
+      "epoch": 3.4756910657007967,
+      "grad_norm": 0.23458243906497955,
+      "learning_rate": 1.1244112647498647e-05,
+      "loss": 0.3502,
       "step": 96440
     },
     {
-      "epoch": 3.39,
-      "learning_rate": 1.2378161737752525e-05,
-      "loss": 0.2457,
+      "epoch": 3.475871265362021,
+      "grad_norm": 0.24316877126693726,
+      "learning_rate": 1.1241676071831073e-05,
+      "loss": 0.4056,
       "step": 96445
     },
     {
-      "epoch": 3.39,
-      "learning_rate": 1.237570282247581e-05,
-      "loss": 0.2659,
+      "epoch": 3.4760514650232457,
+      "grad_norm": 0.25940200686454773,
+      "learning_rate": 1.1239239683617997e-05,
+      "loss": 0.3978,
       "step": 96450
     },
     {
-      "epoch": 3.39,
-      "learning_rate": 1.2373244071117285e-05,
-      "loss": 0.2197,
+      "epoch": 3.4762316646844704,
+      "grad_norm": 0.24130047857761383,
+      "learning_rate": 1.123680348289262e-05,
+      "loss": 0.3834,
       "step": 96455
     },
     {
-      "epoch": 3.39,
-      "learning_rate": 1.2370785483708863e-05,
-      "loss": 0.2739,
+      "epoch": 3.476411864345695,
+      "grad_norm": 0.1958744078874588,
+      "learning_rate": 1.1234367469688118e-05,
+      "loss": 0.3965,
       "step": 96460
     },
     {
-      "epoch": 3.39,
-      "learning_rate": 1.2368327060282466e-05,
-      "loss": 0.2702,
+      "epoch": 3.47659206400692,
+      "grad_norm": 0.29350563883781433,
+      "learning_rate": 1.1231931644037701e-05,
+      "loss": 0.3916,
       "step": 96465
     },
     {
-      "epoch": 3.39,
-      "learning_rate": 1.2365868800870011e-05,
-      "loss": 0.2533,
+      "epoch": 3.476772263668144,
+      "grad_norm": 0.21596458554267883,
+      "learning_rate": 1.1229496005974554e-05,
+      "loss": 0.372,
       "step": 96470
     },
     {
-      "epoch": 3.39,
-      "learning_rate": 1.2363410705503433e-05,
-      "loss": 0.2643,
+      "epoch": 3.476952463329369,
+      "grad_norm": 0.24967950582504272,
+      "learning_rate": 1.1227060555531858e-05,
+      "loss": 0.3712,
       "step": 96475
     },
     {
-      "epoch": 3.39,
-      "learning_rate": 1.2360952774214629e-05,
-      "loss": 0.2736,
+      "epoch": 3.4771326629905936,
+      "grad_norm": 0.20245712995529175,
+      "learning_rate": 1.1224625292742794e-05,
+      "loss": 0.344,
       "step": 96480
     },
     {
-      "epoch": 3.39,
-      "learning_rate": 1.2358495007035534e-05,
-      "loss": 0.2449,
+      "epoch": 3.4773128626518184,
+      "grad_norm": 0.26654964685440063,
+      "learning_rate": 1.1222190217640535e-05,
+      "loss": 0.4232,
       "step": 96485
     },
     {
-      "epoch": 3.39,
-      "learning_rate": 1.2356037403998042e-05,
-      "loss": 0.264,
+      "epoch": 3.4774930623130427,
+      "grad_norm": 0.2164486050605774,
+      "learning_rate": 1.1219755330258274e-05,
+      "loss": 0.3632,
       "step": 96490
     },
     {
-      "epoch": 3.39,
-      "learning_rate": 1.235357996513408e-05,
-      "loss": 0.246,
+      "epoch": 3.4776732619742674,
+      "grad_norm": 0.2758568525314331,
+      "learning_rate": 1.1217320630629184e-05,
+      "loss": 0.3687,
       "step": 96495
     },
     {
-      "epoch": 3.4,
-      "learning_rate": 1.2351122690475544e-05,
-      "loss": 0.2577,
+      "epoch": 3.477853461635492,
+      "grad_norm": 0.24314354360103607,
+      "learning_rate": 1.1214886118786425e-05,
+      "loss": 0.3426,
       "step": 96500
     },
     {
-      "epoch": 3.4,
-      "eval_loss": 0.25399091839790344,
-      "eval_runtime": 10.5362,
-      "eval_samples_per_second": 9.491,
-      "eval_steps_per_second": 9.491,
+      "epoch": 3.477853461635492,
+      "eval_loss": 0.42988404631614685,
+      "eval_runtime": 3.5336,
+      "eval_samples_per_second": 28.3,
+      "eval_steps_per_second": 7.075,
       "step": 96500
     },
     {
-      "epoch": 3.4,
-      "learning_rate": 1.2348665580054347e-05,
-      "loss": 0.2472,
+      "epoch": 3.478033661296717,
+      "grad_norm": 0.19991134107112885,
+      "learning_rate": 1.1212451794763178e-05,
+      "loss": 0.3617,
       "step": 96505
     },
     {
-      "epoch": 3.4,
-      "learning_rate": 1.234620863390238e-05,
-      "loss": 0.2501,
+      "epoch": 3.4782138609579416,
+      "grad_norm": 0.23575247824192047,
+      "learning_rate": 1.1210017658592605e-05,
+      "loss": 0.4334,
       "step": 96510
     },
     {
-      "epoch": 3.4,
-      "learning_rate": 1.2343751852051566e-05,
-      "loss": 0.2679,
+      "epoch": 3.478394060619166,
+      "grad_norm": 0.23354025185108185,
+      "learning_rate": 1.1207583710307861e-05,
+      "loss": 0.3685,
       "step": 96515
     },
     {
-      "epoch": 3.4,
-      "learning_rate": 1.2341295234533792e-05,
-      "loss": 0.2335,
+      "epoch": 3.4785742602803906,
+      "grad_norm": 0.27086594700813293,
+      "learning_rate": 1.1205149949942139e-05,
+      "loss": 0.418,
       "step": 96520
     },
     {
-      "epoch": 3.4,
-      "learning_rate": 1.2338838781380957e-05,
-      "loss": 0.2598,
+      "epoch": 3.4787544599416154,
+      "grad_norm": 0.2325955182313919,
+      "learning_rate": 1.120271637752856e-05,
+      "loss": 0.3802,
       "step": 96525
     },
     {
-      "epoch": 3.4,
-      "learning_rate": 1.2336382492624948e-05,
-      "loss": 0.2633,
+      "epoch": 3.47893465960284,
+      "grad_norm": 0.25425323843955994,
+      "learning_rate": 1.1200282993100305e-05,
+      "loss": 0.3639,
       "step": 96530
     },
     {
-      "epoch": 3.4,
-      "learning_rate": 1.2333926368297666e-05,
-      "loss": 0.2346,
+      "epoch": 3.4791148592640644,
+      "grad_norm": 0.21064163744449615,
+      "learning_rate": 1.1197849796690527e-05,
+      "loss": 0.3845,
       "step": 96535
     },
     {
-      "epoch": 3.4,
-      "learning_rate": 1.2331470408431012e-05,
-      "loss": 0.2565,
+      "epoch": 3.479295058925289,
+      "grad_norm": 0.20380394160747528,
+      "learning_rate": 1.1195416788332371e-05,
+      "loss": 0.3789,
       "step": 96540
     },
     {
-      "epoch": 3.4,
-      "learning_rate": 1.2329014613056855e-05,
-      "loss": 0.277,
+      "epoch": 3.479475258586514,
+      "grad_norm": 0.24418340623378754,
+      "learning_rate": 1.119298396805899e-05,
+      "loss": 0.3595,
       "step": 96545
     },
     {
-      "epoch": 3.4,
-      "learning_rate": 1.2326558982207103e-05,
-      "loss": 0.2606,
+      "epoch": 3.4796554582477386,
+      "grad_norm": 0.19111719727516174,
+      "learning_rate": 1.1190551335903526e-05,
+      "loss": 0.3371,
       "step": 96550
     },
     {
-      "epoch": 3.4,
-      "learning_rate": 1.2324103515913632e-05,
-      "loss": 0.2622,
+      "epoch": 3.4798356579089633,
+      "grad_norm": 0.1976500004529953,
+      "learning_rate": 1.1188118891899122e-05,
+      "loss": 0.3705,
       "step": 96555
     },
     {
-      "epoch": 3.4,
-      "learning_rate": 1.2321648214208322e-05,
-      "loss": 0.2664,
+      "epoch": 3.4800158575701876,
+      "grad_norm": 0.21880120038986206,
+      "learning_rate": 1.118568663607893e-05,
+      "loss": 0.3656,
       "step": 96560
     },
     {
-      "epoch": 3.4,
-      "learning_rate": 1.2319193077123048e-05,
-      "loss": 0.2556,
+      "epoch": 3.4801960572314123,
+      "grad_norm": 0.23799078166484833,
+      "learning_rate": 1.1183254568476084e-05,
+      "loss": 0.4213,
       "step": 96565
     },
     {
-      "epoch": 3.4,
-      "learning_rate": 1.2316738104689702e-05,
-      "loss": 0.2709,
+      "epoch": 3.480376256892637,
+      "grad_norm": 0.2790437638759613,
+      "learning_rate": 1.1180822689123719e-05,
+      "loss": 0.3503,
       "step": 96570
     },
     {
-      "epoch": 3.4,
-      "learning_rate": 1.2314283296940151e-05,
-      "loss": 0.2433,
+      "epoch": 3.480556456553862,
+      "grad_norm": 0.2113523781299591,
+      "learning_rate": 1.1178390998054968e-05,
+      "loss": 0.3869,
       "step": 96575
     },
     {
-      "epoch": 3.4,
-      "learning_rate": 1.2311828653906274e-05,
-      "loss": 0.2474,
+      "epoch": 3.480736656215086,
+      "grad_norm": 0.20490826666355133,
+      "learning_rate": 1.1175959495302957e-05,
+      "loss": 0.3975,
       "step": 96580
     },
     {
-      "epoch": 3.4,
-      "learning_rate": 1.2309374175619932e-05,
-      "loss": 0.2536,
+      "epoch": 3.480916855876311,
+      "grad_norm": 0.19894984364509583,
+      "learning_rate": 1.117352818090083e-05,
+      "loss": 0.364,
       "step": 96585
     },
     {
-      "epoch": 3.4,
-      "learning_rate": 1.2306919862112998e-05,
-      "loss": 0.2532,
+      "epoch": 3.4810970555375356,
+      "grad_norm": 0.2029849737882614,
+      "learning_rate": 1.1171097054881705e-05,
+      "loss": 0.3436,
       "step": 96590
     },
     {
-      "epoch": 3.4,
-      "learning_rate": 1.2304465713417357e-05,
-      "loss": 0.2661,
+      "epoch": 3.4812772551987603,
+      "grad_norm": 0.21581901609897614,
+      "learning_rate": 1.1168666117278704e-05,
+      "loss": 0.3586,
       "step": 96595
     },
     {
-      "epoch": 3.4,
-      "learning_rate": 1.2302011729564861e-05,
-      "loss": 0.2787,
+      "epoch": 3.481457454859985,
+      "grad_norm": 0.27598705887794495,
+      "learning_rate": 1.116623536812495e-05,
+      "loss": 0.3926,
       "step": 96600
     },
     {
-      "epoch": 3.4,
-      "learning_rate": 1.2299557910587366e-05,
-      "loss": 0.2458,
+      "epoch": 3.4816376545212093,
+      "grad_norm": 0.24678505957126617,
+      "learning_rate": 1.1163804807453551e-05,
+      "loss": 0.3838,
       "step": 96605
     },
     {
-      "epoch": 3.4,
-      "learning_rate": 1.229710425651675e-05,
-      "loss": 0.2451,
+      "epoch": 3.481817854182434,
+      "grad_norm": 0.1914645880460739,
+      "learning_rate": 1.1161374435297653e-05,
+      "loss": 0.3738,
       "step": 96610
     },
     {
-      "epoch": 3.4,
-      "learning_rate": 1.2294650767384866e-05,
-      "loss": 0.2549,
+      "epoch": 3.481998053843659,
+      "grad_norm": 0.22241747379302979,
+      "learning_rate": 1.1158944251690337e-05,
+      "loss": 0.3985,
       "step": 96615
     },
     {
-      "epoch": 3.4,
-      "learning_rate": 1.2292197443223567e-05,
-      "loss": 0.2621,
+      "epoch": 3.4821782535048835,
+      "grad_norm": 0.2281738668680191,
+      "learning_rate": 1.1156514256664719e-05,
+      "loss": 0.3981,
       "step": 96620
     },
     {
-      "epoch": 3.4,
-      "learning_rate": 1.22897442840647e-05,
-      "loss": 0.2517,
+      "epoch": 3.482358453166108,
+      "grad_norm": 0.20726355910301208,
+      "learning_rate": 1.1154084450253924e-05,
+      "loss": 0.3826,
       "step": 96625
     },
     {
-      "epoch": 3.4,
-      "learning_rate": 1.2287291289940137e-05,
-      "loss": 0.2579,
+      "epoch": 3.4825386528273325,
+      "grad_norm": 0.2736766040325165,
+      "learning_rate": 1.1151654832491034e-05,
+      "loss": 0.3828,
       "step": 96630
     },
     {
-      "epoch": 3.4,
-      "learning_rate": 1.2284838460881718e-05,
-      "loss": 0.2619,
+      "epoch": 3.4827188524885573,
+      "grad_norm": 0.15977123379707336,
+      "learning_rate": 1.1149225403409189e-05,
+      "loss": 0.3602,
       "step": 96635
     },
     {
-      "epoch": 3.4,
-      "learning_rate": 1.2282385796921284e-05,
-      "loss": 0.2806,
+      "epoch": 3.482899052149782,
+      "grad_norm": 0.25133785605430603,
+      "learning_rate": 1.1146796163041456e-05,
+      "loss": 0.3763,
       "step": 96640
     },
     {
-      "epoch": 3.4,
-      "learning_rate": 1.2279933298090695e-05,
-      "loss": 0.2447,
+      "epoch": 3.4830792518110067,
+      "grad_norm": 0.2637987434864044,
+      "learning_rate": 1.1144367111420934e-05,
+      "loss": 0.3561,
       "step": 96645
     },
     {
-      "epoch": 3.4,
-      "learning_rate": 1.2277480964421784e-05,
-      "loss": 0.2726,
+      "epoch": 3.483259451472231,
+      "grad_norm": 0.262717604637146,
+      "learning_rate": 1.1141938248580736e-05,
+      "loss": 0.3806,
       "step": 96650
     },
     {
-      "epoch": 3.4,
-      "learning_rate": 1.2275028795946405e-05,
-      "loss": 0.264,
+      "epoch": 3.4834396511334558,
+      "grad_norm": 0.21309638023376465,
+      "learning_rate": 1.1139509574553944e-05,
+      "loss": 0.3634,
       "step": 96655
     },
     {
-      "epoch": 3.4,
-      "learning_rate": 1.2272576792696383e-05,
-      "loss": 0.2581,
+      "epoch": 3.4836198507946805,
+      "grad_norm": 0.22101840376853943,
+      "learning_rate": 1.1137081089373655e-05,
+      "loss": 0.3906,
       "step": 96660
     },
     {
-      "epoch": 3.4,
-      "learning_rate": 1.2270124954703575e-05,
-      "loss": 0.247,
+      "epoch": 3.4838000504559052,
+      "grad_norm": 0.25005871057510376,
+      "learning_rate": 1.113465279307295e-05,
+      "loss": 0.3673,
       "step": 96665
     },
     {
-      "epoch": 3.4,
-      "learning_rate": 1.22676732819998e-05,
-      "loss": 0.2573,
+      "epoch": 3.48398025011713,
+      "grad_norm": 0.2766312062740326,
+      "learning_rate": 1.1132224685684906e-05,
+      "loss": 0.3975,
       "step": 96670
     },
     {
-      "epoch": 3.4,
-      "learning_rate": 1.22652217746169e-05,
-      "loss": 0.2721,
+      "epoch": 3.4841604497783543,
+      "grad_norm": 0.23252266645431519,
+      "learning_rate": 1.1129796767242625e-05,
+      "loss": 0.3683,
       "step": 96675
     },
     {
-      "epoch": 3.4,
-      "learning_rate": 1.2262770432586692e-05,
-      "loss": 0.2413,
+      "epoch": 3.484340649439579,
+      "grad_norm": 0.3283499777317047,
+      "learning_rate": 1.1127369037779182e-05,
+      "loss": 0.3911,
       "step": 96680
     },
     {
-      "epoch": 3.4,
-      "learning_rate": 1.2260319255941028e-05,
-      "loss": 0.2581,
+      "epoch": 3.4845208491008037,
+      "grad_norm": 0.31998345255851746,
+      "learning_rate": 1.1124941497327646e-05,
+      "loss": 0.3783,
       "step": 96685
     },
     {
-      "epoch": 3.4,
-      "learning_rate": 1.2257868244711717e-05,
-      "loss": 0.2703,
+      "epoch": 3.4847010487620285,
+      "grad_norm": 0.21715426445007324,
+      "learning_rate": 1.1122514145921097e-05,
+      "loss": 0.3533,
       "step": 96690
     },
     {
-      "epoch": 3.4,
-      "learning_rate": 1.2255417398930586e-05,
-      "loss": 0.2916,
+      "epoch": 3.484881248423253,
+      "grad_norm": 0.23155249655246735,
+      "learning_rate": 1.1120086983592606e-05,
+      "loss": 0.3644,
       "step": 96695
     },
     {
-      "epoch": 3.4,
-      "learning_rate": 1.225296671862947e-05,
-      "loss": 0.2592,
+      "epoch": 3.4850614480844775,
+      "grad_norm": 0.2212996780872345,
+      "learning_rate": 1.1117660010375233e-05,
+      "loss": 0.3898,
       "step": 96700
     },
     {
-      "epoch": 3.4,
-      "learning_rate": 1.2250516203840167e-05,
-      "loss": 0.2426,
+      "epoch": 3.485241647745702,
+      "grad_norm": 0.18644365668296814,
+      "learning_rate": 1.1115233226302074e-05,
+      "loss": 0.3834,
       "step": 96705
     },
     {
-      "epoch": 3.4,
-      "learning_rate": 1.224806585459452e-05,
-      "loss": 0.2486,
+      "epoch": 3.485421847406927,
+      "grad_norm": 0.21037942171096802,
+      "learning_rate": 1.1112806631406153e-05,
+      "loss": 0.3391,
       "step": 96710
     },
     {
-      "epoch": 3.4,
-      "learning_rate": 1.2245615670924334e-05,
-      "loss": 0.2327,
+      "epoch": 3.4856020470681517,
+      "grad_norm": 0.25046810507774353,
+      "learning_rate": 1.1110380225720565e-05,
+      "loss": 0.3941,
       "step": 96715
     },
     {
-      "epoch": 3.4,
-      "learning_rate": 1.2243165652861416e-05,
-      "loss": 0.2592,
+      "epoch": 3.485782246729376,
+      "grad_norm": 0.23813064396381378,
+      "learning_rate": 1.1107954009278357e-05,
+      "loss": 0.386,
       "step": 96720
     },
     {
-      "epoch": 3.4,
-      "learning_rate": 1.2240715800437591e-05,
-      "loss": 0.2651,
+      "epoch": 3.4859624463906007,
+      "grad_norm": 0.2435433268547058,
+      "learning_rate": 1.1105527982112584e-05,
+      "loss": 0.3704,
       "step": 96725
     },
     {
-      "epoch": 3.4,
-      "learning_rate": 1.2238266113684666e-05,
-      "loss": 0.2399,
+      "epoch": 3.4861426460518254,
+      "grad_norm": 0.219720259308815,
+      "learning_rate": 1.1103102144256305e-05,
+      "loss": 0.3791,
       "step": 96730
     },
     {
-      "epoch": 3.4,
-      "learning_rate": 1.2235816592634441e-05,
-      "loss": 0.2465,
+      "epoch": 3.48632284571305,
+      "grad_norm": 0.24152056872844696,
+      "learning_rate": 1.1100676495742568e-05,
+      "loss": 0.3654,
       "step": 96735
     },
     {
-      "epoch": 3.4,
-      "learning_rate": 1.2233367237318718e-05,
-      "loss": 0.251,
+      "epoch": 3.486503045374275,
+      "grad_norm": 0.22711649537086487,
+      "learning_rate": 1.1098251036604413e-05,
+      "loss": 0.397,
       "step": 96740
     },
     {
-      "epoch": 3.4,
-      "learning_rate": 1.2230918047769319e-05,
-      "loss": 0.2387,
+      "epoch": 3.486683245035499,
+      "grad_norm": 0.22654442489147186,
+      "learning_rate": 1.1095825766874904e-05,
+      "loss": 0.3741,
       "step": 96745
     },
     {
-      "epoch": 3.4,
-      "learning_rate": 1.2228469024018024e-05,
-      "loss": 0.24,
+      "epoch": 3.486863444696724,
+      "grad_norm": 0.20491252839565277,
+      "learning_rate": 1.109340068658708e-05,
+      "loss": 0.3717,
       "step": 96750
     },
     {
-      "epoch": 3.4,
-      "learning_rate": 1.222602016609665e-05,
-      "loss": 0.2775,
+      "epoch": 3.4870436443579487,
+      "grad_norm": 0.200746089220047,
+      "learning_rate": 1.1090975795773979e-05,
+      "loss": 0.3603,
       "step": 96755
     },
     {
-      "epoch": 3.4,
-      "learning_rate": 1.2223571474036976e-05,
-      "loss": 0.2399,
+      "epoch": 3.4872238440191734,
+      "grad_norm": 0.2560223639011383,
+      "learning_rate": 1.1088551094468636e-05,
+      "loss": 0.3774,
       "step": 96760
     },
     {
-      "epoch": 3.4,
-      "learning_rate": 1.222112294787082e-05,
-      "loss": 0.2629,
+      "epoch": 3.4874040436803977,
+      "grad_norm": 0.22208933532238007,
+      "learning_rate": 1.1086126582704085e-05,
+      "loss": 0.3932,
       "step": 96765
     },
     {
-      "epoch": 3.4,
-      "learning_rate": 1.2218674587629959e-05,
-      "loss": 0.2504,
+      "epoch": 3.4875842433416224,
+      "grad_norm": 0.25349971652030945,
+      "learning_rate": 1.1083702260513373e-05,
+      "loss": 0.3906,
       "step": 96770
     },
     {
-      "epoch": 3.4,
-      "learning_rate": 1.2216226393346173e-05,
-      "loss": 0.2398,
+      "epoch": 3.487764443002847,
+      "grad_norm": 0.20808152854442596,
+      "learning_rate": 1.1081278127929534e-05,
+      "loss": 0.3931,
       "step": 96775
     },
     {
-      "epoch": 3.4,
-      "learning_rate": 1.2213778365051273e-05,
-      "loss": 0.2523,
+      "epoch": 3.487944642664072,
+      "grad_norm": 0.18402156233787537,
+      "learning_rate": 1.1078854184985567e-05,
+      "loss": 0.3678,
       "step": 96780
     },
     {
-      "epoch": 3.41,
-      "learning_rate": 1.2211330502777036e-05,
-      "loss": 0.2766,
+      "epoch": 3.4881248423252966,
+      "grad_norm": 0.22728796303272247,
+      "learning_rate": 1.1076430431714526e-05,
+      "loss": 0.3698,
       "step": 96785
     },
     {
-      "epoch": 3.41,
-      "learning_rate": 1.2208882806555243e-05,
-      "loss": 0.2539,
+      "epoch": 3.488305041986521,
+      "grad_norm": 0.2671549618244171,
+      "learning_rate": 1.1074006868149413e-05,
+      "loss": 0.3913,
       "step": 96790
     },
     {
-      "epoch": 3.41,
-      "learning_rate": 1.2206435276417669e-05,
-      "loss": 0.2499,
+      "epoch": 3.4884852416477456,
+      "grad_norm": 0.2308701127767563,
+      "learning_rate": 1.1071583494323274e-05,
+      "loss": 0.3614,
       "step": 96795
     },
     {
-      "epoch": 3.41,
-      "learning_rate": 1.2203987912396108e-05,
-      "loss": 0.2583,
+      "epoch": 3.4886654413089704,
+      "grad_norm": 0.22043147683143616,
+      "learning_rate": 1.106916031026912e-05,
+      "loss": 0.378,
       "step": 96800
     },
     {
-      "epoch": 3.41,
-      "learning_rate": 1.2201540714522336e-05,
-      "loss": 0.2588,
+      "epoch": 3.488845640970195,
+      "grad_norm": 0.28679531812667847,
+      "learning_rate": 1.106673731601994e-05,
+      "loss": 0.3863,
       "step": 96805
     },
     {
-      "epoch": 3.41,
-      "learning_rate": 1.2199093682828111e-05,
-      "loss": 0.2623,
+      "epoch": 3.4890258406314194,
+      "grad_norm": 0.25224077701568604,
+      "learning_rate": 1.1064314511608778e-05,
+      "loss": 0.381,
       "step": 96810
     },
     {
-      "epoch": 3.41,
-      "learning_rate": 1.2196646817345217e-05,
-      "loss": 0.2326,
+      "epoch": 3.489206040292644,
+      "grad_norm": 0.252714604139328,
+      "learning_rate": 1.1061891897068624e-05,
+      "loss": 0.37,
       "step": 96815
     },
     {
-      "epoch": 3.41,
-      "learning_rate": 1.219420011810544e-05,
-      "loss": 0.2569,
+      "epoch": 3.489386239953869,
+      "grad_norm": 0.2400377094745636,
+      "learning_rate": 1.105946947243251e-05,
+      "loss": 0.3898,
       "step": 96820
     },
     {
-      "epoch": 3.41,
-      "learning_rate": 1.219175358514053e-05,
-      "loss": 0.2758,
+      "epoch": 3.4895664396150936,
+      "grad_norm": 0.26662567257881165,
+      "learning_rate": 1.1057047237733417e-05,
+      "loss": 0.3624,
       "step": 96825
     },
     {
-      "epoch": 3.41,
-      "learning_rate": 1.2189307218482263e-05,
-      "loss": 0.2601,
+      "epoch": 3.4897466392763183,
+      "grad_norm": 0.219301238656044,
+      "learning_rate": 1.1054625193004347e-05,
+      "loss": 0.3836,
       "step": 96830
     },
     {
-      "epoch": 3.41,
-      "learning_rate": 1.2186861018162387e-05,
-      "loss": 0.2643,
+      "epoch": 3.4899268389375426,
+      "grad_norm": 0.2580451965332031,
+      "learning_rate": 1.1052203338278319e-05,
+      "loss": 0.3923,
       "step": 96835
     },
     {
-      "epoch": 3.41,
-      "learning_rate": 1.2184414984212689e-05,
-      "loss": 0.2625,
+      "epoch": 3.4901070385987674,
+      "grad_norm": 0.23948605358600616,
+      "learning_rate": 1.104978167358832e-05,
+      "loss": 0.3659,
       "step": 96840
     },
     {
-      "epoch": 3.41,
-      "learning_rate": 1.2181969116664912e-05,
-      "loss": 0.2553,
+      "epoch": 3.490287238259992,
+      "grad_norm": 0.20990577340126038,
+      "learning_rate": 1.1047360198967344e-05,
+      "loss": 0.389,
       "step": 96845
     },
     {
-      "epoch": 3.41,
-      "learning_rate": 1.2179523415550822e-05,
-      "loss": 0.2407,
+      "epoch": 3.490467437921217,
+      "grad_norm": 0.2310859113931656,
+      "learning_rate": 1.1044938914448385e-05,
+      "loss": 0.3828,
       "step": 96850
     },
     {
-      "epoch": 3.41,
-      "learning_rate": 1.2177077880902162e-05,
-      "loss": 0.2398,
+      "epoch": 3.490647637582441,
+      "grad_norm": 0.19339106976985931,
+      "learning_rate": 1.104251782006442e-05,
+      "loss": 0.3703,
       "step": 96855
     },
     {
-      "epoch": 3.41,
-      "learning_rate": 1.2174632512750706e-05,
-      "loss": 0.2668,
+      "epoch": 3.490827837243666,
+      "grad_norm": 0.23496823012828827,
+      "learning_rate": 1.104009691584846e-05,
+      "loss": 0.3635,
       "step": 96860
     },
     {
-      "epoch": 3.41,
-      "learning_rate": 1.2172187311128181e-05,
-      "loss": 0.2671,
+      "epoch": 3.4910080369048906,
+      "grad_norm": 0.28526195883750916,
+      "learning_rate": 1.1037676201833474e-05,
+      "loss": 0.4448,
       "step": 96865
     },
     {
-      "epoch": 3.41,
-      "learning_rate": 1.2169742276066362e-05,
-      "loss": 0.2285,
+      "epoch": 3.4911882365661153,
+      "grad_norm": 0.23180098831653595,
+      "learning_rate": 1.103525567805245e-05,
+      "loss": 0.3416,
       "step": 96870
     },
     {
-      "epoch": 3.41,
-      "learning_rate": 1.2167297407596972e-05,
-      "loss": 0.2851,
+      "epoch": 3.49136843622734,
+      "grad_norm": 0.20511779189109802,
+      "learning_rate": 1.1032835344538362e-05,
+      "loss": 0.3618,
       "step": 96875
     },
     {
-      "epoch": 3.41,
-      "learning_rate": 1.2164852705751782e-05,
-      "loss": 0.2614,
+      "epoch": 3.4915486358885643,
+      "grad_norm": 0.22699345648288727,
+      "learning_rate": 1.1030415201324188e-05,
+      "loss": 0.3726,
       "step": 96880
     },
     {
-      "epoch": 3.41,
-      "learning_rate": 1.2162408170562518e-05,
-      "loss": 0.257,
+      "epoch": 3.491728835549789,
+      "grad_norm": 0.24510350823402405,
+      "learning_rate": 1.1027995248442891e-05,
+      "loss": 0.3828,
       "step": 96885
     },
     {
-      "epoch": 3.41,
-      "learning_rate": 1.2159963802060925e-05,
-      "loss": 0.2346,
+      "epoch": 3.491909035211014,
+      "grad_norm": 0.2351260483264923,
+      "learning_rate": 1.1025575485927476e-05,
+      "loss": 0.3986,
       "step": 96890
     },
     {
-      "epoch": 3.41,
-      "learning_rate": 1.215751960027873e-05,
-      "loss": 0.249,
+      "epoch": 3.4920892348722385,
+      "grad_norm": 0.24812495708465576,
+      "learning_rate": 1.1023155913810868e-05,
+      "loss": 0.3722,
       "step": 96895
     },
     {
-      "epoch": 3.41,
-      "learning_rate": 1.2155075565247686e-05,
-      "loss": 0.2442,
+      "epoch": 3.492269434533463,
+      "grad_norm": 0.26317039132118225,
+      "learning_rate": 1.1020736532126063e-05,
+      "loss": 0.3707,
       "step": 96900
     },
     {
-      "epoch": 3.41,
-      "learning_rate": 1.2152631696999525e-05,
-      "loss": 0.2477,
+      "epoch": 3.4924496341946876,
+      "grad_norm": 0.2346991002559662,
+      "learning_rate": 1.101831734090602e-05,
+      "loss": 0.3692,
       "step": 96905
     },
     {
-      "epoch": 3.41,
-      "learning_rate": 1.2150187995565965e-05,
-      "loss": 0.262,
+      "epoch": 3.4926298338559123,
+      "grad_norm": 0.2354336977005005,
+      "learning_rate": 1.1015898340183684e-05,
+      "loss": 0.3982,
       "step": 96910
     },
     {
-      "epoch": 3.41,
-      "learning_rate": 1.2147744460978753e-05,
-      "loss": 0.2629,
+      "epoch": 3.492810033517137,
+      "grad_norm": 0.2742219567298889,
+      "learning_rate": 1.1013479529992047e-05,
+      "loss": 0.3985,
       "step": 96915
     },
     {
-      "epoch": 3.41,
-      "learning_rate": 1.2145301093269598e-05,
-      "loss": 0.2606,
+      "epoch": 3.4929902331783618,
+      "grad_norm": 0.24369125068187714,
+      "learning_rate": 1.1011060910364033e-05,
+      "loss": 0.4218,
       "step": 96920
     },
     {
-      "epoch": 3.41,
-      "learning_rate": 1.214285789247025e-05,
-      "loss": 0.2496,
+      "epoch": 3.4931704328395865,
+      "grad_norm": 0.2888088822364807,
+      "learning_rate": 1.1008642481332596e-05,
+      "loss": 0.3919,
       "step": 96925
     },
     {
-      "epoch": 3.41,
-      "learning_rate": 1.214041485861241e-05,
-      "loss": 0.2718,
+      "epoch": 3.493350632500811,
+      "grad_norm": 0.2732216715812683,
+      "learning_rate": 1.1006224242930705e-05,
+      "loss": 0.3953,
       "step": 96930
     },
     {
-      "epoch": 3.41,
-      "learning_rate": 1.2137971991727815e-05,
-      "loss": 0.245,
+      "epoch": 3.4935308321620355,
+      "grad_norm": 0.2236909568309784,
+      "learning_rate": 1.1003806195191298e-05,
+      "loss": 0.3726,
       "step": 96935
     },
     {
-      "epoch": 3.41,
-      "learning_rate": 1.2135529291848177e-05,
-      "loss": 0.2643,
+      "epoch": 3.4937110318232603,
+      "grad_norm": 0.21694554388523102,
+      "learning_rate": 1.1001388338147326e-05,
+      "loss": 0.3782,
       "step": 96940
     },
     {
-      "epoch": 3.41,
-      "learning_rate": 1.2133086759005217e-05,
-      "loss": 0.2774,
+      "epoch": 3.493891231484485,
+      "grad_norm": 0.2243148237466812,
+      "learning_rate": 1.0998970671831726e-05,
+      "loss": 0.3572,
       "step": 96945
     },
     {
-      "epoch": 3.41,
-      "learning_rate": 1.213064439323063e-05,
-      "loss": 0.2857,
+      "epoch": 3.4940714311457093,
+      "grad_norm": 0.22870691120624542,
+      "learning_rate": 1.099655319627743e-05,
+      "loss": 0.3504,
       "step": 96950
     },
     {
-      "epoch": 3.41,
-      "learning_rate": 1.212820219455616e-05,
-      "loss": 0.2753,
+      "epoch": 3.494251630806934,
+      "grad_norm": 0.2252253144979477,
+      "learning_rate": 1.0994135911517396e-05,
+      "loss": 0.3817,
       "step": 96955
     },
     {
-      "epoch": 3.41,
-      "learning_rate": 1.2125760163013497e-05,
-      "loss": 0.241,
+      "epoch": 3.4944318304681588,
+      "grad_norm": 0.21160556375980377,
+      "learning_rate": 1.0991718817584549e-05,
+      "loss": 0.3751,
       "step": 96960
     },
     {
-      "epoch": 3.41,
-      "learning_rate": 1.2123318298634356e-05,
-      "loss": 0.2469,
+      "epoch": 3.4946120301293835,
+      "grad_norm": 0.229017436504364,
+      "learning_rate": 1.0989301914511818e-05,
+      "loss": 0.3454,
       "step": 96965
     },
     {
-      "epoch": 3.41,
-      "learning_rate": 1.2120876601450432e-05,
-      "loss": 0.2252,
+      "epoch": 3.494792229790608,
+      "grad_norm": 0.20936749875545502,
+      "learning_rate": 1.0986885202332136e-05,
+      "loss": 0.3819,
       "step": 96970
     },
     {
-      "epoch": 3.41,
-      "learning_rate": 1.211843507149344e-05,
-      "loss": 0.2413,
+      "epoch": 3.4949724294518325,
+      "grad_norm": 0.21880267560482025,
+      "learning_rate": 1.0984468681078424e-05,
+      "loss": 0.3691,
       "step": 96975
     },
     {
-      "epoch": 3.41,
-      "learning_rate": 1.2115993708795085e-05,
-      "loss": 0.2777,
+      "epoch": 3.4951526291130572,
+      "grad_norm": 0.22391989827156067,
+      "learning_rate": 1.098205235078362e-05,
+      "loss": 0.4176,
       "step": 96980
     },
     {
-      "epoch": 3.41,
-      "learning_rate": 1.2113552513387061e-05,
-      "loss": 0.2447,
+      "epoch": 3.495332828774282,
+      "grad_norm": 0.28297480940818787,
+      "learning_rate": 1.0979636211480648e-05,
+      "loss": 0.3618,
       "step": 96985
     },
     {
-      "epoch": 3.41,
-      "learning_rate": 1.2111111485301054e-05,
-      "loss": 0.259,
+      "epoch": 3.4955130284355067,
+      "grad_norm": 0.22317856550216675,
+      "learning_rate": 1.09772202632024e-05,
+      "loss": 0.3755,
       "step": 96990
     },
     {
-      "epoch": 3.41,
-      "learning_rate": 1.2108670624568786e-05,
-      "loss": 0.2624,
+      "epoch": 3.495693228096731,
+      "grad_norm": 0.2054082453250885,
+      "learning_rate": 1.0974804505981822e-05,
+      "loss": 0.3781,
       "step": 96995
     },
     {
-      "epoch": 3.41,
-      "learning_rate": 1.2106229931221929e-05,
-      "loss": 0.2697,
+      "epoch": 3.4958734277579557,
+      "grad_norm": 0.2986920177936554,
+      "learning_rate": 1.0972388939851804e-05,
+      "loss": 0.3982,
       "step": 97000
     },
     {
-      "epoch": 3.41,
-      "eval_loss": 0.2534578740596771,
-      "eval_runtime": 10.5485,
-      "eval_samples_per_second": 9.48,
-      "eval_steps_per_second": 9.48,
+      "epoch": 3.4958734277579557,
+      "eval_loss": 0.4296031594276428,
+      "eval_runtime": 3.5247,
+      "eval_samples_per_second": 28.371,
+      "eval_steps_per_second": 7.093,
       "step": 97000
     },
     {
-      "epoch": 3.41,
-      "learning_rate": 1.210378940529218e-05,
-      "loss": 0.2617,
+      "epoch": 3.4960536274191805,
+      "grad_norm": 0.25335046648979187,
+      "learning_rate": 1.0969973564845293e-05,
+      "loss": 0.4044,
       "step": 97005
     },
     {
-      "epoch": 3.41,
-      "learning_rate": 1.2101349046811216e-05,
-      "loss": 0.2551,
+      "epoch": 3.496233827080405,
+      "grad_norm": 0.23634032905101776,
+      "learning_rate": 1.0967558380995165e-05,
+      "loss": 0.4013,
       "step": 97010
     },
     {
-      "epoch": 3.41,
-      "learning_rate": 1.2098908855810745e-05,
-      "loss": 0.2594,
+      "epoch": 3.49641402674163,
+      "grad_norm": 0.2781918942928314,
+      "learning_rate": 1.0965143388334329e-05,
+      "loss": 0.3848,
       "step": 97015
     },
     {
-      "epoch": 3.41,
-      "learning_rate": 1.209646883232244e-05,
-      "loss": 0.2491,
+      "epoch": 3.496594226402854,
+      "grad_norm": 0.266989141702652,
+      "learning_rate": 1.0962728586895706e-05,
+      "loss": 0.3611,
       "step": 97020
     },
     {
-      "epoch": 3.41,
-      "learning_rate": 1.2094028976377978e-05,
-      "loss": 0.2569,
+      "epoch": 3.496774426064079,
+      "grad_norm": 0.2467830330133438,
+      "learning_rate": 1.0960313976712188e-05,
+      "loss": 0.3863,
       "step": 97025
     },
     {
-      "epoch": 3.41,
-      "learning_rate": 1.209158928800904e-05,
-      "loss": 0.2469,
+      "epoch": 3.4969546257253037,
+      "grad_norm": 0.2517973482608795,
+      "learning_rate": 1.0957899557816676e-05,
+      "loss": 0.3733,
       "step": 97030
     },
     {
-      "epoch": 3.41,
-      "learning_rate": 1.2089149767247317e-05,
-      "loss": 0.2686,
+      "epoch": 3.4971348253865284,
+      "grad_norm": 0.2016320675611496,
+      "learning_rate": 1.095548533024206e-05,
+      "loss": 0.4017,
       "step": 97035
     },
     {
-      "epoch": 3.41,
-      "learning_rate": 1.2086710414124477e-05,
-      "loss": 0.2714,
+      "epoch": 3.4973150250477527,
+      "grad_norm": 0.2199011892080307,
+      "learning_rate": 1.095307129402123e-05,
+      "loss": 0.3199,
       "step": 97040
     },
     {
-      "epoch": 3.41,
-      "learning_rate": 1.2084271228672183e-05,
-      "loss": 0.261,
+      "epoch": 3.4974952247089774,
+      "grad_norm": 0.23464830219745636,
+      "learning_rate": 1.0950657449187094e-05,
+      "loss": 0.3899,
       "step": 97045
     },
     {
-      "epoch": 3.41,
-      "learning_rate": 1.2081832210922125e-05,
-      "loss": 0.2628,
+      "epoch": 3.497675424370202,
+      "grad_norm": 0.2523633539676666,
+      "learning_rate": 1.0948243795772528e-05,
+      "loss": 0.3891,
       "step": 97050
     },
     {
-      "epoch": 3.41,
-      "learning_rate": 1.207939336090596e-05,
-      "loss": 0.2624,
+      "epoch": 3.497855624031427,
+      "grad_norm": 0.26079466938972473,
+      "learning_rate": 1.0945830333810423e-05,
+      "loss": 0.3812,
       "step": 97055
     },
     {
-      "epoch": 3.41,
-      "learning_rate": 1.2076954678655361e-05,
-      "loss": 0.2594,
+      "epoch": 3.4980358236926516,
+      "grad_norm": 0.2218736708164215,
+      "learning_rate": 1.0943417063333655e-05,
+      "loss": 0.3631,
       "step": 97060
     },
     {
-      "epoch": 3.42,
-      "learning_rate": 1.2074516164201977e-05,
-      "loss": 0.2724,
+      "epoch": 3.498216023353876,
+      "grad_norm": 0.19591349363327026,
+      "learning_rate": 1.0941003984375112e-05,
+      "loss": 0.3689,
       "step": 97065
     },
     {
-      "epoch": 3.42,
-      "learning_rate": 1.2072077817577493e-05,
-      "loss": 0.2579,
+      "epoch": 3.4983962230151007,
+      "grad_norm": 0.2568768858909607,
+      "learning_rate": 1.0938591096967657e-05,
+      "loss": 0.4084,
       "step": 97070
     },
     {
-      "epoch": 3.42,
-      "learning_rate": 1.2069639638813557e-05,
-      "loss": 0.23,
+      "epoch": 3.4985764226763254,
+      "grad_norm": 0.21716725826263428,
+      "learning_rate": 1.0936178401144184e-05,
+      "loss": 0.387,
       "step": 97075
     },
     {
-      "epoch": 3.42,
-      "learning_rate": 1.206720162794182e-05,
-      "loss": 0.2357,
+      "epoch": 3.49875662233755,
+      "grad_norm": 0.24818527698516846,
+      "learning_rate": 1.0933765896937556e-05,
+      "loss": 0.402,
       "step": 97080
     },
     {
-      "epoch": 3.42,
-      "learning_rate": 1.2064763784993958e-05,
-      "loss": 0.248,
+      "epoch": 3.4989368219987744,
+      "grad_norm": 0.23442403972148895,
+      "learning_rate": 1.093135358438065e-05,
+      "loss": 0.373,
       "step": 97085
     },
     {
-      "epoch": 3.42,
-      "learning_rate": 1.2062326110001603e-05,
-      "loss": 0.2556,
+      "epoch": 3.499117021659999,
+      "grad_norm": 0.2103729248046875,
+      "learning_rate": 1.0928941463506322e-05,
+      "loss": 0.3536,
       "step": 97090
     },
     {
-      "epoch": 3.42,
-      "learning_rate": 1.2059888602996428e-05,
-      "loss": 0.2812,
+      "epoch": 3.499297221321224,
+      "grad_norm": 0.21677030622959137,
+      "learning_rate": 1.0926529534347435e-05,
+      "loss": 0.3647,
       "step": 97095
     },
     {
-      "epoch": 3.42,
-      "learning_rate": 1.205745126401007e-05,
-      "loss": 0.2516,
+      "epoch": 3.4994774209824486,
+      "grad_norm": 0.24412192404270172,
+      "learning_rate": 1.0924117796936878e-05,
+      "loss": 0.4133,
       "step": 97100
     },
     {
-      "epoch": 3.42,
-      "learning_rate": 1.2055014093074168e-05,
-      "loss": 0.2561,
+      "epoch": 3.4996576206436734,
+      "grad_norm": 0.247679203748703,
+      "learning_rate": 1.0921706251307481e-05,
+      "loss": 0.4191,
       "step": 97105
     },
     {
-      "epoch": 3.42,
-      "learning_rate": 1.2052577090220387e-05,
-      "loss": 0.2539,
+      "epoch": 3.4998378203048977,
+      "grad_norm": 0.23608912527561188,
+      "learning_rate": 1.09192948974921e-05,
+      "loss": 0.3869,
       "step": 97110
     },
     {
-      "epoch": 3.42,
-      "learning_rate": 1.2050140255480361e-05,
-      "loss": 0.2494,
+      "epoch": 3.5000180199661224,
+      "grad_norm": 0.23635424673557281,
+      "learning_rate": 1.0916883735523612e-05,
+      "loss": 0.3906,
       "step": 97115
     },
     {
-      "epoch": 3.42,
-      "learning_rate": 1.2047703588885728e-05,
-      "loss": 0.2374,
+      "epoch": 3.500198219627347,
+      "grad_norm": 0.3293451964855194,
+      "learning_rate": 1.0914472765434852e-05,
+      "loss": 0.3959,
       "step": 97120
     },
     {
-      "epoch": 3.42,
-      "learning_rate": 1.2045267090468117e-05,
-      "loss": 0.245,
+      "epoch": 3.500378419288572,
+      "grad_norm": 0.21557281911373138,
+      "learning_rate": 1.091206198725868e-05,
+      "loss": 0.36,
       "step": 97125
     },
     {
-      "epoch": 3.42,
-      "learning_rate": 1.2042830760259189e-05,
-      "loss": 0.2559,
+      "epoch": 3.500558618949796,
+      "grad_norm": 0.2755977511405945,
+      "learning_rate": 1.090965140102793e-05,
+      "loss": 0.3945,
       "step": 97130
     },
     {
-      "epoch": 3.42,
-      "learning_rate": 1.2040394598290553e-05,
-      "loss": 0.2556,
+      "epoch": 3.500738818611021,
+      "grad_norm": 0.18358829617500305,
+      "learning_rate": 1.0907241006775445e-05,
+      "loss": 0.3519,
       "step": 97135
     },
     {
-      "epoch": 3.42,
-      "learning_rate": 1.203795860459386e-05,
-      "loss": 0.2499,
+      "epoch": 3.5009190182722456,
+      "grad_norm": 0.2956106960773468,
+      "learning_rate": 1.090483080453408e-05,
+      "loss": 0.3928,
       "step": 97140
     },
     {
-      "epoch": 3.42,
-      "learning_rate": 1.2035522779200725e-05,
-      "loss": 0.2616,
+      "epoch": 3.5010992179334703,
+      "grad_norm": 0.23157207667827606,
+      "learning_rate": 1.090242079433667e-05,
+      "loss": 0.4038,
       "step": 97145
     },
     {
-      "epoch": 3.42,
-      "learning_rate": 1.203308712214279e-05,
-      "loss": 0.2649,
+      "epoch": 3.501279417594695,
+      "grad_norm": 0.1804695725440979,
+      "learning_rate": 1.0900010976216046e-05,
+      "loss": 0.3982,
       "step": 97150
     },
     {
-      "epoch": 3.42,
-      "learning_rate": 1.2030651633451676e-05,
-      "loss": 0.2547,
+      "epoch": 3.50145961725592,
+      "grad_norm": 0.25014370679855347,
+      "learning_rate": 1.0897601350205042e-05,
+      "loss": 0.3616,
       "step": 97155
     },
     {
-      "epoch": 3.42,
-      "learning_rate": 1.2028216313158994e-05,
-      "loss": 0.2481,
+      "epoch": 3.501639816917144,
+      "grad_norm": 0.29266318678855896,
+      "learning_rate": 1.0895191916336477e-05,
+      "loss": 0.3894,
       "step": 97160
     },
     {
-      "epoch": 3.42,
-      "learning_rate": 1.2025781161296382e-05,
-      "loss": 0.2559,
+      "epoch": 3.501820016578369,
+      "grad_norm": 0.2770882248878479,
+      "learning_rate": 1.0892782674643207e-05,
+      "loss": 0.3823,
       "step": 97165
     },
     {
-      "epoch": 3.42,
-      "learning_rate": 1.2023346177895453e-05,
-      "loss": 0.2568,
+      "epoch": 3.5020002162395936,
+      "grad_norm": 0.2866399586200714,
+      "learning_rate": 1.0890373625158046e-05,
+      "loss": 0.41,
       "step": 97170
     },
     {
-      "epoch": 3.42,
-      "learning_rate": 1.2020911362987822e-05,
-      "loss": 0.2789,
+      "epoch": 3.502180415900818,
+      "grad_norm": 0.2779446542263031,
+      "learning_rate": 1.0887964767913796e-05,
+      "loss": 0.3801,
       "step": 97175
     },
     {
-      "epoch": 3.42,
-      "learning_rate": 1.2018476716605092e-05,
-      "loss": 0.2556,
+      "epoch": 3.5023606155620426,
+      "grad_norm": 0.2382117062807083,
+      "learning_rate": 1.0885556102943303e-05,
+      "loss": 0.3715,
       "step": 97180
     },
     {
-      "epoch": 3.42,
-      "learning_rate": 1.2016042238778896e-05,
-      "loss": 0.2463,
+      "epoch": 3.5025408152232673,
+      "grad_norm": 0.23480451107025146,
+      "learning_rate": 1.0883147630279367e-05,
+      "loss": 0.3713,
       "step": 97185
     },
     {
-      "epoch": 3.42,
-      "learning_rate": 1.2013607929540826e-05,
-      "loss": 0.2639,
+      "epoch": 3.502721014884492,
+      "grad_norm": 0.2746983468532562,
+      "learning_rate": 1.0880739349954829e-05,
+      "loss": 0.3993,
       "step": 97190
     },
     {
-      "epoch": 3.42,
-      "learning_rate": 1.2011173788922508e-05,
-      "loss": 0.259,
+      "epoch": 3.502901214545717,
+      "grad_norm": 0.27288299798965454,
+      "learning_rate": 1.0878331262002475e-05,
+      "loss": 0.3596,
       "step": 97195
     },
     {
-      "epoch": 3.42,
-      "learning_rate": 1.200873981695553e-05,
-      "loss": 0.2368,
+      "epoch": 3.5030814142069415,
+      "grad_norm": 0.25127673149108887,
+      "learning_rate": 1.0875923366455113e-05,
+      "loss": 0.3874,
       "step": 97200
     },
     {
-      "epoch": 3.42,
-      "learning_rate": 1.2006306013671515e-05,
-      "loss": 0.2639,
+      "epoch": 3.503261613868166,
+      "grad_norm": 0.29258519411087036,
+      "learning_rate": 1.0873515663345572e-05,
+      "loss": 0.411,
       "step": 97205
     },
     {
-      "epoch": 3.42,
-      "learning_rate": 1.2003872379102049e-05,
-      "loss": 0.2753,
+      "epoch": 3.5034418135293905,
+      "grad_norm": 0.2521027624607086,
+      "learning_rate": 1.0871108152706644e-05,
+      "loss": 0.3625,
       "step": 97210
     },
     {
-      "epoch": 3.42,
-      "learning_rate": 1.2001438913278739e-05,
-      "loss": 0.2755,
+      "epoch": 3.5036220131906153,
+      "grad_norm": 0.24310770630836487,
+      "learning_rate": 1.086870083457113e-05,
+      "loss": 0.3946,
       "step": 97215
     },
     {
-      "epoch": 3.42,
-      "learning_rate": 1.1999005616233166e-05,
-      "loss": 0.2661,
+      "epoch": 3.5038022128518396,
+      "grad_norm": 0.23873953521251678,
+      "learning_rate": 1.0866293708971834e-05,
+      "loss": 0.3985,
       "step": 97220
     },
     {
-      "epoch": 3.42,
-      "learning_rate": 1.199657248799695e-05,
-      "loss": 0.2821,
+      "epoch": 3.5039824125130643,
+      "grad_norm": 0.25413691997528076,
+      "learning_rate": 1.086388677594154e-05,
+      "loss": 0.3693,
       "step": 97225
     },
     {
-      "epoch": 3.42,
-      "learning_rate": 1.1994139528601667e-05,
-      "loss": 0.2348,
+      "epoch": 3.504162612174289,
+      "grad_norm": 0.24815772473812103,
+      "learning_rate": 1.086148003551306e-05,
+      "loss": 0.3974,
       "step": 97230
     },
     {
-      "epoch": 3.42,
-      "learning_rate": 1.1991706738078912e-05,
-      "loss": 0.2759,
+      "epoch": 3.5043428118355138,
+      "grad_norm": 0.2222224920988083,
+      "learning_rate": 1.085907348771918e-05,
+      "loss": 0.3328,
       "step": 97235
     },
     {
-      "epoch": 3.42,
-      "learning_rate": 1.1989274116460264e-05,
-      "loss": 0.2274,
+      "epoch": 3.5045230114967385,
+      "grad_norm": 0.18813829123973846,
+      "learning_rate": 1.0856667132592683e-05,
+      "loss": 0.3828,
       "step": 97240
     },
     {
-      "epoch": 3.42,
-      "learning_rate": 1.1986841663777326e-05,
-      "loss": 0.2486,
+      "epoch": 3.5047032111579632,
+      "grad_norm": 0.2500969171524048,
+      "learning_rate": 1.0854260970166358e-05,
+      "loss": 0.3769,
       "step": 97245
     },
     {
-      "epoch": 3.42,
-      "learning_rate": 1.1984409380061665e-05,
-      "loss": 0.2528,
+      "epoch": 3.5048834108191875,
+      "grad_norm": 0.3164437711238861,
+      "learning_rate": 1.085185500047299e-05,
+      "loss": 0.3717,
       "step": 97250
     },
     {
-      "epoch": 3.42,
-      "learning_rate": 1.1981977265344868e-05,
-      "loss": 0.2428,
+      "epoch": 3.5050636104804123,
+      "grad_norm": 0.23495697975158691,
+      "learning_rate": 1.084944922354535e-05,
+      "loss": 0.3808,
       "step": 97255
     },
     {
-      "epoch": 3.42,
-      "learning_rate": 1.1979545319658527e-05,
-      "loss": 0.2572,
+      "epoch": 3.505243810141637,
+      "grad_norm": 0.2900897264480591,
+      "learning_rate": 1.0847043639416233e-05,
+      "loss": 0.3814,
       "step": 97260
     },
     {
-      "epoch": 3.42,
-      "learning_rate": 1.197711354303421e-05,
-      "loss": 0.2478,
+      "epoch": 3.5054240098028617,
+      "grad_norm": 0.2515815794467926,
+      "learning_rate": 1.0844638248118405e-05,
+      "loss": 0.3928,
       "step": 97265
     },
     {
-      "epoch": 3.42,
-      "learning_rate": 1.1974681935503493e-05,
-      "loss": 0.2638,
+      "epoch": 3.505604209464086,
+      "grad_norm": 0.23748047649860382,
+      "learning_rate": 1.0842233049684642e-05,
+      "loss": 0.4071,
       "step": 97270
     },
     {
-      "epoch": 3.42,
-      "learning_rate": 1.1972250497097933e-05,
-      "loss": 0.2489,
+      "epoch": 3.5057844091253108,
+      "grad_norm": 0.2293088734149933,
+      "learning_rate": 1.0839828044147712e-05,
+      "loss": 0.4278,
       "step": 97275
     },
     {
-      "epoch": 3.42,
-      "learning_rate": 1.196981922784913e-05,
-      "loss": 0.2545,
+      "epoch": 3.5059646087865355,
+      "grad_norm": 0.2869160771369934,
+      "learning_rate": 1.0837423231540375e-05,
+      "loss": 0.3777,
       "step": 97280
     },
     {
-      "epoch": 3.42,
-      "learning_rate": 1.1967388127788634e-05,
-      "loss": 0.2686,
+      "epoch": 3.50614480844776,
+      "grad_norm": 0.2737312912940979,
+      "learning_rate": 1.083501861189542e-05,
+      "loss": 0.3707,
       "step": 97285
     },
     {
-      "epoch": 3.42,
-      "learning_rate": 1.1964957196948012e-05,
-      "loss": 0.2514,
+      "epoch": 3.506325008108985,
+      "grad_norm": 0.2278176099061966,
+      "learning_rate": 1.0832614185245587e-05,
+      "loss": 0.362,
       "step": 97290
     },
     {
-      "epoch": 3.42,
-      "learning_rate": 1.1962526435358825e-05,
-      "loss": 0.2623,
+      "epoch": 3.5065052077702092,
+      "grad_norm": 0.26734185218811035,
+      "learning_rate": 1.0830209951623635e-05,
+      "loss": 0.3986,
       "step": 97295
     },
     {
-      "epoch": 3.42,
-      "learning_rate": 1.1960095843052647e-05,
-      "loss": 0.2366,
+      "epoch": 3.506685407431434,
+      "grad_norm": 0.2176479548215866,
+      "learning_rate": 1.0827805911062336e-05,
+      "loss": 0.3585,
       "step": 97300
     },
     {
-      "epoch": 3.42,
-      "learning_rate": 1.1957665420061023e-05,
-      "loss": 0.2658,
+      "epoch": 3.5068656070926587,
+      "grad_norm": 0.22162985801696777,
+      "learning_rate": 1.082540206359444e-05,
+      "loss": 0.3828,
       "step": 97305
     },
     {
-      "epoch": 3.42,
-      "learning_rate": 1.1955235166415526e-05,
-      "loss": 0.2777,
+      "epoch": 3.5070458067538834,
+      "grad_norm": 0.21119628846645355,
+      "learning_rate": 1.0822998409252694e-05,
+      "loss": 0.3715,
       "step": 97310
     },
     {
-      "epoch": 3.42,
-      "learning_rate": 1.1952805082147697e-05,
-      "loss": 0.2421,
+      "epoch": 3.5072260064151077,
+      "grad_norm": 0.25533100962638855,
+      "learning_rate": 1.082059494806985e-05,
+      "loss": 0.4005,
       "step": 97315
     },
     {
-      "epoch": 3.42,
-      "learning_rate": 1.1950375167289102e-05,
-      "loss": 0.2495,
+      "epoch": 3.5074062060763325,
+      "grad_norm": 0.3065352439880371,
+      "learning_rate": 1.0818191680078648e-05,
+      "loss": 0.3915,
       "step": 97320
     },
     {
-      "epoch": 3.42,
-      "learning_rate": 1.1947945421871288e-05,
-      "loss": 0.237,
+      "epoch": 3.507586405737557,
+      "grad_norm": 0.23930329084396362,
+      "learning_rate": 1.0815788605311846e-05,
+      "loss": 0.3843,
       "step": 97325
     },
     {
-      "epoch": 3.42,
-      "learning_rate": 1.1945515845925798e-05,
-      "loss": 0.2598,
+      "epoch": 3.507766605398782,
+      "grad_norm": 0.2522366940975189,
+      "learning_rate": 1.0813385723802177e-05,
+      "loss": 0.3506,
       "step": 97330
     },
     {
-      "epoch": 3.42,
-      "learning_rate": 1.1943086439484174e-05,
-      "loss": 0.2453,
+      "epoch": 3.5079468050600067,
+      "grad_norm": 0.22099296748638153,
+      "learning_rate": 1.0810983035582384e-05,
+      "loss": 0.3578,
       "step": 97335
     },
     {
-      "epoch": 3.42,
-      "learning_rate": 1.1940657202577977e-05,
-      "loss": 0.271,
+      "epoch": 3.5081270047212314,
+      "grad_norm": 0.2783578038215637,
+      "learning_rate": 1.0808580540685198e-05,
+      "loss": 0.409,
       "step": 97340
     },
     {
-      "epoch": 3.42,
-      "learning_rate": 1.1938228135238741e-05,
-      "loss": 0.2707,
+      "epoch": 3.5083072043824557,
+      "grad_norm": 0.282536119222641,
+      "learning_rate": 1.0806178239143347e-05,
+      "loss": 0.392,
       "step": 97345
     },
     {
-      "epoch": 3.43,
-      "learning_rate": 1.1935799237498002e-05,
-      "loss": 0.2652,
+      "epoch": 3.5084874040436804,
+      "grad_norm": 0.2658419609069824,
+      "learning_rate": 1.0803776130989577e-05,
+      "loss": 0.4034,
       "step": 97350
     },
     {
-      "epoch": 3.43,
-      "learning_rate": 1.1933370509387293e-05,
-      "loss": 0.2505,
+      "epoch": 3.508667603704905,
+      "grad_norm": 0.23582491278648376,
+      "learning_rate": 1.0801374216256619e-05,
+      "loss": 0.3993,
       "step": 97355
     },
     {
-      "epoch": 3.43,
-      "learning_rate": 1.193142764905355e-05,
-      "loss": 0.2565,
+      "epoch": 3.5088478033661294,
+      "grad_norm": 0.22847378253936768,
+      "learning_rate": 1.0798972494977167e-05,
+      "loss": 0.3643,
       "step": 97360
     },
     {
-      "epoch": 3.43,
-      "learning_rate": 1.1928999226356375e-05,
-      "loss": 0.2557,
+      "epoch": 3.509028003027354,
+      "grad_norm": 0.24764804542064667,
+      "learning_rate": 1.0796570967183978e-05,
+      "loss": 0.3575,
       "step": 97365
     },
     {
-      "epoch": 3.43,
-      "learning_rate": 1.192657097337754e-05,
-      "loss": 0.2775,
+      "epoch": 3.509208202688579,
+      "grad_norm": 0.2016579508781433,
+      "learning_rate": 1.079416963290976e-05,
+      "loss": 0.3777,
       "step": 97370
     },
     {
-      "epoch": 3.43,
-      "learning_rate": 1.1924142890148566e-05,
-      "loss": 0.2629,
+      "epoch": 3.5093884023498036,
+      "grad_norm": 0.2673712968826294,
+      "learning_rate": 1.0791768492187218e-05,
+      "loss": 0.3758,
       "step": 97375
     },
     {
-      "epoch": 3.43,
-      "learning_rate": 1.1921714976700979e-05,
-      "loss": 0.2723,
+      "epoch": 3.5095686020110284,
+      "grad_norm": 0.23791760206222534,
+      "learning_rate": 1.07893675450491e-05,
+      "loss": 0.3433,
       "step": 97380
     },
     {
-      "epoch": 3.43,
-      "learning_rate": 1.1919287233066295e-05,
-      "loss": 0.2582,
+      "epoch": 3.509748801672253,
+      "grad_norm": 0.208527609705925,
+      "learning_rate": 1.0786966791528078e-05,
+      "loss": 0.3444,
       "step": 97385
     },
     {
-      "epoch": 3.43,
-      "learning_rate": 1.1916859659276055e-05,
-      "loss": 0.2901,
+      "epoch": 3.5099290013334774,
+      "grad_norm": 0.2786830961704254,
+      "learning_rate": 1.0784566231656893e-05,
+      "loss": 0.3838,
       "step": 97390
     },
     {
-      "epoch": 3.43,
-      "learning_rate": 1.191443225536176e-05,
-      "loss": 0.2493,
+      "epoch": 3.510109200994702,
+      "grad_norm": 0.2595866024494171,
+      "learning_rate": 1.078216586546824e-05,
+      "loss": 0.3882,
       "step": 97395
     },
     {
-      "epoch": 3.43,
-      "learning_rate": 1.191200502135495e-05,
-      "loss": 0.24,
+      "epoch": 3.510289400655927,
+      "grad_norm": 0.2380652129650116,
+      "learning_rate": 1.0779765692994826e-05,
+      "loss": 0.3906,
       "step": 97400
     },
     {
-      "epoch": 3.43,
-      "learning_rate": 1.1909577957287119e-05,
-      "loss": 0.2765,
+      "epoch": 3.510469600317151,
+      "grad_norm": 0.2838803231716156,
+      "learning_rate": 1.0777365714269347e-05,
+      "loss": 0.3842,
       "step": 97405
     },
     {
-      "epoch": 3.43,
-      "learning_rate": 1.1907151063189797e-05,
-      "loss": 0.2789,
+      "epoch": 3.510649799978376,
+      "grad_norm": 0.21511715650558472,
+      "learning_rate": 1.0774965929324502e-05,
+      "loss": 0.3744,
       "step": 97410
     },
     {
-      "epoch": 3.43,
-      "learning_rate": 1.1904724339094494e-05,
-      "loss": 0.2526,
+      "epoch": 3.5108299996396006,
+      "grad_norm": 0.22063112258911133,
+      "learning_rate": 1.0772566338193e-05,
+      "loss": 0.3887,
       "step": 97415
     },
     {
-      "epoch": 3.43,
-      "learning_rate": 1.1902297785032715e-05,
-      "loss": 0.2481,
+      "epoch": 3.5110101993008254,
+      "grad_norm": 0.21338684856891632,
+      "learning_rate": 1.0770166940907524e-05,
+      "loss": 0.3964,
       "step": 97420
     },
     {
-      "epoch": 3.43,
-      "learning_rate": 1.1899871401035958e-05,
-      "loss": 0.2529,
+      "epoch": 3.51119039896205,
+      "grad_norm": 0.21853932738304138,
+      "learning_rate": 1.0767767737500772e-05,
+      "loss": 0.3588,
       "step": 97425
     },
     {
-      "epoch": 3.43,
-      "learning_rate": 1.1897445187135747e-05,
-      "loss": 0.2799,
+      "epoch": 3.511370598623275,
+      "grad_norm": 0.20588691532611847,
+      "learning_rate": 1.0765368728005429e-05,
+      "loss": 0.4019,
       "step": 97430
     },
     {
-      "epoch": 3.43,
-      "learning_rate": 1.1895019143363576e-05,
-      "loss": 0.2723,
+      "epoch": 3.511550798284499,
+      "grad_norm": 0.27712467312812805,
+      "learning_rate": 1.076296991245418e-05,
+      "loss": 0.4012,
       "step": 97435
     },
     {
-      "epoch": 3.43,
-      "learning_rate": 1.1892593269750935e-05,
-      "loss": 0.2673,
+      "epoch": 3.511730997945724,
+      "grad_norm": 0.2563231289386749,
+      "learning_rate": 1.07605712908797e-05,
+      "loss": 0.3952,
       "step": 97440
     },
     {
-      "epoch": 3.43,
-      "learning_rate": 1.189016756632934e-05,
-      "loss": 0.2484,
+      "epoch": 3.5119111976069486,
+      "grad_norm": 0.1871892213821411,
+      "learning_rate": 1.0758172863314689e-05,
+      "loss": 0.3994,
       "step": 97445
     },
     {
-      "epoch": 3.43,
-      "learning_rate": 1.188774203313027e-05,
-      "loss": 0.2577,
+      "epoch": 3.512091397268173,
+      "grad_norm": 0.24314607679843903,
+      "learning_rate": 1.0755774629791815e-05,
+      "loss": 0.384,
       "step": 97450
     },
     {
-      "epoch": 3.43,
-      "learning_rate": 1.188531667018524e-05,
-      "loss": 0.2783,
+      "epoch": 3.5122715969293976,
+      "grad_norm": 0.23912304639816284,
+      "learning_rate": 1.0753376590343755e-05,
+      "loss": 0.3468,
       "step": 97455
     },
     {
-      "epoch": 3.43,
-      "learning_rate": 1.188289147752572e-05,
-      "loss": 0.2529,
+      "epoch": 3.5124517965906223,
+      "grad_norm": 0.25849324464797974,
+      "learning_rate": 1.0750978745003181e-05,
+      "loss": 0.395,
       "step": 97460
     },
     {
-      "epoch": 3.43,
-      "learning_rate": 1.188046645518322e-05,
-      "loss": 0.2641,
+      "epoch": 3.512631996251847,
+      "grad_norm": 0.21953187882900238,
+      "learning_rate": 1.0748581093802753e-05,
+      "loss": 0.3912,
       "step": 97465
     },
     {
-      "epoch": 3.43,
-      "learning_rate": 1.1878041603189213e-05,
-      "loss": 0.2413,
+      "epoch": 3.512812195913072,
+      "grad_norm": 0.231736421585083,
+      "learning_rate": 1.0746183636775167e-05,
+      "loss": 0.3748,
       "step": 97470
     },
     {
-      "epoch": 3.43,
-      "learning_rate": 1.1875616921575191e-05,
-      "loss": 0.2463,
+      "epoch": 3.5129923955742965,
+      "grad_norm": 0.27500176429748535,
+      "learning_rate": 1.0743786373953061e-05,
+      "loss": 0.3845,
       "step": 97475
     },
     {
-      "epoch": 3.43,
-      "learning_rate": 1.1873192410372628e-05,
-      "loss": 0.2614,
+      "epoch": 3.513172595235521,
+      "grad_norm": 0.2197420597076416,
+      "learning_rate": 1.0741389305369093e-05,
+      "loss": 0.3836,
       "step": 97480
     },
     {
-      "epoch": 3.43,
-      "learning_rate": 1.1870768069613014e-05,
-      "loss": 0.2712,
+      "epoch": 3.5133527948967456,
+      "grad_norm": 0.22646686434745789,
+      "learning_rate": 1.0738992431055948e-05,
+      "loss": 0.347,
       "step": 97485
     },
     {
-      "epoch": 3.43,
-      "learning_rate": 1.1868343899327827e-05,
-      "loss": 0.2498,
+      "epoch": 3.5135329945579703,
+      "grad_norm": 0.22990435361862183,
+      "learning_rate": 1.073659575104626e-05,
+      "loss": 0.4269,
       "step": 97490
     },
     {
-      "epoch": 3.43,
-      "learning_rate": 1.1865919899548542e-05,
-      "loss": 0.2691,
+      "epoch": 3.5137131942191946,
+      "grad_norm": 0.29172033071517944,
+      "learning_rate": 1.0734678546958963e-05,
+      "loss": 0.3727,
       "step": 97495
     },
     {
-      "epoch": 3.43,
-      "learning_rate": 1.1863496070306618e-05,
-      "loss": 0.2492,
+      "epoch": 3.5138933938804193,
+      "grad_norm": 0.24528679251670837,
+      "learning_rate": 1.0732282216777811e-05,
+      "loss": 0.3703,
       "step": 97500
     },
     {
-      "epoch": 3.43,
-      "eval_loss": 0.25291621685028076,
-      "eval_runtime": 10.5567,
-      "eval_samples_per_second": 9.473,
-      "eval_steps_per_second": 9.473,
+      "epoch": 3.5138933938804193,
+      "eval_loss": 0.4301924407482147,
+      "eval_runtime": 3.5317,
+      "eval_samples_per_second": 28.315,
+      "eval_steps_per_second": 7.079,
       "step": 97500
     },
     {
-      "epoch": 3.43,
-      "learning_rate": 1.1861072411633542e-05,
-      "loss": 0.2492,
+      "epoch": 3.514073593541644,
+      "grad_norm": 0.20823730528354645,
+      "learning_rate": 1.0729886080991553e-05,
+      "loss": 0.3724,
       "step": 97505
     },
     {
-      "epoch": 3.43,
-      "learning_rate": 1.1858648923560792e-05,
-      "loss": 0.2829,
+      "epoch": 3.514253793202869,
+      "grad_norm": 0.23355242609977722,
+      "learning_rate": 1.0727490139632824e-05,
+      "loss": 0.4024,
       "step": 97510
     },
     {
-      "epoch": 3.43,
-      "learning_rate": 1.1856225606119825e-05,
-      "loss": 0.2566,
+      "epoch": 3.5144339928640935,
+      "grad_norm": 0.2617993950843811,
+      "learning_rate": 1.0725094392734289e-05,
+      "loss": 0.3831,
       "step": 97515
     },
     {
-      "epoch": 3.43,
-      "learning_rate": 1.1853802459342098e-05,
-      "loss": 0.2599,
+      "epoch": 3.5146141925253183,
+      "grad_norm": 0.2531090974807739,
+      "learning_rate": 1.0722698840328576e-05,
+      "loss": 0.3876,
       "step": 97520
     },
     {
-      "epoch": 3.43,
-      "learning_rate": 1.185137948325909e-05,
-      "loss": 0.259,
+      "epoch": 3.5147943921865425,
+      "grad_norm": 0.24823562800884247,
+      "learning_rate": 1.0720303482448333e-05,
+      "loss": 0.3846,
       "step": 97525
     },
     {
-      "epoch": 3.43,
-      "learning_rate": 1.1848956677902254e-05,
-      "loss": 0.2482,
+      "epoch": 3.5149745918477673,
+      "grad_norm": 0.2218950390815735,
+      "learning_rate": 1.0717908319126185e-05,
+      "loss": 0.3883,
       "step": 97530
     },
     {
-      "epoch": 3.43,
-      "learning_rate": 1.184653404330305e-05,
-      "loss": 0.2563,
+      "epoch": 3.515154791508992,
+      "grad_norm": 0.23106147348880768,
+      "learning_rate": 1.0715513350394762e-05,
+      "loss": 0.4091,
       "step": 97535
     },
     {
-      "epoch": 3.43,
-      "learning_rate": 1.1844111579492923e-05,
-      "loss": 0.2484,
+      "epoch": 3.5153349911702167,
+      "grad_norm": 0.31263554096221924,
+      "learning_rate": 1.0713118576286716e-05,
+      "loss": 0.3521,
       "step": 97540
     },
     {
-      "epoch": 3.43,
-      "learning_rate": 1.1841689286503347e-05,
-      "loss": 0.2535,
+      "epoch": 3.515515190831441,
+      "grad_norm": 0.25022903084754944,
+      "learning_rate": 1.0710723996834671e-05,
+      "loss": 0.3958,
       "step": 97545
     },
     {
-      "epoch": 3.43,
-      "learning_rate": 1.1839267164365764e-05,
-      "loss": 0.2615,
+      "epoch": 3.5156953904926658,
+      "grad_norm": 0.25562384724617004,
+      "learning_rate": 1.0708329612071227e-05,
+      "loss": 0.3729,
       "step": 97550
     },
     {
-      "epoch": 3.43,
-      "learning_rate": 1.1836845213111614e-05,
-      "loss": 0.26,
+      "epoch": 3.5158755901538905,
+      "grad_norm": 0.25449755787849426,
+      "learning_rate": 1.0705935422029034e-05,
+      "loss": 0.3987,
       "step": 97555
     },
     {
-      "epoch": 3.43,
-      "learning_rate": 1.1834423432772354e-05,
-      "loss": 0.2566,
+      "epoch": 3.5160557898151152,
+      "grad_norm": 0.20006752014160156,
+      "learning_rate": 1.0703541426740697e-05,
+      "loss": 0.3625,
       "step": 97560
     },
     {
-      "epoch": 3.43,
-      "learning_rate": 1.1832001823379438e-05,
-      "loss": 0.2528,
+      "epoch": 3.51623598947634,
+      "grad_norm": 0.20985792577266693,
+      "learning_rate": 1.0701147626238856e-05,
+      "loss": 0.3947,
       "step": 97565
     },
     {
-      "epoch": 3.43,
-      "learning_rate": 1.18295803849643e-05,
-      "loss": 0.264,
+      "epoch": 3.5164161891375643,
+      "grad_norm": 0.24245929718017578,
+      "learning_rate": 1.0698754020556101e-05,
+      "loss": 0.3838,
       "step": 97570
     },
     {
-      "epoch": 3.43,
-      "learning_rate": 1.1827159117558373e-05,
-      "loss": 0.2678,
+      "epoch": 3.516596388798789,
+      "grad_norm": 0.23297759890556335,
+      "learning_rate": 1.0696360609725044e-05,
+      "loss": 0.3709,
       "step": 97575
     },
     {
-      "epoch": 3.43,
-      "learning_rate": 1.1824738021193113e-05,
-      "loss": 0.2706,
+      "epoch": 3.5167765884600137,
+      "grad_norm": 0.27019697427749634,
+      "learning_rate": 1.0693967393778315e-05,
+      "loss": 0.3778,
       "step": 97580
     },
     {
-      "epoch": 3.43,
-      "learning_rate": 1.1822317095899946e-05,
-      "loss": 0.2569,
+      "epoch": 3.5169567881212385,
+      "grad_norm": 0.26086610555648804,
+      "learning_rate": 1.0691574372748509e-05,
+      "loss": 0.393,
       "step": 97585
     },
     {
-      "epoch": 3.43,
-      "learning_rate": 1.1819896341710307e-05,
-      "loss": 0.2325,
+      "epoch": 3.5171369877824628,
+      "grad_norm": 0.20790447294712067,
+      "learning_rate": 1.0689181546668234e-05,
+      "loss": 0.342,
       "step": 97590
     },
     {
-      "epoch": 3.43,
-      "learning_rate": 1.1817475758655618e-05,
-      "loss": 0.261,
+      "epoch": 3.5173171874436875,
+      "grad_norm": 0.24701309204101562,
+      "learning_rate": 1.0686788915570088e-05,
+      "loss": 0.3964,
       "step": 97595
     },
     {
-      "epoch": 3.43,
-      "learning_rate": 1.1815055346767332e-05,
-      "loss": 0.2611,
+      "epoch": 3.517497387104912,
+      "grad_norm": 0.2746676802635193,
+      "learning_rate": 1.0684396479486664e-05,
+      "loss": 0.3973,
       "step": 97600
     },
     {
-      "epoch": 3.43,
-      "learning_rate": 1.181263510607686e-05,
-      "loss": 0.2512,
+      "epoch": 3.517677586766137,
+      "grad_norm": 0.3433469831943512,
+      "learning_rate": 1.0682004238450574e-05,
+      "loss": 0.3645,
       "step": 97605
     },
     {
-      "epoch": 3.43,
-      "learning_rate": 1.1810215036615622e-05,
-      "loss": 0.2693,
+      "epoch": 3.5178577864273617,
+      "grad_norm": 0.22473633289337158,
+      "learning_rate": 1.0679612192494403e-05,
+      "loss": 0.3702,
       "step": 97610
     },
     {
-      "epoch": 3.43,
-      "learning_rate": 1.1807795138415048e-05,
-      "loss": 0.2683,
+      "epoch": 3.5180379860885864,
+      "grad_norm": 0.26055383682250977,
+      "learning_rate": 1.0677220341650747e-05,
+      "loss": 0.3831,
       "step": 97615
     },
     {
-      "epoch": 3.43,
-      "learning_rate": 1.180537541150657e-05,
-      "loss": 0.2503,
+      "epoch": 3.5182181857498107,
+      "grad_norm": 0.22649826109409332,
+      "learning_rate": 1.067482868595219e-05,
+      "loss": 0.3515,
       "step": 97620
     },
     {
-      "epoch": 3.43,
-      "learning_rate": 1.1802955855921596e-05,
-      "loss": 0.2489,
+      "epoch": 3.5183983854110354,
+      "grad_norm": 0.24779903888702393,
+      "learning_rate": 1.067243722543131e-05,
+      "loss": 0.3627,
       "step": 97625
     },
     {
-      "epoch": 3.43,
-      "learning_rate": 1.1800536471691545e-05,
-      "loss": 0.2223,
+      "epoch": 3.51857858507226,
+      "grad_norm": 0.2685101330280304,
+      "learning_rate": 1.0670045960120707e-05,
+      "loss": 0.3862,
       "step": 97630
     },
     {
-      "epoch": 3.44,
-      "learning_rate": 1.1798117258847818e-05,
-      "loss": 0.2399,
+      "epoch": 3.5187587847334845,
+      "grad_norm": 0.2545015513896942,
+      "learning_rate": 1.0667654890052962e-05,
+      "loss": 0.3848,
       "step": 97635
     },
     {
-      "epoch": 3.44,
-      "learning_rate": 1.1795698217421849e-05,
-      "loss": 0.2515,
+      "epoch": 3.518938984394709,
+      "grad_norm": 0.2047766149044037,
+      "learning_rate": 1.0665264015260626e-05,
+      "loss": 0.3676,
       "step": 97640
     },
     {
-      "epoch": 3.44,
-      "learning_rate": 1.1793279347445036e-05,
-      "loss": 0.2619,
+      "epoch": 3.519119184055934,
+      "grad_norm": 0.24359673261642456,
+      "learning_rate": 1.0662873335776302e-05,
+      "loss": 0.3863,
       "step": 97645
     },
     {
-      "epoch": 3.44,
-      "learning_rate": 1.1790860648948787e-05,
-      "loss": 0.269,
+      "epoch": 3.5192993837171587,
+      "grad_norm": 0.2065386027097702,
+      "learning_rate": 1.0660482851632553e-05,
+      "loss": 0.402,
       "step": 97650
     },
     {
-      "epoch": 3.44,
-      "learning_rate": 1.1788442121964496e-05,
-      "loss": 0.2618,
+      "epoch": 3.5194795833783834,
+      "grad_norm": 0.3513331711292267,
+      "learning_rate": 1.065809256286194e-05,
+      "loss": 0.3652,
       "step": 97655
     },
     {
-      "epoch": 3.44,
-      "learning_rate": 1.178602376652359e-05,
-      "loss": 0.2519,
+      "epoch": 3.519659783039608,
+      "grad_norm": 0.23848022520542145,
+      "learning_rate": 1.0655702469497057e-05,
+      "loss": 0.3883,
       "step": 97660
     },
     {
-      "epoch": 3.44,
-      "learning_rate": 1.1783605582657444e-05,
-      "loss": 0.2638,
+      "epoch": 3.5198399827008324,
+      "grad_norm": 0.1824946254491806,
+      "learning_rate": 1.0653312571570434e-05,
+      "loss": 0.384,
       "step": 97665
     },
     {
-      "epoch": 3.44,
-      "learning_rate": 1.1781187570397483e-05,
-      "loss": 0.2562,
+      "epoch": 3.520020182362057,
+      "grad_norm": 0.28067702054977417,
+      "learning_rate": 1.0650922869114658e-05,
+      "loss": 0.3883,
       "step": 97670
     },
     {
-      "epoch": 3.44,
-      "learning_rate": 1.177876972977508e-05,
-      "loss": 0.2606,
+      "epoch": 3.520200382023282,
+      "grad_norm": 0.23458154499530792,
+      "learning_rate": 1.0648533362162277e-05,
+      "loss": 0.4112,
       "step": 97675
     },
     {
-      "epoch": 3.44,
-      "learning_rate": 1.1776352060821646e-05,
-      "loss": 0.254,
+      "epoch": 3.520380581684506,
+      "grad_norm": 0.2893427908420563,
+      "learning_rate": 1.0646144050745854e-05,
+      "loss": 0.3922,
       "step": 97680
     },
     {
-      "epoch": 3.44,
-      "learning_rate": 1.1773934563568568e-05,
-      "loss": 0.2534,
+      "epoch": 3.520560781345731,
+      "grad_norm": 0.17397251725196838,
+      "learning_rate": 1.0643754934897937e-05,
+      "loss": 0.3335,
       "step": 97685
     },
     {
-      "epoch": 3.44,
-      "learning_rate": 1.1771517238047223e-05,
-      "loss": 0.2525,
+      "epoch": 3.5207409810069556,
+      "grad_norm": 0.25946879386901855,
+      "learning_rate": 1.064136601465108e-05,
+      "loss": 0.3578,
       "step": 97690
     },
     {
-      "epoch": 3.44,
-      "learning_rate": 1.1769100084289017e-05,
-      "loss": 0.2712,
+      "epoch": 3.5209211806681804,
+      "grad_norm": 0.28153377771377563,
+      "learning_rate": 1.0638977290037825e-05,
+      "loss": 0.3993,
       "step": 97695
     },
     {
-      "epoch": 3.44,
-      "learning_rate": 1.1766683102325326e-05,
-      "loss": 0.2546,
+      "epoch": 3.521101380329405,
+      "grad_norm": 0.2764267325401306,
+      "learning_rate": 1.063658876109073e-05,
+      "loss": 0.3849,
       "step": 97700
     },
     {
-      "epoch": 3.44,
-      "learning_rate": 1.1764266292187538e-05,
-      "loss": 0.2382,
+      "epoch": 3.52128157999063,
+      "grad_norm": 0.26901406049728394,
+      "learning_rate": 1.0634200427842334e-05,
+      "loss": 0.3975,
       "step": 97705
     },
     {
-      "epoch": 3.44,
-      "learning_rate": 1.1761849653907017e-05,
-      "loss": 0.2577,
+      "epoch": 3.521461779651854,
+      "grad_norm": 0.22816415131092072,
+      "learning_rate": 1.0631812290325174e-05,
+      "loss": 0.3826,
       "step": 97710
     },
     {
-      "epoch": 3.44,
-      "learning_rate": 1.1759433187515163e-05,
-      "loss": 0.2438,
+      "epoch": 3.521641979313079,
+      "grad_norm": 0.22323520481586456,
+      "learning_rate": 1.062942434857179e-05,
+      "loss": 0.3701,
       "step": 97715
     },
     {
-      "epoch": 3.44,
-      "learning_rate": 1.1757016893043335e-05,
-      "loss": 0.2657,
+      "epoch": 3.5218221789743036,
+      "grad_norm": 0.18023869395256042,
+      "learning_rate": 1.062703660261471e-05,
+      "loss": 0.3809,
       "step": 97720
     },
     {
-      "epoch": 3.44,
-      "learning_rate": 1.1754600770522925e-05,
-      "loss": 0.2439,
+      "epoch": 3.522002378635528,
+      "grad_norm": 0.2793877124786377,
+      "learning_rate": 1.062464905248648e-05,
+      "loss": 0.3815,
       "step": 97725
     },
     {
-      "epoch": 3.44,
-      "learning_rate": 1.1752184819985285e-05,
-      "loss": 0.2766,
+      "epoch": 3.5221825782967526,
+      "grad_norm": 0.19938796758651733,
+      "learning_rate": 1.0622261698219634e-05,
+      "loss": 0.3712,
       "step": 97730
     },
     {
-      "epoch": 3.44,
-      "learning_rate": 1.1749769041461803e-05,
-      "loss": 0.2521,
+      "epoch": 3.5223627779579774,
+      "grad_norm": 0.21415963768959045,
+      "learning_rate": 1.0619874539846673e-05,
+      "loss": 0.3845,
       "step": 97735
     },
     {
-      "epoch": 3.44,
-      "learning_rate": 1.1747353434983841e-05,
-      "loss": 0.2627,
+      "epoch": 3.522542977619202,
+      "grad_norm": 0.2542286515235901,
+      "learning_rate": 1.0617487577400143e-05,
+      "loss": 0.3956,
       "step": 97740
     },
     {
-      "epoch": 3.44,
-      "learning_rate": 1.1744938000582756e-05,
-      "loss": 0.252,
+      "epoch": 3.522723177280427,
+      "grad_norm": 0.25228896737098694,
+      "learning_rate": 1.0615100810912551e-05,
+      "loss": 0.365,
       "step": 97745
     },
     {
-      "epoch": 3.44,
-      "learning_rate": 1.174252273828991e-05,
-      "loss": 0.2612,
+      "epoch": 3.5229033769416516,
+      "grad_norm": 0.24575848877429962,
+      "learning_rate": 1.0612714240416444e-05,
+      "loss": 0.3517,
       "step": 97750
     },
     {
-      "epoch": 3.44,
-      "learning_rate": 1.1740107648136677e-05,
-      "loss": 0.2575,
+      "epoch": 3.523083576602876,
+      "grad_norm": 0.2416459023952484,
+      "learning_rate": 1.0610327865944311e-05,
+      "loss": 0.3597,
       "step": 97755
     },
     {
-      "epoch": 3.44,
-      "learning_rate": 1.1737692730154409e-05,
-      "loss": 0.2804,
+      "epoch": 3.5232637762641006,
+      "grad_norm": 0.2579323351383209,
+      "learning_rate": 1.0607941687528669e-05,
+      "loss": 0.4051,
       "step": 97760
     },
     {
-      "epoch": 3.44,
-      "learning_rate": 1.173527798437446e-05,
-      "loss": 0.2669,
+      "epoch": 3.5234439759253253,
+      "grad_norm": 0.3103504478931427,
+      "learning_rate": 1.0605555705202041e-05,
+      "loss": 0.3566,
       "step": 97765
     },
     {
-      "epoch": 3.44,
-      "learning_rate": 1.1732863410828174e-05,
-      "loss": 0.2583,
+      "epoch": 3.52362417558655,
+      "grad_norm": 0.22384612262248993,
+      "learning_rate": 1.060316991899693e-05,
+      "loss": 0.3673,
       "step": 97770
     },
     {
-      "epoch": 3.44,
-      "learning_rate": 1.1730449009546915e-05,
-      "loss": 0.2429,
+      "epoch": 3.5238043752477743,
+      "grad_norm": 0.22075200080871582,
+      "learning_rate": 1.0600784328945843e-05,
+      "loss": 0.3654,
       "step": 97775
     },
     {
-      "epoch": 3.44,
-      "learning_rate": 1.1728034780562038e-05,
-      "loss": 0.2754,
+      "epoch": 3.523984574908999,
+      "grad_norm": 0.24411165714263916,
+      "learning_rate": 1.059839893508128e-05,
+      "loss": 0.3677,
       "step": 97780
     },
     {
-      "epoch": 3.44,
-      "learning_rate": 1.1725620723904877e-05,
-      "loss": 0.2643,
+      "epoch": 3.524164774570224,
+      "grad_norm": 0.2727641761302948,
+      "learning_rate": 1.0596013737435734e-05,
+      "loss": 0.3663,
       "step": 97785
     },
     {
-      "epoch": 3.44,
-      "learning_rate": 1.1723206839606787e-05,
-      "loss": 0.2565,
+      "epoch": 3.5243449742314485,
+      "grad_norm": 0.21201664209365845,
+      "learning_rate": 1.0593628736041722e-05,
+      "loss": 0.3997,
       "step": 97790
     },
     {
-      "epoch": 3.44,
-      "learning_rate": 1.172079312769911e-05,
-      "loss": 0.2546,
+      "epoch": 3.5245251738926733,
+      "grad_norm": 0.2514842450618744,
+      "learning_rate": 1.0591243930931729e-05,
+      "loss": 0.3711,
       "step": 97795
     },
     {
-      "epoch": 3.44,
-      "learning_rate": 1.1718379588213182e-05,
-      "loss": 0.2491,
+      "epoch": 3.5247053735538976,
+      "grad_norm": 0.21594151854515076,
+      "learning_rate": 1.0588859322138247e-05,
+      "loss": 0.3714,
       "step": 97800
     },
     {
-      "epoch": 3.44,
-      "learning_rate": 1.1715966221180333e-05,
-      "loss": 0.2714,
+      "epoch": 3.5248855732151223,
+      "grad_norm": 0.21471446752548218,
+      "learning_rate": 1.058647490969377e-05,
+      "loss": 0.3749,
       "step": 97805
     },
     {
-      "epoch": 3.44,
-      "learning_rate": 1.1713553026631918e-05,
-      "loss": 0.2475,
+      "epoch": 3.525065772876347,
+      "grad_norm": 0.2319725900888443,
+      "learning_rate": 1.0584090693630778e-05,
+      "loss": 0.3581,
       "step": 97810
     },
     {
-      "epoch": 3.44,
-      "learning_rate": 1.1711140004599262e-05,
-      "loss": 0.2586,
+      "epoch": 3.5252459725375718,
+      "grad_norm": 0.18940207362174988,
+      "learning_rate": 1.0581706673981753e-05,
+      "loss": 0.3634,
       "step": 97815
     },
     {
-      "epoch": 3.44,
-      "learning_rate": 1.1708727155113692e-05,
-      "loss": 0.2523,
+      "epoch": 3.525426172198796,
+      "grad_norm": 0.22552742063999176,
+      "learning_rate": 1.0579322850779188e-05,
+      "loss": 0.3836,
       "step": 97820
     },
     {
-      "epoch": 3.44,
-      "learning_rate": 1.1706314478206531e-05,
-      "loss": 0.2639,
+      "epoch": 3.525606371860021,
+      "grad_norm": 0.22318586707115173,
+      "learning_rate": 1.0576939224055563e-05,
+      "loss": 0.3915,
       "step": 97825
     },
     {
-      "epoch": 3.44,
-      "learning_rate": 1.1703901973909129e-05,
-      "loss": 0.2606,
+      "epoch": 3.5257865715212455,
+      "grad_norm": 0.22033105790615082,
+      "learning_rate": 1.0574555793843345e-05,
+      "loss": 0.3772,
       "step": 97830
     },
     {
-      "epoch": 3.44,
-      "learning_rate": 1.1701489642252786e-05,
-      "loss": 0.2492,
+      "epoch": 3.5259667711824703,
+      "grad_norm": 0.19937936961650848,
+      "learning_rate": 1.0572172560175011e-05,
+      "loss": 0.379,
       "step": 97835
     },
     {
-      "epoch": 3.44,
-      "learning_rate": 1.1699077483268844e-05,
-      "loss": 0.2723,
+      "epoch": 3.526146970843695,
+      "grad_norm": 0.19125044345855713,
+      "learning_rate": 1.0569789523083026e-05,
+      "loss": 0.4136,
       "step": 97840
     },
     {
-      "epoch": 3.44,
-      "learning_rate": 1.1696665496988606e-05,
-      "loss": 0.2448,
+      "epoch": 3.5263271705049197,
+      "grad_norm": 0.2663952708244324,
+      "learning_rate": 1.056740668259988e-05,
+      "loss": 0.3793,
       "step": 97845
     },
     {
-      "epoch": 3.44,
-      "learning_rate": 1.1694253683443413e-05,
-      "loss": 0.2654,
+      "epoch": 3.526507370166144,
+      "grad_norm": 0.21519611775875092,
+      "learning_rate": 1.0565024038758009e-05,
+      "loss": 0.3712,
       "step": 97850
     },
     {
-      "epoch": 3.44,
-      "learning_rate": 1.169184204266456e-05,
-      "loss": 0.2614,
+      "epoch": 3.5266875698273688,
+      "grad_norm": 0.2259238213300705,
+      "learning_rate": 1.0562641591589898e-05,
+      "loss": 0.3741,
       "step": 97855
     },
     {
-      "epoch": 3.44,
-      "learning_rate": 1.1689430574683375e-05,
-      "loss": 0.2569,
+      "epoch": 3.5268677694885935,
+      "grad_norm": 0.22167803347110748,
+      "learning_rate": 1.0560259341128e-05,
+      "loss": 0.3672,
       "step": 97860
     },
     {
-      "epoch": 3.44,
-      "learning_rate": 1.1687019279531147e-05,
-      "loss": 0.2485,
+      "epoch": 3.5270479691498178,
+      "grad_norm": 0.1899726837873459,
+      "learning_rate": 1.0557877287404774e-05,
+      "loss": 0.375,
       "step": 97865
     },
     {
-      "epoch": 3.44,
-      "learning_rate": 1.1684608157239214e-05,
-      "loss": 0.2755,
+      "epoch": 3.5272281688110425,
+      "grad_norm": 0.23604774475097656,
+      "learning_rate": 1.0555495430452673e-05,
+      "loss": 0.355,
       "step": 97870
     },
     {
-      "epoch": 3.44,
-      "learning_rate": 1.1682197207838866e-05,
-      "loss": 0.2449,
+      "epoch": 3.5274083684722672,
+      "grad_norm": 0.22828394174575806,
+      "learning_rate": 1.0553113770304152e-05,
+      "loss": 0.3715,
       "step": 97875
     },
     {
-      "epoch": 3.44,
-      "learning_rate": 1.1679786431361401e-05,
-      "loss": 0.2429,
+      "epoch": 3.527588568133492,
+      "grad_norm": 0.1918339878320694,
+      "learning_rate": 1.0550732306991648e-05,
+      "loss": 0.3772,
       "step": 97880
     },
     {
-      "epoch": 3.44,
-      "learning_rate": 1.1677375827838142e-05,
-      "loss": 0.2594,
+      "epoch": 3.5277687677947167,
+      "grad_norm": 0.21536006033420563,
+      "learning_rate": 1.0548351040547628e-05,
+      "loss": 0.348,
       "step": 97885
     },
     {
-      "epoch": 3.44,
-      "learning_rate": 1.1674965397300367e-05,
-      "loss": 0.2398,
+      "epoch": 3.5279489674559414,
+      "grad_norm": 0.2499334067106247,
+      "learning_rate": 1.0545969971004527e-05,
+      "loss": 0.3779,
       "step": 97890
     },
     {
-      "epoch": 3.44,
-      "learning_rate": 1.1672555139779392e-05,
-      "loss": 0.2532,
+      "epoch": 3.5281291671171657,
+      "grad_norm": 0.22823038697242737,
+      "learning_rate": 1.0543589098394784e-05,
+      "loss": 0.3632,
       "step": 97895
     },
     {
-      "epoch": 3.44,
-      "learning_rate": 1.16701450553065e-05,
-      "loss": 0.2591,
+      "epoch": 3.5283093667783905,
+      "grad_norm": 0.20469775795936584,
+      "learning_rate": 1.0541208422750846e-05,
+      "loss": 0.3779,
       "step": 97900
     },
     {
-      "epoch": 3.44,
-      "learning_rate": 1.1667735143912998e-05,
-      "loss": 0.2373,
+      "epoch": 3.528489566439615,
+      "grad_norm": 0.2718513309955597,
+      "learning_rate": 1.053882794410513e-05,
+      "loss": 0.3781,
       "step": 97905
     },
     {
-      "epoch": 3.44,
-      "learning_rate": 1.1665325405630167e-05,
-      "loss": 0.2712,
+      "epoch": 3.5286697661008395,
+      "grad_norm": 0.1913396418094635,
+      "learning_rate": 1.0536447662490097e-05,
+      "loss": 0.3742,
       "step": 97910
     },
     {
-      "epoch": 3.44,
-      "learning_rate": 1.16629158404893e-05,
-      "loss": 0.2522,
+      "epoch": 3.5288499657620642,
+      "grad_norm": 0.24252444505691528,
+      "learning_rate": 1.0534067577938172e-05,
+      "loss": 0.3849,
       "step": 97915
     },
     {
-      "epoch": 3.45,
-      "learning_rate": 1.1660506448521672e-05,
-      "loss": 0.2578,
+      "epoch": 3.529030165423289,
+      "grad_norm": 0.26426470279693604,
+      "learning_rate": 1.0531687690481757e-05,
+      "loss": 0.3961,
       "step": 97920
     },
     {
-      "epoch": 3.45,
-      "learning_rate": 1.1658097229758586e-05,
-      "loss": 0.2519,
+      "epoch": 3.5292103650845137,
+      "grad_norm": 0.1625565141439438,
+      "learning_rate": 1.0529308000153304e-05,
+      "loss": 0.3535,
       "step": 97925
     },
     {
-      "epoch": 3.45,
-      "learning_rate": 1.1655688184231314e-05,
-      "loss": 0.2566,
+      "epoch": 3.5293905647457384,
+      "grad_norm": 0.2652257978916168,
+      "learning_rate": 1.052692850698522e-05,
+      "loss": 0.3936,
       "step": 97930
     },
     {
-      "epoch": 3.45,
-      "learning_rate": 1.1653279311971135e-05,
-      "loss": 0.2403,
+      "epoch": 3.529570764406963,
+      "grad_norm": 0.2854976952075958,
+      "learning_rate": 1.052454921100995e-05,
+      "loss": 0.382,
       "step": 97935
     },
     {
-      "epoch": 3.45,
-      "learning_rate": 1.1650870613009321e-05,
-      "loss": 0.2754,
+      "epoch": 3.5297509640681874,
+      "grad_norm": 0.24219295382499695,
+      "learning_rate": 1.0522170112259887e-05,
+      "loss": 0.3635,
       "step": 97940
     },
     {
-      "epoch": 3.45,
-      "learning_rate": 1.1648462087377154e-05,
-      "loss": 0.2559,
+      "epoch": 3.529931163729412,
+      "grad_norm": 0.19204115867614746,
+      "learning_rate": 1.0519791210767446e-05,
+      "loss": 0.3773,
       "step": 97945
     },
     {
-      "epoch": 3.45,
-      "learning_rate": 1.1646053735105919e-05,
-      "loss": 0.2337,
+      "epoch": 3.530111363390637,
+      "grad_norm": 0.30104586482048035,
+      "learning_rate": 1.0517412506565052e-05,
+      "loss": 0.4092,
       "step": 97950
     },
     {
-      "epoch": 3.45,
-      "learning_rate": 1.1643645556226876e-05,
-      "loss": 0.2396,
+      "epoch": 3.530291563051861,
+      "grad_norm": 0.26225852966308594,
+      "learning_rate": 1.0515033999685109e-05,
+      "loss": 0.3847,
       "step": 97955
     },
     {
-      "epoch": 3.45,
-      "learning_rate": 1.164123755077128e-05,
-      "loss": 0.2341,
+      "epoch": 3.530471762713086,
+      "grad_norm": 0.2835884690284729,
+      "learning_rate": 1.0512655690160025e-05,
+      "loss": 0.4016,
       "step": 97960
     },
     {
-      "epoch": 3.45,
-      "learning_rate": 1.1638829718770421e-05,
-      "loss": 0.241,
+      "epoch": 3.5306519623743107,
+      "grad_norm": 0.21307377517223358,
+      "learning_rate": 1.05102775780222e-05,
+      "loss": 0.3708,
       "step": 97965
     },
     {
-      "epoch": 3.45,
-      "learning_rate": 1.1636422060255555e-05,
-      "loss": 0.2725,
+      "epoch": 3.5308321620355354,
+      "grad_norm": 0.22064463794231415,
+      "learning_rate": 1.050789966330403e-05,
+      "loss": 0.4074,
       "step": 97970
     },
     {
-      "epoch": 3.45,
-      "learning_rate": 1.1634014575257943e-05,
-      "loss": 0.2507,
+      "epoch": 3.53101236169676,
+      "grad_norm": 0.22840745747089386,
+      "learning_rate": 1.050552194603793e-05,
+      "loss": 0.4148,
       "step": 97975
     },
     {
-      "epoch": 3.45,
-      "learning_rate": 1.163160726380883e-05,
-      "loss": 0.2498,
+      "epoch": 3.531192561357985,
+      "grad_norm": 0.22727341949939728,
+      "learning_rate": 1.050314442625629e-05,
+      "loss": 0.3656,
       "step": 97980
     },
     {
-      "epoch": 3.45,
-      "learning_rate": 1.1629200125939498e-05,
-      "loss": 0.2688,
+      "epoch": 3.531372761019209,
+      "grad_norm": 0.22287560999393463,
+      "learning_rate": 1.0500767103991496e-05,
+      "loss": 0.4111,
       "step": 97985
     },
     {
-      "epoch": 3.45,
-      "learning_rate": 1.1626793161681191e-05,
-      "loss": 0.2578,
+      "epoch": 3.531552960680434,
+      "grad_norm": 0.22651442885398865,
+      "learning_rate": 1.0498389979275947e-05,
+      "loss": 0.3876,
       "step": 97990
     },
     {
-      "epoch": 3.45,
-      "learning_rate": 1.162438637106515e-05,
-      "loss": 0.2772,
+      "epoch": 3.5317331603416586,
+      "grad_norm": 0.21187719702720642,
+      "learning_rate": 1.0496013052142027e-05,
+      "loss": 0.3932,
       "step": 97995
     },
     {
-      "epoch": 3.45,
-      "learning_rate": 1.162197975412264e-05,
-      "loss": 0.2662,
+      "epoch": 3.531913360002883,
+      "grad_norm": 0.23431113362312317,
+      "learning_rate": 1.0493636322622108e-05,
+      "loss": 0.3845,
       "step": 98000
     },
     {
-      "epoch": 3.45,
-      "eval_loss": 0.25310197472572327,
-      "eval_runtime": 10.5476,
-      "eval_samples_per_second": 9.481,
-      "eval_steps_per_second": 9.481,
+      "epoch": 3.531913360002883,
+      "eval_loss": 0.42959845066070557,
+      "eval_runtime": 3.5275,
+      "eval_samples_per_second": 28.349,
+      "eval_steps_per_second": 7.087,
       "step": 98000
     },
     {
-      "epoch": 3.45,
-      "learning_rate": 1.1619573310884913e-05,
-      "loss": 0.2549,
+      "epoch": 3.5320935596641077,
+      "grad_norm": 0.28880631923675537,
+      "learning_rate": 1.0491259790748597e-05,
+      "loss": 0.3756,
       "step": 98005
     },
     {
-      "epoch": 3.45,
-      "learning_rate": 1.1617167041383208e-05,
-      "loss": 0.2417,
+      "epoch": 3.5322737593253324,
+      "grad_norm": 0.2546299397945404,
+      "learning_rate": 1.0488883456553861e-05,
+      "loss": 0.3563,
       "step": 98010
     },
     {
-      "epoch": 3.45,
-      "learning_rate": 1.1614760945648765e-05,
-      "loss": 0.2693,
+      "epoch": 3.532453958986557,
+      "grad_norm": 0.2502884566783905,
+      "learning_rate": 1.0486507320070279e-05,
+      "loss": 0.362,
       "step": 98015
     },
     {
-      "epoch": 3.45,
-      "learning_rate": 1.1612355023712834e-05,
-      "loss": 0.2729,
+      "epoch": 3.532634158647782,
+      "grad_norm": 0.23691552877426147,
+      "learning_rate": 1.0484131381330228e-05,
+      "loss": 0.3751,
       "step": 98020
     },
     {
-      "epoch": 3.45,
-      "learning_rate": 1.1609949275606651e-05,
-      "loss": 0.2596,
+      "epoch": 3.5328143583090066,
+      "grad_norm": 0.2482050359249115,
+      "learning_rate": 1.0481755640366065e-05,
+      "loss": 0.3846,
       "step": 98025
     },
     {
-      "epoch": 3.45,
-      "learning_rate": 1.1607543701361455e-05,
-      "loss": 0.2607,
+      "epoch": 3.532994557970231,
+      "grad_norm": 0.24961546063423157,
+      "learning_rate": 1.0479380097210187e-05,
+      "loss": 0.3872,
       "step": 98030
     },
     {
-      "epoch": 3.45,
-      "learning_rate": 1.1605138301008466e-05,
-      "loss": 0.2725,
+      "epoch": 3.5331747576314556,
+      "grad_norm": 0.17989207804203033,
+      "learning_rate": 1.0477004751894929e-05,
+      "loss": 0.3843,
       "step": 98035
     },
     {
-      "epoch": 3.45,
-      "learning_rate": 1.1602733074578939e-05,
-      "loss": 0.263,
+      "epoch": 3.5333549572926803,
+      "grad_norm": 0.2397972047328949,
+      "learning_rate": 1.0474629604452676e-05,
+      "loss": 0.3869,
       "step": 98040
     },
     {
-      "epoch": 3.45,
-      "learning_rate": 1.1600328022104095e-05,
-      "loss": 0.2431,
+      "epoch": 3.533535156953905,
+      "grad_norm": 0.272879958152771,
+      "learning_rate": 1.0472254654915784e-05,
+      "loss": 0.407,
       "step": 98045
     },
     {
-      "epoch": 3.45,
-      "learning_rate": 1.1597923143615152e-05,
-      "loss": 0.2503,
+      "epoch": 3.5337153566151294,
+      "grad_norm": 0.25433728098869324,
+      "learning_rate": 1.046987990331661e-05,
+      "loss": 0.3856,
       "step": 98050
     },
     {
-      "epoch": 3.45,
-      "learning_rate": 1.1595518439143344e-05,
-      "loss": 0.2678,
+      "epoch": 3.533895556276354,
+      "grad_norm": 0.2679862082004547,
+      "learning_rate": 1.046750534968751e-05,
+      "loss": 0.4406,
       "step": 98055
     },
     {
-      "epoch": 3.45,
-      "learning_rate": 1.1593113908719902e-05,
-      "loss": 0.2826,
+      "epoch": 3.534075755937579,
+      "grad_norm": 0.18556173145771027,
+      "learning_rate": 1.0465130994060831e-05,
+      "loss": 0.3705,
       "step": 98060
     },
     {
-      "epoch": 3.45,
-      "learning_rate": 1.1590709552376042e-05,
-      "loss": 0.232,
+      "epoch": 3.5342559555988036,
+      "grad_norm": 0.23661454021930695,
+      "learning_rate": 1.0462756836468924e-05,
+      "loss": 0.3722,
       "step": 98065
     },
     {
-      "epoch": 3.45,
-      "learning_rate": 1.1588305370142982e-05,
-      "loss": 0.2463,
+      "epoch": 3.5344361552600283,
+      "grad_norm": 0.21671940386295319,
+      "learning_rate": 1.046038287694415e-05,
+      "loss": 0.3898,
       "step": 98070
     },
     {
-      "epoch": 3.45,
-      "learning_rate": 1.1585901362051929e-05,
-      "loss": 0.3027,
+      "epoch": 3.5346163549212526,
+      "grad_norm": 0.21144139766693115,
+      "learning_rate": 1.0458009115518841e-05,
+      "loss": 0.364,
       "step": 98075
     },
     {
-      "epoch": 3.45,
-      "learning_rate": 1.1583497528134116e-05,
-      "loss": 0.2829,
+      "epoch": 3.5347965545824773,
+      "grad_norm": 0.2512389123439789,
+      "learning_rate": 1.0455635552225343e-05,
+      "loss": 0.3769,
       "step": 98080
     },
     {
-      "epoch": 3.45,
-      "learning_rate": 1.1581093868420745e-05,
-      "loss": 0.2723,
+      "epoch": 3.534976754243702,
+      "grad_norm": 0.2403516322374344,
+      "learning_rate": 1.0453262187095996e-05,
+      "loss": 0.3693,
       "step": 98085
     },
     {
-      "epoch": 3.45,
-      "learning_rate": 1.1578690382943027e-05,
-      "loss": 0.2583,
+      "epoch": 3.535156953904927,
+      "grad_norm": 0.2084726095199585,
+      "learning_rate": 1.0450889020163126e-05,
+      "loss": 0.3775,
       "step": 98090
     },
     {
-      "epoch": 3.45,
-      "learning_rate": 1.1576287071732158e-05,
-      "loss": 0.2432,
+      "epoch": 3.535337153566151,
+      "grad_norm": 0.2216814011335373,
+      "learning_rate": 1.0448516051459087e-05,
+      "loss": 0.4085,
       "step": 98095
     },
     {
-      "epoch": 3.45,
-      "learning_rate": 1.1573883934819369e-05,
-      "loss": 0.2546,
+      "epoch": 3.535517353227376,
+      "grad_norm": 0.2486451119184494,
+      "learning_rate": 1.0446143281016205e-05,
+      "loss": 0.3845,
       "step": 98100
     },
     {
-      "epoch": 3.45,
-      "learning_rate": 1.1571480972235834e-05,
-      "loss": 0.2613,
+      "epoch": 3.5356975528886005,
+      "grad_norm": 0.22463186085224152,
+      "learning_rate": 1.0443770708866787e-05,
+      "loss": 0.3779,
       "step": 98105
     },
     {
-      "epoch": 3.45,
-      "learning_rate": 1.1569078184012782e-05,
-      "loss": 0.2659,
+      "epoch": 3.5358777525498253,
+      "grad_norm": 0.232293039560318,
+      "learning_rate": 1.0441398335043184e-05,
+      "loss": 0.3514,
       "step": 98110
     },
     {
-      "epoch": 3.45,
-      "learning_rate": 1.156667557018139e-05,
-      "loss": 0.2676,
+      "epoch": 3.53605795221105,
+      "grad_norm": 0.29086434841156006,
+      "learning_rate": 1.0439026159577703e-05,
+      "loss": 0.3531,
       "step": 98115
     },
     {
-      "epoch": 3.45,
-      "learning_rate": 1.1564273130772868e-05,
-      "loss": 0.2521,
+      "epoch": 3.5362381518722747,
+      "grad_norm": 0.2319757342338562,
+      "learning_rate": 1.0436654182502678e-05,
+      "loss": 0.4037,
       "step": 98120
     },
     {
-      "epoch": 3.45,
-      "learning_rate": 1.1561870865818408e-05,
-      "loss": 0.2378,
+      "epoch": 3.536418351533499,
+      "grad_norm": 0.23496223986148834,
+      "learning_rate": 1.0434282403850429e-05,
+      "loss": 0.3831,
       "step": 98125
     },
     {
-      "epoch": 3.45,
-      "learning_rate": 1.1559468775349197e-05,
-      "loss": 0.2688,
+      "epoch": 3.5365985511947238,
+      "grad_norm": 0.21282298862934113,
+      "learning_rate": 1.0431910823653244e-05,
+      "loss": 0.3574,
       "step": 98130
     },
     {
-      "epoch": 3.45,
-      "learning_rate": 1.155706685939642e-05,
-      "loss": 0.2599,
+      "epoch": 3.5367787508559485,
+      "grad_norm": 0.24302206933498383,
+      "learning_rate": 1.0429539441943464e-05,
+      "loss": 0.3586,
       "step": 98135
     },
     {
-      "epoch": 3.45,
-      "learning_rate": 1.1554665117991275e-05,
-      "loss": 0.2635,
+      "epoch": 3.536958950517173,
+      "grad_norm": 0.2705324590206146,
+      "learning_rate": 1.0427168258753386e-05,
+      "loss": 0.3779,
       "step": 98140
     },
     {
-      "epoch": 3.45,
-      "learning_rate": 1.1552263551164947e-05,
-      "loss": 0.245,
+      "epoch": 3.5371391501783975,
+      "grad_norm": 0.23419471085071564,
+      "learning_rate": 1.0424797274115322e-05,
+      "loss": 0.3901,
       "step": 98145
     },
     {
-      "epoch": 3.45,
-      "learning_rate": 1.1549862158948604e-05,
-      "loss": 0.2681,
+      "epoch": 3.5373193498396223,
+      "grad_norm": 0.20766755938529968,
+      "learning_rate": 1.0422426488061572e-05,
+      "loss": 0.386,
       "step": 98150
     },
     {
-      "epoch": 3.45,
-      "learning_rate": 1.1547460941373445e-05,
-      "loss": 0.2921,
+      "epoch": 3.537499549500847,
+      "grad_norm": 0.23583157360553741,
+      "learning_rate": 1.042005590062443e-05,
+      "loss": 0.3403,
       "step": 98155
     },
     {
-      "epoch": 3.45,
-      "learning_rate": 1.1545059898470634e-05,
-      "loss": 0.2582,
+      "epoch": 3.5376797491620717,
+      "grad_norm": 0.2042224109172821,
+      "learning_rate": 1.0417685511836212e-05,
+      "loss": 0.3912,
       "step": 98160
     },
     {
-      "epoch": 3.45,
-      "learning_rate": 1.1542659030271358e-05,
-      "loss": 0.2489,
+      "epoch": 3.5378599488232965,
+      "grad_norm": 0.2094784826040268,
+      "learning_rate": 1.0415315321729205e-05,
+      "loss": 0.3683,
       "step": 98165
     },
     {
-      "epoch": 3.45,
-      "learning_rate": 1.1540258336806778e-05,
-      "loss": 0.2556,
+      "epoch": 3.5380401484845208,
+      "grad_norm": 0.24115018546581268,
+      "learning_rate": 1.0412945330335705e-05,
+      "loss": 0.3984,
       "step": 98170
     },
     {
-      "epoch": 3.45,
-      "learning_rate": 1.1537857818108082e-05,
-      "loss": 0.2798,
+      "epoch": 3.5382203481457455,
+      "grad_norm": 0.216557577252388,
+      "learning_rate": 1.0410575537688e-05,
+      "loss": 0.3695,
       "step": 98175
     },
     {
-      "epoch": 3.45,
-      "learning_rate": 1.1535457474206433e-05,
-      "loss": 0.2701,
+      "epoch": 3.53840054780697,
+      "grad_norm": 0.22672514617443085,
+      "learning_rate": 1.0408205943818378e-05,
+      "loss": 0.3619,
       "step": 98180
     },
     {
-      "epoch": 3.45,
-      "learning_rate": 1.1533057305132995e-05,
-      "loss": 0.2527,
+      "epoch": 3.5385807474681945,
+      "grad_norm": 0.23298558592796326,
+      "learning_rate": 1.0405836548759117e-05,
+      "loss": 0.414,
       "step": 98185
     },
     {
-      "epoch": 3.45,
-      "learning_rate": 1.1530657310918922e-05,
-      "loss": 0.2337,
+      "epoch": 3.5387609471294192,
+      "grad_norm": 0.20260320603847504,
+      "learning_rate": 1.0403467352542515e-05,
+      "loss": 0.3736,
       "step": 98190
     },
     {
-      "epoch": 3.45,
-      "learning_rate": 1.1528257491595395e-05,
-      "loss": 0.24,
+      "epoch": 3.538941146790644,
+      "grad_norm": 0.21936598420143127,
+      "learning_rate": 1.0401098355200848e-05,
+      "loss": 0.3585,
       "step": 98195
     },
     {
-      "epoch": 3.45,
-      "learning_rate": 1.1525857847193568e-05,
-      "loss": 0.2533,
+      "epoch": 3.5391213464518687,
+      "grad_norm": 0.2278914600610733,
+      "learning_rate": 1.0398729556766385e-05,
+      "loss": 0.3446,
       "step": 98200
     },
     {
-      "epoch": 3.46,
-      "learning_rate": 1.1523458377744598e-05,
-      "loss": 0.2501,
+      "epoch": 3.5393015461130934,
+      "grad_norm": 0.24078984558582306,
+      "learning_rate": 1.0396360957271405e-05,
+      "loss": 0.4154,
       "step": 98205
     },
     {
-      "epoch": 3.46,
-      "learning_rate": 1.1521059083279629e-05,
-      "loss": 0.2454,
+      "epoch": 3.539481745774318,
+      "grad_norm": 0.24573035538196564,
+      "learning_rate": 1.0393992556748172e-05,
+      "loss": 0.3924,
       "step": 98210
     },
     {
-      "epoch": 3.46,
-      "learning_rate": 1.1518659963829826e-05,
-      "loss": 0.2358,
+      "epoch": 3.5396619454355425,
+      "grad_norm": 0.25402477383613586,
+      "learning_rate": 1.0391624355228982e-05,
+      "loss": 0.3632,
       "step": 98215
     },
     {
-      "epoch": 3.46,
-      "learning_rate": 1.1516261019426345e-05,
-      "loss": 0.2607,
+      "epoch": 3.539842145096767,
+      "grad_norm": 0.19696156680583954,
+      "learning_rate": 1.0389256352746063e-05,
+      "loss": 0.3623,
       "step": 98220
     },
     {
-      "epoch": 3.46,
-      "learning_rate": 1.1513862250100333e-05,
-      "loss": 0.2938,
+      "epoch": 3.540022344757992,
+      "grad_norm": 0.24171549081802368,
+      "learning_rate": 1.0386888549331706e-05,
+      "loss": 0.4052,
       "step": 98225
     },
     {
-      "epoch": 3.46,
-      "learning_rate": 1.1511463655882917e-05,
-      "loss": 0.2692,
+      "epoch": 3.5402025444192162,
+      "grad_norm": 0.25626784563064575,
+      "learning_rate": 1.0384520945018164e-05,
+      "loss": 0.3945,
       "step": 98230
     },
     {
-      "epoch": 3.46,
-      "learning_rate": 1.150906523680527e-05,
-      "loss": 0.2624,
+      "epoch": 3.540382744080441,
+      "grad_norm": 0.2716904282569885,
+      "learning_rate": 1.0382153539837686e-05,
+      "loss": 0.3824,
       "step": 98235
     },
     {
-      "epoch": 3.46,
-      "learning_rate": 1.1506666992898519e-05,
-      "loss": 0.2545,
+      "epoch": 3.5405629437416657,
+      "grad_norm": 0.20084236562252045,
+      "learning_rate": 1.0379786333822552e-05,
+      "loss": 0.3576,
       "step": 98240
     },
     {
-      "epoch": 3.46,
-      "learning_rate": 1.1504268924193803e-05,
-      "loss": 0.2491,
+      "epoch": 3.5407431434028904,
+      "grad_norm": 0.21598944067955017,
+      "learning_rate": 1.037741932700499e-05,
+      "loss": 0.3622,
       "step": 98245
     },
     {
-      "epoch": 3.46,
-      "learning_rate": 1.1501871030722256e-05,
-      "loss": 0.2624,
+      "epoch": 3.540923343064115,
+      "grad_norm": 0.2069658786058426,
+      "learning_rate": 1.037505251941725e-05,
+      "loss": 0.375,
       "step": 98250
     },
     {
-      "epoch": 3.46,
-      "learning_rate": 1.1499473312515025e-05,
-      "loss": 0.2537,
+      "epoch": 3.54110354272534,
+      "grad_norm": 0.2307698279619217,
+      "learning_rate": 1.0372685911091598e-05,
+      "loss": 0.3512,
       "step": 98255
     },
     {
-      "epoch": 3.46,
-      "learning_rate": 1.1497075769603233e-05,
-      "loss": 0.2647,
+      "epoch": 3.541283742386564,
+      "grad_norm": 0.23482072353363037,
+      "learning_rate": 1.0370319502060267e-05,
+      "loss": 0.3773,
       "step": 98260
     },
     {
-      "epoch": 3.46,
-      "learning_rate": 1.1494678402018008e-05,
-      "loss": 0.2603,
+      "epoch": 3.541463942047789,
+      "grad_norm": 0.23283249139785767,
+      "learning_rate": 1.03679532923555e-05,
+      "loss": 0.4013,
       "step": 98265
     },
     {
-      "epoch": 3.46,
-      "learning_rate": 1.149228120979049e-05,
-      "loss": 0.2296,
+      "epoch": 3.5416441417090136,
+      "grad_norm": 0.23621907830238342,
+      "learning_rate": 1.0365587282009539e-05,
+      "loss": 0.404,
       "step": 98270
     },
     {
-      "epoch": 3.46,
-      "learning_rate": 1.148988419295179e-05,
-      "loss": 0.274,
+      "epoch": 3.5418243413702384,
+      "grad_norm": 0.2579342722892761,
+      "learning_rate": 1.0363221471054607e-05,
+      "loss": 0.4041,
       "step": 98275
     },
     {
-      "epoch": 3.46,
-      "learning_rate": 1.1487487351533051e-05,
-      "loss": 0.2443,
+      "epoch": 3.5420045410314627,
+      "grad_norm": 0.2534416615962982,
+      "learning_rate": 1.0360855859522958e-05,
+      "loss": 0.3824,
       "step": 98280
     },
     {
-      "epoch": 3.46,
-      "learning_rate": 1.1485090685565375e-05,
-      "loss": 0.2331,
+      "epoch": 3.5421847406926874,
+      "grad_norm": 0.20750831067562103,
+      "learning_rate": 1.0358490447446815e-05,
+      "loss": 0.3827,
       "step": 98285
     },
     {
-      "epoch": 3.46,
-      "learning_rate": 1.1482694195079897e-05,
-      "loss": 0.2856,
+      "epoch": 3.542364940353912,
+      "grad_norm": 0.22172388434410095,
+      "learning_rate": 1.0356125234858405e-05,
+      "loss": 0.3838,
       "step": 98290
     },
     {
-      "epoch": 3.46,
-      "learning_rate": 1.1480297880107726e-05,
-      "loss": 0.2525,
+      "epoch": 3.542545140015137,
+      "grad_norm": 0.2661237120628357,
+      "learning_rate": 1.0353760221789951e-05,
+      "loss": 0.3559,
       "step": 98295
     },
     {
-      "epoch": 3.46,
-      "learning_rate": 1.147790174067998e-05,
-      "loss": 0.2415,
+      "epoch": 3.5427253396763616,
+      "grad_norm": 0.24932733178138733,
+      "learning_rate": 1.035139540827367e-05,
+      "loss": 0.3511,
       "step": 98300
     },
     {
-      "epoch": 3.46,
-      "learning_rate": 1.1475505776827757e-05,
-      "loss": 0.2257,
+      "epoch": 3.542905539337586,
+      "grad_norm": 0.21312791109085083,
+      "learning_rate": 1.0349030794341802e-05,
+      "loss": 0.3607,
       "step": 98305
     },
     {
-      "epoch": 3.46,
-      "learning_rate": 1.147310998858219e-05,
-      "loss": 0.2586,
+      "epoch": 3.5430857389988106,
+      "grad_norm": 0.2608811855316162,
+      "learning_rate": 1.0346666380026559e-05,
+      "loss": 0.3928,
       "step": 98310
     },
     {
-      "epoch": 3.46,
-      "learning_rate": 1.1470714375974373e-05,
-      "loss": 0.2403,
+      "epoch": 3.5432659386600354,
+      "grad_norm": 0.21913045644760132,
+      "learning_rate": 1.0344302165360134e-05,
+      "loss": 0.4061,
       "step": 98315
     },
     {
-      "epoch": 3.46,
-      "learning_rate": 1.1468318939035405e-05,
-      "loss": 0.2549,
+      "epoch": 3.54344613832126,
+      "grad_norm": 0.23208434879779816,
+      "learning_rate": 1.0341938150374761e-05,
+      "loss": 0.3448,
       "step": 98320
     },
     {
-      "epoch": 3.46,
-      "learning_rate": 1.146592367779641e-05,
-      "loss": 0.2518,
+      "epoch": 3.5436263379824844,
+      "grad_norm": 0.22456184029579163,
+      "learning_rate": 1.0339574335102645e-05,
+      "loss": 0.3587,
       "step": 98325
     },
     {
-      "epoch": 3.46,
-      "learning_rate": 1.1463528592288464e-05,
-      "loss": 0.2237,
+      "epoch": 3.543806537643709,
+      "grad_norm": 0.23729225993156433,
+      "learning_rate": 1.033721071957599e-05,
+      "loss": 0.3741,
       "step": 98330
     },
     {
-      "epoch": 3.46,
-      "learning_rate": 1.1461133682542687e-05,
-      "loss": 0.2486,
+      "epoch": 3.543986737304934,
+      "grad_norm": 0.26063045859336853,
+      "learning_rate": 1.0334847303827e-05,
+      "loss": 0.3996,
       "step": 98335
     },
     {
-      "epoch": 3.46,
-      "learning_rate": 1.1458738948590166e-05,
-      "loss": 0.2736,
+      "epoch": 3.5441669369661586,
+      "grad_norm": 0.24216891825199127,
+      "learning_rate": 1.0332484087887867e-05,
+      "loss": 0.4231,
       "step": 98340
     },
     {
-      "epoch": 3.46,
-      "learning_rate": 1.145634439046199e-05,
-      "loss": 0.2724,
+      "epoch": 3.5443471366273833,
+      "grad_norm": 0.25515010952949524,
+      "learning_rate": 1.0330121071790808e-05,
+      "loss": 0.3934,
       "step": 98345
     },
     {
-      "epoch": 3.46,
-      "learning_rate": 1.1453950008189266e-05,
-      "loss": 0.2465,
+      "epoch": 3.544527336288608,
+      "grad_norm": 0.18739314377307892,
+      "learning_rate": 1.0327758255568007e-05,
+      "loss": 0.3672,
       "step": 98350
     },
     {
-      "epoch": 3.46,
-      "learning_rate": 1.1451555801803071e-05,
-      "loss": 0.2525,
+      "epoch": 3.5447075359498323,
+      "grad_norm": 0.2008611410856247,
+      "learning_rate": 1.0325395639251661e-05,
+      "loss": 0.3601,
       "step": 98355
     },
     {
-      "epoch": 3.46,
-      "learning_rate": 1.1449161771334496e-05,
-      "loss": 0.225,
+      "epoch": 3.544887735611057,
+      "grad_norm": 0.20512117445468903,
+      "learning_rate": 1.0323033222873956e-05,
+      "loss": 0.3739,
       "step": 98360
     },
     {
-      "epoch": 3.46,
-      "learning_rate": 1.1446767916814618e-05,
-      "loss": 0.2672,
+      "epoch": 3.545067935272282,
+      "grad_norm": 0.25009825825691223,
+      "learning_rate": 1.0320671006467086e-05,
+      "loss": 0.3642,
       "step": 98365
     },
     {
-      "epoch": 3.46,
-      "learning_rate": 1.1444374238274533e-05,
-      "loss": 0.2658,
+      "epoch": 3.545248134933506,
+      "grad_norm": 0.27992385625839233,
+      "learning_rate": 1.031830899006322e-05,
+      "loss": 0.4017,
       "step": 98370
     },
     {
-      "epoch": 3.46,
-      "learning_rate": 1.1441980735745317e-05,
-      "loss": 0.2542,
+      "epoch": 3.545428334594731,
+      "grad_norm": 0.2249823361635208,
+      "learning_rate": 1.0315947173694562e-05,
+      "loss": 0.3877,
       "step": 98375
     },
     {
-      "epoch": 3.46,
-      "learning_rate": 1.1439587409258035e-05,
-      "loss": 0.2588,
+      "epoch": 3.5456085342559556,
+      "grad_norm": 0.26109376549720764,
+      "learning_rate": 1.031358555739328e-05,
+      "loss": 0.3812,
       "step": 98380
     },
     {
-      "epoch": 3.46,
-      "learning_rate": 1.1437194258843775e-05,
-      "loss": 0.28,
+      "epoch": 3.5457887339171803,
+      "grad_norm": 0.23642925918102264,
+      "learning_rate": 1.031122414119155e-05,
+      "loss": 0.3841,
       "step": 98385
     },
     {
-      "epoch": 3.46,
-      "learning_rate": 1.1434801284533619e-05,
-      "loss": 0.2688,
+      "epoch": 3.545968933578405,
+      "grad_norm": 0.2479097694158554,
+      "learning_rate": 1.0308862925121548e-05,
+      "loss": 0.3603,
       "step": 98390
     },
     {
-      "epoch": 3.46,
-      "learning_rate": 1.1432408486358628e-05,
-      "loss": 0.2388,
+      "epoch": 3.5461491332396298,
+      "grad_norm": 0.30400943756103516,
+      "learning_rate": 1.0306501909215438e-05,
+      "loss": 0.3651,
       "step": 98395
     },
     {
-      "epoch": 3.46,
-      "learning_rate": 1.1430015864349863e-05,
-      "loss": 0.2906,
+      "epoch": 3.546329332900854,
+      "grad_norm": 0.1932227909564972,
+      "learning_rate": 1.0304141093505401e-05,
+      "loss": 0.4042,
       "step": 98400
     },
     {
-      "epoch": 3.46,
-      "learning_rate": 1.1427623418538409e-05,
-      "loss": 0.2705,
+      "epoch": 3.546509532562079,
+      "grad_norm": 0.25808045268058777,
+      "learning_rate": 1.0301780478023607e-05,
+      "loss": 0.3855,
       "step": 98405
     },
     {
-      "epoch": 3.46,
-      "learning_rate": 1.1425231148955318e-05,
-      "loss": 0.255,
+      "epoch": 3.5466897322233035,
+      "grad_norm": 0.17466764152050018,
+      "learning_rate": 1.0299420062802187e-05,
+      "loss": 0.3894,
       "step": 98410
     },
     {
-      "epoch": 3.46,
-      "learning_rate": 1.1422839055631654e-05,
-      "loss": 0.2498,
+      "epoch": 3.546869931884528,
+      "grad_norm": 0.2541663944721222,
+      "learning_rate": 1.0297059847873334e-05,
+      "loss": 0.3537,
       "step": 98415
     },
     {
-      "epoch": 3.46,
-      "learning_rate": 1.142044713859847e-05,
-      "loss": 0.2692,
+      "epoch": 3.5470501315457525,
+      "grad_norm": 0.277024507522583,
+      "learning_rate": 1.0294699833269186e-05,
+      "loss": 0.3547,
       "step": 98420
     },
     {
-      "epoch": 3.46,
-      "learning_rate": 1.1418055397886837e-05,
-      "loss": 0.239,
+      "epoch": 3.5472303312069773,
+      "grad_norm": 0.2566564381122589,
+      "learning_rate": 1.029234001902192e-05,
+      "loss": 0.3823,
       "step": 98425
     },
     {
-      "epoch": 3.46,
-      "learning_rate": 1.1415663833527807e-05,
-      "loss": 0.2592,
+      "epoch": 3.547410530868202,
+      "grad_norm": 0.1786002814769745,
+      "learning_rate": 1.0289980405163668e-05,
+      "loss": 0.366,
       "step": 98430
     },
     {
-      "epoch": 3.46,
-      "learning_rate": 1.141327244555242e-05,
-      "loss": 0.2535,
+      "epoch": 3.5475907305294268,
+      "grad_norm": 0.24739274382591248,
+      "learning_rate": 1.0287620991726577e-05,
+      "loss": 0.3887,
       "step": 98435
     },
     {
-      "epoch": 3.46,
-      "learning_rate": 1.1410881233991736e-05,
-      "loss": 0.2289,
+      "epoch": 3.5477709301906515,
+      "grad_norm": 0.2730603516101837,
+      "learning_rate": 1.0285261778742808e-05,
+      "loss": 0.3606,
       "step": 98440
     },
     {
-      "epoch": 3.46,
-      "learning_rate": 1.1408490198876812e-05,
-      "loss": 0.2741,
+      "epoch": 3.5479511298518758,
+      "grad_norm": 0.27662402391433716,
+      "learning_rate": 1.0282902766244498e-05,
+      "loss": 0.4084,
       "step": 98445
     },
     {
-      "epoch": 3.46,
-      "learning_rate": 1.1406099340238686e-05,
-      "loss": 0.2558,
+      "epoch": 3.5481313295131005,
+      "grad_norm": 0.18735605478286743,
+      "learning_rate": 1.0280543954263792e-05,
+      "loss": 0.3771,
       "step": 98450
     },
     {
-      "epoch": 3.46,
-      "learning_rate": 1.1403708658108404e-05,
-      "loss": 0.2714,
+      "epoch": 3.5483115291743252,
+      "grad_norm": 0.2412840873003006,
+      "learning_rate": 1.0278185342832821e-05,
+      "loss": 0.4203,
       "step": 98455
     },
     {
-      "epoch": 3.46,
-      "learning_rate": 1.1401318152516991e-05,
-      "loss": 0.2539,
+      "epoch": 3.5484917288355495,
+      "grad_norm": 0.2625180184841156,
+      "learning_rate": 1.0275826931983718e-05,
+      "loss": 0.4129,
       "step": 98460
     },
     {
-      "epoch": 3.46,
-      "learning_rate": 1.1398927823495512e-05,
-      "loss": 0.2633,
+      "epoch": 3.5486719284967743,
+      "grad_norm": 0.23540963232517242,
+      "learning_rate": 1.0273468721748631e-05,
+      "loss": 0.362,
       "step": 98465
     },
     {
-      "epoch": 3.46,
-      "learning_rate": 1.139653767107499e-05,
-      "loss": 0.2585,
+      "epoch": 3.548852128157999,
+      "grad_norm": 0.23965346813201904,
+      "learning_rate": 1.0271110712159679e-05,
+      "loss": 0.3762,
       "step": 98470
     },
     {
-      "epoch": 3.46,
-      "learning_rate": 1.1394147695286461e-05,
-      "loss": 0.2629,
+      "epoch": 3.5490323278192237,
+      "grad_norm": 0.23907659947872162,
+      "learning_rate": 1.0268752903248995e-05,
+      "loss": 0.3859,
       "step": 98475
     },
     {
-      "epoch": 3.46,
-      "learning_rate": 1.139175789616095e-05,
-      "loss": 0.2507,
+      "epoch": 3.5492125274804485,
+      "grad_norm": 0.2294924557209015,
+      "learning_rate": 1.0266395295048701e-05,
+      "loss": 0.3549,
       "step": 98480
     },
     {
-      "epoch": 3.46,
-      "learning_rate": 1.1389368273729504e-05,
-      "loss": 0.265,
+      "epoch": 3.549392727141673,
+      "grad_norm": 0.25225427746772766,
+      "learning_rate": 1.0264037887590907e-05,
+      "loss": 0.3969,
       "step": 98485
     },
     {
-      "epoch": 3.47,
-      "learning_rate": 1.1386978828023132e-05,
-      "loss": 0.249,
+      "epoch": 3.5495729268028975,
+      "grad_norm": 0.2280511111021042,
+      "learning_rate": 1.0261680680907754e-05,
+      "loss": 0.3913,
       "step": 98490
     },
     {
-      "epoch": 3.47,
-      "learning_rate": 1.1384589559072866e-05,
-      "loss": 0.2654,
+      "epoch": 3.549753126464122,
+      "grad_norm": 0.2504872977733612,
+      "learning_rate": 1.0259323675031357e-05,
+      "loss": 0.3917,
       "step": 98495
     },
     {
-      "epoch": 3.47,
-      "learning_rate": 1.1382200466909745e-05,
-      "loss": 0.2584,
+      "epoch": 3.549933326125347,
+      "grad_norm": 0.21838301420211792,
+      "learning_rate": 1.0256966869993804e-05,
+      "loss": 0.3925,
       "step": 98500
     },
     {
-      "epoch": 3.47,
-      "eval_loss": 0.2523847818374634,
-      "eval_runtime": 10.5582,
-      "eval_samples_per_second": 9.471,
-      "eval_steps_per_second": 9.471,
+      "epoch": 3.549933326125347,
+      "eval_loss": 0.42997604608535767,
+      "eval_runtime": 3.5334,
+      "eval_samples_per_second": 28.301,
+      "eval_steps_per_second": 7.075,
       "step": 98500
     },
     {
-      "epoch": 3.47,
-      "learning_rate": 1.1379811551564773e-05,
-      "loss": 0.2517,
+      "epoch": 3.5501135257865712,
+      "grad_norm": 0.226282998919487,
+      "learning_rate": 1.025461026582723e-05,
+      "loss": 0.3838,
       "step": 98505
     },
     {
-      "epoch": 3.47,
-      "learning_rate": 1.1377422813068975e-05,
-      "loss": 0.2734,
+      "epoch": 3.550293725447796,
+      "grad_norm": 0.2352711260318756,
+      "learning_rate": 1.0252253862563738e-05,
+      "loss": 0.3911,
       "step": 98510
     },
     {
-      "epoch": 3.47,
-      "learning_rate": 1.1375034251453353e-05,
-      "loss": 0.2441,
+      "epoch": 3.5504739251090207,
+      "grad_norm": 0.20618538558483124,
+      "learning_rate": 1.024989766023543e-05,
+      "loss": 0.3656,
       "step": 98515
     },
     {
-      "epoch": 3.47,
-      "learning_rate": 1.137264586674894e-05,
-      "loss": 0.2562,
+      "epoch": 3.5506541247702454,
+      "grad_norm": 0.2552780210971832,
+      "learning_rate": 1.0247541658874412e-05,
+      "loss": 0.3744,
       "step": 98520
     },
     {
-      "epoch": 3.47,
-      "learning_rate": 1.1370257658986743e-05,
-      "loss": 0.2511,
+      "epoch": 3.55083432443147,
+      "grad_norm": 0.21368472278118134,
+      "learning_rate": 1.0245185858512777e-05,
+      "loss": 0.3951,
       "step": 98525
     },
     {
-      "epoch": 3.47,
-      "learning_rate": 1.1367869628197766e-05,
-      "loss": 0.2513,
+      "epoch": 3.551014524092695,
+      "grad_norm": 0.257311075925827,
+      "learning_rate": 1.024283025918263e-05,
+      "loss": 0.3753,
       "step": 98530
     },
     {
-      "epoch": 3.47,
-      "learning_rate": 1.1365481774413009e-05,
-      "loss": 0.2531,
+      "epoch": 3.551194723753919,
+      "grad_norm": 0.1774069368839264,
+      "learning_rate": 1.0240474860916068e-05,
+      "loss": 0.3849,
       "step": 98535
     },
     {
-      "epoch": 3.47,
-      "learning_rate": 1.1363094097663495e-05,
-      "loss": 0.2581,
+      "epoch": 3.551374923415144,
+      "grad_norm": 0.20663419365882874,
+      "learning_rate": 1.023811966374518e-05,
+      "loss": 0.3717,
       "step": 98540
     },
     {
-      "epoch": 3.47,
-      "learning_rate": 1.1360706597980208e-05,
-      "loss": 0.2849,
+      "epoch": 3.5515551230763687,
+      "grad_norm": 0.23598462343215942,
+      "learning_rate": 1.0235764667702053e-05,
+      "loss": 0.4101,
       "step": 98545
     },
     {
-      "epoch": 3.47,
-      "learning_rate": 1.1358319275394167e-05,
-      "loss": 0.2515,
+      "epoch": 3.5517353227375934,
+      "grad_norm": 0.25389420986175537,
+      "learning_rate": 1.0233409872818772e-05,
+      "loss": 0.3636,
       "step": 98550
     },
     {
-      "epoch": 3.47,
-      "learning_rate": 1.1355932129936348e-05,
-      "loss": 0.2516,
+      "epoch": 3.5519155223988177,
+      "grad_norm": 0.18197034299373627,
+      "learning_rate": 1.0231055279127414e-05,
+      "loss": 0.3946,
       "step": 98555
     },
     {
-      "epoch": 3.47,
-      "learning_rate": 1.1353545161637771e-05,
-      "loss": 0.2606,
+      "epoch": 3.5520957220600424,
+      "grad_norm": 0.2080104500055313,
+      "learning_rate": 1.0228700886660078e-05,
+      "loss": 0.361,
       "step": 98560
     },
     {
-      "epoch": 3.47,
-      "learning_rate": 1.1351158370529414e-05,
-      "loss": 0.2511,
+      "epoch": 3.552275921721267,
+      "grad_norm": 0.2673806846141815,
+      "learning_rate": 1.0226346695448832e-05,
+      "loss": 0.3957,
       "step": 98565
     },
     {
-      "epoch": 3.47,
-      "learning_rate": 1.1348771756642273e-05,
-      "loss": 0.2594,
+      "epoch": 3.552456121382492,
+      "grad_norm": 0.25703608989715576,
+      "learning_rate": 1.0223992705525753e-05,
+      "loss": 0.3573,
       "step": 98570
     },
     {
-      "epoch": 3.47,
-      "learning_rate": 1.1346385320007324e-05,
-      "loss": 0.2801,
+      "epoch": 3.5526363210437166,
+      "grad_norm": 0.24846360087394714,
+      "learning_rate": 1.0221638916922909e-05,
+      "loss": 0.3862,
       "step": 98575
     },
     {
-      "epoch": 3.47,
-      "learning_rate": 1.1343999060655571e-05,
-      "loss": 0.2821,
+      "epoch": 3.552816520704941,
+      "grad_norm": 0.25549542903900146,
+      "learning_rate": 1.021928532967237e-05,
+      "loss": 0.3834,
       "step": 98580
     },
     {
-      "epoch": 3.47,
-      "learning_rate": 1.1341612978617994e-05,
-      "loss": 0.2317,
+      "epoch": 3.5529967203661657,
+      "grad_norm": 0.2898486256599426,
+      "learning_rate": 1.0216931943806213e-05,
+      "loss": 0.3971,
       "step": 98585
     },
     {
-      "epoch": 3.47,
-      "learning_rate": 1.133922707392557e-05,
-      "loss": 0.2723,
+      "epoch": 3.5531769200273904,
+      "grad_norm": 0.25332796573638916,
+      "learning_rate": 1.0214578759356504e-05,
+      "loss": 0.3667,
       "step": 98590
     },
     {
-      "epoch": 3.47,
-      "learning_rate": 1.133684134660927e-05,
-      "loss": 0.2516,
+      "epoch": 3.553357119688615,
+      "grad_norm": 0.225131094455719,
+      "learning_rate": 1.021222577635528e-05,
+      "loss": 0.3701,
       "step": 98595
     },
     {
-      "epoch": 3.47,
-      "learning_rate": 1.1334455796700081e-05,
-      "loss": 0.2523,
+      "epoch": 3.5535373193498394,
+      "grad_norm": 0.18378715217113495,
+      "learning_rate": 1.0209872994834627e-05,
+      "loss": 0.3875,
       "step": 98600
     },
     {
-      "epoch": 3.47,
-      "learning_rate": 1.1332070424228985e-05,
-      "loss": 0.2486,
+      "epoch": 3.553717519011064,
+      "grad_norm": 0.21205966174602509,
+      "learning_rate": 1.0207520414826583e-05,
+      "loss": 0.3717,
       "step": 98605
     },
     {
-      "epoch": 3.47,
-      "learning_rate": 1.1329685229226941e-05,
-      "loss": 0.2606,
+      "epoch": 3.553897718672289,
+      "grad_norm": 0.21734385192394257,
+      "learning_rate": 1.0205168036363225e-05,
+      "loss": 0.3397,
       "step": 98610
     },
     {
-      "epoch": 3.47,
-      "learning_rate": 1.132730021172493e-05,
-      "loss": 0.256,
+      "epoch": 3.5540779183335136,
+      "grad_norm": 0.22054389119148254,
+      "learning_rate": 1.0202815859476577e-05,
+      "loss": 0.4035,
       "step": 98615
     },
     {
-      "epoch": 3.47,
-      "learning_rate": 1.132491537175392e-05,
-      "loss": 0.2728,
+      "epoch": 3.5542581179947383,
+      "grad_norm": 0.24117785692214966,
+      "learning_rate": 1.0200463884198693e-05,
+      "loss": 0.3737,
       "step": 98620
     },
     {
-      "epoch": 3.47,
-      "learning_rate": 1.132253070934487e-05,
-      "loss": 0.2528,
+      "epoch": 3.554438317655963,
+      "grad_norm": 0.20071789622306824,
+      "learning_rate": 1.0198112110561631e-05,
+      "loss": 0.3685,
       "step": 98625
     },
     {
-      "epoch": 3.47,
-      "learning_rate": 1.1320146224528735e-05,
-      "loss": 0.2654,
+      "epoch": 3.5546185173171874,
+      "grad_norm": 0.29259100556373596,
+      "learning_rate": 1.0195760538597426e-05,
+      "loss": 0.3778,
       "step": 98630
     },
     {
-      "epoch": 3.47,
-      "learning_rate": 1.1317761917336494e-05,
-      "loss": 0.2539,
+      "epoch": 3.554798716978412,
+      "grad_norm": 0.24713879823684692,
+      "learning_rate": 1.0193409168338116e-05,
+      "loss": 0.3813,
       "step": 98635
     },
     {
-      "epoch": 3.47,
-      "learning_rate": 1.13153777877991e-05,
-      "loss": 0.2663,
+      "epoch": 3.554978916639637,
+      "grad_norm": 0.22376933693885803,
+      "learning_rate": 1.0191057999815743e-05,
+      "loss": 0.3789,
       "step": 98640
     },
     {
-      "epoch": 3.47,
-      "learning_rate": 1.1312993835947506e-05,
-      "loss": 0.2685,
+      "epoch": 3.555159116300861,
+      "grad_norm": 0.2563563287258148,
+      "learning_rate": 1.0188707033062325e-05,
+      "loss": 0.3876,
       "step": 98645
     },
     {
-      "epoch": 3.47,
-      "learning_rate": 1.1310610061812657e-05,
-      "loss": 0.2615,
+      "epoch": 3.555339315962086,
+      "grad_norm": 0.2513370215892792,
+      "learning_rate": 1.0186356268109917e-05,
+      "loss": 0.3808,
       "step": 98650
     },
     {
-      "epoch": 3.47,
-      "learning_rate": 1.1308226465425516e-05,
-      "loss": 0.2569,
+      "epoch": 3.5555195156233106,
+      "grad_norm": 0.23794013261795044,
+      "learning_rate": 1.0184005704990538e-05,
+      "loss": 0.3962,
       "step": 98655
     },
     {
-      "epoch": 3.47,
-      "learning_rate": 1.1305843046817041e-05,
-      "loss": 0.2481,
+      "epoch": 3.5556997152845353,
+      "grad_norm": 0.23575954139232635,
+      "learning_rate": 1.018165534373621e-05,
+      "loss": 0.3906,
       "step": 98660
     },
     {
-      "epoch": 3.47,
-      "learning_rate": 1.130345980601817e-05,
-      "loss": 0.2643,
+      "epoch": 3.55587991494576,
+      "grad_norm": 0.20875148475170135,
+      "learning_rate": 1.0179305184378959e-05,
+      "loss": 0.3759,
       "step": 98665
     },
     {
-      "epoch": 3.47,
-      "learning_rate": 1.1301076743059833e-05,
-      "loss": 0.2595,
+      "epoch": 3.556060114606985,
+      "grad_norm": 0.2343914955854416,
+      "learning_rate": 1.0176955226950799e-05,
+      "loss": 0.3565,
       "step": 98670
     },
     {
-      "epoch": 3.47,
-      "learning_rate": 1.1298693857973e-05,
-      "loss": 0.2522,
+      "epoch": 3.556240314268209,
+      "grad_norm": 0.24285919964313507,
+      "learning_rate": 1.0174605471483761e-05,
+      "loss": 0.3683,
       "step": 98675
     },
     {
-      "epoch": 3.47,
-      "learning_rate": 1.1296311150788596e-05,
-      "loss": 0.2526,
+      "epoch": 3.556420513929434,
+      "grad_norm": 0.24529211223125458,
+      "learning_rate": 1.0172255918009862e-05,
+      "loss": 0.3582,
       "step": 98680
     },
     {
-      "epoch": 3.47,
-      "learning_rate": 1.1293928621537561e-05,
-      "loss": 0.2947,
+      "epoch": 3.5566007135906585,
+      "grad_norm": 0.23087753355503082,
+      "learning_rate": 1.0169906566561087e-05,
+      "loss": 0.3616,
       "step": 98685
     },
     {
-      "epoch": 3.47,
-      "learning_rate": 1.129154627025082e-05,
-      "loss": 0.2512,
+      "epoch": 3.556780913251883,
+      "grad_norm": 0.2296365201473236,
+      "learning_rate": 1.0167557417169476e-05,
+      "loss": 0.3671,
       "step": 98690
     },
     {
-      "epoch": 3.47,
-      "learning_rate": 1.1289164096959326e-05,
-      "loss": 0.2723,
+      "epoch": 3.5569611129131076,
+      "grad_norm": 0.27826911211013794,
+      "learning_rate": 1.0165208469867022e-05,
+      "loss": 0.3988,
       "step": 98695
     },
     {
-      "epoch": 3.47,
-      "learning_rate": 1.1286782101694002e-05,
-      "loss": 0.2568,
+      "epoch": 3.5571413125743323,
+      "grad_norm": 0.21545450389385223,
+      "learning_rate": 1.0162859724685723e-05,
+      "loss": 0.3612,
       "step": 98700
     },
     {
-      "epoch": 3.47,
-      "learning_rate": 1.1284400284485766e-05,
-      "loss": 0.2901,
+      "epoch": 3.557321512235557,
+      "grad_norm": 0.23605626821517944,
+      "learning_rate": 1.0160511181657604e-05,
+      "loss": 0.358,
       "step": 98705
     },
     {
-      "epoch": 3.47,
-      "learning_rate": 1.128201864536556e-05,
-      "loss": 0.2345,
+      "epoch": 3.5575017118967818,
+      "grad_norm": 0.2647882103919983,
+      "learning_rate": 1.0158162840814627e-05,
+      "loss": 0.4287,
       "step": 98710
     },
     {
-      "epoch": 3.47,
-      "learning_rate": 1.1279637184364294e-05,
-      "loss": 0.2344,
+      "epoch": 3.5576819115580065,
+      "grad_norm": 0.2228686660528183,
+      "learning_rate": 1.0155814702188818e-05,
+      "loss": 0.414,
       "step": 98715
     },
     {
-      "epoch": 3.47,
-      "learning_rate": 1.1277255901512907e-05,
-      "loss": 0.2376,
+      "epoch": 3.557862111219231,
+      "grad_norm": 0.23151619732379913,
+      "learning_rate": 1.0153466765812161e-05,
+      "loss": 0.3869,
       "step": 98720
     },
     {
-      "epoch": 3.47,
-      "learning_rate": 1.1274874796842296e-05,
-      "loss": 0.2682,
+      "epoch": 3.5580423108804555,
+      "grad_norm": 0.24438922107219696,
+      "learning_rate": 1.0151119031716646e-05,
+      "loss": 0.3947,
       "step": 98725
     },
     {
-      "epoch": 3.47,
-      "learning_rate": 1.1272493870383404e-05,
-      "loss": 0.2536,
+      "epoch": 3.5582225105416803,
+      "grad_norm": 0.2741881012916565,
+      "learning_rate": 1.0148771499934257e-05,
+      "loss": 0.3601,
       "step": 98730
     },
     {
-      "epoch": 3.47,
-      "learning_rate": 1.1270113122167131e-05,
-      "loss": 0.2728,
+      "epoch": 3.5584027102029046,
+      "grad_norm": 0.3307461440563202,
+      "learning_rate": 1.0146424170496982e-05,
+      "loss": 0.3836,
       "step": 98735
     },
     {
-      "epoch": 3.47,
-      "learning_rate": 1.1267732552224388e-05,
-      "loss": 0.2463,
+      "epoch": 3.5585829098641293,
+      "grad_norm": 0.2043096274137497,
+      "learning_rate": 1.0144077043436792e-05,
+      "loss": 0.3482,
       "step": 98740
     },
     {
-      "epoch": 3.47,
-      "learning_rate": 1.1265352160586082e-05,
-      "loss": 0.2672,
+      "epoch": 3.558763109525354,
+      "grad_norm": 0.24508656561374664,
+      "learning_rate": 1.0141730118785687e-05,
+      "loss": 0.3842,
       "step": 98745
     },
     {
-      "epoch": 3.47,
-      "learning_rate": 1.1262971947283135e-05,
-      "loss": 0.247,
+      "epoch": 3.5589433091865788,
+      "grad_norm": 0.20151925086975098,
+      "learning_rate": 1.013938339657563e-05,
+      "loss": 0.3545,
       "step": 98750
     },
     {
-      "epoch": 3.47,
-      "learning_rate": 1.1260591912346444e-05,
-      "loss": 0.2392,
+      "epoch": 3.5591235088478035,
+      "grad_norm": 0.22490449249744415,
+      "learning_rate": 1.0137036876838598e-05,
+      "loss": 0.3762,
       "step": 98755
     },
     {
-      "epoch": 3.47,
-      "learning_rate": 1.1258212055806904e-05,
-      "loss": 0.2306,
+      "epoch": 3.559303708509028,
+      "grad_norm": 0.21991267800331116,
+      "learning_rate": 1.0134690559606563e-05,
+      "loss": 0.3916,
       "step": 98760
     },
     {
-      "epoch": 3.47,
-      "learning_rate": 1.1255832377695433e-05,
-      "loss": 0.2488,
+      "epoch": 3.5594839081702525,
+      "grad_norm": 0.3030828833580017,
+      "learning_rate": 1.0132344444911482e-05,
+      "loss": 0.3619,
       "step": 98765
     },
     {
-      "epoch": 3.48,
-      "learning_rate": 1.1253452878042912e-05,
-      "loss": 0.2726,
+      "epoch": 3.5596641078314772,
+      "grad_norm": 0.22489933669567108,
+      "learning_rate": 1.0129998532785337e-05,
+      "loss": 0.3944,
       "step": 98770
     },
     {
-      "epoch": 3.48,
-      "learning_rate": 1.1251073556880254e-05,
-      "loss": 0.2537,
+      "epoch": 3.559844307492702,
+      "grad_norm": 0.21111252903938293,
+      "learning_rate": 1.0127652823260092e-05,
+      "loss": 0.3725,
       "step": 98775
     },
     {
-      "epoch": 3.48,
-      "learning_rate": 1.1248694414238346e-05,
-      "loss": 0.2447,
+      "epoch": 3.5600245071539267,
+      "grad_norm": 0.23496191203594208,
+      "learning_rate": 1.012530731636768e-05,
+      "loss": 0.3468,
       "step": 98780
     },
     {
-      "epoch": 3.48,
-      "learning_rate": 1.1246315450148067e-05,
-      "loss": 0.2559,
+      "epoch": 3.560204706815151,
+      "grad_norm": 0.22039222717285156,
+      "learning_rate": 1.0122962012140083e-05,
+      "loss": 0.3789,
       "step": 98785
     },
     {
-      "epoch": 3.48,
-      "learning_rate": 1.124393666464033e-05,
-      "loss": 0.2508,
+      "epoch": 3.5603849064763757,
+      "grad_norm": 0.2323601096868515,
+      "learning_rate": 1.0120616910609243e-05,
+      "loss": 0.3234,
       "step": 98790
     },
     {
-      "epoch": 3.48,
-      "learning_rate": 1.124155805774601e-05,
-      "loss": 0.2415,
+      "epoch": 3.5605651061376005,
+      "grad_norm": 0.21386387944221497,
+      "learning_rate": 1.0118272011807134e-05,
+      "loss": 0.405,
       "step": 98795
     },
     {
-      "epoch": 3.48,
-      "learning_rate": 1.123917962949599e-05,
-      "loss": 0.2632,
+      "epoch": 3.560745305798825,
+      "grad_norm": 0.22486340999603271,
+      "learning_rate": 1.0115927315765678e-05,
+      "loss": 0.3728,
       "step": 98800
     },
     {
-      "epoch": 3.48,
-      "learning_rate": 1.1236801379921145e-05,
-      "loss": 0.2436,
+      "epoch": 3.56092550546005,
+      "grad_norm": 0.22867098450660706,
+      "learning_rate": 1.011358282251682e-05,
+      "loss": 0.4076,
       "step": 98805
     },
     {
-      "epoch": 3.48,
-      "learning_rate": 1.1234423309052371e-05,
-      "loss": 0.2469,
+      "epoch": 3.5611057051212742,
+      "grad_norm": 0.2371281385421753,
+      "learning_rate": 1.0111238532092524e-05,
+      "loss": 0.3776,
       "step": 98810
     },
     {
-      "epoch": 3.48,
-      "learning_rate": 1.1232045416920542e-05,
-      "loss": 0.2319,
+      "epoch": 3.561285904782499,
+      "grad_norm": 0.29785680770874023,
+      "learning_rate": 1.0108894444524713e-05,
+      "loss": 0.4038,
       "step": 98815
     },
     {
-      "epoch": 3.48,
-      "learning_rate": 1.1229667703556521e-05,
-      "loss": 0.2872,
+      "epoch": 3.5614661044437237,
+      "grad_norm": 0.24299155175685883,
+      "learning_rate": 1.0106550559845348e-05,
+      "loss": 0.3717,
       "step": 98820
     },
     {
-      "epoch": 3.48,
-      "learning_rate": 1.1227290168991187e-05,
-      "loss": 0.2447,
+      "epoch": 3.5616463041049484,
+      "grad_norm": 0.19322209060192108,
+      "learning_rate": 1.0104206878086343e-05,
+      "loss": 0.3855,
       "step": 98825
     },
     {
-      "epoch": 3.48,
-      "learning_rate": 1.1224912813255422e-05,
-      "loss": 0.2335,
+      "epoch": 3.5618265037661727,
+      "grad_norm": 0.20190706849098206,
+      "learning_rate": 1.0101863399279621e-05,
+      "loss": 0.3884,
       "step": 98830
     },
     {
-      "epoch": 3.48,
-      "learning_rate": 1.1222535636380089e-05,
-      "loss": 0.2465,
+      "epoch": 3.5620067034273974,
+      "grad_norm": 0.2435564547777176,
+      "learning_rate": 1.0099520123457138e-05,
+      "loss": 0.4033,
       "step": 98835
     },
     {
-      "epoch": 3.48,
-      "learning_rate": 1.1220158638396042e-05,
-      "loss": 0.2574,
+      "epoch": 3.562186903088622,
+      "grad_norm": 0.2607412040233612,
+      "learning_rate": 1.0097177050650808e-05,
+      "loss": 0.3681,
       "step": 98840
     },
     {
-      "epoch": 3.48,
-      "learning_rate": 1.1217781819334161e-05,
-      "loss": 0.2807,
+      "epoch": 3.562367102749847,
+      "grad_norm": 0.20941239595413208,
+      "learning_rate": 1.0094834180892554e-05,
+      "loss": 0.3891,
       "step": 98845
     },
     {
-      "epoch": 3.48,
-      "learning_rate": 1.12154051792253e-05,
-      "loss": 0.2698,
+      "epoch": 3.5625473024110716,
+      "grad_norm": 0.2150138020515442,
+      "learning_rate": 1.0092491514214301e-05,
+      "loss": 0.3664,
       "step": 98850
     },
     {
-      "epoch": 3.48,
-      "learning_rate": 1.1213028718100319e-05,
-      "loss": 0.2538,
+      "epoch": 3.562727502072296,
+      "grad_norm": 0.2666082978248596,
+      "learning_rate": 1.0090149050647955e-05,
+      "loss": 0.3944,
       "step": 98855
     },
     {
-      "epoch": 3.48,
-      "learning_rate": 1.1210652435990065e-05,
-      "loss": 0.2862,
+      "epoch": 3.5629077017335207,
+      "grad_norm": 0.27906569838523865,
+      "learning_rate": 1.0087806790225451e-05,
+      "loss": 0.3951,
       "step": 98860
     },
     {
-      "epoch": 3.48,
-      "learning_rate": 1.120827633292541e-05,
-      "loss": 0.2738,
+      "epoch": 3.5630879013947454,
+      "grad_norm": 0.2231929451227188,
+      "learning_rate": 1.0085464732978692e-05,
+      "loss": 0.3646,
       "step": 98865
     },
     {
-      "epoch": 3.48,
-      "learning_rate": 1.12059004089372e-05,
-      "loss": 0.2577,
+      "epoch": 3.56326810105597,
+      "grad_norm": 0.22479088604450226,
+      "learning_rate": 1.0083122878939588e-05,
+      "loss": 0.4146,
       "step": 98870
     },
     {
-      "epoch": 3.48,
-      "learning_rate": 1.1203524664056273e-05,
-      "loss": 0.2518,
+      "epoch": 3.5634483007171944,
+      "grad_norm": 0.2482202798128128,
+      "learning_rate": 1.0080781228140045e-05,
+      "loss": 0.3679,
       "step": 98875
     },
     {
-      "epoch": 3.48,
-      "learning_rate": 1.1201149098313485e-05,
-      "loss": 0.2623,
+      "epoch": 3.563628500378419,
+      "grad_norm": 0.3211915194988251,
+      "learning_rate": 1.0078439780611973e-05,
+      "loss": 0.3653,
       "step": 98880
     },
     {
-      "epoch": 3.48,
-      "learning_rate": 1.1198773711739691e-05,
-      "loss": 0.2556,
+      "epoch": 3.563808700039644,
+      "grad_norm": 0.2703056037425995,
+      "learning_rate": 1.007609853638726e-05,
+      "loss": 0.3871,
       "step": 98885
     },
     {
-      "epoch": 3.48,
-      "learning_rate": 1.1196398504365727e-05,
-      "loss": 0.2792,
+      "epoch": 3.5639888997008686,
+      "grad_norm": 0.21698278188705444,
+      "learning_rate": 1.0073757495497832e-05,
+      "loss": 0.3286,
       "step": 98890
     },
     {
-      "epoch": 3.48,
-      "learning_rate": 1.1194023476222429e-05,
-      "loss": 0.2372,
+      "epoch": 3.5641690993620934,
+      "grad_norm": 0.2787526547908783,
+      "learning_rate": 1.007141665797555e-05,
+      "loss": 0.3737,
       "step": 98895
     },
     {
-      "epoch": 3.48,
-      "learning_rate": 1.1191648627340628e-05,
-      "loss": 0.2521,
+      "epoch": 3.564349299023318,
+      "grad_norm": 0.23529410362243652,
+      "learning_rate": 1.0069076023852337e-05,
+      "loss": 0.4027,
       "step": 98900
     },
     {
-      "epoch": 3.48,
-      "learning_rate": 1.1189273957751181e-05,
-      "loss": 0.2586,
+      "epoch": 3.5645294986845424,
+      "grad_norm": 0.22988703846931458,
+      "learning_rate": 1.006673559316007e-05,
+      "loss": 0.3583,
       "step": 98905
     },
     {
-      "epoch": 3.48,
-      "learning_rate": 1.118689946748491e-05,
-      "loss": 0.2539,
+      "epoch": 3.564709698345767,
+      "grad_norm": 0.25721538066864014,
+      "learning_rate": 1.006439536593064e-05,
+      "loss": 0.3557,
       "step": 98910
     },
     {
-      "epoch": 3.48,
-      "learning_rate": 1.1184525156572646e-05,
-      "loss": 0.2706,
+      "epoch": 3.564889898006992,
+      "grad_norm": 0.3110162317752838,
+      "learning_rate": 1.006205534219593e-05,
+      "loss": 0.3782,
       "step": 98915
     },
     {
-      "epoch": 3.48,
-      "learning_rate": 1.1182151025045206e-05,
-      "loss": 0.2535,
+      "epoch": 3.565070097668216,
+      "grad_norm": 0.2260642945766449,
+      "learning_rate": 1.0059715521987829e-05,
+      "loss": 0.3644,
       "step": 98920
     },
     {
-      "epoch": 3.48,
-      "learning_rate": 1.1179777072933443e-05,
-      "loss": 0.2416,
+      "epoch": 3.565250297329441,
+      "grad_norm": 0.2566840350627899,
+      "learning_rate": 1.0057375905338199e-05,
+      "loss": 0.3579,
       "step": 98925
     },
     {
-      "epoch": 3.48,
-      "learning_rate": 1.1177403300268152e-05,
-      "loss": 0.2593,
+      "epoch": 3.5654304969906656,
+      "grad_norm": 0.19224494695663452,
+      "learning_rate": 1.0055036492278938e-05,
+      "loss": 0.3589,
       "step": 98930
     },
     {
-      "epoch": 3.48,
-      "learning_rate": 1.1175029707080184e-05,
-      "loss": 0.259,
+      "epoch": 3.5656106966518903,
+      "grad_norm": 0.24910667538642883,
+      "learning_rate": 1.005269728284191e-05,
+      "loss": 0.3914,
       "step": 98935
     },
     {
-      "epoch": 3.48,
-      "learning_rate": 1.1172656293400333e-05,
-      "loss": 0.2734,
+      "epoch": 3.565790896313115,
+      "grad_norm": 0.23995643854141235,
+      "learning_rate": 1.0050358277058991e-05,
+      "loss": 0.3903,
       "step": 98940
     },
     {
-      "epoch": 3.48,
-      "learning_rate": 1.1170283059259434e-05,
-      "loss": 0.2653,
+      "epoch": 3.56597109597434,
+      "grad_norm": 0.29442286491394043,
+      "learning_rate": 1.0048019474962044e-05,
+      "loss": 0.3919,
       "step": 98945
     },
     {
-      "epoch": 3.48,
-      "learning_rate": 1.1167910004688298e-05,
-      "loss": 0.2646,
+      "epoch": 3.566151295635564,
+      "grad_norm": 0.22082147002220154,
+      "learning_rate": 1.004568087658293e-05,
+      "loss": 0.346,
       "step": 98950
     },
     {
-      "epoch": 3.48,
-      "learning_rate": 1.1165537129717727e-05,
-      "loss": 0.2462,
+      "epoch": 3.566331495296789,
+      "grad_norm": 0.24365797638893127,
+      "learning_rate": 1.0043342481953525e-05,
+      "loss": 0.3524,
       "step": 98955
     },
     {
-      "epoch": 3.48,
-      "learning_rate": 1.1163164434378545e-05,
-      "loss": 0.2366,
+      "epoch": 3.5665116949580136,
+      "grad_norm": 0.21891754865646362,
+      "learning_rate": 1.0041004291105693e-05,
+      "loss": 0.3475,
       "step": 98960
     },
     {
-      "epoch": 3.48,
-      "learning_rate": 1.1160791918701557e-05,
-      "loss": 0.2477,
+      "epoch": 3.566691894619238,
+      "grad_norm": 0.22469094395637512,
+      "learning_rate": 1.0038666304071265e-05,
+      "loss": 0.4066,
       "step": 98965
     },
     {
-      "epoch": 3.48,
-      "learning_rate": 1.1158419582717563e-05,
-      "loss": 0.2542,
+      "epoch": 3.5668720942804626,
+      "grad_norm": 0.2404364049434662,
+      "learning_rate": 1.0036328520882119e-05,
+      "loss": 0.4111,
       "step": 98970
     },
     {
-      "epoch": 3.48,
-      "learning_rate": 1.1156047426457363e-05,
-      "loss": 0.2579,
+      "epoch": 3.5670522939416873,
+      "grad_norm": 0.22921812534332275,
+      "learning_rate": 1.0033990941570093e-05,
+      "loss": 0.3701,
       "step": 98975
     },
     {
-      "epoch": 3.48,
-      "learning_rate": 1.115367544995177e-05,
-      "loss": 0.2721,
+      "epoch": 3.567232493602912,
+      "grad_norm": 0.20173293352127075,
+      "learning_rate": 1.0031653566167048e-05,
+      "loss": 0.373,
       "step": 98980
     },
     {
-      "epoch": 3.48,
-      "learning_rate": 1.1151303653231568e-05,
-      "loss": 0.2704,
+      "epoch": 3.567412693264137,
+      "grad_norm": 0.23841802775859833,
+      "learning_rate": 1.0029316394704839e-05,
+      "loss": 0.3717,
       "step": 98985
     },
     {
-      "epoch": 3.48,
-      "learning_rate": 1.1148932036327573e-05,
-      "loss": 0.2792,
+      "epoch": 3.5675928929253615,
+      "grad_norm": 0.25082719326019287,
+      "learning_rate": 1.0026979427215275e-05,
+      "loss": 0.4165,
       "step": 98990
     },
     {
-      "epoch": 3.48,
-      "learning_rate": 1.1146560599270553e-05,
-      "loss": 0.252,
+      "epoch": 3.567773092586586,
+      "grad_norm": 0.23590582609176636,
+      "learning_rate": 1.0024642663730227e-05,
+      "loss": 0.3838,
       "step": 98995
     },
     {
-      "epoch": 3.48,
-      "learning_rate": 1.1144189342091328e-05,
-      "loss": 0.2534,
+      "epoch": 3.5679532922478105,
+      "grad_norm": 0.25584080815315247,
+      "learning_rate": 1.0022306104281523e-05,
+      "loss": 0.3765,
       "step": 99000
     },
     {
-      "epoch": 3.48,
-      "eval_loss": 0.25207310914993286,
-      "eval_runtime": 10.551,
-      "eval_samples_per_second": 9.478,
-      "eval_steps_per_second": 9.478,
+      "epoch": 3.5679532922478105,
+      "eval_loss": 0.43000587821006775,
+      "eval_runtime": 3.5299,
+      "eval_samples_per_second": 28.33,
+      "eval_steps_per_second": 7.082,
       "step": 99000
     },
     {
-      "epoch": 3.48,
-      "learning_rate": 1.114181826482067e-05,
-      "loss": 0.2452,
+      "epoch": 3.5681334919090353,
+      "grad_norm": 0.22017063200473785,
+      "learning_rate": 1.0019969748900998e-05,
+      "loss": 0.3714,
       "step": 99005
     },
     {
-      "epoch": 3.48,
-      "learning_rate": 1.113944736748937e-05,
-      "loss": 0.2669,
+      "epoch": 3.5683136915702596,
+      "grad_norm": 0.25801563262939453,
+      "learning_rate": 1.0017633597620485e-05,
+      "loss": 0.3674,
       "step": 99010
     },
     {
-      "epoch": 3.48,
-      "learning_rate": 1.11370766501282e-05,
-      "loss": 0.2615,
+      "epoch": 3.5684938912314843,
+      "grad_norm": 0.2582071125507355,
+      "learning_rate": 1.0015297650471805e-05,
+      "loss": 0.3786,
       "step": 99015
     },
     {
-      "epoch": 3.48,
-      "learning_rate": 1.1134706112767965e-05,
-      "loss": 0.2689,
+      "epoch": 3.568674090892709,
+      "grad_norm": 0.2171230912208557,
+      "learning_rate": 1.0012961907486804e-05,
+      "loss": 0.3997,
       "step": 99020
     },
     {
-      "epoch": 3.48,
-      "learning_rate": 1.1132335755439435e-05,
-      "loss": 0.2541,
+      "epoch": 3.5688542905539338,
+      "grad_norm": 0.20874454081058502,
+      "learning_rate": 1.0010626368697292e-05,
+      "loss": 0.4006,
       "step": 99025
     },
     {
-      "epoch": 3.48,
-      "learning_rate": 1.1129965578173385e-05,
-      "loss": 0.27,
+      "epoch": 3.5690344902151585,
+      "grad_norm": 0.21824629604816437,
+      "learning_rate": 1.0008291034135098e-05,
+      "loss": 0.3733,
       "step": 99030
     },
     {
-      "epoch": 3.48,
-      "learning_rate": 1.112759558100058e-05,
-      "loss": 0.2776,
+      "epoch": 3.5692146898763832,
+      "grad_norm": 0.3257184326648712,
+      "learning_rate": 1.0005955903832031e-05,
+      "loss": 0.3853,
       "step": 99035
     },
     {
-      "epoch": 3.48,
-      "learning_rate": 1.1125225763951805e-05,
-      "loss": 0.2625,
+      "epoch": 3.5693948895376075,
+      "grad_norm": 0.21955911815166473,
+      "learning_rate": 1.0003620977819908e-05,
+      "loss": 0.3398,
       "step": 99040
     },
     {
-      "epoch": 3.48,
-      "learning_rate": 1.112285612705784e-05,
-      "loss": 0.2843,
+      "epoch": 3.5695750891988323,
+      "grad_norm": 0.2739407420158386,
+      "learning_rate": 1.0001286256130551e-05,
+      "loss": 0.3881,
       "step": 99045
     },
     {
-      "epoch": 3.48,
-      "learning_rate": 1.1120486670349442e-05,
-      "loss": 0.2572,
+      "epoch": 3.569755288860057,
+      "grad_norm": 0.2548768222332001,
+      "learning_rate": 9.998951738795767e-06,
+      "loss": 0.3858,
       "step": 99050
     },
     {
-      "epoch": 3.49,
-      "learning_rate": 1.1118117393857369e-05,
-      "loss": 0.2688,
+      "epoch": 3.5699354885212817,
+      "grad_norm": 0.22001412510871887,
+      "learning_rate": 9.996617425847363e-06,
+      "loss": 0.3921,
       "step": 99055
     },
     {
-      "epoch": 3.49,
-      "learning_rate": 1.1115748297612403e-05,
-      "loss": 0.2671,
+      "epoch": 3.570115688182506,
+      "grad_norm": 0.2573753595352173,
+      "learning_rate": 9.994283317317138e-06,
+      "loss": 0.3646,
       "step": 99060
     },
     {
-      "epoch": 3.49,
-      "learning_rate": 1.1113379381645294e-05,
-      "loss": 0.2572,
+      "epoch": 3.5702958878437308,
+      "grad_norm": 0.30579379200935364,
+      "learning_rate": 9.991949413236898e-06,
+      "loss": 0.4174,
       "step": 99065
     },
     {
-      "epoch": 3.49,
-      "learning_rate": 1.1111010645986794e-05,
-      "loss": 0.2505,
+      "epoch": 3.5704760875049555,
+      "grad_norm": 0.22438612580299377,
+      "learning_rate": 9.989615713638434e-06,
+      "loss": 0.3636,
       "step": 99070
     },
     {
-      "epoch": 3.49,
-      "learning_rate": 1.1108642090667676e-05,
-      "loss": 0.2631,
+      "epoch": 3.57065628716618,
+      "grad_norm": 0.25755026936531067,
+      "learning_rate": 9.987282218553568e-06,
+      "loss": 0.3919,
       "step": 99075
     },
     {
-      "epoch": 3.49,
-      "learning_rate": 1.1106273715718685e-05,
-      "loss": 0.2634,
+      "epoch": 3.570836486827405,
+      "grad_norm": 0.20849241316318512,
+      "learning_rate": 9.984948928014057e-06,
+      "loss": 0.3864,
       "step": 99080
     },
     {
-      "epoch": 3.49,
-      "learning_rate": 1.1103905521170574e-05,
-      "loss": 0.2694,
+      "epoch": 3.5710166864886292,
+      "grad_norm": 0.2606984078884125,
+      "learning_rate": 9.982615842051719e-06,
+      "loss": 0.3855,
       "step": 99085
     },
     {
-      "epoch": 3.49,
-      "learning_rate": 1.1101537507054083e-05,
-      "loss": 0.2526,
+      "epoch": 3.571196886149854,
+      "grad_norm": 0.224889874458313,
+      "learning_rate": 9.98028296069833e-06,
+      "loss": 0.3739,
       "step": 99090
     },
     {
-      "epoch": 3.49,
-      "learning_rate": 1.1099169673399967e-05,
-      "loss": 0.2336,
+      "epoch": 3.5713770858110787,
+      "grad_norm": 0.27115222811698914,
+      "learning_rate": 9.977950283985673e-06,
+      "loss": 0.3709,
       "step": 99095
     },
     {
-      "epoch": 3.49,
-      "learning_rate": 1.1096802020238984e-05,
-      "loss": 0.2396,
+      "epoch": 3.5715572854723034,
+      "grad_norm": 0.22326913475990295,
+      "learning_rate": 9.97561781194555e-06,
+      "loss": 0.3937,
       "step": 99100
     },
     {
-      "epoch": 3.49,
-      "learning_rate": 1.1094434547601862e-05,
-      "loss": 0.2503,
+      "epoch": 3.5717374851335277,
+      "grad_norm": 0.22230251133441925,
+      "learning_rate": 9.97328554460972e-06,
+      "loss": 0.362,
       "step": 99105
     },
     {
-      "epoch": 3.49,
-      "learning_rate": 1.1092067255519333e-05,
-      "loss": 0.2572,
+      "epoch": 3.5719176847947525,
+      "grad_norm": 0.252175509929657,
+      "learning_rate": 9.970953482009953e-06,
+      "loss": 0.3838,
       "step": 99110
     },
     {
-      "epoch": 3.49,
-      "learning_rate": 1.1089700144022155e-05,
-      "loss": 0.2532,
+      "epoch": 3.572097884455977,
+      "grad_norm": 0.24773818254470825,
+      "learning_rate": 9.968621624178046e-06,
+      "loss": 0.3791,
       "step": 99115
     },
     {
-      "epoch": 3.49,
-      "learning_rate": 1.1087333213141055e-05,
-      "loss": 0.2654,
+      "epoch": 3.572278084117202,
+      "grad_norm": 0.2345571517944336,
+      "learning_rate": 9.966289971145756e-06,
+      "loss": 0.3671,
       "step": 99120
     },
     {
-      "epoch": 3.49,
-      "learning_rate": 1.1084966462906765e-05,
-      "loss": 0.2429,
+      "epoch": 3.5724582837784267,
+      "grad_norm": 0.2684936821460724,
+      "learning_rate": 9.963958522944858e-06,
+      "loss": 0.3666,
       "step": 99125
     },
     {
-      "epoch": 3.49,
-      "learning_rate": 1.1082599893350005e-05,
-      "loss": 0.248,
+      "epoch": 3.5726384834396514,
+      "grad_norm": 0.2345341295003891,
+      "learning_rate": 9.961627279607111e-06,
+      "loss": 0.3676,
       "step": 99130
     },
     {
-      "epoch": 3.49,
-      "learning_rate": 1.1080233504501519e-05,
-      "loss": 0.2411,
+      "epoch": 3.5728186831008757,
+      "grad_norm": 0.22453191876411438,
+      "learning_rate": 9.959296241164273e-06,
+      "loss": 0.3522,
       "step": 99135
     },
     {
-      "epoch": 3.49,
-      "learning_rate": 1.107786729639203e-05,
-      "loss": 0.2428,
+      "epoch": 3.5729988827621004,
+      "grad_norm": 0.26160648465156555,
+      "learning_rate": 9.956965407648122e-06,
+      "loss": 0.3757,
       "step": 99140
     },
     {
-      "epoch": 3.49,
-      "learning_rate": 1.1075501269052252e-05,
-      "loss": 0.2668,
+      "epoch": 3.573179082423325,
+      "grad_norm": 0.24852387607097626,
+      "learning_rate": 9.954634779090404e-06,
+      "loss": 0.4034,
       "step": 99145
     },
     {
-      "epoch": 3.49,
-      "learning_rate": 1.107313542251292e-05,
-      "loss": 0.2729,
+      "epoch": 3.5733592820845494,
+      "grad_norm": 0.22410407662391663,
+      "learning_rate": 9.952304355522876e-06,
+      "loss": 0.4013,
       "step": 99150
     },
     {
-      "epoch": 3.49,
-      "learning_rate": 1.1070769756804739e-05,
-      "loss": 0.2708,
+      "epoch": 3.573539481745774,
+      "grad_norm": 0.2551601529121399,
+      "learning_rate": 9.949974136977286e-06,
+      "loss": 0.3943,
       "step": 99155
     },
     {
-      "epoch": 3.49,
-      "learning_rate": 1.106840427195844e-05,
-      "loss": 0.2568,
+      "epoch": 3.573719681406999,
+      "grad_norm": 0.20935653150081635,
+      "learning_rate": 9.947644123485376e-06,
+      "loss": 0.3551,
       "step": 99160
     },
     {
-      "epoch": 3.49,
-      "learning_rate": 1.1066038968004734e-05,
-      "loss": 0.2346,
+      "epoch": 3.5738998810682236,
+      "grad_norm": 0.24979546666145325,
+      "learning_rate": 9.945314315078907e-06,
+      "loss": 0.3857,
       "step": 99165
     },
     {
-      "epoch": 3.49,
-      "learning_rate": 1.1063673844974318e-05,
-      "loss": 0.2624,
+      "epoch": 3.5740800807294484,
+      "grad_norm": 0.2232057750225067,
+      "learning_rate": 9.94298471178963e-06,
+      "loss": 0.3694,
       "step": 99170
     },
     {
-      "epoch": 3.49,
-      "learning_rate": 1.1061308902897922e-05,
-      "loss": 0.2456,
+      "epoch": 3.574260280390673,
+      "grad_norm": 0.20639793574810028,
+      "learning_rate": 9.94065531364925e-06,
+      "loss": 0.3668,
       "step": 99175
     },
     {
-      "epoch": 3.49,
-      "learning_rate": 1.1058944141806243e-05,
-      "loss": 0.2458,
+      "epoch": 3.5744404800518974,
+      "grad_norm": 0.2616964876651764,
+      "learning_rate": 9.938326120689534e-06,
+      "loss": 0.3901,
       "step": 99180
     },
     {
-      "epoch": 3.49,
-      "learning_rate": 1.1056579561729976e-05,
-      "loss": 0.27,
+      "epoch": 3.574620679713122,
+      "grad_norm": 0.2324073314666748,
+      "learning_rate": 9.93599713294221e-06,
+      "loss": 0.3556,
       "step": 99185
     },
     {
-      "epoch": 3.49,
-      "learning_rate": 1.1054215162699846e-05,
-      "loss": 0.2613,
+      "epoch": 3.574800879374347,
+      "grad_norm": 0.23615048825740814,
+      "learning_rate": 9.933668350439008e-06,
+      "loss": 0.3786,
       "step": 99190
     },
     {
-      "epoch": 3.49,
-      "learning_rate": 1.1051850944746543e-05,
-      "loss": 0.2632,
+      "epoch": 3.574981079035571,
+      "grad_norm": 0.21338878571987152,
+      "learning_rate": 9.931339773211657e-06,
+      "loss": 0.3847,
       "step": 99195
     },
     {
-      "epoch": 3.49,
-      "learning_rate": 1.1049486907900752e-05,
-      "loss": 0.2627,
+      "epoch": 3.575161278696796,
+      "grad_norm": 0.2527485489845276,
+      "learning_rate": 9.929011401291877e-06,
+      "loss": 0.384,
       "step": 99200
     },
     {
-      "epoch": 3.49,
-      "learning_rate": 1.104712305219319e-05,
-      "loss": 0.2485,
+      "epoch": 3.5753414783580206,
+      "grad_norm": 0.20677971839904785,
+      "learning_rate": 9.926683234711406e-06,
+      "loss": 0.4078,
       "step": 99205
     },
     {
-      "epoch": 3.49,
-      "learning_rate": 1.104475937765453e-05,
-      "loss": 0.2587,
+      "epoch": 3.5755216780192454,
+      "grad_norm": 0.2947938144207001,
+      "learning_rate": 9.92435527350196e-06,
+      "loss": 0.3631,
       "step": 99210
     },
     {
-      "epoch": 3.49,
-      "learning_rate": 1.1042395884315481e-05,
-      "loss": 0.2514,
+      "epoch": 3.57570187768047,
+      "grad_norm": 0.21625371277332306,
+      "learning_rate": 9.922027517695253e-06,
+      "loss": 0.3614,
       "step": 99215
     },
     {
-      "epoch": 3.49,
-      "learning_rate": 1.1040032572206724e-05,
-      "loss": 0.2589,
+      "epoch": 3.575882077341695,
+      "grad_norm": 0.21706773340702057,
+      "learning_rate": 9.919699967323001e-06,
+      "loss": 0.3756,
       "step": 99220
     },
     {
-      "epoch": 3.49,
-      "learning_rate": 1.1037669441358933e-05,
-      "loss": 0.2746,
+      "epoch": 3.576062277002919,
+      "grad_norm": 0.24126207828521729,
+      "learning_rate": 9.917372622416912e-06,
+      "loss": 0.3679,
       "step": 99225
     },
     {
-      "epoch": 3.49,
-      "learning_rate": 1.1035306491802815e-05,
-      "loss": 0.2571,
+      "epoch": 3.576242476664144,
+      "grad_norm": 0.20370317995548248,
+      "learning_rate": 9.915045483008706e-06,
+      "loss": 0.3477,
       "step": 99230
     },
     {
-      "epoch": 3.49,
-      "learning_rate": 1.1032943723569039e-05,
-      "loss": 0.2562,
+      "epoch": 3.5764226763253686,
+      "grad_norm": 0.2700237035751343,
+      "learning_rate": 9.912718549130088e-06,
+      "loss": 0.3631,
       "step": 99235
     },
     {
-      "epoch": 3.49,
-      "learning_rate": 1.103058113668828e-05,
-      "loss": 0.2581,
+      "epoch": 3.576602875986593,
+      "grad_norm": 0.21943305432796478,
+      "learning_rate": 9.910391820812756e-06,
+      "loss": 0.3704,
       "step": 99240
     },
     {
-      "epoch": 3.49,
-      "learning_rate": 1.1028218731191215e-05,
-      "loss": 0.2801,
+      "epoch": 3.5767830756478176,
+      "grad_norm": 0.22363239526748657,
+      "learning_rate": 9.908065298088414e-06,
+      "loss": 0.3915,
       "step": 99245
     },
     {
-      "epoch": 3.49,
-      "learning_rate": 1.1025856507108529e-05,
-      "loss": 0.2654,
+      "epoch": 3.5769632753090423,
+      "grad_norm": 0.2681196928024292,
+      "learning_rate": 9.905738980988763e-06,
+      "loss": 0.3785,
       "step": 99250
     },
     {
-      "epoch": 3.49,
-      "learning_rate": 1.1023494464470885e-05,
-      "loss": 0.2478,
+      "epoch": 3.577143474970267,
+      "grad_norm": 0.2544713318347931,
+      "learning_rate": 9.903412869545484e-06,
+      "loss": 0.3965,
       "step": 99255
     },
     {
-      "epoch": 3.49,
-      "learning_rate": 1.1021132603308949e-05,
-      "loss": 0.2558,
+      "epoch": 3.577323674631492,
+      "grad_norm": 0.254402220249176,
+      "learning_rate": 9.901086963790294e-06,
+      "loss": 0.4389,
       "step": 99260
     },
     {
-      "epoch": 3.49,
-      "learning_rate": 1.1018770923653391e-05,
-      "loss": 0.269,
+      "epoch": 3.5775038742927165,
+      "grad_norm": 0.2791866064071655,
+      "learning_rate": 9.89876126375487e-06,
+      "loss": 0.4082,
       "step": 99265
     },
     {
-      "epoch": 3.49,
-      "learning_rate": 1.1016409425534888e-05,
-      "loss": 0.2642,
+      "epoch": 3.577684073953941,
+      "grad_norm": 0.2096082866191864,
+      "learning_rate": 9.896435769470897e-06,
+      "loss": 0.3456,
       "step": 99270
     },
     {
-      "epoch": 3.49,
-      "learning_rate": 1.1014048108984093e-05,
-      "loss": 0.2572,
+      "epoch": 3.5778642736151656,
+      "grad_norm": 0.20750059187412262,
+      "learning_rate": 9.894110480970064e-06,
+      "loss": 0.3667,
       "step": 99275
     },
     {
-      "epoch": 3.49,
-      "learning_rate": 1.1011686974031668e-05,
-      "loss": 0.2507,
+      "epoch": 3.5780444732763903,
+      "grad_norm": 0.22872807085514069,
+      "learning_rate": 9.891785398284045e-06,
+      "loss": 0.348,
       "step": 99280
     },
     {
-      "epoch": 3.49,
-      "learning_rate": 1.1009326020708257e-05,
-      "loss": 0.2368,
+      "epoch": 3.578224672937615,
+      "grad_norm": 0.2133398801088333,
+      "learning_rate": 9.889460521444541e-06,
+      "loss": 0.3775,
       "step": 99285
     },
     {
-      "epoch": 3.49,
-      "learning_rate": 1.100696524904454e-05,
-      "loss": 0.2572,
+      "epoch": 3.5784048725988393,
+      "grad_norm": 0.24460799992084503,
+      "learning_rate": 9.8871358504832e-06,
+      "loss": 0.3606,
       "step": 99290
     },
     {
-      "epoch": 3.49,
-      "learning_rate": 1.1004604659071155e-05,
-      "loss": 0.2535,
+      "epoch": 3.578585072260064,
+      "grad_norm": 0.25514861941337585,
+      "learning_rate": 9.884811385431703e-06,
+      "loss": 0.3779,
       "step": 99295
     },
     {
-      "epoch": 3.49,
-      "learning_rate": 1.1002244250818746e-05,
-      "loss": 0.2556,
+      "epoch": 3.578765271921289,
+      "grad_norm": 0.2582806944847107,
+      "learning_rate": 9.88248712632173e-06,
+      "loss": 0.3921,
       "step": 99300
     },
     {
-      "epoch": 3.49,
-      "learning_rate": 1.099988402431798e-05,
-      "loss": 0.2535,
+      "epoch": 3.5789454715825135,
+      "grad_norm": 0.21507005393505096,
+      "learning_rate": 9.880163073184945e-06,
+      "loss": 0.3696,
       "step": 99305
     },
     {
-      "epoch": 3.49,
-      "learning_rate": 1.0997523979599494e-05,
-      "loss": 0.2563,
+      "epoch": 3.5791256712437383,
+      "grad_norm": 0.2389105260372162,
+      "learning_rate": 9.87783922605301e-06,
+      "loss": 0.3775,
       "step": 99310
     },
     {
-      "epoch": 3.49,
-      "learning_rate": 1.0995164116693924e-05,
-      "loss": 0.261,
+      "epoch": 3.5793058709049625,
+      "grad_norm": 0.24397464096546173,
+      "learning_rate": 9.875515584957587e-06,
+      "loss": 0.4102,
       "step": 99315
     },
     {
-      "epoch": 3.49,
-      "learning_rate": 1.0992804435631915e-05,
-      "loss": 0.2546,
+      "epoch": 3.5794860705661873,
+      "grad_norm": 0.262635201215744,
+      "learning_rate": 9.87319214993033e-06,
+      "loss": 0.3625,
       "step": 99320
     },
     {
-      "epoch": 3.49,
-      "learning_rate": 1.099044493644412e-05,
-      "loss": 0.2593,
+      "epoch": 3.579666270227412,
+      "grad_norm": 0.23571620881557465,
+      "learning_rate": 9.870868921002908e-06,
+      "loss": 0.4029,
       "step": 99325
     },
     {
-      "epoch": 3.49,
-      "learning_rate": 1.0988085619161165e-05,
-      "loss": 0.2948,
+      "epoch": 3.5798464698886368,
+      "grad_norm": 0.20199549198150635,
+      "learning_rate": 9.868545898206968e-06,
+      "loss": 0.3573,
       "step": 99330
     },
     {
-      "epoch": 3.49,
-      "learning_rate": 1.0985726483813683e-05,
-      "loss": 0.2426,
+      "epoch": 3.580026669549861,
+      "grad_norm": 0.262723445892334,
+      "learning_rate": 9.86622308157416e-06,
+      "loss": 0.3837,
       "step": 99335
     },
     {
-      "epoch": 3.5,
-      "learning_rate": 1.0983367530432298e-05,
-      "loss": 0.2673,
+      "epoch": 3.5802068692110858,
+      "grad_norm": 0.2516125440597534,
+      "learning_rate": 9.863900471136135e-06,
+      "loss": 0.367,
       "step": 99340
     },
     {
-      "epoch": 3.5,
-      "learning_rate": 1.098100875904766e-05,
-      "loss": 0.2426,
+      "epoch": 3.5803870688723105,
+      "grad_norm": 0.2330031394958496,
+      "learning_rate": 9.861578066924524e-06,
+      "loss": 0.3491,
       "step": 99345
     },
     {
-      "epoch": 3.5,
-      "learning_rate": 1.097865016969038e-05,
-      "loss": 0.259,
+      "epoch": 3.5805672685335352,
+      "grad_norm": 0.2417108565568924,
+      "learning_rate": 9.85925586897099e-06,
+      "loss": 0.4232,
       "step": 99350
     },
     {
-      "epoch": 3.5,
-      "learning_rate": 1.0976291762391092e-05,
-      "loss": 0.2576,
+      "epoch": 3.58074746819476,
+      "grad_norm": 0.3049721121788025,
+      "learning_rate": 9.856933877307173e-06,
+      "loss": 0.3506,
       "step": 99355
     },
     {
-      "epoch": 3.5,
-      "learning_rate": 1.09739335371804e-05,
-      "loss": 0.2639,
+      "epoch": 3.5809276678559843,
+      "grad_norm": 0.20108012855052948,
+      "learning_rate": 9.854612091964683e-06,
+      "loss": 0.3765,
       "step": 99360
     },
     {
-      "epoch": 3.5,
-      "learning_rate": 1.0971575494088948e-05,
-      "loss": 0.2561,
+      "epoch": 3.581107867517209,
+      "grad_norm": 0.18804548680782318,
+      "learning_rate": 9.852290512975179e-06,
+      "loss": 0.3871,
       "step": 99365
     },
     {
-      "epoch": 3.5,
-      "learning_rate": 1.0969217633147336e-05,
-      "loss": 0.2426,
+      "epoch": 3.5812880671784337,
+      "grad_norm": 0.24155212938785553,
+      "learning_rate": 9.849969140370286e-06,
+      "loss": 0.3778,
       "step": 99370
     },
     {
-      "epoch": 3.5,
-      "learning_rate": 1.0966859954386194e-05,
-      "loss": 0.2757,
+      "epoch": 3.5814682668396585,
+      "grad_norm": 0.2706049680709839,
+      "learning_rate": 9.847647974181626e-06,
+      "loss": 0.3831,
       "step": 99375
     },
     {
-      "epoch": 3.5,
-      "learning_rate": 1.0964502457836118e-05,
-      "loss": 0.2638,
+      "epoch": 3.5816484665008828,
+      "grad_norm": 0.26998743414878845,
+      "learning_rate": 9.845327014440834e-06,
+      "loss": 0.4087,
       "step": 99380
     },
     {
-      "epoch": 3.5,
-      "learning_rate": 1.0962145143527735e-05,
-      "loss": 0.2468,
+      "epoch": 3.5818286661621075,
+      "grad_norm": 0.21241755783557892,
+      "learning_rate": 9.843006261179513e-06,
+      "loss": 0.388,
       "step": 99385
     },
     {
-      "epoch": 3.5,
-      "learning_rate": 1.0959788011491649e-05,
-      "loss": 0.2661,
+      "epoch": 3.5820088658233322,
+      "grad_norm": 0.261828750371933,
+      "learning_rate": 9.84068571442931e-06,
+      "loss": 0.3801,
       "step": 99390
     },
     {
-      "epoch": 3.5,
-      "learning_rate": 1.095743106175846e-05,
-      "loss": 0.2553,
+      "epoch": 3.582189065484557,
+      "grad_norm": 0.26328206062316895,
+      "learning_rate": 9.838365374221827e-06,
+      "loss": 0.4188,
       "step": 99395
     },
     {
-      "epoch": 3.5,
-      "learning_rate": 1.0955074294358769e-05,
-      "loss": 0.2562,
+      "epoch": 3.5823692651457817,
+      "grad_norm": 0.26441022753715515,
+      "learning_rate": 9.836045240588684e-06,
+      "loss": 0.3694,
       "step": 99400
     },
     {
-      "epoch": 3.5,
-      "learning_rate": 1.095271770932319e-05,
-      "loss": 0.256,
+      "epoch": 3.5825494648070064,
+      "grad_norm": 0.21081748604774475,
+      "learning_rate": 9.833725313561487e-06,
+      "loss": 0.33,
       "step": 99405
     },
     {
-      "epoch": 3.5,
-      "learning_rate": 1.0950361306682313e-05,
-      "loss": 0.2602,
+      "epoch": 3.5827296644682307,
+      "grad_norm": 0.23671510815620422,
+      "learning_rate": 9.831405593171836e-06,
+      "loss": 0.4302,
       "step": 99410
     },
     {
-      "epoch": 3.5,
-      "learning_rate": 1.0948005086466737e-05,
-      "loss": 0.2556,
+      "epoch": 3.5829098641294554,
+      "grad_norm": 0.2610369920730591,
+      "learning_rate": 9.829086079451357e-06,
+      "loss": 0.368,
       "step": 99415
     },
     {
-      "epoch": 3.5,
-      "learning_rate": 1.0945649048707044e-05,
-      "loss": 0.2603,
+      "epoch": 3.58309006379068,
+      "grad_norm": 0.22345009446144104,
+      "learning_rate": 9.826766772431643e-06,
+      "loss": 0.3939,
       "step": 99420
     },
     {
-      "epoch": 3.5,
-      "learning_rate": 1.0943293193433834e-05,
-      "loss": 0.2637,
+      "epoch": 3.5832702634519045,
+      "grad_norm": 0.28314733505249023,
+      "learning_rate": 9.824447672144293e-06,
+      "loss": 0.3636,
       "step": 99425
     },
     {
-      "epoch": 3.5,
-      "learning_rate": 1.094093752067771e-05,
-      "loss": 0.2701,
+      "epoch": 3.583450463113129,
+      "grad_norm": 0.26532143354415894,
+      "learning_rate": 9.822128778620907e-06,
+      "loss": 0.386,
       "step": 99430
     },
     {
-      "epoch": 3.5,
-      "learning_rate": 1.0938582030469238e-05,
-      "loss": 0.2635,
+      "epoch": 3.583630662774354,
+      "grad_norm": 0.23965269327163696,
+      "learning_rate": 9.819810091893078e-06,
+      "loss": 0.4179,
       "step": 99435
     },
     {
-      "epoch": 3.5,
-      "learning_rate": 1.0936226722839018e-05,
-      "loss": 0.2507,
+      "epoch": 3.5838108624355787,
+      "grad_norm": 0.23256781697273254,
+      "learning_rate": 9.817491611992386e-06,
+      "loss": 0.3622,
       "step": 99440
     },
     {
-      "epoch": 3.5,
-      "learning_rate": 1.093387159781763e-05,
-      "loss": 0.2497,
+      "epoch": 3.5839910620968034,
+      "grad_norm": 0.2113410234451294,
+      "learning_rate": 9.815173338950442e-06,
+      "loss": 0.3808,
       "step": 99445
     },
     {
-      "epoch": 3.5,
-      "learning_rate": 1.0931516655435645e-05,
-      "loss": 0.253,
+      "epoch": 3.584171261758028,
+      "grad_norm": 0.21958471834659576,
+      "learning_rate": 9.812855272798822e-06,
+      "loss": 0.3558,
       "step": 99450
     },
     {
-      "epoch": 3.5,
-      "learning_rate": 1.0929161895723639e-05,
-      "loss": 0.2477,
+      "epoch": 3.5843514614192524,
+      "grad_norm": 0.21652553975582123,
+      "learning_rate": 9.810537413569107e-06,
+      "loss": 0.3874,
       "step": 99455
     },
     {
-      "epoch": 3.5,
-      "learning_rate": 1.0926807318712199e-05,
-      "loss": 0.2548,
+      "epoch": 3.584531661080477,
+      "grad_norm": 0.23006175458431244,
+      "learning_rate": 9.80821976129288e-06,
+      "loss": 0.3678,
       "step": 99460
     },
     {
-      "epoch": 3.5,
-      "learning_rate": 1.0924452924431895e-05,
-      "loss": 0.2588,
+      "epoch": 3.584711860741702,
+      "grad_norm": 0.22379863262176514,
+      "learning_rate": 9.805902316001712e-06,
+      "loss": 0.3814,
       "step": 99465
     },
     {
-      "epoch": 3.5,
-      "learning_rate": 1.0922098712913289e-05,
-      "loss": 0.2448,
+      "epoch": 3.584892060402926,
+      "grad_norm": 0.26252281665802,
+      "learning_rate": 9.803585077727195e-06,
+      "loss": 0.3827,
       "step": 99470
     },
     {
-      "epoch": 3.5,
-      "learning_rate": 1.0919744684186947e-05,
-      "loss": 0.2646,
+      "epoch": 3.585072260064151,
+      "grad_norm": 0.2835235893726349,
+      "learning_rate": 9.801268046500884e-06,
+      "loss": 0.3678,
       "step": 99475
     },
     {
-      "epoch": 3.5,
-      "learning_rate": 1.0917390838283442e-05,
-      "loss": 0.2531,
+      "epoch": 3.5852524597253757,
+      "grad_norm": 0.25666457414627075,
+      "learning_rate": 9.798951222354344e-06,
+      "loss": 0.3539,
       "step": 99480
     },
     {
-      "epoch": 3.5,
-      "learning_rate": 1.0915037175233345e-05,
-      "loss": 0.2822,
+      "epoch": 3.5854326593866004,
+      "grad_norm": 0.26046478748321533,
+      "learning_rate": 9.796634605319158e-06,
+      "loss": 0.3828,
       "step": 99485
     },
     {
-      "epoch": 3.5,
-      "learning_rate": 1.0912683695067205e-05,
-      "loss": 0.2679,
+      "epoch": 3.585612859047825,
+      "grad_norm": 0.23460224270820618,
+      "learning_rate": 9.794318195426882e-06,
+      "loss": 0.3791,
       "step": 99490
     },
     {
-      "epoch": 3.5,
-      "learning_rate": 1.0910330397815576e-05,
-      "loss": 0.2748,
+      "epoch": 3.58579305870905,
+      "grad_norm": 0.28312990069389343,
+      "learning_rate": 9.79200199270908e-06,
+      "loss": 0.4117,
       "step": 99495
     },
     {
-      "epoch": 3.5,
-      "learning_rate": 1.0907977283509032e-05,
-      "loss": 0.2566,
+      "epoch": 3.585973258370274,
+      "grad_norm": 0.2613278925418854,
+      "learning_rate": 9.7896859971973e-06,
+      "loss": 0.3978,
       "step": 99500
     },
     {
-      "epoch": 3.5,
-      "eval_loss": 0.25208380818367004,
-      "eval_runtime": 10.545,
-      "eval_samples_per_second": 9.483,
-      "eval_steps_per_second": 9.483,
+      "epoch": 3.585973258370274,
+      "eval_loss": 0.4298833906650543,
+      "eval_runtime": 3.5272,
+      "eval_samples_per_second": 28.351,
+      "eval_steps_per_second": 7.088,
       "step": 99500
     },
     {
-      "epoch": 3.5,
-      "learning_rate": 1.0905624352178115e-05,
-      "loss": 0.2539,
+      "epoch": 3.586153458031499,
+      "grad_norm": 0.2599956691265106,
+      "learning_rate": 9.787370208923099e-06,
+      "loss": 0.393,
       "step": 99505
     },
     {
-      "epoch": 3.5,
-      "learning_rate": 1.0903271603853376e-05,
-      "loss": 0.2521,
+      "epoch": 3.5863336576927236,
+      "grad_norm": 0.2776644229888916,
+      "learning_rate": 9.785054627918044e-06,
+      "loss": 0.4205,
       "step": 99510
     },
     {
-      "epoch": 3.5,
-      "learning_rate": 1.0900919038565357e-05,
-      "loss": 0.2631,
+      "epoch": 3.586513857353948,
+      "grad_norm": 0.27284055948257446,
+      "learning_rate": 9.782739254213672e-06,
+      "loss": 0.3999,
       "step": 99515
     },
     {
-      "epoch": 3.5,
-      "learning_rate": 1.0898566656344622e-05,
-      "loss": 0.269,
+      "epoch": 3.5866940570151726,
+      "grad_norm": 0.24026501178741455,
+      "learning_rate": 9.780424087841533e-06,
+      "loss": 0.378,
       "step": 99520
     },
     {
-      "epoch": 3.5,
-      "learning_rate": 1.0896214457221706e-05,
-      "loss": 0.2594,
+      "epoch": 3.5868742566763974,
+      "grad_norm": 0.26283571124076843,
+      "learning_rate": 9.778109128833166e-06,
+      "loss": 0.3963,
       "step": 99525
     },
     {
-      "epoch": 3.5,
-      "learning_rate": 1.089386244122714e-05,
-      "loss": 0.2525,
+      "epoch": 3.587054456337622,
+      "grad_norm": 0.22547613084316254,
+      "learning_rate": 9.77579437722011e-06,
+      "loss": 0.3896,
       "step": 99530
     },
     {
-      "epoch": 3.5,
-      "learning_rate": 1.0891510608391475e-05,
-      "loss": 0.2582,
+      "epoch": 3.587234655998847,
+      "grad_norm": 0.2306743711233139,
+      "learning_rate": 9.773479833033913e-06,
+      "loss": 0.4029,
       "step": 99535
     },
     {
-      "epoch": 3.5,
-      "learning_rate": 1.0889158958745257e-05,
-      "loss": 0.2852,
+      "epoch": 3.5874148556600716,
+      "grad_norm": 0.24043156206607819,
+      "learning_rate": 9.771165496306118e-06,
+      "loss": 0.3885,
       "step": 99540
     },
     {
-      "epoch": 3.5,
-      "learning_rate": 1.088680749231901e-05,
-      "loss": 0.2508,
+      "epoch": 3.587595055321296,
+      "grad_norm": 0.18582390248775482,
+      "learning_rate": 9.768851367068224e-06,
+      "loss": 0.3691,
       "step": 99545
     },
     {
-      "epoch": 3.5,
-      "learning_rate": 1.0884456209143257e-05,
-      "loss": 0.263,
+      "epoch": 3.5877752549825206,
+      "grad_norm": 0.24359643459320068,
+      "learning_rate": 9.766537445351792e-06,
+      "loss": 0.402,
       "step": 99550
     },
     {
-      "epoch": 3.5,
-      "learning_rate": 1.0882105109248547e-05,
-      "loss": 0.2546,
+      "epoch": 3.5879554546437453,
+      "grad_norm": 0.16479088366031647,
+      "learning_rate": 9.764223731188337e-06,
+      "loss": 0.398,
       "step": 99555
     },
     {
-      "epoch": 3.5,
-      "learning_rate": 1.0879754192665397e-05,
-      "loss": 0.2612,
+      "epoch": 3.58813565430497,
+      "grad_norm": 0.23543262481689453,
+      "learning_rate": 9.761910224609374e-06,
+      "loss": 0.3722,
       "step": 99560
     },
     {
-      "epoch": 3.5,
-      "learning_rate": 1.0877403459424335e-05,
-      "loss": 0.2559,
+      "epoch": 3.5883158539661943,
+      "grad_norm": 0.26243576407432556,
+      "learning_rate": 9.759596925646456e-06,
+      "loss": 0.3961,
       "step": 99565
     },
     {
-      "epoch": 3.5,
-      "learning_rate": 1.0875052909555872e-05,
-      "loss": 0.2799,
+      "epoch": 3.588496053627419,
+      "grad_norm": 0.28197166323661804,
+      "learning_rate": 9.757283834331057e-06,
+      "loss": 0.3743,
       "step": 99570
     },
     {
-      "epoch": 3.5,
-      "learning_rate": 1.0872702543090547e-05,
-      "loss": 0.2395,
+      "epoch": 3.588676253288644,
+      "grad_norm": 0.24312157928943634,
+      "learning_rate": 9.754970950694725e-06,
+      "loss": 0.4079,
       "step": 99575
     },
     {
-      "epoch": 3.5,
-      "learning_rate": 1.087035236005887e-05,
-      "loss": 0.251,
+      "epoch": 3.5888564529498685,
+      "grad_norm": 0.21838116645812988,
+      "learning_rate": 9.752658274768964e-06,
+      "loss": 0.365,
       "step": 99580
     },
     {
-      "epoch": 3.5,
-      "learning_rate": 1.0868002360491344e-05,
-      "loss": 0.2693,
+      "epoch": 3.5890366526110933,
+      "grad_norm": 0.2255815863609314,
+      "learning_rate": 9.75034580658528e-06,
+      "loss": 0.3944,
       "step": 99585
     },
     {
-      "epoch": 3.5,
-      "learning_rate": 1.0865652544418506e-05,
-      "loss": 0.2612,
+      "epoch": 3.5892168522723176,
+      "grad_norm": 0.2602185308933258,
+      "learning_rate": 9.748033546175182e-06,
+      "loss": 0.3865,
       "step": 99590
     },
     {
-      "epoch": 3.5,
-      "learning_rate": 1.0863302911870846e-05,
-      "loss": 0.2632,
+      "epoch": 3.5893970519335423,
+      "grad_norm": 0.1989058256149292,
+      "learning_rate": 9.745721493570176e-06,
+      "loss": 0.3411,
       "step": 99595
     },
     {
-      "epoch": 3.5,
-      "learning_rate": 1.0860953462878889e-05,
-      "loss": 0.2443,
+      "epoch": 3.589577251594767,
+      "grad_norm": 0.25254446268081665,
+      "learning_rate": 9.743409648801749e-06,
+      "loss": 0.389,
       "step": 99600
     },
     {
-      "epoch": 3.5,
-      "learning_rate": 1.0858604197473135e-05,
-      "loss": 0.2425,
+      "epoch": 3.5897574512559918,
+      "grad_norm": 0.2625841498374939,
+      "learning_rate": 9.741098011901423e-06,
+      "loss": 0.379,
       "step": 99605
     },
     {
-      "epoch": 3.5,
-      "learning_rate": 1.0856255115684075e-05,
-      "loss": 0.2392,
+      "epoch": 3.589937650917216,
+      "grad_norm": 0.215394988656044,
+      "learning_rate": 9.738786582900684e-06,
+      "loss": 0.3997,
       "step": 99610
     },
     {
-      "epoch": 3.5,
-      "learning_rate": 1.0853906217542231e-05,
-      "loss": 0.2635,
+      "epoch": 3.590117850578441,
+      "grad_norm": 0.2351488173007965,
+      "learning_rate": 9.736475361831019e-06,
+      "loss": 0.3684,
       "step": 99615
     },
     {
-      "epoch": 3.5,
-      "learning_rate": 1.085155750307809e-05,
-      "loss": 0.2571,
+      "epoch": 3.5902980502396655,
+      "grad_norm": 0.2385423183441162,
+      "learning_rate": 9.734164348723922e-06,
+      "loss": 0.3604,
       "step": 99620
     },
     {
-      "epoch": 3.51,
-      "learning_rate": 1.0849208972322156e-05,
-      "loss": 0.2595,
+      "epoch": 3.5904782499008903,
+      "grad_norm": 0.17893251776695251,
+      "learning_rate": 9.731853543610876e-06,
+      "loss": 0.3638,
       "step": 99625
     },
     {
-      "epoch": 3.51,
-      "learning_rate": 1.0846860625304906e-05,
-      "loss": 0.2321,
+      "epoch": 3.590658449562115,
+      "grad_norm": 0.23732203245162964,
+      "learning_rate": 9.729542946523376e-06,
+      "loss": 0.3772,
       "step": 99630
     },
     {
-      "epoch": 3.51,
-      "learning_rate": 1.084451246205685e-05,
-      "loss": 0.2596,
+      "epoch": 3.5908386492233397,
+      "grad_norm": 0.21697406470775604,
+      "learning_rate": 9.727232557492896e-06,
+      "loss": 0.3953,
       "step": 99635
     },
     {
-      "epoch": 3.51,
-      "learning_rate": 1.0842164482608463e-05,
-      "loss": 0.2488,
+      "epoch": 3.591018848884564,
+      "grad_norm": 0.21853552758693695,
+      "learning_rate": 9.724922376550915e-06,
+      "loss": 0.3793,
       "step": 99640
     },
     {
-      "epoch": 3.51,
-      "learning_rate": 1.083981668699025e-05,
-      "loss": 0.2511,
+      "epoch": 3.5911990485457888,
+      "grad_norm": 0.23639477789402008,
+      "learning_rate": 9.722612403728912e-06,
+      "loss": 0.4109,
       "step": 99645
     },
     {
-      "epoch": 3.51,
-      "learning_rate": 1.0837469075232678e-05,
-      "loss": 0.2557,
+      "epoch": 3.5913792482070135,
+      "grad_norm": 0.26770761609077454,
+      "learning_rate": 9.720302639058349e-06,
+      "loss": 0.3729,
       "step": 99650
     },
     {
-      "epoch": 3.51,
-      "learning_rate": 1.0835121647366243e-05,
-      "loss": 0.2735,
+      "epoch": 3.5915594478682378,
+      "grad_norm": 0.21606463193893433,
+      "learning_rate": 9.717993082570716e-06,
+      "loss": 0.3727,
       "step": 99655
     },
     {
-      "epoch": 3.51,
-      "learning_rate": 1.083277440342142e-05,
-      "loss": 0.2476,
+      "epoch": 3.5917396475294625,
+      "grad_norm": 0.24439482390880585,
+      "learning_rate": 9.715683734297466e-06,
+      "loss": 0.3812,
       "step": 99660
     },
     {
-      "epoch": 3.51,
-      "learning_rate": 1.0830427343428673e-05,
-      "loss": 0.2678,
+      "epoch": 3.5919198471906872,
+      "grad_norm": 0.24273943901062012,
+      "learning_rate": 9.713374594270055e-06,
+      "loss": 0.4021,
       "step": 99665
     },
     {
-      "epoch": 3.51,
-      "learning_rate": 1.08280804674185e-05,
-      "loss": 0.2673,
+      "epoch": 3.592100046851912,
+      "grad_norm": 0.2814248502254486,
+      "learning_rate": 9.711065662519964e-06,
+      "loss": 0.385,
       "step": 99670
     },
     {
-      "epoch": 3.51,
-      "learning_rate": 1.0825733775421362e-05,
-      "loss": 0.2607,
+      "epoch": 3.5922802465131367,
+      "grad_norm": 0.23897801339626312,
+      "learning_rate": 9.708756939078634e-06,
+      "loss": 0.3566,
       "step": 99675
     },
     {
-      "epoch": 3.51,
-      "learning_rate": 1.0823387267467733e-05,
-      "loss": 0.2653,
+      "epoch": 3.5924604461743614,
+      "grad_norm": 0.24239744246006012,
+      "learning_rate": 9.70644842397755e-06,
+      "loss": 0.4005,
       "step": 99680
     },
     {
-      "epoch": 3.51,
-      "learning_rate": 1.0821040943588066e-05,
-      "loss": 0.2476,
+      "epoch": 3.5926406458355857,
+      "grad_norm": 0.2517867684364319,
+      "learning_rate": 9.704140117248134e-06,
+      "loss": 0.3916,
       "step": 99685
     },
     {
-      "epoch": 3.51,
-      "learning_rate": 1.0818694803812846e-05,
-      "loss": 0.2518,
+      "epoch": 3.5928208454968105,
+      "grad_norm": 0.315687894821167,
+      "learning_rate": 9.701832018921839e-06,
+      "loss": 0.3683,
       "step": 99690
     },
     {
-      "epoch": 3.51,
-      "learning_rate": 1.0816348848172528e-05,
-      "loss": 0.2531,
+      "epoch": 3.593001045158035,
+      "grad_norm": 0.29640886187553406,
+      "learning_rate": 9.69952412903013e-06,
+      "loss": 0.375,
       "step": 99695
     },
     {
-      "epoch": 3.51,
-      "learning_rate": 1.0814003076697565e-05,
-      "loss": 0.2656,
+      "epoch": 3.5931812448192595,
+      "grad_norm": 0.22682395577430725,
+      "learning_rate": 9.697216447604444e-06,
+      "loss": 0.3987,
       "step": 99700
     },
     {
-      "epoch": 3.51,
-      "learning_rate": 1.0811657489418423e-05,
-      "loss": 0.2415,
+      "epoch": 3.5933614444804842,
+      "grad_norm": 0.30215203762054443,
+      "learning_rate": 9.69490897467622e-06,
+      "loss": 0.3749,
       "step": 99705
     },
     {
-      "epoch": 3.51,
-      "learning_rate": 1.0809312086365566e-05,
-      "loss": 0.2528,
+      "epoch": 3.593541644141709,
+      "grad_norm": 0.23141373693943024,
+      "learning_rate": 9.692601710276897e-06,
+      "loss": 0.3501,
       "step": 99710
     },
     {
-      "epoch": 3.51,
-      "learning_rate": 1.0806966867569441e-05,
-      "loss": 0.2445,
+      "epoch": 3.5937218438029337,
+      "grad_norm": 0.238219752907753,
+      "learning_rate": 9.690294654437907e-06,
+      "loss": 0.3775,
       "step": 99715
     },
     {
-      "epoch": 3.51,
-      "learning_rate": 1.0804621833060496e-05,
-      "loss": 0.2759,
+      "epoch": 3.5939020434641584,
+      "grad_norm": 0.1689501702785492,
+      "learning_rate": 9.687987807190693e-06,
+      "loss": 0.3641,
       "step": 99720
     },
     {
-      "epoch": 3.51,
-      "learning_rate": 1.0802276982869176e-05,
-      "loss": 0.2405,
+      "epoch": 3.594082243125383,
+      "grad_norm": 0.21029123663902283,
+      "learning_rate": 9.685681168566683e-06,
+      "loss": 0.3353,
       "step": 99725
     },
     {
-      "epoch": 3.51,
-      "learning_rate": 1.0799932317025937e-05,
-      "loss": 0.262,
+      "epoch": 3.5942624427866074,
+      "grad_norm": 0.22430087625980377,
+      "learning_rate": 9.6833747385973e-06,
+      "loss": 0.3909,
       "step": 99730
     },
     {
-      "epoch": 3.51,
-      "learning_rate": 1.079758783556122e-05,
-      "loss": 0.2699,
+      "epoch": 3.594442642447832,
+      "grad_norm": 0.21610116958618164,
+      "learning_rate": 9.681068517313973e-06,
+      "loss": 0.3581,
       "step": 99735
     },
     {
-      "epoch": 3.51,
-      "learning_rate": 1.0795243538505468e-05,
-      "loss": 0.2695,
+      "epoch": 3.594622842109057,
+      "grad_norm": 0.2914960980415344,
+      "learning_rate": 9.678762504748121e-06,
+      "loss": 0.3932,
       "step": 99740
     },
     {
-      "epoch": 3.51,
-      "learning_rate": 1.0792899425889105e-05,
-      "loss": 0.2483,
+      "epoch": 3.594803041770281,
+      "grad_norm": 0.24280034005641937,
+      "learning_rate": 9.676456700931152e-06,
+      "loss": 0.3773,
       "step": 99745
     },
     {
-      "epoch": 3.51,
-      "learning_rate": 1.079055549774259e-05,
-      "loss": 0.2657,
+      "epoch": 3.594983241431506,
+      "grad_norm": 0.2517662048339844,
+      "learning_rate": 9.674151105894516e-06,
+      "loss": 0.3723,
       "step": 99750
     },
     {
-      "epoch": 3.51,
-      "learning_rate": 1.078821175409634e-05,
-      "loss": 0.2653,
+      "epoch": 3.5951634410927307,
+      "grad_norm": 0.23000794649124146,
+      "learning_rate": 9.671845719669584e-06,
+      "loss": 0.3735,
       "step": 99755
     },
     {
-      "epoch": 3.51,
-      "learning_rate": 1.0785868194980803e-05,
-      "loss": 0.2405,
+      "epoch": 3.5953436407539554,
+      "grad_norm": 0.2036644071340561,
+      "learning_rate": 9.669540542287795e-06,
+      "loss": 0.3355,
       "step": 99760
     },
     {
-      "epoch": 3.51,
-      "learning_rate": 1.078352482042639e-05,
-      "loss": 0.268,
+      "epoch": 3.59552384041518,
+      "grad_norm": 0.210072860121727,
+      "learning_rate": 9.667235573780547e-06,
+      "loss": 0.3654,
       "step": 99765
     },
     {
-      "epoch": 3.51,
-      "learning_rate": 1.0781181630463547e-05,
-      "loss": 0.2504,
+      "epoch": 3.595704040076405,
+      "grad_norm": 0.23076112568378448,
+      "learning_rate": 9.664930814179248e-06,
+      "loss": 0.381,
       "step": 99770
     },
     {
-      "epoch": 3.51,
-      "learning_rate": 1.0778838625122692e-05,
-      "loss": 0.2692,
+      "epoch": 3.595884239737629,
+      "grad_norm": 0.22967948019504547,
+      "learning_rate": 9.662626263515298e-06,
+      "loss": 0.3657,
       "step": 99775
     },
     {
-      "epoch": 3.51,
-      "learning_rate": 1.0776495804434233e-05,
-      "loss": 0.2375,
+      "epoch": 3.596064439398854,
+      "grad_norm": 0.21021999418735504,
+      "learning_rate": 9.660321921820095e-06,
+      "loss": 0.3699,
       "step": 99780
     },
     {
-      "epoch": 3.51,
-      "learning_rate": 1.0774153168428616e-05,
-      "loss": 0.2549,
+      "epoch": 3.5962446390600786,
+      "grad_norm": 0.2712535858154297,
+      "learning_rate": 9.658017789125026e-06,
+      "loss": 0.4032,
       "step": 99785
     },
     {
-      "epoch": 3.51,
-      "learning_rate": 1.0771810717136242e-05,
-      "loss": 0.2362,
+      "epoch": 3.5964248387213034,
+      "grad_norm": 0.19251175224781036,
+      "learning_rate": 9.655713865461505e-06,
+      "loss": 0.3645,
       "step": 99790
     },
     {
-      "epoch": 3.51,
-      "learning_rate": 1.0769468450587528e-05,
-      "loss": 0.2604,
+      "epoch": 3.5966050383825277,
+      "grad_norm": 0.22021165490150452,
+      "learning_rate": 9.653410150860912e-06,
+      "loss": 0.3683,
       "step": 99795
     },
     {
-      "epoch": 3.51,
-      "learning_rate": 1.076712636881288e-05,
-      "loss": 0.2494,
+      "epoch": 3.5967852380437524,
+      "grad_norm": 0.20990362763404846,
+      "learning_rate": 9.651106645354632e-06,
+      "loss": 0.3702,
       "step": 99800
     },
     {
-      "epoch": 3.51,
-      "learning_rate": 1.0764784471842726e-05,
-      "loss": 0.2472,
+      "epoch": 3.596965437704977,
+      "grad_norm": 0.25066566467285156,
+      "learning_rate": 9.648803348974054e-06,
+      "loss": 0.3669,
       "step": 99805
     },
     {
-      "epoch": 3.51,
-      "learning_rate": 1.0762442759707456e-05,
-      "loss": 0.233,
+      "epoch": 3.597145637366202,
+      "grad_norm": 0.21980679035186768,
+      "learning_rate": 9.64650026175055e-06,
+      "loss": 0.3826,
       "step": 99810
     },
     {
-      "epoch": 3.51,
-      "learning_rate": 1.0760101232437492e-05,
-      "loss": 0.2798,
+      "epoch": 3.5973258370274266,
+      "grad_norm": 0.20478756725788116,
+      "learning_rate": 9.644197383715514e-06,
+      "loss": 0.3417,
       "step": 99815
     },
     {
-      "epoch": 3.51,
-      "learning_rate": 1.0757759890063219e-05,
-      "loss": 0.253,
+      "epoch": 3.597506036688651,
+      "grad_norm": 0.2611001431941986,
+      "learning_rate": 9.641894714900316e-06,
+      "loss": 0.3901,
       "step": 99820
     },
     {
-      "epoch": 3.51,
-      "learning_rate": 1.0755418732615059e-05,
-      "loss": 0.243,
+      "epoch": 3.5976862363498756,
+      "grad_norm": 0.2687753140926361,
+      "learning_rate": 9.63959225533633e-06,
+      "loss": 0.3791,
       "step": 99825
     },
     {
-      "epoch": 3.51,
-      "learning_rate": 1.0753077760123399e-05,
-      "loss": 0.2646,
+      "epoch": 3.5978664360111003,
+      "grad_norm": 0.20144647359848022,
+      "learning_rate": 9.637290005054928e-06,
+      "loss": 0.3827,
       "step": 99830
     },
     {
-      "epoch": 3.51,
-      "learning_rate": 1.0750736972618633e-05,
-      "loss": 0.2589,
+      "epoch": 3.598046635672325,
+      "grad_norm": 0.2408836930990219,
+      "learning_rate": 9.634987964087464e-06,
+      "loss": 0.3959,
       "step": 99835
     },
     {
-      "epoch": 3.51,
-      "learning_rate": 1.0748396370131147e-05,
-      "loss": 0.2586,
+      "epoch": 3.5982268353335494,
+      "grad_norm": 0.23702372610569,
+      "learning_rate": 9.632686132465321e-06,
+      "loss": 0.404,
       "step": 99840
     },
     {
-      "epoch": 3.51,
-      "learning_rate": 1.0746055952691353e-05,
-      "loss": 0.2589,
+      "epoch": 3.598407034994774,
+      "grad_norm": 0.2729646563529968,
+      "learning_rate": 9.630384510219867e-06,
+      "loss": 0.3612,
       "step": 99845
     },
     {
-      "epoch": 3.51,
-      "learning_rate": 1.0743715720329628e-05,
-      "loss": 0.2689,
+      "epoch": 3.598587234655999,
+      "grad_norm": 0.25669410824775696,
+      "learning_rate": 9.628083097382428e-06,
+      "loss": 0.373,
       "step": 99850
     },
     {
-      "epoch": 3.51,
-      "learning_rate": 1.074137567307636e-05,
-      "loss": 0.257,
+      "epoch": 3.5987674343172236,
+      "grad_norm": 0.18720833957195282,
+      "learning_rate": 9.625781893984392e-06,
+      "loss": 0.3708,
       "step": 99855
     },
     {
-      "epoch": 3.51,
-      "learning_rate": 1.0739035810961923e-05,
-      "loss": 0.2485,
+      "epoch": 3.5989476339784483,
+      "grad_norm": 0.2443101853132248,
+      "learning_rate": 9.623480900057092e-06,
+      "loss": 0.38,
       "step": 99860
     },
     {
-      "epoch": 3.51,
-      "learning_rate": 1.0736696134016708e-05,
-      "loss": 0.2514,
+      "epoch": 3.5991278336396726,
+      "grad_norm": 0.24887430667877197,
+      "learning_rate": 9.621180115631904e-06,
+      "loss": 0.3652,
       "step": 99865
     },
     {
-      "epoch": 3.51,
-      "learning_rate": 1.07343566422711e-05,
-      "loss": 0.2608,
+      "epoch": 3.5993080333008973,
+      "grad_norm": 0.22554521262645721,
+      "learning_rate": 9.61887954074015e-06,
+      "loss": 0.3778,
       "step": 99870
     },
     {
-      "epoch": 3.51,
-      "learning_rate": 1.0732017335755473e-05,
-      "loss": 0.2521,
+      "epoch": 3.599488232962122,
+      "grad_norm": 0.2344188541173935,
+      "learning_rate": 9.616579175413176e-06,
+      "loss": 0.3619,
       "step": 99875
     },
     {
-      "epoch": 3.51,
-      "learning_rate": 1.0729678214500189e-05,
-      "loss": 0.2462,
+      "epoch": 3.599668432623347,
+      "grad_norm": 0.30154213309288025,
+      "learning_rate": 9.614279019682343e-06,
+      "loss": 0.4012,
       "step": 99880
     },
     {
-      "epoch": 3.51,
-      "learning_rate": 1.0727339278535636e-05,
-      "loss": 0.2633,
+      "epoch": 3.599848632284571,
+      "grad_norm": 0.2288644164800644,
+      "learning_rate": 9.611979073578977e-06,
+      "loss": 0.3882,
       "step": 99885
     },
     {
-      "epoch": 3.51,
-      "learning_rate": 1.072500052789218e-05,
-      "loss": 0.2736,
+      "epoch": 3.600028831945796,
+      "grad_norm": 0.20435230433940887,
+      "learning_rate": 9.60967933713442e-06,
+      "loss": 0.408,
       "step": 99890
     },
     {
-      "epoch": 3.51,
-      "learning_rate": 1.0722661962600169e-05,
-      "loss": 0.2583,
+      "epoch": 3.6002090316070205,
+      "grad_norm": 0.24859078228473663,
+      "learning_rate": 9.607379810379998e-06,
+      "loss": 0.3833,
       "step": 99895
     },
     {
-      "epoch": 3.51,
-      "learning_rate": 1.0720323582689998e-05,
-      "loss": 0.2506,
+      "epoch": 3.6003892312682453,
+      "grad_norm": 0.28969287872314453,
+      "learning_rate": 9.605080493347041e-06,
+      "loss": 0.3902,
       "step": 99900
     },
     {
-      "epoch": 3.51,
-      "learning_rate": 1.0717985388192012e-05,
-      "loss": 0.2657,
+      "epoch": 3.60056943092947,
+      "grad_norm": 0.2676686644554138,
+      "learning_rate": 9.602781386066889e-06,
+      "loss": 0.3858,
       "step": 99905
     },
     {
-      "epoch": 3.52,
-      "learning_rate": 1.0715647379136575e-05,
-      "loss": 0.2707,
+      "epoch": 3.6007496305906947,
+      "grad_norm": 0.2682179808616638,
+      "learning_rate": 9.600482488570862e-06,
+      "loss": 0.3932,
       "step": 99910
     },
     {
-      "epoch": 3.52,
-      "learning_rate": 1.0713309555554032e-05,
-      "loss": 0.2525,
+      "epoch": 3.600929830251919,
+      "grad_norm": 0.21893753111362457,
+      "learning_rate": 9.598183800890276e-06,
+      "loss": 0.3843,
       "step": 99915
     },
     {
-      "epoch": 3.52,
-      "learning_rate": 1.0710971917474752e-05,
-      "loss": 0.2493,
+      "epoch": 3.6011100299131438,
+      "grad_norm": 0.2323058843612671,
+      "learning_rate": 9.595885323056455e-06,
+      "loss": 0.3797,
       "step": 99920
     },
     {
-      "epoch": 3.52,
-      "learning_rate": 1.0708634464929091e-05,
-      "loss": 0.2662,
+      "epoch": 3.6012902295743685,
+      "grad_norm": 0.23504719138145447,
+      "learning_rate": 9.593587055100717e-06,
+      "loss": 0.3858,
       "step": 99925
     },
     {
-      "epoch": 3.52,
-      "learning_rate": 1.0706297197947393e-05,
-      "loss": 0.2626,
+      "epoch": 3.601470429235593,
+      "grad_norm": 0.2661448121070862,
+      "learning_rate": 9.591288997054359e-06,
+      "loss": 0.3477,
       "step": 99930
     },
     {
-      "epoch": 3.52,
-      "learning_rate": 1.0703960116559999e-05,
-      "loss": 0.2644,
+      "epoch": 3.6016506288968175,
+      "grad_norm": 0.261135458946228,
+      "learning_rate": 9.588991148948726e-06,
+      "loss": 0.3728,
       "step": 99935
     },
     {
-      "epoch": 3.52,
-      "learning_rate": 1.0701623220797272e-05,
-      "loss": 0.256,
+      "epoch": 3.6018308285580423,
+      "grad_norm": 0.2633829712867737,
+      "learning_rate": 9.586693510815088e-06,
+      "loss": 0.3891,
       "step": 99940
     },
     {
-      "epoch": 3.52,
-      "learning_rate": 1.0699286510689541e-05,
-      "loss": 0.2534,
+      "epoch": 3.602011028219267,
+      "grad_norm": 0.23481690883636475,
+      "learning_rate": 9.584396082684775e-06,
+      "loss": 0.3821,
       "step": 99945
     },
     {
-      "epoch": 3.52,
-      "learning_rate": 1.0696949986267152e-05,
-      "loss": 0.2299,
+      "epoch": 3.6021912278804917,
+      "grad_norm": 0.23687255382537842,
+      "learning_rate": 9.582098864589078e-06,
+      "loss": 0.3825,
       "step": 99950
     },
     {
-      "epoch": 3.52,
-      "learning_rate": 1.069461364756043e-05,
-      "loss": 0.2468,
+      "epoch": 3.6023714275417165,
+      "grad_norm": 0.24096225202083588,
+      "learning_rate": 9.579801856559292e-06,
+      "loss": 0.35,
       "step": 99955
     },
     {
-      "epoch": 3.52,
-      "learning_rate": 1.0692277494599734e-05,
-      "loss": 0.2435,
+      "epoch": 3.6025516272029408,
+      "grad_norm": 0.23858147859573364,
+      "learning_rate": 9.577505058626738e-06,
+      "loss": 0.3736,
       "step": 99960
     },
     {
-      "epoch": 3.52,
-      "learning_rate": 1.0689941527415384e-05,
-      "loss": 0.2582,
+      "epoch": 3.6027318268641655,
+      "grad_norm": 0.24173545837402344,
+      "learning_rate": 9.575208470822682e-06,
+      "loss": 0.3797,
       "step": 99965
     },
     {
-      "epoch": 3.52,
-      "learning_rate": 1.0687605746037705e-05,
-      "loss": 0.2606,
+      "epoch": 3.60291202652539,
+      "grad_norm": 0.19363081455230713,
+      "learning_rate": 9.572912093178419e-06,
+      "loss": 0.3779,
       "step": 99970
     },
     {
-      "epoch": 3.52,
-      "learning_rate": 1.0685270150497032e-05,
-      "loss": 0.2559,
+      "epoch": 3.6030922261866145,
+      "grad_norm": 0.2636258006095886,
+      "learning_rate": 9.570615925725246e-06,
+      "loss": 0.4388,
       "step": 99975
     },
     {
-      "epoch": 3.52,
-      "learning_rate": 1.06829347408237e-05,
-      "loss": 0.2676,
+      "epoch": 3.6032724258478392,
+      "grad_norm": 0.2344840168952942,
+      "learning_rate": 9.568319968494446e-06,
+      "loss": 0.3784,
       "step": 99980
     },
     {
-      "epoch": 3.52,
-      "learning_rate": 1.0681066546929894e-05,
-      "loss": 0.2738,
+      "epoch": 3.603452625509064,
+      "grad_norm": 0.22249801456928253,
+      "learning_rate": 9.5660242215173e-06,
+      "loss": 0.3607,
       "step": 99985
     },
     {
-      "epoch": 3.52,
-      "learning_rate": 1.0678731471894177e-05,
-      "loss": 0.2751,
+      "epoch": 3.6036328251702887,
+      "grad_norm": 0.19639311730861664,
+      "learning_rate": 9.563728684825082e-06,
+      "loss": 0.3788,
       "step": 99990
     },
     {
-      "epoch": 3.52,
-      "learning_rate": 1.0676396582810693e-05,
-      "loss": 0.268,
+      "epoch": 3.6038130248315134,
+      "grad_norm": 0.2106081247329712,
+      "learning_rate": 9.561433358449068e-06,
+      "loss": 0.3919,
       "step": 99995
     },
     {
-      "epoch": 3.52,
-      "learning_rate": 1.0674061879709762e-05,
-      "loss": 0.2599,
+      "epoch": 3.603993224492738,
+      "grad_norm": 0.2347351461648941,
+      "learning_rate": 9.559138242420544e-06,
+      "loss": 0.3608,
       "step": 100000
     },
     {
-      "epoch": 3.52,
-      "eval_loss": 0.2520838975906372,
-      "eval_runtime": 10.5365,
-      "eval_samples_per_second": 9.491,
-      "eval_steps_per_second": 9.491,
+      "epoch": 3.603993224492738,
+      "eval_loss": 0.42966705560684204,
+      "eval_runtime": 3.5294,
+      "eval_samples_per_second": 28.334,
+      "eval_steps_per_second": 7.083,
       "step": 100000
     },
     {
-      "epoch": 3.52,
-      "learning_rate": 1.0671727362621686e-05,
-      "loss": 0.2531,
+      "epoch": 3.6041734241539625,
+      "grad_norm": 0.23393255472183228,
+      "learning_rate": 9.556843336770768e-06,
+      "loss": 0.3584,
       "step": 100005
     },
     {
-      "epoch": 3.52,
-      "learning_rate": 1.0669393031576788e-05,
-      "loss": 0.2713,
+      "epoch": 3.604353623815187,
+      "grad_norm": 0.22265978157520294,
+      "learning_rate": 9.554548641531014e-06,
+      "loss": 0.355,
       "step": 100010
     },
     {
-      "epoch": 3.52,
-      "learning_rate": 1.0667058886605388e-05,
-      "loss": 0.2459,
+      "epoch": 3.604533823476412,
+      "grad_norm": 0.2124675214290619,
+      "learning_rate": 9.552254156732543e-06,
+      "loss": 0.3352,
       "step": 100015
     },
     {
-      "epoch": 3.52,
-      "learning_rate": 1.0664724927737783e-05,
-      "loss": 0.2487,
+      "epoch": 3.6047140231376362,
+      "grad_norm": 0.20015810430049896,
+      "learning_rate": 9.549959882406611e-06,
+      "loss": 0.3809,
       "step": 100020
     },
     {
-      "epoch": 3.52,
-      "learning_rate": 1.0662391155004265e-05,
-      "loss": 0.2499,
+      "epoch": 3.604894222798861,
+      "grad_norm": 0.24774090945720673,
+      "learning_rate": 9.547665818584491e-06,
+      "loss": 0.3974,
       "step": 100025
     },
     {
-      "epoch": 3.52,
-      "learning_rate": 1.0660057568435158e-05,
-      "loss": 0.2693,
+      "epoch": 3.6050744224600857,
+      "grad_norm": 0.23569168150424957,
+      "learning_rate": 9.545371965297445e-06,
+      "loss": 0.3889,
       "step": 100030
     },
     {
-      "epoch": 3.52,
-      "learning_rate": 1.0657724168060756e-05,
-      "loss": 0.282,
+      "epoch": 3.6052546221213104,
+      "grad_norm": 0.25077682733535767,
+      "learning_rate": 9.543078322576696e-06,
+      "loss": 0.3749,
       "step": 100035
     },
     {
-      "epoch": 3.52,
-      "learning_rate": 1.0655390953911352e-05,
-      "loss": 0.2716,
+      "epoch": 3.605434821782535,
+      "grad_norm": 0.30293580889701843,
+      "learning_rate": 9.54078489045352e-06,
+      "loss": 0.3622,
       "step": 100040
     },
     {
-      "epoch": 3.52,
-      "learning_rate": 1.0653057926017235e-05,
-      "loss": 0.2725,
+      "epoch": 3.60561502144376,
+      "grad_norm": 0.23128095269203186,
+      "learning_rate": 9.538491668959146e-06,
+      "loss": 0.3876,
       "step": 100045
     },
     {
-      "epoch": 3.52,
-      "learning_rate": 1.0650725084408713e-05,
-      "loss": 0.2495,
+      "epoch": 3.605795221104984,
+      "grad_norm": 0.20295816659927368,
+      "learning_rate": 9.536198658124849e-06,
+      "loss": 0.3585,
       "step": 100050
     },
     {
-      "epoch": 3.52,
-      "learning_rate": 1.0648392429116072e-05,
-      "loss": 0.232,
+      "epoch": 3.605975420766209,
+      "grad_norm": 0.23228393495082855,
+      "learning_rate": 9.533905857981843e-06,
+      "loss": 0.4119,
       "step": 100055
     },
     {
-      "epoch": 3.52,
-      "learning_rate": 1.0646059960169585e-05,
-      "loss": 0.2373,
+      "epoch": 3.6061556204274337,
+      "grad_norm": 0.2623964548110962,
+      "learning_rate": 9.531613268561365e-06,
+      "loss": 0.3794,
       "step": 100060
     },
     {
-      "epoch": 3.52,
-      "learning_rate": 1.0643727677599553e-05,
-      "loss": 0.2393,
+      "epoch": 3.6063358200886584,
+      "grad_norm": 0.21798618137836456,
+      "learning_rate": 9.529320889894671e-06,
+      "loss": 0.391,
       "step": 100065
     },
     {
-      "epoch": 3.52,
-      "learning_rate": 1.0641395581436264e-05,
-      "loss": 0.2558,
+      "epoch": 3.6065160197498827,
+      "grad_norm": 0.21800731122493744,
+      "learning_rate": 9.527028722012985e-06,
+      "loss": 0.3605,
       "step": 100070
     },
     {
-      "epoch": 3.52,
-      "learning_rate": 1.063906367170999e-05,
-      "loss": 0.2498,
+      "epoch": 3.6066962194111074,
+      "grad_norm": 0.2361879199743271,
+      "learning_rate": 9.524736764947537e-06,
+      "loss": 0.3766,
       "step": 100075
     },
     {
-      "epoch": 3.52,
-      "learning_rate": 1.0636731948451001e-05,
-      "loss": 0.2549,
+      "epoch": 3.606876419072332,
+      "grad_norm": 0.23150278627872467,
+      "learning_rate": 9.522445018729556e-06,
+      "loss": 0.3902,
       "step": 100080
     },
     {
-      "epoch": 3.52,
-      "learning_rate": 1.063440041168959e-05,
-      "loss": 0.2481,
+      "epoch": 3.607056618733557,
+      "grad_norm": 0.25107669830322266,
+      "learning_rate": 9.520153483390253e-06,
+      "loss": 0.3765,
       "step": 100085
     },
     {
-      "epoch": 3.52,
-      "learning_rate": 1.0632069061456024e-05,
-      "loss": 0.2654,
+      "epoch": 3.6072368183947816,
+      "grad_norm": 0.22123172879219055,
+      "learning_rate": 9.517862158960873e-06,
+      "loss": 0.3759,
       "step": 100090
     },
     {
-      "epoch": 3.52,
-      "learning_rate": 1.0629737897780574e-05,
-      "loss": 0.2545,
+      "epoch": 3.607417018056006,
+      "grad_norm": 0.2164187878370285,
+      "learning_rate": 9.515571045472623e-06,
+      "loss": 0.3721,
       "step": 100095
     },
     {
-      "epoch": 3.52,
-      "learning_rate": 1.0627406920693494e-05,
-      "loss": 0.2416,
+      "epoch": 3.6075972177172306,
+      "grad_norm": 0.23865751922130585,
+      "learning_rate": 9.513280142956718e-06,
+      "loss": 0.3667,
       "step": 100100
     },
     {
-      "epoch": 3.52,
-      "learning_rate": 1.0625076130225075e-05,
-      "loss": 0.2567,
+      "epoch": 3.6077774173784554,
+      "grad_norm": 0.22169511020183563,
+      "learning_rate": 9.510989451444374e-06,
+      "loss": 0.3906,
       "step": 100105
     },
     {
-      "epoch": 3.52,
-      "learning_rate": 1.0622745526405566e-05,
-      "loss": 0.2655,
+      "epoch": 3.60795761703968,
+      "grad_norm": 0.2418065220117569,
+      "learning_rate": 9.5086989709668e-06,
+      "loss": 0.3893,
       "step": 100110
     },
     {
-      "epoch": 3.52,
-      "learning_rate": 1.0620415109265228e-05,
-      "loss": 0.2459,
+      "epoch": 3.6081378167009044,
+      "grad_norm": 0.274650514125824,
+      "learning_rate": 9.506408701555197e-06,
+      "loss": 0.3756,
       "step": 100115
     },
     {
-      "epoch": 3.52,
-      "learning_rate": 1.061808487883432e-05,
-      "loss": 0.2647,
+      "epoch": 3.608318016362129,
+      "grad_norm": 0.25587350130081177,
+      "learning_rate": 9.504118643240781e-06,
+      "loss": 0.3975,
       "step": 100120
     },
     {
-      "epoch": 3.52,
-      "learning_rate": 1.061575483514311e-05,
-      "loss": 0.2402,
+      "epoch": 3.608498216023354,
+      "grad_norm": 0.35312119126319885,
+      "learning_rate": 9.501828796054751e-06,
+      "loss": 0.3939,
       "step": 100125
     },
     {
-      "epoch": 3.52,
-      "learning_rate": 1.0613424978221845e-05,
-      "loss": 0.2637,
+      "epoch": 3.6086784156845786,
+      "grad_norm": 0.24378515779972076,
+      "learning_rate": 9.499539160028301e-06,
+      "loss": 0.4169,
       "step": 100130
     },
     {
-      "epoch": 3.52,
-      "learning_rate": 1.0611095308100774e-05,
-      "loss": 0.2624,
+      "epoch": 3.6088586153458033,
+      "grad_norm": 0.2463681548833847,
+      "learning_rate": 9.497249735192628e-06,
+      "loss": 0.3423,
       "step": 100135
     },
     {
-      "epoch": 3.52,
-      "learning_rate": 1.060876582481014e-05,
-      "loss": 0.2645,
+      "epoch": 3.609038815007028,
+      "grad_norm": 0.2869650423526764,
+      "learning_rate": 9.494960521578922e-06,
+      "loss": 0.3543,
       "step": 100140
     },
     {
-      "epoch": 3.52,
-      "learning_rate": 1.0606436528380207e-05,
-      "loss": 0.2463,
+      "epoch": 3.6092190146682523,
+      "grad_norm": 0.29217445850372314,
+      "learning_rate": 9.49267151921839e-06,
+      "loss": 0.37,
       "step": 100145
     },
     {
-      "epoch": 3.52,
-      "learning_rate": 1.0604107418841209e-05,
-      "loss": 0.2536,
+      "epoch": 3.609399214329477,
+      "grad_norm": 0.21554695069789886,
+      "learning_rate": 9.4903827281422e-06,
+      "loss": 0.3741,
       "step": 100150
     },
     {
-      "epoch": 3.52,
-      "learning_rate": 1.0601778496223389e-05,
-      "loss": 0.2507,
+      "epoch": 3.609579413990702,
+      "grad_norm": 0.17673182487487793,
+      "learning_rate": 9.488094148381533e-06,
+      "loss": 0.3679,
       "step": 100155
     },
     {
-      "epoch": 3.52,
-      "learning_rate": 1.0599449760556978e-05,
-      "loss": 0.2748,
+      "epoch": 3.609759613651926,
+      "grad_norm": 0.2148004174232483,
+      "learning_rate": 9.485805779967588e-06,
+      "loss": 0.3473,
       "step": 100160
     },
     {
-      "epoch": 3.52,
-      "learning_rate": 1.0597121211872232e-05,
-      "loss": 0.253,
+      "epoch": 3.609939813313151,
+      "grad_norm": 0.33437803387641907,
+      "learning_rate": 9.483517622931534e-06,
+      "loss": 0.3883,
       "step": 100165
     },
     {
-      "epoch": 3.52,
-      "learning_rate": 1.0594792850199363e-05,
-      "loss": 0.2503,
+      "epoch": 3.6101200129743756,
+      "grad_norm": 0.31750041246414185,
+      "learning_rate": 9.48122967730455e-06,
+      "loss": 0.3693,
       "step": 100170
     },
     {
-      "epoch": 3.52,
-      "learning_rate": 1.0592464675568627e-05,
-      "loss": 0.2762,
+      "epoch": 3.6103002126356003,
+      "grad_norm": 0.2354346662759781,
+      "learning_rate": 9.478941943117809e-06,
+      "loss": 0.388,
       "step": 100175
     },
     {
-      "epoch": 3.52,
-      "learning_rate": 1.0590136688010232e-05,
-      "loss": 0.2657,
+      "epoch": 3.610480412296825,
+      "grad_norm": 0.2933277189731598,
+      "learning_rate": 9.47665442040247e-06,
+      "loss": 0.4218,
       "step": 100180
     },
     {
-      "epoch": 3.52,
-      "learning_rate": 1.0587808887554427e-05,
-      "loss": 0.2391,
+      "epoch": 3.6106606119580498,
+      "grad_norm": 0.2739449441432953,
+      "learning_rate": 9.474367109189716e-06,
+      "loss": 0.3977,
       "step": 100185
     },
     {
-      "epoch": 3.52,
-      "learning_rate": 1.0585481274231421e-05,
-      "loss": 0.2671,
+      "epoch": 3.610840811619274,
+      "grad_norm": 0.2737182676792145,
+      "learning_rate": 9.472080009510707e-06,
+      "loss": 0.3892,
       "step": 100190
     },
     {
-      "epoch": 3.53,
-      "learning_rate": 1.0583153848071437e-05,
-      "loss": 0.2434,
+      "epoch": 3.611021011280499,
+      "grad_norm": 0.2394844889640808,
+      "learning_rate": 9.469793121396597e-06,
+      "loss": 0.366,
       "step": 100195
     },
     {
-      "epoch": 3.53,
-      "learning_rate": 1.0580826609104704e-05,
-      "loss": 0.2563,
+      "epoch": 3.6112012109417235,
+      "grad_norm": 0.20772317051887512,
+      "learning_rate": 9.46796376325298e-06,
+      "loss": 0.333,
       "step": 100200
     },
     {
-      "epoch": 3.53,
-      "learning_rate": 1.0578499557361438e-05,
-      "loss": 0.2772,
+      "epoch": 3.611381410602948,
+      "grad_norm": 0.19267010688781738,
+      "learning_rate": 9.465677256034215e-06,
+      "loss": 0.3884,
       "step": 100205
     },
     {
-      "epoch": 3.53,
-      "learning_rate": 1.057617269287185e-05,
-      "loss": 0.2472,
+      "epoch": 3.6115616102641726,
+      "grad_norm": 0.28297874331474304,
+      "learning_rate": 9.463390960467583e-06,
+      "loss": 0.3827,
       "step": 100210
     },
     {
-      "epoch": 3.53,
-      "learning_rate": 1.0573846015666145e-05,
-      "loss": 0.2396,
+      "epoch": 3.6117418099253973,
+      "grad_norm": 0.25464269518852234,
+      "learning_rate": 9.461104876584256e-06,
+      "loss": 0.3932,
       "step": 100215
     },
     {
-      "epoch": 3.53,
-      "learning_rate": 1.057151952577455e-05,
-      "loss": 0.2741,
+      "epoch": 3.611922009586622,
+      "grad_norm": 0.22311927378177643,
+      "learning_rate": 9.458819004415361e-06,
+      "loss": 0.37,
       "step": 100220
     },
     {
-      "epoch": 3.53,
-      "learning_rate": 1.0569193223227259e-05,
-      "loss": 0.2355,
+      "epoch": 3.6121022092478468,
+      "grad_norm": 0.22702719271183014,
+      "learning_rate": 9.456533343992039e-06,
+      "loss": 0.4145,
       "step": 100225
     },
     {
-      "epoch": 3.53,
-      "learning_rate": 1.056686710805449e-05,
-      "loss": 0.2578,
+      "epoch": 3.6122824089090715,
+      "grad_norm": 0.20826251804828644,
+      "learning_rate": 9.454247895345447e-06,
+      "loss": 0.3434,
       "step": 100230
     },
     {
-      "epoch": 3.53,
-      "learning_rate": 1.0564541180286434e-05,
-      "loss": 0.2587,
+      "epoch": 3.6124626085702958,
+      "grad_norm": 0.2535870671272278,
+      "learning_rate": 9.451962658506722e-06,
+      "loss": 0.3997,
       "step": 100235
     },
     {
-      "epoch": 3.53,
-      "learning_rate": 1.0562215439953302e-05,
-      "loss": 0.2684,
+      "epoch": 3.6126428082315205,
+      "grad_norm": 0.27190902829170227,
+      "learning_rate": 9.449677633506992e-06,
+      "loss": 0.3874,
       "step": 100240
     },
     {
-      "epoch": 3.53,
-      "learning_rate": 1.0559889887085288e-05,
-      "loss": 0.2392,
+      "epoch": 3.6128230078927452,
+      "grad_norm": 0.2581278383731842,
+      "learning_rate": 9.447392820377397e-06,
+      "loss": 0.3805,
       "step": 100245
     },
     {
-      "epoch": 3.53,
-      "learning_rate": 1.0557564521712593e-05,
-      "loss": 0.2883,
+      "epoch": 3.6130032075539695,
+      "grad_norm": 0.2271093875169754,
+      "learning_rate": 9.445108219149052e-06,
+      "loss": 0.3962,
       "step": 100250
     },
     {
-      "epoch": 3.53,
-      "learning_rate": 1.0555239343865392e-05,
-      "loss": 0.2483,
+      "epoch": 3.6131834072151943,
+      "grad_norm": 0.251756489276886,
+      "learning_rate": 9.442823829853103e-06,
+      "loss": 0.3868,
       "step": 100255
     },
     {
-      "epoch": 3.53,
-      "learning_rate": 1.0552914353573897e-05,
-      "loss": 0.2482,
+      "epoch": 3.613363606876419,
+      "grad_norm": 0.20969605445861816,
+      "learning_rate": 9.440539652520671e-06,
+      "loss": 0.3637,
       "step": 100260
     },
     {
-      "epoch": 3.53,
-      "learning_rate": 1.0550589550868289e-05,
-      "loss": 0.293,
+      "epoch": 3.6135438065376437,
+      "grad_norm": 0.2561705410480499,
+      "learning_rate": 9.438255687182873e-06,
+      "loss": 0.3433,
       "step": 100265
     },
     {
-      "epoch": 3.53,
-      "learning_rate": 1.0548264935778754e-05,
-      "loss": 0.2732,
+      "epoch": 3.6137240061988685,
+      "grad_norm": 0.19795164465904236,
+      "learning_rate": 9.435971933870827e-06,
+      "loss": 0.3794,
       "step": 100270
     },
     {
-      "epoch": 3.53,
-      "learning_rate": 1.0545940508335464e-05,
-      "loss": 0.2652,
+      "epoch": 3.613904205860093,
+      "grad_norm": 0.267392098903656,
+      "learning_rate": 9.433688392615644e-06,
+      "loss": 0.3855,
       "step": 100275
     },
     {
-      "epoch": 3.53,
-      "learning_rate": 1.0544081101506425e-05,
-      "loss": 0.2416,
+      "epoch": 3.6140844055213175,
+      "grad_norm": 0.27009448409080505,
+      "learning_rate": 9.431405063448451e-06,
+      "loss": 0.3847,
       "step": 100280
     },
     {
-      "epoch": 3.53,
-      "learning_rate": 1.0541757011902451e-05,
-      "loss": 0.2487,
+      "epoch": 3.6142646051825422,
+      "grad_norm": 0.22548210620880127,
+      "learning_rate": 9.429121946400358e-06,
+      "loss": 0.3959,
       "step": 100285
     },
     {
-      "epoch": 3.53,
-      "learning_rate": 1.0539433110029242e-05,
-      "loss": 0.2596,
+      "epoch": 3.614444804843767,
+      "grad_norm": 0.2359258383512497,
+      "learning_rate": 9.426839041502444e-06,
+      "loss": 0.3798,
       "step": 100290
     },
     {
-      "epoch": 3.53,
-      "learning_rate": 1.0537109395916966e-05,
-      "loss": 0.2557,
+      "epoch": 3.6146250045049917,
+      "grad_norm": 0.20709139108657837,
+      "learning_rate": 9.424556348785846e-06,
+      "loss": 0.3797,
       "step": 100295
     },
     {
-      "epoch": 3.53,
-      "learning_rate": 1.0534785869595793e-05,
-      "loss": 0.2529,
+      "epoch": 3.614805204166216,
+      "grad_norm": 0.22372165322303772,
+      "learning_rate": 9.422273868281639e-06,
+      "loss": 0.3794,
       "step": 100300
     },
     {
-      "epoch": 3.53,
-      "learning_rate": 1.0532462531095883e-05,
-      "loss": 0.2673,
+      "epoch": 3.6149854038274407,
+      "grad_norm": 0.17889077961444855,
+      "learning_rate": 9.419991600020947e-06,
+      "loss": 0.3699,
       "step": 100305
     },
     {
-      "epoch": 3.53,
-      "learning_rate": 1.0530139380447421e-05,
-      "loss": 0.2484,
+      "epoch": 3.6151656034886654,
+      "grad_norm": 0.17842791974544525,
+      "learning_rate": 9.417709544034862e-06,
+      "loss": 0.3769,
       "step": 100310
     },
     {
-      "epoch": 3.53,
-      "learning_rate": 1.0527816417680555e-05,
-      "loss": 0.2821,
+      "epoch": 3.61534580314989,
+      "grad_norm": 0.2319183498620987,
+      "learning_rate": 9.415427700354446e-06,
+      "loss": 0.3625,
       "step": 100315
     },
     {
-      "epoch": 3.53,
-      "learning_rate": 1.0525493642825462e-05,
-      "loss": 0.251,
+      "epoch": 3.615526002811115,
+      "grad_norm": 0.22687672078609467,
+      "learning_rate": 9.413146069010822e-06,
+      "loss": 0.387,
       "step": 100320
     },
     {
-      "epoch": 3.53,
-      "learning_rate": 1.052317105591229e-05,
-      "loss": 0.266,
+      "epoch": 3.615706202472339,
+      "grad_norm": 0.25115665793418884,
+      "learning_rate": 9.410864650035064e-06,
+      "loss": 0.3584,
       "step": 100325
     },
     {
-      "epoch": 3.53,
-      "learning_rate": 1.0520848656971208e-05,
-      "loss": 0.3008,
+      "epoch": 3.615886402133564,
+      "grad_norm": 0.23549871146678925,
+      "learning_rate": 9.408583443458257e-06,
+      "loss": 0.3989,
       "step": 100330
     },
     {
-      "epoch": 3.53,
-      "learning_rate": 1.0518526446032365e-05,
-      "loss": 0.2552,
+      "epoch": 3.6160666017947887,
+      "grad_norm": 0.19801755249500275,
+      "learning_rate": 9.406302449311485e-06,
+      "loss": 0.3762,
       "step": 100335
     },
     {
-      "epoch": 3.53,
-      "learning_rate": 1.051620442312591e-05,
-      "loss": 0.2618,
+      "epoch": 3.6162468014560134,
+      "grad_norm": 0.25041335821151733,
+      "learning_rate": 9.404021667625812e-06,
+      "loss": 0.3706,
       "step": 100340
     },
     {
-      "epoch": 3.53,
-      "learning_rate": 1.051388258828199e-05,
-      "loss": 0.2678,
+      "epoch": 3.6164270011172377,
+      "grad_norm": 0.2323254495859146,
+      "learning_rate": 9.401741098432332e-06,
+      "loss": 0.3995,
       "step": 100345
     },
     {
-      "epoch": 3.53,
-      "learning_rate": 1.0511560941530765e-05,
-      "loss": 0.2644,
+      "epoch": 3.6166072007784624,
+      "grad_norm": 0.20460768043994904,
+      "learning_rate": 9.399460741762111e-06,
+      "loss": 0.3853,
       "step": 100350
     },
     {
-      "epoch": 3.53,
-      "learning_rate": 1.0509239482902373e-05,
-      "loss": 0.2646,
+      "epoch": 3.616787400439687,
+      "grad_norm": 0.30746299028396606,
+      "learning_rate": 9.397180597646218e-06,
+      "loss": 0.387,
       "step": 100355
     },
     {
-      "epoch": 3.53,
-      "learning_rate": 1.0506918212426956e-05,
-      "loss": 0.3034,
+      "epoch": 3.616967600100912,
+      "grad_norm": 0.24463587999343872,
+      "learning_rate": 9.394900666115716e-06,
+      "loss": 0.3957,
       "step": 100360
     },
     {
-      "epoch": 3.53,
-      "learning_rate": 1.0504597130134647e-05,
-      "loss": 0.2515,
+      "epoch": 3.6171477997621366,
+      "grad_norm": 0.19743074476718903,
+      "learning_rate": 9.392620947201675e-06,
+      "loss": 0.3711,
       "step": 100365
     },
     {
-      "epoch": 3.53,
-      "learning_rate": 1.0502276236055589e-05,
-      "loss": 0.2508,
+      "epoch": 3.617327999423361,
+      "grad_norm": 0.23998847603797913,
+      "learning_rate": 9.390341440935138e-06,
+      "loss": 0.3986,
       "step": 100370
     },
     {
-      "epoch": 3.53,
-      "learning_rate": 1.049995553021993e-05,
-      "loss": 0.2688,
+      "epoch": 3.6175081990845857,
+      "grad_norm": 0.2761986553668976,
+      "learning_rate": 9.38806214734719e-06,
+      "loss": 0.3685,
       "step": 100375
     },
     {
-      "epoch": 3.53,
-      "learning_rate": 1.0497635012657783e-05,
-      "loss": 0.2463,
+      "epoch": 3.6176883987458104,
+      "grad_norm": 0.209417924284935,
+      "learning_rate": 9.38578306646887e-06,
+      "loss": 0.3551,
       "step": 100380
     },
     {
-      "epoch": 3.53,
-      "learning_rate": 1.0495314683399298e-05,
-      "loss": 0.2498,
+      "epoch": 3.617868598407035,
+      "grad_norm": 0.2800207734107971,
+      "learning_rate": 9.383504198331233e-06,
+      "loss": 0.38,
       "step": 100385
     },
     {
-      "epoch": 3.53,
-      "learning_rate": 1.0492994542474593e-05,
-      "loss": 0.2574,
+      "epoch": 3.6180487980682594,
+      "grad_norm": 0.2673444151878357,
+      "learning_rate": 9.38122554296533e-06,
+      "loss": 0.3892,
       "step": 100390
     },
     {
-      "epoch": 3.53,
-      "learning_rate": 1.0490674589913794e-05,
-      "loss": 0.2614,
+      "epoch": 3.618228997729484,
+      "grad_norm": 0.26103273034095764,
+      "learning_rate": 9.378947100402194e-06,
+      "loss": 0.3775,
       "step": 100395
     },
     {
-      "epoch": 3.53,
-      "learning_rate": 1.0488354825747013e-05,
-      "loss": 0.2473,
+      "epoch": 3.618409197390709,
+      "grad_norm": 0.20183852314949036,
+      "learning_rate": 9.376668870672897e-06,
+      "loss": 0.3525,
       "step": 100400
     },
     {
-      "epoch": 3.53,
-      "learning_rate": 1.0486035250004392e-05,
-      "loss": 0.2613,
+      "epoch": 3.6185893970519336,
+      "grad_norm": 0.22087271511554718,
+      "learning_rate": 9.374390853808454e-06,
+      "loss": 0.3442,
       "step": 100405
     },
     {
-      "epoch": 3.53,
-      "learning_rate": 1.0483715862716037e-05,
-      "loss": 0.2537,
+      "epoch": 3.6187695967131583,
+      "grad_norm": 0.23035688698291779,
+      "learning_rate": 9.372113049839903e-06,
+      "loss": 0.3829,
       "step": 100410
     },
     {
-      "epoch": 3.53,
-      "learning_rate": 1.0481396663912063e-05,
-      "loss": 0.2646,
+      "epoch": 3.618949796374383,
+      "grad_norm": 0.199892058968544,
+      "learning_rate": 9.369835458798293e-06,
+      "loss": 0.3724,
       "step": 100415
     },
     {
-      "epoch": 3.53,
-      "learning_rate": 1.0479077653622579e-05,
-      "loss": 0.2528,
+      "epoch": 3.6191299960356074,
+      "grad_norm": 0.2462371587753296,
+      "learning_rate": 9.367558080714639e-06,
+      "loss": 0.3667,
       "step": 100420
     },
     {
-      "epoch": 3.53,
-      "learning_rate": 1.047675883187771e-05,
-      "loss": 0.2698,
+      "epoch": 3.619310195696832,
+      "grad_norm": 0.21908694505691528,
+      "learning_rate": 9.36528091562e-06,
+      "loss": 0.3298,
       "step": 100425
     },
     {
-      "epoch": 3.53,
-      "learning_rate": 1.0474440198707546e-05,
-      "loss": 0.2347,
+      "epoch": 3.619490395358057,
+      "grad_norm": 0.22547689080238342,
+      "learning_rate": 9.36300396354537e-06,
+      "loss": 0.3721,
       "step": 100430
     },
     {
-      "epoch": 3.53,
-      "learning_rate": 1.0472121754142213e-05,
-      "loss": 0.2728,
+      "epoch": 3.619670595019281,
+      "grad_norm": 0.28035834431648254,
+      "learning_rate": 9.360727224521775e-06,
+      "loss": 0.3714,
       "step": 100435
     },
     {
-      "epoch": 3.53,
-      "learning_rate": 1.0469803498211797e-05,
-      "loss": 0.2621,
+      "epoch": 3.619850794680506,
+      "grad_norm": 0.2472325563430786,
+      "learning_rate": 9.358450698580254e-06,
+      "loss": 0.3513,
       "step": 100440
     },
     {
-      "epoch": 3.53,
-      "learning_rate": 1.0467485430946416e-05,
-      "loss": 0.2744,
+      "epoch": 3.6200309943417306,
+      "grad_norm": 0.24515065550804138,
+      "learning_rate": 9.356174385751815e-06,
+      "loss": 0.3279,
       "step": 100445
     },
     {
-      "epoch": 3.53,
-      "learning_rate": 1.0465167552376159e-05,
-      "loss": 0.2605,
+      "epoch": 3.6202111940029553,
+      "grad_norm": 0.29932889342308044,
+      "learning_rate": 9.35389828606747e-06,
+      "loss": 0.3565,
       "step": 100450
     },
     {
-      "epoch": 3.53,
-      "learning_rate": 1.046284986253112e-05,
-      "loss": 0.2494,
+      "epoch": 3.62039139366418,
+      "grad_norm": 0.2718527913093567,
+      "learning_rate": 9.35162239955823e-06,
+      "loss": 0.3859,
       "step": 100455
     },
     {
-      "epoch": 3.53,
-      "learning_rate": 1.0460532361441391e-05,
-      "loss": 0.25,
+      "epoch": 3.620571593325405,
+      "grad_norm": 0.21936644613742828,
+      "learning_rate": 9.349346726255098e-06,
+      "loss": 0.3857,
       "step": 100460
     },
     {
-      "epoch": 3.53,
-      "learning_rate": 1.0458215049137074e-05,
-      "loss": 0.2382,
+      "epoch": 3.620751792986629,
+      "grad_norm": 0.2328893095254898,
+      "learning_rate": 9.347071266189095e-06,
+      "loss": 0.333,
       "step": 100465
     },
     {
-      "epoch": 3.53,
-      "learning_rate": 1.045589792564825e-05,
-      "loss": 0.2719,
+      "epoch": 3.620931992647854,
+      "grad_norm": 0.22715623676776886,
+      "learning_rate": 9.344796019391217e-06,
+      "loss": 0.3534,
       "step": 100470
     },
     {
-      "epoch": 3.53,
-      "learning_rate": 1.0453580991004999e-05,
-      "loss": 0.2378,
+      "epoch": 3.6211121923090785,
+      "grad_norm": 0.2767705023288727,
+      "learning_rate": 9.342520985892461e-06,
+      "loss": 0.3693,
       "step": 100475
     },
     {
-      "epoch": 3.54,
-      "learning_rate": 1.0451264245237422e-05,
-      "loss": 0.2533,
+      "epoch": 3.621292391970303,
+      "grad_norm": 0.22328001260757446,
+      "learning_rate": 9.340246165723826e-06,
+      "loss": 0.4045,
       "step": 100480
     },
     {
-      "epoch": 3.54,
-      "learning_rate": 1.044894768837558e-05,
-      "loss": 0.2508,
+      "epoch": 3.6214725916315276,
+      "grad_norm": 0.25353187322616577,
+      "learning_rate": 9.337971558916295e-06,
+      "loss": 0.3343,
       "step": 100485
     },
     {
-      "epoch": 3.54,
-      "learning_rate": 1.0446631320449571e-05,
-      "loss": 0.2745,
+      "epoch": 3.6216527912927523,
+      "grad_norm": 0.2742082178592682,
+      "learning_rate": 9.335697165500878e-06,
+      "loss": 0.3792,
       "step": 100490
     },
     {
-      "epoch": 3.54,
-      "learning_rate": 1.0444315141489458e-05,
-      "loss": 0.248,
+      "epoch": 3.621832990953977,
+      "grad_norm": 0.28070586919784546,
+      "learning_rate": 9.333422985508563e-06,
+      "loss": 0.3781,
       "step": 100495
     },
     {
-      "epoch": 3.54,
-      "learning_rate": 1.0441999151525326e-05,
-      "loss": 0.2534,
+      "epoch": 3.6220131906152018,
+      "grad_norm": 0.20703862607479095,
+      "learning_rate": 9.331149018970311e-06,
+      "loss": 0.3754,
       "step": 100500
     },
     {
-      "epoch": 3.54,
-      "eval_loss": 0.25199830532073975,
-      "eval_runtime": 10.5358,
-      "eval_samples_per_second": 9.491,
-      "eval_steps_per_second": 9.491,
+      "epoch": 3.6220131906152018,
+      "eval_loss": 0.42909446358680725,
+      "eval_runtime": 3.5291,
+      "eval_samples_per_second": 28.336,
+      "eval_steps_per_second": 7.084,
       "step": 100500
     },
     {
-      "epoch": 3.54,
-      "learning_rate": 1.0439683350587243e-05,
-      "loss": 0.2719,
+      "epoch": 3.6221933902764265,
+      "grad_norm": 0.23609666526317596,
+      "learning_rate": 9.328875265917128e-06,
+      "loss": 0.3835,
       "step": 100505
     },
     {
-      "epoch": 3.54,
-      "learning_rate": 1.0437367738705273e-05,
-      "loss": 0.2638,
+      "epoch": 3.622373589937651,
+      "grad_norm": 0.22293473780155182,
+      "learning_rate": 9.326601726379986e-06,
+      "loss": 0.4105,
       "step": 100510
     },
     {
-      "epoch": 3.54,
-      "learning_rate": 1.043505231590948e-05,
-      "loss": 0.2566,
+      "epoch": 3.6225537895988755,
+      "grad_norm": 0.251966267824173,
+      "learning_rate": 9.324328400389858e-06,
+      "loss": 0.3797,
       "step": 100515
     },
     {
-      "epoch": 3.54,
-      "learning_rate": 1.043273708222994e-05,
-      "loss": 0.256,
+      "epoch": 3.6227339892601003,
+      "grad_norm": 0.22561317682266235,
+      "learning_rate": 9.322055287977724e-06,
+      "loss": 0.3619,
       "step": 100520
     },
     {
-      "epoch": 3.54,
-      "learning_rate": 1.0430422037696713e-05,
-      "loss": 0.2569,
+      "epoch": 3.6229141889213246,
+      "grad_norm": 0.22810781002044678,
+      "learning_rate": 9.319782389174542e-06,
+      "loss": 0.3784,
       "step": 100525
     },
     {
-      "epoch": 3.54,
-      "learning_rate": 1.042810718233984e-05,
-      "loss": 0.2712,
+      "epoch": 3.6230943885825493,
+      "grad_norm": 0.20727618038654327,
+      "learning_rate": 9.317509704011298e-06,
+      "loss": 0.3684,
       "step": 100530
     },
     {
-      "epoch": 3.54,
-      "learning_rate": 1.0425792516189406e-05,
-      "loss": 0.2612,
+      "epoch": 3.623274588243774,
+      "grad_norm": 0.3055703341960907,
+      "learning_rate": 9.315237232518948e-06,
+      "loss": 0.3809,
       "step": 100535
     },
     {
-      "epoch": 3.54,
-      "learning_rate": 1.0423478039275438e-05,
-      "loss": 0.2688,
+      "epoch": 3.6234547879049988,
+      "grad_norm": 0.2843535840511322,
+      "learning_rate": 9.312964974728453e-06,
+      "loss": 0.3996,
       "step": 100540
     },
     {
-      "epoch": 3.54,
-      "learning_rate": 1.0421163751628016e-05,
-      "loss": 0.2562,
+      "epoch": 3.6236349875662235,
+      "grad_norm": 0.22682887315750122,
+      "learning_rate": 9.310692930670773e-06,
+      "loss": 0.3469,
       "step": 100545
     },
     {
-      "epoch": 3.54,
-      "learning_rate": 1.0418849653277171e-05,
-      "loss": 0.2504,
+      "epoch": 3.623815187227448,
+      "grad_norm": 0.19296708703041077,
+      "learning_rate": 9.308421100376865e-06,
+      "loss": 0.3504,
       "step": 100550
     },
     {
-      "epoch": 3.54,
-      "learning_rate": 1.0416535744252945e-05,
-      "loss": 0.2534,
+      "epoch": 3.6239953868886725,
+      "grad_norm": 0.3073079586029053,
+      "learning_rate": 9.306149483877674e-06,
+      "loss": 0.4082,
       "step": 100555
     },
     {
-      "epoch": 3.54,
-      "learning_rate": 1.0414222024585402e-05,
-      "loss": 0.2655,
+      "epoch": 3.6241755865498972,
+      "grad_norm": 0.2964288294315338,
+      "learning_rate": 9.303878081204165e-06,
+      "loss": 0.4123,
       "step": 100560
     },
     {
-      "epoch": 3.54,
-      "learning_rate": 1.041190849430457e-05,
-      "loss": 0.2611,
+      "epoch": 3.624355786211122,
+      "grad_norm": 0.2711564600467682,
+      "learning_rate": 9.301606892387276e-06,
+      "loss": 0.3638,
       "step": 100565
     },
     {
-      "epoch": 3.54,
-      "learning_rate": 1.0409595153440497e-05,
-      "loss": 0.2421,
+      "epoch": 3.6245359858723467,
+      "grad_norm": 0.2122730016708374,
+      "learning_rate": 9.299335917457958e-06,
+      "loss": 0.376,
       "step": 100570
     },
     {
-      "epoch": 3.54,
-      "learning_rate": 1.0407282002023203e-05,
-      "loss": 0.2549,
+      "epoch": 3.624716185533571,
+      "grad_norm": 0.21361200511455536,
+      "learning_rate": 9.297065156447147e-06,
+      "loss": 0.3711,
       "step": 100575
     },
     {
-      "epoch": 3.54,
-      "learning_rate": 1.0404969040082746e-05,
-      "loss": 0.2332,
+      "epoch": 3.6248963851947957,
+      "grad_norm": 0.24864503741264343,
+      "learning_rate": 9.294794609385773e-06,
+      "loss": 0.3661,
       "step": 100580
     },
     {
-      "epoch": 3.54,
-      "learning_rate": 1.0402656267649148e-05,
-      "loss": 0.23,
+      "epoch": 3.6250765848560205,
+      "grad_norm": 0.23032629489898682,
+      "learning_rate": 9.292524276304792e-06,
+      "loss": 0.3687,
       "step": 100585
     },
     {
-      "epoch": 3.54,
-      "learning_rate": 1.0400343684752428e-05,
-      "loss": 0.2538,
+      "epoch": 3.625256784517245,
+      "grad_norm": 0.31852540373802185,
+      "learning_rate": 9.290254157235135e-06,
+      "loss": 0.3777,
       "step": 100590
     },
     {
-      "epoch": 3.54,
-      "learning_rate": 1.0398031291422621e-05,
-      "loss": 0.2485,
+      "epoch": 3.62543698417847,
+      "grad_norm": 0.196120947599411,
+      "learning_rate": 9.287984252207707e-06,
+      "loss": 0.3826,
       "step": 100595
     },
     {
-      "epoch": 3.54,
-      "learning_rate": 1.0395719087689768e-05,
-      "loss": 0.2665,
+      "epoch": 3.6256171838396942,
+      "grad_norm": 0.17373962700366974,
+      "learning_rate": 9.285714561253458e-06,
+      "loss": 0.3801,
       "step": 100600
     },
     {
-      "epoch": 3.54,
-      "learning_rate": 1.0393407073583877e-05,
-      "loss": 0.2479,
+      "epoch": 3.625797383500919,
+      "grad_norm": 0.21692752838134766,
+      "learning_rate": 9.2834450844033e-06,
+      "loss": 0.3712,
       "step": 100605
     },
     {
-      "epoch": 3.54,
-      "learning_rate": 1.0391095249134969e-05,
-      "loss": 0.2536,
+      "epoch": 3.6259775831621437,
+      "grad_norm": 0.23615795373916626,
+      "learning_rate": 9.281175821688177e-06,
+      "loss": 0.3827,
       "step": 100610
     },
     {
-      "epoch": 3.54,
-      "learning_rate": 1.0388783614373052e-05,
-      "loss": 0.2444,
+      "epoch": 3.6261577828233684,
+      "grad_norm": 0.21718338131904602,
+      "learning_rate": 9.278906773138979e-06,
+      "loss": 0.3901,
       "step": 100615
     },
     {
-      "epoch": 3.54,
-      "learning_rate": 1.0386472169328159e-05,
-      "loss": 0.2464,
+      "epoch": 3.6263379824845927,
+      "grad_norm": 0.2522786557674408,
+      "learning_rate": 9.276637938786626e-06,
+      "loss": 0.342,
       "step": 100620
     },
     {
-      "epoch": 3.54,
-      "learning_rate": 1.0384160914030297e-05,
-      "loss": 0.2735,
+      "epoch": 3.6265181821458174,
+      "grad_norm": 0.31845128536224365,
+      "learning_rate": 9.274369318662044e-06,
+      "loss": 0.3978,
       "step": 100625
     },
     {
-      "epoch": 3.54,
-      "learning_rate": 1.0381849848509461e-05,
-      "loss": 0.2655,
+      "epoch": 3.626698381807042,
+      "grad_norm": 0.2796033024787903,
+      "learning_rate": 9.272100912796137e-06,
+      "loss": 0.3601,
       "step": 100630
     },
     {
-      "epoch": 3.54,
-      "learning_rate": 1.0379538972795683e-05,
-      "loss": 0.259,
+      "epoch": 3.626878581468267,
+      "grad_norm": 0.22418884932994843,
+      "learning_rate": 9.26983272121981e-06,
+      "loss": 0.3808,
       "step": 100635
     },
     {
-      "epoch": 3.54,
-      "learning_rate": 1.0377228286918956e-05,
-      "loss": 0.2766,
+      "epoch": 3.6270587811294916,
+      "grad_norm": 0.2793503403663635,
+      "learning_rate": 9.267564743963963e-06,
+      "loss": 0.4101,
       "step": 100640
     },
     {
-      "epoch": 3.54,
-      "learning_rate": 1.0374917790909272e-05,
-      "loss": 0.2667,
+      "epoch": 3.6272389807907164,
+      "grad_norm": 0.23813951015472412,
+      "learning_rate": 9.265296981059496e-06,
+      "loss": 0.4087,
       "step": 100645
     },
     {
-      "epoch": 3.54,
-      "learning_rate": 1.0372607484796643e-05,
-      "loss": 0.2517,
+      "epoch": 3.6274191804519407,
+      "grad_norm": 0.27661383152008057,
+      "learning_rate": 9.263029432537317e-06,
+      "loss": 0.3816,
       "step": 100650
     },
     {
-      "epoch": 3.54,
-      "learning_rate": 1.0370297368611073e-05,
-      "loss": 0.28,
+      "epoch": 3.6275993801131654,
+      "grad_norm": 0.25448471307754517,
+      "learning_rate": 9.260762098428319e-06,
+      "loss": 0.3711,
       "step": 100655
     },
     {
-      "epoch": 3.54,
-      "learning_rate": 1.0367987442382554e-05,
-      "loss": 0.2532,
+      "epoch": 3.62777957977439,
+      "grad_norm": 0.18283936381340027,
+      "learning_rate": 9.258494978763387e-06,
+      "loss": 0.3951,
       "step": 100660
     },
     {
-      "epoch": 3.54,
-      "learning_rate": 1.0365677706141072e-05,
-      "loss": 0.2579,
+      "epoch": 3.6279597794356144,
+      "grad_norm": 0.22228430211544037,
+      "learning_rate": 9.256228073573413e-06,
+      "loss": 0.3452,
       "step": 100665
     },
     {
-      "epoch": 3.54,
-      "learning_rate": 1.0363368159916614e-05,
-      "loss": 0.2502,
+      "epoch": 3.628139979096839,
+      "grad_norm": 0.185472771525383,
+      "learning_rate": 9.253961382889278e-06,
+      "loss": 0.4253,
       "step": 100670
     },
     {
-      "epoch": 3.54,
-      "learning_rate": 1.0361058803739182e-05,
-      "loss": 0.2588,
+      "epoch": 3.628320178758064,
+      "grad_norm": 0.18725287914276123,
+      "learning_rate": 9.251694906741879e-06,
+      "loss": 0.3513,
       "step": 100675
     },
     {
-      "epoch": 3.54,
-      "learning_rate": 1.0358749637638754e-05,
-      "loss": 0.2443,
+      "epoch": 3.6285003784192886,
+      "grad_norm": 0.22151315212249756,
+      "learning_rate": 9.249428645162095e-06,
+      "loss": 0.3352,
       "step": 100680
     },
     {
-      "epoch": 3.54,
-      "learning_rate": 1.0356440661645313e-05,
-      "loss": 0.2402,
+      "epoch": 3.6286805780805134,
+      "grad_norm": 0.22003361582756042,
+      "learning_rate": 9.247162598180777e-06,
+      "loss": 0.3763,
       "step": 100685
     },
     {
-      "epoch": 3.54,
-      "learning_rate": 1.0354131875788833e-05,
-      "loss": 0.2557,
+      "epoch": 3.628860777741738,
+      "grad_norm": 0.280362993478775,
+      "learning_rate": 9.244896765828831e-06,
+      "loss": 0.3839,
       "step": 100690
     },
     {
-      "epoch": 3.54,
-      "learning_rate": 1.0351823280099304e-05,
-      "loss": 0.2556,
+      "epoch": 3.6290409774029624,
+      "grad_norm": 0.24455983936786652,
+      "learning_rate": 9.242631148137114e-06,
+      "loss": 0.3854,
       "step": 100695
     },
     {
-      "epoch": 3.54,
-      "learning_rate": 1.0349514874606692e-05,
-      "loss": 0.2534,
+      "epoch": 3.629221177064187,
+      "grad_norm": 0.28553998470306396,
+      "learning_rate": 9.240365745136498e-06,
+      "loss": 0.4387,
       "step": 100700
     },
     {
-      "epoch": 3.54,
-      "learning_rate": 1.034720665934098e-05,
-      "loss": 0.255,
+      "epoch": 3.629401376725412,
+      "grad_norm": 0.22999370098114014,
+      "learning_rate": 9.238100556857847e-06,
+      "loss": 0.3486,
       "step": 100705
     },
     {
-      "epoch": 3.54,
-      "learning_rate": 1.0344898634332123e-05,
-      "loss": 0.258,
+      "epoch": 3.629581576386636,
+      "grad_norm": 0.24775420129299164,
+      "learning_rate": 9.235835583332017e-06,
+      "loss": 0.4171,
       "step": 100710
     },
     {
-      "epoch": 3.54,
-      "learning_rate": 1.0342590799610111e-05,
-      "loss": 0.2445,
+      "epoch": 3.629761776047861,
+      "grad_norm": 0.30555811524391174,
+      "learning_rate": 9.233570824589881e-06,
+      "loss": 0.4051,
       "step": 100715
     },
     {
-      "epoch": 3.54,
-      "learning_rate": 1.0340283155204895e-05,
-      "loss": 0.2602,
+      "epoch": 3.6299419757090856,
+      "grad_norm": 0.22852908074855804,
+      "learning_rate": 9.23130628066229e-06,
+      "loss": 0.3686,
       "step": 100720
     },
     {
-      "epoch": 3.54,
-      "learning_rate": 1.033797570114644e-05,
-      "loss": 0.2574,
+      "epoch": 3.6301221753703103,
+      "grad_norm": 0.20418477058410645,
+      "learning_rate": 9.229041951580101e-06,
+      "loss": 0.3467,
       "step": 100725
     },
     {
-      "epoch": 3.54,
-      "learning_rate": 1.0335668437464705e-05,
-      "loss": 0.2721,
+      "epoch": 3.630302375031535,
+      "grad_norm": 0.2210427224636078,
+      "learning_rate": 9.226777837374163e-06,
+      "loss": 0.3732,
       "step": 100730
     },
     {
-      "epoch": 3.54,
-      "learning_rate": 1.0333361364189652e-05,
-      "loss": 0.2454,
+      "epoch": 3.63048257469276,
+      "grad_norm": 0.2393556535243988,
+      "learning_rate": 9.224513938075318e-06,
+      "loss": 0.4281,
       "step": 100735
     },
     {
-      "epoch": 3.54,
-      "learning_rate": 1.0331054481351244e-05,
-      "loss": 0.2615,
+      "epoch": 3.630662774353984,
+      "grad_norm": 0.2593443691730499,
+      "learning_rate": 9.222250253714413e-06,
+      "loss": 0.3624,
       "step": 100740
     },
     {
-      "epoch": 3.54,
-      "learning_rate": 1.0328747788979412e-05,
-      "loss": 0.2755,
+      "epoch": 3.630842974015209,
+      "grad_norm": 0.2505268454551697,
+      "learning_rate": 9.219986784322299e-06,
+      "loss": 0.3829,
       "step": 100745
     },
     {
-      "epoch": 3.54,
-      "learning_rate": 1.0326441287104132e-05,
-      "loss": 0.244,
+      "epoch": 3.6310231736764336,
+      "grad_norm": 0.2201002836227417,
+      "learning_rate": 9.217723529929812e-06,
+      "loss": 0.405,
       "step": 100750
     },
     {
-      "epoch": 3.54,
-      "learning_rate": 1.0324134975755334e-05,
-      "loss": 0.2653,
+      "epoch": 3.631203373337658,
+      "grad_norm": 0.2460431605577469,
+      "learning_rate": 9.215460490567784e-06,
+      "loss": 0.3749,
       "step": 100755
     },
     {
-      "epoch": 3.55,
-      "learning_rate": 1.0321828854962978e-05,
-      "loss": 0.2529,
+      "epoch": 3.6313835729988826,
+      "grad_norm": 0.2341911494731903,
+      "learning_rate": 9.213197666267052e-06,
+      "loss": 0.4055,
       "step": 100760
     },
     {
-      "epoch": 3.55,
-      "learning_rate": 1.0319522924756994e-05,
-      "loss": 0.2557,
+      "epoch": 3.6315637726601073,
+      "grad_norm": 0.25453901290893555,
+      "learning_rate": 9.210935057058437e-06,
+      "loss": 0.4276,
       "step": 100765
     },
     {
-      "epoch": 3.55,
-      "learning_rate": 1.0317217185167338e-05,
-      "loss": 0.2609,
+      "epoch": 3.631743972321332,
+      "grad_norm": 0.22719573974609375,
+      "learning_rate": 9.208672662972783e-06,
+      "loss": 0.3768,
       "step": 100770
     },
     {
-      "epoch": 3.55,
-      "learning_rate": 1.0314911636223942e-05,
-      "loss": 0.2805,
+      "epoch": 3.631924171982557,
+      "grad_norm": 0.22030307352542877,
+      "learning_rate": 9.206410484040914e-06,
+      "loss": 0.3629,
       "step": 100775
     },
     {
-      "epoch": 3.55,
-      "learning_rate": 1.0312606277956738e-05,
-      "loss": 0.2377,
+      "epoch": 3.6321043716437815,
+      "grad_norm": 0.27278268337249756,
+      "learning_rate": 9.204148520293629e-06,
+      "loss": 0.3757,
       "step": 100780
     },
     {
-      "epoch": 3.55,
-      "learning_rate": 1.0310301110395657e-05,
-      "loss": 0.2782,
+      "epoch": 3.632284571305006,
+      "grad_norm": 0.2132975459098816,
+      "learning_rate": 9.201886771761772e-06,
+      "loss": 0.3837,
       "step": 100785
     },
     {
-      "epoch": 3.55,
-      "learning_rate": 1.030799613357064e-05,
-      "loss": 0.2543,
+      "epoch": 3.6324647709662305,
+      "grad_norm": 0.2718239724636078,
+      "learning_rate": 9.199625238476139e-06,
+      "loss": 0.3942,
       "step": 100790
     },
     {
-      "epoch": 3.55,
-      "learning_rate": 1.0305691347511612e-05,
-      "loss": 0.2685,
+      "epoch": 3.6326449706274553,
+      "grad_norm": 0.2489858865737915,
+      "learning_rate": 9.19736392046757e-06,
+      "loss": 0.3506,
       "step": 100795
     },
     {
-      "epoch": 3.55,
-      "learning_rate": 1.0303386752248497e-05,
-      "loss": 0.2632,
+      "epoch": 3.63282517028868,
+      "grad_norm": 0.2690345048904419,
+      "learning_rate": 9.195102817766851e-06,
+      "loss": 0.365,
       "step": 100800
     },
     {
-      "epoch": 3.55,
-      "learning_rate": 1.0301082347811214e-05,
-      "loss": 0.2473,
+      "epoch": 3.6330053699499043,
+      "grad_norm": 0.25727832317352295,
+      "learning_rate": 9.192841930404786e-06,
+      "loss": 0.3702,
       "step": 100805
     },
     {
-      "epoch": 3.55,
-      "learning_rate": 1.0298778134229689e-05,
-      "loss": 0.2704,
+      "epoch": 3.633185569611129,
+      "grad_norm": 0.28264495730400085,
+      "learning_rate": 9.190581258412198e-06,
+      "loss": 0.3825,
       "step": 100810
     },
     {
-      "epoch": 3.55,
-      "learning_rate": 1.0296474111533849e-05,
-      "loss": 0.2618,
+      "epoch": 3.6333657692723538,
+      "grad_norm": 0.2584122121334076,
+      "learning_rate": 9.18832080181988e-06,
+      "loss": 0.4279,
       "step": 100815
     },
     {
-      "epoch": 3.55,
-      "learning_rate": 1.0294170279753602e-05,
-      "loss": 0.2552,
+      "epoch": 3.6335459689335785,
+      "grad_norm": 0.2514478862285614,
+      "learning_rate": 9.18606056065863e-06,
+      "loss": 0.3661,
       "step": 100820
     },
     {
-      "epoch": 3.55,
-      "learning_rate": 1.0291866638918854e-05,
-      "loss": 0.2588,
+      "epoch": 3.6337261685948032,
+      "grad_norm": 0.2434535175561905,
+      "learning_rate": 9.183800534959245e-06,
+      "loss": 0.3543,
       "step": 100825
     },
     {
-      "epoch": 3.55,
-      "learning_rate": 1.0289563189059534e-05,
-      "loss": 0.2607,
+      "epoch": 3.6339063682560275,
+      "grad_norm": 0.225295290350914,
+      "learning_rate": 9.181540724752504e-06,
+      "loss": 0.3573,
       "step": 100830
     },
     {
-      "epoch": 3.55,
-      "learning_rate": 1.0287259930205542e-05,
-      "loss": 0.2637,
+      "epoch": 3.6340865679172523,
+      "grad_norm": 0.24885191023349762,
+      "learning_rate": 9.179281130069217e-06,
+      "loss": 0.3833,
       "step": 100835
     },
     {
-      "epoch": 3.55,
-      "learning_rate": 1.0284956862386783e-05,
-      "loss": 0.2253,
+      "epoch": 3.634266767578477,
+      "grad_norm": 0.2610785961151123,
+      "learning_rate": 9.177021750940162e-06,
+      "loss": 0.3577,
       "step": 100840
     },
     {
-      "epoch": 3.55,
-      "learning_rate": 1.0282653985633154e-05,
-      "loss": 0.2524,
+      "epoch": 3.6344469672397017,
+      "grad_norm": 0.2569064497947693,
+      "learning_rate": 9.174762587396124e-06,
+      "loss": 0.3981,
       "step": 100845
     },
     {
-      "epoch": 3.55,
-      "learning_rate": 1.0280351299974572e-05,
-      "loss": 0.243,
+      "epoch": 3.634627166900926,
+      "grad_norm": 0.308135986328125,
+      "learning_rate": 9.17250363946788e-06,
+      "loss": 0.4227,
       "step": 100850
     },
     {
-      "epoch": 3.55,
-      "learning_rate": 1.0278048805440926e-05,
-      "loss": 0.2449,
+      "epoch": 3.6348073665621508,
+      "grad_norm": 0.2591281235218048,
+      "learning_rate": 9.170244907186201e-06,
+      "loss": 0.3902,
       "step": 100855
     },
     {
-      "epoch": 3.55,
-      "learning_rate": 1.0275746502062108e-05,
-      "loss": 0.2679,
+      "epoch": 3.6349875662233755,
+      "grad_norm": 0.24686266481876373,
+      "learning_rate": 9.167986390581879e-06,
+      "loss": 0.3907,
       "step": 100860
     },
     {
-      "epoch": 3.55,
-      "learning_rate": 1.0273444389868025e-05,
-      "loss": 0.2617,
+      "epoch": 3.6351677658846,
+      "grad_norm": 0.21753914654254913,
+      "learning_rate": 9.165728089685687e-06,
+      "loss": 0.398,
       "step": 100865
     },
     {
-      "epoch": 3.55,
-      "learning_rate": 1.0271142468888553e-05,
-      "loss": 0.2594,
+      "epoch": 3.635347965545825,
+      "grad_norm": 0.23476707935333252,
+      "learning_rate": 9.163470004528366e-06,
+      "loss": 0.3879,
       "step": 100870
     },
     {
-      "epoch": 3.55,
-      "learning_rate": 1.0268840739153598e-05,
-      "loss": 0.2612,
+      "epoch": 3.6355281652070492,
+      "grad_norm": 0.2765940725803375,
+      "learning_rate": 9.16121213514071e-06,
+      "loss": 0.4109,
       "step": 100875
     },
     {
-      "epoch": 3.55,
-      "learning_rate": 1.026653920069303e-05,
-      "loss": 0.2597,
+      "epoch": 3.635708364868274,
+      "grad_norm": 0.2919420301914215,
+      "learning_rate": 9.15895448155347e-06,
+      "loss": 0.3816,
       "step": 100880
     },
     {
-      "epoch": 3.55,
-      "learning_rate": 1.0264237853536748e-05,
-      "loss": 0.2657,
+      "epoch": 3.6358885645294987,
+      "grad_norm": 0.2056131362915039,
+      "learning_rate": 9.156697043797402e-06,
+      "loss": 0.3565,
       "step": 100885
     },
     {
-      "epoch": 3.55,
-      "learning_rate": 1.0261936697714628e-05,
-      "loss": 0.259,
+      "epoch": 3.6360687641907234,
+      "grad_norm": 0.29402926564216614,
+      "learning_rate": 9.154439821903286e-06,
+      "loss": 0.4164,
       "step": 100890
     },
     {
-      "epoch": 3.55,
-      "learning_rate": 1.0259635733256548e-05,
-      "loss": 0.254,
+      "epoch": 3.6362489638519477,
+      "grad_norm": 0.2530692517757416,
+      "learning_rate": 9.152182815901841e-06,
+      "loss": 0.3703,
       "step": 100895
     },
     {
-      "epoch": 3.55,
-      "learning_rate": 1.0257334960192372e-05,
-      "loss": 0.2554,
+      "epoch": 3.6364291635131725,
+      "grad_norm": 0.24608898162841797,
+      "learning_rate": 9.149926025823846e-06,
+      "loss": 0.3821,
       "step": 100900
     },
     {
-      "epoch": 3.55,
-      "learning_rate": 1.0255034378551994e-05,
-      "loss": 0.2457,
+      "epoch": 3.636609363174397,
+      "grad_norm": 0.1967172771692276,
+      "learning_rate": 9.147669451700042e-06,
+      "loss": 0.3682,
       "step": 100905
     },
     {
-      "epoch": 3.55,
-      "learning_rate": 1.0252733988365281e-05,
-      "loss": 0.2942,
+      "epoch": 3.636789562835622,
+      "grad_norm": 0.2919192910194397,
+      "learning_rate": 9.14541309356117e-06,
+      "loss": 0.379,
       "step": 100910
     },
     {
-      "epoch": 3.55,
-      "learning_rate": 1.0250433789662087e-05,
-      "loss": 0.2503,
+      "epoch": 3.6369697624968467,
+      "grad_norm": 0.22142526507377625,
+      "learning_rate": 9.143156951437976e-06,
+      "loss": 0.3938,
       "step": 100915
     },
     {
-      "epoch": 3.55,
-      "learning_rate": 1.0248133782472299e-05,
-      "loss": 0.2548,
+      "epoch": 3.6371499621580714,
+      "grad_norm": 0.22058270871639252,
+      "learning_rate": 9.140901025361198e-06,
+      "loss": 0.3691,
       "step": 100920
     },
     {
-      "epoch": 3.55,
-      "learning_rate": 1.0245833966825766e-05,
-      "loss": 0.2613,
+      "epoch": 3.6373301618192957,
+      "grad_norm": 0.2393786460161209,
+      "learning_rate": 9.138645315361566e-06,
+      "loss": 0.3929,
       "step": 100925
     },
     {
-      "epoch": 3.55,
-      "learning_rate": 1.0243534342752362e-05,
-      "loss": 0.2564,
+      "epoch": 3.6375103614805204,
+      "grad_norm": 0.26111987233161926,
+      "learning_rate": 9.13638982146983e-06,
+      "loss": 0.4086,
       "step": 100930
     },
     {
-      "epoch": 3.55,
-      "learning_rate": 1.0241234910281938e-05,
-      "loss": 0.2618,
+      "epoch": 3.637690561141745,
+      "grad_norm": 0.2450508177280426,
+      "learning_rate": 9.134134543716711e-06,
+      "loss": 0.3776,
       "step": 100935
     },
     {
-      "epoch": 3.55,
-      "learning_rate": 1.0238935669444347e-05,
-      "loss": 0.2408,
+      "epoch": 3.6378707608029694,
+      "grad_norm": 0.26415956020355225,
+      "learning_rate": 9.131879482132935e-06,
+      "loss": 0.4035,
       "step": 100940
     },
     {
-      "epoch": 3.55,
-      "learning_rate": 1.0236636620269455e-05,
-      "loss": 0.2501,
+      "epoch": 3.638050960464194,
+      "grad_norm": 0.21556943655014038,
+      "learning_rate": 9.12962463674923e-06,
+      "loss": 0.3617,
       "step": 100945
     },
     {
-      "epoch": 3.55,
-      "learning_rate": 1.0234337762787107e-05,
-      "loss": 0.2572,
+      "epoch": 3.638231160125419,
+      "grad_norm": 0.26129695773124695,
+      "learning_rate": 9.127370007596308e-06,
+      "loss": 0.3823,
       "step": 100950
     },
     {
-      "epoch": 3.55,
-      "learning_rate": 1.0232039097027151e-05,
-      "loss": 0.2402,
+      "epoch": 3.6384113597866437,
+      "grad_norm": 0.2311808317899704,
+      "learning_rate": 9.125115594704905e-06,
+      "loss": 0.3735,
       "step": 100955
     },
     {
-      "epoch": 3.55,
-      "learning_rate": 1.0229740623019429e-05,
-      "loss": 0.2628,
+      "epoch": 3.6385915594478684,
+      "grad_norm": 0.24276402592658997,
+      "learning_rate": 9.122861398105736e-06,
+      "loss": 0.3719,
       "step": 100960
     },
     {
-      "epoch": 3.55,
-      "learning_rate": 1.0227442340793799e-05,
-      "loss": 0.2499,
+      "epoch": 3.638771759109093,
+      "grad_norm": 0.2854870557785034,
+      "learning_rate": 9.12060741782949e-06,
+      "loss": 0.3717,
       "step": 100965
     },
     {
-      "epoch": 3.55,
-      "learning_rate": 1.0225144250380086e-05,
-      "loss": 0.2416,
+      "epoch": 3.6389519587703174,
+      "grad_norm": 0.2724928557872772,
+      "learning_rate": 9.118353653906905e-06,
+      "loss": 0.4111,
       "step": 100970
     },
     {
-      "epoch": 3.55,
-      "learning_rate": 1.0222846351808146e-05,
-      "loss": 0.2722,
+      "epoch": 3.639132158431542,
+      "grad_norm": 0.22794464230537415,
+      "learning_rate": 9.116100106368667e-06,
+      "loss": 0.3872,
       "step": 100975
     },
     {
-      "epoch": 3.55,
-      "learning_rate": 1.02205486451078e-05,
-      "loss": 0.251,
+      "epoch": 3.639312358092767,
+      "grad_norm": 0.2465512901544571,
+      "learning_rate": 9.113846775245506e-06,
+      "loss": 0.3988,
       "step": 100980
     },
     {
-      "epoch": 3.55,
-      "learning_rate": 1.0218251130308898e-05,
-      "loss": 0.2553,
+      "epoch": 3.639492557753991,
+      "grad_norm": 0.24606558680534363,
+      "learning_rate": 9.111593660568099e-06,
+      "loss": 0.3761,
       "step": 100985
     },
     {
-      "epoch": 3.55,
-      "learning_rate": 1.0215953807441262e-05,
-      "loss": 0.2722,
+      "epoch": 3.639672757415216,
+      "grad_norm": 0.1907157152891159,
+      "learning_rate": 9.109340762367141e-06,
+      "loss": 0.3817,
       "step": 100990
     },
     {
-      "epoch": 3.55,
-      "learning_rate": 1.0213656676534717e-05,
-      "loss": 0.2825,
+      "epoch": 3.6398529570764406,
+      "grad_norm": 0.2616572976112366,
+      "learning_rate": 9.107088080673351e-06,
+      "loss": 0.4135,
       "step": 100995
     },
     {
-      "epoch": 3.55,
-      "learning_rate": 1.02113597376191e-05,
-      "loss": 0.2753,
+      "epoch": 3.6400331567376654,
+      "grad_norm": 0.2173600196838379,
+      "learning_rate": 9.104835615517406e-06,
+      "loss": 0.3634,
       "step": 101000
     },
     {
-      "epoch": 3.55,
-      "eval_loss": 0.2520409822463989,
-      "eval_runtime": 10.5341,
-      "eval_samples_per_second": 9.493,
-      "eval_steps_per_second": 9.493,
+      "epoch": 3.6400331567376654,
+      "eval_loss": 0.4288954436779022,
+      "eval_runtime": 3.5295,
+      "eval_samples_per_second": 28.333,
+      "eval_steps_per_second": 7.083,
       "step": 101000
     },
     {
-      "epoch": 3.55,
-      "learning_rate": 1.0209062990724233e-05,
-      "loss": 0.2884,
+      "epoch": 3.64021335639889,
+      "grad_norm": 0.2579386234283447,
+      "learning_rate": 9.10258336693e-06,
+      "loss": 0.3848,
       "step": 101005
     },
     {
-      "epoch": 3.55,
-      "learning_rate": 1.0206766435879937e-05,
-      "loss": 0.2577,
+      "epoch": 3.640393556060115,
+      "grad_norm": 0.21844492852687836,
+      "learning_rate": 9.10033133494182e-06,
+      "loss": 0.4074,
       "step": 101010
     },
     {
-      "epoch": 3.55,
-      "learning_rate": 1.0204470073116021e-05,
-      "loss": 0.2315,
+      "epoch": 3.640573755721339,
+      "grad_norm": 0.25436314940452576,
+      "learning_rate": 9.098079519583536e-06,
+      "loss": 0.3451,
       "step": 101015
     },
     {
-      "epoch": 3.55,
-      "learning_rate": 1.0202173902462317e-05,
-      "loss": 0.2592,
+      "epoch": 3.640753955382564,
+      "grad_norm": 0.237601637840271,
+      "learning_rate": 9.095827920885847e-06,
+      "loss": 0.3692,
       "step": 101020
     },
     {
-      "epoch": 3.55,
-      "learning_rate": 1.0199877923948634e-05,
-      "loss": 0.2662,
+      "epoch": 3.6409341550437886,
+      "grad_norm": 0.1813516467809677,
+      "learning_rate": 9.093576538879425e-06,
+      "loss": 0.408,
       "step": 101025
     },
     {
-      "epoch": 3.55,
-      "learning_rate": 1.0197582137604773e-05,
-      "loss": 0.2575,
+      "epoch": 3.641114354705013,
+      "grad_norm": 0.23129382729530334,
+      "learning_rate": 9.091325373594944e-06,
+      "loss": 0.3885,
       "step": 101030
     },
     {
-      "epoch": 3.55,
-      "learning_rate": 1.0195286543460553e-05,
-      "loss": 0.2617,
+      "epoch": 3.6412945543662376,
+      "grad_norm": 0.24837276339530945,
+      "learning_rate": 9.089074425063074e-06,
+      "loss": 0.4024,
       "step": 101035
     },
     {
-      "epoch": 3.55,
-      "learning_rate": 1.0192991141545788e-05,
-      "loss": 0.2393,
+      "epoch": 3.6414747540274623,
+      "grad_norm": 0.23222190141677856,
+      "learning_rate": 9.086823693314476e-06,
+      "loss": 0.3736,
       "step": 101040
     },
     {
-      "epoch": 3.56,
-      "learning_rate": 1.0190695931890277e-05,
-      "loss": 0.2726,
+      "epoch": 3.641654953688687,
+      "grad_norm": 0.2549395263195038,
+      "learning_rate": 9.084573178379832e-06,
+      "loss": 0.3688,
       "step": 101045
     },
     {
-      "epoch": 3.56,
-      "learning_rate": 1.0188400914523818e-05,
-      "loss": 0.2583,
+      "epoch": 3.641835153349912,
+      "grad_norm": 0.3221529722213745,
+      "learning_rate": 9.082322880289798e-06,
+      "loss": 0.4044,
       "step": 101050
     },
     {
-      "epoch": 3.56,
-      "learning_rate": 1.0186106089476205e-05,
-      "loss": 0.2676,
+      "epoch": 3.6420153530111365,
+      "grad_norm": 0.19385772943496704,
+      "learning_rate": 9.080072799075033e-06,
+      "loss": 0.3395,
       "step": 101055
     },
     {
-      "epoch": 3.56,
-      "learning_rate": 1.018381145677725e-05,
-      "loss": 0.2549,
+      "epoch": 3.642195552672361,
+      "grad_norm": 0.21461910009384155,
+      "learning_rate": 9.077822934766194e-06,
+      "loss": 0.3626,
       "step": 101060
     },
     {
-      "epoch": 3.56,
-      "learning_rate": 1.0181517016456738e-05,
-      "loss": 0.2461,
+      "epoch": 3.6423757523335856,
+      "grad_norm": 0.2746890187263489,
+      "learning_rate": 9.075573287393935e-06,
+      "loss": 0.3974,
       "step": 101065
     },
     {
-      "epoch": 3.56,
-      "learning_rate": 1.0179222768544463e-05,
-      "loss": 0.2421,
+      "epoch": 3.6425559519948103,
+      "grad_norm": 0.22592508792877197,
+      "learning_rate": 9.073323856988898e-06,
+      "loss": 0.3628,
       "step": 101070
     },
     {
-      "epoch": 3.56,
-      "learning_rate": 1.0176928713070205e-05,
-      "loss": 0.2284,
+      "epoch": 3.642736151656035,
+      "grad_norm": 0.17644916474819183,
+      "learning_rate": 9.071074643581757e-06,
+      "loss": 0.3879,
       "step": 101075
     },
     {
-      "epoch": 3.56,
-      "learning_rate": 1.0174634850063765e-05,
-      "loss": 0.2318,
+      "epoch": 3.6429163513172593,
+      "grad_norm": 0.20565937459468842,
+      "learning_rate": 9.068825647203122e-06,
+      "loss": 0.3596,
       "step": 101080
     },
     {
-      "epoch": 3.56,
-      "learning_rate": 1.0172341179554915e-05,
-      "loss": 0.2498,
+      "epoch": 3.643096550978484,
+      "grad_norm": 0.2097647488117218,
+      "learning_rate": 9.066576867883664e-06,
+      "loss": 0.3634,
       "step": 101085
     },
     {
-      "epoch": 3.56,
-      "learning_rate": 1.017004770157344e-05,
-      "loss": 0.2441,
+      "epoch": 3.643276750639709,
+      "grad_norm": 0.24162964522838593,
+      "learning_rate": 9.06432830565401e-06,
+      "loss": 0.3715,
       "step": 101090
     },
     {
-      "epoch": 3.56,
-      "learning_rate": 1.0167754416149134e-05,
-      "loss": 0.2806,
+      "epoch": 3.6434569503009335,
+      "grad_norm": 0.2646183371543884,
+      "learning_rate": 9.062079960544798e-06,
+      "loss": 0.3756,
       "step": 101095
     },
     {
-      "epoch": 3.56,
-      "learning_rate": 1.0165461323311762e-05,
-      "loss": 0.2589,
+      "epoch": 3.6436371499621583,
+      "grad_norm": 0.20252090692520142,
+      "learning_rate": 9.05983183258666e-06,
+      "loss": 0.3804,
       "step": 101100
     },
     {
-      "epoch": 3.56,
-      "learning_rate": 1.0163168423091096e-05,
-      "loss": 0.2643,
+      "epoch": 3.6438173496233826,
+      "grad_norm": 0.2239752858877182,
+      "learning_rate": 9.057583921810225e-06,
+      "loss": 0.34,
       "step": 101105
     },
     {
-      "epoch": 3.56,
-      "learning_rate": 1.0160875715516902e-05,
-      "loss": 0.2619,
+      "epoch": 3.6439975492846073,
+      "grad_norm": 0.23255230486392975,
+      "learning_rate": 9.055336228246119e-06,
+      "loss": 0.3719,
       "step": 101110
     },
     {
-      "epoch": 3.56,
-      "learning_rate": 1.0158583200618967e-05,
-      "loss": 0.2684,
+      "epoch": 3.644177748945832,
+      "grad_norm": 0.23270903527736664,
+      "learning_rate": 9.053088751924976e-06,
+      "loss": 0.3866,
       "step": 101115
     },
     {
-      "epoch": 3.56,
-      "learning_rate": 1.015629087842705e-05,
-      "loss": 0.2496,
+      "epoch": 3.6443579486070568,
+      "grad_norm": 0.27508658170700073,
+      "learning_rate": 9.05084149287741e-06,
+      "loss": 0.4123,
       "step": 101120
     },
     {
-      "epoch": 3.56,
-      "learning_rate": 1.015399874897091e-05,
-      "loss": 0.2581,
+      "epoch": 3.644538148268281,
+      "grad_norm": 0.19277596473693848,
+      "learning_rate": 9.048594451134042e-06,
+      "loss": 0.3398,
       "step": 101125
     },
     {
-      "epoch": 3.56,
-      "learning_rate": 1.0151706812280306e-05,
-      "loss": 0.2518,
+      "epoch": 3.6447183479295058,
+      "grad_norm": 0.28608769178390503,
+      "learning_rate": 9.046347626725487e-06,
+      "loss": 0.4222,
       "step": 101130
     },
     {
-      "epoch": 3.56,
-      "learning_rate": 1.0149415068385013e-05,
-      "loss": 0.2583,
+      "epoch": 3.6448985475907305,
+      "grad_norm": 0.21985885500907898,
+      "learning_rate": 9.04410101968235e-06,
+      "loss": 0.3681,
       "step": 101135
     },
     {
-      "epoch": 3.56,
-      "learning_rate": 1.0147123517314765e-05,
-      "loss": 0.2328,
+      "epoch": 3.6450787472519552,
+      "grad_norm": 0.2108525186777115,
+      "learning_rate": 9.04185463003525e-06,
+      "loss": 0.338,
       "step": 101140
     },
     {
-      "epoch": 3.56,
-      "learning_rate": 1.0144832159099343e-05,
-      "loss": 0.2369,
+      "epoch": 3.64525894691318,
+      "grad_norm": 0.28811293840408325,
+      "learning_rate": 9.039608457814805e-06,
+      "loss": 0.4011,
       "step": 101145
     },
     {
-      "epoch": 3.56,
-      "learning_rate": 1.0142540993768476e-05,
-      "loss": 0.2563,
+      "epoch": 3.6454391465744047,
+      "grad_norm": 0.2849256992340088,
+      "learning_rate": 9.037362503051585e-06,
+      "loss": 0.4098,
       "step": 101150
     },
     {
-      "epoch": 3.56,
-      "learning_rate": 1.0140250021351932e-05,
-      "loss": 0.2836,
+      "epoch": 3.645619346235629,
+      "grad_norm": 0.2216586172580719,
+      "learning_rate": 9.035116765776223e-06,
+      "loss": 0.355,
       "step": 101155
     },
     {
-      "epoch": 3.56,
-      "learning_rate": 1.013795924187945e-05,
-      "loss": 0.2418,
+      "epoch": 3.6457995458968537,
+      "grad_norm": 0.22768335044384003,
+      "learning_rate": 9.032871246019292e-06,
+      "loss": 0.368,
       "step": 101160
     },
     {
-      "epoch": 3.56,
-      "learning_rate": 1.013566865538077e-05,
-      "loss": 0.2433,
+      "epoch": 3.6459797455580785,
+      "grad_norm": 0.23428839445114136,
+      "learning_rate": 9.030625943811408e-06,
+      "loss": 0.3678,
       "step": 101165
     },
     {
-      "epoch": 3.56,
-      "learning_rate": 1.0133378261885626e-05,
-      "loss": 0.2598,
+      "epoch": 3.6461599452193028,
+      "grad_norm": 0.2779921889305115,
+      "learning_rate": 9.028380859183164e-06,
+      "loss": 0.3978,
       "step": 101170
     },
     {
-      "epoch": 3.56,
-      "learning_rate": 1.0131088061423779e-05,
-      "loss": 0.2743,
+      "epoch": 3.6463401448805275,
+      "grad_norm": 0.30054476857185364,
+      "learning_rate": 9.02613599216512e-06,
+      "loss": 0.3748,
       "step": 101175
     },
     {
-      "epoch": 3.56,
-      "learning_rate": 1.0128798054024954e-05,
-      "loss": 0.2507,
+      "epoch": 3.6465203445417522,
+      "grad_norm": 0.27546700835227966,
+      "learning_rate": 9.023891342787888e-06,
+      "loss": 0.3651,
       "step": 101180
     },
     {
-      "epoch": 3.56,
-      "learning_rate": 1.0126508239718888e-05,
-      "loss": 0.2571,
+      "epoch": 3.646700544202977,
+      "grad_norm": 0.23770831525325775,
+      "learning_rate": 9.021646911082044e-06,
+      "loss": 0.3828,
       "step": 101185
     },
     {
-      "epoch": 3.56,
-      "learning_rate": 1.01242186185353e-05,
-      "loss": 0.2496,
+      "epoch": 3.6468807438642017,
+      "grad_norm": 0.2967265248298645,
+      "learning_rate": 9.019402697078167e-06,
+      "loss": 0.3796,
       "step": 101190
     },
     {
-      "epoch": 3.56,
-      "learning_rate": 1.0121929190503929e-05,
-      "loss": 0.2665,
+      "epoch": 3.6470609435254264,
+      "grad_norm": 0.22645854949951172,
+      "learning_rate": 9.017158700806835e-06,
+      "loss": 0.3777,
       "step": 101195
     },
     {
-      "epoch": 3.56,
-      "learning_rate": 1.0119639955654512e-05,
-      "loss": 0.2489,
+      "epoch": 3.6472411431866507,
+      "grad_norm": 0.21501436829566956,
+      "learning_rate": 9.014914922298612e-06,
+      "loss": 0.3483,
       "step": 101200
     },
     {
-      "epoch": 3.56,
-      "learning_rate": 1.0117350914016754e-05,
-      "loss": 0.2636,
+      "epoch": 3.6474213428478754,
+      "grad_norm": 0.25435328483581543,
+      "learning_rate": 9.012671361584088e-06,
+      "loss": 0.4062,
       "step": 101205
     },
     {
-      "epoch": 3.56,
-      "learning_rate": 1.0115062065620398e-05,
-      "loss": 0.2412,
+      "epoch": 3.6476015425091,
+      "grad_norm": 0.21039626002311707,
+      "learning_rate": 9.010428018693823e-06,
+      "loss": 0.3534,
       "step": 101210
     },
     {
-      "epoch": 3.56,
-      "learning_rate": 1.011277341049515e-05,
-      "loss": 0.2654,
+      "epoch": 3.6477817421703245,
+      "grad_norm": 0.24590708315372467,
+      "learning_rate": 9.008184893658378e-06,
+      "loss": 0.3897,
       "step": 101215
     },
     {
-      "epoch": 3.56,
-      "learning_rate": 1.0110484948670726e-05,
-      "loss": 0.2726,
+      "epoch": 3.647961941831549,
+      "grad_norm": 0.23743712902069092,
+      "learning_rate": 9.005941986508318e-06,
+      "loss": 0.375,
       "step": 101220
     },
     {
-      "epoch": 3.56,
-      "learning_rate": 1.0108196680176835e-05,
-      "loss": 0.251,
+      "epoch": 3.648142141492774,
+      "grad_norm": 0.23340876400470734,
+      "learning_rate": 9.0036992972742e-06,
+      "loss": 0.368,
       "step": 101225
     },
     {
-      "epoch": 3.56,
-      "learning_rate": 1.0105908605043204e-05,
-      "loss": 0.2586,
+      "epoch": 3.6483223411539987,
+      "grad_norm": 0.2544988691806793,
+      "learning_rate": 9.001456825986579e-06,
+      "loss": 0.3809,
       "step": 101230
     },
     {
-      "epoch": 3.56,
-      "learning_rate": 1.0103620723299537e-05,
-      "loss": 0.2657,
+      "epoch": 3.6485025408152234,
+      "grad_norm": 0.24296405911445618,
+      "learning_rate": 8.999214572676015e-06,
+      "loss": 0.3803,
       "step": 101235
     },
     {
-      "epoch": 3.56,
-      "learning_rate": 1.0101333034975535e-05,
-      "loss": 0.2426,
+      "epoch": 3.648682740476448,
+      "grad_norm": 0.2581973969936371,
+      "learning_rate": 8.996972537373057e-06,
+      "loss": 0.364,
       "step": 101240
     },
     {
-      "epoch": 3.56,
-      "learning_rate": 1.0099045540100898e-05,
-      "loss": 0.261,
+      "epoch": 3.6488629401376724,
+      "grad_norm": 0.22508592903614044,
+      "learning_rate": 8.99473072010825e-06,
+      "loss": 0.3848,
       "step": 101245
     },
     {
-      "epoch": 3.56,
-      "learning_rate": 1.0096758238705334e-05,
-      "loss": 0.2671,
+      "epoch": 3.649043139798897,
+      "grad_norm": 0.2491467446088791,
+      "learning_rate": 8.992489120912138e-06,
+      "loss": 0.3736,
       "step": 101250
     },
     {
-      "epoch": 3.56,
-      "learning_rate": 1.009447113081855e-05,
-      "loss": 0.2714,
+      "epoch": 3.649223339460122,
+      "grad_norm": 0.24909654259681702,
+      "learning_rate": 8.990247739815252e-06,
+      "loss": 0.3896,
       "step": 101255
     },
     {
-      "epoch": 3.56,
-      "learning_rate": 1.0092184216470238e-05,
-      "loss": 0.254,
+      "epoch": 3.649403539121346,
+      "grad_norm": 0.25092822313308716,
+      "learning_rate": 8.988006576848159e-06,
+      "loss": 0.3498,
       "step": 101260
     },
     {
-      "epoch": 3.56,
-      "learning_rate": 1.0089897495690079e-05,
-      "loss": 0.2622,
+      "epoch": 3.649583738782571,
+      "grad_norm": 0.2587062120437622,
+      "learning_rate": 8.98576563204136e-06,
+      "loss": 0.3853,
       "step": 101265
     },
     {
-      "epoch": 3.56,
-      "learning_rate": 1.0087610968507786e-05,
-      "loss": 0.2505,
+      "epoch": 3.6497639384437957,
+      "grad_norm": 0.27524474263191223,
+      "learning_rate": 8.983524905425413e-06,
+      "loss": 0.3667,
       "step": 101270
     },
     {
-      "epoch": 3.56,
-      "learning_rate": 1.0085324634953033e-05,
-      "loss": 0.2544,
+      "epoch": 3.6499441381050204,
+      "grad_norm": 0.2666233479976654,
+      "learning_rate": 8.981284397030837e-06,
+      "loss": 0.37,
       "step": 101275
     },
     {
-      "epoch": 3.56,
-      "learning_rate": 1.0083038495055511e-05,
-      "loss": 0.2556,
+      "epoch": 3.650124337766245,
+      "grad_norm": 0.21499104797840118,
+      "learning_rate": 8.979044106888152e-06,
+      "loss": 0.3844,
       "step": 101280
     },
     {
-      "epoch": 3.56,
-      "learning_rate": 1.0080752548844895e-05,
-      "loss": 0.241,
+      "epoch": 3.65030453742747,
+      "grad_norm": 0.22629836201667786,
+      "learning_rate": 8.976804035027905e-06,
+      "loss": 0.3803,
       "step": 101285
     },
     {
-      "epoch": 3.56,
-      "learning_rate": 1.0078466796350885e-05,
-      "loss": 0.2487,
+      "epoch": 3.650484737088694,
+      "grad_norm": 0.24577264487743378,
+      "learning_rate": 8.974564181480594e-06,
+      "loss": 0.3886,
       "step": 101290
     },
     {
-      "epoch": 3.56,
-      "learning_rate": 1.0076181237603152e-05,
-      "loss": 0.2585,
+      "epoch": 3.650664936749919,
+      "grad_norm": 0.22071078419685364,
+      "learning_rate": 8.972324546276733e-06,
+      "loss": 0.402,
       "step": 101295
     },
     {
-      "epoch": 3.56,
-      "learning_rate": 1.0073895872631359e-05,
-      "loss": 0.265,
+      "epoch": 3.6508451364111436,
+      "grad_norm": 0.2849634289741516,
+      "learning_rate": 8.97008512944686e-06,
+      "loss": 0.3814,
       "step": 101300
     },
     {
-      "epoch": 3.56,
-      "learning_rate": 1.00716107014652e-05,
-      "loss": 0.2519,
+      "epoch": 3.6510253360723683,
+      "grad_norm": 0.22348150610923767,
+      "learning_rate": 8.967845931021469e-06,
+      "loss": 0.375,
       "step": 101305
     },
     {
-      "epoch": 3.56,
-      "learning_rate": 1.0069325724134328e-05,
-      "loss": 0.2672,
+      "epoch": 3.6512055357335926,
+      "grad_norm": 0.25602954626083374,
+      "learning_rate": 8.965606951031074e-06,
+      "loss": 0.3992,
       "step": 101310
     },
     {
-      "epoch": 3.56,
-      "learning_rate": 1.0067040940668435e-05,
-      "loss": 0.241,
+      "epoch": 3.6513857353948174,
+      "grad_norm": 0.270134836435318,
+      "learning_rate": 8.96336818950618e-06,
+      "loss": 0.3896,
       "step": 101315
     },
     {
-      "epoch": 3.56,
-      "learning_rate": 1.0064756351097163e-05,
-      "loss": 0.2439,
+      "epoch": 3.651565935056042,
+      "grad_norm": 0.2468041628599167,
+      "learning_rate": 8.961129646477281e-06,
+      "loss": 0.3846,
       "step": 101320
     },
     {
-      "epoch": 3.56,
-      "learning_rate": 1.0062471955450194e-05,
-      "loss": 0.2713,
+      "epoch": 3.651746134717267,
+      "grad_norm": 0.22694005072116852,
+      "learning_rate": 8.958891321974896e-06,
+      "loss": 0.3719,
       "step": 101325
     },
     {
-      "epoch": 3.57,
-      "learning_rate": 1.0060187753757183e-05,
-      "loss": 0.25,
+      "epoch": 3.6519263343784916,
+      "grad_norm": 0.3231867253780365,
+      "learning_rate": 8.956653216029509e-06,
+      "loss": 0.4239,
       "step": 101330
     },
     {
-      "epoch": 3.57,
-      "learning_rate": 1.005790374604779e-05,
-      "loss": 0.2872,
+      "epoch": 3.652106534039716,
+      "grad_norm": 0.23950925469398499,
+      "learning_rate": 8.954415328671617e-06,
+      "loss": 0.3259,
       "step": 101335
     },
     {
-      "epoch": 3.57,
-      "learning_rate": 1.005561993235166e-05,
-      "loss": 0.2552,
+      "epoch": 3.6522867337009406,
+      "grad_norm": 0.30801981687545776,
+      "learning_rate": 8.95217765993171e-06,
+      "loss": 0.3592,
       "step": 101340
     },
     {
-      "epoch": 3.57,
-      "learning_rate": 1.0053336312698466e-05,
-      "loss": 0.2638,
+      "epoch": 3.6524669333621653,
+      "grad_norm": 0.23010605573654175,
+      "learning_rate": 8.949940209840266e-06,
+      "loss": 0.3641,
       "step": 101345
     },
     {
-      "epoch": 3.57,
-      "learning_rate": 1.0051052887117848e-05,
-      "loss": 0.2532,
+      "epoch": 3.65264713302339,
+      "grad_norm": 0.2432517111301422,
+      "learning_rate": 8.947702978427786e-06,
+      "loss": 0.3871,
       "step": 101350
     },
     {
-      "epoch": 3.57,
-      "learning_rate": 1.004876965563945e-05,
-      "loss": 0.2728,
+      "epoch": 3.6528273326846143,
+      "grad_norm": 0.21140114963054657,
+      "learning_rate": 8.945465965724756e-06,
+      "loss": 0.3849,
       "step": 101355
     },
     {
-      "epoch": 3.57,
-      "learning_rate": 1.0046486618292933e-05,
-      "loss": 0.2583,
+      "epoch": 3.653007532345839,
+      "grad_norm": 0.263443261384964,
+      "learning_rate": 8.943229171761628e-06,
+      "loss": 0.4131,
       "step": 101360
     },
     {
-      "epoch": 3.57,
-      "learning_rate": 1.0044203775107921e-05,
-      "loss": 0.2902,
+      "epoch": 3.653187732007064,
+      "grad_norm": 0.19522389769554138,
+      "learning_rate": 8.940992596568897e-06,
+      "loss": 0.3953,
       "step": 101365
     },
     {
-      "epoch": 3.57,
-      "learning_rate": 1.0041921126114079e-05,
-      "loss": 0.2379,
+      "epoch": 3.6533679316682885,
+      "grad_norm": 0.26923754811286926,
+      "learning_rate": 8.938756240177037e-06,
+      "loss": 0.3825,
       "step": 101370
     },
     {
-      "epoch": 3.57,
-      "learning_rate": 1.0039638671341031e-05,
-      "loss": 0.2592,
+      "epoch": 3.6535481313295133,
+      "grad_norm": 0.24230341613292694,
+      "learning_rate": 8.936520102616513e-06,
+      "loss": 0.3723,
       "step": 101375
     },
     {
-      "epoch": 3.57,
-      "learning_rate": 1.003735641081841e-05,
-      "loss": 0.2424,
+      "epoch": 3.6537283309907376,
+      "grad_norm": 0.27861183881759644,
+      "learning_rate": 8.934284183917793e-06,
+      "loss": 0.3888,
       "step": 101380
     },
     {
-      "epoch": 3.57,
-      "learning_rate": 1.0035074344575863e-05,
-      "loss": 0.2278,
+      "epoch": 3.6539085306519623,
+      "grad_norm": 0.18770521879196167,
+      "learning_rate": 8.932048484111333e-06,
+      "loss": 0.3635,
       "step": 101385
     },
     {
-      "epoch": 3.57,
-      "learning_rate": 1.003279247264301e-05,
-      "loss": 0.2786,
+      "epoch": 3.654088730313187,
+      "grad_norm": 0.2484419047832489,
+      "learning_rate": 8.929813003227608e-06,
+      "loss": 0.3887,
       "step": 101390
     },
     {
-      "epoch": 3.57,
-      "learning_rate": 1.0030510795049486e-05,
-      "loss": 0.2683,
+      "epoch": 3.6542689299744118,
+      "grad_norm": 0.21241828799247742,
+      "learning_rate": 8.927577741297072e-06,
+      "loss": 0.3797,
       "step": 101395
     },
     {
-      "epoch": 3.57,
-      "learning_rate": 1.0028229311824904e-05,
-      "loss": 0.2549,
+      "epoch": 3.654449129635636,
+      "grad_norm": 0.2524920701980591,
+      "learning_rate": 8.925342698350175e-06,
+      "loss": 0.3909,
       "step": 101400
     },
     {
-      "epoch": 3.57,
-      "learning_rate": 1.0025948022998908e-05,
-      "loss": 0.2537,
+      "epoch": 3.654629329296861,
+      "grad_norm": 0.2676452696323395,
+      "learning_rate": 8.923107874417372e-06,
+      "loss": 0.3485,
       "step": 101405
     },
     {
-      "epoch": 3.57,
-      "learning_rate": 1.00236669286011e-05,
-      "loss": 0.2422,
+      "epoch": 3.6548095289580855,
+      "grad_norm": 0.21999165415763855,
+      "learning_rate": 8.920873269529112e-06,
+      "loss": 0.3712,
       "step": 101410
     },
     {
-      "epoch": 3.57,
-      "learning_rate": 1.0021386028661115e-05,
-      "loss": 0.2395,
+      "epoch": 3.6549897286193103,
+      "grad_norm": 0.23482544720172882,
+      "learning_rate": 8.918638883715832e-06,
+      "loss": 0.3802,
       "step": 101415
     },
     {
-      "epoch": 3.57,
-      "learning_rate": 1.0019105323208549e-05,
-      "loss": 0.2659,
+      "epoch": 3.655169928280535,
+      "grad_norm": 0.2594209313392639,
+      "learning_rate": 8.916404717007992e-06,
+      "loss": 0.3737,
       "step": 101420
     },
     {
-      "epoch": 3.57,
-      "learning_rate": 1.0016824812273042e-05,
-      "loss": 0.2536,
+      "epoch": 3.6553501279417597,
+      "grad_norm": 0.20219695568084717,
+      "learning_rate": 8.914170769436025e-06,
+      "loss": 0.3846,
       "step": 101425
     },
     {
-      "epoch": 3.57,
-      "learning_rate": 1.0014544495884188e-05,
-      "loss": 0.2819,
+      "epoch": 3.655530327602984,
+      "grad_norm": 0.24486181139945984,
+      "learning_rate": 8.911937041030365e-06,
+      "loss": 0.3891,
       "step": 101430
     },
     {
-      "epoch": 3.57,
-      "learning_rate": 1.001226437407159e-05,
-      "loss": 0.2659,
+      "epoch": 3.6557105272642088,
+      "grad_norm": 0.19330917298793793,
+      "learning_rate": 8.90970353182145e-06,
+      "loss": 0.3704,
       "step": 101435
     },
     {
-      "epoch": 3.57,
-      "learning_rate": 1.000998444686487e-05,
-      "loss": 0.2414,
+      "epoch": 3.6558907269254335,
+      "grad_norm": 0.21298684179782867,
+      "learning_rate": 8.907470241839703e-06,
+      "loss": 0.3495,
       "step": 101440
     },
     {
-      "epoch": 3.57,
-      "learning_rate": 1.0007704714293623e-05,
-      "loss": 0.2716,
+      "epoch": 3.6560709265866578,
+      "grad_norm": 0.25250443816185,
+      "learning_rate": 8.905237171115563e-06,
+      "loss": 0.4116,
       "step": 101445
     },
     {
-      "epoch": 3.57,
-      "learning_rate": 1.0005425176387448e-05,
-      "loss": 0.2525,
+      "epoch": 3.6562511262478825,
+      "grad_norm": 0.20300345122814178,
+      "learning_rate": 8.90300431967945e-06,
+      "loss": 0.3616,
       "step": 101450
     },
     {
-      "epoch": 3.57,
-      "learning_rate": 1.000314583317594e-05,
-      "loss": 0.2505,
+      "epoch": 3.6564313259091072,
+      "grad_norm": 0.2571481466293335,
+      "learning_rate": 8.900771687561787e-06,
+      "loss": 0.3844,
       "step": 101455
     },
     {
-      "epoch": 3.57,
-      "learning_rate": 1.0000866684688706e-05,
-      "loss": 0.2487,
+      "epoch": 3.656611525570332,
+      "grad_norm": 0.18634074926376343,
+      "learning_rate": 8.898539274792997e-06,
+      "loss": 0.3538,
       "step": 101460
     },
     {
-      "epoch": 3.57,
-      "learning_rate": 9.998587730955336e-06,
-      "loss": 0.2531,
+      "epoch": 3.6567917252315567,
+      "grad_norm": 0.2489531934261322,
+      "learning_rate": 8.896307081403479e-06,
+      "loss": 0.3721,
       "step": 101465
     },
     {
-      "epoch": 3.57,
-      "learning_rate": 9.996308972005408e-06,
-      "loss": 0.2609,
+      "epoch": 3.6569719248927814,
+      "grad_norm": 0.2489020973443985,
+      "learning_rate": 8.894075107423678e-06,
+      "loss": 0.3765,
       "step": 101470
     },
     {
-      "epoch": 3.57,
-      "learning_rate": 9.99403040786852e-06,
-      "loss": 0.26,
+      "epoch": 3.6571521245540057,
+      "grad_norm": 0.19710054993629456,
+      "learning_rate": 8.891843352883978e-06,
+      "loss": 0.3843,
       "step": 101475
     },
     {
-      "epoch": 3.57,
-      "learning_rate": 9.991752038574267e-06,
-      "loss": 0.2382,
+      "epoch": 3.6573323242152305,
+      "grad_norm": 0.22611059248447418,
+      "learning_rate": 8.889611817814783e-06,
+      "loss": 0.3832,
       "step": 101480
     },
     {
-      "epoch": 3.57,
-      "learning_rate": 9.989473864152221e-06,
-      "loss": 0.2853,
+      "epoch": 3.657512523876455,
+      "grad_norm": 0.2789013683795929,
+      "learning_rate": 8.887380502246517e-06,
+      "loss": 0.3685,
       "step": 101485
     },
     {
-      "epoch": 3.57,
-      "learning_rate": 9.987195884631967e-06,
-      "loss": 0.2495,
+      "epoch": 3.6576927235376795,
+      "grad_norm": 0.17864690721035004,
+      "learning_rate": 8.885149406209573e-06,
+      "loss": 0.3661,
       "step": 101490
     },
     {
-      "epoch": 3.57,
-      "learning_rate": 9.984918100043072e-06,
-      "loss": 0.2519,
+      "epoch": 3.6578729231989042,
+      "grad_norm": 0.23747530579566956,
+      "learning_rate": 8.882918529734346e-06,
+      "loss": 0.4134,
       "step": 101495
     },
     {
-      "epoch": 3.57,
-      "learning_rate": 9.982640510415128e-06,
-      "loss": 0.2421,
+      "epoch": 3.658053122860129,
+      "grad_norm": 0.25165247917175293,
+      "learning_rate": 8.880687872851237e-06,
+      "loss": 0.3606,
       "step": 101500
     },
     {
-      "epoch": 3.57,
-      "eval_loss": 0.25163590908050537,
-      "eval_runtime": 10.5398,
-      "eval_samples_per_second": 9.488,
-      "eval_steps_per_second": 9.488,
+      "epoch": 3.658053122860129,
+      "eval_loss": 0.4290784001350403,
+      "eval_runtime": 3.529,
+      "eval_samples_per_second": 28.337,
+      "eval_steps_per_second": 7.084,
       "step": 101500
     },
     {
-      "epoch": 3.57,
-      "learning_rate": 9.9803631157777e-06,
-      "loss": 0.2317,
+      "epoch": 3.6582333225213537,
+      "grad_norm": 0.21503256261348724,
+      "learning_rate": 8.878457435590626e-06,
+      "loss": 0.3689,
       "step": 101505
     },
     {
-      "epoch": 3.57,
-      "learning_rate": 9.978085916160359e-06,
-      "loss": 0.265,
+      "epoch": 3.6584135221825784,
+      "grad_norm": 0.1987353265285492,
+      "learning_rate": 8.876227217982916e-06,
+      "loss": 0.3373,
       "step": 101510
     },
     {
-      "epoch": 3.57,
-      "learning_rate": 9.975808911592662e-06,
-      "loss": 0.2652,
+      "epoch": 3.658593721843803,
+      "grad_norm": 0.29664599895477295,
+      "learning_rate": 8.87399722005849e-06,
+      "loss": 0.3837,
       "step": 101515
     },
     {
-      "epoch": 3.57,
-      "learning_rate": 9.973532102104197e-06,
-      "loss": 0.2534,
+      "epoch": 3.6587739215050274,
+      "grad_norm": 0.22414027154445648,
+      "learning_rate": 8.871767441847734e-06,
+      "loss": 0.3662,
       "step": 101520
     },
     {
-      "epoch": 3.57,
-      "learning_rate": 9.971255487724504e-06,
-      "loss": 0.2545,
+      "epoch": 3.658954121166252,
+      "grad_norm": 0.2331104874610901,
+      "learning_rate": 8.869537883381022e-06,
+      "loss": 0.3745,
       "step": 101525
     },
     {
-      "epoch": 3.57,
-      "learning_rate": 9.968979068483164e-06,
-      "loss": 0.2848,
+      "epoch": 3.659134320827477,
+      "grad_norm": 0.2712317407131195,
+      "learning_rate": 8.86730854468872e-06,
+      "loss": 0.3619,
       "step": 101530
     },
     {
-      "epoch": 3.57,
-      "learning_rate": 9.966702844409714e-06,
-      "loss": 0.2729,
+      "epoch": 3.659314520488701,
+      "grad_norm": 0.22348475456237793,
+      "learning_rate": 8.865079425801228e-06,
+      "loss": 0.3921,
       "step": 101535
     },
     {
-      "epoch": 3.57,
-      "learning_rate": 9.964426815533732e-06,
-      "loss": 0.2398,
+      "epoch": 3.659494720149926,
+      "grad_norm": 0.2932550311088562,
+      "learning_rate": 8.862850526748917e-06,
+      "loss": 0.3714,
       "step": 101540
     },
     {
-      "epoch": 3.57,
-      "learning_rate": 9.962150981884757e-06,
-      "loss": 0.2565,
+      "epoch": 3.6596749198111507,
+      "grad_norm": 0.26502054929733276,
+      "learning_rate": 8.860621847562123e-06,
+      "loss": 0.3736,
       "step": 101545
     },
     {
-      "epoch": 3.57,
-      "learning_rate": 9.959875343492334e-06,
-      "loss": 0.2747,
+      "epoch": 3.6598551194723754,
+      "grad_norm": 0.25092580914497375,
+      "learning_rate": 8.858393388271238e-06,
+      "loss": 0.3978,
       "step": 101550
     },
     {
-      "epoch": 3.57,
-      "learning_rate": 9.957599900386024e-06,
-      "loss": 0.2581,
+      "epoch": 3.6600353191336,
+      "grad_norm": 0.2842051684856415,
+      "learning_rate": 8.85616514890662e-06,
+      "loss": 0.3674,
       "step": 101555
     },
     {
-      "epoch": 3.57,
-      "learning_rate": 9.955324652595369e-06,
-      "loss": 0.2666,
+      "epoch": 3.660215518794825,
+      "grad_norm": 0.2661273777484894,
+      "learning_rate": 8.853937129498627e-06,
+      "loss": 0.3835,
       "step": 101560
     },
     {
-      "epoch": 3.57,
-      "learning_rate": 9.953049600149908e-06,
-      "loss": 0.2615,
+      "epoch": 3.660395718456049,
+      "grad_norm": 0.23841172456741333,
+      "learning_rate": 8.851709330077615e-06,
+      "loss": 0.3654,
       "step": 101565
     },
     {
-      "epoch": 3.57,
-      "learning_rate": 9.950774743079172e-06,
-      "loss": 0.2477,
+      "epoch": 3.660575918117274,
+      "grad_norm": 0.20587527751922607,
+      "learning_rate": 8.849481750673927e-06,
+      "loss": 0.383,
       "step": 101570
     },
     {
-      "epoch": 3.57,
-      "learning_rate": 9.948500081412717e-06,
-      "loss": 0.2734,
+      "epoch": 3.6607561177784986,
+      "grad_norm": 0.2173176258802414,
+      "learning_rate": 8.84725439131793e-06,
+      "loss": 0.4062,
       "step": 101575
     },
     {
-      "epoch": 3.57,
-      "learning_rate": 9.94622561518006e-06,
-      "loss": 0.2722,
+      "epoch": 3.6609363174397234,
+      "grad_norm": 0.201410710811615,
+      "learning_rate": 8.845027252039969e-06,
+      "loss": 0.3691,
       "step": 101580
     },
     {
-      "epoch": 3.57,
-      "learning_rate": 9.943951344410751e-06,
-      "loss": 0.248,
+      "epoch": 3.6611165171009477,
+      "grad_norm": 0.24154742062091827,
+      "learning_rate": 8.842800332870382e-06,
+      "loss": 0.3661,
       "step": 101585
     },
     {
-      "epoch": 3.57,
-      "learning_rate": 9.941677269134303e-06,
-      "loss": 0.259,
+      "epoch": 3.6612967167621724,
+      "grad_norm": 0.18455544114112854,
+      "learning_rate": 8.84057363383951e-06,
+      "loss": 0.3598,
       "step": 101590
     },
     {
-      "epoch": 3.57,
-      "learning_rate": 9.939403389380261e-06,
-      "loss": 0.2713,
+      "epoch": 3.661476916423397,
+      "grad_norm": 0.24699454009532928,
+      "learning_rate": 8.838347154977697e-06,
+      "loss": 0.3765,
       "step": 101595
     },
     {
-      "epoch": 3.57,
-      "learning_rate": 9.937129705178142e-06,
-      "loss": 0.2572,
+      "epoch": 3.661657116084622,
+      "grad_norm": 0.20563150942325592,
+      "learning_rate": 8.836120896315267e-06,
+      "loss": 0.3507,
       "step": 101600
     },
     {
-      "epoch": 3.57,
-      "learning_rate": 9.934856216557465e-06,
-      "loss": 0.2642,
+      "epoch": 3.6618373157458466,
+      "grad_norm": 0.21455486118793488,
+      "learning_rate": 8.833894857882566e-06,
+      "loss": 0.365,
       "step": 101605
     },
     {
-      "epoch": 3.57,
-      "learning_rate": 9.932582923547745e-06,
-      "loss": 0.2454,
+      "epoch": 3.662017515407071,
+      "grad_norm": 0.237900048494339,
+      "learning_rate": 8.83166903970992e-06,
+      "loss": 0.3663,
       "step": 101610
     },
     {
-      "epoch": 3.58,
-      "learning_rate": 9.930309826178513e-06,
-      "loss": 0.2612,
+      "epoch": 3.6621977150682956,
+      "grad_norm": 0.24725571274757385,
+      "learning_rate": 8.82944344182765e-06,
+      "loss": 0.382,
       "step": 101615
     },
     {
-      "epoch": 3.58,
-      "learning_rate": 9.928036924479276e-06,
-      "loss": 0.2666,
+      "epoch": 3.6623779147295203,
+      "grad_norm": 0.28484535217285156,
+      "learning_rate": 8.827218064266085e-06,
+      "loss": 0.4055,
       "step": 101620
     },
     {
-      "epoch": 3.58,
-      "learning_rate": 9.925764218479549e-06,
-      "loss": 0.2315,
+      "epoch": 3.662558114390745,
+      "grad_norm": 0.2544311583042145,
+      "learning_rate": 8.824992907055534e-06,
+      "loss": 0.3613,
       "step": 101625
     },
     {
-      "epoch": 3.58,
-      "learning_rate": 9.923491708208829e-06,
-      "loss": 0.2454,
+      "epoch": 3.6627383140519694,
+      "grad_norm": 0.21593233942985535,
+      "learning_rate": 8.822767970226332e-06,
+      "loss": 0.3756,
       "step": 101630
     },
     {
-      "epoch": 3.58,
-      "learning_rate": 9.921219393696634e-06,
-      "loss": 0.2555,
+      "epoch": 3.662918513713194,
+      "grad_norm": 0.2815852463245392,
+      "learning_rate": 8.820543253808783e-06,
+      "loss": 0.4064,
       "step": 101635
     },
     {
-      "epoch": 3.58,
-      "learning_rate": 9.918947274972476e-06,
-      "loss": 0.2696,
+      "epoch": 3.663098713374419,
+      "grad_norm": 0.299152135848999,
+      "learning_rate": 8.8183187578332e-06,
+      "loss": 0.4224,
       "step": 101640
     },
     {
-      "epoch": 3.58,
-      "learning_rate": 9.916675352065849e-06,
-      "loss": 0.2541,
+      "epoch": 3.6632789130356436,
+      "grad_norm": 0.24634763598442078,
+      "learning_rate": 8.81609448232989e-06,
+      "loss": 0.3834,
       "step": 101645
     },
     {
-      "epoch": 3.58,
-      "learning_rate": 9.914403625006245e-06,
-      "loss": 0.244,
+      "epoch": 3.6634591126968683,
+      "grad_norm": 0.22788767516613007,
+      "learning_rate": 8.813870427329155e-06,
+      "loss": 0.3795,
       "step": 101650
     },
     {
-      "epoch": 3.58,
-      "learning_rate": 9.912132093823173e-06,
-      "loss": 0.2263,
+      "epoch": 3.663639312358093,
+      "grad_norm": 0.26291418075561523,
+      "learning_rate": 8.811646592861314e-06,
+      "loss": 0.376,
       "step": 101655
     },
     {
-      "epoch": 3.58,
-      "learning_rate": 9.909860758546128e-06,
-      "loss": 0.2481,
+      "epoch": 3.6638195120193173,
+      "grad_norm": 0.191041499376297,
+      "learning_rate": 8.809422978956647e-06,
+      "loss": 0.3674,
       "step": 101660
     },
     {
-      "epoch": 3.58,
-      "learning_rate": 9.907589619204583e-06,
-      "loss": 0.2358,
+      "epoch": 3.663999711680542,
+      "grad_norm": 0.28010818362236023,
+      "learning_rate": 8.80719958564545e-06,
+      "loss": 0.4071,
       "step": 101665
     },
     {
-      "epoch": 3.58,
-      "learning_rate": 9.905318675828052e-06,
-      "loss": 0.2522,
+      "epoch": 3.664179911341767,
+      "grad_norm": 0.21684035658836365,
+      "learning_rate": 8.804976412958029e-06,
+      "loss": 0.3826,
       "step": 101670
     },
     {
-      "epoch": 3.58,
-      "learning_rate": 9.903047928446007e-06,
-      "loss": 0.2481,
+      "epoch": 3.664360111002991,
+      "grad_norm": 0.26138171553611755,
+      "learning_rate": 8.802753460924674e-06,
+      "loss": 0.3812,
       "step": 101675
     },
     {
-      "epoch": 3.58,
-      "learning_rate": 9.900777377087939e-06,
-      "loss": 0.2524,
+      "epoch": 3.664540310664216,
+      "grad_norm": 0.20288319885730743,
+      "learning_rate": 8.800530729575665e-06,
+      "loss": 0.3448,
       "step": 101680
     },
     {
-      "epoch": 3.58,
-      "learning_rate": 9.898507021783315e-06,
-      "loss": 0.2466,
+      "epoch": 3.6647205103254406,
+      "grad_norm": 0.2686285674571991,
+      "learning_rate": 8.798308218941287e-06,
+      "loss": 0.414,
       "step": 101685
     },
     {
-      "epoch": 3.58,
-      "learning_rate": 9.896236862561625e-06,
-      "loss": 0.2591,
+      "epoch": 3.6649007099866653,
+      "grad_norm": 0.26645994186401367,
+      "learning_rate": 8.796085929051814e-06,
+      "loss": 0.3848,
       "step": 101690
     },
     {
-      "epoch": 3.58,
-      "learning_rate": 9.893966899452356e-06,
-      "loss": 0.2576,
+      "epoch": 3.66508090964789,
+      "grad_norm": 0.22347326576709747,
+      "learning_rate": 8.793863859937543e-06,
+      "loss": 0.3875,
       "step": 101695
     },
     {
-      "epoch": 3.58,
-      "learning_rate": 9.89169713248497e-06,
-      "loss": 0.2258,
+      "epoch": 3.6652611093091148,
+      "grad_norm": 0.21688346564769745,
+      "learning_rate": 8.79164201162874e-06,
+      "loss": 0.3964,
       "step": 101700
     },
     {
-      "epoch": 3.58,
-      "learning_rate": 9.889427561688933e-06,
-      "loss": 0.2534,
+      "epoch": 3.665441308970339,
+      "grad_norm": 0.2277715802192688,
+      "learning_rate": 8.789420384155675e-06,
+      "loss": 0.388,
       "step": 101705
     },
     {
-      "epoch": 3.58,
-      "learning_rate": 9.88715818709373e-06,
-      "loss": 0.2556,
+      "epoch": 3.6656215086315638,
+      "grad_norm": 0.23789222538471222,
+      "learning_rate": 8.78719897754862e-06,
+      "loss": 0.3821,
       "step": 101710
     },
     {
-      "epoch": 3.58,
-      "learning_rate": 9.884889008728814e-06,
-      "loss": 0.239,
+      "epoch": 3.6658017082927885,
+      "grad_norm": 0.24796278774738312,
+      "learning_rate": 8.784977791837831e-06,
+      "loss": 0.3832,
       "step": 101715
     },
     {
-      "epoch": 3.58,
-      "learning_rate": 9.882620026623657e-06,
-      "loss": 0.2524,
+      "epoch": 3.665981907954013,
+      "grad_norm": 0.28839823603630066,
+      "learning_rate": 8.782756827053588e-06,
+      "loss": 0.3859,
       "step": 101720
     },
     {
-      "epoch": 3.58,
-      "learning_rate": 9.880351240807707e-06,
-      "loss": 0.2553,
+      "epoch": 3.6661621076152375,
+      "grad_norm": 0.2025182694196701,
+      "learning_rate": 8.780536083226154e-06,
+      "loss": 0.3639,
       "step": 101725
     },
     {
-      "epoch": 3.58,
-      "learning_rate": 9.878082651310439e-06,
-      "loss": 0.2408,
+      "epoch": 3.6663423072764623,
+      "grad_norm": 0.24186572432518005,
+      "learning_rate": 8.778315560385756e-06,
+      "loss": 0.3622,
       "step": 101730
     },
     {
-      "epoch": 3.58,
-      "learning_rate": 9.875814258161303e-06,
-      "loss": 0.2744,
+      "epoch": 3.666522506937687,
+      "grad_norm": 0.19818326830863953,
+      "learning_rate": 8.776095258562677e-06,
+      "loss": 0.4044,
       "step": 101735
     },
     {
-      "epoch": 3.58,
-      "learning_rate": 9.873546061389743e-06,
-      "loss": 0.2682,
+      "epoch": 3.6667027065989117,
+      "grad_norm": 0.2045317143201828,
+      "learning_rate": 8.773875177787161e-06,
+      "loss": 0.3493,
       "step": 101740
     },
     {
-      "epoch": 3.58,
-      "learning_rate": 9.871278061025227e-06,
-      "loss": 0.2772,
+      "epoch": 3.6668829062601365,
+      "grad_norm": 0.2028067409992218,
+      "learning_rate": 8.771655318089445e-06,
+      "loss": 0.3869,
       "step": 101745
     },
     {
-      "epoch": 3.58,
-      "learning_rate": 9.869010257097185e-06,
-      "loss": 0.2773,
+      "epoch": 3.6670631059213608,
+      "grad_norm": 0.24124419689178467,
+      "learning_rate": 8.769435679499798e-06,
+      "loss": 0.3822,
       "step": 101750
     },
     {
-      "epoch": 3.58,
-      "learning_rate": 9.866742649635085e-06,
-      "loss": 0.2547,
+      "epoch": 3.6672433055825855,
+      "grad_norm": 0.2072051465511322,
+      "learning_rate": 8.767216262048433e-06,
+      "loss": 0.3652,
       "step": 101755
     },
     {
-      "epoch": 3.58,
-      "learning_rate": 9.864475238668358e-06,
-      "loss": 0.2462,
+      "epoch": 3.6674235052438102,
+      "grad_norm": 0.22216108441352844,
+      "learning_rate": 8.764997065765612e-06,
+      "loss": 0.404,
       "step": 101760
     },
     {
-      "epoch": 3.58,
-      "learning_rate": 9.862208024226435e-06,
-      "loss": 0.2654,
+      "epoch": 3.6676037049050345,
+      "grad_norm": 0.23680134117603302,
+      "learning_rate": 8.762778090681562e-06,
+      "loss": 0.3964,
       "step": 101765
     },
     {
-      "epoch": 3.58,
-      "learning_rate": 9.859941006338773e-06,
-      "loss": 0.2604,
+      "epoch": 3.6677839045662592,
+      "grad_norm": 0.29067909717559814,
+      "learning_rate": 8.760559336826519e-06,
+      "loss": 0.3897,
       "step": 101770
     },
     {
-      "epoch": 3.58,
-      "learning_rate": 9.8576741850348e-06,
-      "loss": 0.258,
+      "epoch": 3.667964104227484,
+      "grad_norm": 0.244569793343544,
+      "learning_rate": 8.758340804230709e-06,
+      "loss": 0.3949,
       "step": 101775
     },
     {
-      "epoch": 3.58,
-      "learning_rate": 9.855407560343941e-06,
-      "loss": 0.2535,
+      "epoch": 3.6681443038887087,
+      "grad_norm": 0.22354869544506073,
+      "learning_rate": 8.75612249292436e-06,
+      "loss": 0.3871,
       "step": 101780
     },
     {
-      "epoch": 3.58,
-      "learning_rate": 9.853141132295643e-06,
-      "loss": 0.2421,
+      "epoch": 3.6683245035499334,
+      "grad_norm": 0.2619014084339142,
+      "learning_rate": 8.75390440293769e-06,
+      "loss": 0.4095,
       "step": 101785
     },
     {
-      "epoch": 3.58,
-      "learning_rate": 9.850874900919325e-06,
-      "loss": 0.2586,
+      "epoch": 3.668504703211158,
+      "grad_norm": 0.2473766952753067,
+      "learning_rate": 8.751686534300934e-06,
+      "loss": 0.37,
       "step": 101790
     },
     {
-      "epoch": 3.58,
-      "learning_rate": 9.848608866244403e-06,
-      "loss": 0.2664,
+      "epoch": 3.6686849028723825,
+      "grad_norm": 0.24038511514663696,
+      "learning_rate": 8.749468887044308e-06,
+      "loss": 0.3576,
       "step": 101795
     },
     {
-      "epoch": 3.58,
-      "learning_rate": 9.846343028300318e-06,
-      "loss": 0.2558,
+      "epoch": 3.668865102533607,
+      "grad_norm": 0.25563234090805054,
+      "learning_rate": 8.747251461198016e-06,
+      "loss": 0.4066,
       "step": 101800
     },
     {
-      "epoch": 3.58,
-      "learning_rate": 9.844077387116474e-06,
-      "loss": 0.2521,
+      "epoch": 3.669045302194832,
+      "grad_norm": 0.2780168354511261,
+      "learning_rate": 8.745034256792281e-06,
+      "loss": 0.3688,
       "step": 101805
     },
     {
-      "epoch": 3.58,
-      "learning_rate": 9.841811942722306e-06,
-      "loss": 0.2885,
+      "epoch": 3.6692255018560567,
+      "grad_norm": 0.2796255350112915,
+      "learning_rate": 8.742817273857295e-06,
+      "loss": 0.3828,
       "step": 101810
     },
     {
-      "epoch": 3.58,
-      "learning_rate": 9.83954669514722e-06,
-      "loss": 0.2562,
+      "epoch": 3.669405701517281,
+      "grad_norm": 0.2719496190547943,
+      "learning_rate": 8.740600512423289e-06,
+      "loss": 0.3553,
       "step": 101815
     },
     {
-      "epoch": 3.58,
-      "learning_rate": 9.837281644420621e-06,
-      "loss": 0.2398,
+      "epoch": 3.6695859011785057,
+      "grad_norm": 0.22607314586639404,
+      "learning_rate": 8.73838397252046e-06,
+      "loss": 0.368,
       "step": 101820
     },
     {
-      "epoch": 3.58,
-      "learning_rate": 9.835016790571935e-06,
-      "loss": 0.2481,
+      "epoch": 3.6697661008397304,
+      "grad_norm": 0.20450633764266968,
+      "learning_rate": 8.73616765417898e-06,
+      "loss": 0.3893,
       "step": 101825
     },
     {
-      "epoch": 3.58,
-      "learning_rate": 9.832752133630563e-06,
-      "loss": 0.279,
+      "epoch": 3.669946300500955,
+      "grad_norm": 0.20725052058696747,
+      "learning_rate": 8.73395155742908e-06,
+      "loss": 0.3655,
       "step": 101830
     },
     {
-      "epoch": 3.58,
-      "learning_rate": 9.830487673625907e-06,
-      "loss": 0.2565,
+      "epoch": 3.67012650016218,
+      "grad_norm": 0.2661026418209076,
+      "learning_rate": 8.731735682300932e-06,
+      "loss": 0.3748,
       "step": 101835
     },
     {
-      "epoch": 3.58,
-      "learning_rate": 9.828223410587364e-06,
-      "loss": 0.2632,
+      "epoch": 3.670306699823404,
+      "grad_norm": 0.20922096073627472,
+      "learning_rate": 8.72952002882475e-06,
+      "loss": 0.3746,
       "step": 101840
     },
     {
-      "epoch": 3.58,
-      "learning_rate": 9.825959344544353e-06,
-      "loss": 0.2412,
+      "epoch": 3.670486899484629,
+      "grad_norm": 0.2124052345752716,
+      "learning_rate": 8.7273045970307e-06,
+      "loss": 0.3466,
       "step": 101845
     },
     {
-      "epoch": 3.58,
-      "learning_rate": 9.823695475526245e-06,
-      "loss": 0.2662,
+      "epoch": 3.6706670991458537,
+      "grad_norm": 0.24056388437747955,
+      "learning_rate": 8.725089386948967e-06,
+      "loss": 0.3639,
       "step": 101850
     },
     {
-      "epoch": 3.58,
-      "learning_rate": 9.821431803562465e-06,
-      "loss": 0.2596,
+      "epoch": 3.6708472988070784,
+      "grad_norm": 0.21469198167324066,
+      "learning_rate": 8.722874398609749e-06,
+      "loss": 0.4017,
       "step": 101855
     },
     {
-      "epoch": 3.58,
-      "learning_rate": 9.819168328682377e-06,
-      "loss": 0.2805,
+      "epoch": 3.6710274984683027,
+      "grad_norm": 0.23535849153995514,
+      "learning_rate": 8.720659632043207e-06,
+      "loss": 0.4005,
       "step": 101860
     },
     {
-      "epoch": 3.58,
-      "learning_rate": 9.816905050915392e-06,
-      "loss": 0.2598,
+      "epoch": 3.6712076981295274,
+      "grad_norm": 0.2353014349937439,
+      "learning_rate": 8.718445087279541e-06,
+      "loss": 0.341,
       "step": 101865
     },
     {
-      "epoch": 3.58,
-      "learning_rate": 9.814641970290891e-06,
-      "loss": 0.2627,
+      "epoch": 3.671387897790752,
+      "grad_norm": 0.24978400766849518,
+      "learning_rate": 8.716230764348901e-06,
+      "loss": 0.3621,
       "step": 101870
     },
     {
-      "epoch": 3.58,
-      "learning_rate": 9.812379086838255e-06,
-      "loss": 0.2407,
+      "epoch": 3.671568097451977,
+      "grad_norm": 0.20733198523521423,
+      "learning_rate": 8.714016663281457e-06,
+      "loss": 0.3481,
       "step": 101875
     },
     {
-      "epoch": 3.58,
-      "learning_rate": 9.810116400586861e-06,
-      "loss": 0.2531,
+      "epoch": 3.6717482971132016,
+      "grad_norm": 0.23899643123149872,
+      "learning_rate": 8.71180278410739e-06,
+      "loss": 0.3813,
       "step": 101880
     },
     {
-      "epoch": 3.58,
-      "learning_rate": 9.807853911566103e-06,
-      "loss": 0.2484,
+      "epoch": 3.671928496774426,
+      "grad_norm": 0.27464571595191956,
+      "learning_rate": 8.709589126856857e-06,
+      "loss": 0.4055,
       "step": 101885
     },
     {
-      "epoch": 3.58,
-      "learning_rate": 9.80559161980535e-06,
-      "loss": 0.2519,
+      "epoch": 3.6721086964356506,
+      "grad_norm": 0.27018100023269653,
+      "learning_rate": 8.707375691560018e-06,
+      "loss": 0.3866,
       "step": 101890
     },
     {
-      "epoch": 3.58,
-      "learning_rate": 9.803329525333977e-06,
-      "loss": 0.2567,
+      "epoch": 3.6722888960968754,
+      "grad_norm": 0.22291234135627747,
+      "learning_rate": 8.70516247824703e-06,
+      "loss": 0.3894,
       "step": 101895
     },
     {
-      "epoch": 3.59,
-      "learning_rate": 9.801067628181343e-06,
-      "loss": 0.2459,
+      "epoch": 3.6724690957581,
+      "grad_norm": 0.24809099733829498,
+      "learning_rate": 8.702949486948042e-06,
+      "loss": 0.4137,
       "step": 101900
     },
     {
-      "epoch": 3.59,
-      "learning_rate": 9.798805928376839e-06,
-      "loss": 0.2837,
+      "epoch": 3.6726492954193244,
+      "grad_norm": 0.27931684255599976,
+      "learning_rate": 8.70073671769322e-06,
+      "loss": 0.3755,
       "step": 101905
     },
     {
-      "epoch": 3.59,
-      "learning_rate": 9.796544425949817e-06,
-      "loss": 0.2564,
+      "epoch": 3.672829495080549,
+      "grad_norm": 0.2072281837463379,
+      "learning_rate": 8.698524170512703e-06,
+      "loss": 0.4101,
       "step": 101910
     },
     {
-      "epoch": 3.59,
-      "learning_rate": 9.794283120929643e-06,
-      "loss": 0.2478,
+      "epoch": 3.673009694741774,
+      "grad_norm": 0.20641744136810303,
+      "learning_rate": 8.696311845436641e-06,
+      "loss": 0.3667,
       "step": 101915
     },
     {
-      "epoch": 3.59,
-      "learning_rate": 9.792022013345692e-06,
-      "loss": 0.2534,
+      "epoch": 3.6731898944029986,
+      "grad_norm": 0.28100159764289856,
+      "learning_rate": 8.69409974249517e-06,
+      "loss": 0.3929,
       "step": 101920
     },
     {
-      "epoch": 3.59,
-      "learning_rate": 9.78976110322731e-06,
-      "loss": 0.2457,
+      "epoch": 3.6733700940642233,
+      "grad_norm": 0.2308221310377121,
+      "learning_rate": 8.691887861718437e-06,
+      "loss": 0.3521,
       "step": 101925
     },
     {
-      "epoch": 3.59,
-      "learning_rate": 9.78750039060386e-06,
-      "loss": 0.2703,
+      "epoch": 3.673550293725448,
+      "grad_norm": 0.23538222908973694,
+      "learning_rate": 8.689676203136565e-06,
+      "loss": 0.3795,
       "step": 101930
     },
     {
-      "epoch": 3.59,
-      "learning_rate": 9.785239875504682e-06,
-      "loss": 0.2379,
+      "epoch": 3.6737304933866723,
+      "grad_norm": 0.25849369168281555,
+      "learning_rate": 8.687464766779712e-06,
+      "loss": 0.3653,
       "step": 101935
     },
     {
-      "epoch": 3.59,
-      "learning_rate": 9.782979557959142e-06,
-      "loss": 0.2682,
+      "epoch": 3.673910693047897,
+      "grad_norm": 0.19717663526535034,
+      "learning_rate": 8.685253552677978e-06,
+      "loss": 0.3281,
       "step": 101940
     },
     {
-      "epoch": 3.59,
-      "learning_rate": 9.780719437996588e-06,
-      "loss": 0.2586,
+      "epoch": 3.674090892709122,
+      "grad_norm": 0.21411502361297607,
+      "learning_rate": 8.683042560861512e-06,
+      "loss": 0.367,
       "step": 101945
     },
     {
-      "epoch": 3.59,
-      "learning_rate": 9.778459515646363e-06,
-      "loss": 0.2426,
+      "epoch": 3.674271092370346,
+      "grad_norm": 0.23891699314117432,
+      "learning_rate": 8.680831791360433e-06,
+      "loss": 0.3794,
       "step": 101950
     },
     {
-      "epoch": 3.59,
-      "learning_rate": 9.776199790937798e-06,
-      "loss": 0.2492,
+      "epoch": 3.674451292031571,
+      "grad_norm": 0.2534111440181732,
+      "learning_rate": 8.678621244204863e-06,
+      "loss": 0.4035,
       "step": 101955
     },
     {
-      "epoch": 3.59,
-      "learning_rate": 9.773940263900258e-06,
-      "loss": 0.2677,
+      "epoch": 3.6746314916927956,
+      "grad_norm": 0.21654155850410461,
+      "learning_rate": 8.676410919424917e-06,
+      "loss": 0.3668,
       "step": 101960
     },
     {
-      "epoch": 3.59,
-      "learning_rate": 9.77168093456306e-06,
-      "loss": 0.2639,
+      "epoch": 3.6748116913540203,
+      "grad_norm": 0.2257722020149231,
+      "learning_rate": 8.674200817050712e-06,
+      "loss": 0.3829,
       "step": 101965
     },
     {
-      "epoch": 3.59,
-      "learning_rate": 9.769421802955556e-06,
-      "loss": 0.2521,
+      "epoch": 3.674991891015245,
+      "grad_norm": 0.20194071531295776,
+      "learning_rate": 8.671990937112354e-06,
+      "loss": 0.3617,
       "step": 101970
     },
     {
-      "epoch": 3.59,
-      "learning_rate": 9.767162869107063e-06,
-      "loss": 0.2546,
+      "epoch": 3.6751720906764698,
+      "grad_norm": 0.2552418112754822,
+      "learning_rate": 8.669781279639963e-06,
+      "loss": 0.3767,
       "step": 101975
     },
     {
-      "epoch": 3.59,
-      "learning_rate": 9.764904133046934e-06,
-      "loss": 0.2585,
+      "epoch": 3.675352290337694,
+      "grad_norm": 0.1989719569683075,
+      "learning_rate": 8.667571844663644e-06,
+      "loss": 0.3989,
       "step": 101980
     },
     {
-      "epoch": 3.59,
-      "learning_rate": 9.762645594804481e-06,
-      "loss": 0.2574,
+      "epoch": 3.675532489998919,
+      "grad_norm": 0.2214037925004959,
+      "learning_rate": 8.665362632213494e-06,
+      "loss": 0.3783,
       "step": 101985
     },
     {
-      "epoch": 3.59,
-      "learning_rate": 9.760387254409034e-06,
-      "loss": 0.2536,
+      "epoch": 3.6757126896601435,
+      "grad_norm": 0.23372018337249756,
+      "learning_rate": 8.663153642319616e-06,
+      "loss": 0.3871,
       "step": 101990
     },
     {
-      "epoch": 3.59,
-      "learning_rate": 9.758129111889908e-06,
-      "loss": 0.2695,
+      "epoch": 3.675892889321368,
+      "grad_norm": 0.20797449350357056,
+      "learning_rate": 8.6609448750121e-06,
+      "loss": 0.3736,
       "step": 101995
     },
     {
-      "epoch": 3.59,
-      "learning_rate": 9.75587116727644e-06,
-      "loss": 0.2769,
+      "epoch": 3.6760730889825926,
+      "grad_norm": 0.2792084515094757,
+      "learning_rate": 8.658736330321051e-06,
+      "loss": 0.3913,
       "step": 102000
     },
     {
-      "epoch": 3.59,
-      "eval_loss": 0.25186842679977417,
-      "eval_runtime": 10.5576,
-      "eval_samples_per_second": 9.472,
-      "eval_steps_per_second": 9.472,
+      "epoch": 3.6760730889825926,
+      "eval_loss": 0.4288567304611206,
+      "eval_runtime": 3.5476,
+      "eval_samples_per_second": 28.188,
+      "eval_steps_per_second": 7.047,
       "step": 102000
     },
     {
-      "epoch": 3.59,
-      "learning_rate": 9.753613420597935e-06,
-      "loss": 0.2926,
+      "epoch": 3.6762532886438173,
+      "grad_norm": 0.2272174209356308,
+      "learning_rate": 8.656528008276568e-06,
+      "loss": 0.3661,
       "step": 102005
     },
     {
-      "epoch": 3.59,
-      "learning_rate": 9.751355871883713e-06,
-      "loss": 0.2503,
+      "epoch": 3.676433488305042,
+      "grad_norm": 0.21868175268173218,
+      "learning_rate": 8.654319908908709e-06,
+      "loss": 0.3666,
       "step": 102010
     },
     {
-      "epoch": 3.59,
-      "learning_rate": 9.74909852116308e-06,
-      "loss": 0.2665,
+      "epoch": 3.6766136879662668,
+      "grad_norm": 0.23744815587997437,
+      "learning_rate": 8.652112032247587e-06,
+      "loss": 0.369,
       "step": 102015
     },
     {
-      "epoch": 3.59,
-      "learning_rate": 9.746841368465351e-06,
-      "loss": 0.2326,
+      "epoch": 3.6767938876274915,
+      "grad_norm": 0.26380473375320435,
+      "learning_rate": 8.649904378323262e-06,
+      "loss": 0.3666,
       "step": 102020
     },
     {
-      "epoch": 3.59,
-      "learning_rate": 9.74458441381984e-06,
-      "loss": 0.2563,
+      "epoch": 3.6769740872887158,
+      "grad_norm": 0.2755199372768402,
+      "learning_rate": 8.647696947165834e-06,
+      "loss": 0.3548,
       "step": 102025
     },
     {
-      "epoch": 3.59,
-      "learning_rate": 9.742327657255843e-06,
-      "loss": 0.248,
+      "epoch": 3.6771542869499405,
+      "grad_norm": 0.2914251685142517,
+      "learning_rate": 8.645489738805376e-06,
+      "loss": 0.3747,
       "step": 102030
     },
     {
-      "epoch": 3.59,
-      "learning_rate": 9.740071098802672e-06,
-      "loss": 0.2442,
+      "epoch": 3.6773344866111652,
+      "grad_norm": 0.18029101192951202,
+      "learning_rate": 8.64328275327194e-06,
+      "loss": 0.3884,
       "step": 102035
     },
     {
-      "epoch": 3.59,
-      "learning_rate": 9.73781473848962e-06,
-      "loss": 0.2555,
+      "epoch": 3.6775146862723895,
+      "grad_norm": 0.22522252798080444,
+      "learning_rate": 8.641075990595615e-06,
+      "loss": 0.3904,
       "step": 102040
     },
     {
-      "epoch": 3.59,
-      "learning_rate": 9.735558576345988e-06,
-      "loss": 0.2531,
+      "epoch": 3.6776948859336143,
+      "grad_norm": 0.2504473924636841,
+      "learning_rate": 8.638869450806455e-06,
+      "loss": 0.3892,
       "step": 102045
     },
     {
-      "epoch": 3.59,
-      "learning_rate": 9.733302612401058e-06,
-      "loss": 0.2493,
+      "epoch": 3.677875085594839,
+      "grad_norm": 0.28689128160476685,
+      "learning_rate": 8.63666313393455e-06,
+      "loss": 0.3805,
       "step": 102050
     },
     {
-      "epoch": 3.59,
-      "learning_rate": 9.731046846684142e-06,
-      "loss": 0.2575,
+      "epoch": 3.6780552852560637,
+      "grad_norm": 0.3135242164134979,
+      "learning_rate": 8.634457040009932e-06,
+      "loss": 0.373,
       "step": 102055
     },
     {
-      "epoch": 3.59,
-      "learning_rate": 9.728791279224519e-06,
-      "loss": 0.2549,
+      "epoch": 3.6782354849172885,
+      "grad_norm": 0.2535361051559448,
+      "learning_rate": 8.63225116906266e-06,
+      "loss": 0.3851,
       "step": 102060
     },
     {
-      "epoch": 3.59,
-      "learning_rate": 9.726535910051477e-06,
-      "loss": 0.2429,
+      "epoch": 3.678415684578513,
+      "grad_norm": 0.27366381883621216,
+      "learning_rate": 8.630045521122806e-06,
+      "loss": 0.3767,
       "step": 102065
     },
     {
-      "epoch": 3.59,
-      "learning_rate": 9.724280739194294e-06,
-      "loss": 0.2302,
+      "epoch": 3.6785958842397375,
+      "grad_norm": 0.2666814923286438,
+      "learning_rate": 8.62784009622041e-06,
+      "loss": 0.3867,
       "step": 102070
     },
     {
-      "epoch": 3.59,
-      "learning_rate": 9.722025766682254e-06,
-      "loss": 0.2686,
+      "epoch": 3.6787760839009622,
+      "grad_norm": 0.27532005310058594,
+      "learning_rate": 8.625634894385525e-06,
+      "loss": 0.3659,
       "step": 102075
     },
     {
-      "epoch": 3.59,
-      "learning_rate": 9.719770992544655e-06,
-      "loss": 0.2265,
+      "epoch": 3.678956283562187,
+      "grad_norm": 0.2169187366962433,
+      "learning_rate": 8.623429915648195e-06,
+      "loss": 0.3966,
       "step": 102080
     },
     {
-      "epoch": 3.59,
-      "learning_rate": 9.717516416810756e-06,
-      "loss": 0.2452,
+      "epoch": 3.6791364832234117,
+      "grad_norm": 0.20126007497310638,
+      "learning_rate": 8.62122516003845e-06,
+      "loss": 0.3569,
       "step": 102085
     },
     {
-      "epoch": 3.59,
-      "learning_rate": 9.715262039509829e-06,
-      "loss": 0.2526,
+      "epoch": 3.679316682884636,
+      "grad_norm": 0.22194866836071014,
+      "learning_rate": 8.619020627586355e-06,
+      "loss": 0.3845,
       "step": 102090
     },
     {
-      "epoch": 3.59,
-      "learning_rate": 9.713007860671158e-06,
-      "loss": 0.2746,
+      "epoch": 3.6794968825458607,
+      "grad_norm": 0.2033412605524063,
+      "learning_rate": 8.61681631832193e-06,
+      "loss": 0.3834,
       "step": 102095
     },
     {
-      "epoch": 3.59,
-      "learning_rate": 9.710753880324005e-06,
-      "loss": 0.2545,
+      "epoch": 3.6796770822070854,
+      "grad_norm": 0.23441757261753082,
+      "learning_rate": 8.61461223227521e-06,
+      "loss": 0.4229,
       "step": 102100
     },
     {
-      "epoch": 3.59,
-      "learning_rate": 9.708500098497638e-06,
-      "loss": 0.2662,
+      "epoch": 3.67985728186831,
+      "grad_norm": 0.2877858877182007,
+      "learning_rate": 8.612408369476225e-06,
+      "loss": 0.3905,
       "step": 102105
     },
     {
-      "epoch": 3.59,
-      "learning_rate": 9.706246515221312e-06,
-      "loss": 0.2388,
+      "epoch": 3.680037481529535,
+      "grad_norm": 0.2478102594614029,
+      "learning_rate": 8.610204729955005e-06,
+      "loss": 0.3757,
       "step": 102110
     },
     {
-      "epoch": 3.59,
-      "learning_rate": 9.703993130524302e-06,
-      "loss": 0.2357,
+      "epoch": 3.680217681190759,
+      "grad_norm": 0.2399631142616272,
+      "learning_rate": 8.608001313741562e-06,
+      "loss": 0.3803,
       "step": 102115
     },
     {
-      "epoch": 3.59,
-      "learning_rate": 9.701739944435864e-06,
-      "loss": 0.2314,
+      "epoch": 3.680397880851984,
+      "grad_norm": 0.2504538595676422,
+      "learning_rate": 8.605798120865946e-06,
+      "loss": 0.381,
       "step": 102120
     },
     {
-      "epoch": 3.59,
-      "learning_rate": 9.69948695698524e-06,
-      "loss": 0.249,
+      "epoch": 3.6805780805132087,
+      "grad_norm": 0.21540558338165283,
+      "learning_rate": 8.60359515135814e-06,
+      "loss": 0.4004,
       "step": 102125
     },
     {
-      "epoch": 3.59,
-      "learning_rate": 9.697234168201696e-06,
-      "loss": 0.262,
+      "epoch": 3.6807582801744334,
+      "grad_norm": 0.28474679589271545,
+      "learning_rate": 8.601392405248186e-06,
+      "loss": 0.4054,
       "step": 102130
     },
     {
-      "epoch": 3.59,
-      "learning_rate": 9.69498157811449e-06,
-      "loss": 0.2401,
+      "epoch": 3.6809384798356577,
+      "grad_norm": 0.24228844046592712,
+      "learning_rate": 8.59918988256608e-06,
+      "loss": 0.4117,
       "step": 102135
     },
     {
-      "epoch": 3.59,
-      "learning_rate": 9.69272918675286e-06,
-      "loss": 0.2485,
+      "epoch": 3.6811186794968824,
+      "grad_norm": 0.2137162685394287,
+      "learning_rate": 8.596987583341842e-06,
+      "loss": 0.3891,
       "step": 102140
     },
     {
-      "epoch": 3.59,
-      "learning_rate": 9.690476994146044e-06,
-      "loss": 0.2577,
+      "epoch": 3.681298879158107,
+      "grad_norm": 0.22238872945308685,
+      "learning_rate": 8.594785507605468e-06,
+      "loss": 0.3851,
       "step": 102145
     },
     {
-      "epoch": 3.59,
-      "learning_rate": 9.688225000323306e-06,
-      "loss": 0.2478,
+      "epoch": 3.681479078819332,
+      "grad_norm": 0.21729840338230133,
+      "learning_rate": 8.592583655386969e-06,
+      "loss": 0.3649,
       "step": 102150
     },
     {
-      "epoch": 3.59,
-      "learning_rate": 9.685973205313876e-06,
-      "loss": 0.2574,
+      "epoch": 3.6816592784805566,
+      "grad_norm": 0.205953449010849,
+      "learning_rate": 8.590382026716331e-06,
+      "loss": 0.3353,
       "step": 102155
     },
     {
-      "epoch": 3.59,
-      "learning_rate": 9.683721609146989e-06,
-      "loss": 0.255,
+      "epoch": 3.6818394781417814,
+      "grad_norm": 0.2748815715312958,
+      "learning_rate": 8.58818062162357e-06,
+      "loss": 0.425,
       "step": 102160
     },
     {
-      "epoch": 3.59,
-      "learning_rate": 9.681470211851878e-06,
-      "loss": 0.2443,
+      "epoch": 3.6820196778030057,
+      "grad_norm": 0.29391714930534363,
+      "learning_rate": 8.58597944013867e-06,
+      "loss": 0.4082,
       "step": 102165
     },
     {
-      "epoch": 3.59,
-      "learning_rate": 9.679219013457789e-06,
-      "loss": 0.2414,
+      "epoch": 3.6821998774642304,
+      "grad_norm": 0.24622267484664917,
+      "learning_rate": 8.583778482291621e-06,
+      "loss": 0.3797,
       "step": 102170
     },
     {
-      "epoch": 3.59,
-      "learning_rate": 9.676968013993943e-06,
-      "loss": 0.2342,
+      "epoch": 3.682380077125455,
+      "grad_norm": 0.27090001106262207,
+      "learning_rate": 8.581577748112416e-06,
+      "loss": 0.37,
       "step": 102175
     },
     {
-      "epoch": 3.59,
-      "learning_rate": 9.67471721348956e-06,
-      "loss": 0.2552,
+      "epoch": 3.6825602767866794,
+      "grad_norm": 0.24336892366409302,
+      "learning_rate": 8.579377237631022e-06,
+      "loss": 0.3865,
       "step": 102180
     },
     {
-      "epoch": 3.6,
-      "learning_rate": 9.672466611973882e-06,
-      "loss": 0.2323,
+      "epoch": 3.682740476447904,
+      "grad_norm": 0.18202580511569977,
+      "learning_rate": 8.577176950877444e-06,
+      "loss": 0.3838,
       "step": 102185
     },
     {
-      "epoch": 3.6,
-      "learning_rate": 9.670216209476118e-06,
-      "loss": 0.2672,
+      "epoch": 3.682920676109129,
+      "grad_norm": 0.25919443368911743,
+      "learning_rate": 8.574976887881653e-06,
+      "loss": 0.379,
       "step": 102190
     },
     {
-      "epoch": 3.6,
-      "learning_rate": 9.6679660060255e-06,
-      "loss": 0.259,
+      "epoch": 3.6831008757703536,
+      "grad_norm": 0.2314002364873886,
+      "learning_rate": 8.572777048673619e-06,
+      "loss": 0.3772,
       "step": 102195
     },
     {
-      "epoch": 3.6,
-      "learning_rate": 9.66571600165124e-06,
-      "loss": 0.2511,
+      "epoch": 3.6832810754315783,
+      "grad_norm": 0.22203220427036285,
+      "learning_rate": 8.57057743328332e-06,
+      "loss": 0.3834,
       "step": 102200
     },
     {
-      "epoch": 3.6,
-      "learning_rate": 9.663466196382544e-06,
-      "loss": 0.2588,
+      "epoch": 3.683461275092803,
+      "grad_norm": 0.24144263565540314,
+      "learning_rate": 8.568378041740712e-06,
+      "loss": 0.3648,
       "step": 102205
     },
     {
-      "epoch": 3.6,
-      "learning_rate": 9.66121659024864e-06,
-      "loss": 0.2726,
+      "epoch": 3.6836414747540274,
+      "grad_norm": 0.22700710594654083,
+      "learning_rate": 8.566178874075781e-06,
+      "loss": 0.4007,
       "step": 102210
     },
     {
-      "epoch": 3.6,
-      "learning_rate": 9.658967183278731e-06,
-      "loss": 0.2671,
+      "epoch": 3.683821674415252,
+      "grad_norm": 0.22427678108215332,
+      "learning_rate": 8.563979930318489e-06,
+      "loss": 0.3775,
       "step": 102215
     },
     {
-      "epoch": 3.6,
-      "learning_rate": 9.656717975502023e-06,
-      "loss": 0.2524,
+      "epoch": 3.684001874076477,
+      "grad_norm": 0.2833479344844818,
+      "learning_rate": 8.561781210498773e-06,
+      "loss": 0.3957,
       "step": 102220
     },
     {
-      "epoch": 3.6,
-      "learning_rate": 9.654468966947711e-06,
-      "loss": 0.2584,
+      "epoch": 3.684182073737701,
+      "grad_norm": 0.25942087173461914,
+      "learning_rate": 8.559582714646613e-06,
+      "loss": 0.3475,
       "step": 102225
     },
     {
-      "epoch": 3.6,
-      "learning_rate": 9.652220157645014e-06,
-      "loss": 0.2484,
+      "epoch": 3.684362273398926,
+      "grad_norm": 0.2398940771818161,
+      "learning_rate": 8.557384442791947e-06,
+      "loss": 0.3867,
       "step": 102230
     },
     {
-      "epoch": 3.6,
-      "learning_rate": 9.649971547623119e-06,
-      "loss": 0.2767,
+      "epoch": 3.6845424730601506,
+      "grad_norm": 0.25691062211990356,
+      "learning_rate": 8.55518639496475e-06,
+      "loss": 0.4184,
       "step": 102235
     },
     {
-      "epoch": 3.6,
-      "learning_rate": 9.647723136911233e-06,
-      "loss": 0.2603,
+      "epoch": 3.6847226727213753,
+      "grad_norm": 0.3172163665294647,
+      "learning_rate": 8.552988571194948e-06,
+      "loss": 0.4036,
       "step": 102240
     },
     {
-      "epoch": 3.6,
-      "learning_rate": 9.645474925538536e-06,
-      "loss": 0.2558,
+      "epoch": 3.6849028723826,
+      "grad_norm": 0.2886956036090851,
+      "learning_rate": 8.550790971512484e-06,
+      "loss": 0.3484,
       "step": 102245
     },
     {
-      "epoch": 3.6,
-      "learning_rate": 9.643226913534236e-06,
-      "loss": 0.2626,
+      "epoch": 3.685083072043825,
+      "grad_norm": 0.26184308528900146,
+      "learning_rate": 8.548593595947315e-06,
+      "loss": 0.3872,
       "step": 102250
     },
     {
-      "epoch": 3.6,
-      "learning_rate": 9.640979100927517e-06,
-      "loss": 0.267,
+      "epoch": 3.685263271705049,
+      "grad_norm": 0.21521615982055664,
+      "learning_rate": 8.546396444529375e-06,
+      "loss": 0.3831,
       "step": 102255
     },
     {
-      "epoch": 3.6,
-      "learning_rate": 9.63873148774755e-06,
-      "loss": 0.2714,
+      "epoch": 3.685443471366274,
+      "grad_norm": 0.2815273404121399,
+      "learning_rate": 8.544199517288599e-06,
+      "loss": 0.3915,
       "step": 102260
     },
     {
-      "epoch": 3.6,
-      "learning_rate": 9.636484074023542e-06,
-      "loss": 0.2546,
+      "epoch": 3.6856236710274985,
+      "grad_norm": 0.24685360491275787,
+      "learning_rate": 8.542002814254918e-06,
+      "loss": 0.3755,
       "step": 102265
     },
     {
-      "epoch": 3.6,
-      "learning_rate": 9.634236859784663e-06,
-      "loss": 0.2572,
+      "epoch": 3.685803870688723,
+      "grad_norm": 0.21425403654575348,
+      "learning_rate": 8.539806335458253e-06,
+      "loss": 0.3594,
       "step": 102270
     },
     {
-      "epoch": 3.6,
-      "learning_rate": 9.631989845060094e-06,
-      "loss": 0.2706,
+      "epoch": 3.6859840703499476,
+      "grad_norm": 0.20560602843761444,
+      "learning_rate": 8.537610080928548e-06,
+      "loss": 0.3628,
       "step": 102275
     },
     {
-      "epoch": 3.6,
-      "learning_rate": 9.629743029878995e-06,
-      "loss": 0.2668,
+      "epoch": 3.6861642700111723,
+      "grad_norm": 0.2760855555534363,
+      "learning_rate": 8.53541405069572e-06,
+      "loss": 0.4007,
       "step": 102280
     },
     {
-      "epoch": 3.6,
-      "learning_rate": 9.627496414270568e-06,
-      "loss": 0.2564,
+      "epoch": 3.686344469672397,
+      "grad_norm": 0.24321669340133667,
+      "learning_rate": 8.533218244789683e-06,
+      "loss": 0.3586,
       "step": 102285
     },
     {
-      "epoch": 3.6,
-      "learning_rate": 9.625249998263958e-06,
-      "loss": 0.245,
+      "epoch": 3.6865246693336218,
+      "grad_norm": 0.23994041979312897,
+      "learning_rate": 8.531022663240366e-06,
+      "loss": 0.3742,
       "step": 102290
     },
     {
-      "epoch": 3.6,
-      "learning_rate": 9.62300378188835e-06,
-      "loss": 0.2856,
+      "epoch": 3.6867048689948465,
+      "grad_norm": 0.23098506033420563,
+      "learning_rate": 8.528827306077672e-06,
+      "loss": 0.3835,
       "step": 102295
     },
     {
-      "epoch": 3.6,
-      "learning_rate": 9.6207577651729e-06,
-      "loss": 0.2775,
+      "epoch": 3.686885068656071,
+      "grad_norm": 0.21299470961093903,
+      "learning_rate": 8.526632173331511e-06,
+      "loss": 0.3864,
       "step": 102300
     },
     {
-      "epoch": 3.6,
-      "learning_rate": 9.618511948146783e-06,
-      "loss": 0.281,
+      "epoch": 3.6870652683172955,
+      "grad_norm": 0.20803943276405334,
+      "learning_rate": 8.524437265031815e-06,
+      "loss": 0.4014,
       "step": 102305
     },
     {
-      "epoch": 3.6,
-      "learning_rate": 9.61626633083915e-06,
-      "loss": 0.2514,
+      "epoch": 3.6872454679785203,
+      "grad_norm": 0.26612281799316406,
+      "learning_rate": 8.522242581208451e-06,
+      "loss": 0.3805,
       "step": 102310
     },
     {
-      "epoch": 3.6,
-      "learning_rate": 9.614020913279161e-06,
-      "loss": 0.2598,
+      "epoch": 3.687425667639745,
+      "grad_norm": 0.26680296659469604,
+      "learning_rate": 8.520048121891352e-06,
+      "loss": 0.3989,
       "step": 102315
     },
     {
-      "epoch": 3.6,
-      "learning_rate": 9.611775695495961e-06,
-      "loss": 0.2906,
+      "epoch": 3.6876058673009693,
+      "grad_norm": 0.26344892382621765,
+      "learning_rate": 8.517853887110408e-06,
+      "loss": 0.3972,
       "step": 102320
     },
     {
-      "epoch": 3.6,
-      "learning_rate": 9.609530677518724e-06,
-      "loss": 0.2555,
+      "epoch": 3.687786066962194,
+      "grad_norm": 0.2039593607187271,
+      "learning_rate": 8.515659876895504e-06,
+      "loss": 0.3739,
       "step": 102325
     },
     {
-      "epoch": 3.6,
-      "learning_rate": 9.607285859376587e-06,
-      "loss": 0.2633,
+      "epoch": 3.6879662666234188,
+      "grad_norm": 0.24437233805656433,
+      "learning_rate": 8.51346609127656e-06,
+      "loss": 0.3616,
       "step": 102330
     },
     {
-      "epoch": 3.6,
-      "learning_rate": 9.605041241098697e-06,
-      "loss": 0.2663,
+      "epoch": 3.6881464662846435,
+      "grad_norm": 0.23033076524734497,
+      "learning_rate": 8.51127253028344e-06,
+      "loss": 0.4245,
       "step": 102335
     },
     {
-      "epoch": 3.6,
-      "learning_rate": 9.602796822714194e-06,
-      "loss": 0.2656,
+      "epoch": 3.688326665945868,
+      "grad_norm": 0.2611679136753082,
+      "learning_rate": 8.509079193946032e-06,
+      "loss": 0.3759,
       "step": 102340
     },
     {
-      "epoch": 3.6,
-      "learning_rate": 9.600552604252237e-06,
-      "loss": 0.257,
+      "epoch": 3.6885068656070925,
+      "grad_norm": 0.19574002921581268,
+      "learning_rate": 8.506886082294233e-06,
+      "loss": 0.3695,
       "step": 102345
     },
     {
-      "epoch": 3.6,
-      "learning_rate": 9.598308585741944e-06,
-      "loss": 0.2543,
+      "epoch": 3.6886870652683172,
+      "grad_norm": 0.2866314947605133,
+      "learning_rate": 8.504693195357921e-06,
+      "loss": 0.3685,
       "step": 102350
     },
     {
-      "epoch": 3.6,
-      "learning_rate": 9.596064767212477e-06,
-      "loss": 0.2677,
+      "epoch": 3.688867264929542,
+      "grad_norm": 0.23495450615882874,
+      "learning_rate": 8.502500533166968e-06,
+      "loss": 0.4131,
       "step": 102355
     },
     {
-      "epoch": 3.6,
-      "learning_rate": 9.593821148692945e-06,
-      "loss": 0.2726,
+      "epoch": 3.6890474645907667,
+      "grad_norm": 0.2283242791891098,
+      "learning_rate": 8.500308095751253e-06,
+      "loss": 0.3596,
       "step": 102360
     },
     {
-      "epoch": 3.6,
-      "learning_rate": 9.591577730212503e-06,
-      "loss": 0.2637,
+      "epoch": 3.689227664251991,
+      "grad_norm": 0.3276098370552063,
+      "learning_rate": 8.498115883140637e-06,
+      "loss": 0.3671,
       "step": 102365
     },
     {
-      "epoch": 3.6,
-      "learning_rate": 9.58933451180027e-06,
-      "loss": 0.2391,
+      "epoch": 3.6894078639132157,
+      "grad_norm": 0.2699391543865204,
+      "learning_rate": 8.495923895365004e-06,
+      "loss": 0.3776,
       "step": 102370
     },
     {
-      "epoch": 3.6,
-      "learning_rate": 9.58709149348536e-06,
-      "loss": 0.2445,
+      "epoch": 3.6895880635744405,
+      "grad_norm": 0.2524622082710266,
+      "learning_rate": 8.493732132454215e-06,
+      "loss": 0.3558,
       "step": 102375
     },
     {
-      "epoch": 3.6,
-      "learning_rate": 9.58484867529692e-06,
-      "loss": 0.2422,
+      "epoch": 3.689768263235665,
+      "grad_norm": 0.26742005348205566,
+      "learning_rate": 8.49154059443813e-06,
+      "loss": 0.3457,
       "step": 102380
     },
     {
-      "epoch": 3.6,
-      "learning_rate": 9.582606057264062e-06,
-      "loss": 0.2746,
+      "epoch": 3.68994846289689,
+      "grad_norm": 0.2498975545167923,
+      "learning_rate": 8.489349281346606e-06,
+      "loss": 0.3697,
       "step": 102385
     },
     {
-      "epoch": 3.6,
-      "learning_rate": 9.580363639415905e-06,
-      "loss": 0.2527,
+      "epoch": 3.6901286625581142,
+      "grad_norm": 0.20344579219818115,
+      "learning_rate": 8.487158193209497e-06,
+      "loss": 0.3884,
       "step": 102390
     },
     {
-      "epoch": 3.6,
-      "learning_rate": 9.578121421781552e-06,
-      "loss": 0.2454,
+      "epoch": 3.690308862219339,
+      "grad_norm": 0.2646947205066681,
+      "learning_rate": 8.484967330056665e-06,
+      "loss": 0.3891,
       "step": 102395
     },
     {
-      "epoch": 3.6,
-      "learning_rate": 9.57587940439014e-06,
-      "loss": 0.2562,
+      "epoch": 3.6904890618805637,
+      "grad_norm": 0.23882755637168884,
+      "learning_rate": 8.482776691917966e-06,
+      "loss": 0.3553,
       "step": 102400
     },
     {
-      "epoch": 3.6,
-      "learning_rate": 9.573637587270759e-06,
-      "loss": 0.2745,
+      "epoch": 3.6906692615417884,
+      "grad_norm": 0.20374254882335663,
+      "learning_rate": 8.480586278823219e-06,
+      "loss": 0.3649,
       "step": 102405
     },
     {
-      "epoch": 3.6,
-      "learning_rate": 9.571395970452537e-06,
-      "loss": 0.2321,
+      "epoch": 3.6908494612030127,
+      "grad_norm": 0.18348126113414764,
+      "learning_rate": 8.478396090802294e-06,
+      "loss": 0.338,
       "step": 102410
     },
     {
-      "epoch": 3.6,
-      "learning_rate": 9.569154553964563e-06,
-      "loss": 0.2629,
+      "epoch": 3.6910296608642374,
+      "grad_norm": 0.22053396701812744,
+      "learning_rate": 8.476206127885026e-06,
+      "loss": 0.3812,
       "step": 102415
     },
     {
-      "epoch": 3.6,
-      "learning_rate": 9.566913337835956e-06,
-      "loss": 0.2616,
+      "epoch": 3.691209860525462,
+      "grad_norm": 0.19936566054821014,
+      "learning_rate": 8.474016390101247e-06,
+      "loss": 0.3908,
       "step": 102420
     },
     {
-      "epoch": 3.6,
-      "learning_rate": 9.564672322095808e-06,
-      "loss": 0.2612,
+      "epoch": 3.691390060186687,
+      "grad_norm": 0.20026947557926178,
+      "learning_rate": 8.471826877480795e-06,
+      "loss": 0.3673,
       "step": 102425
     },
     {
-      "epoch": 3.6,
-      "learning_rate": 9.562431506773217e-06,
-      "loss": 0.2632,
+      "epoch": 3.6915702598479117,
+      "grad_norm": 0.24228374660015106,
+      "learning_rate": 8.469637590053497e-06,
+      "loss": 0.3657,
       "step": 102430
     },
     {
-      "epoch": 3.6,
-      "learning_rate": 9.560190891897272e-06,
-      "loss": 0.2614,
+      "epoch": 3.6917504595091364,
+      "grad_norm": 0.2570672035217285,
+      "learning_rate": 8.467448527849192e-06,
+      "loss": 0.344,
       "step": 102435
     },
     {
-      "epoch": 3.6,
-      "learning_rate": 9.55795047749708e-06,
-      "loss": 0.244,
+      "epoch": 3.6919306591703607,
+      "grad_norm": 0.255535364151001,
+      "learning_rate": 8.4652596908977e-06,
+      "loss": 0.377,
       "step": 102440
     },
     {
-      "epoch": 3.6,
-      "learning_rate": 9.555710263601728e-06,
-      "loss": 0.2503,
+      "epoch": 3.6921108588315854,
+      "grad_norm": 0.2862887680530548,
+      "learning_rate": 8.463071079228846e-06,
+      "loss": 0.3874,
       "step": 102445
     },
     {
-      "epoch": 3.6,
-      "learning_rate": 9.553470250240296e-06,
-      "loss": 0.2715,
+      "epoch": 3.69229105849281,
+      "grad_norm": 0.2560741901397705,
+      "learning_rate": 8.460882692872446e-06,
+      "loss": 0.3996,
       "step": 102450
     },
     {
-      "epoch": 3.6,
-      "learning_rate": 9.551230437441866e-06,
-      "loss": 0.2537,
+      "epoch": 3.6924712581540344,
+      "grad_norm": 0.1878066509962082,
+      "learning_rate": 8.458694531858308e-06,
+      "loss": 0.3576,
       "step": 102455
     },
     {
-      "epoch": 3.6,
-      "learning_rate": 9.548990825235527e-06,
-      "loss": 0.2516,
+      "epoch": 3.692651457815259,
+      "grad_norm": 0.2039957493543625,
+      "learning_rate": 8.456506596216262e-06,
+      "loss": 0.3603,
       "step": 102460
     },
     {
-      "epoch": 3.61,
-      "learning_rate": 9.546751413650368e-06,
-      "loss": 0.25,
+      "epoch": 3.692831657476484,
+      "grad_norm": 0.2823259234428406,
+      "learning_rate": 8.454318885976113e-06,
+      "loss": 0.3855,
       "step": 102465
     },
     {
-      "epoch": 3.61,
-      "learning_rate": 9.544512202715458e-06,
-      "loss": 0.2575,
+      "epoch": 3.6930118571377086,
+      "grad_norm": 0.20816364884376526,
+      "learning_rate": 8.452131401167665e-06,
+      "loss": 0.3641,
       "step": 102470
     },
     {
-      "epoch": 3.61,
-      "learning_rate": 9.542273192459863e-06,
-      "loss": 0.2565,
+      "epoch": 3.6931920567989334,
+      "grad_norm": 0.248849555850029,
+      "learning_rate": 8.449944141820723e-06,
+      "loss": 0.3656,
       "step": 102475
     },
     {
-      "epoch": 3.61,
-      "learning_rate": 9.540034382912673e-06,
-      "loss": 0.2556,
+      "epoch": 3.693372256460158,
+      "grad_norm": 0.21117424964904785,
+      "learning_rate": 8.447757107965088e-06,
+      "loss": 0.3622,
       "step": 102480
     },
     {
-      "epoch": 3.61,
-      "learning_rate": 9.537795774102948e-06,
-      "loss": 0.2792,
+      "epoch": 3.6935524561213824,
+      "grad_norm": 0.21693001687526703,
+      "learning_rate": 8.445570299630548e-06,
+      "loss": 0.3607,
       "step": 102485
     },
     {
-      "epoch": 3.61,
-      "learning_rate": 9.535557366059747e-06,
-      "loss": 0.236,
+      "epoch": 3.693732655782607,
+      "grad_norm": 0.24611179530620575,
+      "learning_rate": 8.443383716846917e-06,
+      "loss": 0.3802,
       "step": 102490
     },
     {
-      "epoch": 3.61,
-      "learning_rate": 9.533319158812148e-06,
-      "loss": 0.2493,
+      "epoch": 3.693912855443832,
+      "grad_norm": 0.199144646525383,
+      "learning_rate": 8.441197359643977e-06,
+      "loss": 0.3146,
       "step": 102495
     },
     {
-      "epoch": 3.61,
-      "learning_rate": 9.53108115238921e-06,
-      "loss": 0.2487,
+      "epoch": 3.694093055105056,
+      "grad_norm": 0.2069302648305893,
+      "learning_rate": 8.439011228051515e-06,
+      "loss": 0.3344,
       "step": 102500
     },
     {
-      "epoch": 3.61,
-      "eval_loss": 0.2519163489341736,
-      "eval_runtime": 10.5511,
-      "eval_samples_per_second": 9.478,
-      "eval_steps_per_second": 9.478,
+      "epoch": 3.694093055105056,
+      "eval_loss": 0.4291780889034271,
+      "eval_runtime": 3.5329,
+      "eval_samples_per_second": 28.305,
+      "eval_steps_per_second": 7.076,
       "step": 102500
     },
     {
-      "epoch": 3.61,
-      "learning_rate": 9.528843346819989e-06,
-      "loss": 0.2756,
+      "epoch": 3.694273254766281,
+      "grad_norm": 0.22869972884655,
+      "learning_rate": 8.436825322099324e-06,
+      "loss": 0.402,
       "step": 102505
     },
     {
-      "epoch": 3.61,
-      "learning_rate": 9.526605742133529e-06,
-      "loss": 0.2386,
+      "epoch": 3.6944534544275056,
+      "grad_norm": 0.21545152366161346,
+      "learning_rate": 8.434639641817168e-06,
+      "loss": 0.3588,
       "step": 102510
     },
     {
-      "epoch": 3.61,
-      "learning_rate": 9.524368338358897e-06,
-      "loss": 0.2588,
+      "epoch": 3.6946336540887303,
+      "grad_norm": 0.20983858406543732,
+      "learning_rate": 8.43245418723486e-06,
+      "loss": 0.3756,
       "step": 102515
     },
     {
-      "epoch": 3.61,
-      "learning_rate": 9.522131135525153e-06,
-      "loss": 0.2416,
+      "epoch": 3.694813853749955,
+      "grad_norm": 0.23540320992469788,
+      "learning_rate": 8.430268958382146e-06,
+      "loss": 0.3386,
       "step": 102520
     },
     {
-      "epoch": 3.61,
-      "learning_rate": 9.519894133661337e-06,
-      "loss": 0.2571,
+      "epoch": 3.69499405341118,
+      "grad_norm": 0.2099265158176422,
+      "learning_rate": 8.428083955288801e-06,
+      "loss": 0.3688,
       "step": 102525
     },
     {
-      "epoch": 3.61,
-      "learning_rate": 9.517657332796485e-06,
-      "loss": 0.2645,
+      "epoch": 3.695174253072404,
+      "grad_norm": 0.2085736095905304,
+      "learning_rate": 8.425899177984611e-06,
+      "loss": 0.3659,
       "step": 102530
     },
     {
-      "epoch": 3.61,
-      "learning_rate": 9.515420732959654e-06,
-      "loss": 0.2454,
+      "epoch": 3.695354452733629,
+      "grad_norm": 0.23389971256256104,
+      "learning_rate": 8.423714626499338e-06,
+      "loss": 0.355,
       "step": 102535
     },
     {
-      "epoch": 3.61,
-      "learning_rate": 9.513184334179883e-06,
-      "loss": 0.2603,
+      "epoch": 3.6955346523948536,
+      "grad_norm": 0.23065683245658875,
+      "learning_rate": 8.421530300862743e-06,
+      "loss": 0.3645,
       "step": 102540
     },
     {
-      "epoch": 3.61,
-      "learning_rate": 9.510948136486209e-06,
-      "loss": 0.2355,
+      "epoch": 3.695714852056078,
+      "grad_norm": 0.21496419608592987,
+      "learning_rate": 8.419346201104588e-06,
+      "loss": 0.3849,
       "step": 102545
     },
     {
-      "epoch": 3.61,
-      "learning_rate": 9.508712139907654e-06,
-      "loss": 0.2529,
+      "epoch": 3.6958950517173026,
+      "grad_norm": 0.22192232310771942,
+      "learning_rate": 8.41716232725462e-06,
+      "loss": 0.3783,
       "step": 102550
     },
     {
-      "epoch": 3.61,
-      "learning_rate": 9.506476344473272e-06,
-      "loss": 0.2456,
+      "epoch": 3.6960752513785273,
+      "grad_norm": 0.21217724680900574,
+      "learning_rate": 8.414978679342617e-06,
+      "loss": 0.3721,
       "step": 102555
     },
     {
-      "epoch": 3.61,
-      "learning_rate": 9.504240750212088e-06,
-      "loss": 0.2642,
+      "epoch": 3.696255451039752,
+      "grad_norm": 0.27816903591156006,
+      "learning_rate": 8.412795257398318e-06,
+      "loss": 0.3908,
       "step": 102560
     },
     {
-      "epoch": 3.61,
-      "learning_rate": 9.502005357153113e-06,
-      "loss": 0.2516,
+      "epoch": 3.696435650700977,
+      "grad_norm": 0.25314852595329285,
+      "learning_rate": 8.410612061451473e-06,
+      "loss": 0.3816,
       "step": 102565
     },
     {
-      "epoch": 3.61,
-      "learning_rate": 9.499770165325386e-06,
-      "loss": 0.2802,
+      "epoch": 3.6966158503622015,
+      "grad_norm": 0.25173550844192505,
+      "learning_rate": 8.408429091531825e-06,
+      "loss": 0.3694,
       "step": 102570
     },
     {
-      "epoch": 3.61,
-      "learning_rate": 9.497535174757934e-06,
-      "loss": 0.2455,
+      "epoch": 3.696796050023426,
+      "grad_norm": 0.2536860704421997,
+      "learning_rate": 8.406246347669108e-06,
+      "loss": 0.3842,
       "step": 102575
     },
     {
-      "epoch": 3.61,
-      "learning_rate": 9.495300385479777e-06,
-      "loss": 0.2484,
+      "epoch": 3.6969762496846506,
+      "grad_norm": 0.22247686982154846,
+      "learning_rate": 8.404063829893083e-06,
+      "loss": 0.4039,
       "step": 102580
     },
     {
-      "epoch": 3.61,
-      "learning_rate": 9.493065797519923e-06,
-      "loss": 0.2567,
+      "epoch": 3.6971564493458753,
+      "grad_norm": 0.20821170508861542,
+      "learning_rate": 8.401881538233483e-06,
+      "loss": 0.3945,
       "step": 102585
     },
     {
-      "epoch": 3.61,
-      "learning_rate": 9.490831410907382e-06,
-      "loss": 0.2646,
+      "epoch": 3.6973366490071,
+      "grad_norm": 0.252389132976532,
+      "learning_rate": 8.399699472720019e-06,
+      "loss": 0.3411,
       "step": 102590
     },
     {
-      "epoch": 3.61,
-      "learning_rate": 9.488597225671184e-06,
-      "loss": 0.2687,
+      "epoch": 3.6975168486683243,
+      "grad_norm": 0.2539133131504059,
+      "learning_rate": 8.397517633382441e-06,
+      "loss": 0.3972,
       "step": 102595
     },
     {
-      "epoch": 3.61,
-      "learning_rate": 9.486363241840326e-06,
-      "loss": 0.2501,
+      "epoch": 3.697697048329549,
+      "grad_norm": 0.17788057029247284,
+      "learning_rate": 8.395336020250472e-06,
+      "loss": 0.3556,
       "step": 102600
     },
     {
-      "epoch": 3.61,
-      "learning_rate": 9.48412945944381e-06,
-      "loss": 0.2685,
+      "epoch": 3.6978772479907738,
+      "grad_norm": 0.20301634073257446,
+      "learning_rate": 8.393154633353825e-06,
+      "loss": 0.3843,
       "step": 102605
     },
     {
-      "epoch": 3.61,
-      "learning_rate": 9.481895878510658e-06,
-      "loss": 0.2516,
+      "epoch": 3.6980574476519985,
+      "grad_norm": 0.2251247763633728,
+      "learning_rate": 8.39097347272225e-06,
+      "loss": 0.3663,
       "step": 102610
     },
     {
-      "epoch": 3.61,
-      "learning_rate": 9.47966249906986e-06,
-      "loss": 0.264,
+      "epoch": 3.6982376473132232,
+      "grad_norm": 0.285604327917099,
+      "learning_rate": 8.388792538385429e-06,
+      "loss": 0.3637,
       "step": 102615
     },
     {
-      "epoch": 3.61,
-      "learning_rate": 9.477429321150409e-06,
-      "loss": 0.248,
+      "epoch": 3.6984178469744475,
+      "grad_norm": 0.2661183178424835,
+      "learning_rate": 8.3866118303731e-06,
+      "loss": 0.4143,
       "step": 102620
     },
     {
-      "epoch": 3.61,
-      "learning_rate": 9.475196344781315e-06,
-      "loss": 0.2655,
+      "epoch": 3.6985980466356723,
+      "grad_norm": 0.2364463359117508,
+      "learning_rate": 8.384431348714972e-06,
+      "loss": 0.3785,
       "step": 102625
     },
     {
-      "epoch": 3.61,
-      "learning_rate": 9.472963569991555e-06,
-      "loss": 0.2697,
+      "epoch": 3.698778246296897,
+      "grad_norm": 0.21531599760055542,
+      "learning_rate": 8.382251093440747e-06,
+      "loss": 0.3636,
       "step": 102630
     },
     {
-      "epoch": 3.61,
-      "learning_rate": 9.470730996810142e-06,
-      "loss": 0.2444,
+      "epoch": 3.6989584459581217,
+      "grad_norm": 0.23172800242900848,
+      "learning_rate": 8.380071064580133e-06,
+      "loss": 0.3799,
       "step": 102635
     },
     {
-      "epoch": 3.61,
-      "learning_rate": 9.468498625266051e-06,
-      "loss": 0.2492,
+      "epoch": 3.699138645619346,
+      "grad_norm": 0.2571316063404083,
+      "learning_rate": 8.377891262162827e-06,
+      "loss": 0.3758,
       "step": 102640
     },
     {
-      "epoch": 3.61,
-      "learning_rate": 9.466266455388262e-06,
-      "loss": 0.2685,
+      "epoch": 3.6993188452805708,
+      "grad_norm": 0.22325757145881653,
+      "learning_rate": 8.37571168621854e-06,
+      "loss": 0.3447,
       "step": 102645
     },
     {
-      "epoch": 3.61,
-      "learning_rate": 9.464034487205775e-06,
-      "loss": 0.2759,
+      "epoch": 3.6994990449417955,
+      "grad_norm": 0.2073674499988556,
+      "learning_rate": 8.373532336776965e-06,
+      "loss": 0.4042,
       "step": 102650
     },
     {
-      "epoch": 3.61,
-      "learning_rate": 9.461802720747562e-06,
-      "loss": 0.2395,
+      "epoch": 3.6996792446030202,
+      "grad_norm": 0.22034046053886414,
+      "learning_rate": 8.371353213867792e-06,
+      "loss": 0.386,
       "step": 102655
     },
     {
-      "epoch": 3.61,
-      "learning_rate": 9.459571156042599e-06,
-      "loss": 0.2341,
+      "epoch": 3.699859444264245,
+      "grad_norm": 0.23101651668548584,
+      "learning_rate": 8.369174317520714e-06,
+      "loss": 0.4023,
       "step": 102660
     },
     {
-      "epoch": 3.61,
-      "learning_rate": 9.457339793119858e-06,
-      "loss": 0.2506,
+      "epoch": 3.7000396439254697,
+      "grad_norm": 0.2687579095363617,
+      "learning_rate": 8.366995647765413e-06,
+      "loss": 0.3909,
       "step": 102665
     },
     {
-      "epoch": 3.61,
-      "learning_rate": 9.45510863200832e-06,
-      "loss": 0.2364,
+      "epoch": 3.700219843586694,
+      "grad_norm": 0.24219973385334015,
+      "learning_rate": 8.364817204631569e-06,
+      "loss": 0.3828,
       "step": 102670
     },
     {
-      "epoch": 3.61,
-      "learning_rate": 9.452877672736946e-06,
-      "loss": 0.2614,
+      "epoch": 3.7004000432479187,
+      "grad_norm": 0.2693396508693695,
+      "learning_rate": 8.36263898814888e-06,
+      "loss": 0.3629,
       "step": 102675
     },
     {
-      "epoch": 3.61,
-      "learning_rate": 9.450646915334718e-06,
-      "loss": 0.2472,
+      "epoch": 3.7005802429091434,
+      "grad_norm": 0.20992809534072876,
+      "learning_rate": 8.36046099834701e-06,
+      "loss": 0.3739,
       "step": 102680
     },
     {
-      "epoch": 3.61,
-      "learning_rate": 9.448416359830583e-06,
-      "loss": 0.2616,
+      "epoch": 3.7007604425703677,
+      "grad_norm": 0.22777998447418213,
+      "learning_rate": 8.35828323525564e-06,
+      "loss": 0.3974,
       "step": 102685
     },
     {
-      "epoch": 3.61,
-      "learning_rate": 9.446186006253523e-06,
-      "loss": 0.258,
+      "epoch": 3.7009406422315925,
+      "grad_norm": 0.25267812609672546,
+      "learning_rate": 8.35610569890444e-06,
+      "loss": 0.3982,
       "step": 102690
     },
     {
-      "epoch": 3.61,
-      "learning_rate": 9.443955854632488e-06,
-      "loss": 0.2586,
+      "epoch": 3.701120841892817,
+      "grad_norm": 0.24532164633274078,
+      "learning_rate": 8.353928389323064e-06,
+      "loss": 0.3529,
       "step": 102695
     },
     {
-      "epoch": 3.61,
-      "learning_rate": 9.441725904996431e-06,
-      "loss": 0.2646,
+      "epoch": 3.701301041554042,
+      "grad_norm": 0.23484160006046295,
+      "learning_rate": 8.351751306541215e-06,
+      "loss": 0.3878,
       "step": 102700
     },
     {
-      "epoch": 3.61,
-      "learning_rate": 9.439496157374303e-06,
-      "loss": 0.2476,
+      "epoch": 3.7014812412152667,
+      "grad_norm": 0.2185060977935791,
+      "learning_rate": 8.349574450588518e-06,
+      "loss": 0.3754,
       "step": 102705
     },
     {
-      "epoch": 3.61,
-      "learning_rate": 9.437266611795073e-06,
-      "loss": 0.2542,
+      "epoch": 3.7016614408764914,
+      "grad_norm": 0.28905272483825684,
+      "learning_rate": 8.347397821494637e-06,
+      "loss": 0.3616,
       "step": 102710
     },
     {
-      "epoch": 3.61,
-      "learning_rate": 9.435037268287677e-06,
-      "loss": 0.2702,
+      "epoch": 3.7018416405377157,
+      "grad_norm": 0.25226137042045593,
+      "learning_rate": 8.345221419289247e-06,
+      "loss": 0.387,
       "step": 102715
     },
     {
-      "epoch": 3.61,
-      "learning_rate": 9.432808126881057e-06,
-      "loss": 0.2632,
+      "epoch": 3.7020218401989404,
+      "grad_norm": 0.22827744483947754,
+      "learning_rate": 8.343045244001982e-06,
+      "loss": 0.3774,
       "step": 102720
     },
     {
-      "epoch": 3.61,
-      "learning_rate": 9.430579187604174e-06,
-      "loss": 0.2578,
+      "epoch": 3.702202039860165,
+      "grad_norm": 0.2841804623603821,
+      "learning_rate": 8.340869295662517e-06,
+      "loss": 0.4008,
       "step": 102725
     },
     {
-      "epoch": 3.61,
-      "learning_rate": 9.428350450485951e-06,
-      "loss": 0.251,
+      "epoch": 3.7023822395213895,
+      "grad_norm": 0.2542993724346161,
+      "learning_rate": 8.338693574300474e-06,
+      "loss": 0.3553,
       "step": 102730
     },
     {
-      "epoch": 3.61,
-      "learning_rate": 9.426121915555341e-06,
-      "loss": 0.2416,
+      "epoch": 3.702562439182614,
+      "grad_norm": 0.21926385164260864,
+      "learning_rate": 8.336518079945497e-06,
+      "loss": 0.4037,
       "step": 102735
     },
     {
-      "epoch": 3.61,
-      "learning_rate": 9.423893582841268e-06,
-      "loss": 0.2561,
+      "epoch": 3.702742638843839,
+      "grad_norm": 0.2608276903629303,
+      "learning_rate": 8.334342812627244e-06,
+      "loss": 0.3645,
       "step": 102740
     },
     {
-      "epoch": 3.61,
-      "learning_rate": 9.42166545237268e-06,
-      "loss": 0.2564,
+      "epoch": 3.7029228385050637,
+      "grad_norm": 0.2661699950695038,
+      "learning_rate": 8.332167772375344e-06,
+      "loss": 0.3456,
       "step": 102745
     },
     {
-      "epoch": 3.62,
-      "learning_rate": 9.419437524178499e-06,
-      "loss": 0.2659,
+      "epoch": 3.7031030381662884,
+      "grad_norm": 0.18789133429527283,
+      "learning_rate": 8.32999295921943e-06,
+      "loss": 0.3741,
       "step": 102750
     },
     {
-      "epoch": 3.62,
-      "learning_rate": 9.417209798287654e-06,
-      "loss": 0.2413,
+      "epoch": 3.703283237827513,
+      "grad_norm": 0.276394248008728,
+      "learning_rate": 8.327818373189133e-06,
+      "loss": 0.3662,
       "step": 102755
     },
     {
-      "epoch": 3.62,
-      "learning_rate": 9.414982274729061e-06,
-      "loss": 0.2587,
+      "epoch": 3.7034634374887374,
+      "grad_norm": 0.20188158750534058,
+      "learning_rate": 8.325644014314077e-06,
+      "loss": 0.3959,
       "step": 102760
     },
     {
-      "epoch": 3.62,
-      "learning_rate": 9.412754953531663e-06,
-      "loss": 0.2574,
+      "epoch": 3.703643637149962,
+      "grad_norm": 0.22388440370559692,
+      "learning_rate": 8.323469882623899e-06,
+      "loss": 0.3678,
       "step": 102765
     },
     {
-      "epoch": 3.62,
-      "learning_rate": 9.410527834724369e-06,
-      "loss": 0.2834,
+      "epoch": 3.703823836811187,
+      "grad_norm": 0.2965695559978485,
+      "learning_rate": 8.321295978148217e-06,
+      "loss": 0.3772,
       "step": 102770
     },
     {
-      "epoch": 3.62,
-      "learning_rate": 9.408300918336097e-06,
-      "loss": 0.248,
+      "epoch": 3.704004036472411,
+      "grad_norm": 0.2891329824924469,
+      "learning_rate": 8.319122300916649e-06,
+      "loss": 0.3989,
       "step": 102775
     },
     {
-      "epoch": 3.62,
-      "learning_rate": 9.406074204395754e-06,
-      "loss": 0.2823,
+      "epoch": 3.704184236133636,
+      "grad_norm": 0.2619994580745697,
+      "learning_rate": 8.316948850958809e-06,
+      "loss": 0.3902,
       "step": 102780
     },
     {
-      "epoch": 3.62,
-      "learning_rate": 9.403847692932268e-06,
-      "loss": 0.2686,
+      "epoch": 3.7043644357948606,
+      "grad_norm": 0.24362410604953766,
+      "learning_rate": 8.31477562830431e-06,
+      "loss": 0.3854,
       "step": 102785
     },
     {
-      "epoch": 3.62,
-      "learning_rate": 9.401621383974532e-06,
-      "loss": 0.2514,
+      "epoch": 3.7045446354560854,
+      "grad_norm": 0.27109575271606445,
+      "learning_rate": 8.312602632982756e-06,
+      "loss": 0.373,
       "step": 102790
     },
     {
-      "epoch": 3.62,
-      "learning_rate": 9.399395277551473e-06,
-      "loss": 0.2579,
+      "epoch": 3.70472483511731,
+      "grad_norm": 0.24202807247638702,
+      "learning_rate": 8.310429865023775e-06,
+      "loss": 0.3879,
       "step": 102795
     },
     {
-      "epoch": 3.62,
-      "learning_rate": 9.397169373691977e-06,
-      "loss": 0.2614,
+      "epoch": 3.704905034778535,
+      "grad_norm": 0.245781809091568,
+      "learning_rate": 8.308257324456942e-06,
+      "loss": 0.3847,
       "step": 102800
     },
     {
-      "epoch": 3.62,
-      "learning_rate": 9.394943672424964e-06,
-      "loss": 0.2541,
+      "epoch": 3.705085234439759,
+      "grad_norm": 0.2280445694923401,
+      "learning_rate": 8.306085011311878e-06,
+      "loss": 0.3645,
       "step": 102805
     },
     {
-      "epoch": 3.62,
-      "learning_rate": 9.392718173779319e-06,
-      "loss": 0.2648,
+      "epoch": 3.705265434100984,
+      "grad_norm": 0.2140251249074936,
+      "learning_rate": 8.303912925618174e-06,
+      "loss": 0.3751,
       "step": 102810
     },
     {
-      "epoch": 3.62,
-      "learning_rate": 9.390492877783946e-06,
-      "loss": 0.2603,
+      "epoch": 3.7054456337622086,
+      "grad_norm": 0.3311702311038971,
+      "learning_rate": 8.301741067405424e-06,
+      "loss": 0.3822,
       "step": 102815
     },
     {
-      "epoch": 3.62,
-      "learning_rate": 9.388267784467725e-06,
-      "loss": 0.2555,
+      "epoch": 3.7056258334234333,
+      "grad_norm": 0.2583707273006439,
+      "learning_rate": 8.29956943670322e-06,
+      "loss": 0.3744,
       "step": 102820
     },
     {
-      "epoch": 3.62,
-      "learning_rate": 9.38604289385957e-06,
-      "loss": 0.2853,
+      "epoch": 3.7058060330846576,
+      "grad_norm": 0.2845836877822876,
+      "learning_rate": 8.297398033541137e-06,
+      "loss": 0.389,
       "step": 102825
     },
     {
-      "epoch": 3.62,
-      "learning_rate": 9.383818205988356e-06,
-      "loss": 0.2637,
+      "epoch": 3.7059862327458823,
+      "grad_norm": 0.23895393311977386,
+      "learning_rate": 8.295226857948785e-06,
+      "loss": 0.3877,
       "step": 102830
     },
     {
-      "epoch": 3.62,
-      "learning_rate": 9.381593720882961e-06,
-      "loss": 0.2786,
+      "epoch": 3.706166432407107,
+      "grad_norm": 0.25052353739738464,
+      "learning_rate": 8.29305590995573e-06,
+      "loss": 0.4122,
       "step": 102835
     },
     {
-      "epoch": 3.62,
-      "learning_rate": 9.379369438572292e-06,
-      "loss": 0.2633,
+      "epoch": 3.706346632068332,
+      "grad_norm": 0.23934602737426758,
+      "learning_rate": 8.290885189591555e-06,
+      "loss": 0.4222,
       "step": 102840
     },
     {
-      "epoch": 3.62,
-      "learning_rate": 9.377145359085205e-06,
-      "loss": 0.2493,
+      "epoch": 3.7065268317295565,
+      "grad_norm": 0.24071356654167175,
+      "learning_rate": 8.288714696885835e-06,
+      "loss": 0.3534,
       "step": 102845
     },
     {
-      "epoch": 3.62,
-      "learning_rate": 9.374921482450597e-06,
-      "loss": 0.2476,
+      "epoch": 3.706707031390781,
+      "grad_norm": 0.250177264213562,
+      "learning_rate": 8.28654443186814e-06,
+      "loss": 0.3517,
       "step": 102850
     },
     {
-      "epoch": 3.62,
-      "learning_rate": 9.372697808697328e-06,
-      "loss": 0.2641,
+      "epoch": 3.7068872310520056,
+      "grad_norm": 0.25618675351142883,
+      "learning_rate": 8.284374394568034e-06,
+      "loss": 0.3571,
       "step": 102855
     },
     {
-      "epoch": 3.62,
-      "learning_rate": 9.37047433785429e-06,
-      "loss": 0.2473,
+      "epoch": 3.7070674307132303,
+      "grad_norm": 0.2533912658691406,
+      "learning_rate": 8.282204585015098e-06,
+      "loss": 0.384,
       "step": 102860
     },
     {
-      "epoch": 3.62,
-      "learning_rate": 9.36825106995034e-06,
-      "loss": 0.2858,
+      "epoch": 3.707247630374455,
+      "grad_norm": 0.24974510073661804,
+      "learning_rate": 8.280035003238889e-06,
+      "loss": 0.397,
       "step": 102865
     },
     {
-      "epoch": 3.62,
-      "learning_rate": 9.366028005014348e-06,
-      "loss": 0.2718,
+      "epoch": 3.7074278300356793,
+      "grad_norm": 0.2762242555618286,
+      "learning_rate": 8.277865649268965e-06,
+      "loss": 0.3575,
       "step": 102870
     },
     {
-      "epoch": 3.62,
-      "learning_rate": 9.363805143075175e-06,
-      "loss": 0.2565,
+      "epoch": 3.707608029696904,
+      "grad_norm": 0.28377920389175415,
+      "learning_rate": 8.275696523134885e-06,
+      "loss": 0.3622,
       "step": 102875
     },
     {
-      "epoch": 3.62,
-      "learning_rate": 9.36158248416169e-06,
-      "loss": 0.2571,
+      "epoch": 3.707788229358129,
+      "grad_norm": 0.2506207525730133,
+      "learning_rate": 8.273527624866192e-06,
+      "loss": 0.4064,
       "step": 102880
     },
     {
-      "epoch": 3.62,
-      "learning_rate": 9.359360028302755e-06,
-      "loss": 0.2721,
+      "epoch": 3.7079684290193535,
+      "grad_norm": 0.27845442295074463,
+      "learning_rate": 8.271358954492458e-06,
+      "loss": 0.3983,
       "step": 102885
     },
     {
-      "epoch": 3.62,
-      "learning_rate": 9.357137775527219e-06,
-      "loss": 0.2719,
+      "epoch": 3.7081486286805783,
+      "grad_norm": 0.1891564130783081,
+      "learning_rate": 8.269190512043226e-06,
+      "loss": 0.3888,
       "step": 102890
     },
     {
-      "epoch": 3.62,
-      "learning_rate": 9.354915725863936e-06,
-      "loss": 0.2675,
+      "epoch": 3.7083288283418026,
+      "grad_norm": 0.21435974538326263,
+      "learning_rate": 8.267022297548016e-06,
+      "loss": 0.3931,
       "step": 102895
     },
     {
-      "epoch": 3.62,
-      "learning_rate": 9.352693879341756e-06,
-      "loss": 0.2651,
+      "epoch": 3.7085090280030273,
+      "grad_norm": 0.2660645842552185,
+      "learning_rate": 8.264854311036399e-06,
+      "loss": 0.3827,
       "step": 102900
     },
     {
-      "epoch": 3.62,
-      "learning_rate": 9.350472235989547e-06,
-      "loss": 0.2441,
+      "epoch": 3.708689227664252,
+      "grad_norm": 0.23603610694408417,
+      "learning_rate": 8.262686552537894e-06,
+      "loss": 0.3614,
       "step": 102905
     },
     {
-      "epoch": 3.62,
-      "learning_rate": 9.348250795836141e-06,
-      "loss": 0.251,
+      "epoch": 3.7088694273254768,
+      "grad_norm": 0.24526463449001312,
+      "learning_rate": 8.260519022082058e-06,
+      "loss": 0.3742,
       "step": 102910
     },
     {
-      "epoch": 3.62,
-      "learning_rate": 9.346029558910375e-06,
-      "loss": 0.2562,
+      "epoch": 3.709049626986701,
+      "grad_norm": 0.24888326227664948,
+      "learning_rate": 8.258351719698401e-06,
+      "loss": 0.3826,
       "step": 102915
     },
     {
-      "epoch": 3.62,
-      "learning_rate": 9.343808525241108e-06,
-      "loss": 0.2616,
+      "epoch": 3.7092298266479258,
+      "grad_norm": 0.2145439237356186,
+      "learning_rate": 8.256184645416453e-06,
+      "loss": 0.3911,
       "step": 102920
     },
     {
-      "epoch": 3.62,
-      "learning_rate": 9.341587694857168e-06,
-      "loss": 0.2518,
+      "epoch": 3.7094100263091505,
+      "grad_norm": 0.2232615351676941,
+      "learning_rate": 8.254017799265757e-06,
+      "loss": 0.3904,
       "step": 102925
     },
     {
-      "epoch": 3.62,
-      "learning_rate": 9.339367067787393e-06,
-      "loss": 0.2519,
+      "epoch": 3.7095902259703752,
+      "grad_norm": 0.2325001209974289,
+      "learning_rate": 8.251851181275824e-06,
+      "loss": 0.3509,
       "step": 102930
     },
     {
-      "epoch": 3.62,
-      "learning_rate": 9.33714664406061e-06,
-      "loss": 0.2531,
+      "epoch": 3.7097704256316,
+      "grad_norm": 0.2643667459487915,
+      "learning_rate": 8.249684791476178e-06,
+      "loss": 0.3894,
       "step": 102935
     },
     {
-      "epoch": 3.62,
-      "learning_rate": 9.334926423705662e-06,
-      "loss": 0.2655,
+      "epoch": 3.7099506252928247,
+      "grad_norm": 0.22574815154075623,
+      "learning_rate": 8.247518629896334e-06,
+      "loss": 0.3611,
       "step": 102940
     },
     {
-      "epoch": 3.62,
-      "learning_rate": 9.33270640675137e-06,
-      "loss": 0.2587,
+      "epoch": 3.710130824954049,
+      "grad_norm": 0.26316380500793457,
+      "learning_rate": 8.245352696565797e-06,
+      "loss": 0.3905,
       "step": 102945
     },
     {
-      "epoch": 3.62,
-      "learning_rate": 9.330486593226553e-06,
-      "loss": 0.2436,
+      "epoch": 3.7103110246152737,
+      "grad_norm": 0.20727650821208954,
+      "learning_rate": 8.243186991514092e-06,
+      "loss": 0.3788,
       "step": 102950
     },
     {
-      "epoch": 3.62,
-      "learning_rate": 9.32826698316004e-06,
-      "loss": 0.2479,
+      "epoch": 3.7104912242764985,
+      "grad_norm": 0.24111421406269073,
+      "learning_rate": 8.241021514770721e-06,
+      "loss": 0.364,
       "step": 102955
     },
     {
-      "epoch": 3.62,
-      "learning_rate": 9.32604757658066e-06,
-      "loss": 0.2617,
+      "epoch": 3.7106714239377228,
+      "grad_norm": 0.21225392818450928,
+      "learning_rate": 8.238856266365188e-06,
+      "loss": 0.3928,
       "step": 102960
     },
     {
-      "epoch": 3.62,
-      "learning_rate": 9.323828373517224e-06,
-      "loss": 0.2575,
+      "epoch": 3.7108516235989475,
+      "grad_norm": 0.27659305930137634,
+      "learning_rate": 8.236691246326994e-06,
+      "loss": 0.3714,
       "step": 102965
     },
     {
-      "epoch": 3.62,
-      "learning_rate": 9.321609373998538e-06,
-      "loss": 0.276,
+      "epoch": 3.7110318232601722,
+      "grad_norm": 0.26336660981178284,
+      "learning_rate": 8.234526454685634e-06,
+      "loss": 0.3199,
       "step": 102970
     },
     {
-      "epoch": 3.62,
-      "learning_rate": 9.319390578053427e-06,
-      "loss": 0.2469,
+      "epoch": 3.711212022921397,
+      "grad_norm": 0.2715068459510803,
+      "learning_rate": 8.232361891470599e-06,
+      "loss": 0.3795,
       "step": 102975
     },
     {
-      "epoch": 3.62,
-      "learning_rate": 9.3171719857107e-06,
-      "loss": 0.2463,
+      "epoch": 3.7113922225826217,
+      "grad_norm": 0.2276494801044464,
+      "learning_rate": 8.230197556711403e-06,
+      "loss": 0.3569,
       "step": 102980
     },
     {
-      "epoch": 3.62,
-      "learning_rate": 9.314953596999156e-06,
-      "loss": 0.2471,
+      "epoch": 3.7115724222438464,
+      "grad_norm": 0.25353577733039856,
+      "learning_rate": 8.228033450437503e-06,
+      "loss": 0.406,
       "step": 102985
     },
     {
-      "epoch": 3.62,
-      "learning_rate": 9.312735411947597e-06,
-      "loss": 0.2415,
+      "epoch": 3.7117526219050707,
+      "grad_norm": 0.20923548936843872,
+      "learning_rate": 8.225869572678405e-06,
+      "loss": 0.3721,
       "step": 102990
     },
     {
-      "epoch": 3.62,
-      "learning_rate": 9.310517430584836e-06,
-      "loss": 0.2512,
+      "epoch": 3.7119328215662954,
+      "grad_norm": 0.2902141809463501,
+      "learning_rate": 8.22370592346359e-06,
+      "loss": 0.4074,
       "step": 102995
     },
     {
-      "epoch": 3.62,
-      "learning_rate": 9.308299652939666e-06,
-      "loss": 0.2563,
+      "epoch": 3.71211302122752,
+      "grad_norm": 0.26280632615089417,
+      "learning_rate": 8.221542502822533e-06,
+      "loss": 0.3569,
       "step": 103000
     },
     {
-      "epoch": 3.62,
-      "eval_loss": 0.25192898511886597,
-      "eval_runtime": 10.5381,
-      "eval_samples_per_second": 9.489,
-      "eval_steps_per_second": 9.489,
+      "epoch": 3.71211302122752,
+      "eval_loss": 0.42930299043655396,
+      "eval_runtime": 3.5329,
+      "eval_samples_per_second": 28.306,
+      "eval_steps_per_second": 7.076,
       "step": 103000
     },
     {
-      "epoch": 3.62,
-      "learning_rate": 9.306082079040879e-06,
-      "loss": 0.2784,
+      "epoch": 3.7122932208887445,
+      "grad_norm": 0.24260425567626953,
+      "learning_rate": 8.219379310784708e-06,
+      "loss": 0.3687,
       "step": 103005
     },
     {
-      "epoch": 3.62,
-      "learning_rate": 9.30386470891727e-06,
-      "loss": 0.2625,
+      "epoch": 3.712473420549969,
+      "grad_norm": 0.2277086079120636,
+      "learning_rate": 8.217216347379594e-06,
+      "loss": 0.392,
       "step": 103010
     },
     {
-      "epoch": 3.62,
-      "learning_rate": 9.301647542597643e-06,
-      "loss": 0.2697,
+      "epoch": 3.712653620211194,
+      "grad_norm": 0.28349313139915466,
+      "learning_rate": 8.21505361263665e-06,
+      "loss": 0.3905,
       "step": 103015
     },
     {
-      "epoch": 3.62,
-      "learning_rate": 9.299430580110776e-06,
-      "loss": 0.2611,
+      "epoch": 3.7128338198724187,
+      "grad_norm": 0.2919948399066925,
+      "learning_rate": 8.212891106585357e-06,
+      "loss": 0.3954,
       "step": 103020
     },
     {
-      "epoch": 3.62,
-      "learning_rate": 9.29721382148546e-06,
-      "loss": 0.2729,
+      "epoch": 3.7130140195336434,
+      "grad_norm": 0.20646890997886658,
+      "learning_rate": 8.210728829255173e-06,
+      "loss": 0.3824,
       "step": 103025
     },
     {
-      "epoch": 3.62,
-      "learning_rate": 9.29499726675046e-06,
-      "loss": 0.2328,
+      "epoch": 3.713194219194868,
+      "grad_norm": 0.2770639657974243,
+      "learning_rate": 8.208566780675561e-06,
+      "loss": 0.3688,
       "step": 103030
     },
     {
-      "epoch": 3.63,
-      "learning_rate": 9.292780915934585e-06,
-      "loss": 0.2722,
+      "epoch": 3.7133744188560924,
+      "grad_norm": 0.31089577078819275,
+      "learning_rate": 8.206404960875972e-06,
+      "loss": 0.3768,
       "step": 103035
     },
     {
-      "epoch": 3.63,
-      "learning_rate": 9.290564769066592e-06,
-      "loss": 0.278,
+      "epoch": 3.713554618517317,
+      "grad_norm": 0.20020338892936707,
+      "learning_rate": 8.204243369885859e-06,
+      "loss": 0.3446,
       "step": 103040
     },
     {
-      "epoch": 3.63,
-      "learning_rate": 9.288348826175266e-06,
-      "loss": 0.2406,
+      "epoch": 3.713734818178542,
+      "grad_norm": 0.25456637144088745,
+      "learning_rate": 8.202082007734684e-06,
+      "loss": 0.3762,
       "step": 103045
     },
     {
-      "epoch": 3.63,
-      "learning_rate": 9.28613308728937e-06,
-      "loss": 0.2726,
+      "epoch": 3.713915017839766,
+      "grad_norm": 0.23088417947292328,
+      "learning_rate": 8.19992087445189e-06,
+      "loss": 0.3943,
       "step": 103050
     },
     {
-      "epoch": 3.63,
-      "learning_rate": 9.283917552437684e-06,
-      "loss": 0.2624,
+      "epoch": 3.714095217500991,
+      "grad_norm": 0.26196154952049255,
+      "learning_rate": 8.197759970066923e-06,
+      "loss": 0.4015,
       "step": 103055
     },
     {
-      "epoch": 3.63,
-      "learning_rate": 9.281702221648966e-06,
-      "loss": 0.2461,
+      "epoch": 3.7142754171622157,
+      "grad_norm": 0.21619191765785217,
+      "learning_rate": 8.195599294609222e-06,
+      "loss": 0.3914,
       "step": 103060
     },
     {
-      "epoch": 3.63,
-      "learning_rate": 9.279487094951991e-06,
-      "loss": 0.2531,
+      "epoch": 3.7144556168234404,
+      "grad_norm": 0.29650604724884033,
+      "learning_rate": 8.19343884810822e-06,
+      "loss": 0.3818,
       "step": 103065
     },
     {
-      "epoch": 3.63,
-      "learning_rate": 9.277272172375511e-06,
-      "loss": 0.2718,
+      "epoch": 3.714635816484665,
+      "grad_norm": 0.26321521401405334,
+      "learning_rate": 8.191278630593367e-06,
+      "loss": 0.3734,
       "step": 103070
     },
     {
-      "epoch": 3.63,
-      "learning_rate": 9.275057453948297e-06,
-      "loss": 0.2636,
+      "epoch": 3.71481601614589,
+      "grad_norm": 0.2702058255672455,
+      "learning_rate": 8.189118642094095e-06,
+      "loss": 0.3759,
       "step": 103075
     },
     {
-      "epoch": 3.63,
-      "learning_rate": 9.2728429396991e-06,
-      "loss": 0.2607,
+      "epoch": 3.714996215807114,
+      "grad_norm": 0.2815110683441162,
+      "learning_rate": 8.186958882639813e-06,
+      "loss": 0.3757,
       "step": 103080
     },
     {
-      "epoch": 3.63,
-      "learning_rate": 9.270628629656663e-06,
-      "loss": 0.2376,
+      "epoch": 3.715176415468339,
+      "grad_norm": 0.28946641087532043,
+      "learning_rate": 8.18479935225997e-06,
+      "loss": 0.3692,
       "step": 103085
     },
     {
-      "epoch": 3.63,
-      "learning_rate": 9.268414523849755e-06,
-      "loss": 0.2535,
+      "epoch": 3.7153566151295636,
+      "grad_norm": 0.3027191758155823,
+      "learning_rate": 8.18264005098397e-06,
+      "loss": 0.3738,
       "step": 103090
     },
     {
-      "epoch": 3.63,
-      "learning_rate": 9.26620062230712e-06,
-      "loss": 0.2569,
+      "epoch": 3.7155368147907883,
+      "grad_norm": 0.22006292641162872,
+      "learning_rate": 8.18048097884126e-06,
+      "loss": 0.398,
       "step": 103095
     },
     {
-      "epoch": 3.63,
-      "learning_rate": 9.263986925057497e-06,
-      "loss": 0.2596,
+      "epoch": 3.7157170144520126,
+      "grad_norm": 0.27258020639419556,
+      "learning_rate": 8.178322135861233e-06,
+      "loss": 0.3971,
       "step": 103100
     },
     {
-      "epoch": 3.63,
-      "learning_rate": 9.261773432129628e-06,
-      "loss": 0.2691,
+      "epoch": 3.7158972141132374,
+      "grad_norm": 0.22416990995407104,
+      "learning_rate": 8.176163522073302e-06,
+      "loss": 0.3712,
       "step": 103105
     },
     {
-      "epoch": 3.63,
-      "learning_rate": 9.259560143552265e-06,
-      "loss": 0.2405,
+      "epoch": 3.716077413774462,
+      "grad_norm": 0.22454127669334412,
+      "learning_rate": 8.174005137506894e-06,
+      "loss": 0.3709,
       "step": 103110
     },
     {
-      "epoch": 3.63,
-      "learning_rate": 9.257347059354132e-06,
-      "loss": 0.288,
+      "epoch": 3.716257613435687,
+      "grad_norm": 0.21970023214817047,
+      "learning_rate": 8.171846982191409e-06,
+      "loss": 0.3468,
       "step": 103115
     },
     {
-      "epoch": 3.63,
-      "learning_rate": 9.255134179563985e-06,
-      "loss": 0.2593,
+      "epoch": 3.7164378130969116,
+      "grad_norm": 0.23587019741535187,
+      "learning_rate": 8.169689056156249e-06,
+      "loss": 0.3688,
       "step": 103120
     },
     {
-      "epoch": 3.63,
-      "learning_rate": 9.252921504210533e-06,
-      "loss": 0.2573,
+      "epoch": 3.716618012758136,
+      "grad_norm": 0.2582648992538452,
+      "learning_rate": 8.167531359430815e-06,
+      "loss": 0.3482,
       "step": 103125
     },
     {
-      "epoch": 3.63,
-      "learning_rate": 9.250709033322524e-06,
-      "loss": 0.2445,
+      "epoch": 3.7167982124193606,
+      "grad_norm": 0.23627053201198578,
+      "learning_rate": 8.165373892044503e-06,
+      "loss": 0.3601,
       "step": 103130
     },
     {
-      "epoch": 3.63,
-      "learning_rate": 9.248496766928683e-06,
-      "loss": 0.2501,
+      "epoch": 3.7169784120805853,
+      "grad_norm": 0.2360253483057022,
+      "learning_rate": 8.163216654026721e-06,
+      "loss": 0.3759,
       "step": 103135
     },
     {
-      "epoch": 3.63,
-      "learning_rate": 9.246284705057728e-06,
-      "loss": 0.2482,
+      "epoch": 3.71715861174181,
+      "grad_norm": 0.32699692249298096,
+      "learning_rate": 8.16105964540685e-06,
+      "loss": 0.4162,
       "step": 103140
     },
     {
-      "epoch": 3.63,
-      "learning_rate": 9.244072847738377e-06,
-      "loss": 0.2378,
+      "epoch": 3.7173388114030343,
+      "grad_norm": 0.2977791726589203,
+      "learning_rate": 8.158902866214282e-06,
+      "loss": 0.3725,
       "step": 103145
     },
     {
-      "epoch": 3.63,
-      "learning_rate": 9.241861194999365e-06,
-      "loss": 0.2562,
+      "epoch": 3.717519011064259,
+      "grad_norm": 0.29648059606552124,
+      "learning_rate": 8.156746316478403e-06,
+      "loss": 0.412,
       "step": 103150
     },
     {
-      "epoch": 3.63,
-      "learning_rate": 9.2396497468694e-06,
-      "loss": 0.2551,
+      "epoch": 3.717699210725484,
+      "grad_norm": 0.2452070415019989,
+      "learning_rate": 8.154589996228595e-06,
+      "loss": 0.3724,
       "step": 103155
     },
     {
-      "epoch": 3.63,
-      "learning_rate": 9.237438503377196e-06,
-      "loss": 0.2423,
+      "epoch": 3.7178794103867085,
+      "grad_norm": 0.23447571694850922,
+      "learning_rate": 8.152433905494228e-06,
+      "loss": 0.3573,
       "step": 103160
     },
     {
-      "epoch": 3.63,
-      "learning_rate": 9.235227464551457e-06,
-      "loss": 0.2807,
+      "epoch": 3.7180596100479333,
+      "grad_norm": 0.22790288925170898,
+      "learning_rate": 8.150278044304702e-06,
+      "loss": 0.3927,
       "step": 103165
     },
     {
-      "epoch": 3.63,
-      "learning_rate": 9.233016630420898e-06,
-      "loss": 0.2741,
+      "epoch": 3.718239809709158,
+      "grad_norm": 0.24579019844532013,
+      "learning_rate": 8.14812241268936e-06,
+      "loss": 0.3558,
       "step": 103170
     },
     {
-      "epoch": 3.63,
-      "learning_rate": 9.230806001014238e-06,
-      "loss": 0.2461,
+      "epoch": 3.7184200093703823,
+      "grad_norm": 0.2221144735813141,
+      "learning_rate": 8.145967010677597e-06,
+      "loss": 0.3759,
       "step": 103175
     },
     {
-      "epoch": 3.63,
-      "learning_rate": 9.228595576360172e-06,
-      "loss": 0.2494,
+      "epoch": 3.718600209031607,
+      "grad_norm": 0.2838151454925537,
+      "learning_rate": 8.143811838298767e-06,
+      "loss": 0.3491,
       "step": 103180
     },
     {
-      "epoch": 3.63,
-      "learning_rate": 9.226385356487388e-06,
-      "loss": 0.2675,
+      "epoch": 3.7187804086928318,
+      "grad_norm": 0.20898236334323883,
+      "learning_rate": 8.141656895582233e-06,
+      "loss": 0.3515,
       "step": 103185
     },
     {
-      "epoch": 3.63,
-      "learning_rate": 9.2241753414246e-06,
-      "loss": 0.2761,
+      "epoch": 3.718960608354056,
+      "grad_norm": 0.2016228437423706,
+      "learning_rate": 8.139502182557373e-06,
+      "loss": 0.3518,
       "step": 103190
     },
     {
-      "epoch": 3.63,
-      "learning_rate": 9.221965531200504e-06,
-      "loss": 0.2476,
+      "epoch": 3.719140808015281,
+      "grad_norm": 0.29854917526245117,
+      "learning_rate": 8.137347699253526e-06,
+      "loss": 0.3839,
       "step": 103195
     },
     {
-      "epoch": 3.63,
-      "learning_rate": 9.219755925843776e-06,
-      "loss": 0.2674,
+      "epoch": 3.7193210076765055,
+      "grad_norm": 0.24756811559200287,
+      "learning_rate": 8.135193445700043e-06,
+      "loss": 0.3774,
       "step": 103200
     },
     {
-      "epoch": 3.63,
-      "learning_rate": 9.217546525383128e-06,
-      "loss": 0.2556,
+      "epoch": 3.7195012073377303,
+      "grad_norm": 0.25717464089393616,
+      "learning_rate": 8.133039421926291e-06,
+      "loss": 0.3913,
       "step": 103205
     },
     {
-      "epoch": 3.63,
-      "learning_rate": 9.215337329847238e-06,
-      "loss": 0.263,
+      "epoch": 3.719681406998955,
+      "grad_norm": 0.216623917222023,
+      "learning_rate": 8.130885627961612e-06,
+      "loss": 0.401,
       "step": 103210
     },
     {
-      "epoch": 3.63,
-      "learning_rate": 9.213128339264788e-06,
-      "loss": 0.2573,
+      "epoch": 3.7198616066601797,
+      "grad_norm": 0.2196447253227234,
+      "learning_rate": 8.12873206383535e-06,
+      "loss": 0.3996,
       "step": 103215
     },
     {
-      "epoch": 3.63,
-      "learning_rate": 9.210919553664454e-06,
-      "loss": 0.237,
+      "epoch": 3.720041806321404,
+      "grad_norm": 0.17287389934062958,
+      "learning_rate": 8.126578729576851e-06,
+      "loss": 0.3962,
       "step": 103220
     },
     {
-      "epoch": 3.63,
-      "learning_rate": 9.208710973074932e-06,
-      "loss": 0.2578,
+      "epoch": 3.7202220059826288,
+      "grad_norm": 0.23201312124729156,
+      "learning_rate": 8.124425625215437e-06,
+      "loss": 0.3583,
       "step": 103225
     },
     {
-      "epoch": 3.63,
-      "learning_rate": 9.206502597524886e-06,
-      "loss": 0.2596,
+      "epoch": 3.7204022056438535,
+      "grad_norm": 0.21326382458209991,
+      "learning_rate": 8.122272750780465e-06,
+      "loss": 0.38,
       "step": 103230
     },
     {
-      "epoch": 3.63,
-      "learning_rate": 9.204294427043e-06,
-      "loss": 0.2492,
+      "epoch": 3.7205824053050778,
+      "grad_norm": 0.2922148108482361,
+      "learning_rate": 8.120120106301263e-06,
+      "loss": 0.3748,
       "step": 103235
     },
     {
-      "epoch": 3.63,
-      "learning_rate": 9.202086461657935e-06,
-      "loss": 0.229,
+      "epoch": 3.7207626049663025,
+      "grad_norm": 0.22092339396476746,
+      "learning_rate": 8.117967691807155e-06,
+      "loss": 0.4146,
       "step": 103240
     },
     {
-      "epoch": 3.63,
-      "learning_rate": 9.199878701398376e-06,
-      "loss": 0.2658,
+      "epoch": 3.7209428046275272,
+      "grad_norm": 0.2587362825870514,
+      "learning_rate": 8.115815507327465e-06,
+      "loss": 0.3996,
       "step": 103245
     },
     {
-      "epoch": 3.63,
-      "learning_rate": 9.197671146292977e-06,
-      "loss": 0.2651,
+      "epoch": 3.721123004288752,
+      "grad_norm": 0.2722143232822418,
+      "learning_rate": 8.113663552891516e-06,
+      "loss": 0.4322,
       "step": 103250
     },
     {
-      "epoch": 3.63,
-      "learning_rate": 9.195463796370405e-06,
-      "loss": 0.2714,
+      "epoch": 3.7213032039499767,
+      "grad_norm": 0.21530699729919434,
+      "learning_rate": 8.11151182852864e-06,
+      "loss": 0.3856,
       "step": 103255
     },
     {
-      "epoch": 3.63,
-      "learning_rate": 9.19325665165931e-06,
-      "loss": 0.2282,
+      "epoch": 3.7214834036112014,
+      "grad_norm": 0.22060318291187286,
+      "learning_rate": 8.10936033426815e-06,
+      "loss": 0.3657,
       "step": 103260
     },
     {
-      "epoch": 3.63,
-      "learning_rate": 9.191049712188371e-06,
-      "loss": 0.2691,
+      "epoch": 3.7216636032724257,
+      "grad_norm": 0.36303722858428955,
+      "learning_rate": 8.107209070139341e-06,
+      "loss": 0.4083,
       "step": 103265
     },
     {
-      "epoch": 3.63,
-      "learning_rate": 9.18884297798623e-06,
-      "loss": 0.2691,
+      "epoch": 3.7218438029336505,
+      "grad_norm": 0.189983531832695,
+      "learning_rate": 8.105058036171547e-06,
+      "loss": 0.3606,
       "step": 103270
     },
     {
-      "epoch": 3.63,
-      "learning_rate": 9.18663644908154e-06,
-      "loss": 0.252,
+      "epoch": 3.722024002594875,
+      "grad_norm": 0.2879542410373688,
+      "learning_rate": 8.102907232394057e-06,
+      "loss": 0.3928,
       "step": 103275
     },
     {
-      "epoch": 3.63,
-      "learning_rate": 9.184430125502958e-06,
-      "loss": 0.2643,
+      "epoch": 3.7222042022560995,
+      "grad_norm": 0.2692555785179138,
+      "learning_rate": 8.100756658836202e-06,
+      "loss": 0.3752,
       "step": 103280
     },
     {
-      "epoch": 3.63,
-      "learning_rate": 9.182224007279122e-06,
-      "loss": 0.2822,
+      "epoch": 3.7223844019173242,
+      "grad_norm": 0.298684298992157,
+      "learning_rate": 8.098606315527258e-06,
+      "loss": 0.3915,
       "step": 103285
     },
     {
-      "epoch": 3.63,
-      "learning_rate": 9.18001809443869e-06,
-      "loss": 0.2773,
+      "epoch": 3.722564601578549,
+      "grad_norm": 0.2624755799770355,
+      "learning_rate": 8.096456202496518e-06,
+      "loss": 0.3932,
       "step": 103290
     },
     {
-      "epoch": 3.63,
-      "learning_rate": 9.177812387010295e-06,
-      "loss": 0.2452,
+      "epoch": 3.7227448012397737,
+      "grad_norm": 0.25200989842414856,
+      "learning_rate": 8.094306319773304e-06,
+      "loss": 0.3696,
       "step": 103295
     },
     {
-      "epoch": 3.63,
-      "learning_rate": 9.175606885022575e-06,
-      "loss": 0.254,
+      "epoch": 3.7229250009009984,
+      "grad_norm": 0.26927831768989563,
+      "learning_rate": 8.092156667386891e-06,
+      "loss": 0.3891,
       "step": 103300
     },
     {
-      "epoch": 3.63,
-      "learning_rate": 9.173401588504177e-06,
-      "loss": 0.278,
+      "epoch": 3.723105200562223,
+      "grad_norm": 0.21889813244342804,
+      "learning_rate": 8.090007245366567e-06,
+      "loss": 0.3829,
       "step": 103305
     },
     {
-      "epoch": 3.63,
-      "learning_rate": 9.171196497483731e-06,
-      "loss": 0.2877,
+      "epoch": 3.7232854002234474,
+      "grad_norm": 0.24276183545589447,
+      "learning_rate": 8.087858053741626e-06,
+      "loss": 0.3566,
       "step": 103310
     },
     {
-      "epoch": 3.63,
-      "learning_rate": 9.168991611989855e-06,
-      "loss": 0.2475,
+      "epoch": 3.723465599884672,
+      "grad_norm": 0.2647343873977661,
+      "learning_rate": 8.085709092541332e-06,
+      "loss": 0.3916,
       "step": 103315
     },
     {
-      "epoch": 3.64,
-      "learning_rate": 9.166786932051203e-06,
-      "loss": 0.2479,
+      "epoch": 3.723645799545897,
+      "grad_norm": 0.23813170194625854,
+      "learning_rate": 8.083560361794988e-06,
+      "loss": 0.3831,
       "step": 103320
     },
     {
-      "epoch": 3.64,
-      "learning_rate": 9.164582457696384e-06,
-      "loss": 0.2354,
+      "epoch": 3.7238259992071217,
+      "grad_norm": 0.21774239838123322,
+      "learning_rate": 8.081411861531856e-06,
+      "loss": 0.3904,
       "step": 103325
     },
     {
-      "epoch": 3.64,
-      "learning_rate": 9.162378188954027e-06,
-      "loss": 0.2533,
+      "epoch": 3.724006198868346,
+      "grad_norm": 0.2522357106208801,
+      "learning_rate": 8.079263591781213e-06,
+      "loss": 0.3394,
       "step": 103330
     },
     {
-      "epoch": 3.64,
-      "learning_rate": 9.160174125852745e-06,
-      "loss": 0.2438,
+      "epoch": 3.7241863985295707,
+      "grad_norm": 0.2412281036376953,
+      "learning_rate": 8.077115552572328e-06,
+      "loss": 0.3927,
       "step": 103335
     },
     {
-      "epoch": 3.64,
-      "learning_rate": 9.157970268421162e-06,
-      "loss": 0.2633,
+      "epoch": 3.7243665981907954,
+      "grad_norm": 0.20674298703670502,
+      "learning_rate": 8.074967743934466e-06,
+      "loss": 0.3638,
       "step": 103340
     },
     {
-      "epoch": 3.64,
-      "learning_rate": 9.155766616687902e-06,
-      "loss": 0.2715,
+      "epoch": 3.72454679785202,
+      "grad_norm": 0.2086464762687683,
+      "learning_rate": 8.072820165896886e-06,
+      "loss": 0.3662,
       "step": 103345
     },
     {
-      "epoch": 3.64,
-      "learning_rate": 9.153563170681573e-06,
-      "loss": 0.2678,
+      "epoch": 3.724726997513245,
+      "grad_norm": 0.24110788106918335,
+      "learning_rate": 8.07067281848886e-06,
+      "loss": 0.4261,
       "step": 103350
     },
     {
-      "epoch": 3.64,
-      "learning_rate": 9.151359930430772e-06,
-      "loss": 0.2665,
+      "epoch": 3.724907197174469,
+      "grad_norm": 0.2587282955646515,
+      "learning_rate": 8.06852570173964e-06,
+      "loss": 0.3886,
       "step": 103355
     },
     {
-      "epoch": 3.64,
-      "learning_rate": 9.149156895964128e-06,
-      "loss": 0.2473,
+      "epoch": 3.725087396835694,
+      "grad_norm": 0.20811423659324646,
+      "learning_rate": 8.06637881567848e-06,
+      "loss": 0.3289,
       "step": 103360
     },
     {
-      "epoch": 3.64,
-      "learning_rate": 9.146954067310237e-06,
-      "loss": 0.2512,
+      "epoch": 3.7252675964969186,
+      "grad_norm": 0.19246096909046173,
+      "learning_rate": 8.064232160334632e-06,
+      "loss": 0.351,
       "step": 103365
     },
     {
-      "epoch": 3.64,
-      "learning_rate": 9.144751444497698e-06,
-      "loss": 0.2683,
+      "epoch": 3.7254477961581434,
+      "grad_norm": 0.2847633361816406,
+      "learning_rate": 8.062085735737329e-06,
+      "loss": 0.3723,
       "step": 103370
     },
     {
-      "epoch": 3.64,
-      "learning_rate": 9.142549027555103e-06,
-      "loss": 0.2704,
+      "epoch": 3.7256279958193677,
+      "grad_norm": 0.21437424421310425,
+      "learning_rate": 8.059939541915847e-06,
+      "loss": 0.3805,
       "step": 103375
     },
     {
-      "epoch": 3.64,
-      "learning_rate": 9.140346816511069e-06,
-      "loss": 0.2578,
+      "epoch": 3.7258081954805924,
+      "grad_norm": 0.23143932223320007,
+      "learning_rate": 8.057793578899403e-06,
+      "loss": 0.3444,
       "step": 103380
     },
     {
-      "epoch": 3.64,
-      "learning_rate": 9.138144811394178e-06,
-      "loss": 0.2615,
+      "epoch": 3.725988395141817,
+      "grad_norm": 0.18867234885692596,
+      "learning_rate": 8.055647846717232e-06,
+      "loss": 0.367,
       "step": 103385
     },
     {
-      "epoch": 3.64,
-      "learning_rate": 9.135943012233017e-06,
-      "loss": 0.2751,
+      "epoch": 3.726168594803042,
+      "grad_norm": 0.22278887033462524,
+      "learning_rate": 8.053502345398583e-06,
+      "loss": 0.3791,
       "step": 103390
     },
     {
-      "epoch": 3.64,
-      "learning_rate": 9.13374141905618e-06,
-      "loss": 0.2699,
+      "epoch": 3.7263487944642666,
+      "grad_norm": 0.2264564037322998,
+      "learning_rate": 8.051357074972688e-06,
+      "loss": 0.3769,
       "step": 103395
     },
     {
-      "epoch": 3.64,
-      "learning_rate": 9.131540031892261e-06,
-      "loss": 0.2361,
+      "epoch": 3.726528994125491,
+      "grad_norm": 0.3085385262966156,
+      "learning_rate": 8.04921203546877e-06,
+      "loss": 0.3945,
       "step": 103400
     },
     {
-      "epoch": 3.64,
-      "learning_rate": 9.12933885076984e-06,
-      "loss": 0.2633,
+      "epoch": 3.7267091937867156,
+      "grad_norm": 0.1784570962190628,
+      "learning_rate": 8.047067226916058e-06,
+      "loss": 0.3429,
       "step": 103405
     },
     {
-      "epoch": 3.64,
-      "learning_rate": 9.127137875717492e-06,
-      "loss": 0.2574,
+      "epoch": 3.7268893934479403,
+      "grad_norm": 0.1708642989397049,
+      "learning_rate": 8.044922649343762e-06,
+      "loss": 0.3599,
       "step": 103410
     },
     {
-      "epoch": 3.64,
-      "learning_rate": 9.124937106763793e-06,
-      "loss": 0.2722,
+      "epoch": 3.727069593109165,
+      "grad_norm": 0.30440521240234375,
+      "learning_rate": 8.042778302781123e-06,
+      "loss": 0.4162,
       "step": 103415
     },
     {
-      "epoch": 3.64,
-      "learning_rate": 9.12273654393733e-06,
-      "loss": 0.2493,
+      "epoch": 3.7272497927703894,
+      "grad_norm": 0.28747257590293884,
+      "learning_rate": 8.040634187257345e-06,
+      "loss": 0.3721,
       "step": 103420
     },
     {
-      "epoch": 3.64,
-      "learning_rate": 9.120536187266667e-06,
-      "loss": 0.2745,
+      "epoch": 3.727429992431614,
+      "grad_norm": 0.24325698614120483,
+      "learning_rate": 8.038490302801641e-06,
+      "loss": 0.3838,
       "step": 103425
     },
     {
-      "epoch": 3.64,
-      "learning_rate": 9.118336036780373e-06,
-      "loss": 0.263,
+      "epoch": 3.727610192092839,
+      "grad_norm": 0.22188496589660645,
+      "learning_rate": 8.036346649443224e-06,
+      "loss": 0.3684,
       "step": 103430
     },
     {
-      "epoch": 3.64,
-      "learning_rate": 9.116136092507024e-06,
-      "loss": 0.2741,
+      "epoch": 3.7277903917540636,
+      "grad_norm": 0.23214270174503326,
+      "learning_rate": 8.034203227211292e-06,
+      "loss": 0.3609,
       "step": 103435
     },
     {
-      "epoch": 3.64,
-      "learning_rate": 9.113936354475183e-06,
-      "loss": 0.2627,
+      "epoch": 3.7279705914152883,
+      "grad_norm": 0.2736099660396576,
+      "learning_rate": 8.032060036135067e-06,
+      "loss": 0.3663,
       "step": 103440
     },
     {
-      "epoch": 3.64,
-      "learning_rate": 9.1117368227134e-06,
-      "loss": 0.2345,
+      "epoch": 3.728150791076513,
+      "grad_norm": 0.26921337842941284,
+      "learning_rate": 8.029917076243745e-06,
+      "loss": 0.3804,
       "step": 103445
     },
     {
-      "epoch": 3.64,
-      "learning_rate": 9.109537497250242e-06,
-      "loss": 0.2622,
+      "epoch": 3.7283309907377373,
+      "grad_norm": 0.25673896074295044,
+      "learning_rate": 8.027774347566499e-06,
+      "loss": 0.3767,
       "step": 103450
     },
     {
-      "epoch": 3.64,
-      "learning_rate": 9.107338378114277e-06,
-      "loss": 0.2581,
+      "epoch": 3.728511190398962,
+      "grad_norm": 0.288904070854187,
+      "learning_rate": 8.025631850132554e-06,
+      "loss": 0.3969,
       "step": 103455
     },
     {
-      "epoch": 3.64,
-      "learning_rate": 9.105139465334048e-06,
-      "loss": 0.2879,
+      "epoch": 3.728691390060187,
+      "grad_norm": 0.23645661771297455,
+      "learning_rate": 8.023489583971077e-06,
+      "loss": 0.3695,
       "step": 103460
     },
     {
-      "epoch": 3.64,
-      "learning_rate": 9.102940758938109e-06,
-      "loss": 0.2679,
+      "epoch": 3.728871589721411,
+      "grad_norm": 0.24535304307937622,
+      "learning_rate": 8.021347549111277e-06,
+      "loss": 0.4008,
       "step": 103465
     },
     {
-      "epoch": 3.64,
-      "learning_rate": 9.100742258955e-06,
-      "loss": 0.2528,
+      "epoch": 3.729051789382636,
+      "grad_norm": 0.25152429938316345,
+      "learning_rate": 8.019205745582337e-06,
+      "loss": 0.3894,
       "step": 103470
     },
     {
-      "epoch": 3.64,
-      "learning_rate": 9.098543965413281e-06,
-      "loss": 0.2925,
+      "epoch": 3.7292319890438606,
+      "grad_norm": 0.2640352249145508,
+      "learning_rate": 8.017064173413416e-06,
+      "loss": 0.3442,
       "step": 103475
     },
     {
-      "epoch": 3.64,
-      "learning_rate": 9.096345878341492e-06,
-      "loss": 0.2465,
+      "epoch": 3.7294121887050853,
+      "grad_norm": 0.22095341980457306,
+      "learning_rate": 8.015351082277115e-06,
+      "loss": 0.3711,
       "step": 103480
     },
     {
-      "epoch": 3.64,
-      "learning_rate": 9.094147997768168e-06,
-      "loss": 0.271,
+      "epoch": 3.72959238836631,
+      "grad_norm": 0.23479041457176208,
+      "learning_rate": 8.013209926629791e-06,
+      "loss": 0.3823,
       "step": 103485
     },
     {
-      "epoch": 3.64,
-      "learning_rate": 9.091950323721842e-06,
-      "loss": 0.2876,
+      "epoch": 3.7297725880275348,
+      "grad_norm": 0.2575128972530365,
+      "learning_rate": 8.011069002424194e-06,
+      "loss": 0.3876,
       "step": 103490
     },
     {
-      "epoch": 3.64,
-      "learning_rate": 9.089752856231069e-06,
-      "loss": 0.2557,
+      "epoch": 3.729952787688759,
+      "grad_norm": 0.25867316126823425,
+      "learning_rate": 8.00892830968949e-06,
+      "loss": 0.3505,
       "step": 103495
     },
     {
-      "epoch": 3.64,
-      "learning_rate": 9.08755559532436e-06,
-      "loss": 0.2575,
+      "epoch": 3.7301329873499838,
+      "grad_norm": 0.24454909563064575,
+      "learning_rate": 8.00678784845485e-06,
+      "loss": 0.4008,
       "step": 103500
     },
     {
-      "epoch": 3.64,
-      "eval_loss": 0.2520705461502075,
-      "eval_runtime": 10.549,
-      "eval_samples_per_second": 9.48,
-      "eval_steps_per_second": 9.48,
+      "epoch": 3.7301329873499838,
+      "eval_loss": 0.4287916421890259,
+      "eval_runtime": 3.5358,
+      "eval_samples_per_second": 28.282,
+      "eval_steps_per_second": 7.071,
       "step": 103500
     },
     {
-      "epoch": 3.64,
-      "learning_rate": 9.085358541030262e-06,
-      "loss": 0.2715,
+      "epoch": 3.7303131870112085,
+      "grad_norm": 0.2569969892501831,
+      "learning_rate": 8.004647618749424e-06,
+      "loss": 0.3734,
       "step": 103505
     },
     {
-      "epoch": 3.64,
-      "learning_rate": 9.083161693377288e-06,
-      "loss": 0.2492,
+      "epoch": 3.730493386672433,
+      "grad_norm": 0.26036691665649414,
+      "learning_rate": 8.002507620602398e-06,
+      "loss": 0.3745,
       "step": 103510
     },
     {
-      "epoch": 3.64,
-      "learning_rate": 9.080965052393974e-06,
-      "loss": 0.2546,
+      "epoch": 3.7306735863336575,
+      "grad_norm": 0.2165229618549347,
+      "learning_rate": 8.000367854042911e-06,
+      "loss": 0.3696,
       "step": 103515
     },
     {
-      "epoch": 3.64,
-      "learning_rate": 9.078768618108843e-06,
-      "loss": 0.2508,
+      "epoch": 3.7308537859948823,
+      "grad_norm": 0.20541858673095703,
+      "learning_rate": 7.99822831910012e-06,
+      "loss": 0.3753,
       "step": 103520
     },
     {
-      "epoch": 3.64,
-      "learning_rate": 9.07657239055041e-06,
-      "loss": 0.2488,
+      "epoch": 3.731033985656107,
+      "grad_norm": 0.23357829451560974,
+      "learning_rate": 7.99608901580318e-06,
+      "loss": 0.3534,
       "step": 103525
     },
     {
-      "epoch": 3.64,
-      "learning_rate": 9.074376369747176e-06,
-      "loss": 0.2609,
+      "epoch": 3.7312141853173317,
+      "grad_norm": 0.2972395420074463,
+      "learning_rate": 7.993949944181223e-06,
+      "loss": 0.3948,
       "step": 103530
     },
     {
-      "epoch": 3.64,
-      "learning_rate": 9.072180555727683e-06,
-      "loss": 0.2645,
+      "epoch": 3.7313943849785565,
+      "grad_norm": 0.22406528890132904,
+      "learning_rate": 7.991811104263417e-06,
+      "loss": 0.3747,
       "step": 103535
     },
     {
-      "epoch": 3.64,
-      "learning_rate": 9.069984948520428e-06,
-      "loss": 0.2386,
+      "epoch": 3.7315745846397808,
+      "grad_norm": 0.21077711880207062,
+      "learning_rate": 7.989672496078899e-06,
+      "loss": 0.3868,
       "step": 103540
     },
     {
-      "epoch": 3.64,
-      "learning_rate": 9.06778954815391e-06,
-      "loss": 0.2696,
+      "epoch": 3.7317547843010055,
+      "grad_norm": 0.2398710399866104,
+      "learning_rate": 7.98753411965678e-06,
+      "loss": 0.3643,
       "step": 103545
     },
     {
-      "epoch": 3.64,
-      "learning_rate": 9.065594354656653e-06,
-      "loss": 0.2695,
+      "epoch": 3.7319349839622302,
+      "grad_norm": 0.29205483198165894,
+      "learning_rate": 7.985395975026227e-06,
+      "loss": 0.3413,
       "step": 103550
     },
     {
-      "epoch": 3.64,
-      "learning_rate": 9.063399368057143e-06,
-      "loss": 0.2459,
+      "epoch": 3.7321151836234545,
+      "grad_norm": 0.24142169952392578,
+      "learning_rate": 7.983258062216361e-06,
+      "loss": 0.3876,
       "step": 103555
     },
     {
-      "epoch": 3.64,
-      "learning_rate": 9.061204588383901e-06,
-      "loss": 0.2661,
+      "epoch": 3.7322953832846792,
+      "grad_norm": 0.2126045972108841,
+      "learning_rate": 7.981120381256308e-06,
+      "loss": 0.3878,
       "step": 103560
     },
     {
-      "epoch": 3.64,
-      "learning_rate": 9.059010015665403e-06,
-      "loss": 0.275,
+      "epoch": 3.732475582945904,
+      "grad_norm": 0.21912820637226105,
+      "learning_rate": 7.978982932175195e-06,
+      "loss": 0.3782,
       "step": 103565
     },
     {
-      "epoch": 3.64,
-      "learning_rate": 9.056815649930165e-06,
-      "loss": 0.266,
+      "epoch": 3.7326557826071287,
+      "grad_norm": 0.3162360191345215,
+      "learning_rate": 7.976845715002131e-06,
+      "loss": 0.4018,
       "step": 103570
     },
     {
-      "epoch": 3.64,
-      "learning_rate": 9.054621491206667e-06,
-      "loss": 0.2611,
+      "epoch": 3.7328359822683534,
+      "grad_norm": 0.20726588368415833,
+      "learning_rate": 7.974708729766262e-06,
+      "loss": 0.3311,
       "step": 103575
     },
     {
-      "epoch": 3.64,
-      "learning_rate": 9.052427539523401e-06,
-      "loss": 0.249,
+      "epoch": 3.733016181929578,
+      "grad_norm": 0.30441704392433167,
+      "learning_rate": 7.97257197649669e-06,
+      "loss": 0.3811,
       "step": 103580
     },
     {
-      "epoch": 3.64,
-      "learning_rate": 9.050233794908844e-06,
-      "loss": 0.2448,
+      "epoch": 3.7331963815908025,
+      "grad_norm": 0.23164492845535278,
+      "learning_rate": 7.970435455222528e-06,
+      "loss": 0.3566,
       "step": 103585
     },
     {
-      "epoch": 3.64,
-      "learning_rate": 9.048040257391501e-06,
-      "loss": 0.2596,
+      "epoch": 3.733376581252027,
+      "grad_norm": 0.26288750767707825,
+      "learning_rate": 7.968299165972884e-06,
+      "loss": 0.3842,
       "step": 103590
     },
     {
-      "epoch": 3.64,
-      "learning_rate": 9.04584692699984e-06,
-      "loss": 0.2596,
+      "epoch": 3.733556780913252,
+      "grad_norm": 0.23692747950553894,
+      "learning_rate": 7.966163108776866e-06,
+      "loss": 0.3784,
       "step": 103595
     },
     {
-      "epoch": 3.64,
-      "learning_rate": 9.043653803762345e-06,
-      "loss": 0.2496,
+      "epoch": 3.7337369805744767,
+      "grad_norm": 0.19637106359004974,
+      "learning_rate": 7.964027283663573e-06,
+      "loss": 0.3741,
       "step": 103600
     },
     {
-      "epoch": 3.65,
-      "learning_rate": 9.041460887707478e-06,
-      "loss": 0.2707,
+      "epoch": 3.733917180235701,
+      "grad_norm": 0.2366388440132141,
+      "learning_rate": 7.961891690662115e-06,
+      "loss": 0.3927,
       "step": 103605
     },
     {
-      "epoch": 3.65,
-      "learning_rate": 9.039268178863727e-06,
-      "loss": 0.2813,
+      "epoch": 3.7340973798969257,
+      "grad_norm": 0.24893350899219513,
+      "learning_rate": 7.959756329801588e-06,
+      "loss": 0.3664,
       "step": 103610
     },
     {
-      "epoch": 3.65,
-      "learning_rate": 9.037075677259569e-06,
-      "loss": 0.257,
+      "epoch": 3.7342775795581504,
+      "grad_norm": 0.26749271154403687,
+      "learning_rate": 7.957621201111079e-06,
+      "loss": 0.3876,
       "step": 103615
     },
     {
-      "epoch": 3.65,
-      "learning_rate": 9.034883382923459e-06,
-      "loss": 0.2434,
+      "epoch": 3.734457779219375,
+      "grad_norm": 0.23498766124248505,
+      "learning_rate": 7.95548630461968e-06,
+      "loss": 0.3765,
       "step": 103620
     },
     {
-      "epoch": 3.65,
-      "learning_rate": 9.032691295883863e-06,
-      "loss": 0.2592,
+      "epoch": 3.7346379788806,
+      "grad_norm": 0.23466086387634277,
+      "learning_rate": 7.953351640356475e-06,
+      "loss": 0.3897,
       "step": 103625
     },
     {
-      "epoch": 3.65,
-      "learning_rate": 9.030499416169256e-06,
-      "loss": 0.2716,
+      "epoch": 3.734818178541824,
+      "grad_norm": 0.27381378412246704,
+      "learning_rate": 7.951217208350567e-06,
+      "loss": 0.3869,
       "step": 103630
     },
     {
-      "epoch": 3.65,
-      "learning_rate": 9.028307743808087e-06,
-      "loss": 0.2672,
+      "epoch": 3.734998378203049,
+      "grad_norm": 0.23438720405101776,
+      "learning_rate": 7.94908300863102e-06,
+      "loss": 0.3729,
       "step": 103635
     },
     {
-      "epoch": 3.65,
-      "learning_rate": 9.026116278828816e-06,
-      "loss": 0.2589,
+      "epoch": 3.7351785778642737,
+      "grad_norm": 0.2471228539943695,
+      "learning_rate": 7.946949041226904e-06,
+      "loss": 0.4086,
       "step": 103640
     },
     {
-      "epoch": 3.65,
-      "learning_rate": 9.023925021259894e-06,
-      "loss": 0.244,
+      "epoch": 3.7353587775254984,
+      "grad_norm": 0.2612572908401489,
+      "learning_rate": 7.944815306167314e-06,
+      "loss": 0.3758,
       "step": 103645
     },
     {
-      "epoch": 3.65,
-      "learning_rate": 9.021733971129782e-06,
-      "loss": 0.2679,
+      "epoch": 3.7355389771867227,
+      "grad_norm": 0.31286609172821045,
+      "learning_rate": 7.942681803481309e-06,
+      "loss": 0.3795,
       "step": 103650
     },
     {
-      "epoch": 3.65,
-      "learning_rate": 9.019543128466922e-06,
-      "loss": 0.2594,
+      "epoch": 3.7357191768479474,
+      "grad_norm": 0.2046850472688675,
+      "learning_rate": 7.940548533197973e-06,
+      "loss": 0.3579,
       "step": 103655
     },
     {
-      "epoch": 3.65,
-      "learning_rate": 9.017352493299757e-06,
-      "loss": 0.2675,
+      "epoch": 3.735899376509172,
+      "grad_norm": 0.22326718270778656,
+      "learning_rate": 7.938415495346354e-06,
+      "loss": 0.3635,
       "step": 103660
     },
     {
-      "epoch": 3.65,
-      "learning_rate": 9.01560013458205e-06,
-      "loss": 0.2589,
+      "epoch": 3.736079576170397,
+      "grad_norm": 0.27218303084373474,
+      "learning_rate": 7.936282689955515e-06,
+      "loss": 0.3704,
       "step": 103665
     },
     {
-      "epoch": 3.65,
-      "learning_rate": 9.013409872978834e-06,
-      "loss": 0.2636,
+      "epoch": 3.7362597758316216,
+      "grad_norm": 0.2567422688007355,
+      "learning_rate": 7.934150117054528e-06,
+      "loss": 0.3358,
       "step": 103670
     },
     {
-      "epoch": 3.65,
-      "learning_rate": 9.011219818950945e-06,
-      "loss": 0.2685,
+      "epoch": 3.7364399754928463,
+      "grad_norm": 0.21654558181762695,
+      "learning_rate": 7.93201777667244e-06,
+      "loss": 0.3797,
       "step": 103675
     },
     {
-      "epoch": 3.65,
-      "learning_rate": 9.009029972526839e-06,
-      "loss": 0.2679,
+      "epoch": 3.7366201751540706,
+      "grad_norm": 0.26781266927719116,
+      "learning_rate": 7.929885668838305e-06,
+      "loss": 0.3711,
       "step": 103680
     },
     {
-      "epoch": 3.65,
-      "learning_rate": 9.00684033373494e-06,
-      "loss": 0.2409,
+      "epoch": 3.7368003748152954,
+      "grad_norm": 0.23572920262813568,
+      "learning_rate": 7.927753793581174e-06,
+      "loss": 0.3858,
       "step": 103685
     },
     {
-      "epoch": 3.65,
-      "learning_rate": 9.004650902603679e-06,
-      "loss": 0.2331,
+      "epoch": 3.73698057447652,
+      "grad_norm": 0.2272387593984604,
+      "learning_rate": 7.925622150930085e-06,
+      "loss": 0.4051,
       "step": 103690
     },
     {
-      "epoch": 3.65,
-      "learning_rate": 9.002461679161475e-06,
-      "loss": 0.2469,
+      "epoch": 3.7371607741377444,
+      "grad_norm": 0.24096395075321198,
+      "learning_rate": 7.923490740914097e-06,
+      "loss": 0.3699,
       "step": 103695
     },
     {
-      "epoch": 3.65,
-      "learning_rate": 9.000272663436765e-06,
-      "loss": 0.2537,
+      "epoch": 3.737340973798969,
+      "grad_norm": 0.2557925581932068,
+      "learning_rate": 7.92135956356224e-06,
+      "loss": 0.4079,
       "step": 103700
     },
     {
-      "epoch": 3.65,
-      "learning_rate": 8.998083855457979e-06,
-      "loss": 0.247,
+      "epoch": 3.737521173460194,
+      "grad_norm": 0.2199949324131012,
+      "learning_rate": 7.919228618903551e-06,
+      "loss": 0.3751,
       "step": 103705
     },
     {
-      "epoch": 3.65,
-      "learning_rate": 8.995895255253529e-06,
-      "loss": 0.2679,
+      "epoch": 3.7377013731214186,
+      "grad_norm": 0.2655465602874756,
+      "learning_rate": 7.91709790696707e-06,
+      "loss": 0.4015,
       "step": 103710
     },
     {
-      "epoch": 3.65,
-      "learning_rate": 8.99370686285182e-06,
-      "loss": 0.2445,
+      "epoch": 3.7378815727826433,
+      "grad_norm": 0.25117290019989014,
+      "learning_rate": 7.91496742778181e-06,
+      "loss": 0.3899,
       "step": 103715
     },
     {
-      "epoch": 3.65,
-      "learning_rate": 8.991518678281292e-06,
-      "loss": 0.2439,
+      "epoch": 3.738061772443868,
+      "grad_norm": 0.30154553055763245,
+      "learning_rate": 7.91283718137682e-06,
+      "loss": 0.424,
       "step": 103720
     },
     {
-      "epoch": 3.65,
-      "learning_rate": 8.989330701570339e-06,
-      "loss": 0.2615,
+      "epoch": 3.7382419721050923,
+      "grad_norm": 0.27134832739830017,
+      "learning_rate": 7.910707167781126e-06,
+      "loss": 0.3404,
       "step": 103725
     },
     {
-      "epoch": 3.65,
-      "learning_rate": 8.98714293274737e-06,
-      "loss": 0.2648,
+      "epoch": 3.738422171766317,
+      "grad_norm": 0.2242002934217453,
+      "learning_rate": 7.908577387023719e-06,
+      "loss": 0.38,
       "step": 103730
     },
     {
-      "epoch": 3.65,
-      "learning_rate": 8.984955371840806e-06,
-      "loss": 0.2536,
+      "epoch": 3.738602371427542,
+      "grad_norm": 0.2798958122730255,
+      "learning_rate": 7.906447839133643e-06,
+      "loss": 0.3884,
       "step": 103735
     },
     {
-      "epoch": 3.65,
-      "learning_rate": 8.98276801887904e-06,
-      "loss": 0.2468,
+      "epoch": 3.738782571088766,
+      "grad_norm": 0.2313423752784729,
+      "learning_rate": 7.90431852413991e-06,
+      "loss": 0.3852,
       "step": 103740
     },
     {
-      "epoch": 3.65,
-      "learning_rate": 8.980580873890477e-06,
-      "loss": 0.2553,
+      "epoch": 3.738962770749991,
+      "grad_norm": 0.24077194929122925,
+      "learning_rate": 7.902189442071525e-06,
+      "loss": 0.3847,
       "step": 103745
     },
     {
-      "epoch": 3.65,
-      "learning_rate": 8.978393936903507e-06,
-      "loss": 0.2512,
+      "epoch": 3.7391429704112156,
+      "grad_norm": 0.24760626256465912,
+      "learning_rate": 7.900060592957498e-06,
+      "loss": 0.3997,
       "step": 103750
     },
     {
-      "epoch": 3.65,
-      "learning_rate": 8.97620720794653e-06,
-      "loss": 0.2637,
+      "epoch": 3.7393231700724403,
+      "grad_norm": 0.19511263072490692,
+      "learning_rate": 7.897931976826827e-06,
+      "loss": 0.3953,
       "step": 103755
     },
     {
-      "epoch": 3.65,
-      "learning_rate": 8.974020687047955e-06,
-      "loss": 0.2473,
+      "epoch": 3.739503369733665,
+      "grad_norm": 0.1964816004037857,
+      "learning_rate": 7.895803593708532e-06,
+      "loss": 0.3628,
       "step": 103760
     },
     {
-      "epoch": 3.65,
-      "learning_rate": 8.971834374236155e-06,
-      "loss": 0.2655,
+      "epoch": 3.7396835693948898,
+      "grad_norm": 0.2270769327878952,
+      "learning_rate": 7.893675443631599e-06,
+      "loss": 0.3912,
       "step": 103765
     },
     {
-      "epoch": 3.65,
-      "learning_rate": 8.969648269539515e-06,
-      "loss": 0.2595,
+      "epoch": 3.739863769056114,
+      "grad_norm": 0.21808400750160217,
+      "learning_rate": 7.891547526625026e-06,
+      "loss": 0.3762,
       "step": 103770
     },
     {
-      "epoch": 3.65,
-      "learning_rate": 8.96746237298644e-06,
-      "loss": 0.2466,
+      "epoch": 3.740043968717339,
+      "grad_norm": 0.29697415232658386,
+      "learning_rate": 7.889419842717807e-06,
+      "loss": 0.364,
       "step": 103775
     },
     {
-      "epoch": 3.65,
-      "learning_rate": 8.965276684605297e-06,
-      "loss": 0.2621,
+      "epoch": 3.7402241683785635,
+      "grad_norm": 0.26142972707748413,
+      "learning_rate": 7.88729239193893e-06,
+      "loss": 0.3906,
       "step": 103780
     },
     {
-      "epoch": 3.65,
-      "learning_rate": 8.96309120442447e-06,
-      "loss": 0.2389,
+      "epoch": 3.740404368039788,
+      "grad_norm": 0.3567783534526825,
+      "learning_rate": 7.885165174317374e-06,
+      "loss": 0.3906,
       "step": 103785
     },
     {
-      "epoch": 3.65,
-      "learning_rate": 8.960905932472327e-06,
-      "loss": 0.2661,
+      "epoch": 3.7405845677010126,
+      "grad_norm": 0.2854664921760559,
+      "learning_rate": 7.883038189882137e-06,
+      "loss": 0.4332,
       "step": 103790
     },
     {
-      "epoch": 3.65,
-      "learning_rate": 8.95872086877726e-06,
-      "loss": 0.274,
+      "epoch": 3.7407647673622373,
+      "grad_norm": 0.2237059473991394,
+      "learning_rate": 7.880911438662186e-06,
+      "loss": 0.38,
       "step": 103795
     },
     {
-      "epoch": 3.65,
-      "learning_rate": 8.956536013367628e-06,
-      "loss": 0.2816,
+      "epoch": 3.740944967023462,
+      "grad_norm": 0.1987508237361908,
+      "learning_rate": 7.878784920686509e-06,
+      "loss": 0.4177,
       "step": 103800
     },
     {
-      "epoch": 3.65,
-      "learning_rate": 8.954351366271796e-06,
-      "loss": 0.2454,
+      "epoch": 3.7411251666846868,
+      "grad_norm": 0.2105659693479538,
+      "learning_rate": 7.876658635984065e-06,
+      "loss": 0.3688,
       "step": 103805
     },
     {
-      "epoch": 3.65,
-      "learning_rate": 8.952166927518146e-06,
-      "loss": 0.2581,
+      "epoch": 3.7413053663459115,
+      "grad_norm": 0.2706106901168823,
+      "learning_rate": 7.87453258458383e-06,
+      "loss": 0.3544,
       "step": 103810
     },
     {
-      "epoch": 3.65,
-      "learning_rate": 8.949982697135024e-06,
-      "loss": 0.2761,
+      "epoch": 3.7414855660071358,
+      "grad_norm": 0.2130729705095291,
+      "learning_rate": 7.872406766514779e-06,
+      "loss": 0.4017,
       "step": 103815
     },
     {
-      "epoch": 3.65,
-      "learning_rate": 8.947798675150806e-06,
-      "loss": 0.2835,
+      "epoch": 3.7416657656683605,
+      "grad_norm": 0.191060870885849,
+      "learning_rate": 7.870281181805877e-06,
+      "loss": 0.3528,
       "step": 103820
     },
     {
-      "epoch": 3.65,
-      "learning_rate": 8.945614861593843e-06,
-      "loss": 0.2625,
+      "epoch": 3.7418459653295852,
+      "grad_norm": 0.26372456550598145,
+      "learning_rate": 7.868155830486063e-06,
+      "loss": 0.363,
       "step": 103825
     },
     {
-      "epoch": 3.65,
-      "learning_rate": 8.943431256492484e-06,
-      "loss": 0.2508,
+      "epoch": 3.74202616499081,
+      "grad_norm": 0.2443712055683136,
+      "learning_rate": 7.866030712584318e-06,
+      "loss": 0.3847,
       "step": 103830
     },
     {
-      "epoch": 3.65,
-      "learning_rate": 8.941247859875095e-06,
-      "loss": 0.2554,
+      "epoch": 3.7422063646520343,
+      "grad_norm": 0.30589404702186584,
+      "learning_rate": 7.863905828129577e-06,
+      "loss": 0.3667,
       "step": 103835
     },
     {
-      "epoch": 3.65,
-      "learning_rate": 8.939064671770023e-06,
-      "loss": 0.2452,
+      "epoch": 3.742386564313259,
+      "grad_norm": 0.3021865785121918,
+      "learning_rate": 7.861781177150818e-06,
+      "loss": 0.4225,
       "step": 103840
     },
     {
-      "epoch": 3.65,
-      "learning_rate": 8.936881692205601e-06,
-      "loss": 0.2538,
+      "epoch": 3.7425667639744837,
+      "grad_norm": 0.21095475554466248,
+      "learning_rate": 7.859656759676964e-06,
+      "loss": 0.3872,
       "step": 103845
     },
     {
-      "epoch": 3.65,
-      "learning_rate": 8.934698921210195e-06,
-      "loss": 0.2836,
+      "epoch": 3.7427469636357085,
+      "grad_norm": 0.25870993733406067,
+      "learning_rate": 7.857532575736961e-06,
+      "loss": 0.3782,
       "step": 103850
     },
     {
-      "epoch": 3.65,
-      "learning_rate": 8.932516358812132e-06,
-      "loss": 0.2783,
+      "epoch": 3.742927163296933,
+      "grad_norm": 0.2628803849220276,
+      "learning_rate": 7.855408625359768e-06,
+      "loss": 0.3724,
       "step": 103855
     },
     {
-      "epoch": 3.65,
-      "learning_rate": 8.93033400503975e-06,
-      "loss": 0.2501,
+      "epoch": 3.7431073629581575,
+      "grad_norm": 0.2707422375679016,
+      "learning_rate": 7.85328490857431e-06,
+      "loss": 0.3451,
       "step": 103860
     },
     {
-      "epoch": 3.65,
-      "learning_rate": 8.928151859921397e-06,
-      "loss": 0.2647,
+      "epoch": 3.7432875626193822,
+      "grad_norm": 0.21672846376895905,
+      "learning_rate": 7.851161425409525e-06,
+      "loss": 0.3629,
       "step": 103865
     },
     {
-      "epoch": 3.65,
-      "learning_rate": 8.925969923485391e-06,
-      "loss": 0.2741,
+      "epoch": 3.743467762280607,
+      "grad_norm": 0.2119687795639038,
+      "learning_rate": 7.849038175894346e-06,
+      "loss": 0.3829,
       "step": 103870
     },
     {
-      "epoch": 3.65,
-      "learning_rate": 8.923788195760086e-06,
-      "loss": 0.2295,
+      "epoch": 3.7436479619418317,
+      "grad_norm": 0.3262704610824585,
+      "learning_rate": 7.846915160057694e-06,
+      "loss": 0.3918,
       "step": 103875
     },
     {
-      "epoch": 3.65,
-      "learning_rate": 8.921606676773791e-06,
-      "loss": 0.2456,
+      "epoch": 3.743828161603056,
+      "grad_norm": 0.2228257954120636,
+      "learning_rate": 7.84479237792851e-06,
+      "loss": 0.3837,
       "step": 103880
     },
     {
-      "epoch": 3.65,
-      "learning_rate": 8.919425366554832e-06,
-      "loss": 0.2479,
+      "epoch": 3.7440083612642807,
+      "grad_norm": 0.24818168580532074,
+      "learning_rate": 7.842669829535709e-06,
+      "loss": 0.3801,
       "step": 103885
     },
     {
-      "epoch": 3.66,
-      "learning_rate": 8.917244265131543e-06,
-      "loss": 0.2524,
+      "epoch": 3.7441885609255054,
+      "grad_norm": 0.18585266172885895,
+      "learning_rate": 7.840547514908209e-06,
+      "loss": 0.3744,
       "step": 103890
     },
     {
-      "epoch": 3.66,
-      "learning_rate": 8.915063372532242e-06,
-      "loss": 0.2658,
+      "epoch": 3.74436876058673,
+      "grad_norm": 0.25262919068336487,
+      "learning_rate": 7.838425434074925e-06,
+      "loss": 0.3952,
       "step": 103895
     },
     {
-      "epoch": 3.66,
-      "learning_rate": 8.91288268878524e-06,
-      "loss": 0.28,
+      "epoch": 3.744548960247955,
+      "grad_norm": 0.21435338258743286,
+      "learning_rate": 7.836303587064766e-06,
+      "loss": 0.3595,
       "step": 103900
     },
     {
-      "epoch": 3.66,
-      "learning_rate": 8.910702213918848e-06,
-      "loss": 0.2514,
+      "epoch": 3.744729159909179,
+      "grad_norm": 0.23021239042282104,
+      "learning_rate": 7.834181973906656e-06,
+      "loss": 0.4104,
       "step": 103905
     },
     {
-      "epoch": 3.66,
-      "learning_rate": 8.90852194796139e-06,
-      "loss": 0.2709,
+      "epoch": 3.744909359570404,
+      "grad_norm": 0.2677323818206787,
+      "learning_rate": 7.832060594629504e-06,
+      "loss": 0.3823,
       "step": 103910
     },
     {
-      "epoch": 3.66,
-      "learning_rate": 8.906341890941164e-06,
-      "loss": 0.2609,
+      "epoch": 3.7450895592316287,
+      "grad_norm": 0.22377659380435944,
+      "learning_rate": 7.82993944926218e-06,
+      "loss": 0.3476,
       "step": 103915
     },
     {
-      "epoch": 3.66,
-      "learning_rate": 8.904162042886487e-06,
-      "loss": 0.2631,
+      "epoch": 3.7452697588928534,
+      "grad_norm": 0.2559918761253357,
+      "learning_rate": 7.82781853783362e-06,
+      "loss": 0.3891,
       "step": 103920
     },
     {
-      "epoch": 3.66,
-      "learning_rate": 8.901982403825654e-06,
-      "loss": 0.2406,
+      "epoch": 3.7454499585540777,
+      "grad_norm": 0.201356440782547,
+      "learning_rate": 7.825697860372705e-06,
+      "loss": 0.3572,
       "step": 103925
     },
     {
-      "epoch": 3.66,
-      "learning_rate": 8.899802973786977e-06,
-      "loss": 0.2491,
+      "epoch": 3.7456301582153024,
+      "grad_norm": 0.26732808351516724,
+      "learning_rate": 7.823577416908325e-06,
+      "loss": 0.3886,
       "step": 103930
     },
     {
-      "epoch": 3.66,
-      "learning_rate": 8.89762375279875e-06,
-      "loss": 0.2554,
+      "epoch": 3.745810357876527,
+      "grad_norm": 0.22887802124023438,
+      "learning_rate": 7.821457207469392e-06,
+      "loss": 0.378,
       "step": 103935
     },
     {
-      "epoch": 3.66,
-      "learning_rate": 8.895444740889267e-06,
-      "loss": 0.2572,
+      "epoch": 3.745990557537752,
+      "grad_norm": 0.26050490140914917,
+      "learning_rate": 7.819337232084764e-06,
+      "loss": 0.3839,
       "step": 103940
     },
     {
-      "epoch": 3.66,
-      "learning_rate": 8.893265938086808e-06,
-      "loss": 0.2622,
+      "epoch": 3.7461707571989766,
+      "grad_norm": 0.23908868432044983,
+      "learning_rate": 7.817217490783346e-06,
+      "loss": 0.4197,
       "step": 103945
     },
     {
-      "epoch": 3.66,
-      "learning_rate": 8.891087344419688e-06,
-      "loss": 0.2487,
+      "epoch": 3.7463509568602014,
+      "grad_norm": 0.23192131519317627,
+      "learning_rate": 7.815097983594016e-06,
+      "loss": 0.3885,
       "step": 103950
     },
     {
-      "epoch": 3.66,
-      "learning_rate": 8.888908959916182e-06,
-      "loss": 0.2296,
+      "epoch": 3.7465311565214257,
+      "grad_norm": 0.31525352597236633,
+      "learning_rate": 7.812978710545646e-06,
+      "loss": 0.3664,
       "step": 103955
     },
     {
-      "epoch": 3.66,
-      "learning_rate": 8.886730784604565e-06,
-      "loss": 0.2806,
+      "epoch": 3.7467113561826504,
+      "grad_norm": 0.2983095347881317,
+      "learning_rate": 7.810859671667118e-06,
+      "loss": 0.4093,
       "step": 103960
     },
     {
-      "epoch": 3.66,
-      "learning_rate": 8.884552818513143e-06,
-      "loss": 0.2633,
+      "epoch": 3.746891555843875,
+      "grad_norm": 0.30987173318862915,
+      "learning_rate": 7.808740866987293e-06,
+      "loss": 0.4,
       "step": 103965
     },
     {
-      "epoch": 3.66,
-      "learning_rate": 8.882375061670178e-06,
-      "loss": 0.2681,
+      "epoch": 3.7470717555050994,
+      "grad_norm": 0.2129722535610199,
+      "learning_rate": 7.80662229653504e-06,
+      "loss": 0.3731,
       "step": 103970
     },
     {
-      "epoch": 3.66,
-      "learning_rate": 8.880197514103944e-06,
-      "loss": 0.2587,
+      "epoch": 3.747251955166324,
+      "grad_norm": 0.2828996181488037,
+      "learning_rate": 7.804503960339238e-06,
+      "loss": 0.3715,
       "step": 103975
     },
     {
-      "epoch": 3.66,
-      "learning_rate": 8.878020175842722e-06,
-      "loss": 0.2471,
+      "epoch": 3.747432154827549,
+      "grad_norm": 0.26484811305999756,
+      "learning_rate": 7.80238585842874e-06,
+      "loss": 0.3518,
       "step": 103980
     },
     {
-      "epoch": 3.66,
-      "learning_rate": 8.875843046914792e-06,
-      "loss": 0.2488,
+      "epoch": 3.7476123544887736,
+      "grad_norm": 0.23784774541854858,
+      "learning_rate": 7.800267990832404e-06,
+      "loss": 0.4227,
       "step": 103985
     },
     {
-      "epoch": 3.66,
-      "learning_rate": 8.873666127348412e-06,
-      "loss": 0.258,
+      "epoch": 3.7477925541499983,
+      "grad_norm": 0.22352570295333862,
+      "learning_rate": 7.79815035757909e-06,
+      "loss": 0.3722,
       "step": 103990
     },
     {
-      "epoch": 3.66,
-      "learning_rate": 8.871489417171853e-06,
-      "loss": 0.2769,
+      "epoch": 3.747972753811223,
+      "grad_norm": 0.1883036494255066,
+      "learning_rate": 7.796032958697635e-06,
+      "loss": 0.3852,
       "step": 103995
     },
     {
-      "epoch": 3.66,
-      "learning_rate": 8.869312916413363e-06,
-      "loss": 0.2779,
+      "epoch": 3.7481529534724474,
+      "grad_norm": 0.29498735070228577,
+      "learning_rate": 7.79391579421691e-06,
+      "loss": 0.3987,
       "step": 104000
     },
     {
-      "epoch": 3.66,
-      "eval_loss": 0.25182366371154785,
-      "eval_runtime": 10.5433,
-      "eval_samples_per_second": 9.485,
-      "eval_steps_per_second": 9.485,
+      "epoch": 3.7481529534724474,
+      "eval_loss": 0.4285476803779602,
+      "eval_runtime": 3.5328,
+      "eval_samples_per_second": 28.306,
+      "eval_steps_per_second": 7.076,
       "step": 104000
     },
     {
-      "epoch": 3.66,
-      "learning_rate": 8.867136625101225e-06,
-      "loss": 0.2541,
+      "epoch": 3.748333153133672,
+      "grad_norm": 0.2650872766971588,
+      "learning_rate": 7.79179886416576e-06,
+      "loss": 0.3802,
       "step": 104005
     },
     {
-      "epoch": 3.66,
-      "learning_rate": 8.864960543263684e-06,
-      "loss": 0.2648,
+      "epoch": 3.748513352794897,
+      "grad_norm": 0.19966596364974976,
+      "learning_rate": 7.789682168573004e-06,
+      "loss": 0.3448,
       "step": 104010
     },
     {
-      "epoch": 3.66,
-      "learning_rate": 8.862784670928998e-06,
-      "loss": 0.2607,
+      "epoch": 3.748693552456121,
+      "grad_norm": 0.2707030177116394,
+      "learning_rate": 7.7875657074675e-06,
+      "loss": 0.3787,
       "step": 104015
     },
     {
-      "epoch": 3.66,
-      "learning_rate": 8.860609008125408e-06,
-      "loss": 0.2858,
+      "epoch": 3.748873752117346,
+      "grad_norm": 0.21822407841682434,
+      "learning_rate": 7.785449480878076e-06,
+      "loss": 0.4043,
       "step": 104020
     },
     {
-      "epoch": 3.66,
-      "learning_rate": 8.858433554881185e-06,
-      "loss": 0.2665,
+      "epoch": 3.7490539517785706,
+      "grad_norm": 0.23433195054531097,
+      "learning_rate": 7.783333488833585e-06,
+      "loss": 0.3704,
       "step": 104025
     },
     {
-      "epoch": 3.66,
-      "learning_rate": 8.856258311224556e-06,
-      "loss": 0.2776,
+      "epoch": 3.7492341514397953,
+      "grad_norm": 0.26820576190948486,
+      "learning_rate": 7.781217731362834e-06,
+      "loss": 0.372,
       "step": 104030
     },
     {
-      "epoch": 3.66,
-      "learning_rate": 8.85408327718378e-06,
-      "loss": 0.2577,
+      "epoch": 3.74941435110102,
+      "grad_norm": 0.27072396874427795,
+      "learning_rate": 7.779102208494648e-06,
+      "loss": 0.3665,
       "step": 104035
     },
     {
-      "epoch": 3.66,
-      "learning_rate": 8.851908452787086e-06,
-      "loss": 0.2582,
+      "epoch": 3.749594550762245,
+      "grad_norm": 0.20719735324382782,
+      "learning_rate": 7.776986920257873e-06,
+      "loss": 0.3721,
       "step": 104040
     },
     {
-      "epoch": 3.66,
-      "learning_rate": 8.849733838062724e-06,
-      "loss": 0.2478,
+      "epoch": 3.749774750423469,
+      "grad_norm": 0.2387949824333191,
+      "learning_rate": 7.774871866681313e-06,
+      "loss": 0.3456,
       "step": 104045
     },
     {
-      "epoch": 3.66,
-      "learning_rate": 8.847559433038927e-06,
-      "loss": 0.2542,
+      "epoch": 3.749954950084694,
+      "grad_norm": 0.23849116265773773,
+      "learning_rate": 7.772757047793792e-06,
+      "loss": 0.3932,
       "step": 104050
     },
     {
-      "epoch": 3.66,
-      "learning_rate": 8.845385237743924e-06,
-      "loss": 0.2527,
+      "epoch": 3.7501351497459186,
+      "grad_norm": 0.25094231963157654,
+      "learning_rate": 7.770642463624117e-06,
+      "loss": 0.3632,
       "step": 104055
     },
     {
-      "epoch": 3.66,
-      "learning_rate": 8.843211252205944e-06,
-      "loss": 0.2481,
+      "epoch": 3.750315349407143,
+      "grad_norm": 0.21101891994476318,
+      "learning_rate": 7.768528114201095e-06,
+      "loss": 0.3791,
       "step": 104060
     },
     {
-      "epoch": 3.66,
-      "learning_rate": 8.84103747645322e-06,
-      "loss": 0.2688,
+      "epoch": 3.7504955490683676,
+      "grad_norm": 0.2257104218006134,
+      "learning_rate": 7.766413999553552e-06,
+      "loss": 0.3644,
       "step": 104065
     },
     {
-      "epoch": 3.66,
-      "learning_rate": 8.83886391051398e-06,
-      "loss": 0.2292,
+      "epoch": 3.7506757487295923,
+      "grad_norm": 0.28515586256980896,
+      "learning_rate": 7.76430011971028e-06,
+      "loss": 0.3604,
       "step": 104070
     },
     {
-      "epoch": 3.66,
-      "learning_rate": 8.836690554416433e-06,
-      "loss": 0.2643,
+      "epoch": 3.750855948390817,
+      "grad_norm": 0.23402278125286102,
+      "learning_rate": 7.762186474700084e-06,
+      "loss": 0.3458,
       "step": 104075
     },
     {
-      "epoch": 3.66,
-      "learning_rate": 8.834517408188814e-06,
-      "loss": 0.2564,
+      "epoch": 3.7510361480520418,
+      "grad_norm": 0.239648699760437,
+      "learning_rate": 7.760073064551757e-06,
+      "loss": 0.3902,
       "step": 104080
     },
     {
-      "epoch": 3.66,
-      "learning_rate": 8.832344471859324e-06,
-      "loss": 0.2784,
+      "epoch": 3.7512163477132665,
+      "grad_norm": 0.19721876084804535,
+      "learning_rate": 7.75795988929409e-06,
+      "loss": 0.3715,
       "step": 104085
     },
     {
-      "epoch": 3.66,
-      "learning_rate": 8.830171745456195e-06,
-      "loss": 0.2447,
+      "epoch": 3.751396547374491,
+      "grad_norm": 0.29029005765914917,
+      "learning_rate": 7.755846948955889e-06,
+      "loss": 0.3953,
       "step": 104090
     },
     {
-      "epoch": 3.66,
-      "learning_rate": 8.827999229007622e-06,
-      "loss": 0.2457,
+      "epoch": 3.7515767470357155,
+      "grad_norm": 0.23127782344818115,
+      "learning_rate": 7.753734243565935e-06,
+      "loss": 0.3737,
       "step": 104095
     },
     {
-      "epoch": 3.66,
-      "learning_rate": 8.825826922541827e-06,
-      "loss": 0.2656,
+      "epoch": 3.7517569466969403,
+      "grad_norm": 0.2870927155017853,
+      "learning_rate": 7.751621773153014e-06,
+      "loss": 0.415,
       "step": 104100
     },
     {
-      "epoch": 3.66,
-      "learning_rate": 8.823654826087014e-06,
-      "loss": 0.2321,
+      "epoch": 3.751937146358165,
+      "grad_norm": 0.2417393922805786,
+      "learning_rate": 7.749509537745906e-06,
+      "loss": 0.3357,
       "step": 104105
     },
     {
-      "epoch": 3.66,
-      "learning_rate": 8.82148293967138e-06,
-      "loss": 0.251,
+      "epoch": 3.7521173460193893,
+      "grad_norm": 0.226087749004364,
+      "learning_rate": 7.74739753737339e-06,
+      "loss": 0.3768,
       "step": 104110
     },
     {
-      "epoch": 3.66,
-      "learning_rate": 8.819311263323121e-06,
-      "loss": 0.27,
+      "epoch": 3.752297545680614,
+      "grad_norm": 0.25405430793762207,
+      "learning_rate": 7.745285772064237e-06,
+      "loss": 0.3824,
       "step": 104115
     },
     {
-      "epoch": 3.66,
-      "learning_rate": 8.817139797070447e-06,
-      "loss": 0.2412,
+      "epoch": 3.7524777453418388,
+      "grad_norm": 0.24911725521087646,
+      "learning_rate": 7.743174241847237e-06,
+      "loss": 0.3721,
       "step": 104120
     },
     {
-      "epoch": 3.66,
-      "learning_rate": 8.814968540941551e-06,
-      "loss": 0.2398,
+      "epoch": 3.7526579450030635,
+      "grad_norm": 0.2514585852622986,
+      "learning_rate": 7.741062946751135e-06,
+      "loss": 0.4184,
       "step": 104125
     },
     {
-      "epoch": 3.66,
-      "learning_rate": 8.812797494964622e-06,
-      "loss": 0.2572,
+      "epoch": 3.7528381446642882,
+      "grad_norm": 0.2331729680299759,
+      "learning_rate": 7.738951886804713e-06,
+      "loss": 0.4288,
       "step": 104130
     },
     {
-      "epoch": 3.66,
-      "learning_rate": 8.810626659167842e-06,
-      "loss": 0.2635,
+      "epoch": 3.7530183443255125,
+      "grad_norm": 0.2459155172109604,
+      "learning_rate": 7.736841062036731e-06,
+      "loss": 0.3804,
       "step": 104135
     },
     {
-      "epoch": 3.66,
-      "learning_rate": 8.808456033579405e-06,
-      "loss": 0.2611,
+      "epoch": 3.7531985439867372,
+      "grad_norm": 0.2611176073551178,
+      "learning_rate": 7.734730472475948e-06,
+      "loss": 0.3821,
       "step": 104140
     },
     {
-      "epoch": 3.66,
-      "learning_rate": 8.806285618227503e-06,
-      "loss": 0.2458,
+      "epoch": 3.753378743647962,
+      "grad_norm": 0.29315176606178284,
+      "learning_rate": 7.73262011815112e-06,
+      "loss": 0.3522,
       "step": 104145
     },
     {
-      "epoch": 3.66,
-      "learning_rate": 8.80411541314031e-06,
-      "loss": 0.2607,
+      "epoch": 3.7535589433091867,
+      "grad_norm": 0.20626281201839447,
+      "learning_rate": 7.730509999090999e-06,
+      "loss": 0.3707,
       "step": 104150
     },
     {
-      "epoch": 3.66,
-      "learning_rate": 8.801945418345997e-06,
-      "loss": 0.2331,
+      "epoch": 3.753739142970411,
+      "grad_norm": 0.19073475897312164,
+      "learning_rate": 7.72840011532433e-06,
+      "loss": 0.3713,
       "step": 104155
     },
     {
-      "epoch": 3.66,
-      "learning_rate": 8.799775633872756e-06,
-      "loss": 0.2686,
+      "epoch": 3.7539193426316357,
+      "grad_norm": 0.2597355246543884,
+      "learning_rate": 7.726290466879872e-06,
+      "loss": 0.3736,
       "step": 104160
     },
     {
-      "epoch": 3.66,
-      "learning_rate": 8.797606059748754e-06,
-      "loss": 0.2482,
+      "epoch": 3.7540995422928605,
+      "grad_norm": 0.19545309245586395,
+      "learning_rate": 7.724181053786361e-06,
+      "loss": 0.3695,
       "step": 104165
     },
     {
-      "epoch": 3.66,
-      "learning_rate": 8.795436696002157e-06,
-      "loss": 0.2513,
+      "epoch": 3.754279741954085,
+      "grad_norm": 0.2396879643201828,
+      "learning_rate": 7.722071876072538e-06,
+      "loss": 0.4125,
       "step": 104170
     },
     {
-      "epoch": 3.67,
-      "learning_rate": 8.79326754266113e-06,
-      "loss": 0.2615,
+      "epoch": 3.75445994161531,
+      "grad_norm": 0.22774870693683624,
+      "learning_rate": 7.71996293376714e-06,
+      "loss": 0.3802,
       "step": 104175
     },
     {
-      "epoch": 3.67,
-      "learning_rate": 8.791098599753848e-06,
-      "loss": 0.2588,
+      "epoch": 3.7546401412765347,
+      "grad_norm": 0.20402881503105164,
+      "learning_rate": 7.717854226898897e-06,
+      "loss": 0.34,
       "step": 104180
     },
     {
-      "epoch": 3.67,
-      "learning_rate": 8.78892986730847e-06,
-      "loss": 0.2491,
+      "epoch": 3.754820340937759,
+      "grad_norm": 0.22189156711101532,
+      "learning_rate": 7.71574575549655e-06,
+      "loss": 0.3903,
       "step": 104185
     },
     {
-      "epoch": 3.67,
-      "learning_rate": 8.786761345353148e-06,
-      "loss": 0.2622,
+      "epoch": 3.7550005405989837,
+      "grad_norm": 0.19797289371490479,
+      "learning_rate": 7.713637519588829e-06,
+      "loss": 0.3864,
       "step": 104190
     },
     {
-      "epoch": 3.67,
-      "learning_rate": 8.784593033916041e-06,
-      "loss": 0.2478,
+      "epoch": 3.7551807402602084,
+      "grad_norm": 0.24208073318004608,
+      "learning_rate": 7.711529519204435e-06,
+      "loss": 0.3533,
       "step": 104195
     },
     {
-      "epoch": 3.67,
-      "learning_rate": 8.782424933025317e-06,
-      "loss": 0.2468,
+      "epoch": 3.7553609399214327,
+      "grad_norm": 0.24616724252700806,
+      "learning_rate": 7.709421754372112e-06,
+      "loss": 0.3659,
       "step": 104200
     },
     {
-      "epoch": 3.67,
-      "learning_rate": 8.78025704270912e-06,
-      "loss": 0.2527,
+      "epoch": 3.7555411395826575,
+      "grad_norm": 0.2978845238685608,
+      "learning_rate": 7.70731422512056e-06,
+      "loss": 0.4112,
       "step": 104205
     },
     {
-      "epoch": 3.67,
-      "learning_rate": 8.778089362995585e-06,
-      "loss": 0.2665,
+      "epoch": 3.755721339243882,
+      "grad_norm": 0.22275815904140472,
+      "learning_rate": 7.705206931478512e-06,
+      "loss": 0.3825,
       "step": 104210
     },
     {
-      "epoch": 3.67,
-      "learning_rate": 8.775921893912878e-06,
-      "loss": 0.2505,
+      "epoch": 3.755901538905107,
+      "grad_norm": 0.3553224205970764,
+      "learning_rate": 7.703099873474681e-06,
+      "loss": 0.3811,
       "step": 104215
     },
     {
-      "epoch": 3.67,
-      "learning_rate": 8.773754635489135e-06,
-      "loss": 0.2476,
+      "epoch": 3.7560817385663317,
+      "grad_norm": 0.2287319004535675,
+      "learning_rate": 7.700993051137751e-06,
+      "loss": 0.3758,
       "step": 104220
     },
     {
-      "epoch": 3.67,
-      "learning_rate": 8.771587587752491e-06,
-      "loss": 0.2766,
+      "epoch": 3.7562619382275564,
+      "grad_norm": 0.22735942900180817,
+      "learning_rate": 7.698886464496446e-06,
+      "loss": 0.4105,
       "step": 104225
     },
     {
-      "epoch": 3.67,
-      "learning_rate": 8.76942075073108e-06,
-      "loss": 0.2894,
+      "epoch": 3.7564421378887807,
+      "grad_norm": 0.18838316202163696,
+      "learning_rate": 7.696780113579464e-06,
+      "loss": 0.3754,
       "step": 104230
     },
     {
-      "epoch": 3.67,
-      "learning_rate": 8.767254124453052e-06,
-      "loss": 0.2504,
+      "epoch": 3.7566223375500054,
+      "grad_norm": 0.2099224328994751,
+      "learning_rate": 7.694673998415503e-06,
+      "loss": 0.3815,
       "step": 104235
     },
     {
-      "epoch": 3.67,
-      "learning_rate": 8.765087708946534e-06,
-      "loss": 0.2704,
+      "epoch": 3.75680253721123,
+      "grad_norm": 0.3171490728855133,
+      "learning_rate": 7.692568119033258e-06,
+      "loss": 0.3913,
       "step": 104240
     },
     {
-      "epoch": 3.67,
-      "learning_rate": 8.762921504239643e-06,
-      "loss": 0.2435,
+      "epoch": 3.7569827368724544,
+      "grad_norm": 0.25509047508239746,
+      "learning_rate": 7.690462475461416e-06,
+      "loss": 0.4001,
       "step": 104245
     },
     {
-      "epoch": 3.67,
-      "learning_rate": 8.760755510360525e-06,
-      "loss": 0.2439,
+      "epoch": 3.757162936533679,
+      "grad_norm": 0.2789323627948761,
+      "learning_rate": 7.688357067728676e-06,
+      "loss": 0.3647,
       "step": 104250
     },
     {
-      "epoch": 3.67,
-      "learning_rate": 8.758589727337286e-06,
-      "loss": 0.2392,
+      "epoch": 3.757343136194904,
+      "grad_norm": 0.21449807286262512,
+      "learning_rate": 7.686251895863721e-06,
+      "loss": 0.3643,
       "step": 104255
     },
     {
-      "epoch": 3.67,
-      "learning_rate": 8.756424155198064e-06,
-      "loss": 0.2783,
+      "epoch": 3.7575233358561286,
+      "grad_norm": 0.25625357031822205,
+      "learning_rate": 7.684146959895233e-06,
+      "loss": 0.3751,
       "step": 104260
     },
     {
-      "epoch": 3.67,
-      "learning_rate": 8.754258793970968e-06,
-      "loss": 0.2382,
+      "epoch": 3.7577035355173534,
+      "grad_norm": 0.25232619047164917,
+      "learning_rate": 7.68204225985189e-06,
+      "loss": 0.415,
       "step": 104265
     },
     {
-      "epoch": 3.67,
-      "learning_rate": 8.752093643684109e-06,
-      "loss": 0.258,
+      "epoch": 3.757883735178578,
+      "grad_norm": 0.24988026916980743,
+      "learning_rate": 7.67993779576236e-06,
+      "loss": 0.3681,
       "step": 104270
     },
     {
-      "epoch": 3.67,
-      "learning_rate": 8.749928704365611e-06,
-      "loss": 0.2632,
+      "epoch": 3.7580639348398024,
+      "grad_norm": 0.21227630972862244,
+      "learning_rate": 7.677833567655331e-06,
+      "loss": 0.3763,
       "step": 104275
     },
     {
-      "epoch": 3.67,
-      "learning_rate": 8.747763976043585e-06,
-      "loss": 0.2571,
+      "epoch": 3.758244134501027,
+      "grad_norm": 0.2227819859981537,
+      "learning_rate": 7.675729575559468e-06,
+      "loss": 0.3833,
       "step": 104280
     },
     {
-      "epoch": 3.67,
-      "learning_rate": 8.74559945874613e-06,
-      "loss": 0.2561,
+      "epoch": 3.758424334162252,
+      "grad_norm": 0.3054330348968506,
+      "learning_rate": 7.673625819503433e-06,
+      "loss": 0.3833,
       "step": 104285
     },
     {
-      "epoch": 3.67,
-      "learning_rate": 8.743435152501344e-06,
-      "loss": 0.2723,
+      "epoch": 3.758604533823476,
+      "grad_norm": 0.2734309136867523,
+      "learning_rate": 7.671522299515893e-06,
+      "loss": 0.3916,
       "step": 104290
     },
     {
-      "epoch": 3.67,
-      "learning_rate": 8.741271057337352e-06,
-      "loss": 0.2665,
+      "epoch": 3.758784733484701,
+      "grad_norm": 0.24472157657146454,
+      "learning_rate": 7.669419015625507e-06,
+      "loss": 0.3786,
       "step": 104295
     },
     {
-      "epoch": 3.67,
-      "learning_rate": 8.73910717328223e-06,
-      "loss": 0.2651,
+      "epoch": 3.7589649331459256,
+      "grad_norm": 0.21005383133888245,
+      "learning_rate": 7.667315967860925e-06,
+      "loss": 0.38,
       "step": 104300
     },
     {
-      "epoch": 3.67,
-      "learning_rate": 8.736943500364092e-06,
-      "loss": 0.2573,
+      "epoch": 3.7591451328071503,
+      "grad_norm": 0.18134789168834686,
+      "learning_rate": 7.665213156250819e-06,
+      "loss": 0.3707,
       "step": 104305
     },
     {
-      "epoch": 3.67,
-      "learning_rate": 8.734780038611018e-06,
-      "loss": 0.2369,
+      "epoch": 3.759325332468375,
+      "grad_norm": 0.26828300952911377,
+      "learning_rate": 7.663110580823816e-06,
+      "loss": 0.3727,
       "step": 104310
     },
     {
-      "epoch": 3.67,
-      "learning_rate": 8.732616788051113e-06,
-      "loss": 0.2747,
+      "epoch": 3.7595055321296,
+      "grad_norm": 0.28588294982910156,
+      "learning_rate": 7.661008241608581e-06,
+      "loss": 0.3755,
       "step": 104315
     },
     {
-      "epoch": 3.67,
-      "learning_rate": 8.730453748712458e-06,
-      "loss": 0.2685,
+      "epoch": 3.759685731790824,
+      "grad_norm": 0.27312424778938293,
+      "learning_rate": 7.65890613863375e-06,
+      "loss": 0.4112,
       "step": 104320
     },
     {
-      "epoch": 3.67,
-      "learning_rate": 8.728290920623133e-06,
-      "loss": 0.2375,
+      "epoch": 3.759865931452049,
+      "grad_norm": 0.2866198420524597,
+      "learning_rate": 7.65680427192797e-06,
+      "loss": 0.4028,
       "step": 104325
     },
     {
-      "epoch": 3.67,
-      "learning_rate": 8.726128303811234e-06,
-      "loss": 0.2635,
+      "epoch": 3.7600461311132736,
+      "grad_norm": 0.2471124827861786,
+      "learning_rate": 7.654702641519871e-06,
+      "loss": 0.3696,
       "step": 104330
     },
     {
-      "epoch": 3.67,
-      "learning_rate": 8.723965898304833e-06,
-      "loss": 0.2421,
+      "epoch": 3.7602263307744983,
+      "grad_norm": 0.24191972613334656,
+      "learning_rate": 7.65260124743809e-06,
+      "loss": 0.3817,
       "step": 104335
     },
     {
-      "epoch": 3.67,
-      "learning_rate": 8.721803704132011e-06,
-      "loss": 0.2507,
+      "epoch": 3.7604065304357226,
+      "grad_norm": 0.22872976958751678,
+      "learning_rate": 7.650500089711252e-06,
+      "loss": 0.4256,
       "step": 104340
     },
     {
-      "epoch": 3.67,
-      "learning_rate": 8.719641721320831e-06,
-      "loss": 0.2881,
+      "epoch": 3.7605867300969473,
+      "grad_norm": 0.3062474727630615,
+      "learning_rate": 7.648399168367998e-06,
+      "loss": 0.3657,
       "step": 104345
     },
     {
-      "epoch": 3.67,
-      "learning_rate": 8.717479949899381e-06,
-      "loss": 0.2453,
+      "epoch": 3.760766929758172,
+      "grad_norm": 0.2952815592288971,
+      "learning_rate": 7.646298483436946e-06,
+      "loss": 0.3974,
       "step": 104350
     },
     {
-      "epoch": 3.67,
-      "learning_rate": 8.715318389895715e-06,
-      "loss": 0.2716,
+      "epoch": 3.760947129419397,
+      "grad_norm": 0.30336183309555054,
+      "learning_rate": 7.644198034946718e-06,
+      "loss": 0.369,
       "step": 104355
     },
     {
-      "epoch": 3.67,
-      "learning_rate": 8.713157041337919e-06,
-      "loss": 0.2413,
+      "epoch": 3.7611273290806215,
+      "grad_norm": 0.24275535345077515,
+      "learning_rate": 7.642097822925932e-06,
+      "loss": 0.3679,
       "step": 104360
     },
     {
-      "epoch": 3.67,
-      "learning_rate": 8.710995904254036e-06,
-      "loss": 0.2537,
+      "epoch": 3.761307528741846,
+      "grad_norm": 0.20958095788955688,
+      "learning_rate": 7.639997847403194e-06,
+      "loss": 0.4052,
       "step": 104365
     },
     {
-      "epoch": 3.67,
-      "learning_rate": 8.708834978672142e-06,
-      "loss": 0.2658,
+      "epoch": 3.7614877284030706,
+      "grad_norm": 0.27302390336990356,
+      "learning_rate": 7.637898108407132e-06,
+      "loss": 0.3837,
       "step": 104370
     },
     {
-      "epoch": 3.67,
-      "learning_rate": 8.70667426462029e-06,
-      "loss": 0.2662,
+      "epoch": 3.7616679280642953,
+      "grad_norm": 0.3011491894721985,
+      "learning_rate": 7.635798605966346e-06,
+      "loss": 0.4018,
       "step": 104375
     },
     {
-      "epoch": 3.67,
-      "learning_rate": 8.704513762126537e-06,
-      "loss": 0.2634,
+      "epoch": 3.76184812772552,
+      "grad_norm": 0.2879173159599304,
+      "learning_rate": 7.633699340109443e-06,
+      "loss": 0.4226,
       "step": 104380
     },
     {
-      "epoch": 3.67,
-      "learning_rate": 8.702353471218924e-06,
-      "loss": 0.2611,
+      "epoch": 3.7620283273867443,
+      "grad_norm": 0.2275601178407669,
+      "learning_rate": 7.631600310865025e-06,
+      "loss": 0.4083,
       "step": 104385
     },
     {
-      "epoch": 3.67,
-      "learning_rate": 8.70019339192552e-06,
-      "loss": 0.2609,
+      "epoch": 3.762208527047969,
+      "grad_norm": 0.208656445145607,
+      "learning_rate": 7.62950151826168e-06,
+      "loss": 0.3667,
       "step": 104390
     },
     {
-      "epoch": 3.67,
-      "learning_rate": 8.698033524274362e-06,
-      "loss": 0.2498,
+      "epoch": 3.7623887267091938,
+      "grad_norm": 0.2310076206922531,
+      "learning_rate": 7.627402962328026e-06,
+      "loss": 0.3664,
       "step": 104395
     },
     {
-      "epoch": 3.67,
-      "learning_rate": 8.695873868293495e-06,
-      "loss": 0.2725,
+      "epoch": 3.7625689263704185,
+      "grad_norm": 0.2891443073749542,
+      "learning_rate": 7.625304643092648e-06,
+      "loss": 0.3927,
       "step": 104400
     },
     {
-      "epoch": 3.67,
-      "learning_rate": 8.693714424010954e-06,
-      "loss": 0.2618,
+      "epoch": 3.7627491260316432,
+      "grad_norm": 0.254215806722641,
+      "learning_rate": 7.623206560584115e-06,
+      "loss": 0.3493,
       "step": 104405
     },
     {
-      "epoch": 3.67,
-      "learning_rate": 8.691555191454793e-06,
-      "loss": 0.2633,
+      "epoch": 3.7629293256928675,
+      "grad_norm": 0.19437402486801147,
+      "learning_rate": 7.621108714831038e-06,
+      "loss": 0.3739,
       "step": 104410
     },
     {
-      "epoch": 3.67,
-      "learning_rate": 8.68939617065303e-06,
-      "loss": 0.2531,
+      "epoch": 3.7631095253540923,
+      "grad_norm": 0.23699922859668732,
+      "learning_rate": 7.619011105861987e-06,
+      "loss": 0.4086,
       "step": 104415
     },
     {
-      "epoch": 3.67,
-      "learning_rate": 8.687237361633719e-06,
-      "loss": 0.2509,
+      "epoch": 3.763289725015317,
+      "grad_norm": 0.21388274431228638,
+      "learning_rate": 7.616913733705547e-06,
+      "loss": 0.3858,
       "step": 104420
     },
     {
-      "epoch": 3.67,
-      "learning_rate": 8.68507876442487e-06,
-      "loss": 0.2635,
+      "epoch": 3.7634699246765417,
+      "grad_norm": 0.22161832451820374,
+      "learning_rate": 7.61481659839029e-06,
+      "loss": 0.3757,
       "step": 104425
     },
     {
-      "epoch": 3.67,
-      "learning_rate": 8.682920379054533e-06,
-      "loss": 0.2755,
+      "epoch": 3.763650124337766,
+      "grad_norm": 0.2681003212928772,
+      "learning_rate": 7.612719699944784e-06,
+      "loss": 0.4036,
       "step": 104430
     },
     {
-      "epoch": 3.67,
-      "learning_rate": 8.680762205550718e-06,
-      "loss": 0.2635,
+      "epoch": 3.7638303239989908,
+      "grad_norm": 0.2696530818939209,
+      "learning_rate": 7.610623038397613e-06,
+      "loss": 0.408,
       "step": 104435
     },
     {
-      "epoch": 3.67,
-      "learning_rate": 8.678604243941446e-06,
-      "loss": 0.2531,
+      "epoch": 3.7640105236602155,
+      "grad_norm": 0.25873225927352905,
+      "learning_rate": 7.608526613777339e-06,
+      "loss": 0.4039,
       "step": 104440
     },
     {
-      "epoch": 3.67,
-      "learning_rate": 8.676446494254748e-06,
-      "loss": 0.2754,
+      "epoch": 3.7641907233214402,
+      "grad_norm": 0.1770642250776291,
+      "learning_rate": 7.6064304261125205e-06,
+      "loss": 0.3661,
       "step": 104445
     },
     {
-      "epoch": 3.67,
-      "learning_rate": 8.674288956518633e-06,
-      "loss": 0.2745,
+      "epoch": 3.764370922982665,
+      "grad_norm": 0.25411689281463623,
+      "learning_rate": 7.604334475431721e-06,
+      "loss": 0.3954,
       "step": 104450
     },
     {
-      "epoch": 3.68,
-      "learning_rate": 8.672131630761119e-06,
-      "loss": 0.2658,
+      "epoch": 3.7645511226438897,
+      "grad_norm": 0.19758695363998413,
+      "learning_rate": 7.602238761763486e-06,
+      "loss": 0.3648,
       "step": 104455
     },
     {
-      "epoch": 3.68,
-      "learning_rate": 8.669974517010207e-06,
-      "loss": 0.2687,
+      "epoch": 3.764731322305114,
+      "grad_norm": 0.24316218495368958,
+      "learning_rate": 7.600143285136391e-06,
+      "loss": 0.4164,
       "step": 104460
     },
     {
-      "epoch": 3.68,
-      "learning_rate": 8.667817615293922e-06,
-      "loss": 0.2609,
+      "epoch": 3.7649115219663387,
+      "grad_norm": 0.23017525672912598,
+      "learning_rate": 7.598048045578973e-06,
+      "loss": 0.3761,
       "step": 104465
     },
     {
-      "epoch": 3.68,
-      "learning_rate": 8.665660925640254e-06,
-      "loss": 0.246,
+      "epoch": 3.7650917216275634,
+      "grad_norm": 0.2317330241203308,
+      "learning_rate": 7.595953043119783e-06,
+      "loss": 0.3907,
       "step": 104470
     },
     {
-      "epoch": 3.68,
-      "learning_rate": 8.663504448077222e-06,
-      "loss": 0.2521,
+      "epoch": 3.7652719212887877,
+      "grad_norm": 0.21798720955848694,
+      "learning_rate": 7.593858277787361e-06,
+      "loss": 0.3797,
       "step": 104475
     },
     {
-      "epoch": 3.68,
-      "learning_rate": 8.661348182632812e-06,
-      "loss": 0.2602,
+      "epoch": 3.7654521209500125,
+      "grad_norm": 0.18874719738960266,
+      "learning_rate": 7.59176374961025e-06,
+      "loss": 0.3733,
       "step": 104480
     },
     {
-      "epoch": 3.68,
-      "learning_rate": 8.659192129335034e-06,
-      "loss": 0.2542,
+      "epoch": 3.765632320611237,
+      "grad_norm": 0.24172750115394592,
+      "learning_rate": 7.58966945861698e-06,
+      "loss": 0.3748,
       "step": 104485
     },
     {
-      "epoch": 3.68,
-      "learning_rate": 8.657036288211878e-06,
-      "loss": 0.2705,
+      "epoch": 3.765812520272462,
+      "grad_norm": 0.32291868329048157,
+      "learning_rate": 7.5875754048361114e-06,
+      "loss": 0.4205,
       "step": 104490
     },
     {
-      "epoch": 3.68,
-      "learning_rate": 8.654880659291335e-06,
-      "loss": 0.2608,
+      "epoch": 3.7659927199336867,
+      "grad_norm": 0.23808036744594574,
+      "learning_rate": 7.5854815882961364e-06,
+      "loss": 0.3663,
       "step": 104495
     },
     {
-      "epoch": 3.68,
-      "learning_rate": 8.652725242601386e-06,
-      "loss": 0.256,
+      "epoch": 3.7661729195949114,
+      "grad_norm": 0.2094922959804535,
+      "learning_rate": 7.583388009025616e-06,
+      "loss": 0.3745,
       "step": 104500
     },
     {
-      "epoch": 3.68,
-      "eval_loss": 0.2519655227661133,
-      "eval_runtime": 10.5307,
-      "eval_samples_per_second": 9.496,
-      "eval_steps_per_second": 9.496,
+      "epoch": 3.7661729195949114,
+      "eval_loss": 0.4286915063858032,
+      "eval_runtime": 3.5312,
+      "eval_samples_per_second": 28.319,
+      "eval_steps_per_second": 7.08,
       "step": 104500
     },
     {
-      "epoch": 3.68,
-      "learning_rate": 8.650570038170037e-06,
-      "loss": 0.2585,
+      "epoch": 3.7663531192561357,
+      "grad_norm": 0.25394493341445923,
+      "learning_rate": 7.581294667053057e-06,
+      "loss": 0.3741,
       "step": 104505
     },
     {
-      "epoch": 3.68,
-      "learning_rate": 8.64841504602526e-06,
-      "loss": 0.2414,
+      "epoch": 3.7665333189173604,
+      "grad_norm": 0.22518283128738403,
+      "learning_rate": 7.579201562406982e-06,
+      "loss": 0.372,
       "step": 104510
     },
     {
-      "epoch": 3.68,
-      "learning_rate": 8.646260266195036e-06,
-      "loss": 0.2293,
+      "epoch": 3.766713518578585,
+      "grad_norm": 0.24221821129322052,
+      "learning_rate": 7.577108695115928e-06,
+      "loss": 0.4004,
       "step": 104515
     },
     {
-      "epoch": 3.68,
-      "learning_rate": 8.644105698707339e-06,
-      "loss": 0.2631,
+      "epoch": 3.7668937182398095,
+      "grad_norm": 0.25630852580070496,
+      "learning_rate": 7.575016065208385e-06,
+      "loss": 0.3834,
       "step": 104520
     },
     {
-      "epoch": 3.68,
-      "learning_rate": 8.641951343590149e-06,
-      "loss": 0.2891,
+      "epoch": 3.767073917901034,
+      "grad_norm": 0.21267221868038177,
+      "learning_rate": 7.5729236727128674e-06,
+      "loss": 0.3436,
       "step": 104525
     },
     {
-      "epoch": 3.68,
-      "learning_rate": 8.63979720087145e-06,
-      "loss": 0.2713,
+      "epoch": 3.767254117562259,
+      "grad_norm": 0.24416276812553406,
+      "learning_rate": 7.5708315176579e-06,
+      "loss": 0.4102,
       "step": 104530
     },
     {
-      "epoch": 3.68,
-      "learning_rate": 8.637643270579204e-06,
-      "loss": 0.2612,
+      "epoch": 3.7674343172234837,
+      "grad_norm": 0.2707764506340027,
+      "learning_rate": 7.568739600071978e-06,
+      "loss": 0.3707,
       "step": 104535
     },
     {
-      "epoch": 3.68,
-      "learning_rate": 8.635489552741366e-06,
-      "loss": 0.241,
+      "epoch": 3.7676145168847084,
+      "grad_norm": 0.2370613068342209,
+      "learning_rate": 7.566647919983602e-06,
+      "loss": 0.3701,
       "step": 104540
     },
     {
-      "epoch": 3.68,
-      "learning_rate": 8.633336047385921e-06,
-      "loss": 0.2513,
+      "epoch": 3.767794716545933,
+      "grad_norm": 0.2817910611629486,
+      "learning_rate": 7.564556477421275e-06,
+      "loss": 0.3784,
       "step": 104545
     },
     {
-      "epoch": 3.68,
-      "learning_rate": 8.63118275454082e-06,
-      "loss": 0.2451,
+      "epoch": 3.7679749162071574,
+      "grad_norm": 0.27115681767463684,
+      "learning_rate": 7.562465272413483e-06,
+      "loss": 0.4333,
       "step": 104550
     },
     {
-      "epoch": 3.68,
-      "learning_rate": 8.629029674234015e-06,
-      "loss": 0.2679,
+      "epoch": 3.768155115868382,
+      "grad_norm": 0.23443850874900818,
+      "learning_rate": 7.560374304988732e-06,
+      "loss": 0.4106,
       "step": 104555
     },
     {
-      "epoch": 3.68,
-      "learning_rate": 8.62687680649348e-06,
-      "loss": 0.2696,
+      "epoch": 3.768335315529607,
+      "grad_norm": 0.2954671382904053,
+      "learning_rate": 7.5582835751755064e-06,
+      "loss": 0.3693,
       "step": 104560
     },
     {
-      "epoch": 3.68,
-      "learning_rate": 8.62472415134716e-06,
-      "loss": 0.2559,
+      "epoch": 3.768515515190831,
+      "grad_norm": 0.2391144335269928,
+      "learning_rate": 7.556193083002291e-06,
+      "loss": 0.3376,
       "step": 104565
     },
     {
-      "epoch": 3.68,
-      "learning_rate": 8.622571708823007e-06,
-      "loss": 0.2507,
+      "epoch": 3.768695714852056,
+      "grad_norm": 0.29020795226097107,
+      "learning_rate": 7.554102828497564e-06,
+      "loss": 0.4022,
       "step": 104570
     },
     {
-      "epoch": 3.68,
-      "learning_rate": 8.620419478948957e-06,
-      "loss": 0.2527,
+      "epoch": 3.7688759145132806,
+      "grad_norm": 0.24601967632770538,
+      "learning_rate": 7.552012811689804e-06,
+      "loss": 0.3811,
       "step": 104575
     },
     {
-      "epoch": 3.68,
-      "learning_rate": 8.618267461752965e-06,
-      "loss": 0.2299,
+      "epoch": 3.7690561141745054,
+      "grad_norm": 0.23416286706924438,
+      "learning_rate": 7.549923032607498e-06,
+      "loss": 0.3815,
       "step": 104580
     },
     {
-      "epoch": 3.68,
-      "learning_rate": 8.616115657262982e-06,
-      "loss": 0.2722,
+      "epoch": 3.76923631383573,
+      "grad_norm": 0.2573948800563812,
+      "learning_rate": 7.547833491279119e-06,
+      "loss": 0.3997,
       "step": 104585
     },
     {
-      "epoch": 3.68,
-      "learning_rate": 8.61396406550694e-06,
-      "loss": 0.2582,
+      "epoch": 3.769416513496955,
+      "grad_norm": 0.23901304602622986,
+      "learning_rate": 7.5457441877331145e-06,
+      "loss": 0.3577,
       "step": 104590
     },
     {
-      "epoch": 3.68,
-      "learning_rate": 8.611812686512769e-06,
-      "loss": 0.273,
+      "epoch": 3.769596713158179,
+      "grad_norm": 0.1927175372838974,
+      "learning_rate": 7.543655121997975e-06,
+      "loss": 0.3581,
       "step": 104595
     },
     {
-      "epoch": 3.68,
-      "learning_rate": 8.609661520308417e-06,
-      "loss": 0.2449,
+      "epoch": 3.769776912819404,
+      "grad_norm": 0.30516329407691956,
+      "learning_rate": 7.541566294102154e-06,
+      "loss": 0.3853,
       "step": 104600
     },
     {
-      "epoch": 3.68,
-      "learning_rate": 8.607510566921809e-06,
-      "loss": 0.2516,
+      "epoch": 3.7699571124806286,
+      "grad_norm": 0.22798460721969604,
+      "learning_rate": 7.539895403048933e-06,
+      "loss": 0.3772,
       "step": 104605
     },
     {
-      "epoch": 3.68,
-      "learning_rate": 8.605359826380873e-06,
-      "loss": 0.2548,
+      "epoch": 3.7701373121418533,
+      "grad_norm": 0.24452564120292664,
+      "learning_rate": 7.537807003335604e-06,
+      "loss": 0.3669,
       "step": 104610
     },
     {
-      "epoch": 3.68,
-      "learning_rate": 8.603209298713527e-06,
-      "loss": 0.2399,
+      "epoch": 3.7703175118030776,
+      "grad_norm": 0.22427910566329956,
+      "learning_rate": 7.5357188415412724e-06,
+      "loss": 0.3694,
       "step": 104615
     },
     {
-      "epoch": 3.68,
-      "learning_rate": 8.60105898394771e-06,
-      "loss": 0.232,
+      "epoch": 3.7704977114643023,
+      "grad_norm": 0.2747547924518585,
+      "learning_rate": 7.5336309176943845e-06,
+      "loss": 0.3733,
       "step": 104620
     },
     {
-      "epoch": 3.68,
-      "learning_rate": 8.598908882111336e-06,
-      "loss": 0.2542,
+      "epoch": 3.770677911125527,
+      "grad_norm": 0.22400154173374176,
+      "learning_rate": 7.531543231823399e-06,
+      "loss": 0.3961,
       "step": 104625
     },
     {
-      "epoch": 3.68,
-      "learning_rate": 8.596758993232312e-06,
-      "loss": 0.257,
+      "epoch": 3.770858110786752,
+      "grad_norm": 0.2119477391242981,
+      "learning_rate": 7.529455783956757e-06,
+      "loss": 0.3566,
       "step": 104630
     },
     {
-      "epoch": 3.68,
-      "learning_rate": 8.594609317338565e-06,
-      "loss": 0.2478,
+      "epoch": 3.7710383104479765,
+      "grad_norm": 0.22162562608718872,
+      "learning_rate": 7.5273685741228976e-06,
+      "loss": 0.3925,
       "step": 104635
     },
     {
-      "epoch": 3.68,
-      "learning_rate": 8.59245985445801e-06,
-      "loss": 0.255,
+      "epoch": 3.771218510109201,
+      "grad_norm": 0.20573687553405762,
+      "learning_rate": 7.525281602350259e-06,
+      "loss": 0.3899,
       "step": 104640
     },
     {
-      "epoch": 3.68,
-      "learning_rate": 8.59031060461855e-06,
-      "loss": 0.2627,
+      "epoch": 3.7713987097704256,
+      "grad_norm": 0.26889345049858093,
+      "learning_rate": 7.523194868667266e-06,
+      "loss": 0.407,
       "step": 104645
     },
     {
-      "epoch": 3.68,
-      "learning_rate": 8.588161567848095e-06,
-      "loss": 0.2289,
+      "epoch": 3.7715789094316503,
+      "grad_norm": 0.23474839329719543,
+      "learning_rate": 7.521108373102367e-06,
+      "loss": 0.3577,
       "step": 104650
     },
     {
-      "epoch": 3.68,
-      "learning_rate": 8.586012744174538e-06,
-      "loss": 0.255,
+      "epoch": 3.771759109092875,
+      "grad_norm": 0.2511729896068573,
+      "learning_rate": 7.519022115683994e-06,
+      "loss": 0.3994,
       "step": 104655
     },
     {
-      "epoch": 3.68,
-      "learning_rate": 8.583864133625794e-06,
-      "loss": 0.2671,
+      "epoch": 3.7719393087540993,
+      "grad_norm": 0.24130114912986755,
+      "learning_rate": 7.5169360964405415e-06,
+      "loss": 0.3788,
       "step": 104660
     },
     {
-      "epoch": 3.68,
-      "learning_rate": 8.581715736229756e-06,
-      "loss": 0.2716,
+      "epoch": 3.772119508415324,
+      "grad_norm": 0.2123662829399109,
+      "learning_rate": 7.514850315400457e-06,
+      "loss": 0.3786,
       "step": 104665
     },
     {
-      "epoch": 3.68,
-      "learning_rate": 8.579567552014312e-06,
-      "loss": 0.2696,
+      "epoch": 3.772299708076549,
+      "grad_norm": 0.21604010462760925,
+      "learning_rate": 7.512764772592151e-06,
+      "loss": 0.3807,
       "step": 104670
     },
     {
-      "epoch": 3.68,
-      "learning_rate": 8.57741958100737e-06,
-      "loss": 0.2524,
+      "epoch": 3.7724799077377735,
+      "grad_norm": 0.26000019907951355,
+      "learning_rate": 7.510679468044035e-06,
+      "loss": 0.3552,
       "step": 104675
     },
     {
-      "epoch": 3.68,
-      "learning_rate": 8.575271823236813e-06,
-      "loss": 0.265,
+      "epoch": 3.7726601073989983,
+      "grad_norm": 0.2929229140281677,
+      "learning_rate": 7.508594401784538e-06,
+      "loss": 0.3689,
       "step": 104680
     },
     {
-      "epoch": 3.68,
-      "learning_rate": 8.573124278730516e-06,
-      "loss": 0.2618,
+      "epoch": 3.772840307060223,
+      "grad_norm": 0.26968440413475037,
+      "learning_rate": 7.506509573842041e-06,
+      "loss": 0.3501,
       "step": 104685
     },
     {
-      "epoch": 3.68,
-      "learning_rate": 8.570976947516385e-06,
-      "loss": 0.2743,
+      "epoch": 3.7730205067214473,
+      "grad_norm": 0.2252168208360672,
+      "learning_rate": 7.5044249842449735e-06,
+      "loss": 0.3709,
       "step": 104690
     },
     {
-      "epoch": 3.68,
-      "learning_rate": 8.568829829622283e-06,
-      "loss": 0.2527,
+      "epoch": 3.773200706382672,
+      "grad_norm": 0.20150010287761688,
+      "learning_rate": 7.502340633021726e-06,
+      "loss": 0.3637,
       "step": 104695
     },
     {
-      "epoch": 3.68,
-      "learning_rate": 8.566682925076106e-06,
-      "loss": 0.2584,
+      "epoch": 3.7733809060438968,
+      "grad_norm": 0.2643103003501892,
+      "learning_rate": 7.500256520200702e-06,
+      "loss": 0.4388,
       "step": 104700
     },
     {
-      "epoch": 3.68,
-      "learning_rate": 8.564536233905718e-06,
-      "loss": 0.2596,
+      "epoch": 3.773561105705121,
+      "grad_norm": 0.2742420732975006,
+      "learning_rate": 7.498172645810292e-06,
+      "loss": 0.3994,
       "step": 104705
     },
     {
-      "epoch": 3.68,
-      "learning_rate": 8.562389756138988e-06,
-      "loss": 0.276,
+      "epoch": 3.7737413053663458,
+      "grad_norm": 0.3029593229293823,
+      "learning_rate": 7.496089009878884e-06,
+      "loss": 0.3941,
       "step": 104710
     },
     {
-      "epoch": 3.68,
-      "learning_rate": 8.560243491803805e-06,
-      "loss": 0.2643,
+      "epoch": 3.7739215050275705,
+      "grad_norm": 0.23510831594467163,
+      "learning_rate": 7.494005612434885e-06,
+      "loss": 0.3873,
       "step": 104715
     },
     {
-      "epoch": 3.68,
-      "learning_rate": 8.558097440928026e-06,
-      "loss": 0.2524,
+      "epoch": 3.7741017046887952,
+      "grad_norm": 0.2507026195526123,
+      "learning_rate": 7.49192245350667e-06,
+      "loss": 0.3756,
       "step": 104720
     },
     {
-      "epoch": 3.68,
-      "learning_rate": 8.555951603539513e-06,
-      "loss": 0.2529,
+      "epoch": 3.77428190435002,
+      "grad_norm": 0.2193816602230072,
+      "learning_rate": 7.48983953312262e-06,
+      "loss": 0.3564,
       "step": 104725
     },
     {
-      "epoch": 3.68,
-      "learning_rate": 8.553805979666126e-06,
-      "loss": 0.2391,
+      "epoch": 3.7744621040112447,
+      "grad_norm": 0.2253447026014328,
+      "learning_rate": 7.487756851311114e-06,
+      "loss": 0.3607,
       "step": 104730
     },
     {
-      "epoch": 3.68,
-      "learning_rate": 8.551660569335737e-06,
-      "loss": 0.2702,
+      "epoch": 3.774642303672469,
+      "grad_norm": 0.27351540327072144,
+      "learning_rate": 7.485674408100535e-06,
+      "loss": 0.3961,
       "step": 104735
     },
     {
-      "epoch": 3.69,
-      "learning_rate": 8.54951537257619e-06,
-      "loss": 0.2369,
+      "epoch": 3.7748225033336937,
+      "grad_norm": 0.20876996219158173,
+      "learning_rate": 7.483592203519241e-06,
+      "loss": 0.3839,
       "step": 104740
     },
     {
-      "epoch": 3.69,
-      "learning_rate": 8.547370389415349e-06,
-      "loss": 0.2496,
+      "epoch": 3.7750027029949185,
+      "grad_norm": 0.23797959089279175,
+      "learning_rate": 7.481510237595621e-06,
+      "loss": 0.3904,
       "step": 104745
     },
     {
-      "epoch": 3.69,
-      "learning_rate": 8.545225619881053e-06,
-      "loss": 0.2909,
+      "epoch": 3.7751829026561428,
+      "grad_norm": 0.21280419826507568,
+      "learning_rate": 7.47942851035803e-06,
+      "loss": 0.3748,
       "step": 104750
     },
     {
-      "epoch": 3.69,
-      "learning_rate": 8.543081064001168e-06,
-      "loss": 0.2438,
+      "epoch": 3.7753631023173675,
+      "grad_norm": 0.22197531163692474,
+      "learning_rate": 7.477347021834838e-06,
+      "loss": 0.3514,
       "step": 104755
     },
     {
-      "epoch": 3.69,
-      "learning_rate": 8.54093672180353e-06,
-      "loss": 0.2567,
+      "epoch": 3.7755433019785922,
+      "grad_norm": 0.2479541003704071,
+      "learning_rate": 7.475265772054396e-06,
+      "loss": 0.3469,
       "step": 104760
     },
     {
-      "epoch": 3.69,
-      "learning_rate": 8.538792593315981e-06,
-      "loss": 0.2717,
+      "epoch": 3.775723501639817,
+      "grad_norm": 0.2767854630947113,
+      "learning_rate": 7.4731847610450604e-06,
+      "loss": 0.3672,
       "step": 104765
     },
     {
-      "epoch": 3.69,
-      "learning_rate": 8.536648678566353e-06,
-      "loss": 0.2731,
+      "epoch": 3.7759037013010417,
+      "grad_norm": 0.21771368384361267,
+      "learning_rate": 7.471103988835202e-06,
+      "loss": 0.3716,
       "step": 104770
     },
     {
-      "epoch": 3.69,
-      "learning_rate": 8.534504977582502e-06,
-      "loss": 0.2562,
+      "epoch": 3.7760839009622664,
+      "grad_norm": 0.25155210494995117,
+      "learning_rate": 7.46902345545315e-06,
+      "loss": 0.3624,
       "step": 104775
     },
     {
-      "epoch": 3.69,
-      "learning_rate": 8.532361490392252e-06,
-      "loss": 0.2409,
+      "epoch": 3.7762641006234907,
+      "grad_norm": 0.27159664034843445,
+      "learning_rate": 7.466943160927253e-06,
+      "loss": 0.3326,
       "step": 104780
     },
     {
-      "epoch": 3.69,
-      "learning_rate": 8.530218217023428e-06,
-      "loss": 0.2501,
+      "epoch": 3.7764443002847154,
+      "grad_norm": 0.22621986269950867,
+      "learning_rate": 7.464863105285868e-06,
+      "loss": 0.3833,
       "step": 104785
     },
     {
-      "epoch": 3.69,
-      "learning_rate": 8.528075157503873e-06,
-      "loss": 0.254,
+      "epoch": 3.77662449994594,
+      "grad_norm": 0.21097923815250397,
+      "learning_rate": 7.462783288557329e-06,
+      "loss": 0.378,
       "step": 104790
     },
     {
-      "epoch": 3.69,
-      "learning_rate": 8.525932311861399e-06,
-      "loss": 0.2703,
+      "epoch": 3.7768046996071645,
+      "grad_norm": 0.21815836429595947,
+      "learning_rate": 7.460703710769973e-06,
+      "loss": 0.364,
       "step": 104795
     },
     {
-      "epoch": 3.69,
-      "learning_rate": 8.523789680123844e-06,
-      "loss": 0.2657,
+      "epoch": 3.776984899268389,
+      "grad_norm": 0.22789250314235687,
+      "learning_rate": 7.458624371952133e-06,
+      "loss": 0.3781,
       "step": 104800
     },
     {
-      "epoch": 3.69,
-      "learning_rate": 8.521647262319016e-06,
-      "loss": 0.2476,
+      "epoch": 3.777165098929614,
+      "grad_norm": 0.28179875016212463,
+      "learning_rate": 7.456545272132132e-06,
+      "loss": 0.417,
       "step": 104805
     },
     {
-      "epoch": 3.69,
-      "learning_rate": 8.519505058474747e-06,
-      "loss": 0.254,
+      "epoch": 3.7773452985908387,
+      "grad_norm": 0.21157479286193848,
+      "learning_rate": 7.45446641133831e-06,
+      "loss": 0.3719,
       "step": 104810
     },
     {
-      "epoch": 3.69,
-      "learning_rate": 8.517363068618842e-06,
-      "loss": 0.2672,
+      "epoch": 3.7775254982520634,
+      "grad_norm": 0.24585333466529846,
+      "learning_rate": 7.452387789598988e-06,
+      "loss": 0.399,
       "step": 104815
     },
     {
-      "epoch": 3.69,
-      "learning_rate": 8.515221292779115e-06,
-      "loss": 0.2628,
+      "epoch": 3.777705697913288,
+      "grad_norm": 0.2581721842288971,
+      "learning_rate": 7.450309406942488e-06,
+      "loss": 0.4285,
       "step": 104820
     },
     {
-      "epoch": 3.69,
-      "learning_rate": 8.51307973098337e-06,
-      "loss": 0.2794,
+      "epoch": 3.7778858975745124,
+      "grad_norm": 0.2738182842731476,
+      "learning_rate": 7.448231263397121e-06,
+      "loss": 0.4033,
       "step": 104825
     },
     {
-      "epoch": 3.69,
-      "learning_rate": 8.510938383259426e-06,
-      "loss": 0.267,
+      "epoch": 3.778066097235737,
+      "grad_norm": 0.22549976408481598,
+      "learning_rate": 7.446153358991198e-06,
+      "loss": 0.4013,
       "step": 104830
     },
     {
-      "epoch": 3.69,
-      "learning_rate": 8.508797249635078e-06,
-      "loss": 0.2309,
+      "epoch": 3.778246296896962,
+      "grad_norm": 0.20595544576644897,
+      "learning_rate": 7.444075693753044e-06,
+      "loss": 0.3896,
       "step": 104835
     },
     {
-      "epoch": 3.69,
-      "learning_rate": 8.506656330138132e-06,
-      "loss": 0.2529,
+      "epoch": 3.7784264965581866,
+      "grad_norm": 0.24291178584098816,
+      "learning_rate": 7.441998267710962e-06,
+      "loss": 0.3715,
       "step": 104840
     },
     {
-      "epoch": 3.69,
-      "learning_rate": 8.504515624796378e-06,
-      "loss": 0.2605,
+      "epoch": 3.778606696219411,
+      "grad_norm": 0.248873770236969,
+      "learning_rate": 7.439921080893253e-06,
+      "loss": 0.3732,
       "step": 104845
     },
     {
-      "epoch": 3.69,
-      "learning_rate": 8.502375133637621e-06,
-      "loss": 0.257,
+      "epoch": 3.7787868958806357,
+      "grad_norm": 0.28607362508773804,
+      "learning_rate": 7.43784413332822e-06,
+      "loss": 0.3769,
       "step": 104850
     },
     {
-      "epoch": 3.69,
-      "learning_rate": 8.500234856689646e-06,
-      "loss": 0.2644,
+      "epoch": 3.7789670955418604,
+      "grad_norm": 0.254102498292923,
+      "learning_rate": 7.43576742504416e-06,
+      "loss": 0.3671,
       "step": 104855
     },
     {
-      "epoch": 3.69,
-      "learning_rate": 8.498094793980254e-06,
-      "loss": 0.247,
+      "epoch": 3.779147295203085,
+      "grad_norm": 0.21698035299777985,
+      "learning_rate": 7.433690956069361e-06,
+      "loss": 0.3563,
       "step": 104860
     },
     {
-      "epoch": 3.69,
-      "learning_rate": 8.495954945537218e-06,
-      "loss": 0.2457,
+      "epoch": 3.77932749486431,
+      "grad_norm": 0.25203150510787964,
+      "learning_rate": 7.431614726432137e-06,
+      "loss": 0.3384,
       "step": 104865
     },
     {
-      "epoch": 3.69,
-      "learning_rate": 8.49381531138834e-06,
-      "loss": 0.2419,
+      "epoch": 3.779507694525534,
+      "grad_norm": 0.24114732444286346,
+      "learning_rate": 7.429538736160746e-06,
+      "loss": 0.4012,
       "step": 104870
     },
     {
-      "epoch": 3.69,
-      "learning_rate": 8.491675891561393e-06,
-      "loss": 0.2537,
+      "epoch": 3.779687894186759,
+      "grad_norm": 0.26797887682914734,
+      "learning_rate": 7.4274629852834955e-06,
+      "loss": 0.4196,
       "step": 104875
     },
     {
-      "epoch": 3.69,
-      "learning_rate": 8.489536686084152e-06,
-      "loss": 0.2623,
+      "epoch": 3.7798680938479836,
+      "grad_norm": 0.22066861391067505,
+      "learning_rate": 7.425387473828657e-06,
+      "loss": 0.3429,
       "step": 104880
     },
     {
-      "epoch": 3.69,
-      "learning_rate": 8.487397694984388e-06,
-      "loss": 0.2529,
+      "epoch": 3.7800482935092083,
+      "grad_norm": 0.23529654741287231,
+      "learning_rate": 7.423312201824514e-06,
+      "loss": 0.3881,
       "step": 104885
     },
     {
-      "epoch": 3.69,
-      "learning_rate": 8.485258918289895e-06,
-      "loss": 0.2585,
+      "epoch": 3.7802284931704326,
+      "grad_norm": 0.2249959409236908,
+      "learning_rate": 7.421237169299341e-06,
+      "loss": 0.3611,
       "step": 104890
     },
     {
-      "epoch": 3.69,
-      "learning_rate": 8.483120356028429e-06,
-      "loss": 0.2664,
+      "epoch": 3.7804086928316574,
+      "grad_norm": 0.24566762149333954,
+      "learning_rate": 7.419162376281397e-06,
+      "loss": 0.4263,
       "step": 104895
     },
     {
-      "epoch": 3.69,
-      "learning_rate": 8.48098200822775e-06,
-      "loss": 0.2778,
+      "epoch": 3.780588892492882,
+      "grad_norm": 0.18927571177482605,
+      "learning_rate": 7.417087822798971e-06,
+      "loss": 0.3778,
       "step": 104900
     },
     {
-      "epoch": 3.69,
-      "learning_rate": 8.478843874915646e-06,
-      "loss": 0.2379,
+      "epoch": 3.780769092154107,
+      "grad_norm": 0.2276502400636673,
+      "learning_rate": 7.415013508880319e-06,
+      "loss": 0.3857,
       "step": 104905
     },
     {
-      "epoch": 3.69,
-      "learning_rate": 8.476705956119854e-06,
-      "loss": 0.2932,
+      "epoch": 3.7809492918153316,
+      "grad_norm": 0.2894652485847473,
+      "learning_rate": 7.412939434553707e-06,
+      "loss": 0.3817,
       "step": 104910
     },
     {
-      "epoch": 3.69,
-      "learning_rate": 8.474568251868157e-06,
-      "loss": 0.2642,
+      "epoch": 3.781129491476556,
+      "grad_norm": 0.2344924807548523,
+      "learning_rate": 7.410865599847386e-06,
+      "loss": 0.3711,
       "step": 104915
     },
     {
-      "epoch": 3.69,
-      "learning_rate": 8.472430762188289e-06,
-      "loss": 0.2508,
+      "epoch": 3.7813096911377806,
+      "grad_norm": 0.2306702584028244,
+      "learning_rate": 7.408792004789616e-06,
+      "loss": 0.3803,
       "step": 104920
     },
     {
-      "epoch": 3.69,
-      "learning_rate": 8.470293487108024e-06,
-      "loss": 0.2744,
+      "epoch": 3.7814898907990053,
+      "grad_norm": 0.25984832644462585,
+      "learning_rate": 7.4067186494086425e-06,
+      "loss": 0.4154,
       "step": 104925
     },
     {
-      "epoch": 3.69,
-      "learning_rate": 8.468156426655108e-06,
-      "loss": 0.275,
+      "epoch": 3.78167009046023,
+      "grad_norm": 0.23578700423240662,
+      "learning_rate": 7.404645533732729e-06,
+      "loss": 0.3523,
       "step": 104930
     },
     {
-      "epoch": 3.69,
-      "learning_rate": 8.466019580857284e-06,
-      "loss": 0.2619,
+      "epoch": 3.7818502901214543,
+      "grad_norm": 0.2648398280143738,
+      "learning_rate": 7.4025726577901135e-06,
+      "loss": 0.3804,
       "step": 104935
     },
     {
-      "epoch": 3.69,
-      "learning_rate": 8.463882949742292e-06,
-      "loss": 0.2579,
+      "epoch": 3.782030489782679,
+      "grad_norm": 0.22717751562595367,
+      "learning_rate": 7.400500021609038e-06,
+      "loss": 0.376,
       "step": 104940
     },
     {
-      "epoch": 3.69,
-      "learning_rate": 8.461746533337888e-06,
-      "loss": 0.2501,
+      "epoch": 3.782210689443904,
+      "grad_norm": 0.2581784129142761,
+      "learning_rate": 7.398427625217743e-06,
+      "loss": 0.3915,
       "step": 104945
     },
     {
-      "epoch": 3.69,
-      "learning_rate": 8.459610331671807e-06,
-      "loss": 0.2773,
+      "epoch": 3.7823908891051286,
+      "grad_norm": 0.263120174407959,
+      "learning_rate": 7.3963554686444556e-06,
+      "loss": 0.4054,
       "step": 104950
     },
     {
-      "epoch": 3.69,
-      "learning_rate": 8.457474344771787e-06,
-      "loss": 0.2684,
+      "epoch": 3.7825710887663533,
+      "grad_norm": 0.24363566935062408,
+      "learning_rate": 7.394283551917433e-06,
+      "loss": 0.3712,
       "step": 104955
     },
     {
-      "epoch": 3.69,
-      "learning_rate": 8.455338572665553e-06,
-      "loss": 0.256,
+      "epoch": 3.782751288427578,
+      "grad_norm": 0.2304638773202896,
+      "learning_rate": 7.39221187506488e-06,
+      "loss": 0.3958,
       "step": 104960
     },
     {
-      "epoch": 3.69,
-      "learning_rate": 8.453203015380842e-06,
-      "loss": 0.2457,
+      "epoch": 3.7829314880888023,
+      "grad_norm": 0.3181205093860626,
+      "learning_rate": 7.3901404381150255e-06,
+      "loss": 0.405,
       "step": 104965
     },
     {
-      "epoch": 3.69,
-      "learning_rate": 8.451067672945395e-06,
-      "loss": 0.2383,
+      "epoch": 3.783111687750027,
+      "grad_norm": 0.29112708568573,
+      "learning_rate": 7.3880692410961045e-06,
+      "loss": 0.38,
       "step": 104970
     },
     {
-      "epoch": 3.69,
-      "learning_rate": 8.448932545386928e-06,
-      "loss": 0.2567,
+      "epoch": 3.7832918874112518,
+      "grad_norm": 0.25521332025527954,
+      "learning_rate": 7.385998284036322e-06,
+      "loss": 0.4297,
       "step": 104975
     },
     {
-      "epoch": 3.69,
-      "learning_rate": 8.446797632733156e-06,
-      "loss": 0.2452,
+      "epoch": 3.783472087072476,
+      "grad_norm": 0.2132871150970459,
+      "learning_rate": 7.383927566963919e-06,
+      "loss": 0.3451,
       "step": 104980
     },
     {
-      "epoch": 3.69,
-      "learning_rate": 8.444662935011815e-06,
-      "loss": 0.2533,
+      "epoch": 3.783652286733701,
+      "grad_norm": 0.25532466173171997,
+      "learning_rate": 7.381857089907082e-06,
+      "loss": 0.4008,
       "step": 104985
     },
     {
-      "epoch": 3.69,
-      "learning_rate": 8.442528452250614e-06,
-      "loss": 0.2415,
+      "epoch": 3.7838324863949255,
+      "grad_norm": 0.28475886583328247,
+      "learning_rate": 7.379786852894027e-06,
+      "loss": 0.3948,
       "step": 104990
     },
     {
-      "epoch": 3.69,
-      "learning_rate": 8.440394184477274e-06,
-      "loss": 0.2855,
+      "epoch": 3.7840126860561503,
+      "grad_norm": 0.24241603910923004,
+      "learning_rate": 7.377716855952971e-06,
+      "loss": 0.3718,
       "step": 104995
     },
     {
-      "epoch": 3.69,
-      "learning_rate": 8.438260131719489e-06,
-      "loss": 0.2472,
+      "epoch": 3.784192885717375,
+      "grad_norm": 0.22656336426734924,
+      "learning_rate": 7.37564709911211e-06,
+      "loss": 0.3624,
       "step": 105000
     },
     {
-      "epoch": 3.69,
-      "eval_loss": 0.25183290243148804,
-      "eval_runtime": 10.5414,
-      "eval_samples_per_second": 9.486,
-      "eval_steps_per_second": 9.486,
+      "epoch": 3.784192885717375,
+      "eval_loss": 0.4286513924598694,
+      "eval_runtime": 3.5365,
+      "eval_samples_per_second": 28.277,
+      "eval_steps_per_second": 7.069,
       "step": 105000
     },
     {
-      "epoch": 3.69,
-      "learning_rate": 8.436126294004992e-06,
-      "loss": 0.2619,
+      "epoch": 3.7843730853785997,
+      "grad_norm": 0.2644498944282532,
+      "learning_rate": 7.3735775823996465e-06,
+      "loss": 0.387,
       "step": 105005
     },
     {
-      "epoch": 3.69,
-      "learning_rate": 8.43399267136148e-06,
-      "loss": 0.2468,
+      "epoch": 3.784553285039824,
+      "grad_norm": 0.2483258843421936,
+      "learning_rate": 7.371508305843775e-06,
+      "loss": 0.3835,
       "step": 105010
     },
     {
-      "epoch": 3.69,
-      "learning_rate": 8.431859263816647e-06,
-      "loss": 0.2627,
+      "epoch": 3.7847334847010488,
+      "grad_norm": 0.22739523649215698,
+      "learning_rate": 7.3694392694726835e-06,
+      "loss": 0.3869,
       "step": 105015
     },
     {
-      "epoch": 3.69,
-      "learning_rate": 8.429726071398202e-06,
-      "loss": 0.2629,
+      "epoch": 3.7849136843622735,
+      "grad_norm": 0.306366503238678,
+      "learning_rate": 7.367370473314575e-06,
+      "loss": 0.3759,
       "step": 105020
     },
     {
-      "epoch": 3.7,
-      "learning_rate": 8.42759309413385e-06,
-      "loss": 0.2444,
+      "epoch": 3.785093884023498,
+      "grad_norm": 0.21845044195652008,
+      "learning_rate": 7.365301917397629e-06,
+      "loss": 0.3692,
       "step": 105025
     },
     {
-      "epoch": 3.7,
-      "learning_rate": 8.425460332051286e-06,
-      "loss": 0.2447,
+      "epoch": 3.7852740836847225,
+      "grad_norm": 0.2691507041454315,
+      "learning_rate": 7.363233601750033e-06,
+      "loss": 0.3929,
       "step": 105030
     },
     {
-      "epoch": 3.7,
-      "learning_rate": 8.423327785178183e-06,
-      "loss": 0.2685,
+      "epoch": 3.7854542833459472,
+      "grad_norm": 0.23008035123348236,
+      "learning_rate": 7.361165526399963e-06,
+      "loss": 0.3841,
       "step": 105035
     },
     {
-      "epoch": 3.7,
-      "learning_rate": 8.421195453542255e-06,
-      "loss": 0.2732,
+      "epoch": 3.785634483007172,
+      "grad_norm": 0.21267521381378174,
+      "learning_rate": 7.359097691375596e-06,
+      "loss": 0.3342,
       "step": 105040
     },
     {
-      "epoch": 3.7,
-      "learning_rate": 8.419063337171176e-06,
-      "loss": 0.2691,
+      "epoch": 3.7858146826683967,
+      "grad_norm": 0.25716471672058105,
+      "learning_rate": 7.357030096705103e-06,
+      "loss": 0.4005,
       "step": 105045
     },
     {
-      "epoch": 3.7,
-      "learning_rate": 8.416931436092635e-06,
-      "loss": 0.2742,
+      "epoch": 3.7859948823296214,
+      "grad_norm": 0.25578397512435913,
+      "learning_rate": 7.354962742416674e-06,
+      "loss": 0.3783,
       "step": 105050
     },
     {
-      "epoch": 3.7,
-      "learning_rate": 8.4147997503343e-06,
-      "loss": 0.2566,
+      "epoch": 3.7861750819908457,
+      "grad_norm": 0.25349554419517517,
+      "learning_rate": 7.352895628538445e-06,
+      "loss": 0.3775,
       "step": 105055
     },
     {
-      "epoch": 3.7,
-      "learning_rate": 8.412668279923868e-06,
-      "loss": 0.248,
+      "epoch": 3.7863552816520705,
+      "grad_norm": 0.22552327811717987,
+      "learning_rate": 7.350828755098604e-06,
+      "loss": 0.3725,
       "step": 105060
     },
     {
-      "epoch": 3.7,
-      "learning_rate": 8.41053702488901e-06,
-      "loss": 0.2583,
+      "epoch": 3.786535481313295,
+      "grad_norm": 0.20090581476688385,
+      "learning_rate": 7.348762122125305e-06,
+      "loss": 0.3946,
       "step": 105065
     },
     {
-      "epoch": 3.7,
-      "learning_rate": 8.408405985257387e-06,
-      "loss": 0.2467,
+      "epoch": 3.7867156809745195,
+      "grad_norm": 0.24854734539985657,
+      "learning_rate": 7.346695729646705e-06,
+      "loss": 0.402,
       "step": 105070
     },
     {
-      "epoch": 3.7,
-      "learning_rate": 8.406275161056675e-06,
-      "loss": 0.2544,
+      "epoch": 3.7868958806357442,
+      "grad_norm": 0.2817821502685547,
+      "learning_rate": 7.344629577690956e-06,
+      "loss": 0.3533,
       "step": 105075
     },
     {
-      "epoch": 3.7,
-      "learning_rate": 8.404144552314555e-06,
-      "loss": 0.2703,
+      "epoch": 3.787076080296969,
+      "grad_norm": 0.24129050970077515,
+      "learning_rate": 7.3425636662862115e-06,
+      "loss": 0.3783,
       "step": 105080
     },
     {
-      "epoch": 3.7,
-      "learning_rate": 8.402014159058683e-06,
-      "loss": 0.2647,
+      "epoch": 3.7872562799581937,
+      "grad_norm": 0.2622377574443817,
+      "learning_rate": 7.340497995460613e-06,
+      "loss": 0.4121,
       "step": 105085
     },
     {
-      "epoch": 3.7,
-      "learning_rate": 8.399883981316713e-06,
-      "loss": 0.2664,
+      "epoch": 3.7874364796194184,
+      "grad_norm": 0.22572429478168488,
+      "learning_rate": 7.338432565242314e-06,
+      "loss": 0.3947,
       "step": 105090
     },
     {
-      "epoch": 3.7,
-      "learning_rate": 8.397754019116307e-06,
-      "loss": 0.2453,
+      "epoch": 3.787616679280643,
+      "grad_norm": 0.23703832924365997,
+      "learning_rate": 7.3363673756594555e-06,
+      "loss": 0.3583,
       "step": 105095
     },
     {
-      "epoch": 3.7,
-      "learning_rate": 8.395624272485128e-06,
-      "loss": 0.2837,
+      "epoch": 3.7877968789418675,
+      "grad_norm": 0.24465423822402954,
+      "learning_rate": 7.3343024267401725e-06,
+      "loss": 0.4008,
       "step": 105100
     },
     {
-      "epoch": 3.7,
-      "learning_rate": 8.393494741450828e-06,
-      "loss": 0.2511,
+      "epoch": 3.787977078603092,
+      "grad_norm": 0.22674879431724548,
+      "learning_rate": 7.332237718512594e-06,
+      "loss": 0.3801,
       "step": 105105
     },
     {
-      "epoch": 3.7,
-      "learning_rate": 8.391365426041054e-06,
-      "loss": 0.2686,
+      "epoch": 3.788157278264317,
+      "grad_norm": 0.24080802500247955,
+      "learning_rate": 7.330173251004851e-06,
+      "loss": 0.3732,
       "step": 105110
     },
     {
-      "epoch": 3.7,
-      "learning_rate": 8.389236326283445e-06,
-      "loss": 0.2467,
+      "epoch": 3.7883374779255417,
+      "grad_norm": 0.28059399127960205,
+      "learning_rate": 7.328109024245086e-06,
+      "loss": 0.4045,
       "step": 105115
     },
     {
-      "epoch": 3.7,
-      "learning_rate": 8.387107442205663e-06,
-      "loss": 0.2503,
+      "epoch": 3.788517677586766,
+      "grad_norm": 0.19681565463542938,
+      "learning_rate": 7.326045038261411e-06,
+      "loss": 0.3517,
       "step": 105120
     },
     {
-      "epoch": 3.7,
-      "learning_rate": 8.384978773835336e-06,
-      "loss": 0.247,
+      "epoch": 3.7886978772479907,
+      "grad_norm": 0.2852540612220764,
+      "learning_rate": 7.3239812930819524e-06,
+      "loss": 0.4153,
       "step": 105125
     },
     {
-      "epoch": 3.7,
-      "learning_rate": 8.38285032120012e-06,
-      "loss": 0.2505,
+      "epoch": 3.7888780769092154,
+      "grad_norm": 0.28406253457069397,
+      "learning_rate": 7.321917788734825e-06,
+      "loss": 0.3899,
       "step": 105130
     },
     {
-      "epoch": 3.7,
-      "learning_rate": 8.38072208432763e-06,
-      "loss": 0.2562,
+      "epoch": 3.78905827657044,
+      "grad_norm": 0.2611198127269745,
+      "learning_rate": 7.31985452524814e-06,
+      "loss": 0.3949,
       "step": 105135
     },
     {
-      "epoch": 3.7,
-      "learning_rate": 8.37859406324552e-06,
-      "loss": 0.2503,
+      "epoch": 3.789238476231665,
+      "grad_norm": 0.24562934041023254,
+      "learning_rate": 7.31779150265002e-06,
+      "loss": 0.3793,
       "step": 105140
     },
     {
-      "epoch": 3.7,
-      "learning_rate": 8.376466257981413e-06,
-      "loss": 0.259,
+      "epoch": 3.789418675892889,
+      "grad_norm": 0.27355340123176575,
+      "learning_rate": 7.315728720968576e-06,
+      "loss": 0.3941,
       "step": 105145
     },
     {
-      "epoch": 3.7,
-      "learning_rate": 8.374338668562928e-06,
-      "loss": 0.2572,
+      "epoch": 3.789598875554114,
+      "grad_norm": 0.20609021186828613,
+      "learning_rate": 7.313666180231888e-06,
+      "loss": 0.3649,
       "step": 105150
     },
     {
-      "epoch": 3.7,
-      "learning_rate": 8.372211295017707e-06,
-      "loss": 0.262,
+      "epoch": 3.7897790752153386,
+      "grad_norm": 0.21863946318626404,
+      "learning_rate": 7.311603880468082e-06,
+      "loss": 0.4042,
       "step": 105155
     },
     {
-      "epoch": 3.7,
-      "learning_rate": 8.370084137373363e-06,
-      "loss": 0.2496,
+      "epoch": 3.7899592748765634,
+      "grad_norm": 0.22044524550437927,
+      "learning_rate": 7.3095418217052405e-06,
+      "loss": 0.3829,
       "step": 105160
     },
     {
-      "epoch": 3.7,
-      "learning_rate": 8.367957195657517e-06,
-      "loss": 0.2508,
+      "epoch": 3.7901394745377877,
+      "grad_norm": 0.23821088671684265,
+      "learning_rate": 7.3074800039714844e-06,
+      "loss": 0.3746,
       "step": 105165
     },
     {
-      "epoch": 3.7,
-      "learning_rate": 8.365830469897778e-06,
-      "loss": 0.2497,
+      "epoch": 3.7903196741990124,
+      "grad_norm": 0.21885710954666138,
+      "learning_rate": 7.305418427294877e-06,
+      "loss": 0.342,
       "step": 105170
     },
     {
-      "epoch": 3.7,
-      "learning_rate": 8.363703960121777e-06,
-      "loss": 0.2672,
+      "epoch": 3.790499873860237,
+      "grad_norm": 0.26055842638015747,
+      "learning_rate": 7.303357091703511e-06,
+      "loss": 0.3668,
       "step": 105175
     },
     {
-      "epoch": 3.7,
-      "learning_rate": 8.36157766635711e-06,
-      "loss": 0.2408,
+      "epoch": 3.790680073521462,
+      "grad_norm": 0.21390856802463531,
+      "learning_rate": 7.301295997225488e-06,
+      "loss": 0.3783,
       "step": 105180
     },
     {
-      "epoch": 3.7,
-      "learning_rate": 8.359451588631397e-06,
-      "loss": 0.2604,
+      "epoch": 3.7908602731826866,
+      "grad_norm": 0.20063291490077972,
+      "learning_rate": 7.299235143888878e-06,
+      "loss": 0.373,
       "step": 105185
     },
     {
-      "epoch": 3.7,
-      "learning_rate": 8.357325726972231e-06,
-      "loss": 0.2554,
+      "epoch": 3.7910404728439113,
+      "grad_norm": 0.19434435665607452,
+      "learning_rate": 7.297174531721762e-06,
+      "loss": 0.4074,
       "step": 105190
     },
     {
-      "epoch": 3.7,
-      "learning_rate": 8.35520008140723e-06,
-      "loss": 0.2394,
+      "epoch": 3.7912206725051356,
+      "grad_norm": 0.2075873613357544,
+      "learning_rate": 7.295114160752217e-06,
+      "loss": 0.3656,
       "step": 105195
     },
     {
-      "epoch": 3.7,
-      "learning_rate": 8.353074651963987e-06,
-      "loss": 0.239,
+      "epoch": 3.7914008721663603,
+      "grad_norm": 0.2969796657562256,
+      "learning_rate": 7.293054031008306e-06,
+      "loss": 0.4043,
       "step": 105200
     },
     {
-      "epoch": 3.7,
-      "learning_rate": 8.3509494386701e-06,
-      "loss": 0.2537,
+      "epoch": 3.791581071827585,
+      "grad_norm": 0.3087260127067566,
+      "learning_rate": 7.290994142518115e-06,
+      "loss": 0.3599,
       "step": 105205
     },
     {
-      "epoch": 3.7,
-      "learning_rate": 8.348824441553155e-06,
-      "loss": 0.2547,
+      "epoch": 3.7917612714888094,
+      "grad_norm": 0.2378859966993332,
+      "learning_rate": 7.288934495309699e-06,
+      "loss": 0.3884,
       "step": 105210
     },
     {
-      "epoch": 3.7,
-      "learning_rate": 8.346699660640759e-06,
-      "loss": 0.2668,
+      "epoch": 3.791941471150034,
+      "grad_norm": 0.292811781167984,
+      "learning_rate": 7.286875089411119e-06,
+      "loss": 0.3768,
       "step": 105215
     },
     {
-      "epoch": 3.7,
-      "learning_rate": 8.344575095960491e-06,
-      "loss": 0.2744,
+      "epoch": 3.792121670811259,
+      "grad_norm": 0.2711928188800812,
+      "learning_rate": 7.284815924850441e-06,
+      "loss": 0.3985,
       "step": 105220
     },
     {
-      "epoch": 3.7,
-      "learning_rate": 8.34245074753994e-06,
-      "loss": 0.2562,
+      "epoch": 3.7923018704724836,
+      "grad_norm": 0.23554451763629913,
+      "learning_rate": 7.282757001655713e-06,
+      "loss": 0.3652,
       "step": 105225
     },
     {
-      "epoch": 3.7,
-      "learning_rate": 8.34032661540668e-06,
-      "loss": 0.2451,
+      "epoch": 3.7924820701337083,
+      "grad_norm": 0.24152545630931854,
+      "learning_rate": 7.280698319854984e-06,
+      "loss": 0.4081,
       "step": 105230
     },
     {
-      "epoch": 3.7,
-      "learning_rate": 8.338202699588299e-06,
-      "loss": 0.2601,
+      "epoch": 3.792662269794933,
+      "grad_norm": 0.30207541584968567,
+      "learning_rate": 7.2786398794763235e-06,
+      "loss": 0.4096,
       "step": 105235
     },
     {
-      "epoch": 3.7,
-      "learning_rate": 8.336079000112385e-06,
-      "loss": 0.244,
+      "epoch": 3.7928424694561573,
+      "grad_norm": 0.24256369471549988,
+      "learning_rate": 7.27658168054775e-06,
+      "loss": 0.3791,
       "step": 105240
     },
     {
-      "epoch": 3.7,
-      "learning_rate": 8.333955517006496e-06,
-      "loss": 0.266,
+      "epoch": 3.793022669117382,
+      "grad_norm": 0.221241295337677,
+      "learning_rate": 7.274523723097329e-06,
+      "loss": 0.3753,
       "step": 105245
     },
     {
-      "epoch": 3.7,
-      "learning_rate": 8.331832250298215e-06,
-      "loss": 0.2506,
+      "epoch": 3.793202868778607,
+      "grad_norm": 0.2517562508583069,
+      "learning_rate": 7.272466007153086e-06,
+      "loss": 0.4043,
       "step": 105250
     },
     {
-      "epoch": 3.7,
-      "learning_rate": 8.329709200015111e-06,
-      "loss": 0.269,
+      "epoch": 3.793383068439831,
+      "grad_norm": 0.2802235782146454,
+      "learning_rate": 7.270408532743059e-06,
+      "loss": 0.3501,
       "step": 105255
     },
     {
-      "epoch": 3.7,
-      "learning_rate": 8.327586366184747e-06,
-      "loss": 0.2455,
+      "epoch": 3.793563268101056,
+      "grad_norm": 0.28124722838401794,
+      "learning_rate": 7.268351299895295e-06,
+      "loss": 0.3683,
       "step": 105260
     },
     {
-      "epoch": 3.7,
-      "learning_rate": 8.325463748834677e-06,
-      "loss": 0.2603,
+      "epoch": 3.7937434677622806,
+      "grad_norm": 0.21457895636558533,
+      "learning_rate": 7.266294308637805e-06,
+      "loss": 0.3829,
       "step": 105265
     },
     {
-      "epoch": 3.7,
-      "learning_rate": 8.323341347992481e-06,
-      "loss": 0.2528,
+      "epoch": 3.7939236674235053,
+      "grad_norm": 0.2182459533214569,
+      "learning_rate": 7.264237558998615e-06,
+      "loss": 0.3997,
       "step": 105270
     },
     {
-      "epoch": 3.7,
-      "learning_rate": 8.321219163685706e-06,
-      "loss": 0.2537,
+      "epoch": 3.79410386708473,
+      "grad_norm": 0.22549273073673248,
+      "learning_rate": 7.262181051005762e-06,
+      "loss": 0.3851,
       "step": 105275
     },
     {
-      "epoch": 3.7,
-      "learning_rate": 8.31909719594191e-06,
-      "loss": 0.2582,
+      "epoch": 3.7942840667459548,
+      "grad_norm": 0.23273131251335144,
+      "learning_rate": 7.260124784687256e-06,
+      "loss": 0.4087,
       "step": 105280
     },
     {
-      "epoch": 3.7,
-      "learning_rate": 8.316975444788632e-06,
-      "loss": 0.254,
+      "epoch": 3.794464266407179,
+      "grad_norm": 0.22543184459209442,
+      "learning_rate": 7.258068760071115e-06,
+      "loss": 0.3824,
       "step": 105285
     },
     {
-      "epoch": 3.7,
-      "learning_rate": 8.314853910253445e-06,
-      "loss": 0.2834,
+      "epoch": 3.7946444660684038,
+      "grad_norm": 0.39207062125205994,
+      "learning_rate": 7.256012977185356e-06,
+      "loss": 0.4229,
       "step": 105290
     },
     {
-      "epoch": 3.7,
-      "learning_rate": 8.312732592363876e-06,
-      "loss": 0.2707,
+      "epoch": 3.7948246657296285,
+      "grad_norm": 0.24338999390602112,
+      "learning_rate": 7.253957436057973e-06,
+      "loss": 0.3539,
       "step": 105295
     },
     {
-      "epoch": 3.7,
-      "learning_rate": 8.310611491147482e-06,
-      "loss": 0.2676,
+      "epoch": 3.795004865390853,
+      "grad_norm": 0.20618541538715363,
+      "learning_rate": 7.251902136716996e-06,
+      "loss": 0.3645,
       "step": 105300
     },
     {
-      "epoch": 3.7,
-      "learning_rate": 8.308490606631794e-06,
-      "loss": 0.2672,
+      "epoch": 3.7951850650520775,
+      "grad_norm": 0.24109040200710297,
+      "learning_rate": 7.249847079190414e-06,
+      "loss": 0.3805,
       "step": 105305
     },
     {
-      "epoch": 3.71,
-      "learning_rate": 8.306369938844364e-06,
-      "loss": 0.2639,
+      "epoch": 3.7953652647133023,
+      "grad_norm": 0.2463836818933487,
+      "learning_rate": 7.247792263506228e-06,
+      "loss": 0.3682,
       "step": 105310
     },
     {
-      "epoch": 3.71,
-      "learning_rate": 8.304249487812715e-06,
-      "loss": 0.2749,
+      "epoch": 3.795545464374527,
+      "grad_norm": 0.23828503489494324,
+      "learning_rate": 7.2457376896924365e-06,
+      "loss": 0.3785,
       "step": 105315
     },
     {
-      "epoch": 3.71,
-      "learning_rate": 8.302129253564385e-06,
-      "loss": 0.2716,
+      "epoch": 3.7957256640357517,
+      "grad_norm": 0.2362687587738037,
+      "learning_rate": 7.243683357777023e-06,
+      "loss": 0.3743,
       "step": 105320
     },
     {
-      "epoch": 3.71,
-      "learning_rate": 8.300009236126897e-06,
-      "loss": 0.2357,
+      "epoch": 3.7959058636969765,
+      "grad_norm": 0.2598309814929962,
+      "learning_rate": 7.2416292677879946e-06,
+      "loss": 0.3675,
       "step": 105325
     },
     {
-      "epoch": 3.71,
-      "learning_rate": 8.29788943552779e-06,
-      "loss": 0.2526,
+      "epoch": 3.7960860633582008,
+      "grad_norm": 0.2267717868089676,
+      "learning_rate": 7.239575419753339e-06,
+      "loss": 0.3687,
       "step": 105330
     },
     {
-      "epoch": 3.71,
-      "learning_rate": 8.29576985179458e-06,
-      "loss": 0.2675,
+      "epoch": 3.7962662630194255,
+      "grad_norm": 0.20179912447929382,
+      "learning_rate": 7.237521813701012e-06,
+      "loss": 0.3649,
       "step": 105335
     },
     {
-      "epoch": 3.71,
-      "learning_rate": 8.293650484954782e-06,
-      "loss": 0.2506,
+      "epoch": 3.7964464626806502,
+      "grad_norm": 0.29818323254585266,
+      "learning_rate": 7.235468449659019e-06,
+      "loss": 0.3729,
       "step": 105340
     },
     {
-      "epoch": 3.71,
-      "learning_rate": 8.291531335035932e-06,
-      "loss": 0.2853,
+      "epoch": 3.796626662341875,
+      "grad_norm": 0.26517441868782043,
+      "learning_rate": 7.233415327655321e-06,
+      "loss": 0.3762,
       "step": 105345
     },
     {
-      "epoch": 3.71,
-      "learning_rate": 8.289412402065527e-06,
-      "loss": 0.2501,
+      "epoch": 3.7968068620030992,
+      "grad_norm": 0.22521330416202545,
+      "learning_rate": 7.231362447717915e-06,
+      "loss": 0.3459,
       "step": 105350
     },
     {
-      "epoch": 3.71,
-      "learning_rate": 8.287293686071099e-06,
-      "loss": 0.2818,
+      "epoch": 3.796987061664324,
+      "grad_norm": 0.20847706496715546,
+      "learning_rate": 7.229309809874749e-06,
+      "loss": 0.3975,
       "step": 105355
     },
     {
-      "epoch": 3.71,
-      "learning_rate": 8.28517518708014e-06,
-      "loss": 0.2614,
+      "epoch": 3.7971672613255487,
+      "grad_norm": 0.22983822226524353,
+      "learning_rate": 7.22725741415379e-06,
+      "loss": 0.3864,
       "step": 105360
     },
     {
-      "epoch": 3.71,
-      "learning_rate": 8.283056905120173e-06,
-      "loss": 0.2488,
+      "epoch": 3.7973474609867734,
+      "grad_norm": 0.24204209446907043,
+      "learning_rate": 7.225205260583013e-06,
+      "loss": 0.3853,
       "step": 105365
     },
     {
-      "epoch": 3.71,
-      "learning_rate": 8.280938840218697e-06,
-      "loss": 0.2743,
+      "epoch": 3.797527660647998,
+      "grad_norm": 0.3181317448616028,
+      "learning_rate": 7.223153349190373e-06,
+      "loss": 0.388,
       "step": 105370
     },
     {
-      "epoch": 3.71,
-      "learning_rate": 8.278820992403207e-06,
-      "loss": 0.2463,
+      "epoch": 3.7977078603092225,
+      "grad_norm": 0.22117455303668976,
+      "learning_rate": 7.221101680003828e-06,
+      "loss": 0.4012,
       "step": 105375
     },
     {
-      "epoch": 3.71,
-      "learning_rate": 8.276703361701205e-06,
-      "loss": 0.2606,
+      "epoch": 3.797888059970447,
+      "grad_norm": 0.1853034645318985,
+      "learning_rate": 7.219050253051329e-06,
+      "loss": 0.3584,
       "step": 105380
     },
     {
-      "epoch": 3.71,
-      "learning_rate": 8.274585948140193e-06,
-      "loss": 0.262,
+      "epoch": 3.798068259631672,
+      "grad_norm": 0.20408424735069275,
+      "learning_rate": 7.216999068360822e-06,
+      "loss": 0.3564,
       "step": 105385
     },
     {
-      "epoch": 3.71,
-      "learning_rate": 8.272468751747661e-06,
-      "loss": 0.2712,
+      "epoch": 3.7982484592928967,
+      "grad_norm": 0.2169559895992279,
+      "learning_rate": 7.214948125960266e-06,
+      "loss": 0.3362,
       "step": 105390
     },
     {
-      "epoch": 3.71,
-      "learning_rate": 8.270351772551096e-06,
-      "loss": 0.2589,
+      "epoch": 3.798428658954121,
+      "grad_norm": 0.2969491481781006,
+      "learning_rate": 7.212897425877599e-06,
+      "loss": 0.3837,
       "step": 105395
     },
     {
-      "epoch": 3.71,
-      "learning_rate": 8.268235010577985e-06,
-      "loss": 0.2597,
+      "epoch": 3.7986088586153457,
+      "grad_norm": 0.2599042057991028,
+      "learning_rate": 7.2108469681407605e-06,
+      "loss": 0.3692,
       "step": 105400
     },
     {
-      "epoch": 3.71,
-      "learning_rate": 8.266118465855813e-06,
-      "loss": 0.2744,
+      "epoch": 3.7987890582765704,
+      "grad_norm": 0.19389642775058746,
+      "learning_rate": 7.208796752777691e-06,
+      "loss": 0.356,
       "step": 105405
     },
     {
-      "epoch": 3.71,
-      "learning_rate": 8.264002138412071e-06,
-      "loss": 0.2636,
+      "epoch": 3.798969257937795,
+      "grad_norm": 0.22755640745162964,
+      "learning_rate": 7.206746779816317e-06,
+      "loss": 0.4012,
       "step": 105410
     },
     {
-      "epoch": 3.71,
-      "learning_rate": 8.261886028274232e-06,
-      "loss": 0.2698,
+      "epoch": 3.79914945759902,
+      "grad_norm": 0.25583744049072266,
+      "learning_rate": 7.204697049284567e-06,
+      "loss": 0.3835,
       "step": 105415
     },
     {
-      "epoch": 3.71,
-      "learning_rate": 8.259770135469765e-06,
-      "loss": 0.255,
+      "epoch": 3.799329657260244,
+      "grad_norm": 0.26095953583717346,
+      "learning_rate": 7.202647561210382e-06,
+      "loss": 0.373,
       "step": 105420
     },
     {
-      "epoch": 3.71,
-      "learning_rate": 8.257654460026162e-06,
-      "loss": 0.2355,
+      "epoch": 3.799509856921469,
+      "grad_norm": 0.26644977927207947,
+      "learning_rate": 7.200598315621679e-06,
+      "loss": 0.4386,
       "step": 105425
     },
     {
-      "epoch": 3.71,
-      "learning_rate": 8.255539001970877e-06,
-      "loss": 0.2424,
+      "epoch": 3.7996900565826937,
+      "grad_norm": 0.22700461745262146,
+      "learning_rate": 7.198549312546379e-06,
+      "loss": 0.4308,
       "step": 105430
     },
     {
-      "epoch": 3.71,
-      "learning_rate": 8.253423761331386e-06,
-      "loss": 0.2442,
+      "epoch": 3.7998702562439184,
+      "grad_norm": 0.20378275215625763,
+      "learning_rate": 7.196500552012397e-06,
+      "loss": 0.3812,
       "step": 105435
     },
     {
-      "epoch": 3.71,
-      "learning_rate": 8.251308738135144e-06,
-      "loss": 0.2664,
+      "epoch": 3.8000504559051427,
+      "grad_norm": 0.2918340861797333,
+      "learning_rate": 7.194452034047639e-06,
+      "loss": 0.3411,
       "step": 105440
     },
     {
-      "epoch": 3.71,
-      "learning_rate": 8.249193932409626e-06,
-      "loss": 0.262,
+      "epoch": 3.8002306555663674,
+      "grad_norm": 0.19135360419750214,
+      "learning_rate": 7.1924037586800415e-06,
+      "loss": 0.3652,
       "step": 105445
     },
     {
-      "epoch": 3.71,
-      "learning_rate": 8.247079344182284e-06,
-      "loss": 0.2698,
+      "epoch": 3.800410855227592,
+      "grad_norm": 0.24061061441898346,
+      "learning_rate": 7.190355725937487e-06,
+      "loss": 0.3725,
       "step": 105450
     },
     {
-      "epoch": 3.71,
-      "learning_rate": 8.244964973480573e-06,
-      "loss": 0.2772,
+      "epoch": 3.800591054888817,
+      "grad_norm": 0.27103692293167114,
+      "learning_rate": 7.1883079358478825e-06,
+      "loss": 0.3687,
       "step": 105455
     },
     {
-      "epoch": 3.71,
-      "learning_rate": 8.242850820331946e-06,
-      "loss": 0.2527,
+      "epoch": 3.8007712545500416,
+      "grad_norm": 0.2294936627149582,
+      "learning_rate": 7.186260388439137e-06,
+      "loss": 0.3556,
       "step": 105460
     },
     {
-      "epoch": 3.71,
-      "learning_rate": 8.240736884763866e-06,
-      "loss": 0.2283,
+      "epoch": 3.8009514542112663,
+      "grad_norm": 0.2741897702217102,
+      "learning_rate": 7.184213083739147e-06,
+      "loss": 0.3703,
       "step": 105465
     },
     {
-      "epoch": 3.71,
-      "learning_rate": 8.238623166803775e-06,
-      "loss": 0.251,
+      "epoch": 3.8011316538724906,
+      "grad_norm": 0.2626726031303406,
+      "learning_rate": 7.182166021775805e-06,
+      "loss": 0.379,
       "step": 105470
     },
     {
-      "epoch": 3.71,
-      "learning_rate": 8.236509666479106e-06,
-      "loss": 0.2775,
+      "epoch": 3.8013118535337154,
+      "grad_norm": 0.292074054479599,
+      "learning_rate": 7.1801192025770015e-06,
+      "loss": 0.3627,
       "step": 105475
     },
     {
-      "epoch": 3.71,
-      "learning_rate": 8.23439638381732e-06,
-      "loss": 0.2506,
+      "epoch": 3.80149205319494,
+      "grad_norm": 0.25104910135269165,
+      "learning_rate": 7.178072626170615e-06,
+      "loss": 0.3784,
       "step": 105480
     },
     {
-      "epoch": 3.71,
-      "learning_rate": 8.232283318845849e-06,
-      "loss": 0.2641,
+      "epoch": 3.8016722528561644,
+      "grad_norm": 0.24408739805221558,
+      "learning_rate": 7.176026292584548e-06,
+      "loss": 0.3792,
       "step": 105485
     },
     {
-      "epoch": 3.71,
-      "learning_rate": 8.230170471592127e-06,
-      "loss": 0.2681,
+      "epoch": 3.801852452517389,
+      "grad_norm": 0.238882914185524,
+      "learning_rate": 7.1739802018466695e-06,
+      "loss": 0.3824,
       "step": 105490
     },
     {
-      "epoch": 3.71,
-      "learning_rate": 8.228057842083584e-06,
-      "loss": 0.2476,
+      "epoch": 3.802032652178614,
+      "grad_norm": 0.22461140155792236,
+      "learning_rate": 7.171934353984863e-06,
+      "loss": 0.3616,
       "step": 105495
     },
     {
-      "epoch": 3.71,
-      "learning_rate": 8.225945430347662e-06,
-      "loss": 0.2553,
+      "epoch": 3.8022128518398386,
+      "grad_norm": 0.2754288911819458,
+      "learning_rate": 7.169888749026995e-06,
+      "loss": 0.3749,
       "step": 105500
     },
     {
-      "epoch": 3.71,
-      "eval_loss": 0.2516160011291504,
-      "eval_runtime": 10.5476,
-      "eval_samples_per_second": 9.481,
-      "eval_steps_per_second": 9.481,
+      "epoch": 3.8022128518398386,
+      "eval_loss": 0.4286453127861023,
+      "eval_runtime": 3.5307,
+      "eval_samples_per_second": 28.323,
+      "eval_steps_per_second": 7.081,
       "step": 105500
     },
     {
-      "epoch": 3.71,
-      "learning_rate": 8.223833236411784e-06,
-      "loss": 0.2675,
+      "epoch": 3.8023930515010633,
+      "grad_norm": 0.1857483983039856,
+      "learning_rate": 7.167843387000936e-06,
+      "loss": 0.4028,
       "step": 105505
     },
     {
-      "epoch": 3.71,
-      "learning_rate": 8.221721260303373e-06,
-      "loss": 0.2427,
+      "epoch": 3.802573251162288,
+      "grad_norm": 0.18854497373104095,
+      "learning_rate": 7.165798267934565e-06,
+      "loss": 0.3894,
       "step": 105510
     },
     {
-      "epoch": 3.71,
-      "learning_rate": 8.219609502049846e-06,
-      "loss": 0.242,
+      "epoch": 3.8027534508235123,
+      "grad_norm": 0.2537713050842285,
+      "learning_rate": 7.163753391855749e-06,
+      "loss": 0.3789,
       "step": 105515
     },
     {
-      "epoch": 3.71,
-      "learning_rate": 8.217497961678644e-06,
-      "loss": 0.2726,
+      "epoch": 3.802933650484737,
+      "grad_norm": 0.2627655565738678,
+      "learning_rate": 7.161708758792324e-06,
+      "loss": 0.3729,
       "step": 105520
     },
     {
-      "epoch": 3.71,
-      "learning_rate": 8.21538663921717e-06,
-      "loss": 0.2623,
+      "epoch": 3.803113850145962,
+      "grad_norm": 0.2551818788051605,
+      "learning_rate": 7.15966436877217e-06,
+      "loss": 0.3955,
       "step": 105525
     },
     {
-      "epoch": 3.71,
-      "learning_rate": 8.213275534692841e-06,
-      "loss": 0.2397,
+      "epoch": 3.803294049807186,
+      "grad_norm": 0.2461671382188797,
+      "learning_rate": 7.157620221823127e-06,
+      "loss": 0.3978,
       "step": 105530
     },
     {
-      "epoch": 3.71,
-      "learning_rate": 8.211164648133055e-06,
-      "loss": 0.2779,
+      "epoch": 3.803474249468411,
+      "grad_norm": 0.29957154393196106,
+      "learning_rate": 7.155576317973061e-06,
+      "loss": 0.4128,
       "step": 105535
     },
     {
-      "epoch": 3.71,
-      "learning_rate": 8.20905397956524e-06,
-      "loss": 0.2635,
+      "epoch": 3.8036544491296356,
+      "grad_norm": 0.2094818353652954,
+      "learning_rate": 7.153532657249823e-06,
+      "loss": 0.3306,
       "step": 105540
     },
     {
-      "epoch": 3.71,
-      "learning_rate": 8.206943529016794e-06,
-      "loss": 0.2533,
+      "epoch": 3.8038346487908603,
+      "grad_norm": 0.21276958286762238,
+      "learning_rate": 7.1514892396812335e-06,
+      "loss": 0.3973,
       "step": 105545
     },
     {
-      "epoch": 3.71,
-      "learning_rate": 8.204833296515122e-06,
-      "loss": 0.2796,
+      "epoch": 3.804014848452085,
+      "grad_norm": 0.2489434778690338,
+      "learning_rate": 7.149446065295151e-06,
+      "loss": 0.3623,
       "step": 105550
     },
     {
-      "epoch": 3.71,
-      "learning_rate": 8.20272328208761e-06,
-      "loss": 0.2774,
+      "epoch": 3.8041950481133098,
+      "grad_norm": 0.2391531616449356,
+      "learning_rate": 7.147403134119412e-06,
+      "loss": 0.3777,
       "step": 105555
     },
     {
-      "epoch": 3.71,
-      "learning_rate": 8.200613485761677e-06,
-      "loss": 0.2724,
+      "epoch": 3.804375247774534,
+      "grad_norm": 0.2393515408039093,
+      "learning_rate": 7.145360446181848e-06,
+      "loss": 0.3775,
       "step": 105560
     },
     {
-      "epoch": 3.71,
-      "learning_rate": 8.198503907564694e-06,
-      "loss": 0.2661,
+      "epoch": 3.804555447435759,
+      "grad_norm": 0.26830413937568665,
+      "learning_rate": 7.1433180015102936e-06,
+      "loss": 0.3694,
       "step": 105565
     },
     {
-      "epoch": 3.71,
-      "learning_rate": 8.196394547524078e-06,
-      "loss": 0.2391,
+      "epoch": 3.8047356470969835,
+      "grad_norm": 0.3024479150772095,
+      "learning_rate": 7.141275800132563e-06,
+      "loss": 0.3865,
       "step": 105570
     },
     {
-      "epoch": 3.71,
-      "learning_rate": 8.194285405667193e-06,
-      "loss": 0.2699,
+      "epoch": 3.804915846758208,
+      "grad_norm": 0.21776083111763,
+      "learning_rate": 7.1392338420765005e-06,
+      "loss": 0.3866,
       "step": 105575
     },
     {
-      "epoch": 3.71,
-      "learning_rate": 8.192176482021446e-06,
-      "loss": 0.2464,
+      "epoch": 3.8050960464194326,
+      "grad_norm": 0.24332380294799805,
+      "learning_rate": 7.137192127369921e-06,
+      "loss": 0.3977,
       "step": 105580
     },
     {
-      "epoch": 3.71,
-      "learning_rate": 8.190067776614214e-06,
-      "loss": 0.2504,
+      "epoch": 3.8052762460806573,
+      "grad_norm": 0.1978878378868103,
+      "learning_rate": 7.13515065604064e-06,
+      "loss": 0.3775,
       "step": 105585
     },
     {
-      "epoch": 3.71,
-      "learning_rate": 8.187959289472869e-06,
-      "loss": 0.256,
+      "epoch": 3.805456445741882,
+      "grad_norm": 0.22366763651371002,
+      "learning_rate": 7.1331094281164715e-06,
+      "loss": 0.3749,
       "step": 105590
     },
     {
-      "epoch": 3.72,
-      "learning_rate": 8.185851020624788e-06,
-      "loss": 0.2432,
+      "epoch": 3.8056366454031068,
+      "grad_norm": 0.20870821177959442,
+      "learning_rate": 7.13106844362523e-06,
+      "loss": 0.4025,
       "step": 105595
     },
     {
-      "epoch": 3.72,
-      "learning_rate": 8.183742970097357e-06,
-      "loss": 0.2638,
+      "epoch": 3.8058168450643315,
+      "grad_norm": 0.2318304181098938,
+      "learning_rate": 7.129027702594713e-06,
+      "loss": 0.4165,
       "step": 105600
     },
     {
-      "epoch": 3.72,
-      "learning_rate": 8.181635137917942e-06,
-      "loss": 0.2533,
+      "epoch": 3.8059970447255558,
+      "grad_norm": 0.22088027000427246,
+      "learning_rate": 7.126987205052738e-06,
+      "loss": 0.3893,
       "step": 105605
     },
     {
-      "epoch": 3.72,
-      "learning_rate": 8.179527524113904e-06,
-      "loss": 0.244,
+      "epoch": 3.8061772443867805,
+      "grad_norm": 0.27171891927719116,
+      "learning_rate": 7.124946951027103e-06,
+      "loss": 0.3769,
       "step": 105610
     },
     {
-      "epoch": 3.72,
-      "learning_rate": 8.17742012871262e-06,
-      "loss": 0.2672,
+      "epoch": 3.8063574440480052,
+      "grad_norm": 0.21456721425056458,
+      "learning_rate": 7.12290694054561e-06,
+      "loss": 0.3962,
       "step": 105615
     },
     {
-      "epoch": 3.72,
-      "learning_rate": 8.175312951741446e-06,
-      "loss": 0.2632,
+      "epoch": 3.80653764370923,
+      "grad_norm": 0.2533538043498993,
+      "learning_rate": 7.120867173636042e-06,
+      "loss": 0.3834,
       "step": 105620
     },
     {
-      "epoch": 3.72,
-      "learning_rate": 8.173205993227751e-06,
-      "loss": 0.259,
+      "epoch": 3.8067178433704543,
+      "grad_norm": 0.2441474199295044,
+      "learning_rate": 7.118827650326193e-06,
+      "loss": 0.3718,
       "step": 105625
     },
     {
-      "epoch": 3.72,
-      "learning_rate": 8.17109925319888e-06,
-      "loss": 0.2579,
+      "epoch": 3.806898043031679,
+      "grad_norm": 0.2645309865474701,
+      "learning_rate": 7.116788370643873e-06,
+      "loss": 0.3572,
       "step": 105630
     },
     {
-      "epoch": 3.72,
-      "learning_rate": 8.168992731682202e-06,
-      "loss": 0.2476,
+      "epoch": 3.8070782426929037,
+      "grad_norm": 0.2198515087366104,
+      "learning_rate": 7.1147493346168385e-06,
+      "loss": 0.3757,
       "step": 105635
     },
     {
-      "epoch": 3.72,
-      "learning_rate": 8.166886428705058e-06,
-      "loss": 0.2701,
+      "epoch": 3.8072584423541285,
+      "grad_norm": 0.23640519380569458,
+      "learning_rate": 7.112710542272874e-06,
+      "loss": 0.3703,
       "step": 105640
     },
     {
-      "epoch": 3.72,
-      "learning_rate": 8.164780344294806e-06,
-      "loss": 0.2374,
+      "epoch": 3.807438642015353,
+      "grad_norm": 0.28243306279182434,
+      "learning_rate": 7.110671993639772e-06,
+      "loss": 0.427,
       "step": 105645
     },
     {
-      "epoch": 3.72,
-      "learning_rate": 8.162674478478774e-06,
-      "loss": 0.2599,
+      "epoch": 3.8076188416765775,
+      "grad_norm": 0.2905183434486389,
+      "learning_rate": 7.108633688745303e-06,
+      "loss": 0.4278,
       "step": 105650
     },
     {
-      "epoch": 3.72,
-      "learning_rate": 8.16056883128433e-06,
-      "loss": 0.2632,
+      "epoch": 3.8077990413378022,
+      "grad_norm": 0.2647572457790375,
+      "learning_rate": 7.106595627617235e-06,
+      "loss": 0.3962,
       "step": 105655
     },
     {
-      "epoch": 3.72,
-      "learning_rate": 8.158463402738798e-06,
-      "loss": 0.227,
+      "epoch": 3.807979240999027,
+      "grad_norm": 0.22848838567733765,
+      "learning_rate": 7.104557810283338e-06,
+      "loss": 0.3891,
       "step": 105660
     },
     {
-      "epoch": 3.72,
-      "learning_rate": 8.156358192869521e-06,
-      "loss": 0.2211,
+      "epoch": 3.8081594406602517,
+      "grad_norm": 0.23743629455566406,
+      "learning_rate": 7.102520236771368e-06,
+      "loss": 0.3594,
       "step": 105665
     },
     {
-      "epoch": 3.72,
-      "learning_rate": 8.154253201703824e-06,
-      "loss": 0.2356,
+      "epoch": 3.808339640321476,
+      "grad_norm": 0.2258959710597992,
+      "learning_rate": 7.100482907109102e-06,
+      "loss": 0.3635,
       "step": 105670
     },
     {
-      "epoch": 3.72,
-      "learning_rate": 8.15214842926905e-06,
-      "loss": 0.2578,
+      "epoch": 3.8085198399827007,
+      "grad_norm": 0.2394886165857315,
+      "learning_rate": 7.098445821324293e-06,
+      "loss": 0.3791,
       "step": 105675
     },
     {
-      "epoch": 3.72,
-      "learning_rate": 8.15004387559253e-06,
-      "loss": 0.2664,
+      "epoch": 3.8087000396439255,
+      "grad_norm": 0.2250211089849472,
+      "learning_rate": 7.096408979444691e-06,
+      "loss": 0.403,
       "step": 105680
     },
     {
-      "epoch": 3.72,
-      "learning_rate": 8.147939540701588e-06,
-      "loss": 0.2439,
+      "epoch": 3.80888023930515,
+      "grad_norm": 0.1884530633687973,
+      "learning_rate": 7.094372381498052e-06,
+      "loss": 0.3308,
       "step": 105685
     },
     {
-      "epoch": 3.72,
-      "learning_rate": 8.145835424623539e-06,
-      "loss": 0.2558,
+      "epoch": 3.809060438966375,
+      "grad_norm": 0.24882015585899353,
+      "learning_rate": 7.092336027512115e-06,
+      "loss": 0.3892,
       "step": 105690
     },
     {
-      "epoch": 3.72,
-      "learning_rate": 8.143731527385711e-06,
-      "loss": 0.2669,
+      "epoch": 3.8092406386275997,
+      "grad_norm": 0.21837691962718964,
+      "learning_rate": 7.0902999175146396e-06,
+      "loss": 0.3885,
       "step": 105695
     },
     {
-      "epoch": 3.72,
-      "learning_rate": 8.141627849015427e-06,
-      "loss": 0.2658,
+      "epoch": 3.809420838288824,
+      "grad_norm": 0.27154818177223206,
+      "learning_rate": 7.0882640515333585e-06,
+      "loss": 0.3667,
       "step": 105700
     },
     {
-      "epoch": 3.72,
-      "learning_rate": 8.139524389539993e-06,
-      "loss": 0.2569,
+      "epoch": 3.8096010379500487,
+      "grad_norm": 0.19823803007602692,
+      "learning_rate": 7.0862284295960144e-06,
+      "loss": 0.3472,
       "step": 105705
     },
     {
-      "epoch": 3.72,
-      "learning_rate": 8.137421148986716e-06,
-      "loss": 0.251,
+      "epoch": 3.8097812376112734,
+      "grad_norm": 0.2487536519765854,
+      "learning_rate": 7.08419305173034e-06,
+      "loss": 0.3867,
       "step": 105710
     },
     {
-      "epoch": 3.72,
-      "learning_rate": 8.135318127382917e-06,
-      "loss": 0.2741,
+      "epoch": 3.8099614372724977,
+      "grad_norm": 0.2396322339773178,
+      "learning_rate": 7.082157917964058e-06,
+      "loss": 0.3602,
       "step": 105715
     },
     {
-      "epoch": 3.72,
-      "learning_rate": 8.133215324755903e-06,
-      "loss": 0.2686,
+      "epoch": 3.8101416369337224,
+      "grad_norm": 0.22832445800304413,
+      "learning_rate": 7.0801230283249145e-06,
+      "loss": 0.3922,
       "step": 105720
     },
     {
-      "epoch": 3.72,
-      "learning_rate": 8.131112741132962e-06,
-      "loss": 0.2404,
+      "epoch": 3.810321836594947,
+      "grad_norm": 0.22182156145572662,
+      "learning_rate": 7.078088382840631e-06,
+      "loss": 0.3613,
       "step": 105725
     },
     {
-      "epoch": 3.72,
-      "learning_rate": 8.129010376541412e-06,
-      "loss": 0.271,
+      "epoch": 3.810502036256172,
+      "grad_norm": 0.26968759298324585,
+      "learning_rate": 7.0760539815389075e-06,
+      "loss": 0.4139,
       "step": 105730
     },
     {
-      "epoch": 3.72,
-      "learning_rate": 8.126908231008536e-06,
-      "loss": 0.2655,
+      "epoch": 3.8106822359173966,
+      "grad_norm": 0.19509482383728027,
+      "learning_rate": 7.0740198244474895e-06,
+      "loss": 0.356,
       "step": 105735
     },
     {
-      "epoch": 3.72,
-      "learning_rate": 8.124806304561647e-06,
-      "loss": 0.2301,
+      "epoch": 3.8108624355786214,
+      "grad_norm": 0.24696923792362213,
+      "learning_rate": 7.071985911594078e-06,
+      "loss": 0.3766,
       "step": 105740
     },
     {
-      "epoch": 3.72,
-      "learning_rate": 8.122704597228017e-06,
-      "loss": 0.2741,
+      "epoch": 3.8110426352398457,
+      "grad_norm": 0.2208133488893509,
+      "learning_rate": 7.069952243006389e-06,
+      "loss": 0.3709,
       "step": 105745
     },
     {
-      "epoch": 3.72,
-      "learning_rate": 8.120603109034954e-06,
-      "loss": 0.2468,
+      "epoch": 3.8112228349010704,
+      "grad_norm": 0.2347523719072342,
+      "learning_rate": 7.067918818712127e-06,
+      "loss": 0.3837,
       "step": 105750
     },
     {
-      "epoch": 3.72,
-      "learning_rate": 8.118501840009738e-06,
-      "loss": 0.269,
+      "epoch": 3.811403034562295,
+      "grad_norm": 0.25173941254615784,
+      "learning_rate": 7.065885638738995e-06,
+      "loss": 0.4095,
       "step": 105755
     },
     {
-      "epoch": 3.72,
-      "learning_rate": 8.116400790179648e-06,
-      "loss": 0.2589,
+      "epoch": 3.8115832342235194,
+      "grad_norm": 0.19811011850833893,
+      "learning_rate": 7.063852703114704e-06,
+      "loss": 0.4196,
       "step": 105760
     },
     {
-      "epoch": 3.72,
-      "learning_rate": 8.114299959571959e-06,
-      "loss": 0.2432,
+      "epoch": 3.811763433884744,
+      "grad_norm": 0.22163262963294983,
+      "learning_rate": 7.061820011866949e-06,
+      "loss": 0.3861,
       "step": 105765
     },
     {
-      "epoch": 3.72,
-      "learning_rate": 8.112199348213965e-06,
-      "loss": 0.2401,
+      "epoch": 3.811943633545969,
+      "grad_norm": 0.21935562789440155,
+      "learning_rate": 7.059787565023421e-06,
+      "loss": 0.3796,
       "step": 105770
     },
     {
-      "epoch": 3.72,
-      "learning_rate": 8.110098956132934e-06,
-      "loss": 0.2524,
+      "epoch": 3.8121238332071936,
+      "grad_norm": 0.1883041262626648,
+      "learning_rate": 7.0577553626118145e-06,
+      "loss": 0.3783,
       "step": 105775
     },
     {
-      "epoch": 3.72,
-      "learning_rate": 8.107998783356128e-06,
-      "loss": 0.2679,
+      "epoch": 3.8123040328684183,
+      "grad_norm": 0.2626688778400421,
+      "learning_rate": 7.05572340465982e-06,
+      "loss": 0.4043,
       "step": 105780
     },
     {
-      "epoch": 3.72,
-      "learning_rate": 8.105898829910833e-06,
-      "loss": 0.2746,
+      "epoch": 3.812484232529643,
+      "grad_norm": 0.2265564501285553,
+      "learning_rate": 7.053691691195111e-06,
+      "loss": 0.3838,
       "step": 105785
     },
     {
-      "epoch": 3.72,
-      "learning_rate": 8.1037990958243e-06,
-      "loss": 0.253,
+      "epoch": 3.8126644321908674,
+      "grad_norm": 0.3275195360183716,
+      "learning_rate": 7.051660222245388e-06,
+      "loss": 0.3578,
       "step": 105790
     },
     {
-      "epoch": 3.72,
-      "learning_rate": 8.101699581123809e-06,
-      "loss": 0.2622,
+      "epoch": 3.812844631852092,
+      "grad_norm": 0.24849045276641846,
+      "learning_rate": 7.049628997838315e-06,
+      "loss": 0.3943,
       "step": 105795
     },
     {
-      "epoch": 3.72,
-      "learning_rate": 8.099600285836614e-06,
-      "loss": 0.2589,
+      "epoch": 3.813024831513317,
+      "grad_norm": 0.22452738881111145,
+      "learning_rate": 7.047598018001575e-06,
+      "loss": 0.3827,
       "step": 105800
     },
     {
-      "epoch": 3.72,
-      "learning_rate": 8.097501209989963e-06,
-      "loss": 0.2516,
+      "epoch": 3.813205031174541,
+      "grad_norm": 0.2721640467643738,
+      "learning_rate": 7.045567282762836e-06,
+      "loss": 0.3972,
       "step": 105805
     },
     {
-      "epoch": 3.72,
-      "learning_rate": 8.095402353611125e-06,
-      "loss": 0.2557,
+      "epoch": 3.813385230835766,
+      "grad_norm": 0.22152206301689148,
+      "learning_rate": 7.043536792149757e-06,
+      "loss": 0.354,
       "step": 105810
     },
     {
-      "epoch": 3.72,
-      "learning_rate": 8.093303716727351e-06,
-      "loss": 0.2641,
+      "epoch": 3.8135654304969906,
+      "grad_norm": 0.19255799055099487,
+      "learning_rate": 7.04150654619003e-06,
+      "loss": 0.3711,
       "step": 105815
     },
     {
-      "epoch": 3.72,
-      "learning_rate": 8.091205299365884e-06,
-      "loss": 0.2774,
+      "epoch": 3.8137456301582153,
+      "grad_norm": 0.2830601930618286,
+      "learning_rate": 7.039476544911291e-06,
+      "loss": 0.38,
       "step": 105820
     },
     {
-      "epoch": 3.72,
-      "learning_rate": 8.089107101553964e-06,
-      "loss": 0.2497,
+      "epoch": 3.81392582981944,
+      "grad_norm": 0.20766660571098328,
+      "learning_rate": 7.037446788341198e-06,
+      "loss": 0.3799,
       "step": 105825
     },
     {
-      "epoch": 3.72,
-      "learning_rate": 8.087009123318853e-06,
-      "loss": 0.2542,
+      "epoch": 3.814106029480665,
+      "grad_norm": 0.2430235594511032,
+      "learning_rate": 7.035417276507425e-06,
+      "loss": 0.3873,
       "step": 105830
     },
     {
-      "epoch": 3.72,
-      "learning_rate": 8.084911364687784e-06,
-      "loss": 0.2485,
+      "epoch": 3.814286229141889,
+      "grad_norm": 0.29118457436561584,
+      "learning_rate": 7.0333880094376055e-06,
+      "loss": 0.3836,
       "step": 105835
     },
     {
-      "epoch": 3.72,
-      "learning_rate": 8.082813825687982e-06,
-      "loss": 0.2743,
+      "epoch": 3.814466428803114,
+      "grad_norm": 0.20183739066123962,
+      "learning_rate": 7.031358987159409e-06,
+      "loss": 0.3727,
       "step": 105840
     },
     {
-      "epoch": 3.72,
-      "learning_rate": 8.080716506346696e-06,
-      "loss": 0.2506,
+      "epoch": 3.8146466284643386,
+      "grad_norm": 0.22086220979690552,
+      "learning_rate": 7.029330209700463e-06,
+      "loss": 0.3697,
       "step": 105845
     },
     {
-      "epoch": 3.72,
-      "learning_rate": 8.078619406691163e-06,
-      "loss": 0.2794,
+      "epoch": 3.8148268281255633,
+      "grad_norm": 0.24414370954036713,
+      "learning_rate": 7.0273016770884045e-06,
+      "loss": 0.3763,
       "step": 105850
     },
     {
-      "epoch": 3.72,
-      "learning_rate": 8.076522526748605e-06,
-      "loss": 0.261,
+      "epoch": 3.8150070277867876,
+      "grad_norm": 0.20783692598342896,
+      "learning_rate": 7.025273389350886e-06,
+      "loss": 0.3389,
       "step": 105855
     },
     {
-      "epoch": 3.72,
-      "learning_rate": 8.074425866546242e-06,
-      "loss": 0.238,
+      "epoch": 3.8151872274480123,
+      "grad_norm": 0.29034414887428284,
+      "learning_rate": 7.023245346515541e-06,
+      "loss": 0.4052,
       "step": 105860
     },
     {
-      "epoch": 3.72,
-      "learning_rate": 8.072329426111313e-06,
-      "loss": 0.2335,
+      "epoch": 3.815367427109237,
+      "grad_norm": 0.27145785093307495,
+      "learning_rate": 7.021217548609999e-06,
+      "loss": 0.3902,
       "step": 105865
     },
     {
-      "epoch": 3.72,
-      "learning_rate": 8.070233205471031e-06,
-      "loss": 0.2563,
+      "epoch": 3.8155476267704618,
+      "grad_norm": 0.2717854976654053,
+      "learning_rate": 7.019189995661884e-06,
+      "loss": 0.374,
       "step": 105870
     },
     {
-      "epoch": 3.72,
-      "learning_rate": 8.068137204652612e-06,
-      "loss": 0.2484,
+      "epoch": 3.8157278264316865,
+      "grad_norm": 0.30218860507011414,
+      "learning_rate": 7.017162687698817e-06,
+      "loss": 0.3719,
       "step": 105875
     },
     {
-      "epoch": 3.73,
-      "learning_rate": 8.066041423683265e-06,
-      "loss": 0.2721,
+      "epoch": 3.815908026092911,
+      "grad_norm": 0.2221565693616867,
+      "learning_rate": 7.015135624748434e-06,
+      "loss": 0.3678,
       "step": 105880
     },
     {
-      "epoch": 3.73,
-      "learning_rate": 8.063945862590219e-06,
-      "loss": 0.2546,
+      "epoch": 3.8160882257541355,
+      "grad_norm": 0.2444906085729599,
+      "learning_rate": 7.013108806838348e-06,
+      "loss": 0.4113,
       "step": 105885
     },
     {
-      "epoch": 3.73,
-      "learning_rate": 8.061850521400676e-06,
-      "loss": 0.2546,
+      "epoch": 3.8162684254153603,
+      "grad_norm": 0.2825545370578766,
+      "learning_rate": 7.011082233996169e-06,
+      "loss": 0.3969,
       "step": 105890
     },
     {
-      "epoch": 3.73,
-      "learning_rate": 8.059755400141833e-06,
-      "loss": 0.2708,
+      "epoch": 3.816448625076585,
+      "grad_norm": 0.226121723651886,
+      "learning_rate": 7.009055906249515e-06,
+      "loss": 0.3798,
       "step": 105895
     },
     {
-      "epoch": 3.73,
-      "learning_rate": 8.0576604988409e-06,
-      "loss": 0.2522,
+      "epoch": 3.8166288247378093,
+      "grad_norm": 0.2496255487203598,
+      "learning_rate": 7.007029823625982e-06,
+      "loss": 0.3724,
       "step": 105900
     },
     {
-      "epoch": 3.73,
-      "learning_rate": 8.055565817525088e-06,
-      "loss": 0.2467,
+      "epoch": 3.816809024399034,
+      "grad_norm": 0.2551683187484741,
+      "learning_rate": 7.0050039861531915e-06,
+      "loss": 0.3912,
       "step": 105905
     },
     {
-      "epoch": 3.73,
-      "learning_rate": 8.053471356221589e-06,
-      "loss": 0.2535,
+      "epoch": 3.8169892240602588,
+      "grad_norm": 0.24302713572978973,
+      "learning_rate": 7.002978393858747e-06,
+      "loss": 0.376,
       "step": 105910
     },
     {
-      "epoch": 3.73,
-      "learning_rate": 8.051377114957595e-06,
-      "loss": 0.2481,
+      "epoch": 3.8171694237214835,
+      "grad_norm": 0.25622105598449707,
+      "learning_rate": 7.00095304677022e-06,
+      "loss": 0.374,
       "step": 105915
     },
     {
-      "epoch": 3.73,
-      "learning_rate": 8.049283093760294e-06,
-      "loss": 0.2395,
+      "epoch": 3.8173496233827082,
+      "grad_norm": 0.23344071209430695,
+      "learning_rate": 6.9989279449152316e-06,
+      "loss": 0.3773,
       "step": 105920
     },
     {
-      "epoch": 3.73,
-      "learning_rate": 8.047189292656885e-06,
-      "loss": 0.222,
+      "epoch": 3.8175298230439325,
+      "grad_norm": 0.24558846652507782,
+      "learning_rate": 6.996903088321366e-06,
+      "loss": 0.3956,
       "step": 105925
     },
     {
-      "epoch": 3.73,
-      "learning_rate": 8.045095711674553e-06,
-      "loss": 0.278,
+      "epoch": 3.8177100227051572,
+      "grad_norm": 0.2149125039577484,
+      "learning_rate": 6.994878477016209e-06,
+      "loss": 0.3839,
       "step": 105930
     },
     {
-      "epoch": 3.73,
-      "learning_rate": 8.043002350840479e-06,
-      "loss": 0.2369,
+      "epoch": 3.817890222366382,
+      "grad_norm": 0.2317209243774414,
+      "learning_rate": 6.992854111027347e-06,
+      "loss": 0.403,
       "step": 105935
     },
     {
-      "epoch": 3.73,
-      "learning_rate": 8.040909210181838e-06,
-      "loss": 0.2597,
+      "epoch": 3.8180704220276067,
+      "grad_norm": 0.29228413105010986,
+      "learning_rate": 6.9908299903823555e-06,
+      "loss": 0.3927,
       "step": 105940
     },
     {
-      "epoch": 3.73,
-      "learning_rate": 8.038816289725821e-06,
-      "loss": 0.2459,
+      "epoch": 3.818250621688831,
+      "grad_norm": 0.20377010107040405,
+      "learning_rate": 6.988806115108826e-06,
+      "loss": 0.3415,
       "step": 105945
     },
     {
-      "epoch": 3.73,
-      "learning_rate": 8.036723589499586e-06,
-      "loss": 0.2616,
+      "epoch": 3.8184308213500557,
+      "grad_norm": 0.24830646812915802,
+      "learning_rate": 6.986782485234322e-06,
+      "loss": 0.38,
       "step": 105950
     },
     {
-      "epoch": 3.73,
-      "learning_rate": 8.034631109530319e-06,
-      "loss": 0.2529,
+      "epoch": 3.8186110210112805,
+      "grad_norm": 0.2752242982387543,
+      "learning_rate": 6.9847591007864225e-06,
+      "loss": 0.3827,
       "step": 105955
     },
     {
-      "epoch": 3.73,
-      "learning_rate": 8.032538849845195e-06,
-      "loss": 0.2854,
+      "epoch": 3.818791220672505,
+      "grad_norm": 0.2355724275112152,
+      "learning_rate": 6.9827359617926945e-06,
+      "loss": 0.3878,
       "step": 105960
     },
     {
-      "epoch": 3.73,
-      "learning_rate": 8.030446810471371e-06,
-      "loss": 0.2783,
+      "epoch": 3.81897142033373,
+      "grad_norm": 0.21855174005031586,
+      "learning_rate": 6.980713068280698e-06,
+      "loss": 0.3636,
       "step": 105965
     },
     {
-      "epoch": 3.73,
-      "learning_rate": 8.028354991436014e-06,
-      "loss": 0.2637,
+      "epoch": 3.8191516199949547,
+      "grad_norm": 0.24916981160640717,
+      "learning_rate": 6.978690420277989e-06,
+      "loss": 0.3912,
       "step": 105970
     },
     {
-      "epoch": 3.73,
-      "learning_rate": 8.02626339276627e-06,
-      "loss": 0.2745,
+      "epoch": 3.819331819656179,
+      "grad_norm": 0.25882112979888916,
+      "learning_rate": 6.976668017812144e-06,
+      "loss": 0.4247,
       "step": 105975
     },
     {
-      "epoch": 3.73,
-      "learning_rate": 8.02417201448932e-06,
-      "loss": 0.2596,
+      "epoch": 3.8195120193174037,
+      "grad_norm": 0.24076975882053375,
+      "learning_rate": 6.974645860910706e-06,
+      "loss": 0.3626,
       "step": 105980
     },
     {
-      "epoch": 3.73,
-      "learning_rate": 8.022080856632312e-06,
-      "loss": 0.261,
+      "epoch": 3.8196922189786284,
+      "grad_norm": 0.25258931517601013,
+      "learning_rate": 6.9726239496012285e-06,
+      "loss": 0.3866,
       "step": 105985
     },
     {
-      "epoch": 3.73,
-      "learning_rate": 8.019989919222393e-06,
-      "loss": 0.2877,
+      "epoch": 3.8198724186398527,
+      "grad_norm": 0.24138487875461578,
+      "learning_rate": 6.97060228391126e-06,
+      "loss": 0.3732,
       "step": 105990
     },
     {
-      "epoch": 3.73,
-      "learning_rate": 8.017899202286705e-06,
-      "loss": 0.2326,
+      "epoch": 3.8200526183010775,
+      "grad_norm": 0.23699279129505157,
+      "learning_rate": 6.968580863868334e-06,
+      "loss": 0.3936,
       "step": 105995
     },
     {
-      "epoch": 3.73,
-      "learning_rate": 8.015808705852412e-06,
-      "loss": 0.2643,
+      "epoch": 3.820232817962302,
+      "grad_norm": 0.28243473172187805,
+      "learning_rate": 6.9665596895000155e-06,
+      "loss": 0.3651,
       "step": 106000
     },
     {
-      "epoch": 3.73,
-      "eval_loss": 0.25173264741897583,
-      "eval_runtime": 10.5541,
-      "eval_samples_per_second": 9.475,
-      "eval_steps_per_second": 9.475,
+      "epoch": 3.820232817962302,
+      "eval_loss": 0.42869871854782104,
+      "eval_runtime": 3.5415,
+      "eval_samples_per_second": 28.237,
+      "eval_steps_per_second": 7.059,
       "step": 106000
     },
     {
-      "epoch": 3.73,
-      "learning_rate": 8.013718429946645e-06,
-      "loss": 0.2656,
+      "epoch": 3.820413017623527,
+      "grad_norm": 0.2416362762451172,
+      "learning_rate": 6.964538760833836e-06,
+      "loss": 0.3882,
       "step": 106005
     },
     {
-      "epoch": 3.73,
-      "learning_rate": 8.011628374596555e-06,
-      "loss": 0.2496,
+      "epoch": 3.8205932172847517,
+      "grad_norm": 0.2267039567232132,
+      "learning_rate": 6.962518077897306e-06,
+      "loss": 0.392,
       "step": 106010
     },
     {
-      "epoch": 3.73,
-      "learning_rate": 8.009538539829267e-06,
-      "loss": 0.2472,
+      "epoch": 3.8207734169459764,
+      "grad_norm": 0.24580471217632294,
+      "learning_rate": 6.960497640717986e-06,
+      "loss": 0.3695,
       "step": 106015
     },
     {
-      "epoch": 3.73,
-      "learning_rate": 8.007448925671934e-06,
-      "loss": 0.2544,
+      "epoch": 3.8209536166072007,
+      "grad_norm": 0.23100297152996063,
+      "learning_rate": 6.958477449323384e-06,
+      "loss": 0.389,
       "step": 106020
     },
     {
-      "epoch": 3.73,
-      "learning_rate": 8.005359532151677e-06,
-      "loss": 0.2508,
+      "epoch": 3.8211338162684254,
+      "grad_norm": 0.25452324748039246,
+      "learning_rate": 6.956457503741049e-06,
+      "loss": 0.3714,
       "step": 106025
     },
     {
-      "epoch": 3.73,
-      "learning_rate": 8.003270359295628e-06,
-      "loss": 0.2504,
+      "epoch": 3.82131401592965,
+      "grad_norm": 0.20147554576396942,
+      "learning_rate": 6.954437803998479e-06,
+      "loss": 0.3673,
       "step": 106030
     },
     {
-      "epoch": 3.73,
-      "learning_rate": 8.001181407130903e-06,
-      "loss": 0.2553,
+      "epoch": 3.8214942155908744,
+      "grad_norm": 0.26207849383354187,
+      "learning_rate": 6.952418350123194e-06,
+      "loss": 0.3893,
       "step": 106035
     },
     {
-      "epoch": 3.73,
-      "learning_rate": 7.99909267568464e-06,
-      "loss": 0.2288,
+      "epoch": 3.821674415252099,
+      "grad_norm": 0.25192660093307495,
+      "learning_rate": 6.950399142142722e-06,
+      "loss": 0.4313,
       "step": 106040
     },
     {
-      "epoch": 3.73,
-      "learning_rate": 7.99700416498396e-06,
-      "loss": 0.2656,
+      "epoch": 3.821854614913324,
+      "grad_norm": 0.2420777529478073,
+      "learning_rate": 6.9483801800845685e-06,
+      "loss": 0.3608,
       "step": 106045
     },
     {
-      "epoch": 3.73,
-      "learning_rate": 7.994915875055971e-06,
-      "loss": 0.2688,
+      "epoch": 3.8220348145745486,
+      "grad_norm": 0.19895605742931366,
+      "learning_rate": 6.9463614639762366e-06,
+      "loss": 0.3602,
       "step": 106050
     },
     {
-      "epoch": 3.73,
-      "learning_rate": 7.992827805927785e-06,
-      "loss": 0.2497,
+      "epoch": 3.8222150142357734,
+      "grad_norm": 0.3234860599040985,
+      "learning_rate": 6.944342993845237e-06,
+      "loss": 0.3955,
       "step": 106055
     },
     {
-      "epoch": 3.73,
-      "learning_rate": 7.990739957626525e-06,
-      "loss": 0.2433,
+      "epoch": 3.822395213896998,
+      "grad_norm": 0.23543211817741394,
+      "learning_rate": 6.942324769719061e-06,
+      "loss": 0.3795,
       "step": 106060
     },
     {
-      "epoch": 3.73,
-      "learning_rate": 7.988652330179303e-06,
-      "loss": 0.268,
+      "epoch": 3.8225754135582224,
+      "grad_norm": 0.18973630666732788,
+      "learning_rate": 6.9403067916252205e-06,
+      "loss": 0.3613,
       "step": 106065
     },
     {
-      "epoch": 3.73,
-      "learning_rate": 7.986564923613214e-06,
-      "loss": 0.2484,
+      "epoch": 3.822755613219447,
+      "grad_norm": 0.242222398519516,
+      "learning_rate": 6.938289059591205e-06,
+      "loss": 0.3813,
       "step": 106070
     },
     {
-      "epoch": 3.73,
-      "learning_rate": 7.98447773795537e-06,
-      "loss": 0.2794,
+      "epoch": 3.822935812880672,
+      "grad_norm": 0.23931699991226196,
+      "learning_rate": 6.936271573644501e-06,
+      "loss": 0.3862,
       "step": 106075
     },
     {
-      "epoch": 3.73,
-      "learning_rate": 7.982390773232873e-06,
-      "loss": 0.2717,
+      "epoch": 3.823116012541896,
+      "grad_norm": 0.23971091210842133,
+      "learning_rate": 6.934254333812601e-06,
+      "loss": 0.4044,
       "step": 106080
     },
     {
-      "epoch": 3.73,
-      "learning_rate": 7.980304029472816e-06,
-      "loss": 0.2683,
+      "epoch": 3.823296212203121,
+      "grad_norm": 0.26793283224105835,
+      "learning_rate": 6.932237340122982e-06,
+      "loss": 0.4126,
       "step": 106085
     },
     {
-      "epoch": 3.73,
-      "learning_rate": 7.978217506702284e-06,
-      "loss": 0.2694,
+      "epoch": 3.8234764118643456,
+      "grad_norm": 0.24026966094970703,
+      "learning_rate": 6.930220592603137e-06,
+      "loss": 0.4094,
       "step": 106090
     },
     {
-      "epoch": 3.73,
-      "learning_rate": 7.976131204948393e-06,
-      "loss": 0.2623,
+      "epoch": 3.8236566115255703,
+      "grad_norm": 0.24495449662208557,
+      "learning_rate": 6.928204091280549e-06,
+      "loss": 0.4011,
       "step": 106095
     },
     {
-      "epoch": 3.73,
-      "learning_rate": 7.974045124238217e-06,
-      "loss": 0.2564,
+      "epoch": 3.823836811186795,
+      "grad_norm": 0.20799663662910461,
+      "learning_rate": 6.926187836182663e-06,
+      "loss": 0.3974,
       "step": 106100
     },
     {
-      "epoch": 3.73,
-      "learning_rate": 7.971959264598844e-06,
-      "loss": 0.2586,
+      "epoch": 3.82401701084802,
+      "grad_norm": 0.2459585964679718,
+      "learning_rate": 6.924171827336975e-06,
+      "loss": 0.3619,
       "step": 106105
     },
     {
-      "epoch": 3.73,
-      "learning_rate": 7.969873626057348e-06,
-      "loss": 0.2623,
+      "epoch": 3.824197210509244,
+      "grad_norm": 0.2440251260995865,
+      "learning_rate": 6.9221560647709485e-06,
+      "loss": 0.3862,
       "step": 106110
     },
     {
-      "epoch": 3.73,
-      "learning_rate": 7.967788208640823e-06,
-      "loss": 0.2551,
+      "epoch": 3.824377410170469,
+      "grad_norm": 0.29908034205436707,
+      "learning_rate": 6.920140548512038e-06,
+      "loss": 0.4097,
       "step": 106115
     },
     {
-      "epoch": 3.73,
-      "learning_rate": 7.96570301237635e-06,
-      "loss": 0.2527,
+      "epoch": 3.8245576098316936,
+      "grad_norm": 0.304340660572052,
+      "learning_rate": 6.9181252785877285e-06,
+      "loss": 0.3573,
       "step": 106120
     },
     {
-      "epoch": 3.73,
-      "learning_rate": 7.963618037290999e-06,
-      "loss": 0.2826,
+      "epoch": 3.8247378094929183,
+      "grad_norm": 0.20545925199985504,
+      "learning_rate": 6.916110255025443e-06,
+      "loss": 0.365,
       "step": 106125
     },
     {
-      "epoch": 3.73,
-      "learning_rate": 7.961533283411829e-06,
-      "loss": 0.2554,
+      "epoch": 3.8249180091541426,
+      "grad_norm": 0.20274123549461365,
+      "learning_rate": 6.914095477852664e-06,
+      "loss": 0.3642,
       "step": 106130
     },
     {
-      "epoch": 3.73,
-      "learning_rate": 7.95944875076593e-06,
-      "loss": 0.2638,
+      "epoch": 3.8250982088153673,
+      "grad_norm": 0.21836356818675995,
+      "learning_rate": 6.912080947096833e-06,
+      "loss": 0.3782,
       "step": 106135
     },
     {
-      "epoch": 3.73,
-      "learning_rate": 7.95736443938036e-06,
-      "loss": 0.27,
+      "epoch": 3.825278408476592,
+      "grad_norm": 0.21938830614089966,
+      "learning_rate": 6.910066662785394e-06,
+      "loss": 0.389,
       "step": 106140
     },
     {
-      "epoch": 3.73,
-      "learning_rate": 7.955280349282183e-06,
-      "loss": 0.2426,
+      "epoch": 3.825458608137817,
+      "grad_norm": 0.22280332446098328,
+      "learning_rate": 6.908052624945796e-06,
+      "loss": 0.3447,
       "step": 106145
     },
     {
-      "epoch": 3.73,
-      "learning_rate": 7.953196480498446e-06,
-      "loss": 0.247,
+      "epoch": 3.8256388077990415,
+      "grad_norm": 0.25814318656921387,
+      "learning_rate": 6.90603883360548e-06,
+      "loss": 0.3924,
       "step": 106150
     },
     {
-      "epoch": 3.73,
-      "learning_rate": 7.951112833056226e-06,
-      "loss": 0.2563,
+      "epoch": 3.825819007460266,
+      "grad_norm": 0.28167471289634705,
+      "learning_rate": 6.904025288791874e-06,
+      "loss": 0.3889,
       "step": 106155
     },
     {
-      "epoch": 3.74,
-      "learning_rate": 7.949029406982574e-06,
-      "loss": 0.2629,
+      "epoch": 3.8259992071214906,
+      "grad_norm": 0.21864792704582214,
+      "learning_rate": 6.902011990532425e-06,
+      "loss": 0.3953,
       "step": 106160
     },
     {
-      "epoch": 3.74,
-      "learning_rate": 7.946946202304528e-06,
-      "loss": 0.253,
+      "epoch": 3.8261794067827153,
+      "grad_norm": 0.23761038482189178,
+      "learning_rate": 6.89999893885456e-06,
+      "loss": 0.3828,
       "step": 106165
     },
     {
-      "epoch": 3.74,
-      "learning_rate": 7.944863219049155e-06,
-      "loss": 0.2575,
+      "epoch": 3.82635960644394,
+      "grad_norm": 0.23446708917617798,
+      "learning_rate": 6.8979861337857055e-06,
+      "loss": 0.3481,
       "step": 106170
     },
     {
-      "epoch": 3.74,
-      "learning_rate": 7.942780457243487e-06,
-      "loss": 0.268,
+      "epoch": 3.8265398061051643,
+      "grad_norm": 0.25455090403556824,
+      "learning_rate": 6.895973575353287e-06,
+      "loss": 0.38,
       "step": 106175
     },
     {
-      "epoch": 3.74,
-      "learning_rate": 7.94069791691458e-06,
-      "loss": 0.2613,
+      "epoch": 3.826720005766389,
+      "grad_norm": 0.2646729648113251,
+      "learning_rate": 6.893961263584714e-06,
+      "loss": 0.3828,
       "step": 106180
     },
     {
-      "epoch": 3.74,
-      "learning_rate": 7.93861559808946e-06,
-      "loss": 0.2466,
+      "epoch": 3.8269002054276138,
+      "grad_norm": 0.25168514251708984,
+      "learning_rate": 6.891949198507419e-06,
+      "loss": 0.376,
       "step": 106185
     },
     {
-      "epoch": 3.74,
-      "learning_rate": 7.936533500795182e-06,
-      "loss": 0.2552,
+      "epoch": 3.8270804050888385,
+      "grad_norm": 0.24938587844371796,
+      "learning_rate": 6.88993738014882e-06,
+      "loss": 0.3898,
       "step": 106190
     },
     {
-      "epoch": 3.74,
-      "learning_rate": 7.934451625058769e-06,
-      "loss": 0.2545,
+      "epoch": 3.8272606047500632,
+      "grad_norm": 0.24917519092559814,
+      "learning_rate": 6.8879258085363025e-06,
+      "loss": 0.3947,
       "step": 106195
     },
     {
-      "epoch": 3.74,
-      "learning_rate": 7.932369970907256e-06,
-      "loss": 0.2565,
+      "epoch": 3.827440804411288,
+      "grad_norm": 0.2233772575855255,
+      "learning_rate": 6.8859144836972976e-06,
+      "loss": 0.3719,
       "step": 106200
     },
     {
-      "epoch": 3.74,
-      "learning_rate": 7.930288538367665e-06,
-      "loss": 0.2552,
+      "epoch": 3.8276210040725123,
+      "grad_norm": 0.2717728018760681,
+      "learning_rate": 6.883903405659192e-06,
+      "loss": 0.3885,
       "step": 106205
     },
     {
-      "epoch": 3.74,
-      "learning_rate": 7.928207327467033e-06,
-      "loss": 0.2505,
+      "epoch": 3.827801203733737,
+      "grad_norm": 0.24201686680316925,
+      "learning_rate": 6.881892574449411e-06,
+      "loss": 0.3567,
       "step": 106210
     },
     {
-      "epoch": 3.74,
-      "learning_rate": 7.926126338232378e-06,
-      "loss": 0.2556,
+      "epoch": 3.8279814033949617,
+      "grad_norm": 0.22028078138828278,
+      "learning_rate": 6.87988199009533e-06,
+      "loss": 0.3652,
       "step": 106215
     },
     {
-      "epoch": 3.74,
-      "learning_rate": 7.924045570690714e-06,
-      "loss": 0.2675,
+      "epoch": 3.828161603056186,
+      "grad_norm": 0.21828703582286835,
+      "learning_rate": 6.87787165262434e-06,
+      "loss": 0.3752,
       "step": 106220
     },
     {
-      "epoch": 3.74,
-      "learning_rate": 7.921965024869074e-06,
-      "loss": 0.2662,
+      "epoch": 3.8283418027174108,
+      "grad_norm": 0.22384034097194672,
+      "learning_rate": 6.875861562063846e-06,
+      "loss": 0.387,
       "step": 106225
     },
     {
-      "epoch": 3.74,
-      "learning_rate": 7.919884700794452e-06,
-      "loss": 0.2472,
+      "epoch": 3.8285220023786355,
+      "grad_norm": 0.38978999853134155,
+      "learning_rate": 6.873851718441232e-06,
+      "loss": 0.3725,
       "step": 106230
     },
     {
-      "epoch": 3.74,
-      "learning_rate": 7.91780459849388e-06,
-      "loss": 0.2526,
+      "epoch": 3.8287022020398602,
+      "grad_norm": 0.30686643719673157,
+      "learning_rate": 6.8718421217838754e-06,
+      "loss": 0.3894,
       "step": 106235
     },
     {
-      "epoch": 3.74,
-      "learning_rate": 7.91572471799436e-06,
-      "loss": 0.2511,
+      "epoch": 3.828882401701085,
+      "grad_norm": 0.24971897900104523,
+      "learning_rate": 6.869832772119164e-06,
+      "loss": 0.3785,
       "step": 106240
     },
     {
-      "epoch": 3.74,
-      "learning_rate": 7.913645059322886e-06,
-      "loss": 0.244,
+      "epoch": 3.8290626013623097,
+      "grad_norm": 0.26769372820854187,
+      "learning_rate": 6.86782366947446e-06,
+      "loss": 0.3451,
       "step": 106245
     },
     {
-      "epoch": 3.74,
-      "learning_rate": 7.911565622506478e-06,
-      "loss": 0.2755,
+      "epoch": 3.829242801023534,
+      "grad_norm": 0.30026867985725403,
+      "learning_rate": 6.865814813877158e-06,
+      "loss": 0.4113,
       "step": 106250
     },
     {
-      "epoch": 3.74,
-      "learning_rate": 7.909486407572132e-06,
-      "loss": 0.2588,
+      "epoch": 3.8294230006847587,
+      "grad_norm": 0.2826428711414337,
+      "learning_rate": 6.863806205354617e-06,
+      "loss": 0.4001,
       "step": 106255
     },
     {
-      "epoch": 3.74,
-      "learning_rate": 7.907407414546836e-06,
-      "loss": 0.2589,
+      "epoch": 3.8296032003459834,
+      "grad_norm": 0.24307632446289062,
+      "learning_rate": 6.861797843934206e-06,
+      "loss": 0.3537,
       "step": 106260
     },
     {
-      "epoch": 3.74,
-      "learning_rate": 7.905328643457586e-06,
-      "loss": 0.2673,
+      "epoch": 3.8297834000072077,
+      "grad_norm": 0.277296245098114,
+      "learning_rate": 6.859789729643287e-06,
+      "loss": 0.3807,
       "step": 106265
     },
     {
-      "epoch": 3.74,
-      "learning_rate": 7.903250094331385e-06,
-      "loss": 0.2476,
+      "epoch": 3.8299635996684325,
+      "grad_norm": 0.2420312762260437,
+      "learning_rate": 6.857781862509221e-06,
+      "loss": 0.3795,
       "step": 106270
     },
     {
-      "epoch": 3.74,
-      "learning_rate": 7.901171767195215e-06,
-      "loss": 0.2565,
+      "epoch": 3.830143799329657,
+      "grad_norm": 0.17696630954742432,
+      "learning_rate": 6.855774242559359e-06,
+      "loss": 0.3511,
       "step": 106275
     },
     {
-      "epoch": 3.74,
-      "learning_rate": 7.899093662076052e-06,
-      "loss": 0.2415,
+      "epoch": 3.830323998990882,
+      "grad_norm": 0.26137152314186096,
+      "learning_rate": 6.853766869821066e-06,
+      "loss": 0.3418,
       "step": 106280
     },
     {
-      "epoch": 3.74,
-      "learning_rate": 7.897015779000887e-06,
-      "loss": 0.2595,
+      "epoch": 3.8305041986521067,
+      "grad_norm": 0.21335354447364807,
+      "learning_rate": 6.851759744321687e-06,
+      "loss": 0.3742,
       "step": 106285
     },
     {
-      "epoch": 3.74,
-      "learning_rate": 7.894938117996711e-06,
-      "loss": 0.2557,
+      "epoch": 3.8306843983133314,
+      "grad_norm": 0.23139753937721252,
+      "learning_rate": 6.849752866088566e-06,
+      "loss": 0.3503,
       "step": 106290
     },
     {
-      "epoch": 3.74,
-      "learning_rate": 7.89286067909049e-06,
-      "loss": 0.2347,
+      "epoch": 3.8308645979745557,
+      "grad_norm": 0.21770185232162476,
+      "learning_rate": 6.84774623514905e-06,
+      "loss": 0.3839,
       "step": 106295
     },
     {
-      "epoch": 3.74,
-      "learning_rate": 7.890783462309193e-06,
-      "loss": 0.2701,
+      "epoch": 3.8310447976357804,
+      "grad_norm": 0.2551557719707489,
+      "learning_rate": 6.845739851530469e-06,
+      "loss": 0.3913,
       "step": 106300
     },
     {
-      "epoch": 3.74,
-      "learning_rate": 7.888706467679805e-06,
-      "loss": 0.2512,
+      "epoch": 3.831224997297005,
+      "grad_norm": 0.2286597192287445,
+      "learning_rate": 6.843733715260181e-06,
+      "loss": 0.413,
       "step": 106305
     },
     {
-      "epoch": 3.74,
-      "learning_rate": 7.886629695229284e-06,
-      "loss": 0.2537,
+      "epoch": 3.8314051969582295,
+      "grad_norm": 0.22169090807437897,
+      "learning_rate": 6.841727826365493e-06,
+      "loss": 0.379,
       "step": 106310
     },
     {
-      "epoch": 3.74,
-      "learning_rate": 7.884553144984599e-06,
-      "loss": 0.264,
+      "epoch": 3.831585396619454,
+      "grad_norm": 0.240523561835289,
+      "learning_rate": 6.839722184873757e-06,
+      "loss": 0.3675,
       "step": 106315
     },
     {
-      "epoch": 3.74,
-      "learning_rate": 7.882476816972706e-06,
-      "loss": 0.2409,
+      "epoch": 3.831765596280679,
+      "grad_norm": 0.2571893632411957,
+      "learning_rate": 6.8377167908122876e-06,
+      "loss": 0.382,
       "step": 106320
     },
     {
-      "epoch": 3.74,
-      "learning_rate": 7.880400711220576e-06,
-      "loss": 0.2468,
+      "epoch": 3.8319457959419037,
+      "grad_norm": 0.24433764815330505,
+      "learning_rate": 6.8357116442084115e-06,
+      "loss": 0.3755,
       "step": 106325
     },
     {
-      "epoch": 3.74,
-      "learning_rate": 7.878324827755163e-06,
-      "loss": 0.2689,
+      "epoch": 3.8321259956031284,
+      "grad_norm": 0.1973889172077179,
+      "learning_rate": 6.833706745089446e-06,
+      "loss": 0.3574,
       "step": 106330
     },
     {
-      "epoch": 3.74,
-      "learning_rate": 7.876249166603408e-06,
-      "loss": 0.2537,
+      "epoch": 3.832306195264353,
+      "grad_norm": 0.28018200397491455,
+      "learning_rate": 6.831702093482711e-06,
+      "loss": 0.3909,
       "step": 106335
     },
     {
-      "epoch": 3.74,
-      "learning_rate": 7.874173727792275e-06,
-      "loss": 0.2725,
+      "epoch": 3.8324863949255774,
+      "grad_norm": 0.2281164973974228,
+      "learning_rate": 6.82969768941551e-06,
+      "loss": 0.368,
       "step": 106340
     },
     {
-      "epoch": 3.74,
-      "learning_rate": 7.872098511348716e-06,
-      "loss": 0.26,
+      "epoch": 3.832666594586802,
+      "grad_norm": 0.25604715943336487,
+      "learning_rate": 6.827693532915166e-06,
+      "loss": 0.3783,
       "step": 106345
     },
     {
-      "epoch": 3.74,
-      "learning_rate": 7.87002351729967e-06,
-      "loss": 0.258,
+      "epoch": 3.832846794248027,
+      "grad_norm": 0.21953798830509186,
+      "learning_rate": 6.82568962400898e-06,
+      "loss": 0.377,
       "step": 106350
     },
     {
-      "epoch": 3.74,
-      "learning_rate": 7.86794874567208e-06,
-      "loss": 0.2454,
+      "epoch": 3.8330269939092516,
+      "grad_norm": 0.22131898999214172,
+      "learning_rate": 6.823685962724255e-06,
+      "loss": 0.3739,
       "step": 106355
     },
     {
-      "epoch": 3.74,
-      "learning_rate": 7.865874196492878e-06,
-      "loss": 0.2593,
+      "epoch": 3.833207193570476,
+      "grad_norm": 0.3169558048248291,
+      "learning_rate": 6.8216825490882875e-06,
+      "loss": 0.4028,
       "step": 106360
     },
     {
-      "epoch": 3.74,
-      "learning_rate": 7.863799869789016e-06,
-      "loss": 0.2799,
+      "epoch": 3.8333873932317006,
+      "grad_norm": 0.2238478660583496,
+      "learning_rate": 6.819679383128372e-06,
+      "loss": 0.4006,
       "step": 106365
     },
     {
-      "epoch": 3.74,
-      "learning_rate": 7.861725765587419e-06,
-      "loss": 0.2782,
+      "epoch": 3.8335675928929254,
+      "grad_norm": 0.28951504826545715,
+      "learning_rate": 6.817676464871808e-06,
+      "loss": 0.361,
       "step": 106370
     },
     {
-      "epoch": 3.74,
-      "learning_rate": 7.859651883915018e-06,
-      "loss": 0.2602,
+      "epoch": 3.83374779255415,
+      "grad_norm": 0.2536328136920929,
+      "learning_rate": 6.815673794345895e-06,
+      "loss": 0.3859,
       "step": 106375
     },
     {
-      "epoch": 3.74,
-      "learning_rate": 7.857578224798731e-06,
-      "loss": 0.25,
+      "epoch": 3.833927992215375,
+      "grad_norm": 0.2552483379840851,
+      "learning_rate": 6.8136713715778875e-06,
+      "loss": 0.4012,
       "step": 106380
     },
     {
-      "epoch": 3.74,
-      "learning_rate": 7.855504788265505e-06,
-      "loss": 0.2351,
+      "epoch": 3.834108191876599,
+      "grad_norm": 0.2368234395980835,
+      "learning_rate": 6.811669196595094e-06,
+      "loss": 0.3601,
       "step": 106385
     },
     {
-      "epoch": 3.74,
-      "learning_rate": 7.85343157434224e-06,
-      "loss": 0.2632,
+      "epoch": 3.834288391537824,
+      "grad_norm": 0.2486155778169632,
+      "learning_rate": 6.80966726942478e-06,
+      "loss": 0.3964,
       "step": 106390
     },
     {
-      "epoch": 3.74,
-      "learning_rate": 7.851358583055875e-06,
-      "loss": 0.2442,
+      "epoch": 3.8344685911990486,
+      "grad_norm": 0.2343214601278305,
+      "learning_rate": 6.807665590094242e-06,
+      "loss": 0.408,
       "step": 106395
     },
     {
-      "epoch": 3.74,
-      "learning_rate": 7.849285814433312e-06,
-      "loss": 0.2653,
+      "epoch": 3.8346487908602733,
+      "grad_norm": 0.22257140278816223,
+      "learning_rate": 6.805664158630728e-06,
+      "loss": 0.3616,
       "step": 106400
     },
     {
-      "epoch": 3.74,
-      "learning_rate": 7.847213268501478e-06,
-      "loss": 0.2266,
+      "epoch": 3.8348289905214976,
+      "grad_norm": 0.26378926634788513,
+      "learning_rate": 6.803662975061515e-06,
+      "loss": 0.4036,
       "step": 106405
     },
     {
-      "epoch": 3.74,
-      "learning_rate": 7.845140945287271e-06,
-      "loss": 0.2704,
+      "epoch": 3.8350091901827223,
+      "grad_norm": 0.2561560273170471,
+      "learning_rate": 6.801662039413875e-06,
+      "loss": 0.3486,
       "step": 106410
     },
     {
-      "epoch": 3.74,
-      "learning_rate": 7.843068844817597e-06,
-      "loss": 0.2662,
+      "epoch": 3.835189389843947,
+      "grad_norm": 0.24770770967006683,
+      "learning_rate": 6.799661351715067e-06,
+      "loss": 0.417,
       "step": 106415
     },
     {
-      "epoch": 3.74,
-      "learning_rate": 7.840996967119377e-06,
-      "loss": 0.2551,
+      "epoch": 3.835369589505172,
+      "grad_norm": 0.27001357078552246,
+      "learning_rate": 6.797660911992351e-06,
+      "loss": 0.398,
       "step": 106420
     },
     {
-      "epoch": 3.74,
-      "learning_rate": 7.838925312219503e-06,
-      "loss": 0.2598,
+      "epoch": 3.8355497891663966,
+      "grad_norm": 0.2153816670179367,
+      "learning_rate": 6.795660720272978e-06,
+      "loss": 0.3642,
       "step": 106425
     },
     {
-      "epoch": 3.74,
-      "learning_rate": 7.836853880144874e-06,
-      "loss": 0.2275,
+      "epoch": 3.835729988827621,
+      "grad_norm": 0.2308131456375122,
+      "learning_rate": 6.7936607765841985e-06,
+      "loss": 0.4077,
       "step": 106430
     },
     {
-      "epoch": 3.74,
-      "learning_rate": 7.834782670922377e-06,
-      "loss": 0.2677,
+      "epoch": 3.8359101884888456,
+      "grad_norm": 0.19024352729320526,
+      "learning_rate": 6.791661080953274e-06,
+      "loss": 0.3708,
       "step": 106435
     },
     {
-      "epoch": 3.74,
-      "learning_rate": 7.832711684578925e-06,
-      "loss": 0.279,
+      "epoch": 3.8360903881500703,
+      "grad_norm": 0.15937311947345734,
+      "learning_rate": 6.78966163340744e-06,
+      "loss": 0.3621,
       "step": 106440
     },
     {
-      "epoch": 3.75,
-      "learning_rate": 7.830640921141388e-06,
-      "loss": 0.2623,
+      "epoch": 3.836270587811295,
+      "grad_norm": 0.29799744486808777,
+      "learning_rate": 6.78766243397394e-06,
+      "loss": 0.371,
       "step": 106445
     },
     {
-      "epoch": 3.75,
-      "learning_rate": 7.828570380636672e-06,
-      "loss": 0.2451,
+      "epoch": 3.8364507874725193,
+      "grad_norm": 0.20933188498020172,
+      "learning_rate": 6.785663482680016e-06,
+      "loss": 0.3618,
       "step": 106450
     },
     {
-      "epoch": 3.75,
-      "learning_rate": 7.826500063091646e-06,
-      "loss": 0.2776,
+      "epoch": 3.836630987133744,
+      "grad_norm": 0.25528839230537415,
+      "learning_rate": 6.7836647795528976e-06,
+      "loss": 0.3922,
       "step": 106455
     },
     {
-      "epoch": 3.75,
-      "learning_rate": 7.824429968533207e-06,
-      "loss": 0.2688,
+      "epoch": 3.836811186794969,
+      "grad_norm": 0.231916606426239,
+      "learning_rate": 6.781666324619815e-06,
+      "loss": 0.379,
       "step": 106460
     },
     {
-      "epoch": 3.75,
-      "learning_rate": 7.822360096988224e-06,
-      "loss": 0.2786,
+      "epoch": 3.8369913864561935,
+      "grad_norm": 0.21450571715831757,
+      "learning_rate": 6.779668117908008e-06,
+      "loss": 0.3953,
       "step": 106465
     },
     {
-      "epoch": 3.75,
-      "learning_rate": 7.820290448483574e-06,
-      "loss": 0.2598,
+      "epoch": 3.8371715861174183,
+      "grad_norm": 0.23598310351371765,
+      "learning_rate": 6.777670159444696e-06,
+      "loss": 0.367,
       "step": 106470
     },
     {
-      "epoch": 3.75,
-      "learning_rate": 7.818221023046127e-06,
-      "loss": 0.2687,
+      "epoch": 3.837351785778643,
+      "grad_norm": 0.18300804495811462,
+      "learning_rate": 6.775672449257098e-06,
+      "loss": 0.3378,
       "step": 106475
     },
     {
-      "epoch": 3.75,
-      "learning_rate": 7.816151820702761e-06,
-      "loss": 0.2457,
+      "epoch": 3.8375319854398673,
+      "grad_norm": 0.27081814408302307,
+      "learning_rate": 6.773674987372436e-06,
+      "loss": 0.389,
       "step": 106480
     },
     {
-      "epoch": 3.75,
-      "learning_rate": 7.81408284148034e-06,
-      "loss": 0.2553,
+      "epoch": 3.837712185101092,
+      "grad_norm": 0.22525940835475922,
+      "learning_rate": 6.771677773817917e-06,
+      "loss": 0.398,
       "step": 106485
     },
     {
-      "epoch": 3.75,
-      "learning_rate": 7.812014085405728e-06,
-      "loss": 0.2522,
+      "epoch": 3.8378923847623168,
+      "grad_norm": 0.23411720991134644,
+      "learning_rate": 6.769680808620774e-06,
+      "loss": 0.4028,
       "step": 106490
     },
     {
-      "epoch": 3.75,
-      "learning_rate": 7.809945552505779e-06,
-      "loss": 0.2643,
+      "epoch": 3.838072584423541,
+      "grad_norm": 0.2922147214412689,
+      "learning_rate": 6.7676840918081825e-06,
+      "loss": 0.4001,
       "step": 106495
     },
     {
-      "epoch": 3.75,
-      "learning_rate": 7.807877242807357e-06,
-      "loss": 0.2496,
+      "epoch": 3.838252784084766,
+      "grad_norm": 0.23292958736419678,
+      "learning_rate": 6.765687623407377e-06,
+      "loss": 0.3461,
       "step": 106500
     },
     {
-      "epoch": 3.75,
-      "eval_loss": 0.2516409158706665,
-      "eval_runtime": 10.5352,
-      "eval_samples_per_second": 9.492,
-      "eval_steps_per_second": 9.492,
+      "epoch": 3.838252784084766,
+      "eval_loss": 0.4283367097377777,
+      "eval_runtime": 3.5256,
+      "eval_samples_per_second": 28.364,
+      "eval_steps_per_second": 7.091,
       "step": 106500
     },
     {
-      "epoch": 3.75,
-      "learning_rate": 7.805809156337327e-06,
-      "loss": 0.2422,
+      "epoch": 3.8384329837459905,
+      "grad_norm": 0.2138739675283432,
+      "learning_rate": 6.763691403445543e-06,
+      "loss": 0.3657,
       "step": 106505
     },
     {
-      "epoch": 3.75,
-      "learning_rate": 7.803741293122533e-06,
-      "loss": 0.2345,
+      "epoch": 3.8386131834072152,
+      "grad_norm": 0.20448392629623413,
+      "learning_rate": 6.761695431949888e-06,
+      "loss": 0.3634,
       "step": 106510
     },
     {
-      "epoch": 3.75,
-      "learning_rate": 7.801673653189818e-06,
-      "loss": 0.2534,
+      "epoch": 3.83879338306844,
+      "grad_norm": 0.269741415977478,
+      "learning_rate": 6.759699708947598e-06,
+      "loss": 0.3845,
       "step": 106515
     },
     {
-      "epoch": 3.75,
-      "learning_rate": 7.799606236566045e-06,
-      "loss": 0.271,
+      "epoch": 3.8389735827296647,
+      "grad_norm": 0.20338128507137299,
+      "learning_rate": 6.757704234465869e-06,
+      "loss": 0.3784,
       "step": 106520
     },
     {
-      "epoch": 3.75,
-      "learning_rate": 7.797539043278048e-06,
-      "loss": 0.2369,
+      "epoch": 3.839153782390889,
+      "grad_norm": 0.2528837025165558,
+      "learning_rate": 6.75570900853188e-06,
+      "loss": 0.4003,
       "step": 106525
     },
     {
-      "epoch": 3.75,
-      "learning_rate": 7.795472073352664e-06,
-      "loss": 0.2472,
+      "epoch": 3.8393339820521137,
+      "grad_norm": 0.22932448983192444,
+      "learning_rate": 6.753714031172834e-06,
+      "loss": 0.3484,
       "step": 106530
     },
     {
-      "epoch": 3.75,
-      "learning_rate": 7.793405326816744e-06,
-      "loss": 0.2607,
+      "epoch": 3.8395141817133385,
+      "grad_norm": 0.2735101282596588,
+      "learning_rate": 6.751719302415898e-06,
+      "loss": 0.4048,
       "step": 106535
     },
     {
-      "epoch": 3.75,
-      "learning_rate": 7.791338803697115e-06,
-      "loss": 0.2395,
+      "epoch": 3.8396943813745628,
+      "grad_norm": 0.2353210300207138,
+      "learning_rate": 6.749724822288256e-06,
+      "loss": 0.3939,
       "step": 106540
     },
     {
-      "epoch": 3.75,
-      "learning_rate": 7.789272504020611e-06,
-      "loss": 0.2539,
+      "epoch": 3.8398745810357875,
+      "grad_norm": 0.2255113422870636,
+      "learning_rate": 6.747730590817078e-06,
+      "loss": 0.3748,
       "step": 106545
     },
     {
-      "epoch": 3.75,
-      "learning_rate": 7.787206427814053e-06,
-      "loss": 0.2468,
+      "epoch": 3.8400547806970122,
+      "grad_norm": 0.2259177565574646,
+      "learning_rate": 6.7457366080295296e-06,
+      "loss": 0.3593,
       "step": 106550
     },
     {
-      "epoch": 3.75,
-      "learning_rate": 7.785140575104277e-06,
-      "loss": 0.2856,
+      "epoch": 3.840234980358237,
+      "grad_norm": 0.27952998876571655,
+      "learning_rate": 6.743742873952794e-06,
+      "loss": 0.3721,
       "step": 106555
     },
     {
-      "epoch": 3.75,
-      "learning_rate": 7.783074945918115e-06,
-      "loss": 0.2525,
+      "epoch": 3.8404151800194617,
+      "grad_norm": 0.23785199224948883,
+      "learning_rate": 6.741749388614027e-06,
+      "loss": 0.3657,
       "step": 106560
     },
     {
-      "epoch": 3.75,
-      "learning_rate": 7.781009540282378e-06,
-      "loss": 0.2338,
+      "epoch": 3.8405953796806864,
+      "grad_norm": 0.19771380722522736,
+      "learning_rate": 6.739756152040391e-06,
+      "loss": 0.3827,
       "step": 106565
     },
     {
-      "epoch": 3.75,
-      "learning_rate": 7.778944358223875e-06,
-      "loss": 0.2603,
+      "epoch": 3.8407755793419107,
+      "grad_norm": 0.23672428727149963,
+      "learning_rate": 6.7377631642590395e-06,
+      "loss": 0.4001,
       "step": 106570
     },
     {
-      "epoch": 3.75,
-      "learning_rate": 7.776879399769438e-06,
-      "loss": 0.2385,
+      "epoch": 3.8409557790031355,
+      "grad_norm": 0.28637582063674927,
+      "learning_rate": 6.7357704252971245e-06,
+      "loss": 0.381,
       "step": 106575
     },
     {
-      "epoch": 3.75,
-      "learning_rate": 7.77481466494587e-06,
-      "loss": 0.2633,
+      "epoch": 3.84113597866436,
+      "grad_norm": 0.23448318243026733,
+      "learning_rate": 6.733777935181806e-06,
+      "loss": 0.4347,
       "step": 106580
     },
     {
-      "epoch": 3.75,
-      "learning_rate": 7.772750153779985e-06,
-      "loss": 0.2625,
+      "epoch": 3.8413161783255845,
+      "grad_norm": 0.24527902901172638,
+      "learning_rate": 6.731785693940237e-06,
+      "loss": 0.405,
       "step": 106585
     },
     {
-      "epoch": 3.75,
-      "learning_rate": 7.770685866298577e-06,
-      "loss": 0.2417,
+      "epoch": 3.841496377986809,
+      "grad_norm": 0.2648589015007019,
+      "learning_rate": 6.729793701599535e-06,
+      "loss": 0.3911,
       "step": 106590
     },
     {
-      "epoch": 3.75,
-      "learning_rate": 7.768621802528467e-06,
-      "loss": 0.2505,
+      "epoch": 3.841676577648034,
+      "grad_norm": 0.28334563970565796,
+      "learning_rate": 6.727801958186864e-06,
+      "loss": 0.4194,
       "step": 106595
     },
     {
-      "epoch": 3.75,
-      "learning_rate": 7.766557962496445e-06,
-      "loss": 0.2446,
+      "epoch": 3.8418567773092587,
+      "grad_norm": 0.23570017516613007,
+      "learning_rate": 6.725810463729354e-06,
+      "loss": 0.3873,
       "step": 106600
     },
     {
-      "epoch": 3.75,
-      "learning_rate": 7.764494346229307e-06,
-      "loss": 0.2627,
+      "epoch": 3.8420369769704834,
+      "grad_norm": 0.254119873046875,
+      "learning_rate": 6.723819218254138e-06,
+      "loss": 0.3425,
       "step": 106605
     },
     {
-      "epoch": 3.75,
-      "learning_rate": 7.762430953753855e-06,
-      "loss": 0.2617,
+      "epoch": 3.842217176631708,
+      "grad_norm": 0.23003673553466797,
+      "learning_rate": 6.721828221788346e-06,
+      "loss": 0.3899,
       "step": 106610
     },
     {
-      "epoch": 3.75,
-      "learning_rate": 7.760367785096869e-06,
-      "loss": 0.2554,
+      "epoch": 3.8423973762929324,
+      "grad_norm": 0.21980217099189758,
+      "learning_rate": 6.719837474359098e-06,
+      "loss": 0.372,
       "step": 106615
     },
     {
-      "epoch": 3.75,
-      "learning_rate": 7.758304840285158e-06,
-      "loss": 0.253,
+      "epoch": 3.842577575954157,
+      "grad_norm": 0.25185590982437134,
+      "learning_rate": 6.717846975993536e-06,
+      "loss": 0.4302,
       "step": 106620
     },
     {
-      "epoch": 3.75,
-      "learning_rate": 7.75624211934549e-06,
-      "loss": 0.2655,
+      "epoch": 3.842757775615382,
+      "grad_norm": 0.23943038284778595,
+      "learning_rate": 6.715856726718767e-06,
+      "loss": 0.4065,
       "step": 106625
     },
     {
-      "epoch": 3.75,
-      "learning_rate": 7.75417962230465e-06,
-      "loss": 0.2666,
+      "epoch": 3.8429379752766066,
+      "grad_norm": 0.2441278100013733,
+      "learning_rate": 6.71386672656191e-06,
+      "loss": 0.3898,
       "step": 106630
     },
     {
-      "epoch": 3.75,
-      "learning_rate": 7.75211734918943e-06,
-      "loss": 0.2492,
+      "epoch": 3.843118174937831,
+      "grad_norm": 0.2758260667324066,
+      "learning_rate": 6.711876975550077e-06,
+      "loss": 0.3623,
       "step": 106635
     },
     {
-      "epoch": 3.75,
-      "learning_rate": 7.750055300026599e-06,
-      "loss": 0.2651,
+      "epoch": 3.8432983745990557,
+      "grad_norm": 0.20836462080478668,
+      "learning_rate": 6.7098874737103825e-06,
+      "loss": 0.372,
       "step": 106640
     },
     {
-      "epoch": 3.75,
-      "learning_rate": 7.747993474842923e-06,
-      "loss": 0.286,
+      "epoch": 3.8434785742602804,
+      "grad_norm": 0.25424110889434814,
+      "learning_rate": 6.707898221069922e-06,
+      "loss": 0.3941,
       "step": 106645
     },
     {
-      "epoch": 3.75,
-      "learning_rate": 7.745931873665194e-06,
-      "loss": 0.2526,
+      "epoch": 3.843658773921505,
+      "grad_norm": 0.25770169496536255,
+      "learning_rate": 6.705909217655815e-06,
+      "loss": 0.3841,
       "step": 106650
     },
     {
-      "epoch": 3.75,
-      "learning_rate": 7.743870496520165e-06,
-      "loss": 0.246,
+      "epoch": 3.84383897358273,
+      "grad_norm": 0.296792596578598,
+      "learning_rate": 6.703920463495151e-06,
+      "loss": 0.3746,
       "step": 106655
     },
     {
-      "epoch": 3.75,
-      "learning_rate": 7.741809343434598e-06,
-      "loss": 0.254,
+      "epoch": 3.844019173243954,
+      "grad_norm": 0.2855359613895416,
+      "learning_rate": 6.701931958615029e-06,
+      "loss": 0.4001,
       "step": 106660
     },
     {
-      "epoch": 3.75,
-      "learning_rate": 7.739748414435272e-06,
-      "loss": 0.2589,
+      "epoch": 3.844199372905179,
+      "grad_norm": 0.2668587267398834,
+      "learning_rate": 6.699943703042541e-06,
+      "loss": 0.3571,
       "step": 106665
     },
     {
-      "epoch": 3.75,
-      "learning_rate": 7.737687709548927e-06,
-      "loss": 0.2474,
+      "epoch": 3.8443795725664036,
+      "grad_norm": 0.2563954293727875,
+      "learning_rate": 6.69795569680477e-06,
+      "loss": 0.3489,
       "step": 106670
     },
     {
-      "epoch": 3.75,
-      "learning_rate": 7.735627228802342e-06,
-      "loss": 0.2452,
+      "epoch": 3.8445597722276283,
+      "grad_norm": 0.3133787214756012,
+      "learning_rate": 6.6959679399288265e-06,
+      "loss": 0.3753,
       "step": 106675
     },
     {
-      "epoch": 3.75,
-      "learning_rate": 7.733566972222259e-06,
-      "loss": 0.2682,
+      "epoch": 3.8447399718888526,
+      "grad_norm": 0.22209054231643677,
+      "learning_rate": 6.693980432441757e-06,
+      "loss": 0.3987,
       "step": 106680
     },
     {
-      "epoch": 3.75,
-      "learning_rate": 7.73150693983542e-06,
-      "loss": 0.2653,
+      "epoch": 3.8449201715500774,
+      "grad_norm": 0.22669175267219543,
+      "learning_rate": 6.69199317437067e-06,
+      "loss": 0.3612,
       "step": 106685
     },
     {
-      "epoch": 3.75,
-      "learning_rate": 7.729447131668593e-06,
-      "loss": 0.2691,
+      "epoch": 3.845100371211302,
+      "grad_norm": 0.24316006898880005,
+      "learning_rate": 6.6900061657426325e-06,
+      "loss": 0.3768,
       "step": 106690
     },
     {
-      "epoch": 3.75,
-      "learning_rate": 7.727387547748513e-06,
-      "loss": 0.2659,
+      "epoch": 3.845280570872527,
+      "grad_norm": 0.26429635286331177,
+      "learning_rate": 6.688019406584706e-06,
+      "loss": 0.3953,
       "step": 106695
     },
     {
-      "epoch": 3.75,
-      "learning_rate": 7.725328188101925e-06,
-      "loss": 0.2828,
+      "epoch": 3.8454607705337516,
+      "grad_norm": 0.24208183586597443,
+      "learning_rate": 6.686032896923986e-06,
+      "loss": 0.3652,
       "step": 106700
     },
     {
-      "epoch": 3.75,
-      "learning_rate": 7.723269052755553e-06,
-      "loss": 0.2698,
+      "epoch": 3.8456409701949763,
+      "grad_norm": 0.21602210402488708,
+      "learning_rate": 6.684046636787514e-06,
+      "loss": 0.3948,
       "step": 106705
     },
     {
-      "epoch": 3.75,
-      "learning_rate": 7.721210141736156e-06,
-      "loss": 0.2539,
+      "epoch": 3.8458211698562006,
+      "grad_norm": 0.21090590953826904,
+      "learning_rate": 6.682060626202355e-06,
+      "loss": 0.3688,
       "step": 106710
     },
     {
-      "epoch": 3.75,
-      "learning_rate": 7.719151455070462e-06,
-      "loss": 0.2575,
+      "epoch": 3.8460013695174253,
+      "grad_norm": 0.23994767665863037,
+      "learning_rate": 6.6800748651955795e-06,
+      "loss": 0.3735,
       "step": 106715
     },
     {
-      "epoch": 3.75,
-      "learning_rate": 7.717092992785185e-06,
-      "loss": 0.2708,
+      "epoch": 3.84618156917865,
+      "grad_norm": 0.24536341428756714,
+      "learning_rate": 6.678089353794237e-06,
+      "loss": 0.3965,
       "step": 106720
     },
     {
-      "epoch": 3.75,
-      "learning_rate": 7.715034754907068e-06,
-      "loss": 0.2411,
+      "epoch": 3.8463617688398744,
+      "grad_norm": 0.24873821437358856,
+      "learning_rate": 6.676104092025378e-06,
+      "loss": 0.3989,
       "step": 106725
     },
     {
-      "epoch": 3.76,
-      "learning_rate": 7.71297674146284e-06,
-      "loss": 0.2614,
+      "epoch": 3.846541968501099,
+      "grad_norm": 0.2966354191303253,
+      "learning_rate": 6.674119079916056e-06,
+      "loss": 0.404,
       "step": 106730
     },
     {
-      "epoch": 3.76,
-      "learning_rate": 7.71091895247922e-06,
-      "loss": 0.2691,
+      "epoch": 3.846722168162324,
+      "grad_norm": 0.24333995580673218,
+      "learning_rate": 6.672134317493303e-06,
+      "loss": 0.3602,
       "step": 106735
     },
     {
-      "epoch": 3.76,
-      "learning_rate": 7.708861387982921e-06,
-      "loss": 0.2591,
+      "epoch": 3.8469023678235486,
+      "grad_norm": 0.2882740795612335,
+      "learning_rate": 6.670149804784181e-06,
+      "loss": 0.3889,
       "step": 106740
     },
     {
-      "epoch": 3.76,
-      "learning_rate": 7.706804048000654e-06,
-      "loss": 0.2456,
+      "epoch": 3.8470825674847733,
+      "grad_norm": 0.2220769077539444,
+      "learning_rate": 6.668165541815721e-06,
+      "loss": 0.3715,
       "step": 106745
     },
     {
-      "epoch": 3.76,
-      "learning_rate": 7.704746932559148e-06,
-      "loss": 0.2298,
+      "epoch": 3.847262767145998,
+      "grad_norm": 0.26409411430358887,
+      "learning_rate": 6.666181528614954e-06,
+      "loss": 0.356,
       "step": 106750
     },
     {
-      "epoch": 3.76,
-      "learning_rate": 7.702690041685105e-06,
-      "loss": 0.2538,
+      "epoch": 3.8474429668072223,
+      "grad_norm": 0.24379925429821014,
+      "learning_rate": 6.6641977652089155e-06,
+      "loss": 0.3866,
       "step": 106755
     },
     {
-      "epoch": 3.76,
-      "learning_rate": 7.70063337540523e-06,
-      "loss": 0.2592,
+      "epoch": 3.847623166468447,
+      "grad_norm": 0.2612653076648712,
+      "learning_rate": 6.662214251624624e-06,
+      "loss": 0.3836,
       "step": 106760
     },
     {
-      "epoch": 3.76,
-      "learning_rate": 7.698576933746225e-06,
-      "loss": 0.2535,
+      "epoch": 3.8478033661296718,
+      "grad_norm": 0.20070244371891022,
+      "learning_rate": 6.660230987889121e-06,
+      "loss": 0.3629,
       "step": 106765
     },
     {
-      "epoch": 3.76,
-      "learning_rate": 7.696520716734804e-06,
-      "loss": 0.2577,
+      "epoch": 3.847983565790896,
+      "grad_norm": 0.1985878348350525,
+      "learning_rate": 6.658247974029427e-06,
+      "loss": 0.3356,
       "step": 106770
     },
     {
-      "epoch": 3.76,
-      "learning_rate": 7.694464724397649e-06,
-      "loss": 0.256,
+      "epoch": 3.848163765452121,
+      "grad_norm": 0.24218730628490448,
+      "learning_rate": 6.656265210072537e-06,
+      "loss": 0.396,
       "step": 106775
     },
     {
-      "epoch": 3.76,
-      "learning_rate": 7.692408956761466e-06,
-      "loss": 0.2563,
+      "epoch": 3.8483439651133455,
+      "grad_norm": 0.29759272933006287,
+      "learning_rate": 6.6542826960454915e-06,
+      "loss": 0.3804,
       "step": 106780
     },
     {
-      "epoch": 3.76,
-      "learning_rate": 7.690353413852952e-06,
-      "loss": 0.2454,
+      "epoch": 3.8485241647745703,
+      "grad_norm": 0.340660035610199,
+      "learning_rate": 6.652300431975292e-06,
+      "loss": 0.4202,
       "step": 106785
     },
     {
-      "epoch": 3.76,
-      "learning_rate": 7.688298095698792e-06,
-      "loss": 0.2379,
+      "epoch": 3.848704364435795,
+      "grad_norm": 0.24566273391246796,
+      "learning_rate": 6.650318417888948e-06,
+      "loss": 0.396,
       "step": 106790
     },
     {
-      "epoch": 3.76,
-      "learning_rate": 7.686243002325675e-06,
-      "loss": 0.2398,
+      "epoch": 3.8488845640970197,
+      "grad_norm": 0.2540030777454376,
+      "learning_rate": 6.648336653813461e-06,
+      "loss": 0.3947,
       "step": 106795
     },
     {
-      "epoch": 3.76,
-      "learning_rate": 7.684188133760272e-06,
-      "loss": 0.2724,
+      "epoch": 3.849064763758244,
+      "grad_norm": 0.27518022060394287,
+      "learning_rate": 6.646355139775828e-06,
+      "loss": 0.3916,
       "step": 106800
     },
     {
-      "epoch": 3.76,
-      "learning_rate": 7.682133490029286e-06,
-      "loss": 0.246,
+      "epoch": 3.8492449634194688,
+      "grad_norm": 0.19195982813835144,
+      "learning_rate": 6.644373875803059e-06,
+      "loss": 0.3582,
       "step": 106805
     },
     {
-      "epoch": 3.76,
-      "learning_rate": 7.680079071159379e-06,
-      "loss": 0.2528,
+      "epoch": 3.8494251630806935,
+      "grad_norm": 0.17849093675613403,
+      "learning_rate": 6.642392861922145e-06,
+      "loss": 0.4018,
       "step": 106810
     },
     {
-      "epoch": 3.76,
-      "learning_rate": 7.678024877177234e-06,
-      "loss": 0.26,
+      "epoch": 3.849605362741918,
+      "grad_norm": 0.20975105464458466,
+      "learning_rate": 6.64041209816007e-06,
+      "loss": 0.3944,
       "step": 106815
     },
     {
-      "epoch": 3.76,
-      "learning_rate": 7.675970908109512e-06,
-      "loss": 0.2595,
+      "epoch": 3.8497855624031425,
+      "grad_norm": 0.20660048723220825,
+      "learning_rate": 6.638431584543827e-06,
+      "loss": 0.3759,
       "step": 106820
     },
     {
-      "epoch": 3.76,
-      "learning_rate": 7.6739171639829e-06,
-      "loss": 0.2548,
+      "epoch": 3.8499657620643672,
+      "grad_norm": 0.20970280468463898,
+      "learning_rate": 6.636451321100401e-06,
+      "loss": 0.3418,
       "step": 106825
     },
     {
-      "epoch": 3.76,
-      "learning_rate": 7.671863644824045e-06,
-      "loss": 0.2683,
+      "epoch": 3.850145961725592,
+      "grad_norm": 0.20445165038108826,
+      "learning_rate": 6.634471307856763e-06,
+      "loss": 0.3869,
       "step": 106830
     },
     {
-      "epoch": 3.76,
-      "learning_rate": 7.669810350659629e-06,
-      "loss": 0.2656,
+      "epoch": 3.8503261613868167,
+      "grad_norm": 0.2625582814216614,
+      "learning_rate": 6.632491544839903e-06,
+      "loss": 0.4001,
       "step": 106835
     },
     {
-      "epoch": 3.76,
-      "learning_rate": 7.667757281516296e-06,
-      "loss": 0.2655,
+      "epoch": 3.8505063610480414,
+      "grad_norm": 0.2839062213897705,
+      "learning_rate": 6.6305120320767885e-06,
+      "loss": 0.3926,
       "step": 106840
     },
     {
-      "epoch": 3.76,
-      "learning_rate": 7.665704437420724e-06,
-      "loss": 0.2698,
+      "epoch": 3.8506865607092657,
+      "grad_norm": 0.22392812371253967,
+      "learning_rate": 6.628532769594395e-06,
+      "loss": 0.4024,
       "step": 106845
     },
     {
-      "epoch": 3.76,
-      "learning_rate": 7.663651818399553e-06,
-      "loss": 0.2581,
+      "epoch": 3.8508667603704905,
+      "grad_norm": 0.2890661060810089,
+      "learning_rate": 6.626553757419682e-06,
+      "loss": 0.3862,
       "step": 106850
     },
     {
-      "epoch": 3.76,
-      "learning_rate": 7.661599424479437e-06,
-      "loss": 0.2746,
+      "epoch": 3.851046960031715,
+      "grad_norm": 0.29909321665763855,
+      "learning_rate": 6.62457499557961e-06,
+      "loss": 0.3887,
       "step": 106855
     },
     {
-      "epoch": 3.76,
-      "learning_rate": 7.659547255687023e-06,
-      "loss": 0.2813,
+      "epoch": 3.85122715969294,
+      "grad_norm": 0.22355806827545166,
+      "learning_rate": 6.622596484101156e-06,
+      "loss": 0.388,
       "step": 106860
     },
     {
-      "epoch": 3.76,
-      "learning_rate": 7.657495312048963e-06,
-      "loss": 0.2596,
+      "epoch": 3.8514073593541642,
+      "grad_norm": 0.2812059223651886,
+      "learning_rate": 6.620618223011274e-06,
+      "loss": 0.3845,
       "step": 106865
     },
     {
-      "epoch": 3.76,
-      "learning_rate": 7.6554435935919e-06,
-      "loss": 0.2277,
+      "epoch": 3.851587559015389,
+      "grad_norm": 0.28787899017333984,
+      "learning_rate": 6.618640212336893e-06,
+      "loss": 0.3785,
       "step": 106870
     },
     {
-      "epoch": 3.76,
-      "learning_rate": 7.65339210034247e-06,
-      "loss": 0.2635,
+      "epoch": 3.8517677586766137,
+      "grad_norm": 0.21656519174575806,
+      "learning_rate": 6.616662452104991e-06,
+      "loss": 0.393,
       "step": 106875
     },
     {
-      "epoch": 3.76,
-      "learning_rate": 7.651340832327305e-06,
-      "loss": 0.2561,
+      "epoch": 3.8519479583378384,
+      "grad_norm": 0.2218136340379715,
+      "learning_rate": 6.6146849423424946e-06,
+      "loss": 0.3646,
       "step": 106880
     },
     {
-      "epoch": 3.76,
-      "learning_rate": 7.649289789573044e-06,
-      "loss": 0.2629,
+      "epoch": 3.852128157999063,
+      "grad_norm": 0.2481302171945572,
+      "learning_rate": 6.612707683076369e-06,
+      "loss": 0.3755,
       "step": 106885
     },
     {
-      "epoch": 3.76,
-      "learning_rate": 7.64723897210633e-06,
-      "loss": 0.2495,
+      "epoch": 3.8523083576602875,
+      "grad_norm": 0.23201486468315125,
+      "learning_rate": 6.610730674333537e-06,
+      "loss": 0.3652,
       "step": 106890
     },
     {
-      "epoch": 3.76,
-      "learning_rate": 7.645188379953774e-06,
-      "loss": 0.2622,
+      "epoch": 3.852488557321512,
+      "grad_norm": 0.2411048710346222,
+      "learning_rate": 6.6087539161409305e-06,
+      "loss": 0.3625,
       "step": 106895
     },
     {
-      "epoch": 3.76,
-      "learning_rate": 7.643138013142018e-06,
-      "loss": 0.2551,
+      "epoch": 3.852668756982737,
+      "grad_norm": 0.24224650859832764,
+      "learning_rate": 6.6067774085254995e-06,
+      "loss": 0.354,
       "step": 106900
     },
     {
-      "epoch": 3.76,
-      "learning_rate": 7.641087871697677e-06,
-      "loss": 0.2712,
+      "epoch": 3.8528489566439617,
+      "grad_norm": 0.25247785449028015,
+      "learning_rate": 6.6048011515141646e-06,
+      "loss": 0.3811,
       "step": 106905
     },
     {
-      "epoch": 3.76,
-      "learning_rate": 7.63903795564737e-06,
-      "loss": 0.2408,
+      "epoch": 3.853029156305186,
+      "grad_norm": 0.2523113191127777,
+      "learning_rate": 6.602825145133854e-06,
+      "loss": 0.4067,
       "step": 106910
     },
     {
-      "epoch": 3.76,
-      "learning_rate": 7.636988265017702e-06,
-      "loss": 0.2498,
+      "epoch": 3.8532093559664107,
+      "grad_norm": 0.21480904519557953,
+      "learning_rate": 6.600849389411487e-06,
+      "loss": 0.388,
       "step": 106915
     },
     {
-      "epoch": 3.76,
-      "learning_rate": 7.634938799835312e-06,
-      "loss": 0.2667,
+      "epoch": 3.8533895556276354,
+      "grad_norm": 0.2217121422290802,
+      "learning_rate": 6.59887388437398e-06,
+      "loss": 0.3805,
       "step": 106920
     },
     {
-      "epoch": 3.76,
-      "learning_rate": 7.632889560126794e-06,
-      "loss": 0.2521,
+      "epoch": 3.85356975528886,
+      "grad_norm": 0.2531042695045471,
+      "learning_rate": 6.5968986300482595e-06,
+      "loss": 0.3892,
       "step": 106925
     },
     {
-      "epoch": 3.76,
-      "learning_rate": 7.630840545918761e-06,
-      "loss": 0.2465,
+      "epoch": 3.853749954950085,
+      "grad_norm": 0.27398139238357544,
+      "learning_rate": 6.594923626461233e-06,
+      "loss": 0.3802,
       "step": 106930
     },
     {
-      "epoch": 3.76,
-      "learning_rate": 7.628791757237811e-06,
-      "loss": 0.2646,
+      "epoch": 3.853930154611309,
+      "grad_norm": 0.24913030862808228,
+      "learning_rate": 6.592948873639807e-06,
+      "loss": 0.3898,
       "step": 106935
     },
     {
-      "epoch": 3.76,
-      "learning_rate": 7.626743194110547e-06,
-      "loss": 0.241,
+      "epoch": 3.854110354272534,
+      "grad_norm": 0.23287741839885712,
+      "learning_rate": 6.590974371610889e-06,
+      "loss": 0.3705,
       "step": 106940
     },
     {
-      "epoch": 3.76,
-      "learning_rate": 7.624694856563585e-06,
-      "loss": 0.2578,
+      "epoch": 3.8542905539337586,
+      "grad_norm": 0.23462989926338196,
+      "learning_rate": 6.589000120401375e-06,
+      "loss": 0.3583,
       "step": 106945
     },
     {
-      "epoch": 3.76,
-      "learning_rate": 7.6226467446235075e-06,
-      "loss": 0.2587,
+      "epoch": 3.8544707535949834,
+      "grad_norm": 0.2538296580314636,
+      "learning_rate": 6.587026120038178e-06,
+      "loss": 0.393,
       "step": 106950
     },
     {
-      "epoch": 3.76,
-      "learning_rate": 7.620598858316905e-06,
-      "loss": 0.263,
+      "epoch": 3.8546509532562077,
+      "grad_norm": 0.23619996011257172,
+      "learning_rate": 6.585052370548192e-06,
+      "loss": 0.3613,
       "step": 106955
     },
     {
-      "epoch": 3.76,
-      "learning_rate": 7.618551197670376e-06,
-      "loss": 0.2499,
+      "epoch": 3.8548311529174324,
+      "grad_norm": 0.20444704592227936,
+      "learning_rate": 6.583078871958287e-06,
+      "loss": 0.3983,
       "step": 106960
     },
     {
-      "epoch": 3.76,
-      "learning_rate": 7.616503762710509e-06,
-      "loss": 0.2554,
+      "epoch": 3.855011352578657,
+      "grad_norm": 0.2089342325925827,
+      "learning_rate": 6.581105624295372e-06,
+      "loss": 0.3489,
       "step": 106965
     },
     {
-      "epoch": 3.76,
-      "learning_rate": 7.6144565534638835e-06,
-      "loss": 0.2798,
+      "epoch": 3.855191552239882,
+      "grad_norm": 0.2452775537967682,
+      "learning_rate": 6.579132627586329e-06,
+      "loss": 0.3685,
       "step": 106970
     },
     {
-      "epoch": 3.76,
-      "learning_rate": 7.612409569957074e-06,
-      "loss": 0.2447,
+      "epoch": 3.8553717519011066,
+      "grad_norm": 0.23251672089099884,
+      "learning_rate": 6.577159881858031e-06,
+      "loss": 0.3592,
       "step": 106975
     },
     {
-      "epoch": 3.76,
-      "learning_rate": 7.610362812216676e-06,
-      "loss": 0.2339,
+      "epoch": 3.8555519515623313,
+      "grad_norm": 0.21264806389808655,
+      "learning_rate": 6.575187387137377e-06,
+      "loss": 0.3709,
       "step": 106980
     },
     {
-      "epoch": 3.76,
-      "learning_rate": 7.608316280269253e-06,
-      "loss": 0.236,
+      "epoch": 3.8557321512235556,
+      "grad_norm": 0.20880858600139618,
+      "learning_rate": 6.5732151434512115e-06,
+      "loss": 0.36,
       "step": 106985
     },
     {
-      "epoch": 3.76,
-      "learning_rate": 7.6062699741413775e-06,
-      "loss": 0.2403,
+      "epoch": 3.8559123508847803,
+      "grad_norm": 0.2800189256668091,
+      "learning_rate": 6.571243150826431e-06,
+      "loss": 0.3584,
       "step": 106990
     },
     {
-      "epoch": 3.76,
-      "learning_rate": 7.604223893859622e-06,
-      "loss": 0.2719,
+      "epoch": 3.856092550546005,
+      "grad_norm": 0.27210569381713867,
+      "learning_rate": 6.569271409289895e-06,
+      "loss": 0.3647,
       "step": 106995
     },
     {
-      "epoch": 3.76,
-      "learning_rate": 7.602178039450563e-06,
-      "loss": 0.2621,
+      "epoch": 3.8562727502072294,
+      "grad_norm": 0.22882099449634552,
+      "learning_rate": 6.567299918868467e-06,
+      "loss": 0.3507,
       "step": 107000
     },
     {
-      "epoch": 3.76,
-      "eval_loss": 0.251350462436676,
-      "eval_runtime": 10.5372,
-      "eval_samples_per_second": 9.49,
-      "eval_steps_per_second": 9.49,
+      "epoch": 3.8562727502072294,
+      "eval_loss": 0.42817628383636475,
+      "eval_runtime": 3.5212,
+      "eval_samples_per_second": 28.399,
+      "eval_steps_per_second": 7.1,
       "step": 107000
     },
     {
-      "epoch": 3.76,
-      "learning_rate": 7.600132410940758e-06,
-      "loss": 0.271,
+      "epoch": 3.856452949868454,
+      "grad_norm": 0.22581203281879425,
+      "learning_rate": 6.565328679589008e-06,
+      "loss": 0.3751,
       "step": 107005
     },
     {
-      "epoch": 3.76,
-      "learning_rate": 7.598087008356758e-06,
-      "loss": 0.2491,
+      "epoch": 3.856633149529679,
+      "grad_norm": 0.23361347615718842,
+      "learning_rate": 6.563357691478378e-06,
+      "loss": 0.3699,
       "step": 107010
     },
     {
-      "epoch": 3.77,
-      "learning_rate": 7.596041831725137e-06,
-      "loss": 0.2713,
+      "epoch": 3.8568133491909036,
+      "grad_norm": 0.2552787959575653,
+      "learning_rate": 6.561386954563423e-06,
+      "loss": 0.368,
       "step": 107015
     },
     {
-      "epoch": 3.77,
-      "learning_rate": 7.593996881072443e-06,
-      "loss": 0.2486,
+      "epoch": 3.8569935488521283,
+      "grad_norm": 0.19680064916610718,
+      "learning_rate": 6.559416468871008e-06,
+      "loss": 0.4097,
       "step": 107020
     },
     {
-      "epoch": 3.77,
-      "learning_rate": 7.591952156425231e-06,
-      "loss": 0.2624,
+      "epoch": 3.857173748513353,
+      "grad_norm": 0.300270676612854,
+      "learning_rate": 6.557446234427972e-06,
+      "loss": 0.3709,
       "step": 107025
     },
     {
-      "epoch": 3.77,
-      "learning_rate": 7.589907657810039e-06,
-      "loss": 0.2571,
+      "epoch": 3.8573539481745773,
+      "grad_norm": 0.2358800619840622,
+      "learning_rate": 6.555476251261161e-06,
+      "loss": 0.3658,
       "step": 107030
     },
     {
-      "epoch": 3.77,
-      "learning_rate": 7.587863385253427e-06,
-      "loss": 0.2679,
+      "epoch": 3.857534147835802,
+      "grad_norm": 0.2335239201784134,
+      "learning_rate": 6.553506519397417e-06,
+      "loss": 0.3828,
       "step": 107035
     },
     {
-      "epoch": 3.77,
-      "learning_rate": 7.585819338781938e-06,
-      "loss": 0.2466,
+      "epoch": 3.857714347497027,
+      "grad_norm": 0.3097918629646301,
+      "learning_rate": 6.551537038863567e-06,
+      "loss": 0.3893,
       "step": 107040
     },
     {
-      "epoch": 3.77,
-      "learning_rate": 7.583775518422098e-06,
-      "loss": 0.2415,
+      "epoch": 3.857894547158251,
+      "grad_norm": 0.22679081559181213,
+      "learning_rate": 6.549567809686458e-06,
+      "loss": 0.3852,
       "step": 107045
     },
     {
-      "epoch": 3.77,
-      "learning_rate": 7.581731924200464e-06,
-      "loss": 0.2564,
+      "epoch": 3.858074746819476,
+      "grad_norm": 0.23170988261699677,
+      "learning_rate": 6.547598831892926e-06,
+      "loss": 0.3755,
       "step": 107050
     },
     {
-      "epoch": 3.77,
-      "learning_rate": 7.579688556143555e-06,
-      "loss": 0.2617,
+      "epoch": 3.8582549464807006,
+      "grad_norm": 0.24494971334934235,
+      "learning_rate": 6.545630105509771e-06,
+      "loss": 0.3614,
       "step": 107055
     },
     {
-      "epoch": 3.77,
-      "learning_rate": 7.577645414277917e-06,
-      "loss": 0.2695,
+      "epoch": 3.8584351461419253,
+      "grad_norm": 0.22469878196716309,
+      "learning_rate": 6.543661630563841e-06,
+      "loss": 0.37,
       "step": 107060
     },
     {
-      "epoch": 3.77,
-      "learning_rate": 7.575602498630072e-06,
-      "loss": 0.2421,
+      "epoch": 3.85861534580315,
+      "grad_norm": 0.27316057682037354,
+      "learning_rate": 6.541693407081939e-06,
+      "loss": 0.3548,
       "step": 107065
     },
     {
-      "epoch": 3.77,
-      "learning_rate": 7.573559809226536e-06,
-      "loss": 0.2467,
+      "epoch": 3.8587955454643748,
+      "grad_norm": 0.2721122205257416,
+      "learning_rate": 6.53972543509091e-06,
+      "loss": 0.3677,
       "step": 107070
     },
     {
-      "epoch": 3.77,
-      "learning_rate": 7.571517346093851e-06,
-      "loss": 0.2511,
+      "epoch": 3.858975745125599,
+      "grad_norm": 0.2867799699306488,
+      "learning_rate": 6.537757714617537e-06,
+      "loss": 0.367,
       "step": 107075
     },
     {
-      "epoch": 3.77,
-      "learning_rate": 7.569475109258525e-06,
-      "loss": 0.2721,
+      "epoch": 3.8591559447868238,
+      "grad_norm": 0.2439620941877365,
+      "learning_rate": 6.535790245688633e-06,
+      "loss": 0.3661,
       "step": 107080
     },
     {
-      "epoch": 3.77,
-      "learning_rate": 7.567433098747079e-06,
-      "loss": 0.2477,
+      "epoch": 3.8593361444480485,
+      "grad_norm": 0.21324770152568817,
+      "learning_rate": 6.533823028331021e-06,
+      "loss": 0.3907,
       "step": 107085
     },
     {
-      "epoch": 3.77,
-      "learning_rate": 7.565391314586018e-06,
-      "loss": 0.2662,
+      "epoch": 3.859516344109273,
+      "grad_norm": 0.24692684412002563,
+      "learning_rate": 6.5318560625714924e-06,
+      "loss": 0.3767,
       "step": 107090
     },
     {
-      "epoch": 3.77,
-      "learning_rate": 7.563349756801869e-06,
-      "loss": 0.2606,
+      "epoch": 3.8596965437704975,
+      "grad_norm": 0.2432447373867035,
+      "learning_rate": 6.529889348436852e-06,
+      "loss": 0.3643,
       "step": 107095
     },
     {
-      "epoch": 3.77,
-      "learning_rate": 7.561308425421124e-06,
-      "loss": 0.2439,
+      "epoch": 3.8598767434317223,
+      "grad_norm": 0.2154054492712021,
+      "learning_rate": 6.5279228859538935e-06,
+      "loss": 0.3863,
       "step": 107100
     },
     {
-      "epoch": 3.77,
-      "learning_rate": 7.559267320470304e-06,
-      "loss": 0.2411,
+      "epoch": 3.860056943092947,
+      "grad_norm": 0.23707520961761475,
+      "learning_rate": 6.5259566751494e-06,
+      "loss": 0.3823,
       "step": 107105
     },
     {
-      "epoch": 3.77,
-      "learning_rate": 7.557226441975898e-06,
-      "loss": 0.2375,
+      "epoch": 3.8602371427541717,
+      "grad_norm": 0.23373231291770935,
+      "learning_rate": 6.523990716050179e-06,
+      "loss": 0.393,
       "step": 107110
     },
     {
-      "epoch": 3.77,
-      "learning_rate": 7.555185789964417e-06,
-      "loss": 0.258,
+      "epoch": 3.8604173424153965,
+      "grad_norm": 0.26256629824638367,
+      "learning_rate": 6.522025008683008e-06,
+      "loss": 0.3853,
       "step": 107115
     },
     {
-      "epoch": 3.77,
-      "learning_rate": 7.5531453644623525e-06,
-      "loss": 0.2532,
+      "epoch": 3.8605975420766208,
+      "grad_norm": 0.23125460743904114,
+      "learning_rate": 6.520059553074667e-06,
+      "loss": 0.3477,
       "step": 107120
     },
     {
-      "epoch": 3.77,
-      "learning_rate": 7.5511051654961915e-06,
-      "loss": 0.2477,
+      "epoch": 3.8607777417378455,
+      "grad_norm": 0.23365911841392517,
+      "learning_rate": 6.518094349251938e-06,
+      "loss": 0.3883,
       "step": 107125
     },
     {
-      "epoch": 3.77,
-      "learning_rate": 7.549065193092436e-06,
-      "loss": 0.2492,
+      "epoch": 3.8609579413990702,
+      "grad_norm": 0.18686018884181976,
+      "learning_rate": 6.516129397241588e-06,
+      "loss": 0.376,
       "step": 107130
     },
     {
-      "epoch": 3.77,
-      "learning_rate": 7.54702544727757e-06,
-      "loss": 0.2551,
+      "epoch": 3.861138141060295,
+      "grad_norm": 0.23002989590168,
+      "learning_rate": 6.5141646970704015e-06,
+      "loss": 0.3877,
       "step": 107135
     },
     {
-      "epoch": 3.77,
-      "learning_rate": 7.544985928078077e-06,
-      "loss": 0.2316,
+      "epoch": 3.8613183407215192,
+      "grad_norm": 0.3117145299911499,
+      "learning_rate": 6.512200248765146e-06,
+      "loss": 0.4153,
       "step": 107140
     },
     {
-      "epoch": 3.77,
-      "learning_rate": 7.542946635520429e-06,
-      "loss": 0.2486,
+      "epoch": 3.861498540382744,
+      "grad_norm": 0.2764013111591339,
+      "learning_rate": 6.510236052352581e-06,
+      "loss": 0.398,
       "step": 107145
     },
     {
-      "epoch": 3.77,
-      "learning_rate": 7.5409075696311225e-06,
-      "loss": 0.2684,
+      "epoch": 3.8616787400439687,
+      "grad_norm": 0.3027382493019104,
+      "learning_rate": 6.508272107859467e-06,
+      "loss": 0.3804,
       "step": 107150
     },
     {
-      "epoch": 3.77,
-      "learning_rate": 7.538868730436627e-06,
-      "loss": 0.2515,
+      "epoch": 3.8618589397051934,
+      "grad_norm": 0.20171962678432465,
+      "learning_rate": 6.50630841531257e-06,
+      "loss": 0.3427,
       "step": 107155
     },
     {
-      "epoch": 3.77,
-      "learning_rate": 7.536830117963403e-06,
-      "loss": 0.2779,
+      "epoch": 3.862039139366418,
+      "grad_norm": 0.2870292663574219,
+      "learning_rate": 6.50434497473863e-06,
+      "loss": 0.3651,
       "step": 107160
     },
     {
-      "epoch": 3.77,
-      "learning_rate": 7.534791732237934e-06,
-      "loss": 0.2644,
+      "epoch": 3.8622193390276425,
+      "grad_norm": 0.2380640059709549,
+      "learning_rate": 6.502381786164424e-06,
+      "loss": 0.3692,
       "step": 107165
     },
     {
-      "epoch": 3.77,
-      "learning_rate": 7.532753573286691e-06,
-      "loss": 0.2374,
+      "epoch": 3.862399538688867,
+      "grad_norm": 0.2691294550895691,
+      "learning_rate": 6.500418849616671e-06,
+      "loss": 0.3916,
       "step": 107170
     },
     {
-      "epoch": 3.77,
-      "learning_rate": 7.530715641136129e-06,
-      "loss": 0.2743,
+      "epoch": 3.862579738350092,
+      "grad_norm": 0.27460694313049316,
+      "learning_rate": 6.498456165122139e-06,
+      "loss": 0.379,
       "step": 107175
     },
     {
-      "epoch": 3.77,
-      "learning_rate": 7.528677935812714e-06,
-      "loss": 0.2512,
+      "epoch": 3.8627599380113167,
+      "grad_norm": 0.24331451952457428,
+      "learning_rate": 6.496493732707556e-06,
+      "loss": 0.3881,
       "step": 107180
     },
     {
-      "epoch": 3.77,
-      "learning_rate": 7.526640457342896e-06,
-      "loss": 0.2593,
+      "epoch": 3.862940137672541,
+      "grad_norm": 0.2511681020259857,
+      "learning_rate": 6.494531552399666e-06,
+      "loss": 0.371,
       "step": 107185
     },
     {
-      "epoch": 3.77,
-      "learning_rate": 7.5246032057531415e-06,
-      "loss": 0.246,
+      "epoch": 3.8631203373337657,
+      "grad_norm": 0.23705346882343292,
+      "learning_rate": 6.4925696242251975e-06,
+      "loss": 0.3931,
       "step": 107190
     },
     {
-      "epoch": 3.77,
-      "learning_rate": 7.522566181069899e-06,
-      "loss": 0.2391,
+      "epoch": 3.8633005369949904,
+      "grad_norm": 0.24382388591766357,
+      "learning_rate": 6.490607948210889e-06,
+      "loss": 0.3663,
       "step": 107195
     },
     {
-      "epoch": 3.77,
-      "learning_rate": 7.520529383319616e-06,
-      "loss": 0.2643,
+      "epoch": 3.863480736656215,
+      "grad_norm": 0.29807329177856445,
+      "learning_rate": 6.488646524383454e-06,
+      "loss": 0.3872,
       "step": 107200
     },
     {
-      "epoch": 3.77,
-      "learning_rate": 7.518492812528732e-06,
-      "loss": 0.2446,
+      "epoch": 3.86366093631744,
+      "grad_norm": 0.18684354424476624,
+      "learning_rate": 6.486685352769634e-06,
+      "loss": 0.363,
       "step": 107205
     },
     {
-      "epoch": 3.77,
-      "learning_rate": 7.516456468723707e-06,
-      "loss": 0.2642,
+      "epoch": 3.8638411359786646,
+      "grad_norm": 0.20785370469093323,
+      "learning_rate": 6.484724433396141e-06,
+      "loss": 0.3894,
       "step": 107210
     },
     {
-      "epoch": 3.77,
-      "learning_rate": 7.514420351930962e-06,
-      "loss": 0.284,
+      "epoch": 3.864021335639889,
+      "grad_norm": 0.2543993890285492,
+      "learning_rate": 6.482763766289693e-06,
+      "loss": 0.3472,
       "step": 107215
     },
     {
-      "epoch": 3.77,
-      "learning_rate": 7.512384462176955e-06,
-      "loss": 0.2778,
+      "epoch": 3.8642015353011137,
+      "grad_norm": 0.31936973333358765,
+      "learning_rate": 6.480803351477005e-06,
+      "loss": 0.3948,
       "step": 107220
     },
     {
-      "epoch": 3.77,
-      "learning_rate": 7.510348799488104e-06,
-      "loss": 0.2793,
+      "epoch": 3.8643817349623384,
+      "grad_norm": 0.33362480998039246,
+      "learning_rate": 6.478843188984776e-06,
+      "loss": 0.3973,
       "step": 107225
     },
     {
-      "epoch": 3.77,
-      "learning_rate": 7.508313363890853e-06,
-      "loss": 0.2586,
+      "epoch": 3.8645619346235627,
+      "grad_norm": 0.2270936667919159,
+      "learning_rate": 6.476883278839732e-06,
+      "loss": 0.3745,
       "step": 107230
     },
     {
-      "epoch": 3.77,
-      "learning_rate": 7.506278155411628e-06,
-      "loss": 0.2428,
+      "epoch": 3.8647421342847874,
+      "grad_norm": 0.22412335872650146,
+      "learning_rate": 6.474923621068574e-06,
+      "loss": 0.3999,
       "step": 107235
     },
     {
-      "epoch": 3.77,
-      "learning_rate": 7.504243174076841e-06,
-      "loss": 0.2601,
+      "epoch": 3.864922333946012,
+      "grad_norm": 0.23207208514213562,
+      "learning_rate": 6.472964215697982e-06,
+      "loss": 0.4043,
       "step": 107240
     },
     {
-      "epoch": 3.77,
-      "learning_rate": 7.502208419912937e-06,
-      "loss": 0.2664,
+      "epoch": 3.865102533607237,
+      "grad_norm": 0.23929710686206818,
+      "learning_rate": 6.47100506275467e-06,
+      "loss": 0.3865,
       "step": 107245
     },
     {
-      "epoch": 3.77,
-      "learning_rate": 7.500173892946328e-06,
-      "loss": 0.2744,
+      "epoch": 3.8652827332684616,
+      "grad_norm": 0.22255569696426392,
+      "learning_rate": 6.469046162265322e-06,
+      "loss": 0.3684,
       "step": 107250
     },
     {
-      "epoch": 3.77,
-      "learning_rate": 7.498139593203424e-06,
-      "loss": 0.2545,
+      "epoch": 3.8654629329296863,
+      "grad_norm": 0.2753349840641022,
+      "learning_rate": 6.467087514256645e-06,
+      "loss": 0.3796,
       "step": 107255
     },
     {
-      "epoch": 3.77,
-      "learning_rate": 7.496105520710636e-06,
-      "loss": 0.2582,
+      "epoch": 3.8656431325909106,
+      "grad_norm": 0.2791849374771118,
+      "learning_rate": 6.465129118755309e-06,
+      "loss": 0.3707,
       "step": 107260
     },
     {
-      "epoch": 3.77,
-      "learning_rate": 7.494071675494391e-06,
-      "loss": 0.263,
+      "epoch": 3.8658233322521354,
+      "grad_norm": 0.2520217299461365,
+      "learning_rate": 6.4631709757879885e-06,
+      "loss": 0.3908,
       "step": 107265
     },
     {
-      "epoch": 3.77,
-      "learning_rate": 7.49203805758108e-06,
-      "loss": 0.2693,
+      "epoch": 3.86600353191336,
+      "grad_norm": 0.2589038014411926,
+      "learning_rate": 6.461213085381384e-06,
+      "loss": 0.3644,
       "step": 107270
     },
     {
-      "epoch": 3.77,
-      "learning_rate": 7.490004666997125e-06,
-      "loss": 0.2386,
+      "epoch": 3.8661837315745844,
+      "grad_norm": 0.21371279656887054,
+      "learning_rate": 6.459255447562155e-06,
+      "loss": 0.3742,
       "step": 107275
     },
     {
-      "epoch": 3.77,
-      "learning_rate": 7.487971503768909e-06,
-      "loss": 0.2503,
+      "epoch": 3.866363931235809,
+      "grad_norm": 0.24295492470264435,
+      "learning_rate": 6.457298062356995e-06,
+      "loss": 0.3742,
       "step": 107280
     },
     {
-      "epoch": 3.77,
-      "learning_rate": 7.485938567922854e-06,
-      "loss": 0.2404,
+      "epoch": 3.866544130897034,
+      "grad_norm": 0.2283410131931305,
+      "learning_rate": 6.45534092979255e-06,
+      "loss": 0.3884,
       "step": 107285
     },
     {
-      "epoch": 3.77,
-      "learning_rate": 7.483905859485344e-06,
-      "loss": 0.2657,
+      "epoch": 3.8667243305582586,
+      "grad_norm": 0.20489716529846191,
+      "learning_rate": 6.453384049895489e-06,
+      "loss": 0.3688,
       "step": 107290
     },
     {
-      "epoch": 3.77,
-      "learning_rate": 7.481873378482771e-06,
-      "loss": 0.2587,
+      "epoch": 3.8669045302194833,
+      "grad_norm": 0.3497338891029358,
+      "learning_rate": 6.451427422692485e-06,
+      "loss": 0.3847,
       "step": 107295
     },
     {
-      "epoch": 3.78,
-      "learning_rate": 7.47984112494152e-06,
-      "loss": 0.2653,
+      "epoch": 3.867084729880708,
+      "grad_norm": 0.29012593626976013,
+      "learning_rate": 6.449471048210193e-06,
+      "loss": 0.3362,
       "step": 107300
     },
     {
-      "epoch": 3.78,
-      "learning_rate": 7.477809098887994e-06,
-      "loss": 0.2493,
+      "epoch": 3.8672649295419324,
+      "grad_norm": 0.24133215844631195,
+      "learning_rate": 6.4475149264752675e-06,
+      "loss": 0.3658,
       "step": 107305
     },
     {
-      "epoch": 3.78,
-      "learning_rate": 7.475777300348568e-06,
-      "loss": 0.2677,
+      "epoch": 3.867445129203157,
+      "grad_norm": 0.26846662163734436,
+      "learning_rate": 6.445559057514358e-06,
+      "loss": 0.3976,
       "step": 107310
     },
     {
-      "epoch": 3.78,
-      "learning_rate": 7.473745729349627e-06,
-      "loss": 0.2736,
+      "epoch": 3.867625328864382,
+      "grad_norm": 0.2922466993331909,
+      "learning_rate": 6.443603441354107e-06,
+      "loss": 0.404,
       "step": 107315
     },
     {
-      "epoch": 3.78,
-      "learning_rate": 7.471714385917539e-06,
-      "loss": 0.253,
+      "epoch": 3.867805528525606,
+      "grad_norm": 0.24670326709747314,
+      "learning_rate": 6.441648078021173e-06,
+      "loss": 0.3723,
       "step": 107320
     },
     {
-      "epoch": 3.78,
-      "learning_rate": 7.46968327007869e-06,
-      "loss": 0.2603,
+      "epoch": 3.867985728186831,
+      "grad_norm": 0.29100772738456726,
+      "learning_rate": 6.439692967542191e-06,
+      "loss": 0.4027,
       "step": 107325
     },
     {
-      "epoch": 3.78,
-      "learning_rate": 7.467652381859455e-06,
-      "loss": 0.2642,
+      "epoch": 3.8681659278480556,
+      "grad_norm": 0.21996453404426575,
+      "learning_rate": 6.4377381099438e-06,
+      "loss": 0.4048,
       "step": 107330
     },
     {
-      "epoch": 3.78,
-      "learning_rate": 7.4656217212862e-06,
-      "loss": 0.241,
+      "epoch": 3.8683461275092803,
+      "grad_norm": 0.2679681181907654,
+      "learning_rate": 6.435783505252632e-06,
+      "loss": 0.3685,
       "step": 107335
     },
     {
-      "epoch": 3.78,
-      "learning_rate": 7.463591288385285e-06,
-      "loss": 0.2435,
+      "epoch": 3.868526327170505,
+      "grad_norm": 0.2263759821653366,
+      "learning_rate": 6.4338291534953215e-06,
+      "loss": 0.3657,
       "step": 107340
     },
     {
-      "epoch": 3.78,
-      "learning_rate": 7.461561083183088e-06,
-      "loss": 0.2613,
+      "epoch": 3.8687065268317298,
+      "grad_norm": 0.2599135637283325,
+      "learning_rate": 6.431875054698486e-06,
+      "loss": 0.3653,
       "step": 107345
     },
     {
-      "epoch": 3.78,
-      "learning_rate": 7.459531105705964e-06,
-      "loss": 0.2122,
+      "epoch": 3.868886726492954,
+      "grad_norm": 0.28875550627708435,
+      "learning_rate": 6.429921208888773e-06,
+      "loss": 0.3754,
       "step": 107350
     },
     {
-      "epoch": 3.78,
-      "learning_rate": 7.457501355980256e-06,
-      "loss": 0.2452,
+      "epoch": 3.869066926154179,
+      "grad_norm": 0.23925039172172546,
+      "learning_rate": 6.4279676160927725e-06,
+      "loss": 0.4144,
       "step": 107355
     },
     {
-      "epoch": 3.78,
-      "learning_rate": 7.455471834032343e-06,
-      "loss": 0.2441,
+      "epoch": 3.8692471258154035,
+      "grad_norm": 0.3068399727344513,
+      "learning_rate": 6.426014276337125e-06,
+      "loss": 0.3856,
       "step": 107360
     },
     {
-      "epoch": 3.78,
-      "learning_rate": 7.453442539888566e-06,
-      "loss": 0.2422,
+      "epoch": 3.8694273254766283,
+      "grad_norm": 0.18254274129867554,
+      "learning_rate": 6.4240611896484365e-06,
+      "loss": 0.3931,
       "step": 107365
     },
     {
-      "epoch": 3.78,
-      "learning_rate": 7.451413473575275e-06,
-      "loss": 0.258,
+      "epoch": 3.8696075251378526,
+      "grad_norm": 0.20957380533218384,
+      "learning_rate": 6.422108356053319e-06,
+      "loss": 0.3773,
       "step": 107370
     },
     {
-      "epoch": 3.78,
-      "learning_rate": 7.449384635118805e-06,
-      "loss": 0.2488,
+      "epoch": 3.8697877247990773,
+      "grad_norm": 0.20488250255584717,
+      "learning_rate": 6.420155775578379e-06,
+      "loss": 0.3621,
       "step": 107375
     },
     {
-      "epoch": 3.78,
-      "learning_rate": 7.44735602454551e-06,
-      "loss": 0.2523,
+      "epoch": 3.869967924460302,
+      "grad_norm": 0.30153316259384155,
+      "learning_rate": 6.418203448250218e-06,
+      "loss": 0.4062,
       "step": 107380
     },
     {
-      "epoch": 3.78,
-      "learning_rate": 7.445327641881739e-06,
-      "loss": 0.2399,
+      "epoch": 3.8701481241215268,
+      "grad_norm": 0.27384403347969055,
+      "learning_rate": 6.416251374095431e-06,
+      "loss": 0.4122,
       "step": 107385
     },
     {
-      "epoch": 3.78,
-      "learning_rate": 7.443299487153818e-06,
-      "loss": 0.2626,
+      "epoch": 3.8703283237827515,
+      "grad_norm": 0.317130446434021,
+      "learning_rate": 6.414299553140629e-06,
+      "loss": 0.4039,
       "step": 107390
     },
     {
-      "epoch": 3.78,
-      "learning_rate": 7.4412715603880745e-06,
-      "loss": 0.274,
+      "epoch": 3.870508523443976,
+      "grad_norm": 0.21292860805988312,
+      "learning_rate": 6.412347985412395e-06,
+      "loss": 0.3729,
       "step": 107395
     },
     {
-      "epoch": 3.78,
-      "learning_rate": 7.439243861610856e-06,
-      "loss": 0.2693,
+      "epoch": 3.8706887231052005,
+      "grad_norm": 0.2204897701740265,
+      "learning_rate": 6.410396670937325e-06,
+      "loss": 0.3625,
       "step": 107400
     },
     {
-      "epoch": 3.78,
-      "learning_rate": 7.437216390848484e-06,
-      "loss": 0.2787,
+      "epoch": 3.8708689227664252,
+      "grad_norm": 0.2745380997657776,
+      "learning_rate": 6.4084456097419976e-06,
+      "loss": 0.4188,
       "step": 107405
     },
     {
-      "epoch": 3.78,
-      "learning_rate": 7.435189148127283e-06,
-      "loss": 0.2647,
+      "epoch": 3.87104912242765,
+      "grad_norm": 0.28390398621559143,
+      "learning_rate": 6.4064948018529915e-06,
+      "loss": 0.39,
       "step": 107410
     },
     {
-      "epoch": 3.78,
-      "learning_rate": 7.433162133473568e-06,
-      "loss": 0.2889,
+      "epoch": 3.8712293220888743,
+      "grad_norm": 0.2293710857629776,
+      "learning_rate": 6.404544247296903e-06,
+      "loss": 0.3506,
       "step": 107415
     },
     {
-      "epoch": 3.78,
-      "learning_rate": 7.43113534691367e-06,
-      "loss": 0.2834,
+      "epoch": 3.871409521750099,
+      "grad_norm": 0.22023165225982666,
+      "learning_rate": 6.4025939461003075e-06,
+      "loss": 0.3471,
       "step": 107420
     },
     {
-      "epoch": 3.78,
-      "learning_rate": 7.429108788473904e-06,
-      "loss": 0.2552,
+      "epoch": 3.8715897214113237,
+      "grad_norm": 0.25889191031455994,
+      "learning_rate": 6.400643898289751e-06,
+      "loss": 0.3967,
       "step": 107425
     },
     {
-      "epoch": 3.78,
-      "learning_rate": 7.427082458180573e-06,
-      "loss": 0.2517,
+      "epoch": 3.8717699210725485,
+      "grad_norm": 0.2643846273422241,
+      "learning_rate": 6.3986941038918305e-06,
+      "loss": 0.394,
       "step": 107430
     },
     {
-      "epoch": 3.78,
-      "learning_rate": 7.425056356059995e-06,
-      "loss": 0.2573,
+      "epoch": 3.871950120733773,
+      "grad_norm": 0.205980584025383,
+      "learning_rate": 6.396744562933094e-06,
+      "loss": 0.3505,
       "step": 107435
     },
     {
-      "epoch": 3.78,
-      "learning_rate": 7.423030482138485e-06,
-      "loss": 0.2585,
+      "epoch": 3.8721303203949975,
+      "grad_norm": 0.24297955632209778,
+      "learning_rate": 6.394795275440118e-06,
+      "loss": 0.3931,
       "step": 107440
     },
     {
-      "epoch": 3.78,
-      "learning_rate": 7.421004836442341e-06,
-      "loss": 0.2431,
+      "epoch": 3.8723105200562222,
+      "grad_norm": 0.22801950573921204,
+      "learning_rate": 6.392846241439462e-06,
+      "loss": 0.3853,
       "step": 107445
     },
     {
-      "epoch": 3.78,
-      "learning_rate": 7.418979418997865e-06,
-      "loss": 0.2421,
+      "epoch": 3.872490719717447,
+      "grad_norm": 0.20344804227352142,
+      "learning_rate": 6.390897460957657e-06,
+      "loss": 0.3943,
       "step": 107450
     },
     {
-      "epoch": 3.78,
-      "learning_rate": 7.416954229831347e-06,
-      "loss": 0.2353,
+      "epoch": 3.8726709193786717,
+      "grad_norm": 0.24186420440673828,
+      "learning_rate": 6.38894893402128e-06,
+      "loss": 0.368,
       "step": 107455
     },
     {
-      "epoch": 3.78,
-      "learning_rate": 7.414929268969095e-06,
-      "loss": 0.2701,
+      "epoch": 3.872851119039896,
+      "grad_norm": 0.28862589597702026,
+      "learning_rate": 6.387000660656869e-06,
+      "loss": 0.4025,
       "step": 107460
     },
     {
-      "epoch": 3.78,
-      "learning_rate": 7.412904536437401e-06,
-      "loss": 0.2685,
+      "epoch": 3.8730313187011207,
+      "grad_norm": 0.2653532922267914,
+      "learning_rate": 6.385052640890973e-06,
+      "loss": 0.4228,
       "step": 107465
     },
     {
-      "epoch": 3.78,
-      "learning_rate": 7.410880032262543e-06,
-      "loss": 0.2601,
+      "epoch": 3.8732115183623455,
+      "grad_norm": 0.20793728530406952,
+      "learning_rate": 6.383104874750129e-06,
+      "loss": 0.3892,
       "step": 107470
     },
     {
-      "epoch": 3.78,
-      "learning_rate": 7.408855756470823e-06,
-      "loss": 0.249,
+      "epoch": 3.87339171802357,
+      "grad_norm": 0.20338398218154907,
+      "learning_rate": 6.38115736226087e-06,
+      "loss": 0.3623,
       "step": 107475
     },
     {
-      "epoch": 3.78,
-      "learning_rate": 7.406831709088519e-06,
-      "loss": 0.2613,
+      "epoch": 3.873571917684795,
+      "grad_norm": 0.27274417877197266,
+      "learning_rate": 6.3792101034497454e-06,
+      "loss": 0.4013,
       "step": 107480
     },
     {
-      "epoch": 3.78,
-      "learning_rate": 7.4048078901419045e-06,
-      "loss": 0.2371,
+      "epoch": 3.8737521173460197,
+      "grad_norm": 0.21529635787010193,
+      "learning_rate": 6.3772630983432776e-06,
+      "loss": 0.3646,
       "step": 107485
     },
     {
-      "epoch": 3.78,
-      "learning_rate": 7.40278429965727e-06,
-      "loss": 0.2581,
+      "epoch": 3.873932317007244,
+      "grad_norm": 0.24685950577259064,
+      "learning_rate": 6.375316346967994e-06,
+      "loss": 0.3973,
       "step": 107490
     },
     {
-      "epoch": 3.78,
-      "learning_rate": 7.400760937660875e-06,
-      "loss": 0.28,
+      "epoch": 3.8741125166684687,
+      "grad_norm": 0.1983158439397812,
+      "learning_rate": 6.373369849350419e-06,
+      "loss": 0.368,
       "step": 107495
     },
     {
-      "epoch": 3.78,
-      "learning_rate": 7.39873780417901e-06,
-      "loss": 0.2608,
+      "epoch": 3.8742927163296934,
+      "grad_norm": 0.27724209427833557,
+      "learning_rate": 6.371423605517066e-06,
+      "loss": 0.3422,
       "step": 107500
     },
     {
-      "epoch": 3.78,
-      "eval_loss": 0.25113674998283386,
-      "eval_runtime": 10.5445,
-      "eval_samples_per_second": 9.484,
-      "eval_steps_per_second": 9.484,
+      "epoch": 3.8742927163296934,
+      "eval_loss": 0.4283193349838257,
+      "eval_runtime": 3.5283,
+      "eval_samples_per_second": 28.342,
+      "eval_steps_per_second": 7.086,
       "step": 107500
     },
     {
-      "epoch": 3.78,
-      "learning_rate": 7.3967148992379356e-06,
-      "loss": 0.2622,
+      "epoch": 3.8744729159909177,
+      "grad_norm": 0.25930055975914,
+      "learning_rate": 6.369477615494468e-06,
+      "loss": 0.3701,
       "step": 107505
     },
     {
-      "epoch": 3.78,
-      "learning_rate": 7.394692222863908e-06,
-      "loss": 0.249,
+      "epoch": 3.8746531156521424,
+      "grad_norm": 0.25892987847328186,
+      "learning_rate": 6.3675318793091295e-06,
+      "loss": 0.3995,
       "step": 107510
     },
     {
-      "epoch": 3.78,
-      "learning_rate": 7.3926697750832105e-06,
-      "loss": 0.2459,
+      "epoch": 3.874833315313367,
+      "grad_norm": 0.22906829416751862,
+      "learning_rate": 6.365586396987563e-06,
+      "loss": 0.3731,
       "step": 107515
     },
     {
-      "epoch": 3.78,
-      "learning_rate": 7.390647555922089e-06,
-      "loss": 0.2483,
+      "epoch": 3.875013514974592,
+      "grad_norm": 0.2710282802581787,
+      "learning_rate": 6.363641168556275e-06,
+      "loss": 0.4196,
       "step": 107520
     },
     {
-      "epoch": 3.78,
-      "learning_rate": 7.388625565406806e-06,
-      "loss": 0.2719,
+      "epoch": 3.8751937146358166,
+      "grad_norm": 0.24969340860843658,
+      "learning_rate": 6.361696194041766e-06,
+      "loss": 0.4083,
       "step": 107525
     },
     {
-      "epoch": 3.78,
-      "learning_rate": 7.386603803563605e-06,
-      "loss": 0.2514,
+      "epoch": 3.8753739142970414,
+      "grad_norm": 0.28726011514663696,
+      "learning_rate": 6.359751473470532e-06,
+      "loss": 0.4036,
       "step": 107530
     },
     {
-      "epoch": 3.78,
-      "learning_rate": 7.384582270418753e-06,
-      "loss": 0.2525,
+      "epoch": 3.8755541139582657,
+      "grad_norm": 0.2590154707431793,
+      "learning_rate": 6.35780700686909e-06,
+      "loss": 0.3776,
       "step": 107535
     },
     {
-      "epoch": 3.78,
-      "learning_rate": 7.382560965998483e-06,
-      "loss": 0.2545,
+      "epoch": 3.8757343136194904,
+      "grad_norm": 0.22933447360992432,
+      "learning_rate": 6.355862794263903e-06,
+      "loss": 0.3781,
       "step": 107540
     },
     {
-      "epoch": 3.78,
-      "learning_rate": 7.380539890329058e-06,
-      "loss": 0.2962,
+      "epoch": 3.875914513280715,
+      "grad_norm": 0.24526602029800415,
+      "learning_rate": 6.353918835681483e-06,
+      "loss": 0.3529,
       "step": 107545
     },
     {
-      "epoch": 3.78,
-      "learning_rate": 7.378519043436704e-06,
-      "loss": 0.2555,
+      "epoch": 3.8760947129419394,
+      "grad_norm": 0.24101245403289795,
+      "learning_rate": 6.351975131148308e-06,
+      "loss": 0.3841,
       "step": 107550
     },
     {
-      "epoch": 3.78,
-      "learning_rate": 7.3764984253476724e-06,
-      "loss": 0.2605,
+      "epoch": 3.876274912603164,
+      "grad_norm": 0.23179291188716888,
+      "learning_rate": 6.350031680690854e-06,
+      "loss": 0.3638,
       "step": 107555
     },
     {
-      "epoch": 3.78,
-      "learning_rate": 7.374478036088195e-06,
-      "loss": 0.2702,
+      "epoch": 3.876455112264389,
+      "grad_norm": 0.25088703632354736,
+      "learning_rate": 6.348088484335624e-06,
+      "loss": 0.3617,
       "step": 107560
     },
     {
-      "epoch": 3.78,
-      "learning_rate": 7.3724578756845076e-06,
-      "loss": 0.2675,
+      "epoch": 3.8766353119256136,
+      "grad_norm": 0.2836175262928009,
+      "learning_rate": 6.346145542109069e-06,
+      "loss": 0.3649,
       "step": 107565
     },
     {
-      "epoch": 3.78,
-      "learning_rate": 7.370437944162825e-06,
-      "loss": 0.2398,
+      "epoch": 3.8768155115868383,
+      "grad_norm": 0.22787512838840485,
+      "learning_rate": 6.344202854037662e-06,
+      "loss": 0.3655,
       "step": 107570
     },
     {
-      "epoch": 3.78,
-      "learning_rate": 7.368418241549399e-06,
-      "loss": 0.272,
+      "epoch": 3.876995711248063,
+      "grad_norm": 0.24610376358032227,
+      "learning_rate": 6.342260420147889e-06,
+      "loss": 0.3755,
       "step": 107575
     },
     {
-      "epoch": 3.78,
-      "learning_rate": 7.366398767870441e-06,
-      "loss": 0.2472,
+      "epoch": 3.8771759109092874,
+      "grad_norm": 0.270090788602829,
+      "learning_rate": 6.340318240466203e-06,
+      "loss": 0.3885,
       "step": 107580
     },
     {
-      "epoch": 3.79,
-      "learning_rate": 7.3643795231521675e-06,
-      "loss": 0.2548,
+      "epoch": 3.877356110570512,
+      "grad_norm": 0.24736785888671875,
+      "learning_rate": 6.3383763150190675e-06,
+      "loss": 0.3802,
       "step": 107585
     },
     {
-      "epoch": 3.79,
-      "learning_rate": 7.362360507420809e-06,
-      "loss": 0.2573,
+      "epoch": 3.877536310231737,
+      "grad_norm": 0.22846059501171112,
+      "learning_rate": 6.336434643832942e-06,
+      "loss": 0.3796,
       "step": 107590
     },
     {
-      "epoch": 3.79,
-      "learning_rate": 7.360341720702577e-06,
-      "loss": 0.2434,
+      "epoch": 3.877716509892961,
+      "grad_norm": 0.20484845340251923,
+      "learning_rate": 6.334493226934276e-06,
+      "loss": 0.3389,
       "step": 107595
     },
     {
-      "epoch": 3.79,
-      "learning_rate": 7.358323163023675e-06,
-      "loss": 0.2234,
+      "epoch": 3.877896709554186,
+      "grad_norm": 0.2559860050678253,
+      "learning_rate": 6.3325520643495314e-06,
+      "loss": 0.3823,
       "step": 107600
     },
     {
-      "epoch": 3.79,
-      "learning_rate": 7.356304834410321e-06,
-      "loss": 0.2456,
+      "epoch": 3.8780769092154106,
+      "grad_norm": 0.20136897265911102,
+      "learning_rate": 6.330611156105151e-06,
+      "loss": 0.38,
       "step": 107605
     },
     {
-      "epoch": 3.79,
-      "learning_rate": 7.354286734888729e-06,
-      "loss": 0.2644,
+      "epoch": 3.8782571088766353,
+      "grad_norm": 0.20288234949111938,
+      "learning_rate": 6.328670502227579e-06,
+      "loss": 0.3778,
       "step": 107610
     },
     {
-      "epoch": 3.79,
-      "learning_rate": 7.352268864485093e-06,
-      "loss": 0.2681,
+      "epoch": 3.87843730853786,
+      "grad_norm": 0.2542252838611603,
+      "learning_rate": 6.326730102743259e-06,
+      "loss": 0.3886,
       "step": 107615
     },
     {
-      "epoch": 3.79,
-      "learning_rate": 7.350251223225619e-06,
-      "loss": 0.2773,
+      "epoch": 3.878617508199085,
+      "grad_norm": 0.2526249587535858,
+      "learning_rate": 6.324789957678617e-06,
+      "loss": 0.3643,
       "step": 107620
     },
     {
-      "epoch": 3.79,
-      "learning_rate": 7.34823381113649e-06,
-      "loss": 0.2726,
+      "epoch": 3.878797707860309,
+      "grad_norm": 0.20341147482395172,
+      "learning_rate": 6.3228500670601015e-06,
+      "loss": 0.389,
       "step": 107625
     },
     {
-      "epoch": 3.79,
-      "learning_rate": 7.346216628243921e-06,
-      "loss": 0.2355,
+      "epoch": 3.878977907521534,
+      "grad_norm": 0.2713795602321625,
+      "learning_rate": 6.320910430914148e-06,
+      "loss": 0.4011,
       "step": 107630
     },
     {
-      "epoch": 3.79,
-      "learning_rate": 7.344199674574093e-06,
-      "loss": 0.2691,
+      "epoch": 3.8791581071827586,
+      "grad_norm": 0.2428542822599411,
+      "learning_rate": 6.318971049267159e-06,
+      "loss": 0.3911,
       "step": 107635
     },
     {
-      "epoch": 3.79,
-      "learning_rate": 7.342182950153198e-06,
-      "loss": 0.2433,
+      "epoch": 3.8793383068439833,
+      "grad_norm": 0.2431861311197281,
+      "learning_rate": 6.317031922145577e-06,
+      "loss": 0.4293,
       "step": 107640
     },
     {
-      "epoch": 3.79,
-      "learning_rate": 7.340166455007414e-06,
-      "loss": 0.2368,
+      "epoch": 3.8795185065052076,
+      "grad_norm": 0.23291923105716705,
+      "learning_rate": 6.315093049575821e-06,
+      "loss": 0.3808,
       "step": 107645
     },
     {
-      "epoch": 3.79,
-      "learning_rate": 7.338150189162937e-06,
-      "loss": 0.2437,
+      "epoch": 3.8796987061664323,
+      "grad_norm": 0.287264347076416,
+      "learning_rate": 6.313154431584303e-06,
+      "loss": 0.4031,
       "step": 107650
     },
     {
-      "epoch": 3.79,
-      "learning_rate": 7.336134152645932e-06,
-      "loss": 0.264,
+      "epoch": 3.879878905827657,
+      "grad_norm": 0.24084652960300446,
+      "learning_rate": 6.31121606819744e-06,
+      "loss": 0.3864,
       "step": 107655
     },
     {
-      "epoch": 3.79,
-      "learning_rate": 7.334118345482593e-06,
-      "loss": 0.2433,
+      "epoch": 3.8800591054888818,
+      "grad_norm": 0.25108230113983154,
+      "learning_rate": 6.309277959441629e-06,
+      "loss": 0.4094,
       "step": 107660
     },
     {
-      "epoch": 3.79,
-      "learning_rate": 7.332102767699076e-06,
-      "loss": 0.2492,
+      "epoch": 3.8802393051501065,
+      "grad_norm": 0.24399013817310333,
+      "learning_rate": 6.307340105343298e-06,
+      "loss": 0.3851,
       "step": 107665
     },
     {
-      "epoch": 3.79,
-      "learning_rate": 7.3300874193215715e-06,
-      "loss": 0.2647,
+      "epoch": 3.880419504811331,
+      "grad_norm": 0.20561757683753967,
+      "learning_rate": 6.305402505928837e-06,
+      "loss": 0.3682,
       "step": 107670
     },
     {
-      "epoch": 3.79,
-      "learning_rate": 7.328072300376234e-06,
-      "loss": 0.2439,
+      "epoch": 3.8805997044725555,
+      "grad_norm": 0.2230025976896286,
+      "learning_rate": 6.3034651612246476e-06,
+      "loss": 0.4135,
       "step": 107675
     },
     {
-      "epoch": 3.79,
-      "learning_rate": 7.326057410889231e-06,
-      "loss": 0.2655,
+      "epoch": 3.8807799041337803,
+      "grad_norm": 0.2320537567138672,
+      "learning_rate": 6.301528071257127e-06,
+      "loss": 0.3782,
       "step": 107680
     },
     {
-      "epoch": 3.79,
-      "learning_rate": 7.32404275088672e-06,
-      "loss": 0.2685,
+      "epoch": 3.880960103795005,
+      "grad_norm": 0.28236204385757446,
+      "learning_rate": 6.299591236052657e-06,
+      "loss": 0.4315,
       "step": 107685
     },
     {
-      "epoch": 3.79,
-      "learning_rate": 7.32202832039487e-06,
-      "loss": 0.2541,
+      "epoch": 3.8811403034562293,
+      "grad_norm": 0.24441473186016083,
+      "learning_rate": 6.297654655637644e-06,
+      "loss": 0.3819,
       "step": 107690
     },
     {
-      "epoch": 3.79,
-      "learning_rate": 7.320014119439833e-06,
-      "loss": 0.254,
+      "epoch": 3.881320503117454,
+      "grad_norm": 0.2429056018590927,
+      "learning_rate": 6.295718330038466e-06,
+      "loss": 0.4267,
       "step": 107695
     },
     {
-      "epoch": 3.79,
-      "learning_rate": 7.318000148047754e-06,
-      "loss": 0.2618,
+      "epoch": 3.8815007027786788,
+      "grad_norm": 0.26638370752334595,
+      "learning_rate": 6.293782259281503e-06,
+      "loss": 0.3785,
       "step": 107700
     },
     {
-      "epoch": 3.79,
-      "learning_rate": 7.315986406244798e-06,
-      "loss": 0.255,
+      "epoch": 3.8816809024399035,
+      "grad_norm": 0.23976294696331024,
+      "learning_rate": 6.291846443393137e-06,
+      "loss": 0.3971,
       "step": 107705
     },
     {
-      "epoch": 3.79,
-      "learning_rate": 7.313972894057094e-06,
-      "loss": 0.2475,
+      "epoch": 3.8818611021011282,
+      "grad_norm": 0.2684692144393921,
+      "learning_rate": 6.289910882399741e-06,
+      "loss": 0.4069,
       "step": 107710
     },
     {
-      "epoch": 3.79,
-      "learning_rate": 7.311959611510807e-06,
-      "loss": 0.2343,
+      "epoch": 3.882041301762353,
+      "grad_norm": 0.2549930512905121,
+      "learning_rate": 6.287975576327679e-06,
+      "loss": 0.379,
       "step": 107715
     },
     {
-      "epoch": 3.79,
-      "learning_rate": 7.309946558632061e-06,
-      "loss": 0.2428,
+      "epoch": 3.8822215014235772,
+      "grad_norm": 0.24471694231033325,
+      "learning_rate": 6.286040525203332e-06,
+      "loss": 0.3727,
       "step": 107720
     },
     {
-      "epoch": 3.79,
-      "learning_rate": 7.307933735447006e-06,
-      "loss": 0.2493,
+      "epoch": 3.882401701084802,
+      "grad_norm": 0.2403818666934967,
+      "learning_rate": 6.284105729053064e-06,
+      "loss": 0.3695,
       "step": 107725
     },
     {
-      "epoch": 3.79,
-      "learning_rate": 7.305921141981772e-06,
-      "loss": 0.2648,
+      "epoch": 3.8825819007460267,
+      "grad_norm": 0.24353429675102234,
+      "learning_rate": 6.28217118790323e-06,
+      "loss": 0.394,
       "step": 107730
     },
     {
-      "epoch": 3.79,
-      "learning_rate": 7.303908778262491e-06,
-      "loss": 0.2529,
+      "epoch": 3.882762100407251,
+      "grad_norm": 0.17580963671207428,
+      "learning_rate": 6.28023690178019e-06,
+      "loss": 0.3487,
       "step": 107735
     },
     {
-      "epoch": 3.79,
-      "learning_rate": 7.301896644315287e-06,
-      "loss": 0.2375,
+      "epoch": 3.8829423000684757,
+      "grad_norm": 0.21670427918434143,
+      "learning_rate": 6.27830287071029e-06,
+      "loss": 0.3557,
       "step": 107740
     },
     {
-      "epoch": 3.79,
-      "learning_rate": 7.299884740166297e-06,
-      "loss": 0.2622,
+      "epoch": 3.8831224997297005,
+      "grad_norm": 0.26354971528053284,
+      "learning_rate": 6.276369094719903e-06,
+      "loss": 0.3939,
       "step": 107745
     },
     {
-      "epoch": 3.79,
-      "learning_rate": 7.29787306584164e-06,
-      "loss": 0.2582,
+      "epoch": 3.883302699390925,
+      "grad_norm": 0.24392694234848022,
+      "learning_rate": 6.274435573835355e-06,
+      "loss": 0.3638,
       "step": 107750
     },
     {
-      "epoch": 3.79,
-      "learning_rate": 7.295861621367434e-06,
-      "loss": 0.2783,
+      "epoch": 3.88348289905215,
+      "grad_norm": 0.2409679889678955,
+      "learning_rate": 6.272502308082989e-06,
+      "loss": 0.3691,
       "step": 107755
     },
     {
-      "epoch": 3.79,
-      "learning_rate": 7.293850406769789e-06,
-      "loss": 0.2627,
+      "epoch": 3.8836630987133747,
+      "grad_norm": 0.24243175983428955,
+      "learning_rate": 6.270569297489162e-06,
+      "loss": 0.3579,
       "step": 107760
     },
     {
-      "epoch": 3.79,
-      "learning_rate": 7.2918394220748306e-06,
-      "loss": 0.2553,
+      "epoch": 3.883843298374599,
+      "grad_norm": 0.19919145107269287,
+      "learning_rate": 6.2686365420802e-06,
+      "loss": 0.3814,
       "step": 107765
     },
     {
-      "epoch": 3.79,
-      "learning_rate": 7.289828667308673e-06,
-      "loss": 0.2623,
+      "epoch": 3.8840234980358237,
+      "grad_norm": 0.2675543427467346,
+      "learning_rate": 6.266704041882443e-06,
+      "loss": 0.3378,
       "step": 107770
     },
     {
-      "epoch": 3.79,
-      "learning_rate": 7.287818142497418e-06,
-      "loss": 0.2408,
+      "epoch": 3.8842036976970484,
+      "grad_norm": 0.2584518492221832,
+      "learning_rate": 6.264771796922212e-06,
+      "loss": 0.361,
       "step": 107775
     },
     {
-      "epoch": 3.79,
-      "learning_rate": 7.285807847667167e-06,
-      "loss": 0.2554,
+      "epoch": 3.8843838973582727,
+      "grad_norm": 0.23442451655864716,
+      "learning_rate": 6.262839807225834e-06,
+      "loss": 0.3821,
       "step": 107780
     },
     {
-      "epoch": 3.79,
-      "learning_rate": 7.283797782844032e-06,
-      "loss": 0.2503,
+      "epoch": 3.8845640970194975,
+      "grad_norm": 0.24751421809196472,
+      "learning_rate": 6.260908072819641e-06,
+      "loss": 0.3798,
       "step": 107785
     },
     {
-      "epoch": 3.79,
-      "learning_rate": 7.281787948054109e-06,
-      "loss": 0.2431,
+      "epoch": 3.884744296680722,
+      "grad_norm": 0.2060132473707199,
+      "learning_rate": 6.258976593729948e-06,
+      "loss": 0.3664,
       "step": 107790
     },
     {
-      "epoch": 3.79,
-      "learning_rate": 7.27977834332349e-06,
-      "loss": 0.2696,
+      "epoch": 3.884924496341947,
+      "grad_norm": 0.18985074758529663,
+      "learning_rate": 6.2570453699830725e-06,
+      "loss": 0.417,
       "step": 107795
     },
     {
-      "epoch": 3.79,
-      "learning_rate": 7.277768968678264e-06,
-      "loss": 0.2607,
+      "epoch": 3.8851046960031717,
+      "grad_norm": 0.21168529987335205,
+      "learning_rate": 6.2551144016053224e-06,
+      "loss": 0.371,
       "step": 107800
     },
     {
-      "epoch": 3.79,
-      "learning_rate": 7.27575982414454e-06,
-      "loss": 0.2608,
+      "epoch": 3.8852848956643964,
+      "grad_norm": 0.21965354681015015,
+      "learning_rate": 6.253183688623002e-06,
+      "loss": 0.3586,
       "step": 107805
     },
     {
-      "epoch": 3.79,
-      "learning_rate": 7.2737509097483894e-06,
-      "loss": 0.274,
+      "epoch": 3.8854650953256207,
+      "grad_norm": 0.2647111117839813,
+      "learning_rate": 6.251253231062435e-06,
+      "loss": 0.4069,
       "step": 107810
     },
     {
-      "epoch": 3.79,
-      "learning_rate": 7.271742225515896e-06,
-      "loss": 0.2475,
+      "epoch": 3.8856452949868454,
+      "grad_norm": 0.2614315152168274,
+      "learning_rate": 6.249323028949916e-06,
+      "loss": 0.4049,
       "step": 107815
     },
     {
-      "epoch": 3.79,
-      "learning_rate": 7.269733771473147e-06,
-      "loss": 0.2381,
+      "epoch": 3.88582549464807,
+      "grad_norm": 0.24572202563285828,
+      "learning_rate": 6.247393082311728e-06,
+      "loss": 0.4089,
       "step": 107820
     },
     {
-      "epoch": 3.79,
-      "learning_rate": 7.2677255476462255e-06,
-      "loss": 0.2625,
+      "epoch": 3.8860056943092944,
+      "grad_norm": 0.21743349730968475,
+      "learning_rate": 6.245463391174186e-06,
+      "loss": 0.3464,
       "step": 107825
     },
     {
-      "epoch": 3.79,
-      "learning_rate": 7.265717554061205e-06,
-      "loss": 0.248,
+      "epoch": 3.886185893970519,
+      "grad_norm": 0.21934011578559875,
+      "learning_rate": 6.243533955563574e-06,
+      "loss": 0.3717,
       "step": 107830
     },
     {
-      "epoch": 3.79,
-      "learning_rate": 7.2637097907441444e-06,
-      "loss": 0.2609,
+      "epoch": 3.886366093631744,
+      "grad_norm": 0.2820332646369934,
+      "learning_rate": 6.241604775506174e-06,
+      "loss": 0.3887,
       "step": 107835
     },
     {
-      "epoch": 3.79,
-      "learning_rate": 7.261702257721131e-06,
-      "loss": 0.2537,
+      "epoch": 3.8865462932929686,
+      "grad_norm": 0.2698332965373993,
+      "learning_rate": 6.2396758510282895e-06,
+      "loss": 0.3956,
       "step": 107840
     },
     {
-      "epoch": 3.79,
-      "learning_rate": 7.259694955018226e-06,
-      "loss": 0.2211,
+      "epoch": 3.8867264929541934,
+      "grad_norm": 0.2532370090484619,
+      "learning_rate": 6.237747182156178e-06,
+      "loss": 0.3619,
       "step": 107845
     },
     {
-      "epoch": 3.79,
-      "learning_rate": 7.25768788266149e-06,
-      "loss": 0.2568,
+      "epoch": 3.886906692615418,
+      "grad_norm": 0.275337278842926,
+      "learning_rate": 6.2358187689161356e-06,
+      "loss": 0.3903,
       "step": 107850
     },
     {
-      "epoch": 3.79,
-      "learning_rate": 7.255681040676976e-06,
-      "loss": 0.2581,
+      "epoch": 3.8870868922766424,
+      "grad_norm": 0.2451581060886383,
+      "learning_rate": 6.233890611334428e-06,
+      "loss": 0.3933,
       "step": 107855
     },
     {
-      "epoch": 3.79,
-      "learning_rate": 7.253674429090756e-06,
-      "loss": 0.2605,
+      "epoch": 3.887267091937867,
+      "grad_norm": 0.2677507698535919,
+      "learning_rate": 6.231962709437328e-06,
+      "loss": 0.3806,
       "step": 107860
     },
     {
-      "epoch": 3.79,
-      "learning_rate": 7.25166804792888e-06,
-      "loss": 0.2435,
+      "epoch": 3.887447291599092,
+      "grad_norm": 0.26012924313545227,
+      "learning_rate": 6.230035063251102e-06,
+      "loss": 0.3458,
       "step": 107865
     },
     {
-      "epoch": 3.8,
-      "learning_rate": 7.249661897217391e-06,
-      "loss": 0.26,
+      "epoch": 3.8876274912603166,
+      "grad_norm": 0.2477564662694931,
+      "learning_rate": 6.2281076728020085e-06,
+      "loss": 0.394,
       "step": 107870
     },
     {
-      "epoch": 3.8,
-      "learning_rate": 7.247655976982343e-06,
-      "loss": 0.2477,
+      "epoch": 3.887807690921541,
+      "grad_norm": 0.2495788335800171,
+      "learning_rate": 6.226180538116319e-06,
+      "loss": 0.3995,
       "step": 107875
     },
     {
-      "epoch": 3.8,
-      "learning_rate": 7.245650287249791e-06,
-      "loss": 0.2548,
+      "epoch": 3.8879878905827656,
+      "grad_norm": 0.2793433666229248,
+      "learning_rate": 6.224253659220286e-06,
+      "loss": 0.3702,
       "step": 107880
     },
     {
-      "epoch": 3.8,
-      "learning_rate": 7.243644828045768e-06,
-      "loss": 0.2486,
+      "epoch": 3.8881680902439903,
+      "grad_norm": 0.19013087451457977,
+      "learning_rate": 6.22232703614016e-06,
+      "loss": 0.3745,
       "step": 107885
     },
     {
-      "epoch": 3.8,
-      "learning_rate": 7.241639599396316e-06,
-      "loss": 0.2293,
+      "epoch": 3.888348289905215,
+      "grad_norm": 0.2680392265319824,
+      "learning_rate": 6.220400668902196e-06,
+      "loss": 0.3855,
       "step": 107890
     },
     {
-      "epoch": 3.8,
-      "learning_rate": 7.2396346013274635e-06,
-      "loss": 0.2371,
+      "epoch": 3.88852848956644,
+      "grad_norm": 0.29071950912475586,
+      "learning_rate": 6.218474557532633e-06,
+      "loss": 0.4051,
       "step": 107895
     },
     {
-      "epoch": 3.8,
-      "learning_rate": 7.237629833865259e-06,
-      "loss": 0.2498,
+      "epoch": 3.888708689227664,
+      "grad_norm": 0.27325236797332764,
+      "learning_rate": 6.216548702057715e-06,
+      "loss": 0.3856,
       "step": 107900
     },
     {
-      "epoch": 3.8,
-      "learning_rate": 7.235625297035725e-06,
-      "loss": 0.2501,
+      "epoch": 3.888888888888889,
+      "grad_norm": 0.23505853116512299,
+      "learning_rate": 6.214623102503689e-06,
+      "loss": 0.3783,
       "step": 107905
     },
     {
-      "epoch": 3.8,
-      "learning_rate": 7.233620990864889e-06,
-      "loss": 0.2624,
+      "epoch": 3.8890690885501136,
+      "grad_norm": 0.23786808550357819,
+      "learning_rate": 6.212697758896788e-06,
+      "loss": 0.3829,
       "step": 107910
     },
     {
-      "epoch": 3.8,
-      "learning_rate": 7.231616915378769e-06,
-      "loss": 0.2747,
+      "epoch": 3.8892492882113383,
+      "grad_norm": 0.4009203612804413,
+      "learning_rate": 6.2107726712632404e-06,
+      "loss": 0.3708,
       "step": 107915
     },
     {
-      "epoch": 3.8,
-      "learning_rate": 7.2296130706033996e-06,
-      "loss": 0.2571,
+      "epoch": 3.8894294878725626,
+      "grad_norm": 0.20991608500480652,
+      "learning_rate": 6.208847839629278e-06,
+      "loss": 0.3835,
       "step": 107920
     },
     {
-      "epoch": 3.8,
-      "learning_rate": 7.227609456564788e-06,
-      "loss": 0.2528,
+      "epoch": 3.8896096875337873,
+      "grad_norm": 0.22346235811710358,
+      "learning_rate": 6.206923264021119e-06,
+      "loss": 0.3753,
       "step": 107925
     },
     {
-      "epoch": 3.8,
-      "learning_rate": 7.225606073288962e-06,
-      "loss": 0.2426,
+      "epoch": 3.889789887195012,
+      "grad_norm": 0.22620312869548798,
+      "learning_rate": 6.204998944465007e-06,
+      "loss": 0.3604,
       "step": 107930
     },
     {
-      "epoch": 3.8,
-      "learning_rate": 7.223602920801919e-06,
-      "loss": 0.2591,
+      "epoch": 3.889970086856237,
+      "grad_norm": 0.2643108665943146,
+      "learning_rate": 6.203074880987137e-06,
+      "loss": 0.3826,
       "step": 107935
     },
     {
-      "epoch": 3.8,
-      "learning_rate": 7.221599999129686e-06,
-      "loss": 0.2634,
+      "epoch": 3.8901502865174615,
+      "grad_norm": 0.25667524337768555,
+      "learning_rate": 6.201151073613726e-06,
+      "loss": 0.3917,
       "step": 107940
     },
     {
-      "epoch": 3.8,
-      "learning_rate": 7.219597308298259e-06,
-      "loss": 0.2832,
+      "epoch": 3.890330486178686,
+      "grad_norm": 0.20191670954227448,
+      "learning_rate": 6.199227522371001e-06,
+      "loss": 0.3856,
       "step": 107945
     },
     {
-      "epoch": 3.8,
-      "learning_rate": 7.217594848333636e-06,
-      "loss": 0.2459,
+      "epoch": 3.8905106858399106,
+      "grad_norm": 0.25496187806129456,
+      "learning_rate": 6.197304227285158e-06,
+      "loss": 0.3649,
       "step": 107950
     },
     {
-      "epoch": 3.8,
-      "learning_rate": 7.215592619261832e-06,
-      "loss": 0.257,
+      "epoch": 3.8906908855011353,
+      "grad_norm": 0.2135312408208847,
+      "learning_rate": 6.195381188382407e-06,
+      "loss": 0.3425,
       "step": 107955
     },
     {
-      "epoch": 3.8,
-      "learning_rate": 7.21359062110884e-06,
-      "loss": 0.2516,
+      "epoch": 3.89087108516236,
+      "grad_norm": 0.20012623071670532,
+      "learning_rate": 6.193458405688945e-06,
+      "loss": 0.3803,
       "step": 107960
     },
     {
-      "epoch": 3.8,
-      "learning_rate": 7.21158885390065e-06,
-      "loss": 0.2523,
+      "epoch": 3.8910512848235843,
+      "grad_norm": 0.2358315885066986,
+      "learning_rate": 6.191535879230964e-06,
+      "loss": 0.3862,
       "step": 107965
     },
     {
-      "epoch": 3.8,
-      "learning_rate": 7.209587317663246e-06,
-      "loss": 0.2596,
+      "epoch": 3.891231484484809,
+      "grad_norm": 0.2576768398284912,
+      "learning_rate": 6.1896136090346755e-06,
+      "loss": 0.407,
       "step": 107970
     },
     {
-      "epoch": 3.8,
-      "learning_rate": 7.207586012422637e-06,
-      "loss": 0.2645,
+      "epoch": 3.891411684146034,
+      "grad_norm": 0.23550166189670563,
+      "learning_rate": 6.187691595126255e-06,
+      "loss": 0.3778,
       "step": 107975
     },
     {
-      "epoch": 3.8,
-      "learning_rate": 7.20558493820479e-06,
-      "loss": 0.2577,
+      "epoch": 3.8915918838072585,
+      "grad_norm": 0.22159500420093536,
+      "learning_rate": 6.185769837531899e-06,
+      "loss": 0.3908,
       "step": 107980
     },
     {
-      "epoch": 3.8,
-      "learning_rate": 7.203584095035704e-06,
-      "loss": 0.2718,
+      "epoch": 3.8917720834684832,
+      "grad_norm": 0.27781954407691956,
+      "learning_rate": 6.183848336277784e-06,
+      "loss": 0.3946,
       "step": 107985
     },
     {
-      "epoch": 3.8,
-      "learning_rate": 7.2015834829413406e-06,
-      "loss": 0.246,
+      "epoch": 3.891952283129708,
+      "grad_norm": 0.23066557943820953,
+      "learning_rate": 6.1819270913900826e-06,
+      "loss": 0.394,
       "step": 107990
     },
     {
-      "epoch": 3.8,
-      "learning_rate": 7.199583101947696e-06,
-      "loss": 0.2511,
+      "epoch": 3.8921324827909323,
+      "grad_norm": 0.26918134093284607,
+      "learning_rate": 6.1800061028949916e-06,
+      "loss": 0.3916,
       "step": 107995
     },
     {
-      "epoch": 3.8,
-      "learning_rate": 7.197582952080731e-06,
-      "loss": 0.2522,
+      "epoch": 3.892312682452157,
+      "grad_norm": 0.24214857816696167,
+      "learning_rate": 6.178085370818676e-06,
+      "loss": 0.3739,
       "step": 108000
     },
     {
-      "epoch": 3.8,
-      "eval_loss": 0.25110575556755066,
-      "eval_runtime": 10.5564,
-      "eval_samples_per_second": 9.473,
-      "eval_steps_per_second": 9.473,
+      "epoch": 3.892312682452157,
+      "eval_loss": 0.42808079719543457,
+      "eval_runtime": 3.5205,
+      "eval_samples_per_second": 28.405,
+      "eval_steps_per_second": 7.101,
       "step": 108000
     },
     {
-      "epoch": 3.8,
-      "learning_rate": 7.195583033366421e-06,
-      "loss": 0.252,
+      "epoch": 3.8924928821133817,
+      "grad_norm": 0.18488271534442902,
+      "learning_rate": 6.1761648951873e-06,
+      "loss": 0.3708,
       "step": 108005
     },
     {
-      "epoch": 3.8,
-      "learning_rate": 7.193583345830723e-06,
-      "loss": 0.2597,
+      "epoch": 3.892673081774606,
+      "grad_norm": 0.3259855806827545,
+      "learning_rate": 6.174244676027033e-06,
+      "loss": 0.36,
       "step": 108010
     },
     {
-      "epoch": 3.8,
-      "learning_rate": 7.19158388949962e-06,
-      "loss": 0.2421,
+      "epoch": 3.8928532814358308,
+      "grad_norm": 0.25019964575767517,
+      "learning_rate": 6.172324713364039e-06,
+      "loss": 0.3678,
       "step": 108015
     },
     {
-      "epoch": 3.8,
-      "learning_rate": 7.189584664399063e-06,
-      "loss": 0.2783,
+      "epoch": 3.8930334810970555,
+      "grad_norm": 0.24694189429283142,
+      "learning_rate": 6.170405007224467e-06,
+      "loss": 0.3751,
       "step": 108020
     },
     {
-      "epoch": 3.8,
-      "learning_rate": 7.187585670555011e-06,
-      "loss": 0.2669,
+      "epoch": 3.8932136807582802,
+      "grad_norm": 0.1833256483078003,
+      "learning_rate": 6.168485557634496e-06,
+      "loss": 0.3654,
       "step": 108025
     },
     {
-      "epoch": 3.8,
-      "learning_rate": 7.185586907993413e-06,
-      "loss": 0.2566,
+      "epoch": 3.893393880419505,
+      "grad_norm": 0.2395724505186081,
+      "learning_rate": 6.166566364620249e-06,
+      "loss": 0.3517,
       "step": 108030
     },
     {
-      "epoch": 3.8,
-      "learning_rate": 7.1835883767402385e-06,
-      "loss": 0.2678,
+      "epoch": 3.8935740800807297,
+      "grad_norm": 0.2357412576675415,
+      "learning_rate": 6.164647428207896e-06,
+      "loss": 0.3885,
       "step": 108035
     },
     {
-      "epoch": 3.8,
-      "learning_rate": 7.181590076821418e-06,
-      "loss": 0.2312,
+      "epoch": 3.893754279741954,
+      "grad_norm": 0.25304263830184937,
+      "learning_rate": 6.162728748423577e-06,
+      "loss": 0.3779,
       "step": 108040
     },
     {
-      "epoch": 3.8,
-      "learning_rate": 7.1795920082629144e-06,
-      "loss": 0.2733,
+      "epoch": 3.8939344794031787,
+      "grad_norm": 0.23290584981441498,
+      "learning_rate": 6.160810325293429e-06,
+      "loss": 0.3866,
       "step": 108045
     },
     {
-      "epoch": 3.8,
-      "learning_rate": 7.177594171090657e-06,
-      "loss": 0.2529,
+      "epoch": 3.8941146790644035,
+      "grad_norm": 0.25585252046585083,
+      "learning_rate": 6.158892158843593e-06,
+      "loss": 0.3682,
       "step": 108050
     },
     {
-      "epoch": 3.8,
-      "learning_rate": 7.1755965653306026e-06,
-      "loss": 0.2586,
+      "epoch": 3.8942948787256277,
+      "grad_norm": 0.24071577191352844,
+      "learning_rate": 6.1569742491002054e-06,
+      "loss": 0.3762,
       "step": 108055
     },
     {
-      "epoch": 3.8,
-      "learning_rate": 7.173599191008678e-06,
-      "loss": 0.2694,
+      "epoch": 3.8944750783868525,
+      "grad_norm": 0.21371334791183472,
+      "learning_rate": 6.155056596089387e-06,
+      "loss": 0.3549,
       "step": 108060
     },
     {
-      "epoch": 3.8,
-      "learning_rate": 7.171602048150816e-06,
-      "loss": 0.2616,
+      "epoch": 3.894655278048077,
+      "grad_norm": 0.23668819665908813,
+      "learning_rate": 6.153139199837282e-06,
+      "loss": 0.383,
       "step": 108065
     },
     {
-      "epoch": 3.8,
-      "learning_rate": 7.169605136782956e-06,
-      "loss": 0.2726,
+      "epoch": 3.894835477709302,
+      "grad_norm": 0.23611487448215485,
+      "learning_rate": 6.151222060370007e-06,
+      "loss": 0.3706,
       "step": 108070
     },
     {
-      "epoch": 3.8,
-      "learning_rate": 7.167608456931024e-06,
-      "loss": 0.2387,
+      "epoch": 3.8950156773705267,
+      "grad_norm": 0.22055992484092712,
+      "learning_rate": 6.149305177713682e-06,
+      "loss": 0.4095,
       "step": 108075
     },
     {
-      "epoch": 3.8,
-      "learning_rate": 7.165612008620945e-06,
-      "loss": 0.2768,
+      "epoch": 3.8951958770317514,
+      "grad_norm": 0.20526739954948425,
+      "learning_rate": 6.147388551894423e-06,
+      "loss": 0.3806,
       "step": 108080
     },
     {
-      "epoch": 3.8,
-      "learning_rate": 7.163615791878634e-06,
-      "loss": 0.2449,
+      "epoch": 3.8953760766929757,
+      "grad_norm": 0.27778366208076477,
+      "learning_rate": 6.145472182938339e-06,
+      "loss": 0.4111,
       "step": 108085
     },
     {
-      "epoch": 3.8,
-      "learning_rate": 7.1616198067300215e-06,
-      "loss": 0.2485,
+      "epoch": 3.8955562763542004,
+      "grad_norm": 0.2274710088968277,
+      "learning_rate": 6.143556070871554e-06,
+      "loss": 0.3905,
       "step": 108090
     },
     {
-      "epoch": 3.8,
-      "learning_rate": 7.1596240532010136e-06,
-      "loss": 0.2662,
+      "epoch": 3.895736476015425,
+      "grad_norm": 0.29294195771217346,
+      "learning_rate": 6.141640215720165e-06,
+      "loss": 0.3839,
       "step": 108095
     },
     {
-      "epoch": 3.8,
-      "learning_rate": 7.157628531317539e-06,
-      "loss": 0.239,
+      "epoch": 3.8959166756766495,
+      "grad_norm": 0.2358580380678177,
+      "learning_rate": 6.139724617510279e-06,
+      "loss": 0.3886,
       "step": 108100
     },
     {
-      "epoch": 3.8,
-      "learning_rate": 7.1556332411054915e-06,
-      "loss": 0.2711,
+      "epoch": 3.896096875337874,
+      "grad_norm": 0.2260809987783432,
+      "learning_rate": 6.137809276267992e-06,
+      "loss": 0.348,
       "step": 108105
     },
     {
-      "epoch": 3.8,
-      "learning_rate": 7.153638182590794e-06,
-      "loss": 0.2437,
+      "epoch": 3.896277074999099,
+      "grad_norm": 0.22863958775997162,
+      "learning_rate": 6.1358941920193955e-06,
+      "loss": 0.368,
       "step": 108110
     },
     {
-      "epoch": 3.8,
-      "learning_rate": 7.1516433557993426e-06,
-      "loss": 0.2689,
+      "epoch": 3.8964572746603237,
+      "grad_norm": 0.22534294426441193,
+      "learning_rate": 6.133979364790601e-06,
+      "loss": 0.3866,
       "step": 108115
     },
     {
-      "epoch": 3.8,
-      "learning_rate": 7.149648760757041e-06,
-      "loss": 0.2491,
+      "epoch": 3.8966374743215484,
+      "grad_norm": 0.22448866069316864,
+      "learning_rate": 6.132064794607681e-06,
+      "loss": 0.3379,
       "step": 108120
     },
     {
-      "epoch": 3.8,
-      "learning_rate": 7.147654397489778e-06,
-      "loss": 0.2512,
+      "epoch": 3.896817673982773,
+      "grad_norm": 0.3200676143169403,
+      "learning_rate": 6.130150481496716e-06,
+      "loss": 0.3729,
       "step": 108125
     },
     {
-      "epoch": 3.8,
-      "learning_rate": 7.145660266023466e-06,
-      "loss": 0.2689,
+      "epoch": 3.8969978736439974,
+      "grad_norm": 0.2216705083847046,
+      "learning_rate": 6.128236425483805e-06,
+      "loss": 0.3849,
       "step": 108130
     },
     {
-      "epoch": 3.8,
-      "learning_rate": 7.1436663663839855e-06,
-      "loss": 0.2484,
+      "epoch": 3.897178073305222,
+      "grad_norm": 0.3131025433540344,
+      "learning_rate": 6.12632262659501e-06,
+      "loss": 0.358,
       "step": 108135
     },
     {
-      "epoch": 3.8,
-      "learning_rate": 7.1416726985972264e-06,
-      "loss": 0.2558,
+      "epoch": 3.897358272966447,
+      "grad_norm": 0.23903706669807434,
+      "learning_rate": 6.124409084856433e-06,
+      "loss": 0.3908,
       "step": 108140
     },
     {
-      "epoch": 3.8,
-      "learning_rate": 7.139679262689083e-06,
-      "loss": 0.274,
+      "epoch": 3.8975384726276716,
+      "grad_norm": 0.3344826400279999,
+      "learning_rate": 6.122495800294117e-06,
+      "loss": 0.4013,
       "step": 108145
     },
     {
-      "epoch": 3.81,
-      "learning_rate": 7.137686058685428e-06,
-      "loss": 0.2741,
+      "epoch": 3.897718672288896,
+      "grad_norm": 0.248307466506958,
+      "learning_rate": 6.120582772934136e-06,
+      "loss": 0.377,
       "step": 108150
     },
     {
-      "epoch": 3.81,
-      "learning_rate": 7.135693086612155e-06,
-      "loss": 0.2981,
+      "epoch": 3.8978988719501206,
+      "grad_norm": 0.21172760426998138,
+      "learning_rate": 6.118670002802568e-06,
+      "loss": 0.3875,
       "step": 108155
     },
     {
-      "epoch": 3.81,
-      "learning_rate": 7.133700346495134e-06,
-      "loss": 0.2628,
+      "epoch": 3.8980790716113454,
+      "grad_norm": 0.24647603929042816,
+      "learning_rate": 6.116757489925462e-06,
+      "loss": 0.3586,
       "step": 108160
     },
     {
-      "epoch": 3.81,
-      "learning_rate": 7.131707838360233e-06,
-      "loss": 0.2518,
+      "epoch": 3.89825927127257,
+      "grad_norm": 0.20428353548049927,
+      "learning_rate": 6.114845234328881e-06,
+      "loss": 0.3424,
       "step": 108165
     },
     {
-      "epoch": 3.81,
-      "learning_rate": 7.1297155622333356e-06,
-      "loss": 0.2673,
+      "epoch": 3.898439470933795,
+      "grad_norm": 0.2739085257053375,
+      "learning_rate": 6.112933236038878e-06,
+      "loss": 0.3761,
       "step": 108170
     },
     {
-      "epoch": 3.81,
-      "learning_rate": 7.1277235181403054e-06,
-      "loss": 0.2507,
+      "epoch": 3.898619670595019,
+      "grad_norm": 0.21551024913787842,
+      "learning_rate": 6.111021495081496e-06,
+      "loss": 0.3632,
       "step": 108175
     },
     {
-      "epoch": 3.81,
-      "learning_rate": 7.1257317061070004e-06,
-      "loss": 0.2517,
+      "epoch": 3.898799870256244,
+      "grad_norm": 0.2526455819606781,
+      "learning_rate": 6.109110011482797e-06,
+      "loss": 0.3779,
       "step": 108180
     },
     {
-      "epoch": 3.81,
-      "learning_rate": 7.123740126159298e-06,
-      "loss": 0.2719,
+      "epoch": 3.8989800699174686,
+      "grad_norm": 0.2527839243412018,
+      "learning_rate": 6.107198785268814e-06,
+      "loss": 0.3568,
       "step": 108185
     },
     {
-      "epoch": 3.81,
-      "learning_rate": 7.121748778323048e-06,
-      "loss": 0.2713,
+      "epoch": 3.8991602695786933,
+      "grad_norm": 0.24524226784706116,
+      "learning_rate": 6.105287816465591e-06,
+      "loss": 0.3872,
       "step": 108190
     },
     {
-      "epoch": 3.81,
-      "learning_rate": 7.119757662624107e-06,
-      "loss": 0.2425,
+      "epoch": 3.8993404692399176,
+      "grad_norm": 0.20940566062927246,
+      "learning_rate": 6.103377105099165e-06,
+      "loss": 0.3772,
       "step": 108195
     },
     {
-      "epoch": 3.81,
-      "learning_rate": 7.117766779088322e-06,
-      "loss": 0.2567,
+      "epoch": 3.8995206689011424,
+      "grad_norm": 0.2865338921546936,
+      "learning_rate": 6.101466651195564e-06,
+      "loss": 0.4084,
       "step": 108200
     },
     {
-      "epoch": 3.81,
-      "learning_rate": 7.115776127741553e-06,
-      "loss": 0.2699,
+      "epoch": 3.899700868562367,
+      "grad_norm": 0.2509433925151825,
+      "learning_rate": 6.099556454780817e-06,
+      "loss": 0.3579,
       "step": 108205
     },
     {
-      "epoch": 3.81,
-      "learning_rate": 7.113785708609649e-06,
-      "loss": 0.2453,
+      "epoch": 3.899881068223592,
+      "grad_norm": 0.271535187959671,
+      "learning_rate": 6.097646515880967e-06,
+      "loss": 0.4306,
       "step": 108210
     },
     {
-      "epoch": 3.81,
-      "learning_rate": 7.11179552171845e-06,
-      "loss": 0.277,
+      "epoch": 3.9000612678848166,
+      "grad_norm": 0.2970694899559021,
+      "learning_rate": 6.095736834522009e-06,
+      "loss": 0.3645,
       "step": 108215
     },
     {
-      "epoch": 3.81,
-      "learning_rate": 7.109805567093792e-06,
-      "loss": 0.2807,
+      "epoch": 3.9002414675460413,
+      "grad_norm": 0.2571132779121399,
+      "learning_rate": 6.093827410729985e-06,
+      "loss": 0.4004,
       "step": 108220
     },
     {
-      "epoch": 3.81,
-      "learning_rate": 7.107815844761525e-06,
-      "loss": 0.2494,
+      "epoch": 3.9004216672072656,
+      "grad_norm": 0.23718644678592682,
+      "learning_rate": 6.091918244530903e-06,
+      "loss": 0.3924,
       "step": 108225
     },
     {
-      "epoch": 3.81,
-      "learning_rate": 7.10582635474748e-06,
-      "loss": 0.255,
+      "epoch": 3.9006018668684903,
+      "grad_norm": 0.24922458827495575,
+      "learning_rate": 6.090009335950772e-06,
+      "loss": 0.3443,
       "step": 108230
     },
     {
-      "epoch": 3.81,
-      "learning_rate": 7.103837097077484e-06,
-      "loss": 0.2396,
+      "epoch": 3.900782066529715,
+      "grad_norm": 0.2874678075313568,
+      "learning_rate": 6.088100685015605e-06,
+      "loss": 0.3818,
       "step": 108235
     },
     {
-      "epoch": 3.81,
-      "learning_rate": 7.101848071777364e-06,
-      "loss": 0.2441,
+      "epoch": 3.9009622661909393,
+      "grad_norm": 0.23355595767498016,
+      "learning_rate": 6.0861922917514024e-06,
+      "loss": 0.391,
       "step": 108240
     },
     {
-      "epoch": 3.81,
-      "learning_rate": 7.099859278872958e-06,
-      "loss": 0.2754,
+      "epoch": 3.901142465852164,
+      "grad_norm": 0.2503456473350525,
+      "learning_rate": 6.084284156184161e-06,
+      "loss": 0.3889,
       "step": 108245
     },
     {
-      "epoch": 3.81,
-      "learning_rate": 7.097870718390084e-06,
-      "loss": 0.2485,
+      "epoch": 3.901322665513389,
+      "grad_norm": 0.2787312865257263,
+      "learning_rate": 6.082376278339893e-06,
+      "loss": 0.4176,
       "step": 108250
     },
     {
-      "epoch": 3.81,
-      "learning_rate": 7.095882390354553e-06,
-      "loss": 0.2541,
+      "epoch": 3.9015028651746135,
+      "grad_norm": 0.24109527468681335,
+      "learning_rate": 6.080468658244587e-06,
+      "loss": 0.3879,
       "step": 108255
     },
     {
-      "epoch": 3.81,
-      "learning_rate": 7.093894294792186e-06,
-      "loss": 0.2545,
+      "epoch": 3.9016830648358383,
+      "grad_norm": 0.24064235389232635,
+      "learning_rate": 6.078561295924232e-06,
+      "loss": 0.4142,
       "step": 108260
     },
     {
-      "epoch": 3.81,
-      "learning_rate": 7.091906431728812e-06,
-      "loss": 0.2585,
+      "epoch": 3.901863264497063,
+      "grad_norm": 0.20909632742404938,
+      "learning_rate": 6.076654191404815e-06,
+      "loss": 0.3915,
       "step": 108265
     },
     {
-      "epoch": 3.81,
-      "learning_rate": 7.089918801190232e-06,
-      "loss": 0.2477,
+      "epoch": 3.9020434641582873,
+      "grad_norm": 0.21806110441684723,
+      "learning_rate": 6.074747344712314e-06,
+      "loss": 0.3774,
       "step": 108270
     },
     {
-      "epoch": 3.81,
-      "learning_rate": 7.087931403202253e-06,
-      "loss": 0.2772,
+      "epoch": 3.902223663819512,
+      "grad_norm": 0.21959331631660461,
+      "learning_rate": 6.0728407558727246e-06,
+      "loss": 0.375,
       "step": 108275
     },
     {
-      "epoch": 3.81,
-      "learning_rate": 7.08594423779067e-06,
-      "loss": 0.2597,
+      "epoch": 3.9024038634807368,
+      "grad_norm": 0.20421810448169708,
+      "learning_rate": 6.070934424912014e-06,
+      "loss": 0.3855,
       "step": 108280
     },
     {
-      "epoch": 3.81,
-      "learning_rate": 7.083957304981303e-06,
-      "loss": 0.2318,
+      "epoch": 3.902584063141961,
+      "grad_norm": 0.18650266528129578,
+      "learning_rate": 6.06902835185616e-06,
+      "loss": 0.3661,
       "step": 108285
     },
     {
-      "epoch": 3.81,
-      "learning_rate": 7.081970604799945e-06,
-      "loss": 0.2652,
+      "epoch": 3.902764262803186,
+      "grad_norm": 0.23381759226322174,
+      "learning_rate": 6.067122536731126e-06,
+      "loss": 0.4026,
       "step": 108290
     },
     {
-      "epoch": 3.81,
-      "learning_rate": 7.079984137272383e-06,
-      "loss": 0.2515,
+      "epoch": 3.9029444624644105,
+      "grad_norm": 0.19069373607635498,
+      "learning_rate": 6.065216979562877e-06,
+      "loss": 0.3993,
       "step": 108295
     },
     {
-      "epoch": 3.81,
-      "learning_rate": 7.0779979024244245e-06,
-      "loss": 0.2409,
+      "epoch": 3.9031246621256352,
+      "grad_norm": 0.22146674990653992,
+      "learning_rate": 6.063311680377387e-06,
+      "loss": 0.3903,
       "step": 108300
     },
     {
-      "epoch": 3.81,
-      "learning_rate": 7.076011900281851e-06,
-      "loss": 0.2472,
+      "epoch": 3.90330486178686,
+      "grad_norm": 0.24637946486473083,
+      "learning_rate": 6.061406639200617e-06,
+      "loss": 0.3661,
       "step": 108305
     },
     {
-      "epoch": 3.81,
-      "learning_rate": 7.074026130870445e-06,
-      "loss": 0.264,
+      "epoch": 3.9034850614480847,
+      "grad_norm": 0.2656784653663635,
+      "learning_rate": 6.0595018560585e-06,
+      "loss": 0.3357,
       "step": 108310
     },
     {
-      "epoch": 3.81,
-      "learning_rate": 7.072040594215995e-06,
-      "loss": 0.2683,
+      "epoch": 3.903665261109309,
+      "grad_norm": 0.21832560002803802,
+      "learning_rate": 6.057597330977011e-06,
+      "loss": 0.348,
       "step": 108315
     },
     {
-      "epoch": 3.81,
-      "learning_rate": 7.070055290344291e-06,
-      "loss": 0.2634,
+      "epoch": 3.9038454607705337,
+      "grad_norm": 0.215951070189476,
+      "learning_rate": 6.055693063982082e-06,
+      "loss": 0.3295,
       "step": 108320
     },
     {
-      "epoch": 3.81,
-      "learning_rate": 7.0680702192811e-06,
-      "loss": 0.2489,
+      "epoch": 3.9040256604317585,
+      "grad_norm": 0.23906680941581726,
+      "learning_rate": 6.053789055099685e-06,
+      "loss": 0.3876,
       "step": 108325
     },
     {
-      "epoch": 3.81,
-      "learning_rate": 7.066085381052204e-06,
-      "loss": 0.2795,
+      "epoch": 3.9042058600929828,
+      "grad_norm": 0.2423383742570877,
+      "learning_rate": 6.051885304355734e-06,
+      "loss": 0.3528,
       "step": 108330
     },
     {
-      "epoch": 3.81,
-      "learning_rate": 7.064100775683358e-06,
-      "loss": 0.24,
+      "epoch": 3.9043860597542075,
+      "grad_norm": 0.20754733681678772,
+      "learning_rate": 6.04998181177617e-06,
+      "loss": 0.3529,
       "step": 108335
     },
     {
-      "epoch": 3.81,
-      "learning_rate": 7.0621164032003555e-06,
-      "loss": 0.2562,
+      "epoch": 3.9045662594154322,
+      "grad_norm": 0.23662780225276947,
+      "learning_rate": 6.048078577386945e-06,
+      "loss": 0.3659,
       "step": 108340
     },
     {
-      "epoch": 3.81,
-      "learning_rate": 7.060132263628946e-06,
-      "loss": 0.261,
+      "epoch": 3.904746459076657,
+      "grad_norm": 0.3288610577583313,
+      "learning_rate": 6.046175601213977e-06,
+      "loss": 0.4009,
       "step": 108345
     },
     {
-      "epoch": 3.81,
-      "learning_rate": 7.058148356994898e-06,
-      "loss": 0.2728,
+      "epoch": 3.9049266587378817,
+      "grad_norm": 0.30411192774772644,
+      "learning_rate": 6.044272883283198e-06,
+      "loss": 0.3728,
       "step": 108350
     },
     {
-      "epoch": 3.81,
-      "learning_rate": 7.056164683323962e-06,
-      "loss": 0.2721,
+      "epoch": 3.9051068583991064,
+      "grad_norm": 0.2672629654407501,
+      "learning_rate": 6.04237042362053e-06,
+      "loss": 0.3629,
       "step": 108355
     },
     {
-      "epoch": 3.81,
-      "learning_rate": 7.054181242641911e-06,
-      "loss": 0.276,
+      "epoch": 3.9052870580603307,
+      "grad_norm": 0.22953931987285614,
+      "learning_rate": 6.040468222251891e-06,
+      "loss": 0.3809,
       "step": 108360
     },
     {
-      "epoch": 3.81,
-      "learning_rate": 7.05219803497448e-06,
-      "loss": 0.2347,
+      "epoch": 3.9054672577215555,
+      "grad_norm": 0.22573940455913544,
+      "learning_rate": 6.038566279203206e-06,
+      "loss": 0.3482,
       "step": 108365
     },
     {
-      "epoch": 3.81,
-      "learning_rate": 7.05021506034744e-06,
-      "loss": 0.2393,
+      "epoch": 3.90564745738278,
+      "grad_norm": 0.29565203189849854,
+      "learning_rate": 6.036664594500385e-06,
+      "loss": 0.3928,
       "step": 108370
     },
     {
-      "epoch": 3.81,
-      "learning_rate": 7.048232318786519e-06,
-      "loss": 0.2607,
+      "epoch": 3.905827657044005,
+      "grad_norm": 0.2615247070789337,
+      "learning_rate": 6.0347631681693384e-06,
+      "loss": 0.3485,
       "step": 108375
     },
     {
-      "epoch": 3.81,
-      "learning_rate": 7.046249810317476e-06,
-      "loss": 0.2788,
+      "epoch": 3.906007856705229,
+      "grad_norm": 0.3057554364204407,
+      "learning_rate": 6.0328620002359695e-06,
+      "loss": 0.3922,
       "step": 108380
     },
     {
-      "epoch": 3.81,
-      "learning_rate": 7.044267534966051e-06,
-      "loss": 0.248,
+      "epoch": 3.906188056366454,
+      "grad_norm": 0.23635023832321167,
+      "learning_rate": 6.030961090726186e-06,
+      "loss": 0.4425,
       "step": 108385
     },
     {
-      "epoch": 3.81,
-      "learning_rate": 7.042285492757975e-06,
-      "loss": 0.2547,
+      "epoch": 3.9063682560276787,
+      "grad_norm": 0.16913343966007233,
+      "learning_rate": 6.029060439665876e-06,
+      "loss": 0.3574,
       "step": 108390
     },
     {
-      "epoch": 3.81,
-      "learning_rate": 7.040303683718982e-06,
-      "loss": 0.2323,
+      "epoch": 3.9065484556889034,
+      "grad_norm": 0.20368297398090363,
+      "learning_rate": 6.0271600470809595e-06,
+      "loss": 0.3818,
       "step": 108395
     },
     {
-      "epoch": 3.81,
-      "learning_rate": 7.038322107874817e-06,
-      "loss": 0.2419,
+      "epoch": 3.906728655350128,
+      "grad_norm": 0.27825161814689636,
+      "learning_rate": 6.0252599129973e-06,
+      "loss": 0.3998,
       "step": 108400
     },
     {
-      "epoch": 3.81,
-      "learning_rate": 7.036340765251204e-06,
-      "loss": 0.2638,
+      "epoch": 3.9069088550113524,
+      "grad_norm": 0.2604582905769348,
+      "learning_rate": 6.023360037440809e-06,
+      "loss": 0.4151,
       "step": 108405
     },
     {
-      "epoch": 3.81,
-      "learning_rate": 7.034359655873857e-06,
-      "loss": 0.2656,
+      "epoch": 3.907089054672577,
+      "grad_norm": 0.27497240900993347,
+      "learning_rate": 6.021460420437364e-06,
+      "loss": 0.382,
       "step": 108410
     },
     {
-      "epoch": 3.81,
-      "learning_rate": 7.03237877976852e-06,
-      "loss": 0.2576,
+      "epoch": 3.907269254333802,
+      "grad_norm": 0.18511195480823517,
+      "learning_rate": 6.0195610620128354e-06,
+      "loss": 0.389,
       "step": 108415
     },
     {
-      "epoch": 3.81,
-      "learning_rate": 7.030398136960894e-06,
-      "loss": 0.2746,
+      "epoch": 3.9074494539950266,
+      "grad_norm": 0.24932751059532166,
+      "learning_rate": 6.01766196219313e-06,
+      "loss": 0.3824,
       "step": 108420
     },
     {
-      "epoch": 3.81,
-      "learning_rate": 7.0284177274767135e-06,
-      "loss": 0.2601,
+      "epoch": 3.907629653656251,
+      "grad_norm": 0.24645911157131195,
+      "learning_rate": 6.0157631210040975e-06,
+      "loss": 0.3741,
       "step": 108425
     },
     {
-      "epoch": 3.81,
-      "learning_rate": 7.0264375513416776e-06,
-      "loss": 0.2409,
+      "epoch": 3.9078098533174757,
+      "grad_norm": 0.24892014265060425,
+      "learning_rate": 6.0138645384716085e-06,
+      "loss": 0.368,
       "step": 108430
     },
     {
-      "epoch": 3.82,
-      "learning_rate": 7.024457608581514e-06,
-      "loss": 0.2737,
+      "epoch": 3.9079900529787004,
+      "grad_norm": 0.22853389382362366,
+      "learning_rate": 6.0119662146215475e-06,
+      "loss": 0.3671,
       "step": 108435
     },
     {
-      "epoch": 3.82,
-      "learning_rate": 7.0224778992219214e-06,
-      "loss": 0.2879,
+      "epoch": 3.908170252639925,
+      "grad_norm": 0.2502167820930481,
+      "learning_rate": 6.010068149479772e-06,
+      "loss": 0.4101,
       "step": 108440
     },
     {
-      "epoch": 3.82,
-      "learning_rate": 7.020498423288604e-06,
-      "loss": 0.2708,
+      "epoch": 3.90835045230115,
+      "grad_norm": 0.25325870513916016,
+      "learning_rate": 6.008170343072139e-06,
+      "loss": 0.3676,
       "step": 108445
     },
     {
-      "epoch": 3.82,
-      "learning_rate": 7.018519180807262e-06,
-      "loss": 0.2499,
+      "epoch": 3.908530651962374,
+      "grad_norm": 0.269979864358902,
+      "learning_rate": 6.006272795424511e-06,
+      "loss": 0.3602,
       "step": 108450
     },
     {
-      "epoch": 3.82,
-      "learning_rate": 7.016540171803601e-06,
-      "loss": 0.2447,
+      "epoch": 3.908710851623599,
+      "grad_norm": 0.21623685956001282,
+      "learning_rate": 6.004375506562729e-06,
+      "loss": 0.3969,
       "step": 108455
     },
     {
-      "epoch": 3.82,
-      "learning_rate": 7.014561396303318e-06,
-      "loss": 0.2602,
+      "epoch": 3.9088910512848236,
+      "grad_norm": 0.20743024349212646,
+      "learning_rate": 6.002478476512663e-06,
+      "loss": 0.3587,
       "step": 108460
     },
     {
-      "epoch": 3.82,
-      "learning_rate": 7.012582854332101e-06,
-      "loss": 0.2647,
+      "epoch": 3.9090712509460483,
+      "grad_norm": 0.26383909583091736,
+      "learning_rate": 6.0005817053001485e-06,
+      "loss": 0.3759,
       "step": 108465
     },
     {
-      "epoch": 3.82,
-      "learning_rate": 7.0106045459156334e-06,
-      "loss": 0.2633,
+      "epoch": 3.9092514506072726,
+      "grad_norm": 0.22323699295520782,
+      "learning_rate": 5.998685192951029e-06,
+      "loss": 0.371,
       "step": 108470
     },
     {
-      "epoch": 3.82,
-      "learning_rate": 7.008626471079619e-06,
-      "loss": 0.2714,
+      "epoch": 3.9094316502684974,
+      "grad_norm": 0.2346304953098297,
+      "learning_rate": 5.996788939491143e-06,
+      "loss": 0.3776,
       "step": 108475
     },
     {
-      "epoch": 3.82,
-      "learning_rate": 7.006648629849724e-06,
-      "loss": 0.2677,
+      "epoch": 3.909611849929722,
+      "grad_norm": 0.25308549404144287,
+      "learning_rate": 5.994892944946326e-06,
+      "loss": 0.3947,
       "step": 108480
     },
     {
-      "epoch": 3.82,
-      "learning_rate": 7.004671022251649e-06,
-      "loss": 0.2508,
+      "epoch": 3.909792049590947,
+      "grad_norm": 0.24598398804664612,
+      "learning_rate": 5.992997209342416e-06,
+      "loss": 0.3543,
       "step": 108485
     },
     {
-      "epoch": 3.82,
-      "learning_rate": 7.002693648311051e-06,
-      "loss": 0.2567,
+      "epoch": 3.9099722492521716,
+      "grad_norm": 0.21500460803508759,
+      "learning_rate": 5.991101732705248e-06,
+      "loss": 0.3669,
       "step": 108490
     },
     {
-      "epoch": 3.82,
-      "learning_rate": 7.000716508053623e-06,
-      "loss": 0.2636,
+      "epoch": 3.9101524489133963,
+      "grad_norm": 0.2395174652338028,
+      "learning_rate": 5.989206515060625e-06,
+      "loss": 0.3834,
       "step": 108495
     },
     {
-      "epoch": 3.82,
-      "learning_rate": 6.99873960150503e-06,
-      "loss": 0.2607,
+      "epoch": 3.9103326485746206,
+      "grad_norm": 0.20296072959899902,
+      "learning_rate": 5.987311556434391e-06,
+      "loss": 0.4221,
       "step": 108500
     },
     {
-      "epoch": 3.82,
-      "eval_loss": 0.25093138217926025,
-      "eval_runtime": 10.5444,
-      "eval_samples_per_second": 9.484,
-      "eval_steps_per_second": 9.484,
+      "epoch": 3.9103326485746206,
+      "eval_loss": 0.4281059205532074,
+      "eval_runtime": 3.5316,
+      "eval_samples_per_second": 28.316,
+      "eval_steps_per_second": 7.079,
       "step": 108500
     },
     {
-      "epoch": 3.82,
-      "learning_rate": 6.9967629286909416e-06,
-      "loss": 0.241,
+      "epoch": 3.9105128482358453,
+      "grad_norm": 0.22557717561721802,
+      "learning_rate": 5.985416856852347e-06,
+      "loss": 0.3589,
       "step": 108505
     },
     {
-      "epoch": 3.82,
-      "learning_rate": 6.994786489637012e-06,
-      "loss": 0.26,
+      "epoch": 3.91069304789707,
+      "grad_norm": 0.22067588567733765,
+      "learning_rate": 5.9835224163403315e-06,
+      "loss": 0.3639,
       "step": 108510
     },
     {
-      "epoch": 3.82,
-      "learning_rate": 6.9928102843689225e-06,
-      "loss": 0.2456,
+      "epoch": 3.9108732475582944,
+      "grad_norm": 0.2571333050727844,
+      "learning_rate": 5.9816282349241335e-06,
+      "loss": 0.3619,
       "step": 108515
     },
     {
-      "epoch": 3.82,
-      "learning_rate": 6.990834312912323e-06,
-      "loss": 0.2521,
+      "epoch": 3.911053447219519,
+      "grad_norm": 0.24989810585975647,
+      "learning_rate": 5.979734312629562e-06,
+      "loss": 0.4091,
       "step": 108520
     },
     {
-      "epoch": 3.82,
-      "learning_rate": 6.988858575292867e-06,
-      "loss": 0.2463,
+      "epoch": 3.911233646880744,
+      "grad_norm": 0.2283269464969635,
+      "learning_rate": 5.977840649482435e-06,
+      "loss": 0.3804,
       "step": 108525
     },
     {
-      "epoch": 3.82,
-      "learning_rate": 6.9868830715362185e-06,
-      "loss": 0.2718,
+      "epoch": 3.9114138465419686,
+      "grad_norm": 0.2447991967201233,
+      "learning_rate": 5.975947245508548e-06,
+      "loss": 0.3941,
       "step": 108530
     },
     {
-      "epoch": 3.82,
-      "learning_rate": 6.98530283692935e-06,
-      "loss": 0.2599,
+      "epoch": 3.9115940462031933,
+      "grad_norm": 0.27993232011795044,
+      "learning_rate": 5.9740541007336974e-06,
+      "loss": 0.3952,
       "step": 108535
     },
     {
-      "epoch": 3.82,
-      "learning_rate": 6.983327754190374e-06,
-      "loss": 0.2566,
+      "epoch": 3.911774245864418,
+      "grad_norm": 0.21602769196033478,
+      "learning_rate": 5.972161215183672e-06,
+      "loss": 0.3507,
       "step": 108540
     },
     {
-      "epoch": 3.82,
-      "learning_rate": 6.981352905386021e-06,
-      "loss": 0.2568,
+      "epoch": 3.9119544455256423,
+      "grad_norm": 0.2476750910282135,
+      "learning_rate": 5.970268588884262e-06,
+      "loss": 0.3867,
       "step": 108545
     },
     {
-      "epoch": 3.82,
-      "learning_rate": 6.979378290541927e-06,
-      "loss": 0.2647,
+      "epoch": 3.912134645186867,
+      "grad_norm": 0.2227691113948822,
+      "learning_rate": 5.968376221861266e-06,
+      "loss": 0.3658,
       "step": 108550
     },
     {
-      "epoch": 3.82,
-      "learning_rate": 6.97740390968373e-06,
-      "loss": 0.2363,
+      "epoch": 3.9123148448480918,
+      "grad_norm": 0.19756188988685608,
+      "learning_rate": 5.9664841141404585e-06,
+      "loss": 0.3666,
       "step": 108555
     },
     {
-      "epoch": 3.82,
-      "learning_rate": 6.975429762837058e-06,
-      "loss": 0.2646,
+      "epoch": 3.912495044509316,
+      "grad_norm": 0.2740320563316345,
+      "learning_rate": 5.964592265747617e-06,
+      "loss": 0.3738,
       "step": 108560
     },
     {
-      "epoch": 3.82,
-      "learning_rate": 6.97345585002755e-06,
-      "loss": 0.258,
+      "epoch": 3.912675244170541,
+      "grad_norm": 0.28433406352996826,
+      "learning_rate": 5.9627006767085244e-06,
+      "loss": 0.3794,
       "step": 108565
     },
     {
-      "epoch": 3.82,
-      "learning_rate": 6.971482171280846e-06,
-      "loss": 0.2637,
+      "epoch": 3.9128554438317655,
+      "grad_norm": 0.3014092445373535,
+      "learning_rate": 5.960809347048948e-06,
+      "loss": 0.3592,
       "step": 108570
     },
     {
-      "epoch": 3.82,
-      "learning_rate": 6.969508726622565e-06,
-      "loss": 0.2563,
+      "epoch": 3.9130356434929903,
+      "grad_norm": 0.2842918038368225,
+      "learning_rate": 5.95891827679465e-06,
+      "loss": 0.3823,
       "step": 108575
     },
     {
-      "epoch": 3.82,
-      "learning_rate": 6.967535516078325e-06,
-      "loss": 0.2778,
+      "epoch": 3.913215843154215,
+      "grad_norm": 0.27905863523483276,
+      "learning_rate": 5.957027465971413e-06,
+      "loss": 0.3583,
       "step": 108580
     },
     {
-      "epoch": 3.82,
-      "learning_rate": 6.965562539673756e-06,
-      "loss": 0.2618,
+      "epoch": 3.9133960428154397,
+      "grad_norm": 0.27240729331970215,
+      "learning_rate": 5.955136914604989e-06,
+      "loss": 0.3527,
       "step": 108585
     },
     {
-      "epoch": 3.82,
-      "learning_rate": 6.963589797434475e-06,
-      "loss": 0.2527,
+      "epoch": 3.913576242476664,
+      "grad_norm": 0.2016393542289734,
+      "learning_rate": 5.953246622721137e-06,
+      "loss": 0.3709,
       "step": 108590
     },
     {
-      "epoch": 3.82,
-      "learning_rate": 6.9616172893860885e-06,
-      "loss": 0.2589,
+      "epoch": 3.9137564421378888,
+      "grad_norm": 0.24369855225086212,
+      "learning_rate": 5.951356590345611e-06,
+      "loss": 0.3883,
       "step": 108595
     },
     {
-      "epoch": 3.82,
-      "learning_rate": 6.959645015554217e-06,
-      "loss": 0.2867,
+      "epoch": 3.9139366417991135,
+      "grad_norm": 0.20250004529953003,
+      "learning_rate": 5.94946681750416e-06,
+      "loss": 0.3649,
       "step": 108600
     },
     {
-      "epoch": 3.82,
-      "learning_rate": 6.957672975964468e-06,
-      "loss": 0.2487,
+      "epoch": 3.914116841460338,
+      "grad_norm": 0.26689743995666504,
+      "learning_rate": 5.9475773042225455e-06,
+      "loss": 0.3699,
       "step": 108605
     },
     {
-      "epoch": 3.82,
-      "learning_rate": 6.955701170642445e-06,
-      "loss": 0.267,
+      "epoch": 3.9142970411215625,
+      "grad_norm": 0.24879872798919678,
+      "learning_rate": 5.945688050526496e-06,
+      "loss": 0.3921,
       "step": 108610
     },
     {
-      "epoch": 3.82,
-      "learning_rate": 6.953729599613743e-06,
-      "loss": 0.2538,
+      "epoch": 3.9144772407827872,
+      "grad_norm": 0.20798632502555847,
+      "learning_rate": 5.943799056441751e-06,
+      "loss": 0.3478,
       "step": 108615
     },
     {
-      "epoch": 3.82,
-      "learning_rate": 6.951758262903976e-06,
-      "loss": 0.2294,
+      "epoch": 3.914657440444012,
+      "grad_norm": 0.2264632135629654,
+      "learning_rate": 5.941910321994063e-06,
+      "loss": 0.3668,
       "step": 108620
     },
     {
-      "epoch": 3.82,
-      "learning_rate": 6.949787160538729e-06,
-      "loss": 0.2474,
+      "epoch": 3.9148376401052367,
+      "grad_norm": 0.26185837388038635,
+      "learning_rate": 5.940021847209157e-06,
+      "loss": 0.3667,
       "step": 108625
     },
     {
-      "epoch": 3.82,
-      "learning_rate": 6.947816292543605e-06,
-      "loss": 0.2506,
+      "epoch": 3.9150178397664614,
+      "grad_norm": 0.20407342910766602,
+      "learning_rate": 5.93813363211276e-06,
+      "loss": 0.3491,
       "step": 108630
     },
     {
-      "epoch": 3.82,
-      "learning_rate": 6.945845658944183e-06,
-      "loss": 0.2478,
+      "epoch": 3.9151980394276857,
+      "grad_norm": 0.21757426857948303,
+      "learning_rate": 5.936245676730603e-06,
+      "loss": 0.3534,
       "step": 108635
     },
     {
-      "epoch": 3.82,
-      "learning_rate": 6.943875259766064e-06,
-      "loss": 0.295,
+      "epoch": 3.9153782390889105,
+      "grad_norm": 0.23258854448795319,
+      "learning_rate": 5.934357981088403e-06,
+      "loss": 0.4074,
       "step": 108640
     },
     {
-      "epoch": 3.82,
-      "learning_rate": 6.941905095034826e-06,
-      "loss": 0.2573,
+      "epoch": 3.915558438750135,
+      "grad_norm": 0.31023046374320984,
+      "learning_rate": 5.932470545211891e-06,
+      "loss": 0.4156,
       "step": 108645
     },
     {
-      "epoch": 3.82,
-      "learning_rate": 6.9399351647760485e-06,
-      "loss": 0.253,
+      "epoch": 3.91573863841136,
+      "grad_norm": 0.2955925166606903,
+      "learning_rate": 5.930583369126774e-06,
+      "loss": 0.4184,
       "step": 108650
     },
     {
-      "epoch": 3.82,
-      "learning_rate": 6.937965469015304e-06,
-      "loss": 0.2792,
+      "epoch": 3.9159188380725842,
+      "grad_norm": 0.25111132860183716,
+      "learning_rate": 5.928696452858768e-06,
+      "loss": 0.3435,
       "step": 108655
     },
     {
-      "epoch": 3.82,
-      "learning_rate": 6.935996007778184e-06,
-      "loss": 0.251,
+      "epoch": 3.916099037733809,
+      "grad_norm": 0.25099194049835205,
+      "learning_rate": 5.92680979643358e-06,
+      "loss": 0.3505,
       "step": 108660
     },
     {
-      "epoch": 3.82,
-      "learning_rate": 6.93402678109025e-06,
-      "loss": 0.2515,
+      "epoch": 3.9162792373950337,
+      "grad_norm": 0.23228740692138672,
+      "learning_rate": 5.924923399876908e-06,
+      "loss": 0.3887,
       "step": 108665
     },
     {
-      "epoch": 3.82,
-      "learning_rate": 6.932057788977062e-06,
-      "loss": 0.264,
+      "epoch": 3.9164594370562584,
+      "grad_norm": 0.26973646879196167,
+      "learning_rate": 5.9230372632144705e-06,
+      "loss": 0.3786,
       "step": 108670
     },
     {
-      "epoch": 3.82,
-      "learning_rate": 6.930089031464207e-06,
-      "loss": 0.2479,
+      "epoch": 3.916639636717483,
+      "grad_norm": 0.2194148302078247,
+      "learning_rate": 5.921151386471962e-06,
+      "loss": 0.3936,
       "step": 108675
     },
     {
-      "epoch": 3.82,
-      "learning_rate": 6.928120508577229e-06,
-      "loss": 0.2568,
+      "epoch": 3.9168198363787075,
+      "grad_norm": 0.21613870561122894,
+      "learning_rate": 5.919265769675059e-06,
+      "loss": 0.3644,
       "step": 108680
     },
     {
-      "epoch": 3.82,
-      "learning_rate": 6.926152220341703e-06,
-      "loss": 0.2518,
+      "epoch": 3.917000036039932,
+      "grad_norm": 0.2264699786901474,
+      "learning_rate": 5.917380412849474e-06,
+      "loss": 0.3825,
       "step": 108685
     },
     {
-      "epoch": 3.82,
-      "learning_rate": 6.924184166783182e-06,
-      "loss": 0.2506,
+      "epoch": 3.917180235701157,
+      "grad_norm": 0.21265465021133423,
+      "learning_rate": 5.9154953160208776e-06,
+      "loss": 0.3764,
       "step": 108690
     },
     {
-      "epoch": 3.82,
-      "learning_rate": 6.922216347927208e-06,
-      "loss": 0.2643,
+      "epoch": 3.9173604353623817,
+      "grad_norm": 0.2905406653881073,
+      "learning_rate": 5.913610479214976e-06,
+      "loss": 0.3926,
       "step": 108695
     },
     {
-      "epoch": 3.82,
-      "learning_rate": 6.920248763799348e-06,
-      "loss": 0.2458,
+      "epoch": 3.917540635023606,
+      "grad_norm": 0.30452674627304077,
+      "learning_rate": 5.911725902457432e-06,
+      "loss": 0.3959,
       "step": 108700
     },
     {
-      "epoch": 3.82,
-      "learning_rate": 6.918281414425143e-06,
-      "loss": 0.2576,
+      "epoch": 3.9177208346848307,
+      "grad_norm": 0.20968550443649292,
+      "learning_rate": 5.909841585773917e-06,
+      "loss": 0.3671,
       "step": 108705
     },
     {
-      "epoch": 3.82,
-      "learning_rate": 6.916314299830129e-06,
-      "loss": 0.2654,
+      "epoch": 3.9179010343460554,
+      "grad_norm": 0.20306989550590515,
+      "learning_rate": 5.9079575291901254e-06,
+      "loss": 0.3766,
       "step": 108710
     },
     {
-      "epoch": 3.82,
-      "learning_rate": 6.914347420039865e-06,
-      "loss": 0.2416,
+      "epoch": 3.91808123400728,
+      "grad_norm": 0.20528307557106018,
+      "learning_rate": 5.906073732731712e-06,
+      "loss": 0.3609,
       "step": 108715
     },
     {
-      "epoch": 3.83,
-      "learning_rate": 6.912380775079882e-06,
-      "loss": 0.2236,
+      "epoch": 3.918261433668505,
+      "grad_norm": 0.32778772711753845,
+      "learning_rate": 5.90419019642435e-06,
+      "loss": 0.3861,
       "step": 108720
     },
     {
-      "epoch": 3.83,
-      "learning_rate": 6.9104143649757065e-06,
-      "loss": 0.2305,
+      "epoch": 3.9184416333297296,
+      "grad_norm": 0.25821900367736816,
+      "learning_rate": 5.9023069202936984e-06,
+      "loss": 0.3839,
       "step": 108725
     },
     {
-      "epoch": 3.83,
-      "learning_rate": 6.9084481897528835e-06,
-      "loss": 0.2546,
+      "epoch": 3.918621832990954,
+      "grad_norm": 0.2765897810459137,
+      "learning_rate": 5.90042390436541e-06,
+      "loss": 0.3741,
       "step": 108730
     },
     {
-      "epoch": 3.83,
-      "learning_rate": 6.906482249436933e-06,
-      "loss": 0.2442,
+      "epoch": 3.9188020326521786,
+      "grad_norm": 0.22229650616645813,
+      "learning_rate": 5.898541148665154e-06,
+      "loss": 0.3783,
       "step": 108735
     },
     {
-      "epoch": 3.83,
-      "learning_rate": 6.904516544053394e-06,
-      "loss": 0.2563,
+      "epoch": 3.9189822323134034,
+      "grad_norm": 0.32153648138046265,
+      "learning_rate": 5.896658653218578e-06,
+      "loss": 0.3865,
       "step": 108740
     },
     {
-      "epoch": 3.83,
-      "learning_rate": 6.902551073627783e-06,
-      "loss": 0.2573,
+      "epoch": 3.9191624319746277,
+      "grad_norm": 0.23668645322322845,
+      "learning_rate": 5.8947764180513316e-06,
+      "loss": 0.3962,
       "step": 108745
     },
     {
-      "epoch": 3.83,
-      "learning_rate": 6.9005858381856105e-06,
-      "loss": 0.2586,
+      "epoch": 3.9193426316358524,
+      "grad_norm": 0.26345136761665344,
+      "learning_rate": 5.892894443189056e-06,
+      "loss": 0.3455,
       "step": 108750
     },
     {
-      "epoch": 3.83,
-      "learning_rate": 6.898620837752412e-06,
-      "loss": 0.2463,
+      "epoch": 3.919522831297077,
+      "grad_norm": 0.23843061923980713,
+      "learning_rate": 5.891012728657394e-06,
+      "loss": 0.3981,
       "step": 108755
     },
     {
-      "epoch": 3.83,
-      "learning_rate": 6.896656072353691e-06,
-      "loss": 0.2651,
+      "epoch": 3.919703030958302,
+      "grad_norm": 0.25555703043937683,
+      "learning_rate": 5.889131274481977e-06,
+      "loss": 0.3639,
       "step": 108760
     },
     {
-      "epoch": 3.83,
-      "learning_rate": 6.894691542014961e-06,
-      "loss": 0.2895,
+      "epoch": 3.9198832306195266,
+      "grad_norm": 0.18954876065254211,
+      "learning_rate": 5.887250080688453e-06,
+      "loss": 0.3554,
       "step": 108765
     },
     {
-      "epoch": 3.83,
-      "learning_rate": 6.892727246761721e-06,
-      "loss": 0.2465,
+      "epoch": 3.9200634302807513,
+      "grad_norm": 0.2582615613937378,
+      "learning_rate": 5.885369147302447e-06,
+      "loss": 0.4017,
       "step": 108770
     },
     {
-      "epoch": 3.83,
-      "learning_rate": 6.890763186619495e-06,
-      "loss": 0.2639,
+      "epoch": 3.9202436299419756,
+      "grad_norm": 0.20174746215343475,
+      "learning_rate": 5.883488474349586e-06,
+      "loss": 0.381,
       "step": 108775
     },
     {
-      "epoch": 3.83,
-      "learning_rate": 6.888799361613771e-06,
-      "loss": 0.2673,
+      "epoch": 3.9204238296032003,
+      "grad_norm": 0.2733262777328491,
+      "learning_rate": 5.881608061855492e-06,
+      "loss": 0.39,
       "step": 108780
     },
     {
-      "epoch": 3.83,
-      "learning_rate": 6.8868357717700446e-06,
-      "loss": 0.2721,
+      "epoch": 3.920604029264425,
+      "grad_norm": 0.21888548135757446,
+      "learning_rate": 5.879727909845781e-06,
+      "loss": 0.3547,
       "step": 108785
     },
     {
-      "epoch": 3.83,
-      "learning_rate": 6.884872417113822e-06,
-      "loss": 0.2598,
+      "epoch": 3.9207842289256494,
+      "grad_norm": 0.26759272813796997,
+      "learning_rate": 5.8778480183460885e-06,
+      "loss": 0.4001,
       "step": 108790
     },
     {
-      "epoch": 3.83,
-      "learning_rate": 6.882909297670598e-06,
-      "loss": 0.2458,
+      "epoch": 3.920964428586874,
+      "grad_norm": 0.26541879773139954,
+      "learning_rate": 5.875968387382008e-06,
+      "loss": 0.342,
       "step": 108795
     },
     {
-      "epoch": 3.83,
-      "learning_rate": 6.880946413465858e-06,
-      "loss": 0.2658,
+      "epoch": 3.921144628248099,
+      "grad_norm": 0.2672159671783447,
+      "learning_rate": 5.874089016979151e-06,
+      "loss": 0.3564,
       "step": 108800
     },
     {
-      "epoch": 3.83,
-      "learning_rate": 6.878983764525085e-06,
-      "loss": 0.2494,
+      "epoch": 3.9213248279093236,
+      "grad_norm": 0.18863865733146667,
+      "learning_rate": 5.872209907163132e-06,
+      "loss": 0.3994,
       "step": 108805
     },
     {
-      "epoch": 3.83,
-      "learning_rate": 6.87702135087376e-06,
-      "loss": 0.256,
+      "epoch": 3.9215050275705483,
+      "grad_norm": 0.2630614638328552,
+      "learning_rate": 5.870331057959552e-06,
+      "loss": 0.3765,
       "step": 108810
     },
     {
-      "epoch": 3.83,
-      "learning_rate": 6.875059172537379e-06,
-      "loss": 0.2424,
+      "epoch": 3.921685227231773,
+      "grad_norm": 0.2330998033285141,
+      "learning_rate": 5.868452469394006e-06,
+      "loss": 0.3701,
       "step": 108815
     },
     {
-      "epoch": 3.83,
-      "learning_rate": 6.873097229541406e-06,
-      "loss": 0.2528,
+      "epoch": 3.9218654268929973,
+      "grad_norm": 0.20276689529418945,
+      "learning_rate": 5.86657414149209e-06,
+      "loss": 0.3582,
       "step": 108820
     },
     {
-      "epoch": 3.83,
-      "learning_rate": 6.871135521911315e-06,
-      "loss": 0.2658,
+      "epoch": 3.922045626554222,
+      "grad_norm": 0.2714731991291046,
+      "learning_rate": 5.864696074279394e-06,
+      "loss": 0.3724,
       "step": 108825
     },
     {
-      "epoch": 3.83,
-      "learning_rate": 6.869174049672589e-06,
-      "loss": 0.2566,
+      "epoch": 3.922225826215447,
+      "grad_norm": 0.2654849886894226,
+      "learning_rate": 5.862818267781514e-06,
+      "loss": 0.3798,
       "step": 108830
     },
     {
-      "epoch": 3.83,
-      "learning_rate": 6.8672128128506874e-06,
-      "loss": 0.2662,
+      "epoch": 3.922406025876671,
+      "grad_norm": 0.18133863806724548,
+      "learning_rate": 5.860940722024027e-06,
+      "loss": 0.4061,
       "step": 108835
     },
     {
-      "epoch": 3.83,
-      "learning_rate": 6.865251811471071e-06,
-      "loss": 0.2603,
+      "epoch": 3.922586225537896,
+      "grad_norm": 0.2509717345237732,
+      "learning_rate": 5.8590634370325224e-06,
+      "loss": 0.4112,
       "step": 108840
     },
     {
-      "epoch": 3.83,
-      "learning_rate": 6.863291045559209e-06,
-      "loss": 0.2534,
+      "epoch": 3.9227664251991206,
+      "grad_norm": 0.23945482075214386,
+      "learning_rate": 5.857186412832569e-06,
+      "loss": 0.3412,
       "step": 108845
     },
     {
-      "epoch": 3.83,
-      "learning_rate": 6.861330515140568e-06,
-      "loss": 0.257,
+      "epoch": 3.9229466248603453,
+      "grad_norm": 0.30383405089378357,
+      "learning_rate": 5.855309649449739e-06,
+      "loss": 0.4108,
       "step": 108850
     },
     {
-      "epoch": 3.83,
-      "learning_rate": 6.859370220240594e-06,
-      "loss": 0.2316,
+      "epoch": 3.92312682452157,
+      "grad_norm": 0.20874951779842377,
+      "learning_rate": 5.853433146909618e-06,
+      "loss": 0.4059,
       "step": 108855
     },
     {
-      "epoch": 3.83,
-      "learning_rate": 6.857410160884745e-06,
-      "loss": 0.2485,
+      "epoch": 3.9233070241827948,
+      "grad_norm": 0.3230457901954651,
+      "learning_rate": 5.851556905237768e-06,
+      "loss": 0.4288,
       "step": 108860
     },
     {
-      "epoch": 3.83,
-      "learning_rate": 6.8554503370984595e-06,
-      "loss": 0.2623,
+      "epoch": 3.923487223844019,
+      "grad_norm": 0.28123703598976135,
+      "learning_rate": 5.849680924459733e-06,
+      "loss": 0.4122,
       "step": 108865
     },
     {
-      "epoch": 3.83,
-      "learning_rate": 6.853490748907204e-06,
-      "loss": 0.264,
+      "epoch": 3.923667423505244,
+      "grad_norm": 0.2393689751625061,
+      "learning_rate": 5.847805204601095e-06,
+      "loss": 0.3825,
       "step": 108870
     },
     {
-      "epoch": 3.83,
-      "learning_rate": 6.8515313963364105e-06,
-      "loss": 0.2575,
+      "epoch": 3.9238476231664685,
+      "grad_norm": 0.23288077116012573,
+      "learning_rate": 5.845929745687398e-06,
+      "loss": 0.3605,
       "step": 108875
     },
     {
-      "epoch": 3.83,
-      "learning_rate": 6.84957227941152e-06,
-      "loss": 0.269,
+      "epoch": 3.9240278228276932,
+      "grad_norm": 0.23712435364723206,
+      "learning_rate": 5.844054547744204e-06,
+      "loss": 0.3832,
       "step": 108880
     },
     {
-      "epoch": 3.83,
-      "learning_rate": 6.847613398157965e-06,
-      "loss": 0.2572,
+      "epoch": 3.9242080224889175,
+      "grad_norm": 0.23456718027591705,
+      "learning_rate": 5.842179610797066e-06,
+      "loss": 0.3643,
       "step": 108885
     },
     {
-      "epoch": 3.83,
-      "learning_rate": 6.845654752601197e-06,
-      "loss": 0.2508,
+      "epoch": 3.9243882221501423,
+      "grad_norm": 0.2397494912147522,
+      "learning_rate": 5.84030493487151e-06,
+      "loss": 0.3329,
       "step": 108890
     },
     {
-      "epoch": 3.83,
-      "learning_rate": 6.843696342766628e-06,
-      "loss": 0.247,
+      "epoch": 3.924568421811367,
+      "grad_norm": 0.22742457687854767,
+      "learning_rate": 5.838430519993096e-06,
+      "loss": 0.3847,
       "step": 108895
     },
     {
-      "epoch": 3.83,
-      "learning_rate": 6.841738168679704e-06,
-      "loss": 0.2632,
+      "epoch": 3.9247486214725917,
+      "grad_norm": 0.21573381125926971,
+      "learning_rate": 5.836556366187354e-06,
+      "loss": 0.3471,
       "step": 108900
     },
     {
-      "epoch": 3.83,
-      "learning_rate": 6.839780230365836e-06,
-      "loss": 0.2564,
+      "epoch": 3.9249288211338165,
+      "grad_norm": 0.2524566650390625,
+      "learning_rate": 5.834682473479824e-06,
+      "loss": 0.35,
       "step": 108905
     },
     {
-      "epoch": 3.83,
-      "learning_rate": 6.837822527850463e-06,
-      "loss": 0.2442,
+      "epoch": 3.9251090207950408,
+      "grad_norm": 0.22824294865131378,
+      "learning_rate": 5.832808841896034e-06,
+      "loss": 0.3783,
       "step": 108910
     },
     {
-      "epoch": 3.83,
-      "learning_rate": 6.835865061158994e-06,
-      "loss": 0.2493,
+      "epoch": 3.9252892204562655,
+      "grad_norm": 0.2146436721086502,
+      "learning_rate": 5.830935471461504e-06,
+      "loss": 0.3896,
       "step": 108915
     },
     {
-      "epoch": 3.83,
-      "learning_rate": 6.833907830316846e-06,
-      "loss": 0.2528,
+      "epoch": 3.9254694201174902,
+      "grad_norm": 0.2600581645965576,
+      "learning_rate": 5.829062362201776e-06,
+      "loss": 0.3775,
       "step": 108920
     },
     {
-      "epoch": 3.83,
-      "learning_rate": 6.831950835349427e-06,
-      "loss": 0.2608,
+      "epoch": 3.925649619778715,
+      "grad_norm": 0.3017957806587219,
+      "learning_rate": 5.827189514142361e-06,
+      "loss": 0.4071,
       "step": 108925
     },
     {
-      "epoch": 3.83,
-      "learning_rate": 6.829994076282159e-06,
-      "loss": 0.2436,
+      "epoch": 3.9258298194399393,
+      "grad_norm": 0.2875199615955353,
+      "learning_rate": 5.8253169273087774e-06,
+      "loss": 0.4051,
       "step": 108930
     },
     {
-      "epoch": 3.83,
-      "learning_rate": 6.828037553140443e-06,
-      "loss": 0.2438,
+      "epoch": 3.926010019101164,
+      "grad_norm": 0.22084161639213562,
+      "learning_rate": 5.823444601726538e-06,
+      "loss": 0.3576,
       "step": 108935
     },
     {
-      "epoch": 3.83,
-      "learning_rate": 6.826081265949677e-06,
-      "loss": 0.2773,
+      "epoch": 3.9261902187623887,
+      "grad_norm": 0.19947287440299988,
+      "learning_rate": 5.8215725374211525e-06,
+      "loss": 0.3679,
       "step": 108940
     },
     {
-      "epoch": 3.83,
-      "learning_rate": 6.824125214735274e-06,
-      "loss": 0.2594,
+      "epoch": 3.9263704184236135,
+      "grad_norm": 0.2296856790781021,
+      "learning_rate": 5.819700734418121e-06,
+      "loss": 0.4062,
       "step": 108945
     },
     {
-      "epoch": 3.83,
-      "learning_rate": 6.822169399522624e-06,
-      "loss": 0.2415,
+      "epoch": 3.926550618084838,
+      "grad_norm": 0.258223295211792,
+      "learning_rate": 5.817829192742963e-06,
+      "loss": 0.3719,
       "step": 108950
     },
     {
-      "epoch": 3.83,
-      "learning_rate": 6.820213820337129e-06,
-      "loss": 0.272,
+      "epoch": 3.9267308177460625,
+      "grad_norm": 0.27443790435791016,
+      "learning_rate": 5.8159579124211665e-06,
+      "loss": 0.3654,
       "step": 108955
     },
     {
-      "epoch": 3.83,
-      "learning_rate": 6.81825847720417e-06,
-      "loss": 0.2607,
+      "epoch": 3.926911017407287,
+      "grad_norm": 0.23533061146736145,
+      "learning_rate": 5.8140868934782295e-06,
+      "loss": 0.3518,
       "step": 108960
     },
     {
-      "epoch": 3.83,
-      "learning_rate": 6.816303370149149e-06,
-      "loss": 0.2469,
+      "epoch": 3.927091217068512,
+      "grad_norm": 0.21623454988002777,
+      "learning_rate": 5.812216135939644e-06,
+      "loss": 0.3867,
       "step": 108965
     },
     {
-      "epoch": 3.83,
-      "learning_rate": 6.814348499197445e-06,
-      "loss": 0.2684,
+      "epoch": 3.9272714167297367,
+      "grad_norm": 0.2719837725162506,
+      "learning_rate": 5.810345639830891e-06,
+      "loss": 0.3659,
       "step": 108970
     },
     {
-      "epoch": 3.83,
-      "learning_rate": 6.812393864374442e-06,
-      "loss": 0.2724,
+      "epoch": 3.927451616390961,
+      "grad_norm": 0.2205159217119217,
+      "learning_rate": 5.80847540517748e-06,
+      "loss": 0.3625,
       "step": 108975
     },
     {
-      "epoch": 3.83,
-      "learning_rate": 6.810439465705512e-06,
-      "loss": 0.2911,
+      "epoch": 3.9276318160521857,
+      "grad_norm": 0.23744551837444305,
+      "learning_rate": 5.806605432004866e-06,
+      "loss": 0.3772,
       "step": 108980
     },
     {
-      "epoch": 3.83,
-      "learning_rate": 6.808485303216044e-06,
-      "loss": 0.2469,
+      "epoch": 3.9278120157134104,
+      "grad_norm": 0.235315203666687,
+      "learning_rate": 5.804735720338531e-06,
+      "loss": 0.3935,
       "step": 108985
     },
     {
-      "epoch": 3.83,
-      "learning_rate": 6.806531376931408e-06,
-      "loss": 0.2637,
+      "epoch": 3.927992215374635,
+      "grad_norm": 0.26385051012039185,
+      "learning_rate": 5.802866270203964e-06,
+      "loss": 0.3864,
       "step": 108990
     },
     {
-      "epoch": 3.83,
-      "learning_rate": 6.80457768687697e-06,
-      "loss": 0.2581,
+      "epoch": 3.92817241503586,
+      "grad_norm": 0.234380304813385,
+      "learning_rate": 5.800997081626619e-06,
+      "loss": 0.3736,
       "step": 108995
     },
     {
-      "epoch": 3.83,
-      "learning_rate": 6.802624233078092e-06,
-      "loss": 0.2617,
+      "epoch": 3.9283526146970846,
+      "grad_norm": 0.21200546622276306,
+      "learning_rate": 5.799128154631986e-06,
+      "loss": 0.3632,
       "step": 109000
     },
     {
-      "epoch": 3.83,
-      "eval_loss": 0.2511043846607208,
-      "eval_runtime": 10.5525,
-      "eval_samples_per_second": 9.476,
-      "eval_steps_per_second": 9.476,
+      "epoch": 3.9283526146970846,
+      "eval_loss": 0.42800185084342957,
+      "eval_runtime": 3.5264,
+      "eval_samples_per_second": 28.357,
+      "eval_steps_per_second": 7.089,
       "step": 109000
     },
     {
-      "epoch": 3.84,
-      "learning_rate": 6.800671015560145e-06,
-      "loss": 0.2452,
+      "epoch": 3.928532814358309,
+      "grad_norm": 0.24245581030845642,
+      "learning_rate": 5.797259489245508e-06,
+      "loss": 0.3871,
       "step": 109005
     },
     {
-      "epoch": 3.84,
-      "learning_rate": 6.7987180343485005e-06,
-      "loss": 0.25,
+      "epoch": 3.9287130140195337,
+      "grad_norm": 0.29158252477645874,
+      "learning_rate": 5.795391085492644e-06,
+      "loss": 0.4042,
       "step": 109010
     },
     {
-      "epoch": 3.84,
-      "learning_rate": 6.796765289468507e-06,
-      "loss": 0.26,
+      "epoch": 3.9288932136807584,
+      "grad_norm": 0.2737109959125519,
+      "learning_rate": 5.793522943398863e-06,
+      "loss": 0.4132,
       "step": 109015
     },
     {
-      "epoch": 3.84,
-      "learning_rate": 6.7948127809455115e-06,
-      "loss": 0.2422,
+      "epoch": 3.9290734133419827,
+      "grad_norm": 0.255136638879776,
+      "learning_rate": 5.791655062989618e-06,
+      "loss": 0.3969,
       "step": 109020
     },
     {
-      "epoch": 3.84,
-      "learning_rate": 6.792860508804882e-06,
-      "loss": 0.2714,
+      "epoch": 3.9292536130032074,
+      "grad_norm": 0.2431129515171051,
+      "learning_rate": 5.789787444290351e-06,
+      "loss": 0.3924,
       "step": 109025
     },
     {
-      "epoch": 3.84,
-      "learning_rate": 6.790908473071961e-06,
-      "loss": 0.3057,
+      "epoch": 3.929433812664432,
+      "grad_norm": 0.27715519070625305,
+      "learning_rate": 5.787920087326512e-06,
+      "loss": 0.3818,
       "step": 109030
     },
     {
-      "epoch": 3.84,
-      "learning_rate": 6.7889566737720896e-06,
-      "loss": 0.2673,
+      "epoch": 3.929614012325657,
+      "grad_norm": 0.24491608142852783,
+      "learning_rate": 5.786052992123533e-06,
+      "loss": 0.3742,
       "step": 109035
     },
     {
-      "epoch": 3.84,
-      "learning_rate": 6.787005110930611e-06,
-      "loss": 0.2768,
+      "epoch": 3.9297942119868816,
+      "grad_norm": 0.21611642837524414,
+      "learning_rate": 5.784186158706872e-06,
+      "loss": 0.3943,
       "step": 109040
     },
     {
-      "epoch": 3.84,
-      "learning_rate": 6.785053784572873e-06,
-      "loss": 0.2606,
+      "epoch": 3.9299744116481063,
+      "grad_norm": 0.2787199914455414,
+      "learning_rate": 5.782319587101953e-06,
+      "loss": 0.3888,
       "step": 109045
     },
     {
-      "epoch": 3.84,
-      "learning_rate": 6.783102694724211e-06,
-      "loss": 0.2553,
+      "epoch": 3.9301546113093306,
+      "grad_norm": 0.26569807529449463,
+      "learning_rate": 5.780453277334208e-06,
+      "loss": 0.3812,
       "step": 109050
     },
     {
-      "epoch": 3.84,
-      "learning_rate": 6.781151841409944e-06,
-      "loss": 0.2538,
+      "epoch": 3.9303348109705554,
+      "grad_norm": 0.25157374143600464,
+      "learning_rate": 5.778587229429069e-06,
+      "loss": 0.4124,
       "step": 109055
     },
     {
-      "epoch": 3.84,
-      "learning_rate": 6.7792012246554235e-06,
-      "loss": 0.2512,
+      "epoch": 3.93051501063178,
+      "grad_norm": 0.21944718062877655,
+      "learning_rate": 5.776721443411956e-06,
+      "loss": 0.3392,
       "step": 109060
     },
     {
-      "epoch": 3.84,
-      "learning_rate": 6.7772508444859575e-06,
-      "loss": 0.2475,
+      "epoch": 3.9306952102930044,
+      "grad_norm": 0.25366106629371643,
+      "learning_rate": 5.774855919308284e-06,
+      "loss": 0.4012,
       "step": 109065
     },
     {
-      "epoch": 3.84,
-      "learning_rate": 6.77530070092689e-06,
-      "loss": 0.2504,
+      "epoch": 3.930875409954229,
+      "grad_norm": 0.24010884761810303,
+      "learning_rate": 5.772990657143492e-06,
+      "loss": 0.3748,
       "step": 109070
     },
     {
-      "epoch": 3.84,
-      "learning_rate": 6.773350794003522e-06,
-      "loss": 0.2552,
+      "epoch": 3.931055609615454,
+      "grad_norm": 0.2395809143781662,
+      "learning_rate": 5.771125656942966e-06,
+      "loss": 0.3855,
       "step": 109075
     },
     {
-      "epoch": 3.84,
-      "learning_rate": 6.771401123741192e-06,
-      "loss": 0.2618,
+      "epoch": 3.9312358092766786,
+      "grad_norm": 0.22961130738258362,
+      "learning_rate": 5.769260918732139e-06,
+      "loss": 0.3671,
       "step": 109080
     },
     {
-      "epoch": 3.84,
-      "learning_rate": 6.7694516901652045e-06,
-      "loss": 0.2591,
+      "epoch": 3.9314160089379033,
+      "grad_norm": 0.21910971403121948,
+      "learning_rate": 5.767396442536404e-06,
+      "loss": 0.3404,
       "step": 109085
     },
     {
-      "epoch": 3.84,
-      "learning_rate": 6.767502493300873e-06,
-      "loss": 0.2501,
+      "epoch": 3.931596208599128,
+      "grad_norm": 0.2396189421415329,
+      "learning_rate": 5.765532228381174e-06,
+      "loss": 0.404,
       "step": 109090
     },
     {
-      "epoch": 3.84,
-      "learning_rate": 6.765553533173497e-06,
-      "loss": 0.2763,
+      "epoch": 3.9317764082603524,
+      "grad_norm": 0.25804564356803894,
+      "learning_rate": 5.763668276291842e-06,
+      "loss": 0.3575,
       "step": 109095
     },
     {
-      "epoch": 3.84,
-      "learning_rate": 6.7636048098084e-06,
-      "loss": 0.2461,
+      "epoch": 3.931956607921577,
+      "grad_norm": 0.2125440388917923,
+      "learning_rate": 5.761804586293798e-06,
+      "loss": 0.3658,
       "step": 109100
     },
     {
-      "epoch": 3.84,
-      "learning_rate": 6.761656323230875e-06,
-      "loss": 0.2704,
+      "epoch": 3.932136807582802,
+      "grad_norm": 0.20259714126586914,
+      "learning_rate": 5.759941158412449e-06,
+      "loss": 0.3936,
       "step": 109105
     },
     {
-      "epoch": 3.84,
-      "learning_rate": 6.759708073466217e-06,
-      "loss": 0.2636,
+      "epoch": 3.932317007244026,
+      "grad_norm": 0.26884329319000244,
+      "learning_rate": 5.758077992673175e-06,
+      "loss": 0.3642,
       "step": 109110
     },
     {
-      "epoch": 3.84,
-      "learning_rate": 6.7577600605397365e-06,
-      "loss": 0.2704,
+      "epoch": 3.932497206905251,
+      "grad_norm": 0.22424453496932983,
+      "learning_rate": 5.756215089101366e-06,
+      "loss": 0.404,
       "step": 109115
     },
     {
-      "epoch": 3.84,
-      "learning_rate": 6.755812284476712e-06,
-      "loss": 0.2373,
+      "epoch": 3.9326774065664756,
+      "grad_norm": 0.3316144645214081,
+      "learning_rate": 5.7543524477224e-06,
+      "loss": 0.3845,
       "step": 109120
     },
     {
-      "epoch": 3.84,
-      "learning_rate": 6.753864745302449e-06,
-      "loss": 0.2521,
+      "epoch": 3.9328576062277003,
+      "grad_norm": 0.2477588653564453,
+      "learning_rate": 5.752490068561653e-06,
+      "loss": 0.3847,
       "step": 109125
     },
     {
-      "epoch": 3.84,
-      "learning_rate": 6.7519174430422275e-06,
-      "loss": 0.2599,
+      "epoch": 3.933037805888925,
+      "grad_norm": 0.21736137568950653,
+      "learning_rate": 5.750627951644496e-06,
+      "loss": 0.3864,
       "step": 109130
     },
     {
-      "epoch": 3.84,
-      "learning_rate": 6.749970377721323e-06,
-      "loss": 0.2376,
+      "epoch": 3.9332180055501498,
+      "grad_norm": 0.253451406955719,
+      "learning_rate": 5.748766096996316e-06,
+      "loss": 0.3711,
       "step": 109135
     },
     {
-      "epoch": 3.84,
-      "learning_rate": 6.748023549365034e-06,
-      "loss": 0.2462,
+      "epoch": 3.933398205211374,
+      "grad_norm": 0.2794763147830963,
+      "learning_rate": 5.74690450464247e-06,
+      "loss": 0.3803,
       "step": 109140
     },
     {
-      "epoch": 3.84,
-      "learning_rate": 6.746076957998631e-06,
-      "loss": 0.2321,
+      "epoch": 3.933578404872599,
+      "grad_norm": 0.24337390065193176,
+      "learning_rate": 5.745043174608322e-06,
+      "loss": 0.3818,
       "step": 109145
     },
     {
-      "epoch": 3.84,
-      "learning_rate": 6.744130603647392e-06,
-      "loss": 0.2397,
+      "epoch": 3.9337586045338235,
+      "grad_norm": 0.2820989489555359,
+      "learning_rate": 5.7431821069192345e-06,
+      "loss": 0.4063,
       "step": 109150
     },
     {
-      "epoch": 3.84,
-      "learning_rate": 6.742184486336576e-06,
-      "loss": 0.2366,
+      "epoch": 3.9339388041950483,
+      "grad_norm": 0.27026164531707764,
+      "learning_rate": 5.741321301600553e-06,
+      "loss": 0.3897,
       "step": 109155
     },
     {
-      "epoch": 3.84,
-      "learning_rate": 6.740238606091468e-06,
-      "loss": 0.2545,
+      "epoch": 3.9341190038562726,
+      "grad_norm": 0.24505634605884552,
+      "learning_rate": 5.7394607586776475e-06,
+      "loss": 0.3734,
       "step": 109160
     },
     {
-      "epoch": 3.84,
-      "learning_rate": 6.738292962937323e-06,
-      "loss": 0.2597,
+      "epoch": 3.9342992035174973,
+      "grad_norm": 0.2610364258289337,
+      "learning_rate": 5.73760047817587e-06,
+      "loss": 0.3896,
       "step": 109165
     },
     {
-      "epoch": 3.84,
-      "learning_rate": 6.736347556899414e-06,
-      "loss": 0.2432,
+      "epoch": 3.934479403178722,
+      "grad_norm": 0.2157258540391922,
+      "learning_rate": 5.735740460120539e-06,
+      "loss": 0.3791,
       "step": 109170
     },
     {
-      "epoch": 3.84,
-      "learning_rate": 6.734402388002992e-06,
-      "loss": 0.2496,
+      "epoch": 3.9346596028399468,
+      "grad_norm": 0.2157232165336609,
+      "learning_rate": 5.733880704537023e-06,
+      "loss": 0.3816,
       "step": 109175
     },
     {
-      "epoch": 3.84,
-      "learning_rate": 6.7324574562733225e-06,
-      "loss": 0.2766,
+      "epoch": 3.9348398025011715,
+      "grad_norm": 0.22246497869491577,
+      "learning_rate": 5.732021211450647e-06,
+      "loss": 0.3887,
       "step": 109180
     },
     {
-      "epoch": 3.84,
-      "learning_rate": 6.730512761735655e-06,
-      "loss": 0.242,
+      "epoch": 3.935020002162396,
+      "grad_norm": 0.1965298354625702,
+      "learning_rate": 5.730161980886764e-06,
+      "loss": 0.373,
       "step": 109185
     },
     {
-      "epoch": 3.84,
-      "learning_rate": 6.728568304415231e-06,
-      "loss": 0.262,
+      "epoch": 3.9352002018236205,
+      "grad_norm": 0.2733939588069916,
+      "learning_rate": 5.7283030128706865e-06,
+      "loss": 0.3991,
       "step": 109190
     },
     {
-      "epoch": 3.84,
-      "learning_rate": 6.726624084337313e-06,
-      "loss": 0.2766,
+      "epoch": 3.9353804014848452,
+      "grad_norm": 0.22759440541267395,
+      "learning_rate": 5.726444307427742e-06,
+      "loss": 0.3825,
       "step": 109195
     },
     {
-      "epoch": 3.84,
-      "learning_rate": 6.724680101527139e-06,
-      "loss": 0.253,
+      "epoch": 3.93556060114607,
+      "grad_norm": 0.24440236389636993,
+      "learning_rate": 5.724585864583271e-06,
+      "loss": 0.3675,
       "step": 109200
     },
     {
-      "epoch": 3.84,
-      "learning_rate": 6.722736356009951e-06,
-      "loss": 0.2514,
+      "epoch": 3.9357408008072943,
+      "grad_norm": 0.24098166823387146,
+      "learning_rate": 5.722727684362583e-06,
+      "loss": 0.3957,
       "step": 109205
     },
     {
-      "epoch": 3.84,
-      "learning_rate": 6.720792847810977e-06,
-      "loss": 0.2806,
+      "epoch": 3.935921000468519,
+      "grad_norm": 0.1858464926481247,
+      "learning_rate": 5.720869766790999e-06,
+      "loss": 0.3768,
       "step": 109210
     },
     {
-      "epoch": 3.84,
-      "learning_rate": 6.718849576955471e-06,
-      "loss": 0.2738,
+      "epoch": 3.9361012001297437,
+      "grad_norm": 0.2523440420627594,
+      "learning_rate": 5.719012111893832e-06,
+      "loss": 0.3604,
       "step": 109215
     },
     {
-      "epoch": 3.84,
-      "learning_rate": 6.716906543468657e-06,
-      "loss": 0.2753,
+      "epoch": 3.9362813997909685,
+      "grad_norm": 0.3191049098968506,
+      "learning_rate": 5.7171547196963854e-06,
+      "loss": 0.3716,
       "step": 109220
     },
     {
-      "epoch": 3.84,
-      "learning_rate": 6.714963747375752e-06,
-      "loss": 0.2665,
+      "epoch": 3.936461599452193,
+      "grad_norm": 0.22533096373081207,
+      "learning_rate": 5.715297590223981e-06,
+      "loss": 0.3648,
       "step": 109225
     },
     {
-      "epoch": 3.84,
-      "learning_rate": 6.713021188701993e-06,
-      "loss": 0.2478,
+      "epoch": 3.936641799113418,
+      "grad_norm": 0.23449839651584625,
+      "learning_rate": 5.713440723501912e-06,
+      "loss": 0.4066,
       "step": 109230
     },
     {
-      "epoch": 3.84,
-      "learning_rate": 6.711078867472612e-06,
-      "loss": 0.2545,
+      "epoch": 3.9368219987746422,
+      "grad_norm": 0.2854231894016266,
+      "learning_rate": 5.711584119555477e-06,
+      "loss": 0.3489,
       "step": 109235
     },
     {
-      "epoch": 3.84,
-      "learning_rate": 6.709136783712816e-06,
-      "loss": 0.2623,
+      "epoch": 3.937002198435867,
+      "grad_norm": 0.21907241642475128,
+      "learning_rate": 5.709727778409976e-06,
+      "loss": 0.378,
       "step": 109240
     },
     {
-      "epoch": 3.84,
-      "learning_rate": 6.707194937447825e-06,
-      "loss": 0.2383,
+      "epoch": 3.9371823980970917,
+      "grad_norm": 0.2457626909017563,
+      "learning_rate": 5.7078717000907e-06,
+      "loss": 0.3533,
       "step": 109245
     },
     {
-      "epoch": 3.84,
-      "learning_rate": 6.705253328702846e-06,
-      "loss": 0.2522,
+      "epoch": 3.937362597758316,
+      "grad_norm": 0.24565333127975464,
+      "learning_rate": 5.706015884622928e-06,
+      "loss": 0.3805,
       "step": 109250
     },
     {
-      "epoch": 3.84,
-      "learning_rate": 6.703311957503103e-06,
-      "loss": 0.2479,
+      "epoch": 3.9375427974195407,
+      "grad_norm": 0.23259061574935913,
+      "learning_rate": 5.704160332031969e-06,
+      "loss": 0.373,
       "step": 109255
     },
     {
-      "epoch": 3.84,
-      "learning_rate": 6.7013708238737954e-06,
-      "loss": 0.2549,
+      "epoch": 3.9377229970807655,
+      "grad_norm": 0.2456471472978592,
+      "learning_rate": 5.702305042343075e-06,
+      "loss": 0.357,
       "step": 109260
     },
     {
-      "epoch": 3.84,
-      "learning_rate": 6.699429927840131e-06,
-      "loss": 0.254,
+      "epoch": 3.93790319674199,
+      "grad_norm": 0.27309781312942505,
+      "learning_rate": 5.700450015581549e-06,
+      "loss": 0.3559,
       "step": 109265
     },
     {
-      "epoch": 3.84,
-      "learning_rate": 6.697489269427298e-06,
-      "loss": 0.2509,
+      "epoch": 3.938083396403215,
+      "grad_norm": 0.2628590166568756,
+      "learning_rate": 5.698595251772651e-06,
+      "loss": 0.3645,
       "step": 109270
     },
     {
-      "epoch": 3.84,
-      "learning_rate": 6.695548848660515e-06,
-      "loss": 0.249,
+      "epoch": 3.9382635960644397,
+      "grad_norm": 0.19301342964172363,
+      "learning_rate": 5.696740750941651e-06,
+      "loss": 0.3669,
       "step": 109275
     },
     {
-      "epoch": 3.84,
-      "learning_rate": 6.693608665564957e-06,
-      "loss": 0.2779,
+      "epoch": 3.938443795725664,
+      "grad_norm": 0.2241153120994568,
+      "learning_rate": 5.6948865131138344e-06,
+      "loss": 0.3542,
       "step": 109280
     },
     {
-      "epoch": 3.84,
-      "learning_rate": 6.691668720165836e-06,
-      "loss": 0.2439,
+      "epoch": 3.9386239953868887,
+      "grad_norm": 0.21480773389339447,
+      "learning_rate": 5.6930325383144386e-06,
+      "loss": 0.3573,
       "step": 109285
     },
     {
-      "epoch": 3.85,
-      "learning_rate": 6.6897290124883216e-06,
-      "loss": 0.2579,
+      "epoch": 3.9388041950481134,
+      "grad_norm": 0.27999842166900635,
+      "learning_rate": 5.6911788265687435e-06,
+      "loss": 0.4112,
       "step": 109290
     },
     {
-      "epoch": 3.85,
-      "learning_rate": 6.687789542557618e-06,
-      "loss": 0.2478,
+      "epoch": 3.9389843947093377,
+      "grad_norm": 0.2639464735984802,
+      "learning_rate": 5.689325377901997e-06,
+      "loss": 0.3861,
       "step": 109295
     },
     {
-      "epoch": 3.85,
-      "learning_rate": 6.6858503103989e-06,
-      "loss": 0.2513,
+      "epoch": 3.9391645943705624,
+      "grad_norm": 0.21820655465126038,
+      "learning_rate": 5.6874721923394545e-06,
+      "loss": 0.3692,
       "step": 109300
     },
     {
-      "epoch": 3.85,
-      "learning_rate": 6.683911316037336e-06,
-      "loss": 0.2413,
+      "epoch": 3.939344794031787,
+      "grad_norm": 0.2508092522621155,
+      "learning_rate": 5.685619269906364e-06,
+      "loss": 0.3893,
       "step": 109305
     },
     {
-      "epoch": 3.85,
-      "learning_rate": 6.681972559498123e-06,
-      "loss": 0.2327,
+      "epoch": 3.939524993693012,
+      "grad_norm": 0.22443386912345886,
+      "learning_rate": 5.683766610627972e-06,
+      "loss": 0.3788,
       "step": 109310
     },
     {
-      "epoch": 3.85,
-      "learning_rate": 6.680034040806421e-06,
-      "loss": 0.256,
+      "epoch": 3.9397051933542366,
+      "grad_norm": 0.23551128804683685,
+      "learning_rate": 5.681914214529513e-06,
+      "loss": 0.3793,
       "step": 109315
     },
     {
-      "epoch": 3.85,
-      "learning_rate": 6.6780957599874075e-06,
-      "loss": 0.2294,
+      "epoch": 3.9398853930154614,
+      "grad_norm": 0.2493138313293457,
+      "learning_rate": 5.6800620816362395e-06,
+      "loss": 0.3987,
       "step": 109320
     },
     {
-      "epoch": 3.85,
-      "learning_rate": 6.676157717066237e-06,
-      "loss": 0.2734,
+      "epoch": 3.9400655926766857,
+      "grad_norm": 0.27576449513435364,
+      "learning_rate": 5.678210211973378e-06,
+      "loss": 0.3981,
       "step": 109325
     },
     {
-      "epoch": 3.85,
-      "learning_rate": 6.6742199120680885e-06,
-      "loss": 0.2602,
+      "epoch": 3.9402457923379104,
+      "grad_norm": 0.2820398509502411,
+      "learning_rate": 5.676358605566165e-06,
+      "loss": 0.346,
       "step": 109330
     },
     {
-      "epoch": 3.85,
-      "learning_rate": 6.672282345018111e-06,
-      "loss": 0.2599,
+      "epoch": 3.940425991999135,
+      "grad_norm": 0.21052910387516022,
+      "learning_rate": 5.674507262439821e-06,
+      "loss": 0.3848,
       "step": 109335
     },
     {
-      "epoch": 3.85,
-      "learning_rate": 6.670345015941476e-06,
-      "loss": 0.2409,
+      "epoch": 3.9406061916603594,
+      "grad_norm": 0.2435581535100937,
+      "learning_rate": 5.67265618261957e-06,
+      "loss": 0.3897,
       "step": 109340
     },
     {
-      "epoch": 3.85,
-      "learning_rate": 6.668407924863324e-06,
-      "loss": 0.2501,
+      "epoch": 3.940786391321584,
+      "grad_norm": 0.23726899921894073,
+      "learning_rate": 5.670805366130644e-06,
+      "loss": 0.3792,
       "step": 109345
     },
     {
-      "epoch": 3.85,
-      "learning_rate": 6.666471071808822e-06,
-      "loss": 0.2568,
+      "epoch": 3.940966590982809,
+      "grad_norm": 0.26930510997772217,
+      "learning_rate": 5.6689548129982565e-06,
+      "loss": 0.3823,
       "step": 109350
     },
     {
-      "epoch": 3.85,
-      "learning_rate": 6.664534456803109e-06,
-      "loss": 0.2652,
+      "epoch": 3.9411467906440336,
+      "grad_norm": 0.2104661613702774,
+      "learning_rate": 5.667104523247605e-06,
+      "loss": 0.4122,
       "step": 109355
     },
     {
-      "epoch": 3.85,
-      "learning_rate": 6.662598079871335e-06,
-      "loss": 0.2532,
+      "epoch": 3.9413269903052583,
+      "grad_norm": 0.2654925286769867,
+      "learning_rate": 5.665254496903919e-06,
+      "loss": 0.3602,
       "step": 109360
     },
     {
-      "epoch": 3.85,
-      "learning_rate": 6.660661941038632e-06,
-      "loss": 0.2475,
+      "epoch": 3.941507189966483,
+      "grad_norm": 0.26603254675865173,
+      "learning_rate": 5.663404733992389e-06,
+      "loss": 0.3944,
       "step": 109365
     },
     {
-      "epoch": 3.85,
-      "learning_rate": 6.658726040330155e-06,
-      "loss": 0.2356,
+      "epoch": 3.9416873896277074,
+      "grad_norm": 0.25029662251472473,
+      "learning_rate": 5.6615552345382405e-06,
+      "loss": 0.3695,
       "step": 109370
     },
     {
-      "epoch": 3.85,
-      "learning_rate": 6.6567903777710326e-06,
-      "loss": 0.2341,
+      "epoch": 3.941867589288932,
+      "grad_norm": 0.28742098808288574,
+      "learning_rate": 5.65970599856665e-06,
+      "loss": 0.3896,
       "step": 109375
     },
     {
-      "epoch": 3.85,
-      "learning_rate": 6.6548549533863995e-06,
-      "loss": 0.2409,
+      "epoch": 3.942047788950157,
+      "grad_norm": 0.24512720108032227,
+      "learning_rate": 5.657857026102814e-06,
+      "loss": 0.3886,
       "step": 109380
     },
     {
-      "epoch": 3.85,
-      "learning_rate": 6.652919767201377e-06,
-      "loss": 0.2327,
+      "epoch": 3.9422279886113816,
+      "grad_norm": 0.23668235540390015,
+      "learning_rate": 5.65600831717194e-06,
+      "loss": 0.3648,
       "step": 109385
     },
     {
-      "epoch": 3.85,
-      "learning_rate": 6.650984819241099e-06,
-      "loss": 0.2545,
+      "epoch": 3.942408188272606,
+      "grad_norm": 0.22627753019332886,
+      "learning_rate": 5.654159871799206e-06,
+      "loss": 0.3937,
       "step": 109390
     },
     {
-      "epoch": 3.85,
-      "learning_rate": 6.6490501095307e-06,
-      "loss": 0.2491,
+      "epoch": 3.9425883879338306,
+      "grad_norm": 0.2681983709335327,
+      "learning_rate": 5.652311690009798e-06,
+      "loss": 0.365,
       "step": 109395
     },
     {
-      "epoch": 3.85,
-      "learning_rate": 6.647115638095288e-06,
-      "loss": 0.2579,
+      "epoch": 3.9427685875950553,
+      "grad_norm": 0.19054733216762543,
+      "learning_rate": 5.650463771828898e-06,
+      "loss": 0.3927,
       "step": 109400
     },
     {
-      "epoch": 3.85,
-      "learning_rate": 6.6451814049599825e-06,
-      "loss": 0.2572,
+      "epoch": 3.94294878725628,
+      "grad_norm": 0.279351145029068,
+      "learning_rate": 5.648616117281677e-06,
+      "loss": 0.3602,
       "step": 109405
     },
     {
-      "epoch": 3.85,
-      "learning_rate": 6.643247410149903e-06,
-      "loss": 0.2585,
+      "epoch": 3.943128986917505,
+      "grad_norm": 0.20294272899627686,
+      "learning_rate": 5.64676872639332e-06,
+      "loss": 0.365,
       "step": 109410
     },
     {
-      "epoch": 3.85,
-      "learning_rate": 6.641313653690162e-06,
-      "loss": 0.2468,
+      "epoch": 3.943309186578729,
+      "grad_norm": 0.3039834201335907,
+      "learning_rate": 5.6449215991889934e-06,
+      "loss": 0.3659,
       "step": 109415
     },
     {
-      "epoch": 3.85,
-      "learning_rate": 6.639380135605855e-06,
-      "loss": 0.2684,
+      "epoch": 3.943489386239954,
+      "grad_norm": 0.3200457990169525,
+      "learning_rate": 5.643074735693862e-06,
+      "loss": 0.3948,
       "step": 109420
     },
     {
-      "epoch": 3.85,
-      "learning_rate": 6.6374468559221025e-06,
-      "loss": 0.2608,
+      "epoch": 3.9436695859011786,
+      "grad_norm": 0.2976410984992981,
+      "learning_rate": 5.641228135933091e-06,
+      "loss": 0.353,
       "step": 109425
     },
     {
-      "epoch": 3.85,
-      "learning_rate": 6.635513814664004e-06,
-      "loss": 0.255,
+      "epoch": 3.9438497855624033,
+      "grad_norm": 0.23301257193088531,
+      "learning_rate": 5.6393817999318365e-06,
+      "loss": 0.3733,
       "step": 109430
     },
     {
-      "epoch": 3.85,
-      "learning_rate": 6.6335810118566545e-06,
-      "loss": 0.2507,
+      "epoch": 3.9440299852236276,
+      "grad_norm": 0.2382785677909851,
+      "learning_rate": 5.637535727715251e-06,
+      "loss": 0.3598,
       "step": 109435
     },
     {
-      "epoch": 3.85,
-      "learning_rate": 6.631648447525144e-06,
-      "loss": 0.2536,
+      "epoch": 3.9442101848848523,
+      "grad_norm": 0.2509336471557617,
+      "learning_rate": 5.635689919308498e-06,
+      "loss": 0.3935,
       "step": 109440
     },
     {
-      "epoch": 3.85,
-      "learning_rate": 6.6297161216945735e-06,
-      "loss": 0.2414,
+      "epoch": 3.944390384546077,
+      "grad_norm": 0.2762444317340851,
+      "learning_rate": 5.633844374736724e-06,
+      "loss": 0.4127,
       "step": 109445
     },
     {
-      "epoch": 3.85,
-      "learning_rate": 6.6277840343900415e-06,
-      "loss": 0.2569,
+      "epoch": 3.9445705842073018,
+      "grad_norm": 0.2096366137266159,
+      "learning_rate": 5.6319990940250696e-06,
+      "loss": 0.393,
       "step": 109450
     },
     {
-      "epoch": 3.85,
-      "learning_rate": 6.625852185636625e-06,
-      "loss": 0.2659,
+      "epoch": 3.9447507838685265,
+      "grad_norm": 0.23750926554203033,
+      "learning_rate": 5.630154077198677e-06,
+      "loss": 0.371,
       "step": 109455
     },
     {
-      "epoch": 3.85,
-      "learning_rate": 6.623920575459403e-06,
-      "loss": 0.2423,
+      "epoch": 3.944930983529751,
+      "grad_norm": 0.22931966185569763,
+      "learning_rate": 5.628309324282676e-06,
+      "loss": 0.3814,
       "step": 109460
     },
     {
-      "epoch": 3.85,
-      "learning_rate": 6.6219892038834666e-06,
-      "loss": 0.2552,
+      "epoch": 3.9451111831909755,
+      "grad_norm": 0.23210839927196503,
+      "learning_rate": 5.626464835302228e-06,
+      "loss": 0.3851,
       "step": 109465
     },
     {
-      "epoch": 3.85,
-      "learning_rate": 6.62005807093389e-06,
-      "loss": 0.2629,
+      "epoch": 3.9452913828522003,
+      "grad_norm": 0.3315199911594391,
+      "learning_rate": 5.62462061028243e-06,
+      "loss": 0.3627,
       "step": 109470
     },
     {
-      "epoch": 3.85,
-      "learning_rate": 6.618127176635747e-06,
-      "loss": 0.2408,
+      "epoch": 3.945471582513425,
+      "grad_norm": 0.20502960681915283,
+      "learning_rate": 5.6227766492484315e-06,
+      "loss": 0.391,
       "step": 109475
     },
     {
-      "epoch": 3.85,
-      "learning_rate": 6.6161965210141e-06,
-      "loss": 0.2507,
+      "epoch": 3.9456517821746493,
+      "grad_norm": 0.23437687754631042,
+      "learning_rate": 5.62093295222535e-06,
+      "loss": 0.3435,
       "step": 109480
     },
     {
-      "epoch": 3.85,
-      "learning_rate": 6.614266104094035e-06,
-      "loss": 0.2448,
+      "epoch": 3.945831981835874,
+      "grad_norm": 0.2661973237991333,
+      "learning_rate": 5.619089519238305e-06,
+      "loss": 0.3835,
       "step": 109485
     },
     {
-      "epoch": 3.85,
-      "learning_rate": 6.612335925900606e-06,
-      "loss": 0.2761,
+      "epoch": 3.9460121814970988,
+      "grad_norm": 0.2659934163093567,
+      "learning_rate": 5.617246350312414e-06,
+      "loss": 0.4016,
       "step": 109490
     },
     {
-      "epoch": 3.85,
-      "learning_rate": 6.6104059864588705e-06,
-      "loss": 0.2613,
+      "epoch": 3.9461923811583235,
+      "grad_norm": 0.24436159431934357,
+      "learning_rate": 5.6154034454727885e-06,
+      "loss": 0.3518,
       "step": 109495
     },
     {
-      "epoch": 3.85,
-      "learning_rate": 6.608476285793902e-06,
-      "loss": 0.2673,
+      "epoch": 3.9463725808195482,
+      "grad_norm": 0.23105373978614807,
+      "learning_rate": 5.6135608047445306e-06,
+      "loss": 0.3574,
       "step": 109500
     },
     {
-      "epoch": 3.85,
-      "eval_loss": 0.2506560981273651,
-      "eval_runtime": 10.5745,
-      "eval_samples_per_second": 9.457,
-      "eval_steps_per_second": 9.457,
+      "epoch": 3.9463725808195482,
+      "eval_loss": 0.4281553030014038,
+      "eval_runtime": 3.5325,
+      "eval_samples_per_second": 28.309,
+      "eval_steps_per_second": 7.077,
       "step": 109500
     },
     {
-      "epoch": 3.85,
-      "learning_rate": 6.60654682393074e-06,
-      "loss": 0.2786,
+      "epoch": 3.946552780480773,
+      "grad_norm": 0.23463231325149536,
+      "learning_rate": 5.611718428152759e-06,
+      "loss": 0.372,
       "step": 109505
     },
     {
-      "epoch": 3.85,
-      "learning_rate": 6.6046176008944555e-06,
-      "loss": 0.2617,
+      "epoch": 3.9467329801419972,
+      "grad_norm": 0.19474101066589355,
+      "learning_rate": 5.609876315722573e-06,
+      "loss": 0.3851,
       "step": 109510
     },
     {
-      "epoch": 3.85,
-      "learning_rate": 6.602688616710087e-06,
-      "loss": 0.2634,
+      "epoch": 3.946913179803222,
+      "grad_norm": 0.27828896045684814,
+      "learning_rate": 5.6080344674790676e-06,
+      "loss": 0.389,
       "step": 109515
     },
     {
-      "epoch": 3.85,
-      "learning_rate": 6.600759871402673e-06,
-      "loss": 0.251,
+      "epoch": 3.9470933794644467,
+      "grad_norm": 0.22116512060165405,
+      "learning_rate": 5.606192883447336e-06,
+      "loss": 0.3621,
       "step": 109520
     },
     {
-      "epoch": 3.85,
-      "learning_rate": 6.598831364997276e-06,
-      "loss": 0.2717,
+      "epoch": 3.947273579125671,
+      "grad_norm": 0.27116748690605164,
+      "learning_rate": 5.604351563652466e-06,
+      "loss": 0.3818,
       "step": 109525
     },
     {
-      "epoch": 3.85,
-      "learning_rate": 6.596903097518928e-06,
-      "loss": 0.2611,
+      "epoch": 3.9474537787868957,
+      "grad_norm": 0.2014303356409073,
+      "learning_rate": 5.60251050811956e-06,
+      "loss": 0.3935,
       "step": 109530
     },
     {
-      "epoch": 3.85,
-      "learning_rate": 6.594975068992657e-06,
-      "loss": 0.2771,
+      "epoch": 3.9476339784481205,
+      "grad_norm": 0.30118808150291443,
+      "learning_rate": 5.600669716873697e-06,
+      "loss": 0.4039,
       "step": 109535
     },
     {
-      "epoch": 3.85,
-      "learning_rate": 6.593047279443512e-06,
-      "loss": 0.2661,
+      "epoch": 3.947814178109345,
+      "grad_norm": 0.23722492158412933,
+      "learning_rate": 5.5988291899399385e-06,
+      "loss": 0.4329,
       "step": 109540
     },
     {
-      "epoch": 3.85,
-      "learning_rate": 6.5911197288965165e-06,
-      "loss": 0.2525,
+      "epoch": 3.94799437777057,
+      "grad_norm": 0.25418469309806824,
+      "learning_rate": 5.596988927343386e-06,
+      "loss": 0.3885,
       "step": 109545
     },
     {
-      "epoch": 3.85,
-      "learning_rate": 6.589192417376694e-06,
-      "loss": 0.2556,
+      "epoch": 3.9481745774317947,
+      "grad_norm": 0.21690379083156586,
+      "learning_rate": 5.595148929109095e-06,
+      "loss": 0.3706,
       "step": 109550
     },
     {
-      "epoch": 3.85,
-      "learning_rate": 6.58726534490908e-06,
-      "loss": 0.2301,
+      "epoch": 3.948354777093019,
+      "grad_norm": 0.2371807098388672,
+      "learning_rate": 5.593309195262156e-06,
+      "loss": 0.401,
       "step": 109555
     },
     {
-      "epoch": 3.85,
-      "learning_rate": 6.585338511518685e-06,
-      "loss": 0.2549,
+      "epoch": 3.9485349767542437,
+      "grad_norm": 0.2515186369419098,
+      "learning_rate": 5.591469725827614e-06,
+      "loss": 0.3795,
       "step": 109560
     },
     {
-      "epoch": 3.85,
-      "learning_rate": 6.583411917230542e-06,
-      "loss": 0.2957,
+      "epoch": 3.9487151764154684,
+      "grad_norm": 0.22983525693416595,
+      "learning_rate": 5.589630520830536e-06,
+      "loss": 0.4015,
       "step": 109565
     },
     {
-      "epoch": 3.85,
-      "learning_rate": 6.581485562069656e-06,
-      "loss": 0.227,
+      "epoch": 3.9488953760766927,
+      "grad_norm": 0.2533568739891052,
+      "learning_rate": 5.587791580295987e-06,
+      "loss": 0.3962,
       "step": 109570
     },
     {
-      "epoch": 3.86,
-      "learning_rate": 6.579559446061037e-06,
-      "loss": 0.2694,
+      "epoch": 3.9490755757379175,
+      "grad_norm": 0.24301256239414215,
+      "learning_rate": 5.58595290424902e-06,
+      "loss": 0.3958,
       "step": 109575
     },
     {
-      "epoch": 3.86,
-      "learning_rate": 6.577633569229705e-06,
-      "loss": 0.2556,
+      "epoch": 3.949255775399142,
+      "grad_norm": 0.2538679838180542,
+      "learning_rate": 5.5841144927146875e-06,
+      "loss": 0.4044,
       "step": 109580
     },
     {
-      "epoch": 3.86,
-      "learning_rate": 6.575707931600661e-06,
-      "loss": 0.2774,
+      "epoch": 3.949435975060367,
+      "grad_norm": 0.2607662081718445,
+      "learning_rate": 5.582276345718037e-06,
+      "loss": 0.3572,
       "step": 109585
     },
     {
-      "epoch": 3.86,
-      "learning_rate": 6.573782533198908e-06,
-      "loss": 0.2739,
+      "epoch": 3.9496161747215917,
+      "grad_norm": 0.23929943144321442,
+      "learning_rate": 5.5804384632841044e-06,
+      "loss": 0.3594,
       "step": 109590
     },
     {
-      "epoch": 3.86,
-      "learning_rate": 6.5718573740494365e-06,
-      "loss": 0.2586,
+      "epoch": 3.9497963743828164,
+      "grad_norm": 0.29421868920326233,
+      "learning_rate": 5.578600845437942e-06,
+      "loss": 0.3578,
       "step": 109595
     },
     {
-      "epoch": 3.86,
-      "learning_rate": 6.569932454177261e-06,
-      "loss": 0.2453,
+      "epoch": 3.9499765740440407,
+      "grad_norm": 0.19604270160198212,
+      "learning_rate": 5.576763492204587e-06,
+      "loss": 0.3804,
       "step": 109600
     },
     {
-      "epoch": 3.86,
-      "learning_rate": 6.568007773607357e-06,
-      "loss": 0.2472,
+      "epoch": 3.9501567737052654,
+      "grad_norm": 0.2703362703323364,
+      "learning_rate": 5.574926403609066e-06,
+      "loss": 0.41,
       "step": 109605
     },
     {
-      "epoch": 3.86,
-      "learning_rate": 6.566083332364734e-06,
-      "loss": 0.277,
+      "epoch": 3.95033697336649,
+      "grad_norm": 0.336032509803772,
+      "learning_rate": 5.573089579676413e-06,
+      "loss": 0.3946,
       "step": 109610
     },
     {
-      "epoch": 3.86,
-      "learning_rate": 6.5641591304743625e-06,
-      "loss": 0.2531,
+      "epoch": 3.9505171730277144,
+      "grad_norm": 0.3432498574256897,
+      "learning_rate": 5.5712530204316545e-06,
+      "loss": 0.3912,
       "step": 109615
     },
     {
-      "epoch": 3.86,
-      "learning_rate": 6.5622351679612424e-06,
-      "loss": 0.2533,
+      "epoch": 3.950697372688939,
+      "grad_norm": 0.24581557512283325,
+      "learning_rate": 5.569416725899804e-06,
+      "loss": 0.3591,
       "step": 109620
     },
     {
-      "epoch": 3.86,
-      "learning_rate": 6.560311444850351e-06,
-      "loss": 0.2605,
+      "epoch": 3.950877572350164,
+      "grad_norm": 0.2500128448009491,
+      "learning_rate": 5.567580696105895e-06,
+      "loss": 0.4141,
       "step": 109625
     },
     {
-      "epoch": 3.86,
-      "learning_rate": 6.558387961166659e-06,
-      "loss": 0.2622,
+      "epoch": 3.9510577720113886,
+      "grad_norm": 0.2446908950805664,
+      "learning_rate": 5.565744931074937e-06,
+      "loss": 0.3828,
       "step": 109630
     },
     {
-      "epoch": 3.86,
-      "learning_rate": 6.55646471693514e-06,
-      "loss": 0.2792,
+      "epoch": 3.9512379716726134,
+      "grad_norm": 0.22978517413139343,
+      "learning_rate": 5.563909430831943e-06,
+      "loss": 0.3831,
       "step": 109635
     },
     {
-      "epoch": 3.86,
-      "learning_rate": 6.554541712180778e-06,
-      "loss": 0.2643,
+      "epoch": 3.951418171333838,
+      "grad_norm": 0.2268865555524826,
+      "learning_rate": 5.56207419540192e-06,
+      "loss": 0.3924,
       "step": 109640
     },
     {
-      "epoch": 3.86,
-      "learning_rate": 6.5526189469285375e-06,
-      "loss": 0.2605,
+      "epoch": 3.9515983709950624,
+      "grad_norm": 0.22920677065849304,
+      "learning_rate": 5.560239224809866e-06,
+      "loss": 0.3564,
       "step": 109645
     },
     {
-      "epoch": 3.86,
-      "learning_rate": 6.5506964212033754e-06,
-      "loss": 0.2747,
+      "epoch": 3.951778570656287,
+      "grad_norm": 0.24135887622833252,
+      "learning_rate": 5.558404519080801e-06,
+      "loss": 0.3767,
       "step": 109650
     },
     {
-      "epoch": 3.86,
-      "learning_rate": 6.548774135030269e-06,
-      "loss": 0.2637,
+      "epoch": 3.951958770317512,
+      "grad_norm": 0.27055665850639343,
+      "learning_rate": 5.556570078239703e-06,
+      "loss": 0.4004,
       "step": 109655
     },
     {
-      "epoch": 3.86,
-      "learning_rate": 6.546852088434172e-06,
-      "loss": 0.2543,
+      "epoch": 3.9521389699787366,
+      "grad_norm": 0.19351933896541595,
+      "learning_rate": 5.554735902311567e-06,
+      "loss": 0.3738,
       "step": 109660
     },
     {
-      "epoch": 3.86,
-      "learning_rate": 6.544930281440028e-06,
-      "loss": 0.2289,
+      "epoch": 3.952319169639961,
+      "grad_norm": 0.2451026439666748,
+      "learning_rate": 5.552901991321399e-06,
+      "loss": 0.3815,
       "step": 109665
     },
     {
-      "epoch": 3.86,
-      "learning_rate": 6.543008714072804e-06,
-      "loss": 0.2327,
+      "epoch": 3.9524993693011856,
+      "grad_norm": 0.2261337786912918,
+      "learning_rate": 5.5510683452941745e-06,
+      "loss": 0.3699,
       "step": 109670
     },
     {
-      "epoch": 3.86,
-      "learning_rate": 6.541087386357459e-06,
-      "loss": 0.2571,
+      "epoch": 3.9526795689624104,
+      "grad_norm": 0.22415819764137268,
+      "learning_rate": 5.549234964254879e-06,
+      "loss": 0.383,
       "step": 109675
     },
     {
-      "epoch": 3.86,
-      "learning_rate": 6.539166298318924e-06,
-      "loss": 0.2668,
+      "epoch": 3.952859768623635,
+      "grad_norm": 0.23757857084274292,
+      "learning_rate": 5.547401848228489e-06,
+      "loss": 0.3905,
       "step": 109680
     },
     {
-      "epoch": 3.86,
-      "learning_rate": 6.5372454499821515e-06,
-      "loss": 0.246,
+      "epoch": 3.95303996828486,
+      "grad_norm": 0.22890903055667877,
+      "learning_rate": 5.545568997239978e-06,
+      "loss": 0.3978,
       "step": 109685
     },
     {
-      "epoch": 3.86,
-      "learning_rate": 6.53532484137207e-06,
-      "loss": 0.2797,
+      "epoch": 3.953220167946084,
+      "grad_norm": 0.24089115858078003,
+      "learning_rate": 5.543736411314329e-06,
+      "loss": 0.3726,
       "step": 109690
     },
     {
-      "epoch": 3.86,
-      "learning_rate": 6.5334044725136375e-06,
-      "loss": 0.2634,
+      "epoch": 3.953400367607309,
+      "grad_norm": 0.2351144254207611,
+      "learning_rate": 5.541904090476505e-06,
+      "loss": 0.3726,
       "step": 109695
     },
     {
-      "epoch": 3.86,
-      "learning_rate": 6.5314843434317764e-06,
-      "loss": 0.2428,
+      "epoch": 3.9535805672685336,
+      "grad_norm": 0.20729920268058777,
+      "learning_rate": 5.54007203475147e-06,
+      "loss": 0.3672,
       "step": 109700
     },
     {
-      "epoch": 3.86,
-      "learning_rate": 6.529564454151418e-06,
-      "loss": 0.2481,
+      "epoch": 3.9537607669297583,
+      "grad_norm": 0.231421560049057,
+      "learning_rate": 5.538240244164186e-06,
+      "loss": 0.3376,
       "step": 109705
     },
     {
-      "epoch": 3.86,
-      "learning_rate": 6.527644804697489e-06,
-      "loss": 0.2572,
+      "epoch": 3.9539409665909826,
+      "grad_norm": 0.2679100036621094,
+      "learning_rate": 5.536408718739605e-06,
+      "loss": 0.3995,
       "step": 109710
     },
     {
-      "epoch": 3.86,
-      "learning_rate": 6.525725395094923e-06,
-      "loss": 0.254,
+      "epoch": 3.9541211662522073,
+      "grad_norm": 0.2754489481449127,
+      "learning_rate": 5.534577458502693e-06,
+      "loss": 0.3953,
       "step": 109715
     },
     {
-      "epoch": 3.86,
-      "learning_rate": 6.5238062253686314e-06,
-      "loss": 0.2514,
+      "epoch": 3.954301365913432,
+      "grad_norm": 0.21285389363765717,
+      "learning_rate": 5.532746463478403e-06,
+      "loss": 0.3766,
       "step": 109720
     },
     {
-      "epoch": 3.86,
-      "learning_rate": 6.5218872955435465e-06,
-      "loss": 0.2573,
+      "epoch": 3.954481565574657,
+      "grad_norm": 0.24366551637649536,
+      "learning_rate": 5.5309157336916575e-06,
+      "loss": 0.3781,
       "step": 109725
     },
     {
-      "epoch": 3.86,
-      "learning_rate": 6.519968605644569e-06,
-      "loss": 0.2663,
+      "epoch": 3.9546617652358815,
+      "grad_norm": 0.20494620501995087,
+      "learning_rate": 5.529085269167422e-06,
+      "loss": 0.3657,
       "step": 109730
     },
     {
-      "epoch": 3.86,
-      "learning_rate": 6.518050155696631e-06,
-      "loss": 0.2713,
+      "epoch": 3.9548419648971063,
+      "grad_norm": 0.268721342086792,
+      "learning_rate": 5.527255069930626e-06,
+      "loss": 0.3734,
       "step": 109735
     },
     {
-      "epoch": 3.86,
-      "learning_rate": 6.516131945724629e-06,
-      "loss": 0.279,
+      "epoch": 3.9550221645583306,
+      "grad_norm": 0.22930163145065308,
+      "learning_rate": 5.525425136006216e-06,
+      "loss": 0.368,
       "step": 109740
     },
     {
-      "epoch": 3.86,
-      "learning_rate": 6.514213975753472e-06,
-      "loss": 0.2773,
+      "epoch": 3.9552023642195553,
+      "grad_norm": 0.25767385959625244,
+      "learning_rate": 5.5235954674191235e-06,
+      "loss": 0.3905,
       "step": 109745
     },
     {
-      "epoch": 3.86,
-      "learning_rate": 6.5122962458080586e-06,
-      "loss": 0.2522,
+      "epoch": 3.95538256388078,
+      "grad_norm": 0.24452823400497437,
+      "learning_rate": 5.5217660641942585e-06,
+      "loss": 0.38,
       "step": 109750
     },
     {
-      "epoch": 3.86,
-      "learning_rate": 6.510378755913302e-06,
-      "loss": 0.2602,
+      "epoch": 3.9555627635420043,
+      "grad_norm": 0.2989307940006256,
+      "learning_rate": 5.519936926356567e-06,
+      "loss": 0.3936,
       "step": 109755
     },
     {
-      "epoch": 3.86,
-      "learning_rate": 6.5084615060940905e-06,
-      "loss": 0.2345,
+      "epoch": 3.955742963203229,
+      "grad_norm": 0.23451735079288483,
+      "learning_rate": 5.518108053930962e-06,
+      "loss": 0.3972,
       "step": 109760
     },
     {
-      "epoch": 3.86,
-      "learning_rate": 6.5065444963753134e-06,
-      "loss": 0.2481,
+      "epoch": 3.955923162864454,
+      "grad_norm": 0.27269425988197327,
+      "learning_rate": 5.516279446942366e-06,
+      "loss": 0.379,
       "step": 109765
     },
     {
-      "epoch": 3.86,
-      "learning_rate": 6.504627726781875e-06,
-      "loss": 0.2813,
+      "epoch": 3.9561033625256785,
+      "grad_norm": 0.2326863706111908,
+      "learning_rate": 5.5144511054156885e-06,
+      "loss": 0.3866,
       "step": 109770
     },
     {
-      "epoch": 3.86,
-      "learning_rate": 6.502711197338651e-06,
-      "loss": 0.2379,
+      "epoch": 3.9562835621869032,
+      "grad_norm": 0.2507339119911194,
+      "learning_rate": 5.512623029375835e-06,
+      "loss": 0.3986,
       "step": 109775
     },
     {
-      "epoch": 3.86,
-      "learning_rate": 6.500794908070537e-06,
-      "loss": 0.247,
+      "epoch": 3.956463761848128,
+      "grad_norm": 0.23485083878040314,
+      "learning_rate": 5.510795218847725e-06,
+      "loss": 0.4092,
       "step": 109780
     },
     {
-      "epoch": 3.86,
-      "learning_rate": 6.498878859002405e-06,
-      "loss": 0.2438,
+      "epoch": 3.9566439615093523,
+      "grad_norm": 0.22537368535995483,
+      "learning_rate": 5.5089676738562606e-06,
+      "loss": 0.3579,
       "step": 109785
     },
     {
-      "epoch": 3.86,
-      "learning_rate": 6.496963050159144e-06,
-      "loss": 0.2464,
+      "epoch": 3.956824161170577,
+      "grad_norm": 0.2544284760951996,
+      "learning_rate": 5.507140394426335e-06,
+      "loss": 0.3826,
       "step": 109790
     },
     {
-      "epoch": 3.86,
-      "learning_rate": 6.495047481565625e-06,
-      "loss": 0.253,
+      "epoch": 3.9570043608318017,
+      "grad_norm": 0.2154984176158905,
+      "learning_rate": 5.50531338058285e-06,
+      "loss": 0.3775,
       "step": 109795
     },
     {
-      "epoch": 3.86,
-      "learning_rate": 6.493132153246717e-06,
-      "loss": 0.2706,
+      "epoch": 3.957184560493026,
+      "grad_norm": 0.29244160652160645,
+      "learning_rate": 5.503486632350696e-06,
+      "loss": 0.3983,
       "step": 109800
     },
     {
-      "epoch": 3.86,
-      "learning_rate": 6.491217065227287e-06,
-      "loss": 0.2471,
+      "epoch": 3.9573647601542508,
+      "grad_norm": 0.21299979090690613,
+      "learning_rate": 5.501660149754753e-06,
+      "loss": 0.36,
       "step": 109805
     },
     {
-      "epoch": 3.86,
-      "learning_rate": 6.489302217532212e-06,
-      "loss": 0.25,
+      "epoch": 3.9575449598154755,
+      "grad_norm": 0.21175287663936615,
+      "learning_rate": 5.4998339328199255e-06,
+      "loss": 0.3463,
       "step": 109810
     },
     {
-      "epoch": 3.86,
-      "learning_rate": 6.487387610186346e-06,
-      "loss": 0.2396,
+      "epoch": 3.9577251594767002,
+      "grad_norm": 0.19887837767601013,
+      "learning_rate": 5.498007981571082e-06,
+      "loss": 0.3646,
       "step": 109815
     },
     {
-      "epoch": 3.86,
-      "learning_rate": 6.485473243214554e-06,
-      "loss": 0.2484,
+      "epoch": 3.957905359137925,
+      "grad_norm": 0.22215895354747772,
+      "learning_rate": 5.4961822960331064e-06,
+      "loss": 0.3875,
       "step": 109820
     },
     {
-      "epoch": 3.86,
-      "learning_rate": 6.483559116641683e-06,
-      "loss": 0.2551,
+      "epoch": 3.9580855587991497,
+      "grad_norm": 0.2197951376438141,
+      "learning_rate": 5.494356876230869e-06,
+      "loss": 0.3831,
       "step": 109825
     },
     {
-      "epoch": 3.86,
-      "learning_rate": 6.481645230492594e-06,
-      "loss": 0.272,
+      "epoch": 3.958265758460374,
+      "grad_norm": 0.27231666445732117,
+      "learning_rate": 5.492531722189237e-06,
+      "loss": 0.3833,
       "step": 109830
     },
     {
-      "epoch": 3.86,
-      "learning_rate": 6.479731584792145e-06,
-      "loss": 0.2526,
+      "epoch": 3.9584459581215987,
+      "grad_norm": 0.22005848586559296,
+      "learning_rate": 5.490706833933096e-06,
+      "loss": 0.3738,
       "step": 109835
     },
     {
-      "epoch": 3.86,
-      "learning_rate": 6.477818179565176e-06,
-      "loss": 0.2607,
+      "epoch": 3.9586261577828235,
+      "grad_norm": 0.19976355135440826,
+      "learning_rate": 5.488882211487292e-06,
+      "loss": 0.4119,
       "step": 109840
     },
     {
-      "epoch": 3.86,
-      "learning_rate": 6.475905014836525e-06,
-      "loss": 0.2703,
+      "epoch": 3.9588063574440477,
+      "grad_norm": 0.24296000599861145,
+      "learning_rate": 5.487057854876682e-06,
+      "loss": 0.3737,
       "step": 109845
     },
     {
-      "epoch": 3.86,
-      "learning_rate": 6.4739920906310456e-06,
-      "loss": 0.2606,
+      "epoch": 3.9589865571052725,
+      "grad_norm": 0.32272353768348694,
+      "learning_rate": 5.485233764126138e-06,
+      "loss": 0.421,
       "step": 109850
     },
     {
-      "epoch": 3.87,
-      "learning_rate": 6.472079406973569e-06,
-      "loss": 0.2238,
+      "epoch": 3.959166756766497,
+      "grad_norm": 0.23841454088687897,
+      "learning_rate": 5.4834099392605e-06,
+      "loss": 0.3983,
       "step": 109855
     },
     {
-      "epoch": 3.87,
-      "learning_rate": 6.470166963888932e-06,
-      "loss": 0.2683,
+      "epoch": 3.959346956427722,
+      "grad_norm": 0.2716634273529053,
+      "learning_rate": 5.481586380304637e-06,
+      "loss": 0.3676,
       "step": 109860
     },
     {
-      "epoch": 3.87,
-      "learning_rate": 6.4682547614019596e-06,
-      "loss": 0.2836,
+      "epoch": 3.9595271560889467,
+      "grad_norm": 0.2504705488681793,
+      "learning_rate": 5.479763087283374e-06,
+      "loss": 0.3868,
       "step": 109865
     },
     {
-      "epoch": 3.87,
-      "learning_rate": 6.466342799537492e-06,
-      "loss": 0.2543,
+      "epoch": 3.9597073557501714,
+      "grad_norm": 0.2552550137042999,
+      "learning_rate": 5.4779400602215505e-06,
+      "loss": 0.3778,
       "step": 109870
     },
     {
-      "epoch": 3.87,
-      "learning_rate": 6.4644310783203485e-06,
-      "loss": 0.2591,
+      "epoch": 3.9598875554113957,
+      "grad_norm": 0.24792692065238953,
+      "learning_rate": 5.476117299144024e-06,
+      "loss": 0.3571,
       "step": 109875
     },
     {
-      "epoch": 3.87,
-      "learning_rate": 6.462519597775346e-06,
-      "loss": 0.2443,
+      "epoch": 3.9600677550726204,
+      "grad_norm": 0.2851993441581726,
+      "learning_rate": 5.474294804075617e-06,
+      "loss": 0.3578,
       "step": 109880
     },
     {
-      "epoch": 3.87,
-      "learning_rate": 6.460608357927309e-06,
-      "loss": 0.2622,
+      "epoch": 3.960247954733845,
+      "grad_norm": 0.28467345237731934,
+      "learning_rate": 5.472472575041166e-06,
+      "loss": 0.3829,
       "step": 109885
     },
     {
-      "epoch": 3.87,
-      "learning_rate": 6.458697358801061e-06,
-      "loss": 0.2791,
+      "epoch": 3.96042815439507,
+      "grad_norm": 0.26601019501686096,
+      "learning_rate": 5.470650612065492e-06,
+      "loss": 0.3478,
       "step": 109890
     },
     {
-      "epoch": 3.87,
-      "learning_rate": 6.456786600421411e-06,
-      "loss": 0.2643,
+      "epoch": 3.960608354056294,
+      "grad_norm": 0.24324598908424377,
+      "learning_rate": 5.468828915173416e-06,
+      "loss": 0.3622,
       "step": 109895
     },
     {
-      "epoch": 3.87,
-      "learning_rate": 6.454876082813158e-06,
-      "loss": 0.2539,
+      "epoch": 3.960788553717519,
+      "grad_norm": 0.21530494093894958,
+      "learning_rate": 5.4670074843897715e-06,
+      "loss": 0.3667,
       "step": 109900
     },
     {
-      "epoch": 3.87,
-      "learning_rate": 6.452965806001124e-06,
-      "loss": 0.2467,
+      "epoch": 3.9609687533787437,
+      "grad_norm": 0.2549532651901245,
+      "learning_rate": 5.465186319739371e-06,
+      "loss": 0.4011,
       "step": 109905
     },
     {
-      "epoch": 3.87,
-      "learning_rate": 6.451055770010106e-06,
-      "loss": 0.277,
+      "epoch": 3.9611489530399684,
+      "grad_norm": 0.26102617383003235,
+      "learning_rate": 5.463365421247024e-06,
+      "loss": 0.3918,
       "step": 109910
     },
     {
-      "epoch": 3.87,
-      "learning_rate": 6.449145974864906e-06,
-      "loss": 0.2755,
+      "epoch": 3.961329152701193,
+      "grad_norm": 0.24327853322029114,
+      "learning_rate": 5.46154478893754e-06,
+      "loss": 0.3286,
       "step": 109915
     },
     {
-      "epoch": 3.87,
-      "learning_rate": 6.447236420590311e-06,
-      "loss": 0.2534,
+      "epoch": 3.9615093523624174,
+      "grad_norm": 0.26538389921188354,
+      "learning_rate": 5.459724422835716e-06,
+      "loss": 0.3881,
       "step": 109920
     },
     {
-      "epoch": 3.87,
-      "learning_rate": 6.445327107211132e-06,
-      "loss": 0.2553,
+      "epoch": 3.961689552023642,
+      "grad_norm": 0.18152877688407898,
+      "learning_rate": 5.457904322966373e-06,
+      "loss": 0.4096,
       "step": 109925
     },
     {
-      "epoch": 3.87,
-      "learning_rate": 6.443418034752155e-06,
-      "loss": 0.277,
+      "epoch": 3.961869751684867,
+      "grad_norm": 0.2658418118953705,
+      "learning_rate": 5.456084489354307e-06,
+      "loss": 0.3956,
       "step": 109930
     },
     {
-      "epoch": 3.87,
-      "learning_rate": 6.441509203238155e-06,
-      "loss": 0.2739,
+      "epoch": 3.9620499513460916,
+      "grad_norm": 0.23971152305603027,
+      "learning_rate": 5.454264922024294e-06,
+      "loss": 0.4017,
       "step": 109935
     },
     {
-      "epoch": 3.87,
-      "learning_rate": 6.439600612693935e-06,
-      "loss": 0.2784,
+      "epoch": 3.962230151007316,
+      "grad_norm": 0.2468896061182022,
+      "learning_rate": 5.4524456210011425e-06,
+      "loss": 0.3612,
       "step": 109940
     },
     {
-      "epoch": 3.87,
-      "learning_rate": 6.437692263144266e-06,
-      "loss": 0.2548,
+      "epoch": 3.9624103506685406,
+      "grad_norm": 0.29942211508750916,
+      "learning_rate": 5.450626586309634e-06,
+      "loss": 0.3639,
       "step": 109945
     },
     {
-      "epoch": 3.87,
-      "learning_rate": 6.435784154613933e-06,
-      "loss": 0.2584,
+      "epoch": 3.9625905503297654,
+      "grad_norm": 0.22130388021469116,
+      "learning_rate": 5.448807817974555e-06,
+      "loss": 0.3656,
       "step": 109950
     },
     {
-      "epoch": 3.87,
-      "learning_rate": 6.43387628712771e-06,
-      "loss": 0.2587,
+      "epoch": 3.96277074999099,
+      "grad_norm": 0.24621397256851196,
+      "learning_rate": 5.4469893160206815e-06,
+      "loss": 0.3605,
       "step": 109955
     },
     {
-      "epoch": 3.87,
-      "learning_rate": 6.431968660710361e-06,
-      "loss": 0.2411,
+      "epoch": 3.962950949652215,
+      "grad_norm": 0.2520377039909363,
+      "learning_rate": 5.4451710804727885e-06,
+      "loss": 0.3968,
       "step": 109960
     },
     {
-      "epoch": 3.87,
-      "learning_rate": 6.430061275386667e-06,
-      "loss": 0.2401,
+      "epoch": 3.963131149313439,
+      "grad_norm": 0.20365937054157257,
+      "learning_rate": 5.443353111355659e-06,
+      "loss": 0.3776,
       "step": 109965
     },
     {
-      "epoch": 3.87,
-      "learning_rate": 6.428154131181393e-06,
-      "loss": 0.2648,
+      "epoch": 3.963311348974664,
+      "grad_norm": 0.2284347116947174,
+      "learning_rate": 5.44153540869406e-06,
+      "loss": 0.3964,
       "step": 109970
     },
     {
-      "epoch": 3.87,
-      "learning_rate": 6.426247228119297e-06,
-      "loss": 0.252,
+      "epoch": 3.9634915486358886,
+      "grad_norm": 0.2807764708995819,
+      "learning_rate": 5.439717972512751e-06,
+      "loss": 0.3695,
       "step": 109975
     },
     {
-      "epoch": 3.87,
-      "learning_rate": 6.424340566225132e-06,
-      "loss": 0.2508,
+      "epoch": 3.9636717482971133,
+      "grad_norm": 0.2894369959831238,
+      "learning_rate": 5.437900802836499e-06,
+      "loss": 0.377,
       "step": 109980
     },
     {
-      "epoch": 3.87,
-      "learning_rate": 6.422434145523671e-06,
-      "loss": 0.2496,
+      "epoch": 3.9638519479583376,
+      "grad_norm": 0.2694891095161438,
+      "learning_rate": 5.436083899690061e-06,
+      "loss": 0.3832,
       "step": 109985
     },
     {
-      "epoch": 3.87,
-      "learning_rate": 6.420527966039652e-06,
-      "loss": 0.2349,
+      "epoch": 3.9640321476195624,
+      "grad_norm": 0.2883671224117279,
+      "learning_rate": 5.4342672630981865e-06,
+      "loss": 0.3855,
       "step": 109990
     },
     {
-      "epoch": 3.87,
-      "learning_rate": 6.418622027797838e-06,
-      "loss": 0.2563,
+      "epoch": 3.964212347280787,
+      "grad_norm": 0.2609039545059204,
+      "learning_rate": 5.432450893085639e-06,
+      "loss": 0.3817,
       "step": 109995
     },
     {
-      "epoch": 3.87,
-      "learning_rate": 6.416716330822967e-06,
-      "loss": 0.2675,
+      "epoch": 3.964392546942012,
+      "grad_norm": 0.20613409578800201,
+      "learning_rate": 5.430634789677158e-06,
+      "loss": 0.3783,
       "step": 110000
     },
     {
-      "epoch": 3.87,
-      "eval_loss": 0.2505649924278259,
-      "eval_runtime": 10.5578,
-      "eval_samples_per_second": 9.472,
-      "eval_steps_per_second": 9.472,
+      "epoch": 3.964392546942012,
+      "eval_loss": 0.427830308675766,
+      "eval_runtime": 3.5225,
+      "eval_samples_per_second": 28.389,
+      "eval_steps_per_second": 7.097,
       "step": 110000
     },
     {
-      "epoch": 3.87,
-      "learning_rate": 6.414810875139796e-06,
-      "loss": 0.2671,
+      "epoch": 3.9645727466032366,
+      "grad_norm": 0.23606936633586884,
+      "learning_rate": 5.428818952897491e-06,
+      "loss": 0.3744,
       "step": 110005
     },
     {
-      "epoch": 3.87,
-      "learning_rate": 6.4129056607730555e-06,
-      "loss": 0.2459,
+      "epoch": 3.9647529462644613,
+      "grad_norm": 0.2058301568031311,
+      "learning_rate": 5.427003382771376e-06,
+      "loss": 0.3679,
       "step": 110010
     },
     {
-      "epoch": 3.87,
-      "learning_rate": 6.411000687747478e-06,
-      "loss": 0.2513,
+      "epoch": 3.9649331459256856,
+      "grad_norm": 0.2392200529575348,
+      "learning_rate": 5.425188079323543e-06,
+      "loss": 0.3732,
       "step": 110015
     },
     {
-      "epoch": 3.87,
-      "learning_rate": 6.409095956087813e-06,
-      "loss": 0.2715,
+      "epoch": 3.9651133455869103,
+      "grad_norm": 0.22312918305397034,
+      "learning_rate": 5.423373042578742e-06,
+      "loss": 0.3765,
       "step": 110020
     },
     {
-      "epoch": 3.87,
-      "learning_rate": 6.407191465818788e-06,
-      "loss": 0.2547,
+      "epoch": 3.965293545248135,
+      "grad_norm": 0.2611244320869446,
+      "learning_rate": 5.421558272561697e-06,
+      "loss": 0.3729,
       "step": 110025
     },
     {
-      "epoch": 3.87,
-      "learning_rate": 6.405287216965125e-06,
-      "loss": 0.2419,
+      "epoch": 3.9654737449093593,
+      "grad_norm": 0.22937054932117462,
+      "learning_rate": 5.419743769297117e-06,
+      "loss": 0.3881,
       "step": 110030
     },
     {
-      "epoch": 3.87,
-      "learning_rate": 6.403383209551547e-06,
-      "loss": 0.2474,
+      "epoch": 3.965653944570584,
+      "grad_norm": 0.2127593755722046,
+      "learning_rate": 5.417929532809746e-06,
+      "loss": 0.3301,
       "step": 110035
     },
     {
-      "epoch": 3.87,
-      "learning_rate": 6.401479443602792e-06,
-      "loss": 0.2438,
+      "epoch": 3.965834144231809,
+      "grad_norm": 0.25653398036956787,
+      "learning_rate": 5.416115563124283e-06,
+      "loss": 0.3757,
       "step": 110040
     },
     {
-      "epoch": 3.87,
-      "learning_rate": 6.39957591914356e-06,
-      "loss": 0.2548,
+      "epoch": 3.9660143438930335,
+      "grad_norm": 0.23030655086040497,
+      "learning_rate": 5.414301860265472e-06,
+      "loss": 0.3658,
       "step": 110045
     },
     {
-      "epoch": 3.87,
-      "learning_rate": 6.397672636198584e-06,
-      "loss": 0.2529,
+      "epoch": 3.9661945435542583,
+      "grad_norm": 0.28496742248535156,
+      "learning_rate": 5.412488424257997e-06,
+      "loss": 0.3403,
       "step": 110050
     },
     {
-      "epoch": 3.87,
-      "learning_rate": 6.395769594792564e-06,
-      "loss": 0.2649,
+      "epoch": 3.966374743215483,
+      "grad_norm": 0.21361872553825378,
+      "learning_rate": 5.410675255126568e-06,
+      "loss": 0.3534,
       "step": 110055
     },
     {
-      "epoch": 3.87,
-      "learning_rate": 6.3938667949502205e-06,
-      "loss": 0.2417,
+      "epoch": 3.9665549428767073,
+      "grad_norm": 0.22324657440185547,
+      "learning_rate": 5.408862352895905e-06,
+      "loss": 0.3898,
       "step": 110060
     },
     {
-      "epoch": 3.87,
-      "learning_rate": 6.3919642366962545e-06,
-      "loss": 0.2702,
+      "epoch": 3.966735142537932,
+      "grad_norm": 0.2588847875595093,
+      "learning_rate": 5.407049717590698e-06,
+      "loss": 0.3481,
       "step": 110065
     },
     {
-      "epoch": 3.87,
-      "learning_rate": 6.3900619200553705e-06,
-      "loss": 0.2615,
+      "epoch": 3.9669153421991568,
+      "grad_norm": 0.23955267667770386,
+      "learning_rate": 5.405237349235645e-06,
+      "loss": 0.3918,
       "step": 110070
     },
     {
-      "epoch": 3.87,
-      "learning_rate": 6.38815984505226e-06,
-      "loss": 0.2561,
+      "epoch": 3.967095541860381,
+      "grad_norm": 0.2915831208229065,
+      "learning_rate": 5.40342524785544e-06,
+      "loss": 0.3471,
       "step": 110075
     },
     {
-      "epoch": 3.87,
-      "learning_rate": 6.386258011711638e-06,
-      "loss": 0.2494,
+      "epoch": 3.967275741521606,
+      "grad_norm": 0.24253830313682556,
+      "learning_rate": 5.401613413474762e-06,
+      "loss": 0.371,
       "step": 110080
     },
     {
-      "epoch": 3.87,
-      "learning_rate": 6.384356420058188e-06,
-      "loss": 0.2514,
+      "epoch": 3.9674559411828305,
+      "grad_norm": 0.25190943479537964,
+      "learning_rate": 5.399801846118318e-06,
+      "loss": 0.3664,
       "step": 110085
     },
     {
-      "epoch": 3.87,
-      "learning_rate": 6.382455070116599e-06,
-      "loss": 0.2737,
+      "epoch": 3.9676361408440552,
+      "grad_norm": 0.2514355480670929,
+      "learning_rate": 5.397990545810777e-06,
+      "loss": 0.3753,
       "step": 110090
     },
     {
-      "epoch": 3.87,
-      "learning_rate": 6.380553961911556e-06,
-      "loss": 0.2449,
+      "epoch": 3.96781634050528,
+      "grad_norm": 0.2575605809688568,
+      "learning_rate": 5.396179512576821e-06,
+      "loss": 0.3553,
       "step": 110095
     },
     {
-      "epoch": 3.87,
-      "learning_rate": 6.378653095467757e-06,
-      "loss": 0.2506,
+      "epoch": 3.9679965401665047,
+      "grad_norm": 0.24077442288398743,
+      "learning_rate": 5.394368746441125e-06,
+      "loss": 0.3611,
       "step": 110100
     },
     {
-      "epoch": 3.87,
-      "learning_rate": 6.376752470809869e-06,
-      "loss": 0.2576,
+      "epoch": 3.968176739827729,
+      "grad_norm": 0.2745268940925598,
+      "learning_rate": 5.39255824742835e-06,
+      "loss": 0.3721,
       "step": 110105
     },
     {
-      "epoch": 3.87,
-      "learning_rate": 6.374852087962574e-06,
-      "loss": 0.243,
+      "epoch": 3.9683569394889537,
+      "grad_norm": 0.22541183233261108,
+      "learning_rate": 5.39074801556318e-06,
+      "loss": 0.373,
       "step": 110110
     },
     {
-      "epoch": 3.87,
-      "learning_rate": 6.372951946950559e-06,
-      "loss": 0.2739,
+      "epoch": 3.9685371391501785,
+      "grad_norm": 0.21690218150615692,
+      "learning_rate": 5.388938050870279e-06,
+      "loss": 0.3838,
       "step": 110115
     },
     {
-      "epoch": 3.87,
-      "learning_rate": 6.37105204779849e-06,
-      "loss": 0.2525,
+      "epoch": 3.9687173388114028,
+      "grad_norm": 0.26479387283325195,
+      "learning_rate": 5.387128353374285e-06,
+      "loss": 0.3713,
       "step": 110120
     },
     {
-      "epoch": 3.87,
-      "learning_rate": 6.369152390531031e-06,
-      "loss": 0.2615,
+      "epoch": 3.9688975384726275,
+      "grad_norm": 0.2510485351085663,
+      "learning_rate": 5.385318923099877e-06,
+      "loss": 0.3695,
       "step": 110125
     },
     {
-      "epoch": 3.87,
-      "learning_rate": 6.367252975172841e-06,
-      "loss": 0.2457,
+      "epoch": 3.9690777381338522,
+      "grad_norm": 0.23300889134407043,
+      "learning_rate": 5.383509760071703e-06,
+      "loss": 0.3741,
       "step": 110130
     },
     {
-      "epoch": 3.87,
-      "learning_rate": 6.3653538017486e-06,
-      "loss": 0.2682,
+      "epoch": 3.969257937795077,
+      "grad_norm": 0.21996130049228668,
+      "learning_rate": 5.3817008643144096e-06,
+      "loss": 0.3319,
       "step": 110135
     },
     {
-      "epoch": 3.88,
-      "learning_rate": 6.363454870282959e-06,
-      "loss": 0.266,
+      "epoch": 3.9694381374563017,
+      "grad_norm": 0.24466222524642944,
+      "learning_rate": 5.379892235852646e-06,
+      "loss": 0.3745,
       "step": 110140
     },
     {
-      "epoch": 3.88,
-      "learning_rate": 6.361556180800577e-06,
-      "loss": 0.2501,
+      "epoch": 3.9696183371175264,
+      "grad_norm": 0.2463037371635437,
+      "learning_rate": 5.378083874711043e-06,
+      "loss": 0.3904,
       "step": 110145
     },
     {
-      "epoch": 3.88,
-      "learning_rate": 6.3596577333260935e-06,
-      "loss": 0.261,
+      "epoch": 3.9697985367787507,
+      "grad_norm": 0.23316043615341187,
+      "learning_rate": 5.3762757809142555e-06,
+      "loss": 0.3767,
       "step": 110150
     },
     {
-      "epoch": 3.88,
-      "learning_rate": 6.35775952788418e-06,
-      "loss": 0.2517,
+      "epoch": 3.9699787364399755,
+      "grad_norm": 0.24817629158496857,
+      "learning_rate": 5.374467954486911e-06,
+      "loss": 0.3823,
       "step": 110155
     },
     {
-      "epoch": 3.88,
-      "learning_rate": 6.355861564499466e-06,
-      "loss": 0.283,
+      "epoch": 3.9701589361012,
+      "grad_norm": 0.26238492131233215,
+      "learning_rate": 5.372660395453641e-06,
+      "loss": 0.4208,
       "step": 110160
     },
     {
-      "epoch": 3.88,
-      "learning_rate": 6.353963843196609e-06,
-      "loss": 0.2687,
+      "epoch": 3.970339135762425,
+      "grad_norm": 0.2136494368314743,
+      "learning_rate": 5.370853103839071e-06,
+      "loss": 0.3898,
       "step": 110165
     },
     {
-      "epoch": 3.88,
-      "learning_rate": 6.352066364000236e-06,
-      "loss": 0.2655,
+      "epoch": 3.970519335423649,
+      "grad_norm": 0.28680065274238586,
+      "learning_rate": 5.36904607966783e-06,
+      "loss": 0.4224,
       "step": 110170
     },
     {
-      "epoch": 3.88,
-      "learning_rate": 6.350169126935001e-06,
-      "loss": 0.2467,
+      "epoch": 3.970699535084874,
+      "grad_norm": 0.35283222794532776,
+      "learning_rate": 5.367239322964529e-06,
+      "loss": 0.3185,
       "step": 110175
     },
     {
-      "epoch": 3.88,
-      "learning_rate": 6.348272132025526e-06,
-      "loss": 0.2478,
+      "epoch": 3.9708797347460987,
+      "grad_norm": 0.2509290277957916,
+      "learning_rate": 5.365432833753797e-06,
+      "loss": 0.3933,
       "step": 110180
     },
     {
-      "epoch": 3.88,
-      "learning_rate": 6.3463753792964485e-06,
-      "loss": 0.2508,
+      "epoch": 3.9710599344073234,
+      "grad_norm": 0.2476135492324829,
+      "learning_rate": 5.363626612060241e-06,
+      "loss": 0.3734,
       "step": 110185
     },
     {
-      "epoch": 3.88,
-      "learning_rate": 6.344478868772386e-06,
-      "loss": 0.2584,
+      "epoch": 3.971240134068548,
+      "grad_norm": 0.2526455223560333,
+      "learning_rate": 5.36182065790847e-06,
+      "loss": 0.3917,
       "step": 110190
     },
     {
-      "epoch": 3.88,
-      "learning_rate": 6.342582600477975e-06,
-      "loss": 0.2522,
+      "epoch": 3.9714203337297724,
+      "grad_norm": 0.2520613670349121,
+      "learning_rate": 5.360014971323094e-06,
+      "loss": 0.413,
       "step": 110195
     },
     {
-      "epoch": 3.88,
-      "learning_rate": 6.340686574437835e-06,
-      "loss": 0.2353,
+      "epoch": 3.971600533390997,
+      "grad_norm": 0.24791032075881958,
+      "learning_rate": 5.358209552328699e-06,
+      "loss": 0.3903,
       "step": 110200
     },
     {
-      "epoch": 3.88,
-      "learning_rate": 6.338790790676585e-06,
-      "loss": 0.2505,
+      "epoch": 3.971780733052222,
+      "grad_norm": 0.22999384999275208,
+      "learning_rate": 5.356404400949908e-06,
+      "loss": 0.3685,
       "step": 110205
     },
     {
-      "epoch": 3.88,
-      "learning_rate": 6.336895249218827e-06,
-      "loss": 0.2479,
+      "epoch": 3.9719609327134466,
+      "grad_norm": 0.25010523200035095,
+      "learning_rate": 5.354599517211309e-06,
+      "loss": 0.3803,
       "step": 110210
     },
     {
-      "epoch": 3.88,
-      "learning_rate": 6.334999950089188e-06,
-      "loss": 0.2476,
+      "epoch": 3.972141132374671,
+      "grad_norm": 0.2954365313053131,
+      "learning_rate": 5.3527949011374724e-06,
+      "loss": 0.4006,
       "step": 110215
     },
     {
-      "epoch": 3.88,
-      "learning_rate": 6.33310489331228e-06,
-      "loss": 0.2702,
+      "epoch": 3.9723213320358957,
+      "grad_norm": 0.23827119171619415,
+      "learning_rate": 5.350990552753013e-06,
+      "loss": 0.347,
       "step": 110220
     },
     {
-      "epoch": 3.88,
-      "learning_rate": 6.3312100789126975e-06,
-      "loss": 0.2398,
+      "epoch": 3.9725015316971204,
+      "grad_norm": 0.23179586231708527,
+      "learning_rate": 5.349186472082493e-06,
+      "loss": 0.3856,
       "step": 110225
     },
     {
-      "epoch": 3.88,
-      "learning_rate": 6.329315506915054e-06,
-      "loss": 0.2507,
+      "epoch": 3.972681731358345,
+      "grad_norm": 0.21170353889465332,
+      "learning_rate": 5.347382659150516e-06,
+      "loss": 0.3922,
       "step": 110230
     },
     {
-      "epoch": 3.88,
-      "learning_rate": 6.327421177343945e-06,
-      "loss": 0.2615,
+      "epoch": 3.97286193101957,
+      "grad_norm": 0.2426007241010666,
+      "learning_rate": 5.345579113981641e-06,
+      "loss": 0.3793,
       "step": 110235
     },
     {
-      "epoch": 3.88,
-      "learning_rate": 6.325527090223965e-06,
-      "loss": 0.2568,
+      "epoch": 3.9730421306807946,
+      "grad_norm": 0.28864866495132446,
+      "learning_rate": 5.343775836600437e-06,
+      "loss": 0.3969,
       "step": 110240
     },
     {
-      "epoch": 3.88,
-      "learning_rate": 6.323633245579702e-06,
-      "loss": 0.2687,
+      "epoch": 3.973222330342019,
+      "grad_norm": 0.27137455344200134,
+      "learning_rate": 5.341972827031488e-06,
+      "loss": 0.4164,
       "step": 110245
     },
     {
-      "epoch": 3.88,
-      "learning_rate": 6.321739643435762e-06,
-      "loss": 0.2595,
+      "epoch": 3.9734025300032436,
+      "grad_norm": 0.286211758852005,
+      "learning_rate": 5.340170085299353e-06,
+      "loss": 0.393,
       "step": 110250
     },
     {
-      "epoch": 3.88,
-      "learning_rate": 6.319846283816721e-06,
-      "loss": 0.2578,
+      "epoch": 3.9735827296644683,
+      "grad_norm": 0.21608895063400269,
+      "learning_rate": 5.338367611428596e-06,
+      "loss": 0.3818,
       "step": 110255
     },
     {
-      "epoch": 3.88,
-      "learning_rate": 6.317953166747168e-06,
-      "loss": 0.2479,
+      "epoch": 3.9737629293256926,
+      "grad_norm": 0.24834184348583221,
+      "learning_rate": 5.336565405443772e-06,
+      "loss": 0.3731,
       "step": 110260
     },
     {
-      "epoch": 3.88,
-      "learning_rate": 6.316060292251675e-06,
-      "loss": 0.2583,
+      "epoch": 3.9739431289869174,
+      "grad_norm": 0.24085558950901031,
+      "learning_rate": 5.334763467369433e-06,
+      "loss": 0.3548,
       "step": 110265
     },
     {
-      "epoch": 3.88,
-      "learning_rate": 6.3141676603548226e-06,
-      "loss": 0.2528,
+      "epoch": 3.974123328648142,
+      "grad_norm": 0.28163549304008484,
+      "learning_rate": 5.332961797230138e-06,
+      "loss": 0.3742,
       "step": 110270
     },
     {
-      "epoch": 3.88,
-      "learning_rate": 6.312275271081197e-06,
-      "loss": 0.2585,
+      "epoch": 3.974303528309367,
+      "grad_norm": 0.26481056213378906,
+      "learning_rate": 5.331160395050433e-06,
+      "loss": 0.4114,
       "step": 110275
     },
     {
-      "epoch": 3.88,
-      "learning_rate": 6.31038312445536e-06,
-      "loss": 0.2335,
+      "epoch": 3.9744837279705916,
+      "grad_norm": 0.252930223941803,
+      "learning_rate": 5.3293592608548606e-06,
+      "loss": 0.3308,
       "step": 110280
     },
     {
-      "epoch": 3.88,
-      "learning_rate": 6.308491220501874e-06,
-      "loss": 0.2517,
+      "epoch": 3.9746639276318163,
+      "grad_norm": 0.2916033864021301,
+      "learning_rate": 5.327558394667958e-06,
+      "loss": 0.3712,
       "step": 110285
     },
     {
-      "epoch": 3.88,
-      "learning_rate": 6.306599559245322e-06,
-      "loss": 0.2758,
+      "epoch": 3.9748441272930406,
+      "grad_norm": 0.27334362268447876,
+      "learning_rate": 5.325757796514258e-06,
+      "loss": 0.3883,
       "step": 110290
     },
     {
-      "epoch": 3.88,
-      "learning_rate": 6.304708140710253e-06,
-      "loss": 0.2717,
+      "epoch": 3.9750243269542653,
+      "grad_norm": 0.2750318944454193,
+      "learning_rate": 5.323957466418303e-06,
+      "loss": 0.3864,
       "step": 110295
     },
     {
-      "epoch": 3.88,
-      "learning_rate": 6.302816964921226e-06,
-      "loss": 0.2278,
+      "epoch": 3.97520452661549,
+      "grad_norm": 0.2912689447402954,
+      "learning_rate": 5.322157404404629e-06,
+      "loss": 0.3845,
       "step": 110300
     },
     {
-      "epoch": 3.88,
-      "learning_rate": 6.300926031902796e-06,
-      "loss": 0.2578,
+      "epoch": 3.9753847262767144,
+      "grad_norm": 0.26538118720054626,
+      "learning_rate": 5.320357610497734e-06,
+      "loss": 0.3879,
       "step": 110305
     },
     {
-      "epoch": 3.88,
-      "learning_rate": 6.299035341679524e-06,
-      "loss": 0.2574,
+      "epoch": 3.975564925937939,
+      "grad_norm": 0.20100893080234528,
+      "learning_rate": 5.3185580847221635e-06,
+      "loss": 0.4083,
       "step": 110310
     },
     {
-      "epoch": 3.88,
-      "learning_rate": 6.297144894275952e-06,
-      "loss": 0.2487,
+      "epoch": 3.975745125599164,
+      "grad_norm": 0.2257593870162964,
+      "learning_rate": 5.31675882710243e-06,
+      "loss": 0.3935,
       "step": 110315
     },
     {
-      "epoch": 3.88,
-      "learning_rate": 6.29525468971662e-06,
-      "loss": 0.2582,
+      "epoch": 3.9759253252603886,
+      "grad_norm": 0.21260130405426025,
+      "learning_rate": 5.314959837663039e-06,
+      "loss": 0.3592,
       "step": 110320
     },
     {
-      "epoch": 3.88,
-      "learning_rate": 6.293364728026082e-06,
-      "loss": 0.2558,
+      "epoch": 3.9761055249216133,
+      "grad_norm": 0.2437485307455063,
+      "learning_rate": 5.313161116428522e-06,
+      "loss": 0.3784,
       "step": 110325
     },
     {
-      "epoch": 3.88,
-      "learning_rate": 6.291475009228881e-06,
-      "loss": 0.2433,
+      "epoch": 3.976285724582838,
+      "grad_norm": 0.25405043363571167,
+      "learning_rate": 5.311362663423361e-06,
+      "loss": 0.3999,
       "step": 110330
     },
     {
-      "epoch": 3.88,
-      "learning_rate": 6.289585533349548e-06,
-      "loss": 0.2345,
+      "epoch": 3.9764659242440623,
+      "grad_norm": 0.20990101993083954,
+      "learning_rate": 5.30956447867208e-06,
+      "loss": 0.351,
       "step": 110335
     },
     {
-      "epoch": 3.88,
-      "learning_rate": 6.287696300412607e-06,
-      "loss": 0.2536,
+      "epoch": 3.976646123905287,
+      "grad_norm": 0.22888407111167908,
+      "learning_rate": 5.307766562199168e-06,
+      "loss": 0.4054,
       "step": 110340
     },
     {
-      "epoch": 3.88,
-      "learning_rate": 6.2858073104426066e-06,
-      "loss": 0.2563,
+      "epoch": 3.976826323566512,
+      "grad_norm": 0.19908618927001953,
+      "learning_rate": 5.305968914029125e-06,
+      "loss": 0.3972,
       "step": 110345
     },
     {
-      "epoch": 3.88,
-      "learning_rate": 6.283918563464067e-06,
-      "loss": 0.2594,
+      "epoch": 3.977006523227736,
+      "grad_norm": 0.26034486293792725,
+      "learning_rate": 5.3041715341864415e-06,
+      "loss": 0.345,
       "step": 110350
     },
     {
-      "epoch": 3.88,
-      "learning_rate": 6.282030059501506e-06,
-      "loss": 0.2484,
+      "epoch": 3.977186722888961,
+      "grad_norm": 0.2160859853029251,
+      "learning_rate": 5.3023744226956106e-06,
+      "loss": 0.3957,
       "step": 110355
     },
     {
-      "epoch": 3.88,
-      "learning_rate": 6.280141798579445e-06,
-      "loss": 0.2491,
+      "epoch": 3.9773669225501855,
+      "grad_norm": 0.23834609985351562,
+      "learning_rate": 5.300577579581107e-06,
+      "loss": 0.387,
       "step": 110360
     },
     {
-      "epoch": 3.88,
-      "learning_rate": 6.2782537807224125e-06,
-      "loss": 0.2373,
+      "epoch": 3.9775471222114103,
+      "grad_norm": 0.2771296203136444,
+      "learning_rate": 5.298781004867425e-06,
+      "loss": 0.383,
       "step": 110365
     },
     {
-      "epoch": 3.88,
-      "learning_rate": 6.276366005954917e-06,
-      "loss": 0.2352,
+      "epoch": 3.977727321872635,
+      "grad_norm": 0.2532643973827362,
+      "learning_rate": 5.296984698579038e-06,
+      "loss": 0.3822,
       "step": 110370
     },
     {
-      "epoch": 3.88,
-      "learning_rate": 6.274478474301463e-06,
-      "loss": 0.2756,
+      "epoch": 3.9779075215338597,
+      "grad_norm": 0.24019809067249298,
+      "learning_rate": 5.295188660740422e-06,
+      "loss": 0.3729,
       "step": 110375
     },
     {
-      "epoch": 3.88,
-      "learning_rate": 6.27259118578657e-06,
-      "loss": 0.2509,
+      "epoch": 3.978087721195084,
+      "grad_norm": 0.20523947477340698,
+      "learning_rate": 5.2933928913760426e-06,
+      "loss": 0.3622,
       "step": 110380
     },
     {
-      "epoch": 3.88,
-      "learning_rate": 6.270704140434733e-06,
-      "loss": 0.2575,
+      "epoch": 3.9782679208563088,
+      "grad_norm": 0.21291251480579376,
+      "learning_rate": 5.291597390510364e-06,
+      "loss": 0.3355,
       "step": 110385
     },
     {
-      "epoch": 3.88,
-      "learning_rate": 6.268817338270466e-06,
-      "loss": 0.2689,
+      "epoch": 3.9784481205175335,
+      "grad_norm": 0.18715035915374756,
+      "learning_rate": 5.289802158167862e-06,
+      "loss": 0.3336,
       "step": 110390
     },
     {
-      "epoch": 3.88,
-      "learning_rate": 6.266930779318262e-06,
-      "loss": 0.2609,
+      "epoch": 3.9786283201787582,
+      "grad_norm": 0.24453963339328766,
+      "learning_rate": 5.288007194372996e-06,
+      "loss": 0.4144,
       "step": 110395
     },
     {
-      "epoch": 3.88,
-      "learning_rate": 6.26504446360261e-06,
-      "loss": 0.2579,
+      "epoch": 3.9788085198399825,
+      "grad_norm": 0.24042107164859772,
+      "learning_rate": 5.2862124991502035e-06,
+      "loss": 0.3996,
       "step": 110400
     },
     {
-      "epoch": 3.88,
-      "learning_rate": 6.2631583911480164e-06,
-      "loss": 0.2455,
+      "epoch": 3.9789887195012072,
+      "grad_norm": 0.2552798390388489,
+      "learning_rate": 5.284418072523953e-06,
+      "loss": 0.3834,
       "step": 110405
     },
     {
-      "epoch": 3.88,
-      "learning_rate": 6.261272561978962e-06,
-      "loss": 0.2613,
+      "epoch": 3.979168919162432,
+      "grad_norm": 0.273781418800354,
+      "learning_rate": 5.282623914518686e-06,
+      "loss": 0.3961,
       "step": 110410
     },
     {
-      "epoch": 3.88,
-      "learning_rate": 6.259386976119932e-06,
-      "loss": 0.2372,
+      "epoch": 3.9793491188236567,
+      "grad_norm": 0.2784363329410553,
+      "learning_rate": 5.280830025158861e-06,
+      "loss": 0.3527,
       "step": 110415
     },
     {
-      "epoch": 3.88,
-      "learning_rate": 6.257501633595408e-06,
-      "loss": 0.2716,
+      "epoch": 3.9795293184848815,
+      "grad_norm": 0.28789547085762024,
+      "learning_rate": 5.279036404468904e-06,
+      "loss": 0.3491,
       "step": 110420
     },
     {
-      "epoch": 3.89,
-      "learning_rate": 6.2556165344298785e-06,
-      "loss": 0.2526,
+      "epoch": 3.9797095181461057,
+      "grad_norm": 0.20534764230251312,
+      "learning_rate": 5.277243052473252e-06,
+      "loss": 0.3759,
       "step": 110425
     },
     {
-      "epoch": 3.89,
-      "learning_rate": 6.253731678647809e-06,
-      "loss": 0.2588,
+      "epoch": 3.9798897178073305,
+      "grad_norm": 0.29807037115097046,
+      "learning_rate": 5.27544996919635e-06,
+      "loss": 0.3851,
       "step": 110430
     },
     {
-      "epoch": 3.89,
-      "learning_rate": 6.251847066273686e-06,
-      "loss": 0.2479,
+      "epoch": 3.980069917468555,
+      "grad_norm": 0.2329607754945755,
+      "learning_rate": 5.273657154662626e-06,
+      "loss": 0.3944,
       "step": 110435
     },
     {
-      "epoch": 3.89,
-      "learning_rate": 6.249962697331968e-06,
-      "loss": 0.2687,
+      "epoch": 3.98025011712978,
+      "grad_norm": 0.2554316520690918,
+      "learning_rate": 5.271864608896501e-06,
+      "loss": 0.4016,
       "step": 110440
     },
     {
-      "epoch": 3.89,
-      "learning_rate": 6.248078571847133e-06,
-      "loss": 0.2849,
+      "epoch": 3.9804303167910042,
+      "grad_norm": 0.2084946185350418,
+      "learning_rate": 5.270072331922405e-06,
+      "loss": 0.3763,
       "step": 110445
     },
     {
-      "epoch": 3.89,
-      "learning_rate": 6.24619468984364e-06,
-      "loss": 0.2562,
+      "epoch": 3.980610516452229,
+      "grad_norm": 0.2905483841896057,
+      "learning_rate": 5.268280323764744e-06,
+      "loss": 0.3805,
       "step": 110450
     },
     {
-      "epoch": 3.89,
-      "learning_rate": 6.244311051345941e-06,
-      "loss": 0.2657,
+      "epoch": 3.9807907161134537,
+      "grad_norm": 0.2571609616279602,
+      "learning_rate": 5.266488584447949e-06,
+      "loss": 0.3814,
       "step": 110455
     },
     {
-      "epoch": 3.89,
-      "learning_rate": 6.242427656378514e-06,
-      "loss": 0.2758,
+      "epoch": 3.9809709157746784,
+      "grad_norm": 0.25238341093063354,
+      "learning_rate": 5.264697113996428e-06,
+      "loss": 0.396,
       "step": 110460
     },
     {
-      "epoch": 3.89,
-      "learning_rate": 6.240544504965798e-06,
-      "loss": 0.2533,
+      "epoch": 3.981151115435903,
+      "grad_norm": 0.21064473688602448,
+      "learning_rate": 5.2629059124345875e-06,
+      "loss": 0.3894,
       "step": 110465
     },
     {
-      "epoch": 3.89,
-      "learning_rate": 6.23866159713225e-06,
-      "loss": 0.2474,
+      "epoch": 3.9813313150971275,
+      "grad_norm": 0.19198505580425262,
+      "learning_rate": 5.261114979786832e-06,
+      "loss": 0.3529,
       "step": 110470
     },
     {
-      "epoch": 3.89,
-      "learning_rate": 6.23677893290231e-06,
-      "loss": 0.2443,
+      "epoch": 3.981511514758352,
+      "grad_norm": 0.20793737471103668,
+      "learning_rate": 5.259324316077561e-06,
+      "loss": 0.3754,
       "step": 110475
     },
     {
-      "epoch": 3.89,
-      "learning_rate": 6.234896512300437e-06,
-      "loss": 0.2682,
+      "epoch": 3.981691714419577,
+      "grad_norm": 0.2338850349187851,
+      "learning_rate": 5.257533921331176e-06,
+      "loss": 0.3651,
       "step": 110480
     },
     {
-      "epoch": 3.89,
-      "learning_rate": 6.233014335351059e-06,
-      "loss": 0.2568,
+      "epoch": 3.9818719140808017,
+      "grad_norm": 0.2929813265800476,
+      "learning_rate": 5.255743795572071e-06,
+      "loss": 0.3788,
       "step": 110485
     },
     {
-      "epoch": 3.89,
-      "learning_rate": 6.231132402078629e-06,
-      "loss": 0.2572,
+      "epoch": 3.982052113742026,
+      "grad_norm": 0.22587592899799347,
+      "learning_rate": 5.253953938824635e-06,
+      "loss": 0.3693,
       "step": 110490
     },
     {
-      "epoch": 3.89,
-      "learning_rate": 6.22925071250757e-06,
-      "loss": 0.262,
+      "epoch": 3.9822323134032507,
+      "grad_norm": 0.25955894589424133,
+      "learning_rate": 5.252164351113254e-06,
+      "loss": 0.3895,
       "step": 110495
     },
     {
-      "epoch": 3.89,
-      "learning_rate": 6.2273692666623255e-06,
-      "loss": 0.269,
+      "epoch": 3.9824125130644754,
+      "grad_norm": 0.22721461951732635,
+      "learning_rate": 5.250375032462307e-06,
+      "loss": 0.3641,
       "step": 110500
     },
     {
-      "epoch": 3.89,
-      "eval_loss": 0.25054579973220825,
-      "eval_runtime": 10.5376,
-      "eval_samples_per_second": 9.49,
-      "eval_steps_per_second": 9.49,
+      "epoch": 3.9824125130644754,
+      "eval_loss": 0.4280896484851837,
+      "eval_runtime": 3.5308,
+      "eval_samples_per_second": 28.322,
+      "eval_steps_per_second": 7.081,
       "step": 110500
     },
     {
-      "epoch": 3.89,
-      "learning_rate": 6.225488064567322e-06,
-      "loss": 0.2626,
+      "epoch": 3.9825927127257,
+      "grad_norm": 0.21194064617156982,
+      "learning_rate": 5.248585982896173e-06,
+      "loss": 0.3549,
       "step": 110505
     },
     {
-      "epoch": 3.89,
-      "learning_rate": 6.223607106246981e-06,
-      "loss": 0.2529,
+      "epoch": 3.982772912386925,
+      "grad_norm": 0.24252504110336304,
+      "learning_rate": 5.2467972024392435e-06,
+      "loss": 0.3785,
       "step": 110510
     },
     {
-      "epoch": 3.89,
-      "learning_rate": 6.2217263917257215e-06,
-      "loss": 0.2658,
+      "epoch": 3.9829531120481496,
+      "grad_norm": 0.28674301505088806,
+      "learning_rate": 5.245008691115863e-06,
+      "loss": 0.3665,
       "step": 110515
     },
     {
-      "epoch": 3.89,
-      "learning_rate": 6.219845921027975e-06,
-      "loss": 0.2449,
+      "epoch": 3.983133311709374,
+      "grad_norm": 0.25369998812675476,
+      "learning_rate": 5.243220448950423e-06,
+      "loss": 0.3448,
       "step": 110520
     },
     {
-      "epoch": 3.89,
-      "learning_rate": 6.217965694178152e-06,
-      "loss": 0.263,
+      "epoch": 3.9833135113705986,
+      "grad_norm": 0.19632992148399353,
+      "learning_rate": 5.241432475967278e-06,
+      "loss": 0.3655,
       "step": 110525
     },
     {
-      "epoch": 3.89,
-      "learning_rate": 6.21608571120067e-06,
-      "loss": 0.2426,
+      "epoch": 3.9834937110318234,
+      "grad_norm": 0.2441524714231491,
+      "learning_rate": 5.239644772190791e-06,
+      "loss": 0.3927,
       "step": 110530
     },
     {
-      "epoch": 3.89,
-      "learning_rate": 6.214205972119924e-06,
-      "loss": 0.2637,
+      "epoch": 3.9836739106930477,
+      "grad_norm": 0.2212579846382141,
+      "learning_rate": 5.237857337645319e-06,
+      "loss": 0.3874,
       "step": 110535
     },
     {
-      "epoch": 3.89,
-      "learning_rate": 6.212326476960345e-06,
-      "loss": 0.267,
+      "epoch": 3.9838541103542724,
+      "grad_norm": 0.2771473228931427,
+      "learning_rate": 5.236070172355212e-06,
+      "loss": 0.3673,
       "step": 110540
     },
     {
-      "epoch": 3.89,
-      "learning_rate": 6.210447225746316e-06,
-      "loss": 0.2568,
+      "epoch": 3.984034310015497,
+      "grad_norm": 0.2671329379081726,
+      "learning_rate": 5.234283276344818e-06,
+      "loss": 0.3517,
       "step": 110545
     },
     {
-      "epoch": 3.89,
-      "learning_rate": 6.208568218502253e-06,
-      "loss": 0.2487,
+      "epoch": 3.984214509676722,
+      "grad_norm": 0.20370586216449738,
+      "learning_rate": 5.232496649638494e-06,
+      "loss": 0.3667,
       "step": 110550
     },
     {
-      "epoch": 3.89,
-      "learning_rate": 6.206689455252543e-06,
-      "loss": 0.2381,
+      "epoch": 3.9843947093379466,
+      "grad_norm": 0.25102975964546204,
+      "learning_rate": 5.230710292260576e-06,
+      "loss": 0.382,
       "step": 110555
     },
     {
-      "epoch": 3.89,
-      "learning_rate": 6.20481093602159e-06,
-      "loss": 0.2342,
+      "epoch": 3.9845749089991713,
+      "grad_norm": 0.2473146915435791,
+      "learning_rate": 5.2289242042354025e-06,
+      "loss": 0.4072,
       "step": 110560
     },
     {
-      "epoch": 3.89,
-      "learning_rate": 6.202932660833783e-06,
-      "loss": 0.2762,
+      "epoch": 3.9847551086603956,
+      "grad_norm": 0.2885691523551941,
+      "learning_rate": 5.22713838558731e-06,
+      "loss": 0.3524,
       "step": 110565
     },
     {
-      "epoch": 3.89,
-      "learning_rate": 6.201054629713507e-06,
-      "loss": 0.2464,
+      "epoch": 3.9849353083216204,
+      "grad_norm": 0.2779085338115692,
+      "learning_rate": 5.225352836340622e-06,
+      "loss": 0.3521,
       "step": 110570
     },
     {
-      "epoch": 3.89,
-      "learning_rate": 6.199176842685139e-06,
-      "loss": 0.2611,
+      "epoch": 3.985115507982845,
+      "grad_norm": 0.2686401903629303,
+      "learning_rate": 5.223567556519679e-06,
+      "loss": 0.3986,
       "step": 110575
     },
     {
-      "epoch": 3.89,
-      "learning_rate": 6.197299299773077e-06,
-      "loss": 0.2453,
+      "epoch": 3.9852957076440694,
+      "grad_norm": 0.20243895053863525,
+      "learning_rate": 5.221782546148804e-06,
+      "loss": 0.358,
       "step": 110580
     },
     {
-      "epoch": 3.89,
-      "learning_rate": 6.195422001001694e-06,
-      "loss": 0.2531,
+      "epoch": 3.985475907305294,
+      "grad_norm": 0.2190602421760559,
+      "learning_rate": 5.219997805252303e-06,
+      "loss": 0.3865,
       "step": 110585
     },
     {
-      "epoch": 3.89,
-      "learning_rate": 6.193544946395355e-06,
-      "loss": 0.2641,
+      "epoch": 3.985656106966519,
+      "grad_norm": 0.2563530504703522,
+      "learning_rate": 5.2182133338545074e-06,
+      "loss": 0.3668,
       "step": 110590
     },
     {
-      "epoch": 3.89,
-      "learning_rate": 6.191668135978448e-06,
-      "loss": 0.2675,
+      "epoch": 3.9858363066277436,
+      "grad_norm": 0.26702284812927246,
+      "learning_rate": 5.216429131979717e-06,
+      "loss": 0.3666,
       "step": 110595
     },
     {
-      "epoch": 3.89,
-      "learning_rate": 6.189791569775327e-06,
-      "loss": 0.2526,
+      "epoch": 3.9860165062889683,
+      "grad_norm": 0.243973970413208,
+      "learning_rate": 5.214645199652257e-06,
+      "loss": 0.3657,
       "step": 110600
     },
     {
-      "epoch": 3.89,
-      "learning_rate": 6.187915247810375e-06,
-      "loss": 0.2349,
+      "epoch": 3.986196705950193,
+      "grad_norm": 0.22171549499034882,
+      "learning_rate": 5.212861536896435e-06,
+      "loss": 0.4154,
       "step": 110605
     },
     {
-      "epoch": 3.89,
-      "learning_rate": 6.1860391701079376e-06,
-      "loss": 0.253,
+      "epoch": 3.9863769056114173,
+      "grad_norm": 0.224653422832489,
+      "learning_rate": 5.211078143736528e-06,
+      "loss": 0.3901,
       "step": 110610
     },
     {
-      "epoch": 3.89,
-      "learning_rate": 6.184163336692392e-06,
-      "loss": 0.2703,
+      "epoch": 3.986557105272642,
+      "grad_norm": 0.2409631460905075,
+      "learning_rate": 5.209295020196855e-06,
+      "loss": 0.3659,
       "step": 110615
     },
     {
-      "epoch": 3.89,
-      "learning_rate": 6.182287747588084e-06,
-      "loss": 0.2514,
+      "epoch": 3.986737304933867,
+      "grad_norm": 0.21369391679763794,
+      "learning_rate": 5.207512166301709e-06,
+      "loss": 0.3474,
       "step": 110620
     },
     {
-      "epoch": 3.89,
-      "learning_rate": 6.180412402819366e-06,
-      "loss": 0.2467,
+      "epoch": 3.986917504595091,
+      "grad_norm": 0.24725468456745148,
+      "learning_rate": 5.205729582075375e-06,
+      "loss": 0.3711,
       "step": 110625
     },
     {
-      "epoch": 3.89,
-      "learning_rate": 6.178537302410587e-06,
-      "loss": 0.2518,
+      "epoch": 3.987097704256316,
+      "grad_norm": 0.23448602855205536,
+      "learning_rate": 5.203947267542145e-06,
+      "loss": 0.3823,
       "step": 110630
     },
     {
-      "epoch": 3.89,
-      "learning_rate": 6.1766624463861e-06,
-      "loss": 0.2674,
+      "epoch": 3.9872779039175406,
+      "grad_norm": 0.20930229127407074,
+      "learning_rate": 5.202165222726294e-06,
+      "loss": 0.3424,
       "step": 110635
     },
     {
-      "epoch": 3.89,
-      "learning_rate": 6.174787834770251e-06,
-      "loss": 0.2545,
+      "epoch": 3.9874581035787653,
+      "grad_norm": 0.2615196406841278,
+      "learning_rate": 5.200383447652115e-06,
+      "loss": 0.3677,
       "step": 110640
     },
     {
-      "epoch": 3.89,
-      "learning_rate": 6.172913467587374e-06,
-      "loss": 0.259,
+      "epoch": 3.98763830323999,
+      "grad_norm": 0.27675536274909973,
+      "learning_rate": 5.198601942343878e-06,
+      "loss": 0.3756,
       "step": 110645
     },
     {
-      "epoch": 3.89,
-      "learning_rate": 6.171039344861798e-06,
-      "loss": 0.2691,
+      "epoch": 3.9878185029012148,
+      "grad_norm": 0.2948794364929199,
+      "learning_rate": 5.196820706825856e-06,
+      "loss": 0.4064,
       "step": 110650
     },
     {
-      "epoch": 3.89,
-      "learning_rate": 6.169165466617871e-06,
-      "loss": 0.2591,
+      "epoch": 3.987998702562439,
+      "grad_norm": 0.2301834374666214,
+      "learning_rate": 5.195039741122318e-06,
+      "loss": 0.337,
       "step": 110655
     },
     {
-      "epoch": 3.89,
-      "learning_rate": 6.167291832879926e-06,
-      "loss": 0.2503,
+      "epoch": 3.988178902223664,
+      "grad_norm": 0.27218741178512573,
+      "learning_rate": 5.193259045257523e-06,
+      "loss": 0.4075,
       "step": 110660
     },
     {
-      "epoch": 3.89,
-      "learning_rate": 6.165418443672285e-06,
-      "loss": 0.2583,
+      "epoch": 3.9883591018848885,
+      "grad_norm": 0.2404523491859436,
+      "learning_rate": 5.191478619255746e-06,
+      "loss": 0.3845,
       "step": 110665
     },
     {
-      "epoch": 3.89,
-      "learning_rate": 6.163545299019266e-06,
-      "loss": 0.2442,
+      "epoch": 3.9885393015461132,
+      "grad_norm": 0.2734452486038208,
+      "learning_rate": 5.189698463141237e-06,
+      "loss": 0.3988,
       "step": 110670
     },
     {
-      "epoch": 3.89,
-      "learning_rate": 6.1616723989452064e-06,
-      "loss": 0.282,
+      "epoch": 3.9887195012073375,
+      "grad_norm": 0.2366897016763687,
+      "learning_rate": 5.18791857693825e-06,
+      "loss": 0.3859,
       "step": 110675
     },
     {
-      "epoch": 3.89,
-      "learning_rate": 6.159799743474415e-06,
-      "loss": 0.2496,
+      "epoch": 3.9888997008685623,
+      "grad_norm": 0.22594159841537476,
+      "learning_rate": 5.186138960671039e-06,
+      "loss": 0.3872,
       "step": 110680
     },
     {
-      "epoch": 3.89,
-      "learning_rate": 6.1579273326312085e-06,
-      "loss": 0.2592,
+      "epoch": 3.989079900529787,
+      "grad_norm": 0.24368955194950104,
+      "learning_rate": 5.184359614363846e-06,
+      "loss": 0.3196,
       "step": 110685
     },
     {
-      "epoch": 3.89,
-      "learning_rate": 6.156055166439889e-06,
-      "loss": 0.248,
+      "epoch": 3.9892601001910117,
+      "grad_norm": 0.25890570878982544,
+      "learning_rate": 5.18258053804091e-06,
+      "loss": 0.3746,
       "step": 110690
     },
     {
-      "epoch": 3.89,
-      "learning_rate": 6.154183244924783e-06,
-      "loss": 0.271,
+      "epoch": 3.9894402998522365,
+      "grad_norm": 0.23485039174556732,
+      "learning_rate": 5.180801731726493e-06,
+      "loss": 0.364,
       "step": 110695
     },
     {
-      "epoch": 3.89,
-      "learning_rate": 6.152311568110186e-06,
-      "loss": 0.2695,
+      "epoch": 3.9896204995134608,
+      "grad_norm": 0.18861232697963715,
+      "learning_rate": 5.179023195444802e-06,
+      "loss": 0.3839,
       "step": 110700
     },
     {
-      "epoch": 3.89,
-      "learning_rate": 6.150440136020397e-06,
-      "loss": 0.2529,
+      "epoch": 3.9898006991746855,
+      "grad_norm": 0.23490877449512482,
+      "learning_rate": 5.177244929220088e-06,
+      "loss": 0.3705,
       "step": 110705
     },
     {
-      "epoch": 3.9,
-      "learning_rate": 6.14856894867972e-06,
-      "loss": 0.2594,
+      "epoch": 3.9899808988359102,
+      "grad_norm": 0.2072230577468872,
+      "learning_rate": 5.175466933076573e-06,
+      "loss": 0.3797,
       "step": 110710
     },
     {
-      "epoch": 3.9,
-      "learning_rate": 6.1466980061124586e-06,
-      "loss": 0.2701,
+      "epoch": 3.990161098497135,
+      "grad_norm": 0.2669602334499359,
+      "learning_rate": 5.173689207038479e-06,
+      "loss": 0.3559,
       "step": 110715
     },
     {
-      "epoch": 3.9,
-      "learning_rate": 6.144827308342899e-06,
-      "loss": 0.2438,
+      "epoch": 3.9903412981583593,
+      "grad_norm": 0.2188437432050705,
+      "learning_rate": 5.17191175113004e-06,
+      "loss": 0.3732,
       "step": 110720
     },
     {
-      "epoch": 3.9,
-      "learning_rate": 6.142956855395324e-06,
-      "loss": 0.2585,
+      "epoch": 3.990521497819584,
+      "grad_norm": 0.2595865726470947,
+      "learning_rate": 5.1701345653754615e-06,
+      "loss": 0.3975,
       "step": 110725
     },
     {
-      "epoch": 3.9,
-      "learning_rate": 6.141086647294034e-06,
-      "loss": 0.2527,
+      "epoch": 3.9907016974808087,
+      "grad_norm": 0.22355596721172333,
+      "learning_rate": 5.168357649798952e-06,
+      "loss": 0.3346,
       "step": 110730
     },
     {
-      "epoch": 3.9,
-      "learning_rate": 6.139216684063307e-06,
-      "loss": 0.2485,
+      "epoch": 3.9908818971420335,
+      "grad_norm": 0.19183094799518585,
+      "learning_rate": 5.1665810044247395e-06,
+      "loss": 0.3851,
       "step": 110735
     },
     {
-      "epoch": 3.9,
-      "learning_rate": 6.137346965727417e-06,
-      "loss": 0.2606,
+      "epoch": 3.991062096803258,
+      "grad_norm": 0.2600308060646057,
+      "learning_rate": 5.164804629277018e-06,
+      "loss": 0.3942,
       "step": 110740
     },
     {
-      "epoch": 3.9,
-      "learning_rate": 6.135477492310643e-06,
-      "loss": 0.257,
+      "epoch": 3.991242296464483,
+      "grad_norm": 0.2572632431983948,
+      "learning_rate": 5.163028524379995e-06,
+      "loss": 0.4191,
       "step": 110745
     },
     {
-      "epoch": 3.9,
-      "learning_rate": 6.133608263837265e-06,
-      "loss": 0.2391,
+      "epoch": 3.991422496125707,
+      "grad_norm": 0.21313636004924774,
+      "learning_rate": 5.161252689757867e-06,
+      "loss": 0.3822,
       "step": 110750
     },
     {
-      "epoch": 3.9,
-      "learning_rate": 6.131739280331553e-06,
-      "loss": 0.2296,
+      "epoch": 3.991602695786932,
+      "grad_norm": 0.24319961667060852,
+      "learning_rate": 5.159477125434822e-06,
+      "loss": 0.3934,
       "step": 110755
     },
     {
-      "epoch": 3.9,
-      "learning_rate": 6.129870541817765e-06,
-      "loss": 0.2693,
+      "epoch": 3.9917828954481567,
+      "grad_norm": 0.2313625067472458,
+      "learning_rate": 5.157701831435069e-06,
+      "loss": 0.3736,
       "step": 110760
     },
     {
-      "epoch": 3.9,
-      "learning_rate": 6.128002048320169e-06,
-      "loss": 0.2424,
+      "epoch": 3.991963095109381,
+      "grad_norm": 0.2143564522266388,
+      "learning_rate": 5.155926807782785e-06,
+      "loss": 0.3559,
       "step": 110765
     },
     {
-      "epoch": 3.9,
-      "learning_rate": 6.12613379986304e-06,
-      "loss": 0.2491,
+      "epoch": 3.9921432947706057,
+      "grad_norm": 0.28823596239089966,
+      "learning_rate": 5.15415205450216e-06,
+      "loss": 0.3423,
       "step": 110770
     },
     {
-      "epoch": 3.9,
-      "learning_rate": 6.124265796470624e-06,
-      "loss": 0.2508,
+      "epoch": 3.9923234944318304,
+      "grad_norm": 0.21266886591911316,
+      "learning_rate": 5.152377571617368e-06,
+      "loss": 0.367,
       "step": 110775
     },
     {
-      "epoch": 3.9,
-      "learning_rate": 6.122398038167176e-06,
-      "loss": 0.2627,
+      "epoch": 3.992503694093055,
+      "grad_norm": 0.25234442949295044,
+      "learning_rate": 5.150603359152581e-06,
+      "loss": 0.3763,
       "step": 110780
     },
     {
-      "epoch": 3.9,
-      "learning_rate": 6.120530524976942e-06,
-      "loss": 0.2383,
+      "epoch": 3.99268389375428,
+      "grad_norm": 0.2379533350467682,
+      "learning_rate": 5.14882941713199e-06,
+      "loss": 0.3801,
       "step": 110785
     },
     {
-      "epoch": 3.9,
-      "learning_rate": 6.118663256924187e-06,
-      "loss": 0.262,
+      "epoch": 3.9928640934155046,
+      "grad_norm": 0.26140525937080383,
+      "learning_rate": 5.147410458251575e-06,
+      "loss": 0.3709,
       "step": 110790
     },
     {
-      "epoch": 3.9,
-      "learning_rate": 6.116796234033143e-06,
-      "loss": 0.2547,
+      "epoch": 3.993044293076729,
+      "grad_norm": 0.24928440153598785,
+      "learning_rate": 5.1456370030914245e-06,
+      "loss": 0.3754,
       "step": 110795
     },
     {
-      "epoch": 3.9,
-      "learning_rate": 6.114929456328056e-06,
-      "loss": 0.2659,
+      "epoch": 3.9932244927379537,
+      "grad_norm": 0.2500706613063812,
+      "learning_rate": 5.1438638184431175e-06,
+      "loss": 0.3691,
       "step": 110800
     },
     {
-      "epoch": 3.9,
-      "learning_rate": 6.113062923833157e-06,
-      "loss": 0.2456,
+      "epoch": 3.9934046923991784,
+      "grad_norm": 0.2307901233434677,
+      "learning_rate": 5.142090904330829e-06,
+      "loss": 0.3391,
       "step": 110805
     },
     {
-      "epoch": 3.9,
-      "learning_rate": 6.111196636572697e-06,
-      "loss": 0.2617,
+      "epoch": 3.9935848920604027,
+      "grad_norm": 0.2293684333562851,
+      "learning_rate": 5.1403182607787065e-06,
+      "loss": 0.3591,
       "step": 110810
     },
     {
-      "epoch": 3.9,
-      "learning_rate": 6.109330594570892e-06,
-      "loss": 0.2244,
+      "epoch": 3.9937650917216274,
+      "grad_norm": 0.21017733216285706,
+      "learning_rate": 5.138545887810903e-06,
+      "loss": 0.3675,
       "step": 110815
     },
     {
-      "epoch": 3.9,
-      "learning_rate": 6.107464797851989e-06,
-      "loss": 0.2801,
+      "epoch": 3.993945291382852,
+      "grad_norm": 0.26224109530448914,
+      "learning_rate": 5.136773785451565e-06,
+      "loss": 0.3531,
       "step": 110820
     },
     {
-      "epoch": 3.9,
-      "learning_rate": 6.105599246440197e-06,
-      "loss": 0.2624,
+      "epoch": 3.994125491044077,
+      "grad_norm": 0.20869319140911102,
+      "learning_rate": 5.135001953724833e-06,
+      "loss": 0.3558,
       "step": 110825
     },
     {
-      "epoch": 3.9,
-      "learning_rate": 6.103733940359754e-06,
-      "loss": 0.2427,
+      "epoch": 3.9943056907053016,
+      "grad_norm": 0.3143329918384552,
+      "learning_rate": 5.133230392654861e-06,
+      "loss": 0.3956,
       "step": 110830
     },
     {
-      "epoch": 3.9,
-      "learning_rate": 6.101868879634873e-06,
-      "loss": 0.265,
+      "epoch": 3.9944858903665263,
+      "grad_norm": 0.25546544790267944,
+      "learning_rate": 5.131459102265779e-06,
+      "loss": 0.3808,
       "step": 110835
     },
     {
-      "epoch": 3.9,
-      "learning_rate": 6.10000406428976e-06,
-      "loss": 0.2538,
+      "epoch": 3.9946660900277506,
+      "grad_norm": 0.20834296941757202,
+      "learning_rate": 5.1296880825817185e-06,
+      "loss": 0.3834,
       "step": 110840
     },
     {
-      "epoch": 3.9,
-      "learning_rate": 6.098139494348651e-06,
-      "loss": 0.249,
+      "epoch": 3.9948462896889754,
+      "grad_norm": 0.2101668119430542,
+      "learning_rate": 5.127917333626811e-06,
+      "loss": 0.3633,
       "step": 110845
     },
     {
-      "epoch": 3.9,
-      "learning_rate": 6.096275169835742e-06,
-      "loss": 0.2571,
+      "epoch": 3.9950264893502,
+      "grad_norm": 0.2534985840320587,
+      "learning_rate": 5.126146855425176e-06,
+      "loss": 0.3666,
       "step": 110850
     },
     {
-      "epoch": 3.9,
-      "learning_rate": 6.09441109077524e-06,
-      "loss": 0.2541,
+      "epoch": 3.9952066890114244,
+      "grad_norm": 0.1858637034893036,
+      "learning_rate": 5.12437664800095e-06,
+      "loss": 0.3767,
       "step": 110855
     },
     {
-      "epoch": 3.9,
-      "learning_rate": 6.092547257191344e-06,
-      "loss": 0.2631,
+      "epoch": 3.995386888672649,
+      "grad_norm": 0.25924816727638245,
+      "learning_rate": 5.12260671137825e-06,
+      "loss": 0.3932,
       "step": 110860
     },
     {
-      "epoch": 3.9,
-      "learning_rate": 6.0906836691082685e-06,
-      "loss": 0.2344,
+      "epoch": 3.995567088333874,
+      "grad_norm": 0.2486696094274521,
+      "learning_rate": 5.12083704558117e-06,
+      "loss": 0.3783,
       "step": 110865
     },
     {
-      "epoch": 3.9,
-      "learning_rate": 6.088820326550199e-06,
-      "loss": 0.2398,
+      "epoch": 3.9957472879950986,
+      "grad_norm": 0.24414058029651642,
+      "learning_rate": 5.119067650633847e-06,
+      "loss": 0.3898,
       "step": 110870
     },
     {
-      "epoch": 3.9,
-      "learning_rate": 6.0869572295413405e-06,
-      "loss": 0.2803,
+      "epoch": 3.9959274876563233,
+      "grad_norm": 0.22701597213745117,
+      "learning_rate": 5.117298526560374e-06,
+      "loss": 0.3538,
       "step": 110875
     },
     {
-      "epoch": 3.9,
-      "learning_rate": 6.085094378105868e-06,
-      "loss": 0.2657,
+      "epoch": 3.996107687317548,
+      "grad_norm": 0.2523609697818756,
+      "learning_rate": 5.115529673384861e-06,
+      "loss": 0.3745,
       "step": 110880
     },
     {
-      "epoch": 3.9,
-      "learning_rate": 6.083231772267991e-06,
-      "loss": 0.2532,
+      "epoch": 3.9962878869787724,
+      "grad_norm": 0.21400229632854462,
+      "learning_rate": 5.113761091131408e-06,
+      "loss": 0.3759,
       "step": 110885
     },
     {
-      "epoch": 3.9,
-      "learning_rate": 6.081369412051882e-06,
-      "loss": 0.2559,
+      "epoch": 3.996468086639997,
+      "grad_norm": 0.2355312556028366,
+      "learning_rate": 5.111992779824101e-06,
+      "loss": 0.3256,
       "step": 110890
     },
     {
-      "epoch": 3.9,
-      "learning_rate": 6.079507297481723e-06,
-      "loss": 0.2637,
+      "epoch": 3.996648286301222,
+      "grad_norm": 0.231663778424263,
+      "learning_rate": 5.110224739487051e-06,
+      "loss": 0.4152,
       "step": 110895
     },
     {
-      "epoch": 3.9,
-      "learning_rate": 6.0776454285816856e-06,
-      "loss": 0.2566,
+      "epoch": 3.9968284859624466,
+      "grad_norm": 0.22095178067684174,
+      "learning_rate": 5.108456970144338e-06,
+      "loss": 0.3665,
       "step": 110900
     },
     {
-      "epoch": 3.9,
-      "learning_rate": 6.075783805375956e-06,
-      "loss": 0.2534,
+      "epoch": 3.997008685623671,
+      "grad_norm": 0.31155142188072205,
+      "learning_rate": 5.106689471820045e-06,
+      "loss": 0.3996,
       "step": 110905
     },
     {
-      "epoch": 3.9,
-      "learning_rate": 6.073922427888707e-06,
-      "loss": 0.2611,
+      "epoch": 3.9971888852848956,
+      "grad_norm": 0.18749769032001495,
+      "learning_rate": 5.10492224453826e-06,
+      "loss": 0.3583,
       "step": 110910
     },
     {
-      "epoch": 3.9,
-      "learning_rate": 6.072061296144099e-06,
-      "loss": 0.2659,
+      "epoch": 3.9973690849461203,
+      "grad_norm": 0.28036782145500183,
+      "learning_rate": 5.1031552883230475e-06,
+      "loss": 0.3738,
       "step": 110915
     },
     {
-      "epoch": 3.9,
-      "learning_rate": 6.070200410166297e-06,
-      "loss": 0.2797,
+      "epoch": 3.997549284607345,
+      "grad_norm": 0.2641852796077728,
+      "learning_rate": 5.101388603198501e-06,
+      "loss": 0.3823,
       "step": 110920
     },
     {
-      "epoch": 3.9,
-      "learning_rate": 6.068339769979465e-06,
-      "loss": 0.2757,
+      "epoch": 3.9977294842685698,
+      "grad_norm": 0.22281073033809662,
+      "learning_rate": 5.099622189188679e-06,
+      "loss": 0.3982,
       "step": 110925
     },
     {
-      "epoch": 3.9,
-      "learning_rate": 6.066479375607773e-06,
-      "loss": 0.2475,
+      "epoch": 3.997909683929794,
+      "grad_norm": 0.2702135741710663,
+      "learning_rate": 5.097856046317656e-06,
+      "loss": 0.3821,
       "step": 110930
     },
     {
-      "epoch": 3.9,
-      "learning_rate": 6.06461922707536e-06,
-      "loss": 0.2344,
+      "epoch": 3.998089883591019,
+      "grad_norm": 0.25473445653915405,
+      "learning_rate": 5.096090174609489e-06,
+      "loss": 0.3727,
       "step": 110935
     },
     {
-      "epoch": 3.9,
-      "learning_rate": 6.062759324406397e-06,
-      "loss": 0.2698,
+      "epoch": 3.9982700832522435,
+      "grad_norm": 0.2873947024345398,
+      "learning_rate": 5.0943245740882414e-06,
+      "loss": 0.3805,
       "step": 110940
     },
     {
-      "epoch": 3.9,
-      "learning_rate": 6.060899667625023e-06,
-      "loss": 0.2541,
+      "epoch": 3.9984502829134683,
+      "grad_norm": 0.211199089884758,
+      "learning_rate": 5.092559244777958e-06,
+      "loss": 0.358,
       "step": 110945
     },
     {
-      "epoch": 3.9,
-      "learning_rate": 6.059040256755386e-06,
-      "loss": 0.2707,
+      "epoch": 3.9986304825746926,
+      "grad_norm": 0.256854772567749,
+      "learning_rate": 5.090794186702711e-06,
+      "loss": 0.3848,
       "step": 110950
     },
     {
-      "epoch": 3.9,
-      "learning_rate": 6.057181091821621e-06,
-      "loss": 0.2364,
+      "epoch": 3.9988106822359173,
+      "grad_norm": 0.22104725241661072,
+      "learning_rate": 5.0890293998865355e-06,
+      "loss": 0.3522,
       "step": 110955
     },
     {
-      "epoch": 3.9,
-      "learning_rate": 6.055322172847882e-06,
-      "loss": 0.2434,
+      "epoch": 3.998990881897142,
+      "grad_norm": 0.2235444188117981,
+      "learning_rate": 5.087264884353482e-06,
+      "loss": 0.3764,
       "step": 110960
     },
     {
-      "epoch": 3.9,
-      "learning_rate": 6.0534634998583e-06,
-      "loss": 0.2586,
+      "epoch": 3.9991710815583668,
+      "grad_norm": 0.21866776049137115,
+      "learning_rate": 5.085500640127588e-06,
+      "loss": 0.366,
       "step": 110965
     },
     {
-      "epoch": 3.9,
-      "learning_rate": 6.051605072877009e-06,
-      "loss": 0.2427,
+      "epoch": 3.9993512812195915,
+      "grad_norm": 0.2265380173921585,
+      "learning_rate": 5.083736667232886e-06,
+      "loss": 0.3968,
       "step": 110970
     },
     {
-      "epoch": 3.9,
-      "learning_rate": 6.049746891928132e-06,
-      "loss": 0.2499,
+      "epoch": 3.999531480880816,
+      "grad_norm": 0.19983336329460144,
+      "learning_rate": 5.081972965693429e-06,
+      "loss": 0.3938,
       "step": 110975
     },
     {
-      "epoch": 3.9,
-      "learning_rate": 6.047888957035808e-06,
-      "loss": 0.2687,
+      "epoch": 3.9997116805420405,
+      "grad_norm": 0.2481551617383957,
+      "learning_rate": 5.080209535533229e-06,
+      "loss": 0.3913,
       "step": 110980
     },
     {
-      "epoch": 3.9,
-      "learning_rate": 6.04603126822415e-06,
-      "loss": 0.2442,
+      "epoch": 3.9998918802032652,
+      "grad_norm": 0.20951221883296967,
+      "learning_rate": 5.078446376776308e-06,
+      "loss": 0.3391,
       "step": 110985
     },
     {
-      "epoch": 3.9,
-      "learning_rate": 6.0441738255172905e-06,
-      "loss": 0.2566,
+      "epoch": 4.0000720798644895,
+      "grad_norm": 0.25803592801094055,
+      "learning_rate": 5.076683489446707e-06,
+      "loss": 0.3555,
       "step": 110990
     },
     {
-      "epoch": 3.91,
-      "learning_rate": 6.042316628939337e-06,
-      "loss": 0.2521,
+      "epoch": 4.000252279525714,
+      "grad_norm": 0.17808286845684052,
+      "learning_rate": 5.074920873568434e-06,
+      "loss": 0.3527,
       "step": 110995
     },
     {
-      "epoch": 3.91,
-      "learning_rate": 6.0404596785144155e-06,
-      "loss": 0.2526,
+      "epoch": 4.000432479186939,
+      "grad_norm": 0.2654048502445221,
+      "learning_rate": 5.073158529165508e-06,
+      "loss": 0.3552,
       "step": 111000
     },
     {
-      "epoch": 3.91,
-      "eval_loss": 0.25022944808006287,
-      "eval_runtime": 10.5381,
-      "eval_samples_per_second": 9.489,
-      "eval_steps_per_second": 9.489,
+      "epoch": 4.000432479186939,
+      "eval_loss": 0.42817291617393494,
+      "eval_runtime": 3.521,
+      "eval_samples_per_second": 28.401,
+      "eval_steps_per_second": 7.1,
       "step": 111000
     },
     {
-      "epoch": 3.91,
-      "learning_rate": 6.038602974266627e-06,
-      "loss": 0.2202,
+      "epoch": 4.000612678848164,
+      "grad_norm": 0.34447306394577026,
+      "learning_rate": 5.07139645626194e-06,
+      "loss": 0.3778,
       "step": 111005
     },
     {
-      "epoch": 3.91,
-      "learning_rate": 6.036746516220088e-06,
-      "loss": 0.2684,
+      "epoch": 4.0007928785093885,
+      "grad_norm": 0.22575746476650238,
+      "learning_rate": 5.069634654881728e-06,
+      "loss": 0.4257,
       "step": 111010
     },
     {
-      "epoch": 3.91,
-      "learning_rate": 6.034890304398888e-06,
-      "loss": 0.2788,
+      "epoch": 4.000973078170613,
+      "grad_norm": 0.23802292346954346,
+      "learning_rate": 5.067873125048894e-06,
+      "loss": 0.3794,
       "step": 111015
     },
     {
-      "epoch": 3.91,
-      "learning_rate": 6.033034338827146e-06,
-      "loss": 0.2533,
+      "epoch": 4.001153277831838,
+      "grad_norm": 0.20055340230464935,
+      "learning_rate": 5.066111866787429e-06,
+      "loss": 0.3623,
       "step": 111020
     },
     {
-      "epoch": 3.91,
-      "learning_rate": 6.031178619528951e-06,
-      "loss": 0.2626,
+      "epoch": 4.001333477493063,
+      "grad_norm": 0.23983781039714813,
+      "learning_rate": 5.06435088012133e-06,
+      "loss": 0.3751,
       "step": 111025
     },
     {
-      "epoch": 3.91,
-      "learning_rate": 6.0293231465283986e-06,
-      "loss": 0.2392,
+      "epoch": 4.0015136771542865,
+      "grad_norm": 0.27245986461639404,
+      "learning_rate": 5.062590165074591e-06,
+      "loss": 0.3687,
       "step": 111030
     },
     {
-      "epoch": 3.91,
-      "learning_rate": 6.0274679198495856e-06,
-      "loss": 0.236,
+      "epoch": 4.001693876815511,
+      "grad_norm": 0.20769277215003967,
+      "learning_rate": 5.060829721671193e-06,
+      "loss": 0.365,
       "step": 111035
     },
     {
-      "epoch": 3.91,
-      "learning_rate": 6.025612939516595e-06,
-      "loss": 0.2892,
+      "epoch": 4.001874076476736,
+      "grad_norm": 0.2103443592786789,
+      "learning_rate": 5.059069549935139e-06,
+      "loss": 0.349,
       "step": 111040
     },
     {
-      "epoch": 3.91,
-      "learning_rate": 6.0237582055535205e-06,
-      "loss": 0.2554,
+      "epoch": 4.002054276137961,
+      "grad_norm": 0.26758506894111633,
+      "learning_rate": 5.057309649890407e-06,
+      "loss": 0.3724,
       "step": 111045
     },
     {
-      "epoch": 3.91,
-      "learning_rate": 6.021903717984434e-06,
-      "loss": 0.2808,
+      "epoch": 4.0022344757991855,
+      "grad_norm": 0.22477580606937408,
+      "learning_rate": 5.055550021560956e-06,
+      "loss": 0.3396,
       "step": 111050
     },
     {
-      "epoch": 3.91,
-      "learning_rate": 6.020049476833428e-06,
-      "loss": 0.2824,
+      "epoch": 4.00241467546041,
+      "grad_norm": 0.2671130299568176,
+      "learning_rate": 5.053790664970781e-06,
+      "loss": 0.3947,
       "step": 111055
     },
     {
-      "epoch": 3.91,
-      "learning_rate": 6.0181954821245695e-06,
-      "loss": 0.2596,
+      "epoch": 4.002594875121635,
+      "grad_norm": 0.24974533915519714,
+      "learning_rate": 5.052031580143848e-06,
+      "loss": 0.3721,
       "step": 111060
     },
     {
-      "epoch": 3.91,
-      "learning_rate": 6.016341733881936e-06,
-      "loss": 0.2509,
+      "epoch": 4.00277507478286,
+      "grad_norm": 0.2837058901786804,
+      "learning_rate": 5.050272767104114e-06,
+      "loss": 0.3729,
       "step": 111065
     },
     {
-      "epoch": 3.91,
-      "learning_rate": 6.014488232129584e-06,
-      "loss": 0.2612,
+      "epoch": 4.002955274444084,
+      "grad_norm": 0.29399046301841736,
+      "learning_rate": 5.048514225875567e-06,
+      "loss": 0.4003,
       "step": 111070
     },
     {
-      "epoch": 3.91,
-      "learning_rate": 6.012634976891598e-06,
-      "loss": 0.2241,
+      "epoch": 4.003135474105309,
+      "grad_norm": 0.2518676221370697,
+      "learning_rate": 5.046755956482135e-06,
+      "loss": 0.386,
       "step": 111075
     },
     {
-      "epoch": 3.91,
-      "learning_rate": 6.010781968192037e-06,
-      "loss": 0.2892,
+      "epoch": 4.003315673766533,
+      "grad_norm": 0.3091764748096466,
+      "learning_rate": 5.044997958947801e-06,
+      "loss": 0.3661,
       "step": 111080
     },
     {
-      "epoch": 3.91,
-      "learning_rate": 6.008929206054956e-06,
-      "loss": 0.2441,
+      "epoch": 4.003495873427758,
+      "grad_norm": 0.24037063121795654,
+      "learning_rate": 5.0432402332965005e-06,
+      "loss": 0.3483,
       "step": 111085
     },
     {
-      "epoch": 3.91,
-      "learning_rate": 6.007076690504404e-06,
-      "loss": 0.2606,
+      "epoch": 4.003676073088982,
+      "grad_norm": 0.2921180725097656,
+      "learning_rate": 5.041482779552192e-06,
+      "loss": 0.3889,
       "step": 111090
     },
     {
-      "epoch": 3.91,
-      "learning_rate": 6.0052244215644484e-06,
-      "loss": 0.2838,
+      "epoch": 4.003856272750207,
+      "grad_norm": 0.20005546510219574,
+      "learning_rate": 5.039725597738815e-06,
+      "loss": 0.3702,
       "step": 111095
     },
     {
-      "epoch": 3.91,
-      "learning_rate": 6.003372399259138e-06,
-      "loss": 0.2421,
+      "epoch": 4.004036472411432,
+      "grad_norm": 0.25480973720550537,
+      "learning_rate": 5.037968687880306e-06,
+      "loss": 0.363,
       "step": 111100
     },
     {
-      "epoch": 3.91,
-      "learning_rate": 6.001520623612522e-06,
-      "loss": 0.2662,
+      "epoch": 4.004216672072657,
+      "grad_norm": 0.2529853284358978,
+      "learning_rate": 5.036212050000616e-06,
+      "loss": 0.3698,
       "step": 111105
     },
     {
-      "epoch": 3.91,
-      "learning_rate": 5.999669094648633e-06,
-      "loss": 0.2496,
+      "epoch": 4.004396871733881,
+      "grad_norm": 0.2999667525291443,
+      "learning_rate": 5.034455684123673e-06,
+      "loss": 0.3507,
       "step": 111110
     },
     {
-      "epoch": 3.91,
-      "learning_rate": 5.997817812391523e-06,
-      "loss": 0.2536,
+      "epoch": 4.004577071395106,
+      "grad_norm": 0.21827971935272217,
+      "learning_rate": 5.032699590273404e-06,
+      "loss": 0.3973,
       "step": 111115
     },
     {
-      "epoch": 3.91,
-      "learning_rate": 5.995966776865228e-06,
-      "loss": 0.2571,
+      "epoch": 4.004757271056331,
+      "grad_norm": 0.2813790440559387,
+      "learning_rate": 5.030943768473736e-06,
+      "loss": 0.3708,
       "step": 111120
     },
     {
-      "epoch": 3.91,
-      "learning_rate": 5.994115988093782e-06,
-      "loss": 0.2573,
+      "epoch": 4.004937470717555,
+      "grad_norm": 0.28172221779823303,
+      "learning_rate": 5.029188218748595e-06,
+      "loss": 0.38,
       "step": 111125
     },
     {
-      "epoch": 3.91,
-      "learning_rate": 5.992265446101203e-06,
-      "loss": 0.2578,
+      "epoch": 4.005117670378779,
+      "grad_norm": 0.21738965809345245,
+      "learning_rate": 5.027432941121893e-06,
+      "loss": 0.3671,
       "step": 111130
     },
     {
-      "epoch": 3.91,
-      "learning_rate": 5.990415150911541e-06,
-      "loss": 0.297,
+      "epoch": 4.005297870040004,
+      "grad_norm": 0.24970416724681854,
+      "learning_rate": 5.025677935617554e-06,
+      "loss": 0.3419,
       "step": 111135
     },
     {
-      "epoch": 3.91,
-      "learning_rate": 5.98856510254881e-06,
-      "loss": 0.2555,
+      "epoch": 4.005478069701229,
+      "grad_norm": Infinity,
+      "learning_rate": 5.024274127158257e-06,
+      "loss": 0.3606,
       "step": 111140
     },
     {
-      "epoch": 3.91,
-      "learning_rate": 5.986715301037024e-06,
-      "loss": 0.2434,
+      "epoch": 4.005658269362454,
+      "grad_norm": 0.289569228887558,
+      "learning_rate": 5.0225196115344124e-06,
+      "loss": 0.3755,
       "step": 111145
     },
     {
-      "epoch": 3.91,
-      "learning_rate": 5.984865746400211e-06,
-      "loss": 0.266,
+      "epoch": 4.005838469023678,
+      "grad_norm": 0.2360028475522995,
+      "learning_rate": 5.020765368099878e-06,
+      "loss": 0.3216,
       "step": 111150
     },
     {
-      "epoch": 3.91,
-      "learning_rate": 5.983016438662392e-06,
-      "loss": 0.2509,
+      "epoch": 4.006018668684903,
+      "grad_norm": 0.26487815380096436,
+      "learning_rate": 5.019011396878548e-06,
+      "loss": 0.3865,
       "step": 111155
     },
     {
-      "epoch": 3.91,
-      "learning_rate": 5.981167377847571e-06,
-      "loss": 0.2096,
+      "epoch": 4.006198868346128,
+      "grad_norm": 0.19243910908699036,
+      "learning_rate": 5.017257697894321e-06,
+      "loss": 0.3547,
       "step": 111160
     },
     {
-      "epoch": 3.91,
-      "learning_rate": 5.97931856397975e-06,
-      "loss": 0.2772,
+      "epoch": 4.0063790680073526,
+      "grad_norm": 0.2855333983898163,
+      "learning_rate": 5.015504271171087e-06,
+      "loss": 0.3976,
       "step": 111165
     },
     {
-      "epoch": 3.91,
-      "learning_rate": 5.977469997082949e-06,
-      "loss": 0.2639,
+      "epoch": 4.006559267668576,
+      "grad_norm": 0.2589762508869171,
+      "learning_rate": 5.013751116732732e-06,
+      "loss": 0.3747,
       "step": 111170
     },
     {
-      "epoch": 3.91,
-      "learning_rate": 5.975621677181165e-06,
-      "loss": 0.2814,
+      "epoch": 4.006739467329801,
+      "grad_norm": 0.22709418833255768,
+      "learning_rate": 5.011998234603155e-06,
+      "loss": 0.357,
       "step": 111175
     },
     {
-      "epoch": 3.91,
-      "learning_rate": 5.973773604298397e-06,
-      "loss": 0.2605,
+      "epoch": 4.006919666991026,
+      "grad_norm": 0.2505840063095093,
+      "learning_rate": 5.010245624806232e-06,
+      "loss": 0.3812,
       "step": 111180
     },
     {
-      "epoch": 3.91,
-      "learning_rate": 5.9719257784586316e-06,
-      "loss": 0.265,
+      "epoch": 4.007099866652251,
+      "grad_norm": 0.23559889197349548,
+      "learning_rate": 5.008493287365842e-06,
+      "loss": 0.3925,
       "step": 111185
     },
     {
-      "epoch": 3.91,
-      "learning_rate": 5.970078199685877e-06,
-      "loss": 0.2361,
+      "epoch": 4.007280066313475,
+      "grad_norm": 0.22356297075748444,
+      "learning_rate": 5.006741222305861e-06,
+      "loss": 0.3622,
       "step": 111190
     },
     {
-      "epoch": 3.91,
-      "learning_rate": 5.968230868004115e-06,
-      "loss": 0.258,
+      "epoch": 4.0074602659747,
+      "grad_norm": 0.1752263903617859,
+      "learning_rate": 5.004989429650161e-06,
+      "loss": 0.3287,
       "step": 111195
     },
     {
-      "epoch": 3.91,
-      "learning_rate": 5.966383783437329e-06,
-      "loss": 0.2375,
+      "epoch": 4.007640465635925,
+      "grad_norm": 0.21405918896198273,
+      "learning_rate": 5.003237909422601e-06,
+      "loss": 0.3745,
       "step": 111200
     },
     {
-      "epoch": 3.91,
-      "learning_rate": 5.964536946009502e-06,
-      "loss": 0.241,
+      "epoch": 4.0078206652971495,
+      "grad_norm": 0.27426615357398987,
+      "learning_rate": 5.001486661647059e-06,
+      "loss": 0.401,
       "step": 111205
     },
     {
-      "epoch": 3.91,
-      "learning_rate": 5.962690355744627e-06,
-      "loss": 0.2537,
+      "epoch": 4.008000864958374,
+      "grad_norm": 0.21843719482421875,
+      "learning_rate": 4.99973568634739e-06,
+      "loss": 0.3893,
       "step": 111210
     },
     {
-      "epoch": 3.91,
-      "learning_rate": 5.960844012666672e-06,
-      "loss": 0.2603,
+      "epoch": 4.008181064619598,
+      "grad_norm": 0.20329011976718903,
+      "learning_rate": 4.9979849835474515e-06,
+      "loss": 0.3976,
       "step": 111215
     },
     {
-      "epoch": 3.91,
-      "learning_rate": 5.958997916799611e-06,
-      "loss": 0.2443,
+      "epoch": 4.008361264280823,
+      "grad_norm": 0.21530331671237946,
+      "learning_rate": 4.996234553271092e-06,
+      "loss": 0.3503,
       "step": 111220
     },
     {
-      "epoch": 3.91,
-      "learning_rate": 5.957152068167404e-06,
-      "loss": 0.2563,
+      "epoch": 4.008541463942048,
+      "grad_norm": 0.23485726118087769,
+      "learning_rate": 4.994484395542159e-06,
+      "loss": 0.3453,
       "step": 111225
     },
     {
-      "epoch": 3.91,
-      "learning_rate": 5.9553064667940366e-06,
-      "loss": 0.2483,
+      "epoch": 4.008721663603272,
+      "grad_norm": 0.22897350788116455,
+      "learning_rate": 4.992734510384511e-06,
+      "loss": 0.3664,
       "step": 111230
     },
     {
-      "epoch": 3.91,
-      "learning_rate": 5.953461112703462e-06,
-      "loss": 0.2582,
+      "epoch": 4.008901863264497,
+      "grad_norm": 0.2363205850124359,
+      "learning_rate": 4.990984897821988e-06,
+      "loss": 0.3858,
       "step": 111235
     },
     {
-      "epoch": 3.91,
-      "learning_rate": 5.951616005919644e-06,
-      "loss": 0.2488,
+      "epoch": 4.009082062925722,
+      "grad_norm": 0.2695505917072296,
+      "learning_rate": 4.989235557878408e-06,
+      "loss": 0.3598,
       "step": 111240
     },
     {
-      "epoch": 3.91,
-      "learning_rate": 5.949771146466532e-06,
-      "loss": 0.2573,
+      "epoch": 4.0092622625869465,
+      "grad_norm": 0.22455400228500366,
+      "learning_rate": 4.987486490577626e-06,
+      "loss": 0.3521,
       "step": 111245
     },
     {
-      "epoch": 3.91,
-      "learning_rate": 5.947926534368093e-06,
-      "loss": 0.2767,
+      "epoch": 4.009442462248171,
+      "grad_norm": 0.23414896428585052,
+      "learning_rate": 4.985737695943457e-06,
+      "loss": 0.3757,
       "step": 111250
     },
     {
-      "epoch": 3.91,
-      "learning_rate": 5.9460821696482645e-06,
-      "loss": 0.2423,
+      "epoch": 4.009622661909396,
+      "grad_norm": 0.26559486985206604,
+      "learning_rate": 4.9839891739997535e-06,
+      "loss": 0.3847,
       "step": 111255
     },
     {
-      "epoch": 3.91,
-      "learning_rate": 5.944238052331011e-06,
-      "loss": 0.2467,
+      "epoch": 4.00980286157062,
+      "grad_norm": 0.2793545126914978,
+      "learning_rate": 4.982240924770315e-06,
+      "loss": 0.4138,
       "step": 111260
     },
     {
-      "epoch": 3.91,
-      "learning_rate": 5.942394182440261e-06,
-      "loss": 0.2543,
+      "epoch": 4.009983061231845,
+      "grad_norm": 0.28454217314720154,
+      "learning_rate": 4.980492948278961e-06,
+      "loss": 0.3402,
       "step": 111265
     },
     {
-      "epoch": 3.91,
-      "learning_rate": 5.940550559999969e-06,
-      "loss": 0.2854,
+      "epoch": 4.010163260893069,
+      "grad_norm": 0.1981167495250702,
+      "learning_rate": 4.978745244549521e-06,
+      "loss": 0.3759,
       "step": 111270
     },
     {
-      "epoch": 3.91,
-      "learning_rate": 5.938707185034067e-06,
-      "loss": 0.2625,
+      "epoch": 4.010343460554294,
+      "grad_norm": 0.2657647132873535,
+      "learning_rate": 4.976997813605802e-06,
+      "loss": 0.3501,
       "step": 111275
     },
     {
-      "epoch": 3.92,
-      "learning_rate": 5.936864057566485e-06,
-      "loss": 0.255,
+      "epoch": 4.010523660215519,
+      "grad_norm": 0.20731034874916077,
+      "learning_rate": 4.975250655471611e-06,
+      "loss": 0.3307,
       "step": 111280
     },
     {
-      "epoch": 3.92,
-      "learning_rate": 5.9350211776211665e-06,
-      "loss": 0.2553,
+      "epoch": 4.0107038598767435,
+      "grad_norm": 0.3427305221557617,
+      "learning_rate": 4.973503770170751e-06,
+      "loss": 0.3818,
       "step": 111285
     },
     {
-      "epoch": 3.92,
-      "learning_rate": 5.933178545222034e-06,
-      "loss": 0.2503,
+      "epoch": 4.010884059537968,
+      "grad_norm": 0.2292642891407013,
+      "learning_rate": 4.971757157727019e-06,
+      "loss": 0.3537,
       "step": 111290
     },
     {
-      "epoch": 3.92,
-      "learning_rate": 5.931336160393014e-06,
-      "loss": 0.2601,
+      "epoch": 4.011064259199193,
+      "grad_norm": 0.2728843092918396,
+      "learning_rate": 4.970010818164225e-06,
+      "loss": 0.3926,
       "step": 111295
     },
     {
-      "epoch": 3.92,
-      "learning_rate": 5.929494023158016e-06,
-      "loss": 0.2451,
+      "epoch": 4.011244458860418,
+      "grad_norm": 0.2139047533273697,
+      "learning_rate": 4.968264751506157e-06,
+      "loss": 0.3357,
       "step": 111300
     },
     {
-      "epoch": 3.92,
-      "learning_rate": 5.927652133540981e-06,
-      "loss": 0.2553,
+      "epoch": 4.0114246585216415,
+      "grad_norm": 0.2937036156654358,
+      "learning_rate": 4.966518957776603e-06,
+      "loss": 0.3816,
       "step": 111305
     },
     {
-      "epoch": 3.92,
-      "learning_rate": 5.9258104915658054e-06,
-      "loss": 0.2729,
+      "epoch": 4.011604858182866,
+      "grad_norm": 0.31727510690689087,
+      "learning_rate": 4.964773436999348e-06,
+      "loss": 0.3726,
       "step": 111310
     },
     {
-      "epoch": 3.92,
-      "learning_rate": 5.9239690972564184e-06,
-      "loss": 0.2575,
+      "epoch": 4.011785057844091,
+      "grad_norm": 0.2226351499557495,
+      "learning_rate": 4.96302818919818e-06,
+      "loss": 0.3629,
       "step": 111315
     },
     {
-      "epoch": 3.92,
-      "learning_rate": 5.922127950636711e-06,
-      "loss": 0.2593,
+      "epoch": 4.011965257505316,
+      "grad_norm": 0.28360915184020996,
+      "learning_rate": 4.961283214396864e-06,
+      "loss": 0.3353,
       "step": 111320
     },
     {
-      "epoch": 3.92,
-      "learning_rate": 5.920287051730608e-06,
-      "loss": 0.2671,
+      "epoch": 4.0121454571665405,
+      "grad_norm": 0.25234130024909973,
+      "learning_rate": 4.959538512619197e-06,
+      "loss": 0.3666,
       "step": 111325
     },
     {
-      "epoch": 3.92,
-      "learning_rate": 5.918446400562005e-06,
-      "loss": 0.2497,
+      "epoch": 4.012325656827765,
+      "grad_norm": 0.357035368680954,
+      "learning_rate": 4.9577940838889255e-06,
+      "loss": 0.3633,
       "step": 111330
     },
     {
-      "epoch": 3.92,
-      "learning_rate": 5.9166059971548e-06,
-      "loss": 0.2913,
+      "epoch": 4.01250585648899,
+      "grad_norm": 0.26695460081100464,
+      "learning_rate": 4.956049928229836e-06,
+      "loss": 0.371,
       "step": 111335
     },
     {
-      "epoch": 3.92,
-      "learning_rate": 5.914765841532883e-06,
-      "loss": 0.2321,
+      "epoch": 4.012686056150215,
+      "grad_norm": 0.24045346677303314,
+      "learning_rate": 4.954306045665686e-06,
+      "loss": 0.384,
       "step": 111340
     },
     {
-      "epoch": 3.92,
-      "learning_rate": 5.912925933720159e-06,
-      "loss": 0.2588,
+      "epoch": 4.012866255811439,
+      "grad_norm": 0.22014078497886658,
+      "learning_rate": 4.952562436220234e-06,
+      "loss": 0.3773,
       "step": 111345
     },
     {
-      "epoch": 3.92,
-      "learning_rate": 5.911086273740512e-06,
-      "loss": 0.2605,
+      "epoch": 4.013046455472664,
+      "grad_norm": 0.22839348018169403,
+      "learning_rate": 4.950819099917239e-06,
+      "loss": 0.3454,
       "step": 111350
     },
     {
-      "epoch": 3.92,
-      "learning_rate": 5.909246861617834e-06,
-      "loss": 0.2494,
+      "epoch": 4.013226655133888,
+      "grad_norm": 0.2525436282157898,
+      "learning_rate": 4.949076036780445e-06,
+      "loss": 0.3878,
       "step": 111355
     },
     {
-      "epoch": 3.92,
-      "learning_rate": 5.907407697375994e-06,
-      "loss": 0.2646,
+      "epoch": 4.013406854795113,
+      "grad_norm": 0.23207542300224304,
+      "learning_rate": 4.947333246833613e-06,
+      "loss": 0.3541,
       "step": 111360
     },
     {
-      "epoch": 3.92,
-      "learning_rate": 5.905568781038884e-06,
-      "loss": 0.2306,
+      "epoch": 4.0135870544563375,
+      "grad_norm": 0.2273051142692566,
+      "learning_rate": 4.945590730100486e-06,
+      "loss": 0.3643,
       "step": 111365
     },
     {
-      "epoch": 3.92,
-      "learning_rate": 5.903730112630385e-06,
-      "loss": 0.2534,
+      "epoch": 4.013767254117562,
+      "grad_norm": 0.31350627541542053,
+      "learning_rate": 4.943848486604802e-06,
+      "loss": 0.3552,
       "step": 111370
     },
     {
-      "epoch": 3.92,
-      "learning_rate": 5.901891692174366e-06,
-      "loss": 0.258,
+      "epoch": 4.013947453778787,
+      "grad_norm": 0.26370009779930115,
+      "learning_rate": 4.942106516370298e-06,
+      "loss": 0.3816,
       "step": 111375
     },
     {
-      "epoch": 3.92,
-      "learning_rate": 5.90005351969469e-06,
-      "loss": 0.2528,
+      "epoch": 4.014127653440012,
+      "grad_norm": 0.24444960057735443,
+      "learning_rate": 4.940364819420709e-06,
+      "loss": 0.38,
       "step": 111380
     },
     {
-      "epoch": 3.92,
-      "learning_rate": 5.898215595215239e-06,
-      "loss": 0.2637,
+      "epoch": 4.014307853101236,
+      "grad_norm": 0.1919577717781067,
+      "learning_rate": 4.93862339577976e-06,
+      "loss": 0.3385,
       "step": 111385
     },
     {
-      "epoch": 3.92,
-      "learning_rate": 5.896377918759871e-06,
-      "loss": 0.2638,
+      "epoch": 4.014488052762461,
+      "grad_norm": 0.2629943788051605,
+      "learning_rate": 4.936882245471192e-06,
+      "loss": 0.3585,
       "step": 111390
     },
     {
-      "epoch": 3.92,
-      "learning_rate": 5.8945404903524404e-06,
-      "loss": 0.268,
+      "epoch": 4.014668252423686,
+      "grad_norm": 0.2678518295288086,
+      "learning_rate": 4.935141368518717e-06,
+      "loss": 0.3823,
       "step": 111395
     },
     {
-      "epoch": 3.92,
-      "learning_rate": 5.892703310016814e-06,
-      "loss": 0.2561,
+      "epoch": 4.01484845208491,
+      "grad_norm": 0.2841361165046692,
+      "learning_rate": 4.933400764946056e-06,
+      "loss": 0.3928,
       "step": 111400
     },
     {
-      "epoch": 3.92,
-      "learning_rate": 5.890866377776844e-06,
-      "loss": 0.247,
+      "epoch": 4.015028651746134,
+      "grad_norm": 0.2338554561138153,
+      "learning_rate": 4.931660434776925e-06,
+      "loss": 0.3745,
       "step": 111405
     },
     {
-      "epoch": 3.92,
-      "learning_rate": 5.889029693656381e-06,
-      "loss": 0.2453,
+      "epoch": 4.015208851407359,
+      "grad_norm": 0.2238939106464386,
+      "learning_rate": 4.929920378035027e-06,
+      "loss": 0.3918,
       "step": 111410
     },
     {
-      "epoch": 3.92,
-      "learning_rate": 5.887193257679266e-06,
-      "loss": 0.2482,
+      "epoch": 4.015389051068584,
+      "grad_norm": 0.28193482756614685,
+      "learning_rate": 4.9281805947440865e-06,
+      "loss": 0.3874,
       "step": 111415
     },
     {
-      "epoch": 3.92,
-      "learning_rate": 5.88535706986936e-06,
-      "loss": 0.2626,
+      "epoch": 4.015569250729809,
+      "grad_norm": 0.216234028339386,
+      "learning_rate": 4.926441084927805e-06,
+      "loss": 0.3888,
       "step": 111420
     },
     {
-      "epoch": 3.92,
-      "learning_rate": 5.883521130250485e-06,
-      "loss": 0.2708,
+      "epoch": 4.015749450391033,
+      "grad_norm": 0.22301048040390015,
+      "learning_rate": 4.924701848609864e-06,
+      "loss": 0.3606,
       "step": 111425
     },
     {
-      "epoch": 3.92,
-      "learning_rate": 5.881685438846498e-06,
-      "loss": 0.253,
+      "epoch": 4.015929650052258,
+      "grad_norm": 0.24179665744304657,
+      "learning_rate": 4.922962885813981e-06,
+      "loss": 0.3597,
       "step": 111430
     },
     {
-      "epoch": 3.92,
-      "learning_rate": 5.879849995681219e-06,
-      "loss": 0.2609,
+      "epoch": 4.016109849713483,
+      "grad_norm": 0.24188639223575592,
+      "learning_rate": 4.9212241965638365e-06,
+      "loss": 0.3479,
       "step": 111435
     },
     {
-      "epoch": 3.92,
-      "learning_rate": 5.878014800778495e-06,
-      "loss": 0.2458,
+      "epoch": 4.016290049374708,
+      "grad_norm": 0.26836642622947693,
+      "learning_rate": 4.919485780883135e-06,
+      "loss": 0.3606,
       "step": 111440
     },
     {
-      "epoch": 3.92,
-      "learning_rate": 5.876179854162145e-06,
-      "loss": 0.2702,
+      "epoch": 4.016470249035931,
+      "grad_norm": 0.2224971503019333,
+      "learning_rate": 4.9177476387955475e-06,
+      "loss": 0.3699,
       "step": 111445
     },
     {
-      "epoch": 3.92,
-      "learning_rate": 5.874345155855998e-06,
-      "loss": 0.277,
+      "epoch": 4.016650448697156,
+      "grad_norm": 0.27810531854629517,
+      "learning_rate": 4.916009770324753e-06,
+      "loss": 0.3726,
       "step": 111450
     },
     {
-      "epoch": 3.92,
-      "learning_rate": 5.872510705883868e-06,
-      "loss": 0.2534,
+      "epoch": 4.016830648358381,
+      "grad_norm": 0.21480928361415863,
+      "learning_rate": 4.9142721754944446e-06,
+      "loss": 0.3529,
       "step": 111455
     },
     {
-      "epoch": 3.92,
-      "learning_rate": 5.870676504269587e-06,
-      "loss": 0.275,
+      "epoch": 4.017010848019606,
+      "grad_norm": 0.24362391233444214,
+      "learning_rate": 4.9125348543282875e-06,
+      "loss": 0.3412,
       "step": 111460
     },
     {
-      "epoch": 3.92,
-      "learning_rate": 5.868842551036966e-06,
-      "loss": 0.2626,
+      "epoch": 4.01719104768083,
+      "grad_norm": 0.2782408893108368,
+      "learning_rate": 4.910797806849956e-06,
+      "loss": 0.3839,
       "step": 111465
     },
     {
-      "epoch": 3.92,
-      "learning_rate": 5.867008846209809e-06,
-      "loss": 0.2749,
+      "epoch": 4.017371247342055,
+      "grad_norm": 0.22316765785217285,
+      "learning_rate": 4.909061033083112e-06,
+      "loss": 0.3899,
       "step": 111470
     },
     {
-      "epoch": 3.92,
-      "learning_rate": 5.865175389811942e-06,
-      "loss": 0.2634,
+      "epoch": 4.01755144700328,
+      "grad_norm": 0.2562723457813263,
+      "learning_rate": 4.907324533051419e-06,
+      "loss": 0.3515,
       "step": 111475
     },
     {
-      "epoch": 3.92,
-      "learning_rate": 5.8633421818671515e-06,
-      "loss": 0.2567,
+      "epoch": 4.0177316466645046,
+      "grad_norm": 0.2701679468154907,
+      "learning_rate": 4.905588306778544e-06,
+      "loss": 0.3532,
       "step": 111480
     },
     {
-      "epoch": 3.92,
-      "learning_rate": 5.861509222399261e-06,
-      "loss": 0.2652,
+      "epoch": 4.017911846325729,
+      "grad_norm": 0.20503626763820648,
+      "learning_rate": 4.9038523542881355e-06,
+      "loss": 0.3923,
       "step": 111485
     },
     {
-      "epoch": 3.92,
-      "learning_rate": 5.8596765114320605e-06,
-      "loss": 0.2739,
+      "epoch": 4.018092045986953,
+      "grad_norm": 0.2625948488712311,
+      "learning_rate": 4.902116675603852e-06,
+      "loss": 0.3843,
       "step": 111490
     },
     {
-      "epoch": 3.92,
-      "learning_rate": 5.8578440489893385e-06,
-      "loss": 0.2396,
+      "epoch": 4.018272245648178,
+      "grad_norm": 0.24299366772174835,
+      "learning_rate": 4.900381270749335e-06,
+      "loss": 0.3681,
       "step": 111495
     },
     {
-      "epoch": 3.92,
-      "learning_rate": 5.856011835094902e-06,
-      "loss": 0.2528,
+      "epoch": 4.018452445309403,
+      "grad_norm": 0.19307754933834076,
+      "learning_rate": 4.898646139748234e-06,
+      "loss": 0.3657,
       "step": 111500
     },
     {
-      "epoch": 3.92,
-      "eval_loss": 0.2502204477787018,
-      "eval_runtime": 10.5596,
-      "eval_samples_per_second": 9.47,
-      "eval_steps_per_second": 9.47,
+      "epoch": 4.018452445309403,
+      "eval_loss": 0.42932766675949097,
+      "eval_runtime": 3.5358,
+      "eval_samples_per_second": 28.283,
+      "eval_steps_per_second": 7.071,
       "step": 111500
     },
     {
-      "epoch": 3.92,
-      "learning_rate": 5.8541798697725384e-06,
-      "loss": 0.2533,
+      "epoch": 4.018632644970627,
+      "grad_norm": 0.22898030281066895,
+      "learning_rate": 4.896911282624178e-06,
+      "loss": 0.3802,
       "step": 111505
     },
     {
-      "epoch": 3.92,
-      "learning_rate": 5.852348153046022e-06,
-      "loss": 0.2447,
+      "epoch": 4.018812844631852,
+      "grad_norm": 0.21475541591644287,
+      "learning_rate": 4.8951766994008274e-06,
+      "loss": 0.3608,
       "step": 111510
     },
     {
-      "epoch": 3.92,
-      "learning_rate": 5.8505166849391555e-06,
-      "loss": 0.259,
+      "epoch": 4.018993044293077,
+      "grad_norm": 0.2441704273223877,
+      "learning_rate": 4.893442390101791e-06,
+      "loss": 0.3579,
       "step": 111515
     },
     {
-      "epoch": 3.92,
-      "learning_rate": 5.848685465475709e-06,
-      "loss": 0.2621,
+      "epoch": 4.0191732439543015,
+      "grad_norm": 0.24041332304477692,
+      "learning_rate": 4.891708354750716e-06,
+      "loss": 0.3862,
       "step": 111520
     },
     {
-      "epoch": 3.92,
-      "learning_rate": 5.84685449467946e-06,
-      "loss": 0.2557,
+      "epoch": 4.019353443615526,
+      "grad_norm": 0.262669175863266,
+      "learning_rate": 4.889974593371218e-06,
+      "loss": 0.3809,
       "step": 111525
     },
     {
-      "epoch": 3.92,
-      "learning_rate": 5.845023772574177e-06,
-      "loss": 0.2671,
+      "epoch": 4.019533643276751,
+      "grad_norm": 0.3033019006252289,
+      "learning_rate": 4.888241105986918e-06,
+      "loss": 0.401,
       "step": 111530
     },
     {
-      "epoch": 3.92,
-      "learning_rate": 5.843193299183636e-06,
-      "loss": 0.2392,
+      "epoch": 4.019713842937975,
+      "grad_norm": 0.22395041584968567,
+      "learning_rate": 4.886507892621453e-06,
+      "loss": 0.3718,
       "step": 111535
     },
     {
-      "epoch": 3.92,
-      "learning_rate": 5.841363074531611e-06,
-      "loss": 0.2612,
+      "epoch": 4.0198940425992,
+      "grad_norm": 0.2800545394420624,
+      "learning_rate": 4.88477495329841e-06,
+      "loss": 0.3998,
       "step": 111540
     },
     {
-      "epoch": 3.92,
-      "learning_rate": 5.8398990739176875e-06,
-      "loss": 0.2569,
+      "epoch": 4.020074242260424,
+      "grad_norm": 0.2535509765148163,
+      "learning_rate": 4.883042288041423e-06,
+      "loss": 0.3666,
       "step": 111545
     },
     {
-      "epoch": 3.92,
-      "learning_rate": 5.838069297054866e-06,
-      "loss": 0.2386,
+      "epoch": 4.020254441921649,
+      "grad_norm": 0.26017653942108154,
+      "learning_rate": 4.881309896874087e-06,
+      "loss": 0.4077,
       "step": 111550
     },
     {
-      "epoch": 3.92,
-      "learning_rate": 5.836239768997084e-06,
-      "loss": 0.2383,
+      "epoch": 4.020434641582874,
+      "grad_norm": 0.2357221096754074,
+      "learning_rate": 4.879577779820007e-06,
+      "loss": 0.3764,
       "step": 111555
     },
     {
-      "epoch": 3.92,
-      "learning_rate": 5.834410489768097e-06,
-      "loss": 0.252,
+      "epoch": 4.0206148412440985,
+      "grad_norm": 0.24376040697097778,
+      "learning_rate": 4.877845936902789e-06,
+      "loss": 0.3291,
       "step": 111560
     },
     {
-      "epoch": 3.93,
-      "learning_rate": 5.832581459391662e-06,
-      "loss": 0.2548,
+      "epoch": 4.020795040905323,
+      "grad_norm": 0.22934941947460175,
+      "learning_rate": 4.87611436814602e-06,
+      "loss": 0.3561,
       "step": 111565
     },
     {
-      "epoch": 3.93,
-      "learning_rate": 5.830752677891521e-06,
-      "loss": 0.2536,
+      "epoch": 4.020975240566548,
+      "grad_norm": 0.25413674116134644,
+      "learning_rate": 4.874383073573294e-06,
+      "loss": 0.3447,
       "step": 111570
     },
     {
-      "epoch": 3.93,
-      "learning_rate": 5.8289241452914285e-06,
-      "loss": 0.246,
+      "epoch": 4.021155440227773,
+      "grad_norm": 0.23448479175567627,
+      "learning_rate": 4.872652053208207e-06,
+      "loss": 0.3587,
       "step": 111575
     },
     {
-      "epoch": 3.93,
-      "learning_rate": 5.827095861615114e-06,
-      "loss": 0.2682,
+      "epoch": 4.0213356398889974,
+      "grad_norm": 0.2433057725429535,
+      "learning_rate": 4.870921307074339e-06,
+      "loss": 0.3762,
       "step": 111580
     },
     {
-      "epoch": 3.93,
-      "learning_rate": 5.825267826886333e-06,
-      "loss": 0.2444,
+      "epoch": 4.021515839550221,
+      "grad_norm": 0.22767940163612366,
+      "learning_rate": 4.86919083519527e-06,
+      "loss": 0.383,
       "step": 111585
     },
     {
-      "epoch": 3.93,
-      "learning_rate": 5.823440041128813e-06,
-      "loss": 0.2722,
+      "epoch": 4.021696039211446,
+      "grad_norm": 0.24196745455265045,
+      "learning_rate": 4.867460637594579e-06,
+      "loss": 0.3517,
       "step": 111590
     },
     {
-      "epoch": 3.93,
-      "learning_rate": 5.821612504366286e-06,
-      "loss": 0.2466,
+      "epoch": 4.021876238872671,
+      "grad_norm": 0.27073025703430176,
+      "learning_rate": 4.865730714295832e-06,
+      "loss": 0.3839,
       "step": 111595
     },
     {
-      "epoch": 3.93,
-      "learning_rate": 5.819785216622473e-06,
-      "loss": 0.2542,
+      "epoch": 4.0220564385338955,
+      "grad_norm": 0.30549201369285583,
+      "learning_rate": 4.864001065322616e-06,
+      "loss": 0.377,
       "step": 111600
     },
     {
-      "epoch": 3.93,
-      "learning_rate": 5.817958177921118e-06,
-      "loss": 0.2506,
+      "epoch": 4.02223663819512,
+      "grad_norm": 0.23959122598171234,
+      "learning_rate": 4.862271690698489e-06,
+      "loss": 0.3811,
       "step": 111605
     },
     {
-      "epoch": 3.93,
-      "learning_rate": 5.816131388285934e-06,
-      "loss": 0.2846,
+      "epoch": 4.022416837856345,
+      "grad_norm": 0.2952626943588257,
+      "learning_rate": 4.8605425904470005e-06,
+      "loss": 0.3632,
       "step": 111610
     },
     {
-      "epoch": 3.93,
-      "learning_rate": 5.814304847740631e-06,
-      "loss": 0.2556,
+      "epoch": 4.02259703751757,
+      "grad_norm": 0.24323879182338715,
+      "learning_rate": 4.858813764591727e-06,
+      "loss": 0.4091,
       "step": 111615
     },
     {
-      "epoch": 3.93,
-      "learning_rate": 5.812478556308948e-06,
-      "loss": 0.2741,
+      "epoch": 4.022777237178794,
+      "grad_norm": 0.23510761559009552,
+      "learning_rate": 4.857085213156209e-06,
+      "loss": 0.3739,
       "step": 111620
     },
     {
-      "epoch": 3.93,
-      "learning_rate": 5.810652514014575e-06,
-      "loss": 0.2559,
+      "epoch": 4.022957436840019,
+      "grad_norm": 0.2673884630203247,
+      "learning_rate": 4.855356936164018e-06,
+      "loss": 0.3963,
       "step": 111625
     },
     {
-      "epoch": 3.93,
-      "learning_rate": 5.808826720881242e-06,
-      "loss": 0.236,
+      "epoch": 4.023137636501243,
+      "grad_norm": 0.25060799717903137,
+      "learning_rate": 4.853628933638682e-06,
+      "loss": 0.3889,
       "step": 111630
     },
     {
-      "epoch": 3.93,
-      "learning_rate": 5.807001176932644e-06,
-      "loss": 0.2694,
+      "epoch": 4.023317836162468,
+      "grad_norm": 0.2960754930973053,
+      "learning_rate": 4.851901205603746e-06,
+      "loss": 0.3883,
       "step": 111635
     },
     {
-      "epoch": 3.93,
-      "learning_rate": 5.805175882192484e-06,
-      "loss": 0.259,
+      "epoch": 4.0234980358236925,
+      "grad_norm": 0.2322222739458084,
+      "learning_rate": 4.85017375208276e-06,
+      "loss": 0.3718,
       "step": 111640
     },
     {
-      "epoch": 3.93,
-      "learning_rate": 5.8033508366844705e-06,
-      "loss": 0.2497,
+      "epoch": 4.023678235484917,
+      "grad_norm": 0.2459927797317505,
+      "learning_rate": 4.848446573099258e-06,
+      "loss": 0.3736,
       "step": 111645
     },
     {
-      "epoch": 3.93,
-      "learning_rate": 5.801526040432295e-06,
-      "loss": 0.2416,
+      "epoch": 4.023858435146142,
+      "grad_norm": 0.26722821593284607,
+      "learning_rate": 4.8467196686767695e-06,
+      "loss": 0.3832,
       "step": 111650
     },
     {
-      "epoch": 3.93,
-      "learning_rate": 5.79970149345965e-06,
-      "loss": 0.243,
+      "epoch": 4.024038634807367,
+      "grad_norm": 0.2405874878168106,
+      "learning_rate": 4.8449930388388215e-06,
+      "loss": 0.3748,
       "step": 111655
     },
     {
-      "epoch": 3.93,
-      "learning_rate": 5.797877195790222e-06,
-      "loss": 0.2511,
+      "epoch": 4.024218834468591,
+      "grad_norm": 0.2542579472064972,
+      "learning_rate": 4.843266683608936e-06,
+      "loss": 0.395,
       "step": 111660
     },
     {
-      "epoch": 3.93,
-      "learning_rate": 5.796053147447711e-06,
-      "loss": 0.2293,
+      "epoch": 4.024399034129816,
+      "grad_norm": 0.24059006571769714,
+      "learning_rate": 4.8415406030106465e-06,
+      "loss": 0.381,
       "step": 111665
     },
     {
-      "epoch": 3.93,
-      "learning_rate": 5.794229348455796e-06,
-      "loss": 0.2502,
+      "epoch": 4.024579233791041,
+      "grad_norm": 0.2607281804084778,
+      "learning_rate": 4.839814797067463e-06,
+      "loss": 0.3987,
       "step": 111670
     },
     {
-      "epoch": 3.93,
-      "learning_rate": 5.792405798838147e-06,
-      "loss": 0.2516,
+      "epoch": 4.024759433452265,
+      "grad_norm": 0.2265724539756775,
+      "learning_rate": 4.838089265802901e-06,
+      "loss": 0.3876,
       "step": 111675
     },
     {
-      "epoch": 3.93,
-      "learning_rate": 5.790582498618449e-06,
-      "loss": 0.2522,
+      "epoch": 4.0249396331134895,
+      "grad_norm": 0.28544244170188904,
+      "learning_rate": 4.836364009240465e-06,
+      "loss": 0.3896,
       "step": 111680
     },
     {
-      "epoch": 3.93,
-      "learning_rate": 5.788759447820386e-06,
-      "loss": 0.2582,
+      "epoch": 4.025119832774714,
+      "grad_norm": 0.21306416392326355,
+      "learning_rate": 4.83463902740367e-06,
+      "loss": 0.3687,
       "step": 111685
     },
     {
-      "epoch": 3.93,
-      "learning_rate": 5.786936646467622e-06,
-      "loss": 0.2405,
+      "epoch": 4.025300032435939,
+      "grad_norm": 0.22182679176330566,
+      "learning_rate": 4.832914320316006e-06,
+      "loss": 0.3602,
       "step": 111690
     },
     {
-      "epoch": 3.93,
-      "learning_rate": 5.785114094583815e-06,
-      "loss": 0.2332,
+      "epoch": 4.025480232097164,
+      "grad_norm": 0.2823246121406555,
+      "learning_rate": 4.831189888000986e-06,
+      "loss": 0.3927,
       "step": 111695
     },
     {
-      "epoch": 3.93,
-      "learning_rate": 5.783291792192647e-06,
-      "loss": 0.2697,
+      "epoch": 4.025660431758388,
+      "grad_norm": 0.2230641394853592,
+      "learning_rate": 4.8294657304821e-06,
+      "loss": 0.3698,
       "step": 111700
     },
     {
-      "epoch": 3.93,
-      "learning_rate": 5.781469739317769e-06,
-      "loss": 0.2611,
+      "epoch": 4.025840631419613,
+      "grad_norm": 0.2776728570461273,
+      "learning_rate": 4.827741847782838e-06,
+      "loss": 0.3692,
       "step": 111705
     },
     {
-      "epoch": 3.93,
-      "learning_rate": 5.779647935982843e-06,
-      "loss": 0.2595,
+      "epoch": 4.026020831080838,
+      "grad_norm": 0.2971155643463135,
+      "learning_rate": 4.826018239926689e-06,
+      "loss": 0.3711,
       "step": 111710
     },
     {
-      "epoch": 3.93,
-      "learning_rate": 5.777826382211513e-06,
-      "loss": 0.2782,
+      "epoch": 4.026201030742063,
+      "grad_norm": 0.23264330625534058,
+      "learning_rate": 4.824294906937126e-06,
+      "loss": 0.4002,
       "step": 111715
     },
     {
-      "epoch": 3.93,
-      "learning_rate": 5.776005078027447e-06,
-      "loss": 0.2558,
+      "epoch": 4.026381230403286,
+      "grad_norm": 0.20618966221809387,
+      "learning_rate": 4.8225718488376535e-06,
+      "loss": 0.3391,
       "step": 111720
     },
     {
-      "epoch": 3.93,
-      "learning_rate": 5.774184023454287e-06,
-      "loss": 0.2604,
+      "epoch": 4.026561430064511,
+      "grad_norm": 0.23295053839683533,
+      "learning_rate": 4.8208490656517225e-06,
+      "loss": 0.3322,
       "step": 111725
     },
     {
-      "epoch": 3.93,
-      "learning_rate": 5.77236321851567e-06,
-      "loss": 0.2573,
+      "epoch": 4.026741629725736,
+      "grad_norm": 0.19821318984031677,
+      "learning_rate": 4.81912655740282e-06,
+      "loss": 0.351,
       "step": 111730
     },
     {
-      "epoch": 3.93,
-      "learning_rate": 5.7705426632352426e-06,
-      "loss": 0.2599,
+      "epoch": 4.026921829386961,
+      "grad_norm": 0.3166259825229645,
+      "learning_rate": 4.81740432411441e-06,
+      "loss": 0.3433,
       "step": 111735
     },
     {
-      "epoch": 3.93,
-      "learning_rate": 5.768722357636658e-06,
-      "loss": 0.2361,
+      "epoch": 4.027102029048185,
+      "grad_norm": 0.2291676253080368,
+      "learning_rate": 4.815682365809959e-06,
+      "loss": 0.3965,
       "step": 111740
     },
     {
-      "epoch": 3.93,
-      "learning_rate": 5.766902301743534e-06,
-      "loss": 0.2601,
+      "epoch": 4.02728222870941,
+      "grad_norm": 0.2493552565574646,
+      "learning_rate": 4.81396068251293e-06,
+      "loss": 0.3637,
       "step": 111745
     },
     {
-      "epoch": 3.93,
-      "learning_rate": 5.765082495579513e-06,
-      "loss": 0.2411,
+      "epoch": 4.027462428370635,
+      "grad_norm": 0.2596898376941681,
+      "learning_rate": 4.812239274246777e-06,
+      "loss": 0.3494,
       "step": 111750
     },
     {
-      "epoch": 3.93,
-      "learning_rate": 5.76326293916821e-06,
-      "loss": 0.2592,
+      "epoch": 4.02764262803186,
+      "grad_norm": 0.246169313788414,
+      "learning_rate": 4.810518141034951e-06,
+      "loss": 0.4282,
       "step": 111755
     },
     {
-      "epoch": 3.93,
-      "learning_rate": 5.761443632533267e-06,
-      "loss": 0.2562,
+      "epoch": 4.027822827693084,
+      "grad_norm": 0.29572200775146484,
+      "learning_rate": 4.808797282900912e-06,
+      "loss": 0.3703,
       "step": 111760
     },
     {
-      "epoch": 3.93,
-      "learning_rate": 5.759624575698302e-06,
-      "loss": 0.2676,
+      "epoch": 4.028003027354308,
+      "grad_norm": 0.23195022344589233,
+      "learning_rate": 4.807076699868099e-06,
+      "loss": 0.3474,
       "step": 111765
     },
     {
-      "epoch": 3.93,
-      "learning_rate": 5.75780576868693e-06,
-      "loss": 0.2544,
+      "epoch": 4.028183227015533,
+      "grad_norm": 0.21716444194316864,
+      "learning_rate": 4.805356391959959e-06,
+      "loss": 0.3792,
       "step": 111770
     },
     {
-      "epoch": 3.93,
-      "learning_rate": 5.755987211522765e-06,
-      "loss": 0.2397,
+      "epoch": 4.028363426676758,
+      "grad_norm": 0.2253197580575943,
+      "learning_rate": 4.8036363591999255e-06,
+      "loss": 0.3733,
       "step": 111775
     },
     {
-      "epoch": 3.93,
-      "learning_rate": 5.754168904229426e-06,
-      "loss": 0.2674,
+      "epoch": 4.028543626337982,
+      "grad_norm": 0.26673805713653564,
+      "learning_rate": 4.801916601611433e-06,
+      "loss": 0.3862,
       "step": 111780
     },
     {
-      "epoch": 3.93,
-      "learning_rate": 5.7523508468305175e-06,
-      "loss": 0.2483,
+      "epoch": 4.028723825999207,
+      "grad_norm": 0.2358836978673935,
+      "learning_rate": 4.8001971192179226e-06,
+      "loss": 0.3647,
       "step": 111785
     },
     {
-      "epoch": 3.93,
-      "learning_rate": 5.750533039349654e-06,
-      "loss": 0.255,
+      "epoch": 4.028904025660432,
+      "grad_norm": 0.2774098813533783,
+      "learning_rate": 4.79847791204282e-06,
+      "loss": 0.3863,
       "step": 111790
     },
     {
-      "epoch": 3.93,
-      "learning_rate": 5.748715481810427e-06,
-      "loss": 0.2782,
+      "epoch": 4.0290842253216566,
+      "grad_norm": 0.287156879901886,
+      "learning_rate": 4.7967589801095335e-06,
+      "loss": 0.3558,
       "step": 111795
     },
     {
-      "epoch": 3.93,
-      "learning_rate": 5.746898174236448e-06,
-      "loss": 0.2647,
+      "epoch": 4.029264424982881,
+      "grad_norm": 0.24171940982341766,
+      "learning_rate": 4.7950403234414995e-06,
+      "loss": 0.358,
       "step": 111800
     },
     {
-      "epoch": 3.93,
-      "learning_rate": 5.745081116651307e-06,
-      "loss": 0.2554,
+      "epoch": 4.029444624644106,
+      "grad_norm": 0.23957575857639313,
+      "learning_rate": 4.793321942062124e-06,
+      "loss": 0.3846,
       "step": 111805
     },
     {
-      "epoch": 3.93,
-      "learning_rate": 5.7432643090786e-06,
-      "loss": 0.2611,
+      "epoch": 4.02962482430533,
+      "grad_norm": 0.28689080476760864,
+      "learning_rate": 4.791603835994829e-06,
+      "loss": 0.3987,
       "step": 111810
     },
     {
-      "epoch": 3.93,
-      "learning_rate": 5.741447751541909e-06,
-      "loss": 0.2397,
+      "epoch": 4.029805023966555,
+      "grad_norm": 0.28590720891952515,
+      "learning_rate": 4.789886005263028e-06,
+      "loss": 0.3648,
       "step": 111815
     },
     {
-      "epoch": 3.93,
-      "learning_rate": 5.739631444064833e-06,
-      "loss": 0.2665,
+      "epoch": 4.029985223627779,
+      "grad_norm": 0.22683300077915192,
+      "learning_rate": 4.788168449890104e-06,
+      "loss": 0.3772,
       "step": 111820
     },
     {
-      "epoch": 3.93,
-      "learning_rate": 5.7378153866709476e-06,
-      "loss": 0.2722,
+      "epoch": 4.030165423289004,
+      "grad_norm": 0.2961829900741577,
+      "learning_rate": 4.786451169899478e-06,
+      "loss": 0.3508,
       "step": 111825
     },
     {
-      "epoch": 3.93,
-      "learning_rate": 5.73599957938383e-06,
-      "loss": 0.2563,
+      "epoch": 4.030345622950229,
+      "grad_norm": 0.2695339322090149,
+      "learning_rate": 4.784734165314542e-06,
+      "loss": 0.3903,
       "step": 111830
     },
     {
-      "epoch": 3.93,
-      "learning_rate": 5.734184022227071e-06,
-      "loss": 0.2429,
+      "epoch": 4.0305258226114535,
+      "grad_norm": 0.1870175302028656,
+      "learning_rate": 4.783017436158691e-06,
+      "loss": 0.4212,
       "step": 111835
     },
     {
-      "epoch": 3.93,
-      "learning_rate": 5.732368715224226e-06,
-      "loss": 0.2568,
+      "epoch": 4.030706022272678,
+      "grad_norm": 0.26787975430488586,
+      "learning_rate": 4.781300982455309e-06,
+      "loss": 0.3925,
       "step": 111840
     },
     {
-      "epoch": 3.94,
-      "learning_rate": 5.730553658398885e-06,
-      "loss": 0.2663,
+      "epoch": 4.030886221933903,
+      "grad_norm": 0.2814364731311798,
+      "learning_rate": 4.779584804227783e-06,
+      "loss": 0.348,
       "step": 111845
     },
     {
-      "epoch": 3.94,
-      "learning_rate": 5.728738851774598e-06,
-      "loss": 0.2366,
+      "epoch": 4.031066421595128,
+      "grad_norm": 0.3020128309726715,
+      "learning_rate": 4.777868901499505e-06,
+      "loss": 0.3637,
       "step": 111850
     },
     {
-      "epoch": 3.94,
-      "learning_rate": 5.7269242953749416e-06,
-      "loss": 0.2618,
+      "epoch": 4.0312466212563525,
+      "grad_norm": 0.22931928932666779,
+      "learning_rate": 4.776153274293848e-06,
+      "loss": 0.3362,
       "step": 111855
     },
     {
-      "epoch": 3.94,
-      "learning_rate": 5.725109989223474e-06,
-      "loss": 0.2438,
+      "epoch": 4.031426820917576,
+      "grad_norm": 0.24356307089328766,
+      "learning_rate": 4.774437922634187e-06,
+      "loss": 0.3895,
       "step": 111860
     },
     {
-      "epoch": 3.94,
-      "learning_rate": 5.72329593334375e-06,
-      "loss": 0.237,
+      "epoch": 4.031607020578801,
+      "grad_norm": 0.2754194438457489,
+      "learning_rate": 4.772722846543892e-06,
+      "loss": 0.3886,
       "step": 111865
     },
     {
-      "epoch": 3.94,
-      "learning_rate": 5.721482127759317e-06,
-      "loss": 0.2486,
+      "epoch": 4.031787220240026,
+      "grad_norm": 0.2473660260438919,
+      "learning_rate": 4.7710080460463356e-06,
+      "loss": 0.3681,
       "step": 111870
     },
     {
-      "epoch": 3.94,
-      "learning_rate": 5.719668572493742e-06,
-      "loss": 0.2308,
+      "epoch": 4.0319674199012505,
+      "grad_norm": 0.24248690903186798,
+      "learning_rate": 4.769293521164869e-06,
+      "loss": 0.3893,
       "step": 111875
     },
     {
-      "epoch": 3.94,
-      "learning_rate": 5.717855267570563e-06,
-      "loss": 0.2412,
+      "epoch": 4.032147619562475,
+      "grad_norm": 0.27788907289505005,
+      "learning_rate": 4.76757927192287e-06,
+      "loss": 0.3653,
       "step": 111880
     },
     {
-      "epoch": 3.94,
-      "learning_rate": 5.716042213013326e-06,
-      "loss": 0.2283,
+      "epoch": 4.0323278192237,
+      "grad_norm": 0.22080229222774506,
+      "learning_rate": 4.765865298343686e-06,
+      "loss": 0.4033,
       "step": 111885
     },
     {
-      "epoch": 3.94,
-      "learning_rate": 5.714229408845567e-06,
-      "loss": 0.2597,
+      "epoch": 4.032508018884925,
+      "grad_norm": 0.26164865493774414,
+      "learning_rate": 4.764151600450667e-06,
+      "loss": 0.3896,
       "step": 111890
     },
     {
-      "epoch": 3.94,
-      "learning_rate": 5.712416855090827e-06,
-      "loss": 0.2631,
+      "epoch": 4.0326882185461495,
+      "grad_norm": 0.24852848052978516,
+      "learning_rate": 4.7624381782671675e-06,
+      "loss": 0.3699,
       "step": 111895
     },
     {
-      "epoch": 3.94,
-      "learning_rate": 5.7106045517726525e-06,
-      "loss": 0.2607,
+      "epoch": 4.032868418207374,
+      "grad_norm": 0.18736465275287628,
+      "learning_rate": 4.760725031816521e-06,
+      "loss": 0.3585,
       "step": 111900
     },
     {
-      "epoch": 3.94,
-      "learning_rate": 5.708792498914564e-06,
-      "loss": 0.2574,
+      "epoch": 4.033048617868598,
+      "grad_norm": 0.3258678913116455,
+      "learning_rate": 4.759012161122092e-06,
+      "loss": 0.4135,
       "step": 111905
     },
     {
-      "epoch": 3.94,
-      "learning_rate": 5.706980696540084e-06,
-      "loss": 0.2504,
+      "epoch": 4.033228817529823,
+      "grad_norm": 0.23651060461997986,
+      "learning_rate": 4.757299566207196e-06,
+      "loss": 0.3723,
       "step": 111910
     },
     {
-      "epoch": 3.94,
-      "learning_rate": 5.705169144672753e-06,
-      "loss": 0.2306,
+      "epoch": 4.0334090171910475,
+      "grad_norm": 0.2316112071275711,
+      "learning_rate": 4.75558724709517e-06,
+      "loss": 0.3634,
       "step": 111915
     },
     {
-      "epoch": 3.94,
-      "learning_rate": 5.703357843336085e-06,
-      "loss": 0.2684,
+      "epoch": 4.033589216852272,
+      "grad_norm": 0.22371111810207367,
+      "learning_rate": 4.753875203809352e-06,
+      "loss": 0.3619,
       "step": 111920
     },
     {
-      "epoch": 3.94,
-      "learning_rate": 5.701546792553597e-06,
-      "loss": 0.2615,
+      "epoch": 4.033769416513497,
+      "grad_norm": 0.2516692578792572,
+      "learning_rate": 4.752163436373061e-06,
+      "loss": 0.3887,
       "step": 111925
     },
     {
-      "epoch": 3.94,
-      "learning_rate": 5.6997359923488005e-06,
-      "loss": 0.2572,
+      "epoch": 4.033949616174722,
+      "grad_norm": 0.24128705263137817,
+      "learning_rate": 4.750451944809634e-06,
+      "loss": 0.3456,
       "step": 111930
     },
     {
-      "epoch": 3.94,
-      "learning_rate": 5.6979254427452195e-06,
-      "loss": 0.2564,
+      "epoch": 4.034129815835946,
+      "grad_norm": 0.24484743177890778,
+      "learning_rate": 4.748740729142373e-06,
+      "loss": 0.3982,
       "step": 111935
     },
     {
-      "epoch": 3.94,
-      "learning_rate": 5.696115143766356e-06,
-      "loss": 0.2539,
+      "epoch": 4.034310015497171,
+      "grad_norm": 0.21860742568969727,
+      "learning_rate": 4.747029789394591e-06,
+      "loss": 0.3646,
       "step": 111940
     },
     {
-      "epoch": 3.94,
-      "learning_rate": 5.694305095435709e-06,
-      "loss": 0.2691,
+      "epoch": 4.034490215158396,
+      "grad_norm": 0.24374550580978394,
+      "learning_rate": 4.7453191255896145e-06,
+      "loss": 0.3578,
       "step": 111945
     },
     {
-      "epoch": 3.94,
-      "learning_rate": 5.692495297776787e-06,
-      "loss": 0.2637,
+      "epoch": 4.03467041481962,
+      "grad_norm": 0.29540616273880005,
+      "learning_rate": 4.743608737750746e-06,
+      "loss": 0.3785,
       "step": 111950
     },
     {
-      "epoch": 3.94,
-      "learning_rate": 5.690685750813096e-06,
-      "loss": 0.255,
+      "epoch": 4.0348506144808445,
+      "grad_norm": 0.19289089739322662,
+      "learning_rate": 4.7418986259012836e-06,
+      "loss": 0.3744,
       "step": 111955
     },
     {
-      "epoch": 3.94,
-      "learning_rate": 5.688876454568126e-06,
-      "loss": 0.2362,
+      "epoch": 4.035030814142069,
+      "grad_norm": 0.2452259063720703,
+      "learning_rate": 4.740188790064531e-06,
+      "loss": 0.3566,
       "step": 111960
     },
     {
-      "epoch": 3.94,
-      "learning_rate": 5.6870674090653604e-06,
-      "loss": 0.2559,
+      "epoch": 4.035211013803294,
+      "grad_norm": 0.2256636768579483,
+      "learning_rate": 4.738479230263776e-06,
+      "loss": 0.3578,
       "step": 111965
     },
     {
-      "epoch": 3.94,
-      "learning_rate": 5.685258614328304e-06,
-      "loss": 0.2687,
+      "epoch": 4.035391213464519,
+      "grad_norm": 0.21084842085838318,
+      "learning_rate": 4.736769946522326e-06,
+      "loss": 0.359,
       "step": 111970
     },
     {
-      "epoch": 3.94,
-      "learning_rate": 5.683450070380436e-06,
-      "loss": 0.2709,
+      "epoch": 4.035571413125743,
+      "grad_norm": 0.22610744833946228,
+      "learning_rate": 4.735060938863464e-06,
+      "loss": 0.3715,
       "step": 111975
     },
     {
-      "epoch": 3.94,
-      "learning_rate": 5.681641777245239e-06,
-      "loss": 0.2485,
+      "epoch": 4.035751612786968,
+      "grad_norm": 0.2904773950576782,
+      "learning_rate": 4.733352207310473e-06,
+      "loss": 0.3575,
       "step": 111980
     },
     {
-      "epoch": 3.94,
-      "learning_rate": 5.679833734946186e-06,
-      "loss": 0.2666,
+      "epoch": 4.035931812448193,
+      "grad_norm": 0.2337644398212433,
+      "learning_rate": 4.731643751886633e-06,
+      "loss": 0.3408,
       "step": 111985
     },
     {
-      "epoch": 3.94,
-      "learning_rate": 5.678025943506765e-06,
-      "loss": 0.2412,
+      "epoch": 4.036112012109418,
+      "grad_norm": 0.2608318030834198,
+      "learning_rate": 4.729935572615219e-06,
+      "loss": 0.3671,
       "step": 111990
     },
     {
-      "epoch": 3.94,
-      "learning_rate": 5.676218402950445e-06,
-      "loss": 0.2448,
+      "epoch": 4.0362922117706415,
+      "grad_norm": 0.2300008088350296,
+      "learning_rate": 4.728227669519511e-06,
+      "loss": 0.3985,
       "step": 111995
     },
     {
-      "epoch": 3.94,
-      "learning_rate": 5.674411113300688e-06,
-      "loss": 0.2586,
+      "epoch": 4.036472411431866,
+      "grad_norm": 0.24639463424682617,
+      "learning_rate": 4.726520042622784e-06,
+      "loss": 0.364,
       "step": 112000
     },
     {
-      "epoch": 3.94,
-      "eval_loss": 0.2502914071083069,
-      "eval_runtime": 10.543,
-      "eval_samples_per_second": 9.485,
-      "eval_steps_per_second": 9.485,
+      "epoch": 4.036472411431866,
+      "eval_loss": 0.4291331470012665,
+      "eval_runtime": 3.5175,
+      "eval_samples_per_second": 28.429,
+      "eval_steps_per_second": 7.107,
       "step": 112000
     },
     {
-      "epoch": 3.94,
-      "learning_rate": 5.67260407458097e-06,
-      "loss": 0.26,
+      "epoch": 4.036652611093091,
+      "grad_norm": 0.2528611123561859,
+      "learning_rate": 4.724812691948285e-06,
+      "loss": 0.3803,
       "step": 112005
     },
     {
-      "epoch": 3.94,
-      "learning_rate": 5.670797286814747e-06,
-      "loss": 0.2595,
+      "epoch": 4.036832810754316,
+      "grad_norm": 0.21292369067668915,
+      "learning_rate": 4.723105617519294e-06,
+      "loss": 0.3711,
       "step": 112010
     },
     {
-      "epoch": 3.94,
-      "learning_rate": 5.668990750025488e-06,
-      "loss": 0.2258,
+      "epoch": 4.03701301041554,
+      "grad_norm": 0.238296777009964,
+      "learning_rate": 4.72139881935906e-06,
+      "loss": 0.4014,
       "step": 112015
     },
     {
-      "epoch": 3.94,
-      "learning_rate": 5.667184464236644e-06,
-      "loss": 0.2686,
+      "epoch": 4.037193210076765,
+      "grad_norm": 0.24008990824222565,
+      "learning_rate": 4.719692297490844e-06,
+      "loss": 0.3731,
       "step": 112020
     },
     {
-      "epoch": 3.94,
-      "learning_rate": 5.665378429471663e-06,
-      "loss": 0.274,
+      "epoch": 4.03737340973799,
+      "grad_norm": 0.2941230535507202,
+      "learning_rate": 4.717986051937895e-06,
+      "loss": 0.362,
       "step": 112025
     },
     {
-      "epoch": 3.94,
-      "learning_rate": 5.663572645754009e-06,
-      "loss": 0.2668,
+      "epoch": 4.037553609399215,
+      "grad_norm": 0.2386900782585144,
+      "learning_rate": 4.716280082723451e-06,
+      "loss": 0.3563,
       "step": 112030
     },
     {
-      "epoch": 3.94,
-      "learning_rate": 5.661767113107119e-06,
-      "loss": 0.2795,
+      "epoch": 4.037733809060439,
+      "grad_norm": 0.22380246222019196,
+      "learning_rate": 4.714574389870768e-06,
+      "loss": 0.3797,
       "step": 112035
     },
     {
-      "epoch": 3.94,
-      "learning_rate": 5.659961831554442e-06,
-      "loss": 0.2584,
+      "epoch": 4.037914008721663,
+      "grad_norm": 0.22900426387786865,
+      "learning_rate": 4.712868973403087e-06,
+      "loss": 0.3573,
       "step": 112040
     },
     {
-      "epoch": 3.94,
-      "learning_rate": 5.658156801119404e-06,
-      "loss": 0.25,
+      "epoch": 4.038094208382888,
+      "grad_norm": 0.21817940473556519,
+      "learning_rate": 4.711163833343635e-06,
+      "loss": 0.3323,
       "step": 112045
     },
     {
-      "epoch": 3.94,
-      "learning_rate": 5.656352021825464e-06,
-      "loss": 0.2491,
+      "epoch": 4.038274408044113,
+      "grad_norm": 0.2606887221336365,
+      "learning_rate": 4.709458969715646e-06,
+      "loss": 0.3829,
       "step": 112050
     },
     {
-      "epoch": 3.94,
-      "learning_rate": 5.654547493696036e-06,
-      "loss": 0.2243,
+      "epoch": 4.038454607705337,
+      "grad_norm": 0.17615626752376556,
+      "learning_rate": 4.7077543825423535e-06,
+      "loss": 0.3649,
       "step": 112055
     },
     {
-      "epoch": 3.94,
-      "learning_rate": 5.652743216754566e-06,
-      "loss": 0.2689,
+      "epoch": 4.038634807366562,
+      "grad_norm": 0.3894023299217224,
+      "learning_rate": 4.706050071846971e-06,
+      "loss": 0.3597,
       "step": 112060
     },
     {
-      "epoch": 3.94,
-      "learning_rate": 5.650939191024468e-06,
-      "loss": 0.2536,
+      "epoch": 4.038815007027787,
+      "grad_norm": 0.1984912008047104,
+      "learning_rate": 4.704346037652735e-06,
+      "loss": 0.3796,
       "step": 112065
     },
     {
-      "epoch": 3.94,
-      "learning_rate": 5.649135416529183e-06,
-      "loss": 0.2666,
+      "epoch": 4.038995206689012,
+      "grad_norm": 0.3245207667350769,
+      "learning_rate": 4.702642279982855e-06,
+      "loss": 0.4189,
       "step": 112070
     },
     {
-      "epoch": 3.94,
-      "learning_rate": 5.6473318932921185e-06,
-      "loss": 0.2499,
+      "epoch": 4.039175406350236,
+      "grad_norm": 0.2791898548603058,
+      "learning_rate": 4.700938798860544e-06,
+      "loss": 0.3301,
       "step": 112075
     },
     {
-      "epoch": 3.94,
-      "learning_rate": 5.6455286213366905e-06,
-      "loss": 0.2671,
+      "epoch": 4.039355606011461,
+      "grad_norm": 0.20591488480567932,
+      "learning_rate": 4.6992355943090095e-06,
+      "loss": 0.3515,
       "step": 112080
     },
     {
-      "epoch": 3.94,
-      "learning_rate": 5.643725600686326e-06,
-      "loss": 0.2577,
+      "epoch": 4.039535805672686,
+      "grad_norm": 0.20929506421089172,
+      "learning_rate": 4.6975326663514535e-06,
+      "loss": 0.349,
       "step": 112085
     },
     {
-      "epoch": 3.94,
-      "learning_rate": 5.641922831364424e-06,
-      "loss": 0.2648,
+      "epoch": 4.03971600533391,
+      "grad_norm": 0.27950236201286316,
+      "learning_rate": 4.695830015011094e-06,
+      "loss": 0.3617,
       "step": 112090
     },
     {
-      "epoch": 3.94,
-      "learning_rate": 5.6401203133944005e-06,
-      "loss": 0.2578,
+      "epoch": 4.039896204995134,
+      "grad_norm": 0.21270787715911865,
+      "learning_rate": 4.6941276403111245e-06,
+      "loss": 0.3668,
       "step": 112095
     },
     {
-      "epoch": 3.94,
-      "learning_rate": 5.6383180467996486e-06,
-      "loss": 0.2655,
+      "epoch": 4.040076404656359,
+      "grad_norm": 0.24949444830417633,
+      "learning_rate": 4.69242554227472e-06,
+      "loss": 0.3616,
       "step": 112100
     },
     {
-      "epoch": 3.94,
-      "learning_rate": 5.636516031603583e-06,
-      "loss": 0.2529,
+      "epoch": 4.040256604317584,
+      "grad_norm": 0.20137085020542145,
+      "learning_rate": 4.690723720925094e-06,
+      "loss": 0.347,
       "step": 112105
     },
     {
-      "epoch": 3.94,
-      "learning_rate": 5.6347142678295966e-06,
-      "loss": 0.2483,
+      "epoch": 4.0404368039788086,
+      "grad_norm": 0.2264707386493683,
+      "learning_rate": 4.689022176285418e-06,
+      "loss": 0.4139,
       "step": 112110
     },
     {
-      "epoch": 3.94,
-      "learning_rate": 5.632912755501077e-06,
-      "loss": 0.2485,
+      "epoch": 4.040617003640033,
+      "grad_norm": 0.25556209683418274,
+      "learning_rate": 4.687320908378895e-06,
+      "loss": 0.345,
       "step": 112115
     },
     {
-      "epoch": 3.94,
-      "learning_rate": 5.63111149464142e-06,
-      "loss": 0.2457,
+      "epoch": 4.040797203301258,
+      "grad_norm": 0.2599429488182068,
+      "learning_rate": 4.685619917228687e-06,
+      "loss": 0.3659,
       "step": 112120
     },
     {
-      "epoch": 3.94,
-      "learning_rate": 5.629310485274022e-06,
-      "loss": 0.2464,
+      "epoch": 4.040977402962483,
+      "grad_norm": 0.19548028707504272,
+      "learning_rate": 4.6839192028579674e-06,
+      "loss": 0.3567,
       "step": 112125
     },
     {
-      "epoch": 3.95,
-      "learning_rate": 5.627509727422265e-06,
-      "loss": 0.2581,
+      "epoch": 4.0411576026237075,
+      "grad_norm": 0.23381386697292328,
+      "learning_rate": 4.68221876528992e-06,
+      "loss": 0.3942,
       "step": 112130
     },
     {
-      "epoch": 3.95,
-      "learning_rate": 5.6257092211095236e-06,
-      "loss": 0.2541,
+      "epoch": 4.041337802284931,
+      "grad_norm": 0.21480177342891693,
+      "learning_rate": 4.68051860454771e-06,
+      "loss": 0.3674,
       "step": 112135
     },
     {
-      "epoch": 3.95,
-      "learning_rate": 5.623908966359176e-06,
-      "loss": 0.2483,
+      "epoch": 4.041518001946156,
+      "grad_norm": 0.20809517800807953,
+      "learning_rate": 4.678818720654502e-06,
+      "loss": 0.3405,
       "step": 112140
     },
     {
-      "epoch": 3.95,
-      "learning_rate": 5.622108963194605e-06,
-      "loss": 0.259,
+      "epoch": 4.041698201607381,
+      "grad_norm": 0.24788372218608856,
+      "learning_rate": 4.677119113633452e-06,
+      "loss": 0.4156,
       "step": 112145
     },
     {
-      "epoch": 3.95,
-      "learning_rate": 5.620309211639179e-06,
-      "loss": 0.2469,
+      "epoch": 4.0418784012686055,
+      "grad_norm": 0.24612867832183838,
+      "learning_rate": 4.675419783507712e-06,
+      "loss": 0.3738,
       "step": 112150
     },
     {
-      "epoch": 3.95,
-      "learning_rate": 5.618509711716269e-06,
-      "loss": 0.2391,
+      "epoch": 4.04205860092983,
+      "grad_norm": 0.2980947494506836,
+      "learning_rate": 4.673720730300451e-06,
+      "loss": 0.4167,
       "step": 112155
     },
     {
-      "epoch": 3.95,
-      "learning_rate": 5.616710463449229e-06,
-      "loss": 0.2729,
+      "epoch": 4.042238800591055,
+      "grad_norm": 0.2769397497177124,
+      "learning_rate": 4.672021954034811e-06,
+      "loss": 0.394,
       "step": 112160
     },
     {
-      "epoch": 3.95,
-      "learning_rate": 5.614911466861436e-06,
-      "loss": 0.2461,
+      "epoch": 4.04241900025228,
+      "grad_norm": 0.24730835855007172,
+      "learning_rate": 4.670323454733933e-06,
+      "loss": 0.3496,
       "step": 112165
     },
     {
-      "epoch": 3.95,
-      "learning_rate": 5.6131127219762355e-06,
-      "loss": 0.2695,
+      "epoch": 4.0425991999135045,
+      "grad_norm": 0.2504551410675049,
+      "learning_rate": 4.668625232420965e-06,
+      "loss": 0.3607,
       "step": 112170
     },
     {
-      "epoch": 3.95,
-      "learning_rate": 5.6113142288169916e-06,
-      "loss": 0.2568,
+      "epoch": 4.042779399574729,
+      "grad_norm": 0.2242858111858368,
+      "learning_rate": 4.666927287119032e-06,
+      "loss": 0.3641,
       "step": 112175
     },
     {
-      "epoch": 3.95,
-      "learning_rate": 5.60951598740706e-06,
-      "loss": 0.2366,
+      "epoch": 4.042959599235953,
+      "grad_norm": 0.22944757342338562,
+      "learning_rate": 4.665229618851289e-06,
+      "loss": 0.3853,
       "step": 112180
     },
     {
-      "epoch": 3.95,
-      "learning_rate": 5.607717997769785e-06,
-      "loss": 0.2724,
+      "epoch": 4.043139798897178,
+      "grad_norm": 0.24846546351909637,
+      "learning_rate": 4.663532227640857e-06,
+      "loss": 0.3888,
       "step": 112185
     },
     {
-      "epoch": 3.95,
-      "learning_rate": 5.605920259928512e-06,
-      "loss": 0.261,
+      "epoch": 4.0433199985584025,
+      "grad_norm": 0.2676353454589844,
+      "learning_rate": 4.661835113510851e-06,
+      "loss": 0.3324,
       "step": 112190
     },
     {
-      "epoch": 3.95,
-      "learning_rate": 5.604122773906576e-06,
-      "loss": 0.2338,
+      "epoch": 4.043500198219627,
+      "grad_norm": 0.39060455560684204,
+      "learning_rate": 4.6601382764844105e-06,
+      "loss": 0.3952,
       "step": 112195
     },
     {
-      "epoch": 3.95,
-      "learning_rate": 5.6023255397273295e-06,
-      "loss": 0.2412,
+      "epoch": 4.043680397880852,
+      "grad_norm": 0.2727122902870178,
+      "learning_rate": 4.6584417165846495e-06,
+      "loss": 0.3762,
       "step": 112200
     },
     {
-      "epoch": 3.95,
-      "learning_rate": 5.600528557414103e-06,
-      "loss": 0.2643,
+      "epoch": 4.043860597542077,
+      "grad_norm": 0.19775696098804474,
+      "learning_rate": 4.6567454338346804e-06,
+      "loss": 0.3758,
       "step": 112205
     },
     {
-      "epoch": 3.95,
-      "learning_rate": 5.598731826990231e-06,
-      "loss": 0.2565,
+      "epoch": 4.0440407972033015,
+      "grad_norm": 0.25784486532211304,
+      "learning_rate": 4.655049428257615e-06,
+      "loss": 0.3843,
       "step": 112210
     },
     {
-      "epoch": 3.95,
-      "learning_rate": 5.596935348479029e-06,
-      "loss": 0.2487,
+      "epoch": 4.044220996864526,
+      "grad_norm": 0.2990073263645172,
+      "learning_rate": 4.6533536998765555e-06,
+      "loss": 0.3875,
       "step": 112215
     },
     {
-      "epoch": 3.95,
-      "learning_rate": 5.595139121903844e-06,
-      "loss": 0.2487,
+      "epoch": 4.044401196525751,
+      "grad_norm": 0.24339328706264496,
+      "learning_rate": 4.651658248714621e-06,
+      "loss": 0.3816,
       "step": 112220
     },
     {
-      "epoch": 3.95,
-      "learning_rate": 5.593343147287983e-06,
-      "loss": 0.2763,
+      "epoch": 4.044581396186975,
+      "grad_norm": 0.2621822953224182,
+      "learning_rate": 4.649963074794902e-06,
+      "loss": 0.3381,
       "step": 112225
     },
     {
-      "epoch": 3.95,
-      "learning_rate": 5.591547424654775e-06,
-      "loss": 0.2291,
+      "epoch": 4.0447615958481995,
+      "grad_norm": 0.25351327657699585,
+      "learning_rate": 4.648268178140497e-06,
+      "loss": 0.3854,
       "step": 112230
     },
     {
-      "epoch": 3.95,
-      "learning_rate": 5.5897519540275305e-06,
-      "loss": 0.2449,
+      "epoch": 4.044941795509424,
+      "grad_norm": 0.3043707013130188,
+      "learning_rate": 4.646573558774497e-06,
+      "loss": 0.3838,
       "step": 112235
     },
     {
-      "epoch": 3.95,
-      "learning_rate": 5.587956735429569e-06,
-      "loss": 0.27,
+      "epoch": 4.045121995170649,
+      "grad_norm": 0.24836041033267975,
+      "learning_rate": 4.644879216719994e-06,
+      "loss": 0.3588,
       "step": 112240
     },
     {
-      "epoch": 3.95,
-      "learning_rate": 5.586161768884196e-06,
-      "loss": 0.2645,
+      "epoch": 4.045302194831874,
+      "grad_norm": 0.2699911594390869,
+      "learning_rate": 4.643185152000062e-06,
+      "loss": 0.3773,
       "step": 112245
     },
     {
-      "epoch": 3.95,
-      "learning_rate": 5.584367054414721e-06,
-      "loss": 0.2603,
+      "epoch": 4.045482394493098,
+      "grad_norm": 0.25113409757614136,
+      "learning_rate": 4.641491364637798e-06,
+      "loss": 0.3652,
       "step": 112250
     },
     {
-      "epoch": 3.95,
-      "learning_rate": 5.582572592044436e-06,
-      "loss": 0.2299,
+      "epoch": 4.045662594154323,
+      "grad_norm": 0.24034260213375092,
+      "learning_rate": 4.6397978546562744e-06,
+      "loss": 0.3806,
       "step": 112255
     },
     {
-      "epoch": 3.95,
-      "learning_rate": 5.580778381796656e-06,
-      "loss": 0.2727,
+      "epoch": 4.045842793815548,
+      "grad_norm": 0.2835821211338043,
+      "learning_rate": 4.6381046220785595e-06,
+      "loss": 0.3634,
       "step": 112260
     },
     {
-      "epoch": 3.95,
-      "learning_rate": 5.578984423694672e-06,
-      "loss": 0.2459,
+      "epoch": 4.046022993476773,
+      "grad_norm": 0.20180295407772064,
+      "learning_rate": 4.636411666927731e-06,
+      "loss": 0.3606,
       "step": 112265
     },
     {
-      "epoch": 3.95,
-      "learning_rate": 5.577190717761774e-06,
-      "loss": 0.244,
+      "epoch": 4.0462031931379965,
+      "grad_norm": 0.24634915590286255,
+      "learning_rate": 4.634718989226841e-06,
+      "loss": 0.3853,
       "step": 112270
     },
     {
-      "epoch": 3.95,
-      "learning_rate": 5.575397264021248e-06,
-      "loss": 0.2651,
+      "epoch": 4.046383392799221,
+      "grad_norm": 0.23739095032215118,
+      "learning_rate": 4.63302658899897e-06,
+      "loss": 0.3726,
       "step": 112275
     },
     {
-      "epoch": 3.95,
-      "learning_rate": 5.57360406249639e-06,
-      "loss": 0.281,
+      "epoch": 4.046563592460446,
+      "grad_norm": 0.32305994629859924,
+      "learning_rate": 4.631334466267176e-06,
+      "loss": 0.4008,
       "step": 112280
     },
     {
-      "epoch": 3.95,
-      "learning_rate": 5.571811113210484e-06,
-      "loss": 0.2392,
+      "epoch": 4.046743792121671,
+      "grad_norm": 0.2640374004840851,
+      "learning_rate": 4.629642621054492e-06,
+      "loss": 0.3762,
       "step": 112285
     },
     {
-      "epoch": 3.95,
-      "learning_rate": 5.570018416186804e-06,
-      "loss": 0.2478,
+      "epoch": 4.046923991782895,
+      "grad_norm": 0.25261178612709045,
+      "learning_rate": 4.627951053383991e-06,
+      "loss": 0.3538,
       "step": 112290
     },
     {
-      "epoch": 3.95,
-      "learning_rate": 5.568225971448634e-06,
-      "loss": 0.2503,
+      "epoch": 4.04710419144412,
+      "grad_norm": 0.25557366013526917,
+      "learning_rate": 4.626259763278707e-06,
+      "loss": 0.4039,
       "step": 112295
     },
     {
-      "epoch": 3.95,
-      "learning_rate": 5.566433779019245e-06,
-      "loss": 0.2471,
+      "epoch": 4.047284391105345,
+      "grad_norm": 0.23059487342834473,
+      "learning_rate": 4.624568750761699e-06,
+      "loss": 0.3653,
       "step": 112300
     },
     {
-      "epoch": 3.95,
-      "learning_rate": 5.564641838921905e-06,
-      "loss": 0.2399,
+      "epoch": 4.04746459076657,
+      "grad_norm": 0.1989499032497406,
+      "learning_rate": 4.622878015855994e-06,
+      "loss": 0.3872,
       "step": 112305
     },
     {
-      "epoch": 3.95,
-      "learning_rate": 5.562850151179877e-06,
-      "loss": 0.2553,
+      "epoch": 4.047644790427794,
+      "grad_norm": 0.2884153127670288,
+      "learning_rate": 4.621187558584622e-06,
+      "loss": 0.3838,
       "step": 112310
     },
     {
-      "epoch": 3.95,
-      "learning_rate": 5.561058715816436e-06,
-      "loss": 0.2661,
+      "epoch": 4.047824990089018,
+      "grad_norm": 0.26490432024002075,
+      "learning_rate": 4.61949737897063e-06,
+      "loss": 0.3588,
       "step": 112315
     },
     {
-      "epoch": 3.95,
-      "learning_rate": 5.5592675328548365e-06,
-      "loss": 0.2591,
+      "epoch": 4.048005189750243,
+      "grad_norm": 0.2709238827228546,
+      "learning_rate": 4.6178074770370424e-06,
+      "loss": 0.4018,
       "step": 112320
     },
     {
-      "epoch": 3.95,
-      "learning_rate": 5.557476602318337e-06,
-      "loss": 0.2437,
+      "epoch": 4.048185389411468,
+      "grad_norm": 0.2417963743209839,
+      "learning_rate": 4.616117852806881e-06,
+      "loss": 0.3779,
       "step": 112325
     },
     {
-      "epoch": 3.95,
-      "learning_rate": 5.5556859242301865e-06,
-      "loss": 0.2619,
+      "epoch": 4.048365589072692,
+      "grad_norm": 0.257938414812088,
+      "learning_rate": 4.614428506303168e-06,
+      "loss": 0.3659,
       "step": 112330
     },
     {
-      "epoch": 3.95,
-      "learning_rate": 5.5538954986136346e-06,
-      "loss": 0.2565,
+      "epoch": 4.048545788733917,
+      "grad_norm": 0.2447778433561325,
+      "learning_rate": 4.612739437548913e-06,
+      "loss": 0.363,
       "step": 112335
     },
     {
-      "epoch": 3.95,
-      "learning_rate": 5.5521053254919435e-06,
-      "loss": 0.252,
+      "epoch": 4.048725988395142,
+      "grad_norm": 0.26171186566352844,
+      "learning_rate": 4.611050646567142e-06,
+      "loss": 0.4081,
       "step": 112340
     },
     {
-      "epoch": 3.95,
-      "learning_rate": 5.550315404888348e-06,
-      "loss": 0.2521,
+      "epoch": 4.048906188056367,
+      "grad_norm": 0.27486851811408997,
+      "learning_rate": 4.60936213338086e-06,
+      "loss": 0.3324,
       "step": 112345
     },
     {
-      "epoch": 3.95,
-      "learning_rate": 5.5485257368260855e-06,
-      "loss": 0.2571,
+      "epoch": 4.049086387717591,
+      "grad_norm": 0.24321185052394867,
+      "learning_rate": 4.6076738980130705e-06,
+      "loss": 0.3454,
       "step": 112350
     },
     {
-      "epoch": 3.95,
-      "learning_rate": 5.5467363213283995e-06,
-      "loss": 0.2457,
+      "epoch": 4.049266587378816,
+      "grad_norm": 0.334555059671402,
+      "learning_rate": 4.605985940486776e-06,
+      "loss": 0.3937,
       "step": 112355
     },
     {
-      "epoch": 3.95,
-      "learning_rate": 5.5449471584185234e-06,
-      "loss": 0.2381,
+      "epoch": 4.049446787040041,
+      "grad_norm": 0.30166709423065186,
+      "learning_rate": 4.604298260824965e-06,
+      "loss": 0.3657,
       "step": 112360
     },
     {
-      "epoch": 3.95,
-      "learning_rate": 5.54315824811969e-06,
-      "loss": 0.2456,
+      "epoch": 4.049626986701265,
+      "grad_norm": 0.2711981236934662,
+      "learning_rate": 4.602610859050652e-06,
+      "loss": 0.3539,
       "step": 112365
     },
     {
-      "epoch": 3.95,
-      "learning_rate": 5.541369590455114e-06,
-      "loss": 0.2348,
+      "epoch": 4.049807186362489,
+      "grad_norm": 0.30052125453948975,
+      "learning_rate": 4.600923735186824e-06,
+      "loss": 0.3794,
       "step": 112370
     },
     {
-      "epoch": 3.95,
-      "learning_rate": 5.5395811854480365e-06,
-      "loss": 0.2713,
+      "epoch": 4.049987386023714,
+      "grad_norm": 0.24002404510974884,
+      "learning_rate": 4.5992368892564446e-06,
+      "loss": 0.3768,
       "step": 112375
     },
     {
-      "epoch": 3.95,
-      "learning_rate": 5.537793033121674e-06,
-      "loss": 0.2595,
+      "epoch": 4.050167585684939,
+      "grad_norm": 0.26016783714294434,
+      "learning_rate": 4.5975503212825206e-06,
+      "loss": 0.3781,
       "step": 112380
     },
     {
-      "epoch": 3.95,
-      "learning_rate": 5.536005133499236e-06,
-      "loss": 0.2677,
+      "epoch": 4.050347785346164,
+      "grad_norm": 0.2607904374599457,
+      "learning_rate": 4.595864031288025e-06,
+      "loss": 0.3555,
       "step": 112385
     },
     {
-      "epoch": 3.95,
-      "learning_rate": 5.534217486603946e-06,
-      "loss": 0.2486,
+      "epoch": 4.050527985007388,
+      "grad_norm": 0.2204258143901825,
+      "learning_rate": 4.594178019295922e-06,
+      "loss": 0.3897,
       "step": 112390
     },
     {
-      "epoch": 3.95,
-      "learning_rate": 5.532430092459017e-06,
-      "loss": 0.2567,
+      "epoch": 4.050708184668613,
+      "grad_norm": 0.2713705003261566,
+      "learning_rate": 4.592492285329208e-06,
+      "loss": 0.3818,
       "step": 112395
     },
     {
-      "epoch": 3.95,
-      "learning_rate": 5.530642951087656e-06,
-      "loss": 0.2534,
+      "epoch": 4.050888384329838,
+      "grad_norm": 0.271567165851593,
+      "learning_rate": 4.590806829410824e-06,
+      "loss": 0.3609,
       "step": 112400
     },
     {
-      "epoch": 3.95,
-      "learning_rate": 5.528856062513058e-06,
-      "loss": 0.2319,
+      "epoch": 4.0510685839910625,
+      "grad_norm": 0.24618300795555115,
+      "learning_rate": 4.589121651563749e-06,
+      "loss": 0.3876,
       "step": 112405
     },
     {
-      "epoch": 3.95,
-      "learning_rate": 5.527069426758438e-06,
-      "loss": 0.2461,
+      "epoch": 4.051248783652286,
+      "grad_norm": 0.29348066449165344,
+      "learning_rate": 4.587436751810942e-06,
+      "loss": 0.3868,
       "step": 112410
     },
     {
-      "epoch": 3.96,
-      "learning_rate": 5.525283043846988e-06,
-      "loss": 0.2407,
+      "epoch": 4.051428983313511,
+      "grad_norm": 0.21088933944702148,
+      "learning_rate": 4.585752130175355e-06,
+      "loss": 0.3823,
       "step": 112415
     },
     {
-      "epoch": 3.96,
-      "learning_rate": 5.523496913801904e-06,
-      "loss": 0.2517,
+      "epoch": 4.051609182974736,
+      "grad_norm": 0.23839104175567627,
+      "learning_rate": 4.5840677866799435e-06,
+      "loss": 0.3536,
       "step": 112420
     },
     {
-      "epoch": 3.96,
-      "learning_rate": 5.521711036646371e-06,
-      "loss": 0.2763,
+      "epoch": 4.051789382635961,
+      "grad_norm": 0.2764272391796112,
+      "learning_rate": 4.582383721347658e-06,
+      "loss": 0.3414,
       "step": 112425
     },
     {
-      "epoch": 3.96,
-      "learning_rate": 5.51992541240359e-06,
-      "loss": 0.243,
+      "epoch": 4.051969582297185,
+      "grad_norm": 0.2566319406032562,
+      "learning_rate": 4.580699934201432e-06,
+      "loss": 0.374,
       "step": 112430
     },
     {
-      "epoch": 3.96,
-      "learning_rate": 5.518140041096739e-06,
-      "loss": 0.2605,
+      "epoch": 4.05214978195841,
+      "grad_norm": 0.24363850057125092,
+      "learning_rate": 4.579016425264224e-06,
+      "loss": 0.3331,
       "step": 112435
     },
     {
-      "epoch": 3.96,
-      "learning_rate": 5.516354922748996e-06,
-      "loss": 0.2414,
+      "epoch": 4.052329981619635,
+      "grad_norm": 0.25705766677856445,
+      "learning_rate": 4.577333194558964e-06,
+      "loss": 0.3527,
       "step": 112440
     },
     {
-      "epoch": 3.96,
-      "learning_rate": 5.514570057383547e-06,
-      "loss": 0.2687,
+      "epoch": 4.0525101812808595,
+      "grad_norm": 0.23861709237098694,
+      "learning_rate": 4.575650242108584e-06,
+      "loss": 0.3601,
       "step": 112445
     },
     {
-      "epoch": 3.96,
-      "learning_rate": 5.512785445023561e-06,
-      "loss": 0.2659,
+      "epoch": 4.052690380942084,
+      "grad_norm": 0.23494866490364075,
+      "learning_rate": 4.573967567936016e-06,
+      "loss": 0.3682,
       "step": 112450
     },
     {
-      "epoch": 3.96,
-      "learning_rate": 5.511001085692219e-06,
-      "loss": 0.2549,
+      "epoch": 4.052870580603308,
+      "grad_norm": 0.29604148864746094,
+      "learning_rate": 4.572285172064181e-06,
+      "loss": 0.3502,
       "step": 112455
     },
     {
-      "epoch": 3.96,
-      "learning_rate": 5.509216979412685e-06,
-      "loss": 0.2681,
+      "epoch": 4.053050780264533,
+      "grad_norm": 0.2703550457954407,
+      "learning_rate": 4.570603054516009e-06,
+      "loss": 0.372,
       "step": 112460
     },
     {
-      "epoch": 3.96,
-      "learning_rate": 5.507433126208114e-06,
-      "loss": 0.2681,
+      "epoch": 4.0532309799257575,
+      "grad_norm": 0.20241078734397888,
+      "learning_rate": 4.568921215314423e-06,
+      "loss": 0.3367,
       "step": 112465
     },
     {
-      "epoch": 3.96,
-      "learning_rate": 5.5056495261016886e-06,
-      "loss": 0.2617,
+      "epoch": 4.053411179586982,
+      "grad_norm": 0.2732201814651489,
+      "learning_rate": 4.567239654482313e-06,
+      "loss": 0.3518,
       "step": 112470
     },
     {
-      "epoch": 3.96,
-      "learning_rate": 5.5038661791165545e-06,
-      "loss": 0.2559,
+      "epoch": 4.053591379248207,
+      "grad_norm": 0.2432149201631546,
+      "learning_rate": 4.5655583720426195e-06,
+      "loss": 0.3639,
       "step": 112475
     },
     {
-      "epoch": 3.96,
-      "learning_rate": 5.502083085275869e-06,
-      "loss": 0.2439,
+      "epoch": 4.053771578909432,
+      "grad_norm": 0.28690987825393677,
+      "learning_rate": 4.563877368018227e-06,
+      "loss": 0.395,
       "step": 112480
     },
     {
-      "epoch": 3.96,
-      "learning_rate": 5.500300244602779e-06,
-      "loss": 0.2442,
+      "epoch": 4.0539517785706565,
+      "grad_norm": 0.24714258313179016,
+      "learning_rate": 4.562196642432063e-06,
+      "loss": 0.3719,
       "step": 112485
     },
     {
-      "epoch": 3.96,
-      "learning_rate": 5.498517657120447e-06,
-      "loss": 0.2508,
+      "epoch": 4.054131978231881,
+      "grad_norm": 0.22608081996440887,
+      "learning_rate": 4.560516195307005e-06,
+      "loss": 0.3674,
       "step": 112490
     },
     {
-      "epoch": 3.96,
-      "learning_rate": 5.496735322852004e-06,
-      "loss": 0.2434,
+      "epoch": 4.054312177893106,
+      "grad_norm": 0.26828816533088684,
+      "learning_rate": 4.5588360266659495e-06,
+      "loss": 0.3886,
       "step": 112495
     },
     {
-      "epoch": 3.96,
-      "learning_rate": 5.494953241820608e-06,
-      "loss": 0.2763,
+      "epoch": 4.05449237755433,
+      "grad_norm": 0.24400553107261658,
+      "learning_rate": 4.557156136531804e-06,
+      "loss": 0.3551,
       "step": 112500
     },
     {
-      "epoch": 3.96,
-      "eval_loss": 0.25041884183883667,
-      "eval_runtime": 10.5467,
-      "eval_samples_per_second": 9.482,
-      "eval_steps_per_second": 9.482,
+      "epoch": 4.05449237755433,
+      "eval_loss": 0.42917296290397644,
+      "eval_runtime": 3.5255,
+      "eval_samples_per_second": 28.365,
+      "eval_steps_per_second": 7.091,
       "step": 112500
     },
     {
-      "epoch": 3.96,
-      "learning_rate": 5.49317141404938e-06,
-      "loss": 0.2521,
+      "epoch": 4.0546725772155545,
+      "grad_norm": 0.27448979020118713,
+      "learning_rate": 4.555476524927449e-06,
+      "loss": 0.378,
       "step": 112505
     },
     {
-      "epoch": 3.96,
-      "learning_rate": 5.491389839561476e-06,
-      "loss": 0.2585,
+      "epoch": 4.054852776876779,
+      "grad_norm": 0.24059316515922546,
+      "learning_rate": 4.553797191875767e-06,
+      "loss": 0.3708,
       "step": 112510
     },
     {
-      "epoch": 3.96,
-      "learning_rate": 5.489608518380016e-06,
-      "loss": 0.2675,
+      "epoch": 4.055032976538004,
+      "grad_norm": 0.26339319348335266,
+      "learning_rate": 4.552118137399639e-06,
+      "loss": 0.3886,
       "step": 112515
     },
     {
-      "epoch": 3.96,
-      "learning_rate": 5.487827450528124e-06,
-      "loss": 0.2587,
+      "epoch": 4.055213176199229,
+      "grad_norm": 0.2619602084159851,
+      "learning_rate": 4.550439361521935e-06,
+      "loss": 0.3898,
       "step": 112520
     },
     {
-      "epoch": 3.96,
-      "learning_rate": 5.486046636028941e-06,
-      "loss": 0.2654,
+      "epoch": 4.0553933758604535,
+      "grad_norm": 0.25006812810897827,
+      "learning_rate": 4.548760864265545e-06,
+      "loss": 0.3685,
       "step": 112525
     },
     {
-      "epoch": 3.96,
-      "learning_rate": 5.484266074905584e-06,
-      "loss": 0.2562,
+      "epoch": 4.055573575521678,
+      "grad_norm": 0.2471635937690735,
+      "learning_rate": 4.547082645653328e-06,
+      "loss": 0.3887,
       "step": 112530
     },
     {
-      "epoch": 3.96,
-      "learning_rate": 5.4824857671811695e-06,
-      "loss": 0.2806,
+      "epoch": 4.055753775182903,
+      "grad_norm": 0.2514253258705139,
+      "learning_rate": 4.545404705708148e-06,
+      "loss": 0.3463,
       "step": 112535
     },
     {
-      "epoch": 3.96,
-      "learning_rate": 5.4807057128788054e-06,
-      "loss": 0.2583,
+      "epoch": 4.055933974844128,
+      "grad_norm": 0.21959030628204346,
+      "learning_rate": 4.543727044452873e-06,
+      "loss": 0.4068,
       "step": 112540
     },
     {
-      "epoch": 3.96,
-      "learning_rate": 5.478925912021623e-06,
-      "loss": 0.253,
+      "epoch": 4.0561141745053515,
+      "grad_norm": 0.21514563262462616,
+      "learning_rate": 4.542049661910344e-06,
+      "loss": 0.337,
       "step": 112545
     },
     {
-      "epoch": 3.96,
-      "learning_rate": 5.477146364632724e-06,
-      "loss": 0.2511,
+      "epoch": 4.056294374166576,
+      "grad_norm": 0.1881556361913681,
+      "learning_rate": 4.540372558103439e-06,
+      "loss": 0.3604,
       "step": 112550
     },
     {
-      "epoch": 3.96,
-      "learning_rate": 5.475367070735207e-06,
-      "loss": 0.25,
+      "epoch": 4.056474573827801,
+      "grad_norm": 0.2358875721693039,
+      "learning_rate": 4.538695733054995e-06,
+      "loss": 0.3601,
       "step": 112555
     },
     {
-      "epoch": 3.96,
-      "learning_rate": 5.473588030352178e-06,
-      "loss": 0.2861,
+      "epoch": 4.056654773489026,
+      "grad_norm": 0.2567630410194397,
+      "learning_rate": 4.537019186787861e-06,
+      "loss": 0.3783,
       "step": 112560
     },
     {
-      "epoch": 3.96,
-      "learning_rate": 5.471809243506751e-06,
-      "loss": 0.2481,
+      "epoch": 4.05683497315025,
+      "grad_norm": 0.26848697662353516,
+      "learning_rate": 4.535342919324878e-06,
+      "loss": 0.4059,
       "step": 112565
     },
     {
-      "epoch": 3.96,
-      "learning_rate": 5.470030710222007e-06,
-      "loss": 0.2609,
+      "epoch": 4.057015172811475,
+      "grad_norm": 0.22706463932991028,
+      "learning_rate": 4.533666930688885e-06,
+      "loss": 0.368,
       "step": 112570
     },
     {
-      "epoch": 3.96,
-      "learning_rate": 5.468252430521048e-06,
-      "loss": 0.2488,
+      "epoch": 4.0571953724727,
+      "grad_norm": 0.26052287220954895,
+      "learning_rate": 4.531991220902712e-06,
+      "loss": 0.3455,
       "step": 112575
     },
     {
-      "epoch": 3.96,
-      "learning_rate": 5.4664744044269495e-06,
-      "loss": 0.279,
+      "epoch": 4.057375572133925,
+      "grad_norm": 0.2673739194869995,
+      "learning_rate": 4.530315789989209e-06,
+      "loss": 0.3275,
       "step": 112580
     },
     {
-      "epoch": 3.96,
-      "learning_rate": 5.464696631962815e-06,
-      "loss": 0.2666,
+      "epoch": 4.057555771795149,
+      "grad_norm": 0.22586947679519653,
+      "learning_rate": 4.528640637971177e-06,
+      "loss": 0.3832,
       "step": 112585
     },
     {
-      "epoch": 3.96,
-      "learning_rate": 5.462919113151721e-06,
-      "loss": 0.2647,
+      "epoch": 4.057735971456374,
+      "grad_norm": 0.28457507491111755,
+      "learning_rate": 4.5269657648714605e-06,
+      "loss": 0.3725,
       "step": 112590
     },
     {
-      "epoch": 3.96,
-      "learning_rate": 5.461141848016746e-06,
-      "loss": 0.2778,
+      "epoch": 4.057916171117598,
+      "grad_norm": 0.27478164434432983,
+      "learning_rate": 4.52529117071287e-06,
+      "loss": 0.3783,
       "step": 112595
     },
     {
-      "epoch": 3.96,
-      "learning_rate": 5.45936483658096e-06,
-      "loss": 0.2522,
+      "epoch": 4.058096370778823,
+      "grad_norm": 0.22563715279102325,
+      "learning_rate": 4.523616855518226e-06,
+      "loss": 0.3952,
       "step": 112600
     },
     {
-      "epoch": 3.96,
-      "learning_rate": 5.4575880788674534e-06,
-      "loss": 0.2357,
+      "epoch": 4.058276570440047,
+      "grad_norm": 0.26424163579940796,
+      "learning_rate": 4.5219428193103345e-06,
+      "loss": 0.3737,
       "step": 112605
     },
     {
-      "epoch": 3.96,
-      "learning_rate": 5.455811574899278e-06,
-      "loss": 0.2347,
+      "epoch": 4.058456770101272,
+      "grad_norm": 0.2882944643497467,
+      "learning_rate": 4.5202690621120095e-06,
+      "loss": 0.3523,
       "step": 112610
     },
     {
-      "epoch": 3.96,
-      "learning_rate": 5.454035324699517e-06,
-      "loss": 0.2538,
+      "epoch": 4.058636969762497,
+      "grad_norm": 0.2373085916042328,
+      "learning_rate": 4.518595583946048e-06,
+      "loss": 0.3757,
       "step": 112615
     },
     {
-      "epoch": 3.96,
-      "learning_rate": 5.452259328291218e-06,
-      "loss": 0.2651,
+      "epoch": 4.058817169423722,
+      "grad_norm": 0.2473360151052475,
+      "learning_rate": 4.516922384835259e-06,
+      "loss": 0.3808,
       "step": 112620
     },
     {
-      "epoch": 3.96,
-      "learning_rate": 5.450483585697458e-06,
-      "loss": 0.2506,
+      "epoch": 4.058997369084946,
+      "grad_norm": 0.19893379509449005,
+      "learning_rate": 4.51524946480244e-06,
+      "loss": 0.3353,
       "step": 112625
     },
     {
-      "epoch": 3.96,
-      "learning_rate": 5.448708096941282e-06,
-      "loss": 0.2621,
+      "epoch": 4.059177568746171,
+      "grad_norm": 0.3020991086959839,
+      "learning_rate": 4.51357682387038e-06,
+      "loss": 0.3551,
       "step": 112630
     },
     {
-      "epoch": 3.96,
-      "learning_rate": 5.446932862045742e-06,
-      "loss": 0.2529,
+      "epoch": 4.059357768407396,
+      "grad_norm": 0.2574337422847748,
+      "learning_rate": 4.511904462061873e-06,
+      "loss": 0.3703,
       "step": 112635
     },
     {
-      "epoch": 3.96,
-      "learning_rate": 5.445157881033899e-06,
-      "loss": 0.2539,
+      "epoch": 4.05953796806862,
+      "grad_norm": 0.2121739387512207,
+      "learning_rate": 4.51023237939969e-06,
+      "loss": 0.3629,
       "step": 112640
     },
     {
-      "epoch": 3.96,
-      "learning_rate": 5.443383153928794e-06,
-      "loss": 0.2768,
+      "epoch": 4.059718167729844,
+      "grad_norm": 0.2671159505844116,
+      "learning_rate": 4.508560575906631e-06,
+      "loss": 0.3642,
       "step": 112645
     },
     {
-      "epoch": 3.96,
-      "learning_rate": 5.441608680753471e-06,
-      "loss": 0.2464,
+      "epoch": 4.059898367391069,
+      "grad_norm": 0.18342912197113037,
+      "learning_rate": 4.506889051605473e-06,
+      "loss": 0.3798,
       "step": 112650
     },
     {
-      "epoch": 3.96,
-      "learning_rate": 5.439834461530963e-06,
-      "loss": 0.2607,
+      "epoch": 4.060078567052294,
+      "grad_norm": 0.22590021789073944,
+      "learning_rate": 4.505217806518975e-06,
+      "loss": 0.3921,
       "step": 112655
     },
     {
-      "epoch": 3.96,
-      "learning_rate": 5.43806049628432e-06,
-      "loss": 0.2392,
+      "epoch": 4.060258766713519,
+      "grad_norm": 0.24707923829555511,
+      "learning_rate": 4.50354684066992e-06,
+      "loss": 0.3438,
       "step": 112660
     },
     {
-      "epoch": 3.96,
-      "learning_rate": 5.436286785036562e-06,
-      "loss": 0.2353,
+      "epoch": 4.060438966374743,
+      "grad_norm": 0.22925209999084473,
+      "learning_rate": 4.501876154081064e-06,
+      "loss": 0.3645,
       "step": 112665
     },
     {
-      "epoch": 3.96,
-      "learning_rate": 5.434513327810733e-06,
-      "loss": 0.2523,
+      "epoch": 4.060619166035968,
+      "grad_norm": 0.21116822957992554,
+      "learning_rate": 4.500205746775185e-06,
+      "loss": 0.3326,
       "step": 112670
     },
     {
-      "epoch": 3.96,
-      "learning_rate": 5.432740124629851e-06,
-      "loss": 0.2582,
+      "epoch": 4.060799365697193,
+      "grad_norm": 0.25304093956947327,
+      "learning_rate": 4.498535618775037e-06,
+      "loss": 0.3963,
       "step": 112675
     },
     {
-      "epoch": 3.96,
-      "learning_rate": 5.430967175516949e-06,
-      "loss": 0.2597,
+      "epoch": 4.0609795653584175,
+      "grad_norm": 0.25455543398857117,
+      "learning_rate": 4.496865770103362e-06,
+      "loss": 0.3797,
       "step": 112680
     },
     {
-      "epoch": 3.96,
-      "learning_rate": 5.42919448049504e-06,
-      "loss": 0.2584,
+      "epoch": 4.061159765019641,
+      "grad_norm": 0.24712957441806793,
+      "learning_rate": 4.495196200782928e-06,
+      "loss": 0.3822,
       "step": 112685
     },
     {
-      "epoch": 3.96,
-      "learning_rate": 5.427422039587144e-06,
-      "loss": 0.2459,
+      "epoch": 4.061339964680866,
+      "grad_norm": 0.21119093894958496,
+      "learning_rate": 4.493526910836476e-06,
+      "loss": 0.3993,
       "step": 112690
     },
     {
-      "epoch": 3.96,
-      "learning_rate": 5.425649852816267e-06,
-      "loss": 0.2611,
+      "epoch": 4.061520164342091,
+      "grad_norm": 0.2761557698249817,
+      "learning_rate": 4.491857900286747e-06,
+      "loss": 0.3782,
       "step": 112695
     },
     {
-      "epoch": 3.97,
-      "learning_rate": 5.423877920205434e-06,
-      "loss": 0.2471,
+      "epoch": 4.061700364003316,
+      "grad_norm": 0.2396336793899536,
+      "learning_rate": 4.490189169156487e-06,
+      "loss": 0.3892,
       "step": 112700
     },
     {
-      "epoch": 3.97,
-      "learning_rate": 5.422106241777644e-06,
-      "loss": 0.2433,
+      "epoch": 4.06188056366454,
+      "grad_norm": 0.3060736656188965,
+      "learning_rate": 4.488520717468419e-06,
+      "loss": 0.3525,
       "step": 112705
     },
     {
-      "epoch": 3.97,
-      "learning_rate": 5.420334817555903e-06,
-      "loss": 0.2592,
+      "epoch": 4.062060763325765,
+      "grad_norm": 0.2690458595752716,
+      "learning_rate": 4.486852545245296e-06,
+      "loss": 0.3654,
       "step": 112710
     },
     {
-      "epoch": 3.97,
-      "learning_rate": 5.418563647563205e-06,
-      "loss": 0.2581,
+      "epoch": 4.06224096298699,
+      "grad_norm": 0.25593844056129456,
+      "learning_rate": 4.4851846525098305e-06,
+      "loss": 0.3875,
       "step": 112715
     },
     {
-      "epoch": 3.97,
-      "learning_rate": 5.416792731822551e-06,
-      "loss": 0.269,
+      "epoch": 4.0624211626482145,
+      "grad_norm": 0.276349812746048,
+      "learning_rate": 4.483517039284754e-06,
+      "loss": 0.367,
       "step": 112720
     },
     {
-      "epoch": 3.97,
-      "learning_rate": 5.415022070356948e-06,
-      "loss": 0.2512,
+      "epoch": 4.062601362309439,
+      "grad_norm": 0.3029731810092926,
+      "learning_rate": 4.481849705592789e-06,
+      "loss": 0.3476,
       "step": 112725
     },
     {
-      "epoch": 3.97,
-      "learning_rate": 5.413251663189373e-06,
-      "loss": 0.249,
+      "epoch": 4.062781561970663,
+      "grad_norm": 0.2233833223581314,
+      "learning_rate": 4.480182651456638e-06,
+      "loss": 0.3712,
       "step": 112730
     },
     {
-      "epoch": 3.97,
-      "learning_rate": 5.411481510342811e-06,
-      "loss": 0.257,
+      "epoch": 4.062961761631888,
+      "grad_norm": 0.29494714736938477,
+      "learning_rate": 4.4785158768990365e-06,
+      "loss": 0.4097,
       "step": 112735
     },
     {
-      "epoch": 3.97,
-      "learning_rate": 5.40971161184026e-06,
-      "loss": 0.2505,
+      "epoch": 4.063141961293113,
+      "grad_norm": 0.23293974995613098,
+      "learning_rate": 4.47684938194268e-06,
+      "loss": 0.362,
       "step": 112740
     },
     {
-      "epoch": 3.97,
-      "learning_rate": 5.407941967704694e-06,
-      "loss": 0.2469,
+      "epoch": 4.063322160954337,
+      "grad_norm": 0.2608863115310669,
+      "learning_rate": 4.475183166610278e-06,
+      "loss": 0.3822,
       "step": 112745
     },
     {
-      "epoch": 3.97,
-      "learning_rate": 5.406172577959082e-06,
-      "loss": 0.2598,
+      "epoch": 4.063502360615562,
+      "grad_norm": 0.2746981978416443,
+      "learning_rate": 4.473517230924532e-06,
+      "loss": 0.369,
       "step": 112750
     },
     {
-      "epoch": 3.97,
-      "learning_rate": 5.404403442626413e-06,
-      "loss": 0.2777,
+      "epoch": 4.063682560276787,
+      "grad_norm": 0.24028368294239044,
+      "learning_rate": 4.47185157490814e-06,
+      "loss": 0.402,
       "step": 112755
     },
     {
-      "epoch": 3.97,
-      "learning_rate": 5.4026345617296495e-06,
-      "loss": 0.2368,
+      "epoch": 4.0638627599380115,
+      "grad_norm": 0.2710413634777069,
+      "learning_rate": 4.47018619858379e-06,
+      "loss": 0.3931,
       "step": 112760
     },
     {
-      "epoch": 3.97,
-      "learning_rate": 5.400865935291763e-06,
-      "loss": 0.2426,
+      "epoch": 4.064042959599236,
+      "grad_norm": 0.22267523407936096,
+      "learning_rate": 4.468521101974188e-06,
+      "loss": 0.3561,
       "step": 112765
     },
     {
-      "epoch": 3.97,
-      "learning_rate": 5.399097563335709e-06,
-      "loss": 0.251,
+      "epoch": 4.064223159260461,
+      "grad_norm": 0.22819946706295013,
+      "learning_rate": 4.466856285102e-06,
+      "loss": 0.3752,
       "step": 112770
     },
     {
-      "epoch": 3.97,
-      "learning_rate": 5.397329445884456e-06,
-      "loss": 0.2764,
+      "epoch": 4.064403358921685,
+      "grad_norm": 0.2634032964706421,
+      "learning_rate": 4.465191747989927e-06,
+      "loss": 0.3828,
       "step": 112775
     },
     {
-      "epoch": 3.97,
-      "learning_rate": 5.395561582960968e-06,
-      "loss": 0.2656,
+      "epoch": 4.0645835585829095,
+      "grad_norm": 0.2825946807861328,
+      "learning_rate": 4.463527490660641e-06,
+      "loss": 0.3642,
       "step": 112780
     },
     {
-      "epoch": 3.97,
-      "learning_rate": 5.3937939745881935e-06,
-      "loss": 0.2502,
+      "epoch": 4.064763758244134,
+      "grad_norm": 0.24241934716701508,
+      "learning_rate": 4.461863513136816e-06,
+      "loss": 0.3458,
       "step": 112785
     },
     {
-      "epoch": 3.97,
-      "learning_rate": 5.392026620789075e-06,
-      "loss": 0.2647,
+      "epoch": 4.064943957905359,
+      "grad_norm": 0.27716174721717834,
+      "learning_rate": 4.460199815441124e-06,
+      "loss": 0.3882,
       "step": 112790
     },
     {
-      "epoch": 3.97,
-      "learning_rate": 5.390259521586577e-06,
-      "loss": 0.2678,
+      "epoch": 4.065124157566584,
+      "grad_norm": 0.24334949254989624,
+      "learning_rate": 4.458536397596233e-06,
+      "loss": 0.3789,
       "step": 112795
     },
     {
-      "epoch": 3.97,
-      "learning_rate": 5.388492677003634e-06,
-      "loss": 0.2412,
+      "epoch": 4.0653043572278085,
+      "grad_norm": 0.26515430212020874,
+      "learning_rate": 4.456873259624802e-06,
+      "loss": 0.4021,
       "step": 112800
     },
     {
-      "epoch": 3.97,
-      "learning_rate": 5.38672608706319e-06,
-      "loss": 0.2462,
+      "epoch": 4.065484556889033,
+      "grad_norm": 0.28932276368141174,
+      "learning_rate": 4.455210401549501e-06,
+      "loss": 0.3874,
       "step": 112805
     },
     {
-      "epoch": 3.97,
-      "learning_rate": 5.384959751788176e-06,
-      "loss": 0.2595,
+      "epoch": 4.065664756550258,
+      "grad_norm": 0.19949932396411896,
+      "learning_rate": 4.4535478233929786e-06,
+      "loss": 0.3483,
       "step": 112810
     },
     {
-      "epoch": 3.97,
-      "learning_rate": 5.383193671201539e-06,
-      "loss": 0.2423,
+      "epoch": 4.065844956211483,
+      "grad_norm": 0.2426164299249649,
+      "learning_rate": 4.45188552517789e-06,
+      "loss": 0.3642,
       "step": 112815
     },
     {
-      "epoch": 3.97,
-      "learning_rate": 5.381427845326206e-06,
-      "loss": 0.2729,
+      "epoch": 4.0660251558727065,
+      "grad_norm": 0.24339383840560913,
+      "learning_rate": 4.450223506926884e-06,
+      "loss": 0.3921,
       "step": 112820
     },
     {
-      "epoch": 3.97,
-      "learning_rate": 5.379662274185093e-06,
-      "loss": 0.2904,
+      "epoch": 4.066205355533931,
+      "grad_norm": 0.2848939001560211,
+      "learning_rate": 4.448561768662599e-06,
+      "loss": 0.3685,
       "step": 112825
     },
     {
-      "epoch": 3.97,
-      "learning_rate": 5.377896957801137e-06,
-      "loss": 0.2528,
+      "epoch": 4.066385555195156,
+      "grad_norm": 0.23407739400863647,
+      "learning_rate": 4.4469003104076865e-06,
+      "loss": 0.3936,
       "step": 112830
     },
     {
-      "epoch": 3.97,
-      "learning_rate": 5.376131896197267e-06,
-      "loss": 0.2373,
+      "epoch": 4.066565754856381,
+      "grad_norm": 0.34121012687683105,
+      "learning_rate": 4.445239132184778e-06,
+      "loss": 0.3816,
       "step": 112835
     },
     {
-      "epoch": 3.97,
-      "learning_rate": 5.374367089396387e-06,
-      "loss": 0.2667,
+      "epoch": 4.0667459545176055,
+      "grad_norm": 0.2433071732521057,
+      "learning_rate": 4.443578234016504e-06,
+      "loss": 0.3895,
       "step": 112840
     },
     {
-      "epoch": 3.97,
-      "learning_rate": 5.37260253742142e-06,
-      "loss": 0.271,
+      "epoch": 4.06692615417883,
+      "grad_norm": 0.24688871204853058,
+      "learning_rate": 4.4419176159255e-06,
+      "loss": 0.4009,
       "step": 112845
     },
     {
-      "epoch": 3.97,
-      "learning_rate": 5.370838240295267e-06,
-      "loss": 0.2607,
+      "epoch": 4.067106353840055,
+      "grad_norm": 0.255391389131546,
+      "learning_rate": 4.440257277934381e-06,
+      "loss": 0.3783,
       "step": 112850
     },
     {
-      "epoch": 3.97,
-      "learning_rate": 5.36907419804085e-06,
-      "loss": 0.2656,
+      "epoch": 4.06728655350128,
+      "grad_norm": 0.323646605014801,
+      "learning_rate": 4.438597220065782e-06,
+      "loss": 0.3655,
       "step": 112855
     },
     {
-      "epoch": 3.97,
-      "learning_rate": 5.367310410681064e-06,
-      "loss": 0.2527,
+      "epoch": 4.067466753162504,
+      "grad_norm": 0.3119712769985199,
+      "learning_rate": 4.436937442342323e-06,
+      "loss": 0.357,
       "step": 112860
     },
     {
-      "epoch": 3.97,
-      "learning_rate": 5.365546878238811e-06,
-      "loss": 0.2566,
+      "epoch": 4.067646952823729,
+      "grad_norm": 0.2642781138420105,
+      "learning_rate": 4.435277944786595e-06,
+      "loss": 0.3654,
       "step": 112865
     },
     {
-      "epoch": 3.97,
-      "learning_rate": 5.363783600736997e-06,
-      "loss": 0.239,
+      "epoch": 4.067827152484953,
+      "grad_norm": 0.19354425370693207,
+      "learning_rate": 4.433618727421232e-06,
+      "loss": 0.3937,
       "step": 112870
     },
     {
-      "epoch": 3.97,
-      "learning_rate": 5.362020578198512e-06,
-      "loss": 0.2389,
+      "epoch": 4.068007352146178,
+      "grad_norm": 0.227387472987175,
+      "learning_rate": 4.431959790268831e-06,
+      "loss": 0.3485,
       "step": 112875
     },
     {
-      "epoch": 3.97,
-      "learning_rate": 5.360257810646244e-06,
-      "loss": 0.2692,
+      "epoch": 4.068187551807402,
+      "grad_norm": 0.256632000207901,
+      "learning_rate": 4.430301133351997e-06,
+      "loss": 0.3545,
       "step": 112880
     },
     {
-      "epoch": 3.97,
-      "learning_rate": 5.358495298103089e-06,
-      "loss": 0.2588,
+      "epoch": 4.068367751468627,
+      "grad_norm": 0.21638993918895721,
+      "learning_rate": 4.428642756693324e-06,
+      "loss": 0.3527,
       "step": 112885
     },
     {
-      "epoch": 3.97,
-      "learning_rate": 5.356733040591921e-06,
-      "loss": 0.2516,
+      "epoch": 4.068547951129852,
+      "grad_norm": 0.30779585242271423,
+      "learning_rate": 4.426984660315409e-06,
+      "loss": 0.3595,
       "step": 112890
     },
     {
-      "epoch": 3.97,
-      "learning_rate": 5.354971038135639e-06,
-      "loss": 0.2483,
+      "epoch": 4.068728150791077,
+      "grad_norm": 0.2729836702346802,
+      "learning_rate": 4.425326844240849e-06,
+      "loss": 0.419,
       "step": 112895
     },
     {
-      "epoch": 3.97,
-      "learning_rate": 5.353209290757108e-06,
-      "loss": 0.2648,
+      "epoch": 4.068908350452301,
+      "grad_norm": 0.26512590050697327,
+      "learning_rate": 4.423669308492228e-06,
+      "loss": 0.3389,
       "step": 112900
     },
     {
-      "epoch": 3.97,
-      "learning_rate": 5.351447798479203e-06,
-      "loss": 0.2546,
+      "epoch": 4.069088550113526,
+      "grad_norm": 0.27653393149375916,
+      "learning_rate": 4.422012053092128e-06,
+      "loss": 0.3723,
       "step": 112905
     },
     {
-      "epoch": 3.97,
-      "learning_rate": 5.349686561324805e-06,
-      "loss": 0.2499,
+      "epoch": 4.069268749774751,
+      "grad_norm": 0.22129282355308533,
+      "learning_rate": 4.420355078063129e-06,
+      "loss": 0.3573,
       "step": 112910
     },
     {
-      "epoch": 3.97,
-      "learning_rate": 5.347925579316776e-06,
-      "loss": 0.2448,
+      "epoch": 4.069448949435975,
+      "grad_norm": 0.22182343900203705,
+      "learning_rate": 4.41869838342781e-06,
+      "loss": 0.3635,
       "step": 112915
     },
     {
-      "epoch": 3.97,
-      "learning_rate": 5.346164852477984e-06,
-      "loss": 0.2545,
+      "epoch": 4.069629149097199,
+      "grad_norm": 0.24553145468235016,
+      "learning_rate": 4.417041969208735e-06,
+      "loss": 0.3854,
       "step": 112920
     },
     {
-      "epoch": 3.97,
-      "learning_rate": 5.34440438083128e-06,
-      "loss": 0.2552,
+      "epoch": 4.069809348758424,
+      "grad_norm": 0.24355679750442505,
+      "learning_rate": 4.4153858354284814e-06,
+      "loss": 0.3762,
       "step": 112925
     },
     {
-      "epoch": 3.97,
-      "learning_rate": 5.34264416439954e-06,
-      "loss": 0.2438,
+      "epoch": 4.069989548419649,
+      "grad_norm": 0.24160782992839813,
+      "learning_rate": 4.413729982109613e-06,
+      "loss": 0.3698,
       "step": 112930
     },
     {
-      "epoch": 3.97,
-      "learning_rate": 5.340884203205604e-06,
-      "loss": 0.2661,
+      "epoch": 4.070169748080874,
+      "grad_norm": 0.23975780606269836,
+      "learning_rate": 4.412074409274689e-06,
+      "loss": 0.3852,
       "step": 112935
     },
     {
-      "epoch": 3.97,
-      "learning_rate": 5.339124497272338e-06,
-      "loss": 0.2654,
+      "epoch": 4.070349947742098,
+      "grad_norm": 0.19808252155780792,
+      "learning_rate": 4.4104191169462625e-06,
+      "loss": 0.341,
       "step": 112940
     },
     {
-      "epoch": 3.97,
-      "learning_rate": 5.337365046622575e-06,
-      "loss": 0.2698,
+      "epoch": 4.070530147403323,
+      "grad_norm": 0.2225724160671234,
+      "learning_rate": 4.4087641051468846e-06,
+      "loss": 0.3836,
       "step": 112945
     },
     {
-      "epoch": 3.97,
-      "learning_rate": 5.3356058512791765e-06,
-      "loss": 0.235,
+      "epoch": 4.070710347064548,
+      "grad_norm": 0.2074027806520462,
+      "learning_rate": 4.4071093738991125e-06,
+      "loss": 0.3487,
       "step": 112950
     },
     {
-      "epoch": 3.97,
-      "learning_rate": 5.333846911264975e-06,
-      "loss": 0.248,
+      "epoch": 4.0708905467257726,
+      "grad_norm": 0.271662175655365,
+      "learning_rate": 4.405454923225491e-06,
+      "loss": 0.3715,
       "step": 112955
     },
     {
-      "epoch": 3.97,
-      "learning_rate": 5.3320882266028135e-06,
-      "loss": 0.2477,
+      "epoch": 4.071070746386996,
+      "grad_norm": 0.23517721891403198,
+      "learning_rate": 4.40380075314856e-06,
+      "loss": 0.3949,
       "step": 112960
     },
     {
-      "epoch": 3.97,
-      "learning_rate": 5.330329797315517e-06,
-      "loss": 0.2536,
+      "epoch": 4.071250946048221,
+      "grad_norm": 0.24322974681854248,
+      "learning_rate": 4.402146863690854e-06,
+      "loss": 0.384,
       "step": 112965
     },
     {
-      "epoch": 3.97,
-      "learning_rate": 5.3285716234259316e-06,
-      "loss": 0.2584,
+      "epoch": 4.071431145709446,
+      "grad_norm": 0.2979622781276703,
+      "learning_rate": 4.400493254874902e-06,
+      "loss": 0.3898,
       "step": 112970
     },
     {
-      "epoch": 3.97,
-      "learning_rate": 5.326813704956879e-06,
-      "loss": 0.2535,
+      "epoch": 4.071611345370671,
+      "grad_norm": 0.19745075702667236,
+      "learning_rate": 4.3988399267232555e-06,
+      "loss": 0.3579,
       "step": 112975
     },
     {
-      "epoch": 3.97,
-      "learning_rate": 5.325056041931187e-06,
-      "loss": 0.2336,
+      "epoch": 4.071791545031895,
+      "grad_norm": 0.23466679453849792,
+      "learning_rate": 4.397186879258419e-06,
+      "loss": 0.3682,
       "step": 112980
     },
     {
-      "epoch": 3.98,
-      "learning_rate": 5.323298634371668e-06,
-      "loss": 0.2672,
+      "epoch": 4.07197174469312,
+      "grad_norm": 0.18801245093345642,
+      "learning_rate": 4.39553411250292e-06,
+      "loss": 0.3595,
       "step": 112985
     },
     {
-      "epoch": 3.98,
-      "learning_rate": 5.321541482301157e-06,
-      "loss": 0.2798,
+      "epoch": 4.072151944354345,
+      "grad_norm": 0.23627303540706635,
+      "learning_rate": 4.393881626479282e-06,
+      "loss": 0.3774,
       "step": 112990
     },
     {
-      "epoch": 3.98,
-      "learning_rate": 5.319784585742455e-06,
-      "loss": 0.2321,
+      "epoch": 4.0723321440155695,
+      "grad_norm": 0.28023478388786316,
+      "learning_rate": 4.3922294212100175e-06,
+      "loss": 0.3824,
       "step": 112995
     },
     {
-      "epoch": 3.98,
-      "learning_rate": 5.318027944718379e-06,
-      "loss": 0.2488,
+      "epoch": 4.072512343676794,
+      "grad_norm": 0.2806805372238159,
+      "learning_rate": 4.390577496717638e-06,
+      "loss": 0.4054,
       "step": 113000
     },
     {
-      "epoch": 3.98,
-      "eval_loss": 0.25009918212890625,
-      "eval_runtime": 10.5513,
-      "eval_samples_per_second": 9.478,
-      "eval_steps_per_second": 9.478,
+      "epoch": 4.072512343676794,
+      "eval_loss": 0.4291308522224426,
+      "eval_runtime": 3.5302,
+      "eval_samples_per_second": 28.327,
+      "eval_steps_per_second": 7.082,
       "step": 113000
     },
     {
-      "epoch": 3.98,
-      "learning_rate": 5.316271559251745e-06,
-      "loss": 0.243,
+      "epoch": 4.072692543338018,
+      "grad_norm": 0.30133917927742004,
+      "learning_rate": 4.3889258530246505e-06,
+      "loss": 0.3692,
       "step": 113005
     },
     {
-      "epoch": 3.98,
-      "learning_rate": 5.314515429365355e-06,
-      "loss": 0.2529,
+      "epoch": 4.072872742999243,
+      "grad_norm": 0.22192399203777313,
+      "learning_rate": 4.387274490153551e-06,
+      "loss": 0.3635,
       "step": 113010
     },
     {
-      "epoch": 3.98,
-      "learning_rate": 5.312759555082006e-06,
-      "loss": 0.2655,
+      "epoch": 4.073052942660468,
+      "grad_norm": 0.23747487366199493,
+      "learning_rate": 4.385623408126852e-06,
+      "loss": 0.3788,
       "step": 113015
     },
     {
-      "epoch": 3.98,
-      "learning_rate": 5.311003936424497e-06,
-      "loss": 0.2353,
+      "epoch": 4.073233142321692,
+      "grad_norm": 0.25482213497161865,
+      "learning_rate": 4.383972606967044e-06,
+      "loss": 0.3497,
       "step": 113020
     },
     {
-      "epoch": 3.98,
-      "learning_rate": 5.309248573415629e-06,
-      "loss": 0.2534,
+      "epoch": 4.073413341982917,
+      "grad_norm": 0.2351936399936676,
+      "learning_rate": 4.382322086696616e-06,
+      "loss": 0.3412,
       "step": 113025
     },
     {
-      "epoch": 3.98,
-      "learning_rate": 5.307493466078195e-06,
-      "loss": 0.2704,
+      "epoch": 4.073593541644142,
+      "grad_norm": 0.21274252235889435,
+      "learning_rate": 4.380671847338061e-06,
+      "loss": 0.3827,
       "step": 113030
     },
     {
-      "epoch": 3.98,
-      "learning_rate": 5.305738614434977e-06,
-      "loss": 0.262,
+      "epoch": 4.0737737413053665,
+      "grad_norm": 0.26760604977607727,
+      "learning_rate": 4.379021888913851e-06,
+      "loss": 0.4002,
       "step": 113035
     },
     {
-      "epoch": 3.98,
-      "learning_rate": 5.303984018508759e-06,
-      "loss": 0.254,
+      "epoch": 4.073953940966591,
+      "grad_norm": 0.28649207949638367,
+      "learning_rate": 4.377372211446482e-06,
+      "loss": 0.3985,
       "step": 113040
     },
     {
-      "epoch": 3.98,
-      "learning_rate": 5.302229678322338e-06,
-      "loss": 0.2676,
+      "epoch": 4.074134140627816,
+      "grad_norm": 0.23005342483520508,
+      "learning_rate": 4.37572281495843e-06,
+      "loss": 0.3485,
       "step": 113045
     },
     {
-      "epoch": 3.98,
-      "learning_rate": 5.300475593898471e-06,
-      "loss": 0.2601,
+      "epoch": 4.07431434028904,
+      "grad_norm": 0.24847397208213806,
+      "learning_rate": 4.37407369947215e-06,
+      "loss": 0.3606,
       "step": 113050
     },
     {
-      "epoch": 3.98,
-      "learning_rate": 5.298721765259956e-06,
-      "loss": 0.27,
+      "epoch": 4.074494539950265,
+      "grad_norm": 0.2495976686477661,
+      "learning_rate": 4.37242486501013e-06,
+      "loss": 0.4039,
       "step": 113055
     },
     {
-      "epoch": 3.98,
-      "learning_rate": 5.296968192429549e-06,
-      "loss": 0.2447,
+      "epoch": 4.074674739611489,
+      "grad_norm": 0.25559067726135254,
+      "learning_rate": 4.370776311594826e-06,
+      "loss": 0.3642,
       "step": 113060
     },
     {
-      "epoch": 3.98,
-      "learning_rate": 5.295214875430032e-06,
-      "loss": 0.249,
+      "epoch": 4.074854939272714,
+      "grad_norm": 0.22897782921791077,
+      "learning_rate": 4.369128039248699e-06,
+      "loss": 0.3976,
       "step": 113065
     },
     {
-      "epoch": 3.98,
-      "learning_rate": 5.293461814284162e-06,
-      "loss": 0.25,
+      "epoch": 4.075035138933939,
+      "grad_norm": 0.24675941467285156,
+      "learning_rate": 4.36748004799421e-06,
+      "loss": 0.4089,
       "step": 113070
     },
     {
-      "epoch": 3.98,
-      "learning_rate": 5.291709009014706e-06,
-      "loss": 0.2598,
+      "epoch": 4.0752153385951635,
+      "grad_norm": 0.30933111906051636,
+      "learning_rate": 4.365832337853806e-06,
+      "loss": 0.3718,
       "step": 113075
     },
     {
-      "epoch": 3.98,
-      "learning_rate": 5.28995645964441e-06,
-      "loss": 0.2613,
+      "epoch": 4.075395538256388,
+      "grad_norm": 0.2843681871891022,
+      "learning_rate": 4.364184908849947e-06,
+      "loss": 0.3674,
       "step": 113080
     },
     {
-      "epoch": 3.98,
-      "learning_rate": 5.288204166196048e-06,
-      "loss": 0.2532,
+      "epoch": 4.075575737917613,
+      "grad_norm": 0.20400385558605194,
+      "learning_rate": 4.362537761005073e-06,
+      "loss": 0.4067,
       "step": 113085
     },
     {
-      "epoch": 3.98,
-      "learning_rate": 5.286452128692365e-06,
-      "loss": 0.2574,
+      "epoch": 4.075755937578838,
+      "grad_norm": 0.2532506585121155,
+      "learning_rate": 4.3608908943416264e-06,
+      "loss": 0.3598,
       "step": 113090
     },
     {
-      "epoch": 3.98,
-      "learning_rate": 5.284700347156105e-06,
-      "loss": 0.2614,
+      "epoch": 4.0759361372400615,
+      "grad_norm": 0.23435381054878235,
+      "learning_rate": 4.359244308882043e-06,
+      "loss": 0.3748,
       "step": 113095
     },
     {
-      "epoch": 3.98,
-      "learning_rate": 5.28294882161002e-06,
-      "loss": 0.2657,
+      "epoch": 4.076116336901286,
+      "grad_norm": 0.2436995804309845,
+      "learning_rate": 4.357598004648763e-06,
+      "loss": 0.3937,
       "step": 113100
     },
     {
-      "epoch": 3.98,
-      "learning_rate": 5.281197552076847e-06,
-      "loss": 0.2509,
+      "epoch": 4.076296536562511,
+      "grad_norm": 0.26135680079460144,
+      "learning_rate": 4.355951981664208e-06,
+      "loss": 0.3677,
       "step": 113105
     },
     {
-      "epoch": 3.98,
-      "learning_rate": 5.279446538579336e-06,
-      "loss": 0.2641,
+      "epoch": 4.076476736223736,
+      "grad_norm": 0.3116925060749054,
+      "learning_rate": 4.354306239950814e-06,
+      "loss": 0.3898,
       "step": 113110
     },
     {
-      "epoch": 3.98,
-      "learning_rate": 5.2776957811402044e-06,
-      "loss": 0.2451,
+      "epoch": 4.0766569358849605,
+      "grad_norm": 0.25382331013679504,
+      "learning_rate": 4.352660779531001e-06,
+      "loss": 0.4214,
       "step": 113115
     },
     {
-      "epoch": 3.98,
-      "learning_rate": 5.275945279782207e-06,
-      "loss": 0.259,
+      "epoch": 4.076837135546185,
+      "grad_norm": 0.2189731001853943,
+      "learning_rate": 4.3510156004271886e-06,
+      "loss": 0.3645,
       "step": 113120
     },
     {
-      "epoch": 3.98,
-      "learning_rate": 5.27419503452806e-06,
-      "loss": 0.2733,
+      "epoch": 4.07701733520741,
+      "grad_norm": 0.26065686345100403,
+      "learning_rate": 4.3493707026617894e-06,
+      "loss": 0.3576,
       "step": 113125
     },
     {
-      "epoch": 3.98,
-      "learning_rate": 5.272445045400489e-06,
-      "loss": 0.273,
+      "epoch": 4.077197534868635,
+      "grad_norm": 0.25834450125694275,
+      "learning_rate": 4.347726086257212e-06,
+      "loss": 0.3675,
       "step": 113130
     },
     {
-      "epoch": 3.98,
-      "learning_rate": 5.270695312422214e-06,
-      "loss": 0.2471,
+      "epoch": 4.077377734529859,
+      "grad_norm": 0.28615930676460266,
+      "learning_rate": 4.346081751235873e-06,
+      "loss": 0.3697,
       "step": 113135
     },
     {
-      "epoch": 3.98,
-      "learning_rate": 5.268945835615962e-06,
-      "loss": 0.247,
+      "epoch": 4.077557934191084,
+      "grad_norm": 0.27379459142684937,
+      "learning_rate": 4.344437697620174e-06,
+      "loss": 0.3802,
       "step": 113140
     },
     {
-      "epoch": 3.98,
-      "learning_rate": 5.267196615004446e-06,
-      "loss": 0.2455,
+      "epoch": 4.077738133852308,
+      "grad_norm": 0.27598029375076294,
+      "learning_rate": 4.342793925432509e-06,
+      "loss": 0.3636,
       "step": 113145
     },
     {
-      "epoch": 3.98,
-      "learning_rate": 5.265447650610381e-06,
-      "loss": 0.2811,
+      "epoch": 4.077918333513533,
+      "grad_norm": 0.2820260226726532,
+      "learning_rate": 4.341150434695279e-06,
+      "loss": 0.3939,
       "step": 113150
     },
     {
-      "epoch": 3.98,
-      "learning_rate": 5.263698942456463e-06,
-      "loss": 0.255,
+      "epoch": 4.0780985331747575,
+      "grad_norm": 0.25555455684661865,
+      "learning_rate": 4.339507225430867e-06,
+      "loss": 0.3348,
       "step": 113155
     },
     {
-      "epoch": 3.98,
-      "learning_rate": 5.261950490565407e-06,
-      "loss": 0.2532,
+      "epoch": 4.078278732835982,
+      "grad_norm": 0.2568057179450989,
+      "learning_rate": 4.337864297661684e-06,
+      "loss": 0.3276,
       "step": 113160
     },
     {
-      "epoch": 3.98,
-      "learning_rate": 5.260202294959923e-06,
-      "loss": 0.2497,
+      "epoch": 4.078458932497207,
+      "grad_norm": 0.2430773675441742,
+      "learning_rate": 4.336221651410091e-06,
+      "loss": 0.3626,
       "step": 113165
     },
     {
-      "epoch": 3.98,
-      "learning_rate": 5.258454355662704e-06,
-      "loss": 0.2666,
+      "epoch": 4.078639132158432,
+      "grad_norm": 0.26786041259765625,
+      "learning_rate": 4.334579286698473e-06,
+      "loss": 0.3538,
       "step": 113170
     },
     {
-      "epoch": 3.98,
-      "learning_rate": 5.256706672696438e-06,
-      "loss": 0.2644,
+      "epoch": 4.078819331819656,
+      "grad_norm": 0.23630788922309875,
+      "learning_rate": 4.3329372035492145e-06,
+      "loss": 0.3774,
       "step": 113175
     },
     {
-      "epoch": 3.98,
-      "learning_rate": 5.254959246083832e-06,
-      "loss": 0.2605,
+      "epoch": 4.078999531480881,
+      "grad_norm": 0.2112296223640442,
+      "learning_rate": 4.3312954019846894e-06,
+      "loss": 0.3677,
       "step": 113180
     },
     {
-      "epoch": 3.98,
-      "learning_rate": 5.253212075847566e-06,
-      "loss": 0.2268,
+      "epoch": 4.079179731142106,
+      "grad_norm": 0.20694564282894135,
+      "learning_rate": 4.32965388202726e-06,
+      "loss": 0.3376,
       "step": 113185
     },
     {
-      "epoch": 3.98,
-      "learning_rate": 5.251465162010327e-06,
-      "loss": 0.271,
+      "epoch": 4.07935993080333,
+      "grad_norm": 0.19612739980220795,
+      "learning_rate": 4.3280126436992944e-06,
+      "loss": 0.3649,
       "step": 113190
     },
     {
-      "epoch": 3.98,
-      "learning_rate": 5.249718504594794e-06,
-      "loss": 0.2515,
+      "epoch": 4.079540130464554,
+      "grad_norm": 0.2163010984659195,
+      "learning_rate": 4.32637168702315e-06,
+      "loss": 0.3567,
       "step": 113195
     },
     {
-      "epoch": 3.98,
-      "learning_rate": 5.247972103623655e-06,
-      "loss": 0.2726,
+      "epoch": 4.079720330125779,
+      "grad_norm": 0.27228453755378723,
+      "learning_rate": 4.324731012021193e-06,
+      "loss": 0.3773,
       "step": 113200
     },
     {
-      "epoch": 3.98,
-      "learning_rate": 5.246225959119583e-06,
-      "loss": 0.2534,
+      "epoch": 4.079900529787004,
+      "grad_norm": 0.28055036067962646,
+      "learning_rate": 4.323090618715775e-06,
+      "loss": 0.3608,
       "step": 113205
     },
     {
-      "epoch": 3.98,
-      "learning_rate": 5.244480071105243e-06,
-      "loss": 0.258,
+      "epoch": 4.080080729448229,
+      "grad_norm": 0.2570863366127014,
+      "learning_rate": 4.321450507129243e-06,
+      "loss": 0.3704,
       "step": 113210
     },
     {
-      "epoch": 3.98,
-      "learning_rate": 5.242734439603306e-06,
-      "loss": 0.2804,
+      "epoch": 4.080260929109453,
+      "grad_norm": 0.262403666973114,
+      "learning_rate": 4.319810677283945e-06,
+      "loss": 0.3805,
       "step": 113215
     },
     {
-      "epoch": 3.98,
-      "learning_rate": 5.240989064636451e-06,
-      "loss": 0.246,
+      "epoch": 4.080441128770678,
+      "grad_norm": 0.31988710165023804,
+      "learning_rate": 4.318171129202217e-06,
+      "loss": 0.3807,
       "step": 113220
     },
     {
-      "epoch": 3.98,
-      "learning_rate": 5.2392439462273315e-06,
-      "loss": 0.2575,
+      "epoch": 4.080621328431903,
+      "grad_norm": 0.2832499146461487,
+      "learning_rate": 4.316531862906409e-06,
+      "loss": 0.3679,
       "step": 113225
     },
     {
-      "epoch": 3.98,
-      "learning_rate": 5.2374990843985975e-06,
-      "loss": 0.2712,
+      "epoch": 4.080801528093128,
+      "grad_norm": 0.3199155330657959,
+      "learning_rate": 4.314892878418855e-06,
+      "loss": 0.3587,
       "step": 113230
     },
     {
-      "epoch": 3.98,
-      "learning_rate": 5.235754479172925e-06,
-      "loss": 0.2583,
+      "epoch": 4.080981727754351,
+      "grad_norm": 0.3142116069793701,
+      "learning_rate": 4.313254175761869e-06,
+      "loss": 0.3666,
       "step": 113235
     },
     {
-      "epoch": 3.98,
-      "learning_rate": 5.23401013057295e-06,
-      "loss": 0.2454,
+      "epoch": 4.081161927415576,
+      "grad_norm": 0.2442038655281067,
+      "learning_rate": 4.311615754957796e-06,
+      "loss": 0.3704,
       "step": 113240
     },
     {
-      "epoch": 3.98,
-      "learning_rate": 5.2322660386213306e-06,
-      "loss": 0.2527,
+      "epoch": 4.081342127076801,
+      "grad_norm": 0.2437644749879837,
+      "learning_rate": 4.309977616028954e-06,
+      "loss": 0.3603,
       "step": 113245
     },
     {
-      "epoch": 3.98,
-      "learning_rate": 5.2305222033407015e-06,
-      "loss": 0.2681,
+      "epoch": 4.081522326738026,
+      "grad_norm": 0.24333834648132324,
+      "learning_rate": 4.3083397589976535e-06,
+      "loss": 0.3627,
       "step": 113250
     },
     {
-      "epoch": 3.98,
-      "learning_rate": 5.2287786247537215e-06,
-      "loss": 0.2444,
+      "epoch": 4.08170252639925,
+      "grad_norm": 0.2943533957004547,
+      "learning_rate": 4.30670218388623e-06,
+      "loss": 0.3846,
       "step": 113255
     },
     {
-      "epoch": 3.98,
-      "learning_rate": 5.227035302883018e-06,
-      "loss": 0.2593,
+      "epoch": 4.081882726060475,
+      "grad_norm": 0.24343356490135193,
+      "learning_rate": 4.305064890716973e-06,
+      "loss": 0.3757,
       "step": 113260
     },
     {
-      "epoch": 3.98,
-      "learning_rate": 5.225292237751225e-06,
-      "loss": 0.2631,
+      "epoch": 4.0820629257217,
+      "grad_norm": 0.22839723527431488,
+      "learning_rate": 4.3034278795122035e-06,
+      "loss": 0.3877,
       "step": 113265
     },
     {
-      "epoch": 3.99,
-      "learning_rate": 5.223549429380983e-06,
-      "loss": 0.2598,
+      "epoch": 4.0822431253829246,
+      "grad_norm": 0.3169880211353302,
+      "learning_rate": 4.301791150294224e-06,
+      "loss": 0.4014,
       "step": 113270
     },
     {
-      "epoch": 3.99,
-      "learning_rate": 5.221806877794922e-06,
-      "loss": 0.2705,
+      "epoch": 4.082423325044149,
+      "grad_norm": 0.24656805396080017,
+      "learning_rate": 4.300154703085332e-06,
+      "loss": 0.3673,
       "step": 113275
     },
     {
-      "epoch": 3.99,
-      "learning_rate": 5.220064583015666e-06,
-      "loss": 0.2551,
+      "epoch": 4.082603524705373,
+      "grad_norm": 0.2953994870185852,
+      "learning_rate": 4.298518537907826e-06,
+      "loss": 0.3797,
       "step": 113280
     },
     {
-      "epoch": 3.99,
-      "learning_rate": 5.218322545065834e-06,
-      "loss": 0.2478,
+      "epoch": 4.082783724366598,
+      "grad_norm": 0.2698920965194702,
+      "learning_rate": 4.296882654783993e-06,
+      "loss": 0.3937,
       "step": 113285
     },
     {
-      "epoch": 3.99,
-      "learning_rate": 5.216580763968043e-06,
-      "loss": 0.2449,
+      "epoch": 4.082963924027823,
+      "grad_norm": 0.27484050393104553,
+      "learning_rate": 4.295247053736124e-06,
+      "loss": 0.3645,
       "step": 113290
     },
     {
-      "epoch": 3.99,
-      "learning_rate": 5.214839239744917e-06,
-      "loss": 0.2426,
+      "epoch": 4.083144123689047,
+      "grad_norm": 0.22406524419784546,
+      "learning_rate": 4.293611734786507e-06,
+      "loss": 0.3469,
       "step": 113295
     },
     {
-      "epoch": 3.99,
-      "learning_rate": 5.2130979724190644e-06,
-      "loss": 0.2532,
+      "epoch": 4.083324323350272,
+      "grad_norm": 0.22183670103549957,
+      "learning_rate": 4.291976697957425e-06,
+      "loss": 0.3374,
       "step": 113300
     },
     {
-      "epoch": 3.99,
-      "learning_rate": 5.211356962013098e-06,
-      "loss": 0.2717,
+      "epoch": 4.083504523011497,
+      "grad_norm": 0.25425711274147034,
+      "learning_rate": 4.290341943271151e-06,
+      "loss": 0.3478,
       "step": 113305
     },
     {
-      "epoch": 3.99,
-      "learning_rate": 5.209616208549611e-06,
-      "loss": 0.259,
+      "epoch": 4.0836847226727215,
+      "grad_norm": 0.26116448640823364,
+      "learning_rate": 4.288707470749956e-06,
+      "loss": 0.3491,
       "step": 113310
     },
     {
-      "epoch": 3.99,
-      "learning_rate": 5.2078757120512215e-06,
-      "loss": 0.2383,
+      "epoch": 4.083864922333946,
+      "grad_norm": 0.21055485308170319,
+      "learning_rate": 4.287073280416107e-06,
+      "loss": 0.3569,
       "step": 113315
     },
     {
-      "epoch": 3.99,
-      "learning_rate": 5.206135472540519e-06,
-      "loss": 0.2458,
+      "epoch": 4.084045121995171,
+      "grad_norm": 0.24054095149040222,
+      "learning_rate": 4.28543937229188e-06,
+      "loss": 0.3747,
       "step": 113320
     },
     {
-      "epoch": 3.99,
-      "learning_rate": 5.204395490040106e-06,
-      "loss": 0.2585,
+      "epoch": 4.084225321656396,
+      "grad_norm": 0.2076614648103714,
+      "learning_rate": 4.283805746399533e-06,
+      "loss": 0.374,
       "step": 113325
     },
     {
-      "epoch": 3.99,
-      "learning_rate": 5.202655764572567e-06,
-      "loss": 0.2655,
+      "epoch": 4.08440552131762,
+      "grad_norm": 0.2705661356449127,
+      "learning_rate": 4.2821724027613224e-06,
+      "loss": 0.3778,
       "step": 113330
     },
     {
-      "epoch": 3.99,
-      "learning_rate": 5.200916296160502e-06,
-      "loss": 0.2825,
+      "epoch": 4.084585720978844,
+      "grad_norm": 0.20731863379478455,
+      "learning_rate": 4.280539341399497e-06,
+      "loss": 0.3833,
       "step": 113335
     },
     {
-      "epoch": 3.99,
-      "learning_rate": 5.1991770848264894e-06,
-      "loss": 0.2488,
+      "epoch": 4.084765920640069,
+      "grad_norm": 0.2571890950202942,
+      "learning_rate": 4.27890656233631e-06,
+      "loss": 0.3689,
       "step": 113340
     },
     {
-      "epoch": 3.99,
-      "learning_rate": 5.197438130593107e-06,
-      "loss": 0.2558,
+      "epoch": 4.084946120301294,
+      "grad_norm": 0.2876233756542206,
+      "learning_rate": 4.27727406559402e-06,
+      "loss": 0.3643,
       "step": 113345
     },
     {
-      "epoch": 3.99,
-      "learning_rate": 5.195699433482948e-06,
-      "loss": 0.2687,
+      "epoch": 4.0851263199625185,
+      "grad_norm": 0.21950480341911316,
+      "learning_rate": 4.275641851194853e-06,
+      "loss": 0.3595,
       "step": 113350
     },
     {
-      "epoch": 3.99,
-      "learning_rate": 5.193960993518579e-06,
-      "loss": 0.2517,
+      "epoch": 4.085306519623743,
+      "grad_norm": 0.2071562260389328,
+      "learning_rate": 4.274009919161048e-06,
+      "loss": 0.3788,
       "step": 113355
     },
     {
-      "epoch": 3.99,
-      "learning_rate": 5.1922228107225746e-06,
-      "loss": 0.2554,
+      "epoch": 4.085486719284968,
+      "grad_norm": 0.2707284688949585,
+      "learning_rate": 4.272378269514851e-06,
+      "loss": 0.3466,
       "step": 113360
     },
     {
-      "epoch": 3.99,
-      "learning_rate": 5.190484885117497e-06,
-      "loss": 0.2562,
+      "epoch": 4.085666918946193,
+      "grad_norm": 0.22272071242332458,
+      "learning_rate": 4.2707469022784804e-06,
+      "loss": 0.3734,
       "step": 113365
     },
     {
-      "epoch": 3.99,
-      "learning_rate": 5.188747216725925e-06,
-      "loss": 0.2593,
+      "epoch": 4.0858471186074174,
+      "grad_norm": 0.26984214782714844,
+      "learning_rate": 4.269115817474181e-06,
+      "loss": 0.3943,
       "step": 113370
     },
     {
-      "epoch": 3.99,
-      "learning_rate": 5.187009805570406e-06,
-      "loss": 0.2627,
+      "epoch": 4.086027318268641,
+      "grad_norm": 0.23475557565689087,
+      "learning_rate": 4.267485015124162e-06,
+      "loss": 0.4002,
       "step": 113375
     },
     {
-      "epoch": 3.99,
-      "learning_rate": 5.1852726516735175e-06,
-      "loss": 0.2581,
+      "epoch": 4.086207517929866,
+      "grad_norm": 0.2445220947265625,
+      "learning_rate": 4.265854495250638e-06,
+      "loss": 0.3855,
       "step": 113380
     },
     {
-      "epoch": 3.99,
-      "learning_rate": 5.1835357550577945e-06,
-      "loss": 0.2622,
+      "epoch": 4.086387717591091,
+      "grad_norm": 0.3065510392189026,
+      "learning_rate": 4.264224257875837e-06,
+      "loss": 0.3648,
       "step": 113385
     },
     {
-      "epoch": 3.99,
-      "learning_rate": 5.181799115745809e-06,
-      "loss": 0.2618,
+      "epoch": 4.0865679172523155,
+      "grad_norm": 0.25749513506889343,
+      "learning_rate": 4.262594303021966e-06,
+      "loss": 0.3851,
       "step": 113390
     },
     {
-      "epoch": 3.99,
-      "learning_rate": 5.180062733760102e-06,
-      "loss": 0.2673,
+      "epoch": 4.08674811691354,
+      "grad_norm": 0.307963103055954,
+      "learning_rate": 4.260964630711234e-06,
+      "loss": 0.3856,
       "step": 113395
     },
     {
-      "epoch": 3.99,
-      "learning_rate": 5.178326609123216e-06,
-      "loss": 0.2467,
+      "epoch": 4.086928316574765,
+      "grad_norm": 0.2993742525577545,
+      "learning_rate": 4.2593352409658445e-06,
+      "loss": 0.4103,
       "step": 113400
     },
     {
-      "epoch": 3.99,
-      "learning_rate": 5.176590741857689e-06,
-      "loss": 0.28,
+      "epoch": 4.08710851623599,
+      "grad_norm": 0.28788864612579346,
+      "learning_rate": 4.25770613380799e-06,
+      "loss": 0.3724,
       "step": 113405
     },
     {
-      "epoch": 3.99,
-      "learning_rate": 5.174855131986079e-06,
-      "loss": 0.2534,
+      "epoch": 4.087288715897214,
+      "grad_norm": 0.22574250400066376,
+      "learning_rate": 4.2560773092598786e-06,
+      "loss": 0.3732,
       "step": 113410
     },
     {
-      "epoch": 3.99,
-      "learning_rate": 5.173119779530905e-06,
-      "loss": 0.2566,
+      "epoch": 4.087468915558439,
+      "grad_norm": 0.21905584633350372,
+      "learning_rate": 4.2544487673436996e-06,
+      "loss": 0.368,
       "step": 113415
     },
     {
-      "epoch": 3.99,
-      "learning_rate": 5.171384684514705e-06,
-      "loss": 0.2594,
+      "epoch": 4.087649115219663,
+      "grad_norm": 0.20254629850387573,
+      "learning_rate": 4.252820508081637e-06,
+      "loss": 0.3668,
       "step": 113420
     },
     {
-      "epoch": 3.99,
-      "learning_rate": 5.169649846960001e-06,
-      "loss": 0.2377,
+      "epoch": 4.087829314880888,
+      "grad_norm": 0.24264578521251678,
+      "learning_rate": 4.251192531495879e-06,
+      "loss": 0.3633,
       "step": 113425
     },
     {
-      "epoch": 3.99,
-      "learning_rate": 5.1679152668893255e-06,
-      "loss": 0.2462,
+      "epoch": 4.0880095145421125,
+      "grad_norm": 0.25658249855041504,
+      "learning_rate": 4.249564837608608e-06,
+      "loss": 0.3669,
       "step": 113430
     },
     {
-      "epoch": 3.99,
-      "learning_rate": 5.166180944325208e-06,
-      "loss": 0.2307,
+      "epoch": 4.088189714203337,
+      "grad_norm": 0.27399855852127075,
+      "learning_rate": 4.247937426441989e-06,
+      "loss": 0.3615,
       "step": 113435
     },
     {
-      "epoch": 3.99,
-      "learning_rate": 5.164446879290158e-06,
-      "loss": 0.2457,
+      "epoch": 4.088369913864562,
+      "grad_norm": 0.22151198983192444,
+      "learning_rate": 4.2463102980182216e-06,
+      "loss": 0.3419,
       "step": 113440
     },
     {
-      "epoch": 3.99,
-      "learning_rate": 5.162713071806688e-06,
-      "loss": 0.2578,
+      "epoch": 4.088550113525787,
+      "grad_norm": 0.23478932678699493,
+      "learning_rate": 4.244683452359443e-06,
+      "loss": 0.3594,
       "step": 113445
     },
     {
-      "epoch": 3.99,
-      "learning_rate": 5.160979521897322e-06,
-      "loss": 0.2747,
+      "epoch": 4.088730313187011,
+      "grad_norm": 0.19280430674552917,
+      "learning_rate": 4.24305688948784e-06,
+      "loss": 0.3518,
       "step": 113450
     },
     {
-      "epoch": 3.99,
-      "learning_rate": 5.159246229584561e-06,
-      "loss": 0.2453,
+      "epoch": 4.088910512848236,
+      "grad_norm": 0.2612258195877075,
+      "learning_rate": 4.24143060942557e-06,
+      "loss": 0.3779,
       "step": 113455
     },
     {
-      "epoch": 3.99,
-      "learning_rate": 5.157513194890906e-06,
-      "loss": 0.2429,
+      "epoch": 4.089090712509461,
+      "grad_norm": 0.24851427972316742,
+      "learning_rate": 4.239804612194787e-06,
+      "loss": 0.3803,
       "step": 113460
     },
     {
-      "epoch": 3.99,
-      "learning_rate": 5.155780417838874e-06,
-      "loss": 0.2773,
+      "epoch": 4.089270912170685,
+      "grad_norm": 0.24412818253040314,
+      "learning_rate": 4.238178897817646e-06,
+      "loss": 0.3736,
       "step": 113465
     },
     {
-      "epoch": 3.99,
-      "learning_rate": 5.154047898450954e-06,
-      "loss": 0.2238,
+      "epoch": 4.0894511118319095,
+      "grad_norm": 0.2470037341117859,
+      "learning_rate": 4.2365534663163005e-06,
+      "loss": 0.3824,
       "step": 113470
     },
     {
-      "epoch": 3.99,
-      "learning_rate": 5.152315636749641e-06,
-      "loss": 0.2749,
+      "epoch": 4.089631311493134,
+      "grad_norm": 0.27560508251190186,
+      "learning_rate": 4.234928317712888e-06,
+      "loss": 0.3997,
       "step": 113475
     },
     {
-      "epoch": 3.99,
-      "learning_rate": 5.150583632757427e-06,
-      "loss": 0.2671,
+      "epoch": 4.089811511154359,
+      "grad_norm": 0.29321983456611633,
+      "learning_rate": 4.233303452029561e-06,
+      "loss": 0.3753,
       "step": 113480
     },
     {
-      "epoch": 3.99,
-      "learning_rate": 5.148851886496809e-06,
-      "loss": 0.2504,
+      "epoch": 4.089991710815584,
+      "grad_norm": 0.26840469241142273,
+      "learning_rate": 4.231678869288455e-06,
+      "loss": 0.3921,
       "step": 113485
     },
     {
-      "epoch": 3.99,
-      "learning_rate": 5.147120397990257e-06,
-      "loss": 0.2435,
+      "epoch": 4.090171910476808,
+      "grad_norm": 0.18927091360092163,
+      "learning_rate": 4.230054569511705e-06,
+      "loss": 0.3568,
       "step": 113490
     },
     {
-      "epoch": 3.99,
-      "learning_rate": 5.145389167260273e-06,
-      "loss": 0.2397,
+      "epoch": 4.090352110138033,
+      "grad_norm": 0.2522876262664795,
+      "learning_rate": 4.228430552721438e-06,
+      "loss": 0.351,
       "step": 113495
     },
     {
-      "epoch": 3.99,
-      "learning_rate": 5.143658194329318e-06,
-      "loss": 0.2781,
+      "epoch": 4.090532309799258,
+      "grad_norm": 0.30679717659950256,
+      "learning_rate": 4.226806818939777e-06,
+      "loss": 0.3625,
       "step": 113500
     },
     {
-      "epoch": 3.99,
-      "eval_loss": 0.2501179575920105,
-      "eval_runtime": 10.5566,
-      "eval_samples_per_second": 9.473,
-      "eval_steps_per_second": 9.473,
+      "epoch": 4.090532309799258,
+      "eval_loss": 0.4292847514152527,
+      "eval_runtime": 3.5606,
+      "eval_samples_per_second": 28.085,
+      "eval_steps_per_second": 7.021,
       "step": 113500
     },
     {
-      "epoch": 3.99,
-      "learning_rate": 5.14192747921988e-06,
-      "loss": 0.2406,
+      "epoch": 4.090712509460483,
+      "grad_norm": 0.28912264108657837,
+      "learning_rate": 4.225183368188859e-06,
+      "loss": 0.3747,
       "step": 113505
     },
     {
-      "epoch": 3.99,
-      "learning_rate": 5.1401970219544296e-06,
-      "loss": 0.2449,
+      "epoch": 4.090892709121706,
+      "grad_norm": 0.2808322012424469,
+      "learning_rate": 4.223560200490801e-06,
+      "loss": 0.355,
       "step": 113510
     },
     {
-      "epoch": 3.99,
-      "learning_rate": 5.138466822555432e-06,
-      "loss": 0.2581,
+      "epoch": 4.091072908782931,
+      "grad_norm": 0.27322959899902344,
+      "learning_rate": 4.221937315867702e-06,
+      "loss": 0.3809,
       "step": 113515
     },
     {
-      "epoch": 3.99,
-      "learning_rate": 5.136736881045348e-06,
-      "loss": 0.2583,
+      "epoch": 4.091253108444156,
+      "grad_norm": 0.22654719650745392,
+      "learning_rate": 4.220314714341689e-06,
+      "loss": 0.3452,
       "step": 113520
     },
     {
-      "epoch": 3.99,
-      "learning_rate": 5.13500719744665e-06,
-      "loss": 0.2618,
+      "epoch": 4.091433308105381,
+      "grad_norm": 0.28565314412117004,
+      "learning_rate": 4.218692395934856e-06,
+      "loss": 0.3838,
       "step": 113525
     },
     {
-      "epoch": 3.99,
-      "learning_rate": 5.133277771781794e-06,
-      "loss": 0.2484,
+      "epoch": 4.091613507766605,
+      "grad_norm": 0.24002893269062042,
+      "learning_rate": 4.217070360669325e-06,
+      "loss": 0.3882,
       "step": 113530
     },
     {
-      "epoch": 3.99,
-      "learning_rate": 5.131548604073225e-06,
-      "loss": 0.2693,
+      "epoch": 4.09179370742783,
+      "grad_norm": 0.28343817591667175,
+      "learning_rate": 4.215448608567193e-06,
+      "loss": 0.3902,
       "step": 113535
     },
     {
-      "epoch": 3.99,
-      "learning_rate": 5.12981969434341e-06,
-      "loss": 0.247,
+      "epoch": 4.091973907089055,
+      "grad_norm": 0.2013174593448639,
+      "learning_rate": 4.213827139650536e-06,
+      "loss": 0.3459,
       "step": 113540
     },
     {
-      "epoch": 3.99,
-      "learning_rate": 5.128091042614785e-06,
-      "loss": 0.2388,
+      "epoch": 4.09215410675028,
+      "grad_norm": 0.3252928555011749,
+      "learning_rate": 4.212205953941467e-06,
+      "loss": 0.3473,
       "step": 113545
     },
     {
-      "epoch": 4.0,
-      "learning_rate": 5.1263626489098106e-06,
-      "loss": 0.2597,
+      "epoch": 4.092334306411504,
+      "grad_norm": 0.20931507647037506,
+      "learning_rate": 4.2105850514620625e-06,
+      "loss": 0.3574,
       "step": 113550
     },
     {
-      "epoch": 4.0,
-      "learning_rate": 5.124634513250917e-06,
-      "loss": 0.2461,
+      "epoch": 4.092514506072728,
+      "grad_norm": 0.21451285481452942,
+      "learning_rate": 4.208964432234422e-06,
+      "loss": 0.4097,
       "step": 113555
     },
     {
-      "epoch": 4.0,
-      "learning_rate": 5.12290663566054e-06,
-      "loss": 0.2549,
+      "epoch": 4.092694705733953,
+      "grad_norm": 0.23355019092559814,
+      "learning_rate": 4.20734409628061e-06,
+      "loss": 0.348,
       "step": 113560
     },
     {
-      "epoch": 4.0,
-      "learning_rate": 5.121179016161129e-06,
-      "loss": 0.2586,
+      "epoch": 4.092874905395178,
+      "grad_norm": 0.21140533685684204,
+      "learning_rate": 4.2057240436227025e-06,
+      "loss": 0.3602,
       "step": 113565
     },
     {
-      "epoch": 4.0,
-      "learning_rate": 5.119451654775106e-06,
-      "loss": 0.2343,
+      "epoch": 4.093055105056402,
+      "grad_norm": 0.22797791659832,
+      "learning_rate": 4.2041042742827855e-06,
+      "loss": 0.3894,
       "step": 113570
     },
     {
-      "epoch": 4.0,
-      "learning_rate": 5.117724551524897e-06,
-      "loss": 0.2777,
+      "epoch": 4.093235304717627,
+      "grad_norm": 0.2819134593009949,
+      "learning_rate": 4.202484788282923e-06,
+      "loss": 0.3677,
       "step": 113575
     },
     {
-      "epoch": 4.0,
-      "learning_rate": 5.115997706432937e-06,
-      "loss": 0.2766,
+      "epoch": 4.093415504378852,
+      "grad_norm": 0.27418237924575806,
+      "learning_rate": 4.200865585645178e-06,
+      "loss": 0.3588,
       "step": 113580
     },
     {
-      "epoch": 4.0,
-      "learning_rate": 5.114271119521644e-06,
-      "loss": 0.2655,
+      "epoch": 4.0935957040400766,
+      "grad_norm": 0.258755087852478,
+      "learning_rate": 4.19924666639161e-06,
+      "loss": 0.3634,
       "step": 113585
     },
     {
-      "epoch": 4.0,
-      "learning_rate": 5.1125447908134335e-06,
-      "loss": 0.2774,
+      "epoch": 4.093775903701301,
+      "grad_norm": 0.2556750476360321,
+      "learning_rate": 4.197628030544276e-06,
+      "loss": 0.3824,
       "step": 113590
     },
     {
-      "epoch": 4.0,
-      "learning_rate": 5.110818720330715e-06,
-      "loss": 0.2562,
+      "epoch": 4.093956103362526,
+      "grad_norm": 0.19411814212799072,
+      "learning_rate": 4.196009678125237e-06,
+      "loss": 0.3567,
       "step": 113595
     },
     {
-      "epoch": 4.0,
-      "learning_rate": 5.109092908095911e-06,
-      "loss": 0.2677,
+      "epoch": 4.094136303023751,
+      "grad_norm": 0.23082207143306732,
+      "learning_rate": 4.19439160915654e-06,
+      "loss": 0.3516,
       "step": 113600
     },
     {
-      "epoch": 4.0,
-      "learning_rate": 5.107367354131432e-06,
-      "loss": 0.2523,
+      "epoch": 4.094316502684975,
+      "grad_norm": 0.26558494567871094,
+      "learning_rate": 4.192773823660229e-06,
+      "loss": 0.4062,
       "step": 113605
     },
     {
-      "epoch": 4.0,
-      "learning_rate": 5.105642058459678e-06,
-      "loss": 0.2611,
+      "epoch": 4.094496702346199,
+      "grad_norm": 0.2534783184528351,
+      "learning_rate": 4.191156321658343e-06,
+      "loss": 0.3816,
       "step": 113610
     },
     {
-      "epoch": 4.0,
-      "learning_rate": 5.103917021103044e-06,
-      "loss": 0.2413,
+      "epoch": 4.094676902007424,
+      "grad_norm": 0.2528476417064667,
+      "learning_rate": 4.189539103172926e-06,
+      "loss": 0.355,
       "step": 113615
     },
     {
-      "epoch": 4.0,
-      "learning_rate": 5.102192242083945e-06,
-      "loss": 0.2559,
+      "epoch": 4.094857101668649,
+      "grad_norm": 0.29152292013168335,
+      "learning_rate": 4.1879221682260054e-06,
+      "loss": 0.3748,
       "step": 113620
     },
     {
-      "epoch": 4.0,
-      "learning_rate": 5.100467721424765e-06,
-      "loss": 0.2795,
+      "epoch": 4.0950373013298735,
+      "grad_norm": 0.22224609553813934,
+      "learning_rate": 4.186305516839625e-06,
+      "loss": 0.3615,
       "step": 113625
     },
     {
-      "epoch": 4.0,
-      "learning_rate": 5.098743459147895e-06,
-      "loss": 0.2676,
+      "epoch": 4.095217500991098,
+      "grad_norm": 0.2683086395263672,
+      "learning_rate": 4.184689149035792e-06,
+      "loss": 0.4004,
       "step": 113630
     },
     {
-      "epoch": 4.0,
-      "learning_rate": 5.09701945527572e-06,
-      "loss": 0.2559,
+      "epoch": 4.095397700652323,
+      "grad_norm": 0.2543949782848358,
+      "learning_rate": 4.183073064836545e-06,
+      "loss": 0.3536,
       "step": 113635
     },
     {
-      "epoch": 4.0,
-      "learning_rate": 5.095295709830638e-06,
-      "loss": 0.2421,
+      "epoch": 4.095577900313548,
+      "grad_norm": 0.28799569606781006,
+      "learning_rate": 4.181457264263897e-06,
+      "loss": 0.3799,
       "step": 113640
     },
     {
-      "epoch": 4.0,
-      "learning_rate": 5.093572222835025e-06,
-      "loss": 0.2599,
+      "epoch": 4.0957580999747725,
+      "grad_norm": 0.21432629227638245,
+      "learning_rate": 4.179841747339864e-06,
+      "loss": 0.3454,
       "step": 113645
     },
     {
-      "epoch": 4.0,
-      "learning_rate": 5.091848994311252e-06,
-      "loss": 0.253,
+      "epoch": 4.095938299635996,
+      "grad_norm": 0.256199449300766,
+      "learning_rate": 4.178226514086453e-06,
+      "loss": 0.3815,
       "step": 113650
     },
     {
-      "epoch": 4.0,
-      "learning_rate": 5.090126024281697e-06,
-      "loss": 0.2594,
+      "epoch": 4.096118499297221,
+      "grad_norm": 0.19477510452270508,
+      "learning_rate": 4.176611564525679e-06,
+      "loss": 0.3582,
       "step": 113655
     },
     {
-      "epoch": 4.0,
-      "learning_rate": 5.088403312768741e-06,
-      "loss": 0.275,
+      "epoch": 4.096298698958446,
+      "grad_norm": 0.2505654990673065,
+      "learning_rate": 4.1749968986795316e-06,
+      "loss": 0.3512,
       "step": 113660
     },
     {
-      "epoch": 4.0,
-      "learning_rate": 5.086680859794748e-06,
-      "loss": 0.2643,
+      "epoch": 4.0964788986196705,
+      "grad_norm": 0.26357826590538025,
+      "learning_rate": 4.173382516570026e-06,
+      "loss": 0.361,
       "step": 113665
     },
     {
-      "epoch": 4.0,
-      "learning_rate": 5.084958665382078e-06,
-      "loss": 0.2585,
+      "epoch": 4.096659098280895,
+      "grad_norm": 0.22375109791755676,
+      "learning_rate": 4.171768418219152e-06,
+      "loss": 0.3643,
       "step": 113670
     },
     {
-      "epoch": 4.0,
-      "learning_rate": 5.08323672955309e-06,
-      "loss": 0.2616,
+      "epoch": 4.09683929794212,
+      "grad_norm": 0.2819434404373169,
+      "learning_rate": 4.170154603648901e-06,
+      "loss": 0.3794,
       "step": 113675
     },
     {
-      "epoch": 4.0,
-      "learning_rate": 5.081515052330154e-06,
-      "loss": 0.2442,
+      "epoch": 4.097019497603345,
+      "grad_norm": 0.24838387966156006,
+      "learning_rate": 4.168541072881263e-06,
+      "loss": 0.3224,
       "step": 113680
     },
     {
-      "epoch": 4.0,
-      "learning_rate": 5.079793633735619e-06,
-      "loss": 0.2518,
+      "epoch": 4.0971996972645695,
+      "grad_norm": 0.2065441608428955,
+      "learning_rate": 4.1669278259382104e-06,
+      "loss": 0.343,
       "step": 113685
     },
     {
-      "epoch": 4.0,
-      "learning_rate": 5.0780724737918285e-06,
-      "loss": 0.2479,
+      "epoch": 4.097379896925794,
+      "grad_norm": 0.2739092707633972,
+      "learning_rate": 4.16531486284174e-06,
+      "loss": 0.4081,
       "step": 113690
     },
     {
-      "epoch": 4.0,
-      "learning_rate": 5.076351572521146e-06,
-      "loss": 0.2412,
+      "epoch": 4.097560096587018,
+      "grad_norm": 0.2650372385978699,
+      "learning_rate": 4.1637021836138215e-06,
+      "loss": 0.3675,
       "step": 113695
     },
     {
-      "epoch": 4.0,
-      "learning_rate": 5.074630929945906e-06,
-      "loss": 0.2516,
+      "epoch": 4.097740296248243,
+      "grad_norm": 0.23538993299007416,
+      "learning_rate": 4.162089788276424e-06,
+      "loss": 0.397,
       "step": 113700
     },
     {
-      "epoch": 4.0,
-      "learning_rate": 5.0729105460884464e-06,
-      "loss": 0.2561,
+      "epoch": 4.0979204959094675,
+      "grad_norm": 0.2418733686208725,
+      "learning_rate": 4.160477676851523e-06,
+      "loss": 0.3946,
       "step": 113705
     },
     {
-      "epoch": 4.0,
-      "learning_rate": 5.071190420971112e-06,
-      "loss": 0.2434,
+      "epoch": 4.098100695570692,
+      "grad_norm": 0.2708457112312317,
+      "learning_rate": 4.158865849361071e-06,
+      "loss": 0.3647,
       "step": 113710
     },
     {
-      "epoch": 4.0,
-      "learning_rate": 5.06947055461624e-06,
-      "loss": 0.2354,
+      "epoch": 4.098280895231917,
+      "grad_norm": 0.24828830361366272,
+      "learning_rate": 4.157254305827041e-06,
+      "loss": 0.3341,
       "step": 113715
     },
     {
-      "epoch": 4.0,
-      "learning_rate": 5.06775094704616e-06,
-      "loss": 0.2413,
+      "epoch": 4.098461094893142,
+      "grad_norm": 0.27252304553985596,
+      "learning_rate": 4.1556430462713954e-06,
+      "loss": 0.3764,
       "step": 113720
     },
     {
-      "epoch": 4.0,
-      "learning_rate": 5.066031598283197e-06,
-      "loss": 0.2387,
+      "epoch": 4.098641294554366,
+      "grad_norm": 0.21500465273857117,
+      "learning_rate": 4.154032070716063e-06,
+      "loss": 0.3764,
       "step": 113725
     },
     {
-      "epoch": 4.0,
-      "learning_rate": 5.064312508349672e-06,
-      "loss": 0.2448,
+      "epoch": 4.098821494215591,
+      "grad_norm": 0.2509201467037201,
+      "learning_rate": 4.152421379183013e-06,
+      "loss": 0.3784,
       "step": 113730
     },
     {
-      "epoch": 4.0,
-      "learning_rate": 5.062593677267916e-06,
-      "loss": 0.2661,
+      "epoch": 4.099001693876816,
+      "grad_norm": 0.24733483791351318,
+      "learning_rate": 4.150810971694183e-06,
+      "loss": 0.3653,
       "step": 113735
     },
     {
-      "epoch": 4.0,
-      "learning_rate": 5.060875105060242e-06,
-      "loss": 0.2472,
+      "epoch": 4.09918189353804,
+      "grad_norm": 0.22112403810024261,
+      "learning_rate": 4.1492008482715254e-06,
+      "loss": 0.3734,
       "step": 113740
     },
     {
-      "epoch": 4.0,
-      "learning_rate": 5.059156791748965e-06,
-      "loss": 0.2515,
+      "epoch": 4.0993620931992645,
+      "grad_norm": 0.22997091710567474,
+      "learning_rate": 4.147591008936966e-06,
+      "loss": 0.3426,
       "step": 113745
     },
     {
-      "epoch": 4.0,
-      "learning_rate": 5.05743873735639e-06,
-      "loss": 0.258,
+      "epoch": 4.099542292860489,
+      "grad_norm": 0.21955984830856323,
+      "learning_rate": 4.145981453712436e-06,
+      "loss": 0.3695,
       "step": 113750
     },
     {
-      "epoch": 4.0,
-      "learning_rate": 5.055720941904835e-06,
-      "loss": 0.2551,
+      "epoch": 4.099722492521714,
+      "grad_norm": 0.2155844122171402,
+      "learning_rate": 4.144372182619874e-06,
+      "loss": 0.3533,
       "step": 113755
     },
     {
-      "epoch": 4.0,
-      "learning_rate": 5.0540034054165934e-06,
-      "loss": 0.2431,
+      "epoch": 4.099902692182939,
+      "grad_norm": 0.2931585907936096,
+      "learning_rate": 4.1430849703355755e-06,
+      "loss": 0.4126,
       "step": 113760
     },
     {
-      "epoch": 4.0,
-      "learning_rate": 5.05228612791398e-06,
-      "loss": 0.2497,
+      "epoch": 4.100082891844163,
+      "grad_norm": 0.23086297512054443,
+      "learning_rate": 4.141476210735803e-06,
+      "loss": 0.3405,
       "step": 113765
     },
     {
-      "epoch": 4.0,
-      "learning_rate": 5.050569109419279e-06,
-      "loss": 0.2363,
+      "epoch": 4.100263091505388,
+      "grad_norm": 0.259778767824173,
+      "learning_rate": 4.139867735329381e-06,
+      "loss": 0.3706,
       "step": 113770
     },
     {
-      "epoch": 4.0,
-      "learning_rate": 5.048852349954797e-06,
-      "loss": 0.2531,
+      "epoch": 4.100443291166613,
+      "grad_norm": 0.22859430313110352,
+      "learning_rate": 4.13825954413822e-06,
+      "loss": 0.3301,
       "step": 113775
     },
     {
-      "epoch": 4.0,
-      "learning_rate": 5.04713584954282e-06,
-      "loss": 0.2519,
+      "epoch": 4.100623490827838,
+      "grad_norm": 0.25174832344055176,
+      "learning_rate": 4.13665163718423e-06,
+      "loss": 0.3392,
       "step": 113780
     },
     {
-      "epoch": 4.0,
-      "learning_rate": 5.045419608205632e-06,
-      "loss": 0.2519,
+      "epoch": 4.1008036904890615,
+      "grad_norm": 0.2478947937488556,
+      "learning_rate": 4.135044014489326e-06,
+      "loss": 0.3599,
       "step": 113785
     },
     {
-      "epoch": 4.0,
-      "learning_rate": 5.0437036259655155e-06,
-      "loss": 0.2466,
+      "epoch": 4.100983890150286,
+      "grad_norm": 0.2401927262544632,
+      "learning_rate": 4.133436676075414e-06,
+      "loss": 0.3812,
       "step": 113790
     },
     {
-      "epoch": 4.0,
-      "learning_rate": 5.041987902844761e-06,
-      "loss": 0.2551,
+      "epoch": 4.101164089811511,
+      "grad_norm": 0.2504352331161499,
+      "learning_rate": 4.131829621964378e-06,
+      "loss": 0.4121,
       "step": 113795
     },
     {
-      "epoch": 4.0,
-      "learning_rate": 5.040272438865642e-06,
-      "loss": 0.2258,
+      "epoch": 4.101344289472736,
+      "grad_norm": 0.21674078702926636,
+      "learning_rate": 4.130222852178129e-06,
+      "loss": 0.323,
       "step": 113800
     },
     {
-      "epoch": 4.0,
-      "learning_rate": 5.038557234050423e-06,
-      "loss": 0.2406,
+      "epoch": 4.10152448913396,
+      "grad_norm": 0.2414158135652542,
+      "learning_rate": 4.128616366738544e-06,
+      "loss": 0.3697,
       "step": 113805
     },
     {
-      "epoch": 4.0,
-      "learning_rate": 5.036842288421392e-06,
-      "loss": 0.2453,
+      "epoch": 4.101704688795185,
+      "grad_norm": 0.2412973940372467,
+      "learning_rate": 4.127010165667533e-06,
+      "loss": 0.3529,
       "step": 113810
     },
     {
-      "epoch": 4.0,
-      "learning_rate": 5.0351276020008e-06,
-      "loss": 0.2551,
+      "epoch": 4.10188488845641,
+      "grad_norm": 0.24448582530021667,
+      "learning_rate": 4.125404248986961e-06,
+      "loss": 0.3587,
       "step": 113815
     },
     {
-      "epoch": 4.0,
-      "learning_rate": 5.0334131748109245e-06,
-      "loss": 0.254,
+      "epoch": 4.102065088117635,
+      "grad_norm": 0.2748193144798279,
+      "learning_rate": 4.12379861671871e-06,
+      "loss": 0.374,
       "step": 113820
     },
     {
-      "epoch": 4.0,
-      "learning_rate": 5.0316990068740165e-06,
-      "loss": 0.2592,
+      "epoch": 4.102245287778859,
+      "grad_norm": 0.28702595829963684,
+      "learning_rate": 4.1221932688846664e-06,
+      "loss": 0.3794,
       "step": 113825
     },
     {
-      "epoch": 4.0,
-      "learning_rate": 5.029985098212342e-06,
-      "loss": 0.2589,
+      "epoch": 4.102425487440083,
+      "grad_norm": 0.23407040536403656,
+      "learning_rate": 4.120588205506698e-06,
+      "loss": 0.3652,
       "step": 113830
     },
     {
-      "epoch": 4.01,
-      "learning_rate": 5.028271448848149e-06,
-      "loss": 0.257,
+      "epoch": 4.102605687101308,
+      "grad_norm": 0.28252798318862915,
+      "learning_rate": 4.1189834266066735e-06,
+      "loss": 0.3586,
       "step": 113835
     },
     {
-      "epoch": 4.01,
-      "learning_rate": 5.026558058803691e-06,
-      "loss": 0.2377,
+      "epoch": 4.102785886762533,
+      "grad_norm": 0.2878764569759369,
+      "learning_rate": 4.117378932206456e-06,
+      "loss": 0.4031,
       "step": 113840
     },
     {
-      "epoch": 4.01,
-      "learning_rate": 5.024844928101208e-06,
-      "loss": 0.2496,
+      "epoch": 4.102966086423757,
+      "grad_norm": 0.21540088951587677,
+      "learning_rate": 4.1157747223279034e-06,
+      "loss": 0.3806,
       "step": 113845
     },
     {
-      "epoch": 4.01,
-      "learning_rate": 5.023132056762953e-06,
-      "loss": 0.2402,
+      "epoch": 4.103146286084982,
+      "grad_norm": 0.24554520845413208,
+      "learning_rate": 4.114170796992881e-06,
+      "loss": 0.3937,
       "step": 113850
     },
     {
-      "epoch": 4.01,
-      "learning_rate": 5.021419444811165e-06,
-      "loss": 0.2561,
+      "epoch": 4.103326485746207,
+      "grad_norm": 0.2406504899263382,
+      "learning_rate": 4.112567156223237e-06,
+      "loss": 0.3786,
       "step": 113855
     },
     {
-      "epoch": 4.01,
-      "learning_rate": 5.01970709226808e-06,
-      "loss": 0.259,
+      "epoch": 4.103506685407432,
+      "grad_norm": 0.25493624806404114,
+      "learning_rate": 4.110963800040824e-06,
+      "loss": 0.3673,
       "step": 113860
     },
     {
-      "epoch": 4.01,
-      "learning_rate": 5.017994999155923e-06,
-      "loss": 0.2348,
+      "epoch": 4.103686885068656,
+      "grad_norm": 0.20379377901554108,
+      "learning_rate": 4.109360728467485e-06,
+      "loss": 0.3536,
       "step": 113865
     },
     {
-      "epoch": 4.01,
-      "learning_rate": 5.016283165496933e-06,
-      "loss": 0.2411,
+      "epoch": 4.103867084729881,
+      "grad_norm": 0.25673457980155945,
+      "learning_rate": 4.107757941525059e-06,
+      "loss": 0.3743,
       "step": 113870
     },
     {
-      "epoch": 4.01,
-      "learning_rate": 5.014571591313341e-06,
-      "loss": 0.2645,
+      "epoch": 4.104047284391106,
+      "grad_norm": 0.2972078025341034,
+      "learning_rate": 4.1061554392353816e-06,
+      "loss": 0.3751,
       "step": 113875
     },
     {
-      "epoch": 4.01,
-      "learning_rate": 5.012860276627368e-06,
-      "loss": 0.2558,
+      "epoch": 4.10422748405233,
+      "grad_norm": 0.2244071513414383,
+      "learning_rate": 4.104553221620297e-06,
+      "loss": 0.3879,
       "step": 113880
     },
     {
-      "epoch": 4.01,
-      "learning_rate": 5.0111492214612254e-06,
-      "loss": 0.2518,
+      "epoch": 4.104407683713554,
+      "grad_norm": 0.20610791444778442,
+      "learning_rate": 4.1029512887016286e-06,
+      "loss": 0.3761,
       "step": 113885
     },
     {
-      "epoch": 4.01,
-      "learning_rate": 5.009438425837143e-06,
-      "loss": 0.2665,
+      "epoch": 4.104587883374779,
+      "grad_norm": 0.23367708921432495,
+      "learning_rate": 4.101349640501206e-06,
+      "loss": 0.3863,
       "step": 113890
     },
     {
-      "epoch": 4.01,
-      "learning_rate": 5.007727889777327e-06,
-      "loss": 0.2439,
+      "epoch": 4.104768083036004,
+      "grad_norm": 0.26050618290901184,
+      "learning_rate": 4.099748277040846e-06,
+      "loss": 0.4031,
       "step": 113895
     },
     {
-      "epoch": 4.01,
-      "learning_rate": 5.006017613303993e-06,
-      "loss": 0.2468,
+      "epoch": 4.104948282697229,
+      "grad_norm": 0.2746144235134125,
+      "learning_rate": 4.098147198342364e-06,
+      "loss": 0.3751,
       "step": 113900
     },
     {
-      "epoch": 4.01,
-      "learning_rate": 5.004307596439334e-06,
-      "loss": 0.2549,
+      "epoch": 4.105128482358453,
+      "grad_norm": 0.22158581018447876,
+      "learning_rate": 4.09654640442759e-06,
+      "loss": 0.3932,
       "step": 113905
     },
     {
-      "epoch": 4.01,
-      "learning_rate": 5.002597839205572e-06,
-      "loss": 0.2707,
+      "epoch": 4.105308682019678,
+      "grad_norm": 0.21861512959003448,
+      "learning_rate": 4.09494589531832e-06,
+      "loss": 0.3805,
       "step": 113910
     },
     {
-      "epoch": 4.01,
-      "learning_rate": 5.000888341624896e-06,
-      "loss": 0.2407,
+      "epoch": 4.105488881680903,
+      "grad_norm": 0.2370881885290146,
+      "learning_rate": 4.0933456710363585e-06,
+      "loss": 0.3719,
       "step": 113915
     },
     {
-      "epoch": 4.01,
-      "learning_rate": 4.999179103719501e-06,
-      "loss": 0.2382,
+      "epoch": 4.1056690813421275,
+      "grad_norm": 0.23769058287143707,
+      "learning_rate": 4.091745731603519e-06,
+      "loss": 0.3728,
       "step": 113920
     },
     {
-      "epoch": 4.01,
-      "learning_rate": 4.997470125511588e-06,
-      "loss": 0.2536,
+      "epoch": 4.105849281003351,
+      "grad_norm": 0.19477471709251404,
+      "learning_rate": 4.090146077041598e-06,
+      "loss": 0.3491,
       "step": 113925
     },
     {
-      "epoch": 4.01,
-      "learning_rate": 4.995761407023339e-06,
-      "loss": 0.2392,
+      "epoch": 4.106029480664576,
+      "grad_norm": 0.310822457075119,
+      "learning_rate": 4.088546707372387e-06,
+      "loss": 0.3381,
       "step": 113930
     },
     {
-      "epoch": 4.01,
-      "learning_rate": 4.994052948276948e-06,
-      "loss": 0.2406,
+      "epoch": 4.106209680325801,
+      "grad_norm": 0.25079676508903503,
+      "learning_rate": 4.086947622617682e-06,
+      "loss": 0.3781,
       "step": 113935
     },
     {
-      "epoch": 4.01,
-      "learning_rate": 4.992344749294592e-06,
-      "loss": 0.2357,
+      "epoch": 4.1063898799870255,
+      "grad_norm": 0.2638763189315796,
+      "learning_rate": 4.085348822799257e-06,
+      "loss": 0.4023,
       "step": 113940
     },
     {
-      "epoch": 4.01,
-      "learning_rate": 4.990636810098459e-06,
-      "loss": 0.2519,
+      "epoch": 4.10657007964825,
+      "grad_norm": 0.25837767124176025,
+      "learning_rate": 4.083750307938911e-06,
+      "loss": 0.3652,
       "step": 113945
     },
     {
-      "epoch": 4.01,
-      "learning_rate": 4.98892913071072e-06,
-      "loss": 0.2271,
+      "epoch": 4.106750279309475,
+      "grad_norm": 0.2736017405986786,
+      "learning_rate": 4.082152078058419e-06,
+      "loss": 0.3973,
       "step": 113950
     },
     {
-      "epoch": 4.01,
-      "learning_rate": 4.98722171115355e-06,
-      "loss": 0.2694,
+      "epoch": 4.1069304789707,
+      "grad_norm": 0.2712678611278534,
+      "learning_rate": 4.080554133179554e-06,
+      "loss": 0.381,
       "step": 113955
     },
     {
-      "epoch": 4.01,
-      "learning_rate": 4.985514551449108e-06,
-      "loss": 0.2582,
+      "epoch": 4.1071106786319245,
+      "grad_norm": 0.26710009574890137,
+      "learning_rate": 4.078956473324091e-06,
+      "loss": 0.3954,
       "step": 113960
     },
     {
-      "epoch": 4.01,
-      "learning_rate": 4.9838076516195796e-06,
-      "loss": 0.2609,
+      "epoch": 4.107290878293149,
+      "grad_norm": 0.2664423882961273,
+      "learning_rate": 4.077359098513789e-06,
+      "loss": 0.3671,
       "step": 113965
     },
     {
-      "epoch": 4.01,
-      "learning_rate": 4.982101011687115e-06,
-      "loss": 0.2533,
+      "epoch": 4.107471077954373,
+      "grad_norm": 0.2606585919857025,
+      "learning_rate": 4.075762008770423e-06,
+      "loss": 0.4075,
       "step": 113970
     },
     {
-      "epoch": 4.01,
-      "learning_rate": 4.980394631673871e-06,
-      "loss": 0.2564,
+      "epoch": 4.107651277615598,
+      "grad_norm": 0.26195093989372253,
+      "learning_rate": 4.074165204115754e-06,
+      "loss": 0.3666,
       "step": 113975
     },
     {
-      "epoch": 4.01,
-      "learning_rate": 4.978688511602014e-06,
-      "loss": 0.2653,
+      "epoch": 4.1078314772768225,
+      "grad_norm": 0.23841117322444916,
+      "learning_rate": 4.072568684571524e-06,
+      "loss": 0.3521,
       "step": 113980
     },
     {
-      "epoch": 4.01,
-      "learning_rate": 4.976982651493689e-06,
-      "loss": 0.2505,
+      "epoch": 4.108011676938047,
+      "grad_norm": 0.24487237632274628,
+      "learning_rate": 4.070972450159497e-06,
+      "loss": 0.3861,
       "step": 113985
     },
     {
-      "epoch": 4.01,
-      "learning_rate": 4.9752770513710524e-06,
-      "loss": 0.2538,
+      "epoch": 4.108191876599272,
+      "grad_norm": 0.22291886806488037,
+      "learning_rate": 4.069376500901414e-06,
+      "loss": 0.3597,
       "step": 113990
     },
     {
-      "epoch": 4.01,
-      "learning_rate": 4.9735717112562475e-06,
-      "loss": 0.2483,
+      "epoch": 4.108372076260497,
+      "grad_norm": 0.2304575890302658,
+      "learning_rate": 4.06778083681903e-06,
+      "loss": 0.362,
       "step": 113995
     },
     {
-      "epoch": 4.01,
-      "learning_rate": 4.971866631171409e-06,
-      "loss": 0.2482,
+      "epoch": 4.1085522759217215,
+      "grad_norm": 0.2351878434419632,
+      "learning_rate": 4.066185457934083e-06,
+      "loss": 0.3816,
       "step": 114000
     },
     {
-      "epoch": 4.01,
-      "eval_loss": 0.25011008977890015,
-      "eval_runtime": 10.5643,
-      "eval_samples_per_second": 9.466,
-      "eval_steps_per_second": 9.466,
+      "epoch": 4.1085522759217215,
+      "eval_loss": 0.4291282296180725,
+      "eval_runtime": 3.5373,
+      "eval_samples_per_second": 28.27,
+      "eval_steps_per_second": 7.068,
       "step": 114000
     },
     {
-      "epoch": 4.01,
-      "learning_rate": 4.9701618111386906e-06,
-      "loss": 0.2451,
+      "epoch": 4.108732475582946,
+      "grad_norm": 0.25444066524505615,
+      "learning_rate": 4.064590364268298e-06,
+      "loss": 0.3812,
       "step": 114005
     },
     {
-      "epoch": 4.01,
-      "learning_rate": 4.968457251180222e-06,
-      "loss": 0.2401,
+      "epoch": 4.108912675244171,
+      "grad_norm": 0.23914211988449097,
+      "learning_rate": 4.062995555843419e-06,
+      "loss": 0.3443,
       "step": 114010
     },
     {
-      "epoch": 4.01,
-      "learning_rate": 4.966752951318132e-06,
-      "loss": 0.2673,
+      "epoch": 4.109092874905395,
+      "grad_norm": 0.25930920243263245,
+      "learning_rate": 4.061401032681172e-06,
+      "loss": 0.3715,
       "step": 114015
     },
     {
-      "epoch": 4.01,
-      "learning_rate": 4.965048911574547e-06,
-      "loss": 0.2462,
+      "epoch": 4.1092730745666195,
+      "grad_norm": 0.37500235438346863,
+      "learning_rate": 4.059806794803283e-06,
+      "loss": 0.4553,
       "step": 114020
     },
     {
-      "epoch": 4.01,
-      "learning_rate": 4.963345131971606e-06,
-      "loss": 0.2481,
+      "epoch": 4.109453274227844,
+      "grad_norm": 0.26621177792549133,
+      "learning_rate": 4.058212842231474e-06,
+      "loss": 0.3756,
       "step": 114025
     },
     {
-      "epoch": 4.01,
-      "learning_rate": 4.961641612531423e-06,
-      "loss": 0.2547,
+      "epoch": 4.109633473889069,
+      "grad_norm": 0.22278466820716858,
+      "learning_rate": 4.056619174987453e-06,
+      "loss": 0.3579,
       "step": 114030
     },
     {
-      "epoch": 4.01,
-      "learning_rate": 4.959938353276111e-06,
-      "loss": 0.257,
+      "epoch": 4.109813673550294,
+      "grad_norm": 0.2507738769054413,
+      "learning_rate": 4.055025793092945e-06,
+      "loss": 0.3704,
       "step": 114035
     },
     {
-      "epoch": 4.01,
-      "learning_rate": 4.958235354227791e-06,
-      "loss": 0.2402,
+      "epoch": 4.109993873211518,
+      "grad_norm": 0.29194319248199463,
+      "learning_rate": 4.053432696569659e-06,
+      "loss": 0.3727,
       "step": 114040
     },
     {
-      "epoch": 4.01,
-      "learning_rate": 4.956532615408588e-06,
-      "loss": 0.2731,
+      "epoch": 4.110174072872743,
+      "grad_norm": 0.22868956625461578,
+      "learning_rate": 4.0518398854392955e-06,
+      "loss": 0.3543,
       "step": 114045
     },
     {
-      "epoch": 4.01,
-      "learning_rate": 4.954830136840597e-06,
-      "loss": 0.2485,
+      "epoch": 4.110354272533968,
+      "grad_norm": 0.24495558440685272,
+      "learning_rate": 4.050247359723558e-06,
+      "loss": 0.3589,
       "step": 114050
     },
     {
-      "epoch": 4.01,
-      "learning_rate": 4.9531279185459186e-06,
-      "loss": 0.2538,
+      "epoch": 4.110534472195193,
+      "grad_norm": 0.22219400107860565,
+      "learning_rate": 4.0486551194441445e-06,
+      "loss": 0.3735,
       "step": 114055
     },
     {
-      "epoch": 4.01,
-      "learning_rate": 4.9514259605466715e-06,
-      "loss": 0.2468,
+      "epoch": 4.1107146718564165,
+      "grad_norm": 0.240044966340065,
+      "learning_rate": 4.0470631646227416e-06,
+      "loss": 0.3688,
       "step": 114060
     },
     {
-      "epoch": 4.01,
-      "learning_rate": 4.949724262864944e-06,
-      "loss": 0.2409,
+      "epoch": 4.110894871517641,
+      "grad_norm": 0.2267971634864807,
+      "learning_rate": 4.045471495281053e-06,
+      "loss": 0.3558,
       "step": 114065
     },
     {
-      "epoch": 4.01,
-      "learning_rate": 4.948022825522835e-06,
-      "loss": 0.2396,
+      "epoch": 4.111075071178866,
+      "grad_norm": 0.24570490419864655,
+      "learning_rate": 4.043880111440759e-06,
+      "loss": 0.3887,
       "step": 114070
     },
     {
-      "epoch": 4.01,
-      "learning_rate": 4.946321648542429e-06,
-      "loss": 0.2473,
+      "epoch": 4.111255270840091,
+      "grad_norm": 0.24875159561634064,
+      "learning_rate": 4.04228901312354e-06,
+      "loss": 0.3644,
       "step": 114075
     },
     {
-      "epoch": 4.01,
-      "learning_rate": 4.944620731945826e-06,
-      "loss": 0.2257,
+      "epoch": 4.111435470501315,
+      "grad_norm": 0.2332112193107605,
+      "learning_rate": 4.040698200351076e-06,
+      "loss": 0.3569,
       "step": 114080
     },
     {
-      "epoch": 4.01,
-      "learning_rate": 4.942920075755103e-06,
-      "loss": 0.2723,
+      "epoch": 4.11161567016254,
+      "grad_norm": 0.2375115007162094,
+      "learning_rate": 4.039107673145038e-06,
+      "loss": 0.3451,
       "step": 114085
     },
     {
-      "epoch": 4.01,
-      "learning_rate": 4.9412196799923424e-06,
-      "loss": 0.2209,
+      "epoch": 4.111795869823765,
+      "grad_norm": 0.26559481024742126,
+      "learning_rate": 4.037517431527108e-06,
+      "loss": 0.3891,
       "step": 114090
     },
     {
-      "epoch": 4.01,
-      "learning_rate": 4.939519544679622e-06,
-      "loss": 0.2604,
+      "epoch": 4.11197606948499,
+      "grad_norm": 0.26684683561325073,
+      "learning_rate": 4.0359274755189386e-06,
+      "loss": 0.3918,
       "step": 114095
     },
     {
-      "epoch": 4.01,
-      "learning_rate": 4.937819669839028e-06,
-      "loss": 0.2597,
+      "epoch": 4.112156269146214,
+      "grad_norm": 0.288522332906723,
+      "learning_rate": 4.034337805142193e-06,
+      "loss": 0.3604,
       "step": 114100
     },
     {
-      "epoch": 4.01,
-      "learning_rate": 4.936120055492624e-06,
-      "loss": 0.2312,
+      "epoch": 4.112336468807438,
+      "grad_norm": 0.23801618814468384,
+      "learning_rate": 4.032748420418545e-06,
+      "loss": 0.3558,
       "step": 114105
     },
     {
-      "epoch": 4.01,
-      "learning_rate": 4.934420701662479e-06,
-      "loss": 0.2435,
+      "epoch": 4.112516668468663,
+      "grad_norm": 0.26979947090148926,
+      "learning_rate": 4.031159321369637e-06,
+      "loss": 0.3942,
       "step": 114110
     },
     {
-      "epoch": 4.01,
-      "learning_rate": 4.93272160837065e-06,
-      "loss": 0.2504,
+      "epoch": 4.112696868129888,
+      "grad_norm": 0.24331551790237427,
+      "learning_rate": 4.029570508017125e-06,
+      "loss": 0.4005,
       "step": 114115
     },
     {
-      "epoch": 4.02,
-      "learning_rate": 4.931022775639213e-06,
-      "loss": 0.2324,
+      "epoch": 4.112877067791112,
+      "grad_norm": 0.24675403535366058,
+      "learning_rate": 4.027981980382656e-06,
+      "loss": 0.3982,
       "step": 114120
     },
     {
-      "epoch": 4.02,
-      "learning_rate": 4.929324203490221e-06,
-      "loss": 0.2535,
+      "epoch": 4.113057267452337,
+      "grad_norm": 0.18268905580043793,
+      "learning_rate": 4.026393738487863e-06,
+      "loss": 0.3782,
       "step": 114125
     },
     {
-      "epoch": 4.02,
-      "learning_rate": 4.927625891945725e-06,
-      "loss": 0.2404,
+      "epoch": 4.113237467113562,
+      "grad_norm": 0.24003072082996368,
+      "learning_rate": 4.024805782354402e-06,
+      "loss": 0.3864,
       "step": 114130
     },
     {
-      "epoch": 4.02,
-      "learning_rate": 4.925927841027772e-06,
-      "loss": 0.2518,
+      "epoch": 4.113417666774787,
+      "grad_norm": 0.26609840989112854,
+      "learning_rate": 4.023218112003902e-06,
+      "loss": 0.3561,
       "step": 114135
     },
     {
-      "epoch": 4.02,
-      "learning_rate": 4.924230050758425e-06,
-      "loss": 0.2626,
+      "epoch": 4.113597866436011,
+      "grad_norm": 0.29449740052223206,
+      "learning_rate": 4.021630727457995e-06,
+      "loss": 0.4234,
       "step": 114140
     },
     {
-      "epoch": 4.02,
-      "learning_rate": 4.922532521159712e-06,
-      "loss": 0.267,
+      "epoch": 4.113778066097236,
+      "grad_norm": 0.20891326665878296,
+      "learning_rate": 4.020043628738304e-06,
+      "loss": 0.3911,
       "step": 114145
     },
     {
-      "epoch": 4.02,
-      "learning_rate": 4.920835252253683e-06,
-      "loss": 0.2593,
+      "epoch": 4.113958265758461,
+      "grad_norm": 0.20913873612880707,
+      "learning_rate": 4.018456815866453e-06,
+      "loss": 0.3592,
       "step": 114150
     },
     {
-      "epoch": 4.02,
-      "learning_rate": 4.9191382440623826e-06,
-      "loss": 0.2333,
+      "epoch": 4.114138465419685,
+      "grad_norm": 0.22680309414863586,
+      "learning_rate": 4.016870288864072e-06,
+      "loss": 0.367,
       "step": 114155
     },
     {
-      "epoch": 4.02,
-      "learning_rate": 4.917441496607836e-06,
-      "loss": 0.2776,
+      "epoch": 4.114318665080909,
+      "grad_norm": 0.2160423845052719,
+      "learning_rate": 4.015284047752771e-06,
+      "loss": 0.3595,
       "step": 114160
     },
     {
-      "epoch": 4.02,
-      "learning_rate": 4.915745009912079e-06,
-      "loss": 0.2543,
+      "epoch": 4.114498864742134,
+      "grad_norm": 0.2502204179763794,
+      "learning_rate": 4.013698092554163e-06,
+      "loss": 0.3754,
       "step": 114165
     },
     {
-      "epoch": 4.02,
-      "learning_rate": 4.91404878399713e-06,
-      "loss": 0.255,
+      "epoch": 4.114679064403359,
+      "grad_norm": 0.2723202109336853,
+      "learning_rate": 4.012112423289855e-06,
+      "loss": 0.3667,
       "step": 114170
     },
     {
-      "epoch": 4.02,
-      "learning_rate": 4.912352818885027e-06,
-      "loss": 0.2524,
+      "epoch": 4.114859264064584,
+      "grad_norm": 0.279547780752182,
+      "learning_rate": 4.010527039981443e-06,
+      "loss": 0.3663,
       "step": 114175
     },
     {
-      "epoch": 4.02,
-      "learning_rate": 4.910657114597786e-06,
-      "loss": 0.2635,
+      "epoch": 4.115039463725808,
+      "grad_norm": 0.2516183853149414,
+      "learning_rate": 4.008941942650546e-06,
+      "loss": 0.368,
       "step": 114180
     },
     {
-      "epoch": 4.02,
-      "learning_rate": 4.908961671157422e-06,
-      "loss": 0.2681,
+      "epoch": 4.115219663387033,
+      "grad_norm": 0.277414470911026,
+      "learning_rate": 4.007357131318753e-06,
+      "loss": 0.3524,
       "step": 114185
     },
     {
-      "epoch": 4.02,
-      "learning_rate": 4.9072664885859435e-06,
-      "loss": 0.2215,
+      "epoch": 4.115399863048258,
+      "grad_norm": 0.237200066447258,
+      "learning_rate": 4.005772606007646e-06,
+      "loss": 0.3486,
       "step": 114190
     },
     {
-      "epoch": 4.02,
-      "learning_rate": 4.905571566905373e-06,
-      "loss": 0.229,
+      "epoch": 4.1155800627094825,
+      "grad_norm": 0.2728269100189209,
+      "learning_rate": 4.004188366738829e-06,
+      "loss": 0.3794,
       "step": 114195
     },
     {
-      "epoch": 4.02,
-      "learning_rate": 4.9038769061377085e-06,
-      "loss": 0.2318,
+      "epoch": 4.115760262370706,
+      "grad_norm": 0.2168535739183426,
+      "learning_rate": 4.002604413533878e-06,
+      "loss": 0.3768,
       "step": 114200
     },
     {
-      "epoch": 4.02,
-      "learning_rate": 4.902182506304964e-06,
-      "loss": 0.2536,
+      "epoch": 4.115940462031931,
+      "grad_norm": 0.22027429938316345,
+      "learning_rate": 4.00102074641438e-06,
+      "loss": 0.3745,
       "step": 114205
     },
     {
-      "epoch": 4.02,
-      "learning_rate": 4.900488367429129e-06,
-      "loss": 0.2424,
+      "epoch": 4.116120661693156,
+      "grad_norm": 0.23675896227359772,
+      "learning_rate": 3.999437365401906e-06,
+      "loss": 0.3777,
       "step": 114210
     },
     {
-      "epoch": 4.02,
-      "learning_rate": 4.898794489532213e-06,
-      "loss": 0.248,
+      "epoch": 4.116300861354381,
+      "grad_norm": 0.24391767382621765,
+      "learning_rate": 3.997854270518026e-06,
+      "loss": 0.3857,
       "step": 114215
     },
     {
-      "epoch": 4.02,
-      "learning_rate": 4.897100872636201e-06,
-      "loss": 0.2413,
+      "epoch": 4.116481061015605,
+      "grad_norm": 0.239417165517807,
+      "learning_rate": 3.996271461784324e-06,
+      "loss": 0.3692,
       "step": 114220
     },
     {
-      "epoch": 4.02,
-      "learning_rate": 4.895407516763087e-06,
-      "loss": 0.2437,
+      "epoch": 4.11666126067683,
+      "grad_norm": 0.2657260596752167,
+      "learning_rate": 3.9946889392223545e-06,
+      "loss": 0.3556,
       "step": 114225
     },
     {
-      "epoch": 4.02,
-      "learning_rate": 4.893714421934853e-06,
-      "loss": 0.2403,
+      "epoch": 4.116841460338055,
+      "grad_norm": 0.2049977332353592,
+      "learning_rate": 3.993106702853683e-06,
+      "loss": 0.3461,
       "step": 114230
     },
     {
-      "epoch": 4.02,
-      "learning_rate": 4.89202158817349e-06,
-      "loss": 0.2628,
+      "epoch": 4.1170216599992795,
+      "grad_norm": 0.24377499520778656,
+      "learning_rate": 3.9915247526998625e-06,
+      "loss": 0.3724,
       "step": 114235
     },
     {
-      "epoch": 4.02,
-      "learning_rate": 4.890329015500977e-06,
-      "loss": 0.2512,
+      "epoch": 4.117201859660504,
+      "grad_norm": 0.23424182832241058,
+      "learning_rate": 3.989943088782453e-06,
+      "loss": 0.3779,
       "step": 114240
     },
     {
-      "epoch": 4.02,
-      "learning_rate": 4.888636703939289e-06,
-      "loss": 0.2466,
+      "epoch": 4.117382059321728,
+      "grad_norm": 0.21818037331104279,
+      "learning_rate": 3.9883617111229955e-06,
+      "loss": 0.3498,
       "step": 114245
     },
     {
-      "epoch": 4.02,
-      "learning_rate": 4.886944653510392e-06,
-      "loss": 0.2495,
+      "epoch": 4.117562258982953,
+      "grad_norm": 0.25411292910575867,
+      "learning_rate": 3.986780619743047e-06,
+      "loss": 0.3329,
       "step": 114250
     },
     {
-      "epoch": 4.02,
-      "learning_rate": 4.885252864236264e-06,
-      "loss": 0.2409,
+      "epoch": 4.1177424586441775,
+      "grad_norm": 0.2751554846763611,
+      "learning_rate": 3.985199814664142e-06,
+      "loss": 0.3821,
       "step": 114255
     },
     {
-      "epoch": 4.02,
-      "learning_rate": 4.883561336138878e-06,
-      "loss": 0.2546,
+      "epoch": 4.117922658305402,
+      "grad_norm": 0.20654655992984772,
+      "learning_rate": 3.9836192959078225e-06,
+      "loss": 0.3981,
       "step": 114260
     },
     {
-      "epoch": 4.02,
-      "learning_rate": 4.881870069240191e-06,
-      "loss": 0.2439,
+      "epoch": 4.118102857966627,
+      "grad_norm": 0.21761824190616608,
+      "learning_rate": 3.982039063495621e-06,
+      "loss": 0.3587,
       "step": 114265
     },
     {
-      "epoch": 4.02,
-      "learning_rate": 4.880179063562154e-06,
-      "loss": 0.2482,
+      "epoch": 4.118283057627852,
+      "grad_norm": 0.3250308334827423,
+      "learning_rate": 3.980459117449065e-06,
+      "loss": 0.3831,
       "step": 114270
     },
     {
-      "epoch": 4.02,
-      "learning_rate": 4.878826447113366e-06,
-      "loss": 0.2398,
+      "epoch": 4.1184632572890765,
+      "grad_norm": 0.22311455011367798,
+      "learning_rate": 3.978879457789686e-06,
+      "loss": 0.3779,
       "step": 114275
     },
     {
-      "epoch": 4.02,
-      "learning_rate": 4.8771359116878464e-06,
-      "loss": 0.2486,
+      "epoch": 4.118643456950301,
+      "grad_norm": 0.28070372343063354,
+      "learning_rate": 3.977300084539013e-06,
+      "loss": 0.3387,
       "step": 114280
     },
     {
-      "epoch": 4.02,
-      "learning_rate": 4.875445637544462e-06,
-      "loss": 0.2419,
+      "epoch": 4.118823656611526,
+      "grad_norm": 0.19799430668354034,
+      "learning_rate": 3.975720997718544e-06,
+      "loss": 0.3412,
       "step": 114285
     },
     {
-      "epoch": 4.02,
-      "learning_rate": 4.873755624705145e-06,
-      "loss": 0.2332,
+      "epoch": 4.11900385627275,
+      "grad_norm": 0.24587881565093994,
+      "learning_rate": 3.974142197349809e-06,
+      "loss": 0.3441,
       "step": 114290
     },
     {
-      "epoch": 4.02,
-      "learning_rate": 4.872065873191855e-06,
-      "loss": 0.2522,
+      "epoch": 4.1191840559339745,
+      "grad_norm": 0.26680803298950195,
+      "learning_rate": 3.972563683454314e-06,
+      "loss": 0.3738,
       "step": 114295
     },
     {
-      "epoch": 4.02,
-      "learning_rate": 4.870376383026518e-06,
-      "loss": 0.2662,
+      "epoch": 4.119364255595199,
+      "grad_norm": 0.22729459404945374,
+      "learning_rate": 3.970985456053578e-06,
+      "loss": 0.38,
       "step": 114300
     },
     {
-      "epoch": 4.02,
-      "learning_rate": 4.868687154231083e-06,
-      "loss": 0.2625,
+      "epoch": 4.119544455256424,
+      "grad_norm": 0.28258028626441956,
+      "learning_rate": 3.969407515169088e-06,
+      "loss": 0.3575,
       "step": 114305
     },
     {
-      "epoch": 4.02,
-      "learning_rate": 4.866998186827482e-06,
-      "loss": 0.2336,
+      "epoch": 4.119724654917649,
+      "grad_norm": 0.2133554369211197,
+      "learning_rate": 3.967829860822342e-06,
+      "loss": 0.3429,
       "step": 114310
     },
     {
-      "epoch": 4.02,
-      "learning_rate": 4.865309480837638e-06,
-      "loss": 0.2593,
+      "epoch": 4.1199048545788735,
+      "grad_norm": 0.2711166441440582,
+      "learning_rate": 3.966252493034847e-06,
+      "loss": 0.3196,
       "step": 114315
     },
     {
-      "epoch": 4.02,
-      "learning_rate": 4.863621036283475e-06,
-      "loss": 0.2451,
+      "epoch": 4.120085054240098,
+      "grad_norm": 0.2520091235637665,
+      "learning_rate": 3.964675411828092e-06,
+      "loss": 0.387,
       "step": 114320
     },
     {
-      "epoch": 4.02,
-      "learning_rate": 4.861932853186929e-06,
-      "loss": 0.2312,
+      "epoch": 4.120265253901323,
+      "grad_norm": 0.2586155831813812,
+      "learning_rate": 3.963098617223562e-06,
+      "loss": 0.3943,
       "step": 114325
     },
     {
-      "epoch": 4.02,
-      "learning_rate": 4.860244931569915e-06,
-      "loss": 0.2612,
+      "epoch": 4.120445453562548,
+      "grad_norm": 0.23721350729465485,
+      "learning_rate": 3.961522109242741e-06,
+      "loss": 0.3705,
       "step": 114330
     },
     {
-      "epoch": 4.02,
-      "learning_rate": 4.858557271454348e-06,
-      "loss": 0.2539,
+      "epoch": 4.120625653223772,
+      "grad_norm": 0.24257147312164307,
+      "learning_rate": 3.959945887907099e-06,
+      "loss": 0.3615,
       "step": 114335
     },
     {
-      "epoch": 4.02,
-      "learning_rate": 4.856869872862132e-06,
-      "loss": 0.2428,
+      "epoch": 4.120805852884996,
+      "grad_norm": 0.2581164240837097,
+      "learning_rate": 3.9583699532381304e-06,
+      "loss": 0.3599,
       "step": 114340
     },
     {
-      "epoch": 4.02,
-      "learning_rate": 4.855182735815189e-06,
-      "loss": 0.2413,
+      "epoch": 4.120986052546221,
+      "grad_norm": 0.2084408849477768,
+      "learning_rate": 3.956794305257294e-06,
+      "loss": 0.3728,
       "step": 114345
     },
     {
-      "epoch": 4.02,
-      "learning_rate": 4.853495860335428e-06,
-      "loss": 0.2506,
+      "epoch": 4.121166252207446,
+      "grad_norm": 0.22756318747997284,
+      "learning_rate": 3.955218943986064e-06,
+      "loss": 0.3401,
       "step": 114350
     },
     {
-      "epoch": 4.02,
-      "learning_rate": 4.851809246444739e-06,
-      "loss": 0.2711,
+      "epoch": 4.12134645186867,
+      "grad_norm": 0.21954157948493958,
+      "learning_rate": 3.9536438694459e-06,
+      "loss": 0.394,
       "step": 114355
     },
     {
-      "epoch": 4.02,
-      "learning_rate": 4.850122894165035e-06,
-      "loss": 0.2475,
+      "epoch": 4.121526651529895,
+      "grad_norm": 0.24439333379268646,
+      "learning_rate": 3.952069081658258e-06,
+      "loss": 0.3647,
       "step": 114360
     },
     {
-      "epoch": 4.02,
-      "learning_rate": 4.848436803518205e-06,
-      "loss": 0.2499,
+      "epoch": 4.12170685119112,
+      "grad_norm": 0.2532067596912384,
+      "learning_rate": 3.950494580644606e-06,
+      "loss": 0.3597,
       "step": 114365
     },
     {
-      "epoch": 4.02,
-      "learning_rate": 4.846750974526146e-06,
-      "loss": 0.2573,
+      "epoch": 4.121887050852345,
+      "grad_norm": 0.23494374752044678,
+      "learning_rate": 3.9489203664263955e-06,
+      "loss": 0.3872,
       "step": 114370
     },
     {
-      "epoch": 4.02,
-      "learning_rate": 4.845065407210733e-06,
-      "loss": 0.23,
+      "epoch": 4.122067250513569,
+      "grad_norm": 0.2250523865222931,
+      "learning_rate": 3.947346439025057e-06,
+      "loss": 0.3852,
       "step": 114375
     },
     {
-      "epoch": 4.02,
-      "learning_rate": 4.843380101593872e-06,
-      "loss": 0.2409,
+      "epoch": 4.122247450174794,
+      "grad_norm": 0.21566912531852722,
+      "learning_rate": 3.9457727984620524e-06,
+      "loss": 0.3862,
       "step": 114380
     },
     {
-      "epoch": 4.02,
-      "learning_rate": 4.8416950576974375e-06,
-      "loss": 0.2442,
+      "epoch": 4.122427649836018,
+      "grad_norm": 0.33429858088493347,
+      "learning_rate": 3.94419944475882e-06,
+      "loss": 0.3911,
       "step": 114385
     },
     {
-      "epoch": 4.02,
-      "learning_rate": 4.840010275543308e-06,
-      "loss": 0.2409,
+      "epoch": 4.122607849497243,
+      "grad_norm": 0.28775933384895325,
+      "learning_rate": 3.942626377936793e-06,
+      "loss": 0.3773,
       "step": 114390
     },
     {
-      "epoch": 4.02,
-      "learning_rate": 4.838325755153353e-06,
-      "loss": 0.2386,
+      "epoch": 4.122788049158467,
+      "grad_norm": 0.23396413028240204,
+      "learning_rate": 3.941053598017402e-06,
+      "loss": 0.3944,
       "step": 114395
     },
     {
-      "epoch": 4.02,
-      "learning_rate": 4.836641496549449e-06,
-      "loss": 0.2406,
+      "epoch": 4.122968248819692,
+      "grad_norm": 0.2518860101699829,
+      "learning_rate": 3.939481105022075e-06,
+      "loss": 0.3333,
       "step": 114400
     },
     {
-      "epoch": 4.03,
-      "learning_rate": 4.834957499753473e-06,
-      "loss": 0.2305,
+      "epoch": 4.123148448480917,
+      "grad_norm": 0.2610318064689636,
+      "learning_rate": 3.937908898972248e-06,
+      "loss": 0.3537,
       "step": 114405
     },
     {
-      "epoch": 4.03,
-      "learning_rate": 4.833273764787286e-06,
-      "loss": 0.2556,
+      "epoch": 4.123328648142142,
+      "grad_norm": 0.2574523687362671,
+      "learning_rate": 3.936336979889332e-06,
+      "loss": 0.3709,
       "step": 114410
     },
     {
-      "epoch": 4.03,
-      "learning_rate": 4.831590291672744e-06,
-      "loss": 0.2356,
+      "epoch": 4.123508847803366,
+      "grad_norm": 0.2853260040283203,
+      "learning_rate": 3.934765347794747e-06,
+      "loss": 0.372,
       "step": 114415
     },
     {
-      "epoch": 4.03,
-      "learning_rate": 4.8299070804317145e-06,
-      "loss": 0.2411,
+      "epoch": 4.123689047464591,
+      "grad_norm": 0.22084416449069977,
+      "learning_rate": 3.933194002709906e-06,
+      "loss": 0.3494,
       "step": 114420
     },
     {
-      "epoch": 4.03,
-      "learning_rate": 4.828224131086051e-06,
-      "loss": 0.2539,
+      "epoch": 4.123869247125816,
+      "grad_norm": 0.23949111998081207,
+      "learning_rate": 3.9316229446562184e-06,
+      "loss": 0.3449,
       "step": 114425
     },
     {
-      "epoch": 4.03,
-      "learning_rate": 4.826541443657601e-06,
-      "loss": 0.2573,
+      "epoch": 4.12404944678704,
+      "grad_norm": 0.2440653145313263,
+      "learning_rate": 3.9300521736550825e-06,
+      "loss": 0.358,
       "step": 114430
     },
     {
-      "epoch": 4.03,
-      "learning_rate": 4.824859018168209e-06,
-      "loss": 0.2431,
+      "epoch": 4.124229646448264,
+      "grad_norm": 0.2716491222381592,
+      "learning_rate": 3.928481689727911e-06,
+      "loss": 0.3932,
       "step": 114435
     },
     {
-      "epoch": 4.03,
-      "learning_rate": 4.823176854639735e-06,
-      "loss": 0.2348,
+      "epoch": 4.124409846109489,
+      "grad_norm": 0.26395681500434875,
+      "learning_rate": 3.926911492896098e-06,
+      "loss": 0.3841,
       "step": 114440
     },
     {
-      "epoch": 4.03,
-      "learning_rate": 4.821494953094013e-06,
-      "loss": 0.2431,
+      "epoch": 4.124590045770714,
+      "grad_norm": 0.2175927311182022,
+      "learning_rate": 3.925341583181039e-06,
+      "loss": 0.3782,
       "step": 114445
     },
     {
-      "epoch": 4.03,
-      "learning_rate": 4.819813313552873e-06,
-      "loss": 0.2549,
+      "epoch": 4.124770245431939,
+      "grad_norm": 0.26144078373908997,
+      "learning_rate": 3.923771960604117e-06,
+      "loss": 0.3954,
       "step": 114450
     },
     {
-      "epoch": 4.03,
-      "learning_rate": 4.818131936038159e-06,
-      "loss": 0.2597,
+      "epoch": 4.124950445093163,
+      "grad_norm": 0.24796469509601593,
+      "learning_rate": 3.922202625186719e-06,
+      "loss": 0.3914,
       "step": 114455
     },
     {
-      "epoch": 4.03,
-      "learning_rate": 4.816450820571708e-06,
-      "loss": 0.2243,
+      "epoch": 4.125130644754388,
+      "grad_norm": 0.24708044528961182,
+      "learning_rate": 3.920633576950234e-06,
+      "loss": 0.391,
       "step": 114460
     },
     {
-      "epoch": 4.03,
-      "learning_rate": 4.814769967175339e-06,
-      "loss": 0.2375,
+      "epoch": 4.125310844415613,
+      "grad_norm": 0.23979832231998444,
+      "learning_rate": 3.91906481591604e-06,
+      "loss": 0.3627,
       "step": 114465
     },
     {
-      "epoch": 4.03,
-      "learning_rate": 4.813089375870878e-06,
-      "loss": 0.2516,
+      "epoch": 4.1254910440768375,
+      "grad_norm": 0.2370292693376541,
+      "learning_rate": 3.917496342105495e-06,
+      "loss": 0.3845,
       "step": 114470
     },
     {
-      "epoch": 4.03,
-      "learning_rate": 4.811409046680149e-06,
-      "loss": 0.258,
+      "epoch": 4.125671243738061,
+      "grad_norm": 0.2842269837856293,
+      "learning_rate": 3.915928155539986e-06,
+      "loss": 0.3571,
       "step": 114475
     },
     {
-      "epoch": 4.03,
-      "learning_rate": 4.809728979624972e-06,
-      "loss": 0.2288,
+      "epoch": 4.125851443399286,
+      "grad_norm": 0.23658975958824158,
+      "learning_rate": 3.914360256240871e-06,
+      "loss": 0.357,
       "step": 114480
     },
     {
-      "epoch": 4.03,
-      "learning_rate": 4.808049174727155e-06,
-      "loss": 0.2605,
+      "epoch": 4.126031643060511,
+      "grad_norm": 0.24424517154693604,
+      "learning_rate": 3.912792644229524e-06,
+      "loss": 0.3581,
       "step": 114485
     },
     {
-      "epoch": 4.03,
-      "learning_rate": 4.806369632008509e-06,
-      "loss": 0.2512,
+      "epoch": 4.126211842721736,
+      "grad_norm": 0.3009839951992035,
+      "learning_rate": 3.91122531952729e-06,
+      "loss": 0.3738,
       "step": 114490
     },
     {
-      "epoch": 4.03,
-      "learning_rate": 4.804690351490851e-06,
-      "loss": 0.2438,
+      "epoch": 4.12639204238296,
+      "grad_norm": 0.1834869086742401,
+      "learning_rate": 3.9096582821555204e-06,
+      "loss": 0.3381,
       "step": 114495
     },
     {
-      "epoch": 4.03,
-      "learning_rate": 4.8030113331959805e-06,
-      "loss": 0.2472,
+      "epoch": 4.126572242044185,
+      "grad_norm": 0.19505488872528076,
+      "learning_rate": 3.908091532135583e-06,
+      "loss": 0.3742,
       "step": 114500
     },
     {
-      "epoch": 4.03,
-      "eval_loss": 0.24984058737754822,
-      "eval_runtime": 10.5589,
-      "eval_samples_per_second": 9.471,
-      "eval_steps_per_second": 9.471,
+      "epoch": 4.126572242044185,
+      "eval_loss": 0.42937061190605164,
+      "eval_runtime": 3.5336,
+      "eval_samples_per_second": 28.3,
+      "eval_steps_per_second": 7.075,
       "step": 114500
     },
     {
-      "epoch": 4.03,
-      "learning_rate": 4.80133257714569e-06,
-      "loss": 0.2481,
+      "epoch": 4.12675244170541,
+      "grad_norm": 0.21231359243392944,
+      "learning_rate": 3.906525069488812e-06,
+      "loss": 0.3716,
       "step": 114505
     },
     {
-      "epoch": 4.03,
-      "learning_rate": 4.79965408336179e-06,
-      "loss": 0.2397,
+      "epoch": 4.1269326413666345,
+      "grad_norm": 0.255136638879776,
+      "learning_rate": 3.904958894236554e-06,
+      "loss": 0.3851,
       "step": 114510
     },
     {
-      "epoch": 4.03,
-      "learning_rate": 4.797975851866063e-06,
-      "loss": 0.2519,
+      "epoch": 4.127112841027859,
+      "grad_norm": 0.22166618704795837,
+      "learning_rate": 3.903393006400147e-06,
+      "loss": 0.3521,
       "step": 114515
     },
     {
-      "epoch": 4.03,
-      "learning_rate": 4.796297882680315e-06,
-      "loss": 0.229,
+      "epoch": 4.127293040689083,
+      "grad_norm": 0.23030775785446167,
+      "learning_rate": 3.901827406000918e-06,
+      "loss": 0.348,
       "step": 114520
     },
     {
-      "epoch": 4.03,
-      "learning_rate": 4.794620175826323e-06,
-      "loss": 0.2386,
+      "epoch": 4.127473240350308,
+      "grad_norm": 0.20609252154827118,
+      "learning_rate": 3.900262093060214e-06,
+      "loss": 0.3331,
       "step": 114525
     },
     {
-      "epoch": 4.03,
-      "learning_rate": 4.792942731325864e-06,
-      "loss": 0.2488,
+      "epoch": 4.127653440011533,
+      "grad_norm": 0.2601858377456665,
+      "learning_rate": 3.898697067599355e-06,
+      "loss": 0.3415,
       "step": 114530
     },
     {
-      "epoch": 4.03,
-      "learning_rate": 4.791265549200735e-06,
-      "loss": 0.2647,
+      "epoch": 4.127833639672757,
+      "grad_norm": 0.20611776411533356,
+      "learning_rate": 3.897132329639661e-06,
+      "loss": 0.3781,
       "step": 114535
     },
     {
-      "epoch": 4.03,
-      "learning_rate": 4.789588629472702e-06,
-      "loss": 0.2424,
+      "epoch": 4.128013839333982,
+      "grad_norm": 0.2544640898704529,
+      "learning_rate": 3.895567879202452e-06,
+      "loss": 0.3592,
       "step": 114540
     },
     {
-      "epoch": 4.03,
-      "learning_rate": 4.7879119721635444e-06,
-      "loss": 0.2584,
+      "epoch": 4.128194038995207,
+      "grad_norm": 0.2738896608352661,
+      "learning_rate": 3.894003716309047e-06,
+      "loss": 0.3508,
       "step": 114545
     },
     {
-      "epoch": 4.03,
-      "learning_rate": 4.786235577295023e-06,
-      "loss": 0.2679,
+      "epoch": 4.1283742386564315,
+      "grad_norm": 0.20672179758548737,
+      "learning_rate": 3.89243984098075e-06,
+      "loss": 0.3829,
       "step": 114550
     },
     {
-      "epoch": 4.03,
-      "learning_rate": 4.784559444888917e-06,
-      "loss": 0.2353,
+      "epoch": 4.128554438317656,
+      "grad_norm": 0.31415534019470215,
+      "learning_rate": 3.890876253238884e-06,
+      "loss": 0.3825,
       "step": 114555
     },
     {
-      "epoch": 4.03,
-      "learning_rate": 4.782883574966979e-06,
-      "loss": 0.2396,
+      "epoch": 4.128734637978881,
+      "grad_norm": 0.2770949900150299,
+      "learning_rate": 3.8893129531047314e-06,
+      "loss": 0.3637,
       "step": 114560
     },
     {
-      "epoch": 4.03,
-      "learning_rate": 4.781207967550977e-06,
-      "loss": 0.2575,
+      "epoch": 4.128914837640105,
+      "grad_norm": 0.20876848697662354,
+      "learning_rate": 3.887749940599608e-06,
+      "loss": 0.3571,
       "step": 114565
     },
     {
-      "epoch": 4.03,
-      "learning_rate": 4.779532622662661e-06,
-      "loss": 0.2447,
+      "epoch": 4.1290950373013295,
+      "grad_norm": 0.2529737055301666,
+      "learning_rate": 3.886187215744805e-06,
+      "loss": 0.3665,
       "step": 114570
     },
     {
-      "epoch": 4.03,
-      "learning_rate": 4.777857540323794e-06,
-      "loss": 0.2481,
+      "epoch": 4.129275236962554,
+      "grad_norm": 0.27046170830726624,
+      "learning_rate": 3.884624778561605e-06,
+      "loss": 0.368,
       "step": 114575
     },
     {
-      "epoch": 4.03,
-      "learning_rate": 4.776182720556121e-06,
-      "loss": 0.2639,
+      "epoch": 4.129455436623779,
+      "grad_norm": 0.2750345766544342,
+      "learning_rate": 3.8830626290713185e-06,
+      "loss": 0.3764,
       "step": 114580
     },
     {
-      "epoch": 4.03,
-      "learning_rate": 4.774508163381381e-06,
-      "loss": 0.2617,
+      "epoch": 4.129635636285004,
+      "grad_norm": 0.24169829487800598,
+      "learning_rate": 3.881500767295201e-06,
+      "loss": 0.363,
       "step": 114585
     },
     {
-      "epoch": 4.03,
-      "learning_rate": 4.772833868821328e-06,
-      "loss": 0.2352,
+      "epoch": 4.1298158359462285,
+      "grad_norm": 0.24228541553020477,
+      "learning_rate": 3.8799391932545555e-06,
+      "loss": 0.3773,
       "step": 114590
     },
     {
-      "epoch": 4.03,
-      "learning_rate": 4.771159836897699e-06,
-      "loss": 0.2407,
+      "epoch": 4.129996035607453,
+      "grad_norm": 0.21988023817539215,
+      "learning_rate": 3.878377906970648e-06,
+      "loss": 0.3664,
       "step": 114595
     },
     {
-      "epoch": 4.03,
-      "learning_rate": 4.769486067632226e-06,
-      "loss": 0.2519,
+      "epoch": 4.130176235268678,
+      "grad_norm": 0.2830277979373932,
+      "learning_rate": 3.876816908464753e-06,
+      "loss": 0.4126,
       "step": 114600
     },
     {
-      "epoch": 4.03,
-      "learning_rate": 4.767812561046639e-06,
-      "loss": 0.2444,
+      "epoch": 4.130356434929903,
+      "grad_norm": 0.2855124771595001,
+      "learning_rate": 3.875256197758137e-06,
+      "loss": 0.3822,
       "step": 114605
     },
     {
-      "epoch": 4.03,
-      "learning_rate": 4.76613931716268e-06,
-      "loss": 0.2578,
+      "epoch": 4.130536634591127,
+      "grad_norm": 0.314847856760025,
+      "learning_rate": 3.873695774872066e-06,
+      "loss": 0.3807,
       "step": 114610
     },
     {
-      "epoch": 4.03,
-      "learning_rate": 4.764466336002066e-06,
-      "loss": 0.2422,
+      "epoch": 4.130716834252351,
+      "grad_norm": 0.26585081219673157,
+      "learning_rate": 3.872135639827795e-06,
+      "loss": 0.3741,
       "step": 114615
     },
     {
-      "epoch": 4.03,
-      "learning_rate": 4.762793617586514e-06,
-      "loss": 0.2292,
+      "epoch": 4.130897033913576,
+      "grad_norm": 0.2731850743293762,
+      "learning_rate": 3.870575792646591e-06,
+      "loss": 0.3896,
       "step": 114620
     },
     {
-      "epoch": 4.03,
-      "learning_rate": 4.761121161937751e-06,
-      "loss": 0.2377,
+      "epoch": 4.131077233574801,
+      "grad_norm": 0.2461189180612564,
+      "learning_rate": 3.8690162333497006e-06,
+      "loss": 0.4125,
       "step": 114625
     },
     {
-      "epoch": 4.03,
-      "learning_rate": 4.759448969077496e-06,
-      "loss": 0.2634,
+      "epoch": 4.1312574332360255,
+      "grad_norm": 0.2130453884601593,
+      "learning_rate": 3.867456961958374e-06,
+      "loss": 0.3652,
       "step": 114630
     },
     {
-      "epoch": 4.03,
-      "learning_rate": 4.757777039027458e-06,
-      "loss": 0.2383,
+      "epoch": 4.13143763289725,
+      "grad_norm": 0.2346867322921753,
+      "learning_rate": 3.865897978493851e-06,
+      "loss": 0.3495,
       "step": 114635
     },
     {
-      "epoch": 4.03,
-      "learning_rate": 4.756105371809347e-06,
-      "loss": 0.2616,
+      "epoch": 4.131617832558475,
+      "grad_norm": 0.23216021060943604,
+      "learning_rate": 3.864339282977375e-06,
+      "loss": 0.367,
       "step": 114640
     },
     {
-      "epoch": 4.03,
-      "learning_rate": 4.75443396744486e-06,
-      "loss": 0.2624,
+      "epoch": 4.1317980322197,
+      "grad_norm": 0.28052276372909546,
+      "learning_rate": 3.862780875430188e-06,
+      "loss": 0.3296,
       "step": 114645
     },
     {
-      "epoch": 4.03,
-      "learning_rate": 4.752762825955711e-06,
-      "loss": 0.2428,
+      "epoch": 4.131978231880924,
+      "grad_norm": 0.26213961839675903,
+      "learning_rate": 3.861222755873526e-06,
+      "loss": 0.4356,
       "step": 114650
     },
     {
-      "epoch": 4.03,
-      "learning_rate": 4.751091947363592e-06,
-      "loss": 0.2472,
+      "epoch": 4.132158431542149,
+      "grad_norm": 0.3032136857509613,
+      "learning_rate": 3.859664924328599e-06,
+      "loss": 0.3864,
       "step": 114655
     },
     {
-      "epoch": 4.03,
-      "learning_rate": 4.749421331690202e-06,
-      "loss": 0.2706,
+      "epoch": 4.132338631203373,
+      "grad_norm": 0.297122061252594,
+      "learning_rate": 3.85810738081665e-06,
+      "loss": 0.3805,
       "step": 114660
     },
     {
-      "epoch": 4.03,
-      "learning_rate": 4.747750978957221e-06,
-      "loss": 0.2544,
+      "epoch": 4.132518830864598,
+      "grad_norm": 0.1998542994260788,
+      "learning_rate": 3.85655012535889e-06,
+      "loss": 0.3474,
       "step": 114665
     },
     {
-      "epoch": 4.03,
-      "learning_rate": 4.746080889186355e-06,
-      "loss": 0.2538,
+      "epoch": 4.132699030525822,
+      "grad_norm": 0.21667265892028809,
+      "learning_rate": 3.854993157976553e-06,
+      "loss": 0.3602,
       "step": 114670
     },
     {
-      "epoch": 4.03,
-      "learning_rate": 4.744411062399276e-06,
-      "loss": 0.2644,
+      "epoch": 4.132879230187047,
+      "grad_norm": 0.2543310523033142,
+      "learning_rate": 3.853436478690833e-06,
+      "loss": 0.3556,
       "step": 114675
     },
     {
-      "epoch": 4.03,
-      "learning_rate": 4.742741498617676e-06,
-      "loss": 0.2504,
+      "epoch": 4.133059429848272,
+      "grad_norm": 0.2114337682723999,
+      "learning_rate": 3.851880087522941e-06,
+      "loss": 0.3946,
       "step": 114680
     },
     {
-      "epoch": 4.03,
-      "learning_rate": 4.74107219786322e-06,
-      "loss": 0.2516,
+      "epoch": 4.133239629509497,
+      "grad_norm": 0.2727842330932617,
+      "learning_rate": 3.850323984494095e-06,
+      "loss": 0.3585,
       "step": 114685
     },
     {
-      "epoch": 4.04,
-      "learning_rate": 4.739403160157596e-06,
-      "loss": 0.2469,
+      "epoch": 4.133419829170721,
+      "grad_norm": 0.27654939889907837,
+      "learning_rate": 3.848768169625491e-06,
+      "loss": 0.362,
       "step": 114690
     },
     {
-      "epoch": 4.04,
-      "learning_rate": 4.737734385522471e-06,
-      "loss": 0.2598,
+      "epoch": 4.133600028831946,
+      "grad_norm": 0.26321300864219666,
+      "learning_rate": 3.8472126429383246e-06,
+      "loss": 0.3651,
       "step": 114695
     },
     {
-      "epoch": 4.04,
-      "learning_rate": 4.736065873979506e-06,
-      "loss": 0.2376,
+      "epoch": 4.133780228493171,
+      "grad_norm": 0.23769168555736542,
+      "learning_rate": 3.84565740445379e-06,
+      "loss": 0.3345,
       "step": 114700
     },
     {
-      "epoch": 4.04,
-      "learning_rate": 4.734397625550377e-06,
-      "loss": 0.2479,
+      "epoch": 4.133960428154395,
+      "grad_norm": 0.2931036353111267,
+      "learning_rate": 3.844102454193072e-06,
+      "loss": 0.3837,
       "step": 114705
     },
     {
-      "epoch": 4.04,
-      "learning_rate": 4.732729640256739e-06,
-      "loss": 0.2631,
+      "epoch": 4.134140627815619,
+      "grad_norm": 0.24765877425670624,
+      "learning_rate": 3.84254779217737e-06,
+      "loss": 0.3692,
       "step": 114710
     },
     {
-      "epoch": 4.04,
-      "learning_rate": 4.7310619181202495e-06,
-      "loss": 0.2706,
+      "epoch": 4.134320827476844,
+      "grad_norm": 0.23711945116519928,
+      "learning_rate": 3.840993418427855e-06,
+      "loss": 0.3612,
       "step": 114715
     },
     {
-      "epoch": 4.04,
-      "learning_rate": 4.729394459162556e-06,
-      "loss": 0.2574,
+      "epoch": 4.134501027138069,
+      "grad_norm": 0.23325876891613007,
+      "learning_rate": 3.839439332965711e-06,
+      "loss": 0.3809,
       "step": 114720
     },
     {
-      "epoch": 4.04,
-      "learning_rate": 4.727727263405326e-06,
-      "loss": 0.2489,
+      "epoch": 4.134681226799294,
+      "grad_norm": 0.2400001585483551,
+      "learning_rate": 3.837885535812108e-06,
+      "loss": 0.3972,
       "step": 114725
     },
     {
-      "epoch": 4.04,
-      "learning_rate": 4.7260603308701885e-06,
-      "loss": 0.2596,
+      "epoch": 4.134861426460518,
+      "grad_norm": 0.24221131205558777,
+      "learning_rate": 3.836332026988218e-06,
+      "loss": 0.3731,
       "step": 114730
     },
     {
-      "epoch": 4.04,
-      "learning_rate": 4.7243936615788045e-06,
-      "loss": 0.2392,
+      "epoch": 4.135041626121743,
+      "grad_norm": 0.22964808344841003,
+      "learning_rate": 3.834778806515199e-06,
+      "loss": 0.3772,
       "step": 114735
     },
     {
-      "epoch": 4.04,
-      "learning_rate": 4.7227272555528e-06,
-      "loss": 0.2669,
+      "epoch": 4.135221825782968,
+      "grad_norm": 0.2366577684879303,
+      "learning_rate": 3.833225874414228e-06,
+      "loss": 0.3839,
       "step": 114740
     },
     {
-      "epoch": 4.04,
-      "learning_rate": 4.721061112813827e-06,
-      "loss": 0.2326,
+      "epoch": 4.1354020254441926,
+      "grad_norm": 0.281572550535202,
+      "learning_rate": 3.8316732307064566e-06,
+      "loss": 0.3524,
       "step": 114745
     },
     {
-      "epoch": 4.04,
-      "learning_rate": 4.719395233383509e-06,
-      "loss": 0.2521,
+      "epoch": 4.135582225105416,
+      "grad_norm": 0.24455080926418304,
+      "learning_rate": 3.83012087541304e-06,
+      "loss": 0.3965,
       "step": 114750
     },
     {
-      "epoch": 4.04,
-      "learning_rate": 4.717729617283478e-06,
-      "loss": 0.2751,
+      "epoch": 4.135762424766641,
+      "grad_norm": 0.2673299312591553,
+      "learning_rate": 3.828568808555127e-06,
+      "loss": 0.3553,
       "step": 114755
     },
     {
-      "epoch": 4.04,
-      "learning_rate": 4.716064264535356e-06,
-      "loss": 0.2576,
+      "epoch": 4.135942624427866,
+      "grad_norm": 0.29149943590164185,
+      "learning_rate": 3.827017030153859e-06,
+      "loss": 0.3808,
       "step": 114760
     },
     {
-      "epoch": 4.04,
-      "learning_rate": 4.714399175160777e-06,
-      "loss": 0.2684,
+      "epoch": 4.136122824089091,
+      "grad_norm": 0.2999207377433777,
+      "learning_rate": 3.825465540230397e-06,
+      "loss": 0.4054,
       "step": 114765
     },
     {
-      "epoch": 4.04,
-      "learning_rate": 4.712734349181358e-06,
-      "loss": 0.2575,
+      "epoch": 4.136303023750315,
+      "grad_norm": 0.2576257288455963,
+      "learning_rate": 3.823914338805856e-06,
+      "loss": 0.3723,
       "step": 114770
     },
     {
-      "epoch": 4.04,
-      "learning_rate": 4.7110697866187135e-06,
-      "loss": 0.248,
+      "epoch": 4.13648322341154,
+      "grad_norm": 0.2315170168876648,
+      "learning_rate": 3.822363425901388e-06,
+      "loss": 0.3923,
       "step": 114775
     },
     {
-      "epoch": 4.04,
-      "learning_rate": 4.709405487494447e-06,
-      "loss": 0.2375,
+      "epoch": 4.136663423072765,
+      "grad_norm": 0.24673189222812653,
+      "learning_rate": 3.8208128015381176e-06,
+      "loss": 0.3754,
       "step": 114780
     },
     {
-      "epoch": 4.04,
-      "learning_rate": 4.707741451830183e-06,
-      "loss": 0.2677,
+      "epoch": 4.1368436227339895,
+      "grad_norm": 0.24493588507175446,
+      "learning_rate": 3.8192624657371725e-06,
+      "loss": 0.3749,
       "step": 114785
     },
     {
-      "epoch": 4.04,
-      "learning_rate": 4.706077679647528e-06,
-      "loss": 0.2777,
+      "epoch": 4.137023822395214,
+      "grad_norm": 0.2173871397972107,
+      "learning_rate": 3.817712418519676e-06,
+      "loss": 0.3839,
       "step": 114790
     },
     {
-      "epoch": 4.04,
-      "learning_rate": 4.704414170968077e-06,
-      "loss": 0.2592,
+      "epoch": 4.137204022056438,
+      "grad_norm": 0.3121466040611267,
+      "learning_rate": 3.816162659906747e-06,
+      "loss": 0.3741,
       "step": 114795
     },
     {
-      "epoch": 4.04,
-      "learning_rate": 4.702750925813432e-06,
-      "loss": 0.252,
+      "epoch": 4.137384221717663,
+      "grad_norm": 0.2459418624639511,
+      "learning_rate": 3.8146131899194942e-06,
+      "loss": 0.3915,
       "step": 114800
     },
     {
-      "epoch": 4.04,
-      "learning_rate": 4.70108794420519e-06,
-      "loss": 0.2279,
+      "epoch": 4.137564421378888,
+      "grad_norm": 0.2291506677865982,
+      "learning_rate": 3.8130640085790396e-06,
+      "loss": 0.3682,
       "step": 114805
     },
     {
-      "epoch": 4.04,
-      "learning_rate": 4.69942522616495e-06,
-      "loss": 0.2586,
+      "epoch": 4.137744621040112,
+      "grad_norm": 0.2138427197933197,
+      "learning_rate": 3.811515115906489e-06,
+      "loss": 0.3368,
       "step": 114810
     },
     {
-      "epoch": 4.04,
-      "learning_rate": 4.697762771714284e-06,
-      "loss": 0.2462,
+      "epoch": 4.137924820701337,
+      "grad_norm": 0.28790679574012756,
+      "learning_rate": 3.8099665119229382e-06,
+      "loss": 0.3658,
       "step": 114815
     },
     {
-      "epoch": 4.04,
-      "learning_rate": 4.696100580874796e-06,
-      "loss": 0.2519,
+      "epoch": 4.138105020362562,
+      "grad_norm": 0.2095593363046646,
+      "learning_rate": 3.8084181966494966e-06,
+      "loss": 0.3348,
       "step": 114820
     },
     {
-      "epoch": 4.04,
-      "learning_rate": 4.694438653668062e-06,
-      "loss": 0.2567,
+      "epoch": 4.1382852200237865,
+      "grad_norm": 0.28139370679855347,
+      "learning_rate": 3.8068701701072436e-06,
+      "loss": 0.3597,
       "step": 114825
     },
     {
-      "epoch": 4.04,
-      "learning_rate": 4.692776990115661e-06,
-      "loss": 0.2444,
+      "epoch": 4.138465419685011,
+      "grad_norm": 0.2298731952905655,
+      "learning_rate": 3.8053224323172904e-06,
+      "loss": 0.3687,
       "step": 114830
     },
     {
-      "epoch": 4.04,
-      "learning_rate": 4.691115590239159e-06,
-      "loss": 0.2809,
+      "epoch": 4.138645619346236,
+      "grad_norm": 0.2664845287799835,
+      "learning_rate": 3.8037749833007203e-06,
+      "loss": 0.3159,
       "step": 114835
     },
     {
-      "epoch": 4.04,
-      "learning_rate": 4.689454454060138e-06,
-      "loss": 0.2277,
+      "epoch": 4.13882581900746,
+      "grad_norm": 0.21755090355873108,
+      "learning_rate": 3.8022278230785996e-06,
+      "loss": 0.3509,
       "step": 114840
     },
     {
-      "epoch": 4.04,
-      "learning_rate": 4.6877935816001725e-06,
-      "loss": 0.2642,
+      "epoch": 4.139006018668685,
+      "grad_norm": 0.29724082350730896,
+      "learning_rate": 3.8006809516720282e-06,
+      "loss": 0.3392,
       "step": 114845
     },
     {
-      "epoch": 4.04,
-      "learning_rate": 4.686132972880819e-06,
-      "loss": 0.2511,
+      "epoch": 4.139186218329909,
+      "grad_norm": 0.23176833987236023,
+      "learning_rate": 3.79913436910207e-06,
+      "loss": 0.3852,
       "step": 114850
     },
     {
-      "epoch": 4.04,
-      "learning_rate": 4.684472627923639e-06,
-      "loss": 0.236,
+      "epoch": 4.139366417991134,
+      "grad_norm": 0.2422746866941452,
+      "learning_rate": 3.7975880753898045e-06,
+      "loss": 0.3726,
       "step": 114855
     },
     {
-      "epoch": 4.04,
-      "learning_rate": 4.682812546750198e-06,
-      "loss": 0.2635,
+      "epoch": 4.139546617652359,
+      "grad_norm": 0.23337143659591675,
+      "learning_rate": 3.796042070556302e-06,
+      "loss": 0.3838,
       "step": 114860
     },
     {
-      "epoch": 4.04,
-      "learning_rate": 4.681152729382046e-06,
-      "loss": 0.2551,
+      "epoch": 4.1397268173135835,
+      "grad_norm": 0.2680877447128296,
+      "learning_rate": 3.794496354622612e-06,
+      "loss": 0.4008,
       "step": 114865
     },
     {
-      "epoch": 4.04,
-      "learning_rate": 4.6794931758407354e-06,
-      "loss": 0.2644,
+      "epoch": 4.139907016974808,
+      "grad_norm": 0.26528939604759216,
+      "learning_rate": 3.792950927609809e-06,
+      "loss": 0.389,
       "step": 114870
     },
     {
-      "epoch": 4.04,
-      "learning_rate": 4.677833886147809e-06,
-      "loss": 0.2385,
+      "epoch": 4.140087216636033,
+      "grad_norm": 0.2818238139152527,
+      "learning_rate": 3.7914057895389433e-06,
+      "loss": 0.4002,
       "step": 114875
     },
     {
-      "epoch": 4.04,
-      "learning_rate": 4.676174860324822e-06,
-      "loss": 0.2558,
+      "epoch": 4.140267416297258,
+      "grad_norm": 0.2836932837963104,
+      "learning_rate": 3.789860940431067e-06,
+      "loss": 0.3686,
       "step": 114880
     },
     {
-      "epoch": 4.04,
-      "learning_rate": 4.674516098393311e-06,
-      "loss": 0.2599,
+      "epoch": 4.140447615958482,
+      "grad_norm": 0.22321908175945282,
+      "learning_rate": 3.7883163803072277e-06,
+      "loss": 0.3532,
       "step": 114885
     },
     {
-      "epoch": 4.04,
-      "learning_rate": 4.672857600374805e-06,
-      "loss": 0.2463,
+      "epoch": 4.140627815619706,
+      "grad_norm": 0.18690916895866394,
+      "learning_rate": 3.786772109188466e-06,
+      "loss": 0.3564,
       "step": 114890
     },
     {
-      "epoch": 4.04,
-      "learning_rate": 4.671199366290849e-06,
-      "loss": 0.2445,
+      "epoch": 4.140808015280931,
+      "grad_norm": 0.21489687263965607,
+      "learning_rate": 3.7852281270958357e-06,
+      "loss": 0.3558,
       "step": 114895
     },
     {
-      "epoch": 4.04,
-      "learning_rate": 4.66954139616298e-06,
-      "loss": 0.2477,
+      "epoch": 4.140988214942156,
+      "grad_norm": 0.2863061726093292,
+      "learning_rate": 3.7836844340503637e-06,
+      "loss": 0.3493,
       "step": 114900
     },
     {
-      "epoch": 4.04,
-      "learning_rate": 4.6678836900127145e-06,
-      "loss": 0.2558,
+      "epoch": 4.1411684146033805,
+      "grad_norm": 0.22299538552761078,
+      "learning_rate": 3.782141030073083e-06,
+      "loss": 0.356,
       "step": 114905
     },
     {
-      "epoch": 4.04,
-      "learning_rate": 4.666226247861583e-06,
-      "loss": 0.2577,
+      "epoch": 4.141348614264605,
+      "grad_norm": 0.2980153262615204,
+      "learning_rate": 3.7805979151850217e-06,
+      "loss": 0.3812,
       "step": 114910
     },
     {
-      "epoch": 4.04,
-      "learning_rate": 4.6645690697310925e-06,
-      "loss": 0.2541,
+      "epoch": 4.14152881392583,
+      "grad_norm": 0.20422306656837463,
+      "learning_rate": 3.7790550894072073e-06,
+      "loss": 0.362,
       "step": 114915
     },
     {
-      "epoch": 4.04,
-      "learning_rate": 4.662912155642782e-06,
-      "loss": 0.2404,
+      "epoch": 4.141709013587055,
+      "grad_norm": 0.2799051105976105,
+      "learning_rate": 3.7775125527606534e-06,
+      "loss": 0.37,
       "step": 114920
     },
     {
-      "epoch": 4.04,
-      "learning_rate": 4.661255505618153e-06,
-      "loss": 0.2522,
+      "epoch": 4.141889213248279,
+      "grad_norm": 0.2169763743877411,
+      "learning_rate": 3.775970305266385e-06,
+      "loss": 0.3863,
       "step": 114925
     },
     {
-      "epoch": 4.04,
-      "learning_rate": 4.659599119678709e-06,
-      "loss": 0.2446,
+      "epoch": 4.142069412909504,
+      "grad_norm": 0.27121099829673767,
+      "learning_rate": 3.774428346945416e-06,
+      "loss": 0.3563,
       "step": 114930
     },
     {
-      "epoch": 4.04,
-      "learning_rate": 4.657942997845974e-06,
-      "loss": 0.2606,
+      "epoch": 4.142249612570728,
+      "grad_norm": 0.28678664565086365,
+      "learning_rate": 3.772886677818749e-06,
+      "loss": 0.4166,
       "step": 114935
     },
     {
-      "epoch": 4.04,
-      "learning_rate": 4.656287140141444e-06,
-      "loss": 0.248,
+      "epoch": 4.142429812231953,
+      "grad_norm": 0.24329231679439545,
+      "learning_rate": 3.7713452979073913e-06,
+      "loss": 0.372,
       "step": 114940
     },
     {
-      "epoch": 4.04,
-      "learning_rate": 4.65463154658661e-06,
-      "loss": 0.2571,
+      "epoch": 4.1426100118931775,
+      "grad_norm": 0.25663426518440247,
+      "learning_rate": 3.7698042072323386e-06,
+      "loss": 0.3657,
       "step": 114945
     },
     {
-      "epoch": 4.04,
-      "learning_rate": 4.652976217202987e-06,
-      "loss": 0.2576,
+      "epoch": 4.142790211554402,
+      "grad_norm": 0.29921120405197144,
+      "learning_rate": 3.7682634058146065e-06,
+      "loss": 0.3919,
       "step": 114950
     },
     {
-      "epoch": 4.04,
-      "learning_rate": 4.651321152012048e-06,
-      "loss": 0.2447,
+      "epoch": 4.142970411215627,
+      "grad_norm": 0.25063279271125793,
+      "learning_rate": 3.766722893675162e-06,
+      "loss": 0.3886,
       "step": 114955
     },
     {
-      "epoch": 4.04,
-      "learning_rate": 4.649666351035304e-06,
-      "loss": 0.2749,
+      "epoch": 4.143150610876852,
+      "grad_norm": 0.24971787631511688,
+      "learning_rate": 3.76518267083501e-06,
+      "loss": 0.3546,
       "step": 114960
     },
     {
-      "epoch": 4.04,
-      "learning_rate": 4.648011814294229e-06,
-      "loss": 0.2494,
+      "epoch": 4.143330810538076,
+      "grad_norm": 0.25325024127960205,
+      "learning_rate": 3.763642737315137e-06,
+      "loss": 0.3863,
       "step": 114965
     },
     {
-      "epoch": 4.04,
-      "learning_rate": 4.646357541810301e-06,
-      "loss": 0.2668,
+      "epoch": 4.143511010199301,
+      "grad_norm": 0.24295130372047424,
+      "learning_rate": 3.762103093136518e-06,
+      "loss": 0.3941,
       "step": 114970
     },
     {
-      "epoch": 4.05,
-      "learning_rate": 4.644703533605013e-06,
-      "loss": 0.2353,
+      "epoch": 4.143691209860526,
+      "grad_norm": 0.19150157272815704,
+      "learning_rate": 3.7605637383201357e-06,
+      "loss": 0.3463,
       "step": 114975
     },
     {
-      "epoch": 4.05,
-      "learning_rate": 4.643049789699835e-06,
-      "loss": 0.2493,
+      "epoch": 4.14387140952175,
+      "grad_norm": 0.24536427855491638,
+      "learning_rate": 3.759024672886957e-06,
+      "loss": 0.3587,
       "step": 114980
     },
     {
-      "epoch": 4.05,
-      "learning_rate": 4.641396310116241e-06,
-      "loss": 0.2338,
+      "epoch": 4.144051609182974,
+      "grad_norm": 0.3299379348754883,
+      "learning_rate": 3.7574858968579513e-06,
+      "loss": 0.3832,
       "step": 114985
     },
     {
-      "epoch": 4.05,
-      "learning_rate": 4.639743094875693e-06,
-      "loss": 0.2765,
+      "epoch": 4.144231808844199,
+      "grad_norm": 0.2293405532836914,
+      "learning_rate": 3.7559474102540905e-06,
+      "loss": 0.3422,
       "step": 114990
     },
     {
-      "epoch": 4.05,
-      "learning_rate": 4.638090143999668e-06,
-      "loss": 0.2424,
+      "epoch": 4.144412008505424,
+      "grad_norm": 0.29431861639022827,
+      "learning_rate": 3.7544092130963355e-06,
+      "loss": 0.3949,
       "step": 114995
     },
     {
-      "epoch": 4.05,
-      "learning_rate": 4.636437457509618e-06,
-      "loss": 0.2276,
+      "epoch": 4.144592208166649,
+      "grad_norm": 0.23299720883369446,
+      "learning_rate": 3.752871305405642e-06,
+      "loss": 0.3659,
       "step": 115000
     },
     {
-      "epoch": 4.05,
-      "eval_loss": 0.24977658689022064,
-      "eval_runtime": 10.5663,
-      "eval_samples_per_second": 9.464,
-      "eval_steps_per_second": 9.464,
+      "epoch": 4.144592208166649,
+      "eval_loss": 0.4293991029262543,
+      "eval_runtime": 3.5316,
+      "eval_samples_per_second": 28.316,
+      "eval_steps_per_second": 7.079,
       "step": 115000
     },
     {
-      "epoch": 4.05,
-      "learning_rate": 4.634785035427014e-06,
-      "loss": 0.2404,
+      "epoch": 4.144772407827873,
+      "grad_norm": 0.23517471551895142,
+      "learning_rate": 3.7513336872029625e-06,
+      "loss": 0.381,
       "step": 115005
     },
     {
-      "epoch": 4.05,
-      "learning_rate": 4.6331328777733e-06,
-      "loss": 0.2405,
+      "epoch": 4.144952607489098,
+      "grad_norm": 0.24651336669921875,
+      "learning_rate": 3.7497963585092415e-06,
+      "loss": 0.4108,
       "step": 115010
     },
     {
-      "epoch": 4.05,
-      "learning_rate": 4.6314809845699355e-06,
-      "loss": 0.2592,
+      "epoch": 4.145132807150323,
+      "grad_norm": 0.2341834455728531,
+      "learning_rate": 3.7482593193454375e-06,
+      "loss": 0.3443,
       "step": 115015
     },
     {
-      "epoch": 4.05,
-      "learning_rate": 4.629829355838372e-06,
-      "loss": 0.2474,
+      "epoch": 4.145313006811548,
+      "grad_norm": 0.2706274092197418,
+      "learning_rate": 3.7467225697324833e-06,
+      "loss": 0.3762,
       "step": 115020
     },
     {
-      "epoch": 4.05,
-      "learning_rate": 4.628177991600047e-06,
-      "loss": 0.2443,
+      "epoch": 4.145493206472771,
+      "grad_norm": 0.2560116946697235,
+      "learning_rate": 3.7451861096913234e-06,
+      "loss": 0.3493,
       "step": 115025
     },
     {
-      "epoch": 4.05,
-      "learning_rate": 4.626526891876398e-06,
-      "loss": 0.2654,
+      "epoch": 4.145673406133996,
+      "grad_norm": 0.23122893273830414,
+      "learning_rate": 3.7436499392428855e-06,
+      "loss": 0.3732,
       "step": 115030
     },
     {
-      "epoch": 4.05,
-      "learning_rate": 4.624876056688879e-06,
-      "loss": 0.2398,
+      "epoch": 4.145853605795221,
+      "grad_norm": 0.2816644012928009,
+      "learning_rate": 3.7421140584080942e-06,
+      "loss": 0.378,
       "step": 115035
     },
     {
-      "epoch": 4.05,
-      "learning_rate": 4.623225486058918e-06,
-      "loss": 0.2607,
+      "epoch": 4.146033805456446,
+      "grad_norm": 0.3167472183704376,
+      "learning_rate": 3.740578467207892e-06,
+      "loss": 0.4059,
       "step": 115040
     },
     {
-      "epoch": 4.05,
-      "learning_rate": 4.621575180007937e-06,
-      "loss": 0.2391,
+      "epoch": 4.14621400511767,
+      "grad_norm": 0.22045278549194336,
+      "learning_rate": 3.7390431656631975e-06,
+      "loss": 0.3764,
       "step": 115045
     },
     {
-      "epoch": 4.05,
-      "learning_rate": 4.61992513855738e-06,
-      "loss": 0.2658,
+      "epoch": 4.146394204778895,
+      "grad_norm": 0.25917506217956543,
+      "learning_rate": 3.737508153794911e-06,
+      "loss": 0.3721,
       "step": 115050
     },
     {
-      "epoch": 4.05,
-      "learning_rate": 4.61827536172866e-06,
-      "loss": 0.2454,
+      "epoch": 4.14657440444012,
+      "grad_norm": 0.2379886656999588,
+      "learning_rate": 3.7359734316239653e-06,
+      "loss": 0.3326,
       "step": 115055
     },
     {
-      "epoch": 4.05,
-      "learning_rate": 4.616625849543199e-06,
-      "loss": 0.271,
+      "epoch": 4.1467546041013446,
+      "grad_norm": 0.3082965612411499,
+      "learning_rate": 3.7344389991712666e-06,
+      "loss": 0.339,
       "step": 115060
     },
     {
-      "epoch": 4.05,
-      "learning_rate": 4.61497660202242e-06,
-      "loss": 0.2539,
+      "epoch": 4.146934803762569,
+      "grad_norm": 0.25842636823654175,
+      "learning_rate": 3.732904856457717e-06,
+      "loss": 0.3674,
       "step": 115065
     },
     {
-      "epoch": 4.05,
-      "learning_rate": 4.6133276191877405e-06,
-      "loss": 0.2412,
+      "epoch": 4.147115003423793,
+      "grad_norm": 0.2803257405757904,
+      "learning_rate": 3.731371003504222e-06,
+      "loss": 0.3743,
       "step": 115070
     },
     {
-      "epoch": 4.05,
-      "learning_rate": 4.611678901060565e-06,
-      "loss": 0.2474,
+      "epoch": 4.147295203085018,
+      "grad_norm": 0.1919623762369156,
+      "learning_rate": 3.7298374403316737e-06,
+      "loss": 0.3497,
       "step": 115075
     },
     {
-      "epoch": 4.05,
-      "learning_rate": 4.610030447662303e-06,
-      "loss": 0.2609,
+      "epoch": 4.147475402746243,
+      "grad_norm": 0.25834715366363525,
+      "learning_rate": 3.7283041669609775e-06,
+      "loss": 0.3668,
       "step": 115080
     },
     {
-      "epoch": 4.05,
-      "learning_rate": 4.608382259014355e-06,
-      "loss": 0.2499,
+      "epoch": 4.147655602407467,
+      "grad_norm": 0.2559783458709717,
+      "learning_rate": 3.7267711834130194e-06,
+      "loss": 0.3553,
       "step": 115085
     },
     {
-      "epoch": 4.05,
-      "learning_rate": 4.606734335138127e-06,
-      "loss": 0.2305,
+      "epoch": 4.147835802068692,
+      "grad_norm": 0.30941301584243774,
+      "learning_rate": 3.725238489708685e-06,
+      "loss": 0.3845,
       "step": 115090
     },
     {
-      "epoch": 4.05,
-      "learning_rate": 4.6050866760550145e-06,
-      "loss": 0.2512,
+      "epoch": 4.148016001729917,
+      "grad_norm": 0.22858378291130066,
+      "learning_rate": 3.7237060858688584e-06,
+      "loss": 0.3625,
       "step": 115095
     },
     {
-      "epoch": 4.05,
-      "learning_rate": 4.6034392817864145e-06,
-      "loss": 0.2668,
+      "epoch": 4.1481962013911415,
+      "grad_norm": 0.2147953361272812,
+      "learning_rate": 3.722173971914417e-06,
+      "loss": 0.3643,
       "step": 115100
     },
     {
-      "epoch": 4.05,
-      "learning_rate": 4.601792152353704e-06,
-      "loss": 0.2516,
+      "epoch": 4.148376401052366,
+      "grad_norm": 0.3191777169704437,
+      "learning_rate": 3.7206421478662294e-06,
+      "loss": 0.3437,
       "step": 115105
     },
     {
-      "epoch": 4.05,
-      "learning_rate": 4.600145287778285e-06,
-      "loss": 0.248,
+      "epoch": 4.148556600713591,
+      "grad_norm": 0.2847544848918915,
+      "learning_rate": 3.7191106137451776e-06,
+      "loss": 0.3535,
       "step": 115110
     },
     {
-      "epoch": 4.05,
-      "learning_rate": 4.598498688081532e-06,
-      "loss": 0.2313,
+      "epoch": 4.148736800374815,
+      "grad_norm": 0.3304956555366516,
+      "learning_rate": 3.7175793695721244e-06,
+      "loss": 0.3658,
       "step": 115115
     },
     {
-      "epoch": 4.05,
-      "learning_rate": 4.596852353284834e-06,
-      "loss": 0.2647,
+      "epoch": 4.14891700003604,
+      "grad_norm": 0.2144591212272644,
+      "learning_rate": 3.7160484153679313e-06,
+      "loss": 0.3673,
       "step": 115120
     },
     {
-      "epoch": 4.05,
-      "learning_rate": 4.595206283409554e-06,
-      "loss": 0.2435,
+      "epoch": 4.149097199697264,
+      "grad_norm": 0.2552328407764435,
+      "learning_rate": 3.714517751153457e-06,
+      "loss": 0.3804,
       "step": 115125
     },
     {
-      "epoch": 4.05,
-      "learning_rate": 4.59356047847708e-06,
-      "loss": 0.2797,
+      "epoch": 4.149277399358489,
+      "grad_norm": 0.19830626249313354,
+      "learning_rate": 3.7129873769495534e-06,
+      "loss": 0.3935,
       "step": 115130
     },
     {
-      "epoch": 4.05,
-      "learning_rate": 4.591914938508776e-06,
-      "loss": 0.2365,
+      "epoch": 4.149457599019714,
+      "grad_norm": 0.22259441018104553,
+      "learning_rate": 3.7114572927770875e-06,
+      "loss": 0.3558,
       "step": 115135
     },
     {
-      "epoch": 4.05,
-      "learning_rate": 4.590269663526006e-06,
-      "loss": 0.2377,
+      "epoch": 4.1496377986809385,
+      "grad_norm": 0.26004859805107117,
+      "learning_rate": 3.709927498656887e-06,
+      "loss": 0.3926,
       "step": 115140
     },
     {
-      "epoch": 4.05,
-      "learning_rate": 4.5886246535501275e-06,
-      "loss": 0.2622,
+      "epoch": 4.149817998342163,
+      "grad_norm": 0.2216162085533142,
+      "learning_rate": 3.708397994609797e-06,
+      "loss": 0.3725,
       "step": 115145
     },
     {
-      "epoch": 4.05,
-      "learning_rate": 4.586979908602515e-06,
-      "loss": 0.2367,
+      "epoch": 4.149998198003388,
+      "grad_norm": 0.23738591372966766,
+      "learning_rate": 3.7068687806566694e-06,
+      "loss": 0.3529,
       "step": 115150
     },
     {
-      "epoch": 4.05,
-      "learning_rate": 4.585335428704512e-06,
-      "loss": 0.254,
+      "epoch": 4.150178397664613,
+      "grad_norm": 0.2530551254749298,
+      "learning_rate": 3.705339856818324e-06,
+      "loss": 0.3268,
       "step": 115155
     },
     {
-      "epoch": 4.05,
-      "learning_rate": 4.583691213877472e-06,
-      "loss": 0.2522,
+      "epoch": 4.1503585973258375,
+      "grad_norm": 0.20809151232242584,
+      "learning_rate": 3.7038112231156134e-06,
+      "loss": 0.3447,
       "step": 115160
     },
     {
-      "epoch": 4.05,
-      "learning_rate": 4.5820472641427495e-06,
-      "loss": 0.249,
+      "epoch": 4.150538796987061,
+      "grad_norm": 0.22564588487148285,
+      "learning_rate": 3.7022828795693467e-06,
+      "loss": 0.3774,
       "step": 115165
     },
     {
-      "epoch": 4.05,
-      "learning_rate": 4.580403579521683e-06,
-      "loss": 0.2377,
+      "epoch": 4.150718996648286,
+      "grad_norm": 0.2013123780488968,
+      "learning_rate": 3.700754826200345e-06,
+      "loss": 0.3752,
       "step": 115170
     },
     {
-      "epoch": 4.05,
-      "learning_rate": 4.578760160035625e-06,
-      "loss": 0.2376,
+      "epoch": 4.150899196309511,
+      "grad_norm": 0.29589053988456726,
+      "learning_rate": 3.6992270630294396e-06,
+      "loss": 0.345,
       "step": 115175
     },
     {
-      "epoch": 4.05,
-      "learning_rate": 4.577117005705902e-06,
-      "loss": 0.2611,
+      "epoch": 4.1510793959707355,
+      "grad_norm": 0.2514277994632721,
+      "learning_rate": 3.697699590077444e-06,
+      "loss": 0.4192,
       "step": 115180
     },
     {
-      "epoch": 4.05,
-      "learning_rate": 4.575474116553863e-06,
-      "loss": 0.2428,
+      "epoch": 4.15125959563196,
+      "grad_norm": 0.23917201161384583,
+      "learning_rate": 3.696172407365167e-06,
+      "loss": 0.3958,
       "step": 115185
     },
     {
-      "epoch": 4.05,
-      "learning_rate": 4.573831492600831e-06,
-      "loss": 0.283,
+      "epoch": 4.151439795293185,
+      "grad_norm": 0.3023105561733246,
+      "learning_rate": 3.694645514913417e-06,
+      "loss": 0.3855,
       "step": 115190
     },
     {
-      "epoch": 4.05,
-      "learning_rate": 4.5721891338681345e-06,
-      "loss": 0.2481,
+      "epoch": 4.15161999495441,
+      "grad_norm": 0.20139765739440918,
+      "learning_rate": 3.693118912742988e-06,
+      "loss": 0.3553,
       "step": 115195
     },
     {
-      "epoch": 4.05,
-      "learning_rate": 4.570547040377096e-06,
-      "loss": 0.2675,
+      "epoch": 4.151800194615634,
+      "grad_norm": 0.25707682967185974,
+      "learning_rate": 3.6915926008746964e-06,
+      "loss": 0.3776,
       "step": 115200
     },
     {
-      "epoch": 4.05,
-      "learning_rate": 4.568905212149047e-06,
-      "loss": 0.2363,
+      "epoch": 4.151980394276859,
+      "grad_norm": 0.2819069027900696,
+      "learning_rate": 3.690066579329332e-06,
+      "loss": 0.4,
       "step": 115205
     },
     {
-      "epoch": 4.05,
-      "learning_rate": 4.567263649205297e-06,
-      "loss": 0.2414,
+      "epoch": 4.152160593938083,
+      "grad_norm": 0.24341073632240295,
+      "learning_rate": 3.6885408481276833e-06,
+      "loss": 0.3162,
       "step": 115210
     },
     {
-      "epoch": 4.05,
-      "learning_rate": 4.565622351567164e-06,
-      "loss": 0.2436,
+      "epoch": 4.152340793599308,
+      "grad_norm": 0.262385755777359,
+      "learning_rate": 3.6870154072905396e-06,
+      "loss": 0.4045,
       "step": 115215
     },
     {
-      "epoch": 4.05,
-      "learning_rate": 4.56398131925595e-06,
-      "loss": 0.2647,
+      "epoch": 4.1525209932605325,
+      "grad_norm": 0.2288028746843338,
+      "learning_rate": 3.6854902568386777e-06,
+      "loss": 0.3662,
       "step": 115220
     },
     {
-      "epoch": 4.05,
-      "learning_rate": 4.5623405522929735e-06,
-      "loss": 0.2445,
+      "epoch": 4.152701192921757,
+      "grad_norm": 0.3026867210865021,
+      "learning_rate": 3.6839653967928907e-06,
+      "loss": 0.3821,
       "step": 115225
     },
     {
-      "epoch": 4.05,
-      "learning_rate": 4.56070005069954e-06,
-      "loss": 0.2502,
+      "epoch": 4.152881392582982,
+      "grad_norm": 0.2670912742614746,
+      "learning_rate": 3.6824408271739552e-06,
+      "loss": 0.3924,
       "step": 115230
     },
     {
-      "epoch": 4.05,
-      "learning_rate": 4.559059814496947e-06,
-      "loss": 0.2407,
+      "epoch": 4.153061592244207,
+      "grad_norm": 0.2640323340892792,
+      "learning_rate": 3.6809165480026246e-06,
+      "loss": 0.3696,
       "step": 115235
     },
     {
-      "epoch": 4.05,
-      "learning_rate": 4.557419843706484e-06,
-      "loss": 0.2386,
+      "epoch": 4.153241791905431,
+      "grad_norm": 0.272617369890213,
+      "learning_rate": 3.6793925592996825e-06,
+      "loss": 0.3597,
       "step": 115240
     },
     {
-      "epoch": 4.05,
-      "learning_rate": 4.555780138349461e-06,
-      "loss": 0.2654,
+      "epoch": 4.153421991566656,
+      "grad_norm": 0.2459690123796463,
+      "learning_rate": 3.6778688610858896e-06,
+      "loss": 0.3587,
       "step": 115245
     },
     {
-      "epoch": 4.05,
-      "learning_rate": 4.55414069844716e-06,
-      "loss": 0.2552,
+      "epoch": 4.153602191227881,
+      "grad_norm": 0.22347575426101685,
+      "learning_rate": 3.6763454533820075e-06,
+      "loss": 0.3799,
       "step": 115250
     },
     {
-      "epoch": 4.05,
-      "learning_rate": 4.552501524020863e-06,
-      "loss": 0.2426,
+      "epoch": 4.153782390889105,
+      "grad_norm": 0.2681044042110443,
+      "learning_rate": 3.6748223362087886e-06,
+      "loss": 0.3968,
       "step": 115255
     },
     {
-      "epoch": 4.06,
-      "learning_rate": 4.550862615091855e-06,
-      "loss": 0.2663,
+      "epoch": 4.1539625905503295,
+      "grad_norm": 0.2700466215610504,
+      "learning_rate": 3.6732995095869806e-06,
+      "loss": 0.3832,
       "step": 115260
     },
     {
-      "epoch": 4.06,
-      "learning_rate": 4.549223971681426e-06,
-      "loss": 0.2714,
+      "epoch": 4.154142790211554,
+      "grad_norm": 0.24011053144931793,
+      "learning_rate": 3.671776973537344e-06,
+      "loss": 0.3449,
       "step": 115265
     },
     {
-      "epoch": 4.06,
-      "learning_rate": 4.547585593810846e-06,
-      "loss": 0.2595,
+      "epoch": 4.154322989872779,
+      "grad_norm": 0.26308053731918335,
+      "learning_rate": 3.670254728080616e-06,
+      "loss": 0.3731,
       "step": 115270
     },
     {
-      "epoch": 4.06,
-      "learning_rate": 4.545947481501384e-06,
-      "loss": 0.2411,
+      "epoch": 4.154503189534004,
+      "grad_norm": 0.19285693764686584,
+      "learning_rate": 3.6687327732375376e-06,
+      "loss": 0.3644,
       "step": 115275
     },
     {
-      "epoch": 4.06,
-      "learning_rate": 4.544309634774313e-06,
-      "loss": 0.2261,
+      "epoch": 4.154683389195228,
+      "grad_norm": 0.327217698097229,
+      "learning_rate": 3.667211109028848e-06,
+      "loss": 0.4314,
       "step": 115280
     },
     {
-      "epoch": 4.06,
-      "learning_rate": 4.542672053650904e-06,
-      "loss": 0.2438,
+      "epoch": 4.154863588856453,
+      "grad_norm": 0.22068528831005096,
+      "learning_rate": 3.665689735475275e-06,
+      "loss": 0.3825,
       "step": 115285
     },
     {
-      "epoch": 4.06,
-      "learning_rate": 4.541034738152419e-06,
-      "loss": 0.2659,
+      "epoch": 4.155043788517678,
+      "grad_norm": 0.23468096554279327,
+      "learning_rate": 3.664168652597541e-06,
+      "loss": 0.3646,
       "step": 115290
     },
     {
-      "epoch": 4.06,
-      "learning_rate": 4.539397688300109e-06,
-      "loss": 0.2445,
+      "epoch": 4.155223988178903,
+      "grad_norm": 0.2511104941368103,
+      "learning_rate": 3.6626478604163877e-06,
+      "loss": 0.3713,
       "step": 115295
     },
     {
-      "epoch": 4.06,
-      "learning_rate": 4.537760904115243e-06,
-      "loss": 0.2389,
+      "epoch": 4.155404187840126,
+      "grad_norm": 0.2715943157672882,
+      "learning_rate": 3.6611273589525236e-06,
+      "loss": 0.3386,
       "step": 115300
     },
     {
-      "epoch": 4.06,
-      "learning_rate": 4.5361243856190645e-06,
-      "loss": 0.2672,
+      "epoch": 4.155584387501351,
+      "grad_norm": 0.2659306526184082,
+      "learning_rate": 3.659607148226671e-06,
+      "loss": 0.3665,
       "step": 115305
     },
     {
-      "epoch": 4.06,
-      "learning_rate": 4.534488132832826e-06,
-      "loss": 0.2355,
+      "epoch": 4.155764587162576,
+      "grad_norm": 0.2604070007801056,
+      "learning_rate": 3.658087228259538e-06,
+      "loss": 0.4022,
       "step": 115310
     },
     {
-      "epoch": 4.06,
-      "learning_rate": 4.532852145777763e-06,
-      "loss": 0.2406,
+      "epoch": 4.155944786823801,
+      "grad_norm": 0.25114676356315613,
+      "learning_rate": 3.6565675990718277e-06,
+      "loss": 0.3772,
       "step": 115315
     },
     {
-      "epoch": 4.06,
-      "learning_rate": 4.531216424475135e-06,
-      "loss": 0.2558,
+      "epoch": 4.156124986485025,
+      "grad_norm": 0.25636595487594604,
+      "learning_rate": 3.6550482606842572e-06,
+      "loss": 0.3437,
       "step": 115320
     },
     {
-      "epoch": 4.06,
-      "learning_rate": 4.529580968946173e-06,
-      "loss": 0.235,
+      "epoch": 4.15630518614625,
+      "grad_norm": 0.23536555469036102,
+      "learning_rate": 3.653529213117529e-06,
+      "loss": 0.391,
       "step": 115325
     },
     {
-      "epoch": 4.06,
-      "learning_rate": 4.5279457792121034e-06,
-      "loss": 0.2618,
+      "epoch": 4.156485385807475,
+      "grad_norm": 0.26349982619285583,
+      "learning_rate": 3.6520104563923213e-06,
+      "loss": 0.3576,
       "step": 115330
     },
     {
-      "epoch": 4.06,
-      "learning_rate": 4.526310855294166e-06,
-      "loss": 0.2475,
+      "epoch": 4.1566655854687,
+      "grad_norm": 0.26209887862205505,
+      "learning_rate": 3.6504919905293422e-06,
+      "loss": 0.3723,
       "step": 115335
     },
     {
-      "epoch": 4.06,
-      "learning_rate": 4.524676197213598e-06,
-      "loss": 0.2574,
+      "epoch": 4.156845785129924,
+      "grad_norm": 0.21833176910877228,
+      "learning_rate": 3.6489738155492696e-06,
+      "loss": 0.3723,
       "step": 115340
     },
     {
-      "epoch": 4.06,
-      "learning_rate": 4.523041804991615e-06,
-      "loss": 0.235,
+      "epoch": 4.157025984791149,
+      "grad_norm": 0.29597973823547363,
+      "learning_rate": 3.6474559314728097e-06,
+      "loss": 0.3672,
       "step": 115345
     },
     {
-      "epoch": 4.06,
-      "learning_rate": 4.521407678649439e-06,
-      "loss": 0.2392,
+      "epoch": 4.157206184452373,
+      "grad_norm": 0.24725015461444855,
+      "learning_rate": 3.6459383383206202e-06,
+      "loss": 0.3619,
       "step": 115350
     },
     {
-      "epoch": 4.06,
-      "learning_rate": 4.519773818208284e-06,
-      "loss": 0.2594,
+      "epoch": 4.157386384113598,
+      "grad_norm": 0.29851892590522766,
+      "learning_rate": 3.6444210361133824e-06,
+      "loss": 0.3806,
       "step": 115355
     },
     {
-      "epoch": 4.06,
-      "learning_rate": 4.518140223689371e-06,
-      "loss": 0.249,
+      "epoch": 4.157566583774822,
+      "grad_norm": 0.2327483892440796,
+      "learning_rate": 3.6429040248717795e-06,
+      "loss": 0.3437,
       "step": 115360
     },
     {
-      "epoch": 4.06,
-      "learning_rate": 4.5165068951139115e-06,
-      "loss": 0.2587,
+      "epoch": 4.157746783436047,
+      "grad_norm": 0.23190578818321228,
+      "learning_rate": 3.6413873046164728e-06,
+      "loss": 0.3671,
       "step": 115365
     },
     {
-      "epoch": 4.06,
-      "learning_rate": 4.514873832503111e-06,
-      "loss": 0.2529,
+      "epoch": 4.157926983097272,
+      "grad_norm": 0.21011842787265778,
+      "learning_rate": 3.6398708753681286e-06,
+      "loss": 0.3481,
       "step": 115370
     },
     {
-      "epoch": 4.06,
-      "learning_rate": 4.513241035878168e-06,
-      "loss": 0.2408,
+      "epoch": 4.158107182758497,
+      "grad_norm": 0.3232230544090271,
+      "learning_rate": 3.638354737147412e-06,
+      "loss": 0.4116,
       "step": 115375
     },
     {
-      "epoch": 4.06,
-      "learning_rate": 4.511608505260295e-06,
-      "loss": 0.2692,
+      "epoch": 4.158287382419721,
+      "grad_norm": 0.23676584661006927,
+      "learning_rate": 3.6368388899749667e-06,
+      "loss": 0.3751,
       "step": 115380
     },
     {
-      "epoch": 4.06,
-      "learning_rate": 4.509976240670677e-06,
-      "loss": 0.2393,
+      "epoch": 4.158467582080946,
+      "grad_norm": 0.2382199913263321,
+      "learning_rate": 3.6353233338714626e-06,
+      "loss": 0.3765,
       "step": 115385
     },
     {
-      "epoch": 4.06,
-      "learning_rate": 4.508344242130519e-06,
-      "loss": 0.2522,
+      "epoch": 4.15864778174217,
+      "grad_norm": 0.2705291509628296,
+      "learning_rate": 3.6338080688575415e-06,
+      "loss": 0.3765,
       "step": 115390
     },
     {
-      "epoch": 4.06,
-      "learning_rate": 4.506712509661002e-06,
-      "loss": 0.2416,
+      "epoch": 4.158827981403395,
+      "grad_norm": 0.25305843353271484,
+      "learning_rate": 3.632293094953848e-06,
+      "loss": 0.3927,
       "step": 115395
     },
     {
-      "epoch": 4.06,
-      "learning_rate": 4.505081043283321e-06,
-      "loss": 0.2572,
+      "epoch": 4.159008181064619,
+      "grad_norm": 0.20932044088840485,
+      "learning_rate": 3.6307784121810267e-06,
+      "loss": 0.3489,
       "step": 115400
     },
     {
-      "epoch": 4.06,
-      "learning_rate": 4.503449843018659e-06,
-      "loss": 0.2597,
+      "epoch": 4.159188380725844,
+      "grad_norm": 0.22791607677936554,
+      "learning_rate": 3.6292640205597055e-06,
+      "loss": 0.3658,
       "step": 115405
     },
     {
-      "epoch": 4.06,
-      "learning_rate": 4.501818908888184e-06,
-      "loss": 0.2414,
+      "epoch": 4.159368580387069,
+      "grad_norm": 0.24676699936389923,
+      "learning_rate": 3.627749920110529e-06,
+      "loss": 0.3737,
       "step": 115410
     },
     {
-      "epoch": 4.06,
-      "learning_rate": 4.500188240913089e-06,
-      "loss": 0.2503,
+      "epoch": 4.1595487800482935,
+      "grad_norm": 0.26706597208976746,
+      "learning_rate": 3.626236110854128e-06,
+      "loss": 0.3784,
       "step": 115415
     },
     {
-      "epoch": 4.06,
-      "learning_rate": 4.498557839114542e-06,
-      "loss": 0.2564,
+      "epoch": 4.159728979709518,
+      "grad_norm": 0.22861981391906738,
+      "learning_rate": 3.624722592811111e-06,
+      "loss": 0.3555,
       "step": 115420
     },
     {
-      "epoch": 4.06,
-      "learning_rate": 4.496927703513709e-06,
-      "loss": 0.2322,
+      "epoch": 4.159909179370743,
+      "grad_norm": 0.24083659052848816,
+      "learning_rate": 3.623209366002117e-06,
+      "loss": 0.3431,
       "step": 115425
     },
     {
-      "epoch": 4.06,
-      "learning_rate": 4.495297834131751e-06,
-      "loss": 0.2533,
+      "epoch": 4.160089379031968,
+      "grad_norm": 0.19510973989963531,
+      "learning_rate": 3.6216964304477547e-06,
+      "loss": 0.3485,
       "step": 115430
     },
     {
-      "epoch": 4.06,
-      "learning_rate": 4.493668230989842e-06,
-      "loss": 0.2587,
+      "epoch": 4.1602695786931925,
+      "grad_norm": 0.27537184953689575,
+      "learning_rate": 3.620183786168635e-06,
+      "loss": 0.3949,
       "step": 115435
     },
     {
-      "epoch": 4.06,
-      "learning_rate": 4.492038894109133e-06,
-      "loss": 0.2619,
+      "epoch": 4.160449778354416,
+      "grad_norm": 0.23948287963867188,
+      "learning_rate": 3.618671433185383e-06,
+      "loss": 0.3744,
       "step": 115440
     },
     {
-      "epoch": 4.06,
-      "learning_rate": 4.490409823510791e-06,
-      "loss": 0.2575,
+      "epoch": 4.160629978015641,
+      "grad_norm": 0.2727479040622711,
+      "learning_rate": 3.617159371518583e-06,
+      "loss": 0.3983,
       "step": 115445
     },
     {
-      "epoch": 4.06,
-      "learning_rate": 4.488781019215951e-06,
-      "loss": 0.2596,
+      "epoch": 4.160810177676866,
+      "grad_norm": 0.2685748040676117,
+      "learning_rate": 3.6156476011888514e-06,
+      "loss": 0.3499,
       "step": 115450
     },
     {
-      "epoch": 4.06,
-      "learning_rate": 4.487152481245782e-06,
-      "loss": 0.2451,
+      "epoch": 4.1609903773380905,
+      "grad_norm": 0.24433422088623047,
+      "learning_rate": 3.6141361222167826e-06,
+      "loss": 0.3799,
       "step": 115455
     },
     {
-      "epoch": 4.06,
-      "learning_rate": 4.4855242096214195e-06,
-      "loss": 0.2547,
+      "epoch": 4.161170576999315,
+      "grad_norm": 0.24008378386497498,
+      "learning_rate": 3.6126249346229656e-06,
+      "loss": 0.3337,
       "step": 115460
     },
     {
-      "epoch": 4.06,
-      "learning_rate": 4.483896204364002e-06,
-      "loss": 0.248,
+      "epoch": 4.16135077666054,
+      "grad_norm": 0.28376707434654236,
+      "learning_rate": 3.6111140384279955e-06,
+      "loss": 0.3982,
       "step": 115465
     },
     {
-      "epoch": 4.06,
-      "learning_rate": 4.482268465494669e-06,
-      "loss": 0.2582,
+      "epoch": 4.161530976321765,
+      "grad_norm": 0.22135163843631744,
+      "learning_rate": 3.609603433652453e-06,
+      "loss": 0.3829,
       "step": 115470
     },
     {
-      "epoch": 4.06,
-      "learning_rate": 4.480640993034563e-06,
-      "loss": 0.2328,
+      "epoch": 4.1617111759829895,
+      "grad_norm": 0.2689472436904907,
+      "learning_rate": 3.608093120316919e-06,
+      "loss": 0.366,
       "step": 115475
     },
     {
-      "epoch": 4.06,
-      "learning_rate": 4.479013787004813e-06,
-      "loss": 0.2704,
+      "epoch": 4.161891375644214,
+      "grad_norm": 0.24480800330638885,
+      "learning_rate": 3.606583098441982e-06,
+      "loss": 0.3782,
       "step": 115480
     },
     {
-      "epoch": 4.06,
-      "learning_rate": 4.477386847426543e-06,
-      "loss": 0.2622,
+      "epoch": 4.162071575305438,
+      "grad_norm": 0.27388742566108704,
+      "learning_rate": 3.6050733680482063e-06,
+      "loss": 0.3936,
       "step": 115485
     },
     {
-      "epoch": 4.06,
-      "learning_rate": 4.475760174320872e-06,
-      "loss": 0.2616,
+      "epoch": 4.162251774966663,
+      "grad_norm": 0.2704862356185913,
+      "learning_rate": 3.603563929156162e-06,
+      "loss": 0.3431,
       "step": 115490
     },
     {
-      "epoch": 4.06,
-      "learning_rate": 4.474133767708938e-06,
-      "loss": 0.2608,
+      "epoch": 4.1624319746278875,
+      "grad_norm": 0.6965771913528442,
+      "learning_rate": 3.602054781786418e-06,
+      "loss": 0.4249,
       "step": 115495
     },
     {
-      "epoch": 4.06,
-      "learning_rate": 4.472507627611844e-06,
-      "loss": 0.2762,
+      "epoch": 4.162612174289112,
+      "grad_norm": 0.32166483998298645,
+      "learning_rate": 3.600545925959531e-06,
+      "loss": 0.3667,
       "step": 115500
     },
     {
-      "epoch": 4.06,
-      "eval_loss": 0.2497061789035797,
-      "eval_runtime": 10.5326,
-      "eval_samples_per_second": 9.494,
-      "eval_steps_per_second": 9.494,
+      "epoch": 4.162612174289112,
+      "eval_loss": 0.42964106798171997,
+      "eval_runtime": 3.5337,
+      "eval_samples_per_second": 28.299,
+      "eval_steps_per_second": 7.075,
       "step": 115500
     },
     {
-      "epoch": 4.06,
-      "learning_rate": 4.470881754050716e-06,
-      "loss": 0.254,
+      "epoch": 4.162792373950337,
+      "grad_norm": 0.19025029242038727,
+      "learning_rate": 3.599037361696067e-06,
+      "loss": 0.3388,
       "step": 115505
     },
     {
-      "epoch": 4.06,
-      "learning_rate": 4.469256147046652e-06,
-      "loss": 0.2564,
+      "epoch": 4.162972573611562,
+      "grad_norm": 0.27333715558052063,
+      "learning_rate": 3.5975290890165796e-06,
+      "loss": 0.3742,
       "step": 115510
     },
     {
-      "epoch": 4.06,
-      "learning_rate": 4.467630806620773e-06,
-      "loss": 0.2523,
+      "epoch": 4.163152773272786,
+      "grad_norm": 0.28083357214927673,
+      "learning_rate": 3.596021107941605e-06,
+      "loss": 0.3722,
       "step": 115515
     },
     {
-      "epoch": 4.06,
-      "learning_rate": 4.4660057327941765e-06,
-      "loss": 0.2488,
+      "epoch": 4.163332972934011,
+      "grad_norm": 0.26710939407348633,
+      "learning_rate": 3.5945134184917046e-06,
+      "loss": 0.3906,
       "step": 115520
     },
     {
-      "epoch": 4.06,
-      "learning_rate": 4.464380925587958e-06,
-      "loss": 0.2506,
+      "epoch": 4.163513172595236,
+      "grad_norm": 0.23718206584453583,
+      "learning_rate": 3.5930060206874087e-06,
+      "loss": 0.3612,
       "step": 115525
     },
     {
-      "epoch": 4.06,
-      "learning_rate": 4.462756385023226e-06,
-      "loss": 0.2627,
+      "epoch": 4.16369337225646,
+      "grad_norm": 0.23605412244796753,
+      "learning_rate": 3.591498914549274e-06,
+      "loss": 0.359,
       "step": 115530
     },
     {
-      "epoch": 4.06,
-      "learning_rate": 4.461132111121066e-06,
-      "loss": 0.2383,
+      "epoch": 4.1638735719176845,
+      "grad_norm": 0.2375323325395584,
+      "learning_rate": 3.5899921000978137e-06,
+      "loss": 0.4013,
       "step": 115535
     },
     {
-      "epoch": 4.07,
-      "learning_rate": 4.459508103902573e-06,
-      "loss": 0.2519,
+      "epoch": 4.164053771578909,
+      "grad_norm": 0.19733993709087372,
+      "learning_rate": 3.588485577353562e-06,
+      "loss": 0.3864,
       "step": 115540
     },
     {
-      "epoch": 4.07,
-      "learning_rate": 4.457884363388823e-06,
-      "loss": 0.2319,
+      "epoch": 4.164233971240134,
+      "grad_norm": 0.28855153918266296,
+      "learning_rate": 3.5869793463370547e-06,
+      "loss": 0.3374,
       "step": 115545
     },
     {
-      "epoch": 4.07,
-      "learning_rate": 4.456260889600913e-06,
-      "loss": 0.2517,
+      "epoch": 4.164414170901359,
+      "grad_norm": 0.24822647869586945,
+      "learning_rate": 3.585473407068801e-06,
+      "loss": 0.3587,
       "step": 115550
     },
     {
-      "epoch": 4.07,
-      "learning_rate": 4.454637682559909e-06,
-      "loss": 0.2411,
+      "epoch": 4.164594370562583,
+      "grad_norm": 0.26548629999160767,
+      "learning_rate": 3.58396775956934e-06,
+      "loss": 0.3545,
       "step": 115555
     },
     {
-      "epoch": 4.07,
-      "learning_rate": 4.453014742286901e-06,
-      "loss": 0.2421,
+      "epoch": 4.164774570223808,
+      "grad_norm": 0.20807038247585297,
+      "learning_rate": 3.582462403859163e-06,
+      "loss": 0.3993,
       "step": 115560
     },
     {
-      "epoch": 4.07,
-      "learning_rate": 4.451392068802951e-06,
-      "loss": 0.2497,
+      "epoch": 4.164954769885033,
+      "grad_norm": 0.20517727732658386,
+      "learning_rate": 3.5809573399587846e-06,
+      "loss": 0.3603,
       "step": 115565
     },
     {
-      "epoch": 4.07,
-      "learning_rate": 4.449769662129138e-06,
-      "loss": 0.2512,
+      "epoch": 4.165134969546258,
+      "grad_norm": 0.27609434723854065,
+      "learning_rate": 3.579452567888722e-06,
+      "loss": 0.387,
       "step": 115570
     },
     {
-      "epoch": 4.07,
-      "learning_rate": 4.448147522286525e-06,
-      "loss": 0.256,
+      "epoch": 4.1653151692074815,
+      "grad_norm": 0.24407240748405457,
+      "learning_rate": 3.577948087669472e-06,
+      "loss": 0.4059,
       "step": 115575
     },
     {
-      "epoch": 4.07,
-      "learning_rate": 4.446525649296168e-06,
-      "loss": 0.2621,
+      "epoch": 4.165495368868706,
+      "grad_norm": 0.2198261171579361,
+      "learning_rate": 3.5764438993215298e-06,
+      "loss": 0.3422,
       "step": 115580
     },
     {
-      "epoch": 4.07,
-      "learning_rate": 4.444904043179127e-06,
-      "loss": 0.2608,
+      "epoch": 4.165675568529931,
+      "grad_norm": 0.2635994255542755,
+      "learning_rate": 3.5749400028653927e-06,
+      "loss": 0.3966,
       "step": 115585
     },
     {
-      "epoch": 4.07,
-      "learning_rate": 4.443282703956467e-06,
-      "loss": 0.2618,
+      "epoch": 4.165855768191156,
+      "grad_norm": 0.2824068069458008,
+      "learning_rate": 3.5734363983215448e-06,
+      "loss": 0.36,
       "step": 115590
     },
     {
-      "epoch": 4.07,
-      "learning_rate": 4.4416616316492326e-06,
-      "loss": 0.2595,
+      "epoch": 4.16603596785238,
+      "grad_norm": 0.3001638650894165,
+      "learning_rate": 3.571933085710483e-06,
+      "loss": 0.3725,
       "step": 115595
     },
     {
-      "epoch": 4.07,
-      "learning_rate": 4.440040826278468e-06,
-      "loss": 0.2439,
+      "epoch": 4.166216167513605,
+      "grad_norm": 0.27790218591690063,
+      "learning_rate": 3.5704300650526823e-06,
+      "loss": 0.3894,
       "step": 115600
     },
     {
-      "epoch": 4.07,
-      "learning_rate": 4.438420287865231e-06,
-      "loss": 0.2538,
+      "epoch": 4.16639636717483,
+      "grad_norm": 0.2569793462753296,
+      "learning_rate": 3.568927336368627e-06,
+      "loss": 0.3765,
       "step": 115605
     },
     {
-      "epoch": 4.07,
-      "learning_rate": 4.436800016430548e-06,
-      "loss": 0.2538,
+      "epoch": 4.166576566836055,
+      "grad_norm": 0.25097736716270447,
+      "learning_rate": 3.567424899678784e-06,
+      "loss": 0.3612,
       "step": 115610
     },
     {
-      "epoch": 4.07,
-      "learning_rate": 4.435180011995471e-06,
-      "loss": 0.2486,
+      "epoch": 4.166756766497279,
+      "grad_norm": 0.2786616086959839,
+      "learning_rate": 3.5659227550036284e-06,
+      "loss": 0.3388,
       "step": 115615
     },
     {
-      "epoch": 4.07,
-      "learning_rate": 4.433560274581031e-06,
-      "loss": 0.2631,
+      "epoch": 4.166936966158504,
+      "grad_norm": 0.22307883203029633,
+      "learning_rate": 3.5644209023636214e-06,
+      "loss": 0.3804,
       "step": 115620
     },
     {
-      "epoch": 4.07,
-      "learning_rate": 4.431940804208248e-06,
-      "loss": 0.227,
+      "epoch": 4.167117165819728,
+      "grad_norm": 0.27132895588874817,
+      "learning_rate": 3.562919341779239e-06,
+      "loss": 0.3481,
       "step": 115625
     },
     {
-      "epoch": 4.07,
-      "learning_rate": 4.4303216008981685e-06,
-      "loss": 0.2532,
+      "epoch": 4.167297365480953,
+      "grad_norm": 0.2521323263645172,
+      "learning_rate": 3.5614180732709194e-06,
+      "loss": 0.409,
       "step": 115630
     },
     {
-      "epoch": 4.07,
-      "learning_rate": 4.428702664671805e-06,
-      "loss": 0.2379,
+      "epoch": 4.167477565142177,
+      "grad_norm": 0.2317046821117401,
+      "learning_rate": 3.5599170968591334e-06,
+      "loss": 0.3321,
       "step": 115635
     },
     {
-      "epoch": 4.07,
-      "learning_rate": 4.427083995550174e-06,
-      "loss": 0.2311,
+      "epoch": 4.167657764803402,
+      "grad_norm": 0.2945384979248047,
+      "learning_rate": 3.558416412564325e-06,
+      "loss": 0.347,
       "step": 115640
     },
     {
-      "epoch": 4.07,
-      "learning_rate": 4.425465593554304e-06,
-      "loss": 0.2583,
+      "epoch": 4.167837964464627,
+      "grad_norm": 0.26886409521102905,
+      "learning_rate": 3.5569160204069446e-06,
+      "loss": 0.391,
       "step": 115645
     },
     {
-      "epoch": 4.07,
-      "learning_rate": 4.423847458705207e-06,
-      "loss": 0.2367,
+      "epoch": 4.168018164125852,
+      "grad_norm": 0.25826171040534973,
+      "learning_rate": 3.555415920407429e-06,
+      "loss": 0.3669,
       "step": 115650
     },
     {
-      "epoch": 4.07,
-      "learning_rate": 4.422229591023891e-06,
-      "loss": 0.2599,
+      "epoch": 4.168198363787076,
+      "grad_norm": 0.3123614192008972,
+      "learning_rate": 3.5539161125862226e-06,
+      "loss": 0.3831,
       "step": 115655
     },
     {
-      "epoch": 4.07,
-      "learning_rate": 4.420611990531353e-06,
-      "loss": 0.2391,
+      "epoch": 4.168378563448301,
+      "grad_norm": 0.21039628982543945,
+      "learning_rate": 3.552416596963748e-06,
+      "loss": 0.3678,
       "step": 115660
     },
     {
-      "epoch": 4.07,
-      "learning_rate": 4.418994657248607e-06,
-      "loss": 0.2353,
+      "epoch": 4.168558763109526,
+      "grad_norm": 0.21605625748634338,
+      "learning_rate": 3.550917373560453e-06,
+      "loss": 0.3595,
       "step": 115665
     },
     {
-      "epoch": 4.07,
-      "learning_rate": 4.417377591196659e-06,
-      "loss": 0.2539,
+      "epoch": 4.16873896277075,
+      "grad_norm": 0.1955236792564392,
+      "learning_rate": 3.549418442396757e-06,
+      "loss": 0.3606,
       "step": 115670
     },
     {
-      "epoch": 4.07,
-      "learning_rate": 4.4157607923965006e-06,
-      "loss": 0.2533,
+      "epoch": 4.168919162431974,
+      "grad_norm": 0.2682759165763855,
+      "learning_rate": 3.54791980349308e-06,
+      "loss": 0.4046,
       "step": 115675
     },
     {
-      "epoch": 4.07,
-      "learning_rate": 4.414144260869113e-06,
-      "loss": 0.2263,
+      "epoch": 4.169099362093199,
+      "grad_norm": 0.26579999923706055,
+      "learning_rate": 3.546421456869842e-06,
+      "loss": 0.3668,
       "step": 115680
     },
     {
-      "epoch": 4.07,
-      "learning_rate": 4.412527996635504e-06,
-      "loss": 0.251,
+      "epoch": 4.169279561754424,
+      "grad_norm": 0.2342423051595688,
+      "learning_rate": 3.5449234025474536e-06,
+      "loss": 0.3764,
       "step": 115685
     },
     {
-      "epoch": 4.07,
-      "learning_rate": 4.41091199971665e-06,
-      "loss": 0.2435,
+      "epoch": 4.169459761415649,
+      "grad_norm": 0.24451355636119843,
+      "learning_rate": 3.543425640546333e-06,
+      "loss": 0.4006,
       "step": 115690
     },
     {
-      "epoch": 4.07,
-      "learning_rate": 4.409296270133537e-06,
-      "loss": 0.2552,
+      "epoch": 4.169639961076873,
+      "grad_norm": 0.26802754402160645,
+      "learning_rate": 3.5419281708868933e-06,
+      "loss": 0.3725,
       "step": 115695
     },
     {
-      "epoch": 4.07,
-      "learning_rate": 4.407680807907133e-06,
-      "loss": 0.2533,
+      "epoch": 4.169820160738098,
+      "grad_norm": 0.21809948980808258,
+      "learning_rate": 3.5404309935895134e-06,
+      "loss": 0.3756,
       "step": 115700
     },
     {
-      "epoch": 4.07,
-      "learning_rate": 4.40606561305843e-06,
-      "loss": 0.2538,
+      "epoch": 4.170000360399323,
+      "grad_norm": 0.23009651899337769,
+      "learning_rate": 3.5389341086746158e-06,
+      "loss": 0.3687,
       "step": 115705
     },
     {
-      "epoch": 4.07,
-      "learning_rate": 4.404450685608394e-06,
-      "loss": 0.2393,
+      "epoch": 4.1701805600605475,
+      "grad_norm": 0.2777594327926636,
+      "learning_rate": 3.537437516162578e-06,
+      "loss": 0.3974,
       "step": 115710
     },
     {
-      "epoch": 4.07,
-      "learning_rate": 4.402836025577986e-06,
-      "loss": 0.2787,
+      "epoch": 4.170360759721771,
+      "grad_norm": 0.22270822525024414,
+      "learning_rate": 3.5359412160738044e-06,
+      "loss": 0.3573,
       "step": 115715
     },
     {
-      "epoch": 4.07,
-      "learning_rate": 4.401221632988178e-06,
-      "loss": 0.2305,
+      "epoch": 4.170540959382996,
+      "grad_norm": 0.2586876451969147,
+      "learning_rate": 3.5344452084286826e-06,
+      "loss": 0.3584,
       "step": 115720
     },
     {
-      "epoch": 4.07,
-      "learning_rate": 4.3996075078599405e-06,
-      "loss": 0.2592,
+      "epoch": 4.170721159044221,
+      "grad_norm": 0.2567586600780487,
+      "learning_rate": 3.5329494932475754e-06,
+      "loss": 0.3843,
       "step": 115725
     },
     {
-      "epoch": 4.07,
-      "learning_rate": 4.3979936502142186e-06,
-      "loss": 0.2396,
+      "epoch": 4.1709013587054455,
+      "grad_norm": 0.24362897872924805,
+      "learning_rate": 3.5314540705508823e-06,
+      "loss": 0.3739,
       "step": 115730
     },
     {
-      "epoch": 4.07,
-      "learning_rate": 4.396380060071975e-06,
-      "loss": 0.2475,
+      "epoch": 4.17108155836667,
+      "grad_norm": 0.3008570373058319,
+      "learning_rate": 3.529958940358971e-06,
+      "loss": 0.3583,
       "step": 115735
     },
     {
-      "epoch": 4.07,
-      "learning_rate": 4.3947667374541534e-06,
-      "loss": 0.2531,
+      "epoch": 4.171261758027895,
+      "grad_norm": 0.272904634475708,
+      "learning_rate": 3.5284641026922137e-06,
+      "loss": 0.3998,
       "step": 115740
     },
     {
-      "epoch": 4.07,
-      "learning_rate": 4.3931536823817095e-06,
-      "loss": 0.2214,
+      "epoch": 4.17144195768912,
+      "grad_norm": 0.27428102493286133,
+      "learning_rate": 3.526969557570975e-06,
+      "loss": 0.3883,
       "step": 115745
     },
     {
-      "epoch": 4.07,
-      "learning_rate": 4.391540894875587e-06,
-      "loss": 0.2422,
+      "epoch": 4.1716221573503445,
+      "grad_norm": 0.25134575366973877,
+      "learning_rate": 3.525475305015613e-06,
+      "loss": 0.3813,
       "step": 115750
     },
     {
-      "epoch": 4.07,
-      "learning_rate": 4.389928374956717e-06,
-      "loss": 0.2586,
+      "epoch": 4.171802357011569,
+      "grad_norm": 0.2150566279888153,
+      "learning_rate": 3.5239813450464983e-06,
+      "loss": 0.3792,
       "step": 115755
     },
     {
-      "epoch": 4.07,
-      "learning_rate": 4.38831612264605e-06,
-      "loss": 0.2234,
+      "epoch": 4.171982556672793,
+      "grad_norm": 0.2204793393611908,
+      "learning_rate": 3.5224876776839806e-06,
+      "loss": 0.3873,
       "step": 115760
     },
     {
-      "epoch": 4.07,
-      "learning_rate": 4.386704137964517e-06,
-      "loss": 0.261,
+      "epoch": 4.172162756334018,
+      "grad_norm": 0.2468349039554596,
+      "learning_rate": 3.5209943029484077e-06,
+      "loss": 0.4102,
       "step": 115765
     },
     {
-      "epoch": 4.07,
-      "learning_rate": 4.385092420933037e-06,
-      "loss": 0.2488,
+      "epoch": 4.1723429559952425,
+      "grad_norm": 0.2205950766801834,
+      "learning_rate": 3.5195012208601303e-06,
+      "loss": 0.3288,
       "step": 115770
     },
     {
-      "epoch": 4.07,
-      "learning_rate": 4.383480971572549e-06,
-      "loss": 0.2711,
+      "epoch": 4.172523155656467,
+      "grad_norm": 0.28197482228279114,
+      "learning_rate": 3.5180084314394845e-06,
+      "loss": 0.3871,
       "step": 115775
     },
     {
-      "epoch": 4.07,
-      "learning_rate": 4.381869789903978e-06,
-      "loss": 0.2592,
+      "epoch": 4.172703355317692,
+      "grad_norm": 0.28598877787590027,
+      "learning_rate": 3.516515934706821e-06,
+      "loss": 0.379,
       "step": 115780
     },
     {
-      "epoch": 4.07,
-      "learning_rate": 4.38025887594824e-06,
-      "loss": 0.2521,
+      "epoch": 4.172883554978917,
+      "grad_norm": 0.22734612226486206,
+      "learning_rate": 3.515023730682468e-06,
+      "loss": 0.3956,
       "step": 115785
     },
     {
-      "epoch": 4.07,
-      "learning_rate": 4.378648229726251e-06,
-      "loss": 0.2311,
+      "epoch": 4.1730637546401415,
+      "grad_norm": 0.2860606610774994,
+      "learning_rate": 3.5135318193867563e-06,
+      "loss": 0.4039,
       "step": 115790
     },
     {
-      "epoch": 4.07,
-      "learning_rate": 4.377037851258919e-06,
-      "loss": 0.2464,
+      "epoch": 4.173243954301366,
+      "grad_norm": 0.263313889503479,
+      "learning_rate": 3.5120402008400167e-06,
+      "loss": 0.3714,
       "step": 115795
     },
     {
-      "epoch": 4.07,
-      "learning_rate": 4.375427740567167e-06,
-      "loss": 0.2263,
+      "epoch": 4.173424153962591,
+      "grad_norm": 0.28966277837753296,
+      "learning_rate": 3.5105488750625666e-06,
+      "loss": 0.3525,
       "step": 115800
     },
     {
-      "epoch": 4.07,
-      "learning_rate": 4.373817897671895e-06,
-      "loss": 0.2361,
+      "epoch": 4.173604353623815,
+      "grad_norm": 0.27305835485458374,
+      "learning_rate": 3.50905784207472e-06,
+      "loss": 0.361,
       "step": 115805
     },
     {
-      "epoch": 4.07,
-      "learning_rate": 4.372208322594001e-06,
-      "loss": 0.2518,
+      "epoch": 4.1737845532850395,
+      "grad_norm": 0.28714773058891296,
+      "learning_rate": 3.5075671018968136e-06,
+      "loss": 0.3572,
       "step": 115810
     },
     {
-      "epoch": 4.07,
-      "learning_rate": 4.370599015354382e-06,
-      "loss": 0.2433,
+      "epoch": 4.173964752946264,
+      "grad_norm": 0.23891617357730865,
+      "learning_rate": 3.506076654549134e-06,
+      "loss": 0.3833,
       "step": 115815
     },
     {
-      "epoch": 4.07,
-      "learning_rate": 4.368989975973947e-06,
-      "loss": 0.2632,
+      "epoch": 4.174144952607489,
+      "grad_norm": 0.2725772261619568,
+      "learning_rate": 3.504586500052001e-06,
+      "loss": 0.3409,
       "step": 115820
     },
     {
-      "epoch": 4.08,
-      "learning_rate": 4.367381204473575e-06,
-      "loss": 0.2546,
+      "epoch": 4.174325152268714,
+      "grad_norm": 0.2952682375907898,
+      "learning_rate": 3.503096638425718e-06,
+      "loss": 0.3602,
       "step": 115825
     },
     {
-      "epoch": 4.08,
-      "learning_rate": 4.365772700874165e-06,
-      "loss": 0.2687,
+      "epoch": 4.174505351929938,
+      "grad_norm": 0.2583956718444824,
+      "learning_rate": 3.5016070696905794e-06,
+      "loss": 0.3786,
       "step": 115830
     },
     {
-      "epoch": 4.08,
-      "learning_rate": 4.364164465196594e-06,
-      "loss": 0.241,
+      "epoch": 4.174685551591163,
+      "grad_norm": 0.2635524868965149,
+      "learning_rate": 3.5001177938668834e-06,
+      "loss": 0.3726,
       "step": 115835
     },
     {
-      "epoch": 4.08,
-      "learning_rate": 4.36255649746175e-06,
-      "loss": 0.2409,
+      "epoch": 4.174865751252388,
+      "grad_norm": 0.3067127764225006,
+      "learning_rate": 3.4986288109749217e-06,
+      "loss": 0.41,
       "step": 115840
     },
     {
-      "epoch": 4.08,
-      "learning_rate": 4.360948797690515e-06,
-      "loss": 0.2332,
+      "epoch": 4.175045950913613,
+      "grad_norm": 0.2670503556728363,
+      "learning_rate": 3.4971401210349724e-06,
+      "loss": 0.3842,
       "step": 115845
     },
     {
-      "epoch": 4.08,
-      "learning_rate": 4.3593413659037544e-06,
-      "loss": 0.2577,
+      "epoch": 4.1752261505748365,
+      "grad_norm": 0.2181471437215805,
+      "learning_rate": 3.4956517240673307e-06,
+      "loss": 0.4022,
       "step": 115850
     },
     {
-      "epoch": 4.08,
-      "learning_rate": 4.357734202122338e-06,
-      "loss": 0.2678,
+      "epoch": 4.175406350236061,
+      "grad_norm": 0.21090255677700043,
+      "learning_rate": 3.4941636200922695e-06,
+      "loss": 0.3688,
       "step": 115855
     },
     {
-      "epoch": 4.08,
-      "learning_rate": 4.356127306367147e-06,
-      "loss": 0.257,
+      "epoch": 4.175586549897286,
+      "grad_norm": 0.2528001368045807,
+      "learning_rate": 3.4926758091300694e-06,
+      "loss": 0.3877,
       "step": 115860
     },
     {
-      "epoch": 4.08,
-      "learning_rate": 4.354520678659035e-06,
-      "loss": 0.2372,
+      "epoch": 4.175766749558511,
+      "grad_norm": 0.2146507054567337,
+      "learning_rate": 3.4911882912009947e-06,
+      "loss": 0.392,
       "step": 115865
     },
     {
-      "epoch": 4.08,
-      "learning_rate": 4.352914319018864e-06,
-      "loss": 0.2447,
+      "epoch": 4.175946949219735,
+      "grad_norm": 0.2687932848930359,
+      "learning_rate": 3.48970106632531e-06,
+      "loss": 0.3346,
       "step": 115870
     },
     {
-      "epoch": 4.08,
-      "learning_rate": 4.351308227467496e-06,
-      "loss": 0.2486,
+      "epoch": 4.17612714888096,
+      "grad_norm": 0.33123210072517395,
+      "learning_rate": 3.4882141345232903e-06,
+      "loss": 0.3685,
       "step": 115875
     },
     {
-      "epoch": 4.08,
-      "learning_rate": 4.349702404025777e-06,
-      "loss": 0.2786,
+      "epoch": 4.176307348542185,
+      "grad_norm": 0.2785130739212036,
+      "learning_rate": 3.486727495815187e-06,
+      "loss": 0.3529,
       "step": 115880
     },
     {
-      "epoch": 4.08,
-      "learning_rate": 4.348096848714573e-06,
-      "loss": 0.2499,
+      "epoch": 4.17648754820341,
+      "grad_norm": 0.31324052810668945,
+      "learning_rate": 3.485241150221258e-06,
+      "loss": 0.3905,
       "step": 115885
     },
     {
-      "epoch": 4.08,
-      "learning_rate": 4.346491561554711e-06,
-      "loss": 0.252,
+      "epoch": 4.176667747864634,
+      "grad_norm": 0.23870332539081573,
+      "learning_rate": 3.483755097761751e-06,
+      "loss": 0.3257,
       "step": 115890
     },
     {
-      "epoch": 4.08,
-      "learning_rate": 4.344886542567053e-06,
-      "loss": 0.264,
+      "epoch": 4.176847947525859,
+      "grad_norm": 0.296524316072464,
+      "learning_rate": 3.4822693384569087e-06,
+      "loss": 0.3858,
       "step": 115895
     },
     {
-      "epoch": 4.08,
-      "learning_rate": 4.34328179177243e-06,
-      "loss": 0.2619,
+      "epoch": 4.177028147187083,
+      "grad_norm": 0.22270824015140533,
+      "learning_rate": 3.4807838723269896e-06,
+      "loss": 0.3931,
       "step": 115900
     },
     {
-      "epoch": 4.08,
-      "learning_rate": 4.34167730919168e-06,
-      "loss": 0.2441,
+      "epoch": 4.177208346848308,
+      "grad_norm": 0.25900474190711975,
+      "learning_rate": 3.479298699392228e-06,
+      "loss": 0.3714,
       "step": 115905
     },
     {
-      "epoch": 4.08,
-      "learning_rate": 4.340073094845631e-06,
-      "loss": 0.2411,
+      "epoch": 4.177388546509532,
+      "grad_norm": 0.28385141491889954,
+      "learning_rate": 3.4778138196728425e-06,
+      "loss": 0.3734,
       "step": 115910
     },
     {
-      "epoch": 4.08,
-      "learning_rate": 4.338469148755123e-06,
-      "loss": 0.2495,
+      "epoch": 4.177568746170757,
+      "grad_norm": 0.2589219808578491,
+      "learning_rate": 3.476329233189085e-06,
+      "loss": 0.3638,
       "step": 115915
     },
     {
-      "epoch": 4.08,
-      "learning_rate": 4.336865470940976e-06,
-      "loss": 0.2487,
+      "epoch": 4.177748945831982,
+      "grad_norm": 0.23942036926746368,
+      "learning_rate": 3.4748449399611747e-06,
+      "loss": 0.3584,
       "step": 115920
     },
     {
-      "epoch": 4.08,
-      "learning_rate": 4.335262061424011e-06,
-      "loss": 0.2509,
+      "epoch": 4.177929145493207,
+      "grad_norm": 0.2529667317867279,
+      "learning_rate": 3.473360940009332e-06,
+      "loss": 0.3587,
       "step": 115925
     },
     {
-      "epoch": 4.08,
-      "learning_rate": 4.333658920225048e-06,
-      "loss": 0.2359,
+      "epoch": 4.178109345154431,
+      "grad_norm": 0.22213000059127808,
+      "learning_rate": 3.471877233353782e-06,
+      "loss": 0.3889,
       "step": 115930
     },
     {
-      "epoch": 4.08,
-      "learning_rate": 4.332056047364908e-06,
-      "loss": 0.263,
+      "epoch": 4.178289544815656,
+      "grad_norm": 0.2727386951446533,
+      "learning_rate": 3.470393820014728e-06,
+      "loss": 0.4031,
       "step": 115935
     },
     {
-      "epoch": 4.08,
-      "learning_rate": 4.3304534428643935e-06,
-      "loss": 0.2865,
+      "epoch": 4.178469744476881,
+      "grad_norm": 0.20552094280719757,
+      "learning_rate": 3.4689107000123982e-06,
+      "loss": 0.3649,
       "step": 115940
     },
     {
-      "epoch": 4.08,
-      "learning_rate": 4.328851106744325e-06,
-      "loss": 0.2596,
+      "epoch": 4.178649944138105,
+      "grad_norm": 0.25130918622016907,
+      "learning_rate": 3.467427873366988e-06,
+      "loss": 0.3563,
       "step": 115945
     },
     {
-      "epoch": 4.08,
-      "learning_rate": 4.327249039025497e-06,
-      "loss": 0.2376,
+      "epoch": 4.178830143799329,
+      "grad_norm": 0.2697368562221527,
+      "learning_rate": 3.4659453400987084e-06,
+      "loss": 0.388,
       "step": 115950
     },
     {
-      "epoch": 4.08,
-      "learning_rate": 4.3256472397287224e-06,
-      "loss": 0.247,
+      "epoch": 4.179010343460554,
+      "grad_norm": 0.2934541404247284,
+      "learning_rate": 3.4644631002277523e-06,
+      "loss": 0.3953,
       "step": 115955
     },
     {
-      "epoch": 4.08,
-      "learning_rate": 4.324045708874793e-06,
-      "loss": 0.252,
+      "epoch": 4.179190543121779,
+      "grad_norm": 0.20682430267333984,
+      "learning_rate": 3.462981153774311e-06,
+      "loss": 0.3416,
       "step": 115960
     },
     {
-      "epoch": 4.08,
-      "learning_rate": 4.322444446484505e-06,
-      "loss": 0.2502,
+      "epoch": 4.179370742783004,
+      "grad_norm": 0.22139784693717957,
+      "learning_rate": 3.461499500758589e-06,
+      "loss": 0.3593,
       "step": 115965
     },
     {
-      "epoch": 4.08,
-      "learning_rate": 4.32084345257864e-06,
-      "loss": 0.2715,
+      "epoch": 4.179550942444228,
+      "grad_norm": 0.26105570793151855,
+      "learning_rate": 3.4600181412007633e-06,
+      "loss": 0.3664,
       "step": 115970
     },
     {
-      "epoch": 4.08,
-      "learning_rate": 4.319242727178005e-06,
-      "loss": 0.2224,
+      "epoch": 4.179731142105453,
+      "grad_norm": 0.25991979241371155,
+      "learning_rate": 3.458537075121024e-06,
+      "loss": 0.394,
       "step": 115975
     },
     {
-      "epoch": 4.08,
-      "learning_rate": 4.317642270303371e-06,
-      "loss": 0.2457,
+      "epoch": 4.179911341766678,
+      "grad_norm": 0.24015338718891144,
+      "learning_rate": 3.4570563025395435e-06,
+      "loss": 0.3873,
       "step": 115980
     },
     {
-      "epoch": 4.08,
-      "learning_rate": 4.316042081975519e-06,
-      "loss": 0.2356,
+      "epoch": 4.1800915414279025,
+      "grad_norm": 0.24884821474552155,
+      "learning_rate": 3.4555758234765006e-06,
+      "loss": 0.3393,
       "step": 115985
     },
     {
-      "epoch": 4.08,
-      "learning_rate": 4.314442162215232e-06,
-      "loss": 0.2499,
+      "epoch": 4.180271741089126,
+      "grad_norm": 0.211832195520401,
+      "learning_rate": 3.4540956379520623e-06,
+      "loss": 0.3627,
       "step": 115990
     },
     {
-      "epoch": 4.08,
-      "learning_rate": 4.312842511043274e-06,
-      "loss": 0.2539,
+      "epoch": 4.180451940750351,
+      "grad_norm": 0.28997382521629333,
+      "learning_rate": 3.4526157459864093e-06,
+      "loss": 0.3957,
       "step": 115995
     },
     {
-      "epoch": 4.08,
-      "learning_rate": 4.311243128480433e-06,
-      "loss": 0.2663,
+      "epoch": 4.180632140411576,
+      "grad_norm": 0.23022478818893433,
+      "learning_rate": 3.4511361475996843e-06,
+      "loss": 0.3877,
       "step": 116000
     },
     {
-      "epoch": 4.08,
-      "eval_loss": 0.24984586238861084,
-      "eval_runtime": 10.5713,
-      "eval_samples_per_second": 9.46,
-      "eval_steps_per_second": 9.46,
+      "epoch": 4.180632140411576,
+      "eval_loss": 0.42928194999694824,
+      "eval_runtime": 3.5376,
+      "eval_samples_per_second": 28.268,
+      "eval_steps_per_second": 7.067,
       "step": 116000
     },
     {
-      "epoch": 4.08,
-      "learning_rate": 4.309644014547457e-06,
-      "loss": 0.2392,
+      "epoch": 4.180812340072801,
+      "grad_norm": 0.23664060235023499,
+      "learning_rate": 3.449656842812063e-06,
+      "loss": 0.3887,
       "step": 116005
     },
     {
-      "epoch": 4.08,
-      "learning_rate": 4.308045169265126e-06,
-      "loss": 0.2409,
+      "epoch": 4.180992539734025,
+      "grad_norm": 0.31014373898506165,
+      "learning_rate": 3.4481778316436925e-06,
+      "loss": 0.3901,
       "step": 116010
     },
     {
-      "epoch": 4.08,
-      "learning_rate": 4.306446592654189e-06,
-      "loss": 0.2452,
+      "epoch": 4.18117273939525,
+      "grad_norm": 0.26111266016960144,
+      "learning_rate": 3.4466991141147238e-06,
+      "loss": 0.4023,
       "step": 116015
     },
     {
-      "epoch": 4.08,
-      "learning_rate": 4.304848284735408e-06,
-      "loss": 0.2598,
+      "epoch": 4.181352939056475,
+      "grad_norm": 0.20176957547664642,
+      "learning_rate": 3.445220690245318e-06,
+      "loss": 0.3669,
       "step": 116020
     },
     {
-      "epoch": 4.08,
-      "learning_rate": 4.3032502455295275e-06,
-      "loss": 0.2368,
+      "epoch": 4.1815331387176995,
+      "grad_norm": 0.23669221997261047,
+      "learning_rate": 3.4437425600555988e-06,
+      "loss": 0.3726,
       "step": 116025
     },
     {
-      "epoch": 4.08,
-      "learning_rate": 4.301652475057308e-06,
-      "loss": 0.246,
+      "epoch": 4.181713338378924,
+      "grad_norm": 0.22191861271858215,
+      "learning_rate": 3.4422647235657105e-06,
+      "loss": 0.3698,
       "step": 116030
     },
     {
-      "epoch": 4.08,
-      "learning_rate": 4.3000549733394905e-06,
-      "loss": 0.2407,
+      "epoch": 4.181893538040148,
+      "grad_norm": 0.24171018600463867,
+      "learning_rate": 3.4407871807957957e-06,
+      "loss": 0.3699,
       "step": 116035
     },
     {
-      "epoch": 4.08,
-      "learning_rate": 4.29845774039681e-06,
-      "loss": 0.2708,
+      "epoch": 4.182073737701373,
+      "grad_norm": 0.22796885669231415,
+      "learning_rate": 3.439309931765983e-06,
+      "loss": 0.3784,
       "step": 116040
     },
     {
-      "epoch": 4.08,
-      "learning_rate": 4.29686077625002e-06,
-      "loss": 0.2366,
+      "epoch": 4.1822539373625975,
+      "grad_norm": 0.344339519739151,
+      "learning_rate": 3.4378329764963995e-06,
+      "loss": 0.3725,
       "step": 116045
     },
     {
-      "epoch": 4.08,
-      "learning_rate": 4.295264080919842e-06,
-      "loss": 0.2562,
+      "epoch": 4.182434137023822,
+      "grad_norm": 0.2166433185338974,
+      "learning_rate": 3.4363563150071636e-06,
+      "loss": 0.3542,
       "step": 116050
     },
     {
-      "epoch": 4.08,
-      "learning_rate": 4.293667654427022e-06,
-      "loss": 0.2507,
+      "epoch": 4.182614336685047,
+      "grad_norm": 0.2705228924751282,
+      "learning_rate": 3.434879947318395e-06,
+      "loss": 0.3529,
       "step": 116055
     },
     {
-      "epoch": 4.08,
-      "learning_rate": 4.292071496792282e-06,
-      "loss": 0.2261,
+      "epoch": 4.182794536346272,
+      "grad_norm": 0.2719237506389618,
+      "learning_rate": 3.4334038734502166e-06,
+      "loss": 0.3851,
       "step": 116060
     },
     {
-      "epoch": 4.08,
-      "learning_rate": 4.29047560803634e-06,
-      "loss": 0.2649,
+      "epoch": 4.1829747360074965,
+      "grad_norm": 0.2510291635990143,
+      "learning_rate": 3.431928093422737e-06,
+      "loss": 0.3846,
       "step": 116065
     },
     {
-      "epoch": 4.08,
-      "learning_rate": 4.2888799881799305e-06,
-      "loss": 0.2358,
+      "epoch": 4.183154935668721,
+      "grad_norm": 0.2954877018928528,
+      "learning_rate": 3.43045260725606e-06,
+      "loss": 0.3558,
       "step": 116070
     },
     {
-      "epoch": 4.08,
-      "learning_rate": 4.287284637243763e-06,
-      "loss": 0.2658,
+      "epoch": 4.183335135329946,
+      "grad_norm": 0.2507972717285156,
+      "learning_rate": 3.4289774149702887e-06,
+      "loss": 0.3594,
       "step": 116075
     },
     {
-      "epoch": 4.08,
-      "learning_rate": 4.2856895552485565e-06,
-      "loss": 0.2595,
+      "epoch": 4.18351533499117,
+      "grad_norm": 0.24110138416290283,
+      "learning_rate": 3.427502516585521e-06,
+      "loss": 0.3744,
       "step": 116080
     },
     {
-      "epoch": 4.08,
-      "learning_rate": 4.284094742215014e-06,
-      "loss": 0.2434,
+      "epoch": 4.1836955346523945,
+      "grad_norm": 0.25332164764404297,
+      "learning_rate": 3.426027912121857e-06,
+      "loss": 0.4066,
       "step": 116085
     },
     {
-      "epoch": 4.08,
-      "learning_rate": 4.282500198163855e-06,
-      "loss": 0.2826,
+      "epoch": 4.183875734313619,
+      "grad_norm": 0.22243022918701172,
+      "learning_rate": 3.424553601599395e-06,
+      "loss": 0.3625,
       "step": 116090
     },
     {
-      "epoch": 4.08,
-      "learning_rate": 4.280905923115777e-06,
-      "loss": 0.2464,
+      "epoch": 4.184055933974844,
+      "grad_norm": 0.23225410282611847,
+      "learning_rate": 3.4230795850381998e-06,
+      "loss": 0.3428,
       "step": 116095
     },
     {
-      "epoch": 4.08,
-      "learning_rate": 4.279311917091475e-06,
-      "loss": 0.2642,
+      "epoch": 4.184236133636069,
+      "grad_norm": 0.28310906887054443,
+      "learning_rate": 3.4216058624583743e-06,
+      "loss": 0.3719,
       "step": 116100
     },
     {
-      "epoch": 4.08,
-      "learning_rate": 4.277718180111653e-06,
-      "loss": 0.2515,
+      "epoch": 4.1844163332972935,
+      "grad_norm": 0.2678923010826111,
+      "learning_rate": 3.4201324338799884e-06,
+      "loss": 0.3407,
       "step": 116105
     },
     {
-      "epoch": 4.09,
-      "learning_rate": 4.276124712197011e-06,
-      "loss": 0.2432,
+      "epoch": 4.184596532958518,
+      "grad_norm": 0.2929144501686096,
+      "learning_rate": 3.418659299323124e-06,
+      "loss": 0.3569,
       "step": 116110
     },
     {
-      "epoch": 4.09,
-      "learning_rate": 4.2745315133682295e-06,
-      "loss": 0.2789,
+      "epoch": 4.184776732619743,
+      "grad_norm": 0.2045121192932129,
+      "learning_rate": 3.4171864588078474e-06,
+      "loss": 0.3586,
       "step": 116115
     },
     {
-      "epoch": 4.09,
-      "learning_rate": 4.2729385836459915e-06,
-      "loss": 0.2398,
+      "epoch": 4.184956932280968,
+      "grad_norm": 0.21932172775268555,
+      "learning_rate": 3.415713912354221e-06,
+      "loss": 0.3525,
       "step": 116120
     },
     {
-      "epoch": 4.09,
-      "learning_rate": 4.2713459230509955e-06,
-      "loss": 0.2717,
+      "epoch": 4.1851371319421915,
+      "grad_norm": 0.21695736050605774,
+      "learning_rate": 3.4142416599823175e-06,
+      "loss": 0.3583,
       "step": 116125
     },
     {
-      "epoch": 4.09,
-      "learning_rate": 4.269753531603909e-06,
-      "loss": 0.255,
+      "epoch": 4.185317331603416,
+      "grad_norm": 0.2124013900756836,
+      "learning_rate": 3.4127697017121928e-06,
+      "loss": 0.3525,
       "step": 116130
     },
     {
-      "epoch": 4.09,
-      "learning_rate": 4.268161409325413e-06,
-      "loss": 0.2511,
+      "epoch": 4.185497531264641,
+      "grad_norm": 0.2589491307735443,
+      "learning_rate": 3.4112980375639035e-06,
+      "loss": 0.3682,
       "step": 116135
     },
     {
-      "epoch": 4.09,
-      "learning_rate": 4.2665695562361715e-06,
-      "loss": 0.2584,
+      "epoch": 4.185677730925866,
+      "grad_norm": 0.2505156099796295,
+      "learning_rate": 3.4098266675574974e-06,
+      "loss": 0.3868,
       "step": 116140
     },
     {
-      "epoch": 4.09,
-      "learning_rate": 4.264977972356868e-06,
-      "loss": 0.249,
+      "epoch": 4.18585793058709,
+      "grad_norm": 0.2609826326370239,
+      "learning_rate": 3.4083555917130165e-06,
+      "loss": 0.3701,
       "step": 116145
     },
     {
-      "epoch": 4.09,
-      "learning_rate": 4.2633866577081576e-06,
-      "loss": 0.2773,
+      "epoch": 4.186038130248315,
+      "grad_norm": 0.2588742971420288,
+      "learning_rate": 3.4068848100505176e-06,
+      "loss": 0.3641,
       "step": 116150
     },
     {
-      "epoch": 4.09,
-      "learning_rate": 4.261795612310701e-06,
-      "loss": 0.2374,
+      "epoch": 4.18621832990954,
+      "grad_norm": 0.20735150575637817,
+      "learning_rate": 3.4054143225900335e-06,
+      "loss": 0.3315,
       "step": 116155
     },
     {
-      "epoch": 4.09,
-      "learning_rate": 4.260204836185161e-06,
-      "loss": 0.2514,
+      "epoch": 4.186398529570765,
+      "grad_norm": 0.23937851190567017,
+      "learning_rate": 3.4039441293515966e-06,
+      "loss": 0.3675,
       "step": 116160
     },
     {
-      "epoch": 4.09,
-      "learning_rate": 4.2586143293522005e-06,
-      "loss": 0.2469,
+      "epoch": 4.186578729231989,
+      "grad_norm": 0.24689562618732452,
+      "learning_rate": 3.4024742303552405e-06,
+      "loss": 0.3567,
       "step": 116165
     },
     {
-      "epoch": 4.09,
-      "learning_rate": 4.257024091832463e-06,
-      "loss": 0.2366,
+      "epoch": 4.186758928893214,
+      "grad_norm": 0.24008159339427948,
+      "learning_rate": 3.4010046256209906e-06,
+      "loss": 0.3822,
       "step": 116170
     },
     {
-      "epoch": 4.09,
-      "learning_rate": 4.255434123646598e-06,
-      "loss": 0.2403,
+      "epoch": 4.186939128554438,
+      "grad_norm": 0.1905672252178192,
+      "learning_rate": 3.3995353151688667e-06,
+      "loss": 0.3537,
       "step": 116175
     },
     {
-      "epoch": 4.09,
-      "learning_rate": 4.253844424815243e-06,
-      "loss": 0.2289,
+      "epoch": 4.187119328215663,
+      "grad_norm": 0.20124223828315735,
+      "learning_rate": 3.3980662990188956e-06,
+      "loss": 0.3373,
       "step": 116180
     },
     {
-      "epoch": 4.09,
-      "learning_rate": 4.252254995359051e-06,
-      "loss": 0.2692,
+      "epoch": 4.187299527876887,
+      "grad_norm": 0.29924681782722473,
+      "learning_rate": 3.396597577191091e-06,
+      "loss": 0.3908,
       "step": 116185
     },
     {
-      "epoch": 4.09,
-      "learning_rate": 4.250665835298656e-06,
-      "loss": 0.2502,
+      "epoch": 4.187479727538112,
+      "grad_norm": 0.2561193108558655,
+      "learning_rate": 3.3951291497054616e-06,
+      "loss": 0.364,
       "step": 116190
     },
     {
-      "epoch": 4.09,
-      "learning_rate": 4.249076944654692e-06,
-      "loss": 0.246,
+      "epoch": 4.187659927199337,
+      "grad_norm": 0.23734471201896667,
+      "learning_rate": 3.3936610165820144e-06,
+      "loss": 0.4054,
       "step": 116195
     },
     {
-      "epoch": 4.09,
-      "learning_rate": 4.247488323447779e-06,
-      "loss": 0.2582,
+      "epoch": 4.187840126860562,
+      "grad_norm": 0.21448448300361633,
+      "learning_rate": 3.3921931778407467e-06,
+      "loss": 0.3877,
       "step": 116200
     },
     {
-      "epoch": 4.09,
-      "learning_rate": 4.245899971698564e-06,
-      "loss": 0.2559,
+      "epoch": 4.188020326521786,
+      "grad_norm": 0.25979727506637573,
+      "learning_rate": 3.390725633501676e-06,
+      "loss": 0.3934,
       "step": 116205
     },
     {
-      "epoch": 4.09,
-      "learning_rate": 4.244311889427652e-06,
-      "loss": 0.2409,
+      "epoch": 4.188200526183011,
+      "grad_norm": 0.25088486075401306,
+      "learning_rate": 3.3892583835847786e-06,
+      "loss": 0.3796,
       "step": 116210
     },
     {
-      "epoch": 4.09,
-      "learning_rate": 4.242724076655671e-06,
-      "loss": 0.2466,
+      "epoch": 4.188380725844236,
+      "grad_norm": 0.24051101505756378,
+      "learning_rate": 3.387791428110046e-06,
+      "loss": 0.3435,
       "step": 116215
     },
     {
-      "epoch": 4.09,
-      "learning_rate": 4.241136533403245e-06,
-      "loss": 0.2795,
+      "epoch": 4.18856092550546,
+      "grad_norm": 0.265553742647171,
+      "learning_rate": 3.3863247670974764e-06,
+      "loss": 0.3918,
       "step": 116220
     },
     {
-      "epoch": 4.09,
-      "learning_rate": 4.2395492596909806e-06,
-      "loss": 0.2432,
+      "epoch": 4.188741125166684,
+      "grad_norm": 0.22547340393066406,
+      "learning_rate": 3.384858400567045e-06,
+      "loss": 0.3579,
       "step": 116225
     },
     {
-      "epoch": 4.09,
-      "learning_rate": 4.237962255539488e-06,
-      "loss": 0.2446,
+      "epoch": 4.188921324827909,
+      "grad_norm": 0.2896125614643097,
+      "learning_rate": 3.383392328538737e-06,
+      "loss": 0.3668,
       "step": 116230
     },
     {
-      "epoch": 4.09,
-      "learning_rate": 4.236375520969366e-06,
-      "loss": 0.2807,
+      "epoch": 4.189101524489134,
+      "grad_norm": 0.24138803780078888,
+      "learning_rate": 3.3819265510325214e-06,
+      "loss": 0.3922,
       "step": 116235
     },
     {
-      "epoch": 4.09,
-      "learning_rate": 4.234789056001231e-06,
-      "loss": 0.2676,
+      "epoch": 4.189281724150359,
+      "grad_norm": 0.29337865114212036,
+      "learning_rate": 3.3804610680683665e-06,
+      "loss": 0.4072,
       "step": 116240
     },
     {
-      "epoch": 4.09,
-      "learning_rate": 4.233202860655675e-06,
-      "loss": 0.2475,
+      "epoch": 4.189461923811583,
+      "grad_norm": 0.29773908853530884,
+      "learning_rate": 3.3789958796662496e-06,
+      "loss": 0.3651,
       "step": 116245
     },
     {
-      "epoch": 4.09,
-      "learning_rate": 4.231616934953292e-06,
-      "loss": 0.2703,
+      "epoch": 4.189642123472808,
+      "grad_norm": 0.25638502836227417,
+      "learning_rate": 3.3775309858461253e-06,
+      "loss": 0.3669,
       "step": 116250
     },
     {
-      "epoch": 4.09,
-      "learning_rate": 4.230031278914673e-06,
-      "loss": 0.2574,
+      "epoch": 4.189822323134033,
+      "grad_norm": 0.23928916454315186,
+      "learning_rate": 3.376066386627957e-06,
+      "loss": 0.3569,
       "step": 116255
     },
     {
-      "epoch": 4.09,
-      "learning_rate": 4.228445892560415e-06,
-      "loss": 0.2597,
+      "epoch": 4.1900025227952575,
+      "grad_norm": 0.24761387705802917,
+      "learning_rate": 3.3746020820316972e-06,
+      "loss": 0.4,
       "step": 116260
     },
     {
-      "epoch": 4.09,
-      "learning_rate": 4.226860775911093e-06,
-      "loss": 0.2352,
+      "epoch": 4.190182722456481,
+      "grad_norm": 0.2518233358860016,
+      "learning_rate": 3.373138072077292e-06,
+      "loss": 0.3886,
       "step": 116265
     },
     {
-      "epoch": 4.09,
-      "learning_rate": 4.225592876793025e-06,
-      "loss": 0.2428,
+      "epoch": 4.190362922117706,
+      "grad_norm": 0.2131795734167099,
+      "learning_rate": 3.3716743567847015e-06,
+      "loss": 0.3559,
       "step": 116270
     },
     {
-      "epoch": 4.09,
-      "learning_rate": 4.224008245664465e-06,
-      "loss": 0.2457,
+      "epoch": 4.190543121778931,
+      "grad_norm": 0.27569085359573364,
+      "learning_rate": 3.370210936173862e-06,
+      "loss": 0.3568,
       "step": 116275
     },
     {
-      "epoch": 4.09,
-      "learning_rate": 4.222423884298468e-06,
-      "loss": 0.2538,
+      "epoch": 4.190723321440156,
+      "grad_norm": 0.24571947753429413,
+      "learning_rate": 3.3687478102647025e-06,
+      "loss": 0.3836,
       "step": 116280
     },
     {
-      "epoch": 4.09,
-      "learning_rate": 4.220839792715597e-06,
-      "loss": 0.2322,
+      "epoch": 4.19090352110138,
+      "grad_norm": 0.25763317942619324,
+      "learning_rate": 3.3672849790771704e-06,
+      "loss": 0.3327,
       "step": 116285
     },
     {
-      "epoch": 4.09,
-      "learning_rate": 4.219255970936431e-06,
-      "loss": 0.2299,
+      "epoch": 4.191083720762605,
+      "grad_norm": 0.21428771317005157,
+      "learning_rate": 3.365822442631192e-06,
+      "loss": 0.3846,
       "step": 116290
     },
     {
-      "epoch": 4.09,
-      "learning_rate": 4.217672418981533e-06,
-      "loss": 0.2667,
+      "epoch": 4.19126392042383,
+      "grad_norm": 0.24545690417289734,
+      "learning_rate": 3.3643602009466897e-06,
+      "loss": 0.3348,
       "step": 116295
     },
     {
-      "epoch": 4.09,
-      "learning_rate": 4.216089136871457e-06,
-      "loss": 0.2434,
+      "epoch": 4.1914441200850545,
+      "grad_norm": 0.23609334230422974,
+      "learning_rate": 3.3628982540436004e-06,
+      "loss": 0.3521,
       "step": 116300
     },
     {
-      "epoch": 4.09,
-      "learning_rate": 4.214506124626766e-06,
-      "loss": 0.2448,
+      "epoch": 4.191624319746279,
+      "grad_norm": 0.28154176473617554,
+      "learning_rate": 3.3614366019418255e-06,
+      "loss": 0.3723,
       "step": 116305
     },
     {
-      "epoch": 4.09,
-      "learning_rate": 4.212923382268022e-06,
-      "loss": 0.2586,
+      "epoch": 4.191804519407503,
+      "grad_norm": 0.3257730007171631,
+      "learning_rate": 3.3599752446612896e-06,
+      "loss": 0.3708,
       "step": 116310
     },
     {
-      "epoch": 4.09,
-      "learning_rate": 4.211340909815767e-06,
-      "loss": 0.2356,
+      "epoch": 4.191984719068728,
+      "grad_norm": 0.2922912836074829,
+      "learning_rate": 3.358514182221903e-06,
+      "loss": 0.3717,
       "step": 116315
     },
     {
-      "epoch": 4.09,
-      "learning_rate": 4.209758707290551e-06,
-      "loss": 0.2419,
+      "epoch": 4.192164918729953,
+      "grad_norm": 0.2203517109155655,
+      "learning_rate": 3.357053414643571e-06,
+      "loss": 0.3921,
       "step": 116320
     },
     {
-      "epoch": 4.09,
-      "learning_rate": 4.20817677471291e-06,
-      "loss": 0.232,
+      "epoch": 4.192345118391177,
+      "grad_norm": 0.27214643359184265,
+      "learning_rate": 3.355592941946195e-06,
+      "loss": 0.3657,
       "step": 116325
     },
     {
-      "epoch": 4.09,
-      "learning_rate": 4.206595112103398e-06,
-      "loss": 0.2493,
+      "epoch": 4.192525318052402,
+      "grad_norm": 0.22760187089443207,
+      "learning_rate": 3.3541327641496754e-06,
+      "loss": 0.3207,
       "step": 116330
     },
     {
-      "epoch": 4.09,
-      "learning_rate": 4.205013719482545e-06,
-      "loss": 0.262,
+      "epoch": 4.192705517713627,
+      "grad_norm": 0.2242773473262787,
+      "learning_rate": 3.3526728812738994e-06,
+      "loss": 0.3882,
       "step": 116335
     },
     {
-      "epoch": 4.09,
-      "learning_rate": 4.203432596870885e-06,
-      "loss": 0.2578,
+      "epoch": 4.1928857173748515,
+      "grad_norm": 0.2846883237361908,
+      "learning_rate": 3.35121329333877e-06,
+      "loss": 0.3775,
       "step": 116340
     },
     {
-      "epoch": 4.09,
-      "learning_rate": 4.20185174428894e-06,
-      "loss": 0.2718,
+      "epoch": 4.193065917036076,
+      "grad_norm": 0.2596665918827057,
+      "learning_rate": 3.349754000364169e-06,
+      "loss": 0.3672,
       "step": 116345
     },
     {
-      "epoch": 4.09,
-      "learning_rate": 4.200271161757249e-06,
-      "loss": 0.2511,
+      "epoch": 4.193246116697301,
+      "grad_norm": 0.2771369516849518,
+      "learning_rate": 3.3482950023699777e-06,
+      "loss": 0.3545,
       "step": 116350
     },
     {
-      "epoch": 4.09,
-      "learning_rate": 4.198690849296325e-06,
-      "loss": 0.2662,
+      "epoch": 4.193426316358526,
+      "grad_norm": 0.2566249966621399,
+      "learning_rate": 3.346836299376077e-06,
+      "loss": 0.3985,
       "step": 116355
     },
     {
-      "epoch": 4.09,
-      "learning_rate": 4.197110806926696e-06,
-      "loss": 0.2592,
+      "epoch": 4.1936065160197495,
+      "grad_norm": 0.24861952662467957,
+      "learning_rate": 3.345377891402332e-06,
+      "loss": 0.3828,
       "step": 116360
     },
     {
-      "epoch": 4.09,
-      "learning_rate": 4.19553103466887e-06,
-      "loss": 0.2554,
+      "epoch": 4.193786715680974,
+      "grad_norm": 0.2329019010066986,
+      "learning_rate": 3.3439197784686267e-06,
+      "loss": 0.4036,
       "step": 116365
     },
     {
-      "epoch": 4.09,
-      "learning_rate": 4.1939515325433665e-06,
-      "loss": 0.2606,
+      "epoch": 4.193966915342199,
+      "grad_norm": 0.20926065742969513,
+      "learning_rate": 3.342461960594823e-06,
+      "loss": 0.3728,
       "step": 116370
     },
     {
-      "epoch": 4.09,
-      "learning_rate": 4.192372300570693e-06,
-      "loss": 0.2593,
+      "epoch": 4.194147115003424,
+      "grad_norm": 0.29187721014022827,
+      "learning_rate": 3.341004437800779e-06,
+      "loss": 0.3899,
       "step": 116375
     },
     {
-      "epoch": 4.09,
-      "learning_rate": 4.190793338771351e-06,
-      "loss": 0.2644,
+      "epoch": 4.1943273146646485,
+      "grad_norm": 0.25245538353919983,
+      "learning_rate": 3.3395472101063584e-06,
+      "loss": 0.3589,
       "step": 116380
     },
     {
-      "epoch": 4.09,
-      "learning_rate": 4.189214647165837e-06,
-      "loss": 0.2494,
+      "epoch": 4.194507514325873,
+      "grad_norm": 0.24126659333705902,
+      "learning_rate": 3.338090277531408e-06,
+      "loss": 0.38,
       "step": 116385
     },
     {
-      "epoch": 4.09,
-      "learning_rate": 4.187636225774663e-06,
-      "loss": 0.2551,
+      "epoch": 4.194687713987098,
+      "grad_norm": 0.31214639544487,
+      "learning_rate": 3.336633640095793e-06,
+      "loss": 0.3653,
       "step": 116390
     },
     {
-      "epoch": 4.1,
-      "learning_rate": 4.186058074618318e-06,
-      "loss": 0.2457,
+      "epoch": 4.194867913648323,
+      "grad_norm": 0.2557271122932434,
+      "learning_rate": 3.3351772978193472e-06,
+      "loss": 0.3487,
       "step": 116395
     },
     {
-      "epoch": 4.1,
-      "learning_rate": 4.184480193717285e-06,
-      "loss": 0.2555,
+      "epoch": 4.1950481133095465,
+      "grad_norm": 0.22002442181110382,
+      "learning_rate": 3.3337212507219077e-06,
+      "loss": 0.3418,
       "step": 116400
     },
     {
-      "epoch": 4.1,
-      "learning_rate": 4.182902583092063e-06,
-      "loss": 0.2545,
+      "epoch": 4.195228312970771,
+      "grad_norm": 0.3052828907966614,
+      "learning_rate": 3.332265498823331e-06,
+      "loss": 0.3924,
       "step": 116405
     },
     {
-      "epoch": 4.1,
-      "learning_rate": 4.181325242763129e-06,
-      "loss": 0.2385,
+      "epoch": 4.195408512631996,
+      "grad_norm": 0.2319956123828888,
+      "learning_rate": 3.3308100421434397e-06,
+      "loss": 0.4031,
       "step": 116410
     },
     {
-      "epoch": 4.1,
-      "learning_rate": 4.179748172750972e-06,
-      "loss": 0.229,
+      "epoch": 4.195588712293221,
+      "grad_norm": 0.2713097631931305,
+      "learning_rate": 3.3293548807020654e-06,
+      "loss": 0.3555,
       "step": 116415
     },
     {
-      "epoch": 4.1,
-      "learning_rate": 4.178171373076056e-06,
-      "loss": 0.2596,
+      "epoch": 4.1957689119544455,
+      "grad_norm": 0.296252578496933,
+      "learning_rate": 3.3279000145190366e-06,
+      "loss": 0.3809,
       "step": 116420
     },
     {
-      "epoch": 4.1,
-      "learning_rate": 4.176594843758869e-06,
-      "loss": 0.2502,
+      "epoch": 4.19594911161567,
+      "grad_norm": 0.29615938663482666,
+      "learning_rate": 3.3264454436141677e-06,
+      "loss": 0.382,
       "step": 116425
     },
     {
-      "epoch": 4.1,
-      "learning_rate": 4.175018584819876e-06,
-      "loss": 0.2371,
+      "epoch": 4.196129311276895,
+      "grad_norm": 0.2384418398141861,
+      "learning_rate": 3.324991168007291e-06,
+      "loss": 0.3745,
       "step": 116430
     },
     {
-      "epoch": 4.1,
-      "learning_rate": 4.173442596279542e-06,
-      "loss": 0.234,
+      "epoch": 4.19630951093812,
+      "grad_norm": 0.26839661598205566,
+      "learning_rate": 3.323537187718212e-06,
+      "loss": 0.3893,
       "step": 116435
     },
     {
-      "epoch": 4.1,
-      "learning_rate": 4.171866878158326e-06,
-      "loss": 0.2441,
+      "epoch": 4.196489710599344,
+      "grad_norm": 0.3010355830192566,
+      "learning_rate": 3.322083502766743e-06,
+      "loss": 0.3965,
       "step": 116440
     },
     {
-      "epoch": 4.1,
-      "learning_rate": 4.170291430476697e-06,
-      "loss": 0.2283,
+      "epoch": 4.196669910260569,
+      "grad_norm": 0.25463324785232544,
+      "learning_rate": 3.3206301131726875e-06,
+      "loss": 0.3634,
       "step": 116445
     },
     {
-      "epoch": 4.1,
-      "learning_rate": 4.168716253255106e-06,
-      "loss": 0.2577,
+      "epoch": 4.196850109921793,
+      "grad_norm": 0.2410764843225479,
+      "learning_rate": 3.3191770189558434e-06,
+      "loss": 0.3554,
       "step": 116450
     },
     {
-      "epoch": 4.1,
-      "learning_rate": 4.167141346514006e-06,
-      "loss": 0.2455,
+      "epoch": 4.197030309583018,
+      "grad_norm": 0.315395325422287,
+      "learning_rate": 3.317724220136023e-06,
+      "loss": 0.3935,
       "step": 116455
     },
     {
-      "epoch": 4.1,
-      "learning_rate": 4.165566710273841e-06,
-      "loss": 0.2543,
+      "epoch": 4.197210509244242,
+      "grad_norm": 0.27405762672424316,
+      "learning_rate": 3.3162717167330103e-06,
+      "loss": 0.379,
       "step": 116460
     },
     {
-      "epoch": 4.1,
-      "learning_rate": 4.163992344555062e-06,
-      "loss": 0.2596,
+      "epoch": 4.197390708905467,
+      "grad_norm": 0.21512751281261444,
+      "learning_rate": 3.314819508766598e-06,
+      "loss": 0.358,
       "step": 116465
     },
     {
-      "epoch": 4.1,
-      "learning_rate": 4.162418249378117e-06,
-      "loss": 0.2453,
+      "epoch": 4.197570908566692,
+      "grad_norm": 0.2765870988368988,
+      "learning_rate": 3.3133675962565723e-06,
+      "loss": 0.3922,
       "step": 116470
     },
     {
-      "epoch": 4.1,
-      "learning_rate": 4.160844424763438e-06,
-      "loss": 0.2464,
+      "epoch": 4.197751108227917,
+      "grad_norm": 0.20413513481616974,
+      "learning_rate": 3.311915979222713e-06,
+      "loss": 0.3726,
       "step": 116475
     },
     {
-      "epoch": 4.1,
-      "learning_rate": 4.159270870731455e-06,
-      "loss": 0.2531,
+      "epoch": 4.197931307889141,
+      "grad_norm": 0.28501495718955994,
+      "learning_rate": 3.3104646576847947e-06,
+      "loss": 0.3839,
       "step": 116480
     },
     {
-      "epoch": 4.1,
-      "learning_rate": 4.157697587302612e-06,
-      "loss": 0.2168,
+      "epoch": 4.198111507550366,
+      "grad_norm": 0.2470470815896988,
+      "learning_rate": 3.309013631662608e-06,
+      "loss": 0.3825,
       "step": 116485
     },
     {
-      "epoch": 4.1,
-      "learning_rate": 4.156124574497327e-06,
-      "loss": 0.2791,
+      "epoch": 4.198291707211591,
+      "grad_norm": 0.2344074696302414,
+      "learning_rate": 3.3075629011759007e-06,
+      "loss": 0.37,
       "step": 116490
     },
     {
-      "epoch": 4.1,
-      "learning_rate": 4.154551832336031e-06,
-      "loss": 0.2586,
+      "epoch": 4.198471906872815,
+      "grad_norm": 0.2440483570098877,
+      "learning_rate": 3.3061124662444513e-06,
+      "loss": 0.3667,
       "step": 116495
     },
     {
-      "epoch": 4.1,
-      "learning_rate": 4.152979360839135e-06,
-      "loss": 0.2233,
+      "epoch": 4.198652106534039,
+      "grad_norm": 0.22694821655750275,
+      "learning_rate": 3.3046623268880224e-06,
+      "loss": 0.3171,
       "step": 116500
     },
     {
-      "epoch": 4.1,
-      "eval_loss": 0.2499181181192398,
-      "eval_runtime": 10.5421,
-      "eval_samples_per_second": 9.486,
-      "eval_steps_per_second": 9.486,
+      "epoch": 4.198652106534039,
+      "eval_loss": 0.42932018637657166,
+      "eval_runtime": 3.531,
+      "eval_samples_per_second": 28.32,
+      "eval_steps_per_second": 7.08,
       "step": 116500
     },
     {
-      "epoch": 4.1,
-      "learning_rate": 4.151407160027068e-06,
-      "loss": 0.2409,
+      "epoch": 4.198832306195264,
+      "grad_norm": 0.25988444685935974,
+      "learning_rate": 3.3032124831263668e-06,
+      "loss": 0.3446,
       "step": 116505
     },
     {
-      "epoch": 4.1,
-      "learning_rate": 4.149835229920241e-06,
-      "loss": 0.2562,
+      "epoch": 4.199012505856489,
+      "grad_norm": 0.2985076308250427,
+      "learning_rate": 3.301762934979244e-06,
+      "loss": 0.3893,
       "step": 116510
     },
     {
-      "epoch": 4.1,
-      "learning_rate": 4.148263570539057e-06,
-      "loss": 0.2317,
+      "epoch": 4.199192705517714,
+      "grad_norm": 0.24939928948879242,
+      "learning_rate": 3.3003136824663965e-06,
+      "loss": 0.3893,
       "step": 116515
     },
     {
-      "epoch": 4.1,
-      "learning_rate": 4.146692181903933e-06,
-      "loss": 0.2491,
+      "epoch": 4.199372905178938,
+      "grad_norm": 0.24855536222457886,
+      "learning_rate": 3.29886472560757e-06,
+      "loss": 0.3757,
       "step": 116520
     },
     {
-      "epoch": 4.1,
-      "learning_rate": 4.1451210640352625e-06,
-      "loss": 0.2983,
+      "epoch": 4.199553104840163,
+      "grad_norm": 0.23109517991542816,
+      "learning_rate": 3.297416064422515e-06,
+      "loss": 0.3638,
       "step": 116525
     },
     {
-      "epoch": 4.1,
-      "learning_rate": 4.1435502169534595e-06,
-      "loss": 0.251,
+      "epoch": 4.199733304501388,
+      "grad_norm": 0.3385915160179138,
+      "learning_rate": 3.2959676989309657e-06,
+      "loss": 0.3723,
       "step": 116530
     },
     {
-      "epoch": 4.1,
-      "learning_rate": 4.141979640678903e-06,
-      "loss": 0.2506,
+      "epoch": 4.1999135041626126,
+      "grad_norm": 0.24903911352157593,
+      "learning_rate": 3.2945196291526535e-06,
+      "loss": 0.3575,
       "step": 116535
     },
     {
-      "epoch": 4.1,
-      "learning_rate": 4.140409335232004e-06,
-      "loss": 0.2547,
+      "epoch": 4.200093703823836,
+      "grad_norm": 0.28424081206321716,
+      "learning_rate": 3.29307185510731e-06,
+      "loss": 0.3789,
       "step": 116540
     },
     {
-      "epoch": 4.1,
-      "learning_rate": 4.13883930063314e-06,
-      "loss": 0.242,
+      "epoch": 4.200273903485061,
+      "grad_norm": 0.2683737277984619,
+      "learning_rate": 3.2916243768146533e-06,
+      "loss": 0.3787,
       "step": 116545
     },
     {
-      "epoch": 4.1,
-      "learning_rate": 4.137269536902699e-06,
-      "loss": 0.2716,
+      "epoch": 4.200454103146286,
+      "grad_norm": 0.26493752002716064,
+      "learning_rate": 3.2901771942944167e-06,
+      "loss": 0.3706,
       "step": 116550
     },
     {
-      "epoch": 4.1,
-      "learning_rate": 4.135700044061061e-06,
-      "loss": 0.2277,
+      "epoch": 4.200634302807511,
+      "grad_norm": 0.33468732237815857,
+      "learning_rate": 3.288730307566312e-06,
+      "loss": 0.3563,
       "step": 116555
     },
     {
-      "epoch": 4.1,
-      "learning_rate": 4.134130822128612e-06,
-      "loss": 0.2366,
+      "epoch": 4.200814502468735,
+      "grad_norm": 0.26545238494873047,
+      "learning_rate": 3.2872837166500555e-06,
+      "loss": 0.3556,
       "step": 116560
     },
     {
-      "epoch": 4.1,
-      "learning_rate": 4.132561871125723e-06,
-      "loss": 0.2422,
+      "epoch": 4.20099470212996,
+      "grad_norm": 0.2382139265537262,
+      "learning_rate": 3.2858374215653547e-06,
+      "loss": 0.3753,
       "step": 116565
     },
     {
-      "epoch": 4.1,
-      "learning_rate": 4.1309931910727604e-06,
-      "loss": 0.2761,
+      "epoch": 4.201174901791185,
+      "grad_norm": 0.18432195484638214,
+      "learning_rate": 3.284391422331909e-06,
+      "loss": 0.3495,
       "step": 116570
     },
     {
-      "epoch": 4.1,
-      "learning_rate": 4.129424781990104e-06,
-      "loss": 0.243,
+      "epoch": 4.2013551014524095,
+      "grad_norm": 0.2506224513053894,
+      "learning_rate": 3.282945718969435e-06,
+      "loss": 0.3927,
       "step": 116575
     },
     {
-      "epoch": 4.1,
-      "learning_rate": 4.127856643898109e-06,
-      "loss": 0.2461,
+      "epoch": 4.201535301113634,
+      "grad_norm": 0.270535945892334,
+      "learning_rate": 3.2815003114976177e-06,
+      "loss": 0.383,
       "step": 116580
     },
     {
-      "epoch": 4.1,
-      "learning_rate": 4.126288776817142e-06,
-      "loss": 0.2552,
+      "epoch": 4.201715500774858,
+      "grad_norm": 0.26242825388908386,
+      "learning_rate": 3.2800551999361497e-06,
+      "loss": 0.3206,
       "step": 116585
     },
     {
-      "epoch": 4.1,
-      "learning_rate": 4.1247211807675625e-06,
-      "loss": 0.2589,
+      "epoch": 4.201895700436083,
+      "grad_norm": 0.29855820536613464,
+      "learning_rate": 3.2786103843047254e-06,
+      "loss": 0.3921,
       "step": 116590
     },
     {
-      "epoch": 4.1,
-      "learning_rate": 4.123153855769712e-06,
-      "loss": 0.2355,
+      "epoch": 4.202075900097308,
+      "grad_norm": 0.25370094180107117,
+      "learning_rate": 3.2771658646230274e-06,
+      "loss": 0.3744,
       "step": 116595
     },
     {
-      "epoch": 4.1,
-      "learning_rate": 4.121586801843958e-06,
-      "loss": 0.267,
+      "epoch": 4.202256099758532,
+      "grad_norm": 0.22978107631206512,
+      "learning_rate": 3.2757216409107506e-06,
+      "loss": 0.3969,
       "step": 116600
     },
     {
-      "epoch": 4.1,
-      "learning_rate": 4.12002001901064e-06,
-      "loss": 0.2511,
+      "epoch": 4.202436299419757,
+      "grad_norm": 0.2947256565093994,
+      "learning_rate": 3.2742777131875514e-06,
+      "loss": 0.3395,
       "step": 116605
     },
     {
-      "epoch": 4.1,
-      "learning_rate": 4.118453507290102e-06,
-      "loss": 0.2624,
+      "epoch": 4.202616499080982,
+      "grad_norm": 0.22788827121257782,
+      "learning_rate": 3.2728340814731082e-06,
+      "loss": 0.354,
       "step": 116610
     },
     {
-      "epoch": 4.1,
-      "learning_rate": 4.116887266702676e-06,
-      "loss": 0.2692,
+      "epoch": 4.2027966987422065,
+      "grad_norm": 0.3038329482078552,
+      "learning_rate": 3.2713907457871006e-06,
+      "loss": 0.3497,
       "step": 116615
     },
     {
-      "epoch": 4.1,
-      "learning_rate": 4.115321297268715e-06,
-      "loss": 0.2561,
+      "epoch": 4.202976898403431,
+      "grad_norm": 0.3307643234729767,
+      "learning_rate": 3.2699477061491873e-06,
+      "loss": 0.3688,
       "step": 116620
     },
     {
-      "epoch": 4.1,
-      "learning_rate": 4.113755599008537e-06,
-      "loss": 0.2559,
+      "epoch": 4.203157098064656,
+      "grad_norm": 0.2802697718143463,
+      "learning_rate": 3.2685049625790276e-06,
+      "loss": 0.4305,
       "step": 116625
     },
     {
-      "epoch": 4.1,
-      "learning_rate": 4.112190171942482e-06,
-      "loss": 0.2299,
+      "epoch": 4.203337297725881,
+      "grad_norm": 0.24683642387390137,
+      "learning_rate": 3.2670625150962806e-06,
+      "loss": 0.3701,
       "step": 116630
     },
     {
-      "epoch": 4.1,
-      "learning_rate": 4.110625016090866e-06,
-      "loss": 0.2546,
+      "epoch": 4.203517497387105,
+      "grad_norm": 0.22279055416584015,
+      "learning_rate": 3.265620363720595e-06,
+      "loss": 0.3596,
       "step": 116635
     },
     {
-      "epoch": 4.1,
-      "learning_rate": 4.109060131474026e-06,
-      "loss": 0.2338,
+      "epoch": 4.203697697048329,
+      "grad_norm": 0.21414312720298767,
+      "learning_rate": 3.2641785084716296e-06,
+      "loss": 0.4076,
       "step": 116640
     },
     {
-      "epoch": 4.1,
-      "learning_rate": 4.10749551811227e-06,
-      "loss": 0.2297,
+      "epoch": 4.203877896709554,
+      "grad_norm": 0.2691706717014313,
+      "learning_rate": 3.2627369493690245e-06,
+      "loss": 0.3498,
       "step": 116645
     },
     {
-      "epoch": 4.1,
-      "learning_rate": 4.105931176025912e-06,
-      "loss": 0.2499,
+      "epoch": 4.204058096370779,
+      "grad_norm": 0.24428191781044006,
+      "learning_rate": 3.261295686432417e-06,
+      "loss": 0.4001,
       "step": 116650
     },
     {
-      "epoch": 4.1,
-      "learning_rate": 4.104367105235274e-06,
-      "loss": 0.268,
+      "epoch": 4.2042382960320035,
+      "grad_norm": 0.23344086110591888,
+      "learning_rate": 3.2598547196814487e-06,
+      "loss": 0.3345,
       "step": 116655
     },
     {
-      "epoch": 4.1,
-      "learning_rate": 4.102803305760658e-06,
-      "loss": 0.2344,
+      "epoch": 4.204418495693228,
+      "grad_norm": 0.2314455211162567,
+      "learning_rate": 3.2584140491357523e-06,
+      "loss": 0.3732,
       "step": 116660
     },
     {
-      "epoch": 4.1,
-      "learning_rate": 4.101239777622368e-06,
-      "loss": 0.2388,
+      "epoch": 4.204598695354453,
+      "grad_norm": 0.2983637750148773,
+      "learning_rate": 3.2569736748149477e-06,
+      "loss": 0.4034,
       "step": 116665
     },
     {
-      "epoch": 4.1,
-      "learning_rate": 4.099676520840701e-06,
-      "loss": 0.2223,
+      "epoch": 4.204778895015678,
+      "grad_norm": 0.2737831473350525,
+      "learning_rate": 3.2555335967386804e-06,
+      "loss": 0.334,
       "step": 116670
     },
     {
-      "epoch": 4.1,
-      "learning_rate": 4.098113535435968e-06,
-      "loss": 0.2801,
+      "epoch": 4.204959094676902,
+      "grad_norm": 0.24176688492298126,
+      "learning_rate": 3.2540938149265453e-06,
+      "loss": 0.3541,
       "step": 116675
     },
     {
-      "epoch": 4.11,
-      "learning_rate": 4.096550821428455e-06,
-      "loss": 0.259,
+      "epoch": 4.205139294338126,
+      "grad_norm": 0.21122761070728302,
+      "learning_rate": 3.2526543293981775e-06,
+      "loss": 0.3499,
       "step": 116680
     },
     {
-      "epoch": 4.11,
-      "learning_rate": 4.094988378838446e-06,
-      "loss": 0.2475,
+      "epoch": 4.205319493999351,
+      "grad_norm": 0.23312951624393463,
+      "learning_rate": 3.2512151401731833e-06,
+      "loss": 0.4094,
       "step": 116685
     },
     {
-      "epoch": 4.11,
-      "learning_rate": 4.093426207686238e-06,
-      "loss": 0.233,
+      "epoch": 4.205499693660576,
+      "grad_norm": 0.24352633953094482,
+      "learning_rate": 3.249776247271172e-06,
+      "loss": 0.3615,
       "step": 116690
     },
     {
-      "epoch": 4.11,
-      "learning_rate": 4.091864307992119e-06,
-      "loss": 0.2482,
+      "epoch": 4.2056798933218005,
+      "grad_norm": 0.23205436766147614,
+      "learning_rate": 3.24833765071175e-06,
+      "loss": 0.3529,
       "step": 116695
     },
     {
-      "epoch": 4.11,
-      "learning_rate": 4.090302679776361e-06,
-      "loss": 0.2608,
+      "epoch": 4.205860092983025,
+      "grad_norm": 0.23461531102657318,
+      "learning_rate": 3.246899350514515e-06,
+      "loss": 0.3595,
       "step": 116700
     },
     {
-      "epoch": 4.11,
-      "learning_rate": 4.0887413230592444e-06,
-      "loss": 0.2399,
+      "epoch": 4.20604029264425,
+      "grad_norm": 0.26459741592407227,
+      "learning_rate": 3.245461346699061e-06,
+      "loss": 0.408,
       "step": 116705
     },
     {
-      "epoch": 4.11,
-      "learning_rate": 4.087180237861035e-06,
-      "loss": 0.2476,
+      "epoch": 4.206220492305475,
+      "grad_norm": 0.2987162172794342,
+      "learning_rate": 3.2440236392849877e-06,
+      "loss": 0.3556,
       "step": 116710
     },
     {
-      "epoch": 4.11,
-      "learning_rate": 4.0856194242020144e-06,
-      "loss": 0.2735,
+      "epoch": 4.206400691966699,
+      "grad_norm": 0.22682152688503265,
+      "learning_rate": 3.24258622829188e-06,
+      "loss": 0.361,
       "step": 116715
     },
     {
-      "epoch": 4.11,
-      "learning_rate": 4.08405888210244e-06,
-      "loss": 0.2621,
+      "epoch": 4.206580891627924,
+      "grad_norm": 0.2210111767053604,
+      "learning_rate": 3.241149113739325e-06,
+      "loss": 0.3509,
       "step": 116720
     },
     {
-      "epoch": 4.11,
-      "learning_rate": 4.08249861158258e-06,
-      "loss": 0.244,
+      "epoch": 4.206761091289148,
+      "grad_norm": 0.2724922001361847,
+      "learning_rate": 3.239712295646899e-06,
+      "loss": 0.3745,
       "step": 116725
     },
     {
-      "epoch": 4.11,
-      "learning_rate": 4.080938612662682e-06,
-      "loss": 0.2618,
+      "epoch": 4.206941290950373,
+      "grad_norm": 0.25349199771881104,
+      "learning_rate": 3.238275774034175e-06,
+      "loss": 0.3653,
       "step": 116730
     },
     {
-      "epoch": 4.11,
-      "learning_rate": 4.079378885363017e-06,
-      "loss": 0.2448,
+      "epoch": 4.2071214906115975,
+      "grad_norm": 0.29536816477775574,
+      "learning_rate": 3.2368395489207366e-06,
+      "loss": 0.383,
       "step": 116735
     },
     {
-      "epoch": 4.11,
-      "learning_rate": 4.0778194297038245e-06,
-      "loss": 0.2634,
+      "epoch": 4.207301690272822,
+      "grad_norm": 0.2729000151157379,
+      "learning_rate": 3.235403620326144e-06,
+      "loss": 0.354,
       "step": 116740
     },
     {
-      "epoch": 4.11,
-      "learning_rate": 4.076260245705362e-06,
-      "loss": 0.2495,
+      "epoch": 4.207481889934047,
+      "grad_norm": 0.27360138297080994,
+      "learning_rate": 3.233967988269965e-06,
+      "loss": 0.3776,
       "step": 116745
     },
     {
-      "epoch": 4.11,
-      "learning_rate": 4.074701333387867e-06,
-      "loss": 0.2487,
+      "epoch": 4.207662089595272,
+      "grad_norm": 0.20422430336475372,
+      "learning_rate": 3.2325326527717582e-06,
+      "loss": 0.3573,
       "step": 116750
     },
     {
-      "epoch": 4.11,
-      "learning_rate": 4.073142692771592e-06,
-      "loss": 0.2425,
+      "epoch": 4.207842289256496,
+      "grad_norm": 0.31160596013069153,
+      "learning_rate": 3.2310976138510724e-06,
+      "loss": 0.3787,
       "step": 116755
     },
     {
-      "epoch": 4.11,
-      "learning_rate": 4.071584323876765e-06,
-      "loss": 0.2397,
+      "epoch": 4.208022488917721,
+      "grad_norm": 0.2688230574131012,
+      "learning_rate": 3.2296628715274723e-06,
+      "loss": 0.3671,
       "step": 116760
     },
     {
-      "epoch": 4.11,
-      "learning_rate": 4.0700262267236195e-06,
-      "loss": 0.267,
+      "epoch": 4.208202688578946,
+      "grad_norm": 0.2496272772550583,
+      "learning_rate": 3.2282284258205093e-06,
+      "loss": 0.3467,
       "step": 116765
     },
     {
-      "epoch": 4.11,
-      "learning_rate": 4.068468401332395e-06,
-      "loss": 0.2278,
+      "epoch": 4.20838288824017,
+      "grad_norm": 0.2692166268825531,
+      "learning_rate": 3.226794276749706e-06,
+      "loss": 0.3607,
       "step": 116770
     },
     {
-      "epoch": 4.11,
-      "learning_rate": 4.066910847723315e-06,
-      "loss": 0.2298,
+      "epoch": 4.208563087901394,
+      "grad_norm": 0.27493536472320557,
+      "learning_rate": 3.2253604243346196e-06,
+      "loss": 0.3849,
       "step": 116775
     },
     {
-      "epoch": 4.11,
-      "learning_rate": 4.065353565916599e-06,
-      "loss": 0.2571,
+      "epoch": 4.208743287562619,
+      "grad_norm": 0.25217947363853455,
+      "learning_rate": 3.2239268685947787e-06,
+      "loss": 0.3608,
       "step": 116780
     },
     {
-      "epoch": 4.11,
-      "learning_rate": 4.063796555932467e-06,
-      "loss": 0.2576,
+      "epoch": 4.208923487223844,
+      "grad_norm": 0.24530518054962158,
+      "learning_rate": 3.222493609549726e-06,
+      "loss": 0.4019,
       "step": 116785
     },
     {
-      "epoch": 4.11,
-      "learning_rate": 4.062239817791147e-06,
-      "loss": 0.2294,
+      "epoch": 4.209103686885069,
+      "grad_norm": 0.2516281008720398,
+      "learning_rate": 3.2210606472189792e-06,
+      "loss": 0.3362,
       "step": 116790
     },
     {
-      "epoch": 4.11,
-      "learning_rate": 4.060683351512837e-06,
-      "loss": 0.2432,
+      "epoch": 4.209283886546293,
+      "grad_norm": 0.237876757979393,
+      "learning_rate": 3.2196279816220587e-06,
+      "loss": 0.3502,
       "step": 116795
     },
     {
-      "epoch": 4.11,
-      "learning_rate": 4.0591271571177596e-06,
-      "loss": 0.2427,
+      "epoch": 4.209464086207518,
+      "grad_norm": 0.242570698261261,
+      "learning_rate": 3.2181956127784933e-06,
+      "loss": 0.3677,
       "step": 116800
     },
     {
-      "epoch": 4.11,
-      "learning_rate": 4.057571234626109e-06,
-      "loss": 0.2379,
+      "epoch": 4.209644285868743,
+      "grad_norm": 0.2549905478954315,
+      "learning_rate": 3.2167635407077982e-06,
+      "loss": 0.3541,
       "step": 116805
     },
     {
-      "epoch": 4.11,
-      "learning_rate": 4.056015584058101e-06,
-      "loss": 0.2645,
+      "epoch": 4.209824485529968,
+      "grad_norm": 0.33291956782341003,
+      "learning_rate": 3.2153317654294803e-06,
+      "loss": 0.4255,
       "step": 116810
     },
     {
-      "epoch": 4.11,
-      "learning_rate": 4.054460205433927e-06,
-      "loss": 0.2201,
+      "epoch": 4.210004685191191,
+      "grad_norm": 0.2316872775554657,
+      "learning_rate": 3.213900286963051e-06,
+      "loss": 0.3734,
       "step": 116815
     },
     {
-      "epoch": 4.11,
-      "learning_rate": 4.052905098773785e-06,
-      "loss": 0.2621,
+      "epoch": 4.210184884852416,
+      "grad_norm": 0.24276618659496307,
+      "learning_rate": 3.212469105328006e-06,
+      "loss": 0.402,
       "step": 116820
     },
     {
-      "epoch": 4.11,
-      "learning_rate": 4.0513502640978584e-06,
-      "loss": 0.2348,
+      "epoch": 4.210365084513641,
+      "grad_norm": 0.20943395793437958,
+      "learning_rate": 3.211038220543858e-06,
+      "loss": 0.3512,
       "step": 116825
     },
     {
-      "epoch": 4.11,
-      "learning_rate": 4.049795701426351e-06,
-      "loss": 0.2692,
+      "epoch": 4.210545284174866,
+      "grad_norm": 0.263275682926178,
+      "learning_rate": 3.2096076326300933e-06,
+      "loss": 0.3615,
       "step": 116830
     },
     {
-      "epoch": 4.11,
-      "learning_rate": 4.048241410779438e-06,
-      "loss": 0.2655,
+      "epoch": 4.21072548383609,
+      "grad_norm": 0.234664648771286,
+      "learning_rate": 3.2081773416062083e-06,
+      "loss": 0.3926,
       "step": 116835
     },
     {
-      "epoch": 4.11,
-      "learning_rate": 4.046687392177303e-06,
-      "loss": 0.2686,
+      "epoch": 4.210905683497315,
+      "grad_norm": 0.2711605429649353,
+      "learning_rate": 3.206747347491687e-06,
+      "loss": 0.3848,
       "step": 116840
     },
     {
-      "epoch": 4.11,
-      "learning_rate": 4.045133645640118e-06,
-      "loss": 0.2371,
+      "epoch": 4.21108588315854,
+      "grad_norm": 0.22846084833145142,
+      "learning_rate": 3.205317650306014e-06,
+      "loss": 0.3744,
       "step": 116845
     },
     {
-      "epoch": 4.11,
-      "learning_rate": 4.043580171188061e-06,
-      "loss": 0.2566,
+      "epoch": 4.2112660828197646,
+      "grad_norm": 0.2208365499973297,
+      "learning_rate": 3.2038882500686625e-06,
+      "loss": 0.3703,
       "step": 116850
     },
     {
-      "epoch": 4.11,
-      "learning_rate": 4.042026968841311e-06,
-      "loss": 0.2576,
+      "epoch": 4.211446282480989,
+      "grad_norm": 0.3054751753807068,
+      "learning_rate": 3.202459146799125e-06,
+      "loss": 0.3998,
       "step": 116855
     },
     {
-      "epoch": 4.11,
-      "learning_rate": 4.040474038620032e-06,
-      "loss": 0.2379,
+      "epoch": 4.211626482142213,
+      "grad_norm": 0.23200340569019318,
+      "learning_rate": 3.201030340516853e-06,
+      "loss": 0.3711,
       "step": 116860
     },
     {
-      "epoch": 4.11,
-      "learning_rate": 4.0389213805443775e-06,
-      "loss": 0.2348,
+      "epoch": 4.211806681803438,
+      "grad_norm": 0.29509085416793823,
+      "learning_rate": 3.199601831241325e-06,
+      "loss": 0.385,
       "step": 116865
     },
     {
-      "epoch": 4.11,
-      "learning_rate": 4.037368994634519e-06,
-      "loss": 0.2544,
+      "epoch": 4.211986881464663,
+      "grad_norm": 0.22504891455173492,
+      "learning_rate": 3.198173618992001e-06,
+      "loss": 0.3803,
       "step": 116870
     },
     {
-      "epoch": 4.11,
-      "learning_rate": 4.035816880910612e-06,
-      "loss": 0.2621,
+      "epoch": 4.212167081125887,
+      "grad_norm": 0.2349441796541214,
+      "learning_rate": 3.196745703788334e-06,
+      "loss": 0.3666,
       "step": 116875
     },
     {
-      "epoch": 4.11,
-      "learning_rate": 4.034265039392798e-06,
-      "loss": 0.2743,
+      "epoch": 4.212347280787112,
+      "grad_norm": 0.3012699484825134,
+      "learning_rate": 3.1953180856497984e-06,
+      "loss": 0.3774,
       "step": 116880
     },
     {
-      "epoch": 4.11,
-      "learning_rate": 4.032713470101243e-06,
-      "loss": 0.2546,
+      "epoch": 4.212527480448337,
+      "grad_norm": 0.23085784912109375,
+      "learning_rate": 3.1938907645958275e-06,
+      "loss": 0.3879,
       "step": 116885
     },
     {
-      "epoch": 4.11,
-      "learning_rate": 4.031162173056086e-06,
-      "loss": 0.2264,
+      "epoch": 4.2127076801095615,
+      "grad_norm": 0.20077353715896606,
+      "learning_rate": 3.192463740645868e-06,
+      "loss": 0.3781,
       "step": 116890
     },
     {
-      "epoch": 4.11,
-      "learning_rate": 4.029611148277468e-06,
-      "loss": 0.2662,
+      "epoch": 4.212887879770786,
+      "grad_norm": 0.26119643449783325,
+      "learning_rate": 3.19103701381937e-06,
+      "loss": 0.3959,
       "step": 116895
     },
     {
-      "epoch": 4.11,
-      "learning_rate": 4.028060395785524e-06,
-      "loss": 0.2493,
+      "epoch": 4.213068079432011,
+      "grad_norm": 0.28771156072616577,
+      "learning_rate": 3.1896105841357745e-06,
+      "loss": 0.3555,
       "step": 116900
     },
     {
-      "epoch": 4.11,
-      "learning_rate": 4.026509915600393e-06,
-      "loss": 0.2831,
+      "epoch": 4.213248279093236,
+      "grad_norm": 0.23709803819656372,
+      "learning_rate": 3.188184451614509e-06,
+      "loss": 0.3638,
       "step": 116905
     },
     {
-      "epoch": 4.11,
-      "learning_rate": 4.024959707742218e-06,
-      "loss": 0.2659,
+      "epoch": 4.21342847875446,
+      "grad_norm": 0.25012511014938354,
+      "learning_rate": 3.1867586162750095e-06,
+      "loss": 0.3855,
       "step": 116910
     },
     {
-      "epoch": 4.11,
-      "learning_rate": 4.023409772231115e-06,
-      "loss": 0.246,
+      "epoch": 4.213608678415684,
+      "grad_norm": 0.24966490268707275,
+      "learning_rate": 3.185333078136693e-06,
+      "loss": 0.3769,
       "step": 116915
     },
     {
-      "epoch": 4.11,
-      "learning_rate": 4.021860109087208e-06,
-      "loss": 0.2348,
+      "epoch": 4.213788878076909,
+      "grad_norm": 0.2194073498249054,
+      "learning_rate": 3.1839078372189966e-06,
+      "loss": 0.3601,
       "step": 116920
     },
     {
-      "epoch": 4.11,
-      "learning_rate": 4.020310718330628e-06,
-      "loss": 0.2468,
+      "epoch": 4.213969077738134,
+      "grad_norm": 0.2655618488788605,
+      "learning_rate": 3.182482893541333e-06,
+      "loss": 0.4166,
       "step": 116925
     },
     {
-      "epoch": 4.11,
-      "learning_rate": 4.018761599981488e-06,
-      "loss": 0.2415,
+      "epoch": 4.2141492773993585,
+      "grad_norm": 0.23362277448177338,
+      "learning_rate": 3.1810582471231143e-06,
+      "loss": 0.3466,
       "step": 116930
     },
     {
-      "epoch": 4.11,
-      "learning_rate": 4.017212754059901e-06,
-      "loss": 0.2619,
+      "epoch": 4.214329477060583,
+      "grad_norm": 0.2612302303314209,
+      "learning_rate": 3.179633897983755e-06,
+      "loss": 0.3442,
       "step": 116935
     },
     {
-      "epoch": 4.11,
-      "learning_rate": 4.015664180585971e-06,
-      "loss": 0.2512,
+      "epoch": 4.214509676721808,
+      "grad_norm": 0.258493572473526,
+      "learning_rate": 3.1782098461426516e-06,
+      "loss": 0.3749,
       "step": 116940
     },
     {
-      "epoch": 4.11,
-      "learning_rate": 4.0141158795798185e-06,
-      "loss": 0.2675,
+      "epoch": 4.214689876383033,
+      "grad_norm": 0.297879159450531,
+      "learning_rate": 3.1767860916192216e-06,
+      "loss": 0.395,
       "step": 116945
     },
     {
-      "epoch": 4.11,
-      "learning_rate": 4.012567851061544e-06,
-      "loss": 0.243,
+      "epoch": 4.2148700760442575,
+      "grad_norm": 0.23272161185741425,
+      "learning_rate": 3.17536263443286e-06,
+      "loss": 0.3777,
       "step": 116950
     },
     {
-      "epoch": 4.11,
-      "learning_rate": 4.011020095051235e-06,
-      "loss": 0.2618,
+      "epoch": 4.215050275705481,
+      "grad_norm": 0.21118395030498505,
+      "learning_rate": 3.1739394746029465e-06,
+      "loss": 0.3803,
       "step": 116955
     },
     {
-      "epoch": 4.11,
-      "learning_rate": 4.009472611569006e-06,
-      "loss": 0.2237,
+      "epoch": 4.215230475366706,
+      "grad_norm": 0.1826326549053192,
+      "learning_rate": 3.1725166121488875e-06,
+      "loss": 0.3826,
       "step": 116960
     },
     {
-      "epoch": 4.12,
-      "learning_rate": 4.007925400634935e-06,
-      "loss": 0.2513,
+      "epoch": 4.215410675027931,
+      "grad_norm": 0.23955535888671875,
+      "learning_rate": 3.171094047090059e-06,
+      "loss": 0.3674,
       "step": 116965
     },
     {
-      "epoch": 4.12,
-      "learning_rate": 4.0063784622691255e-06,
-      "loss": 0.2646,
+      "epoch": 4.2155908746891555,
+      "grad_norm": 0.27745673060417175,
+      "learning_rate": 3.169671779445857e-06,
+      "loss": 0.3951,
       "step": 116970
     },
     {
-      "epoch": 4.12,
-      "learning_rate": 4.004831796491653e-06,
-      "loss": 0.2411,
+      "epoch": 4.21577107435038,
+      "grad_norm": 0.1933916211128235,
+      "learning_rate": 3.168249809235646e-06,
+      "loss": 0.3331,
       "step": 116975
     },
     {
-      "epoch": 4.12,
-      "learning_rate": 4.003285403322602e-06,
-      "loss": 0.244,
+      "epoch": 4.215951274011605,
+      "grad_norm": 0.30480435490608215,
+      "learning_rate": 3.1668281364788e-06,
+      "loss": 0.3533,
       "step": 116980
     },
     {
-      "epoch": 4.12,
-      "learning_rate": 4.001739282782055e-06,
-      "loss": 0.2428,
+      "epoch": 4.21613147367283,
+      "grad_norm": 0.2264707088470459,
+      "learning_rate": 3.1654067611946974e-06,
+      "loss": 0.3712,
       "step": 116985
     },
     {
-      "epoch": 4.12,
-      "learning_rate": 4.000193434890087e-06,
-      "loss": 0.2518,
+      "epoch": 4.216311673334054,
+      "grad_norm": 0.25249165296554565,
+      "learning_rate": 3.1639856834027036e-06,
+      "loss": 0.4058,
       "step": 116990
     },
     {
-      "epoch": 4.12,
-      "learning_rate": 3.99864785966676e-06,
-      "loss": 0.2488,
+      "epoch": 4.216491872995279,
+      "grad_norm": 0.22679844498634338,
+      "learning_rate": 3.1625649031221725e-06,
+      "loss": 0.3795,
       "step": 116995
     },
     {
-      "epoch": 4.12,
-      "learning_rate": 3.997102557132157e-06,
-      "loss": 0.25,
+      "epoch": 4.216672072656503,
+      "grad_norm": 0.29318711161613464,
+      "learning_rate": 3.1611444203724695e-06,
+      "loss": 0.3723,
       "step": 117000
     },
     {
-      "epoch": 4.12,
-      "eval_loss": 0.24963203072547913,
-      "eval_runtime": 10.5439,
-      "eval_samples_per_second": 9.484,
-      "eval_steps_per_second": 9.484,
+      "epoch": 4.216672072656503,
+      "eval_loss": 0.4291633367538452,
+      "eval_runtime": 3.5298,
+      "eval_samples_per_second": 28.33,
+      "eval_steps_per_second": 7.083,
       "step": 117000
     },
     {
-      "epoch": 4.12,
-      "learning_rate": 3.995557527306337e-06,
-      "loss": 0.2382,
+      "epoch": 4.216852272317728,
+      "grad_norm": 0.26223427057266235,
+      "learning_rate": 3.1597242351729425e-06,
+      "loss": 0.3777,
       "step": 117005
     },
     {
-      "epoch": 4.12,
-      "learning_rate": 3.994012770209354e-06,
-      "loss": 0.2457,
+      "epoch": 4.2170324719789525,
+      "grad_norm": 0.24939653277397156,
+      "learning_rate": 3.1583043475429453e-06,
+      "loss": 0.3564,
       "step": 117010
     },
     {
-      "epoch": 4.12,
-      "learning_rate": 3.992468285861275e-06,
-      "loss": 0.2589,
+      "epoch": 4.217212671640177,
+      "grad_norm": 0.27071085572242737,
+      "learning_rate": 3.156884757501824e-06,
+      "loss": 0.3712,
       "step": 117015
     },
     {
-      "epoch": 4.12,
-      "learning_rate": 3.9909240742821484e-06,
-      "loss": 0.2411,
+      "epoch": 4.217392871301402,
+      "grad_norm": 0.2698659598827362,
+      "learning_rate": 3.1554654650689215e-06,
+      "loss": 0.394,
       "step": 117020
     },
     {
-      "epoch": 4.12,
-      "learning_rate": 3.989380135492032e-06,
-      "loss": 0.2512,
+      "epoch": 4.217573070962627,
+      "grad_norm": 0.28294146060943604,
+      "learning_rate": 3.154046470263569e-06,
+      "loss": 0.3858,
       "step": 117025
     },
     {
-      "epoch": 4.12,
-      "learning_rate": 3.987836469510967e-06,
-      "loss": 0.2572,
+      "epoch": 4.217753270623851,
+      "grad_norm": 0.28658536076545715,
+      "learning_rate": 3.1526277731051074e-06,
+      "loss": 0.3811,
       "step": 117030
     },
     {
-      "epoch": 4.12,
-      "learning_rate": 3.986293076358993e-06,
-      "loss": 0.2433,
+      "epoch": 4.217933470285076,
+      "grad_norm": 0.22426393628120422,
+      "learning_rate": 3.1512093736128538e-06,
+      "loss": 0.3315,
       "step": 117035
     },
     {
-      "epoch": 4.12,
-      "learning_rate": 3.9847499560561605e-06,
-      "loss": 0.2575,
+      "epoch": 4.218113669946301,
+      "grad_norm": 0.22765573859214783,
+      "learning_rate": 3.149791271806152e-06,
+      "loss": 0.3562,
       "step": 117040
     },
     {
-      "epoch": 4.12,
-      "learning_rate": 3.9832071086225e-06,
-      "loss": 0.2379,
+      "epoch": 4.218293869607525,
+      "grad_norm": 0.2529604434967041,
+      "learning_rate": 3.1483734677043102e-06,
+      "loss": 0.3777,
       "step": 117045
     },
     {
-      "epoch": 4.12,
-      "learning_rate": 3.981664534078045e-06,
-      "loss": 0.2512,
+      "epoch": 4.2184740692687495,
+      "grad_norm": 0.21222354471683502,
+      "learning_rate": 3.14695596132665e-06,
+      "loss": 0.3615,
       "step": 117050
     },
     {
-      "epoch": 4.12,
-      "learning_rate": 3.980122232442817e-06,
-      "loss": 0.2441,
+      "epoch": 4.218654268929974,
+      "grad_norm": 0.28866201639175415,
+      "learning_rate": 3.145538752692484e-06,
+      "loss": 0.3926,
       "step": 117055
     },
     {
-      "epoch": 4.12,
-      "learning_rate": 3.978580203736857e-06,
-      "loss": 0.2501,
+      "epoch": 4.218834468591199,
+      "grad_norm": 0.27018246054649353,
+      "learning_rate": 3.1441218418211155e-06,
+      "loss": 0.3999,
       "step": 117060
     },
     {
-      "epoch": 4.12,
-      "learning_rate": 3.977038447980172e-06,
-      "loss": 0.2599,
+      "epoch": 4.219014668252424,
+      "grad_norm": 0.2012593001127243,
+      "learning_rate": 3.142705228731868e-06,
+      "loss": 0.3466,
       "step": 117065
     },
     {
-      "epoch": 4.12,
-      "learning_rate": 3.975496965192796e-06,
-      "loss": 0.2381,
+      "epoch": 4.219194867913648,
+      "grad_norm": 0.33578363060951233,
+      "learning_rate": 3.141288913444024e-06,
+      "loss": 0.3932,
       "step": 117070
     },
     {
-      "epoch": 4.12,
-      "learning_rate": 3.9739557553947285e-06,
-      "loss": 0.2545,
+      "epoch": 4.219375067574873,
+      "grad_norm": 0.24772273004055023,
+      "learning_rate": 3.139872895976881e-06,
+      "loss": 0.3928,
       "step": 117075
     },
     {
-      "epoch": 4.12,
-      "learning_rate": 3.972414818605996e-06,
-      "loss": 0.2488,
+      "epoch": 4.219555267236098,
+      "grad_norm": 0.2516150176525116,
+      "learning_rate": 3.1384571763497435e-06,
+      "loss": 0.3273,
       "step": 117080
     },
     {
-      "epoch": 4.12,
-      "learning_rate": 3.970874154846599e-06,
-      "loss": 0.2561,
+      "epoch": 4.219735466897323,
+      "grad_norm": 0.3026273250579834,
+      "learning_rate": 3.1370417545818933e-06,
+      "loss": 0.388,
       "step": 117085
     },
     {
-      "epoch": 4.12,
-      "learning_rate": 3.969333764136543e-06,
-      "loss": 0.2487,
+      "epoch": 4.219915666558546,
+      "grad_norm": 0.2576843202114105,
+      "learning_rate": 3.1356266306926175e-06,
+      "loss": 0.3755,
       "step": 117090
     },
     {
-      "epoch": 4.12,
-      "learning_rate": 3.967793646495821e-06,
-      "loss": 0.2425,
+      "epoch": 4.220095866219771,
+      "grad_norm": 0.24807590246200562,
+      "learning_rate": 3.1342118047011955e-06,
+      "loss": 0.3744,
       "step": 117095
     },
     {
-      "epoch": 4.12,
-      "learning_rate": 3.966253801944444e-06,
-      "loss": 0.2653,
+      "epoch": 4.220276065880996,
+      "grad_norm": 0.25404757261276245,
+      "learning_rate": 3.1327972766268977e-06,
+      "loss": 0.3697,
       "step": 117100
     },
     {
-      "epoch": 4.12,
-      "learning_rate": 3.9647142305024e-06,
-      "loss": 0.2509,
+      "epoch": 4.220456265542221,
+      "grad_norm": 0.22809305787086487,
+      "learning_rate": 3.1313830464890085e-06,
+      "loss": 0.3502,
       "step": 117105
     },
     {
-      "epoch": 4.12,
-      "learning_rate": 3.963174932189672e-06,
-      "loss": 0.241,
+      "epoch": 4.220636465203445,
+      "grad_norm": 0.2612314224243164,
+      "learning_rate": 3.1299691143067937e-06,
+      "loss": 0.3539,
       "step": 117110
     },
     {
-      "epoch": 4.12,
-      "learning_rate": 3.9616359070262625e-06,
-      "loss": 0.2357,
+      "epoch": 4.22081666486467,
+      "grad_norm": 0.3267393708229065,
+      "learning_rate": 3.1285554800995127e-06,
+      "loss": 0.389,
       "step": 117115
     },
     {
-      "epoch": 4.12,
-      "learning_rate": 3.9600971550321444e-06,
-      "loss": 0.2503,
+      "epoch": 4.220996864525895,
+      "grad_norm": 0.23879055678844452,
+      "learning_rate": 3.127142143886433e-06,
+      "loss": 0.4019,
       "step": 117120
     },
     {
-      "epoch": 4.12,
-      "learning_rate": 3.958558676227292e-06,
-      "loss": 0.2639,
+      "epoch": 4.22117706418712,
+      "grad_norm": 0.27509593963623047,
+      "learning_rate": 3.1257291056867976e-06,
+      "loss": 0.377,
       "step": 117125
     },
     {
-      "epoch": 4.12,
-      "learning_rate": 3.9570204706316895e-06,
-      "loss": 0.2484,
+      "epoch": 4.221357263848344,
+      "grad_norm": 0.3244267702102661,
+      "learning_rate": 3.124316365519875e-06,
+      "loss": 0.3989,
       "step": 117130
     },
     {
-      "epoch": 4.12,
-      "learning_rate": 3.955482538265312e-06,
-      "loss": 0.2749,
+      "epoch": 4.221537463509568,
+      "grad_norm": 0.2525143623352051,
+      "learning_rate": 3.122903923404913e-06,
+      "loss": 0.3723,
       "step": 117135
     },
     {
-      "epoch": 4.12,
-      "learning_rate": 3.953944879148128e-06,
-      "loss": 0.2528,
+      "epoch": 4.221717663170793,
+      "grad_norm": 0.2540621757507324,
+      "learning_rate": 3.1214917793611355e-06,
+      "loss": 0.3725,
       "step": 117140
     },
     {
-      "epoch": 4.12,
-      "learning_rate": 3.9524074933000974e-06,
-      "loss": 0.2484,
+      "epoch": 4.221897862832018,
+      "grad_norm": 0.2635304033756256,
+      "learning_rate": 3.1200799334078044e-06,
+      "loss": 0.346,
       "step": 117145
     },
     {
-      "epoch": 4.12,
-      "learning_rate": 3.950870380741181e-06,
-      "loss": 0.245,
+      "epoch": 4.222078062493242,
+      "grad_norm": 0.24621783196926117,
+      "learning_rate": 3.118668385564144e-06,
+      "loss": 0.3364,
       "step": 117150
     },
     {
-      "epoch": 4.12,
-      "learning_rate": 3.949333541491346e-06,
-      "loss": 0.2538,
+      "epoch": 4.222258262154467,
+      "grad_norm": 0.2203269898891449,
+      "learning_rate": 3.117257135849394e-06,
+      "loss": 0.3385,
       "step": 117155
     },
     {
-      "epoch": 4.12,
-      "learning_rate": 3.947796975570542e-06,
-      "loss": 0.258,
+      "epoch": 4.222438461815692,
+      "grad_norm": 0.2397884875535965,
+      "learning_rate": 3.1158461842827833e-06,
+      "loss": 0.3664,
       "step": 117160
     },
     {
-      "epoch": 4.12,
-      "learning_rate": 3.946260682998717e-06,
-      "loss": 0.2483,
+      "epoch": 4.222618661476917,
+      "grad_norm": 0.2857881188392639,
+      "learning_rate": 3.114435530883522e-06,
+      "loss": 0.3521,
       "step": 117165
     },
     {
-      "epoch": 4.12,
-      "learning_rate": 3.94472466379582e-06,
-      "loss": 0.2473,
+      "epoch": 4.222798861138141,
+      "grad_norm": 0.25649380683898926,
+      "learning_rate": 3.113025175670847e-06,
+      "loss": 0.3339,
       "step": 117170
     },
     {
-      "epoch": 4.12,
-      "learning_rate": 3.943188917981802e-06,
-      "loss": 0.2507,
+      "epoch": 4.222979060799366,
+      "grad_norm": 0.30346161127090454,
+      "learning_rate": 3.111615118663963e-06,
+      "loss": 0.3665,
       "step": 117175
     },
     {
-      "epoch": 4.12,
-      "learning_rate": 3.941653445576596e-06,
-      "loss": 0.2639,
+      "epoch": 4.223159260460591,
+      "grad_norm": 0.29739436507225037,
+      "learning_rate": 3.1102053598820874e-06,
+      "loss": 0.3754,
       "step": 117180
     },
     {
-      "epoch": 4.12,
-      "learning_rate": 3.940118246600144e-06,
-      "loss": 0.2472,
+      "epoch": 4.223339460121815,
+      "grad_norm": 0.23481184244155884,
+      "learning_rate": 3.108795899344424e-06,
+      "loss": 0.3601,
       "step": 117185
     },
     {
-      "epoch": 4.12,
-      "learning_rate": 3.938583321072373e-06,
-      "loss": 0.2471,
+      "epoch": 4.223519659783039,
+      "grad_norm": 0.2552500069141388,
+      "learning_rate": 3.107386737070175e-06,
+      "loss": 0.4017,
       "step": 117190
     },
     {
-      "epoch": 4.12,
-      "learning_rate": 3.937048669013224e-06,
-      "loss": 0.247,
+      "epoch": 4.223699859444264,
+      "grad_norm": 0.29149770736694336,
+      "learning_rate": 3.105977873078547e-06,
+      "loss": 0.3966,
       "step": 117195
     },
     {
-      "epoch": 4.12,
-      "learning_rate": 3.935514290442616e-06,
-      "loss": 0.2579,
+      "epoch": 4.223880059105489,
+      "grad_norm": 0.28181055188179016,
+      "learning_rate": 3.1045693073887327e-06,
+      "loss": 0.4144,
       "step": 117200
     },
     {
-      "epoch": 4.12,
-      "learning_rate": 3.933980185380473e-06,
-      "loss": 0.261,
+      "epoch": 4.2240602587667135,
+      "grad_norm": 0.25641298294067383,
+      "learning_rate": 3.1031610400199225e-06,
+      "loss": 0.3707,
       "step": 117205
     },
     {
-      "epoch": 4.12,
-      "learning_rate": 3.932446353846708e-06,
-      "loss": 0.2598,
+      "epoch": 4.224240458427938,
+      "grad_norm": 0.20379073917865753,
+      "learning_rate": 3.1017530709913033e-06,
+      "loss": 0.3584,
       "step": 117210
     },
     {
-      "epoch": 4.12,
-      "learning_rate": 3.930912795861249e-06,
-      "loss": 0.262,
+      "epoch": 4.224420658089163,
+      "grad_norm": 0.2511664927005768,
+      "learning_rate": 3.1003454003220577e-06,
+      "loss": 0.374,
       "step": 117215
     },
     {
-      "epoch": 4.12,
-      "learning_rate": 3.929379511444006e-06,
-      "loss": 0.2617,
+      "epoch": 4.224600857750388,
+      "grad_norm": 0.21009521186351776,
+      "learning_rate": 3.098938028031362e-06,
+      "loss": 0.3763,
       "step": 117220
     },
     {
-      "epoch": 4.12,
-      "learning_rate": 3.9278465006148735e-06,
-      "loss": 0.2531,
+      "epoch": 4.2247810574116125,
+      "grad_norm": 0.21885792911052704,
+      "learning_rate": 3.0975309541384e-06,
+      "loss": 0.3636,
       "step": 117225
     },
     {
-      "epoch": 4.12,
-      "learning_rate": 3.926313763393774e-06,
-      "loss": 0.2529,
+      "epoch": 4.224961257072836,
+      "grad_norm": 0.2734248638153076,
+      "learning_rate": 3.096124178662338e-06,
+      "loss": 0.3794,
       "step": 117230
     },
     {
-      "epoch": 4.12,
-      "learning_rate": 3.924781299800598e-06,
-      "loss": 0.2453,
+      "epoch": 4.225141456734061,
+      "grad_norm": 0.24762408435344696,
+      "learning_rate": 3.094717701622346e-06,
+      "loss": 0.3325,
       "step": 117235
     },
     {
-      "epoch": 4.12,
-      "learning_rate": 3.9232491098552535e-06,
-      "loss": 0.2564,
+      "epoch": 4.225321656395286,
+      "grad_norm": 0.3509252071380615,
+      "learning_rate": 3.0933115230375812e-06,
+      "loss": 0.3739,
       "step": 117240
     },
     {
-      "epoch": 4.13,
-      "learning_rate": 3.921717193577621e-06,
-      "loss": 0.2345,
+      "epoch": 4.2255018560565105,
+      "grad_norm": 0.23616372048854828,
+      "learning_rate": 3.0919056429272037e-06,
+      "loss": 0.3129,
       "step": 117245
     },
     {
-      "epoch": 4.13,
-      "learning_rate": 3.920185550987607e-06,
-      "loss": 0.2375,
+      "epoch": 4.225682055717735,
+      "grad_norm": 0.33164361119270325,
+      "learning_rate": 3.090500061310378e-06,
+      "loss": 0.3645,
       "step": 117250
     },
     {
-      "epoch": 4.13,
-      "learning_rate": 3.918654182105092e-06,
-      "loss": 0.2377,
+      "epoch": 4.22586225537896,
+      "grad_norm": 0.2735852003097534,
+      "learning_rate": 3.0890947782062416e-06,
+      "loss": 0.3664,
       "step": 117255
     },
     {
-      "epoch": 4.13,
-      "learning_rate": 3.917123086949959e-06,
-      "loss": 0.2494,
+      "epoch": 4.226042455040185,
+      "grad_norm": 0.22918978333473206,
+      "learning_rate": 3.087689793633944e-06,
+      "loss": 0.3726,
       "step": 117260
     },
     {
-      "epoch": 4.13,
-      "learning_rate": 3.915592265542081e-06,
-      "loss": 0.2634,
+      "epoch": 4.2262226547014095,
+      "grad_norm": 0.2797573506832123,
+      "learning_rate": 3.0862851076126325e-06,
+      "loss": 0.3745,
       "step": 117265
     },
     {
-      "epoch": 4.13,
-      "learning_rate": 3.914061717901352e-06,
-      "loss": 0.2385,
+      "epoch": 4.226402854362634,
+      "grad_norm": 0.26490533351898193,
+      "learning_rate": 3.084880720161448e-06,
+      "loss": 0.3739,
       "step": 117270
     },
     {
-      "epoch": 4.13,
-      "learning_rate": 3.9125314440476316e-06,
-      "loss": 0.2662,
+      "epoch": 4.226583054023858,
+      "grad_norm": 0.2413850724697113,
+      "learning_rate": 3.0834766312995174e-06,
+      "loss": 0.3652,
       "step": 117275
     },
     {
-      "epoch": 4.13,
-      "learning_rate": 3.911001444000795e-06,
-      "loss": 0.2587,
+      "epoch": 4.226763253685083,
+      "grad_norm": 0.2586573660373688,
+      "learning_rate": 3.082072841045977e-06,
+      "loss": 0.3765,
       "step": 117280
     },
     {
-      "epoch": 4.13,
-      "learning_rate": 3.9094717177806995e-06,
-      "loss": 0.2776,
+      "epoch": 4.2269434533463075,
+      "grad_norm": 0.2556487023830414,
+      "learning_rate": 3.080669349419943e-06,
+      "loss": 0.4075,
       "step": 117285
     },
     {
-      "epoch": 4.13,
-      "learning_rate": 3.907942265407216e-06,
-      "loss": 0.2637,
+      "epoch": 4.227123653007532,
+      "grad_norm": 0.2673538029193878,
+      "learning_rate": 3.0792661564405524e-06,
+      "loss": 0.4263,
       "step": 117290
     },
     {
-      "epoch": 4.13,
-      "learning_rate": 3.906413086900207e-06,
-      "loss": 0.2112,
+      "epoch": 4.227303852668757,
+      "grad_norm": 0.2932944893836975,
+      "learning_rate": 3.077863262126915e-06,
+      "loss": 0.3598,
       "step": 117295
     },
     {
-      "epoch": 4.13,
-      "learning_rate": 3.904884182279523e-06,
-      "loss": 0.259,
+      "epoch": 4.227484052329982,
+      "grad_norm": 0.22899527847766876,
+      "learning_rate": 3.076460666498146e-06,
+      "loss": 0.3761,
       "step": 117300
     },
     {
-      "epoch": 4.13,
-      "learning_rate": 3.903355551565011e-06,
-      "loss": 0.2503,
+      "epoch": 4.227664251991206,
+      "grad_norm": 0.26272207498550415,
+      "learning_rate": 3.075058369573358e-06,
+      "loss": 0.3494,
       "step": 117305
     },
     {
-      "epoch": 4.13,
-      "learning_rate": 3.9018271947765274e-06,
-      "loss": 0.2463,
+      "epoch": 4.227844451652431,
+      "grad_norm": 0.28200453519821167,
+      "learning_rate": 3.07365637137165e-06,
+      "loss": 0.3753,
       "step": 117310
     },
     {
-      "epoch": 4.13,
-      "learning_rate": 3.900299111933917e-06,
-      "loss": 0.2535,
+      "epoch": 4.228024651313656,
+      "grad_norm": 0.2724056839942932,
+      "learning_rate": 3.072254671912131e-06,
+      "loss": 0.3659,
       "step": 117315
     },
     {
-      "epoch": 4.13,
-      "learning_rate": 3.898771303057014e-06,
-      "loss": 0.2524,
+      "epoch": 4.22820485097488,
+      "grad_norm": 0.19723065197467804,
+      "learning_rate": 3.0708532712139e-06,
+      "loss": 0.3401,
       "step": 117320
     },
     {
-      "epoch": 4.13,
-      "learning_rate": 3.8972437681656535e-06,
-      "loss": 0.258,
+      "epoch": 4.2283850506361045,
+      "grad_norm": 0.23599091172218323,
+      "learning_rate": 3.0694521692960444e-06,
+      "loss": 0.4004,
       "step": 117325
     },
     {
-      "epoch": 4.13,
-      "learning_rate": 3.895716507279684e-06,
-      "loss": 0.2496,
+      "epoch": 4.228565250297329,
+      "grad_norm": 0.2825251817703247,
+      "learning_rate": 3.06805136617766e-06,
+      "loss": 0.3491,
       "step": 117330
     },
     {
-      "epoch": 4.13,
-      "learning_rate": 3.8941895204189256e-06,
-      "loss": 0.2401,
+      "epoch": 4.228745449958554,
+      "grad_norm": 0.2366117537021637,
+      "learning_rate": 3.066650861877823e-06,
+      "loss": 0.3718,
       "step": 117335
     },
     {
-      "epoch": 4.13,
-      "learning_rate": 3.8926628076032e-06,
-      "loss": 0.2656,
+      "epoch": 4.228925649619779,
+      "grad_norm": 0.252418577671051,
+      "learning_rate": 3.0652506564156275e-06,
+      "loss": 0.4073,
       "step": 117340
     },
     {
-      "epoch": 4.13,
-      "learning_rate": 3.891136368852341e-06,
-      "loss": 0.2565,
+      "epoch": 4.229105849281003,
+      "grad_norm": 0.25921630859375,
+      "learning_rate": 3.063850749810149e-06,
+      "loss": 0.4102,
       "step": 117345
     },
     {
-      "epoch": 4.13,
-      "learning_rate": 3.8896102041861695e-06,
-      "loss": 0.2567,
+      "epoch": 4.229286048942228,
+      "grad_norm": 0.3047523498535156,
+      "learning_rate": 3.062451142080447e-06,
+      "loss": 0.3875,
       "step": 117350
     },
     {
-      "epoch": 4.13,
-      "learning_rate": 3.888084313624496e-06,
-      "loss": 0.2358,
+      "epoch": 4.229466248603453,
+      "grad_norm": 0.2350996434688568,
+      "learning_rate": 3.061051833245607e-06,
+      "loss": 0.3618,
       "step": 117355
     },
     {
-      "epoch": 4.13,
-      "learning_rate": 3.886558697187129e-06,
-      "loss": 0.2781,
+      "epoch": 4.229646448264678,
+      "grad_norm": 0.19591708481311798,
+      "learning_rate": 3.059652823324688e-06,
+      "loss": 0.3718,
       "step": 117360
     },
     {
-      "epoch": 4.13,
-      "learning_rate": 3.885033354893888e-06,
-      "loss": 0.2804,
+      "epoch": 4.229826647925902,
+      "grad_norm": 0.26747336983680725,
+      "learning_rate": 3.058254112336753e-06,
+      "loss": 0.378,
       "step": 117365
     },
     {
-      "epoch": 4.13,
-      "learning_rate": 3.883508286764576e-06,
-      "loss": 0.257,
+      "epoch": 4.230006847587126,
+      "grad_norm": 0.25700604915618896,
+      "learning_rate": 3.0568557003008535e-06,
+      "loss": 0.3677,
       "step": 117370
     },
     {
-      "epoch": 4.13,
-      "learning_rate": 3.881983492818991e-06,
-      "loss": 0.2531,
+      "epoch": 4.230187047248351,
+      "grad_norm": 0.21863490343093872,
+      "learning_rate": 3.0554575872360463e-06,
+      "loss": 0.4036,
       "step": 117375
     },
     {
-      "epoch": 4.13,
-      "learning_rate": 3.880458973076928e-06,
-      "loss": 0.2463,
+      "epoch": 4.230367246909576,
+      "grad_norm": 0.2407938539981842,
+      "learning_rate": 3.054059773161383e-06,
+      "loss": 0.3465,
       "step": 117380
     },
     {
-      "epoch": 4.13,
-      "learning_rate": 3.878934727558192e-06,
-      "loss": 0.2458,
+      "epoch": 4.2305474465708,
+      "grad_norm": 0.27205365896224976,
+      "learning_rate": 3.0526622580959063e-06,
+      "loss": 0.3897,
       "step": 117385
     },
     {
-      "epoch": 4.13,
-      "learning_rate": 3.877410756282571e-06,
-      "loss": 0.2487,
+      "epoch": 4.230727646232025,
+      "grad_norm": 0.21772317588329315,
+      "learning_rate": 3.0512650420586595e-06,
+      "loss": 0.3788,
       "step": 117390
     },
     {
-      "epoch": 4.13,
-      "learning_rate": 3.875887059269842e-06,
-      "loss": 0.2371,
+      "epoch": 4.23090784589325,
+      "grad_norm": 0.24121488630771637,
+      "learning_rate": 3.049868125068675e-06,
+      "loss": 0.3711,
       "step": 117395
     },
     {
-      "epoch": 4.13,
-      "learning_rate": 3.874363636539808e-06,
-      "loss": 0.2452,
+      "epoch": 4.231088045554475,
+      "grad_norm": 0.22327007353305817,
+      "learning_rate": 3.04847150714499e-06,
+      "loss": 0.3705,
       "step": 117400
     },
     {
-      "epoch": 4.13,
-      "learning_rate": 3.872840488112231e-06,
-      "loss": 0.2494,
+      "epoch": 4.231268245215699,
+      "grad_norm": 0.21851101517677307,
+      "learning_rate": 3.0470751883066223e-06,
+      "loss": 0.3382,
       "step": 117405
     },
     {
-      "epoch": 4.13,
-      "learning_rate": 3.871317614006903e-06,
-      "loss": 0.2555,
+      "epoch": 4.231448444876923,
+      "grad_norm": 0.3136238753795624,
+      "learning_rate": 3.0456791685726126e-06,
+      "loss": 0.3754,
       "step": 117410
     },
     {
-      "epoch": 4.13,
-      "learning_rate": 3.869795014243591e-06,
-      "loss": 0.2455,
+      "epoch": 4.231628644538148,
+      "grad_norm": 0.24977144598960876,
+      "learning_rate": 3.0442834479619707e-06,
+      "loss": 0.3548,
       "step": 117415
     },
     {
-      "epoch": 4.13,
-      "learning_rate": 3.868272688842062e-06,
-      "loss": 0.2585,
+      "epoch": 4.231808844199373,
+      "grad_norm": 0.24291959404945374,
+      "learning_rate": 3.0428880264937204e-06,
+      "loss": 0.3734,
       "step": 117420
     },
     {
-      "epoch": 4.13,
-      "learning_rate": 3.866750637822089e-06,
-      "loss": 0.2314,
+      "epoch": 4.231989043860597,
+      "grad_norm": 0.2255113422870636,
+      "learning_rate": 3.0414929041868657e-06,
+      "loss": 0.3476,
       "step": 117425
     },
     {
-      "epoch": 4.13,
-      "learning_rate": 3.865228861203435e-06,
-      "loss": 0.2348,
+      "epoch": 4.232169243521822,
+      "grad_norm": 0.5030208230018616,
+      "learning_rate": 3.0400980810604164e-06,
+      "loss": 0.3713,
       "step": 117430
     },
     {
-      "epoch": 4.13,
-      "learning_rate": 3.863707359005852e-06,
-      "loss": 0.2387,
+      "epoch": 4.232349443183047,
+      "grad_norm": 0.22084102034568787,
+      "learning_rate": 3.0387035571333906e-06,
+      "loss": 0.365,
       "step": 117435
     },
     {
-      "epoch": 4.13,
-      "learning_rate": 3.862186131249096e-06,
-      "loss": 0.2661,
+      "epoch": 4.232529642844272,
+      "grad_norm": 0.20808632671833038,
+      "learning_rate": 3.0373093324247676e-06,
+      "loss": 0.3707,
       "step": 117440
     },
     {
-      "epoch": 4.13,
-      "learning_rate": 3.860665177952927e-06,
-      "loss": 0.2644,
+      "epoch": 4.232709842505496,
+      "grad_norm": 0.2130783647298813,
+      "learning_rate": 3.035915406953552e-06,
+      "loss": 0.3532,
       "step": 117445
     },
     {
-      "epoch": 4.13,
-      "learning_rate": 3.859144499137085e-06,
-      "loss": 0.2466,
+      "epoch": 4.232890042166721,
+      "grad_norm": 0.205308198928833,
+      "learning_rate": 3.0345217807387387e-06,
+      "loss": 0.35,
       "step": 117450
     },
     {
-      "epoch": 4.13,
-      "learning_rate": 3.857624094821324e-06,
-      "loss": 0.2712,
+      "epoch": 4.233070241827946,
+      "grad_norm": 0.2538002133369446,
+      "learning_rate": 3.033128453799308e-06,
+      "loss": 0.3864,
       "step": 117455
     },
     {
-      "epoch": 4.13,
-      "learning_rate": 3.856103965025376e-06,
-      "loss": 0.2314,
+      "epoch": 4.23325044148917,
+      "grad_norm": 0.27659669518470764,
+      "learning_rate": 3.0317354261542613e-06,
+      "loss": 0.3974,
       "step": 117460
     },
     {
-      "epoch": 4.13,
-      "learning_rate": 3.854584109768991e-06,
-      "loss": 0.2546,
+      "epoch": 4.233430641150394,
+      "grad_norm": 0.2688102722167969,
+      "learning_rate": 3.0303426978225607e-06,
+      "loss": 0.3609,
       "step": 117465
     },
     {
-      "epoch": 4.13,
-      "learning_rate": 3.853064529071895e-06,
-      "loss": 0.2517,
+      "epoch": 4.233610840811619,
+      "grad_norm": 0.25696924328804016,
+      "learning_rate": 3.028950268823183e-06,
+      "loss": 0.3414,
       "step": 117470
     },
     {
-      "epoch": 4.13,
-      "learning_rate": 3.851545222953812e-06,
-      "loss": 0.2374,
+      "epoch": 4.233791040472844,
+      "grad_norm": 0.19637776911258698,
+      "learning_rate": 3.0275581391751075e-06,
+      "loss": 0.3418,
       "step": 117475
     },
     {
-      "epoch": 4.13,
-      "learning_rate": 3.850026191434483e-06,
-      "loss": 0.2398,
+      "epoch": 4.233971240134069,
+      "grad_norm": 0.25194182991981506,
+      "learning_rate": 3.026166308897299e-06,
+      "loss": 0.3505,
       "step": 117480
     },
     {
-      "epoch": 4.13,
-      "learning_rate": 3.848507434533627e-06,
-      "loss": 0.2402,
+      "epoch": 4.234151439795293,
+      "grad_norm": 0.3067736029624939,
+      "learning_rate": 3.024774778008721e-06,
+      "loss": 0.3806,
       "step": 117485
     },
     {
-      "epoch": 4.13,
-      "learning_rate": 3.846988952270964e-06,
-      "loss": 0.2551,
+      "epoch": 4.234331639456518,
+      "grad_norm": 0.2704439163208008,
+      "learning_rate": 3.023383546528333e-06,
+      "loss": 0.4056,
       "step": 117490
     },
     {
-      "epoch": 4.13,
-      "learning_rate": 3.8454707446661995e-06,
-      "loss": 0.2419,
+      "epoch": 4.234511839117743,
+      "grad_norm": 0.23782427608966827,
+      "learning_rate": 3.021992614475083e-06,
+      "loss": 0.3591,
       "step": 117495
     },
     {
-      "epoch": 4.13,
-      "learning_rate": 3.8439528117390625e-06,
-      "loss": 0.2405,
+      "epoch": 4.2346920387789675,
+      "grad_norm": 0.21595478057861328,
+      "learning_rate": 3.020601981867932e-06,
+      "loss": 0.3604,
       "step": 117500
     },
     {
-      "epoch": 4.13,
-      "eval_loss": 0.2497539073228836,
-      "eval_runtime": 10.5514,
-      "eval_samples_per_second": 9.477,
-      "eval_steps_per_second": 9.477,
+      "epoch": 4.2346920387789675,
+      "eval_loss": 0.42911702394485474,
+      "eval_runtime": 3.5344,
+      "eval_samples_per_second": 28.293,
+      "eval_steps_per_second": 7.073,
       "step": 117500
     },
     {
-      "epoch": 4.13,
-      "learning_rate": 3.842435153509252e-06,
-      "loss": 0.2553,
+      "epoch": 4.234872238440191,
+      "grad_norm": 0.2689270079135895,
+      "learning_rate": 3.0192116487258252e-06,
+      "loss": 0.378,
       "step": 117505
     },
     {
-      "epoch": 4.13,
-      "learning_rate": 3.840917769996483e-06,
-      "loss": 0.231,
+      "epoch": 4.235052438101416,
+      "grad_norm": 0.24561098217964172,
+      "learning_rate": 3.017821615067706e-06,
+      "loss": 0.349,
       "step": 117510
     },
     {
-      "epoch": 4.13,
-      "learning_rate": 3.839400661220446e-06,
-      "loss": 0.2546,
+      "epoch": 4.235232637762641,
+      "grad_norm": 0.2601664364337921,
+      "learning_rate": 3.01643188091251e-06,
+      "loss": 0.3717,
       "step": 117515
     },
     {
-      "epoch": 4.13,
-      "learning_rate": 3.837883827200853e-06,
-      "loss": 0.2302,
+      "epoch": 4.2354128374238655,
+      "grad_norm": 0.21555201709270477,
+      "learning_rate": 3.0150424462791706e-06,
+      "loss": 0.404,
       "step": 117520
     },
     {
-      "epoch": 4.13,
-      "learning_rate": 3.836367267957391e-06,
-      "loss": 0.2676,
+      "epoch": 4.23559303708509,
+      "grad_norm": 0.30406418442726135,
+      "learning_rate": 3.0136533111866206e-06,
+      "loss": 0.3917,
       "step": 117525
     },
     {
-      "epoch": 4.14,
-      "learning_rate": 3.834850983509752e-06,
-      "loss": 0.2365,
+      "epoch": 4.235773236746315,
+      "grad_norm": 0.21674159169197083,
+      "learning_rate": 3.012264475653795e-06,
+      "loss": 0.365,
       "step": 117530
     },
     {
-      "epoch": 4.14,
-      "learning_rate": 3.833334973877623e-06,
-      "loss": 0.26,
+      "epoch": 4.23595343640754,
+      "grad_norm": 0.27085331082344055,
+      "learning_rate": 3.0108759396996004e-06,
+      "loss": 0.3447,
       "step": 117535
     },
     {
-      "epoch": 4.14,
-      "learning_rate": 3.831819239080692e-06,
-      "loss": 0.2557,
+      "epoch": 4.2361336360687645,
+      "grad_norm": 0.20267651975154877,
+      "learning_rate": 3.0094877033429667e-06,
+      "loss": 0.3808,
       "step": 117540
     },
     {
-      "epoch": 4.14,
-      "learning_rate": 3.830303779138639e-06,
-      "loss": 0.2491,
+      "epoch": 4.236313835729989,
+      "grad_norm": 0.26923617720603943,
+      "learning_rate": 3.0080997666028095e-06,
+      "loss": 0.3694,
       "step": 117545
     },
     {
-      "epoch": 4.14,
-      "learning_rate": 3.828788594071137e-06,
-      "loss": 0.2355,
+      "epoch": 4.236494035391213,
+      "grad_norm": 0.21540358662605286,
+      "learning_rate": 3.0067121294980323e-06,
+      "loss": 0.3543,
       "step": 117550
     },
     {
-      "epoch": 4.14,
-      "learning_rate": 3.827273683897861e-06,
-      "loss": 0.2416,
+      "epoch": 4.236674235052438,
+      "grad_norm": 0.23205356299877167,
+      "learning_rate": 3.0053247920475452e-06,
+      "loss": 0.4063,
       "step": 117555
     },
     {
-      "epoch": 4.14,
-      "learning_rate": 3.825759048638486e-06,
-      "loss": 0.2717,
+      "epoch": 4.2368544347136625,
+      "grad_norm": 0.20041492581367493,
+      "learning_rate": 3.0039377542702417e-06,
+      "loss": 0.3609,
       "step": 117560
     },
     {
-      "epoch": 4.14,
-      "learning_rate": 3.824244688312673e-06,
-      "loss": 0.2813,
+      "epoch": 4.237034634374887,
+      "grad_norm": 0.3079770803451538,
+      "learning_rate": 3.0025510161850373e-06,
+      "loss": 0.3538,
       "step": 117565
     },
     {
-      "epoch": 4.14,
-      "learning_rate": 3.8227306029400834e-06,
-      "loss": 0.2594,
+      "epoch": 4.237214834036112,
+      "grad_norm": 0.28639087080955505,
+      "learning_rate": 3.0011645778108137e-06,
+      "loss": 0.3659,
       "step": 117570
     },
     {
-      "epoch": 4.14,
-      "learning_rate": 3.821216792540388e-06,
-      "loss": 0.2486,
+      "epoch": 4.237395033697337,
+      "grad_norm": 0.2304767519235611,
+      "learning_rate": 2.9997784391664647e-06,
+      "loss": 0.4097,
       "step": 117575
     },
     {
-      "epoch": 4.14,
-      "learning_rate": 3.819703257133236e-06,
-      "loss": 0.251,
+      "epoch": 4.2375752333585615,
+      "grad_norm": 0.2690269947052002,
+      "learning_rate": 2.9983926002708746e-06,
+      "loss": 0.3809,
       "step": 117580
     },
     {
-      "epoch": 4.14,
-      "learning_rate": 3.818189996738278e-06,
-      "loss": 0.2698,
+      "epoch": 4.237755433019786,
+      "grad_norm": 0.26503047347068787,
+      "learning_rate": 2.9970070611429257e-06,
+      "loss": 0.3656,
       "step": 117585
     },
     {
-      "epoch": 4.14,
-      "learning_rate": 3.8166770113751585e-06,
-      "loss": 0.2715,
+      "epoch": 4.237935632681011,
+      "grad_norm": 0.1937418282032013,
+      "learning_rate": 2.9956218218014893e-06,
+      "loss": 0.3704,
       "step": 117590
     },
     {
-      "epoch": 4.14,
-      "learning_rate": 3.815164301063531e-06,
-      "loss": 0.2503,
+      "epoch": 4.238115832342235,
+      "grad_norm": 0.24524284899234772,
+      "learning_rate": 2.9942368822654527e-06,
+      "loss": 0.4087,
       "step": 117595
     },
     {
-      "epoch": 4.14,
-      "learning_rate": 3.8136518658230367e-06,
-      "loss": 0.2364,
+      "epoch": 4.2382960320034595,
+      "grad_norm": 0.27891620993614197,
+      "learning_rate": 2.9928522425536757e-06,
+      "loss": 0.3533,
       "step": 117600
     },
     {
-      "epoch": 4.14,
-      "learning_rate": 3.8121397056733093e-06,
-      "loss": 0.2451,
+      "epoch": 4.238476231664684,
+      "grad_norm": 0.2644476890563965,
+      "learning_rate": 2.991467902685027e-06,
+      "loss": 0.3914,
       "step": 117605
     },
     {
-      "epoch": 4.14,
-      "learning_rate": 3.81062782063398e-06,
-      "loss": 0.2568,
+      "epoch": 4.238656431325909,
+      "grad_norm": 0.2886243164539337,
+      "learning_rate": 2.9900838626783693e-06,
+      "loss": 0.3735,
       "step": 117610
     },
     {
-      "epoch": 4.14,
-      "learning_rate": 3.809116210724689e-06,
-      "loss": 0.2539,
+      "epoch": 4.238836630987134,
+      "grad_norm": 0.2373829036951065,
+      "learning_rate": 2.9887001225525514e-06,
+      "loss": 0.3564,
       "step": 117615
     },
     {
-      "epoch": 4.14,
-      "learning_rate": 3.807604875965051e-06,
-      "loss": 0.2518,
+      "epoch": 4.239016830648358,
+      "grad_norm": 0.22608888149261475,
+      "learning_rate": 2.987316682326438e-06,
+      "loss": 0.3711,
       "step": 117620
     },
     {
-      "epoch": 4.14,
-      "learning_rate": 3.806093816374706e-06,
-      "loss": 0.2304,
+      "epoch": 4.239197030309583,
+      "grad_norm": 0.24686965346336365,
+      "learning_rate": 2.985933542018876e-06,
+      "loss": 0.3849,
       "step": 117625
     },
     {
-      "epoch": 4.14,
-      "learning_rate": 3.8045830319732577e-06,
-      "loss": 0.2379,
+      "epoch": 4.239377229970808,
+      "grad_norm": 0.22954131662845612,
+      "learning_rate": 2.9845507016487e-06,
+      "loss": 0.3794,
       "step": 117630
     },
     {
-      "epoch": 4.14,
-      "learning_rate": 3.8030725227803377e-06,
-      "loss": 0.2314,
+      "epoch": 4.239557429632033,
+      "grad_norm": 0.2016417235136032,
+      "learning_rate": 2.9831681612347616e-06,
+      "loss": 0.3694,
       "step": 117635
     },
     {
-      "epoch": 4.14,
-      "learning_rate": 3.8015622888155523e-06,
-      "loss": 0.2765,
+      "epoch": 4.239737629293257,
+      "grad_norm": 0.21057425439357758,
+      "learning_rate": 2.9817859207958905e-06,
+      "loss": 0.3543,
       "step": 117640
     },
     {
-      "epoch": 4.14,
-      "learning_rate": 3.800052330098508e-06,
-      "loss": 0.2572,
+      "epoch": 4.239917828954481,
+      "grad_norm": 0.23975323140621185,
+      "learning_rate": 2.980403980350932e-06,
+      "loss": 0.3292,
       "step": 117645
     },
     {
-      "epoch": 4.14,
-      "learning_rate": 3.798542646648809e-06,
-      "loss": 0.2456,
+      "epoch": 4.240098028615706,
+      "grad_norm": 0.23370075225830078,
+      "learning_rate": 2.979022339918702e-06,
+      "loss": 0.3572,
       "step": 117650
     },
     {
-      "epoch": 4.14,
-      "learning_rate": 3.7970332384860642e-06,
-      "loss": 0.2484,
+      "epoch": 4.240278228276931,
+      "grad_norm": 0.22406595945358276,
+      "learning_rate": 2.977640999518025e-06,
+      "loss": 0.3527,
       "step": 117655
     },
     {
-      "epoch": 4.14,
-      "learning_rate": 3.795524105629872e-06,
-      "loss": 0.2325,
+      "epoch": 4.240458427938155,
+      "grad_norm": 0.27608826756477356,
+      "learning_rate": 2.976259959167732e-06,
+      "loss": 0.3598,
       "step": 117660
     },
     {
-      "epoch": 4.14,
-      "learning_rate": 3.794015248099825e-06,
-      "loss": 0.249,
+      "epoch": 4.24063862759938,
+      "grad_norm": 0.286156564950943,
+      "learning_rate": 2.9748792188866308e-06,
+      "loss": 0.3439,
       "step": 117665
     },
     {
-      "epoch": 4.14,
-      "learning_rate": 3.79250666591551e-06,
-      "loss": 0.2444,
+      "epoch": 4.240818827260605,
+      "grad_norm": 0.3072472810745239,
+      "learning_rate": 2.973498778693537e-06,
+      "loss": 0.3777,
       "step": 117670
     },
     {
-      "epoch": 4.14,
-      "learning_rate": 3.7909983590965176e-06,
-      "loss": 0.2437,
+      "epoch": 4.24099902692183,
+      "grad_norm": 0.2714073956012726,
+      "learning_rate": 2.9721186386072573e-06,
+      "loss": 0.3941,
       "step": 117675
     },
     {
-      "epoch": 4.14,
-      "learning_rate": 3.78949032766244e-06,
-      "loss": 0.2645,
+      "epoch": 4.241179226583054,
+      "grad_norm": 0.22121752798557281,
+      "learning_rate": 2.970738798646591e-06,
+      "loss": 0.354,
       "step": 117680
     },
     {
-      "epoch": 4.14,
-      "learning_rate": 3.7879825716328475e-06,
-      "loss": 0.2362,
+      "epoch": 4.241359426244279,
+      "grad_norm": 0.2743302285671234,
+      "learning_rate": 2.9693592588303482e-06,
+      "loss": 0.3862,
       "step": 117685
     },
     {
-      "epoch": 4.14,
-      "learning_rate": 3.786475091027328e-06,
-      "loss": 0.2244,
+      "epoch": 4.241539625905503,
+      "grad_norm": 0.2942706048488617,
+      "learning_rate": 2.9679800191773215e-06,
+      "loss": 0.388,
       "step": 117690
     },
     {
-      "epoch": 4.14,
-      "learning_rate": 3.784967885865451e-06,
-      "loss": 0.2653,
+      "epoch": 4.241719825566728,
+      "grad_norm": 0.26952454447746277,
+      "learning_rate": 2.966601079706302e-06,
+      "loss": 0.381,
       "step": 117695
     },
     {
-      "epoch": 4.14,
-      "learning_rate": 3.7834609561667817e-06,
-      "loss": 0.2634,
+      "epoch": 4.241900025227952,
+      "grad_norm": 0.17778491973876953,
+      "learning_rate": 2.9652224404360747e-06,
+      "loss": 0.3192,
       "step": 117700
     },
     {
-      "epoch": 4.14,
-      "learning_rate": 3.7819543019508852e-06,
-      "loss": 0.2359,
+      "epoch": 4.242080224889177,
+      "grad_norm": 0.21804936230182648,
+      "learning_rate": 2.9638441013854268e-06,
+      "loss": 0.3632,
       "step": 117705
     },
     {
-      "epoch": 4.14,
-      "learning_rate": 3.7804479232373345e-06,
-      "loss": 0.2654,
+      "epoch": 4.242260424550402,
+      "grad_norm": 0.25280699133872986,
+      "learning_rate": 2.9624660625731325e-06,
+      "loss": 0.3887,
       "step": 117710
     },
     {
-      "epoch": 4.14,
-      "learning_rate": 3.778941820045681e-06,
-      "loss": 0.2537,
+      "epoch": 4.242440624211627,
+      "grad_norm": 0.2739708721637726,
+      "learning_rate": 2.96108832401798e-06,
+      "loss": 0.3653,
       "step": 117715
     },
     {
-      "epoch": 4.14,
-      "learning_rate": 3.7774359923954866e-06,
-      "loss": 0.2491,
+      "epoch": 4.242620823872851,
+      "grad_norm": 0.19621554017066956,
+      "learning_rate": 2.959710885738723e-06,
+      "loss": 0.3652,
       "step": 117720
     },
     {
-      "epoch": 4.14,
-      "learning_rate": 3.7759304403062914e-06,
-      "loss": 0.2456,
+      "epoch": 4.242801023534076,
+      "grad_norm": 0.25483089685440063,
+      "learning_rate": 2.9583337477541437e-06,
+      "loss": 0.3788,
       "step": 117725
     },
     {
-      "epoch": 4.14,
-      "learning_rate": 3.774425163797654e-06,
-      "loss": 0.2637,
+      "epoch": 4.242981223195301,
+      "grad_norm": 0.25373703241348267,
+      "learning_rate": 2.9569569100830003e-06,
+      "loss": 0.3526,
       "step": 117730
     },
     {
-      "epoch": 4.14,
-      "learning_rate": 3.7729201628891207e-06,
-      "loss": 0.2687,
+      "epoch": 4.243161422856525,
+      "grad_norm": 0.29680660367012024,
+      "learning_rate": 2.9555803727440436e-06,
+      "loss": 0.4243,
       "step": 117735
     },
     {
-      "epoch": 4.14,
-      "learning_rate": 3.7714154376002313e-06,
-      "loss": 0.2358,
+      "epoch": 4.243341622517749,
+      "grad_norm": 0.24780967831611633,
+      "learning_rate": 2.95420413575605e-06,
+      "loss": 0.3667,
       "step": 117740
     },
     {
-      "epoch": 4.14,
-      "learning_rate": 3.7699109879505173e-06,
-      "loss": 0.268,
+      "epoch": 4.243521822178974,
+      "grad_norm": 0.22331710159778595,
+      "learning_rate": 2.9528281991377443e-06,
+      "loss": 0.3457,
       "step": 117745
     },
     {
-      "epoch": 4.14,
-      "learning_rate": 3.7684068139595243e-06,
-      "loss": 0.2769,
+      "epoch": 4.243702021840199,
+      "grad_norm": 0.20010043680667877,
+      "learning_rate": 2.9514525629078915e-06,
+      "loss": 0.3773,
       "step": 117750
     },
     {
-      "epoch": 4.14,
-      "learning_rate": 3.766902915646775e-06,
-      "loss": 0.2521,
+      "epoch": 4.243882221501424,
+      "grad_norm": 0.27777811884880066,
+      "learning_rate": 2.9500772270852268e-06,
+      "loss": 0.3659,
       "step": 117755
     },
     {
-      "epoch": 4.14,
-      "learning_rate": 3.7653992930317987e-06,
-      "loss": 0.2455,
+      "epoch": 4.244062421162648,
+      "grad_norm": 0.25709882378578186,
+      "learning_rate": 2.948702191688493e-06,
+      "loss": 0.385,
       "step": 117760
     },
     {
-      "epoch": 4.14,
-      "learning_rate": 3.7638959461341133e-06,
-      "loss": 0.2522,
+      "epoch": 4.244242620823873,
+      "grad_norm": 0.3200331926345825,
+      "learning_rate": 2.9473274567364234e-06,
+      "loss": 0.3751,
       "step": 117765
     },
     {
-      "epoch": 4.14,
-      "learning_rate": 3.76239287497325e-06,
-      "loss": 0.2544,
+      "epoch": 4.244422820485098,
+      "grad_norm": 0.1971009075641632,
+      "learning_rate": 2.9459530222477464e-06,
+      "loss": 0.3772,
       "step": 117770
     },
     {
-      "epoch": 4.14,
-      "learning_rate": 3.7608900795687186e-06,
-      "loss": 0.2571,
+      "epoch": 4.2446030201463225,
+      "grad_norm": 0.23797380924224854,
+      "learning_rate": 2.944578888241184e-06,
+      "loss": 0.3566,
       "step": 117775
     },
     {
-      "epoch": 4.14,
-      "learning_rate": 3.759387559940025e-06,
-      "loss": 0.2311,
+      "epoch": 4.244783219807546,
+      "grad_norm": 0.2632618844509125,
+      "learning_rate": 2.9432050547354706e-06,
+      "loss": 0.3802,
       "step": 117780
     },
     {
-      "epoch": 4.14,
-      "learning_rate": 3.7578853161066878e-06,
-      "loss": 0.2418,
+      "epoch": 4.244963419468771,
+      "grad_norm": 0.30606886744499207,
+      "learning_rate": 2.9418315217493165e-06,
+      "loss": 0.3788,
       "step": 117785
     },
     {
-      "epoch": 4.14,
-      "learning_rate": 3.756383348088216e-06,
-      "loss": 0.2437,
+      "epoch": 4.245143619129996,
+      "grad_norm": 0.25161272287368774,
+      "learning_rate": 2.94045828930144e-06,
+      "loss": 0.3892,
       "step": 117790
     },
     {
-      "epoch": 4.14,
-      "learning_rate": 3.754881655904105e-06,
-      "loss": 0.2508,
+      "epoch": 4.245323818791221,
+      "grad_norm": 0.24037761986255646,
+      "learning_rate": 2.9390853574105455e-06,
+      "loss": 0.3661,
       "step": 117795
     },
     {
-      "epoch": 4.14,
-      "learning_rate": 3.7533802395738503e-06,
-      "loss": 0.2473,
+      "epoch": 4.245504018452445,
+      "grad_norm": 0.2599053680896759,
+      "learning_rate": 2.9377127260953374e-06,
+      "loss": 0.3867,
       "step": 117800
     },
     {
-      "epoch": 4.14,
-      "learning_rate": 3.7518790991169563e-06,
-      "loss": 0.2565,
+      "epoch": 4.24568421811367,
+      "grad_norm": 0.2276243269443512,
+      "learning_rate": 2.936340395374529e-06,
+      "loss": 0.4001,
       "step": 117805
     },
     {
-      "epoch": 4.14,
-      "learning_rate": 3.7503782345529094e-06,
-      "loss": 0.2237,
+      "epoch": 4.245864417774895,
+      "grad_norm": 0.3195003271102905,
+      "learning_rate": 2.934968365266816e-06,
+      "loss": 0.351,
       "step": 117810
     },
     {
-      "epoch": 4.15,
-      "learning_rate": 3.7488776459011972e-06,
-      "loss": 0.2534,
+      "epoch": 4.2460446174361195,
+      "grad_norm": 0.24188397824764252,
+      "learning_rate": 2.933596635790875e-06,
+      "loss": 0.3847,
       "step": 117815
     },
     {
-      "epoch": 4.15,
-      "learning_rate": 3.7473773331812982e-06,
-      "loss": 0.249,
+      "epoch": 4.246224817097344,
+      "grad_norm": 0.23399367928504944,
+      "learning_rate": 2.9322252069654137e-06,
+      "loss": 0.3921,
       "step": 117820
     },
     {
-      "epoch": 4.15,
-      "learning_rate": 3.745877296412706e-06,
-      "loss": 0.2373,
+      "epoch": 4.246405016758568,
+      "grad_norm": 0.25146761536598206,
+      "learning_rate": 2.930854078809106e-06,
+      "loss": 0.3672,
       "step": 117825
     },
     {
-      "epoch": 4.15,
-      "learning_rate": 3.7443775356148875e-06,
-      "loss": 0.2657,
+      "epoch": 4.246585216419793,
+      "grad_norm": 0.27147242426872253,
+      "learning_rate": 2.929483251340648e-06,
+      "loss": 0.4172,
       "step": 117830
     },
     {
-      "epoch": 4.15,
-      "learning_rate": 3.7428780508073134e-06,
-      "loss": 0.287,
+      "epoch": 4.2467654160810175,
+      "grad_norm": 0.21517179906368256,
+      "learning_rate": 2.9281127245787047e-06,
+      "loss": 0.3483,
       "step": 117835
     },
     {
-      "epoch": 4.15,
-      "learning_rate": 3.741378842009466e-06,
-      "loss": 0.2629,
+      "epoch": 4.246945615742242,
+      "grad_norm": 0.1976982206106186,
+      "learning_rate": 2.9267424985419457e-06,
+      "loss": 0.3928,
       "step": 117840
     },
     {
-      "epoch": 4.15,
-      "learning_rate": 3.7398799092407956e-06,
-      "loss": 0.2411,
+      "epoch": 4.247125815403467,
+      "grad_norm": 0.2845209836959839,
+      "learning_rate": 2.925372573249052e-06,
+      "loss": 0.3635,
       "step": 117845
     },
     {
-      "epoch": 4.15,
-      "learning_rate": 3.738381252520781e-06,
-      "loss": 0.2443,
+      "epoch": 4.247306015064692,
+      "grad_norm": 0.31492358446121216,
+      "learning_rate": 2.924002948718685e-06,
+      "loss": 0.3828,
       "step": 117850
     },
     {
-      "epoch": 4.15,
-      "learning_rate": 3.736882871868877e-06,
-      "loss": 0.2538,
+      "epoch": 4.2474862147259165,
+      "grad_norm": 0.3145192265510559,
+      "learning_rate": 2.922633624969501e-06,
+      "loss": 0.3752,
       "step": 117855
     },
     {
-      "epoch": 4.15,
-      "learning_rate": 3.7353847673045255e-06,
-      "loss": 0.2542,
+      "epoch": 4.247666414387141,
+      "grad_norm": 0.26329880952835083,
+      "learning_rate": 2.9212646020201602e-06,
+      "loss": 0.3715,
       "step": 117860
     },
     {
-      "epoch": 4.15,
-      "learning_rate": 3.7338869388471975e-06,
-      "loss": 0.2587,
+      "epoch": 4.247846614048366,
+      "grad_norm": 0.22250141203403473,
+      "learning_rate": 2.9198958798893115e-06,
+      "loss": 0.403,
       "step": 117865
     },
     {
-      "epoch": 4.15,
-      "learning_rate": 3.732389386516333e-06,
-      "loss": 0.2654,
+      "epoch": 4.24802681370959,
+      "grad_norm": 0.25044476985931396,
+      "learning_rate": 2.9185274585956102e-06,
+      "loss": 0.368,
       "step": 117870
     },
     {
-      "epoch": 4.15,
-      "learning_rate": 3.7308921103313742e-06,
-      "loss": 0.253,
+      "epoch": 4.2482070133708145,
+      "grad_norm": 0.22099538147449493,
+      "learning_rate": 2.9171593381576962e-06,
+      "loss": 0.3449,
       "step": 117875
     },
     {
-      "epoch": 4.15,
-      "learning_rate": 3.729395110311759e-06,
-      "loss": 0.2443,
+      "epoch": 4.248387213032039,
+      "grad_norm": 0.26484134793281555,
+      "learning_rate": 2.915791518594213e-06,
+      "loss": 0.3742,
       "step": 117880
     },
     {
-      "epoch": 4.15,
-      "learning_rate": 3.7278983864769386e-06,
-      "loss": 0.2693,
+      "epoch": 4.248567412693264,
+      "grad_norm": 0.2437736839056015,
+      "learning_rate": 2.914423999923793e-06,
+      "loss": 0.3779,
       "step": 117885
     },
     {
-      "epoch": 4.15,
-      "learning_rate": 3.72640193884633e-06,
-      "loss": 0.2623,
+      "epoch": 4.248747612354489,
+      "grad_norm": 0.2136683613061905,
+      "learning_rate": 2.9130567821650717e-06,
+      "loss": 0.3602,
       "step": 117890
     },
     {
-      "epoch": 4.15,
-      "learning_rate": 3.7249057674393797e-06,
-      "loss": 0.251,
+      "epoch": 4.2489278120157135,
+      "grad_norm": 0.2902207374572754,
+      "learning_rate": 2.9116898653366697e-06,
+      "loss": 0.3767,
       "step": 117895
     },
     {
-      "epoch": 4.15,
-      "learning_rate": 3.7234098722755017e-06,
-      "loss": 0.2495,
+      "epoch": 4.249108011676938,
+      "grad_norm": 0.263248473405838,
+      "learning_rate": 2.9103232494572198e-06,
+      "loss": 0.3361,
       "step": 117900
     },
     {
-      "epoch": 4.15,
-      "learning_rate": 3.7219142533741288e-06,
-      "loss": 0.2352,
+      "epoch": 4.249288211338163,
+      "grad_norm": 0.31815701723098755,
+      "learning_rate": 2.9089569345453425e-06,
+      "loss": 0.3547,
       "step": 117905
     },
     {
-      "epoch": 4.15,
-      "learning_rate": 3.7204189107546777e-06,
-      "loss": 0.2399,
+      "epoch": 4.249468410999388,
+      "grad_norm": 0.27344661951065063,
+      "learning_rate": 2.907590920619649e-06,
+      "loss": 0.3585,
       "step": 117910
     },
     {
-      "epoch": 4.15,
-      "learning_rate": 3.718923844436559e-06,
-      "loss": 0.2328,
+      "epoch": 4.249648610660612,
+      "grad_norm": 0.25131598114967346,
+      "learning_rate": 2.9062252076987512e-06,
+      "loss": 0.3798,
       "step": 117915
     },
     {
-      "epoch": 4.15,
-      "learning_rate": 3.717429054439195e-06,
-      "loss": 0.247,
+      "epoch": 4.249828810321836,
+      "grad_norm": 0.24895480275154114,
+      "learning_rate": 2.9048597958012515e-06,
+      "loss": 0.3788,
       "step": 117920
     },
     {
-      "epoch": 4.15,
-      "learning_rate": 3.7159345407819874e-06,
-      "loss": 0.2389,
+      "epoch": 4.250009009983061,
+      "grad_norm": 0.2820626497268677,
+      "learning_rate": 2.9034946849457707e-06,
+      "loss": 0.3905,
       "step": 117925
     },
     {
-      "epoch": 4.15,
-      "learning_rate": 3.714440303484343e-06,
-      "loss": 0.2324,
+      "epoch": 4.250189209644286,
+      "grad_norm": 0.24140560626983643,
+      "learning_rate": 2.9021298751508856e-06,
+      "loss": 0.3902,
       "step": 117930
     },
     {
-      "epoch": 4.15,
-      "learning_rate": 3.712946342565657e-06,
-      "loss": 0.2551,
+      "epoch": 4.25036940930551,
+      "grad_norm": 0.19780555367469788,
+      "learning_rate": 2.9007653664352095e-06,
+      "loss": 0.3747,
       "step": 117935
     },
     {
-      "epoch": 4.15,
-      "learning_rate": 3.7114526580453416e-06,
-      "loss": 0.245,
+      "epoch": 4.250549608966735,
+      "grad_norm": 0.23243926465511322,
+      "learning_rate": 2.899401158817325e-06,
+      "loss": 0.3512,
       "step": 117940
     },
     {
-      "epoch": 4.15,
-      "learning_rate": 3.7099592499427764e-06,
-      "loss": 0.2531,
+      "epoch": 4.25072980862796,
+      "grad_norm": 0.19363261759281158,
+      "learning_rate": 2.898037252315822e-06,
+      "loss": 0.3274,
       "step": 117945
     },
     {
-      "epoch": 4.15,
-      "learning_rate": 3.708466118277368e-06,
-      "loss": 0.2384,
+      "epoch": 4.250910008289185,
+      "grad_norm": 0.22952446341514587,
+      "learning_rate": 2.8966736469492826e-06,
+      "loss": 0.3882,
       "step": 117950
     },
     {
-      "epoch": 4.15,
-      "learning_rate": 3.706973263068489e-06,
-      "loss": 0.2494,
+      "epoch": 4.251090207950409,
+      "grad_norm": 0.2463994026184082,
+      "learning_rate": 2.8953103427362847e-06,
+      "loss": 0.3941,
       "step": 117955
     },
     {
-      "epoch": 4.15,
-      "learning_rate": 3.7054806843355328e-06,
-      "loss": 0.2568,
+      "epoch": 4.251270407611634,
+      "grad_norm": 0.2101328819990158,
+      "learning_rate": 2.893947339695399e-06,
+      "loss": 0.3343,
       "step": 117960
     },
     {
-      "epoch": 4.15,
-      "learning_rate": 3.703988382097878e-06,
-      "loss": 0.2533,
+      "epoch": 4.251450607272858,
+      "grad_norm": 0.22500717639923096,
+      "learning_rate": 2.8925846378452056e-06,
+      "loss": 0.3845,
       "step": 117965
     },
     {
-      "epoch": 4.15,
-      "learning_rate": 3.702496356374899e-06,
-      "loss": 0.2504,
+      "epoch": 4.251630806934083,
+      "grad_norm": 0.2579551935195923,
+      "learning_rate": 2.8912222372042663e-06,
+      "loss": 0.3704,
       "step": 117970
     },
     {
-      "epoch": 4.15,
-      "learning_rate": 3.701004607185965e-06,
-      "loss": 0.2526,
+      "epoch": 4.251811006595307,
+      "grad_norm": 0.28678232431411743,
+      "learning_rate": 2.889860137791145e-06,
+      "loss": 0.3729,
       "step": 117975
     },
     {
-      "epoch": 4.15,
-      "learning_rate": 3.699513134550453e-06,
-      "loss": 0.2547,
+      "epoch": 4.251991206256532,
+      "grad_norm": 0.24297846853733063,
+      "learning_rate": 2.8884983396243988e-06,
+      "loss": 0.3957,
       "step": 117980
     },
     {
-      "epoch": 4.15,
-      "learning_rate": 3.698021938487728e-06,
-      "loss": 0.2365,
+      "epoch": 4.252171405917757,
+      "grad_norm": 0.2564467191696167,
+      "learning_rate": 2.8871368427225736e-06,
+      "loss": 0.3645,
       "step": 117985
     },
     {
-      "epoch": 4.15,
-      "learning_rate": 3.696531019017149e-06,
-      "loss": 0.2438,
+      "epoch": 4.252351605578982,
+      "grad_norm": 0.21953606605529785,
+      "learning_rate": 2.8857756471042357e-06,
+      "loss": 0.3874,
       "step": 117990
     },
     {
-      "epoch": 4.15,
-      "learning_rate": 3.6950403761580683e-06,
-      "loss": 0.2765,
+      "epoch": 4.252531805240206,
+      "grad_norm": 0.249432772397995,
+      "learning_rate": 2.8844147527879283e-06,
+      "loss": 0.4071,
       "step": 117995
     },
     {
-      "epoch": 4.15,
-      "learning_rate": 3.6935500099298524e-06,
-      "loss": 0.2527,
+      "epoch": 4.252712004901431,
+      "grad_norm": 0.2628020942211151,
+      "learning_rate": 2.8830541597921755e-06,
+      "loss": 0.3869,
       "step": 118000
     },
     {
-      "epoch": 4.15,
-      "eval_loss": 0.24940206110477448,
-      "eval_runtime": 10.5669,
-      "eval_samples_per_second": 9.464,
-      "eval_steps_per_second": 9.464,
+      "epoch": 4.252712004901431,
+      "eval_loss": 0.42919114232063293,
+      "eval_runtime": 3.5338,
+      "eval_samples_per_second": 28.298,
+      "eval_steps_per_second": 7.075,
       "step": 118000
     },
     {
-      "epoch": 4.15,
-      "learning_rate": 3.6920599203518453e-06,
-      "loss": 0.2677,
+      "epoch": 4.252892204562656,
+      "grad_norm": 0.2826399505138397,
+      "learning_rate": 2.881693868135535e-06,
+      "loss": 0.3887,
       "step": 118005
     },
     {
-      "epoch": 4.15,
-      "learning_rate": 3.690570107443403e-06,
-      "loss": 0.2606,
+      "epoch": 4.25307240422388,
+      "grad_norm": 0.24618491530418396,
+      "learning_rate": 2.8803338778365273e-06,
+      "loss": 0.3729,
       "step": 118010
     },
     {
-      "epoch": 4.15,
-      "learning_rate": 3.6890805712238576e-06,
-      "loss": 0.2524,
+      "epoch": 4.253252603885104,
+      "grad_norm": 0.2526216208934784,
+      "learning_rate": 2.8789741889136994e-06,
+      "loss": 0.4333,
       "step": 118015
     },
     {
-      "epoch": 4.15,
-      "learning_rate": 3.6875913117125626e-06,
-      "loss": 0.2526,
+      "epoch": 4.253432803546329,
+      "grad_norm": 0.3000968098640442,
+      "learning_rate": 2.877614801385559e-06,
+      "loss": 0.3544,
       "step": 118020
     },
     {
-      "epoch": 4.15,
-      "learning_rate": 3.6861023289288476e-06,
-      "loss": 0.2534,
+      "epoch": 4.253613003207554,
+      "grad_norm": 0.23581233620643616,
+      "learning_rate": 2.8762557152706286e-06,
+      "loss": 0.3761,
       "step": 118025
     },
     {
-      "epoch": 4.15,
-      "learning_rate": 3.6846136228920497e-06,
-      "loss": 0.2687,
+      "epoch": 4.253793202868779,
+      "grad_norm": 0.2421904057264328,
+      "learning_rate": 2.8748969305874367e-06,
+      "loss": 0.3642,
       "step": 118030
     },
     {
-      "epoch": 4.15,
-      "learning_rate": 3.6831251936214894e-06,
-      "loss": 0.2521,
+      "epoch": 4.253973402530003,
+      "grad_norm": 0.20584893226623535,
+      "learning_rate": 2.87353844735449e-06,
+      "loss": 0.3414,
       "step": 118035
     },
     {
-      "epoch": 4.15,
-      "learning_rate": 3.681637041136507e-06,
-      "loss": 0.2366,
+      "epoch": 4.254153602191228,
+      "grad_norm": 0.23724088072776794,
+      "learning_rate": 2.872180265590299e-06,
+      "loss": 0.3598,
       "step": 118040
     },
     {
-      "epoch": 4.15,
-      "learning_rate": 3.6801491654564174e-06,
-      "loss": 0.2429,
+      "epoch": 4.254333801852453,
+      "grad_norm": 0.23216982185840607,
+      "learning_rate": 2.870822385313368e-06,
+      "loss": 0.3497,
       "step": 118045
     },
     {
-      "epoch": 4.15,
-      "learning_rate": 3.678661566600533e-06,
-      "loss": 0.2454,
+      "epoch": 4.2545140015136775,
+      "grad_norm": 0.27174580097198486,
+      "learning_rate": 2.8694648065421907e-06,
+      "loss": 0.3824,
       "step": 118050
     },
     {
-      "epoch": 4.15,
-      "learning_rate": 3.677174244588186e-06,
-      "loss": 0.2515,
+      "epoch": 4.254694201174901,
+      "grad_norm": 0.31109705567359924,
+      "learning_rate": 2.8681075292952776e-06,
+      "loss": 0.3833,
       "step": 118055
     },
     {
-      "epoch": 4.15,
-      "learning_rate": 3.675687199438671e-06,
-      "loss": 0.2501,
+      "epoch": 4.254874400836126,
+      "grad_norm": 0.2299165278673172,
+      "learning_rate": 2.866750553591116e-06,
+      "loss": 0.3445,
       "step": 118060
     },
     {
-      "epoch": 4.15,
-      "learning_rate": 3.6742004311713123e-06,
-      "loss": 0.2654,
+      "epoch": 4.255054600497351,
+      "grad_norm": 0.30145031213760376,
+      "learning_rate": 2.865393879448189e-06,
+      "loss": 0.3773,
       "step": 118065
     },
     {
-      "epoch": 4.15,
-      "learning_rate": 3.6727139398054e-06,
-      "loss": 0.2568,
+      "epoch": 4.255234800158576,
+      "grad_norm": 0.22781316936016083,
+      "learning_rate": 2.8640375068849867e-06,
+      "loss": 0.3582,
       "step": 118070
     },
     {
-      "epoch": 4.15,
-      "learning_rate": 3.671227725360249e-06,
-      "loss": 0.2438,
+      "epoch": 4.2554149998198,
+      "grad_norm": 0.27182450890541077,
+      "learning_rate": 2.8626814359199894e-06,
+      "loss": 0.3855,
       "step": 118075
     },
     {
-      "epoch": 4.15,
-      "learning_rate": 3.6697417878551465e-06,
-      "loss": 0.2377,
+      "epoch": 4.255595199481025,
+      "grad_norm": 0.2907634377479553,
+      "learning_rate": 2.8613256665716653e-06,
+      "loss": 0.37,
       "step": 118080
     },
     {
-      "epoch": 4.15,
-      "learning_rate": 3.6682561273093914e-06,
-      "loss": 0.2439,
+      "epoch": 4.25577539914225,
+      "grad_norm": 0.2272307574748993,
+      "learning_rate": 2.859970198858497e-06,
+      "loss": 0.3706,
       "step": 118085
     },
     {
-      "epoch": 4.15,
-      "learning_rate": 3.6667707437422677e-06,
-      "loss": 0.2687,
+      "epoch": 4.2559555988034745,
+      "grad_norm": 0.20825766026973724,
+      "learning_rate": 2.85861503279895e-06,
+      "loss": 0.3394,
       "step": 118090
     },
     {
-      "epoch": 4.15,
-      "learning_rate": 3.6652856371730715e-06,
-      "loss": 0.2454,
+      "epoch": 4.256135798464699,
+      "grad_norm": 0.21996937692165375,
+      "learning_rate": 2.8572601684114853e-06,
+      "loss": 0.3639,
       "step": 118095
     },
     {
-      "epoch": 4.16,
-      "learning_rate": 3.663800807621082e-06,
-      "loss": 0.2303,
+      "epoch": 4.256315998125924,
+      "grad_norm": 0.23290881514549255,
+      "learning_rate": 2.855905605714565e-06,
+      "loss": 0.3777,
       "step": 118100
     },
     {
-      "epoch": 4.16,
-      "learning_rate": 3.6623162551055774e-06,
-      "loss": 0.2616,
+      "epoch": 4.256496197787148,
+      "grad_norm": 0.21647751331329346,
+      "learning_rate": 2.854551344726636e-06,
+      "loss": 0.3746,
       "step": 118105
     },
     {
-      "epoch": 4.16,
-      "learning_rate": 3.660831979645829e-06,
-      "loss": 0.2408,
+      "epoch": 4.256676397448373,
+      "grad_norm": 0.2545035183429718,
+      "learning_rate": 2.853197385466169e-06,
+      "loss": 0.3716,
       "step": 118110
     },
     {
-      "epoch": 4.16,
-      "learning_rate": 3.6593479812611133e-06,
-      "loss": 0.2452,
+      "epoch": 4.256856597109597,
+      "grad_norm": 0.20453687012195587,
+      "learning_rate": 2.851843727951595e-06,
+      "loss": 0.3852,
       "step": 118115
     },
     {
-      "epoch": 4.16,
-      "learning_rate": 3.657864259970706e-06,
-      "loss": 0.2708,
+      "epoch": 4.257036796770822,
+      "grad_norm": 0.24812854826450348,
+      "learning_rate": 2.8504903722013587e-06,
+      "loss": 0.3759,
       "step": 118120
     },
     {
-      "epoch": 4.16,
-      "learning_rate": 3.6563808157938666e-06,
-      "loss": 0.2449,
+      "epoch": 4.257216996432047,
+      "grad_norm": 0.29127225279808044,
+      "learning_rate": 2.8491373182339048e-06,
+      "loss": 0.3276,
       "step": 118125
     },
     {
-      "epoch": 4.16,
-      "learning_rate": 3.6548976487498498e-06,
-      "loss": 0.2404,
+      "epoch": 4.2573971960932715,
+      "grad_norm": 0.2810105085372925,
+      "learning_rate": 2.8477845660676684e-06,
+      "loss": 0.4026,
       "step": 118130
     },
     {
-      "epoch": 4.16,
-      "learning_rate": 3.6534147588579258e-06,
-      "loss": 0.257,
+      "epoch": 4.257577395754496,
+      "grad_norm": 0.29022449254989624,
+      "learning_rate": 2.846432115721079e-06,
+      "loss": 0.3805,
       "step": 118135
     },
     {
-      "epoch": 4.16,
-      "learning_rate": 3.651932146137346e-06,
-      "loss": 0.2464,
+      "epoch": 4.257757595415721,
+      "grad_norm": 0.25384798645973206,
+      "learning_rate": 2.8450799672125607e-06,
+      "loss": 0.361,
       "step": 118140
     },
     {
-      "epoch": 4.16,
-      "learning_rate": 3.6504498106073564e-06,
-      "loss": 0.2475,
+      "epoch": 4.257937795076945,
+      "grad_norm": 0.23675154149532318,
+      "learning_rate": 2.843728120560535e-06,
+      "loss": 0.3651,
       "step": 118145
     },
     {
-      "epoch": 4.16,
-      "learning_rate": 3.6489677522871997e-06,
-      "loss": 0.259,
+      "epoch": 4.2581179947381695,
+      "grad_norm": 0.20214100182056427,
+      "learning_rate": 2.8423765757834287e-06,
+      "loss": 0.3901,
       "step": 118150
     },
     {
-      "epoch": 4.16,
-      "learning_rate": 3.6474859711961324e-06,
-      "loss": 0.2459,
+      "epoch": 4.258298194399394,
+      "grad_norm": 0.2030409872531891,
+      "learning_rate": 2.8410253328996523e-06,
+      "loss": 0.3818,
       "step": 118155
     },
     {
-      "epoch": 4.16,
-      "learning_rate": 3.6460044673533865e-06,
-      "loss": 0.2547,
+      "epoch": 4.258478394060619,
+      "grad_norm": 0.2594359815120697,
+      "learning_rate": 2.839674391927613e-06,
+      "loss": 0.3604,
       "step": 118160
     },
     {
-      "epoch": 4.16,
-      "learning_rate": 3.644523240778197e-06,
-      "loss": 0.2544,
+      "epoch": 4.258658593721844,
+      "grad_norm": 0.2291530966758728,
+      "learning_rate": 2.8383237528857243e-06,
+      "loss": 0.3353,
       "step": 118165
     },
     {
-      "epoch": 4.16,
-      "learning_rate": 3.6430422914897984e-06,
-      "loss": 0.2793,
+      "epoch": 4.2588387933830685,
+      "grad_norm": 0.23758092522621155,
+      "learning_rate": 2.836973415792374e-06,
+      "loss": 0.3764,
       "step": 118170
     },
     {
-      "epoch": 4.16,
-      "learning_rate": 3.6415616195074246e-06,
-      "loss": 0.2845,
+      "epoch": 4.259018993044293,
+      "grad_norm": 0.27188360691070557,
+      "learning_rate": 2.8356233806659777e-06,
+      "loss": 0.3948,
       "step": 118175
     },
     {
-      "epoch": 4.16,
-      "learning_rate": 3.6400812248502976e-06,
-      "loss": 0.2606,
+      "epoch": 4.259199192705518,
+      "grad_norm": 0.21040531992912292,
+      "learning_rate": 2.834273647524921e-06,
+      "loss": 0.3403,
       "step": 118180
     },
     {
-      "epoch": 4.16,
-      "learning_rate": 3.638601107537637e-06,
-      "loss": 0.2469,
+      "epoch": 4.259379392366743,
+      "grad_norm": 0.28796496987342834,
+      "learning_rate": 2.832924216387595e-06,
+      "loss": 0.4118,
       "step": 118185
     },
     {
-      "epoch": 4.16,
-      "learning_rate": 3.6371212675886668e-06,
-      "loss": 0.2576,
+      "epoch": 4.259559592027967,
+      "grad_norm": 0.23446249961853027,
+      "learning_rate": 2.831575087272387e-06,
+      "loss": 0.3733,
       "step": 118190
     },
     {
-      "epoch": 4.16,
-      "learning_rate": 3.6356417050225967e-06,
-      "loss": 0.2511,
+      "epoch": 4.259739791689191,
+      "grad_norm": 0.28546759486198425,
+      "learning_rate": 2.8302262601976687e-06,
+      "loss": 0.3581,
       "step": 118195
     },
     {
-      "epoch": 4.16,
-      "learning_rate": 3.6341624198586416e-06,
-      "loss": 0.264,
+      "epoch": 4.259919991350416,
+      "grad_norm": 0.1949034482240677,
+      "learning_rate": 2.8288777351818287e-06,
+      "loss": 0.3575,
       "step": 118200
     },
     {
-      "epoch": 4.16,
-      "learning_rate": 3.6326834121160002e-06,
-      "loss": 0.2621,
+      "epoch": 4.260100191011641,
+      "grad_norm": 0.28568407893180847,
+      "learning_rate": 2.827529512243246e-06,
+      "loss": 0.3726,
       "step": 118205
     },
     {
-      "epoch": 4.16,
-      "learning_rate": 3.6312046818138877e-06,
-      "loss": 0.2457,
+      "epoch": 4.2602803906728655,
+      "grad_norm": 0.24735015630722046,
+      "learning_rate": 2.8261815914002675e-06,
+      "loss": 0.376,
       "step": 118210
     },
     {
-      "epoch": 4.16,
-      "learning_rate": 3.6297262289715028e-06,
-      "loss": 0.2514,
+      "epoch": 4.26046059033409,
+      "grad_norm": 0.18040810525417328,
+      "learning_rate": 2.8248339726712787e-06,
+      "loss": 0.3756,
       "step": 118215
     },
     {
-      "epoch": 4.16,
-      "learning_rate": 3.62824805360803e-06,
-      "loss": 0.269,
+      "epoch": 4.260640789995315,
+      "grad_norm": 0.20753014087677002,
+      "learning_rate": 2.823486656074634e-06,
+      "loss": 0.3616,
       "step": 118220
     },
     {
-      "epoch": 4.16,
-      "learning_rate": 3.626770155742676e-06,
-      "loss": 0.2475,
+      "epoch": 4.26082098965654,
+      "grad_norm": 0.21727077662944794,
+      "learning_rate": 2.8221396416286905e-06,
+      "loss": 0.3591,
       "step": 118225
     },
     {
-      "epoch": 4.16,
-      "learning_rate": 3.6252925353946284e-06,
-      "loss": 0.2417,
+      "epoch": 4.261001189317764,
+      "grad_norm": 0.24121779203414917,
+      "learning_rate": 2.820792929351798e-06,
+      "loss": 0.3535,
       "step": 118230
     },
     {
-      "epoch": 4.16,
-      "learning_rate": 3.6238151925830716e-06,
-      "loss": 0.2555,
+      "epoch": 4.261181388978989,
+      "grad_norm": 0.22880592942237854,
+      "learning_rate": 2.819446519262306e-06,
+      "loss": 0.3734,
       "step": 118235
     },
     {
-      "epoch": 4.16,
-      "learning_rate": 3.6223381273271876e-06,
-      "loss": 0.2482,
+      "epoch": 4.261361588640213,
+      "grad_norm": 0.26039087772369385,
+      "learning_rate": 2.8181004113785632e-06,
+      "loss": 0.3805,
       "step": 118240
     },
     {
-      "epoch": 4.16,
-      "learning_rate": 3.6208613396461476e-06,
-      "loss": 0.2454,
+      "epoch": 4.261541788301438,
+      "grad_norm": 0.21298900246620178,
+      "learning_rate": 2.81675460571891e-06,
+      "loss": 0.3944,
       "step": 118245
     },
     {
-      "epoch": 4.16,
-      "learning_rate": 3.6193848295591416e-06,
-      "loss": 0.2589,
+      "epoch": 4.261721987962662,
+      "grad_norm": 0.24916015565395355,
+      "learning_rate": 2.8154091023016805e-06,
+      "loss": 0.3752,
       "step": 118250
     },
     {
-      "epoch": 4.16,
-      "learning_rate": 3.6179085970853317e-06,
-      "loss": 0.2389,
+      "epoch": 4.261902187623887,
+      "grad_norm": 0.31941652297973633,
+      "learning_rate": 2.8140639011452057e-06,
+      "loss": 0.3874,
       "step": 118255
     },
     {
-      "epoch": 4.16,
-      "learning_rate": 3.6164326422438886e-06,
-      "loss": 0.2575,
+      "epoch": 4.262082387285112,
+      "grad_norm": 0.2693467140197754,
+      "learning_rate": 2.8127190022678135e-06,
+      "loss": 0.3512,
       "step": 118260
     },
     {
-      "epoch": 4.16,
-      "learning_rate": 3.614956965053967e-06,
-      "loss": 0.2208,
+      "epoch": 4.262262586946337,
+      "grad_norm": 0.28551846742630005,
+      "learning_rate": 2.8113744056878277e-06,
+      "loss": 0.3449,
       "step": 118265
     },
     {
-      "epoch": 4.16,
-      "learning_rate": 3.613481565534743e-06,
-      "loss": 0.2806,
+      "epoch": 4.262442786607561,
+      "grad_norm": 0.24838872253894806,
+      "learning_rate": 2.8100301114235726e-06,
+      "loss": 0.354,
       "step": 118270
     },
     {
-      "epoch": 4.16,
-      "learning_rate": 3.6120064437053625e-06,
-      "loss": 0.2387,
+      "epoch": 4.262622986268786,
+      "grad_norm": 0.20312124490737915,
+      "learning_rate": 2.8086861194933616e-06,
+      "loss": 0.3844,
       "step": 118275
     },
     {
-      "epoch": 4.16,
-      "learning_rate": 3.6105315995849876e-06,
-      "loss": 0.248,
+      "epoch": 4.262803185930011,
+      "grad_norm": 0.29347795248031616,
+      "learning_rate": 2.807342429915505e-06,
+      "loss": 0.3986,
       "step": 118280
     },
     {
-      "epoch": 4.16,
-      "learning_rate": 3.6090570331927586e-06,
-      "loss": 0.2652,
+      "epoch": 4.262983385591235,
+      "grad_norm": 0.23310750722885132,
+      "learning_rate": 2.8059990427083127e-06,
+      "loss": 0.3745,
       "step": 118285
     },
     {
-      "epoch": 4.16,
-      "learning_rate": 3.60758274454783e-06,
-      "loss": 0.2443,
+      "epoch": 4.263163585252459,
+      "grad_norm": 0.2755783200263977,
+      "learning_rate": 2.8046559578900815e-06,
+      "loss": 0.3808,
       "step": 118290
     },
     {
-      "epoch": 4.16,
-      "learning_rate": 3.6061087336693444e-06,
-      "loss": 0.2452,
+      "epoch": 4.263343784913684,
+      "grad_norm": 0.21833378076553345,
+      "learning_rate": 2.8033131754791246e-06,
+      "loss": 0.3937,
       "step": 118295
     },
     {
-      "epoch": 4.16,
-      "learning_rate": 3.6046350005764314e-06,
-      "loss": 0.2472,
+      "epoch": 4.263523984574909,
+      "grad_norm": 0.2588220536708832,
+      "learning_rate": 2.801970695493722e-06,
+      "loss": 0.372,
       "step": 118300
     },
     {
-      "epoch": 4.16,
-      "learning_rate": 3.6031615452882388e-06,
-      "loss": 0.2673,
+      "epoch": 4.263704184236134,
+      "grad_norm": 0.23471491038799286,
+      "learning_rate": 2.8006285179521696e-06,
+      "loss": 0.3778,
       "step": 118305
     },
     {
-      "epoch": 4.16,
-      "learning_rate": 3.601688367823891e-06,
-      "loss": 0.2678,
+      "epoch": 4.263884383897358,
+      "grad_norm": 0.31038230657577515,
+      "learning_rate": 2.7992866428727587e-06,
+      "loss": 0.3946,
       "step": 118310
     },
     {
-      "epoch": 4.16,
-      "learning_rate": 3.6002154682025167e-06,
-      "loss": 0.2513,
+      "epoch": 4.264064583558583,
+      "grad_norm": 0.2159845381975174,
+      "learning_rate": 2.7979450702737637e-06,
+      "loss": 0.3621,
       "step": 118315
     },
     {
-      "epoch": 4.16,
-      "learning_rate": 3.59874284644324e-06,
-      "loss": 0.2497,
+      "epoch": 4.264244783219808,
+      "grad_norm": 0.1845831722021103,
+      "learning_rate": 2.7966038001734777e-06,
+      "loss": 0.3498,
       "step": 118320
     },
     {
-      "epoch": 4.16,
-      "learning_rate": 3.5972705025651863e-06,
-      "loss": 0.2727,
+      "epoch": 4.2644249828810326,
+      "grad_norm": 0.21577754616737366,
+      "learning_rate": 2.7952628325901647e-06,
+      "loss": 0.3733,
       "step": 118325
     },
     {
-      "epoch": 4.16,
-      "learning_rate": 3.5957984365874637e-06,
-      "loss": 0.2371,
+      "epoch": 4.264605182542256,
+      "grad_norm": 0.234993115067482,
+      "learning_rate": 2.793922167542087e-06,
+      "loss": 0.3823,
       "step": 118330
     },
     {
-      "epoch": 4.16,
-      "learning_rate": 3.5943266485292005e-06,
-      "loss": 0.2343,
+      "epoch": 4.264785382203481,
+      "grad_norm": 0.22045071423053741,
+      "learning_rate": 2.7925818050475284e-06,
+      "loss": 0.3653,
       "step": 118335
     },
     {
-      "epoch": 4.16,
-      "learning_rate": 3.5928551384094926e-06,
-      "loss": 0.2563,
+      "epoch": 4.264965581864706,
+      "grad_norm": 0.2358977198600769,
+      "learning_rate": 2.7912417451247453e-06,
+      "loss": 0.3503,
       "step": 118340
     },
     {
-      "epoch": 4.16,
-      "learning_rate": 3.5913839062474584e-06,
-      "loss": 0.2377,
+      "epoch": 4.265145781525931,
+      "grad_norm": 0.2254144698381424,
+      "learning_rate": 2.78990198779199e-06,
+      "loss": 0.3497,
       "step": 118345
     },
     {
-      "epoch": 4.16,
-      "learning_rate": 3.589912952062194e-06,
-      "loss": 0.2378,
+      "epoch": 4.265325981187155,
+      "grad_norm": 0.2657366693019867,
+      "learning_rate": 2.7885625330675206e-06,
+      "loss": 0.3555,
       "step": 118350
     },
     {
-      "epoch": 4.16,
-      "learning_rate": 3.5884422758728026e-06,
-      "loss": 0.28,
+      "epoch": 4.26550618084838,
+      "grad_norm": 0.21268507838249207,
+      "learning_rate": 2.7872233809695837e-06,
+      "loss": 0.366,
       "step": 118355
     },
     {
-      "epoch": 4.16,
-      "learning_rate": 3.5869718776983694e-06,
-      "loss": 0.2559,
+      "epoch": 4.265686380509605,
+      "grad_norm": 0.2723730504512787,
+      "learning_rate": 2.7858845315164307e-06,
+      "loss": 0.3865,
       "step": 118360
     },
     {
-      "epoch": 4.16,
-      "learning_rate": 3.5855017575580017e-06,
-      "loss": 0.252,
+      "epoch": 4.2658665801708295,
+      "grad_norm": 0.1851958930492401,
+      "learning_rate": 2.7845459847263e-06,
+      "loss": 0.3302,
       "step": 118365
     },
     {
-      "epoch": 4.16,
-      "learning_rate": 3.5840319154707815e-06,
-      "loss": 0.2521,
+      "epoch": 4.266046779832054,
+      "grad_norm": 0.30947184562683105,
+      "learning_rate": 2.78320774061743e-06,
+      "loss": 0.3843,
       "step": 118370
     },
     {
-      "epoch": 4.16,
-      "learning_rate": 3.582562351455793e-06,
-      "loss": 0.2526,
+      "epoch": 4.266226979493279,
+      "grad_norm": 0.2673863172531128,
+      "learning_rate": 2.7818697992080536e-06,
+      "loss": 0.3744,
       "step": 118375
     },
     {
-      "epoch": 4.16,
-      "learning_rate": 3.5810930655321125e-06,
-      "loss": 0.2539,
+      "epoch": 4.266407179154503,
+      "grad_norm": 0.22295500338077545,
+      "learning_rate": 2.780532160516394e-06,
+      "loss": 0.4008,
       "step": 118380
     },
     {
-      "epoch": 4.17,
-      "learning_rate": 3.5796240577188196e-06,
-      "loss": 0.2334,
+      "epoch": 4.266587378815728,
+      "grad_norm": 0.2774406969547272,
+      "learning_rate": 2.779194824560688e-06,
+      "loss": 0.3602,
       "step": 118385
     },
     {
-      "epoch": 4.17,
-      "learning_rate": 3.5781553280350016e-06,
-      "loss": 0.272,
+      "epoch": 4.266767578476952,
+      "grad_norm": 0.3125916123390198,
+      "learning_rate": 2.7778577913591534e-06,
+      "loss": 0.3976,
       "step": 118390
     },
     {
-      "epoch": 4.17,
-      "learning_rate": 3.5766868764997103e-06,
-      "loss": 0.2575,
+      "epoch": 4.266947778138177,
+      "grad_norm": 0.1951741874217987,
+      "learning_rate": 2.7765210609299958e-06,
+      "loss": 0.3666,
       "step": 118395
     },
     {
-      "epoch": 4.17,
-      "learning_rate": 3.57521870313203e-06,
-      "loss": 0.2603,
+      "epoch": 4.267127977799402,
+      "grad_norm": 0.2636171877384186,
+      "learning_rate": 2.775184633291439e-06,
+      "loss": 0.3816,
       "step": 118400
     },
     {
-      "epoch": 4.17,
-      "learning_rate": 3.5737508079510125e-06,
-      "loss": 0.2526,
+      "epoch": 4.2673081774606265,
+      "grad_norm": 0.25289151072502136,
+      "learning_rate": 2.7738485084616904e-06,
+      "loss": 0.383,
       "step": 118405
     },
     {
-      "epoch": 4.17,
-      "learning_rate": 3.5722831909757225e-06,
-      "loss": 0.267,
+      "epoch": 4.267488377121851,
+      "grad_norm": 0.23693720996379852,
+      "learning_rate": 2.7725126864589527e-06,
+      "loss": 0.3504,
       "step": 118410
     },
     {
-      "epoch": 4.17,
-      "learning_rate": 3.5708158522252067e-06,
-      "loss": 0.237,
+      "epoch": 4.267668576783076,
+      "grad_norm": 0.31816625595092773,
+      "learning_rate": 2.7711771673014253e-06,
+      "loss": 0.3926,
       "step": 118415
     },
     {
-      "epoch": 4.17,
-      "learning_rate": 3.5693487917185325e-06,
-      "loss": 0.2494,
+      "epoch": 4.2678487764443,
+      "grad_norm": 0.20201784372329712,
+      "learning_rate": 2.7698419510073014e-06,
+      "loss": 0.3648,
       "step": 118420
     },
     {
-      "epoch": 4.17,
-      "learning_rate": 3.567882009474738e-06,
-      "loss": 0.2606,
+      "epoch": 4.268028976105525,
+      "grad_norm": 0.28270724415779114,
+      "learning_rate": 2.768507037594781e-06,
+      "loss": 0.3847,
       "step": 118425
     },
     {
-      "epoch": 4.17,
-      "learning_rate": 3.566415505512874e-06,
-      "loss": 0.24,
+      "epoch": 4.268209175766749,
+      "grad_norm": 0.2311401665210724,
+      "learning_rate": 2.7671724270820487e-06,
+      "loss": 0.3913,
       "step": 118430
     },
     {
-      "epoch": 4.17,
-      "learning_rate": 3.564949279851973e-06,
-      "loss": 0.2432,
+      "epoch": 4.268389375427974,
+      "grad_norm": 0.2863185405731201,
+      "learning_rate": 2.765838119487288e-06,
+      "loss": 0.4017,
       "step": 118435
     },
     {
-      "epoch": 4.17,
-      "learning_rate": 3.5634833325110835e-06,
-      "loss": 0.2706,
+      "epoch": 4.268569575089199,
+      "grad_norm": 0.3267385959625244,
+      "learning_rate": 2.764504114828678e-06,
+      "loss": 0.3994,
       "step": 118440
     },
     {
-      "epoch": 4.17,
-      "learning_rate": 3.5620176635092317e-06,
-      "loss": 0.2564,
+      "epoch": 4.2687497747504235,
+      "grad_norm": 0.350425660610199,
+      "learning_rate": 2.7631704131243967e-06,
+      "loss": 0.3746,
       "step": 118445
     },
     {
-      "epoch": 4.17,
-      "learning_rate": 3.5605522728654584e-06,
-      "loss": 0.2298,
+      "epoch": 4.268929974411648,
+      "grad_norm": 0.2516806125640869,
+      "learning_rate": 2.761837014392604e-06,
+      "loss": 0.3326,
       "step": 118450
     },
     {
-      "epoch": 4.17,
-      "learning_rate": 3.5590871605987756e-06,
-      "loss": 0.2499,
+      "epoch": 4.269110174072873,
+      "grad_norm": 0.27395808696746826,
+      "learning_rate": 2.7605039186514858e-06,
+      "loss": 0.3911,
       "step": 118455
     },
     {
-      "epoch": 4.17,
-      "learning_rate": 3.5576223267282242e-06,
-      "loss": 0.2324,
+      "epoch": 4.269290373734098,
+      "grad_norm": 0.24805009365081787,
+      "learning_rate": 2.7591711259191938e-06,
+      "loss": 0.3596,
       "step": 118460
     },
     {
-      "epoch": 4.17,
-      "learning_rate": 3.5561577712728168e-06,
-      "loss": 0.2419,
+      "epoch": 4.269470573395322,
+      "grad_norm": 0.23260460793972015,
+      "learning_rate": 2.7578386362138886e-06,
+      "loss": 0.3581,
       "step": 118465
     },
     {
-      "epoch": 4.17,
-      "learning_rate": 3.554693494251568e-06,
-      "loss": 0.2381,
+      "epoch": 4.269650773056546,
+      "grad_norm": 0.24661628901958466,
+      "learning_rate": 2.756506449553725e-06,
+      "loss": 0.3633,
       "step": 118470
     },
     {
-      "epoch": 4.17,
-      "learning_rate": 3.5532294956834854e-06,
-      "loss": 0.2537,
+      "epoch": 4.269830972717771,
+      "grad_norm": 0.25693389773368835,
+      "learning_rate": 2.7551745659568527e-06,
+      "loss": 0.3823,
       "step": 118475
     },
     {
-      "epoch": 4.17,
-      "learning_rate": 3.55176577558759e-06,
-      "loss": 0.2592,
+      "epoch": 4.270011172378996,
+      "grad_norm": 0.2382129281759262,
+      "learning_rate": 2.7538429854414238e-06,
+      "loss": 0.3639,
       "step": 118480
     },
     {
-      "epoch": 4.17,
-      "learning_rate": 3.55030233398288e-06,
-      "loss": 0.245,
+      "epoch": 4.2701913720402205,
+      "grad_norm": 0.2255300134420395,
+      "learning_rate": 2.7525117080255817e-06,
+      "loss": 0.3625,
       "step": 118485
     },
     {
-      "epoch": 4.17,
-      "learning_rate": 3.548839170888352e-06,
-      "loss": 0.2571,
+      "epoch": 4.270371571701445,
+      "grad_norm": 0.2332393378019333,
+      "learning_rate": 2.7511807337274513e-06,
+      "loss": 0.3748,
       "step": 118490
     },
     {
-      "epoch": 4.17,
-      "learning_rate": 3.5473762863230176e-06,
-      "loss": 0.2119,
+      "epoch": 4.27055177136267,
+      "grad_norm": 0.18017563223838806,
+      "learning_rate": 2.749850062565179e-06,
+      "loss": 0.4011,
       "step": 118495
     },
     {
-      "epoch": 4.17,
-      "learning_rate": 3.5459136803058597e-06,
-      "loss": 0.2487,
+      "epoch": 4.270731971023895,
+      "grad_norm": 0.2834778428077698,
+      "learning_rate": 2.748519694556889e-06,
+      "loss": 0.3585,
       "step": 118500
     },
     {
-      "epoch": 4.17,
-      "eval_loss": 0.2493087649345398,
-      "eval_runtime": 10.5519,
-      "eval_samples_per_second": 9.477,
-      "eval_steps_per_second": 9.477,
+      "epoch": 4.270731971023895,
+      "eval_loss": 0.42914897203445435,
+      "eval_runtime": 3.5317,
+      "eval_samples_per_second": 28.315,
+      "eval_steps_per_second": 7.079,
       "step": 118500
     },
     {
-      "epoch": 4.17,
-      "learning_rate": 3.5444513528558766e-06,
-      "loss": 0.2779,
+      "epoch": 4.270912170685119,
+      "grad_norm": 0.2678777575492859,
+      "learning_rate": 2.747189629720717e-06,
+      "loss": 0.4119,
       "step": 118505
     },
     {
-      "epoch": 4.17,
-      "learning_rate": 3.54298930399205e-06,
-      "loss": 0.2735,
+      "epoch": 4.271092370346344,
+      "grad_norm": 0.23227082192897797,
+      "learning_rate": 2.7458598680747764e-06,
+      "loss": 0.3705,
       "step": 118510
     },
     {
-      "epoch": 4.17,
-      "learning_rate": 3.5415275337333714e-06,
-      "loss": 0.2558,
+      "epoch": 4.271272570007568,
+      "grad_norm": 0.26510533690452576,
+      "learning_rate": 2.7445304096371803e-06,
+      "loss": 0.3262,
       "step": 118515
     },
     {
-      "epoch": 4.17,
-      "learning_rate": 3.5400660420988136e-06,
-      "loss": 0.2297,
+      "epoch": 4.271452769668793,
+      "grad_norm": 0.22343812882900238,
+      "learning_rate": 2.7432012544260533e-06,
+      "loss": 0.3612,
       "step": 118520
     },
     {
-      "epoch": 4.17,
-      "learning_rate": 3.538604829107359e-06,
-      "loss": 0.2475,
+      "epoch": 4.2716329693300175,
+      "grad_norm": 0.2528297007083893,
+      "learning_rate": 2.7418724024595e-06,
+      "loss": 0.3779,
       "step": 118525
     },
     {
-      "epoch": 4.17,
-      "learning_rate": 3.5371438947779694e-06,
-      "loss": 0.2564,
+      "epoch": 4.271813168991242,
+      "grad_norm": 0.28484654426574707,
+      "learning_rate": 2.7405438537556284e-06,
+      "loss": 0.3164,
       "step": 118530
     },
     {
-      "epoch": 4.17,
-      "learning_rate": 3.535683239129628e-06,
-      "loss": 0.2665,
+      "epoch": 4.271993368652467,
+      "grad_norm": 0.2651745676994324,
+      "learning_rate": 2.739215608332535e-06,
+      "loss": 0.4037,
       "step": 118535
     },
     {
-      "epoch": 4.17,
-      "learning_rate": 3.534222862181297e-06,
-      "loss": 0.2458,
+      "epoch": 4.272173568313692,
+      "grad_norm": 0.2041269987821579,
+      "learning_rate": 2.737887666208314e-06,
+      "loss": 0.3386,
       "step": 118540
     },
     {
-      "epoch": 4.17,
-      "learning_rate": 3.532762763951933e-06,
-      "loss": 0.2273,
+      "epoch": 4.272353767974916,
+      "grad_norm": 0.22839802503585815,
+      "learning_rate": 2.7365600274010694e-06,
+      "loss": 0.3724,
       "step": 118545
     },
     {
-      "epoch": 4.17,
-      "learning_rate": 3.531302944460491e-06,
-      "loss": 0.2514,
+      "epoch": 4.272533967636141,
+      "grad_norm": 0.2205089032649994,
+      "learning_rate": 2.7352326919288824e-06,
+      "loss": 0.3454,
       "step": 118550
     },
     {
-      "epoch": 4.17,
-      "learning_rate": 3.5298434037259327e-06,
-      "loss": 0.2274,
+      "epoch": 4.272714167297366,
+      "grad_norm": 0.25789710879325867,
+      "learning_rate": 2.7339056598098407e-06,
+      "loss": 0.3494,
       "step": 118555
     },
     {
-      "epoch": 4.17,
-      "learning_rate": 3.528384141767213e-06,
-      "loss": 0.2526,
+      "epoch": 4.27289436695859,
+      "grad_norm": 0.3210128843784332,
+      "learning_rate": 2.7325789310620212e-06,
+      "loss": 0.3931,
       "step": 118560
     },
     {
-      "epoch": 4.17,
-      "learning_rate": 3.526925158603275e-06,
-      "loss": 0.2429,
+      "epoch": 4.273074566619814,
+      "grad_norm": 0.24505449831485748,
+      "learning_rate": 2.731252505703502e-06,
+      "loss": 0.3628,
       "step": 118565
     },
     {
-      "epoch": 4.17,
-      "learning_rate": 3.5254664542530596e-06,
-      "loss": 0.2546,
+      "epoch": 4.273254766281039,
+      "grad_norm": 0.2696022689342499,
+      "learning_rate": 2.729926383752357e-06,
+      "loss": 0.3555,
       "step": 118570
     },
     {
-      "epoch": 4.17,
-      "learning_rate": 3.524008028735512e-06,
-      "loss": 0.2668,
+      "epoch": 4.273434965942264,
+      "grad_norm": 0.23730942606925964,
+      "learning_rate": 2.7286005652266604e-06,
+      "loss": 0.4053,
       "step": 118575
     },
     {
-      "epoch": 4.17,
-      "learning_rate": 3.5225498820695703e-06,
-      "loss": 0.2632,
+      "epoch": 4.273615165603489,
+      "grad_norm": 0.2668841779232025,
+      "learning_rate": 2.7272750501444594e-06,
+      "loss": 0.38,
       "step": 118580
     },
     {
-      "epoch": 4.17,
-      "learning_rate": 3.5210920142741605e-06,
-      "loss": 0.2648,
+      "epoch": 4.273795365264713,
+      "grad_norm": 0.23594842851161957,
+      "learning_rate": 2.7259498385238282e-06,
+      "loss": 0.3971,
       "step": 118585
     },
     {
-      "epoch": 4.17,
-      "learning_rate": 3.5196344253682124e-06,
-      "loss": 0.2559,
+      "epoch": 4.273975564925938,
+      "grad_norm": 0.2008124589920044,
+      "learning_rate": 2.724624930382819e-06,
+      "loss": 0.3583,
       "step": 118590
     },
     {
-      "epoch": 4.17,
-      "learning_rate": 3.5181771153706605e-06,
-      "loss": 0.2444,
+      "epoch": 4.274155764587163,
+      "grad_norm": 0.21719537675380707,
+      "learning_rate": 2.723300325739475e-06,
+      "loss": 0.3669,
       "step": 118595
     },
     {
-      "epoch": 4.17,
-      "learning_rate": 3.5167200843004233e-06,
-      "loss": 0.2436,
+      "epoch": 4.274335964248388,
+      "grad_norm": 0.25217390060424805,
+      "learning_rate": 2.7219760246118638e-06,
+      "loss": 0.3523,
       "step": 118600
     },
     {
-      "epoch": 4.17,
-      "learning_rate": 3.5152633321764107e-06,
-      "loss": 0.259,
+      "epoch": 4.274516163909611,
+      "grad_norm": 0.19717147946357727,
+      "learning_rate": 2.720652027018006e-06,
+      "loss": 0.3957,
       "step": 118605
     },
     {
-      "epoch": 4.17,
-      "learning_rate": 3.5138068590175435e-06,
-      "loss": 0.2611,
+      "epoch": 4.274696363570836,
+      "grad_norm": 0.2760673761367798,
+      "learning_rate": 2.7193283329759537e-06,
+      "loss": 0.3723,
       "step": 118610
     },
     {
-      "epoch": 4.17,
-      "learning_rate": 3.5123506648427425e-06,
-      "loss": 0.2542,
+      "epoch": 4.274876563232061,
+      "grad_norm": 0.2848314344882965,
+      "learning_rate": 2.7180049425037407e-06,
+      "loss": 0.3912,
       "step": 118615
     },
     {
-      "epoch": 4.17,
-      "learning_rate": 3.5108947496709065e-06,
-      "loss": 0.242,
+      "epoch": 4.275056762893286,
+      "grad_norm": 0.30649277567863464,
+      "learning_rate": 2.7166818556193963e-06,
+      "loss": 0.396,
       "step": 118620
     },
     {
-      "epoch": 4.17,
-      "learning_rate": 3.509439113520932e-06,
-      "loss": 0.2552,
+      "epoch": 4.27523696255451,
+      "grad_norm": 0.31178387999534607,
+      "learning_rate": 2.7153590723409478e-06,
+      "loss": 0.4219,
       "step": 118625
     },
     {
-      "epoch": 4.17,
-      "learning_rate": 3.507983756411737e-06,
-      "loss": 0.2484,
+      "epoch": 4.275417162215735,
+      "grad_norm": 0.2467115819454193,
+      "learning_rate": 2.714036592686417e-06,
+      "loss": 0.3704,
       "step": 118630
     },
     {
-      "epoch": 4.17,
-      "learning_rate": 3.506528678362206e-06,
-      "loss": 0.2656,
+      "epoch": 4.27559736187696,
+      "grad_norm": 0.2777189314365387,
+      "learning_rate": 2.7127144166738173e-06,
+      "loss": 0.3898,
       "step": 118635
     },
     {
-      "epoch": 4.17,
-      "learning_rate": 3.505073879391238e-06,
-      "loss": 0.2556,
+      "epoch": 4.275777561538185,
+      "grad_norm": 0.23906902968883514,
+      "learning_rate": 2.711392544321176e-06,
+      "loss": 0.3649,
       "step": 118640
     },
     {
-      "epoch": 4.17,
-      "learning_rate": 3.5036193595177123e-06,
-      "loss": 0.2626,
+      "epoch": 4.275957761199409,
+      "grad_norm": 0.19675643742084503,
+      "learning_rate": 2.710070975646492e-06,
+      "loss": 0.3649,
       "step": 118645
     },
     {
-      "epoch": 4.17,
-      "learning_rate": 3.502165118760531e-06,
-      "loss": 0.2427,
+      "epoch": 4.276137960860634,
+      "grad_norm": 0.2644674777984619,
+      "learning_rate": 2.7087497106677793e-06,
+      "loss": 0.3944,
       "step": 118650
     },
     {
-      "epoch": 4.17,
-      "learning_rate": 3.500711157138567e-06,
-      "loss": 0.2403,
+      "epoch": 4.276318160521858,
+      "grad_norm": 0.2813570201396942,
+      "learning_rate": 2.707428749403035e-06,
+      "loss": 0.3593,
       "step": 118655
     },
     {
-      "epoch": 4.17,
-      "learning_rate": 3.499257474670692e-06,
-      "loss": 0.2379,
+      "epoch": 4.276498360183083,
+      "grad_norm": 0.24779802560806274,
+      "learning_rate": 2.706108091870252e-06,
+      "loss": 0.3836,
       "step": 118660
     },
     {
-      "epoch": 4.17,
-      "learning_rate": 3.4978040713757902e-06,
-      "loss": 0.2363,
+      "epoch": 4.276678559844307,
+      "grad_norm": 0.2426891177892685,
+      "learning_rate": 2.704787738087436e-06,
+      "loss": 0.3918,
       "step": 118665
     },
     {
-      "epoch": 4.18,
-      "learning_rate": 3.4963509472727392e-06,
-      "loss": 0.2586,
+      "epoch": 4.276858759505532,
+      "grad_norm": 0.3096999228000641,
+      "learning_rate": 2.703467688072575e-06,
+      "loss": 0.3599,
       "step": 118670
     },
     {
-      "epoch": 4.18,
-      "learning_rate": 3.4948981023804007e-06,
-      "loss": 0.2377,
+      "epoch": 4.277038959166757,
+      "grad_norm": 0.20364230871200562,
+      "learning_rate": 2.702147941843641e-06,
+      "loss": 0.3373,
       "step": 118675
     },
     {
-      "epoch": 4.18,
-      "learning_rate": 3.493445536717635e-06,
-      "loss": 0.2561,
+      "epoch": 4.2772191588279815,
+      "grad_norm": 0.24588048458099365,
+      "learning_rate": 2.7008284994186284e-06,
+      "loss": 0.3355,
       "step": 118680
     },
     {
-      "epoch": 4.18,
-      "learning_rate": 3.4919932503033022e-06,
-      "loss": 0.2447,
+      "epoch": 4.277399358489206,
+      "grad_norm": 0.24682247638702393,
+      "learning_rate": 2.6995093608155053e-06,
+      "loss": 0.3409,
       "step": 118685
     },
     {
-      "epoch": 4.18,
-      "learning_rate": 3.4905412431562675e-06,
-      "loss": 0.2475,
+      "epoch": 4.277579558150431,
+      "grad_norm": 0.28250157833099365,
+      "learning_rate": 2.6981905260522607e-06,
+      "loss": 0.3628,
       "step": 118690
     },
     {
-      "epoch": 4.18,
-      "learning_rate": 3.4890895152953794e-06,
-      "loss": 0.258,
+      "epoch": 4.277759757811655,
+      "grad_norm": 0.25315603613853455,
+      "learning_rate": 2.6968719951468464e-06,
+      "loss": 0.3906,
       "step": 118695
     },
     {
-      "epoch": 4.18,
-      "learning_rate": 3.4876380667394874e-06,
-      "loss": 0.2361,
+      "epoch": 4.27793995747288,
+      "grad_norm": 0.2779964506626129,
+      "learning_rate": 2.695553768117226e-06,
+      "loss": 0.3717,
       "step": 118700
     },
     {
-      "epoch": 4.18,
-      "learning_rate": 3.486186897507432e-06,
-      "loss": 0.2671,
+      "epoch": 4.278120157134104,
+      "grad_norm": 0.24498099088668823,
+      "learning_rate": 2.6942358449813745e-06,
+      "loss": 0.3657,
       "step": 118705
     },
     {
-      "epoch": 4.18,
-      "learning_rate": 3.484736007618067e-06,
-      "loss": 0.2538,
+      "epoch": 4.278300356795329,
+      "grad_norm": 0.2811969816684723,
+      "learning_rate": 2.6929182257572432e-06,
+      "loss": 0.3594,
       "step": 118710
     },
     {
-      "epoch": 4.18,
-      "learning_rate": 3.4832853970902218e-06,
-      "loss": 0.2454,
+      "epoch": 4.278480556456554,
+      "grad_norm": 0.2293609082698822,
+      "learning_rate": 2.6916009104627797e-06,
+      "loss": 0.3518,
       "step": 118715
     },
     {
-      "epoch": 4.18,
-      "learning_rate": 3.48183506594274e-06,
-      "loss": 0.2589,
+      "epoch": 4.2786607561177785,
+      "grad_norm": 0.3065091669559479,
+      "learning_rate": 2.6902838991159334e-06,
+      "loss": 0.3964,
       "step": 118720
     },
     {
-      "epoch": 4.18,
-      "learning_rate": 3.4803850141944426e-06,
-      "loss": 0.2587,
+      "epoch": 4.278840955779003,
+      "grad_norm": 0.21707165241241455,
+      "learning_rate": 2.6889671917346483e-06,
+      "loss": 0.3497,
       "step": 118725
     },
     {
-      "epoch": 4.18,
-      "learning_rate": 3.4789352418641728e-06,
-      "loss": 0.2427,
+      "epoch": 4.279021155440228,
+      "grad_norm": 0.3095008432865143,
+      "learning_rate": 2.687650788336868e-06,
+      "loss": 0.3616,
       "step": 118730
     },
     {
-      "epoch": 4.18,
-      "learning_rate": 3.4774857489707435e-06,
-      "loss": 0.2502,
+      "epoch": 4.279201355101453,
+      "grad_norm": 0.2820426821708679,
+      "learning_rate": 2.6863346889405255e-06,
+      "loss": 0.388,
       "step": 118735
     },
     {
-      "epoch": 4.18,
-      "learning_rate": 3.4760365355329725e-06,
-      "loss": 0.2455,
+      "epoch": 4.2793815547626775,
+      "grad_norm": 0.24508492648601532,
+      "learning_rate": 2.685018893563554e-06,
+      "loss": 0.394,
       "step": 118740
     },
     {
-      "epoch": 4.18,
-      "learning_rate": 3.4745876015696872e-06,
-      "loss": 0.2785,
+      "epoch": 4.279561754423901,
+      "grad_norm": 0.2219163477420807,
+      "learning_rate": 2.6837034022238806e-06,
+      "loss": 0.3613,
       "step": 118745
     },
     {
-      "epoch": 4.18,
-      "learning_rate": 3.473138947099697e-06,
-      "loss": 0.2643,
+      "epoch": 4.279741954085126,
+      "grad_norm": 0.23274736106395721,
+      "learning_rate": 2.6823882149394215e-06,
+      "loss": 0.3737,
       "step": 118750
     },
     {
-      "epoch": 4.18,
-      "learning_rate": 3.4716905721418114e-06,
-      "loss": 0.2686,
+      "epoch": 4.279922153746351,
+      "grad_norm": 0.27821585536003113,
+      "learning_rate": 2.6810733317281095e-06,
+      "loss": 0.3587,
       "step": 118755
     },
     {
-      "epoch": 4.18,
-      "learning_rate": 3.47024247671483e-06,
-      "loss": 0.255,
+      "epoch": 4.2801023534075755,
+      "grad_norm": 0.2513722777366638,
+      "learning_rate": 2.67975875260785e-06,
+      "loss": 0.3816,
       "step": 118760
     },
     {
-      "epoch": 4.18,
-      "learning_rate": 3.4687946608375655e-06,
-      "loss": 0.2389,
+      "epoch": 4.2802825530688,
+      "grad_norm": 0.26026418805122375,
+      "learning_rate": 2.6784444775965563e-06,
+      "loss": 0.3802,
       "step": 118765
     },
     {
-      "epoch": 4.18,
-      "learning_rate": 3.4673471245288083e-06,
-      "loss": 0.2675,
+      "epoch": 4.280462752730025,
+      "grad_norm": 0.2638653814792633,
+      "learning_rate": 2.6771305067121363e-06,
+      "loss": 0.3756,
       "step": 118770
     },
     {
-      "epoch": 4.18,
-      "learning_rate": 3.465899867807365e-06,
-      "loss": 0.2451,
+      "epoch": 4.28064295239125,
+      "grad_norm": 0.2352520376443863,
+      "learning_rate": 2.675816839972489e-06,
+      "loss": 0.4026,
       "step": 118775
     },
     {
-      "epoch": 4.18,
-      "learning_rate": 3.4644528906920132e-06,
-      "loss": 0.233,
+      "epoch": 4.280823152052474,
+      "grad_norm": 0.3046009838581085,
+      "learning_rate": 2.6745034773955095e-06,
+      "loss": 0.3791,
       "step": 118780
     },
     {
-      "epoch": 4.18,
-      "learning_rate": 3.463006193201554e-06,
-      "loss": 0.2499,
+      "epoch": 4.281003351713699,
+      "grad_norm": 0.21931016445159912,
+      "learning_rate": 2.6731904189991108e-06,
+      "loss": 0.3673,
       "step": 118785
     },
     {
-      "epoch": 4.18,
-      "learning_rate": 3.461559775354767e-06,
-      "loss": 0.2304,
+      "epoch": 4.281183551374923,
+      "grad_norm": 0.21778719127178192,
+      "learning_rate": 2.671877664801159e-06,
+      "loss": 0.3712,
       "step": 118790
     },
     {
-      "epoch": 4.18,
-      "learning_rate": 3.4601136371704314e-06,
-      "loss": 0.2727,
+      "epoch": 4.281363751036148,
+      "grad_norm": 0.22375808656215668,
+      "learning_rate": 2.6705652148195566e-06,
+      "loss": 0.3566,
       "step": 118795
     },
     {
-      "epoch": 4.18,
-      "learning_rate": 3.458667778667318e-06,
-      "loss": 0.26,
+      "epoch": 4.2815439506973725,
+      "grad_norm": 0.30791062116622925,
+      "learning_rate": 2.6692530690721777e-06,
+      "loss": 0.3972,
       "step": 118800
     },
     {
-      "epoch": 4.18,
-      "learning_rate": 3.4572221998642128e-06,
-      "loss": 0.2551,
+      "epoch": 4.281724150358597,
+      "grad_norm": 0.259032279253006,
+      "learning_rate": 2.667941227576906e-06,
+      "loss": 0.3812,
       "step": 118805
     },
     {
-      "epoch": 4.18,
-      "learning_rate": 3.4557769007798807e-06,
-      "loss": 0.2581,
+      "epoch": 4.281904350019822,
+      "grad_norm": 0.2583853006362915,
+      "learning_rate": 2.66662969035161e-06,
+      "loss": 0.3902,
       "step": 118810
     },
     {
-      "epoch": 4.18,
-      "learning_rate": 3.45433188143309e-06,
-      "loss": 0.2619,
+      "epoch": 4.282084549681047,
+      "grad_norm": 0.3107074499130249,
+      "learning_rate": 2.6653184574141594e-06,
+      "loss": 0.4077,
       "step": 118815
     },
     {
-      "epoch": 4.18,
-      "learning_rate": 3.4528871418425922e-06,
-      "loss": 0.2278,
+      "epoch": 4.282264749342271,
+      "grad_norm": 0.2655295729637146,
+      "learning_rate": 2.6640075287824166e-06,
+      "loss": 0.3709,
       "step": 118820
     },
     {
-      "epoch": 4.18,
-      "learning_rate": 3.451442682027156e-06,
-      "loss": 0.2565,
+      "epoch": 4.282444949003496,
+      "grad_norm": 0.2224242240190506,
+      "learning_rate": 2.6626969044742518e-06,
+      "loss": 0.3534,
       "step": 118825
     },
     {
-      "epoch": 4.18,
-      "learning_rate": 3.4499985020055386e-06,
-      "loss": 0.2589,
+      "epoch": 4.282625148664721,
+      "grad_norm": 0.25910016894340515,
+      "learning_rate": 2.6613865845075164e-06,
+      "loss": 0.3621,
       "step": 118830
     },
     {
-      "epoch": 4.18,
-      "learning_rate": 3.4485546017964916e-06,
-      "loss": 0.2446,
+      "epoch": 4.282805348325946,
+      "grad_norm": 0.24160423874855042,
+      "learning_rate": 2.6600765689000634e-06,
+      "loss": 0.3737,
       "step": 118835
     },
     {
-      "epoch": 4.18,
-      "learning_rate": 3.4471109814187552e-06,
-      "loss": 0.2836,
+      "epoch": 4.2829855479871695,
+      "grad_norm": 0.2217995524406433,
+      "learning_rate": 2.658766857669745e-06,
+      "loss": 0.3841,
       "step": 118840
     },
     {
-      "epoch": 4.18,
-      "learning_rate": 3.4456676408910814e-06,
-      "loss": 0.2464,
+      "epoch": 4.283165747648394,
+      "grad_norm": 0.25542962551116943,
+      "learning_rate": 2.657457450834394e-06,
+      "loss": 0.4073,
       "step": 118845
     },
     {
-      "epoch": 4.18,
-      "learning_rate": 3.444224580232211e-06,
-      "loss": 0.2539,
+      "epoch": 4.283345947309619,
+      "grad_norm": 0.26905080676078796,
+      "learning_rate": 2.656148348411866e-06,
+      "loss": 0.3513,
       "step": 118850
     },
     {
-      "epoch": 4.18,
-      "learning_rate": 3.4427817994608702e-06,
-      "loss": 0.2403,
+      "epoch": 4.283526146970844,
+      "grad_norm": 0.2789619266986847,
+      "learning_rate": 2.6548395504199963e-06,
+      "loss": 0.3766,
       "step": 118855
     },
     {
-      "epoch": 4.18,
-      "learning_rate": 3.441339298595808e-06,
-      "loss": 0.2511,
+      "epoch": 4.283706346632068,
+      "grad_norm": 0.21455584466457367,
+      "learning_rate": 2.6535310568766013e-06,
+      "loss": 0.3954,
       "step": 118860
     },
     {
-      "epoch": 4.18,
-      "learning_rate": 3.439897077655749e-06,
-      "loss": 0.2489,
+      "epoch": 4.283886546293293,
+      "grad_norm": 0.3085833787918091,
+      "learning_rate": 2.6522228677995252e-06,
+      "loss": 0.4161,
       "step": 118865
     },
     {
-      "epoch": 4.18,
-      "learning_rate": 3.438455136659416e-06,
-      "loss": 0.2584,
+      "epoch": 4.284066745954518,
+      "grad_norm": 0.22385522723197937,
+      "learning_rate": 2.6509149832065817e-06,
+      "loss": 0.3982,
       "step": 118870
     },
     {
-      "epoch": 4.18,
-      "learning_rate": 3.437013475625528e-06,
-      "loss": 0.2549,
+      "epoch": 4.284246945615743,
+      "grad_norm": 0.2983032464981079,
+      "learning_rate": 2.6496074031156004e-06,
+      "loss": 0.3327,
       "step": 118875
     },
     {
-      "epoch": 4.18,
-      "learning_rate": 3.435572094572817e-06,
-      "loss": 0.2346,
+      "epoch": 4.284427145276966,
+      "grad_norm": 0.27663928270339966,
+      "learning_rate": 2.6483001275443896e-06,
+      "loss": 0.3619,
       "step": 118880
     },
     {
-      "epoch": 4.18,
-      "learning_rate": 3.434130993519982e-06,
-      "loss": 0.269,
+      "epoch": 4.284607344938191,
+      "grad_norm": 0.28087663650512695,
+      "learning_rate": 2.6469931565107575e-06,
+      "loss": 0.3947,
       "step": 118885
     },
     {
-      "epoch": 4.18,
-      "learning_rate": 3.43269017248575e-06,
-      "loss": 0.2394,
+      "epoch": 4.284787544599416,
+      "grad_norm": 0.27490758895874023,
+      "learning_rate": 2.6456864900325174e-06,
+      "loss": 0.3937,
       "step": 118890
     },
     {
-      "epoch": 4.18,
-      "learning_rate": 3.4312496314888198e-06,
-      "loss": 0.2548,
+      "epoch": 4.284967744260641,
+      "grad_norm": 0.28742140531539917,
+      "learning_rate": 2.644380128127474e-06,
+      "loss": 0.3938,
       "step": 118895
     },
     {
-      "epoch": 4.18,
-      "learning_rate": 3.4298093705479012e-06,
-      "loss": 0.2422,
+      "epoch": 4.285147943921865,
+      "grad_norm": 0.2435084581375122,
+      "learning_rate": 2.6430740708134226e-06,
+      "loss": 0.3626,
       "step": 118900
     },
     {
-      "epoch": 4.18,
-      "learning_rate": 3.428369389681696e-06,
-      "loss": 0.24,
+      "epoch": 4.28532814358309,
+      "grad_norm": 0.27407246828079224,
+      "learning_rate": 2.6417683181081592e-06,
+      "loss": 0.3699,
       "step": 118905
     },
     {
-      "epoch": 4.18,
-      "learning_rate": 3.4269296889088946e-06,
-      "loss": 0.2584,
+      "epoch": 4.285508343244315,
+      "grad_norm": 0.2970322370529175,
+      "learning_rate": 2.640462870029467e-06,
+      "loss": 0.3878,
       "step": 118910
     },
     {
-      "epoch": 4.18,
-      "learning_rate": 3.425490268248191e-06,
-      "loss": 0.2567,
+      "epoch": 4.28568854290554,
+      "grad_norm": 0.27050116658210754,
+      "learning_rate": 2.6391577265951455e-06,
+      "loss": 0.4109,
       "step": 118915
     },
     {
-      "epoch": 4.18,
-      "learning_rate": 3.4240511277182808e-06,
-      "loss": 0.2364,
+      "epoch": 4.285868742566764,
+      "grad_norm": 0.18248991668224335,
+      "learning_rate": 2.6378528878229697e-06,
+      "loss": 0.3786,
       "step": 118920
     },
     {
-      "epoch": 4.18,
-      "learning_rate": 3.422612267337852e-06,
-      "loss": 0.25,
+      "epoch": 4.286048942227989,
+      "grad_norm": 0.24371817708015442,
+      "learning_rate": 2.6365483537307197e-06,
+      "loss": 0.3315,
       "step": 118925
     },
     {
-      "epoch": 4.18,
-      "learning_rate": 3.421173687125573e-06,
-      "loss": 0.2475,
+      "epoch": 4.286229141889213,
+      "grad_norm": 0.2470201700925827,
+      "learning_rate": 2.6352441243361698e-06,
+      "loss": 0.3701,
       "step": 118930
     },
     {
-      "epoch": 4.18,
-      "learning_rate": 3.41973538710014e-06,
-      "loss": 0.27,
+      "epoch": 4.286409341550438,
+      "grad_norm": 0.19732847809791565,
+      "learning_rate": 2.6339401996570813e-06,
+      "loss": 0.3634,
       "step": 118935
     },
     {
-      "epoch": 4.18,
-      "learning_rate": 3.4182973672802153e-06,
-      "loss": 0.2454,
+      "epoch": 4.286589541211662,
+      "grad_norm": 0.21893630921840668,
+      "learning_rate": 2.632636579711234e-06,
+      "loss": 0.3664,
       "step": 118940
     },
     {
-      "epoch": 4.18,
-      "learning_rate": 3.4168596276844823e-06,
-      "loss": 0.2377,
+      "epoch": 4.286769740872887,
+      "grad_norm": 0.26393207907676697,
+      "learning_rate": 2.6313332645163836e-06,
+      "loss": 0.3683,
       "step": 118945
     },
     {
-      "epoch": 4.18,
-      "learning_rate": 3.415709637781853e-06,
-      "loss": 0.241,
+      "epoch": 4.286949940534112,
+      "grad_norm": 0.2583698630332947,
+      "learning_rate": 2.630030254090285e-06,
+      "loss": 0.3745,
       "step": 118950
     },
     {
-      "epoch": 4.19,
-      "learning_rate": 3.4142724026366953e-06,
-      "loss": 0.2411,
+      "epoch": 4.287130140195337,
+      "grad_norm": 0.22545069456100464,
+      "learning_rate": 2.6287275484506934e-06,
+      "loss": 0.4017,
       "step": 118955
     },
     {
-      "epoch": 4.19,
-      "learning_rate": 3.412835447767987e-06,
-      "loss": 0.2634,
+      "epoch": 4.287310339856561,
+      "grad_norm": 0.3334513008594513,
+      "learning_rate": 2.6274251476153587e-06,
+      "loss": 0.382,
       "step": 118960
     },
     {
-      "epoch": 4.19,
-      "learning_rate": 3.4113987731943743e-06,
-      "loss": 0.2555,
+      "epoch": 4.287490539517786,
+      "grad_norm": 0.32256385684013367,
+      "learning_rate": 2.62612305160202e-06,
+      "loss": 0.3868,
       "step": 118965
     },
     {
-      "epoch": 4.19,
-      "learning_rate": 3.409962378934525e-06,
-      "loss": 0.2322,
+      "epoch": 4.287670739179011,
+      "grad_norm": 0.28778645396232605,
+      "learning_rate": 2.6248212604284313e-06,
+      "loss": 0.3516,
       "step": 118970
     },
     {
-      "epoch": 4.19,
-      "learning_rate": 3.408526265007089e-06,
-      "loss": 0.2418,
+      "epoch": 4.287850938840235,
+      "grad_norm": 0.21005696058273315,
+      "learning_rate": 2.6235197741123122e-06,
+      "loss": 0.3425,
       "step": 118975
     },
     {
-      "epoch": 4.19,
-      "learning_rate": 3.407090431430712e-06,
-      "loss": 0.2518,
+      "epoch": 4.288031138501459,
+      "grad_norm": 0.20069657266139984,
+      "learning_rate": 2.6222185926714076e-06,
+      "loss": 0.3665,
       "step": 118980
     },
     {
-      "epoch": 4.19,
-      "learning_rate": 3.4056548782240265e-06,
-      "loss": 0.2366,
+      "epoch": 4.288211338162684,
+      "grad_norm": 0.30215156078338623,
+      "learning_rate": 2.6209177161234445e-06,
+      "loss": 0.3804,
       "step": 118985
     },
     {
-      "epoch": 4.19,
-      "learning_rate": 3.4042196054056896e-06,
-      "loss": 0.2806,
+      "epoch": 4.288391537823909,
+      "grad_norm": 0.22637207806110382,
+      "learning_rate": 2.6196171444861417e-06,
+      "loss": 0.3878,
       "step": 118990
     },
     {
-      "epoch": 4.19,
-      "learning_rate": 3.402784612994328e-06,
-      "loss": 0.2627,
+      "epoch": 4.2885717374851335,
+      "grad_norm": 0.28523603081703186,
+      "learning_rate": 2.6183168777772244e-06,
+      "loss": 0.3649,
       "step": 118995
     },
     {
-      "epoch": 4.19,
-      "learning_rate": 3.401349901008577e-06,
-      "loss": 0.2632,
+      "epoch": 4.288751937146358,
+      "grad_norm": 0.22392310202121735,
+      "learning_rate": 2.617016916014406e-06,
+      "loss": 0.3961,
       "step": 119000
     },
     {
-      "epoch": 4.19,
-      "eval_loss": 0.2492852360010147,
-      "eval_runtime": 10.5487,
-      "eval_samples_per_second": 9.48,
-      "eval_steps_per_second": 9.48,
+      "epoch": 4.288751937146358,
+      "eval_loss": 0.4290127456188202,
+      "eval_runtime": 3.5355,
+      "eval_samples_per_second": 28.284,
+      "eval_steps_per_second": 7.071,
       "step": 119000
     },
     {
-      "epoch": 4.19,
-      "learning_rate": 3.3999154694670548e-06,
-      "loss": 0.254,
+      "epoch": 4.288932136807583,
+      "grad_norm": 0.2049478143453598,
+      "learning_rate": 2.6157172592153944e-06,
+      "loss": 0.3615,
       "step": 119005
     },
     {
-      "epoch": 4.19,
-      "learning_rate": 3.3984813183884022e-06,
-      "loss": 0.2624,
+      "epoch": 4.289112336468808,
+      "grad_norm": 0.23813143372535706,
+      "learning_rate": 2.614417907397906e-06,
+      "loss": 0.3763,
       "step": 119010
     },
     {
-      "epoch": 4.19,
-      "learning_rate": 3.397047447791235e-06,
-      "loss": 0.2382,
+      "epoch": 4.2892925361300325,
+      "grad_norm": 0.29590684175491333,
+      "learning_rate": 2.6131188605796384e-06,
+      "loss": 0.3898,
       "step": 119015
     },
     {
-      "epoch": 4.19,
-      "learning_rate": 3.3956138576941633e-06,
-      "loss": 0.258,
+      "epoch": 4.289472735791256,
+      "grad_norm": 0.24687814712524414,
+      "learning_rate": 2.6118201187782937e-06,
+      "loss": 0.3756,
       "step": 119020
     },
     {
-      "epoch": 4.19,
-      "learning_rate": 3.3941805481158133e-06,
-      "loss": 0.2312,
+      "epoch": 4.289652935452481,
+      "grad_norm": 0.2702934443950653,
+      "learning_rate": 2.6105216820115657e-06,
+      "loss": 0.3699,
       "step": 119025
     },
     {
-      "epoch": 4.19,
-      "learning_rate": 3.392747519074785e-06,
-      "loss": 0.245,
+      "epoch": 4.289833135113706,
+      "grad_norm": 0.24939516186714172,
+      "learning_rate": 2.609223550297141e-06,
+      "loss": 0.3705,
       "step": 119030
     },
     {
-      "epoch": 4.19,
-      "learning_rate": 3.391314770589696e-06,
-      "loss": 0.2552,
+      "epoch": 4.2900133347749305,
+      "grad_norm": 0.2283768504858017,
+      "learning_rate": 2.607925723652713e-06,
+      "loss": 0.3792,
       "step": 119035
     },
     {
-      "epoch": 4.19,
-      "learning_rate": 3.389882302679137e-06,
-      "loss": 0.2511,
+      "epoch": 4.290193534436155,
+      "grad_norm": 0.24190238118171692,
+      "learning_rate": 2.6066282020959678e-06,
+      "loss": 0.3742,
       "step": 119040
     },
     {
-      "epoch": 4.19,
-      "learning_rate": 3.3884501153617214e-06,
-      "loss": 0.2263,
+      "epoch": 4.29037373409738,
+      "grad_norm": 0.1982034593820572,
+      "learning_rate": 2.605330985644566e-06,
+      "loss": 0.3406,
       "step": 119045
     },
     {
-      "epoch": 4.19,
-      "learning_rate": 3.3870182086560396e-06,
-      "loss": 0.268,
+      "epoch": 4.290553933758605,
+      "grad_norm": 0.2854166030883789,
+      "learning_rate": 2.6040340743162e-06,
+      "loss": 0.3875,
       "step": 119050
     },
     {
-      "epoch": 4.19,
-      "learning_rate": 3.3855865825806797e-06,
-      "loss": 0.2514,
+      "epoch": 4.2907341334198295,
+      "grad_norm": 0.2228785753250122,
+      "learning_rate": 2.60273746812853e-06,
+      "loss": 0.3492,
       "step": 119055
     },
     {
-      "epoch": 4.19,
-      "learning_rate": 3.3841552371542297e-06,
-      "loss": 0.2475,
+      "epoch": 4.290914333081054,
+      "grad_norm": 0.3077186346054077,
+      "learning_rate": 2.6014411670992305e-06,
+      "loss": 0.4009,
       "step": 119060
     },
     {
-      "epoch": 4.19,
-      "learning_rate": 3.382724172395285e-06,
-      "loss": 0.2654,
+      "epoch": 4.291094532742278,
+      "grad_norm": 0.26383304595947266,
+      "learning_rate": 2.60014517124596e-06,
+      "loss": 0.3999,
       "step": 119065
     },
     {
-      "epoch": 4.19,
-      "learning_rate": 3.381293388322421e-06,
-      "loss": 0.2376,
+      "epoch": 4.291274732403503,
+      "grad_norm": 0.2802242934703827,
+      "learning_rate": 2.5988494805863682e-06,
+      "loss": 0.3834,
       "step": 119070
     },
     {
-      "epoch": 4.19,
-      "learning_rate": 3.3798628849542072e-06,
-      "loss": 0.2617,
+      "epoch": 4.2914549320647275,
+      "grad_norm": 0.21582341194152832,
+      "learning_rate": 2.597554095138122e-06,
+      "loss": 0.367,
       "step": 119075
     },
     {
-      "epoch": 4.19,
-      "learning_rate": 3.3784326623092333e-06,
-      "loss": 0.2533,
+      "epoch": 4.291635131725952,
+      "grad_norm": 0.2001202404499054,
+      "learning_rate": 2.5962590149188616e-06,
+      "loss": 0.3211,
       "step": 119080
     },
     {
-      "epoch": 4.19,
-      "learning_rate": 3.3770027204060555e-06,
-      "loss": 0.2464,
+      "epoch": 4.291815331387177,
+      "grad_norm": 0.27572277188301086,
+      "learning_rate": 2.594964239946232e-06,
+      "loss": 0.3591,
       "step": 119085
     },
     {
-      "epoch": 4.19,
-      "learning_rate": 3.375573059263254e-06,
-      "loss": 0.2422,
+      "epoch": 4.291995531048402,
+      "grad_norm": 0.2560023069381714,
+      "learning_rate": 2.5936697702378804e-06,
+      "loss": 0.3773,
       "step": 119090
     },
     {
-      "epoch": 4.19,
-      "learning_rate": 3.374143678899383e-06,
-      "loss": 0.2526,
+      "epoch": 4.292175730709626,
+      "grad_norm": 0.23489350080490112,
+      "learning_rate": 2.592375605811434e-06,
+      "loss": 0.3515,
       "step": 119095
     },
     {
-      "epoch": 4.19,
-      "learning_rate": 3.372714579333003e-06,
-      "loss": 0.2593,
+      "epoch": 4.292355930370851,
+      "grad_norm": 0.25158292055130005,
+      "learning_rate": 2.591081746684537e-06,
+      "loss": 0.3998,
       "step": 119100
     },
     {
-      "epoch": 4.19,
-      "learning_rate": 3.3712857605826738e-06,
-      "loss": 0.255,
+      "epoch": 4.292536130032076,
+      "grad_norm": 0.23154456913471222,
+      "learning_rate": 2.5897881928748145e-06,
+      "loss": 0.3961,
       "step": 119105
     },
     {
-      "epoch": 4.19,
-      "learning_rate": 3.3698572226669446e-06,
-      "loss": 0.2451,
+      "epoch": 4.292716329693301,
+      "grad_norm": 0.2787790298461914,
+      "learning_rate": 2.5884949443998857e-06,
+      "loss": 0.375,
       "step": 119110
     },
     {
-      "epoch": 4.19,
-      "learning_rate": 3.3684289656043645e-06,
-      "loss": 0.2499,
+      "epoch": 4.2928965293545245,
+      "grad_norm": 0.23555888235569,
+      "learning_rate": 2.587202001277378e-06,
+      "loss": 0.3809,
       "step": 119115
     },
     {
-      "epoch": 4.19,
-      "learning_rate": 3.3670009894134743e-06,
-      "loss": 0.2308,
+      "epoch": 4.293076729015749,
+      "grad_norm": 0.25257548689842224,
+      "learning_rate": 2.5859093635248964e-06,
+      "loss": 0.3726,
       "step": 119120
     },
     {
-      "epoch": 4.19,
-      "learning_rate": 3.3655732941128255e-06,
-      "loss": 0.2485,
+      "epoch": 4.293256928676974,
+      "grad_norm": 0.23325851559638977,
+      "learning_rate": 2.5846170311600637e-06,
+      "loss": 0.3535,
       "step": 119125
     },
     {
-      "epoch": 4.19,
-      "learning_rate": 3.364145879720948e-06,
-      "loss": 0.2657,
+      "epoch": 4.293437128338199,
+      "grad_norm": 0.2508438527584076,
+      "learning_rate": 2.5833250042004876e-06,
+      "loss": 0.3674,
       "step": 119130
     },
     {
-      "epoch": 4.19,
-      "learning_rate": 3.362718746256374e-06,
-      "loss": 0.2777,
+      "epoch": 4.293617327999423,
+      "grad_norm": 0.3049331605434418,
+      "learning_rate": 2.582033282663765e-06,
+      "loss": 0.409,
       "step": 119135
     },
     {
-      "epoch": 4.19,
-      "learning_rate": 3.3612918937376366e-06,
-      "loss": 0.2472,
+      "epoch": 4.293797527660648,
+      "grad_norm": 0.22652621567249298,
+      "learning_rate": 2.5807418665675017e-06,
+      "loss": 0.3541,
       "step": 119140
     },
     {
-      "epoch": 4.19,
-      "learning_rate": 3.35986532218327e-06,
-      "loss": 0.2415,
+      "epoch": 4.293977727321873,
+      "grad_norm": 0.2690187096595764,
+      "learning_rate": 2.5794507559292885e-06,
+      "loss": 0.3622,
       "step": 119145
     },
     {
-      "epoch": 4.19,
-      "learning_rate": 3.3584390316117904e-06,
-      "loss": 0.2415,
+      "epoch": 4.294157926983098,
+      "grad_norm": 0.24629640579223633,
+      "learning_rate": 2.578159950766712e-06,
+      "loss": 0.3736,
       "step": 119150
     },
     {
-      "epoch": 4.19,
-      "learning_rate": 3.3570130220417107e-06,
-      "loss": 0.2461,
+      "epoch": 4.2943381266443215,
+      "grad_norm": 0.2614523470401764,
+      "learning_rate": 2.576869451097377e-06,
+      "loss": 0.384,
       "step": 119155
     },
     {
-      "epoch": 4.19,
-      "learning_rate": 3.3555872934915604e-06,
-      "loss": 0.2749,
+      "epoch": 4.294518326305546,
+      "grad_norm": 0.2685030996799469,
+      "learning_rate": 2.5755792569388455e-06,
+      "loss": 0.3787,
       "step": 119160
     },
     {
-      "epoch": 4.19,
-      "learning_rate": 3.3541618459798447e-06,
-      "loss": 0.2483,
+      "epoch": 4.294698525966771,
+      "grad_norm": 0.23941963911056519,
+      "learning_rate": 2.5742893683087076e-06,
+      "loss": 0.4015,
       "step": 119165
     },
     {
-      "epoch": 4.19,
-      "learning_rate": 3.352736679525073e-06,
-      "loss": 0.2565,
+      "epoch": 4.294878725627996,
+      "grad_norm": 0.2448320984840393,
+      "learning_rate": 2.5729997852245363e-06,
+      "loss": 0.4058,
       "step": 119170
     },
     {
-      "epoch": 4.19,
-      "learning_rate": 3.3513117941457424e-06,
-      "loss": 0.2477,
+      "epoch": 4.29505892528922,
+      "grad_norm": 0.2499290406703949,
+      "learning_rate": 2.571710507703895e-06,
+      "loss": 0.3464,
       "step": 119175
     },
     {
-      "epoch": 4.19,
-      "learning_rate": 3.3498871898603677e-06,
-      "loss": 0.2491,
+      "epoch": 4.295239124950445,
+      "grad_norm": 0.263079971075058,
+      "learning_rate": 2.5704215357643666e-06,
+      "loss": 0.3939,
       "step": 119180
     },
     {
-      "epoch": 4.19,
-      "learning_rate": 3.3484628666874405e-06,
-      "loss": 0.254,
+      "epoch": 4.29541932461167,
+      "grad_norm": 0.22637499868869781,
+      "learning_rate": 2.5691328694234964e-06,
+      "loss": 0.393,
       "step": 119185
     },
     {
-      "epoch": 4.19,
-      "learning_rate": 3.347038824645446e-06,
-      "loss": 0.2663,
+      "epoch": 4.295599524272895,
+      "grad_norm": 0.29773595929145813,
+      "learning_rate": 2.567844508698844e-06,
+      "loss": 0.4092,
       "step": 119190
     },
     {
-      "epoch": 4.19,
-      "learning_rate": 3.3456150637528856e-06,
-      "loss": 0.2688,
+      "epoch": 4.295779723934119,
+      "grad_norm": 0.25655168294906616,
+      "learning_rate": 2.566556453607974e-06,
+      "loss": 0.3502,
       "step": 119195
     },
     {
-      "epoch": 4.19,
-      "learning_rate": 3.3441915840282473e-06,
-      "loss": 0.274,
+      "epoch": 4.295959923595344,
+      "grad_norm": 0.28525644540786743,
+      "learning_rate": 2.5652687041684247e-06,
+      "loss": 0.3887,
       "step": 119200
     },
     {
-      "epoch": 4.19,
-      "learning_rate": 3.342768385490008e-06,
-      "loss": 0.2508,
+      "epoch": 4.296140123256568,
+      "grad_norm": 0.2969033718109131,
+      "learning_rate": 2.563981260397749e-06,
+      "loss": 0.3525,
       "step": 119205
     },
     {
-      "epoch": 4.19,
-      "learning_rate": 3.341345468156651e-06,
-      "loss": 0.2563,
+      "epoch": 4.296320322917793,
+      "grad_norm": 0.20924274623394012,
+      "learning_rate": 2.562694122313483e-06,
+      "loss": 0.3388,
       "step": 119210
     },
     {
-      "epoch": 4.19,
-      "learning_rate": 3.339922832046641e-06,
-      "loss": 0.2641,
+      "epoch": 4.296500522579017,
+      "grad_norm": 0.24764671921730042,
+      "learning_rate": 2.5614072899331625e-06,
+      "loss": 0.3648,
       "step": 119215
     },
     {
-      "epoch": 4.19,
-      "learning_rate": 3.3385004771784663e-06,
-      "loss": 0.263,
+      "epoch": 4.296680722240242,
+      "grad_norm": 0.2710559368133545,
+      "learning_rate": 2.560120763274329e-06,
+      "loss": 0.342,
       "step": 119220
     },
     {
-      "epoch": 4.19,
-      "learning_rate": 3.337078403570587e-06,
-      "loss": 0.2462,
+      "epoch": 4.296860921901467,
+      "grad_norm": 0.3236762285232544,
+      "learning_rate": 2.5588345423545046e-06,
+      "loss": 0.3871,
       "step": 119225
     },
     {
-      "epoch": 4.19,
-      "learning_rate": 3.3356566112414695e-06,
-      "loss": 0.2361,
+      "epoch": 4.297041121562692,
+      "grad_norm": 0.2507827877998352,
+      "learning_rate": 2.5575486271912143e-06,
+      "loss": 0.3975,
       "step": 119230
     },
     {
-      "epoch": 4.2,
-      "learning_rate": 3.334235100209565e-06,
-      "loss": 0.2614,
+      "epoch": 4.297221321223916,
+      "grad_norm": 0.3020266890525818,
+      "learning_rate": 2.556263017801977e-06,
+      "loss": 0.3805,
       "step": 119235
     },
     {
-      "epoch": 4.2,
-      "learning_rate": 3.332813870493348e-06,
-      "loss": 0.2713,
+      "epoch": 4.297401520885141,
+      "grad_norm": 0.2670553922653198,
+      "learning_rate": 2.55497771420431e-06,
+      "loss": 0.3863,
       "step": 119240
     },
     {
-      "epoch": 4.2,
-      "learning_rate": 3.3313929221112564e-06,
-      "loss": 0.2877,
+      "epoch": 4.297581720546366,
+      "grad_norm": 0.29104623198509216,
+      "learning_rate": 2.553692716415729e-06,
+      "loss": 0.3768,
       "step": 119245
     },
     {
-      "epoch": 4.2,
-      "learning_rate": 3.3299722550817536e-06,
-      "loss": 0.2484,
+      "epoch": 4.29776192020759,
+      "grad_norm": 0.2508760392665863,
+      "learning_rate": 2.552408024453742e-06,
+      "loss": 0.3921,
       "step": 119250
     },
     {
-      "epoch": 4.2,
-      "learning_rate": 3.328551869423274e-06,
-      "loss": 0.2591,
+      "epoch": 4.297942119868814,
+      "grad_norm": 0.27478477358818054,
+      "learning_rate": 2.5511236383358422e-06,
+      "loss": 0.3742,
       "step": 119255
     },
     {
-      "epoch": 4.2,
-      "learning_rate": 3.3271317651542733e-06,
-      "loss": 0.2664,
+      "epoch": 4.298122319530039,
+      "grad_norm": 0.28668543696403503,
+      "learning_rate": 2.549839558079542e-06,
+      "loss": 0.374,
       "step": 119260
     },
     {
-      "epoch": 4.2,
-      "learning_rate": 3.325711942293183e-06,
-      "loss": 0.2699,
+      "epoch": 4.298302519191264,
+      "grad_norm": 0.22271794080734253,
+      "learning_rate": 2.5485557837023303e-06,
+      "loss": 0.3471,
       "step": 119265
     },
     {
-      "epoch": 4.2,
-      "learning_rate": 3.324292400858439e-06,
-      "loss": 0.2607,
+      "epoch": 4.298482718852489,
+      "grad_norm": 0.22481966018676758,
+      "learning_rate": 2.5472723152216992e-06,
+      "loss": 0.3503,
       "step": 119270
     },
     {
-      "epoch": 4.2,
-      "learning_rate": 3.3228731408684654e-06,
-      "loss": 0.2423,
+      "epoch": 4.298662918513713,
+      "grad_norm": 0.22791920602321625,
+      "learning_rate": 2.545989152655137e-06,
+      "loss": 0.3945,
       "step": 119275
     },
     {
-      "epoch": 4.2,
-      "learning_rate": 3.321454162341708e-06,
-      "loss": 0.2607,
+      "epoch": 4.298843118174938,
+      "grad_norm": 0.3118433952331543,
+      "learning_rate": 2.544706296020119e-06,
+      "loss": 0.368,
       "step": 119280
     },
     {
-      "epoch": 4.2,
-      "learning_rate": 3.320035465296578e-06,
-      "loss": 0.2525,
+      "epoch": 4.299023317836163,
+      "grad_norm": 0.2143528163433075,
+      "learning_rate": 2.5434237453341364e-06,
+      "loss": 0.3645,
       "step": 119285
     },
     {
-      "epoch": 4.2,
-      "learning_rate": 3.3186170497514963e-06,
-      "loss": 0.2772,
+      "epoch": 4.2992035174973875,
+      "grad_norm": 0.28790873289108276,
+      "learning_rate": 2.5421415006146558e-06,
+      "loss": 0.391,
       "step": 119290
     },
     {
-      "epoch": 4.2,
-      "learning_rate": 3.3171989157248897e-06,
-      "loss": 0.2663,
+      "epoch": 4.299383717158611,
+      "grad_norm": 0.25662392377853394,
+      "learning_rate": 2.5408595618791497e-06,
+      "loss": 0.3487,
       "step": 119295
     },
     {
-      "epoch": 4.2,
-      "learning_rate": 3.3157810632351597e-06,
-      "loss": 0.2536,
+      "epoch": 4.299563916819836,
+      "grad_norm": 0.26446598768234253,
+      "learning_rate": 2.5395779291450846e-06,
+      "loss": 0.3973,
       "step": 119300
     },
     {
-      "epoch": 4.2,
-      "learning_rate": 3.314363492300726e-06,
-      "loss": 0.2519,
+      "epoch": 4.299744116481061,
+      "grad_norm": 0.24997925758361816,
+      "learning_rate": 2.538296602429921e-06,
+      "loss": 0.3521,
       "step": 119305
     },
     {
-      "epoch": 4.2,
-      "learning_rate": 3.312946202939987e-06,
-      "loss": 0.2738,
+      "epoch": 4.2999243161422855,
+      "grad_norm": 0.22187812626361847,
+      "learning_rate": 2.537015581751115e-06,
+      "loss": 0.3569,
       "step": 119310
     },
     {
-      "epoch": 4.2,
-      "learning_rate": 3.311529195171356e-06,
-      "loss": 0.2578,
+      "epoch": 4.30010451580351,
+      "grad_norm": 0.23413535952568054,
+      "learning_rate": 2.5357348671261246e-06,
+      "loss": 0.3708,
       "step": 119315
     },
     {
-      "epoch": 4.2,
-      "learning_rate": 3.310112469013224e-06,
-      "loss": 0.2636,
+      "epoch": 4.300284715464735,
+      "grad_norm": 0.2146236151456833,
+      "learning_rate": 2.5344544585723997e-06,
+      "loss": 0.3627,
       "step": 119320
     },
     {
-      "epoch": 4.2,
-      "learning_rate": 3.308696024483987e-06,
-      "loss": 0.2374,
+      "epoch": 4.30046491512596,
+      "grad_norm": 0.24262899160385132,
+      "learning_rate": 2.5331743561073816e-06,
+      "loss": 0.3694,
       "step": 119325
     },
     {
-      "epoch": 4.2,
-      "learning_rate": 3.3072798616020306e-06,
-      "loss": 0.2472,
+      "epoch": 4.3006451147871845,
+      "grad_norm": 0.21525375545024872,
+      "learning_rate": 2.5318945597485125e-06,
+      "loss": 0.359,
       "step": 119330
     },
     {
-      "epoch": 4.2,
-      "learning_rate": 3.3058639803857568e-06,
-      "loss": 0.2488,
+      "epoch": 4.300825314448409,
+      "grad_norm": 0.25374144315719604,
+      "learning_rate": 2.5306150695132286e-06,
+      "loss": 0.3611,
       "step": 119335
     },
     {
-      "epoch": 4.2,
-      "learning_rate": 3.304448380853542e-06,
-      "loss": 0.2489,
+      "epoch": 4.301005514109633,
+      "grad_norm": 0.2709214389324188,
+      "learning_rate": 2.529335885418965e-06,
+      "loss": 0.391,
       "step": 119340
     },
     {
-      "epoch": 4.2,
-      "learning_rate": 3.3030330630237666e-06,
-      "loss": 0.2465,
+      "epoch": 4.301185713770858,
+      "grad_norm": 0.30983975529670715,
+      "learning_rate": 2.5280570074831505e-06,
+      "loss": 0.3709,
       "step": 119345
     },
     {
-      "epoch": 4.2,
-      "learning_rate": 3.3016180269148018e-06,
-      "loss": 0.2632,
+      "epoch": 4.3013659134320825,
+      "grad_norm": 0.23917073011398315,
+      "learning_rate": 2.5267784357232123e-06,
+      "loss": 0.3768,
       "step": 119350
     },
     {
-      "epoch": 4.2,
-      "learning_rate": 3.3002032725450274e-06,
-      "loss": 0.2499,
+      "epoch": 4.301546113093307,
+      "grad_norm": 0.22569625079631805,
+      "learning_rate": 2.525500170156564e-06,
+      "loss": 0.3792,
       "step": 119355
     },
     {
-      "epoch": 4.2,
-      "learning_rate": 3.2987887999328173e-06,
-      "loss": 0.2446,
+      "epoch": 4.301726312754532,
+      "grad_norm": 0.19801470637321472,
+      "learning_rate": 2.5242222108006196e-06,
+      "loss": 0.3409,
       "step": 119360
     },
     {
-      "epoch": 4.2,
-      "learning_rate": 3.2973746090965347e-06,
-      "loss": 0.2313,
+      "epoch": 4.301906512415757,
+      "grad_norm": 0.3329550623893738,
+      "learning_rate": 2.5229445576728044e-06,
+      "loss": 0.339,
       "step": 119365
     },
     {
-      "epoch": 4.2,
-      "learning_rate": 3.2959607000545346e-06,
-      "loss": 0.2592,
+      "epoch": 4.3020867120769815,
+      "grad_norm": 0.2501632273197174,
+      "learning_rate": 2.5216672107905153e-06,
+      "loss": 0.3723,
       "step": 119370
     },
     {
-      "epoch": 4.2,
-      "learning_rate": 3.2945470728251855e-06,
-      "loss": 0.2602,
+      "epoch": 4.302266911738206,
+      "grad_norm": 0.255783349275589,
+      "learning_rate": 2.5203901701711525e-06,
+      "loss": 0.3433,
       "step": 119375
     },
     {
-      "epoch": 4.2,
-      "learning_rate": 3.2931337274268393e-06,
-      "loss": 0.2268,
+      "epoch": 4.302447111399431,
+      "grad_norm": 0.30260223150253296,
+      "learning_rate": 2.5191134358321267e-06,
+      "loss": 0.3702,
       "step": 119380
     },
     {
-      "epoch": 4.2,
-      "learning_rate": 3.2917206638778453e-06,
-      "loss": 0.2442,
+      "epoch": 4.302627311060656,
+      "grad_norm": 0.33803072571754456,
+      "learning_rate": 2.5178370077908297e-06,
+      "loss": 0.374,
       "step": 119385
     },
     {
-      "epoch": 4.2,
-      "learning_rate": 3.29030788219655e-06,
-      "loss": 0.2654,
+      "epoch": 4.3028075107218795,
+      "grad_norm": 0.20354852080345154,
+      "learning_rate": 2.5165608860646476e-06,
+      "loss": 0.3368,
       "step": 119390
     },
     {
-      "epoch": 4.2,
-      "learning_rate": 3.2888953824013026e-06,
-      "loss": 0.2611,
+      "epoch": 4.302987710383104,
+      "grad_norm": 0.2563473880290985,
+      "learning_rate": 2.515285070670972e-06,
+      "loss": 0.4105,
       "step": 119395
     },
     {
-      "epoch": 4.2,
-      "learning_rate": 3.2874831645104383e-06,
-      "loss": 0.2603,
+      "epoch": 4.303167910044329,
+      "grad_norm": 0.2289261668920517,
+      "learning_rate": 2.5140095616271816e-06,
+      "loss": 0.3794,
       "step": 119400
     },
     {
-      "epoch": 4.2,
-      "learning_rate": 3.286071228542295e-06,
-      "loss": 0.2546,
+      "epoch": 4.303348109705554,
+      "grad_norm": 0.1935594230890274,
+      "learning_rate": 2.5127343589506607e-06,
+      "loss": 0.3595,
       "step": 119405
     },
     {
-      "epoch": 4.2,
-      "learning_rate": 3.2846595745152033e-06,
-      "loss": 0.2498,
+      "epoch": 4.303528309366778,
+      "grad_norm": 0.2854008078575134,
+      "learning_rate": 2.51145946265878e-06,
+      "loss": 0.399,
       "step": 119410
     },
     {
-      "epoch": 4.2,
-      "learning_rate": 3.283248202447503e-06,
-      "loss": 0.277,
+      "epoch": 4.303708509028003,
+      "grad_norm": 0.22041873633861542,
+      "learning_rate": 2.510184872768909e-06,
+      "loss": 0.3513,
       "step": 119415
     },
     {
-      "epoch": 4.2,
-      "learning_rate": 3.2818371123575136e-06,
-      "loss": 0.2537,
+      "epoch": 4.303888708689228,
+      "grad_norm": 0.2398824393749237,
+      "learning_rate": 2.508910589298416e-06,
+      "loss": 0.3586,
       "step": 119420
     },
     {
-      "epoch": 4.2,
-      "learning_rate": 3.280426304263548e-06,
-      "loss": 0.2423,
+      "epoch": 4.304068908350453,
+      "grad_norm": 0.2502867877483368,
+      "learning_rate": 2.5076366122646577e-06,
+      "loss": 0.3853,
       "step": 119425
     },
     {
-      "epoch": 4.2,
-      "learning_rate": 3.2790157781839414e-06,
-      "loss": 0.2577,
+      "epoch": 4.3042491080116765,
+      "grad_norm": 0.29783013463020325,
+      "learning_rate": 2.5063629416850006e-06,
+      "loss": 0.3828,
       "step": 119430
     },
     {
-      "epoch": 4.2,
-      "learning_rate": 3.2776055341370014e-06,
-      "loss": 0.2648,
+      "epoch": 4.304429307672901,
+      "grad_norm": 0.22847159206867218,
+      "learning_rate": 2.505089577576797e-06,
+      "loss": 0.3618,
       "step": 119435
     },
     {
-      "epoch": 4.2,
-      "learning_rate": 3.2761955721410357e-06,
-      "loss": 0.225,
+      "epoch": 4.304609507334126,
+      "grad_norm": 0.2972005009651184,
+      "learning_rate": 2.503816519957383e-06,
+      "loss": 0.3382,
       "step": 119440
     },
     {
-      "epoch": 4.2,
-      "learning_rate": 3.2747858922143493e-06,
-      "loss": 0.2659,
+      "epoch": 4.304789706995351,
+      "grad_norm": 0.26005420088768005,
+      "learning_rate": 2.5025437688441206e-06,
+      "loss": 0.3663,
       "step": 119445
     },
     {
-      "epoch": 4.2,
-      "learning_rate": 3.273376494375255e-06,
-      "loss": 0.28,
+      "epoch": 4.304969906656575,
+      "grad_norm": 0.23355317115783691,
+      "learning_rate": 2.5012713242543422e-06,
+      "loss": 0.3487,
       "step": 119450
     },
     {
-      "epoch": 4.2,
-      "learning_rate": 3.271967378642049e-06,
-      "loss": 0.2796,
+      "epoch": 4.3051501063178,
+      "grad_norm": 0.23067983984947205,
+      "learning_rate": 2.4999991862053818e-06,
+      "loss": 0.3627,
       "step": 119455
     },
     {
-      "epoch": 4.2,
-      "learning_rate": 3.270558545033023e-06,
-      "loss": 0.2658,
+      "epoch": 4.305330305979025,
+      "grad_norm": 0.28559330105781555,
+      "learning_rate": 2.498727354714586e-06,
+      "loss": 0.3992,
       "step": 119460
     },
     {
-      "epoch": 4.2,
-      "learning_rate": 3.2691499935664784e-06,
-      "loss": 0.2435,
+      "epoch": 4.30551050564025,
+      "grad_norm": 0.20493915677070618,
+      "learning_rate": 2.497455829799261e-06,
+      "loss": 0.3717,
       "step": 119465
     },
     {
-      "epoch": 4.2,
-      "learning_rate": 3.267741724260695e-06,
-      "loss": 0.2611,
+      "epoch": 4.305690705301474,
+      "grad_norm": 0.2629398703575134,
+      "learning_rate": 2.496184611476754e-06,
+      "loss": 0.3634,
       "step": 119470
     },
     {
-      "epoch": 4.2,
-      "learning_rate": 3.266333737133967e-06,
-      "loss": 0.2511,
+      "epoch": 4.305870904962699,
+      "grad_norm": 0.2835977077484131,
+      "learning_rate": 2.49491369976437e-06,
+      "loss": 0.3968,
       "step": 119475
     },
     {
-      "epoch": 4.2,
-      "learning_rate": 3.264926032204574e-06,
-      "loss": 0.2607,
+      "epoch": 4.306051104623923,
+      "grad_norm": 0.29381829500198364,
+      "learning_rate": 2.4936430946794322e-06,
+      "loss": 0.39,
       "step": 119480
     },
     {
-      "epoch": 4.2,
-      "learning_rate": 3.2635186094907872e-06,
-      "loss": 0.2444,
+      "epoch": 4.306231304285148,
+      "grad_norm": 0.21818934381008148,
+      "learning_rate": 2.492372796239251e-06,
+      "loss": 0.3773,
       "step": 119485
     },
     {
-      "epoch": 4.2,
-      "learning_rate": 3.262111469010892e-06,
-      "loss": 0.2625,
+      "epoch": 4.306411503946372,
+      "grad_norm": 0.3064677119255066,
+      "learning_rate": 2.491102804461129e-06,
+      "loss": 0.3716,
       "step": 119490
     },
     {
-      "epoch": 4.2,
-      "learning_rate": 3.2607046107831522e-06,
-      "loss": 0.2604,
+      "epoch": 4.306591703607597,
+      "grad_norm": 0.20513369143009186,
+      "learning_rate": 2.4898331193623724e-06,
+      "loss": 0.3567,
       "step": 119495
     },
     {
-      "epoch": 4.2,
-      "learning_rate": 3.259298034825836e-06,
-      "loss": 0.2672,
+      "epoch": 4.306771903268822,
+      "grad_norm": 0.28164172172546387,
+      "learning_rate": 2.488563740960284e-06,
+      "loss": 0.3665,
       "step": 119500
     },
     {
-      "epoch": 4.2,
-      "eval_loss": 0.24933363497257233,
-      "eval_runtime": 10.5337,
-      "eval_samples_per_second": 9.493,
-      "eval_steps_per_second": 9.493,
+      "epoch": 4.306771903268822,
+      "eval_loss": 0.4290541708469391,
+      "eval_runtime": 3.5337,
+      "eval_samples_per_second": 28.299,
+      "eval_steps_per_second": 7.075,
       "step": 119500
     },
     {
-      "epoch": 4.2,
-      "learning_rate": 3.257891741157204e-06,
-      "loss": 0.2517,
+      "epoch": 4.306952102930047,
+      "grad_norm": 0.24295520782470703,
+      "learning_rate": 2.4872946692721554e-06,
+      "loss": 0.346,
       "step": 119505
     },
     {
-      "epoch": 4.2,
-      "learning_rate": 3.2564857297955247e-06,
-      "loss": 0.265,
+      "epoch": 4.307132302591271,
+      "grad_norm": 0.29755252599716187,
+      "learning_rate": 2.4860259043152813e-06,
+      "loss": 0.3668,
       "step": 119510
     },
     {
-      "epoch": 4.2,
-      "learning_rate": 3.255080000759042e-06,
-      "loss": 0.2615,
+      "epoch": 4.307312502252496,
+      "grad_norm": 0.24635331332683563,
+      "learning_rate": 2.4847574461069446e-06,
+      "loss": 0.3949,
       "step": 119515
     },
     {
-      "epoch": 4.21,
-      "learning_rate": 3.253674554066019e-06,
-      "loss": 0.2522,
+      "epoch": 4.307492701913721,
+      "grad_norm": 0.2744773328304291,
+      "learning_rate": 2.483489294664423e-06,
+      "loss": 0.4012,
       "step": 119520
     },
     {
-      "epoch": 4.21,
-      "learning_rate": 3.252269389734697e-06,
-      "loss": 0.2523,
+      "epoch": 4.307672901574945,
+      "grad_norm": 0.26855769753456116,
+      "learning_rate": 2.482221450005004e-06,
+      "loss": 0.37,
       "step": 119525
     },
     {
-      "epoch": 4.21,
-      "learning_rate": 3.25086450778333e-06,
-      "loss": 0.2653,
+      "epoch": 4.307853101236169,
+      "grad_norm": 0.2621409595012665,
+      "learning_rate": 2.4809539121459613e-06,
+      "loss": 0.3522,
       "step": 119530
     },
     {
-      "epoch": 4.21,
-      "learning_rate": 3.249459908230154e-06,
-      "loss": 0.2465,
+      "epoch": 4.308033300897394,
+      "grad_norm": 0.2507140040397644,
+      "learning_rate": 2.4796866811045595e-06,
+      "loss": 0.3455,
       "step": 119535
     },
     {
-      "epoch": 4.21,
-      "learning_rate": 3.2480555910934017e-06,
-      "loss": 0.2678,
+      "epoch": 4.308213500558619,
+      "grad_norm": 0.26044005155563354,
+      "learning_rate": 2.478419756898065e-06,
+      "loss": 0.3529,
       "step": 119540
     },
     {
-      "epoch": 4.21,
-      "learning_rate": 3.2466515563913167e-06,
-      "loss": 0.2462,
+      "epoch": 4.308393700219844,
+      "grad_norm": 0.2704107165336609,
+      "learning_rate": 2.4771531395437394e-06,
+      "loss": 0.3842,
       "step": 119545
     },
     {
-      "epoch": 4.21,
-      "learning_rate": 3.2452478041421263e-06,
-      "loss": 0.2608,
+      "epoch": 4.308573899881068,
+      "grad_norm": 0.23864923417568207,
+      "learning_rate": 2.4758868290588487e-06,
+      "loss": 0.371,
       "step": 119550
     },
     {
-      "epoch": 4.21,
-      "learning_rate": 3.243844334364057e-06,
-      "loss": 0.247,
+      "epoch": 4.308754099542293,
+      "grad_norm": 0.3333123028278351,
+      "learning_rate": 2.4746208254606386e-06,
+      "loss": 0.3874,
       "step": 119555
     },
     {
-      "epoch": 4.21,
-      "learning_rate": 3.242441147075323e-06,
-      "loss": 0.2504,
+      "epoch": 4.308934299203518,
+      "grad_norm": 0.236053004860878,
+      "learning_rate": 2.47335512876635e-06,
+      "loss": 0.3794,
       "step": 119560
     },
     {
-      "epoch": 4.21,
-      "learning_rate": 3.241038242294159e-06,
-      "loss": 0.2384,
+      "epoch": 4.3091144988647425,
+      "grad_norm": 0.24085481464862823,
+      "learning_rate": 2.4720897389932413e-06,
+      "loss": 0.3613,
       "step": 119565
     },
     {
-      "epoch": 4.21,
-      "learning_rate": 3.239635620038775e-06,
-      "loss": 0.2452,
+      "epoch": 4.309294698525966,
+      "grad_norm": 0.2732737064361572,
+      "learning_rate": 2.470824656158549e-06,
+      "loss": 0.383,
       "step": 119570
     },
     {
-      "epoch": 4.21,
-      "learning_rate": 3.238233280327374e-06,
-      "loss": 0.2408,
+      "epoch": 4.309474898187191,
+      "grad_norm": 0.23329199850559235,
+      "learning_rate": 2.4695598802795096e-06,
+      "loss": 0.3829,
       "step": 119575
     },
     {
-      "epoch": 4.21,
-      "learning_rate": 3.2368312231781746e-06,
-      "loss": 0.2511,
+      "epoch": 4.309655097848416,
+      "grad_norm": 0.22309060394763947,
+      "learning_rate": 2.4682954113733557e-06,
+      "loss": 0.369,
       "step": 119580
     },
     {
-      "epoch": 4.21,
-      "learning_rate": 3.235429448609384e-06,
-      "loss": 0.2409,
+      "epoch": 4.309835297509641,
+      "grad_norm": 0.24578943848609924,
+      "learning_rate": 2.4670312494573076e-06,
+      "loss": 0.4047,
       "step": 119585
     },
     {
-      "epoch": 4.21,
-      "learning_rate": 3.234027956639199e-06,
-      "loss": 0.2468,
+      "epoch": 4.310015497170865,
+      "grad_norm": 0.304373562335968,
+      "learning_rate": 2.4657673945486045e-06,
+      "loss": 0.3447,
       "step": 119590
     },
     {
-      "epoch": 4.21,
-      "learning_rate": 3.2326267472858164e-06,
-      "loss": 0.2512,
+      "epoch": 4.31019569683209,
+      "grad_norm": 0.32049259543418884,
+      "learning_rate": 2.4645038466644538e-06,
+      "loss": 0.4057,
       "step": 119595
     },
     {
-      "epoch": 4.21,
-      "learning_rate": 3.231225820567424e-06,
-      "loss": 0.2381,
+      "epoch": 4.310375896493315,
+      "grad_norm": 0.22794209420681,
+      "learning_rate": 2.4632406058220758e-06,
+      "loss": 0.3567,
       "step": 119600
     },
     {
-      "epoch": 4.21,
-      "learning_rate": 3.2298251765022243e-06,
-      "loss": 0.2532,
+      "epoch": 4.3105560961545395,
+      "grad_norm": 0.21800605952739716,
+      "learning_rate": 2.4619776720386846e-06,
+      "loss": 0.3863,
       "step": 119605
     },
     {
-      "epoch": 4.21,
-      "learning_rate": 3.2284248151083967e-06,
-      "loss": 0.2436,
+      "epoch": 4.310736295815764,
+      "grad_norm": 0.23417772352695465,
+      "learning_rate": 2.460715045331477e-06,
+      "loss": 0.3833,
       "step": 119610
     },
     {
-      "epoch": 4.21,
-      "learning_rate": 3.2270247364041246e-06,
-      "loss": 0.2452,
+      "epoch": 4.310916495476988,
+      "grad_norm": 0.26621463894844055,
+      "learning_rate": 2.4594527257176676e-06,
+      "loss": 0.3761,
       "step": 119615
     },
     {
-      "epoch": 4.21,
-      "learning_rate": 3.2256249404075846e-06,
-      "loss": 0.2474,
+      "epoch": 4.311096695138213,
+      "grad_norm": 0.23449282348155975,
+      "learning_rate": 2.4581907132144506e-06,
+      "loss": 0.35,
       "step": 119620
     },
     {
-      "epoch": 4.21,
-      "learning_rate": 3.2242254271369593e-06,
-      "loss": 0.2411,
+      "epoch": 4.3112768947994375,
+      "grad_norm": 0.2845596671104431,
+      "learning_rate": 2.4569290078390205e-06,
+      "loss": 0.3927,
       "step": 119625
     },
     {
-      "epoch": 4.21,
-      "learning_rate": 3.2228261966104097e-06,
-      "loss": 0.2612,
+      "epoch": 4.311457094460662,
+      "grad_norm": 0.24761004745960236,
+      "learning_rate": 2.4556676096085696e-06,
+      "loss": 0.3448,
       "step": 119630
     },
     {
-      "epoch": 4.21,
-      "learning_rate": 3.2214272488461096e-06,
-      "loss": 0.2593,
+      "epoch": 4.311637294121887,
+      "grad_norm": 0.30386558175086975,
+      "learning_rate": 2.4544065185402837e-06,
+      "loss": 0.3739,
       "step": 119635
     },
     {
-      "epoch": 4.21,
-      "learning_rate": 3.2200285838622305e-06,
-      "loss": 0.2482,
+      "epoch": 4.311817493783112,
+      "grad_norm": 0.32800036668777466,
+      "learning_rate": 2.4531457346513382e-06,
+      "loss": 0.3819,
       "step": 119640
     },
     {
-      "epoch": 4.21,
-      "learning_rate": 3.2186302016769256e-06,
-      "loss": 0.2464,
+      "epoch": 4.3119976934443365,
+      "grad_norm": 0.24963368475437164,
+      "learning_rate": 2.4518852579589247e-06,
+      "loss": 0.3603,
       "step": 119645
     },
     {
-      "epoch": 4.21,
-      "learning_rate": 3.217232102308354e-06,
-      "loss": 0.253,
+      "epoch": 4.312177893105561,
+      "grad_norm": 0.24996010959148407,
+      "learning_rate": 2.450625088480202e-06,
+      "loss": 0.3518,
       "step": 119650
     },
     {
-      "epoch": 4.21,
-      "learning_rate": 3.2158342857746632e-06,
-      "loss": 0.2417,
+      "epoch": 4.312358092766786,
+      "grad_norm": 0.22332359850406647,
+      "learning_rate": 2.44936522623235e-06,
+      "loss": 0.4102,
       "step": 119655
     },
     {
-      "epoch": 4.21,
-      "learning_rate": 3.2144367520940135e-06,
-      "loss": 0.2467,
+      "epoch": 4.312538292428011,
+      "grad_norm": 0.20659637451171875,
+      "learning_rate": 2.448105671232531e-06,
+      "loss": 0.3485,
       "step": 119660
     },
     {
-      "epoch": 4.21,
-      "learning_rate": 3.2130395012845455e-06,
-      "loss": 0.2488,
+      "epoch": 4.3127184920892345,
+      "grad_norm": 0.24972330033779144,
+      "learning_rate": 2.4468464234979063e-06,
+      "loss": 0.3618,
       "step": 119665
     },
     {
-      "epoch": 4.21,
-      "learning_rate": 3.2116425333644007e-06,
-      "loss": 0.2447,
+      "epoch": 4.312898691750459,
+      "grad_norm": 0.2645004987716675,
+      "learning_rate": 2.4455874830456334e-06,
+      "loss": 0.3474,
       "step": 119670
     },
     {
-      "epoch": 4.21,
-      "learning_rate": 3.210245848351712e-06,
-      "loss": 0.2906,
+      "epoch": 4.313078891411684,
+      "grad_norm": 0.20380602777004242,
+      "learning_rate": 2.444328849892863e-06,
+      "loss": 0.3755,
       "step": 119675
     },
     {
-      "epoch": 4.21,
-      "learning_rate": 3.208849446264628e-06,
-      "loss": 0.2576,
+      "epoch": 4.313259091072909,
+      "grad_norm": 0.23130224645137787,
+      "learning_rate": 2.4430705240567425e-06,
+      "loss": 0.3706,
       "step": 119680
     },
     {
-      "epoch": 4.21,
-      "learning_rate": 3.207453327121268e-06,
-      "loss": 0.2508,
+      "epoch": 4.3134392907341335,
+      "grad_norm": 0.23018963634967804,
+      "learning_rate": 2.441812505554422e-06,
+      "loss": 0.3971,
       "step": 119685
     },
     {
-      "epoch": 4.21,
-      "learning_rate": 3.2060574909397677e-06,
-      "loss": 0.2648,
+      "epoch": 4.313619490395358,
+      "grad_norm": 0.26062634587287903,
+      "learning_rate": 2.4405547944030383e-06,
+      "loss": 0.3614,
       "step": 119690
     },
     {
-      "epoch": 4.21,
-      "learning_rate": 3.204661937738246e-06,
-      "loss": 0.2435,
+      "epoch": 4.313799690056583,
+      "grad_norm": 0.2006555050611496,
+      "learning_rate": 2.4392973906197297e-06,
+      "loss": 0.3731,
       "step": 119695
     },
     {
-      "epoch": 4.21,
-      "learning_rate": 3.2032666675348266e-06,
-      "loss": 0.2678,
+      "epoch": 4.313979889717808,
+      "grad_norm": 0.2953641712665558,
+      "learning_rate": 2.4380402942216245e-06,
+      "loss": 0.3982,
       "step": 119700
     },
     {
-      "epoch": 4.21,
-      "learning_rate": 3.201871680347629e-06,
-      "loss": 0.2452,
+      "epoch": 4.3141600893790315,
+      "grad_norm": 0.23595255613327026,
+      "learning_rate": 2.436783505225848e-06,
+      "loss": 0.3646,
       "step": 119705
     },
     {
-      "epoch": 4.21,
-      "learning_rate": 3.2004769761947604e-06,
-      "loss": 0.2479,
+      "epoch": 4.314340289040256,
+      "grad_norm": 0.265407919883728,
+      "learning_rate": 2.4355270236495338e-06,
+      "loss": 0.3473,
       "step": 119710
     },
     {
-      "epoch": 4.21,
-      "learning_rate": 3.1990825550943264e-06,
-      "loss": 0.2468,
+      "epoch": 4.314520488701481,
+      "grad_norm": 0.19755345582962036,
+      "learning_rate": 2.434270849509793e-06,
+      "loss": 0.4052,
       "step": 119715
     },
     {
-      "epoch": 4.21,
-      "learning_rate": 3.197688417064443e-06,
-      "loss": 0.2438,
+      "epoch": 4.314700688362706,
+      "grad_norm": 0.24743589758872986,
+      "learning_rate": 2.433014982823745e-06,
+      "loss": 0.3699,
       "step": 119720
     },
     {
-      "epoch": 4.21,
-      "learning_rate": 3.1962945621232087e-06,
-      "loss": 0.229,
+      "epoch": 4.31488088802393,
+      "grad_norm": 0.38127291202545166,
+      "learning_rate": 2.431759423608496e-06,
+      "loss": 0.3905,
       "step": 119725
     },
     {
-      "epoch": 4.21,
-      "learning_rate": 3.1949009902887185e-06,
-      "loss": 0.248,
+      "epoch": 4.315061087685155,
+      "grad_norm": 0.24170427024364471,
+      "learning_rate": 2.430504171881154e-06,
+      "loss": 0.382,
       "step": 119730
     },
     {
-      "epoch": 4.21,
-      "learning_rate": 3.1935077015790655e-06,
-      "loss": 0.2514,
+      "epoch": 4.31524128734638,
+      "grad_norm": 0.2505660951137543,
+      "learning_rate": 2.4292492276588306e-06,
+      "loss": 0.3779,
       "step": 119735
     },
     {
-      "epoch": 4.21,
-      "learning_rate": 3.1921146960123437e-06,
-      "loss": 0.2503,
+      "epoch": 4.315421487007605,
+      "grad_norm": 0.2789454162120819,
+      "learning_rate": 2.4279945909586125e-06,
+      "loss": 0.3524,
       "step": 119740
     },
     {
-      "epoch": 4.21,
-      "learning_rate": 3.1907219736066448e-06,
-      "loss": 0.242,
+      "epoch": 4.315601686668829,
+      "grad_norm": 0.2678326964378357,
+      "learning_rate": 2.4267402617975944e-06,
+      "loss": 0.3817,
       "step": 119745
     },
     {
-      "epoch": 4.21,
-      "learning_rate": 3.1893295343800424e-06,
-      "loss": 0.2572,
+      "epoch": 4.315781886330054,
+      "grad_norm": 0.2486988753080368,
+      "learning_rate": 2.4254862401928728e-06,
+      "loss": 0.3714,
       "step": 119750
     },
     {
-      "epoch": 4.21,
-      "learning_rate": 3.187937378350628e-06,
-      "loss": 0.2471,
+      "epoch": 4.315962085991278,
+      "grad_norm": 0.27803459763526917,
+      "learning_rate": 2.4242325261615267e-06,
+      "loss": 0.37,
       "step": 119755
     },
     {
-      "epoch": 4.21,
-      "learning_rate": 3.186545505536473e-06,
-      "loss": 0.2571,
+      "epoch": 4.316142285652503,
+      "grad_norm": 0.22846895456314087,
+      "learning_rate": 2.4229791197206502e-06,
+      "loss": 0.3555,
       "step": 119760
     },
     {
-      "epoch": 4.21,
-      "learning_rate": 3.185153915955649e-06,
-      "loss": 0.262,
+      "epoch": 4.316322485313727,
+      "grad_norm": 0.2598210573196411,
+      "learning_rate": 2.4217260208873077e-06,
+      "loss": 0.3441,
       "step": 119765
     },
     {
-      "epoch": 4.21,
-      "learning_rate": 3.1837626096262175e-06,
-      "loss": 0.2488,
+      "epoch": 4.316502684974952,
+      "grad_norm": 0.25162217020988464,
+      "learning_rate": 2.4204732296785743e-06,
+      "loss": 0.3937,
       "step": 119770
     },
     {
-      "epoch": 4.21,
-      "learning_rate": 3.18237158656626e-06,
-      "loss": 0.2683,
+      "epoch": 4.316682884636177,
+      "grad_norm": 0.25310778617858887,
+      "learning_rate": 2.419220746111525e-06,
+      "loss": 0.3502,
       "step": 119775
     },
     {
-      "epoch": 4.21,
-      "learning_rate": 3.180980846793827e-06,
-      "loss": 0.2407,
+      "epoch": 4.316863084297402,
+      "grad_norm": 0.2426055669784546,
+      "learning_rate": 2.417968570203219e-06,
+      "loss": 0.365,
       "step": 119780
     },
     {
-      "epoch": 4.21,
-      "learning_rate": 3.179590390326978e-06,
-      "loss": 0.2662,
+      "epoch": 4.317043283958626,
+      "grad_norm": Infinity,
+      "learning_rate": 2.416967051002339e-06,
+      "loss": 0.3506,
       "step": 119785
     },
     {
-      "epoch": 4.21,
-      "learning_rate": 3.178200217183763e-06,
-      "loss": 0.2616,
+      "epoch": 4.317223483619851,
+      "grad_norm": 0.2902944087982178,
+      "learning_rate": 2.415715428922766e-06,
+      "loss": 0.3559,
       "step": 119790
     },
     {
-      "epoch": 4.21,
-      "learning_rate": 3.1768103273822377e-06,
-      "loss": 0.2381,
+      "epoch": 4.317403683281076,
+      "grad_norm": 0.2499387264251709,
+      "learning_rate": 2.4144641145497e-06,
+      "loss": 0.3648,
       "step": 119795
     },
     {
-      "epoch": 4.21,
-      "learning_rate": 3.17542072094045e-06,
-      "loss": 0.2518,
+      "epoch": 4.3175838829423,
+      "grad_norm": 0.25404685735702515,
+      "learning_rate": 2.4132131079001807e-06,
+      "loss": 0.3391,
       "step": 119800
     },
     {
-      "epoch": 4.22,
-      "learning_rate": 3.174031397876445e-06,
-      "loss": 0.238,
+      "epoch": 4.317764082603524,
+      "grad_norm": 0.26094871759414673,
+      "learning_rate": 2.411962408991264e-06,
+      "loss": 0.3919,
       "step": 119805
     },
     {
-      "epoch": 4.22,
-      "learning_rate": 3.1726423582082524e-06,
-      "loss": 0.2465,
+      "epoch": 4.317944282264749,
+      "grad_norm": 0.2552245855331421,
+      "learning_rate": 2.4107120178399945e-06,
+      "loss": 0.3532,
       "step": 119810
     },
     {
-      "epoch": 4.22,
-      "learning_rate": 3.171253601953919e-06,
-      "loss": 0.2735,
+      "epoch": 4.318124481925974,
+      "grad_norm": 0.24921070039272308,
+      "learning_rate": 2.409461934463389e-06,
+      "loss": 0.3427,
       "step": 119815
     },
     {
-      "epoch": 4.22,
-      "learning_rate": 3.1698651291314717e-06,
-      "loss": 0.2575,
+      "epoch": 4.318304681587199,
+      "grad_norm": 0.23261171579360962,
+      "learning_rate": 2.408212158878495e-06,
+      "loss": 0.3322,
       "step": 119820
     },
     {
-      "epoch": 4.22,
-      "learning_rate": 3.168476939758941e-06,
-      "loss": 0.2336,
+      "epoch": 4.318484881248423,
+      "grad_norm": 0.23947398364543915,
+      "learning_rate": 2.4069626911023325e-06,
+      "loss": 0.3588,
       "step": 119825
     },
     {
-      "epoch": 4.22,
-      "learning_rate": 3.1670890338543425e-06,
-      "loss": 0.2427,
+      "epoch": 4.318665080909648,
+      "grad_norm": 0.2614345848560333,
+      "learning_rate": 2.40571353115194e-06,
+      "loss": 0.3382,
       "step": 119830
     },
     {
-      "epoch": 4.22,
-      "learning_rate": 3.1657014114357093e-06,
-      "loss": 0.2528,
+      "epoch": 4.318845280570873,
+      "grad_norm": 0.30958092212677,
+      "learning_rate": 2.4044646790443188e-06,
+      "loss": 0.3832,
       "step": 119835
     },
     {
-      "epoch": 4.22,
-      "learning_rate": 3.1643140725210546e-06,
-      "loss": 0.241,
+      "epoch": 4.3190254802320975,
+      "grad_norm": 0.26829853653907776,
+      "learning_rate": 2.4032161347964875e-06,
+      "loss": 0.3675,
       "step": 119840
     },
     {
-      "epoch": 4.22,
-      "learning_rate": 3.1629270171283864e-06,
-      "loss": 0.2535,
+      "epoch": 4.319205679893322,
+      "grad_norm": 0.24423977732658386,
+      "learning_rate": 2.4019678984254696e-06,
+      "loss": 0.3462,
       "step": 119845
     },
     {
-      "epoch": 4.22,
-      "learning_rate": 3.161540245275721e-06,
-      "loss": 0.2415,
+      "epoch": 4.319385879554546,
+      "grad_norm": 0.3319460153579712,
+      "learning_rate": 2.400719969948259e-06,
+      "loss": 0.389,
       "step": 119850
     },
     {
-      "epoch": 4.22,
-      "learning_rate": 3.1601537569810655e-06,
-      "loss": 0.2592,
+      "epoch": 4.319566079215771,
+      "grad_norm": 0.26566433906555176,
+      "learning_rate": 2.399472349381868e-06,
+      "loss": 0.3806,
       "step": 119855
     },
     {
-      "epoch": 4.22,
-      "learning_rate": 3.1587675522624233e-06,
-      "loss": 0.2557,
+      "epoch": 4.319746278876996,
+      "grad_norm": 0.32363948225975037,
+      "learning_rate": 2.3982250367432874e-06,
+      "loss": 0.3628,
       "step": 119860
     },
     {
-      "epoch": 4.22,
-      "learning_rate": 3.1573816311377847e-06,
-      "loss": 0.2379,
+      "epoch": 4.31992647853822,
+      "grad_norm": 0.2810172736644745,
+      "learning_rate": 2.396978032049513e-06,
+      "loss": 0.3626,
       "step": 119865
     },
     {
-      "epoch": 4.22,
-      "learning_rate": 3.155995993625155e-06,
-      "loss": 0.2406,
+      "epoch": 4.320106678199445,
+      "grad_norm": 0.262736439704895,
+      "learning_rate": 2.3957313353175387e-06,
+      "loss": 0.3693,
       "step": 119870
     },
     {
-      "epoch": 4.22,
-      "learning_rate": 3.154610639742522e-06,
-      "loss": 0.2616,
+      "epoch": 4.32028687786067,
+      "grad_norm": 0.2636507749557495,
+      "learning_rate": 2.3944849465643488e-06,
+      "loss": 0.3753,
       "step": 119875
     },
     {
-      "epoch": 4.22,
-      "learning_rate": 3.1532255695078725e-06,
-      "loss": 0.2408,
+      "epoch": 4.3204670775218945,
+      "grad_norm": 0.26590320467948914,
+      "learning_rate": 2.3932388658069264e-06,
+      "loss": 0.3929,
       "step": 119880
     },
     {
-      "epoch": 4.22,
-      "learning_rate": 3.151840782939186e-06,
-      "loss": 0.2526,
+      "epoch": 4.320647277183119,
+      "grad_norm": 0.232346773147583,
+      "learning_rate": 2.3919930930622476e-06,
+      "loss": 0.3505,
       "step": 119885
     },
     {
-      "epoch": 4.22,
-      "learning_rate": 3.150456280054456e-06,
-      "loss": 0.2352,
+      "epoch": 4.320827476844343,
+      "grad_norm": 0.23552198708057404,
+      "learning_rate": 2.390747628347284e-06,
+      "loss": 0.3903,
       "step": 119890
     },
     {
-      "epoch": 4.22,
-      "learning_rate": 3.1490720608716493e-06,
-      "loss": 0.2672,
+      "epoch": 4.321007676505568,
+      "grad_norm": 0.2921289801597595,
+      "learning_rate": 2.389502471679003e-06,
+      "loss": 0.3556,
       "step": 119895
     },
     {
-      "epoch": 4.22,
-      "learning_rate": 3.1476881254087374e-06,
-      "loss": 0.2511,
+      "epoch": 4.321187876166793,
+      "grad_norm": 0.2172670215368271,
+      "learning_rate": 2.3882576230743832e-06,
+      "loss": 0.3701,
       "step": 119900
     },
     {
-      "epoch": 4.22,
-      "learning_rate": 3.1463044736837007e-06,
-      "loss": 0.2399,
+      "epoch": 4.321368075828017,
+      "grad_norm": 0.23092396557331085,
+      "learning_rate": 2.3870130825503656e-06,
+      "loss": 0.3806,
       "step": 119905
     },
     {
-      "epoch": 4.22,
-      "learning_rate": 3.1449211057144905e-06,
-      "loss": 0.248,
+      "epoch": 4.321548275489242,
+      "grad_norm": 0.2696508467197418,
+      "learning_rate": 2.3857688501239234e-06,
+      "loss": 0.375,
       "step": 119910
     },
     {
-      "epoch": 4.22,
-      "learning_rate": 3.1435380215190853e-06,
-      "loss": 0.2494,
+      "epoch": 4.321728475150467,
+      "grad_norm": 0.30084672570228577,
+      "learning_rate": 2.3845249258120012e-06,
+      "loss": 0.3592,
       "step": 119915
     },
     {
-      "epoch": 4.22,
-      "learning_rate": 3.142155221115431e-06,
-      "loss": 0.2433,
+      "epoch": 4.3219086748116915,
+      "grad_norm": 0.23035989701747894,
+      "learning_rate": 2.383281309631544e-06,
+      "loss": 0.4045,
       "step": 119920
     },
     {
-      "epoch": 4.22,
-      "learning_rate": 3.140772704521483e-06,
-      "loss": 0.2539,
+      "epoch": 4.322088874472916,
+      "grad_norm": 0.2685616910457611,
+      "learning_rate": 2.382038001599507e-06,
+      "loss": 0.3809,
       "step": 119925
     },
     {
-      "epoch": 4.22,
-      "learning_rate": 3.139390471755202e-06,
-      "loss": 0.251,
+      "epoch": 4.322269074134141,
+      "grad_norm": 0.2256428748369217,
+      "learning_rate": 2.3807950017328217e-06,
+      "loss": 0.3882,
       "step": 119930
     },
     {
-      "epoch": 4.22,
-      "learning_rate": 3.138008522834526e-06,
-      "loss": 0.2298,
+      "epoch": 4.322449273795366,
+      "grad_norm": 0.2647817134857178,
+      "learning_rate": 2.379552310048419e-06,
+      "loss": 0.3421,
       "step": 119935
     },
     {
-      "epoch": 4.22,
-      "learning_rate": 3.1366268577774057e-06,
-      "loss": 0.2414,
+      "epoch": 4.3226294734565895,
+      "grad_norm": 0.25552016496658325,
+      "learning_rate": 2.378309926563241e-06,
+      "loss": 0.3568,
       "step": 119940
     },
     {
-      "epoch": 4.22,
-      "learning_rate": 3.1352454766017698e-06,
-      "loss": 0.2607,
+      "epoch": 4.322809673117814,
+      "grad_norm": 0.23187574744224548,
+      "learning_rate": 2.3770678512942097e-06,
+      "loss": 0.379,
       "step": 119945
     },
     {
-      "epoch": 4.22,
-      "learning_rate": 3.133864379325566e-06,
-      "loss": 0.232,
+      "epoch": 4.322989872779039,
+      "grad_norm": 0.2112554907798767,
+      "learning_rate": 2.375826084258251e-06,
+      "loss": 0.3505,
       "step": 119950
     },
     {
-      "epoch": 4.22,
-      "learning_rate": 3.132483565966718e-06,
-      "loss": 0.2561,
+      "epoch": 4.323170072440264,
+      "grad_norm": 0.30000877380371094,
+      "learning_rate": 2.3745846254722816e-06,
+      "loss": 0.3723,
       "step": 119955
     },
     {
-      "epoch": 4.22,
-      "learning_rate": 3.1311030365431675e-06,
-      "loss": 0.2499,
+      "epoch": 4.3233502721014885,
+      "grad_norm": 0.20632152259349823,
+      "learning_rate": 2.3733434749532104e-06,
+      "loss": 0.3575,
       "step": 119960
     },
     {
-      "epoch": 4.22,
-      "learning_rate": 3.129722791072823e-06,
-      "loss": 0.2691,
+      "epoch": 4.323530471762713,
+      "grad_norm": 0.2675947844982147,
+      "learning_rate": 2.3721026327179595e-06,
+      "loss": 0.4144,
       "step": 119965
     },
     {
-      "epoch": 4.22,
-      "learning_rate": 3.1283428295736243e-06,
-      "loss": 0.2471,
+      "epoch": 4.323710671423938,
+      "grad_norm": 0.2205601930618286,
+      "learning_rate": 2.37086209878343e-06,
+      "loss": 0.3685,
       "step": 119970
     },
     {
-      "epoch": 4.22,
-      "learning_rate": 3.126963152063478e-06,
-      "loss": 0.2565,
+      "epoch": 4.323890871085163,
+      "grad_norm": 0.24030627310276031,
+      "learning_rate": 2.369621873166522e-06,
+      "loss": 0.3793,
       "step": 119975
     },
     {
-      "epoch": 4.22,
-      "learning_rate": 3.1255837585602936e-06,
-      "loss": 0.2482,
+      "epoch": 4.324071070746387,
+      "grad_norm": 0.2809261679649353,
+      "learning_rate": 2.368381955884136e-06,
+      "loss": 0.3722,
       "step": 119980
     },
     {
-      "epoch": 4.22,
-      "learning_rate": 3.1242046490819935e-06,
-      "loss": 0.2418,
+      "epoch": 4.324251270407611,
+      "grad_norm": 0.2488822638988495,
+      "learning_rate": 2.367142346953158e-06,
+      "loss": 0.3601,
       "step": 119985
     },
     {
-      "epoch": 4.22,
-      "learning_rate": 3.122825823646483e-06,
-      "loss": 0.2433,
+      "epoch": 4.324431470068836,
+      "grad_norm": 0.23414598405361176,
+      "learning_rate": 2.3659030463904884e-06,
+      "loss": 0.3784,
       "step": 119990
     },
     {
-      "epoch": 4.22,
-      "learning_rate": 3.1214472822716584e-06,
-      "loss": 0.2518,
+      "epoch": 4.324611669730061,
+      "grad_norm": 0.27667903900146484,
+      "learning_rate": 2.3646640542130117e-06,
+      "loss": 0.3638,
       "step": 119995
     },
     {
-      "epoch": 4.22,
-      "learning_rate": 3.1200690249754193e-06,
-      "loss": 0.2385,
+      "epoch": 4.3247918693912855,
+      "grad_norm": 0.2562429904937744,
+      "learning_rate": 2.3634253704375967e-06,
+      "loss": 0.3664,
       "step": 120000
     },
     {
-      "epoch": 4.22,
-      "eval_loss": 0.2491784542798996,
-      "eval_runtime": 10.5428,
-      "eval_samples_per_second": 9.485,
-      "eval_steps_per_second": 9.485,
+      "epoch": 4.3247918693912855,
+      "eval_loss": 0.4289790987968445,
+      "eval_runtime": 3.5378,
+      "eval_samples_per_second": 28.266,
+      "eval_steps_per_second": 7.067,
       "step": 120000
     },
     {
-      "epoch": 4.22,
-      "learning_rate": 3.1186910517756684e-06,
-      "loss": 0.2656,
+      "epoch": 4.32497206905251,
+      "grad_norm": 0.27249255776405334,
+      "learning_rate": 2.362186995081134e-06,
+      "loss": 0.3678,
       "step": 120005
     },
     {
-      "epoch": 4.22,
-      "learning_rate": 3.117313362690294e-06,
-      "loss": 0.2626,
+      "epoch": 4.325152268713735,
+      "grad_norm": 0.29627206921577454,
+      "learning_rate": 2.3609489281604837e-06,
+      "loss": 0.3616,
       "step": 120010
     },
     {
-      "epoch": 4.22,
-      "learning_rate": 3.1159359577371787e-06,
-      "loss": 0.2739,
+      "epoch": 4.32533246837496,
+      "grad_norm": 0.27116361260414124,
+      "learning_rate": 2.3597111696925327e-06,
+      "loss": 0.3541,
       "step": 120015
     },
     {
-      "epoch": 4.22,
-      "learning_rate": 3.1145588369342117e-06,
-      "loss": 0.2564,
+      "epoch": 4.325512668036184,
+      "grad_norm": 0.26802965998649597,
+      "learning_rate": 2.3584737196941266e-06,
+      "loss": 0.3786,
       "step": 120020
     },
     {
-      "epoch": 4.22,
-      "learning_rate": 3.1131820002992835e-06,
-      "loss": 0.2396,
+      "epoch": 4.325692867697409,
+      "grad_norm": 0.22563250362873077,
+      "learning_rate": 2.3572365781821286e-06,
+      "loss": 0.4056,
       "step": 120025
     },
     {
-      "epoch": 4.22,
-      "learning_rate": 3.1118054478502605e-06,
-      "loss": 0.2415,
+      "epoch": 4.325873067358633,
+      "grad_norm": 0.205821692943573,
+      "learning_rate": 2.355999745173404e-06,
+      "loss": 0.3754,
       "step": 120030
     },
     {
-      "epoch": 4.22,
-      "learning_rate": 3.1104291796050204e-06,
-      "loss": 0.2329,
+      "epoch": 4.326053267019858,
+      "grad_norm": 0.24381496012210846,
+      "learning_rate": 2.3547632206847998e-06,
+      "loss": 0.3969,
       "step": 120035
     },
     {
-      "epoch": 4.22,
-      "learning_rate": 3.109053195581424e-06,
-      "loss": 0.2654,
+      "epoch": 4.326233466681082,
+      "grad_norm": 0.2163795679807663,
+      "learning_rate": 2.353527004733161e-06,
+      "loss": 0.3793,
       "step": 120040
     },
     {
-      "epoch": 4.22,
-      "learning_rate": 3.1076774957973537e-06,
-      "loss": 0.2385,
+      "epoch": 4.326413666342307,
+      "grad_norm": 0.257968544960022,
+      "learning_rate": 2.3522910973353324e-06,
+      "loss": 0.3811,
       "step": 120045
     },
     {
-      "epoch": 4.22,
-      "learning_rate": 3.1063020802706623e-06,
-      "loss": 0.2725,
+      "epoch": 4.326593866003532,
+      "grad_norm": 0.229737788438797,
+      "learning_rate": 2.3510554985081477e-06,
+      "loss": 0.3886,
       "step": 120050
     },
     {
-      "epoch": 4.22,
-      "learning_rate": 3.1049269490192102e-06,
-      "loss": 0.2649,
+      "epoch": 4.326774065664757,
+      "grad_norm": 0.3103449046611786,
+      "learning_rate": 2.349820208268455e-06,
+      "loss": 0.3495,
       "step": 120055
     },
     {
-      "epoch": 4.22,
-      "learning_rate": 3.1035521020608444e-06,
-      "loss": 0.2447,
+      "epoch": 4.326954265325981,
+      "grad_norm": 0.2852785289287567,
+      "learning_rate": 2.348585226633074e-06,
+      "loss": 0.3366,
       "step": 120060
     },
     {
-      "epoch": 4.22,
-      "learning_rate": 3.1021775394134314e-06,
-      "loss": 0.2564,
+      "epoch": 4.327134464987206,
+      "grad_norm": 0.26147526502609253,
+      "learning_rate": 2.3473505536188324e-06,
+      "loss": 0.327,
       "step": 120065
     },
     {
-      "epoch": 4.22,
-      "learning_rate": 3.100803261094809e-06,
-      "loss": 0.252,
+      "epoch": 4.327314664648431,
+      "grad_norm": 0.24618807435035706,
+      "learning_rate": 2.346116189242556e-06,
+      "loss": 0.3706,
       "step": 120070
     },
     {
-      "epoch": 4.22,
-      "learning_rate": 3.0994292671228247e-06,
-      "loss": 0.2462,
+      "epoch": 4.327494864309655,
+      "grad_norm": 0.3258048892021179,
+      "learning_rate": 2.344882133521059e-06,
+      "loss": 0.3921,
       "step": 120075
     },
     {
-      "epoch": 4.22,
-      "learning_rate": 3.098055557515317e-06,
-      "loss": 0.2526,
+      "epoch": 4.327675063970879,
+      "grad_norm": 0.2521205544471741,
+      "learning_rate": 2.3436483864711535e-06,
+      "loss": 0.3588,
       "step": 120080
     },
     {
-      "epoch": 4.22,
-      "learning_rate": 3.096682132290127e-06,
-      "loss": 0.2599,
+      "epoch": 4.327855263632104,
+      "grad_norm": 0.23046831786632538,
+      "learning_rate": 2.342414948109656e-06,
+      "loss": 0.3982,
       "step": 120085
     },
     {
-      "epoch": 4.23,
-      "learning_rate": 3.0953089914650875e-06,
-      "loss": 0.2409,
+      "epoch": 4.328035463293329,
+      "grad_norm": 0.260557621717453,
+      "learning_rate": 2.3411818184533698e-06,
+      "loss": 0.3612,
       "step": 120090
     },
     {
-      "epoch": 4.23,
-      "learning_rate": 3.093936135058015e-06,
-      "loss": 0.2726,
+      "epoch": 4.328215662954554,
+      "grad_norm": 0.20361223816871643,
+      "learning_rate": 2.3399489975190923e-06,
+      "loss": 0.3768,
       "step": 120095
     },
     {
-      "epoch": 4.23,
-      "learning_rate": 3.0925635630867534e-06,
-      "loss": 0.2622,
+      "epoch": 4.328395862615778,
+      "grad_norm": 0.24504904448986053,
+      "learning_rate": 2.3387164853236247e-06,
+      "loss": 0.3511,
       "step": 120100
     },
     {
-      "epoch": 4.23,
-      "learning_rate": 3.0911912755691163e-06,
-      "loss": 0.237,
+      "epoch": 4.328576062277003,
+      "grad_norm": 0.22165676951408386,
+      "learning_rate": 2.3374842818837507e-06,
+      "loss": 0.3731,
       "step": 120105
     },
     {
-      "epoch": 4.23,
-      "learning_rate": 3.089819272522923e-06,
-      "loss": 0.2613,
+      "epoch": 4.328756261938228,
+      "grad_norm": 0.29873546957969666,
+      "learning_rate": 2.3362523872162756e-06,
+      "loss": 0.3618,
       "step": 120110
     },
     {
-      "epoch": 4.23,
-      "learning_rate": 3.0884475539659813e-06,
-      "loss": 0.2473,
+      "epoch": 4.328936461599453,
+      "grad_norm": 0.22169798612594604,
+      "learning_rate": 2.33502080133797e-06,
+      "loss": 0.37,
       "step": 120115
     },
     {
-      "epoch": 4.23,
-      "learning_rate": 3.08707611991611e-06,
-      "loss": 0.2497,
+      "epoch": 4.329116661260677,
+      "grad_norm": 0.19561375677585602,
+      "learning_rate": 2.3337895242656144e-06,
+      "loss": 0.3759,
       "step": 120120
     },
     {
-      "epoch": 4.23,
-      "learning_rate": 3.0857049703911124e-06,
-      "loss": 0.2457,
+      "epoch": 4.329296860921901,
+      "grad_norm": 0.3174193799495697,
+      "learning_rate": 2.33255855601599e-06,
+      "loss": 0.3906,
       "step": 120125
     },
     {
-      "epoch": 4.23,
-      "learning_rate": 3.0843341054087982e-06,
-      "loss": 0.2464,
+      "epoch": 4.329477060583126,
+      "grad_norm": 0.2505806088447571,
+      "learning_rate": 2.33132789660587e-06,
+      "loss": 0.4405,
       "step": 120130
     },
     {
-      "epoch": 4.23,
-      "learning_rate": 3.0829635249869566e-06,
-      "loss": 0.2472,
+      "epoch": 4.329657260244351,
+      "grad_norm": 0.2282123565673828,
+      "learning_rate": 2.330097546052015e-06,
+      "loss": 0.3469,
       "step": 120135
     },
     {
-      "epoch": 4.23,
-      "learning_rate": 3.0815932291433952e-06,
-      "loss": 0.256,
+      "epoch": 4.329837459905575,
+      "grad_norm": 0.2437511831521988,
+      "learning_rate": 2.3288675043711933e-06,
+      "loss": 0.3587,
       "step": 120140
     },
     {
-      "epoch": 4.23,
-      "learning_rate": 3.080223217895903e-06,
-      "loss": 0.2195,
+      "epoch": 4.3300176595668,
+      "grad_norm": 0.20224423706531525,
+      "learning_rate": 2.3276377715801603e-06,
+      "loss": 0.3455,
       "step": 120145
     },
     {
-      "epoch": 4.23,
-      "learning_rate": 3.0788534912622654e-06,
-      "loss": 0.2621,
+      "epoch": 4.330197859228025,
+      "grad_norm": 0.23955215513706207,
+      "learning_rate": 2.3264083476956743e-06,
+      "loss": 0.3537,
       "step": 120150
     },
     {
-      "epoch": 4.23,
-      "learning_rate": 3.0774840492602625e-06,
-      "loss": 0.2544,
+      "epoch": 4.3303780588892495,
+      "grad_norm": 0.23328891396522522,
+      "learning_rate": 2.3251792327344867e-06,
+      "loss": 0.3624,
       "step": 120155
     },
     {
-      "epoch": 4.23,
-      "learning_rate": 3.076114891907686e-06,
-      "loss": 0.2692,
+      "epoch": 4.330558258550474,
+      "grad_norm": 0.20525789260864258,
+      "learning_rate": 2.323950426713342e-06,
+      "loss": 0.3493,
       "step": 120160
     },
     {
-      "epoch": 4.23,
-      "learning_rate": 3.0747460192223103e-06,
-      "loss": 0.2575,
+      "epoch": 4.330738458211698,
+      "grad_norm": 0.252485066652298,
+      "learning_rate": 2.3227219296489827e-06,
+      "loss": 0.3537,
       "step": 120165
     },
     {
-      "epoch": 4.23,
-      "learning_rate": 3.0733774312219104e-06,
-      "loss": 0.265,
+      "epoch": 4.330918657872923,
+      "grad_norm": 0.22659099102020264,
+      "learning_rate": 2.3214937415581424e-06,
+      "loss": 0.3759,
       "step": 120170
     },
     {
-      "epoch": 4.23,
-      "learning_rate": 3.072009127924247e-06,
-      "loss": 0.2286,
+      "epoch": 4.331098857534148,
+      "grad_norm": 0.24077731370925903,
+      "learning_rate": 2.320265862457563e-06,
+      "loss": 0.3789,
       "step": 120175
     },
     {
-      "epoch": 4.23,
-      "learning_rate": 3.0706411093470944e-06,
-      "loss": 0.2429,
+      "epoch": 4.331279057195372,
+      "grad_norm": 0.2883579432964325,
+      "learning_rate": 2.319038292363973e-06,
+      "loss": 0.3513,
       "step": 120180
     },
     {
-      "epoch": 4.23,
-      "learning_rate": 3.0692733755082194e-06,
-      "loss": 0.2679,
+      "epoch": 4.331459256856597,
+      "grad_norm": 0.25911572575569153,
+      "learning_rate": 2.31781103129409e-06,
+      "loss": 0.3691,
       "step": 120185
     },
     {
-      "epoch": 4.23,
-      "learning_rate": 3.0679059264253795e-06,
-      "loss": 0.2519,
+      "epoch": 4.331639456517822,
+      "grad_norm": 0.2714954912662506,
+      "learning_rate": 2.3165840792646414e-06,
+      "loss": 0.3817,
       "step": 120190
     },
     {
-      "epoch": 4.23,
-      "learning_rate": 3.0665387621163192e-06,
-      "loss": 0.2533,
+      "epoch": 4.3318196561790465,
+      "grad_norm": 0.2899724841117859,
+      "learning_rate": 2.3153574362923395e-06,
+      "loss": 0.3776,
       "step": 120195
     },
     {
-      "epoch": 4.23,
-      "learning_rate": 3.065171882598805e-06,
-      "loss": 0.259,
+      "epoch": 4.331999855840271,
+      "grad_norm": 0.2563213109970093,
+      "learning_rate": 2.31413110239391e-06,
+      "loss": 0.4069,
       "step": 120200
     },
     {
-      "epoch": 4.23,
-      "learning_rate": 3.063805287890578e-06,
-      "loss": 0.2365,
+      "epoch": 4.332180055501496,
+      "grad_norm": 0.261004239320755,
+      "learning_rate": 2.3129050775860473e-06,
+      "loss": 0.3644,
       "step": 120205
     },
     {
-      "epoch": 4.23,
-      "learning_rate": 3.062438978009377e-06,
-      "loss": 0.2513,
+      "epoch": 4.332360255162721,
+      "grad_norm": 0.2524876296520233,
+      "learning_rate": 2.3116793618854576e-06,
+      "loss": 0.3632,
       "step": 120210
     },
     {
-      "epoch": 4.23,
-      "learning_rate": 3.061072952972954e-06,
-      "loss": 0.2644,
+      "epoch": 4.332540454823945,
+      "grad_norm": 0.24342192709445953,
+      "learning_rate": 2.3104539553088474e-06,
+      "loss": 0.3615,
       "step": 120215
     },
     {
-      "epoch": 4.23,
-      "learning_rate": 3.05970721279904e-06,
-      "loss": 0.2378,
+      "epoch": 4.332720654485169,
+      "grad_norm": 0.27565255761146545,
+      "learning_rate": 2.3092288578729083e-06,
+      "loss": 0.3985,
       "step": 120220
     },
     {
-      "epoch": 4.23,
-      "learning_rate": 3.05834175750537e-06,
-      "loss": 0.2433,
+      "epoch": 4.332900854146394,
+      "grad_norm": 0.22608241438865662,
+      "learning_rate": 2.308004069594333e-06,
+      "loss": 0.3917,
       "step": 120225
     },
     {
-      "epoch": 4.23,
-      "learning_rate": 3.0569765871096666e-06,
-      "loss": 0.2603,
+      "epoch": 4.333081053807619,
+      "grad_norm": 0.2556675374507904,
+      "learning_rate": 2.3067795904898103e-06,
+      "loss": 0.3498,
       "step": 120230
     },
     {
-      "epoch": 4.23,
-      "learning_rate": 3.0556117016296597e-06,
-      "loss": 0.2841,
+      "epoch": 4.3332612534688435,
+      "grad_norm": 0.2671878933906555,
+      "learning_rate": 2.3055554205760168e-06,
+      "loss": 0.3587,
       "step": 120235
     },
     {
-      "epoch": 4.23,
-      "learning_rate": 3.05424710108308e-06,
-      "loss": 0.2475,
+      "epoch": 4.333441453130068,
+      "grad_norm": 0.232125386595726,
+      "learning_rate": 2.3043315598696406e-06,
+      "loss": 0.3345,
       "step": 120240
     },
     {
-      "epoch": 4.23,
-      "learning_rate": 3.0528827854876374e-06,
-      "loss": 0.2428,
+      "epoch": 4.333621652791293,
+      "grad_norm": 0.268929123878479,
+      "learning_rate": 2.3031080083873556e-06,
+      "loss": 0.4026,
       "step": 120245
     },
     {
-      "epoch": 4.23,
-      "learning_rate": 3.0515187548610437e-06,
-      "loss": 0.2155,
+      "epoch": 4.333801852452518,
+      "grad_norm": 0.2586337625980377,
+      "learning_rate": 2.3018847661458285e-06,
+      "loss": 0.3407,
       "step": 120250
     },
     {
-      "epoch": 4.23,
-      "learning_rate": 3.0501550092210203e-06,
-      "loss": 0.2604,
+      "epoch": 4.333982052113742,
+      "grad_norm": 0.20876720547676086,
+      "learning_rate": 2.300661833161724e-06,
+      "loss": 0.3956,
       "step": 120255
     },
     {
-      "epoch": 4.23,
-      "learning_rate": 3.048791548585267e-06,
-      "loss": 0.2346,
+      "epoch": 4.334162251774966,
+      "grad_norm": 0.26193374395370483,
+      "learning_rate": 2.299439209451712e-06,
+      "loss": 0.3588,
       "step": 120260
     },
     {
-      "epoch": 4.23,
-      "learning_rate": 3.047428372971492e-06,
-      "loss": 0.2381,
+      "epoch": 4.334342451436191,
+      "grad_norm": 0.29305386543273926,
+      "learning_rate": 2.2982168950324374e-06,
+      "loss": 0.3434,
       "step": 120265
     },
     {
-      "epoch": 4.23,
-      "learning_rate": 3.0460654823973837e-06,
-      "loss": 0.2488,
+      "epoch": 4.334522651097416,
+      "grad_norm": 0.1981567144393921,
+      "learning_rate": 2.29699488992057e-06,
+      "loss": 0.3667,
       "step": 120270
     },
     {
-      "epoch": 4.23,
-      "learning_rate": 3.0447028768806535e-06,
-      "loss": 0.2498,
+      "epoch": 4.3347028507586405,
+      "grad_norm": 0.2516523599624634,
+      "learning_rate": 2.29577319413275e-06,
+      "loss": 0.3551,
       "step": 120275
     },
     {
-      "epoch": 4.23,
-      "learning_rate": 3.0433405564389866e-06,
-      "loss": 0.2293,
+      "epoch": 4.334883050419865,
+      "grad_norm": 0.2072906494140625,
+      "learning_rate": 2.2945518076856243e-06,
+      "loss": 0.3863,
       "step": 120280
     },
     {
-      "epoch": 4.23,
-      "learning_rate": 3.041978521090069e-06,
-      "loss": 0.251,
+      "epoch": 4.33506325008109,
+      "grad_norm": 0.2831047475337982,
+      "learning_rate": 2.2933307305958356e-06,
+      "loss": 0.344,
       "step": 120285
     },
     {
-      "epoch": 4.23,
-      "learning_rate": 3.040616770851587e-06,
-      "loss": 0.2576,
+      "epoch": 4.335243449742315,
+      "grad_norm": 0.2764025330543518,
+      "learning_rate": 2.2921099628800126e-06,
+      "loss": 0.4017,
       "step": 120290
     },
     {
-      "epoch": 4.23,
-      "learning_rate": 3.0392553057412316e-06,
-      "loss": 0.2502,
+      "epoch": 4.335423649403539,
+      "grad_norm": 0.2790067195892334,
+      "learning_rate": 2.290889504554808e-06,
+      "loss": 0.3866,
       "step": 120295
     },
     {
-      "epoch": 4.23,
-      "learning_rate": 3.0378941257766745e-06,
-      "loss": 0.2585,
+      "epoch": 4.335603849064764,
+      "grad_norm": 0.20841774344444275,
+      "learning_rate": 2.2896693556368284e-06,
+      "loss": 0.3428,
       "step": 120300
     },
     {
-      "epoch": 4.23,
-      "learning_rate": 3.0365332309755856e-06,
-      "loss": 0.2348,
+      "epoch": 4.335784048725988,
+      "grad_norm": 0.2399793118238449,
+      "learning_rate": 2.2884495161427043e-06,
+      "loss": 0.3393,
       "step": 120305
     },
     {
-      "epoch": 4.23,
-      "learning_rate": 3.0351726213556364e-06,
-      "loss": 0.2497,
+      "epoch": 4.335964248387213,
+      "grad_norm": 0.2118580937385559,
+      "learning_rate": 2.287229986089065e-06,
+      "loss": 0.3936,
       "step": 120310
     },
     {
-      "epoch": 4.23,
-      "learning_rate": 3.033812296934499e-06,
-      "loss": 0.2333,
+      "epoch": 4.3361444480484375,
+      "grad_norm": 0.23589551448822021,
+      "learning_rate": 2.286010765492516e-06,
+      "loss": 0.376,
       "step": 120315
     },
     {
-      "epoch": 4.23,
-      "learning_rate": 3.0324522577298312e-06,
-      "loss": 0.2393,
+      "epoch": 4.336324647709662,
+      "grad_norm": 0.240196093916893,
+      "learning_rate": 2.284791854369675e-06,
+      "loss": 0.3936,
       "step": 120320
     },
     {
-      "epoch": 4.23,
-      "learning_rate": 3.031092503759289e-06,
-      "loss": 0.2626,
+      "epoch": 4.336504847370887,
+      "grad_norm": 0.3579064607620239,
+      "learning_rate": 2.2835732527371478e-06,
+      "loss": 0.3894,
       "step": 120325
     },
     {
-      "epoch": 4.23,
-      "learning_rate": 3.0297330350405382e-06,
-      "loss": 0.226,
+      "epoch": 4.336685047032112,
+      "grad_norm": 0.21416036784648895,
+      "learning_rate": 2.2823549606115324e-06,
+      "loss": 0.3483,
       "step": 120330
     },
     {
-      "epoch": 4.23,
-      "learning_rate": 3.0283738515912262e-06,
-      "loss": 0.2447,
+      "epoch": 4.336865246693336,
+      "grad_norm": 0.2611549496650696,
+      "learning_rate": 2.281136978009435e-06,
+      "loss": 0.4018,
       "step": 120335
     },
     {
-      "epoch": 4.23,
-      "learning_rate": 3.027014953428994e-06,
-      "loss": 0.25,
+      "epoch": 4.337045446354561,
+      "grad_norm": 0.2718494236469269,
+      "learning_rate": 2.2799193049474505e-06,
+      "loss": 0.373,
       "step": 120340
     },
     {
-      "epoch": 4.23,
-      "learning_rate": 3.025656340571495e-06,
-      "loss": 0.2542,
+      "epoch": 4.337225646015786,
+      "grad_norm": 0.2340647578239441,
+      "learning_rate": 2.2787019414421627e-06,
+      "loss": 0.364,
       "step": 120345
     },
     {
-      "epoch": 4.23,
-      "learning_rate": 3.024298013036364e-06,
-      "loss": 0.2484,
+      "epoch": 4.33740584567701,
+      "grad_norm": 0.1981993019580841,
+      "learning_rate": 2.2774848875101647e-06,
+      "loss": 0.3663,
       "step": 120350
     },
     {
-      "epoch": 4.23,
-      "learning_rate": 3.022939970841246e-06,
-      "loss": 0.2477,
+      "epoch": 4.337586045338234,
+      "grad_norm": 0.32858866453170776,
+      "learning_rate": 2.276268143168031e-06,
+      "loss": 0.3422,
       "step": 120355
     },
     {
-      "epoch": 4.23,
-      "learning_rate": 3.021582214003768e-06,
-      "loss": 0.2303,
+      "epoch": 4.337766244999459,
+      "grad_norm": 0.24654696881771088,
+      "learning_rate": 2.275051708432349e-06,
+      "loss": 0.3722,
       "step": 120360
     },
     {
-      "epoch": 4.23,
-      "learning_rate": 3.0202247425415553e-06,
-      "loss": 0.2471,
+      "epoch": 4.337946444660684,
+      "grad_norm": 0.20134994387626648,
+      "learning_rate": 2.2738355833196916e-06,
+      "loss": 0.3416,
       "step": 120365
     },
     {
-      "epoch": 4.23,
-      "learning_rate": 3.0188675564722463e-06,
-      "loss": 0.2564,
+      "epoch": 4.338126644321909,
+      "grad_norm": 0.23259273171424866,
+      "learning_rate": 2.2726197678466178e-06,
+      "loss": 0.3483,
       "step": 120370
     },
     {
-      "epoch": 4.24,
-      "learning_rate": 3.0175106558134546e-06,
-      "loss": 0.2469,
+      "epoch": 4.338306843983133,
+      "grad_norm": 0.25836434960365295,
+      "learning_rate": 2.271404262029703e-06,
+      "loss": 0.3657,
       "step": 120375
     },
     {
-      "epoch": 4.24,
-      "learning_rate": 3.0161540405828e-06,
-      "loss": 0.257,
+      "epoch": 4.338487043644358,
+      "grad_norm": 0.2423553764820099,
+      "learning_rate": 2.2701890658854975e-06,
+      "loss": 0.3692,
       "step": 120380
     },
     {
-      "epoch": 4.24,
-      "learning_rate": 3.0147977107978926e-06,
-      "loss": 0.2453,
+      "epoch": 4.338667243305583,
+      "grad_norm": 0.29900985956192017,
+      "learning_rate": 2.268974179430572e-06,
+      "loss": 0.3787,
       "step": 120385
     },
     {
-      "epoch": 4.24,
-      "learning_rate": 3.0134416664763522e-06,
-      "loss": 0.2795,
+      "epoch": 4.338847442966808,
+      "grad_norm": 0.2342345416545868,
+      "learning_rate": 2.267759602681477e-06,
+      "loss": 0.4216,
       "step": 120390
     },
     {
-      "epoch": 4.24,
-      "learning_rate": 3.0120859076357755e-06,
-      "loss": 0.2609,
+      "epoch": 4.339027642628032,
+      "grad_norm": Infinity,
+      "learning_rate": 2.266788164281511e-06,
+      "loss": 0.399,
       "step": 120395
     },
     {
-      "epoch": 4.24,
-      "learning_rate": 3.010730434293782e-06,
-      "loss": 0.2433,
+      "epoch": 4.339207842289256,
+      "grad_norm": 0.22955337166786194,
+      "learning_rate": 2.265574145044594e-06,
+      "loss": 0.389,
       "step": 120400
     },
     {
-      "epoch": 4.24,
-      "learning_rate": 3.009375246467955e-06,
-      "loss": 0.26,
+      "epoch": 4.339388041950481,
+      "grad_norm": 0.25099387764930725,
+      "learning_rate": 2.264360435559831e-06,
+      "loss": 0.3777,
       "step": 120405
     },
     {
-      "epoch": 4.24,
-      "learning_rate": 3.008020344175902e-06,
-      "loss": 0.2411,
+      "epoch": 4.339568241611706,
+      "grad_norm": 0.23291358351707458,
+      "learning_rate": 2.2631470358437545e-06,
+      "loss": 0.3862,
       "step": 120410
     },
     {
-      "epoch": 4.24,
-      "learning_rate": 3.006665727435212e-06,
-      "loss": 0.2562,
+      "epoch": 4.33974844127293,
+      "grad_norm": 0.2527669370174408,
+      "learning_rate": 2.2619339459128975e-06,
+      "loss": 0.3817,
       "step": 120415
     },
     {
-      "epoch": 4.24,
-      "learning_rate": 3.0053113962634744e-06,
-      "loss": 0.2415,
+      "epoch": 4.339928640934155,
+      "grad_norm": 0.28029200434684753,
+      "learning_rate": 2.2607211657837886e-06,
+      "loss": 0.34,
       "step": 120420
     },
     {
-      "epoch": 4.24,
-      "learning_rate": 3.003957350678269e-06,
-      "loss": 0.2475,
+      "epoch": 4.34010884059538,
+      "grad_norm": 0.2483205646276474,
+      "learning_rate": 2.2595086954729476e-06,
+      "loss": 0.3732,
       "step": 120425
     },
     {
-      "epoch": 4.24,
-      "learning_rate": 3.002603590697184e-06,
-      "loss": 0.256,
+      "epoch": 4.340289040256605,
+      "grad_norm": 0.26079532504081726,
+      "learning_rate": 2.2582965349969035e-06,
+      "loss": 0.4245,
       "step": 120430
     },
     {
-      "epoch": 4.24,
-      "learning_rate": 3.0012501163377974e-06,
-      "loss": 0.2479,
+      "epoch": 4.340469239917829,
+      "grad_norm": 0.3403790593147278,
+      "learning_rate": 2.257084684372163e-06,
+      "loss": 0.347,
       "step": 120435
     },
     {
-      "epoch": 4.24,
-      "learning_rate": 2.999896927617679e-06,
-      "loss": 0.2468,
+      "epoch": 4.340649439579053,
+      "grad_norm": 0.2807793915271759,
+      "learning_rate": 2.255873143615242e-06,
+      "loss": 0.3802,
       "step": 120440
     },
     {
-      "epoch": 4.24,
-      "learning_rate": 2.998544024554395e-06,
-      "loss": 0.242,
+      "epoch": 4.340829639240278,
+      "grad_norm": 0.2575584948062897,
+      "learning_rate": 2.2546619127426477e-06,
+      "loss": 0.3579,
       "step": 120445
     },
     {
-      "epoch": 4.24,
-      "learning_rate": 2.9971914071655227e-06,
-      "loss": 0.2531,
+      "epoch": 4.341009838901503,
+      "grad_norm": 0.2251080423593521,
+      "learning_rate": 2.2534509917708753e-06,
+      "loss": 0.3715,
       "step": 120450
     },
     {
-      "epoch": 4.24,
-      "learning_rate": 2.9958390754686148e-06,
-      "loss": 0.2293,
+      "epoch": 4.341190038562727,
+      "grad_norm": 0.2660796046257019,
+      "learning_rate": 2.2522403807164333e-06,
+      "loss": 0.3834,
       "step": 120455
     },
     {
-      "epoch": 4.24,
-      "learning_rate": 2.9944870294812326e-06,
-      "loss": 0.2333,
+      "epoch": 4.341370238223952,
+      "grad_norm": 0.277935653924942,
+      "learning_rate": 2.251030079595817e-06,
+      "loss": 0.3487,
       "step": 120460
     },
     {
-      "epoch": 4.24,
-      "learning_rate": 2.9931352692209423e-06,
-      "loss": 0.2495,
+      "epoch": 4.341550437885177,
+      "grad_norm": 0.24934552609920502,
+      "learning_rate": 2.249820088425503e-06,
+      "loss": 0.3702,
       "step": 120465
     },
     {
-      "epoch": 4.24,
-      "learning_rate": 2.9917837947052885e-06,
-      "loss": 0.2532,
+      "epoch": 4.3417306375464015,
+      "grad_norm": 0.22956299781799316,
+      "learning_rate": 2.2486104072219905e-06,
+      "loss": 0.3344,
       "step": 120470
     },
     {
-      "epoch": 4.24,
-      "learning_rate": 2.990432605951815e-06,
-      "loss": 0.2596,
+      "epoch": 4.341910837207626,
+      "grad_norm": 0.2705623507499695,
+      "learning_rate": 2.247401036001753e-06,
+      "loss": 0.3638,
       "step": 120475
     },
     {
-      "epoch": 4.24,
-      "learning_rate": 2.9890817029780666e-06,
-      "loss": 0.2376,
+      "epoch": 4.342091036868851,
+      "grad_norm": 0.29147830605506897,
+      "learning_rate": 2.2461919747812753e-06,
+      "loss": 0.3837,
       "step": 120480
     },
     {
-      "epoch": 4.24,
-      "learning_rate": 2.98773108580159e-06,
-      "loss": 0.2323,
+      "epoch": 4.342271236530076,
+      "grad_norm": 0.23217852413654327,
+      "learning_rate": 2.244983223577024e-06,
+      "loss": 0.3905,
       "step": 120485
     },
     {
-      "epoch": 4.24,
-      "learning_rate": 2.9863807544399186e-06,
-      "loss": 0.2409,
+      "epoch": 4.3424514361913,
+      "grad_norm": 0.22803989052772522,
+      "learning_rate": 2.2437747824054694e-06,
+      "loss": 0.3589,
       "step": 120490
     },
     {
-      "epoch": 4.24,
-      "learning_rate": 2.9850307089105856e-06,
-      "loss": 0.24,
+      "epoch": 4.342631635852524,
+      "grad_norm": 0.22938886284828186,
+      "learning_rate": 2.242566651283079e-06,
+      "loss": 0.3986,
       "step": 120495
     },
     {
-      "epoch": 4.24,
-      "learning_rate": 2.9836809492311156e-06,
-      "loss": 0.2306,
+      "epoch": 4.342811835513749,
+      "grad_norm": 0.2516026198863983,
+      "learning_rate": 2.2413588302263143e-06,
+      "loss": 0.3489,
       "step": 120500
     },
     {
-      "epoch": 4.24,
-      "eval_loss": 0.24912191927433014,
-      "eval_runtime": 10.5509,
-      "eval_samples_per_second": 9.478,
-      "eval_steps_per_second": 9.478,
+      "epoch": 4.342811835513749,
+      "eval_loss": 0.4290280044078827,
+      "eval_runtime": 3.5287,
+      "eval_samples_per_second": 28.339,
+      "eval_steps_per_second": 7.085,
       "step": 120500
     },
     {
-      "epoch": 4.24,
-      "learning_rate": 2.982331475419045e-06,
-      "loss": 0.2357,
+      "epoch": 4.342992035174974,
+      "grad_norm": 0.2596825361251831,
+      "learning_rate": 2.2401513192516267e-06,
+      "loss": 0.3906,
       "step": 120505
     },
     {
-      "epoch": 4.24,
-      "learning_rate": 2.980982287491885e-06,
-      "loss": 0.2507,
+      "epoch": 4.3431722348361985,
+      "grad_norm": 0.26794061064720154,
+      "learning_rate": 2.2389441183754745e-06,
+      "loss": 0.4,
       "step": 120510
     },
     {
-      "epoch": 4.24,
-      "learning_rate": 2.979633385467165e-06,
-      "loss": 0.2242,
+      "epoch": 4.343352434497423,
+      "grad_norm": 0.2583240568637848,
+      "learning_rate": 2.2377372276143003e-06,
+      "loss": 0.3953,
       "step": 120515
     },
     {
-      "epoch": 4.24,
-      "learning_rate": 2.9782847693623857e-06,
-      "loss": 0.2441,
+      "epoch": 4.343532634158648,
+      "grad_norm": 0.27646011114120483,
+      "learning_rate": 2.236530646984544e-06,
+      "loss": 0.3915,
       "step": 120520
     },
     {
-      "epoch": 4.24,
-      "learning_rate": 2.9769364391950717e-06,
-      "loss": 0.2661,
+      "epoch": 4.343712833819873,
+      "grad_norm": 0.2590472400188446,
+      "learning_rate": 2.235324376502654e-06,
+      "loss": 0.334,
       "step": 120525
     },
     {
-      "epoch": 4.24,
-      "learning_rate": 2.9755883949827255e-06,
-      "loss": 0.2413,
+      "epoch": 4.3438930334810975,
+      "grad_norm": 0.2593589723110199,
+      "learning_rate": 2.234118416185066e-06,
+      "loss": 0.408,
       "step": 120530
     },
     {
-      "epoch": 4.24,
-      "learning_rate": 2.9742406367428477e-06,
-      "loss": 0.25,
+      "epoch": 4.344073233142321,
+      "grad_norm": 0.2586502134799957,
+      "learning_rate": 2.2329127660482036e-06,
+      "loss": 0.3995,
       "step": 120535
     },
     {
-      "epoch": 4.24,
-      "learning_rate": 2.9728931644929348e-06,
-      "loss": 0.2511,
+      "epoch": 4.344253432803546,
+      "grad_norm": Infinity,
+      "learning_rate": 2.2319484692798763e-06,
+      "loss": 0.3567,
       "step": 120540
     },
     {
-      "epoch": 4.24,
-      "learning_rate": 2.971545978250495e-06,
-      "loss": 0.2275,
+      "epoch": 4.344433632464771,
+      "grad_norm": 0.2285531461238861,
+      "learning_rate": 2.2307433775097196e-06,
+      "loss": 0.3395,
       "step": 120545
     },
     {
-      "epoch": 4.24,
-      "learning_rate": 2.970199078033012e-06,
-      "loss": 0.2579,
+      "epoch": 4.3446138321259955,
+      "grad_norm": 0.28931424021720886,
+      "learning_rate": 2.229538595966277e-06,
+      "loss": 0.3916,
       "step": 120550
     },
     {
-      "epoch": 4.24,
-      "learning_rate": 2.9688524638579712e-06,
-      "loss": 0.232,
+      "epoch": 4.34479403178722,
+      "grad_norm": 0.19079504907131195,
+      "learning_rate": 2.2283341246659623e-06,
+      "loss": 0.3671,
       "step": 120555
     },
     {
-      "epoch": 4.24,
-      "learning_rate": 2.9675061357428647e-06,
-      "loss": 0.2694,
+      "epoch": 4.344974231448445,
+      "grad_norm": 0.33871787786483765,
+      "learning_rate": 2.2271299636251795e-06,
+      "loss": 0.3929,
       "step": 120560
     },
     {
-      "epoch": 4.24,
-      "learning_rate": 2.96616009370517e-06,
-      "loss": 0.2594,
+      "epoch": 4.34515443110967,
+      "grad_norm": 0.2659500241279602,
+      "learning_rate": 2.225926112860349e-06,
+      "loss": 0.3407,
       "step": 120565
     },
     {
-      "epoch": 4.24,
-      "learning_rate": 2.964814337762367e-06,
-      "loss": 0.2464,
+      "epoch": 4.345334630770894,
+      "grad_norm": 0.2704176902770996,
+      "learning_rate": 2.2247225723878663e-06,
+      "loss": 0.3852,
       "step": 120570
     },
     {
-      "epoch": 4.24,
-      "learning_rate": 2.9634688679319263e-06,
-      "loss": 0.2456,
+      "epoch": 4.345514830432119,
+      "grad_norm": 0.27689260244369507,
+      "learning_rate": 2.2235193422241283e-06,
+      "loss": 0.3471,
       "step": 120575
     },
     {
-      "epoch": 4.24,
-      "learning_rate": 2.9621236842313217e-06,
-      "loss": 0.2684,
+      "epoch": 4.345695030093343,
+      "grad_norm": 0.21617692708969116,
+      "learning_rate": 2.2223164223855287e-06,
+      "loss": 0.36,
       "step": 120580
     },
     {
-      "epoch": 4.24,
-      "learning_rate": 2.9607787866780202e-06,
-      "loss": 0.2601,
+      "epoch": 4.345875229754568,
+      "grad_norm": 0.32375189661979675,
+      "learning_rate": 2.221113812888459e-06,
+      "loss": 0.3506,
       "step": 120585
     },
     {
-      "epoch": 4.24,
-      "learning_rate": 2.95943417528948e-06,
-      "loss": 0.2679,
+      "epoch": 4.3460554294157925,
+      "grad_norm": 0.26374006271362305,
+      "learning_rate": 2.2199115137492983e-06,
+      "loss": 0.3691,
       "step": 120590
     },
     {
-      "epoch": 4.24,
-      "learning_rate": 2.9580898500831565e-06,
-      "loss": 0.2576,
+      "epoch": 4.346235629077017,
+      "grad_norm": 0.2784136235713959,
+      "learning_rate": 2.2187095249844385e-06,
+      "loss": 0.3435,
       "step": 120595
     },
     {
-      "epoch": 4.24,
-      "learning_rate": 2.9567458110765166e-06,
-      "loss": 0.2452,
+      "epoch": 4.346415828738242,
+      "grad_norm": 0.23929151892662048,
+      "learning_rate": 2.2175078466102477e-06,
+      "loss": 0.3381,
       "step": 120600
     },
     {
-      "epoch": 4.24,
-      "learning_rate": 2.9554020582870046e-06,
-      "loss": 0.2327,
+      "epoch": 4.346596028399467,
+      "grad_norm": 0.29352447390556335,
+      "learning_rate": 2.216306478643104e-06,
+      "loss": 0.3898,
       "step": 120605
     },
     {
-      "epoch": 4.24,
-      "learning_rate": 2.9540585917320673e-06,
-      "loss": 0.2696,
+      "epoch": 4.346776228060691,
+      "grad_norm": 0.2777535021305084,
+      "learning_rate": 2.2151054210993754e-06,
+      "loss": 0.3625,
       "step": 120610
     },
     {
-      "epoch": 4.24,
-      "learning_rate": 2.9527154114291495e-06,
-      "loss": 0.2686,
+      "epoch": 4.346956427721916,
+      "grad_norm": 0.2904403507709503,
+      "learning_rate": 2.213904673995418e-06,
+      "loss": 0.3525,
       "step": 120615
     },
     {
-      "epoch": 4.24,
-      "learning_rate": 2.9513725173956897e-06,
-      "loss": 0.2216,
+      "epoch": 4.347136627383141,
+      "grad_norm": 0.31510671973228455,
+      "learning_rate": 2.2127042373476016e-06,
+      "loss": 0.3433,
       "step": 120620
     },
     {
-      "epoch": 4.24,
-      "learning_rate": 2.9500299096491323e-06,
-      "loss": 0.2457,
+      "epoch": 4.347316827044365,
+      "grad_norm": 0.26004233956336975,
+      "learning_rate": 2.211504111172283e-06,
+      "loss": 0.3828,
       "step": 120625
     },
     {
-      "epoch": 4.24,
-      "learning_rate": 2.9486875882069027e-06,
-      "loss": 0.2603,
+      "epoch": 4.3474970267055895,
+      "grad_norm": 0.2599365711212158,
+      "learning_rate": 2.210304295485799e-06,
+      "loss": 0.4023,
       "step": 120630
     },
     {
-      "epoch": 4.24,
-      "learning_rate": 2.947345553086431e-06,
-      "loss": 0.2277,
+      "epoch": 4.347677226366814,
+      "grad_norm": 0.2279340624809265,
+      "learning_rate": 2.2091047903045142e-06,
+      "loss": 0.3625,
       "step": 120635
     },
     {
-      "epoch": 4.24,
-      "learning_rate": 2.9460038043051475e-06,
-      "loss": 0.265,
+      "epoch": 4.347857426028039,
+      "grad_norm": 0.2609095871448517,
+      "learning_rate": 2.2079055956447574e-06,
+      "loss": 0.3503,
       "step": 120640
     },
     {
-      "epoch": 4.24,
-      "learning_rate": 2.944662341880472e-06,
-      "loss": 0.2402,
+      "epoch": 4.348037625689264,
+      "grad_norm": 0.2672995328903198,
+      "learning_rate": 2.206706711522882e-06,
+      "loss": 0.3367,
       "step": 120645
     },
     {
-      "epoch": 4.24,
-      "learning_rate": 2.943321165829818e-06,
-      "loss": 0.2363,
+      "epoch": 4.348217825350488,
+      "grad_norm": 0.2778962552547455,
+      "learning_rate": 2.205508137955212e-06,
+      "loss": 0.3601,
       "step": 120650
     },
     {
-      "epoch": 4.24,
-      "learning_rate": 2.9419802761706e-06,
-      "loss": 0.2539,
+      "epoch": 4.348398025011713,
+      "grad_norm": 0.2293660044670105,
+      "learning_rate": 2.204309874958077e-06,
+      "loss": 0.4096,
       "step": 120655
     },
     {
-      "epoch": 4.25,
-      "learning_rate": 2.9406396729202367e-06,
-      "loss": 0.248,
+      "epoch": 4.348578224672938,
+      "grad_norm": 0.22069476544857025,
+      "learning_rate": 2.2031119225478096e-06,
+      "loss": 0.3899,
       "step": 120660
     },
     {
-      "epoch": 4.25,
-      "learning_rate": 2.9392993560961287e-06,
-      "loss": 0.2392,
+      "epoch": 4.348758424334163,
+      "grad_norm": 0.22815968096256256,
+      "learning_rate": 2.201914280740727e-06,
+      "loss": 0.3738,
       "step": 120665
     },
     {
-      "epoch": 4.25,
-      "learning_rate": 2.9379593257156753e-06,
-      "loss": 0.2712,
+      "epoch": 4.348938623995387,
+      "grad_norm": 0.28191712498664856,
+      "learning_rate": 2.2007169495531527e-06,
+      "loss": 0.3879,
       "step": 120670
     },
     {
-      "epoch": 4.25,
-      "learning_rate": 2.9366195817962793e-06,
-      "loss": 0.2678,
+      "epoch": 4.349118823656611,
+      "grad_norm": 0.2501116991043091,
+      "learning_rate": 2.1995199290013926e-06,
+      "loss": 0.3743,
       "step": 120675
     },
     {
-      "epoch": 4.25,
-      "learning_rate": 2.935280124355344e-06,
-      "loss": 0.2323,
+      "epoch": 4.349299023317836,
+      "grad_norm": 0.3013772666454315,
+      "learning_rate": 2.1983232191017556e-06,
+      "loss": 0.3554,
       "step": 120680
     },
     {
-      "epoch": 4.25,
-      "learning_rate": 2.933940953410255e-06,
-      "loss": 0.2584,
+      "epoch": 4.349479222979061,
+      "grad_norm": 0.2710776627063751,
+      "learning_rate": 2.1971268198705543e-06,
+      "loss": 0.3778,
       "step": 120685
     },
     {
-      "epoch": 4.25,
-      "learning_rate": 2.9326020689783963e-06,
-      "loss": 0.247,
+      "epoch": 4.349659422640285,
+      "grad_norm": 0.28486597537994385,
+      "learning_rate": 2.195930731324086e-06,
+      "loss": 0.3988,
       "step": 120690
     },
     {
-      "epoch": 4.25,
-      "learning_rate": 2.9312634710771613e-06,
-      "loss": 0.2425,
+      "epoch": 4.34983962230151,
+      "grad_norm": 0.27696672081947327,
+      "learning_rate": 2.1947349534786443e-06,
+      "loss": 0.3684,
       "step": 120695
     },
     {
-      "epoch": 4.25,
-      "learning_rate": 2.9299251597239256e-06,
-      "loss": 0.2444,
+      "epoch": 4.350019821962735,
+      "grad_norm": 0.2695504426956177,
+      "learning_rate": 2.1935394863505265e-06,
+      "loss": 0.3804,
       "step": 120700
     },
     {
-      "epoch": 4.25,
-      "learning_rate": 2.928587134936067e-06,
-      "loss": 0.2666,
+      "epoch": 4.35020002162396,
+      "grad_norm": 0.2949710786342621,
+      "learning_rate": 2.1923443299560113e-06,
+      "loss": 0.3809,
       "step": 120705
     },
     {
-      "epoch": 4.25,
-      "learning_rate": 2.927249396730955e-06,
-      "loss": 0.2563,
+      "epoch": 4.350380221285184,
+      "grad_norm": 0.29531538486480713,
+      "learning_rate": 2.1911494843113943e-06,
+      "loss": 0.38,
       "step": 120710
     },
     {
-      "epoch": 4.25,
-      "learning_rate": 2.925911945125967e-06,
-      "loss": 0.2358,
+      "epoch": 4.350560420946408,
+      "grad_norm": 0.2712008059024811,
+      "learning_rate": 2.1899549494329513e-06,
+      "loss": 0.358,
       "step": 120715
     },
     {
-      "epoch": 4.25,
-      "learning_rate": 2.9245747801384644e-06,
-      "loss": 0.2709,
+      "epoch": 4.350740620607633,
+      "grad_norm": 0.3000184893608093,
+      "learning_rate": 2.1887607253369525e-06,
+      "loss": 0.3792,
       "step": 120720
     },
     {
-      "epoch": 4.25,
-      "learning_rate": 2.923237901785808e-06,
-      "loss": 0.2481,
+      "epoch": 4.350920820268858,
+      "grad_norm": 0.28387993574142456,
+      "learning_rate": 2.187566812039676e-06,
+      "loss": 0.3642,
       "step": 120725
     },
     {
-      "epoch": 4.25,
-      "learning_rate": 2.9219013100853536e-06,
-      "loss": 0.2693,
+      "epoch": 4.351101019930082,
+      "grad_norm": 0.21037955582141876,
+      "learning_rate": 2.1863732095573825e-06,
+      "loss": 0.3565,
       "step": 120730
     },
     {
-      "epoch": 4.25,
-      "learning_rate": 2.920565005054471e-06,
-      "loss": 0.2394,
+      "epoch": 4.351281219591307,
+      "grad_norm": 0.2745780646800995,
+      "learning_rate": 2.1851799179063326e-06,
+      "loss": 0.3818,
       "step": 120735
     },
     {
-      "epoch": 4.25,
-      "learning_rate": 2.9192289867104983e-06,
-      "loss": 0.2328,
+      "epoch": 4.351461419252532,
+      "grad_norm": 0.25862976908683777,
+      "learning_rate": 2.1839869371028e-06,
+      "loss": 0.3547,
       "step": 120740
     },
     {
-      "epoch": 4.25,
-      "learning_rate": 2.9178932550707834e-06,
-      "loss": 0.2582,
+      "epoch": 4.351641618913757,
+      "grad_norm": 0.2394440770149231,
+      "learning_rate": 2.1827942671630184e-06,
+      "loss": 0.3666,
       "step": 120745
     },
     {
-      "epoch": 4.25,
-      "learning_rate": 2.9165578101526703e-06,
-      "loss": 0.2498,
+      "epoch": 4.351821818574981,
+      "grad_norm": 0.26800236105918884,
+      "learning_rate": 2.181601908103251e-06,
+      "loss": 0.3695,
       "step": 120750
     },
     {
-      "epoch": 4.25,
-      "learning_rate": 2.9152226519735036e-06,
-      "loss": 0.2736,
+      "epoch": 4.352002018236206,
+      "grad_norm": 0.24717827141284943,
+      "learning_rate": 2.1804098599397418e-06,
+      "loss": 0.3604,
       "step": 120755
     },
     {
-      "epoch": 4.25,
-      "learning_rate": 2.913887780550617e-06,
-      "loss": 0.2409,
+      "epoch": 4.352182217897431,
+      "grad_norm": 0.24239477515220642,
+      "learning_rate": 2.179218122688728e-06,
+      "loss": 0.3599,
       "step": 120760
     },
     {
-      "epoch": 4.25,
-      "learning_rate": 2.9125531959013434e-06,
-      "loss": 0.243,
+      "epoch": 4.352362417558655,
+      "grad_norm": 0.28817063570022583,
+      "learning_rate": 2.17802669636645e-06,
+      "loss": 0.4087,
       "step": 120765
     },
     {
-      "epoch": 4.25,
-      "learning_rate": 2.911218898043003e-06,
-      "loss": 0.2517,
+      "epoch": 4.352542617219879,
+      "grad_norm": 0.25521016120910645,
+      "learning_rate": 2.1768355809891396e-06,
+      "loss": 0.3246,
       "step": 120770
     },
     {
-      "epoch": 4.25,
-      "learning_rate": 2.9098848869929334e-06,
-      "loss": 0.2564,
+      "epoch": 4.352722816881104,
+      "grad_norm": 0.23344235122203827,
+      "learning_rate": 2.1756447765730214e-06,
+      "loss": 0.3409,
       "step": 120775
     },
     {
-      "epoch": 4.25,
-      "learning_rate": 2.908551162768447e-06,
-      "loss": 0.2724,
+      "epoch": 4.352903016542329,
+      "grad_norm": 0.27485767006874084,
+      "learning_rate": 2.174454283134328e-06,
+      "loss": 0.3866,
       "step": 120780
     },
     {
-      "epoch": 4.25,
-      "learning_rate": 2.907217725386871e-06,
-      "loss": 0.2543,
+      "epoch": 4.3530832162035535,
+      "grad_norm": 0.24092566967010498,
+      "learning_rate": 2.1732641006892763e-06,
+      "loss": 0.3861,
       "step": 120785
     },
     {
-      "epoch": 4.25,
-      "learning_rate": 2.9058845748655056e-06,
-      "loss": 0.2508,
+      "epoch": 4.353263415864778,
+      "grad_norm": 0.23562981188297272,
+      "learning_rate": 2.1720742292540814e-06,
+      "loss": 0.3637,
       "step": 120790
     },
     {
-      "epoch": 4.25,
-      "learning_rate": 2.9045517112216757e-06,
-      "loss": 0.2684,
+      "epoch": 4.353443615526003,
+      "grad_norm": 0.30748188495635986,
+      "learning_rate": 2.1708846688449553e-06,
+      "loss": 0.3486,
       "step": 120795
     },
     {
-      "epoch": 4.25,
-      "learning_rate": 2.903219134472679e-06,
-      "loss": 0.2584,
+      "epoch": 4.353623815187228,
+      "grad_norm": 0.263886421918869,
+      "learning_rate": 2.1696954194781016e-06,
+      "loss": 0.3717,
       "step": 120800
     },
     {
-      "epoch": 4.25,
-      "learning_rate": 2.901886844635812e-06,
-      "loss": 0.2492,
+      "epoch": 4.3538040148484525,
+      "grad_norm": 0.22052530944347382,
+      "learning_rate": 2.1685064811697325e-06,
+      "loss": 0.3637,
       "step": 120805
     },
     {
-      "epoch": 4.25,
-      "learning_rate": 2.9005548417283897e-06,
-      "loss": 0.2458,
+      "epoch": 4.353984214509676,
+      "grad_norm": 0.2120153307914734,
+      "learning_rate": 2.1673178539360488e-06,
+      "loss": 0.3608,
       "step": 120810
     },
     {
-      "epoch": 4.25,
-      "learning_rate": 2.8992231257676974e-06,
-      "loss": 0.2482,
+      "epoch": 4.354164414170901,
+      "grad_norm": 0.26720255613327026,
+      "learning_rate": 2.1661295377932267e-06,
+      "loss": 0.3854,
       "step": 120815
     },
     {
-      "epoch": 4.25,
-      "learning_rate": 2.8978916967710297e-06,
-      "loss": 0.2641,
+      "epoch": 4.354344613832126,
+      "grad_norm": 0.22795835137367249,
+      "learning_rate": 2.164941532757475e-06,
+      "loss": 0.3628,
       "step": 120820
     },
     {
-      "epoch": 4.25,
-      "learning_rate": 2.8965605547556646e-06,
-      "loss": 0.2626,
+      "epoch": 4.3545248134933505,
+      "grad_norm": 0.33458447456359863,
+      "learning_rate": 2.1637538388449674e-06,
+      "loss": 0.3986,
       "step": 120825
     },
     {
-      "epoch": 4.25,
-      "learning_rate": 2.895229699738902e-06,
-      "loss": 0.26,
+      "epoch": 4.354705013154575,
+      "grad_norm": 0.2997223436832428,
+      "learning_rate": 2.1625664560719043e-06,
+      "loss": 0.3833,
       "step": 120830
     },
     {
-      "epoch": 4.25,
-      "learning_rate": 2.8938991317380065e-06,
-      "loss": 0.2568,
+      "epoch": 4.3548852128158,
+      "grad_norm": 0.2829420864582062,
+      "learning_rate": 2.1613793844544454e-06,
+      "loss": 0.3839,
       "step": 120835
     },
     {
-      "epoch": 4.25,
-      "learning_rate": 2.8925688507702692e-06,
-      "loss": 0.2586,
+      "epoch": 4.355065412477025,
+      "grad_norm": 0.2701405882835388,
+      "learning_rate": 2.160192624008767e-06,
+      "loss": 0.3709,
       "step": 120840
     },
     {
-      "epoch": 4.25,
-      "learning_rate": 2.8912388568529485e-06,
-      "loss": 0.2451,
+      "epoch": 4.3552456121382495,
+      "grad_norm": 0.3344108462333679,
+      "learning_rate": 2.1590061747510465e-06,
+      "loss": 0.3555,
       "step": 120845
     },
     {
-      "epoch": 4.25,
-      "learning_rate": 2.889909150003328e-06,
-      "loss": 0.2543,
+      "epoch": 4.355425811799474,
+      "grad_norm": 0.20211008191108704,
+      "learning_rate": 2.1578200366974386e-06,
+      "loss": 0.4124,
       "step": 120850
     },
     {
-      "epoch": 4.25,
-      "learning_rate": 2.8885797302386665e-06,
-      "loss": 0.2514,
+      "epoch": 4.355606011460699,
+      "grad_norm": 0.25280535221099854,
+      "learning_rate": 2.1566342098641224e-06,
+      "loss": 0.3437,
       "step": 120855
     },
     {
-      "epoch": 4.25,
-      "learning_rate": 2.8872505975762244e-06,
-      "loss": 0.2537,
+      "epoch": 4.355786211121923,
+      "grad_norm": 0.29722049832344055,
+      "learning_rate": 2.1554486942672367e-06,
+      "loss": 0.37,
       "step": 120860
     },
     {
-      "epoch": 4.25,
-      "learning_rate": 2.885921752033255e-06,
-      "loss": 0.248,
+      "epoch": 4.3559664107831475,
+      "grad_norm": 0.22865262627601624,
+      "learning_rate": 2.1542634899229337e-06,
+      "loss": 0.3594,
       "step": 120865
     },
     {
-      "epoch": 4.25,
-      "learning_rate": 2.8845931936270247e-06,
-      "loss": 0.2462,
+      "epoch": 4.356146610444372,
+      "grad_norm": 0.2604643702507019,
+      "learning_rate": 2.1530785968473744e-06,
+      "loss": 0.3679,
       "step": 120870
     },
     {
-      "epoch": 4.25,
-      "learning_rate": 2.883264922374776e-06,
-      "loss": 0.2396,
+      "epoch": 4.356326810105597,
+      "grad_norm": 0.23937147855758667,
+      "learning_rate": 2.1518940150566937e-06,
+      "loss": 0.3803,
       "step": 120875
     },
     {
-      "epoch": 4.25,
-      "learning_rate": 2.8819369382937555e-06,
-      "loss": 0.2467,
+      "epoch": 4.356507009766822,
+      "grad_norm": 0.2754795253276825,
+      "learning_rate": 2.1507097445670337e-06,
+      "loss": 0.3809,
       "step": 120880
     },
     {
-      "epoch": 4.25,
-      "learning_rate": 2.880609241401205e-06,
-      "loss": 0.2532,
+      "epoch": 4.356687209428046,
+      "grad_norm": 0.2818678617477417,
+      "learning_rate": 2.149525785394532e-06,
+      "loss": 0.3701,
       "step": 120885
     },
     {
-      "epoch": 4.25,
-      "learning_rate": 2.8792818317143698e-06,
-      "loss": 0.253,
+      "epoch": 4.356867409089271,
+      "grad_norm": 0.28676289319992065,
+      "learning_rate": 2.148342137555312e-06,
+      "loss": 0.3622,
       "step": 120890
     },
     {
-      "epoch": 4.25,
-      "learning_rate": 2.877954709250477e-06,
-      "loss": 0.2502,
+      "epoch": 4.357047608750496,
+      "grad_norm": 0.23003120720386505,
+      "learning_rate": 2.1471588010655103e-06,
+      "loss": 0.3463,
       "step": 120895
     },
     {
-      "epoch": 4.25,
-      "learning_rate": 2.876627874026769e-06,
-      "loss": 0.2448,
+      "epoch": 4.35722780841172,
+      "grad_norm": 0.2534288465976715,
+      "learning_rate": 2.145975775941245e-06,
+      "loss": 0.359,
       "step": 120900
     },
     {
-      "epoch": 4.25,
-      "learning_rate": 2.875301326060462e-06,
-      "loss": 0.2594,
+      "epoch": 4.3574080080729445,
+      "grad_norm": 0.22617748379707336,
+      "learning_rate": 2.144793062198636e-06,
+      "loss": 0.3883,
       "step": 120905
     },
     {
-      "epoch": 4.25,
-      "learning_rate": 2.873975065368792e-06,
-      "loss": 0.2519,
+      "epoch": 4.357588207734169,
+      "grad_norm": 0.23090511560440063,
+      "learning_rate": 2.143610659853795e-06,
+      "loss": 0.3662,
       "step": 120910
     },
     {
-      "epoch": 4.25,
-      "learning_rate": 2.872649091968976e-06,
-      "loss": 0.2296,
+      "epoch": 4.357768407395394,
+      "grad_norm": 0.26390355825424194,
+      "learning_rate": 2.142428568922836e-06,
+      "loss": 0.4108,
       "step": 120915
     },
     {
-      "epoch": 4.25,
-      "learning_rate": 2.8713234058782207e-06,
-      "loss": 0.2255,
+      "epoch": 4.357948607056619,
+      "grad_norm": 0.2305891066789627,
+      "learning_rate": 2.141246789421855e-06,
+      "loss": 0.3628,
       "step": 120920
     },
     {
-      "epoch": 4.25,
-      "learning_rate": 2.8699980071137554e-06,
-      "loss": 0.26,
+      "epoch": 4.358128806717843,
+      "grad_norm": 0.29624882340431213,
+      "learning_rate": 2.1400653213669715e-06,
+      "loss": 0.4234,
       "step": 120925
     },
     {
-      "epoch": 4.25,
-      "learning_rate": 2.8686728956927813e-06,
-      "loss": 0.2547,
+      "epoch": 4.358309006379068,
+      "grad_norm": 0.24561457335948944,
+      "learning_rate": 2.138884164774263e-06,
+      "loss": 0.3441,
       "step": 120930
     },
     {
-      "epoch": 4.25,
-      "learning_rate": 2.867348071632503e-06,
-      "loss": 0.2509,
+      "epoch": 4.358489206040293,
+      "grad_norm": 0.30524396896362305,
+      "learning_rate": 2.1377033196598367e-06,
+      "loss": 0.3792,
       "step": 120935
     },
     {
-      "epoch": 4.26,
-      "learning_rate": 2.8660235349501213e-06,
-      "loss": 0.2581,
+      "epoch": 4.358669405701518,
+      "grad_norm": 0.22376736998558044,
+      "learning_rate": 2.1365227860397767e-06,
+      "loss": 0.3578,
       "step": 120940
     },
     {
-      "epoch": 4.26,
-      "learning_rate": 2.864699285662839e-06,
-      "loss": 0.2587,
+      "epoch": 4.358849605362742,
+      "grad_norm": 0.2799444794654846,
+      "learning_rate": 2.135342563930165e-06,
+      "loss": 0.396,
       "step": 120945
     },
     {
-      "epoch": 4.26,
-      "learning_rate": 2.863375323787848e-06,
-      "loss": 0.2342,
+      "epoch": 4.359029805023966,
+      "grad_norm": 0.26537925004959106,
+      "learning_rate": 2.1341626533470855e-06,
+      "loss": 0.397,
       "step": 120950
     },
     {
-      "epoch": 4.26,
-      "learning_rate": 2.8620516493423427e-06,
-      "loss": 0.2549,
+      "epoch": 4.359210004685191,
+      "grad_norm": 0.30358922481536865,
+      "learning_rate": 2.1329830543066144e-06,
+      "loss": 0.3729,
       "step": 120955
     },
     {
-      "epoch": 4.26,
-      "learning_rate": 2.8607282623435013e-06,
-      "loss": 0.2453,
+      "epoch": 4.359390204346416,
+      "grad_norm": 0.25468435883522034,
+      "learning_rate": 2.1318037668248165e-06,
+      "loss": 0.39,
       "step": 120960
     },
     {
-      "epoch": 4.26,
-      "learning_rate": 2.859405162808518e-06,
-      "loss": 0.2539,
+      "epoch": 4.35957040400764,
+      "grad_norm": 0.32225528359413147,
+      "learning_rate": 2.1306247909177677e-06,
+      "loss": 0.3589,
       "step": 120965
     },
     {
-      "epoch": 4.26,
-      "learning_rate": 2.858082350754568e-06,
-      "loss": 0.2283,
+      "epoch": 4.359750603668865,
+      "grad_norm": 0.28411632776260376,
+      "learning_rate": 2.1294461266015307e-06,
+      "loss": 0.3462,
       "step": 120970
     },
     {
-      "epoch": 4.26,
-      "learning_rate": 2.8567598261988267e-06,
-      "loss": 0.2295,
+      "epoch": 4.35993080333009,
+      "grad_norm": 0.23393802344799042,
+      "learning_rate": 2.128267773892162e-06,
+      "loss": 0.3878,
       "step": 120975
     },
     {
-      "epoch": 4.26,
-      "learning_rate": 2.8554375891584634e-06,
-      "loss": 0.2649,
+      "epoch": 4.360111002991315,
+      "grad_norm": 0.27324995398521423,
+      "learning_rate": 2.127089732805715e-06,
+      "loss": 0.4043,
       "step": 120980
     },
     {
-      "epoch": 4.26,
-      "learning_rate": 2.854115639650651e-06,
-      "loss": 0.2505,
+      "epoch": 4.360291202652539,
+      "grad_norm": 0.24664054811000824,
+      "learning_rate": 2.125912003358238e-06,
+      "loss": 0.3737,
       "step": 120985
     },
     {
-      "epoch": 4.26,
-      "learning_rate": 2.852793977692553e-06,
-      "loss": 0.2589,
+      "epoch": 4.360471402313763,
+      "grad_norm": 0.22538726031780243,
+      "learning_rate": 2.124734585565788e-06,
+      "loss": 0.3723,
       "step": 120990
     },
     {
-      "epoch": 4.26,
-      "learning_rate": 2.851472603301325e-06,
-      "loss": 0.2311,
+      "epoch": 4.360651601974988,
+      "grad_norm": 0.27115336060523987,
+      "learning_rate": 2.1235574794444046e-06,
+      "loss": 0.3447,
       "step": 120995
     },
     {
-      "epoch": 4.26,
-      "learning_rate": 2.8501515164941346e-06,
-      "loss": 0.2492,
+      "epoch": 4.360831801636213,
+      "grad_norm": 0.19281277060508728,
+      "learning_rate": 2.1223806850101114e-06,
+      "loss": 0.3592,
       "step": 121000
     },
     {
-      "epoch": 4.26,
-      "eval_loss": 0.2490115761756897,
-      "eval_runtime": 10.5538,
-      "eval_samples_per_second": 9.475,
-      "eval_steps_per_second": 9.475,
+      "epoch": 4.360831801636213,
+      "eval_loss": 0.42894721031188965,
+      "eval_runtime": 3.5365,
+      "eval_samples_per_second": 28.276,
+      "eval_steps_per_second": 7.069,
       "step": 121000
     },
     {
-      "epoch": 4.26,
-      "learning_rate": 2.848830717288123e-06,
-      "loss": 0.247,
+      "epoch": 4.361012001297437,
+      "grad_norm": 0.24042321741580963,
+      "learning_rate": 2.1212042022789562e-06,
+      "loss": 0.3635,
       "step": 121005
     },
     {
-      "epoch": 4.26,
-      "learning_rate": 2.8475102057004542e-06,
-      "loss": 0.2259,
+      "epoch": 4.361192200958662,
+      "grad_norm": 0.2706429660320282,
+      "learning_rate": 2.12002803126696e-06,
+      "loss": 0.3986,
       "step": 121010
     },
     {
-      "epoch": 4.26,
-      "learning_rate": 2.8461899817482625e-06,
-      "loss": 0.2403,
+      "epoch": 4.361372400619887,
+      "grad_norm": 0.21886184811592102,
+      "learning_rate": 2.1188521719901574e-06,
+      "loss": 0.328,
       "step": 121015
     },
     {
-      "epoch": 4.26,
-      "learning_rate": 2.8448700454486917e-06,
-      "loss": 0.2519,
+      "epoch": 4.361552600281112,
+      "grad_norm": 0.23916944861412048,
+      "learning_rate": 2.117676624464568e-06,
+      "loss": 0.3908,
       "step": 121020
     },
     {
-      "epoch": 4.26,
-      "learning_rate": 2.8435503968188863e-06,
-      "loss": 0.2418,
+      "epoch": 4.361732799942336,
+      "grad_norm": 0.24652530252933502,
+      "learning_rate": 2.1165013887061963e-06,
+      "loss": 0.3795,
       "step": 121025
     },
     {
-      "epoch": 4.26,
-      "learning_rate": 2.842231035875978e-06,
-      "loss": 0.2397,
+      "epoch": 4.361912999603561,
+      "grad_norm": 0.20936556160449982,
+      "learning_rate": 2.1153264647310654e-06,
+      "loss": 0.3649,
       "step": 121030
     },
     {
-      "epoch": 4.26,
-      "learning_rate": 2.8409119626370913e-06,
-      "loss": 0.2589,
+      "epoch": 4.362093199264786,
+      "grad_norm": 0.2558532953262329,
+      "learning_rate": 2.1141518525551766e-06,
+      "loss": 0.3569,
       "step": 121035
     },
     {
-      "epoch": 4.26,
-      "learning_rate": 2.839593177119365e-06,
-      "loss": 0.2376,
+      "epoch": 4.36227339892601,
+      "grad_norm": 0.2516135573387146,
+      "learning_rate": 2.1129775521945476e-06,
+      "loss": 0.3817,
       "step": 121040
     },
     {
-      "epoch": 4.26,
-      "learning_rate": 2.8382746793399166e-06,
-      "loss": 0.259,
+      "epoch": 4.362453598587234,
+      "grad_norm": 0.22439998388290405,
+      "learning_rate": 2.111803563665166e-06,
+      "loss": 0.3627,
       "step": 121045
     },
     {
-      "epoch": 4.26,
-      "learning_rate": 2.8369564693158684e-06,
-      "loss": 0.2618,
+      "epoch": 4.362633798248459,
+      "grad_norm": 0.2665523886680603,
+      "learning_rate": 2.110629886983023e-06,
+      "loss": 0.3555,
       "step": 121050
     },
     {
-      "epoch": 4.26,
-      "learning_rate": 2.835638547064329e-06,
-      "loss": 0.2712,
+      "epoch": 4.362813997909684,
+      "grad_norm": 0.19356732070446014,
+      "learning_rate": 2.109456522164119e-06,
+      "loss": 0.3745,
       "step": 121055
     },
     {
-      "epoch": 4.26,
-      "learning_rate": 2.8343209126024147e-06,
-      "loss": 0.2358,
+      "epoch": 4.362994197570909,
+      "grad_norm": 0.3014737069606781,
+      "learning_rate": 2.1082834692244424e-06,
+      "loss": 0.3489,
       "step": 121060
     },
     {
-      "epoch": 4.26,
-      "learning_rate": 2.83300356594724e-06,
-      "loss": 0.2489,
+      "epoch": 4.363174397232133,
+      "grad_norm": 0.21742402017116547,
+      "learning_rate": 2.1071107281799675e-06,
+      "loss": 0.3669,
       "step": 121065
     },
     {
-      "epoch": 4.26,
-      "learning_rate": 2.8316865071159078e-06,
-      "loss": 0.2154,
+      "epoch": 4.363354596893358,
+      "grad_norm": 0.23088960349559784,
+      "learning_rate": 2.105938299046678e-06,
+      "loss": 0.3678,
       "step": 121070
     },
     {
-      "epoch": 4.26,
-      "learning_rate": 2.83036973612551e-06,
-      "loss": 0.2405,
+      "epoch": 4.363534796554583,
+      "grad_norm": 0.2994691729545593,
+      "learning_rate": 2.1047661818405423e-06,
+      "loss": 0.4163,
       "step": 121075
     },
     {
-      "epoch": 4.26,
-      "learning_rate": 2.829053252993155e-06,
-      "loss": 0.2571,
+      "epoch": 4.3637149962158075,
+      "grad_norm": 0.22919879853725433,
+      "learning_rate": 2.103594376577542e-06,
+      "loss": 0.3619,
       "step": 121080
     },
     {
-      "epoch": 4.26,
-      "learning_rate": 2.8277370577359354e-06,
-      "loss": 0.2402,
+      "epoch": 4.363895195877031,
+      "grad_norm": 0.21188896894454956,
+      "learning_rate": 2.1024228832736332e-06,
+      "loss": 0.3756,
       "step": 121085
     },
     {
-      "epoch": 4.26,
-      "learning_rate": 2.8264211503709335e-06,
-      "loss": 0.2652,
+      "epoch": 4.364075395538256,
+      "grad_norm": 0.18373993039131165,
+      "learning_rate": 2.1012517019447786e-06,
+      "loss": 0.3687,
       "step": 121090
     },
     {
-      "epoch": 4.26,
-      "learning_rate": 2.8251055309152396e-06,
-      "loss": 0.2483,
+      "epoch": 4.364255595199481,
+      "grad_norm": 0.27971792221069336,
+      "learning_rate": 2.100080832606935e-06,
+      "loss": 0.4032,
       "step": 121095
     },
     {
-      "epoch": 4.26,
-      "learning_rate": 2.8237901993859396e-06,
-      "loss": 0.2549,
+      "epoch": 4.3644357948607055,
+      "grad_norm": 0.24079079926013947,
+      "learning_rate": 2.0989102752760587e-06,
+      "loss": 0.3679,
       "step": 121100
     },
     {
-      "epoch": 4.26,
-      "learning_rate": 2.8224751558001084e-06,
-      "loss": 0.2425,
+      "epoch": 4.36461599452193,
+      "grad_norm": 0.21578475832939148,
+      "learning_rate": 2.0977400299680927e-06,
+      "loss": 0.377,
       "step": 121105
     },
     {
-      "epoch": 4.26,
-      "learning_rate": 2.821160400174819e-06,
-      "loss": 0.2424,
+      "epoch": 4.364796194183155,
+      "grad_norm": 0.24278196692466736,
+      "learning_rate": 2.096570096698991e-06,
+      "loss": 0.3817,
       "step": 121110
     },
     {
-      "epoch": 4.26,
-      "learning_rate": 2.819845932527143e-06,
-      "loss": 0.2504,
+      "epoch": 4.36497639384438,
+      "grad_norm": 0.2502589523792267,
+      "learning_rate": 2.0954004754846795e-06,
+      "loss": 0.3749,
       "step": 121115
     },
     {
-      "epoch": 4.26,
-      "learning_rate": 2.818531752874157e-06,
-      "loss": 0.2361,
+      "epoch": 4.3651565935056045,
+      "grad_norm": 0.24439960718154907,
+      "learning_rate": 2.0942311663411095e-06,
+      "loss": 0.3807,
       "step": 121120
     },
     {
-      "epoch": 4.26,
-      "learning_rate": 2.817217861232918e-06,
-      "loss": 0.2291,
+      "epoch": 4.365336793166829,
+      "grad_norm": 0.2953266501426697,
+      "learning_rate": 2.093062169284202e-06,
+      "loss": 0.3785,
       "step": 121125
     },
     {
-      "epoch": 4.26,
-      "learning_rate": 2.815904257620483e-06,
-      "loss": 0.2484,
+      "epoch": 4.365516992828054,
+      "grad_norm": 0.21381710469722748,
+      "learning_rate": 2.091893484329882e-06,
+      "loss": 0.3554,
       "step": 121130
     },
     {
-      "epoch": 4.26,
-      "learning_rate": 2.8145909420539103e-06,
-      "loss": 0.2387,
+      "epoch": 4.365697192489278,
+      "grad_norm": 0.2584659457206726,
+      "learning_rate": 2.0907251114940907e-06,
+      "loss": 0.3765,
       "step": 121135
     },
     {
-      "epoch": 4.26,
-      "learning_rate": 2.813277914550258e-06,
-      "loss": 0.2293,
+      "epoch": 4.3658773921505025,
+      "grad_norm": 0.205842986702919,
+      "learning_rate": 2.089557050792726e-06,
+      "loss": 0.364,
       "step": 121140
     },
     {
-      "epoch": 4.26,
-      "learning_rate": 2.8119651751265687e-06,
-      "loss": 0.2186,
+      "epoch": 4.366057591811727,
+      "grad_norm": 0.22627507150173187,
+      "learning_rate": 2.088389302241708e-06,
+      "loss": 0.3628,
       "step": 121145
     },
     {
-      "epoch": 4.26,
-      "learning_rate": 2.8106527237998835e-06,
-      "loss": 0.2495,
+      "epoch": 4.366237791472952,
+      "grad_norm": 0.22469063103199005,
+      "learning_rate": 2.087221865856956e-06,
+      "loss": 0.328,
       "step": 121150
     },
     {
-      "epoch": 4.26,
-      "learning_rate": 2.8093405605872565e-06,
-      "loss": 0.2678,
+      "epoch": 4.366417991134177,
+      "grad_norm": 0.24385787546634674,
+      "learning_rate": 2.0860547416543697e-06,
+      "loss": 0.3852,
       "step": 121155
     },
     {
-      "epoch": 4.26,
-      "learning_rate": 2.8080286855057174e-06,
-      "loss": 0.2364,
+      "epoch": 4.3665981907954015,
+      "grad_norm": 0.2274542599916458,
+      "learning_rate": 2.084887929649848e-06,
+      "loss": 0.3781,
       "step": 121160
     },
     {
-      "epoch": 4.26,
-      "learning_rate": 2.806717098572295e-06,
-      "loss": 0.2825,
+      "epoch": 4.366778390456626,
+      "grad_norm": 0.31590333580970764,
+      "learning_rate": 2.0837214298592948e-06,
+      "loss": 0.3794,
       "step": 121165
     },
     {
-      "epoch": 4.26,
-      "learning_rate": 2.8054057998040255e-06,
-      "loss": 0.2593,
+      "epoch": 4.366958590117851,
+      "grad_norm": 0.24526174366474152,
+      "learning_rate": 2.082555242298595e-06,
+      "loss": 0.3573,
       "step": 121170
     },
     {
-      "epoch": 4.26,
-      "learning_rate": 2.8040947892179424e-06,
-      "loss": 0.2576,
+      "epoch": 4.367138789779075,
+      "grad_norm": 0.3067219853401184,
+      "learning_rate": 2.0813893669836495e-06,
+      "loss": 0.388,
       "step": 121175
     },
     {
-      "epoch": 4.26,
-      "learning_rate": 2.802784066831057e-06,
-      "loss": 0.2451,
+      "epoch": 4.3673189894402995,
+      "grad_norm": 0.21108821034431458,
+      "learning_rate": 2.080223803930334e-06,
+      "loss": 0.3336,
       "step": 121180
     },
     {
-      "epoch": 4.26,
-      "learning_rate": 2.801473632660395e-06,
-      "loss": 0.2642,
+      "epoch": 4.367499189101524,
+      "grad_norm": 0.312544584274292,
+      "learning_rate": 2.0790585531545336e-06,
+      "loss": 0.3726,
       "step": 121185
     },
     {
-      "epoch": 4.26,
-      "learning_rate": 2.800163486722962e-06,
-      "loss": 0.2485,
+      "epoch": 4.367679388762749,
+      "grad_norm": 0.2698698937892914,
+      "learning_rate": 2.077893614672122e-06,
+      "loss": 0.3531,
       "step": 121190
     },
     {
-      "epoch": 4.26,
-      "learning_rate": 2.79885362903578e-06,
-      "loss": 0.2436,
+      "epoch": 4.367859588423974,
+      "grad_norm": 0.2712307274341583,
+      "learning_rate": 2.076728988498966e-06,
+      "loss": 0.352,
       "step": 121195
     },
     {
-      "epoch": 4.26,
-      "learning_rate": 2.7975440596158533e-06,
-      "loss": 0.2518,
+      "epoch": 4.368039788085198,
+      "grad_norm": 0.2807067036628723,
+      "learning_rate": 2.0755646746509453e-06,
+      "loss": 0.3675,
       "step": 121200
     },
     {
-      "epoch": 4.26,
-      "learning_rate": 2.796234778480186e-06,
-      "loss": 0.2564,
+      "epoch": 4.368219987746423,
+      "grad_norm": 0.23370806872844696,
+      "learning_rate": 2.0744006731439225e-06,
+      "loss": 0.3685,
       "step": 121205
     },
     {
-      "epoch": 4.26,
-      "learning_rate": 2.7949257856457715e-06,
-      "loss": 0.281,
+      "epoch": 4.368400187407648,
+      "grad_norm": 0.30535852909088135,
+      "learning_rate": 2.0732369839937428e-06,
+      "loss": 0.3524,
       "step": 121210
     },
     {
-      "epoch": 4.26,
-      "learning_rate": 2.793617081129618e-06,
-      "loss": 0.2407,
+      "epoch": 4.368580387068873,
+      "grad_norm": 0.21354763209819794,
+      "learning_rate": 2.072073607216274e-06,
+      "loss": 0.3562,
       "step": 121215
     },
     {
-      "epoch": 4.26,
-      "learning_rate": 2.792308664948706e-06,
-      "loss": 0.2698,
+      "epoch": 4.368760586730097,
+      "grad_norm": 0.23636168241500854,
+      "learning_rate": 2.0709105428273647e-06,
+      "loss": 0.3307,
       "step": 121220
     },
     {
-      "epoch": 4.27,
-      "learning_rate": 2.7910005371200337e-06,
-      "loss": 0.2251,
+      "epoch": 4.368940786391321,
+      "grad_norm": 0.22024032473564148,
+      "learning_rate": 2.0697477908428602e-06,
+      "loss": 0.3996,
       "step": 121225
     },
     {
-      "epoch": 4.27,
-      "learning_rate": 2.7896926976605813e-06,
-      "loss": 0.2399,
+      "epoch": 4.369120986052546,
+      "grad_norm": 0.19595114886760712,
+      "learning_rate": 2.0685853512786013e-06,
+      "loss": 0.3299,
       "step": 121230
     },
     {
-      "epoch": 4.27,
-      "learning_rate": 2.7883851465873356e-06,
-      "loss": 0.2554,
+      "epoch": 4.369301185713771,
+      "grad_norm": 0.28025031089782715,
+      "learning_rate": 2.0674232241504223e-06,
+      "loss": 0.4299,
       "step": 121235
     },
     {
-      "epoch": 4.27,
-      "learning_rate": 2.7870778839172718e-06,
-      "loss": 0.2535,
+      "epoch": 4.369481385374995,
+      "grad_norm": 0.2547728419303894,
+      "learning_rate": 2.0662614094741683e-06,
+      "loss": 0.3679,
       "step": 121240
     },
     {
-      "epoch": 4.27,
-      "learning_rate": 2.7857709096673623e-06,
-      "loss": 0.261,
+      "epoch": 4.36966158503622,
+      "grad_norm": 0.22356577217578888,
+      "learning_rate": 2.0650999072656606e-06,
+      "loss": 0.3665,
       "step": 121245
     },
     {
-      "epoch": 4.27,
-      "learning_rate": 2.7844642238545717e-06,
-      "loss": 0.2599,
+      "epoch": 4.369841784697445,
+      "grad_norm": 0.2514247000217438,
+      "learning_rate": 2.063938717540728e-06,
+      "loss": 0.3732,
       "step": 121250
     },
     {
-      "epoch": 4.27,
-      "learning_rate": 2.78315782649588e-06,
-      "loss": 0.2503,
+      "epoch": 4.37002198435867,
+      "grad_norm": 0.2653522193431854,
+      "learning_rate": 2.0627778403151886e-06,
+      "loss": 0.364,
       "step": 121255
     },
     {
-      "epoch": 4.27,
-      "learning_rate": 2.7818517176082388e-06,
-      "loss": 0.2507,
+      "epoch": 4.370202184019894,
+      "grad_norm": 0.28731030225753784,
+      "learning_rate": 2.0616172756048573e-06,
+      "loss": 0.3676,
       "step": 121260
     },
     {
-      "epoch": 4.27,
-      "learning_rate": 2.7805458972086084e-06,
-      "loss": 0.2684,
+      "epoch": 4.370382383681119,
+      "grad_norm": 0.2413557469844818,
+      "learning_rate": 2.060457023425552e-06,
+      "loss": 0.3525,
       "step": 121265
     },
     {
-      "epoch": 4.27,
-      "learning_rate": 2.779240365313951e-06,
-      "loss": 0.246,
+      "epoch": 4.370562583342343,
+      "grad_norm": 0.27309590578079224,
+      "learning_rate": 2.0592970837930825e-06,
+      "loss": 0.3655,
       "step": 121270
     },
     {
-      "epoch": 4.27,
-      "learning_rate": 2.777935121941208e-06,
-      "loss": 0.2402,
+      "epoch": 4.370742783003568,
+      "grad_norm": 0.2248447686433792,
+      "learning_rate": 2.0581374567232443e-06,
+      "loss": 0.3994,
       "step": 121275
     },
     {
-      "epoch": 4.27,
-      "learning_rate": 2.7766301671073386e-06,
-      "loss": 0.2546,
+      "epoch": 4.370922982664792,
+      "grad_norm": 0.1932300329208374,
+      "learning_rate": 2.056978142231844e-06,
+      "loss": 0.3831,
       "step": 121280
     },
     {
-      "epoch": 4.27,
-      "learning_rate": 2.7753255008292763e-06,
-      "loss": 0.2616,
+      "epoch": 4.371103182326017,
+      "grad_norm": 0.22570841014385223,
+      "learning_rate": 2.055819140334675e-06,
+      "loss": 0.3929,
       "step": 121285
     },
     {
-      "epoch": 4.27,
-      "learning_rate": 2.774021123123968e-06,
-      "loss": 0.2429,
+      "epoch": 4.371283381987242,
+      "grad_norm": 0.2672460973262787,
+      "learning_rate": 2.0546604510475213e-06,
+      "loss": 0.3756,
       "step": 121290
     },
     {
-      "epoch": 4.27,
-      "learning_rate": 2.7727170340083487e-06,
-      "loss": 0.2257,
+      "epoch": 4.371463581648467,
+      "grad_norm": 0.2674674689769745,
+      "learning_rate": 2.0535020743861822e-06,
+      "loss": 0.3614,
       "step": 121295
     },
     {
-      "epoch": 4.27,
-      "learning_rate": 2.771413233499351e-06,
-      "loss": 0.2453,
+      "epoch": 4.371643781309691,
+      "grad_norm": 0.2942127585411072,
+      "learning_rate": 2.0523440103664358e-06,
+      "loss": 0.3904,
       "step": 121300
     },
     {
-      "epoch": 4.27,
-      "learning_rate": 2.7701097216139006e-06,
-      "loss": 0.2324,
+      "epoch": 4.371823980970916,
+      "grad_norm": 0.210496187210083,
+      "learning_rate": 2.0511862590040566e-06,
+      "loss": 0.3949,
       "step": 121305
     },
     {
-      "epoch": 4.27,
-      "learning_rate": 2.7688064983689253e-06,
-      "loss": 0.247,
+      "epoch": 4.372004180632141,
+      "grad_norm": 0.2888723909854889,
+      "learning_rate": 2.0500288203148254e-06,
+      "loss": 0.3525,
       "step": 121310
     },
     {
-      "epoch": 4.27,
-      "learning_rate": 2.767503563781351e-06,
-      "loss": 0.2538,
+      "epoch": 4.372184380293365,
+      "grad_norm": 0.2629977762699127,
+      "learning_rate": 2.0488716943145025e-06,
+      "loss": 0.3809,
       "step": 121315
     },
     {
-      "epoch": 4.27,
-      "learning_rate": 2.766200917868089e-06,
-      "loss": 0.2301,
+      "epoch": 4.372364579954589,
+      "grad_norm": 0.2915158271789551,
+      "learning_rate": 2.047714881018867e-06,
+      "loss": 0.3755,
       "step": 121320
     },
     {
-      "epoch": 4.27,
-      "learning_rate": 2.764898560646051e-06,
-      "loss": 0.2277,
+      "epoch": 4.372544779615814,
+      "grad_norm": 0.2206728309392929,
+      "learning_rate": 2.0465583804436668e-06,
+      "loss": 0.3814,
       "step": 121325
     },
     {
-      "epoch": 4.27,
-      "learning_rate": 2.763596492132159e-06,
-      "loss": 0.2596,
+      "epoch": 4.372724979277039,
+      "grad_norm": 0.27409932017326355,
+      "learning_rate": 2.0454021926046645e-06,
+      "loss": 0.3721,
       "step": 121330
     },
     {
-      "epoch": 4.27,
-      "learning_rate": 2.762294712343305e-06,
-      "loss": 0.2412,
+      "epoch": 4.372905178938264,
+      "grad_norm": 0.2713205814361572,
+      "learning_rate": 2.0442463175176147e-06,
+      "loss": 0.3677,
       "step": 121335
     },
     {
-      "epoch": 4.27,
-      "learning_rate": 2.760993221296404e-06,
-      "loss": 0.2399,
+      "epoch": 4.373085378599488,
+      "grad_norm": 0.2275678962469101,
+      "learning_rate": 2.0430907551982654e-06,
+      "loss": 0.365,
       "step": 121340
     },
     {
-      "epoch": 4.27,
-      "learning_rate": 2.7596920190083445e-06,
-      "loss": 0.2316,
+      "epoch": 4.373265578260713,
+      "grad_norm": 0.2382095456123352,
+      "learning_rate": 2.0419355056623597e-06,
+      "loss": 0.3718,
       "step": 121345
     },
     {
-      "epoch": 4.27,
-      "learning_rate": 2.758391105496033e-06,
-      "loss": 0.2384,
+      "epoch": 4.373445777921938,
+      "grad_norm": 0.207700714468956,
+      "learning_rate": 2.0407805689256374e-06,
+      "loss": 0.3636,
       "step": 121350
     },
     {
-      "epoch": 4.27,
-      "learning_rate": 2.757090480776353e-06,
-      "loss": 0.2528,
+      "epoch": 4.3736259775831625,
+      "grad_norm": 0.2512191832065582,
+      "learning_rate": 2.0396259450038315e-06,
+      "loss": 0.3407,
       "step": 121355
     },
     {
-      "epoch": 4.27,
-      "learning_rate": 2.7557901448661964e-06,
-      "loss": 0.2692,
+      "epoch": 4.373806177244386,
+      "grad_norm": 0.24360640347003937,
+      "learning_rate": 2.038471633912678e-06,
+      "loss": 0.3854,
       "step": 121360
     },
     {
-      "epoch": 4.27,
-      "learning_rate": 2.7544900977824416e-06,
-      "loss": 0.251,
+      "epoch": 4.373986376905611,
+      "grad_norm": 0.2918420433998108,
+      "learning_rate": 2.0373176356679067e-06,
+      "loss": 0.3846,
       "step": 121365
     },
     {
-      "epoch": 4.27,
-      "learning_rate": 2.753190339541978e-06,
-      "loss": 0.2458,
+      "epoch": 4.374166576566836,
+      "grad_norm": 0.21557246148586273,
+      "learning_rate": 2.0361639502852358e-06,
+      "loss": 0.3759,
       "step": 121370
     },
     {
-      "epoch": 4.27,
-      "learning_rate": 2.751890870161675e-06,
-      "loss": 0.2533,
+      "epoch": 4.374346776228061,
+      "grad_norm": 0.25849926471710205,
+      "learning_rate": 2.0350105777803856e-06,
+      "loss": 0.3576,
       "step": 121375
     },
     {
-      "epoch": 4.27,
-      "learning_rate": 2.7505916896583993e-06,
-      "loss": 0.2551,
+      "epoch": 4.374526975889285,
+      "grad_norm": 0.19965314865112305,
+      "learning_rate": 2.033857518169066e-06,
+      "loss": 0.3659,
       "step": 121380
     },
     {
-      "epoch": 4.27,
-      "learning_rate": 2.7492927980490356e-06,
-      "loss": 0.2596,
+      "epoch": 4.37470717555051,
+      "grad_norm": 0.2662714421749115,
+      "learning_rate": 2.032704771466995e-06,
+      "loss": 0.3814,
       "step": 121385
     },
     {
-      "epoch": 4.27,
-      "learning_rate": 2.7479941953504366e-06,
-      "loss": 0.2748,
+      "epoch": 4.374887375211735,
+      "grad_norm": 0.2764718532562256,
+      "learning_rate": 2.0315523376898765e-06,
+      "loss": 0.3941,
       "step": 121390
     },
     {
-      "epoch": 4.27,
-      "learning_rate": 2.7466958815794723e-06,
-      "loss": 0.2447,
+      "epoch": 4.3750675748729595,
+      "grad_norm": 0.2914084494113922,
+      "learning_rate": 2.0304002168534065e-06,
+      "loss": 0.3853,
       "step": 121395
     },
     {
-      "epoch": 4.27,
-      "learning_rate": 2.745397856752993e-06,
-      "loss": 0.2598,
+      "epoch": 4.375247774534184,
+      "grad_norm": 0.29716944694519043,
+      "learning_rate": 2.029248408973289e-06,
+      "loss": 0.3849,
       "step": 121400
     },
     {
-      "epoch": 4.27,
-      "learning_rate": 2.7441001208878575e-06,
-      "loss": 0.2256,
+      "epoch": 4.375427974195409,
+      "grad_norm": 0.21052998304367065,
+      "learning_rate": 2.028096914065214e-06,
+      "loss": 0.3677,
       "step": 121405
     },
     {
-      "epoch": 4.27,
-      "learning_rate": 2.742802674000919e-06,
-      "loss": 0.2501,
+      "epoch": 4.375608173856633,
+      "grad_norm": 0.20469604432582855,
+      "learning_rate": 2.026945732144872e-06,
+      "loss": 0.3207,
       "step": 121410
     },
     {
-      "epoch": 4.27,
-      "learning_rate": 2.7415055161090166e-06,
-      "loss": 0.2437,
+      "epoch": 4.3757883735178575,
+      "grad_norm": 0.24874727427959442,
+      "learning_rate": 2.025794863227945e-06,
+      "loss": 0.3688,
       "step": 121415
     },
     {
-      "epoch": 4.27,
-      "learning_rate": 2.740208647228992e-06,
-      "loss": 0.247,
+      "epoch": 4.375968573179082,
+      "grad_norm": 0.2315155416727066,
+      "learning_rate": 2.024644307330112e-06,
+      "loss": 0.3849,
       "step": 121420
     },
     {
-      "epoch": 4.27,
-      "learning_rate": 2.738912067377694e-06,
-      "loss": 0.2348,
+      "epoch": 4.376148772840307,
+      "grad_norm": 0.27341365814208984,
+      "learning_rate": 2.0234940644670576e-06,
+      "loss": 0.3887,
       "step": 121425
     },
     {
-      "epoch": 4.27,
-      "learning_rate": 2.737615776571953e-06,
-      "loss": 0.2418,
+      "epoch": 4.376328972501532,
+      "grad_norm": 0.2433520257472992,
+      "learning_rate": 2.0223441346544468e-06,
+      "loss": 0.3943,
       "step": 121430
     },
     {
-      "epoch": 4.27,
-      "learning_rate": 2.7363197748285942e-06,
-      "loss": 0.2489,
+      "epoch": 4.3765091721627565,
+      "grad_norm": 0.24576754868030548,
+      "learning_rate": 2.0211945179079483e-06,
+      "loss": 0.371,
       "step": 121435
     },
     {
-      "epoch": 4.27,
-      "learning_rate": 2.735024062164454e-06,
-      "loss": 0.2592,
+      "epoch": 4.376689371823981,
+      "grad_norm": 0.19979894161224365,
+      "learning_rate": 2.020045214243227e-06,
+      "loss": 0.3891,
       "step": 121440
     },
     {
-      "epoch": 4.27,
-      "learning_rate": 2.7337286385963503e-06,
-      "loss": 0.2512,
+      "epoch": 4.376869571485206,
+      "grad_norm": 0.27705785632133484,
+      "learning_rate": 2.0188962236759342e-06,
+      "loss": 0.3707,
       "step": 121445
     },
     {
-      "epoch": 4.27,
-      "learning_rate": 2.7324335041411105e-06,
-      "loss": 0.2565,
+      "epoch": 4.37704977114643,
+      "grad_norm": 0.26884958148002625,
+      "learning_rate": 2.0177475462217376e-06,
+      "loss": 0.3791,
       "step": 121450
     },
     {
-      "epoch": 4.27,
-      "learning_rate": 2.7311386588155464e-06,
-      "loss": 0.2642,
+      "epoch": 4.3772299708076545,
+      "grad_norm": 0.26522353291511536,
+      "learning_rate": 2.0165991818962784e-06,
+      "loss": 0.3957,
       "step": 121455
     },
     {
-      "epoch": 4.27,
-      "learning_rate": 2.7298441026364667e-06,
-      "loss": 0.2381,
+      "epoch": 4.377410170468879,
+      "grad_norm": 0.2895340323448181,
+      "learning_rate": 2.0154511307152066e-06,
+      "loss": 0.3571,
       "step": 121460
     },
     {
-      "epoch": 4.27,
-      "learning_rate": 2.728549835620689e-06,
-      "loss": 0.2478,
+      "epoch": 4.377590370130104,
+      "grad_norm": 0.23953476548194885,
+      "learning_rate": 2.014303392694164e-06,
+      "loss": 0.361,
       "step": 121465
     },
     {
-      "epoch": 4.27,
-      "learning_rate": 2.727255857785016e-06,
-      "loss": 0.255,
+      "epoch": 4.377770569791329,
+      "grad_norm": 0.26138606667518616,
+      "learning_rate": 2.01315596784879e-06,
+      "loss": 0.3653,
       "step": 121470
     },
     {
-      "epoch": 4.27,
-      "learning_rate": 2.7259621691462454e-06,
-      "loss": 0.2477,
+      "epoch": 4.3779507694525535,
+      "grad_norm": 0.2306900918483734,
+      "learning_rate": 2.0120088561947082e-06,
+      "loss": 0.3363,
       "step": 121475
     },
     {
-      "epoch": 4.27,
-      "learning_rate": 2.7246687697211728e-06,
-      "loss": 0.2686,
+      "epoch": 4.378130969113778,
+      "grad_norm": 0.24305300414562225,
+      "learning_rate": 2.010862057747559e-06,
+      "loss": 0.394,
       "step": 121480
     },
     {
-      "epoch": 4.27,
-      "learning_rate": 2.7233756595266003e-06,
-      "loss": 0.2574,
+      "epoch": 4.378311168775003,
+      "grad_norm": 0.2668958604335785,
+      "learning_rate": 2.0097155725229634e-06,
+      "loss": 0.3817,
       "step": 121485
     },
     {
-      "epoch": 4.27,
-      "learning_rate": 2.722082838579315e-06,
-      "loss": 0.2669,
+      "epoch": 4.378491368436228,
+      "grad_norm": 0.23452064394950867,
+      "learning_rate": 2.0085694005365446e-06,
+      "loss": 0.361,
       "step": 121490
     },
     {
-      "epoch": 4.27,
-      "learning_rate": 2.7207903068960982e-06,
-      "loss": 0.2532,
+      "epoch": 4.378671568097452,
+      "grad_norm": 0.2706201672554016,
+      "learning_rate": 2.007423541803913e-06,
+      "loss": 0.3506,
       "step": 121495
     },
     {
-      "epoch": 4.27,
-      "learning_rate": 2.7194980644937335e-06,
-      "loss": 0.2605,
+      "epoch": 4.378851767758676,
+      "grad_norm": 0.2533268332481384,
+      "learning_rate": 2.006277996340683e-06,
+      "loss": 0.3454,
       "step": 121500
     },
     {
-      "epoch": 4.27,
-      "eval_loss": 0.2489774525165558,
-      "eval_runtime": 10.5356,
-      "eval_samples_per_second": 9.492,
-      "eval_steps_per_second": 9.492,
+      "epoch": 4.378851767758676,
+      "eval_loss": 0.42894676327705383,
+      "eval_runtime": 3.5453,
+      "eval_samples_per_second": 28.206,
+      "eval_steps_per_second": 7.052,
       "step": 121500
     },
     {
-      "epoch": 4.27,
-      "learning_rate": 2.71820611138901e-06,
-      "loss": 0.2496,
+      "epoch": 4.379031967419901,
+      "grad_norm": 0.23799777030944824,
+      "learning_rate": 2.0051327641624706e-06,
+      "loss": 0.3548,
       "step": 121505
     },
     {
-      "epoch": 4.28,
-      "learning_rate": 2.716914447598698e-06,
-      "loss": 0.2578,
+      "epoch": 4.379212167081126,
+      "grad_norm": 0.302617609500885,
+      "learning_rate": 2.0039878452848687e-06,
+      "loss": 0.3609,
       "step": 121510
     },
     {
-      "epoch": 4.28,
-      "learning_rate": 2.715623073139559e-06,
-      "loss": 0.2563,
+      "epoch": 4.37939236674235,
+      "grad_norm": 0.21825525164604187,
+      "learning_rate": 2.0028432397234758e-06,
+      "loss": 0.3464,
       "step": 121515
     },
     {
-      "epoch": 4.28,
-      "learning_rate": 2.7143319880283735e-06,
-      "loss": 0.2564,
+      "epoch": 4.379572566403575,
+      "grad_norm": 0.26662659645080566,
+      "learning_rate": 2.0016989474938934e-06,
+      "loss": 0.3305,
       "step": 121520
     },
     {
-      "epoch": 4.28,
-      "learning_rate": 2.713041192281901e-06,
-      "loss": 0.2646,
+      "epoch": 4.3797527660648,
+      "grad_norm": 0.31008052825927734,
+      "learning_rate": 2.000554968611712e-06,
+      "loss": 0.3879,
       "step": 121525
     },
     {
-      "epoch": 4.28,
-      "learning_rate": 2.7117506859169028e-06,
-      "loss": 0.254,
+      "epoch": 4.379932965726025,
+      "grad_norm": 0.2644076943397522,
+      "learning_rate": 1.9994113030925188e-06,
+      "loss": 0.3359,
       "step": 121530
     },
     {
-      "epoch": 4.28,
-      "learning_rate": 2.710460468950127e-06,
-      "loss": 0.2599,
+      "epoch": 4.380113165387249,
+      "grad_norm": 0.25222399830818176,
+      "learning_rate": 1.9982679509518905e-06,
+      "loss": 0.3419,
       "step": 121535
     },
     {
-      "epoch": 4.28,
-      "learning_rate": 2.709170541398337e-06,
-      "loss": 0.2422,
+      "epoch": 4.380293365048474,
+      "grad_norm": 0.2816098928451538,
+      "learning_rate": 1.997124912205403e-06,
+      "loss": 0.3983,
       "step": 121540
     },
     {
-      "epoch": 4.28,
-      "learning_rate": 2.707880903278276e-06,
-      "loss": 0.2631,
+      "epoch": 4.380473564709698,
+      "grad_norm": 0.5892481803894043,
+      "learning_rate": 1.9959821868686416e-06,
+      "loss": 0.3839,
       "step": 121545
     },
     {
-      "epoch": 4.28,
-      "learning_rate": 2.7065915546066883e-06,
-      "loss": 0.2415,
+      "epoch": 4.380653764370923,
+      "grad_norm": 0.2820974886417389,
+      "learning_rate": 1.994839774957169e-06,
+      "loss": 0.3707,
       "step": 121550
     },
     {
-      "epoch": 4.28,
-      "learning_rate": 2.7053024954003163e-06,
-      "loss": 0.2481,
+      "epoch": 4.380833964032147,
+      "grad_norm": 0.24034863710403442,
+      "learning_rate": 1.9936976764865502e-06,
+      "loss": 0.3653,
       "step": 121555
     },
     {
-      "epoch": 4.28,
-      "learning_rate": 2.7040137256759022e-06,
-      "loss": 0.2766,
+      "epoch": 4.381014163693372,
+      "grad_norm": 0.2928817570209503,
+      "learning_rate": 1.9925558914723475e-06,
+      "loss": 0.3839,
       "step": 121560
     },
     {
-      "epoch": 4.28,
-      "learning_rate": 2.7027252454501772e-06,
-      "loss": 0.2354,
+      "epoch": 4.381194363354597,
+      "grad_norm": 0.2788287401199341,
+      "learning_rate": 1.99141441993011e-06,
+      "loss": 0.3397,
       "step": 121565
     },
     {
-      "epoch": 4.28,
-      "learning_rate": 2.701437054739872e-06,
-      "loss": 0.2468,
+      "epoch": 4.381374563015822,
+      "grad_norm": 0.2363235503435135,
+      "learning_rate": 1.9902732618753998e-06,
+      "loss": 0.3395,
       "step": 121570
     },
     {
-      "epoch": 4.28,
-      "learning_rate": 2.700149153561704e-06,
-      "loss": 0.2653,
+      "epoch": 4.381554762677046,
+      "grad_norm": 0.2198476344347,
+      "learning_rate": 1.989132417323769e-06,
+      "loss": 0.3649,
       "step": 121575
     },
     {
-      "epoch": 4.28,
-      "learning_rate": 2.69886154193241e-06,
-      "loss": 0.2803,
+      "epoch": 4.381734962338271,
+      "grad_norm": 0.2580127716064453,
+      "learning_rate": 1.987991886290744e-06,
+      "loss": 0.3786,
       "step": 121580
     },
     {
-      "epoch": 4.28,
-      "learning_rate": 2.697574219868701e-06,
-      "loss": 0.256,
+      "epoch": 4.381915161999496,
+      "grad_norm": 0.225581094622612,
+      "learning_rate": 1.986851668791878e-06,
+      "loss": 0.3706,
       "step": 121585
     },
     {
-      "epoch": 4.28,
-      "learning_rate": 2.696287187387292e-06,
-      "loss": 0.2505,
+      "epoch": 4.38209536166072,
+      "grad_norm": 0.22968202829360962,
+      "learning_rate": 1.9857117648427044e-06,
+      "loss": 0.3682,
       "step": 121590
     },
     {
-      "epoch": 4.28,
-      "learning_rate": 2.6950004445048917e-06,
-      "loss": 0.2512,
+      "epoch": 4.382275561321944,
+      "grad_norm": 0.26144370436668396,
+      "learning_rate": 1.984572174458746e-06,
+      "loss": 0.3679,
       "step": 121595
     },
     {
-      "epoch": 4.28,
-      "learning_rate": 2.6937139912382142e-06,
-      "loss": 0.255,
+      "epoch": 4.382455760983169,
+      "grad_norm": 0.2481224089860916,
+      "learning_rate": 1.983432897655546e-06,
+      "loss": 0.3604,
       "step": 121600
     },
     {
-      "epoch": 4.28,
-      "learning_rate": 2.692427827603958e-06,
-      "loss": 0.2581,
+      "epoch": 4.382635960644394,
+      "grad_norm": 0.24234378337860107,
+      "learning_rate": 1.9822939344486063e-06,
+      "loss": 0.389,
       "step": 121605
     },
     {
-      "epoch": 4.28,
-      "learning_rate": 2.6911419536188228e-06,
-      "loss": 0.2419,
+      "epoch": 4.382816160305619,
+      "grad_norm": 0.27252396941185,
+      "learning_rate": 1.9811552848534587e-06,
+      "loss": 0.3839,
       "step": 121610
     },
     {
-      "epoch": 4.28,
-      "learning_rate": 2.689856369299512e-06,
-      "loss": 0.2618,
+      "epoch": 4.382996359966843,
+      "grad_norm": 0.30419376492500305,
+      "learning_rate": 1.980016948885616e-06,
+      "loss": 0.3901,
       "step": 121615
     },
     {
-      "epoch": 4.28,
-      "learning_rate": 2.688571074662716e-06,
-      "loss": 0.2373,
+      "epoch": 4.383176559628068,
+      "grad_norm": 0.3380153477191925,
+      "learning_rate": 1.9788789265605844e-06,
+      "loss": 0.3617,
       "step": 121620
     },
     {
-      "epoch": 4.28,
-      "learning_rate": 2.68728606972512e-06,
-      "loss": 0.2525,
+      "epoch": 4.383356759289293,
+      "grad_norm": 0.23734219372272491,
+      "learning_rate": 1.9777412178938722e-06,
+      "loss": 0.3463,
       "step": 121625
     },
     {
-      "epoch": 4.28,
-      "learning_rate": 2.6860013545034034e-06,
-      "loss": 0.2449,
+      "epoch": 4.3835369589505175,
+      "grad_norm": 0.2698742747306824,
+      "learning_rate": 1.976603822900974e-06,
+      "loss": 0.3741,
       "step": 121630
     },
     {
-      "epoch": 4.28,
-      "learning_rate": 2.6847169290142583e-06,
-      "loss": 0.2477,
+      "epoch": 4.383717158611741,
+      "grad_norm": 0.2732264995574951,
+      "learning_rate": 1.9754667415973953e-06,
+      "loss": 0.3122,
       "step": 121635
     },
     {
-      "epoch": 4.28,
-      "learning_rate": 2.6834327932743598e-06,
-      "loss": 0.2501,
+      "epoch": 4.383897358272966,
+      "grad_norm": 0.2620375156402588,
+      "learning_rate": 1.974329973998623e-06,
+      "loss": 0.3475,
       "step": 121640
     },
     {
-      "epoch": 4.28,
-      "learning_rate": 2.682148947300378e-06,
-      "loss": 0.2325,
+      "epoch": 4.384077557934191,
+      "grad_norm": 0.26369425654411316,
+      "learning_rate": 1.973193520120148e-06,
+      "loss": 0.3816,
       "step": 121645
     },
     {
-      "epoch": 4.28,
-      "learning_rate": 2.6808653911089775e-06,
-      "loss": 0.2497,
+      "epoch": 4.384257757595416,
+      "grad_norm": 0.26098334789276123,
+      "learning_rate": 1.9720573799774542e-06,
+      "loss": 0.3592,
       "step": 121650
     },
     {
-      "epoch": 4.28,
-      "learning_rate": 2.6795821247168394e-06,
-      "loss": 0.2622,
+      "epoch": 4.38443795725664,
+      "grad_norm": 0.20638631284236908,
+      "learning_rate": 1.970921553586019e-06,
+      "loss": 0.3724,
       "step": 121655
     },
     {
-      "epoch": 4.28,
-      "learning_rate": 2.678299148140609e-06,
-      "loss": 0.2621,
+      "epoch": 4.384618156917865,
+      "grad_norm": 0.24286890029907227,
+      "learning_rate": 1.9697860409613133e-06,
+      "loss": 0.3677,
       "step": 121660
     },
     {
-      "epoch": 4.28,
-      "learning_rate": 2.677016461396961e-06,
-      "loss": 0.2589,
+      "epoch": 4.38479835657909,
+      "grad_norm": 0.2833465039730072,
+      "learning_rate": 1.9686508421188214e-06,
+      "loss": 0.4158,
       "step": 121665
     },
     {
-      "epoch": 4.28,
-      "learning_rate": 2.6757340645025387e-06,
-      "loss": 0.2481,
+      "epoch": 4.3849785562403145,
+      "grad_norm": 0.2680220901966095,
+      "learning_rate": 1.967515957074001e-06,
+      "loss": 0.3728,
       "step": 121670
     },
     {
-      "epoch": 4.28,
-      "learning_rate": 2.674451957474e-06,
-      "loss": 0.2638,
+      "epoch": 4.385158755901539,
+      "grad_norm": 0.26703187823295593,
+      "learning_rate": 1.9663813858423167e-06,
+      "loss": 0.3861,
       "step": 121675
     },
     {
-      "epoch": 4.28,
-      "learning_rate": 2.673170140327991e-06,
-      "loss": 0.268,
+      "epoch": 4.385338955562764,
+      "grad_norm": 0.2260323315858841,
+      "learning_rate": 1.9652471284392265e-06,
+      "loss": 0.3477,
       "step": 121680
     },
     {
-      "epoch": 4.28,
-      "learning_rate": 2.671888613081153e-06,
-      "loss": 0.2838,
+      "epoch": 4.385519155223988,
+      "grad_norm": 0.23601694405078888,
+      "learning_rate": 1.9641131848801785e-06,
+      "loss": 0.4099,
       "step": 121685
     },
     {
-      "epoch": 4.28,
-      "learning_rate": 2.6706073757501227e-06,
-      "loss": 0.2576,
+      "epoch": 4.385699354885213,
+      "grad_norm": 0.26613062620162964,
+      "learning_rate": 1.9629795551806414e-06,
+      "loss": 0.3624,
       "step": 121690
     },
     {
-      "epoch": 4.28,
-      "learning_rate": 2.669326428351543e-06,
-      "loss": 0.2445,
+      "epoch": 4.385879554546437,
+      "grad_norm": 0.23017358779907227,
+      "learning_rate": 1.961846239356038e-06,
+      "loss": 0.3831,
       "step": 121695
     },
     {
-      "epoch": 4.28,
-      "learning_rate": 2.6680457709020447e-06,
-      "loss": 0.2423,
+      "epoch": 4.386059754207662,
+      "grad_norm": 0.3070394992828369,
+      "learning_rate": 1.9607132374218184e-06,
+      "loss": 0.349,
       "step": 121700
     },
     {
-      "epoch": 4.28,
-      "learning_rate": 2.666765403418256e-06,
-      "loss": 0.2709,
+      "epoch": 4.386239953868887,
+      "grad_norm": 0.22875505685806274,
+      "learning_rate": 1.959580549393425e-06,
+      "loss": 0.3717,
       "step": 121705
     },
     {
-      "epoch": 4.28,
-      "learning_rate": 2.6654853259167943e-06,
-      "loss": 0.2508,
+      "epoch": 4.3864201535301115,
+      "grad_norm": 0.3101567327976227,
+      "learning_rate": 1.9584481752862816e-06,
+      "loss": 0.3494,
       "step": 121710
     },
     {
-      "epoch": 4.28,
-      "learning_rate": 2.6642055384142857e-06,
-      "loss": 0.2431,
+      "epoch": 4.386600353191336,
+      "grad_norm": 0.2821701467037201,
+      "learning_rate": 1.9573161151158288e-06,
+      "loss": 0.3666,
       "step": 121715
     },
     {
-      "epoch": 4.28,
-      "learning_rate": 2.662926040927352e-06,
-      "loss": 0.2804,
+      "epoch": 4.386780552852561,
+      "grad_norm": 0.2166757434606552,
+      "learning_rate": 1.9561843688974764e-06,
+      "loss": 0.4189,
       "step": 121720
     },
     {
-      "epoch": 4.28,
-      "learning_rate": 2.661646833472606e-06,
-      "loss": 0.2608,
+      "epoch": 4.386960752513785,
+      "grad_norm": 0.24579738080501556,
+      "learning_rate": 1.955052936646648e-06,
+      "loss": 0.3535,
       "step": 121725
     },
     {
-      "epoch": 4.28,
-      "learning_rate": 2.660367916066647e-06,
-      "loss": 0.2494,
+      "epoch": 4.3871409521750095,
+      "grad_norm": 0.2603980600833893,
+      "learning_rate": 1.953921818378768e-06,
+      "loss": 0.3511,
       "step": 121730
     },
     {
-      "epoch": 4.28,
-      "learning_rate": 2.6590892887260963e-06,
-      "loss": 0.232,
+      "epoch": 4.387321151836234,
+      "grad_norm": 0.23183012008666992,
+      "learning_rate": 1.952791014109237e-06,
+      "loss": 0.3724,
       "step": 121735
     },
     {
-      "epoch": 4.28,
-      "learning_rate": 2.657810951467546e-06,
-      "loss": 0.2577,
+      "epoch": 4.387501351497459,
+      "grad_norm": 0.1970612108707428,
+      "learning_rate": 1.9516605238534686e-06,
+      "loss": 0.3725,
       "step": 121740
     },
     {
-      "epoch": 4.28,
-      "learning_rate": 2.6565329043075905e-06,
-      "loss": 0.2576,
+      "epoch": 4.387681551158684,
+      "grad_norm": 0.283931702375412,
+      "learning_rate": 1.950530347626864e-06,
+      "loss": 0.371,
       "step": 121745
     },
     {
-      "epoch": 4.28,
-      "learning_rate": 2.6552551472628363e-06,
-      "loss": 0.253,
+      "epoch": 4.3878617508199085,
+      "grad_norm": 0.32697275280952454,
+      "learning_rate": 1.9494004854448163e-06,
+      "loss": 0.3709,
       "step": 121750
     },
     {
-      "epoch": 4.28,
-      "learning_rate": 2.65397768034987e-06,
-      "loss": 0.2486,
+      "epoch": 4.388041950481133,
+      "grad_norm": 0.2545589506626129,
+      "learning_rate": 1.948270937322727e-06,
+      "loss": 0.3777,
       "step": 121755
     },
     {
-      "epoch": 4.28,
-      "learning_rate": 2.652700503585276e-06,
-      "loss": 0.2248,
+      "epoch": 4.388222150142358,
+      "grad_norm": 0.2229977399110794,
+      "learning_rate": 1.947141703275987e-06,
+      "loss": 0.3423,
       "step": 121760
     },
     {
-      "epoch": 4.28,
-      "learning_rate": 2.651423616985632e-06,
-      "loss": 0.2486,
+      "epoch": 4.388402349803583,
+      "grad_norm": 0.2144192010164261,
+      "learning_rate": 1.9460127833199756e-06,
+      "loss": 0.3872,
       "step": 121765
     },
     {
-      "epoch": 4.28,
-      "learning_rate": 2.6501470205675283e-06,
-      "loss": 0.2499,
+      "epoch": 4.388582549464807,
+      "grad_norm": 0.25160834193229675,
+      "learning_rate": 1.9448841774700744e-06,
+      "loss": 0.4037,
       "step": 121770
     },
     {
-      "epoch": 4.28,
-      "learning_rate": 2.6488707143475343e-06,
-      "loss": 0.2508,
+      "epoch": 4.388762749126031,
+      "grad_norm": 0.2952007055282593,
+      "learning_rate": 1.943755885741666e-06,
+      "loss": 0.3825,
       "step": 121775
     },
     {
-      "epoch": 4.28,
-      "learning_rate": 2.647594698342229e-06,
-      "loss": 0.2659,
+      "epoch": 4.388942948787256,
+      "grad_norm": 0.25669974088668823,
+      "learning_rate": 1.942627908150113e-06,
+      "loss": 0.3635,
       "step": 121780
     },
     {
-      "epoch": 4.28,
-      "learning_rate": 2.6463189725681683e-06,
-      "loss": 0.2229,
+      "epoch": 4.389123148448481,
+      "grad_norm": 0.28366559743881226,
+      "learning_rate": 1.9415002447108004e-06,
+      "loss": 0.3789,
       "step": 121785
     },
     {
-      "epoch": 4.28,
-      "learning_rate": 2.64504353704193e-06,
-      "loss": 0.2552,
+      "epoch": 4.3893033481097055,
+      "grad_norm": 0.23855380713939667,
+      "learning_rate": 1.9403728954390745e-06,
+      "loss": 0.4061,
       "step": 121790
     },
     {
-      "epoch": 4.29,
-      "learning_rate": 2.643768391780069e-06,
-      "loss": 0.2331,
+      "epoch": 4.38948354777093,
+      "grad_norm": 0.2798866033554077,
+      "learning_rate": 1.939245860350308e-06,
+      "loss": 0.3814,
       "step": 121795
     },
     {
-      "epoch": 4.29,
-      "learning_rate": 2.6424935367991425e-06,
-      "loss": 0.2623,
+      "epoch": 4.389663747432155,
+      "grad_norm": 0.2672715485095978,
+      "learning_rate": 1.9381191394598514e-06,
+      "loss": 0.3589,
       "step": 121800
     },
     {
-      "epoch": 4.29,
-      "learning_rate": 2.6412189721157e-06,
-      "loss": 0.2548,
+      "epoch": 4.38984394709338,
+      "grad_norm": 0.22907915711402893,
+      "learning_rate": 1.936992732783058e-06,
+      "loss": 0.3436,
       "step": 121805
     },
     {
-      "epoch": 4.29,
-      "learning_rate": 2.6399446977462968e-06,
-      "loss": 0.2427,
+      "epoch": 4.390024146754604,
+      "grad_norm": 0.3034953474998474,
+      "learning_rate": 1.935866640335271e-06,
+      "loss": 0.3956,
       "step": 121810
     },
     {
-      "epoch": 4.29,
-      "learning_rate": 2.638670713707475e-06,
-      "loss": 0.2621,
+      "epoch": 4.390204346415829,
+      "grad_norm": 0.21499215066432953,
+      "learning_rate": 1.934740862131837e-06,
+      "loss": 0.3986,
       "step": 121815
     },
     {
-      "epoch": 4.29,
-      "learning_rate": 2.637397020015775e-06,
-      "loss": 0.2476,
+      "epoch": 4.390384546077053,
+      "grad_norm": 0.2739037275314331,
+      "learning_rate": 1.9336153981880906e-06,
+      "loss": 0.3885,
       "step": 121820
     },
     {
-      "epoch": 4.29,
-      "learning_rate": 2.6361236166877414e-06,
-      "loss": 0.2595,
+      "epoch": 4.390564745738278,
+      "grad_norm": 0.21741308271884918,
+      "learning_rate": 1.9324902485193696e-06,
+      "loss": 0.3895,
       "step": 121825
     },
     {
-      "epoch": 4.29,
-      "learning_rate": 2.6348505037398995e-06,
-      "loss": 0.2507,
+      "epoch": 4.390744945399502,
+      "grad_norm": 0.24937617778778076,
+      "learning_rate": 1.9313654131410035e-06,
+      "loss": 0.3741,
       "step": 121830
     },
     {
-      "epoch": 4.29,
-      "learning_rate": 2.633577681188787e-06,
-      "loss": 0.2412,
+      "epoch": 4.390925145060727,
+      "grad_norm": 0.2421903759241104,
+      "learning_rate": 1.930240892068319e-06,
+      "loss": 0.3458,
       "step": 121835
     },
     {
-      "epoch": 4.29,
-      "learning_rate": 2.6323051490509314e-06,
-      "loss": 0.2364,
+      "epoch": 4.391105344721952,
+      "grad_norm": 0.3107787072658539,
+      "learning_rate": 1.929116685316634e-06,
+      "loss": 0.3806,
       "step": 121840
     },
     {
-      "epoch": 4.29,
-      "learning_rate": 2.6310329073428454e-06,
-      "loss": 0.2592,
+      "epoch": 4.391285544383177,
+      "grad_norm": 0.23685042560100555,
+      "learning_rate": 1.9279927929012648e-06,
+      "loss": 0.3384,
       "step": 121845
     },
     {
-      "epoch": 4.29,
-      "learning_rate": 2.629760956081062e-06,
-      "loss": 0.2489,
+      "epoch": 4.391465744044401,
+      "grad_norm": 0.27006104588508606,
+      "learning_rate": 1.926869214837532e-06,
+      "loss": 0.3696,
       "step": 121850
     },
     {
-      "epoch": 4.29,
-      "learning_rate": 2.628489295282091e-06,
-      "loss": 0.2743,
+      "epoch": 4.391645943705626,
+      "grad_norm": 0.23979541659355164,
+      "learning_rate": 1.9257459511407366e-06,
+      "loss": 0.3836,
       "step": 121855
     },
     {
-      "epoch": 4.29,
-      "learning_rate": 2.6272179249624355e-06,
-      "loss": 0.2614,
+      "epoch": 4.391826143366851,
+      "grad_norm": 0.2638680040836334,
+      "learning_rate": 1.9246230018261867e-06,
+      "loss": 0.3689,
       "step": 121860
     },
     {
-      "epoch": 4.29,
-      "learning_rate": 2.625946845138616e-06,
-      "loss": 0.2532,
+      "epoch": 4.392006343028076,
+      "grad_norm": 0.23276658356189728,
+      "learning_rate": 1.923500366909181e-06,
+      "loss": 0.3862,
       "step": 121865
     },
     {
-      "epoch": 4.29,
-      "learning_rate": 2.624676055827133e-06,
-      "loss": 0.2724,
+      "epoch": 4.392186542689299,
+      "grad_norm": 0.3164710998535156,
+      "learning_rate": 1.9223780464050103e-06,
+      "loss": 0.3862,
       "step": 121870
     },
     {
-      "epoch": 4.29,
-      "learning_rate": 2.623405557044481e-06,
-      "loss": 0.2745,
+      "epoch": 4.392366742350524,
+      "grad_norm": 0.2991268038749695,
+      "learning_rate": 1.9212560403289752e-06,
+      "loss": 0.4063,
       "step": 121875
     },
     {
-      "epoch": 4.29,
-      "learning_rate": 2.622135348807164e-06,
-      "loss": 0.2399,
+      "epoch": 4.392546942011749,
+      "grad_norm": 0.2558140456676483,
+      "learning_rate": 1.9201343486963646e-06,
+      "loss": 0.3915,
       "step": 121880
     },
     {
-      "epoch": 4.29,
-      "learning_rate": 2.6208654311316694e-06,
-      "loss": 0.2491,
+      "epoch": 4.392727141672974,
+      "grad_norm": 0.2522118389606476,
+      "learning_rate": 1.9190129715224435e-06,
+      "loss": 0.3721,
       "step": 121885
     },
     {
-      "epoch": 4.29,
-      "learning_rate": 2.619595804034494e-06,
-      "loss": 0.2312,
+      "epoch": 4.392907341334198,
+      "grad_norm": 0.22641310095787048,
+      "learning_rate": 1.9178919088225086e-06,
+      "loss": 0.3486,
       "step": 121890
     },
     {
-      "epoch": 4.29,
-      "learning_rate": 2.6183264675321145e-06,
-      "loss": 0.2636,
+      "epoch": 4.393087540995423,
+      "grad_norm": 0.19269070029258728,
+      "learning_rate": 1.9167711606118216e-06,
+      "loss": 0.3699,
       "step": 121895
     },
     {
-      "epoch": 4.29,
-      "learning_rate": 2.617057421641014e-06,
-      "loss": 0.2386,
+      "epoch": 4.393267740656648,
+      "grad_norm": 0.24431343376636505,
+      "learning_rate": 1.9156507269056683e-06,
+      "loss": 0.3885,
       "step": 121900
     },
     {
-      "epoch": 4.29,
-      "learning_rate": 2.615788666377675e-06,
-      "loss": 0.2337,
+      "epoch": 4.393447940317873,
+      "grad_norm": 0.25327545404434204,
+      "learning_rate": 1.9145306077192976e-06,
+      "loss": 0.382,
       "step": 121905
     },
     {
-      "epoch": 4.29,
-      "learning_rate": 2.614520201758569e-06,
-      "loss": 0.2488,
+      "epoch": 4.393628139979096,
+      "grad_norm": 0.24400508403778076,
+      "learning_rate": 1.9134108030679774e-06,
+      "loss": 0.3645,
       "step": 121910
     },
     {
-      "epoch": 4.29,
-      "learning_rate": 2.6132520278001672e-06,
-      "loss": 0.2382,
+      "epoch": 4.393808339640321,
+      "grad_norm": 0.23288530111312866,
+      "learning_rate": 1.9122913129669657e-06,
+      "loss": 0.3578,
       "step": 121915
     },
     {
-      "epoch": 4.29,
-      "learning_rate": 2.611984144518928e-06,
-      "loss": 0.2425,
+      "epoch": 4.393988539301546,
+      "grad_norm": 0.2210901528596878,
+      "learning_rate": 1.911172137431519e-06,
+      "loss": 0.3833,
       "step": 121920
     },
     {
-      "epoch": 4.29,
-      "learning_rate": 2.610716551931325e-06,
-      "loss": 0.2491,
+      "epoch": 4.394168738962771,
+      "grad_norm": 0.2453107088804245,
+      "learning_rate": 1.910053276476878e-06,
+      "loss": 0.3617,
       "step": 121925
     },
     {
-      "epoch": 4.29,
-      "learning_rate": 2.609449250053814e-06,
-      "loss": 0.2611,
+      "epoch": 4.394348938623995,
+      "grad_norm": 0.2022264003753662,
+      "learning_rate": 1.9089347301182945e-06,
+      "loss": 0.3762,
       "step": 121930
     },
     {
-      "epoch": 4.29,
-      "learning_rate": 2.6081822389028427e-06,
-      "loss": 0.2375,
+      "epoch": 4.39452913828522,
+      "grad_norm": 0.23200011253356934,
+      "learning_rate": 1.9078164983709984e-06,
+      "loss": 0.3415,
       "step": 121935
     },
     {
-      "epoch": 4.29,
-      "learning_rate": 2.6069155184948677e-06,
-      "loss": 0.2616,
+      "epoch": 4.394709337946445,
+      "grad_norm": 0.24369412660598755,
+      "learning_rate": 1.9066985812502353e-06,
+      "loss": 0.3742,
       "step": 121940
     },
     {
-      "epoch": 4.29,
-      "learning_rate": 2.605649088846343e-06,
-      "loss": 0.2383,
+      "epoch": 4.3948895376076695,
+      "grad_norm": 0.27759307622909546,
+      "learning_rate": 1.9055809787712348e-06,
+      "loss": 0.3759,
       "step": 121945
     },
     {
-      "epoch": 4.29,
-      "learning_rate": 2.604382949973705e-06,
-      "loss": 0.2471,
+      "epoch": 4.395069737268894,
+      "grad_norm": 0.28923094272613525,
+      "learning_rate": 1.9044636909492208e-06,
+      "loss": 0.3658,
       "step": 121950
     },
     {
-      "epoch": 4.29,
-      "learning_rate": 2.603117101893396e-06,
-      "loss": 0.2555,
+      "epoch": 4.395249936930119,
+      "grad_norm": 0.3017236888408661,
+      "learning_rate": 1.9033467177994174e-06,
+      "loss": 0.3987,
       "step": 121955
     },
     {
-      "epoch": 4.29,
-      "learning_rate": 2.601851544621847e-06,
-      "loss": 0.2565,
+      "epoch": 4.395430136591343,
+      "grad_norm": 0.31043195724487305,
+      "learning_rate": 1.902230059337043e-06,
+      "loss": 0.3689,
       "step": 121960
     },
     {
-      "epoch": 4.29,
-      "learning_rate": 2.6005862781754986e-06,
-      "loss": 0.242,
+      "epoch": 4.395610336252568,
+      "grad_norm": 0.25115767121315,
+      "learning_rate": 1.9011137155773101e-06,
+      "loss": 0.3732,
       "step": 121965
     },
     {
-      "epoch": 4.29,
-      "learning_rate": 2.599321302570776e-06,
-      "loss": 0.2707,
+      "epoch": 4.395790535913792,
+      "grad_norm": 0.23707129061222076,
+      "learning_rate": 1.8999976865354374e-06,
+      "loss": 0.37,
       "step": 121970
     },
     {
-      "epoch": 4.29,
-      "learning_rate": 2.598056617824099e-06,
-      "loss": 0.2544,
+      "epoch": 4.395970735575017,
+      "grad_norm": 0.2299506962299347,
+      "learning_rate": 1.8988819722266154e-06,
+      "loss": 0.3943,
       "step": 121975
     },
     {
-      "epoch": 4.29,
-      "learning_rate": 2.5967922239518966e-06,
-      "loss": 0.2463,
+      "epoch": 4.396150935236242,
+      "grad_norm": 0.2858457565307617,
+      "learning_rate": 1.8977665726660598e-06,
+      "loss": 0.3703,
       "step": 121980
     },
     {
-      "epoch": 4.29,
-      "learning_rate": 2.595528120970586e-06,
-      "loss": 0.2752,
+      "epoch": 4.3963311348974665,
+      "grad_norm": 0.2686900198459625,
+      "learning_rate": 1.8966514878689612e-06,
+      "loss": 0.3801,
       "step": 121985
     },
     {
-      "epoch": 4.29,
-      "learning_rate": 2.5942643088965712e-06,
-      "loss": 0.2545,
+      "epoch": 4.396511334558691,
+      "grad_norm": 0.25534695386886597,
+      "learning_rate": 1.8955367178505073e-06,
+      "loss": 0.3953,
       "step": 121990
     },
     {
-      "epoch": 4.29,
-      "learning_rate": 2.593000787746269e-06,
-      "loss": 0.2346,
+      "epoch": 4.396691534219916,
+      "grad_norm": 0.23531198501586914,
+      "learning_rate": 1.8944222626259028e-06,
+      "loss": 0.3857,
       "step": 121995
     },
     {
-      "epoch": 4.29,
-      "learning_rate": 2.591737557536089e-06,
-      "loss": 0.2512,
+      "epoch": 4.39687173388114,
+      "grad_norm": 0.23012003302574158,
+      "learning_rate": 1.8933081222103189e-06,
+      "loss": 0.4072,
       "step": 122000
     },
     {
-      "epoch": 4.29,
-      "eval_loss": 0.24890007078647614,
-      "eval_runtime": 10.5474,
-      "eval_samples_per_second": 9.481,
-      "eval_steps_per_second": 9.481,
+      "epoch": 4.39687173388114,
+      "eval_loss": 0.42878466844558716,
+      "eval_runtime": 3.5304,
+      "eval_samples_per_second": 28.325,
+      "eval_steps_per_second": 7.081,
       "step": 122000
     },
     {
-      "epoch": 4.29,
-      "learning_rate": 2.5904746182824292e-06,
-      "loss": 0.2254,
+      "epoch": 4.397051933542365,
+      "grad_norm": 0.2736528515815735,
+      "learning_rate": 1.8921942966189322e-06,
+      "loss": 0.4175,
       "step": 122005
     },
     {
-      "epoch": 4.29,
-      "learning_rate": 2.589211970001687e-06,
-      "loss": 0.2532,
+      "epoch": 4.397232133203589,
+      "grad_norm": 0.31495052576065063,
+      "learning_rate": 1.8910807858669277e-06,
+      "loss": 0.326,
       "step": 122010
     },
     {
-      "epoch": 4.29,
-      "learning_rate": 2.5879496127102556e-06,
-      "loss": 0.2646,
+      "epoch": 4.397412332864814,
+      "grad_norm": 0.27840882539749146,
+      "learning_rate": 1.889967589969474e-06,
+      "loss": 0.3971,
       "step": 122015
     },
     {
-      "epoch": 4.29,
-      "learning_rate": 2.586687546424532e-06,
-      "loss": 0.248,
+      "epoch": 4.397592532526039,
+      "grad_norm": 0.21493427455425262,
+      "learning_rate": 1.8888547089417368e-06,
+      "loss": 0.3331,
       "step": 122020
     },
     {
-      "epoch": 4.29,
-      "learning_rate": 2.585425771160899e-06,
-      "loss": 0.2387,
+      "epoch": 4.3977727321872635,
+      "grad_norm": 0.259475976228714,
+      "learning_rate": 1.8877421427988818e-06,
+      "loss": 0.3773,
       "step": 122025
     },
     {
-      "epoch": 4.29,
-      "learning_rate": 2.584164286935742e-06,
-      "loss": 0.2512,
+      "epoch": 4.397952931848488,
+      "grad_norm": 0.25516223907470703,
+      "learning_rate": 1.8866298915560604e-06,
+      "loss": 0.3608,
       "step": 122030
     },
     {
-      "epoch": 4.29,
-      "learning_rate": 2.582903093765432e-06,
-      "loss": 0.2743,
+      "epoch": 4.398133131509713,
+      "grad_norm": 0.25606194138526917,
+      "learning_rate": 1.8855179552284357e-06,
+      "loss": 0.3888,
       "step": 122035
     },
     {
-      "epoch": 4.29,
-      "learning_rate": 2.5816421916663585e-06,
-      "loss": 0.2504,
+      "epoch": 4.398313331170938,
+      "grad_norm": 0.260857492685318,
+      "learning_rate": 1.8844063338311542e-06,
+      "loss": 0.3505,
       "step": 122040
     },
     {
-      "epoch": 4.29,
-      "learning_rate": 2.5803815806548835e-06,
-      "loss": 0.2552,
+      "epoch": 4.398493530832162,
+      "grad_norm": 0.22685441374778748,
+      "learning_rate": 1.8832950273793586e-06,
+      "loss": 0.3711,
       "step": 122045
     },
     {
-      "epoch": 4.29,
-      "learning_rate": 2.5791212607473804e-06,
-      "loss": 0.2443,
+      "epoch": 4.398673730493386,
+      "grad_norm": 0.24304035305976868,
+      "learning_rate": 1.882184035888196e-06,
+      "loss": 0.3749,
       "step": 122050
     },
     {
-      "epoch": 4.29,
-      "learning_rate": 2.57786123196021e-06,
-      "loss": 0.2674,
+      "epoch": 4.398853930154611,
+      "grad_norm": 0.31736481189727783,
+      "learning_rate": 1.8810733593727952e-06,
+      "loss": 0.352,
       "step": 122055
     },
     {
-      "epoch": 4.29,
-      "learning_rate": 2.5766014943097385e-06,
-      "loss": 0.2529,
+      "epoch": 4.399034129815836,
+      "grad_norm": 0.190985307097435,
+      "learning_rate": 1.8799629978482975e-06,
+      "loss": 0.3598,
       "step": 122060
     },
     {
-      "epoch": 4.29,
-      "learning_rate": 2.575342047812318e-06,
-      "loss": 0.2512,
+      "epoch": 4.3992143294770605,
+      "grad_norm": 0.26111340522766113,
+      "learning_rate": 1.878852951329832e-06,
+      "loss": 0.3519,
       "step": 122065
     },
     {
-      "epoch": 4.29,
-      "learning_rate": 2.574082892484303e-06,
-      "loss": 0.2352,
+      "epoch": 4.399394529138285,
+      "grad_norm": 0.20273645222187042,
+      "learning_rate": 1.877743219832509e-06,
+      "loss": 0.3546,
       "step": 122070
     },
     {
-      "epoch": 4.29,
-      "learning_rate": 2.5728240283420383e-06,
-      "loss": 0.2629,
+      "epoch": 4.39957472879951,
+      "grad_norm": 0.21357549726963043,
+      "learning_rate": 1.8766338033714636e-06,
+      "loss": 0.3868,
       "step": 122075
     },
     {
-      "epoch": 4.3,
-      "learning_rate": 2.571565455401878e-06,
-      "loss": 0.2581,
+      "epoch": 4.399754928460735,
+      "grad_norm": 0.2500903010368347,
+      "learning_rate": 1.8755247019618034e-06,
+      "loss": 0.3884,
       "step": 122080
     },
     {
-      "epoch": 4.3,
-      "learning_rate": 2.570307173680159e-06,
-      "loss": 0.274,
+      "epoch": 4.399935128121959,
+      "grad_norm": 0.23067040741443634,
+      "learning_rate": 1.8744159156186492e-06,
+      "loss": 0.3731,
       "step": 122085
     },
     {
-      "epoch": 4.3,
-      "learning_rate": 2.5690491831932147e-06,
-      "loss": 0.2403,
+      "epoch": 4.400115327783184,
+      "grad_norm": 0.23591923713684082,
+      "learning_rate": 1.8733074443570947e-06,
+      "loss": 0.3813,
       "step": 122090
     },
     {
-      "epoch": 4.3,
-      "learning_rate": 2.567791483957388e-06,
-      "loss": 0.2349,
+      "epoch": 4.400295527444408,
+      "grad_norm": 0.251615434885025,
+      "learning_rate": 1.8721992881922473e-06,
+      "loss": 0.3614,
       "step": 122095
     },
     {
-      "epoch": 4.3,
-      "learning_rate": 2.5665340759890024e-06,
-      "loss": 0.2499,
+      "epoch": 4.400475727105633,
+      "grad_norm": 0.2689308524131775,
+      "learning_rate": 1.8710914471392088e-06,
+      "loss": 0.4076,
       "step": 122100
     },
     {
-      "epoch": 4.3,
-      "learning_rate": 2.5652769593043912e-06,
-      "loss": 0.261,
+      "epoch": 4.4006559267668575,
+      "grad_norm": 0.2680433392524719,
+      "learning_rate": 1.8699839212130727e-06,
+      "loss": 0.3672,
       "step": 122105
     },
     {
-      "epoch": 4.3,
-      "learning_rate": 2.5640201339198697e-06,
-      "loss": 0.2513,
+      "epoch": 4.400836126428082,
+      "grad_norm": 0.2669503092765808,
+      "learning_rate": 1.8688767104289267e-06,
+      "loss": 0.3589,
       "step": 122110
     },
     {
-      "epoch": 4.3,
-      "learning_rate": 2.5627635998517666e-06,
-      "loss": 0.2698,
+      "epoch": 4.401016326089307,
+      "grad_norm": 0.23472002148628235,
+      "learning_rate": 1.8677698148018591e-06,
+      "loss": 0.3815,
       "step": 122115
     },
     {
-      "epoch": 4.3,
-      "learning_rate": 2.561507357116391e-06,
-      "loss": 0.2702,
+      "epoch": 4.401196525750532,
+      "grad_norm": 0.2634426951408386,
+      "learning_rate": 1.8666632343469437e-06,
+      "loss": 0.3885,
       "step": 122120
     },
     {
-      "epoch": 4.3,
-      "learning_rate": 2.560251405730055e-06,
-      "loss": 0.242,
+      "epoch": 4.401376725411756,
+      "grad_norm": 0.22323672473430634,
+      "learning_rate": 1.8655569690792686e-06,
+      "loss": 0.3725,
       "step": 122125
     },
     {
-      "epoch": 4.3,
-      "learning_rate": 2.5589957457090586e-06,
-      "loss": 0.2438,
+      "epoch": 4.401556925072981,
+      "grad_norm": 0.1853046715259552,
+      "learning_rate": 1.8644510190138992e-06,
+      "loss": 0.3773,
       "step": 122130
     },
     {
-      "epoch": 4.3,
-      "learning_rate": 2.55774037706972e-06,
-      "loss": 0.2389,
+      "epoch": 4.401737124734206,
+      "grad_norm": 0.260140061378479,
+      "learning_rate": 1.8633453841659042e-06,
+      "loss": 0.351,
       "step": 122135
     },
     {
-      "epoch": 4.3,
-      "learning_rate": 2.5564852998283344e-06,
-      "loss": 0.2274,
+      "epoch": 4.401917324395431,
+      "grad_norm": 0.26455867290496826,
+      "learning_rate": 1.8622400645503525e-06,
+      "loss": 0.3379,
       "step": 122140
     },
     {
-      "epoch": 4.3,
-      "learning_rate": 2.5552305140011944e-06,
-      "loss": 0.2233,
+      "epoch": 4.402097524056654,
+      "grad_norm": 0.33009883761405945,
+      "learning_rate": 1.861135060182298e-06,
+      "loss": 0.3934,
       "step": 122145
     },
     {
-      "epoch": 4.3,
-      "learning_rate": 2.5539760196045896e-06,
-      "loss": 0.2624,
+      "epoch": 4.402277723717879,
+      "grad_norm": 0.2440415918827057,
+      "learning_rate": 1.860030371076793e-06,
+      "loss": 0.3411,
       "step": 122150
     },
     {
-      "epoch": 4.3,
-      "learning_rate": 2.5527218166548154e-06,
-      "loss": 0.2347,
+      "epoch": 4.402457923379104,
+      "grad_norm": 0.2990182340145111,
+      "learning_rate": 1.8589259972489003e-06,
+      "loss": 0.4079,
       "step": 122155
     },
     {
-      "epoch": 4.3,
-      "learning_rate": 2.551467905168159e-06,
-      "loss": 0.2507,
+      "epoch": 4.402638123040329,
+      "grad_norm": 0.2691243588924408,
+      "learning_rate": 1.8578219387136608e-06,
+      "loss": 0.379,
       "step": 122160
     },
     {
-      "epoch": 4.3,
-      "learning_rate": 2.550214285160896e-06,
-      "loss": 0.2562,
+      "epoch": 4.402818322701553,
+      "grad_norm": 0.23940818011760712,
+      "learning_rate": 1.8567181954861152e-06,
+      "loss": 0.3714,
       "step": 122165
     },
     {
-      "epoch": 4.3,
-      "learning_rate": 2.5489609566493015e-06,
-      "loss": 0.2428,
+      "epoch": 4.402998522362778,
+      "grad_norm": 0.23924623429775238,
+      "learning_rate": 1.8556147675813068e-06,
+      "loss": 0.354,
       "step": 122170
     },
     {
-      "epoch": 4.3,
-      "learning_rate": 2.5477079196496584e-06,
-      "loss": 0.282,
+      "epoch": 4.403178722024003,
+      "grad_norm": 0.2857334017753601,
+      "learning_rate": 1.85451165501426e-06,
+      "loss": 0.398,
       "step": 122175
     },
     {
-      "epoch": 4.3,
-      "learning_rate": 2.5464551741782334e-06,
-      "loss": 0.2664,
+      "epoch": 4.403358921685228,
+      "grad_norm": 0.26370975375175476,
+      "learning_rate": 1.8534088578000181e-06,
+      "loss": 0.3762,
       "step": 122180
     },
     {
-      "epoch": 4.3,
-      "learning_rate": 2.5452027202512883e-06,
-      "loss": 0.2606,
+      "epoch": 4.403539121346451,
+      "grad_norm": 0.24754032492637634,
+      "learning_rate": 1.8523063759535969e-06,
+      "loss": 0.3867,
       "step": 122185
     },
     {
-      "epoch": 4.3,
-      "learning_rate": 2.543950557885083e-06,
-      "loss": 0.2485,
+      "epoch": 4.403719321007676,
+      "grad_norm": 0.24139243364334106,
+      "learning_rate": 1.851204209490015e-06,
+      "loss": 0.3443,
       "step": 122190
     },
     {
-      "epoch": 4.3,
-      "learning_rate": 2.542698687095885e-06,
-      "loss": 0.2526,
+      "epoch": 4.403899520668901,
+      "grad_norm": 0.2864011526107788,
+      "learning_rate": 1.8501023584242965e-06,
+      "loss": 0.3501,
       "step": 122195
     },
     {
-      "epoch": 4.3,
-      "learning_rate": 2.541447107899947e-06,
-      "loss": 0.2412,
+      "epoch": 4.404079720330126,
+      "grad_norm": 0.2613162398338318,
+      "learning_rate": 1.8490008227714545e-06,
+      "loss": 0.3541,
       "step": 122200
     },
     {
-      "epoch": 4.3,
-      "learning_rate": 2.5401958203135097e-06,
-      "loss": 0.2406,
+      "epoch": 4.40425991999135,
+      "grad_norm": 0.2530914843082428,
+      "learning_rate": 1.8478996025464933e-06,
+      "loss": 0.3385,
       "step": 122205
     },
     {
-      "epoch": 4.3,
-      "learning_rate": 2.5389448243528345e-06,
-      "loss": 0.2505,
+      "epoch": 4.404440119652575,
+      "grad_norm": 0.2663853168487549,
+      "learning_rate": 1.8467986977644153e-06,
+      "loss": 0.3808,
       "step": 122210
     },
     {
-      "epoch": 4.3,
-      "learning_rate": 2.5376941200341538e-06,
-      "loss": 0.261,
+      "epoch": 4.4046203193138,
+      "grad_norm": 0.26814600825309753,
+      "learning_rate": 1.8456981084402192e-06,
+      "loss": 0.3838,
       "step": 122215
     },
     {
-      "epoch": 4.3,
-      "learning_rate": 2.5364437073737146e-06,
-      "loss": 0.2548,
+      "epoch": 4.404800518975025,
+      "grad_norm": 0.2669377326965332,
+      "learning_rate": 1.8445978345889097e-06,
+      "loss": 0.381,
       "step": 122220
     },
     {
-      "epoch": 4.3,
-      "learning_rate": 2.5351935863877464e-06,
-      "loss": 0.241,
+      "epoch": 4.404980718636249,
+      "grad_norm": 0.2761158049106598,
+      "learning_rate": 1.8434978762254695e-06,
+      "loss": 0.3659,
       "step": 122225
     },
     {
-      "epoch": 4.3,
-      "learning_rate": 2.533943757092491e-06,
-      "loss": 0.2481,
+      "epoch": 4.405160918297474,
+      "grad_norm": 0.24499890208244324,
+      "learning_rate": 1.8423982333648893e-06,
+      "loss": 0.3993,
       "step": 122230
     },
     {
-      "epoch": 4.3,
-      "learning_rate": 2.532694219504167e-06,
-      "loss": 0.2549,
+      "epoch": 4.405341117958698,
+      "grad_norm": 0.24173754453659058,
+      "learning_rate": 1.8412989060221485e-06,
+      "loss": 0.3923,
       "step": 122235
     },
     {
-      "epoch": 4.3,
-      "learning_rate": 2.531444973639005e-06,
-      "loss": 0.2555,
+      "epoch": 4.405521317619923,
+      "grad_norm": 0.2643168568611145,
+      "learning_rate": 1.8401998942122218e-06,
+      "loss": 0.3681,
       "step": 122240
     },
     {
-      "epoch": 4.3,
-      "learning_rate": 2.5301960195132147e-06,
-      "loss": 0.2465,
+      "epoch": 4.405701517281147,
+      "grad_norm": 0.24129536747932434,
+      "learning_rate": 1.8391011979500939e-06,
+      "loss": 0.3466,
       "step": 122245
     },
     {
-      "epoch": 4.3,
-      "learning_rate": 2.52894735714303e-06,
-      "loss": 0.2605,
+      "epoch": 4.405881716942372,
+      "grad_norm": 0.2655704915523529,
+      "learning_rate": 1.8380028172507308e-06,
+      "loss": 0.3822,
       "step": 122250
     },
     {
-      "epoch": 4.3,
-      "learning_rate": 2.5276989865446522e-06,
-      "loss": 0.2623,
+      "epoch": 4.406061916603597,
+      "grad_norm": 0.2480999380350113,
+      "learning_rate": 1.8369047521290872e-06,
+      "loss": 0.3705,
       "step": 122255
     },
     {
-      "epoch": 4.3,
-      "learning_rate": 2.5264509077342875e-06,
-      "loss": 0.2271,
+      "epoch": 4.4062421162648215,
+      "grad_norm": 0.3369860053062439,
+      "learning_rate": 1.8358070026001345e-06,
+      "loss": 0.3736,
       "step": 122260
     },
     {
-      "epoch": 4.3,
-      "learning_rate": 2.5252031207281534e-06,
-      "loss": 0.2602,
+      "epoch": 4.406422315926046,
+      "grad_norm": 0.3019421100616455,
+      "learning_rate": 1.8347095686788247e-06,
+      "loss": 0.3587,
       "step": 122265
     },
     {
-      "epoch": 4.3,
-      "learning_rate": 2.52395562554244e-06,
-      "loss": 0.261,
+      "epoch": 4.406602515587271,
+      "grad_norm": 0.26029887795448303,
+      "learning_rate": 1.8336124503801179e-06,
+      "loss": 0.3561,
       "step": 122270
     },
     {
-      "epoch": 4.3,
-      "learning_rate": 2.5227084221933567e-06,
-      "loss": 0.2442,
+      "epoch": 4.406782715248496,
+      "grad_norm": 0.31131497025489807,
+      "learning_rate": 1.8325156477189547e-06,
+      "loss": 0.3614,
       "step": 122275
     },
     {
-      "epoch": 4.3,
-      "learning_rate": 2.52146151069709e-06,
-      "loss": 0.2489,
+      "epoch": 4.40696291490972,
+      "grad_norm": 0.24049854278564453,
+      "learning_rate": 1.8314191607102738e-06,
+      "loss": 0.3702,
       "step": 122280
     },
     {
-      "epoch": 4.3,
-      "learning_rate": 2.5202148910698276e-06,
-      "loss": 0.2632,
+      "epoch": 4.407143114570944,
+      "grad_norm": 0.23440882563591003,
+      "learning_rate": 1.8303229893690266e-06,
+      "loss": 0.3774,
       "step": 122285
     },
     {
-      "epoch": 4.3,
-      "learning_rate": 2.5189685633277644e-06,
-      "loss": 0.2268,
+      "epoch": 4.407323314232169,
+      "grad_norm": 0.22003592550754547,
+      "learning_rate": 1.829227133710143e-06,
+      "loss": 0.3919,
       "step": 122290
     },
     {
-      "epoch": 4.3,
-      "learning_rate": 2.5177225274870792e-06,
-      "loss": 0.2557,
+      "epoch": 4.407503513893394,
+      "grad_norm": 0.1988513171672821,
+      "learning_rate": 1.8281315937485527e-06,
+      "loss": 0.3903,
       "step": 122295
     },
     {
-      "epoch": 4.3,
-      "learning_rate": 2.516476783563951e-06,
-      "loss": 0.2423,
+      "epoch": 4.4076837135546185,
+      "grad_norm": 0.19561129808425903,
+      "learning_rate": 1.8270363694991855e-06,
+      "loss": 0.3364,
       "step": 122300
     },
     {
-      "epoch": 4.3,
-      "learning_rate": 2.5152313315745496e-06,
-      "loss": 0.2396,
+      "epoch": 4.407863913215843,
+      "grad_norm": 0.2583327889442444,
+      "learning_rate": 1.825941460976957e-06,
+      "loss": 0.3895,
       "step": 122305
     },
     {
-      "epoch": 4.3,
-      "learning_rate": 2.5139861715350545e-06,
-      "loss": 0.2371,
+      "epoch": 4.408044112877068,
+      "grad_norm": 0.2395971715450287,
+      "learning_rate": 1.8248468681967916e-06,
+      "loss": 0.3592,
       "step": 122310
     },
     {
-      "epoch": 4.3,
-      "learning_rate": 2.512741303461627e-06,
-      "loss": 0.2349,
+      "epoch": 4.408224312538293,
+      "grad_norm": 0.25403663516044617,
+      "learning_rate": 1.8237525911736025e-06,
+      "loss": 0.3655,
       "step": 122315
     },
     {
-      "epoch": 4.3,
-      "learning_rate": 2.511496727370438e-06,
-      "loss": 0.2572,
+      "epoch": 4.4084045121995175,
+      "grad_norm": 0.27949368953704834,
+      "learning_rate": 1.8226586299222943e-06,
+      "loss": 0.3404,
       "step": 122320
     },
     {
-      "epoch": 4.3,
-      "learning_rate": 2.5102524432776386e-06,
-      "loss": 0.2773,
+      "epoch": 4.408584711860741,
+      "grad_norm": 0.2646028399467468,
+      "learning_rate": 1.821564984457777e-06,
+      "loss": 0.3935,
       "step": 122325
     },
     {
-      "epoch": 4.3,
-      "learning_rate": 2.509008451199396e-06,
-      "loss": 0.2501,
+      "epoch": 4.408764911521966,
+      "grad_norm": 0.32724955677986145,
+      "learning_rate": 1.8204716547949502e-06,
+      "loss": 0.3568,
       "step": 122330
     },
     {
-      "epoch": 4.3,
-      "learning_rate": 2.5077647511518557e-06,
-      "loss": 0.2553,
+      "epoch": 4.408945111183191,
+      "grad_norm": 0.26123833656311035,
+      "learning_rate": 1.8193786409487046e-06,
+      "loss": 0.3589,
       "step": 122335
     },
     {
-      "epoch": 4.3,
-      "learning_rate": 2.5065213431511637e-06,
-      "loss": 0.238,
+      "epoch": 4.4091253108444155,
+      "grad_norm": 0.2541702389717102,
+      "learning_rate": 1.8182859429339422e-06,
+      "loss": 0.3606,
       "step": 122340
     },
     {
-      "epoch": 4.3,
-      "learning_rate": 2.50527822721347e-06,
-      "loss": 0.2413,
+      "epoch": 4.40930551050564,
+      "grad_norm": 0.25994282960891724,
+      "learning_rate": 1.8171935607655427e-06,
+      "loss": 0.3641,
       "step": 122345
     },
     {
-      "epoch": 4.3,
-      "learning_rate": 2.504035403354915e-06,
-      "loss": 0.2534,
+      "epoch": 4.409485710166865,
+      "grad_norm": 0.31410351395606995,
+      "learning_rate": 1.816101494458397e-06,
+      "loss": 0.3891,
       "step": 122350
     },
     {
-      "epoch": 4.3,
-      "learning_rate": 2.502792871591636e-06,
-      "loss": 0.2371,
+      "epoch": 4.40966590982809,
+      "grad_norm": 0.23764289915561676,
+      "learning_rate": 1.8150097440273767e-06,
+      "loss": 0.407,
       "step": 122355
     },
     {
-      "epoch": 4.3,
-      "learning_rate": 2.5015506319397615e-06,
-      "loss": 0.2468,
+      "epoch": 4.409846109489314,
+      "grad_norm": 0.24512486159801483,
+      "learning_rate": 1.8139183094873558e-06,
+      "loss": 0.3627,
       "step": 122360
     },
     {
-      "epoch": 4.31,
-      "learning_rate": 2.5003086844154283e-06,
-      "loss": 0.2515,
+      "epoch": 4.410026309150539,
+      "grad_norm": 0.2727164328098297,
+      "learning_rate": 1.812827190853217e-06,
+      "loss": 0.3738,
       "step": 122365
     },
     {
-      "epoch": 4.31,
-      "learning_rate": 2.49906702903476e-06,
-      "loss": 0.2508,
+      "epoch": 4.410206508811763,
+      "grad_norm": 0.2821141183376312,
+      "learning_rate": 1.8117363881398174e-06,
+      "loss": 0.3829,
       "step": 122370
     },
     {
-      "epoch": 4.31,
-      "learning_rate": 2.4978256658138715e-06,
-      "loss": 0.2537,
+      "epoch": 4.410386708472988,
+      "grad_norm": 0.22460602223873138,
+      "learning_rate": 1.8106459013620126e-06,
+      "loss": 0.3681,
       "step": 122375
     },
     {
-      "epoch": 4.31,
-      "learning_rate": 2.496584594768889e-06,
-      "loss": 0.2531,
+      "epoch": 4.4105669081342125,
+      "grad_norm": 0.24663448333740234,
+      "learning_rate": 1.8095557305346734e-06,
+      "loss": 0.3807,
       "step": 122380
     },
     {
-      "epoch": 4.31,
-      "learning_rate": 2.4953438159159267e-06,
-      "loss": 0.244,
+      "epoch": 4.410747107795437,
+      "grad_norm": 0.32760992646217346,
+      "learning_rate": 1.8084658756726463e-06,
+      "loss": 0.3769,
       "step": 122385
     },
     {
-      "epoch": 4.31,
-      "learning_rate": 2.494103329271097e-06,
-      "loss": 0.2435,
+      "epoch": 4.410927307456662,
+      "grad_norm": 0.26017165184020996,
+      "learning_rate": 1.8073763367907836e-06,
+      "loss": 0.4191,
       "step": 122390
     },
     {
-      "epoch": 4.31,
-      "learning_rate": 2.492863134850501e-06,
-      "loss": 0.2725,
+      "epoch": 4.411107507117887,
+      "grad_norm": 0.2591916620731354,
+      "learning_rate": 1.806287113903929e-06,
+      "loss": 0.3955,
       "step": 122395
     },
     {
-      "epoch": 4.31,
-      "learning_rate": 2.4916232326702428e-06,
-      "loss": 0.2411,
+      "epoch": 4.411287706779111,
+      "grad_norm": 0.2518169581890106,
+      "learning_rate": 1.8051982070269147e-06,
+      "loss": 0.3571,
       "step": 122400
     },
     {
-      "epoch": 4.31,
-      "learning_rate": 2.4903836227464258e-06,
-      "loss": 0.2417,
+      "epoch": 4.411467906440336,
+      "grad_norm": 0.3138507604598999,
+      "learning_rate": 1.8041096161745902e-06,
+      "loss": 0.3739,
       "step": 122405
     },
     {
-      "epoch": 4.31,
-      "learning_rate": 2.489144305095145e-06,
-      "loss": 0.2568,
+      "epoch": 4.411648106101561,
+      "grad_norm": 0.22780956327915192,
+      "learning_rate": 1.8030213413617798e-06,
+      "loss": 0.3652,
       "step": 122410
     },
     {
-      "epoch": 4.31,
-      "learning_rate": 2.487905279732489e-06,
-      "loss": 0.2252,
+      "epoch": 4.411828305762786,
+      "grad_norm": 0.2590252161026001,
+      "learning_rate": 1.801933382603313e-06,
+      "loss": 0.3928,
       "step": 122415
     },
     {
-      "epoch": 4.31,
-      "learning_rate": 2.4866665466745404e-06,
-      "loss": 0.2488,
+      "epoch": 4.4120085054240095,
+      "grad_norm": 0.2547787129878998,
+      "learning_rate": 1.8008457399140144e-06,
+      "loss": 0.3554,
       "step": 122420
     },
     {
-      "epoch": 4.31,
-      "learning_rate": 2.4854281059373984e-06,
-      "loss": 0.2501,
+      "epoch": 4.412188705085234,
+      "grad_norm": 0.29435890913009644,
+      "learning_rate": 1.7997584133086943e-06,
+      "loss": 0.4092,
       "step": 122425
     },
     {
-      "epoch": 4.31,
-      "learning_rate": 2.484189957537128e-06,
-      "loss": 0.2605,
+      "epoch": 4.412368904746459,
+      "grad_norm": 0.23160386085510254,
+      "learning_rate": 1.7986714028021794e-06,
+      "loss": 0.3793,
       "step": 122430
     },
     {
-      "epoch": 4.31,
-      "learning_rate": 2.4829521014898103e-06,
-      "loss": 0.2593,
+      "epoch": 4.412549104407684,
+      "grad_norm": 0.1971481442451477,
+      "learning_rate": 1.7975847084092806e-06,
+      "loss": 0.3858,
       "step": 122435
     },
     {
-      "epoch": 4.31,
-      "learning_rate": 2.4817145378115274e-06,
-      "loss": 0.2526,
+      "epoch": 4.412729304068908,
+      "grad_norm": 0.29055434465408325,
+      "learning_rate": 1.7964983301447886e-06,
+      "loss": 0.4172,
       "step": 122440
     },
     {
-      "epoch": 4.31,
-      "learning_rate": 2.480477266518341e-06,
-      "loss": 0.249,
+      "epoch": 4.412909503730133,
+      "grad_norm": 0.2608969509601593,
+      "learning_rate": 1.7954122680235164e-06,
+      "loss": 0.3736,
       "step": 122445
     },
     {
-      "epoch": 4.31,
-      "learning_rate": 2.4792402876263165e-06,
-      "loss": 0.2751,
+      "epoch": 4.413089703391358,
+      "grad_norm": 0.2528577148914337,
+      "learning_rate": 1.794326522060258e-06,
+      "loss": 0.3502,
       "step": 122450
     },
     {
-      "epoch": 4.31,
-      "learning_rate": 2.47800360115151e-06,
-      "loss": 0.2493,
+      "epoch": 4.413269903052583,
+      "grad_norm": 0.21361927688121796,
+      "learning_rate": 1.7932410922698122e-06,
+      "loss": 0.3278,
       "step": 122455
     },
     {
-      "epoch": 4.31,
-      "learning_rate": 2.4767672071099897e-06,
-      "loss": 0.2384,
+      "epoch": 4.4134501027138064,
+      "grad_norm": 0.33277052640914917,
+      "learning_rate": 1.7921559786669679e-06,
+      "loss": 0.3626,
       "step": 122460
     },
     {
-      "epoch": 4.31,
-      "learning_rate": 2.475531105517803e-06,
-      "loss": 0.257,
+      "epoch": 4.413630302375031,
+      "grad_norm": 0.21200527250766754,
+      "learning_rate": 1.7910711812664931e-06,
+      "loss": 0.3842,
       "step": 122465
     },
     {
-      "epoch": 4.31,
-      "learning_rate": 2.474295296391002e-06,
-      "loss": 0.2513,
+      "epoch": 4.413810502036256,
+      "grad_norm": 0.2637445628643036,
+      "learning_rate": 1.7899867000831878e-06,
+      "loss": 0.3866,
       "step": 122470
     },
     {
-      "epoch": 4.31,
-      "learning_rate": 2.473059779745626e-06,
-      "loss": 0.2359,
+      "epoch": 4.413990701697481,
+      "grad_norm": 0.3103804588317871,
+      "learning_rate": 1.78890253513182e-06,
+      "loss": 0.3754,
       "step": 122475
     },
     {
-      "epoch": 4.31,
-      "learning_rate": 2.4718245555977266e-06,
-      "loss": 0.2462,
+      "epoch": 4.414170901358705,
+      "grad_norm": 0.25578418374061584,
+      "learning_rate": 1.7878186864271618e-06,
+      "loss": 0.3487,
       "step": 122480
     },
     {
-      "epoch": 4.31,
-      "learning_rate": 2.4705896239633347e-06,
-      "loss": 0.2384,
+      "epoch": 4.41435110101993,
+      "grad_norm": 0.23292893171310425,
+      "learning_rate": 1.786735153983976e-06,
+      "loss": 0.3688,
       "step": 122485
     },
     {
-      "epoch": 4.31,
-      "learning_rate": 2.469354984858496e-06,
-      "loss": 0.2534,
+      "epoch": 4.414531300681155,
+      "grad_norm": 0.23511111736297607,
+      "learning_rate": 1.7856519378170289e-06,
+      "loss": 0.3837,
       "step": 122490
     },
     {
-      "epoch": 4.31,
-      "learning_rate": 2.468120638299226e-06,
-      "loss": 0.2218,
+      "epoch": 4.41471150034238,
+      "grad_norm": 0.3069644272327423,
+      "learning_rate": 1.7845690379410835e-06,
+      "loss": 0.3729,
       "step": 122495
     },
     {
-      "epoch": 4.31,
-      "learning_rate": 2.466886584301567e-06,
-      "loss": 0.2444,
+      "epoch": 4.414891700003604,
+      "grad_norm": 0.26386865973472595,
+      "learning_rate": 1.783486454370889e-06,
+      "loss": 0.379,
       "step": 122500
     },
     {
-      "epoch": 4.31,
-      "eval_loss": 0.2489146739244461,
-      "eval_runtime": 10.5432,
-      "eval_samples_per_second": 9.485,
-      "eval_steps_per_second": 9.485,
+      "epoch": 4.414891700003604,
+      "eval_loss": 0.42894604802131653,
+      "eval_runtime": 3.5255,
+      "eval_samples_per_second": 28.365,
+      "eval_steps_per_second": 7.091,
       "step": 122500
     },
     {
-      "epoch": 4.31,
-      "learning_rate": 2.465652822881537e-06,
-      "loss": 0.2727,
+      "epoch": 4.415071899664829,
+      "grad_norm": 0.2563088834285736,
+      "learning_rate": 1.7824041871211976e-06,
+      "loss": 0.3939,
       "step": 122505
     },
     {
-      "epoch": 4.31,
-      "learning_rate": 2.464419354055153e-06,
-      "loss": 0.262,
+      "epoch": 4.415252099326053,
+      "grad_norm": 0.2134285569190979,
+      "learning_rate": 1.781322236206756e-06,
+      "loss": 0.3831,
       "step": 122510
     },
     {
-      "epoch": 4.31,
-      "learning_rate": 2.463186177838428e-06,
-      "loss": 0.2561,
+      "epoch": 4.415432298987278,
+      "grad_norm": 0.21774479746818542,
+      "learning_rate": 1.7802406016423022e-06,
+      "loss": 0.3738,
       "step": 122515
     },
     {
-      "epoch": 4.31,
-      "learning_rate": 2.461953294247382e-06,
-      "loss": 0.2547,
+      "epoch": 4.415612498648502,
+      "grad_norm": 0.19313018023967743,
+      "learning_rate": 1.779159283442569e-06,
+      "loss": 0.34,
       "step": 122520
     },
     {
-      "epoch": 4.31,
-      "learning_rate": 2.4607207032980195e-06,
-      "loss": 0.2372,
+      "epoch": 4.415792698309727,
+      "grad_norm": 0.24598465859889984,
+      "learning_rate": 1.7780782816223001e-06,
+      "loss": 0.3575,
       "step": 122525
     },
     {
-      "epoch": 4.31,
-      "learning_rate": 2.4594884050063465e-06,
-      "loss": 0.244,
+      "epoch": 4.415972897970952,
+      "grad_norm": 0.2550036311149597,
+      "learning_rate": 1.77699759619622e-06,
+      "loss": 0.381,
       "step": 122530
     },
     {
-      "epoch": 4.31,
-      "learning_rate": 2.4582563993883563e-06,
-      "loss": 0.265,
+      "epoch": 4.416153097632177,
+      "grad_norm": 0.3060535192489624,
+      "learning_rate": 1.7759172271790498e-06,
+      "loss": 0.4224,
       "step": 122535
     },
     {
-      "epoch": 4.31,
-      "learning_rate": 2.4570246864600493e-06,
-      "loss": 0.2328,
+      "epoch": 4.416333297293401,
+      "grad_norm": 0.3150867223739624,
+      "learning_rate": 1.7748371745855114e-06,
+      "loss": 0.3928,
       "step": 122540
     },
     {
-      "epoch": 4.31,
-      "learning_rate": 2.4557932662374273e-06,
-      "loss": 0.2509,
+      "epoch": 4.416513496954626,
+      "grad_norm": 0.24567945301532745,
+      "learning_rate": 1.7737574384303152e-06,
+      "loss": 0.3674,
       "step": 122545
     },
     {
-      "epoch": 4.31,
-      "learning_rate": 2.4545621387364658e-06,
-      "loss": 0.2652,
+      "epoch": 4.416693696615851,
+      "grad_norm": 0.19101324677467346,
+      "learning_rate": 1.7726780187281854e-06,
+      "loss": 0.3462,
       "step": 122550
     },
     {
-      "epoch": 4.31,
-      "learning_rate": 2.453331303973164e-06,
-      "loss": 0.2345,
+      "epoch": 4.416873896277075,
+      "grad_norm": 0.3299705982208252,
+      "learning_rate": 1.771598915493819e-06,
+      "loss": 0.3893,
       "step": 122555
     },
     {
-      "epoch": 4.31,
-      "learning_rate": 2.4521007619634965e-06,
-      "loss": 0.2505,
+      "epoch": 4.417054095938299,
+      "grad_norm": 0.29562848806381226,
+      "learning_rate": 1.7705201287419121e-06,
+      "loss": 0.3921,
       "step": 122560
     },
     {
-      "epoch": 4.31,
-      "learning_rate": 2.4508705127234384e-06,
-      "loss": 0.243,
+      "epoch": 4.417234295599524,
+      "grad_norm": 0.2895338535308838,
+      "learning_rate": 1.769441658487178e-06,
+      "loss": 0.4102,
       "step": 122565
     },
     {
-      "epoch": 4.31,
-      "learning_rate": 2.4496405562689617e-06,
-      "loss": 0.2588,
+      "epoch": 4.417414495260749,
+      "grad_norm": 0.2296198308467865,
+      "learning_rate": 1.7683635047442998e-06,
+      "loss": 0.3691,
       "step": 122570
     },
     {
-      "epoch": 4.31,
-      "learning_rate": 2.4484108926160486e-06,
-      "loss": 0.2478,
+      "epoch": 4.4175946949219735,
+      "grad_norm": 0.24252089858055115,
+      "learning_rate": 1.7672856675279791e-06,
+      "loss": 0.4019,
       "step": 122575
     },
     {
-      "epoch": 4.31,
-      "learning_rate": 2.447181521780656e-06,
-      "loss": 0.2422,
+      "epoch": 4.417774894583198,
+      "grad_norm": 0.24175925552845,
+      "learning_rate": 1.7662081468528879e-06,
+      "loss": 0.3703,
       "step": 122580
     },
     {
-      "epoch": 4.31,
-      "learning_rate": 2.4459524437787485e-06,
-      "loss": 0.249,
+      "epoch": 4.417955094244423,
+      "grad_norm": 0.22977383434772491,
+      "learning_rate": 1.7651309427337087e-06,
+      "loss": 0.4019,
       "step": 122585
     },
     {
-      "epoch": 4.31,
-      "learning_rate": 2.44472365862628e-06,
-      "loss": 0.2487,
+      "epoch": 4.418135293905648,
+      "grad_norm": 0.28015124797821045,
+      "learning_rate": 1.764054055185127e-06,
+      "loss": 0.3954,
       "step": 122590
     },
     {
-      "epoch": 4.31,
-      "learning_rate": 2.443495166339213e-06,
-      "loss": 0.2348,
+      "epoch": 4.4183154935668725,
+      "grad_norm": 0.231642484664917,
+      "learning_rate": 1.7629774842218089e-06,
+      "loss": 0.3531,
       "step": 122595
     },
     {
-      "epoch": 4.31,
-      "learning_rate": 2.4422669669334986e-06,
-      "loss": 0.2635,
+      "epoch": 4.418495693228096,
+      "grad_norm": 0.24805359542369843,
+      "learning_rate": 1.7619012298584259e-06,
+      "loss": 0.4105,
       "step": 122600
     },
     {
-      "epoch": 4.31,
-      "learning_rate": 2.4410390604250795e-06,
-      "loss": 0.2381,
+      "epoch": 4.418675892889321,
+      "grad_norm": 0.29648274183273315,
+      "learning_rate": 1.7608252921096357e-06,
+      "loss": 0.3828,
       "step": 122605
     },
     {
-      "epoch": 4.31,
-      "learning_rate": 2.4398114468298985e-06,
-      "loss": 0.2357,
+      "epoch": 4.418856092550546,
+      "grad_norm": 0.28801605105400085,
+      "learning_rate": 1.7597496709901017e-06,
+      "loss": 0.3365,
       "step": 122610
     },
     {
-      "epoch": 4.31,
-      "learning_rate": 2.4385841261639035e-06,
-      "loss": 0.2647,
+      "epoch": 4.4190362922117705,
+      "grad_norm": 0.25377213954925537,
+      "learning_rate": 1.7586743665144816e-06,
+      "loss": 0.3936,
       "step": 122615
     },
     {
-      "epoch": 4.31,
-      "learning_rate": 2.4373570984430242e-06,
-      "loss": 0.2329,
+      "epoch": 4.419216491872995,
+      "grad_norm": 0.2505911588668823,
+      "learning_rate": 1.7575993786974221e-06,
+      "loss": 0.3729,
       "step": 122620
     },
     {
-      "epoch": 4.31,
-      "learning_rate": 2.436130363683195e-06,
-      "loss": 0.2431,
+      "epoch": 4.41939669153422,
+      "grad_norm": 0.2358839064836502,
+      "learning_rate": 1.7565247075535724e-06,
+      "loss": 0.3607,
       "step": 122625
     },
     {
-      "epoch": 4.31,
-      "learning_rate": 2.434903921900336e-06,
-      "loss": 0.2418,
+      "epoch": 4.419576891195445,
+      "grad_norm": 0.2155429720878601,
+      "learning_rate": 1.7554503530975735e-06,
+      "loss": 0.3934,
       "step": 122630
     },
     {
-      "epoch": 4.31,
-      "learning_rate": 2.4336777731103847e-06,
-      "loss": 0.2545,
+      "epoch": 4.4197570908566695,
+      "grad_norm": 0.21443898975849152,
+      "learning_rate": 1.7543763153440585e-06,
+      "loss": 0.3833,
       "step": 122635
     },
     {
-      "epoch": 4.31,
-      "learning_rate": 2.4324519173292565e-06,
-      "loss": 0.2493,
+      "epoch": 4.419937290517894,
+      "grad_norm": 0.2692784368991852,
+      "learning_rate": 1.7533025943076708e-06,
+      "loss": 0.3753,
       "step": 122640
     },
     {
-      "epoch": 4.31,
-      "learning_rate": 2.4312263545728605e-06,
-      "loss": 0.2496,
+      "epoch": 4.420117490179118,
+      "grad_norm": 0.1909954696893692,
+      "learning_rate": 1.752229190003038e-06,
+      "loss": 0.3936,
       "step": 122645
     },
     {
-      "epoch": 4.32,
-      "learning_rate": 2.4300010848571225e-06,
-      "loss": 0.2447,
+      "epoch": 4.420297689840343,
+      "grad_norm": 0.26652899384498596,
+      "learning_rate": 1.751156102444776e-06,
+      "loss": 0.3609,
       "step": 122650
     },
     {
-      "epoch": 4.32,
-      "learning_rate": 2.4287761081979445e-06,
-      "loss": 0.2414,
+      "epoch": 4.4204778895015675,
+      "grad_norm": 0.220448300242424,
+      "learning_rate": 1.7500833316475118e-06,
+      "loss": 0.3454,
       "step": 122655
     },
     {
-      "epoch": 4.32,
-      "learning_rate": 2.4275514246112353e-06,
-      "loss": 0.2578,
+      "epoch": 4.420658089162792,
+      "grad_norm": 0.2539389431476593,
+      "learning_rate": 1.7490108776258617e-06,
+      "loss": 0.3466,
       "step": 122660
     },
     {
-      "epoch": 4.32,
-      "learning_rate": 2.426327034112891e-06,
-      "loss": 0.266,
+      "epoch": 4.420838288824017,
+      "grad_norm": 0.2393910139799118,
+      "learning_rate": 1.7479387403944363e-06,
+      "loss": 0.345,
       "step": 122665
     },
     {
-      "epoch": 4.32,
-      "learning_rate": 2.425102936718818e-06,
-      "loss": 0.261,
+      "epoch": 4.421018488485242,
+      "grad_norm": 0.27179139852523804,
+      "learning_rate": 1.7468669199678462e-06,
+      "loss": 0.3895,
       "step": 122670
     },
     {
-      "epoch": 4.32,
-      "learning_rate": 2.423879132444906e-06,
-      "loss": 0.2403,
+      "epoch": 4.421198688146466,
+      "grad_norm": 0.2731141149997711,
+      "learning_rate": 1.7457954163606848e-06,
+      "loss": 0.3681,
       "step": 122675
     },
     {
-      "epoch": 4.32,
-      "learning_rate": 2.422655621307046e-06,
-      "loss": 0.2579,
+      "epoch": 4.421378887807691,
+      "grad_norm": 0.23041914403438568,
+      "learning_rate": 1.744724229587566e-06,
+      "loss": 0.3366,
       "step": 122680
     },
     {
-      "epoch": 4.32,
-      "learning_rate": 2.4214324033211167e-06,
-      "loss": 0.2365,
+      "epoch": 4.421559087468916,
+      "grad_norm": 0.2509845793247223,
+      "learning_rate": 1.743653359663075e-06,
+      "loss": 0.3587,
       "step": 122685
     },
     {
-      "epoch": 4.32,
-      "learning_rate": 2.420209478503016e-06,
-      "loss": 0.2446,
+      "epoch": 4.421739287130141,
+      "grad_norm": 0.250013530254364,
+      "learning_rate": 1.7425828066018056e-06,
+      "loss": 0.3569,
       "step": 122690
     },
     {
-      "epoch": 4.32,
-      "learning_rate": 2.4189868468686124e-06,
-      "loss": 0.2614,
+      "epoch": 4.4219194867913645,
+      "grad_norm": 0.27683958411216736,
+      "learning_rate": 1.7415125704183411e-06,
+      "loss": 0.3784,
       "step": 122695
     },
     {
-      "epoch": 4.32,
-      "learning_rate": 2.4177645084337787e-06,
-      "loss": 0.2611,
+      "epoch": 4.422099686452589,
+      "grad_norm": 0.2434561401605606,
+      "learning_rate": 1.7404426511272664e-06,
+      "loss": 0.3768,
       "step": 122700
     },
     {
-      "epoch": 4.32,
-      "learning_rate": 2.416542463214394e-06,
-      "loss": 0.2676,
+      "epoch": 4.422279886113814,
+      "grad_norm": 0.30197739601135254,
+      "learning_rate": 1.7393730487431536e-06,
+      "loss": 0.3838,
       "step": 122705
     },
     {
-      "epoch": 4.32,
-      "learning_rate": 2.415320711226321e-06,
-      "loss": 0.251,
+      "epoch": 4.422460085775039,
+      "grad_norm": 0.28687089681625366,
+      "learning_rate": 1.7383037632805826e-06,
+      "loss": 0.3882,
       "step": 122710
     },
     {
-      "epoch": 4.32,
-      "learning_rate": 2.414099252485427e-06,
-      "loss": 0.2726,
+      "epoch": 4.422640285436263,
+      "grad_norm": 0.21703818440437317,
+      "learning_rate": 1.7372347947541194e-06,
+      "loss": 0.3486,
       "step": 122715
     },
     {
-      "epoch": 4.32,
-      "learning_rate": 2.4128780870075695e-06,
-      "loss": 0.2336,
+      "epoch": 4.422820485097488,
+      "grad_norm": 0.27013471722602844,
+      "learning_rate": 1.7361661431783304e-06,
+      "loss": 0.3774,
       "step": 122720
     },
     {
-      "epoch": 4.32,
-      "learning_rate": 2.411657214808602e-06,
-      "loss": 0.2599,
+      "epoch": 4.423000684758713,
+      "grad_norm": 0.25899776816368103,
+      "learning_rate": 1.7350978085677704e-06,
+      "loss": 0.3916,
       "step": 122725
     },
     {
-      "epoch": 4.32,
-      "learning_rate": 2.410436635904381e-06,
-      "loss": 0.2276,
+      "epoch": 4.423180884419938,
+      "grad_norm": 0.24092505872249603,
+      "learning_rate": 1.7340297909369973e-06,
+      "loss": 0.3897,
       "step": 122730
     },
     {
-      "epoch": 4.32,
-      "learning_rate": 2.4092163503107556e-06,
-      "loss": 0.2533,
+      "epoch": 4.4233610840811615,
+      "grad_norm": 0.25030139088630676,
+      "learning_rate": 1.7329620903005661e-06,
+      "loss": 0.3741,
       "step": 122735
     },
     {
-      "epoch": 4.32,
-      "learning_rate": 2.4079963580435656e-06,
-      "loss": 0.2604,
+      "epoch": 4.423541283742386,
+      "grad_norm": 0.218690425157547,
+      "learning_rate": 1.7318947066730263e-06,
+      "loss": 0.3597,
       "step": 122740
     },
     {
-      "epoch": 4.32,
-      "learning_rate": 2.406776659118651e-06,
-      "loss": 0.258,
+      "epoch": 4.423721483403611,
+      "grad_norm": 0.2734704613685608,
+      "learning_rate": 1.7308276400689077e-06,
+      "loss": 0.3735,
       "step": 122745
     },
     {
-      "epoch": 4.32,
-      "learning_rate": 2.4055572535518574e-06,
-      "loss": 0.2369,
+      "epoch": 4.423901683064836,
+      "grad_norm": 0.37903356552124023,
+      "learning_rate": 1.7297608905027602e-06,
+      "loss": 0.3653,
       "step": 122750
     },
     {
-      "epoch": 4.32,
-      "learning_rate": 2.4043381413590083e-06,
-      "loss": 0.2541,
+      "epoch": 4.42408188272606,
+      "grad_norm": 0.29031357169151306,
+      "learning_rate": 1.7286944579891134e-06,
+      "loss": 0.411,
       "step": 122755
     },
     {
-      "epoch": 4.32,
-      "learning_rate": 2.403119322555941e-06,
-      "loss": 0.2411,
+      "epoch": 4.424262082387285,
+      "grad_norm": 0.24027031660079956,
+      "learning_rate": 1.7276283425425088e-06,
+      "loss": 0.3777,
       "step": 122760
     },
     {
-      "epoch": 4.32,
-      "learning_rate": 2.401900797158471e-06,
-      "loss": 0.2334,
+      "epoch": 4.42444228204851,
+      "grad_norm": 0.26298728585243225,
+      "learning_rate": 1.726562544177454e-06,
+      "loss": 0.3635,
       "step": 122765
     },
     {
-      "epoch": 4.32,
-      "learning_rate": 2.4006825651824326e-06,
-      "loss": 0.2497,
+      "epoch": 4.424622481709735,
+      "grad_norm": 0.24697257578372955,
+      "learning_rate": 1.7254970629084765e-06,
+      "loss": 0.3843,
       "step": 122770
     },
     {
-      "epoch": 4.32,
-      "learning_rate": 2.399464626643638e-06,
-      "loss": 0.2425,
+      "epoch": 4.424802681370959,
+      "grad_norm": 0.26722240447998047,
+      "learning_rate": 1.7244318987501007e-06,
+      "loss": 0.3959,
       "step": 122775
     },
     {
-      "epoch": 4.32,
-      "learning_rate": 2.3982469815578972e-06,
-      "loss": 0.2526,
+      "epoch": 4.424982881032184,
+      "grad_norm": 0.23795725405216217,
+      "learning_rate": 1.7233670517168344e-06,
+      "loss": 0.3746,
       "step": 122780
     },
     {
-      "epoch": 4.32,
-      "learning_rate": 2.3970296299410277e-06,
-      "loss": 0.2486,
+      "epoch": 4.425163080693408,
+      "grad_norm": 0.31174471974372864,
+      "learning_rate": 1.7223025218231826e-06,
+      "loss": 0.3546,
       "step": 122785
     },
     {
-      "epoch": 4.32,
-      "learning_rate": 2.395812571808831e-06,
-      "loss": 0.2415,
+      "epoch": 4.425343280354633,
+      "grad_norm": 0.2947745621204376,
+      "learning_rate": 1.7212383090836558e-06,
+      "loss": 0.3637,
       "step": 122790
     },
     {
-      "epoch": 4.32,
-      "learning_rate": 2.3945958071771134e-06,
-      "loss": 0.2286,
+      "epoch": 4.425523480015857,
+      "grad_norm": 0.24382677674293518,
+      "learning_rate": 1.7201744135127428e-06,
+      "loss": 0.3861,
       "step": 122795
     },
     {
-      "epoch": 4.32,
-      "learning_rate": 2.393379336061666e-06,
-      "loss": 0.2465,
+      "epoch": 4.425703679677082,
+      "grad_norm": 0.2894967794418335,
+      "learning_rate": 1.7191108351249513e-06,
+      "loss": 0.3632,
       "step": 122800
     },
     {
-      "epoch": 4.32,
-      "learning_rate": 2.3921631584782942e-06,
-      "loss": 0.2524,
+      "epoch": 4.425883879338307,
+      "grad_norm": 0.23039253056049347,
+      "learning_rate": 1.7180475739347695e-06,
+      "loss": 0.3493,
       "step": 122805
     },
     {
-      "epoch": 4.32,
-      "learning_rate": 2.390947274442787e-06,
-      "loss": 0.2628,
+      "epoch": 4.426064078999532,
+      "grad_norm": 0.295876681804657,
+      "learning_rate": 1.7169846299566806e-06,
+      "loss": 0.3588,
       "step": 122810
     },
     {
-      "epoch": 4.32,
-      "learning_rate": 2.389731683970922e-06,
-      "loss": 0.2426,
+      "epoch": 4.426244278660756,
+      "grad_norm": 0.23762252926826477,
+      "learning_rate": 1.715922003205167e-06,
+      "loss": 0.3371,
       "step": 122815
     },
     {
-      "epoch": 4.32,
-      "learning_rate": 2.3885163870784907e-06,
-      "loss": 0.2562,
+      "epoch": 4.426424478321981,
+      "grad_norm": 0.22657039761543274,
+      "learning_rate": 1.7148596936947064e-06,
+      "loss": 0.3865,
       "step": 122820
     },
     {
-      "epoch": 4.32,
-      "learning_rate": 2.387301383781279e-06,
-      "loss": 0.2338,
+      "epoch": 4.426604677983206,
+      "grad_norm": 0.22004631161689758,
+      "learning_rate": 1.713797701439776e-06,
+      "loss": 0.3896,
       "step": 122825
     },
     {
-      "epoch": 4.32,
-      "learning_rate": 2.3860866740950562e-06,
-      "loss": 0.237,
+      "epoch": 4.42678487764443,
+      "grad_norm": 0.2557975947856903,
+      "learning_rate": 1.7127360264548475e-06,
+      "loss": 0.3903,
       "step": 122830
     },
     {
-      "epoch": 4.32,
-      "learning_rate": 2.384872258035592e-06,
-      "loss": 0.2568,
+      "epoch": 4.426965077305654,
+      "grad_norm": 0.2727845311164856,
+      "learning_rate": 1.711674668754376e-06,
+      "loss": 0.3312,
       "step": 122835
     },
     {
-      "epoch": 4.32,
-      "learning_rate": 2.383658135618655e-06,
-      "loss": 0.2528,
+      "epoch": 4.427145276966879,
+      "grad_norm": 0.24959102272987366,
+      "learning_rate": 1.7106136283528336e-06,
+      "loss": 0.3425,
       "step": 122840
     },
     {
-      "epoch": 4.32,
-      "learning_rate": 2.3824443068600182e-06,
-      "loss": 0.2568,
+      "epoch": 4.427325476628104,
+      "grad_norm": 0.25791552662849426,
+      "learning_rate": 1.709552905264669e-06,
+      "loss": 0.3923,
       "step": 122845
     },
     {
-      "epoch": 4.32,
-      "learning_rate": 2.3812307717754333e-06,
-      "loss": 0.2535,
+      "epoch": 4.427505676289329,
+      "grad_norm": 0.22956585884094238,
+      "learning_rate": 1.7084924995043355e-06,
+      "loss": 0.3631,
       "step": 122850
     },
     {
-      "epoch": 4.32,
-      "learning_rate": 2.3800175303806625e-06,
-      "loss": 0.2579,
+      "epoch": 4.427685875950553,
+      "grad_norm": 0.25182798504829407,
+      "learning_rate": 1.707432411086285e-06,
+      "loss": 0.3912,
       "step": 122855
     },
     {
-      "epoch": 4.32,
-      "learning_rate": 2.3788045826914545e-06,
-      "loss": 0.2386,
+      "epoch": 4.427866075611778,
+      "grad_norm": 0.2180362045764923,
+      "learning_rate": 1.7063726400249531e-06,
+      "loss": 0.3683,
       "step": 122860
     },
     {
-      "epoch": 4.32,
-      "learning_rate": 2.3775919287235633e-06,
-      "loss": 0.2235,
+      "epoch": 4.428046275273003,
+      "grad_norm": 0.24969041347503662,
+      "learning_rate": 1.7053131863347893e-06,
+      "loss": 0.3792,
       "step": 122865
     },
     {
-      "epoch": 4.32,
-      "learning_rate": 2.3763795684927286e-06,
-      "loss": 0.2408,
+      "epoch": 4.4282264749342275,
+      "grad_norm": 0.3251616060733795,
+      "learning_rate": 1.704254050030224e-06,
+      "loss": 0.3662,
       "step": 122870
     },
     {
-      "epoch": 4.32,
-      "learning_rate": 2.3751675020146996e-06,
-      "loss": 0.261,
+      "epoch": 4.428406674595452,
+      "grad_norm": 0.2780360281467438,
+      "learning_rate": 1.7031952311256845e-06,
+      "loss": 0.4135,
       "step": 122875
     },
     {
-      "epoch": 4.32,
-      "learning_rate": 2.373955729305205e-06,
-      "loss": 0.2728,
+      "epoch": 4.428586874256676,
+      "grad_norm": 0.23237118124961853,
+      "learning_rate": 1.7021367296356005e-06,
+      "loss": 0.3865,
       "step": 122880
     },
     {
-      "epoch": 4.32,
-      "learning_rate": 2.372744250379988e-06,
-      "loss": 0.2683,
+      "epoch": 4.428767073917901,
+      "grad_norm": 0.21650679409503937,
+      "learning_rate": 1.7010785455743943e-06,
+      "loss": 0.3883,
       "step": 122885
     },
     {
-      "epoch": 4.32,
-      "learning_rate": 2.3715330652547774e-06,
-      "loss": 0.2324,
+      "epoch": 4.4289472735791255,
+      "grad_norm": 0.2289179116487503,
+      "learning_rate": 1.7000206789564765e-06,
+      "loss": 0.3903,
       "step": 122890
     },
     {
-      "epoch": 4.32,
-      "learning_rate": 2.370322173945291e-06,
-      "loss": 0.2572,
+      "epoch": 4.42912747324035,
+      "grad_norm": 0.27513277530670166,
+      "learning_rate": 1.6989631297962717e-06,
+      "loss": 0.3564,
       "step": 122895
     },
     {
-      "epoch": 4.32,
-      "learning_rate": 2.369111576467262e-06,
-      "loss": 0.2803,
+      "epoch": 4.429307672901575,
+      "grad_norm": 0.2146390825510025,
+      "learning_rate": 1.6979058981081819e-06,
+      "loss": 0.3701,
       "step": 122900
     },
     {
-      "epoch": 4.32,
-      "learning_rate": 2.3679012728364063e-06,
-      "loss": 0.256,
+      "epoch": 4.4294878725628,
+      "grad_norm": 0.36598315834999084,
+      "learning_rate": 1.6968489839066125e-06,
+      "loss": 0.3704,
       "step": 122905
     },
     {
-      "epoch": 4.32,
-      "learning_rate": 2.366691263068438e-06,
-      "loss": 0.2598,
+      "epoch": 4.4296680722240245,
+      "grad_norm": 0.2601289749145508,
+      "learning_rate": 1.6957923872059634e-06,
+      "loss": 0.3677,
       "step": 122910
     },
     {
-      "epoch": 4.32,
-      "learning_rate": 2.3654815471790604e-06,
-      "loss": 0.2567,
+      "epoch": 4.429848271885249,
+      "grad_norm": 0.25093069672584534,
+      "learning_rate": 1.6947361080206282e-06,
+      "loss": 0.3744,
       "step": 122915
     },
     {
-      "epoch": 4.32,
-      "learning_rate": 2.3642721251839946e-06,
-      "loss": 0.2443,
+      "epoch": 4.430028471546473,
+      "grad_norm": 0.3232559859752655,
+      "learning_rate": 1.6936801463650066e-06,
+      "loss": 0.3804,
       "step": 122920
     },
     {
-      "epoch": 4.32,
-      "learning_rate": 2.3630629970989332e-06,
-      "loss": 0.2458,
+      "epoch": 4.430208671207698,
+      "grad_norm": 0.26259225606918335,
+      "learning_rate": 1.6926245022534842e-06,
+      "loss": 0.3743,
       "step": 122925
     },
     {
-      "epoch": 4.33,
-      "learning_rate": 2.3618541629395833e-06,
-      "loss": 0.2581,
+      "epoch": 4.4303888708689225,
+      "grad_norm": 0.25600606203079224,
+      "learning_rate": 1.6915691757004303e-06,
+      "loss": 0.3987,
       "step": 122930
     },
     {
-      "epoch": 4.33,
-      "learning_rate": 2.3606456227216357e-06,
-      "loss": 0.2582,
+      "epoch": 4.430569070530147,
+      "grad_norm": 0.2362159788608551,
+      "learning_rate": 1.690514166720239e-06,
+      "loss": 0.3972,
       "step": 122935
     },
     {
-      "epoch": 4.33,
-      "learning_rate": 2.359437376460788e-06,
-      "loss": 0.2519,
+      "epoch": 4.430749270191372,
+      "grad_norm": 0.21137534081935883,
+      "learning_rate": 1.6894594753272763e-06,
+      "loss": 0.3447,
       "step": 122940
     },
     {
-      "epoch": 4.33,
-      "learning_rate": 2.3582294241727284e-06,
-      "loss": 0.2416,
+      "epoch": 4.430929469852597,
+      "grad_norm": 0.23811127245426178,
+      "learning_rate": 1.6884051015359226e-06,
+      "loss": 0.3834,
       "step": 122945
     },
     {
-      "epoch": 4.33,
-      "learning_rate": 2.3570217658731352e-06,
-      "loss": 0.254,
+      "epoch": 4.4311096695138215,
+      "grad_norm": 0.22823196649551392,
+      "learning_rate": 1.687351045360533e-06,
+      "loss": 0.3802,
       "step": 122950
     },
     {
-      "epoch": 4.33,
-      "learning_rate": 2.3558144015776858e-06,
-      "loss": 0.2403,
+      "epoch": 4.431289869175046,
+      "grad_norm": 0.23941099643707275,
+      "learning_rate": 1.6862973068154653e-06,
+      "loss": 0.3415,
       "step": 122955
     },
     {
-      "epoch": 4.33,
-      "learning_rate": 2.354607331302072e-06,
-      "loss": 0.2813,
+      "epoch": 4.431470068836271,
+      "grad_norm": 0.26783496141433716,
+      "learning_rate": 1.6852438859150887e-06,
+      "loss": 0.3696,
       "step": 122960
     },
     {
-      "epoch": 4.33,
-      "learning_rate": 2.3534005550619547e-06,
-      "loss": 0.2667,
+      "epoch": 4.431650268497496,
+      "grad_norm": 0.31337860226631165,
+      "learning_rate": 1.6841907826737503e-06,
+      "loss": 0.3926,
       "step": 122965
     },
     {
-      "epoch": 4.33,
-      "learning_rate": 2.3521940728730095e-06,
-      "loss": 0.2722,
+      "epoch": 4.4318304681587195,
+      "grad_norm": 0.2646377384662628,
+      "learning_rate": 1.6831379971057992e-06,
+      "loss": 0.3866,
       "step": 122970
     },
     {
-      "epoch": 4.33,
-      "learning_rate": 2.350987884750891e-06,
-      "loss": 0.263,
+      "epoch": 4.432010667819944,
+      "grad_norm": 0.27343621850013733,
+      "learning_rate": 1.6820855292255772e-06,
+      "loss": 0.3935,
       "step": 122975
     },
     {
-      "epoch": 4.33,
-      "learning_rate": 2.3497819907112722e-06,
-      "loss": 0.2437,
+      "epoch": 4.432190867481169,
+      "grad_norm": 0.3007572591304779,
+      "learning_rate": 1.6810333790474226e-06,
+      "loss": 0.3962,
       "step": 122980
     },
     {
-      "epoch": 4.33,
-      "learning_rate": 2.3485763907698076e-06,
-      "loss": 0.2371,
+      "epoch": 4.432371067142394,
+      "grad_norm": 0.24796468019485474,
+      "learning_rate": 1.6799815465856767e-06,
+      "loss": 0.365,
       "step": 122985
     },
     {
-      "epoch": 4.33,
-      "learning_rate": 2.3473710849421542e-06,
-      "loss": 0.2622,
+      "epoch": 4.432551266803618,
+      "grad_norm": 0.2737863063812256,
+      "learning_rate": 1.6789300318546697e-06,
+      "loss": 0.4177,
       "step": 122990
     },
     {
-      "epoch": 4.33,
-      "learning_rate": 2.3461660732439524e-06,
-      "loss": 0.2443,
+      "epoch": 4.432731466464843,
+      "grad_norm": 0.298483282327652,
+      "learning_rate": 1.6778788348687235e-06,
+      "loss": 0.3822,
       "step": 122995
     },
     {
-      "epoch": 4.33,
-      "learning_rate": 2.3449613556908613e-06,
-      "loss": 0.2414,
+      "epoch": 4.432911666126068,
+      "grad_norm": 0.23559249937534332,
+      "learning_rate": 1.6768279556421628e-06,
+      "loss": 0.3506,
       "step": 123000
     },
     {
-      "epoch": 4.33,
-      "eval_loss": 0.24904577434062958,
-      "eval_runtime": 10.5244,
-      "eval_samples_per_second": 9.502,
-      "eval_steps_per_second": 9.502,
+      "epoch": 4.432911666126068,
+      "eval_loss": 0.42897456884384155,
+      "eval_runtime": 3.5635,
+      "eval_samples_per_second": 28.062,
+      "eval_steps_per_second": 7.016,
       "step": 123000
     },
     {
-      "epoch": 4.33,
-      "learning_rate": 2.343756932298513e-06,
-      "loss": 0.2657,
+      "epoch": 4.433091865787293,
+      "grad_norm": 0.2772116959095001,
+      "learning_rate": 1.6757773941893068e-06,
+      "loss": 0.4011,
       "step": 123005
     },
     {
-      "epoch": 4.33,
-      "learning_rate": 2.3425528030825538e-06,
-      "loss": 0.2416,
+      "epoch": 4.4332720654485165,
+      "grad_norm": 0.2690448760986328,
+      "learning_rate": 1.6747271505244634e-06,
+      "loss": 0.3637,
       "step": 123010
     },
     {
-      "epoch": 4.33,
-      "learning_rate": 2.341348968058607e-06,
-      "loss": 0.2707,
+      "epoch": 4.433452265109741,
+      "grad_norm": 0.23633429408073425,
+      "learning_rate": 1.6736772246619543e-06,
+      "loss": 0.3459,
       "step": 123015
     },
     {
-      "epoch": 4.33,
-      "learning_rate": 2.3401454272423183e-06,
-      "loss": 0.2352,
+      "epoch": 4.433632464770966,
+      "grad_norm": 0.2238181084394455,
+      "learning_rate": 1.6726276166160681e-06,
+      "loss": 0.3827,
       "step": 123020
     },
     {
-      "epoch": 4.33,
-      "learning_rate": 2.338942180649306e-06,
-      "loss": 0.2549,
+      "epoch": 4.433812664432191,
+      "grad_norm": 0.24402472376823425,
+      "learning_rate": 1.6715783264011188e-06,
+      "loss": 0.388,
       "step": 123025
     },
     {
-      "epoch": 4.33,
-      "learning_rate": 2.337739228295191e-06,
-      "loss": 0.276,
+      "epoch": 4.433992864093415,
+      "grad_norm": 0.2225305289030075,
+      "learning_rate": 1.6705293540313998e-06,
+      "loss": 0.3813,
       "step": 123030
     },
     {
-      "epoch": 4.33,
-      "learning_rate": 2.3365365701955993e-06,
-      "loss": 0.2473,
+      "epoch": 4.43417306375464,
+      "grad_norm": 0.2935526967048645,
+      "learning_rate": 1.6694806995211976e-06,
+      "loss": 0.3678,
       "step": 123035
     },
     {
-      "epoch": 4.33,
-      "learning_rate": 2.3353342063661466e-06,
-      "loss": 0.2493,
+      "epoch": 4.434353263415865,
+      "grad_norm": 0.24194492399692535,
+      "learning_rate": 1.6684323628848113e-06,
+      "loss": 0.3804,
       "step": 123040
     },
     {
-      "epoch": 4.33,
-      "learning_rate": 2.3341321368224422e-06,
-      "loss": 0.2441,
+      "epoch": 4.43453346307709,
+      "grad_norm": 0.22049802541732788,
+      "learning_rate": 1.6673843441365077e-06,
+      "loss": 0.3625,
       "step": 123045
     },
     {
-      "epoch": 4.33,
-      "learning_rate": 2.3329303615800906e-06,
-      "loss": 0.2545,
+      "epoch": 4.434713662738314,
+      "grad_norm": 0.28984788060188293,
+      "learning_rate": 1.6663366432905809e-06,
+      "loss": 0.3471,
       "step": 123050
     },
     {
-      "epoch": 4.33,
-      "learning_rate": 2.331728880654707e-06,
-      "loss": 0.2505,
+      "epoch": 4.434893862399539,
+      "grad_norm": 0.2572941780090332,
+      "learning_rate": 1.6652892603612968e-06,
+      "loss": 0.378,
       "step": 123055
     },
     {
-      "epoch": 4.33,
-      "learning_rate": 2.330527694061882e-06,
-      "loss": 0.2575,
+      "epoch": 4.435074062060763,
+      "grad_norm": 0.2342897206544876,
+      "learning_rate": 1.6642421953629305e-06,
+      "loss": 0.3621,
       "step": 123060
     },
     {
-      "epoch": 4.33,
-      "learning_rate": 2.3293268018172165e-06,
-      "loss": 0.2356,
+      "epoch": 4.435254261721988,
+      "grad_norm": 0.2670467793941498,
+      "learning_rate": 1.6631954483097483e-06,
+      "loss": 0.3689,
       "step": 123065
     },
     {
-      "epoch": 4.33,
-      "learning_rate": 2.3281262039362984e-06,
-      "loss": 0.2797,
+      "epoch": 4.435434461383212,
+      "grad_norm": 0.29698285460472107,
+      "learning_rate": 1.6621490192160082e-06,
+      "loss": 0.3722,
       "step": 123070
     },
     {
-      "epoch": 4.33,
-      "learning_rate": 2.326925900434723e-06,
-      "loss": 0.254,
+      "epoch": 4.435614661044437,
+      "grad_norm": 0.21422305703163147,
+      "learning_rate": 1.6611029080959655e-06,
+      "loss": 0.356,
       "step": 123075
     },
     {
-      "epoch": 4.33,
-      "learning_rate": 2.3257258913280733e-06,
-      "loss": 0.2655,
+      "epoch": 4.435794860705662,
+      "grad_norm": 0.2958966791629791,
+      "learning_rate": 1.660057114963881e-06,
+      "loss": 0.3852,
       "step": 123080
     },
     {
-      "epoch": 4.33,
-      "learning_rate": 2.3245261766319247e-06,
-      "loss": 0.2621,
+      "epoch": 4.435975060366887,
+      "grad_norm": 0.20923343300819397,
+      "learning_rate": 1.659011639833999e-06,
+      "loss": 0.3784,
       "step": 123085
     },
     {
-      "epoch": 4.33,
-      "learning_rate": 2.32332675636186e-06,
-      "loss": 0.2525,
+      "epoch": 4.436155260028111,
+      "grad_norm": 0.2286258041858673,
+      "learning_rate": 1.657966482720566e-06,
+      "loss": 0.3705,
       "step": 123090
     },
     {
-      "epoch": 4.33,
-      "learning_rate": 2.3221276305334603e-06,
-      "loss": 0.2358,
+      "epoch": 4.436335459689336,
+      "grad_norm": 0.25596699118614197,
+      "learning_rate": 1.656921643637821e-06,
+      "loss": 0.3921,
       "step": 123095
     },
     {
-      "epoch": 4.33,
-      "learning_rate": 2.3209287991622865e-06,
-      "loss": 0.2428,
+      "epoch": 4.436515659350561,
+      "grad_norm": 0.2769662141799927,
+      "learning_rate": 1.6558771225999942e-06,
+      "loss": 0.3694,
       "step": 123100
     },
     {
-      "epoch": 4.33,
-      "learning_rate": 2.3197302622639027e-06,
-      "loss": 0.2508,
+      "epoch": 4.436695859011785,
+      "grad_norm": 0.2575506865978241,
+      "learning_rate": 1.6548329196213242e-06,
+      "loss": 0.3762,
       "step": 123105
     },
     {
-      "epoch": 4.33,
-      "learning_rate": 2.318532019853875e-06,
-      "loss": 0.2375,
+      "epoch": 4.436876058673009,
+      "grad_norm": 0.2907906174659729,
+      "learning_rate": 1.6537890347160412e-06,
+      "loss": 0.3744,
       "step": 123110
     },
     {
-      "epoch": 4.33,
-      "learning_rate": 2.3173340719477632e-06,
-      "loss": 0.2348,
+      "epoch": 4.437056258334234,
+      "grad_norm": 0.2527605891227722,
+      "learning_rate": 1.652745467898356e-06,
+      "loss": 0.3641,
       "step": 123115
     },
     {
-      "epoch": 4.33,
-      "learning_rate": 2.316136418561121e-06,
-      "loss": 0.242,
+      "epoch": 4.437236457995459,
+      "grad_norm": 0.31895288825035095,
+      "learning_rate": 1.6517022191824965e-06,
+      "loss": 0.3943,
       "step": 123120
     },
     {
-      "epoch": 4.33,
-      "learning_rate": 2.3149390597094976e-06,
-      "loss": 0.2382,
+      "epoch": 4.437416657656684,
+      "grad_norm": 0.23418550193309784,
+      "learning_rate": 1.6506592885826704e-06,
+      "loss": 0.3661,
       "step": 123125
     },
     {
-      "epoch": 4.33,
-      "learning_rate": 2.313741995408436e-06,
-      "loss": 0.2548,
+      "epoch": 4.437596857317908,
+      "grad_norm": 0.2372836172580719,
+      "learning_rate": 1.649616676113097e-06,
+      "loss": 0.3527,
       "step": 123130
     },
     {
-      "epoch": 4.33,
-      "learning_rate": 2.3125452256734877e-06,
-      "loss": 0.2716,
+      "epoch": 4.437777056979133,
+      "grad_norm": 0.33892858028411865,
+      "learning_rate": 1.6485743817879735e-06,
+      "loss": 0.3835,
       "step": 123135
     },
     {
-      "epoch": 4.33,
-      "learning_rate": 2.3113487505201843e-06,
-      "loss": 0.2397,
+      "epoch": 4.437957256640358,
+      "grad_norm": 0.32817450165748596,
+      "learning_rate": 1.6475324056214992e-06,
+      "loss": 0.3921,
       "step": 123140
     },
     {
-      "epoch": 4.33,
-      "learning_rate": 2.3101525699640693e-06,
-      "loss": 0.263,
+      "epoch": 4.4381374563015825,
+      "grad_norm": 0.30701467394828796,
+      "learning_rate": 1.6466990537720912e-06,
+      "loss": 0.3915,
       "step": 123145
     },
     {
-      "epoch": 4.33,
-      "learning_rate": 2.308956684020666e-06,
-      "loss": 0.2503,
+      "epoch": 4.438317655962807,
+      "grad_norm": 0.21967414021492004,
+      "learning_rate": 1.6456576503269705e-06,
+      "loss": 0.361,
       "step": 123150
     },
     {
-      "epoch": 4.33,
-      "learning_rate": 2.30776109270551e-06,
-      "loss": 0.2225,
+      "epoch": 4.438497855624031,
+      "grad_norm": 0.23973840475082397,
+      "learning_rate": 1.6446165650802398e-06,
+      "loss": 0.3769,
       "step": 123155
     },
     {
-      "epoch": 4.33,
-      "learning_rate": 2.306565796034124e-06,
-      "loss": 0.257,
+      "epoch": 4.438678055285256,
+      "grad_norm": 0.26962411403656006,
+      "learning_rate": 1.6435757980460876e-06,
+      "loss": 0.35,
       "step": 123160
     },
     {
-      "epoch": 4.33,
-      "learning_rate": 2.3053707940220188e-06,
-      "loss": 0.2497,
+      "epoch": 4.438858254946481,
+      "grad_norm": 0.21383236348628998,
+      "learning_rate": 1.642535349238694e-06,
+      "loss": 0.3269,
       "step": 123165
     },
     {
-      "epoch": 4.33,
-      "learning_rate": 2.30417608668472e-06,
-      "loss": 0.2441,
+      "epoch": 4.439038454607705,
+      "grad_norm": 0.23623572289943695,
+      "learning_rate": 1.6414952186722287e-06,
+      "loss": 0.3909,
       "step": 123170
     },
     {
-      "epoch": 4.33,
-      "learning_rate": 2.302981674037741e-06,
-      "loss": 0.2544,
+      "epoch": 4.43921865426893,
+      "grad_norm": 0.24700896441936493,
+      "learning_rate": 1.6404554063608746e-06,
+      "loss": 0.3601,
       "step": 123175
     },
     {
-      "epoch": 4.33,
-      "learning_rate": 2.301787556096585e-06,
-      "loss": 0.2273,
+      "epoch": 4.439398853930155,
+      "grad_norm": 0.24526774883270264,
+      "learning_rate": 1.6394159123187953e-06,
+      "loss": 0.3701,
       "step": 123180
     },
     {
-      "epoch": 4.33,
-      "learning_rate": 2.300593732876757e-06,
-      "loss": 0.2501,
+      "epoch": 4.4395790535913795,
+      "grad_norm": 0.20414763689041138,
+      "learning_rate": 1.6383767365601433e-06,
+      "loss": 0.3538,
       "step": 123185
     },
     {
-      "epoch": 4.33,
-      "learning_rate": 2.2994002043937617e-06,
-      "loss": 0.2558,
+      "epoch": 4.439759253252604,
+      "grad_norm": 0.19671322405338287,
+      "learning_rate": 1.6373378790990906e-06,
+      "loss": 0.3585,
       "step": 123190
     },
     {
-      "epoch": 4.33,
-      "learning_rate": 2.2982069706630916e-06,
-      "loss": 0.2451,
+      "epoch": 4.439939452913828,
+      "grad_norm": 0.2665334939956665,
+      "learning_rate": 1.6362993399497816e-06,
+      "loss": 0.3915,
       "step": 123195
     },
     {
-      "epoch": 4.33,
-      "learning_rate": 2.2970140317002455e-06,
-      "loss": 0.2257,
+      "epoch": 4.440119652575053,
+      "grad_norm": 0.21728205680847168,
+      "learning_rate": 1.6352611191263745e-06,
+      "loss": 0.3661,
       "step": 123200
     },
     {
-      "epoch": 4.33,
-      "learning_rate": 2.295821387520708e-06,
-      "loss": 0.2497,
+      "epoch": 4.4402998522362775,
+      "grad_norm": 0.23190873861312866,
+      "learning_rate": 1.634223216643016e-06,
+      "loss": 0.3862,
       "step": 123205
     },
     {
-      "epoch": 4.33,
-      "learning_rate": 2.2946290381399698e-06,
-      "loss": 0.2344,
+      "epoch": 4.440480051897502,
+      "grad_norm": 0.28229013085365295,
+      "learning_rate": 1.633185632513834e-06,
+      "loss": 0.364,
       "step": 123210
     },
     {
-      "epoch": 4.34,
-      "learning_rate": 2.2934369835735132e-06,
-      "loss": 0.2731,
+      "epoch": 4.440660251558727,
+      "grad_norm": 0.2951110303401947,
+      "learning_rate": 1.632148366752978e-06,
+      "loss": 0.4201,
       "step": 123215
     },
     {
-      "epoch": 4.34,
-      "learning_rate": 2.2922452238368115e-06,
-      "loss": 0.2455,
+      "epoch": 4.440840451219952,
+      "grad_norm": 0.2822306156158447,
+      "learning_rate": 1.6311114193745757e-06,
+      "loss": 0.3883,
       "step": 123220
     },
     {
-      "epoch": 4.34,
-      "learning_rate": 2.2910537589453355e-06,
-      "loss": 0.2615,
+      "epoch": 4.4410206508811765,
+      "grad_norm": 0.280408650636673,
+      "learning_rate": 1.6300747903927576e-06,
+      "loss": 0.4214,
       "step": 123225
     },
     {
-      "epoch": 4.34,
-      "learning_rate": 2.2898625889145653e-06,
-      "loss": 0.2281,
+      "epoch": 4.441200850542401,
+      "grad_norm": 0.23964573442935944,
+      "learning_rate": 1.6290384798216428e-06,
+      "loss": 0.3636,
       "step": 123230
     },
     {
-      "epoch": 4.34,
-      "learning_rate": 2.2886717137599626e-06,
-      "loss": 0.2764,
+      "epoch": 4.441381050203626,
+      "grad_norm": 0.2101919949054718,
+      "learning_rate": 1.6280024876753535e-06,
+      "loss": 0.3383,
       "step": 123235
     },
     {
-      "epoch": 4.34,
-      "learning_rate": 2.287481133496991e-06,
-      "loss": 0.2407,
+      "epoch": 4.441561249864851,
+      "grad_norm": 0.24973724782466888,
+      "learning_rate": 1.626966813968006e-06,
+      "loss": 0.3767,
       "step": 123240
     },
     {
-      "epoch": 4.34,
-      "learning_rate": 2.2862908481411037e-06,
-      "loss": 0.2603,
+      "epoch": 4.4417414495260745,
+      "grad_norm": 0.23754936456680298,
+      "learning_rate": 1.6259314587137114e-06,
+      "loss": 0.3969,
       "step": 123245
     },
     {
-      "epoch": 4.34,
-      "learning_rate": 2.285100857707764e-06,
-      "loss": 0.2479,
+      "epoch": 4.441921649187299,
+      "grad_norm": 0.28546053171157837,
+      "learning_rate": 1.6248964219265777e-06,
+      "loss": 0.3638,
       "step": 123250
     },
     {
-      "epoch": 4.34,
-      "learning_rate": 2.2839111622124184e-06,
-      "loss": 0.2535,
+      "epoch": 4.442101848848524,
+      "grad_norm": 0.2416713684797287,
+      "learning_rate": 1.623861703620702e-06,
+      "loss": 0.3669,
       "step": 123255
     },
     {
-      "epoch": 4.34,
-      "learning_rate": 2.2827217616705145e-06,
-      "loss": 0.2414,
+      "epoch": 4.442282048509749,
+      "grad_norm": 0.26729458570480347,
+      "learning_rate": 1.6228273038101815e-06,
+      "loss": 0.3851,
       "step": 123260
     },
     {
-      "epoch": 4.34,
-      "learning_rate": 2.2815326560975015e-06,
-      "loss": 0.2627,
+      "epoch": 4.4424622481709735,
+      "grad_norm": 0.24362029135227203,
+      "learning_rate": 1.6217932225091187e-06,
+      "loss": 0.405,
       "step": 123265
     },
     {
-      "epoch": 4.34,
-      "learning_rate": 2.280343845508817e-06,
-      "loss": 0.2377,
+      "epoch": 4.442642447832198,
+      "grad_norm": 0.24386849999427795,
+      "learning_rate": 1.620759459731594e-06,
+      "loss": 0.4102,
       "step": 123270
     },
     {
-      "epoch": 4.34,
-      "learning_rate": 2.2791553299198936e-06,
-      "loss": 0.2502,
+      "epoch": 4.442822647493423,
+      "grad_norm": 0.27166756987571716,
+      "learning_rate": 1.619726015491696e-06,
+      "loss": 0.3791,
       "step": 123275
     },
     {
-      "epoch": 4.34,
-      "learning_rate": 2.2779671093461574e-06,
-      "loss": 0.2412,
+      "epoch": 4.443002847154648,
+      "grad_norm": 0.2460469901561737,
+      "learning_rate": 1.6186928898035082e-06,
+      "loss": 0.3231,
       "step": 123280
     },
     {
-      "epoch": 4.34,
-      "learning_rate": 2.2767791838030515e-06,
-      "loss": 0.2599,
+      "epoch": 4.443183046815872,
+      "grad_norm": 0.21668177843093872,
+      "learning_rate": 1.6176600826810994e-06,
+      "loss": 0.3385,
       "step": 123285
     },
     {
-      "epoch": 4.34,
-      "learning_rate": 2.275591553305992e-06,
-      "loss": 0.2541,
+      "epoch": 4.443363246477096,
+      "grad_norm": 0.27097898721694946,
+      "learning_rate": 1.6166275941385422e-06,
+      "loss": 0.3826,
       "step": 123290
     },
     {
-      "epoch": 4.34,
-      "learning_rate": 2.274404217870402e-06,
-      "loss": 0.2362,
+      "epoch": 4.443543446138321,
+      "grad_norm": 0.2606200575828552,
+      "learning_rate": 1.615595424189914e-06,
+      "loss": 0.3223,
       "step": 123295
     },
     {
-      "epoch": 4.34,
-      "learning_rate": 2.2732171775116884e-06,
-      "loss": 0.2717,
+      "epoch": 4.443723645799546,
+      "grad_norm": 0.26379239559173584,
+      "learning_rate": 1.6145635728492619e-06,
+      "loss": 0.3753,
       "step": 123300
     },
     {
-      "epoch": 4.34,
-      "learning_rate": 2.2720304322452786e-06,
-      "loss": 0.2824,
+      "epoch": 4.44390384546077,
+      "grad_norm": 0.2528885006904602,
+      "learning_rate": 1.6135320401306608e-06,
+      "loss": 0.3818,
       "step": 123305
     },
     {
-      "epoch": 4.34,
-      "learning_rate": 2.2708439820865706e-06,
-      "loss": 0.2555,
+      "epoch": 4.444084045121995,
+      "grad_norm": 0.22059816122055054,
+      "learning_rate": 1.612500826048155e-06,
+      "loss": 0.3668,
       "step": 123310
     },
     {
-      "epoch": 4.34,
-      "learning_rate": 2.2696578270509804e-06,
-      "loss": 0.261,
+      "epoch": 4.44426424478322,
+      "grad_norm": 0.285345196723938,
+      "learning_rate": 1.6114699306157944e-06,
+      "loss": 0.356,
       "step": 123315
     },
     {
-      "epoch": 4.34,
-      "learning_rate": 2.2684719671538983e-06,
-      "loss": 0.2478,
+      "epoch": 4.444444444444445,
+      "grad_norm": 0.29682835936546326,
+      "learning_rate": 1.610439353847637e-06,
+      "loss": 0.3959,
       "step": 123320
     },
     {
-      "epoch": 4.34,
-      "learning_rate": 2.2672864024107287e-06,
-      "loss": 0.2405,
+      "epoch": 4.444624644105669,
+      "grad_norm": 0.2404378205537796,
+      "learning_rate": 1.6094090957577079e-06,
+      "loss": 0.372,
       "step": 123325
     },
     {
-      "epoch": 4.34,
-      "learning_rate": 2.2661011328368674e-06,
-      "loss": 0.2556,
+      "epoch": 4.444804843766894,
+      "grad_norm": 0.31720393896102905,
+      "learning_rate": 1.6083791563600514e-06,
+      "loss": 0.3816,
       "step": 123330
     },
     {
-      "epoch": 4.34,
-      "learning_rate": 2.2649161584476998e-06,
-      "loss": 0.2402,
+      "epoch": 4.444985043428118,
+      "grad_norm": 0.19714432954788208,
+      "learning_rate": 1.6073495356687008e-06,
+      "loss": 0.3454,
       "step": 123335
     },
     {
-      "epoch": 4.34,
-      "learning_rate": 2.263731479258607e-06,
-      "loss": 0.2424,
+      "epoch": 4.445165243089343,
+      "grad_norm": 0.2751675546169281,
+      "learning_rate": 1.6063202336976862e-06,
+      "loss": 0.4002,
       "step": 123340
     },
     {
-      "epoch": 4.34,
-      "learning_rate": 2.2625470952849836e-06,
-      "loss": 0.258,
+      "epoch": 4.445345442750567,
+      "grad_norm": 0.2857431471347809,
+      "learning_rate": 1.6052912504610273e-06,
+      "loss": 0.4002,
       "step": 123345
     },
     {
-      "epoch": 4.34,
-      "learning_rate": 2.2613630065421997e-06,
-      "loss": 0.2495,
+      "epoch": 4.445525642411792,
+      "grad_norm": 0.2135014832019806,
+      "learning_rate": 1.6042625859727461e-06,
+      "loss": 0.3727,
       "step": 123350
     },
     {
-      "epoch": 4.34,
-      "learning_rate": 2.2601792130456295e-06,
-      "loss": 0.2442,
+      "epoch": 4.445705842073017,
+      "grad_norm": 0.2780386507511139,
+      "learning_rate": 1.6032342402468536e-06,
+      "loss": 0.3768,
       "step": 123355
     },
     {
-      "epoch": 4.34,
-      "learning_rate": 2.2589957148106493e-06,
-      "loss": 0.2352,
+      "epoch": 4.445886041734242,
+      "grad_norm": 0.24021285772323608,
+      "learning_rate": 1.6022062132973692e-06,
+      "loss": 0.4057,
       "step": 123360
     },
     {
-      "epoch": 4.34,
-      "learning_rate": 2.2578125118526166e-06,
-      "loss": 0.2491,
+      "epoch": 4.446066241395466,
+      "grad_norm": 0.2460976094007492,
+      "learning_rate": 1.6011785051382956e-06,
+      "loss": 0.3774,
       "step": 123365
     },
     {
-      "epoch": 4.34,
-      "learning_rate": 2.256629604186908e-06,
-      "loss": 0.2595,
+      "epoch": 4.446246441056691,
+      "grad_norm": 0.25038421154022217,
+      "learning_rate": 1.6001511157836352e-06,
+      "loss": 0.3843,
       "step": 123370
     },
     {
-      "epoch": 4.34,
-      "learning_rate": 2.255446991828869e-06,
-      "loss": 0.236,
+      "epoch": 4.446426640717916,
+      "grad_norm": 0.3057589828968048,
+      "learning_rate": 1.5991240452473854e-06,
+      "loss": 0.3501,
       "step": 123375
     },
     {
-      "epoch": 4.34,
-      "learning_rate": 2.2542646747938682e-06,
-      "loss": 0.2478,
+      "epoch": 4.44660684037914,
+      "grad_norm": 0.2682446241378784,
+      "learning_rate": 1.5980972935435351e-06,
+      "loss": 0.3708,
       "step": 123380
     },
     {
-      "epoch": 4.34,
-      "learning_rate": 2.2530826530972493e-06,
-      "loss": 0.244,
+      "epoch": 4.446787040040364,
+      "grad_norm": 0.2691839337348938,
+      "learning_rate": 1.5970708606860813e-06,
+      "loss": 0.3683,
       "step": 123385
     },
     {
-      "epoch": 4.34,
-      "learning_rate": 2.2519009267543605e-06,
-      "loss": 0.2586,
+      "epoch": 4.446967239701589,
+      "grad_norm": 0.26984599232673645,
+      "learning_rate": 1.5960447466890128e-06,
+      "loss": 0.4096,
       "step": 123390
     },
     {
-      "epoch": 4.34,
-      "learning_rate": 2.2507194957805455e-06,
-      "loss": 0.2511,
+      "epoch": 4.447147439362814,
+      "grad_norm": 0.2700972557067871,
+      "learning_rate": 1.5950189515662934e-06,
+      "loss": 0.3566,
       "step": 123395
     },
     {
-      "epoch": 4.34,
-      "learning_rate": 2.2495383601911474e-06,
-      "loss": 0.2539,
+      "epoch": 4.447327639024039,
+      "grad_norm": 0.2507593333721161,
+      "learning_rate": 1.5939934753319146e-06,
+      "loss": 0.4151,
       "step": 123400
     },
     {
-      "epoch": 4.34,
-      "learning_rate": 2.248357520001501e-06,
-      "loss": 0.2443,
+      "epoch": 4.447507838685263,
+      "grad_norm": 0.2393806129693985,
+      "learning_rate": 1.5929683179998434e-06,
+      "loss": 0.367,
       "step": 123405
     },
     {
-      "epoch": 4.34,
-      "learning_rate": 2.247176975226939e-06,
-      "loss": 0.2526,
+      "epoch": 4.447688038346488,
+      "grad_norm": 0.2460200935602188,
+      "learning_rate": 1.591943479584046e-06,
+      "loss": 0.4182,
       "step": 123410
     },
     {
-      "epoch": 4.34,
-      "learning_rate": 2.245996725882782e-06,
-      "loss": 0.2532,
+      "epoch": 4.447868238007713,
+      "grad_norm": 0.24296921491622925,
+      "learning_rate": 1.5909189600984865e-06,
+      "loss": 0.3385,
       "step": 123415
     },
     {
-      "epoch": 4.34,
-      "learning_rate": 2.244816771984365e-06,
-      "loss": 0.2536,
+      "epoch": 4.4480484376689375,
+      "grad_norm": 0.24117806553840637,
+      "learning_rate": 1.58989475955712e-06,
+      "loss": 0.3475,
       "step": 123420
     },
     {
-      "epoch": 4.34,
-      "learning_rate": 2.243637113547009e-06,
-      "loss": 0.2413,
+      "epoch": 4.448228637330162,
+      "grad_norm": 0.21672047674655914,
+      "learning_rate": 1.588870877973911e-06,
+      "loss": 0.3675,
       "step": 123425
     },
     {
-      "epoch": 4.34,
-      "learning_rate": 2.24245775058603e-06,
-      "loss": 0.2432,
+      "epoch": 4.448408836991386,
+      "grad_norm": 0.29379647970199585,
+      "learning_rate": 1.5878473153628004e-06,
+      "loss": 0.3689,
       "step": 123430
     },
     {
-      "epoch": 4.34,
-      "learning_rate": 2.241278683116732e-06,
-      "loss": 0.2602,
+      "epoch": 4.448589036652611,
+      "grad_norm": 0.2538691461086273,
+      "learning_rate": 1.5868240717377413e-06,
+      "loss": 0.387,
       "step": 123435
     },
     {
-      "epoch": 4.34,
-      "learning_rate": 2.2400999111544392e-06,
-      "loss": 0.2423,
+      "epoch": 4.448769236313836,
+      "grad_norm": 0.22580280900001526,
+      "learning_rate": 1.58580114711267e-06,
+      "loss": 0.3792,
       "step": 123440
     },
     {
-      "epoch": 4.34,
-      "learning_rate": 2.238921434714447e-06,
-      "loss": 0.2605,
+      "epoch": 4.44894943597506,
+      "grad_norm": 0.24915678799152374,
+      "learning_rate": 1.584778541501522e-06,
+      "loss": 0.3549,
       "step": 123445
     },
     {
-      "epoch": 4.34,
-      "learning_rate": 2.237743253812061e-06,
-      "loss": 0.2329,
+      "epoch": 4.449129635636285,
+      "grad_norm": 0.28229835629463196,
+      "learning_rate": 1.583756254918231e-06,
+      "loss": 0.3472,
       "step": 123450
     },
     {
-      "epoch": 4.34,
-      "learning_rate": 2.2365653684625732e-06,
-      "loss": 0.2502,
+      "epoch": 4.44930983529751,
+      "grad_norm": 0.2428547888994217,
+      "learning_rate": 1.5827342873767332e-06,
+      "loss": 0.3503,
       "step": 123455
     },
     {
-      "epoch": 4.34,
-      "learning_rate": 2.2353877786812887e-06,
-      "loss": 0.2332,
+      "epoch": 4.4494900349587345,
+      "grad_norm": 0.3034095764160156,
+      "learning_rate": 1.5817126388909448e-06,
+      "loss": 0.3411,
       "step": 123460
     },
     {
-      "epoch": 4.34,
-      "learning_rate": 2.23421048448349e-06,
-      "loss": 0.2649,
+      "epoch": 4.449670234619959,
+      "grad_norm": 0.3237626254558563,
+      "learning_rate": 1.5806913094747883e-06,
+      "loss": 0.3616,
       "step": 123465
     },
     {
-      "epoch": 4.34,
-      "learning_rate": 2.2330334858844615e-06,
-      "loss": 0.2413,
+      "epoch": 4.449850434281183,
+      "grad_norm": 0.26890459656715393,
+      "learning_rate": 1.5796702991421802e-06,
+      "loss": 0.407,
       "step": 123470
     },
     {
-      "epoch": 4.34,
-      "learning_rate": 2.2318567828994858e-06,
-      "loss": 0.2451,
+      "epoch": 4.450030633942408,
+      "grad_norm": 0.2624853849411011,
+      "learning_rate": 1.5786496079070262e-06,
+      "loss": 0.3613,
       "step": 123475
     },
     {
-      "epoch": 4.34,
-      "learning_rate": 2.2306803755438533e-06,
-      "loss": 0.2572,
+      "epoch": 4.450210833603633,
+      "grad_norm": 0.20167163014411926,
+      "learning_rate": 1.5776292357832374e-06,
+      "loss": 0.3715,
       "step": 123480
     },
     {
-      "epoch": 4.34,
-      "learning_rate": 2.2295042638328294e-06,
-      "loss": 0.2657,
+      "epoch": 4.450391033264857,
+      "grad_norm": 0.33518025279045105,
+      "learning_rate": 1.576609182784719e-06,
+      "loss": 0.3528,
       "step": 123485
     },
     {
-      "epoch": 4.34,
-      "learning_rate": 2.22832844778168e-06,
-      "loss": 0.2541,
+      "epoch": 4.450571232926082,
+      "grad_norm": 0.2814764082431793,
+      "learning_rate": 1.575589448925363e-06,
+      "loss": 0.3977,
       "step": 123490
     },
     {
-      "epoch": 4.34,
-      "learning_rate": 2.2271529274056873e-06,
-      "loss": 0.2277,
+      "epoch": 4.450751432587307,
+      "grad_norm": 0.2996358871459961,
+      "learning_rate": 1.5745700342190667e-06,
+      "loss": 0.3737,
       "step": 123495
     },
     {
-      "epoch": 4.35,
-      "learning_rate": 2.2259777027201027e-06,
-      "loss": 0.2265,
+      "epoch": 4.4509316322485315,
+      "grad_norm": 0.2564932107925415,
+      "learning_rate": 1.5735509386797132e-06,
+      "loss": 0.3465,
       "step": 123500
     },
     {
-      "epoch": 4.35,
-      "eval_loss": 0.24895407259464264,
-      "eval_runtime": 10.5376,
-      "eval_samples_per_second": 9.49,
-      "eval_steps_per_second": 9.49,
+      "epoch": 4.4509316322485315,
+      "eval_loss": 0.4289168119430542,
+      "eval_runtime": 3.5469,
+      "eval_samples_per_second": 28.194,
+      "eval_steps_per_second": 7.048,
       "step": 123500
     },
     {
-      "epoch": 4.35,
-      "learning_rate": 2.2248027737401896e-06,
-      "loss": 0.2476,
+      "epoch": 4.451111831909756,
+      "grad_norm": 0.24683259427547455,
+      "learning_rate": 1.5725321623212025e-06,
+      "loss": 0.3773,
       "step": 123505
     },
     {
-      "epoch": 4.35,
-      "learning_rate": 2.2236281404811988e-06,
-      "loss": 0.256,
+      "epoch": 4.451292031570981,
+      "grad_norm": 0.22746077179908752,
+      "learning_rate": 1.5715137051574013e-06,
+      "loss": 0.3871,
       "step": 123510
     },
     {
-      "epoch": 4.35,
-      "learning_rate": 2.222453802958391e-06,
-      "loss": 0.2702,
+      "epoch": 4.451472231232206,
+      "grad_norm": 0.304383784532547,
+      "learning_rate": 1.5704955672021848e-06,
+      "loss": 0.3756,
       "step": 123515
     },
     {
-      "epoch": 4.35,
-      "learning_rate": 2.221279761187009e-06,
-      "loss": 0.2348,
+      "epoch": 4.4516524308934295,
+      "grad_norm": 0.2046549767255783,
+      "learning_rate": 1.5694777484694334e-06,
+      "loss": 0.3864,
       "step": 123520
     },
     {
-      "epoch": 4.35,
-      "learning_rate": 2.220106015182294e-06,
-      "loss": 0.2549,
+      "epoch": 4.451832630554654,
+      "grad_norm": 0.2234576940536499,
+      "learning_rate": 1.5684602489730137e-06,
+      "loss": 0.3441,
       "step": 123525
     },
     {
-      "epoch": 4.35,
-      "learning_rate": 2.218932564959489e-06,
-      "loss": 0.2425,
+      "epoch": 4.452012830215879,
+      "grad_norm": 0.2881526052951813,
+      "learning_rate": 1.5674430687267843e-06,
+      "loss": 0.3734,
       "step": 123530
     },
     {
-      "epoch": 4.35,
-      "learning_rate": 2.2177594105338346e-06,
-      "loss": 0.2566,
+      "epoch": 4.452193029877104,
+      "grad_norm": 0.22432564198970795,
+      "learning_rate": 1.5664262077446091e-06,
+      "loss": 0.3581,
       "step": 123535
     },
     {
-      "epoch": 4.35,
-      "learning_rate": 2.216586551920563e-06,
-      "loss": 0.2586,
+      "epoch": 4.4523732295383285,
+      "grad_norm": 0.2805180549621582,
+      "learning_rate": 1.565409666040335e-06,
+      "loss": 0.3709,
       "step": 123540
     },
     {
-      "epoch": 4.35,
-      "learning_rate": 2.21564847802509e-06,
-      "loss": 0.2426,
+      "epoch": 4.452553429199553,
+      "grad_norm": 0.22590774297714233,
+      "learning_rate": 1.5643934436278207e-06,
+      "loss": 0.3828,
       "step": 123545
     },
     {
-      "epoch": 4.35,
-      "learning_rate": 2.2144761519124757e-06,
-      "loss": 0.2442,
+      "epoch": 4.452733628860778,
+      "grad_norm": 0.3044675588607788,
+      "learning_rate": 1.5633775405209079e-06,
+      "loss": 0.3755,
       "step": 123550
     },
     {
-      "epoch": 4.35,
-      "learning_rate": 2.2133041216548717e-06,
-      "loss": 0.2617,
+      "epoch": 4.452913828522003,
+      "grad_norm": 0.22538445889949799,
+      "learning_rate": 1.562361956733438e-06,
+      "loss": 0.3691,
       "step": 123555
     },
     {
-      "epoch": 4.35,
-      "learning_rate": 2.2121323872674932e-06,
-      "loss": 0.2533,
+      "epoch": 4.453094028183227,
+      "grad_norm": 0.21017488837242126,
+      "learning_rate": 1.5613466922792502e-06,
+      "loss": 0.3595,
       "step": 123560
     },
     {
-      "epoch": 4.35,
-      "learning_rate": 2.210960948765556e-06,
-      "loss": 0.2569,
+      "epoch": 4.453274227844451,
+      "grad_norm": 0.24644504487514496,
+      "learning_rate": 1.5603317471721723e-06,
+      "loss": 0.3807,
       "step": 123565
     },
     {
-      "epoch": 4.35,
-      "learning_rate": 2.209789806164278e-06,
-      "loss": 0.2602,
+      "epoch": 4.453454427505676,
+      "grad_norm": 0.289897084236145,
+      "learning_rate": 1.5593171214260377e-06,
+      "loss": 0.3825,
       "step": 123570
     },
     {
-      "epoch": 4.35,
-      "learning_rate": 2.208618959478859e-06,
-      "loss": 0.2582,
+      "epoch": 4.453634627166901,
+      "grad_norm": 0.26970839500427246,
+      "learning_rate": 1.558302815054677e-06,
+      "loss": 0.371,
       "step": 123575
     },
     {
-      "epoch": 4.35,
-      "learning_rate": 2.2074484087244996e-06,
-      "loss": 0.2369,
+      "epoch": 4.4538148268281255,
+      "grad_norm": 0.2419775426387787,
+      "learning_rate": 1.55728882807189e-06,
+      "loss": 0.3658,
       "step": 123580
     },
     {
-      "epoch": 4.35,
-      "learning_rate": 2.206278153916405e-06,
-      "loss": 0.2599,
+      "epoch": 4.45399502648935,
+      "grad_norm": 0.25539085268974304,
+      "learning_rate": 1.5562751604915105e-06,
+      "loss": 0.3781,
       "step": 123585
     },
     {
-      "epoch": 4.35,
-      "learning_rate": 2.205108195069769e-06,
-      "loss": 0.2393,
+      "epoch": 4.454175226150575,
+      "grad_norm": 0.2259790003299713,
+      "learning_rate": 1.5552618123273438e-06,
+      "loss": 0.3886,
       "step": 123590
     },
     {
-      "epoch": 4.35,
-      "learning_rate": 2.2039385321997784e-06,
-      "loss": 0.2642,
+      "epoch": 4.4543554258118,
+      "grad_norm": 0.2829951047897339,
+      "learning_rate": 1.554248783593193e-06,
+      "loss": 0.3758,
       "step": 123595
     },
     {
-      "epoch": 4.35,
-      "learning_rate": 2.2027691653216194e-06,
-      "loss": 0.2589,
+      "epoch": 4.454535625473024,
+      "grad_norm": 0.2783590257167816,
+      "learning_rate": 1.5532360743028662e-06,
+      "loss": 0.386,
       "step": 123600
     },
     {
-      "epoch": 4.35,
-      "learning_rate": 2.2016000944504842e-06,
-      "loss": 0.268,
+      "epoch": 4.454715825134249,
+      "grad_norm": 0.27192917466163635,
+      "learning_rate": 1.5522236844701527e-06,
+      "loss": 0.3568,
       "step": 123605
     },
     {
-      "epoch": 4.35,
-      "learning_rate": 2.2004313196015454e-06,
-      "loss": 0.2664,
+      "epoch": 4.454896024795473,
+      "grad_norm": 0.21945591270923615,
+      "learning_rate": 1.551211614108855e-06,
+      "loss": 0.3553,
       "step": 123610
     },
     {
-      "epoch": 4.35,
-      "learning_rate": 2.1992628407899757e-06,
-      "loss": 0.2459,
+      "epoch": 4.455076224456698,
+      "grad_norm": 0.2255585640668869,
+      "learning_rate": 1.5501998632327624e-06,
+      "loss": 0.3524,
       "step": 123615
     },
     {
-      "epoch": 4.35,
-      "learning_rate": 2.1980946580309524e-06,
-      "loss": 0.2592,
+      "epoch": 4.455256424117922,
+      "grad_norm": 0.22020921111106873,
+      "learning_rate": 1.5491884318556526e-06,
+      "loss": 0.3737,
       "step": 123620
     },
     {
-      "epoch": 4.35,
-      "learning_rate": 2.1969267713396495e-06,
-      "loss": 0.256,
+      "epoch": 4.455436623779147,
+      "grad_norm": 0.3251590132713318,
+      "learning_rate": 1.548177319991312e-06,
+      "loss": 0.3473,
       "step": 123625
     },
     {
-      "epoch": 4.35,
-      "learning_rate": 2.1957591807312215e-06,
-      "loss": 0.2237,
+      "epoch": 4.455616823440372,
+      "grad_norm": 0.3003102242946625,
+      "learning_rate": 1.5471665276535151e-06,
+      "loss": 0.3673,
       "step": 123630
     },
     {
-      "epoch": 4.35,
-      "learning_rate": 2.1945918862208343e-06,
-      "loss": 0.2536,
+      "epoch": 4.455797023101597,
+      "grad_norm": 0.22849665582180023,
+      "learning_rate": 1.5461560548560322e-06,
+      "loss": 0.3556,
       "step": 123635
     },
     {
-      "epoch": 4.35,
-      "learning_rate": 2.1934248878236364e-06,
-      "loss": 0.2541,
+      "epoch": 4.455977222762821,
+      "grad_norm": 0.23016008734703064,
+      "learning_rate": 1.5451459016126325e-06,
+      "loss": 0.4272,
       "step": 123640
     },
     {
-      "epoch": 4.35,
-      "learning_rate": 2.192258185554791e-06,
-      "loss": 0.2483,
+      "epoch": 4.456157422424046,
+      "grad_norm": 0.25868502259254456,
+      "learning_rate": 1.5441360679370798e-06,
+      "loss": 0.3825,
       "step": 123645
     },
     {
-      "epoch": 4.35,
-      "learning_rate": 2.191091779429441e-06,
-      "loss": 0.2494,
+      "epoch": 4.456337622085271,
+      "grad_norm": 0.2868441641330719,
+      "learning_rate": 1.5431265538431327e-06,
+      "loss": 0.3849,
       "step": 123650
     },
     {
-      "epoch": 4.35,
-      "learning_rate": 2.189925669462736e-06,
-      "loss": 0.2473,
+      "epoch": 4.456517821746495,
+      "grad_norm": 0.2647748589515686,
+      "learning_rate": 1.5421173593445442e-06,
+      "loss": 0.3783,
       "step": 123655
     },
     {
-      "epoch": 4.35,
-      "learning_rate": 2.188759855669806e-06,
-      "loss": 0.251,
+      "epoch": 4.456698021407719,
+      "grad_norm": 0.26466822624206543,
+      "learning_rate": 1.5411084844550617e-06,
+      "loss": 0.3664,
       "step": 123660
     },
     {
-      "epoch": 4.35,
-      "learning_rate": 2.1875943380658014e-06,
-      "loss": 0.2533,
+      "epoch": 4.456878221068944,
+      "grad_norm": 0.2780085504055023,
+      "learning_rate": 1.5400999291884377e-06,
+      "loss": 0.3951,
       "step": 123665
     },
     {
-      "epoch": 4.35,
-      "learning_rate": 2.186429116665847e-06,
-      "loss": 0.2504,
+      "epoch": 4.457058420730169,
+      "grad_norm": 0.26032882928848267,
+      "learning_rate": 1.5390916935584116e-06,
+      "loss": 0.4208,
       "step": 123670
     },
     {
-      "epoch": 4.35,
-      "learning_rate": 2.1852641914850757e-06,
-      "loss": 0.2285,
+      "epoch": 4.457238620391394,
+      "grad_norm": 0.23003022372722626,
+      "learning_rate": 1.5380837775787194e-06,
+      "loss": 0.3772,
       "step": 123675
     },
     {
-      "epoch": 4.35,
-      "learning_rate": 2.1840995625386163e-06,
-      "loss": 0.26,
+      "epoch": 4.457418820052618,
+      "grad_norm": 0.2510555684566498,
+      "learning_rate": 1.5370761812630918e-06,
+      "loss": 0.3706,
       "step": 123680
     },
     {
-      "epoch": 4.35,
-      "learning_rate": 2.1829352298415875e-06,
-      "loss": 0.2388,
+      "epoch": 4.457599019713843,
+      "grad_norm": 0.2845185697078705,
+      "learning_rate": 1.5360689046252542e-06,
+      "loss": 0.3997,
       "step": 123685
     },
     {
-      "epoch": 4.35,
-      "learning_rate": 2.18177119340911e-06,
-      "loss": 0.2301,
+      "epoch": 4.457779219375068,
+      "grad_norm": 0.2645367980003357,
+      "learning_rate": 1.5350619476789425e-06,
+      "loss": 0.3494,
       "step": 123690
     },
     {
-      "epoch": 4.35,
-      "learning_rate": 2.180607453256289e-06,
-      "loss": 0.2317,
+      "epoch": 4.457959419036293,
+      "grad_norm": 0.32513341307640076,
+      "learning_rate": 1.5340553104378652e-06,
+      "loss": 0.3991,
       "step": 123695
     },
     {
-      "epoch": 4.35,
-      "learning_rate": 2.1794440093982488e-06,
-      "loss": 0.2423,
+      "epoch": 4.458139618697517,
+      "grad_norm": 0.2185758650302887,
+      "learning_rate": 1.5330489929157394e-06,
+      "loss": 0.3575,
       "step": 123700
     },
     {
-      "epoch": 4.35,
-      "learning_rate": 2.1782808618500877e-06,
-      "loss": 0.2753,
+      "epoch": 4.458319818358741,
+      "grad_norm": 0.24072350561618805,
+      "learning_rate": 1.5320429951262788e-06,
+      "loss": 0.3673,
       "step": 123705
     },
     {
-      "epoch": 4.35,
-      "learning_rate": 2.177118010626908e-06,
-      "loss": 0.2499,
+      "epoch": 4.458500018019966,
+      "grad_norm": 0.2706233263015747,
+      "learning_rate": 1.5310373170831922e-06,
+      "loss": 0.4193,
       "step": 123710
     },
     {
-      "epoch": 4.35,
-      "learning_rate": 2.1759554557438083e-06,
-      "loss": 0.266,
+      "epoch": 4.458680217681191,
+      "grad_norm": 0.21717756986618042,
+      "learning_rate": 1.5300319588001766e-06,
+      "loss": 0.3553,
       "step": 123715
     },
     {
-      "epoch": 4.35,
-      "learning_rate": 2.174793197215888e-06,
-      "loss": 0.2148,
+      "epoch": 4.458860417342415,
+      "grad_norm": 0.22631758451461792,
+      "learning_rate": 1.5290269202909297e-06,
+      "loss": 0.3442,
       "step": 123720
     },
     {
-      "epoch": 4.35,
-      "learning_rate": 2.173631235058232e-06,
-      "loss": 0.2433,
+      "epoch": 4.45904061700364,
+      "grad_norm": 0.2610069215297699,
+      "learning_rate": 1.528022201569146e-06,
+      "loss": 0.3812,
       "step": 123725
     },
     {
-      "epoch": 4.35,
-      "learning_rate": 2.172469569285934e-06,
-      "loss": 0.2612,
+      "epoch": 4.459220816664865,
+      "grad_norm": 0.3348616361618042,
+      "learning_rate": 1.5270178026485172e-06,
+      "loss": 0.4102,
       "step": 123730
     },
     {
-      "epoch": 4.35,
-      "learning_rate": 2.171308199914071e-06,
-      "loss": 0.2477,
+      "epoch": 4.4594010163260895,
+      "grad_norm": 0.27878302335739136,
+      "learning_rate": 1.5260137235427268e-06,
+      "loss": 0.3641,
       "step": 123735
     },
     {
-      "epoch": 4.35,
-      "learning_rate": 2.170147126957733e-06,
-      "loss": 0.2531,
+      "epoch": 4.459581215987314,
+      "grad_norm": 0.2503393292427063,
+      "learning_rate": 1.5250099642654558e-06,
+      "loss": 0.3192,
       "step": 123740
     },
     {
-      "epoch": 4.35,
-      "learning_rate": 2.168986350431987e-06,
-      "loss": 0.2647,
+      "epoch": 4.459761415648538,
+      "grad_norm": 0.2925935983657837,
+      "learning_rate": 1.524006524830379e-06,
+      "loss": 0.3556,
       "step": 123745
     },
     {
-      "epoch": 4.35,
-      "learning_rate": 2.167825870351908e-06,
-      "loss": 0.2689,
+      "epoch": 4.459941615309763,
+      "grad_norm": 0.25310271978378296,
+      "learning_rate": 1.5230034052511637e-06,
+      "loss": 0.3719,
       "step": 123750
     },
     {
-      "epoch": 4.35,
-      "learning_rate": 2.166665686732558e-06,
-      "loss": 0.2332,
+      "epoch": 4.460121814970988,
+      "grad_norm": 0.28005677461624146,
+      "learning_rate": 1.5220006055414848e-06,
+      "loss": 0.3673,
       "step": 123755
     },
     {
-      "epoch": 4.35,
-      "learning_rate": 2.1655057995890126e-06,
-      "loss": 0.244,
+      "epoch": 4.460302014632212,
+      "grad_norm": 0.26591956615448,
+      "learning_rate": 1.520998125715009e-06,
+      "loss": 0.3491,
       "step": 123760
     },
     {
-      "epoch": 4.35,
-      "learning_rate": 2.1643462089363243e-06,
-      "loss": 0.2512,
+      "epoch": 4.460482214293437,
+      "grad_norm": 0.24851107597351074,
+      "learning_rate": 1.5199959657853784e-06,
+      "loss": 0.3389,
       "step": 123765
     },
     {
-      "epoch": 4.35,
-      "learning_rate": 2.16318691478955e-06,
-      "loss": 0.2412,
+      "epoch": 4.460662413954662,
+      "grad_norm": 0.24951453506946564,
+      "learning_rate": 1.5189941257662598e-06,
+      "loss": 0.3878,
       "step": 123770
     },
     {
-      "epoch": 4.35,
-      "learning_rate": 2.162027917163742e-06,
-      "loss": 0.2264,
+      "epoch": 4.4608426136158865,
+      "grad_norm": 0.26914355158805847,
+      "learning_rate": 1.5179926056713006e-06,
+      "loss": 0.3913,
       "step": 123775
     },
     {
-      "epoch": 4.35,
-      "learning_rate": 2.1608692160739465e-06,
-      "loss": 0.2461,
+      "epoch": 4.461022813277111,
+      "grad_norm": 0.2922750413417816,
+      "learning_rate": 1.5169914055141427e-06,
+      "loss": 0.3501,
       "step": 123780
     },
     {
-      "epoch": 4.36,
-      "learning_rate": 2.1597108115352204e-06,
-      "loss": 0.2477,
+      "epoch": 4.461203012938336,
+      "grad_norm": 0.22999052703380585,
+      "learning_rate": 1.5159905253084388e-06,
+      "loss": 0.3749,
       "step": 123785
     },
     {
-      "epoch": 4.36,
-      "learning_rate": 2.1585527035625913e-06,
-      "loss": 0.2551,
+      "epoch": 4.461383212599561,
+      "grad_norm": 0.3160970211029053,
+      "learning_rate": 1.514989965067809e-06,
+      "loss": 0.3529,
       "step": 123790
     },
     {
-      "epoch": 4.36,
-      "learning_rate": 2.157394892171108e-06,
-      "loss": 0.2482,
+      "epoch": 4.461563412260785,
+      "grad_norm": 0.21242515742778778,
+      "learning_rate": 1.5139897248058977e-06,
+      "loss": 0.3838,
       "step": 123795
     },
     {
-      "epoch": 4.36,
-      "learning_rate": 2.156237377375797e-06,
-      "loss": 0.2461,
+      "epoch": 4.461743611922009,
+      "grad_norm": 0.24923621118068695,
+      "learning_rate": 1.5129898045363273e-06,
+      "loss": 0.3617,
       "step": 123800
     },
     {
-      "epoch": 4.36,
-      "learning_rate": 2.1550801591916912e-06,
-      "loss": 0.2364,
+      "epoch": 4.461923811583234,
+      "grad_norm": 0.25680771470069885,
+      "learning_rate": 1.5119902042727257e-06,
+      "loss": 0.3805,
       "step": 123805
     },
     {
-      "epoch": 4.36,
-      "learning_rate": 2.1539232376338065e-06,
-      "loss": 0.2472,
+      "epoch": 4.462104011244459,
+      "grad_norm": 0.29142534732818604,
+      "learning_rate": 1.51099092402871e-06,
+      "loss": 0.3772,
       "step": 123810
     },
     {
-      "epoch": 4.36,
-      "learning_rate": 2.1527666127171802e-06,
-      "loss": 0.2429,
+      "epoch": 4.4622842109056835,
+      "grad_norm": 0.212846577167511,
+      "learning_rate": 1.5099919638178972e-06,
+      "loss": 0.3341,
       "step": 123815
     },
     {
-      "epoch": 4.36,
-      "learning_rate": 2.1516102844568204e-06,
-      "loss": 0.2437,
+      "epoch": 4.462464410566908,
+      "grad_norm": 0.22237548232078552,
+      "learning_rate": 1.50899332365389e-06,
+      "loss": 0.354,
       "step": 123820
     },
     {
-      "epoch": 4.36,
-      "learning_rate": 2.150454252867745e-06,
-      "loss": 0.2514,
+      "epoch": 4.462644610228133,
+      "grad_norm": 0.2658112645149231,
+      "learning_rate": 1.5079950035503054e-06,
+      "loss": 0.3755,
       "step": 123825
     },
     {
-      "epoch": 4.36,
-      "learning_rate": 2.149298517964957e-06,
-      "loss": 0.2665,
+      "epoch": 4.462824809889358,
+      "grad_norm": 0.2840268611907959,
+      "learning_rate": 1.506997003520741e-06,
+      "loss": 0.3935,
       "step": 123830
     },
     {
-      "epoch": 4.36,
-      "learning_rate": 2.1481430797634735e-06,
-      "loss": 0.2393,
+      "epoch": 4.463005009550582,
+      "grad_norm": 0.2549954354763031,
+      "learning_rate": 1.5059993235787912e-06,
+      "loss": 0.3839,
       "step": 123835
     },
     {
-      "epoch": 4.36,
-      "learning_rate": 2.146987938278286e-06,
-      "loss": 0.2507,
+      "epoch": 4.463185209211806,
+      "grad_norm": 0.26900264620780945,
+      "learning_rate": 1.5050019637380565e-06,
+      "loss": 0.3575,
       "step": 123840
     },
     {
-      "epoch": 4.36,
-      "learning_rate": 2.145833093524405e-06,
-      "loss": 0.2336,
+      "epoch": 4.463365408873031,
+      "grad_norm": 0.28676745295524597,
+      "learning_rate": 1.5040049240121145e-06,
+      "loss": 0.3925,
       "step": 123845
     },
     {
-      "epoch": 4.36,
-      "learning_rate": 2.144678545516815e-06,
-      "loss": 0.2427,
+      "epoch": 4.463545608534256,
+      "grad_norm": 0.24285760521888733,
+      "learning_rate": 1.5030082044145605e-06,
+      "loss": 0.3623,
       "step": 123850
     },
     {
-      "epoch": 4.36,
-      "learning_rate": 2.1435242942705187e-06,
-      "loss": 0.255,
+      "epoch": 4.4637258081954805,
+      "grad_norm": 0.2883954346179962,
+      "learning_rate": 1.5020118049589722e-06,
+      "loss": 0.3763,
       "step": 123855
     },
     {
-      "epoch": 4.36,
-      "learning_rate": 2.1423703398004954e-06,
-      "loss": 0.243,
+      "epoch": 4.463906007856705,
+      "grad_norm": 0.31590425968170166,
+      "learning_rate": 1.5010157256589218e-06,
+      "loss": 0.3716,
       "step": 123860
     },
     {
-      "epoch": 4.36,
-      "learning_rate": 2.141216682121727e-06,
-      "loss": 0.2506,
+      "epoch": 4.46408620751793,
+      "grad_norm": 0.21175405383110046,
+      "learning_rate": 1.5000199665279824e-06,
+      "loss": 0.3788,
       "step": 123865
     },
     {
-      "epoch": 4.36,
-      "learning_rate": 2.1400633212491944e-06,
-      "loss": 0.2229,
+      "epoch": 4.464266407179155,
+      "grad_norm": 0.23191265761852264,
+      "learning_rate": 1.4990245275797177e-06,
+      "loss": 0.3739,
       "step": 123870
     },
     {
-      "epoch": 4.36,
-      "learning_rate": 2.1389102571978764e-06,
-      "loss": 0.266,
+      "epoch": 4.464446606840379,
+      "grad_norm": 0.23299415409564972,
+      "learning_rate": 1.498029408827703e-06,
+      "loss": 0.3612,
       "step": 123875
     },
     {
-      "epoch": 4.36,
-      "learning_rate": 2.137757489982742e-06,
-      "loss": 0.2635,
+      "epoch": 4.464626806501604,
+      "grad_norm": 0.28248894214630127,
+      "learning_rate": 1.4970346102854831e-06,
+      "loss": 0.3685,
       "step": 123880
     },
     {
-      "epoch": 4.36,
-      "learning_rate": 2.1366050196187564e-06,
-      "loss": 0.2646,
+      "epoch": 4.464807006162829,
+      "grad_norm": 0.27142465114593506,
+      "learning_rate": 1.4960401319666113e-06,
+      "loss": 0.3475,
       "step": 123885
     },
     {
-      "epoch": 4.36,
-      "learning_rate": 2.1354528461208944e-06,
-      "loss": 0.2281,
+      "epoch": 4.464987205824053,
+      "grad_norm": 0.2543891966342926,
+      "learning_rate": 1.4950459738846455e-06,
+      "loss": 0.3551,
       "step": 123890
     },
     {
-      "epoch": 4.36,
-      "learning_rate": 2.134300969504102e-06,
-      "loss": 0.2389,
+      "epoch": 4.4651674054852775,
+      "grad_norm": 0.22357742488384247,
+      "learning_rate": 1.4940521360531257e-06,
+      "loss": 0.396,
       "step": 123895
     },
     {
-      "epoch": 4.36,
-      "learning_rate": 2.1331493897833452e-06,
-      "loss": 0.2276,
+      "epoch": 4.465347605146502,
+      "grad_norm": 0.3259972333908081,
+      "learning_rate": 1.4930586184856016e-06,
+      "loss": 0.3484,
       "step": 123900
     },
     {
-      "epoch": 4.36,
-      "learning_rate": 2.1319981069735706e-06,
-      "loss": 0.246,
+      "epoch": 4.465527804807727,
+      "grad_norm": 0.23107150197029114,
+      "learning_rate": 1.4920654211955987e-06,
+      "loss": 0.3455,
       "step": 123905
     },
     {
-      "epoch": 4.36,
-      "learning_rate": 2.1308471210897352e-06,
-      "loss": 0.2725,
+      "epoch": 4.465708004468952,
+      "grad_norm": 0.2234567403793335,
+      "learning_rate": 1.4910725441966505e-06,
+      "loss": 0.3741,
       "step": 123910
     },
     {
-      "epoch": 4.36,
-      "learning_rate": 2.12969643214678e-06,
-      "loss": 0.2385,
+      "epoch": 4.465888204130176,
+      "grad_norm": 0.23561283946037292,
+      "learning_rate": 1.4900799875022908e-06,
+      "loss": 0.3616,
       "step": 123915
     },
     {
-      "epoch": 4.36,
-      "learning_rate": 2.12854604015964e-06,
-      "loss": 0.2329,
+      "epoch": 4.466068403791401,
+      "grad_norm": 0.2155967354774475,
+      "learning_rate": 1.4890877511260393e-06,
+      "loss": 0.3914,
       "step": 123920
     },
     {
-      "epoch": 4.36,
-      "learning_rate": 2.1273959451432567e-06,
-      "loss": 0.25,
+      "epoch": 4.466248603452626,
+      "grad_norm": 0.22571438550949097,
+      "learning_rate": 1.4880958350814156e-06,
+      "loss": 0.3525,
       "step": 123925
     },
     {
-      "epoch": 4.36,
-      "learning_rate": 2.126246147112565e-06,
-      "loss": 0.2414,
+      "epoch": 4.46642880311385,
+      "grad_norm": 0.2629721462726593,
+      "learning_rate": 1.487104239381934e-06,
+      "loss": 0.4074,
       "step": 123930
     },
     {
-      "epoch": 4.36,
-      "learning_rate": 2.1250966460824943e-06,
-      "loss": 0.255,
+      "epoch": 4.466609002775074,
+      "grad_norm": 0.2295321524143219,
+      "learning_rate": 1.4861129640411004e-06,
+      "loss": 0.3878,
       "step": 123935
     },
     {
-      "epoch": 4.36,
-      "learning_rate": 2.123947442067964e-06,
-      "loss": 0.2632,
+      "epoch": 4.466789202436299,
+      "grad_norm": 0.2172783464193344,
+      "learning_rate": 1.4851220090724288e-06,
+      "loss": 0.3812,
       "step": 123940
     },
     {
-      "epoch": 4.36,
-      "learning_rate": 2.122798535083906e-06,
-      "loss": 0.2888,
+      "epoch": 4.466969402097524,
+      "grad_norm": 0.2786194980144501,
+      "learning_rate": 1.484131374489417e-06,
+      "loss": 0.3771,
       "step": 123945
     },
     {
-      "epoch": 4.36,
-      "learning_rate": 2.121649925145225e-06,
-      "loss": 0.24,
+      "epoch": 4.467149601758749,
+      "grad_norm": 0.25498166680336,
+      "learning_rate": 1.4831410603055651e-06,
+      "loss": 0.3795,
       "step": 123950
     },
     {
-      "epoch": 4.36,
-      "learning_rate": 2.120501612266848e-06,
-      "loss": 0.2417,
+      "epoch": 4.467329801419973,
+      "grad_norm": 0.24183687567710876,
+      "learning_rate": 1.4821510665343595e-06,
+      "loss": 0.4133,
       "step": 123955
     },
     {
-      "epoch": 4.36,
-      "learning_rate": 2.11935359646368e-06,
-      "loss": 0.2631,
+      "epoch": 4.467510001081198,
+      "grad_norm": 0.2495322972536087,
+      "learning_rate": 1.4811613931892953e-06,
+      "loss": 0.3759,
       "step": 123960
     },
     {
-      "epoch": 4.36,
-      "learning_rate": 2.118205877750623e-06,
-      "loss": 0.2194,
+      "epoch": 4.467690200742423,
+      "grad_norm": 0.2298755794763565,
+      "learning_rate": 1.4801720402838503e-06,
+      "loss": 0.3509,
       "step": 123965
     },
     {
-      "epoch": 4.36,
-      "learning_rate": 2.117058456142587e-06,
-      "loss": 0.2435,
+      "epoch": 4.467870400403648,
+      "grad_norm": 0.2513538897037506,
+      "learning_rate": 1.4791830078315139e-06,
+      "loss": 0.3569,
       "step": 123970
     },
     {
-      "epoch": 4.36,
-      "learning_rate": 2.115911331654466e-06,
-      "loss": 0.2673,
+      "epoch": 4.468050600064872,
+      "grad_norm": 0.22999824583530426,
+      "learning_rate": 1.4781942958457474e-06,
+      "loss": 0.3452,
       "step": 123975
     },
     {
-      "epoch": 4.36,
-      "learning_rate": 2.1147645043011533e-06,
-      "loss": 0.247,
+      "epoch": 4.468230799726096,
+      "grad_norm": 0.29147133231163025,
+      "learning_rate": 1.4772059043400317e-06,
+      "loss": 0.3672,
       "step": 123980
     },
     {
-      "epoch": 4.36,
-      "learning_rate": 2.11361797409754e-06,
-      "loss": 0.2425,
+      "epoch": 4.468410999387321,
+      "grad_norm": 0.2764071226119995,
+      "learning_rate": 1.4762178333278337e-06,
+      "loss": 0.365,
       "step": 123985
     },
     {
-      "epoch": 4.36,
-      "learning_rate": 2.1124717410585164e-06,
-      "loss": 0.2498,
+      "epoch": 4.468591199048546,
+      "grad_norm": 0.22821201384067535,
+      "learning_rate": 1.475230082822615e-06,
+      "loss": 0.3937,
       "step": 123990
     },
     {
-      "epoch": 4.36,
-      "learning_rate": 2.111325805198966e-06,
-      "loss": 0.2905,
+      "epoch": 4.46877139870977,
+      "grad_norm": 0.27766650915145874,
+      "learning_rate": 1.474242652837829e-06,
+      "loss": 0.3654,
       "step": 123995
     },
     {
-      "epoch": 4.36,
-      "learning_rate": 2.1101801665337595e-06,
-      "loss": 0.254,
+      "epoch": 4.468951598370995,
+      "grad_norm": 0.21973931789398193,
+      "learning_rate": 1.4732555433869339e-06,
+      "loss": 0.3532,
       "step": 124000
     },
     {
-      "epoch": 4.36,
-      "eval_loss": 0.24889464676380157,
-      "eval_runtime": 10.5436,
-      "eval_samples_per_second": 9.484,
-      "eval_steps_per_second": 9.484,
+      "epoch": 4.468951598370995,
+      "eval_loss": 0.42887455224990845,
+      "eval_runtime": 3.5305,
+      "eval_samples_per_second": 28.324,
+      "eval_steps_per_second": 7.081,
       "step": 124000
     },
     {
-      "epoch": 4.36,
-      "learning_rate": 2.1090348250777797e-06,
-      "loss": 0.2245,
+      "epoch": 4.46913179803222,
+      "grad_norm": 0.18677954375743866,
+      "learning_rate": 1.4722687544833719e-06,
+      "loss": 0.3637,
       "step": 124005
     },
     {
-      "epoch": 4.36,
-      "learning_rate": 2.107889780845904e-06,
-      "loss": 0.2225,
+      "epoch": 4.469311997693445,
+      "grad_norm": 0.22494982182979584,
+      "learning_rate": 1.4712822861405962e-06,
+      "loss": 0.3444,
       "step": 124010
     },
     {
-      "epoch": 4.36,
-      "learning_rate": 2.1067450338529887e-06,
-      "loss": 0.2623,
+      "epoch": 4.469492197354669,
+      "grad_norm": 0.2011212706565857,
+      "learning_rate": 1.470296138372046e-06,
+      "loss": 0.3575,
       "step": 124015
     },
     {
-      "epoch": 4.36,
-      "learning_rate": 2.1056005841139007e-06,
-      "loss": 0.2521,
+      "epoch": 4.469672397015893,
+      "grad_norm": 0.2682434320449829,
+      "learning_rate": 1.4693103111911549e-06,
+      "loss": 0.3655,
       "step": 124020
     },
     {
-      "epoch": 4.36,
-      "learning_rate": 2.104456431643503e-06,
-      "loss": 0.2633,
+      "epoch": 4.469852596677118,
+      "grad_norm": 0.2674734890460968,
+      "learning_rate": 1.468324804611354e-06,
+      "loss": 0.3717,
       "step": 124025
     },
     {
-      "epoch": 4.36,
-      "learning_rate": 2.103312576456651e-06,
-      "loss": 0.2417,
+      "epoch": 4.470032796338343,
+      "grad_norm": 0.24927879869937897,
+      "learning_rate": 1.4673396186460687e-06,
+      "loss": 0.3839,
       "step": 124030
     },
     {
-      "epoch": 4.36,
-      "learning_rate": 2.1021690185681952e-06,
-      "loss": 0.2539,
+      "epoch": 4.470212995999567,
+      "grad_norm": 0.2811104953289032,
+      "learning_rate": 1.4663547533087297e-06,
+      "loss": 0.347,
       "step": 124035
     },
     {
-      "epoch": 4.36,
-      "learning_rate": 2.1010257579929806e-06,
-      "loss": 0.2413,
+      "epoch": 4.470393195660792,
+      "grad_norm": 0.28087154030799866,
+      "learning_rate": 1.4653702086127541e-06,
+      "loss": 0.4091,
       "step": 124040
     },
     {
-      "epoch": 4.36,
-      "learning_rate": 2.099882794745861e-06,
-      "loss": 0.2429,
+      "epoch": 4.470573395322017,
+      "grad_norm": 0.2959977686405182,
+      "learning_rate": 1.464385984571545e-06,
+      "loss": 0.3983,
       "step": 124045
     },
     {
-      "epoch": 4.36,
-      "learning_rate": 2.0987401288416697e-06,
-      "loss": 0.2523,
+      "epoch": 4.4707535949832415,
+      "grad_norm": 0.3171277344226837,
+      "learning_rate": 1.4634020811985255e-06,
+      "loss": 0.36,
       "step": 124050
     },
     {
-      "epoch": 4.36,
-      "learning_rate": 2.0975977602952418e-06,
-      "loss": 0.2338,
+      "epoch": 4.470933794644466,
+      "grad_norm": 0.2086503803730011,
+      "learning_rate": 1.46241849850709e-06,
+      "loss": 0.3771,
       "step": 124055
     },
     {
-      "epoch": 4.36,
-      "learning_rate": 2.0964556891214127e-06,
-      "loss": 0.2462,
+      "epoch": 4.471113994305691,
+      "grad_norm": 0.23770023882389069,
+      "learning_rate": 1.46143523651065e-06,
+      "loss": 0.3546,
       "step": 124060
     },
     {
-      "epoch": 4.36,
-      "learning_rate": 2.0953139153350175e-06,
-      "loss": 0.2491,
+      "epoch": 4.471294193966916,
+      "grad_norm": 0.319051057100296,
+      "learning_rate": 1.4604522952226001e-06,
+      "loss": 0.3848,
       "step": 124065
     },
     {
-      "epoch": 4.37,
-      "learning_rate": 2.0941724389508783e-06,
-      "loss": 0.237,
+      "epoch": 4.47147439362814,
+      "grad_norm": 0.1936783492565155,
+      "learning_rate": 1.4594696746563246e-06,
+      "loss": 0.3537,
       "step": 124070
     },
     {
-      "epoch": 4.37,
-      "learning_rate": 2.0930312599838132e-06,
-      "loss": 0.2343,
+      "epoch": 4.471654593289364,
+      "grad_norm": 0.25065305829048157,
+      "learning_rate": 1.4584873748252208e-06,
+      "loss": 0.3907,
       "step": 124075
     },
     {
-      "epoch": 4.37,
-      "learning_rate": 2.0918903784486355e-06,
-      "loss": 0.2433,
+      "epoch": 4.471834792950589,
+      "grad_norm": 0.23407846689224243,
+      "learning_rate": 1.4575053957426643e-06,
+      "loss": 0.3819,
       "step": 124080
     },
     {
-      "epoch": 4.37,
-      "learning_rate": 2.0907497943601696e-06,
-      "loss": 0.2395,
+      "epoch": 4.472014992611814,
+      "grad_norm": 0.2705335319042206,
+      "learning_rate": 1.456523737422047e-06,
+      "loss": 0.3762,
       "step": 124085
     },
     {
-      "epoch": 4.37,
-      "learning_rate": 2.0896095077332202e-06,
-      "loss": 0.2481,
+      "epoch": 4.4721951922730385,
+      "grad_norm": 0.2572081685066223,
+      "learning_rate": 1.4555423998767304e-06,
+      "loss": 0.3961,
       "step": 124090
     },
     {
-      "epoch": 4.37,
-      "learning_rate": 2.088469518582592e-06,
-      "loss": 0.2496,
+      "epoch": 4.472375391934263,
+      "grad_norm": 0.2333773970603943,
+      "learning_rate": 1.45456138312009e-06,
+      "loss": 0.3601,
       "step": 124095
     },
     {
-      "epoch": 4.37,
-      "learning_rate": 2.0873298269230816e-06,
-      "loss": 0.2363,
+      "epoch": 4.472555591595488,
+      "grad_norm": 0.2367962896823883,
+      "learning_rate": 1.4535806871654927e-06,
+      "loss": 0.3246,
       "step": 124100
     },
     {
-      "epoch": 4.37,
-      "learning_rate": 2.0861904327695e-06,
-      "loss": 0.2617,
+      "epoch": 4.472735791256713,
+      "grad_norm": 0.23657187819480896,
+      "learning_rate": 1.4526003120263005e-06,
+      "loss": 0.387,
       "step": 124105
     },
     {
-      "epoch": 4.37,
-      "learning_rate": 2.0850513361366313e-06,
-      "loss": 0.2712,
+      "epoch": 4.4729159909179375,
+      "grad_norm": 0.2925667464733124,
+      "learning_rate": 1.451620257715869e-06,
+      "loss": 0.3759,
       "step": 124110
     },
     {
-      "epoch": 4.37,
-      "learning_rate": 2.0839125370392732e-06,
-      "loss": 0.2404,
+      "epoch": 4.473096190579161,
+      "grad_norm": 0.30918559432029724,
+      "learning_rate": 1.450640524247554e-06,
+      "loss": 0.3738,
       "step": 124115
     },
     {
-      "epoch": 4.37,
-      "learning_rate": 2.082774035492202e-06,
-      "loss": 0.2523,
+      "epoch": 4.473276390240386,
+      "grad_norm": 0.32785508036613464,
+      "learning_rate": 1.449661111634698e-06,
+      "loss": 0.3538,
       "step": 124120
     },
     {
-      "epoch": 4.37,
-      "learning_rate": 2.0816358315102147e-06,
-      "loss": 0.2495,
+      "epoch": 4.473456589901611,
+      "grad_norm": 0.25581833720207214,
+      "learning_rate": 1.4486820198906542e-06,
+      "loss": 0.3863,
       "step": 124125
     },
     {
-      "epoch": 4.37,
-      "learning_rate": 2.0804979251080795e-06,
-      "loss": 0.2443,
+      "epoch": 4.4736367895628355,
+      "grad_norm": 0.2755228877067566,
+      "learning_rate": 1.447703249028759e-06,
+      "loss": 0.3925,
       "step": 124130
     },
     {
-      "epoch": 4.37,
-      "learning_rate": 2.0793603163005713e-06,
-      "loss": 0.263,
+      "epoch": 4.47381698922406,
+      "grad_norm": 0.29217204451560974,
+      "learning_rate": 1.4467247990623462e-06,
+      "loss": 0.3763,
       "step": 124135
     },
     {
-      "epoch": 4.37,
-      "learning_rate": 2.078223005102467e-06,
-      "loss": 0.26,
+      "epoch": 4.473997188885285,
+      "grad_norm": 0.22582022845745087,
+      "learning_rate": 1.4457466700047496e-06,
+      "loss": 0.3484,
       "step": 124140
     },
     {
-      "epoch": 4.37,
-      "learning_rate": 2.077085991528532e-06,
-      "loss": 0.251,
+      "epoch": 4.47417738854651,
+      "grad_norm": 0.2606388330459595,
+      "learning_rate": 1.4447688618692916e-06,
+      "loss": 0.3651,
       "step": 124145
     },
     {
-      "epoch": 4.37,
-      "learning_rate": 2.0759492755935304e-06,
-      "loss": 0.2627,
+      "epoch": 4.474357588207734,
+      "grad_norm": 0.24026912450790405,
+      "learning_rate": 1.4437913746692955e-06,
+      "loss": 0.3714,
       "step": 124150
     },
     {
-      "epoch": 4.37,
-      "learning_rate": 2.0748128573122138e-06,
-      "loss": 0.2459,
+      "epoch": 4.474537787868959,
+      "grad_norm": 0.28807055950164795,
+      "learning_rate": 1.4428142084180862e-06,
+      "loss": 0.359,
       "step": 124155
     },
     {
-      "epoch": 4.37,
-      "learning_rate": 2.073676736699348e-06,
-      "loss": 0.2482,
+      "epoch": 4.474717987530184,
+      "grad_norm": 0.23027734458446503,
+      "learning_rate": 1.4418373631289671e-06,
+      "loss": 0.3887,
       "step": 124160
     },
     {
-      "epoch": 4.37,
-      "learning_rate": 2.0725409137696774e-06,
-      "loss": 0.2548,
+      "epoch": 4.474898187191408,
+      "grad_norm": 0.2151884138584137,
+      "learning_rate": 1.4408608388152583e-06,
+      "loss": 0.3844,
       "step": 124165
     },
     {
-      "epoch": 4.37,
-      "learning_rate": 2.0714053885379585e-06,
-      "loss": 0.2491,
+      "epoch": 4.4750783868526325,
+      "grad_norm": 0.30616864562034607,
+      "learning_rate": 1.4398846354902574e-06,
+      "loss": 0.376,
       "step": 124170
     },
     {
-      "epoch": 4.37,
-      "learning_rate": 2.0702701610189223e-06,
-      "loss": 0.2515,
+      "epoch": 4.475258586513857,
+      "grad_norm": 0.28647398948669434,
+      "learning_rate": 1.4389087531672674e-06,
+      "loss": 0.3926,
       "step": 124175
     },
     {
-      "epoch": 4.37,
-      "learning_rate": 2.069135231227323e-06,
-      "loss": 0.2509,
+      "epoch": 4.475438786175082,
+      "grad_norm": 0.24233390390872955,
+      "learning_rate": 1.4379331918595835e-06,
+      "loss": 0.3548,
       "step": 124180
     },
     {
-      "epoch": 4.37,
-      "learning_rate": 2.068000599177891e-06,
-      "loss": 0.2418,
+      "epoch": 4.475618985836307,
+      "grad_norm": 0.2877272963523865,
+      "learning_rate": 1.4369579515804975e-06,
+      "loss": 0.3599,
       "step": 124185
     },
     {
-      "epoch": 4.37,
-      "learning_rate": 2.0668662648853587e-06,
-      "loss": 0.2332,
+      "epoch": 4.475799185497531,
+      "grad_norm": 0.2793196737766266,
+      "learning_rate": 1.4359830323432937e-06,
+      "loss": 0.3681,
       "step": 124190
     },
     {
-      "epoch": 4.37,
-      "learning_rate": 2.0657322283644503e-06,
-      "loss": 0.2485,
+      "epoch": 4.475979385158756,
+      "grad_norm": 0.21865501999855042,
+      "learning_rate": 1.435008434161264e-06,
+      "loss": 0.3772,
       "step": 124195
     },
     {
-      "epoch": 4.37,
-      "learning_rate": 2.0645984896298986e-06,
-      "loss": 0.2504,
+      "epoch": 4.476159584819981,
+      "grad_norm": 0.24669122695922852,
+      "learning_rate": 1.434034157047684e-06,
+      "loss": 0.3853,
       "step": 124200
     },
     {
-      "epoch": 4.37,
-      "learning_rate": 2.0634650486964225e-06,
-      "loss": 0.2497,
+      "epoch": 4.476339784481205,
+      "grad_norm": 0.23632431030273438,
+      "learning_rate": 1.4330602010158235e-06,
+      "loss": 0.3896,
       "step": 124205
     },
     {
-      "epoch": 4.37,
-      "learning_rate": 2.062331905578735e-06,
-      "loss": 0.2594,
+      "epoch": 4.4765199841424295,
+      "grad_norm": 0.2615738809108734,
+      "learning_rate": 1.4320865660789551e-06,
+      "loss": 0.3571,
       "step": 124210
     },
     {
-      "epoch": 4.37,
-      "learning_rate": 2.0611990602915464e-06,
-      "loss": 0.2553,
+      "epoch": 4.476700183803654,
+      "grad_norm": 0.23187601566314697,
+      "learning_rate": 1.4311132522503407e-06,
+      "loss": 0.3799,
       "step": 124215
     },
     {
-      "epoch": 4.37,
-      "learning_rate": 2.0600665128495707e-06,
-      "loss": 0.2484,
+      "epoch": 4.476880383464879,
+      "grad_norm": 0.23954427242279053,
+      "learning_rate": 1.4301402595432472e-06,
+      "loss": 0.3563,
       "step": 124220
     },
     {
-      "epoch": 4.37,
-      "learning_rate": 2.0589342632675173e-06,
-      "loss": 0.2325,
+      "epoch": 4.477060583126104,
+      "grad_norm": 0.2520751953125,
+      "learning_rate": 1.429167587970931e-06,
+      "loss": 0.3862,
       "step": 124225
     },
     {
-      "epoch": 4.37,
-      "learning_rate": 2.0578023115600866e-06,
-      "loss": 0.2434,
+      "epoch": 4.477240782787328,
+      "grad_norm": 0.2765583097934723,
+      "learning_rate": 1.4281952375466451e-06,
+      "loss": 0.3983,
       "step": 124230
     },
     {
-      "epoch": 4.37,
-      "learning_rate": 2.056670657741966e-06,
-      "loss": 0.2457,
+      "epoch": 4.477420982448553,
+      "grad_norm": 0.2432951033115387,
+      "learning_rate": 1.4272232082836317e-06,
+      "loss": 0.3751,
       "step": 124235
     },
     {
-      "epoch": 4.37,
-      "learning_rate": 2.0555393018278635e-06,
-      "loss": 0.2459,
+      "epoch": 4.477601182109778,
+      "grad_norm": 0.2171791046857834,
+      "learning_rate": 1.4262515001951389e-06,
+      "loss": 0.3707,
       "step": 124240
     },
     {
-      "epoch": 4.37,
-      "learning_rate": 2.054408243832459e-06,
-      "loss": 0.2463,
+      "epoch": 4.477781381771003,
+      "grad_norm": 0.2439403533935547,
+      "learning_rate": 1.4252801132944055e-06,
+      "loss": 0.3821,
       "step": 124245
     },
     {
-      "epoch": 4.37,
-      "learning_rate": 2.053277483770444e-06,
-      "loss": 0.2368,
+      "epoch": 4.477961581432227,
+      "grad_norm": 0.2420462816953659,
+      "learning_rate": 1.4243090475946713e-06,
+      "loss": 0.3772,
       "step": 124250
     },
     {
-      "epoch": 4.37,
-      "learning_rate": 2.0521470216564946e-06,
-      "loss": 0.2472,
+      "epoch": 4.478141781093451,
+      "grad_norm": 0.28287315368652344,
+      "learning_rate": 1.4233383031091535e-06,
+      "loss": 0.3906,
       "step": 124255
     },
     {
-      "epoch": 4.37,
-      "learning_rate": 2.0510168575052945e-06,
-      "loss": 0.2458,
+      "epoch": 4.478321980754676,
+      "grad_norm": 0.22504858672618866,
+      "learning_rate": 1.4223678798510915e-06,
+      "loss": 0.3988,
       "step": 124260
     },
     {
-      "epoch": 4.37,
-      "learning_rate": 2.0498869913315206e-06,
-      "loss": 0.2398,
+      "epoch": 4.478502180415901,
+      "grad_norm": 0.2430412620306015,
+      "learning_rate": 1.4213977778336995e-06,
+      "loss": 0.3859,
       "step": 124265
     },
     {
-      "epoch": 4.37,
-      "learning_rate": 2.048757423149833e-06,
-      "loss": 0.2281,
+      "epoch": 4.478682380077125,
+      "grad_norm": 0.31164422631263733,
+      "learning_rate": 1.4204279970702034e-06,
+      "loss": 0.4001,
       "step": 124270
     },
     {
-      "epoch": 4.37,
-      "learning_rate": 2.0476281529749092e-06,
-      "loss": 0.268,
+      "epoch": 4.47886257973835,
+      "grad_norm": 0.3133993446826935,
+      "learning_rate": 1.4194585375738061e-06,
+      "loss": 0.3915,
       "step": 124275
     },
     {
-      "epoch": 4.37,
-      "learning_rate": 2.046499180821401e-06,
-      "loss": 0.259,
+      "epoch": 4.479042779399575,
+      "grad_norm": 0.27546271681785583,
+      "learning_rate": 1.4184893993577198e-06,
+      "loss": 0.3829,
       "step": 124280
     },
     {
-      "epoch": 4.37,
-      "learning_rate": 2.045370506703981e-06,
-      "loss": 0.2406,
+      "epoch": 4.4792229790608,
+      "grad_norm": 0.32316166162490845,
+      "learning_rate": 1.417520582435153e-06,
+      "loss": 0.3139,
       "step": 124285
     },
     {
-      "epoch": 4.37,
-      "learning_rate": 2.0442421306372945e-06,
-      "loss": 0.258,
+      "epoch": 4.479403178722024,
+      "grad_norm": 0.23698268830776215,
+      "learning_rate": 1.4165520868193034e-06,
+      "loss": 0.3568,
       "step": 124290
     },
     {
-      "epoch": 4.37,
-      "learning_rate": 2.043114052635997e-06,
-      "loss": 0.2724,
+      "epoch": 4.479583378383249,
+      "grad_norm": 0.26185452938079834,
+      "learning_rate": 1.4155839125233634e-06,
+      "loss": 0.3757,
       "step": 124295
     },
     {
-      "epoch": 4.37,
-      "learning_rate": 2.0419862727147356e-06,
-      "loss": 0.2524,
+      "epoch": 4.479763578044473,
+      "grad_norm": 0.24788565933704376,
+      "learning_rate": 1.414616059560528e-06,
+      "loss": 0.3441,
       "step": 124300
     },
     {
-      "epoch": 4.37,
-      "learning_rate": 2.0408587908881505e-06,
-      "loss": 0.2328,
+      "epoch": 4.479943777705698,
+      "grad_norm": 0.3030286729335785,
+      "learning_rate": 1.4136485279439783e-06,
+      "loss": 0.3465,
       "step": 124305
     },
     {
-      "epoch": 4.37,
-      "learning_rate": 2.0397316071708806e-06,
-      "loss": 0.2231,
+      "epoch": 4.480123977366922,
+      "grad_norm": 0.2569609582424164,
+      "learning_rate": 1.4126813176869036e-06,
+      "loss": 0.3192,
       "step": 124310
     },
     {
-      "epoch": 4.37,
-      "learning_rate": 2.0386047215775692e-06,
-      "loss": 0.2343,
+      "epoch": 4.480304177028147,
+      "grad_norm": 0.22701065242290497,
+      "learning_rate": 1.4117144288024797e-06,
+      "loss": 0.3387,
       "step": 124315
     },
     {
-      "epoch": 4.37,
-      "learning_rate": 2.0374781341228444e-06,
-      "loss": 0.2617,
+      "epoch": 4.480484376689372,
+      "grad_norm": 0.23131400346755981,
+      "learning_rate": 1.4107478613038821e-06,
+      "loss": 0.3596,
       "step": 124320
     },
     {
-      "epoch": 4.37,
-      "learning_rate": 2.036351844821327e-06,
-      "loss": 0.2502,
+      "epoch": 4.480664576350597,
+      "grad_norm": 0.260778546333313,
+      "learning_rate": 1.4097816152042753e-06,
+      "loss": 0.3689,
       "step": 124325
     },
     {
-      "epoch": 4.37,
-      "learning_rate": 2.0352258536876528e-06,
-      "loss": 0.2625,
+      "epoch": 4.480844776011821,
+      "grad_norm": 0.23251162469387054,
+      "learning_rate": 1.4088156905168293e-06,
+      "loss": 0.3959,
       "step": 124330
     },
     {
-      "epoch": 4.37,
-      "learning_rate": 2.03410016073643e-06,
-      "loss": 0.2404,
+      "epoch": 4.481024975673046,
+      "grad_norm": 0.25117242336273193,
+      "learning_rate": 1.4078500872546973e-06,
+      "loss": 0.3879,
       "step": 124335
     },
     {
-      "epoch": 4.37,
-      "learning_rate": 2.0329747659822873e-06,
-      "loss": 0.2491,
+      "epoch": 4.481205175334271,
+      "grad_norm": 0.2599792778491974,
+      "learning_rate": 1.4068848054310469e-06,
+      "loss": 0.3692,
       "step": 124340
     },
     {
-      "epoch": 4.37,
-      "learning_rate": 2.031849669439831e-06,
-      "loss": 0.24,
+      "epoch": 4.481385374995495,
+      "grad_norm": 0.28005966544151306,
+      "learning_rate": 1.4059198450590143e-06,
+      "loss": 0.3937,
       "step": 124345
     },
     {
-      "epoch": 4.37,
-      "learning_rate": 2.030724871123668e-06,
-      "loss": 0.2722,
+      "epoch": 4.481565574656719,
+      "grad_norm": 0.24943847954273224,
+      "learning_rate": 1.4049552061517619e-06,
+      "loss": 0.344,
       "step": 124350
     },
     {
-      "epoch": 4.38,
-      "learning_rate": 2.029600371048407e-06,
-      "loss": 0.2395,
+      "epoch": 4.481745774317944,
+      "grad_norm": 0.2192724198102951,
+      "learning_rate": 1.4039908887224285e-06,
+      "loss": 0.3625,
       "step": 124355
     },
     {
-      "epoch": 4.38,
-      "learning_rate": 2.0284761692286493e-06,
-      "loss": 0.2606,
+      "epoch": 4.481925973979169,
+      "grad_norm": 0.2770889401435852,
+      "learning_rate": 1.4030268927841427e-06,
+      "loss": 0.3775,
       "step": 124360
     },
     {
-      "epoch": 4.38,
-      "learning_rate": 2.027352265678986e-06,
-      "loss": 0.2531,
+      "epoch": 4.4821061736403935,
+      "grad_norm": 0.22299063205718994,
+      "learning_rate": 1.4020632183500583e-06,
+      "loss": 0.3846,
       "step": 124365
     },
     {
-      "epoch": 4.38,
-      "learning_rate": 2.0262286604140144e-06,
-      "loss": 0.2336,
+      "epoch": 4.482286373301618,
+      "grad_norm": 0.2386179119348526,
+      "learning_rate": 1.4010998654332892e-06,
+      "loss": 0.3591,
       "step": 124370
     },
     {
-      "epoch": 4.38,
-      "learning_rate": 2.0251053534483254e-06,
-      "loss": 0.2459,
+      "epoch": 4.482466572962843,
+      "grad_norm": 0.24074918031692505,
+      "learning_rate": 1.4001368340469611e-06,
+      "loss": 0.3799,
       "step": 124375
     },
     {
-      "epoch": 4.38,
-      "learning_rate": 2.0239823447964963e-06,
-      "loss": 0.2519,
+      "epoch": 4.482646772624068,
+      "grad_norm": 0.2530936896800995,
+      "learning_rate": 1.3991741242042027e-06,
+      "loss": 0.3429,
       "step": 124380
     },
     {
-      "epoch": 4.38,
-      "learning_rate": 2.0228596344731205e-06,
-      "loss": 0.2721,
+      "epoch": 4.4828269722852925,
+      "grad_norm": 0.28871747851371765,
+      "learning_rate": 1.3982117359181286e-06,
+      "loss": 0.3948,
       "step": 124385
     },
     {
-      "epoch": 4.38,
-      "learning_rate": 2.021737222492767e-06,
-      "loss": 0.2573,
+      "epoch": 4.483007171946516,
+      "grad_norm": 0.2461950182914734,
+      "learning_rate": 1.3972496692018499e-06,
+      "loss": 0.357,
       "step": 124390
     },
     {
-      "epoch": 4.38,
-      "learning_rate": 2.0206151088700137e-06,
-      "loss": 0.2838,
+      "epoch": 4.483187371607741,
+      "grad_norm": 0.21096403896808624,
+      "learning_rate": 1.3962879240684734e-06,
+      "loss": 0.3569,
       "step": 124395
     },
     {
-      "epoch": 4.38,
-      "learning_rate": 2.019493293619432e-06,
-      "loss": 0.2519,
+      "epoch": 4.483367571268966,
+      "grad_norm": 0.3317560851573944,
+      "learning_rate": 1.3953265005311021e-06,
+      "loss": 0.3861,
       "step": 124400
     },
     {
-      "epoch": 4.38,
-      "learning_rate": 2.018371776755579e-06,
-      "loss": 0.2379,
+      "epoch": 4.4835477709301905,
+      "grad_norm": 0.2790171504020691,
+      "learning_rate": 1.39436539860284e-06,
+      "loss": 0.3131,
       "step": 124405
     },
     {
-      "epoch": 4.38,
-      "learning_rate": 2.0172505582930295e-06,
-      "loss": 0.2365,
+      "epoch": 4.483727970591415,
+      "grad_norm": 0.21916787326335907,
+      "learning_rate": 1.3934046182967814e-06,
+      "loss": 0.3983,
       "step": 124410
     },
     {
-      "epoch": 4.38,
-      "learning_rate": 2.0161296382463335e-06,
-      "loss": 0.2537,
+      "epoch": 4.48390817025264,
+      "grad_norm": 0.25883591175079346,
+      "learning_rate": 1.3924441596260107e-06,
+      "loss": 0.3775,
       "step": 124415
     },
     {
-      "epoch": 4.38,
-      "learning_rate": 2.0150090166300484e-06,
-      "loss": 0.2313,
+      "epoch": 4.484088369913865,
+      "grad_norm": 0.18844369053840637,
+      "learning_rate": 1.3914840226036202e-06,
+      "loss": 0.4003,
       "step": 124420
     },
     {
-      "epoch": 4.38,
-      "learning_rate": 2.013888693458718e-06,
-      "loss": 0.2522,
+      "epoch": 4.4842685695750895,
+      "grad_norm": 0.23868656158447266,
+      "learning_rate": 1.3905242072426854e-06,
+      "loss": 0.3791,
       "step": 124425
     },
     {
-      "epoch": 4.38,
-      "learning_rate": 2.0127686687469004e-06,
-      "loss": 0.2342,
+      "epoch": 4.484448769236314,
+      "grad_norm": 0.27947574853897095,
+      "learning_rate": 1.3895647135562906e-06,
+      "loss": 0.3521,
       "step": 124430
     },
     {
-      "epoch": 4.38,
-      "learning_rate": 2.011648942509131e-06,
-      "loss": 0.2482,
+      "epoch": 4.484628968897539,
+      "grad_norm": 0.23540154099464417,
+      "learning_rate": 1.3886055415575084e-06,
+      "loss": 0.355,
       "step": 124435
     },
     {
-      "epoch": 4.38,
-      "learning_rate": 2.010529514759946e-06,
-      "loss": 0.2342,
+      "epoch": 4.484809168558763,
+      "grad_norm": 0.23615595698356628,
+      "learning_rate": 1.3876466912593976e-06,
+      "loss": 0.3757,
       "step": 124440
     },
     {
-      "epoch": 4.38,
-      "learning_rate": 2.0094103855138852e-06,
-      "loss": 0.23,
+      "epoch": 4.4849893682199875,
+      "grad_norm": 0.22659413516521454,
+      "learning_rate": 1.3866881626750345e-06,
+      "loss": 0.3436,
       "step": 124445
     },
     {
-      "epoch": 4.38,
-      "learning_rate": 2.0082915547854824e-06,
-      "loss": 0.2378,
+      "epoch": 4.485169567881212,
+      "grad_norm": 0.24826356768608093,
+      "learning_rate": 1.385729955817469e-06,
+      "loss": 0.3645,
       "step": 124450
     },
     {
-      "epoch": 4.38,
-      "learning_rate": 2.007173022589262e-06,
-      "loss": 0.2624,
+      "epoch": 4.485349767542437,
+      "grad_norm": 0.24432316422462463,
+      "learning_rate": 1.3847720706997663e-06,
+      "loss": 0.3564,
       "step": 124455
     },
     {
-      "epoch": 4.38,
-      "learning_rate": 2.0060547889397484e-06,
-      "loss": 0.2536,
+      "epoch": 4.485529967203662,
+      "grad_norm": 0.21962685883045197,
+      "learning_rate": 1.3838145073349712e-06,
+      "loss": 0.3265,
       "step": 124460
     },
     {
-      "epoch": 4.38,
-      "learning_rate": 2.004936853851455e-06,
-      "loss": 0.2516,
+      "epoch": 4.485710166864886,
+      "grad_norm": 0.28142571449279785,
+      "learning_rate": 1.382857265736126e-06,
+      "loss": 0.3877,
       "step": 124465
     },
     {
-      "epoch": 4.38,
-      "learning_rate": 2.0038192173389063e-06,
-      "loss": 0.2552,
+      "epoch": 4.485890366526111,
+      "grad_norm": 0.20428895950317383,
+      "learning_rate": 1.3819003459162816e-06,
+      "loss": 0.3767,
       "step": 124470
     },
     {
-      "epoch": 4.38,
-      "learning_rate": 2.0027018794166102e-06,
-      "loss": 0.2291,
+      "epoch": 4.486070566187336,
+      "grad_norm": 0.27324962615966797,
+      "learning_rate": 1.3809437478884719e-06,
+      "loss": 0.3281,
       "step": 124475
     },
     {
-      "epoch": 4.38,
-      "learning_rate": 2.0015848400990716e-06,
-      "loss": 0.257,
+      "epoch": 4.48625076584856,
+      "grad_norm": 0.2558141350746155,
+      "learning_rate": 1.3799874716657307e-06,
+      "loss": 0.384,
       "step": 124480
     },
     {
-      "epoch": 4.38,
-      "learning_rate": 2.000468099400796e-06,
-      "loss": 0.2471,
+      "epoch": 4.4864309655097845,
+      "grad_norm": 0.24941542744636536,
+      "learning_rate": 1.3790315172610896e-06,
+      "loss": 0.3501,
       "step": 124485
     },
     {
-      "epoch": 4.38,
-      "learning_rate": 1.9993516573362904e-06,
-      "loss": 0.2406,
+      "epoch": 4.486611165171009,
+      "grad_norm": 0.2563149333000183,
+      "learning_rate": 1.3780758846875658e-06,
+      "loss": 0.3841,
       "step": 124490
     },
     {
-      "epoch": 4.38,
-      "learning_rate": 1.998235513920038e-06,
-      "loss": 0.2725,
+      "epoch": 4.486791364832234,
+      "grad_norm": 0.2376670092344284,
+      "learning_rate": 1.3771205739581877e-06,
+      "loss": 0.3725,
       "step": 124495
     },
     {
-      "epoch": 4.38,
-      "learning_rate": 1.9971196691665415e-06,
-      "loss": 0.245,
+      "epoch": 4.486971564493459,
+      "grad_norm": 0.30792880058288574,
+      "learning_rate": 1.37616558508597e-06,
+      "loss": 0.3779,
       "step": 124500
     },
     {
-      "epoch": 4.38,
-      "eval_loss": 0.24900418519973755,
-      "eval_runtime": 10.5316,
-      "eval_samples_per_second": 9.495,
-      "eval_steps_per_second": 9.495,
+      "epoch": 4.486971564493459,
+      "eval_loss": 0.42892351746559143,
+      "eval_runtime": 3.5425,
+      "eval_samples_per_second": 28.229,
+      "eval_steps_per_second": 7.057,
       "step": 124500
     },
     {
-      "epoch": 4.38,
-      "learning_rate": 1.9960041230902893e-06,
-      "loss": 0.2467,
+      "epoch": 4.487151764154683,
+      "grad_norm": 0.28269314765930176,
+      "learning_rate": 1.3752109180839217e-06,
+      "loss": 0.4113,
       "step": 124505
     },
     {
-      "epoch": 4.38,
-      "learning_rate": 1.9948888757057637e-06,
-      "loss": 0.2583,
+      "epoch": 4.487331963815908,
+      "grad_norm": 0.27605292201042175,
+      "learning_rate": 1.3742565729650491e-06,
+      "loss": 0.3593,
       "step": 124510
     },
     {
-      "epoch": 4.38,
-      "learning_rate": 1.9937739270274454e-06,
-      "loss": 0.2436,
+      "epoch": 4.487512163477133,
+      "grad_norm": 0.2970362901687622,
+      "learning_rate": 1.3733025497423586e-06,
+      "loss": 0.3946,
       "step": 124515
     },
     {
-      "epoch": 4.38,
-      "learning_rate": 1.9926592770698057e-06,
-      "loss": 0.2666,
+      "epoch": 4.487692363138358,
+      "grad_norm": 0.285276859998703,
+      "learning_rate": 1.3723488484288421e-06,
+      "loss": 0.3863,
       "step": 124520
     },
     {
-      "epoch": 4.38,
-      "learning_rate": 1.9915449258473274e-06,
-      "loss": 0.2608,
+      "epoch": 4.487872562799582,
+      "grad_norm": 0.2632303535938263,
+      "learning_rate": 1.3713954690375035e-06,
+      "loss": 0.4355,
       "step": 124525
     },
     {
-      "epoch": 4.38,
-      "learning_rate": 1.990430873374477e-06,
-      "loss": 0.2429,
+      "epoch": 4.488052762460806,
+      "grad_norm": 0.3081340193748474,
+      "learning_rate": 1.370442411581324e-06,
+      "loss": 0.3526,
       "step": 124530
     },
     {
-      "epoch": 4.38,
-      "learning_rate": 1.989317119665715e-06,
-      "loss": 0.2524,
+      "epoch": 4.488232962122031,
+      "grad_norm": 0.2594117820262909,
+      "learning_rate": 1.3694896760732933e-06,
+      "loss": 0.3852,
       "step": 124535
     },
     {
-      "epoch": 4.38,
-      "learning_rate": 1.9882036647355025e-06,
-      "loss": 0.2692,
+      "epoch": 4.488413161783256,
+      "grad_norm": 0.30716049671173096,
+      "learning_rate": 1.3685372625263899e-06,
+      "loss": 0.3624,
       "step": 124540
     },
     {
-      "epoch": 4.38,
-      "learning_rate": 1.987090508598305e-06,
-      "loss": 0.2596,
+      "epoch": 4.48859336144448,
+      "grad_norm": 0.23581121861934662,
+      "learning_rate": 1.367585170953589e-06,
+      "loss": 0.3477,
       "step": 124545
     },
     {
-      "epoch": 4.38,
-      "learning_rate": 1.9859776512685667e-06,
-      "loss": 0.2396,
+      "epoch": 4.488773561105705,
+      "grad_norm": 0.24266508221626282,
+      "learning_rate": 1.3666334013678729e-06,
+      "loss": 0.4177,
       "step": 124550
     },
     {
-      "epoch": 4.38,
-      "learning_rate": 1.984865092760746e-06,
-      "loss": 0.2326,
+      "epoch": 4.48895376076693,
+      "grad_norm": 0.22927452623844147,
+      "learning_rate": 1.365681953782194e-06,
+      "loss": 0.3946,
       "step": 124555
     },
     {
-      "epoch": 4.38,
-      "learning_rate": 1.983752833089278e-06,
-      "loss": 0.2454,
+      "epoch": 4.489133960428155,
+      "grad_norm": 0.24199864268302917,
+      "learning_rate": 1.3647308282095206e-06,
+      "loss": 0.3845,
       "step": 124560
     },
     {
-      "epoch": 4.38,
-      "learning_rate": 1.9826408722686178e-06,
-      "loss": 0.2596,
+      "epoch": 4.489314160089379,
+      "grad_norm": 0.2787548303604126,
+      "learning_rate": 1.3637800246628197e-06,
+      "loss": 0.385,
       "step": 124565
     },
     {
-      "epoch": 4.38,
-      "learning_rate": 1.9815292103131988e-06,
-      "loss": 0.273,
+      "epoch": 4.489494359750604,
+      "grad_norm": 0.2417118102312088,
+      "learning_rate": 1.3628295431550365e-06,
+      "loss": 0.3755,
       "step": 124570
     },
     {
-      "epoch": 4.38,
-      "learning_rate": 1.9804178472374506e-06,
-      "loss": 0.2503,
+      "epoch": 4.489674559411828,
+      "grad_norm": 0.2551795542240143,
+      "learning_rate": 1.3618793836991273e-06,
+      "loss": 0.3499,
       "step": 124575
     },
     {
-      "epoch": 4.38,
-      "learning_rate": 1.979306783055801e-06,
-      "loss": 0.248,
+      "epoch": 4.489854759073053,
+      "grad_norm": 0.3174431324005127,
+      "learning_rate": 1.3609295463080318e-06,
+      "loss": 0.3807,
       "step": 124580
     },
     {
-      "epoch": 4.38,
-      "learning_rate": 1.978196017782688e-06,
-      "loss": 0.2314,
+      "epoch": 4.490034958734277,
+      "grad_norm": 0.23123787343502045,
+      "learning_rate": 1.3599800309946925e-06,
+      "loss": 0.3648,
       "step": 124585
     },
     {
-      "epoch": 4.38,
-      "learning_rate": 1.9770855514325284e-06,
-      "loss": 0.2342,
+      "epoch": 4.490215158395502,
+      "grad_norm": 0.27223360538482666,
+      "learning_rate": 1.3590308377720546e-06,
+      "loss": 0.3421,
       "step": 124590
     },
     {
-      "epoch": 4.38,
-      "learning_rate": 1.9759753840197374e-06,
-      "loss": 0.2402,
+      "epoch": 4.490395358056727,
+      "grad_norm": 0.25846990942955017,
+      "learning_rate": 1.3580819666530408e-06,
+      "loss": 0.3496,
       "step": 124595
     },
     {
-      "epoch": 4.38,
-      "learning_rate": 1.97486551555873e-06,
-      "loss": 0.2625,
+      "epoch": 4.490575557717952,
+      "grad_norm": 0.25922471284866333,
+      "learning_rate": 1.3571334176505856e-06,
+      "loss": 0.3783,
       "step": 124600
     },
     {
-      "epoch": 4.38,
-      "learning_rate": 1.973755946063918e-06,
-      "loss": 0.2473,
+      "epoch": 4.490755757379176,
+      "grad_norm": 0.2560346722602844,
+      "learning_rate": 1.3561851907776064e-06,
+      "loss": 0.3274,
       "step": 124605
     },
     {
-      "epoch": 4.38,
-      "learning_rate": 1.9726466755497163e-06,
-      "loss": 0.2347,
+      "epoch": 4.490935957040401,
+      "grad_norm": 0.25044140219688416,
+      "learning_rate": 1.355237286047026e-06,
+      "loss": 0.377,
       "step": 124610
     },
     {
-      "epoch": 4.38,
-      "learning_rate": 1.9715377040305182e-06,
-      "loss": 0.2508,
+      "epoch": 4.491116156701626,
+      "grad_norm": 0.25878745317459106,
+      "learning_rate": 1.3542897034717644e-06,
+      "loss": 0.3426,
       "step": 124615
     },
     {
-      "epoch": 4.38,
-      "learning_rate": 1.970429031520729e-06,
-      "loss": 0.2427,
+      "epoch": 4.49129635636285,
+      "grad_norm": 0.23946064710617065,
+      "learning_rate": 1.3533424430647285e-06,
+      "loss": 0.3811,
       "step": 124620
     },
     {
-      "epoch": 4.38,
-      "learning_rate": 1.969320658034743e-06,
-      "loss": 0.2439,
+      "epoch": 4.491476556024074,
+      "grad_norm": 0.282932311296463,
+      "learning_rate": 1.3523955048388188e-06,
+      "loss": 0.3618,
       "step": 124625
     },
     {
-      "epoch": 4.38,
-      "learning_rate": 1.9682125835869507e-06,
-      "loss": 0.249,
+      "epoch": 4.491656755685299,
+      "grad_norm": 0.2500767707824707,
+      "learning_rate": 1.3514488888069443e-06,
+      "loss": 0.3679,
       "step": 124630
     },
     {
-      "epoch": 4.39,
-      "learning_rate": 1.967104808191736e-06,
-      "loss": 0.2222,
+      "epoch": 4.491836955346524,
+      "grad_norm": 0.2235366553068161,
+      "learning_rate": 1.3505025949819978e-06,
+      "loss": 0.3695,
       "step": 124635
     },
     {
-      "epoch": 4.39,
-      "learning_rate": 1.9659973318634896e-06,
-      "loss": 0.2549,
+      "epoch": 4.492017155007749,
+      "grad_norm": 0.31541529297828674,
+      "learning_rate": 1.3495566233768742e-06,
+      "loss": 0.4089,
       "step": 124640
     },
     {
-      "epoch": 4.39,
-      "learning_rate": 1.964890154616586e-06,
-      "loss": 0.2388,
+      "epoch": 4.492197354668973,
+      "grad_norm": 0.3026379942893982,
+      "learning_rate": 1.3486109740044688e-06,
+      "loss": 0.3709,
       "step": 124645
     },
     {
-      "epoch": 4.39,
-      "learning_rate": 1.9637832764654057e-06,
-      "loss": 0.27,
+      "epoch": 4.492377554330198,
+      "grad_norm": 0.26099175214767456,
+      "learning_rate": 1.347665646877655e-06,
+      "loss": 0.3478,
       "step": 124650
     },
     {
-      "epoch": 4.39,
-      "learning_rate": 1.962676697424312e-06,
-      "loss": 0.2355,
+      "epoch": 4.492557753991423,
+      "grad_norm": 0.2050563395023346,
+      "learning_rate": 1.3467206420093191e-06,
+      "loss": 0.367,
       "step": 124655
     },
     {
-      "epoch": 4.39,
-      "learning_rate": 1.96157041750768e-06,
-      "loss": 0.2612,
+      "epoch": 4.4927379536526475,
+      "grad_norm": 0.26844698190689087,
+      "learning_rate": 1.3457759594123348e-06,
+      "loss": 0.3862,
       "step": 124660
     },
     {
-      "epoch": 4.39,
-      "learning_rate": 1.960464436729878e-06,
-      "loss": 0.2444,
+      "epoch": 4.492918153313871,
+      "grad_norm": 0.2691669762134552,
+      "learning_rate": 1.3448315990995746e-06,
+      "loss": 0.3652,
       "step": 124665
     },
     {
-      "epoch": 4.39,
-      "learning_rate": 1.959358755105262e-06,
-      "loss": 0.2326,
+      "epoch": 4.493098352975096,
+      "grad_norm": 0.26715192198753357,
+      "learning_rate": 1.3438875610839035e-06,
+      "loss": 0.3557,
       "step": 124670
     },
     {
-      "epoch": 4.39,
-      "learning_rate": 1.9582533726481833e-06,
-      "loss": 0.2414,
+      "epoch": 4.493278552636321,
+      "grad_norm": 0.27473947405815125,
+      "learning_rate": 1.3429438453781807e-06,
+      "loss": 0.3646,
       "step": 124675
     },
     {
-      "epoch": 4.39,
-      "learning_rate": 1.9571482893730007e-06,
-      "loss": 0.2675,
+      "epoch": 4.4934587522975455,
+      "grad_norm": 0.272559255361557,
+      "learning_rate": 1.3420004519952706e-06,
+      "loss": 0.3924,
       "step": 124680
     },
     {
-      "epoch": 4.39,
-      "learning_rate": 1.9560435052940667e-06,
-      "loss": 0.2569,
+      "epoch": 4.49363895195877,
+      "grad_norm": 0.2625700831413269,
+      "learning_rate": 1.3410573809480242e-06,
+      "loss": 0.4042,
       "step": 124685
     },
     {
-      "epoch": 4.39,
-      "learning_rate": 1.9549390204257194e-06,
-      "loss": 0.2463,
+      "epoch": 4.493819151619995,
+      "grad_norm": 0.2726500332355499,
+      "learning_rate": 1.3401146322492924e-06,
+      "loss": 0.3691,
       "step": 124690
     },
     {
-      "epoch": 4.39,
-      "learning_rate": 1.953834834782295e-06,
-      "loss": 0.2412,
+      "epoch": 4.49399935128122,
+      "grad_norm": 0.23851507902145386,
+      "learning_rate": 1.339172205911915e-06,
+      "loss": 0.3906,
       "step": 124695
     },
     {
-      "epoch": 4.39,
-      "learning_rate": 1.952730948378145e-06,
-      "loss": 0.2457,
+      "epoch": 4.4941795509424445,
+      "grad_norm": 0.212006077170372,
+      "learning_rate": 1.338230101948737e-06,
+      "loss": 0.3953,
       "step": 124700
     },
     {
-      "epoch": 4.39,
-      "learning_rate": 1.951627361227593e-06,
-      "loss": 0.2618,
+      "epoch": 4.494359750603669,
+      "grad_norm": 0.24577181041240692,
+      "learning_rate": 1.33728832037259e-06,
+      "loss": 0.354,
       "step": 124705
     },
     {
-      "epoch": 4.39,
-      "learning_rate": 1.950524073344967e-06,
-      "loss": 0.2526,
+      "epoch": 4.494539950264894,
+      "grad_norm": 0.24666406214237213,
+      "learning_rate": 1.336346861196311e-06,
+      "loss": 0.3328,
       "step": 124710
     },
     {
-      "epoch": 4.39,
-      "learning_rate": 1.949421084744599e-06,
-      "loss": 0.2455,
+      "epoch": 4.494720149926118,
+      "grad_norm": 0.22719623148441315,
+      "learning_rate": 1.3354057244327256e-06,
+      "loss": 0.3629,
       "step": 124715
     },
     {
-      "epoch": 4.39,
-      "learning_rate": 1.948318395440804e-06,
-      "loss": 0.2467,
+      "epoch": 4.4949003495873425,
+      "grad_norm": 0.24336789548397064,
+      "learning_rate": 1.3344649100946543e-06,
+      "loss": 0.3702,
       "step": 124720
     },
     {
-      "epoch": 4.39,
-      "learning_rate": 1.9472160054479077e-06,
-      "loss": 0.2602,
+      "epoch": 4.495080549248567,
+      "grad_norm": 0.2589198350906372,
+      "learning_rate": 1.3335244181949174e-06,
+      "loss": 0.3501,
       "step": 124725
     },
     {
-      "epoch": 4.39,
-      "learning_rate": 1.946113914780212e-06,
-      "loss": 0.2352,
+      "epoch": 4.495260748909792,
+      "grad_norm": 0.2021062821149826,
+      "learning_rate": 1.3325842487463241e-06,
+      "loss": 0.3659,
       "step": 124730
     },
     {
-      "epoch": 4.39,
-      "learning_rate": 1.945012123452042e-06,
-      "loss": 0.2524,
+      "epoch": 4.495440948571017,
+      "grad_norm": 0.22475126385688782,
+      "learning_rate": 1.3316444017616975e-06,
+      "loss": 0.3618,
       "step": 124735
     },
     {
-      "epoch": 4.39,
-      "learning_rate": 1.943910631477694e-06,
-      "loss": 0.245,
+      "epoch": 4.4956211482322415,
+      "grad_norm": 0.29188576340675354,
+      "learning_rate": 1.3307048772538272e-06,
+      "loss": 0.3786,
       "step": 124740
     },
     {
-      "epoch": 4.39,
-      "learning_rate": 1.9428094388714707e-06,
-      "loss": 0.2418,
+      "epoch": 4.495801347893466,
+      "grad_norm": 0.2823890149593353,
+      "learning_rate": 1.3297656752355197e-06,
+      "loss": 0.3519,
       "step": 124745
     },
     {
-      "epoch": 4.39,
-      "learning_rate": 1.9417085456476693e-06,
-      "loss": 0.2375,
+      "epoch": 4.495981547554691,
+      "grad_norm": 0.2329128384590149,
+      "learning_rate": 1.3288267957195733e-06,
+      "loss": 0.3333,
       "step": 124750
     },
     {
-      "epoch": 4.39,
-      "learning_rate": 1.940607951820592e-06,
-      "loss": 0.2728,
+      "epoch": 4.496161747215915,
+      "grad_norm": 0.3408520817756653,
+      "learning_rate": 1.3278882387187775e-06,
+      "loss": 0.3682,
       "step": 124755
     },
     {
-      "epoch": 4.39,
-      "learning_rate": 1.9395076574045213e-06,
-      "loss": 0.2674,
+      "epoch": 4.4963419468771395,
+      "grad_norm": 0.2713524103164673,
+      "learning_rate": 1.3269500042459276e-06,
+      "loss": 0.3608,
       "step": 124760
     },
     {
-      "epoch": 4.39,
-      "learning_rate": 1.938407662413741e-06,
-      "loss": 0.2358,
+      "epoch": 4.496522146538364,
+      "grad_norm": 0.2413674145936966,
+      "learning_rate": 1.3260120923137943e-06,
+      "loss": 0.3557,
       "step": 124765
     },
     {
-      "epoch": 4.39,
-      "learning_rate": 1.937307966862545e-06,
-      "loss": 0.2655,
+      "epoch": 4.496702346199589,
+      "grad_norm": 0.29502227902412415,
+      "learning_rate": 1.325074502935164e-06,
+      "loss": 0.3778,
       "step": 124770
     },
     {
-      "epoch": 4.39,
-      "learning_rate": 1.936208570765202e-06,
-      "loss": 0.2538,
+      "epoch": 4.496882545860814,
+      "grad_norm": 0.21458546817302704,
+      "learning_rate": 1.3241372361228104e-06,
+      "loss": 0.3857,
       "step": 124775
     },
     {
-      "epoch": 4.39,
-      "learning_rate": 1.935109474135996e-06,
-      "loss": 0.2464,
+      "epoch": 4.497062745522038,
+      "grad_norm": 0.241383358836174,
+      "learning_rate": 1.3232002918895036e-06,
+      "loss": 0.4047,
       "step": 124780
     },
     {
-      "epoch": 4.39,
-      "learning_rate": 1.9340106769891924e-06,
-      "loss": 0.2625,
+      "epoch": 4.497242945183263,
+      "grad_norm": 0.2607963979244232,
+      "learning_rate": 1.3222636702480084e-06,
+      "loss": 0.3352,
       "step": 124785
     },
     {
-      "epoch": 4.39,
-      "learning_rate": 1.9329121793390554e-06,
-      "loss": 0.2676,
+      "epoch": 4.497423144844488,
+      "grad_norm": 0.24018988013267517,
+      "learning_rate": 1.3213273712110895e-06,
+      "loss": 0.3648,
       "step": 124790
     },
     {
-      "epoch": 4.39,
-      "learning_rate": 1.9318139811998537e-06,
-      "loss": 0.2386,
+      "epoch": 4.497603344505713,
+      "grad_norm": 0.22829695045948029,
+      "learning_rate": 1.3203913947914953e-06,
+      "loss": 0.373,
       "step": 124795
     },
     {
-      "epoch": 4.39,
-      "learning_rate": 1.9307160825858483e-06,
-      "loss": 0.2545,
+      "epoch": 4.497783544166937,
+      "grad_norm": 0.22519542276859283,
+      "learning_rate": 1.3194557410019875e-06,
+      "loss": 0.3512,
       "step": 124800
     },
     {
-      "epoch": 4.39,
-      "learning_rate": 1.929618483511289e-06,
-      "loss": 0.2529,
+      "epoch": 4.497963743828161,
+      "grad_norm": 0.2739558219909668,
+      "learning_rate": 1.3185204098553089e-06,
+      "loss": 0.3571,
       "step": 124805
     },
     {
-      "epoch": 4.39,
-      "learning_rate": 1.9285211839904256e-06,
-      "loss": 0.2464,
+      "epoch": 4.498143943489386,
+      "grad_norm": 0.29086029529571533,
+      "learning_rate": 1.3175854013642074e-06,
+      "loss": 0.3761,
       "step": 124810
     },
     {
-      "epoch": 4.39,
-      "learning_rate": 1.9274241840375155e-06,
-      "loss": 0.2499,
+      "epoch": 4.498324143150611,
+      "grad_norm": 0.2614690959453583,
+      "learning_rate": 1.3166507155414176e-06,
+      "loss": 0.3772,
       "step": 124815
     },
     {
-      "epoch": 4.39,
-      "learning_rate": 1.9263274836667927e-06,
-      "loss": 0.2586,
+      "epoch": 4.498504342811835,
+      "grad_norm": 0.2470005601644516,
+      "learning_rate": 1.315716352399679e-06,
+      "loss": 0.3845,
       "step": 124820
     },
     {
-      "epoch": 4.39,
-      "learning_rate": 1.9252310828925035e-06,
-      "loss": 0.2359,
+      "epoch": 4.49868454247306,
+      "grad_norm": 0.22588050365447998,
+      "learning_rate": 1.3147823119517122e-06,
+      "loss": 0.3627,
       "step": 124825
     },
     {
-      "epoch": 4.39,
-      "learning_rate": 1.9241349817288812e-06,
-      "loss": 0.2371,
+      "epoch": 4.498864742134285,
+      "grad_norm": 0.187801331281662,
+      "learning_rate": 1.3138485942102625e-06,
+      "loss": 0.3807,
       "step": 124830
     },
     {
-      "epoch": 4.39,
-      "learning_rate": 1.9230391801901588e-06,
-      "loss": 0.2598,
+      "epoch": 4.49904494179551,
+      "grad_norm": 0.23127955198287964,
+      "learning_rate": 1.3129151991880307e-06,
+      "loss": 0.3522,
       "step": 124835
     },
     {
-      "epoch": 4.39,
-      "learning_rate": 1.9219436782905666e-06,
-      "loss": 0.2648,
+      "epoch": 4.499225141456734,
+      "grad_norm": 0.20928794145584106,
+      "learning_rate": 1.3119821268977455e-06,
+      "loss": 0.3553,
       "step": 124840
     },
     {
-      "epoch": 4.39,
-      "learning_rate": 1.9208484760443235e-06,
-      "loss": 0.2486,
+      "epoch": 4.499405341117959,
+      "grad_norm": 0.2646942138671875,
+      "learning_rate": 1.3110493773521193e-06,
+      "loss": 0.3659,
       "step": 124845
     },
     {
-      "epoch": 4.39,
-      "learning_rate": 1.9197535734656545e-06,
-      "loss": 0.2699,
+      "epoch": 4.499585540779183,
+      "grad_norm": 0.2235756665468216,
+      "learning_rate": 1.3101169505638582e-06,
+      "loss": 0.3924,
       "step": 124850
     },
     {
-      "epoch": 4.39,
-      "learning_rate": 1.9186589705687762e-06,
-      "loss": 0.2401,
+      "epoch": 4.499765740440408,
+      "grad_norm": 0.2302202582359314,
+      "learning_rate": 1.3091848465456692e-06,
+      "loss": 0.3704,
       "step": 124855
     },
     {
-      "epoch": 4.39,
-      "learning_rate": 1.917564667367902e-06,
-      "loss": 0.2449,
+      "epoch": 4.499945940101632,
+      "grad_norm": 0.2802170515060425,
+      "learning_rate": 1.3082530653102444e-06,
+      "loss": 0.3699,
       "step": 124860
     },
     {
-      "epoch": 4.39,
-      "learning_rate": 1.9164706638772316e-06,
-      "loss": 0.236,
+      "epoch": 4.500126139762857,
+      "grad_norm": 0.2077890932559967,
+      "learning_rate": 1.3073216068702876e-06,
+      "loss": 0.3576,
       "step": 124865
     },
     {
-      "epoch": 4.39,
-      "learning_rate": 1.9153769601109844e-06,
-      "loss": 0.26,
+      "epoch": 4.500306339424082,
+      "grad_norm": 0.2789739966392517,
+      "learning_rate": 1.3063904712384889e-06,
+      "loss": 0.3412,
       "step": 124870
     },
     {
-      "epoch": 4.39,
-      "learning_rate": 1.9142835560833515e-06,
-      "loss": 0.2426,
+      "epoch": 4.500486539085307,
+      "grad_norm": 0.2757323086261749,
+      "learning_rate": 1.3054596584275325e-06,
+      "loss": 0.3565,
       "step": 124875
     },
     {
-      "epoch": 4.39,
-      "learning_rate": 1.9131904518085276e-06,
-      "loss": 0.2414,
+      "epoch": 4.500666738746531,
+      "grad_norm": 0.18960040807724,
+      "learning_rate": 1.3045291684500999e-06,
+      "loss": 0.3404,
       "step": 124880
     },
     {
-      "epoch": 4.39,
-      "learning_rate": 1.912097647300715e-06,
-      "loss": 0.2513,
+      "epoch": 4.500846938407756,
+      "grad_norm": 0.23652799427509308,
+      "learning_rate": 1.3035990013188698e-06,
+      "loss": 0.3883,
       "step": 124885
     },
     {
-      "epoch": 4.39,
-      "learning_rate": 1.911005142574099e-06,
-      "loss": 0.2624,
+      "epoch": 4.501027138068981,
+      "grad_norm": 0.2553980350494385,
+      "learning_rate": 1.3026691570465128e-06,
+      "loss": 0.3724,
       "step": 124890
     },
     {
-      "epoch": 4.39,
-      "learning_rate": 1.9099129376428658e-06,
-      "loss": 0.2521,
+      "epoch": 4.5012073377302055,
+      "grad_norm": 0.28166186809539795,
+      "learning_rate": 1.3017396356457045e-06,
+      "loss": 0.3743,
       "step": 124895
     },
     {
-      "epoch": 4.39,
-      "learning_rate": 1.908821032521196e-06,
-      "loss": 0.2464,
+      "epoch": 4.501387537391429,
+      "grad_norm": 0.25217875838279724,
+      "learning_rate": 1.3008104371291047e-06,
+      "loss": 0.3456,
       "step": 124900
     },
     {
-      "epoch": 4.39,
-      "learning_rate": 1.907729427223265e-06,
-      "loss": 0.2497,
+      "epoch": 4.501567737052654,
+      "grad_norm": 0.22290779650211334,
+      "learning_rate": 1.299881561509375e-06,
+      "loss": 0.3797,
       "step": 124905
     },
     {
-      "epoch": 4.39,
-      "learning_rate": 1.9066381217632516e-06,
-      "loss": 0.2499,
+      "epoch": 4.501747936713879,
+      "grad_norm": 0.25568974018096924,
+      "learning_rate": 1.2989530087991696e-06,
+      "loss": 0.3846,
       "step": 124910
     },
     {
-      "epoch": 4.39,
-      "learning_rate": 1.9055471161553262e-06,
-      "loss": 0.2384,
+      "epoch": 4.501928136375104,
+      "grad_norm": 0.23674674332141876,
+      "learning_rate": 1.2980247790111367e-06,
+      "loss": 0.4019,
       "step": 124915
     },
     {
-      "epoch": 4.4,
-      "learning_rate": 1.9044564104136492e-06,
-      "loss": 0.2524,
+      "epoch": 4.502108336036328,
+      "grad_norm": 0.22385911643505096,
+      "learning_rate": 1.2970968721579324e-06,
+      "loss": 0.3634,
       "step": 124920
     },
     {
-      "epoch": 4.4,
-      "learning_rate": 1.9033660045523788e-06,
-      "loss": 0.2348,
+      "epoch": 4.502288535697553,
+      "grad_norm": 0.3423471748828888,
+      "learning_rate": 1.296169288252197e-06,
+      "loss": 0.3883,
       "step": 124925
     },
     {
-      "epoch": 4.4,
-      "learning_rate": 1.902275898585687e-06,
-      "loss": 0.2543,
+      "epoch": 4.502468735358778,
+      "grad_norm": 0.29338136315345764,
+      "learning_rate": 1.295242027306559e-06,
+      "loss": 0.3775,
       "step": 124930
     },
     {
-      "epoch": 4.4,
-      "learning_rate": 1.9011860925277153e-06,
-      "loss": 0.2533,
+      "epoch": 4.5026489350200025,
+      "grad_norm": 0.17532098293304443,
+      "learning_rate": 1.2943150893336614e-06,
+      "loss": 0.343,
       "step": 124935
     },
     {
-      "epoch": 4.4,
-      "learning_rate": 1.9000965863926245e-06,
-      "loss": 0.231,
+      "epoch": 4.502829134681226,
+      "grad_norm": 0.27964937686920166,
+      "learning_rate": 1.29338847434613e-06,
+      "loss": 0.3547,
       "step": 124940
     },
     {
-      "epoch": 4.4,
-      "learning_rate": 1.8990073801945508e-06,
-      "loss": 0.2458,
+      "epoch": 4.503009334342451,
+      "grad_norm": 0.2344919592142105,
+      "learning_rate": 1.2924621823565964e-06,
+      "loss": 0.3394,
       "step": 124945
     },
     {
-      "epoch": 4.4,
-      "learning_rate": 1.8979184739476463e-06,
-      "loss": 0.2199,
+      "epoch": 4.503189534003676,
+      "grad_norm": 0.2620304226875305,
+      "learning_rate": 1.2915362133776725e-06,
+      "loss": 0.3778,
       "step": 124950
     },
     {
-      "epoch": 4.4,
-      "learning_rate": 1.8968298676660472e-06,
-      "loss": 0.2734,
+      "epoch": 4.503369733664901,
+      "grad_norm": 0.26436105370521545,
+      "learning_rate": 1.2906105674219737e-06,
+      "loss": 0.3555,
       "step": 124955
     },
     {
-      "epoch": 4.4,
-      "learning_rate": 1.8957415613638813e-06,
-      "loss": 0.2508,
+      "epoch": 4.503549933326125,
+      "grad_norm": 0.25804218649864197,
+      "learning_rate": 1.2896852445021173e-06,
+      "loss": 0.361,
       "step": 124960
     },
     {
-      "epoch": 4.4,
-      "learning_rate": 1.8946535550552897e-06,
-      "loss": 0.2559,
+      "epoch": 4.50373013298735,
+      "grad_norm": 0.2341192215681076,
+      "learning_rate": 1.2887602446307128e-06,
+      "loss": 0.3672,
       "step": 124965
     },
     {
-      "epoch": 4.4,
-      "learning_rate": 1.8935658487543944e-06,
-      "loss": 0.2305,
+      "epoch": 4.503910332648575,
+      "grad_norm": 0.26045674085617065,
+      "learning_rate": 1.2878355678203558e-06,
+      "loss": 0.3609,
       "step": 124970
     },
     {
-      "epoch": 4.4,
-      "learning_rate": 1.8924784424753179e-06,
-      "loss": 0.2515,
+      "epoch": 4.5040905323097995,
+      "grad_norm": 0.2202635258436203,
+      "learning_rate": 1.2869112140836525e-06,
+      "loss": 0.38,
       "step": 124975
     },
     {
-      "epoch": 4.4,
-      "learning_rate": 1.891391336232176e-06,
-      "loss": 0.2453,
+      "epoch": 4.504270731971024,
+      "grad_norm": 0.2518618404865265,
+      "learning_rate": 1.285987183433185e-06,
+      "loss": 0.3294,
       "step": 124980
     },
     {
-      "epoch": 4.4,
-      "learning_rate": 1.8903045300390942e-06,
-      "loss": 0.244,
+      "epoch": 4.504450931632249,
+      "grad_norm": 0.22585217654705048,
+      "learning_rate": 1.2850634758815567e-06,
+      "loss": 0.3649,
       "step": 124985
     },
     {
-      "epoch": 4.4,
-      "learning_rate": 1.8892180239101747e-06,
-      "loss": 0.2466,
+      "epoch": 4.504631131293473,
+      "grad_norm": 0.24542665481567383,
+      "learning_rate": 1.2841400914413465e-06,
+      "loss": 0.3599,
       "step": 124990
     },
     {
-      "epoch": 4.4,
-      "learning_rate": 1.8881318178595314e-06,
-      "loss": 0.2572,
+      "epoch": 4.5048113309546975,
+      "grad_norm": 0.23829962313175201,
+      "learning_rate": 1.2832170301251362e-06,
+      "loss": 0.379,
       "step": 124995
     },
     {
-      "epoch": 4.4,
-      "learning_rate": 1.8870459119012612e-06,
-      "loss": 0.2675,
+      "epoch": 4.504991530615922,
+      "grad_norm": 0.30551934242248535,
+      "learning_rate": 1.2822942919455045e-06,
+      "loss": 0.4066,
       "step": 125000
     },
     {
-      "epoch": 4.4,
-      "eval_loss": 0.24892328679561615,
-      "eval_runtime": 10.538,
-      "eval_samples_per_second": 9.489,
-      "eval_steps_per_second": 9.489,
+      "epoch": 4.504991530615922,
+      "eval_loss": 0.42903777956962585,
+      "eval_runtime": 3.5368,
+      "eval_samples_per_second": 28.274,
+      "eval_steps_per_second": 7.068,
       "step": 125000
     },
     {
-      "epoch": 4.4,
-      "learning_rate": 1.8859603060494695e-06,
-      "loss": 0.2666,
+      "epoch": 4.505171730277147,
+      "grad_norm": 0.2084054797887802,
+      "learning_rate": 1.281371876915019e-06,
+      "loss": 0.351,
       "step": 125005
     },
     {
-      "epoch": 4.4,
-      "learning_rate": 1.8848750003182537e-06,
-      "loss": 0.2405,
+      "epoch": 4.505351929938372,
+      "grad_norm": 0.20273062586784363,
+      "learning_rate": 1.280449785046245e-06,
+      "loss": 0.366,
       "step": 125010
     },
     {
-      "epoch": 4.4,
-      "learning_rate": 1.883789994721702e-06,
-      "loss": 0.2561,
+      "epoch": 4.5055321295995965,
+      "grad_norm": 0.2866862714290619,
+      "learning_rate": 1.2795280163517581e-06,
+      "loss": 0.413,
       "step": 125015
     },
     {
-      "epoch": 4.4,
-      "learning_rate": 1.8827052892738979e-06,
-      "loss": 0.2539,
+      "epoch": 4.505712329260821,
+      "grad_norm": 0.2754046618938446,
+      "learning_rate": 1.2786065708441042e-06,
+      "loss": 0.3871,
       "step": 125020
     },
     {
-      "epoch": 4.4,
-      "learning_rate": 1.8816208839889382e-06,
-      "loss": 0.2762,
+      "epoch": 4.505892528922046,
+      "grad_norm": 0.2317970097064972,
+      "learning_rate": 1.2776854485358453e-06,
+      "loss": 0.3907,
       "step": 125025
     },
     {
-      "epoch": 4.4,
-      "learning_rate": 1.8805367788808925e-06,
-      "loss": 0.2596,
+      "epoch": 4.50607272858327,
+      "grad_norm": 0.19607658684253693,
+      "learning_rate": 1.2767646494395296e-06,
+      "loss": 0.3727,
       "step": 125030
     },
     {
-      "epoch": 4.4,
-      "learning_rate": 1.879452973963841e-06,
-      "loss": 0.2361,
+      "epoch": 4.5062529282444945,
+      "grad_norm": 0.18511025607585907,
+      "learning_rate": 1.2758441735677028e-06,
+      "loss": 0.3539,
       "step": 125035
     },
     {
-      "epoch": 4.4,
-      "learning_rate": 1.8783694692518506e-06,
-      "loss": 0.2608,
+      "epoch": 4.506433127905719,
+      "grad_norm": 0.27356868982315063,
+      "learning_rate": 1.2749240209329049e-06,
+      "loss": 0.3884,
       "step": 125040
     },
     {
-      "epoch": 4.4,
-      "learning_rate": 1.8772862647589957e-06,
-      "loss": 0.2547,
+      "epoch": 4.506613327566944,
+      "grad_norm": 0.22257938981056213,
+      "learning_rate": 1.274004191547673e-06,
+      "loss": 0.3783,
       "step": 125045
     },
     {
-      "epoch": 4.4,
-      "learning_rate": 1.8762033604993457e-06,
-      "loss": 0.2269,
+      "epoch": 4.506793527228169,
+      "grad_norm": 0.28302866220474243,
+      "learning_rate": 1.2730846854245416e-06,
+      "loss": 0.3706,
       "step": 125050
     },
     {
-      "epoch": 4.4,
-      "learning_rate": 1.875120756486956e-06,
-      "loss": 0.2755,
+      "epoch": 4.5069737268893935,
+      "grad_norm": 0.33327221870422363,
+      "learning_rate": 1.2721655025760365e-06,
+      "loss": 0.387,
       "step": 125055
     },
     {
-      "epoch": 4.4,
-      "learning_rate": 1.8740384527358767e-06,
-      "loss": 0.2498,
+      "epoch": 4.507153926550618,
+      "grad_norm": 0.24522875249385834,
+      "learning_rate": 1.2712466430146868e-06,
+      "loss": 0.3623,
       "step": 125060
     },
     {
-      "epoch": 4.4,
-      "learning_rate": 1.8729564492601742e-06,
-      "loss": 0.2621,
+      "epoch": 4.507334126211843,
+      "grad_norm": 0.2788931131362915,
+      "learning_rate": 1.270328106753005e-06,
+      "loss": 0.3606,
       "step": 125065
     },
     {
-      "epoch": 4.4,
-      "learning_rate": 1.87187474607389e-06,
-      "loss": 0.249,
+      "epoch": 4.507514325873068,
+      "grad_norm": 0.23487606644630432,
+      "learning_rate": 1.2694098938035082e-06,
+      "loss": 0.3898,
       "step": 125070
     },
     {
-      "epoch": 4.4,
-      "learning_rate": 1.8707933431910657e-06,
-      "loss": 0.2657,
+      "epoch": 4.507694525534292,
+      "grad_norm": 0.24076257646083832,
+      "learning_rate": 1.2684920041787034e-06,
+      "loss": 0.3831,
       "step": 125075
     },
     {
-      "epoch": 4.4,
-      "learning_rate": 1.869712240625751e-06,
-      "loss": 0.251,
+      "epoch": 4.507874725195517,
+      "grad_norm": 0.22647157311439514,
+      "learning_rate": 1.2675744378911058e-06,
+      "loss": 0.3621,
       "step": 125080
     },
     {
-      "epoch": 4.4,
-      "learning_rate": 1.8686314383919794e-06,
-      "loss": 0.2563,
+      "epoch": 4.508054924856741,
+      "grad_norm": 0.28119003772735596,
+      "learning_rate": 1.2666571949532109e-06,
+      "loss": 0.3778,
       "step": 125085
     },
     {
-      "epoch": 4.4,
-      "learning_rate": 1.867550936503784e-06,
-      "loss": 0.2529,
+      "epoch": 4.508235124517966,
+      "grad_norm": 0.20872727036476135,
+      "learning_rate": 1.2657402753775165e-06,
+      "loss": 0.3657,
       "step": 125090
     },
     {
-      "epoch": 4.4,
-      "learning_rate": 1.8664707349751926e-06,
-      "loss": 0.2425,
+      "epoch": 4.50841532417919,
+      "grad_norm": 0.2806181311607361,
+      "learning_rate": 1.2648236791765161e-06,
+      "loss": 0.3691,
       "step": 125095
     },
     {
-      "epoch": 4.4,
-      "learning_rate": 1.8653908338202325e-06,
-      "loss": 0.2547,
+      "epoch": 4.508595523840415,
+      "grad_norm": 0.2301296591758728,
+      "learning_rate": 1.2639074063626939e-06,
+      "loss": 0.376,
       "step": 125100
     },
     {
-      "epoch": 4.4,
-      "learning_rate": 1.8643112330529288e-06,
-      "loss": 0.2334,
+      "epoch": 4.50877572350164,
+      "grad_norm": 0.31790241599082947,
+      "learning_rate": 1.2629914569485396e-06,
+      "loss": 0.3857,
       "step": 125105
     },
     {
-      "epoch": 4.4,
-      "learning_rate": 1.8632319326872982e-06,
-      "loss": 0.2442,
+      "epoch": 4.508955923162865,
+      "grad_norm": 0.24608851969242096,
+      "learning_rate": 1.262075830946538e-06,
+      "loss": 0.3611,
       "step": 125110
     },
     {
-      "epoch": 4.4,
-      "learning_rate": 1.8621529327373488e-06,
-      "loss": 0.2678,
+      "epoch": 4.509136122824089,
+      "grad_norm": 0.2477652132511139,
+      "learning_rate": 1.2611605283691485e-06,
+      "loss": 0.3455,
       "step": 125115
     },
     {
-      "epoch": 4.4,
-      "learning_rate": 1.8610742332171e-06,
-      "loss": 0.2398,
+      "epoch": 4.509316322485314,
+      "grad_norm": 0.2697392702102661,
+      "learning_rate": 1.2602455492288556e-06,
+      "loss": 0.3734,
       "step": 125120
     },
     {
-      "epoch": 4.4,
-      "learning_rate": 1.8599958341405515e-06,
-      "loss": 0.2348,
+      "epoch": 4.509496522146538,
+      "grad_norm": 0.29126474261283875,
+      "learning_rate": 1.259330893538116e-06,
+      "loss": 0.3937,
       "step": 125125
     },
     {
-      "epoch": 4.4,
-      "learning_rate": 1.8589177355217092e-06,
-      "loss": 0.2419,
+      "epoch": 4.509676721807763,
+      "grad_norm": 0.281927227973938,
+      "learning_rate": 1.2584165613094029e-06,
+      "loss": 0.3374,
       "step": 125130
     },
     {
-      "epoch": 4.4,
-      "learning_rate": 1.8578399373745643e-06,
-      "loss": 0.2521,
+      "epoch": 4.509856921468987,
+      "grad_norm": 0.30632874369621277,
+      "learning_rate": 1.2575025525551648e-06,
+      "loss": 0.3651,
       "step": 125135
     },
     {
-      "epoch": 4.4,
-      "learning_rate": 1.8567624397131195e-06,
-      "loss": 0.2476,
+      "epoch": 4.510037121130212,
+      "grad_norm": 0.2316470891237259,
+      "learning_rate": 1.2565888672878556e-06,
+      "loss": 0.3384,
       "step": 125140
     },
     {
-      "epoch": 4.4,
-      "learning_rate": 1.8556852425513637e-06,
-      "loss": 0.2725,
+      "epoch": 4.510217320791437,
+      "grad_norm": 0.2622465193271637,
+      "learning_rate": 1.2556755055199292e-06,
+      "loss": 0.3726,
       "step": 125145
     },
     {
-      "epoch": 4.4,
-      "learning_rate": 1.8546083459032775e-06,
-      "loss": 0.2639,
+      "epoch": 4.510397520452662,
+      "grad_norm": 0.25914567708969116,
+      "learning_rate": 1.254762467263826e-06,
+      "loss": 0.3963,
       "step": 125150
     },
     {
-      "epoch": 4.4,
-      "learning_rate": 1.8535317497828552e-06,
-      "loss": 0.2557,
+      "epoch": 4.510577720113886,
+      "grad_norm": 0.26706215739250183,
+      "learning_rate": 1.2538497525319882e-06,
+      "loss": 0.3942,
       "step": 125155
     },
     {
-      "epoch": 4.4,
-      "learning_rate": 1.8524554542040634e-06,
-      "loss": 0.2451,
+      "epoch": 4.510757919775111,
+      "grad_norm": 0.23364363610744476,
+      "learning_rate": 1.2529373613368506e-06,
+      "loss": 0.3678,
       "step": 125160
     },
     {
-      "epoch": 4.4,
-      "learning_rate": 1.8513794591808853e-06,
-      "loss": 0.244,
+      "epoch": 4.510938119436336,
+      "grad_norm": 0.28146228194236755,
+      "learning_rate": 1.2520252936908394e-06,
+      "loss": 0.3441,
       "step": 125165
     },
     {
-      "epoch": 4.4,
-      "learning_rate": 1.8503037647272902e-06,
-      "loss": 0.2563,
+      "epoch": 4.511118319097561,
+      "grad_norm": 0.25890275835990906,
+      "learning_rate": 1.251113549606389e-06,
+      "loss": 0.3795,
       "step": 125170
     },
     {
-      "epoch": 4.4,
-      "learning_rate": 1.8492283708572422e-06,
-      "loss": 0.2534,
+      "epoch": 4.511298518758784,
+      "grad_norm": 0.24945032596588135,
+      "learning_rate": 1.2502021290959177e-06,
+      "loss": 0.3772,
       "step": 125175
     },
     {
-      "epoch": 4.4,
-      "learning_rate": 1.8481532775847077e-06,
-      "loss": 0.2703,
+      "epoch": 4.511478718420009,
+      "grad_norm": 0.2131379246711731,
+      "learning_rate": 1.2492910321718453e-06,
+      "loss": 0.3562,
       "step": 125180
     },
     {
-      "epoch": 4.4,
-      "learning_rate": 1.8470784849236479e-06,
-      "loss": 0.2378,
+      "epoch": 4.511658918081234,
+      "grad_norm": 0.32095032930374146,
+      "learning_rate": 1.2483802588465848e-06,
+      "loss": 0.3529,
       "step": 125185
     },
     {
-      "epoch": 4.4,
-      "learning_rate": 1.8460039928880125e-06,
-      "loss": 0.2739,
+      "epoch": 4.511839117742459,
+      "grad_norm": 0.2094382345676422,
+      "learning_rate": 1.2474698091325454e-06,
+      "loss": 0.3687,
       "step": 125190
     },
     {
-      "epoch": 4.4,
-      "learning_rate": 1.8449298014917572e-06,
-      "loss": 0.2582,
+      "epoch": 4.512019317403683,
+      "grad_norm": 0.279956191778183,
+      "learning_rate": 1.2465596830421255e-06,
+      "loss": 0.426,
       "step": 125195
     },
     {
-      "epoch": 4.4,
-      "learning_rate": 1.843855910748829e-06,
-      "loss": 0.2524,
+      "epoch": 4.512199517064908,
+      "grad_norm": 0.2812948226928711,
+      "learning_rate": 1.2456498805877404e-06,
+      "loss": 0.339,
       "step": 125200
     },
     {
-      "epoch": 4.41,
-      "learning_rate": 1.8427823206731699e-06,
-      "loss": 0.2527,
+      "epoch": 4.512379716726133,
+      "grad_norm": 0.29968753457069397,
+      "learning_rate": 1.2447404017817687e-06,
+      "loss": 0.3693,
       "step": 125205
     },
     {
-      "epoch": 4.41,
-      "learning_rate": 1.841709031278724e-06,
-      "loss": 0.2494,
+      "epoch": 4.5125599163873575,
+      "grad_norm": 0.2712326943874359,
+      "learning_rate": 1.243831246636612e-06,
+      "loss": 0.3906,
       "step": 125210
     },
     {
-      "epoch": 4.41,
-      "learning_rate": 1.840636042579419e-06,
-      "loss": 0.2502,
+      "epoch": 4.512740116048581,
+      "grad_norm": 0.24692299962043762,
+      "learning_rate": 1.2429224151646574e-06,
+      "loss": 0.3267,
       "step": 125215
     },
     {
-      "epoch": 4.41,
-      "learning_rate": 1.8395633545891965e-06,
-      "loss": 0.2505,
+      "epoch": 4.512920315709806,
+      "grad_norm": 0.25159400701522827,
+      "learning_rate": 1.242013907378281e-06,
+      "loss": 0.3634,
       "step": 125220
     },
     {
-      "epoch": 4.41,
-      "learning_rate": 1.8384909673219818e-06,
-      "loss": 0.2543,
+      "epoch": 4.513100515371031,
+      "grad_norm": 0.2895723581314087,
+      "learning_rate": 1.2411057232898733e-06,
+      "loss": 0.3826,
       "step": 125225
     },
     {
-      "epoch": 4.41,
-      "learning_rate": 1.8374188807916942e-06,
-      "loss": 0.2378,
+      "epoch": 4.513280715032256,
+      "grad_norm": 0.23947159945964813,
+      "learning_rate": 1.2401978629117934e-06,
+      "loss": 0.3526,
       "step": 125230
     },
     {
-      "epoch": 4.41,
-      "learning_rate": 1.8363470950122614e-06,
-      "loss": 0.2551,
+      "epoch": 4.51346091469348,
+      "grad_norm": 0.2706504166126251,
+      "learning_rate": 1.2392903262564176e-06,
+      "loss": 0.3864,
       "step": 125235
     },
     {
-      "epoch": 4.41,
-      "learning_rate": 1.835275609997597e-06,
-      "loss": 0.2633,
+      "epoch": 4.513641114354705,
+      "grad_norm": 0.2251925766468048,
+      "learning_rate": 1.238383113336114e-06,
+      "loss": 0.3673,
       "step": 125240
     },
     {
-      "epoch": 4.41,
-      "learning_rate": 1.8342044257616097e-06,
-      "loss": 0.2501,
+      "epoch": 4.51382131401593,
+      "grad_norm": 0.2925741374492645,
+      "learning_rate": 1.2374762241632393e-06,
+      "loss": 0.3739,
       "step": 125245
     },
     {
-      "epoch": 4.41,
-      "learning_rate": 1.8331335423182106e-06,
-      "loss": 0.2508,
+      "epoch": 4.5140015136771545,
+      "grad_norm": 0.24656470119953156,
+      "learning_rate": 1.2365696587501502e-06,
+      "loss": 0.3843,
       "step": 125250
     },
     {
-      "epoch": 4.41,
-      "learning_rate": 1.8320629596813078e-06,
-      "loss": 0.2496,
+      "epoch": 4.514181713338379,
+      "grad_norm": 0.2367469221353531,
+      "learning_rate": 1.235663417109198e-06,
+      "loss": 0.3662,
       "step": 125255
     },
     {
-      "epoch": 4.41,
-      "learning_rate": 1.8309926778647928e-06,
-      "loss": 0.2322,
+      "epoch": 4.514361912999604,
+      "grad_norm": 0.23144228756427765,
+      "learning_rate": 1.2347574992527283e-06,
+      "loss": 0.3484,
       "step": 125260
     },
     {
-      "epoch": 4.41,
-      "learning_rate": 1.8299226968825745e-06,
-      "loss": 0.2623,
+      "epoch": 4.514542112660828,
+      "grad_norm": 0.3216107189655304,
+      "learning_rate": 1.2338519051930925e-06,
+      "loss": 0.3521,
       "step": 125265
     },
     {
-      "epoch": 4.41,
-      "learning_rate": 1.8288530167485385e-06,
-      "loss": 0.2539,
+      "epoch": 4.514722312322053,
+      "grad_norm": 0.3389444649219513,
+      "learning_rate": 1.2329466349426194e-06,
+      "loss": 0.4033,
       "step": 125270
     },
     {
-      "epoch": 4.41,
-      "learning_rate": 1.8277836374765767e-06,
-      "loss": 0.2352,
+      "epoch": 4.514902511983277,
+      "grad_norm": 0.21677029132843018,
+      "learning_rate": 1.2320416885136493e-06,
+      "loss": 0.3743,
       "step": 125275
     },
     {
-      "epoch": 4.41,
-      "learning_rate": 1.826714559080575e-06,
-      "loss": 0.2707,
+      "epoch": 4.515082711644502,
+      "grad_norm": 0.2557178735733032,
+      "learning_rate": 1.2311370659185084e-06,
+      "loss": 0.3911,
       "step": 125280
     },
     {
-      "epoch": 4.41,
-      "learning_rate": 1.8256457815744117e-06,
-      "loss": 0.2852,
+      "epoch": 4.515262911305727,
+      "grad_norm": 0.2284223735332489,
+      "learning_rate": 1.2302327671695201e-06,
+      "loss": 0.3583,
       "step": 125285
     },
     {
-      "epoch": 4.41,
-      "learning_rate": 1.8245773049719639e-06,
-      "loss": 0.2284,
+      "epoch": 4.5154431109669515,
+      "grad_norm": 0.23775717616081238,
+      "learning_rate": 1.2293287922790108e-06,
+      "loss": 0.3642,
       "step": 125290
     },
     {
-      "epoch": 4.41,
-      "learning_rate": 1.823509129287107e-06,
-      "loss": 0.2353,
+      "epoch": 4.515623310628176,
+      "grad_norm": 0.24203971028327942,
+      "learning_rate": 1.2284251412593011e-06,
+      "loss": 0.3632,
       "step": 125295
     },
     {
-      "epoch": 4.41,
-      "learning_rate": 1.8224412545337132e-06,
-      "loss": 0.2682,
+      "epoch": 4.515803510289401,
+      "grad_norm": 0.30471011996269226,
+      "learning_rate": 1.2275218141226868e-06,
+      "loss": 0.3797,
       "step": 125300
     },
     {
-      "epoch": 4.41,
-      "learning_rate": 1.8213736807256382e-06,
-      "loss": 0.262,
+      "epoch": 4.515983709950625,
+      "grad_norm": 0.2442302256822586,
+      "learning_rate": 1.2266188108814886e-06,
+      "loss": 0.3587,
       "step": 125305
     },
     {
-      "epoch": 4.41,
-      "learning_rate": 1.8203064078767568e-06,
-      "loss": 0.284,
+      "epoch": 4.5161639096118495,
+      "grad_norm": 0.2527487874031067,
+      "learning_rate": 1.2257161315480048e-06,
+      "loss": 0.3862,
       "step": 125310
     },
     {
-      "epoch": 4.41,
-      "learning_rate": 1.8192394360009218e-06,
-      "loss": 0.239,
+      "epoch": 4.516344109273074,
+      "grad_norm": 0.2215377241373062,
+      "learning_rate": 1.2248137761345424e-06,
+      "loss": 0.3901,
       "step": 125315
     },
     {
-      "epoch": 4.41,
-      "learning_rate": 1.8181727651119807e-06,
-      "loss": 0.245,
+      "epoch": 4.516524308934299,
+      "grad_norm": 0.23120597004890442,
+      "learning_rate": 1.2239117446533833e-06,
+      "loss": 0.4032,
       "step": 125320
     },
     {
-      "epoch": 4.41,
-      "learning_rate": 1.8171063952237888e-06,
-      "loss": 0.2323,
+      "epoch": 4.516704508595524,
+      "grad_norm": 0.2772935926914215,
+      "learning_rate": 1.2230100371168229e-06,
+      "loss": 0.3507,
       "step": 125325
     },
     {
-      "epoch": 4.41,
-      "learning_rate": 1.8160403263501963e-06,
-      "loss": 0.2541,
+      "epoch": 4.5168847082567485,
+      "grad_norm": 0.31849685311317444,
+      "learning_rate": 1.222108653537149e-06,
+      "loss": 0.3966,
       "step": 125330
     },
     {
-      "epoch": 4.41,
-      "learning_rate": 1.8149745585050421e-06,
-      "loss": 0.2466,
+      "epoch": 4.517064907917973,
+      "grad_norm": 0.23283176124095917,
+      "learning_rate": 1.2212075939266432e-06,
+      "loss": 0.3609,
       "step": 125335
     },
     {
-      "epoch": 4.41,
-      "learning_rate": 1.8139090917021623e-06,
-      "loss": 0.2559,
+      "epoch": 4.517245107579198,
+      "grad_norm": 0.268390953540802,
+      "learning_rate": 1.220306858297579e-06,
+      "loss": 0.3574,
       "step": 125340
     },
     {
-      "epoch": 4.41,
-      "learning_rate": 1.8128439259553875e-06,
-      "loss": 0.2354,
+      "epoch": 4.517425307240423,
+      "grad_norm": 0.315220445394516,
+      "learning_rate": 1.21940644666223e-06,
+      "loss": 0.3818,
       "step": 125345
     },
     {
-      "epoch": 4.41,
-      "learning_rate": 1.8117790612785596e-06,
-      "loss": 0.2612,
+      "epoch": 4.517605506901647,
+      "grad_norm": 0.243333101272583,
+      "learning_rate": 1.2185063590328616e-06,
+      "loss": 0.381,
       "step": 125350
     },
     {
-      "epoch": 4.41,
-      "learning_rate": 1.8107144976855006e-06,
-      "loss": 0.251,
+      "epoch": 4.517785706562872,
+      "grad_norm": 0.22509504854679108,
+      "learning_rate": 1.2176065954217414e-06,
+      "loss": 0.3696,
       "step": 125355
     },
     {
-      "epoch": 4.41,
-      "learning_rate": 1.8096502351900274e-06,
-      "loss": 0.2372,
+      "epoch": 4.517965906224096,
+      "grad_norm": 0.27468258142471313,
+      "learning_rate": 1.2167071558411263e-06,
+      "loss": 0.3988,
       "step": 125360
     },
     {
-      "epoch": 4.41,
-      "learning_rate": 1.8085862738059622e-06,
-      "loss": 0.274,
+      "epoch": 4.518146105885321,
+      "grad_norm": 0.3038482367992401,
+      "learning_rate": 1.2158080403032734e-06,
+      "loss": 0.3707,
       "step": 125365
     },
     {
-      "epoch": 4.41,
-      "learning_rate": 1.8075226135471247e-06,
-      "loss": 0.2536,
+      "epoch": 4.5183263055465455,
+      "grad_norm": 0.23414720594882965,
+      "learning_rate": 1.2149092488204311e-06,
+      "loss": 0.339,
       "step": 125370
     },
     {
-      "epoch": 4.41,
-      "learning_rate": 1.8064592544273146e-06,
-      "loss": 0.2435,
+      "epoch": 4.51850650520777,
+      "grad_norm": 0.2685023546218872,
+      "learning_rate": 1.214010781404845e-06,
+      "loss": 0.3801,
       "step": 125375
     },
     {
-      "epoch": 4.41,
-      "learning_rate": 1.8053961964603517e-06,
-      "loss": 0.2335,
+      "epoch": 4.518686704868995,
+      "grad_norm": 0.25739866495132446,
+      "learning_rate": 1.2131126380687558e-06,
+      "loss": 0.3869,
       "step": 125380
     },
     {
-      "epoch": 4.41,
-      "learning_rate": 1.8043334396600304e-06,
-      "loss": 0.2588,
+      "epoch": 4.51886690453022,
+      "grad_norm": 0.2300902158021927,
+      "learning_rate": 1.2122148188244004e-06,
+      "loss": 0.3804,
       "step": 125385
     },
     {
-      "epoch": 4.41,
-      "learning_rate": 1.8032709840401563e-06,
-      "loss": 0.2283,
+      "epoch": 4.519047104191444,
+      "grad_norm": 0.2580128610134125,
+      "learning_rate": 1.2113173236840164e-06,
+      "loss": 0.3291,
       "step": 125390
     },
     {
-      "epoch": 4.41,
-      "learning_rate": 1.8022088296145212e-06,
-      "loss": 0.2606,
+      "epoch": 4.519227303852669,
+      "grad_norm": 0.27533698081970215,
+      "learning_rate": 1.2104201526598274e-06,
+      "loss": 0.3588,
       "step": 125395
     },
     {
-      "epoch": 4.41,
-      "learning_rate": 1.801146976396914e-06,
-      "loss": 0.2641,
+      "epoch": 4.519407503513893,
+      "grad_norm": 0.2444005161523819,
+      "learning_rate": 1.209523305764057e-06,
+      "loss": 0.364,
       "step": 125400
     },
     {
-      "epoch": 4.41,
-      "learning_rate": 1.8000854244011239e-06,
-      "loss": 0.2468,
+      "epoch": 4.519587703175118,
+      "grad_norm": 0.25520777702331543,
+      "learning_rate": 1.208626783008923e-06,
+      "loss": 0.3582,
       "step": 125405
     },
     {
-      "epoch": 4.41,
-      "learning_rate": 1.7990241736409369e-06,
-      "loss": 0.244,
+      "epoch": 4.519767902836342,
+      "grad_norm": 0.2750135064125061,
+      "learning_rate": 1.2077305844066494e-06,
+      "loss": 0.3841,
       "step": 125410
     },
     {
-      "epoch": 4.41,
-      "learning_rate": 1.7979632241301337e-06,
-      "loss": 0.2569,
+      "epoch": 4.519948102497567,
+      "grad_norm": 0.24095915257930756,
+      "learning_rate": 1.206834709969437e-06,
+      "loss": 0.3664,
       "step": 125415
     },
     {
-      "epoch": 4.41,
-      "learning_rate": 1.7969025758824864e-06,
-      "loss": 0.2388,
+      "epoch": 4.520128302158792,
+      "grad_norm": 0.2322002798318863,
+      "learning_rate": 1.2059391597094905e-06,
+      "loss": 0.3576,
       "step": 125420
     },
     {
-      "epoch": 4.41,
-      "learning_rate": 1.795842228911762e-06,
-      "loss": 0.2548,
+      "epoch": 4.520308501820017,
+      "grad_norm": 0.2865905463695526,
+      "learning_rate": 1.2050439336390217e-06,
+      "loss": 0.3656,
       "step": 125425
     },
     {
-      "epoch": 4.41,
-      "learning_rate": 1.7947821832317357e-06,
-      "loss": 0.243,
+      "epoch": 4.520488701481241,
+      "grad_norm": 0.26659440994262695,
+      "learning_rate": 1.2041490317702214e-06,
+      "loss": 0.4096,
       "step": 125430
     },
     {
-      "epoch": 4.41,
-      "learning_rate": 1.793722438856174e-06,
-      "loss": 0.257,
+      "epoch": 4.520668901142466,
+      "grad_norm": 0.29634758830070496,
+      "learning_rate": 1.2032544541152851e-06,
+      "loss": 0.3838,
       "step": 125435
     },
     {
-      "epoch": 4.41,
-      "learning_rate": 1.7926629957988272e-06,
-      "loss": 0.244,
+      "epoch": 4.520849100803691,
+      "grad_norm": 0.2613474428653717,
+      "learning_rate": 1.2023602006863976e-06,
+      "loss": 0.3772,
       "step": 125440
     },
     {
-      "epoch": 4.41,
-      "learning_rate": 1.7916038540734648e-06,
-      "loss": 0.2558,
+      "epoch": 4.521029300464916,
+      "grad_norm": 0.2532717287540436,
+      "learning_rate": 1.2014662714957409e-06,
+      "loss": 0.3852,
       "step": 125445
     },
     {
-      "epoch": 4.41,
-      "learning_rate": 1.7905450136938284e-06,
-      "loss": 0.2988,
+      "epoch": 4.521209500126139,
+      "grad_norm": 0.22666624188423157,
+      "learning_rate": 1.2005726665555051e-06,
+      "loss": 0.3694,
       "step": 125450
     },
     {
-      "epoch": 4.41,
-      "learning_rate": 1.7894864746736684e-06,
-      "loss": 0.274,
+      "epoch": 4.521389699787364,
+      "grad_norm": 0.22881825268268585,
+      "learning_rate": 1.1996793858778554e-06,
+      "loss": 0.2936,
       "step": 125455
     },
     {
-      "epoch": 4.41,
-      "learning_rate": 1.7884282370267291e-06,
-      "loss": 0.2545,
+      "epoch": 4.521569899448589,
+      "grad_norm": 0.24387423694133759,
+      "learning_rate": 1.1987864294749685e-06,
+      "loss": 0.3588,
       "step": 125460
     },
     {
-      "epoch": 4.41,
-      "learning_rate": 1.7873703007667553e-06,
-      "loss": 0.2523,
+      "epoch": 4.521750099109814,
+      "grad_norm": 0.2651515305042267,
+      "learning_rate": 1.1978937973590092e-06,
+      "loss": 0.3621,
       "step": 125465
     },
     {
-      "epoch": 4.41,
-      "learning_rate": 1.7863126659074802e-06,
-      "loss": 0.2514,
+      "epoch": 4.521930298771038,
+      "grad_norm": 0.21903403103351593,
+      "learning_rate": 1.197001489542135e-06,
+      "loss": 0.3943,
       "step": 125470
     },
     {
-      "epoch": 4.41,
-      "learning_rate": 1.7852553324626376e-06,
-      "loss": 0.2401,
+      "epoch": 4.522110498432263,
+      "grad_norm": 0.2606528103351593,
+      "learning_rate": 1.196109506036508e-06,
+      "loss": 0.353,
       "step": 125475
     },
     {
-      "epoch": 4.41,
-      "learning_rate": 1.7841983004459495e-06,
-      "loss": 0.2538,
+      "epoch": 4.522290698093488,
+      "grad_norm": 0.22265635430812836,
+      "learning_rate": 1.1952178468542852e-06,
+      "loss": 0.3417,
       "step": 125480
     },
     {
-      "epoch": 4.41,
-      "learning_rate": 1.7831415698711467e-06,
-      "loss": 0.2336,
+      "epoch": 4.522470897754713,
+      "grad_norm": 0.24026912450790405,
+      "learning_rate": 1.1943265120076042e-06,
+      "loss": 0.3699,
       "step": 125485
     },
     {
-      "epoch": 4.42,
-      "learning_rate": 1.7820851407519572e-06,
-      "loss": 0.273,
+      "epoch": 4.522651097415936,
+      "grad_norm": 0.23182328045368195,
+      "learning_rate": 1.193435501508619e-06,
+      "loss": 0.3662,
       "step": 125490
     },
     {
-      "epoch": 4.42,
-      "learning_rate": 1.7810290131020868e-06,
-      "loss": 0.2575,
+      "epoch": 4.522831297077161,
+      "grad_norm": 0.26761478185653687,
+      "learning_rate": 1.1925448153694619e-06,
+      "loss": 0.3798,
       "step": 125495
     },
     {
-      "epoch": 4.42,
-      "learning_rate": 1.779973186935252e-06,
-      "loss": 0.2638,
+      "epoch": 4.523011496738386,
+      "grad_norm": 0.23140046000480652,
+      "learning_rate": 1.1916544536022783e-06,
+      "loss": 0.3844,
       "step": 125500
     },
     {
-      "epoch": 4.42,
-      "eval_loss": 0.24885064363479614,
-      "eval_runtime": 10.5403,
-      "eval_samples_per_second": 9.487,
-      "eval_steps_per_second": 9.487,
+      "epoch": 4.523011496738386,
+      "eval_loss": 0.428926557302475,
+      "eval_runtime": 3.5228,
+      "eval_samples_per_second": 28.386,
+      "eval_steps_per_second": 7.097,
       "step": 125500
     },
     {
-      "epoch": 4.42,
-      "learning_rate": 1.7789176622651642e-06,
-      "loss": 0.2569,
+      "epoch": 4.523191696399611,
+      "grad_norm": 0.2863670885562897,
+      "learning_rate": 1.1907644162191922e-06,
+      "loss": 0.3947,
       "step": 125505
     },
     {
-      "epoch": 4.42,
-      "learning_rate": 1.7778624391055293e-06,
-      "loss": 0.2524,
+      "epoch": 4.523371896060835,
+      "grad_norm": 0.1922617256641388,
+      "learning_rate": 1.189874703232327e-06,
+      "loss": 0.3523,
       "step": 125510
     },
     {
-      "epoch": 4.42,
-      "learning_rate": 1.776807517470047e-06,
-      "loss": 0.2501,
+      "epoch": 4.52355209572206,
+      "grad_norm": 0.26853513717651367,
+      "learning_rate": 1.1889853146538148e-06,
+      "loss": 0.3827,
       "step": 125515
     },
     {
-      "epoch": 4.42,
-      "learning_rate": 1.7757528973724096e-06,
-      "loss": 0.2501,
+      "epoch": 4.523732295383285,
+      "grad_norm": 0.31399276852607727,
+      "learning_rate": 1.1880962504957655e-06,
+      "loss": 0.39,
       "step": 125520
     },
     {
-      "epoch": 4.42,
-      "learning_rate": 1.7746985788263199e-06,
-      "loss": 0.2399,
+      "epoch": 4.5239124950445095,
+      "grad_norm": 0.2765243649482727,
+      "learning_rate": 1.187207510770294e-06,
+      "loss": 0.3471,
       "step": 125525
     },
     {
-      "epoch": 4.42,
-      "learning_rate": 1.773644561845461e-06,
-      "loss": 0.2688,
+      "epoch": 4.524092694705734,
+      "grad_norm": 0.2347140908241272,
+      "learning_rate": 1.1863190954895104e-06,
+      "loss": 0.3601,
       "step": 125530
     },
     {
-      "epoch": 4.42,
-      "learning_rate": 1.7725908464435198e-06,
-      "loss": 0.262,
+      "epoch": 4.524272894366959,
+      "grad_norm": 0.25988760590553284,
+      "learning_rate": 1.1854310046655158e-06,
+      "loss": 0.3578,
       "step": 125535
     },
     {
-      "epoch": 4.42,
-      "learning_rate": 1.7715374326341767e-06,
-      "loss": 0.2531,
+      "epoch": 4.524453094028183,
+      "grad_norm": 0.26233744621276855,
+      "learning_rate": 1.184543238310415e-06,
+      "loss": 0.3576,
       "step": 125540
     },
     {
-      "epoch": 4.42,
-      "learning_rate": 1.7704843204311179e-06,
-      "loss": 0.2601,
+      "epoch": 4.524633293689408,
+      "grad_norm": 0.24501554667949677,
+      "learning_rate": 1.1836557964363032e-06,
+      "loss": 0.3731,
       "step": 125545
     },
     {
-      "epoch": 4.42,
-      "learning_rate": 1.7694315098480102e-06,
-      "loss": 0.2438,
+      "epoch": 4.524813493350632,
+      "grad_norm": 0.20214052498340607,
+      "learning_rate": 1.1827686790552711e-06,
+      "loss": 0.3533,
       "step": 125550
     },
     {
-      "epoch": 4.42,
-      "learning_rate": 1.7683790008985206e-06,
-      "loss": 0.2567,
+      "epoch": 4.524993693011857,
+      "grad_norm": 0.27101585268974304,
+      "learning_rate": 1.1818818861794007e-06,
+      "loss": 0.3721,
       "step": 125555
     },
     {
-      "epoch": 4.42,
-      "learning_rate": 1.7675372109243276e-06,
-      "loss": 0.2601,
+      "epoch": 4.525173892673082,
+      "grad_norm": 0.22587630152702332,
+      "learning_rate": 1.1809954178207821e-06,
+      "loss": 0.3643,
       "step": 125560
     },
     {
-      "epoch": 4.42,
-      "learning_rate": 1.7664852449497975e-06,
-      "loss": 0.2437,
+      "epoch": 4.5253540923343065,
+      "grad_norm": 0.293062686920166,
+      "learning_rate": 1.1801092739914838e-06,
+      "loss": 0.3687,
       "step": 125565
     },
     {
-      "epoch": 4.42,
-      "learning_rate": 1.7654335806471439e-06,
-      "loss": 0.2388,
+      "epoch": 4.525534291995531,
+      "grad_norm": 0.3036471903324127,
+      "learning_rate": 1.1792234547035903e-06,
+      "loss": 0.3325,
       "step": 125570
     },
     {
-      "epoch": 4.42,
-      "learning_rate": 1.7643822180300196e-06,
-      "loss": 0.2325,
+      "epoch": 4.525714491656756,
+      "grad_norm": 0.2614469826221466,
+      "learning_rate": 1.178337959969164e-06,
+      "loss": 0.3833,
       "step": 125575
     },
     {
-      "epoch": 4.42,
-      "learning_rate": 1.763331157112086e-06,
-      "loss": 0.2332,
+      "epoch": 4.52589469131798,
+      "grad_norm": 0.26248598098754883,
+      "learning_rate": 1.1774527898002707e-06,
+      "loss": 0.3678,
       "step": 125580
     },
     {
-      "epoch": 4.42,
-      "learning_rate": 1.762280397906982e-06,
-      "loss": 0.2463,
+      "epoch": 4.526074890979205,
+      "grad_norm": 0.22794273495674133,
+      "learning_rate": 1.1765679442089728e-06,
+      "loss": 0.3575,
       "step": 125585
     },
     {
-      "epoch": 4.42,
-      "learning_rate": 1.7612299404283494e-06,
-      "loss": 0.2564,
+      "epoch": 4.526255090640429,
+      "grad_norm": 0.2476157546043396,
+      "learning_rate": 1.1756834232073189e-06,
+      "loss": 0.3442,
       "step": 125590
     },
     {
-      "epoch": 4.42,
-      "learning_rate": 1.760179784689836e-06,
-      "loss": 0.2581,
+      "epoch": 4.526435290301654,
+      "grad_norm": 0.236716166138649,
+      "learning_rate": 1.1747992268073743e-06,
+      "loss": 0.3679,
       "step": 125595
     },
     {
-      "epoch": 4.42,
-      "learning_rate": 1.7591299307050723e-06,
-      "loss": 0.266,
+      "epoch": 4.526615489962879,
+      "grad_norm": 0.295484334230423,
+      "learning_rate": 1.173915355021174e-06,
+      "loss": 0.3566,
       "step": 125600
     },
     {
-      "epoch": 4.42,
-      "learning_rate": 1.7580803784876893e-06,
-      "loss": 0.2848,
+      "epoch": 4.5267956896241035,
+      "grad_norm": 0.237171933054924,
+      "learning_rate": 1.1730318078607584e-06,
+      "loss": 0.3829,
       "step": 125605
     },
     {
-      "epoch": 4.42,
-      "learning_rate": 1.757031128051312e-06,
-      "loss": 0.2355,
+      "epoch": 4.526975889285328,
+      "grad_norm": 0.2715757191181183,
+      "learning_rate": 1.1721485853381787e-06,
+      "loss": 0.3776,
       "step": 125610
     },
     {
-      "epoch": 4.42,
-      "learning_rate": 1.755982179409571e-06,
-      "loss": 0.2384,
+      "epoch": 4.527156088946553,
+      "grad_norm": 0.21538300812244415,
+      "learning_rate": 1.1712656874654586e-06,
+      "loss": 0.3446,
       "step": 125615
     },
     {
-      "epoch": 4.42,
-      "learning_rate": 1.7549335325760834e-06,
-      "loss": 0.2755,
+      "epoch": 4.527336288607778,
+      "grad_norm": 0.21776896715164185,
+      "learning_rate": 1.1703831142546306e-06,
+      "loss": 0.3613,
       "step": 125620
     },
     {
-      "epoch": 4.42,
-      "learning_rate": 1.7538851875644606e-06,
-      "loss": 0.2615,
+      "epoch": 4.527516488269002,
+      "grad_norm": 0.22091469168663025,
+      "learning_rate": 1.1695008657177182e-06,
+      "loss": 0.3677,
       "step": 125625
     },
     {
-      "epoch": 4.42,
-      "learning_rate": 1.7528371443883195e-06,
-      "loss": 0.2423,
+      "epoch": 4.527696687930227,
+      "grad_norm": 0.24481423199176788,
+      "learning_rate": 1.1686189418667393e-06,
+      "loss": 0.3669,
       "step": 125630
     },
     {
-      "epoch": 4.42,
-      "learning_rate": 1.751789403061274e-06,
-      "loss": 0.2409,
+      "epoch": 4.527876887591451,
+      "grad_norm": 0.24965307116508484,
+      "learning_rate": 1.1677373427137178e-06,
+      "loss": 0.366,
       "step": 125635
     },
     {
-      "epoch": 4.42,
-      "learning_rate": 1.750741963596919e-06,
-      "loss": 0.2515,
+      "epoch": 4.528057087252676,
+      "grad_norm": 0.2324383556842804,
+      "learning_rate": 1.1668560682706608e-06,
+      "loss": 0.3827,
       "step": 125640
     },
     {
-      "epoch": 4.42,
-      "learning_rate": 1.7496948260088547e-06,
-      "loss": 0.2494,
+      "epoch": 4.5282372869139005,
+      "grad_norm": 0.2629086971282959,
+      "learning_rate": 1.1659751185495754e-06,
+      "loss": 0.3757,
       "step": 125645
     },
     {
-      "epoch": 4.42,
-      "learning_rate": 1.748647990310684e-06,
-      "loss": 0.2517,
+      "epoch": 4.528417486575125,
+      "grad_norm": 0.2776727080345154,
+      "learning_rate": 1.1650944935624658e-06,
+      "loss": 0.3613,
       "step": 125650
     },
     {
-      "epoch": 4.42,
-      "learning_rate": 1.747601456515996e-06,
-      "loss": 0.2673,
+      "epoch": 4.52859768623635,
+      "grad_norm": 0.28128179907798767,
+      "learning_rate": 1.1642141933213253e-06,
+      "loss": 0.3698,
       "step": 125655
     },
     {
-      "epoch": 4.42,
-      "learning_rate": 1.746555224638377e-06,
-      "loss": 0.2398,
+      "epoch": 4.528777885897575,
+      "grad_norm": 0.2046898454427719,
+      "learning_rate": 1.163334217838155e-06,
+      "loss": 0.3678,
       "step": 125660
     },
     {
-      "epoch": 4.42,
-      "learning_rate": 1.7455092946914136e-06,
-      "loss": 0.2578,
+      "epoch": 4.528958085558799,
+      "grad_norm": 0.21193328499794006,
+      "learning_rate": 1.1624545671249405e-06,
+      "loss": 0.3739,
       "step": 125665
     },
     {
-      "epoch": 4.42,
-      "learning_rate": 1.7444636666886894e-06,
-      "loss": 0.2711,
+      "epoch": 4.529138285220024,
+      "grad_norm": 0.30449825525283813,
+      "learning_rate": 1.161575241193666e-06,
+      "loss": 0.3799,
       "step": 125670
     },
     {
-      "epoch": 4.42,
-      "learning_rate": 1.7434183406437793e-06,
-      "loss": 0.2458,
+      "epoch": 4.529318484881248,
+      "grad_norm": 0.28137627243995667,
+      "learning_rate": 1.1606962400563166e-06,
+      "loss": 0.3485,
       "step": 125675
     },
     {
-      "epoch": 4.42,
-      "learning_rate": 1.7423733165702478e-06,
-      "loss": 0.2469,
+      "epoch": 4.529498684542473,
+      "grad_norm": 0.26191192865371704,
+      "learning_rate": 1.1598175637248605e-06,
+      "loss": 0.3948,
       "step": 125680
     },
     {
-      "epoch": 4.42,
-      "learning_rate": 1.7413285944816754e-06,
-      "loss": 0.2754,
+      "epoch": 4.5296788842036975,
+      "grad_norm": 0.21634413301944733,
+      "learning_rate": 1.1589392122112769e-06,
+      "loss": 0.3728,
       "step": 125685
     },
     {
-      "epoch": 4.42,
-      "learning_rate": 1.7402841743916266e-06,
-      "loss": 0.2554,
+      "epoch": 4.529859083864922,
+      "grad_norm": 0.2029135525226593,
+      "learning_rate": 1.1580611855275341e-06,
+      "loss": 0.3492,
       "step": 125690
     },
     {
-      "epoch": 4.42,
-      "learning_rate": 1.7392400563136568e-06,
-      "loss": 0.2446,
+      "epoch": 4.530039283526147,
+      "grad_norm": 0.2269791066646576,
+      "learning_rate": 1.1571834836855866e-06,
+      "loss": 0.3489,
       "step": 125695
     },
     {
-      "epoch": 4.42,
-      "learning_rate": 1.7381962402613277e-06,
-      "loss": 0.2441,
+      "epoch": 4.530219483187372,
+      "grad_norm": 0.2346084862947464,
+      "learning_rate": 1.1563061066974024e-06,
+      "loss": 0.3775,
       "step": 125700
     },
     {
-      "epoch": 4.42,
-      "learning_rate": 1.737152726248184e-06,
-      "loss": 0.2483,
+      "epoch": 4.530399682848596,
+      "grad_norm": 0.21212127804756165,
+      "learning_rate": 1.1554290545749273e-06,
+      "loss": 0.3526,
       "step": 125705
     },
     {
-      "epoch": 4.42,
-      "learning_rate": 1.736109514287787e-06,
-      "loss": 0.2573,
+      "epoch": 4.530579882509821,
+      "grad_norm": 0.22559018433094025,
+      "learning_rate": 1.154552327330119e-06,
+      "loss": 0.3672,
       "step": 125710
     },
     {
-      "epoch": 4.42,
-      "learning_rate": 1.7350666043936786e-06,
-      "loss": 0.2383,
+      "epoch": 4.530760082171046,
+      "grad_norm": 0.24675384163856506,
+      "learning_rate": 1.1536759249749173e-06,
+      "loss": 0.3749,
       "step": 125715
     },
     {
-      "epoch": 4.42,
-      "learning_rate": 1.7340239965793954e-06,
-      "loss": 0.2672,
+      "epoch": 4.530940281832271,
+      "grad_norm": 0.20101101696491241,
+      "learning_rate": 1.1527998475212604e-06,
+      "loss": 0.3611,
       "step": 125720
     },
     {
-      "epoch": 4.42,
-      "learning_rate": 1.7329816908584734e-06,
-      "loss": 0.2704,
+      "epoch": 4.5311204814934944,
+      "grad_norm": 0.24855537712574005,
+      "learning_rate": 1.151924094981091e-06,
+      "loss": 0.4015,
       "step": 125725
     },
     {
-      "epoch": 4.42,
-      "learning_rate": 1.731939687244455e-06,
-      "loss": 0.2382,
+      "epoch": 4.531300681154719,
+      "grad_norm": 0.2826426029205322,
+      "learning_rate": 1.151048667366339e-06,
+      "loss": 0.3815,
       "step": 125730
     },
     {
-      "epoch": 4.42,
-      "learning_rate": 1.7308979857508623e-06,
-      "loss": 0.2316,
+      "epoch": 4.531480880815944,
+      "grad_norm": 0.2761218547821045,
+      "learning_rate": 1.1501735646889332e-06,
+      "loss": 0.3988,
       "step": 125735
     },
     {
-      "epoch": 4.42,
-      "learning_rate": 1.729856586391229e-06,
-      "loss": 0.2461,
+      "epoch": 4.531661080477169,
+      "grad_norm": 0.2532573938369751,
+      "learning_rate": 1.1492987869607952e-06,
+      "loss": 0.3876,
       "step": 125740
     },
     {
-      "epoch": 4.42,
-      "learning_rate": 1.728815489179067e-06,
-      "loss": 0.2689,
+      "epoch": 4.531841280138393,
+      "grad_norm": 0.2559954524040222,
+      "learning_rate": 1.1484243341938427e-06,
+      "loss": 0.367,
       "step": 125745
     },
     {
-      "epoch": 4.42,
-      "learning_rate": 1.7277746941279038e-06,
-      "loss": 0.2622,
+      "epoch": 4.532021479799618,
+      "grad_norm": 0.22612114250659943,
+      "learning_rate": 1.147550206399986e-06,
+      "loss": 0.3544,
       "step": 125750
     },
     {
-      "epoch": 4.42,
-      "learning_rate": 1.7267342012512484e-06,
-      "loss": 0.2526,
+      "epoch": 4.532201679460843,
+      "grad_norm": 0.24155735969543457,
+      "learning_rate": 1.146676403591146e-06,
+      "loss": 0.4076,
       "step": 125755
     },
     {
-      "epoch": 4.42,
-      "learning_rate": 1.725694010562609e-06,
-      "loss": 0.2388,
+      "epoch": 4.532381879122068,
+      "grad_norm": 0.3148764967918396,
+      "learning_rate": 1.1458029257792185e-06,
+      "loss": 0.3787,
       "step": 125760
     },
     {
-      "epoch": 4.42,
-      "learning_rate": 1.7246541220754975e-06,
-      "loss": 0.2314,
+      "epoch": 4.532562078783291,
+      "grad_norm": 0.2415306717157364,
+      "learning_rate": 1.144929772976111e-06,
+      "loss": 0.3981,
       "step": 125765
     },
     {
-      "epoch": 4.42,
-      "learning_rate": 1.723614535803414e-06,
-      "loss": 0.2311,
+      "epoch": 4.532742278444516,
+      "grad_norm": 0.2210559993982315,
+      "learning_rate": 1.1440569451937166e-06,
+      "loss": 0.3602,
       "step": 125770
     },
     {
-      "epoch": 4.43,
-      "learning_rate": 1.7225752517598558e-06,
-      "loss": 0.2388,
+      "epoch": 4.532922478105741,
+      "grad_norm": 0.22382661700248718,
+      "learning_rate": 1.143184442443923e-06,
+      "loss": 0.337,
       "step": 125775
     },
     {
-      "epoch": 4.43,
-      "learning_rate": 1.7215362699583154e-06,
-      "loss": 0.2332,
+      "epoch": 4.533102677766966,
+      "grad_norm": 0.24328698217868805,
+      "learning_rate": 1.1423122647386291e-06,
+      "loss": 0.3835,
       "step": 125780
     },
     {
-      "epoch": 4.43,
-      "learning_rate": 1.7204975904122872e-06,
-      "loss": 0.2405,
+      "epoch": 4.53328287742819,
+      "grad_norm": 0.2541979253292084,
+      "learning_rate": 1.1414404120897087e-06,
+      "loss": 0.38,
       "step": 125785
     },
     {
-      "epoch": 4.43,
-      "learning_rate": 1.719459213135252e-06,
-      "loss": 0.2476,
+      "epoch": 4.533463077089415,
+      "grad_norm": 0.24431033432483673,
+      "learning_rate": 1.1405688845090383e-06,
+      "loss": 0.3388,
       "step": 125790
     },
     {
-      "epoch": 4.43,
-      "learning_rate": 1.7184211381407023e-06,
-      "loss": 0.2752,
+      "epoch": 4.53364327675064,
+      "grad_norm": 0.24632686376571655,
+      "learning_rate": 1.139697682008503e-06,
+      "loss": 0.351,
       "step": 125795
     },
     {
-      "epoch": 4.43,
-      "learning_rate": 1.7173833654421046e-06,
-      "loss": 0.2647,
+      "epoch": 4.533823476411865,
+      "grad_norm": 0.24334456026554108,
+      "learning_rate": 1.1388268045999655e-06,
+      "loss": 0.342,
       "step": 125800
     },
     {
-      "epoch": 4.43,
-      "learning_rate": 1.7163458950529454e-06,
-      "loss": 0.2338,
+      "epoch": 4.534003676073089,
+      "grad_norm": 0.22440917789936066,
+      "learning_rate": 1.137956252295297e-06,
+      "loss": 0.3438,
       "step": 125805
     },
     {
-      "epoch": 4.43,
-      "learning_rate": 1.715308726986692e-06,
-      "loss": 0.228,
+      "epoch": 4.534183875734314,
+      "grad_norm": 0.25877583026885986,
+      "learning_rate": 1.1370860251063515e-06,
+      "loss": 0.3857,
       "step": 125810
     },
     {
-      "epoch": 4.43,
-      "learning_rate": 1.7142718612568053e-06,
-      "loss": 0.2639,
+      "epoch": 4.534364075395538,
+      "grad_norm": 0.2521132826805115,
+      "learning_rate": 1.1362161230449863e-06,
+      "loss": 0.3869,
       "step": 125815
     },
     {
-      "epoch": 4.43,
-      "learning_rate": 1.713235297876753e-06,
-      "loss": 0.2404,
+      "epoch": 4.534544275056763,
+      "grad_norm": 0.23100094497203827,
+      "learning_rate": 1.1353465461230616e-06,
+      "loss": 0.3642,
       "step": 125820
     },
     {
-      "epoch": 4.43,
-      "learning_rate": 1.7121990368599961e-06,
-      "loss": 0.2597,
+      "epoch": 4.534724474717987,
+      "grad_norm": 0.2277524769306183,
+      "learning_rate": 1.134477294352418e-06,
+      "loss": 0.3368,
       "step": 125825
     },
     {
-      "epoch": 4.43,
-      "learning_rate": 1.711163078219985e-06,
-      "loss": 0.2799,
+      "epoch": 4.534904674379212,
+      "grad_norm": 0.2838972210884094,
+      "learning_rate": 1.1336083677448984e-06,
+      "loss": 0.4055,
       "step": 125830
     },
     {
-      "epoch": 4.43,
-      "learning_rate": 1.7101274219701756e-06,
-      "loss": 0.2636,
+      "epoch": 4.535084874040437,
+      "grad_norm": 0.26135292649269104,
+      "learning_rate": 1.132739766312349e-06,
+      "loss": 0.3795,
       "step": 125835
     },
     {
-      "epoch": 4.43,
-      "learning_rate": 1.7090920681240102e-06,
-      "loss": 0.2583,
+      "epoch": 4.5352650737016615,
+      "grad_norm": 0.24618162214756012,
+      "learning_rate": 1.131871490066591e-06,
+      "loss": 0.3496,
       "step": 125840
     },
     {
-      "epoch": 4.43,
-      "learning_rate": 1.7080570166949333e-06,
-      "loss": 0.2649,
+      "epoch": 4.535445273362886,
+      "grad_norm": 0.28488776087760925,
+      "learning_rate": 1.1310035390194706e-06,
+      "loss": 0.3829,
       "step": 125845
     },
     {
-      "epoch": 4.43,
-      "learning_rate": 1.7070222676963897e-06,
-      "loss": 0.2289,
+      "epoch": 4.535625473024111,
+      "grad_norm": 0.28462615609169006,
+      "learning_rate": 1.1301359131828032e-06,
+      "loss": 0.3896,
       "step": 125850
     },
     {
-      "epoch": 4.43,
-      "learning_rate": 1.7059878211418107e-06,
-      "loss": 0.2291,
+      "epoch": 4.535805672685335,
+      "grad_norm": 0.295928955078125,
+      "learning_rate": 1.1292686125684127e-06,
+      "loss": 0.4013,
       "step": 125855
     },
     {
-      "epoch": 4.43,
-      "learning_rate": 1.7049536770446295e-06,
-      "loss": 0.2542,
+      "epoch": 4.53598587234656,
+      "grad_norm": 0.27669599652290344,
+      "learning_rate": 1.1284016371881173e-06,
+      "loss": 0.3411,
       "step": 125860
     },
     {
-      "epoch": 4.43,
-      "learning_rate": 1.7039198354182718e-06,
-      "loss": 0.2395,
+      "epoch": 4.536166072007784,
+      "grad_norm": 0.31112879514694214,
+      "learning_rate": 1.1275349870537243e-06,
+      "loss": 0.3581,
       "step": 125865
     },
     {
-      "epoch": 4.43,
-      "learning_rate": 1.7028862962761654e-06,
-      "loss": 0.2525,
+      "epoch": 4.536346271669009,
+      "grad_norm": 0.30249273777008057,
+      "learning_rate": 1.1266686621770466e-06,
+      "loss": 0.3908,
       "step": 125870
     },
     {
-      "epoch": 4.43,
-      "learning_rate": 1.7018530596317195e-06,
-      "loss": 0.2491,
+      "epoch": 4.536526471330234,
+      "grad_norm": 0.260932058095932,
+      "learning_rate": 1.1258026625698914e-06,
+      "loss": 0.3768,
       "step": 125875
     },
     {
-      "epoch": 4.43,
-      "learning_rate": 1.7008201254983647e-06,
-      "loss": 0.2712,
+      "epoch": 4.5367066709914585,
+      "grad_norm": 0.2824000418186188,
+      "learning_rate": 1.1249369882440464e-06,
+      "loss": 0.3298,
       "step": 125880
     },
     {
-      "epoch": 4.43,
-      "learning_rate": 1.6997874938895043e-06,
-      "loss": 0.2482,
+      "epoch": 4.536886870652683,
+      "grad_norm": 0.2413741648197174,
+      "learning_rate": 1.124071639211316e-06,
+      "loss": 0.3432,
       "step": 125885
     },
     {
-      "epoch": 4.43,
-      "learning_rate": 1.6987551648185495e-06,
-      "loss": 0.2483,
+      "epoch": 4.537067070313908,
+      "grad_norm": 0.2007882446050644,
+      "learning_rate": 1.1232066154834852e-06,
+      "loss": 0.3411,
       "step": 125890
     },
     {
-      "epoch": 4.43,
-      "learning_rate": 1.6977231382988956e-06,
-      "loss": 0.2378,
+      "epoch": 4.537247269975133,
+      "grad_norm": 0.2404080480337143,
+      "learning_rate": 1.122341917072342e-06,
+      "loss": 0.3577,
       "step": 125895
     },
     {
-      "epoch": 4.43,
-      "learning_rate": 1.6966914143439566e-06,
-      "loss": 0.2375,
+      "epoch": 4.5374274696363575,
+      "grad_norm": 0.2568254768848419,
+      "learning_rate": 1.1214775439896686e-06,
+      "loss": 0.3915,
       "step": 125900
     },
     {
-      "epoch": 4.43,
-      "learning_rate": 1.695659992967119e-06,
-      "loss": 0.2365,
+      "epoch": 4.537607669297582,
+      "grad_norm": 0.26964160799980164,
+      "learning_rate": 1.1206134962472387e-06,
+      "loss": 0.3875,
       "step": 125905
     },
     {
-      "epoch": 4.43,
-      "learning_rate": 1.6946288741817806e-06,
-      "loss": 0.2516,
+      "epoch": 4.537787868958806,
+      "grad_norm": 0.26435330510139465,
+      "learning_rate": 1.1197497738568264e-06,
+      "loss": 0.3631,
       "step": 125910
     },
     {
-      "epoch": 4.43,
-      "learning_rate": 1.6935980580013221e-06,
-      "loss": 0.2581,
+      "epoch": 4.537968068620031,
+      "grad_norm": 0.27454784512519836,
+      "learning_rate": 1.1188863768302027e-06,
+      "loss": 0.3484,
       "step": 125915
     },
     {
-      "epoch": 4.43,
-      "learning_rate": 1.6925675444391386e-06,
-      "loss": 0.262,
+      "epoch": 4.5381482682812555,
+      "grad_norm": 0.2890207767486572,
+      "learning_rate": 1.1180233051791279e-06,
+      "loss": 0.3799,
       "step": 125920
     },
     {
-      "epoch": 4.43,
-      "learning_rate": 1.6915373335086054e-06,
-      "loss": 0.2466,
+      "epoch": 4.53832846794248,
+      "grad_norm": 0.27648934721946716,
+      "learning_rate": 1.1171605589153616e-06,
+      "loss": 0.3902,
       "step": 125925
     },
     {
-      "epoch": 4.43,
-      "learning_rate": 1.6905074252231007e-06,
-      "loss": 0.2644,
+      "epoch": 4.538508667603705,
+      "grad_norm": 0.2612933814525604,
+      "learning_rate": 1.1162981380506587e-06,
+      "loss": 0.3866,
       "step": 125930
     },
     {
-      "epoch": 4.43,
-      "learning_rate": 1.6894778195959887e-06,
-      "loss": 0.2811,
+      "epoch": 4.53868886726493,
+      "grad_norm": 0.23902583122253418,
+      "learning_rate": 1.1154360425967652e-06,
+      "loss": 0.3764,
       "step": 125935
     },
     {
-      "epoch": 4.43,
-      "learning_rate": 1.6884485166406504e-06,
-      "loss": 0.2388,
+      "epoch": 4.538869066926154,
+      "grad_norm": 0.2665193974971771,
+      "learning_rate": 1.1145742725654357e-06,
+      "loss": 0.3214,
       "step": 125940
     },
     {
-      "epoch": 4.43,
-      "learning_rate": 1.6874195163704475e-06,
-      "loss": 0.2606,
+      "epoch": 4.539049266587379,
+      "grad_norm": 0.3075253665447235,
+      "learning_rate": 1.1137128279684078e-06,
+      "loss": 0.3819,
       "step": 125945
     },
     {
-      "epoch": 4.43,
-      "learning_rate": 1.6863908187987305e-06,
-      "loss": 0.2765,
+      "epoch": 4.539229466248603,
+      "grad_norm": 0.25425055623054504,
+      "learning_rate": 1.1128517088174195e-06,
+      "loss": 0.3686,
       "step": 125950
     },
     {
-      "epoch": 4.43,
-      "learning_rate": 1.6853624239388716e-06,
-      "loss": 0.2499,
+      "epoch": 4.539409665909828,
+      "grad_norm": 0.2914336025714874,
+      "learning_rate": 1.1119909151242003e-06,
+      "loss": 0.4104,
       "step": 125955
     },
     {
-      "epoch": 4.43,
-      "learning_rate": 1.6843343318042104e-06,
-      "loss": 0.2536,
+      "epoch": 4.5395898655710525,
+      "grad_norm": 0.23320336639881134,
+      "learning_rate": 1.1111304469004769e-06,
+      "loss": 0.3615,
       "step": 125960
     },
     {
-      "epoch": 4.43,
-      "learning_rate": 1.6833065424081086e-06,
-      "loss": 0.2379,
+      "epoch": 4.539770065232277,
+      "grad_norm": 0.2035553902387619,
+      "learning_rate": 1.1102703041579787e-06,
+      "loss": 0.3544,
       "step": 125965
     },
     {
-      "epoch": 4.43,
-      "learning_rate": 1.6822790557638995e-06,
-      "loss": 0.2696,
+      "epoch": 4.539950264893502,
+      "grad_norm": 0.27693793177604675,
+      "learning_rate": 1.1094104869084242e-06,
+      "loss": 0.3813,
       "step": 125970
     },
     {
-      "epoch": 4.43,
-      "learning_rate": 1.681251871884934e-06,
-      "loss": 0.2442,
+      "epoch": 4.540130464554727,
+      "grad_norm": 0.2863492965698242,
+      "learning_rate": 1.1085509951635236e-06,
+      "loss": 0.3601,
       "step": 125975
     },
     {
-      "epoch": 4.43,
-      "learning_rate": 1.6802249907845457e-06,
-      "loss": 0.2633,
+      "epoch": 4.540310664215951,
+      "grad_norm": 0.22666563093662262,
+      "learning_rate": 1.1076918289349924e-06,
+      "loss": 0.3503,
       "step": 125980
     },
     {
-      "epoch": 4.43,
-      "learning_rate": 1.6791984124760657e-06,
-      "loss": 0.2424,
+      "epoch": 4.540490863877176,
+      "grad_norm": 0.25621458888053894,
+      "learning_rate": 1.1068329882345296e-06,
+      "loss": 0.3974,
       "step": 125985
     },
     {
-      "epoch": 4.43,
-      "learning_rate": 1.6781721369728248e-06,
-      "loss": 0.2456,
+      "epoch": 4.540671063538401,
+      "grad_norm": 0.23761415481567383,
+      "learning_rate": 1.105974473073848e-06,
+      "loss": 0.3689,
       "step": 125990
     },
     {
-      "epoch": 4.43,
-      "learning_rate": 1.6771461642881515e-06,
-      "loss": 0.2539,
+      "epoch": 4.540851263199626,
+      "grad_norm": 0.27705660462379456,
+      "learning_rate": 1.1051162834646356e-06,
+      "loss": 0.4032,
       "step": 125995
     },
     {
-      "epoch": 4.43,
-      "learning_rate": 1.6761204944353626e-06,
-      "loss": 0.2602,
+      "epoch": 4.5410314628608495,
+      "grad_norm": 0.25998204946517944,
+      "learning_rate": 1.1042584194185857e-06,
+      "loss": 0.3646,
       "step": 126000
     },
     {
-      "epoch": 4.43,
-      "eval_loss": 0.24884916841983795,
-      "eval_runtime": 10.5407,
-      "eval_samples_per_second": 9.487,
-      "eval_steps_per_second": 9.487,
+      "epoch": 4.5410314628608495,
+      "eval_loss": 0.42872315645217896,
+      "eval_runtime": 3.5299,
+      "eval_samples_per_second": 28.33,
+      "eval_steps_per_second": 7.082,
       "step": 126000
     },
     {
-      "epoch": 4.43,
-      "learning_rate": 1.675095127427781e-06,
-      "loss": 0.236,
+      "epoch": 4.541211662522074,
+      "grad_norm": 0.2178216576576233,
+      "learning_rate": 1.1034008809473916e-06,
+      "loss": 0.365,
       "step": 126005
     },
     {
-      "epoch": 4.43,
-      "learning_rate": 1.6740700632787127e-06,
-      "loss": 0.2544,
+      "epoch": 4.541391862183299,
+      "grad_norm": 0.2195737063884735,
+      "learning_rate": 1.1025436680627332e-06,
+      "loss": 0.3845,
       "step": 126010
     },
     {
-      "epoch": 4.43,
-      "learning_rate": 1.6730453020014718e-06,
-      "loss": 0.2377,
+      "epoch": 4.541572061844524,
+      "grad_norm": 0.24314382672309875,
+      "learning_rate": 1.10168678077629e-06,
+      "loss": 0.3472,
       "step": 126015
     },
     {
-      "epoch": 4.43,
-      "learning_rate": 1.6720208436093705e-06,
-      "loss": 0.2608,
+      "epoch": 4.541752261505748,
+      "grad_norm": 0.2001175880432129,
+      "learning_rate": 1.1008302190997383e-06,
+      "loss": 0.3606,
       "step": 126020
     },
     {
-      "epoch": 4.43,
-      "learning_rate": 1.670996688115703e-06,
-      "loss": 0.2792,
+      "epoch": 4.541932461166973,
+      "grad_norm": 0.24921679496765137,
+      "learning_rate": 1.0999739830447441e-06,
+      "loss": 0.3683,
       "step": 126025
     },
     {
-      "epoch": 4.43,
-      "learning_rate": 1.6699728355337645e-06,
-      "loss": 0.2465,
+      "epoch": 4.542112660828198,
+      "grad_norm": 0.2515079379081726,
+      "learning_rate": 1.0991180726229789e-06,
+      "loss": 0.3579,
       "step": 126030
     },
     {
-      "epoch": 4.43,
-      "learning_rate": 1.6689492858768584e-06,
-      "loss": 0.2511,
+      "epoch": 4.542292860489423,
+      "grad_norm": 0.2457331418991089,
+      "learning_rate": 1.0982624878461051e-06,
+      "loss": 0.3869,
       "step": 126035
     },
     {
-      "epoch": 4.43,
-      "learning_rate": 1.6679260391582713e-06,
-      "loss": 0.2508,
+      "epoch": 4.5424730601506464,
+      "grad_norm": 0.22210632264614105,
+      "learning_rate": 1.0974072287257775e-06,
+      "loss": 0.3974,
       "step": 126040
     },
     {
-      "epoch": 4.43,
-      "learning_rate": 1.6669030953912869e-06,
-      "loss": 0.2346,
+      "epoch": 4.542653259811871,
+      "grad_norm": 0.256584107875824,
+      "learning_rate": 1.0965522952736478e-06,
+      "loss": 0.3527,
       "step": 126045
     },
     {
-      "epoch": 4.43,
-      "learning_rate": 1.665880454589186e-06,
-      "loss": 0.2555,
+      "epoch": 4.542833459473096,
+      "grad_norm": 0.25087007880210876,
+      "learning_rate": 1.0956976875013598e-06,
+      "loss": 0.3566,
       "step": 126050
     },
     {
-      "epoch": 4.43,
-      "learning_rate": 1.6648581167652498e-06,
-      "loss": 0.2365,
+      "epoch": 4.543013659134321,
+      "grad_norm": 0.2532568871974945,
+      "learning_rate": 1.0948434054205704e-06,
+      "loss": 0.3707,
       "step": 126055
     },
     {
-      "epoch": 4.44,
-      "learning_rate": 1.6638360819327569e-06,
-      "loss": 0.2425,
+      "epoch": 4.543193858795545,
+      "grad_norm": 0.22819988429546356,
+      "learning_rate": 1.093989449042912e-06,
+      "loss": 0.3395,
       "step": 126060
     },
     {
-      "epoch": 4.44,
-      "learning_rate": 1.6628143501049687e-06,
-      "loss": 0.2536,
+      "epoch": 4.54337405845677,
+      "grad_norm": 0.26745355129241943,
+      "learning_rate": 1.0931358183800117e-06,
+      "loss": 0.3549,
       "step": 126065
     },
     {
-      "epoch": 4.44,
-      "learning_rate": 1.661792921295155e-06,
-      "loss": 0.2817,
+      "epoch": 4.543554258117995,
+      "grad_norm": 0.23124882578849792,
+      "learning_rate": 1.0922825134435127e-06,
+      "loss": 0.3598,
       "step": 126070
     },
     {
-      "epoch": 4.44,
-      "learning_rate": 1.660771795516583e-06,
-      "loss": 0.2533,
+      "epoch": 4.54373445777922,
+      "grad_norm": 0.32220420241355896,
+      "learning_rate": 1.0914295342450337e-06,
+      "loss": 0.361,
       "step": 126075
     },
     {
-      "epoch": 4.44,
-      "learning_rate": 1.659750972782509e-06,
-      "loss": 0.2689,
+      "epoch": 4.543914657440444,
+      "grad_norm": 0.24584202468395233,
+      "learning_rate": 1.0905768807961987e-06,
+      "loss": 0.3809,
       "step": 126080
     },
     {
-      "epoch": 4.44,
-      "learning_rate": 1.6587304531061831e-06,
-      "loss": 0.2694,
+      "epoch": 4.544094857101669,
+      "grad_norm": 0.24640391767024994,
+      "learning_rate": 1.0897245531086287e-06,
+      "loss": 0.369,
       "step": 126085
     },
     {
-      "epoch": 4.44,
-      "learning_rate": 1.6577102365008644e-06,
-      "loss": 0.2329,
+      "epoch": 4.544275056762894,
+      "grad_norm": 0.2793191373348236,
+      "learning_rate": 1.0888725511939258e-06,
+      "loss": 0.3747,
       "step": 126090
     },
     {
-      "epoch": 4.44,
-      "learning_rate": 1.6566903229797954e-06,
-      "loss": 0.2602,
+      "epoch": 4.544455256424118,
+      "grad_norm": 0.2635978162288666,
+      "learning_rate": 1.0880208750637083e-06,
+      "loss": 0.3755,
       "step": 126095
     },
     {
-      "epoch": 4.44,
-      "learning_rate": 1.6556707125562177e-06,
-      "loss": 0.2298,
+      "epoch": 4.544635456085342,
+      "grad_norm": 0.2227722406387329,
+      "learning_rate": 1.0871695247295782e-06,
+      "loss": 0.3653,
       "step": 126100
     },
     {
-      "epoch": 4.44,
-      "learning_rate": 1.654651405243368e-06,
-      "loss": 0.2428,
+      "epoch": 4.544815655746567,
+      "grad_norm": 0.23662735521793365,
+      "learning_rate": 1.0863185002031317e-06,
+      "loss": 0.3871,
       "step": 126105
     },
     {
-      "epoch": 4.44,
-      "learning_rate": 1.653632401054489e-06,
-      "loss": 0.2488,
+      "epoch": 4.544995855407792,
+      "grad_norm": 0.25815528631210327,
+      "learning_rate": 1.0854678014959679e-06,
+      "loss": 0.3697,
       "step": 126110
     },
     {
-      "epoch": 4.44,
-      "learning_rate": 1.6526137000028058e-06,
-      "loss": 0.2422,
+      "epoch": 4.545176055069017,
+      "grad_norm": 0.2185879647731781,
+      "learning_rate": 1.084617428619672e-06,
+      "loss": 0.3678,
       "step": 126115
     },
     {
-      "epoch": 4.44,
-      "learning_rate": 1.6515953021015441e-06,
-      "loss": 0.2439,
+      "epoch": 4.545356254730241,
+      "grad_norm": 0.2667752504348755,
+      "learning_rate": 1.0837673815858345e-06,
+      "loss": 0.3512,
       "step": 126120
     },
     {
-      "epoch": 4.44,
-      "learning_rate": 1.6505772073639291e-06,
-      "loss": 0.2258,
+      "epoch": 4.545536454391466,
+      "grad_norm": 0.20545603334903717,
+      "learning_rate": 1.0829176604060353e-06,
+      "loss": 0.373,
       "step": 126125
     },
     {
-      "epoch": 4.44,
-      "learning_rate": 1.6495594158031868e-06,
-      "loss": 0.2332,
+      "epoch": 4.545716654052691,
+      "grad_norm": 0.2735787332057953,
+      "learning_rate": 1.0820682650918569e-06,
+      "loss": 0.4048,
       "step": 126130
     },
     {
-      "epoch": 4.44,
-      "learning_rate": 1.6485419274325259e-06,
-      "loss": 0.2456,
+      "epoch": 4.545896853713915,
+      "grad_norm": 0.27663978934288025,
+      "learning_rate": 1.0812191956548644e-06,
+      "loss": 0.4058,
       "step": 126135
     },
     {
-      "epoch": 4.44,
-      "learning_rate": 1.6475247422651608e-06,
-      "loss": 0.2648,
+      "epoch": 4.546077053375139,
+      "grad_norm": 0.33949437737464905,
+      "learning_rate": 1.0803704521066298e-06,
+      "loss": 0.367,
       "step": 126140
     },
     {
-      "epoch": 4.44,
-      "learning_rate": 1.646507860314292e-06,
-      "loss": 0.2597,
+      "epoch": 4.546257253036364,
+      "grad_norm": 0.28756633400917053,
+      "learning_rate": 1.0795220344587131e-06,
+      "loss": 0.3924,
       "step": 126145
     },
     {
-      "epoch": 4.44,
-      "learning_rate": 1.6454912815931312e-06,
-      "loss": 0.2382,
+      "epoch": 4.546437452697589,
+      "grad_norm": 0.2777945101261139,
+      "learning_rate": 1.0786739427226827e-06,
+      "loss": 0.3827,
       "step": 126150
     },
     {
-      "epoch": 4.44,
-      "learning_rate": 1.6444750061148762e-06,
-      "loss": 0.277,
+      "epoch": 4.5466176523588135,
+      "grad_norm": 0.2093537449836731,
+      "learning_rate": 1.0778261769100905e-06,
+      "loss": 0.3825,
       "step": 126155
     },
     {
-      "epoch": 4.44,
-      "learning_rate": 1.6434590338927191e-06,
-      "loss": 0.2687,
+      "epoch": 4.546797852020038,
+      "grad_norm": 0.2346353679895401,
+      "learning_rate": 1.0769787370324802e-06,
+      "loss": 0.3685,
       "step": 126160
     },
     {
-      "epoch": 4.44,
-      "learning_rate": 1.6424433649398524e-06,
-      "loss": 0.2389,
+      "epoch": 4.546978051681263,
+      "grad_norm": 0.23039868474006653,
+      "learning_rate": 1.0761316231014061e-06,
+      "loss": 0.3617,
       "step": 126165
     },
     {
-      "epoch": 4.44,
-      "learning_rate": 1.6414279992694682e-06,
-      "loss": 0.2455,
+      "epoch": 4.547158251342488,
+      "grad_norm": 0.23052442073822021,
+      "learning_rate": 1.0752848351284011e-06,
+      "loss": 0.3888,
       "step": 126170
     },
     {
-      "epoch": 4.44,
-      "learning_rate": 1.6404129368947445e-06,
-      "loss": 0.2309,
+      "epoch": 4.5473384510037125,
+      "grad_norm": 0.26234328746795654,
+      "learning_rate": 1.0744383731250168e-06,
+      "loss": 0.3682,
       "step": 126175
     },
     {
-      "epoch": 4.44,
-      "learning_rate": 1.6393981778288658e-06,
-      "loss": 0.2248,
+      "epoch": 4.547518650664937,
+      "grad_norm": 0.25200456380844116,
+      "learning_rate": 1.0735922371027746e-06,
+      "loss": 0.3748,
       "step": 126180
     },
     {
-      "epoch": 4.44,
-      "learning_rate": 1.6383837220850045e-06,
-      "loss": 0.2573,
+      "epoch": 4.547698850326161,
+      "grad_norm": 0.31448420882225037,
+      "learning_rate": 1.0727464270732013e-06,
+      "loss": 0.3791,
       "step": 126185
     },
     {
-      "epoch": 4.44,
-      "learning_rate": 1.6373695696763391e-06,
-      "loss": 0.2841,
+      "epoch": 4.547879049987386,
+      "grad_norm": 0.22912736237049103,
+      "learning_rate": 1.0719009430478294e-06,
+      "loss": 0.3655,
       "step": 126190
     },
     {
-      "epoch": 4.44,
-      "learning_rate": 1.636355720616034e-06,
-      "loss": 0.2451,
+      "epoch": 4.5480592496486105,
+      "grad_norm": 0.23380719125270844,
+      "learning_rate": 1.0710557850381747e-06,
+      "loss": 0.3199,
       "step": 126195
     },
     {
-      "epoch": 4.44,
-      "learning_rate": 1.6353421749172453e-06,
-      "loss": 0.2418,
+      "epoch": 4.548239449309835,
+      "grad_norm": 0.23022237420082092,
+      "learning_rate": 1.0702109530557502e-06,
+      "loss": 0.3819,
       "step": 126200
     },
     {
-      "epoch": 4.44,
-      "learning_rate": 1.6343289325931489e-06,
-      "loss": 0.2396,
+      "epoch": 4.54841964897106,
+      "grad_norm": 0.2493658810853958,
+      "learning_rate": 1.069366447112069e-06,
+      "loss": 0.4007,
       "step": 126205
     },
     {
-      "epoch": 4.44,
-      "learning_rate": 1.6333159936568893e-06,
-      "loss": 0.2439,
+      "epoch": 4.548599848632285,
+      "grad_norm": 0.21301326155662537,
+      "learning_rate": 1.0685222672186357e-06,
+      "loss": 0.3805,
       "step": 126210
     },
     {
-      "epoch": 4.44,
-      "learning_rate": 1.6323033581216228e-06,
-      "loss": 0.2512,
+      "epoch": 4.5487800482935095,
+      "grad_norm": 0.299532949924469,
+      "learning_rate": 1.067678413386955e-06,
+      "loss": 0.3399,
       "step": 126215
     },
     {
-      "epoch": 4.44,
-      "learning_rate": 1.6312910260004949e-06,
-      "loss": 0.2396,
+      "epoch": 4.548960247954734,
+      "grad_norm": 0.21883751451969147,
+      "learning_rate": 1.0668348856285231e-06,
+      "loss": 0.3422,
       "step": 126220
     },
     {
-      "epoch": 4.44,
-      "learning_rate": 1.6302789973066556e-06,
-      "loss": 0.2277,
+      "epoch": 4.549140447615958,
+      "grad_norm": 0.2648196518421173,
+      "learning_rate": 1.0659916839548313e-06,
+      "loss": 0.3592,
       "step": 126225
     },
     {
-      "epoch": 4.44,
-      "learning_rate": 1.6292672720532365e-06,
-      "loss": 0.2511,
+      "epoch": 4.549320647277183,
+      "grad_norm": 0.20908887684345245,
+      "learning_rate": 1.0651488083773697e-06,
+      "loss": 0.3723,
       "step": 126230
     },
     {
-      "epoch": 4.44,
-      "learning_rate": 1.6282558502533879e-06,
-      "loss": 0.2438,
+      "epoch": 4.5495008469384075,
+      "grad_norm": 0.25661706924438477,
+      "learning_rate": 1.0643062589076186e-06,
+      "loss": 0.3885,
       "step": 126235
     },
     {
-      "epoch": 4.44,
-      "learning_rate": 1.6272447319202272e-06,
-      "loss": 0.248,
+      "epoch": 4.549681046599632,
+      "grad_norm": 0.2967999279499054,
+      "learning_rate": 1.0634640355570658e-06,
+      "loss": 0.3833,
       "step": 126240
     },
     {
-      "epoch": 4.44,
-      "learning_rate": 1.6262339170668966e-06,
-      "loss": 0.245,
+      "epoch": 4.549861246260857,
+      "grad_norm": 0.3139897286891937,
+      "learning_rate": 1.062622138337177e-06,
+      "loss": 0.3911,
       "step": 126245
     },
     {
-      "epoch": 4.44,
-      "learning_rate": 1.6252234057065136e-06,
-      "loss": 0.2582,
+      "epoch": 4.550041445922082,
+      "grad_norm": 0.2221304029226303,
+      "learning_rate": 1.0617805672594295e-06,
+      "loss": 0.369,
       "step": 126250
     },
     {
-      "epoch": 4.44,
-      "learning_rate": 1.6242131978522007e-06,
-      "loss": 0.2563,
+      "epoch": 4.550221645583306,
+      "grad_norm": 0.2807883322238922,
+      "learning_rate": 1.0609393223352887e-06,
+      "loss": 0.372,
       "step": 126255
     },
     {
-      "epoch": 4.44,
-      "learning_rate": 1.62320329351707e-06,
-      "loss": 0.2532,
+      "epoch": 4.550401845244531,
+      "grad_norm": 0.2690640389919281,
+      "learning_rate": 1.0600984035762124e-06,
+      "loss": 0.3985,
       "step": 126260
     },
     {
-      "epoch": 4.44,
-      "learning_rate": 1.6221936927142412e-06,
-      "loss": 0.2486,
+      "epoch": 4.550582044905756,
+      "grad_norm": 0.2200150489807129,
+      "learning_rate": 1.059257810993658e-06,
+      "loss": 0.3896,
       "step": 126265
     },
     {
-      "epoch": 4.44,
-      "learning_rate": 1.6211843954568206e-06,
-      "loss": 0.258,
+      "epoch": 4.550762244566981,
+      "grad_norm": 0.21496446430683136,
+      "learning_rate": 1.0584175445990884e-06,
+      "loss": 0.3742,
       "step": 126270
     },
     {
-      "epoch": 4.44,
-      "learning_rate": 1.6201754017579147e-06,
-      "loss": 0.2594,
+      "epoch": 4.5509424442282045,
+      "grad_norm": 0.27852797508239746,
+      "learning_rate": 1.0575776044039366e-06,
+      "loss": 0.3906,
       "step": 126275
     },
     {
-      "epoch": 4.44,
-      "learning_rate": 1.6191667116306208e-06,
-      "loss": 0.2616,
+      "epoch": 4.551122643889429,
+      "grad_norm": 0.27898815274238586,
+      "learning_rate": 1.0567379904196567e-06,
+      "loss": 0.3217,
       "step": 126280
     },
     {
-      "epoch": 4.44,
-      "learning_rate": 1.6181583250880372e-06,
-      "loss": 0.2547,
+      "epoch": 4.551302843550654,
+      "grad_norm": 0.25052666664123535,
+      "learning_rate": 1.0558987026576873e-06,
+      "loss": 0.3821,
       "step": 126285
     },
     {
-      "epoch": 4.44,
-      "learning_rate": 1.6171502421432615e-06,
-      "loss": 0.2448,
+      "epoch": 4.551483043211879,
+      "grad_norm": 0.306427538394928,
+      "learning_rate": 1.0550597411294633e-06,
+      "loss": 0.3712,
       "step": 126290
     },
     {
-      "epoch": 4.44,
-      "learning_rate": 1.6161424628093808e-06,
-      "loss": 0.2552,
+      "epoch": 4.551663242873103,
+      "grad_norm": 0.30519047379493713,
+      "learning_rate": 1.0542211058464146e-06,
+      "loss": 0.416,
       "step": 126295
     },
     {
-      "epoch": 4.44,
-      "learning_rate": 1.615134987099476e-06,
-      "loss": 0.2536,
+      "epoch": 4.551843442534328,
+      "grad_norm": 0.2536962032318115,
+      "learning_rate": 1.0533827968199656e-06,
+      "loss": 0.3969,
       "step": 126300
     },
     {
-      "epoch": 4.44,
-      "learning_rate": 1.6141278150266364e-06,
-      "loss": 0.2477,
+      "epoch": 4.552023642195553,
+      "grad_norm": 0.27985987067222595,
+      "learning_rate": 1.0525448140615374e-06,
+      "loss": 0.3781,
       "step": 126305
     },
     {
-      "epoch": 4.44,
-      "learning_rate": 1.6131209466039354e-06,
-      "loss": 0.2782,
+      "epoch": 4.552203841856778,
+      "grad_norm": 0.2177259922027588,
+      "learning_rate": 1.0517071575825544e-06,
+      "loss": 0.3797,
       "step": 126310
     },
     {
-      "epoch": 4.44,
-      "learning_rate": 1.6121143818444429e-06,
-      "loss": 0.2656,
+      "epoch": 4.5523840415180015,
+      "grad_norm": 0.25428178906440735,
+      "learning_rate": 1.0508698273944212e-06,
+      "loss": 0.3856,
       "step": 126315
     },
     {
-      "epoch": 4.44,
-      "learning_rate": 1.6111081207612344e-06,
-      "loss": 0.2611,
+      "epoch": 4.552564241179226,
+      "grad_norm": 0.235152468085289,
+      "learning_rate": 1.0500328235085538e-06,
+      "loss": 0.36,
       "step": 126320
     },
     {
-      "epoch": 4.44,
-      "learning_rate": 1.6101021633673747e-06,
-      "loss": 0.2602,
+      "epoch": 4.552744440840451,
+      "grad_norm": 0.22013433277606964,
+      "learning_rate": 1.0491961459363515e-06,
+      "loss": 0.3802,
       "step": 126325
     },
     {
-      "epoch": 4.44,
-      "learning_rate": 1.6090965096759225e-06,
-      "loss": 0.2545,
+      "epoch": 4.552924640501676,
+      "grad_norm": 0.277230829000473,
+      "learning_rate": 1.0483597946892104e-06,
+      "loss": 0.3904,
       "step": 126330
     },
     {
-      "epoch": 4.44,
-      "learning_rate": 1.6080911596999316e-06,
-      "loss": 0.2484,
+      "epoch": 4.5531048401629,
+      "grad_norm": 0.25709518790245056,
+      "learning_rate": 1.0475237697785328e-06,
+      "loss": 0.3708,
       "step": 126335
     },
     {
-      "epoch": 4.44,
-      "learning_rate": 1.6070861134524661e-06,
-      "loss": 0.2543,
+      "epoch": 4.553285039824125,
+      "grad_norm": 0.2280193567276001,
+      "learning_rate": 1.0466880712157095e-06,
+      "loss": 0.3783,
       "step": 126340
     },
     {
-      "epoch": 4.45,
-      "learning_rate": 1.6060813709465688e-06,
-      "loss": 0.2743,
+      "epoch": 4.55346523948535,
+      "grad_norm": 0.2616412937641144,
+      "learning_rate": 1.0458526990121175e-06,
+      "loss": 0.4052,
       "step": 126345
     },
     {
-      "epoch": 4.45,
-      "learning_rate": 1.6050769321952903e-06,
-      "loss": 0.2608,
+      "epoch": 4.553645439146575,
+      "grad_norm": 0.33539026975631714,
+      "learning_rate": 1.0450176531791478e-06,
+      "loss": 0.3937,
       "step": 126350
     },
     {
-      "epoch": 4.45,
-      "learning_rate": 1.604072797211667e-06,
-      "loss": 0.2474,
+      "epoch": 4.553825638807799,
+      "grad_norm": 0.21199902892112732,
+      "learning_rate": 1.0441829337281744e-06,
+      "loss": 0.3597,
       "step": 126355
     },
     {
-      "epoch": 4.45,
-      "learning_rate": 1.6030689660087416e-06,
-      "loss": 0.2777,
+      "epoch": 4.554005838469024,
+      "grad_norm": 0.24767298996448517,
+      "learning_rate": 1.0433485406705718e-06,
+      "loss": 0.3845,
       "step": 126360
     },
     {
-      "epoch": 4.45,
-      "learning_rate": 1.6020654385995482e-06,
-      "loss": 0.2635,
+      "epoch": 4.554186038130249,
+      "grad_norm": 0.2962040305137634,
+      "learning_rate": 1.0425144740177085e-06,
+      "loss": 0.3648,
       "step": 126365
     },
     {
-      "epoch": 4.45,
-      "learning_rate": 1.6010622149971178e-06,
-      "loss": 0.2895,
+      "epoch": 4.554366237791473,
+      "grad_norm": 0.2016834169626236,
+      "learning_rate": 1.041680733780942e-06,
+      "loss": 0.3578,
       "step": 126370
     },
     {
-      "epoch": 4.45,
-      "learning_rate": 1.600059295214465e-06,
-      "loss": 0.2526,
+      "epoch": 4.554546437452697,
+      "grad_norm": 0.21861585974693298,
+      "learning_rate": 1.040847319971641e-06,
+      "loss": 0.3705,
       "step": 126375
     },
     {
-      "epoch": 4.45,
-      "learning_rate": 1.5990566792646293e-06,
-      "loss": 0.2474,
+      "epoch": 4.554726637113922,
+      "grad_norm": 0.3241893947124481,
+      "learning_rate": 1.040014232601158e-06,
+      "loss": 0.3971,
       "step": 126380
     },
     {
-      "epoch": 4.45,
-      "learning_rate": 1.5980543671606173e-06,
-      "loss": 0.2581,
+      "epoch": 4.554906836775147,
+      "grad_norm": 0.24871119856834412,
+      "learning_rate": 1.0391814716808391e-06,
+      "loss": 0.3911,
       "step": 126385
     },
     {
-      "epoch": 4.45,
-      "learning_rate": 1.5970523589154462e-06,
-      "loss": 0.2419,
+      "epoch": 4.555087036436372,
+      "grad_norm": 0.20087426900863647,
+      "learning_rate": 1.0383490372220361e-06,
+      "loss": 0.3938,
       "step": 126390
     },
     {
-      "epoch": 4.45,
-      "learning_rate": 1.5960506545421305e-06,
-      "loss": 0.2387,
+      "epoch": 4.555267236097596,
+      "grad_norm": 0.2455914169549942,
+      "learning_rate": 1.0375169292360847e-06,
+      "loss": 0.3256,
       "step": 126395
     },
     {
-      "epoch": 4.45,
-      "learning_rate": 1.5950492540536681e-06,
-      "loss": 0.2547,
+      "epoch": 4.555447435758821,
+      "grad_norm": 0.2423105537891388,
+      "learning_rate": 1.0366851477343286e-06,
+      "loss": 0.3616,
       "step": 126400
     },
     {
-      "epoch": 4.45,
-      "learning_rate": 1.5940481574630712e-06,
-      "loss": 0.2519,
+      "epoch": 4.555627635420046,
+      "grad_norm": 0.2773745059967041,
+      "learning_rate": 1.0358536927280977e-06,
+      "loss": 0.3839,
       "step": 126405
     },
     {
-      "epoch": 4.45,
-      "learning_rate": 1.5930473647833315e-06,
-      "loss": 0.2622,
+      "epoch": 4.55580783508127,
+      "grad_norm": 0.2075750231742859,
+      "learning_rate": 1.0350225642287215e-06,
+      "loss": 0.3333,
       "step": 126410
     },
     {
-      "epoch": 4.45,
-      "learning_rate": 1.5920468760274449e-06,
-      "loss": 0.2612,
+      "epoch": 4.555988034742494,
+      "grad_norm": 0.28897416591644287,
+      "learning_rate": 1.0341917622475216e-06,
+      "loss": 0.3451,
       "step": 126415
     },
     {
-      "epoch": 4.45,
-      "learning_rate": 1.591046691208406e-06,
-      "loss": 0.2612,
+      "epoch": 4.556168234403719,
+      "grad_norm": 0.2481168508529663,
+      "learning_rate": 1.0333612867958197e-06,
+      "loss": 0.3703,
       "step": 126420
     },
     {
-      "epoch": 4.45,
-      "learning_rate": 1.5900468103391992e-06,
-      "loss": 0.2397,
+      "epoch": 4.556348434064944,
+      "grad_norm": 0.23715002834796906,
+      "learning_rate": 1.032531137884926e-06,
+      "loss": 0.3816,
       "step": 126425
     },
     {
-      "epoch": 4.45,
-      "learning_rate": 1.5890472334328028e-06,
-      "loss": 0.2415,
+      "epoch": 4.556528633726169,
+      "grad_norm": 0.24655555188655853,
+      "learning_rate": 1.0317013155261595e-06,
+      "loss": 0.3749,
       "step": 126430
     },
     {
-      "epoch": 4.45,
-      "learning_rate": 1.5880479605022037e-06,
-      "loss": 0.2416,
+      "epoch": 4.556708833387393,
+      "grad_norm": 0.20687216520309448,
+      "learning_rate": 1.030871819730822e-06,
+      "loss": 0.3603,
       "step": 126435
     },
     {
-      "epoch": 4.45,
-      "learning_rate": 1.587048991560372e-06,
-      "loss": 0.2499,
+      "epoch": 4.556889033048618,
+      "grad_norm": 0.2607882022857666,
+      "learning_rate": 1.0300426505102156e-06,
+      "loss": 0.3703,
       "step": 126440
     },
     {
-      "epoch": 4.45,
-      "learning_rate": 1.586050326620281e-06,
-      "loss": 0.2477,
+      "epoch": 4.557069232709843,
+      "grad_norm": 0.2784349322319031,
+      "learning_rate": 1.0292138078756396e-06,
+      "loss": 0.3719,
       "step": 126445
     },
     {
-      "epoch": 4.45,
-      "learning_rate": 1.585051965694892e-06,
-      "loss": 0.2689,
+      "epoch": 4.5572494323710675,
+      "grad_norm": 0.261934757232666,
+      "learning_rate": 1.0283852918383768e-06,
+      "loss": 0.3879,
       "step": 126450
     },
     {
-      "epoch": 4.45,
-      "learning_rate": 1.5840539087971701e-06,
-      "loss": 0.2361,
+      "epoch": 4.557429632032292,
+      "grad_norm": 0.27113890647888184,
+      "learning_rate": 1.0275571024097348e-06,
+      "loss": 0.3513,
       "step": 126455
     },
     {
-      "epoch": 4.45,
-      "learning_rate": 1.5830561559400825e-06,
-      "loss": 0.2429,
+      "epoch": 4.557609831693516,
+      "grad_norm": 0.2984657287597656,
+      "learning_rate": 1.0267292396009764e-06,
+      "loss": 0.3412,
       "step": 126460
     },
     {
-      "epoch": 4.45,
-      "learning_rate": 1.5820587071365772e-06,
-      "loss": 0.2768,
+      "epoch": 4.557790031354741,
+      "grad_norm": 0.2706381380558014,
+      "learning_rate": 1.0259017034233932e-06,
+      "loss": 0.3714,
       "step": 126465
     },
     {
-      "epoch": 4.45,
-      "learning_rate": 1.581061562399605e-06,
-      "loss": 0.234,
+      "epoch": 4.5579702310159655,
+      "grad_norm": 0.27217623591423035,
+      "learning_rate": 1.025074493888259e-06,
+      "loss": 0.3975,
       "step": 126470
     },
     {
-      "epoch": 4.45,
-      "learning_rate": 1.5800647217421166e-06,
-      "loss": 0.275,
+      "epoch": 4.55815043067719,
+      "grad_norm": 0.22170761227607727,
+      "learning_rate": 1.0242476110068426e-06,
+      "loss": 0.384,
       "step": 126475
     },
     {
-      "epoch": 4.45,
-      "learning_rate": 1.5790681851770544e-06,
-      "loss": 0.2451,
+      "epoch": 4.558330630338415,
+      "grad_norm": 0.2368282973766327,
+      "learning_rate": 1.0234210547904132e-06,
+      "loss": 0.3971,
       "step": 126480
     },
     {
-      "epoch": 4.45,
-      "learning_rate": 1.5780719527173555e-06,
-      "loss": 0.2236,
+      "epoch": 4.55851082999964,
+      "grad_norm": 0.22124750912189484,
+      "learning_rate": 1.0225948252502283e-06,
+      "loss": 0.3928,
       "step": 126485
     },
     {
-      "epoch": 4.45,
-      "learning_rate": 1.5770760243759536e-06,
-      "loss": 0.2471,
+      "epoch": 4.5586910296608645,
+      "grad_norm": 0.27596476674079895,
+      "learning_rate": 1.0217689223975425e-06,
+      "loss": 0.3498,
       "step": 126490
     },
     {
-      "epoch": 4.45,
-      "learning_rate": 1.5760804001657858e-06,
-      "loss": 0.2583,
+      "epoch": 4.558871229322089,
+      "grad_norm": 0.2869780659675598,
+      "learning_rate": 1.0209433462436164e-06,
+      "loss": 0.3905,
       "step": 126495
     },
     {
-      "epoch": 4.45,
-      "learning_rate": 1.575085080099778e-06,
-      "loss": 0.2628,
+      "epoch": 4.559051428983313,
+      "grad_norm": 0.2928743362426758,
+      "learning_rate": 1.0201180967996937e-06,
+      "loss": 0.3722,
       "step": 126500
     },
     {
-      "epoch": 4.45,
-      "eval_loss": 0.2488119751214981,
-      "eval_runtime": 10.5384,
-      "eval_samples_per_second": 9.489,
-      "eval_steps_per_second": 9.489,
+      "epoch": 4.559051428983313,
+      "eval_loss": 0.42879581451416016,
+      "eval_runtime": 3.5284,
+      "eval_samples_per_second": 28.342,
+      "eval_steps_per_second": 7.085,
       "step": 126500
     },
     {
-      "epoch": 4.45,
-      "learning_rate": 1.5740900641908501e-06,
-      "loss": 0.2759,
+      "epoch": 4.559231628644538,
+      "grad_norm": 0.2935590147972107,
+      "learning_rate": 1.0192931740770213e-06,
+      "loss": 0.388,
       "step": 126505
     },
     {
-      "epoch": 4.45,
-      "learning_rate": 1.5730953524519227e-06,
-      "loss": 0.2509,
+      "epoch": 4.5594118283057625,
+      "grad_norm": 0.22944243252277374,
+      "learning_rate": 1.018468578086837e-06,
+      "loss": 0.3727,
       "step": 126510
     },
     {
-      "epoch": 4.45,
-      "learning_rate": 1.5721009448959184e-06,
-      "loss": 0.242,
+      "epoch": 4.559592027966987,
+      "grad_norm": 0.25734102725982666,
+      "learning_rate": 1.0176443088403708e-06,
+      "loss": 0.3495,
       "step": 126515
     },
     {
-      "epoch": 4.45,
-      "learning_rate": 1.571106841535744e-06,
-      "loss": 0.2429,
+      "epoch": 4.559772227628212,
+      "grad_norm": 0.2919704020023346,
+      "learning_rate": 1.0168203663488612e-06,
+      "loss": 0.3703,
       "step": 126520
     },
     {
-      "epoch": 4.45,
-      "learning_rate": 1.5701130423843053e-06,
-      "loss": 0.2498,
+      "epoch": 4.559952427289437,
+      "grad_norm": 0.29848772287368774,
+      "learning_rate": 1.0159967506235297e-06,
+      "loss": 0.3914,
       "step": 126525
     },
     {
-      "epoch": 4.45,
-      "learning_rate": 1.5691195474545062e-06,
-      "loss": 0.2651,
+      "epoch": 4.5601326269506615,
+      "grad_norm": 0.28151485323905945,
+      "learning_rate": 1.0151734616756003e-06,
+      "loss": 0.356,
       "step": 126530
     },
     {
-      "epoch": 4.45,
-      "learning_rate": 1.5681263567592502e-06,
-      "loss": 0.2381,
+      "epoch": 4.560312826611886,
+      "grad_norm": 0.387460857629776,
+      "learning_rate": 1.0143504995162895e-06,
+      "loss": 0.3724,
       "step": 126535
     },
     {
-      "epoch": 4.45,
-      "learning_rate": 1.5671334703114327e-06,
-      "loss": 0.2313,
+      "epoch": 4.560493026273111,
+      "grad_norm": 0.221677765250206,
+      "learning_rate": 1.0135278641568046e-06,
+      "loss": 0.3535,
       "step": 126540
     },
     {
-      "epoch": 4.45,
-      "learning_rate": 1.5661408881239404e-06,
-      "loss": 0.2446,
+      "epoch": 4.560673225934336,
+      "grad_norm": 0.3218268156051636,
+      "learning_rate": 1.012705555608362e-06,
+      "loss": 0.3796,
       "step": 126545
     },
     {
-      "epoch": 4.45,
-      "learning_rate": 1.5651486102096686e-06,
-      "loss": 0.2733,
+      "epoch": 4.5608534255955595,
+      "grad_norm": 0.2881567180156708,
+      "learning_rate": 1.0118835738821664e-06,
+      "loss": 0.3839,
       "step": 126550
     },
     {
-      "epoch": 4.45,
-      "learning_rate": 1.5641566365814958e-06,
-      "loss": 0.262,
+      "epoch": 4.561033625256784,
+      "grad_norm": 0.25800442695617676,
+      "learning_rate": 1.0110619189894061e-06,
+      "loss": 0.3872,
       "step": 126555
     },
     {
-      "epoch": 4.45,
-      "learning_rate": 1.5631649672523036e-06,
-      "loss": 0.2356,
+      "epoch": 4.561213824918009,
+      "grad_norm": 0.2594134509563446,
+      "learning_rate": 1.010240590941286e-06,
+      "loss": 0.3691,
       "step": 126560
     },
     {
-      "epoch": 4.45,
-      "learning_rate": 1.5621736022349676e-06,
-      "loss": 0.2432,
+      "epoch": 4.561394024579234,
+      "grad_norm": 0.22755463421344757,
+      "learning_rate": 1.0094195897489944e-06,
+      "loss": 0.3766,
       "step": 126565
     },
     {
-      "epoch": 4.45,
-      "learning_rate": 1.5611825415423637e-06,
-      "loss": 0.269,
+      "epoch": 4.561574224240458,
+      "grad_norm": 0.21717192232608795,
+      "learning_rate": 1.008598915423714e-06,
+      "loss": 0.3823,
       "step": 126570
     },
     {
-      "epoch": 4.45,
-      "learning_rate": 1.5601917851873593e-06,
-      "loss": 0.2697,
+      "epoch": 4.561754423901683,
+      "grad_norm": 0.21237117052078247,
+      "learning_rate": 1.0077785679766304e-06,
+      "loss": 0.3484,
       "step": 126575
     },
     {
-      "epoch": 4.45,
-      "learning_rate": 1.5592013331828193e-06,
-      "loss": 0.2502,
+      "epoch": 4.561934623562908,
+      "grad_norm": 0.24171051383018494,
+      "learning_rate": 1.0069585474189126e-06,
+      "loss": 0.4059,
       "step": 126580
     },
     {
-      "epoch": 4.45,
-      "learning_rate": 1.5582111855415943e-06,
-      "loss": 0.23,
+      "epoch": 4.562114823224133,
+      "grad_norm": 0.24933667480945587,
+      "learning_rate": 1.0061388537617456e-06,
+      "loss": 0.3449,
       "step": 126585
     },
     {
-      "epoch": 4.45,
-      "learning_rate": 1.5572213422765548e-06,
-      "loss": 0.2646,
+      "epoch": 4.5622950228853565,
+      "grad_norm": 0.27278992533683777,
+      "learning_rate": 1.0053194870162901e-06,
+      "loss": 0.3666,
       "step": 126590
     },
     {
-      "epoch": 4.45,
-      "learning_rate": 1.5562318034005457e-06,
-      "loss": 0.2634,
+      "epoch": 4.562475222546581,
+      "grad_norm": 0.21958884596824646,
+      "learning_rate": 1.0045004471937125e-06,
+      "loss": 0.3687,
       "step": 126595
     },
     {
-      "epoch": 4.45,
-      "learning_rate": 1.5552425689264156e-06,
-      "loss": 0.2501,
+      "epoch": 4.562655422207806,
+      "grad_norm": 0.2822725474834442,
+      "learning_rate": 1.0036817343051674e-06,
+      "loss": 0.3441,
       "step": 126600
     },
     {
-      "epoch": 4.45,
-      "learning_rate": 1.5542536388670065e-06,
-      "loss": 0.2284,
+      "epoch": 4.562835621869031,
+      "grad_norm": 0.2642861604690552,
+      "learning_rate": 1.0028633483618154e-06,
+      "loss": 0.3919,
       "step": 126605
     },
     {
-      "epoch": 4.45,
-      "learning_rate": 1.5532650132351667e-06,
-      "loss": 0.2591,
+      "epoch": 4.563015821530255,
+      "grad_norm": 0.2586185336112976,
+      "learning_rate": 1.0020452893748006e-06,
+      "loss": 0.3887,
       "step": 126610
     },
     {
-      "epoch": 4.45,
-      "learning_rate": 1.5522766920437275e-06,
-      "loss": 0.2355,
+      "epoch": 4.56319602119148,
+      "grad_norm": 0.2781605124473572,
+      "learning_rate": 1.0012275573552748e-06,
+      "loss": 0.3715,
       "step": 126615
     },
     {
-      "epoch": 4.45,
-      "learning_rate": 1.5512886753055233e-06,
-      "loss": 0.2518,
+      "epoch": 4.563376220852705,
+      "grad_norm": 0.27998635172843933,
+      "learning_rate": 1.0004101523143794e-06,
+      "loss": 0.383,
       "step": 126620
     },
     {
-      "epoch": 4.46,
-      "learning_rate": 1.5503009630333799e-06,
-      "loss": 0.2802,
+      "epoch": 4.56355642051393,
+      "grad_norm": 0.29203617572784424,
+      "learning_rate": 9.99593074263247e-07,
+      "loss": 0.3644,
       "step": 126625
     },
     {
-      "epoch": 4.46,
-      "learning_rate": 1.5493135552401284e-06,
-      "loss": 0.2514,
+      "epoch": 4.563736620175154,
+      "grad_norm": 0.2884252965450287,
+      "learning_rate": 9.987763232130132e-07,
+      "loss": 0.4022,
       "step": 126630
     },
     {
-      "epoch": 4.46,
-      "learning_rate": 1.5483264519385864e-06,
-      "loss": 0.2342,
+      "epoch": 4.563916819836379,
+      "grad_norm": 0.26518526673316956,
+      "learning_rate": 9.979598991748023e-07,
+      "loss": 0.3545,
       "step": 126635
     },
     {
-      "epoch": 4.46,
-      "learning_rate": 1.547339653141569e-06,
-      "loss": 0.2499,
+      "epoch": 4.564097019497604,
+      "grad_norm": 0.28502172231674194,
+      "learning_rate": 9.971438021597473e-07,
+      "loss": 0.3894,
       "step": 126640
     },
     {
-      "epoch": 4.46,
-      "learning_rate": 1.5463531588618878e-06,
-      "loss": 0.2665,
+      "epoch": 4.564277219158828,
+      "grad_norm": 0.2587156593799591,
+      "learning_rate": 9.96328032178953e-07,
+      "loss": 0.3615,
       "step": 126645
     },
     {
-      "epoch": 4.46,
-      "learning_rate": 1.545366969112355e-06,
-      "loss": 0.2529,
+      "epoch": 4.564457418820052,
+      "grad_norm": 0.2818557024002075,
+      "learning_rate": 9.955125892435468e-07,
+      "loss": 0.3558,
       "step": 126650
     },
     {
-      "epoch": 4.46,
-      "learning_rate": 1.544381083905777e-06,
-      "loss": 0.2425,
+      "epoch": 4.564637618481277,
+      "grad_norm": 0.2464769035577774,
+      "learning_rate": 9.946974733646335e-07,
+      "loss": 0.3825,
       "step": 126655
     },
     {
-      "epoch": 4.46,
-      "learning_rate": 1.5433955032549546e-06,
-      "loss": 0.248,
+      "epoch": 4.564817818142502,
+      "grad_norm": 0.21170103549957275,
+      "learning_rate": 9.938826845533183e-07,
+      "loss": 0.3545,
       "step": 126660
     },
     {
-      "epoch": 4.46,
-      "learning_rate": 1.5424102271726776e-06,
-      "loss": 0.2423,
+      "epoch": 4.564998017803727,
+      "grad_norm": 0.22021377086639404,
+      "learning_rate": 9.930682228207089e-07,
+      "loss": 0.3835,
       "step": 126665
     },
     {
-      "epoch": 4.46,
-      "learning_rate": 1.541425255671744e-06,
-      "loss": 0.2585,
+      "epoch": 4.565178217464951,
+      "grad_norm": 0.24116267263889313,
+      "learning_rate": 9.922540881778936e-07,
+      "loss": 0.3767,
       "step": 126670
     },
     {
-      "epoch": 4.46,
-      "learning_rate": 1.5404405887649493e-06,
-      "loss": 0.2577,
+      "epoch": 4.565358417126176,
+      "grad_norm": 0.20193111896514893,
+      "learning_rate": 9.914402806359635e-07,
+      "loss": 0.3498,
       "step": 126675
     },
     {
-      "epoch": 4.46,
-      "learning_rate": 1.5394562264650665e-06,
-      "loss": 0.2412,
+      "epoch": 4.565538616787401,
+      "grad_norm": 0.3153676986694336,
+      "learning_rate": 9.906268002060155e-07,
+      "loss": 0.3707,
       "step": 126680
     },
     {
-      "epoch": 4.46,
-      "learning_rate": 1.5384721687848908e-06,
-      "loss": 0.2255,
+      "epoch": 4.565718816448625,
+      "grad_norm": 0.21089020371437073,
+      "learning_rate": 9.898136468991293e-07,
+      "loss": 0.362,
       "step": 126685
     },
     {
-      "epoch": 4.46,
-      "learning_rate": 1.53748841573719e-06,
-      "loss": 0.2443,
+      "epoch": 4.565899016109849,
+      "grad_norm": 0.22067415714263916,
+      "learning_rate": 9.890008207263852e-07,
+      "loss": 0.3148,
       "step": 126690
     },
     {
-      "epoch": 4.46,
-      "learning_rate": 1.5365049673347397e-06,
-      "loss": 0.2799,
+      "epoch": 4.566079215771074,
+      "grad_norm": 0.27838194370269775,
+      "learning_rate": 9.881883216988547e-07,
+      "loss": 0.3914,
       "step": 126695
     },
     {
-      "epoch": 4.46,
-      "learning_rate": 1.5355218235903051e-06,
-      "loss": 0.2415,
+      "epoch": 4.566259415432299,
+      "grad_norm": 0.2392108291387558,
+      "learning_rate": 9.87376149827607e-07,
+      "loss": 0.3841,
       "step": 126700
     },
     {
-      "epoch": 4.46,
-      "learning_rate": 1.534538984516662e-06,
-      "loss": 0.2554,
+      "epoch": 4.566439615093524,
+      "grad_norm": 0.2691670358181,
+      "learning_rate": 9.865643051237134e-07,
+      "loss": 0.3909,
       "step": 126705
     },
     {
-      "epoch": 4.46,
-      "learning_rate": 1.5335564501265637e-06,
-      "loss": 0.2415,
+      "epoch": 4.566619814754748,
+      "grad_norm": 0.29195481538772583,
+      "learning_rate": 9.857527875982292e-07,
+      "loss": 0.4007,
       "step": 126710
     },
     {
-      "epoch": 4.46,
-      "learning_rate": 1.5325742204327725e-06,
-      "loss": 0.2521,
+      "epoch": 4.566800014415973,
+      "grad_norm": 0.2075537145137787,
+      "learning_rate": 9.849415972622178e-07,
+      "loss": 0.3993,
       "step": 126715
     },
     {
-      "epoch": 4.46,
-      "learning_rate": 1.5315922954480338e-06,
-      "loss": 0.2632,
+      "epoch": 4.566980214077198,
+      "grad_norm": 0.24173662066459656,
+      "learning_rate": 9.841307341267258e-07,
+      "loss": 0.3896,
       "step": 126720
     },
     {
-      "epoch": 4.46,
-      "learning_rate": 1.5306106751851013e-06,
-      "loss": 0.2467,
+      "epoch": 4.5671604137384225,
+      "grad_norm": 0.21936601400375366,
+      "learning_rate": 9.833201982028e-07,
+      "loss": 0.369,
       "step": 126725
     },
     {
-      "epoch": 4.46,
-      "learning_rate": 1.5296293596567284e-06,
-      "loss": 0.2427,
+      "epoch": 4.567340613399647,
+      "grad_norm": 0.2615318298339844,
+      "learning_rate": 9.8250998950149e-07,
+      "loss": 0.3511,
       "step": 126730
     },
     {
-      "epoch": 4.46,
-      "learning_rate": 1.5286483488756499e-06,
-      "loss": 0.2531,
+      "epoch": 4.567520813060871,
+      "grad_norm": 0.23326557874679565,
+      "learning_rate": 9.81700108033834e-07,
+      "loss": 0.347,
       "step": 126735
     },
     {
-      "epoch": 4.46,
-      "learning_rate": 1.5276676428546022e-06,
-      "loss": 0.2361,
+      "epoch": 4.567701012722096,
+      "grad_norm": 0.26399147510528564,
+      "learning_rate": 9.808905538108593e-07,
+      "loss": 0.3568,
       "step": 126740
     },
     {
-      "epoch": 4.46,
-      "learning_rate": 1.5266872416063226e-06,
-      "loss": 0.2521,
+      "epoch": 4.567881212383321,
+      "grad_norm": 0.23725526034832,
+      "learning_rate": 9.800813268436016e-07,
+      "loss": 0.4013,
       "step": 126745
     },
     {
-      "epoch": 4.46,
-      "learning_rate": 1.525707145143543e-06,
-      "loss": 0.2528,
+      "epoch": 4.568061412044545,
+      "grad_norm": 0.23820684850215912,
+      "learning_rate": 9.792724271430881e-07,
+      "loss": 0.3677,
       "step": 126750
     },
     {
-      "epoch": 4.46,
-      "learning_rate": 1.524727353478983e-06,
-      "loss": 0.2141,
+      "epoch": 4.56824161170577,
+      "grad_norm": 0.23823130130767822,
+      "learning_rate": 9.78463854720335e-07,
+      "loss": 0.3627,
       "step": 126755
     },
     {
-      "epoch": 4.46,
-      "learning_rate": 1.5237478666253635e-06,
-      "loss": 0.2405,
+      "epoch": 4.568421811366995,
+      "grad_norm": 0.2619285583496094,
+      "learning_rate": 9.776556095863615e-07,
+      "loss": 0.3665,
       "step": 126760
     },
     {
-      "epoch": 4.46,
-      "learning_rate": 1.522768684595413e-06,
-      "loss": 0.2489,
+      "epoch": 4.5686020110282195,
+      "grad_norm": 0.2296546995639801,
+      "learning_rate": 9.768476917521752e-07,
+      "loss": 0.3725,
       "step": 126765
     },
     {
-      "epoch": 4.46,
-      "learning_rate": 1.5217898074018356e-06,
-      "loss": 0.2644,
+      "epoch": 4.568782210689444,
+      "grad_norm": 0.24734491109848022,
+      "learning_rate": 9.76040101228795e-07,
+      "loss": 0.3575,
       "step": 126770
     },
     {
-      "epoch": 4.46,
-      "learning_rate": 1.5208112350573428e-06,
-      "loss": 0.2701,
+      "epoch": 4.568962410350668,
+      "grad_norm": 0.24005775153636932,
+      "learning_rate": 9.752328380272125e-07,
+      "loss": 0.3752,
       "step": 126775
     },
     {
-      "epoch": 4.46,
-      "learning_rate": 1.519832967574647e-06,
-      "loss": 0.2505,
+      "epoch": 4.569142610011893,
+      "grad_norm": 0.27801743149757385,
+      "learning_rate": 9.74425902158435e-07,
+      "loss": 0.4004,
       "step": 126780
     },
     {
-      "epoch": 4.46,
-      "learning_rate": 1.5188550049664434e-06,
-      "loss": 0.2438,
+      "epoch": 4.5693228096731175,
+      "grad_norm": 0.23164783418178558,
+      "learning_rate": 9.736192936334516e-07,
+      "loss": 0.3666,
       "step": 126785
     },
     {
-      "epoch": 4.46,
-      "learning_rate": 1.5178773472454333e-06,
-      "loss": 0.2178,
+      "epoch": 4.569503009334342,
+      "grad_norm": 0.2679736912250519,
+      "learning_rate": 9.72813012463253e-07,
+      "loss": 0.4026,
       "step": 126790
     },
     {
-      "epoch": 4.46,
-      "learning_rate": 1.5168999944243089e-06,
-      "loss": 0.2501,
+      "epoch": 4.569683208995567,
+      "grad_norm": 0.21138164401054382,
+      "learning_rate": 9.72007058658822e-07,
+      "loss": 0.4035,
       "step": 126795
     },
     {
-      "epoch": 4.46,
-      "learning_rate": 1.515922946515766e-06,
-      "loss": 0.2513,
+      "epoch": 4.569863408656792,
+      "grad_norm": 0.27566009759902954,
+      "learning_rate": 9.712014322311475e-07,
+      "loss": 0.3793,
       "step": 126800
     },
     {
-      "epoch": 4.46,
-      "learning_rate": 1.5149462035324857e-06,
-      "loss": 0.2651,
+      "epoch": 4.5700436083180165,
+      "grad_norm": 0.20621134340763092,
+      "learning_rate": 9.703961331912009e-07,
+      "loss": 0.367,
       "step": 126805
     },
     {
-      "epoch": 4.46,
-      "learning_rate": 1.5139697654871527e-06,
-      "loss": 0.2726,
+      "epoch": 4.570223807979241,
+      "grad_norm": 0.2810385227203369,
+      "learning_rate": 9.695911615499542e-07,
+      "loss": 0.3406,
       "step": 126810
     },
     {
-      "epoch": 4.46,
-      "learning_rate": 1.5129936323924426e-06,
-      "loss": 0.2419,
+      "epoch": 4.570404007640466,
+      "grad_norm": 0.24304364621639252,
+      "learning_rate": 9.687865173183762e-07,
+      "loss": 0.3867,
       "step": 126815
     },
     {
-      "epoch": 4.46,
-      "learning_rate": 1.5120178042610317e-06,
-      "loss": 0.2407,
+      "epoch": 4.570584207301691,
+      "grad_norm": 0.2781292200088501,
+      "learning_rate": 9.679822005074251e-07,
+      "loss": 0.4125,
       "step": 126820
     },
     {
-      "epoch": 4.46,
-      "learning_rate": 1.511042281105593e-06,
-      "loss": 0.2472,
+      "epoch": 4.5707644069629145,
+      "grad_norm": 0.2525552809238434,
+      "learning_rate": 9.671782111280698e-07,
+      "loss": 0.3521,
       "step": 126825
     },
     {
-      "epoch": 4.46,
-      "learning_rate": 1.5100670629387858e-06,
-      "loss": 0.2563,
+      "epoch": 4.570944606624139,
+      "grad_norm": 0.19801495969295502,
+      "learning_rate": 9.66374549191254e-07,
+      "loss": 0.3962,
       "step": 126830
     },
     {
-      "epoch": 4.46,
-      "learning_rate": 1.5090921497732808e-06,
-      "loss": 0.2501,
+      "epoch": 4.571124806285364,
+      "grad_norm": 0.26284393668174744,
+      "learning_rate": 9.65571214707936e-07,
+      "loss": 0.3921,
       "step": 126835
     },
     {
-      "epoch": 4.46,
-      "learning_rate": 1.5081175416217313e-06,
-      "loss": 0.2744,
+      "epoch": 4.571305005946589,
+      "grad_norm": 0.24263347685337067,
+      "learning_rate": 9.647682076890541e-07,
+      "loss": 0.3704,
       "step": 126840
     },
     {
-      "epoch": 4.46,
-      "learning_rate": 1.507143238496797e-06,
-      "loss": 0.2663,
+      "epoch": 4.5714852056078135,
+      "grad_norm": 0.2582213580608368,
+      "learning_rate": 9.639655281455496e-07,
+      "loss": 0.3755,
       "step": 126845
     },
     {
-      "epoch": 4.46,
-      "learning_rate": 1.5061692404111232e-06,
-      "loss": 0.2494,
+      "epoch": 4.571665405269038,
+      "grad_norm": 0.2347324639558792,
+      "learning_rate": 9.631631760883692e-07,
+      "loss": 0.3776,
       "step": 126850
     },
     {
-      "epoch": 4.46,
-      "learning_rate": 1.5051955473773556e-06,
-      "loss": 0.2423,
+      "epoch": 4.571845604930263,
+      "grad_norm": 0.26129037141799927,
+      "learning_rate": 9.62361151528432e-07,
+      "loss": 0.386,
       "step": 126855
     },
     {
-      "epoch": 4.46,
-      "learning_rate": 1.5042221594081447e-06,
-      "loss": 0.2425,
+      "epoch": 4.572025804591488,
+      "grad_norm": 0.3019455373287201,
+      "learning_rate": 9.615594544766681e-07,
+      "loss": 0.3658,
       "step": 126860
     },
     {
-      "epoch": 4.46,
-      "learning_rate": 1.503249076516125e-06,
-      "loss": 0.2373,
+      "epoch": 4.5722060042527115,
+      "grad_norm": 0.24731415510177612,
+      "learning_rate": 9.60758084944005e-07,
+      "loss": 0.3629,
       "step": 126865
     },
     {
-      "epoch": 4.46,
-      "learning_rate": 1.502276298713931e-06,
-      "loss": 0.2575,
+      "epoch": 4.572386203913936,
+      "grad_norm": 0.2329123318195343,
+      "learning_rate": 9.599570429413591e-07,
+      "loss": 0.3584,
       "step": 126870
     },
     {
-      "epoch": 4.46,
-      "learning_rate": 1.5013038260141887e-06,
-      "loss": 0.239,
+      "epoch": 4.572566403575161,
+      "grad_norm": 0.270871102809906,
+      "learning_rate": 9.591563284796435e-07,
+      "loss": 0.3416,
       "step": 126875
     },
     {
-      "epoch": 4.46,
-      "learning_rate": 1.5003316584295352e-06,
-      "loss": 0.2341,
+      "epoch": 4.572746603236386,
+      "grad_norm": 0.2525206208229065,
+      "learning_rate": 9.58355941569769e-07,
+      "loss": 0.3663,
       "step": 126880
     },
     {
-      "epoch": 4.46,
-      "learning_rate": 1.499359795972588e-06,
-      "loss": 0.2364,
+      "epoch": 4.57292680289761,
+      "grad_norm": 0.25082695484161377,
+      "learning_rate": 9.575558822226354e-07,
+      "loss": 0.3574,
       "step": 126885
     },
     {
-      "epoch": 4.46,
-      "learning_rate": 1.4983882386559622e-06,
-      "loss": 0.2315,
+      "epoch": 4.573107002558835,
+      "grad_norm": 0.25247716903686523,
+      "learning_rate": 9.567561504491535e-07,
+      "loss": 0.3674,
       "step": 126890
     },
     {
-      "epoch": 4.46,
-      "learning_rate": 1.4974169864922782e-06,
-      "loss": 0.2696,
+      "epoch": 4.57328720222006,
+      "grad_norm": 0.3109354078769684,
+      "learning_rate": 9.559567462602143e-07,
+      "loss": 0.3811,
       "step": 126895
     },
     {
-      "epoch": 4.46,
-      "learning_rate": 1.4964460394941482e-06,
-      "loss": 0.2497,
+      "epoch": 4.573467401881285,
+      "grad_norm": 0.26343855261802673,
+      "learning_rate": 9.551576696667092e-07,
+      "loss": 0.3876,
       "step": 126900
     },
     {
-      "epoch": 4.46,
-      "learning_rate": 1.4954753976741786e-06,
-      "loss": 0.2626,
+      "epoch": 4.573647601542509,
+      "grad_norm": 0.28081122040748596,
+      "learning_rate": 9.545186442679322e-07,
+      "loss": 0.3957,
       "step": 126905
     },
     {
-      "epoch": 4.47,
-      "learning_rate": 1.494505061044965e-06,
-      "loss": 0.2441,
+      "epoch": 4.573827801203734,
+      "grad_norm": 0.2132662832736969,
+      "learning_rate": 9.537201573736415e-07,
+      "loss": 0.373,
       "step": 126910
     },
     {
-      "epoch": 4.47,
-      "learning_rate": 1.493535029619117e-06,
-      "loss": 0.2493,
+      "epoch": 4.574008000864959,
+      "grad_norm": 0.3238392770290375,
+      "learning_rate": 9.52921998105255e-07,
+      "loss": 0.3838,
       "step": 126915
     },
     {
-      "epoch": 4.47,
-      "learning_rate": 1.4925653034092267e-06,
-      "loss": 0.2521,
+      "epoch": 4.574188200526183,
+      "grad_norm": 0.30945807695388794,
+      "learning_rate": 9.521241664736558e-07,
+      "loss": 0.3599,
       "step": 126920
     },
     {
-      "epoch": 4.47,
-      "learning_rate": 1.491595882427882e-06,
-      "loss": 0.2596,
+      "epoch": 4.574368400187407,
+      "grad_norm": 0.28242048621177673,
+      "learning_rate": 9.513266624897072e-07,
+      "loss": 0.3748,
       "step": 126925
     },
     {
-      "epoch": 4.47,
-      "learning_rate": 1.4906267666876695e-06,
-      "loss": 0.2547,
+      "epoch": 4.574548599848632,
+      "grad_norm": 0.24541223049163818,
+      "learning_rate": 9.505294861642727e-07,
+      "loss": 0.386,
       "step": 126930
     },
     {
-      "epoch": 4.47,
-      "learning_rate": 1.4896579562011793e-06,
-      "loss": 0.256,
+      "epoch": 4.574728799509857,
+      "grad_norm": 0.23810042440891266,
+      "learning_rate": 9.497326375082216e-07,
+      "loss": 0.3933,
       "step": 126935
     },
     {
-      "epoch": 4.47,
-      "learning_rate": 1.4886894509809845e-06,
-      "loss": 0.2824,
+      "epoch": 4.574908999171082,
+      "grad_norm": 0.2471175640821457,
+      "learning_rate": 9.489361165324062e-07,
+      "loss": 0.3648,
       "step": 126940
     },
     {
-      "epoch": 4.47,
-      "learning_rate": 1.4877212510396615e-06,
-      "loss": 0.2579,
+      "epoch": 4.575089198832306,
+      "grad_norm": 0.2977747321128845,
+      "learning_rate": 9.481399232476817e-07,
+      "loss": 0.4237,
       "step": 126945
     },
     {
-      "epoch": 4.47,
-      "learning_rate": 1.4867533563897807e-06,
-      "loss": 0.259,
+      "epoch": 4.575269398493531,
+      "grad_norm": 0.2620290517807007,
+      "learning_rate": 9.473440576648923e-07,
+      "loss": 0.3927,
       "step": 126950
     },
     {
-      "epoch": 4.47,
-      "learning_rate": 1.4857857670439152e-06,
-      "loss": 0.2363,
+      "epoch": 4.575449598154756,
+      "grad_norm": 0.2539403736591339,
+      "learning_rate": 9.46548519794882e-07,
+      "loss": 0.3549,
       "step": 126955
     },
     {
-      "epoch": 4.47,
-      "learning_rate": 1.4848184830146272e-06,
-      "loss": 0.2565,
+      "epoch": 4.57562979781598,
+      "grad_norm": 0.2643360197544098,
+      "learning_rate": 9.457533096484922e-07,
+      "loss": 0.3877,
       "step": 126960
     },
     {
-      "epoch": 4.47,
-      "learning_rate": 1.4838515043144734e-06,
-      "loss": 0.2329,
+      "epoch": 4.575809997477204,
+      "grad_norm": 0.22208766639232635,
+      "learning_rate": 9.449584272365585e-07,
+      "loss": 0.3733,
       "step": 126965
     },
     {
-      "epoch": 4.47,
-      "learning_rate": 1.482884830956005e-06,
-      "loss": 0.254,
+      "epoch": 4.575990197138429,
+      "grad_norm": 0.22070464491844177,
+      "learning_rate": 9.441638725699059e-07,
+      "loss": 0.3158,
       "step": 126970
     },
     {
-      "epoch": 4.47,
-      "learning_rate": 1.481918462951784e-06,
-      "loss": 0.2786,
+      "epoch": 4.576170396799654,
+      "grad_norm": 0.2639147937297821,
+      "learning_rate": 9.433696456593671e-07,
+      "loss": 0.3675,
       "step": 126975
     },
     {
-      "epoch": 4.47,
-      "learning_rate": 1.4809524003143532e-06,
-      "loss": 0.2525,
+      "epoch": 4.576350596460879,
+      "grad_norm": 0.2987234592437744,
+      "learning_rate": 9.425757465157531e-07,
+      "loss": 0.4199,
       "step": 126980
     },
     {
-      "epoch": 4.47,
-      "learning_rate": 1.4799866430562554e-06,
-      "loss": 0.2582,
+      "epoch": 4.576530796122103,
+      "grad_norm": 0.24704287946224213,
+      "learning_rate": 9.417821751498912e-07,
+      "loss": 0.3624,
       "step": 126985
     },
     {
-      "epoch": 4.47,
-      "learning_rate": 1.4790211911900277e-06,
-      "loss": 0.2658,
+      "epoch": 4.576710995783328,
+      "grad_norm": 0.23351232707500458,
+      "learning_rate": 9.409889315725895e-07,
+      "loss": 0.3265,
       "step": 126990
     },
     {
-      "epoch": 4.47,
-      "learning_rate": 1.47805604472821e-06,
-      "loss": 0.2376,
+      "epoch": 4.576891195444553,
+      "grad_norm": 0.23723867535591125,
+      "learning_rate": 9.40196015794656e-07,
+      "loss": 0.4,
       "step": 126995
     },
     {
-      "epoch": 4.47,
-      "learning_rate": 1.4770912036833313e-06,
-      "loss": 0.2564,
+      "epoch": 4.5770713951057775,
+      "grad_norm": 0.24983817338943481,
+      "learning_rate": 9.394034278268931e-07,
+      "loss": 0.3816,
       "step": 127000
     },
     {
-      "epoch": 4.47,
-      "eval_loss": 0.2487480491399765,
-      "eval_runtime": 10.5483,
-      "eval_samples_per_second": 9.48,
-      "eval_steps_per_second": 9.48,
+      "epoch": 4.5770713951057775,
+      "eval_loss": 0.42867863178253174,
+      "eval_runtime": 3.5177,
+      "eval_samples_per_second": 28.428,
+      "eval_steps_per_second": 7.107,
       "step": 127000
     },
     {
-      "epoch": 4.47,
-      "learning_rate": 1.476126668067923e-06,
-      "loss": 0.2605,
+      "epoch": 4.577251594767002,
+      "grad_norm": 0.2775065302848816,
+      "learning_rate": 9.386111676801007e-07,
+      "loss": 0.414,
       "step": 127005
     },
     {
-      "epoch": 4.47,
-      "learning_rate": 1.475162437894506e-06,
-      "loss": 0.2506,
+      "epoch": 4.577431794428226,
+      "grad_norm": 0.31633272767066956,
+      "learning_rate": 9.3781923536507e-07,
+      "loss": 0.3846,
       "step": 127010
     },
     {
-      "epoch": 4.47,
-      "learning_rate": 1.4741985131756037e-06,
-      "loss": 0.2621,
+      "epoch": 4.577611994089451,
+      "grad_norm": 0.2062579095363617,
+      "learning_rate": 9.37027630892598e-07,
+      "loss": 0.3812,
       "step": 127015
     },
     {
-      "epoch": 4.47,
-      "learning_rate": 1.4732348939237307e-06,
-      "loss": 0.2506,
+      "epoch": 4.577792193750676,
+      "grad_norm": 0.3232049345970154,
+      "learning_rate": 9.36236354273462e-07,
+      "loss": 0.3796,
       "step": 127020
     },
     {
-      "epoch": 4.47,
-      "learning_rate": 1.4722715801513937e-06,
-      "loss": 0.2394,
+      "epoch": 4.5779723934119,
+      "grad_norm": 0.24797019362449646,
+      "learning_rate": 9.35445405518448e-07,
+      "loss": 0.384,
       "step": 127025
     },
     {
-      "epoch": 4.47,
-      "learning_rate": 1.4713085718711078e-06,
-      "loss": 0.2506,
+      "epoch": 4.578152593073125,
+      "grad_norm": 0.23074999451637268,
+      "learning_rate": 9.346547846383308e-07,
+      "loss": 0.3794,
       "step": 127030
     },
     {
-      "epoch": 4.47,
-      "learning_rate": 1.4703458690953742e-06,
-      "loss": 0.2659,
+      "epoch": 4.57833279273435,
+      "grad_norm": 0.2560942769050598,
+      "learning_rate": 9.338644916438849e-07,
+      "loss": 0.3497,
       "step": 127035
     },
     {
-      "epoch": 4.47,
-      "learning_rate": 1.469383471836694e-06,
-      "loss": 0.2494,
+      "epoch": 4.5785129923955745,
+      "grad_norm": 0.20177873969078064,
+      "learning_rate": 9.33074526545874e-07,
+      "loss": 0.3644,
       "step": 127040
     },
     {
-      "epoch": 4.47,
-      "learning_rate": 1.468421380107557e-06,
-      "loss": 0.238,
+      "epoch": 4.578693192056799,
+      "grad_norm": 0.2853398025035858,
+      "learning_rate": 9.322848893550645e-07,
+      "loss": 0.3395,
       "step": 127045
     },
     {
-      "epoch": 4.47,
-      "learning_rate": 1.4674595939204643e-06,
-      "loss": 0.2531,
+      "epoch": 4.578873391718023,
+      "grad_norm": 0.26799461245536804,
+      "learning_rate": 9.314955800822117e-07,
+      "loss": 0.3263,
       "step": 127050
     },
     {
-      "epoch": 4.47,
-      "learning_rate": 1.466498113287898e-06,
-      "loss": 0.2549,
+      "epoch": 4.579053591379248,
+      "grad_norm": 0.2538415491580963,
+      "learning_rate": 9.307065987380736e-07,
+      "loss": 0.356,
       "step": 127055
     },
     {
-      "epoch": 4.47,
-      "learning_rate": 1.4655369382223478e-06,
-      "loss": 0.2312,
+      "epoch": 4.579233791040473,
+      "grad_norm": 0.2968897819519043,
+      "learning_rate": 9.299179453333973e-07,
+      "loss": 0.3384,
       "step": 127060
     },
     {
-      "epoch": 4.47,
-      "learning_rate": 1.4645760687362842e-06,
-      "loss": 0.2283,
+      "epoch": 4.579413990701697,
+      "grad_norm": 0.2750302255153656,
+      "learning_rate": 9.291296198789296e-07,
+      "loss": 0.3623,
       "step": 127065
     },
     {
-      "epoch": 4.47,
-      "learning_rate": 1.4636155048421974e-06,
-      "loss": 0.2538,
+      "epoch": 4.579594190362922,
+      "grad_norm": 0.2515812814235687,
+      "learning_rate": 9.283416223854119e-07,
+      "loss": 0.3995,
       "step": 127070
     },
     {
-      "epoch": 4.47,
-      "learning_rate": 1.4626552465525495e-06,
-      "loss": 0.2794,
+      "epoch": 4.579774390024147,
+      "grad_norm": 0.2370171695947647,
+      "learning_rate": 9.275539528635746e-07,
+      "loss": 0.3846,
       "step": 127075
     },
     {
-      "epoch": 4.47,
-      "learning_rate": 1.461695293879814e-06,
-      "loss": 0.274,
+      "epoch": 4.5799545896853715,
+      "grad_norm": 0.2444002330303192,
+      "learning_rate": 9.267666113241563e-07,
+      "loss": 0.3358,
       "step": 127080
     },
     {
-      "epoch": 4.47,
-      "learning_rate": 1.4607356468364475e-06,
-      "loss": 0.2674,
+      "epoch": 4.580134789346596,
+      "grad_norm": 0.25110915303230286,
+      "learning_rate": 9.259795977778846e-07,
+      "loss": 0.3625,
       "step": 127085
     },
     {
-      "epoch": 4.47,
-      "learning_rate": 1.4597763054349179e-06,
-      "loss": 0.2542,
+      "epoch": 4.580314989007821,
+      "grad_norm": 0.2334173619747162,
+      "learning_rate": 9.251929122354785e-07,
+      "loss": 0.3718,
       "step": 127090
     },
     {
-      "epoch": 4.47,
-      "learning_rate": 1.4588172696876817e-06,
-      "loss": 0.261,
+      "epoch": 4.580495188669046,
+      "grad_norm": 0.3335031270980835,
+      "learning_rate": 9.244065547076574e-07,
+      "loss": 0.3461,
       "step": 127095
     },
     {
-      "epoch": 4.47,
-      "learning_rate": 1.4578585396071904e-06,
-      "loss": 0.2515,
+      "epoch": 4.58067538833027,
+      "grad_norm": 0.286377489566803,
+      "learning_rate": 9.23620525205135e-07,
+      "loss": 0.3706,
       "step": 127100
     },
     {
-      "epoch": 4.47,
-      "learning_rate": 1.4569001152058837e-06,
-      "loss": 0.2512,
+      "epoch": 4.580855587991494,
+      "grad_norm": 0.28314265608787537,
+      "learning_rate": 9.228348237386248e-07,
+      "loss": 0.3618,
       "step": 127105
     },
     {
-      "epoch": 4.47,
-      "learning_rate": 1.4559419964962156e-06,
-      "loss": 0.272,
+      "epoch": 4.581035787652719,
+      "grad_norm": 0.29264768958091736,
+      "learning_rate": 9.220494503188265e-07,
+      "loss": 0.3632,
       "step": 127110
     },
     {
-      "epoch": 4.47,
-      "learning_rate": 1.4549841834906292e-06,
-      "loss": 0.235,
+      "epoch": 4.581215987313944,
+      "grad_norm": 0.22267203032970428,
+      "learning_rate": 9.212644049564401e-07,
+      "loss": 0.3707,
       "step": 127115
     },
     {
-      "epoch": 4.47,
-      "learning_rate": 1.4540266762015531e-06,
-      "loss": 0.2546,
+      "epoch": 4.5813961869751685,
+      "grad_norm": 0.26017507910728455,
+      "learning_rate": 9.204796876621679e-07,
+      "loss": 0.3522,
       "step": 127120
     },
     {
-      "epoch": 4.47,
-      "learning_rate": 1.453069474641422e-06,
-      "loss": 0.239,
+      "epoch": 4.581576386636393,
+      "grad_norm": 0.297837495803833,
+      "learning_rate": 9.196952984466961e-07,
+      "loss": 0.3786,
       "step": 127125
     },
     {
-      "epoch": 4.47,
-      "learning_rate": 1.4521125788226703e-06,
-      "loss": 0.2469,
+      "epoch": 4.581756586297618,
+      "grad_norm": 0.23576004803180695,
+      "learning_rate": 9.189112373207188e-07,
+      "loss": 0.3523,
       "step": 127130
     },
     {
-      "epoch": 4.47,
-      "learning_rate": 1.4511559887577158e-06,
-      "loss": 0.2428,
+      "epoch": 4.581936785958843,
+      "grad_norm": 0.22233489155769348,
+      "learning_rate": 9.181275042949078e-07,
+      "loss": 0.3784,
       "step": 127135
     },
     {
-      "epoch": 4.47,
-      "learning_rate": 1.4501997044589793e-06,
-      "loss": 0.2392,
+      "epoch": 4.582116985620067,
+      "grad_norm": 0.19266854226589203,
+      "learning_rate": 9.173440993799492e-07,
+      "loss": 0.3736,
       "step": 127140
     },
     {
-      "epoch": 4.47,
-      "learning_rate": 1.449243725938884e-06,
-      "loss": 0.2742,
+      "epoch": 4.582297185281291,
+      "grad_norm": 0.2523898482322693,
+      "learning_rate": 9.165610225865152e-07,
+      "loss": 0.3829,
       "step": 127145
     },
     {
-      "epoch": 4.47,
-      "learning_rate": 1.4482880532098369e-06,
-      "loss": 0.2383,
+      "epoch": 4.582477384942516,
+      "grad_norm": 0.26051437854766846,
+      "learning_rate": 9.157782739252718e-07,
+      "loss": 0.3785,
       "step": 127150
     },
     {
-      "epoch": 4.47,
-      "learning_rate": 1.4473326862842474e-06,
-      "loss": 0.2642,
+      "epoch": 4.582657584603741,
+      "grad_norm": 0.3132672607898712,
+      "learning_rate": 9.149958534068915e-07,
+      "loss": 0.349,
       "step": 127155
     },
     {
-      "epoch": 4.47,
-      "learning_rate": 1.4463776251745191e-06,
-      "loss": 0.2464,
+      "epoch": 4.5828377842649655,
+      "grad_norm": 0.24046078324317932,
+      "learning_rate": 9.142137610420265e-07,
+      "loss": 0.3656,
       "step": 127160
     },
     {
-      "epoch": 4.47,
-      "learning_rate": 1.4454228698930567e-06,
-      "loss": 0.2648,
+      "epoch": 4.58301798392619,
+      "grad_norm": 0.2993525564670563,
+      "learning_rate": 9.134319968413325e-07,
+      "loss": 0.3774,
       "step": 127165
     },
     {
-      "epoch": 4.47,
-      "learning_rate": 1.4444684204522551e-06,
-      "loss": 0.2388,
+      "epoch": 4.583198183587415,
+      "grad_norm": 0.24848264455795288,
+      "learning_rate": 9.1265056081547e-07,
+      "loss": 0.4103,
       "step": 127170
     },
     {
-      "epoch": 4.47,
-      "learning_rate": 1.4435142768645106e-06,
-      "loss": 0.2547,
+      "epoch": 4.58337838324864,
+      "grad_norm": 0.24963483214378357,
+      "learning_rate": 9.118694529750782e-07,
+      "loss": 0.3906,
       "step": 127175
     },
     {
-      "epoch": 4.47,
-      "learning_rate": 1.4425604391422043e-06,
-      "loss": 0.265,
+      "epoch": 4.583558582909864,
+      "grad_norm": 0.3283641040325165,
+      "learning_rate": 9.11088673330801e-07,
+      "loss": 0.4036,
       "step": 127180
     },
     {
-      "epoch": 4.47,
-      "learning_rate": 1.4416069072977266e-06,
-      "loss": 0.2466,
+      "epoch": 4.583738782571089,
+      "grad_norm": 0.24759909510612488,
+      "learning_rate": 9.103082218932774e-07,
+      "loss": 0.3573,
       "step": 127185
     },
     {
-      "epoch": 4.47,
-      "learning_rate": 1.4406536813434624e-06,
-      "loss": 0.2674,
+      "epoch": 4.583918982232314,
+      "grad_norm": 0.2345927208662033,
+      "learning_rate": 9.09528098673143e-07,
+      "loss": 0.3574,
       "step": 127190
     },
     {
-      "epoch": 4.48,
-      "learning_rate": 1.4397007612917817e-06,
-      "loss": 0.2396,
+      "epoch": 4.584099181893538,
+      "grad_norm": 0.2747328579425812,
+      "learning_rate": 9.087483036810174e-07,
+      "loss": 0.3511,
       "step": 127195
     },
     {
-      "epoch": 4.48,
-      "learning_rate": 1.4387481471550557e-06,
-      "loss": 0.2615,
+      "epoch": 4.5842793815547624,
+      "grad_norm": 0.2693836987018585,
+      "learning_rate": 9.07968836927539e-07,
+      "loss": 0.3638,
       "step": 127200
     },
     {
-      "epoch": 4.48,
-      "learning_rate": 1.4377958389456604e-06,
-      "loss": 0.2407,
+      "epoch": 4.584459581215987,
+      "grad_norm": 0.29041051864624023,
+      "learning_rate": 9.071896984233163e-07,
+      "loss": 0.4105,
       "step": 127205
     },
     {
-      "epoch": 4.48,
-      "learning_rate": 1.4368438366759606e-06,
-      "loss": 0.2464,
+      "epoch": 4.584639780877212,
+      "grad_norm": 0.23431363701820374,
+      "learning_rate": 9.064108881789712e-07,
+      "loss": 0.3591,
       "step": 127210
     },
     {
-      "epoch": 4.48,
-      "learning_rate": 1.435892140358308e-06,
-      "loss": 0.2432,
+      "epoch": 4.584819980538437,
+      "grad_norm": 0.21707414090633392,
+      "learning_rate": 9.056324062051147e-07,
+      "loss": 0.3548,
       "step": 127215
     },
     {
-      "epoch": 4.48,
-      "learning_rate": 1.4349407500050731e-06,
-      "loss": 0.2628,
+      "epoch": 4.585000180199661,
+      "grad_norm": 0.2566896378993988,
+      "learning_rate": 9.048542525123493e-07,
+      "loss": 0.3801,
       "step": 127220
     },
     {
-      "epoch": 4.48,
-      "learning_rate": 1.4339896656285989e-06,
-      "loss": 0.2422,
+      "epoch": 4.585180379860886,
+      "grad_norm": 0.22780990600585938,
+      "learning_rate": 9.040764271112806e-07,
+      "loss": 0.3584,
       "step": 127225
     },
     {
-      "epoch": 4.48,
-      "learning_rate": 1.4330388872412392e-06,
-      "loss": 0.2442,
+      "epoch": 4.585360579522111,
+      "grad_norm": 0.28133270144462585,
+      "learning_rate": 9.032989300125055e-07,
+      "loss": 0.3961,
       "step": 127230
     },
     {
-      "epoch": 4.48,
-      "learning_rate": 1.4320884148553427e-06,
-      "loss": 0.2485,
+      "epoch": 4.585540779183335,
+      "grad_norm": 0.24993495643138885,
+      "learning_rate": 9.025217612266157e-07,
+      "loss": 0.3699,
       "step": 127235
     },
     {
-      "epoch": 4.48,
-      "learning_rate": 1.431138248483238e-06,
-      "loss": 0.2313,
+      "epoch": 4.585720978844559,
+      "grad_norm": 0.20704279839992523,
+      "learning_rate": 9.017449207642026e-07,
+      "loss": 0.3575,
       "step": 127240
     },
     {
-      "epoch": 4.48,
-      "learning_rate": 1.4301883881372768e-06,
-      "loss": 0.2543,
+      "epoch": 4.585901178505784,
+      "grad_norm": 0.22936898469924927,
+      "learning_rate": 9.009684086358494e-07,
+      "loss": 0.3541,
       "step": 127245
     },
     {
-      "epoch": 4.48,
-      "learning_rate": 1.4292388338297852e-06,
-      "loss": 0.2537,
+      "epoch": 4.586081378167009,
+      "grad_norm": 0.26734253764152527,
+      "learning_rate": 9.001922248521366e-07,
+      "loss": 0.399,
       "step": 127250
     },
     {
-      "epoch": 4.48,
-      "learning_rate": 1.428289585573092e-06,
-      "loss": 0.2665,
+      "epoch": 4.586261577828234,
+      "grad_norm": 0.26072803139686584,
+      "learning_rate": 8.994163694236391e-07,
+      "loss": 0.3603,
       "step": 127255
     },
     {
-      "epoch": 4.48,
-      "learning_rate": 1.4273406433795267e-06,
-      "loss": 0.223,
+      "epoch": 4.586441777489458,
+      "grad_norm": 0.21624024212360382,
+      "learning_rate": 8.98640842360926e-07,
+      "loss": 0.3603,
       "step": 127260
     },
     {
-      "epoch": 4.48,
-      "learning_rate": 1.4263920072614068e-06,
-      "loss": 0.2692,
+      "epoch": 4.586621977150683,
+      "grad_norm": 0.2746710479259491,
+      "learning_rate": 8.97865643674567e-07,
+      "loss": 0.3696,
       "step": 127265
     },
     {
-      "epoch": 4.48,
-      "learning_rate": 1.4254436772310503e-06,
-      "loss": 0.2491,
+      "epoch": 4.586802176811908,
+      "grad_norm": 0.2964709997177124,
+      "learning_rate": 8.970907733751199e-07,
+      "loss": 0.3768,
       "step": 127270
     },
     {
-      "epoch": 4.48,
-      "learning_rate": 1.4244956533007726e-06,
-      "loss": 0.2294,
+      "epoch": 4.586982376473133,
+      "grad_norm": 0.2429097592830658,
+      "learning_rate": 8.96316231473146e-07,
+      "loss": 0.3656,
       "step": 127275
     },
     {
-      "epoch": 4.48,
-      "learning_rate": 1.42354793548288e-06,
-      "loss": 0.2604,
+      "epoch": 4.587162576134357,
+      "grad_norm": 0.22786717116832733,
+      "learning_rate": 8.955420179791979e-07,
+      "loss": 0.3619,
       "step": 127280
     },
     {
-      "epoch": 4.48,
-      "learning_rate": 1.4226005237896827e-06,
-      "loss": 0.2476,
+      "epoch": 4.587342775795581,
+      "grad_norm": 0.30810287594795227,
+      "learning_rate": 8.947681329038198e-07,
+      "loss": 0.3812,
       "step": 127285
     },
     {
-      "epoch": 4.48,
-      "learning_rate": 1.421653418233479e-06,
-      "loss": 0.254,
+      "epoch": 4.587522975456806,
+      "grad_norm": 0.2238711714744568,
+      "learning_rate": 8.939945762575619e-07,
+      "loss": 0.3629,
       "step": 127290
     },
     {
-      "epoch": 4.48,
-      "learning_rate": 1.4207066188265645e-06,
-      "loss": 0.256,
+      "epoch": 4.587703175118031,
+      "grad_norm": 0.22540000081062317,
+      "learning_rate": 8.932213480509627e-07,
+      "loss": 0.3756,
       "step": 127295
     },
     {
-      "epoch": 4.48,
-      "learning_rate": 1.4197601255812376e-06,
-      "loss": 0.2693,
+      "epoch": 4.587883374779255,
+      "grad_norm": 0.2444496601819992,
+      "learning_rate": 8.9244844829455e-07,
+      "loss": 0.3786,
       "step": 127300
     },
     {
-      "epoch": 4.48,
-      "learning_rate": 1.418813938509786e-06,
-      "loss": 0.2401,
+      "epoch": 4.58806357444048,
+      "grad_norm": 0.2592555582523346,
+      "learning_rate": 8.916758769988626e-07,
+      "loss": 0.3718,
       "step": 127305
     },
     {
-      "epoch": 4.48,
-      "learning_rate": 1.4178680576244967e-06,
-      "loss": 0.234,
+      "epoch": 4.588243774101705,
+      "grad_norm": 0.26862582564353943,
+      "learning_rate": 8.909036341744226e-07,
+      "loss": 0.373,
       "step": 127310
     },
     {
-      "epoch": 4.48,
-      "learning_rate": 1.4169224829376438e-06,
-      "loss": 0.2692,
+      "epoch": 4.5884239737629295,
+      "grad_norm": 0.24260741472244263,
+      "learning_rate": 8.901317198317577e-07,
+      "loss": 0.3717,
       "step": 127315
     },
     {
-      "epoch": 4.48,
-      "learning_rate": 1.4159772144615142e-06,
-      "loss": 0.2418,
+      "epoch": 4.588604173424154,
+      "grad_norm": 0.24818958342075348,
+      "learning_rate": 8.893601339813762e-07,
+      "loss": 0.3762,
       "step": 127320
     },
     {
-      "epoch": 4.48,
-      "learning_rate": 1.4150322522083787e-06,
-      "loss": 0.227,
+      "epoch": 4.588784373085378,
+      "grad_norm": 0.2344391644001007,
+      "learning_rate": 8.885888766337947e-07,
+      "loss": 0.39,
       "step": 127325
     },
     {
-      "epoch": 4.48,
-      "learning_rate": 1.4140875961905026e-06,
-      "loss": 0.2393,
+      "epoch": 4.588964572746603,
+      "grad_norm": 0.28583937883377075,
+      "learning_rate": 8.878179477995214e-07,
+      "loss": 0.3784,
       "step": 127330
     },
     {
-      "epoch": 4.48,
-      "learning_rate": 1.4131432464201565e-06,
-      "loss": 0.2483,
+      "epoch": 4.589144772407828,
+      "grad_norm": 0.24101144075393677,
+      "learning_rate": 8.870473474890617e-07,
+      "loss": 0.3615,
       "step": 127335
     },
     {
-      "epoch": 4.48,
-      "learning_rate": 1.4121992029096031e-06,
-      "loss": 0.2745,
+      "epoch": 4.589324972069052,
+      "grad_norm": 0.2540043890476227,
+      "learning_rate": 8.86277075712913e-07,
+      "loss": 0.3619,
       "step": 127340
     },
     {
-      "epoch": 4.48,
-      "learning_rate": 1.4112554656710992e-06,
-      "loss": 0.2644,
+      "epoch": 4.589505171730277,
+      "grad_norm": 0.2533954381942749,
+      "learning_rate": 8.855071324815723e-07,
+      "loss": 0.3504,
       "step": 127345
     },
     {
-      "epoch": 4.48,
-      "learning_rate": 1.4103120347168958e-06,
-      "loss": 0.2589,
+      "epoch": 4.589685371391502,
+      "grad_norm": 0.20439958572387695,
+      "learning_rate": 8.847375178055228e-07,
+      "loss": 0.3585,
       "step": 127350
     },
     {
-      "epoch": 4.48,
-      "learning_rate": 1.4093689100592416e-06,
-      "loss": 0.2414,
+      "epoch": 4.5898655710527265,
+      "grad_norm": 0.22959500551223755,
+      "learning_rate": 8.839682316952619e-07,
+      "loss": 0.3976,
       "step": 127355
     },
     {
-      "epoch": 4.48,
-      "learning_rate": 1.4084260917103881e-06,
-      "loss": 0.2681,
+      "epoch": 4.590045770713951,
+      "grad_norm": 0.22936657071113586,
+      "learning_rate": 8.831992741612616e-07,
+      "loss": 0.3839,
       "step": 127360
     },
     {
-      "epoch": 4.48,
-      "learning_rate": 1.4074835796825753e-06,
-      "loss": 0.2525,
+      "epoch": 4.590225970375176,
+      "grad_norm": 0.2100255936384201,
+      "learning_rate": 8.824306452140025e-07,
+      "loss": 0.3651,
       "step": 127365
     },
     {
-      "epoch": 4.48,
-      "learning_rate": 1.406541373988038e-06,
-      "loss": 0.2467,
+      "epoch": 4.590406170036401,
+      "grad_norm": 0.26398009061813354,
+      "learning_rate": 8.816623448639594e-07,
+      "loss": 0.3644,
       "step": 127370
     },
     {
-      "epoch": 4.48,
-      "learning_rate": 1.4055994746390132e-06,
-      "loss": 0.2416,
+      "epoch": 4.5905863696976255,
+      "grad_norm": 0.2545202374458313,
+      "learning_rate": 8.808943731215935e-07,
+      "loss": 0.3634,
       "step": 127375
     },
     {
-      "epoch": 4.48,
-      "learning_rate": 1.4046578816477306e-06,
-      "loss": 0.2252,
+      "epoch": 4.590766569358849,
+      "grad_norm": 0.2735772430896759,
+      "learning_rate": 8.801267299973715e-07,
+      "loss": 0.3916,
       "step": 127380
     },
     {
-      "epoch": 4.48,
-      "learning_rate": 1.4037165950264136e-06,
-      "loss": 0.2356,
+      "epoch": 4.590946769020074,
+      "grad_norm": 0.2763000428676605,
+      "learning_rate": 8.793594155017598e-07,
+      "loss": 0.3839,
       "step": 127385
     },
     {
-      "epoch": 4.48,
-      "learning_rate": 1.4027756147872856e-06,
-      "loss": 0.2436,
+      "epoch": 4.591126968681299,
+      "grad_norm": 0.26428860425949097,
+      "learning_rate": 8.785924296451975e-07,
+      "loss": 0.381,
       "step": 127390
     },
     {
-      "epoch": 4.48,
-      "learning_rate": 1.4018349409425674e-06,
-      "loss": 0.2723,
+      "epoch": 4.5913071683425235,
+      "grad_norm": 0.2190413475036621,
+      "learning_rate": 8.778257724381483e-07,
+      "loss": 0.357,
       "step": 127395
     },
     {
-      "epoch": 4.48,
-      "learning_rate": 1.4008945735044716e-06,
-      "loss": 0.2374,
+      "epoch": 4.591487368003748,
+      "grad_norm": 0.2818199694156647,
+      "learning_rate": 8.770594438910512e-07,
+      "loss": 0.4062,
       "step": 127400
     },
     {
-      "epoch": 4.48,
-      "learning_rate": 1.3999545124852077e-06,
-      "loss": 0.2445,
+      "epoch": 4.591667567664973,
+      "grad_norm": 0.25005903840065,
+      "learning_rate": 8.762934440143478e-07,
+      "loss": 0.3673,
       "step": 127405
     },
     {
-      "epoch": 4.48,
-      "learning_rate": 1.399014757896977e-06,
-      "loss": 0.2511,
+      "epoch": 4.591847767326198,
+      "grad_norm": 0.2070559412240982,
+      "learning_rate": 8.755277728184796e-07,
+      "loss": 0.3492,
       "step": 127410
     },
     {
-      "epoch": 4.48,
-      "learning_rate": 1.3980753097519894e-06,
-      "loss": 0.2439,
+      "epoch": 4.592027966987422,
+      "grad_norm": 0.23622746765613556,
+      "learning_rate": 8.747624303138746e-07,
+      "loss": 0.3909,
       "step": 127415
     },
     {
-      "epoch": 4.48,
-      "learning_rate": 1.3971361680624406e-06,
-      "loss": 0.2561,
+      "epoch": 4.592208166648646,
+      "grad_norm": 0.32990074157714844,
+      "learning_rate": 8.739974165109549e-07,
+      "loss": 0.3831,
       "step": 127420
     },
     {
-      "epoch": 4.48,
-      "learning_rate": 1.3961973328405236e-06,
-      "loss": 0.2491,
+      "epoch": 4.592388366309871,
+      "grad_norm": 0.21549347043037415,
+      "learning_rate": 8.732327314201539e-07,
+      "loss": 0.3512,
       "step": 127425
     },
     {
-      "epoch": 4.48,
-      "learning_rate": 1.3952588040984232e-06,
-      "loss": 0.2633,
+      "epoch": 4.592568565971096,
+      "grad_norm": 0.23620551824569702,
+      "learning_rate": 8.724683750518853e-07,
+      "loss": 0.3524,
       "step": 127430
     },
     {
-      "epoch": 4.48,
-      "learning_rate": 1.394320581848338e-06,
-      "loss": 0.2349,
+      "epoch": 4.5927487656323205,
+      "grad_norm": 0.2451809197664261,
+      "learning_rate": 8.717043474165659e-07,
+      "loss": 0.3549,
       "step": 127435
     },
     {
-      "epoch": 4.48,
-      "learning_rate": 1.3933826661024386e-06,
-      "loss": 0.2662,
+      "epoch": 4.592928965293545,
+      "grad_norm": 0.2065710574388504,
+      "learning_rate": 8.709406485246013e-07,
+      "loss": 0.3637,
       "step": 127440
     },
     {
-      "epoch": 4.48,
-      "learning_rate": 1.3924450568729098e-06,
-      "loss": 0.2524,
+      "epoch": 4.59310916495477,
+      "grad_norm": 0.2208905667066574,
+      "learning_rate": 8.701772783863943e-07,
+      "loss": 0.3542,
       "step": 127445
     },
     {
-      "epoch": 4.48,
-      "learning_rate": 1.3915077541719224e-06,
-      "loss": 0.2417,
+      "epoch": 4.593289364615995,
+      "grad_norm": 0.2570354640483856,
+      "learning_rate": 8.69414237012356e-07,
+      "loss": 0.3631,
       "step": 127450
     },
     {
-      "epoch": 4.48,
-      "learning_rate": 1.39057075801165e-06,
-      "loss": 0.2401,
+      "epoch": 4.593469564277219,
+      "grad_norm": 0.2893807888031006,
+      "learning_rate": 8.686515244128752e-07,
+      "loss": 0.3994,
       "step": 127455
     },
     {
-      "epoch": 4.48,
-      "learning_rate": 1.389634068404258e-06,
-      "loss": 0.2373,
+      "epoch": 4.593649763938444,
+      "grad_norm": 0.2452847957611084,
+      "learning_rate": 8.678891405983464e-07,
+      "loss": 0.367,
       "step": 127460
     },
     {
-      "epoch": 4.48,
-      "learning_rate": 1.3886976853619087e-06,
-      "loss": 0.2342,
+      "epoch": 4.593829963599669,
+      "grad_norm": 0.28028956055641174,
+      "learning_rate": 8.671270855791558e-07,
+      "loss": 0.388,
       "step": 127465
     },
     {
-      "epoch": 4.48,
-      "learning_rate": 1.3877616088967565e-06,
-      "loss": 0.2402,
+      "epoch": 4.594010163260893,
+      "grad_norm": 0.2985093295574188,
+      "learning_rate": 8.66365359365684e-07,
+      "loss": 0.4046,
       "step": 127470
     },
     {
-      "epoch": 4.48,
-      "learning_rate": 1.386825839020961e-06,
-      "loss": 0.2766,
+      "epoch": 4.5941903629221175,
+      "grad_norm": 0.2665720283985138,
+      "learning_rate": 8.656039619683143e-07,
+      "loss": 0.3886,
       "step": 127475
     },
     {
-      "epoch": 4.49,
-      "learning_rate": 1.3858903757466706e-06,
-      "loss": 0.2329,
+      "epoch": 4.594370562583342,
+      "grad_norm": 0.21940386295318604,
+      "learning_rate": 8.64842893397419e-07,
+      "loss": 0.3497,
       "step": 127480
     },
     {
-      "epoch": 4.49,
-      "learning_rate": 1.3849552190860259e-06,
-      "loss": 0.2594,
+      "epoch": 4.594550762244567,
+      "grad_norm": 0.21801802515983582,
+      "learning_rate": 8.640821536633647e-07,
+      "loss": 0.3911,
       "step": 127485
     },
     {
-      "epoch": 4.49,
-      "learning_rate": 1.384020369051181e-06,
-      "loss": 0.2463,
+      "epoch": 4.594730961905792,
+      "grad_norm": 0.22866007685661316,
+      "learning_rate": 8.633217427765184e-07,
+      "loss": 0.3706,
       "step": 127490
     },
     {
-      "epoch": 4.49,
-      "learning_rate": 1.3830858256542623e-06,
-      "loss": 0.2369,
+      "epoch": 4.594911161567016,
+      "grad_norm": 0.21963122487068176,
+      "learning_rate": 8.625616607472381e-07,
+      "loss": 0.3706,
       "step": 127495
     },
     {
-      "epoch": 4.49,
-      "learning_rate": 1.3821515889074154e-06,
-      "loss": 0.2386,
+      "epoch": 4.595091361228241,
+      "grad_norm": 0.22869311273097992,
+      "learning_rate": 8.618019075858852e-07,
+      "loss": 0.3886,
       "step": 127500
     },
     {
-      "epoch": 4.49,
-      "eval_loss": 0.2487904578447342,
-      "eval_runtime": 10.5517,
-      "eval_samples_per_second": 9.477,
-      "eval_steps_per_second": 9.477,
+      "epoch": 4.595091361228241,
+      "eval_loss": 0.42893698811531067,
+      "eval_runtime": 3.5318,
+      "eval_samples_per_second": 28.314,
+      "eval_steps_per_second": 7.078,
       "step": 127500
     },
     {
-      "epoch": 4.49,
-      "learning_rate": 1.3812176588227615e-06,
-      "loss": 0.2426,
+      "epoch": 4.595271560889466,
+      "grad_norm": 0.26364636421203613,
+      "learning_rate": 8.610424833028097e-07,
+      "loss": 0.3719,
       "step": 127505
     },
     {
-      "epoch": 4.49,
-      "learning_rate": 1.3802840354124353e-06,
-      "loss": 0.2382,
+      "epoch": 4.59545176055069,
+      "grad_norm": 0.2287474125623703,
+      "learning_rate": 8.602833879083505e-07,
+      "loss": 0.382,
       "step": 127510
     },
     {
-      "epoch": 4.49,
-      "learning_rate": 1.3793507186885546e-06,
-      "loss": 0.2469,
+      "epoch": 4.5956319602119144,
+      "grad_norm": 0.2660192549228668,
+      "learning_rate": 8.595246214128605e-07,
+      "loss": 0.3737,
       "step": 127515
     },
     {
-      "epoch": 4.49,
-      "learning_rate": 1.3784177086632377e-06,
-      "loss": 0.2207,
+      "epoch": 4.595812159873139,
+      "grad_norm": 0.2630321681499481,
+      "learning_rate": 8.587661838266731e-07,
+      "loss": 0.3626,
       "step": 127520
     },
     {
-      "epoch": 4.49,
-      "learning_rate": 1.3774850053485944e-06,
-      "loss": 0.2492,
+      "epoch": 4.595992359534364,
+      "grad_norm": 0.21891890466213226,
+      "learning_rate": 8.580080751601244e-07,
+      "loss": 0.3754,
       "step": 127525
     },
     {
-      "epoch": 4.49,
-      "learning_rate": 1.3765526087567455e-06,
-      "loss": 0.2256,
+      "epoch": 4.596172559195589,
+      "grad_norm": 0.2554580271244049,
+      "learning_rate": 8.572502954235422e-07,
+      "loss": 0.4112,
       "step": 127530
     },
     {
-      "epoch": 4.49,
-      "learning_rate": 1.3756205188997922e-06,
-      "loss": 0.2693,
+      "epoch": 4.596352758856813,
+      "grad_norm": 0.2666684091091156,
+      "learning_rate": 8.564928446272463e-07,
+      "loss": 0.3948,
       "step": 127535
     },
     {
-      "epoch": 4.49,
-      "learning_rate": 1.3746887357898392e-06,
-      "loss": 0.2465,
+      "epoch": 4.596532958518038,
+      "grad_norm": 0.22191660106182098,
+      "learning_rate": 8.557357227815615e-07,
+      "loss": 0.3486,
       "step": 127540
     },
     {
-      "epoch": 4.49,
-      "learning_rate": 1.3737572594389765e-06,
-      "loss": 0.254,
+      "epoch": 4.596713158179263,
+      "grad_norm": 0.2158440500497818,
+      "learning_rate": 8.549789298968075e-07,
+      "loss": 0.3562,
       "step": 127545
     },
     {
-      "epoch": 4.49,
-      "learning_rate": 1.3728260898593082e-06,
-      "loss": 0.2624,
+      "epoch": 4.596893357840488,
+      "grad_norm": 0.2409515678882599,
+      "learning_rate": 8.542224659832871e-07,
+      "loss": 0.3832,
       "step": 127550
     },
     {
-      "epoch": 4.49,
-      "learning_rate": 1.371895227062925e-06,
-      "loss": 0.2333,
+      "epoch": 4.597073557501712,
+      "grad_norm": 0.3206073045730591,
+      "learning_rate": 8.534663310513141e-07,
+      "loss": 0.3694,
       "step": 127555
     },
     {
-      "epoch": 4.49,
-      "learning_rate": 1.3709646710619113e-06,
-      "loss": 0.2297,
+      "epoch": 4.597253757162936,
+      "grad_norm": 0.2606680393218994,
+      "learning_rate": 8.527105251111833e-07,
+      "loss": 0.3742,
       "step": 127560
     },
     {
-      "epoch": 4.49,
-      "learning_rate": 1.3700344218683437e-06,
-      "loss": 0.2366,
+      "epoch": 4.597433956824161,
+      "grad_norm": 0.3032326102256775,
+      "learning_rate": 8.519550481731975e-07,
+      "loss": 0.3228,
       "step": 127565
     },
     {
-      "epoch": 4.49,
-      "learning_rate": 1.3691044794943125e-06,
-      "loss": 0.2414,
+      "epoch": 4.597614156485386,
+      "grad_norm": 0.303232342004776,
+      "learning_rate": 8.511999002476512e-07,
+      "loss": 0.396,
       "step": 127570
     },
     {
-      "epoch": 4.49,
-      "learning_rate": 1.368174843951886e-06,
-      "loss": 0.2431,
+      "epoch": 4.59779435614661,
+      "grad_norm": 0.2779444754123688,
+      "learning_rate": 8.504450813448306e-07,
+      "loss": 0.3345,
       "step": 127575
     },
     {
-      "epoch": 4.49,
-      "learning_rate": 1.367245515253135e-06,
-      "loss": 0.2397,
+      "epoch": 4.597974555807835,
+      "grad_norm": 0.2656551003456116,
+      "learning_rate": 8.496905914750192e-07,
+      "loss": 0.3711,
       "step": 127580
     },
     {
-      "epoch": 4.49,
-      "learning_rate": 1.3663164934101219e-06,
-      "loss": 0.2398,
+      "epoch": 4.59815475546906,
+      "grad_norm": 0.2177533507347107,
+      "learning_rate": 8.489364306484976e-07,
+      "loss": 0.3436,
       "step": 127585
     },
     {
-      "epoch": 4.49,
-      "learning_rate": 1.3653877784349178e-06,
-      "loss": 0.2439,
+      "epoch": 4.598334955130285,
+      "grad_norm": 0.25983306765556335,
+      "learning_rate": 8.481825988755382e-07,
+      "loss": 0.3416,
       "step": 127590
     },
     {
-      "epoch": 4.49,
-      "learning_rate": 1.364459370339577e-06,
-      "loss": 0.2407,
+      "epoch": 4.598515154791509,
+      "grad_norm": 0.23284339904785156,
+      "learning_rate": 8.474290961664244e-07,
+      "loss": 0.333,
       "step": 127595
     },
     {
-      "epoch": 4.49,
-      "learning_rate": 1.3635312691361508e-06,
-      "loss": 0.2473,
+      "epoch": 4.598695354452733,
+      "grad_norm": 0.2582654356956482,
+      "learning_rate": 8.466759225314063e-07,
+      "loss": 0.33,
       "step": 127600
     },
     {
-      "epoch": 4.49,
-      "learning_rate": 1.3626034748366962e-06,
-      "loss": 0.2755,
+      "epoch": 4.598875554113958,
+      "grad_norm": 0.31043708324432373,
+      "learning_rate": 8.459230779807508e-07,
+      "loss": 0.3663,
       "step": 127605
     },
     {
-      "epoch": 4.49,
-      "learning_rate": 1.3616759874532592e-06,
-      "loss": 0.258,
+      "epoch": 4.599055753775183,
+      "grad_norm": 0.25590458512306213,
+      "learning_rate": 8.451705625247191e-07,
+      "loss": 0.3797,
       "step": 127610
     },
     {
-      "epoch": 4.49,
-      "learning_rate": 1.360748806997883e-06,
-      "loss": 0.2314,
+      "epoch": 4.599235953436407,
+      "grad_norm": 0.2824951410293579,
+      "learning_rate": 8.444183761735641e-07,
+      "loss": 0.385,
       "step": 127615
     },
     {
-      "epoch": 4.49,
-      "learning_rate": 1.3598219334826022e-06,
-      "loss": 0.238,
+      "epoch": 4.599416153097632,
+      "grad_norm": 0.3160955309867859,
+      "learning_rate": 8.436665189375303e-07,
+      "loss": 0.351,
       "step": 127620
     },
     {
-      "epoch": 4.49,
-      "learning_rate": 1.3588953669194548e-06,
-      "loss": 0.2529,
+      "epoch": 4.599596352758857,
+      "grad_norm": 0.29997727274894714,
+      "learning_rate": 8.429149908268652e-07,
+      "loss": 0.3823,
       "step": 127625
     },
     {
-      "epoch": 4.49,
-      "learning_rate": 1.3579691073204726e-06,
-      "loss": 0.2615,
+      "epoch": 4.5997765524200815,
+      "grad_norm": 0.2773001194000244,
+      "learning_rate": 8.421637918518022e-07,
+      "loss": 0.3737,
       "step": 127630
     },
     {
-      "epoch": 4.49,
-      "learning_rate": 1.3570431546976792e-06,
-      "loss": 0.244,
+      "epoch": 4.599956752081306,
+      "grad_norm": 0.2861831486225128,
+      "learning_rate": 8.41412922022583e-07,
+      "loss": 0.3375,
       "step": 127635
     },
     {
-      "epoch": 4.49,
-      "learning_rate": 1.3561175090630984e-06,
-      "loss": 0.2943,
+      "epoch": 4.600136951742531,
+      "grad_norm": 0.27983224391937256,
+      "learning_rate": 8.406623813494358e-07,
+      "loss": 0.365,
       "step": 127640
     },
     {
-      "epoch": 4.49,
-      "learning_rate": 1.3551921704287513e-06,
-      "loss": 0.237,
+      "epoch": 4.600317151403756,
+      "grad_norm": 0.26674070954322815,
+      "learning_rate": 8.399121698425855e-07,
+      "loss": 0.3933,
       "step": 127645
     },
     {
-      "epoch": 4.49,
-      "learning_rate": 1.354267138806653e-06,
-      "loss": 0.2492,
+      "epoch": 4.6004973510649805,
+      "grad_norm": 0.23168079555034637,
+      "learning_rate": 8.391622875122545e-07,
+      "loss": 0.3665,
       "step": 127650
     },
     {
-      "epoch": 4.49,
-      "learning_rate": 1.353342414208808e-06,
-      "loss": 0.2577,
+      "epoch": 4.600677550726204,
+      "grad_norm": 0.2457718551158905,
+      "learning_rate": 8.384127343686599e-07,
+      "loss": 0.3455,
       "step": 127655
     },
     {
-      "epoch": 4.49,
-      "learning_rate": 1.3524179966472317e-06,
-      "loss": 0.2467,
+      "epoch": 4.600857750387429,
+      "grad_norm": 0.25445955991744995,
+      "learning_rate": 8.376635104220126e-07,
+      "loss": 0.3738,
       "step": 127660
     },
     {
-      "epoch": 4.49,
-      "learning_rate": 1.3514938861339199e-06,
-      "loss": 0.2468,
+      "epoch": 4.601037950048654,
+      "grad_norm": 0.23274029791355133,
+      "learning_rate": 8.369146156825269e-07,
+      "loss": 0.3727,
       "step": 127665
     },
     {
-      "epoch": 4.49,
-      "learning_rate": 1.350570082680877e-06,
-      "loss": 0.236,
+      "epoch": 4.6012181497098785,
+      "grad_norm": 0.2078734189271927,
+      "learning_rate": 8.361660501603946e-07,
+      "loss": 0.3685,
       "step": 127670
     },
     {
-      "epoch": 4.49,
-      "learning_rate": 1.349646586300099e-06,
-      "loss": 0.251,
+      "epoch": 4.601398349371103,
+      "grad_norm": 0.24284294247627258,
+      "learning_rate": 8.354178138658269e-07,
+      "loss": 0.3768,
       "step": 127675
     },
     {
-      "epoch": 4.49,
-      "learning_rate": 1.3487233970035679e-06,
-      "loss": 0.2442,
+      "epoch": 4.601578549032328,
+      "grad_norm": 0.3138650059700012,
+      "learning_rate": 8.346699068090074e-07,
+      "loss": 0.3789,
       "step": 127680
     },
     {
-      "epoch": 4.49,
-      "learning_rate": 1.3478005148032824e-06,
-      "loss": 0.2409,
+      "epoch": 4.601758748693553,
+      "grad_norm": 0.23819488286972046,
+      "learning_rate": 8.339223290001363e-07,
+      "loss": 0.3685,
       "step": 127685
     },
     {
-      "epoch": 4.49,
-      "learning_rate": 1.3468779397112191e-06,
-      "loss": 0.2736,
+      "epoch": 4.6019389483547775,
+      "grad_norm": 0.2773071825504303,
+      "learning_rate": 8.331750804493971e-07,
+      "loss": 0.373,
       "step": 127690
     },
     {
-      "epoch": 4.49,
-      "learning_rate": 1.3459556717393574e-06,
-      "loss": 0.2542,
+      "epoch": 4.602119148016001,
+      "grad_norm": 0.2507835924625397,
+      "learning_rate": 8.324281611669621e-07,
+      "loss": 0.3704,
       "step": 127695
     },
     {
-      "epoch": 4.49,
-      "learning_rate": 1.345033710899668e-06,
-      "loss": 0.2343,
+      "epoch": 4.602299347677226,
+      "grad_norm": 0.1873769313097,
+      "learning_rate": 8.316815711630205e-07,
+      "loss": 0.3327,
       "step": 127700
     },
     {
-      "epoch": 4.49,
-      "learning_rate": 1.3441120572041305e-06,
-      "loss": 0.244,
+      "epoch": 4.602479547338451,
+      "grad_norm": 0.23078793287277222,
+      "learning_rate": 8.309353104477335e-07,
+      "loss": 0.3727,
       "step": 127705
     },
     {
-      "epoch": 4.49,
-      "learning_rate": 1.3431907106647073e-06,
-      "loss": 0.2564,
+      "epoch": 4.6026597469996755,
+      "grad_norm": 0.23477856814861298,
+      "learning_rate": 8.301893790312765e-07,
+      "loss": 0.3795,
       "step": 127710
     },
     {
-      "epoch": 4.49,
-      "learning_rate": 1.3422696712933642e-06,
-      "loss": 0.242,
+      "epoch": 4.6028399466609,
+      "grad_norm": 0.27823832631111145,
+      "learning_rate": 8.294437769238106e-07,
+      "loss": 0.3449,
       "step": 127715
     },
     {
-      "epoch": 4.49,
-      "learning_rate": 1.3413489391020551e-06,
-      "loss": 0.2257,
+      "epoch": 4.603020146322125,
+      "grad_norm": 0.27423784136772156,
+      "learning_rate": 8.286985041354889e-07,
+      "loss": 0.3604,
       "step": 127720
     },
     {
-      "epoch": 4.49,
-      "learning_rate": 1.34042851410274e-06,
-      "loss": 0.269,
+      "epoch": 4.60320034598335,
+      "grad_norm": 0.2629181146621704,
+      "learning_rate": 8.279535606764755e-07,
+      "loss": 0.3842,
       "step": 127725
     },
     {
-      "epoch": 4.49,
-      "learning_rate": 1.3395083963073707e-06,
-      "loss": 0.2503,
+      "epoch": 4.603380545644574,
+      "grad_norm": 0.25341519713401794,
+      "learning_rate": 8.272089465569121e-07,
+      "loss": 0.3454,
       "step": 127730
     },
     {
-      "epoch": 4.49,
-      "learning_rate": 1.3385885857278874e-06,
-      "loss": 0.2621,
+      "epoch": 4.603560745305799,
+      "grad_norm": 0.2602289617061615,
+      "learning_rate": 8.264646617869493e-07,
+      "loss": 0.3771,
       "step": 127735
     },
     {
-      "epoch": 4.49,
-      "learning_rate": 1.3376690823762417e-06,
-      "loss": 0.2473,
+      "epoch": 4.603740944967024,
+      "grad_norm": 0.2383667379617691,
+      "learning_rate": 8.25720706376723e-07,
+      "loss": 0.3977,
       "step": 127740
     },
     {
-      "epoch": 4.49,
-      "learning_rate": 1.3367498862643685e-06,
-      "loss": 0.2326,
+      "epoch": 4.603921144628248,
+      "grad_norm": 0.2735196650028229,
+      "learning_rate": 8.249770803363727e-07,
+      "loss": 0.3894,
       "step": 127745
     },
     {
-      "epoch": 4.49,
-      "learning_rate": 1.3358309974042027e-06,
-      "loss": 0.2687,
+      "epoch": 4.6041013442894725,
+      "grad_norm": 0.2749096751213074,
+      "learning_rate": 8.242337836760261e-07,
+      "loss": 0.3741,
       "step": 127750
     },
     {
-      "epoch": 4.49,
-      "learning_rate": 1.3349124158076736e-06,
-      "loss": 0.2496,
+      "epoch": 4.604281543950697,
+      "grad_norm": 0.22528064250946045,
+      "learning_rate": 8.234908164058169e-07,
+      "loss": 0.3538,
       "step": 127755
     },
     {
-      "epoch": 4.49,
-      "learning_rate": 1.3339941414867162e-06,
-      "loss": 0.2329,
+      "epoch": 4.604461743611922,
+      "grad_norm": 0.22581934928894043,
+      "learning_rate": 8.227481785358648e-07,
+      "loss": 0.3696,
       "step": 127760
     },
     {
-      "epoch": 4.5,
-      "learning_rate": 1.333076174453246e-06,
-      "loss": 0.2658,
+      "epoch": 4.604641943273147,
+      "grad_norm": 0.31339600682258606,
+      "learning_rate": 8.220058700762894e-07,
+      "loss": 0.4024,
       "step": 127765
     },
     {
-      "epoch": 4.5,
-      "learning_rate": 1.3321585147191812e-06,
-      "loss": 0.2538,
+      "epoch": 4.604822142934371,
+      "grad_norm": 0.2499316930770874,
+      "learning_rate": 8.212638910372023e-07,
+      "loss": 0.386,
       "step": 127770
     },
     {
-      "epoch": 4.5,
-      "learning_rate": 1.33124116229644e-06,
-      "loss": 0.2431,
+      "epoch": 4.605002342595596,
+      "grad_norm": 0.27650460600852966,
+      "learning_rate": 8.205222414287089e-07,
+      "loss": 0.378,
       "step": 127775
     },
     {
-      "epoch": 4.5,
-      "learning_rate": 1.3303241171969406e-06,
-      "loss": 0.2353,
+      "epoch": 4.605182542256821,
+      "grad_norm": 0.2796940505504608,
+      "learning_rate": 8.197809212609236e-07,
+      "loss": 0.3827,
       "step": 127780
     },
     {
-      "epoch": 4.5,
-      "learning_rate": 1.3294073794325818e-06,
-      "loss": 0.2328,
+      "epoch": 4.605362741918045,
+      "grad_norm": 0.24295328557491302,
+      "learning_rate": 8.190399305439412e-07,
+      "loss": 0.3327,
       "step": 127785
     },
     {
-      "epoch": 4.5,
-      "learning_rate": 1.3284909490152681e-06,
-      "loss": 0.2453,
+      "epoch": 4.6055429415792695,
+      "grad_norm": 0.2779107689857483,
+      "learning_rate": 8.182992692878561e-07,
+      "loss": 0.3532,
       "step": 127790
     },
     {
-      "epoch": 4.5,
-      "learning_rate": 1.3275748259568955e-06,
-      "loss": 0.2366,
+      "epoch": 4.605723141240494,
+      "grad_norm": 0.22981327772140503,
+      "learning_rate": 8.175589375027631e-07,
+      "loss": 0.3642,
       "step": 127795
     },
     {
-      "epoch": 4.5,
-      "learning_rate": 1.3266590102693655e-06,
-      "loss": 0.254,
+      "epoch": 4.605903340901719,
+      "grad_norm": 0.2713334262371063,
+      "learning_rate": 8.168189351987488e-07,
+      "loss": 0.3618,
       "step": 127800
     },
     {
-      "epoch": 4.5,
-      "learning_rate": 1.3257435019645687e-06,
-      "loss": 0.2509,
+      "epoch": 4.606083540562944,
+      "grad_norm": 0.2509160339832306,
+      "learning_rate": 8.160792623858909e-07,
+      "loss": 0.3942,
       "step": 127805
     },
     {
-      "epoch": 4.5,
-      "learning_rate": 1.3248283010543927e-06,
-      "loss": 0.2444,
+      "epoch": 4.606263740224168,
+      "grad_norm": 0.3234541416168213,
+      "learning_rate": 8.153399190742761e-07,
+      "loss": 0.3853,
       "step": 127810
     },
     {
-      "epoch": 4.5,
-      "learning_rate": 1.3240963616583557e-06,
-      "loss": 0.2366,
+      "epoch": 4.606443939885393,
+      "grad_norm": 0.35334426164627075,
+      "learning_rate": 8.146009052739656e-07,
+      "loss": 0.4016,
       "step": 127815
     },
     {
-      "epoch": 4.5,
-      "learning_rate": 1.3231817140884339e-06,
-      "loss": 0.2728,
+      "epoch": 4.606624139546618,
+      "grad_norm": 0.3184077739715576,
+      "learning_rate": 8.138622209950403e-07,
+      "loss": 0.3873,
       "step": 127820
     },
     {
-      "epoch": 4.5,
-      "learning_rate": 1.3222673739463926e-06,
-      "loss": 0.2542,
+      "epoch": 4.606804339207843,
+      "grad_norm": 0.20487551391124725,
+      "learning_rate": 8.131238662475588e-07,
+      "loss": 0.329,
       "step": 127825
     },
     {
-      "epoch": 4.5,
-      "learning_rate": 1.3213533412441054e-06,
-      "loss": 0.2533,
+      "epoch": 4.606984538869067,
+      "grad_norm": 0.24536176025867462,
+      "learning_rate": 8.123858410415825e-07,
+      "loss": 0.3856,
       "step": 127830
     },
     {
-      "epoch": 4.5,
-      "learning_rate": 1.3204396159934406e-06,
-      "loss": 0.2539,
+      "epoch": 4.607164738530291,
+      "grad_norm": 0.27180904150009155,
+      "learning_rate": 8.116481453871672e-07,
+      "loss": 0.3782,
       "step": 127835
     },
     {
-      "epoch": 4.5,
-      "learning_rate": 1.319526198206264e-06,
-      "loss": 0.251,
+      "epoch": 4.607344938191516,
+      "grad_norm": 0.27243977785110474,
+      "learning_rate": 8.109107792943577e-07,
+      "loss": 0.3721,
       "step": 127840
     },
     {
-      "epoch": 4.5,
-      "learning_rate": 1.3186130878944296e-06,
-      "loss": 0.2461,
+      "epoch": 4.607525137852741,
+      "grad_norm": 0.2593880295753479,
+      "learning_rate": 8.101737427732098e-07,
+      "loss": 0.3698,
       "step": 127845
     },
     {
-      "epoch": 4.5,
-      "learning_rate": 1.3177002850698006e-06,
-      "loss": 0.2506,
+      "epoch": 4.607705337513965,
+      "grad_norm": 0.2483491450548172,
+      "learning_rate": 8.0943703583376e-07,
+      "loss": 0.3611,
       "step": 127850
     },
     {
-      "epoch": 4.5,
-      "learning_rate": 1.3167877897442227e-06,
-      "loss": 0.2423,
+      "epoch": 4.60788553717519,
+      "grad_norm": 0.28499743342399597,
+      "learning_rate": 8.087006584860501e-07,
+      "loss": 0.3617,
       "step": 127855
     },
     {
-      "epoch": 4.5,
-      "learning_rate": 1.315875601929553e-06,
-      "loss": 0.2454,
+      "epoch": 4.608065736836415,
+      "grad_norm": 0.29936861991882324,
+      "learning_rate": 8.079646107401085e-07,
+      "loss": 0.3898,
       "step": 127860
     },
     {
-      "epoch": 4.5,
-      "learning_rate": 1.3149637216376243e-06,
-      "loss": 0.256,
+      "epoch": 4.60824593649764,
+      "grad_norm": 0.25267350673675537,
+      "learning_rate": 8.072288926059601e-07,
+      "loss": 0.3722,
       "step": 127865
     },
     {
-      "epoch": 4.5,
-      "learning_rate": 1.3140521488802876e-06,
-      "loss": 0.2749,
+      "epoch": 4.608426136158864,
+      "grad_norm": 0.23155082762241364,
+      "learning_rate": 8.064935040936417e-07,
+      "loss": 0.3845,
       "step": 127870
     },
     {
-      "epoch": 4.5,
-      "learning_rate": 1.3131408836693726e-06,
-      "loss": 0.2542,
+      "epoch": 4.608606335820088,
+      "grad_norm": 0.21807222068309784,
+      "learning_rate": 8.057584452131644e-07,
+      "loss": 0.3625,
       "step": 127875
     },
     {
-      "epoch": 4.5,
-      "learning_rate": 1.3122299260167143e-06,
-      "loss": 0.2321,
+      "epoch": 4.608786535481313,
+      "grad_norm": 0.21520547568798065,
+      "learning_rate": 8.0502371597454e-07,
+      "loss": 0.388,
       "step": 127880
     },
     {
-      "epoch": 4.5,
-      "learning_rate": 1.3113192759341363e-06,
-      "loss": 0.2417,
+      "epoch": 4.608966735142538,
+      "grad_norm": 0.3115110993385315,
+      "learning_rate": 8.042893163877852e-07,
+      "loss": 0.3558,
       "step": 127885
     },
     {
-      "epoch": 4.5,
-      "learning_rate": 1.3104089334334686e-06,
-      "loss": 0.2423,
+      "epoch": 4.609146934803762,
+      "grad_norm": 0.2501071095466614,
+      "learning_rate": 8.035552464629059e-07,
+      "loss": 0.3805,
       "step": 127890
     },
     {
-      "epoch": 4.5,
-      "learning_rate": 1.3094988985265287e-06,
-      "loss": 0.2395,
+      "epoch": 4.609327134464987,
+      "grad_norm": 0.27980107069015503,
+      "learning_rate": 8.028215062098998e-07,
+      "loss": 0.35,
       "step": 127895
     },
     {
-      "epoch": 4.5,
-      "learning_rate": 1.308589171225133e-06,
-      "loss": 0.2731,
+      "epoch": 4.609507334126212,
+      "grad_norm": 0.35373973846435547,
+      "learning_rate": 8.020880956387672e-07,
+      "loss": 0.3654,
       "step": 127900
     },
     {
-      "epoch": 4.5,
-      "learning_rate": 1.307679751541091e-06,
-      "loss": 0.2482,
+      "epoch": 4.609687533787437,
+      "grad_norm": 0.3000072240829468,
+      "learning_rate": 8.013550147594945e-07,
+      "loss": 0.3785,
       "step": 127905
     },
     {
-      "epoch": 4.5,
-      "learning_rate": 1.3067706394862156e-06,
-      "loss": 0.2195,
+      "epoch": 4.609867733448661,
+      "grad_norm": 0.258821576833725,
+      "learning_rate": 8.006222635820792e-07,
+      "loss": 0.3589,
       "step": 127910
     },
     {
-      "epoch": 4.5,
-      "learning_rate": 1.3058618350723085e-06,
-      "loss": 0.2456,
+      "epoch": 4.610047933109886,
+      "grad_norm": 0.2523718774318695,
+      "learning_rate": 7.998898421165025e-07,
+      "loss": 0.3906,
       "step": 127915
     },
     {
-      "epoch": 4.5,
-      "learning_rate": 1.3049533383111685e-06,
-      "loss": 0.251,
+      "epoch": 4.610228132771111,
+      "grad_norm": 0.2900547385215759,
+      "learning_rate": 7.991577503727366e-07,
+      "loss": 0.3752,
       "step": 127920
     },
     {
-      "epoch": 4.5,
-      "learning_rate": 1.3040451492145972e-06,
-      "loss": 0.2635,
+      "epoch": 4.6104083324323355,
+      "grad_norm": 0.23753364384174347,
+      "learning_rate": 7.984259883607653e-07,
+      "loss": 0.343,
       "step": 127925
     },
     {
-      "epoch": 4.5,
-      "learning_rate": 1.3031372677943827e-06,
-      "loss": 0.2577,
+      "epoch": 4.610588532093559,
+      "grad_norm": 0.24349209666252136,
+      "learning_rate": 7.97694556090553e-07,
+      "loss": 0.3692,
       "step": 127930
     },
     {
-      "epoch": 4.5,
-      "learning_rate": 1.3022296940623153e-06,
-      "loss": 0.2534,
+      "epoch": 4.610768731754784,
+      "grad_norm": 0.2928297519683838,
+      "learning_rate": 7.969634535720639e-07,
+      "loss": 0.3934,
       "step": 127935
     },
     {
-      "epoch": 4.5,
-      "learning_rate": 1.3013224280301745e-06,
-      "loss": 0.2489,
+      "epoch": 4.610948931416009,
+      "grad_norm": 0.29633548855781555,
+      "learning_rate": 7.962326808152621e-07,
+      "loss": 0.3514,
       "step": 127940
     },
     {
-      "epoch": 4.5,
-      "learning_rate": 1.3004154697097454e-06,
-      "loss": 0.2611,
+      "epoch": 4.6111291310772335,
+      "grad_norm": 0.26136893033981323,
+      "learning_rate": 7.955022378301064e-07,
+      "loss": 0.3772,
       "step": 127945
     },
     {
-      "epoch": 4.5,
-      "learning_rate": 1.2995088191128046e-06,
-      "loss": 0.2496,
+      "epoch": 4.611309330738458,
+      "grad_norm": 0.24420562386512756,
+      "learning_rate": 7.947721246265472e-07,
+      "loss": 0.352,
       "step": 127950
     },
     {
-      "epoch": 4.5,
-      "learning_rate": 1.2986024762511235e-06,
-      "loss": 0.2583,
+      "epoch": 4.611489530399683,
+      "grad_norm": 0.24144157767295837,
+      "learning_rate": 7.940423412145292e-07,
+      "loss": 0.4035,
       "step": 127955
     },
     {
-      "epoch": 4.5,
-      "learning_rate": 1.2976964411364644e-06,
-      "loss": 0.2547,
+      "epoch": 4.611669730060908,
+      "grad_norm": 0.19324125349521637,
+      "learning_rate": 7.933128876039974e-07,
+      "loss": 0.3594,
       "step": 127960
     },
     {
-      "epoch": 4.5,
-      "learning_rate": 1.296790713780599e-06,
-      "loss": 0.2385,
+      "epoch": 4.6118499297221325,
+      "grad_norm": 0.2407853752374649,
+      "learning_rate": 7.925837638048967e-07,
+      "loss": 0.394,
       "step": 127965
     },
     {
-      "epoch": 4.5,
-      "learning_rate": 1.2958852941952843e-06,
-      "loss": 0.2369,
+      "epoch": 4.612030129383356,
+      "grad_norm": 0.2682248055934906,
+      "learning_rate": 7.918549698271494e-07,
+      "loss": 0.36,
       "step": 127970
     },
     {
-      "epoch": 4.5,
-      "learning_rate": 1.2949801823922802e-06,
-      "loss": 0.2403,
+      "epoch": 4.612210329044581,
+      "grad_norm": 0.21902526915073395,
+      "learning_rate": 7.911265056806921e-07,
+      "loss": 0.3495,
       "step": 127975
     },
     {
-      "epoch": 4.5,
-      "learning_rate": 1.2940753783833303e-06,
-      "loss": 0.2501,
+      "epoch": 4.612390528705806,
+      "grad_norm": 0.25772321224212646,
+      "learning_rate": 7.903983713754503e-07,
+      "loss": 0.3791,
       "step": 127980
     },
     {
-      "epoch": 4.5,
-      "learning_rate": 1.2931708821801946e-06,
-      "loss": 0.2313,
+      "epoch": 4.6125707283670305,
+      "grad_norm": 0.23041696846485138,
+      "learning_rate": 7.896705669213411e-07,
+      "loss": 0.4039,
       "step": 127985
     },
     {
-      "epoch": 4.5,
-      "learning_rate": 1.292266693794611e-06,
-      "loss": 0.2635,
+      "epoch": 4.612750928028255,
+      "grad_norm": 0.2209273725748062,
+      "learning_rate": 7.889430923282898e-07,
+      "loss": 0.3744,
       "step": 127990
     },
     {
-      "epoch": 4.5,
-      "learning_rate": 1.2913628132383226e-06,
-      "loss": 0.2462,
+      "epoch": 4.61293112768948,
+      "grad_norm": 0.23536401987075806,
+      "learning_rate": 7.882159476061968e-07,
+      "loss": 0.3478,
       "step": 127995
     },
     {
-      "epoch": 4.5,
-      "learning_rate": 1.2904592405230565e-06,
-      "loss": 0.2526,
+      "epoch": 4.613111327350705,
+      "grad_norm": 0.29315903782844543,
+      "learning_rate": 7.874891327649709e-07,
+      "loss": 0.4023,
       "step": 128000
     },
     {
-      "epoch": 4.5,
-      "eval_loss": 0.24867750704288483,
-      "eval_runtime": 10.5298,
-      "eval_samples_per_second": 9.497,
-      "eval_steps_per_second": 9.497,
+      "epoch": 4.613111327350705,
+      "eval_loss": 0.42875730991363525,
+      "eval_runtime": 3.5339,
+      "eval_samples_per_second": 28.298,
+      "eval_steps_per_second": 7.074,
       "step": 128000
     },
     {
-      "epoch": 4.5,
-      "learning_rate": 1.2895559756605586e-06,
-      "loss": 0.255,
+      "epoch": 4.6132915270119295,
+      "grad_norm": 0.25371137261390686,
+      "learning_rate": 7.867626478145235e-07,
+      "loss": 0.3724,
       "step": 128005
     },
     {
-      "epoch": 4.5,
-      "learning_rate": 1.2886530186625473e-06,
-      "loss": 0.2401,
+      "epoch": 4.613471726673154,
+      "grad_norm": 0.2467198222875595,
+      "learning_rate": 7.86036492764744e-07,
+      "loss": 0.386,
       "step": 128010
     },
     {
-      "epoch": 4.5,
-      "learning_rate": 1.2877503695407495e-06,
-      "loss": 0.2524,
+      "epoch": 4.613651926334379,
+      "grad_norm": 0.36411789059638977,
+      "learning_rate": 7.853106676255328e-07,
+      "loss": 0.3526,
       "step": 128015
     },
     {
-      "epoch": 4.5,
-      "learning_rate": 1.286848028306889e-06,
-      "loss": 0.2278,
+      "epoch": 4.613832125995603,
+      "grad_norm": 0.31409749388694763,
+      "learning_rate": 7.84585172406771e-07,
+      "loss": 0.4179,
       "step": 128020
     },
     {
-      "epoch": 4.5,
-      "learning_rate": 1.285945994972676e-06,
-      "loss": 0.2532,
+      "epoch": 4.6140123256568275,
+      "grad_norm": 0.23237422108650208,
+      "learning_rate": 7.838600071183478e-07,
+      "loss": 0.3654,
       "step": 128025
     },
     {
-      "epoch": 4.5,
-      "learning_rate": 1.2850442695498287e-06,
-      "loss": 0.2428,
+      "epoch": 4.614192525318052,
+      "grad_norm": 0.2930370271205902,
+      "learning_rate": 7.831351717701469e-07,
+      "loss": 0.3874,
       "step": 128030
     },
     {
-      "epoch": 4.5,
-      "learning_rate": 1.2841428520500492e-06,
-      "loss": 0.2495,
+      "epoch": 4.614372724979277,
+      "grad_norm": 0.24373428523540497,
+      "learning_rate": 7.82410666372041e-07,
+      "loss": 0.3839,
       "step": 128035
     },
     {
-      "epoch": 4.5,
-      "learning_rate": 1.28324174248505e-06,
-      "loss": 0.2322,
+      "epoch": 4.614552924640502,
+      "grad_norm": 0.31576600670814514,
+      "learning_rate": 7.816864909339e-07,
+      "loss": 0.384,
       "step": 128040
     },
     {
-      "epoch": 4.5,
-      "learning_rate": 1.282340940866525e-06,
-      "loss": 0.2704,
+      "epoch": 4.614733124301726,
+      "grad_norm": 0.2532004714012146,
+      "learning_rate": 7.809626454655911e-07,
+      "loss": 0.3523,
       "step": 128045
     },
     {
-      "epoch": 4.51,
-      "learning_rate": 1.2814404472061752e-06,
-      "loss": 0.2555,
+      "epoch": 4.614913323962951,
+      "grad_norm": 0.2986234724521637,
+      "learning_rate": 7.802391299769784e-07,
+      "loss": 0.3987,
       "step": 128050
     },
     {
-      "epoch": 4.51,
-      "learning_rate": 1.2805402615156837e-06,
-      "loss": 0.2477,
+      "epoch": 4.615093523624176,
+      "grad_norm": 0.24925264716148376,
+      "learning_rate": 7.795159444779154e-07,
+      "loss": 0.3647,
       "step": 128055
     },
     {
-      "epoch": 4.51,
-      "learning_rate": 1.2796403838067517e-06,
-      "loss": 0.2378,
+      "epoch": 4.6152737232854,
+      "grad_norm": 0.19276399910449982,
+      "learning_rate": 7.787930889782663e-07,
+      "loss": 0.3356,
       "step": 128060
     },
     {
-      "epoch": 4.51,
-      "learning_rate": 1.2787408140910535e-06,
-      "loss": 0.2618,
+      "epoch": 4.6154539229466245,
+      "grad_norm": 0.23550952970981598,
+      "learning_rate": 7.780705634878621e-07,
+      "loss": 0.4043,
       "step": 128065
     },
     {
-      "epoch": 4.51,
-      "learning_rate": 1.2778415523802767e-06,
-      "loss": 0.2622,
+      "epoch": 4.615634122607849,
+      "grad_norm": 0.29155659675598145,
+      "learning_rate": 7.773483680165617e-07,
+      "loss": 0.3733,
       "step": 128070
     },
     {
-      "epoch": 4.51,
-      "learning_rate": 1.2769425986860873e-06,
-      "loss": 0.253,
+      "epoch": 4.615814322269074,
+      "grad_norm": 0.2625998854637146,
+      "learning_rate": 7.766265025742014e-07,
+      "loss": 0.3544,
       "step": 128075
     },
     {
-      "epoch": 4.51,
-      "learning_rate": 1.2760439530201645e-06,
-      "loss": 0.2437,
+      "epoch": 4.615994521930299,
+      "grad_norm": 0.26157450675964355,
+      "learning_rate": 7.759049671706125e-07,
+      "loss": 0.3456,
       "step": 128080
     },
     {
-      "epoch": 4.51,
-      "learning_rate": 1.27514561539418e-06,
-      "loss": 0.2546,
+      "epoch": 4.616174721591523,
+      "grad_norm": 0.2234465479850769,
+      "learning_rate": 7.751837618156316e-07,
+      "loss": 0.4037,
       "step": 128085
     },
     {
-      "epoch": 4.51,
-      "learning_rate": 1.2742475858197961e-06,
-      "loss": 0.2535,
+      "epoch": 4.616354921252748,
+      "grad_norm": 0.2910038232803345,
+      "learning_rate": 7.744628865190784e-07,
+      "loss": 0.384,
       "step": 128090
     },
     {
-      "epoch": 4.51,
-      "learning_rate": 1.2733498643086677e-06,
-      "loss": 0.2389,
+      "epoch": 4.616535120913973,
+      "grad_norm": 0.28692278265953064,
+      "learning_rate": 7.737423412907785e-07,
+      "loss": 0.3577,
       "step": 128095
     },
     {
-      "epoch": 4.51,
-      "learning_rate": 1.272452450872458e-06,
-      "loss": 0.2523,
+      "epoch": 4.616715320575198,
+      "grad_norm": 0.23942017555236816,
+      "learning_rate": 7.730221261405518e-07,
+      "loss": 0.3879,
       "step": 128100
     },
     {
-      "epoch": 4.51,
-      "learning_rate": 1.2715553455228151e-06,
-      "loss": 0.243,
+      "epoch": 4.616895520236422,
+      "grad_norm": 0.23349037766456604,
+      "learning_rate": 7.723022410782044e-07,
+      "loss": 0.3775,
       "step": 128105
     },
     {
-      "epoch": 4.51,
-      "learning_rate": 1.2706585482713889e-06,
-      "loss": 0.2398,
+      "epoch": 4.617075719897647,
+      "grad_norm": 0.2587268352508545,
+      "learning_rate": 7.715826861135505e-07,
+      "loss": 0.4072,
       "step": 128110
     },
     {
-      "epoch": 4.51,
-      "learning_rate": 1.2697620591298199e-06,
-      "loss": 0.2464,
+      "epoch": 4.617255919558871,
+      "grad_norm": 0.24212481081485748,
+      "learning_rate": 7.708634612563936e-07,
+      "loss": 0.3423,
       "step": 128115
     },
     {
-      "epoch": 4.51,
-      "learning_rate": 1.268865878109754e-06,
-      "loss": 0.243,
+      "epoch": 4.617436119220096,
+      "grad_norm": 0.2429906278848648,
+      "learning_rate": 7.701445665165258e-07,
+      "loss": 0.3702,
       "step": 128120
     },
     {
-      "epoch": 4.51,
-      "learning_rate": 1.2679700052228265e-06,
-      "loss": 0.26,
+      "epoch": 4.61761631888132,
+      "grad_norm": 0.2274543046951294,
+      "learning_rate": 7.694260019037502e-07,
+      "loss": 0.3496,
       "step": 128125
     },
     {
-      "epoch": 4.51,
-      "learning_rate": 1.2670744404806645e-06,
-      "loss": 0.2382,
+      "epoch": 4.617796518542545,
+      "grad_norm": 0.27825266122817993,
+      "learning_rate": 7.687077674278564e-07,
+      "loss": 0.4096,
       "step": 128130
     },
     {
-      "epoch": 4.51,
-      "learning_rate": 1.2661791838949028e-06,
-      "loss": 0.2534,
+      "epoch": 4.61797671820377,
+      "grad_norm": 0.2594509720802307,
+      "learning_rate": 7.679898630986254e-07,
+      "loss": 0.3866,
       "step": 128135
     },
     {
-      "epoch": 4.51,
-      "learning_rate": 1.2652842354771627e-06,
-      "loss": 0.2408,
+      "epoch": 4.618156917864995,
+      "grad_norm": 0.2519370913505554,
+      "learning_rate": 7.67272288925841e-07,
+      "loss": 0.3612,
       "step": 128140
     },
     {
-      "epoch": 4.51,
-      "learning_rate": 1.2643895952390682e-06,
-      "loss": 0.2671,
+      "epoch": 4.618337117526219,
+      "grad_norm": 0.26362189650535583,
+      "learning_rate": 7.665550449192788e-07,
+      "loss": 0.3569,
       "step": 128145
     },
     {
-      "epoch": 4.51,
-      "learning_rate": 1.263495263192227e-06,
-      "loss": 0.2214,
+      "epoch": 4.618517317187444,
+      "grad_norm": 0.20409050583839417,
+      "learning_rate": 7.658381310887142e-07,
+      "loss": 0.4079,
       "step": 128150
     },
     {
-      "epoch": 4.51,
-      "learning_rate": 1.26260123934826e-06,
-      "loss": 0.2392,
+      "epoch": 4.618697516848668,
+      "grad_norm": 0.24394989013671875,
+      "learning_rate": 7.651215474439172e-07,
+      "loss": 0.3567,
       "step": 128155
     },
     {
-      "epoch": 4.51,
-      "learning_rate": 1.2617075237187748e-06,
-      "loss": 0.2451,
+      "epoch": 4.618877716509893,
+      "grad_norm": 0.2227751463651657,
+      "learning_rate": 7.644052939946411e-07,
+      "loss": 0.4044,
       "step": 128160
     },
     {
-      "epoch": 4.51,
-      "learning_rate": 1.260814116315373e-06,
-      "loss": 0.2509,
+      "epoch": 4.619057916171117,
+      "grad_norm": 0.2579880952835083,
+      "learning_rate": 7.636893707506532e-07,
+      "loss": 0.3804,
       "step": 128165
     },
     {
-      "epoch": 4.51,
-      "learning_rate": 1.2599210171496484e-06,
-      "loss": 0.2544,
+      "epoch": 4.619238115832342,
+      "grad_norm": 0.2594741880893707,
+      "learning_rate": 7.629737777217011e-07,
+      "loss": 0.3805,
       "step": 128170
     },
     {
-      "epoch": 4.51,
-      "learning_rate": 1.2590282262332109e-06,
-      "loss": 0.2544,
+      "epoch": 4.619418315493567,
+      "grad_norm": 0.2516389787197113,
+      "learning_rate": 7.622585149175465e-07,
+      "loss": 0.3612,
       "step": 128175
     },
     {
-      "epoch": 4.51,
-      "learning_rate": 1.2581357435776458e-06,
-      "loss": 0.2139,
+      "epoch": 4.619598515154792,
+      "grad_norm": 0.24397249519824982,
+      "learning_rate": 7.615435823479233e-07,
+      "loss": 0.3967,
       "step": 128180
     },
     {
-      "epoch": 4.51,
-      "learning_rate": 1.2572435691945355e-06,
-      "loss": 0.2569,
+      "epoch": 4.619778714816016,
+      "grad_norm": 0.2816333472728729,
+      "learning_rate": 7.608289800225737e-07,
+      "loss": 0.4114,
       "step": 128185
     },
     {
-      "epoch": 4.51,
-      "learning_rate": 1.2563517030954735e-06,
-      "loss": 0.2502,
+      "epoch": 4.619958914477241,
+      "grad_norm": 0.19909071922302246,
+      "learning_rate": 7.601147079512399e-07,
+      "loss": 0.3857,
       "step": 128190
     },
     {
-      "epoch": 4.51,
-      "learning_rate": 1.2554601452920394e-06,
-      "loss": 0.2751,
+      "epoch": 4.620139114138466,
+      "grad_norm": 0.2591770589351654,
+      "learning_rate": 7.594007661436475e-07,
+      "loss": 0.3494,
       "step": 128195
     },
     {
-      "epoch": 4.51,
-      "learning_rate": 1.2545688957958046e-06,
-      "loss": 0.2483,
+      "epoch": 4.6203193137996905,
+      "grad_norm": 0.2472904771566391,
+      "learning_rate": 7.586871546095303e-07,
+      "loss": 0.3945,
       "step": 128200
     },
     {
-      "epoch": 4.51,
-      "learning_rate": 1.2536779546183485e-06,
-      "loss": 0.2366,
+      "epoch": 4.620499513460914,
+      "grad_norm": 0.32332003116607666,
+      "learning_rate": 7.579738733586056e-07,
+      "loss": 0.3713,
       "step": 128205
     },
     {
-      "epoch": 4.51,
-      "learning_rate": 1.2527873217712287e-06,
-      "loss": 0.2486,
+      "epoch": 4.620679713122139,
+      "grad_norm": 0.25053974986076355,
+      "learning_rate": 7.572609224005905e-07,
+      "loss": 0.3696,
       "step": 128210
     },
     {
-      "epoch": 4.51,
-      "learning_rate": 1.2518969972660194e-06,
-      "loss": 0.2633,
+      "epoch": 4.620859912783364,
+      "grad_norm": 0.2518098056316376,
+      "learning_rate": 7.565483017452024e-07,
+      "loss": 0.3672,
       "step": 128215
     },
     {
-      "epoch": 4.51,
-      "learning_rate": 1.2510069811142749e-06,
-      "loss": 0.2338,
+      "epoch": 4.621040112444589,
+      "grad_norm": 0.25849613547325134,
+      "learning_rate": 7.558360114021528e-07,
+      "loss": 0.3614,
       "step": 128220
     },
     {
-      "epoch": 4.51,
-      "learning_rate": 1.2501172733275557e-06,
-      "loss": 0.2527,
+      "epoch": 4.621220312105813,
+      "grad_norm": 0.19711540639400482,
+      "learning_rate": 7.551240513811425e-07,
+      "loss": 0.3363,
       "step": 128225
     },
     {
-      "epoch": 4.51,
-      "learning_rate": 1.2492278739174079e-06,
-      "loss": 0.2317,
+      "epoch": 4.621400511767038,
+      "grad_norm": 0.28743451833724976,
+      "learning_rate": 7.544124216918746e-07,
+      "loss": 0.3525,
       "step": 128230
     },
     {
-      "epoch": 4.51,
-      "learning_rate": 1.248338782895389e-06,
-      "loss": 0.2424,
+      "epoch": 4.621580711428263,
+      "grad_norm": 0.2442832887172699,
+      "learning_rate": 7.537011223440415e-07,
+      "loss": 0.3821,
       "step": 128235
     },
     {
-      "epoch": 4.51,
-      "learning_rate": 1.2474500002730316e-06,
-      "loss": 0.2638,
+      "epoch": 4.6217609110894875,
+      "grad_norm": 0.23622441291809082,
+      "learning_rate": 7.529901533473355e-07,
+      "loss": 0.3843,
       "step": 128240
     },
     {
-      "epoch": 4.51,
-      "learning_rate": 1.2465615260618874e-06,
-      "loss": 0.2505,
+      "epoch": 4.621941110750711,
+      "grad_norm": 0.274114191532135,
+      "learning_rate": 7.522795147114486e-07,
+      "loss": 0.364,
       "step": 128245
     },
     {
-      "epoch": 4.51,
-      "learning_rate": 1.245673360273486e-06,
-      "loss": 0.2524,
+      "epoch": 4.622121310411936,
+      "grad_norm": 0.28449776768684387,
+      "learning_rate": 7.515692064460539e-07,
+      "loss": 0.3848,
       "step": 128250
     },
     {
-      "epoch": 4.51,
-      "learning_rate": 1.2447855029193628e-06,
-      "loss": 0.2336,
+      "epoch": 4.622301510073161,
+      "grad_norm": 0.2581150531768799,
+      "learning_rate": 7.508592285608351e-07,
+      "loss": 0.3404,
       "step": 128255
     },
     {
-      "epoch": 4.51,
-      "learning_rate": 1.2438979540110445e-06,
-      "loss": 0.2408,
+      "epoch": 4.6224817097343855,
+      "grad_norm": 0.25341930985450745,
+      "learning_rate": 7.501495810654679e-07,
+      "loss": 0.3227,
       "step": 128260
     },
     {
-      "epoch": 4.51,
-      "learning_rate": 1.2430107135600527e-06,
-      "loss": 0.249,
+      "epoch": 4.62266190939561,
+      "grad_norm": 0.24932630360126495,
+      "learning_rate": 7.494402639696113e-07,
+      "loss": 0.3482,
       "step": 128265
     },
     {
-      "epoch": 4.51,
-      "learning_rate": 1.2421237815779112e-06,
-      "loss": 0.2256,
+      "epoch": 4.622842109056835,
+      "grad_norm": 0.2252834588289261,
+      "learning_rate": 7.487312772829436e-07,
+      "loss": 0.356,
       "step": 128270
     },
     {
-      "epoch": 4.51,
-      "learning_rate": 1.241237158076139e-06,
-      "loss": 0.2579,
+      "epoch": 4.62302230871806,
+      "grad_norm": 0.23649568855762482,
+      "learning_rate": 7.480226210151098e-07,
+      "loss": 0.3956,
       "step": 128275
     },
     {
-      "epoch": 4.51,
-      "learning_rate": 1.240350843066243e-06,
-      "loss": 0.2842,
+      "epoch": 4.6232025083792845,
+      "grad_norm": 0.20822374522686005,
+      "learning_rate": 7.473142951757772e-07,
+      "loss": 0.3501,
       "step": 128280
     },
     {
-      "epoch": 4.51,
-      "learning_rate": 1.2394648365597283e-06,
-      "loss": 0.2446,
+      "epoch": 4.623382708040509,
+      "grad_norm": 0.2828695476055145,
+      "learning_rate": 7.466062997745909e-07,
+      "loss": 0.356,
       "step": 128285
     },
     {
-      "epoch": 4.51,
-      "learning_rate": 1.2385791385681105e-06,
-      "loss": 0.2459,
+      "epoch": 4.623562907701734,
+      "grad_norm": 0.2521428167819977,
+      "learning_rate": 7.458986348211988e-07,
+      "loss": 0.4002,
       "step": 128290
     },
     {
-      "epoch": 4.51,
-      "learning_rate": 1.2376937491028778e-06,
-      "loss": 0.2724,
+      "epoch": 4.623743107362958,
+      "grad_norm": 0.25071465969085693,
+      "learning_rate": 7.45191300325243e-07,
+      "loss": 0.4166,
       "step": 128295
     },
     {
-      "epoch": 4.51,
-      "learning_rate": 1.2368086681755347e-06,
-      "loss": 0.2479,
+      "epoch": 4.6239233070241825,
+      "grad_norm": 0.2567957937717438,
+      "learning_rate": 7.444842962963577e-07,
+      "loss": 0.3731,
       "step": 128300
     },
     {
-      "epoch": 4.51,
-      "learning_rate": 1.2359238957975695e-06,
-      "loss": 0.2443,
+      "epoch": 4.624103506685407,
+      "grad_norm": 0.2071949541568756,
+      "learning_rate": 7.437776227441767e-07,
+      "loss": 0.3451,
       "step": 128305
     },
     {
-      "epoch": 4.51,
-      "learning_rate": 1.2350394319804726e-06,
-      "loss": 0.2539,
+      "epoch": 4.624283706346632,
+      "grad_norm": 0.2758256793022156,
+      "learning_rate": 7.430712796783312e-07,
+      "loss": 0.3545,
       "step": 128310
     },
     {
-      "epoch": 4.51,
-      "learning_rate": 1.2341552767357296e-06,
-      "loss": 0.2659,
+      "epoch": 4.624463906007857,
+      "grad_norm": 0.26326316595077515,
+      "learning_rate": 7.423652671084441e-07,
+      "loss": 0.3581,
       "step": 128315
     },
     {
-      "epoch": 4.51,
-      "learning_rate": 1.2332714300748144e-06,
-      "loss": 0.267,
+      "epoch": 4.6246441056690815,
+      "grad_norm": 0.28544881939888,
+      "learning_rate": 7.416595850441327e-07,
+      "loss": 0.3611,
       "step": 128320
     },
     {
-      "epoch": 4.51,
-      "learning_rate": 1.232387892009207e-06,
-      "loss": 0.283,
+      "epoch": 4.624824305330306,
+      "grad_norm": 0.2547943592071533,
+      "learning_rate": 7.409542334950143e-07,
+      "loss": 0.3622,
       "step": 128325
     },
     {
-      "epoch": 4.52,
-      "learning_rate": 1.2315046625503785e-06,
-      "loss": 0.2495,
+      "epoch": 4.625004504991531,
+      "grad_norm": 0.24495618045330048,
+      "learning_rate": 7.402492124706922e-07,
+      "loss": 0.3712,
       "step": 128330
     },
     {
-      "epoch": 4.52,
-      "learning_rate": 1.2306217417098004e-06,
-      "loss": 0.2439,
+      "epoch": 4.625184704652755,
+      "grad_norm": 0.33649563789367676,
+      "learning_rate": 7.395445219807839e-07,
+      "loss": 0.3987,
       "step": 128335
     },
     {
-      "epoch": 4.52,
-      "learning_rate": 1.2297391294989357e-06,
-      "loss": 0.2537,
+      "epoch": 4.6253649043139795,
+      "grad_norm": 0.2949081361293793,
+      "learning_rate": 7.388401620348844e-07,
+      "loss": 0.3823,
       "step": 128340
     },
     {
-      "epoch": 4.52,
-      "learning_rate": 1.2288568259292394e-06,
-      "loss": 0.2471,
+      "epoch": 4.625545103975204,
+      "grad_norm": 0.23140224814414978,
+      "learning_rate": 7.381361326425862e-07,
+      "loss": 0.3957,
       "step": 128345
     },
     {
-      "epoch": 4.52,
-      "learning_rate": 1.2279748310121742e-06,
-      "loss": 0.246,
+      "epoch": 4.625725303636429,
+      "grad_norm": 0.2955685257911682,
+      "learning_rate": 7.374324338134897e-07,
+      "loss": 0.3733,
       "step": 128350
     },
     {
-      "epoch": 4.52,
-      "learning_rate": 1.2270931447591838e-06,
-      "loss": 0.2501,
+      "epoch": 4.625905503297654,
+      "grad_norm": 0.2494993358850479,
+      "learning_rate": 7.367290655571762e-07,
+      "loss": 0.3525,
       "step": 128355
     },
     {
-      "epoch": 4.52,
-      "learning_rate": 1.2262117671817286e-06,
-      "loss": 0.2576,
+      "epoch": 4.626085702958878,
+      "grad_norm": 0.21933454275131226,
+      "learning_rate": 7.360260278832353e-07,
+      "loss": 0.3365,
       "step": 128360
     },
     {
-      "epoch": 4.52,
-      "learning_rate": 1.225330698291241e-06,
-      "loss": 0.2712,
+      "epoch": 4.626265902620103,
+      "grad_norm": 0.26655271649360657,
+      "learning_rate": 7.3532332080124e-07,
+      "loss": 0.3693,
       "step": 128365
     },
     {
-      "epoch": 4.52,
-      "learning_rate": 1.2244499380991704e-06,
-      "loss": 0.2883,
+      "epoch": 4.626446102281328,
+      "grad_norm": 0.25093698501586914,
+      "learning_rate": 7.346209443207657e-07,
+      "loss": 0.364,
       "step": 128370
     },
     {
-      "epoch": 4.52,
-      "learning_rate": 1.223569486616949e-06,
-      "loss": 0.2656,
+      "epoch": 4.626626301942553,
+      "grad_norm": 0.2572937309741974,
+      "learning_rate": 7.339188984513828e-07,
+      "loss": 0.369,
       "step": 128375
     },
     {
-      "epoch": 4.52,
-      "learning_rate": 1.2226893438560038e-06,
-      "loss": 0.2379,
+      "epoch": 4.626806501603777,
+      "grad_norm": 0.2407890409231186,
+      "learning_rate": 7.332171832026586e-07,
+      "loss": 0.3885,
       "step": 128380
     },
     {
-      "epoch": 4.52,
-      "learning_rate": 1.2218095098277733e-06,
-      "loss": 0.2576,
+      "epoch": 4.626986701265002,
+      "grad_norm": 0.2307957410812378,
+      "learning_rate": 7.325157985841518e-07,
+      "loss": 0.3407,
       "step": 128385
     },
     {
-      "epoch": 4.52,
-      "learning_rate": 1.2209299845436727e-06,
-      "loss": 0.2686,
+      "epoch": 4.627166900926226,
+      "grad_norm": 0.3046651780605316,
+      "learning_rate": 7.318147446054191e-07,
+      "loss": 0.398,
       "step": 128390
     },
     {
-      "epoch": 4.52,
-      "learning_rate": 1.2200507680151268e-06,
-      "loss": 0.2639,
+      "epoch": 4.627347100587451,
+      "grad_norm": 0.2512384355068207,
+      "learning_rate": 7.311140212760109e-07,
+      "loss": 0.3883,
       "step": 128395
     },
     {
-      "epoch": 4.52,
-      "learning_rate": 1.2191718602535484e-06,
-      "loss": 0.2408,
+      "epoch": 4.627527300248675,
+      "grad_norm": 0.33214032649993896,
+      "learning_rate": 7.304136286054753e-07,
+      "loss": 0.3687,
       "step": 128400
     },
     {
-      "epoch": 4.52,
-      "learning_rate": 1.2182932612703507e-06,
-      "loss": 0.2649,
+      "epoch": 4.6277074999099,
+      "grad_norm": 0.22773124277591705,
+      "learning_rate": 7.297135666033572e-07,
+      "loss": 0.382,
       "step": 128405
     },
     {
-      "epoch": 4.52,
-      "learning_rate": 1.217414971076941e-06,
-      "loss": 0.2392,
+      "epoch": 4.627887699571125,
+      "grad_norm": 0.28903457522392273,
+      "learning_rate": 7.290138352791936e-07,
+      "loss": 0.365,
       "step": 128410
     },
     {
-      "epoch": 4.52,
-      "learning_rate": 1.2165369896847302e-06,
-      "loss": 0.2655,
+      "epoch": 4.62806789923235,
+      "grad_norm": 0.26188942790031433,
+      "learning_rate": 7.283144346425158e-07,
+      "loss": 0.3868,
       "step": 128415
     },
     {
-      "epoch": 4.52,
-      "learning_rate": 1.2156593171051055e-06,
-      "loss": 0.2692,
+      "epoch": 4.628248098893574,
+      "grad_norm": 0.2626602053642273,
+      "learning_rate": 7.27615364702855e-07,
+      "loss": 0.3717,
       "step": 128420
     },
     {
-      "epoch": 4.52,
-      "learning_rate": 1.2147819533494753e-06,
-      "loss": 0.2493,
+      "epoch": 4.628428298554799,
+      "grad_norm": 0.21673083305358887,
+      "learning_rate": 7.269166254697313e-07,
+      "loss": 0.405,
       "step": 128425
     },
     {
-      "epoch": 4.52,
-      "learning_rate": 1.2139048984292273e-06,
-      "loss": 0.2598,
+      "epoch": 4.628608498216023,
+      "grad_norm": 0.2583613991737366,
+      "learning_rate": 7.26218216952676e-07,
+      "loss": 0.372,
       "step": 128430
     },
     {
-      "epoch": 4.52,
-      "learning_rate": 1.2130281523557468e-06,
-      "loss": 0.2561,
+      "epoch": 4.628788697877248,
+      "grad_norm": 0.2558477818965912,
+      "learning_rate": 7.255201391611954e-07,
+      "loss": 0.3432,
       "step": 128435
     },
     {
-      "epoch": 4.52,
-      "learning_rate": 1.2121517151404193e-06,
-      "loss": 0.2514,
+      "epoch": 4.628968897538472,
+      "grad_norm": 0.23397044837474823,
+      "learning_rate": 7.248223921048014e-07,
+      "loss": 0.3572,
       "step": 128440
     },
     {
-      "epoch": 4.52,
-      "learning_rate": 1.2112755867946246e-06,
-      "loss": 0.2554,
+      "epoch": 4.629149097199697,
+      "grad_norm": 0.23054839670658112,
+      "learning_rate": 7.241249757930057e-07,
+      "loss": 0.352,
       "step": 128445
     },
     {
-      "epoch": 4.52,
-      "learning_rate": 1.2103997673297423e-06,
-      "loss": 0.2531,
+      "epoch": 4.629329296860922,
+      "grad_norm": 0.20861117541790009,
+      "learning_rate": 7.234278902353036e-07,
+      "loss": 0.3908,
       "step": 128450
     },
     {
-      "epoch": 4.52,
-      "learning_rate": 1.2095242567571358e-06,
-      "loss": 0.2466,
+      "epoch": 4.629509496522147,
+      "grad_norm": 0.25009605288505554,
+      "learning_rate": 7.227311354412014e-07,
+      "loss": 0.3717,
       "step": 128455
     },
     {
-      "epoch": 4.52,
-      "learning_rate": 1.2086490550881818e-06,
-      "loss": 0.2515,
+      "epoch": 4.629689696183371,
+      "grad_norm": 0.20757490396499634,
+      "learning_rate": 7.220347114201803e-07,
+      "loss": 0.3404,
       "step": 128460
     },
     {
-      "epoch": 4.52,
-      "learning_rate": 1.2077741623342354e-06,
-      "loss": 0.2532,
+      "epoch": 4.629869895844596,
+      "grad_norm": 0.21684591472148895,
+      "learning_rate": 7.213386181817411e-07,
+      "loss": 0.3976,
       "step": 128465
     },
     {
-      "epoch": 4.52,
-      "learning_rate": 1.2068995785066678e-06,
-      "loss": 0.282,
+      "epoch": 4.630050095505821,
+      "grad_norm": 0.2429099678993225,
+      "learning_rate": 7.206428557353623e-07,
+      "loss": 0.3773,
       "step": 128470
     },
     {
-      "epoch": 4.52,
-      "learning_rate": 1.2060253036168284e-06,
-      "loss": 0.2421,
+      "epoch": 4.6302302951670455,
+      "grad_norm": 0.22892560064792633,
+      "learning_rate": 7.199474240905224e-07,
+      "loss": 0.3536,
       "step": 128475
     },
     {
-      "epoch": 4.52,
-      "learning_rate": 1.2051513376760664e-06,
-      "loss": 0.2655,
+      "epoch": 4.630410494828269,
+      "grad_norm": 0.2792302072048187,
+      "learning_rate": 7.192523232567e-07,
+      "loss": 0.3563,
       "step": 128480
     },
     {
-      "epoch": 4.52,
-      "learning_rate": 1.2042776806957368e-06,
-      "loss": 0.2595,
+      "epoch": 4.630590694489494,
+      "grad_norm": 0.3034243881702423,
+      "learning_rate": 7.185575532433625e-07,
+      "loss": 0.3868,
       "step": 128485
     },
     {
-      "epoch": 4.52,
-      "learning_rate": 1.2034043326871775e-06,
-      "loss": 0.2237,
+      "epoch": 4.630770894150719,
+      "grad_norm": 0.2696426212787628,
+      "learning_rate": 7.178631140599746e-07,
+      "loss": 0.3712,
       "step": 128490
     },
     {
-      "epoch": 4.52,
-      "learning_rate": 1.2025312936617295e-06,
-      "loss": 0.2359,
+      "epoch": 4.630951093811944,
+      "grad_norm": 0.2592863440513611,
+      "learning_rate": 7.171690057160035e-07,
+      "loss": 0.3662,
       "step": 128495
     },
     {
-      "epoch": 4.52,
-      "learning_rate": 1.201658563630731e-06,
-      "loss": 0.2484,
+      "epoch": 4.631131293473168,
+      "grad_norm": 0.27280524373054504,
+      "learning_rate": 7.164752282209031e-07,
+      "loss": 0.381,
       "step": 128500
     },
     {
-      "epoch": 4.52,
-      "eval_loss": 0.24868270754814148,
-      "eval_runtime": 10.5541,
-      "eval_samples_per_second": 9.475,
-      "eval_steps_per_second": 9.475,
+      "epoch": 4.631131293473168,
+      "eval_loss": 0.42883244156837463,
+      "eval_runtime": 3.5159,
+      "eval_samples_per_second": 28.443,
+      "eval_steps_per_second": 7.111,
       "step": 128500
     },
     {
-      "epoch": 4.52,
-      "learning_rate": 1.2007861426055146e-06,
-      "loss": 0.2519,
+      "epoch": 4.631311493134393,
+      "grad_norm": 0.27146467566490173,
+      "learning_rate": 7.157817815841267e-07,
+      "loss": 0.3605,
       "step": 128505
     },
     {
-      "epoch": 4.52,
-      "learning_rate": 1.1999140305974043e-06,
-      "loss": 0.243,
+      "epoch": 4.631491692795618,
+      "grad_norm": 0.2853062152862549,
+      "learning_rate": 7.150886658151224e-07,
+      "loss": 0.3779,
       "step": 128510
     },
     {
-      "epoch": 4.52,
-      "learning_rate": 1.1990422276177221e-06,
-      "loss": 0.2495,
+      "epoch": 4.6316718924568425,
+      "grad_norm": 0.28604602813720703,
+      "learning_rate": 7.143958809233297e-07,
+      "loss": 0.3426,
       "step": 128515
     },
     {
-      "epoch": 4.52,
-      "learning_rate": 1.1981707336777893e-06,
-      "loss": 0.2229,
+      "epoch": 4.631852092118066,
+      "grad_norm": 0.230341836810112,
+      "learning_rate": 7.137034269181969e-07,
+      "loss": 0.3912,
       "step": 128520
     },
     {
-      "epoch": 4.52,
-      "learning_rate": 1.1972995487889299e-06,
-      "loss": 0.255,
+      "epoch": 4.632032291779291,
+      "grad_norm": 0.24526306986808777,
+      "learning_rate": 7.130113038091524e-07,
+      "loss": 0.386,
       "step": 128525
     },
     {
-      "epoch": 4.52,
-      "learning_rate": 1.1964286729624492e-06,
-      "loss": 0.2373,
+      "epoch": 4.632212491440516,
+      "grad_norm": 0.2500110864639282,
+      "learning_rate": 7.12319511605622e-07,
+      "loss": 0.3485,
       "step": 128530
     },
     {
-      "epoch": 4.52,
-      "learning_rate": 1.1955581062096488e-06,
-      "loss": 0.2358,
+      "epoch": 4.632392691101741,
+      "grad_norm": 0.23886556923389435,
+      "learning_rate": 7.116280503170398e-07,
+      "loss": 0.381,
       "step": 128535
     },
     {
-      "epoch": 4.52,
-      "learning_rate": 1.194687848541845e-06,
-      "loss": 0.2546,
+      "epoch": 4.632572890762965,
+      "grad_norm": 0.19476771354675293,
+      "learning_rate": 7.109369199528177e-07,
+      "loss": 0.3623,
       "step": 128540
     },
     {
-      "epoch": 4.52,
-      "learning_rate": 1.1938178999703287e-06,
-      "loss": 0.2521,
+      "epoch": 4.63275309042419,
+      "grad_norm": 0.23774556815624237,
+      "learning_rate": 7.102461205223843e-07,
+      "loss": 0.3641,
       "step": 128545
     },
     {
-      "epoch": 4.52,
-      "learning_rate": 1.1929482605064018e-06,
-      "loss": 0.2572,
+      "epoch": 4.632933290085415,
+      "grad_norm": 0.26775142550468445,
+      "learning_rate": 7.095556520351432e-07,
+      "loss": 0.3751,
       "step": 128550
     },
     {
-      "epoch": 4.52,
-      "learning_rate": 1.1920789301613443e-06,
-      "loss": 0.2352,
+      "epoch": 4.6331134897466395,
+      "grad_norm": 0.24131718277931213,
+      "learning_rate": 7.088655145005008e-07,
+      "loss": 0.3709,
       "step": 128555
     },
     {
-      "epoch": 4.52,
-      "learning_rate": 1.1912099089464584e-06,
-      "loss": 0.2382,
+      "epoch": 4.633293689407864,
+      "grad_norm": 0.2844558358192444,
+      "learning_rate": 7.081757079278662e-07,
+      "loss": 0.3807,
       "step": 128560
     },
     {
-      "epoch": 4.52,
-      "learning_rate": 1.190341196873021e-06,
-      "loss": 0.2519,
+      "epoch": 4.633473889069089,
+      "grad_norm": 0.303622305393219,
+      "learning_rate": 7.074862323266318e-07,
+      "loss": 0.3549,
       "step": 128565
     },
     {
-      "epoch": 4.52,
-      "learning_rate": 1.1894727939523093e-06,
-      "loss": 0.2644,
+      "epoch": 4.633654088730313,
+      "grad_norm": 0.21193251013755798,
+      "learning_rate": 7.067970877061958e-07,
+      "loss": 0.3675,
       "step": 128570
     },
     {
-      "epoch": 4.52,
-      "learning_rate": 1.1886047001956003e-06,
-      "loss": 0.2342,
+      "epoch": 4.6338342883915375,
+      "grad_norm": 0.2827516496181488,
+      "learning_rate": 7.061082740759478e-07,
+      "loss": 0.3776,
       "step": 128575
     },
     {
-      "epoch": 4.52,
-      "learning_rate": 1.1879104477958952e-06,
-      "loss": 0.2639,
+      "epoch": 4.634014488052762,
+      "grad_norm": 0.2597757577896118,
+      "learning_rate": 7.054197914452664e-07,
+      "loss": 0.3736,
       "step": 128580
     },
     {
-      "epoch": 4.52,
-      "learning_rate": 1.187042910562794e-06,
-      "loss": 0.2323,
+      "epoch": 4.634194687713987,
+      "grad_norm": 0.23663856089115143,
+      "learning_rate": 7.047316398235387e-07,
+      "loss": 0.3633,
       "step": 128585
     },
     {
-      "epoch": 4.52,
-      "learning_rate": 1.1861756825252534e-06,
-      "loss": 0.2592,
+      "epoch": 4.634374887375212,
+      "grad_norm": 0.22754615545272827,
+      "learning_rate": 7.040438192201403e-07,
+      "loss": 0.3843,
       "step": 128590
     },
     {
-      "epoch": 4.52,
-      "learning_rate": 1.185308763694526e-06,
-      "loss": 0.2256,
+      "epoch": 4.6345550870364365,
+      "grad_norm": 0.28135913610458374,
+      "learning_rate": 7.033563296444417e-07,
+      "loss": 0.354,
       "step": 128595
     },
     {
-      "epoch": 4.52,
-      "learning_rate": 1.1844421540818662e-06,
-      "loss": 0.2473,
+      "epoch": 4.634735286697661,
+      "grad_norm": 0.31850412487983704,
+      "learning_rate": 7.026691711058076e-07,
+      "loss": 0.3579,
       "step": 128600
     },
     {
-      "epoch": 4.52,
-      "learning_rate": 1.1835758536985347e-06,
-      "loss": 0.2467,
+      "epoch": 4.634915486358886,
+      "grad_norm": 0.24080584943294525,
+      "learning_rate": 7.019823436136024e-07,
+      "loss": 0.3839,
       "step": 128605
     },
     {
-      "epoch": 4.52,
-      "learning_rate": 1.1827098625557698e-06,
-      "loss": 0.2349,
+      "epoch": 4.63509568602011,
+      "grad_norm": 0.2874895930290222,
+      "learning_rate": 7.012958471771803e-07,
+      "loss": 0.3642,
       "step": 128610
     },
     {
-      "epoch": 4.53,
-      "learning_rate": 1.1818441806648262e-06,
-      "loss": 0.2594,
+      "epoch": 4.6352758856813345,
+      "grad_norm": 0.24868187308311462,
+      "learning_rate": 7.006096818059e-07,
+      "loss": 0.3436,
       "step": 128615
     },
     {
-      "epoch": 4.53,
-      "learning_rate": 1.1809788080369393e-06,
-      "loss": 0.2482,
+      "epoch": 4.635456085342559,
+      "grad_norm": 0.24118977785110474,
+      "learning_rate": 6.999238475091096e-07,
+      "loss": 0.3891,
       "step": 128620
     },
     {
-      "epoch": 4.53,
-      "learning_rate": 1.1801137446833422e-06,
-      "loss": 0.2539,
+      "epoch": 4.635636285003784,
+      "grad_norm": 0.202513188123703,
+      "learning_rate": 6.99238344296152e-07,
+      "loss": 0.3655,
       "step": 128625
     },
     {
-      "epoch": 4.53,
-      "learning_rate": 1.1792489906152726e-06,
-      "loss": 0.2359,
+      "epoch": 4.635816484665009,
+      "grad_norm": 0.24123841524124146,
+      "learning_rate": 6.985531721763666e-07,
+      "loss": 0.4058,
       "step": 128630
     },
     {
-      "epoch": 4.53,
-      "learning_rate": 1.1783845458439552e-06,
-      "loss": 0.2332,
+      "epoch": 4.6359966843262335,
+      "grad_norm": 0.2426537424325943,
+      "learning_rate": 6.978683311590877e-07,
+      "loss": 0.3598,
       "step": 128635
     },
     {
-      "epoch": 4.53,
-      "learning_rate": 1.177520410380617e-06,
-      "loss": 0.2727,
+      "epoch": 4.636176883987458,
+      "grad_norm": 0.2862418591976166,
+      "learning_rate": 6.971838212536524e-07,
+      "loss": 0.3541,
       "step": 128640
     },
     {
-      "epoch": 4.53,
-      "learning_rate": 1.176656584236474e-06,
-      "loss": 0.246,
+      "epoch": 4.636357083648683,
+      "grad_norm": 0.24865272641181946,
+      "learning_rate": 6.96499642469381e-07,
+      "loss": 0.3766,
       "step": 128645
     },
     {
-      "epoch": 4.53,
-      "learning_rate": 1.175793067422748e-06,
-      "loss": 0.2481,
+      "epoch": 4.636537283309908,
+      "grad_norm": 0.31950682401657104,
+      "learning_rate": 6.958157948155936e-07,
+      "loss": 0.347,
       "step": 128650
     },
     {
-      "epoch": 4.53,
-      "learning_rate": 1.1749298599506464e-06,
-      "loss": 0.2231,
+      "epoch": 4.636717482971132,
+      "grad_norm": 0.2226303666830063,
+      "learning_rate": 6.951322783016107e-07,
+      "loss": 0.3515,
       "step": 128655
     },
     {
-      "epoch": 4.53,
-      "learning_rate": 1.174066961831377e-06,
-      "loss": 0.2451,
+      "epoch": 4.636897682632357,
+      "grad_norm": 0.22242146730422974,
+      "learning_rate": 6.944490929367469e-07,
+      "loss": 0.3747,
       "step": 128660
     },
     {
-      "epoch": 4.53,
-      "learning_rate": 1.1732043730761478e-06,
-      "loss": 0.2431,
+      "epoch": 4.637077882293581,
+      "grad_norm": 0.3434227705001831,
+      "learning_rate": 6.93766238730309e-07,
+      "loss": 0.399,
       "step": 128665
     },
     {
-      "epoch": 4.53,
-      "learning_rate": 1.1723420936961576e-06,
-      "loss": 0.2217,
+      "epoch": 4.637258081954806,
+      "grad_norm": 0.27048400044441223,
+      "learning_rate": 6.930837156916004e-07,
+      "loss": 0.3702,
       "step": 128670
     },
     {
-      "epoch": 4.53,
-      "learning_rate": 1.1714801237026036e-06,
-      "loss": 0.252,
+      "epoch": 4.6374382816160304,
+      "grad_norm": 0.2766757011413574,
+      "learning_rate": 6.924015238299164e-07,
+      "loss": 0.373,
       "step": 128675
     },
     {
-      "epoch": 4.53,
-      "learning_rate": 1.1706184631066737e-06,
-      "loss": 0.2514,
+      "epoch": 4.637618481277255,
+      "grad_norm": 0.22949692606925964,
+      "learning_rate": 6.91719663154558e-07,
+      "loss": 0.3402,
       "step": 128680
     },
     {
-      "epoch": 4.53,
-      "learning_rate": 1.1697571119195643e-06,
-      "loss": 0.229,
+      "epoch": 4.63779868093848,
+      "grad_norm": 0.2594715654850006,
+      "learning_rate": 6.910381336748123e-07,
+      "loss": 0.3437,
       "step": 128685
     },
     {
-      "epoch": 4.53,
-      "learning_rate": 1.1688960701524503e-06,
-      "loss": 0.2412,
+      "epoch": 4.637978880599705,
+      "grad_norm": 0.24005542695522308,
+      "learning_rate": 6.903569353999661e-07,
+      "loss": 0.3574,
       "step": 128690
     },
     {
-      "epoch": 4.53,
-      "learning_rate": 1.1680353378165172e-06,
-      "loss": 0.2556,
+      "epoch": 4.638159080260929,
+      "grad_norm": 0.2983018755912781,
+      "learning_rate": 6.896760683393011e-07,
+      "loss": 0.366,
       "step": 128695
     },
     {
-      "epoch": 4.53,
-      "learning_rate": 1.167174914922936e-06,
-      "loss": 0.2428,
+      "epoch": 4.638339279922154,
+      "grad_norm": 0.2457517385482788,
+      "learning_rate": 6.889955325020903e-07,
+      "loss": 0.3647,
       "step": 128700
     },
     {
-      "epoch": 4.53,
-      "learning_rate": 1.1663148014828873e-06,
-      "loss": 0.2401,
+      "epoch": 4.638519479583378,
+      "grad_norm": 0.27278947830200195,
+      "learning_rate": 6.883153278976123e-07,
+      "loss": 0.3514,
       "step": 128705
     },
     {
-      "epoch": 4.53,
-      "learning_rate": 1.1654549975075314e-06,
-      "loss": 0.2426,
+      "epoch": 4.638699679244603,
+      "grad_norm": 0.3145797848701477,
+      "learning_rate": 6.876354545351293e-07,
+      "loss": 0.3964,
       "step": 128710
     },
     {
-      "epoch": 4.53,
-      "learning_rate": 1.1645955030080313e-06,
-      "loss": 0.2618,
+      "epoch": 4.638879878905827,
+      "grad_norm": 0.26128581166267395,
+      "learning_rate": 6.869559124239061e-07,
+      "loss": 0.3822,
       "step": 128715
     },
     {
-      "epoch": 4.53,
-      "learning_rate": 1.1637363179955535e-06,
-      "loss": 0.2434,
+      "epoch": 4.639060078567052,
+      "grad_norm": 0.2640693187713623,
+      "learning_rate": 6.862767015732019e-07,
+      "loss": 0.3944,
       "step": 128720
     },
     {
-      "epoch": 4.53,
-      "learning_rate": 1.1628774424812528e-06,
-      "loss": 0.2437,
+      "epoch": 4.639240278228277,
+      "grad_norm": 0.24374765157699585,
+      "learning_rate": 6.855978219922676e-07,
+      "loss": 0.3894,
       "step": 128725
     },
     {
-      "epoch": 4.53,
-      "learning_rate": 1.1620188764762785e-06,
-      "loss": 0.2514,
+      "epoch": 4.639420477889502,
+      "grad_norm": 0.24732322990894318,
+      "learning_rate": 6.849192736903598e-07,
+      "loss": 0.3685,
       "step": 128730
     },
     {
-      "epoch": 4.53,
-      "learning_rate": 1.1611606199917773e-06,
-      "loss": 0.2378,
+      "epoch": 4.639600677550726,
+      "grad_norm": 0.307091623544693,
+      "learning_rate": 6.84241056676721e-07,
+      "loss": 0.4219,
       "step": 128735
     },
     {
-      "epoch": 4.53,
-      "learning_rate": 1.1603026730388962e-06,
-      "loss": 0.2316,
+      "epoch": 4.639780877211951,
+      "grad_norm": 0.2447475790977478,
+      "learning_rate": 6.835631709605856e-07,
+      "loss": 0.3844,
       "step": 128740
     },
     {
-      "epoch": 4.53,
-      "learning_rate": 1.1594450356287727e-06,
-      "loss": 0.2539,
+      "epoch": 4.639961076873176,
+      "grad_norm": 0.21926872432231903,
+      "learning_rate": 6.828856165511932e-07,
+      "loss": 0.3567,
       "step": 128745
     },
     {
-      "epoch": 4.53,
-      "learning_rate": 1.158587707772546e-06,
-      "loss": 0.2615,
+      "epoch": 4.640141276534401,
+      "grad_norm": 0.27030664682388306,
+      "learning_rate": 6.822083934577811e-07,
+      "loss": 0.4035,
       "step": 128750
     },
     {
-      "epoch": 4.53,
-      "learning_rate": 1.1577306894813483e-06,
-      "loss": 0.2508,
+      "epoch": 4.640321476195624,
+      "grad_norm": 0.28935354948043823,
+      "learning_rate": 6.815315016895696e-07,
+      "loss": 0.3833,
       "step": 128755
     },
     {
-      "epoch": 4.53,
-      "learning_rate": 1.1568739807662988e-06,
-      "loss": 0.2585,
+      "epoch": 4.640501675856849,
+      "grad_norm": 0.25917571783065796,
+      "learning_rate": 6.808549412557819e-07,
+      "loss": 0.3641,
       "step": 128760
     },
     {
-      "epoch": 4.53,
-      "learning_rate": 1.1560175816385298e-06,
-      "loss": 0.2464,
+      "epoch": 4.640681875518074,
+      "grad_norm": 0.2617967426776886,
+      "learning_rate": 6.801787121656355e-07,
+      "loss": 0.3447,
       "step": 128765
     },
     {
-      "epoch": 4.53,
-      "learning_rate": 1.155161492109158e-06,
-      "loss": 0.253,
+      "epoch": 4.640862075179299,
+      "grad_norm": 0.23066262900829315,
+      "learning_rate": 6.795028144283483e-07,
+      "loss": 0.3868,
       "step": 128770
     },
     {
-      "epoch": 4.53,
-      "learning_rate": 1.154305712189302e-06,
-      "loss": 0.2679,
+      "epoch": 4.641042274840523,
+      "grad_norm": 0.20748838782310486,
+      "learning_rate": 6.788272480531266e-07,
+      "loss": 0.3589,
       "step": 128775
     },
     {
-      "epoch": 4.53,
-      "learning_rate": 1.1534502418900667e-06,
-      "loss": 0.2551,
+      "epoch": 4.641222474501748,
+      "grad_norm": 0.2675074636936188,
+      "learning_rate": 6.781520130491742e-07,
+      "loss": 0.3952,
       "step": 128780
     },
     {
-      "epoch": 4.53,
-      "learning_rate": 1.152595081222571e-06,
-      "loss": 0.2369,
+      "epoch": 4.641402674162973,
+      "grad_norm": 0.25951436161994934,
+      "learning_rate": 6.774771094256949e-07,
+      "loss": 0.3544,
       "step": 128785
     },
     {
-      "epoch": 4.53,
-      "learning_rate": 1.1517402301979092e-06,
-      "loss": 0.233,
+      "epoch": 4.6415828738241975,
+      "grad_norm": 0.23791882395744324,
+      "learning_rate": 6.768025371918785e-07,
+      "loss": 0.379,
       "step": 128790
     },
     {
-      "epoch": 4.53,
-      "learning_rate": 1.150885688827183e-06,
-      "loss": 0.2555,
+      "epoch": 4.641763073485421,
+      "grad_norm": 0.2497406154870987,
+      "learning_rate": 6.761282963569149e-07,
+      "loss": 0.3833,
       "step": 128795
     },
     {
-      "epoch": 4.53,
-      "learning_rate": 1.1500314571214921e-06,
-      "loss": 0.2557,
+      "epoch": 4.641943273146646,
+      "grad_norm": 0.2554693818092346,
+      "learning_rate": 6.754543869299996e-07,
+      "loss": 0.3814,
       "step": 128800
     },
     {
-      "epoch": 4.53,
-      "learning_rate": 1.149177535091922e-06,
-      "loss": 0.2527,
+      "epoch": 4.642123472807871,
+      "grad_norm": 0.22991681098937988,
+      "learning_rate": 6.747808089203056e-07,
+      "loss": 0.3698,
       "step": 128805
     },
     {
-      "epoch": 4.53,
-      "learning_rate": 1.1483239227495668e-06,
-      "loss": 0.2333,
+      "epoch": 4.642303672469096,
+      "grad_norm": 0.2903478741645813,
+      "learning_rate": 6.741075623370147e-07,
+      "loss": 0.3504,
       "step": 128810
     },
     {
-      "epoch": 4.53,
-      "learning_rate": 1.1474706201055007e-06,
-      "loss": 0.2304,
+      "epoch": 4.64248387213032,
+      "grad_norm": 0.21784360706806183,
+      "learning_rate": 6.734346471893e-07,
+      "loss": 0.3714,
       "step": 128815
     },
     {
-      "epoch": 4.53,
-      "learning_rate": 1.146617627170815e-06,
-      "loss": 0.2461,
+      "epoch": 4.642664071791545,
+      "grad_norm": 0.2957072854042053,
+      "learning_rate": 6.727620634863235e-07,
+      "loss": 0.3464,
       "step": 128820
     },
     {
-      "epoch": 4.53,
-      "learning_rate": 1.1457649439565733e-06,
-      "loss": 0.2539,
+      "epoch": 4.64284427145277,
+      "grad_norm": 0.2421649545431137,
+      "learning_rate": 6.720898112372614e-07,
+      "loss": 0.354,
       "step": 128825
     },
     {
-      "epoch": 4.53,
-      "learning_rate": 1.1449125704738606e-06,
-      "loss": 0.2542,
+      "epoch": 4.6430244711139945,
+      "grad_norm": 0.21480095386505127,
+      "learning_rate": 6.714178904512619e-07,
+      "loss": 0.3823,
       "step": 128830
     },
     {
-      "epoch": 4.53,
-      "learning_rate": 1.1440605067337295e-06,
-      "loss": 0.2524,
+      "epoch": 4.643204670775219,
+      "grad_norm": 0.22311148047447205,
+      "learning_rate": 6.707463011374815e-07,
+      "loss": 0.3979,
       "step": 128835
     },
     {
-      "epoch": 4.53,
-      "learning_rate": 1.1432087527472574e-06,
-      "loss": 0.2413,
+      "epoch": 4.643384870436444,
+      "grad_norm": 0.2693723440170288,
+      "learning_rate": 6.70075043305074e-07,
+      "loss": 0.3854,
       "step": 128840
     },
     {
-      "epoch": 4.53,
-      "learning_rate": 1.1423573085254963e-06,
-      "loss": 0.2529,
+      "epoch": 4.643565070097668,
+      "grad_norm": 0.29857292771339417,
+      "learning_rate": 6.69404116963182e-07,
+      "loss": 0.3685,
       "step": 128845
     },
     {
-      "epoch": 4.53,
-      "learning_rate": 1.1415061740795014e-06,
-      "loss": 0.2289,
+      "epoch": 4.643745269758893,
+      "grad_norm": 0.2625799775123596,
+      "learning_rate": 6.68733522120954e-07,
+      "loss": 0.3603,
       "step": 128850
     },
     {
-      "epoch": 4.53,
-      "learning_rate": 1.1406553494203247e-06,
-      "loss": 0.2514,
+      "epoch": 4.643925469420117,
+      "grad_norm": 0.2390749603509903,
+      "learning_rate": 6.680632587875185e-07,
+      "loss": 0.3602,
       "step": 128855
     },
     {
-      "epoch": 4.53,
-      "learning_rate": 1.1398048345590162e-06,
-      "loss": 0.255,
+      "epoch": 4.644105669081342,
+      "grad_norm": 0.28638026118278503,
+      "learning_rate": 6.673933269720072e-07,
+      "loss": 0.3673,
       "step": 128860
     },
     {
-      "epoch": 4.53,
-      "learning_rate": 1.1389546295066166e-06,
-      "loss": 0.2527,
+      "epoch": 4.644285868742567,
+      "grad_norm": 0.2660246789455414,
+      "learning_rate": 6.667237266835519e-07,
+      "loss": 0.3327,
       "step": 128865
     },
     {
-      "epoch": 4.53,
-      "learning_rate": 1.1381047342741674e-06,
-      "loss": 0.2329,
+      "epoch": 4.6444660684037915,
+      "grad_norm": 0.218661829829216,
+      "learning_rate": 6.660544579312755e-07,
+      "loss": 0.3471,
       "step": 128870
     },
     {
-      "epoch": 4.53,
-      "learning_rate": 1.1372551488726985e-06,
-      "loss": 0.2546,
+      "epoch": 4.644646268065016,
+      "grad_norm": 0.2507161498069763,
+      "learning_rate": 6.65385520724296e-07,
+      "loss": 0.3942,
       "step": 128875
     },
     {
-      "epoch": 4.53,
-      "learning_rate": 1.1364058733132483e-06,
-      "loss": 0.2448,
+      "epoch": 4.644826467726241,
+      "grad_norm": 0.2658718526363373,
+      "learning_rate": 6.64716915071728e-07,
+      "loss": 0.3769,
       "step": 128880
     },
     {
-      "epoch": 4.53,
-      "learning_rate": 1.1355569076068411e-06,
-      "loss": 0.2494,
+      "epoch": 4.645006667387465,
+      "grad_norm": 0.24641652405261993,
+      "learning_rate": 6.640486409826785e-07,
+      "loss": 0.3482,
       "step": 128885
     },
     {
-      "epoch": 4.53,
-      "learning_rate": 1.134708251764502e-06,
-      "loss": 0.2349,
+      "epoch": 4.6451868670486895,
+      "grad_norm": 0.27895164489746094,
+      "learning_rate": 6.633806984662566e-07,
+      "loss": 0.3843,
       "step": 128890
     },
     {
-      "epoch": 4.53,
-      "learning_rate": 1.1338599057972437e-06,
-      "loss": 0.245,
+      "epoch": 4.645367066709914,
+      "grad_norm": 0.2246672362089157,
+      "learning_rate": 6.627130875315607e-07,
+      "loss": 0.3708,
       "step": 128895
     },
     {
-      "epoch": 4.54,
-      "learning_rate": 1.1330118697160914e-06,
-      "loss": 0.2504,
+      "epoch": 4.645547266371139,
+      "grad_norm": 0.2589907944202423,
+      "learning_rate": 6.620458081876891e-07,
+      "loss": 0.364,
       "step": 128900
     },
     {
-      "epoch": 4.54,
-      "learning_rate": 1.1321641435320495e-06,
-      "loss": 0.2522,
+      "epoch": 4.645727466032364,
+      "grad_norm": 0.31070613861083984,
+      "learning_rate": 6.613788604437288e-07,
+      "loss": 0.3812,
       "step": 128905
     },
     {
-      "epoch": 4.54,
-      "learning_rate": 1.1313167272561238e-06,
-      "loss": 0.242,
+      "epoch": 4.6459076656935885,
+      "grad_norm": 0.24168938398361206,
+      "learning_rate": 6.607122443087672e-07,
+      "loss": 0.3644,
       "step": 128910
     },
     {
-      "epoch": 4.54,
-      "learning_rate": 1.1304696208993215e-06,
-      "loss": 0.2493,
+      "epoch": 4.646087865354813,
+      "grad_norm": 0.24737270176410675,
+      "learning_rate": 6.60045959791894e-07,
+      "loss": 0.361,
       "step": 128915
     },
     {
-      "epoch": 4.54,
-      "learning_rate": 1.1296228244726426e-06,
-      "loss": 0.2431,
+      "epoch": 4.646268065016038,
+      "grad_norm": 0.24914908409118652,
+      "learning_rate": 6.593800069021827e-07,
+      "loss": 0.3888,
       "step": 128920
     },
     {
-      "epoch": 4.54,
-      "learning_rate": 1.1287763379870809e-06,
-      "loss": 0.2553,
+      "epoch": 4.646448264677263,
+      "grad_norm": 0.24401962757110596,
+      "learning_rate": 6.587143856487011e-07,
+      "loss": 0.3582,
       "step": 128925
     },
     {
-      "epoch": 4.54,
-      "learning_rate": 1.127930161453622e-06,
-      "loss": 0.2363,
+      "epoch": 4.646628464338487,
+      "grad_norm": 0.24380113184452057,
+      "learning_rate": 6.58049096040525e-07,
+      "loss": 0.3765,
       "step": 128930
     },
     {
-      "epoch": 4.54,
-      "learning_rate": 1.1270842948832599e-06,
-      "loss": 0.2522,
+      "epoch": 4.646808663999712,
+      "grad_norm": 0.241029292345047,
+      "learning_rate": 6.573841380867197e-07,
+      "loss": 0.3522,
       "step": 128935
     },
     {
-      "epoch": 4.54,
-      "learning_rate": 1.1262387382869778e-06,
-      "loss": 0.254,
+      "epoch": 4.646988863660936,
+      "grad_norm": 0.229360893368721,
+      "learning_rate": 6.567195117963415e-07,
+      "loss": 0.36,
       "step": 128940
     },
     {
-      "epoch": 4.54,
-      "learning_rate": 1.1253934916757496e-06,
-      "loss": 0.2429,
+      "epoch": 4.647169063322161,
+      "grad_norm": 0.2620939016342163,
+      "learning_rate": 6.560552171784473e-07,
+      "loss": 0.4135,
       "step": 128945
     },
     {
-      "epoch": 4.54,
-      "learning_rate": 1.1245485550605534e-06,
-      "loss": 0.2606,
+      "epoch": 4.6473492629833855,
+      "grad_norm": 0.2812032997608185,
+      "learning_rate": 6.553912542420826e-07,
+      "loss": 0.3603,
       "step": 128950
     },
     {
-      "epoch": 4.54,
-      "learning_rate": 1.1237039284523605e-06,
-      "loss": 0.2483,
+      "epoch": 4.64752946264461,
+      "grad_norm": 0.27785706520080566,
+      "learning_rate": 6.54727622996304e-07,
+      "loss": 0.3793,
       "step": 128955
     },
     {
-      "epoch": 4.54,
-      "learning_rate": 1.122859611862137e-06,
-      "loss": 0.2245,
+      "epoch": 4.647709662305835,
+      "grad_norm": 0.2647446095943451,
+      "learning_rate": 6.540643234501487e-07,
+      "loss": 0.3661,
       "step": 128960
     },
     {
-      "epoch": 4.54,
-      "learning_rate": 1.1220156053008468e-06,
-      "loss": 0.2884,
+      "epoch": 4.64788986196706,
+      "grad_norm": 0.28079158067703247,
+      "learning_rate": 6.534013556126511e-07,
+      "loss": 0.3658,
       "step": 128965
     },
     {
-      "epoch": 4.54,
-      "learning_rate": 1.1211719087794447e-06,
-      "loss": 0.2425,
+      "epoch": 4.648070061628284,
+      "grad_norm": 0.291545569896698,
+      "learning_rate": 6.527387194928486e-07,
+      "loss": 0.3871,
       "step": 128970
     },
     {
-      "epoch": 4.54,
-      "learning_rate": 1.1203285223088915e-06,
-      "loss": 0.2316,
+      "epoch": 4.648250261289509,
+      "grad_norm": 0.24639180302619934,
+      "learning_rate": 6.520764150997644e-07,
+      "loss": 0.3754,
       "step": 128975
     },
     {
-      "epoch": 4.54,
-      "learning_rate": 1.119485445900137e-06,
-      "loss": 0.2686,
+      "epoch": 4.648430460950733,
+      "grad_norm": 0.1946740746498108,
+      "learning_rate": 6.514144424424246e-07,
+      "loss": 0.3622,
       "step": 128980
     },
     {
-      "epoch": 4.54,
-      "learning_rate": 1.1186426795641192e-06,
-      "loss": 0.2521,
+      "epoch": 4.648610660611958,
+      "grad_norm": 0.24029649794101715,
+      "learning_rate": 6.507528015298525e-07,
+      "loss": 0.3713,
       "step": 128985
     },
     {
-      "epoch": 4.54,
-      "learning_rate": 1.1178002233117934e-06,
-      "loss": 0.2504,
+      "epoch": 4.6487908602731824,
+      "grad_norm": 0.22794067859649658,
+      "learning_rate": 6.500914923710577e-07,
+      "loss": 0.3872,
       "step": 128990
     },
     {
-      "epoch": 4.54,
-      "learning_rate": 1.11695807715409e-06,
-      "loss": 0.2675,
+      "epoch": 4.648971059934407,
+      "grad_norm": 0.2424376904964447,
+      "learning_rate": 6.494305149750524e-07,
+      "loss": 0.3783,
       "step": 128995
     },
     {
-      "epoch": 4.54,
-      "learning_rate": 1.1161162411019471e-06,
-      "loss": 0.2357,
+      "epoch": 4.649151259595632,
+      "grad_norm": 0.27544233202934265,
+      "learning_rate": 6.487698693508432e-07,
+      "loss": 0.3789,
       "step": 129000
     },
     {
-      "epoch": 4.54,
-      "eval_loss": 0.24872316420078278,
-      "eval_runtime": 10.5448,
-      "eval_samples_per_second": 9.483,
-      "eval_steps_per_second": 9.483,
+      "epoch": 4.649151259595632,
+      "eval_loss": 0.4288814663887024,
+      "eval_runtime": 3.5274,
+      "eval_samples_per_second": 28.349,
+      "eval_steps_per_second": 7.087,
       "step": 129000
     },
     {
-      "epoch": 4.54,
-      "learning_rate": 1.115274715166298e-06,
-      "loss": 0.2395,
+      "epoch": 4.649331459256857,
+      "grad_norm": 0.23391346633434296,
+      "learning_rate": 6.481095555074257e-07,
+      "loss": 0.3823,
       "step": 129005
     },
     {
-      "epoch": 4.54,
-      "learning_rate": 1.1144334993580613e-06,
-      "loss": 0.2523,
+      "epoch": 4.649511658918081,
+      "grad_norm": 0.3378024697303772,
+      "learning_rate": 6.474495734538038e-07,
+      "loss": 0.381,
       "step": 129010
     },
     {
-      "epoch": 4.54,
-      "learning_rate": 1.1135925936881674e-06,
-      "loss": 0.2347,
+      "epoch": 4.649691858579306,
+      "grad_norm": 0.23834291100502014,
+      "learning_rate": 6.467899231989704e-07,
+      "loss": 0.3852,
       "step": 129015
     },
     {
-      "epoch": 4.54,
-      "learning_rate": 1.1127519981675299e-06,
-      "loss": 0.2594,
+      "epoch": 4.649872058240531,
+      "grad_norm": 0.2810116410255432,
+      "learning_rate": 6.461306047519017e-07,
+      "loss": 0.3648,
       "step": 129020
     },
     {
-      "epoch": 4.54,
-      "learning_rate": 1.1119117128070649e-06,
-      "loss": 0.2511,
+      "epoch": 4.650052257901756,
+      "grad_norm": 0.2411430925130844,
+      "learning_rate": 6.45471618121593e-07,
+      "loss": 0.3573,
       "step": 129025
     },
     {
-      "epoch": 4.54,
-      "learning_rate": 1.111071737617686e-06,
-      "loss": 0.2486,
+      "epoch": 4.650232457562979,
+      "grad_norm": 0.22861650586128235,
+      "learning_rate": 6.448129633170153e-07,
+      "loss": 0.3441,
       "step": 129030
     },
     {
-      "epoch": 4.54,
-      "learning_rate": 1.1102320726102983e-06,
-      "loss": 0.2397,
+      "epoch": 4.650412657224204,
+      "grad_norm": 0.29254698753356934,
+      "learning_rate": 6.441546403471499e-07,
+      "loss": 0.3794,
       "step": 129035
     },
     {
-      "epoch": 4.54,
-      "learning_rate": 1.1093927177957985e-06,
-      "loss": 0.241,
+      "epoch": 4.650592856885429,
+      "grad_norm": 0.2996603548526764,
+      "learning_rate": 6.434966492209621e-07,
+      "loss": 0.375,
       "step": 129040
     },
     {
-      "epoch": 4.54,
-      "learning_rate": 1.1085536731850947e-06,
-      "loss": 0.2557,
+      "epoch": 4.650773056546654,
+      "grad_norm": 0.29424965381622314,
+      "learning_rate": 6.428389899474113e-07,
+      "loss": 0.3789,
       "step": 129045
     },
     {
-      "epoch": 4.54,
-      "learning_rate": 1.10771493878907e-06,
-      "loss": 0.2746,
+      "epoch": 4.650953256207878,
+      "grad_norm": 0.22816717624664307,
+      "learning_rate": 6.421816625354682e-07,
+      "loss": 0.3889,
       "step": 129050
     },
     {
-      "epoch": 4.54,
-      "learning_rate": 1.1068765146186266e-06,
-      "loss": 0.2543,
+      "epoch": 4.651133455869103,
+      "grad_norm": 0.29083213210105896,
+      "learning_rate": 6.415246669940838e-07,
+      "loss": 0.3914,
       "step": 129055
     },
     {
-      "epoch": 4.54,
-      "learning_rate": 1.1060384006846475e-06,
-      "loss": 0.2526,
+      "epoch": 4.651313655530328,
+      "grad_norm": 0.24897891283035278,
+      "learning_rate": 6.408680033322096e-07,
+      "loss": 0.382,
       "step": 129060
     },
     {
-      "epoch": 4.54,
-      "learning_rate": 1.1052005969980073e-06,
-      "loss": 0.2341,
+      "epoch": 4.651493855191553,
+      "grad_norm": 0.2584581971168518,
+      "learning_rate": 6.402116715587908e-07,
+      "loss": 0.41,
       "step": 129065
     },
     {
-      "epoch": 4.54,
-      "learning_rate": 1.1043631035695918e-06,
-      "loss": 0.2538,
+      "epoch": 4.651674054852776,
+      "grad_norm": 0.26626285910606384,
+      "learning_rate": 6.395556716827705e-07,
+      "loss": 0.3482,
       "step": 129070
     },
     {
-      "epoch": 4.54,
-      "learning_rate": 1.1035259204102756e-06,
-      "loss": 0.2729,
+      "epoch": 4.651854254514001,
+      "grad_norm": 0.29574957489967346,
+      "learning_rate": 6.389000037130916e-07,
+      "loss": 0.3603,
       "step": 129075
     },
     {
-      "epoch": 4.54,
-      "learning_rate": 1.1026890475309277e-06,
-      "loss": 0.2602,
+      "epoch": 4.652034454175226,
+      "grad_norm": 0.21474981307983398,
+      "learning_rate": 6.382446676586828e-07,
+      "loss": 0.3585,
       "step": 129080
     },
     {
-      "epoch": 4.54,
-      "learning_rate": 1.1018524849424089e-06,
-      "loss": 0.2441,
+      "epoch": 4.652214653836451,
+      "grad_norm": 0.25126680731773376,
+      "learning_rate": 6.375896635284734e-07,
+      "loss": 0.385,
       "step": 129085
     },
     {
-      "epoch": 4.54,
-      "learning_rate": 1.1010162326555912e-06,
-      "loss": 0.2472,
+      "epoch": 4.652394853497675,
+      "grad_norm": 0.20911748707294464,
+      "learning_rate": 6.369349913313893e-07,
+      "loss": 0.3573,
       "step": 129090
     },
     {
-      "epoch": 4.54,
-      "learning_rate": 1.100180290681327e-06,
-      "loss": 0.2702,
+      "epoch": 4.6525750531589,
+      "grad_norm": 0.24984316527843475,
+      "learning_rate": 6.362806510763458e-07,
+      "loss": 0.3599,
       "step": 129095
     },
     {
-      "epoch": 4.54,
-      "learning_rate": 1.0993446590304658e-06,
-      "loss": 0.257,
+      "epoch": 4.652755252820125,
+      "grad_norm": 0.20926958322525024,
+      "learning_rate": 6.356266427722663e-07,
+      "loss": 0.37,
       "step": 129100
     },
     {
-      "epoch": 4.54,
-      "learning_rate": 1.0985093377138656e-06,
-      "loss": 0.2181,
+      "epoch": 4.6529354524813495,
+      "grad_norm": 0.23557382822036743,
+      "learning_rate": 6.349729664280546e-07,
+      "loss": 0.3507,
       "step": 129105
     },
     {
-      "epoch": 4.54,
-      "learning_rate": 1.0976743267423705e-06,
-      "loss": 0.2417,
+      "epoch": 4.653115652142574,
+      "grad_norm": 0.27930939197540283,
+      "learning_rate": 6.343196220526176e-07,
+      "loss": 0.3653,
       "step": 129110
     },
     {
-      "epoch": 4.54,
-      "learning_rate": 1.096839626126825e-06,
-      "loss": 0.2334,
+      "epoch": 4.653295851803799,
+      "grad_norm": 0.20511141419410706,
+      "learning_rate": 6.336666096548593e-07,
+      "loss": 0.3598,
       "step": 129115
     },
     {
-      "epoch": 4.54,
-      "learning_rate": 1.0960052358780615e-06,
-      "loss": 0.2583,
+      "epoch": 4.653476051465024,
+      "grad_norm": 0.2338797003030777,
+      "learning_rate": 6.330139292436782e-07,
+      "loss": 0.3693,
       "step": 129120
     },
     {
-      "epoch": 4.54,
-      "learning_rate": 1.0951711560069161e-06,
-      "loss": 0.254,
+      "epoch": 4.653656251126248,
+      "grad_norm": 0.29057690501213074,
+      "learning_rate": 6.323615808279587e-07,
+      "loss": 0.3527,
       "step": 129125
     },
     {
-      "epoch": 4.54,
-      "learning_rate": 1.0943373865242218e-06,
-      "loss": 0.2625,
+      "epoch": 4.653836450787472,
+      "grad_norm": 0.26193034648895264,
+      "learning_rate": 6.317095644166021e-07,
+      "loss": 0.379,
       "step": 129130
     },
     {
-      "epoch": 4.54,
-      "learning_rate": 1.0935039274408003e-06,
-      "loss": 0.2485,
+      "epoch": 4.654016650448697,
+      "grad_norm": 0.2552226781845093,
+      "learning_rate": 6.310578800184791e-07,
+      "loss": 0.3968,
       "step": 129135
     },
     {
-      "epoch": 4.54,
-      "learning_rate": 1.0926707787674767e-06,
-      "loss": 0.2437,
+      "epoch": 4.654196850109922,
+      "grad_norm": 0.22633424401283264,
+      "learning_rate": 6.30406527642477e-07,
+      "loss": 0.3954,
       "step": 129140
     },
     {
-      "epoch": 4.54,
-      "learning_rate": 1.091837940515064e-06,
-      "loss": 0.2593,
+      "epoch": 4.6543770497711465,
+      "grad_norm": 0.2168351113796234,
+      "learning_rate": 6.297555072974692e-07,
+      "loss": 0.4123,
       "step": 129145
     },
     {
-      "epoch": 4.54,
-      "learning_rate": 1.0910054126943846e-06,
-      "loss": 0.2597,
+      "epoch": 4.654557249432371,
+      "grad_norm": 0.2150443196296692,
+      "learning_rate": 6.291048189923238e-07,
+      "loss": 0.3593,
       "step": 129150
     },
     {
-      "epoch": 4.54,
-      "learning_rate": 1.090173195316238e-06,
-      "loss": 0.2385,
+      "epoch": 4.654737449093596,
+      "grad_norm": 0.2785971462726593,
+      "learning_rate": 6.284544627359057e-07,
+      "loss": 0.365,
       "step": 129155
     },
     {
-      "epoch": 4.54,
-      "learning_rate": 1.0893412883914378e-06,
-      "loss": 0.2366,
+      "epoch": 4.654917648754821,
+      "grad_norm": 0.28096702694892883,
+      "learning_rate": 6.278044385370774e-07,
+      "loss": 0.3482,
       "step": 129160
     },
     {
-      "epoch": 4.54,
-      "learning_rate": 1.0885096919307835e-06,
-      "loss": 0.2441,
+      "epoch": 4.655097848416045,
+      "grad_norm": 0.2804014980792999,
+      "learning_rate": 6.271547464046928e-07,
+      "loss": 0.3895,
       "step": 129165
     },
     {
-      "epoch": 4.54,
-      "learning_rate": 1.0876784059450722e-06,
-      "loss": 0.2596,
+      "epoch": 4.655278048077269,
+      "grad_norm": 0.26343923807144165,
+      "learning_rate": 6.265053863476089e-07,
+      "loss": 0.4001,
       "step": 129170
     },
     {
-      "epoch": 4.54,
-      "learning_rate": 1.086847430445101e-06,
-      "loss": 0.25,
+      "epoch": 4.655458247738494,
+      "grad_norm": 0.23427803814411163,
+      "learning_rate": 6.258563583746713e-07,
+      "loss": 0.3745,
       "step": 129175
     },
     {
-      "epoch": 4.54,
-      "learning_rate": 1.0860167654416525e-06,
-      "loss": 0.2292,
+      "epoch": 4.655638447399719,
+      "grad_norm": 0.24222131073474884,
+      "learning_rate": 6.252076624947201e-07,
+      "loss": 0.3622,
       "step": 129180
     },
     {
-      "epoch": 4.55,
-      "learning_rate": 1.0851864109455184e-06,
-      "loss": 0.2542,
+      "epoch": 4.6558186470609435,
+      "grad_norm": 0.25545430183410645,
+      "learning_rate": 6.245592987165955e-07,
+      "loss": 0.388,
       "step": 129185
     },
     {
-      "epoch": 4.55,
-      "learning_rate": 1.084356366967479e-06,
-      "loss": 0.2411,
+      "epoch": 4.655998846722168,
+      "grad_norm": 0.30208820104599,
+      "learning_rate": 6.239112670491293e-07,
+      "loss": 0.4266,
       "step": 129190
     },
     {
-      "epoch": 4.55,
-      "learning_rate": 1.0835266335183113e-06,
-      "loss": 0.257,
+      "epoch": 4.656179046383393,
+      "grad_norm": 0.26093167066574097,
+      "learning_rate": 6.232635675011562e-07,
+      "loss": 0.3532,
       "step": 129195
     },
     {
-      "epoch": 4.55,
-      "learning_rate": 1.082697210608785e-06,
-      "loss": 0.2618,
+      "epoch": 4.656359246044618,
+      "grad_norm": 0.3261514902114868,
+      "learning_rate": 6.226162000814967e-07,
+      "loss": 0.3773,
       "step": 129200
     },
     {
-      "epoch": 4.55,
-      "learning_rate": 1.0818680982496749e-06,
-      "loss": 0.2573,
+      "epoch": 4.656539445705842,
+      "grad_norm": 0.22967924177646637,
+      "learning_rate": 6.219691647989689e-07,
+      "loss": 0.3761,
       "step": 129205
     },
     {
-      "epoch": 4.55,
-      "learning_rate": 1.081039296451744e-06,
-      "loss": 0.2352,
+      "epoch": 4.656719645367067,
+      "grad_norm": 0.23772135376930237,
+      "learning_rate": 6.213224616623964e-07,
+      "loss": 0.3674,
       "step": 129210
     },
     {
-      "epoch": 4.55,
-      "learning_rate": 1.0802108052257592e-06,
-      "loss": 0.2664,
+      "epoch": 4.656899845028291,
+      "grad_norm": 0.24498851597309113,
+      "learning_rate": 6.206760906805803e-07,
+      "loss": 0.3609,
       "step": 129215
     },
     {
-      "epoch": 4.55,
-      "learning_rate": 1.0793826245824674e-06,
-      "loss": 0.2425,
+      "epoch": 4.657080044689516,
+      "grad_norm": 0.22295355796813965,
+      "learning_rate": 6.200300518623387e-07,
+      "loss": 0.3527,
       "step": 129220
     },
     {
-      "epoch": 4.55,
-      "learning_rate": 1.0785547545326318e-06,
-      "loss": 0.2493,
+      "epoch": 4.6572602443507405,
+      "grad_norm": 0.2808474600315094,
+      "learning_rate": 6.193843452164616e-07,
+      "loss": 0.3905,
       "step": 129225
     },
     {
-      "epoch": 4.55,
-      "learning_rate": 1.0777271950869972e-06,
-      "loss": 0.245,
+      "epoch": 4.657440444011965,
+      "grad_norm": 0.2916402816772461,
+      "learning_rate": 6.187389707517532e-07,
+      "loss": 0.385,
       "step": 129230
     },
     {
-      "epoch": 4.55,
-      "learning_rate": 1.0768999462563128e-06,
-      "loss": 0.2364,
+      "epoch": 4.65762064367319,
+      "grad_norm": 0.26787781715393066,
+      "learning_rate": 6.180939284770093e-07,
+      "loss": 0.3638,
       "step": 129235
     },
     {
-      "epoch": 4.55,
-      "learning_rate": 1.0760730080513093e-06,
-      "loss": 0.2364,
+      "epoch": 4.657800843334415,
+      "grad_norm": 0.29607486724853516,
+      "learning_rate": 6.174492184010117e-07,
+      "loss": 0.3698,
       "step": 129240
     },
     {
-      "epoch": 4.55,
-      "learning_rate": 1.0752463804827361e-06,
-      "loss": 0.2683,
+      "epoch": 4.657981042995639,
+      "grad_norm": 0.2748851478099823,
+      "learning_rate": 6.168048405325505e-07,
+      "loss": 0.3977,
       "step": 129245
     },
     {
-      "epoch": 4.55,
-      "learning_rate": 1.074420063561324e-06,
-      "loss": 0.2563,
+      "epoch": 4.658161242656864,
+      "grad_norm": 0.2305481731891632,
+      "learning_rate": 6.161607948804021e-07,
+      "loss": 0.3268,
       "step": 129250
     },
     {
-      "epoch": 4.55,
-      "learning_rate": 1.073594057297797e-06,
-      "loss": 0.2401,
+      "epoch": 4.658341442318088,
+      "grad_norm": 0.20638981461524963,
+      "learning_rate": 6.155170814533401e-07,
+      "loss": 0.3884,
       "step": 129255
     },
     {
-      "epoch": 4.55,
-      "learning_rate": 1.0727683617028805e-06,
-      "loss": 0.2371,
+      "epoch": 4.658521641979313,
+      "grad_norm": 0.25829213857650757,
+      "learning_rate": 6.148737002601379e-07,
+      "loss": 0.4005,
       "step": 129260
     },
     {
-      "epoch": 4.55,
-      "learning_rate": 1.071942976787299e-06,
-      "loss": 0.2316,
+      "epoch": 4.6587018416405375,
+      "grad_norm": 0.28095561265945435,
+      "learning_rate": 6.142306513095608e-07,
+      "loss": 0.373,
       "step": 129265
     },
     {
-      "epoch": 4.55,
-      "learning_rate": 1.0711179025617719e-06,
-      "loss": 0.249,
+      "epoch": 4.658882041301762,
+      "grad_norm": 0.22143559157848358,
+      "learning_rate": 6.135879346103712e-07,
+      "loss": 0.359,
       "step": 129270
     },
     {
-      "epoch": 4.55,
-      "learning_rate": 1.070293139037004e-06,
-      "loss": 0.2508,
+      "epoch": 4.659062240962987,
+      "grad_norm": 0.24161243438720703,
+      "learning_rate": 6.129455501713233e-07,
+      "loss": 0.3702,
       "step": 129275
     },
     {
-      "epoch": 4.55,
-      "learning_rate": 1.0694686862237152e-06,
-      "loss": 0.2644,
+      "epoch": 4.659242440624212,
+      "grad_norm": 0.20303134620189667,
+      "learning_rate": 6.123034980011683e-07,
+      "loss": 0.3933,
       "step": 129280
     },
     {
-      "epoch": 4.55,
-      "learning_rate": 1.068644544132602e-06,
-      "loss": 0.2427,
+      "epoch": 4.659422640285436,
+      "grad_norm": 0.29990285634994507,
+      "learning_rate": 6.116617781086603e-07,
+      "loss": 0.387,
       "step": 129285
     },
     {
-      "epoch": 4.55,
-      "learning_rate": 1.06782071277437e-06,
-      "loss": 0.2476,
+      "epoch": 4.659602839946661,
+      "grad_norm": 0.24720498919487,
+      "learning_rate": 6.110203905025397e-07,
+      "loss": 0.3566,
       "step": 129290
     },
     {
-      "epoch": 4.55,
-      "learning_rate": 1.0669971921597105e-06,
-      "loss": 0.263,
+      "epoch": 4.659783039607886,
+      "grad_norm": 0.2561245560646057,
+      "learning_rate": 6.10379335191541e-07,
+      "loss": 0.3665,
       "step": 129295
     },
     {
-      "epoch": 4.55,
-      "learning_rate": 1.0661739822993234e-06,
-      "loss": 0.2553,
+      "epoch": 4.659963239269111,
+      "grad_norm": 0.24636390805244446,
+      "learning_rate": 6.097386121844045e-07,
+      "loss": 0.3752,
       "step": 129300
     },
     {
-      "epoch": 4.55,
-      "learning_rate": 1.0653510832038943e-06,
-      "loss": 0.2611,
+      "epoch": 4.6601434389303344,
+      "grad_norm": 0.22616252303123474,
+      "learning_rate": 6.090982214898567e-07,
+      "loss": 0.3541,
       "step": 129305
     },
     {
-      "epoch": 4.55,
-      "learning_rate": 1.0645284948841067e-06,
-      "loss": 0.2628,
+      "epoch": 4.660323638591559,
+      "grad_norm": 0.2326437532901764,
+      "learning_rate": 6.084581631166208e-07,
+      "loss": 0.4103,
       "step": 129310
     },
     {
-      "epoch": 4.55,
-      "learning_rate": 1.0637062173506406e-06,
-      "loss": 0.2506,
+      "epoch": 4.660503838252784,
+      "grad_norm": 0.26222485303878784,
+      "learning_rate": 6.078184370734236e-07,
+      "loss": 0.3368,
       "step": 129315
     },
     {
-      "epoch": 4.55,
-      "learning_rate": 1.062884250614174e-06,
-      "loss": 0.2878,
+      "epoch": 4.660684037914009,
+      "grad_norm": 0.20036746561527252,
+      "learning_rate": 6.071790433689744e-07,
+      "loss": 0.3811,
       "step": 129320
     },
     {
-      "epoch": 4.55,
-      "learning_rate": 1.0620625946853841e-06,
-      "loss": 0.2495,
+      "epoch": 4.660864237575233,
+      "grad_norm": 0.3011195659637451,
+      "learning_rate": 6.065399820119888e-07,
+      "loss": 0.3701,
       "step": 129325
     },
     {
-      "epoch": 4.55,
-      "learning_rate": 1.0612412495749348e-06,
-      "loss": 0.2503,
+      "epoch": 4.661044437236458,
+      "grad_norm": 0.3017103672027588,
+      "learning_rate": 6.059012530111763e-07,
+      "loss": 0.3723,
       "step": 129330
     },
     {
-      "epoch": 4.55,
-      "learning_rate": 1.0604202152934895e-06,
-      "loss": 0.2626,
+      "epoch": 4.661224636897683,
+      "grad_norm": 0.2928502857685089,
+      "learning_rate": 6.052628563752355e-07,
+      "loss": 0.4019,
       "step": 129335
     },
     {
-      "epoch": 4.55,
-      "learning_rate": 1.0595994918517121e-06,
-      "loss": 0.237,
+      "epoch": 4.661404836558908,
+      "grad_norm": 0.26223188638687134,
+      "learning_rate": 6.046247921128623e-07,
+      "loss": 0.376,
       "step": 129340
     },
     {
-      "epoch": 4.55,
-      "learning_rate": 1.0587790792602608e-06,
-      "loss": 0.2439,
+      "epoch": 4.661585036220131,
+      "grad_norm": 0.25763529539108276,
+      "learning_rate": 6.039870602327552e-07,
+      "loss": 0.383,
       "step": 129345
     },
     {
-      "epoch": 4.55,
-      "learning_rate": 1.0579589775297827e-06,
-      "loss": 0.2621,
+      "epoch": 4.661765235881356,
+      "grad_norm": 0.3097740709781647,
+      "learning_rate": 6.03349660743599e-07,
+      "loss": 0.3831,
       "step": 129350
     },
     {
-      "epoch": 4.55,
-      "learning_rate": 1.0571391866709275e-06,
-      "loss": 0.2516,
+      "epoch": 4.661945435542581,
+      "grad_norm": 0.21880759298801422,
+      "learning_rate": 6.027125936540839e-07,
+      "loss": 0.325,
       "step": 129355
     },
     {
-      "epoch": 4.55,
-      "learning_rate": 1.0563197066943447e-06,
-      "loss": 0.2473,
+      "epoch": 4.662125635203806,
+      "grad_norm": 0.27281638979911804,
+      "learning_rate": 6.020758589728864e-07,
+      "loss": 0.3572,
       "step": 129360
     },
     {
-      "epoch": 4.55,
-      "learning_rate": 1.0555005376106708e-06,
-      "loss": 0.2242,
+      "epoch": 4.66230583486503,
+      "grad_norm": 0.26208505034446716,
+      "learning_rate": 6.014394567086801e-07,
+      "loss": 0.3569,
       "step": 129365
     },
     {
-      "epoch": 4.55,
-      "learning_rate": 1.0546816794305358e-06,
-      "loss": 0.263,
+      "epoch": 4.662486034526255,
+      "grad_norm": 0.2751930058002472,
+      "learning_rate": 6.008033868701357e-07,
+      "loss": 0.3527,
       "step": 129370
     },
     {
-      "epoch": 4.55,
-      "learning_rate": 1.0538631321645815e-06,
-      "loss": 0.2499,
+      "epoch": 4.66266623418748,
+      "grad_norm": 0.2358606606721878,
+      "learning_rate": 6.001676494659214e-07,
+      "loss": 0.3303,
       "step": 129375
     },
     {
-      "epoch": 4.55,
-      "learning_rate": 1.0530448958234378e-06,
-      "loss": 0.2622,
+      "epoch": 4.662846433848705,
+      "grad_norm": 0.22096312046051025,
+      "learning_rate": 5.995322445047025e-07,
+      "loss": 0.3917,
       "step": 129380
     },
     {
-      "epoch": 4.55,
-      "learning_rate": 1.0522269704177218e-06,
-      "loss": 0.2467,
+      "epoch": 4.663026633509929,
+      "grad_norm": 0.2579168677330017,
+      "learning_rate": 5.988971719951331e-07,
+      "loss": 0.41,
       "step": 129385
     },
     {
-      "epoch": 4.55,
-      "learning_rate": 1.0514093559580552e-06,
-      "loss": 0.2427,
+      "epoch": 4.663206833171154,
+      "grad_norm": 0.29768040776252747,
+      "learning_rate": 5.982624319458618e-07,
+      "loss": 0.3481,
       "step": 129390
     },
     {
-      "epoch": 4.55,
-      "learning_rate": 1.0505920524550571e-06,
-      "loss": 0.2242,
+      "epoch": 4.663387032832379,
+      "grad_norm": 0.24240122735500336,
+      "learning_rate": 5.976280243655403e-07,
+      "loss": 0.3567,
       "step": 129395
     },
     {
-      "epoch": 4.55,
-      "learning_rate": 1.049775059919339e-06,
-      "loss": 0.251,
+      "epoch": 4.663567232493603,
+      "grad_norm": 0.3015746772289276,
+      "learning_rate": 5.96993949262814e-07,
+      "loss": 0.3612,
       "step": 129400
     },
     {
-      "epoch": 4.55,
-      "learning_rate": 1.0489583783615086e-06,
-      "loss": 0.2533,
+      "epoch": 4.663747432154827,
+      "grad_norm": 0.2500225901603699,
+      "learning_rate": 5.963602066463236e-07,
+      "loss": 0.3912,
       "step": 129405
     },
     {
-      "epoch": 4.55,
-      "learning_rate": 1.0481420077921634e-06,
-      "loss": 0.2605,
+      "epoch": 4.663927631816052,
+      "grad_norm": 0.25465816259384155,
+      "learning_rate": 5.957267965246982e-07,
+      "loss": 0.3754,
       "step": 129410
     },
     {
-      "epoch": 4.55,
-      "learning_rate": 1.047325948221914e-06,
-      "loss": 0.206,
+      "epoch": 4.664107831477277,
+      "grad_norm": 0.24899233877658844,
+      "learning_rate": 5.95093718906567e-07,
+      "loss": 0.3582,
       "step": 129415
     },
     {
-      "epoch": 4.55,
-      "learning_rate": 1.0465101996613518e-06,
-      "loss": 0.2622,
+      "epoch": 4.6642880311385015,
+      "grad_norm": 0.22887267172336578,
+      "learning_rate": 5.944609738005618e-07,
+      "loss": 0.3409,
       "step": 129420
     },
     {
-      "epoch": 4.55,
-      "learning_rate": 1.045694762121066e-06,
-      "loss": 0.2663,
+      "epoch": 4.664468230799726,
+      "grad_norm": 0.26909008622169495,
+      "learning_rate": 5.93828561215301e-07,
+      "loss": 0.3867,
       "step": 129425
     },
     {
-      "epoch": 4.55,
-      "learning_rate": 1.0448796356116508e-06,
-      "loss": 0.2672,
+      "epoch": 4.664648430460951,
+      "grad_norm": 0.2556828558444977,
+      "learning_rate": 5.931964811594026e-07,
+      "loss": 0.356,
       "step": 129430
     },
     {
-      "epoch": 4.55,
-      "learning_rate": 1.0440648201436836e-06,
-      "loss": 0.2433,
+      "epoch": 4.664828630122176,
+      "grad_norm": 0.25458240509033203,
+      "learning_rate": 5.925647336414735e-07,
+      "loss": 0.4164,
       "step": 129435
     },
     {
-      "epoch": 4.55,
-      "learning_rate": 1.043250315727748e-06,
-      "loss": 0.2424,
+      "epoch": 4.6650088297834,
+      "grad_norm": 0.26027604937553406,
+      "learning_rate": 5.919333186701265e-07,
+      "loss": 0.3915,
       "step": 129440
     },
     {
-      "epoch": 4.55,
-      "learning_rate": 1.042436122374421e-06,
-      "loss": 0.2538,
+      "epoch": 4.665189029444624,
+      "grad_norm": 0.23065000772476196,
+      "learning_rate": 5.9130223625396e-07,
+      "loss": 0.3516,
       "step": 129445
     },
     {
-      "epoch": 4.55,
-      "learning_rate": 1.0416222400942698e-06,
-      "loss": 0.2515,
+      "epoch": 4.665369229105849,
+      "grad_norm": 0.2661188542842865,
+      "learning_rate": 5.906714864015783e-07,
+      "loss": 0.3747,
       "step": 129450
     },
     {
-      "epoch": 4.55,
-      "learning_rate": 1.0408086688978663e-06,
-      "loss": 0.2418,
+      "epoch": 4.665549428767074,
+      "grad_norm": 0.25803518295288086,
+      "learning_rate": 5.900410691215719e-07,
+      "loss": 0.3264,
       "step": 129455
     },
     {
-      "epoch": 4.55,
-      "learning_rate": 1.0399954087957714e-06,
-      "loss": 0.2414,
+      "epoch": 4.6657296284282985,
+      "grad_norm": 0.24731706082820892,
+      "learning_rate": 5.894109844225281e-07,
+      "loss": 0.3546,
       "step": 129460
     },
     {
-      "epoch": 4.55,
-      "learning_rate": 1.039182459798549e-06,
-      "loss": 0.2363,
+      "epoch": 4.665909828089523,
+      "grad_norm": 0.21868941187858582,
+      "learning_rate": 5.88781232313032e-07,
+      "loss": 0.3743,
       "step": 129465
     },
     {
-      "epoch": 4.56,
-      "learning_rate": 1.0383698219167459e-06,
-      "loss": 0.2648,
+      "epoch": 4.666090027750748,
+      "grad_norm": 0.2019900679588318,
+      "learning_rate": 5.881518128016655e-07,
+      "loss": 0.3616,
       "step": 129470
     },
     {
-      "epoch": 4.56,
-      "learning_rate": 1.0375574951609234e-06,
-      "loss": 0.2565,
+      "epoch": 4.666270227411973,
+      "grad_norm": 0.37569132447242737,
+      "learning_rate": 5.875227258970078e-07,
+      "loss": 0.3833,
       "step": 129475
     },
     {
-      "epoch": 4.56,
-      "learning_rate": 1.0367454795416205e-06,
-      "loss": 0.2511,
+      "epoch": 4.6664504270731975,
+      "grad_norm": 0.2718876600265503,
+      "learning_rate": 5.868939716076244e-07,
+      "loss": 0.3947,
       "step": 129480
     },
     {
-      "epoch": 4.56,
-      "learning_rate": 1.0359337750693893e-06,
-      "loss": 0.242,
+      "epoch": 4.666630626734422,
+      "grad_norm": 0.23850391805171967,
+      "learning_rate": 5.86265549942086e-07,
+      "loss": 0.3544,
       "step": 129485
     },
     {
-      "epoch": 4.56,
-      "learning_rate": 1.0351223817547633e-06,
-      "loss": 0.2405,
+      "epoch": 4.666810826395646,
+      "grad_norm": 0.28313085436820984,
+      "learning_rate": 5.856374609089526e-07,
+      "loss": 0.3506,
       "step": 129490
     },
     {
-      "epoch": 4.56,
-      "learning_rate": 1.0343112996082815e-06,
-      "loss": 0.2332,
+      "epoch": 4.666991026056871,
+      "grad_norm": 0.21799631416797638,
+      "learning_rate": 5.850097045167785e-07,
+      "loss": 0.3657,
       "step": 129495
     },
     {
-      "epoch": 4.56,
-      "learning_rate": 1.0335005286404742e-06,
-      "loss": 0.2607,
+      "epoch": 4.6671712257180955,
+      "grad_norm": 0.25782206654548645,
+      "learning_rate": 5.843822807741261e-07,
+      "loss": 0.3886,
       "step": 129500
     },
     {
-      "epoch": 4.56,
-      "eval_loss": 0.24865983426570892,
-      "eval_runtime": 10.5527,
-      "eval_samples_per_second": 9.476,
-      "eval_steps_per_second": 9.476,
+      "epoch": 4.6671712257180955,
+      "eval_loss": 0.4288308322429657,
+      "eval_runtime": 3.5293,
+      "eval_samples_per_second": 28.335,
+      "eval_steps_per_second": 7.084,
       "step": 129500
     },
     {
-      "epoch": 4.56,
-      "learning_rate": 1.0326900688618635e-06,
-      "loss": 0.256,
+      "epoch": 4.66735142537932,
+      "grad_norm": 0.2763291001319885,
+      "learning_rate": 5.837551896895305e-07,
+      "loss": 0.373,
       "step": 129505
     },
     {
-      "epoch": 4.56,
-      "learning_rate": 1.0318799202829827e-06,
-      "loss": 0.2483,
+      "epoch": 4.667531625040545,
+      "grad_norm": 0.3070751130580902,
+      "learning_rate": 5.831284312715485e-07,
+      "loss": 0.3634,
       "step": 129510
     },
     {
-      "epoch": 4.56,
-      "learning_rate": 1.0310700829143427e-06,
-      "loss": 0.238,
+      "epoch": 4.66771182470177,
+      "grad_norm": 0.3483225703239441,
+      "learning_rate": 5.82502005528715e-07,
+      "loss": 0.3704,
       "step": 129515
     },
     {
-      "epoch": 4.56,
-      "learning_rate": 1.0302605567664631e-06,
-      "loss": 0.2675,
+      "epoch": 4.667892024362994,
+      "grad_norm": 0.20731894671916962,
+      "learning_rate": 5.818759124695622e-07,
+      "loss": 0.3615,
       "step": 129520
     },
     {
-      "epoch": 4.56,
-      "learning_rate": 1.029451341849852e-06,
-      "loss": 0.2344,
+      "epoch": 4.668072224024219,
+      "grad_norm": 0.25448137521743774,
+      "learning_rate": 5.81250152102622e-07,
+      "loss": 0.3756,
       "step": 129525
     },
     {
-      "epoch": 4.56,
-      "learning_rate": 1.0286424381750175e-06,
-      "loss": 0.267,
+      "epoch": 4.668252423685443,
+      "grad_norm": 0.3217816948890686,
+      "learning_rate": 5.806247244364238e-07,
+      "loss": 0.4025,
       "step": 129530
     },
     {
-      "epoch": 4.56,
-      "learning_rate": 1.0278338457524655e-06,
-      "loss": 0.2232,
+      "epoch": 4.668432623346668,
+      "grad_norm": 0.21159349381923676,
+      "learning_rate": 5.799996294794801e-07,
+      "loss": 0.382,
       "step": 129535
     },
     {
-      "epoch": 4.56,
-      "learning_rate": 1.0270255645926902e-06,
-      "loss": 0.2512,
+      "epoch": 4.6686128230078925,
+      "grad_norm": 0.2236149162054062,
+      "learning_rate": 5.793748672403204e-07,
+      "loss": 0.3662,
       "step": 129540
     },
     {
-      "epoch": 4.56,
-      "learning_rate": 1.0262175947061886e-06,
-      "loss": 0.276,
+      "epoch": 4.668793022669117,
+      "grad_norm": 0.262952595949173,
+      "learning_rate": 5.78750437727446e-07,
+      "loss": 0.3876,
       "step": 129545
     },
     {
-      "epoch": 4.56,
-      "learning_rate": 1.0254099361034553e-06,
-      "loss": 0.2621,
+      "epoch": 4.668973222330342,
+      "grad_norm": 0.27944156527519226,
+      "learning_rate": 5.781263409493698e-07,
+      "loss": 0.3889,
       "step": 129550
     },
     {
-      "epoch": 4.56,
-      "learning_rate": 1.0246025887949734e-06,
-      "loss": 0.2586,
+      "epoch": 4.669153421991567,
+      "grad_norm": 0.24670124053955078,
+      "learning_rate": 5.775025769145959e-07,
+      "loss": 0.3498,
       "step": 129555
     },
     {
-      "epoch": 4.56,
-      "learning_rate": 1.0237955527912264e-06,
-      "loss": 0.2416,
+      "epoch": 4.669333621652791,
+      "grad_norm": 0.2797219455242157,
+      "learning_rate": 5.768791456316175e-07,
+      "loss": 0.3675,
       "step": 129560
     },
     {
-      "epoch": 4.56,
-      "learning_rate": 1.0229888281026916e-06,
-      "loss": 0.2527,
+      "epoch": 4.669513821314016,
+      "grad_norm": 0.19079472124576569,
+      "learning_rate": 5.762560471089334e-07,
+      "loss": 0.3674,
       "step": 129565
     },
     {
-      "epoch": 4.56,
-      "learning_rate": 1.0221824147398474e-06,
-      "loss": 0.2497,
+      "epoch": 4.669694020975241,
+      "grad_norm": 0.20952509343624115,
+      "learning_rate": 5.756332813550369e-07,
+      "loss": 0.3384,
       "step": 129570
     },
     {
-      "epoch": 4.56,
-      "learning_rate": 1.0213763127131625e-06,
-      "loss": 0.2474,
+      "epoch": 4.669874220636466,
+      "grad_norm": 0.2480088174343109,
+      "learning_rate": 5.750108483784017e-07,
+      "loss": 0.3942,
       "step": 129575
     },
     {
-      "epoch": 4.56,
-      "learning_rate": 1.0205705220331042e-06,
-      "loss": 0.2455,
+      "epoch": 4.6700544202976895,
+      "grad_norm": 0.19479909539222717,
+      "learning_rate": 5.743887481875182e-07,
+      "loss": 0.3673,
       "step": 129580
     },
     {
-      "epoch": 4.56,
-      "learning_rate": 1.0197650427101302e-06,
-      "loss": 0.2321,
+      "epoch": 4.670234619958914,
+      "grad_norm": 0.27210843563079834,
+      "learning_rate": 5.737669807908546e-07,
+      "loss": 0.374,
       "step": 129585
     },
     {
-      "epoch": 4.56,
-      "learning_rate": 1.0189598747547074e-06,
-      "loss": 0.2573,
+      "epoch": 4.670414819620139,
+      "grad_norm": 0.25238466262817383,
+      "learning_rate": 5.731455461968932e-07,
+      "loss": 0.3692,
       "step": 129590
     },
     {
-      "epoch": 4.56,
-      "learning_rate": 1.0181550181772858e-06,
-      "loss": 0.2412,
+      "epoch": 4.670595019281364,
+      "grad_norm": 0.3293459117412567,
+      "learning_rate": 5.725244444140937e-07,
+      "loss": 0.365,
       "step": 129595
     },
     {
-      "epoch": 4.56,
-      "learning_rate": 1.0173504729883182e-06,
-      "loss": 0.2519,
+      "epoch": 4.670775218942588,
+      "grad_norm": 0.25170597434043884,
+      "learning_rate": 5.719036754509161e-07,
+      "loss": 0.4202,
       "step": 129600
     },
     {
-      "epoch": 4.56,
-      "learning_rate": 1.0165462391982432e-06,
-      "loss": 0.2473,
+      "epoch": 4.670955418603813,
+      "grad_norm": 0.22425323724746704,
+      "learning_rate": 5.712832393158229e-07,
+      "loss": 0.3636,
       "step": 129605
     },
     {
-      "epoch": 4.56,
-      "learning_rate": 1.0157423168175167e-06,
-      "loss": 0.2461,
+      "epoch": 4.671135618265038,
+      "grad_norm": 0.3065611720085144,
+      "learning_rate": 5.706631360172659e-07,
+      "loss": 0.3638,
       "step": 129610
     },
     {
-      "epoch": 4.56,
-      "learning_rate": 1.014938705856569e-06,
-      "loss": 0.2548,
+      "epoch": 4.671315817926263,
+      "grad_norm": 0.2282622754573822,
+      "learning_rate": 5.700433655636939e-07,
+      "loss": 0.3536,
       "step": 129615
     },
     {
-      "epoch": 4.56,
-      "learning_rate": 1.0141354063258335e-06,
-      "loss": 0.2538,
+      "epoch": 4.6714960175874864,
+      "grad_norm": 0.22002369165420532,
+      "learning_rate": 5.694239279635527e-07,
+      "loss": 0.3777,
       "step": 129620
     },
     {
-      "epoch": 4.56,
-      "learning_rate": 1.0133324182357434e-06,
-      "loss": 0.2641,
+      "epoch": 4.671676217248711,
+      "grad_norm": 0.263820081949234,
+      "learning_rate": 5.688048232252746e-07,
+      "loss": 0.3732,
       "step": 129625
     },
     {
-      "epoch": 4.56,
-      "learning_rate": 1.012529741596724e-06,
-      "loss": 0.2556,
+      "epoch": 4.671856416909936,
+      "grad_norm": 0.24897272884845734,
+      "learning_rate": 5.681860513573084e-07,
+      "loss": 0.3497,
       "step": 129630
     },
     {
-      "epoch": 4.56,
-      "learning_rate": 1.0117273764191998e-06,
-      "loss": 0.2564,
+      "epoch": 4.672036616571161,
+      "grad_norm": 0.2410743087530136,
+      "learning_rate": 5.675676123680724e-07,
+      "loss": 0.3739,
       "step": 129635
     },
     {
-      "epoch": 4.56,
-      "learning_rate": 1.0109253227135824e-06,
-      "loss": 0.2577,
+      "epoch": 4.672216816232385,
+      "grad_norm": 0.28402799367904663,
+      "learning_rate": 5.669495062660013e-07,
+      "loss": 0.3734,
       "step": 129640
     },
     {
-      "epoch": 4.56,
-      "learning_rate": 1.010123580490291e-06,
-      "loss": 0.2703,
+      "epoch": 4.67239701589361,
+      "grad_norm": 0.22168543934822083,
+      "learning_rate": 5.663317330595108e-07,
+      "loss": 0.3817,
       "step": 129645
     },
     {
-      "epoch": 4.56,
-      "learning_rate": 1.0093221497597367e-06,
-      "loss": 0.2258,
+      "epoch": 4.672577215554835,
+      "grad_norm": 0.25576576590538025,
+      "learning_rate": 5.657142927570163e-07,
+      "loss": 0.4007,
       "step": 129650
     },
     {
-      "epoch": 4.56,
-      "learning_rate": 1.0085210305323223e-06,
-      "loss": 0.2672,
+      "epoch": 4.67275741521606,
+      "grad_norm": 0.21974904835224152,
+      "learning_rate": 5.65097185366939e-07,
+      "loss": 0.399,
       "step": 129655
     },
     {
-      "epoch": 4.56,
-      "learning_rate": 1.0077202228184507e-06,
-      "loss": 0.25,
+      "epoch": 4.672937614877284,
+      "grad_norm": 0.3177945911884308,
+      "learning_rate": 5.644804108976804e-07,
+      "loss": 0.3589,
       "step": 129660
     },
     {
-      "epoch": 4.56,
-      "learning_rate": 1.0069197266285245e-06,
-      "loss": 0.2373,
+      "epoch": 4.673117814538509,
+      "grad_norm": 0.3027752637863159,
+      "learning_rate": 5.638639693576447e-07,
+      "loss": 0.3754,
       "step": 129665
     },
     {
-      "epoch": 4.56,
-      "learning_rate": 1.0061195419729302e-06,
-      "loss": 0.2499,
+      "epoch": 4.673298014199734,
+      "grad_norm": 0.3068021237850189,
+      "learning_rate": 5.63247860755231e-07,
+      "loss": 0.3751,
       "step": 129670
     },
     {
-      "epoch": 4.56,
-      "learning_rate": 1.0053196688620647e-06,
-      "loss": 0.2464,
+      "epoch": 4.673478213860958,
+      "grad_norm": 0.2814277410507202,
+      "learning_rate": 5.626320850988353e-07,
+      "loss": 0.3641,
       "step": 129675
     },
     {
-      "epoch": 4.56,
-      "learning_rate": 1.004520107306306e-06,
-      "loss": 0.2409,
+      "epoch": 4.673658413522182,
+      "grad_norm": 0.2657163441181183,
+      "learning_rate": 5.620166423968454e-07,
+      "loss": 0.3845,
       "step": 129680
     },
     {
-      "epoch": 4.56,
-      "learning_rate": 1.0037208573160428e-06,
-      "loss": 0.2213,
+      "epoch": 4.673838613183407,
+      "grad_norm": 0.21708977222442627,
+      "learning_rate": 5.614015326576489e-07,
+      "loss": 0.3781,
       "step": 129685
     },
     {
-      "epoch": 4.56,
-      "learning_rate": 1.0029219189016503e-06,
-      "loss": 0.2502,
+      "epoch": 4.674018812844632,
+      "grad_norm": 0.2891923189163208,
+      "learning_rate": 5.607867558896224e-07,
+      "loss": 0.3721,
       "step": 129690
     },
     {
-      "epoch": 4.56,
-      "learning_rate": 1.0021232920735008e-06,
-      "loss": 0.2466,
+      "epoch": 4.674199012505857,
+      "grad_norm": 0.25727081298828125,
+      "learning_rate": 5.601723121011482e-07,
+      "loss": 0.32,
       "step": 129695
     },
     {
-      "epoch": 4.56,
-      "learning_rate": 1.0013249768419635e-06,
-      "loss": 0.2498,
+      "epoch": 4.674379212167081,
+      "grad_norm": 0.18455742299556732,
+      "learning_rate": 5.595582013005918e-07,
+      "loss": 0.3469,
       "step": 129700
     },
     {
-      "epoch": 4.56,
-      "learning_rate": 1.0005269732174055e-06,
-      "loss": 0.2472,
+      "epoch": 4.674559411828306,
+      "grad_norm": 0.27723222970962524,
+      "learning_rate": 5.589444234963214e-07,
+      "loss": 0.3841,
       "step": 129705
     },
     {
-      "epoch": 4.56,
-      "learning_rate": 9.997292812101905e-07,
-      "loss": 0.2651,
+      "epoch": 4.674739611489531,
+      "grad_norm": 0.2724769115447998,
+      "learning_rate": 5.583309786967084e-07,
+      "loss": 0.3844,
       "step": 129710
     },
     {
-      "epoch": 4.56,
-      "learning_rate": 9.98931900830677e-07,
-      "loss": 0.235,
+      "epoch": 4.674919811150755,
+      "grad_norm": 0.25393152236938477,
+      "learning_rate": 5.577178669100986e-07,
+      "loss": 0.4236,
       "step": 129715
     },
     {
-      "epoch": 4.56,
-      "learning_rate": 9.981348320892093e-07,
-      "loss": 0.2582,
+      "epoch": 4.675100010811979,
+      "grad_norm": 0.2612922489643097,
+      "learning_rate": 5.571050881448492e-07,
+      "loss": 0.3766,
       "step": 129720
     },
     {
-      "epoch": 4.56,
-      "learning_rate": 9.97338074996146e-07,
-      "loss": 0.2462,
+      "epoch": 4.675280210473204,
+      "grad_norm": 0.2447577714920044,
+      "learning_rate": 5.56492642409312e-07,
+      "loss": 0.3888,
       "step": 129725
     },
     {
-      "epoch": 4.56,
-      "learning_rate": 9.965416295618313e-07,
-      "loss": 0.2511,
+      "epoch": 4.675460410134429,
+      "grad_norm": 0.30114662647247314,
+      "learning_rate": 5.55880529711833e-07,
+      "loss": 0.3967,
       "step": 129730
     },
     {
-      "epoch": 4.56,
-      "learning_rate": 9.957454957966017e-07,
-      "loss": 0.2487,
+      "epoch": 4.6756406097956535,
+      "grad_norm": 0.24598009884357452,
+      "learning_rate": 5.552687500607473e-07,
+      "loss": 0.376,
       "step": 129735
     },
     {
-      "epoch": 4.56,
-      "learning_rate": 9.949496737107988e-07,
-      "loss": 0.2294,
+      "epoch": 4.675820809456878,
+      "grad_norm": 0.23333467543125153,
+      "learning_rate": 5.546573034643926e-07,
+      "loss": 0.373,
       "step": 129740
     },
     {
-      "epoch": 4.56,
-      "learning_rate": 9.941541633147562e-07,
-      "loss": 0.2585,
+      "epoch": 4.676001009118103,
+      "grad_norm": 0.3097272515296936,
+      "learning_rate": 5.540461899310956e-07,
+      "loss": 0.4054,
       "step": 129745
     },
     {
-      "epoch": 4.56,
-      "learning_rate": 9.933589646187985e-07,
-      "loss": 0.2421,
+      "epoch": 4.676181208779328,
+      "grad_norm": 0.2813061773777008,
+      "learning_rate": 5.534354094691912e-07,
+      "loss": 0.3875,
       "step": 129750
     },
     {
-      "epoch": 4.57,
-      "learning_rate": 9.92564077633254e-07,
-      "loss": 0.2785,
+      "epoch": 4.6763614084405525,
+      "grad_norm": 0.27265501022338867,
+      "learning_rate": 5.528249620869952e-07,
+      "loss": 0.3326,
       "step": 129755
     },
     {
-      "epoch": 4.57,
-      "learning_rate": 9.91769502368442e-07,
-      "loss": 0.2468,
+      "epoch": 4.676541608101777,
+      "grad_norm": 0.23373085260391235,
+      "learning_rate": 5.522148477928257e-07,
+      "loss": 0.355,
       "step": 129760
     },
     {
-      "epoch": 4.57,
-      "learning_rate": 9.90975238834682e-07,
-      "loss": 0.2277,
+      "epoch": 4.676721807763001,
+      "grad_norm": 0.2653125822544098,
+      "learning_rate": 5.516050665949956e-07,
+      "loss": 0.374,
       "step": 129765
     },
     {
-      "epoch": 4.57,
-      "learning_rate": 9.901812870422884e-07,
-      "loss": 0.2453,
+      "epoch": 4.676902007424226,
+      "grad_norm": 0.2680438160896301,
+      "learning_rate": 5.509956185018123e-07,
+      "loss": 0.3668,
       "step": 129770
     },
     {
-      "epoch": 4.57,
-      "learning_rate": 9.893876470015661e-07,
-      "loss": 0.2453,
+      "epoch": 4.6770822070854505,
+      "grad_norm": 0.24834401905536652,
+      "learning_rate": 5.503865035215799e-07,
+      "loss": 0.3672,
       "step": 129775
     },
     {
-      "epoch": 4.57,
-      "learning_rate": 9.885943187228214e-07,
-      "loss": 0.2807,
+      "epoch": 4.677262406746675,
+      "grad_norm": 0.2544165551662445,
+      "learning_rate": 5.497777216626004e-07,
+      "loss": 0.3523,
       "step": 129780
     },
     {
-      "epoch": 4.57,
-      "learning_rate": 9.878013022163567e-07,
-      "loss": 0.2355,
+      "epoch": 4.6774426064079,
+      "grad_norm": 0.2632615864276886,
+      "learning_rate": 5.491692729331643e-07,
+      "loss": 0.3629,
       "step": 129785
     },
     {
-      "epoch": 4.57,
-      "learning_rate": 9.870085974924698e-07,
-      "loss": 0.2383,
+      "epoch": 4.677622806069125,
+      "grad_norm": 0.27221032977104187,
+      "learning_rate": 5.485611573415622e-07,
+      "loss": 0.3558,
       "step": 129790
     },
     {
-      "epoch": 4.57,
-      "learning_rate": 9.862162045614436e-07,
-      "loss": 0.2452,
+      "epoch": 4.6778030057303495,
+      "grad_norm": 0.23004166781902313,
+      "learning_rate": 5.479533748960819e-07,
+      "loss": 0.3677,
       "step": 129795
     },
     {
-      "epoch": 4.57,
-      "learning_rate": 9.854241234335814e-07,
-      "loss": 0.265,
+      "epoch": 4.677983205391574,
+      "grad_norm": 0.23140105605125427,
+      "learning_rate": 5.473459256050029e-07,
+      "loss": 0.3682,
       "step": 129800
     },
     {
-      "epoch": 4.57,
-      "learning_rate": 9.846323541191582e-07,
-      "loss": 0.2254,
+      "epoch": 4.678163405052798,
+      "grad_norm": 0.25781282782554626,
+      "learning_rate": 5.46738809476599e-07,
+      "loss": 0.3737,
       "step": 129805
     },
     {
-      "epoch": 4.57,
-      "learning_rate": 9.838408966284545e-07,
-      "loss": 0.2428,
+      "epoch": 4.678343604714023,
+      "grad_norm": 0.28814247250556946,
+      "learning_rate": 5.461320265191445e-07,
+      "loss": 0.4094,
       "step": 129810
     },
     {
-      "epoch": 4.57,
-      "learning_rate": 9.830497509717511e-07,
-      "loss": 0.2269,
+      "epoch": 4.6785238043752475,
+      "grad_norm": 0.2671186327934265,
+      "learning_rate": 5.455255767409101e-07,
+      "loss": 0.3495,
       "step": 129815
     },
     {
-      "epoch": 4.57,
-      "learning_rate": 9.822589171593206e-07,
-      "loss": 0.2352,
+      "epoch": 4.678704004036472,
+      "grad_norm": 0.25495436787605286,
+      "learning_rate": 5.449194601501534e-07,
+      "loss": 0.4176,
       "step": 129820
     },
     {
-      "epoch": 4.57,
-      "learning_rate": 9.814683952014292e-07,
-      "loss": 0.2565,
+      "epoch": 4.678884203697697,
+      "grad_norm": 0.24299226701259613,
+      "learning_rate": 5.443136767551343e-07,
+      "loss": 0.3457,
       "step": 129825
     },
     {
-      "epoch": 4.57,
-      "learning_rate": 9.806781851083414e-07,
-      "loss": 0.2418,
+      "epoch": 4.679064403358922,
+      "grad_norm": 0.2552716135978699,
+      "learning_rate": 5.437082265641075e-07,
+      "loss": 0.4135,
       "step": 129830
     },
     {
-      "epoch": 4.57,
-      "learning_rate": 9.79888286890318e-07,
-      "loss": 0.2701,
+      "epoch": 4.679244603020146,
+      "grad_norm": 0.2803386151790619,
+      "learning_rate": 5.431031095853189e-07,
+      "loss": 0.3884,
       "step": 129835
     },
     {
-      "epoch": 4.57,
-      "learning_rate": 9.790987005576152e-07,
-      "loss": 0.2583,
+      "epoch": 4.679424802681371,
+      "grad_norm": 0.20010852813720703,
+      "learning_rate": 5.42498325827015e-07,
+      "loss": 0.3422,
       "step": 129840
     },
     {
-      "epoch": 4.57,
-      "learning_rate": 9.783094261204884e-07,
-      "loss": 0.2461,
+      "epoch": 4.679605002342596,
+      "grad_norm": 0.28240904211997986,
+      "learning_rate": 5.41893875297439e-07,
+      "loss": 0.3968,
       "step": 129845
     },
     {
-      "epoch": 4.57,
-      "learning_rate": 9.775204635891766e-07,
-      "loss": 0.2528,
+      "epoch": 4.679785202003821,
+      "grad_norm": 0.28269240260124207,
+      "learning_rate": 5.412897580048231e-07,
+      "loss": 0.3503,
       "step": 129850
     },
     {
-      "epoch": 4.57,
-      "learning_rate": 9.767318129739355e-07,
-      "loss": 0.2441,
+      "epoch": 4.6799654016650445,
+      "grad_norm": 0.295452743768692,
+      "learning_rate": 5.406859739574e-07,
+      "loss": 0.3895,
       "step": 129855
     },
     {
-      "epoch": 4.57,
-      "learning_rate": 9.759434742849988e-07,
-      "loss": 0.263,
+      "epoch": 4.680145601326269,
+      "grad_norm": 0.2563684284687042,
+      "learning_rate": 5.400825231633932e-07,
+      "loss": 0.38,
       "step": 129860
     },
     {
-      "epoch": 4.57,
-      "learning_rate": 9.751554475326025e-07,
-      "loss": 0.2548,
+      "epoch": 4.680325800987494,
+      "grad_norm": 0.22655951976776123,
+      "learning_rate": 5.394794056310243e-07,
+      "loss": 0.3844,
       "step": 129865
     },
     {
-      "epoch": 4.57,
-      "learning_rate": 9.743677327269802e-07,
-      "loss": 0.2489,
+      "epoch": 4.680506000648719,
+      "grad_norm": 0.2583646774291992,
+      "learning_rate": 5.38876621368517e-07,
+      "loss": 0.3703,
       "step": 129870
     },
     {
-      "epoch": 4.57,
-      "learning_rate": 9.735803298783569e-07,
-      "loss": 0.2698,
+      "epoch": 4.680686200309943,
+      "grad_norm": 0.24790343642234802,
+      "learning_rate": 5.382741703840788e-07,
+      "loss": 0.3648,
       "step": 129875
     },
     {
-      "epoch": 4.57,
-      "learning_rate": 9.727932389969636e-07,
-      "loss": 0.2638,
+      "epoch": 4.680866399971168,
+      "grad_norm": 0.2448999434709549,
+      "learning_rate": 5.376720526859197e-07,
+      "loss": 0.3749,
       "step": 129880
     },
     {
-      "epoch": 4.57,
-      "learning_rate": 9.72006460093014e-07,
-      "loss": 0.2491,
+      "epoch": 4.681046599632393,
+      "grad_norm": 0.2374943196773529,
+      "learning_rate": 5.370702682822415e-07,
+      "loss": 0.3917,
       "step": 129885
     },
     {
-      "epoch": 4.57,
-      "learning_rate": 9.712199931767223e-07,
-      "loss": 0.2499,
+      "epoch": 4.681226799293618,
+      "grad_norm": 0.23962445557117462,
+      "learning_rate": 5.364688171812432e-07,
+      "loss": 0.3672,
       "step": 129890
     },
     {
-      "epoch": 4.57,
-      "learning_rate": 9.704338382583084e-07,
-      "loss": 0.2712,
+      "epoch": 4.6814069989548415,
+      "grad_norm": 0.2775137424468994,
+      "learning_rate": 5.358676993911238e-07,
+      "loss": 0.391,
       "step": 129895
     },
     {
-      "epoch": 4.57,
-      "learning_rate": 9.69647995347972e-07,
-      "loss": 0.2672,
+      "epoch": 4.681587198616066,
+      "grad_norm": 0.23689554631710052,
+      "learning_rate": 5.352669149200711e-07,
+      "loss": 0.3584,
       "step": 129900
     },
     {
-      "epoch": 4.57,
-      "learning_rate": 9.688624644559218e-07,
-      "loss": 0.2716,
+      "epoch": 4.681767398277291,
+      "grad_norm": 0.22408610582351685,
+      "learning_rate": 5.34666463776265e-07,
+      "loss": 0.363,
       "step": 129905
     },
     {
-      "epoch": 4.57,
-      "learning_rate": 9.680772455923526e-07,
-      "loss": 0.2433,
+      "epoch": 4.681947597938516,
+      "grad_norm": 0.2599274218082428,
+      "learning_rate": 5.340663459678958e-07,
+      "loss": 0.3446,
       "step": 129910
     },
     {
-      "epoch": 4.57,
-      "learning_rate": 9.672923387674644e-07,
-      "loss": 0.2583,
+      "epoch": 4.68212779759974,
+      "grad_norm": 0.2526589632034302,
+      "learning_rate": 5.334665615031376e-07,
+      "loss": 0.3397,
       "step": 129915
     },
     {
-      "epoch": 4.57,
-      "learning_rate": 9.665077439914461e-07,
-      "loss": 0.2455,
+      "epoch": 4.682307997260965,
+      "grad_norm": 0.23241139948368073,
+      "learning_rate": 5.328671103901589e-07,
+      "loss": 0.3638,
       "step": 129920
     },
     {
-      "epoch": 4.57,
-      "learning_rate": 9.65723461274487e-07,
-      "loss": 0.2558,
+      "epoch": 4.68248819692219,
+      "grad_norm": 0.2700398862361908,
+      "learning_rate": 5.322679926371282e-07,
+      "loss": 0.3878,
       "step": 129925
     },
     {
-      "epoch": 4.57,
-      "learning_rate": 9.649394906267677e-07,
-      "loss": 0.25,
+      "epoch": 4.682668396583415,
+      "grad_norm": 0.24214357137680054,
+      "learning_rate": 5.316692082522057e-07,
+      "loss": 0.3821,
       "step": 129930
     },
     {
-      "epoch": 4.57,
-      "learning_rate": 9.64155832058472e-07,
-      "loss": 0.2606,
+      "epoch": 4.682848596244639,
+      "grad_norm": 0.1930561661720276,
+      "learning_rate": 5.310707572435569e-07,
+      "loss": 0.3813,
       "step": 129935
     },
     {
-      "epoch": 4.57,
-      "learning_rate": 9.63372485579775e-07,
-      "loss": 0.2532,
+      "epoch": 4.683028795905864,
+      "grad_norm": 0.28772681951522827,
+      "learning_rate": 5.30472639619331e-07,
+      "loss": 0.3628,
       "step": 129940
     },
     {
-      "epoch": 4.57,
-      "learning_rate": 9.625894512008432e-07,
-      "loss": 0.2402,
+      "epoch": 4.683208995567089,
+      "grad_norm": 0.2775450050830841,
+      "learning_rate": 5.298748553876797e-07,
+      "loss": 0.4035,
       "step": 129945
     },
     {
-      "epoch": 4.57,
-      "learning_rate": 9.618067289318412e-07,
-      "loss": 0.2522,
+      "epoch": 4.683389195228313,
+      "grad_norm": 0.3269951045513153,
+      "learning_rate": 5.292774045567439e-07,
+      "loss": 0.3668,
       "step": 129950
     },
     {
-      "epoch": 4.57,
-      "learning_rate": 9.610243187829438e-07,
-      "loss": 0.2268,
+      "epoch": 4.683569394889537,
+      "grad_norm": 0.27343136072158813,
+      "learning_rate": 5.286802871346641e-07,
+      "loss": 0.3986,
       "step": 129955
     },
     {
-      "epoch": 4.57,
-      "learning_rate": 9.602422207643013e-07,
-      "loss": 0.2497,
+      "epoch": 4.683749594550762,
+      "grad_norm": 0.2822220027446747,
+      "learning_rate": 5.280835031295811e-07,
+      "loss": 0.3583,
       "step": 129960
     },
     {
-      "epoch": 4.57,
-      "learning_rate": 9.59460434886067e-07,
-      "loss": 0.2595,
+      "epoch": 4.683929794211987,
+      "grad_norm": 0.26671725511550903,
+      "learning_rate": 5.274870525496245e-07,
+      "loss": 0.3761,
       "step": 129965
     },
     {
-      "epoch": 4.57,
-      "learning_rate": 9.58678961158399e-07,
-      "loss": 0.2403,
+      "epoch": 4.684109993873212,
+      "grad_norm": 0.2931014597415924,
+      "learning_rate": 5.268909354029127e-07,
+      "loss": 0.3933,
       "step": 129970
     },
     {
-      "epoch": 4.57,
-      "learning_rate": 9.578977995914424e-07,
-      "loss": 0.2467,
+      "epoch": 4.684290193534436,
+      "grad_norm": 0.218561589717865,
+      "learning_rate": 5.262951516975756e-07,
+      "loss": 0.4029,
       "step": 129975
     },
     {
-      "epoch": 4.57,
-      "learning_rate": 9.57116950195336e-07,
-      "loss": 0.2558,
+      "epoch": 4.684470393195661,
+      "grad_norm": 0.25005042552948,
+      "learning_rate": 5.256997014417315e-07,
+      "loss": 0.3756,
       "step": 129980
     },
     {
-      "epoch": 4.57,
-      "learning_rate": 9.563364129802187e-07,
-      "loss": 0.2688,
+      "epoch": 4.684650592856886,
+      "grad_norm": 0.2712259590625763,
+      "learning_rate": 5.251045846434876e-07,
+      "loss": 0.3682,
       "step": 129985
     },
     {
-      "epoch": 4.57,
-      "learning_rate": 9.55556187956233e-07,
-      "loss": 0.2545,
+      "epoch": 4.68483079251811,
+      "grad_norm": 0.2651374936103821,
+      "learning_rate": 5.245098013109573e-07,
+      "loss": 0.3631,
       "step": 129990
     },
     {
-      "epoch": 4.57,
-      "learning_rate": 9.547762751335038e-07,
-      "loss": 0.2514,
+      "epoch": 4.685010992179334,
+      "grad_norm": 0.27646490931510925,
+      "learning_rate": 5.239153514522393e-07,
+      "loss": 0.3945,
       "step": 129995
     },
     {
-      "epoch": 4.57,
-      "learning_rate": 9.539966745221563e-07,
-      "loss": 0.251,
+      "epoch": 4.685191191840559,
+      "grad_norm": 0.24400193989276886,
+      "learning_rate": 5.233212350754385e-07,
+      "loss": 0.3355,
       "step": 130000
     },
     {
-      "epoch": 4.57,
-      "eval_loss": 0.24862876534461975,
-      "eval_runtime": 10.5504,
-      "eval_samples_per_second": 9.478,
-      "eval_steps_per_second": 9.478,
+      "epoch": 4.685191191840559,
+      "eval_loss": 0.4288380742073059,
+      "eval_runtime": 3.5366,
+      "eval_samples_per_second": 28.276,
+      "eval_steps_per_second": 7.069,
       "step": 130000
     },
     {
-      "epoch": 4.57,
-      "learning_rate": 9.532173861323129e-07,
-      "loss": 0.2545,
+      "epoch": 4.685371391501784,
+      "grad_norm": 0.22407753765583038,
+      "learning_rate": 5.227274521886483e-07,
+      "loss": 0.3887,
       "step": 130005
     },
     {
-      "epoch": 4.57,
-      "learning_rate": 9.524384099740991e-07,
-      "loss": 0.2479,
+      "epoch": 4.685551591163009,
+      "grad_norm": 0.2578125298023224,
+      "learning_rate": 5.221340027999566e-07,
+      "loss": 0.3554,
       "step": 130010
     },
     {
-      "epoch": 4.57,
-      "learning_rate": 9.516597460576232e-07,
-      "loss": 0.2336,
+      "epoch": 4.685731790824233,
+      "grad_norm": 0.26388269662857056,
+      "learning_rate": 5.215408869174487e-07,
+      "loss": 0.3703,
       "step": 130015
     },
     {
-      "epoch": 4.57,
-      "learning_rate": 9.508813943929967e-07,
-      "loss": 0.2482,
+      "epoch": 4.685911990485458,
+      "grad_norm": 0.3199911117553711,
+      "learning_rate": 5.20948104549207e-07,
+      "loss": 0.3835,
       "step": 130020
     },
     {
-      "epoch": 4.57,
-      "learning_rate": 9.501033549903227e-07,
-      "loss": 0.2513,
+      "epoch": 4.686092190146683,
+      "grad_norm": 0.2785996198654175,
+      "learning_rate": 5.203556557033085e-07,
+      "loss": 0.3619,
       "step": 130025
     },
     {
-      "epoch": 4.57,
-      "learning_rate": 9.493256278597124e-07,
-      "loss": 0.2543,
+      "epoch": 4.6862723898079075,
+      "grad_norm": 0.2726043462753296,
+      "learning_rate": 5.197635403878243e-07,
+      "loss": 0.3626,
       "step": 130030
     },
     {
-      "epoch": 4.57,
-      "learning_rate": 9.485482130112549e-07,
-      "loss": 0.2624,
+      "epoch": 4.686452589469132,
+      "grad_norm": 0.2747058868408203,
+      "learning_rate": 5.19171758610823e-07,
+      "loss": 0.3786,
       "step": 130035
     },
     {
-      "epoch": 4.58,
-      "learning_rate": 9.477711104550507e-07,
-      "loss": 0.2444,
+      "epoch": 4.686632789130356,
+      "grad_norm": 0.2753622233867645,
+      "learning_rate": 5.185803103803677e-07,
+      "loss": 0.4035,
       "step": 130040
     },
     {
-      "epoch": 4.58,
-      "learning_rate": 9.46994320201186e-07,
-      "loss": 0.2581,
+      "epoch": 4.686812988791581,
+      "grad_norm": 0.3083060383796692,
+      "learning_rate": 5.179891957045158e-07,
+      "loss": 0.3657,
       "step": 130045
     },
     {
-      "epoch": 4.58,
-      "learning_rate": 9.4621784225975e-07,
-      "loss": 0.2551,
+      "epoch": 4.6869931884528055,
+      "grad_norm": 0.23582348227500916,
+      "learning_rate": 5.173984145913191e-07,
+      "loss": 0.3944,
       "step": 130050
     },
     {
-      "epoch": 4.58,
-      "learning_rate": 9.454416766408236e-07,
-      "loss": 0.2578,
+      "epoch": 4.68717338811403,
+      "grad_norm": 0.2400096207857132,
+      "learning_rate": 5.168079670488296e-07,
+      "loss": 0.3639,
       "step": 130055
     },
     {
-      "epoch": 4.58,
-      "learning_rate": 9.446658233544847e-07,
-      "loss": 0.2461,
+      "epoch": 4.687353587775255,
+      "grad_norm": 0.2639099657535553,
+      "learning_rate": 5.162178530850937e-07,
+      "loss": 0.3944,
       "step": 130060
     },
     {
-      "epoch": 4.58,
-      "learning_rate": 9.438902824108059e-07,
-      "loss": 0.2454,
+      "epoch": 4.68753378743648,
+      "grad_norm": 0.21031659841537476,
+      "learning_rate": 5.156280727081492e-07,
+      "loss": 0.3775,
       "step": 130065
     },
     {
-      "epoch": 4.58,
-      "learning_rate": 9.431150538198597e-07,
-      "loss": 0.2429,
+      "epoch": 4.6877139870977045,
+      "grad_norm": 0.2745669484138489,
+      "learning_rate": 5.150386259260314e-07,
+      "loss": 0.3419,
       "step": 130070
     },
     {
-      "epoch": 4.58,
-      "learning_rate": 9.423401375917101e-07,
-      "loss": 0.2445,
+      "epoch": 4.687894186758929,
+      "grad_norm": 0.25483670830726624,
+      "learning_rate": 5.144495127467675e-07,
+      "loss": 0.3432,
       "step": 130075
     },
     {
-      "epoch": 4.58,
-      "learning_rate": 9.41565533736416e-07,
-      "loss": 0.2653,
+      "epoch": 4.688074386420153,
+      "grad_norm": 0.27415239810943604,
+      "learning_rate": 5.138607331783951e-07,
+      "loss": 0.4003,
       "step": 130080
     },
     {
-      "epoch": 4.58,
-      "learning_rate": 9.407912422640414e-07,
-      "loss": 0.2538,
+      "epoch": 4.688254586081378,
+      "grad_norm": 0.33515846729278564,
+      "learning_rate": 5.132722872289275e-07,
+      "loss": 0.3553,
       "step": 130085
     },
     {
-      "epoch": 4.58,
-      "learning_rate": 9.400172631846337e-07,
-      "loss": 0.2458,
+      "epoch": 4.6884347857426025,
+      "grad_norm": 0.2727743089199066,
+      "learning_rate": 5.126841749063805e-07,
+      "loss": 0.3815,
       "step": 130090
     },
     {
-      "epoch": 4.58,
-      "learning_rate": 9.392435965082491e-07,
-      "loss": 0.2516,
+      "epoch": 4.688614985403827,
+      "grad_norm": 0.239381343126297,
+      "learning_rate": 5.120963962187753e-07,
+      "loss": 0.373,
       "step": 130095
     },
     {
-      "epoch": 4.58,
-      "learning_rate": 9.384702422449265e-07,
-      "loss": 0.2256,
+      "epoch": 4.688795185065052,
+      "grad_norm": 0.23639050126075745,
+      "learning_rate": 5.115089511741139e-07,
+      "loss": 0.3469,
       "step": 130100
     },
     {
-      "epoch": 4.58,
-      "learning_rate": 9.376972004047136e-07,
-      "loss": 0.2582,
+      "epoch": 4.688975384726277,
+      "grad_norm": 0.2537215054035187,
+      "learning_rate": 5.109218397804011e-07,
+      "loss": 0.3599,
       "step": 130105
     },
     {
-      "epoch": 4.58,
-      "learning_rate": 9.369244709976465e-07,
-      "loss": 0.239,
+      "epoch": 4.6891555843875015,
+      "grad_norm": 0.268621027469635,
+      "learning_rate": 5.103350620456387e-07,
+      "loss": 0.4099,
       "step": 130110
     },
     {
-      "epoch": 4.58,
-      "learning_rate": 9.361520540337564e-07,
-      "loss": 0.2414,
+      "epoch": 4.689335784048726,
+      "grad_norm": 0.22903098165988922,
+      "learning_rate": 5.097486179778177e-07,
+      "loss": 0.3794,
       "step": 130115
     },
     {
-      "epoch": 4.58,
-      "learning_rate": 9.353799495230687e-07,
-      "loss": 0.2354,
+      "epoch": 4.689515983709951,
+      "grad_norm": 0.2540788948535919,
+      "learning_rate": 5.091625075849316e-07,
+      "loss": 0.3508,
       "step": 130120
     },
     {
-      "epoch": 4.58,
-      "learning_rate": 9.346081574756166e-07,
-      "loss": 0.2258,
+      "epoch": 4.689696183371176,
+      "grad_norm": 0.2338607758283615,
+      "learning_rate": 5.085767308749629e-07,
+      "loss": 0.366,
       "step": 130125
     },
     {
-      "epoch": 4.58,
-      "learning_rate": 9.338366779014174e-07,
-      "loss": 0.2356,
+      "epoch": 4.6898763830324,
+      "grad_norm": 0.2733408808708191,
+      "learning_rate": 5.079912878558968e-07,
+      "loss": 0.3504,
       "step": 130130
     },
     {
-      "epoch": 4.58,
-      "learning_rate": 9.330655108104907e-07,
-      "loss": 0.237,
+      "epoch": 4.690056582693624,
+      "grad_norm": 0.26694294810295105,
+      "learning_rate": 5.074061785357076e-07,
+      "loss": 0.3895,
       "step": 130135
     },
     {
-      "epoch": 4.58,
-      "learning_rate": 9.322946562128426e-07,
-      "loss": 0.2546,
+      "epoch": 4.690236782354849,
+      "grad_norm": 0.2584874629974365,
+      "learning_rate": 5.068214029223639e-07,
+      "loss": 0.376,
       "step": 130140
     },
     {
-      "epoch": 4.58,
-      "learning_rate": 9.3152411411849e-07,
-      "loss": 0.2514,
+      "epoch": 4.690416982016074,
+      "grad_norm": 0.2830393314361572,
+      "learning_rate": 5.062369610238399e-07,
+      "loss": 0.3739,
       "step": 130145
     },
     {
-      "epoch": 4.58,
-      "learning_rate": 9.307538845374358e-07,
-      "loss": 0.2565,
+      "epoch": 4.690597181677298,
+      "grad_norm": 0.2988766133785248,
+      "learning_rate": 5.056528528480958e-07,
+      "loss": 0.3467,
       "step": 130150
     },
     {
-      "epoch": 4.58,
-      "learning_rate": 9.299839674796806e-07,
-      "loss": 0.2609,
+      "epoch": 4.690777381338523,
+      "grad_norm": 0.2446690797805786,
+      "learning_rate": 5.050690784030892e-07,
+      "loss": 0.393,
       "step": 130155
     },
     {
-      "epoch": 4.58,
-      "learning_rate": 9.292143629552191e-07,
-      "loss": 0.2513,
+      "epoch": 4.690957580999748,
+      "grad_norm": 0.23521770536899567,
+      "learning_rate": 5.044856376967721e-07,
+      "loss": 0.387,
       "step": 130160
     },
     {
-      "epoch": 4.58,
-      "learning_rate": 9.284450709740461e-07,
-      "loss": 0.2503,
+      "epoch": 4.691137780660973,
+      "grad_norm": 0.26140204071998596,
+      "learning_rate": 5.039025307370965e-07,
+      "loss": 0.3775,
       "step": 130165
     },
     {
-      "epoch": 4.58,
-      "learning_rate": 9.276760915461508e-07,
-      "loss": 0.2504,
+      "epoch": 4.691317980322197,
+      "grad_norm": 0.3232240676879883,
+      "learning_rate": 5.033197575320059e-07,
+      "loss": 0.3765,
       "step": 130170
     },
     {
-      "epoch": 4.58,
-      "learning_rate": 9.269074246815195e-07,
-      "loss": 0.2378,
+      "epoch": 4.691498179983421,
+      "grad_norm": 0.2332460880279541,
+      "learning_rate": 5.027373180894441e-07,
+      "loss": 0.375,
       "step": 130175
     },
     {
-      "epoch": 4.58,
-      "learning_rate": 9.261390703901251e-07,
-      "loss": 0.2432,
+      "epoch": 4.691678379644646,
+      "grad_norm": 0.25555670261383057,
+      "learning_rate": 5.021552124173379e-07,
+      "loss": 0.3493,
       "step": 130180
     },
     {
-      "epoch": 4.58,
-      "learning_rate": 9.25371028681954e-07,
-      "loss": 0.2401,
+      "epoch": 4.691858579305871,
+      "grad_norm": 0.2844722270965576,
+      "learning_rate": 5.015734405236284e-07,
+      "loss": 0.3675,
       "step": 130185
     },
     {
-      "epoch": 4.58,
-      "learning_rate": 9.246032995669728e-07,
-      "loss": 0.2513,
+      "epoch": 4.692038778967095,
+      "grad_norm": 0.282840758562088,
+      "learning_rate": 5.009920024162368e-07,
+      "loss": 0.4054,
       "step": 130190
     },
     {
-      "epoch": 4.58,
-      "learning_rate": 9.238358830551519e-07,
-      "loss": 0.2554,
+      "epoch": 4.69221897862832,
+      "grad_norm": 0.22690938413143158,
+      "learning_rate": 5.004108981030847e-07,
+      "loss": 0.3435,
       "step": 130195
     },
     {
-      "epoch": 4.58,
-      "learning_rate": 9.230687791564524e-07,
-      "loss": 0.2605,
+      "epoch": 4.692399178289545,
+      "grad_norm": 0.24626252055168152,
+      "learning_rate": 4.998301275920936e-07,
+      "loss": 0.3498,
       "step": 130200
     },
     {
-      "epoch": 4.58,
-      "learning_rate": 9.223019878808415e-07,
-      "loss": 0.2507,
+      "epoch": 4.69257937795077,
+      "grad_norm": 0.23748280107975006,
+      "learning_rate": 4.99249690891171e-07,
+      "loss": 0.401,
       "step": 130205
     },
     {
-      "epoch": 4.58,
-      "learning_rate": 9.215355092382694e-07,
-      "loss": 0.2304,
+      "epoch": 4.692759577611994,
+      "grad_norm": 0.2302643209695816,
+      "learning_rate": 4.98669588008227e-07,
+      "loss": 0.3868,
       "step": 130210
     },
     {
-      "epoch": 4.58,
-      "learning_rate": 9.207693432386893e-07,
-      "loss": 0.248,
+      "epoch": 4.692939777273219,
+      "grad_norm": 0.2521742582321167,
+      "learning_rate": 4.980898189511668e-07,
+      "loss": 0.3973,
       "step": 130215
     },
     {
-      "epoch": 4.58,
-      "learning_rate": 9.200034898920546e-07,
-      "loss": 0.2407,
+      "epoch": 4.693119976934444,
+      "grad_norm": 0.24064558744430542,
+      "learning_rate": 4.975103837278921e-07,
+      "loss": 0.3831,
       "step": 130220
     },
     {
-      "epoch": 4.58,
-      "learning_rate": 9.192379492083042e-07,
-      "loss": 0.2477,
+      "epoch": 4.693300176595668,
+      "grad_norm": 0.233879953622818,
+      "learning_rate": 4.969312823462912e-07,
+      "loss": 0.3357,
       "step": 130225
     },
     {
-      "epoch": 4.58,
-      "learning_rate": 9.184727211973776e-07,
-      "loss": 0.2558,
+      "epoch": 4.693480376256892,
+      "grad_norm": 0.27745091915130615,
+      "learning_rate": 4.963525148142606e-07,
+      "loss": 0.3951,
       "step": 130230
     },
     {
-      "epoch": 4.58,
-      "learning_rate": 9.177078058692112e-07,
-      "loss": 0.2516,
+      "epoch": 4.693660575918117,
+      "grad_norm": 0.2301134169101715,
+      "learning_rate": 4.9577408113968e-07,
+      "loss": 0.381,
       "step": 130235
     },
     {
-      "epoch": 4.58,
-      "learning_rate": 9.169432032337416e-07,
-      "loss": 0.2311,
+      "epoch": 4.693840775579342,
+      "grad_norm": 0.2147936373949051,
+      "learning_rate": 4.951959813304346e-07,
+      "loss": 0.3665,
       "step": 130240
     },
     {
-      "epoch": 4.58,
-      "learning_rate": 9.161789133008914e-07,
-      "loss": 0.2586,
+      "epoch": 4.694020975240567,
+      "grad_norm": 0.24145711958408356,
+      "learning_rate": 4.94618215394399e-07,
+      "loss": 0.4068,
       "step": 130245
     },
     {
-      "epoch": 4.58,
-      "learning_rate": 9.15414936080583e-07,
-      "loss": 0.2382,
+      "epoch": 4.694201174901791,
+      "grad_norm": 0.23923490941524506,
+      "learning_rate": 4.940407833394473e-07,
+      "loss": 0.4088,
       "step": 130250
     },
     {
-      "epoch": 4.58,
-      "learning_rate": 9.146512715827421e-07,
-      "loss": 0.2545,
+      "epoch": 4.694381374563016,
+      "grad_norm": 0.24783268570899963,
+      "learning_rate": 4.934636851734453e-07,
+      "loss": 0.3628,
       "step": 130255
     },
     {
-      "epoch": 4.58,
-      "learning_rate": 9.1388791981728e-07,
-      "loss": 0.237,
+      "epoch": 4.694561574224241,
+      "grad_norm": 0.2156895250082016,
+      "learning_rate": 4.928869209042536e-07,
+      "loss": 0.3927,
       "step": 130260
     },
     {
-      "epoch": 4.58,
-      "learning_rate": 9.13124880794114e-07,
-      "loss": 0.2335,
+      "epoch": 4.694741773885465,
+      "grad_norm": 0.28466811776161194,
+      "learning_rate": 4.923104905397379e-07,
+      "loss": 0.4006,
       "step": 130265
     },
     {
-      "epoch": 4.58,
-      "learning_rate": 9.12362154523147e-07,
-      "loss": 0.2433,
+      "epoch": 4.694921973546689,
+      "grad_norm": 0.25431156158447266,
+      "learning_rate": 4.917343940877423e-07,
+      "loss": 0.3977,
       "step": 130270
     },
     {
-      "epoch": 4.58,
-      "learning_rate": 9.115997410142796e-07,
-      "loss": 0.2211,
+      "epoch": 4.695102173207914,
+      "grad_norm": 0.3024144470691681,
+      "learning_rate": 4.911586315561212e-07,
+      "loss": 0.3901,
       "step": 130275
     },
     {
-      "epoch": 4.58,
-      "learning_rate": 9.108376402774177e-07,
-      "loss": 0.2445,
+      "epoch": 4.695282372869139,
+      "grad_norm": 0.29389986395835876,
+      "learning_rate": 4.905832029527186e-07,
+      "loss": 0.3691,
       "step": 130280
     },
     {
-      "epoch": 4.58,
-      "learning_rate": 9.100758523224534e-07,
-      "loss": 0.2432,
+      "epoch": 4.695462572530364,
+      "grad_norm": 0.26747405529022217,
+      "learning_rate": 4.900081082853753e-07,
+      "loss": 0.4077,
       "step": 130285
     },
     {
-      "epoch": 4.58,
-      "learning_rate": 9.09314377159276e-07,
-      "loss": 0.2565,
+      "epoch": 4.695642772191588,
+      "grad_norm": 0.259396493434906,
+      "learning_rate": 4.894333475619295e-07,
+      "loss": 0.3603,
       "step": 130290
     },
     {
-      "epoch": 4.58,
-      "learning_rate": 9.085532147977749e-07,
-      "loss": 0.2818,
+      "epoch": 4.695822971852813,
+      "grad_norm": 0.25687748193740845,
+      "learning_rate": 4.888589207902056e-07,
+      "loss": 0.374,
       "step": 130295
     },
     {
-      "epoch": 4.58,
-      "learning_rate": 9.077923652478365e-07,
-      "loss": 0.2358,
+      "epoch": 4.696003171514038,
+      "grad_norm": 0.2638847231864929,
+      "learning_rate": 4.882848279780306e-07,
+      "loss": 0.4057,
       "step": 130300
     },
     {
-      "epoch": 4.58,
-      "learning_rate": 9.070318285193336e-07,
-      "loss": 0.2563,
+      "epoch": 4.6961833711752625,
+      "grad_norm": 0.24750365316867828,
+      "learning_rate": 4.877110691332343e-07,
+      "loss": 0.3794,
       "step": 130305
     },
     {
-      "epoch": 4.58,
-      "learning_rate": 9.06271604622147e-07,
-      "loss": 0.2659,
+      "epoch": 4.696363570836487,
+      "grad_norm": 0.22299759089946747,
+      "learning_rate": 4.871376442636272e-07,
+      "loss": 0.379,
       "step": 130310
     },
     {
-      "epoch": 4.58,
-      "learning_rate": 9.055116935661439e-07,
-      "loss": 0.2431,
+      "epoch": 4.696543770497711,
+      "grad_norm": 0.24523359537124634,
+      "learning_rate": 4.865645533770224e-07,
+      "loss": 0.3825,
       "step": 130315
     },
     {
-      "epoch": 4.59,
-      "learning_rate": 9.047520953611943e-07,
-      "loss": 0.2275,
+      "epoch": 4.696723970158936,
+      "grad_norm": 0.24802939593791962,
+      "learning_rate": 4.859917964812305e-07,
+      "loss": 0.4062,
       "step": 130320
     },
     {
-      "epoch": 4.59,
-      "learning_rate": 9.039928100171569e-07,
-      "loss": 0.2477,
+      "epoch": 4.696904169820161,
+      "grad_norm": 0.25451767444610596,
+      "learning_rate": 4.854193735840534e-07,
+      "loss": 0.3949,
       "step": 130325
     },
     {
-      "epoch": 4.59,
-      "learning_rate": 9.032338375438931e-07,
-      "loss": 0.2529,
+      "epoch": 4.697084369481385,
+      "grad_norm": 0.2552180588245392,
+      "learning_rate": 4.848472846932933e-07,
+      "loss": 0.3683,
       "step": 130330
     },
     {
-      "epoch": 4.59,
-      "learning_rate": 9.024751779512592e-07,
-      "loss": 0.255,
+      "epoch": 4.69726456914261,
+      "grad_norm": 0.2642126679420471,
+      "learning_rate": 4.842755298167412e-07,
+      "loss": 0.3984,
       "step": 130335
     },
     {
-      "epoch": 4.59,
-      "learning_rate": 9.017168312491053e-07,
-      "loss": 0.23,
+      "epoch": 4.697444768803835,
+      "grad_norm": 0.2669326961040497,
+      "learning_rate": 4.837041089621908e-07,
+      "loss": 0.3565,
       "step": 130340
     },
     {
-      "epoch": 4.59,
-      "learning_rate": 9.009587974472767e-07,
-      "loss": 0.2536,
+      "epoch": 4.6976249684650595,
+      "grad_norm": 0.27341264486312866,
+      "learning_rate": 4.831330221374248e-07,
+      "loss": 0.4032,
       "step": 130345
     },
     {
-      "epoch": 4.59,
-      "learning_rate": 9.002010765556124e-07,
-      "loss": 0.23,
+      "epoch": 4.697805168126284,
+      "grad_norm": 0.2548832297325134,
+      "learning_rate": 4.82562269350223e-07,
+      "loss": 0.3906,
       "step": 130350
     },
     {
-      "epoch": 4.59,
-      "learning_rate": 8.994436685839602e-07,
-      "loss": 0.2501,
+      "epoch": 4.697985367787508,
+      "grad_norm": 0.29564833641052246,
+      "learning_rate": 4.819918506083627e-07,
+      "loss": 0.3535,
       "step": 130355
     },
     {
-      "epoch": 4.59,
-      "learning_rate": 8.986865735421429e-07,
-      "loss": 0.2553,
+      "epoch": 4.698165567448733,
+      "grad_norm": 0.29607510566711426,
+      "learning_rate": 4.814217659196207e-07,
+      "loss": 0.339,
       "step": 130360
     },
     {
-      "epoch": 4.59,
-      "learning_rate": 8.979297914400026e-07,
-      "loss": 0.2381,
+      "epoch": 4.6983457671099575,
+      "grad_norm": 0.2381865680217743,
+      "learning_rate": 4.80852015291755e-07,
+      "loss": 0.3721,
       "step": 130365
     },
     {
-      "epoch": 4.59,
-      "learning_rate": 8.971733222873563e-07,
-      "loss": 0.2512,
+      "epoch": 4.698525966771182,
+      "grad_norm": 0.2766072154045105,
+      "learning_rate": 4.802825987325371e-07,
+      "loss": 0.3651,
       "step": 130370
     },
     {
-      "epoch": 4.59,
-      "learning_rate": 8.964171660940323e-07,
-      "loss": 0.2411,
+      "epoch": 4.698706166432407,
+      "grad_norm": 0.26471713185310364,
+      "learning_rate": 4.797135162497219e-07,
+      "loss": 0.3566,
       "step": 130375
     },
     {
-      "epoch": 4.59,
-      "learning_rate": 8.956613228698479e-07,
-      "loss": 0.253,
+      "epoch": 4.698886366093632,
+      "grad_norm": 0.2708904445171356,
+      "learning_rate": 4.791447678510613e-07,
+      "loss": 0.39,
       "step": 130380
     },
     {
-      "epoch": 4.59,
-      "learning_rate": 8.949057926246146e-07,
-      "loss": 0.2265,
+      "epoch": 4.6990665657548565,
+      "grad_norm": 0.23361581563949585,
+      "learning_rate": 4.785763535443078e-07,
+      "loss": 0.3869,
       "step": 130385
     },
     {
-      "epoch": 4.59,
-      "learning_rate": 8.941505753681412e-07,
-      "loss": 0.2592,
+      "epoch": 4.699246765416081,
+      "grad_norm": 0.19293129444122314,
+      "learning_rate": 4.780082733372021e-07,
+      "loss": 0.3327,
       "step": 130390
     },
     {
-      "epoch": 4.59,
-      "learning_rate": 8.933956711102393e-07,
-      "loss": 0.2514,
+      "epoch": 4.699426965077306,
+      "grad_norm": 0.2304544597864151,
+      "learning_rate": 4.774405272374827e-07,
+      "loss": 0.3904,
       "step": 130395
     },
     {
-      "epoch": 4.59,
-      "learning_rate": 8.926410798607066e-07,
-      "loss": 0.2592,
+      "epoch": 4.699607164738531,
+      "grad_norm": 0.2673514187335968,
+      "learning_rate": 4.768731152528933e-07,
+      "loss": 0.4132,
       "step": 130400
     },
     {
-      "epoch": 4.59,
-      "learning_rate": 8.918868016293435e-07,
-      "loss": 0.2251,
+      "epoch": 4.699787364399755,
+      "grad_norm": 0.27514368295669556,
+      "learning_rate": 4.763060373911582e-07,
+      "loss": 0.3417,
       "step": 130405
     },
     {
-      "epoch": 4.59,
-      "learning_rate": 8.911328364259397e-07,
-      "loss": 0.2537,
+      "epoch": 4.699967564060979,
+      "grad_norm": 0.24293185770511627,
+      "learning_rate": 4.7573929366000746e-07,
+      "loss": 0.3843,
       "step": 130410
     },
     {
-      "epoch": 4.59,
-      "learning_rate": 8.903791842602899e-07,
-      "loss": 0.2395,
+      "epoch": 4.700147763722204,
+      "grad_norm": 0.27645057439804077,
+      "learning_rate": 4.751728840671599e-07,
+      "loss": 0.3684,
       "step": 130415
     },
     {
-      "epoch": 4.59,
-      "learning_rate": 8.896258451421752e-07,
-      "loss": 0.2495,
+      "epoch": 4.700327963383429,
+      "grad_norm": 0.23572184145450592,
+      "learning_rate": 4.7460680862033146e-07,
+      "loss": 0.3592,
       "step": 130420
     },
     {
-      "epoch": 4.59,
-      "learning_rate": 8.888728190813822e-07,
-      "loss": 0.2558,
+      "epoch": 4.7005081630446535,
+      "grad_norm": 0.26540112495422363,
+      "learning_rate": 4.740410673272383e-07,
+      "loss": 0.3695,
       "step": 130425
     },
     {
-      "epoch": 4.59,
-      "learning_rate": 8.881201060876837e-07,
-      "loss": 0.2303,
+      "epoch": 4.700688362705878,
+      "grad_norm": 0.29115819931030273,
+      "learning_rate": 4.7347566019558807e-07,
+      "loss": 0.3774,
       "step": 130430
     },
     {
-      "epoch": 4.59,
-      "learning_rate": 8.873677061708551e-07,
-      "loss": 0.2421,
+      "epoch": 4.700868562367103,
+      "grad_norm": 0.239894837141037,
+      "learning_rate": 4.72910587233083e-07,
+      "loss": 0.3866,
       "step": 130435
     },
     {
-      "epoch": 4.59,
-      "learning_rate": 8.866156193406694e-07,
-      "loss": 0.2329,
+      "epoch": 4.701048762028328,
+      "grad_norm": 0.2407197803258896,
+      "learning_rate": 4.723458484474225e-07,
+      "loss": 0.3989,
       "step": 130440
     },
     {
-      "epoch": 4.59,
-      "learning_rate": 8.858638456068852e-07,
-      "loss": 0.2282,
+      "epoch": 4.701228961689552,
+      "grad_norm": 0.20911681652069092,
+      "learning_rate": 4.717814438462975e-07,
+      "loss": 0.3579,
       "step": 130445
     },
     {
-      "epoch": 4.59,
-      "learning_rate": 8.851123849792698e-07,
-      "loss": 0.2595,
+      "epoch": 4.701409161350776,
+      "grad_norm": 0.3005968928337097,
+      "learning_rate": 4.712173734374048e-07,
+      "loss": 0.3647,
       "step": 130450
     },
     {
-      "epoch": 4.59,
-      "learning_rate": 8.843612374675764e-07,
-      "loss": 0.2371,
+      "epoch": 4.701589361012001,
+      "grad_norm": 0.24574686586856842,
+      "learning_rate": 4.706536372284298e-07,
+      "loss": 0.3752,
       "step": 130455
     },
     {
-      "epoch": 4.59,
-      "learning_rate": 8.836104030815612e-07,
-      "loss": 0.2655,
+      "epoch": 4.701769560673226,
+      "grad_norm": 0.33672651648521423,
+      "learning_rate": 4.700902352270442e-07,
+      "loss": 0.4029,
       "step": 130460
     },
     {
-      "epoch": 4.59,
-      "learning_rate": 8.828598818309663e-07,
-      "loss": 0.248,
+      "epoch": 4.7019497603344504,
+      "grad_norm": 0.22763417661190033,
+      "learning_rate": 4.695271674409335e-07,
+      "loss": 0.3716,
       "step": 130465
     },
     {
-      "epoch": 4.59,
-      "learning_rate": 8.82109673725548e-07,
-      "loss": 0.2432,
+      "epoch": 4.702129959995675,
+      "grad_norm": 0.24694277346134186,
+      "learning_rate": 4.689644338777638e-07,
+      "loss": 0.3485,
       "step": 130470
     },
     {
-      "epoch": 4.59,
-      "learning_rate": 8.813597787750372e-07,
-      "loss": 0.2633,
+      "epoch": 4.7023101596569,
+      "grad_norm": 0.21548400819301605,
+      "learning_rate": 4.684020345452067e-07,
+      "loss": 0.3963,
       "step": 130475
     },
     {
-      "epoch": 4.59,
-      "learning_rate": 8.806101969891761e-07,
-      "loss": 0.2779,
+      "epoch": 4.702490359318125,
+      "grad_norm": 0.260619580745697,
+      "learning_rate": 4.678399694509228e-07,
+      "loss": 0.4,
       "step": 130480
     },
     {
-      "epoch": 4.59,
-      "learning_rate": 8.79860928377696e-07,
-      "loss": 0.2375,
+      "epoch": 4.702670558979349,
+      "grad_norm": 0.31047627329826355,
+      "learning_rate": 4.67278238602567e-07,
+      "loss": 0.3997,
       "step": 130485
     },
     {
-      "epoch": 4.59,
-      "learning_rate": 8.791119729503278e-07,
-      "loss": 0.2311,
+      "epoch": 4.702850758640574,
+      "grad_norm": 0.2056533545255661,
+      "learning_rate": 4.6671684200779717e-07,
+      "loss": 0.3476,
       "step": 130490
     },
     {
-      "epoch": 4.59,
-      "learning_rate": 8.783633307167943e-07,
-      "loss": 0.2546,
+      "epoch": 4.703030958301799,
+      "grad_norm": 0.24057982861995697,
+      "learning_rate": 4.6615577967425996e-07,
+      "loss": 0.3588,
       "step": 130495
     },
     {
-      "epoch": 4.59,
-      "learning_rate": 8.776150016868157e-07,
-      "loss": 0.2328,
+      "epoch": 4.703211157963023,
+      "grad_norm": 0.25371670722961426,
+      "learning_rate": 4.6559505160959916e-07,
+      "loss": 0.4112,
       "step": 130500
     },
     {
-      "epoch": 4.59,
-      "eval_loss": 0.24856102466583252,
-      "eval_runtime": 10.5535,
-      "eval_samples_per_second": 9.475,
-      "eval_steps_per_second": 9.475,
+      "epoch": 4.703211157963023,
+      "eval_loss": 0.42873358726501465,
+      "eval_runtime": 3.5322,
+      "eval_samples_per_second": 28.311,
+      "eval_steps_per_second": 7.078,
       "step": 130500
     },
     {
-      "epoch": 4.59,
-      "learning_rate": 8.768669858701061e-07,
-      "loss": 0.2428,
+      "epoch": 4.703391357624247,
+      "grad_norm": 0.3009524345397949,
+      "learning_rate": 4.6503465782145873e-07,
+      "loss": 0.3473,
       "step": 130505
     },
     {
-      "epoch": 4.59,
-      "learning_rate": 8.761192832763859e-07,
-      "loss": 0.2483,
+      "epoch": 4.703571557285472,
+      "grad_norm": 0.27046623826026917,
+      "learning_rate": 4.6447459831746586e-07,
+      "loss": 0.395,
       "step": 130510
     },
     {
-      "epoch": 4.59,
-      "learning_rate": 8.753718939153555e-07,
-      "loss": 0.2551,
+      "epoch": 4.703751756946697,
+      "grad_norm": 0.2447580248117447,
+      "learning_rate": 4.639148731052617e-07,
+      "loss": 0.3678,
       "step": 130515
     },
     {
-      "epoch": 4.59,
-      "learning_rate": 8.746248177967209e-07,
-      "loss": 0.2682,
+      "epoch": 4.703931956607922,
+      "grad_norm": 0.28649699687957764,
+      "learning_rate": 4.6335548219246517e-07,
+      "loss": 0.3448,
       "step": 130520
     },
     {
-      "epoch": 4.59,
-      "learning_rate": 8.738780549301884e-07,
-      "loss": 0.2427,
+      "epoch": 4.704112156269146,
+      "grad_norm": 0.28706684708595276,
+      "learning_rate": 4.6279642558670064e-07,
+      "loss": 0.3887,
       "step": 130525
     },
     {
-      "epoch": 4.59,
-      "learning_rate": 8.731316053254446e-07,
-      "loss": 0.2679,
+      "epoch": 4.704292355930371,
+      "grad_norm": 0.20302236080169678,
+      "learning_rate": 4.622377032955816e-07,
+      "loss": 0.3479,
       "step": 130530
     },
     {
-      "epoch": 4.59,
-      "learning_rate": 8.723854689921901e-07,
-      "loss": 0.2462,
+      "epoch": 4.704472555591596,
+      "grad_norm": 0.2422175258398056,
+      "learning_rate": 4.616793153267268e-07,
+      "loss": 0.3707,
       "step": 130535
     },
     {
-      "epoch": 4.59,
-      "learning_rate": 8.716396459401088e-07,
-      "loss": 0.2441,
+      "epoch": 4.70465275525282,
+      "grad_norm": 0.25601354241371155,
+      "learning_rate": 4.6112126168773587e-07,
+      "loss": 0.3576,
       "step": 130540
     },
     {
-      "epoch": 4.59,
-      "learning_rate": 8.708941361788847e-07,
-      "loss": 0.2305,
+      "epoch": 4.704832954914044,
+      "grad_norm": 0.2792492210865021,
+      "learning_rate": 4.60563542386222e-07,
+      "loss": 0.3739,
       "step": 130545
     },
     {
-      "epoch": 4.59,
-      "learning_rate": 8.701489397181989e-07,
-      "loss": 0.2483,
+      "epoch": 4.705013154575269,
+      "grad_norm": 0.22106266021728516,
+      "learning_rate": 4.600061574297765e-07,
+      "loss": 0.3579,
       "step": 130550
     },
     {
-      "epoch": 4.59,
-      "learning_rate": 8.694040565677297e-07,
-      "loss": 0.238,
+      "epoch": 4.705193354236494,
+      "grad_norm": 0.2654285430908203,
+      "learning_rate": 4.5944910682599597e-07,
+      "loss": 0.3411,
       "step": 130555
     },
     {
-      "epoch": 4.59,
-      "learning_rate": 8.686594867371417e-07,
-      "loss": 0.2371,
+      "epoch": 4.705373553897719,
+      "grad_norm": 0.32557204365730286,
+      "learning_rate": 4.588923905824688e-07,
+      "loss": 0.3711,
       "step": 130560
     },
     {
-      "epoch": 4.59,
-      "learning_rate": 8.679152302361104e-07,
-      "loss": 0.2677,
+      "epoch": 4.705553753558943,
+      "grad_norm": 0.308724969625473,
+      "learning_rate": 4.583360087067834e-07,
+      "loss": 0.3891,
       "step": 130565
     },
     {
-      "epoch": 4.59,
-      "learning_rate": 8.671712870742948e-07,
-      "loss": 0.2529,
+      "epoch": 4.705733953220168,
+      "grad_norm": 0.2696959376335144,
+      "learning_rate": 4.577799612065198e-07,
+      "loss": 0.3632,
       "step": 130570
     },
     {
-      "epoch": 4.59,
-      "learning_rate": 8.664276572613595e-07,
-      "loss": 0.2513,
+      "epoch": 4.705914152881393,
+      "grad_norm": 0.280472069978714,
+      "learning_rate": 4.572242480892497e-07,
+      "loss": 0.3685,
       "step": 130575
     },
     {
-      "epoch": 4.59,
-      "learning_rate": 8.656843408069493e-07,
-      "loss": 0.2493,
+      "epoch": 4.7060943525426175,
+      "grad_norm": 0.26516270637512207,
+      "learning_rate": 4.566688693625476e-07,
+      "loss": 0.3442,
       "step": 130580
     },
     {
-      "epoch": 4.59,
-      "learning_rate": 8.64941337720726e-07,
-      "loss": 0.2527,
+      "epoch": 4.706274552203842,
+      "grad_norm": 0.2824823558330536,
+      "learning_rate": 4.5611382503398246e-07,
+      "loss": 0.372,
       "step": 130585
     },
     {
-      "epoch": 4.59,
-      "learning_rate": 8.641986480123349e-07,
-      "loss": 0.2418,
+      "epoch": 4.706454751865066,
+      "grad_norm": 0.2615325450897217,
+      "learning_rate": 4.5555911511111495e-07,
+      "loss": 0.3817,
       "step": 130590
     },
     {
-      "epoch": 4.59,
-      "learning_rate": 8.634562716914179e-07,
-      "loss": 0.2504,
+      "epoch": 4.706634951526291,
+      "grad_norm": 0.2668514549732208,
+      "learning_rate": 4.5500473960150004e-07,
+      "loss": 0.3485,
       "step": 130595
     },
     {
-      "epoch": 4.59,
-      "learning_rate": 8.627142087676121e-07,
-      "loss": 0.257,
+      "epoch": 4.706815151187516,
+      "grad_norm": 0.286246657371521,
+      "learning_rate": 4.5445069851269564e-07,
+      "loss": 0.4019,
       "step": 130600
     },
     {
-      "epoch": 4.6,
-      "learning_rate": 8.619724592505596e-07,
-      "loss": 0.2459,
+      "epoch": 4.70699535084874,
+      "grad_norm": 0.26146531105041504,
+      "learning_rate": 4.5389699185224564e-07,
+      "loss": 0.3399,
       "step": 130605
     },
     {
-      "epoch": 4.6,
-      "learning_rate": 8.61231023149886e-07,
-      "loss": 0.2586,
+      "epoch": 4.707175550509965,
+      "grad_norm": 0.27129244804382324,
+      "learning_rate": 4.5334361962769965e-07,
+      "loss": 0.384,
       "step": 130610
     },
     {
-      "epoch": 4.6,
-      "learning_rate": 8.604899004752198e-07,
-      "loss": 0.2216,
+      "epoch": 4.70735575017119,
+      "grad_norm": 0.21053126454353333,
+      "learning_rate": 4.527905818465933e-07,
+      "loss": 0.3281,
       "step": 130615
     },
     {
-      "epoch": 4.6,
-      "learning_rate": 8.59749091236181e-07,
-      "loss": 0.2588,
+      "epoch": 4.7075359498324145,
+      "grad_norm": 0.287487655878067,
+      "learning_rate": 4.5223787851646215e-07,
+      "loss": 0.3927,
       "step": 130620
     },
     {
-      "epoch": 4.6,
-      "learning_rate": 8.590085954423955e-07,
-      "loss": 0.2517,
+      "epoch": 4.707716149493639,
+      "grad_norm": 0.25808537006378174,
+      "learning_rate": 4.5168550964483915e-07,
+      "loss": 0.3529,
       "step": 130625
     },
     {
-      "epoch": 4.6,
-      "learning_rate": 8.582684131034718e-07,
-      "loss": 0.2439,
+      "epoch": 4.707896349154863,
+      "grad_norm": 0.21509693562984467,
+      "learning_rate": 4.51133475239246e-07,
+      "loss": 0.3739,
       "step": 130630
     },
     {
-      "epoch": 4.6,
-      "learning_rate": 8.575285442290193e-07,
-      "loss": 0.2547,
+      "epoch": 4.708076548816088,
+      "grad_norm": 0.3704529404640198,
+      "learning_rate": 4.5058177530720724e-07,
+      "loss": 0.3907,
       "step": 130635
     },
     {
-      "epoch": 4.6,
-      "learning_rate": 8.567889888286495e-07,
-      "loss": 0.265,
+      "epoch": 4.708256748477313,
+      "grad_norm": 0.25836381316185,
+      "learning_rate": 4.500304098562419e-07,
+      "loss": 0.3895,
       "step": 130640
     },
     {
-      "epoch": 4.6,
-      "learning_rate": 8.56049746911966e-07,
-      "loss": 0.2511,
+      "epoch": 4.708436948138537,
+      "grad_norm": 0.21811331808567047,
+      "learning_rate": 4.4947937889385507e-07,
+      "loss": 0.382,
       "step": 130645
     },
     {
-      "epoch": 4.6,
-      "learning_rate": 8.553108184885666e-07,
-      "loss": 0.2476,
+      "epoch": 4.708617147799762,
+      "grad_norm": 0.24641865491867065,
+      "learning_rate": 4.489286824275601e-07,
+      "loss": 0.3452,
       "step": 130650
     },
     {
-      "epoch": 4.6,
-      "learning_rate": 8.545722035680409e-07,
-      "loss": 0.2505,
+      "epoch": 4.708797347460987,
+      "grad_norm": 0.23483498394489288,
+      "learning_rate": 4.483783204648595e-07,
+      "loss": 0.3834,
       "step": 130655
     },
     {
-      "epoch": 4.6,
-      "learning_rate": 8.538339021599839e-07,
-      "loss": 0.2566,
+      "epoch": 4.7089775471222115,
+      "grad_norm": 0.28817734122276306,
+      "learning_rate": 4.4782829301324993e-07,
+      "loss": 0.3866,
       "step": 130660
     },
     {
-      "epoch": 4.6,
-      "learning_rate": 8.530959142739797e-07,
-      "loss": 0.2655,
+      "epoch": 4.709157746783436,
+      "grad_norm": 0.2239607572555542,
+      "learning_rate": 4.4738851190498645e-07,
+      "loss": 0.3863,
       "step": 130665
     },
     {
-      "epoch": 4.6,
-      "learning_rate": 8.523582399196151e-07,
-      "loss": 0.2626,
+      "epoch": 4.709337946444661,
+      "grad_norm": 0.22553732991218567,
+      "learning_rate": 4.4683908659222396e-07,
+      "loss": 0.3521,
       "step": 130670
     },
     {
-      "epoch": 4.6,
-      "learning_rate": 8.5162087910646e-07,
-      "loss": 0.2215,
+      "epoch": 4.709518146105886,
+      "grad_norm": 0.21854856610298157,
+      "learning_rate": 4.4628999581152463e-07,
+      "loss": 0.3594,
       "step": 130675
     },
     {
-      "epoch": 4.6,
-      "learning_rate": 8.508838318440959e-07,
-      "loss": 0.2521,
+      "epoch": 4.70969834576711,
+      "grad_norm": 0.22778260707855225,
+      "learning_rate": 4.457412395703714e-07,
+      "loss": 0.3684,
       "step": 130680
     },
     {
-      "epoch": 4.6,
-      "learning_rate": 8.501470981420928e-07,
-      "loss": 0.26,
+      "epoch": 4.709878545428334,
+      "grad_norm": 0.20570991933345795,
+      "learning_rate": 4.451928178762416e-07,
+      "loss": 0.3913,
       "step": 130685
     },
     {
-      "epoch": 4.6,
-      "learning_rate": 8.494106780100098e-07,
-      "loss": 0.229,
+      "epoch": 4.710058745089559,
+      "grad_norm": 0.26178503036499023,
+      "learning_rate": 4.446447307366042e-07,
+      "loss": 0.3578,
       "step": 130690
     },
     {
-      "epoch": 4.6,
-      "learning_rate": 8.48674571457414e-07,
-      "loss": 0.2412,
+      "epoch": 4.710238944750784,
+      "grad_norm": 0.2906634211540222,
+      "learning_rate": 4.4409697815892827e-07,
+      "loss": 0.3814,
       "step": 130695
     },
     {
-      "epoch": 4.6,
-      "learning_rate": 8.479387784938647e-07,
-      "loss": 0.2515,
+      "epoch": 4.7104191444120085,
+      "grad_norm": 0.24634572863578796,
+      "learning_rate": 4.435495601506745e-07,
+      "loss": 0.3709,
       "step": 130700
     },
     {
-      "epoch": 4.6,
-      "learning_rate": 8.472032991289125e-07,
-      "loss": 0.2331,
+      "epoch": 4.710599344073233,
+      "grad_norm": 0.205975741147995,
+      "learning_rate": 4.430024767193064e-07,
+      "loss": 0.3627,
       "step": 130705
     },
     {
-      "epoch": 4.6,
-      "learning_rate": 8.464681333721109e-07,
-      "loss": 0.2427,
+      "epoch": 4.710779543734458,
+      "grad_norm": 0.24753928184509277,
+      "learning_rate": 4.4245572787227905e-07,
+      "loss": 0.3751,
       "step": 130710
     },
     {
-      "epoch": 4.6,
-      "learning_rate": 8.457332812329966e-07,
-      "loss": 0.26,
+      "epoch": 4.710959743395683,
+      "grad_norm": 0.24739667773246765,
+      "learning_rate": 4.41909313617031e-07,
+      "loss": 0.3678,
       "step": 130715
     },
     {
-      "epoch": 4.6,
-      "learning_rate": 8.449987427211231e-07,
-      "loss": 0.2556,
+      "epoch": 4.711139943056907,
+      "grad_norm": 0.19426771998405457,
+      "learning_rate": 4.4136323396101743e-07,
+      "loss": 0.3344,
       "step": 130720
     },
     {
-      "epoch": 4.6,
-      "learning_rate": 8.442645178460217e-07,
-      "loss": 0.2508,
+      "epoch": 4.711320142718131,
+      "grad_norm": 0.2439907044172287,
+      "learning_rate": 4.408174889116712e-07,
+      "loss": 0.3633,
       "step": 130725
     },
     {
-      "epoch": 4.6,
-      "learning_rate": 8.435306066172238e-07,
-      "loss": 0.2447,
+      "epoch": 4.711500342379356,
+      "grad_norm": 0.2416488081216812,
+      "learning_rate": 4.402720784764391e-07,
+      "loss": 0.3851,
       "step": 130730
     },
     {
-      "epoch": 4.6,
-      "learning_rate": 8.427970090442605e-07,
-      "loss": 0.2612,
+      "epoch": 4.711680542040581,
+      "grad_norm": 0.23560959100723267,
+      "learning_rate": 4.3972700266274035e-07,
+      "loss": 0.3564,
       "step": 130735
     },
     {
-      "epoch": 4.6,
-      "learning_rate": 8.420637251366604e-07,
-      "loss": 0.2698,
+      "epoch": 4.7118607417018055,
+      "grad_norm": 0.24532419443130493,
+      "learning_rate": 4.3918226147800503e-07,
+      "loss": 0.3914,
       "step": 130740
     },
     {
-      "epoch": 4.6,
-      "learning_rate": 8.413307549039406e-07,
-      "loss": 0.2544,
+      "epoch": 4.71204094136303,
+      "grad_norm": 0.25437724590301514,
+      "learning_rate": 4.386378549296577e-07,
+      "loss": 0.3805,
       "step": 130745
     },
     {
-      "epoch": 4.6,
-      "learning_rate": 8.405980983556216e-07,
-      "loss": 0.2537,
+      "epoch": 4.712221141024255,
+      "grad_norm": 0.2842262089252472,
+      "learning_rate": 4.380937830251175e-07,
+      "loss": 0.3494,
       "step": 130750
     },
     {
-      "epoch": 4.6,
-      "learning_rate": 8.398657555012123e-07,
-      "loss": 0.2554,
+      "epoch": 4.71240134068548,
+      "grad_norm": 0.21695636212825775,
+      "learning_rate": 4.375500457717896e-07,
+      "loss": 0.3494,
       "step": 130755
     },
     {
-      "epoch": 4.6,
-      "learning_rate": 8.391337263502247e-07,
-      "loss": 0.2342,
+      "epoch": 4.712581540346704,
+      "grad_norm": 0.26148954033851624,
+      "learning_rate": 4.3700664317709026e-07,
+      "loss": 0.3834,
       "step": 130760
     },
     {
-      "epoch": 4.6,
-      "learning_rate": 8.384020109121649e-07,
-      "loss": 0.2464,
+      "epoch": 4.712761740007929,
+      "grad_norm": 0.31055474281311035,
+      "learning_rate": 4.364635752484164e-07,
+      "loss": 0.3661,
       "step": 130765
     },
     {
-      "epoch": 4.6,
-      "learning_rate": 8.376706091965309e-07,
-      "loss": 0.2523,
+      "epoch": 4.712941939669154,
+      "grad_norm": 0.32295408844947815,
+      "learning_rate": 4.3592084199317316e-07,
+      "loss": 0.3913,
       "step": 130770
     },
     {
-      "epoch": 4.6,
-      "learning_rate": 8.369395212128206e-07,
-      "loss": 0.2362,
+      "epoch": 4.713122139330378,
+      "grad_norm": 0.239763081073761,
+      "learning_rate": 4.353784434187519e-07,
+      "loss": 0.3283,
       "step": 130775
     },
     {
-      "epoch": 4.6,
-      "learning_rate": 8.362087469705265e-07,
-      "loss": 0.2474,
+      "epoch": 4.7133023389916024,
+      "grad_norm": 0.2873198986053467,
+      "learning_rate": 4.3483637953254384e-07,
+      "loss": 0.3751,
       "step": 130780
     },
     {
-      "epoch": 4.6,
-      "learning_rate": 8.354782864791383e-07,
-      "loss": 0.2297,
+      "epoch": 4.713482538652827,
+      "grad_norm": 0.25636720657348633,
+      "learning_rate": 4.3429465034193485e-07,
+      "loss": 0.3496,
       "step": 130785
     },
     {
-      "epoch": 4.6,
-      "learning_rate": 8.347481397481344e-07,
-      "loss": 0.2467,
+      "epoch": 4.713662738314052,
+      "grad_norm": 0.29088860750198364,
+      "learning_rate": 4.33753255854305e-07,
+      "loss": 0.3963,
       "step": 130790
     },
     {
-      "epoch": 4.6,
-      "learning_rate": 8.340183067870072e-07,
-      "loss": 0.2582,
+      "epoch": 4.713842937975277,
+      "grad_norm": 0.2415892779827118,
+      "learning_rate": 4.3321219607702634e-07,
+      "loss": 0.3655,
       "step": 130795
     },
     {
-      "epoch": 4.6,
-      "learning_rate": 8.332887876052215e-07,
-      "loss": 0.2395,
+      "epoch": 4.714023137636501,
+      "grad_norm": 0.2409953773021698,
+      "learning_rate": 4.3267147101747896e-07,
+      "loss": 0.3836,
       "step": 130800
     },
     {
-      "epoch": 4.6,
-      "learning_rate": 8.325595822122556e-07,
-      "loss": 0.273,
+      "epoch": 4.714203337297726,
+      "grad_norm": 0.2819431722164154,
+      "learning_rate": 4.321310806830292e-07,
+      "loss": 0.3651,
       "step": 130805
     },
     {
-      "epoch": 4.6,
-      "learning_rate": 8.318306906175771e-07,
-      "loss": 0.2424,
+      "epoch": 4.714383536958951,
+      "grad_norm": 0.24938331544399261,
+      "learning_rate": 4.3159102508103233e-07,
+      "loss": 0.3752,
       "step": 130810
     },
     {
-      "epoch": 4.6,
-      "learning_rate": 8.311021128306506e-07,
-      "loss": 0.235,
+      "epoch": 4.714563736620175,
+      "grad_norm": 0.2521766424179077,
+      "learning_rate": 4.3105130421885463e-07,
+      "loss": 0.3826,
       "step": 130815
     },
     {
-      "epoch": 4.6,
-      "learning_rate": 8.303738488609353e-07,
-      "loss": 0.2673,
+      "epoch": 4.714743936281399,
+      "grad_norm": 0.25512444972991943,
+      "learning_rate": 4.3051191810384304e-07,
+      "loss": 0.3638,
       "step": 130820
     },
     {
-      "epoch": 4.6,
-      "learning_rate": 8.296458987178873e-07,
-      "loss": 0.2706,
+      "epoch": 4.714924135942624,
+      "grad_norm": 0.2716846466064453,
+      "learning_rate": 4.299728667433528e-07,
+      "loss": 0.3844,
       "step": 130825
     },
     {
-      "epoch": 4.6,
-      "learning_rate": 8.289182624109549e-07,
-      "loss": 0.275,
+      "epoch": 4.715104335603849,
+      "grad_norm": 0.2582548260688782,
+      "learning_rate": 4.2943415014472523e-07,
+      "loss": 0.3979,
       "step": 130830
     },
     {
-      "epoch": 4.6,
-      "learning_rate": 8.28190939949594e-07,
-      "loss": 0.2573,
+      "epoch": 4.715284535265074,
+      "grad_norm": 0.2618710398674011,
+      "learning_rate": 4.288957683152961e-07,
+      "loss": 0.3923,
       "step": 130835
     },
     {
-      "epoch": 4.6,
-      "learning_rate": 8.274639313432419e-07,
-      "loss": 0.2329,
+      "epoch": 4.715464734926298,
+      "grad_norm": 0.24619565904140472,
+      "learning_rate": 4.283577212624096e-07,
+      "loss": 0.389,
       "step": 130840
     },
     {
-      "epoch": 4.6,
-      "learning_rate": 8.267372366013409e-07,
-      "loss": 0.2559,
+      "epoch": 4.715644934587523,
+      "grad_norm": 0.28034818172454834,
+      "learning_rate": 4.2782000899339035e-07,
+      "loss": 0.3825,
       "step": 130845
     },
     {
-      "epoch": 4.6,
-      "learning_rate": 8.260108557333251e-07,
-      "loss": 0.2578,
+      "epoch": 4.715825134248748,
+      "grad_norm": 0.27038708329200745,
+      "learning_rate": 4.272826315155659e-07,
+      "loss": 0.398,
       "step": 130850
     },
     {
-      "epoch": 4.6,
-      "learning_rate": 8.252847887486287e-07,
-      "loss": 0.2443,
+      "epoch": 4.716005333909973,
+      "grad_norm": 0.2420760691165924,
+      "learning_rate": 4.267455888362581e-07,
+      "loss": 0.387,
       "step": 130855
     },
     {
-      "epoch": 4.6,
-      "learning_rate": 8.245590356566745e-07,
-      "loss": 0.253,
+      "epoch": 4.716185533571197,
+      "grad_norm": 0.26630204916000366,
+      "learning_rate": 4.2620888096278346e-07,
+      "loss": 0.4072,
       "step": 130860
     },
     {
-      "epoch": 4.6,
-      "learning_rate": 8.238335964668914e-07,
-      "loss": 0.2273,
+      "epoch": 4.716365733232421,
+      "grad_norm": 0.2816385328769684,
+      "learning_rate": 4.256725079024554e-07,
+      "loss": 0.372,
       "step": 130865
     },
     {
-      "epoch": 4.6,
-      "learning_rate": 8.231084711886939e-07,
-      "loss": 0.2434,
+      "epoch": 4.716545932893646,
+      "grad_norm": 0.23565801978111267,
+      "learning_rate": 4.25136469662582e-07,
+      "loss": 0.3533,
       "step": 130870
     },
     {
-      "epoch": 4.6,
-      "learning_rate": 8.223836598315049e-07,
-      "loss": 0.2605,
+      "epoch": 4.716726132554871,
+      "grad_norm": 0.2744382321834564,
+      "learning_rate": 4.2460076625046865e-07,
+      "loss": 0.377,
       "step": 130875
     },
     {
-      "epoch": 4.6,
-      "learning_rate": 8.216591624047282e-07,
-      "loss": 0.2579,
+      "epoch": 4.716906332216095,
+      "grad_norm": 0.28961044549942017,
+      "learning_rate": 4.2406539767340934e-07,
+      "loss": 0.3608,
       "step": 130880
     },
     {
-      "epoch": 4.6,
-      "learning_rate": 8.209349789177728e-07,
-      "loss": 0.2405,
+      "epoch": 4.71708653187732,
+      "grad_norm": 0.24323128163814545,
+      "learning_rate": 4.2353036393870116e-07,
+      "loss": 0.3578,
       "step": 130885
     },
     {
-      "epoch": 4.61,
-      "learning_rate": 8.202111093800396e-07,
-      "loss": 0.236,
+      "epoch": 4.717266731538545,
+      "grad_norm": 0.2941163182258606,
+      "learning_rate": 4.2299566505363266e-07,
+      "loss": 0.3813,
       "step": 130890
     },
     {
-      "epoch": 4.61,
-      "learning_rate": 8.194875538009322e-07,
-      "loss": 0.2623,
+      "epoch": 4.7174469311997695,
+      "grad_norm": 0.28495699167251587,
+      "learning_rate": 4.224613010254952e-07,
+      "loss": 0.4055,
       "step": 130895
     },
     {
-      "epoch": 4.61,
-      "learning_rate": 8.187643121898458e-07,
-      "loss": 0.259,
+      "epoch": 4.717627130860994,
+      "grad_norm": 0.36021918058395386,
+      "learning_rate": 4.21927271861558e-07,
+      "loss": 0.3711,
       "step": 130900
     },
     {
-      "epoch": 4.61,
-      "learning_rate": 8.180413845561646e-07,
-      "loss": 0.2347,
+      "epoch": 4.717807330522218,
+      "grad_norm": 0.23534350097179413,
+      "learning_rate": 4.213935775691041e-07,
+      "loss": 0.3555,
       "step": 130905
     },
     {
-      "epoch": 4.61,
-      "learning_rate": 8.173187709092811e-07,
-      "loss": 0.2467,
+      "epoch": 4.717987530183443,
+      "grad_norm": 0.2570185959339142,
+      "learning_rate": 4.2086021815540553e-07,
+      "loss": 0.3603,
       "step": 130910
     },
     {
-      "epoch": 4.61,
-      "learning_rate": 8.165964712585739e-07,
-      "loss": 0.2806,
+      "epoch": 4.718167729844668,
+      "grad_norm": 0.207231804728508,
+      "learning_rate": 4.2032719362772864e-07,
+      "loss": 0.3445,
       "step": 130915
     },
     {
-      "epoch": 4.61,
-      "learning_rate": 8.158744856134271e-07,
-      "loss": 0.2558,
+      "epoch": 4.718347929505892,
+      "grad_norm": 0.24607731401920319,
+      "learning_rate": 4.1979450399333434e-07,
+      "loss": 0.3545,
       "step": 130920
     },
     {
-      "epoch": 4.61,
-      "learning_rate": 8.151528139832109e-07,
-      "loss": 0.25,
+      "epoch": 4.718528129167117,
+      "grad_norm": 0.19284264743328094,
+      "learning_rate": 4.1926214925948057e-07,
+      "loss": 0.3441,
       "step": 130925
     },
     {
-      "epoch": 4.61,
-      "learning_rate": 8.144314563772987e-07,
-      "loss": 0.2682,
+      "epoch": 4.718708328828342,
+      "grad_norm": 0.2840045988559723,
+      "learning_rate": 4.1873012943342007e-07,
+      "loss": 0.3717,
       "step": 130930
     },
     {
-      "epoch": 4.61,
-      "learning_rate": 8.137104128050549e-07,
-      "loss": 0.2685,
+      "epoch": 4.7188885284895665,
+      "grad_norm": 0.26149991154670715,
+      "learning_rate": 4.181984445224052e-07,
+      "loss": 0.352,
       "step": 130935
     },
     {
-      "epoch": 4.61,
-      "learning_rate": 8.129896832758416e-07,
-      "loss": 0.2251,
+      "epoch": 4.719068728150791,
+      "grad_norm": 0.2628280818462372,
+      "learning_rate": 4.176670945336747e-07,
+      "loss": 0.3827,
       "step": 130940
     },
     {
-      "epoch": 4.61,
-      "learning_rate": 8.122692677990151e-07,
-      "loss": 0.2476,
+      "epoch": 4.719248927812016,
+      "grad_norm": 0.29585734009742737,
+      "learning_rate": 4.1713607947447276e-07,
+      "loss": 0.3666,
       "step": 130945
     },
     {
-      "epoch": 4.61,
-      "learning_rate": 8.115491663839348e-07,
-      "loss": 0.2486,
+      "epoch": 4.719429127473241,
+      "grad_norm": 0.26492440700531006,
+      "learning_rate": 4.166053993520297e-07,
+      "loss": 0.3757,
       "step": 130950
     },
     {
-      "epoch": 4.61,
-      "learning_rate": 8.108293790399485e-07,
-      "loss": 0.2653,
+      "epoch": 4.7196093271344655,
+      "grad_norm": 0.24920345842838287,
+      "learning_rate": 4.160750541735814e-07,
+      "loss": 0.3477,
       "step": 130955
     },
     {
-      "epoch": 4.61,
-      "learning_rate": 8.101099057763989e-07,
-      "loss": 0.2658,
+      "epoch": 4.719789526795689,
+      "grad_norm": 0.2741335928440094,
+      "learning_rate": 4.1554504394634984e-07,
+      "loss": 0.3818,
       "step": 130960
     },
     {
-      "epoch": 4.61,
-      "learning_rate": 8.093907466026313e-07,
-      "loss": 0.2512,
+      "epoch": 4.719969726456914,
+      "grad_norm": 0.2235456109046936,
+      "learning_rate": 4.1501536867755706e-07,
+      "loss": 0.3711,
       "step": 130965
     },
     {
-      "epoch": 4.61,
-      "learning_rate": 8.086719015279825e-07,
-      "loss": 0.2522,
+      "epoch": 4.720149926118139,
+      "grad_norm": 0.27255016565322876,
+      "learning_rate": 4.144860283744223e-07,
+      "loss": 0.381,
       "step": 130970
     },
     {
-      "epoch": 4.61,
-      "learning_rate": 8.07953370561787e-07,
-      "loss": 0.2256,
+      "epoch": 4.7203301257793635,
+      "grad_norm": 0.3171219229698181,
+      "learning_rate": 4.1395702304415364e-07,
+      "loss": 0.3889,
       "step": 130975
     },
     {
-      "epoch": 4.61,
-      "learning_rate": 8.072351537133732e-07,
-      "loss": 0.275,
+      "epoch": 4.720510325440588,
+      "grad_norm": 0.2652469873428345,
+      "learning_rate": 4.134283526939592e-07,
+      "loss": 0.3715,
       "step": 130980
     },
     {
-      "epoch": 4.61,
-      "learning_rate": 8.065172509920671e-07,
-      "loss": 0.2418,
+      "epoch": 4.720690525101813,
+      "grad_norm": 0.2219124734401703,
+      "learning_rate": 4.129000173310443e-07,
+      "loss": 0.3463,
       "step": 130985
     },
     {
-      "epoch": 4.61,
-      "learning_rate": 8.057996624071917e-07,
-      "loss": 0.2487,
+      "epoch": 4.720870724763038,
+      "grad_norm": 0.24788452684879303,
+      "learning_rate": 4.123720169626061e-07,
+      "loss": 0.3595,
       "step": 130990
     },
     {
-      "epoch": 4.61,
-      "learning_rate": 8.050823879680619e-07,
-      "loss": 0.2264,
+      "epoch": 4.721050924424262,
+      "grad_norm": 0.21029269695281982,
+      "learning_rate": 4.1184435159584143e-07,
+      "loss": 0.3676,
       "step": 130995
     },
     {
-      "epoch": 4.61,
-      "learning_rate": 8.043654276839951e-07,
-      "loss": 0.2602,
+      "epoch": 4.721231124085486,
+      "grad_norm": 0.28664112091064453,
+      "learning_rate": 4.1131702123793357e-07,
+      "loss": 0.3637,
       "step": 131000
     },
     {
-      "epoch": 4.61,
-      "eval_loss": 0.24853986501693726,
-      "eval_runtime": 10.534,
-      "eval_samples_per_second": 9.493,
-      "eval_steps_per_second": 9.493,
+      "epoch": 4.721231124085486,
+      "eval_loss": 0.42884311079978943,
+      "eval_runtime": 3.5394,
+      "eval_samples_per_second": 28.253,
+      "eval_steps_per_second": 7.063,
       "step": 131000
     },
     {
-      "epoch": 4.61,
-      "learning_rate": 8.036487815642923e-07,
-      "loss": 0.2605,
+      "epoch": 4.721411323746711,
+      "grad_norm": 0.3436447083950043,
+      "learning_rate": 4.107900258960712e-07,
+      "loss": 0.39,
       "step": 131005
     },
     {
-      "epoch": 4.61,
-      "learning_rate": 8.029324496182683e-07,
-      "loss": 0.2415,
+      "epoch": 4.721591523407936,
+      "grad_norm": 0.20727182924747467,
+      "learning_rate": 4.102633655774374e-07,
+      "loss": 0.3699,
       "step": 131010
     },
     {
-      "epoch": 4.61,
-      "learning_rate": 8.022164318552184e-07,
-      "loss": 0.2655,
+      "epoch": 4.7217717230691605,
+      "grad_norm": 0.24989864230155945,
+      "learning_rate": 4.0973704028920423e-07,
+      "loss": 0.3706,
       "step": 131015
     },
     {
-      "epoch": 4.61,
-      "learning_rate": 8.015007282844405e-07,
-      "loss": 0.2641,
+      "epoch": 4.721951922730385,
+      "grad_norm": 0.24569204449653625,
+      "learning_rate": 4.092110500385382e-07,
+      "loss": 0.3611,
       "step": 131020
     },
     {
-      "epoch": 4.61,
-      "learning_rate": 8.007853389152276e-07,
-      "loss": 0.2411,
+      "epoch": 4.72213212239161,
+      "grad_norm": 0.2474668025970459,
+      "learning_rate": 4.0868539483261413e-07,
+      "loss": 0.3665,
       "step": 131025
     },
     {
-      "epoch": 4.61,
-      "learning_rate": 8.000702637568719e-07,
-      "loss": 0.2311,
+      "epoch": 4.722312322052835,
+      "grad_norm": 0.2459608018398285,
+      "learning_rate": 4.081600746785874e-07,
+      "loss": 0.3716,
       "step": 131030
     },
     {
-      "epoch": 4.61,
-      "learning_rate": 7.99355502818655e-07,
-      "loss": 0.2515,
+      "epoch": 4.722492521714059,
+      "grad_norm": 0.2828958034515381,
+      "learning_rate": 4.076350895836245e-07,
+      "loss": 0.3439,
       "step": 131035
     },
     {
-      "epoch": 4.61,
-      "learning_rate": 7.986410561098556e-07,
-      "loss": 0.2698,
+      "epoch": 4.722672721375284,
+      "grad_norm": 0.4490903317928314,
+      "learning_rate": 4.0711043955486974e-07,
+      "loss": 0.3709,
       "step": 131040
     },
     {
-      "epoch": 4.61,
-      "learning_rate": 7.979269236397552e-07,
-      "loss": 0.2797,
+      "epoch": 4.722852921036509,
+      "grad_norm": 0.30186668038368225,
+      "learning_rate": 4.065861245994701e-07,
+      "loss": 0.3289,
       "step": 131045
     },
     {
-      "epoch": 4.61,
-      "learning_rate": 7.97213105417624e-07,
-      "loss": 0.2432,
+      "epoch": 4.723033120697733,
+      "grad_norm": 0.2187676876783371,
+      "learning_rate": 4.0606214472457826e-07,
+      "loss": 0.3377,
       "step": 131050
     },
     {
-      "epoch": 4.61,
-      "learning_rate": 7.964996014527298e-07,
-      "loss": 0.2413,
+      "epoch": 4.7232133203589575,
+      "grad_norm": 0.21422752737998962,
+      "learning_rate": 4.055384999373246e-07,
+      "loss": 0.3654,
       "step": 131055
     },
     {
-      "epoch": 4.61,
-      "learning_rate": 7.957864117543373e-07,
-      "loss": 0.2622,
+      "epoch": 4.723393520020182,
+      "grad_norm": 0.20950256288051605,
+      "learning_rate": 4.050151902448507e-07,
+      "loss": 0.375,
       "step": 131060
     },
     {
-      "epoch": 4.61,
-      "learning_rate": 7.950735363317085e-07,
-      "loss": 0.2344,
+      "epoch": 4.723573719681407,
+      "grad_norm": 0.23349595069885254,
+      "learning_rate": 4.044922156542813e-07,
+      "loss": 0.371,
       "step": 131065
     },
     {
-      "epoch": 4.61,
-      "learning_rate": 7.943609751940972e-07,
-      "loss": 0.2349,
+      "epoch": 4.723753919342632,
+      "grad_norm": 0.2319730669260025,
+      "learning_rate": 4.0396957617274134e-07,
+      "loss": 0.3757,
       "step": 131070
     },
     {
-      "epoch": 4.61,
-      "learning_rate": 7.936487283507543e-07,
-      "loss": 0.2351,
+      "epoch": 4.723934119003856,
+      "grad_norm": 0.22478604316711426,
+      "learning_rate": 4.034472718073556e-07,
+      "loss": 0.3753,
       "step": 131075
     },
     {
-      "epoch": 4.61,
-      "learning_rate": 7.929367958109307e-07,
-      "loss": 0.2361,
+      "epoch": 4.724114318665081,
+      "grad_norm": 0.19446243345737457,
+      "learning_rate": 4.02925302565238e-07,
+      "loss": 0.3447,
       "step": 131080
     },
     {
-      "epoch": 4.61,
-      "learning_rate": 7.922251775838718e-07,
-      "loss": 0.2419,
+      "epoch": 4.724294518326306,
+      "grad_norm": 0.2668735682964325,
+      "learning_rate": 4.0240366845350485e-07,
+      "loss": 0.3687,
       "step": 131085
     },
     {
-      "epoch": 4.61,
-      "learning_rate": 7.915138736788147e-07,
-      "loss": 0.261,
+      "epoch": 4.72447471798753,
+      "grad_norm": 0.22886505722999573,
+      "learning_rate": 4.018823694792562e-07,
+      "loss": 0.3632,
       "step": 131090
     },
     {
-      "epoch": 4.61,
-      "learning_rate": 7.908028841049992e-07,
-      "loss": 0.2685,
+      "epoch": 4.7246549176487544,
+      "grad_norm": 0.25273075699806213,
+      "learning_rate": 4.0136140564959735e-07,
+      "loss": 0.3876,
       "step": 131095
     },
     {
-      "epoch": 4.61,
-      "learning_rate": 7.900922088716483e-07,
-      "loss": 0.2618,
+      "epoch": 4.724835117309979,
+      "grad_norm": 0.2730422914028168,
+      "learning_rate": 4.0084077697162826e-07,
+      "loss": 0.4057,
       "step": 131100
     },
     {
-      "epoch": 4.61,
-      "learning_rate": 7.893818479879966e-07,
-      "loss": 0.2682,
+      "epoch": 4.725015316971204,
+      "grad_norm": 0.21895499527454376,
+      "learning_rate": 4.003204834524432e-07,
+      "loss": 0.3848,
       "step": 131105
     },
     {
-      "epoch": 4.61,
-      "learning_rate": 7.886718014632699e-07,
-      "loss": 0.2335,
+      "epoch": 4.725195516632429,
+      "grad_norm": 0.23920688033103943,
+      "learning_rate": 3.9980052509912267e-07,
+      "loss": 0.3527,
       "step": 131110
     },
     {
-      "epoch": 4.61,
-      "learning_rate": 7.879620693066803e-07,
-      "loss": 0.2395,
+      "epoch": 4.725375716293653,
+      "grad_norm": 0.2694016695022583,
+      "learning_rate": 3.99280901918761e-07,
+      "loss": 0.3571,
       "step": 131115
     },
     {
-      "epoch": 4.61,
-      "learning_rate": 7.872526515274481e-07,
-      "loss": 0.268,
+      "epoch": 4.725555915954878,
+      "grad_norm": 0.2753653824329376,
+      "learning_rate": 3.987616139184358e-07,
+      "loss": 0.3968,
       "step": 131120
     },
     {
-      "epoch": 4.61,
-      "learning_rate": 7.865435481347827e-07,
-      "loss": 0.2562,
+      "epoch": 4.725736115616103,
+      "grad_norm": 0.24241603910923004,
+      "learning_rate": 3.9824266110521924e-07,
+      "loss": 0.3743,
       "step": 131125
     },
     {
-      "epoch": 4.61,
-      "learning_rate": 7.858347591378907e-07,
-      "loss": 0.228,
+      "epoch": 4.725916315277328,
+      "grad_norm": 0.28032630681991577,
+      "learning_rate": 3.977240434861834e-07,
+      "loss": 0.382,
       "step": 131130
     },
     {
-      "epoch": 4.61,
-      "learning_rate": 7.851262845459756e-07,
-      "loss": 0.2396,
+      "epoch": 4.726096514938552,
+      "grad_norm": 0.2554890215396881,
+      "learning_rate": 3.972057610683921e-07,
+      "loss": 0.4077,
       "step": 131135
     },
     {
-      "epoch": 4.61,
-      "learning_rate": 7.844181243682414e-07,
-      "loss": 0.2498,
+      "epoch": 4.726276714599777,
+      "grad_norm": 0.27855169773101807,
+      "learning_rate": 3.9668781385891194e-07,
+      "loss": 0.3812,
       "step": 131140
     },
     {
-      "epoch": 4.61,
-      "learning_rate": 7.837102786138806e-07,
-      "loss": 0.2512,
+      "epoch": 4.726456914261001,
+      "grad_norm": 0.23848845064640045,
+      "learning_rate": 3.9617020186479557e-07,
+      "loss": 0.3784,
       "step": 131145
     },
     {
-      "epoch": 4.61,
-      "learning_rate": 7.830027472920803e-07,
-      "loss": 0.2554,
+      "epoch": 4.726637113922226,
+      "grad_norm": 0.2614748477935791,
+      "learning_rate": 3.956529250930985e-07,
+      "loss": 0.3939,
       "step": 131150
     },
     {
-      "epoch": 4.61,
-      "learning_rate": 7.822955304120277e-07,
-      "loss": 0.2474,
+      "epoch": 4.72681731358345,
+      "grad_norm": 0.22815276682376862,
+      "learning_rate": 3.95135983550865e-07,
+      "loss": 0.3932,
       "step": 131155
     },
     {
-      "epoch": 4.61,
-      "learning_rate": 7.815886279829098e-07,
-      "loss": 0.2653,
+      "epoch": 4.726997513244675,
+      "grad_norm": 0.28664857149124146,
+      "learning_rate": 3.9461937724514233e-07,
+      "loss": 0.369,
       "step": 131160
     },
     {
-      "epoch": 4.61,
-      "learning_rate": 7.808820400139028e-07,
-      "loss": 0.2367,
+      "epoch": 4.7271777129059,
+      "grad_norm": 0.25510892271995544,
+      "learning_rate": 3.941031061829636e-07,
+      "loss": 0.3868,
       "step": 131165
     },
     {
-      "epoch": 4.61,
-      "learning_rate": 7.801757665141824e-07,
-      "loss": 0.2574,
+      "epoch": 4.727357912567125,
+      "grad_norm": 0.227322056889534,
+      "learning_rate": 3.9358717037136786e-07,
+      "loss": 0.357,
       "step": 131170
     },
     {
-      "epoch": 4.62,
-      "learning_rate": 7.794698074929163e-07,
-      "loss": 0.2576,
+      "epoch": 4.727538112228349,
+      "grad_norm": 0.2149222195148468,
+      "learning_rate": 3.930715698173826e-07,
+      "loss": 0.3443,
       "step": 131175
     },
     {
-      "epoch": 4.62,
-      "learning_rate": 7.787641629592752e-07,
-      "loss": 0.2521,
+      "epoch": 4.727718311889574,
+      "grad_norm": 0.2911987900733948,
+      "learning_rate": 3.925563045280328e-07,
+      "loss": 0.385,
       "step": 131180
     },
     {
-      "epoch": 4.62,
-      "learning_rate": 7.780588329224154e-07,
-      "loss": 0.2378,
+      "epoch": 4.727898511550798,
+      "grad_norm": 0.22979766130447388,
+      "learning_rate": 3.9204137451034076e-07,
+      "loss": 0.3557,
       "step": 131185
     },
     {
-      "epoch": 4.62,
-      "learning_rate": 7.773538173915018e-07,
-      "loss": 0.2399,
+      "epoch": 4.728078711212023,
+      "grad_norm": 0.24125829339027405,
+      "learning_rate": 3.9152677977131734e-07,
+      "loss": 0.3766,
       "step": 131190
     },
     {
-      "epoch": 4.62,
-      "learning_rate": 7.766491163756828e-07,
-      "loss": 0.2391,
+      "epoch": 4.728258910873247,
+      "grad_norm": 0.270080029964447,
+      "learning_rate": 3.910125203179821e-07,
+      "loss": 0.3529,
       "step": 131195
     },
     {
-      "epoch": 4.62,
-      "learning_rate": 7.759447298841149e-07,
-      "loss": 0.2725,
+      "epoch": 4.728439110534472,
+      "grad_norm": 0.20879442989826202,
+      "learning_rate": 3.904985961573349e-07,
+      "loss": 0.3869,
       "step": 131200
     },
     {
-      "epoch": 4.62,
-      "learning_rate": 7.75240657925938e-07,
-      "loss": 0.257,
+      "epoch": 4.728619310195697,
+      "grad_norm": 0.23659254610538483,
+      "learning_rate": 3.899850072963784e-07,
+      "loss": 0.3665,
       "step": 131205
     },
     {
-      "epoch": 4.62,
-      "learning_rate": 7.745369005102976e-07,
-      "loss": 0.2435,
+      "epoch": 4.7287995098569215,
+      "grad_norm": 0.23854908347129822,
+      "learning_rate": 3.8947175374211274e-07,
+      "loss": 0.3918,
       "step": 131210
     },
     {
-      "epoch": 4.62,
-      "learning_rate": 7.73833457646328e-07,
-      "loss": 0.247,
+      "epoch": 4.728979709518146,
+      "grad_norm": 0.2165476530790329,
+      "learning_rate": 3.889588355015267e-07,
+      "loss": 0.3698,
       "step": 131215
     },
     {
-      "epoch": 4.62,
-      "learning_rate": 7.731303293431691e-07,
-      "loss": 0.2392,
+      "epoch": 4.729159909179371,
+      "grad_norm": 0.2346785068511963,
+      "learning_rate": 3.8844625258161737e-07,
+      "loss": 0.3508,
       "step": 131220
     },
     {
-      "epoch": 4.62,
-      "learning_rate": 7.72427515609947e-07,
-      "loss": 0.2595,
+      "epoch": 4.729340108840596,
+      "grad_norm": 0.25256285071372986,
+      "learning_rate": 3.879340049893598e-07,
+      "loss": 0.373,
       "step": 131225
     },
     {
-      "epoch": 4.62,
-      "learning_rate": 7.717250164557849e-07,
-      "loss": 0.2273,
+      "epoch": 4.7295203085018205,
+      "grad_norm": 0.22178764641284943,
+      "learning_rate": 3.874220927317346e-07,
+      "loss": 0.3534,
       "step": 131230
     },
     {
-      "epoch": 4.62,
-      "learning_rate": 7.710228318898033e-07,
-      "loss": 0.254,
+      "epoch": 4.729700508163044,
+      "grad_norm": 0.27392688393592834,
+      "learning_rate": 3.8691051581571933e-07,
+      "loss": 0.3442,
       "step": 131235
     },
     {
-      "epoch": 4.62,
-      "learning_rate": 7.703209619211254e-07,
-      "loss": 0.2277,
+      "epoch": 4.729880707824269,
+      "grad_norm": 0.26804807782173157,
+      "learning_rate": 3.8639927424828635e-07,
+      "loss": 0.3818,
       "step": 131240
     },
     {
-      "epoch": 4.62,
-      "learning_rate": 7.696194065588635e-07,
-      "loss": 0.2502,
+      "epoch": 4.730060907485494,
+      "grad_norm": 0.2786375880241394,
+      "learning_rate": 3.8588836803639397e-07,
+      "loss": 0.4149,
       "step": 131245
     },
     {
-      "epoch": 4.62,
-      "learning_rate": 7.689181658121214e-07,
-      "loss": 0.2546,
+      "epoch": 4.7302411071467185,
+      "grad_norm": 0.2680922746658325,
+      "learning_rate": 3.853777971870087e-07,
+      "loss": 0.3972,
       "step": 131250
     },
     {
-      "epoch": 4.62,
-      "learning_rate": 7.682172396900111e-07,
-      "loss": 0.2075,
+      "epoch": 4.730421306807943,
+      "grad_norm": 0.2476295381784439,
+      "learning_rate": 3.848675617070835e-07,
+      "loss": 0.4079,
       "step": 131255
     },
     {
-      "epoch": 4.62,
-      "learning_rate": 7.67516628201631e-07,
-      "loss": 0.2491,
+      "epoch": 4.730601506469168,
+      "grad_norm": 0.2629091143608093,
+      "learning_rate": 3.843576616035738e-07,
+      "loss": 0.3755,
       "step": 131260
     },
     {
-      "epoch": 4.62,
-      "learning_rate": 7.668163313560795e-07,
-      "loss": 0.2517,
+      "epoch": 4.730781706130393,
+      "grad_norm": 0.26332563161849976,
+      "learning_rate": 3.8384809688342684e-07,
+      "loss": 0.3467,
       "step": 131265
     },
     {
-      "epoch": 4.62,
-      "learning_rate": 7.661163491624434e-07,
-      "loss": 0.2478,
+      "epoch": 4.7309619057916175,
+      "grad_norm": 0.21108414232730865,
+      "learning_rate": 3.833388675535815e-07,
+      "loss": 0.3619,
       "step": 131270
     },
     {
-      "epoch": 4.62,
-      "learning_rate": 7.654166816298159e-07,
-      "loss": 0.239,
+      "epoch": 4.731142105452841,
+      "grad_norm": 0.24823954701423645,
+      "learning_rate": 3.828299736209795e-07,
+      "loss": 0.3529,
       "step": 131275
     },
     {
-      "epoch": 4.62,
-      "learning_rate": 7.647173287672865e-07,
-      "loss": 0.2577,
+      "epoch": 4.731322305114066,
+      "grad_norm": 0.261547714471817,
+      "learning_rate": 3.8232141509255137e-07,
+      "loss": 0.3579,
       "step": 131280
     },
     {
-      "epoch": 4.62,
-      "learning_rate": 7.640182905839261e-07,
-      "loss": 0.2768,
+      "epoch": 4.731502504775291,
+      "grad_norm": 0.22427648305892944,
+      "learning_rate": 3.818131919752277e-07,
+      "loss": 0.3685,
       "step": 131285
     },
     {
-      "epoch": 4.62,
-      "learning_rate": 7.633195670888161e-07,
-      "loss": 0.2681,
+      "epoch": 4.7316827044365155,
+      "grad_norm": 0.27016377449035645,
+      "learning_rate": 3.8130530427593626e-07,
+      "loss": 0.3812,
       "step": 131290
     },
     {
-      "epoch": 4.62,
-      "learning_rate": 7.626211582910326e-07,
-      "loss": 0.2349,
+      "epoch": 4.73186290409774,
+      "grad_norm": 0.2747372090816498,
+      "learning_rate": 3.8079775200159096e-07,
+      "loss": 0.3558,
       "step": 131295
     },
     {
-      "epoch": 4.62,
-      "learning_rate": 7.61923064199635e-07,
-      "loss": 0.2773,
+      "epoch": 4.732043103758965,
+      "grad_norm": 0.2099730223417282,
+      "learning_rate": 3.802905351591113e-07,
+      "loss": 0.3624,
       "step": 131300
     },
     {
-      "epoch": 4.62,
-      "learning_rate": 7.612252848236967e-07,
-      "loss": 0.2587,
+      "epoch": 4.73222330342019,
+      "grad_norm": 0.2557778060436249,
+      "learning_rate": 3.797836537554056e-07,
+      "loss": 0.3553,
       "step": 131305
     },
     {
-      "epoch": 4.62,
-      "learning_rate": 7.605278201722715e-07,
-      "loss": 0.2307,
+      "epoch": 4.732403503081414,
+      "grad_norm": 0.26313820481300354,
+      "learning_rate": 3.792771077973795e-07,
+      "loss": 0.3633,
       "step": 131310
     },
     {
-      "epoch": 4.62,
-      "learning_rate": 7.598306702544217e-07,
-      "loss": 0.2553,
+      "epoch": 4.732583702742639,
+      "grad_norm": 0.25919729471206665,
+      "learning_rate": 3.787708972919385e-07,
+      "loss": 0.4063,
       "step": 131315
     },
     {
-      "epoch": 4.62,
-      "learning_rate": 7.591338350791954e-07,
-      "loss": 0.2531,
+      "epoch": 4.732763902403864,
+      "grad_norm": 0.26296132802963257,
+      "learning_rate": 3.7826502224597716e-07,
+      "loss": 0.3512,
       "step": 131320
     },
     {
-      "epoch": 4.62,
-      "learning_rate": 7.584373146556411e-07,
-      "loss": 0.2391,
+      "epoch": 4.732944102065088,
+      "grad_norm": 0.23905587196350098,
+      "learning_rate": 3.7775948266638717e-07,
+      "loss": 0.389,
       "step": 131325
     },
     {
-      "epoch": 4.62,
-      "learning_rate": 7.577411089927989e-07,
-      "loss": 0.2578,
+      "epoch": 4.7331243017263125,
+      "grad_norm": 0.21689580380916595,
+      "learning_rate": 3.7725427856005747e-07,
+      "loss": 0.3284,
       "step": 131330
     },
     {
-      "epoch": 4.62,
-      "learning_rate": 7.570452180997139e-07,
-      "loss": 0.2569,
+      "epoch": 4.733304501387537,
+      "grad_norm": 0.28716617822647095,
+      "learning_rate": 3.767494099338714e-07,
+      "loss": 0.3821,
       "step": 131335
     },
     {
-      "epoch": 4.62,
-      "learning_rate": 7.56349641985421e-07,
-      "loss": 0.2361,
+      "epoch": 4.733484701048762,
+      "grad_norm": 0.21327391266822815,
+      "learning_rate": 3.762448767947069e-07,
+      "loss": 0.3663,
       "step": 131340
     },
     {
-      "epoch": 4.62,
-      "learning_rate": 7.556543806589489e-07,
-      "loss": 0.2516,
+      "epoch": 4.733664900709987,
+      "grad_norm": 0.2822578251361847,
+      "learning_rate": 3.757406791494389e-07,
+      "loss": 0.3799,
       "step": 131345
     },
     {
-      "epoch": 4.62,
-      "learning_rate": 7.549594341293292e-07,
-      "loss": 0.2621,
+      "epoch": 4.733845100371211,
+      "grad_norm": 0.23147007822990417,
+      "learning_rate": 3.752368170049342e-07,
+      "loss": 0.3403,
       "step": 131350
     },
     {
-      "epoch": 4.62,
-      "learning_rate": 7.542648024055826e-07,
-      "loss": 0.2666,
+      "epoch": 4.734025300032436,
+      "grad_norm": 0.2217051088809967,
+      "learning_rate": 3.747332903680623e-07,
+      "loss": 0.3961,
       "step": 131355
     },
     {
-      "epoch": 4.62,
-      "learning_rate": 7.535704854967296e-07,
-      "loss": 0.2504,
+      "epoch": 4.734205499693661,
+      "grad_norm": 0.23157884180545807,
+      "learning_rate": 3.742300992456843e-07,
+      "loss": 0.3447,
       "step": 131360
     },
     {
-      "epoch": 4.62,
-      "learning_rate": 7.528764834117824e-07,
-      "loss": 0.235,
+      "epoch": 4.734385699354885,
+      "grad_norm": 0.3138419985771179,
+      "learning_rate": 3.737272436446504e-07,
+      "loss": 0.4007,
       "step": 131365
     },
     {
-      "epoch": 4.62,
-      "learning_rate": 7.521827961597561e-07,
-      "loss": 0.2512,
+      "epoch": 4.7345658990161095,
+      "grad_norm": 0.2702825963497162,
+      "learning_rate": 3.7322472357181615e-07,
+      "loss": 0.3663,
       "step": 131370
     },
     {
-      "epoch": 4.62,
-      "learning_rate": 7.514894237496573e-07,
-      "loss": 0.2873,
+      "epoch": 4.734746098677334,
+      "grad_norm": 0.22164510190486908,
+      "learning_rate": 3.7272253903402607e-07,
+      "loss": 0.328,
       "step": 131375
     },
     {
-      "epoch": 4.62,
-      "learning_rate": 7.507963661904871e-07,
-      "loss": 0.2377,
+      "epoch": 4.734926298338559,
+      "grad_norm": 0.267595499753952,
+      "learning_rate": 3.722206900381248e-07,
+      "loss": 0.368,
       "step": 131380
     },
     {
-      "epoch": 4.62,
-      "learning_rate": 7.50103623491244e-07,
-      "loss": 0.2497,
+      "epoch": 4.735106497999784,
+      "grad_norm": 0.2832372486591339,
+      "learning_rate": 3.7171917659094834e-07,
+      "loss": 0.3878,
       "step": 131385
     },
     {
-      "epoch": 4.62,
-      "learning_rate": 7.494111956609262e-07,
-      "loss": 0.2483,
+      "epoch": 4.735286697661008,
+      "grad_norm": 0.22526337206363678,
+      "learning_rate": 3.7121799869933026e-07,
+      "loss": 0.3577,
       "step": 131390
     },
     {
-      "epoch": 4.62,
-      "learning_rate": 7.48719082708521e-07,
-      "loss": 0.2678,
+      "epoch": 4.735466897322233,
+      "grad_norm": 0.3088375926017761,
+      "learning_rate": 3.707171563700956e-07,
+      "loss": 0.4084,
       "step": 131395
     },
     {
-      "epoch": 4.62,
-      "learning_rate": 7.480272846430159e-07,
-      "loss": 0.2581,
+      "epoch": 4.735647096983458,
+      "grad_norm": 0.3013952374458313,
+      "learning_rate": 3.702166496100723e-07,
+      "loss": 0.3899,
       "step": 131400
     },
     {
-      "epoch": 4.62,
-      "learning_rate": 7.473358014733949e-07,
-      "loss": 0.2626,
+      "epoch": 4.735827296644683,
+      "grad_norm": 0.24282796680927277,
+      "learning_rate": 3.697164784260826e-07,
+      "loss": 0.3846,
       "step": 131405
     },
     {
-      "epoch": 4.62,
-      "learning_rate": 7.466446332086347e-07,
-      "loss": 0.2685,
+      "epoch": 4.736007496305907,
+      "grad_norm": 0.21715568006038666,
+      "learning_rate": 3.6921664282493493e-07,
+      "loss": 0.345,
       "step": 131410
     },
     {
-      "epoch": 4.62,
-      "learning_rate": 7.459537798577109e-07,
-      "loss": 0.2595,
+      "epoch": 4.736187695967132,
+      "grad_norm": 0.23071317374706268,
+      "learning_rate": 3.687171428134406e-07,
+      "loss": 0.3733,
       "step": 131415
     },
     {
-      "epoch": 4.62,
-      "learning_rate": 7.452632414295946e-07,
-      "loss": 0.2587,
+      "epoch": 4.736367895628356,
+      "grad_norm": 0.22580982744693756,
+      "learning_rate": 3.68217978398408e-07,
+      "loss": 0.3632,
       "step": 131420
     },
     {
-      "epoch": 4.62,
-      "learning_rate": 7.445730179332478e-07,
-      "loss": 0.2841,
+      "epoch": 4.736548095289581,
+      "grad_norm": 0.2745913863182068,
+      "learning_rate": 3.6771914958663723e-07,
+      "loss": 0.3948,
       "step": 131425
     },
     {
-      "epoch": 4.62,
-      "learning_rate": 7.438831093776383e-07,
-      "loss": 0.2513,
+      "epoch": 4.736728294950805,
+      "grad_norm": 0.22211065888404846,
+      "learning_rate": 3.6722065638492287e-07,
+      "loss": 0.3529,
       "step": 131430
     },
     {
-      "epoch": 4.62,
-      "learning_rate": 7.431935157717201e-07,
-      "loss": 0.2596,
+      "epoch": 4.73690849461203,
+      "grad_norm": 0.2897282838821411,
+      "learning_rate": 3.667224988000567e-07,
+      "loss": 0.3725,
       "step": 131435
     },
     {
-      "epoch": 4.62,
-      "learning_rate": 7.425042371244472e-07,
-      "loss": 0.2492,
+      "epoch": 4.737088694273255,
+      "grad_norm": 0.2873408794403076,
+      "learning_rate": 3.6622467683882777e-07,
+      "loss": 0.3692,
       "step": 131440
     },
     {
-      "epoch": 4.62,
-      "learning_rate": 7.418152734447708e-07,
-      "loss": 0.2612,
+      "epoch": 4.73726889393448,
+      "grad_norm": 0.24957486987113953,
+      "learning_rate": 3.657271905080195e-07,
+      "loss": 0.3915,
       "step": 131445
     },
     {
-      "epoch": 4.62,
-      "learning_rate": 7.411266247416365e-07,
-      "loss": 0.2343,
+      "epoch": 4.737449093595704,
+      "grad_norm": 0.24464282393455505,
+      "learning_rate": 3.652300398144098e-07,
+      "loss": 0.3876,
       "step": 131450
     },
     {
-      "epoch": 4.62,
-      "learning_rate": 7.40438291023987e-07,
-      "loss": 0.244,
+      "epoch": 4.737629293256929,
+      "grad_norm": 0.3088054358959198,
+      "learning_rate": 3.647332247647711e-07,
+      "loss": 0.3674,
       "step": 131455
     },
     {
-      "epoch": 4.63,
-      "learning_rate": 7.397502723007571e-07,
-      "loss": 0.254,
+      "epoch": 4.737809492918153,
+      "grad_norm": 0.26032140851020813,
+      "learning_rate": 3.6423674536587015e-07,
+      "loss": 0.387,
       "step": 131460
     },
     {
-      "epoch": 4.63,
-      "learning_rate": 7.390625685808811e-07,
-      "loss": 0.2319,
+      "epoch": 4.737989692579378,
+      "grad_norm": 0.2570837736129761,
+      "learning_rate": 3.6374060162447653e-07,
+      "loss": 0.3837,
       "step": 131465
     },
     {
-      "epoch": 4.63,
-      "learning_rate": 7.383751798732907e-07,
-      "loss": 0.2528,
+      "epoch": 4.738169892240602,
+      "grad_norm": 0.19201694428920746,
+      "learning_rate": 3.6324479354734595e-07,
+      "loss": 0.3706,
       "step": 131470
     },
     {
-      "epoch": 4.63,
-      "learning_rate": 7.376881061869096e-07,
-      "loss": 0.2371,
+      "epoch": 4.738350091901827,
+      "grad_norm": 0.28280356526374817,
+      "learning_rate": 3.627493211412342e-07,
+      "loss": 0.4057,
       "step": 131475
     },
     {
-      "epoch": 4.63,
-      "learning_rate": 7.370013475306581e-07,
-      "loss": 0.2413,
+      "epoch": 4.738530291563052,
+      "grad_norm": 0.26518356800079346,
+      "learning_rate": 3.6225418441289684e-07,
+      "loss": 0.4072,
       "step": 131480
     },
     {
-      "epoch": 4.63,
-      "learning_rate": 7.363149039134542e-07,
-      "loss": 0.2594,
+      "epoch": 4.738710491224277,
+      "grad_norm": 0.261747270822525,
+      "learning_rate": 3.61759383369073e-07,
+      "loss": 0.3822,
       "step": 131485
     },
     {
-      "epoch": 4.63,
-      "learning_rate": 7.356287753442131e-07,
-      "loss": 0.269,
+      "epoch": 4.738890690885501,
+      "grad_norm": 0.2915024757385254,
+      "learning_rate": 3.612649180165101e-07,
+      "loss": 0.3657,
       "step": 131490
     },
     {
-      "epoch": 4.63,
-      "learning_rate": 7.349429618318416e-07,
-      "loss": 0.2414,
+      "epoch": 4.739070890546726,
+      "grad_norm": 0.2875518202781677,
+      "learning_rate": 3.607707883619388e-07,
+      "loss": 0.3783,
       "step": 131495
     },
     {
-      "epoch": 4.63,
-      "learning_rate": 7.342574633852406e-07,
-      "loss": 0.2318,
+      "epoch": 4.739251090207951,
+      "grad_norm": 0.2599601745605469,
+      "learning_rate": 3.6027699441209815e-07,
+      "loss": 0.3707,
       "step": 131500
     },
     {
-      "epoch": 4.63,
-      "eval_loss": 0.24853749573230743,
-      "eval_runtime": 10.5675,
-      "eval_samples_per_second": 9.463,
-      "eval_steps_per_second": 9.463,
+      "epoch": 4.739251090207951,
+      "eval_loss": 0.42884549498558044,
+      "eval_runtime": 3.5372,
+      "eval_samples_per_second": 28.271,
+      "eval_steps_per_second": 7.068,
       "step": 131500
     },
     {
-      "epoch": 4.63,
-      "learning_rate": 7.335722800133171e-07,
-      "loss": 0.2536,
+      "epoch": 4.7394312898691755,
+      "grad_norm": 0.23225200176239014,
+      "learning_rate": 3.5978353617371065e-07,
+      "loss": 0.3767,
       "step": 131505
     },
     {
-      "epoch": 4.63,
-      "learning_rate": 7.328874117249668e-07,
-      "loss": 0.2504,
+      "epoch": 4.739611489530399,
+      "grad_norm": 0.2974475920200348,
+      "learning_rate": 3.5929041365350413e-07,
+      "loss": 0.3841,
       "step": 131510
     },
     {
-      "epoch": 4.63,
-      "learning_rate": 7.322028585290768e-07,
-      "loss": 0.2521,
+      "epoch": 4.739791689191624,
+      "grad_norm": 0.28682082891464233,
+      "learning_rate": 3.587976268581955e-07,
+      "loss": 0.3533,
       "step": 131515
     },
     {
-      "epoch": 4.63,
-      "learning_rate": 7.315186204345403e-07,
-      "loss": 0.2433,
+      "epoch": 4.739971888852849,
+      "grad_norm": 0.2352001816034317,
+      "learning_rate": 3.5830517579449605e-07,
+      "loss": 0.3648,
       "step": 131520
     },
     {
-      "epoch": 4.63,
-      "learning_rate": 7.308346974502444e-07,
-      "loss": 0.2541,
+      "epoch": 4.7401520885140735,
+      "grad_norm": 0.25316321849823,
+      "learning_rate": 3.5781306046911986e-07,
+      "loss": 0.3353,
       "step": 131525
     },
     {
-      "epoch": 4.63,
-      "learning_rate": 7.301510895850655e-07,
-      "loss": 0.2189,
+      "epoch": 4.740332288175298,
+      "grad_norm": 0.26786184310913086,
+      "learning_rate": 3.573212808887699e-07,
+      "loss": 0.3874,
       "step": 131530
     },
     {
-      "epoch": 4.63,
-      "learning_rate": 7.294677968478797e-07,
-      "loss": 0.2477,
+      "epoch": 4.740512487836523,
+      "grad_norm": 0.27805742621421814,
+      "learning_rate": 3.568298370601436e-07,
+      "loss": 0.4236,
       "step": 131535
     },
     {
-      "epoch": 4.63,
-      "learning_rate": 7.287848192475577e-07,
-      "loss": 0.234,
+      "epoch": 4.740692687497748,
+      "grad_norm": 0.3236870765686035,
+      "learning_rate": 3.5633872898994113e-07,
+      "loss": 0.4053,
       "step": 131540
     },
     {
-      "epoch": 4.63,
-      "learning_rate": 7.281021567929703e-07,
-      "loss": 0.2328,
+      "epoch": 4.7408728871589725,
+      "grad_norm": 0.2795065641403198,
+      "learning_rate": 3.5584795668485435e-07,
+      "loss": 0.3755,
       "step": 131545
     },
     {
-      "epoch": 4.63,
-      "learning_rate": 7.274198094929824e-07,
-      "loss": 0.2481,
+      "epoch": 4.741053086820196,
+      "grad_norm": 0.23501679301261902,
+      "learning_rate": 3.553575201515669e-07,
+      "loss": 0.3576,
       "step": 131550
     },
     {
-      "epoch": 4.63,
-      "learning_rate": 7.267377773564509e-07,
-      "loss": 0.2676,
+      "epoch": 4.741233286481421,
+      "grad_norm": 0.2597077190876007,
+      "learning_rate": 3.5486741939676215e-07,
+      "loss": 0.3755,
       "step": 131555
     },
     {
-      "epoch": 4.63,
-      "learning_rate": 7.2605606039223e-07,
-      "loss": 0.2516,
+      "epoch": 4.741413486142646,
+      "grad_norm": 0.2638082504272461,
+      "learning_rate": 3.5437765442711545e-07,
+      "loss": 0.3753,
       "step": 131560
     },
     {
-      "epoch": 4.63,
-      "learning_rate": 7.253746586091792e-07,
-      "loss": 0.2411,
+      "epoch": 4.7415936858038705,
+      "grad_norm": 0.32323339581489563,
+      "learning_rate": 3.53888225249302e-07,
+      "loss": 0.3974,
       "step": 131565
     },
     {
-      "epoch": 4.63,
-      "learning_rate": 7.246935720161358e-07,
-      "loss": 0.2535,
+      "epoch": 4.741773885465095,
+      "grad_norm": 0.26547950506210327,
+      "learning_rate": 3.5339913186999416e-07,
+      "loss": 0.3688,
       "step": 131570
     },
     {
-      "epoch": 4.63,
-      "learning_rate": 7.24012800621951e-07,
-      "loss": 0.2354,
+      "epoch": 4.74195408512632,
+      "grad_norm": 0.3076135516166687,
+      "learning_rate": 3.5291037429584784e-07,
+      "loss": 0.3674,
       "step": 131575
     },
     {
-      "epoch": 4.63,
-      "learning_rate": 7.233323444354595e-07,
-      "loss": 0.2391,
+      "epoch": 4.742134284787545,
+      "grad_norm": 0.2687076926231384,
+      "learning_rate": 3.5242195253352426e-07,
+      "loss": 0.335,
       "step": 131580
     },
     {
-      "epoch": 4.63,
-      "learning_rate": 7.226522034655015e-07,
-      "loss": 0.2697,
+      "epoch": 4.7423144844487695,
+      "grad_norm": 0.23814557492733002,
+      "learning_rate": 3.519338665896821e-07,
+      "loss": 0.3682,
       "step": 131585
     },
     {
-      "epoch": 4.63,
-      "learning_rate": 7.219723777209058e-07,
-      "loss": 0.277,
+      "epoch": 4.742494684109994,
+      "grad_norm": 0.21820604801177979,
+      "learning_rate": 3.514461164709687e-07,
+      "loss": 0.3579,
       "step": 131590
     },
     {
-      "epoch": 4.63,
-      "learning_rate": 7.212928672104963e-07,
-      "loss": 0.2566,
+      "epoch": 4.742674883771219,
+      "grad_norm": 0.23697353899478912,
+      "learning_rate": 3.509587021840316e-07,
+      "loss": 0.337,
       "step": 131595
     },
     {
-      "epoch": 4.63,
-      "learning_rate": 7.206136719430989e-07,
-      "loss": 0.2366,
+      "epoch": 4.742855083432443,
+      "grad_norm": 0.36326366662979126,
+      "learning_rate": 3.5047162373550433e-07,
+      "loss": 0.4014,
       "step": 131600
     },
     {
-      "epoch": 4.63,
-      "learning_rate": 7.199347919275345e-07,
-      "loss": 0.2788,
+      "epoch": 4.7430352830936675,
+      "grad_norm": 0.2886950373649597,
+      "learning_rate": 3.499848811320344e-07,
+      "loss": 0.3733,
       "step": 131605
     },
     {
-      "epoch": 4.63,
-      "learning_rate": 7.192562271726155e-07,
-      "loss": 0.2382,
+      "epoch": 4.743215482754892,
+      "grad_norm": 0.2526874840259552,
+      "learning_rate": 3.494984743802443e-07,
+      "loss": 0.3709,
       "step": 131610
     },
     {
-      "epoch": 4.63,
-      "learning_rate": 7.185779776871487e-07,
-      "loss": 0.2471,
+      "epoch": 4.743395682416117,
+      "grad_norm": 0.3060406744480133,
+      "learning_rate": 3.4901240348676756e-07,
+      "loss": 0.368,
       "step": 131615
     },
     {
-      "epoch": 4.63,
-      "learning_rate": 7.179000434799493e-07,
-      "loss": 0.2557,
+      "epoch": 4.743575882077342,
+      "grad_norm": 0.23225359618663788,
+      "learning_rate": 3.485266684582239e-07,
+      "loss": 0.3697,
       "step": 131620
     },
     {
-      "epoch": 4.63,
-      "learning_rate": 7.172224245598102e-07,
-      "loss": 0.2397,
+      "epoch": 4.743756081738566,
+      "grad_norm": 0.2793751060962677,
+      "learning_rate": 3.4804126930123026e-07,
+      "loss": 0.3779,
       "step": 131625
     },
     {
-      "epoch": 4.63,
-      "learning_rate": 7.165451209355411e-07,
-      "loss": 0.2317,
+      "epoch": 4.743936281399791,
+      "grad_norm": 0.25954195857048035,
+      "learning_rate": 3.4755620602240357e-07,
+      "loss": 0.3558,
       "step": 131630
     },
     {
-      "epoch": 4.63,
-      "learning_rate": 7.158681326159239e-07,
-      "loss": 0.2725,
+      "epoch": 4.744116481061016,
+      "grad_norm": 0.224226713180542,
+      "learning_rate": 3.4707147862834965e-07,
+      "loss": 0.3742,
       "step": 131635
     },
     {
-      "epoch": 4.63,
-      "learning_rate": 7.151914596097597e-07,
-      "loss": 0.2184,
+      "epoch": 4.74429668072224,
+      "grad_norm": 0.22823956608772278,
+      "learning_rate": 3.4658708712567433e-07,
+      "loss": 0.3546,
       "step": 131640
     },
     {
-      "epoch": 4.63,
-      "learning_rate": 7.145151019258306e-07,
-      "loss": 0.249,
+      "epoch": 4.7444768803834645,
+      "grad_norm": 0.27476179599761963,
+      "learning_rate": 3.461030315209779e-07,
+      "loss": 0.3922,
       "step": 131645
     },
     {
-      "epoch": 4.63,
-      "learning_rate": 7.138390595729183e-07,
-      "loss": 0.2474,
+      "epoch": 4.744657080044689,
+      "grad_norm": 0.26549628376960754,
+      "learning_rate": 3.4561931182085225e-07,
+      "loss": 0.3772,
       "step": 131650
     },
     {
-      "epoch": 4.63,
-      "learning_rate": 7.131633325597964e-07,
-      "loss": 0.2322,
+      "epoch": 4.744837279705914,
+      "grad_norm": 0.29317188262939453,
+      "learning_rate": 3.4513592803188944e-07,
+      "loss": 0.3897,
       "step": 131655
     },
     {
-      "epoch": 4.63,
-      "learning_rate": 7.124879208952468e-07,
-      "loss": 0.2616,
+      "epoch": 4.745017479367139,
+      "grad_norm": 0.2582772970199585,
+      "learning_rate": 3.446528801606785e-07,
+      "loss": 0.3766,
       "step": 131660
     },
     {
-      "epoch": 4.63,
-      "learning_rate": 7.118128245880346e-07,
-      "loss": 0.2338,
+      "epoch": 4.745197679028363,
+      "grad_norm": 0.2573517858982086,
+      "learning_rate": 3.4417016821379765e-07,
+      "loss": 0.3822,
       "step": 131665
     },
     {
-      "epoch": 4.63,
-      "learning_rate": 7.111380436469279e-07,
-      "loss": 0.2484,
+      "epoch": 4.745377878689588,
+      "grad_norm": 0.2170310914516449,
+      "learning_rate": 3.436877921978249e-07,
+      "loss": 0.3824,
       "step": 131670
     },
     {
-      "epoch": 4.63,
-      "learning_rate": 7.104635780806834e-07,
-      "loss": 0.2542,
+      "epoch": 4.745558078350813,
+      "grad_norm": 0.2496137022972107,
+      "learning_rate": 3.432057521193327e-07,
+      "loss": 0.3849,
       "step": 131675
     },
     {
-      "epoch": 4.63,
-      "learning_rate": 7.09789427898061e-07,
-      "loss": 0.2621,
+      "epoch": 4.745738278012038,
+      "grad_norm": 0.285085529088974,
+      "learning_rate": 3.427240479848881e-07,
+      "loss": 0.3786,
       "step": 131680
     },
     {
-      "epoch": 4.63,
-      "learning_rate": 7.091155931078175e-07,
-      "loss": 0.2529,
+      "epoch": 4.745918477673262,
+      "grad_norm": 0.2879694104194641,
+      "learning_rate": 3.4224267980105807e-07,
+      "loss": 0.3835,
       "step": 131685
     },
     {
-      "epoch": 4.63,
-      "learning_rate": 7.084420737187014e-07,
-      "loss": 0.2661,
+      "epoch": 4.746098677334487,
+      "grad_norm": 0.26048213243484497,
+      "learning_rate": 3.417616475743929e-07,
+      "loss": 0.3415,
       "step": 131690
     },
     {
-      "epoch": 4.63,
-      "learning_rate": 7.07768869739453e-07,
-      "loss": 0.2306,
+      "epoch": 4.746278876995711,
+      "grad_norm": 0.29208385944366455,
+      "learning_rate": 3.412809513114512e-07,
+      "loss": 0.362,
       "step": 131695
     },
     {
-      "epoch": 4.63,
-      "learning_rate": 7.070959811788181e-07,
-      "loss": 0.2482,
+      "epoch": 4.746459076656936,
+      "grad_norm": 0.23016516864299774,
+      "learning_rate": 3.4080059101878614e-07,
+      "loss": 0.3496,
       "step": 131700
     },
     {
-      "epoch": 4.63,
-      "learning_rate": 7.06423408045534e-07,
-      "loss": 0.2536,
+      "epoch": 4.74663927631816,
+      "grad_norm": 0.22892868518829346,
+      "learning_rate": 3.403205667029369e-07,
+      "loss": 0.37,
       "step": 131705
     },
     {
-      "epoch": 4.63,
-      "learning_rate": 7.057511503483299e-07,
-      "loss": 0.2493,
+      "epoch": 4.746819475979385,
+      "grad_norm": 0.2265908122062683,
+      "learning_rate": 3.398408783704454e-07,
+      "loss": 0.3758,
       "step": 131710
     },
     {
-      "epoch": 4.63,
-      "learning_rate": 7.050792080959351e-07,
-      "loss": 0.2561,
+      "epoch": 4.74699967564061,
+      "grad_norm": 0.22055870294570923,
+      "learning_rate": 3.393615260278482e-07,
+      "loss": 0.4055,
       "step": 131715
     },
     {
-      "epoch": 4.63,
-      "learning_rate": 7.044075812970785e-07,
-      "loss": 0.261,
+      "epoch": 4.747179875301835,
+      "grad_norm": 0.22395636141300201,
+      "learning_rate": 3.3888250968167335e-07,
+      "loss": 0.4186,
       "step": 131720
     },
     {
-      "epoch": 4.63,
-      "learning_rate": 7.037362699604783e-07,
-      "loss": 0.2364,
+      "epoch": 4.747360074963059,
+      "grad_norm": 0.24537676572799683,
+      "learning_rate": 3.3840382933845174e-07,
+      "loss": 0.3678,
       "step": 131725
     },
     {
-      "epoch": 4.63,
-      "learning_rate": 7.030652740948496e-07,
-      "loss": 0.2581,
+      "epoch": 4.747540274624284,
+      "grad_norm": 0.28544852137565613,
+      "learning_rate": 3.379254850047031e-07,
+      "loss": 0.3729,
       "step": 131730
     },
     {
-      "epoch": 4.63,
-      "learning_rate": 7.023945937089077e-07,
-      "loss": 0.2381,
+      "epoch": 4.747720474285508,
+      "grad_norm": 0.2334669977426529,
+      "learning_rate": 3.3744747668694456e-07,
+      "loss": 0.366,
       "step": 131735
     },
     {
-      "epoch": 4.63,
-      "learning_rate": 7.01724228811354e-07,
-      "loss": 0.2489,
+      "epoch": 4.747900673946733,
+      "grad_norm": 0.26452332735061646,
+      "learning_rate": 3.369698043916902e-07,
+      "loss": 0.3512,
       "step": 131740
     },
     {
-      "epoch": 4.64,
-      "learning_rate": 7.010541794109038e-07,
-      "loss": 0.2608,
+      "epoch": 4.748080873607957,
+      "grad_norm": 0.2582310140132904,
+      "learning_rate": 3.364924681254461e-07,
+      "loss": 0.3748,
       "step": 131745
     },
     {
-      "epoch": 4.64,
-      "learning_rate": 7.003844455162473e-07,
-      "loss": 0.2626,
+      "epoch": 4.748261073269182,
+      "grad_norm": 0.23445448279380798,
+      "learning_rate": 3.36015467894718e-07,
+      "loss": 0.3843,
       "step": 131750
     },
     {
-      "epoch": 4.64,
-      "learning_rate": 6.997150271360886e-07,
-      "loss": 0.2393,
+      "epoch": 4.748441272930407,
+      "grad_norm": 0.25989097356796265,
+      "learning_rate": 3.355388037060064e-07,
+      "loss": 0.3922,
       "step": 131755
     },
     {
-      "epoch": 4.64,
-      "learning_rate": 6.990459242791125e-07,
-      "loss": 0.2539,
+      "epoch": 4.748621472591632,
+      "grad_norm": 0.24597260355949402,
+      "learning_rate": 3.350624755658005e-07,
+      "loss": 0.3349,
       "step": 131760
     },
     {
-      "epoch": 4.64,
-      "learning_rate": 6.983771369540148e-07,
-      "loss": 0.2579,
+      "epoch": 4.748801672252856,
+      "grad_norm": 0.2952145040035248,
+      "learning_rate": 3.345864834805951e-07,
+      "loss": 0.3894,
       "step": 131765
     },
     {
-      "epoch": 4.64,
-      "learning_rate": 6.977086651694692e-07,
-      "loss": 0.2256,
+      "epoch": 4.748981871914081,
+      "grad_norm": 0.25986775755882263,
+      "learning_rate": 3.341108274568683e-07,
+      "loss": 0.3596,
       "step": 131770
     },
     {
-      "epoch": 4.64,
-      "learning_rate": 6.97040508934163e-07,
-      "loss": 0.2416,
+      "epoch": 4.749162071575306,
+      "grad_norm": 0.2798271179199219,
+      "learning_rate": 3.336355075011122e-07,
+      "loss": 0.3774,
       "step": 131775
     },
     {
-      "epoch": 4.64,
-      "learning_rate": 6.963726682567701e-07,
-      "loss": 0.2481,
+      "epoch": 4.7493422712365305,
+      "grad_norm": 0.25875961780548096,
+      "learning_rate": 3.3316052361979655e-07,
+      "loss": 0.3871,
       "step": 131780
     },
     {
-      "epoch": 4.64,
-      "learning_rate": 6.95705143145961e-07,
-      "loss": 0.2529,
+      "epoch": 4.749522470897754,
+      "grad_norm": 0.32598257064819336,
+      "learning_rate": 3.326858758193885e-07,
+      "loss": 0.394,
       "step": 131785
     },
     {
-      "epoch": 4.64,
-      "learning_rate": 6.95037933610404e-07,
-      "loss": 0.2361,
+      "epoch": 4.749702670558979,
+      "grad_norm": 0.2517693042755127,
+      "learning_rate": 3.3221156410636335e-07,
+      "loss": 0.3926,
       "step": 131790
     },
     {
-      "epoch": 4.64,
-      "learning_rate": 6.943710396587589e-07,
-      "loss": 0.2571,
+      "epoch": 4.749882870220204,
+      "grad_norm": 0.21797238290309906,
+      "learning_rate": 3.3173758848717705e-07,
+      "loss": 0.3232,
       "step": 131795
     },
     {
-      "epoch": 4.64,
-      "learning_rate": 6.937044612996906e-07,
-      "loss": 0.2358,
+      "epoch": 4.750063069881429,
+      "grad_norm": 0.1989823281764984,
+      "learning_rate": 3.312639489682939e-07,
+      "loss": 0.3844,
       "step": 131800
     },
     {
-      "epoch": 4.64,
-      "learning_rate": 6.93038198541851e-07,
-      "loss": 0.2303,
+      "epoch": 4.750243269542653,
+      "grad_norm": 0.22494913637638092,
+      "learning_rate": 3.307906455561616e-07,
+      "loss": 0.3855,
       "step": 131805
     },
     {
-      "epoch": 4.64,
-      "learning_rate": 6.923722513938912e-07,
-      "loss": 0.2525,
+      "epoch": 4.750423469203878,
+      "grad_norm": 0.22416415810585022,
+      "learning_rate": 3.303176782572276e-07,
+      "loss": 0.3918,
       "step": 131810
     },
     {
-      "epoch": 4.64,
-      "learning_rate": 6.917066198644601e-07,
-      "loss": 0.2429,
+      "epoch": 4.750603668865103,
+      "grad_norm": 0.2782622277736664,
+      "learning_rate": 3.2984504707794237e-07,
+      "loss": 0.3599,
       "step": 131815
     },
     {
-      "epoch": 4.64,
-      "learning_rate": 6.910413039621977e-07,
-      "loss": 0.2485,
+      "epoch": 4.7507838685263275,
+      "grad_norm": 0.23325783014297485,
+      "learning_rate": 3.2937275202473974e-07,
+      "loss": 0.3743,
       "step": 131820
     },
     {
-      "epoch": 4.64,
-      "learning_rate": 6.903763036957445e-07,
-      "loss": 0.2254,
+      "epoch": 4.750964068187551,
+      "grad_norm": 0.2581305503845215,
+      "learning_rate": 3.2890079310405885e-07,
+      "loss": 0.3498,
       "step": 131825
     },
     {
-      "epoch": 4.64,
-      "learning_rate": 6.897116190737296e-07,
-      "loss": 0.2648,
+      "epoch": 4.751144267848776,
+      "grad_norm": 0.30278897285461426,
+      "learning_rate": 3.2842917032233076e-07,
+      "loss": 0.3916,
       "step": 131830
     },
     {
-      "epoch": 4.64,
-      "learning_rate": 6.890472501047934e-07,
-      "loss": 0.2625,
+      "epoch": 4.751324467510001,
+      "grad_norm": 0.26210495829582214,
+      "learning_rate": 3.279578836859754e-07,
+      "loss": 0.3663,
       "step": 131835
     },
     {
-      "epoch": 4.64,
-      "learning_rate": 6.88383196797554e-07,
-      "loss": 0.2564,
+      "epoch": 4.7515046671712255,
+      "grad_norm": 0.3384738862514496,
+      "learning_rate": 3.274869332014152e-07,
+      "loss": 0.379,
       "step": 131840
     },
     {
-      "epoch": 4.64,
-      "learning_rate": 6.877194591606406e-07,
-      "loss": 0.2375,
+      "epoch": 4.75168486683245,
+      "grad_norm": 0.3079512119293213,
+      "learning_rate": 3.27016318875073e-07,
+      "loss": 0.4015,
       "step": 131845
     },
     {
-      "epoch": 4.64,
-      "learning_rate": 6.870560372026629e-07,
-      "loss": 0.2494,
+      "epoch": 4.751865066493675,
+      "grad_norm": 0.2897801995277405,
+      "learning_rate": 3.265460407133547e-07,
+      "loss": 0.382,
       "step": 131850
     },
     {
-      "epoch": 4.64,
-      "learning_rate": 6.86392930932242e-07,
-      "loss": 0.233,
+      "epoch": 4.7520452661549,
+      "grad_norm": 0.2497911900281906,
+      "learning_rate": 3.260760987226691e-07,
+      "loss": 0.3617,
       "step": 131855
     },
     {
-      "epoch": 4.64,
-      "learning_rate": 6.857301403579875e-07,
-      "loss": 0.2484,
+      "epoch": 4.7522254658161245,
+      "grad_norm": 0.23933720588684082,
+      "learning_rate": 3.2560649290942216e-07,
+      "loss": 0.3815,
       "step": 131860
     },
     {
-      "epoch": 4.64,
-      "learning_rate": 6.850676654885008e-07,
-      "loss": 0.2281,
+      "epoch": 4.752405665477349,
+      "grad_norm": 0.24069957435131073,
+      "learning_rate": 3.2513722328000595e-07,
+      "loss": 0.363,
       "step": 131865
     },
     {
-      "epoch": 4.64,
-      "learning_rate": 6.84405506332389e-07,
-      "loss": 0.2505,
+      "epoch": 4.752585865138574,
+      "grad_norm": 0.3069649636745453,
+      "learning_rate": 3.2466828984082096e-07,
+      "loss": 0.3922,
       "step": 131870
     },
     {
-      "epoch": 4.64,
-      "learning_rate": 6.837436628982452e-07,
-      "loss": 0.2339,
+      "epoch": 4.752766064799798,
+      "grad_norm": 0.21737022697925568,
+      "learning_rate": 3.24199692598251e-07,
+      "loss": 0.3574,
       "step": 131875
     },
     {
-      "epoch": 4.64,
-      "learning_rate": 6.830821351946654e-07,
-      "loss": 0.2606,
+      "epoch": 4.7529462644610225,
+      "grad_norm": 0.22841662168502808,
+      "learning_rate": 3.2373143155868255e-07,
+      "loss": 0.344,
       "step": 131880
     },
     {
-      "epoch": 4.64,
-      "learning_rate": 6.82420923230237e-07,
-      "loss": 0.2293,
+      "epoch": 4.753126464122247,
+      "grad_norm": 0.25134187936782837,
+      "learning_rate": 3.232635067284995e-07,
+      "loss": 0.3917,
       "step": 131885
     },
     {
-      "epoch": 4.64,
-      "learning_rate": 6.817600270135504e-07,
-      "loss": 0.2381,
+      "epoch": 4.753306663783472,
+      "grad_norm": 0.21516527235507965,
+      "learning_rate": 3.2279591811406884e-07,
+      "loss": 0.3816,
       "step": 131890
     },
     {
-      "epoch": 4.64,
-      "learning_rate": 6.810994465531822e-07,
-      "loss": 0.2254,
+      "epoch": 4.753486863444697,
+      "grad_norm": 0.20590807497501373,
+      "learning_rate": 3.223286657217717e-07,
+      "loss": 0.3616,
       "step": 131895
     },
     {
-      "epoch": 4.64,
-      "learning_rate": 6.804391818577089e-07,
-      "loss": 0.2608,
+      "epoch": 4.7536670631059215,
+      "grad_norm": 0.2603895962238312,
+      "learning_rate": 3.21861749557964e-07,
+      "loss": 0.3731,
       "step": 131900
     },
     {
-      "epoch": 4.64,
-      "learning_rate": 6.797792329357039e-07,
-      "loss": 0.2643,
+      "epoch": 4.753847262767146,
+      "grad_norm": 0.2670796513557434,
+      "learning_rate": 3.2139516962901297e-07,
+      "loss": 0.3868,
       "step": 131905
     },
     {
-      "epoch": 4.64,
-      "learning_rate": 6.791195997957439e-07,
-      "loss": 0.2592,
+      "epoch": 4.754027462428371,
+      "grad_norm": 0.2581007480621338,
+      "learning_rate": 3.2092892594127454e-07,
+      "loss": 0.382,
       "step": 131910
     },
     {
-      "epoch": 4.64,
-      "learning_rate": 6.784602824463859e-07,
-      "loss": 0.2448,
+      "epoch": 4.754207662089595,
+      "grad_norm": 0.29957446455955505,
+      "learning_rate": 3.2046301850110484e-07,
+      "loss": 0.3801,
       "step": 131915
     },
     {
-      "epoch": 4.64,
-      "learning_rate": 6.778012808961925e-07,
-      "loss": 0.246,
+      "epoch": 4.7543878617508195,
+      "grad_norm": 0.21832120418548584,
+      "learning_rate": 3.199974473148459e-07,
+      "loss": 0.3558,
       "step": 131920
     },
     {
-      "epoch": 4.64,
-      "learning_rate": 6.77142595153718e-07,
-      "loss": 0.236,
+      "epoch": 4.754568061412044,
+      "grad_norm": 0.3172387182712555,
+      "learning_rate": 3.195322123888428e-07,
+      "loss": 0.3774,
       "step": 131925
     },
     {
-      "epoch": 4.64,
-      "learning_rate": 6.764842252275195e-07,
-      "loss": 0.249,
+      "epoch": 4.754748261073269,
+      "grad_norm": 0.2534702718257904,
+      "learning_rate": 3.1906731372943476e-07,
+      "loss": 0.386,
       "step": 131930
     },
     {
-      "epoch": 4.64,
-      "learning_rate": 6.758261711261455e-07,
-      "loss": 0.2457,
+      "epoch": 4.754928460734494,
+      "grad_norm": 0.21878264844417572,
+      "learning_rate": 3.186027513429585e-07,
+      "loss": 0.3593,
       "step": 131935
     },
     {
-      "epoch": 4.64,
-      "learning_rate": 6.751684328581364e-07,
-      "loss": 0.2405,
+      "epoch": 4.7551086603957184,
+      "grad_norm": 0.2851618528366089,
+      "learning_rate": 3.181385252357394e-07,
+      "loss": 0.4047,
       "step": 131940
     },
     {
-      "epoch": 4.64,
-      "learning_rate": 6.745110104320329e-07,
-      "loss": 0.2463,
+      "epoch": 4.755288860056943,
+      "grad_norm": 0.24602822959423065,
+      "learning_rate": 3.176746354141058e-07,
+      "loss": 0.3892,
       "step": 131945
     },
     {
-      "epoch": 4.64,
-      "learning_rate": 6.738539038563752e-07,
-      "loss": 0.2291,
+      "epoch": 4.755469059718168,
+      "grad_norm": 0.25003623962402344,
+      "learning_rate": 3.172110818843749e-07,
+      "loss": 0.3674,
       "step": 131950
     },
     {
-      "epoch": 4.64,
-      "learning_rate": 6.731971131396897e-07,
-      "loss": 0.2499,
+      "epoch": 4.755649259379393,
+      "grad_norm": 0.282927930355072,
+      "learning_rate": 3.167478646528638e-07,
+      "loss": 0.3198,
       "step": 131955
     },
     {
-      "epoch": 4.64,
-      "learning_rate": 6.725406382905086e-07,
-      "loss": 0.2429,
+      "epoch": 4.755829459040617,
+      "grad_norm": 0.23673711717128754,
+      "learning_rate": 3.162849837258869e-07,
+      "loss": 0.3578,
       "step": 131960
     },
     {
-      "epoch": 4.64,
-      "learning_rate": 6.718844793173585e-07,
-      "loss": 0.2376,
+      "epoch": 4.756009658701842,
+      "grad_norm": 0.2531015872955322,
+      "learning_rate": 3.158224391097503e-07,
+      "loss": 0.3597,
       "step": 131965
     },
     {
-      "epoch": 4.64,
-      "learning_rate": 6.712286362287545e-07,
-      "loss": 0.2402,
+      "epoch": 4.756189858363066,
+      "grad_norm": 0.25352203845977783,
+      "learning_rate": 3.1536023081074894e-07,
+      "loss": 0.4158,
       "step": 131970
     },
     {
-      "epoch": 4.64,
-      "learning_rate": 6.705731090332123e-07,
-      "loss": 0.2352,
+      "epoch": 4.756370058024291,
+      "grad_norm": 0.278499037027359,
+      "learning_rate": 3.1489835883518615e-07,
+      "loss": 0.3489,
       "step": 131975
     },
     {
-      "epoch": 4.64,
-      "learning_rate": 6.699178977392446e-07,
-      "loss": 0.2627,
+      "epoch": 4.756550257685515,
+      "grad_norm": 0.27848923206329346,
+      "learning_rate": 3.144368231893541e-07,
+      "loss": 0.4042,
       "step": 131980
     },
     {
-      "epoch": 4.64,
-      "learning_rate": 6.692630023553581e-07,
-      "loss": 0.2318,
+      "epoch": 4.75673045734674,
+      "grad_norm": 0.25936684012413025,
+      "learning_rate": 3.1397562387954214e-07,
+      "loss": 0.3477,
       "step": 131985
     },
     {
-      "epoch": 4.64,
-      "learning_rate": 6.686084228900602e-07,
-      "loss": 0.2494,
+      "epoch": 4.756910657007965,
+      "grad_norm": 0.21595986187458038,
+      "learning_rate": 3.1351476091203427e-07,
+      "loss": 0.3695,
       "step": 131990
     },
     {
-      "epoch": 4.64,
-      "learning_rate": 6.679541593518468e-07,
-      "loss": 0.2371,
+      "epoch": 4.75709085666919,
+      "grad_norm": 0.2597789466381073,
+      "learning_rate": 3.1305423429310587e-07,
+      "loss": 0.3714,
       "step": 131995
     },
     {
-      "epoch": 4.64,
-      "learning_rate": 6.673002117492083e-07,
-      "loss": 0.262,
+      "epoch": 4.757271056330414,
+      "grad_norm": 0.22275707125663757,
+      "learning_rate": 3.1259404402903536e-07,
+      "loss": 0.3491,
       "step": 132000
     },
     {
-      "epoch": 4.64,
-      "eval_loss": 0.24853284657001495,
-      "eval_runtime": 10.5548,
-      "eval_samples_per_second": 9.474,
-      "eval_steps_per_second": 9.474,
+      "epoch": 4.757271056330414,
+      "eval_loss": 0.4288484454154968,
+      "eval_runtime": 3.5292,
+      "eval_samples_per_second": 28.335,
+      "eval_steps_per_second": 7.084,
       "step": 132000
     },
     {
-      "epoch": 4.64,
-      "learning_rate": 6.666465800906435e-07,
-      "loss": 0.2598,
+      "epoch": 4.757451255991639,
+      "grad_norm": 0.25196489691734314,
+      "learning_rate": 3.121341901260899e-07,
+      "loss": 0.3829,
       "step": 132005
     },
     {
-      "epoch": 4.64,
-      "learning_rate": 6.659932643846345e-07,
-      "loss": 0.2609,
+      "epoch": 4.757631455652863,
+      "grad_norm": 0.21652761101722717,
+      "learning_rate": 3.1167467259053675e-07,
+      "loss": 0.3318,
       "step": 132010
     },
     {
-      "epoch": 4.64,
-      "learning_rate": 6.653402646396717e-07,
-      "loss": 0.2532,
+      "epoch": 4.757811655314088,
+      "grad_norm": 0.2428998500108719,
+      "learning_rate": 3.112154914286375e-07,
+      "loss": 0.357,
       "step": 132015
     },
     {
-      "epoch": 4.64,
-      "learning_rate": 6.646875808642233e-07,
-      "loss": 0.2528,
+      "epoch": 4.757991854975312,
+      "grad_norm": 0.2448771446943283,
+      "learning_rate": 3.1075664664664827e-07,
+      "loss": 0.3707,
       "step": 132020
     },
     {
-      "epoch": 4.65,
-      "learning_rate": 6.640352130667716e-07,
-      "loss": 0.26,
+      "epoch": 4.758172054636537,
+      "grad_norm": 0.2827889621257782,
+      "learning_rate": 3.1029813825081687e-07,
+      "loss": 0.393,
       "step": 132025
     },
     {
-      "epoch": 4.65,
-      "learning_rate": 6.633831612557844e-07,
-      "loss": 0.2451,
+      "epoch": 4.758352254297762,
+      "grad_norm": 0.21509301662445068,
+      "learning_rate": 3.098399662473939e-07,
+      "loss": 0.3583,
       "step": 132030
     },
     {
-      "epoch": 4.65,
-      "learning_rate": 6.627314254397276e-07,
-      "loss": 0.2449,
+      "epoch": 4.758532453958987,
+      "grad_norm": 0.23091164231300354,
+      "learning_rate": 3.0938213064262143e-07,
+      "loss": 0.3593,
       "step": 132035
     },
     {
-      "epoch": 4.65,
-      "learning_rate": 6.620800056270638e-07,
-      "loss": 0.2424,
+      "epoch": 4.758712653620211,
+      "grad_norm": 0.3308013677597046,
+      "learning_rate": 3.089246314427391e-07,
+      "loss": 0.383,
       "step": 132040
     },
     {
-      "epoch": 4.65,
-      "learning_rate": 6.614289018262526e-07,
-      "loss": 0.2506,
+      "epoch": 4.758892853281436,
+      "grad_norm": 0.2075890302658081,
+      "learning_rate": 3.084674686539779e-07,
+      "loss": 0.3666,
       "step": 132045
     },
     {
-      "epoch": 4.65,
-      "learning_rate": 6.607781140457487e-07,
-      "loss": 0.2562,
+      "epoch": 4.759073052942661,
+      "grad_norm": 0.2468065321445465,
+      "learning_rate": 3.080106422825635e-07,
+      "loss": 0.3804,
       "step": 132050
     },
     {
-      "epoch": 4.65,
-      "learning_rate": 6.601276422940006e-07,
-      "loss": 0.2633,
+      "epoch": 4.7592532526038855,
+      "grad_norm": 0.24998079240322113,
+      "learning_rate": 3.0755415233472693e-07,
+      "loss": 0.4156,
       "step": 132055
     },
     {
-      "epoch": 4.65,
-      "learning_rate": 6.594774865794517e-07,
-      "loss": 0.2432,
+      "epoch": 4.759433452265109,
+      "grad_norm": 0.23068219423294067,
+      "learning_rate": 3.0709799881668554e-07,
+      "loss": 0.3924,
       "step": 132060
     },
     {
-      "epoch": 4.65,
-      "learning_rate": 6.588276469105481e-07,
-      "loss": 0.2413,
+      "epoch": 4.759613651926334,
+      "grad_norm": 0.2239476889371872,
+      "learning_rate": 3.066421817346482e-07,
+      "loss": 0.3956,
       "step": 132065
     },
     {
-      "epoch": 4.65,
-      "learning_rate": 6.581781232957273e-07,
-      "loss": 0.2464,
+      "epoch": 4.759793851587559,
+      "grad_norm": 0.2508222162723541,
+      "learning_rate": 3.0618670109483494e-07,
+      "loss": 0.3624,
       "step": 132070
     },
     {
-      "epoch": 4.65,
-      "learning_rate": 6.575289157434189e-07,
-      "loss": 0.2366,
+      "epoch": 4.759974051248784,
+      "grad_norm": 0.21215388178825378,
+      "learning_rate": 3.0573155690344355e-07,
+      "loss": 0.3786,
       "step": 132075
     },
     {
-      "epoch": 4.65,
-      "learning_rate": 6.568800242620577e-07,
-      "loss": 0.2498,
+      "epoch": 4.760154250910008,
+      "grad_norm": 0.23385365307331085,
+      "learning_rate": 3.05276749166683e-07,
+      "loss": 0.3711,
       "step": 132080
     },
     {
-      "epoch": 4.65,
-      "learning_rate": 6.562314488600674e-07,
-      "loss": 0.2505,
+      "epoch": 4.760334450571233,
+      "grad_norm": 0.26072898507118225,
+      "learning_rate": 3.048222778907428e-07,
+      "loss": 0.3787,
       "step": 132085
     },
     {
-      "epoch": 4.65,
-      "learning_rate": 6.555831895458691e-07,
-      "loss": 0.256,
+      "epoch": 4.760514650232458,
+      "grad_norm": 0.27673956751823425,
+      "learning_rate": 3.0436814308181794e-07,
+      "loss": 0.3841,
       "step": 132090
     },
     {
-      "epoch": 4.65,
-      "learning_rate": 6.549352463278757e-07,
-      "loss": 0.2644,
+      "epoch": 4.7606948498936825,
+      "grad_norm": 0.2762385308742523,
+      "learning_rate": 3.0391434474609516e-07,
+      "loss": 0.3801,
       "step": 132095
     },
     {
-      "epoch": 4.65,
-      "learning_rate": 6.542876192145053e-07,
-      "loss": 0.2679,
+      "epoch": 4.760875049554906,
+      "grad_norm": 0.27955085039138794,
+      "learning_rate": 3.0346088288976117e-07,
+      "loss": 0.3601,
       "step": 132100
     },
     {
-      "epoch": 4.65,
-      "learning_rate": 6.536403082141679e-07,
-      "loss": 0.2297,
+      "epoch": 4.761055249216131,
+      "grad_norm": 0.2552419900894165,
+      "learning_rate": 3.030077575189888e-07,
+      "loss": 0.363,
       "step": 132105
     },
     {
-      "epoch": 4.65,
-      "learning_rate": 6.529933133352678e-07,
-      "loss": 0.256,
+      "epoch": 4.761235448877356,
+      "grad_norm": 0.30773216485977173,
+      "learning_rate": 3.0255496863995646e-07,
+      "loss": 0.3942,
       "step": 132110
     },
     {
-      "epoch": 4.65,
-      "learning_rate": 6.523466345861984e-07,
-      "loss": 0.2407,
+      "epoch": 4.761415648538581,
+      "grad_norm": 0.22618763148784637,
+      "learning_rate": 3.021025162588259e-07,
+      "loss": 0.391,
       "step": 132115
     },
     {
-      "epoch": 4.65,
-      "learning_rate": 6.51700271975364e-07,
-      "loss": 0.2212,
+      "epoch": 4.761595848199805,
+      "grad_norm": 0.2810955047607422,
+      "learning_rate": 3.016504003817727e-07,
+      "loss": 0.3841,
       "step": 132120
     },
     {
-      "epoch": 4.65,
-      "learning_rate": 6.510542255111579e-07,
-      "loss": 0.2716,
+      "epoch": 4.76177604786103,
+      "grad_norm": 0.25989750027656555,
+      "learning_rate": 3.0119862101494754e-07,
+      "loss": 0.3548,
       "step": 132125
     },
     {
-      "epoch": 4.65,
-      "learning_rate": 6.504084952019651e-07,
-      "loss": 0.2682,
+      "epoch": 4.761956247522255,
+      "grad_norm": 0.2658855617046356,
+      "learning_rate": 3.0074717816451214e-07,
+      "loss": 0.3629,
       "step": 132130
     },
     {
-      "epoch": 4.65,
-      "learning_rate": 6.497630810561706e-07,
-      "loss": 0.2413,
+      "epoch": 4.7621364471834795,
+      "grad_norm": 0.27331972122192383,
+      "learning_rate": 3.002960718366116e-07,
+      "loss": 0.3689,
       "step": 132135
     },
     {
-      "epoch": 4.65,
-      "learning_rate": 6.491179830821537e-07,
-      "loss": 0.2548,
+      "epoch": 4.762316646844704,
+      "grad_norm": 0.22226150333881378,
+      "learning_rate": 2.998453020373965e-07,
+      "loss": 0.4006,
       "step": 132140
     },
     {
-      "epoch": 4.65,
-      "learning_rate": 6.484732012882966e-07,
-      "loss": 0.2452,
+      "epoch": 4.762496846505929,
+      "grad_norm": 0.24564702808856964,
+      "learning_rate": 2.9939486877300636e-07,
+      "loss": 0.3452,
       "step": 132145
     },
     {
-      "epoch": 4.65,
-      "learning_rate": 6.478287356829649e-07,
-      "loss": 0.2513,
+      "epoch": 4.762677046167154,
+      "grad_norm": 0.3113086521625519,
+      "learning_rate": 2.989447720495808e-07,
+      "loss": 0.3813,
       "step": 132150
     },
     {
-      "epoch": 4.65,
-      "learning_rate": 6.47184586274524e-07,
-      "loss": 0.2597,
+      "epoch": 4.7628572458283776,
+      "grad_norm": 0.24232225120067596,
+      "learning_rate": 2.9849501187324823e-07,
+      "loss": 0.3843,
       "step": 132155
     },
     {
-      "epoch": 4.65,
-      "learning_rate": 6.465407530713452e-07,
-      "loss": 0.2343,
+      "epoch": 4.763037445489602,
+      "grad_norm": 0.21868427097797394,
+      "learning_rate": 2.980455882501398e-07,
+      "loss": 0.3898,
       "step": 132160
     },
     {
-      "epoch": 4.65,
-      "learning_rate": 6.458972360817883e-07,
-      "loss": 0.2566,
+      "epoch": 4.763217645150827,
+      "grad_norm": 0.2546680271625519,
+      "learning_rate": 2.975965011863785e-07,
+      "loss": 0.318,
       "step": 132165
     },
     {
-      "epoch": 4.65,
-      "learning_rate": 6.452540353141995e-07,
-      "loss": 0.249,
+      "epoch": 4.763397844812052,
+      "grad_norm": 0.2764042615890503,
+      "learning_rate": 2.971477506880815e-07,
+      "loss": 0.3723,
       "step": 132170
     },
     {
-      "epoch": 4.65,
-      "learning_rate": 6.446111507769414e-07,
-      "loss": 0.2572,
+      "epoch": 4.7635780444732765,
+      "grad_norm": 0.23360565304756165,
+      "learning_rate": 2.9669933676136354e-07,
+      "loss": 0.3803,
       "step": 132175
     },
     {
-      "epoch": 4.65,
-      "learning_rate": 6.439685824783548e-07,
-      "loss": 0.2377,
+      "epoch": 4.763758244134501,
+      "grad_norm": 0.258620023727417,
+      "learning_rate": 2.9625125941233345e-07,
+      "loss": 0.4101,
       "step": 132180
     },
     {
-      "epoch": 4.65,
-      "learning_rate": 6.433263304267855e-07,
-      "loss": 0.2539,
+      "epoch": 4.763938443795726,
+      "grad_norm": 0.2147550880908966,
+      "learning_rate": 2.9580351864710034e-07,
+      "loss": 0.362,
       "step": 132185
     },
     {
-      "epoch": 4.65,
-      "learning_rate": 6.426843946305716e-07,
-      "loss": 0.2579,
+      "epoch": 4.764118643456951,
+      "grad_norm": 0.2418157309293747,
+      "learning_rate": 2.9535611447175924e-07,
+      "loss": 0.358,
       "step": 132190
     },
     {
-      "epoch": 4.65,
-      "learning_rate": 6.420427750980479e-07,
-      "loss": 0.245,
+      "epoch": 4.7642988431181745,
+      "grad_norm": 0.26910528540611267,
+      "learning_rate": 2.9490904689240815e-07,
+      "loss": 0.4377,
       "step": 132195
     },
     {
-      "epoch": 4.65,
-      "learning_rate": 6.414014718375494e-07,
-      "loss": 0.2459,
+      "epoch": 4.764479042779399,
+      "grad_norm": 0.26641741394996643,
+      "learning_rate": 2.944623159151394e-07,
+      "loss": 0.3293,
       "step": 132200
     },
     {
-      "epoch": 4.65,
-      "learning_rate": 6.407604848574001e-07,
-      "loss": 0.2422,
+      "epoch": 4.764659242440624,
+      "grad_norm": 0.2724882960319519,
+      "learning_rate": 2.9401592154603694e-07,
+      "loss": 0.3949,
       "step": 132205
     },
     {
-      "epoch": 4.65,
-      "learning_rate": 6.401198141659182e-07,
-      "loss": 0.2429,
+      "epoch": 4.764839442101849,
+      "grad_norm": 0.245937317609787,
+      "learning_rate": 2.935698637911849e-07,
+      "loss": 0.3721,
       "step": 132210
     },
     {
-      "epoch": 4.65,
-      "learning_rate": 6.394794597714304e-07,
-      "loss": 0.2692,
+      "epoch": 4.7650196417630735,
+      "grad_norm": 0.25321248173713684,
+      "learning_rate": 2.9312414265665886e-07,
+      "loss": 0.3822,
       "step": 132215
     },
     {
-      "epoch": 4.65,
-      "learning_rate": 6.388394216822469e-07,
-      "loss": 0.2535,
+      "epoch": 4.765199841424298,
+      "grad_norm": 0.27613797783851624,
+      "learning_rate": 2.9267875814853463e-07,
+      "loss": 0.3634,
       "step": 132220
     },
     {
-      "epoch": 4.65,
-      "learning_rate": 6.381996999066775e-07,
-      "loss": 0.2592,
+      "epoch": 4.765380041085523,
+      "grad_norm": 0.2703777849674225,
+      "learning_rate": 2.922337102728795e-07,
+      "loss": 0.3705,
       "step": 132225
     },
     {
-      "epoch": 4.65,
-      "learning_rate": 6.375602944530295e-07,
-      "loss": 0.247,
+      "epoch": 4.765560240746748,
+      "grad_norm": 0.23466283082962036,
+      "learning_rate": 2.917889990357553e-07,
+      "loss": 0.3601,
       "step": 132230
     },
     {
-      "epoch": 4.65,
-      "learning_rate": 6.369212053296076e-07,
-      "loss": 0.2495,
+      "epoch": 4.765740440407972,
+      "grad_norm": 0.26712194085121155,
+      "learning_rate": 2.9134462444322106e-07,
+      "loss": 0.3573,
       "step": 132235
     },
     {
-      "epoch": 4.65,
-      "learning_rate": 6.362824325447076e-07,
-      "loss": 0.2421,
+      "epoch": 4.765920640069197,
+      "grad_norm": 0.23665735125541687,
+      "learning_rate": 2.909005865013331e-07,
+      "loss": 0.3711,
       "step": 132240
     },
     {
-      "epoch": 4.65,
-      "learning_rate": 6.35643976106623e-07,
-      "loss": 0.2519,
+      "epoch": 4.766100839730421,
+      "grad_norm": 0.26551133394241333,
+      "learning_rate": 2.904568852161449e-07,
+      "loss": 0.3894,
       "step": 132245
     },
     {
-      "epoch": 4.65,
-      "learning_rate": 6.350058360236443e-07,
-      "loss": 0.2391,
+      "epoch": 4.766281039391646,
+      "grad_norm": 0.21451817452907562,
+      "learning_rate": 2.900135205936932e-07,
+      "loss": 0.3891,
       "step": 132250
     },
     {
-      "epoch": 4.65,
-      "learning_rate": 6.343680123040597e-07,
-      "loss": 0.2558,
+      "epoch": 4.7664612390528704,
+      "grad_norm": 0.256231427192688,
+      "learning_rate": 2.8957049264002334e-07,
+      "loss": 0.3673,
       "step": 132255
     },
     {
-      "epoch": 4.65,
-      "learning_rate": 6.337305049561481e-07,
-      "loss": 0.2499,
+      "epoch": 4.766641438714095,
+      "grad_norm": 0.2589702904224396,
+      "learning_rate": 2.89127801361172e-07,
+      "loss": 0.4113,
       "step": 132260
     },
     {
-      "epoch": 4.65,
-      "learning_rate": 6.330933139881895e-07,
-      "loss": 0.2328,
+      "epoch": 4.76682163837532,
+      "grad_norm": 0.25339174270629883,
+      "learning_rate": 2.886854467631733e-07,
+      "loss": 0.3935,
       "step": 132265
     },
     {
-      "epoch": 4.65,
-      "learning_rate": 6.32456439408452e-07,
-      "loss": 0.2415,
+      "epoch": 4.767001838036545,
+      "grad_norm": 0.23474030196666718,
+      "learning_rate": 2.882434288520475e-07,
+      "loss": 0.3563,
       "step": 132270
     },
     {
-      "epoch": 4.65,
-      "learning_rate": 6.318198812252124e-07,
-      "loss": 0.244,
+      "epoch": 4.767182037697769,
+      "grad_norm": 0.2822888195514679,
+      "learning_rate": 2.8780174763382296e-07,
+      "loss": 0.3655,
       "step": 132275
     },
     {
-      "epoch": 4.65,
-      "learning_rate": 6.31183639446728e-07,
-      "loss": 0.2548,
+      "epoch": 4.767362237358994,
+      "grad_norm": 0.32862353324890137,
+      "learning_rate": 2.8736040311451163e-07,
+      "loss": 0.3644,
       "step": 132280
     },
     {
-      "epoch": 4.65,
-      "learning_rate": 6.305477140812699e-07,
-      "loss": 0.2634,
+      "epoch": 4.767542437020218,
+      "grad_norm": 0.2945389151573181,
+      "learning_rate": 2.869193953001337e-07,
+      "loss": 0.3493,
       "step": 132285
     },
     {
-      "epoch": 4.65,
-      "learning_rate": 6.299121051370843e-07,
-      "loss": 0.2385,
+      "epoch": 4.767722636681443,
+      "grad_norm": 0.20633207261562347,
+      "learning_rate": 2.8647872419669544e-07,
+      "loss": 0.3379,
       "step": 132290
     },
     {
-      "epoch": 4.65,
-      "learning_rate": 6.292768126224341e-07,
-      "loss": 0.2506,
+      "epoch": 4.767902836342667,
+      "grad_norm": 0.25659510493278503,
+      "learning_rate": 2.8603838981019757e-07,
+      "loss": 0.3413,
       "step": 132295
     },
     {
-      "epoch": 4.65,
-      "learning_rate": 6.286418365455626e-07,
-      "loss": 0.2482,
+      "epoch": 4.768083036003892,
+      "grad_norm": 0.24695339798927307,
+      "learning_rate": 2.8559839214664094e-07,
+      "loss": 0.3909,
       "step": 132300
     },
     {
-      "epoch": 4.65,
-      "learning_rate": 6.280071769147134e-07,
-      "loss": 0.2356,
+      "epoch": 4.768263235665117,
+      "grad_norm": 0.2526165544986725,
+      "learning_rate": 2.8515873121202076e-07,
+      "loss": 0.3805,
       "step": 132305
     },
     {
-      "epoch": 4.66,
-      "learning_rate": 6.273728337381296e-07,
-      "loss": 0.2472,
+      "epoch": 4.768443435326342,
+      "grad_norm": 0.31589943170547485,
+      "learning_rate": 2.8471940701233216e-07,
+      "loss": 0.4108,
       "step": 132310
     },
     {
-      "epoch": 4.66,
-      "learning_rate": 6.267388070240493e-07,
-      "loss": 0.2518,
+      "epoch": 4.768623634987566,
+      "grad_norm": 0.24592305719852448,
+      "learning_rate": 2.8428041955355365e-07,
+      "loss": 0.3881,
       "step": 132315
     },
     {
-      "epoch": 4.66,
-      "learning_rate": 6.261050967807047e-07,
-      "loss": 0.2369,
+      "epoch": 4.768803834648791,
+      "grad_norm": 0.30194249749183655,
+      "learning_rate": 2.8384176884166947e-07,
+      "loss": 0.3638,
       "step": 132320
     },
     {
-      "epoch": 4.66,
-      "learning_rate": 6.254717030163171e-07,
-      "loss": 0.2795,
+      "epoch": 4.768984034310016,
+      "grad_norm": 0.26434436440467834,
+      "learning_rate": 2.834034548826553e-07,
+      "loss": 0.3949,
       "step": 132325
     },
     {
-      "epoch": 4.66,
-      "learning_rate": 6.248386257391187e-07,
-      "loss": 0.2486,
+      "epoch": 4.769164233971241,
+      "grad_norm": 0.24037402868270874,
+      "learning_rate": 2.8296547768248415e-07,
+      "loss": 0.3672,
       "step": 132330
     },
     {
-      "epoch": 4.66,
-      "learning_rate": 6.242058649573279e-07,
-      "loss": 0.247,
+      "epoch": 4.769344433632464,
+      "grad_norm": 0.2738417088985443,
+      "learning_rate": 2.825278372471263e-07,
+      "loss": 0.3967,
       "step": 132335
     },
     {
-      "epoch": 4.66,
-      "learning_rate": 6.235734206791577e-07,
-      "loss": 0.264,
+      "epoch": 4.769524633293689,
+      "grad_norm": 0.2520720958709717,
+      "learning_rate": 2.82090533582538e-07,
+      "loss": 0.3668,
       "step": 132340
     },
     {
-      "epoch": 4.66,
-      "learning_rate": 6.229412929128209e-07,
-      "loss": 0.2515,
+      "epoch": 4.769704832954914,
+      "grad_norm": 0.2859429717063904,
+      "learning_rate": 2.8165356669468126e-07,
+      "loss": 0.3258,
       "step": 132345
     },
     {
-      "epoch": 4.66,
-      "learning_rate": 6.223094816665304e-07,
-      "loss": 0.2593,
+      "epoch": 4.769885032616139,
+      "grad_norm": 0.2699842154979706,
+      "learning_rate": 2.8121693658950955e-07,
+      "loss": 0.352,
       "step": 132350
     },
     {
-      "epoch": 4.66,
-      "learning_rate": 6.216779869484823e-07,
-      "loss": 0.2431,
+      "epoch": 4.770065232277363,
+      "grad_norm": 0.29695814847946167,
+      "learning_rate": 2.807806432729682e-07,
+      "loss": 0.3885,
       "step": 132355
     },
     {
-      "epoch": 4.66,
-      "learning_rate": 6.210468087668842e-07,
-      "loss": 0.2436,
+      "epoch": 4.770245431938588,
+      "grad_norm": 0.2401593029499054,
+      "learning_rate": 2.803446867510079e-07,
+      "loss": 0.3437,
       "step": 132360
     },
     {
-      "epoch": 4.66,
-      "learning_rate": 6.204159471299209e-07,
-      "loss": 0.2392,
+      "epoch": 4.770425631599813,
+      "grad_norm": 0.26505574584007263,
+      "learning_rate": 2.7990906702956566e-07,
+      "loss": 0.3474,
       "step": 132365
     },
     {
-      "epoch": 4.66,
-      "learning_rate": 6.197854020457944e-07,
-      "loss": 0.2254,
+      "epoch": 4.7706058312610375,
+      "grad_norm": 0.7576366662979126,
+      "learning_rate": 2.7947378411457557e-07,
+      "loss": 0.3716,
       "step": 132370
     },
     {
-      "epoch": 4.66,
-      "learning_rate": 6.191551735226869e-07,
-      "loss": 0.2355,
+      "epoch": 4.770786030922261,
+      "grad_norm": 0.3026728332042694,
+      "learning_rate": 2.7903883801196906e-07,
+      "loss": 0.342,
       "step": 132375
     },
     {
-      "epoch": 4.66,
-      "learning_rate": 6.185252615687809e-07,
-      "loss": 0.2576,
+      "epoch": 4.770966230583486,
+      "grad_norm": 0.3232538104057312,
+      "learning_rate": 2.786042287276719e-07,
+      "loss": 0.3823,
       "step": 132380
     },
     {
-      "epoch": 4.66,
-      "learning_rate": 6.178956661922531e-07,
-      "loss": 0.2521,
+      "epoch": 4.771146430244711,
+      "grad_norm": 0.24834728240966797,
+      "learning_rate": 2.781699562676071e-07,
+      "loss": 0.3594,
       "step": 132385
     },
     {
-      "epoch": 4.66,
-      "learning_rate": 6.172663874012857e-07,
-      "loss": 0.2453,
+      "epoch": 4.771326629905936,
+      "grad_norm": 0.23225435614585876,
+      "learning_rate": 2.7773602063768946e-07,
+      "loss": 0.3594,
       "step": 132390
     },
     {
-      "epoch": 4.66,
-      "learning_rate": 6.166374252040418e-07,
-      "loss": 0.2552,
+      "epoch": 4.77150682956716,
+      "grad_norm": 0.27175524830818176,
+      "learning_rate": 2.773024218438308e-07,
+      "loss": 0.3937,
       "step": 132395
     },
     {
-      "epoch": 4.66,
-      "learning_rate": 6.160087796086955e-07,
-      "loss": 0.234,
+      "epoch": 4.771687029228385,
+      "grad_norm": 0.250227153301239,
+      "learning_rate": 2.768691598919432e-07,
+      "loss": 0.3693,
       "step": 132400
     },
     {
-      "epoch": 4.66,
-      "learning_rate": 6.15380450623404e-07,
-      "loss": 0.2648,
+      "epoch": 4.77186722888961,
+      "grad_norm": 0.23589470982551575,
+      "learning_rate": 2.7643623478792456e-07,
+      "loss": 0.369,
       "step": 132405
     },
     {
-      "epoch": 4.66,
-      "learning_rate": 6.147524382563275e-07,
-      "loss": 0.2542,
+      "epoch": 4.7720474285508345,
+      "grad_norm": 0.32625946402549744,
+      "learning_rate": 2.7600364653767584e-07,
+      "loss": 0.3857,
       "step": 132410
     },
     {
-      "epoch": 4.66,
-      "learning_rate": 6.141247425156204e-07,
-      "loss": 0.22,
+      "epoch": 4.772227628212059,
+      "grad_norm": 0.29416128993034363,
+      "learning_rate": 2.755713951470923e-07,
+      "loss": 0.3979,
       "step": 132415
     },
     {
-      "epoch": 4.66,
-      "learning_rate": 6.134973634094321e-07,
-      "loss": 0.2267,
+      "epoch": 4.772407827873284,
+      "grad_norm": 0.22298985719680786,
+      "learning_rate": 2.7513948062205807e-07,
+      "loss": 0.383,
       "step": 132420
     },
     {
-      "epoch": 4.66,
-      "learning_rate": 6.128703009459114e-07,
-      "loss": 0.25,
+      "epoch": 4.772588027534509,
+      "grad_norm": 0.23173697292804718,
+      "learning_rate": 2.747079029684657e-07,
+      "loss": 0.335,
       "step": 132425
     },
     {
-      "epoch": 4.66,
-      "learning_rate": 6.122435551331962e-07,
-      "loss": 0.2679,
+      "epoch": 4.772768227195733,
+      "grad_norm": 0.3118121922016144,
+      "learning_rate": 2.7427666219219104e-07,
+      "loss": 0.411,
       "step": 132430
     },
     {
-      "epoch": 4.66,
-      "learning_rate": 6.116171259794273e-07,
-      "loss": 0.2352,
+      "epoch": 4.772948426856957,
+      "grad_norm": 0.23155774176120758,
+      "learning_rate": 2.738457582991072e-07,
+      "loss": 0.3765,
       "step": 132435
     },
     {
-      "epoch": 4.66,
-      "learning_rate": 6.109910134927344e-07,
-      "loss": 0.2364,
+      "epoch": 4.773128626518182,
+      "grad_norm": 0.25789666175842285,
+      "learning_rate": 2.734151912950872e-07,
+      "loss": 0.3977,
       "step": 132440
     },
     {
-      "epoch": 4.66,
-      "learning_rate": 6.103652176812524e-07,
-      "loss": 0.2653,
+      "epoch": 4.773308826179407,
+      "grad_norm": 0.24202270805835724,
+      "learning_rate": 2.729849611859986e-07,
+      "loss": 0.3611,
       "step": 132445
     },
     {
-      "epoch": 4.66,
-      "learning_rate": 6.097397385531029e-07,
-      "loss": 0.2623,
+      "epoch": 4.7734890258406315,
+      "grad_norm": 0.26847487688064575,
+      "learning_rate": 2.7255506797770615e-07,
+      "loss": 0.3389,
       "step": 132450
     },
     {
-      "epoch": 4.66,
-      "learning_rate": 6.091145761164125e-07,
-      "loss": 0.264,
+      "epoch": 4.773669225501856,
+      "grad_norm": 0.23348930478096008,
+      "learning_rate": 2.721255116760607e-07,
+      "loss": 0.3824,
       "step": 132455
     },
     {
-      "epoch": 4.66,
-      "learning_rate": 6.084897303792914e-07,
-      "loss": 0.2523,
+      "epoch": 4.773849425163081,
+      "grad_norm": 0.23757703602313995,
+      "learning_rate": 2.7169629228691597e-07,
+      "loss": 0.3796,
       "step": 132460
     },
     {
-      "epoch": 4.66,
-      "learning_rate": 6.078652013498609e-07,
-      "loss": 0.2527,
+      "epoch": 4.774029624824306,
+      "grad_norm": 0.2990122139453888,
+      "learning_rate": 2.7126740981612555e-07,
+      "loss": 0.4045,
       "step": 132465
     },
     {
-      "epoch": 4.66,
-      "learning_rate": 6.07240989036223e-07,
-      "loss": 0.2597,
+      "epoch": 4.7742098244855296,
+      "grad_norm": 0.2594546675682068,
+      "learning_rate": 2.708388642695237e-07,
+      "loss": 0.3844,
       "step": 132470
     },
     {
-      "epoch": 4.66,
-      "learning_rate": 6.066170934464877e-07,
-      "loss": 0.2469,
+      "epoch": 4.774390024146754,
+      "grad_norm": 0.2652234435081482,
+      "learning_rate": 2.7041065565296117e-07,
+      "loss": 0.3808,
       "step": 132475
     },
     {
-      "epoch": 4.66,
-      "learning_rate": 6.059935145887485e-07,
-      "loss": 0.2414,
+      "epoch": 4.774570223807979,
+      "grad_norm": 0.22843249142169952,
+      "learning_rate": 2.6998278397226404e-07,
+      "loss": 0.3518,
       "step": 132480
     },
     {
-      "epoch": 4.66,
-      "learning_rate": 6.05370252471113e-07,
-      "loss": 0.2587,
+      "epoch": 4.774750423469204,
+      "grad_norm": 0.24594981968402863,
+      "learning_rate": 2.695552492332609e-07,
+      "loss": 0.3952,
       "step": 132485
     },
     {
-      "epoch": 4.66,
-      "learning_rate": 6.047473071016663e-07,
-      "loss": 0.26,
+      "epoch": 4.7749306231304285,
+      "grad_norm": 0.20578601956367493,
+      "learning_rate": 2.6912805144178046e-07,
+      "loss": 0.3949,
       "step": 132490
     },
     {
-      "epoch": 4.66,
-      "learning_rate": 6.041246784884991e-07,
-      "loss": 0.2351,
+      "epoch": 4.775110822791653,
+      "grad_norm": 0.22884593904018402,
+      "learning_rate": 2.6870119060364295e-07,
+      "loss": 0.366,
       "step": 132495
     },
     {
-      "epoch": 4.66,
-      "learning_rate": 6.035023666396938e-07,
-      "loss": 0.2501,
+      "epoch": 4.775291022452878,
+      "grad_norm": 0.2841814458370209,
+      "learning_rate": 2.682746667246633e-07,
+      "loss": 0.339,
       "step": 132500
     },
     {
-      "epoch": 4.66,
-      "eval_loss": 0.24850960075855255,
-      "eval_runtime": 10.5477,
-      "eval_samples_per_second": 9.481,
-      "eval_steps_per_second": 9.481,
+      "epoch": 4.775291022452878,
+      "eval_loss": 0.4288337230682373,
+      "eval_runtime": 3.5448,
+      "eval_samples_per_second": 28.21,
+      "eval_steps_per_second": 7.053,
       "step": 132500
     },
     {
-      "epoch": 4.66,
-      "learning_rate": 6.028803715633302e-07,
-      "loss": 0.2474,
+      "epoch": 4.775471222114103,
+      "grad_norm": 0.216835618019104,
+      "learning_rate": 2.678484798106534e-07,
+      "loss": 0.381,
       "step": 132505
     },
     {
-      "epoch": 4.66,
-      "learning_rate": 6.022586932674906e-07,
-      "loss": 0.2318,
+      "epoch": 4.775651421775327,
+      "grad_norm": 0.28074148297309875,
+      "learning_rate": 2.6742262986741706e-07,
+      "loss": 0.3725,
       "step": 132510
     },
     {
-      "epoch": 4.66,
-      "learning_rate": 6.016373317602436e-07,
-      "loss": 0.2425,
+      "epoch": 4.775831621436552,
+      "grad_norm": 0.3115323781967163,
+      "learning_rate": 2.669971169007607e-07,
+      "loss": 0.3841,
       "step": 132515
     },
     {
-      "epoch": 4.66,
-      "learning_rate": 6.010162870496522e-07,
-      "loss": 0.24,
+      "epoch": 4.776011821097776,
+      "grad_norm": 0.31163284182548523,
+      "learning_rate": 2.6657194091648243e-07,
+      "loss": 0.3959,
       "step": 132520
     },
     {
-      "epoch": 4.66,
-      "learning_rate": 6.003955591437904e-07,
-      "loss": 0.2605,
+      "epoch": 4.776192020759001,
+      "grad_norm": 0.25598570704460144,
+      "learning_rate": 2.661471019203693e-07,
+      "loss": 0.3576,
       "step": 132525
     },
     {
-      "epoch": 4.66,
-      "learning_rate": 5.997751480507102e-07,
-      "loss": 0.2381,
+      "epoch": 4.7763722204202255,
+      "grad_norm": 0.25508904457092285,
+      "learning_rate": 2.657225999182167e-07,
+      "loss": 0.3747,
       "step": 132530
     },
     {
-      "epoch": 4.66,
-      "learning_rate": 5.991550537784663e-07,
-      "loss": 0.2376,
+      "epoch": 4.77655242008145,
+      "grad_norm": 0.24659277498722076,
+      "learning_rate": 2.652984349158033e-07,
+      "loss": 0.3652,
       "step": 132535
     },
     {
-      "epoch": 4.66,
-      "learning_rate": 5.98535276335116e-07,
-      "loss": 0.2289,
+      "epoch": 4.776732619742675,
+      "grad_norm": 0.2608989179134369,
+      "learning_rate": 2.6487460691890787e-07,
+      "loss": 0.381,
       "step": 132540
     },
     {
-      "epoch": 4.66,
-      "learning_rate": 5.979158157287057e-07,
-      "loss": 0.2477,
+      "epoch": 4.7769128194039,
+      "grad_norm": 0.22622261941432953,
+      "learning_rate": 2.6445111593331187e-07,
+      "loss": 0.3443,
       "step": 132545
     },
     {
-      "epoch": 4.66,
-      "learning_rate": 5.972966719672762e-07,
-      "loss": 0.2415,
+      "epoch": 4.777093019065124,
+      "grad_norm": 0.23941785097122192,
+      "learning_rate": 2.640279619647773e-07,
+      "loss": 0.3482,
       "step": 132550
     },
     {
-      "epoch": 4.66,
-      "learning_rate": 5.966778450588628e-07,
-      "loss": 0.2521,
+      "epoch": 4.777273218726349,
+      "grad_norm": 0.25073572993278503,
+      "learning_rate": 2.6360514501907183e-07,
+      "loss": 0.3629,
       "step": 132555
     },
     {
-      "epoch": 4.66,
-      "learning_rate": 5.96059335011509e-07,
-      "loss": 0.2563,
+      "epoch": 4.777453418387573,
+      "grad_norm": 0.2828969359397888,
+      "learning_rate": 2.631826651019603e-07,
+      "loss": 0.3381,
       "step": 132560
     },
     {
-      "epoch": 4.66,
-      "learning_rate": 5.954411418332418e-07,
-      "loss": 0.255,
+      "epoch": 4.777633618048798,
+      "grad_norm": 0.29804477095603943,
+      "learning_rate": 2.627605222191937e-07,
+      "loss": 0.373,
       "step": 132565
     },
     {
-      "epoch": 4.66,
-      "learning_rate": 5.948232655320879e-07,
-      "loss": 0.2468,
+      "epoch": 4.7778138177100224,
+      "grad_norm": 0.21889689564704895,
+      "learning_rate": 2.6233871637652563e-07,
+      "loss": 0.3954,
       "step": 132570
     },
     {
-      "epoch": 4.66,
-      "learning_rate": 5.942057061160688e-07,
-      "loss": 0.267,
+      "epoch": 4.777994017371247,
+      "grad_norm": 0.32417771220207214,
+      "learning_rate": 2.619172475797044e-07,
+      "loss": 0.4157,
       "step": 132575
     },
     {
-      "epoch": 4.66,
-      "learning_rate": 5.935884635932087e-07,
-      "loss": 0.2616,
+      "epoch": 4.778174217032472,
+      "grad_norm": 0.24297624826431274,
+      "learning_rate": 2.6149611583446975e-07,
+      "loss": 0.374,
       "step": 132580
     },
     {
-      "epoch": 4.66,
-      "learning_rate": 5.929715379715151e-07,
-      "loss": 0.2581,
+      "epoch": 4.778354416693697,
+      "grad_norm": 0.2629621922969818,
+      "learning_rate": 2.6107532114656166e-07,
+      "loss": 0.3859,
       "step": 132585
     },
     {
-      "epoch": 4.66,
-      "learning_rate": 5.923549292590008e-07,
-      "loss": 0.2606,
+      "epoch": 4.778534616354921,
+      "grad_norm": 0.2679719924926758,
+      "learning_rate": 2.6065486352171153e-07,
+      "loss": 0.366,
       "step": 132590
     },
     {
-      "epoch": 4.67,
-      "learning_rate": 5.917386374636735e-07,
-      "loss": 0.2622,
+      "epoch": 4.778714816016146,
+      "grad_norm": 0.25725606083869934,
+      "learning_rate": 2.6023474296565096e-07,
+      "loss": 0.4017,
       "step": 132595
     },
     {
-      "epoch": 4.67,
-      "learning_rate": 5.911226625935351e-07,
-      "loss": 0.236,
+      "epoch": 4.778895015677371,
+      "grad_norm": 0.26632368564605713,
+      "learning_rate": 2.598149594841004e-07,
+      "loss": 0.412,
       "step": 132600
     },
     {
-      "epoch": 4.67,
-      "learning_rate": 5.905070046565819e-07,
-      "loss": 0.2401,
+      "epoch": 4.779075215338596,
+      "grad_norm": 0.20511886477470398,
+      "learning_rate": 2.59395513082783e-07,
+      "loss": 0.3194,
       "step": 132605
     },
     {
-      "epoch": 4.67,
-      "learning_rate": 5.898916636608077e-07,
-      "loss": 0.2449,
+      "epoch": 4.779255414999819,
+      "grad_norm": 0.26528292894363403,
+      "learning_rate": 2.589764037674108e-07,
+      "loss": 0.3529,
       "step": 132610
     },
     {
-      "epoch": 4.67,
-      "learning_rate": 5.892766396142058e-07,
-      "loss": 0.2527,
+      "epoch": 4.779435614661044,
+      "grad_norm": 0.2384311705827713,
+      "learning_rate": 2.5855763154369604e-07,
+      "loss": 0.3273,
       "step": 132615
     },
     {
-      "epoch": 4.67,
-      "learning_rate": 5.886619325247561e-07,
-      "loss": 0.271,
+      "epoch": 4.779615814322269,
+      "grad_norm": 0.28175267577171326,
+      "learning_rate": 2.581391964173424e-07,
+      "loss": 0.3562,
       "step": 132620
     },
     {
-      "epoch": 4.67,
-      "learning_rate": 5.880475424004467e-07,
-      "loss": 0.2544,
+      "epoch": 4.779796013983494,
+      "grad_norm": 0.2850888967514038,
+      "learning_rate": 2.5772109839405367e-07,
+      "loss": 0.39,
       "step": 132625
     },
     {
-      "epoch": 4.67,
-      "learning_rate": 5.874334692492517e-07,
-      "loss": 0.2472,
+      "epoch": 4.779976213644718,
+      "grad_norm": 0.2696197032928467,
+      "learning_rate": 2.573033374795225e-07,
+      "loss": 0.3607,
       "step": 132630
     },
     {
-      "epoch": 4.67,
-      "learning_rate": 5.868197130791397e-07,
-      "loss": 0.2511,
+      "epoch": 4.780156413305943,
+      "grad_norm": 0.2529081702232361,
+      "learning_rate": 2.5688591367944437e-07,
+      "loss": 0.3348,
       "step": 132635
     },
     {
-      "epoch": 4.67,
-      "learning_rate": 5.862062738980906e-07,
-      "loss": 0.2521,
+      "epoch": 4.780336612967168,
+      "grad_norm": 0.2832430601119995,
+      "learning_rate": 2.5646882699950634e-07,
+      "loss": 0.3933,
       "step": 132640
     },
     {
-      "epoch": 4.67,
-      "learning_rate": 5.855931517140617e-07,
-      "loss": 0.249,
+      "epoch": 4.780516812628393,
+      "grad_norm": 0.22554835677146912,
+      "learning_rate": 2.5605207744538727e-07,
+      "loss": 0.3706,
       "step": 132645
     },
     {
-      "epoch": 4.67,
-      "learning_rate": 5.849803465350162e-07,
-      "loss": 0.261,
+      "epoch": 4.780697012289616,
+      "grad_norm": 0.24301104247570038,
+      "learning_rate": 2.556356650227687e-07,
+      "loss": 0.3657,
       "step": 132650
     },
     {
-      "epoch": 4.67,
-      "learning_rate": 5.843678583689116e-07,
-      "loss": 0.2515,
+      "epoch": 4.780877211950841,
+      "grad_norm": 0.2747708559036255,
+      "learning_rate": 2.5521958973732383e-07,
+      "loss": 0.3806,
       "step": 132655
     },
     {
-      "epoch": 4.67,
-      "learning_rate": 5.837556872236999e-07,
-      "loss": 0.248,
+      "epoch": 4.781057411612066,
+      "grad_norm": 0.2549690008163452,
+      "learning_rate": 2.548038515947232e-07,
+      "loss": 0.3701,
       "step": 132660
     },
     {
-      "epoch": 4.67,
-      "learning_rate": 5.831438331073275e-07,
-      "loss": 0.2506,
+      "epoch": 4.781237611273291,
+      "grad_norm": 0.24783384799957275,
+      "learning_rate": 2.5438845060062887e-07,
+      "loss": 0.3806,
       "step": 132665
     },
     {
-      "epoch": 4.67,
-      "learning_rate": 5.825322960277435e-07,
-      "loss": 0.2607,
+      "epoch": 4.781417810934515,
+      "grad_norm": 0.2004653662443161,
+      "learning_rate": 2.5397338676069746e-07,
+      "loss": 0.3775,
       "step": 132670
     },
     {
-      "epoch": 4.67,
-      "learning_rate": 5.819210759928833e-07,
-      "loss": 0.2541,
+      "epoch": 4.78159801059574,
+      "grad_norm": 0.2571395933628082,
+      "learning_rate": 2.5355866008059117e-07,
+      "loss": 0.3793,
       "step": 132675
     },
     {
-      "epoch": 4.67,
-      "learning_rate": 5.813101730106879e-07,
-      "loss": 0.249,
+      "epoch": 4.781778210256965,
+      "grad_norm": 0.2348882555961609,
+      "learning_rate": 2.5314427056595537e-07,
+      "loss": 0.391,
       "step": 132680
     },
     {
-      "epoch": 4.67,
-      "learning_rate": 5.806995870890869e-07,
-      "loss": 0.2453,
+      "epoch": 4.7819584099181895,
+      "grad_norm": 0.26653140783309937,
+      "learning_rate": 2.527302182224384e-07,
+      "loss": 0.3878,
       "step": 132685
     },
     {
-      "epoch": 4.67,
-      "learning_rate": 5.800893182360073e-07,
-      "loss": 0.2662,
+      "epoch": 4.782138609579414,
+      "grad_norm": 0.2474595308303833,
+      "learning_rate": 2.523165030556829e-07,
+      "loss": 0.3423,
       "step": 132690
     },
     {
-      "epoch": 4.67,
-      "learning_rate": 5.794793664593762e-07,
-      "loss": 0.2356,
+      "epoch": 4.782318809240639,
+      "grad_norm": 0.2751745879650116,
+      "learning_rate": 2.519031250713205e-07,
+      "loss": 0.3803,
       "step": 132695
     },
     {
-      "epoch": 4.67,
-      "learning_rate": 5.788697317671149e-07,
-      "loss": 0.2409,
+      "epoch": 4.782499008901864,
+      "grad_norm": 0.27088475227355957,
+      "learning_rate": 2.5149008427498567e-07,
+      "loss": 0.3678,
       "step": 132700
     },
     {
-      "epoch": 4.67,
-      "learning_rate": 5.78260414167131e-07,
-      "loss": 0.2677,
+      "epoch": 4.782679208563088,
+      "grad_norm": 0.2260597050189972,
+      "learning_rate": 2.510773806723099e-07,
+      "loss": 0.3565,
       "step": 132705
     },
     {
-      "epoch": 4.67,
-      "learning_rate": 5.776514136673433e-07,
-      "loss": 0.257,
+      "epoch": 4.782859408224312,
+      "grad_norm": 0.23380564153194427,
+      "learning_rate": 2.5066501426891377e-07,
+      "loss": 0.3805,
       "step": 132710
     },
     {
-      "epoch": 4.67,
-      "learning_rate": 5.770427302756564e-07,
-      "loss": 0.2392,
+      "epoch": 4.783039607885537,
+      "grad_norm": 0.22910314798355103,
+      "learning_rate": 2.5025298507041494e-07,
+      "loss": 0.325,
       "step": 132715
     },
     {
-      "epoch": 4.67,
-      "learning_rate": 5.764343639999725e-07,
-      "loss": 0.2663,
+      "epoch": 4.783219807546762,
+      "grad_norm": 0.2516254782676697,
+      "learning_rate": 2.498412930824257e-07,
+      "loss": 0.372,
       "step": 132720
     },
     {
-      "epoch": 4.67,
-      "learning_rate": 5.758263148481963e-07,
-      "loss": 0.2483,
+      "epoch": 4.7834000072079865,
+      "grad_norm": 0.24351219832897186,
+      "learning_rate": 2.4942993831055803e-07,
+      "loss": 0.3611,
       "step": 132725
     },
     {
-      "epoch": 4.67,
-      "learning_rate": 5.752185828282159e-07,
-      "loss": 0.2377,
+      "epoch": 4.783580206869211,
+      "grad_norm": 0.3001359701156616,
+      "learning_rate": 2.49018920760416e-07,
+      "loss": 0.4113,
       "step": 132730
     },
     {
-      "epoch": 4.67,
-      "learning_rate": 5.746111679479277e-07,
-      "loss": 0.2485,
+      "epoch": 4.783760406530436,
+      "grad_norm": 0.256946861743927,
+      "learning_rate": 2.486082404375978e-07,
+      "loss": 0.3766,
       "step": 132735
     },
     {
-      "epoch": 4.67,
-      "learning_rate": 5.740040702152172e-07,
-      "loss": 0.2491,
+      "epoch": 4.783940606191661,
+      "grad_norm": 0.25258302688598633,
+      "learning_rate": 2.4819789734770174e-07,
+      "loss": 0.3731,
       "step": 132740
     },
     {
-      "epoch": 4.67,
-      "learning_rate": 5.733972896379669e-07,
-      "loss": 0.2482,
+      "epoch": 4.784120805852885,
+      "grad_norm": 0.24291817843914032,
+      "learning_rate": 2.4778789149631773e-07,
+      "loss": 0.3833,
       "step": 132745
     },
     {
-      "epoch": 4.67,
-      "learning_rate": 5.727908262240539e-07,
-      "loss": 0.2355,
+      "epoch": 4.784301005514109,
+      "grad_norm": 0.25857844948768616,
+      "learning_rate": 2.4737822288902756e-07,
+      "loss": 0.3858,
       "step": 132750
     },
     {
-      "epoch": 4.67,
-      "learning_rate": 5.721846799813524e-07,
-      "loss": 0.2578,
+      "epoch": 4.784481205175334,
+      "grad_norm": 0.276473730802536,
+      "learning_rate": 2.4696889153142386e-07,
+      "loss": 0.3602,
       "step": 132755
     },
     {
-      "epoch": 4.67,
-      "learning_rate": 5.715788509177367e-07,
-      "loss": 0.2684,
+      "epoch": 4.784661404836559,
+      "grad_norm": 0.2200738787651062,
+      "learning_rate": 2.465598974290717e-07,
+      "loss": 0.3473,
       "step": 132760
     },
     {
-      "epoch": 4.67,
-      "learning_rate": 5.709733390410698e-07,
-      "loss": 0.281,
+      "epoch": 4.7848416044977835,
+      "grad_norm": 0.22857971489429474,
+      "learning_rate": 2.461512405875471e-07,
+      "loss": 0.3452,
       "step": 132765
     },
     {
-      "epoch": 4.67,
-      "learning_rate": 5.703681443592152e-07,
-      "loss": 0.2661,
+      "epoch": 4.785021804159008,
+      "grad_norm": 0.2619878351688385,
+      "learning_rate": 2.4574292101242347e-07,
+      "loss": 0.3811,
       "step": 132770
     },
     {
-      "epoch": 4.67,
-      "learning_rate": 5.697632668800302e-07,
-      "loss": 0.2444,
+      "epoch": 4.785202003820233,
+      "grad_norm": 0.2593579888343811,
+      "learning_rate": 2.453349387092574e-07,
+      "loss": 0.3729,
       "step": 132775
     },
     {
-      "epoch": 4.67,
-      "learning_rate": 5.691587066113668e-07,
-      "loss": 0.2518,
+      "epoch": 4.785382203481458,
+      "grad_norm": 0.22074739634990692,
+      "learning_rate": 2.449272936836111e-07,
+      "loss": 0.3484,
       "step": 132780
     },
     {
-      "epoch": 4.67,
-      "learning_rate": 5.685544635610773e-07,
-      "loss": 0.256,
+      "epoch": 4.785562403142682,
+      "grad_norm": 0.25108104944229126,
+      "learning_rate": 2.4451998594103854e-07,
+      "loss": 0.3631,
       "step": 132785
     },
     {
-      "epoch": 4.67,
-      "learning_rate": 5.679505377370109e-07,
-      "loss": 0.2404,
+      "epoch": 4.785742602803907,
+      "grad_norm": 0.2209988385438919,
+      "learning_rate": 2.441130154870852e-07,
+      "loss": 0.3648,
       "step": 132790
     },
     {
-      "epoch": 4.67,
-      "learning_rate": 5.673469291470029e-07,
-      "loss": 0.2286,
+      "epoch": 4.785922802465131,
+      "grad_norm": 0.2671109735965729,
+      "learning_rate": 2.437063823273023e-07,
+      "loss": 0.3766,
       "step": 132795
     },
     {
-      "epoch": 4.67,
-      "learning_rate": 5.667436377988944e-07,
-      "loss": 0.2347,
+      "epoch": 4.786103002126356,
+      "grad_norm": 0.25873270630836487,
+      "learning_rate": 2.4330008646722693e-07,
+      "loss": 0.4221,
       "step": 132800
     },
     {
-      "epoch": 4.67,
-      "learning_rate": 5.661406637005151e-07,
-      "loss": 0.2382,
+      "epoch": 4.7862832017875805,
+      "grad_norm": 0.24569159746170044,
+      "learning_rate": 2.428941279123936e-07,
+      "loss": 0.349,
       "step": 132805
     },
     {
-      "epoch": 4.67,
-      "learning_rate": 5.655380068597005e-07,
-      "loss": 0.2313,
+      "epoch": 4.786463401448805,
+      "grad_norm": 0.22799314558506012,
+      "learning_rate": 2.424885066683341e-07,
+      "loss": 0.3881,
       "step": 132810
     },
     {
-      "epoch": 4.67,
-      "learning_rate": 5.649356672842721e-07,
-      "loss": 0.2519,
+      "epoch": 4.78664360111003,
+      "grad_norm": 0.258485347032547,
+      "learning_rate": 2.420832227405745e-07,
+      "loss": 0.3969,
       "step": 132815
     },
     {
-      "epoch": 4.67,
-      "learning_rate": 5.643336449820485e-07,
-      "loss": 0.2517,
+      "epoch": 4.786823800771255,
+      "grad_norm": 0.208878755569458,
+      "learning_rate": 2.416782761346409e-07,
+      "loss": 0.361,
       "step": 132820
     },
     {
-      "epoch": 4.67,
-      "learning_rate": 5.637319399608487e-07,
-      "loss": 0.2417,
+      "epoch": 4.787004000432479,
+      "grad_norm": 0.2455977350473404,
+      "learning_rate": 2.4127366685604834e-07,
+      "loss": 0.3483,
       "step": 132825
     },
     {
-      "epoch": 4.67,
-      "learning_rate": 5.631305522284858e-07,
-      "loss": 0.2561,
+      "epoch": 4.787184200093704,
+      "grad_norm": 0.25862938165664673,
+      "learning_rate": 2.408693949103036e-07,
+      "loss": 0.3815,
       "step": 132830
     },
     {
-      "epoch": 4.67,
-      "learning_rate": 5.625294817927673e-07,
-      "loss": 0.2748,
+      "epoch": 4.787364399754928,
+      "grad_norm": 0.2472279667854309,
+      "learning_rate": 2.4046546030292437e-07,
+      "loss": 0.3725,
       "step": 132835
     },
     {
-      "epoch": 4.67,
-      "learning_rate": 5.619287286615011e-07,
-      "loss": 0.2655,
+      "epoch": 4.787544599416153,
+      "grad_norm": 0.2628474533557892,
+      "learning_rate": 2.400618630394064e-07,
+      "loss": 0.3722,
       "step": 132840
     },
     {
-      "epoch": 4.67,
-      "learning_rate": 5.613282928424806e-07,
-      "loss": 0.2467,
+      "epoch": 4.7877247990773775,
+      "grad_norm": 0.2300945371389389,
+      "learning_rate": 2.3965860312525344e-07,
+      "loss": 0.3503,
       "step": 132845
     },
     {
-      "epoch": 4.67,
-      "learning_rate": 5.607281743435112e-07,
-      "loss": 0.2592,
+      "epoch": 4.787904998738602,
+      "grad_norm": 0.27947181463241577,
+      "learning_rate": 2.392556805659585e-07,
+      "loss": 0.3705,
       "step": 132850
     },
     {
-      "epoch": 4.67,
-      "learning_rate": 5.601283731723806e-07,
-      "loss": 0.2365,
+      "epoch": 4.788085198399827,
+      "grad_norm": 0.22649262845516205,
+      "learning_rate": 2.3885309536700874e-07,
+      "loss": 0.3519,
       "step": 132855
     },
     {
-      "epoch": 4.67,
-      "learning_rate": 5.595288893368744e-07,
-      "loss": 0.25,
+      "epoch": 4.788265398061052,
+      "grad_norm": 0.2503747045993805,
+      "learning_rate": 2.3845084753389426e-07,
+      "loss": 0.3673,
       "step": 132860
     },
     {
-      "epoch": 4.67,
-      "learning_rate": 5.589297228447781e-07,
-      "loss": 0.2686,
+      "epoch": 4.788445597722276,
+      "grad_norm": 0.2534830570220947,
+      "learning_rate": 2.3804893707209396e-07,
+      "loss": 0.3949,
       "step": 132865
     },
     {
-      "epoch": 4.67,
-      "learning_rate": 5.583308737038745e-07,
-      "loss": 0.2597,
+      "epoch": 4.788625797383501,
+      "grad_norm": 0.2691611349582672,
+      "learning_rate": 2.3764736398708133e-07,
+      "loss": 0.3847,
       "step": 132870
     },
     {
-      "epoch": 4.67,
-      "learning_rate": 5.577323419219349e-07,
-      "loss": 0.2482,
+      "epoch": 4.788805997044726,
+      "grad_norm": 0.23171448707580566,
+      "learning_rate": 2.3724612828432968e-07,
+      "loss": 0.3372,
       "step": 132875
     },
     {
-      "epoch": 4.68,
-      "learning_rate": 5.571341275067338e-07,
-      "loss": 0.2365,
+      "epoch": 4.788986196705951,
+      "grad_norm": 0.20922648906707764,
+      "learning_rate": 2.3684522996930137e-07,
+      "loss": 0.3914,
       "step": 132880
     },
     {
-      "epoch": 4.68,
-      "learning_rate": 5.565362304660371e-07,
-      "loss": 0.2679,
+      "epoch": 4.7891663963671744,
+      "grad_norm": 0.24962696433067322,
+      "learning_rate": 2.3644466904746697e-07,
+      "loss": 0.3608,
       "step": 132885
     },
     {
-      "epoch": 4.68,
-      "learning_rate": 5.559386508076053e-07,
-      "loss": 0.2361,
+      "epoch": 4.789346596028399,
+      "grad_norm": 0.3583545982837677,
+      "learning_rate": 2.3604444552427773e-07,
+      "loss": 0.3603,
       "step": 132890
     },
     {
-      "epoch": 4.68,
-      "learning_rate": 5.553413885392044e-07,
-      "loss": 0.2445,
+      "epoch": 4.789526795689624,
+      "grad_norm": 0.26689061522483826,
+      "learning_rate": 2.3564455940519037e-07,
+      "loss": 0.3372,
       "step": 132895
     },
     {
-      "epoch": 4.68,
-      "learning_rate": 5.547444436685866e-07,
-      "loss": 0.2383,
+      "epoch": 4.789706995350849,
+      "grad_norm": 0.24085666239261627,
+      "learning_rate": 2.3524501069565053e-07,
+      "loss": 0.3906,
       "step": 132900
     },
     {
-      "epoch": 4.68,
-      "learning_rate": 5.541478162035013e-07,
-      "loss": 0.2589,
+      "epoch": 4.789887195012073,
+      "grad_norm": 0.2830328345298767,
+      "learning_rate": 2.3484579940110385e-07,
+      "loss": 0.3628,
       "step": 132905
     },
     {
-      "epoch": 4.68,
-      "learning_rate": 5.535515061516977e-07,
-      "loss": 0.2598,
+      "epoch": 4.790067394673298,
+      "grad_norm": 0.30713436007499695,
+      "learning_rate": 2.3444692552698488e-07,
+      "loss": 0.3912,
       "step": 132910
     },
     {
-      "epoch": 4.68,
-      "learning_rate": 5.529555135209142e-07,
-      "loss": 0.2544,
+      "epoch": 4.790247594334523,
+      "grad_norm": 0.21235887706279755,
+      "learning_rate": 2.340483890787337e-07,
+      "loss": 0.3317,
       "step": 132915
     },
     {
-      "epoch": 4.68,
-      "learning_rate": 5.523598383188916e-07,
-      "loss": 0.2531,
+      "epoch": 4.790427793995748,
+      "grad_norm": 0.2915695011615753,
+      "learning_rate": 2.3365019006177936e-07,
+      "loss": 0.3714,
       "step": 132920
     },
     {
-      "epoch": 4.68,
-      "learning_rate": 5.517644805533684e-07,
-      "loss": 0.2508,
+      "epoch": 4.790607993656971,
+      "grad_norm": 0.29393500089645386,
+      "learning_rate": 2.3325232848154522e-07,
+      "loss": 0.3477,
       "step": 132925
     },
     {
-      "epoch": 4.68,
-      "learning_rate": 5.511694402320716e-07,
-      "loss": 0.2439,
+      "epoch": 4.790788193318196,
+      "grad_norm": 0.2399698942899704,
+      "learning_rate": 2.3285480434345474e-07,
+      "loss": 0.3784,
       "step": 132930
     },
     {
-      "epoch": 4.68,
-      "learning_rate": 5.505747173627257e-07,
-      "loss": 0.2572,
+      "epoch": 4.790968392979421,
+      "grad_norm": 0.25534698367118835,
+      "learning_rate": 2.324576176529203e-07,
+      "loss": 0.3805,
       "step": 132935
     },
     {
-      "epoch": 4.68,
-      "learning_rate": 5.499803119530522e-07,
-      "loss": 0.2283,
+      "epoch": 4.791148592640646,
+      "grad_norm": 0.23360496759414673,
+      "learning_rate": 2.3206076841535696e-07,
+      "loss": 0.3425,
       "step": 132940
     },
     {
-      "epoch": 4.68,
-      "learning_rate": 5.4938622401077e-07,
-      "loss": 0.2578,
+      "epoch": 4.79132879230187,
+      "grad_norm": 0.2366493046283722,
+      "learning_rate": 2.3166425663617154e-07,
+      "loss": 0.3638,
       "step": 132945
     },
     {
-      "epoch": 4.68,
-      "learning_rate": 5.487924535435979e-07,
-      "loss": 0.2574,
+      "epoch": 4.791508991963095,
+      "grad_norm": 0.23220838606357574,
+      "learning_rate": 2.3126808232076247e-07,
+      "loss": 0.367,
       "step": 132950
     },
     {
-      "epoch": 4.68,
-      "learning_rate": 5.481990005592435e-07,
-      "loss": 0.2252,
+      "epoch": 4.79168919162432,
+      "grad_norm": 0.22431600093841553,
+      "learning_rate": 2.30872245474531e-07,
+      "loss": 0.3759,
       "step": 132955
     },
     {
-      "epoch": 4.68,
-      "learning_rate": 5.476058650654065e-07,
-      "loss": 0.2651,
+      "epoch": 4.791869391285545,
+      "grad_norm": 0.24509382247924805,
+      "learning_rate": 2.3047674610287007e-07,
+      "loss": 0.3951,
       "step": 132960
     },
     {
-      "epoch": 4.68,
-      "learning_rate": 5.470130470697943e-07,
-      "loss": 0.2429,
+      "epoch": 4.792049590946769,
+      "grad_norm": 0.3701198101043701,
+      "learning_rate": 2.3008158421116977e-07,
+      "loss": 0.3721,
       "step": 132965
     },
     {
-      "epoch": 4.68,
-      "learning_rate": 5.464205465801037e-07,
-      "loss": 0.2431,
+      "epoch": 4.792229790607994,
+      "grad_norm": 0.2326425015926361,
+      "learning_rate": 2.2968675980480913e-07,
+      "loss": 0.3701,
       "step": 132970
     },
     {
-      "epoch": 4.68,
-      "learning_rate": 5.458283636040257e-07,
-      "loss": 0.2552,
+      "epoch": 4.792409990269219,
+      "grad_norm": 0.25888702273368835,
+      "learning_rate": 2.2929227288917278e-07,
+      "loss": 0.3776,
       "step": 132975
     },
     {
-      "epoch": 4.68,
-      "learning_rate": 5.452364981492486e-07,
-      "loss": 0.2313,
+      "epoch": 4.792590189930443,
+      "grad_norm": 0.2808634638786316,
+      "learning_rate": 2.2889812346963134e-07,
+      "loss": 0.3671,
       "step": 132980
     },
     {
-      "epoch": 4.68,
-      "learning_rate": 5.446449502234635e-07,
-      "loss": 0.246,
+      "epoch": 4.792770389591667,
+      "grad_norm": 0.24987556040287018,
+      "learning_rate": 2.2850431155156116e-07,
+      "loss": 0.3775,
       "step": 132985
     },
     {
-      "epoch": 4.68,
-      "learning_rate": 5.440537198343449e-07,
-      "loss": 0.2323,
+      "epoch": 4.792950589252892,
+      "grad_norm": 0.3451884388923645,
+      "learning_rate": 2.2811083714032177e-07,
+      "loss": 0.3949,
       "step": 132990
     },
     {
-      "epoch": 4.68,
-      "learning_rate": 5.434628069895698e-07,
-      "loss": 0.2413,
+      "epoch": 4.793130788914117,
+      "grad_norm": 0.23816993832588196,
+      "learning_rate": 2.2771770024127559e-07,
+      "loss": 0.3669,
       "step": 132995
     },
     {
-      "epoch": 4.68,
-      "learning_rate": 5.428722116968127e-07,
-      "loss": 0.263,
+      "epoch": 4.7933109885753415,
+      "grad_norm": 0.28144383430480957,
+      "learning_rate": 2.2732490085977943e-07,
+      "loss": 0.3463,
       "step": 133000
     },
     {
-      "epoch": 4.68,
-      "eval_loss": 0.24850255250930786,
-      "eval_runtime": 10.5511,
-      "eval_samples_per_second": 9.478,
-      "eval_steps_per_second": 9.478,
+      "epoch": 4.7933109885753415,
+      "eval_loss": 0.428801029920578,
+      "eval_runtime": 3.5412,
+      "eval_samples_per_second": 28.239,
+      "eval_steps_per_second": 7.06,
       "step": 133000
     },
     {
-      "epoch": 4.68,
-      "learning_rate": 5.422819339637453e-07,
-      "loss": 0.2391,
+      "epoch": 4.793491188236566,
+      "grad_norm": 0.29074543714523315,
+      "learning_rate": 2.2693243900118454e-07,
+      "loss": 0.3782,
       "step": 133005
     },
     {
-      "epoch": 4.68,
-      "learning_rate": 5.41691973798028e-07,
-      "loss": 0.2412,
+      "epoch": 4.793671387897791,
+      "grad_norm": 0.25907793641090393,
+      "learning_rate": 2.2654031467084225e-07,
+      "loss": 0.3529,
       "step": 133010
     },
     {
-      "epoch": 4.68,
-      "learning_rate": 5.411023312073188e-07,
-      "loss": 0.2447,
+      "epoch": 4.793851587559016,
+      "grad_norm": 0.2522429823875427,
+      "learning_rate": 2.261485278740899e-07,
+      "loss": 0.3729,
       "step": 133015
     },
     {
-      "epoch": 4.68,
-      "learning_rate": 5.405130061992808e-07,
-      "loss": 0.2358,
+      "epoch": 4.79403178722024,
+      "grad_norm": 0.2468443363904953,
+      "learning_rate": 2.257570786162677e-07,
+      "loss": 0.3641,
       "step": 133020
     },
     {
-      "epoch": 4.68,
-      "learning_rate": 5.399239987815607e-07,
-      "loss": 0.2476,
+      "epoch": 4.794211986881464,
+      "grad_norm": 0.2643374502658844,
+      "learning_rate": 2.2536596690270751e-07,
+      "loss": 0.3705,
       "step": 133025
     },
     {
-      "epoch": 4.68,
-      "learning_rate": 5.393353089618081e-07,
-      "loss": 0.2601,
+      "epoch": 4.794392186542689,
+      "grad_norm": 0.34247177839279175,
+      "learning_rate": 2.2497519273874113e-07,
+      "loss": 0.3723,
       "step": 133030
     },
     {
-      "epoch": 4.68,
-      "learning_rate": 5.387469367476638e-07,
-      "loss": 0.2679,
+      "epoch": 4.794572386203914,
+      "grad_norm": 0.28834789991378784,
+      "learning_rate": 2.245847561296921e-07,
+      "loss": 0.367,
       "step": 133035
     },
     {
-      "epoch": 4.68,
-      "learning_rate": 5.38158882146772e-07,
-      "loss": 0.2602,
+      "epoch": 4.7947525858651385,
+      "grad_norm": 0.2463623285293579,
+      "learning_rate": 2.241946570808756e-07,
+      "loss": 0.3654,
       "step": 133040
     },
     {
-      "epoch": 4.68,
-      "learning_rate": 5.375711451667653e-07,
-      "loss": 0.2363,
+      "epoch": 4.794932785526363,
+      "grad_norm": 0.24550725519657135,
+      "learning_rate": 2.2380489559761242e-07,
+      "loss": 0.3917,
       "step": 133045
     },
     {
-      "epoch": 4.68,
-      "learning_rate": 5.369837258152766e-07,
-      "loss": 0.2532,
+      "epoch": 4.795112985187588,
+      "grad_norm": 0.23969918489456177,
+      "learning_rate": 2.2341547168521215e-07,
+      "loss": 0.3357,
       "step": 133050
     },
     {
-      "epoch": 4.68,
-      "learning_rate": 5.363966240999329e-07,
-      "loss": 0.2625,
+      "epoch": 4.795293184848813,
+      "grad_norm": 0.1967889666557312,
+      "learning_rate": 2.230263853489789e-07,
+      "loss": 0.3329,
       "step": 133055
     },
     {
-      "epoch": 4.68,
-      "learning_rate": 5.358098400283534e-07,
-      "loss": 0.2409,
+      "epoch": 4.7954733845100375,
+      "grad_norm": 0.2610912322998047,
+      "learning_rate": 2.2263763659421123e-07,
+      "loss": 0.3567,
       "step": 133060
     },
     {
-      "epoch": 4.68,
-      "learning_rate": 5.352233736081652e-07,
-      "loss": 0.2583,
+      "epoch": 4.795653584171262,
+      "grad_norm": 0.26562637090682983,
+      "learning_rate": 2.222492254262104e-07,
+      "loss": 0.348,
       "step": 133065
     },
     {
-      "epoch": 4.68,
-      "learning_rate": 5.346372248469761e-07,
-      "loss": 0.2374,
+      "epoch": 4.795833783832486,
+      "grad_norm": 0.2973119020462036,
+      "learning_rate": 2.2186115185026668e-07,
+      "loss": 0.4043,
       "step": 133070
     },
     {
-      "epoch": 4.68,
-      "learning_rate": 5.340513937523994e-07,
-      "loss": 0.2515,
+      "epoch": 4.796013983493711,
+      "grad_norm": 0.28032979369163513,
+      "learning_rate": 2.2147341587166748e-07,
+      "loss": 0.3818,
       "step": 133075
     },
     {
-      "epoch": 4.68,
-      "learning_rate": 5.33465880332043e-07,
-      "loss": 0.2455,
+      "epoch": 4.7961941831549355,
+      "grad_norm": 0.2844066917896271,
+      "learning_rate": 2.2108601749569467e-07,
+      "loss": 0.3509,
       "step": 133080
     },
     {
-      "epoch": 4.68,
-      "learning_rate": 5.328806845935091e-07,
-      "loss": 0.2345,
+      "epoch": 4.79637438281616,
+      "grad_norm": 0.25133219361305237,
+      "learning_rate": 2.206989567276302e-07,
+      "loss": 0.3783,
       "step": 133085
     },
     {
-      "epoch": 4.68,
-      "learning_rate": 5.322958065443944e-07,
-      "loss": 0.2455,
+      "epoch": 4.796554582477385,
+      "grad_norm": 0.2511965036392212,
+      "learning_rate": 2.2031223357274477e-07,
+      "loss": 0.3864,
       "step": 133090
     },
     {
-      "epoch": 4.68,
-      "learning_rate": 5.317112461922902e-07,
-      "loss": 0.2656,
+      "epoch": 4.79673478213861,
+      "grad_norm": 0.27423515915870667,
+      "learning_rate": 2.1992584803630368e-07,
+      "loss": 0.3573,
       "step": 133095
     },
     {
-      "epoch": 4.68,
-      "learning_rate": 5.311270035447957e-07,
-      "loss": 0.2337,
+      "epoch": 4.796914981799834,
+      "grad_norm": 0.24834780395030975,
+      "learning_rate": 2.1953980012357767e-07,
+      "loss": 0.3601,
       "step": 133100
     },
     {
-      "epoch": 4.68,
-      "learning_rate": 5.305430786094856e-07,
-      "loss": 0.2505,
+      "epoch": 4.797095181461059,
+      "grad_norm": 0.20566605031490326,
+      "learning_rate": 2.1915408983982643e-07,
+      "loss": 0.3793,
       "step": 133105
     },
     {
-      "epoch": 4.68,
-      "learning_rate": 5.29959471393951e-07,
-      "loss": 0.2346,
+      "epoch": 4.797275381122283,
+      "grad_norm": 0.30788904428482056,
+      "learning_rate": 2.1876871719030135e-07,
+      "loss": 0.3647,
       "step": 133110
     },
     {
-      "epoch": 4.68,
-      "learning_rate": 5.293761819057663e-07,
-      "loss": 0.2234,
+      "epoch": 4.797455580783508,
+      "grad_norm": 0.2940426766872406,
+      "learning_rate": 2.1838368218025374e-07,
+      "loss": 0.3761,
       "step": 133115
     },
     {
-      "epoch": 4.68,
-      "learning_rate": 5.287932101525061e-07,
-      "loss": 0.2616,
+      "epoch": 4.7976357804447325,
+      "grad_norm": 0.24170757830142975,
+      "learning_rate": 2.1799898481493219e-07,
+      "loss": 0.3499,
       "step": 133120
     },
     {
-      "epoch": 4.68,
-      "learning_rate": 5.282105561417366e-07,
-      "loss": 0.2709,
+      "epoch": 4.797815980105957,
+      "grad_norm": 0.19300325214862823,
+      "learning_rate": 2.1761462509957698e-07,
+      "loss": 0.3965,
       "step": 133125
     },
     {
-      "epoch": 4.68,
-      "learning_rate": 5.276282198810267e-07,
-      "loss": 0.2789,
+      "epoch": 4.797996179767182,
+      "grad_norm": 0.2832929491996765,
+      "learning_rate": 2.1723060303942278e-07,
+      "loss": 0.3731,
       "step": 133130
     },
     {
-      "epoch": 4.68,
-      "learning_rate": 5.270462013779371e-07,
-      "loss": 0.2462,
+      "epoch": 4.798176379428407,
+      "grad_norm": 0.2704760730266571,
+      "learning_rate": 2.1684691863970152e-07,
+      "loss": 0.3728,
       "step": 133135
     },
     {
-      "epoch": 4.68,
-      "learning_rate": 5.264645006400254e-07,
-      "loss": 0.2529,
+      "epoch": 4.798356579089631,
+      "grad_norm": 0.2666027247905731,
+      "learning_rate": 2.1646357190564514e-07,
+      "loss": 0.3435,
       "step": 133140
     },
     {
-      "epoch": 4.68,
-      "learning_rate": 5.258831176748413e-07,
-      "loss": 0.2318,
+      "epoch": 4.798536778750856,
+      "grad_norm": 0.2326856553554535,
+      "learning_rate": 2.1608056284247725e-07,
+      "loss": 0.3331,
       "step": 133145
     },
     {
-      "epoch": 4.68,
-      "learning_rate": 5.253020524899343e-07,
-      "loss": 0.2744,
+      "epoch": 4.798716978412081,
+      "grad_norm": 0.23978599905967712,
+      "learning_rate": 2.1569789145541031e-07,
+      "loss": 0.3723,
       "step": 133150
     },
     {
-      "epoch": 4.68,
-      "learning_rate": 5.248374291488805e-07,
-      "loss": 0.2483,
+      "epoch": 4.798897178073306,
+      "grad_norm": 0.2229306995868683,
+      "learning_rate": 2.1531555774966238e-07,
+      "loss": 0.3922,
       "step": 133155
     },
     {
-      "epoch": 4.68,
-      "learning_rate": 5.242569359874882e-07,
-      "loss": 0.235,
+      "epoch": 4.79907737773453,
+      "grad_norm": 0.26587215065956116,
+      "learning_rate": 2.1493356173043765e-07,
+      "loss": 0.3805,
       "step": 133160
     },
     {
-      "epoch": 4.69,
-      "learning_rate": 5.236767606274867e-07,
-      "loss": 0.2279,
+      "epoch": 4.799257577395754,
+      "grad_norm": 0.21258600056171417,
+      "learning_rate": 2.1455190340294863e-07,
+      "loss": 0.3379,
       "step": 133165
     },
     {
-      "epoch": 4.69,
-      "learning_rate": 5.230969030764143e-07,
-      "loss": 0.2417,
+      "epoch": 4.799437777056979,
+      "grad_norm": 0.2616501450538635,
+      "learning_rate": 2.141705827723911e-07,
+      "loss": 0.3754,
       "step": 133170
     },
     {
-      "epoch": 4.69,
-      "learning_rate": 5.225173633417985e-07,
-      "loss": 0.2398,
+      "epoch": 4.799617976718204,
+      "grad_norm": 0.21897155046463013,
+      "learning_rate": 2.1378959984395818e-07,
+      "loss": 0.3835,
       "step": 133175
     },
     {
-      "epoch": 4.69,
-      "learning_rate": 5.219381414311636e-07,
-      "loss": 0.2554,
+      "epoch": 4.799798176379428,
+      "grad_norm": 0.26772749423980713,
+      "learning_rate": 2.1340895462284571e-07,
+      "loss": 0.3344,
       "step": 133180
     },
     {
-      "epoch": 4.69,
-      "learning_rate": 5.21359237352026e-07,
-      "loss": 0.2657,
+      "epoch": 4.799978376040653,
+      "grad_norm": 0.3039627969264984,
+      "learning_rate": 2.1302864711423564e-07,
+      "loss": 0.4117,
       "step": 133185
     },
     {
-      "epoch": 4.69,
-      "learning_rate": 5.207806511119101e-07,
-      "loss": 0.2319,
+      "epoch": 4.800158575701878,
+      "grad_norm": 0.27932974696159363,
+      "learning_rate": 2.1264867732331272e-07,
+      "loss": 0.3639,
       "step": 133190
     },
     {
-      "epoch": 4.69,
-      "learning_rate": 5.202023827183239e-07,
-      "loss": 0.242,
+      "epoch": 4.800338775363103,
+      "grad_norm": 0.23750893771648407,
+      "learning_rate": 2.1226904525525336e-07,
+      "loss": 0.3618,
       "step": 133195
     },
     {
-      "epoch": 4.69,
-      "learning_rate": 5.19624432178778e-07,
-      "loss": 0.2349,
+      "epoch": 4.800518975024327,
+      "grad_norm": 0.26062673330307007,
+      "learning_rate": 2.1188975091522568e-07,
+      "loss": 0.372,
       "step": 133200
     },
     {
-      "epoch": 4.69,
-      "learning_rate": 5.190467995007747e-07,
-      "loss": 0.2436,
+      "epoch": 4.800699174685551,
+      "grad_norm": 0.250861257314682,
+      "learning_rate": 2.1151079430840325e-07,
+      "loss": 0.393,
       "step": 133205
     },
     {
-      "epoch": 4.69,
-      "learning_rate": 5.184694846918164e-07,
-      "loss": 0.2547,
+      "epoch": 4.800879374346776,
+      "grad_norm": 0.2910214364528656,
+      "learning_rate": 2.111321754399487e-07,
+      "loss": 0.3412,
       "step": 133210
     },
     {
-      "epoch": 4.69,
-      "learning_rate": 5.178924877593971e-07,
-      "loss": 0.2507,
+      "epoch": 4.801059574008001,
+      "grad_norm": 0.23345068097114563,
+      "learning_rate": 2.1075389431501613e-07,
+      "loss": 0.3914,
       "step": 133215
     },
     {
-      "epoch": 4.69,
-      "learning_rate": 5.17315808711008e-07,
-      "loss": 0.2255,
+      "epoch": 4.801239773669225,
+      "grad_norm": 0.22395312786102295,
+      "learning_rate": 2.1037595093876538e-07,
+      "loss": 0.3657,
       "step": 133220
     },
     {
-      "epoch": 4.69,
-      "learning_rate": 5.167394475541404e-07,
-      "loss": 0.2469,
+      "epoch": 4.80141997333045,
+      "grad_norm": 0.21458613872528076,
+      "learning_rate": 2.0999834531633955e-07,
+      "loss": 0.3573,
       "step": 133225
     },
     {
-      "epoch": 4.69,
-      "learning_rate": 5.16163404296277e-07,
-      "loss": 0.2552,
+      "epoch": 4.801600172991675,
+      "grad_norm": 0.2688688635826111,
+      "learning_rate": 2.0962107745288727e-07,
+      "loss": 0.3631,
       "step": 133230
     },
     {
-      "epoch": 4.69,
-      "learning_rate": 5.155876789448955e-07,
-      "loss": 0.2525,
+      "epoch": 4.8017803726529,
+      "grad_norm": 0.2928563952445984,
+      "learning_rate": 2.0924414735355168e-07,
+      "loss": 0.3654,
       "step": 133235
     },
     {
-      "epoch": 4.69,
-      "learning_rate": 5.150122715074701e-07,
-      "loss": 0.2423,
+      "epoch": 4.801960572314124,
+      "grad_norm": 0.360637903213501,
+      "learning_rate": 2.08867555023462e-07,
+      "loss": 0.3995,
       "step": 133240
     },
     {
-      "epoch": 4.69,
-      "learning_rate": 5.144371819914757e-07,
-      "loss": 0.2567,
+      "epoch": 4.802140771975349,
+      "grad_norm": 0.3004918694496155,
+      "learning_rate": 2.08491300467753e-07,
+      "loss": 0.3326,
       "step": 133245
     },
     {
-      "epoch": 4.69,
-      "learning_rate": 5.138624104043754e-07,
-      "loss": 0.2347,
+      "epoch": 4.802320971636574,
+      "grad_norm": 0.25441572070121765,
+      "learning_rate": 2.0811538369155115e-07,
+      "loss": 0.4041,
       "step": 133250
     },
     {
-      "epoch": 4.69,
-      "learning_rate": 5.132879567536386e-07,
-      "loss": 0.2455,
+      "epoch": 4.802501171297798,
+      "grad_norm": 0.22652685642242432,
+      "learning_rate": 2.0773980469997456e-07,
+      "loss": 0.3531,
       "step": 133255
     },
     {
-      "epoch": 4.69,
-      "learning_rate": 5.127138210467175e-07,
-      "loss": 0.2377,
+      "epoch": 4.802681370959022,
+      "grad_norm": 0.2670671343803406,
+      "learning_rate": 2.0736456349814416e-07,
+      "loss": 0.3764,
       "step": 133260
     },
     {
-      "epoch": 4.69,
-      "learning_rate": 5.1214000329107e-07,
-      "loss": 0.2489,
+      "epoch": 4.802861570620247,
+      "grad_norm": 0.2373109608888626,
+      "learning_rate": 2.069896600911725e-07,
+      "loss": 0.3733,
       "step": 133265
     },
     {
-      "epoch": 4.69,
-      "learning_rate": 5.115665034941486e-07,
-      "loss": 0.2669,
+      "epoch": 4.803041770281472,
+      "grad_norm": 0.2623651921749115,
+      "learning_rate": 2.066150944841666e-07,
+      "loss": 0.3785,
       "step": 133270
     },
     {
-      "epoch": 4.69,
-      "learning_rate": 5.109933216633972e-07,
-      "loss": 0.2368,
+      "epoch": 4.803221969942697,
+      "grad_norm": 0.26846858859062195,
+      "learning_rate": 2.0624086668223074e-07,
+      "loss": 0.366,
       "step": 133275
     },
     {
-      "epoch": 4.69,
-      "learning_rate": 5.104204578062571e-07,
-      "loss": 0.2481,
+      "epoch": 4.803402169603921,
+      "grad_norm": 0.2998198866844177,
+      "learning_rate": 2.0586697669046083e-07,
+      "loss": 0.399,
       "step": 133280
     },
     {
-      "epoch": 4.69,
-      "learning_rate": 5.098479119301697e-07,
-      "loss": 0.2422,
+      "epoch": 4.803582369265146,
+      "grad_norm": 0.29226166009902954,
+      "learning_rate": 2.054934245139528e-07,
+      "loss": 0.3444,
       "step": 133285
     },
     {
-      "epoch": 4.69,
-      "learning_rate": 5.092756840425678e-07,
-      "loss": 0.257,
+      "epoch": 4.803762568926371,
+      "grad_norm": 0.24952854216098785,
+      "learning_rate": 2.0512021015779704e-07,
+      "loss": 0.3419,
       "step": 133290
     },
     {
-      "epoch": 4.69,
-      "learning_rate": 5.087037741508816e-07,
-      "loss": 0.2429,
+      "epoch": 4.803942768587595,
+      "grad_norm": 0.28424960374832153,
+      "learning_rate": 2.0474733362708109e-07,
+      "loss": 0.3488,
       "step": 133295
     },
     {
-      "epoch": 4.69,
-      "learning_rate": 5.081321822625329e-07,
-      "loss": 0.2547,
+      "epoch": 4.804122968248819,
+      "grad_norm": 0.3396574854850769,
+      "learning_rate": 2.0437479492687872e-07,
+      "loss": 0.3507,
       "step": 133300
     },
     {
-      "epoch": 4.69,
-      "learning_rate": 5.075609083849491e-07,
-      "loss": 0.2409,
+      "epoch": 4.804303167910044,
+      "grad_norm": 0.23657868802547455,
+      "learning_rate": 2.040025940622692e-07,
+      "loss": 0.3767,
       "step": 133305
     },
     {
-      "epoch": 4.69,
-      "learning_rate": 5.069899525255467e-07,
-      "loss": 0.2559,
+      "epoch": 4.804483367571269,
+      "grad_norm": 0.24988289177417755,
+      "learning_rate": 2.0363073103832619e-07,
+      "loss": 0.3637,
       "step": 133310
     },
     {
-      "epoch": 4.69,
-      "learning_rate": 5.064193146917362e-07,
-      "loss": 0.2601,
+      "epoch": 4.8046635672324935,
+      "grad_norm": 0.26382821798324585,
+      "learning_rate": 2.032592058601124e-07,
+      "loss": 0.3908,
       "step": 133315
     },
     {
-      "epoch": 4.69,
-      "learning_rate": 5.058489948909312e-07,
-      "loss": 0.24,
+      "epoch": 4.804843766893718,
+      "grad_norm": 0.24887283146381378,
+      "learning_rate": 2.0288801853269036e-07,
+      "loss": 0.4073,
       "step": 133320
     },
     {
-      "epoch": 4.69,
-      "learning_rate": 5.052789931305341e-07,
-      "loss": 0.2504,
+      "epoch": 4.805023966554943,
+      "grad_norm": 0.27872660756111145,
+      "learning_rate": 2.0251716906111718e-07,
+      "loss": 0.3677,
       "step": 133325
     },
     {
-      "epoch": 4.69,
-      "learning_rate": 5.047093094179472e-07,
-      "loss": 0.2488,
+      "epoch": 4.805204166216168,
+      "grad_norm": 0.22657455503940582,
+      "learning_rate": 2.0214665745044714e-07,
+      "loss": 0.3901,
       "step": 133330
     },
     {
-      "epoch": 4.69,
-      "learning_rate": 5.041399437605649e-07,
-      "loss": 0.251,
+      "epoch": 4.8053843658773925,
+      "grad_norm": 0.22239211201667786,
+      "learning_rate": 2.0177648370573176e-07,
+      "loss": 0.3766,
       "step": 133335
     },
     {
-      "epoch": 4.69,
-      "learning_rate": 5.035708961657836e-07,
-      "loss": 0.2721,
+      "epoch": 4.805564565538617,
+      "grad_norm": 0.28701508045196533,
+      "learning_rate": 2.014066478320087e-07,
+      "loss": 0.4004,
       "step": 133340
     },
     {
-      "epoch": 4.69,
-      "learning_rate": 5.030021666409895e-07,
-      "loss": 0.2156,
+      "epoch": 4.805744765199841,
+      "grad_norm": 0.31527572870254517,
+      "learning_rate": 2.0103714983431553e-07,
+      "loss": 0.3964,
       "step": 133345
     },
     {
-      "epoch": 4.69,
-      "learning_rate": 5.02433755193571e-07,
-      "loss": 0.2312,
+      "epoch": 4.805924964861066,
+      "grad_norm": 0.27578893303871155,
+      "learning_rate": 2.0066798971769273e-07,
+      "loss": 0.3874,
       "step": 133350
     },
     {
-      "epoch": 4.69,
-      "learning_rate": 5.018656618309025e-07,
-      "loss": 0.266,
+      "epoch": 4.8061051645222905,
+      "grad_norm": 0.25750282406806946,
+      "learning_rate": 2.0029916748716683e-07,
+      "loss": 0.3622,
       "step": 133355
     },
     {
-      "epoch": 4.69,
-      "learning_rate": 5.012978865603646e-07,
-      "loss": 0.2396,
+      "epoch": 4.806285364183515,
+      "grad_norm": 0.2739262580871582,
+      "learning_rate": 1.999306831477643e-07,
+      "loss": 0.3426,
       "step": 133360
     },
     {
-      "epoch": 4.69,
-      "learning_rate": 5.007304293893262e-07,
-      "loss": 0.2459,
+      "epoch": 4.80646556384474,
+      "grad_norm": 0.25903549790382385,
+      "learning_rate": 1.995625367045062e-07,
+      "loss": 0.3852,
       "step": 133365
     },
     {
-      "epoch": 4.69,
-      "learning_rate": 5.001632903251619e-07,
-      "loss": 0.2543,
+      "epoch": 4.806645763505965,
+      "grad_norm": 0.19403640925884247,
+      "learning_rate": 1.9919472816240237e-07,
+      "loss": 0.3446,
       "step": 133370
     },
     {
-      "epoch": 4.69,
-      "learning_rate": 4.99596469375227e-07,
-      "loss": 0.2621,
+      "epoch": 4.8068259631671895,
+      "grad_norm": 0.22920994460582733,
+      "learning_rate": 1.9882725752647102e-07,
+      "loss": 0.3628,
       "step": 133375
     },
     {
-      "epoch": 4.69,
-      "learning_rate": 4.990299665468879e-07,
-      "loss": 0.2688,
+      "epoch": 4.807006162828414,
+      "grad_norm": 0.24877864122390747,
+      "learning_rate": 1.984601248017165e-07,
+      "loss": 0.3912,
       "step": 133380
     },
     {
-      "epoch": 4.69,
-      "learning_rate": 4.984637818474996e-07,
-      "loss": 0.2269,
+      "epoch": 4.807186362489638,
+      "grad_norm": 0.22861328721046448,
+      "learning_rate": 1.9809332999314311e-07,
+      "loss": 0.3505,
       "step": 133385
     },
     {
-      "epoch": 4.69,
-      "learning_rate": 4.978979152844093e-07,
-      "loss": 0.2381,
+      "epoch": 4.807366562150863,
+      "grad_norm": 0.2941596806049347,
+      "learning_rate": 1.9772687310574412e-07,
+      "loss": 0.3911,
       "step": 133390
     },
     {
-      "epoch": 4.69,
-      "learning_rate": 4.973323668649665e-07,
-      "loss": 0.2326,
+      "epoch": 4.8075467618120875,
+      "grad_norm": 0.24592210352420807,
+      "learning_rate": 1.973607541445155e-07,
+      "loss": 0.4009,
       "step": 133395
     },
     {
-      "epoch": 4.69,
-      "learning_rate": 4.967671365965181e-07,
-      "loss": 0.276,
+      "epoch": 4.807726961473312,
+      "grad_norm": 0.26915445923805237,
+      "learning_rate": 1.9699497311444493e-07,
+      "loss": 0.381,
       "step": 133400
     },
     {
-      "epoch": 4.69,
-      "learning_rate": 4.962022244863973e-07,
-      "loss": 0.2454,
+      "epoch": 4.807907161134537,
+      "grad_norm": 0.2586546838283539,
+      "learning_rate": 1.9662953002051455e-07,
+      "loss": 0.3685,
       "step": 133405
     },
     {
-      "epoch": 4.69,
-      "learning_rate": 4.956376305419425e-07,
-      "loss": 0.2617,
+      "epoch": 4.808087360795762,
+      "grad_norm": 0.21903952956199646,
+      "learning_rate": 1.9626442486770647e-07,
+      "loss": 0.4036,
       "step": 133410
     },
     {
-      "epoch": 4.69,
-      "learning_rate": 4.950733547704866e-07,
-      "loss": 0.2602,
+      "epoch": 4.8082675604569864,
+      "grad_norm": 0.2535596191883087,
+      "learning_rate": 1.9589965766099172e-07,
+      "loss": 0.3819,
       "step": 133415
     },
     {
-      "epoch": 4.69,
-      "learning_rate": 4.94509397179349e-07,
-      "loss": 0.2497,
+      "epoch": 4.808447760118211,
+      "grad_norm": 0.28337281942367554,
+      "learning_rate": 1.9553522840534133e-07,
+      "loss": 0.3731,
       "step": 133420
     },
     {
-      "epoch": 4.69,
-      "learning_rate": 4.939457577758627e-07,
-      "loss": 0.2693,
+      "epoch": 4.808627959779436,
+      "grad_norm": 0.2324288785457611,
+      "learning_rate": 1.9517113710572354e-07,
+      "loss": 0.3678,
       "step": 133425
     },
     {
-      "epoch": 4.69,
-      "learning_rate": 4.933824365673356e-07,
-      "loss": 0.2398,
+      "epoch": 4.808808159440661,
+      "grad_norm": 0.31472426652908325,
+      "learning_rate": 1.948073837670955e-07,
+      "loss": 0.3801,
       "step": 133430
     },
     {
-      "epoch": 4.69,
-      "learning_rate": 4.928194335610897e-07,
-      "loss": 0.2697,
+      "epoch": 4.808988359101885,
+      "grad_norm": 0.2125542312860489,
+      "learning_rate": 1.9444396839441436e-07,
+      "loss": 0.3243,
       "step": 133435
     },
     {
-      "epoch": 4.69,
-      "learning_rate": 4.922567487644331e-07,
-      "loss": 0.2417,
+      "epoch": 4.809168558763109,
+      "grad_norm": 0.27946358919143677,
+      "learning_rate": 1.9408089099263172e-07,
+      "loss": 0.3759,
       "step": 133440
     },
     {
-      "epoch": 4.69,
-      "learning_rate": 4.916943821846737e-07,
-      "loss": 0.2278,
+      "epoch": 4.809348758424334,
+      "grad_norm": 0.2657468914985657,
+      "learning_rate": 1.9371815156669358e-07,
+      "loss": 0.385,
       "step": 133445
     },
     {
-      "epoch": 4.7,
-      "learning_rate": 4.911323338291057e-07,
-      "loss": 0.2754,
+      "epoch": 4.809528958085559,
+      "grad_norm": 0.31190598011016846,
+      "learning_rate": 1.9335575012154327e-07,
+      "loss": 0.4146,
       "step": 133450
     },
     {
-      "epoch": 4.7,
-      "learning_rate": 4.905706037050345e-07,
-      "loss": 0.2477,
+      "epoch": 4.809709157746783,
+      "grad_norm": 0.2902090549468994,
+      "learning_rate": 1.9299368666211847e-07,
+      "loss": 0.3892,
       "step": 133455
     },
     {
-      "epoch": 4.7,
-      "learning_rate": 4.900091918197541e-07,
-      "loss": 0.2517,
+      "epoch": 4.809889357408008,
+      "grad_norm": 0.27345964312553406,
+      "learning_rate": 1.9263196119335413e-07,
+      "loss": 0.3975,
       "step": 133460
     },
     {
-      "epoch": 4.7,
-      "learning_rate": 4.894480981805505e-07,
-      "loss": 0.2734,
+      "epoch": 4.810069557069233,
+      "grad_norm": 0.2101958841085434,
+      "learning_rate": 1.9227057372017132e-07,
+      "loss": 0.3399,
       "step": 133465
     },
     {
-      "epoch": 4.7,
-      "learning_rate": 4.888873227947066e-07,
-      "loss": 0.2455,
+      "epoch": 4.810249756730458,
+      "grad_norm": 0.24289266765117645,
+      "learning_rate": 1.9190952424750496e-07,
+      "loss": 0.3661,
       "step": 133470
     },
     {
-      "epoch": 4.7,
-      "learning_rate": 4.883268656695111e-07,
-      "loss": 0.2346,
+      "epoch": 4.810429956391682,
+      "grad_norm": 0.27102965116500854,
+      "learning_rate": 1.9154881278026504e-07,
+      "loss": 0.3835,
       "step": 133475
     },
     {
-      "epoch": 4.7,
-      "learning_rate": 4.87766726812236e-07,
-      "loss": 0.2484,
+      "epoch": 4.810610156052906,
+      "grad_norm": 0.2354983687400818,
+      "learning_rate": 1.9118843932336982e-07,
+      "loss": 0.3508,
       "step": 133480
     },
     {
-      "epoch": 4.7,
-      "learning_rate": 4.872069062301588e-07,
-      "loss": 0.2498,
+      "epoch": 4.810790355714131,
+      "grad_norm": 0.2682526707649231,
+      "learning_rate": 1.908284038817293e-07,
+      "loss": 0.3447,
       "step": 133485
     },
     {
-      "epoch": 4.7,
-      "learning_rate": 4.866474039305402e-07,
-      "loss": 0.2736,
+      "epoch": 4.810970555375356,
+      "grad_norm": 0.2339244782924652,
+      "learning_rate": 1.9046870646024785e-07,
+      "loss": 0.373,
       "step": 133490
     },
     {
-      "epoch": 4.7,
-      "learning_rate": 4.860882199206523e-07,
-      "loss": 0.252,
+      "epoch": 4.81115075503658,
+      "grad_norm": 0.2050226777791977,
+      "learning_rate": 1.901093470638271e-07,
+      "loss": 0.368,
       "step": 133495
     },
     {
-      "epoch": 4.7,
-      "learning_rate": 4.85529354207756e-07,
-      "loss": 0.2395,
+      "epoch": 4.811330954697805,
+      "grad_norm": 0.21445462107658386,
+      "learning_rate": 1.8975032569736318e-07,
+      "loss": 0.3391,
       "step": 133500
     },
     {
-      "epoch": 4.7,
-      "eval_loss": 0.24847909808158875,
-      "eval_runtime": 10.5573,
-      "eval_samples_per_second": 9.472,
-      "eval_steps_per_second": 9.472,
+      "epoch": 4.811330954697805,
+      "eval_loss": 0.4288221001625061,
+      "eval_runtime": 3.5381,
+      "eval_samples_per_second": 28.264,
+      "eval_steps_per_second": 7.066,
       "step": 133500
     },
     {
-      "epoch": 4.7,
-      "learning_rate": 4.849708067991038e-07,
-      "loss": 0.2376,
+      "epoch": 4.81151115435903,
+      "grad_norm": 0.23659485578536987,
+      "learning_rate": 1.8939164236574658e-07,
+      "loss": 0.3578,
       "step": 133505
     },
     {
-      "epoch": 4.7,
-      "learning_rate": 4.844125777019481e-07,
-      "loss": 0.2373,
+      "epoch": 4.811691354020255,
+      "grad_norm": 0.2790174186229706,
+      "learning_rate": 1.8903329707386785e-07,
+      "loss": 0.3562,
       "step": 133510
     },
     {
-      "epoch": 4.7,
-      "learning_rate": 4.838546669235416e-07,
-      "loss": 0.2411,
+      "epoch": 4.811871553681479,
+      "grad_norm": 0.2516117990016937,
+      "learning_rate": 1.8867528982660643e-07,
+      "loss": 0.3731,
       "step": 133515
     },
     {
-      "epoch": 4.7,
-      "learning_rate": 4.832970744711257e-07,
-      "loss": 0.2565,
+      "epoch": 4.812051753342704,
+      "grad_norm": 0.2900271713733673,
+      "learning_rate": 1.8831762062883896e-07,
+      "loss": 0.3736,
       "step": 133520
     },
     {
-      "epoch": 4.7,
-      "learning_rate": 4.827398003519363e-07,
-      "loss": 0.2392,
+      "epoch": 4.812231953003929,
+      "grad_norm": 0.22807054221630096,
+      "learning_rate": 1.8796028948544209e-07,
+      "loss": 0.3544,
       "step": 133525
     },
     {
-      "epoch": 4.7,
-      "learning_rate": 4.821828445732147e-07,
-      "loss": 0.2737,
+      "epoch": 4.812412152665153,
+      "grad_norm": 0.26726317405700684,
+      "learning_rate": 1.8760329640128139e-07,
+      "loss": 0.3583,
       "step": 133530
     },
     {
-      "epoch": 4.7,
-      "learning_rate": 4.816262071421912e-07,
-      "loss": 0.2578,
+      "epoch": 4.812592352326377,
+      "grad_norm": 0.23728644847869873,
+      "learning_rate": 1.8724664138122238e-07,
+      "loss": 0.3509,
       "step": 133535
     },
     {
-      "epoch": 4.7,
-      "learning_rate": 4.810698880660935e-07,
-      "loss": 0.2449,
+      "epoch": 4.812772551987602,
+      "grad_norm": 0.27613380551338196,
+      "learning_rate": 1.8689032443012234e-07,
+      "loss": 0.3841,
       "step": 133540
     },
     {
-      "epoch": 4.7,
-      "learning_rate": 4.805138873521436e-07,
-      "loss": 0.2407,
+      "epoch": 4.812952751648827,
+      "grad_norm": 0.23071031272411346,
+      "learning_rate": 1.8653434555284123e-07,
+      "loss": 0.3659,
       "step": 133545
     },
     {
-      "epoch": 4.7,
-      "learning_rate": 4.799582050075635e-07,
-      "loss": 0.2536,
+      "epoch": 4.813132951310052,
+      "grad_norm": 0.2540687024593353,
+      "learning_rate": 1.8617870475422238e-07,
+      "loss": 0.374,
       "step": 133550
     },
     {
-      "epoch": 4.7,
-      "learning_rate": 4.794028410395695e-07,
-      "loss": 0.2539,
+      "epoch": 4.813313150971276,
+      "grad_norm": 0.2460535317659378,
+      "learning_rate": 1.8582340203911475e-07,
+      "loss": 0.3813,
       "step": 133555
     },
     {
-      "epoch": 4.7,
-      "learning_rate": 4.788477954553672e-07,
-      "loss": 0.2721,
+      "epoch": 4.813493350632501,
+      "grad_norm": 0.18833312392234802,
+      "learning_rate": 1.8546843741236163e-07,
+      "loss": 0.3373,
       "step": 133560
     },
     {
-      "epoch": 4.7,
-      "learning_rate": 4.782930682621644e-07,
-      "loss": 0.2498,
+      "epoch": 4.813673550293726,
+      "grad_norm": 0.2607569694519043,
+      "learning_rate": 1.851138108787953e-07,
+      "loss": 0.3953,
       "step": 133565
     },
     {
-      "epoch": 4.7,
-      "learning_rate": 4.777386594671668e-07,
-      "loss": 0.2588,
+      "epoch": 4.81385374995495,
+      "grad_norm": 0.2937115430831909,
+      "learning_rate": 1.84759522443248e-07,
+      "loss": 0.3231,
       "step": 133570
     },
     {
-      "epoch": 4.7,
-      "learning_rate": 4.771845690775739e-07,
-      "loss": 0.2659,
+      "epoch": 4.814033949616174,
+      "grad_norm": 0.27710872888565063,
+      "learning_rate": 1.844055721105492e-07,
+      "loss": 0.389,
       "step": 133575
     },
     {
-      "epoch": 4.7,
-      "learning_rate": 4.7663079710057455e-07,
-      "loss": 0.2421,
+      "epoch": 4.814214149277399,
+      "grad_norm": 0.19749537110328674,
+      "learning_rate": 1.8405195988552003e-07,
+      "loss": 0.3749,
       "step": 133580
     },
     {
-      "epoch": 4.7,
-      "learning_rate": 4.7607734354336296e-07,
-      "loss": 0.2476,
+      "epoch": 4.814394348938624,
+      "grad_norm": 0.20698542892932892,
+      "learning_rate": 1.8369868577297612e-07,
+      "loss": 0.3486,
       "step": 133585
     },
     {
-      "epoch": 4.7,
-      "learning_rate": 4.7552420841312786e-07,
-      "loss": 0.2486,
+      "epoch": 4.814574548599849,
+      "grad_norm": 0.22517277300357819,
+      "learning_rate": 1.8334574977773577e-07,
+      "loss": 0.3636,
       "step": 133590
     },
     {
-      "epoch": 4.7,
-      "learning_rate": 4.7497139171704954e-07,
-      "loss": 0.2392,
+      "epoch": 4.814754748261073,
+      "grad_norm": 0.2940255105495453,
+      "learning_rate": 1.8299315190460352e-07,
+      "loss": 0.388,
       "step": 133595
     },
     {
-      "epoch": 4.7,
-      "learning_rate": 4.7441889346230284e-07,
-      "loss": 0.2418,
+      "epoch": 4.814934947922298,
+      "grad_norm": 0.21540936827659607,
+      "learning_rate": 1.8264089215838386e-07,
+      "loss": 0.414,
       "step": 133600
     },
     {
-      "epoch": 4.7,
-      "learning_rate": 4.738667136560654e-07,
-      "loss": 0.2407,
+      "epoch": 4.815115147583523,
+      "grad_norm": 0.2051449418067932,
+      "learning_rate": 1.8228897054388128e-07,
+      "loss": 0.3345,
       "step": 133605
     },
     {
-      "epoch": 4.7,
-      "learning_rate": 4.7331485230550363e-07,
-      "loss": 0.2377,
+      "epoch": 4.8152953472447475,
+      "grad_norm": 0.2661552429199219,
+      "learning_rate": 1.8193738706588082e-07,
+      "loss": 0.3696,
       "step": 133610
     },
     {
-      "epoch": 4.7,
-      "learning_rate": 4.727633094177869e-07,
-      "loss": 0.2626,
+      "epoch": 4.815475546905972,
+      "grad_norm": 0.26067331433296204,
+      "learning_rate": 1.815861417291842e-07,
+      "loss": 0.3499,
       "step": 133615
     },
     {
-      "epoch": 4.7,
-      "learning_rate": 4.722120850000733e-07,
-      "loss": 0.2364,
+      "epoch": 4.815655746567196,
+      "grad_norm": 0.23874883353710175,
+      "learning_rate": 1.8123523453856816e-07,
+      "loss": 0.3581,
       "step": 133620
     },
     {
-      "epoch": 4.7,
-      "learning_rate": 4.71661179059521e-07,
-      "loss": 0.2398,
+      "epoch": 4.815835946228421,
+      "grad_norm": 0.264144629240036,
+      "learning_rate": 1.8088466549881778e-07,
+      "loss": 0.3731,
       "step": 133625
     },
     {
-      "epoch": 4.7,
-      "learning_rate": 4.711105916032854e-07,
-      "loss": 0.2548,
+      "epoch": 4.8160161458896455,
+      "grad_norm": 0.2632257342338562,
+      "learning_rate": 1.8053443461470698e-07,
+      "loss": 0.4175,
       "step": 133630
     },
     {
-      "epoch": 4.7,
-      "learning_rate": 4.70560322638508e-07,
-      "loss": 0.2547,
+      "epoch": 4.81619634555087,
+      "grad_norm": 0.2465040385723114,
+      "learning_rate": 1.8018454189101254e-07,
+      "loss": 0.3609,
       "step": 133635
     },
     {
-      "epoch": 4.7,
-      "learning_rate": 4.7001037217234424e-07,
-      "loss": 0.2344,
+      "epoch": 4.816376545212095,
+      "grad_norm": 0.2356046885251999,
+      "learning_rate": 1.7983498733249725e-07,
+      "loss": 0.3647,
       "step": 133640
     },
     {
-      "epoch": 4.7,
-      "learning_rate": 4.6946074021193e-07,
-      "loss": 0.2463,
+      "epoch": 4.81655674487332,
+      "grad_norm": 0.2914571166038513,
+      "learning_rate": 1.7948577094392405e-07,
+      "loss": 0.3861,
       "step": 133645
     },
     {
-      "epoch": 4.7,
-      "learning_rate": 4.6891142676440125e-07,
-      "loss": 0.2588,
+      "epoch": 4.8167369445345445,
+      "grad_norm": 0.2516320049762726,
+      "learning_rate": 1.7913689273005018e-07,
+      "loss": 0.389,
       "step": 133650
     },
     {
-      "epoch": 4.7,
-      "learning_rate": 4.6836243183689123e-07,
-      "loss": 0.2377,
+      "epoch": 4.816917144195769,
+      "grad_norm": 0.2756977379322052,
+      "learning_rate": 1.78788352695633e-07,
+      "loss": 0.397,
       "step": 133655
     },
     {
-      "epoch": 4.7,
-      "learning_rate": 4.6781375543652475e-07,
-      "loss": 0.2665,
+      "epoch": 4.817097343856993,
+      "grad_norm": 0.31001341342926025,
+      "learning_rate": 1.7844015084542143e-07,
+      "loss": 0.3965,
       "step": 133660
     },
     {
-      "epoch": 4.7,
-      "learning_rate": 4.672653975704322e-07,
-      "loss": 0.2636,
+      "epoch": 4.817277543518218,
+      "grad_norm": 0.31293249130249023,
+      "learning_rate": 1.780922871841534e-07,
+      "loss": 0.3876,
       "step": 133665
     },
     {
-      "epoch": 4.7,
-      "learning_rate": 4.667173582457329e-07,
-      "loss": 0.2751,
+      "epoch": 4.8174577431794425,
+      "grad_norm": 0.23081745207309723,
+      "learning_rate": 1.7774476171657229e-07,
+      "loss": 0.3338,
       "step": 133670
     },
     {
-      "epoch": 4.7,
-      "learning_rate": 4.661696374695379e-07,
-      "loss": 0.2446,
+      "epoch": 4.817637942840667,
+      "grad_norm": 0.229642853140831,
+      "learning_rate": 1.7739757444741323e-07,
+      "loss": 0.4005,
       "step": 133675
     },
     {
-      "epoch": 4.7,
-      "learning_rate": 4.656222352489609e-07,
-      "loss": 0.2516,
+      "epoch": 4.817818142501892,
+      "grad_norm": 0.29115715622901917,
+      "learning_rate": 1.7705072538140856e-07,
+      "loss": 0.3731,
       "step": 133680
     },
     {
-      "epoch": 4.7,
-      "learning_rate": 4.650751515911128e-07,
-      "loss": 0.2605,
+      "epoch": 4.817998342163117,
+      "grad_norm": 0.27003014087677,
+      "learning_rate": 1.7670421452328224e-07,
+      "loss": 0.3786,
       "step": 133685
     },
     {
-      "epoch": 4.7,
-      "learning_rate": 4.645283865030936e-07,
-      "loss": 0.2383,
+      "epoch": 4.8181785418243415,
+      "grad_norm": 0.2552355229854584,
+      "learning_rate": 1.7635804187775273e-07,
+      "loss": 0.3423,
       "step": 133690
     },
     {
-      "epoch": 4.7,
-      "learning_rate": 4.639819399920031e-07,
-      "loss": 0.2682,
+      "epoch": 4.818358741485566,
+      "grad_norm": 0.29862159490585327,
+      "learning_rate": 1.760122074495385e-07,
+      "loss": 0.373,
       "step": 133695
     },
     {
-      "epoch": 4.7,
-      "learning_rate": 4.6343581206493844e-07,
-      "loss": 0.251,
+      "epoch": 4.818538941146791,
+      "grad_norm": 0.2521141469478607,
+      "learning_rate": 1.7566671124335242e-07,
+      "loss": 0.3318,
       "step": 133700
     },
     {
-      "epoch": 4.7,
-      "learning_rate": 4.628900027289884e-07,
-      "loss": 0.2662,
+      "epoch": 4.818719140808016,
+      "grad_norm": 0.23437343537807465,
+      "learning_rate": 1.7532155326390464e-07,
+      "loss": 0.373,
       "step": 133705
     },
     {
-      "epoch": 4.7,
-      "learning_rate": 4.623445119912445e-07,
-      "loss": 0.2338,
+      "epoch": 4.81889934046924,
+      "grad_norm": 0.24322153627872467,
+      "learning_rate": 1.749767335158914e-07,
+      "loss": 0.3533,
       "step": 133710
     },
     {
-      "epoch": 4.7,
-      "learning_rate": 4.617993398587844e-07,
-      "loss": 0.2535,
+      "epoch": 4.819079540130464,
+      "grad_norm": 0.29220882058143616,
+      "learning_rate": 1.7463225200401167e-07,
+      "loss": 0.3859,
       "step": 133715
     },
     {
-      "epoch": 4.7,
-      "learning_rate": 4.6125448633868586e-07,
-      "loss": 0.274,
+      "epoch": 4.819259739791689,
+      "grad_norm": 0.19493746757507324,
+      "learning_rate": 1.7428810873296453e-07,
+      "loss": 0.3794,
       "step": 133720
     },
     {
-      "epoch": 4.7,
-      "learning_rate": 4.607099514380292e-07,
-      "loss": 0.253,
+      "epoch": 4.819439939452914,
+      "grad_norm": 0.23761507868766785,
+      "learning_rate": 1.739443037074351e-07,
+      "loss": 0.3677,
       "step": 133725
     },
     {
-      "epoch": 4.7,
-      "learning_rate": 4.601657351638811e-07,
-      "loss": 0.2562,
+      "epoch": 4.8196201391141384,
+      "grad_norm": 0.31488823890686035,
+      "learning_rate": 1.736008369321085e-07,
+      "loss": 0.3997,
       "step": 133730
     },
     {
-      "epoch": 4.71,
-      "learning_rate": 4.59621837523308e-07,
-      "loss": 0.2455,
+      "epoch": 4.819800338775363,
+      "grad_norm": 0.23459258675575256,
+      "learning_rate": 1.7325770841166156e-07,
+      "loss": 0.3629,
       "step": 133735
     },
     {
-      "epoch": 4.71,
-      "learning_rate": 4.5907825852337106e-07,
-      "loss": 0.2697,
+      "epoch": 4.819980538436588,
+      "grad_norm": 0.29168611764907837,
+      "learning_rate": 1.7291491815077388e-07,
+      "loss": 0.3874,
       "step": 133740
     },
     {
-      "epoch": 4.71,
-      "learning_rate": 4.585349981711312e-07,
-      "loss": 0.2471,
+      "epoch": 4.820160738097813,
+      "grad_norm": 0.28949815034866333,
+      "learning_rate": 1.7257246615411393e-07,
+      "loss": 0.3727,
       "step": 133745
     },
     {
-      "epoch": 4.71,
-      "learning_rate": 4.5799205647364116e-07,
-      "loss": 0.2394,
+      "epoch": 4.820340937759037,
+      "grad_norm": 0.23711396753787994,
+      "learning_rate": 1.7223035242634467e-07,
+      "loss": 0.3764,
       "step": 133750
     },
     {
-      "epoch": 4.71,
-      "learning_rate": 4.5744943343795364e-07,
-      "loss": 0.2628,
+      "epoch": 4.820521137420261,
+      "grad_norm": 0.26176103949546814,
+      "learning_rate": 1.7188857697213178e-07,
+      "loss": 0.3257,
       "step": 133755
     },
     {
-      "epoch": 4.71,
-      "learning_rate": 4.5690712907110465e-07,
-      "loss": 0.2439,
+      "epoch": 4.820701337081486,
+      "grad_norm": 0.3358304798603058,
+      "learning_rate": 1.715471397961327e-07,
+      "loss": 0.3996,
       "step": 133760
     },
     {
-      "epoch": 4.71,
-      "learning_rate": 4.563651433801469e-07,
-      "loss": 0.2465,
+      "epoch": 4.820881536742711,
+      "grad_norm": 0.32944080233573914,
+      "learning_rate": 1.7120604090299363e-07,
+      "loss": 0.3569,
       "step": 133765
     },
     {
-      "epoch": 4.71,
-      "learning_rate": 4.5582347637211367e-07,
-      "loss": 0.2499,
+      "epoch": 4.821061736403935,
+      "grad_norm": 0.25890570878982544,
+      "learning_rate": 1.708652802973637e-07,
+      "loss": 0.3859,
       "step": 133770
     },
     {
-      "epoch": 4.71,
-      "learning_rate": 4.55390372222117e-07,
-      "loss": 0.2433,
+      "epoch": 4.82124193606516,
+      "grad_norm": 0.23768959939479828,
+      "learning_rate": 1.7052485798389196e-07,
+      "loss": 0.3734,
       "step": 133775
     },
     {
-      "epoch": 4.71,
-      "learning_rate": 4.548492788610637e-07,
-      "loss": 0.2583,
+      "epoch": 4.821422135726385,
+      "grad_norm": 0.2727597951889038,
+      "learning_rate": 1.70184773967208e-07,
+      "loss": 0.3652,
       "step": 133780
     },
     {
-      "epoch": 4.71,
-      "learning_rate": 4.5430850420261904e-07,
-      "loss": 0.2323,
+      "epoch": 4.82160233538761,
+      "grad_norm": 0.31934478878974915,
+      "learning_rate": 1.698450282519526e-07,
+      "loss": 0.3821,
       "step": 133785
     },
     {
-      "epoch": 4.71,
-      "learning_rate": 4.537680482537998e-07,
-      "loss": 0.2464,
+      "epoch": 4.821782535048834,
+      "grad_norm": 0.2313583940267563,
+      "learning_rate": 1.695056208427498e-07,
+      "loss": 0.3477,
       "step": 133790
     },
     {
-      "epoch": 4.71,
-      "learning_rate": 4.5322791102162807e-07,
-      "loss": 0.2542,
+      "epoch": 4.821962734710059,
+      "grad_norm": 0.28459036350250244,
+      "learning_rate": 1.691665517442237e-07,
+      "loss": 0.3909,
       "step": 133795
     },
     {
-      "epoch": 4.71,
-      "learning_rate": 4.5268809251311496e-07,
-      "loss": 0.259,
+      "epoch": 4.822142934371284,
+      "grad_norm": 0.2991600036621094,
+      "learning_rate": 1.6882782096099836e-07,
+      "loss": 0.3633,
       "step": 133800
     },
     {
-      "epoch": 4.71,
-      "learning_rate": 4.5214859273527146e-07,
-      "loss": 0.2701,
+      "epoch": 4.822323134032508,
+      "grad_norm": 0.2798830270767212,
+      "learning_rate": 1.684894284976868e-07,
+      "loss": 0.3536,
       "step": 133805
     },
     {
-      "epoch": 4.71,
-      "learning_rate": 4.5160941169509476e-07,
-      "loss": 0.2358,
+      "epoch": 4.822503333693732,
+      "grad_norm": 0.2202059030532837,
+      "learning_rate": 1.681513743588964e-07,
+      "loss": 0.3627,
       "step": 133810
     },
     {
-      "epoch": 4.71,
-      "learning_rate": 4.510705493995987e-07,
-      "loss": 0.2536,
+      "epoch": 4.822683533354957,
+      "grad_norm": 0.2335967868566513,
+      "learning_rate": 1.6781365854924014e-07,
+      "loss": 0.3384,
       "step": 133815
     },
     {
-      "epoch": 4.71,
-      "learning_rate": 4.5053200585576937e-07,
-      "loss": 0.2644,
+      "epoch": 4.822863733016182,
+      "grad_norm": 0.21627481281757355,
+      "learning_rate": 1.674762810733116e-07,
+      "loss": 0.3793,
       "step": 133820
     },
     {
-      "epoch": 4.71,
-      "learning_rate": 4.49993781070604e-07,
-      "loss": 0.2591,
+      "epoch": 4.823043932677407,
+      "grad_norm": 0.2823790907859802,
+      "learning_rate": 1.671392419357126e-07,
+      "loss": 0.3695,
       "step": 133825
     },
     {
-      "epoch": 4.71,
-      "learning_rate": 4.494558750510913e-07,
-      "loss": 0.2571,
+      "epoch": 4.823224132338631,
+      "grad_norm": 0.2549593448638916,
+      "learning_rate": 1.66802541141034e-07,
+      "loss": 0.342,
       "step": 133830
     },
     {
-      "epoch": 4.71,
-      "learning_rate": 4.4891828780421194e-07,
-      "loss": 0.2387,
+      "epoch": 4.823404331999856,
+      "grad_norm": 0.21052326261997223,
+      "learning_rate": 1.6646617869386095e-07,
+      "loss": 0.3486,
       "step": 133835
     },
     {
-      "epoch": 4.71,
-      "learning_rate": 4.483810193369492e-07,
-      "loss": 0.2582,
+      "epoch": 4.823584531661081,
+      "grad_norm": 0.24714645743370056,
+      "learning_rate": 1.6613015459877868e-07,
+      "loss": 0.377,
       "step": 133840
     },
     {
-      "epoch": 4.71,
-      "learning_rate": 4.4784406965628077e-07,
-      "loss": 0.2423,
+      "epoch": 4.823764731322305,
+      "grad_norm": 0.24139024317264557,
+      "learning_rate": 1.6579446886036687e-07,
+      "loss": 0.3834,
       "step": 133845
     },
     {
-      "epoch": 4.71,
-      "learning_rate": 4.4730743876917614e-07,
-      "loss": 0.2634,
+      "epoch": 4.823944930983529,
+      "grad_norm": 0.29688748717308044,
+      "learning_rate": 1.6545912148319687e-07,
+      "loss": 0.4207,
       "step": 133850
     },
     {
-      "epoch": 4.71,
-      "learning_rate": 4.4677112668260467e-07,
-      "loss": 0.2468,
+      "epoch": 4.824125130644754,
+      "grad_norm": 0.26805707812309265,
+      "learning_rate": 1.6512411247183724e-07,
+      "loss": 0.3626,
       "step": 133855
     },
     {
-      "epoch": 4.71,
-      "learning_rate": 4.4623513340352753e-07,
-      "loss": 0.2499,
+      "epoch": 4.824305330305979,
+      "grad_norm": 0.2496339976787567,
+      "learning_rate": 1.6478944183085376e-07,
+      "loss": 0.3988,
       "step": 133860
     },
     {
-      "epoch": 4.71,
-      "learning_rate": 4.4569945893890576e-07,
-      "loss": 0.265,
+      "epoch": 4.824485529967204,
+      "grad_norm": 0.20922985672950745,
+      "learning_rate": 1.644551095648067e-07,
+      "loss": 0.3854,
       "step": 133865
     },
     {
-      "epoch": 4.71,
-      "learning_rate": 4.451641032956949e-07,
-      "loss": 0.2362,
+      "epoch": 4.824665729628428,
+      "grad_norm": 0.2915142774581909,
+      "learning_rate": 1.6412111567825074e-07,
+      "loss": 0.3817,
       "step": 133870
     },
     {
-      "epoch": 4.71,
-      "learning_rate": 4.44629066480845e-07,
-      "loss": 0.2777,
+      "epoch": 4.824845929289653,
+      "grad_norm": 0.25543341040611267,
+      "learning_rate": 1.6378746017573222e-07,
+      "loss": 0.3862,
       "step": 133875
     },
     {
-      "epoch": 4.71,
-      "learning_rate": 4.4409434850130607e-07,
-      "loss": 0.2626,
+      "epoch": 4.825026128950878,
+      "grad_norm": 0.2814326882362366,
+      "learning_rate": 1.6345414306180584e-07,
+      "loss": 0.3492,
       "step": 133880
     },
     {
-      "epoch": 4.71,
-      "learning_rate": 4.435599493640169e-07,
-      "loss": 0.2508,
+      "epoch": 4.8252063286121025,
+      "grad_norm": 0.2885473370552063,
+      "learning_rate": 1.6312116434100412e-07,
+      "loss": 0.3942,
       "step": 133885
     },
     {
-      "epoch": 4.71,
-      "learning_rate": 4.4302586907591926e-07,
-      "loss": 0.2379,
+      "epoch": 4.825386528273327,
+      "grad_norm": 0.239909365773201,
+      "learning_rate": 1.6278852401787336e-07,
+      "loss": 0.4162,
       "step": 133890
     },
     {
-      "epoch": 4.71,
-      "learning_rate": 4.4249210764394643e-07,
-      "loss": 0.2597,
+      "epoch": 4.825566727934551,
+      "grad_norm": 0.22957023978233337,
+      "learning_rate": 1.6245622209693778e-07,
+      "loss": 0.3542,
       "step": 133895
     },
     {
-      "epoch": 4.71,
-      "learning_rate": 4.419586650750318e-07,
-      "loss": 0.2523,
+      "epoch": 4.825746927595776,
+      "grad_norm": 0.22540193796157837,
+      "learning_rate": 1.621242585827243e-07,
+      "loss": 0.3753,
       "step": 133900
     },
     {
-      "epoch": 4.71,
-      "learning_rate": 4.414255413760976e-07,
-      "loss": 0.2325,
+      "epoch": 4.825927127257001,
+      "grad_norm": 0.22572508454322815,
+      "learning_rate": 1.617926334797626e-07,
+      "loss": 0.3706,
       "step": 133905
     },
     {
-      "epoch": 4.71,
-      "learning_rate": 4.408927365540688e-07,
-      "loss": 0.2461,
+      "epoch": 4.826107326918225,
+      "grad_norm": 0.2843494117259979,
+      "learning_rate": 1.6146134679256574e-07,
+      "loss": 0.399,
       "step": 133910
     },
     {
-      "epoch": 4.71,
-      "learning_rate": 4.4036025061586216e-07,
-      "loss": 0.254,
+      "epoch": 4.82628752657945,
+      "grad_norm": 0.2644733190536499,
+      "learning_rate": 1.6113039852565237e-07,
+      "loss": 0.3451,
       "step": 133915
     },
     {
-      "epoch": 4.71,
-      "learning_rate": 4.3982808356839434e-07,
-      "loss": 0.2598,
+      "epoch": 4.826467726240675,
+      "grad_norm": 0.25811052322387695,
+      "learning_rate": 1.6079978868352442e-07,
+      "loss": 0.388,
       "step": 133920
     },
     {
-      "epoch": 4.71,
-      "learning_rate": 4.3929623541856813e-07,
-      "loss": 0.244,
+      "epoch": 4.8266479259018995,
+      "grad_norm": 0.2605394721031189,
+      "learning_rate": 1.6046951727069214e-07,
+      "loss": 0.3679,
       "step": 133925
     },
     {
-      "epoch": 4.71,
-      "learning_rate": 4.3876470617330026e-07,
-      "loss": 0.2638,
+      "epoch": 4.826828125563124,
+      "grad_norm": 0.29219627380371094,
+      "learning_rate": 1.60139584291652e-07,
+      "loss": 0.3613,
       "step": 133930
     },
     {
-      "epoch": 4.71,
-      "learning_rate": 4.382334958394824e-07,
-      "loss": 0.25,
+      "epoch": 4.827008325224348,
+      "grad_norm": 0.24204973876476288,
+      "learning_rate": 1.598099897509031e-07,
+      "loss": 0.3677,
       "step": 133935
     },
     {
-      "epoch": 4.71,
-      "learning_rate": 4.377026044240173e-07,
-      "loss": 0.2512,
+      "epoch": 4.827188524885573,
+      "grad_norm": 0.2921901047229767,
+      "learning_rate": 1.5948073365293358e-07,
+      "loss": 0.3456,
       "step": 133940
     },
     {
-      "epoch": 4.71,
-      "learning_rate": 4.3717203193379686e-07,
-      "loss": 0.2599,
+      "epoch": 4.8273687245467976,
+      "grad_norm": 0.2787991166114807,
+      "learning_rate": 1.5915181600222872e-07,
+      "loss": 0.3672,
       "step": 133945
     },
     {
-      "epoch": 4.71,
-      "learning_rate": 4.366417783757071e-07,
-      "loss": 0.2406,
+      "epoch": 4.827548924208022,
+      "grad_norm": 0.2651299238204956,
+      "learning_rate": 1.5882323680327104e-07,
+      "loss": 0.3701,
       "step": 133950
     },
     {
-      "epoch": 4.71,
-      "learning_rate": 4.3611184375663973e-07,
-      "loss": 0.2614,
+      "epoch": 4.827729123869247,
+      "grad_norm": 0.27234312891960144,
+      "learning_rate": 1.5849499606053753e-07,
+      "loss": 0.3357,
       "step": 133955
     },
     {
-      "epoch": 4.71,
-      "learning_rate": 4.3558222808346707e-07,
-      "loss": 0.2675,
+      "epoch": 4.827909323530472,
+      "grad_norm": 0.2102580964565277,
+      "learning_rate": 1.5816709377849957e-07,
+      "loss": 0.3899,
       "step": 133960
     },
     {
-      "epoch": 4.71,
-      "learning_rate": 4.3505293136307523e-07,
-      "loss": 0.2617,
+      "epoch": 4.8280895231916965,
+      "grad_norm": 0.29363569617271423,
+      "learning_rate": 1.5783952996162864e-07,
+      "loss": 0.3687,
       "step": 133965
     },
     {
-      "epoch": 4.71,
-      "learning_rate": 4.34523953602331e-07,
-      "loss": 0.2576,
+      "epoch": 4.828269722852921,
+      "grad_norm": 0.22374111413955688,
+      "learning_rate": 1.5751230461438228e-07,
+      "loss": 0.3412,
       "step": 133970
     },
     {
-      "epoch": 4.71,
-      "learning_rate": 4.339952948081066e-07,
-      "loss": 0.2348,
+      "epoch": 4.828449922514146,
+      "grad_norm": 0.2635408341884613,
+      "learning_rate": 1.5718541774122076e-07,
+      "loss": 0.4143,
       "step": 133975
     },
     {
-      "epoch": 4.71,
-      "learning_rate": 4.3346695498725776e-07,
-      "loss": 0.2357,
+      "epoch": 4.828630122175371,
+      "grad_norm": 0.25199902057647705,
+      "learning_rate": 1.568588693465961e-07,
+      "loss": 0.375,
       "step": 133980
     },
     {
-      "epoch": 4.71,
-      "learning_rate": 4.3293893414665667e-07,
-      "loss": 0.2417,
+      "epoch": 4.828810321836595,
+      "grad_norm": 0.21928834915161133,
+      "learning_rate": 1.5653265943496587e-07,
+      "loss": 0.3635,
       "step": 133985
     },
     {
-      "epoch": 4.71,
-      "learning_rate": 4.324112322931506e-07,
-      "loss": 0.2454,
+      "epoch": 4.828990521497819,
+      "grad_norm": 0.25685426592826843,
+      "learning_rate": 1.562067880107626e-07,
+      "loss": 0.3996,
       "step": 133990
     },
     {
-      "epoch": 4.71,
-      "learning_rate": 4.318838494335925e-07,
-      "loss": 0.2458,
+      "epoch": 4.829170721159044,
+      "grad_norm": 0.24886153638362885,
+      "learning_rate": 1.5588125507843275e-07,
+      "loss": 0.3769,
       "step": 133995
     },
     {
-      "epoch": 4.71,
-      "learning_rate": 4.313567855748324e-07,
-      "loss": 0.2546,
+      "epoch": 4.829350920820269,
+      "grad_norm": 0.2590484321117401,
+      "learning_rate": 1.5555606064241168e-07,
+      "loss": 0.3867,
       "step": 134000
     },
     {
-      "epoch": 4.71,
-      "eval_loss": 0.24852041900157928,
-      "eval_runtime": 10.5387,
-      "eval_samples_per_second": 9.489,
-      "eval_steps_per_second": 9.489,
+      "epoch": 4.829350920820269,
+      "eval_loss": 0.42881202697753906,
+      "eval_runtime": 3.5296,
+      "eval_samples_per_second": 28.332,
+      "eval_steps_per_second": 7.083,
       "step": 134000
     },
     {
-      "epoch": 4.71,
-      "learning_rate": 4.3083004072371203e-07,
-      "loss": 0.2659,
+      "epoch": 4.8295311204814935,
+      "grad_norm": 0.21452315151691437,
+      "learning_rate": 1.5523120470712915e-07,
+      "loss": 0.3729,
       "step": 134005
     },
     {
-      "epoch": 4.71,
-      "learning_rate": 4.303036148870759e-07,
-      "loss": 0.2601,
+      "epoch": 4.829711320142718,
+      "grad_norm": 0.2168400138616562,
+      "learning_rate": 1.5490668727701217e-07,
+      "loss": 0.3683,
       "step": 134010
     },
     {
-      "epoch": 4.72,
-      "learning_rate": 4.29777508071752e-07,
-      "loss": 0.2358,
+      "epoch": 4.829891519803943,
+      "grad_norm": 0.3310934603214264,
+      "learning_rate": 1.5458250835648225e-07,
+      "loss": 0.3743,
       "step": 134015
     },
     {
-      "epoch": 4.72,
-      "learning_rate": 4.2925172028457364e-07,
-      "loss": 0.2433,
+      "epoch": 4.830071719465168,
+      "grad_norm": 0.2809125781059265,
+      "learning_rate": 1.5425866794995247e-07,
+      "loss": 0.377,
       "step": 134020
     },
     {
-      "epoch": 4.72,
-      "learning_rate": 4.2872625153237144e-07,
-      "loss": 0.2395,
+      "epoch": 4.830251919126392,
+      "grad_norm": 0.2371680587530136,
+      "learning_rate": 1.5393516606183878e-07,
+      "loss": 0.3484,
       "step": 134025
     },
     {
-      "epoch": 4.72,
-      "learning_rate": 4.282011018219678e-07,
-      "loss": 0.2385,
+      "epoch": 4.830432118787616,
+      "grad_norm": 0.22989128530025482,
+      "learning_rate": 1.5361200269655153e-07,
+      "loss": 0.3637,
       "step": 134030
     },
     {
-      "epoch": 4.72,
-      "learning_rate": 4.2767627116017674e-07,
-      "loss": 0.2473,
+      "epoch": 4.830612318448841,
+      "grad_norm": 0.250777930021286,
+      "learning_rate": 1.5328917785848719e-07,
+      "loss": 0.3924,
       "step": 134035
     },
     {
-      "epoch": 4.72,
-      "learning_rate": 4.271517595538149e-07,
-      "loss": 0.2381,
+      "epoch": 4.830792518110066,
+      "grad_norm": 0.26133477687835693,
+      "learning_rate": 1.5296669155204778e-07,
+      "loss": 0.3637,
       "step": 134040
     },
     {
-      "epoch": 4.72,
-      "learning_rate": 4.266275670096936e-07,
-      "loss": 0.243,
+      "epoch": 4.8309727177712904,
+      "grad_norm": 0.28973469138145447,
+      "learning_rate": 1.526445437816243e-07,
+      "loss": 0.3738,
       "step": 134045
     },
     {
-      "epoch": 4.72,
-      "learning_rate": 4.261036935346213e-07,
-      "loss": 0.2607,
+      "epoch": 4.831152917432515,
+      "grad_norm": 0.2642209529876709,
+      "learning_rate": 1.5232273455161316e-07,
+      "loss": 0.3481,
       "step": 134050
     },
     {
-      "epoch": 4.72,
-      "learning_rate": 4.255801391353953e-07,
-      "loss": 0.2362,
+      "epoch": 4.83133311709374,
+      "grad_norm": 0.259695827960968,
+      "learning_rate": 1.5200126386639148e-07,
+      "loss": 0.3832,
       "step": 134055
     },
     {
-      "epoch": 4.72,
-      "learning_rate": 4.250569038188129e-07,
-      "loss": 0.256,
+      "epoch": 4.831513316754965,
+      "grad_norm": 0.2560080885887146,
+      "learning_rate": 1.516801317303418e-07,
+      "loss": 0.3417,
       "step": 134060
     },
     {
-      "epoch": 4.72,
-      "learning_rate": 4.2453398759167704e-07,
-      "loss": 0.2454,
+      "epoch": 4.831693516416189,
+      "grad_norm": 0.24287177622318268,
+      "learning_rate": 1.5135933814783844e-07,
+      "loss": 0.3666,
       "step": 134065
     },
     {
-      "epoch": 4.72,
-      "learning_rate": 4.240113904607684e-07,
-      "loss": 0.2605,
+      "epoch": 4.831873716077414,
+      "grad_norm": 0.2676023840904236,
+      "learning_rate": 1.5103888312325566e-07,
+      "loss": 0.3918,
       "step": 134070
     },
     {
-      "epoch": 4.72,
-      "learning_rate": 4.23489112432876e-07,
-      "loss": 0.2446,
+      "epoch": 4.832053915738639,
+      "grad_norm": 0.35573047399520874,
+      "learning_rate": 1.5071876666095385e-07,
+      "loss": 0.3883,
       "step": 134075
     },
     {
-      "epoch": 4.72,
-      "learning_rate": 4.2296715351478335e-07,
-      "loss": 0.2687,
+      "epoch": 4.832234115399863,
+      "grad_norm": 0.26907774806022644,
+      "learning_rate": 1.503989887653018e-07,
+      "loss": 0.3785,
       "step": 134080
     },
     {
-      "epoch": 4.72,
-      "learning_rate": 4.224455137132627e-07,
-      "loss": 0.2663,
+      "epoch": 4.832414315061087,
+      "grad_norm": 0.269654244184494,
+      "learning_rate": 1.500795494406515e-07,
+      "loss": 0.3896,
       "step": 134085
     },
     {
-      "epoch": 4.72,
-      "learning_rate": 4.21924193035092e-07,
-      "loss": 0.2411,
+      "epoch": 4.832594514722312,
+      "grad_norm": 0.3965383470058441,
+      "learning_rate": 1.497604486913551e-07,
+      "loss": 0.377,
       "step": 134090
     },
     {
-      "epoch": 4.72,
-      "learning_rate": 4.2140319148703535e-07,
-      "loss": 0.2379,
+      "epoch": 4.832774714383537,
+      "grad_norm": 0.22493110597133636,
+      "learning_rate": 1.4944168652176183e-07,
+      "loss": 0.3609,
       "step": 134095
     },
     {
-      "epoch": 4.72,
-      "learning_rate": 4.208825090758622e-07,
-      "loss": 0.2565,
+      "epoch": 4.832954914044762,
+      "grad_norm": 0.2781234681606293,
+      "learning_rate": 1.4912326293621549e-07,
+      "loss": 0.3761,
       "step": 134100
     },
     {
-      "epoch": 4.72,
-      "learning_rate": 4.203621458083312e-07,
-      "loss": 0.2397,
+      "epoch": 4.833135113705986,
+      "grad_norm": 0.3429318070411682,
+      "learning_rate": 1.4880517793905148e-07,
+      "loss": 0.3934,
       "step": 134105
     },
     {
-      "epoch": 4.72,
-      "learning_rate": 4.19842101691198e-07,
-      "loss": 0.2508,
+      "epoch": 4.833315313367211,
+      "grad_norm": 0.2164117842912674,
+      "learning_rate": 1.4848743153460522e-07,
+      "loss": 0.3973,
       "step": 134110
     },
     {
-      "epoch": 4.72,
-      "learning_rate": 4.1932237673121545e-07,
-      "loss": 0.2467,
+      "epoch": 4.833495513028436,
+      "grad_norm": 0.32656198740005493,
+      "learning_rate": 1.4817002372720935e-07,
+      "loss": 0.3912,
       "step": 134115
     },
     {
-      "epoch": 4.72,
-      "learning_rate": 4.1880297093513654e-07,
-      "loss": 0.2712,
+      "epoch": 4.83367571268966,
+      "grad_norm": 0.25910449028015137,
+      "learning_rate": 1.4785295452118264e-07,
+      "loss": 0.3827,
       "step": 134120
     },
     {
-      "epoch": 4.72,
-      "learning_rate": 4.1828388430969755e-07,
-      "loss": 0.254,
+      "epoch": 4.833855912350884,
+      "grad_norm": 0.20214806497097015,
+      "learning_rate": 1.4753622392084943e-07,
+      "loss": 0.3517,
       "step": 134125
     },
     {
-      "epoch": 4.72,
-      "learning_rate": 4.177651168616459e-07,
-      "loss": 0.2601,
+      "epoch": 4.834036112012109,
+      "grad_norm": 0.27396053075790405,
+      "learning_rate": 1.472198319305229e-07,
+      "loss": 0.3526,
       "step": 134130
     },
     {
-      "epoch": 4.72,
-      "learning_rate": 4.172466685977094e-07,
-      "loss": 0.2542,
+      "epoch": 4.834216311673334,
+      "grad_norm": 0.20608721673488617,
+      "learning_rate": 1.4690377855451353e-07,
+      "loss": 0.3631,
       "step": 134135
     },
     {
-      "epoch": 4.72,
-      "learning_rate": 4.167285395246273e-07,
-      "loss": 0.2537,
+      "epoch": 4.834396511334559,
+      "grad_norm": 0.26418453454971313,
+      "learning_rate": 1.4658806379712897e-07,
+      "loss": 0.3475,
       "step": 134140
     },
     {
-      "epoch": 4.72,
-      "learning_rate": 4.1621072964912464e-07,
-      "loss": 0.2534,
+      "epoch": 4.834576710995783,
+      "grad_norm": 0.2818256616592407,
+      "learning_rate": 1.4627268766267133e-07,
+      "loss": 0.36,
       "step": 134145
     },
     {
-      "epoch": 4.72,
-      "learning_rate": 4.1569323897792387e-07,
-      "loss": 0.2512,
+      "epoch": 4.834756910657008,
+      "grad_norm": 0.30141425132751465,
+      "learning_rate": 1.4595765015543715e-07,
+      "loss": 0.373,
       "step": 134150
     },
     {
-      "epoch": 4.72,
-      "learning_rate": 4.151760675177446e-07,
-      "loss": 0.2476,
+      "epoch": 4.834937110318233,
+      "grad_norm": 0.309048056602478,
+      "learning_rate": 1.456429512797175e-07,
+      "loss": 0.4023,
       "step": 134155
     },
     {
-      "epoch": 4.72,
-      "learning_rate": 4.1465921527530095e-07,
-      "loss": 0.2304,
+      "epoch": 4.8351173099794575,
+      "grad_norm": 0.28096258640289307,
+      "learning_rate": 1.4532859103980056e-07,
+      "loss": 0.3722,
       "step": 134160
     },
     {
-      "epoch": 4.72,
-      "learning_rate": 4.141426822573041e-07,
-      "loss": 0.2545,
+      "epoch": 4.835297509640682,
+      "grad_norm": 0.30201297998428345,
+      "learning_rate": 1.4501456943996628e-07,
+      "loss": 0.3901,
       "step": 134165
     },
     {
-      "epoch": 4.72,
-      "learning_rate": 4.136264684704655e-07,
-      "loss": 0.2259,
+      "epoch": 4.835477709301906,
+      "grad_norm": 0.2812051773071289,
+      "learning_rate": 1.447008864845001e-07,
+      "loss": 0.4076,
       "step": 134170
     },
     {
-      "epoch": 4.72,
-      "learning_rate": 4.1311057392147966e-07,
-      "loss": 0.248,
+      "epoch": 4.835657908963131,
+      "grad_norm": 0.3078659176826477,
+      "learning_rate": 1.443875421776736e-07,
+      "loss": 0.3531,
       "step": 134175
     },
     {
-      "epoch": 4.72,
-      "learning_rate": 4.1259499861705243e-07,
-      "loss": 0.2331,
+      "epoch": 4.835838108624356,
+      "grad_norm": 0.2857990264892578,
+      "learning_rate": 1.4407453652375002e-07,
+      "loss": 0.389,
       "step": 134180
     },
     {
-      "epoch": 4.72,
-      "learning_rate": 4.120797425638756e-07,
-      "loss": 0.2626,
+      "epoch": 4.83601830828558,
+      "grad_norm": 0.23130671679973602,
+      "learning_rate": 1.43761869527001e-07,
+      "loss": 0.4073,
       "step": 134185
     },
     {
-      "epoch": 4.72,
-      "learning_rate": 4.115648057686411e-07,
-      "loss": 0.2415,
+      "epoch": 4.836198507946805,
+      "grad_norm": 0.2094799429178238,
+      "learning_rate": 1.434495411916842e-07,
+      "loss": 0.3676,
       "step": 134190
     },
     {
-      "epoch": 4.72,
-      "learning_rate": 4.1105018823802966e-07,
-      "loss": 0.247,
+      "epoch": 4.83637870760803,
+      "grad_norm": 0.26732537150382996,
+      "learning_rate": 1.4313755152205455e-07,
+      "loss": 0.3971,
       "step": 134195
     },
     {
-      "epoch": 4.72,
-      "learning_rate": 4.105358899787276e-07,
-      "loss": 0.2723,
+      "epoch": 4.8365589072692545,
+      "grad_norm": 0.2994996905326843,
+      "learning_rate": 1.4282590052236423e-07,
+      "loss": 0.3768,
       "step": 134200
     },
     {
-      "epoch": 4.72,
-      "learning_rate": 4.1002191099741296e-07,
-      "loss": 0.2401,
+      "epoch": 4.836739106930479,
+      "grad_norm": 0.22868464887142181,
+      "learning_rate": 1.4251458819685704e-07,
+      "loss": 0.3547,
       "step": 134205
     },
     {
-      "epoch": 4.72,
-      "learning_rate": 4.095082513007581e-07,
-      "loss": 0.2487,
+      "epoch": 4.836919306591704,
+      "grad_norm": 0.26898080110549927,
+      "learning_rate": 1.4220361454977682e-07,
+      "loss": 0.3795,
       "step": 134210
     },
     {
-      "epoch": 4.72,
-      "learning_rate": 4.0899491089543276e-07,
-      "loss": 0.228,
+      "epoch": 4.837099506252928,
+      "grad_norm": 0.2844333350658417,
+      "learning_rate": 1.4189297958536185e-07,
+      "loss": 0.4158,
       "step": 134215
     },
     {
-      "epoch": 4.72,
-      "learning_rate": 4.08481889788101e-07,
-      "loss": 0.238,
+      "epoch": 4.837279705914153,
+      "grad_norm": 0.22372843325138092,
+      "learning_rate": 1.415826833078393e-07,
+      "loss": 0.3845,
       "step": 134220
     },
     {
-      "epoch": 4.72,
-      "learning_rate": 4.0796918798542694e-07,
-      "loss": 0.2298,
+      "epoch": 4.837459905575377,
+      "grad_norm": 0.3177548050880432,
+      "learning_rate": 1.412727257214419e-07,
+      "loss": 0.4199,
       "step": 134225
     },
     {
-      "epoch": 4.72,
-      "learning_rate": 4.074568054940636e-07,
-      "loss": 0.2284,
+      "epoch": 4.837640105236602,
+      "grad_norm": 0.22313030064105988,
+      "learning_rate": 1.409631068303885e-07,
+      "loss": 0.3804,
       "step": 134230
     },
     {
-      "epoch": 4.72,
-      "learning_rate": 4.069447423206696e-07,
-      "loss": 0.2346,
+      "epoch": 4.837820304897827,
+      "grad_norm": 0.21491451561450958,
+      "learning_rate": 1.4065382663890347e-07,
+      "loss": 0.3386,
       "step": 134235
     },
     {
-      "epoch": 4.72,
-      "learning_rate": 4.064329984718923e-07,
-      "loss": 0.2512,
+      "epoch": 4.8380005045590515,
+      "grad_norm": 0.24090984463691711,
+      "learning_rate": 1.403448851511946e-07,
+      "loss": 0.3939,
       "step": 134240
     },
     {
-      "epoch": 4.72,
-      "learning_rate": 4.059215739543737e-07,
-      "loss": 0.239,
+      "epoch": 4.838180704220276,
+      "grad_norm": 0.3148958384990692,
+      "learning_rate": 1.4003628237147238e-07,
+      "loss": 0.3783,
       "step": 134245
     },
     {
-      "epoch": 4.72,
-      "learning_rate": 4.0541046877475285e-07,
-      "loss": 0.2488,
+      "epoch": 4.838360903881501,
+      "grad_norm": 0.32658737897872925,
+      "learning_rate": 1.3972801830394732e-07,
+      "loss": 0.403,
       "step": 134250
     },
     {
-      "epoch": 4.72,
-      "learning_rate": 4.0489968293966896e-07,
-      "loss": 0.2513,
+      "epoch": 4.838541103542726,
+      "grad_norm": 0.26316890120506287,
+      "learning_rate": 1.3942009295281056e-07,
+      "loss": 0.3716,
       "step": 134255
     },
     {
-      "epoch": 4.72,
-      "learning_rate": 4.0438921645575555e-07,
-      "loss": 0.2548,
+      "epoch": 4.83872130320395,
+      "grad_norm": 0.3213282823562622,
+      "learning_rate": 1.3911250632226146e-07,
+      "loss": 0.3641,
       "step": 134260
     },
     {
-      "epoch": 4.72,
-      "learning_rate": 4.038790693296407e-07,
-      "loss": 0.2413,
+      "epoch": 4.838901502865174,
+      "grad_norm": 0.2604977786540985,
+      "learning_rate": 1.3880525841649673e-07,
+      "loss": 0.361,
       "step": 134265
     },
     {
-      "epoch": 4.72,
-      "learning_rate": 4.033692415679413e-07,
-      "loss": 0.265,
+      "epoch": 4.839081702526399,
+      "grad_norm": 0.22943316400051117,
+      "learning_rate": 1.3849834923969073e-07,
+      "loss": 0.3516,
       "step": 134270
     },
     {
-      "epoch": 4.72,
-      "learning_rate": 4.028597331772854e-07,
-      "loss": 0.2508,
+      "epoch": 4.839261902187624,
+      "grad_norm": 0.29876038432121277,
+      "learning_rate": 1.3819177879603462e-07,
+      "loss": 0.3684,
       "step": 134275
     },
     {
-      "epoch": 4.72,
-      "learning_rate": 4.023505441642872e-07,
-      "loss": 0.225,
+      "epoch": 4.8394421018488485,
+      "grad_norm": 0.2451476752758026,
+      "learning_rate": 1.378855470897028e-07,
+      "loss": 0.3882,
       "step": 134280
     },
     {
-      "epoch": 4.72,
-      "learning_rate": 4.018416745355552e-07,
-      "loss": 0.2656,
+      "epoch": 4.839622301510073,
+      "grad_norm": 0.2849496603012085,
+      "learning_rate": 1.3757965412486696e-07,
+      "loss": 0.3678,
       "step": 134285
     },
     {
-      "epoch": 4.72,
-      "learning_rate": 4.013331242976981e-07,
-      "loss": 0.2575,
+      "epoch": 4.839802501171298,
+      "grad_norm": 0.19778481125831604,
+      "learning_rate": 1.3727409990569319e-07,
+      "loss": 0.3739,
       "step": 134290
     },
     {
-      "epoch": 4.72,
-      "learning_rate": 4.008248934573189e-07,
-      "loss": 0.2526,
+      "epoch": 4.839982700832523,
+      "grad_norm": 0.22249376773834229,
+      "learning_rate": 1.369688844363448e-07,
+      "loss": 0.357,
       "step": 134295
     },
     {
-      "epoch": 4.73,
-      "learning_rate": 4.0031698202101784e-07,
-      "loss": 0.2413,
+      "epoch": 4.840162900493747,
+      "grad_norm": 0.2733737528324127,
+      "learning_rate": 1.366640077209852e-07,
+      "loss": 0.3621,
       "step": 134300
     },
     {
-      "epoch": 4.73,
-      "learning_rate": 3.9980938999538975e-07,
-      "loss": 0.2438,
+      "epoch": 4.840343100154971,
+      "grad_norm": 0.21390418708324432,
+      "learning_rate": 1.363594697637638e-07,
+      "loss": 0.3495,
       "step": 134305
     },
     {
-      "epoch": 4.73,
-      "learning_rate": 3.9930211738702374e-07,
-      "loss": 0.2287,
+      "epoch": 4.840523299816196,
+      "grad_norm": 0.3072846233844757,
+      "learning_rate": 1.3605527056883004e-07,
+      "loss": 0.4137,
       "step": 134310
     },
     {
-      "epoch": 4.73,
-      "learning_rate": 3.987951642025062e-07,
-      "loss": 0.2725,
+      "epoch": 4.840703499477421,
+      "grad_norm": 0.25589028000831604,
+      "learning_rate": 1.3575141014032788e-07,
+      "loss": 0.4014,
       "step": 134315
     },
     {
-      "epoch": 4.73,
-      "learning_rate": 3.982885304484235e-07,
-      "loss": 0.2506,
+      "epoch": 4.8408836991386455,
+      "grad_norm": 0.2160835713148117,
+      "learning_rate": 1.354478884824012e-07,
+      "loss": 0.3557,
       "step": 134320
     },
     {
-      "epoch": 4.73,
-      "learning_rate": 3.9778221613134823e-07,
-      "loss": 0.2348,
+      "epoch": 4.84106389879987,
+      "grad_norm": 0.2661241888999939,
+      "learning_rate": 1.351447055991828e-07,
+      "loss": 0.3846,
       "step": 134325
     },
     {
-      "epoch": 4.73,
-      "learning_rate": 3.972762212578557e-07,
-      "loss": 0.2513,
+      "epoch": 4.841244098461095,
+      "grad_norm": 0.23646846413612366,
+      "learning_rate": 1.3484186149480272e-07,
+      "loss": 0.3725,
       "step": 134330
     },
     {
-      "epoch": 4.73,
-      "learning_rate": 3.967705458345211e-07,
-      "loss": 0.2566,
+      "epoch": 4.84142429812232,
+      "grad_norm": 0.24086101353168488,
+      "learning_rate": 1.3453935617339099e-07,
+      "loss": 0.3693,
       "step": 134335
     },
     {
-      "epoch": 4.73,
-      "learning_rate": 3.962651898679087e-07,
-      "loss": 0.2325,
+      "epoch": 4.841604497783544,
+      "grad_norm": 0.19936615228652954,
+      "learning_rate": 1.3423718963906374e-07,
+      "loss": 0.3585,
       "step": 134340
     },
     {
-      "epoch": 4.73,
-      "learning_rate": 3.957601533645744e-07,
-      "loss": 0.2445,
+      "epoch": 4.841784697444769,
+      "grad_norm": 0.2506243884563446,
+      "learning_rate": 1.339353618959427e-07,
+      "loss": 0.3775,
       "step": 134345
     },
     {
-      "epoch": 4.73,
-      "learning_rate": 3.9525543633108233e-07,
-      "loss": 0.2519,
+      "epoch": 4.841964897105994,
+      "grad_norm": 0.22023634612560272,
+      "learning_rate": 1.3363387294813568e-07,
+      "loss": 0.3357,
       "step": 134350
     },
     {
-      "epoch": 4.73,
-      "learning_rate": 3.9475103877398e-07,
-      "loss": 0.2446,
+      "epoch": 4.842145096767218,
+      "grad_norm": 0.27518585324287415,
+      "learning_rate": 1.3333272279975328e-07,
+      "loss": 0.3789,
       "step": 134355
     },
     {
-      "epoch": 4.73,
-      "learning_rate": 3.9424696069982334e-07,
-      "loss": 0.2499,
+      "epoch": 4.8423252964284424,
+      "grad_norm": 0.20245154201984406,
+      "learning_rate": 1.3303191145490057e-07,
+      "loss": 0.359,
       "step": 134360
     },
     {
-      "epoch": 4.73,
-      "learning_rate": 3.937432021151516e-07,
-      "loss": 0.255,
+      "epoch": 4.842505496089667,
+      "grad_norm": 0.29023030400276184,
+      "learning_rate": 1.3273143891767147e-07,
+      "loss": 0.3642,
       "step": 134365
     },
     {
-      "epoch": 4.73,
-      "learning_rate": 3.9323976302650946e-07,
-      "loss": 0.268,
+      "epoch": 4.842685695750892,
+      "grad_norm": 0.23865652084350586,
+      "learning_rate": 1.3243130519216274e-07,
+      "loss": 0.3794,
       "step": 134370
     },
     {
-      "epoch": 4.73,
-      "learning_rate": 3.9273664344043337e-07,
-      "loss": 0.2577,
+      "epoch": 4.842865895412117,
+      "grad_norm": 0.222527414560318,
+      "learning_rate": 1.3213151028246273e-07,
+      "loss": 0.3462,
       "step": 134375
     },
     {
-      "epoch": 4.73,
-      "learning_rate": 3.9223384336345424e-07,
-      "loss": 0.2665,
+      "epoch": 4.843046095073341,
+      "grad_norm": 0.23358391225337982,
+      "learning_rate": 1.3183205419265708e-07,
+      "loss": 0.3391,
       "step": 134380
     },
     {
-      "epoch": 4.73,
-      "learning_rate": 3.91731362802103e-07,
-      "loss": 0.26,
+      "epoch": 4.843226294734566,
+      "grad_norm": 0.25445687770843506,
+      "learning_rate": 1.315329369268231e-07,
+      "loss": 0.3423,
       "step": 134385
     },
     {
-      "epoch": 4.73,
-      "learning_rate": 3.9122920176289934e-07,
-      "loss": 0.2487,
+      "epoch": 4.843406494395791,
+      "grad_norm": 0.312282532453537,
+      "learning_rate": 1.3123415848903809e-07,
+      "loss": 0.3769,
       "step": 134390
     },
     {
-      "epoch": 4.73,
-      "learning_rate": 3.9072736025236866e-07,
-      "loss": 0.2559,
+      "epoch": 4.843586694057015,
+      "grad_norm": 0.25823071599006653,
+      "learning_rate": 1.3093571888337652e-07,
+      "loss": 0.3843,
       "step": 134395
     },
     {
-      "epoch": 4.73,
-      "learning_rate": 3.902258382770252e-07,
-      "loss": 0.2562,
+      "epoch": 4.843766893718239,
+      "grad_norm": 0.24276918172836304,
+      "learning_rate": 1.3063761811389906e-07,
+      "loss": 0.3791,
       "step": 134400
     },
     {
-      "epoch": 4.73,
-      "learning_rate": 3.897246358433804e-07,
-      "loss": 0.2477,
+      "epoch": 4.843947093379464,
+      "grad_norm": 0.25537583231925964,
+      "learning_rate": 1.3033985618466914e-07,
+      "loss": 0.4017,
       "step": 134405
     },
     {
-      "epoch": 4.73,
-      "learning_rate": 3.892237529579429e-07,
-      "loss": 0.2262,
+      "epoch": 4.844127293040689,
+      "grad_norm": 0.2588360011577606,
+      "learning_rate": 1.3004243309974184e-07,
+      "loss": 0.3756,
       "step": 134410
     },
     {
-      "epoch": 4.73,
-      "learning_rate": 3.8872318962721586e-07,
-      "loss": 0.2661,
+      "epoch": 4.844307492701914,
+      "grad_norm": 0.3001175820827484,
+      "learning_rate": 1.2974534886317225e-07,
+      "loss": 0.3726,
       "step": 134415
     },
     {
-      "epoch": 4.73,
-      "learning_rate": 3.882229458576997e-07,
-      "loss": 0.2361,
+      "epoch": 4.844487692363138,
+      "grad_norm": 0.3025680184364319,
+      "learning_rate": 1.294486034790099e-07,
+      "loss": 0.3936,
       "step": 134420
     },
     {
-      "epoch": 4.73,
-      "learning_rate": 3.877230216558864e-07,
-      "loss": 0.2421,
+      "epoch": 4.844667892024363,
+      "grad_norm": 0.27103391289711,
+      "learning_rate": 1.2915219695129321e-07,
+      "loss": 0.3672,
       "step": 134425
     },
     {
-      "epoch": 4.73,
-      "learning_rate": 3.872234170282707e-07,
-      "loss": 0.2562,
+      "epoch": 4.844848091685588,
+      "grad_norm": 0.27774909138679504,
+      "learning_rate": 1.2885612928406342e-07,
+      "loss": 0.3652,
       "step": 134430
     },
     {
-      "epoch": 4.73,
-      "learning_rate": 3.867241319813364e-07,
-      "loss": 0.2601,
+      "epoch": 4.845028291346813,
+      "grad_norm": 0.2541874945163727,
+      "learning_rate": 1.2856040048135342e-07,
+      "loss": 0.3296,
       "step": 134435
     },
     {
-      "epoch": 4.73,
-      "learning_rate": 3.8622516652156993e-07,
-      "loss": 0.2331,
+      "epoch": 4.845208491008037,
+      "grad_norm": 0.2172260284423828,
+      "learning_rate": 1.2826501054719054e-07,
+      "loss": 0.3267,
       "step": 134440
     },
     {
-      "epoch": 4.73,
-      "learning_rate": 3.857265206554495e-07,
-      "loss": 0.2541,
+      "epoch": 4.845388690669262,
+      "grad_norm": 0.177218958735466,
+      "learning_rate": 1.2796995948560487e-07,
+      "loss": 0.349,
       "step": 134445
     },
     {
-      "epoch": 4.73,
-      "learning_rate": 3.8522819438944757e-07,
-      "loss": 0.2329,
+      "epoch": 4.845568890330486,
+      "grad_norm": 0.26676860451698303,
+      "learning_rate": 1.2767524730061263e-07,
+      "loss": 0.4027,
       "step": 134450
     },
     {
-      "epoch": 4.73,
-      "learning_rate": 3.8473018773003967e-07,
-      "loss": 0.2487,
+      "epoch": 4.845749089991711,
+      "grad_norm": 0.27540668845176697,
+      "learning_rate": 1.2738087399622733e-07,
+      "loss": 0.3745,
       "step": 134455
     },
     {
-      "epoch": 4.73,
-      "learning_rate": 3.8423250068368165e-07,
-      "loss": 0.2591,
+      "epoch": 4.845929289652935,
+      "grad_norm": 0.25069594383239746,
+      "learning_rate": 1.2708683957646238e-07,
+      "loss": 0.3432,
       "step": 134460
     },
     {
-      "epoch": 4.73,
-      "learning_rate": 3.8373513325684897e-07,
-      "loss": 0.25,
+      "epoch": 4.84610948931416,
+      "grad_norm": 0.29599523544311523,
+      "learning_rate": 1.2679314404532572e-07,
+      "loss": 0.3566,
       "step": 134465
     },
     {
-      "epoch": 4.73,
-      "learning_rate": 3.8323808545598915e-07,
-      "loss": 0.2557,
+      "epoch": 4.846289688975385,
+      "grad_norm": 0.2683459520339966,
+      "learning_rate": 1.264997874068169e-07,
+      "loss": 0.3686,
       "step": 134470
     },
     {
-      "epoch": 4.73,
-      "learning_rate": 3.82741357287561e-07,
-      "loss": 0.2413,
+      "epoch": 4.8464698886366095,
+      "grad_norm": 0.250180184841156,
+      "learning_rate": 1.2620676966493272e-07,
+      "loss": 0.3944,
       "step": 134475
     },
     {
-      "epoch": 4.73,
-      "learning_rate": 3.822449487580121e-07,
-      "loss": 0.2602,
+      "epoch": 4.846650088297834,
+      "grad_norm": 0.27705875039100647,
+      "learning_rate": 1.2591409082366445e-07,
+      "loss": 0.3683,
       "step": 134480
     },
     {
-      "epoch": 4.73,
-      "learning_rate": 3.8174885987378725e-07,
-      "loss": 0.2503,
+      "epoch": 4.846830287959059,
+      "grad_norm": 0.267696350812912,
+      "learning_rate": 1.2562175088700057e-07,
+      "loss": 0.3831,
       "step": 134485
     },
     {
-      "epoch": 4.73,
-      "learning_rate": 3.812530906413342e-07,
-      "loss": 0.2533,
+      "epoch": 4.847010487620283,
+      "grad_norm": 0.26516005396842957,
+      "learning_rate": 1.2532974985892398e-07,
+      "loss": 0.3802,
       "step": 134490
     },
     {
-      "epoch": 4.73,
-      "learning_rate": 3.80757641067081e-07,
-      "loss": 0.24,
+      "epoch": 4.847190687281508,
+      "grad_norm": 0.22032172977924347,
+      "learning_rate": 1.2503808774341486e-07,
+      "loss": 0.346,
       "step": 134495
     },
     {
-      "epoch": 4.73,
-      "learning_rate": 3.8026251115746427e-07,
-      "loss": 0.2486,
+      "epoch": 4.847370886942732,
+      "grad_norm": 0.25521814823150635,
+      "learning_rate": 1.2474676454444778e-07,
+      "loss": 0.3357,
       "step": 134500
     },
     {
-      "epoch": 4.73,
-      "eval_loss": 0.24848829209804535,
-      "eval_runtime": 10.5443,
-      "eval_samples_per_second": 9.484,
-      "eval_steps_per_second": 9.484,
+      "epoch": 4.847370886942732,
+      "eval_loss": 0.4288318157196045,
+      "eval_runtime": 3.5358,
+      "eval_samples_per_second": 28.282,
+      "eval_steps_per_second": 7.07,
       "step": 134500
     },
     {
-      "epoch": 4.73,
-      "learning_rate": 3.7976770091891766e-07,
-      "loss": 0.2469,
+      "epoch": 4.847551086603957,
+      "grad_norm": 0.25918158888816833,
+      "learning_rate": 1.2445578026598903e-07,
+      "loss": 0.3733,
       "step": 134505
     },
     {
-      "epoch": 4.73,
-      "learning_rate": 3.7927321035786113e-07,
-      "loss": 0.2598,
+      "epoch": 4.847731286265182,
+      "grad_norm": 0.2234797328710556,
+      "learning_rate": 1.241651349120021e-07,
+      "loss": 0.3643,
       "step": 134510
     },
     {
-      "epoch": 4.73,
-      "learning_rate": 3.7877903948071445e-07,
-      "loss": 0.2534,
+      "epoch": 4.8479114859264065,
+      "grad_norm": 0.30298492312431335,
+      "learning_rate": 1.238748284864505e-07,
+      "loss": 0.361,
       "step": 134515
     },
     {
-      "epoch": 4.73,
-      "learning_rate": 3.782851882938976e-07,
-      "loss": 0.2452,
+      "epoch": 4.848091685587631,
+      "grad_norm": 0.2233441025018692,
+      "learning_rate": 1.2358486099328658e-07,
+      "loss": 0.3642,
       "step": 134520
     },
     {
-      "epoch": 4.73,
-      "learning_rate": 3.777916568038192e-07,
-      "loss": 0.2526,
+      "epoch": 4.848271885248856,
+      "grad_norm": 0.26897063851356506,
+      "learning_rate": 1.2329523243646556e-07,
+      "loss": 0.3709,
       "step": 134525
     },
     {
-      "epoch": 4.73,
-      "learning_rate": 3.772984450168937e-07,
-      "loss": 0.2298,
+      "epoch": 4.848452084910081,
+      "grad_norm": 0.2564341723918915,
+      "learning_rate": 1.2300594281992872e-07,
+      "loss": 0.3741,
       "step": 134530
     },
     {
-      "epoch": 4.73,
-      "learning_rate": 3.768055529395187e-07,
-      "loss": 0.2505,
+      "epoch": 4.8486322845713055,
+      "grad_norm": 0.1938348412513733,
+      "learning_rate": 1.227169921476201e-07,
+      "loss": 0.3406,
       "step": 134535
     },
     {
-      "epoch": 4.73,
-      "learning_rate": 3.763129805780946e-07,
-      "loss": 0.2471,
+      "epoch": 4.848812484232529,
+      "grad_norm": 0.25928547978401184,
+      "learning_rate": 1.224283804234755e-07,
+      "loss": 0.3673,
       "step": 134540
     },
     {
-      "epoch": 4.73,
-      "learning_rate": 3.758207279390191e-07,
-      "loss": 0.2605,
+      "epoch": 4.848992683893754,
+      "grad_norm": 0.3056635558605194,
+      "learning_rate": 1.221401076514306e-07,
+      "loss": 0.3775,
       "step": 134545
     },
     {
-      "epoch": 4.73,
-      "learning_rate": 3.7532879502868433e-07,
-      "loss": 0.2401,
+      "epoch": 4.849172883554979,
+      "grad_norm": 0.2699172794818878,
+      "learning_rate": 1.2185217383540725e-07,
+      "loss": 0.3643,
       "step": 134550
     },
     {
-      "epoch": 4.73,
-      "learning_rate": 3.7483718185347683e-07,
-      "loss": 0.2561,
+      "epoch": 4.8493530832162035,
+      "grad_norm": 0.2580404579639435,
+      "learning_rate": 1.2156457897933014e-07,
+      "loss": 0.3634,
       "step": 134555
     },
     {
-      "epoch": 4.73,
-      "learning_rate": 3.7434588841978034e-07,
-      "loss": 0.2412,
+      "epoch": 4.849533282877428,
+      "grad_norm": 0.2560892701148987,
+      "learning_rate": 1.2127732308712114e-07,
+      "loss": 0.3698,
       "step": 134560
     },
     {
-      "epoch": 4.73,
-      "learning_rate": 3.738549147339732e-07,
-      "loss": 0.2506,
+      "epoch": 4.849713482538653,
+      "grad_norm": 0.2861708998680115,
+      "learning_rate": 1.2099040616269374e-07,
+      "loss": 0.3928,
       "step": 134565
     },
     {
-      "epoch": 4.73,
-      "learning_rate": 3.733642608024307e-07,
-      "loss": 0.2368,
+      "epoch": 4.849893682199878,
+      "grad_norm": 0.21710778772830963,
+      "learning_rate": 1.207038282099532e-07,
+      "loss": 0.384,
       "step": 134570
     },
     {
-      "epoch": 4.73,
-      "learning_rate": 3.7287392663152286e-07,
-      "loss": 0.2515,
+      "epoch": 4.850073881861102,
+      "grad_norm": 0.2481117993593216,
+      "learning_rate": 1.2041758923280465e-07,
+      "loss": 0.3562,
       "step": 134575
     },
     {
-      "epoch": 4.73,
-      "learning_rate": 3.7238391222761957e-07,
-      "loss": 0.2606,
+      "epoch": 4.850254081522326,
+      "grad_norm": 0.24015110731124878,
+      "learning_rate": 1.2013168923515338e-07,
+      "loss": 0.3729,
       "step": 134580
     },
     {
-      "epoch": 4.74,
-      "learning_rate": 3.7189421759708245e-07,
-      "loss": 0.2508,
+      "epoch": 4.850434281183551,
+      "grad_norm": 0.24284091591835022,
+      "learning_rate": 1.1990321330488386e-07,
+      "loss": 0.3809,
       "step": 134585
     },
     {
-      "epoch": 4.74,
-      "learning_rate": 3.7140484274626466e-07,
-      "loss": 0.265,
+      "epoch": 4.850614480844776,
+      "grad_norm": 0.23821890354156494,
+      "learning_rate": 1.196179234801309e-07,
+      "loss": 0.3453,
       "step": 134590
     },
     {
-      "epoch": 4.74,
-      "learning_rate": 3.70915787681525e-07,
-      "loss": 0.263,
+      "epoch": 4.8507946805060005,
+      "grad_norm": 0.2183995395898819,
+      "learning_rate": 1.1933297264576927e-07,
+      "loss": 0.3285,
       "step": 134595
     },
     {
-      "epoch": 4.74,
-      "learning_rate": 3.7042705240921127e-07,
-      "loss": 0.2538,
+      "epoch": 4.850974880167225,
+      "grad_norm": 0.22939856350421906,
+      "learning_rate": 1.1904836080567638e-07,
+      "loss": 0.3785,
       "step": 134600
     },
     {
-      "epoch": 4.74,
-      "learning_rate": 3.699386369356711e-07,
-      "loss": 0.2524,
+      "epoch": 4.85115507982845,
+      "grad_norm": 0.27934911847114563,
+      "learning_rate": 1.187640879637325e-07,
+      "loss": 0.3638,
       "step": 134605
     },
     {
-      "epoch": 4.74,
-      "learning_rate": 3.6945054126724945e-07,
-      "loss": 0.2637,
+      "epoch": 4.851335279489675,
+      "grad_norm": 0.263106107711792,
+      "learning_rate": 1.184801541238123e-07,
+      "loss": 0.3625,
       "step": 134610
     },
     {
-      "epoch": 4.74,
-      "learning_rate": 3.689627654102773e-07,
-      "loss": 0.2613,
+      "epoch": 4.851515479150899,
+      "grad_norm": 0.31745436787605286,
+      "learning_rate": 1.181965592897849e-07,
+      "loss": 0.3788,
       "step": 134615
     },
     {
-      "epoch": 4.74,
-      "learning_rate": 3.6847530937109133e-07,
-      "loss": 0.2424,
+      "epoch": 4.851695678812124,
+      "grad_norm": 0.297881543636322,
+      "learning_rate": 1.1791330346550832e-07,
+      "loss": 0.3824,
       "step": 134620
     },
     {
-      "epoch": 4.74,
-      "learning_rate": 3.679881731560225e-07,
-      "loss": 0.2353,
+      "epoch": 4.851875878473349,
+      "grad_norm": 0.2612767219543457,
+      "learning_rate": 1.1763038665484616e-07,
+      "loss": 0.3582,
       "step": 134625
     },
     {
-      "epoch": 4.74,
-      "learning_rate": 3.675013567713936e-07,
-      "loss": 0.2516,
+      "epoch": 4.852056078134573,
+      "grad_norm": 0.27403807640075684,
+      "learning_rate": 1.1734780886165364e-07,
+      "loss": 0.4069,
       "step": 134630
     },
     {
-      "epoch": 4.74,
-      "learning_rate": 3.670148602235246e-07,
-      "loss": 0.252,
+      "epoch": 4.8522362777957975,
+      "grad_norm": 0.250141978263855,
+      "learning_rate": 1.1706557008978048e-07,
+      "loss": 0.3807,
       "step": 134635
     },
     {
-      "epoch": 4.74,
-      "learning_rate": 3.665286835187354e-07,
-      "loss": 0.2366,
+      "epoch": 4.852416477457022,
+      "grad_norm": 0.24126245081424713,
+      "learning_rate": 1.167836703430708e-07,
+      "loss": 0.3707,
       "step": 134640
     },
     {
-      "epoch": 4.74,
-      "learning_rate": 3.6604282666333754e-07,
-      "loss": 0.2468,
+      "epoch": 4.852596677118247,
+      "grad_norm": 0.2205542027950287,
+      "learning_rate": 1.1650210962536323e-07,
+      "loss": 0.3328,
       "step": 134645
     },
     {
-      "epoch": 4.74,
-      "learning_rate": 3.6555728966363725e-07,
-      "loss": 0.2369,
+      "epoch": 4.852776876779472,
+      "grad_norm": 0.23354937136173248,
+      "learning_rate": 1.1622088794049912e-07,
+      "loss": 0.3718,
       "step": 134650
     },
     {
-      "epoch": 4.74,
-      "learning_rate": 3.6507207252594046e-07,
-      "loss": 0.2279,
+      "epoch": 4.852957076440696,
+      "grad_norm": 0.25610846281051636,
+      "learning_rate": 1.1594000529230875e-07,
+      "loss": 0.3912,
       "step": 134655
     },
     {
-      "epoch": 4.74,
-      "learning_rate": 3.6458717525654773e-07,
-      "loss": 0.2566,
+      "epoch": 4.853137276101921,
+      "grad_norm": 0.3403148055076599,
+      "learning_rate": 1.1565946168461684e-07,
+      "loss": 0.3726,
       "step": 134660
     },
     {
-      "epoch": 4.74,
-      "learning_rate": 3.641025978617568e-07,
-      "loss": 0.2486,
+      "epoch": 4.853317475763146,
+      "grad_norm": 0.2914038598537445,
+      "learning_rate": 1.1537925712124809e-07,
+      "loss": 0.3737,
       "step": 134665
     },
     {
-      "epoch": 4.74,
-      "learning_rate": 3.6361834034785714e-07,
-      "loss": 0.2552,
+      "epoch": 4.85349767542437,
+      "grad_norm": 0.3024619519710541,
+      "learning_rate": 1.150993916060189e-07,
+      "loss": 0.3859,
       "step": 134670
     },
     {
-      "epoch": 4.74,
-      "learning_rate": 3.631344027211381e-07,
-      "loss": 0.2381,
+      "epoch": 4.8536778750855945,
+      "grad_norm": 0.2520431876182556,
+      "learning_rate": 1.1481986514274012e-07,
+      "loss": 0.3754,
       "step": 134675
     },
     {
-      "epoch": 4.74,
-      "learning_rate": 3.6265078498788353e-07,
-      "loss": 0.2367,
+      "epoch": 4.853858074746819,
+      "grad_norm": 0.23025912046432495,
+      "learning_rate": 1.1454067773522537e-07,
+      "loss": 0.3815,
       "step": 134680
     },
     {
-      "epoch": 4.74,
-      "learning_rate": 3.6216748715437176e-07,
-      "loss": 0.2521,
+      "epoch": 4.854038274408044,
+      "grad_norm": 0.2676393687725067,
+      "learning_rate": 1.1426182938727714e-07,
+      "loss": 0.369,
       "step": 134685
     },
     {
-      "epoch": 4.74,
-      "learning_rate": 3.616845092268756e-07,
-      "loss": 0.2516,
+      "epoch": 4.854218474069269,
+      "grad_norm": 0.2332090437412262,
+      "learning_rate": 1.1398332010269241e-07,
+      "loss": 0.3455,
       "step": 134690
     },
     {
-      "epoch": 4.74,
-      "learning_rate": 3.612018512116705e-07,
-      "loss": 0.2339,
+      "epoch": 4.854398673730493,
+      "grad_norm": 0.23634545505046844,
+      "learning_rate": 1.1370514988526538e-07,
+      "loss": 0.4167,
       "step": 134695
     },
     {
-      "epoch": 4.74,
-      "learning_rate": 3.6071951311502374e-07,
-      "loss": 0.2278,
+      "epoch": 4.854578873391718,
+      "grad_norm": 0.25552475452423096,
+      "learning_rate": 1.1342731873878742e-07,
+      "loss": 0.4016,
       "step": 134700
     },
     {
-      "epoch": 4.74,
-      "learning_rate": 3.602374949431941e-07,
-      "loss": 0.2603,
+      "epoch": 4.854759073052943,
+      "grad_norm": 0.21695470809936523,
+      "learning_rate": 1.1314982666704444e-07,
+      "loss": 0.3786,
       "step": 134705
     },
     {
-      "epoch": 4.74,
-      "learning_rate": 3.5975579670244065e-07,
-      "loss": 0.2373,
+      "epoch": 4.854939272714168,
+      "grad_norm": 0.27159446477890015,
+      "learning_rate": 1.1287267367381671e-07,
+      "loss": 0.396,
       "step": 134710
     },
     {
-      "epoch": 4.74,
-      "learning_rate": 3.592744183990221e-07,
-      "loss": 0.2669,
+      "epoch": 4.855119472375392,
+      "grad_norm": 0.2568185329437256,
+      "learning_rate": 1.1259585976288179e-07,
+      "loss": 0.3537,
       "step": 134715
     },
     {
-      "epoch": 4.74,
-      "learning_rate": 3.587933600391863e-07,
-      "loss": 0.2454,
+      "epoch": 4.855299672036617,
+      "grad_norm": 0.29152026772499084,
+      "learning_rate": 1.1231938493800887e-07,
+      "loss": 0.341,
       "step": 134720
     },
     {
-      "epoch": 4.74,
-      "learning_rate": 3.583126216291782e-07,
-      "loss": 0.2355,
+      "epoch": 4.855479871697841,
+      "grad_norm": 0.2555713951587677,
+      "learning_rate": 1.1204324920296716e-07,
+      "loss": 0.3395,
       "step": 134725
     },
     {
-      "epoch": 4.74,
-      "learning_rate": 3.5783220317524293e-07,
-      "loss": 0.2398,
+      "epoch": 4.855660071359066,
+      "grad_norm": 0.2556825280189514,
+      "learning_rate": 1.1176745256151755e-07,
+      "loss": 0.3501,
       "step": 134730
     },
     {
-      "epoch": 4.74,
-      "learning_rate": 3.5735210468361426e-07,
-      "loss": 0.2453,
+      "epoch": 4.85584027102029,
+      "grad_norm": 0.30311036109924316,
+      "learning_rate": 1.1149199501741537e-07,
+      "loss": 0.3778,
       "step": 134735
     },
     {
-      "epoch": 4.74,
-      "learning_rate": 3.5687232616053166e-07,
-      "loss": 0.2474,
+      "epoch": 4.856020470681515,
+      "grad_norm": 0.23065444827079773,
+      "learning_rate": 1.1121687657441871e-07,
+      "loss": 0.373,
       "step": 134740
     },
     {
-      "epoch": 4.74,
-      "learning_rate": 3.563928676122208e-07,
-      "loss": 0.2625,
+      "epoch": 4.85620067034274,
+      "grad_norm": 0.26827138662338257,
+      "learning_rate": 1.1094209723627458e-07,
+      "loss": 0.3632,
       "step": 134745
     },
     {
-      "epoch": 4.74,
-      "learning_rate": 3.559137290449044e-07,
-      "loss": 0.2214,
+      "epoch": 4.856380870003965,
+      "grad_norm": 0.29760560393333435,
+      "learning_rate": 1.1066765700672166e-07,
+      "loss": 0.3913,
       "step": 134750
     },
     {
-      "epoch": 4.74,
-      "learning_rate": 3.5543491046481084e-07,
-      "loss": 0.2508,
+      "epoch": 4.856561069665189,
+      "grad_norm": 0.27333009243011475,
+      "learning_rate": 1.1039355588950695e-07,
+      "loss": 0.3704,
       "step": 134755
     },
     {
-      "epoch": 4.74,
-      "learning_rate": 3.549564118781518e-07,
-      "loss": 0.2269,
+      "epoch": 4.856741269326414,
+      "grad_norm": 0.244674414396286,
+      "learning_rate": 1.101197938883608e-07,
+      "loss": 0.3931,
       "step": 134760
     },
     {
-      "epoch": 4.74,
-      "learning_rate": 3.544782332911417e-07,
-      "loss": 0.2652,
+      "epoch": 4.856921468987638,
+      "grad_norm": 0.2720971405506134,
+      "learning_rate": 1.0984637100701078e-07,
+      "loss": 0.3967,
       "step": 134765
     },
     {
-      "epoch": 4.74,
-      "learning_rate": 3.540003747099896e-07,
-      "loss": 0.2392,
+      "epoch": 4.857101668648863,
+      "grad_norm": 0.17154434323310852,
+      "learning_rate": 1.0957328724918725e-07,
+      "loss": 0.3374,
       "step": 134770
     },
     {
-      "epoch": 4.74,
-      "learning_rate": 3.535228361409015e-07,
-      "loss": 0.2644,
+      "epoch": 4.857281868310087,
+      "grad_norm": 0.27293217182159424,
+      "learning_rate": 1.0930054261860668e-07,
+      "loss": 0.3667,
       "step": 134775
     },
     {
-      "epoch": 4.74,
-      "learning_rate": 3.5304561759007813e-07,
-      "loss": 0.2172,
+      "epoch": 4.857462067971312,
+      "grad_norm": 0.24722565710544586,
+      "learning_rate": 1.0902813711899107e-07,
+      "loss": 0.4101,
       "step": 134780
     },
     {
-      "epoch": 4.74,
-      "learning_rate": 3.525687190637117e-07,
-      "loss": 0.2304,
+      "epoch": 4.857642267632537,
+      "grad_norm": 0.3114294707775116,
+      "learning_rate": 1.0875607075404582e-07,
+      "loss": 0.3313,
       "step": 134785
     },
     {
-      "epoch": 4.74,
-      "learning_rate": 3.5209214056799997e-07,
-      "loss": 0.2548,
+      "epoch": 4.8578224672937615,
+      "grad_norm": 0.3135533332824707,
+      "learning_rate": 1.0848434352747905e-07,
+      "loss": 0.3961,
       "step": 134790
     },
     {
-      "epoch": 4.74,
-      "learning_rate": 3.5161588210912697e-07,
-      "loss": 0.2445,
+      "epoch": 4.858002666954986,
+      "grad_norm": 0.30858752131462097,
+      "learning_rate": 1.0821295544299337e-07,
+      "loss": 0.4051,
       "step": 134795
     },
     {
-      "epoch": 4.74,
-      "learning_rate": 3.511399436932766e-07,
-      "loss": 0.2448,
+      "epoch": 4.858182866616211,
+      "grad_norm": 0.25825393199920654,
+      "learning_rate": 1.0794190650429137e-07,
+      "loss": 0.3879,
       "step": 134800
     },
     {
-      "epoch": 4.74,
-      "learning_rate": 3.5066432532663005e-07,
-      "loss": 0.2474,
+      "epoch": 4.858363066277436,
+      "grad_norm": 0.31138578057289124,
+      "learning_rate": 1.0767119671505899e-07,
+      "loss": 0.3673,
       "step": 134805
     },
     {
-      "epoch": 4.74,
-      "learning_rate": 3.5018902701536296e-07,
-      "loss": 0.2618,
+      "epoch": 4.8585432659386605,
+      "grad_norm": 0.21300721168518066,
+      "learning_rate": 1.0740082607898494e-07,
+      "loss": 0.3528,
       "step": 134810
     },
     {
-      "epoch": 4.74,
-      "learning_rate": 3.4971404876564806e-07,
-      "loss": 0.2629,
+      "epoch": 4.858723465599884,
+      "grad_norm": 0.23539184033870697,
+      "learning_rate": 1.0713079459975795e-07,
+      "loss": 0.3364,
       "step": 134815
     },
     {
-      "epoch": 4.74,
-      "learning_rate": 3.4923939058364994e-07,
-      "loss": 0.2302,
+      "epoch": 4.858903665261109,
+      "grad_norm": 0.2637368440628052,
+      "learning_rate": 1.0686110228105284e-07,
+      "loss": 0.3895,
       "step": 134820
     },
     {
-      "epoch": 4.74,
-      "learning_rate": 3.4876505247553315e-07,
-      "loss": 0.237,
+      "epoch": 4.859083864922334,
+      "grad_norm": 0.24945858120918274,
+      "learning_rate": 1.0659174912655001e-07,
+      "loss": 0.3608,
       "step": 134825
     },
     {
-      "epoch": 4.74,
-      "learning_rate": 3.4829103444745383e-07,
-      "loss": 0.242,
+      "epoch": 4.8592640645835585,
+      "grad_norm": 0.2507854402065277,
+      "learning_rate": 1.0632273513991042e-07,
+      "loss": 0.3788,
       "step": 134830
     },
     {
-      "epoch": 4.74,
-      "learning_rate": 3.4781733650557366e-07,
-      "loss": 0.2565,
+      "epoch": 4.859444264244783,
+      "grad_norm": 0.2690359950065613,
+      "learning_rate": 1.0605406032480614e-07,
+      "loss": 0.3606,
       "step": 134835
     },
     {
-      "epoch": 4.74,
-      "learning_rate": 3.4734395865603787e-07,
-      "loss": 0.2636,
+      "epoch": 4.859624463906008,
+      "grad_norm": 0.2506677210330963,
+      "learning_rate": 1.0578572468489534e-07,
+      "loss": 0.4075,
       "step": 134840
     },
     {
-      "epoch": 4.74,
-      "learning_rate": 3.468709009049914e-07,
-      "loss": 0.2631,
+      "epoch": 4.859804663567233,
+      "grad_norm": 0.2608148455619812,
+      "learning_rate": 1.0551772822383343e-07,
+      "loss": 0.336,
       "step": 134845
     },
     {
-      "epoch": 4.74,
-      "learning_rate": 3.463981632585822e-07,
-      "loss": 0.2821,
+      "epoch": 4.8599848632284575,
+      "grad_norm": 0.26057958602905273,
+      "learning_rate": 1.0525007094527305e-07,
+      "loss": 0.3707,
       "step": 134850
     },
     {
-      "epoch": 4.74,
-      "learning_rate": 3.459257457229442e-07,
-      "loss": 0.2558,
+      "epoch": 4.860165062889681,
+      "grad_norm": 0.3034597337245941,
+      "learning_rate": 1.049827528528613e-07,
+      "loss": 0.3569,
       "step": 134855
     },
     {
-      "epoch": 4.74,
-      "learning_rate": 3.454536483042142e-07,
-      "loss": 0.2601,
+      "epoch": 4.860345262550906,
+      "grad_norm": 0.25797319412231445,
+      "learning_rate": 1.047157739502369e-07,
+      "loss": 0.3628,
       "step": 134860
     },
     {
-      "epoch": 4.74,
-      "learning_rate": 3.449818710085179e-07,
-      "loss": 0.26,
+      "epoch": 4.860525462212131,
+      "grad_norm": 0.2869213819503784,
+      "learning_rate": 1.0444913424104418e-07,
+      "loss": 0.4191,
       "step": 134865
     },
     {
-      "epoch": 4.75,
-      "learning_rate": 3.4451041384198635e-07,
-      "loss": 0.2316,
+      "epoch": 4.8607056618733555,
+      "grad_norm": 0.2047075480222702,
+      "learning_rate": 1.041828337289108e-07,
+      "loss": 0.3554,
       "step": 134870
     },
     {
-      "epoch": 4.75,
-      "learning_rate": 3.440392768107342e-07,
-      "loss": 0.254,
+      "epoch": 4.86088586153458,
+      "grad_norm": 0.2842639088630676,
+      "learning_rate": 1.0391687241746717e-07,
+      "loss": 0.3925,
       "step": 134875
     },
     {
-      "epoch": 4.75,
-      "learning_rate": 3.4356845992088715e-07,
-      "loss": 0.2416,
+      "epoch": 4.861066061195805,
+      "grad_norm": 0.2997124493122101,
+      "learning_rate": 1.036512503103354e-07,
+      "loss": 0.3753,
       "step": 134880
     },
     {
-      "epoch": 4.75,
-      "learning_rate": 3.430979631785514e-07,
-      "loss": 0.2333,
+      "epoch": 4.86124626085703,
+      "grad_norm": 0.23771364986896515,
+      "learning_rate": 1.0338596741113204e-07,
+      "loss": 0.3682,
       "step": 134885
     },
     {
-      "epoch": 4.75,
-      "learning_rate": 3.4262778658984416e-07,
-      "loss": 0.2421,
+      "epoch": 4.8614264605182544,
+      "grad_norm": 0.2713213264942169,
+      "learning_rate": 1.0312102372347921e-07,
+      "loss": 0.3753,
       "step": 134890
     },
     {
-      "epoch": 4.75,
-      "learning_rate": 3.421579301608607e-07,
-      "loss": 0.2682,
+      "epoch": 4.861606660179479,
+      "grad_norm": 0.28782111406326294,
+      "learning_rate": 1.0285641925097956e-07,
+      "loss": 0.3416,
       "step": 134895
     },
     {
-      "epoch": 4.75,
-      "learning_rate": 3.4168839389770725e-07,
-      "loss": 0.2293,
+      "epoch": 4.861786859840704,
+      "grad_norm": 0.21895797550678253,
+      "learning_rate": 1.0259215399724132e-07,
+      "loss": 0.3613,
       "step": 134900
     },
     {
-      "epoch": 4.75,
-      "learning_rate": 3.4121917780648163e-07,
-      "loss": 0.2583,
+      "epoch": 4.861967059501928,
+      "grad_norm": 0.21650052070617676,
+      "learning_rate": 1.0232822796586716e-07,
+      "loss": 0.3367,
       "step": 134905
     },
     {
-      "epoch": 4.75,
-      "learning_rate": 3.407502818932734e-07,
-      "loss": 0.2395,
+      "epoch": 4.8621472591631525,
+      "grad_norm": 0.22131061553955078,
+      "learning_rate": 1.0206464116044868e-07,
+      "loss": 0.4108,
       "step": 134910
     },
     {
-      "epoch": 4.75,
-      "learning_rate": 3.402817061641722e-07,
-      "loss": 0.2444,
+      "epoch": 4.862327458824377,
+      "grad_norm": 0.2649054527282715,
+      "learning_rate": 1.0180139358458018e-07,
+      "loss": 0.3535,
       "step": 134915
     },
     {
-      "epoch": 4.75,
-      "learning_rate": 3.398134506252593e-07,
-      "loss": 0.2272,
+      "epoch": 4.862507658485602,
+      "grad_norm": 0.1987181305885315,
+      "learning_rate": 1.0153848524184494e-07,
+      "loss": 0.4098,
       "step": 134920
     },
     {
-      "epoch": 4.75,
-      "learning_rate": 3.393455152826186e-07,
-      "loss": 0.2682,
+      "epoch": 4.862687858146827,
+      "grad_norm": 0.22203120589256287,
+      "learning_rate": 1.0127591613582899e-07,
+      "loss": 0.369,
       "step": 134925
     },
     {
-      "epoch": 4.75,
-      "learning_rate": 3.388779001423259e-07,
-      "loss": 0.2648,
+      "epoch": 4.862868057808051,
+      "grad_norm": 0.28719407320022583,
+      "learning_rate": 1.010136862701072e-07,
+      "loss": 0.3782,
       "step": 134930
     },
     {
-      "epoch": 4.75,
-      "learning_rate": 3.3841060521044856e-07,
-      "loss": 0.2421,
+      "epoch": 4.863048257469276,
+      "grad_norm": 0.2688189148902893,
+      "learning_rate": 1.0075179564825455e-07,
+      "loss": 0.3805,
       "step": 134935
     },
     {
-      "epoch": 4.75,
-      "learning_rate": 3.3794363049305667e-07,
-      "loss": 0.2558,
+      "epoch": 4.863228457130501,
+      "grad_norm": 0.22982627153396606,
+      "learning_rate": 1.0049024427383758e-07,
+      "loss": 0.3613,
       "step": 134940
     },
     {
-      "epoch": 4.75,
-      "learning_rate": 3.3747697599621764e-07,
-      "loss": 0.2466,
+      "epoch": 4.863408656791725,
+      "grad_norm": 0.18782635033130646,
+      "learning_rate": 1.0022903215042012e-07,
+      "loss": 0.3668,
       "step": 134945
     },
     {
-      "epoch": 4.75,
-      "learning_rate": 3.3701064172598497e-07,
-      "loss": 0.2599,
+      "epoch": 4.8635888564529495,
+      "grad_norm": 0.26904863119125366,
+      "learning_rate": 9.996815928156044e-08,
+      "loss": 0.371,
       "step": 134950
     },
     {
-      "epoch": 4.75,
-      "learning_rate": 3.365446276884149e-07,
-      "loss": 0.2467,
+      "epoch": 4.863769056114174,
+      "grad_norm": 0.29253089427948,
+      "learning_rate": 9.970762567081404e-08,
+      "loss": 0.3492,
       "step": 134955
     },
     {
-      "epoch": 4.75,
-      "learning_rate": 3.3607893388955815e-07,
-      "loss": 0.2531,
+      "epoch": 4.863949255775399,
+      "grad_norm": 0.25840044021606445,
+      "learning_rate": 9.944743132173085e-08,
+      "loss": 0.3703,
       "step": 134960
     },
     {
-      "epoch": 4.75,
-      "learning_rate": 3.3561356033546266e-07,
-      "loss": 0.2548,
+      "epoch": 4.864129455436624,
+      "grad_norm": 0.27990660071372986,
+      "learning_rate": 9.918757623785802e-08,
+      "loss": 0.3802,
       "step": 134965
     },
     {
-      "epoch": 4.75,
-      "learning_rate": 3.351485070321736e-07,
-      "loss": 0.2526,
+      "epoch": 4.864309655097848,
+      "grad_norm": 0.24384404718875885,
+      "learning_rate": 9.892806042273162e-08,
+      "loss": 0.3459,
       "step": 134970
     },
     {
-      "epoch": 4.75,
-      "learning_rate": 3.34683773985725e-07,
-      "loss": 0.2468,
+      "epoch": 4.864489854759073,
+      "grad_norm": 0.189422607421875,
+      "learning_rate": 9.866888387988771e-08,
+      "loss": 0.3627,
       "step": 134975
     },
     {
-      "epoch": 4.75,
-      "learning_rate": 3.34219361202151e-07,
-      "loss": 0.232,
+      "epoch": 4.864670054420298,
+      "grad_norm": 0.2541491985321045,
+      "learning_rate": 9.841004661285957e-08,
+      "loss": 0.3354,
       "step": 134980
     },
     {
-      "epoch": 4.75,
-      "learning_rate": 3.3375526868748284e-07,
-      "loss": 0.2708,
+      "epoch": 4.864850254081523,
+      "grad_norm": 0.24088044464588165,
+      "learning_rate": 9.815154862517495e-08,
+      "loss": 0.3643,
       "step": 134985
     },
     {
-      "epoch": 4.75,
-      "learning_rate": 3.3329149644774904e-07,
-      "loss": 0.2592,
+      "epoch": 4.865030453742747,
+      "grad_norm": 0.19148464500904083,
+      "learning_rate": 9.789338992035324e-08,
+      "loss": 0.3547,
       "step": 134990
     },
     {
-      "epoch": 4.75,
-      "learning_rate": 3.3282804448896697e-07,
-      "loss": 0.2699,
+      "epoch": 4.865210653403972,
+      "grad_norm": 0.2473505735397339,
+      "learning_rate": 9.763557050191385e-08,
+      "loss": 0.3399,
       "step": 134995
     },
     {
-      "epoch": 4.75,
-      "learning_rate": 3.3236491281715685e-07,
-      "loss": 0.2617,
+      "epoch": 4.865390853065196,
+      "grad_norm": 0.2551048696041107,
+      "learning_rate": 9.737809037336787e-08,
+      "loss": 0.3505,
       "step": 135000
     },
     {
-      "epoch": 4.75,
-      "eval_loss": 0.24845683574676514,
-      "eval_runtime": 10.5389,
-      "eval_samples_per_second": 9.489,
-      "eval_steps_per_second": 9.489,
+      "epoch": 4.865390853065196,
+      "eval_loss": 0.42877399921417236,
+      "eval_runtime": 3.5351,
+      "eval_samples_per_second": 28.288,
+      "eval_steps_per_second": 7.072,
       "step": 135000
     },
     {
-      "epoch": 4.75,
-      "learning_rate": 3.3190210143833336e-07,
-      "loss": 0.235,
+      "epoch": 4.865571052726421,
+      "grad_norm": 0.24867664277553558,
+      "learning_rate": 9.71209495382236e-08,
+      "loss": 0.3733,
       "step": 135005
     },
     {
-      "epoch": 4.75,
-      "learning_rate": 3.3143961035850545e-07,
-      "loss": 0.2608,
+      "epoch": 4.865751252387645,
+      "grad_norm": 0.33623263239860535,
+      "learning_rate": 9.686414799998656e-08,
+      "loss": 0.4158,
       "step": 135010
     },
     {
-      "epoch": 4.75,
-      "learning_rate": 3.30977439583674e-07,
-      "loss": 0.2415,
+      "epoch": 4.86593145204887,
+      "grad_norm": 0.301376074552536,
+      "learning_rate": 9.660768576215396e-08,
+      "loss": 0.3439,
       "step": 135015
     },
     {
-      "epoch": 4.75,
-      "learning_rate": 3.3051558911984247e-07,
-      "loss": 0.2501,
+      "epoch": 4.866111651710095,
+      "grad_norm": 0.3229968845844269,
+      "learning_rate": 9.635156282822023e-08,
+      "loss": 0.3624,
       "step": 135020
     },
     {
-      "epoch": 4.75,
-      "learning_rate": 3.3005405897301165e-07,
-      "loss": 0.2308,
+      "epoch": 4.86629185137132,
+      "grad_norm": 0.2948048710823059,
+      "learning_rate": 9.609577920167422e-08,
+      "loss": 0.3791,
       "step": 135025
     },
     {
-      "epoch": 4.75,
-      "learning_rate": 3.295928491491684e-07,
-      "loss": 0.2542,
+      "epoch": 4.866472051032544,
+      "grad_norm": 0.26417121291160583,
+      "learning_rate": 9.584033488600208e-08,
+      "loss": 0.357,
       "step": 135030
     },
     {
-      "epoch": 4.75,
-      "learning_rate": 3.2913195965430245e-07,
-      "loss": 0.2402,
+      "epoch": 4.866652250693769,
+      "grad_norm": 0.234276682138443,
+      "learning_rate": 9.558522988468432e-08,
+      "loss": 0.3794,
       "step": 135035
     },
     {
-      "epoch": 4.75,
-      "learning_rate": 3.2867139049440334e-07,
-      "loss": 0.246,
+      "epoch": 4.866832450354993,
+      "grad_norm": 0.22871927917003632,
+      "learning_rate": 9.533046420119318e-08,
+      "loss": 0.3673,
       "step": 135040
     },
     {
-      "epoch": 4.75,
-      "learning_rate": 3.282111416754441e-07,
-      "loss": 0.2597,
+      "epoch": 4.867012650016218,
+      "grad_norm": 0.2720992863178253,
+      "learning_rate": 9.507603783900366e-08,
+      "loss": 0.3691,
       "step": 135045
     },
     {
-      "epoch": 4.75,
-      "learning_rate": 3.277512132034033e-07,
-      "loss": 0.2654,
+      "epoch": 4.867192849677442,
+      "grad_norm": 0.187312513589859,
+      "learning_rate": 9.482195080158517e-08,
+      "loss": 0.3576,
       "step": 135050
     },
     {
-      "epoch": 4.75,
-      "learning_rate": 3.2729160508425396e-07,
-      "loss": 0.2464,
+      "epoch": 4.867373049338667,
+      "grad_norm": 0.2163442075252533,
+      "learning_rate": 9.456820309239333e-08,
+      "loss": 0.329,
       "step": 135055
     },
     {
-      "epoch": 4.75,
-      "learning_rate": 3.268323173239635e-07,
-      "loss": 0.2554,
+      "epoch": 4.867553248999892,
+      "grad_norm": 0.2720244526863098,
+      "learning_rate": 9.431479471488647e-08,
+      "loss": 0.364,
       "step": 135060
     },
     {
-      "epoch": 4.75,
-      "learning_rate": 3.2637334992849656e-07,
-      "loss": 0.2478,
+      "epoch": 4.867733448661117,
+      "grad_norm": 0.23891782760620117,
+      "learning_rate": 9.406172567252015e-08,
+      "loss": 0.3779,
       "step": 135065
     },
     {
-      "epoch": 4.75,
-      "learning_rate": 3.259147029038123e-07,
-      "loss": 0.2364,
+      "epoch": 4.867913648322341,
+      "grad_norm": 0.21805892884731293,
+      "learning_rate": 9.380899596873882e-08,
+      "loss": 0.3617,
       "step": 135070
     },
     {
-      "epoch": 4.75,
-      "learning_rate": 3.2545637625586154e-07,
-      "loss": 0.2548,
+      "epoch": 4.868093847983566,
+      "grad_norm": 0.23532462120056152,
+      "learning_rate": 9.355660560699254e-08,
+      "loss": 0.3854,
       "step": 135075
     },
     {
-      "epoch": 4.75,
-      "learning_rate": 3.2499836999059786e-07,
-      "loss": 0.2601,
+      "epoch": 4.868274047644791,
+      "grad_norm": 0.3134779632091522,
+      "learning_rate": 9.330455459071185e-08,
+      "loss": 0.3714,
       "step": 135080
     },
     {
-      "epoch": 4.75,
-      "learning_rate": 3.24540684113972e-07,
-      "loss": 0.2602,
+      "epoch": 4.8684542473060155,
+      "grad_norm": 0.24097296595573425,
+      "learning_rate": 9.305284292333572e-08,
+      "loss": 0.3793,
       "step": 135085
     },
     {
-      "epoch": 4.75,
-      "learning_rate": 3.2408331863192363e-07,
-      "loss": 0.2577,
+      "epoch": 4.868634446967239,
+      "grad_norm": 0.24849741160869598,
+      "learning_rate": 9.280147060829192e-08,
+      "loss": 0.3469,
       "step": 135090
     },
     {
-      "epoch": 4.75,
-      "learning_rate": 3.2362627355038975e-07,
-      "loss": 0.241,
+      "epoch": 4.868814646628464,
+      "grad_norm": 0.23304787278175354,
+      "learning_rate": 9.25504376490055e-08,
+      "loss": 0.3691,
       "step": 135095
     },
     {
-      "epoch": 4.75,
-      "learning_rate": 3.231695488753045e-07,
-      "loss": 0.2389,
+      "epoch": 4.868994846289689,
+      "grad_norm": 0.2561473548412323,
+      "learning_rate": 9.229974404889875e-08,
+      "loss": 0.3874,
       "step": 135100
     },
     {
-      "epoch": 4.75,
-      "learning_rate": 3.2271314461260195e-07,
-      "loss": 0.2322,
+      "epoch": 4.8691750459509135,
+      "grad_norm": 0.23009338974952698,
+      "learning_rate": 9.204938981138277e-08,
+      "loss": 0.3618,
       "step": 135105
     },
     {
-      "epoch": 4.75,
-      "learning_rate": 3.2225706076820796e-07,
-      "loss": 0.2624,
+      "epoch": 4.869355245612138,
+      "grad_norm": 0.22427034378051758,
+      "learning_rate": 9.179937493987434e-08,
+      "loss": 0.4131,
       "step": 135110
     },
     {
-      "epoch": 4.75,
-      "learning_rate": 3.2180129734804e-07,
-      "loss": 0.2631,
+      "epoch": 4.869535445273363,
+      "grad_norm": 0.30147793889045715,
+      "learning_rate": 9.154969943777625e-08,
+      "loss": 0.4054,
       "step": 135115
     },
     {
-      "epoch": 4.75,
-      "learning_rate": 3.2134585435802113e-07,
-      "loss": 0.2308,
+      "epoch": 4.869715644934588,
+      "grad_norm": 0.2412918359041214,
+      "learning_rate": 9.130036330849134e-08,
+      "loss": 0.3642,
       "step": 135120
     },
     {
-      "epoch": 4.75,
-      "learning_rate": 3.208907318040605e-07,
-      "loss": 0.2534,
+      "epoch": 4.8698958445958125,
+      "grad_norm": 0.2309122383594513,
+      "learning_rate": 9.105136655541691e-08,
+      "loss": 0.3581,
       "step": 135125
     },
     {
-      "epoch": 4.75,
-      "learning_rate": 3.2043592969207005e-07,
-      "loss": 0.22,
+      "epoch": 4.870076044257036,
+      "grad_norm": 0.29338333010673523,
+      "learning_rate": 9.080270918194466e-08,
+      "loss": 0.3669,
       "step": 135130
     },
     {
-      "epoch": 4.75,
-      "learning_rate": 3.1998144802795616e-07,
-      "loss": 0.2446,
+      "epoch": 4.870256243918261,
+      "grad_norm": 0.2538418471813202,
+      "learning_rate": 9.05543911914608e-08,
+      "loss": 0.3738,
       "step": 135135
     },
     {
-      "epoch": 4.75,
-      "learning_rate": 3.195272868176169e-07,
-      "loss": 0.2606,
+      "epoch": 4.870436443579486,
+      "grad_norm": 0.2361038327217102,
+      "learning_rate": 9.03064125873515e-08,
+      "loss": 0.3925,
       "step": 135140
     },
     {
-      "epoch": 4.75,
-      "learning_rate": 3.190734460669531e-07,
-      "loss": 0.2732,
+      "epoch": 4.8706166432407105,
+      "grad_norm": 0.27490556240081787,
+      "learning_rate": 9.00587733729974e-08,
+      "loss": 0.3542,
       "step": 135145
     },
     {
-      "epoch": 4.75,
-      "learning_rate": 3.186199257818517e-07,
-      "loss": 0.2327,
+      "epoch": 4.870796842901935,
+      "grad_norm": 0.21929150819778442,
+      "learning_rate": 8.9811473551768e-08,
+      "loss": 0.35,
       "step": 135150
     },
     {
-      "epoch": 4.76,
-      "learning_rate": 3.1816672596820527e-07,
-      "loss": 0.2567,
+      "epoch": 4.87097704256316,
+      "grad_norm": 0.234662726521492,
+      "learning_rate": 8.956451312703562e-08,
+      "loss": 0.3558,
       "step": 135155
     },
     {
-      "epoch": 4.76,
-      "learning_rate": 3.1771384663190075e-07,
-      "loss": 0.2659,
+      "epoch": 4.871157242224385,
+      "grad_norm": 0.2809460759162903,
+      "learning_rate": 8.931789210216146e-08,
+      "loss": 0.4014,
       "step": 135160
     },
     {
-      "epoch": 4.76,
-      "learning_rate": 3.172612877788139e-07,
-      "loss": 0.2662,
+      "epoch": 4.8713374418856095,
+      "grad_norm": 0.2753678560256958,
+      "learning_rate": 8.907161048050949e-08,
+      "loss": 0.3782,
       "step": 135165
     },
     {
-      "epoch": 4.76,
-      "learning_rate": 3.168090494148207e-07,
-      "loss": 0.2372,
+      "epoch": 4.871517641546834,
+      "grad_norm": 0.3030547499656677,
+      "learning_rate": 8.882566826543259e-08,
+      "loss": 0.4204,
       "step": 135170
     },
     {
-      "epoch": 4.76,
-      "learning_rate": 3.163571315457997e-07,
-      "loss": 0.268,
+      "epoch": 4.871697841208059,
+      "grad_norm": 0.2812124788761139,
+      "learning_rate": 8.858006546028641e-08,
+      "loss": 0.3582,
       "step": 135175
     },
     {
-      "epoch": 4.76,
-      "learning_rate": 3.1590553417761294e-07,
-      "loss": 0.2421,
+      "epoch": 4.871878040869283,
+      "grad_norm": 0.24554339051246643,
+      "learning_rate": 8.833480206840994e-08,
+      "loss": 0.3686,
       "step": 135180
     },
     {
-      "epoch": 4.76,
-      "learning_rate": 3.154542573161251e-07,
-      "loss": 0.2636,
+      "epoch": 4.8720582405305075,
+      "grad_norm": 0.24936611950397491,
+      "learning_rate": 8.80898780931505e-08,
+      "loss": 0.3608,
       "step": 135185
     },
     {
-      "epoch": 4.76,
-      "learning_rate": 3.150033009671927e-07,
-      "loss": 0.2537,
+      "epoch": 4.872238440191732,
+      "grad_norm": 0.26903268694877625,
+      "learning_rate": 8.784529353784432e-08,
+      "loss": 0.3633,
       "step": 135190
     },
     {
-      "epoch": 4.76,
-      "learning_rate": 3.145526651366776e-07,
-      "loss": 0.2599,
+      "epoch": 4.872418639852957,
+      "grad_norm": 0.2582688629627228,
+      "learning_rate": 8.760104840582207e-08,
+      "loss": 0.3629,
       "step": 135195
     },
     {
-      "epoch": 4.76,
-      "learning_rate": 3.1410234983042796e-07,
-      "loss": 0.2666,
+      "epoch": 4.872598839514182,
+      "grad_norm": 0.24837349355220795,
+      "learning_rate": 8.735714270041162e-08,
+      "loss": 0.3739,
       "step": 135200
     },
     {
-      "epoch": 4.76,
-      "learning_rate": 3.136523550542891e-07,
-      "loss": 0.2494,
+      "epoch": 4.8727790391754064,
+      "grad_norm": 0.2944589853286743,
+      "learning_rate": 8.711357642493812e-08,
+      "loss": 0.3869,
       "step": 135205
     },
     {
-      "epoch": 4.76,
-      "learning_rate": 3.1320268081410365e-07,
-      "loss": 0.2346,
+      "epoch": 4.872959238836631,
+      "grad_norm": 0.2535194456577301,
+      "learning_rate": 8.687034958271833e-08,
+      "loss": 0.3885,
       "step": 135210
     },
     {
-      "epoch": 4.76,
-      "learning_rate": 3.127533271157168e-07,
-      "loss": 0.2193,
+      "epoch": 4.873139438497856,
+      "grad_norm": 0.2899385392665863,
+      "learning_rate": 8.66274621770663e-08,
+      "loss": 0.3827,
       "step": 135215
     },
     {
-      "epoch": 4.76,
-      "learning_rate": 3.123042939649573e-07,
-      "loss": 0.2747,
+      "epoch": 4.873319638159081,
+      "grad_norm": 0.23525474965572357,
+      "learning_rate": 8.638491421129325e-08,
+      "loss": 0.4044,
       "step": 135220
     },
     {
-      "epoch": 4.76,
-      "learning_rate": 3.1185558136765656e-07,
-      "loss": 0.2646,
+      "epoch": 4.8734998378203045,
+      "grad_norm": 0.2679741084575653,
+      "learning_rate": 8.614270568869931e-08,
+      "loss": 0.3682,
       "step": 135225
     },
     {
-      "epoch": 4.76,
-      "learning_rate": 3.114071893296378e-07,
-      "loss": 0.2353,
+      "epoch": 4.873680037481529,
+      "grad_norm": 0.2726483643054962,
+      "learning_rate": 8.590083661259019e-08,
+      "loss": 0.4052,
       "step": 135230
     },
     {
-      "epoch": 4.76,
-      "learning_rate": 3.1095911785672946e-07,
-      "loss": 0.2685,
+      "epoch": 4.873860237142754,
+      "grad_norm": 0.2760205864906311,
+      "learning_rate": 8.565930698625769e-08,
+      "loss": 0.3704,
       "step": 135235
     },
     {
-      "epoch": 4.76,
-      "learning_rate": 3.105113669547466e-07,
-      "loss": 0.2474,
+      "epoch": 4.874040436803979,
+      "grad_norm": 0.21959510445594788,
+      "learning_rate": 8.54181168129936e-08,
+      "loss": 0.3618,
       "step": 135240
     },
     {
-      "epoch": 4.76,
-      "learning_rate": 3.1006393662950106e-07,
-      "loss": 0.2549,
+      "epoch": 4.874220636465203,
+      "grad_norm": 0.2475598305463791,
+      "learning_rate": 8.517726609608424e-08,
+      "loss": 0.3569,
       "step": 135245
     },
     {
-      "epoch": 4.76,
-      "learning_rate": 3.096168268868049e-07,
-      "loss": 0.2491,
+      "epoch": 4.874400836126428,
+      "grad_norm": 0.3407549560070038,
+      "learning_rate": 8.493675483881303e-08,
+      "loss": 0.3741,
       "step": 135250
     },
     {
-      "epoch": 4.76,
-      "learning_rate": 3.091700377324647e-07,
-      "loss": 0.242,
+      "epoch": 4.874581035787653,
+      "grad_norm": 0.2598182260990143,
+      "learning_rate": 8.46965830444496e-08,
+      "loss": 0.373,
       "step": 135255
     },
     {
-      "epoch": 4.76,
-      "learning_rate": 3.0872356917227565e-07,
-      "loss": 0.2686,
+      "epoch": 4.874761235448878,
+      "grad_norm": 0.20785191655158997,
+      "learning_rate": 8.445675071627468e-08,
+      "loss": 0.3365,
       "step": 135260
     },
     {
-      "epoch": 4.76,
-      "learning_rate": 3.082774212120415e-07,
-      "loss": 0.2521,
+      "epoch": 4.874941435110102,
+      "grad_norm": 0.33232206106185913,
+      "learning_rate": 8.421725785755508e-08,
+      "loss": 0.3491,
       "step": 135265
     },
     {
-      "epoch": 4.76,
-      "learning_rate": 3.078315938575521e-07,
-      "loss": 0.2656,
+      "epoch": 4.875121634771327,
+      "grad_norm": 0.26161181926727295,
+      "learning_rate": 8.397810447154653e-08,
+      "loss": 0.3811,
       "step": 135270
     },
     {
-      "epoch": 4.76,
-      "learning_rate": 3.073860871145945e-07,
-      "loss": 0.2623,
+      "epoch": 4.875301834432551,
+      "grad_norm": 0.2408609241247177,
+      "learning_rate": 8.373929056151586e-08,
+      "loss": 0.3769,
       "step": 135275
     },
     {
-      "epoch": 4.76,
-      "learning_rate": 3.0694090098895846e-07,
-      "loss": 0.2488,
+      "epoch": 4.875482034093776,
+      "grad_norm": 0.23764941096305847,
+      "learning_rate": 8.350081613071326e-08,
+      "loss": 0.3892,
       "step": 135280
     },
     {
-      "epoch": 4.76,
-      "learning_rate": 3.064960354864199e-07,
-      "loss": 0.2451,
+      "epoch": 4.875662233755,
+      "grad_norm": 0.2595727741718292,
+      "learning_rate": 8.326268118238612e-08,
+      "loss": 0.3808,
       "step": 135285
     },
     {
-      "epoch": 4.76,
-      "learning_rate": 3.0605149061275763e-07,
-      "loss": 0.2686,
+      "epoch": 4.875842433416225,
+      "grad_norm": 0.25060775876045227,
+      "learning_rate": 8.302488571978185e-08,
+      "loss": 0.3857,
       "step": 135290
     },
     {
-      "epoch": 4.76,
-      "learning_rate": 3.056072663737447e-07,
-      "loss": 0.2509,
+      "epoch": 4.87602263307745,
+      "grad_norm": 0.26845234632492065,
+      "learning_rate": 8.278742974613951e-08,
+      "loss": 0.4005,
       "step": 135295
     },
     {
-      "epoch": 4.76,
-      "learning_rate": 3.0516336277514323e-07,
-      "loss": 0.2616,
+      "epoch": 4.876202832738675,
+      "grad_norm": 0.2484569400548935,
+      "learning_rate": 8.255031326469542e-08,
+      "loss": 0.3708,
       "step": 135300
     },
     {
-      "epoch": 4.76,
-      "learning_rate": 3.047197798227236e-07,
-      "loss": 0.257,
+      "epoch": 4.876383032399899,
+      "grad_norm": 0.22691169381141663,
+      "learning_rate": 8.231353627867755e-08,
+      "loss": 0.379,
       "step": 135305
     },
     {
-      "epoch": 4.76,
-      "learning_rate": 3.042765175222423e-07,
-      "loss": 0.2525,
+      "epoch": 4.876563232061124,
+      "grad_norm": 0.28611308336257935,
+      "learning_rate": 8.207709879131386e-08,
+      "loss": 0.3717,
       "step": 135310
     },
     {
-      "epoch": 4.76,
-      "learning_rate": 3.0383357587945307e-07,
-      "loss": 0.2538,
+      "epoch": 4.876743431722348,
+      "grad_norm": 0.21973417699337006,
+      "learning_rate": 8.184100080582679e-08,
+      "loss": 0.3596,
       "step": 135315
     },
     {
-      "epoch": 4.76,
-      "learning_rate": 3.0339095490011236e-07,
-      "loss": 0.2413,
+      "epoch": 4.876923631383573,
+      "grad_norm": 0.3248337209224701,
+      "learning_rate": 8.160524232543043e-08,
+      "loss": 0.3808,
       "step": 135320
     },
     {
-      "epoch": 4.76,
-      "learning_rate": 3.0294865458996293e-07,
-      "loss": 0.2461,
+      "epoch": 4.877103831044797,
+      "grad_norm": 0.29214221239089966,
+      "learning_rate": 8.136982335333887e-08,
+      "loss": 0.3813,
       "step": 135325
     },
     {
-      "epoch": 4.76,
-      "learning_rate": 3.0250667495475005e-07,
-      "loss": 0.2333,
+      "epoch": 4.877284030706022,
+      "grad_norm": 0.26933974027633667,
+      "learning_rate": 8.113474389275788e-08,
+      "loss": 0.3874,
       "step": 135330
     },
     {
-      "epoch": 4.76,
-      "learning_rate": 3.020650160002109e-07,
-      "loss": 0.2433,
+      "epoch": 4.877464230367247,
+      "grad_norm": 0.2623523473739624,
+      "learning_rate": 8.090000394689324e-08,
+      "loss": 0.4208,
       "step": 135335
     },
     {
-      "epoch": 4.76,
-      "learning_rate": 3.0162367773208244e-07,
-      "loss": 0.236,
+      "epoch": 4.877644430028472,
+      "grad_norm": 0.2307843416929245,
+      "learning_rate": 8.06656035189396e-08,
+      "loss": 0.3574,
       "step": 135340
     },
     {
-      "epoch": 4.76,
-      "learning_rate": 3.011826601560935e-07,
-      "loss": 0.262,
+      "epoch": 4.877824629689696,
+      "grad_norm": 0.26475924253463745,
+      "learning_rate": 8.04315426120944e-08,
+      "loss": 0.3531,
       "step": 135345
     },
     {
-      "epoch": 4.76,
-      "learning_rate": 3.007419632779701e-07,
-      "loss": 0.2551,
+      "epoch": 4.878004829350921,
+      "grad_norm": 0.2883531153202057,
+      "learning_rate": 8.0197821229544e-08,
+      "loss": 0.3882,
       "step": 135350
     },
     {
-      "epoch": 4.76,
-      "learning_rate": 3.0030158710343537e-07,
-      "loss": 0.2543,
+      "epoch": 4.878185029012146,
+      "grad_norm": 0.2579416036605835,
+      "learning_rate": 7.996443937447196e-08,
+      "loss": 0.3979,
       "step": 135355
     },
     {
-      "epoch": 4.76,
-      "learning_rate": 2.9986153163820696e-07,
-      "loss": 0.2461,
+      "epoch": 4.8783652286733705,
+      "grad_norm": 0.21452683210372925,
+      "learning_rate": 7.973139705006183e-08,
+      "loss": 0.3527,
       "step": 135360
     },
     {
-      "epoch": 4.76,
-      "learning_rate": 2.994217968879942e-07,
-      "loss": 0.2425,
+      "epoch": 4.878545428334594,
+      "grad_norm": 0.2238248735666275,
+      "learning_rate": 7.949869425948609e-08,
+      "loss": 0.3585,
       "step": 135365
     },
     {
-      "epoch": 4.76,
-      "learning_rate": 2.9898238285851756e-07,
-      "loss": 0.2627,
+      "epoch": 4.878725627995819,
+      "grad_norm": 0.2821867763996124,
+      "learning_rate": 7.926633100591719e-08,
+      "loss": 0.3877,
       "step": 135370
     },
     {
-      "epoch": 4.76,
-      "learning_rate": 2.985432895554724e-07,
-      "loss": 0.2568,
+      "epoch": 4.878905827657044,
+      "grad_norm": 0.2567353844642639,
+      "learning_rate": 7.903430729251926e-08,
+      "loss": 0.3633,
       "step": 135375
     },
     {
-      "epoch": 4.76,
-      "learning_rate": 2.9810451698456254e-07,
-      "loss": 0.2551,
+      "epoch": 4.879086027318269,
+      "grad_norm": 0.26188910007476807,
+      "learning_rate": 7.880262312245368e-08,
+      "loss": 0.3812,
       "step": 135380
     },
     {
-      "epoch": 4.76,
-      "learning_rate": 2.97666065151489e-07,
-      "loss": 0.2592,
+      "epoch": 4.879266226979493,
+      "grad_norm": 0.2313767969608307,
+      "learning_rate": 7.857127849887624e-08,
+      "loss": 0.3708,
       "step": 135385
     },
     {
-      "epoch": 4.76,
-      "learning_rate": 2.972279340619416e-07,
-      "loss": 0.2342,
+      "epoch": 4.879446426640718,
+      "grad_norm": 0.318751722574234,
+      "learning_rate": 7.834027342494277e-08,
+      "loss": 0.3856,
       "step": 135390
     },
     {
-      "epoch": 4.76,
-      "learning_rate": 2.967901237216103e-07,
-      "loss": 0.2552,
+      "epoch": 4.879626626301943,
+      "grad_norm": 0.2602209150791168,
+      "learning_rate": 7.810960790379517e-08,
+      "loss": 0.3871,
       "step": 135395
     },
     {
-      "epoch": 4.76,
-      "learning_rate": 2.9635263413617664e-07,
-      "loss": 0.2617,
+      "epoch": 4.8798068259631675,
+      "grad_norm": 0.2891538441181183,
+      "learning_rate": 7.787928193858096e-08,
+      "loss": 0.3592,
       "step": 135400
     },
     {
-      "epoch": 4.76,
-      "learning_rate": 2.9591546531132773e-07,
-      "loss": 0.2556,
+      "epoch": 4.879987025624391,
+      "grad_norm": 0.30516698956489563,
+      "learning_rate": 7.76492955324365e-08,
+      "loss": 0.3672,
       "step": 135405
     },
     {
-      "epoch": 4.76,
-      "learning_rate": 2.9547861725273405e-07,
-      "loss": 0.2481,
+      "epoch": 4.880167225285616,
+      "grad_norm": 0.2555546462535858,
+      "learning_rate": 7.741964868849539e-08,
+      "loss": 0.3664,
       "step": 135410
     },
     {
-      "epoch": 4.76,
-      "learning_rate": 2.9504208996606885e-07,
-      "loss": 0.281,
+      "epoch": 4.880347424946841,
+      "grad_norm": 0.25104543566703796,
+      "learning_rate": 7.719034140988569e-08,
+      "loss": 0.3555,
       "step": 135415
     },
     {
-      "epoch": 4.76,
-      "learning_rate": 2.946058834569998e-07,
-      "loss": 0.2579,
+      "epoch": 4.8805276246080656,
+      "grad_norm": 0.2774941027164459,
+      "learning_rate": 7.696137369973266e-08,
+      "loss": 0.3659,
       "step": 135420
     },
     {
-      "epoch": 4.76,
-      "learning_rate": 2.9416999773119456e-07,
-      "loss": 0.2726,
+      "epoch": 4.88070782426929,
+      "grad_norm": 0.31660085916519165,
+      "learning_rate": 7.673274556115328e-08,
+      "loss": 0.3489,
       "step": 135425
     },
     {
-      "epoch": 4.76,
-      "learning_rate": 2.937344327943098e-07,
-      "loss": 0.2413,
+      "epoch": 4.880888023930515,
+      "grad_norm": 0.32428672909736633,
+      "learning_rate": 7.650445699727005e-08,
+      "loss": 0.4014,
       "step": 135430
     },
     {
-      "epoch": 4.76,
-      "learning_rate": 2.9329918865199925e-07,
-      "loss": 0.2489,
+      "epoch": 4.88106822359174,
+      "grad_norm": 0.2850460112094879,
+      "learning_rate": 7.627650801118325e-08,
+      "loss": 0.3708,
       "step": 135435
     },
     {
-      "epoch": 4.77,
-      "learning_rate": 2.928642653099167e-07,
-      "loss": 0.245,
+      "epoch": 4.8812484232529645,
+      "grad_norm": 0.24217717349529266,
+      "learning_rate": 7.604889860600706e-08,
+      "loss": 0.3699,
       "step": 135440
     },
     {
-      "epoch": 4.77,
-      "learning_rate": 2.924296627737133e-07,
-      "loss": 0.2226,
+      "epoch": 4.881428622914189,
+      "grad_norm": 0.2553984224796295,
+      "learning_rate": 7.582162878483623e-08,
+      "loss": 0.4042,
       "step": 135445
     },
     {
-      "epoch": 4.77,
-      "learning_rate": 2.9199538104902335e-07,
-      "loss": 0.2536,
+      "epoch": 4.881608822575414,
+      "grad_norm": 0.23735560476779938,
+      "learning_rate": 7.559469855077383e-08,
+      "loss": 0.3997,
       "step": 135450
     },
     {
-      "epoch": 4.77,
-      "learning_rate": 2.9156142014149237e-07,
-      "loss": 0.2443,
+      "epoch": 4.881789022236639,
+      "grad_norm": 0.2347259223461151,
+      "learning_rate": 7.53681079069063e-08,
+      "loss": 0.3606,
       "step": 135455
     },
     {
-      "epoch": 4.77,
-      "learning_rate": 2.911277800567519e-07,
-      "loss": 0.2464,
+      "epoch": 4.8819692218978625,
+      "grad_norm": 0.24018234014511108,
+      "learning_rate": 7.51418568563228e-08,
+      "loss": 0.3564,
       "step": 135460
     },
     {
-      "epoch": 4.77,
-      "learning_rate": 2.906944608004336e-07,
-      "loss": 0.2466,
+      "epoch": 4.882149421559087,
+      "grad_norm": 0.26346853375434875,
+      "learning_rate": 7.491594540210423e-08,
+      "loss": 0.3791,
       "step": 135465
     },
     {
-      "epoch": 4.77,
-      "learning_rate": 2.9026146237816354e-07,
-      "loss": 0.2641,
+      "epoch": 4.882329621220312,
+      "grad_norm": 0.29124143719673157,
+      "learning_rate": 7.469037354733144e-08,
+      "loss": 0.3546,
       "step": 135470
     },
     {
-      "epoch": 4.77,
-      "learning_rate": 2.89828784795565e-07,
-      "loss": 0.2644,
+      "epoch": 4.882509820881537,
+      "grad_norm": 0.24946916103363037,
+      "learning_rate": 7.446514129507698e-08,
+      "loss": 0.4008,
       "step": 135475
     },
     {
-      "epoch": 4.77,
-      "learning_rate": 2.8939642805825564e-07,
-      "loss": 0.271,
+      "epoch": 4.8826900205427615,
+      "grad_norm": 0.23744343221187592,
+      "learning_rate": 7.424024864841061e-08,
+      "loss": 0.3892,
       "step": 135480
     },
     {
-      "epoch": 4.77,
-      "learning_rate": 2.889643921718449e-07,
-      "loss": 0.2541,
+      "epoch": 4.882870220203986,
+      "grad_norm": 0.25719282031059265,
+      "learning_rate": 7.401569561039379e-08,
+      "loss": 0.3679,
       "step": 135485
     },
     {
-      "epoch": 4.77,
-      "learning_rate": 2.885326771419505e-07,
-      "loss": 0.2691,
+      "epoch": 4.883050419865211,
+      "grad_norm": 0.27071502804756165,
+      "learning_rate": 7.379148218408516e-08,
+      "loss": 0.3677,
       "step": 135490
     },
     {
-      "epoch": 4.77,
-      "learning_rate": 2.8810128297417083e-07,
-      "loss": 0.2418,
+      "epoch": 4.883230619526436,
+      "grad_norm": 0.28900146484375,
+      "learning_rate": 7.356760837254617e-08,
+      "loss": 0.3915,
       "step": 135495
     },
     {
-      "epoch": 4.77,
-      "learning_rate": 2.876702096741124e-07,
-      "loss": 0.262,
+      "epoch": 4.8834108191876595,
+      "grad_norm": 0.2625620663166046,
+      "learning_rate": 7.334407417881883e-08,
+      "loss": 0.389,
       "step": 135500
     },
     {
-      "epoch": 4.77,
-      "eval_loss": 0.24846675992012024,
-      "eval_runtime": 10.5434,
-      "eval_samples_per_second": 9.485,
-      "eval_steps_per_second": 9.485,
+      "epoch": 4.8834108191876595,
+      "eval_loss": 0.4288046956062317,
+      "eval_runtime": 3.533,
+      "eval_samples_per_second": 28.305,
+      "eval_steps_per_second": 7.076,
       "step": 135500
     },
     {
-      "epoch": 4.77,
-      "learning_rate": 2.87239457247368e-07,
-      "loss": 0.2446,
+      "epoch": 4.883591018848884,
+      "grad_norm": 0.21090403199195862,
+      "learning_rate": 7.312087960595348e-08,
+      "loss": 0.3617,
       "step": 135505
     },
     {
-      "epoch": 4.77,
-      "learning_rate": 2.8680902569953595e-07,
-      "loss": 0.2705,
+      "epoch": 4.883771218510109,
+      "grad_norm": 0.300222784280777,
+      "learning_rate": 7.289802465698936e-08,
+      "loss": 0.3612,
       "step": 135510
     },
     {
-      "epoch": 4.77,
-      "learning_rate": 2.8637891503620064e-07,
-      "loss": 0.2406,
+      "epoch": 4.883951418171334,
+      "grad_norm": 0.2370145618915558,
+      "learning_rate": 7.267550933496569e-08,
+      "loss": 0.3661,
       "step": 135515
     },
     {
-      "epoch": 4.77,
-      "learning_rate": 2.8594912526294925e-07,
-      "loss": 0.2783,
+      "epoch": 4.8841316178325584,
+      "grad_norm": 0.3078131377696991,
+      "learning_rate": 7.245333364291063e-08,
+      "loss": 0.3926,
       "step": 135520
     },
     {
-      "epoch": 4.77,
-      "learning_rate": 2.855196563853607e-07,
-      "loss": 0.2388,
+      "epoch": 4.884311817493783,
+      "grad_norm": 0.30718788504600525,
+      "learning_rate": 7.223149758385506e-08,
+      "loss": 0.3866,
       "step": 135525
     },
     {
-      "epoch": 4.77,
-      "learning_rate": 2.850905084090111e-07,
-      "loss": 0.2583,
+      "epoch": 4.884492017155008,
+      "grad_norm": 0.251179575920105,
+      "learning_rate": 7.201000116081602e-08,
+      "loss": 0.3607,
       "step": 135530
     },
     {
-      "epoch": 4.77,
-      "learning_rate": 2.8466168133947367e-07,
-      "loss": 0.2553,
+      "epoch": 4.884672216816233,
+      "grad_norm": 0.21268831193447113,
+      "learning_rate": 7.17888443768161e-08,
+      "loss": 0.3675,
       "step": 135535
     },
     {
-      "epoch": 4.77,
-      "learning_rate": 2.842331751823135e-07,
-      "loss": 0.2385,
+      "epoch": 4.884852416477457,
+      "grad_norm": 0.22792141139507294,
+      "learning_rate": 7.156802723486678e-08,
+      "loss": 0.3899,
       "step": 135540
     },
     {
-      "epoch": 4.77,
-      "learning_rate": 2.8380498994310113e-07,
-      "loss": 0.2284,
+      "epoch": 4.885032616138682,
+      "grad_norm": 0.2687564790248871,
+      "learning_rate": 7.134754973797674e-08,
+      "loss": 0.3699,
       "step": 135545
     },
     {
-      "epoch": 4.77,
-      "learning_rate": 2.8337712562739316e-07,
-      "loss": 0.2567,
+      "epoch": 4.885212815799906,
+      "grad_norm": 0.2454073429107666,
+      "learning_rate": 7.112741188915195e-08,
+      "loss": 0.3885,
       "step": 135550
     },
     {
-      "epoch": 4.77,
-      "learning_rate": 2.8294958224074073e-07,
-      "loss": 0.2624,
+      "epoch": 4.885393015461131,
+      "grad_norm": 0.2492527812719345,
+      "learning_rate": 7.09076136913872e-08,
+      "loss": 0.3554,
       "step": 135555
     },
     {
-      "epoch": 4.77,
-      "learning_rate": 2.8252235978870054e-07,
-      "loss": 0.2405,
+      "epoch": 4.885573215122355,
+      "grad_norm": 0.24872152507305145,
+      "learning_rate": 7.068815514768013e-08,
+      "loss": 0.3674,
       "step": 135560
     },
     {
-      "epoch": 4.77,
-      "learning_rate": 2.820954582768209e-07,
-      "loss": 0.2596,
+      "epoch": 4.88575341478358,
+      "grad_norm": 0.3966946005821228,
+      "learning_rate": 7.046903626101997e-08,
+      "loss": 0.3429,
       "step": 135565
     },
     {
-      "epoch": 4.77,
-      "learning_rate": 2.816688777106391e-07,
-      "loss": 0.2357,
+      "epoch": 4.885933614444805,
+      "grad_norm": 0.2777992784976959,
+      "learning_rate": 7.025025703439325e-08,
+      "loss": 0.3934,
       "step": 135570
     },
     {
-      "epoch": 4.77,
-      "learning_rate": 2.8124261809569785e-07,
-      "loss": 0.2375,
+      "epoch": 4.88611381410603,
+      "grad_norm": 0.21319259703159332,
+      "learning_rate": 7.003181747078091e-08,
+      "loss": 0.373,
       "step": 135575
     },
     {
-      "epoch": 4.77,
-      "learning_rate": 2.8081667943752885e-07,
-      "loss": 0.2561,
+      "epoch": 4.886294013767254,
+      "grad_norm": 0.22784800827503204,
+      "learning_rate": 6.981371757315835e-08,
+      "loss": 0.3694,
       "step": 135580
     },
     {
-      "epoch": 4.77,
-      "learning_rate": 2.803910617416666e-07,
-      "loss": 0.2425,
+      "epoch": 4.886474213428479,
+      "grad_norm": 0.3045668601989746,
+      "learning_rate": 6.959595734449542e-08,
+      "loss": 0.4188,
       "step": 135585
     },
     {
-      "epoch": 4.77,
-      "learning_rate": 2.7996576501363445e-07,
-      "loss": 0.239,
+      "epoch": 4.886654413089703,
+      "grad_norm": 0.22848592698574066,
+      "learning_rate": 6.937853678776474e-08,
+      "loss": 0.3826,
       "step": 135590
     },
     {
-      "epoch": 4.77,
-      "learning_rate": 2.795407892589558e-07,
-      "loss": 0.2516,
+      "epoch": 4.886834612750928,
+      "grad_norm": 0.17041075229644775,
+      "learning_rate": 6.916145590592227e-08,
+      "loss": 0.3806,
       "step": 135595
     },
     {
-      "epoch": 4.77,
-      "learning_rate": 2.791161344831511e-07,
-      "loss": 0.2466,
+      "epoch": 4.887014812412152,
+      "grad_norm": 0.3228053152561188,
+      "learning_rate": 6.894471470192676e-08,
+      "loss": 0.3686,
       "step": 135600
     },
     {
-      "epoch": 4.77,
-      "learning_rate": 2.7869180069172997e-07,
-      "loss": 0.2325,
+      "epoch": 4.887195012073377,
+      "grad_norm": 0.26596736907958984,
+      "learning_rate": 6.872831317873418e-08,
+      "loss": 0.3763,
       "step": 135605
     },
     {
-      "epoch": 4.77,
-      "learning_rate": 2.782677878902018e-07,
-      "loss": 0.2447,
+      "epoch": 4.887375211734602,
+      "grad_norm": 0.2538110017776489,
+      "learning_rate": 6.851225133929217e-08,
+      "loss": 0.3434,
       "step": 135610
     },
     {
-      "epoch": 4.77,
-      "learning_rate": 2.778440960840761e-07,
-      "loss": 0.2456,
+      "epoch": 4.887555411395827,
+      "grad_norm": 0.23110675811767578,
+      "learning_rate": 6.829652918654284e-08,
+      "loss": 0.3542,
       "step": 135615
     },
     {
-      "epoch": 4.77,
-      "learning_rate": 2.774207252788541e-07,
-      "loss": 0.2659,
+      "epoch": 4.887735611057051,
+      "grad_norm": 0.24243219196796417,
+      "learning_rate": 6.808114672342825e-08,
+      "loss": 0.3663,
       "step": 135620
     },
     {
-      "epoch": 4.77,
-      "learning_rate": 2.769976754800313e-07,
-      "loss": 0.2339,
+      "epoch": 4.887915810718276,
+      "grad_norm": 0.27579575777053833,
+      "learning_rate": 6.78661039528794e-08,
+      "loss": 0.3798,
       "step": 135625
     },
     {
-      "epoch": 4.77,
-      "learning_rate": 2.7657494669309783e-07,
-      "loss": 0.2383,
+      "epoch": 4.888096010379501,
+      "grad_norm": 0.2889721393585205,
+      "learning_rate": 6.765140087782729e-08,
+      "loss": 0.3903,
       "step": 135630
     },
     {
-      "epoch": 4.77,
-      "learning_rate": 2.7615253892354643e-07,
-      "loss": 0.2633,
+      "epoch": 4.8882762100407255,
+      "grad_norm": 0.22426536679267883,
+      "learning_rate": 6.743703750120011e-08,
+      "loss": 0.3345,
       "step": 135635
     },
     {
-      "epoch": 4.77,
-      "learning_rate": 2.757304521768617e-07,
-      "loss": 0.2628,
+      "epoch": 4.888456409701949,
+      "grad_norm": 0.3247759938240051,
+      "learning_rate": 6.722301382591223e-08,
+      "loss": 0.4081,
       "step": 135640
     },
     {
-      "epoch": 4.77,
-      "learning_rate": 2.753086864585197e-07,
-      "loss": 0.2574,
+      "epoch": 4.888636609363174,
+      "grad_norm": 0.24855536222457886,
+      "learning_rate": 6.700932985488628e-08,
+      "loss": 0.4023,
       "step": 135645
     },
     {
-      "epoch": 4.77,
-      "learning_rate": 2.748872417740023e-07,
-      "loss": 0.2639,
+      "epoch": 4.888816809024399,
+      "grad_norm": 0.26974859833717346,
+      "learning_rate": 6.679598559103107e-08,
+      "loss": 0.3685,
       "step": 135650
     },
     {
-      "epoch": 4.77,
-      "learning_rate": 2.7446611812877997e-07,
-      "loss": 0.2581,
+      "epoch": 4.888997008685624,
+      "grad_norm": 0.2807910442352295,
+      "learning_rate": 6.658298103725258e-08,
+      "loss": 0.3518,
       "step": 135655
     },
     {
-      "epoch": 4.77,
-      "learning_rate": 2.740453155283207e-07,
-      "loss": 0.2548,
+      "epoch": 4.889177208346848,
+      "grad_norm": 0.271327406167984,
+      "learning_rate": 6.637031619645684e-08,
+      "loss": 0.3636,
       "step": 135660
     },
     {
-      "epoch": 4.77,
-      "learning_rate": 2.7362483397808667e-07,
-      "loss": 0.255,
+      "epoch": 4.889357408008073,
+      "grad_norm": 0.2790520489215851,
+      "learning_rate": 6.61579910715332e-08,
+      "loss": 0.3721,
       "step": 135665
     },
     {
-      "epoch": 4.77,
-      "learning_rate": 2.732046734835403e-07,
-      "loss": 0.2429,
+      "epoch": 4.889537607669298,
+      "grad_norm": 0.2329198122024536,
+      "learning_rate": 6.594600566538212e-08,
+      "loss": 0.356,
       "step": 135670
     },
     {
-      "epoch": 4.77,
-      "learning_rate": 2.7278483405013545e-07,
-      "loss": 0.2406,
+      "epoch": 4.8897178073305225,
+      "grad_norm": 0.23520605266094208,
+      "learning_rate": 6.573435998089018e-08,
+      "loss": 0.3887,
       "step": 135675
     },
     {
-      "epoch": 4.77,
-      "learning_rate": 2.7236531568332334e-07,
-      "loss": 0.2339,
+      "epoch": 4.889898006991746,
+      "grad_norm": 0.27827075123786926,
+      "learning_rate": 6.552305402093838e-08,
+      "loss": 0.3742,
       "step": 135680
     },
     {
-      "epoch": 4.77,
-      "learning_rate": 2.7194611838855243e-07,
-      "loss": 0.2244,
+      "epoch": 4.890078206652971,
+      "grad_norm": 0.2520049214363098,
+      "learning_rate": 6.531208778841059e-08,
+      "loss": 0.379,
       "step": 135685
     },
     {
-      "epoch": 4.77,
-      "learning_rate": 2.715272421712628e-07,
-      "loss": 0.2684,
+      "epoch": 4.890258406314196,
+      "grad_norm": 0.27177202701568604,
+      "learning_rate": 6.510146128617389e-08,
+      "loss": 0.3425,
       "step": 135690
     },
     {
-      "epoch": 4.77,
-      "learning_rate": 2.711086870368973e-07,
-      "loss": 0.2579,
+      "epoch": 4.890438605975421,
+      "grad_norm": 0.2742045819759369,
+      "learning_rate": 6.48911745171038e-08,
+      "loss": 0.4138,
       "step": 135695
     },
     {
-      "epoch": 4.77,
-      "learning_rate": 2.706904529908849e-07,
-      "loss": 0.2527,
+      "epoch": 4.890618805636645,
+      "grad_norm": 0.2716894745826721,
+      "learning_rate": 6.468122748406469e-08,
+      "loss": 0.3458,
       "step": 135700
     },
     {
-      "epoch": 4.77,
-      "learning_rate": 2.70272540038663e-07,
-      "loss": 0.2475,
+      "epoch": 4.89079900529787,
+      "grad_norm": 0.2410261183977127,
+      "learning_rate": 6.447162018991537e-08,
+      "loss": 0.3588,
       "step": 135705
     },
     {
-      "epoch": 4.77,
-      "learning_rate": 2.6985494818565215e-07,
-      "loss": 0.2505,
+      "epoch": 4.890979204959095,
+      "grad_norm": 0.2710227072238922,
+      "learning_rate": 6.426235263751468e-08,
+      "loss": 0.3844,
       "step": 135710
     },
     {
-      "epoch": 4.77,
-      "learning_rate": 2.6943767743728145e-07,
-      "loss": 0.2667,
+      "epoch": 4.8911594046203195,
+      "grad_norm": 0.27327826619148254,
+      "learning_rate": 6.405342482970755e-08,
+      "loss": 0.3764,
       "step": 135715
     },
     {
-      "epoch": 4.78,
-      "learning_rate": 2.6902072779896314e-07,
-      "loss": 0.2407,
+      "epoch": 4.891339604281544,
+      "grad_norm": 0.2611567974090576,
+      "learning_rate": 6.384483676934727e-08,
+      "loss": 0.3772,
       "step": 135720
     },
     {
-      "epoch": 4.78,
-      "learning_rate": 2.686040992761096e-07,
-      "loss": 0.2654,
+      "epoch": 4.891519803942769,
+      "grad_norm": 0.23226478695869446,
+      "learning_rate": 6.363658845927323e-08,
+      "loss": 0.3606,
       "step": 135725
     },
     {
-      "epoch": 4.78,
-      "learning_rate": 2.6818779187413324e-07,
-      "loss": 0.2428,
+      "epoch": 4.891700003603994,
+      "grad_norm": 0.28792256116867065,
+      "learning_rate": 6.342867990232481e-08,
+      "loss": 0.3684,
       "step": 135730
     },
     {
-      "epoch": 4.78,
-      "learning_rate": 2.677718055984435e-07,
-      "loss": 0.249,
+      "epoch": 4.8918802032652176,
+      "grad_norm": 0.24476850032806396,
+      "learning_rate": 6.322111110133033e-08,
+      "loss": 0.3505,
       "step": 135735
     },
     {
-      "epoch": 4.78,
-      "learning_rate": 2.6735614045443615e-07,
-      "loss": 0.2608,
+      "epoch": 4.892060402926442,
+      "grad_norm": 0.25403422117233276,
+      "learning_rate": 6.301388205912084e-08,
+      "loss": 0.3466,
       "step": 135740
     },
     {
-      "epoch": 4.78,
-      "learning_rate": 2.669407964475068e-07,
-      "loss": 0.2707,
+      "epoch": 4.892240602587667,
+      "grad_norm": 0.2334025502204895,
+      "learning_rate": 6.28069927785191e-08,
+      "loss": 0.3667,
       "step": 135745
     },
     {
-      "epoch": 4.78,
-      "learning_rate": 2.66525773583054e-07,
-      "loss": 0.2643,
+      "epoch": 4.892420802248892,
+      "grad_norm": 0.30650824308395386,
+      "learning_rate": 6.260044326234227e-08,
+      "loss": 0.3441,
       "step": 135750
     },
     {
-      "epoch": 4.78,
-      "learning_rate": 2.661110718664622e-07,
-      "loss": 0.2534,
+      "epoch": 4.8926010019101165,
+      "grad_norm": 0.2807537317276001,
+      "learning_rate": 6.239423351341034e-08,
+      "loss": 0.3514,
       "step": 135755
     },
     {
-      "epoch": 4.78,
-      "learning_rate": 2.656966913031217e-07,
-      "loss": 0.2655,
+      "epoch": 4.892781201571341,
+      "grad_norm": 0.28259849548339844,
+      "learning_rate": 6.218836353452662e-08,
+      "loss": 0.3651,
       "step": 135760
     },
     {
-      "epoch": 4.78,
-      "learning_rate": 2.6528263189840583e-07,
-      "loss": 0.2535,
+      "epoch": 4.892961401232566,
+      "grad_norm": 0.2717583179473877,
+      "learning_rate": 6.198283332849719e-08,
+      "loss": 0.3725,
       "step": 135765
     },
     {
-      "epoch": 4.78,
-      "learning_rate": 2.6486889365769374e-07,
-      "loss": 0.2756,
+      "epoch": 4.893141600893791,
+      "grad_norm": 0.24983961880207062,
+      "learning_rate": 6.177764289812538e-08,
+      "loss": 0.3589,
       "step": 135770
     },
     {
-      "epoch": 4.78,
-      "learning_rate": 2.644554765863616e-07,
-      "loss": 0.2767,
+      "epoch": 4.8933218005550145,
+      "grad_norm": 0.2629464566707611,
+      "learning_rate": 6.157279224620616e-08,
+      "loss": 0.361,
       "step": 135775
     },
     {
-      "epoch": 4.78,
-      "learning_rate": 2.6404238068977184e-07,
-      "loss": 0.2491,
+      "epoch": 4.893502000216239,
+      "grad_norm": 0.22154061496257782,
+      "learning_rate": 6.136828137552619e-08,
+      "loss": 0.3872,
       "step": 135780
     },
     {
-      "epoch": 4.78,
-      "learning_rate": 2.6362960597328954e-07,
-      "loss": 0.2487,
+      "epoch": 4.893682199877464,
+      "grad_norm": 0.294659823179245,
+      "learning_rate": 6.11641102888777e-08,
+      "loss": 0.3613,
       "step": 135785
     },
     {
-      "epoch": 4.78,
-      "learning_rate": 2.6321715244227443e-07,
-      "loss": 0.2472,
+      "epoch": 4.893862399538689,
+      "grad_norm": 0.20993182063102722,
+      "learning_rate": 6.0960278989039e-08,
+      "loss": 0.3675,
       "step": 135790
     },
     {
-      "epoch": 4.78,
-      "learning_rate": 2.6280502010208594e-07,
-      "loss": 0.2413,
+      "epoch": 4.8940425991999135,
+      "grad_norm": 0.21128609776496887,
+      "learning_rate": 6.075678747878567e-08,
+      "loss": 0.3701,
       "step": 135795
     },
     {
-      "epoch": 4.78,
-      "learning_rate": 2.6239320895806997e-07,
-      "loss": 0.2438,
+      "epoch": 4.894222798861138,
+      "grad_norm": 0.22223599255084991,
+      "learning_rate": 6.055363576089601e-08,
+      "loss": 0.3529,
       "step": 135800
     },
     {
-      "epoch": 4.78,
-      "learning_rate": 2.6198171901557213e-07,
-      "loss": 0.2769,
+      "epoch": 4.894402998522363,
+      "grad_norm": 0.24055111408233643,
+      "learning_rate": 6.035082383813451e-08,
+      "loss": 0.3486,
       "step": 135805
     },
     {
-      "epoch": 4.78,
-      "learning_rate": 2.615705502799437e-07,
-      "loss": 0.2551,
+      "epoch": 4.894583198183588,
+      "grad_norm": 0.30938276648521423,
+      "learning_rate": 6.014835171326283e-08,
+      "loss": 0.3945,
       "step": 135810
     },
     {
-      "epoch": 4.78,
-      "learning_rate": 2.611597027565138e-07,
-      "loss": 0.2648,
+      "epoch": 4.894763397844812,
+      "grad_norm": 0.2910751700401306,
+      "learning_rate": 5.994621938904266e-08,
+      "loss": 0.337,
       "step": 135815
     },
     {
-      "epoch": 4.78,
-      "learning_rate": 2.6074917645062537e-07,
-      "loss": 0.2381,
+      "epoch": 4.894943597506037,
+      "grad_norm": 0.2789898216724396,
+      "learning_rate": 5.974442686822457e-08,
+      "loss": 0.3547,
       "step": 135820
     },
     {
-      "epoch": 4.78,
-      "learning_rate": 2.6033897136760186e-07,
-      "loss": 0.2514,
+      "epoch": 4.895123797167261,
+      "grad_norm": 0.2609977424144745,
+      "learning_rate": 5.9542974153561916e-08,
+      "loss": 0.3549,
       "step": 135825
     },
     {
-      "epoch": 4.78,
-      "learning_rate": 2.599290875127752e-07,
-      "loss": 0.2506,
+      "epoch": 4.895303996828486,
+      "grad_norm": 0.2318691611289978,
+      "learning_rate": 5.934186124779972e-08,
+      "loss": 0.352,
       "step": 135830
     },
     {
-      "epoch": 4.78,
-      "learning_rate": 2.5951952489146613e-07,
-      "loss": 0.2424,
+      "epoch": 4.8954841964897104,
+      "grad_norm": 0.24600878357887268,
+      "learning_rate": 5.914108815367192e-08,
+      "loss": 0.3723,
       "step": 135835
     },
     {
-      "epoch": 4.78,
-      "learning_rate": 2.5911028350898704e-07,
-      "loss": 0.2303,
+      "epoch": 4.895664396150935,
+      "grad_norm": 0.2980553209781647,
+      "learning_rate": 5.894065487392075e-08,
+      "loss": 0.3656,
       "step": 135840
     },
     {
-      "epoch": 4.78,
-      "learning_rate": 2.5870136337065863e-07,
-      "loss": 0.2387,
+      "epoch": 4.89584459581216,
+      "grad_norm": 0.2521979808807373,
+      "learning_rate": 5.8740561411271824e-08,
+      "loss": 0.3641,
       "step": 135845
     },
     {
-      "epoch": 4.78,
-      "learning_rate": 2.5829276448179063e-07,
-      "loss": 0.284,
+      "epoch": 4.896024795473385,
+      "grad_norm": 0.31992876529693604,
+      "learning_rate": 5.854080776845627e-08,
+      "loss": 0.3926,
       "step": 135850
     },
     {
-      "epoch": 4.78,
-      "learning_rate": 2.578844868476815e-07,
-      "loss": 0.2156,
+      "epoch": 4.896204995134609,
+      "grad_norm": 0.28693369030952454,
+      "learning_rate": 5.8341393948194154e-08,
+      "loss": 0.3604,
       "step": 135855
     },
     {
-      "epoch": 4.78,
-      "learning_rate": 2.574765304736382e-07,
-      "loss": 0.2494,
+      "epoch": 4.896385194795834,
+      "grad_norm": 0.2383836805820465,
+      "learning_rate": 5.814231995319996e-08,
+      "loss": 0.3979,
       "step": 135860
     },
     {
-      "epoch": 4.78,
-      "learning_rate": 2.570688953649564e-07,
-      "loss": 0.2516,
+      "epoch": 4.896565394457058,
+      "grad_norm": 0.32336336374282837,
+      "learning_rate": 5.794358578618819e-08,
+      "loss": 0.3494,
       "step": 135865
     },
     {
-      "epoch": 4.78,
-      "learning_rate": 2.5666158152692907e-07,
-      "loss": 0.2611,
+      "epoch": 4.896745594118283,
+      "grad_norm": 0.2836247682571411,
+      "learning_rate": 5.774519144986501e-08,
+      "loss": 0.3809,
       "step": 135870
     },
     {
-      "epoch": 4.78,
-      "learning_rate": 2.562545889648438e-07,
-      "loss": 0.2535,
+      "epoch": 4.896925793779507,
+      "grad_norm": 0.2685668170452118,
+      "learning_rate": 5.754713694693659e-08,
+      "loss": 0.3776,
       "step": 135875
     },
     {
-      "epoch": 4.78,
-      "learning_rate": 2.558479176839851e-07,
-      "loss": 0.2218,
+      "epoch": 4.897105993440732,
+      "grad_norm": 0.28975343704223633,
+      "learning_rate": 5.734942228009799e-08,
+      "loss": 0.3755,
       "step": 135880
     },
     {
-      "epoch": 4.78,
-      "learning_rate": 2.554415676896349e-07,
-      "loss": 0.2752,
+      "epoch": 4.897286193101957,
+      "grad_norm": 0.3143594264984131,
+      "learning_rate": 5.715204745204428e-08,
+      "loss": 0.3687,
       "step": 135885
     },
     {
-      "epoch": 4.78,
-      "learning_rate": 2.550355389870723e-07,
-      "loss": 0.2576,
+      "epoch": 4.897466392763182,
+      "grad_norm": 0.21849270164966583,
+      "learning_rate": 5.695501246546775e-08,
+      "loss": 0.3852,
       "step": 135890
     },
     {
-      "epoch": 4.78,
-      "learning_rate": 2.546298315815626e-07,
-      "loss": 0.2479,
+      "epoch": 4.897646592424406,
+      "grad_norm": 0.26976725459098816,
+      "learning_rate": 5.6758317323046815e-08,
+      "loss": 0.3919,
       "step": 135895
     },
     {
-      "epoch": 4.78,
-      "learning_rate": 2.5422444547837376e-07,
-      "loss": 0.2629,
+      "epoch": 4.897826792085631,
+      "grad_norm": 0.32330963015556335,
+      "learning_rate": 5.656196202746544e-08,
+      "loss": 0.3806,
       "step": 135900
     },
     {
-      "epoch": 4.78,
-      "learning_rate": 2.5381938068277654e-07,
-      "loss": 0.2659,
+      "epoch": 4.898006991746856,
+      "grad_norm": 0.26278913021087646,
+      "learning_rate": 5.6365946581399265e-08,
+      "loss": 0.3952,
       "step": 135905
     },
     {
-      "epoch": 4.78,
-      "learning_rate": 2.534146372000251e-07,
-      "loss": 0.2494,
+      "epoch": 4.898187191408081,
+      "grad_norm": 0.28790876269340515,
+      "learning_rate": 5.617027098751559e-08,
+      "loss": 0.3506,
       "step": 135910
     },
     {
-      "epoch": 4.78,
-      "learning_rate": 2.530102150353764e-07,
-      "loss": 0.2525,
+      "epoch": 4.898367391069304,
+      "grad_norm": 0.26349297165870667,
+      "learning_rate": 5.597493524848452e-08,
+      "loss": 0.3843,
       "step": 135915
     },
     {
-      "epoch": 4.78,
-      "learning_rate": 2.5260611419408163e-07,
-      "loss": 0.2237,
+      "epoch": 4.898547590730529,
+      "grad_norm": 0.2596375644207001,
+      "learning_rate": 5.577993936696502e-08,
+      "loss": 0.3871,
       "step": 135920
     },
     {
-      "epoch": 4.78,
-      "learning_rate": 2.52202334681384e-07,
-      "loss": 0.2504,
+      "epoch": 4.898727790391754,
+      "grad_norm": 0.22803239524364471,
+      "learning_rate": 5.5585283345613304e-08,
+      "loss": 0.3691,
       "step": 135925
     },
     {
-      "epoch": 4.78,
-      "learning_rate": 2.5179887650253476e-07,
-      "loss": 0.232,
+      "epoch": 4.898907990052979,
+      "grad_norm": 0.22493457794189453,
+      "learning_rate": 5.5390967187085585e-08,
+      "loss": 0.37,
       "step": 135930
     },
     {
-      "epoch": 4.78,
-      "learning_rate": 2.513957396627631e-07,
-      "loss": 0.253,
+      "epoch": 4.899088189714203,
+      "grad_norm": 0.2446439415216446,
+      "learning_rate": 5.519699089402419e-08,
+      "loss": 0.3802,
       "step": 135935
     },
     {
-      "epoch": 4.78,
-      "learning_rate": 2.5099292416730925e-07,
-      "loss": 0.231,
+      "epoch": 4.899268389375428,
+      "grad_norm": 0.2472543567419052,
+      "learning_rate": 5.5003354469077006e-08,
+      "loss": 0.3364,
       "step": 135940
     },
     {
-      "epoch": 4.78,
-      "learning_rate": 2.5059043002140235e-07,
-      "loss": 0.2375,
+      "epoch": 4.899448589036653,
+      "grad_norm": 0.23050671815872192,
+      "learning_rate": 5.481005791487526e-08,
+      "loss": 0.3832,
       "step": 135945
     },
     {
-      "epoch": 4.78,
-      "learning_rate": 2.5018825723026884e-07,
-      "loss": 0.2539,
+      "epoch": 4.8996287886978775,
+      "grad_norm": 0.3573814034461975,
+      "learning_rate": 5.461710123406128e-08,
+      "loss": 0.368,
       "step": 135950
     },
     {
-      "epoch": 4.78,
-      "learning_rate": 2.4978640579912947e-07,
-      "loss": 0.2531,
+      "epoch": 4.899808988359101,
+      "grad_norm": 0.24702803790569305,
+      "learning_rate": 5.442448442925796e-08,
+      "loss": 0.3933,
       "step": 135955
     },
     {
-      "epoch": 4.78,
-      "learning_rate": 2.4938487573320234e-07,
-      "loss": 0.2811,
+      "epoch": 4.899989188020326,
+      "grad_norm": 0.20937135815620422,
+      "learning_rate": 5.423220750309099e-08,
+      "loss": 0.3516,
       "step": 135960
     },
     {
-      "epoch": 4.78,
-      "learning_rate": 2.489836670377027e-07,
-      "loss": 0.2536,
+      "epoch": 4.900169387681551,
+      "grad_norm": 0.30278798937797546,
+      "learning_rate": 5.4040270458180496e-08,
+      "loss": 0.3488,
       "step": 135965
     },
     {
-      "epoch": 4.78,
-      "learning_rate": 2.485827797178375e-07,
-      "loss": 0.2492,
+      "epoch": 4.900349587342776,
+      "grad_norm": 0.2523956298828125,
+      "learning_rate": 5.384867329714105e-08,
+      "loss": 0.3533,
       "step": 135970
     },
     {
-      "epoch": 4.78,
-      "learning_rate": 2.481822137788137e-07,
-      "loss": 0.2356,
+      "epoch": 4.900529787004,
+      "grad_norm": 0.30250051617622375,
+      "learning_rate": 5.365741602258445e-08,
+      "loss": 0.3883,
       "step": 135975
     },
     {
-      "epoch": 4.78,
-      "learning_rate": 2.4778196922582984e-07,
-      "loss": 0.2409,
+      "epoch": 4.900709986665225,
+      "grad_norm": 0.21715682744979858,
+      "learning_rate": 5.3466498637116945e-08,
+      "loss": 0.3592,
       "step": 135980
     },
     {
-      "epoch": 4.78,
-      "learning_rate": 2.4738204606408465e-07,
-      "loss": 0.2646,
+      "epoch": 4.90089018632645,
+      "grad_norm": 0.2727758586406708,
+      "learning_rate": 5.327592114333646e-08,
+      "loss": 0.3527,
       "step": 135985
     },
     {
-      "epoch": 4.78,
-      "learning_rate": 2.469824442987739e-07,
-      "loss": 0.2426,
+      "epoch": 4.9010703859876745,
+      "grad_norm": 0.33506885170936584,
+      "learning_rate": 5.308568354384369e-08,
+      "loss": 0.4078,
       "step": 135990
     },
     {
-      "epoch": 4.78,
-      "learning_rate": 2.465831639350824e-07,
-      "loss": 0.2461,
+      "epoch": 4.901250585648899,
+      "grad_norm": 0.2240988165140152,
+      "learning_rate": 5.289578584122823e-08,
+      "loss": 0.3615,
       "step": 135995
     },
     {
-      "epoch": 4.78,
-      "learning_rate": 2.4618420497819484e-07,
-      "loss": 0.2414,
+      "epoch": 4.901430785310124,
+      "grad_norm": 0.2511599361896515,
+      "learning_rate": 5.27062280380769e-08,
+      "loss": 0.3921,
       "step": 136000
     },
     {
-      "epoch": 4.78,
-      "eval_loss": 0.2484423816204071,
-      "eval_runtime": 10.5512,
-      "eval_samples_per_second": 9.478,
-      "eval_steps_per_second": 9.478,
+      "epoch": 4.901430785310124,
+      "eval_loss": 0.4287901222705841,
+      "eval_runtime": 3.5335,
+      "eval_samples_per_second": 28.3,
+      "eval_steps_per_second": 7.075,
       "step": 136000
     },
     {
-      "epoch": 4.79,
-      "learning_rate": 2.457855674332932e-07,
-      "loss": 0.2498,
+      "epoch": 4.901610984971349,
+      "grad_norm": 0.27596327662467957,
+      "learning_rate": 5.251701013697374e-08,
+      "loss": 0.3939,
       "step": 136005
     },
     {
-      "epoch": 4.79,
-      "learning_rate": 2.453872513055511e-07,
-      "loss": 0.2329,
+      "epoch": 4.901791184632573,
+      "grad_norm": 0.3502033054828644,
+      "learning_rate": 5.2328132140497255e-08,
+      "loss": 0.3705,
       "step": 136010
     },
     {
-      "epoch": 4.79,
-      "learning_rate": 2.449892566001394e-07,
-      "loss": 0.2374,
+      "epoch": 4.901971384293797,
+      "grad_norm": 0.269580602645874,
+      "learning_rate": 5.21395940512176e-08,
+      "loss": 0.3879,
       "step": 136015
     },
     {
-      "epoch": 4.79,
-      "learning_rate": 2.4459158332223173e-07,
-      "loss": 0.2284,
+      "epoch": 4.902151583955022,
+      "grad_norm": 0.24801914393901825,
+      "learning_rate": 5.195139587170772e-08,
+      "loss": 0.3859,
       "step": 136020
     },
     {
-      "epoch": 4.79,
-      "learning_rate": 2.4419423147698793e-07,
-      "loss": 0.244,
+      "epoch": 4.902331783616247,
+      "grad_norm": 0.2341059297323227,
+      "learning_rate": 5.1763537604529453e-08,
+      "loss": 0.371,
       "step": 136025
     },
     {
-      "epoch": 4.79,
-      "learning_rate": 2.437972010695649e-07,
-      "loss": 0.2669,
+      "epoch": 4.9025119832774715,
+      "grad_norm": 0.21445782482624054,
+      "learning_rate": 5.1576019252241866e-08,
+      "loss": 0.3202,
       "step": 136030
     },
     {
-      "epoch": 4.79,
-      "learning_rate": 2.434004921051197e-07,
-      "loss": 0.2353,
+      "epoch": 4.902692182938696,
+      "grad_norm": 0.1884135752916336,
+      "learning_rate": 5.138884081740125e-08,
+      "loss": 0.3182,
       "step": 136035
     },
     {
-      "epoch": 4.79,
-      "learning_rate": 2.4300410458880654e-07,
-      "loss": 0.2529,
+      "epoch": 4.902872382599921,
+      "grad_norm": 0.24446237087249756,
+      "learning_rate": 5.120200230255834e-08,
+      "loss": 0.3877,
       "step": 136040
     },
     {
-      "epoch": 4.79,
-      "learning_rate": 2.426080385257684e-07,
-      "loss": 0.26,
+      "epoch": 4.903052582261146,
+      "grad_norm": 0.26809754967689514,
+      "learning_rate": 5.101550371025832e-08,
+      "loss": 0.3798,
       "step": 136045
     },
     {
-      "epoch": 4.79,
-      "learning_rate": 2.422122939211513e-07,
-      "loss": 0.2691,
+      "epoch": 4.9032327819223696,
+      "grad_norm": 0.2706208825111389,
+      "learning_rate": 5.082934504304082e-08,
+      "loss": 0.3462,
       "step": 136050
     },
     {
-      "epoch": 4.79,
-      "learning_rate": 2.418168707800872e-07,
-      "loss": 0.2767,
+      "epoch": 4.903412981583594,
+      "grad_norm": 0.2580229341983795,
+      "learning_rate": 5.0643526303445485e-08,
+      "loss": 0.3824,
       "step": 136055
     },
     {
-      "epoch": 4.79,
-      "learning_rate": 2.414217691077192e-07,
-      "loss": 0.2778,
+      "epoch": 4.903593181244819,
+      "grad_norm": 0.24854421615600586,
+      "learning_rate": 5.045804749399807e-08,
+      "loss": 0.3535,
       "step": 136060
     },
     {
-      "epoch": 4.79,
-      "learning_rate": 2.41026988909171e-07,
-      "loss": 0.2329,
+      "epoch": 4.903773380906044,
+      "grad_norm": 0.22989878058433533,
+      "learning_rate": 5.0272908617229885e-08,
+      "loss": 0.3991,
       "step": 136065
     },
     {
-      "epoch": 4.79,
-      "learning_rate": 2.40632530189569e-07,
-      "loss": 0.2373,
+      "epoch": 4.9039535805672685,
+      "grad_norm": 0.24265380203723907,
+      "learning_rate": 5.0088109675666684e-08,
+      "loss": 0.3802,
       "step": 136070
     },
     {
-      "epoch": 4.79,
-      "learning_rate": 2.402383929540397e-07,
-      "loss": 0.2503,
+      "epoch": 4.904133780228493,
+      "grad_norm": 0.2347709685564041,
+      "learning_rate": 4.990365067181757e-08,
+      "loss": 0.3465,
       "step": 136075
     },
     {
-      "epoch": 4.79,
-      "learning_rate": 2.398445772076957e-07,
-      "loss": 0.2404,
+      "epoch": 4.904313979889718,
+      "grad_norm": 0.18331743776798248,
+      "learning_rate": 4.9719531608205527e-08,
+      "loss": 0.3867,
       "step": 136080
     },
     {
-      "epoch": 4.79,
-      "learning_rate": 2.394510829556523e-07,
-      "loss": 0.2588,
+      "epoch": 4.904494179550943,
+      "grad_norm": 0.3420829772949219,
+      "learning_rate": 4.9535752487331335e-08,
+      "loss": 0.3615,
       "step": 136085
     },
     {
-      "epoch": 4.79,
-      "learning_rate": 2.390579102030166e-07,
-      "loss": 0.2546,
+      "epoch": 4.904674379212167,
+      "grad_norm": 0.31512126326560974,
+      "learning_rate": 4.9352313311701317e-08,
+      "loss": 0.3836,
       "step": 136090
     },
     {
-      "epoch": 4.79,
-      "learning_rate": 2.3866505895489833e-07,
-      "loss": 0.2649,
+      "epoch": 4.904854578873392,
+      "grad_norm": 0.3358932137489319,
+      "learning_rate": 4.916921408381625e-08,
+      "loss": 0.3708,
       "step": 136095
     },
     {
-      "epoch": 4.79,
-      "learning_rate": 2.382725292163962e-07,
-      "loss": 0.2536,
+      "epoch": 4.905034778534616,
+      "grad_norm": 0.21517984569072723,
+      "learning_rate": 4.898645480617137e-08,
+      "loss": 0.3639,
       "step": 136100
     },
     {
-      "epoch": 4.79,
-      "learning_rate": 2.3788032099260893e-07,
-      "loss": 0.2585,
+      "epoch": 4.905214978195841,
+      "grad_norm": 0.28815269470214844,
+      "learning_rate": 4.880403548125356e-08,
+      "loss": 0.3626,
       "step": 136105
     },
     {
-      "epoch": 4.79,
-      "learning_rate": 2.374884342886241e-07,
-      "loss": 0.237,
+      "epoch": 4.9053951778570655,
+      "grad_norm": 0.2533737123012543,
+      "learning_rate": 4.8621956111549735e-08,
+      "loss": 0.3567,
       "step": 136110
     },
     {
-      "epoch": 4.79,
-      "learning_rate": 2.370968691095321e-07,
-      "loss": 0.2487,
+      "epoch": 4.90557537751829,
+      "grad_norm": 0.2487945407629013,
+      "learning_rate": 4.8440216699544015e-08,
+      "loss": 0.3604,
       "step": 136115
     },
     {
-      "epoch": 4.79,
-      "learning_rate": 2.3670562546042053e-07,
-      "loss": 0.2351,
+      "epoch": 4.905755577179515,
+      "grad_norm": 0.22873222827911377,
+      "learning_rate": 4.8258817247706646e-08,
+      "loss": 0.346,
       "step": 136120
     },
     {
-      "epoch": 4.79,
-      "learning_rate": 2.3631470334636584e-07,
-      "loss": 0.2577,
+      "epoch": 4.90593577684074,
+      "grad_norm": 0.2279924899339676,
+      "learning_rate": 4.807775775851342e-08,
+      "loss": 0.3948,
       "step": 136125
     },
     {
-      "epoch": 4.79,
-      "learning_rate": 2.359241027724418e-07,
-      "loss": 0.2367,
+      "epoch": 4.906115976501964,
+      "grad_norm": 0.24945561587810516,
+      "learning_rate": 4.7897038234429035e-08,
+      "loss": 0.3805,
       "step": 136130
     },
     {
-      "epoch": 4.79,
-      "learning_rate": 2.355338237437277e-07,
-      "loss": 0.2543,
+      "epoch": 4.906296176163189,
+      "grad_norm": 0.19843456149101257,
+      "learning_rate": 4.7716658677918194e-08,
+      "loss": 0.3494,
       "step": 136135
     },
     {
-      "epoch": 4.79,
-      "learning_rate": 2.3514386626528328e-07,
-      "loss": 0.2224,
+      "epoch": 4.906476375824413,
+      "grad_norm": 0.2849428057670593,
+      "learning_rate": 4.753661909143448e-08,
+      "loss": 0.4041,
       "step": 136140
     },
     {
-      "epoch": 4.79,
-      "learning_rate": 2.3475423034217402e-07,
-      "loss": 0.256,
+      "epoch": 4.906656575485638,
+      "grad_norm": 0.2670860290527344,
+      "learning_rate": 4.735691947743703e-08,
+      "loss": 0.3781,
       "step": 136145
     },
     {
-      "epoch": 4.79,
-      "learning_rate": 2.3436491597946253e-07,
-      "loss": 0.2579,
+      "epoch": 4.9068367751468625,
+      "grad_norm": 0.2657918930053711,
+      "learning_rate": 4.717755983836836e-08,
+      "loss": 0.3338,
       "step": 136150
     },
     {
-      "epoch": 4.79,
-      "learning_rate": 2.3397592318220308e-07,
-      "loss": 0.2449,
+      "epoch": 4.907016974808087,
+      "grad_norm": 0.2555624544620514,
+      "learning_rate": 4.699854017667371e-08,
+      "loss": 0.3745,
       "step": 136155
     },
     {
-      "epoch": 4.79,
-      "learning_rate": 2.3358725195544162e-07,
-      "loss": 0.2583,
+      "epoch": 4.907197174469312,
+      "grad_norm": 0.2596377432346344,
+      "learning_rate": 4.6819860494792814e-08,
+      "loss": 0.3466,
       "step": 136160
     },
     {
-      "epoch": 4.79,
-      "learning_rate": 2.3319890230422968e-07,
-      "loss": 0.2448,
+      "epoch": 4.907377374130537,
+      "grad_norm": 0.21930861473083496,
+      "learning_rate": 4.6641520795162596e-08,
+      "loss": 0.3511,
       "step": 136165
     },
     {
-      "epoch": 4.79,
-      "learning_rate": 2.328108742336077e-07,
-      "loss": 0.2609,
+      "epoch": 4.907557573791761,
+      "grad_norm": 0.2237590104341507,
+      "learning_rate": 4.6463521080208906e-08,
+      "loss": 0.3618,
       "step": 136170
     },
     {
-      "epoch": 4.79,
-      "learning_rate": 2.3242316774861604e-07,
-      "loss": 0.2442,
+      "epoch": 4.907737773452986,
+      "grad_norm": 0.2339451014995575,
+      "learning_rate": 4.6285861352357574e-08,
+      "loss": 0.3686,
       "step": 136175
     },
     {
-      "epoch": 4.79,
-      "learning_rate": 2.3203578285428406e-07,
-      "loss": 0.2456,
+      "epoch": 4.907917973114211,
+      "grad_norm": 0.27476266026496887,
+      "learning_rate": 4.610854161403166e-08,
+      "loss": 0.3819,
       "step": 136180
     },
     {
-      "epoch": 4.79,
-      "learning_rate": 2.316487195556466e-07,
-      "loss": 0.2538,
+      "epoch": 4.908098172775436,
+      "grad_norm": 0.2562359869480133,
+      "learning_rate": 4.593156186764591e-08,
+      "loss": 0.3852,
       "step": 136185
     },
     {
-      "epoch": 4.79,
-      "learning_rate": 2.3126197785772742e-07,
-      "loss": 0.2818,
+      "epoch": 4.908278372436659,
+      "grad_norm": 0.2553425133228302,
+      "learning_rate": 4.57549221156095e-08,
+      "loss": 0.3646,
       "step": 136190
     },
     {
-      "epoch": 4.79,
-      "learning_rate": 2.3087555776554747e-07,
-      "loss": 0.2886,
+      "epoch": 4.908458572097884,
+      "grad_norm": 0.25126445293426514,
+      "learning_rate": 4.5578622360334386e-08,
+      "loss": 0.3564,
       "step": 136195
     },
     {
-      "epoch": 4.79,
-      "learning_rate": 2.3048945928412502e-07,
-      "loss": 0.2577,
+      "epoch": 4.908638771759109,
+      "grad_norm": 0.2578040659427643,
+      "learning_rate": 4.540266260421588e-08,
+      "loss": 0.346,
       "step": 136200
     },
     {
-      "epoch": 4.79,
-      "learning_rate": 2.3010368241847268e-07,
-      "loss": 0.2543,
+      "epoch": 4.908818971420334,
+      "grad_norm": 0.2958472967147827,
+      "learning_rate": 4.5227042849654844e-08,
+      "loss": 0.4071,
       "step": 136205
     },
     {
-      "epoch": 4.79,
-      "learning_rate": 2.2971822717360038e-07,
-      "loss": 0.2468,
+      "epoch": 4.908999171081558,
+      "grad_norm": 0.2669653594493866,
+      "learning_rate": 4.505176309904657e-08,
+      "loss": 0.3787,
       "step": 136210
     },
     {
-      "epoch": 4.79,
-      "learning_rate": 2.2933309355451515e-07,
-      "loss": 0.253,
+      "epoch": 4.909179370742783,
+      "grad_norm": 0.294811874628067,
+      "learning_rate": 4.4876823354775275e-08,
+      "loss": 0.3732,
       "step": 136215
     },
     {
-      "epoch": 4.79,
-      "learning_rate": 2.2894828156621305e-07,
-      "loss": 0.2237,
+      "epoch": 4.909359570404008,
+      "grad_norm": 0.2965461015701294,
+      "learning_rate": 4.470222361922516e-08,
+      "loss": 0.3712,
       "step": 136220
     },
     {
-      "epoch": 4.79,
-      "learning_rate": 2.2856379121369288e-07,
-      "loss": 0.2673,
+      "epoch": 4.909539770065233,
+      "grad_norm": 0.24471497535705566,
+      "learning_rate": 4.452796389477487e-08,
+      "loss": 0.355,
       "step": 136225
     },
     {
-      "epoch": 4.79,
-      "learning_rate": 2.281796225019478e-07,
-      "loss": 0.2708,
+      "epoch": 4.909719969726457,
+      "grad_norm": 0.2504851520061493,
+      "learning_rate": 4.435404418380029e-08,
+      "loss": 0.3638,
       "step": 136230
     },
     {
-      "epoch": 4.79,
-      "learning_rate": 2.2779577543596276e-07,
-      "loss": 0.2639,
+      "epoch": 4.909900169387681,
+      "grad_norm": 0.26970013976097107,
+      "learning_rate": 4.418046448867175e-08,
+      "loss": 0.3881,
       "step": 136235
     },
     {
-      "epoch": 4.79,
-      "learning_rate": 2.274122500207282e-07,
-      "loss": 0.2538,
+      "epoch": 4.910080369048906,
+      "grad_norm": 0.31252387166023254,
+      "learning_rate": 4.400722481175401e-08,
+      "loss": 0.4242,
       "step": 136240
     },
     {
-      "epoch": 4.79,
-      "learning_rate": 2.2702904626121514e-07,
-      "loss": 0.2511,
+      "epoch": 4.910260568710131,
+      "grad_norm": 0.27864494919776917,
+      "learning_rate": 4.3834325155403535e-08,
+      "loss": 0.3557,
       "step": 136245
     },
     {
-      "epoch": 4.79,
-      "learning_rate": 2.266461641624057e-07,
-      "loss": 0.2426,
+      "epoch": 4.910440768371355,
+      "grad_norm": 0.2759137451648712,
+      "learning_rate": 4.366176552197676e-08,
+      "loss": 0.3641,
       "step": 136250
     },
     {
-      "epoch": 4.79,
-      "learning_rate": 2.2626360372927092e-07,
-      "loss": 0.269,
+      "epoch": 4.91062096803258,
+      "grad_norm": 0.23578329384326935,
+      "learning_rate": 4.348954591383014e-08,
+      "loss": 0.3712,
       "step": 136255
     },
     {
-      "epoch": 4.79,
-      "learning_rate": 2.2588136496677625e-07,
-      "loss": 0.2443,
+      "epoch": 4.910801167693805,
+      "grad_norm": 0.2522645592689514,
+      "learning_rate": 4.331766633330625e-08,
+      "loss": 0.3695,
       "step": 136260
     },
     {
-      "epoch": 4.79,
-      "learning_rate": 2.2549944787988163e-07,
-      "loss": 0.2544,
+      "epoch": 4.9109813673550295,
+      "grad_norm": 0.300750195980072,
+      "learning_rate": 4.3146126782747655e-08,
+      "loss": 0.3721,
       "step": 136265
     },
     {
-      "epoch": 4.79,
-      "learning_rate": 2.2511785247355533e-07,
-      "loss": 0.2678,
+      "epoch": 4.911161567016254,
+      "grad_norm": 0.26035916805267334,
+      "learning_rate": 4.29749272644886e-08,
+      "loss": 0.3666,
       "step": 136270
     },
     {
-      "epoch": 4.79,
-      "learning_rate": 2.2473657875274334e-07,
-      "loss": 0.2303,
+      "epoch": 4.911341766677479,
+      "grad_norm": 0.2753373086452484,
+      "learning_rate": 4.28040677808661e-08,
+      "loss": 0.372,
       "step": 136275
     },
     {
-      "epoch": 4.79,
-      "learning_rate": 2.2435562672240006e-07,
-      "loss": 0.2569,
+      "epoch": 4.911521966338704,
+      "grad_norm": 0.24623210728168488,
+      "learning_rate": 4.263354833420607e-08,
+      "loss": 0.3655,
       "step": 136280
     },
     {
-      "epoch": 4.79,
-      "learning_rate": 2.2397499638747155e-07,
-      "loss": 0.2722,
+      "epoch": 4.911702165999928,
+      "grad_norm": 0.25700467824935913,
+      "learning_rate": 4.246336892683445e-08,
+      "loss": 0.3749,
       "step": 136285
     },
     {
-      "epoch": 4.8,
-      "learning_rate": 2.2359468775289827e-07,
-      "loss": 0.2658,
+      "epoch": 4.911882365661152,
+      "grad_norm": 0.21152786910533905,
+      "learning_rate": 4.229352956106603e-08,
+      "loss": 0.3809,
       "step": 136290
     },
     {
-      "epoch": 4.8,
-      "learning_rate": 2.2321470082362072e-07,
-      "loss": 0.2705,
+      "epoch": 4.912062565322377,
+      "grad_norm": 0.3113376200199127,
+      "learning_rate": 4.2124030239212855e-08,
+      "loss": 0.3877,
       "step": 136295
     },
     {
-      "epoch": 4.8,
-      "learning_rate": 2.2283503560457108e-07,
-      "loss": 0.2751,
+      "epoch": 4.912242764983602,
+      "grad_norm": 0.22789722681045532,
+      "learning_rate": 4.195487096359252e-08,
+      "loss": 0.3668,
       "step": 136300
     },
     {
-      "epoch": 4.8,
-      "learning_rate": 2.2245569210068152e-07,
-      "loss": 0.2432,
+      "epoch": 4.9124229646448265,
+      "grad_norm": 0.22643516957759857,
+      "learning_rate": 4.178605173650318e-08,
+      "loss": 0.3519,
       "step": 136305
     },
     {
-      "epoch": 4.8,
-      "learning_rate": 2.2207667031687307e-07,
-      "loss": 0.246,
+      "epoch": 4.912603164306051,
+      "grad_norm": 0.1881403774023056,
+      "learning_rate": 4.161757256024579e-08,
+      "loss": 0.3422,
       "step": 136310
     },
     {
-      "epoch": 4.8,
-      "learning_rate": 2.2169797025807239e-07,
-      "loss": 0.2459,
+      "epoch": 4.912783363967276,
+      "grad_norm": 0.2479417622089386,
+      "learning_rate": 4.144943343711849e-08,
+      "loss": 0.4007,
       "step": 136315
     },
     {
-      "epoch": 4.8,
-      "learning_rate": 2.2131959192919215e-07,
-      "loss": 0.2376,
+      "epoch": 4.912963563628501,
+      "grad_norm": 0.23649077117443085,
+      "learning_rate": 4.128163436941113e-08,
+      "loss": 0.3453,
       "step": 136320
     },
     {
-      "epoch": 4.8,
-      "learning_rate": 2.2094153533514794e-07,
-      "loss": 0.253,
+      "epoch": 4.913143763289725,
+      "grad_norm": 0.25313544273376465,
+      "learning_rate": 4.111417535940798e-08,
+      "loss": 0.4058,
       "step": 136325
     },
     {
-      "epoch": 4.8,
-      "learning_rate": 2.2056380048084968e-07,
-      "loss": 0.2476,
+      "epoch": 4.913323962950949,
+      "grad_norm": 0.2785737216472626,
+      "learning_rate": 4.094705640939056e-08,
+      "loss": 0.3633,
       "step": 136330
     },
     {
-      "epoch": 4.8,
-      "learning_rate": 2.2018638737119902e-07,
-      "loss": 0.2345,
+      "epoch": 4.913504162612174,
+      "grad_norm": 0.292338490486145,
+      "learning_rate": 4.0780277521640374e-08,
+      "loss": 0.3945,
       "step": 136335
     },
     {
-      "epoch": 4.8,
-      "learning_rate": 2.1980929601109756e-07,
-      "loss": 0.2509,
+      "epoch": 4.913684362273399,
+      "grad_norm": 0.2278224229812622,
+      "learning_rate": 4.061383869842506e-08,
+      "loss": 0.3593,
       "step": 136340
     },
     {
-      "epoch": 4.8,
-      "learning_rate": 2.1943252640544143e-07,
-      "loss": 0.2467,
+      "epoch": 4.9138645619346235,
+      "grad_norm": 0.2170296460390091,
+      "learning_rate": 4.044773994201223e-08,
+      "loss": 0.3742,
       "step": 136345
     },
     {
-      "epoch": 4.8,
-      "learning_rate": 2.1905607855912115e-07,
-      "loss": 0.2424,
+      "epoch": 4.914044761595848,
+      "grad_norm": 0.2683648467063904,
+      "learning_rate": 4.0281981254669535e-08,
+      "loss": 0.4156,
       "step": 136350
     },
     {
-      "epoch": 4.8,
-      "learning_rate": 2.186799524770272e-07,
-      "loss": 0.2398,
+      "epoch": 4.914224961257073,
+      "grad_norm": 0.2352461963891983,
+      "learning_rate": 4.011656263865071e-08,
+      "loss": 0.3855,
       "step": 136355
     },
     {
-      "epoch": 4.8,
-      "learning_rate": 2.1830414816404466e-07,
-      "loss": 0.2574,
+      "epoch": 4.914405160918298,
+      "grad_norm": 0.28996291756629944,
+      "learning_rate": 3.99514840962123e-08,
+      "loss": 0.4141,
       "step": 136360
     },
     {
-      "epoch": 4.8,
-      "learning_rate": 2.179286656250501e-07,
-      "loss": 0.2354,
+      "epoch": 4.914585360579522,
+      "grad_norm": 0.22475266456604004,
+      "learning_rate": 3.9786745629602494e-08,
+      "loss": 0.3708,
       "step": 136365
     },
     {
-      "epoch": 4.8,
-      "learning_rate": 2.175535048649202e-07,
-      "loss": 0.2391,
+      "epoch": 4.914765560240747,
+      "grad_norm": 0.19460433721542358,
+      "learning_rate": 3.962234724106395e-08,
+      "loss": 0.3589,
       "step": 136370
     },
     {
-      "epoch": 4.8,
-      "learning_rate": 2.171786658885233e-07,
-      "loss": 0.2304,
+      "epoch": 4.914945759901971,
+      "grad_norm": 0.2615772485733032,
+      "learning_rate": 3.945828893284209e-08,
+      "loss": 0.3617,
       "step": 136375
     },
     {
-      "epoch": 4.8,
-      "learning_rate": 2.1680414870073053e-07,
-      "loss": 0.2502,
+      "epoch": 4.915125959563196,
+      "grad_norm": 0.26762276887893677,
+      "learning_rate": 3.929457070716569e-08,
+      "loss": 0.3672,
       "step": 136380
     },
     {
-      "epoch": 4.8,
-      "learning_rate": 2.164299533064046e-07,
-      "loss": 0.2468,
+      "epoch": 4.9153061592244205,
+      "grad_norm": 0.40078943967819214,
+      "learning_rate": 3.913119256626907e-08,
+      "loss": 0.3814,
       "step": 136385
     },
     {
-      "epoch": 4.8,
-      "learning_rate": 2.1605607971040276e-07,
-      "loss": 0.2589,
+      "epoch": 4.915486358885645,
+      "grad_norm": 0.2580387592315674,
+      "learning_rate": 3.896815451237823e-08,
+      "loss": 0.3731,
       "step": 136390
     },
     {
-      "epoch": 4.8,
-      "learning_rate": 2.156825279175767e-07,
-      "loss": 0.2593,
+      "epoch": 4.91566655854687,
+      "grad_norm": 0.2312445044517517,
+      "learning_rate": 3.880545654771084e-08,
+      "loss": 0.3327,
       "step": 136395
     },
     {
-      "epoch": 4.8,
-      "learning_rate": 2.1530929793278088e-07,
-      "loss": 0.2649,
+      "epoch": 4.915846758208095,
+      "grad_norm": 0.22005638480186462,
+      "learning_rate": 3.864309867449012e-08,
+      "loss": 0.3817,
       "step": 136400
     },
     {
-      "epoch": 4.8,
-      "learning_rate": 2.149363897608614e-07,
-      "loss": 0.2745,
+      "epoch": 4.916026957869319,
+      "grad_norm": 0.2570466995239258,
+      "learning_rate": 3.848108089492264e-08,
+      "loss": 0.3434,
       "step": 136405
     },
     {
-      "epoch": 4.8,
-      "learning_rate": 2.145638034066588e-07,
-      "loss": 0.2492,
+      "epoch": 4.916207157530544,
+      "grad_norm": 0.2323221117258072,
+      "learning_rate": 3.8319403211217744e-08,
+      "loss": 0.3523,
       "step": 136410
     },
     {
-      "epoch": 4.8,
-      "learning_rate": 2.141915388750082e-07,
-      "loss": 0.2456,
+      "epoch": 4.916387357191768,
+      "grad_norm": 0.26294851303100586,
+      "learning_rate": 3.815806562557644e-08,
+      "loss": 0.3646,
       "step": 136415
     },
     {
-      "epoch": 4.8,
-      "learning_rate": 2.1381959617074732e-07,
-      "loss": 0.2561,
+      "epoch": 4.916567556852993,
+      "grad_norm": 0.29720500111579895,
+      "learning_rate": 3.799706814020254e-08,
+      "loss": 0.3537,
       "step": 136420
     },
     {
-      "epoch": 4.8,
-      "learning_rate": 2.1344797529870565e-07,
-      "loss": 0.2624,
+      "epoch": 4.9167477565142175,
+      "grad_norm": 0.27285248041152954,
+      "learning_rate": 3.783641075728317e-08,
+      "loss": 0.3456,
       "step": 136425
     },
     {
-      "epoch": 4.8,
-      "learning_rate": 2.130766762637071e-07,
-      "loss": 0.2427,
+      "epoch": 4.916927956175442,
+      "grad_norm": 0.2373269498348236,
+      "learning_rate": 3.767609347901102e-08,
+      "loss": 0.3396,
       "step": 136430
     },
     {
-      "epoch": 4.8,
-      "learning_rate": 2.1270569907057004e-07,
-      "loss": 0.2615,
+      "epoch": 4.917108155836667,
+      "grad_norm": 0.232501819729805,
+      "learning_rate": 3.751611630756768e-08,
+      "loss": 0.3594,
       "step": 136435
     },
     {
-      "epoch": 4.8,
-      "learning_rate": 2.123350437241156e-07,
-      "loss": 0.2741,
+      "epoch": 4.917288355497892,
+      "grad_norm": 0.24298062920570374,
+      "learning_rate": 3.735647924513475e-08,
+      "loss": 0.3462,
       "step": 136440
     },
     {
-      "epoch": 4.8,
-      "learning_rate": 2.1196471022915665e-07,
-      "loss": 0.2443,
+      "epoch": 4.917468555159116,
+      "grad_norm": 0.2818912863731384,
+      "learning_rate": 3.719718229388824e-08,
+      "loss": 0.3911,
       "step": 136445
     },
     {
-      "epoch": 4.8,
-      "learning_rate": 2.1159469859049764e-07,
-      "loss": 0.2567,
+      "epoch": 4.917648754820341,
+      "grad_norm": 0.2380588948726654,
+      "learning_rate": 3.703822545599589e-08,
+      "loss": 0.424,
       "step": 136450
     },
     {
-      "epoch": 4.8,
-      "learning_rate": 2.1122500881294582e-07,
-      "loss": 0.262,
+      "epoch": 4.917828954481566,
+      "grad_norm": 0.23889237642288208,
+      "learning_rate": 3.687960873362539e-08,
+      "loss": 0.3688,
       "step": 136455
     },
     {
-      "epoch": 4.8,
-      "learning_rate": 2.1085564090130294e-07,
-      "loss": 0.2325,
+      "epoch": 4.918009154142791,
+      "grad_norm": 0.2520101070404053,
+      "learning_rate": 3.672133212893891e-08,
+      "loss": 0.3466,
       "step": 136460
     },
     {
-      "epoch": 4.8,
-      "learning_rate": 2.104865948603596e-07,
-      "loss": 0.2378,
+      "epoch": 4.918189353804015,
+      "grad_norm": 0.19939465820789337,
+      "learning_rate": 3.6563395644087504e-08,
+      "loss": 0.3633,
       "step": 136465
     },
     {
-      "epoch": 4.8,
-      "learning_rate": 2.1011787069491195e-07,
-      "loss": 0.2405,
+      "epoch": 4.918369553465239,
+      "grad_norm": 0.21489644050598145,
+      "learning_rate": 3.6405799281230556e-08,
+      "loss": 0.373,
       "step": 136470
     },
     {
-      "epoch": 4.8,
-      "learning_rate": 2.0974946840974786e-07,
-      "loss": 0.2429,
+      "epoch": 4.918549753126464,
+      "grad_norm": 0.22894160449504852,
+      "learning_rate": 3.6248543042508023e-08,
+      "loss": 0.3401,
       "step": 136475
     },
     {
-      "epoch": 4.8,
-      "learning_rate": 2.093813880096468e-07,
-      "loss": 0.2424,
+      "epoch": 4.918729952787689,
+      "grad_norm": 0.2765687108039856,
+      "learning_rate": 3.609162693006818e-08,
+      "loss": 0.3808,
       "step": 136480
     },
     {
-      "epoch": 4.8,
-      "learning_rate": 2.0901362949939384e-07,
-      "loss": 0.2447,
+      "epoch": 4.918910152448913,
+      "grad_norm": 0.2676806151866913,
+      "learning_rate": 3.5935050946045434e-08,
+      "loss": 0.3748,
       "step": 136485
     },
     {
-      "epoch": 4.8,
-      "learning_rate": 2.086461928837602e-07,
-      "loss": 0.2492,
+      "epoch": 4.919090352110138,
+      "grad_norm": 0.1796978861093521,
+      "learning_rate": 3.5778815092576965e-08,
+      "loss": 0.364,
       "step": 136490
     },
     {
-      "epoch": 4.8,
-      "learning_rate": 2.0827907816751702e-07,
-      "loss": 0.2411,
+      "epoch": 4.919270551771363,
+      "grad_norm": 0.2699376344680786,
+      "learning_rate": 3.562291937178608e-08,
+      "loss": 0.3574,
       "step": 136495
     },
     {
-      "epoch": 4.8,
-      "learning_rate": 2.0791228535543272e-07,
-      "loss": 0.2492,
+      "epoch": 4.919450751432588,
+      "grad_norm": 0.2322850078344345,
+      "learning_rate": 3.546736378580162e-08,
+      "loss": 0.375,
       "step": 136500
     },
     {
-      "epoch": 4.8,
-      "eval_loss": 0.2484273761510849,
-      "eval_runtime": 10.5494,
-      "eval_samples_per_second": 9.479,
-      "eval_steps_per_second": 9.479,
+      "epoch": 4.919450751432588,
+      "eval_loss": 0.42881008982658386,
+      "eval_runtime": 3.537,
+      "eval_samples_per_second": 28.272,
+      "eval_steps_per_second": 7.068,
       "step": 136500
     },
     {
-      "epoch": 4.8,
-      "learning_rate": 2.075458144522674e-07,
-      "loss": 0.2473,
+      "epoch": 4.919630951093812,
+      "grad_norm": 0.266796350479126,
+      "learning_rate": 3.531214833673857e-08,
+      "loss": 0.4061,
       "step": 136505
     },
     {
-      "epoch": 4.8,
-      "learning_rate": 2.071796654627839e-07,
-      "loss": 0.2616,
+      "epoch": 4.919811150755036,
+      "grad_norm": 0.23469606041908264,
+      "learning_rate": 3.515727302671468e-08,
+      "loss": 0.3704,
       "step": 136510
     },
     {
-      "epoch": 4.8,
-      "learning_rate": 2.0681383839172842e-07,
-      "loss": 0.2791,
+      "epoch": 4.919991350416261,
+      "grad_norm": 0.27841997146606445,
+      "learning_rate": 3.500273785784214e-08,
+      "loss": 0.3688,
       "step": 136515
     },
     {
-      "epoch": 4.8,
-      "learning_rate": 2.0644833324386104e-07,
-      "loss": 0.2526,
+      "epoch": 4.920171550077486,
+      "grad_norm": 0.23662640154361725,
+      "learning_rate": 3.4848542832222056e-08,
+      "loss": 0.3902,
       "step": 136520
     },
     {
-      "epoch": 4.8,
-      "learning_rate": 2.0608315002391687e-07,
-      "loss": 0.2583,
+      "epoch": 4.92035174973871,
+      "grad_norm": 0.3038434386253357,
+      "learning_rate": 3.46946879519583e-08,
+      "loss": 0.3871,
       "step": 136525
     },
     {
-      "epoch": 4.8,
-      "learning_rate": 2.0571828873664767e-07,
-      "loss": 0.2542,
+      "epoch": 4.920531949399935,
+      "grad_norm": 0.2633965015411377,
+      "learning_rate": 3.454117321914363e-08,
+      "loss": 0.3661,
       "step": 136530
     },
     {
-      "epoch": 4.8,
-      "learning_rate": 2.0535374938678575e-07,
-      "loss": 0.2544,
+      "epoch": 4.92071214906116,
+      "grad_norm": 0.3141336441040039,
+      "learning_rate": 3.438799863587361e-08,
+      "loss": 0.3908,
       "step": 136535
     },
     {
-      "epoch": 4.8,
-      "learning_rate": 2.0498953197906344e-07,
-      "loss": 0.2393,
+      "epoch": 4.920892348722385,
+      "grad_norm": 0.21846546232700348,
+      "learning_rate": 3.423516420423545e-08,
+      "loss": 0.3788,
       "step": 136540
     },
     {
-      "epoch": 4.8,
-      "learning_rate": 2.046256365182131e-07,
-      "loss": 0.246,
+      "epoch": 4.921072548383609,
+      "grad_norm": 0.27012789249420166,
+      "learning_rate": 3.408266992630804e-08,
+      "loss": 0.3757,
       "step": 136545
     },
     {
-      "epoch": 4.8,
-      "learning_rate": 2.042620630089559e-07,
-      "loss": 0.2716,
+      "epoch": 4.921252748044834,
+      "grad_norm": 0.24852919578552246,
+      "learning_rate": 3.393051580417028e-08,
+      "loss": 0.3594,
       "step": 136550
     },
     {
-      "epoch": 4.8,
-      "learning_rate": 2.0389881145601863e-07,
-      "loss": 0.2491,
+      "epoch": 4.921432947706059,
+      "grad_norm": 0.2667607069015503,
+      "learning_rate": 3.377870183989551e-08,
+      "loss": 0.3766,
       "step": 136555
     },
     {
-      "epoch": 4.8,
-      "learning_rate": 2.0353588186411143e-07,
-      "loss": 0.2638,
+      "epoch": 4.921613147367283,
+      "grad_norm": 0.26542437076568604,
+      "learning_rate": 3.3627228035551515e-08,
+      "loss": 0.368,
       "step": 136560
     },
     {
-      "epoch": 4.8,
-      "learning_rate": 2.0317327423794995e-07,
-      "loss": 0.2367,
+      "epoch": 4.921793347028507,
+      "grad_norm": 0.3026573657989502,
+      "learning_rate": 3.3476094393203313e-08,
+      "loss": 0.3884,
       "step": 136565
     },
     {
-      "epoch": 4.8,
-      "learning_rate": 2.0281098858224157e-07,
-      "loss": 0.2575,
+      "epoch": 4.921973546689732,
+      "grad_norm": 0.26088666915893555,
+      "learning_rate": 3.3325300914910374e-08,
+      "loss": 0.3654,
       "step": 136570
     },
     {
-      "epoch": 4.81,
-      "learning_rate": 2.0244902490169081e-07,
-      "loss": 0.2421,
+      "epoch": 4.922153746350957,
+      "grad_norm": 0.22487777471542358,
+      "learning_rate": 3.31748476027266e-08,
+      "loss": 0.3848,
       "step": 136575
     },
     {
-      "epoch": 4.81,
-      "learning_rate": 2.0208738320099675e-07,
-      "loss": 0.2673,
+      "epoch": 4.9223339460121815,
+      "grad_norm": 0.2889478802680969,
+      "learning_rate": 3.302473445870313e-08,
+      "loss": 0.4139,
       "step": 136580
     },
     {
-      "epoch": 4.81,
-      "learning_rate": 2.0172606348485556e-07,
-      "loss": 0.2437,
+      "epoch": 4.922514145673406,
+      "grad_norm": 0.21475286781787872,
+      "learning_rate": 3.2874961484882785e-08,
+      "loss": 0.3401,
       "step": 136585
     },
     {
-      "epoch": 4.81,
-      "learning_rate": 2.0136506575795798e-07,
-      "loss": 0.2469,
+      "epoch": 4.922694345334631,
+      "grad_norm": 0.27408671379089355,
+      "learning_rate": 3.272552868330558e-08,
+      "loss": 0.4108,
       "step": 136590
     },
     {
-      "epoch": 4.81,
-      "learning_rate": 2.0100439002499473e-07,
-      "loss": 0.258,
+      "epoch": 4.922874544995856,
+      "grad_norm": 0.2363937646150589,
+      "learning_rate": 3.257643605601157e-08,
+      "loss": 0.3962,
       "step": 136595
     },
     {
-      "epoch": 4.81,
-      "learning_rate": 2.0064403629064533e-07,
-      "loss": 0.2409,
+      "epoch": 4.92305474465708,
+      "grad_norm": 0.22098080813884735,
+      "learning_rate": 3.2427683605026905e-08,
+      "loss": 0.3596,
       "step": 136600
     },
     {
-      "epoch": 4.81,
-      "learning_rate": 2.002840045595894e-07,
-      "loss": 0.2655,
+      "epoch": 4.923234944318304,
+      "grad_norm": 0.2587352693080902,
+      "learning_rate": 3.227927133238329e-08,
+      "loss": 0.4166,
       "step": 136605
     },
     {
-      "epoch": 4.81,
-      "learning_rate": 1.9992429483650098e-07,
-      "loss": 0.2792,
+      "epoch": 4.923415143979529,
+      "grad_norm": 0.3030674159526825,
+      "learning_rate": 3.213119924010133e-08,
+      "loss": 0.3743,
       "step": 136610
     },
     {
-      "epoch": 4.81,
-      "learning_rate": 1.9956490712605135e-07,
-      "loss": 0.2716,
+      "epoch": 4.923595343640754,
+      "grad_norm": 0.24243710935115814,
+      "learning_rate": 3.1983467330196084e-08,
+      "loss": 0.372,
       "step": 136615
     },
     {
-      "epoch": 4.81,
-      "learning_rate": 1.9920584143290898e-07,
-      "loss": 0.2667,
+      "epoch": 4.9237755433019785,
+      "grad_norm": 0.2680684030056,
+      "learning_rate": 3.1836075604685375e-08,
+      "loss": 0.4072,
       "step": 136620
     },
     {
-      "epoch": 4.81,
-      "learning_rate": 1.9884709776173404e-07,
-      "loss": 0.2351,
+      "epoch": 4.923955742963203,
+      "grad_norm": 0.2981574237346649,
+      "learning_rate": 3.1689024065570397e-08,
+      "loss": 0.3586,
       "step": 136625
     },
     {
-      "epoch": 4.81,
-      "learning_rate": 1.9848867611718393e-07,
-      "loss": 0.2529,
+      "epoch": 4.924135942624428,
+      "grad_norm": 0.24394196271896362,
+      "learning_rate": 3.1542312714860635e-08,
+      "loss": 0.3844,
       "step": 136630
     },
     {
-      "epoch": 4.81,
-      "learning_rate": 1.981305765039132e-07,
-      "loss": 0.2524,
+      "epoch": 4.924316142285653,
+      "grad_norm": 0.26692622900009155,
+      "learning_rate": 3.139594155455172e-08,
+      "loss": 0.3926,
       "step": 136635
     },
     {
-      "epoch": 4.81,
-      "learning_rate": 1.9777279892657375e-07,
-      "loss": 0.2502,
+      "epoch": 4.9244963419468775,
+      "grad_norm": 0.3153393268585205,
+      "learning_rate": 3.1249910586639286e-08,
+      "loss": 0.3689,
       "step": 136640
     },
     {
-      "epoch": 4.81,
-      "learning_rate": 1.9741534338980906e-07,
-      "loss": 0.2542,
+      "epoch": 4.924676541608102,
+      "grad_norm": 0.2018764615058899,
+      "learning_rate": 3.11042198131134e-08,
+      "loss": 0.3361,
       "step": 136645
     },
     {
-      "epoch": 4.81,
-      "learning_rate": 1.9705820989825984e-07,
-      "loss": 0.2528,
+      "epoch": 4.924856741269326,
+      "grad_norm": 0.24366877973079681,
+      "learning_rate": 3.095886923595581e-08,
+      "loss": 0.3578,
       "step": 136650
     },
     {
-      "epoch": 4.81,
-      "learning_rate": 1.967013984565641e-07,
-      "loss": 0.2565,
+      "epoch": 4.925036940930551,
+      "grad_norm": 0.21055671572685242,
+      "learning_rate": 3.081385885715105e-08,
+      "loss": 0.3606,
       "step": 136655
     },
     {
-      "epoch": 4.81,
-      "learning_rate": 1.9634490906935698e-07,
-      "loss": 0.2407,
+      "epoch": 4.9252171405917755,
+      "grad_norm": 0.25222551822662354,
+      "learning_rate": 3.06691886786753e-08,
+      "loss": 0.3939,
       "step": 136660
     },
     {
-      "epoch": 4.81,
-      "learning_rate": 1.9598874174126258e-07,
-      "loss": 0.2541,
+      "epoch": 4.925397340253,
+      "grad_norm": 0.18292470276355743,
+      "learning_rate": 3.052485870249644e-08,
+      "loss": 0.3271,
       "step": 136665
     },
     {
-      "epoch": 4.81,
-      "learning_rate": 1.9563289647690775e-07,
-      "loss": 0.2327,
+      "epoch": 4.925577539914225,
+      "grad_norm": 0.2845352590084076,
+      "learning_rate": 3.038086893057956e-08,
+      "loss": 0.366,
       "step": 136670
     },
     {
-      "epoch": 4.81,
-      "learning_rate": 1.9527737328091378e-07,
-      "loss": 0.2477,
+      "epoch": 4.92575773957545,
+      "grad_norm": 0.26779624819755554,
+      "learning_rate": 3.0237219364892544e-08,
+      "loss": 0.3505,
       "step": 136675
     },
     {
-      "epoch": 4.81,
-      "learning_rate": 1.9492217215789644e-07,
-      "loss": 0.2629,
+      "epoch": 4.9259379392366744,
+      "grad_norm": 0.24947002530097961,
+      "learning_rate": 3.009391000738659e-08,
+      "loss": 0.3692,
       "step": 136680
     },
     {
-      "epoch": 4.81,
-      "learning_rate": 1.945672931124659e-07,
-      "loss": 0.2566,
+      "epoch": 4.926118138897899,
+      "grad_norm": 0.2781367301940918,
+      "learning_rate": 2.995094086001848e-08,
+      "loss": 0.3752,
       "step": 136685
     },
     {
-      "epoch": 4.81,
-      "learning_rate": 1.9421273614923518e-07,
-      "loss": 0.261,
+      "epoch": 4.926298338559123,
+      "grad_norm": 0.24780642986297607,
+      "learning_rate": 2.980831192473388e-08,
+      "loss": 0.3734,
       "step": 136690
     },
     {
-      "epoch": 4.81,
-      "learning_rate": 1.938585012728006e-07,
-      "loss": 0.2365,
+      "epoch": 4.926478538220348,
+      "grad_norm": 0.2541256248950958,
+      "learning_rate": 2.9666023203475667e-08,
+      "loss": 0.3417,
       "step": 136695
     },
     {
-      "epoch": 4.81,
-      "learning_rate": 1.9350458848776675e-07,
-      "loss": 0.2402,
+      "epoch": 4.9266587378815725,
+      "grad_norm": 0.2498217672109604,
+      "learning_rate": 2.9524074698186743e-08,
+      "loss": 0.3829,
       "step": 136700
     },
     {
-      "epoch": 4.81,
-      "learning_rate": 1.9315099779872726e-07,
-      "loss": 0.2434,
+      "epoch": 4.926838937542797,
+      "grad_norm": 0.2726345360279083,
+      "learning_rate": 2.938246641079334e-08,
+      "loss": 0.3742,
       "step": 136705
     },
     {
-      "epoch": 4.81,
-      "learning_rate": 1.927977292102756e-07,
-      "loss": 0.2461,
+      "epoch": 4.927019137204022,
+      "grad_norm": 0.29105934500694275,
+      "learning_rate": 2.9241198343232802e-08,
+      "loss": 0.3852,
       "step": 136710
     },
     {
-      "epoch": 4.81,
-      "learning_rate": 1.9244478272699705e-07,
-      "loss": 0.2583,
+      "epoch": 4.927199336865247,
+      "grad_norm": 0.24921295046806335,
+      "learning_rate": 2.910027049742581e-08,
+      "loss": 0.3763,
       "step": 136715
     },
     {
-      "epoch": 4.81,
-      "learning_rate": 1.9209215835347404e-07,
-      "loss": 0.2627,
+      "epoch": 4.927379536526471,
+      "grad_norm": 0.3070908784866333,
+      "learning_rate": 2.8959682875293047e-08,
+      "loss": 0.3878,
       "step": 136720
     },
     {
-      "epoch": 4.81,
-      "learning_rate": 1.9173985609428348e-07,
-      "loss": 0.2547,
+      "epoch": 4.927559736187696,
+      "grad_norm": 0.319961279630661,
+      "learning_rate": 2.8819435478749657e-08,
+      "loss": 0.4065,
       "step": 136725
     },
     {
-      "epoch": 4.81,
-      "learning_rate": 1.913878759540022e-07,
-      "loss": 0.253,
+      "epoch": 4.927739935848921,
+      "grad_norm": 0.3225443959236145,
+      "learning_rate": 2.8679528309705217e-08,
+      "loss": 0.4024,
       "step": 136730
     },
     {
-      "epoch": 4.81,
-      "learning_rate": 1.9103621793719885e-07,
-      "loss": 0.246,
+      "epoch": 4.927920135510146,
+      "grad_norm": 0.2788389325141907,
+      "learning_rate": 2.8539961370069314e-08,
+      "loss": 0.3578,
       "step": 136735
     },
     {
-      "epoch": 4.81,
-      "learning_rate": 1.9068488204844193e-07,
-      "loss": 0.256,
+      "epoch": 4.92810033517137,
+      "grad_norm": 0.2546929717063904,
+      "learning_rate": 2.84007346617432e-08,
+      "loss": 0.3585,
       "step": 136740
     },
     {
-      "epoch": 4.81,
-      "learning_rate": 1.903338682922917e-07,
-      "loss": 0.2588,
+      "epoch": 4.928280534832594,
+      "grad_norm": 0.23796042799949646,
+      "learning_rate": 2.826184818661981e-08,
+      "loss": 0.3595,
       "step": 136745
     },
     {
-      "epoch": 4.81,
-      "learning_rate": 1.899831766733057e-07,
-      "loss": 0.2472,
+      "epoch": 4.928460734493819,
+      "grad_norm": 0.22895626723766327,
+      "learning_rate": 2.8123301946594847e-08,
+      "loss": 0.3904,
       "step": 136750
     },
     {
-      "epoch": 4.81,
-      "learning_rate": 1.896328071960385e-07,
-      "loss": 0.265,
+      "epoch": 4.928640934155044,
+      "grad_norm": 0.2509956657886505,
+      "learning_rate": 2.7985095943555696e-08,
+      "loss": 0.4023,
       "step": 136755
     },
     {
-      "epoch": 4.81,
-      "learning_rate": 1.892827598650393e-07,
-      "loss": 0.2555,
+      "epoch": 4.928821133816268,
+      "grad_norm": 0.34779104590415955,
+      "learning_rate": 2.7847230179384176e-08,
+      "loss": 0.3417,
       "step": 136760
     },
     {
-      "epoch": 4.81,
-      "learning_rate": 1.889330346848489e-07,
-      "loss": 0.2616,
+      "epoch": 4.929001333477493,
+      "grad_norm": 0.19889460504055023,
+      "learning_rate": 2.7709704655959344e-08,
+      "loss": 0.342,
       "step": 136765
     },
     {
-      "epoch": 4.81,
-      "learning_rate": 1.8858363166001646e-07,
-      "loss": 0.2351,
+      "epoch": 4.929181533138718,
+      "grad_norm": 0.2757225036621094,
+      "learning_rate": 2.757251937515748e-08,
+      "loss": 0.3928,
       "step": 136770
     },
     {
-      "epoch": 4.81,
-      "learning_rate": 1.8823455079507168e-07,
-      "loss": 0.2585,
+      "epoch": 4.929361732799943,
+      "grad_norm": 0.32304972410202026,
+      "learning_rate": 2.7435674338843752e-08,
+      "loss": 0.379,
       "step": 136775
     },
     {
-      "epoch": 4.81,
-      "learning_rate": 1.8788579209454982e-07,
-      "loss": 0.2486,
+      "epoch": 4.929541932461167,
+      "grad_norm": 0.21971245110034943,
+      "learning_rate": 2.729916954888334e-08,
+      "loss": 0.3514,
       "step": 136780
     },
     {
-      "epoch": 4.81,
-      "learning_rate": 1.8753735556298057e-07,
-      "loss": 0.2458,
+      "epoch": 4.929722132122391,
+      "grad_norm": 0.2933555841445923,
+      "learning_rate": 2.7163005007135867e-08,
+      "loss": 0.3583,
       "step": 136785
     },
     {
-      "epoch": 4.81,
-      "learning_rate": 1.8718924120488536e-07,
-      "loss": 0.2521,
+      "epoch": 4.929902331783616,
+      "grad_norm": 0.25747615098953247,
+      "learning_rate": 2.7027180715458177e-08,
+      "loss": 0.3979,
       "step": 136790
     },
     {
-      "epoch": 4.81,
-      "learning_rate": 1.8684144902478552e-07,
-      "loss": 0.2684,
+      "epoch": 4.930082531444841,
+      "grad_norm": 0.27800336480140686,
+      "learning_rate": 2.689169667570157e-08,
+      "loss": 0.3521,
       "step": 136795
     },
     {
-      "epoch": 4.81,
-      "learning_rate": 1.864939790271969e-07,
-      "loss": 0.2521,
+      "epoch": 4.930262731106065,
+      "grad_norm": 0.24476049840450287,
+      "learning_rate": 2.675655288970902e-08,
+      "loss": 0.3589,
       "step": 136800
     },
     {
-      "epoch": 4.81,
-      "learning_rate": 1.8614683121663257e-07,
-      "loss": 0.2539,
+      "epoch": 4.93044293076729,
+      "grad_norm": 0.21203723549842834,
+      "learning_rate": 2.6621749359326263e-08,
+      "loss": 0.3473,
       "step": 136805
     },
     {
-      "epoch": 4.81,
-      "learning_rate": 1.8580000559759726e-07,
-      "loss": 0.2546,
+      "epoch": 4.930623130428515,
+      "grad_norm": 0.25764790177345276,
+      "learning_rate": 2.6487286086385177e-08,
+      "loss": 0.3886,
       "step": 136810
     },
     {
-      "epoch": 4.81,
-      "learning_rate": 1.8545350217459846e-07,
-      "loss": 0.2537,
+      "epoch": 4.93080333008974,
+      "grad_norm": 0.2776869535446167,
+      "learning_rate": 2.63531630727204e-08,
+      "loss": 0.4101,
       "step": 136815
     },
     {
-      "epoch": 4.81,
-      "learning_rate": 1.851073209521298e-07,
-      "loss": 0.2552,
+      "epoch": 4.930983529750964,
+      "grad_norm": 0.23723161220550537,
+      "learning_rate": 2.6219380320158248e-08,
+      "loss": 0.3742,
       "step": 136820
     },
     {
-      "epoch": 4.81,
-      "learning_rate": 1.8476146193468768e-07,
-      "loss": 0.242,
+      "epoch": 4.931163729412189,
+      "grad_norm": 0.28152310848236084,
+      "learning_rate": 2.6085937830522266e-08,
+      "loss": 0.3533,
       "step": 136825
     },
     {
-      "epoch": 4.81,
-      "learning_rate": 1.8441592512676853e-07,
-      "loss": 0.2568,
+      "epoch": 4.931343929073414,
+      "grad_norm": 0.3157213032245636,
+      "learning_rate": 2.595283560562767e-08,
+      "loss": 0.3813,
       "step": 136830
     },
     {
-      "epoch": 4.81,
-      "learning_rate": 1.8407071053285208e-07,
-      "loss": 0.2641,
+      "epoch": 4.931524128734638,
+      "grad_norm": 0.22716785967350006,
+      "learning_rate": 2.582007364729522e-08,
+      "loss": 0.3488,
       "step": 136835
     },
     {
-      "epoch": 4.81,
-      "learning_rate": 1.8372581815742085e-07,
-      "loss": 0.2589,
+      "epoch": 4.931704328395862,
+      "grad_norm": 0.27288514375686646,
+      "learning_rate": 2.568765195732625e-08,
+      "loss": 0.3763,
       "step": 136840
     },
     {
-      "epoch": 4.81,
-      "learning_rate": 1.8338124800496014e-07,
-      "loss": 0.2387,
+      "epoch": 4.931884528057087,
+      "grad_norm": 0.22473356127738953,
+      "learning_rate": 2.5555570537527663e-08,
+      "loss": 0.3811,
       "step": 136845
     },
     {
-      "epoch": 4.81,
-      "learning_rate": 1.8303700007993585e-07,
-      "loss": 0.2526,
+      "epoch": 4.932064727718312,
+      "grad_norm": 0.25203046202659607,
+      "learning_rate": 2.5423829389700783e-08,
+      "loss": 0.3522,
       "step": 136850
     },
     {
-      "epoch": 4.81,
-      "learning_rate": 1.8269307438682214e-07,
-      "loss": 0.2391,
+      "epoch": 4.932244927379537,
+      "grad_norm": 0.24768032133579254,
+      "learning_rate": 2.5292428515638622e-08,
+      "loss": 0.3684,
       "step": 136855
     },
     {
-      "epoch": 4.82,
-      "learning_rate": 1.8234947093008214e-07,
-      "loss": 0.2204,
+      "epoch": 4.932425127040761,
+      "grad_norm": 0.23223379254341125,
+      "learning_rate": 2.5161367917131416e-08,
+      "loss": 0.3809,
       "step": 136860
     },
     {
-      "epoch": 4.82,
-      "learning_rate": 1.8200618971417892e-07,
-      "loss": 0.2157,
+      "epoch": 4.932605326701986,
+      "grad_norm": 0.24357058107852936,
+      "learning_rate": 2.5030647595963853e-08,
+      "loss": 0.3799,
       "step": 136865
     },
     {
-      "epoch": 4.82,
-      "learning_rate": 1.8166323074357284e-07,
-      "loss": 0.2753,
+      "epoch": 4.932785526363211,
+      "grad_norm": 0.3189919590950012,
+      "learning_rate": 2.4900267553920608e-08,
+      "loss": 0.3666,
       "step": 136870
     },
     {
-      "epoch": 4.82,
-      "learning_rate": 1.813205940227103e-07,
-      "loss": 0.2546,
+      "epoch": 4.932965726024435,
+      "grad_norm": 0.2528684437274933,
+      "learning_rate": 2.4770227792775268e-08,
+      "loss": 0.3559,
       "step": 136875
     },
     {
-      "epoch": 4.82,
-      "learning_rate": 1.8097827955604608e-07,
-      "loss": 0.2641,
+      "epoch": 4.933145925685659,
+      "grad_norm": 0.2626766562461853,
+      "learning_rate": 2.464052831429864e-08,
+      "loss": 0.4019,
       "step": 136880
     },
     {
-      "epoch": 4.82,
-      "learning_rate": 1.8063628734802107e-07,
-      "loss": 0.2514,
+      "epoch": 4.933326125346884,
+      "grad_norm": 0.2691434323787689,
+      "learning_rate": 2.4511169120261524e-08,
+      "loss": 0.3848,
       "step": 136885
     },
     {
-      "epoch": 4.82,
-      "learning_rate": 1.8036292561080804e-07,
-      "loss": 0.222,
+      "epoch": 4.933506325008109,
+      "grad_norm": 0.26981598138809204,
+      "learning_rate": 2.4382150212423625e-08,
+      "loss": 0.3692,
       "step": 136890
     },
     {
-      "epoch": 4.82,
-      "learning_rate": 1.8002151347952245e-07,
-      "loss": 0.2635,
+      "epoch": 4.9336865246693336,
+      "grad_norm": 0.2676762342453003,
+      "learning_rate": 2.425347159254465e-08,
+      "loss": 0.3717,
       "step": 136895
     },
     {
-      "epoch": 4.82,
-      "learning_rate": 1.7968042361930227e-07,
-      "loss": 0.2581,
+      "epoch": 4.933866724330558,
+      "grad_norm": 0.2530960738658905,
+      "learning_rate": 2.4125133262373202e-08,
+      "loss": 0.3849,
       "step": 136900
     },
     {
-      "epoch": 4.82,
-      "learning_rate": 1.7933965603457447e-07,
-      "loss": 0.2634,
+      "epoch": 4.934046923991783,
+      "grad_norm": 0.2379455268383026,
+      "learning_rate": 2.3997135223663424e-08,
+      "loss": 0.3896,
       "step": 136905
     },
     {
-      "epoch": 4.82,
-      "learning_rate": 1.7899921072976332e-07,
-      "loss": 0.2852,
+      "epoch": 4.934227123653008,
+      "grad_norm": 0.29824960231781006,
+      "learning_rate": 2.3869477478158374e-08,
+      "loss": 0.3574,
       "step": 136910
     },
     {
-      "epoch": 4.82,
-      "learning_rate": 1.7865908770929306e-07,
-      "loss": 0.2385,
+      "epoch": 4.9344073233142325,
+      "grad_norm": 0.21034584939479828,
+      "learning_rate": 2.374216002759555e-08,
+      "loss": 0.3759,
       "step": 136915
     },
     {
-      "epoch": 4.82,
-      "learning_rate": 1.7831928697757404e-07,
-      "loss": 0.2633,
+      "epoch": 4.934587522975457,
+      "grad_norm": 0.28804388642311096,
+      "learning_rate": 2.3615182873709673e-08,
+      "loss": 0.386,
       "step": 136920
     },
     {
-      "epoch": 4.82,
-      "learning_rate": 1.7797980853901942e-07,
-      "loss": 0.2316,
+      "epoch": 4.934767722636681,
+      "grad_norm": 0.25442805886268616,
+      "learning_rate": 2.3488546018232692e-08,
+      "loss": 0.39,
       "step": 136925
     },
     {
-      "epoch": 4.82,
-      "learning_rate": 1.776406523980395e-07,
-      "loss": 0.2528,
+      "epoch": 4.934947922297906,
+      "grad_norm": 0.2353426218032837,
+      "learning_rate": 2.3362249462885454e-08,
+      "loss": 0.3843,
       "step": 136930
     },
     {
-      "epoch": 4.82,
-      "learning_rate": 1.7730181855903916e-07,
-      "loss": 0.264,
+      "epoch": 4.9351281219591305,
+      "grad_norm": 0.22822360694408417,
+      "learning_rate": 2.3236293209394354e-08,
+      "loss": 0.3774,
       "step": 136935
     },
     {
-      "epoch": 4.82,
-      "learning_rate": 1.7696330702641482e-07,
-      "loss": 0.2626,
+      "epoch": 4.935308321620355,
+      "grad_norm": 0.21662981808185577,
+      "learning_rate": 2.311067725947469e-08,
+      "loss": 0.3724,
       "step": 136940
     },
     {
-      "epoch": 4.82,
-      "learning_rate": 1.7662511780456304e-07,
-      "loss": 0.2429,
+      "epoch": 4.93548852128158,
+      "grad_norm": 0.2427043616771698,
+      "learning_rate": 2.2985401614833422e-08,
+      "loss": 0.3599,
       "step": 136945
     },
     {
-      "epoch": 4.82,
-      "learning_rate": 1.762872508978719e-07,
-      "loss": 0.256,
+      "epoch": 4.935668720942805,
+      "grad_norm": 0.2436215728521347,
+      "learning_rate": 2.2860466277180303e-08,
+      "loss": 0.3463,
       "step": 136950
     },
     {
-      "epoch": 4.82,
-      "learning_rate": 1.7594970631073515e-07,
-      "loss": 0.2664,
+      "epoch": 4.9358489206040295,
+      "grad_norm": 0.18670439720153809,
+      "learning_rate": 2.273587124821952e-08,
+      "loss": 0.3419,
       "step": 136955
     },
     {
-      "epoch": 4.82,
-      "learning_rate": 1.756124840475326e-07,
-      "loss": 0.238,
+      "epoch": 4.936029120265254,
+      "grad_norm": 0.23611606657505035,
+      "learning_rate": 2.2611616529644164e-08,
+      "loss": 0.374,
       "step": 136960
     },
     {
-      "epoch": 4.82,
-      "learning_rate": 1.7527558411263856e-07,
-      "loss": 0.232,
+      "epoch": 4.936209319926478,
+      "grad_norm": 0.27556440234184265,
+      "learning_rate": 2.2487702123152877e-08,
+      "loss": 0.3414,
       "step": 136965
     },
     {
-      "epoch": 4.82,
-      "learning_rate": 1.7493900651043282e-07,
-      "loss": 0.2588,
+      "epoch": 4.936389519587703,
+      "grad_norm": 0.2503463625907898,
+      "learning_rate": 2.2364128030430422e-08,
+      "loss": 0.3384,
       "step": 136970
     },
     {
-      "epoch": 4.82,
-      "learning_rate": 1.7460275124528413e-07,
-      "loss": 0.2515,
+      "epoch": 4.9365697192489275,
+      "grad_norm": 0.2716796398162842,
+      "learning_rate": 2.2240894253158785e-08,
+      "loss": 0.3669,
       "step": 136975
     },
     {
-      "epoch": 4.82,
-      "learning_rate": 1.7426681832155844e-07,
-      "loss": 0.2641,
+      "epoch": 4.936749918910152,
+      "grad_norm": 0.2600260376930237,
+      "learning_rate": 2.2118000793019956e-08,
+      "loss": 0.4075,
       "step": 136980
     },
     {
-      "epoch": 4.82,
-      "learning_rate": 1.7393120774361338e-07,
-      "loss": 0.2433,
+      "epoch": 4.936930118571377,
+      "grad_norm": 0.20133672654628754,
+      "learning_rate": 2.1995447651687595e-08,
+      "loss": 0.3881,
       "step": 136985
     },
     {
-      "epoch": 4.82,
-      "learning_rate": 1.7359591951581489e-07,
-      "loss": 0.259,
+      "epoch": 4.937110318232602,
+      "grad_norm": 0.268351674079895,
+      "learning_rate": 2.187323483083259e-08,
+      "loss": 0.355,
       "step": 136990
     },
     {
-      "epoch": 4.82,
-      "learning_rate": 1.7326095364251226e-07,
-      "loss": 0.2382,
+      "epoch": 4.9372905178938264,
+      "grad_norm": 0.24614043533802032,
+      "learning_rate": 2.1751362332117496e-08,
+      "loss": 0.3716,
       "step": 136995
     },
     {
-      "epoch": 4.82,
-      "learning_rate": 1.7292631012805204e-07,
-      "loss": 0.2548,
+      "epoch": 4.937470717555051,
+      "grad_norm": 0.24108512699604034,
+      "learning_rate": 2.162983015720488e-08,
+      "loss": 0.3567,
       "step": 137000
     },
     {
-      "epoch": 4.82,
-      "eval_loss": 0.24844510853290558,
-      "eval_runtime": 10.5423,
-      "eval_samples_per_second": 9.486,
-      "eval_steps_per_second": 9.486,
+      "epoch": 4.937470717555051,
+      "eval_loss": 0.4288177788257599,
+      "eval_runtime": 3.5288,
+      "eval_samples_per_second": 28.339,
+      "eval_steps_per_second": 7.085,
       "step": 137000
     },
     {
-      "epoch": 4.82,
-      "learning_rate": 1.7259198897678352e-07,
-      "loss": 0.2178,
+      "epoch": 4.937650917216276,
+      "grad_norm": 0.2266797125339508,
+      "learning_rate": 2.150863830774896e-08,
+      "loss": 0.3821,
       "step": 137005
     },
     {
-      "epoch": 4.82,
-      "learning_rate": 1.722579901930449e-07,
-      "loss": 0.2575,
+      "epoch": 4.937831116877501,
+      "grad_norm": 0.2394786775112152,
+      "learning_rate": 2.138778678540121e-08,
+      "loss": 0.3445,
       "step": 137010
     },
     {
-      "epoch": 4.82,
-      "learning_rate": 1.7192431378117714e-07,
-      "loss": 0.2518,
+      "epoch": 4.938011316538725,
+      "grad_norm": 0.2852950394153595,
+      "learning_rate": 2.12672755918103e-08,
+      "loss": 0.3695,
       "step": 137015
     },
     {
-      "epoch": 4.82,
-      "learning_rate": 1.7159095974550731e-07,
-      "loss": 0.2554,
+      "epoch": 4.938191516199949,
+      "grad_norm": 0.270437628030777,
+      "learning_rate": 2.1147104728616584e-08,
+      "loss": 0.3827,
       "step": 137020
     },
     {
-      "epoch": 4.82,
-      "learning_rate": 1.7125792809036533e-07,
-      "loss": 0.2502,
+      "epoch": 4.938371715861174,
+      "grad_norm": 0.26983991265296936,
+      "learning_rate": 2.1027274197457646e-08,
+      "loss": 0.3773,
       "step": 137025
     },
     {
-      "epoch": 4.82,
-      "learning_rate": 1.7092521882007827e-07,
-      "loss": 0.2334,
+      "epoch": 4.938551915522399,
+      "grad_norm": 0.2832012474536896,
+      "learning_rate": 2.0907783999965515e-08,
+      "loss": 0.3651,
       "step": 137030
     },
     {
-      "epoch": 4.82,
-      "learning_rate": 1.7059283193896492e-07,
-      "loss": 0.2553,
+      "epoch": 4.938732115183623,
+      "grad_norm": 0.28883448243141174,
+      "learning_rate": 2.0788634137769437e-08,
+      "loss": 0.3767,
       "step": 137035
     },
     {
-      "epoch": 4.82,
-      "learning_rate": 1.7026076745134123e-07,
-      "loss": 0.2533,
+      "epoch": 4.938912314844848,
+      "grad_norm": 0.29116883873939514,
+      "learning_rate": 2.066982461249034e-08,
+      "loss": 0.3505,
       "step": 137040
     },
     {
-      "epoch": 4.82,
-      "learning_rate": 1.699290253615149e-07,
-      "loss": 0.2623,
+      "epoch": 4.939092514506073,
+      "grad_norm": 0.2609765827655792,
+      "learning_rate": 2.0551355425749154e-08,
+      "loss": 0.3741,
       "step": 137045
     },
     {
-      "epoch": 4.82,
-      "learning_rate": 1.6959760567379912e-07,
-      "loss": 0.2666,
+      "epoch": 4.939272714167298,
+      "grad_norm": 0.23762352764606476,
+      "learning_rate": 2.0433226579161247e-08,
+      "loss": 0.402,
       "step": 137050
     },
     {
-      "epoch": 4.82,
-      "learning_rate": 1.6926650839249325e-07,
-      "loss": 0.245,
+      "epoch": 4.939452913828522,
+      "grad_norm": 0.23419687151908875,
+      "learning_rate": 2.031543807433367e-08,
+      "loss": 0.3582,
       "step": 137055
     },
     {
-      "epoch": 4.82,
-      "learning_rate": 1.6893573352189663e-07,
-      "loss": 0.2505,
+      "epoch": 4.939633113489746,
+      "grad_norm": 0.23861204087734222,
+      "learning_rate": 2.0197989912873473e-08,
+      "loss": 0.3914,
       "step": 137060
     },
     {
-      "epoch": 4.82,
-      "learning_rate": 1.686052810663058e-07,
-      "loss": 0.2386,
+      "epoch": 4.939813313150971,
+      "grad_norm": 0.2407989501953125,
+      "learning_rate": 2.0080882096376595e-08,
+      "loss": 0.3529,
       "step": 137065
     },
     {
-      "epoch": 4.82,
-      "learning_rate": 1.682751510300118e-07,
-      "loss": 0.2312,
+      "epoch": 4.939993512812196,
+      "grad_norm": 0.2740454375743866,
+      "learning_rate": 1.9964114626441764e-08,
+      "loss": 0.3948,
       "step": 137070
     },
     {
-      "epoch": 4.82,
-      "learning_rate": 1.6794534341730006e-07,
-      "loss": 0.2437,
+      "epoch": 4.94017371247342,
+      "grad_norm": 0.28621092438697815,
+      "learning_rate": 1.984768750466215e-08,
+      "loss": 0.3807,
       "step": 137075
     },
     {
-      "epoch": 4.82,
-      "learning_rate": 1.676158582324533e-07,
-      "loss": 0.2555,
+      "epoch": 4.940353912134645,
+      "grad_norm": 0.26705020666122437,
+      "learning_rate": 1.9731600732619816e-08,
+      "loss": 0.4056,
       "step": 137080
     },
     {
-      "epoch": 4.82,
-      "learning_rate": 1.672866954797514e-07,
-      "loss": 0.2692,
+      "epoch": 4.94053411179587,
+      "grad_norm": 0.27954116463661194,
+      "learning_rate": 1.961585431189683e-08,
+      "loss": 0.3494,
       "step": 137085
     },
     {
-      "epoch": 4.82,
-      "learning_rate": 1.6695785516346318e-07,
-      "loss": 0.2638,
+      "epoch": 4.940714311457095,
+      "grad_norm": 0.2583678662776947,
+      "learning_rate": 1.9500448244072487e-08,
+      "loss": 0.4074,
       "step": 137090
     },
     {
-      "epoch": 4.82,
-      "learning_rate": 1.6662933728786302e-07,
-      "loss": 0.2389,
+      "epoch": 4.940894511118319,
+      "grad_norm": 0.27501171827316284,
+      "learning_rate": 1.9385382530717755e-08,
+      "loss": 0.3644,
       "step": 137095
     },
     {
-      "epoch": 4.82,
-      "learning_rate": 1.6630114185721413e-07,
-      "loss": 0.2585,
+      "epoch": 4.941074710779544,
+      "grad_norm": 0.2305932641029358,
+      "learning_rate": 1.9270657173403593e-08,
+      "loss": 0.3497,
       "step": 137100
     },
     {
-      "epoch": 4.82,
-      "learning_rate": 1.659732688757798e-07,
-      "loss": 0.2204,
+      "epoch": 4.941254910440769,
+      "grad_norm": 0.23129644989967346,
+      "learning_rate": 1.9156272173687094e-08,
+      "loss": 0.3861,
       "step": 137105
     },
     {
-      "epoch": 4.82,
-      "learning_rate": 1.6564571834781495e-07,
-      "loss": 0.2584,
+      "epoch": 4.941435110101993,
+      "grad_norm": 0.295730322599411,
+      "learning_rate": 1.9042227533130897e-08,
+      "loss": 0.3406,
       "step": 137110
     },
     {
-      "epoch": 4.82,
-      "learning_rate": 1.6531849027757728e-07,
-      "loss": 0.2436,
+      "epoch": 4.941615309763217,
+      "grad_norm": 0.2581028938293457,
+      "learning_rate": 1.8928523253286536e-08,
+      "loss": 0.3465,
       "step": 137115
     },
     {
-      "epoch": 4.82,
-      "learning_rate": 1.6499158466930786e-07,
-      "loss": 0.2404,
+      "epoch": 4.941795509424442,
+      "grad_norm": 0.21017776429653168,
+      "learning_rate": 1.881515933570832e-08,
+      "loss": 0.3547,
       "step": 137120
     },
     {
-      "epoch": 4.82,
-      "learning_rate": 1.6466500152725882e-07,
-      "loss": 0.225,
+      "epoch": 4.941975709085667,
+      "grad_norm": 0.24874837696552277,
+      "learning_rate": 1.8702135781933917e-08,
+      "loss": 0.3988,
       "step": 137125
     },
     {
-      "epoch": 4.82,
-      "learning_rate": 1.6433874085566848e-07,
-      "loss": 0.2519,
+      "epoch": 4.942155908746892,
+      "grad_norm": 0.2943408191204071,
+      "learning_rate": 1.858945259350653e-08,
+      "loss": 0.3855,
       "step": 137130
     },
     {
-      "epoch": 4.82,
-      "learning_rate": 1.640128026587695e-07,
-      "loss": 0.2528,
+      "epoch": 4.942336108408116,
+      "grad_norm": 0.27087777853012085,
+      "learning_rate": 1.8477109771963817e-08,
+      "loss": 0.381,
       "step": 137135
     },
     {
-      "epoch": 4.82,
-      "learning_rate": 1.6368718694079744e-07,
-      "loss": 0.2454,
+      "epoch": 4.942516308069341,
+      "grad_norm": 0.2375209629535675,
+      "learning_rate": 1.8365107318829568e-08,
+      "loss": 0.3586,
       "step": 137140
     },
     {
-      "epoch": 4.83,
-      "learning_rate": 1.633618937059822e-07,
-      "loss": 0.2469,
+      "epoch": 4.942696507730566,
+      "grad_norm": 0.2501397430896759,
+      "learning_rate": 1.8253445235638654e-08,
+      "loss": 0.3695,
       "step": 137145
     },
     {
-      "epoch": 4.83,
-      "learning_rate": 1.6303692295854268e-07,
-      "loss": 0.2677,
+      "epoch": 4.94287670739179,
+      "grad_norm": 0.2996070981025696,
+      "learning_rate": 1.8142123523903765e-08,
+      "loss": 0.4018,
       "step": 137150
     },
     {
-      "epoch": 4.83,
-      "learning_rate": 1.6271227470270046e-07,
-      "loss": 0.2444,
+      "epoch": 4.943056907053014,
+      "grad_norm": 0.3034278154373169,
+      "learning_rate": 1.8031142185148674e-08,
+      "loss": 0.3636,
       "step": 137155
     },
     {
-      "epoch": 4.83,
-      "learning_rate": 1.6238794894267163e-07,
-      "loss": 0.2543,
+      "epoch": 4.943237106714239,
+      "grad_norm": 0.2226463109254837,
+      "learning_rate": 1.792050122088329e-08,
+      "loss": 0.3754,
       "step": 137160
     },
     {
-      "epoch": 4.83,
-      "learning_rate": 1.6206394568266393e-07,
-      "loss": 0.2486,
+      "epoch": 4.943417306375464,
+      "grad_norm": 0.24234063923358917,
+      "learning_rate": 1.7810200632611963e-08,
+      "loss": 0.3599,
       "step": 137165
     },
     {
-      "epoch": 4.83,
-      "learning_rate": 1.6174026492689065e-07,
-      "loss": 0.2558,
+      "epoch": 4.943597506036689,
+      "grad_norm": 0.25728529691696167,
+      "learning_rate": 1.7700240421841817e-08,
+      "loss": 0.3817,
       "step": 137170
     },
     {
-      "epoch": 4.83,
-      "learning_rate": 1.6141690667954846e-07,
-      "loss": 0.2419,
+      "epoch": 4.943777705697913,
+      "grad_norm": 0.22455373406410217,
+      "learning_rate": 1.759062059006611e-08,
+      "loss": 0.3744,
       "step": 137175
     },
     {
-      "epoch": 4.83,
-      "learning_rate": 1.6109387094483953e-07,
-      "loss": 0.259,
+      "epoch": 4.943957905359138,
+      "grad_norm": 0.2268945425748825,
+      "learning_rate": 1.7481341138783634e-08,
+      "loss": 0.3862,
       "step": 137180
     },
     {
-      "epoch": 4.83,
-      "learning_rate": 1.6077115772695773e-07,
-      "loss": 0.247,
+      "epoch": 4.944138105020363,
+      "grad_norm": 0.27768200635910034,
+      "learning_rate": 1.7372402069482097e-08,
+      "loss": 0.3813,
       "step": 137185
     },
     {
-      "epoch": 4.83,
-      "learning_rate": 1.6044876703009414e-07,
-      "loss": 0.2441,
+      "epoch": 4.9443183046815875,
+      "grad_norm": 0.256106436252594,
+      "learning_rate": 1.726380338364364e-08,
+      "loss": 0.3873,
       "step": 137190
     },
     {
-      "epoch": 4.83,
-      "learning_rate": 1.6012669885843158e-07,
-      "loss": 0.2417,
+      "epoch": 4.944498504342812,
+      "grad_norm": 0.21937619149684906,
+      "learning_rate": 1.7155545082750413e-08,
+      "loss": 0.3762,
       "step": 137195
     },
     {
-      "epoch": 4.83,
-      "learning_rate": 1.5980495321615273e-07,
-      "loss": 0.2485,
+      "epoch": 4.944678704004036,
+      "grad_norm": 0.2467588484287262,
+      "learning_rate": 1.704762716827346e-08,
+      "loss": 0.3593,
       "step": 137200
     },
     {
-      "epoch": 4.83,
-      "learning_rate": 1.5948353010743765e-07,
-      "loss": 0.2444,
+      "epoch": 4.944858903665261,
+      "grad_norm": 0.282312273979187,
+      "learning_rate": 1.6940049641686605e-08,
+      "loss": 0.3672,
       "step": 137205
     },
     {
-      "epoch": 4.83,
-      "learning_rate": 1.59162429536458e-07,
-      "loss": 0.2523,
+      "epoch": 4.9450391033264856,
+      "grad_norm": 0.2458559274673462,
+      "learning_rate": 1.683281250445534e-08,
+      "loss": 0.3546,
       "step": 137210
     },
     {
-      "epoch": 4.83,
-      "learning_rate": 1.5884165150738538e-07,
-      "loss": 0.2692,
+      "epoch": 4.94521930298771,
+      "grad_norm": 0.2797468602657318,
+      "learning_rate": 1.6725915758039612e-08,
+      "loss": 0.3572,
       "step": 137215
     },
     {
-      "epoch": 4.83,
-      "learning_rate": 1.5852119602438043e-07,
-      "loss": 0.2524,
+      "epoch": 4.945399502648935,
+      "grad_norm": 0.2865085005760193,
+      "learning_rate": 1.661935940389936e-08,
+      "loss": 0.3809,
       "step": 137220
     },
     {
-      "epoch": 4.83,
-      "learning_rate": 1.5820106309160642e-07,
-      "loss": 0.2647,
+      "epoch": 4.94557970231016,
+      "grad_norm": 0.2759266495704651,
+      "learning_rate": 1.6513143443477873e-08,
+      "loss": 0.3687,
       "step": 137225
     },
     {
-      "epoch": 4.83,
-      "learning_rate": 1.578812527132212e-07,
-      "loss": 0.247,
+      "epoch": 4.9457599019713845,
+      "grad_norm": 0.2847881615161896,
+      "learning_rate": 1.6407267878232326e-08,
+      "loss": 0.3586,
       "step": 137230
     },
     {
-      "epoch": 4.83,
-      "learning_rate": 1.5756176489337693e-07,
-      "loss": 0.2237,
+      "epoch": 4.945940101632609,
+      "grad_norm": 0.2958735227584839,
+      "learning_rate": 1.630173270960045e-08,
+      "loss": 0.3507,
       "step": 137235
     },
     {
-      "epoch": 4.83,
-      "learning_rate": 1.5724259963622035e-07,
-      "loss": 0.2317,
+      "epoch": 4.946120301293834,
+      "grad_norm": 0.29240357875823975,
+      "learning_rate": 1.619653793901721e-08,
+      "loss": 0.3602,
       "step": 137240
     },
     {
-      "epoch": 4.83,
-      "learning_rate": 1.5692375694589813e-07,
-      "loss": 0.2366,
+      "epoch": 4.946300500955058,
+      "grad_norm": 0.25079208612442017,
+      "learning_rate": 1.609168356792312e-08,
+      "loss": 0.3342,
       "step": 137245
     },
     {
-      "epoch": 4.83,
-      "learning_rate": 1.566052368265486e-07,
-      "loss": 0.261,
+      "epoch": 4.9464807006162825,
+      "grad_norm": 0.2954399287700653,
+      "learning_rate": 1.598716959773927e-08,
+      "loss": 0.3812,
       "step": 137250
     },
     {
-      "epoch": 4.83,
-      "learning_rate": 1.562870392823046e-07,
-      "loss": 0.2254,
+      "epoch": 4.946660900277507,
+      "grad_norm": 0.28976091742515564,
+      "learning_rate": 1.588299602989507e-08,
+      "loss": 0.3823,
       "step": 137255
     },
     {
-      "epoch": 4.83,
-      "learning_rate": 1.5596916431730447e-07,
-      "loss": 0.2487,
+      "epoch": 4.946841099938732,
+      "grad_norm": 0.328617125749588,
+      "learning_rate": 1.5779162865806053e-08,
+      "loss": 0.3833,
       "step": 137260
     },
     {
-      "epoch": 4.83,
-      "learning_rate": 1.556516119356699e-07,
-      "loss": 0.2459,
+      "epoch": 4.947021299599957,
+      "grad_norm": 0.2867159843444824,
+      "learning_rate": 1.5675670106890527e-08,
+      "loss": 0.3928,
       "step": 137265
     },
     {
-      "epoch": 4.83,
-      "learning_rate": 1.553343821415254e-07,
-      "loss": 0.2599,
+      "epoch": 4.9472014992611815,
+      "grad_norm": 0.221615731716156,
+      "learning_rate": 1.557251775455848e-08,
+      "loss": 0.3601,
       "step": 137270
     },
     {
-      "epoch": 4.83,
-      "learning_rate": 1.5501747493899266e-07,
-      "loss": 0.2809,
+      "epoch": 4.947381698922406,
+      "grad_norm": 0.17245341837406158,
+      "learning_rate": 1.5469705810211566e-08,
+      "loss": 0.3856,
       "step": 137275
     },
     {
-      "epoch": 4.83,
-      "learning_rate": 1.5470089033218504e-07,
-      "loss": 0.2483,
+      "epoch": 4.947561898583631,
+      "grad_norm": 0.23257580399513245,
+      "learning_rate": 1.5367234275254215e-08,
+      "loss": 0.391,
       "step": 137280
     },
     {
-      "epoch": 4.83,
-      "learning_rate": 1.543846283252104e-07,
-      "loss": 0.2593,
+      "epoch": 4.947742098244856,
+      "grad_norm": 0.2740226984024048,
+      "learning_rate": 1.5265103151079762e-08,
+      "loss": 0.3524,
       "step": 137285
     },
     {
-      "epoch": 4.83,
-      "learning_rate": 1.5406868892217928e-07,
-      "loss": 0.2499,
+      "epoch": 4.94792229790608,
+      "grad_norm": 0.23505844175815582,
+      "learning_rate": 1.5163312439081533e-08,
+      "loss": 0.3628,
       "step": 137290
     },
     {
-      "epoch": 4.83,
-      "learning_rate": 1.5375307212719125e-07,
-      "loss": 0.2366,
+      "epoch": 4.948102497567304,
+      "grad_norm": 0.26007720828056335,
+      "learning_rate": 1.5061862140644535e-08,
+      "loss": 0.3935,
       "step": 137295
     },
     {
-      "epoch": 4.83,
-      "learning_rate": 1.5343777794434577e-07,
-      "loss": 0.2279,
+      "epoch": 4.948282697228529,
+      "grad_norm": 0.24692949652671814,
+      "learning_rate": 1.4960752257153764e-08,
+      "loss": 0.3707,
       "step": 137300
     },
     {
-      "epoch": 4.83,
-      "learning_rate": 1.531228063777368e-07,
-      "loss": 0.2305,
+      "epoch": 4.948462896889754,
+      "grad_norm": 0.25878119468688965,
+      "learning_rate": 1.4859982789985905e-08,
+      "loss": 0.3508,
       "step": 137305
     },
     {
-      "epoch": 4.83,
-      "learning_rate": 1.5280815743144993e-07,
-      "loss": 0.2283,
+      "epoch": 4.9486430965509784,
+      "grad_norm": 0.3022497892379761,
+      "learning_rate": 1.4759553740512078e-08,
+      "loss": 0.3793,
       "step": 137310
     },
     {
-      "epoch": 4.83,
-      "learning_rate": 1.5249383110957915e-07,
-      "loss": 0.2718,
+      "epoch": 4.948823296212203,
+      "grad_norm": 0.21404872834682465,
+      "learning_rate": 1.4659465110103409e-08,
+      "loss": 0.3658,
       "step": 137315
     },
     {
-      "epoch": 4.83,
-      "learning_rate": 1.5217982741619897e-07,
-      "loss": 0.2493,
+      "epoch": 4.949003495873428,
+      "grad_norm": 0.23077760636806488,
+      "learning_rate": 1.4559716900122699e-08,
+      "loss": 0.3532,
       "step": 137320
     },
     {
-      "epoch": 4.83,
-      "learning_rate": 1.5186614635538664e-07,
-      "loss": 0.2549,
+      "epoch": 4.949183695534653,
+      "grad_norm": 0.29384756088256836,
+      "learning_rate": 1.4460309111927194e-08,
+      "loss": 0.3861,
       "step": 137325
     },
     {
-      "epoch": 4.83,
-      "learning_rate": 1.5155278793121953e-07,
-      "loss": 0.2453,
+      "epoch": 4.949363895195877,
+      "grad_norm": 0.26955410838127136,
+      "learning_rate": 1.4361241746874143e-08,
+      "loss": 0.3721,
       "step": 137330
     },
     {
-      "epoch": 4.83,
-      "learning_rate": 1.5123975214776098e-07,
-      "loss": 0.2457,
+      "epoch": 4.949544094857101,
+      "grad_norm": 0.3485639989376068,
+      "learning_rate": 1.426251480630969e-08,
+      "loss": 0.4169,
       "step": 137335
     },
     {
-      "epoch": 4.83,
-      "learning_rate": 1.5092703900908e-07,
-      "loss": 0.2334,
+      "epoch": 4.949724294518326,
+      "grad_norm": 0.2870672643184662,
+      "learning_rate": 1.4164128291582757e-08,
+      "loss": 0.344,
       "step": 137340
     },
     {
-      "epoch": 4.83,
-      "learning_rate": 1.5061464851923445e-07,
-      "loss": 0.2459,
+      "epoch": 4.949904494179551,
+      "grad_norm": 0.2651885449886322,
+      "learning_rate": 1.4066082204031162e-08,
+      "loss": 0.386,
       "step": 137345
     },
     {
-      "epoch": 4.83,
-      "learning_rate": 1.503025806822822e-07,
-      "loss": 0.251,
+      "epoch": 4.950084693840775,
+      "grad_norm": 0.21898417174816132,
+      "learning_rate": 1.3968376544992723e-08,
+      "loss": 0.3918,
       "step": 137350
     },
     {
-      "epoch": 4.83,
-      "learning_rate": 1.4999083550227277e-07,
-      "loss": 0.2453,
+      "epoch": 4.950264893502,
+      "grad_norm": 0.23120352625846863,
+      "learning_rate": 1.3871011315796934e-08,
+      "loss": 0.3546,
       "step": 137355
     },
     {
-      "epoch": 4.83,
-      "learning_rate": 1.496794129832585e-07,
-      "loss": 0.2514,
+      "epoch": 4.950445093163225,
+      "grad_norm": 0.2565666437149048,
+      "learning_rate": 1.377398651777051e-08,
+      "loss": 0.3648,
       "step": 137360
     },
     {
-      "epoch": 4.83,
-      "learning_rate": 1.4936831312927778e-07,
-      "loss": 0.2441,
+      "epoch": 4.95062529282445,
+      "grad_norm": 0.26287904381752014,
+      "learning_rate": 1.3677302152237393e-08,
+      "loss": 0.3705,
       "step": 137365
     },
     {
-      "epoch": 4.83,
-      "learning_rate": 1.4905753594437465e-07,
-      "loss": 0.2444,
+      "epoch": 4.950805492485674,
+      "grad_norm": 0.2553930878639221,
+      "learning_rate": 1.3580958220513195e-08,
+      "loss": 0.3652,
       "step": 137370
     },
     {
-      "epoch": 4.83,
-      "learning_rate": 1.487470814325792e-07,
-      "loss": 0.2529,
+      "epoch": 4.950985692146899,
+      "grad_norm": 0.2285178005695343,
+      "learning_rate": 1.3484954723910759e-08,
+      "loss": 0.3703,
       "step": 137375
     },
     {
-      "epoch": 4.83,
-      "learning_rate": 1.484369495979271e-07,
-      "loss": 0.2412,
+      "epoch": 4.951165891808124,
+      "grad_norm": 0.2720915377140045,
+      "learning_rate": 1.3389291663737368e-08,
+      "loss": 0.3932,
       "step": 137380
     },
     {
-      "epoch": 4.83,
-      "learning_rate": 1.4812714044444286e-07,
-      "loss": 0.2791,
+      "epoch": 4.951346091469348,
+      "grad_norm": 0.26943936944007874,
+      "learning_rate": 1.3293969041297538e-08,
+      "loss": 0.3849,
       "step": 137385
     },
     {
-      "epoch": 4.83,
-      "learning_rate": 1.4781765397614834e-07,
-      "loss": 0.2492,
+      "epoch": 4.951526291130572,
+      "grad_norm": 0.2463103085756302,
+      "learning_rate": 1.3198986857890227e-08,
+      "loss": 0.3596,
       "step": 137390
     },
     {
-      "epoch": 4.83,
-      "learning_rate": 1.4750849019706526e-07,
-      "loss": 0.2469,
+      "epoch": 4.951706490791797,
+      "grad_norm": 0.3053506016731262,
+      "learning_rate": 1.3104345114808848e-08,
+      "loss": 0.3752,
       "step": 137395
     },
     {
-      "epoch": 4.83,
-      "learning_rate": 1.4719964911120433e-07,
-      "loss": 0.2257,
+      "epoch": 4.951886690453022,
+      "grad_norm": 0.24420666694641113,
+      "learning_rate": 1.301004381334403e-08,
+      "loss": 0.3704,
       "step": 137400
     },
     {
-      "epoch": 4.83,
-      "learning_rate": 1.46891130722579e-07,
-      "loss": 0.2551,
+      "epoch": 4.952066890114247,
+      "grad_norm": 0.26522696018218994,
+      "learning_rate": 1.2916082954780861e-08,
+      "loss": 0.372,
       "step": 137405
     },
     {
-      "epoch": 4.83,
-      "learning_rate": 1.465829350351944e-07,
-      "loss": 0.2612,
+      "epoch": 4.952247089775471,
+      "grad_norm": 0.2860383987426758,
+      "learning_rate": 1.2822462540396096e-08,
+      "loss": 0.3768,
       "step": 137410
     },
     {
-      "epoch": 4.83,
-      "learning_rate": 1.462750620530501e-07,
-      "loss": 0.253,
+      "epoch": 4.952427289436696,
+      "grad_norm": 0.21746990084648132,
+      "learning_rate": 1.2729182571466492e-08,
+      "loss": 0.3822,
       "step": 137415
     },
     {
-      "epoch": 4.83,
-      "learning_rate": 1.4596751178014567e-07,
-      "loss": 0.2282,
+      "epoch": 4.952607489097921,
+      "grad_norm": 0.24215134978294373,
+      "learning_rate": 1.2636243049266028e-08,
+      "loss": 0.3612,
       "step": 137420
     },
     {
-      "epoch": 4.83,
-      "learning_rate": 1.4566028422046962e-07,
-      "loss": 0.265,
+      "epoch": 4.952787688759145,
+      "grad_norm": 0.23102861642837524,
+      "learning_rate": 1.254364397506036e-08,
+      "loss": 0.3624,
       "step": 137425
     },
     {
-      "epoch": 4.84,
-      "learning_rate": 1.4535337937801875e-07,
-      "loss": 0.2416,
+      "epoch": 4.952967888420369,
+      "grad_norm": 0.2532614767551422,
+      "learning_rate": 1.245138535010959e-08,
+      "loss": 0.3882,
       "step": 137430
     },
     {
-      "epoch": 4.84,
-      "learning_rate": 1.4504679725677594e-07,
-      "loss": 0.2468,
+      "epoch": 4.953148088081594,
+      "grad_norm": 0.25287774205207825,
+      "learning_rate": 1.235946717566827e-08,
+      "loss": 0.4318,
       "step": 137435
     },
     {
-      "epoch": 4.84,
-      "learning_rate": 1.4474053786071862e-07,
-      "loss": 0.2627,
+      "epoch": 4.953328287742819,
+      "grad_norm": 0.26038363575935364,
+      "learning_rate": 1.226788945299373e-08,
+      "loss": 0.3708,
       "step": 137440
     },
     {
-      "epoch": 4.84,
-      "learning_rate": 1.4443460119382413e-07,
-      "loss": 0.2615,
+      "epoch": 4.953508487404044,
+      "grad_norm": 0.2988220453262329,
+      "learning_rate": 1.2176652183329418e-08,
+      "loss": 0.3655,
       "step": 137445
     },
     {
-      "epoch": 4.84,
-      "learning_rate": 1.4412898726006708e-07,
-      "loss": 0.2619,
+      "epoch": 4.953688687065268,
+      "grad_norm": 0.3282366096973419,
+      "learning_rate": 1.2085755367921558e-08,
+      "loss": 0.3618,
       "step": 137450
     },
     {
-      "epoch": 4.84,
-      "learning_rate": 1.4382369606341372e-07,
-      "loss": 0.2418,
+      "epoch": 4.953868886726493,
+      "grad_norm": 0.31220465898513794,
+      "learning_rate": 1.1995199008008051e-08,
+      "loss": 0.3505,
       "step": 137455
     },
     {
-      "epoch": 4.84,
-      "learning_rate": 1.4351872760783036e-07,
-      "loss": 0.2381,
+      "epoch": 4.954049086387718,
+      "grad_norm": 0.32093510031700134,
+      "learning_rate": 1.1904983104821243e-08,
+      "loss": 0.3727,
       "step": 137460
     },
     {
-      "epoch": 4.84,
-      "learning_rate": 1.4321408189727214e-07,
-      "loss": 0.2456,
+      "epoch": 4.9542292860489425,
+      "grad_norm": 0.2403687983751297,
+      "learning_rate": 1.1815107659590707e-08,
+      "loss": 0.3588,
       "step": 137465
     },
     {
-      "epoch": 4.84,
-      "learning_rate": 1.4290975893570258e-07,
-      "loss": 0.2547,
+      "epoch": 4.954409485710167,
+      "grad_norm": 0.21455620229244232,
+      "learning_rate": 1.172557267354324e-08,
+      "loss": 0.3521,
       "step": 137470
     },
     {
-      "epoch": 4.84,
-      "learning_rate": 1.426057587270657e-07,
-      "loss": 0.2227,
+      "epoch": 4.954589685371392,
+      "grad_norm": 0.26999422907829285,
+      "learning_rate": 1.1636378147897308e-08,
+      "loss": 0.3487,
       "step": 137475
     },
     {
-      "epoch": 4.84,
-      "learning_rate": 1.423020812753112e-07,
-      "loss": 0.2465,
+      "epoch": 4.954769885032616,
+      "grad_norm": 0.2714780569076538,
+      "learning_rate": 1.1547524083865835e-08,
+      "loss": 0.3968,
       "step": 137480
     },
     {
-      "epoch": 4.84,
-      "learning_rate": 1.4199872658438308e-07,
-      "loss": 0.2424,
+      "epoch": 4.954950084693841,
+      "grad_norm": 0.3368239104747772,
+      "learning_rate": 1.1459010482661735e-08,
+      "loss": 0.3995,
       "step": 137485
     },
     {
-      "epoch": 4.84,
-      "learning_rate": 1.4169569465821708e-07,
-      "loss": 0.2548,
+      "epoch": 4.955130284355065,
+      "grad_norm": 0.24230121076107025,
+      "learning_rate": 1.137083734549238e-08,
+      "loss": 0.3657,
       "step": 137490
     },
     {
-      "epoch": 4.84,
-      "learning_rate": 1.4139298550075174e-07,
-      "loss": 0.2368,
+      "epoch": 4.95531048401629,
+      "grad_norm": 0.20803499221801758,
+      "learning_rate": 1.1283004673554032e-08,
+      "loss": 0.3669,
       "step": 137495
     },
     {
-      "epoch": 4.84,
-      "learning_rate": 1.4109059911591726e-07,
-      "loss": 0.2584,
+      "epoch": 4.955490683677515,
+      "grad_norm": 0.19922496378421783,
+      "learning_rate": 1.1195512468048508e-08,
+      "loss": 0.3611,
       "step": 137500
     },
     {
-      "epoch": 4.84,
-      "eval_loss": 0.24842242896556854,
-      "eval_runtime": 10.5421,
-      "eval_samples_per_second": 9.486,
-      "eval_steps_per_second": 9.486,
+      "epoch": 4.955490683677515,
+      "eval_loss": 0.42881152033805847,
+      "eval_runtime": 3.5375,
+      "eval_samples_per_second": 28.268,
+      "eval_steps_per_second": 7.067,
       "step": 137500
     },
     {
-      "epoch": 4.84,
-      "learning_rate": 1.4078853550763827e-07,
-      "loss": 0.2424,
+      "epoch": 4.9556708833387395,
+      "grad_norm": 0.2823123335838318,
+      "learning_rate": 1.1108360730169299e-08,
+      "loss": 0.3637,
       "step": 137505
     },
     {
-      "epoch": 4.84,
-      "learning_rate": 1.404867946798366e-07,
-      "loss": 0.2437,
+      "epoch": 4.955851082999964,
+      "grad_norm": 0.2755952775478363,
+      "learning_rate": 1.1021549461096015e-08,
+      "loss": 0.3775,
       "step": 137510
     },
     {
-      "epoch": 4.84,
-      "learning_rate": 1.4018537663643138e-07,
-      "loss": 0.2562,
+      "epoch": 4.956031282661189,
+      "grad_norm": 0.24063868820667267,
+      "learning_rate": 1.093507866201937e-08,
+      "loss": 0.3269,
       "step": 137515
     },
     {
-      "epoch": 4.84,
-      "learning_rate": 1.3988428138133614e-07,
-      "loss": 0.2518,
+      "epoch": 4.956211482322413,
+      "grad_norm": 0.20716242492198944,
+      "learning_rate": 1.0848948334113428e-08,
+      "loss": 0.3577,
       "step": 137520
     },
     {
-      "epoch": 4.84,
-      "learning_rate": 1.3958350891845884e-07,
-      "loss": 0.2405,
+      "epoch": 4.9563916819836376,
+      "grad_norm": 0.30676987767219543,
+      "learning_rate": 1.0763158478549474e-08,
+      "loss": 0.3848,
       "step": 137525
     },
     {
-      "epoch": 4.84,
-      "learning_rate": 1.3928305925170748e-07,
-      "loss": 0.265,
+      "epoch": 4.956571881644862,
+      "grad_norm": 0.2790820896625519,
+      "learning_rate": 1.0677709096501565e-08,
+      "loss": 0.354,
       "step": 137530
     },
     {
-      "epoch": 4.84,
-      "learning_rate": 1.3898293238498173e-07,
-      "loss": 0.2407,
+      "epoch": 4.956752081306087,
+      "grad_norm": 0.22698865830898285,
+      "learning_rate": 1.0609594732577389e-08,
+      "loss": 0.3403,
       "step": 137535
     },
     {
-      "epoch": 4.84,
-      "learning_rate": 1.3868312832217566e-07,
-      "loss": 0.243,
+      "epoch": 4.956932280967312,
+      "grad_norm": 0.19917847216129303,
+      "learning_rate": 1.0524758205784913e-08,
+      "loss": 0.3678,
       "step": 137540
     },
     {
-      "epoch": 4.84,
-      "learning_rate": 1.3838364706718898e-07,
-      "loss": 0.2689,
+      "epoch": 4.9571124806285365,
+      "grad_norm": 0.2684897482395172,
+      "learning_rate": 1.0440262155755886e-08,
+      "loss": 0.3663,
       "step": 137545
     },
     {
-      "epoch": 4.84,
-      "learning_rate": 1.3808448862390467e-07,
-      "loss": 0.2632,
+      "epoch": 4.957292680289761,
+      "grad_norm": 0.24667024612426758,
+      "learning_rate": 1.0356106583639391e-08,
+      "loss": 0.3596,
       "step": 137550
     },
     {
-      "epoch": 4.84,
-      "learning_rate": 1.377856529962085e-07,
-      "loss": 0.2625,
+      "epoch": 4.957472879950986,
+      "grad_norm": 0.1977359503507614,
+      "learning_rate": 1.0272291490581731e-08,
+      "loss": 0.3723,
       "step": 137555
     },
     {
-      "epoch": 4.84,
-      "learning_rate": 1.3748714018797792e-07,
-      "loss": 0.2655,
+      "epoch": 4.957653079612211,
+      "grad_norm": 0.26738354563713074,
+      "learning_rate": 1.0188816877726437e-08,
+      "loss": 0.3911,
       "step": 137560
     },
     {
-      "epoch": 4.84,
-      "learning_rate": 1.3718895020309598e-07,
-      "loss": 0.2619,
+      "epoch": 4.957833279273435,
+      "grad_norm": 0.2211945503950119,
+      "learning_rate": 1.0105682746211487e-08,
+      "loss": 0.3196,
       "step": 137565
     },
     {
-      "epoch": 4.84,
-      "learning_rate": 1.3689108304542896e-07,
-      "loss": 0.2728,
+      "epoch": 4.958013478934659,
+      "grad_norm": 0.27656158804893494,
+      "learning_rate": 1.0022889097169308e-08,
+      "loss": 0.348,
       "step": 137570
     },
     {
-      "epoch": 4.84,
-      "learning_rate": 1.3659353871884606e-07,
-      "loss": 0.2463,
+      "epoch": 4.958193678595884,
+      "grad_norm": 0.2519603967666626,
+      "learning_rate": 9.940435931724001e-09,
+      "loss": 0.3757,
       "step": 137575
     },
     {
-      "epoch": 4.84,
-      "learning_rate": 1.3629631722721082e-07,
-      "loss": 0.2852,
+      "epoch": 4.958373878257109,
+      "grad_norm": 0.25226572155952454,
+      "learning_rate": 9.858323251005219e-09,
+      "loss": 0.3778,
       "step": 137580
     },
     {
-      "epoch": 4.84,
-      "learning_rate": 1.3599941857438125e-07,
-      "loss": 0.2604,
+      "epoch": 4.9585540779183335,
+      "grad_norm": 0.20732538402080536,
+      "learning_rate": 9.77655105612596e-09,
+      "loss": 0.3805,
       "step": 137585
     },
     {
-      "epoch": 4.84,
-      "learning_rate": 1.357028427642154e-07,
-      "loss": 0.2438,
+      "epoch": 4.958734277579558,
+      "grad_norm": 0.2726946771144867,
+      "learning_rate": 9.695119348204773e-09,
+      "loss": 0.3624,
       "step": 137590
     },
     {
-      "epoch": 4.84,
-      "learning_rate": 1.354065898005602e-07,
-      "loss": 0.2587,
+      "epoch": 4.958914477240783,
+      "grad_norm": 0.20067138969898224,
+      "learning_rate": 9.614028128346331e-09,
+      "loss": 0.333,
       "step": 137595
     },
     {
-      "epoch": 4.84,
-      "learning_rate": 1.3511065968726255e-07,
-      "loss": 0.2569,
+      "epoch": 4.959094676902008,
+      "grad_norm": 0.2827952206134796,
+      "learning_rate": 9.533277397660856e-09,
+      "loss": 0.3483,
       "step": 137600
     },
     {
-      "epoch": 4.84,
-      "learning_rate": 1.3481505242816661e-07,
-      "loss": 0.2462,
+      "epoch": 4.959274876563232,
+      "grad_norm": 0.23344914615154266,
+      "learning_rate": 9.452867157247469e-09,
+      "loss": 0.3538,
       "step": 137605
     },
     {
-      "epoch": 4.84,
-      "learning_rate": 1.3451976802711376e-07,
-      "loss": 0.2453,
+      "epoch": 4.959455076224456,
+      "grad_norm": 0.2827022969722748,
+      "learning_rate": 9.372797408196965e-09,
+      "loss": 0.3309,
       "step": 137610
     },
     {
-      "epoch": 4.84,
-      "learning_rate": 1.3422480648793146e-07,
-      "loss": 0.2474,
+      "epoch": 4.959635275885681,
+      "grad_norm": 0.2828150987625122,
+      "learning_rate": 9.293068151605689e-09,
+      "loss": 0.3894,
       "step": 137615
     },
     {
-      "epoch": 4.84,
-      "learning_rate": 1.3393016781445278e-07,
-      "loss": 0.26,
+      "epoch": 4.959815475546906,
+      "grad_norm": 0.26365163922309875,
+      "learning_rate": 9.213679388558882e-09,
+      "loss": 0.3679,
       "step": 137620
     },
     {
-      "epoch": 4.84,
-      "learning_rate": 1.336358520105052e-07,
-      "loss": 0.2556,
+      "epoch": 4.9599956752081305,
+      "grad_norm": 0.291981965303421,
+      "learning_rate": 9.134631120136238e-09,
+      "loss": 0.3676,
       "step": 137625
     },
     {
-      "epoch": 4.84,
-      "learning_rate": 1.3334185907990792e-07,
-      "loss": 0.2798,
+      "epoch": 4.960175874869355,
+      "grad_norm": 0.2089691460132599,
+      "learning_rate": 9.055923347414675e-09,
+      "loss": 0.3302,
       "step": 137630
     },
     {
-      "epoch": 4.84,
-      "learning_rate": 1.330481890264801e-07,
-      "loss": 0.2664,
+      "epoch": 4.96035607453058,
+      "grad_norm": 0.33166277408599854,
+      "learning_rate": 8.977556071471105e-09,
+      "loss": 0.352,
       "step": 137635
     },
     {
-      "epoch": 4.84,
-      "learning_rate": 1.3275484185403254e-07,
-      "loss": 0.2458,
+      "epoch": 4.960536274191805,
+      "grad_norm": 0.2414046823978424,
+      "learning_rate": 8.899529293365794e-09,
+      "loss": 0.3702,
       "step": 137640
     },
     {
-      "epoch": 4.84,
-      "learning_rate": 1.324618175663761e-07,
-      "loss": 0.2679,
+      "epoch": 4.960716473853029,
+      "grad_norm": 0.2852274477481842,
+      "learning_rate": 8.821843014170106e-09,
+      "loss": 0.3833,
       "step": 137645
     },
     {
-      "epoch": 4.84,
-      "learning_rate": 1.321691161673133e-07,
-      "loss": 0.2491,
+      "epoch": 4.960896673514254,
+      "grad_norm": 0.23705297708511353,
+      "learning_rate": 8.744497234935978e-09,
+      "loss": 0.3855,
       "step": 137650
     },
     {
-      "epoch": 4.84,
-      "learning_rate": 1.3187673766064667e-07,
-      "loss": 0.2237,
+      "epoch": 4.961076873175479,
+      "grad_norm": 0.28535374999046326,
+      "learning_rate": 8.66749195671812e-09,
+      "loss": 0.3653,
       "step": 137655
     },
     {
-      "epoch": 4.84,
-      "learning_rate": 1.3158468205017038e-07,
-      "loss": 0.2304,
+      "epoch": 4.961257072836703,
+      "grad_norm": 0.23056595027446747,
+      "learning_rate": 8.590827180571248e-09,
+      "loss": 0.3719,
       "step": 137660
     },
     {
-      "epoch": 4.84,
-      "learning_rate": 1.3129294933967862e-07,
-      "loss": 0.2609,
+      "epoch": 4.961437272497927,
+      "grad_norm": 0.24574077129364014,
+      "learning_rate": 8.514502907533418e-09,
+      "loss": 0.397,
       "step": 137665
     },
     {
-      "epoch": 4.84,
-      "learning_rate": 1.3100153953296e-07,
-      "loss": 0.2479,
+      "epoch": 4.961617472159152,
+      "grad_norm": 0.24938498437404633,
+      "learning_rate": 8.438519138645462e-09,
+      "loss": 0.3871,
       "step": 137670
     },
     {
-      "epoch": 4.84,
-      "learning_rate": 1.3071045263379765e-07,
-      "loss": 0.2493,
+      "epoch": 4.961797671820377,
+      "grad_norm": 0.35726431012153625,
+      "learning_rate": 8.362875874945442e-09,
+      "loss": 0.3636,
       "step": 137675
     },
     {
-      "epoch": 4.84,
-      "learning_rate": 1.304196886459691e-07,
-      "loss": 0.2465,
+      "epoch": 4.961977871481602,
+      "grad_norm": 0.2681273818016052,
+      "learning_rate": 8.287573117465863e-09,
+      "loss": 0.3763,
       "step": 137680
     },
     {
-      "epoch": 4.84,
-      "learning_rate": 1.3012924757325184e-07,
-      "loss": 0.2568,
+      "epoch": 4.962158071142826,
+      "grad_norm": 0.25855353474617004,
+      "learning_rate": 8.212610867225356e-09,
+      "loss": 0.3967,
       "step": 137685
     },
     {
-      "epoch": 4.84,
-      "learning_rate": 1.2983912941941789e-07,
-      "loss": 0.2331,
+      "epoch": 4.962338270804051,
+      "grad_norm": 0.22436444461345673,
+      "learning_rate": 8.137989125250877e-09,
+      "loss": 0.3584,
       "step": 137690
     },
     {
-      "epoch": 4.84,
-      "learning_rate": 1.295493341882309e-07,
-      "loss": 0.2528,
+      "epoch": 4.962518470465276,
+      "grad_norm": 0.2069372534751892,
+      "learning_rate": 8.063707892558281e-09,
+      "loss": 0.3502,
       "step": 137695
     },
     {
-      "epoch": 4.84,
-      "learning_rate": 1.292598618834573e-07,
-      "loss": 0.2509,
+      "epoch": 4.9626986701265,
+      "grad_norm": 0.23629046976566315,
+      "learning_rate": 7.989767170157869e-09,
+      "loss": 0.3882,
       "step": 137700
     },
     {
-      "epoch": 4.84,
-      "learning_rate": 1.289707125088524e-07,
-      "loss": 0.2602,
+      "epoch": 4.962878869787724,
+      "grad_norm": 0.24616286158561707,
+      "learning_rate": 7.916166959059944e-09,
+      "loss": 0.3603,
       "step": 137705
     },
     {
-      "epoch": 4.85,
-      "learning_rate": 1.2868188606817156e-07,
-      "loss": 0.2426,
+      "epoch": 4.963059069448949,
+      "grad_norm": 0.22220478951931,
+      "learning_rate": 7.84290726026371e-09,
+      "loss": 0.3726,
       "step": 137710
     },
     {
-      "epoch": 4.85,
-      "learning_rate": 1.2839338256516454e-07,
-      "loss": 0.2452,
+      "epoch": 4.963239269110174,
+      "grad_norm": 0.23332755267620087,
+      "learning_rate": 7.769988074771139e-09,
+      "loss": 0.3892,
       "step": 137715
     },
     {
-      "epoch": 4.85,
-      "learning_rate": 1.281052020035811e-07,
-      "loss": 0.2606,
+      "epoch": 4.963419468771399,
+      "grad_norm": 0.2551371455192566,
+      "learning_rate": 7.697409403573108e-09,
+      "loss": 0.3594,
       "step": 137720
     },
     {
-      "epoch": 4.85,
-      "learning_rate": 1.2781734438715998e-07,
-      "loss": 0.2324,
+      "epoch": 4.963599668432623,
+      "grad_norm": 0.26924464106559753,
+      "learning_rate": 7.625171247657714e-09,
+      "loss": 0.3503,
       "step": 137725
     },
     {
-      "epoch": 4.85,
-      "learning_rate": 1.275298097196398e-07,
-      "loss": 0.2512,
+      "epoch": 4.963779868093848,
+      "grad_norm": 0.25808215141296387,
+      "learning_rate": 7.553273608013057e-09,
+      "loss": 0.3743,
       "step": 137730
     },
     {
-      "epoch": 4.85,
-      "learning_rate": 1.2724259800475092e-07,
-      "loss": 0.2296,
+      "epoch": 4.963960067755073,
+      "grad_norm": 0.30064940452575684,
+      "learning_rate": 7.481716485616131e-09,
+      "loss": 0.3813,
       "step": 137735
     },
     {
-      "epoch": 4.85,
-      "learning_rate": 1.2695570924622924e-07,
-      "loss": 0.247,
+      "epoch": 4.9641402674162975,
+      "grad_norm": 0.23391714692115784,
+      "learning_rate": 7.410499881438382e-09,
+      "loss": 0.3778,
       "step": 137740
     },
     {
-      "epoch": 4.85,
-      "learning_rate": 1.2666914344779402e-07,
-      "loss": 0.2445,
+      "epoch": 4.964320467077522,
+      "grad_norm": 0.25494813919067383,
+      "learning_rate": 7.339623796456807e-09,
+      "loss": 0.3887,
       "step": 137745
     },
     {
-      "epoch": 4.85,
-      "learning_rate": 1.2638290061316727e-07,
-      "loss": 0.2427,
+      "epoch": 4.964500666738747,
+      "grad_norm": 0.30371445417404175,
+      "learning_rate": 7.2690882316345245e-09,
+      "loss": 0.3799,
       "step": 137750
     },
     {
-      "epoch": 4.85,
-      "learning_rate": 1.2609698074606546e-07,
-      "loss": 0.2335,
+      "epoch": 4.964680866399971,
+      "grad_norm": 0.3035849630832672,
+      "learning_rate": 7.198893187931877e-09,
+      "loss": 0.372,
       "step": 137755
     },
     {
-      "epoch": 4.85,
-      "learning_rate": 1.258113838502023e-07,
-      "loss": 0.2391,
+      "epoch": 4.964861066061196,
+      "grad_norm": 0.7903338670730591,
+      "learning_rate": 7.129038666306431e-09,
+      "loss": 0.3787,
       "step": 137760
     },
     {
-      "epoch": 4.85,
-      "learning_rate": 1.2552610992928594e-07,
-      "loss": 0.2573,
+      "epoch": 4.96504126572242,
+      "grad_norm": 0.24585622549057007,
+      "learning_rate": 7.059524667707429e-09,
+      "loss": 0.3528,
       "step": 137765
     },
     {
-      "epoch": 4.85,
-      "learning_rate": 1.2524115898701893e-07,
-      "loss": 0.2355,
+      "epoch": 4.965221465383645,
+      "grad_norm": 0.6391294002532959,
+      "learning_rate": 6.99035119308411e-09,
+      "loss": 0.3562,
       "step": 137770
     },
     {
-      "epoch": 4.85,
-      "learning_rate": 1.249565310271039e-07,
-      "loss": 0.2606,
+      "epoch": 4.96540166504487,
+      "grad_norm": 0.2620919942855835,
+      "learning_rate": 6.9215182433773895e-09,
+      "loss": 0.349,
       "step": 137775
     },
     {
-      "epoch": 4.85,
-      "learning_rate": 1.2467222605323514e-07,
-      "loss": 0.2592,
+      "epoch": 4.9655818647060945,
+      "grad_norm": 0.25720539689064026,
+      "learning_rate": 6.85302581952818e-09,
+      "loss": 0.334,
       "step": 137780
     },
     {
-      "epoch": 4.85,
-      "learning_rate": 1.243882440691041e-07,
-      "loss": 0.2403,
+      "epoch": 4.965762064367319,
+      "grad_norm": 0.23249337077140808,
+      "learning_rate": 6.784873922466295e-09,
+      "loss": 0.3948,
       "step": 137785
     },
     {
-      "epoch": 4.85,
-      "learning_rate": 1.2410458507839673e-07,
-      "loss": 0.2583,
+      "epoch": 4.965942264028544,
+      "grad_norm": 0.29254230856895447,
+      "learning_rate": 6.717062553121545e-09,
+      "loss": 0.3938,
       "step": 137790
     },
     {
-      "epoch": 4.85,
-      "learning_rate": 1.238212490847962e-07,
-      "loss": 0.2581,
+      "epoch": 4.966122463689768,
+      "grad_norm": 0.27492207288742065,
+      "learning_rate": 6.649591712420966e-09,
+      "loss": 0.4013,
       "step": 137795
     },
     {
-      "epoch": 4.85,
-      "learning_rate": 1.2353823609198567e-07,
-      "loss": 0.2451,
+      "epoch": 4.966302663350993,
+      "grad_norm": 0.24353964626789093,
+      "learning_rate": 6.5824614012777174e-09,
+      "loss": 0.3498,
       "step": 137800
     },
     {
-      "epoch": 4.85,
-      "learning_rate": 1.232555461036372e-07,
-      "loss": 0.2673,
+      "epoch": 4.966482863012217,
+      "grad_norm": 0.2534390389919281,
+      "learning_rate": 6.515671620610508e-09,
+      "loss": 0.3779,
       "step": 137805
     },
     {
-      "epoch": 4.85,
-      "learning_rate": 1.229731791234201e-07,
-      "loss": 0.2485,
+      "epoch": 4.966663062673442,
+      "grad_norm": 0.28155553340911865,
+      "learning_rate": 6.449222371326946e-09,
+      "loss": 0.3622,
       "step": 137810
     },
     {
-      "epoch": 4.85,
-      "learning_rate": 1.2269113515500085e-07,
-      "loss": 0.2637,
+      "epoch": 4.966843262334667,
+      "grad_norm": 0.28064605593681335,
+      "learning_rate": 6.383113654334638e-09,
+      "loss": 0.3944,
       "step": 137815
     },
     {
-      "epoch": 4.85,
-      "learning_rate": 1.224094142020432e-07,
-      "loss": 0.2276,
+      "epoch": 4.9670234619958915,
+      "grad_norm": 0.32647833228111267,
+      "learning_rate": 6.31734547053564e-09,
+      "loss": 0.3521,
       "step": 137820
     },
     {
-      "epoch": 4.85,
-      "learning_rate": 1.2212801626820536e-07,
-      "loss": 0.254,
+      "epoch": 4.967203661657116,
+      "grad_norm": 0.2887493073940277,
+      "learning_rate": 6.2519178208209075e-09,
+      "loss": 0.3774,
       "step": 137825
     },
     {
-      "epoch": 4.85,
-      "learning_rate": 1.2184694135713714e-07,
-      "loss": 0.2214,
+      "epoch": 4.967383861318341,
+      "grad_norm": 0.2260408252477646,
+      "learning_rate": 6.186830706086943e-09,
+      "loss": 0.4083,
       "step": 137830
     },
     {
-      "epoch": 4.85,
-      "learning_rate": 1.2156618947249397e-07,
-      "loss": 0.2268,
+      "epoch": 4.967564060979566,
+      "grad_norm": 0.2433057725429535,
+      "learning_rate": 6.122084127216377e-09,
+      "loss": 0.4013,
       "step": 137835
     },
     {
-      "epoch": 4.85,
-      "learning_rate": 1.212857606179174e-07,
-      "loss": 0.2677,
+      "epoch": 4.96774426064079,
+      "grad_norm": 0.2385009527206421,
+      "learning_rate": 6.057678085094609e-09,
+      "loss": 0.3676,
       "step": 137840
     },
     {
-      "epoch": 4.85,
-      "learning_rate": 1.2100565479704896e-07,
-      "loss": 0.2739,
+      "epoch": 4.967924460302014,
+      "grad_norm": 0.20035654306411743,
+      "learning_rate": 5.993612580598717e-09,
+      "loss": 0.3494,
       "step": 137845
     },
     {
-      "epoch": 4.85,
-      "learning_rate": 1.2072587201352736e-07,
-      "loss": 0.2272,
+      "epoch": 4.968104659963239,
+      "grad_norm": 0.28439047932624817,
+      "learning_rate": 5.929887614600227e-09,
+      "loss": 0.3352,
       "step": 137850
     },
     {
-      "epoch": 4.85,
-      "learning_rate": 1.2044641227098308e-07,
-      "loss": 0.2485,
+      "epoch": 4.968284859624464,
+      "grad_norm": 0.2770126163959503,
+      "learning_rate": 5.866503187967886e-09,
+      "loss": 0.3925,
       "step": 137855
     },
     {
-      "epoch": 4.85,
-      "learning_rate": 1.2016727557304652e-07,
-      "loss": 0.2608,
+      "epoch": 4.9684650592856885,
+      "grad_norm": 0.33329346776008606,
+      "learning_rate": 5.803459301564895e-09,
+      "loss": 0.3521,
       "step": 137860
     },
     {
-      "epoch": 4.85,
-      "learning_rate": 1.1988846192334257e-07,
-      "loss": 0.2513,
+      "epoch": 4.968645258946913,
+      "grad_norm": 0.27743661403656006,
+      "learning_rate": 5.7407559562516754e-09,
+      "loss": 0.3377,
       "step": 137865
     },
     {
-      "epoch": 4.85,
-      "learning_rate": 1.1960997132548779e-07,
-      "loss": 0.2354,
+      "epoch": 4.968825458608138,
+      "grad_norm": 0.2770087718963623,
+      "learning_rate": 5.678393152880323e-09,
+      "loss": 0.3618,
       "step": 137870
     },
     {
-      "epoch": 4.85,
-      "learning_rate": 1.1933180378310427e-07,
-      "loss": 0.2553,
+      "epoch": 4.969005658269363,
+      "grad_norm": 0.26707491278648376,
+      "learning_rate": 5.616370892302936e-09,
+      "loss": 0.3735,
       "step": 137875
     },
     {
-      "epoch": 4.85,
-      "learning_rate": 1.1905395929979746e-07,
-      "loss": 0.2457,
+      "epoch": 4.969185857930587,
+      "grad_norm": 0.24807052314281464,
+      "learning_rate": 5.554689175363281e-09,
+      "loss": 0.3664,
       "step": 137880
     },
     {
-      "epoch": 4.85,
-      "learning_rate": 1.1877643787917837e-07,
-      "loss": 0.2558,
+      "epoch": 4.969366057591811,
+      "grad_norm": 0.2263546884059906,
+      "learning_rate": 5.493348002902354e-09,
+      "loss": 0.3802,
       "step": 137885
     },
     {
-      "epoch": 4.85,
-      "learning_rate": 1.1849923952484965e-07,
-      "loss": 0.2364,
+      "epoch": 4.969546257253036,
+      "grad_norm": 0.2367497682571411,
+      "learning_rate": 5.432347375755598e-09,
+      "loss": 0.3439,
       "step": 137890
     },
     {
-      "epoch": 4.85,
-      "learning_rate": 1.1822236424041123e-07,
-      "loss": 0.2577,
+      "epoch": 4.969726456914261,
+      "grad_norm": 0.25867414474487305,
+      "learning_rate": 5.371687294752903e-09,
+      "loss": 0.3795,
       "step": 137895
     },
     {
-      "epoch": 4.85,
-      "learning_rate": 1.1794581202945742e-07,
-      "loss": 0.2297,
+      "epoch": 4.9699066565754855,
+      "grad_norm": 0.2083641141653061,
+      "learning_rate": 5.311367760721387e-09,
+      "loss": 0.3653,
       "step": 137900
     },
     {
-      "epoch": 4.85,
-      "learning_rate": 1.1766958289557983e-07,
-      "loss": 0.2403,
+      "epoch": 4.97008685623671,
+      "grad_norm": 0.24108898639678955,
+      "learning_rate": 5.251388774485388e-09,
+      "loss": 0.3576,
       "step": 137905
     },
     {
-      "epoch": 4.85,
-      "learning_rate": 1.1739367684236446e-07,
-      "loss": 0.2338,
+      "epoch": 4.970267055897935,
+      "grad_norm": 0.2666069567203522,
+      "learning_rate": 5.1917503368581474e-09,
+      "loss": 0.3818,
       "step": 137910
     },
     {
-      "epoch": 4.85,
-      "learning_rate": 1.1711809387339457e-07,
-      "loss": 0.2553,
+      "epoch": 4.97044725555916,
+      "grad_norm": 0.24382546544075012,
+      "learning_rate": 5.132452448655678e-09,
+      "loss": 0.3721,
       "step": 137915
     },
     {
-      "epoch": 4.85,
-      "learning_rate": 1.1684283399224505e-07,
-      "loss": 0.2175,
+      "epoch": 4.970627455220384,
+      "grad_norm": 0.307109534740448,
+      "learning_rate": 5.073495110682891e-09,
+      "loss": 0.4167,
       "step": 137920
     },
     {
-      "epoch": 4.85,
-      "learning_rate": 1.165678972024964e-07,
-      "loss": 0.2418,
+      "epoch": 4.970807654881609,
+      "grad_norm": 0.2820093631744385,
+      "learning_rate": 5.014878323744698e-09,
+      "loss": 0.3609,
       "step": 137925
     },
     {
-      "epoch": 4.85,
-      "learning_rate": 1.162932835077124e-07,
-      "loss": 0.2471,
+      "epoch": 4.970987854542834,
+      "grad_norm": 0.2773934304714203,
+      "learning_rate": 4.956602088640461e-09,
+      "loss": 0.3742,
       "step": 137930
     },
     {
-      "epoch": 4.85,
-      "learning_rate": 1.1601899291146245e-07,
-      "loss": 0.2449,
+      "epoch": 4.971168054204058,
+      "grad_norm": 0.23025000095367432,
+      "learning_rate": 4.898666406163988e-09,
+      "loss": 0.3682,
       "step": 137935
     },
     {
-      "epoch": 4.85,
-      "learning_rate": 1.157450254173048e-07,
-      "loss": 0.2599,
+      "epoch": 4.9713482538652825,
+      "grad_norm": 0.24413655698299408,
+      "learning_rate": 4.841071277103537e-09,
+      "loss": 0.4005,
       "step": 137940
     },
     {
-      "epoch": 4.85,
-      "learning_rate": 1.154713810288005e-07,
-      "loss": 0.2583,
+      "epoch": 4.971528453526507,
+      "grad_norm": 0.23541654646396637,
+      "learning_rate": 4.783816702241817e-09,
+      "loss": 0.3692,
       "step": 137945
     },
     {
-      "epoch": 4.85,
-      "learning_rate": 1.1519805974949948e-07,
-      "loss": 0.2593,
+      "epoch": 4.971708653187732,
+      "grad_norm": 0.2659054398536682,
+      "learning_rate": 4.7269026823643095e-09,
+      "loss": 0.4064,
       "step": 137950
     },
     {
-      "epoch": 4.85,
-      "learning_rate": 1.1492506158295447e-07,
-      "loss": 0.2464,
+      "epoch": 4.971888852848957,
+      "grad_norm": 0.23656433820724487,
+      "learning_rate": 4.670329218242619e-09,
+      "loss": 0.3935,
       "step": 137955
     },
     {
-      "epoch": 4.85,
-      "learning_rate": 1.1465238653270427e-07,
-      "loss": 0.2589,
+      "epoch": 4.972069052510181,
+      "grad_norm": 0.27703598141670227,
+      "learning_rate": 4.614096310648352e-09,
+      "loss": 0.4068,
       "step": 137960
     },
     {
-      "epoch": 4.85,
-      "learning_rate": 1.1438003460229607e-07,
-      "loss": 0.2413,
+      "epoch": 4.972249252171406,
+      "grad_norm": 0.20020613074302673,
+      "learning_rate": 4.558203960347562e-09,
+      "loss": 0.3666,
       "step": 137965
     },
     {
-      "epoch": 4.85,
-      "learning_rate": 1.1410800579526037e-07,
-      "loss": 0.2471,
+      "epoch": 4.972429451832631,
+      "grad_norm": 0.27739599347114563,
+      "learning_rate": 4.5026521681035275e-09,
+      "loss": 0.3525,
       "step": 137970
     },
     {
-      "epoch": 4.85,
-      "learning_rate": 1.1383630011513325e-07,
-      "loss": 0.2523,
+      "epoch": 4.972609651493855,
+      "grad_norm": 0.31181418895721436,
+      "learning_rate": 4.4474409346711984e-09,
+      "loss": 0.412,
       "step": 137975
     },
     {
-      "epoch": 4.85,
-      "learning_rate": 1.1356491756543964e-07,
-      "loss": 0.2421,
+      "epoch": 4.972789851155079,
+      "grad_norm": 0.2787044048309326,
+      "learning_rate": 4.392570260799977e-09,
+      "loss": 0.3389,
       "step": 137980
     },
     {
-      "epoch": 4.85,
-      "learning_rate": 1.1329385814970727e-07,
-      "loss": 0.2348,
+      "epoch": 4.972970050816304,
+      "grad_norm": 0.25257426500320435,
+      "learning_rate": 4.338040147244815e-09,
+      "loss": 0.3701,
       "step": 137985
     },
     {
-      "epoch": 4.85,
-      "learning_rate": 1.1302312187145003e-07,
-      "loss": 0.2566,
+      "epoch": 4.973150250477529,
+      "grad_norm": 0.34394869208335876,
+      "learning_rate": 4.283850594741234e-09,
+      "loss": 0.3779,
       "step": 137990
     },
     {
-      "epoch": 4.86,
-      "learning_rate": 1.1275270873419008e-07,
-      "loss": 0.2439,
+      "epoch": 4.973330450138754,
+      "grad_norm": 0.2812308073043823,
+      "learning_rate": 4.230001604033085e-09,
+      "loss": 0.3751,
       "step": 137995
     },
     {
-      "epoch": 4.86,
-      "learning_rate": 1.1248261874143296e-07,
-      "loss": 0.2405,
+      "epoch": 4.973510649799978,
+      "grad_norm": 0.30743271112442017,
+      "learning_rate": 4.176493175850338e-09,
+      "loss": 0.392,
       "step": 138000
     },
     {
-      "epoch": 4.86,
-      "eval_loss": 0.24844183027744293,
-      "eval_runtime": 10.5317,
-      "eval_samples_per_second": 9.495,
-      "eval_steps_per_second": 9.495,
+      "epoch": 4.973510649799978,
+      "eval_loss": 0.4287694990634918,
+      "eval_runtime": 3.5429,
+      "eval_samples_per_second": 28.226,
+      "eval_steps_per_second": 7.056,
       "step": 138000
     },
     {
-      "epoch": 4.86,
-      "learning_rate": 1.1221285189668696e-07,
-      "loss": 0.2696,
+      "epoch": 4.973690849461203,
+      "grad_norm": 0.20221595466136932,
+      "learning_rate": 4.123325310925741e-09,
+      "loss": 0.374,
       "step": 138005
     },
     {
-      "epoch": 4.86,
-      "learning_rate": 1.1194340820345484e-07,
-      "loss": 0.2456,
+      "epoch": 4.973871049122428,
+      "grad_norm": 0.19391195476055145,
+      "learning_rate": 4.070498009978163e-09,
+      "loss": 0.3898,
       "step": 138010
     },
     {
-      "epoch": 4.86,
-      "learning_rate": 1.1167428766523658e-07,
-      "loss": 0.2656,
+      "epoch": 4.974051248783653,
+      "grad_norm": 0.2685374617576599,
+      "learning_rate": 4.018011273734801e-09,
+      "loss": 0.3656,
       "step": 138015
     },
     {
-      "epoch": 4.86,
-      "learning_rate": 1.114054902855266e-07,
-      "loss": 0.216,
+      "epoch": 4.974231448444877,
+      "grad_norm": 0.2981413006782532,
+      "learning_rate": 3.965865102903421e-09,
+      "loss": 0.339,
       "step": 138020
     },
     {
-      "epoch": 4.86,
-      "learning_rate": 1.1113701606781379e-07,
-      "loss": 0.2607,
+      "epoch": 4.974411648106102,
+      "grad_norm": 0.24895471334457397,
+      "learning_rate": 3.914059498197342e-09,
+      "loss": 0.3855,
       "step": 138025
     },
     {
-      "epoch": 4.86,
-      "learning_rate": 1.1086886501558424e-07,
-      "loss": 0.2399,
+      "epoch": 4.974591847767326,
+      "grad_norm": 0.289763867855072,
+      "learning_rate": 3.862594460324332e-09,
+      "loss": 0.3667,
       "step": 138030
     },
     {
-      "epoch": 4.86,
-      "learning_rate": 1.1060103713232128e-07,
-      "loss": 0.252,
+      "epoch": 4.974772047428551,
+      "grad_norm": 0.2891743779182434,
+      "learning_rate": 3.811469989983829e-09,
+      "loss": 0.3728,
       "step": 138035
     },
     {
-      "epoch": 4.86,
-      "learning_rate": 1.1033353242149713e-07,
-      "loss": 0.2688,
+      "epoch": 4.974952247089775,
+      "grad_norm": 0.30129414796829224,
+      "learning_rate": 3.7606860878725005e-09,
+      "loss": 0.3678,
       "step": 138040
     },
     {
-      "epoch": 4.86,
-      "learning_rate": 1.1006635088659234e-07,
-      "loss": 0.2486,
+      "epoch": 4.975132446751,
+      "grad_norm": 0.3116011619567871,
+      "learning_rate": 3.7102427546842343e-09,
+      "loss": 0.3739,
       "step": 138045
     },
     {
-      "epoch": 4.86,
-      "learning_rate": 1.097994925310708e-07,
-      "loss": 0.2501,
+      "epoch": 4.975312646412225,
+      "grad_norm": 0.24255453050136566,
+      "learning_rate": 3.6601399911018185e-09,
+      "loss": 0.3675,
       "step": 138050
     },
     {
-      "epoch": 4.86,
-      "learning_rate": 1.0953295735840196e-07,
-      "loss": 0.2465,
+      "epoch": 4.9754928460734495,
+      "grad_norm": 0.33993199467658997,
+      "learning_rate": 3.610377797813591e-09,
+      "loss": 0.3956,
       "step": 138055
     },
     {
-      "epoch": 4.86,
-      "learning_rate": 1.0926674537204417e-07,
-      "loss": 0.2444,
+      "epoch": 4.975673045734674,
+      "grad_norm": 0.23433293402194977,
+      "learning_rate": 3.560956175491237e-09,
+      "loss": 0.3584,
       "step": 138060
     },
     {
-      "epoch": 4.86,
-      "learning_rate": 1.09000856575453e-07,
-      "loss": 0.2591,
+      "epoch": 4.975853245395899,
+      "grad_norm": 0.3143203556537628,
+      "learning_rate": 3.511875124811992e-09,
+      "loss": 0.3765,
       "step": 138065
     },
     {
-      "epoch": 4.86,
-      "learning_rate": 1.0873529097208124e-07,
-      "loss": 0.2573,
+      "epoch": 4.976033445057123,
+      "grad_norm": 0.25443294644355774,
+      "learning_rate": 3.4631346464447658e-09,
+      "loss": 0.3904,
       "step": 138070
     },
     {
-      "epoch": 4.86,
-      "learning_rate": 1.084700485653789e-07,
-      "loss": 0.2582,
+      "epoch": 4.976213644718348,
+      "grad_norm": 0.22525587677955627,
+      "learning_rate": 3.4147347410529164e-09,
+      "loss": 0.3743,
       "step": 138075
     },
     {
-      "epoch": 4.86,
-      "learning_rate": 1.0820512935879046e-07,
-      "loss": 0.2638,
+      "epoch": 4.976393844379572,
+      "grad_norm": 0.27883243560791016,
+      "learning_rate": 3.3666754092970266e-09,
+      "loss": 0.3883,
       "step": 138080
     },
     {
-      "epoch": 4.86,
-      "learning_rate": 1.0794053335575205e-07,
-      "loss": 0.2711,
+      "epoch": 4.976574044040797,
+      "grad_norm": 0.2141328603029251,
+      "learning_rate": 3.3189566518293526e-09,
+      "loss": 0.3603,
       "step": 138085
     },
     {
-      "epoch": 4.86,
-      "learning_rate": 1.0767626055969982e-07,
-      "loss": 0.2571,
+      "epoch": 4.976754243702022,
+      "grad_norm": 0.24589018523693085,
+      "learning_rate": 3.2715784692993747e-09,
+      "loss": 0.3851,
       "step": 138090
     },
     {
-      "epoch": 4.86,
-      "learning_rate": 1.074123109740699e-07,
-      "loss": 0.2652,
+      "epoch": 4.9769344433632465,
+      "grad_norm": 0.27227839827537537,
+      "learning_rate": 3.2245408623565733e-09,
+      "loss": 0.3899,
       "step": 138095
     },
     {
-      "epoch": 4.86,
-      "learning_rate": 1.0714868460228456e-07,
-      "loss": 0.2603,
+      "epoch": 4.977114643024471,
+      "grad_norm": 0.2859554588794708,
+      "learning_rate": 3.1778438316393266e-09,
+      "loss": 0.3719,
       "step": 138100
     },
     {
-      "epoch": 4.86,
-      "learning_rate": 1.0688538144776883e-07,
-      "loss": 0.2508,
+      "epoch": 4.977294842685696,
+      "grad_norm": 0.2409016191959381,
+      "learning_rate": 3.1314873777860133e-09,
+      "loss": 0.3319,
       "step": 138105
     },
     {
-      "epoch": 4.86,
-      "learning_rate": 1.0662240151393942e-07,
-      "loss": 0.2563,
+      "epoch": 4.977475042346921,
+      "grad_norm": 0.27938273549079895,
+      "learning_rate": 3.0854715014266845e-09,
+      "loss": 0.3947,
       "step": 138110
     },
     {
-      "epoch": 4.86,
-      "learning_rate": 1.0635974480421585e-07,
-      "loss": 0.2324,
+      "epoch": 4.9776552420081455,
+      "grad_norm": 0.26294222474098206,
+      "learning_rate": 3.039796203185841e-09,
+      "loss": 0.3686,
       "step": 138115
     },
     {
-      "epoch": 4.86,
-      "learning_rate": 1.0609741132200369e-07,
-      "loss": 0.2727,
+      "epoch": 4.977835441669369,
+      "grad_norm": 0.267437607049942,
+      "learning_rate": 2.994461483690758e-09,
+      "loss": 0.367,
       "step": 138120
     },
     {
-      "epoch": 4.86,
-      "learning_rate": 1.0583540107071132e-07,
-      "loss": 0.2503,
+      "epoch": 4.978015641330594,
+      "grad_norm": 0.23144756257534027,
+      "learning_rate": 2.9494673435548346e-09,
+      "loss": 0.3605,
       "step": 138125
     },
     {
-      "epoch": 4.86,
-      "learning_rate": 1.0557371405374162e-07,
-      "loss": 0.2551,
+      "epoch": 4.978195840991819,
+      "grad_norm": 0.24928845465183258,
+      "learning_rate": 2.9048137833942446e-09,
+      "loss": 0.3923,
       "step": 138130
     },
     {
-      "epoch": 4.86,
-      "learning_rate": 1.0531235027448904e-07,
-      "loss": 0.2414,
+      "epoch": 4.9783760406530435,
+      "grad_norm": 0.2760869860649109,
+      "learning_rate": 2.860500803814059e-09,
+      "loss": 0.3978,
       "step": 138135
     },
     {
-      "epoch": 4.86,
-      "learning_rate": 1.0505130973635092e-07,
-      "loss": 0.2511,
+      "epoch": 4.978556240314268,
+      "grad_norm": 0.25941580533981323,
+      "learning_rate": 2.8165284054193498e-09,
+      "loss": 0.3736,
       "step": 138140
     },
     {
-      "epoch": 4.86,
-      "learning_rate": 1.0479059244271616e-07,
-      "loss": 0.2478,
+      "epoch": 4.978736439975493,
+      "grad_norm": 0.26913321018218994,
+      "learning_rate": 2.7728965888124124e-09,
+      "loss": 0.3906,
       "step": 138145
     },
     {
-      "epoch": 4.86,
-      "learning_rate": 1.0453019839696543e-07,
-      "loss": 0.24,
+      "epoch": 4.978916639636718,
+      "grad_norm": 0.24921806156635284,
+      "learning_rate": 2.729605354584441e-09,
+      "loss": 0.3461,
       "step": 138150
     },
     {
-      "epoch": 4.86,
-      "learning_rate": 1.0427012760248489e-07,
-      "loss": 0.2648,
+      "epoch": 4.9790968392979424,
+      "grad_norm": 0.2868900001049042,
+      "learning_rate": 2.68665470332663e-09,
+      "loss": 0.3779,
       "step": 138155
     },
     {
-      "epoch": 4.86,
-      "learning_rate": 1.040103800626524e-07,
-      "loss": 0.2579,
+      "epoch": 4.979277038959166,
+      "grad_norm": 0.2325631082057953,
+      "learning_rate": 2.644044635621845e-09,
+      "loss": 0.384,
       "step": 138160
     },
     {
-      "epoch": 4.86,
-      "learning_rate": 1.0375095578083471e-07,
-      "loss": 0.242,
+      "epoch": 4.979457238620391,
+      "grad_norm": 0.24822711944580078,
+      "learning_rate": 2.601775152052954e-09,
+      "loss": 0.3653,
       "step": 138165
     },
     {
-      "epoch": 4.86,
-      "learning_rate": 1.034918547604069e-07,
-      "loss": 0.2422,
+      "epoch": 4.979637438281616,
+      "grad_norm": 0.2686733603477478,
+      "learning_rate": 2.559846253194498e-09,
+      "loss": 0.3809,
       "step": 138170
     },
     {
-      "epoch": 4.86,
-      "learning_rate": 1.032330770047274e-07,
-      "loss": 0.2489,
+      "epoch": 4.9798176379428405,
+      "grad_norm": 0.27422237396240234,
+      "learning_rate": 2.518257939621016e-09,
+      "loss": 0.3853,
       "step": 138175
     },
     {
-      "epoch": 4.86,
-      "learning_rate": 1.029746225171574e-07,
-      "loss": 0.242,
+      "epoch": 4.979997837604065,
+      "grad_norm": 0.24748918414115906,
+      "learning_rate": 2.4770102118931714e-09,
+      "loss": 0.4023,
       "step": 138180
     },
     {
-      "epoch": 4.86,
-      "learning_rate": 1.0271649130105809e-07,
-      "loss": 0.2356,
+      "epoch": 4.98017803726529,
+      "grad_norm": 0.2908730208873749,
+      "learning_rate": 2.4361030705771782e-09,
+      "loss": 0.3696,
       "step": 138185
     },
     {
-      "epoch": 4.86,
-      "learning_rate": 1.0245868335977405e-07,
-      "loss": 0.2374,
+      "epoch": 4.980358236926515,
+      "grad_norm": 0.31870460510253906,
+      "learning_rate": 2.395536516230923e-09,
+      "loss": 0.3693,
       "step": 138190
     },
     {
-      "epoch": 4.86,
-      "learning_rate": 1.0220119869665812e-07,
-      "loss": 0.2428,
+      "epoch": 4.980538436587739,
+      "grad_norm": 0.25807613134384155,
+      "learning_rate": 2.355310549406742e-09,
+      "loss": 0.4022,
       "step": 138195
     },
     {
-      "epoch": 4.86,
-      "learning_rate": 1.0194403731504932e-07,
-      "loss": 0.2554,
+      "epoch": 4.980718636248964,
+      "grad_norm": 0.269734263420105,
+      "learning_rate": 2.315425170648644e-09,
+      "loss": 0.3582,
       "step": 138200
     },
     {
-      "epoch": 4.86,
-      "learning_rate": 1.016871992182894e-07,
-      "loss": 0.2524,
+      "epoch": 4.980898835910189,
+      "grad_norm": 0.22261251509189606,
+      "learning_rate": 2.27588038050619e-09,
+      "loss": 0.4311,
       "step": 138205
     },
     {
-      "epoch": 4.86,
-      "learning_rate": 1.014306844097146e-07,
-      "loss": 0.2522,
+      "epoch": 4.981079035571413,
+      "grad_norm": 0.25113579630851746,
+      "learning_rate": 2.2366761795150626e-09,
+      "loss": 0.3689,
       "step": 138210
     },
     {
-      "epoch": 4.86,
-      "learning_rate": 1.0117449289265001e-07,
-      "loss": 0.2532,
+      "epoch": 4.9812592352326375,
+      "grad_norm": 0.21979762613773346,
+      "learning_rate": 2.1978125682081683e-09,
+      "loss": 0.3975,
       "step": 138215
     },
     {
-      "epoch": 4.86,
-      "learning_rate": 1.0091862467042912e-07,
-      "loss": 0.2641,
+      "epoch": 4.981439434893862,
+      "grad_norm": 0.2908676564693451,
+      "learning_rate": 2.159289547115639e-09,
+      "loss": 0.3526,
       "step": 138220
     },
     {
-      "epoch": 4.86,
-      "learning_rate": 1.0066307974636868e-07,
-      "loss": 0.2501,
+      "epoch": 4.981619634555087,
+      "grad_norm": 0.2306397557258606,
+      "learning_rate": 2.121107116764831e-09,
+      "loss": 0.3521,
       "step": 138225
     },
     {
-      "epoch": 4.86,
-      "learning_rate": 1.0040785812379106e-07,
-      "loss": 0.2482,
+      "epoch": 4.981799834216312,
+      "grad_norm": 0.2513945698738098,
+      "learning_rate": 2.0832652776719974e-09,
+      "loss": 0.3938,
       "step": 138230
     },
     {
-      "epoch": 4.86,
-      "learning_rate": 1.001529598060047e-07,
-      "loss": 0.2367,
+      "epoch": 4.981980033877536,
+      "grad_norm": 0.2760458290576935,
+      "learning_rate": 2.0457640303561676e-09,
+      "loss": 0.3951,
       "step": 138235
     },
     {
-      "epoch": 4.86,
-      "learning_rate": 9.989838479632641e-08,
-      "loss": 0.2477,
+      "epoch": 4.982160233538761,
+      "grad_norm": 0.2553420066833496,
+      "learning_rate": 2.0086033753280442e-09,
+      "loss": 0.3646,
       "step": 138240
     },
     {
-      "epoch": 4.86,
-      "learning_rate": 9.964413309805353e-08,
-      "loss": 0.2497,
+      "epoch": 4.982340433199986,
+      "grad_norm": 0.25386497378349304,
+      "learning_rate": 1.971783313090003e-09,
+      "loss": 0.3633,
       "step": 138245
     },
     {
-      "epoch": 4.86,
-      "learning_rate": 9.939020471449456e-08,
-      "loss": 0.2469,
+      "epoch": 4.982520632861211,
+      "grad_norm": 0.20062953233718872,
+      "learning_rate": 1.9353038441499716e-09,
+      "loss": 0.3194,
       "step": 138250
     },
     {
-      "epoch": 4.86,
-      "learning_rate": 9.913659964894406e-08,
-      "loss": 0.2403,
+      "epoch": 4.9827008325224345,
+      "grad_norm": 0.26775026321411133,
+      "learning_rate": 1.899164968999223e-09,
+      "loss": 0.4075,
       "step": 138255
     },
     {
-      "epoch": 4.86,
-      "learning_rate": 9.888331790469107e-08,
-      "loss": 0.2633,
+      "epoch": 4.982881032183659,
+      "grad_norm": 0.273628830909729,
+      "learning_rate": 1.8633666881345823e-09,
+      "loss": 0.3616,
       "step": 138260
     },
     {
-      "epoch": 4.86,
-      "learning_rate": 9.863035948503019e-08,
-      "loss": 0.255,
+      "epoch": 4.983061231844884,
+      "grad_norm": 0.2722027003765106,
+      "learning_rate": 1.8279090020417721e-09,
+      "loss": 0.3612,
       "step": 138265
     },
     {
-      "epoch": 4.86,
-      "learning_rate": 9.837772439324488e-08,
-      "loss": 0.2596,
+      "epoch": 4.983241431506109,
+      "grad_norm": 0.2538944184780121,
+      "learning_rate": 1.7927919112009639e-09,
+      "loss": 0.3556,
       "step": 138270
     },
     {
-      "epoch": 4.86,
-      "learning_rate": 9.812541263261033e-08,
-      "loss": 0.2755,
+      "epoch": 4.983421631167333,
+      "grad_norm": 0.264728844165802,
+      "learning_rate": 1.7580154160951047e-09,
+      "loss": 0.386,
       "step": 138275
     },
     {
-      "epoch": 4.87,
-      "learning_rate": 9.787342420640999e-08,
-      "loss": 0.2565,
+      "epoch": 4.983601830828558,
+      "grad_norm": 0.2643708288669586,
+      "learning_rate": 1.7235795171988146e-09,
+      "loss": 0.3676,
       "step": 138280
     },
     {
-      "epoch": 4.87,
-      "learning_rate": 9.762175911791072e-08,
-      "loss": 0.2332,
+      "epoch": 4.983782030489783,
+      "grad_norm": 0.2401455044746399,
+      "learning_rate": 1.689484214975612e-09,
+      "loss": 0.3727,
       "step": 138285
     },
     {
-      "epoch": 4.87,
-      "learning_rate": 9.73704173703821e-08,
-      "loss": 0.2405,
+      "epoch": 4.983962230151008,
+      "grad_norm": 0.2023637741804123,
+      "learning_rate": 1.6557295098945658e-09,
+      "loss": 0.4015,
       "step": 138290
     },
     {
-      "epoch": 4.87,
-      "learning_rate": 9.711939896708544e-08,
-      "loss": 0.2411,
+      "epoch": 4.984142429812232,
+      "grad_norm": 0.26757675409317017,
+      "learning_rate": 1.622315402416419e-09,
+      "loss": 0.3644,
       "step": 138295
     },
     {
-      "epoch": 4.87,
-      "learning_rate": 9.686870391128478e-08,
-      "loss": 0.233,
+      "epoch": 4.984322629473457,
+      "grad_norm": 0.22083128988742828,
+      "learning_rate": 1.5892418929908114e-09,
+      "loss": 0.3287,
       "step": 138300
     },
     {
-      "epoch": 4.87,
-      "learning_rate": 9.661833220623029e-08,
-      "loss": 0.2342,
+      "epoch": 4.984502829134681,
+      "grad_norm": 0.3092709183692932,
+      "learning_rate": 1.5565089820729351e-09,
+      "loss": 0.3711,
       "step": 138305
     },
     {
-      "epoch": 4.87,
-      "learning_rate": 9.636828385517493e-08,
-      "loss": 0.2569,
+      "epoch": 4.984683028795906,
+      "grad_norm": 0.22019417583942413,
+      "learning_rate": 1.5241166701096543e-09,
+      "loss": 0.3608,
       "step": 138310
     },
     {
-      "epoch": 4.87,
-      "learning_rate": 9.611855886136334e-08,
-      "loss": 0.2464,
+      "epoch": 4.98486322845713,
+      "grad_norm": 0.24744448065757751,
+      "learning_rate": 1.4920649575395073e-09,
+      "loss": 0.3614,
       "step": 138315
     },
     {
-      "epoch": 4.87,
-      "learning_rate": 9.58691572280429e-08,
-      "loss": 0.2324,
+      "epoch": 4.985043428118355,
+      "grad_norm": 0.17398010194301605,
+      "learning_rate": 1.4603538448010323e-09,
+      "loss": 0.3591,
       "step": 138320
     },
     {
-      "epoch": 4.87,
-      "learning_rate": 9.562007895844716e-08,
-      "loss": 0.261,
+      "epoch": 4.98522362777958,
+      "grad_norm": 0.22615504264831543,
+      "learning_rate": 1.428983332321665e-09,
+      "loss": 0.363,
       "step": 138325
     },
     {
-      "epoch": 4.87,
-      "learning_rate": 9.537132405581239e-08,
-      "loss": 0.243,
+      "epoch": 4.985403827440805,
+      "grad_norm": 0.25961410999298096,
+      "learning_rate": 1.397953420537168e-09,
+      "loss": 0.3894,
       "step": 138330
     },
     {
-      "epoch": 4.87,
-      "learning_rate": 9.512289252336937e-08,
-      "loss": 0.2475,
+      "epoch": 4.985584027102029,
+      "grad_norm": 0.27555739879608154,
+      "learning_rate": 1.367264109863875e-09,
+      "loss": 0.3741,
       "step": 138335
     },
     {
-      "epoch": 4.87,
-      "learning_rate": 9.487478436434327e-08,
-      "loss": 0.2221,
+      "epoch": 4.985764226763254,
+      "grad_norm": 0.3149866759777069,
+      "learning_rate": 1.336915400720895e-09,
+      "loss": 0.4,
       "step": 138340
     },
     {
-      "epoch": 4.87,
-      "learning_rate": 9.462699958195376e-08,
-      "loss": 0.2452,
+      "epoch": 4.985944426424478,
+      "grad_norm": 0.2356063425540924,
+      "learning_rate": 1.3069072935217863e-09,
+      "loss": 0.3838,
       "step": 138345
     },
     {
-      "epoch": 4.87,
-      "learning_rate": 9.437953817941769e-08,
-      "loss": 0.2392,
+      "epoch": 4.986124626085703,
+      "grad_norm": 0.28220483660697937,
+      "learning_rate": 1.2772397886773313e-09,
+      "loss": 0.3788,
       "step": 138350
     },
     {
-      "epoch": 4.87,
-      "learning_rate": 9.413240015995195e-08,
-      "loss": 0.2473,
+      "epoch": 4.986304825746927,
+      "grad_norm": 0.21516813337802887,
+      "learning_rate": 1.247912886589986e-09,
+      "loss": 0.3715,
       "step": 138355
     },
     {
-      "epoch": 4.87,
-      "learning_rate": 9.388558552676508e-08,
-      "loss": 0.2585,
+      "epoch": 4.986485025408152,
+      "grad_norm": 0.2194688469171524,
+      "learning_rate": 1.2189265876622058e-09,
+      "loss": 0.3802,
       "step": 138360
     },
     {
-      "epoch": 4.87,
-      "learning_rate": 9.363909428305728e-08,
-      "loss": 0.2457,
+      "epoch": 4.986665225069377,
+      "grad_norm": 0.26937180757522583,
+      "learning_rate": 1.1902808922853447e-09,
+      "loss": 0.3905,
       "step": 138365
     },
     {
-      "epoch": 4.87,
-      "learning_rate": 9.339292643203157e-08,
-      "loss": 0.2499,
+      "epoch": 4.9868454247306016,
+      "grad_norm": 0.2718323767185211,
+      "learning_rate": 1.1619758008507564e-09,
+      "loss": 0.3827,
       "step": 138370
     },
     {
-      "epoch": 4.87,
-      "learning_rate": 9.314708197688537e-08,
-      "loss": 0.2475,
+      "epoch": 4.987025624391826,
+      "grad_norm": 0.31076350808143616,
+      "learning_rate": 1.1340113137442431e-09,
+      "loss": 0.3789,
       "step": 138375
     },
     {
-      "epoch": 4.87,
-      "learning_rate": 9.290156092081059e-08,
-      "loss": 0.2387,
+      "epoch": 4.987205824053051,
+      "grad_norm": 0.26941627264022827,
+      "learning_rate": 1.1063874313460564e-09,
+      "loss": 0.3854,
       "step": 138380
     },
     {
-      "epoch": 4.87,
-      "learning_rate": 9.265636326699634e-08,
-      "loss": 0.2408,
+      "epoch": 4.987386023714276,
+      "grad_norm": 0.2722313702106476,
+      "learning_rate": 1.0791041540336722e-09,
+      "loss": 0.3723,
       "step": 138385
     },
     {
-      "epoch": 4.87,
-      "learning_rate": 9.24114890186234e-08,
-      "loss": 0.2603,
+      "epoch": 4.9875662233755005,
+      "grad_norm": 0.2729116976261139,
+      "learning_rate": 1.0521614821790148e-09,
+      "loss": 0.3595,
       "step": 138390
     },
     {
-      "epoch": 4.87,
-      "learning_rate": 9.21669381788698e-08,
-      "loss": 0.2541,
+      "epoch": 4.987746423036724,
+      "grad_norm": 0.25742781162261963,
+      "learning_rate": 1.0255594161484583e-09,
+      "loss": 0.3608,
       "step": 138395
     },
     {
-      "epoch": 4.87,
-      "learning_rate": 9.19227107509163e-08,
-      "loss": 0.2484,
+      "epoch": 4.987926622697949,
+      "grad_norm": 0.27901482582092285,
+      "learning_rate": 9.992979563056003e-10,
+      "loss": 0.3677,
       "step": 138400
     },
     {
-      "epoch": 4.87,
-      "learning_rate": 9.167880673793261e-08,
-      "loss": 0.2568,
+      "epoch": 4.988106822359174,
+      "grad_norm": 0.20445889234542847,
+      "learning_rate": 9.785340251255993e-10,
+      "loss": 0.3751,
       "step": 138405
     },
     {
-      "epoch": 4.87,
-      "learning_rate": 9.143522614308287e-08,
-      "loss": 0.251,
+      "epoch": 4.9882870220203985,
+      "grad_norm": 0.2854103446006775,
+      "learning_rate": 9.528856573182144e-10,
+      "loss": 0.3522,
       "step": 138410
     },
     {
-      "epoch": 4.87,
-      "learning_rate": 9.11919689695312e-08,
-      "loss": 0.2694,
+      "epoch": 4.988467221681623,
+      "grad_norm": 0.25394073128700256,
+      "learning_rate": 9.275778966866267e-10,
+      "loss": 0.3537,
       "step": 138415
     },
     {
-      "epoch": 4.87,
-      "learning_rate": 9.094903522043618e-08,
-      "loss": 0.234,
+      "epoch": 4.988647421342848,
+      "grad_norm": 0.24267399311065674,
+      "learning_rate": 9.026107435750053e-10,
+      "loss": 0.3617,
       "step": 138420
     },
     {
-      "epoch": 4.87,
-      "learning_rate": 9.070642489895087e-08,
-      "loss": 0.2431,
+      "epoch": 4.988827621004073,
+      "grad_norm": 0.21222999691963196,
+      "learning_rate": 8.779841983275195e-10,
+      "loss": 0.3673,
       "step": 138425
     },
     {
-      "epoch": 4.87,
-      "learning_rate": 9.046413800822828e-08,
-      "loss": 0.257,
+      "epoch": 4.9890078206652975,
+      "grad_norm": 0.24381837248802185,
+      "learning_rate": 8.53698261277236e-10,
+      "loss": 0.376,
       "step": 138430
     },
     {
-      "epoch": 4.87,
-      "learning_rate": 9.022217455141314e-08,
-      "loss": 0.2599,
+      "epoch": 4.989188020326521,
+      "grad_norm": 0.3303523659706116,
+      "learning_rate": 8.297529327544462e-10,
+      "loss": 0.3671,
       "step": 138435
     },
     {
-      "epoch": 4.87,
-      "learning_rate": 8.998053453164734e-08,
-      "loss": 0.2709,
+      "epoch": 4.989368219987746,
+      "grad_norm": 0.21352322399616241,
+      "learning_rate": 8.061482130894416e-10,
+      "loss": 0.3689,
       "step": 138440
     },
     {
-      "epoch": 4.87,
-      "learning_rate": 8.97392179520673e-08,
-      "loss": 0.2447,
+      "epoch": 4.989548419648971,
+      "grad_norm": 0.21397538483142853,
+      "learning_rate": 7.8288410259586e-10,
+      "loss": 0.3615,
       "step": 138445
     },
     {
-      "epoch": 4.87,
-      "learning_rate": 8.94982248158066e-08,
-      "loss": 0.2296,
+      "epoch": 4.9897286193101955,
+      "grad_norm": 0.25124630331993103,
+      "learning_rate": 7.599606015984418e-10,
+      "loss": 0.3633,
       "step": 138450
     },
     {
-      "epoch": 4.87,
-      "learning_rate": 8.925755512599609e-08,
-      "loss": 0.2401,
+      "epoch": 4.98990881897142,
+      "grad_norm": 0.31034645438194275,
+      "learning_rate": 7.373777104052737e-10,
+      "loss": 0.3662,
       "step": 138455
     },
     {
-      "epoch": 4.87,
-      "learning_rate": 8.901720888575826e-08,
-      "loss": 0.2526,
+      "epoch": 4.990089018632645,
+      "grad_norm": 0.23552536964416504,
+      "learning_rate": 7.151354293272183e-10,
+      "loss": 0.3589,
       "step": 138460
     },
     {
-      "epoch": 4.87,
-      "learning_rate": 8.877718609821562e-08,
-      "loss": 0.2359,
+      "epoch": 4.99026921829387,
+      "grad_norm": 0.23122437298297882,
+      "learning_rate": 6.932337586640359e-10,
+      "loss": 0.3852,
       "step": 138465
     },
     {
-      "epoch": 4.87,
-      "learning_rate": 8.85374867664851e-08,
-      "loss": 0.2719,
+      "epoch": 4.9904494179550944,
+      "grad_norm": 0.25336378812789917,
+      "learning_rate": 6.716726987154865e-10,
+      "loss": 0.3705,
       "step": 138470
     },
     {
-      "epoch": 4.87,
-      "learning_rate": 8.829811089367811e-08,
-      "loss": 0.2384,
+      "epoch": 4.990629617616319,
+      "grad_norm": 0.2418462485074997,
+      "learning_rate": 6.504522497757792e-10,
+      "loss": 0.395,
       "step": 138475
     },
     {
-      "epoch": 4.87,
-      "learning_rate": 8.805905848290053e-08,
-      "loss": 0.2477,
+      "epoch": 4.990809817277544,
+      "grad_norm": 0.23543506860733032,
+      "learning_rate": 6.295724121335722e-10,
+      "loss": 0.352,
       "step": 138480
     },
     {
-      "epoch": 4.87,
-      "learning_rate": 8.782032953725817e-08,
-      "loss": 0.2318,
+      "epoch": 4.990990016938769,
+      "grad_norm": 0.28832608461380005,
+      "learning_rate": 6.090331860719722e-10,
+      "loss": 0.3692,
       "step": 138485
     },
     {
-      "epoch": 4.87,
-      "learning_rate": 8.758192405985133e-08,
-      "loss": 0.2362,
+      "epoch": 4.9911702165999925,
+      "grad_norm": 0.21373923122882843,
+      "learning_rate": 5.888345718740862e-10,
+      "loss": 0.373,
       "step": 138490
     },
     {
-      "epoch": 4.87,
-      "learning_rate": 8.734384205377755e-08,
-      "loss": 0.2613,
+      "epoch": 4.991350416261217,
+      "grad_norm": 0.28929153084754944,
+      "learning_rate": 5.689765698119187e-10,
+      "loss": 0.4071,
       "step": 138495
     },
     {
-      "epoch": 4.87,
-      "learning_rate": 8.710608352212324e-08,
-      "loss": 0.2318,
+      "epoch": 4.991530615922442,
+      "grad_norm": 0.226441890001297,
+      "learning_rate": 5.4945918016025e-10,
+      "loss": 0.398,
       "step": 138500
     },
     {
-      "epoch": 4.87,
-      "eval_loss": 0.24843639135360718,
-      "eval_runtime": 10.5531,
-      "eval_samples_per_second": 9.476,
-      "eval_steps_per_second": 9.476,
+      "epoch": 4.991530615922442,
+      "eval_loss": 0.428780198097229,
+      "eval_runtime": 3.5301,
+      "eval_samples_per_second": 28.328,
+      "eval_steps_per_second": 7.082,
       "step": 138500
-    },
-    {
-      "epoch": 4.87,
-      "learning_rate": 8.686864846798037e-08,
-      "loss": 0.2708,
-      "step": 138505
-    },
-    {
-      "epoch": 4.87,
-      "learning_rate": 8.663153689442982e-08,
-      "loss": 0.2577,
-      "step": 138510
-    },
-    {
-      "epoch": 4.87,
-      "learning_rate": 8.639474880454968e-08,
-      "loss": 0.2459,
-      "step": 138515
-    },
-    {
-      "epoch": 4.87,
-      "learning_rate": 8.615828420141526e-08,
-      "loss": 0.2296,
-      "step": 138520
-    },
-    {
-      "epoch": 4.87,
-      "learning_rate": 8.592214308809632e-08,
-      "loss": 0.2533,
-      "step": 138525
-    },
-    {
-      "epoch": 4.87,
-      "learning_rate": 8.568632546765987e-08,
-      "loss": 0.2483,
-      "step": 138530
-    },
-    {
-      "epoch": 4.87,
-      "learning_rate": 8.545083134316456e-08,
-      "loss": 0.2506,
-      "step": 138535
-    },
-    {
-      "epoch": 4.87,
-      "learning_rate": 8.521566071767462e-08,
-      "loss": 0.2508,
-      "step": 138540
-    },
-    {
-      "epoch": 4.87,
-      "learning_rate": 8.49808135942376e-08,
-      "loss": 0.2545,
-      "step": 138545
-    },
-    {
-      "epoch": 4.87,
-      "learning_rate": 8.474628997590661e-08,
-      "loss": 0.2407,
-      "step": 138550
-    },
-    {
-      "epoch": 4.87,
-      "learning_rate": 8.451208986572367e-08,
-      "loss": 0.2555,
-      "step": 138555
-    },
-    {
-      "epoch": 4.87,
-      "learning_rate": 8.42782132667308e-08,
-      "loss": 0.258,
-      "step": 138560
-    },
-    {
-      "epoch": 4.88,
-      "learning_rate": 8.404466018196722e-08,
-      "loss": 0.2316,
-      "step": 138565
-    },
-    {
-      "epoch": 4.88,
-      "learning_rate": 8.381143061446383e-08,
-      "loss": 0.2609,
-      "step": 138570
-    },
-    {
-      "epoch": 4.88,
-      "learning_rate": 8.35785245672488e-08,
-      "loss": 0.2565,
-      "step": 138575
-    },
-    {
-      "epoch": 4.88,
-      "learning_rate": 8.334594204334467e-08,
-      "loss": 0.2244,
-      "step": 138580
-    },
-    {
-      "epoch": 4.88,
-      "learning_rate": 8.311368304577128e-08,
-      "loss": 0.2621,
-      "step": 138585
-    },
-    {
-      "epoch": 4.88,
-      "learning_rate": 8.28817475775484e-08,
-      "loss": 0.2418,
-      "step": 138590
-    },
-    {
-      "epoch": 4.88,
-      "learning_rate": 8.265013564168478e-08,
-      "loss": 0.2513,
-      "step": 138595
-    },
-    {
-      "epoch": 4.88,
-      "learning_rate": 8.241884724118631e-08,
-      "loss": 0.2521,
-      "step": 138600
-    },
-    {
-      "epoch": 4.88,
-      "learning_rate": 8.218788237905895e-08,
-      "loss": 0.2319,
-      "step": 138605
-    },
-    {
-      "epoch": 4.88,
-      "learning_rate": 8.195724105830305e-08,
-      "loss": 0.2504,
-      "step": 138610
-    },
-    {
-      "epoch": 4.88,
-      "learning_rate": 8.172692328190512e-08,
-      "loss": 0.2492,
-      "step": 138615
-    },
-    {
-      "epoch": 4.88,
-      "learning_rate": 8.149692905286555e-08,
-      "loss": 0.2471,
-      "step": 138620
-    },
-    {
-      "epoch": 4.88,
-      "learning_rate": 8.126725837416527e-08,
-      "loss": 0.2494,
-      "step": 138625
-    },
-    {
-      "epoch": 4.88,
-      "learning_rate": 8.103791124878801e-08,
-      "loss": 0.2642,
-      "step": 138630
-    },
-    {
-      "epoch": 4.88,
-      "learning_rate": 8.080888767971196e-08,
-      "loss": 0.2405,
-      "step": 138635
-    },
-    {
-      "epoch": 4.88,
-      "learning_rate": 8.058018766990972e-08,
-      "loss": 0.2446,
-      "step": 138640
-    },
-    {
-      "epoch": 4.88,
-      "learning_rate": 8.035181122234837e-08,
-      "loss": 0.244,
-      "step": 138645
-    },
-    {
-      "epoch": 4.88,
-      "learning_rate": 8.012375834000053e-08,
-      "loss": 0.261,
-      "step": 138650
-    },
-    {
-      "epoch": 4.88,
-      "learning_rate": 7.989602902581938e-08,
-      "loss": 0.2475,
-      "step": 138655
-    },
-    {
-      "epoch": 4.88,
-      "learning_rate": 7.966862328276647e-08,
-      "loss": 0.2544,
-      "step": 138660
-    },
-    {
-      "epoch": 4.88,
-      "learning_rate": 7.944154111379498e-08,
-      "loss": 0.2701,
-      "step": 138665
-    },
-    {
-      "epoch": 4.88,
-      "learning_rate": 7.921478252184977e-08,
-      "loss": 0.2404,
-      "step": 138670
-    },
-    {
-      "epoch": 4.88,
-      "learning_rate": 7.898834750987849e-08,
-      "loss": 0.2496,
-      "step": 138675
-    },
-    {
-      "epoch": 4.88,
-      "learning_rate": 7.876223608081767e-08,
-      "loss": 0.2536,
-      "step": 138680
-    },
-    {
-      "epoch": 4.88,
-      "learning_rate": 7.853644823760941e-08,
-      "loss": 0.2337,
-      "step": 138685
-    },
-    {
-      "epoch": 4.88,
-      "learning_rate": 7.831098398317916e-08,
-      "loss": 0.2566,
-      "step": 138690
-    },
-    {
-      "epoch": 4.88,
-      "learning_rate": 7.80858433204551e-08,
-      "loss": 0.2438,
-      "step": 138695
-    },
-    {
-      "epoch": 4.88,
-      "learning_rate": 7.786102625236269e-08,
-      "loss": 0.2492,
-      "step": 138700
-    },
-    {
-      "epoch": 4.88,
-      "learning_rate": 7.763653278182182e-08,
-      "loss": 0.2476,
-      "step": 138705
-    },
-    {
-      "epoch": 4.88,
-      "learning_rate": 7.74123629117468e-08,
-      "loss": 0.2509,
-      "step": 138710
-    },
-    {
-      "epoch": 4.88,
-      "learning_rate": 7.718851664504645e-08,
-      "loss": 0.2566,
-      "step": 138715
-    },
-    {
-      "epoch": 4.88,
-      "learning_rate": 7.696499398462953e-08,
-      "loss": 0.2335,
-      "step": 138720
-    },
-    {
-      "epoch": 4.88,
-      "learning_rate": 7.67417949333965e-08,
-      "loss": 0.2551,
-      "step": 138725
-    },
-    {
-      "epoch": 4.88,
-      "learning_rate": 7.651891949424784e-08,
-      "loss": 0.2579,
-      "step": 138730
-    },
-    {
-      "epoch": 4.88,
-      "learning_rate": 7.629636767007287e-08,
-      "loss": 0.2632,
-      "step": 138735
-    },
-    {
-      "epoch": 4.88,
-      "learning_rate": 7.607413946376651e-08,
-      "loss": 0.2468,
-      "step": 138740
-    },
-    {
-      "epoch": 4.88,
-      "learning_rate": 7.585223487821259e-08,
-      "loss": 0.2377,
-      "step": 138745
-    },
-    {
-      "epoch": 4.88,
-      "learning_rate": 7.563065391628932e-08,
-      "loss": 0.2611,
-      "step": 138750
-    },
-    {
-      "epoch": 4.88,
-      "learning_rate": 7.540939658087776e-08,
-      "loss": 0.2642,
-      "step": 138755
-    },
-    {
-      "epoch": 4.88,
-      "learning_rate": 7.518846287484782e-08,
-      "loss": 0.2432,
-      "step": 138760
-    },
-    {
-      "epoch": 4.88,
-      "learning_rate": 7.496785280107222e-08,
-      "loss": 0.2575,
-      "step": 138765
-    },
-    {
-      "epoch": 4.88,
-      "learning_rate": 7.474756636240976e-08,
-      "loss": 0.2462,
-      "step": 138770
-    },
-    {
-      "epoch": 4.88,
-      "learning_rate": 7.452760356172483e-08,
-      "loss": 0.2464,
-      "step": 138775
-    },
-    {
-      "epoch": 4.88,
-      "learning_rate": 7.430796440187349e-08,
-      "loss": 0.2735,
-      "step": 138780
-    },
-    {
-      "epoch": 4.88,
-      "learning_rate": 7.408864888570621e-08,
-      "loss": 0.2807,
-      "step": 138785
-    },
-    {
-      "epoch": 4.88,
-      "learning_rate": 7.386965701607074e-08,
-      "loss": 0.2442,
-      "step": 138790
-    },
-    {
-      "epoch": 4.88,
-      "learning_rate": 7.3650988795812e-08,
-      "loss": 0.2407,
-      "step": 138795
-    },
-    {
-      "epoch": 4.88,
-      "learning_rate": 7.343264422776663e-08,
-      "loss": 0.2234,
-      "step": 138800
-    },
-    {
-      "epoch": 4.88,
-      "learning_rate": 7.321462331477124e-08,
-      "loss": 0.2519,
-      "step": 138805
-    },
-    {
-      "epoch": 4.88,
-      "learning_rate": 7.299692605965691e-08,
-      "loss": 0.2525,
-      "step": 138810
-    },
-    {
-      "epoch": 4.88,
-      "learning_rate": 7.277955246524915e-08,
-      "loss": 0.2468,
-      "step": 138815
-    },
-    {
-      "epoch": 4.88,
-      "learning_rate": 7.256250253437347e-08,
-      "loss": 0.2615,
-      "step": 138820
-    },
-    {
-      "epoch": 4.88,
-      "learning_rate": 7.234577626984151e-08,
-      "loss": 0.2465,
-      "step": 138825
-    },
-    {
-      "epoch": 4.88,
-      "learning_rate": 7.212937367447603e-08,
-      "loss": 0.2319,
-      "step": 138830
-    },
-    {
-      "epoch": 4.88,
-      "learning_rate": 7.191329475108032e-08,
-      "loss": 0.2629,
-      "step": 138835
-    },
-    {
-      "epoch": 4.88,
-      "learning_rate": 7.169753950246605e-08,
-      "loss": 0.2428,
-      "step": 138840
-    },
-    {
-      "epoch": 4.88,
-      "learning_rate": 7.14821079314254e-08,
-      "loss": 0.2525,
-      "step": 138845
-    },
-    {
-      "epoch": 4.89,
-      "learning_rate": 7.126700004076448e-08,
-      "loss": 0.234,
-      "step": 138850
-    },
-    {
-      "epoch": 4.89,
-      "learning_rate": 7.10522158332727e-08,
-      "loss": 0.2653,
-      "step": 138855
-    },
-    {
-      "epoch": 4.89,
-      "learning_rate": 7.083775531173675e-08,
-      "loss": 0.2505,
-      "step": 138860
-    },
-    {
-      "epoch": 4.89,
-      "learning_rate": 7.062361847894605e-08,
-      "loss": 0.2414,
-      "step": 138865
-    },
-    {
-      "epoch": 4.89,
-      "learning_rate": 7.040980533767894e-08,
-      "loss": 0.2582,
-      "step": 138870
-    },
-    {
-      "epoch": 4.89,
-      "learning_rate": 7.019631589070818e-08,
-      "loss": 0.2623,
-      "step": 138875
-    },
-    {
-      "epoch": 4.89,
-      "learning_rate": 6.998315014081213e-08,
-      "loss": 0.2364,
-      "step": 138880
-    },
-    {
-      "epoch": 4.89,
-      "learning_rate": 6.977030809075247e-08,
-      "loss": 0.2384,
-      "step": 138885
-    },
-    {
-      "epoch": 4.89,
-      "learning_rate": 6.955778974329919e-08,
-      "loss": 0.2242,
-      "step": 138890
-    },
-    {
-      "epoch": 4.89,
-      "learning_rate": 6.934559510120564e-08,
-      "loss": 0.235,
-      "step": 138895
-    },
-    {
-      "epoch": 4.89,
-      "learning_rate": 6.913372416723074e-08,
-      "loss": 0.2355,
-      "step": 138900
-    },
-    {
-      "epoch": 4.89,
-      "learning_rate": 6.892217694412505e-08,
-      "loss": 0.2448,
-      "step": 138905
-    },
-    {
-      "epoch": 4.89,
-      "learning_rate": 6.871095343463362e-08,
-      "loss": 0.2716,
-      "step": 138910
-    },
-    {
-      "epoch": 4.89,
-      "learning_rate": 6.850005364150147e-08,
-      "loss": 0.2524,
-      "step": 138915
-    },
-    {
-      "epoch": 4.89,
-      "learning_rate": 6.828947756746252e-08,
-      "loss": 0.2654,
-      "step": 138920
-    },
-    {
-      "epoch": 4.89,
-      "learning_rate": 6.807922521525622e-08,
-      "loss": 0.2407,
-      "step": 138925
-    },
-    {
-      "epoch": 4.89,
-      "learning_rate": 6.786929658761099e-08,
-      "loss": 0.2612,
-      "step": 138930
-    },
-    {
-      "epoch": 4.89,
-      "learning_rate": 6.765969168724962e-08,
-      "loss": 0.2588,
-      "step": 138935
-    },
-    {
-      "epoch": 4.89,
-      "learning_rate": 6.745041051689772e-08,
-      "loss": 0.2408,
-      "step": 138940
-    },
-    {
-      "epoch": 4.89,
-      "learning_rate": 6.724145307927254e-08,
-      "loss": 0.2578,
-      "step": 138945
-    },
-    {
-      "epoch": 4.89,
-      "learning_rate": 6.703281937708306e-08,
-      "loss": 0.2552,
-      "step": 138950
-    },
-    {
-      "epoch": 4.89,
-      "learning_rate": 6.682450941304374e-08,
-      "loss": 0.2701,
-      "step": 138955
-    },
-    {
-      "epoch": 4.89,
-      "learning_rate": 6.661652318985523e-08,
-      "loss": 0.2682,
-      "step": 138960
-    },
-    {
-      "epoch": 4.89,
-      "learning_rate": 6.640886071021813e-08,
-      "loss": 0.2446,
-      "step": 138965
-    },
-    {
-      "epoch": 4.89,
-      "learning_rate": 6.620152197683305e-08,
-      "loss": 0.2478,
-      "step": 138970
-    },
-    {
-      "epoch": 4.89,
-      "learning_rate": 6.599450699238674e-08,
-      "loss": 0.241,
-      "step": 138975
-    },
-    {
-      "epoch": 4.89,
-      "learning_rate": 6.578781575956872e-08,
-      "loss": 0.2433,
-      "step": 138980
-    },
-    {
-      "epoch": 4.89,
-      "learning_rate": 6.558144828106572e-08,
-      "loss": 0.2669,
-      "step": 138985
-    },
-    {
-      "epoch": 4.89,
-      "learning_rate": 6.537540455955616e-08,
-      "loss": 0.2576,
-      "step": 138990
-    },
-    {
-      "epoch": 4.89,
-      "learning_rate": 6.516968459771289e-08,
-      "loss": 0.2703,
-      "step": 138995
-    },
-    {
-      "epoch": 4.89,
-      "learning_rate": 6.496428839820878e-08,
-      "loss": 0.2401,
-      "step": 139000
-    },
-    {
-      "epoch": 4.89,
-      "eval_loss": 0.24838781356811523,
-      "eval_runtime": 10.5455,
-      "eval_samples_per_second": 9.483,
-      "eval_steps_per_second": 9.483,
-      "step": 139000
-    },
-    {
-      "epoch": 4.89,
-      "learning_rate": 6.475921596370838e-08,
-      "loss": 0.2372,
-      "step": 139005
-    },
-    {
-      "epoch": 4.89,
-      "learning_rate": 6.455446729688175e-08,
-      "loss": 0.2433,
-      "step": 139010
-    },
-    {
-      "epoch": 4.89,
-      "learning_rate": 6.435004240037957e-08,
-      "loss": 0.2662,
-      "step": 139015
-    },
-    {
-      "epoch": 4.89,
-      "learning_rate": 6.414594127685802e-08,
-      "loss": 0.282,
-      "step": 139020
-    },
-    {
-      "epoch": 4.89,
-      "learning_rate": 6.39421639289678e-08,
-      "loss": 0.2483,
-      "step": 139025
-    },
-    {
-      "epoch": 4.89,
-      "learning_rate": 6.373871035935675e-08,
-      "loss": 0.2787,
-      "step": 139030
-    },
-    {
-      "epoch": 4.89,
-      "learning_rate": 6.353558057066445e-08,
-      "loss": 0.2755,
-      "step": 139035
-    },
-    {
-      "epoch": 4.89,
-      "learning_rate": 6.333277456552766e-08,
-      "loss": 0.2572,
-      "step": 139040
-    },
-    {
-      "epoch": 4.89,
-      "learning_rate": 6.313029234658042e-08,
-      "loss": 0.242,
-      "step": 139045
-    },
-    {
-      "epoch": 4.89,
-      "learning_rate": 6.292813391645391e-08,
-      "loss": 0.2458,
-      "step": 139050
-    },
-    {
-      "epoch": 4.89,
-      "learning_rate": 6.272629927777108e-08,
-      "loss": 0.243,
-      "step": 139055
-    },
-    {
-      "epoch": 4.89,
-      "learning_rate": 6.252478843315479e-08,
-      "loss": 0.2483,
-      "step": 139060
-    },
-    {
-      "epoch": 4.89,
-      "learning_rate": 6.232360138521686e-08,
-      "loss": 0.2451,
-      "step": 139065
-    },
-    {
-      "epoch": 4.89,
-      "learning_rate": 6.212273813657465e-08,
-      "loss": 0.2657,
-      "step": 139070
-    },
-    {
-      "epoch": 4.89,
-      "learning_rate": 6.19221986898344e-08,
-      "loss": 0.2732,
-      "step": 139075
-    },
-    {
-      "epoch": 4.89,
-      "learning_rate": 6.17219830475968e-08,
-      "loss": 0.2429,
-      "step": 139080
-    },
-    {
-      "epoch": 4.89,
-      "learning_rate": 6.152209121246532e-08,
-      "loss": 0.2576,
-      "step": 139085
-    },
-    {
-      "epoch": 4.89,
-      "learning_rate": 6.132252318703513e-08,
-      "loss": 0.2451,
-      "step": 139090
-    },
-    {
-      "epoch": 4.89,
-      "learning_rate": 6.11232789738958e-08,
-      "loss": 0.244,
-      "step": 139095
-    },
-    {
-      "epoch": 4.89,
-      "learning_rate": 6.092435857563695e-08,
-      "loss": 0.2442,
-      "step": 139100
-    },
-    {
-      "epoch": 4.89,
-      "learning_rate": 6.072576199483982e-08,
-      "loss": 0.2445,
-      "step": 139105
-    },
-    {
-      "epoch": 4.89,
-      "learning_rate": 6.052748923408291e-08,
-      "loss": 0.249,
-      "step": 139110
-    },
-    {
-      "epoch": 4.89,
-      "learning_rate": 6.032954029593917e-08,
-      "loss": 0.2412,
-      "step": 139115
-    },
-    {
-      "epoch": 4.89,
-      "learning_rate": 6.013191518298156e-08,
-      "loss": 0.2595,
-      "step": 139120
-    },
-    {
-      "epoch": 4.89,
-      "learning_rate": 5.993461389777467e-08,
-      "loss": 0.2366,
-      "step": 139125
-    },
-    {
-      "epoch": 4.89,
-      "learning_rate": 5.973763644288033e-08,
-      "loss": 0.2396,
-      "step": 139130
-    },
-    {
-      "epoch": 4.9,
-      "learning_rate": 5.954098282085763e-08,
-      "loss": 0.2491,
-      "step": 139135
-    },
-    {
-      "epoch": 4.9,
-      "learning_rate": 5.934465303425729e-08,
-      "loss": 0.2339,
-      "step": 139140
-    },
-    {
-      "epoch": 4.9,
-      "learning_rate": 5.914864708563283e-08,
-      "loss": 0.2546,
-      "step": 139145
-    },
-    {
-      "epoch": 4.9,
-      "learning_rate": 5.895296497752389e-08,
-      "loss": 0.2587,
-      "step": 139150
-    },
-    {
-      "epoch": 4.9,
-      "learning_rate": 5.875760671247565e-08,
-      "loss": 0.252,
-      "step": 139155
-    },
-    {
-      "epoch": 4.9,
-      "learning_rate": 5.85625722930222e-08,
-      "loss": 0.2302,
-      "step": 139160
-    },
-    {
-      "epoch": 4.9,
-      "learning_rate": 5.8367861721697617e-08,
-      "loss": 0.2826,
-      "step": 139165
-    },
-    {
-      "epoch": 4.9,
-      "learning_rate": 5.817347500102765e-08,
-      "loss": 0.2645,
-      "step": 139170
-    },
-    {
-      "epoch": 4.9,
-      "learning_rate": 5.797941213353808e-08,
-      "loss": 0.2565,
-      "step": 139175
-    },
-    {
-      "epoch": 4.9,
-      "learning_rate": 5.77856731217491e-08,
-      "loss": 0.2418,
-      "step": 139180
-    },
-    {
-      "epoch": 4.9,
-      "learning_rate": 5.759225796817536e-08,
-      "loss": 0.25,
-      "step": 139185
-    },
-    {
-      "epoch": 4.9,
-      "learning_rate": 5.7399166675328745e-08,
-      "loss": 0.258,
-      "step": 139190
-    },
-    {
-      "epoch": 4.9,
-      "learning_rate": 5.7206399245715584e-08,
-      "loss": 0.2421,
-      "step": 139195
-    },
-    {
-      "epoch": 4.9,
-      "learning_rate": 5.7013955681839426e-08,
-      "loss": 0.2445,
-      "step": 139200
-    },
-    {
-      "epoch": 4.9,
-      "learning_rate": 5.682183598620106e-08,
-      "loss": 0.2674,
-      "step": 139205
-    },
-    {
-      "epoch": 4.9,
-      "learning_rate": 5.6630040161290145e-08,
-      "loss": 0.2495,
-      "step": 139210
-    },
-    {
-      "epoch": 4.9,
-      "learning_rate": 5.643856820960192e-08,
-      "loss": 0.2737,
-      "step": 139215
-    },
-    {
-      "epoch": 4.9,
-      "learning_rate": 5.62474201336205e-08,
-      "loss": 0.262,
-      "step": 139220
-    },
-    {
-      "epoch": 4.9,
-      "learning_rate": 5.6056595935827236e-08,
-      "loss": 0.244,
-      "step": 139225
-    },
-    {
-      "epoch": 4.9,
-      "learning_rate": 5.586609561869793e-08,
-      "loss": 0.2663,
-      "step": 139230
-    },
-    {
-      "epoch": 4.9,
-      "learning_rate": 5.567591918471116e-08,
-      "loss": 0.254,
-      "step": 139235
-    },
-    {
-      "epoch": 4.9,
-      "learning_rate": 5.5486066636334374e-08,
-      "loss": 0.274,
-      "step": 139240
-    },
-    {
-      "epoch": 4.9,
-      "learning_rate": 5.529653797602952e-08,
-      "loss": 0.2628,
-      "step": 139245
-    },
-    {
-      "epoch": 4.9,
-      "learning_rate": 5.5107333206261266e-08,
-      "loss": 0.2555,
-      "step": 139250
-    },
-    {
-      "epoch": 4.9,
-      "learning_rate": 5.4918452329486004e-08,
-      "loss": 0.2574,
-      "step": 139255
-    },
-    {
-      "epoch": 4.9,
-      "learning_rate": 5.4729895348154534e-08,
-      "loss": 0.2486,
-      "step": 139260
-    },
-    {
-      "epoch": 4.9,
-      "learning_rate": 5.454166226471491e-08,
-      "loss": 0.236,
-      "step": 139265
-    },
-    {
-      "epoch": 4.9,
-      "learning_rate": 5.435375308161239e-08,
-      "loss": 0.2445,
-      "step": 139270
-    },
-    {
-      "epoch": 4.9,
-      "learning_rate": 5.4166167801286697e-08,
-      "loss": 0.2466,
-      "step": 139275
-    },
-    {
-      "epoch": 4.9,
-      "learning_rate": 5.397890642617476e-08,
-      "loss": 0.2489,
-      "step": 139280
-    },
-    {
-      "epoch": 4.9,
-      "learning_rate": 5.3791968958705195e-08,
-      "loss": 0.2564,
-      "step": 139285
-    },
-    {
-      "epoch": 4.9,
-      "learning_rate": 5.360535540130662e-08,
-      "loss": 0.2539,
-      "step": 139290
-    },
-    {
-      "epoch": 4.9,
-      "learning_rate": 5.34190657564021e-08,
-      "loss": 0.2812,
-      "step": 139295
-    },
-    {
-      "epoch": 4.9,
-      "learning_rate": 5.3233100026411906e-08,
-      "loss": 0.2725,
-      "step": 139300
-    },
-    {
-      "epoch": 4.9,
-      "learning_rate": 5.3047458213748015e-08,
-      "loss": 0.2347,
-      "step": 139305
-    },
-    {
-      "epoch": 4.9,
-      "learning_rate": 5.286214032082237e-08,
-      "loss": 0.2598,
-      "step": 139310
-    },
-    {
-      "epoch": 4.9,
-      "learning_rate": 5.26771463500414e-08,
-      "loss": 0.2549,
-      "step": 139315
-    },
-    {
-      "epoch": 4.9,
-      "learning_rate": 5.249247630380594e-08,
-      "loss": 0.2379,
-      "step": 139320
-    },
-    {
-      "epoch": 4.9,
-      "learning_rate": 5.230813018451686e-08,
-      "loss": 0.261,
-      "step": 139325
-    },
-    {
-      "epoch": 4.9,
-      "learning_rate": 5.212410799456391e-08,
-      "loss": 0.2443,
-      "step": 139330
-    },
-    {
-      "epoch": 4.9,
-      "learning_rate": 5.1940409736339626e-08,
-      "loss": 0.2517,
-      "step": 139335
-    },
-    {
-      "epoch": 4.9,
-      "learning_rate": 5.175703541222543e-08,
-      "loss": 0.277,
-      "step": 139340
-    },
-    {
-      "epoch": 4.9,
-      "learning_rate": 5.1573985024605534e-08,
-      "loss": 0.2466,
-      "step": 139345
-    },
-    {
-      "epoch": 4.9,
-      "learning_rate": 5.13912585758558e-08,
-      "loss": 0.2624,
-      "step": 139350
-    },
-    {
-      "epoch": 4.9,
-      "learning_rate": 5.120885606834935e-08,
-      "loss": 0.2759,
-      "step": 139355
-    },
-    {
-      "epoch": 4.9,
-      "learning_rate": 5.1026777504453724e-08,
-      "loss": 0.271,
-      "step": 139360
-    },
-    {
-      "epoch": 4.9,
-      "learning_rate": 5.08450228865337e-08,
-      "loss": 0.2514,
-      "step": 139365
-    },
-    {
-      "epoch": 4.9,
-      "learning_rate": 5.0663592216948496e-08,
-      "loss": 0.2702,
-      "step": 139370
-    },
-    {
-      "epoch": 4.9,
-      "learning_rate": 5.0482485498054566e-08,
-      "loss": 0.2748,
-      "step": 139375
-    },
-    {
-      "epoch": 4.9,
-      "learning_rate": 5.03017027322028e-08,
-      "loss": 0.2497,
-      "step": 139380
-    },
-    {
-      "epoch": 4.9,
-      "learning_rate": 5.0121243921741337e-08,
-      "loss": 0.2428,
-      "step": 139385
-    },
-    {
-      "epoch": 4.9,
-      "learning_rate": 4.9941109069012726e-08,
-      "loss": 0.2449,
-      "step": 139390
-    },
-    {
-      "epoch": 4.9,
-      "learning_rate": 4.976129817635677e-08,
-      "loss": 0.2438,
-      "step": 139395
-    },
-    {
-      "epoch": 4.9,
-      "learning_rate": 4.958181124610772e-08,
-      "loss": 0.2722,
-      "step": 139400
-    },
-    {
-      "epoch": 4.9,
-      "learning_rate": 4.9402648280594265e-08,
-      "loss": 0.2562,
-      "step": 139405
-    },
-    {
-      "epoch": 4.9,
-      "learning_rate": 4.922380928214787e-08,
-      "loss": 0.2715,
-      "step": 139410
-    },
-    {
-      "epoch": 4.91,
-      "learning_rate": 4.9045294253083354e-08,
-      "loss": 0.2519,
-      "step": 139415
-    },
-    {
-      "epoch": 4.91,
-      "learning_rate": 4.886710319572385e-08,
-      "loss": 0.2516,
-      "step": 139420
-    },
-    {
-      "epoch": 4.91,
-      "learning_rate": 4.868923611238141e-08,
-      "loss": 0.269,
-      "step": 139425
-    },
-    {
-      "epoch": 4.91,
-      "learning_rate": 4.851169300536529e-08,
-      "loss": 0.2386,
-      "step": 139430
-    },
-    {
-      "epoch": 4.91,
-      "learning_rate": 4.833447387698198e-08,
-      "loss": 0.2752,
-      "step": 139435
-    },
-    {
-      "epoch": 4.91,
-      "learning_rate": 4.815757872952964e-08,
-      "loss": 0.2407,
-      "step": 139440
-    },
-    {
-      "epoch": 4.91,
-      "learning_rate": 4.798100756530921e-08,
-      "loss": 0.2601,
-      "step": 139445
-    },
-    {
-      "epoch": 4.91,
-      "learning_rate": 4.78047603866133e-08,
-      "loss": 0.2495,
-      "step": 139450
-    },
-    {
-      "epoch": 4.91,
-      "learning_rate": 4.762883719572619e-08,
-      "loss": 0.2573,
-      "step": 139455
-    },
-    {
-      "epoch": 4.91,
-      "learning_rate": 4.745323799493495e-08,
-      "loss": 0.2502,
-      "step": 139460
-    },
-    {
-      "epoch": 4.91,
-      "learning_rate": 4.72779627865183e-08,
-      "loss": 0.2501,
-      "step": 139465
-    },
-    {
-      "epoch": 4.91,
-      "learning_rate": 4.7103011572752206e-08,
-      "loss": 0.2381,
-      "step": 139470
-    },
-    {
-      "epoch": 4.91,
-      "learning_rate": 4.692838435590985e-08,
-      "loss": 0.2508,
-      "step": 139475
-    },
-    {
-      "epoch": 4.91,
-      "learning_rate": 4.675408113825885e-08,
-      "loss": 0.2582,
-      "step": 139480
-    },
-    {
-      "epoch": 4.91,
-      "learning_rate": 4.6580101922058527e-08,
-      "loss": 0.251,
-      "step": 139485
-    },
-    {
-      "epoch": 4.91,
-      "learning_rate": 4.640644670957372e-08,
-      "loss": 0.2454,
-      "step": 139490
-    },
-    {
-      "epoch": 4.91,
-      "learning_rate": 4.623311550305542e-08,
-      "loss": 0.2563,
-      "step": 139495
-    },
-    {
-      "epoch": 4.91,
-      "learning_rate": 4.6060108304754604e-08,
-      "loss": 0.2511,
-      "step": 139500
-    },
-    {
-      "epoch": 4.91,
-      "eval_loss": 0.24838365614414215,
-      "eval_runtime": 10.5546,
-      "eval_samples_per_second": 9.475,
-      "eval_steps_per_second": 9.475,
-      "step": 139500
-    },
-    {
-      "epoch": 4.91,
-      "learning_rate": 4.5887425116916685e-08,
-      "loss": 0.2367,
-      "step": 139505
-    },
-    {
-      "epoch": 4.91,
-      "learning_rate": 4.5715065941787095e-08,
-      "loss": 0.2403,
-      "step": 139510
-    },
-    {
-      "epoch": 4.91,
-      "learning_rate": 4.554303078160016e-08,
-      "loss": 0.2501,
-      "step": 139515
-    },
-    {
-      "epoch": 4.91,
-      "learning_rate": 4.537131963859298e-08,
-      "loss": 0.2529,
-      "step": 139520
-    },
-    {
-      "epoch": 4.91,
-      "learning_rate": 4.519993251499155e-08,
-      "loss": 0.2343,
-      "step": 139525
-    },
-    {
-      "epoch": 4.91,
-      "learning_rate": 4.502886941302464e-08,
-      "loss": 0.2539,
-      "step": 139530
-    },
-    {
-      "epoch": 4.91,
-      "learning_rate": 4.485813033491271e-08,
-      "loss": 0.2312,
-      "step": 139535
-    },
-    {
-      "epoch": 4.91,
-      "learning_rate": 4.468771528287064e-08,
-      "loss": 0.2603,
-      "step": 139540
-    },
-    {
-      "epoch": 4.91,
-      "learning_rate": 4.451762425911332e-08,
-      "loss": 0.2469,
-      "step": 139545
-    },
-    {
-      "epoch": 4.91,
-      "learning_rate": 4.434785726584734e-08,
-      "loss": 0.2494,
-      "step": 139550
-    },
-    {
-      "epoch": 4.91,
-      "learning_rate": 4.417841430527925e-08,
-      "loss": 0.247,
-      "step": 139555
-    },
-    {
-      "epoch": 4.91,
-      "learning_rate": 4.40092953796073e-08,
-      "loss": 0.2504,
-      "step": 139560
-    },
-    {
-      "epoch": 4.91,
-      "learning_rate": 4.3840500491026946e-08,
-      "loss": 0.2637,
-      "step": 139565
-    },
-    {
-      "epoch": 4.91,
-      "learning_rate": 4.3672029641730896e-08,
-      "loss": 0.2409,
-      "step": 139570
-    },
-    {
-      "epoch": 4.91,
-      "learning_rate": 4.350388283390905e-08,
-      "loss": 0.2413,
-      "step": 139575
-    },
-    {
-      "epoch": 4.91,
-      "learning_rate": 4.3336060069740224e-08,
-      "loss": 0.2618,
-      "step": 139580
-    },
-    {
-      "epoch": 4.91,
-      "learning_rate": 4.3168561351406014e-08,
-      "loss": 0.2557,
-      "step": 139585
-    },
-    {
-      "epoch": 4.91,
-      "learning_rate": 4.3001386681082444e-08,
-      "loss": 0.2262,
-      "step": 139590
-    },
-    {
-      "epoch": 4.91,
-      "learning_rate": 4.2834536060937235e-08,
-      "loss": 0.2644,
-      "step": 139595
-    },
-    {
-      "epoch": 4.91,
-      "learning_rate": 4.26680094931381e-08,
-      "loss": 0.2536,
-      "step": 139600
-    },
-    {
-      "epoch": 4.91,
-      "learning_rate": 4.250180697984718e-08,
-      "loss": 0.2648,
-      "step": 139605
-    },
-    {
-      "epoch": 4.91,
-      "learning_rate": 4.233592852322388e-08,
-      "loss": 0.2665,
-      "step": 139610
-    },
-    {
-      "epoch": 4.91,
-      "learning_rate": 4.217037412541924e-08,
-      "loss": 0.2514,
-      "step": 139615
-    },
-    {
-      "epoch": 4.91,
-      "learning_rate": 4.200514378858433e-08,
-      "loss": 0.2558,
-      "step": 139620
-    },
-    {
-      "epoch": 4.91,
-      "learning_rate": 4.184023751486743e-08,
-      "loss": 0.2542,
-      "step": 139625
-    },
-    {
-      "epoch": 4.91,
-      "learning_rate": 4.167565530640571e-08,
-      "loss": 0.252,
-      "step": 139630
-    },
-    {
-      "epoch": 4.91,
-      "learning_rate": 4.151139716533636e-08,
-      "loss": 0.2377,
-      "step": 139635
-    },
-    {
-      "epoch": 4.91,
-      "learning_rate": 4.134746309379378e-08,
-      "loss": 0.2441,
-      "step": 139640
-    },
-    {
-      "epoch": 4.91,
-      "learning_rate": 4.118385309390682e-08,
-      "loss": 0.2431,
-      "step": 139645
-    },
-    {
-      "epoch": 4.91,
-      "learning_rate": 4.102056716779601e-08,
-      "loss": 0.2303,
-      "step": 139650
-    },
-    {
-      "epoch": 4.91,
-      "learning_rate": 4.0857605317587424e-08,
-      "loss": 0.2495,
-      "step": 139655
-    },
-    {
-      "epoch": 4.91,
-      "learning_rate": 4.069496754539326e-08,
-      "loss": 0.2484,
-      "step": 139660
-    },
-    {
-      "epoch": 4.91,
-      "learning_rate": 4.0532653853328495e-08,
-      "loss": 0.2516,
-      "step": 139665
-    },
-    {
-      "epoch": 4.91,
-      "learning_rate": 4.037066424349423e-08,
-      "loss": 0.2341,
-      "step": 139670
-    },
-    {
-      "epoch": 4.91,
-      "learning_rate": 4.020899871799988e-08,
-      "loss": 0.2712,
-      "step": 139675
-    },
-    {
-      "epoch": 4.91,
-      "learning_rate": 4.0047657278943774e-08,
-      "loss": 0.2612,
-      "step": 139680
-    },
-    {
-      "epoch": 4.91,
-      "learning_rate": 3.988663992841868e-08,
-      "loss": 0.2186,
-      "step": 139685
-    },
-    {
-      "epoch": 4.91,
-      "learning_rate": 3.972594666851459e-08,
-      "loss": 0.2565,
-      "step": 139690
-    },
-    {
-      "epoch": 4.91,
-      "learning_rate": 3.9565577501321505e-08,
-      "loss": 0.253,
-      "step": 139695
-    },
-    {
-      "epoch": 4.92,
-      "learning_rate": 3.940553242891831e-08,
-      "loss": 0.25,
-      "step": 139700
-    },
-    {
-      "epoch": 4.92,
-      "learning_rate": 3.924581145338668e-08,
-      "loss": 0.2526,
-      "step": 139705
-    },
-    {
-      "epoch": 4.92,
-      "learning_rate": 3.908641457679718e-08,
-      "loss": 0.2363,
-      "step": 139710
-    },
-    {
-      "epoch": 4.92,
-      "learning_rate": 3.892734180122037e-08,
-      "loss": 0.2487,
-      "step": 139715
-    },
-    {
-      "epoch": 4.92,
-      "learning_rate": 3.8768593128724054e-08,
-      "loss": 0.2375,
-      "step": 139720
-    },
-    {
-      "epoch": 4.92,
-      "learning_rate": 3.861016856136768e-08,
-      "loss": 0.2468,
-      "step": 139725
-    },
-    {
-      "epoch": 4.92,
-      "learning_rate": 3.8452068101205165e-08,
-      "loss": 0.2504,
-      "step": 139730
-    },
-    {
-      "epoch": 4.92,
-      "learning_rate": 3.82942917502932e-08,
-      "loss": 0.2475,
-      "step": 139735
-    },
-    {
-      "epoch": 4.92,
-      "learning_rate": 3.8136839510680145e-08,
-      "loss": 0.2352,
-      "step": 139740
-    },
-    {
-      "epoch": 4.92,
-      "learning_rate": 3.7979711384408814e-08,
-      "loss": 0.2396,
-      "step": 139745
-    },
-    {
-      "epoch": 4.92,
-      "learning_rate": 3.782290737352201e-08,
-      "loss": 0.2344,
-      "step": 139750
-    },
-    {
-      "epoch": 4.92,
-      "learning_rate": 3.7666427480054224e-08,
-      "loss": 0.2612,
-      "step": 139755
-    },
-    {
-      "epoch": 4.92,
-      "learning_rate": 3.751027170603438e-08,
-      "loss": 0.2375,
-      "step": 139760
-    },
-    {
-      "epoch": 4.92,
-      "learning_rate": 3.73544400534942e-08,
-      "loss": 0.2605,
-      "step": 139765
-    },
-    {
-      "epoch": 4.92,
-      "learning_rate": 3.719893252445705e-08,
-      "loss": 0.2591,
-      "step": 139770
-    },
-    {
-      "epoch": 4.92,
-      "learning_rate": 3.7043749120938e-08,
-      "loss": 0.2505,
-      "step": 139775
-    },
-    {
-      "epoch": 4.92,
-      "learning_rate": 3.688888984495764e-08,
-      "loss": 0.2627,
-      "step": 139780
-    },
-    {
-      "epoch": 4.92,
-      "learning_rate": 3.673435469851994e-08,
-      "loss": 0.2744,
-      "step": 139785
-    },
-    {
-      "epoch": 4.92,
-      "learning_rate": 3.658014368363993e-08,
-      "loss": 0.2597,
-      "step": 139790
-    },
-    {
-      "epoch": 4.92,
-      "learning_rate": 3.642625680231049e-08,
-      "loss": 0.2467,
-      "step": 139795
-    },
-    {
-      "epoch": 4.92,
-      "learning_rate": 3.627269405653832e-08,
-      "loss": 0.2613,
-      "step": 139800
-    },
-    {
-      "epoch": 4.92,
-      "learning_rate": 3.6119455448310746e-08,
-      "loss": 0.2453,
-      "step": 139805
-    },
-    {
-      "epoch": 4.92,
-      "learning_rate": 3.59665409796206e-08,
-      "loss": 0.2469,
-      "step": 139810
-    },
-    {
-      "epoch": 4.92,
-      "learning_rate": 3.581395065245519e-08,
-      "loss": 0.2677,
-      "step": 139815
-    },
-    {
-      "epoch": 4.92,
-      "learning_rate": 3.566168446879348e-08,
-      "loss": 0.2571,
-      "step": 139820
-    },
-    {
-      "epoch": 4.92,
-      "learning_rate": 3.5509742430608894e-08,
-      "loss": 0.2689,
-      "step": 139825
-    },
-    {
-      "epoch": 4.92,
-      "learning_rate": 3.535812453988319e-08,
-      "loss": 0.2546,
-      "step": 139830
-    },
-    {
-      "epoch": 4.92,
-      "learning_rate": 3.52068307985759e-08,
-      "loss": 0.2396,
-      "step": 139835
-    },
-    {
-      "epoch": 4.92,
-      "learning_rate": 3.505586120865767e-08,
-      "loss": 0.2477,
-      "step": 139840
-    },
-    {
-      "epoch": 4.92,
-      "learning_rate": 3.4905215772085275e-08,
-      "loss": 0.2458,
-      "step": 139845
-    },
-    {
-      "epoch": 4.92,
-      "learning_rate": 3.475489449081826e-08,
-      "loss": 0.2482,
-      "step": 139850
-    },
-    {
-      "epoch": 4.92,
-      "learning_rate": 3.460489736680228e-08,
-      "loss": 0.2831,
-      "step": 139855
-    },
-    {
-      "epoch": 4.92,
-      "learning_rate": 3.445522440199134e-08,
-      "loss": 0.2551,
-      "step": 139860
-    },
-    {
-      "epoch": 4.92,
-      "learning_rate": 3.4305875598325545e-08,
-      "loss": 0.2651,
-      "step": 139865
-    },
-    {
-      "epoch": 4.92,
-      "learning_rate": 3.415685095774779e-08,
-      "loss": 0.2481,
-      "step": 139870
-    },
-    {
-      "epoch": 4.92,
-      "learning_rate": 3.400815048218708e-08,
-      "loss": 0.2391,
-      "step": 139875
-    },
-    {
-      "epoch": 4.92,
-      "learning_rate": 3.3859774173577996e-08,
-      "loss": 0.2464,
-      "step": 139880
-    },
-    {
-      "epoch": 4.92,
-      "learning_rate": 3.371172203384676e-08,
-      "loss": 0.2561,
-      "step": 139885
-    },
-    {
-      "epoch": 4.92,
-      "learning_rate": 3.356399406491684e-08,
-      "loss": 0.264,
-      "step": 139890
-    },
-    {
-      "epoch": 4.92,
-      "learning_rate": 3.3416590268703364e-08,
-      "loss": 0.2389,
-      "step": 139895
-    },
-    {
-      "epoch": 4.92,
-      "learning_rate": 3.3269510647121474e-08,
-      "loss": 0.2743,
-      "step": 139900
-    },
-    {
-      "epoch": 4.92,
-      "learning_rate": 3.312275520208352e-08,
-      "loss": 0.2377,
-      "step": 139905
-    },
-    {
-      "epoch": 4.92,
-      "learning_rate": 3.297632393549077e-08,
-      "loss": 0.2321,
-      "step": 139910
-    },
-    {
-      "epoch": 4.92,
-      "learning_rate": 3.283021684924448e-08,
-      "loss": 0.2727,
-      "step": 139915
-    },
-    {
-      "epoch": 4.92,
-      "learning_rate": 3.268443394524867e-08,
-      "loss": 0.2393,
-      "step": 139920
-    },
-    {
-      "epoch": 4.92,
-      "learning_rate": 3.2538975225387956e-08,
-      "loss": 0.2432,
-      "step": 139925
-    },
-    {
-      "epoch": 4.92,
-      "learning_rate": 3.2393840691555266e-08,
-      "loss": 0.2301,
-      "step": 139930
-    },
-    {
-      "epoch": 4.92,
-      "learning_rate": 3.22490303456352e-08,
-      "loss": 0.2793,
-      "step": 139935
-    },
-    {
-      "epoch": 4.92,
-      "learning_rate": 3.2104544189506816e-08,
-      "loss": 0.251,
-      "step": 139940
-    },
-    {
-      "epoch": 4.92,
-      "learning_rate": 3.196038222504638e-08,
-      "loss": 0.2762,
-      "step": 139945
-    },
-    {
-      "epoch": 4.92,
-      "learning_rate": 3.181654445412463e-08,
-      "loss": 0.2547,
-      "step": 139950
-    },
-    {
-      "epoch": 4.92,
-      "learning_rate": 3.1673030878612286e-08,
-      "loss": 0.2638,
-      "step": 139955
-    },
-    {
-      "epoch": 4.92,
-      "learning_rate": 3.1529841500368975e-08,
-      "loss": 0.2362,
-      "step": 139960
-    },
-    {
-      "epoch": 4.92,
-      "learning_rate": 3.138697632125709e-08,
-      "loss": 0.2567,
-      "step": 139965
-    },
-    {
-      "epoch": 4.92,
-      "learning_rate": 3.1244435343130706e-08,
-      "loss": 0.2716,
-      "step": 139970
-    },
-    {
-      "epoch": 4.92,
-      "learning_rate": 3.110221856783835e-08,
-      "loss": 0.2464,
-      "step": 139975
-    },
-    {
-      "epoch": 4.92,
-      "learning_rate": 3.096032599723131e-08,
-      "loss": 0.2455,
-      "step": 139980
-    },
-    {
-      "epoch": 4.93,
-      "learning_rate": 3.081875763314701e-08,
-      "loss": 0.2536,
-      "step": 139985
-    },
-    {
-      "epoch": 4.93,
-      "learning_rate": 3.067751347742842e-08,
-      "loss": 0.2541,
-      "step": 139990
-    },
-    {
-      "epoch": 4.93,
-      "learning_rate": 3.053659353190463e-08,
-      "loss": 0.2264,
-      "step": 139995
-    },
-    {
-      "epoch": 4.93,
-      "learning_rate": 3.039599779840752e-08,
-      "loss": 0.2472,
-      "step": 140000
-    },
-    {
-      "epoch": 4.93,
-      "eval_loss": 0.2483794391155243,
-      "eval_runtime": 10.5607,
-      "eval_samples_per_second": 9.469,
-      "eval_steps_per_second": 9.469,
-      "step": 140000
-    },
-    {
-      "epoch": 4.93,
-      "learning_rate": 3.025572627876339e-08,
-      "loss": 0.2355,
-      "step": 140005
-    },
-    {
-      "epoch": 4.93,
-      "learning_rate": 3.011577897479301e-08,
-      "loss": 0.2437,
-      "step": 140010
-    },
-    {
-      "epoch": 4.93,
-      "learning_rate": 2.997615588831437e-08,
-      "loss": 0.2383,
-      "step": 140015
-    },
-    {
-      "epoch": 4.93,
-      "learning_rate": 2.9836857021137147e-08,
-      "loss": 0.2328,
-      "step": 140020
-    },
-    {
-      "epoch": 4.93,
-      "learning_rate": 2.969788237507376e-08,
-      "loss": 0.2205,
-      "step": 140025
-    },
-    {
-      "epoch": 4.93,
-      "learning_rate": 2.9559231951928334e-08,
-      "loss": 0.2371,
-      "step": 140030
-    },
-    {
-      "epoch": 4.93,
-      "learning_rate": 2.942090575349943e-08,
-      "loss": 0.2592,
-      "step": 140035
-    },
-    {
-      "epoch": 4.93,
-      "learning_rate": 2.9282903781582828e-08,
-      "loss": 0.2468,
-      "step": 140040
-    },
-    {
-      "epoch": 4.93,
-      "learning_rate": 2.9145226037974317e-08,
-      "loss": 0.2362,
-      "step": 140045
-    },
-    {
-      "epoch": 4.93,
-      "learning_rate": 2.9007872524455805e-08,
-      "loss": 0.2539,
-      "step": 140050
-    },
-    {
-      "epoch": 4.93,
-      "learning_rate": 2.887084324281475e-08,
-      "loss": 0.2358,
-      "step": 140055
-    },
-    {
-      "epoch": 4.93,
-      "learning_rate": 2.873413819482751e-08,
-      "loss": 0.2481,
-      "step": 140060
-    },
-    {
-      "epoch": 4.93,
-      "learning_rate": 2.8597757382273216e-08,
-      "loss": 0.2433,
-      "step": 140065
-    },
-    {
-      "epoch": 4.93,
-      "learning_rate": 2.8461700806917123e-08,
-      "loss": 0.2358,
-      "step": 140070
-    },
-    {
-      "epoch": 4.93,
-      "learning_rate": 2.8325968470532816e-08,
-      "loss": 0.2537,
-      "step": 140075
-    },
-    {
-      "epoch": 4.93,
-      "learning_rate": 2.8190560374877217e-08,
-      "loss": 0.2366,
-      "step": 140080
-    },
-    {
-      "epoch": 4.93,
-      "learning_rate": 2.8055476521710034e-08,
-      "loss": 0.2612,
-      "step": 140085
-    },
-    {
-      "epoch": 4.93,
-      "learning_rate": 2.7920716912785416e-08,
-      "loss": 0.2616,
-      "step": 140090
-    },
-    {
-      "epoch": 4.93,
-      "learning_rate": 2.7786281549851965e-08,
-      "loss": 0.2407,
-      "step": 140095
-    },
-    {
-      "epoch": 4.93,
-      "learning_rate": 2.7652170434658285e-08,
-      "loss": 0.253,
-      "step": 140100
-    },
-    {
-      "epoch": 4.93,
-      "learning_rate": 2.7518383568941875e-08,
-      "loss": 0.2263,
-      "step": 140105
-    },
-    {
-      "epoch": 4.93,
-      "learning_rate": 2.7384920954443006e-08,
-      "loss": 0.2594,
-      "step": 140110
-    },
-    {
-      "epoch": 4.93,
-      "learning_rate": 2.7251782592893627e-08,
-      "loss": 0.2574,
-      "step": 140115
-    },
-    {
-      "epoch": 4.93,
-      "learning_rate": 2.7118968486020136e-08,
-      "loss": 0.2437,
-      "step": 140120
-    },
-    {
-      "epoch": 4.93,
-      "learning_rate": 2.6986478635551705e-08,
-      "loss": 0.2568,
-      "step": 140125
-    },
-    {
-      "epoch": 4.93,
-      "learning_rate": 2.6854313043203627e-08,
-      "loss": 0.2295,
-      "step": 140130
-    },
-    {
-      "epoch": 4.93,
-      "learning_rate": 2.672247171069675e-08,
-      "loss": 0.2409,
-      "step": 140135
-    },
-    {
-      "epoch": 4.93,
-      "learning_rate": 2.6590954639738043e-08,
-      "loss": 0.2298,
-      "step": 140140
-    },
-    {
-      "epoch": 4.93,
-      "learning_rate": 2.645976183204002e-08,
-      "loss": 0.2613,
-      "step": 140145
-    },
-    {
-      "epoch": 4.93,
-      "learning_rate": 2.6328893289301326e-08,
-      "loss": 0.2364,
-      "step": 140150
-    },
-    {
-      "epoch": 4.93,
-      "learning_rate": 2.6198349013226154e-08,
-      "loss": 0.2616,
-      "step": 140155
-    },
-    {
-      "epoch": 4.93,
-      "learning_rate": 2.606812900550759e-08,
-      "loss": 0.252,
-      "step": 140160
-    },
-    {
-      "epoch": 4.93,
-      "learning_rate": 2.593823326783318e-08,
-      "loss": 0.2611,
-      "step": 140165
-    },
-    {
-      "epoch": 4.93,
-      "learning_rate": 2.580866180189323e-08,
-      "loss": 0.2567,
-      "step": 140170
-    },
-    {
-      "epoch": 4.93,
-      "learning_rate": 2.5679414609369734e-08,
-      "loss": 0.2269,
-      "step": 140175
-    },
-    {
-      "epoch": 4.93,
-      "learning_rate": 2.55504916919419e-08,
-      "loss": 0.2433,
-      "step": 140180
-    },
-    {
-      "epoch": 4.93,
-      "learning_rate": 2.5421893051280622e-08,
-      "loss": 0.2451,
-      "step": 140185
-    },
-    {
-      "epoch": 4.93,
-      "learning_rate": 2.5293618689056775e-08,
-      "loss": 0.2458,
-      "step": 140190
-    },
-    {
-      "epoch": 4.93,
-      "learning_rate": 2.51656686069357e-08,
-      "loss": 0.2379,
-      "step": 140195
-    },
-    {
-      "epoch": 4.93,
-      "learning_rate": 2.5038042806579953e-08,
-      "loss": 0.2551,
-      "step": 140200
-    },
-    {
-      "epoch": 4.93,
-      "learning_rate": 2.4910741289646543e-08,
-      "loss": 0.2582,
-      "step": 140205
-    },
-    {
-      "epoch": 4.93,
-      "learning_rate": 2.478376405778693e-08,
-      "loss": 0.243,
-      "step": 140210
-    },
-    {
-      "epoch": 4.93,
-      "learning_rate": 2.4657111112652566e-08,
-      "loss": 0.2432,
-      "step": 140215
-    },
-    {
-      "epoch": 4.93,
-      "learning_rate": 2.4530782455886582e-08,
-      "loss": 0.2447,
-      "step": 140220
-    },
-    {
-      "epoch": 4.93,
-      "learning_rate": 2.4404778089126555e-08,
-      "loss": 0.252,
-      "step": 140225
-    },
-    {
-      "epoch": 4.93,
-      "learning_rate": 2.4279098014012847e-08,
-      "loss": 0.2502,
-      "step": 140230
-    },
-    {
-      "epoch": 4.93,
-      "learning_rate": 2.41537422321747e-08,
-      "loss": 0.2246,
-      "step": 140235
-    },
-    {
-      "epoch": 4.93,
-      "learning_rate": 2.4028710745241378e-08,
-      "loss": 0.2442,
-      "step": 140240
-    },
-    {
-      "epoch": 4.93,
-      "learning_rate": 2.39040035548338e-08,
-      "loss": 0.2282,
-      "step": 140245
-    },
-    {
-      "epoch": 4.93,
-      "learning_rate": 2.3779620662572888e-08,
-      "loss": 0.248,
-      "step": 140250
-    },
-    {
-      "epoch": 4.93,
-      "learning_rate": 2.36555620700768e-08,
-      "loss": 0.2496,
-      "step": 140255
-    },
-    {
-      "epoch": 4.93,
-      "learning_rate": 2.35318277789498e-08,
-      "loss": 0.2691,
-      "step": 140260
-    },
-    {
-      "epoch": 4.93,
-      "learning_rate": 2.3408417790801716e-08,
-      "loss": 0.2883,
-      "step": 140265
-    },
-    {
-      "epoch": 4.94,
-      "learning_rate": 2.3285332107236823e-08,
-      "loss": 0.2487,
-      "step": 140270
-    },
-    {
-      "epoch": 4.94,
-      "learning_rate": 2.316257072985384e-08,
-      "loss": 0.2364,
-      "step": 140275
-    },
-    {
-      "epoch": 4.94,
-      "learning_rate": 2.304013366024038e-08,
-      "loss": 0.2636,
-      "step": 140280
-    },
-    {
-      "epoch": 4.94,
-      "learning_rate": 2.29180208999924e-08,
-      "loss": 0.2453,
-      "step": 140285
-    },
-    {
-      "epoch": 4.94,
-      "learning_rate": 2.2796232450691956e-08,
-      "loss": 0.2455,
-      "step": 140290
-    },
-    {
-      "epoch": 4.94,
-      "learning_rate": 2.2674768313923898e-08,
-      "loss": 0.2399,
-      "step": 140295
-    },
-    {
-      "epoch": 4.94,
-      "learning_rate": 2.255362849126197e-08,
-      "loss": 0.2482,
-      "step": 140300
-    },
-    {
-      "epoch": 4.94,
-      "learning_rate": 2.2432812984279905e-08,
-      "loss": 0.2399,
-      "step": 140305
-    },
-    {
-      "epoch": 4.94,
-      "learning_rate": 2.2312321794548673e-08,
-      "loss": 0.2593,
-      "step": 140310
-    },
-    {
-      "epoch": 4.94,
-      "learning_rate": 2.219215492362814e-08,
-      "loss": 0.239,
-      "step": 140315
-    },
-    {
-      "epoch": 4.94,
-      "learning_rate": 2.2072312373083716e-08,
-      "loss": 0.261,
-      "step": 140320
-    },
-    {
-      "epoch": 4.94,
-      "learning_rate": 2.1952794144469713e-08,
-      "loss": 0.2639,
-      "step": 140325
-    },
-    {
-      "epoch": 4.94,
-      "learning_rate": 2.183360023933767e-08,
-      "loss": 0.2514,
-      "step": 140330
-    },
-    {
-      "epoch": 4.94,
-      "learning_rate": 2.1714730659233573e-08,
-      "loss": 0.2684,
-      "step": 140335
-    },
-    {
-      "epoch": 4.94,
-      "learning_rate": 2.1596185405703407e-08,
-      "loss": 0.2716,
-      "step": 140340
-    },
-    {
-      "epoch": 4.94,
-      "learning_rate": 2.1477964480284827e-08,
-      "loss": 0.2562,
-      "step": 140345
-    },
-    {
-      "epoch": 4.94,
-      "learning_rate": 2.1360067884512725e-08,
-      "loss": 0.248,
-      "step": 140350
-    },
-    {
-      "epoch": 4.94,
-      "learning_rate": 2.12424956199192e-08,
-      "loss": 0.239,
-      "step": 140355
-    },
-    {
-      "epoch": 4.94,
-      "learning_rate": 2.1125247688030812e-08,
-      "loss": 0.2444,
-      "step": 140360
-    },
-    {
-      "epoch": 4.94,
-      "learning_rate": 2.1008324090368568e-08,
-      "loss": 0.2539,
-      "step": 140365
-    },
-    {
-      "epoch": 4.94,
-      "learning_rate": 2.0891724828453476e-08,
-      "loss": 0.2366,
-      "step": 140370
-    },
-    {
-      "epoch": 4.94,
-      "learning_rate": 2.077544990379543e-08,
-      "loss": 0.2539,
-      "step": 140375
-    },
-    {
-      "epoch": 4.94,
-      "learning_rate": 2.0659499317907116e-08,
-      "loss": 0.2419,
-      "step": 140380
-    },
-    {
-      "epoch": 4.94,
-      "learning_rate": 2.0543873072292887e-08,
-      "loss": 0.2431,
-      "step": 140385
-    },
-    {
-      "epoch": 4.94,
-      "learning_rate": 2.0428571168454312e-08,
-      "loss": 0.2521,
-      "step": 140390
-    },
-    {
-      "epoch": 4.94,
-      "learning_rate": 2.0313593607887427e-08,
-      "loss": 0.2359,
-      "step": 140395
-    },
-    {
-      "epoch": 4.94,
-      "learning_rate": 2.019894039208825e-08,
-      "loss": 0.2645,
-      "step": 140400
-    },
-    {
-      "epoch": 4.94,
-      "learning_rate": 2.008461152254171e-08,
-      "loss": 0.2329,
-      "step": 140405
-    },
-    {
-      "epoch": 4.94,
-      "learning_rate": 1.9970607000735497e-08,
-      "loss": 0.2538,
-      "step": 140410
-    },
-    {
-      "epoch": 4.94,
-      "learning_rate": 1.9856926828148993e-08,
-      "loss": 0.2494,
-      "step": 140415
-    },
-    {
-      "epoch": 4.94,
-      "learning_rate": 1.9743571006256014e-08,
-      "loss": 0.2461,
-      "step": 140420
-    },
-    {
-      "epoch": 4.94,
-      "learning_rate": 1.963053953653038e-08,
-      "loss": 0.2636,
-      "step": 140425
-    },
-    {
-      "epoch": 4.94,
-      "learning_rate": 1.9517832420440363e-08,
-      "loss": 0.25,
-      "step": 140430
-    },
-    {
-      "epoch": 4.94,
-      "learning_rate": 1.940544965944868e-08,
-      "loss": 0.2498,
-      "step": 140435
-    },
-    {
-      "epoch": 4.94,
-      "learning_rate": 1.929339125501528e-08,
-      "loss": 0.2806,
-      "step": 140440
-    },
-    {
-      "epoch": 4.94,
-      "learning_rate": 1.9181657208591775e-08,
-      "loss": 0.2557,
-      "step": 140445
-    },
-    {
-      "epoch": 4.94,
-      "learning_rate": 1.9070247521632556e-08,
-      "loss": 0.2491,
-      "step": 140450
-    },
-    {
-      "epoch": 4.94,
-      "learning_rate": 1.895916219558369e-08,
-      "loss": 0.2463,
-      "step": 140455
-    },
-    {
-      "epoch": 4.94,
-      "learning_rate": 1.8848401231885692e-08,
-      "loss": 0.2464,
-      "step": 140460
-    },
-    {
-      "epoch": 4.94,
-      "learning_rate": 1.8737964631979078e-08,
-      "loss": 0.2438,
-      "step": 140465
-    },
-    {
-      "epoch": 4.94,
-      "learning_rate": 1.8627852397298808e-08,
-      "loss": 0.2542,
-      "step": 140470
-    },
-    {
-      "epoch": 4.94,
-      "learning_rate": 1.851806452926874e-08,
-      "loss": 0.2692,
-      "step": 140475
-    },
-    {
-      "epoch": 4.94,
-      "learning_rate": 1.840860102932107e-08,
-      "loss": 0.2289,
-      "step": 140480
-    },
-    {
-      "epoch": 4.94,
-      "learning_rate": 1.829946189887133e-08,
-      "loss": 0.2403,
-      "step": 140485
-    },
-    {
-      "epoch": 4.94,
-      "learning_rate": 1.8190647139343376e-08,
-      "loss": 0.2488,
-      "step": 140490
-    },
-    {
-      "epoch": 4.94,
-      "learning_rate": 1.8082156752141642e-08,
-      "loss": 0.2613,
-      "step": 140495
-    },
-    {
-      "epoch": 4.94,
-      "learning_rate": 1.7973990738684443e-08,
-      "loss": 0.2542,
-      "step": 140500
-    },
-    {
-      "epoch": 4.94,
-      "eval_loss": 0.24837003648281097,
-      "eval_runtime": 10.5523,
-      "eval_samples_per_second": 9.477,
-      "eval_steps_per_second": 9.477,
-      "step": 140500
-    },
-    {
-      "epoch": 4.94,
-      "learning_rate": 1.786614910036788e-08,
-      "loss": 0.2355,
-      "step": 140505
-    },
-    {
-      "epoch": 4.94,
-      "learning_rate": 1.7758631838596385e-08,
-      "loss": 0.2509,
-      "step": 140510
-    },
-    {
-      "epoch": 4.94,
-      "learning_rate": 1.7651438954763287e-08,
-      "loss": 0.2399,
-      "step": 140515
-    },
-    {
-      "epoch": 4.94,
-      "learning_rate": 1.7544570450264697e-08,
-      "loss": 0.2539,
-      "step": 140520
-    },
-    {
-      "epoch": 4.94,
-      "learning_rate": 1.7438026326485612e-08,
-      "loss": 0.262,
-      "step": 140525
-    },
-    {
-      "epoch": 4.94,
-      "learning_rate": 1.7331806584808263e-08,
-      "loss": 0.234,
-      "step": 140530
-    },
-    {
-      "epoch": 4.94,
-      "learning_rate": 1.7225911226614877e-08,
-      "loss": 0.2576,
-      "step": 140535
-    },
-    {
-      "epoch": 4.94,
-      "learning_rate": 1.712034025327658e-08,
-      "loss": 0.2385,
-      "step": 140540
-    },
-    {
-      "epoch": 4.94,
-      "learning_rate": 1.701509366616727e-08,
-      "loss": 0.2688,
-      "step": 140545
-    },
-    {
-      "epoch": 4.94,
-      "learning_rate": 1.69101714666553e-08,
-      "loss": 0.247,
-      "step": 140550
-    },
-    {
-      "epoch": 4.95,
-      "learning_rate": 1.6805573656095142e-08,
-      "loss": 0.2424,
-      "step": 140555
-    },
-    {
-      "epoch": 4.95,
-      "learning_rate": 1.6701300235855144e-08,
-      "loss": 0.2788,
-      "step": 140560
-    },
-    {
-      "epoch": 4.95,
-      "learning_rate": 1.6597351207281452e-08,
-      "loss": 0.2521,
-      "step": 140565
-    },
-    {
-      "epoch": 4.95,
-      "learning_rate": 1.6493726571725766e-08,
-      "loss": 0.2425,
-      "step": 140570
-    },
-    {
-      "epoch": 4.95,
-      "learning_rate": 1.6390426330534226e-08,
-      "loss": 0.2437,
-      "step": 140575
-    },
-    {
-      "epoch": 4.95,
-      "learning_rate": 1.6287450485050205e-08,
-      "loss": 0.2269,
-      "step": 140580
-    },
-    {
-      "epoch": 4.95,
-      "learning_rate": 1.6184799036608745e-08,
-      "loss": 0.2454,
-      "step": 140585
-    },
-    {
-      "epoch": 4.95,
-      "learning_rate": 1.6082471986542115e-08,
-      "loss": 0.2661,
-      "step": 140590
-    },
-    {
-      "epoch": 4.95,
-      "learning_rate": 1.5980469336179805e-08,
-      "loss": 0.2495,
-      "step": 140595
-    },
-    {
-      "epoch": 4.95,
-      "learning_rate": 1.5878791086845758e-08,
-      "loss": 0.2439,
-      "step": 140600
-    },
-    {
-      "epoch": 4.95,
-      "learning_rate": 1.5777437239858363e-08,
-      "loss": 0.248,
-      "step": 140605
-    },
-    {
-      "epoch": 4.95,
-      "learning_rate": 1.5676407796538784e-08,
-      "loss": 0.2583,
-      "step": 140610
-    },
-    {
-      "epoch": 4.95,
-      "learning_rate": 1.557570275819431e-08,
-      "loss": 0.2516,
-      "step": 140615
-    },
-    {
-      "epoch": 4.95,
-      "learning_rate": 1.5475322126135006e-08,
-      "loss": 0.2551,
-      "step": 140620
-    },
-    {
-      "epoch": 4.95,
-      "learning_rate": 1.5375265901662604e-08,
-      "loss": 0.2522,
-      "step": 140625
-    },
-    {
-      "epoch": 4.95,
-      "learning_rate": 1.527553408607607e-08,
-      "loss": 0.2352,
-      "step": 140630
-    },
-    {
-      "epoch": 4.95,
-      "learning_rate": 1.5176126680671587e-08,
-      "loss": 0.2735,
-      "step": 140635
-    },
-    {
-      "epoch": 4.95,
-      "learning_rate": 1.5077043686739788e-08,
-      "loss": 0.2506,
-      "step": 140640
-    },
-    {
-      "epoch": 4.95,
-      "learning_rate": 1.4978285105565758e-08,
-      "loss": 0.2519,
-      "step": 140645
-    },
-    {
-      "epoch": 4.95,
-      "learning_rate": 1.487985093843458e-08,
-      "loss": 0.2482,
-      "step": 140650
-    },
-    {
-      "epoch": 4.95,
-      "learning_rate": 1.478174118662301e-08,
-      "loss": 0.2525,
-      "step": 140655
-    },
-    {
-      "epoch": 4.95,
-      "learning_rate": 1.4683955851402254e-08,
-      "loss": 0.2385,
-      "step": 140660
-    },
-    {
-      "epoch": 4.95,
-      "learning_rate": 1.458649493404629e-08,
-      "loss": 0.2423,
-      "step": 140665
-    },
-    {
-      "epoch": 4.95,
-      "learning_rate": 1.4489358435818e-08,
-      "loss": 0.2522,
-      "step": 140670
-    },
-    {
-      "epoch": 4.95,
-      "learning_rate": 1.4392546357980263e-08,
-      "loss": 0.2519,
-      "step": 140675
-    },
-    {
-      "epoch": 4.95,
-      "learning_rate": 1.429605870178763e-08,
-      "loss": 0.2423,
-      "step": 140680
-    },
-    {
-      "epoch": 4.95,
-      "learning_rate": 1.4199895468497426e-08,
-      "loss": 0.2499,
-      "step": 140685
-    },
-    {
-      "epoch": 4.95,
-      "learning_rate": 1.4104056659353104e-08,
-      "loss": 0.2446,
-      "step": 140690
-    },
-    {
-      "epoch": 4.95,
-      "learning_rate": 1.400854227560089e-08,
-      "loss": 0.2484,
-      "step": 140695
-    },
-    {
-      "epoch": 4.95,
-      "learning_rate": 1.3913352318481453e-08,
-      "loss": 0.2521,
-      "step": 140700
-    },
-    {
-      "epoch": 4.95,
-      "learning_rate": 1.3818486789232699e-08,
-      "loss": 0.2474,
-      "step": 140705
-    },
-    {
-      "epoch": 4.95,
-      "learning_rate": 1.372394568908142e-08,
-      "loss": 0.2445,
-      "step": 140710
-    },
-    {
-      "epoch": 4.95,
-      "learning_rate": 1.3629729019259962e-08,
-      "loss": 0.251,
-      "step": 140715
-    },
-    {
-      "epoch": 4.95,
-      "learning_rate": 1.3535836780989575e-08,
-      "loss": 0.2597,
-      "step": 140720
-    },
-    {
-      "epoch": 4.95,
-      "learning_rate": 1.3442268975488726e-08,
-      "loss": 0.25,
-      "step": 140725
-    },
-    {
-      "epoch": 4.95,
-      "learning_rate": 1.3349025603975884e-08,
-      "loss": 0.2393,
-      "step": 140730
-    },
-    {
-      "epoch": 4.95,
-      "learning_rate": 1.3256106667655644e-08,
-      "loss": 0.2513,
-      "step": 140735
-    },
-    {
-      "epoch": 4.95,
-      "learning_rate": 1.3163512167740922e-08,
-      "loss": 0.2491,
-      "step": 140740
-    },
-    {
-      "epoch": 4.95,
-      "learning_rate": 1.3071242105427983e-08,
-      "loss": 0.2626,
-      "step": 140745
-    },
-    {
-      "epoch": 4.95,
-      "learning_rate": 1.2979296481921422e-08,
-      "loss": 0.2549,
-      "step": 140750
-    },
-    {
-      "epoch": 4.95,
-      "learning_rate": 1.2887675298409175e-08,
-      "loss": 0.2489,
-      "step": 140755
-    },
-    {
-      "epoch": 4.95,
-      "learning_rate": 1.2796378556084732e-08,
-      "loss": 0.2622,
-      "step": 140760
-    },
-    {
-      "epoch": 4.95,
-      "learning_rate": 1.270540625613048e-08,
-      "loss": 0.2323,
-      "step": 140765
-    },
-    {
-      "epoch": 4.95,
-      "learning_rate": 1.2614758399731585e-08,
-      "loss": 0.2562,
-      "step": 140770
-    },
-    {
-      "epoch": 4.95,
-      "learning_rate": 1.2524434988062106e-08,
-      "loss": 0.2434,
-      "step": 140775
-    },
-    {
-      "epoch": 4.95,
-      "learning_rate": 1.2434436022293327e-08,
-      "loss": 0.2521,
-      "step": 140780
-    },
-    {
-      "epoch": 4.95,
-      "learning_rate": 1.2344761503599312e-08,
-      "loss": 0.2525,
-      "step": 140785
-    },
-    {
-      "epoch": 4.95,
-      "learning_rate": 1.2255411433137464e-08,
-      "loss": 0.2441,
-      "step": 140790
-    },
-    {
-      "epoch": 4.95,
-      "learning_rate": 1.2166385812073522e-08,
-      "loss": 0.2572,
-      "step": 140795
-    },
-    {
-      "epoch": 4.95,
-      "learning_rate": 1.2077684641562114e-08,
-      "loss": 0.2391,
-      "step": 140800
-    },
-    {
-      "epoch": 4.95,
-      "learning_rate": 1.1989307922752324e-08,
-      "loss": 0.2428,
-      "step": 140805
-    },
-    {
-      "epoch": 4.95,
-      "learning_rate": 1.1901255656796006e-08,
-      "loss": 0.2312,
-      "step": 140810
-    },
-    {
-      "epoch": 4.95,
-      "learning_rate": 1.1813527844833916e-08,
-      "loss": 0.2521,
-      "step": 140815
-    },
-    {
-      "epoch": 4.95,
-      "learning_rate": 1.1726124488004032e-08,
-      "loss": 0.2576,
-      "step": 140820
-    },
-    {
-      "epoch": 4.95,
-      "learning_rate": 1.1639045587441554e-08,
-      "loss": 0.2525,
-      "step": 140825
-    },
-    {
-      "epoch": 4.95,
-      "learning_rate": 1.1552291144278915e-08,
-      "loss": 0.2638,
-      "step": 140830
-    },
-    {
-      "epoch": 4.95,
-      "learning_rate": 1.1465861159642988e-08,
-      "loss": 0.2554,
-      "step": 140835
-    },
-    {
-      "epoch": 4.96,
-      "learning_rate": 1.1379755634655099e-08,
-      "loss": 0.2675,
-      "step": 140840
-    },
-    {
-      "epoch": 4.96,
-      "learning_rate": 1.1293974570431021e-08,
-      "loss": 0.2464,
-      "step": 140845
-    },
-    {
-      "epoch": 4.96,
-      "learning_rate": 1.120851796808653e-08,
-      "loss": 0.2727,
-      "step": 140850
-    },
-    {
-      "epoch": 4.96,
-      "learning_rate": 1.1123385828734622e-08,
-      "loss": 0.2349,
-      "step": 140855
-    },
-    {
-      "epoch": 4.96,
-      "learning_rate": 1.103857815347442e-08,
-      "loss": 0.257,
-      "step": 140860
-    },
-    {
-      "epoch": 4.96,
-      "learning_rate": 1.0954094943410598e-08,
-      "loss": 0.2598,
-      "step": 140865
-    },
-    {
-      "epoch": 4.96,
-      "learning_rate": 1.086993619963672e-08,
-      "loss": 0.2654,
-      "step": 140870
-    },
-    {
-      "epoch": 4.96,
-      "learning_rate": 1.0786101923251912e-08,
-      "loss": 0.2598,
-      "step": 140875
-    },
-    {
-      "epoch": 4.96,
-      "learning_rate": 1.0702592115338639e-08,
-      "loss": 0.2388,
-      "step": 140880
-    },
-    {
-      "epoch": 4.96,
-      "learning_rate": 1.061940677698492e-08,
-      "loss": 0.2646,
-      "step": 140885
-    },
-    {
-      "epoch": 4.96,
-      "learning_rate": 1.0536545909270446e-08,
-      "loss": 0.2582,
-      "step": 140890
-    },
-    {
-      "epoch": 4.96,
-      "learning_rate": 1.0454009513269359e-08,
-      "loss": 0.2524,
-      "step": 140895
-    },
-    {
-      "epoch": 4.96,
-      "learning_rate": 1.0371797590053023e-08,
-      "loss": 0.2725,
-      "step": 140900
-    },
-    {
-      "epoch": 4.96,
-      "learning_rate": 1.0289910140692804e-08,
-      "loss": 0.226,
-      "step": 140905
-    },
-    {
-      "epoch": 4.96,
-      "learning_rate": 1.020834716624619e-08,
-      "loss": 0.2708,
-      "step": 140910
-    },
-    {
-      "epoch": 4.96,
-      "learning_rate": 1.0127108667778994e-08,
-      "loss": 0.2469,
-      "step": 140915
-    },
-    {
-      "epoch": 4.96,
-      "learning_rate": 1.0046194646340379e-08,
-      "loss": 0.2362,
-      "step": 140920
-    },
-    {
-      "epoch": 4.96,
-      "learning_rate": 9.965605102982278e-09,
-      "loss": 0.2453,
-      "step": 140925
-    },
-    {
-      "epoch": 4.96,
-      "learning_rate": 9.885340038751079e-09,
-      "loss": 0.2572,
-      "step": 140930
-    },
-    {
-      "epoch": 4.96,
-      "learning_rate": 9.805399454690389e-09,
-      "loss": 0.2479,
-      "step": 140935
-    },
-    {
-      "epoch": 4.96,
-      "learning_rate": 9.725783351838268e-09,
-      "loss": 0.2517,
-      "step": 140940
-    },
-    {
-      "epoch": 4.96,
-      "learning_rate": 9.646491731227225e-09,
-      "loss": 0.247,
-      "step": 140945
-    },
-    {
-      "epoch": 4.96,
-      "learning_rate": 9.567524593886989e-09,
-      "loss": 0.2489,
-      "step": 140950
-    },
-    {
-      "epoch": 4.96,
-      "learning_rate": 9.488881940844518e-09,
-      "loss": 0.257,
-      "step": 140955
-    },
-    {
-      "epoch": 4.96,
-      "learning_rate": 9.410563773118442e-09,
-      "loss": 0.2491,
-      "step": 140960
-    },
-    {
-      "epoch": 4.96,
-      "learning_rate": 9.332570091724612e-09,
-      "loss": 0.2267,
-      "step": 140965
-    },
-    {
-      "epoch": 4.96,
-      "learning_rate": 9.254900897681662e-09,
-      "loss": 0.2503,
-      "step": 140970
-    },
-    {
-      "epoch": 4.96,
-      "learning_rate": 9.177556191991565e-09,
-      "loss": 0.2487,
-      "step": 140975
-    },
-    {
-      "epoch": 4.96,
-      "learning_rate": 9.100535975661851e-09,
-      "loss": 0.2439,
-      "step": 140980
-    },
-    {
-      "epoch": 4.96,
-      "learning_rate": 9.023840249694492e-09,
-      "loss": 0.2705,
-      "step": 140985
-    },
-    {
-      "epoch": 4.96,
-      "learning_rate": 8.947469015080367e-09,
-      "loss": 0.2442,
-      "step": 140990
-    },
-    {
-      "epoch": 4.96,
-      "learning_rate": 8.871422272813124e-09,
-      "loss": 0.2463,
-      "step": 140995
-    },
-    {
-      "epoch": 4.96,
-      "learning_rate": 8.795700023883634e-09,
-      "loss": 0.2905,
-      "step": 141000
-    },
-    {
-      "epoch": 4.96,
-      "eval_loss": 0.24837106466293335,
-      "eval_runtime": 10.5397,
-      "eval_samples_per_second": 9.488,
-      "eval_steps_per_second": 9.488,
-      "step": 141000
-    },
-    {
-      "epoch": 4.96,
-      "learning_rate": 8.720302269268899e-09,
-      "loss": 0.2621,
-      "step": 141005
-    },
-    {
-      "epoch": 4.96,
-      "learning_rate": 8.645229009954236e-09,
-      "loss": 0.2356,
-      "step": 141010
-    },
-    {
-      "epoch": 4.96,
-      "learning_rate": 8.570480246908318e-09,
-      "loss": 0.253,
-      "step": 141015
-    },
-    {
-      "epoch": 4.96,
-      "learning_rate": 8.496055981105367e-09,
-      "loss": 0.2589,
-      "step": 141020
-    },
-    {
-      "epoch": 4.96,
-      "learning_rate": 8.421956213514049e-09,
-      "loss": 0.2399,
-      "step": 141025
-    },
-    {
-      "epoch": 4.96,
-      "learning_rate": 8.348180945091932e-09,
-      "loss": 0.2517,
-      "step": 141030
-    },
-    {
-      "epoch": 4.96,
-      "learning_rate": 8.274730176796586e-09,
-      "loss": 0.2442,
-      "step": 141035
-    },
-    {
-      "epoch": 4.96,
-      "learning_rate": 8.201603909585576e-09,
-      "loss": 0.2488,
-      "step": 141040
-    },
-    {
-      "epoch": 4.96,
-      "learning_rate": 8.128802144405367e-09,
-      "loss": 0.261,
-      "step": 141045
-    },
-    {
-      "epoch": 4.96,
-      "learning_rate": 8.056324882202425e-09,
-      "loss": 0.2444,
-      "step": 141050
-    },
-    {
-      "epoch": 4.96,
-      "learning_rate": 7.984172123920441e-09,
-      "loss": 0.2682,
-      "step": 141055
-    },
-    {
-      "epoch": 4.96,
-      "learning_rate": 7.912343870492001e-09,
-      "loss": 0.2582,
-      "step": 141060
-    },
-    {
-      "epoch": 4.96,
-      "learning_rate": 7.84084012284969e-09,
-      "loss": 0.2334,
-      "step": 141065
-    },
-    {
-      "epoch": 4.96,
-      "learning_rate": 7.7696608819261e-09,
-      "loss": 0.268,
-      "step": 141070
-    },
-    {
-      "epoch": 4.96,
-      "learning_rate": 7.698806148639937e-09,
-      "loss": 0.262,
-      "step": 141075
-    },
-    {
-      "epoch": 4.96,
-      "learning_rate": 7.628275923918238e-09,
-      "loss": 0.2562,
-      "step": 141080
-    },
-    {
-      "epoch": 4.96,
-      "learning_rate": 7.558070208671385e-09,
-      "loss": 0.2562,
-      "step": 141085
-    },
-    {
-      "epoch": 4.96,
-      "learning_rate": 7.488189003809765e-09,
-      "loss": 0.2381,
-      "step": 141090
-    },
-    {
-      "epoch": 4.96,
-      "learning_rate": 7.4186323102465315e-09,
-      "loss": 0.2648,
-      "step": 141095
-    },
-    {
-      "epoch": 4.96,
-      "learning_rate": 7.3494001288809676e-09,
-      "loss": 0.2472,
-      "step": 141100
-    },
-    {
-      "epoch": 4.96,
-      "learning_rate": 7.2804924606095785e-09,
-      "loss": 0.2379,
-      "step": 141105
-    },
-    {
-      "epoch": 4.96,
-      "learning_rate": 7.211909306334419e-09,
-      "loss": 0.2699,
-      "step": 141110
-    },
-    {
-      "epoch": 4.96,
-      "learning_rate": 7.1436506669381175e-09,
-      "loss": 0.2614,
-      "step": 141115
-    },
-    {
-      "epoch": 4.96,
-      "learning_rate": 7.0757165433144035e-09,
-      "loss": 0.2625,
-      "step": 141120
-    },
-    {
-      "epoch": 4.97,
-      "learning_rate": 7.008106936337577e-09,
-      "loss": 0.2355,
-      "step": 141125
-    },
-    {
-      "epoch": 4.97,
-      "learning_rate": 6.940821846893042e-09,
-      "loss": 0.2697,
-      "step": 141130
-    },
-    {
-      "epoch": 4.97,
-      "learning_rate": 6.873861275849547e-09,
-      "loss": 0.2454,
-      "step": 141135
-    },
-    {
-      "epoch": 4.97,
-      "learning_rate": 6.807225224078617e-09,
-      "loss": 0.2598,
-      "step": 141140
-    },
-    {
-      "epoch": 4.97,
-      "learning_rate": 6.7409136924434516e-09,
-      "loss": 0.2537,
-      "step": 141145
-    },
-    {
-      "epoch": 4.97,
-      "learning_rate": 6.674926681804472e-09,
-      "loss": 0.2552,
-      "step": 141150
-    },
-    {
-      "epoch": 4.97,
-      "learning_rate": 6.609264193024878e-09,
-      "loss": 0.2453,
-      "step": 141155
-    },
-    {
-      "epoch": 4.97,
-      "learning_rate": 6.543926226948438e-09,
-      "loss": 0.2443,
-      "step": 141160
-    },
-    {
-      "epoch": 4.97,
-      "learning_rate": 6.478912784430025e-09,
-      "loss": 0.241,
-      "step": 141165
-    },
-    {
-      "epoch": 4.97,
-      "learning_rate": 6.414223866313407e-09,
-      "loss": 0.2486,
-      "step": 141170
-    },
-    {
-      "epoch": 4.97,
-      "learning_rate": 6.349859473434028e-09,
-      "loss": 0.2404,
-      "step": 141175
-    },
-    {
-      "epoch": 4.97,
-      "learning_rate": 6.285819606630105e-09,
-      "loss": 0.2474,
-      "step": 141180
-    },
-    {
-      "epoch": 4.97,
-      "learning_rate": 6.2221042667343075e-09,
-      "loss": 0.2452,
-      "step": 141185
-    },
-    {
-      "epoch": 4.97,
-      "learning_rate": 6.15871345457375e-09,
-      "loss": 0.2747,
-      "step": 141190
-    },
-    {
-      "epoch": 4.97,
-      "learning_rate": 6.0956471709699985e-09,
-      "loss": 0.2486,
-      "step": 141195
-    },
-    {
-      "epoch": 4.97,
-      "learning_rate": 6.032905416741841e-09,
-      "loss": 0.2398,
-      "step": 141200
-    },
-    {
-      "epoch": 4.97,
-      "learning_rate": 5.970488192705293e-09,
-      "loss": 0.2424,
-      "step": 141205
-    },
-    {
-      "epoch": 4.97,
-      "learning_rate": 5.908395499670815e-09,
-      "loss": 0.2566,
-      "step": 141210
-    },
-    {
-      "epoch": 4.97,
-      "learning_rate": 5.846627338443322e-09,
-      "loss": 0.2596,
-      "step": 141215
-    },
-    {
-      "epoch": 4.97,
-      "learning_rate": 5.7851837098249486e-09,
-      "loss": 0.2413,
-      "step": 141220
-    },
-    {
-      "epoch": 4.97,
-      "learning_rate": 5.724064614615054e-09,
-      "loss": 0.2422,
-      "step": 141225
-    },
-    {
-      "epoch": 4.97,
-      "learning_rate": 5.6632700536074505e-09,
-      "loss": 0.2516,
-      "step": 141230
-    },
-    {
-      "epoch": 4.97,
-      "learning_rate": 5.602800027587618e-09,
-      "loss": 0.2625,
-      "step": 141235
-    },
-    {
-      "epoch": 4.97,
-      "learning_rate": 5.542654537346592e-09,
-      "loss": 0.2523,
-      "step": 141240
-    },
-    {
-      "epoch": 4.97,
-      "learning_rate": 5.482833583661529e-09,
-      "loss": 0.2497,
-      "step": 141245
-    },
-    {
-      "epoch": 4.97,
-      "learning_rate": 5.423337167309583e-09,
-      "loss": 0.2619,
-      "step": 141250
-    },
-    {
-      "epoch": 4.97,
-      "learning_rate": 5.364165289062362e-09,
-      "loss": 0.253,
-      "step": 141255
-    },
-    {
-      "epoch": 4.97,
-      "learning_rate": 5.305317949691468e-09,
-      "loss": 0.2374,
-      "step": 141260
-    },
-    {
-      "epoch": 4.97,
-      "learning_rate": 5.246795149957407e-09,
-      "loss": 0.2619,
-      "step": 141265
-    },
-    {
-      "epoch": 4.97,
-      "learning_rate": 5.188596890620678e-09,
-      "loss": 0.2507,
-      "step": 141270
-    },
-    {
-      "epoch": 4.97,
-      "learning_rate": 5.130723172439011e-09,
-      "loss": 0.2566,
-      "step": 141275
-    },
-    {
-      "epoch": 4.97,
-      "learning_rate": 5.073173996161806e-09,
-      "loss": 0.2527,
-      "step": 141280
-    },
-    {
-      "epoch": 4.97,
-      "learning_rate": 5.015949362538464e-09,
-      "loss": 0.2586,
-      "step": 141285
-    },
-    {
-      "epoch": 4.97,
-      "learning_rate": 4.959049272310057e-09,
-      "loss": 0.254,
-      "step": 141290
-    },
-    {
-      "epoch": 4.97,
-      "learning_rate": 4.90247372621766e-09,
-      "loss": 0.2495,
-      "step": 141295
-    },
-    {
-      "epoch": 4.97,
-      "learning_rate": 4.846222724991245e-09,
-      "loss": 0.2651,
-      "step": 141300
-    },
-    {
-      "epoch": 4.97,
-      "learning_rate": 4.790296269366334e-09,
-      "loss": 0.2469,
-      "step": 141305
-    },
-    {
-      "epoch": 4.97,
-      "learning_rate": 4.734694360067349e-09,
-      "loss": 0.2607,
-      "step": 141310
-    },
-    {
-      "epoch": 4.97,
-      "learning_rate": 4.679416997813158e-09,
-      "loss": 0.2666,
-      "step": 141315
-    },
-    {
-      "epoch": 4.97,
-      "learning_rate": 4.624464183325406e-09,
-      "loss": 0.2676,
-      "step": 141320
-    },
-    {
-      "epoch": 4.97,
-      "learning_rate": 4.569835917317411e-09,
-      "loss": 0.2595,
-      "step": 141325
-    },
-    {
-      "epoch": 4.97,
-      "learning_rate": 4.5155322004969415e-09,
-      "loss": 0.2523,
-      "step": 141330
-    },
-    {
-      "epoch": 4.97,
-      "learning_rate": 4.461553033571764e-09,
-      "loss": 0.2547,
-      "step": 141335
-    },
-    {
-      "epoch": 4.97,
-      "learning_rate": 4.407898417238543e-09,
-      "loss": 0.2426,
-      "step": 141340
-    },
-    {
-      "epoch": 4.97,
-      "learning_rate": 4.354568352193944e-09,
-      "loss": 0.25,
-      "step": 141345
-    },
-    {
-      "epoch": 4.97,
-      "learning_rate": 4.301562839134632e-09,
-      "loss": 0.2483,
-      "step": 141350
-    },
-    {
-      "epoch": 4.97,
-      "learning_rate": 4.248881878748945e-09,
-      "loss": 0.2567,
-      "step": 141355
-    },
-    {
-      "epoch": 4.97,
-      "learning_rate": 4.196525471714119e-09,
-      "loss": 0.2535,
-      "step": 141360
-    },
-    {
-      "epoch": 4.97,
-      "learning_rate": 4.1444936187184925e-09,
-      "loss": 0.2507,
-      "step": 141365
-    },
-    {
-      "epoch": 4.97,
-      "learning_rate": 4.092786320430974e-09,
-      "loss": 0.2381,
-      "step": 141370
-    },
-    {
-      "epoch": 4.97,
-      "learning_rate": 4.041403577528802e-09,
-      "loss": 0.2421,
-      "step": 141375
-    },
-    {
-      "epoch": 4.97,
-      "learning_rate": 3.9903453906725565e-09,
-      "loss": 0.2582,
-      "step": 141380
-    },
-    {
-      "epoch": 4.97,
-      "learning_rate": 3.939611760531148e-09,
-      "loss": 0.2553,
-      "step": 141385
-    },
-    {
-      "epoch": 4.97,
-      "learning_rate": 3.889202687759607e-09,
-      "loss": 0.2348,
-      "step": 141390
-    },
-    {
-      "epoch": 4.97,
-      "learning_rate": 3.839118173015743e-09,
-      "loss": 0.2817,
-      "step": 141395
-    },
-    {
-      "epoch": 4.97,
-      "learning_rate": 3.789358216946259e-09,
-      "loss": 0.2468,
-      "step": 141400
-    },
-    {
-      "epoch": 4.98,
-      "learning_rate": 3.7399228201978606e-09,
-      "loss": 0.258,
-      "step": 141405
-    },
-    {
-      "epoch": 4.98,
-      "learning_rate": 3.6908119834144772e-09,
-      "loss": 0.2321,
-      "step": 141410
-    },
-    {
-      "epoch": 4.98,
-      "learning_rate": 3.642025707231711e-09,
-      "loss": 0.2494,
-      "step": 141415
-    },
-    {
-      "epoch": 4.98,
-      "learning_rate": 3.593563992285165e-09,
-      "loss": 0.2506,
-      "step": 141420
-    },
-    {
-      "epoch": 4.98,
-      "learning_rate": 3.5454268392048906e-09,
-      "loss": 0.2526,
-      "step": 141425
-    },
-    {
-      "epoch": 4.98,
-      "learning_rate": 3.4976142486098374e-09,
-      "loss": 0.2732,
-      "step": 141430
-    },
-    {
-      "epoch": 4.98,
-      "learning_rate": 3.450126221127281e-09,
-      "loss": 0.224,
-      "step": 141435
-    },
-    {
-      "epoch": 4.98,
-      "learning_rate": 3.40296275737062e-09,
-      "loss": 0.2584,
-      "step": 141440
-    },
-    {
-      "epoch": 4.98,
-      "learning_rate": 3.356123857953253e-09,
-      "loss": 0.2479,
-      "step": 141445
-    },
-    {
-      "epoch": 4.98,
-      "learning_rate": 3.3096095234858016e-09,
-      "loss": 0.2733,
-      "step": 141450
-    },
-    {
-      "epoch": 4.98,
-      "learning_rate": 3.2634197545650115e-09,
-      "loss": 0.2542,
-      "step": 141455
-    },
-    {
-      "epoch": 4.98,
-      "learning_rate": 3.217554551798729e-09,
-      "loss": 0.2575,
-      "step": 141460
-    },
-    {
-      "epoch": 4.98,
-      "learning_rate": 3.1720139157781493e-09,
-      "loss": 0.2401,
-      "step": 141465
-    },
-    {
-      "epoch": 4.98,
-      "learning_rate": 3.126797847094465e-09,
-      "loss": 0.2382,
-      "step": 141470
-    },
-    {
-      "epoch": 4.98,
-      "learning_rate": 3.0819063463360943e-09,
-      "loss": 0.259,
-      "step": 141475
-    },
-    {
-      "epoch": 4.98,
-      "learning_rate": 3.0373394140859048e-09,
-      "loss": 0.2539,
-      "step": 141480
-    },
-    {
-      "epoch": 4.98,
-      "learning_rate": 2.9930970509239874e-09,
-      "loss": 0.243,
-      "step": 141485
-    },
-    {
-      "epoch": 4.98,
-      "learning_rate": 2.9491792574221077e-09,
-      "loss": 0.2471,
-      "step": 141490
-    },
-    {
-      "epoch": 4.98,
-      "learning_rate": 2.9055860341492546e-09,
-      "loss": 0.2529,
-      "step": 141495
-    },
-    {
-      "epoch": 4.98,
-      "learning_rate": 2.8623173816744175e-09,
-      "loss": 0.2396,
-      "step": 141500
-    },
-    {
-      "epoch": 4.98,
-      "eval_loss": 0.2483615279197693,
-      "eval_runtime": 10.5543,
-      "eval_samples_per_second": 9.475,
-      "eval_steps_per_second": 9.475,
-      "step": 141500
-    },
-    {
-      "epoch": 4.98,
-      "learning_rate": 2.8193733005610346e-09,
-      "loss": 0.2468,
-      "step": 141505
-    },
-    {
-      "epoch": 4.98,
-      "learning_rate": 2.7767537913614418e-09,
-      "loss": 0.259,
-      "step": 141510
-    },
-    {
-      "epoch": 4.98,
-      "learning_rate": 2.734458854633526e-09,
-      "loss": 0.2438,
-      "step": 141515
-    },
-    {
-      "epoch": 4.98,
-      "learning_rate": 2.692488490924072e-09,
-      "loss": 0.2723,
-      "step": 141520
-    },
-    {
-      "epoch": 4.98,
-      "learning_rate": 2.6508427007798654e-09,
-      "loss": 0.2448,
-      "step": 141525
-    },
-    {
-      "epoch": 4.98,
-      "learning_rate": 2.6095214847393635e-09,
-      "loss": 0.2505,
-      "step": 141530
-    },
-    {
-      "epoch": 4.98,
-      "learning_rate": 2.568524843341025e-09,
-      "loss": 0.2619,
-      "step": 141535
-    },
-    {
-      "epoch": 4.98,
-      "learning_rate": 2.5278527771177562e-09,
-      "loss": 0.2537,
-      "step": 141540
-    },
-    {
-      "epoch": 4.98,
-      "learning_rate": 2.4875052865941386e-09,
-      "loss": 0.264,
-      "step": 141545
-    },
-    {
-      "epoch": 4.98,
-      "learning_rate": 2.447482372297527e-09,
-      "loss": 0.2323,
-      "step": 141550
-    },
-    {
-      "epoch": 4.98,
-      "learning_rate": 2.4077840347441762e-09,
-      "loss": 0.2491,
-      "step": 141555
-    },
-    {
-      "epoch": 4.98,
-      "learning_rate": 2.36841027445589e-09,
-      "loss": 0.2279,
-      "step": 141560
-    },
-    {
-      "epoch": 4.98,
-      "learning_rate": 2.329361091935045e-09,
-      "loss": 0.2559,
-      "step": 141565
-    },
-    {
-      "epoch": 4.98,
-      "learning_rate": 2.290636487697895e-09,
-      "loss": 0.2693,
-      "step": 141570
-    },
-    {
-      "epoch": 4.98,
-      "learning_rate": 2.2522364622384885e-09,
-      "loss": 0.2513,
-      "step": 141575
-    },
-    {
-      "epoch": 4.98,
-      "learning_rate": 2.2141610160619773e-09,
-      "loss": 0.26,
-      "step": 141580
-    },
-    {
-      "epoch": 4.98,
-      "learning_rate": 2.176410149659636e-09,
-      "loss": 0.2768,
-      "step": 141585
-    },
-    {
-      "epoch": 4.98,
-      "learning_rate": 2.1389838635227366e-09,
-      "loss": 0.2432,
-      "step": 141590
-    },
-    {
-      "epoch": 4.98,
-      "learning_rate": 2.101882158137003e-09,
-      "loss": 0.2819,
-      "step": 141595
-    },
-    {
-      "epoch": 4.98,
-      "learning_rate": 2.0651050339853815e-09,
-      "loss": 0.2579,
-      "step": 141600
-    },
-    {
-      "epoch": 4.98,
-      "learning_rate": 2.028652491542493e-09,
-      "loss": 0.247,
-      "step": 141605
-    },
-    {
-      "epoch": 4.98,
-      "learning_rate": 1.9925245312829577e-09,
-      "loss": 0.2499,
-      "step": 141610
-    },
-    {
-      "epoch": 4.98,
-      "learning_rate": 1.95672115367862e-09,
-      "loss": 0.2774,
-      "step": 141615
-    },
-    {
-      "epoch": 4.98,
-      "learning_rate": 1.921242359190223e-09,
-      "loss": 0.2289,
-      "step": 141620
-    },
-    {
-      "epoch": 4.98,
-      "learning_rate": 1.886088148278509e-09,
-      "loss": 0.2339,
-      "step": 141625
-    },
-    {
-      "epoch": 4.98,
-      "learning_rate": 1.85125852140422e-09,
-      "loss": 0.251,
-      "step": 141630
-    },
-    {
-      "epoch": 4.98,
-      "learning_rate": 1.8167534790169971e-09,
-      "loss": 0.2504,
-      "step": 141635
-    },
-    {
-      "epoch": 4.98,
-      "learning_rate": 1.782573021563705e-09,
-      "loss": 0.2926,
-      "step": 141640
-    },
-    {
-      "epoch": 4.98,
-      "learning_rate": 1.7487171494884325e-09,
-      "loss": 0.242,
-      "step": 141645
-    },
-    {
-      "epoch": 4.98,
-      "learning_rate": 1.7151858632324935e-09,
-      "loss": 0.2579,
-      "step": 141650
-    },
-    {
-      "epoch": 4.98,
-      "learning_rate": 1.6819791632288751e-09,
-      "loss": 0.2323,
-      "step": 141655
-    },
-    {
-      "epoch": 4.98,
-      "learning_rate": 1.6490970499133395e-09,
-      "loss": 0.2269,
-      "step": 141660
-    },
-    {
-      "epoch": 4.98,
-      "learning_rate": 1.6165395237077718e-09,
-      "loss": 0.2234,
-      "step": 141665
-    },
-    {
-      "epoch": 4.98,
-      "learning_rate": 1.5843065850368321e-09,
-      "loss": 0.2495,
-      "step": 141670
-    },
-    {
-      "epoch": 4.98,
-      "learning_rate": 1.5523982343196297e-09,
-      "loss": 0.2542,
-      "step": 141675
-    },
-    {
-      "epoch": 4.98,
-      "learning_rate": 1.5208144719697227e-09,
-      "loss": 0.2523,
-      "step": 141680
-    },
-    {
-      "epoch": 4.98,
-      "learning_rate": 1.4895552983978934e-09,
-      "loss": 0.2379,
-      "step": 141685
-    },
-    {
-      "epoch": 4.99,
-      "learning_rate": 1.4586207140093733e-09,
-      "loss": 0.2469,
-      "step": 141690
-    },
-    {
-      "epoch": 4.99,
-      "learning_rate": 1.4280107192066184e-09,
-      "loss": 0.2736,
-      "step": 141695
-    },
-    {
-      "epoch": 4.99,
-      "learning_rate": 1.3977253143837577e-09,
-      "loss": 0.2477,
-      "step": 141700
-    },
-    {
-      "epoch": 4.99,
-      "learning_rate": 1.3677644999404716e-09,
-      "loss": 0.2599,
-      "step": 141705
-    },
-    {
-      "epoch": 4.99,
-      "learning_rate": 1.3381282762597868e-09,
-      "loss": 0.2482,
-      "step": 141710
-    },
-    {
-      "epoch": 4.99,
-      "learning_rate": 1.3088166437275062e-09,
-      "loss": 0.2591,
-      "step": 141715
-    },
-    {
-      "epoch": 4.99,
-      "learning_rate": 1.2798296027294321e-09,
-      "loss": 0.264,
-      "step": 141720
-    },
-    {
-      "epoch": 4.99,
-      "learning_rate": 1.2511671536347136e-09,
-      "loss": 0.2271,
-      "step": 141725
-    },
-    {
-      "epoch": 4.99,
-      "learning_rate": 1.2228292968208266e-09,
-      "loss": 0.2488,
-      "step": 141730
-    },
-    {
-      "epoch": 4.99,
-      "learning_rate": 1.1948160326513692e-09,
-      "loss": 0.2753,
-      "step": 141735
-    },
-    {
-      "epoch": 4.99,
-      "learning_rate": 1.1671273614927147e-09,
-      "loss": 0.2398,
-      "step": 141740
-    },
-    {
-      "epoch": 4.99,
-      "learning_rate": 1.139763283705686e-09,
-      "loss": 0.246,
-      "step": 141745
-    },
-    {
-      "epoch": 4.99,
-      "learning_rate": 1.1127237996427786e-09,
-      "loss": 0.257,
-      "step": 141750
-    },
-    {
-      "epoch": 4.99,
-      "learning_rate": 1.0860089096564884e-09,
-      "loss": 0.2507,
-      "step": 141755
-    },
-    {
-      "epoch": 4.99,
-      "learning_rate": 1.0596186140937602e-09,
-      "loss": 0.2598,
-      "step": 141760
-    },
-    {
-      "epoch": 4.99,
-      "learning_rate": 1.0335529132987631e-09,
-      "loss": 0.2502,
-      "step": 141765
-    },
-    {
-      "epoch": 4.99,
-      "learning_rate": 1.007811807604564e-09,
-      "loss": 0.2493,
-      "step": 141770
-    },
-    {
-      "epoch": 4.99,
-      "learning_rate": 9.823952973525563e-10,
-      "loss": 0.2634,
-      "step": 141775
-    },
-    {
-      "epoch": 4.99,
-      "learning_rate": 9.573033828674805e-10,
-      "loss": 0.2628,
-      "step": 141780
-    },
-    {
-      "epoch": 4.99,
-      "learning_rate": 9.325360644768522e-10,
-      "loss": 0.247,
-      "step": 141785
-    },
-    {
-      "epoch": 4.99,
-      "learning_rate": 9.080933425026361e-10,
-      "loss": 0.2553,
-      "step": 141790
-    },
-    {
-      "epoch": 4.99,
-      "learning_rate": 8.839752172612459e-10,
-      "loss": 0.2516,
-      "step": 141795
-    },
-    {
-      "epoch": 4.99,
-      "learning_rate": 8.601816890663195e-10,
-      "loss": 0.2546,
-      "step": 141800
-    },
-    {
-      "epoch": 4.99,
-      "learning_rate": 8.367127582259438e-10,
-      "loss": 0.2592,
-      "step": 141805
-    },
-    {
-      "epoch": 4.99,
-      "learning_rate": 8.135684250482056e-10,
-      "loss": 0.2581,
-      "step": 141810
-    },
-    {
-      "epoch": 4.99,
-      "learning_rate": 7.907486898300898e-10,
-      "loss": 0.251,
-      "step": 141815
-    },
-    {
-      "epoch": 4.99,
-      "learning_rate": 7.682535528685808e-10,
-      "loss": 0.2412,
-      "step": 141820
-    },
-    {
-      "epoch": 4.99,
-      "learning_rate": 7.460830144551123e-10,
-      "loss": 0.2584,
-      "step": 141825
-    },
-    {
-      "epoch": 4.99,
-      "learning_rate": 7.242370748811178e-10,
-      "loss": 0.2558,
-      "step": 141830
-    },
-    {
-      "epoch": 4.99,
-      "learning_rate": 7.027157344241531e-10,
-      "loss": 0.2496,
-      "step": 141835
-    },
-    {
-      "epoch": 4.99,
-      "learning_rate": 6.815189933701005e-10,
-      "loss": 0.267,
-      "step": 141840
-    },
-    {
-      "epoch": 4.99,
-      "learning_rate": 6.606468519909648e-10,
-      "loss": 0.2709,
-      "step": 141845
-    },
-    {
-      "epoch": 4.99,
-      "learning_rate": 6.400993105587505e-10,
-      "loss": 0.2453,
-      "step": 141850
-    },
-    {
-      "epoch": 4.99,
-      "learning_rate": 6.198763693399112e-10,
-      "loss": 0.2638,
-      "step": 141855
-    },
-    {
-      "epoch": 4.99,
-      "learning_rate": 5.999780285953493e-10,
-      "loss": 0.2354,
-      "step": 141860
-    },
-    {
-      "epoch": 4.99,
-      "learning_rate": 5.804042885859673e-10,
-      "loss": 0.2624,
-      "step": 141865
-    },
-    {
-      "epoch": 4.99,
-      "learning_rate": 5.611551495643408e-10,
-      "loss": 0.2541,
-      "step": 141870
-    },
-    {
-      "epoch": 4.99,
-      "learning_rate": 5.4223061178027e-10,
-      "loss": 0.2372,
-      "step": 141875
-    },
-    {
-      "epoch": 4.99,
-      "learning_rate": 5.236306754807796e-10,
-      "loss": 0.2381,
-      "step": 141880
-    },
-    {
-      "epoch": 4.99,
-      "learning_rate": 5.053553409045675e-10,
-      "loss": 0.2495,
-      "step": 141885
-    },
-    {
-      "epoch": 4.99,
-      "learning_rate": 4.874046082931072e-10,
-      "loss": 0.2311,
-      "step": 141890
-    },
-    {
-      "epoch": 4.99,
-      "learning_rate": 4.6977847787677e-10,
-      "loss": 0.2347,
-      "step": 141895
-    },
-    {
-      "epoch": 4.99,
-      "learning_rate": 4.524769498859271e-10,
-      "loss": 0.248,
-      "step": 141900
-    },
-    {
-      "epoch": 4.99,
-      "learning_rate": 4.3550002454539886e-10,
-      "loss": 0.2674,
-      "step": 141905
-    },
-    {
-      "epoch": 4.99,
-      "learning_rate": 4.1884770207167855e-10,
-      "loss": 0.2447,
-      "step": 141910
-    },
-    {
-      "epoch": 4.99,
-      "learning_rate": 4.025199826868109e-10,
-      "loss": 0.2451,
-      "step": 141915
-    },
-    {
-      "epoch": 4.99,
-      "learning_rate": 3.865168665989627e-10,
-      "loss": 0.2551,
-      "step": 141920
-    },
-    {
-      "epoch": 4.99,
-      "learning_rate": 3.708383540135252e-10,
-      "loss": 0.2753,
-      "step": 141925
-    },
-    {
-      "epoch": 4.99,
-      "learning_rate": 3.5548444514144075e-10,
-      "loss": 0.2452,
-      "step": 141930
-    },
-    {
-      "epoch": 4.99,
-      "learning_rate": 3.4045514017699845e-10,
-      "loss": 0.2607,
-      "step": 141935
-    },
-    {
-      "epoch": 4.99,
-      "learning_rate": 3.2575043931448723e-10,
-      "loss": 0.2621,
-      "step": 141940
-    },
-    {
-      "epoch": 4.99,
-      "learning_rate": 3.113703427481962e-10,
-      "loss": 0.2502,
-      "step": 141945
-    },
-    {
-      "epoch": 4.99,
-      "learning_rate": 2.973148506640877e-10,
-      "loss": 0.2524,
-      "step": 141950
-    },
-    {
-      "epoch": 4.99,
-      "learning_rate": 2.8358396324257295e-10,
-      "loss": 0.2224,
-      "step": 141955
-    },
-    {
-      "epoch": 4.99,
-      "learning_rate": 2.7017768066406323e-10,
-      "loss": 0.2429,
-      "step": 141960
-    },
-    {
-      "epoch": 4.99,
-      "learning_rate": 2.5709600310341864e-10,
-      "loss": 0.2337,
-      "step": 141965
-    },
-    {
-      "epoch": 4.99,
-      "learning_rate": 2.443389307271726e-10,
-      "loss": 0.2236,
-      "step": 141970
-    },
-    {
-      "epoch": 5.0,
-      "learning_rate": 2.3190646370185865e-10,
-      "loss": 0.2481,
-      "step": 141975
-    },
-    {
-      "epoch": 5.0,
-      "learning_rate": 2.197986021912346e-10,
-      "loss": 0.2354,
-      "step": 141980
-    },
-    {
-      "epoch": 5.0,
-      "learning_rate": 2.0801534635073172e-10,
-      "loss": 0.2381,
-      "step": 141985
-    },
-    {
-      "epoch": 5.0,
-      "learning_rate": 1.9655669633300567e-10,
-      "loss": 0.2499,
-      "step": 141990
-    },
-    {
-      "epoch": 5.0,
-      "learning_rate": 1.8542265228793653e-10,
-      "loss": 0.2834,
-      "step": 141995
-    },
-    {
-      "epoch": 5.0,
-      "learning_rate": 1.7461321435985333e-10,
-      "loss": 0.2573,
-      "step": 142000
-    },
-    {
-      "epoch": 5.0,
-      "eval_loss": 0.24838173389434814,
-      "eval_runtime": 10.5549,
-      "eval_samples_per_second": 9.474,
-      "eval_steps_per_second": 9.474,
-      "step": 142000
     }
   ],
   "logging_steps": 5,
-  "max_steps": 142115,
+  "max_steps": 138735,
   "num_input_tokens_seen": 0,
   "num_train_epochs": 5,
   "save_steps": 500,
-  "total_flos": 4.632594192202221e+19,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 4.156594211931134e+19,
   "train_batch_size": 2,
   "trial_name": null,
   "trial_params": null