diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,64469 @@
+{
+  "best_metric": 0.16217102110385895,
+  "best_model_checkpoint": "wyluilipe/deabuse/checkpoint-9000",
+  "epoch": 2.742032471437162,
+  "eval_steps": 120,
+  "global_step": 9120,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "grad_norm": 13.701467514038086,
+      "learning_rate": 2e-05,
+      "loss": 6.6206,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 39.5434684753418,
+      "learning_rate": 1.9997995389395612e-05,
+      "loss": 4.445,
+      "step": 2
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 83.92737579345703,
+      "learning_rate": 1.9995990778791222e-05,
+      "loss": 4.277,
+      "step": 3
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 19.87366485595703,
+      "learning_rate": 1.9993986168186832e-05,
+      "loss": 4.9229,
+      "step": 4
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 22.43378448486328,
+      "learning_rate": 1.9991981557582442e-05,
+      "loss": 4.7554,
+      "step": 5
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 35.35092544555664,
+      "learning_rate": 1.9989976946978052e-05,
+      "loss": 4.1126,
+      "step": 6
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 133.4622344970703,
+      "learning_rate": 1.998797233637366e-05,
+      "loss": 4.5409,
+      "step": 7
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 20.33665657043457,
+      "learning_rate": 1.9985967725769272e-05,
+      "loss": 4.0595,
+      "step": 8
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 85.67984008789062,
+      "learning_rate": 1.9983963115164882e-05,
+      "loss": 4.9032,
+      "step": 9
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 34.769309997558594,
+      "learning_rate": 1.998195850456049e-05,
+      "loss": 5.5253,
+      "step": 10
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 40.53438949584961,
+      "learning_rate": 1.9979953893956102e-05,
+      "loss": 4.3671,
+      "step": 11
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 21.633525848388672,
+      "learning_rate": 1.997794928335171e-05,
+      "loss": 3.9831,
+      "step": 12
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 20.7197322845459,
+      "learning_rate": 1.997594467274732e-05,
+      "loss": 4.4466,
+      "step": 13
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 110.01347351074219,
+      "learning_rate": 1.9973940062142933e-05,
+      "loss": 3.8518,
+      "step": 14
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 19.917098999023438,
+      "learning_rate": 1.997193545153854e-05,
+      "loss": 3.7986,
+      "step": 15
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 31.238054275512695,
+      "learning_rate": 1.996993084093415e-05,
+      "loss": 3.9456,
+      "step": 16
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 20.226728439331055,
+      "learning_rate": 1.996792623032976e-05,
+      "loss": 4.1858,
+      "step": 17
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 28.916229248046875,
+      "learning_rate": 1.996592161972537e-05,
+      "loss": 4.2707,
+      "step": 18
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 20.103069305419922,
+      "learning_rate": 1.996391700912098e-05,
+      "loss": 5.2733,
+      "step": 19
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 30.481670379638672,
+      "learning_rate": 1.996191239851659e-05,
+      "loss": 3.6844,
+      "step": 20
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 27.793643951416016,
+      "learning_rate": 1.99599077879122e-05,
+      "loss": 4.3403,
+      "step": 21
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 32.221065521240234,
+      "learning_rate": 1.995790317730781e-05,
+      "loss": 3.9117,
+      "step": 22
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 23.525253295898438,
+      "learning_rate": 1.995589856670342e-05,
+      "loss": 3.804,
+      "step": 23
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 16.033878326416016,
+      "learning_rate": 1.995389395609903e-05,
+      "loss": 5.166,
+      "step": 24
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 22.915647506713867,
+      "learning_rate": 1.995188934549464e-05,
+      "loss": 3.5634,
+      "step": 25
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 18.16558837890625,
+      "learning_rate": 1.994988473489025e-05,
+      "loss": 4.9719,
+      "step": 26
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 31.655603408813477,
+      "learning_rate": 1.994788012428586e-05,
+      "loss": 5.1661,
+      "step": 27
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 18.09580421447754,
+      "learning_rate": 1.994587551368147e-05,
+      "loss": 4.4041,
+      "step": 28
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 32.707950592041016,
+      "learning_rate": 1.9943870903077077e-05,
+      "loss": 4.5414,
+      "step": 29
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 53.21809005737305,
+      "learning_rate": 1.994186629247269e-05,
+      "loss": 4.2779,
+      "step": 30
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 36.61603927612305,
+      "learning_rate": 1.9939861681868297e-05,
+      "loss": 4.0579,
+      "step": 31
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 21.24146842956543,
+      "learning_rate": 1.9937857071263907e-05,
+      "loss": 4.4486,
+      "step": 32
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 45.33989715576172,
+      "learning_rate": 1.993585246065952e-05,
+      "loss": 4.7946,
+      "step": 33
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 36.568145751953125,
+      "learning_rate": 1.9933847850055128e-05,
+      "loss": 4.6392,
+      "step": 34
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 19.171186447143555,
+      "learning_rate": 1.9931843239450738e-05,
+      "loss": 4.0399,
+      "step": 35
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 28.191804885864258,
+      "learning_rate": 1.9929838628846348e-05,
+      "loss": 4.9894,
+      "step": 36
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 56.99246597290039,
+      "learning_rate": 1.9927834018241958e-05,
+      "loss": 3.7354,
+      "step": 37
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 25.940950393676758,
+      "learning_rate": 1.9925829407637568e-05,
+      "loss": 3.4817,
+      "step": 38
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 38.012001037597656,
+      "learning_rate": 1.9923824797033178e-05,
+      "loss": 3.5887,
+      "step": 39
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 23.125755310058594,
+      "learning_rate": 1.9921820186428788e-05,
+      "loss": 4.8167,
+      "step": 40
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 25.253414154052734,
+      "learning_rate": 1.9919815575824398e-05,
+      "loss": 3.8167,
+      "step": 41
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 18.642539978027344,
+      "learning_rate": 1.991781096522001e-05,
+      "loss": 3.5483,
+      "step": 42
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 102.15142059326172,
+      "learning_rate": 1.9915806354615615e-05,
+      "loss": 5.1262,
+      "step": 43
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 24.280908584594727,
+      "learning_rate": 1.991380174401123e-05,
+      "loss": 4.2309,
+      "step": 44
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 18.596893310546875,
+      "learning_rate": 1.991179713340684e-05,
+      "loss": 3.3356,
+      "step": 45
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 21.592117309570312,
+      "learning_rate": 1.9909792522802445e-05,
+      "loss": 4.8101,
+      "step": 46
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 19.77571678161621,
+      "learning_rate": 1.990778791219806e-05,
+      "loss": 3.4436,
+      "step": 47
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 47.20704650878906,
+      "learning_rate": 1.9905783301593665e-05,
+      "loss": 4.077,
+      "step": 48
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 21.739601135253906,
+      "learning_rate": 1.9903778690989276e-05,
+      "loss": 4.5314,
+      "step": 49
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 22.48163604736328,
+      "learning_rate": 1.9901774080384886e-05,
+      "loss": 4.1253,
+      "step": 50
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 26.019960403442383,
+      "learning_rate": 1.9899769469780496e-05,
+      "loss": 4.4319,
+      "step": 51
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 21.49262046813965,
+      "learning_rate": 1.9897764859176106e-05,
+      "loss": 4.7591,
+      "step": 52
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 21.852209091186523,
+      "learning_rate": 1.9895760248571716e-05,
+      "loss": 4.5679,
+      "step": 53
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 17.225950241088867,
+      "learning_rate": 1.9893755637967326e-05,
+      "loss": 3.5482,
+      "step": 54
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 17.815263748168945,
+      "learning_rate": 1.9891751027362936e-05,
+      "loss": 3.6293,
+      "step": 55
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 19.188623428344727,
+      "learning_rate": 1.9889746416758546e-05,
+      "loss": 4.1118,
+      "step": 56
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 21.920475006103516,
+      "learning_rate": 1.9887741806154156e-05,
+      "loss": 3.8108,
+      "step": 57
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 24.557518005371094,
+      "learning_rate": 1.9885737195549766e-05,
+      "loss": 3.8317,
+      "step": 58
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 90.77320098876953,
+      "learning_rate": 1.9883732584945376e-05,
+      "loss": 5.1687,
+      "step": 59
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 25.83659553527832,
+      "learning_rate": 1.9881727974340986e-05,
+      "loss": 5.0576,
+      "step": 60
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 19.337127685546875,
+      "learning_rate": 1.9879723363736597e-05,
+      "loss": 3.364,
+      "step": 61
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 19.83724594116211,
+      "learning_rate": 1.9877718753132203e-05,
+      "loss": 3.5426,
+      "step": 62
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 14.76127815246582,
+      "learning_rate": 1.9875714142527817e-05,
+      "loss": 3.8213,
+      "step": 63
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 21.111013412475586,
+      "learning_rate": 1.9873709531923427e-05,
+      "loss": 3.4596,
+      "step": 64
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 63.46014404296875,
+      "learning_rate": 1.9871704921319033e-05,
+      "loss": 4.4058,
+      "step": 65
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 19.21253776550293,
+      "learning_rate": 1.9869700310714647e-05,
+      "loss": 3.8774,
+      "step": 66
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 23.148107528686523,
+      "learning_rate": 1.9867695700110254e-05,
+      "loss": 3.9923,
+      "step": 67
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 25.991294860839844,
+      "learning_rate": 1.9865691089505864e-05,
+      "loss": 4.5205,
+      "step": 68
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 18.949478149414062,
+      "learning_rate": 1.9863686478901477e-05,
+      "loss": 4.678,
+      "step": 69
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 17.614856719970703,
+      "learning_rate": 1.9861681868297084e-05,
+      "loss": 3.758,
+      "step": 70
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 20.783035278320312,
+      "learning_rate": 1.9859677257692694e-05,
+      "loss": 4.5238,
+      "step": 71
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 27.744810104370117,
+      "learning_rate": 1.9857672647088304e-05,
+      "loss": 4.7011,
+      "step": 72
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 22.72287368774414,
+      "learning_rate": 1.9855668036483914e-05,
+      "loss": 3.6715,
+      "step": 73
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 23.866741180419922,
+      "learning_rate": 1.9853663425879524e-05,
+      "loss": 5.4669,
+      "step": 74
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 16.936256408691406,
+      "learning_rate": 1.9851658815275134e-05,
+      "loss": 3.3014,
+      "step": 75
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 22.439456939697266,
+      "learning_rate": 1.9849654204670744e-05,
+      "loss": 3.202,
+      "step": 76
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 16.06572723388672,
+      "learning_rate": 1.9847649594066354e-05,
+      "loss": 3.3253,
+      "step": 77
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 29.42819595336914,
+      "learning_rate": 1.9845644983461965e-05,
+      "loss": 4.1956,
+      "step": 78
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 20.668291091918945,
+      "learning_rate": 1.9843640372857575e-05,
+      "loss": 4.4557,
+      "step": 79
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 21.241540908813477,
+      "learning_rate": 1.9841635762253185e-05,
+      "loss": 2.7819,
+      "step": 80
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 39.1732292175293,
+      "learning_rate": 1.9839631151648795e-05,
+      "loss": 4.044,
+      "step": 81
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 17.961164474487305,
+      "learning_rate": 1.9837626541044405e-05,
+      "loss": 4.0641,
+      "step": 82
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 23.27247428894043,
+      "learning_rate": 1.9835621930440015e-05,
+      "loss": 4.1457,
+      "step": 83
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 33.87324523925781,
+      "learning_rate": 1.983361731983562e-05,
+      "loss": 4.4233,
+      "step": 84
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 27.01508903503418,
+      "learning_rate": 1.9831612709231235e-05,
+      "loss": 3.5391,
+      "step": 85
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 18.453147888183594,
+      "learning_rate": 1.9829608098626842e-05,
+      "loss": 3.8098,
+      "step": 86
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 551.41259765625,
+      "learning_rate": 1.9827603488022452e-05,
+      "loss": 3.3489,
+      "step": 87
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 33.0008544921875,
+      "learning_rate": 1.9825598877418065e-05,
+      "loss": 3.8738,
+      "step": 88
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 80.79717254638672,
+      "learning_rate": 1.9823594266813672e-05,
+      "loss": 3.9937,
+      "step": 89
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 20.0161075592041,
+      "learning_rate": 1.9821589656209282e-05,
+      "loss": 3.2797,
+      "step": 90
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 25.17633056640625,
+      "learning_rate": 1.9819585045604892e-05,
+      "loss": 3.3241,
+      "step": 91
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 50.05910873413086,
+      "learning_rate": 1.9817580435000502e-05,
+      "loss": 5.4063,
+      "step": 92
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 19.56489372253418,
+      "learning_rate": 1.9815575824396112e-05,
+      "loss": 3.3133,
+      "step": 93
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 66.2906265258789,
+      "learning_rate": 1.9813571213791723e-05,
+      "loss": 4.3678,
+      "step": 94
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 17.404077529907227,
+      "learning_rate": 1.9811566603187333e-05,
+      "loss": 3.7368,
+      "step": 95
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 17.154563903808594,
+      "learning_rate": 1.9809561992582943e-05,
+      "loss": 3.6814,
+      "step": 96
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 21.113162994384766,
+      "learning_rate": 1.9807557381978553e-05,
+      "loss": 4.3332,
+      "step": 97
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 35.284725189208984,
+      "learning_rate": 1.9805552771374163e-05,
+      "loss": 3.2572,
+      "step": 98
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 19.46085548400879,
+      "learning_rate": 1.9803548160769773e-05,
+      "loss": 3.5422,
+      "step": 99
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 24.549236297607422,
+      "learning_rate": 1.9801543550165383e-05,
+      "loss": 3.071,
+      "step": 100
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 24.21219253540039,
+      "learning_rate": 1.9799538939560993e-05,
+      "loss": 3.7529,
+      "step": 101
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 19.04281234741211,
+      "learning_rate": 1.9797534328956603e-05,
+      "loss": 3.5407,
+      "step": 102
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 46.35712814331055,
+      "learning_rate": 1.979552971835221e-05,
+      "loss": 3.0938,
+      "step": 103
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 32.1992073059082,
+      "learning_rate": 1.9793525107747823e-05,
+      "loss": 4.1555,
+      "step": 104
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 18.388561248779297,
+      "learning_rate": 1.979152049714343e-05,
+      "loss": 3.25,
+      "step": 105
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 25.422645568847656,
+      "learning_rate": 1.978951588653904e-05,
+      "loss": 3.7947,
+      "step": 106
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 20.530521392822266,
+      "learning_rate": 1.9787511275934654e-05,
+      "loss": 3.3529,
+      "step": 107
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 40.206138610839844,
+      "learning_rate": 1.978550666533026e-05,
+      "loss": 4.6079,
+      "step": 108
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 43.84805679321289,
+      "learning_rate": 1.978350205472587e-05,
+      "loss": 3.6345,
+      "step": 109
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 26.29327392578125,
+      "learning_rate": 1.978149744412148e-05,
+      "loss": 4.9385,
+      "step": 110
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 24.576719284057617,
+      "learning_rate": 1.977949283351709e-05,
+      "loss": 3.7024,
+      "step": 111
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 21.797935485839844,
+      "learning_rate": 1.97774882229127e-05,
+      "loss": 4.0716,
+      "step": 112
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 51.30803298950195,
+      "learning_rate": 1.977548361230831e-05,
+      "loss": 4.4803,
+      "step": 113
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 17.133638381958008,
+      "learning_rate": 1.977347900170392e-05,
+      "loss": 2.996,
+      "step": 114
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 23.30716323852539,
+      "learning_rate": 1.977147439109953e-05,
+      "loss": 3.6316,
+      "step": 115
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 21.08536720275879,
+      "learning_rate": 1.976946978049514e-05,
+      "loss": 3.2101,
+      "step": 116
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 40.39435577392578,
+      "learning_rate": 1.9767465169890748e-05,
+      "loss": 3.9427,
+      "step": 117
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 20.827363967895508,
+      "learning_rate": 1.976546055928636e-05,
+      "loss": 3.7911,
+      "step": 118
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 31.05581283569336,
+      "learning_rate": 1.976345594868197e-05,
+      "loss": 4.0686,
+      "step": 119
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 75.66057586669922,
+      "learning_rate": 1.976145133807758e-05,
+      "loss": 3.659,
+      "step": 120
+    },
+    {
+      "epoch": 0.04,
+      "eval_loss": 1.6635291576385498,
+      "eval_runtime": 43.6414,
+      "eval_samples_per_second": 33.89,
+      "eval_steps_per_second": 33.89,
+      "step": 120
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 14.022979736328125,
+      "learning_rate": 1.975944672747319e-05,
+      "loss": 3.0417,
+      "step": 121
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 38.61516189575195,
+      "learning_rate": 1.9757442116868798e-05,
+      "loss": 4.3155,
+      "step": 122
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 19.140295028686523,
+      "learning_rate": 1.975543750626441e-05,
+      "loss": 4.6368,
+      "step": 123
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 24.628950119018555,
+      "learning_rate": 1.975343289566002e-05,
+      "loss": 4.131,
+      "step": 124
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 32.278045654296875,
+      "learning_rate": 1.975142828505563e-05,
+      "loss": 4.4181,
+      "step": 125
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 25.219432830810547,
+      "learning_rate": 1.9749423674451242e-05,
+      "loss": 3.8382,
+      "step": 126
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 20.332189559936523,
+      "learning_rate": 1.974741906384685e-05,
+      "loss": 2.7729,
+      "step": 127
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 24.03636932373047,
+      "learning_rate": 1.974541445324246e-05,
+      "loss": 3.4281,
+      "step": 128
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 48.14682388305664,
+      "learning_rate": 1.974340984263807e-05,
+      "loss": 3.3327,
+      "step": 129
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 22.612560272216797,
+      "learning_rate": 1.974140523203368e-05,
+      "loss": 2.8426,
+      "step": 130
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 18.185649871826172,
+      "learning_rate": 1.973940062142929e-05,
+      "loss": 3.8089,
+      "step": 131
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 39.714744567871094,
+      "learning_rate": 1.97373960108249e-05,
+      "loss": 3.5672,
+      "step": 132
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 63.07310485839844,
+      "learning_rate": 1.973539140022051e-05,
+      "loss": 3.1297,
+      "step": 133
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 22.210826873779297,
+      "learning_rate": 1.973338678961612e-05,
+      "loss": 5.2622,
+      "step": 134
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 15.51650619506836,
+      "learning_rate": 1.973138217901173e-05,
+      "loss": 3.0659,
+      "step": 135
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 20.847930908203125,
+      "learning_rate": 1.9729377568407336e-05,
+      "loss": 3.1732,
+      "step": 136
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 18.530948638916016,
+      "learning_rate": 1.972737295780295e-05,
+      "loss": 2.7667,
+      "step": 137
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 47.08240509033203,
+      "learning_rate": 1.972536834719856e-05,
+      "loss": 2.5596,
+      "step": 138
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 26.97058868408203,
+      "learning_rate": 1.9723363736594166e-05,
+      "loss": 3.129,
+      "step": 139
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 20.2777099609375,
+      "learning_rate": 1.972135912598978e-05,
+      "loss": 3.4919,
+      "step": 140
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 21.72832679748535,
+      "learning_rate": 1.9719354515385386e-05,
+      "loss": 4.1341,
+      "step": 141
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 17.30972671508789,
+      "learning_rate": 1.9717349904780996e-05,
+      "loss": 3.2444,
+      "step": 142
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 49.03208923339844,
+      "learning_rate": 1.971534529417661e-05,
+      "loss": 4.314,
+      "step": 143
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 26.420650482177734,
+      "learning_rate": 1.9713340683572217e-05,
+      "loss": 3.5411,
+      "step": 144
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 25.890357971191406,
+      "learning_rate": 1.9711336072967827e-05,
+      "loss": 3.741,
+      "step": 145
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 22.66683578491211,
+      "learning_rate": 1.9709331462363437e-05,
+      "loss": 3.116,
+      "step": 146
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 31.934288024902344,
+      "learning_rate": 1.9707326851759047e-05,
+      "loss": 3.8739,
+      "step": 147
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 29.55501937866211,
+      "learning_rate": 1.9705322241154657e-05,
+      "loss": 3.7167,
+      "step": 148
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 21.13062286376953,
+      "learning_rate": 1.9703317630550267e-05,
+      "loss": 3.0495,
+      "step": 149
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 27.444517135620117,
+      "learning_rate": 1.9701313019945877e-05,
+      "loss": 3.2852,
+      "step": 150
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 19.37506675720215,
+      "learning_rate": 1.9699308409341487e-05,
+      "loss": 5.2696,
+      "step": 151
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 49.605247497558594,
+      "learning_rate": 1.9697303798737097e-05,
+      "loss": 3.7375,
+      "step": 152
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 19.60191535949707,
+      "learning_rate": 1.9695299188132707e-05,
+      "loss": 3.6209,
+      "step": 153
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 18.86824607849121,
+      "learning_rate": 1.9693294577528317e-05,
+      "loss": 3.1842,
+      "step": 154
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 30.93740463256836,
+      "learning_rate": 1.9691289966923928e-05,
+      "loss": 3.555,
+      "step": 155
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 24.22440528869629,
+      "learning_rate": 1.9689285356319538e-05,
+      "loss": 3.5906,
+      "step": 156
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 43.011539459228516,
+      "learning_rate": 1.9687280745715148e-05,
+      "loss": 4.5245,
+      "step": 157
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 24.332199096679688,
+      "learning_rate": 1.9685276135110754e-05,
+      "loss": 3.9531,
+      "step": 158
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 37.94855880737305,
+      "learning_rate": 1.9683271524506368e-05,
+      "loss": 4.7006,
+      "step": 159
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 20.99969482421875,
+      "learning_rate": 1.9681266913901975e-05,
+      "loss": 4.7367,
+      "step": 160
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 34.952232360839844,
+      "learning_rate": 1.9679262303297585e-05,
+      "loss": 3.2787,
+      "step": 161
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 25.392833709716797,
+      "learning_rate": 1.9677257692693198e-05,
+      "loss": 5.474,
+      "step": 162
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 20.959766387939453,
+      "learning_rate": 1.9675253082088805e-05,
+      "loss": 4.8595,
+      "step": 163
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 49.37668228149414,
+      "learning_rate": 1.9673248471484415e-05,
+      "loss": 4.2558,
+      "step": 164
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 23.258516311645508,
+      "learning_rate": 1.9671243860880025e-05,
+      "loss": 4.362,
+      "step": 165
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 23.190126419067383,
+      "learning_rate": 1.9669239250275635e-05,
+      "loss": 3.2228,
+      "step": 166
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 18.251544952392578,
+      "learning_rate": 1.9667234639671245e-05,
+      "loss": 4.7009,
+      "step": 167
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 19.142702102661133,
+      "learning_rate": 1.9665230029066855e-05,
+      "loss": 3.9582,
+      "step": 168
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 21.59228515625,
+      "learning_rate": 1.9663225418462465e-05,
+      "loss": 5.7372,
+      "step": 169
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 55.695281982421875,
+      "learning_rate": 1.9661220807858075e-05,
+      "loss": 3.9148,
+      "step": 170
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 68.97881317138672,
+      "learning_rate": 1.9659216197253685e-05,
+      "loss": 4.3682,
+      "step": 171
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 20.77494239807129,
+      "learning_rate": 1.9657211586649296e-05,
+      "loss": 4.0729,
+      "step": 172
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 48.90683364868164,
+      "learning_rate": 1.9655206976044906e-05,
+      "loss": 3.9279,
+      "step": 173
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 27.227113723754883,
+      "learning_rate": 1.9653202365440516e-05,
+      "loss": 4.6083,
+      "step": 174
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 25.539125442504883,
+      "learning_rate": 1.9651197754836126e-05,
+      "loss": 3.8157,
+      "step": 175
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 21.638704299926758,
+      "learning_rate": 1.9649193144231736e-05,
+      "loss": 3.8649,
+      "step": 176
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 29.82769775390625,
+      "learning_rate": 1.9647188533627343e-05,
+      "loss": 4.5611,
+      "step": 177
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 22.91179656982422,
+      "learning_rate": 1.9645183923022956e-05,
+      "loss": 3.083,
+      "step": 178
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 19.676715850830078,
+      "learning_rate": 1.9643179312418563e-05,
+      "loss": 3.8661,
+      "step": 179
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 18.871702194213867,
+      "learning_rate": 1.9641174701814173e-05,
+      "loss": 3.8232,
+      "step": 180
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 33.633079528808594,
+      "learning_rate": 1.9639170091209786e-05,
+      "loss": 3.9483,
+      "step": 181
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 17.3801212310791,
+      "learning_rate": 1.9637165480605393e-05,
+      "loss": 3.4511,
+      "step": 182
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 30.476238250732422,
+      "learning_rate": 1.9635160870001003e-05,
+      "loss": 4.5803,
+      "step": 183
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 20.700347900390625,
+      "learning_rate": 1.9633156259396613e-05,
+      "loss": 3.6991,
+      "step": 184
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 16.463884353637695,
+      "learning_rate": 1.9631151648792223e-05,
+      "loss": 3.2284,
+      "step": 185
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 22.347118377685547,
+      "learning_rate": 1.9629147038187833e-05,
+      "loss": 4.603,
+      "step": 186
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 22.56680679321289,
+      "learning_rate": 1.9627142427583443e-05,
+      "loss": 4.332,
+      "step": 187
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 31.80207061767578,
+      "learning_rate": 1.9625137816979054e-05,
+      "loss": 3.8008,
+      "step": 188
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 22.534713745117188,
+      "learning_rate": 1.9623133206374664e-05,
+      "loss": 3.7199,
+      "step": 189
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 22.314205169677734,
+      "learning_rate": 1.9621128595770274e-05,
+      "loss": 4.3184,
+      "step": 190
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 24.710988998413086,
+      "learning_rate": 1.9619123985165884e-05,
+      "loss": 4.7067,
+      "step": 191
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 34.565242767333984,
+      "learning_rate": 1.9617119374561494e-05,
+      "loss": 3.0013,
+      "step": 192
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 21.252920150756836,
+      "learning_rate": 1.9615114763957104e-05,
+      "loss": 5.5017,
+      "step": 193
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 18.080705642700195,
+      "learning_rate": 1.9613110153352714e-05,
+      "loss": 3.4748,
+      "step": 194
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 20.119901657104492,
+      "learning_rate": 1.9611105542748324e-05,
+      "loss": 4.6922,
+      "step": 195
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 20.794160842895508,
+      "learning_rate": 1.960910093214393e-05,
+      "loss": 4.3524,
+      "step": 196
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 16.995960235595703,
+      "learning_rate": 1.9607096321539544e-05,
+      "loss": 5.311,
+      "step": 197
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 28.07927131652832,
+      "learning_rate": 1.9605091710935154e-05,
+      "loss": 4.5386,
+      "step": 198
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 20.054431915283203,
+      "learning_rate": 1.960308710033076e-05,
+      "loss": 3.8613,
+      "step": 199
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 19.08523941040039,
+      "learning_rate": 1.9601082489726375e-05,
+      "loss": 4.5862,
+      "step": 200
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 20.86153793334961,
+      "learning_rate": 1.959907787912198e-05,
+      "loss": 4.4943,
+      "step": 201
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 17.117141723632812,
+      "learning_rate": 1.959707326851759e-05,
+      "loss": 3.1455,
+      "step": 202
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 20.701730728149414,
+      "learning_rate": 1.95950686579132e-05,
+      "loss": 4.1613,
+      "step": 203
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 17.824867248535156,
+      "learning_rate": 1.959306404730881e-05,
+      "loss": 3.764,
+      "step": 204
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 23.23500633239746,
+      "learning_rate": 1.959105943670442e-05,
+      "loss": 3.647,
+      "step": 205
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 19.766569137573242,
+      "learning_rate": 1.958905482610003e-05,
+      "loss": 3.9145,
+      "step": 206
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 22.38438606262207,
+      "learning_rate": 1.9587050215495642e-05,
+      "loss": 4.4528,
+      "step": 207
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 17.628910064697266,
+      "learning_rate": 1.9585045604891252e-05,
+      "loss": 3.7731,
+      "step": 208
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 18.171613693237305,
+      "learning_rate": 1.9583040994286862e-05,
+      "loss": 4.1904,
+      "step": 209
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 27.142881393432617,
+      "learning_rate": 1.958103638368247e-05,
+      "loss": 4.3347,
+      "step": 210
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 13.47857666015625,
+      "learning_rate": 1.9579031773078082e-05,
+      "loss": 3.2086,
+      "step": 211
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 18.275033950805664,
+      "learning_rate": 1.9577027162473692e-05,
+      "loss": 4.6453,
+      "step": 212
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 17.69455337524414,
+      "learning_rate": 1.95750225518693e-05,
+      "loss": 3.5041,
+      "step": 213
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 16.023883819580078,
+      "learning_rate": 1.9573017941264912e-05,
+      "loss": 4.2656,
+      "step": 214
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 25.941577911376953,
+      "learning_rate": 1.957101333066052e-05,
+      "loss": 3.9683,
+      "step": 215
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 28.56459617614746,
+      "learning_rate": 1.956900872005613e-05,
+      "loss": 4.1389,
+      "step": 216
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 18.201026916503906,
+      "learning_rate": 1.9567004109451743e-05,
+      "loss": 4.2534,
+      "step": 217
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 17.178457260131836,
+      "learning_rate": 1.956499949884735e-05,
+      "loss": 3.8277,
+      "step": 218
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 38.801876068115234,
+      "learning_rate": 1.956299488824296e-05,
+      "loss": 4.1429,
+      "step": 219
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 23.956295013427734,
+      "learning_rate": 1.956099027763857e-05,
+      "loss": 5.4837,
+      "step": 220
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 16.060426712036133,
+      "learning_rate": 1.955898566703418e-05,
+      "loss": 3.7044,
+      "step": 221
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 18.859159469604492,
+      "learning_rate": 1.955698105642979e-05,
+      "loss": 4.168,
+      "step": 222
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 21.508644104003906,
+      "learning_rate": 1.95549764458254e-05,
+      "loss": 4.5285,
+      "step": 223
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 25.023488998413086,
+      "learning_rate": 1.955297183522101e-05,
+      "loss": 3.2762,
+      "step": 224
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 16.727890014648438,
+      "learning_rate": 1.955096722461662e-05,
+      "loss": 3.3409,
+      "step": 225
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 20.456737518310547,
+      "learning_rate": 1.954896261401223e-05,
+      "loss": 4.1665,
+      "step": 226
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 17.329370498657227,
+      "learning_rate": 1.954695800340784e-05,
+      "loss": 3.0992,
+      "step": 227
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 39.43778610229492,
+      "learning_rate": 1.954495339280345e-05,
+      "loss": 4.5199,
+      "step": 228
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 20.120681762695312,
+      "learning_rate": 1.954294878219906e-05,
+      "loss": 4.6025,
+      "step": 229
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 22.665719985961914,
+      "learning_rate": 1.954094417159467e-05,
+      "loss": 4.3815,
+      "step": 230
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 24.487995147705078,
+      "learning_rate": 1.953893956099028e-05,
+      "loss": 4.2686,
+      "step": 231
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 32.88031768798828,
+      "learning_rate": 1.9536934950385887e-05,
+      "loss": 4.3775,
+      "step": 232
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 22.327272415161133,
+      "learning_rate": 1.95349303397815e-05,
+      "loss": 3.7784,
+      "step": 233
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 33.93623352050781,
+      "learning_rate": 1.9532925729177107e-05,
+      "loss": 3.2231,
+      "step": 234
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 24.04096794128418,
+      "learning_rate": 1.9530921118572717e-05,
+      "loss": 4.1805,
+      "step": 235
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 18.985858917236328,
+      "learning_rate": 1.952891650796833e-05,
+      "loss": 4.1365,
+      "step": 236
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 21.616893768310547,
+      "learning_rate": 1.9526911897363937e-05,
+      "loss": 2.4831,
+      "step": 237
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 26.768409729003906,
+      "learning_rate": 1.9524907286759548e-05,
+      "loss": 4.6538,
+      "step": 238
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 21.572690963745117,
+      "learning_rate": 1.9522902676155158e-05,
+      "loss": 3.7805,
+      "step": 239
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 14.675576210021973,
+      "learning_rate": 1.9520898065550768e-05,
+      "loss": 3.3591,
+      "step": 240
+    },
+    {
+      "epoch": 0.07,
+      "eval_loss": 1.399770975112915,
+      "eval_runtime": 43.6698,
+      "eval_samples_per_second": 33.868,
+      "eval_steps_per_second": 33.868,
+      "step": 240
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 39.946598052978516,
+      "learning_rate": 1.9518893454946378e-05,
+      "loss": 3.3287,
+      "step": 241
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 22.521690368652344,
+      "learning_rate": 1.9516888844341988e-05,
+      "loss": 3.3806,
+      "step": 242
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 18.76473045349121,
+      "learning_rate": 1.9514884233737598e-05,
+      "loss": 4.1907,
+      "step": 243
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 17.154766082763672,
+      "learning_rate": 1.9512879623133208e-05,
+      "loss": 3.7032,
+      "step": 244
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 25.60683250427246,
+      "learning_rate": 1.9510875012528818e-05,
+      "loss": 4.4485,
+      "step": 245
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 29.663965225219727,
+      "learning_rate": 1.9508870401924428e-05,
+      "loss": 4.5449,
+      "step": 246
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 24.08624267578125,
+      "learning_rate": 1.950686579132004e-05,
+      "loss": 4.6023,
+      "step": 247
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 17.730571746826172,
+      "learning_rate": 1.950486118071565e-05,
+      "loss": 4.7885,
+      "step": 248
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 16.919239044189453,
+      "learning_rate": 1.950285657011126e-05,
+      "loss": 3.2188,
+      "step": 249
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 16.421016693115234,
+      "learning_rate": 1.950085195950687e-05,
+      "loss": 4.6058,
+      "step": 250
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 26.03248405456543,
+      "learning_rate": 1.9498847348902475e-05,
+      "loss": 5.1777,
+      "step": 251
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 24.347097396850586,
+      "learning_rate": 1.949684273829809e-05,
+      "loss": 4.5525,
+      "step": 252
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 16.834897994995117,
+      "learning_rate": 1.9494838127693695e-05,
+      "loss": 3.7237,
+      "step": 253
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 14.267677307128906,
+      "learning_rate": 1.9492833517089306e-05,
+      "loss": 4.2919,
+      "step": 254
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 14.566679954528809,
+      "learning_rate": 1.949082890648492e-05,
+      "loss": 4.2664,
+      "step": 255
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 20.296588897705078,
+      "learning_rate": 1.9488824295880526e-05,
+      "loss": 4.1636,
+      "step": 256
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 19.508743286132812,
+      "learning_rate": 1.9486819685276136e-05,
+      "loss": 3.5817,
+      "step": 257
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 22.892982482910156,
+      "learning_rate": 1.9484815074671746e-05,
+      "loss": 4.1911,
+      "step": 258
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 20.90638542175293,
+      "learning_rate": 1.9482810464067356e-05,
+      "loss": 3.3314,
+      "step": 259
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 16.314008712768555,
+      "learning_rate": 1.9480805853462966e-05,
+      "loss": 3.5795,
+      "step": 260
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 19.024072647094727,
+      "learning_rate": 1.9478801242858576e-05,
+      "loss": 3.4856,
+      "step": 261
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 15.41294002532959,
+      "learning_rate": 1.9476796632254186e-05,
+      "loss": 3.9168,
+      "step": 262
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 27.279674530029297,
+      "learning_rate": 1.9474792021649796e-05,
+      "loss": 4.4373,
+      "step": 263
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 26.365346908569336,
+      "learning_rate": 1.9472787411045406e-05,
+      "loss": 4.3507,
+      "step": 264
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 18.319107055664062,
+      "learning_rate": 1.9470782800441016e-05,
+      "loss": 4.1936,
+      "step": 265
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 16.385744094848633,
+      "learning_rate": 1.9468778189836627e-05,
+      "loss": 4.3412,
+      "step": 266
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 17.853548049926758,
+      "learning_rate": 1.9466773579232237e-05,
+      "loss": 4.4193,
+      "step": 267
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 23.52976417541504,
+      "learning_rate": 1.9464768968627847e-05,
+      "loss": 5.0003,
+      "step": 268
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 20.375717163085938,
+      "learning_rate": 1.9462764358023457e-05,
+      "loss": 3.6193,
+      "step": 269
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 19.429637908935547,
+      "learning_rate": 1.9460759747419063e-05,
+      "loss": 4.0949,
+      "step": 270
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 16.466793060302734,
+      "learning_rate": 1.9458755136814677e-05,
+      "loss": 3.3825,
+      "step": 271
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 28.024784088134766,
+      "learning_rate": 1.9456750526210287e-05,
+      "loss": 4.1028,
+      "step": 272
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 20.214069366455078,
+      "learning_rate": 1.9454745915605894e-05,
+      "loss": 3.611,
+      "step": 273
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 17.36444854736328,
+      "learning_rate": 1.9452741305001507e-05,
+      "loss": 3.7358,
+      "step": 274
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 17.746870040893555,
+      "learning_rate": 1.9450736694397114e-05,
+      "loss": 3.527,
+      "step": 275
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 15.168383598327637,
+      "learning_rate": 1.9448732083792724e-05,
+      "loss": 3.3192,
+      "step": 276
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 24.655101776123047,
+      "learning_rate": 1.9446727473188334e-05,
+      "loss": 5.3757,
+      "step": 277
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 22.04741668701172,
+      "learning_rate": 1.9444722862583944e-05,
+      "loss": 3.5021,
+      "step": 278
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 15.345871925354004,
+      "learning_rate": 1.9442718251979554e-05,
+      "loss": 4.5815,
+      "step": 279
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 17.076583862304688,
+      "learning_rate": 1.9440713641375164e-05,
+      "loss": 4.7359,
+      "step": 280
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 17.223661422729492,
+      "learning_rate": 1.9438709030770774e-05,
+      "loss": 3.6285,
+      "step": 281
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 19.869312286376953,
+      "learning_rate": 1.9436704420166385e-05,
+      "loss": 3.9027,
+      "step": 282
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 20.547405242919922,
+      "learning_rate": 1.9434699809561995e-05,
+      "loss": 3.6283,
+      "step": 283
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 18.522144317626953,
+      "learning_rate": 1.9432695198957605e-05,
+      "loss": 4.2172,
+      "step": 284
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 30.722158432006836,
+      "learning_rate": 1.9430690588353215e-05,
+      "loss": 3.6669,
+      "step": 285
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 16.493730545043945,
+      "learning_rate": 1.9428685977748825e-05,
+      "loss": 4.1544,
+      "step": 286
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 15.535743713378906,
+      "learning_rate": 1.942668136714443e-05,
+      "loss": 4.1201,
+      "step": 287
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 136.23187255859375,
+      "learning_rate": 1.9424676756540045e-05,
+      "loss": 4.5668,
+      "step": 288
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 15.681511878967285,
+      "learning_rate": 1.9422672145935652e-05,
+      "loss": 4.1992,
+      "step": 289
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 12.06187915802002,
+      "learning_rate": 1.9420667535331262e-05,
+      "loss": 2.6199,
+      "step": 290
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 14.176081657409668,
+      "learning_rate": 1.9418662924726875e-05,
+      "loss": 2.6663,
+      "step": 291
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 16.331350326538086,
+      "learning_rate": 1.9416658314122482e-05,
+      "loss": 4.1395,
+      "step": 292
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 25.406064987182617,
+      "learning_rate": 1.9414653703518095e-05,
+      "loss": 3.2631,
+      "step": 293
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 25.89323616027832,
+      "learning_rate": 1.9412649092913702e-05,
+      "loss": 4.4711,
+      "step": 294
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 12.358001708984375,
+      "learning_rate": 1.9410644482309312e-05,
+      "loss": 4.0686,
+      "step": 295
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 17.452009201049805,
+      "learning_rate": 1.9408639871704922e-05,
+      "loss": 3.6253,
+      "step": 296
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 28.617233276367188,
+      "learning_rate": 1.9406635261100532e-05,
+      "loss": 4.9357,
+      "step": 297
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 17.00152587890625,
+      "learning_rate": 1.9404630650496142e-05,
+      "loss": 3.7579,
+      "step": 298
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 17.73528289794922,
+      "learning_rate": 1.9402626039891753e-05,
+      "loss": 3.7798,
+      "step": 299
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 15.016806602478027,
+      "learning_rate": 1.9400621429287363e-05,
+      "loss": 4.3192,
+      "step": 300
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 12.723051071166992,
+      "learning_rate": 1.9398616818682973e-05,
+      "loss": 3.8939,
+      "step": 301
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 14.875953674316406,
+      "learning_rate": 1.9396612208078583e-05,
+      "loss": 2.8365,
+      "step": 302
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 16.63105583190918,
+      "learning_rate": 1.9394607597474193e-05,
+      "loss": 3.7121,
+      "step": 303
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 22.02267074584961,
+      "learning_rate": 1.9392602986869803e-05,
+      "loss": 5.0273,
+      "step": 304
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 44.309234619140625,
+      "learning_rate": 1.9390598376265413e-05,
+      "loss": 4.5806,
+      "step": 305
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 20.403287887573242,
+      "learning_rate": 1.938859376566102e-05,
+      "loss": 3.2671,
+      "step": 306
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 15.081746101379395,
+      "learning_rate": 1.9386589155056633e-05,
+      "loss": 4.2114,
+      "step": 307
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 16.417646408081055,
+      "learning_rate": 1.938458454445224e-05,
+      "loss": 3.878,
+      "step": 308
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 20.749244689941406,
+      "learning_rate": 1.938257993384785e-05,
+      "loss": 3.3179,
+      "step": 309
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 21.518075942993164,
+      "learning_rate": 1.9380575323243463e-05,
+      "loss": 3.4908,
+      "step": 310
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 17.475610733032227,
+      "learning_rate": 1.937857071263907e-05,
+      "loss": 4.1835,
+      "step": 311
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 20.958995819091797,
+      "learning_rate": 1.937656610203468e-05,
+      "loss": 4.5692,
+      "step": 312
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 16.16900062561035,
+      "learning_rate": 1.937456149143029e-05,
+      "loss": 2.9379,
+      "step": 313
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 29.76364517211914,
+      "learning_rate": 1.93725568808259e-05,
+      "loss": 4.0285,
+      "step": 314
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 14.153215408325195,
+      "learning_rate": 1.937055227022151e-05,
+      "loss": 3.6143,
+      "step": 315
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 25.9129581451416,
+      "learning_rate": 1.936854765961712e-05,
+      "loss": 4.9313,
+      "step": 316
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 12.36880874633789,
+      "learning_rate": 1.936654304901273e-05,
+      "loss": 3.5271,
+      "step": 317
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 20.27505874633789,
+      "learning_rate": 1.936453843840834e-05,
+      "loss": 3.8401,
+      "step": 318
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 20.562847137451172,
+      "learning_rate": 1.936253382780395e-05,
+      "loss": 4.0286,
+      "step": 319
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 23.03611183166504,
+      "learning_rate": 1.936052921719956e-05,
+      "loss": 3.1755,
+      "step": 320
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 24.045217514038086,
+      "learning_rate": 1.935852460659517e-05,
+      "loss": 4.1921,
+      "step": 321
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 15.012109756469727,
+      "learning_rate": 1.935651999599078e-05,
+      "loss": 3.9538,
+      "step": 322
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 16.769580841064453,
+      "learning_rate": 1.935451538538639e-05,
+      "loss": 4.1792,
+      "step": 323
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 15.990212440490723,
+      "learning_rate": 1.9352510774782e-05,
+      "loss": 3.3792,
+      "step": 324
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 17.586837768554688,
+      "learning_rate": 1.9350506164177608e-05,
+      "loss": 3.8851,
+      "step": 325
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 28.66331672668457,
+      "learning_rate": 1.934850155357322e-05,
+      "loss": 4.4007,
+      "step": 326
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 19.1878604888916,
+      "learning_rate": 1.934649694296883e-05,
+      "loss": 3.7238,
+      "step": 327
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 27.458173751831055,
+      "learning_rate": 1.9344492332364438e-05,
+      "loss": 3.6198,
+      "step": 328
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 23.124168395996094,
+      "learning_rate": 1.934248772176005e-05,
+      "loss": 3.9671,
+      "step": 329
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 22.697099685668945,
+      "learning_rate": 1.934048311115566e-05,
+      "loss": 4.6046,
+      "step": 330
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 16.66628074645996,
+      "learning_rate": 1.933847850055127e-05,
+      "loss": 3.4079,
+      "step": 331
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 21.602243423461914,
+      "learning_rate": 1.933647388994688e-05,
+      "loss": 2.6947,
+      "step": 332
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 19.987550735473633,
+      "learning_rate": 1.933446927934249e-05,
+      "loss": 4.0264,
+      "step": 333
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 27.548095703125,
+      "learning_rate": 1.93324646687381e-05,
+      "loss": 3.6234,
+      "step": 334
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 16.985288619995117,
+      "learning_rate": 1.933046005813371e-05,
+      "loss": 3.5273,
+      "step": 335
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 249.91415405273438,
+      "learning_rate": 1.932845544752932e-05,
+      "loss": 3.0628,
+      "step": 336
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 22.47792625427246,
+      "learning_rate": 1.932645083692493e-05,
+      "loss": 3.7402,
+      "step": 337
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 28.15104866027832,
+      "learning_rate": 1.932444622632054e-05,
+      "loss": 3.5902,
+      "step": 338
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 15.687856674194336,
+      "learning_rate": 1.932244161571615e-05,
+      "loss": 3.4816,
+      "step": 339
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 15.60968017578125,
+      "learning_rate": 1.932043700511176e-05,
+      "loss": 3.8098,
+      "step": 340
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 21.058176040649414,
+      "learning_rate": 1.931843239450737e-05,
+      "loss": 4.0355,
+      "step": 341
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 52.26857376098633,
+      "learning_rate": 1.931642778390298e-05,
+      "loss": 5.9303,
+      "step": 342
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 14.664571762084961,
+      "learning_rate": 1.931442317329859e-05,
+      "loss": 3.5722,
+      "step": 343
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 11.48156452178955,
+      "learning_rate": 1.9312418562694196e-05,
+      "loss": 3.8408,
+      "step": 344
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 18.76517677307129,
+      "learning_rate": 1.931041395208981e-05,
+      "loss": 3.5406,
+      "step": 345
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 15.111738204956055,
+      "learning_rate": 1.930840934148542e-05,
+      "loss": 3.6611,
+      "step": 346
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 11.203923225402832,
+      "learning_rate": 1.9306404730881026e-05,
+      "loss": 3.0199,
+      "step": 347
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 21.51028060913086,
+      "learning_rate": 1.930440012027664e-05,
+      "loss": 3.1991,
+      "step": 348
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 21.56675910949707,
+      "learning_rate": 1.9302395509672247e-05,
+      "loss": 3.6963,
+      "step": 349
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 13.743776321411133,
+      "learning_rate": 1.9300390899067857e-05,
+      "loss": 3.6142,
+      "step": 350
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 15.742630004882812,
+      "learning_rate": 1.9298386288463467e-05,
+      "loss": 2.8555,
+      "step": 351
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 14.802862167358398,
+      "learning_rate": 1.9296381677859077e-05,
+      "loss": 2.9698,
+      "step": 352
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 16.21837615966797,
+      "learning_rate": 1.9294377067254687e-05,
+      "loss": 5.3529,
+      "step": 353
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 14.180800437927246,
+      "learning_rate": 1.9292372456650297e-05,
+      "loss": 3.7083,
+      "step": 354
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 16.493022918701172,
+      "learning_rate": 1.9290367846045907e-05,
+      "loss": 2.6302,
+      "step": 355
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 19.73296356201172,
+      "learning_rate": 1.9288363235441517e-05,
+      "loss": 3.4556,
+      "step": 356
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 20.31797218322754,
+      "learning_rate": 1.9286358624837127e-05,
+      "loss": 4.2031,
+      "step": 357
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 24.127195358276367,
+      "learning_rate": 1.9284354014232737e-05,
+      "loss": 4.2493,
+      "step": 358
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 33.403282165527344,
+      "learning_rate": 1.9282349403628347e-05,
+      "loss": 4.6944,
+      "step": 359
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 52.27415466308594,
+      "learning_rate": 1.9280344793023958e-05,
+      "loss": 3.5661,
+      "step": 360
+    },
+    {
+      "epoch": 0.11,
+      "eval_loss": 1.2527186870574951,
+      "eval_runtime": 43.7248,
+      "eval_samples_per_second": 33.825,
+      "eval_steps_per_second": 33.825,
+      "step": 360
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 18.096208572387695,
+      "learning_rate": 1.9278340182419568e-05,
+      "loss": 4.3089,
+      "step": 361
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 12.513582229614258,
+      "learning_rate": 1.9276335571815178e-05,
+      "loss": 3.7448,
+      "step": 362
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 18.849184036254883,
+      "learning_rate": 1.9274330961210784e-05,
+      "loss": 4.8774,
+      "step": 363
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 14.817676544189453,
+      "learning_rate": 1.9272326350606398e-05,
+      "loss": 4.2751,
+      "step": 364
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 24.063844680786133,
+      "learning_rate": 1.9270321740002008e-05,
+      "loss": 3.8558,
+      "step": 365
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 18.589746475219727,
+      "learning_rate": 1.9268317129397615e-05,
+      "loss": 2.9172,
+      "step": 366
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 51.5893669128418,
+      "learning_rate": 1.9266312518793228e-05,
+      "loss": 3.2441,
+      "step": 367
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 18.138118743896484,
+      "learning_rate": 1.9264307908188835e-05,
+      "loss": 2.9894,
+      "step": 368
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 14.857394218444824,
+      "learning_rate": 1.9262303297584445e-05,
+      "loss": 3.1436,
+      "step": 369
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 17.468170166015625,
+      "learning_rate": 1.926029868698006e-05,
+      "loss": 3.2683,
+      "step": 370
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 17.41706657409668,
+      "learning_rate": 1.9258294076375665e-05,
+      "loss": 3.6597,
+      "step": 371
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 15.607809066772461,
+      "learning_rate": 1.9256289465771275e-05,
+      "loss": 3.5317,
+      "step": 372
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 25.48333168029785,
+      "learning_rate": 1.9254284855166885e-05,
+      "loss": 4.017,
+      "step": 373
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 26.36655044555664,
+      "learning_rate": 1.9252280244562495e-05,
+      "loss": 3.1871,
+      "step": 374
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 29.703283309936523,
+      "learning_rate": 1.9250275633958105e-05,
+      "loss": 3.56,
+      "step": 375
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 12.352496147155762,
+      "learning_rate": 1.9248271023353715e-05,
+      "loss": 3.2961,
+      "step": 376
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 24.971912384033203,
+      "learning_rate": 1.9246266412749326e-05,
+      "loss": 3.8872,
+      "step": 377
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 20.017274856567383,
+      "learning_rate": 1.9244261802144936e-05,
+      "loss": 4.3486,
+      "step": 378
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 16.611631393432617,
+      "learning_rate": 1.9242257191540546e-05,
+      "loss": 4.1083,
+      "step": 379
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 94.53077697753906,
+      "learning_rate": 1.9240252580936152e-05,
+      "loss": 3.7256,
+      "step": 380
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 14.142595291137695,
+      "learning_rate": 1.9238247970331766e-05,
+      "loss": 3.9881,
+      "step": 381
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 15.951706886291504,
+      "learning_rate": 1.9236243359727373e-05,
+      "loss": 3.6539,
+      "step": 382
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 16.317209243774414,
+      "learning_rate": 1.9234238749122983e-05,
+      "loss": 4.028,
+      "step": 383
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 15.425968170166016,
+      "learning_rate": 1.9232234138518596e-05,
+      "loss": 2.9222,
+      "step": 384
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 15.79262924194336,
+      "learning_rate": 1.9230229527914203e-05,
+      "loss": 3.1179,
+      "step": 385
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 19.95915985107422,
+      "learning_rate": 1.9228224917309813e-05,
+      "loss": 3.9919,
+      "step": 386
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 14.439689636230469,
+      "learning_rate": 1.9226220306705423e-05,
+      "loss": 3.257,
+      "step": 387
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 16.699209213256836,
+      "learning_rate": 1.9224215696101033e-05,
+      "loss": 3.8755,
+      "step": 388
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 24.074472427368164,
+      "learning_rate": 1.9222211085496643e-05,
+      "loss": 4.0977,
+      "step": 389
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 15.57771110534668,
+      "learning_rate": 1.9220206474892253e-05,
+      "loss": 4.4361,
+      "step": 390
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 13.52245807647705,
+      "learning_rate": 1.9218201864287863e-05,
+      "loss": 3.7686,
+      "step": 391
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 19.746614456176758,
+      "learning_rate": 1.9216197253683473e-05,
+      "loss": 3.819,
+      "step": 392
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 14.887005805969238,
+      "learning_rate": 1.9214192643079084e-05,
+      "loss": 3.8956,
+      "step": 393
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 18.658023834228516,
+      "learning_rate": 1.9212188032474694e-05,
+      "loss": 3.9219,
+      "step": 394
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 14.493846893310547,
+      "learning_rate": 1.9210183421870304e-05,
+      "loss": 4.26,
+      "step": 395
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 16.01069450378418,
+      "learning_rate": 1.9208178811265914e-05,
+      "loss": 2.9605,
+      "step": 396
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 24.826866149902344,
+      "learning_rate": 1.9206174200661524e-05,
+      "loss": 4.3274,
+      "step": 397
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 10.156407356262207,
+      "learning_rate": 1.9204169590057134e-05,
+      "loss": 3.2431,
+      "step": 398
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 21.510562896728516,
+      "learning_rate": 1.920216497945274e-05,
+      "loss": 4.4685,
+      "step": 399
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 25.705717086791992,
+      "learning_rate": 1.9200160368848354e-05,
+      "loss": 4.3723,
+      "step": 400
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 15.125773429870605,
+      "learning_rate": 1.9198155758243964e-05,
+      "loss": 3.9991,
+      "step": 401
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 11.93122386932373,
+      "learning_rate": 1.919615114763957e-05,
+      "loss": 3.5661,
+      "step": 402
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 9.818219184875488,
+      "learning_rate": 1.9194146537035184e-05,
+      "loss": 3.4324,
+      "step": 403
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 23.427623748779297,
+      "learning_rate": 1.919214192643079e-05,
+      "loss": 4.2537,
+      "step": 404
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 18.164058685302734,
+      "learning_rate": 1.91901373158264e-05,
+      "loss": 3.2003,
+      "step": 405
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 20.42031478881836,
+      "learning_rate": 1.918813270522201e-05,
+      "loss": 2.4072,
+      "step": 406
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 21.685937881469727,
+      "learning_rate": 1.918612809461762e-05,
+      "loss": 3.5641,
+      "step": 407
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 19.072772979736328,
+      "learning_rate": 1.918412348401323e-05,
+      "loss": 3.8314,
+      "step": 408
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 11.509551048278809,
+      "learning_rate": 1.918211887340884e-05,
+      "loss": 2.837,
+      "step": 409
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 14.759239196777344,
+      "learning_rate": 1.918011426280445e-05,
+      "loss": 3.501,
+      "step": 410
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 25.07742691040039,
+      "learning_rate": 1.917810965220006e-05,
+      "loss": 3.0863,
+      "step": 411
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 17.365577697753906,
+      "learning_rate": 1.9176105041595672e-05,
+      "loss": 3.6191,
+      "step": 412
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 15.148959159851074,
+      "learning_rate": 1.9174100430991282e-05,
+      "loss": 3.6065,
+      "step": 413
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 22.2659854888916,
+      "learning_rate": 1.9172095820386892e-05,
+      "loss": 3.7687,
+      "step": 414
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 10.9700927734375,
+      "learning_rate": 1.9170091209782502e-05,
+      "loss": 3.6995,
+      "step": 415
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 16.873733520507812,
+      "learning_rate": 1.9168086599178112e-05,
+      "loss": 2.3494,
+      "step": 416
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 22.604717254638672,
+      "learning_rate": 1.9166081988573722e-05,
+      "loss": 3.417,
+      "step": 417
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 16.342870712280273,
+      "learning_rate": 1.916407737796933e-05,
+      "loss": 3.7907,
+      "step": 418
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 43.08720016479492,
+      "learning_rate": 1.9162072767364942e-05,
+      "loss": 2.4359,
+      "step": 419
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 18.63433074951172,
+      "learning_rate": 1.9160068156760552e-05,
+      "loss": 4.5936,
+      "step": 420
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 20.13007164001465,
+      "learning_rate": 1.915806354615616e-05,
+      "loss": 4.5972,
+      "step": 421
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 17.850637435913086,
+      "learning_rate": 1.9156058935551773e-05,
+      "loss": 3.1744,
+      "step": 422
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 15.922115325927734,
+      "learning_rate": 1.915405432494738e-05,
+      "loss": 3.862,
+      "step": 423
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 20.92375946044922,
+      "learning_rate": 1.915204971434299e-05,
+      "loss": 2.907,
+      "step": 424
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 18.767166137695312,
+      "learning_rate": 1.91500451037386e-05,
+      "loss": 4.3281,
+      "step": 425
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 16.2836856842041,
+      "learning_rate": 1.914804049313421e-05,
+      "loss": 3.5196,
+      "step": 426
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 13.612652778625488,
+      "learning_rate": 1.914603588252982e-05,
+      "loss": 4.0671,
+      "step": 427
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 20.497167587280273,
+      "learning_rate": 1.914403127192543e-05,
+      "loss": 4.5793,
+      "step": 428
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 16.230653762817383,
+      "learning_rate": 1.914202666132104e-05,
+      "loss": 4.3463,
+      "step": 429
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 17.90238380432129,
+      "learning_rate": 1.914002205071665e-05,
+      "loss": 3.6869,
+      "step": 430
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 23.716754913330078,
+      "learning_rate": 1.913801744011226e-05,
+      "loss": 3.8698,
+      "step": 431
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 17.085433959960938,
+      "learning_rate": 1.913601282950787e-05,
+      "loss": 2.4927,
+      "step": 432
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 19.192747116088867,
+      "learning_rate": 1.913400821890348e-05,
+      "loss": 3.9132,
+      "step": 433
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 13.243489265441895,
+      "learning_rate": 1.913200360829909e-05,
+      "loss": 2.9527,
+      "step": 434
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 15.914332389831543,
+      "learning_rate": 1.91299989976947e-05,
+      "loss": 3.9814,
+      "step": 435
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 16.82046127319336,
+      "learning_rate": 1.912799438709031e-05,
+      "loss": 3.2976,
+      "step": 436
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 30.610519409179688,
+      "learning_rate": 1.9125989776485917e-05,
+      "loss": 2.8746,
+      "step": 437
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 14.973118782043457,
+      "learning_rate": 1.912398516588153e-05,
+      "loss": 3.4231,
+      "step": 438
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 17.690595626831055,
+      "learning_rate": 1.912198055527714e-05,
+      "loss": 2.8452,
+      "step": 439
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 13.494382858276367,
+      "learning_rate": 1.9119975944672747e-05,
+      "loss": 3.4656,
+      "step": 440
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 15.88420581817627,
+      "learning_rate": 1.911797133406836e-05,
+      "loss": 3.4494,
+      "step": 441
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 15.013689041137695,
+      "learning_rate": 1.9115966723463967e-05,
+      "loss": 3.7962,
+      "step": 442
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 19.49163246154785,
+      "learning_rate": 1.9113962112859578e-05,
+      "loss": 3.5861,
+      "step": 443
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 14.526626586914062,
+      "learning_rate": 1.911195750225519e-05,
+      "loss": 4.3179,
+      "step": 444
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 14.432934761047363,
+      "learning_rate": 1.9109952891650798e-05,
+      "loss": 3.6713,
+      "step": 445
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 17.720027923583984,
+      "learning_rate": 1.9107948281046408e-05,
+      "loss": 2.9935,
+      "step": 446
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 30.397794723510742,
+      "learning_rate": 1.9105943670442018e-05,
+      "loss": 3.4623,
+      "step": 447
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 14.584136009216309,
+      "learning_rate": 1.9103939059837628e-05,
+      "loss": 3.4608,
+      "step": 448
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 10.301027297973633,
+      "learning_rate": 1.9101934449233238e-05,
+      "loss": 3.1884,
+      "step": 449
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 17.015613555908203,
+      "learning_rate": 1.9099929838628848e-05,
+      "loss": 3.9679,
+      "step": 450
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 13.941701889038086,
+      "learning_rate": 1.9097925228024458e-05,
+      "loss": 3.3073,
+      "step": 451
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 16.194982528686523,
+      "learning_rate": 1.909592061742007e-05,
+      "loss": 3.6541,
+      "step": 452
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 19.734464645385742,
+      "learning_rate": 1.909391600681568e-05,
+      "loss": 4.4298,
+      "step": 453
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 12.222415924072266,
+      "learning_rate": 1.9091911396211285e-05,
+      "loss": 2.5833,
+      "step": 454
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 21.09163475036621,
+      "learning_rate": 1.90899067856069e-05,
+      "loss": 3.8562,
+      "step": 455
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 14.021400451660156,
+      "learning_rate": 1.908790217500251e-05,
+      "loss": 3.9132,
+      "step": 456
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 19.492759704589844,
+      "learning_rate": 1.9085897564398115e-05,
+      "loss": 4.7376,
+      "step": 457
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 13.180451393127441,
+      "learning_rate": 1.908389295379373e-05,
+      "loss": 2.8013,
+      "step": 458
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 19.046419143676758,
+      "learning_rate": 1.9081888343189336e-05,
+      "loss": 3.2353,
+      "step": 459
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 24.7363338470459,
+      "learning_rate": 1.9079883732584946e-05,
+      "loss": 4.2447,
+      "step": 460
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 12.855481147766113,
+      "learning_rate": 1.9077879121980556e-05,
+      "loss": 3.3459,
+      "step": 461
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 13.02658462524414,
+      "learning_rate": 1.9075874511376166e-05,
+      "loss": 4.0793,
+      "step": 462
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 17.221975326538086,
+      "learning_rate": 1.907386990077178e-05,
+      "loss": 3.3399,
+      "step": 463
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 22.799137115478516,
+      "learning_rate": 1.9071865290167386e-05,
+      "loss": 3.7892,
+      "step": 464
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 21.327260971069336,
+      "learning_rate": 1.9069860679562996e-05,
+      "loss": 2.5584,
+      "step": 465
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 26.7298583984375,
+      "learning_rate": 1.9067856068958606e-05,
+      "loss": 4.5134,
+      "step": 466
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 25.994600296020508,
+      "learning_rate": 1.9065851458354216e-05,
+      "loss": 2.8459,
+      "step": 467
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 22.67238998413086,
+      "learning_rate": 1.9063846847749826e-05,
+      "loss": 4.0606,
+      "step": 468
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 16.200130462646484,
+      "learning_rate": 1.9061842237145436e-05,
+      "loss": 4.0715,
+      "step": 469
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 12.668680191040039,
+      "learning_rate": 1.9059837626541046e-05,
+      "loss": 2.9427,
+      "step": 470
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 24.67599868774414,
+      "learning_rate": 1.9057833015936657e-05,
+      "loss": 3.3925,
+      "step": 471
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 26.510074615478516,
+      "learning_rate": 1.9055828405332267e-05,
+      "loss": 2.774,
+      "step": 472
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 19.766721725463867,
+      "learning_rate": 1.9053823794727873e-05,
+      "loss": 3.1132,
+      "step": 473
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 14.419114112854004,
+      "learning_rate": 1.9051819184123487e-05,
+      "loss": 2.8139,
+      "step": 474
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 11.945268630981445,
+      "learning_rate": 1.9049814573519097e-05,
+      "loss": 3.4746,
+      "step": 475
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 19.036060333251953,
+      "learning_rate": 1.9047809962914704e-05,
+      "loss": 4.0888,
+      "step": 476
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 18.684444427490234,
+      "learning_rate": 1.9045805352310317e-05,
+      "loss": 3.8154,
+      "step": 477
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 10.9110689163208,
+      "learning_rate": 1.9043800741705924e-05,
+      "loss": 3.7957,
+      "step": 478
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 22.50055503845215,
+      "learning_rate": 1.9041796131101534e-05,
+      "loss": 4.0542,
+      "step": 479
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 14.225284576416016,
+      "learning_rate": 1.9039791520497144e-05,
+      "loss": 2.9864,
+      "step": 480
+    },
+    {
+      "epoch": 0.14,
+      "eval_loss": 0.9622519612312317,
+      "eval_runtime": 43.4567,
+      "eval_samples_per_second": 34.034,
+      "eval_steps_per_second": 34.034,
+      "step": 480
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 14.51484489440918,
+      "learning_rate": 1.9037786909892754e-05,
+      "loss": 3.9772,
+      "step": 481
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 51.21427536010742,
+      "learning_rate": 1.9035782299288364e-05,
+      "loss": 3.8034,
+      "step": 482
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 26.110877990722656,
+      "learning_rate": 1.9033777688683974e-05,
+      "loss": 3.6146,
+      "step": 483
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 16.056385040283203,
+      "learning_rate": 1.9031773078079584e-05,
+      "loss": 3.8006,
+      "step": 484
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 15.42993450164795,
+      "learning_rate": 1.9029768467475194e-05,
+      "loss": 3.0251,
+      "step": 485
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 12.17597484588623,
+      "learning_rate": 1.9027763856870804e-05,
+      "loss": 3.4756,
+      "step": 486
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 11.121559143066406,
+      "learning_rate": 1.9025759246266415e-05,
+      "loss": 2.7291,
+      "step": 487
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 17.4909725189209,
+      "learning_rate": 1.9023754635662025e-05,
+      "loss": 3.6708,
+      "step": 488
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 23.629941940307617,
+      "learning_rate": 1.9021750025057635e-05,
+      "loss": 3.0275,
+      "step": 489
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 12.820405960083008,
+      "learning_rate": 1.9019745414453245e-05,
+      "loss": 3.0267,
+      "step": 490
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 22.951032638549805,
+      "learning_rate": 1.9017740803848855e-05,
+      "loss": 2.393,
+      "step": 491
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 11.71376895904541,
+      "learning_rate": 1.901573619324446e-05,
+      "loss": 2.9705,
+      "step": 492
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 14.28131103515625,
+      "learning_rate": 1.9013731582640075e-05,
+      "loss": 2.727,
+      "step": 493
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 16.59454345703125,
+      "learning_rate": 1.9011726972035685e-05,
+      "loss": 3.5118,
+      "step": 494
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 20.26374053955078,
+      "learning_rate": 1.9009722361431292e-05,
+      "loss": 4.6275,
+      "step": 495
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 13.29510498046875,
+      "learning_rate": 1.9007717750826905e-05,
+      "loss": 3.8611,
+      "step": 496
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 14.79573917388916,
+      "learning_rate": 1.9005713140222512e-05,
+      "loss": 4.632,
+      "step": 497
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 19.15884017944336,
+      "learning_rate": 1.9003708529618122e-05,
+      "loss": 3.5274,
+      "step": 498
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 11.399006843566895,
+      "learning_rate": 1.9001703919013736e-05,
+      "loss": 4.0488,
+      "step": 499
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 12.227916717529297,
+      "learning_rate": 1.8999699308409342e-05,
+      "loss": 3.8005,
+      "step": 500
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 15.450165748596191,
+      "learning_rate": 1.8997694697804952e-05,
+      "loss": 2.9546,
+      "step": 501
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 19.479894638061523,
+      "learning_rate": 1.8995690087200562e-05,
+      "loss": 3.056,
+      "step": 502
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 18.609848022460938,
+      "learning_rate": 1.8993685476596172e-05,
+      "loss": 3.8573,
+      "step": 503
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 19.30628204345703,
+      "learning_rate": 1.8991680865991783e-05,
+      "loss": 2.9196,
+      "step": 504
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 15.707794189453125,
+      "learning_rate": 1.8989676255387393e-05,
+      "loss": 3.6089,
+      "step": 505
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 18.78173065185547,
+      "learning_rate": 1.8987671644783003e-05,
+      "loss": 2.8028,
+      "step": 506
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 17.811344146728516,
+      "learning_rate": 1.8985667034178613e-05,
+      "loss": 3.645,
+      "step": 507
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 15.66889476776123,
+      "learning_rate": 1.8983662423574223e-05,
+      "loss": 3.5172,
+      "step": 508
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 13.496955871582031,
+      "learning_rate": 1.8981657812969833e-05,
+      "loss": 3.2022,
+      "step": 509
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 14.416739463806152,
+      "learning_rate": 1.8979653202365443e-05,
+      "loss": 3.7212,
+      "step": 510
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 13.47453784942627,
+      "learning_rate": 1.897764859176105e-05,
+      "loss": 4.1187,
+      "step": 511
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 18.787689208984375,
+      "learning_rate": 1.8975643981156663e-05,
+      "loss": 3.6262,
+      "step": 512
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 10.850929260253906,
+      "learning_rate": 1.8973639370552273e-05,
+      "loss": 2.937,
+      "step": 513
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 14.07578182220459,
+      "learning_rate": 1.897163475994788e-05,
+      "loss": 4.5924,
+      "step": 514
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 13.420841217041016,
+      "learning_rate": 1.8969630149343493e-05,
+      "loss": 2.5988,
+      "step": 515
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 16.54571533203125,
+      "learning_rate": 1.89676255387391e-05,
+      "loss": 3.835,
+      "step": 516
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 68.53701782226562,
+      "learning_rate": 1.896562092813471e-05,
+      "loss": 5.1961,
+      "step": 517
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 9.695450782775879,
+      "learning_rate": 1.8963616317530324e-05,
+      "loss": 3.4607,
+      "step": 518
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 15.207024574279785,
+      "learning_rate": 1.896161170692593e-05,
+      "loss": 3.3355,
+      "step": 519
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 14.207443237304688,
+      "learning_rate": 1.895960709632154e-05,
+      "loss": 3.5345,
+      "step": 520
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 18.703203201293945,
+      "learning_rate": 1.895760248571715e-05,
+      "loss": 4.1013,
+      "step": 521
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 22.39604949951172,
+      "learning_rate": 1.895559787511276e-05,
+      "loss": 4.8708,
+      "step": 522
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 19.35679817199707,
+      "learning_rate": 1.895359326450837e-05,
+      "loss": 2.9551,
+      "step": 523
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 13.426315307617188,
+      "learning_rate": 1.895158865390398e-05,
+      "loss": 2.9543,
+      "step": 524
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 18.244346618652344,
+      "learning_rate": 1.894958404329959e-05,
+      "loss": 3.8422,
+      "step": 525
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 16.004077911376953,
+      "learning_rate": 1.89475794326952e-05,
+      "loss": 2.7931,
+      "step": 526
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 23.495067596435547,
+      "learning_rate": 1.894557482209081e-05,
+      "loss": 4.6575,
+      "step": 527
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 16.585546493530273,
+      "learning_rate": 1.894357021148642e-05,
+      "loss": 3.2343,
+      "step": 528
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 25.565746307373047,
+      "learning_rate": 1.894156560088203e-05,
+      "loss": 3.6168,
+      "step": 529
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 17.877038955688477,
+      "learning_rate": 1.893956099027764e-05,
+      "loss": 4.1159,
+      "step": 530
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 25.356380462646484,
+      "learning_rate": 1.893755637967325e-05,
+      "loss": 4.1038,
+      "step": 531
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 28.306737899780273,
+      "learning_rate": 1.893555176906886e-05,
+      "loss": 3.7835,
+      "step": 532
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 16.962968826293945,
+      "learning_rate": 1.8933547158464468e-05,
+      "loss": 2.9008,
+      "step": 533
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 20.51753044128418,
+      "learning_rate": 1.8931542547860082e-05,
+      "loss": 4.6783,
+      "step": 534
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 17.836177825927734,
+      "learning_rate": 1.892953793725569e-05,
+      "loss": 2.8979,
+      "step": 535
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 14.355405807495117,
+      "learning_rate": 1.89275333266513e-05,
+      "loss": 3.6143,
+      "step": 536
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 17.681434631347656,
+      "learning_rate": 1.8925528716046912e-05,
+      "loss": 2.4729,
+      "step": 537
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 16.02050018310547,
+      "learning_rate": 1.892352410544252e-05,
+      "loss": 2.3059,
+      "step": 538
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 12.686012268066406,
+      "learning_rate": 1.892151949483813e-05,
+      "loss": 3.307,
+      "step": 539
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 13.235198974609375,
+      "learning_rate": 1.891951488423374e-05,
+      "loss": 4.249,
+      "step": 540
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 16.329849243164062,
+      "learning_rate": 1.891751027362935e-05,
+      "loss": 3.9762,
+      "step": 541
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 18.02645492553711,
+      "learning_rate": 1.891550566302496e-05,
+      "loss": 3.3566,
+      "step": 542
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 15.097955703735352,
+      "learning_rate": 1.891350105242057e-05,
+      "loss": 3.683,
+      "step": 543
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 14.13657283782959,
+      "learning_rate": 1.891149644181618e-05,
+      "loss": 2.5315,
+      "step": 544
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 13.064208984375,
+      "learning_rate": 1.890949183121179e-05,
+      "loss": 3.7249,
+      "step": 545
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 12.626958847045898,
+      "learning_rate": 1.89074872206074e-05,
+      "loss": 3.4621,
+      "step": 546
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 15.529770851135254,
+      "learning_rate": 1.8905482610003006e-05,
+      "loss": 3.6776,
+      "step": 547
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 16.237417221069336,
+      "learning_rate": 1.890347799939862e-05,
+      "loss": 5.0018,
+      "step": 548
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 24.94883155822754,
+      "learning_rate": 1.890147338879423e-05,
+      "loss": 3.5904,
+      "step": 549
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 18.61686134338379,
+      "learning_rate": 1.8899468778189836e-05,
+      "loss": 3.47,
+      "step": 550
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 23.455402374267578,
+      "learning_rate": 1.889746416758545e-05,
+      "loss": 3.8038,
+      "step": 551
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 14.99221420288086,
+      "learning_rate": 1.8895459556981056e-05,
+      "loss": 3.4294,
+      "step": 552
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 15.767837524414062,
+      "learning_rate": 1.8893454946376667e-05,
+      "loss": 3.2229,
+      "step": 553
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 12.863175392150879,
+      "learning_rate": 1.8891450335772277e-05,
+      "loss": 3.6184,
+      "step": 554
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 14.412956237792969,
+      "learning_rate": 1.8889445725167887e-05,
+      "loss": 4.1913,
+      "step": 555
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 24.851470947265625,
+      "learning_rate": 1.8887441114563497e-05,
+      "loss": 3.9418,
+      "step": 556
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 47.26935958862305,
+      "learning_rate": 1.8885436503959107e-05,
+      "loss": 5.1826,
+      "step": 557
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 15.844667434692383,
+      "learning_rate": 1.8883431893354717e-05,
+      "loss": 3.7824,
+      "step": 558
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 11.842270851135254,
+      "learning_rate": 1.8881427282750327e-05,
+      "loss": 2.9349,
+      "step": 559
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 21.0179386138916,
+      "learning_rate": 1.8879422672145937e-05,
+      "loss": 3.2554,
+      "step": 560
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 19.371559143066406,
+      "learning_rate": 1.8877418061541547e-05,
+      "loss": 3.6965,
+      "step": 561
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 17.8704833984375,
+      "learning_rate": 1.8875413450937157e-05,
+      "loss": 3.6738,
+      "step": 562
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 11.61600112915039,
+      "learning_rate": 1.8873408840332767e-05,
+      "loss": 2.5997,
+      "step": 563
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 17.82415199279785,
+      "learning_rate": 1.8871404229728377e-05,
+      "loss": 3.402,
+      "step": 564
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 13.841615676879883,
+      "learning_rate": 1.8869399619123988e-05,
+      "loss": 2.5574,
+      "step": 565
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 17.705219268798828,
+      "learning_rate": 1.8867395008519594e-05,
+      "loss": 3.4216,
+      "step": 566
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 15.329522132873535,
+      "learning_rate": 1.8865390397915208e-05,
+      "loss": 2.5929,
+      "step": 567
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 14.040759086608887,
+      "learning_rate": 1.8863385787310818e-05,
+      "loss": 3.7188,
+      "step": 568
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 20.66543960571289,
+      "learning_rate": 1.8861381176706424e-05,
+      "loss": 3.3664,
+      "step": 569
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 10.702879905700684,
+      "learning_rate": 1.8859376566102038e-05,
+      "loss": 3.0084,
+      "step": 570
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 15.897429466247559,
+      "learning_rate": 1.8857371955497645e-05,
+      "loss": 3.7669,
+      "step": 571
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 11.95020866394043,
+      "learning_rate": 1.8855367344893255e-05,
+      "loss": 3.9482,
+      "step": 572
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 27.125240325927734,
+      "learning_rate": 1.8853362734288868e-05,
+      "loss": 2.7438,
+      "step": 573
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 13.781254768371582,
+      "learning_rate": 1.8851358123684475e-05,
+      "loss": 3.3872,
+      "step": 574
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 29.12127113342285,
+      "learning_rate": 1.8849353513080085e-05,
+      "loss": 4.5567,
+      "step": 575
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 15.671697616577148,
+      "learning_rate": 1.8847348902475695e-05,
+      "loss": 2.5302,
+      "step": 576
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 17.996654510498047,
+      "learning_rate": 1.8845344291871305e-05,
+      "loss": 3.1327,
+      "step": 577
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 16.846689224243164,
+      "learning_rate": 1.8843339681266915e-05,
+      "loss": 2.5453,
+      "step": 578
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 21.583017349243164,
+      "learning_rate": 1.8841335070662525e-05,
+      "loss": 3.5481,
+      "step": 579
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 14.08506965637207,
+      "learning_rate": 1.8839330460058135e-05,
+      "loss": 3.8354,
+      "step": 580
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 12.634489059448242,
+      "learning_rate": 1.8837325849453745e-05,
+      "loss": 3.1067,
+      "step": 581
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 11.801117897033691,
+      "learning_rate": 1.8835321238849356e-05,
+      "loss": 3.4585,
+      "step": 582
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 15.204583168029785,
+      "learning_rate": 1.8833316628244966e-05,
+      "loss": 3.3097,
+      "step": 583
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 15.794861793518066,
+      "learning_rate": 1.8831312017640576e-05,
+      "loss": 3.0038,
+      "step": 584
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 13.39342212677002,
+      "learning_rate": 1.8829307407036186e-05,
+      "loss": 3.4626,
+      "step": 585
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 14.78641414642334,
+      "learning_rate": 1.8827302796431796e-05,
+      "loss": 3.4037,
+      "step": 586
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 16.814594268798828,
+      "learning_rate": 1.8825298185827406e-05,
+      "loss": 3.7882,
+      "step": 587
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 27.2209415435791,
+      "learning_rate": 1.8823293575223013e-05,
+      "loss": 3.2455,
+      "step": 588
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 16.686363220214844,
+      "learning_rate": 1.8821288964618626e-05,
+      "loss": 3.932,
+      "step": 589
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 17.841867446899414,
+      "learning_rate": 1.8819284354014233e-05,
+      "loss": 2.3729,
+      "step": 590
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 13.670007705688477,
+      "learning_rate": 1.8817279743409843e-05,
+      "loss": 3.9139,
+      "step": 591
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 12.825089454650879,
+      "learning_rate": 1.8815275132805456e-05,
+      "loss": 3.4298,
+      "step": 592
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 21.140291213989258,
+      "learning_rate": 1.8813270522201063e-05,
+      "loss": 3.6672,
+      "step": 593
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 19.407217025756836,
+      "learning_rate": 1.8811265911596673e-05,
+      "loss": 3.853,
+      "step": 594
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 17.949951171875,
+      "learning_rate": 1.8809261300992283e-05,
+      "loss": 4.0853,
+      "step": 595
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 14.42286205291748,
+      "learning_rate": 1.8807256690387893e-05,
+      "loss": 3.8624,
+      "step": 596
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 13.524324417114258,
+      "learning_rate": 1.8805252079783503e-05,
+      "loss": 3.4537,
+      "step": 597
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 16.790464401245117,
+      "learning_rate": 1.8803247469179114e-05,
+      "loss": 3.3715,
+      "step": 598
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 28.930021286010742,
+      "learning_rate": 1.8801242858574724e-05,
+      "loss": 4.5601,
+      "step": 599
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 17.764270782470703,
+      "learning_rate": 1.8799238247970334e-05,
+      "loss": 3.0025,
+      "step": 600
+    },
+    {
+      "epoch": 0.18,
+      "eval_loss": 0.8108684420585632,
+      "eval_runtime": 43.7575,
+      "eval_samples_per_second": 33.8,
+      "eval_steps_per_second": 33.8,
+      "step": 600
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 12.37165641784668,
+      "learning_rate": 1.8797233637365944e-05,
+      "loss": 3.7558,
+      "step": 601
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 31.21900177001953,
+      "learning_rate": 1.8795229026761554e-05,
+      "loss": 3.6418,
+      "step": 602
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 12.556445121765137,
+      "learning_rate": 1.8793224416157164e-05,
+      "loss": 2.7222,
+      "step": 603
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 12.976155281066895,
+      "learning_rate": 1.8791219805552774e-05,
+      "loss": 3.1556,
+      "step": 604
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 14.281742095947266,
+      "learning_rate": 1.8789215194948384e-05,
+      "loss": 3.5987,
+      "step": 605
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 17.95466423034668,
+      "learning_rate": 1.8787210584343994e-05,
+      "loss": 2.8724,
+      "step": 606
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 11.985798835754395,
+      "learning_rate": 1.87852059737396e-05,
+      "loss": 3.3135,
+      "step": 607
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 11.839190483093262,
+      "learning_rate": 1.8783201363135214e-05,
+      "loss": 2.7356,
+      "step": 608
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 19.593158721923828,
+      "learning_rate": 1.878119675253082e-05,
+      "loss": 2.48,
+      "step": 609
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 11.183398246765137,
+      "learning_rate": 1.877919214192643e-05,
+      "loss": 2.3999,
+      "step": 610
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 16.24125862121582,
+      "learning_rate": 1.8777187531322045e-05,
+      "loss": 3.8464,
+      "step": 611
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 17.56856346130371,
+      "learning_rate": 1.877518292071765e-05,
+      "loss": 3.5414,
+      "step": 612
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 15.031785011291504,
+      "learning_rate": 1.877317831011326e-05,
+      "loss": 3.4706,
+      "step": 613
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 14.207649230957031,
+      "learning_rate": 1.877117369950887e-05,
+      "loss": 3.0532,
+      "step": 614
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 15.561878204345703,
+      "learning_rate": 1.876916908890448e-05,
+      "loss": 3.5544,
+      "step": 615
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 15.425860404968262,
+      "learning_rate": 1.876716447830009e-05,
+      "loss": 3.2101,
+      "step": 616
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 17.44911766052246,
+      "learning_rate": 1.8765159867695702e-05,
+      "loss": 4.0486,
+      "step": 617
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 16.89653205871582,
+      "learning_rate": 1.8763155257091312e-05,
+      "loss": 3.1268,
+      "step": 618
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 17.181591033935547,
+      "learning_rate": 1.8761150646486922e-05,
+      "loss": 3.0687,
+      "step": 619
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 14.057995796203613,
+      "learning_rate": 1.8759146035882532e-05,
+      "loss": 2.7292,
+      "step": 620
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 62.06224822998047,
+      "learning_rate": 1.875714142527814e-05,
+      "loss": 2.8716,
+      "step": 621
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 15.26970100402832,
+      "learning_rate": 1.8755136814673752e-05,
+      "loss": 3.4212,
+      "step": 622
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 12.904154777526855,
+      "learning_rate": 1.8753132204069362e-05,
+      "loss": 2.5465,
+      "step": 623
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 16.669986724853516,
+      "learning_rate": 1.875112759346497e-05,
+      "loss": 2.8482,
+      "step": 624
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 10.211058616638184,
+      "learning_rate": 1.8749122982860582e-05,
+      "loss": 2.5476,
+      "step": 625
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 21.396862030029297,
+      "learning_rate": 1.874711837225619e-05,
+      "loss": 3.7639,
+      "step": 626
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 20.889707565307617,
+      "learning_rate": 1.87451137616518e-05,
+      "loss": 3.3019,
+      "step": 627
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 25.096691131591797,
+      "learning_rate": 1.874310915104741e-05,
+      "loss": 4.246,
+      "step": 628
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 13.878023147583008,
+      "learning_rate": 1.874110454044302e-05,
+      "loss": 2.9438,
+      "step": 629
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 18.34180450439453,
+      "learning_rate": 1.873909992983863e-05,
+      "loss": 4.3573,
+      "step": 630
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 18.738731384277344,
+      "learning_rate": 1.873709531923424e-05,
+      "loss": 2.7748,
+      "step": 631
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 15.163655281066895,
+      "learning_rate": 1.873509070862985e-05,
+      "loss": 3.4938,
+      "step": 632
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 16.00517463684082,
+      "learning_rate": 1.873308609802546e-05,
+      "loss": 4.1243,
+      "step": 633
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 15.933029174804688,
+      "learning_rate": 1.873108148742107e-05,
+      "loss": 3.8793,
+      "step": 634
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 17.073196411132812,
+      "learning_rate": 1.872907687681668e-05,
+      "loss": 3.3813,
+      "step": 635
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 16.49427604675293,
+      "learning_rate": 1.872707226621229e-05,
+      "loss": 3.9439,
+      "step": 636
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 14.988619804382324,
+      "learning_rate": 1.87250676556079e-05,
+      "loss": 3.4706,
+      "step": 637
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 14.796496391296387,
+      "learning_rate": 1.872306304500351e-05,
+      "loss": 2.572,
+      "step": 638
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 14.737870216369629,
+      "learning_rate": 1.872105843439912e-05,
+      "loss": 4.6644,
+      "step": 639
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 14.593988418579102,
+      "learning_rate": 1.8719053823794727e-05,
+      "loss": 2.9476,
+      "step": 640
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 24.125364303588867,
+      "learning_rate": 1.871704921319034e-05,
+      "loss": 3.0569,
+      "step": 641
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 12.783120155334473,
+      "learning_rate": 1.871504460258595e-05,
+      "loss": 2.7435,
+      "step": 642
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 20.349702835083008,
+      "learning_rate": 1.8713039991981557e-05,
+      "loss": 3.924,
+      "step": 643
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 18.848377227783203,
+      "learning_rate": 1.871103538137717e-05,
+      "loss": 3.6072,
+      "step": 644
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 16.555147171020508,
+      "learning_rate": 1.8709030770772777e-05,
+      "loss": 2.6862,
+      "step": 645
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 93.45834350585938,
+      "learning_rate": 1.8707026160168387e-05,
+      "loss": 3.1187,
+      "step": 646
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 12.666962623596191,
+      "learning_rate": 1.8705021549564e-05,
+      "loss": 3.3528,
+      "step": 647
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 13.548149108886719,
+      "learning_rate": 1.8703016938959608e-05,
+      "loss": 2.8839,
+      "step": 648
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 20.26407241821289,
+      "learning_rate": 1.8701012328355218e-05,
+      "loss": 3.2999,
+      "step": 649
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 16.36279296875,
+      "learning_rate": 1.8699007717750828e-05,
+      "loss": 2.7201,
+      "step": 650
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 15.859626770019531,
+      "learning_rate": 1.8697003107146438e-05,
+      "loss": 3.3978,
+      "step": 651
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 20.308744430541992,
+      "learning_rate": 1.8694998496542048e-05,
+      "loss": 3.7883,
+      "step": 652
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 14.849808692932129,
+      "learning_rate": 1.8692993885937658e-05,
+      "loss": 3.0277,
+      "step": 653
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 10.114745140075684,
+      "learning_rate": 1.8690989275333268e-05,
+      "loss": 2.4441,
+      "step": 654
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 21.375455856323242,
+      "learning_rate": 1.8688984664728878e-05,
+      "loss": 2.9501,
+      "step": 655
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 21.958810806274414,
+      "learning_rate": 1.8686980054124488e-05,
+      "loss": 2.8862,
+      "step": 656
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 14.946088790893555,
+      "learning_rate": 1.86849754435201e-05,
+      "loss": 3.4693,
+      "step": 657
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 13.369003295898438,
+      "learning_rate": 1.868297083291571e-05,
+      "loss": 4.4842,
+      "step": 658
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 13.760272979736328,
+      "learning_rate": 1.868096622231132e-05,
+      "loss": 4.0259,
+      "step": 659
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 18.290048599243164,
+      "learning_rate": 1.867896161170693e-05,
+      "loss": 2.4139,
+      "step": 660
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 20.004146575927734,
+      "learning_rate": 1.867695700110254e-05,
+      "loss": 3.8439,
+      "step": 661
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 15.349987983703613,
+      "learning_rate": 1.8674952390498145e-05,
+      "loss": 2.9132,
+      "step": 662
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 14.10637092590332,
+      "learning_rate": 1.867294777989376e-05,
+      "loss": 2.9145,
+      "step": 663
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 15.24035358428955,
+      "learning_rate": 1.8670943169289366e-05,
+      "loss": 2.748,
+      "step": 664
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 23.505083084106445,
+      "learning_rate": 1.8668938558684976e-05,
+      "loss": 2.8452,
+      "step": 665
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 82.25447845458984,
+      "learning_rate": 1.866693394808059e-05,
+      "loss": 2.8634,
+      "step": 666
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 24.59209632873535,
+      "learning_rate": 1.8664929337476196e-05,
+      "loss": 2.9632,
+      "step": 667
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 18.12932014465332,
+      "learning_rate": 1.8662924726871806e-05,
+      "loss": 3.9964,
+      "step": 668
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 13.9307222366333,
+      "learning_rate": 1.8660920116267416e-05,
+      "loss": 3.37,
+      "step": 669
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 15.804916381835938,
+      "learning_rate": 1.8658915505663026e-05,
+      "loss": 3.7911,
+      "step": 670
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 12.954413414001465,
+      "learning_rate": 1.8656910895058636e-05,
+      "loss": 3.4076,
+      "step": 671
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 19.96038055419922,
+      "learning_rate": 1.8654906284454246e-05,
+      "loss": 2.6706,
+      "step": 672
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 14.459244728088379,
+      "learning_rate": 1.8652901673849856e-05,
+      "loss": 2.6447,
+      "step": 673
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 14.168243408203125,
+      "learning_rate": 1.8650897063245466e-05,
+      "loss": 2.8916,
+      "step": 674
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 12.294631004333496,
+      "learning_rate": 1.8648892452641076e-05,
+      "loss": 3.911,
+      "step": 675
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 16.20514488220215,
+      "learning_rate": 1.8646887842036687e-05,
+      "loss": 4.1275,
+      "step": 676
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 13.80794906616211,
+      "learning_rate": 1.8644883231432297e-05,
+      "loss": 2.5423,
+      "step": 677
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 13.199609756469727,
+      "learning_rate": 1.8642878620827907e-05,
+      "loss": 3.119,
+      "step": 678
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 17.830242156982422,
+      "learning_rate": 1.8640874010223517e-05,
+      "loss": 3.9047,
+      "step": 679
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 17.04703140258789,
+      "learning_rate": 1.8638869399619127e-05,
+      "loss": 2.4903,
+      "step": 680
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 16.005630493164062,
+      "learning_rate": 1.8636864789014734e-05,
+      "loss": 3.5141,
+      "step": 681
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 12.92525577545166,
+      "learning_rate": 1.8634860178410347e-05,
+      "loss": 3.5435,
+      "step": 682
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 13.173433303833008,
+      "learning_rate": 1.8632855567805954e-05,
+      "loss": 3.4652,
+      "step": 683
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 15.04622745513916,
+      "learning_rate": 1.8630850957201564e-05,
+      "loss": 3.2746,
+      "step": 684
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 16.044187545776367,
+      "learning_rate": 1.8628846346597177e-05,
+      "loss": 3.5208,
+      "step": 685
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 12.709644317626953,
+      "learning_rate": 1.8626841735992784e-05,
+      "loss": 2.9898,
+      "step": 686
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 62.63556671142578,
+      "learning_rate": 1.8624837125388394e-05,
+      "loss": 4.3599,
+      "step": 687
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 14.936622619628906,
+      "learning_rate": 1.8622832514784004e-05,
+      "loss": 3.7374,
+      "step": 688
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 18.37685203552246,
+      "learning_rate": 1.8620827904179614e-05,
+      "loss": 2.8972,
+      "step": 689
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 18.769283294677734,
+      "learning_rate": 1.8618823293575224e-05,
+      "loss": 2.7669,
+      "step": 690
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 14.026557922363281,
+      "learning_rate": 1.8616818682970834e-05,
+      "loss": 3.4005,
+      "step": 691
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 15.341630935668945,
+      "learning_rate": 1.8614814072366445e-05,
+      "loss": 3.8892,
+      "step": 692
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 20.44842529296875,
+      "learning_rate": 1.8612809461762055e-05,
+      "loss": 4.4409,
+      "step": 693
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 23.54950523376465,
+      "learning_rate": 1.8610804851157665e-05,
+      "loss": 3.9646,
+      "step": 694
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 13.78188419342041,
+      "learning_rate": 1.860880024055327e-05,
+      "loss": 4.5521,
+      "step": 695
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 21.157424926757812,
+      "learning_rate": 1.8606795629948885e-05,
+      "loss": 3.6927,
+      "step": 696
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 12.406957626342773,
+      "learning_rate": 1.8604791019344495e-05,
+      "loss": 3.2458,
+      "step": 697
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 18.12921142578125,
+      "learning_rate": 1.86027864087401e-05,
+      "loss": 3.7537,
+      "step": 698
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 21.271060943603516,
+      "learning_rate": 1.8600781798135715e-05,
+      "loss": 4.3055,
+      "step": 699
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 13.396303176879883,
+      "learning_rate": 1.8598777187531322e-05,
+      "loss": 3.5275,
+      "step": 700
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 11.93570613861084,
+      "learning_rate": 1.8596772576926935e-05,
+      "loss": 3.0972,
+      "step": 701
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 14.72830581665039,
+      "learning_rate": 1.8594767966322545e-05,
+      "loss": 4.0018,
+      "step": 702
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 23.694110870361328,
+      "learning_rate": 1.8592763355718152e-05,
+      "loss": 3.2218,
+      "step": 703
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 17.545129776000977,
+      "learning_rate": 1.8590758745113766e-05,
+      "loss": 3.2768,
+      "step": 704
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 16.36639976501465,
+      "learning_rate": 1.8588754134509372e-05,
+      "loss": 3.1514,
+      "step": 705
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 11.386521339416504,
+      "learning_rate": 1.8586749523904982e-05,
+      "loss": 2.3357,
+      "step": 706
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 10.972953796386719,
+      "learning_rate": 1.8584744913300592e-05,
+      "loss": 2.8289,
+      "step": 707
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 16.927831649780273,
+      "learning_rate": 1.8582740302696202e-05,
+      "loss": 3.086,
+      "step": 708
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 14.967674255371094,
+      "learning_rate": 1.8580735692091813e-05,
+      "loss": 3.6209,
+      "step": 709
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 16.293773651123047,
+      "learning_rate": 1.8578731081487423e-05,
+      "loss": 4.2359,
+      "step": 710
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 14.0774564743042,
+      "learning_rate": 1.8576726470883033e-05,
+      "loss": 3.7756,
+      "step": 711
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 26.526830673217773,
+      "learning_rate": 1.8574721860278643e-05,
+      "loss": 3.1339,
+      "step": 712
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 13.169925689697266,
+      "learning_rate": 1.8572717249674253e-05,
+      "loss": 2.6322,
+      "step": 713
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 13.879220962524414,
+      "learning_rate": 1.857071263906986e-05,
+      "loss": 2.5253,
+      "step": 714
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 15.33066177368164,
+      "learning_rate": 1.8568708028465473e-05,
+      "loss": 3.3604,
+      "step": 715
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 16.04281997680664,
+      "learning_rate": 1.8566703417861083e-05,
+      "loss": 2.824,
+      "step": 716
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 13.022286415100098,
+      "learning_rate": 1.856469880725669e-05,
+      "loss": 2.6692,
+      "step": 717
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 14.34041976928711,
+      "learning_rate": 1.8562694196652303e-05,
+      "loss": 3.0903,
+      "step": 718
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 17.68020248413086,
+      "learning_rate": 1.856068958604791e-05,
+      "loss": 3.4143,
+      "step": 719
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 13.933920860290527,
+      "learning_rate": 1.855868497544352e-05,
+      "loss": 3.2975,
+      "step": 720
+    },
+    {
+      "epoch": 0.22,
+      "eval_loss": 0.7311321496963501,
+      "eval_runtime": 43.9321,
+      "eval_samples_per_second": 33.666,
+      "eval_steps_per_second": 33.666,
+      "step": 720
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 14.72537612915039,
+      "learning_rate": 1.8556680364839134e-05,
+      "loss": 3.9612,
+      "step": 721
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 26.851642608642578,
+      "learning_rate": 1.855467575423474e-05,
+      "loss": 3.4687,
+      "step": 722
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 16.372344970703125,
+      "learning_rate": 1.855267114363035e-05,
+      "loss": 3.7605,
+      "step": 723
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 17.88177490234375,
+      "learning_rate": 1.855066653302596e-05,
+      "loss": 3.4436,
+      "step": 724
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 16.162473678588867,
+      "learning_rate": 1.854866192242157e-05,
+      "loss": 3.4066,
+      "step": 725
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 12.335894584655762,
+      "learning_rate": 1.854665731181718e-05,
+      "loss": 2.9006,
+      "step": 726
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 19.66415786743164,
+      "learning_rate": 1.854465270121279e-05,
+      "loss": 3.5752,
+      "step": 727
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 11.808516502380371,
+      "learning_rate": 1.85426480906084e-05,
+      "loss": 1.8741,
+      "step": 728
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 15.641440391540527,
+      "learning_rate": 1.854064348000401e-05,
+      "loss": 3.3391,
+      "step": 729
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 13.590608596801758,
+      "learning_rate": 1.853863886939962e-05,
+      "loss": 2.9265,
+      "step": 730
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 18.237276077270508,
+      "learning_rate": 1.853663425879523e-05,
+      "loss": 3.6717,
+      "step": 731
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 41.96309280395508,
+      "learning_rate": 1.853462964819084e-05,
+      "loss": 2.857,
+      "step": 732
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 15.107841491699219,
+      "learning_rate": 1.853262503758645e-05,
+      "loss": 3.9187,
+      "step": 733
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 15.574593544006348,
+      "learning_rate": 1.853062042698206e-05,
+      "loss": 3.1606,
+      "step": 734
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 21.812761306762695,
+      "learning_rate": 1.852861581637767e-05,
+      "loss": 2.8953,
+      "step": 735
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 21.730745315551758,
+      "learning_rate": 1.8526611205773278e-05,
+      "loss": 3.3289,
+      "step": 736
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 33.00816345214844,
+      "learning_rate": 1.852460659516889e-05,
+      "loss": 2.98,
+      "step": 737
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 16.756553649902344,
+      "learning_rate": 1.8522601984564498e-05,
+      "loss": 3.0438,
+      "step": 738
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 24.259780883789062,
+      "learning_rate": 1.852059737396011e-05,
+      "loss": 3.9168,
+      "step": 739
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 10.961197853088379,
+      "learning_rate": 1.8518592763355722e-05,
+      "loss": 2.521,
+      "step": 740
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 11.985860824584961,
+      "learning_rate": 1.851658815275133e-05,
+      "loss": 3.6082,
+      "step": 741
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 13.038666725158691,
+      "learning_rate": 1.851458354214694e-05,
+      "loss": 2.8915,
+      "step": 742
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 15.792445182800293,
+      "learning_rate": 1.851257893154255e-05,
+      "loss": 3.0136,
+      "step": 743
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 11.698639869689941,
+      "learning_rate": 1.851057432093816e-05,
+      "loss": 2.9538,
+      "step": 744
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 14.640687942504883,
+      "learning_rate": 1.850856971033377e-05,
+      "loss": 2.1114,
+      "step": 745
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 21.3159122467041,
+      "learning_rate": 1.850656509972938e-05,
+      "loss": 3.1929,
+      "step": 746
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 15.516459465026855,
+      "learning_rate": 1.850456048912499e-05,
+      "loss": 3.4558,
+      "step": 747
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 12.31661605834961,
+      "learning_rate": 1.85025558785206e-05,
+      "loss": 3.5311,
+      "step": 748
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 17.94566535949707,
+      "learning_rate": 1.850055126791621e-05,
+      "loss": 3.6713,
+      "step": 749
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 13.146416664123535,
+      "learning_rate": 1.849854665731182e-05,
+      "loss": 2.0929,
+      "step": 750
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 15.036728858947754,
+      "learning_rate": 1.849654204670743e-05,
+      "loss": 3.2656,
+      "step": 751
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 12.377967834472656,
+      "learning_rate": 1.849453743610304e-05,
+      "loss": 2.811,
+      "step": 752
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 48.185482025146484,
+      "learning_rate": 1.849253282549865e-05,
+      "loss": 4.3531,
+      "step": 753
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 16.886388778686523,
+      "learning_rate": 1.849052821489426e-05,
+      "loss": 3.3402,
+      "step": 754
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 10.66983699798584,
+      "learning_rate": 1.8488523604289866e-05,
+      "loss": 2.5462,
+      "step": 755
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 13.7207670211792,
+      "learning_rate": 1.848651899368548e-05,
+      "loss": 4.2536,
+      "step": 756
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 11.786395072937012,
+      "learning_rate": 1.8484514383081086e-05,
+      "loss": 3.3007,
+      "step": 757
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 19.73274803161621,
+      "learning_rate": 1.8482509772476697e-05,
+      "loss": 3.2301,
+      "step": 758
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 15.213275909423828,
+      "learning_rate": 1.848050516187231e-05,
+      "loss": 3.0059,
+      "step": 759
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 19.37066078186035,
+      "learning_rate": 1.8478500551267917e-05,
+      "loss": 2.8112,
+      "step": 760
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 48.94891357421875,
+      "learning_rate": 1.8476495940663527e-05,
+      "loss": 4.0552,
+      "step": 761
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 11.666791915893555,
+      "learning_rate": 1.8474491330059137e-05,
+      "loss": 3.0329,
+      "step": 762
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 14.530978202819824,
+      "learning_rate": 1.8472486719454747e-05,
+      "loss": 2.9601,
+      "step": 763
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 15.476301193237305,
+      "learning_rate": 1.8470482108850357e-05,
+      "loss": 3.8685,
+      "step": 764
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 34.36149597167969,
+      "learning_rate": 1.8468477498245967e-05,
+      "loss": 3.6309,
+      "step": 765
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 13.753215789794922,
+      "learning_rate": 1.8466472887641577e-05,
+      "loss": 2.5185,
+      "step": 766
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 14.692520141601562,
+      "learning_rate": 1.8464468277037187e-05,
+      "loss": 3.4938,
+      "step": 767
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 19.021665573120117,
+      "learning_rate": 1.8462463666432797e-05,
+      "loss": 2.4541,
+      "step": 768
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 15.532613754272461,
+      "learning_rate": 1.8460459055828407e-05,
+      "loss": 3.5129,
+      "step": 769
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 19.325273513793945,
+      "learning_rate": 1.8458454445224018e-05,
+      "loss": 3.8114,
+      "step": 770
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 19.3370361328125,
+      "learning_rate": 1.8456449834619628e-05,
+      "loss": 3.7453,
+      "step": 771
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 17.998004913330078,
+      "learning_rate": 1.8454445224015238e-05,
+      "loss": 4.1352,
+      "step": 772
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 14.474387168884277,
+      "learning_rate": 1.8452440613410848e-05,
+      "loss": 3.0304,
+      "step": 773
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 12.42952823638916,
+      "learning_rate": 1.8450436002806454e-05,
+      "loss": 2.8432,
+      "step": 774
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 15.06114387512207,
+      "learning_rate": 1.8448431392202068e-05,
+      "loss": 2.9204,
+      "step": 775
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 19.3477840423584,
+      "learning_rate": 1.8446426781597678e-05,
+      "loss": 2.6387,
+      "step": 776
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 19.3367862701416,
+      "learning_rate": 1.8444422170993285e-05,
+      "loss": 3.7189,
+      "step": 777
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 18.679487228393555,
+      "learning_rate": 1.8442417560388898e-05,
+      "loss": 2.7699,
+      "step": 778
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 17.268980026245117,
+      "learning_rate": 1.8440412949784505e-05,
+      "loss": 3.1304,
+      "step": 779
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 12.782285690307617,
+      "learning_rate": 1.8438408339180115e-05,
+      "loss": 2.1415,
+      "step": 780
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 13.401495933532715,
+      "learning_rate": 1.8436403728575725e-05,
+      "loss": 2.8394,
+      "step": 781
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 17.430240631103516,
+      "learning_rate": 1.8434399117971335e-05,
+      "loss": 3.3336,
+      "step": 782
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 50.752410888671875,
+      "learning_rate": 1.8432394507366945e-05,
+      "loss": 3.1897,
+      "step": 783
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 14.428933143615723,
+      "learning_rate": 1.8430389896762555e-05,
+      "loss": 2.8947,
+      "step": 784
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 60.45037078857422,
+      "learning_rate": 1.8428385286158165e-05,
+      "loss": 2.7407,
+      "step": 785
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 15.153972625732422,
+      "learning_rate": 1.8426380675553776e-05,
+      "loss": 3.9118,
+      "step": 786
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 13.527739524841309,
+      "learning_rate": 1.8424376064949386e-05,
+      "loss": 2.8687,
+      "step": 787
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 13.59441089630127,
+      "learning_rate": 1.8422371454344996e-05,
+      "loss": 3.4082,
+      "step": 788
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 18.013368606567383,
+      "learning_rate": 1.8420366843740606e-05,
+      "loss": 3.1523,
+      "step": 789
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 23.031423568725586,
+      "learning_rate": 1.8418362233136216e-05,
+      "loss": 3.8803,
+      "step": 790
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 12.914963722229004,
+      "learning_rate": 1.8416357622531823e-05,
+      "loss": 3.0994,
+      "step": 791
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 25.980188369750977,
+      "learning_rate": 1.8414353011927436e-05,
+      "loss": 1.9718,
+      "step": 792
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 14.725564002990723,
+      "learning_rate": 1.8412348401323043e-05,
+      "loss": 2.472,
+      "step": 793
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 13.297785758972168,
+      "learning_rate": 1.8410343790718653e-05,
+      "loss": 4.0097,
+      "step": 794
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 15.004578590393066,
+      "learning_rate": 1.8408339180114266e-05,
+      "loss": 2.8653,
+      "step": 795
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 19.27330780029297,
+      "learning_rate": 1.8406334569509873e-05,
+      "loss": 3.1684,
+      "step": 796
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 25.846576690673828,
+      "learning_rate": 1.8404329958905483e-05,
+      "loss": 2.6346,
+      "step": 797
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 9.999025344848633,
+      "learning_rate": 1.8402325348301093e-05,
+      "loss": 1.9912,
+      "step": 798
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 11.801385879516602,
+      "learning_rate": 1.8400320737696703e-05,
+      "loss": 3.3957,
+      "step": 799
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 12.033352851867676,
+      "learning_rate": 1.8398316127092313e-05,
+      "loss": 2.62,
+      "step": 800
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 12.087538719177246,
+      "learning_rate": 1.8396311516487923e-05,
+      "loss": 3.2451,
+      "step": 801
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 9.926271438598633,
+      "learning_rate": 1.8394306905883533e-05,
+      "loss": 2.5499,
+      "step": 802
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 15.592676162719727,
+      "learning_rate": 1.8392302295279144e-05,
+      "loss": 3.7471,
+      "step": 803
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 16.60494613647461,
+      "learning_rate": 1.8390297684674754e-05,
+      "loss": 3.218,
+      "step": 804
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 20.184545516967773,
+      "learning_rate": 1.8388293074070364e-05,
+      "loss": 2.9737,
+      "step": 805
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 12.5513334274292,
+      "learning_rate": 1.8386288463465974e-05,
+      "loss": 2.3381,
+      "step": 806
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 19.2335262298584,
+      "learning_rate": 1.8384283852861584e-05,
+      "loss": 3.5287,
+      "step": 807
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 16.872840881347656,
+      "learning_rate": 1.8382279242257194e-05,
+      "loss": 3.1892,
+      "step": 808
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 22.987356185913086,
+      "learning_rate": 1.8380274631652804e-05,
+      "loss": 2.534,
+      "step": 809
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 18.01645851135254,
+      "learning_rate": 1.837827002104841e-05,
+      "loss": 3.5278,
+      "step": 810
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 23.722457885742188,
+      "learning_rate": 1.8376265410444024e-05,
+      "loss": 3.9305,
+      "step": 811
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 11.13327693939209,
+      "learning_rate": 1.837426079983963e-05,
+      "loss": 3.9406,
+      "step": 812
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 18.1392765045166,
+      "learning_rate": 1.837225618923524e-05,
+      "loss": 2.3532,
+      "step": 813
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 17.197660446166992,
+      "learning_rate": 1.8370251578630854e-05,
+      "loss": 3.8571,
+      "step": 814
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 16.233530044555664,
+      "learning_rate": 1.836824696802646e-05,
+      "loss": 2.8763,
+      "step": 815
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 104.82502746582031,
+      "learning_rate": 1.836624235742207e-05,
+      "loss": 3.3302,
+      "step": 816
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 14.149030685424805,
+      "learning_rate": 1.836423774681768e-05,
+      "loss": 3.2956,
+      "step": 817
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 13.704061508178711,
+      "learning_rate": 1.836223313621329e-05,
+      "loss": 2.7446,
+      "step": 818
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 14.284625053405762,
+      "learning_rate": 1.83602285256089e-05,
+      "loss": 2.9887,
+      "step": 819
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 20.733549118041992,
+      "learning_rate": 1.835822391500451e-05,
+      "loss": 3.3273,
+      "step": 820
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 27.305700302124023,
+      "learning_rate": 1.835621930440012e-05,
+      "loss": 3.061,
+      "step": 821
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 11.420647621154785,
+      "learning_rate": 1.8354214693795732e-05,
+      "loss": 3.1825,
+      "step": 822
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 13.486950874328613,
+      "learning_rate": 1.8352210083191342e-05,
+      "loss": 3.1205,
+      "step": 823
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 21.656282424926758,
+      "learning_rate": 1.8350205472586952e-05,
+      "loss": 2.7608,
+      "step": 824
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 31.164608001708984,
+      "learning_rate": 1.8348200861982562e-05,
+      "loss": 3.3097,
+      "step": 825
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 36.02324676513672,
+      "learning_rate": 1.8346196251378172e-05,
+      "loss": 3.2008,
+      "step": 826
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 15.944084167480469,
+      "learning_rate": 1.8344191640773782e-05,
+      "loss": 2.9221,
+      "step": 827
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 14.965399742126465,
+      "learning_rate": 1.8342187030169392e-05,
+      "loss": 3.1973,
+      "step": 828
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 20.134925842285156,
+      "learning_rate": 1.8340182419565e-05,
+      "loss": 3.521,
+      "step": 829
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 17.916906356811523,
+      "learning_rate": 1.8338177808960612e-05,
+      "loss": 3.8329,
+      "step": 830
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 22.273197174072266,
+      "learning_rate": 1.8336173198356223e-05,
+      "loss": 3.0719,
+      "step": 831
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 18.33119773864746,
+      "learning_rate": 1.833416858775183e-05,
+      "loss": 3.1312,
+      "step": 832
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 17.575429916381836,
+      "learning_rate": 1.8332163977147443e-05,
+      "loss": 2.7529,
+      "step": 833
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 16.5457820892334,
+      "learning_rate": 1.833015936654305e-05,
+      "loss": 2.3905,
+      "step": 834
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 19.157955169677734,
+      "learning_rate": 1.832815475593866e-05,
+      "loss": 3.6365,
+      "step": 835
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 16.239450454711914,
+      "learning_rate": 1.832615014533427e-05,
+      "loss": 2.864,
+      "step": 836
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 24.656652450561523,
+      "learning_rate": 1.832414553472988e-05,
+      "loss": 3.4563,
+      "step": 837
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 21.09084129333496,
+      "learning_rate": 1.832214092412549e-05,
+      "loss": 3.3509,
+      "step": 838
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 14.556492805480957,
+      "learning_rate": 1.83201363135211e-05,
+      "loss": 3.0753,
+      "step": 839
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 19.413835525512695,
+      "learning_rate": 1.831813170291671e-05,
+      "loss": 2.9689,
+      "step": 840
+    },
+    {
+      "epoch": 0.25,
+      "eval_loss": 0.5996547937393188,
+      "eval_runtime": 43.6208,
+      "eval_samples_per_second": 33.906,
+      "eval_steps_per_second": 33.906,
+      "step": 840
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 17.024858474731445,
+      "learning_rate": 1.831612709231232e-05,
+      "loss": 4.6989,
+      "step": 841
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 12.914769172668457,
+      "learning_rate": 1.831412248170793e-05,
+      "loss": 3.6312,
+      "step": 842
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 89.747802734375,
+      "learning_rate": 1.831211787110354e-05,
+      "loss": 2.7291,
+      "step": 843
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 15.173033714294434,
+      "learning_rate": 1.831011326049915e-05,
+      "loss": 3.1805,
+      "step": 844
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 40.06729507446289,
+      "learning_rate": 1.830810864989476e-05,
+      "loss": 4.5534,
+      "step": 845
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 13.402242660522461,
+      "learning_rate": 1.830610403929037e-05,
+      "loss": 2.946,
+      "step": 846
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 18.814979553222656,
+      "learning_rate": 1.830409942868598e-05,
+      "loss": 3.3728,
+      "step": 847
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 12.975325584411621,
+      "learning_rate": 1.8302094818081587e-05,
+      "loss": 2.8497,
+      "step": 848
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 9.28498363494873,
+      "learning_rate": 1.83000902074772e-05,
+      "loss": 2.4016,
+      "step": 849
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 15.675614356994629,
+      "learning_rate": 1.829808559687281e-05,
+      "loss": 2.7848,
+      "step": 850
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 13.038680076599121,
+      "learning_rate": 1.8296080986268417e-05,
+      "loss": 3.4155,
+      "step": 851
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 11.621955871582031,
+      "learning_rate": 1.829407637566403e-05,
+      "loss": 2.8026,
+      "step": 852
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 11.502103805541992,
+      "learning_rate": 1.8292071765059638e-05,
+      "loss": 2.2732,
+      "step": 853
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 27.41511344909668,
+      "learning_rate": 1.8290067154455248e-05,
+      "loss": 4.0694,
+      "step": 854
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 16.437061309814453,
+      "learning_rate": 1.8288062543850858e-05,
+      "loss": 3.0165,
+      "step": 855
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 14.086379051208496,
+      "learning_rate": 1.8286057933246468e-05,
+      "loss": 2.6271,
+      "step": 856
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 18.89293670654297,
+      "learning_rate": 1.8284053322642078e-05,
+      "loss": 3.2895,
+      "step": 857
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 14.886481285095215,
+      "learning_rate": 1.8282048712037688e-05,
+      "loss": 3.3004,
+      "step": 858
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 21.090091705322266,
+      "learning_rate": 1.8280044101433298e-05,
+      "loss": 3.0523,
+      "step": 859
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 13.093547821044922,
+      "learning_rate": 1.8278039490828908e-05,
+      "loss": 2.3511,
+      "step": 860
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 13.202871322631836,
+      "learning_rate": 1.8276034880224518e-05,
+      "loss": 2.7941,
+      "step": 861
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 10.028319358825684,
+      "learning_rate": 1.827403026962013e-05,
+      "loss": 2.6855,
+      "step": 862
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 14.725035667419434,
+      "learning_rate": 1.827202565901574e-05,
+      "loss": 3.3114,
+      "step": 863
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 20.055673599243164,
+      "learning_rate": 1.827002104841135e-05,
+      "loss": 4.0679,
+      "step": 864
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 15.065644264221191,
+      "learning_rate": 1.8268016437806955e-05,
+      "loss": 2.8739,
+      "step": 865
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 11.230770111083984,
+      "learning_rate": 1.826601182720257e-05,
+      "loss": 3.2987,
+      "step": 866
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 27.67386245727539,
+      "learning_rate": 1.8264007216598175e-05,
+      "loss": 3.0173,
+      "step": 867
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 18.033613204956055,
+      "learning_rate": 1.8262002605993785e-05,
+      "loss": 3.4757,
+      "step": 868
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 12.066390037536621,
+      "learning_rate": 1.82599979953894e-05,
+      "loss": 2.815,
+      "step": 869
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 18.521556854248047,
+      "learning_rate": 1.8257993384785006e-05,
+      "loss": 4.0135,
+      "step": 870
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 19.63759422302246,
+      "learning_rate": 1.825598877418062e-05,
+      "loss": 4.0507,
+      "step": 871
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 34.88851547241211,
+      "learning_rate": 1.8253984163576226e-05,
+      "loss": 4.7139,
+      "step": 872
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 15.407425880432129,
+      "learning_rate": 1.8251979552971836e-05,
+      "loss": 3.7188,
+      "step": 873
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 12.915898323059082,
+      "learning_rate": 1.824997494236745e-05,
+      "loss": 2.5801,
+      "step": 874
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 12.690547943115234,
+      "learning_rate": 1.8247970331763056e-05,
+      "loss": 2.3076,
+      "step": 875
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 31.50592041015625,
+      "learning_rate": 1.8245965721158666e-05,
+      "loss": 2.8265,
+      "step": 876
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 13.253562927246094,
+      "learning_rate": 1.8243961110554276e-05,
+      "loss": 2.612,
+      "step": 877
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 21.80856704711914,
+      "learning_rate": 1.8241956499949886e-05,
+      "loss": 3.0365,
+      "step": 878
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 17.941879272460938,
+      "learning_rate": 1.8239951889345496e-05,
+      "loss": 2.9886,
+      "step": 879
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 14.684907913208008,
+      "learning_rate": 1.8237947278741106e-05,
+      "loss": 2.6941,
+      "step": 880
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 19.08102798461914,
+      "learning_rate": 1.8235942668136717e-05,
+      "loss": 3.4722,
+      "step": 881
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 14.615809440612793,
+      "learning_rate": 1.8233938057532327e-05,
+      "loss": 2.9938,
+      "step": 882
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 22.291290283203125,
+      "learning_rate": 1.8231933446927937e-05,
+      "loss": 4.3829,
+      "step": 883
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 16.102741241455078,
+      "learning_rate": 1.8229928836323543e-05,
+      "loss": 3.5552,
+      "step": 884
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 20.495725631713867,
+      "learning_rate": 1.8227924225719157e-05,
+      "loss": 3.3252,
+      "step": 885
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 25.180877685546875,
+      "learning_rate": 1.8225919615114764e-05,
+      "loss": 2.6339,
+      "step": 886
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 14.054975509643555,
+      "learning_rate": 1.8223915004510374e-05,
+      "loss": 3.2697,
+      "step": 887
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 14.858386993408203,
+      "learning_rate": 1.8221910393905987e-05,
+      "loss": 2.3764,
+      "step": 888
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 14.020156860351562,
+      "learning_rate": 1.8219905783301594e-05,
+      "loss": 2.4207,
+      "step": 889
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 20.844745635986328,
+      "learning_rate": 1.8217901172697204e-05,
+      "loss": 2.8965,
+      "step": 890
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 12.82536792755127,
+      "learning_rate": 1.8215896562092814e-05,
+      "loss": 2.1605,
+      "step": 891
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 11.74587345123291,
+      "learning_rate": 1.8213891951488424e-05,
+      "loss": 3.171,
+      "step": 892
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 21.482084274291992,
+      "learning_rate": 1.8211887340884034e-05,
+      "loss": 3.142,
+      "step": 893
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 28.953731536865234,
+      "learning_rate": 1.8209882730279644e-05,
+      "loss": 3.2382,
+      "step": 894
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 17.727638244628906,
+      "learning_rate": 1.8207878119675254e-05,
+      "loss": 3.2516,
+      "step": 895
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 24.647932052612305,
+      "learning_rate": 1.8205873509070864e-05,
+      "loss": 3.1515,
+      "step": 896
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 15.397368431091309,
+      "learning_rate": 1.8203868898466475e-05,
+      "loss": 3.8315,
+      "step": 897
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 10.54749584197998,
+      "learning_rate": 1.8201864287862085e-05,
+      "loss": 1.3177,
+      "step": 898
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 22.014423370361328,
+      "learning_rate": 1.8199859677257695e-05,
+      "loss": 3.8405,
+      "step": 899
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 18.356632232666016,
+      "learning_rate": 1.8197855066653305e-05,
+      "loss": 3.0761,
+      "step": 900
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 19.23956298828125,
+      "learning_rate": 1.8195850456048915e-05,
+      "loss": 3.8153,
+      "step": 901
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 21.569475173950195,
+      "learning_rate": 1.8193845845444525e-05,
+      "loss": 2.913,
+      "step": 902
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 10.557537078857422,
+      "learning_rate": 1.819184123484013e-05,
+      "loss": 2.8867,
+      "step": 903
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 14.843564987182617,
+      "learning_rate": 1.8189836624235745e-05,
+      "loss": 2.9857,
+      "step": 904
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 10.520699501037598,
+      "learning_rate": 1.8187832013631355e-05,
+      "loss": 3.0075,
+      "step": 905
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 18.192188262939453,
+      "learning_rate": 1.8185827403026962e-05,
+      "loss": 4.4008,
+      "step": 906
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 13.926733016967773,
+      "learning_rate": 1.8183822792422575e-05,
+      "loss": 3.1211,
+      "step": 907
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 17.12418556213379,
+      "learning_rate": 1.8181818181818182e-05,
+      "loss": 3.4901,
+      "step": 908
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 16.681673049926758,
+      "learning_rate": 1.8179813571213792e-05,
+      "loss": 3.2524,
+      "step": 909
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 18.500484466552734,
+      "learning_rate": 1.8177808960609402e-05,
+      "loss": 3.7737,
+      "step": 910
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 13.746018409729004,
+      "learning_rate": 1.8175804350005012e-05,
+      "loss": 2.4568,
+      "step": 911
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 12.869297981262207,
+      "learning_rate": 1.8173799739400622e-05,
+      "loss": 2.6976,
+      "step": 912
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 17.630205154418945,
+      "learning_rate": 1.8171795128796232e-05,
+      "loss": 3.2424,
+      "step": 913
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 11.685369491577148,
+      "learning_rate": 1.8169790518191843e-05,
+      "loss": 2.3216,
+      "step": 914
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 23.479877471923828,
+      "learning_rate": 1.8167785907587453e-05,
+      "loss": 3.3934,
+      "step": 915
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 11.527375221252441,
+      "learning_rate": 1.8165781296983063e-05,
+      "loss": 2.9907,
+      "step": 916
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 11.39271354675293,
+      "learning_rate": 1.8163776686378673e-05,
+      "loss": 2.8909,
+      "step": 917
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 18.55435562133789,
+      "learning_rate": 1.8161772075774283e-05,
+      "loss": 2.9944,
+      "step": 918
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 13.561935424804688,
+      "learning_rate": 1.8159767465169893e-05,
+      "loss": 2.385,
+      "step": 919
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 16.513639450073242,
+      "learning_rate": 1.8157762854565503e-05,
+      "loss": 3.0903,
+      "step": 920
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 19.51789093017578,
+      "learning_rate": 1.8155758243961113e-05,
+      "loss": 4.0376,
+      "step": 921
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 16.57154655456543,
+      "learning_rate": 1.815375363335672e-05,
+      "loss": 3.3258,
+      "step": 922
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 12.257933616638184,
+      "learning_rate": 1.8151749022752333e-05,
+      "loss": 2.85,
+      "step": 923
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 19.822603225708008,
+      "learning_rate": 1.8149744412147943e-05,
+      "loss": 2.7183,
+      "step": 924
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 13.144048690795898,
+      "learning_rate": 1.814773980154355e-05,
+      "loss": 2.9005,
+      "step": 925
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 13.302081108093262,
+      "learning_rate": 1.8145735190939164e-05,
+      "loss": 3.5627,
+      "step": 926
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 18.423614501953125,
+      "learning_rate": 1.814373058033477e-05,
+      "loss": 3.7219,
+      "step": 927
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 16.71343231201172,
+      "learning_rate": 1.814172596973038e-05,
+      "loss": 2.3099,
+      "step": 928
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 12.817221641540527,
+      "learning_rate": 1.813972135912599e-05,
+      "loss": 3.1187,
+      "step": 929
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 23.68539047241211,
+      "learning_rate": 1.81377167485216e-05,
+      "loss": 2.8866,
+      "step": 930
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 19.9428768157959,
+      "learning_rate": 1.813571213791721e-05,
+      "loss": 3.3457,
+      "step": 931
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 15.457541465759277,
+      "learning_rate": 1.813370752731282e-05,
+      "loss": 2.5672,
+      "step": 932
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 14.215147972106934,
+      "learning_rate": 1.813170291670843e-05,
+      "loss": 3.2077,
+      "step": 933
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 23.65352439880371,
+      "learning_rate": 1.812969830610404e-05,
+      "loss": 4.0376,
+      "step": 934
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 15.811484336853027,
+      "learning_rate": 1.812769369549965e-05,
+      "loss": 3.4806,
+      "step": 935
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 22.097625732421875,
+      "learning_rate": 1.812568908489526e-05,
+      "loss": 2.9302,
+      "step": 936
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 21.592880249023438,
+      "learning_rate": 1.812368447429087e-05,
+      "loss": 3.5155,
+      "step": 937
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 13.204874992370605,
+      "learning_rate": 1.812167986368648e-05,
+      "loss": 3.172,
+      "step": 938
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 19.575117111206055,
+      "learning_rate": 1.811967525308209e-05,
+      "loss": 3.1032,
+      "step": 939
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 19.439678192138672,
+      "learning_rate": 1.81176706424777e-05,
+      "loss": 3.1572,
+      "step": 940
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 18.003063201904297,
+      "learning_rate": 1.8115666031873308e-05,
+      "loss": 3.6915,
+      "step": 941
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 14.612069129943848,
+      "learning_rate": 1.811366142126892e-05,
+      "loss": 2.6144,
+      "step": 942
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 26.311756134033203,
+      "learning_rate": 1.811165681066453e-05,
+      "loss": 4.1978,
+      "step": 943
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 12.142837524414062,
+      "learning_rate": 1.810965220006014e-05,
+      "loss": 1.9791,
+      "step": 944
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 16.259824752807617,
+      "learning_rate": 1.8107647589455752e-05,
+      "loss": 3.0366,
+      "step": 945
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 15.532299041748047,
+      "learning_rate": 1.810564297885136e-05,
+      "loss": 2.9808,
+      "step": 946
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 31.902145385742188,
+      "learning_rate": 1.810363836824697e-05,
+      "loss": 4.2394,
+      "step": 947
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 20.337215423583984,
+      "learning_rate": 1.8101633757642582e-05,
+      "loss": 3.8077,
+      "step": 948
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 15.51062297821045,
+      "learning_rate": 1.809962914703819e-05,
+      "loss": 1.9024,
+      "step": 949
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 21.733741760253906,
+      "learning_rate": 1.80976245364338e-05,
+      "loss": 3.1563,
+      "step": 950
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 27.472270965576172,
+      "learning_rate": 1.809561992582941e-05,
+      "loss": 3.7646,
+      "step": 951
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 33.5841178894043,
+      "learning_rate": 1.809361531522502e-05,
+      "loss": 3.6314,
+      "step": 952
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 15.488965034484863,
+      "learning_rate": 1.809161070462063e-05,
+      "loss": 3.8817,
+      "step": 953
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 11.746240615844727,
+      "learning_rate": 1.808960609401624e-05,
+      "loss": 2.1518,
+      "step": 954
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 9.70715618133545,
+      "learning_rate": 1.808760148341185e-05,
+      "loss": 2.5758,
+      "step": 955
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 15.384037017822266,
+      "learning_rate": 1.808559687280746e-05,
+      "loss": 3.2143,
+      "step": 956
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 18.4316349029541,
+      "learning_rate": 1.808359226220307e-05,
+      "loss": 4.8752,
+      "step": 957
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 13.498906135559082,
+      "learning_rate": 1.8081587651598676e-05,
+      "loss": 3.2658,
+      "step": 958
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 26.12436866760254,
+      "learning_rate": 1.807958304099429e-05,
+      "loss": 1.9117,
+      "step": 959
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 14.987210273742676,
+      "learning_rate": 1.80775784303899e-05,
+      "loss": 2.1122,
+      "step": 960
+    },
+    {
+      "epoch": 0.29,
+      "eval_loss": 0.5726422667503357,
+      "eval_runtime": 43.6835,
+      "eval_samples_per_second": 33.857,
+      "eval_steps_per_second": 33.857,
+      "step": 960
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 11.189974784851074,
+      "learning_rate": 1.8075573819785506e-05,
+      "loss": 2.7843,
+      "step": 961
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 13.831470489501953,
+      "learning_rate": 1.807356920918112e-05,
+      "loss": 3.8859,
+      "step": 962
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 14.756105422973633,
+      "learning_rate": 1.8071564598576727e-05,
+      "loss": 2.6933,
+      "step": 963
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 13.506275177001953,
+      "learning_rate": 1.8069559987972337e-05,
+      "loss": 2.7927,
+      "step": 964
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 27.999460220336914,
+      "learning_rate": 1.8067555377367947e-05,
+      "loss": 3.3839,
+      "step": 965
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 13.807269096374512,
+      "learning_rate": 1.8065550766763557e-05,
+      "loss": 2.0444,
+      "step": 966
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 15.027080535888672,
+      "learning_rate": 1.8063546156159167e-05,
+      "loss": 2.5851,
+      "step": 967
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 46.652530670166016,
+      "learning_rate": 1.8061541545554777e-05,
+      "loss": 2.5573,
+      "step": 968
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 17.231346130371094,
+      "learning_rate": 1.8059536934950387e-05,
+      "loss": 2.3659,
+      "step": 969
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 17.74903678894043,
+      "learning_rate": 1.8057532324345997e-05,
+      "loss": 2.8746,
+      "step": 970
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 9.483039855957031,
+      "learning_rate": 1.8055527713741607e-05,
+      "loss": 2.6829,
+      "step": 971
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 10.787821769714355,
+      "learning_rate": 1.8053523103137217e-05,
+      "loss": 2.395,
+      "step": 972
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 16.904396057128906,
+      "learning_rate": 1.8051518492532827e-05,
+      "loss": 2.8498,
+      "step": 973
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 18.882564544677734,
+      "learning_rate": 1.8049513881928437e-05,
+      "loss": 3.3251,
+      "step": 974
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 18.128250122070312,
+      "learning_rate": 1.8047509271324048e-05,
+      "loss": 3.1162,
+      "step": 975
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 19.998552322387695,
+      "learning_rate": 1.8045504660719658e-05,
+      "loss": 3.183,
+      "step": 976
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 19.723297119140625,
+      "learning_rate": 1.8043500050115264e-05,
+      "loss": 3.0132,
+      "step": 977
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 18.414817810058594,
+      "learning_rate": 1.8041495439510878e-05,
+      "loss": 2.6902,
+      "step": 978
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 15.55300521850586,
+      "learning_rate": 1.8039490828906488e-05,
+      "loss": 2.0389,
+      "step": 979
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 19.29793357849121,
+      "learning_rate": 1.8037486218302095e-05,
+      "loss": 3.276,
+      "step": 980
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 18.001596450805664,
+      "learning_rate": 1.8035481607697708e-05,
+      "loss": 2.2783,
+      "step": 981
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 12.103147506713867,
+      "learning_rate": 1.8033476997093315e-05,
+      "loss": 3.8083,
+      "step": 982
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 15.265912055969238,
+      "learning_rate": 1.8031472386488925e-05,
+      "loss": 2.1125,
+      "step": 983
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 20.835962295532227,
+      "learning_rate": 1.8029467775884535e-05,
+      "loss": 2.128,
+      "step": 984
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 27.07541847229004,
+      "learning_rate": 1.8027463165280145e-05,
+      "loss": 2.9855,
+      "step": 985
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 17.84712791442871,
+      "learning_rate": 1.8025458554675755e-05,
+      "loss": 2.7988,
+      "step": 986
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 15.661683082580566,
+      "learning_rate": 1.8023453944071365e-05,
+      "loss": 3.1825,
+      "step": 987
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 21.294462203979492,
+      "learning_rate": 1.8021449333466975e-05,
+      "loss": 3.2597,
+      "step": 988
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 14.863018035888672,
+      "learning_rate": 1.8019444722862585e-05,
+      "loss": 2.7602,
+      "step": 989
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 14.493101119995117,
+      "learning_rate": 1.8017440112258195e-05,
+      "loss": 3.5392,
+      "step": 990
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 16.697307586669922,
+      "learning_rate": 1.8015435501653806e-05,
+      "loss": 3.0317,
+      "step": 991
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 21.106035232543945,
+      "learning_rate": 1.8013430891049416e-05,
+      "loss": 2.6286,
+      "step": 992
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 14.834815979003906,
+      "learning_rate": 1.8011426280445026e-05,
+      "loss": 2.2224,
+      "step": 993
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 16.66942596435547,
+      "learning_rate": 1.8009421669840636e-05,
+      "loss": 3.2398,
+      "step": 994
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 20.955211639404297,
+      "learning_rate": 1.8007417059236246e-05,
+      "loss": 3.2694,
+      "step": 995
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 16.78022575378418,
+      "learning_rate": 1.8005412448631853e-05,
+      "loss": 3.3401,
+      "step": 996
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 10.238494873046875,
+      "learning_rate": 1.8003407838027466e-05,
+      "loss": 2.686,
+      "step": 997
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 25.67633628845215,
+      "learning_rate": 1.8001403227423076e-05,
+      "loss": 4.3305,
+      "step": 998
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 18.46412467956543,
+      "learning_rate": 1.7999398616818683e-05,
+      "loss": 4.0384,
+      "step": 999
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 16.967388153076172,
+      "learning_rate": 1.7997394006214296e-05,
+      "loss": 2.752,
+      "step": 1000
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 16.72841453552246,
+      "learning_rate": 1.7995389395609903e-05,
+      "loss": 1.9186,
+      "step": 1001
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 17.65038299560547,
+      "learning_rate": 1.7993384785005513e-05,
+      "loss": 2.8156,
+      "step": 1002
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 15.546381950378418,
+      "learning_rate": 1.7991380174401123e-05,
+      "loss": 2.9373,
+      "step": 1003
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 13.01362133026123,
+      "learning_rate": 1.7989375563796733e-05,
+      "loss": 2.4807,
+      "step": 1004
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 18.31947135925293,
+      "learning_rate": 1.7987370953192343e-05,
+      "loss": 2.6384,
+      "step": 1005
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 19.710500717163086,
+      "learning_rate": 1.7985366342587953e-05,
+      "loss": 2.9314,
+      "step": 1006
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 12.663653373718262,
+      "learning_rate": 1.7983361731983563e-05,
+      "loss": 1.9977,
+      "step": 1007
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 23.557924270629883,
+      "learning_rate": 1.7981357121379174e-05,
+      "loss": 3.6326,
+      "step": 1008
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 12.505498886108398,
+      "learning_rate": 1.7979352510774784e-05,
+      "loss": 2.3971,
+      "step": 1009
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 21.709392547607422,
+      "learning_rate": 1.7977347900170394e-05,
+      "loss": 2.4877,
+      "step": 1010
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 12.365106582641602,
+      "learning_rate": 1.7975343289566004e-05,
+      "loss": 3.1574,
+      "step": 1011
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 18.052745819091797,
+      "learning_rate": 1.7973338678961614e-05,
+      "loss": 3.6156,
+      "step": 1012
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 16.1375732421875,
+      "learning_rate": 1.7971334068357224e-05,
+      "loss": 3.0042,
+      "step": 1013
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 15.52522087097168,
+      "learning_rate": 1.7969329457752834e-05,
+      "loss": 3.3669,
+      "step": 1014
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 18.78600311279297,
+      "learning_rate": 1.796732484714844e-05,
+      "loss": 4.9568,
+      "step": 1015
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 32.19879150390625,
+      "learning_rate": 1.7965320236544054e-05,
+      "loss": 3.4862,
+      "step": 1016
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 22.817644119262695,
+      "learning_rate": 1.7963315625939664e-05,
+      "loss": 3.6447,
+      "step": 1017
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 13.087103843688965,
+      "learning_rate": 1.796131101533527e-05,
+      "loss": 3.1701,
+      "step": 1018
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 15.170973777770996,
+      "learning_rate": 1.7959306404730884e-05,
+      "loss": 2.907,
+      "step": 1019
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 18.237186431884766,
+      "learning_rate": 1.795730179412649e-05,
+      "loss": 3.34,
+      "step": 1020
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 21.131467819213867,
+      "learning_rate": 1.79552971835221e-05,
+      "loss": 3.0868,
+      "step": 1021
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 21.808650970458984,
+      "learning_rate": 1.7953292572917715e-05,
+      "loss": 2.4194,
+      "step": 1022
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 8.451123237609863,
+      "learning_rate": 1.795128796231332e-05,
+      "loss": 2.0371,
+      "step": 1023
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 13.38468074798584,
+      "learning_rate": 1.794928335170893e-05,
+      "loss": 2.1737,
+      "step": 1024
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 11.502458572387695,
+      "learning_rate": 1.794727874110454e-05,
+      "loss": 2.5593,
+      "step": 1025
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 8.056184768676758,
+      "learning_rate": 1.794527413050015e-05,
+      "loss": 2.0176,
+      "step": 1026
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 12.703288078308105,
+      "learning_rate": 1.7943269519895762e-05,
+      "loss": 3.1376,
+      "step": 1027
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 18.933555603027344,
+      "learning_rate": 1.7941264909291372e-05,
+      "loss": 2.6452,
+      "step": 1028
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 19.963056564331055,
+      "learning_rate": 1.7939260298686982e-05,
+      "loss": 3.3245,
+      "step": 1029
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 11.356467247009277,
+      "learning_rate": 1.7937255688082592e-05,
+      "loss": 2.8729,
+      "step": 1030
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 22.96855926513672,
+      "learning_rate": 1.7935251077478202e-05,
+      "loss": 3.4278,
+      "step": 1031
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 9.410229682922363,
+      "learning_rate": 1.793324646687381e-05,
+      "loss": 2.4024,
+      "step": 1032
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 15.325308799743652,
+      "learning_rate": 1.7931241856269422e-05,
+      "loss": 3.0026,
+      "step": 1033
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 17.43353271484375,
+      "learning_rate": 1.7929237245665032e-05,
+      "loss": 3.4274,
+      "step": 1034
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 18.91429328918457,
+      "learning_rate": 1.792723263506064e-05,
+      "loss": 2.8543,
+      "step": 1035
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 13.564133644104004,
+      "learning_rate": 1.7925228024456253e-05,
+      "loss": 1.9484,
+      "step": 1036
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 21.08160972595215,
+      "learning_rate": 1.792322341385186e-05,
+      "loss": 3.0899,
+      "step": 1037
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 11.906721115112305,
+      "learning_rate": 1.792121880324747e-05,
+      "loss": 2.3356,
+      "step": 1038
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 12.753665924072266,
+      "learning_rate": 1.791921419264308e-05,
+      "loss": 1.7421,
+      "step": 1039
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 13.137770652770996,
+      "learning_rate": 1.791720958203869e-05,
+      "loss": 3.0557,
+      "step": 1040
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 18.3424129486084,
+      "learning_rate": 1.7915204971434303e-05,
+      "loss": 2.9359,
+      "step": 1041
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 14.686424255371094,
+      "learning_rate": 1.791320036082991e-05,
+      "loss": 2.947,
+      "step": 1042
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 18.53430938720703,
+      "learning_rate": 1.791119575022552e-05,
+      "loss": 3.6603,
+      "step": 1043
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 12.61064338684082,
+      "learning_rate": 1.790919113962113e-05,
+      "loss": 3.071,
+      "step": 1044
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 16.695178985595703,
+      "learning_rate": 1.790718652901674e-05,
+      "loss": 3.6908,
+      "step": 1045
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 21.377464294433594,
+      "learning_rate": 1.790518191841235e-05,
+      "loss": 2.0132,
+      "step": 1046
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 12.307204246520996,
+      "learning_rate": 1.790317730780796e-05,
+      "loss": 3.3536,
+      "step": 1047
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 18.136398315429688,
+      "learning_rate": 1.790117269720357e-05,
+      "loss": 3.3327,
+      "step": 1048
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 14.812959671020508,
+      "learning_rate": 1.789916808659918e-05,
+      "loss": 2.203,
+      "step": 1049
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 16.73038101196289,
+      "learning_rate": 1.789716347599479e-05,
+      "loss": 2.5606,
+      "step": 1050
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 28.89952278137207,
+      "learning_rate": 1.7895158865390397e-05,
+      "loss": 3.7834,
+      "step": 1051
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 16.880115509033203,
+      "learning_rate": 1.789315425478601e-05,
+      "loss": 2.5196,
+      "step": 1052
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 13.364277839660645,
+      "learning_rate": 1.789114964418162e-05,
+      "loss": 2.9332,
+      "step": 1053
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 34.66802978515625,
+      "learning_rate": 1.7889145033577227e-05,
+      "loss": 3.8191,
+      "step": 1054
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 31.339599609375,
+      "learning_rate": 1.788714042297284e-05,
+      "loss": 2.9259,
+      "step": 1055
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 20.67884063720703,
+      "learning_rate": 1.7885135812368447e-05,
+      "loss": 2.7581,
+      "step": 1056
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 15.182465553283691,
+      "learning_rate": 1.7883131201764058e-05,
+      "loss": 2.8311,
+      "step": 1057
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 18.39655303955078,
+      "learning_rate": 1.7881126591159668e-05,
+      "loss": 2.7088,
+      "step": 1058
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 15.220178604125977,
+      "learning_rate": 1.7879121980555278e-05,
+      "loss": 3.1082,
+      "step": 1059
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 13.494277954101562,
+      "learning_rate": 1.7877117369950888e-05,
+      "loss": 2.8682,
+      "step": 1060
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 19.83290672302246,
+      "learning_rate": 1.7875112759346498e-05,
+      "loss": 3.3298,
+      "step": 1061
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 22.487585067749023,
+      "learning_rate": 1.7873108148742108e-05,
+      "loss": 3.2538,
+      "step": 1062
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 24.89381217956543,
+      "learning_rate": 1.7871103538137718e-05,
+      "loss": 2.5621,
+      "step": 1063
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 17.16600227355957,
+      "learning_rate": 1.7869098927533328e-05,
+      "loss": 2.8225,
+      "step": 1064
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 10.632941246032715,
+      "learning_rate": 1.7867094316928938e-05,
+      "loss": 3.0174,
+      "step": 1065
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 20.951936721801758,
+      "learning_rate": 1.7865089706324548e-05,
+      "loss": 3.5367,
+      "step": 1066
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 15.389318466186523,
+      "learning_rate": 1.786308509572016e-05,
+      "loss": 2.2259,
+      "step": 1067
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 44.66872787475586,
+      "learning_rate": 1.786108048511577e-05,
+      "loss": 4.754,
+      "step": 1068
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 12.777280807495117,
+      "learning_rate": 1.785907587451138e-05,
+      "loss": 3.2503,
+      "step": 1069
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 130.45664978027344,
+      "learning_rate": 1.7857071263906985e-05,
+      "loss": 2.8328,
+      "step": 1070
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 25.45984649658203,
+      "learning_rate": 1.78550666533026e-05,
+      "loss": 2.7516,
+      "step": 1071
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 15.323507308959961,
+      "learning_rate": 1.785306204269821e-05,
+      "loss": 2.0025,
+      "step": 1072
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 12.247516632080078,
+      "learning_rate": 1.7851057432093815e-05,
+      "loss": 2.7504,
+      "step": 1073
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 33.093528747558594,
+      "learning_rate": 1.784905282148943e-05,
+      "loss": 3.8759,
+      "step": 1074
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 19.78754234313965,
+      "learning_rate": 1.7847048210885036e-05,
+      "loss": 3.0278,
+      "step": 1075
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 15.436639785766602,
+      "learning_rate": 1.7845043600280646e-05,
+      "loss": 3.5511,
+      "step": 1076
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 12.693981170654297,
+      "learning_rate": 1.784303898967626e-05,
+      "loss": 3.4346,
+      "step": 1077
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 17.650697708129883,
+      "learning_rate": 1.7841034379071866e-05,
+      "loss": 2.3388,
+      "step": 1078
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 18.479936599731445,
+      "learning_rate": 1.7839029768467476e-05,
+      "loss": 3.4169,
+      "step": 1079
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 14.039420127868652,
+      "learning_rate": 1.7837025157863086e-05,
+      "loss": 3.0089,
+      "step": 1080
+    },
+    {
+      "epoch": 0.32,
+      "eval_loss": 0.5645097494125366,
+      "eval_runtime": 43.9047,
+      "eval_samples_per_second": 33.687,
+      "eval_steps_per_second": 33.687,
+      "step": 1080
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 18.384023666381836,
+      "learning_rate": 1.7835020547258696e-05,
+      "loss": 5.0499,
+      "step": 1081
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 14.556102752685547,
+      "learning_rate": 1.7833015936654306e-05,
+      "loss": 2.7198,
+      "step": 1082
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 16.748477935791016,
+      "learning_rate": 1.7831011326049916e-05,
+      "loss": 2.7809,
+      "step": 1083
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 16.843488693237305,
+      "learning_rate": 1.7829006715445526e-05,
+      "loss": 3.1988,
+      "step": 1084
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 60.20880889892578,
+      "learning_rate": 1.7827002104841137e-05,
+      "loss": 4.0739,
+      "step": 1085
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 22.219043731689453,
+      "learning_rate": 1.7824997494236747e-05,
+      "loss": 3.1217,
+      "step": 1086
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 15.742715835571289,
+      "learning_rate": 1.7822992883632357e-05,
+      "loss": 2.1444,
+      "step": 1087
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 13.159707069396973,
+      "learning_rate": 1.7820988273027967e-05,
+      "loss": 3.7656,
+      "step": 1088
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 18.79963493347168,
+      "learning_rate": 1.7818983662423573e-05,
+      "loss": 3.2334,
+      "step": 1089
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 12.51858139038086,
+      "learning_rate": 1.7816979051819187e-05,
+      "loss": 1.6161,
+      "step": 1090
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 18.549959182739258,
+      "learning_rate": 1.7814974441214797e-05,
+      "loss": 2.2857,
+      "step": 1091
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 17.99864959716797,
+      "learning_rate": 1.7812969830610404e-05,
+      "loss": 2.2779,
+      "step": 1092
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 16.206727981567383,
+      "learning_rate": 1.7810965220006017e-05,
+      "loss": 2.1039,
+      "step": 1093
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 14.30031681060791,
+      "learning_rate": 1.7808960609401624e-05,
+      "loss": 3.3124,
+      "step": 1094
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 48.405094146728516,
+      "learning_rate": 1.7806955998797234e-05,
+      "loss": 3.8335,
+      "step": 1095
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 17.39105224609375,
+      "learning_rate": 1.7804951388192847e-05,
+      "loss": 2.8733,
+      "step": 1096
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 23.60736656188965,
+      "learning_rate": 1.7802946777588454e-05,
+      "loss": 2.9195,
+      "step": 1097
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 18.69573402404785,
+      "learning_rate": 1.7800942166984064e-05,
+      "loss": 3.4923,
+      "step": 1098
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 31.806943893432617,
+      "learning_rate": 1.7798937556379674e-05,
+      "loss": 2.9115,
+      "step": 1099
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 13.861745834350586,
+      "learning_rate": 1.7796932945775284e-05,
+      "loss": 1.3565,
+      "step": 1100
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 23.198631286621094,
+      "learning_rate": 1.7794928335170894e-05,
+      "loss": 3.0973,
+      "step": 1101
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 16.94276237487793,
+      "learning_rate": 1.7792923724566505e-05,
+      "loss": 3.5986,
+      "step": 1102
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 9.507773399353027,
+      "learning_rate": 1.7790919113962115e-05,
+      "loss": 1.7677,
+      "step": 1103
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 61.569091796875,
+      "learning_rate": 1.7788914503357725e-05,
+      "loss": 2.5007,
+      "step": 1104
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 16.81398582458496,
+      "learning_rate": 1.7786909892753335e-05,
+      "loss": 3.5078,
+      "step": 1105
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 15.168441772460938,
+      "learning_rate": 1.778490528214894e-05,
+      "loss": 2.3151,
+      "step": 1106
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 16.73635482788086,
+      "learning_rate": 1.7782900671544555e-05,
+      "loss": 3.0094,
+      "step": 1107
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 16.253849029541016,
+      "learning_rate": 1.7780896060940165e-05,
+      "loss": 2.7682,
+      "step": 1108
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 14.130172729492188,
+      "learning_rate": 1.7778891450335775e-05,
+      "loss": 3.0624,
+      "step": 1109
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 17.684720993041992,
+      "learning_rate": 1.7776886839731385e-05,
+      "loss": 3.3596,
+      "step": 1110
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 29.59230613708496,
+      "learning_rate": 1.7774882229126992e-05,
+      "loss": 3.1807,
+      "step": 1111
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 11.93481159210205,
+      "learning_rate": 1.7772877618522605e-05,
+      "loss": 2.8107,
+      "step": 1112
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 16.962127685546875,
+      "learning_rate": 1.7770873007918212e-05,
+      "loss": 2.9108,
+      "step": 1113
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 13.559270858764648,
+      "learning_rate": 1.7768868397313822e-05,
+      "loss": 3.5231,
+      "step": 1114
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 12.885640144348145,
+      "learning_rate": 1.7766863786709436e-05,
+      "loss": 2.7105,
+      "step": 1115
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 29.005674362182617,
+      "learning_rate": 1.7764859176105042e-05,
+      "loss": 3.413,
+      "step": 1116
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 13.02987289428711,
+      "learning_rate": 1.7762854565500652e-05,
+      "loss": 1.8334,
+      "step": 1117
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 14.34976863861084,
+      "learning_rate": 1.7760849954896263e-05,
+      "loss": 1.5514,
+      "step": 1118
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 22.86460304260254,
+      "learning_rate": 1.7758845344291873e-05,
+      "loss": 3.4641,
+      "step": 1119
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 28.6113224029541,
+      "learning_rate": 1.7756840733687483e-05,
+      "loss": 3.761,
+      "step": 1120
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 16.072839736938477,
+      "learning_rate": 1.7754836123083093e-05,
+      "loss": 1.5352,
+      "step": 1121
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 18.935298919677734,
+      "learning_rate": 1.7752831512478703e-05,
+      "loss": 2.8795,
+      "step": 1122
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 22.4312686920166,
+      "learning_rate": 1.7750826901874313e-05,
+      "loss": 4.1038,
+      "step": 1123
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 13.36023998260498,
+      "learning_rate": 1.7748822291269923e-05,
+      "loss": 2.9175,
+      "step": 1124
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 10.283577919006348,
+      "learning_rate": 1.774681768066553e-05,
+      "loss": 1.6504,
+      "step": 1125
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 16.4921817779541,
+      "learning_rate": 1.7744813070061143e-05,
+      "loss": 3.4366,
+      "step": 1126
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 24.69314193725586,
+      "learning_rate": 1.7742808459456753e-05,
+      "loss": 3.5184,
+      "step": 1127
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 15.775534629821777,
+      "learning_rate": 1.774080384885236e-05,
+      "loss": 3.0247,
+      "step": 1128
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 35.886539459228516,
+      "learning_rate": 1.7738799238247973e-05,
+      "loss": 3.0789,
+      "step": 1129
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 13.481932640075684,
+      "learning_rate": 1.773679462764358e-05,
+      "loss": 2.6866,
+      "step": 1130
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 13.571722030639648,
+      "learning_rate": 1.773479001703919e-05,
+      "loss": 3.6435,
+      "step": 1131
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 12.800562858581543,
+      "learning_rate": 1.77327854064348e-05,
+      "loss": 2.6728,
+      "step": 1132
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 28.864028930664062,
+      "learning_rate": 1.773078079583041e-05,
+      "loss": 1.8873,
+      "step": 1133
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 23.177940368652344,
+      "learning_rate": 1.772877618522602e-05,
+      "loss": 3.7982,
+      "step": 1134
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 18.07777214050293,
+      "learning_rate": 1.772677157462163e-05,
+      "loss": 3.1977,
+      "step": 1135
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 25.78776741027832,
+      "learning_rate": 1.772476696401724e-05,
+      "loss": 2.4591,
+      "step": 1136
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 14.545743942260742,
+      "learning_rate": 1.772276235341285e-05,
+      "loss": 2.5043,
+      "step": 1137
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 12.708026885986328,
+      "learning_rate": 1.772075774280846e-05,
+      "loss": 3.234,
+      "step": 1138
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 15.227660179138184,
+      "learning_rate": 1.771875313220407e-05,
+      "loss": 3.9899,
+      "step": 1139
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 23.074007034301758,
+      "learning_rate": 1.771674852159968e-05,
+      "loss": 1.94,
+      "step": 1140
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 20.678417205810547,
+      "learning_rate": 1.771474391099529e-05,
+      "loss": 3.0762,
+      "step": 1141
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 10.947648048400879,
+      "learning_rate": 1.77127393003909e-05,
+      "loss": 3.0063,
+      "step": 1142
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 12.576761245727539,
+      "learning_rate": 1.771073468978651e-05,
+      "loss": 2.5958,
+      "step": 1143
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 25.190330505371094,
+      "learning_rate": 1.7708730079182118e-05,
+      "loss": 2.9465,
+      "step": 1144
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 14.625896453857422,
+      "learning_rate": 1.770672546857773e-05,
+      "loss": 2.337,
+      "step": 1145
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 13.233895301818848,
+      "learning_rate": 1.770472085797334e-05,
+      "loss": 2.7883,
+      "step": 1146
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 11.731785774230957,
+      "learning_rate": 1.7702716247368948e-05,
+      "loss": 2.4522,
+      "step": 1147
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 13.378613471984863,
+      "learning_rate": 1.770071163676456e-05,
+      "loss": 2.6688,
+      "step": 1148
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 34.357383728027344,
+      "learning_rate": 1.769870702616017e-05,
+      "loss": 3.0263,
+      "step": 1149
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 13.445833206176758,
+      "learning_rate": 1.769670241555578e-05,
+      "loss": 1.7473,
+      "step": 1150
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 13.28929328918457,
+      "learning_rate": 1.7694697804951392e-05,
+      "loss": 2.3902,
+      "step": 1151
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 19.005510330200195,
+      "learning_rate": 1.7692693194347e-05,
+      "loss": 2.506,
+      "step": 1152
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 18.774620056152344,
+      "learning_rate": 1.769068858374261e-05,
+      "loss": 2.8765,
+      "step": 1153
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 21.18247413635254,
+      "learning_rate": 1.768868397313822e-05,
+      "loss": 2.4275,
+      "step": 1154
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 22.34467124938965,
+      "learning_rate": 1.768667936253383e-05,
+      "loss": 2.6174,
+      "step": 1155
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 28.30898094177246,
+      "learning_rate": 1.768467475192944e-05,
+      "loss": 3.1029,
+      "step": 1156
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 12.033406257629395,
+      "learning_rate": 1.768267014132505e-05,
+      "loss": 3.4031,
+      "step": 1157
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 16.54393768310547,
+      "learning_rate": 1.768066553072066e-05,
+      "loss": 2.6641,
+      "step": 1158
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 29.7037410736084,
+      "learning_rate": 1.767866092011627e-05,
+      "loss": 2.8023,
+      "step": 1159
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 17.149572372436523,
+      "learning_rate": 1.767665630951188e-05,
+      "loss": 3.3883,
+      "step": 1160
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 32.76160430908203,
+      "learning_rate": 1.767465169890749e-05,
+      "loss": 3.1329,
+      "step": 1161
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 13.081724166870117,
+      "learning_rate": 1.76726470883031e-05,
+      "loss": 3.1289,
+      "step": 1162
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 29.665246963500977,
+      "learning_rate": 1.767064247769871e-05,
+      "loss": 3.6106,
+      "step": 1163
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 19.989797592163086,
+      "learning_rate": 1.766863786709432e-05,
+      "loss": 3.4249,
+      "step": 1164
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 11.273635864257812,
+      "learning_rate": 1.766663325648993e-05,
+      "loss": 3.0172,
+      "step": 1165
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 25.350624084472656,
+      "learning_rate": 1.7664628645885536e-05,
+      "loss": 2.8628,
+      "step": 1166
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 14.034561157226562,
+      "learning_rate": 1.766262403528115e-05,
+      "loss": 2.7608,
+      "step": 1167
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 20.618732452392578,
+      "learning_rate": 1.7660619424676757e-05,
+      "loss": 2.6732,
+      "step": 1168
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 18.557788848876953,
+      "learning_rate": 1.7658614814072367e-05,
+      "loss": 4.2311,
+      "step": 1169
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 22.748462677001953,
+      "learning_rate": 1.765661020346798e-05,
+      "loss": 3.0337,
+      "step": 1170
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 15.054831504821777,
+      "learning_rate": 1.7654605592863587e-05,
+      "loss": 2.5534,
+      "step": 1171
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 12.077056884765625,
+      "learning_rate": 1.7652600982259197e-05,
+      "loss": 1.9834,
+      "step": 1172
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 17.17857551574707,
+      "learning_rate": 1.7650596371654807e-05,
+      "loss": 3.1654,
+      "step": 1173
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 13.104513168334961,
+      "learning_rate": 1.7648591761050417e-05,
+      "loss": 2.999,
+      "step": 1174
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 14.851997375488281,
+      "learning_rate": 1.7646587150446027e-05,
+      "loss": 3.3295,
+      "step": 1175
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 12.090683937072754,
+      "learning_rate": 1.7644582539841637e-05,
+      "loss": 2.4796,
+      "step": 1176
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 13.724775314331055,
+      "learning_rate": 1.7642577929237247e-05,
+      "loss": 2.9697,
+      "step": 1177
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 26.366907119750977,
+      "learning_rate": 1.7640573318632857e-05,
+      "loss": 3.2335,
+      "step": 1178
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 10.779017448425293,
+      "learning_rate": 1.7638568708028467e-05,
+      "loss": 2.4018,
+      "step": 1179
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 22.231727600097656,
+      "learning_rate": 1.7636564097424078e-05,
+      "loss": 3.2377,
+      "step": 1180
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 18.77667236328125,
+      "learning_rate": 1.7634559486819688e-05,
+      "loss": 3.0199,
+      "step": 1181
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 16.207422256469727,
+      "learning_rate": 1.7632554876215298e-05,
+      "loss": 2.7055,
+      "step": 1182
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 16.809715270996094,
+      "learning_rate": 1.7630550265610908e-05,
+      "loss": 2.9821,
+      "step": 1183
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 16.3651065826416,
+      "learning_rate": 1.7628545655006518e-05,
+      "loss": 2.1699,
+      "step": 1184
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 11.468558311462402,
+      "learning_rate": 1.7626541044402125e-05,
+      "loss": 2.6325,
+      "step": 1185
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 16.051881790161133,
+      "learning_rate": 1.7624536433797738e-05,
+      "loss": 2.3443,
+      "step": 1186
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 12.057586669921875,
+      "learning_rate": 1.7622531823193345e-05,
+      "loss": 2.7623,
+      "step": 1187
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 17.24924659729004,
+      "learning_rate": 1.7620527212588955e-05,
+      "loss": 3.3553,
+      "step": 1188
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 12.06871223449707,
+      "learning_rate": 1.761852260198457e-05,
+      "loss": 1.9598,
+      "step": 1189
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 20.307418823242188,
+      "learning_rate": 1.7616517991380175e-05,
+      "loss": 3.1637,
+      "step": 1190
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 11.176687240600586,
+      "learning_rate": 1.7614513380775785e-05,
+      "loss": 2.5982,
+      "step": 1191
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 14.596364974975586,
+      "learning_rate": 1.7612508770171395e-05,
+      "loss": 2.5782,
+      "step": 1192
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 16.369272232055664,
+      "learning_rate": 1.7610504159567005e-05,
+      "loss": 2.7876,
+      "step": 1193
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 16.84731101989746,
+      "learning_rate": 1.7608499548962615e-05,
+      "loss": 2.9056,
+      "step": 1194
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 51.60824203491211,
+      "learning_rate": 1.7606494938358225e-05,
+      "loss": 2.3254,
+      "step": 1195
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 18.47427749633789,
+      "learning_rate": 1.7604490327753836e-05,
+      "loss": 2.9355,
+      "step": 1196
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 15.123434066772461,
+      "learning_rate": 1.7602485717149446e-05,
+      "loss": 2.588,
+      "step": 1197
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 19.959136962890625,
+      "learning_rate": 1.7600481106545056e-05,
+      "loss": 2.5238,
+      "step": 1198
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 19.62584114074707,
+      "learning_rate": 1.7598476495940662e-05,
+      "loss": 3.2818,
+      "step": 1199
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 19.437580108642578,
+      "learning_rate": 1.7596471885336276e-05,
+      "loss": 2.8019,
+      "step": 1200
+    },
+    {
+      "epoch": 0.36,
+      "eval_loss": 0.5124673843383789,
+      "eval_runtime": 43.4829,
+      "eval_samples_per_second": 34.013,
+      "eval_steps_per_second": 34.013,
+      "step": 1200
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 41.26815414428711,
+      "learning_rate": 1.7594467274731886e-05,
+      "loss": 3.0088,
+      "step": 1201
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 23.42255973815918,
+      "learning_rate": 1.7592462664127493e-05,
+      "loss": 3.0811,
+      "step": 1202
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 24.135406494140625,
+      "learning_rate": 1.7590458053523106e-05,
+      "loss": 3.4351,
+      "step": 1203
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 11.820728302001953,
+      "learning_rate": 1.7588453442918713e-05,
+      "loss": 2.2131,
+      "step": 1204
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 12.589012145996094,
+      "learning_rate": 1.7586448832314323e-05,
+      "loss": 2.5815,
+      "step": 1205
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 14.668851852416992,
+      "learning_rate": 1.7584444221709936e-05,
+      "loss": 2.9765,
+      "step": 1206
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 13.122230529785156,
+      "learning_rate": 1.7582439611105543e-05,
+      "loss": 2.3476,
+      "step": 1207
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 17.157947540283203,
+      "learning_rate": 1.7580435000501153e-05,
+      "loss": 2.8408,
+      "step": 1208
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 21.18256950378418,
+      "learning_rate": 1.7578430389896763e-05,
+      "loss": 2.5619,
+      "step": 1209
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 19.1740779876709,
+      "learning_rate": 1.7576425779292373e-05,
+      "loss": 4.0597,
+      "step": 1210
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 20.59099769592285,
+      "learning_rate": 1.7574421168687983e-05,
+      "loss": 1.8192,
+      "step": 1211
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 31.291309356689453,
+      "learning_rate": 1.7572416558083593e-05,
+      "loss": 3.6861,
+      "step": 1212
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 11.655035018920898,
+      "learning_rate": 1.7570411947479204e-05,
+      "loss": 2.3163,
+      "step": 1213
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 27.356048583984375,
+      "learning_rate": 1.7568407336874814e-05,
+      "loss": 3.9249,
+      "step": 1214
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 22.89837646484375,
+      "learning_rate": 1.7566402726270424e-05,
+      "loss": 2.6824,
+      "step": 1215
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 16.46586799621582,
+      "learning_rate": 1.7564398115666034e-05,
+      "loss": 2.5388,
+      "step": 1216
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 18.528961181640625,
+      "learning_rate": 1.7562393505061644e-05,
+      "loss": 2.435,
+      "step": 1217
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 10.361041069030762,
+      "learning_rate": 1.756038889445725e-05,
+      "loss": 2.5633,
+      "step": 1218
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 17.265287399291992,
+      "learning_rate": 1.7558384283852864e-05,
+      "loss": 2.3523,
+      "step": 1219
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 18.783292770385742,
+      "learning_rate": 1.7556379673248474e-05,
+      "loss": 2.3425,
+      "step": 1220
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 21.62091827392578,
+      "learning_rate": 1.755437506264408e-05,
+      "loss": 2.7889,
+      "step": 1221
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 19.435802459716797,
+      "learning_rate": 1.7552370452039694e-05,
+      "loss": 3.4363,
+      "step": 1222
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 13.962774276733398,
+      "learning_rate": 1.75503658414353e-05,
+      "loss": 1.6731,
+      "step": 1223
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 10.637103080749512,
+      "learning_rate": 1.754836123083091e-05,
+      "loss": 1.9421,
+      "step": 1224
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 12.985700607299805,
+      "learning_rate": 1.7546356620226525e-05,
+      "loss": 2.2144,
+      "step": 1225
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 16.512508392333984,
+      "learning_rate": 1.754435200962213e-05,
+      "loss": 2.4662,
+      "step": 1226
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 20.497188568115234,
+      "learning_rate": 1.754234739901774e-05,
+      "loss": 2.0858,
+      "step": 1227
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 32.291175842285156,
+      "learning_rate": 1.754034278841335e-05,
+      "loss": 3.3415,
+      "step": 1228
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 16.312284469604492,
+      "learning_rate": 1.753833817780896e-05,
+      "loss": 2.2323,
+      "step": 1229
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 23.16538429260254,
+      "learning_rate": 1.753633356720457e-05,
+      "loss": 3.7011,
+      "step": 1230
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 9.652095794677734,
+      "learning_rate": 1.7534328956600182e-05,
+      "loss": 2.5907,
+      "step": 1231
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 16.361186981201172,
+      "learning_rate": 1.7532324345995792e-05,
+      "loss": 3.1489,
+      "step": 1232
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 24.299665451049805,
+      "learning_rate": 1.7530319735391402e-05,
+      "loss": 2.751,
+      "step": 1233
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 11.663421630859375,
+      "learning_rate": 1.7528315124787012e-05,
+      "loss": 2.2756,
+      "step": 1234
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 16.520639419555664,
+      "learning_rate": 1.7526310514182622e-05,
+      "loss": 1.8753,
+      "step": 1235
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 30.01593780517578,
+      "learning_rate": 1.7524305903578232e-05,
+      "loss": 1.1438,
+      "step": 1236
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 28.729244232177734,
+      "learning_rate": 1.7522301292973842e-05,
+      "loss": 3.4131,
+      "step": 1237
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 18.10856819152832,
+      "learning_rate": 1.7520296682369452e-05,
+      "loss": 2.8804,
+      "step": 1238
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 18.52988052368164,
+      "learning_rate": 1.7518292071765062e-05,
+      "loss": 2.57,
+      "step": 1239
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 18.071670532226562,
+      "learning_rate": 1.751628746116067e-05,
+      "loss": 2.5888,
+      "step": 1240
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 9.437912940979004,
+      "learning_rate": 1.7514282850556283e-05,
+      "loss": 2.1698,
+      "step": 1241
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 11.90031623840332,
+      "learning_rate": 1.751227823995189e-05,
+      "loss": 2.5627,
+      "step": 1242
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 12.437149047851562,
+      "learning_rate": 1.75102736293475e-05,
+      "loss": 2.2992,
+      "step": 1243
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 27.860820770263672,
+      "learning_rate": 1.7508269018743113e-05,
+      "loss": 3.5747,
+      "step": 1244
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 26.30794334411621,
+      "learning_rate": 1.750626440813872e-05,
+      "loss": 3.8743,
+      "step": 1245
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 17.703901290893555,
+      "learning_rate": 1.750425979753433e-05,
+      "loss": 2.338,
+      "step": 1246
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 10.545022010803223,
+      "learning_rate": 1.750225518692994e-05,
+      "loss": 1.9899,
+      "step": 1247
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 24.279394149780273,
+      "learning_rate": 1.750025057632555e-05,
+      "loss": 2.9463,
+      "step": 1248
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 21.58678436279297,
+      "learning_rate": 1.749824596572116e-05,
+      "loss": 3.4896,
+      "step": 1249
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 13.904131889343262,
+      "learning_rate": 1.749624135511677e-05,
+      "loss": 3.1997,
+      "step": 1250
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 10.544210433959961,
+      "learning_rate": 1.749423674451238e-05,
+      "loss": 1.9443,
+      "step": 1251
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 16.909954071044922,
+      "learning_rate": 1.749223213390799e-05,
+      "loss": 3.1345,
+      "step": 1252
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 18.661376953125,
+      "learning_rate": 1.74902275233036e-05,
+      "loss": 3.7376,
+      "step": 1253
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 16.603628158569336,
+      "learning_rate": 1.748822291269921e-05,
+      "loss": 2.108,
+      "step": 1254
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 12.670201301574707,
+      "learning_rate": 1.748621830209482e-05,
+      "loss": 3.4042,
+      "step": 1255
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 12.029052734375,
+      "learning_rate": 1.748421369149043e-05,
+      "loss": 2.9556,
+      "step": 1256
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 28.61595916748047,
+      "learning_rate": 1.748220908088604e-05,
+      "loss": 2.0854,
+      "step": 1257
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 22.586305618286133,
+      "learning_rate": 1.748020447028165e-05,
+      "loss": 2.461,
+      "step": 1258
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 14.331897735595703,
+      "learning_rate": 1.7478199859677257e-05,
+      "loss": 3.5532,
+      "step": 1259
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 18.208744049072266,
+      "learning_rate": 1.747619524907287e-05,
+      "loss": 2.59,
+      "step": 1260
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 26.50733757019043,
+      "learning_rate": 1.7474190638468477e-05,
+      "loss": 3.567,
+      "step": 1261
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 17.948020935058594,
+      "learning_rate": 1.7472186027864088e-05,
+      "loss": 2.5004,
+      "step": 1262
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 11.927952766418457,
+      "learning_rate": 1.74701814172597e-05,
+      "loss": 2.1265,
+      "step": 1263
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 26.811809539794922,
+      "learning_rate": 1.7468176806655308e-05,
+      "loss": 2.7967,
+      "step": 1264
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 82.8468246459961,
+      "learning_rate": 1.7466172196050918e-05,
+      "loss": 2.652,
+      "step": 1265
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 13.454938888549805,
+      "learning_rate": 1.7464167585446528e-05,
+      "loss": 2.282,
+      "step": 1266
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 20.55266571044922,
+      "learning_rate": 1.7462162974842138e-05,
+      "loss": 2.3296,
+      "step": 1267
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 10.279245376586914,
+      "learning_rate": 1.7460158364237748e-05,
+      "loss": 1.3453,
+      "step": 1268
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 11.651695251464844,
+      "learning_rate": 1.7458153753633358e-05,
+      "loss": 2.5631,
+      "step": 1269
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 12.197574615478516,
+      "learning_rate": 1.7456149143028968e-05,
+      "loss": 3.0425,
+      "step": 1270
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 24.782493591308594,
+      "learning_rate": 1.7454144532424578e-05,
+      "loss": 3.1392,
+      "step": 1271
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 10.844775199890137,
+      "learning_rate": 1.745213992182019e-05,
+      "loss": 2.9512,
+      "step": 1272
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 13.511686325073242,
+      "learning_rate": 1.7450135311215795e-05,
+      "loss": 2.8739,
+      "step": 1273
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 21.19719886779785,
+      "learning_rate": 1.744813070061141e-05,
+      "loss": 2.8494,
+      "step": 1274
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 17.959545135498047,
+      "learning_rate": 1.744612609000702e-05,
+      "loss": 2.8389,
+      "step": 1275
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 17.719860076904297,
+      "learning_rate": 1.7444121479402625e-05,
+      "loss": 2.7844,
+      "step": 1276
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 15.047807693481445,
+      "learning_rate": 1.744211686879824e-05,
+      "loss": 2.5462,
+      "step": 1277
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 18.358489990234375,
+      "learning_rate": 1.7440112258193846e-05,
+      "loss": 2.8936,
+      "step": 1278
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 20.895231246948242,
+      "learning_rate": 1.743810764758946e-05,
+      "loss": 3.1691,
+      "step": 1279
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 10.4979248046875,
+      "learning_rate": 1.743610303698507e-05,
+      "loss": 1.5841,
+      "step": 1280
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 14.42455768585205,
+      "learning_rate": 1.7434098426380676e-05,
+      "loss": 2.7656,
+      "step": 1281
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 11.873161315917969,
+      "learning_rate": 1.743209381577629e-05,
+      "loss": 2.9719,
+      "step": 1282
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 23.232919692993164,
+      "learning_rate": 1.7430089205171896e-05,
+      "loss": 2.5282,
+      "step": 1283
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 15.954151153564453,
+      "learning_rate": 1.7428084594567506e-05,
+      "loss": 3.3602,
+      "step": 1284
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 15.332113265991211,
+      "learning_rate": 1.7426079983963116e-05,
+      "loss": 3.1367,
+      "step": 1285
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 17.88947296142578,
+      "learning_rate": 1.7424075373358726e-05,
+      "loss": 2.8707,
+      "step": 1286
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 14.455220222473145,
+      "learning_rate": 1.7422070762754336e-05,
+      "loss": 2.7743,
+      "step": 1287
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 12.738212585449219,
+      "learning_rate": 1.7420066152149946e-05,
+      "loss": 2.4288,
+      "step": 1288
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 17.96286392211914,
+      "learning_rate": 1.7418061541545556e-05,
+      "loss": 2.1277,
+      "step": 1289
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 10.881392478942871,
+      "learning_rate": 1.7416056930941167e-05,
+      "loss": 3.3985,
+      "step": 1290
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 9.816652297973633,
+      "learning_rate": 1.7414052320336777e-05,
+      "loss": 1.9979,
+      "step": 1291
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 14.846339225769043,
+      "learning_rate": 1.7412047709732387e-05,
+      "loss": 3.2407,
+      "step": 1292
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 24.287704467773438,
+      "learning_rate": 1.7410043099127997e-05,
+      "loss": 3.1519,
+      "step": 1293
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 18.814102172851562,
+      "learning_rate": 1.7408038488523607e-05,
+      "loss": 3.9634,
+      "step": 1294
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 21.42952537536621,
+      "learning_rate": 1.7406033877919214e-05,
+      "loss": 3.3032,
+      "step": 1295
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 14.19246768951416,
+      "learning_rate": 1.7404029267314827e-05,
+      "loss": 2.3004,
+      "step": 1296
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 32.35289001464844,
+      "learning_rate": 1.7402024656710434e-05,
+      "loss": 4.9394,
+      "step": 1297
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 14.341626167297363,
+      "learning_rate": 1.7400020046106044e-05,
+      "loss": 2.9232,
+      "step": 1298
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 13.265176773071289,
+      "learning_rate": 1.7398015435501657e-05,
+      "loss": 3.7512,
+      "step": 1299
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 15.265084266662598,
+      "learning_rate": 1.7396010824897264e-05,
+      "loss": 3.0586,
+      "step": 1300
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 19.14535903930664,
+      "learning_rate": 1.7394006214292874e-05,
+      "loss": 3.9271,
+      "step": 1301
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 19.937034606933594,
+      "learning_rate": 1.7392001603688484e-05,
+      "loss": 1.9253,
+      "step": 1302
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 15.879293441772461,
+      "learning_rate": 1.7389996993084094e-05,
+      "loss": 2.5909,
+      "step": 1303
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 22.646526336669922,
+      "learning_rate": 1.7387992382479704e-05,
+      "loss": 2.1147,
+      "step": 1304
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 19.122434616088867,
+      "learning_rate": 1.7385987771875314e-05,
+      "loss": 2.6576,
+      "step": 1305
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 16.39354705810547,
+      "learning_rate": 1.7383983161270924e-05,
+      "loss": 2.3048,
+      "step": 1306
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 21.781461715698242,
+      "learning_rate": 1.7381978550666535e-05,
+      "loss": 2.5542,
+      "step": 1307
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 13.316234588623047,
+      "learning_rate": 1.7379973940062145e-05,
+      "loss": 2.669,
+      "step": 1308
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 11.472354888916016,
+      "learning_rate": 1.7377969329457755e-05,
+      "loss": 2.7657,
+      "step": 1309
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 25.965810775756836,
+      "learning_rate": 1.7375964718853365e-05,
+      "loss": 2.8167,
+      "step": 1310
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 17.545780181884766,
+      "learning_rate": 1.7373960108248975e-05,
+      "loss": 3.5917,
+      "step": 1311
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 22.960935592651367,
+      "learning_rate": 1.7371955497644585e-05,
+      "loss": 3.4562,
+      "step": 1312
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 11.906625747680664,
+      "learning_rate": 1.7369950887040195e-05,
+      "loss": 3.293,
+      "step": 1313
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 14.520302772521973,
+      "learning_rate": 1.7367946276435802e-05,
+      "loss": 2.368,
+      "step": 1314
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 9.339329719543457,
+      "learning_rate": 1.7365941665831415e-05,
+      "loss": 2.1992,
+      "step": 1315
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 19.66999626159668,
+      "learning_rate": 1.7363937055227022e-05,
+      "loss": 3.1317,
+      "step": 1316
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 19.995849609375,
+      "learning_rate": 1.7361932444622632e-05,
+      "loss": 2.606,
+      "step": 1317
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 16.427093505859375,
+      "learning_rate": 1.7359927834018245e-05,
+      "loss": 1.7637,
+      "step": 1318
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 14.198533058166504,
+      "learning_rate": 1.7357923223413852e-05,
+      "loss": 2.904,
+      "step": 1319
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 12.009779930114746,
+      "learning_rate": 1.7355918612809462e-05,
+      "loss": 2.2005,
+      "step": 1320
+    },
+    {
+      "epoch": 0.4,
+      "eval_loss": 0.39250195026397705,
+      "eval_runtime": 43.6691,
+      "eval_samples_per_second": 33.868,
+      "eval_steps_per_second": 33.868,
+      "step": 1320
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 11.49736499786377,
+      "learning_rate": 1.7353914002205072e-05,
+      "loss": 2.9098,
+      "step": 1321
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 12.25181770324707,
+      "learning_rate": 1.7351909391600682e-05,
+      "loss": 2.83,
+      "step": 1322
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 17.131059646606445,
+      "learning_rate": 1.7349904780996293e-05,
+      "loss": 2.6729,
+      "step": 1323
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 16.65376853942871,
+      "learning_rate": 1.7347900170391903e-05,
+      "loss": 2.9689,
+      "step": 1324
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 13.16373062133789,
+      "learning_rate": 1.7345895559787513e-05,
+      "loss": 2.5196,
+      "step": 1325
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 15.137256622314453,
+      "learning_rate": 1.7343890949183123e-05,
+      "loss": 3.2273,
+      "step": 1326
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 12.256842613220215,
+      "learning_rate": 1.7341886338578733e-05,
+      "loss": 3.0272,
+      "step": 1327
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 46.50150680541992,
+      "learning_rate": 1.7339881727974343e-05,
+      "loss": 3.4704,
+      "step": 1328
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 23.8785400390625,
+      "learning_rate": 1.7337877117369953e-05,
+      "loss": 3.1262,
+      "step": 1329
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 17.71267318725586,
+      "learning_rate": 1.7335872506765563e-05,
+      "loss": 2.1541,
+      "step": 1330
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 22.225250244140625,
+      "learning_rate": 1.7333867896161173e-05,
+      "loss": 1.5919,
+      "step": 1331
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 16.465585708618164,
+      "learning_rate": 1.7331863285556783e-05,
+      "loss": 2.3168,
+      "step": 1332
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 24.607051849365234,
+      "learning_rate": 1.732985867495239e-05,
+      "loss": 3.1243,
+      "step": 1333
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 19.95473861694336,
+      "learning_rate": 1.7327854064348003e-05,
+      "loss": 2.5079,
+      "step": 1334
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 19.784517288208008,
+      "learning_rate": 1.7325849453743614e-05,
+      "loss": 3.3156,
+      "step": 1335
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 13.695775032043457,
+      "learning_rate": 1.732384484313922e-05,
+      "loss": 3.1165,
+      "step": 1336
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 11.951885223388672,
+      "learning_rate": 1.7321840232534834e-05,
+      "loss": 2.9321,
+      "step": 1337
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 13.507436752319336,
+      "learning_rate": 1.731983562193044e-05,
+      "loss": 2.4538,
+      "step": 1338
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 11.957839965820312,
+      "learning_rate": 1.731783101132605e-05,
+      "loss": 3.1245,
+      "step": 1339
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 14.829548835754395,
+      "learning_rate": 1.731582640072166e-05,
+      "loss": 2.3632,
+      "step": 1340
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 14.587291717529297,
+      "learning_rate": 1.731382179011727e-05,
+      "loss": 2.8577,
+      "step": 1341
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 13.556631088256836,
+      "learning_rate": 1.731181717951288e-05,
+      "loss": 2.0475,
+      "step": 1342
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 13.010260581970215,
+      "learning_rate": 1.730981256890849e-05,
+      "loss": 2.6603,
+      "step": 1343
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 11.09408187866211,
+      "learning_rate": 1.73078079583041e-05,
+      "loss": 2.5755,
+      "step": 1344
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 13.849974632263184,
+      "learning_rate": 1.730580334769971e-05,
+      "loss": 2.8472,
+      "step": 1345
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 21.211544036865234,
+      "learning_rate": 1.730379873709532e-05,
+      "loss": 2.7465,
+      "step": 1346
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 15.509568214416504,
+      "learning_rate": 1.730179412649093e-05,
+      "loss": 3.4552,
+      "step": 1347
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 13.904120445251465,
+      "learning_rate": 1.729978951588654e-05,
+      "loss": 2.9963,
+      "step": 1348
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 17.5559024810791,
+      "learning_rate": 1.729778490528215e-05,
+      "loss": 2.4842,
+      "step": 1349
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 15.476587295532227,
+      "learning_rate": 1.729578029467776e-05,
+      "loss": 2.5928,
+      "step": 1350
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 15.32288932800293,
+      "learning_rate": 1.729377568407337e-05,
+      "loss": 3.4517,
+      "step": 1351
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 15.49992561340332,
+      "learning_rate": 1.7291771073468978e-05,
+      "loss": 3.042,
+      "step": 1352
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 13.955615043640137,
+      "learning_rate": 1.728976646286459e-05,
+      "loss": 2.5861,
+      "step": 1353
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 12.550683975219727,
+      "learning_rate": 1.7287761852260202e-05,
+      "loss": 3.6951,
+      "step": 1354
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 22.899044036865234,
+      "learning_rate": 1.728575724165581e-05,
+      "loss": 2.9966,
+      "step": 1355
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 12.468682289123535,
+      "learning_rate": 1.7283752631051422e-05,
+      "loss": 2.1763,
+      "step": 1356
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 22.829370498657227,
+      "learning_rate": 1.728174802044703e-05,
+      "loss": 3.2563,
+      "step": 1357
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 27.643447875976562,
+      "learning_rate": 1.727974340984264e-05,
+      "loss": 2.7474,
+      "step": 1358
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 17.19843101501465,
+      "learning_rate": 1.727773879923825e-05,
+      "loss": 2.9442,
+      "step": 1359
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 13.734451293945312,
+      "learning_rate": 1.727573418863386e-05,
+      "loss": 2.8518,
+      "step": 1360
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 11.024998664855957,
+      "learning_rate": 1.727372957802947e-05,
+      "loss": 2.2833,
+      "step": 1361
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 19.57155990600586,
+      "learning_rate": 1.727172496742508e-05,
+      "loss": 2.9668,
+      "step": 1362
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 24.14830207824707,
+      "learning_rate": 1.726972035682069e-05,
+      "loss": 2.2259,
+      "step": 1363
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 13.887985229492188,
+      "learning_rate": 1.72677157462163e-05,
+      "loss": 2.4764,
+      "step": 1364
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 11.10314655303955,
+      "learning_rate": 1.726571113561191e-05,
+      "loss": 2.1367,
+      "step": 1365
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 11.575247764587402,
+      "learning_rate": 1.726370652500752e-05,
+      "loss": 2.396,
+      "step": 1366
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 17.140932083129883,
+      "learning_rate": 1.726170191440313e-05,
+      "loss": 3.0233,
+      "step": 1367
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 18.503704071044922,
+      "learning_rate": 1.725969730379874e-05,
+      "loss": 3.2255,
+      "step": 1368
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 14.695094108581543,
+      "learning_rate": 1.7257692693194346e-05,
+      "loss": 3.1332,
+      "step": 1369
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 12.079935073852539,
+      "learning_rate": 1.725568808258996e-05,
+      "loss": 2.4825,
+      "step": 1370
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 23.080753326416016,
+      "learning_rate": 1.7253683471985566e-05,
+      "loss": 2.7521,
+      "step": 1371
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 13.837876319885254,
+      "learning_rate": 1.7251678861381176e-05,
+      "loss": 2.5238,
+      "step": 1372
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 19.977251052856445,
+      "learning_rate": 1.724967425077679e-05,
+      "loss": 3.2955,
+      "step": 1373
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 13.569607734680176,
+      "learning_rate": 1.7247669640172397e-05,
+      "loss": 2.7767,
+      "step": 1374
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 14.404850959777832,
+      "learning_rate": 1.7245665029568007e-05,
+      "loss": 2.6134,
+      "step": 1375
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 16.575212478637695,
+      "learning_rate": 1.7243660418963617e-05,
+      "loss": 1.7531,
+      "step": 1376
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 21.60881233215332,
+      "learning_rate": 1.7241655808359227e-05,
+      "loss": 2.9619,
+      "step": 1377
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 17.051956176757812,
+      "learning_rate": 1.7239651197754837e-05,
+      "loss": 3.0801,
+      "step": 1378
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 37.58320999145508,
+      "learning_rate": 1.7237646587150447e-05,
+      "loss": 2.3241,
+      "step": 1379
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 12.96102237701416,
+      "learning_rate": 1.7235641976546057e-05,
+      "loss": 2.2837,
+      "step": 1380
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 34.093780517578125,
+      "learning_rate": 1.7233637365941667e-05,
+      "loss": 2.8085,
+      "step": 1381
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 17.705703735351562,
+      "learning_rate": 1.7231632755337277e-05,
+      "loss": 3.4047,
+      "step": 1382
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 16.184062957763672,
+      "learning_rate": 1.7229628144732887e-05,
+      "loss": 3.4847,
+      "step": 1383
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 12.74220085144043,
+      "learning_rate": 1.7227623534128497e-05,
+      "loss": 1.8471,
+      "step": 1384
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 26.621505737304688,
+      "learning_rate": 1.7225618923524108e-05,
+      "loss": 3.4508,
+      "step": 1385
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 23.22762107849121,
+      "learning_rate": 1.7223614312919718e-05,
+      "loss": 2.8567,
+      "step": 1386
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 13.235843658447266,
+      "learning_rate": 1.7221609702315328e-05,
+      "loss": 2.368,
+      "step": 1387
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 10.793066024780273,
+      "learning_rate": 1.7219605091710934e-05,
+      "loss": 1.9442,
+      "step": 1388
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 18.45147705078125,
+      "learning_rate": 1.7217600481106548e-05,
+      "loss": 3.3398,
+      "step": 1389
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 14.921577453613281,
+      "learning_rate": 1.7215595870502155e-05,
+      "loss": 2.1594,
+      "step": 1390
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 14.961502075195312,
+      "learning_rate": 1.7213591259897765e-05,
+      "loss": 3.3974,
+      "step": 1391
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 19.07111167907715,
+      "learning_rate": 1.7211586649293378e-05,
+      "loss": 3.1362,
+      "step": 1392
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 10.949179649353027,
+      "learning_rate": 1.7209582038688985e-05,
+      "loss": 2.0993,
+      "step": 1393
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 12.562166213989258,
+      "learning_rate": 1.7207577428084595e-05,
+      "loss": 2.0554,
+      "step": 1394
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 11.35824966430664,
+      "learning_rate": 1.7205572817480205e-05,
+      "loss": 2.1912,
+      "step": 1395
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 16.764873504638672,
+      "learning_rate": 1.7203568206875815e-05,
+      "loss": 2.5602,
+      "step": 1396
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 14.467211723327637,
+      "learning_rate": 1.7201563596271425e-05,
+      "loss": 2.6329,
+      "step": 1397
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 11.931891441345215,
+      "learning_rate": 1.7199558985667035e-05,
+      "loss": 2.2821,
+      "step": 1398
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 22.51726722717285,
+      "learning_rate": 1.7197554375062645e-05,
+      "loss": 2.7927,
+      "step": 1399
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 10.18954086303711,
+      "learning_rate": 1.7195549764458255e-05,
+      "loss": 1.9007,
+      "step": 1400
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 15.76696491241455,
+      "learning_rate": 1.7193545153853866e-05,
+      "loss": 2.0431,
+      "step": 1401
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 11.40339183807373,
+      "learning_rate": 1.7191540543249476e-05,
+      "loss": 2.5742,
+      "step": 1402
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 12.021610260009766,
+      "learning_rate": 1.7189535932645086e-05,
+      "loss": 3.6196,
+      "step": 1403
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 19.796781539916992,
+      "learning_rate": 1.7187531322040696e-05,
+      "loss": 2.8469,
+      "step": 1404
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 14.170371055603027,
+      "learning_rate": 1.7185526711436306e-05,
+      "loss": 2.3962,
+      "step": 1405
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 11.088910102844238,
+      "learning_rate": 1.7183522100831916e-05,
+      "loss": 2.6827,
+      "step": 1406
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 136.58908081054688,
+      "learning_rate": 1.7181517490227523e-05,
+      "loss": 2.3091,
+      "step": 1407
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 12.442429542541504,
+      "learning_rate": 1.7179512879623136e-05,
+      "loss": 2.4494,
+      "step": 1408
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 14.868009567260742,
+      "learning_rate": 1.7177508269018746e-05,
+      "loss": 2.5188,
+      "step": 1409
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 12.164691925048828,
+      "learning_rate": 1.7175503658414353e-05,
+      "loss": 2.378,
+      "step": 1410
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 15.50973129272461,
+      "learning_rate": 1.7173499047809966e-05,
+      "loss": 4.2229,
+      "step": 1411
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 16.49388885498047,
+      "learning_rate": 1.7171494437205573e-05,
+      "loss": 2.5579,
+      "step": 1412
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 14.063419342041016,
+      "learning_rate": 1.7169489826601183e-05,
+      "loss": 3.0148,
+      "step": 1413
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 37.99775314331055,
+      "learning_rate": 1.7167485215996793e-05,
+      "loss": 2.9612,
+      "step": 1414
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 18.0521183013916,
+      "learning_rate": 1.7165480605392403e-05,
+      "loss": 2.871,
+      "step": 1415
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 17.04564094543457,
+      "learning_rate": 1.7163475994788013e-05,
+      "loss": 2.6688,
+      "step": 1416
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 13.493517875671387,
+      "learning_rate": 1.7161471384183624e-05,
+      "loss": 2.1012,
+      "step": 1417
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 13.597625732421875,
+      "learning_rate": 1.7159466773579234e-05,
+      "loss": 3.3529,
+      "step": 1418
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 17.114131927490234,
+      "learning_rate": 1.7157462162974844e-05,
+      "loss": 3.2373,
+      "step": 1419
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 18.538379669189453,
+      "learning_rate": 1.7155457552370454e-05,
+      "loss": 2.4973,
+      "step": 1420
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 23.284835815429688,
+      "learning_rate": 1.7153452941766064e-05,
+      "loss": 3.6339,
+      "step": 1421
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 19.382251739501953,
+      "learning_rate": 1.7151448331161674e-05,
+      "loss": 3.1989,
+      "step": 1422
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 17.688274383544922,
+      "learning_rate": 1.7149443720557284e-05,
+      "loss": 2.1218,
+      "step": 1423
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 24.966703414916992,
+      "learning_rate": 1.7147439109952894e-05,
+      "loss": 2.9954,
+      "step": 1424
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 31.188905715942383,
+      "learning_rate": 1.7145434499348504e-05,
+      "loss": 2.9035,
+      "step": 1425
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 13.262977600097656,
+      "learning_rate": 1.714342988874411e-05,
+      "loss": 1.3303,
+      "step": 1426
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 13.84818172454834,
+      "learning_rate": 1.7141425278139724e-05,
+      "loss": 2.3282,
+      "step": 1427
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 55.73940658569336,
+      "learning_rate": 1.7139420667535334e-05,
+      "loss": 3.5851,
+      "step": 1428
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 11.50109577178955,
+      "learning_rate": 1.713741605693094e-05,
+      "loss": 2.6434,
+      "step": 1429
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 17.894203186035156,
+      "learning_rate": 1.7135411446326555e-05,
+      "loss": 2.5986,
+      "step": 1430
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 33.426055908203125,
+      "learning_rate": 1.713340683572216e-05,
+      "loss": 2.7803,
+      "step": 1431
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 10.518821716308594,
+      "learning_rate": 1.713140222511777e-05,
+      "loss": 2.1774,
+      "step": 1432
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 13.691694259643555,
+      "learning_rate": 1.712939761451338e-05,
+      "loss": 2.4883,
+      "step": 1433
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 19.365102767944336,
+      "learning_rate": 1.712739300390899e-05,
+      "loss": 2.6138,
+      "step": 1434
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 31.23158836364746,
+      "learning_rate": 1.71253883933046e-05,
+      "loss": 2.905,
+      "step": 1435
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 23.67169761657715,
+      "learning_rate": 1.7123383782700212e-05,
+      "loss": 3.4515,
+      "step": 1436
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 11.991448402404785,
+      "learning_rate": 1.7121379172095822e-05,
+      "loss": 2.562,
+      "step": 1437
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 34.79066467285156,
+      "learning_rate": 1.7119374561491432e-05,
+      "loss": 3.1665,
+      "step": 1438
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 24.111236572265625,
+      "learning_rate": 1.7117369950887042e-05,
+      "loss": 3.2707,
+      "step": 1439
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 36.61920928955078,
+      "learning_rate": 1.7115365340282652e-05,
+      "loss": 2.3113,
+      "step": 1440
+    },
+    {
+      "epoch": 0.43,
+      "eval_loss": 0.4355357885360718,
+      "eval_runtime": 43.4342,
+      "eval_samples_per_second": 34.051,
+      "eval_steps_per_second": 34.051,
+      "step": 1440
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 25.45790672302246,
+      "learning_rate": 1.7113360729678262e-05,
+      "loss": 2.8492,
+      "step": 1441
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 8.575899124145508,
+      "learning_rate": 1.7111356119073872e-05,
+      "loss": 2.934,
+      "step": 1442
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 14.703217506408691,
+      "learning_rate": 1.710935150846948e-05,
+      "loss": 2.8837,
+      "step": 1443
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 45.12884521484375,
+      "learning_rate": 1.7107346897865092e-05,
+      "loss": 4.0134,
+      "step": 1444
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 17.04209327697754,
+      "learning_rate": 1.71053422872607e-05,
+      "loss": 2.9208,
+      "step": 1445
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 11.436484336853027,
+      "learning_rate": 1.710333767665631e-05,
+      "loss": 2.5288,
+      "step": 1446
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 11.852707862854004,
+      "learning_rate": 1.7101333066051923e-05,
+      "loss": 2.4064,
+      "step": 1447
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 18.86704444885254,
+      "learning_rate": 1.709932845544753e-05,
+      "loss": 2.924,
+      "step": 1448
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 26.129491806030273,
+      "learning_rate": 1.7097323844843143e-05,
+      "loss": 2.4179,
+      "step": 1449
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 13.224143028259277,
+      "learning_rate": 1.709531923423875e-05,
+      "loss": 3.0095,
+      "step": 1450
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 12.908326148986816,
+      "learning_rate": 1.709331462363436e-05,
+      "loss": 2.6045,
+      "step": 1451
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 17.967126846313477,
+      "learning_rate": 1.7091310013029973e-05,
+      "loss": 1.9252,
+      "step": 1452
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 15.611589431762695,
+      "learning_rate": 1.708930540242558e-05,
+      "loss": 3.118,
+      "step": 1453
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 19.359302520751953,
+      "learning_rate": 1.708730079182119e-05,
+      "loss": 2.8996,
+      "step": 1454
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 15.188103675842285,
+      "learning_rate": 1.70852961812168e-05,
+      "loss": 3.3204,
+      "step": 1455
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 24.560422897338867,
+      "learning_rate": 1.708329157061241e-05,
+      "loss": 2.7835,
+      "step": 1456
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 13.218994140625,
+      "learning_rate": 1.708128696000802e-05,
+      "loss": 2.3175,
+      "step": 1457
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 13.680419921875,
+      "learning_rate": 1.707928234940363e-05,
+      "loss": 2.2045,
+      "step": 1458
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 15.42464828491211,
+      "learning_rate": 1.707727773879924e-05,
+      "loss": 2.7285,
+      "step": 1459
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 25.709156036376953,
+      "learning_rate": 1.707527312819485e-05,
+      "loss": 3.6841,
+      "step": 1460
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 15.961677551269531,
+      "learning_rate": 1.707326851759046e-05,
+      "loss": 3.4276,
+      "step": 1461
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 15.101017951965332,
+      "learning_rate": 1.7071263906986067e-05,
+      "loss": 2.935,
+      "step": 1462
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 15.92037582397461,
+      "learning_rate": 1.706925929638168e-05,
+      "loss": 2.3067,
+      "step": 1463
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 17.06378173828125,
+      "learning_rate": 1.7067254685777287e-05,
+      "loss": 2.2178,
+      "step": 1464
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 15.061570167541504,
+      "learning_rate": 1.7065250075172897e-05,
+      "loss": 2.2647,
+      "step": 1465
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 11.610605239868164,
+      "learning_rate": 1.706324546456851e-05,
+      "loss": 2.2437,
+      "step": 1466
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 9.191241264343262,
+      "learning_rate": 1.7061240853964118e-05,
+      "loss": 2.011,
+      "step": 1467
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 29.908700942993164,
+      "learning_rate": 1.7059236243359728e-05,
+      "loss": 3.4803,
+      "step": 1468
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 31.88568115234375,
+      "learning_rate": 1.7057231632755338e-05,
+      "loss": 1.611,
+      "step": 1469
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 18.888980865478516,
+      "learning_rate": 1.7055227022150948e-05,
+      "loss": 3.0609,
+      "step": 1470
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 15.416746139526367,
+      "learning_rate": 1.7053222411546558e-05,
+      "loss": 2.4115,
+      "step": 1471
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 16.257118225097656,
+      "learning_rate": 1.7051217800942168e-05,
+      "loss": 3.6617,
+      "step": 1472
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 14.70289134979248,
+      "learning_rate": 1.7049213190337778e-05,
+      "loss": 2.0395,
+      "step": 1473
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 14.966999053955078,
+      "learning_rate": 1.7047208579733388e-05,
+      "loss": 2.7337,
+      "step": 1474
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 19.118864059448242,
+      "learning_rate": 1.7045203969128998e-05,
+      "loss": 2.5427,
+      "step": 1475
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 21.44363021850586,
+      "learning_rate": 1.704319935852461e-05,
+      "loss": 3.0899,
+      "step": 1476
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 16.222488403320312,
+      "learning_rate": 1.704119474792022e-05,
+      "loss": 2.3837,
+      "step": 1477
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 13.13603401184082,
+      "learning_rate": 1.703919013731583e-05,
+      "loss": 2.4675,
+      "step": 1478
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 20.3343563079834,
+      "learning_rate": 1.703718552671144e-05,
+      "loss": 2.9424,
+      "step": 1479
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 19.545503616333008,
+      "learning_rate": 1.703518091610705e-05,
+      "loss": 2.9953,
+      "step": 1480
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 18.053640365600586,
+      "learning_rate": 1.7033176305502655e-05,
+      "loss": 2.11,
+      "step": 1481
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 18.4859561920166,
+      "learning_rate": 1.703117169489827e-05,
+      "loss": 3.0607,
+      "step": 1482
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 13.820666313171387,
+      "learning_rate": 1.702916708429388e-05,
+      "loss": 2.8445,
+      "step": 1483
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 15.42257022857666,
+      "learning_rate": 1.7027162473689486e-05,
+      "loss": 2.8965,
+      "step": 1484
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 27.479591369628906,
+      "learning_rate": 1.70251578630851e-05,
+      "loss": 2.9632,
+      "step": 1485
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 21.065073013305664,
+      "learning_rate": 1.7023153252480706e-05,
+      "loss": 2.7442,
+      "step": 1486
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 7.838624477386475,
+      "learning_rate": 1.7021148641876316e-05,
+      "loss": 2.0983,
+      "step": 1487
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 17.929758071899414,
+      "learning_rate": 1.7019144031271926e-05,
+      "loss": 2.3848,
+      "step": 1488
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 10.071520805358887,
+      "learning_rate": 1.7017139420667536e-05,
+      "loss": 1.5338,
+      "step": 1489
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 15.892718315124512,
+      "learning_rate": 1.7015134810063146e-05,
+      "loss": 2.5901,
+      "step": 1490
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 13.212607383728027,
+      "learning_rate": 1.7013130199458756e-05,
+      "loss": 3.0064,
+      "step": 1491
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 21.533849716186523,
+      "learning_rate": 1.7011125588854366e-05,
+      "loss": 2.9632,
+      "step": 1492
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 18.14595603942871,
+      "learning_rate": 1.7009120978249976e-05,
+      "loss": 1.7862,
+      "step": 1493
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 12.642379760742188,
+      "learning_rate": 1.7007116367645586e-05,
+      "loss": 2.7063,
+      "step": 1494
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 18.708799362182617,
+      "learning_rate": 1.7005111757041197e-05,
+      "loss": 2.7772,
+      "step": 1495
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 29.240921020507812,
+      "learning_rate": 1.7003107146436807e-05,
+      "loss": 1.7695,
+      "step": 1496
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 18.173198699951172,
+      "learning_rate": 1.7001102535832417e-05,
+      "loss": 2.528,
+      "step": 1497
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 10.611312866210938,
+      "learning_rate": 1.6999097925228027e-05,
+      "loss": 2.3301,
+      "step": 1498
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 17.193771362304688,
+      "learning_rate": 1.6997093314623637e-05,
+      "loss": 1.8752,
+      "step": 1499
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 17.25593376159668,
+      "learning_rate": 1.6995088704019244e-05,
+      "loss": 3.6294,
+      "step": 1500
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 18.063823699951172,
+      "learning_rate": 1.6993084093414857e-05,
+      "loss": 3.0548,
+      "step": 1501
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 22.38393783569336,
+      "learning_rate": 1.6991079482810467e-05,
+      "loss": 2.6282,
+      "step": 1502
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 11.480286598205566,
+      "learning_rate": 1.6989074872206074e-05,
+      "loss": 1.8335,
+      "step": 1503
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 26.893537521362305,
+      "learning_rate": 1.6987070261601687e-05,
+      "loss": 2.7152,
+      "step": 1504
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 11.621118545532227,
+      "learning_rate": 1.6985065650997294e-05,
+      "loss": 2.3662,
+      "step": 1505
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 15.238231658935547,
+      "learning_rate": 1.6983061040392904e-05,
+      "loss": 2.4623,
+      "step": 1506
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 20.844465255737305,
+      "learning_rate": 1.6981056429788514e-05,
+      "loss": 2.5909,
+      "step": 1507
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 11.749519348144531,
+      "learning_rate": 1.6979051819184124e-05,
+      "loss": 1.8395,
+      "step": 1508
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 15.4244384765625,
+      "learning_rate": 1.6977047208579734e-05,
+      "loss": 3.7336,
+      "step": 1509
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 11.57870864868164,
+      "learning_rate": 1.6975042597975344e-05,
+      "loss": 2.3881,
+      "step": 1510
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 21.28474235534668,
+      "learning_rate": 1.6973037987370954e-05,
+      "loss": 2.8733,
+      "step": 1511
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 13.479619026184082,
+      "learning_rate": 1.6971033376766565e-05,
+      "loss": 1.693,
+      "step": 1512
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 17.223281860351562,
+      "learning_rate": 1.6969028766162175e-05,
+      "loss": 2.2971,
+      "step": 1513
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 31.056154251098633,
+      "learning_rate": 1.6967024155557785e-05,
+      "loss": 2.5655,
+      "step": 1514
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 60.43313980102539,
+      "learning_rate": 1.6965019544953395e-05,
+      "loss": 3.5182,
+      "step": 1515
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 15.120966911315918,
+      "learning_rate": 1.6963014934349005e-05,
+      "loss": 2.7201,
+      "step": 1516
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 19.221208572387695,
+      "learning_rate": 1.6961010323744615e-05,
+      "loss": 3.3837,
+      "step": 1517
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 13.70907974243164,
+      "learning_rate": 1.6959005713140225e-05,
+      "loss": 1.8185,
+      "step": 1518
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 25.266483306884766,
+      "learning_rate": 1.6957001102535832e-05,
+      "loss": 3.0797,
+      "step": 1519
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 16.34473419189453,
+      "learning_rate": 1.6954996491931445e-05,
+      "loss": 2.689,
+      "step": 1520
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 21.672119140625,
+      "learning_rate": 1.6952991881327055e-05,
+      "loss": 2.4117,
+      "step": 1521
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 12.567007064819336,
+      "learning_rate": 1.6950987270722662e-05,
+      "loss": 2.5169,
+      "step": 1522
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 54.59273147583008,
+      "learning_rate": 1.6948982660118275e-05,
+      "loss": 3.4834,
+      "step": 1523
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 18.03861427307129,
+      "learning_rate": 1.6946978049513882e-05,
+      "loss": 2.824,
+      "step": 1524
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 40.20027542114258,
+      "learning_rate": 1.6944973438909492e-05,
+      "loss": 2.5567,
+      "step": 1525
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 12.565523147583008,
+      "learning_rate": 1.6942968828305106e-05,
+      "loss": 3.1967,
+      "step": 1526
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 12.132319450378418,
+      "learning_rate": 1.6940964217700712e-05,
+      "loss": 2.5713,
+      "step": 1527
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 28.399211883544922,
+      "learning_rate": 1.6938959607096323e-05,
+      "loss": 2.7361,
+      "step": 1528
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 20.132675170898438,
+      "learning_rate": 1.6936954996491933e-05,
+      "loss": 2.6361,
+      "step": 1529
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 19.281675338745117,
+      "learning_rate": 1.6934950385887543e-05,
+      "loss": 2.8946,
+      "step": 1530
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 11.40804672241211,
+      "learning_rate": 1.6932945775283153e-05,
+      "loss": 2.4466,
+      "step": 1531
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 19.21536636352539,
+      "learning_rate": 1.6930941164678763e-05,
+      "loss": 2.9093,
+      "step": 1532
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 13.203646659851074,
+      "learning_rate": 1.6928936554074373e-05,
+      "loss": 2.3911,
+      "step": 1533
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 11.949692726135254,
+      "learning_rate": 1.6926931943469983e-05,
+      "loss": 2.8228,
+      "step": 1534
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 14.0568208694458,
+      "learning_rate": 1.6924927332865593e-05,
+      "loss": 2.7099,
+      "step": 1535
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 55.63922882080078,
+      "learning_rate": 1.69229227222612e-05,
+      "loss": 2.8945,
+      "step": 1536
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 9.444130897521973,
+      "learning_rate": 1.6920918111656813e-05,
+      "loss": 1.714,
+      "step": 1537
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 12.79077434539795,
+      "learning_rate": 1.6918913501052423e-05,
+      "loss": 3.4578,
+      "step": 1538
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 12.505064010620117,
+      "learning_rate": 1.691690889044803e-05,
+      "loss": 1.7876,
+      "step": 1539
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 19.741012573242188,
+      "learning_rate": 1.6914904279843644e-05,
+      "loss": 2.2345,
+      "step": 1540
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 26.9569149017334,
+      "learning_rate": 1.691289966923925e-05,
+      "loss": 2.3699,
+      "step": 1541
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 21.60605812072754,
+      "learning_rate": 1.691089505863486e-05,
+      "loss": 3.1069,
+      "step": 1542
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 14.085561752319336,
+      "learning_rate": 1.690889044803047e-05,
+      "loss": 2.5667,
+      "step": 1543
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 19.00534439086914,
+      "learning_rate": 1.690688583742608e-05,
+      "loss": 2.816,
+      "step": 1544
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 14.58003044128418,
+      "learning_rate": 1.690488122682169e-05,
+      "loss": 3.1388,
+      "step": 1545
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 23.24366569519043,
+      "learning_rate": 1.69028766162173e-05,
+      "loss": 1.9925,
+      "step": 1546
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 18.700740814208984,
+      "learning_rate": 1.690087200561291e-05,
+      "loss": 1.9935,
+      "step": 1547
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 20.79172134399414,
+      "learning_rate": 1.689886739500852e-05,
+      "loss": 3.1285,
+      "step": 1548
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 22.56122589111328,
+      "learning_rate": 1.689686278440413e-05,
+      "loss": 3.1273,
+      "step": 1549
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 12.774810791015625,
+      "learning_rate": 1.689485817379974e-05,
+      "loss": 2.8262,
+      "step": 1550
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 16.529850006103516,
+      "learning_rate": 1.689285356319535e-05,
+      "loss": 2.3473,
+      "step": 1551
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 20.23556137084961,
+      "learning_rate": 1.689084895259096e-05,
+      "loss": 1.8937,
+      "step": 1552
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 23.965484619140625,
+      "learning_rate": 1.688884434198657e-05,
+      "loss": 2.5148,
+      "step": 1553
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 11.394997596740723,
+      "learning_rate": 1.688683973138218e-05,
+      "loss": 2.7078,
+      "step": 1554
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 14.20688247680664,
+      "learning_rate": 1.6884835120777788e-05,
+      "loss": 1.9273,
+      "step": 1555
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 8.693448066711426,
+      "learning_rate": 1.68828305101734e-05,
+      "loss": 1.7773,
+      "step": 1556
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 14.1410493850708,
+      "learning_rate": 1.688082589956901e-05,
+      "loss": 2.5604,
+      "step": 1557
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 29.748794555664062,
+      "learning_rate": 1.6878821288964618e-05,
+      "loss": 2.5162,
+      "step": 1558
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 10.367297172546387,
+      "learning_rate": 1.6876816678360232e-05,
+      "loss": 3.0466,
+      "step": 1559
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 16.686874389648438,
+      "learning_rate": 1.687481206775584e-05,
+      "loss": 3.304,
+      "step": 1560
+    },
+    {
+      "epoch": 0.47,
+      "eval_loss": 0.3973664343357086,
+      "eval_runtime": 43.4098,
+      "eval_samples_per_second": 34.071,
+      "eval_steps_per_second": 34.071,
+      "step": 1560
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 14.685885429382324,
+      "learning_rate": 1.687280745715145e-05,
+      "loss": 2.3338,
+      "step": 1561
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 20.013853073120117,
+      "learning_rate": 1.687080284654706e-05,
+      "loss": 2.3406,
+      "step": 1562
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 14.868475914001465,
+      "learning_rate": 1.686879823594267e-05,
+      "loss": 1.8712,
+      "step": 1563
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 20.89446449279785,
+      "learning_rate": 1.686679362533828e-05,
+      "loss": 2.7493,
+      "step": 1564
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 34.56494903564453,
+      "learning_rate": 1.686478901473389e-05,
+      "loss": 2.8977,
+      "step": 1565
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 14.068147659301758,
+      "learning_rate": 1.68627844041295e-05,
+      "loss": 2.5127,
+      "step": 1566
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 18.625877380371094,
+      "learning_rate": 1.686077979352511e-05,
+      "loss": 2.8985,
+      "step": 1567
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 11.287114143371582,
+      "learning_rate": 1.685877518292072e-05,
+      "loss": 1.6863,
+      "step": 1568
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 19.42438316345215,
+      "learning_rate": 1.685677057231633e-05,
+      "loss": 2.5056,
+      "step": 1569
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 13.997237205505371,
+      "learning_rate": 1.685476596171194e-05,
+      "loss": 2.6668,
+      "step": 1570
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 10.12904167175293,
+      "learning_rate": 1.685276135110755e-05,
+      "loss": 2.0919,
+      "step": 1571
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 23.484573364257812,
+      "learning_rate": 1.685075674050316e-05,
+      "loss": 2.9543,
+      "step": 1572
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 21.227386474609375,
+      "learning_rate": 1.684875212989877e-05,
+      "loss": 2.0336,
+      "step": 1573
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 30.29522132873535,
+      "learning_rate": 1.6846747519294376e-05,
+      "loss": 2.6453,
+      "step": 1574
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 25.412912368774414,
+      "learning_rate": 1.684474290868999e-05,
+      "loss": 3.1794,
+      "step": 1575
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 26.188493728637695,
+      "learning_rate": 1.68427382980856e-05,
+      "loss": 2.9667,
+      "step": 1576
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 15.816750526428223,
+      "learning_rate": 1.6840733687481206e-05,
+      "loss": 2.0286,
+      "step": 1577
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 17.47620964050293,
+      "learning_rate": 1.683872907687682e-05,
+      "loss": 3.0095,
+      "step": 1578
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 14.732656478881836,
+      "learning_rate": 1.6836724466272427e-05,
+      "loss": 3.2241,
+      "step": 1579
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 15.800749778747559,
+      "learning_rate": 1.6834719855668037e-05,
+      "loss": 2.4581,
+      "step": 1580
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 14.257530212402344,
+      "learning_rate": 1.683271524506365e-05,
+      "loss": 2.6594,
+      "step": 1581
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 18.350427627563477,
+      "learning_rate": 1.6830710634459257e-05,
+      "loss": 2.8644,
+      "step": 1582
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 17.27747344970703,
+      "learning_rate": 1.6828706023854867e-05,
+      "loss": 2.8629,
+      "step": 1583
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 13.865232467651367,
+      "learning_rate": 1.6826701413250477e-05,
+      "loss": 2.115,
+      "step": 1584
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 26.224536895751953,
+      "learning_rate": 1.6824696802646087e-05,
+      "loss": 3.4931,
+      "step": 1585
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 10.468141555786133,
+      "learning_rate": 1.6822692192041697e-05,
+      "loss": 1.9319,
+      "step": 1586
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 10.993379592895508,
+      "learning_rate": 1.6820687581437307e-05,
+      "loss": 2.8885,
+      "step": 1587
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 18.79608726501465,
+      "learning_rate": 1.6818682970832917e-05,
+      "loss": 3.4616,
+      "step": 1588
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 23.80633544921875,
+      "learning_rate": 1.6816678360228528e-05,
+      "loss": 3.5015,
+      "step": 1589
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 13.126861572265625,
+      "learning_rate": 1.6814673749624138e-05,
+      "loss": 2.2426,
+      "step": 1590
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 18.61846923828125,
+      "learning_rate": 1.6812669139019748e-05,
+      "loss": 2.8252,
+      "step": 1591
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 9.052312850952148,
+      "learning_rate": 1.6810664528415358e-05,
+      "loss": 2.1735,
+      "step": 1592
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 9.982820510864258,
+      "learning_rate": 1.6808659917810964e-05,
+      "loss": 1.8825,
+      "step": 1593
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 18.984390258789062,
+      "learning_rate": 1.6806655307206578e-05,
+      "loss": 3.0772,
+      "step": 1594
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 19.849212646484375,
+      "learning_rate": 1.6804650696602188e-05,
+      "loss": 2.8981,
+      "step": 1595
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 15.359475135803223,
+      "learning_rate": 1.6802646085997795e-05,
+      "loss": 2.7745,
+      "step": 1596
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 13.44256591796875,
+      "learning_rate": 1.6800641475393408e-05,
+      "loss": 2.4322,
+      "step": 1597
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 19.221418380737305,
+      "learning_rate": 1.6798636864789015e-05,
+      "loss": 1.8762,
+      "step": 1598
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 20.303268432617188,
+      "learning_rate": 1.6796632254184625e-05,
+      "loss": 1.7788,
+      "step": 1599
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 19.06624412536621,
+      "learning_rate": 1.679462764358024e-05,
+      "loss": 3.7968,
+      "step": 1600
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 10.663858413696289,
+      "learning_rate": 1.6792623032975845e-05,
+      "loss": 1.1838,
+      "step": 1601
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 16.027814865112305,
+      "learning_rate": 1.6790618422371455e-05,
+      "loss": 2.7883,
+      "step": 1602
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 15.941658020019531,
+      "learning_rate": 1.6788613811767065e-05,
+      "loss": 2.9455,
+      "step": 1603
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 16.636554718017578,
+      "learning_rate": 1.6786609201162675e-05,
+      "loss": 2.8941,
+      "step": 1604
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 11.56485366821289,
+      "learning_rate": 1.6784604590558285e-05,
+      "loss": 2.961,
+      "step": 1605
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 13.037918090820312,
+      "learning_rate": 1.6782599979953896e-05,
+      "loss": 2.292,
+      "step": 1606
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 16.472829818725586,
+      "learning_rate": 1.6780595369349506e-05,
+      "loss": 3.5172,
+      "step": 1607
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 21.085201263427734,
+      "learning_rate": 1.6778590758745116e-05,
+      "loss": 2.8983,
+      "step": 1608
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 43.585323333740234,
+      "learning_rate": 1.6776586148140726e-05,
+      "loss": 2.7387,
+      "step": 1609
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 12.80986213684082,
+      "learning_rate": 1.6774581537536332e-05,
+      "loss": 2.1192,
+      "step": 1610
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 24.946006774902344,
+      "learning_rate": 1.6772576926931946e-05,
+      "loss": 3.2683,
+      "step": 1611
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 10.049246788024902,
+      "learning_rate": 1.6770572316327556e-05,
+      "loss": 2.5447,
+      "step": 1612
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 13.763422012329102,
+      "learning_rate": 1.6768567705723163e-05,
+      "loss": 2.1387,
+      "step": 1613
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 18.731260299682617,
+      "learning_rate": 1.6766563095118776e-05,
+      "loss": 1.9977,
+      "step": 1614
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 11.16000747680664,
+      "learning_rate": 1.6764558484514383e-05,
+      "loss": 1.9304,
+      "step": 1615
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 15.650193214416504,
+      "learning_rate": 1.6762553873909993e-05,
+      "loss": 3.0744,
+      "step": 1616
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 19.635541915893555,
+      "learning_rate": 1.6760549263305603e-05,
+      "loss": 2.1045,
+      "step": 1617
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 17.894126892089844,
+      "learning_rate": 1.6758544652701213e-05,
+      "loss": 2.6789,
+      "step": 1618
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 20.0896053314209,
+      "learning_rate": 1.6756540042096823e-05,
+      "loss": 2.6357,
+      "step": 1619
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 18.415935516357422,
+      "learning_rate": 1.6754535431492433e-05,
+      "loss": 2.9713,
+      "step": 1620
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 22.75372314453125,
+      "learning_rate": 1.6752530820888043e-05,
+      "loss": 3.3434,
+      "step": 1621
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 20.71872901916504,
+      "learning_rate": 1.6750526210283654e-05,
+      "loss": 2.466,
+      "step": 1622
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 15.37236499786377,
+      "learning_rate": 1.6748521599679264e-05,
+      "loss": 3.1419,
+      "step": 1623
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 16.39763641357422,
+      "learning_rate": 1.6746516989074874e-05,
+      "loss": 2.6307,
+      "step": 1624
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 12.221854209899902,
+      "learning_rate": 1.6744512378470484e-05,
+      "loss": 3.121,
+      "step": 1625
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 12.343154907226562,
+      "learning_rate": 1.6742507767866094e-05,
+      "loss": 2.0893,
+      "step": 1626
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 15.398106575012207,
+      "learning_rate": 1.6740503157261704e-05,
+      "loss": 2.126,
+      "step": 1627
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 16.43120574951172,
+      "learning_rate": 1.6738498546657314e-05,
+      "loss": 3.3734,
+      "step": 1628
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 11.764989852905273,
+      "learning_rate": 1.673649393605292e-05,
+      "loss": 1.5352,
+      "step": 1629
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 24.872587203979492,
+      "learning_rate": 1.6734489325448534e-05,
+      "loss": 2.9974,
+      "step": 1630
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 16.981983184814453,
+      "learning_rate": 1.6732484714844144e-05,
+      "loss": 2.1175,
+      "step": 1631
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 20.261165618896484,
+      "learning_rate": 1.673048010423975e-05,
+      "loss": 3.5279,
+      "step": 1632
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 26.34723663330078,
+      "learning_rate": 1.6728475493635364e-05,
+      "loss": 3.3491,
+      "step": 1633
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 18.659168243408203,
+      "learning_rate": 1.672647088303097e-05,
+      "loss": 2.9663,
+      "step": 1634
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 11.186785697937012,
+      "learning_rate": 1.672446627242658e-05,
+      "loss": 2.1397,
+      "step": 1635
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 9.327524185180664,
+      "learning_rate": 1.672246166182219e-05,
+      "loss": 1.744,
+      "step": 1636
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 17.69675064086914,
+      "learning_rate": 1.67204570512178e-05,
+      "loss": 2.9865,
+      "step": 1637
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 21.239151000976562,
+      "learning_rate": 1.671845244061341e-05,
+      "loss": 2.8212,
+      "step": 1638
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 41.4120979309082,
+      "learning_rate": 1.671644783000902e-05,
+      "loss": 2.7785,
+      "step": 1639
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 11.790454864501953,
+      "learning_rate": 1.671444321940463e-05,
+      "loss": 2.2797,
+      "step": 1640
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 17.779918670654297,
+      "learning_rate": 1.6712438608800242e-05,
+      "loss": 3.0112,
+      "step": 1641
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 16.839773178100586,
+      "learning_rate": 1.6710433998195852e-05,
+      "loss": 2.4351,
+      "step": 1642
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 15.281671524047852,
+      "learning_rate": 1.6708429387591462e-05,
+      "loss": 1.9823,
+      "step": 1643
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 19.457735061645508,
+      "learning_rate": 1.6706424776987072e-05,
+      "loss": 3.6049,
+      "step": 1644
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 16.624021530151367,
+      "learning_rate": 1.6704420166382682e-05,
+      "loss": 3.3101,
+      "step": 1645
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 14.08040714263916,
+      "learning_rate": 1.6702415555778292e-05,
+      "loss": 2.4127,
+      "step": 1646
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 23.419265747070312,
+      "learning_rate": 1.6700410945173902e-05,
+      "loss": 3.5622,
+      "step": 1647
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 12.655960083007812,
+      "learning_rate": 1.669840633456951e-05,
+      "loss": 1.7092,
+      "step": 1648
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 10.199254989624023,
+      "learning_rate": 1.6696401723965122e-05,
+      "loss": 1.9251,
+      "step": 1649
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 30.36039161682129,
+      "learning_rate": 1.6694397113360732e-05,
+      "loss": 1.9794,
+      "step": 1650
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 20.610151290893555,
+      "learning_rate": 1.669239250275634e-05,
+      "loss": 2.0625,
+      "step": 1651
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 11.706425666809082,
+      "learning_rate": 1.6690387892151953e-05,
+      "loss": 1.8749,
+      "step": 1652
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 9.867399215698242,
+      "learning_rate": 1.668838328154756e-05,
+      "loss": 2.4526,
+      "step": 1653
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 21.617414474487305,
+      "learning_rate": 1.668637867094317e-05,
+      "loss": 4.5465,
+      "step": 1654
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 21.315481185913086,
+      "learning_rate": 1.6684374060338783e-05,
+      "loss": 2.3748,
+      "step": 1655
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 19.910005569458008,
+      "learning_rate": 1.668236944973439e-05,
+      "loss": 2.4007,
+      "step": 1656
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 11.719648361206055,
+      "learning_rate": 1.668036483913e-05,
+      "loss": 2.5453,
+      "step": 1657
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 14.366843223571777,
+      "learning_rate": 1.667836022852561e-05,
+      "loss": 2.7823,
+      "step": 1658
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 20.43113136291504,
+      "learning_rate": 1.667635561792122e-05,
+      "loss": 3.0117,
+      "step": 1659
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 25.725727081298828,
+      "learning_rate": 1.667435100731683e-05,
+      "loss": 1.9852,
+      "step": 1660
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 12.890719413757324,
+      "learning_rate": 1.667234639671244e-05,
+      "loss": 2.8881,
+      "step": 1661
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 15.555781364440918,
+      "learning_rate": 1.667034178610805e-05,
+      "loss": 3.1708,
+      "step": 1662
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 13.397451400756836,
+      "learning_rate": 1.666833717550366e-05,
+      "loss": 2.0334,
+      "step": 1663
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 12.852067947387695,
+      "learning_rate": 1.666633256489927e-05,
+      "loss": 3.193,
+      "step": 1664
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 17.795087814331055,
+      "learning_rate": 1.666432795429488e-05,
+      "loss": 2.1889,
+      "step": 1665
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 10.84477710723877,
+      "learning_rate": 1.666232334369049e-05,
+      "loss": 2.1249,
+      "step": 1666
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 13.947750091552734,
+      "learning_rate": 1.66603187330861e-05,
+      "loss": 2.1903,
+      "step": 1667
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 15.641617774963379,
+      "learning_rate": 1.665831412248171e-05,
+      "loss": 1.4892,
+      "step": 1668
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 17.841720581054688,
+      "learning_rate": 1.665630951187732e-05,
+      "loss": 2.2156,
+      "step": 1669
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 21.81488800048828,
+      "learning_rate": 1.6654304901272927e-05,
+      "loss": 2.7059,
+      "step": 1670
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 10.093061447143555,
+      "learning_rate": 1.665230029066854e-05,
+      "loss": 2.087,
+      "step": 1671
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 12.38084602355957,
+      "learning_rate": 1.6650295680064148e-05,
+      "loss": 2.6622,
+      "step": 1672
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 13.34916877746582,
+      "learning_rate": 1.6648291069459758e-05,
+      "loss": 3.5172,
+      "step": 1673
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 14.111736297607422,
+      "learning_rate": 1.664628645885537e-05,
+      "loss": 2.5334,
+      "step": 1674
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 19.57823944091797,
+      "learning_rate": 1.6644281848250978e-05,
+      "loss": 2.8923,
+      "step": 1675
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 20.907072067260742,
+      "learning_rate": 1.6642277237646588e-05,
+      "loss": 2.7302,
+      "step": 1676
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 12.42391300201416,
+      "learning_rate": 1.6640272627042198e-05,
+      "loss": 2.4441,
+      "step": 1677
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 39.446815490722656,
+      "learning_rate": 1.6638268016437808e-05,
+      "loss": 3.3596,
+      "step": 1678
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 10.514243125915527,
+      "learning_rate": 1.6636263405833418e-05,
+      "loss": 1.9339,
+      "step": 1679
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 22.982282638549805,
+      "learning_rate": 1.6634258795229028e-05,
+      "loss": 2.6203,
+      "step": 1680
+    },
+    {
+      "epoch": 0.51,
+      "eval_loss": 0.421953409910202,
+      "eval_runtime": 43.6652,
+      "eval_samples_per_second": 33.871,
+      "eval_steps_per_second": 33.871,
+      "step": 1680
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 31.43738555908203,
+      "learning_rate": 1.663225418462464e-05,
+      "loss": 2.5882,
+      "step": 1681
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 10.706387519836426,
+      "learning_rate": 1.663024957402025e-05,
+      "loss": 2.1867,
+      "step": 1682
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 15.15072250366211,
+      "learning_rate": 1.662824496341586e-05,
+      "loss": 2.9213,
+      "step": 1683
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 9.624457359313965,
+      "learning_rate": 1.6626240352811465e-05,
+      "loss": 1.8701,
+      "step": 1684
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 9.49343490600586,
+      "learning_rate": 1.662423574220708e-05,
+      "loss": 1.612,
+      "step": 1685
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 28.716997146606445,
+      "learning_rate": 1.662223113160269e-05,
+      "loss": 3.3154,
+      "step": 1686
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 9.385660171508789,
+      "learning_rate": 1.66202265209983e-05,
+      "loss": 2.5589,
+      "step": 1687
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 14.978555679321289,
+      "learning_rate": 1.661822191039391e-05,
+      "loss": 2.1595,
+      "step": 1688
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 18.335002899169922,
+      "learning_rate": 1.6616217299789516e-05,
+      "loss": 2.4189,
+      "step": 1689
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 16.582286834716797,
+      "learning_rate": 1.661421268918513e-05,
+      "loss": 3.3232,
+      "step": 1690
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 13.659370422363281,
+      "learning_rate": 1.6612208078580736e-05,
+      "loss": 2.2705,
+      "step": 1691
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 13.425541877746582,
+      "learning_rate": 1.6610203467976346e-05,
+      "loss": 2.4391,
+      "step": 1692
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 22.200546264648438,
+      "learning_rate": 1.660819885737196e-05,
+      "loss": 2.4271,
+      "step": 1693
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 14.445297241210938,
+      "learning_rate": 1.6606194246767566e-05,
+      "loss": 2.3692,
+      "step": 1694
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 14.760397911071777,
+      "learning_rate": 1.6604189636163176e-05,
+      "loss": 2.3689,
+      "step": 1695
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 10.618377685546875,
+      "learning_rate": 1.6602185025558786e-05,
+      "loss": 2.1075,
+      "step": 1696
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 17.689512252807617,
+      "learning_rate": 1.6600180414954396e-05,
+      "loss": 3.0902,
+      "step": 1697
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 8.896835327148438,
+      "learning_rate": 1.6598175804350006e-05,
+      "loss": 2.3301,
+      "step": 1698
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 9.717402458190918,
+      "learning_rate": 1.6596171193745616e-05,
+      "loss": 1.9009,
+      "step": 1699
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 16.815967559814453,
+      "learning_rate": 1.6594166583141227e-05,
+      "loss": 3.0338,
+      "step": 1700
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 33.1158447265625,
+      "learning_rate": 1.6592161972536837e-05,
+      "loss": 3.0146,
+      "step": 1701
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 12.088573455810547,
+      "learning_rate": 1.6590157361932447e-05,
+      "loss": 3.0398,
+      "step": 1702
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 24.571584701538086,
+      "learning_rate": 1.6588152751328053e-05,
+      "loss": 3.3241,
+      "step": 1703
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 23.808401107788086,
+      "learning_rate": 1.6586148140723667e-05,
+      "loss": 3.3574,
+      "step": 1704
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 20.748384475708008,
+      "learning_rate": 1.6584143530119277e-05,
+      "loss": 3.0009,
+      "step": 1705
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 24.722551345825195,
+      "learning_rate": 1.6582138919514884e-05,
+      "loss": 2.8073,
+      "step": 1706
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 16.103742599487305,
+      "learning_rate": 1.6580134308910497e-05,
+      "loss": 3.2797,
+      "step": 1707
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 17.112245559692383,
+      "learning_rate": 1.6578129698306104e-05,
+      "loss": 2.8844,
+      "step": 1708
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 12.628554344177246,
+      "learning_rate": 1.6576125087701714e-05,
+      "loss": 2.5837,
+      "step": 1709
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 15.136996269226074,
+      "learning_rate": 1.6574120477097327e-05,
+      "loss": 2.8925,
+      "step": 1710
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 15.211194038391113,
+      "learning_rate": 1.6572115866492934e-05,
+      "loss": 2.2617,
+      "step": 1711
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 13.138132095336914,
+      "learning_rate": 1.6570111255888544e-05,
+      "loss": 2.1384,
+      "step": 1712
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 15.303442001342773,
+      "learning_rate": 1.6568106645284154e-05,
+      "loss": 3.7071,
+      "step": 1713
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 16.974143981933594,
+      "learning_rate": 1.6566102034679764e-05,
+      "loss": 2.0731,
+      "step": 1714
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 22.324352264404297,
+      "learning_rate": 1.6564097424075374e-05,
+      "loss": 2.5875,
+      "step": 1715
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 17.300701141357422,
+      "learning_rate": 1.6562092813470984e-05,
+      "loss": 2.9271,
+      "step": 1716
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 20.104694366455078,
+      "learning_rate": 1.6560088202866595e-05,
+      "loss": 3.3379,
+      "step": 1717
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 11.9177827835083,
+      "learning_rate": 1.6558083592262205e-05,
+      "loss": 2.2438,
+      "step": 1718
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 12.473027229309082,
+      "learning_rate": 1.6556078981657815e-05,
+      "loss": 2.8623,
+      "step": 1719
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 16.957218170166016,
+      "learning_rate": 1.6554074371053425e-05,
+      "loss": 2.7224,
+      "step": 1720
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 9.168384552001953,
+      "learning_rate": 1.6552069760449035e-05,
+      "loss": 1.4801,
+      "step": 1721
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 14.333914756774902,
+      "learning_rate": 1.655006514984464e-05,
+      "loss": 1.834,
+      "step": 1722
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 8.387497901916504,
+      "learning_rate": 1.6548060539240255e-05,
+      "loss": 1.4313,
+      "step": 1723
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 10.79047679901123,
+      "learning_rate": 1.6546055928635865e-05,
+      "loss": 3.2971,
+      "step": 1724
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 10.988299369812012,
+      "learning_rate": 1.6544051318031472e-05,
+      "loss": 2.5413,
+      "step": 1725
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 16.327035903930664,
+      "learning_rate": 1.6542046707427085e-05,
+      "loss": 2.1365,
+      "step": 1726
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 10.375070571899414,
+      "learning_rate": 1.6540042096822692e-05,
+      "loss": 2.7831,
+      "step": 1727
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 11.78280258178711,
+      "learning_rate": 1.6538037486218302e-05,
+      "loss": 2.6977,
+      "step": 1728
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 25.220386505126953,
+      "learning_rate": 1.6536032875613916e-05,
+      "loss": 2.6352,
+      "step": 1729
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 38.015220642089844,
+      "learning_rate": 1.6534028265009522e-05,
+      "loss": 2.9953,
+      "step": 1730
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 16.3960018157959,
+      "learning_rate": 1.6532023654405132e-05,
+      "loss": 2.1179,
+      "step": 1731
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 17.71184730529785,
+      "learning_rate": 1.6530019043800742e-05,
+      "loss": 2.3667,
+      "step": 1732
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 11.10054874420166,
+      "learning_rate": 1.6528014433196353e-05,
+      "loss": 1.6067,
+      "step": 1733
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 19.382043838500977,
+      "learning_rate": 1.6526009822591963e-05,
+      "loss": 3.0277,
+      "step": 1734
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 64.88998413085938,
+      "learning_rate": 1.6524005211987573e-05,
+      "loss": 1.9251,
+      "step": 1735
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 16.99798583984375,
+      "learning_rate": 1.6522000601383183e-05,
+      "loss": 2.2818,
+      "step": 1736
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 11.27112102508545,
+      "learning_rate": 1.6519995990778793e-05,
+      "loss": 2.4658,
+      "step": 1737
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 16.193252563476562,
+      "learning_rate": 1.6517991380174403e-05,
+      "loss": 2.023,
+      "step": 1738
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 11.178709983825684,
+      "learning_rate": 1.6515986769570013e-05,
+      "loss": 2.0991,
+      "step": 1739
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 12.193035125732422,
+      "learning_rate": 1.6513982158965623e-05,
+      "loss": 2.4835,
+      "step": 1740
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 11.212671279907227,
+      "learning_rate": 1.6511977548361233e-05,
+      "loss": 1.8206,
+      "step": 1741
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 7.858899116516113,
+      "learning_rate": 1.6509972937756843e-05,
+      "loss": 2.2607,
+      "step": 1742
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 34.4583740234375,
+      "learning_rate": 1.6507968327152453e-05,
+      "loss": 2.8326,
+      "step": 1743
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 15.727729797363281,
+      "learning_rate": 1.650596371654806e-05,
+      "loss": 3.1574,
+      "step": 1744
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 25.254718780517578,
+      "learning_rate": 1.6503959105943674e-05,
+      "loss": 3.1741,
+      "step": 1745
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 17.325634002685547,
+      "learning_rate": 1.650195449533928e-05,
+      "loss": 2.9572,
+      "step": 1746
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 23.073179244995117,
+      "learning_rate": 1.649994988473489e-05,
+      "loss": 1.7773,
+      "step": 1747
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 16.520126342773438,
+      "learning_rate": 1.6497945274130504e-05,
+      "loss": 3.4431,
+      "step": 1748
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 18.78542709350586,
+      "learning_rate": 1.649594066352611e-05,
+      "loss": 2.6952,
+      "step": 1749
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 19.45966911315918,
+      "learning_rate": 1.649393605292172e-05,
+      "loss": 2.005,
+      "step": 1750
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 11.350322723388672,
+      "learning_rate": 1.649193144231733e-05,
+      "loss": 1.6326,
+      "step": 1751
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 18.4759521484375,
+      "learning_rate": 1.648992683171294e-05,
+      "loss": 2.4786,
+      "step": 1752
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 15.755661010742188,
+      "learning_rate": 1.648792222110855e-05,
+      "loss": 1.7824,
+      "step": 1753
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 46.813907623291016,
+      "learning_rate": 1.648591761050416e-05,
+      "loss": 3.5372,
+      "step": 1754
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 11.6588134765625,
+      "learning_rate": 1.648391299989977e-05,
+      "loss": 1.7987,
+      "step": 1755
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 15.711779594421387,
+      "learning_rate": 1.648190838929538e-05,
+      "loss": 2.7252,
+      "step": 1756
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 9.573955535888672,
+      "learning_rate": 1.647990377869099e-05,
+      "loss": 2.4631,
+      "step": 1757
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 9.92371940612793,
+      "learning_rate": 1.64778991680866e-05,
+      "loss": 2.0796,
+      "step": 1758
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 26.95187759399414,
+      "learning_rate": 1.647589455748221e-05,
+      "loss": 2.8893,
+      "step": 1759
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 9.273636817932129,
+      "learning_rate": 1.647388994687782e-05,
+      "loss": 2.0131,
+      "step": 1760
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 19.220373153686523,
+      "learning_rate": 1.647188533627343e-05,
+      "loss": 3.017,
+      "step": 1761
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 34.35186767578125,
+      "learning_rate": 1.646988072566904e-05,
+      "loss": 3.3928,
+      "step": 1762
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 36.41823959350586,
+      "learning_rate": 1.6467876115064648e-05,
+      "loss": 3.6049,
+      "step": 1763
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 14.849743843078613,
+      "learning_rate": 1.6465871504460262e-05,
+      "loss": 1.9525,
+      "step": 1764
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 20.26209259033203,
+      "learning_rate": 1.646386689385587e-05,
+      "loss": 2.2319,
+      "step": 1765
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 24.784208297729492,
+      "learning_rate": 1.646186228325148e-05,
+      "loss": 2.5943,
+      "step": 1766
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 8.97102165222168,
+      "learning_rate": 1.6459857672647092e-05,
+      "loss": 2.0995,
+      "step": 1767
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 12.025717735290527,
+      "learning_rate": 1.64578530620427e-05,
+      "loss": 1.3348,
+      "step": 1768
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 22.216552734375,
+      "learning_rate": 1.645584845143831e-05,
+      "loss": 2.9476,
+      "step": 1769
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 16.27435874938965,
+      "learning_rate": 1.645384384083392e-05,
+      "loss": 2.9931,
+      "step": 1770
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 23.192441940307617,
+      "learning_rate": 1.645183923022953e-05,
+      "loss": 3.2747,
+      "step": 1771
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 26.984561920166016,
+      "learning_rate": 1.644983461962514e-05,
+      "loss": 3.2546,
+      "step": 1772
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 16.29502296447754,
+      "learning_rate": 1.644783000902075e-05,
+      "loss": 2.9974,
+      "step": 1773
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 15.848845481872559,
+      "learning_rate": 1.644582539841636e-05,
+      "loss": 3.1271,
+      "step": 1774
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 34.28373336791992,
+      "learning_rate": 1.644382078781197e-05,
+      "loss": 3.1274,
+      "step": 1775
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 18.812297821044922,
+      "learning_rate": 1.644181617720758e-05,
+      "loss": 2.0213,
+      "step": 1776
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 23.698867797851562,
+      "learning_rate": 1.6439811566603186e-05,
+      "loss": 3.1921,
+      "step": 1777
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 8.073270797729492,
+      "learning_rate": 1.64378069559988e-05,
+      "loss": 2.1989,
+      "step": 1778
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 18.90115737915039,
+      "learning_rate": 1.643580234539441e-05,
+      "loss": 3.2046,
+      "step": 1779
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 12.170257568359375,
+      "learning_rate": 1.6433797734790016e-05,
+      "loss": 3.0221,
+      "step": 1780
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 19.89399528503418,
+      "learning_rate": 1.643179312418563e-05,
+      "loss": 3.4413,
+      "step": 1781
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 14.685297012329102,
+      "learning_rate": 1.6429788513581237e-05,
+      "loss": 2.2774,
+      "step": 1782
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 29.041440963745117,
+      "learning_rate": 1.6427783902976847e-05,
+      "loss": 2.4153,
+      "step": 1783
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 13.65660285949707,
+      "learning_rate": 1.642577929237246e-05,
+      "loss": 2.6716,
+      "step": 1784
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 11.705273628234863,
+      "learning_rate": 1.6423774681768067e-05,
+      "loss": 1.8001,
+      "step": 1785
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 25.235740661621094,
+      "learning_rate": 1.6421770071163677e-05,
+      "loss": 3.1693,
+      "step": 1786
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 15.884099960327148,
+      "learning_rate": 1.6419765460559287e-05,
+      "loss": 2.4285,
+      "step": 1787
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 52.04803466796875,
+      "learning_rate": 1.6417760849954897e-05,
+      "loss": 4.0988,
+      "step": 1788
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 42.220130920410156,
+      "learning_rate": 1.6415756239350507e-05,
+      "loss": 2.3157,
+      "step": 1789
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 19.528057098388672,
+      "learning_rate": 1.6413751628746117e-05,
+      "loss": 2.153,
+      "step": 1790
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 10.022256851196289,
+      "learning_rate": 1.6411747018141727e-05,
+      "loss": 2.6832,
+      "step": 1791
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 14.05746841430664,
+      "learning_rate": 1.6409742407537337e-05,
+      "loss": 2.6188,
+      "step": 1792
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 17.109386444091797,
+      "learning_rate": 1.6407737796932947e-05,
+      "loss": 2.4536,
+      "step": 1793
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 14.443997383117676,
+      "learning_rate": 1.6405733186328558e-05,
+      "loss": 2.4826,
+      "step": 1794
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 29.712779998779297,
+      "learning_rate": 1.6403728575724168e-05,
+      "loss": 3.0167,
+      "step": 1795
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 19.298267364501953,
+      "learning_rate": 1.6401723965119778e-05,
+      "loss": 2.9807,
+      "step": 1796
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 15.033485412597656,
+      "learning_rate": 1.6399719354515388e-05,
+      "loss": 2.6656,
+      "step": 1797
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 16.12212371826172,
+      "learning_rate": 1.6397714743910998e-05,
+      "loss": 2.3691,
+      "step": 1798
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 26.953134536743164,
+      "learning_rate": 1.6395710133306605e-05,
+      "loss": 3.2797,
+      "step": 1799
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 9.630372047424316,
+      "learning_rate": 1.6393705522702218e-05,
+      "loss": 1.7721,
+      "step": 1800
+    },
+    {
+      "epoch": 0.54,
+      "eval_loss": 0.3416885435581207,
+      "eval_runtime": 43.4913,
+      "eval_samples_per_second": 34.007,
+      "eval_steps_per_second": 34.007,
+      "step": 1800
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 18.797924041748047,
+      "learning_rate": 1.6391700912097825e-05,
+      "loss": 1.5968,
+      "step": 1801
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 31.31778907775879,
+      "learning_rate": 1.6389696301493435e-05,
+      "loss": 2.9072,
+      "step": 1802
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 14.117547035217285,
+      "learning_rate": 1.6387691690889048e-05,
+      "loss": 2.4949,
+      "step": 1803
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 12.579538345336914,
+      "learning_rate": 1.6385687080284655e-05,
+      "loss": 1.8951,
+      "step": 1804
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 23.75373077392578,
+      "learning_rate": 1.6383682469680265e-05,
+      "loss": 1.9489,
+      "step": 1805
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 13.660398483276367,
+      "learning_rate": 1.6381677859075875e-05,
+      "loss": 1.6119,
+      "step": 1806
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 60.78506851196289,
+      "learning_rate": 1.6379673248471485e-05,
+      "loss": 2.7651,
+      "step": 1807
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 14.931727409362793,
+      "learning_rate": 1.6377668637867095e-05,
+      "loss": 2.2028,
+      "step": 1808
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 10.574440956115723,
+      "learning_rate": 1.6375664027262705e-05,
+      "loss": 2.6472,
+      "step": 1809
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 14.71934986114502,
+      "learning_rate": 1.6373659416658315e-05,
+      "loss": 2.3538,
+      "step": 1810
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 16.00834846496582,
+      "learning_rate": 1.6371654806053926e-05,
+      "loss": 2.8576,
+      "step": 1811
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 15.810260772705078,
+      "learning_rate": 1.6369650195449536e-05,
+      "loss": 2.9982,
+      "step": 1812
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 29.83978843688965,
+      "learning_rate": 1.6367645584845146e-05,
+      "loss": 3.0221,
+      "step": 1813
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 16.944915771484375,
+      "learning_rate": 1.6365640974240756e-05,
+      "loss": 2.8306,
+      "step": 1814
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 22.454910278320312,
+      "learning_rate": 1.6363636363636366e-05,
+      "loss": 2.3444,
+      "step": 1815
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 16.41376495361328,
+      "learning_rate": 1.6361631753031976e-05,
+      "loss": 1.8047,
+      "step": 1816
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4576.21875,
+      "learning_rate": 1.6359627142427586e-05,
+      "loss": 2.2836,
+      "step": 1817
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 15.997746467590332,
+      "learning_rate": 1.6357622531823193e-05,
+      "loss": 2.5703,
+      "step": 1818
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 15.037249565124512,
+      "learning_rate": 1.6355617921218806e-05,
+      "loss": 3.2597,
+      "step": 1819
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 13.758872032165527,
+      "learning_rate": 1.6353613310614413e-05,
+      "loss": 3.4616,
+      "step": 1820
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 10.099397659301758,
+      "learning_rate": 1.6351608700010023e-05,
+      "loss": 1.96,
+      "step": 1821
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 32.01512908935547,
+      "learning_rate": 1.6349604089405636e-05,
+      "loss": 4.0462,
+      "step": 1822
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 17.568777084350586,
+      "learning_rate": 1.6347599478801243e-05,
+      "loss": 2.5459,
+      "step": 1823
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 20.525028228759766,
+      "learning_rate": 1.6345594868196853e-05,
+      "loss": 2.6286,
+      "step": 1824
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 18.882827758789062,
+      "learning_rate": 1.6343590257592463e-05,
+      "loss": 2.5098,
+      "step": 1825
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 10.823426246643066,
+      "learning_rate": 1.6341585646988073e-05,
+      "loss": 1.6721,
+      "step": 1826
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 14.602689743041992,
+      "learning_rate": 1.6339581036383684e-05,
+      "loss": 2.6786,
+      "step": 1827
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 17.44394302368164,
+      "learning_rate": 1.6337576425779294e-05,
+      "loss": 2.5762,
+      "step": 1828
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 19.712026596069336,
+      "learning_rate": 1.6335571815174904e-05,
+      "loss": 2.6143,
+      "step": 1829
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 25.916528701782227,
+      "learning_rate": 1.6333567204570514e-05,
+      "loss": 2.5159,
+      "step": 1830
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 17.429548263549805,
+      "learning_rate": 1.6331562593966124e-05,
+      "loss": 3.5722,
+      "step": 1831
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 14.735082626342773,
+      "learning_rate": 1.6329557983361734e-05,
+      "loss": 2.0687,
+      "step": 1832
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 19.132709503173828,
+      "learning_rate": 1.6327553372757344e-05,
+      "loss": 2.9178,
+      "step": 1833
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 20.842416763305664,
+      "learning_rate": 1.6325548762152954e-05,
+      "loss": 3.3049,
+      "step": 1834
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 14.198040008544922,
+      "learning_rate": 1.6323544151548564e-05,
+      "loss": 2.8631,
+      "step": 1835
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 12.003619194030762,
+      "learning_rate": 1.6321539540944174e-05,
+      "loss": 2.0466,
+      "step": 1836
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 14.629396438598633,
+      "learning_rate": 1.631953493033978e-05,
+      "loss": 2.967,
+      "step": 1837
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 15.829573631286621,
+      "learning_rate": 1.6317530319735394e-05,
+      "loss": 2.1286,
+      "step": 1838
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 14.251692771911621,
+      "learning_rate": 1.6315525709131e-05,
+      "loss": 2.4073,
+      "step": 1839
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 25.430139541625977,
+      "learning_rate": 1.631352109852661e-05,
+      "loss": 2.9292,
+      "step": 1840
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 27.490047454833984,
+      "learning_rate": 1.6311516487922225e-05,
+      "loss": 3.4921,
+      "step": 1841
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 9.101457595825195,
+      "learning_rate": 1.630951187731783e-05,
+      "loss": 2.1141,
+      "step": 1842
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 12.929884910583496,
+      "learning_rate": 1.630750726671344e-05,
+      "loss": 1.934,
+      "step": 1843
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 11.764657020568848,
+      "learning_rate": 1.630550265610905e-05,
+      "loss": 2.0135,
+      "step": 1844
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 15.838483810424805,
+      "learning_rate": 1.630349804550466e-05,
+      "loss": 2.7645,
+      "step": 1845
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 19.013673782348633,
+      "learning_rate": 1.6301493434900272e-05,
+      "loss": 2.2205,
+      "step": 1846
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 12.440937042236328,
+      "learning_rate": 1.6299488824295882e-05,
+      "loss": 3.1805,
+      "step": 1847
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 18.47015380859375,
+      "learning_rate": 1.6297484213691492e-05,
+      "loss": 2.9228,
+      "step": 1848
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 26.46478843688965,
+      "learning_rate": 1.6295479603087102e-05,
+      "loss": 3.1306,
+      "step": 1849
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 20.930150985717773,
+      "learning_rate": 1.6293474992482712e-05,
+      "loss": 2.2428,
+      "step": 1850
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 13.267045021057129,
+      "learning_rate": 1.629147038187832e-05,
+      "loss": 2.6697,
+      "step": 1851
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 11.723921775817871,
+      "learning_rate": 1.6289465771273932e-05,
+      "loss": 2.8602,
+      "step": 1852
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 11.42064094543457,
+      "learning_rate": 1.6287461160669542e-05,
+      "loss": 2.9313,
+      "step": 1853
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 31.033870697021484,
+      "learning_rate": 1.628545655006515e-05,
+      "loss": 3.2394,
+      "step": 1854
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 17.858312606811523,
+      "learning_rate": 1.6283451939460762e-05,
+      "loss": 2.1223,
+      "step": 1855
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 13.080681800842285,
+      "learning_rate": 1.628144732885637e-05,
+      "loss": 2.4824,
+      "step": 1856
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 9.161316871643066,
+      "learning_rate": 1.6279442718251983e-05,
+      "loss": 1.5993,
+      "step": 1857
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 15.82925796508789,
+      "learning_rate": 1.6277438107647593e-05,
+      "loss": 2.2007,
+      "step": 1858
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 29.120468139648438,
+      "learning_rate": 1.62754334970432e-05,
+      "loss": 3.2814,
+      "step": 1859
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 20.672487258911133,
+      "learning_rate": 1.6273428886438813e-05,
+      "loss": 2.6652,
+      "step": 1860
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 18.959318161010742,
+      "learning_rate": 1.627142427583442e-05,
+      "loss": 2.5797,
+      "step": 1861
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 10.144336700439453,
+      "learning_rate": 1.626941966523003e-05,
+      "loss": 2.1145,
+      "step": 1862
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 19.466264724731445,
+      "learning_rate": 1.626741505462564e-05,
+      "loss": 2.4139,
+      "step": 1863
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 13.246243476867676,
+      "learning_rate": 1.626541044402125e-05,
+      "loss": 2.2984,
+      "step": 1864
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 15.933266639709473,
+      "learning_rate": 1.626340583341686e-05,
+      "loss": 2.6167,
+      "step": 1865
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 10.605395317077637,
+      "learning_rate": 1.626140122281247e-05,
+      "loss": 2.094,
+      "step": 1866
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 14.293509483337402,
+      "learning_rate": 1.625939661220808e-05,
+      "loss": 2.4193,
+      "step": 1867
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 14.355191230773926,
+      "learning_rate": 1.625739200160369e-05,
+      "loss": 3.3217,
+      "step": 1868
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 11.297273635864258,
+      "learning_rate": 1.62553873909993e-05,
+      "loss": 1.9198,
+      "step": 1869
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 19.611162185668945,
+      "learning_rate": 1.625338278039491e-05,
+      "loss": 2.3439,
+      "step": 1870
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 9.530548095703125,
+      "learning_rate": 1.625137816979052e-05,
+      "loss": 1.1765,
+      "step": 1871
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 20.621294021606445,
+      "learning_rate": 1.624937355918613e-05,
+      "loss": 2.2362,
+      "step": 1872
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 20.514188766479492,
+      "learning_rate": 1.6247368948581737e-05,
+      "loss": 3.2915,
+      "step": 1873
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 11.529969215393066,
+      "learning_rate": 1.624536433797735e-05,
+      "loss": 1.7275,
+      "step": 1874
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 15.720181465148926,
+      "learning_rate": 1.6243359727372957e-05,
+      "loss": 2.0002,
+      "step": 1875
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 20.808303833007812,
+      "learning_rate": 1.6241355116768567e-05,
+      "loss": 3.8721,
+      "step": 1876
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 11.557236671447754,
+      "learning_rate": 1.623935050616418e-05,
+      "loss": 1.6353,
+      "step": 1877
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 13.909945487976074,
+      "learning_rate": 1.6237345895559788e-05,
+      "loss": 2.2762,
+      "step": 1878
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 36.823299407958984,
+      "learning_rate": 1.6235341284955398e-05,
+      "loss": 2.6792,
+      "step": 1879
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 73.27782440185547,
+      "learning_rate": 1.6233336674351008e-05,
+      "loss": 3.6506,
+      "step": 1880
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 15.239794731140137,
+      "learning_rate": 1.6231332063746618e-05,
+      "loss": 2.5237,
+      "step": 1881
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 28.48020362854004,
+      "learning_rate": 1.6229327453142228e-05,
+      "loss": 2.8108,
+      "step": 1882
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 9.497574806213379,
+      "learning_rate": 1.6227322842537838e-05,
+      "loss": 1.4183,
+      "step": 1883
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 12.64692211151123,
+      "learning_rate": 1.6225318231933448e-05,
+      "loss": 2.5578,
+      "step": 1884
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 23.391645431518555,
+      "learning_rate": 1.6223313621329058e-05,
+      "loss": 2.03,
+      "step": 1885
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 25.81462860107422,
+      "learning_rate": 1.622130901072467e-05,
+      "loss": 2.3658,
+      "step": 1886
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 22.7410831451416,
+      "learning_rate": 1.621930440012028e-05,
+      "loss": 2.291,
+      "step": 1887
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 14.532958030700684,
+      "learning_rate": 1.621729978951589e-05,
+      "loss": 2.3839,
+      "step": 1888
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 9.742128372192383,
+      "learning_rate": 1.62152951789115e-05,
+      "loss": 1.9635,
+      "step": 1889
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 11.445870399475098,
+      "learning_rate": 1.621329056830711e-05,
+      "loss": 2.448,
+      "step": 1890
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 17.39885902404785,
+      "learning_rate": 1.621128595770272e-05,
+      "loss": 1.9545,
+      "step": 1891
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 11.735391616821289,
+      "learning_rate": 1.6209281347098325e-05,
+      "loss": 2.49,
+      "step": 1892
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 31.197790145874023,
+      "learning_rate": 1.620727673649394e-05,
+      "loss": 2.4989,
+      "step": 1893
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 28.4355525970459,
+      "learning_rate": 1.6205272125889546e-05,
+      "loss": 2.9879,
+      "step": 1894
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 21.622291564941406,
+      "learning_rate": 1.6203267515285156e-05,
+      "loss": 3.1001,
+      "step": 1895
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 15.58934497833252,
+      "learning_rate": 1.620126290468077e-05,
+      "loss": 2.0499,
+      "step": 1896
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 16.792016983032227,
+      "learning_rate": 1.6199258294076376e-05,
+      "loss": 2.5507,
+      "step": 1897
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 15.837790489196777,
+      "learning_rate": 1.6197253683471986e-05,
+      "loss": 2.5776,
+      "step": 1898
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 13.746726989746094,
+      "learning_rate": 1.6195249072867596e-05,
+      "loss": 1.7229,
+      "step": 1899
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 20.217098236083984,
+      "learning_rate": 1.6193244462263206e-05,
+      "loss": 2.7257,
+      "step": 1900
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 14.193788528442383,
+      "learning_rate": 1.6191239851658816e-05,
+      "loss": 2.4532,
+      "step": 1901
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 12.840964317321777,
+      "learning_rate": 1.6189235241054426e-05,
+      "loss": 1.3137,
+      "step": 1902
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 10.857292175292969,
+      "learning_rate": 1.6187230630450036e-05,
+      "loss": 2.3681,
+      "step": 1903
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 11.725610733032227,
+      "learning_rate": 1.6185226019845646e-05,
+      "loss": 2.0677,
+      "step": 1904
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 16.389385223388672,
+      "learning_rate": 1.6183221409241257e-05,
+      "loss": 2.2652,
+      "step": 1905
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 20.219581604003906,
+      "learning_rate": 1.6181216798636867e-05,
+      "loss": 2.3222,
+      "step": 1906
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 15.775784492492676,
+      "learning_rate": 1.6179212188032477e-05,
+      "loss": 2.0435,
+      "step": 1907
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 11.360732078552246,
+      "learning_rate": 1.6177207577428087e-05,
+      "loss": 2.0024,
+      "step": 1908
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 26.330806732177734,
+      "learning_rate": 1.6175202966823697e-05,
+      "loss": 3.1264,
+      "step": 1909
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 16.440622329711914,
+      "learning_rate": 1.6173198356219307e-05,
+      "loss": 2.6761,
+      "step": 1910
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 14.55276107788086,
+      "learning_rate": 1.6171193745614914e-05,
+      "loss": 2.2345,
+      "step": 1911
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 15.58680534362793,
+      "learning_rate": 1.6169189135010527e-05,
+      "loss": 2.8794,
+      "step": 1912
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 27.979393005371094,
+      "learning_rate": 1.6167184524406137e-05,
+      "loss": 2.5893,
+      "step": 1913
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 14.484015464782715,
+      "learning_rate": 1.6165179913801744e-05,
+      "loss": 2.4981,
+      "step": 1914
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 9.881906509399414,
+      "learning_rate": 1.6163175303197357e-05,
+      "loss": 2.1874,
+      "step": 1915
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 17.09141731262207,
+      "learning_rate": 1.6161170692592964e-05,
+      "loss": 2.9159,
+      "step": 1916
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 18.972490310668945,
+      "learning_rate": 1.6159166081988574e-05,
+      "loss": 1.6856,
+      "step": 1917
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 10.02529525756836,
+      "learning_rate": 1.6157161471384184e-05,
+      "loss": 1.7176,
+      "step": 1918
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 21.166505813598633,
+      "learning_rate": 1.6155156860779794e-05,
+      "loss": 2.5582,
+      "step": 1919
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 34.15064239501953,
+      "learning_rate": 1.6153152250175404e-05,
+      "loss": 4.2157,
+      "step": 1920
+    },
+    {
+      "epoch": 0.58,
+      "eval_loss": 0.339510053396225,
+      "eval_runtime": 43.4353,
+      "eval_samples_per_second": 34.051,
+      "eval_steps_per_second": 34.051,
+      "step": 1920
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 18.826351165771484,
+      "learning_rate": 1.6151147639571015e-05,
+      "loss": 2.8172,
+      "step": 1921
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 28.641447067260742,
+      "learning_rate": 1.6149143028966625e-05,
+      "loss": 3.8235,
+      "step": 1922
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 18.224559783935547,
+      "learning_rate": 1.6147138418362235e-05,
+      "loss": 1.6789,
+      "step": 1923
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 20.033323287963867,
+      "learning_rate": 1.6145133807757845e-05,
+      "loss": 2.2715,
+      "step": 1924
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 18.451032638549805,
+      "learning_rate": 1.6143129197153455e-05,
+      "loss": 3.6021,
+      "step": 1925
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 13.090971946716309,
+      "learning_rate": 1.6141124586549065e-05,
+      "loss": 2.8423,
+      "step": 1926
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 31.57698631286621,
+      "learning_rate": 1.6139119975944675e-05,
+      "loss": 2.1646,
+      "step": 1927
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 19.8615779876709,
+      "learning_rate": 1.6137115365340285e-05,
+      "loss": 2.4651,
+      "step": 1928
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 17.440027236938477,
+      "learning_rate": 1.6135110754735895e-05,
+      "loss": 2.2323,
+      "step": 1929
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 12.359676361083984,
+      "learning_rate": 1.6133106144131502e-05,
+      "loss": 3.0975,
+      "step": 1930
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 7.871943473815918,
+      "learning_rate": 1.6131101533527115e-05,
+      "loss": 1.8966,
+      "step": 1931
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 13.588000297546387,
+      "learning_rate": 1.6129096922922725e-05,
+      "loss": 2.7859,
+      "step": 1932
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 16.730295181274414,
+      "learning_rate": 1.6127092312318332e-05,
+      "loss": 2.4625,
+      "step": 1933
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 13.648134231567383,
+      "learning_rate": 1.6125087701713946e-05,
+      "loss": 2.5004,
+      "step": 1934
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 19.520750045776367,
+      "learning_rate": 1.6123083091109552e-05,
+      "loss": 3.5212,
+      "step": 1935
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 19.73345184326172,
+      "learning_rate": 1.6121078480505162e-05,
+      "loss": 2.2163,
+      "step": 1936
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 11.382865905761719,
+      "learning_rate": 1.6119073869900772e-05,
+      "loss": 1.5394,
+      "step": 1937
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 11.32697868347168,
+      "learning_rate": 1.6117069259296383e-05,
+      "loss": 1.7365,
+      "step": 1938
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 17.160417556762695,
+      "learning_rate": 1.6115064648691993e-05,
+      "loss": 3.0884,
+      "step": 1939
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 10.711620330810547,
+      "learning_rate": 1.6113060038087603e-05,
+      "loss": 2.0598,
+      "step": 1940
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 15.657376289367676,
+      "learning_rate": 1.6111055427483213e-05,
+      "loss": 2.5557,
+      "step": 1941
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 17.509824752807617,
+      "learning_rate": 1.6109050816878823e-05,
+      "loss": 2.2766,
+      "step": 1942
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 10.339923858642578,
+      "learning_rate": 1.6107046206274433e-05,
+      "loss": 1.8058,
+      "step": 1943
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 16.32088851928711,
+      "learning_rate": 1.6105041595670043e-05,
+      "loss": 2.2214,
+      "step": 1944
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 34.30913543701172,
+      "learning_rate": 1.6103036985065653e-05,
+      "loss": 3.2443,
+      "step": 1945
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 13.773527145385742,
+      "learning_rate": 1.6101032374461263e-05,
+      "loss": 3.0382,
+      "step": 1946
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 23.005949020385742,
+      "learning_rate": 1.609902776385687e-05,
+      "loss": 2.0551,
+      "step": 1947
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 32.260467529296875,
+      "learning_rate": 1.6097023153252483e-05,
+      "loss": 2.4883,
+      "step": 1948
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 13.194154739379883,
+      "learning_rate": 1.609501854264809e-05,
+      "loss": 2.7221,
+      "step": 1949
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 29.038406372070312,
+      "learning_rate": 1.60930139320437e-05,
+      "loss": 2.7329,
+      "step": 1950
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 15.281466484069824,
+      "learning_rate": 1.6091009321439314e-05,
+      "loss": 2.5336,
+      "step": 1951
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 14.665143013000488,
+      "learning_rate": 1.608900471083492e-05,
+      "loss": 1.9784,
+      "step": 1952
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 17.575735092163086,
+      "learning_rate": 1.608700010023053e-05,
+      "loss": 1.5022,
+      "step": 1953
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 16.851743698120117,
+      "learning_rate": 1.608499548962614e-05,
+      "loss": 2.1047,
+      "step": 1954
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 17.620439529418945,
+      "learning_rate": 1.608299087902175e-05,
+      "loss": 2.4975,
+      "step": 1955
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 18.211750030517578,
+      "learning_rate": 1.608098626841736e-05,
+      "loss": 2.441,
+      "step": 1956
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 38.04513168334961,
+      "learning_rate": 1.607898165781297e-05,
+      "loss": 3.2777,
+      "step": 1957
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 18.044525146484375,
+      "learning_rate": 1.607697704720858e-05,
+      "loss": 3.4203,
+      "step": 1958
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 14.710494995117188,
+      "learning_rate": 1.607497243660419e-05,
+      "loss": 2.0862,
+      "step": 1959
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 10.158746719360352,
+      "learning_rate": 1.60729678259998e-05,
+      "loss": 1.8249,
+      "step": 1960
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 14.311506271362305,
+      "learning_rate": 1.607096321539541e-05,
+      "loss": 2.0706,
+      "step": 1961
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 9.432246208190918,
+      "learning_rate": 1.606895860479102e-05,
+      "loss": 2.0623,
+      "step": 1962
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 17.04642105102539,
+      "learning_rate": 1.606695399418663e-05,
+      "loss": 2.918,
+      "step": 1963
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 11.562948226928711,
+      "learning_rate": 1.606494938358224e-05,
+      "loss": 2.5815,
+      "step": 1964
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 18.638172149658203,
+      "learning_rate": 1.606294477297785e-05,
+      "loss": 2.995,
+      "step": 1965
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 14.144832611083984,
+      "learning_rate": 1.6060940162373458e-05,
+      "loss": 2.1113,
+      "step": 1966
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 16.268152236938477,
+      "learning_rate": 1.605893555176907e-05,
+      "loss": 2.2524,
+      "step": 1967
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 16.423564910888672,
+      "learning_rate": 1.6056930941164678e-05,
+      "loss": 2.2185,
+      "step": 1968
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 28.028411865234375,
+      "learning_rate": 1.605492633056029e-05,
+      "loss": 2.556,
+      "step": 1969
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 21.24933624267578,
+      "learning_rate": 1.6052921719955902e-05,
+      "loss": 2.3257,
+      "step": 1970
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 15.96957778930664,
+      "learning_rate": 1.605091710935151e-05,
+      "loss": 2.617,
+      "step": 1971
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 10.580945014953613,
+      "learning_rate": 1.604891249874712e-05,
+      "loss": 1.4834,
+      "step": 1972
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 12.444055557250977,
+      "learning_rate": 1.604690788814273e-05,
+      "loss": 2.222,
+      "step": 1973
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 12.718807220458984,
+      "learning_rate": 1.604490327753834e-05,
+      "loss": 1.986,
+      "step": 1974
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 11.073088645935059,
+      "learning_rate": 1.604289866693395e-05,
+      "loss": 2.3385,
+      "step": 1975
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 22.884178161621094,
+      "learning_rate": 1.604089405632956e-05,
+      "loss": 3.075,
+      "step": 1976
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 13.408631324768066,
+      "learning_rate": 1.603888944572517e-05,
+      "loss": 1.867,
+      "step": 1977
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 20.047677993774414,
+      "learning_rate": 1.603688483512078e-05,
+      "loss": 2.3888,
+      "step": 1978
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 56.602203369140625,
+      "learning_rate": 1.603488022451639e-05,
+      "loss": 2.9184,
+      "step": 1979
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 23.749616622924805,
+      "learning_rate": 1.6032875613912e-05,
+      "loss": 2.7749,
+      "step": 1980
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 26.525043487548828,
+      "learning_rate": 1.603087100330761e-05,
+      "loss": 2.1839,
+      "step": 1981
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 32.17291259765625,
+      "learning_rate": 1.602886639270322e-05,
+      "loss": 2.7472,
+      "step": 1982
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 13.829222679138184,
+      "learning_rate": 1.602686178209883e-05,
+      "loss": 2.3262,
+      "step": 1983
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 10.74743366241455,
+      "learning_rate": 1.602485717149444e-05,
+      "loss": 3.0373,
+      "step": 1984
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 19.987829208374023,
+      "learning_rate": 1.6022852560890046e-05,
+      "loss": 2.2646,
+      "step": 1985
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 15.228018760681152,
+      "learning_rate": 1.602084795028566e-05,
+      "loss": 2.8864,
+      "step": 1986
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 39.494258880615234,
+      "learning_rate": 1.601884333968127e-05,
+      "loss": 3.7293,
+      "step": 1987
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 21.58100700378418,
+      "learning_rate": 1.6016838729076877e-05,
+      "loss": 3.238,
+      "step": 1988
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 26.791006088256836,
+      "learning_rate": 1.601483411847249e-05,
+      "loss": 2.7061,
+      "step": 1989
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 26.071443557739258,
+      "learning_rate": 1.6012829507868097e-05,
+      "loss": 2.4725,
+      "step": 1990
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 14.587154388427734,
+      "learning_rate": 1.6010824897263707e-05,
+      "loss": 2.7513,
+      "step": 1991
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 19.951908111572266,
+      "learning_rate": 1.6008820286659317e-05,
+      "loss": 2.3058,
+      "step": 1992
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 18.910688400268555,
+      "learning_rate": 1.6006815676054927e-05,
+      "loss": 2.4444,
+      "step": 1993
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 12.591381072998047,
+      "learning_rate": 1.6004811065450537e-05,
+      "loss": 3.0018,
+      "step": 1994
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 25.672998428344727,
+      "learning_rate": 1.6002806454846147e-05,
+      "loss": 2.6811,
+      "step": 1995
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 24.982601165771484,
+      "learning_rate": 1.6000801844241757e-05,
+      "loss": 3.1955,
+      "step": 1996
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 12.664204597473145,
+      "learning_rate": 1.5998797233637367e-05,
+      "loss": 2.2998,
+      "step": 1997
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 10.164942741394043,
+      "learning_rate": 1.5996792623032977e-05,
+      "loss": 3.4713,
+      "step": 1998
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 11.533797264099121,
+      "learning_rate": 1.5994788012428588e-05,
+      "loss": 2.7492,
+      "step": 1999
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 31.470319747924805,
+      "learning_rate": 1.5992783401824198e-05,
+      "loss": 3.0072,
+      "step": 2000
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 17.791658401489258,
+      "learning_rate": 1.5990778791219808e-05,
+      "loss": 1.9317,
+      "step": 2001
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 31.150108337402344,
+      "learning_rate": 1.5988774180615418e-05,
+      "loss": 2.8161,
+      "step": 2002
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 13.658658027648926,
+      "learning_rate": 1.5986769570011028e-05,
+      "loss": 1.9486,
+      "step": 2003
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 13.074492454528809,
+      "learning_rate": 1.5984764959406635e-05,
+      "loss": 1.8306,
+      "step": 2004
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 13.934167861938477,
+      "learning_rate": 1.5982760348802248e-05,
+      "loss": 2.0469,
+      "step": 2005
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 19.25830078125,
+      "learning_rate": 1.5980755738197858e-05,
+      "loss": 3.0445,
+      "step": 2006
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 12.957606315612793,
+      "learning_rate": 1.5978751127593465e-05,
+      "loss": 2.2492,
+      "step": 2007
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 18.43289566040039,
+      "learning_rate": 1.5976746516989078e-05,
+      "loss": 3.2304,
+      "step": 2008
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 15.042205810546875,
+      "learning_rate": 1.5974741906384685e-05,
+      "loss": 2.5782,
+      "step": 2009
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 32.639183044433594,
+      "learning_rate": 1.5972737295780295e-05,
+      "loss": 2.2884,
+      "step": 2010
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 14.541574478149414,
+      "learning_rate": 1.5970732685175905e-05,
+      "loss": 2.1057,
+      "step": 2011
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 33.099979400634766,
+      "learning_rate": 1.5968728074571515e-05,
+      "loss": 2.8465,
+      "step": 2012
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 29.527706146240234,
+      "learning_rate": 1.5966723463967125e-05,
+      "loss": 1.8221,
+      "step": 2013
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 12.222841262817383,
+      "learning_rate": 1.5964718853362735e-05,
+      "loss": 1.9562,
+      "step": 2014
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 11.430624961853027,
+      "learning_rate": 1.5962714242758345e-05,
+      "loss": 1.8148,
+      "step": 2015
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 12.992792129516602,
+      "learning_rate": 1.5960709632153956e-05,
+      "loss": 2.0858,
+      "step": 2016
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 25.652822494506836,
+      "learning_rate": 1.5958705021549566e-05,
+      "loss": 3.0605,
+      "step": 2017
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 9.427448272705078,
+      "learning_rate": 1.5956700410945176e-05,
+      "loss": 2.0717,
+      "step": 2018
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 18.489501953125,
+      "learning_rate": 1.5954695800340786e-05,
+      "loss": 2.3319,
+      "step": 2019
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 14.960678100585938,
+      "learning_rate": 1.5952691189736396e-05,
+      "loss": 2.3096,
+      "step": 2020
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 17.850452423095703,
+      "learning_rate": 1.5950686579132003e-05,
+      "loss": 2.5948,
+      "step": 2021
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 13.320516586303711,
+      "learning_rate": 1.5948681968527616e-05,
+      "loss": 2.0685,
+      "step": 2022
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 12.267958641052246,
+      "learning_rate": 1.5946677357923223e-05,
+      "loss": 2.1863,
+      "step": 2023
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 17.197010040283203,
+      "learning_rate": 1.5944672747318833e-05,
+      "loss": 2.5286,
+      "step": 2024
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 16.6329288482666,
+      "learning_rate": 1.5942668136714446e-05,
+      "loss": 2.0742,
+      "step": 2025
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 36.75046157836914,
+      "learning_rate": 1.5940663526110053e-05,
+      "loss": 2.7239,
+      "step": 2026
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 14.18947982788086,
+      "learning_rate": 1.5938658915505663e-05,
+      "loss": 2.6615,
+      "step": 2027
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 13.752023696899414,
+      "learning_rate": 1.5936654304901273e-05,
+      "loss": 1.8863,
+      "step": 2028
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 17.094932556152344,
+      "learning_rate": 1.5934649694296883e-05,
+      "loss": 3.2093,
+      "step": 2029
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 16.756057739257812,
+      "learning_rate": 1.5932645083692497e-05,
+      "loss": 2.6765,
+      "step": 2030
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 21.48285484313965,
+      "learning_rate": 1.5930640473088103e-05,
+      "loss": 2.0404,
+      "step": 2031
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 31.83589744567871,
+      "learning_rate": 1.5928635862483714e-05,
+      "loss": 2.708,
+      "step": 2032
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 50.02532958984375,
+      "learning_rate": 1.5926631251879324e-05,
+      "loss": 2.5092,
+      "step": 2033
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 11.736242294311523,
+      "learning_rate": 1.5924626641274934e-05,
+      "loss": 1.6545,
+      "step": 2034
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 22.306257247924805,
+      "learning_rate": 1.5922622030670544e-05,
+      "loss": 2.438,
+      "step": 2035
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 19.088542938232422,
+      "learning_rate": 1.5920617420066154e-05,
+      "loss": 2.5573,
+      "step": 2036
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 43.53116989135742,
+      "learning_rate": 1.5918612809461764e-05,
+      "loss": 2.4721,
+      "step": 2037
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 16.74660873413086,
+      "learning_rate": 1.5916608198857374e-05,
+      "loss": 2.1669,
+      "step": 2038
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 18.83763885498047,
+      "learning_rate": 1.5914603588252984e-05,
+      "loss": 2.0639,
+      "step": 2039
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 12.473159790039062,
+      "learning_rate": 1.591259897764859e-05,
+      "loss": 2.0599,
+      "step": 2040
+    },
+    {
+      "epoch": 0.61,
+      "eval_loss": 0.35515907406806946,
+      "eval_runtime": 43.2723,
+      "eval_samples_per_second": 34.179,
+      "eval_steps_per_second": 34.179,
+      "step": 2040
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 34.54806900024414,
+      "learning_rate": 1.5910594367044204e-05,
+      "loss": 2.8225,
+      "step": 2041
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 11.377670288085938,
+      "learning_rate": 1.5908589756439814e-05,
+      "loss": 1.3387,
+      "step": 2042
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 21.465269088745117,
+      "learning_rate": 1.590658514583542e-05,
+      "loss": 3.1448,
+      "step": 2043
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 13.400733947753906,
+      "learning_rate": 1.5904580535231035e-05,
+      "loss": 1.9627,
+      "step": 2044
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 8.886704444885254,
+      "learning_rate": 1.590257592462664e-05,
+      "loss": 1.8101,
+      "step": 2045
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 17.08548355102539,
+      "learning_rate": 1.590057131402225e-05,
+      "loss": 2.1456,
+      "step": 2046
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 16.1516170501709,
+      "learning_rate": 1.589856670341786e-05,
+      "loss": 2.642,
+      "step": 2047
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 16.168611526489258,
+      "learning_rate": 1.589656209281347e-05,
+      "loss": 2.2961,
+      "step": 2048
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 17.722412109375,
+      "learning_rate": 1.589455748220908e-05,
+      "loss": 2.2299,
+      "step": 2049
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 17.987937927246094,
+      "learning_rate": 1.589255287160469e-05,
+      "loss": 2.1003,
+      "step": 2050
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 13.64321517944336,
+      "learning_rate": 1.5890548261000302e-05,
+      "loss": 2.5651,
+      "step": 2051
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 10.08339786529541,
+      "learning_rate": 1.5888543650395912e-05,
+      "loss": 2.679,
+      "step": 2052
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 19.540874481201172,
+      "learning_rate": 1.5886539039791522e-05,
+      "loss": 1.9754,
+      "step": 2053
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 11.823172569274902,
+      "learning_rate": 1.5884534429187132e-05,
+      "loss": 1.5079,
+      "step": 2054
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 14.114450454711914,
+      "learning_rate": 1.5882529818582742e-05,
+      "loss": 2.3742,
+      "step": 2055
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 9.858566284179688,
+      "learning_rate": 1.5880525207978352e-05,
+      "loss": 3.4989,
+      "step": 2056
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 14.25406551361084,
+      "learning_rate": 1.5878520597373962e-05,
+      "loss": 3.3511,
+      "step": 2057
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 16.247241973876953,
+      "learning_rate": 1.5876515986769572e-05,
+      "loss": 2.1215,
+      "step": 2058
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 13.218579292297363,
+      "learning_rate": 1.587451137616518e-05,
+      "loss": 2.4852,
+      "step": 2059
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 13.075812339782715,
+      "learning_rate": 1.5872506765560793e-05,
+      "loss": 2.3496,
+      "step": 2060
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 26.715604782104492,
+      "learning_rate": 1.5870502154956403e-05,
+      "loss": 2.1378,
+      "step": 2061
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 20.732059478759766,
+      "learning_rate": 1.586849754435201e-05,
+      "loss": 2.008,
+      "step": 2062
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 18.93284797668457,
+      "learning_rate": 1.5866492933747623e-05,
+      "loss": 3.4783,
+      "step": 2063
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 9.962442398071289,
+      "learning_rate": 1.586448832314323e-05,
+      "loss": 2.3671,
+      "step": 2064
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 9.932574272155762,
+      "learning_rate": 1.586248371253884e-05,
+      "loss": 1.7524,
+      "step": 2065
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 16.17888069152832,
+      "learning_rate": 1.586047910193445e-05,
+      "loss": 1.6364,
+      "step": 2066
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 24.62932014465332,
+      "learning_rate": 1.585847449133006e-05,
+      "loss": 1.8643,
+      "step": 2067
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 17.033721923828125,
+      "learning_rate": 1.585646988072567e-05,
+      "loss": 2.2953,
+      "step": 2068
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 24.557415008544922,
+      "learning_rate": 1.585446527012128e-05,
+      "loss": 3.0784,
+      "step": 2069
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 27.415164947509766,
+      "learning_rate": 1.585246065951689e-05,
+      "loss": 2.3325,
+      "step": 2070
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 18.90687370300293,
+      "learning_rate": 1.58504560489125e-05,
+      "loss": 2.4899,
+      "step": 2071
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 12.064643859863281,
+      "learning_rate": 1.584845143830811e-05,
+      "loss": 2.6493,
+      "step": 2072
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 15.442275047302246,
+      "learning_rate": 1.584644682770372e-05,
+      "loss": 1.9658,
+      "step": 2073
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 14.710504531860352,
+      "learning_rate": 1.584444221709933e-05,
+      "loss": 2.2267,
+      "step": 2074
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 9.406432151794434,
+      "learning_rate": 1.584243760649494e-05,
+      "loss": 1.4927,
+      "step": 2075
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 14.804086685180664,
+      "learning_rate": 1.584043299589055e-05,
+      "loss": 1.8073,
+      "step": 2076
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 15.176630973815918,
+      "learning_rate": 1.583842838528616e-05,
+      "loss": 1.8248,
+      "step": 2077
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 20.446643829345703,
+      "learning_rate": 1.5836423774681767e-05,
+      "loss": 2.3386,
+      "step": 2078
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 20.13260269165039,
+      "learning_rate": 1.583441916407738e-05,
+      "loss": 2.7811,
+      "step": 2079
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 18.64963150024414,
+      "learning_rate": 1.583241455347299e-05,
+      "loss": 2.9017,
+      "step": 2080
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 17.931795120239258,
+      "learning_rate": 1.5830409942868597e-05,
+      "loss": 2.3632,
+      "step": 2081
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 19.928274154663086,
+      "learning_rate": 1.582840533226421e-05,
+      "loss": 1.7388,
+      "step": 2082
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 11.930707931518555,
+      "learning_rate": 1.5826400721659818e-05,
+      "loss": 1.5894,
+      "step": 2083
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 24.95637321472168,
+      "learning_rate": 1.5824396111055428e-05,
+      "loss": 3.1926,
+      "step": 2084
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 14.697010040283203,
+      "learning_rate": 1.582239150045104e-05,
+      "loss": 3.0097,
+      "step": 2085
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 9.35490894317627,
+      "learning_rate": 1.5820386889846648e-05,
+      "loss": 1.5664,
+      "step": 2086
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 21.47760581970215,
+      "learning_rate": 1.5818382279242258e-05,
+      "loss": 2.8386,
+      "step": 2087
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 16.634397506713867,
+      "learning_rate": 1.5816377668637868e-05,
+      "loss": 2.8702,
+      "step": 2088
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 12.756913185119629,
+      "learning_rate": 1.5814373058033478e-05,
+      "loss": 2.456,
+      "step": 2089
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 13.137290000915527,
+      "learning_rate": 1.5812368447429088e-05,
+      "loss": 1.8993,
+      "step": 2090
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 23.007200241088867,
+      "learning_rate": 1.58103638368247e-05,
+      "loss": 2.7142,
+      "step": 2091
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 10.685704231262207,
+      "learning_rate": 1.580835922622031e-05,
+      "loss": 1.872,
+      "step": 2092
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 34.11006164550781,
+      "learning_rate": 1.580635461561592e-05,
+      "loss": 2.8411,
+      "step": 2093
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 15.17668628692627,
+      "learning_rate": 1.580435000501153e-05,
+      "loss": 2.318,
+      "step": 2094
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 10.943110466003418,
+      "learning_rate": 1.580234539440714e-05,
+      "loss": 3.1694,
+      "step": 2095
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 17.957090377807617,
+      "learning_rate": 1.580034078380275e-05,
+      "loss": 2.8144,
+      "step": 2096
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 21.06376075744629,
+      "learning_rate": 1.5798336173198355e-05,
+      "loss": 2.4499,
+      "step": 2097
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 17.400442123413086,
+      "learning_rate": 1.579633156259397e-05,
+      "loss": 1.1497,
+      "step": 2098
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 14.588909149169922,
+      "learning_rate": 1.579432695198958e-05,
+      "loss": 3.2946,
+      "step": 2099
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 13.294282913208008,
+      "learning_rate": 1.5792322341385186e-05,
+      "loss": 1.6623,
+      "step": 2100
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 23.731582641601562,
+      "learning_rate": 1.57903177307808e-05,
+      "loss": 3.3506,
+      "step": 2101
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 15.64577579498291,
+      "learning_rate": 1.5788313120176406e-05,
+      "loss": 2.6303,
+      "step": 2102
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 19.747045516967773,
+      "learning_rate": 1.5786308509572016e-05,
+      "loss": 2.5166,
+      "step": 2103
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 42.6046142578125,
+      "learning_rate": 1.578430389896763e-05,
+      "loss": 2.8453,
+      "step": 2104
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 8.715738296508789,
+      "learning_rate": 1.5782299288363236e-05,
+      "loss": 1.9525,
+      "step": 2105
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 31.00543975830078,
+      "learning_rate": 1.5780294677758846e-05,
+      "loss": 3.0852,
+      "step": 2106
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 10.329021453857422,
+      "learning_rate": 1.5778290067154456e-05,
+      "loss": 2.1254,
+      "step": 2107
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 12.255505561828613,
+      "learning_rate": 1.5776285456550066e-05,
+      "loss": 2.1254,
+      "step": 2108
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 13.093355178833008,
+      "learning_rate": 1.5774280845945676e-05,
+      "loss": 2.1318,
+      "step": 2109
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 34.728302001953125,
+      "learning_rate": 1.5772276235341287e-05,
+      "loss": 3.1983,
+      "step": 2110
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 13.47143840789795,
+      "learning_rate": 1.5770271624736897e-05,
+      "loss": 2.0851,
+      "step": 2111
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 13.41025161743164,
+      "learning_rate": 1.5768267014132507e-05,
+      "loss": 2.8133,
+      "step": 2112
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 15.229413986206055,
+      "learning_rate": 1.5766262403528117e-05,
+      "loss": 2.8769,
+      "step": 2113
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 29.792980194091797,
+      "learning_rate": 1.5764257792923724e-05,
+      "loss": 2.6786,
+      "step": 2114
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 12.121004104614258,
+      "learning_rate": 1.5762253182319337e-05,
+      "loss": 2.3366,
+      "step": 2115
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 17.383033752441406,
+      "learning_rate": 1.5760248571714947e-05,
+      "loss": 2.2895,
+      "step": 2116
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 28.382801055908203,
+      "learning_rate": 1.5758243961110554e-05,
+      "loss": 2.5891,
+      "step": 2117
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 14.136621475219727,
+      "learning_rate": 1.5756239350506167e-05,
+      "loss": 2.4559,
+      "step": 2118
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 12.245827674865723,
+      "learning_rate": 1.5754234739901774e-05,
+      "loss": 1.3721,
+      "step": 2119
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 10.706925392150879,
+      "learning_rate": 1.5752230129297384e-05,
+      "loss": 2.2164,
+      "step": 2120
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 16.694616317749023,
+      "learning_rate": 1.5750225518692994e-05,
+      "loss": 3.428,
+      "step": 2121
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 33.90262985229492,
+      "learning_rate": 1.5748220908088604e-05,
+      "loss": 3.3125,
+      "step": 2122
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 12.597078323364258,
+      "learning_rate": 1.5746216297484214e-05,
+      "loss": 1.9381,
+      "step": 2123
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 13.725189208984375,
+      "learning_rate": 1.5744211686879824e-05,
+      "loss": 1.9921,
+      "step": 2124
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 13.38775634765625,
+      "learning_rate": 1.5742207076275434e-05,
+      "loss": 2.3561,
+      "step": 2125
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 15.606164932250977,
+      "learning_rate": 1.5740202465671045e-05,
+      "loss": 3.1076,
+      "step": 2126
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 17.07231330871582,
+      "learning_rate": 1.5738197855066655e-05,
+      "loss": 2.0324,
+      "step": 2127
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 18.28791618347168,
+      "learning_rate": 1.5736193244462265e-05,
+      "loss": 2.8287,
+      "step": 2128
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 20.407081604003906,
+      "learning_rate": 1.5734188633857875e-05,
+      "loss": 2.3472,
+      "step": 2129
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 12.980356216430664,
+      "learning_rate": 1.5732184023253485e-05,
+      "loss": 3.1184,
+      "step": 2130
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 19.251155853271484,
+      "learning_rate": 1.5730179412649095e-05,
+      "loss": 2.2474,
+      "step": 2131
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 8.661418914794922,
+      "learning_rate": 1.5728174802044705e-05,
+      "loss": 1.4024,
+      "step": 2132
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 24.52813148498535,
+      "learning_rate": 1.5726170191440312e-05,
+      "loss": 2.4654,
+      "step": 2133
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 12.824371337890625,
+      "learning_rate": 1.5724165580835925e-05,
+      "loss": 1.5831,
+      "step": 2134
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 11.373884201049805,
+      "learning_rate": 1.5722160970231535e-05,
+      "loss": 2.043,
+      "step": 2135
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 24.369014739990234,
+      "learning_rate": 1.5720156359627142e-05,
+      "loss": 2.4182,
+      "step": 2136
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 12.161373138427734,
+      "learning_rate": 1.5718151749022755e-05,
+      "loss": 1.8717,
+      "step": 2137
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 22.5483341217041,
+      "learning_rate": 1.5716147138418362e-05,
+      "loss": 1.8705,
+      "step": 2138
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 16.376750946044922,
+      "learning_rate": 1.5714142527813972e-05,
+      "loss": 2.359,
+      "step": 2139
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 12.507425308227539,
+      "learning_rate": 1.5712137917209582e-05,
+      "loss": 1.8627,
+      "step": 2140
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 12.094954490661621,
+      "learning_rate": 1.5710133306605192e-05,
+      "loss": 1.5144,
+      "step": 2141
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 12.721661567687988,
+      "learning_rate": 1.5708128696000802e-05,
+      "loss": 2.4135,
+      "step": 2142
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 16.270231246948242,
+      "learning_rate": 1.5706124085396413e-05,
+      "loss": 1.5671,
+      "step": 2143
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 22.51713752746582,
+      "learning_rate": 1.5704119474792023e-05,
+      "loss": 2.8953,
+      "step": 2144
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 14.330571174621582,
+      "learning_rate": 1.5702114864187633e-05,
+      "loss": 1.948,
+      "step": 2145
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 16.00999641418457,
+      "learning_rate": 1.5700110253583243e-05,
+      "loss": 2.0595,
+      "step": 2146
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 23.264543533325195,
+      "learning_rate": 1.5698105642978853e-05,
+      "loss": 2.7985,
+      "step": 2147
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 12.527949333190918,
+      "learning_rate": 1.5696101032374463e-05,
+      "loss": 2.2662,
+      "step": 2148
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 23.13153076171875,
+      "learning_rate": 1.5694096421770073e-05,
+      "loss": 3.2819,
+      "step": 2149
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 11.828460693359375,
+      "learning_rate": 1.5692091811165683e-05,
+      "loss": 1.9632,
+      "step": 2150
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 30.631380081176758,
+      "learning_rate": 1.5690087200561293e-05,
+      "loss": 2.5282,
+      "step": 2151
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 16.613862991333008,
+      "learning_rate": 1.56880825899569e-05,
+      "loss": 2.6759,
+      "step": 2152
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 36.72877502441406,
+      "learning_rate": 1.5686077979352513e-05,
+      "loss": 2.0691,
+      "step": 2153
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 11.451091766357422,
+      "learning_rate": 1.5684073368748123e-05,
+      "loss": 1.2804,
+      "step": 2154
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 13.62600326538086,
+      "learning_rate": 1.568206875814373e-05,
+      "loss": 1.9118,
+      "step": 2155
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 21.245529174804688,
+      "learning_rate": 1.5680064147539344e-05,
+      "loss": 2.7814,
+      "step": 2156
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 22.127309799194336,
+      "learning_rate": 1.567805953693495e-05,
+      "loss": 1.8782,
+      "step": 2157
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 11.301019668579102,
+      "learning_rate": 1.567605492633056e-05,
+      "loss": 3.1518,
+      "step": 2158
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 11.62056827545166,
+      "learning_rate": 1.5674050315726174e-05,
+      "loss": 2.3903,
+      "step": 2159
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 15.686822891235352,
+      "learning_rate": 1.567204570512178e-05,
+      "loss": 2.4537,
+      "step": 2160
+    },
+    {
+      "epoch": 0.65,
+      "eval_loss": 0.3099038302898407,
+      "eval_runtime": 43.3953,
+      "eval_samples_per_second": 34.082,
+      "eval_steps_per_second": 34.082,
+      "step": 2160
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 14.51310920715332,
+      "learning_rate": 1.567004109451739e-05,
+      "loss": 2.5808,
+      "step": 2161
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 15.057507514953613,
+      "learning_rate": 1.5668036483913e-05,
+      "loss": 2.8596,
+      "step": 2162
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 12.982969284057617,
+      "learning_rate": 1.566603187330861e-05,
+      "loss": 2.5042,
+      "step": 2163
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 11.829024314880371,
+      "learning_rate": 1.566402726270422e-05,
+      "loss": 1.9732,
+      "step": 2164
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 14.070128440856934,
+      "learning_rate": 1.566202265209983e-05,
+      "loss": 2.6257,
+      "step": 2165
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 10.464433670043945,
+      "learning_rate": 1.566001804149544e-05,
+      "loss": 2.178,
+      "step": 2166
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 11.11734390258789,
+      "learning_rate": 1.565801343089105e-05,
+      "loss": 1.5285,
+      "step": 2167
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 14.957688331604004,
+      "learning_rate": 1.565600882028666e-05,
+      "loss": 2.1816,
+      "step": 2168
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 16.353477478027344,
+      "learning_rate": 1.565400420968227e-05,
+      "loss": 2.2476,
+      "step": 2169
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 9.818117141723633,
+      "learning_rate": 1.565199959907788e-05,
+      "loss": 2.1336,
+      "step": 2170
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 20.093345642089844,
+      "learning_rate": 1.564999498847349e-05,
+      "loss": 2.6549,
+      "step": 2171
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 18.352497100830078,
+      "learning_rate": 1.56479903778691e-05,
+      "loss": 2.6063,
+      "step": 2172
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 26.143571853637695,
+      "learning_rate": 1.5645985767264712e-05,
+      "loss": 2.8258,
+      "step": 2173
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 18.577131271362305,
+      "learning_rate": 1.564398115666032e-05,
+      "loss": 1.8137,
+      "step": 2174
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 11.283987045288086,
+      "learning_rate": 1.5641976546055932e-05,
+      "loss": 1.9124,
+      "step": 2175
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 19.12086296081543,
+      "learning_rate": 1.563997193545154e-05,
+      "loss": 2.3072,
+      "step": 2176
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 11.322100639343262,
+      "learning_rate": 1.563796732484715e-05,
+      "loss": 1.8458,
+      "step": 2177
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 16.309368133544922,
+      "learning_rate": 1.5635962714242762e-05,
+      "loss": 2.3407,
+      "step": 2178
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 42.83634567260742,
+      "learning_rate": 1.563395810363837e-05,
+      "loss": 2.9269,
+      "step": 2179
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 16.00682830810547,
+      "learning_rate": 1.563195349303398e-05,
+      "loss": 2.2228,
+      "step": 2180
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 10.3877592086792,
+      "learning_rate": 1.562994888242959e-05,
+      "loss": 2.0959,
+      "step": 2181
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 23.170366287231445,
+      "learning_rate": 1.56279442718252e-05,
+      "loss": 2.8749,
+      "step": 2182
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 10.100824356079102,
+      "learning_rate": 1.562593966122081e-05,
+      "loss": 2.0848,
+      "step": 2183
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 17.805927276611328,
+      "learning_rate": 1.562393505061642e-05,
+      "loss": 2.6061,
+      "step": 2184
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 18.987218856811523,
+      "learning_rate": 1.562193044001203e-05,
+      "loss": 2.7442,
+      "step": 2185
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 44.760196685791016,
+      "learning_rate": 1.561992582940764e-05,
+      "loss": 2.3028,
+      "step": 2186
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 17.44770622253418,
+      "learning_rate": 1.561792121880325e-05,
+      "loss": 3.0411,
+      "step": 2187
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 15.366558074951172,
+      "learning_rate": 1.5615916608198856e-05,
+      "loss": 2.4066,
+      "step": 2188
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 14.17463493347168,
+      "learning_rate": 1.561391199759447e-05,
+      "loss": 1.7436,
+      "step": 2189
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 11.214127540588379,
+      "learning_rate": 1.561190738699008e-05,
+      "loss": 1.9083,
+      "step": 2190
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 14.602449417114258,
+      "learning_rate": 1.5609902776385686e-05,
+      "loss": 2.8559,
+      "step": 2191
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 16.023332595825195,
+      "learning_rate": 1.56078981657813e-05,
+      "loss": 1.8949,
+      "step": 2192
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 20.160932540893555,
+      "learning_rate": 1.5605893555176907e-05,
+      "loss": 2.0295,
+      "step": 2193
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 12.389837265014648,
+      "learning_rate": 1.5603888944572517e-05,
+      "loss": 2.2384,
+      "step": 2194
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 18.265491485595703,
+      "learning_rate": 1.5601884333968127e-05,
+      "loss": 2.4571,
+      "step": 2195
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 14.70784854888916,
+      "learning_rate": 1.5599879723363737e-05,
+      "loss": 2.1679,
+      "step": 2196
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 16.466394424438477,
+      "learning_rate": 1.5597875112759347e-05,
+      "loss": 1.9488,
+      "step": 2197
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 21.658117294311523,
+      "learning_rate": 1.5595870502154957e-05,
+      "loss": 2.7079,
+      "step": 2198
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 17.353038787841797,
+      "learning_rate": 1.5593865891550567e-05,
+      "loss": 2.9999,
+      "step": 2199
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 15.819283485412598,
+      "learning_rate": 1.5591861280946177e-05,
+      "loss": 2.1287,
+      "step": 2200
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 13.62539291381836,
+      "learning_rate": 1.5589856670341787e-05,
+      "loss": 1.9569,
+      "step": 2201
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 18.662357330322266,
+      "learning_rate": 1.5587852059737397e-05,
+      "loss": 2.51,
+      "step": 2202
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 26.726213455200195,
+      "learning_rate": 1.5585847449133007e-05,
+      "loss": 1.9874,
+      "step": 2203
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 16.964014053344727,
+      "learning_rate": 1.5583842838528618e-05,
+      "loss": 3.0422,
+      "step": 2204
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 24.55567169189453,
+      "learning_rate": 1.5581838227924228e-05,
+      "loss": 2.4687,
+      "step": 2205
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 12.898763656616211,
+      "learning_rate": 1.5579833617319838e-05,
+      "loss": 1.8264,
+      "step": 2206
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 14.676507949829102,
+      "learning_rate": 1.5577829006715444e-05,
+      "loss": 1.942,
+      "step": 2207
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 21.591907501220703,
+      "learning_rate": 1.5575824396111058e-05,
+      "loss": 2.193,
+      "step": 2208
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 22.116395950317383,
+      "learning_rate": 1.5573819785506668e-05,
+      "loss": 1.8098,
+      "step": 2209
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 17.048828125,
+      "learning_rate": 1.5571815174902275e-05,
+      "loss": 2.4309,
+      "step": 2210
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 17.50075912475586,
+      "learning_rate": 1.5569810564297888e-05,
+      "loss": 3.5984,
+      "step": 2211
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 42.22193145751953,
+      "learning_rate": 1.5567805953693495e-05,
+      "loss": 2.6408,
+      "step": 2212
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 16.618879318237305,
+      "learning_rate": 1.5565801343089105e-05,
+      "loss": 1.9168,
+      "step": 2213
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 20.115299224853516,
+      "learning_rate": 1.5563796732484715e-05,
+      "loss": 2.5345,
+      "step": 2214
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 9.575844764709473,
+      "learning_rate": 1.5561792121880325e-05,
+      "loss": 2.5817,
+      "step": 2215
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 16.084333419799805,
+      "learning_rate": 1.5559787511275935e-05,
+      "loss": 2.8486,
+      "step": 2216
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 9.678840637207031,
+      "learning_rate": 1.5557782900671545e-05,
+      "loss": 2.0502,
+      "step": 2217
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 25.424049377441406,
+      "learning_rate": 1.5555778290067155e-05,
+      "loss": 2.3241,
+      "step": 2218
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 14.452130317687988,
+      "learning_rate": 1.5553773679462765e-05,
+      "loss": 2.2107,
+      "step": 2219
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 34.95116424560547,
+      "learning_rate": 1.5551769068858375e-05,
+      "loss": 2.9797,
+      "step": 2220
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 10.273086547851562,
+      "learning_rate": 1.5549764458253986e-05,
+      "loss": 1.676,
+      "step": 2221
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 16.252601623535156,
+      "learning_rate": 1.5547759847649596e-05,
+      "loss": 2.6206,
+      "step": 2222
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 10.310708045959473,
+      "learning_rate": 1.5545755237045206e-05,
+      "loss": 2.1657,
+      "step": 2223
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 11.262001037597656,
+      "learning_rate": 1.5543750626440816e-05,
+      "loss": 1.9435,
+      "step": 2224
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 20.996740341186523,
+      "learning_rate": 1.5541746015836426e-05,
+      "loss": 2.7842,
+      "step": 2225
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 25.680212020874023,
+      "learning_rate": 1.5539741405232033e-05,
+      "loss": 2.4817,
+      "step": 2226
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 11.317098617553711,
+      "learning_rate": 1.5537736794627646e-05,
+      "loss": 1.2717,
+      "step": 2227
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 14.966314315795898,
+      "learning_rate": 1.5535732184023256e-05,
+      "loss": 2.5357,
+      "step": 2228
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 22.922651290893555,
+      "learning_rate": 1.5533727573418863e-05,
+      "loss": 3.0417,
+      "step": 2229
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 21.57832908630371,
+      "learning_rate": 1.5531722962814476e-05,
+      "loss": 2.1712,
+      "step": 2230
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 22.31909942626953,
+      "learning_rate": 1.5529718352210083e-05,
+      "loss": 2.7337,
+      "step": 2231
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 10.842183113098145,
+      "learning_rate": 1.5527713741605693e-05,
+      "loss": 1.886,
+      "step": 2232
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 17.99361228942871,
+      "learning_rate": 1.5525709131001307e-05,
+      "loss": 2.3479,
+      "step": 2233
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 36.20960998535156,
+      "learning_rate": 1.5523704520396913e-05,
+      "loss": 2.8728,
+      "step": 2234
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 14.748534202575684,
+      "learning_rate": 1.5521699909792523e-05,
+      "loss": 2.0896,
+      "step": 2235
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 10.2194242477417,
+      "learning_rate": 1.5519695299188133e-05,
+      "loss": 2.0784,
+      "step": 2236
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 19.986635208129883,
+      "learning_rate": 1.5517690688583744e-05,
+      "loss": 2.4334,
+      "step": 2237
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 19.66067886352539,
+      "learning_rate": 1.5515686077979354e-05,
+      "loss": 2.684,
+      "step": 2238
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 70.83124542236328,
+      "learning_rate": 1.5513681467374964e-05,
+      "loss": 2.5154,
+      "step": 2239
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 23.612783432006836,
+      "learning_rate": 1.5511676856770574e-05,
+      "loss": 2.886,
+      "step": 2240
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 15.822755813598633,
+      "learning_rate": 1.5509672246166184e-05,
+      "loss": 2.0681,
+      "step": 2241
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 25.98380470275879,
+      "learning_rate": 1.5507667635561794e-05,
+      "loss": 2.8038,
+      "step": 2242
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 27.250049591064453,
+      "learning_rate": 1.5505663024957404e-05,
+      "loss": 2.5481,
+      "step": 2243
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 36.15570831298828,
+      "learning_rate": 1.5503658414353014e-05,
+      "loss": 2.1858,
+      "step": 2244
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 10.935135841369629,
+      "learning_rate": 1.5501653803748624e-05,
+      "loss": 1.1376,
+      "step": 2245
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 21.463272094726562,
+      "learning_rate": 1.5499649193144234e-05,
+      "loss": 2.5708,
+      "step": 2246
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 12.168373107910156,
+      "learning_rate": 1.5497644582539844e-05,
+      "loss": 2.3351,
+      "step": 2247
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 20.284778594970703,
+      "learning_rate": 1.549563997193545e-05,
+      "loss": 2.2741,
+      "step": 2248
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 21.697620391845703,
+      "learning_rate": 1.5493635361331065e-05,
+      "loss": 3.3955,
+      "step": 2249
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 15.422367095947266,
+      "learning_rate": 1.549163075072667e-05,
+      "loss": 2.879,
+      "step": 2250
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 17.34280776977539,
+      "learning_rate": 1.548962614012228e-05,
+      "loss": 2.5866,
+      "step": 2251
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 17.975322723388672,
+      "learning_rate": 1.5487621529517895e-05,
+      "loss": 2.3053,
+      "step": 2252
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 30.330408096313477,
+      "learning_rate": 1.54856169189135e-05,
+      "loss": 3.002,
+      "step": 2253
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 17.363624572753906,
+      "learning_rate": 1.548361230830911e-05,
+      "loss": 2.323,
+      "step": 2254
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 14.340827941894531,
+      "learning_rate": 1.548160769770472e-05,
+      "loss": 1.8183,
+      "step": 2255
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 11.480280876159668,
+      "learning_rate": 1.5479603087100332e-05,
+      "loss": 1.9501,
+      "step": 2256
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 10.279202461242676,
+      "learning_rate": 1.5477598476495942e-05,
+      "loss": 2.1804,
+      "step": 2257
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 7.849579334259033,
+      "learning_rate": 1.5475593865891552e-05,
+      "loss": 2.1096,
+      "step": 2258
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 15.246369361877441,
+      "learning_rate": 1.5473589255287162e-05,
+      "loss": 2.5827,
+      "step": 2259
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 11.893373489379883,
+      "learning_rate": 1.5471584644682772e-05,
+      "loss": 1.4826,
+      "step": 2260
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 22.876691818237305,
+      "learning_rate": 1.5469580034078382e-05,
+      "loss": 2.7584,
+      "step": 2261
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 22.17548179626465,
+      "learning_rate": 1.546757542347399e-05,
+      "loss": 2.8086,
+      "step": 2262
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 19.29637908935547,
+      "learning_rate": 1.5465570812869602e-05,
+      "loss": 2.4008,
+      "step": 2263
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 18.765453338623047,
+      "learning_rate": 1.5463566202265212e-05,
+      "loss": 2.7496,
+      "step": 2264
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 15.087578773498535,
+      "learning_rate": 1.5461561591660823e-05,
+      "loss": 2.6039,
+      "step": 2265
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 9.94806957244873,
+      "learning_rate": 1.5459556981056433e-05,
+      "loss": 1.0356,
+      "step": 2266
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 20.82723617553711,
+      "learning_rate": 1.545755237045204e-05,
+      "loss": 2.3768,
+      "step": 2267
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 11.183643341064453,
+      "learning_rate": 1.5455547759847653e-05,
+      "loss": 2.0118,
+      "step": 2268
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 11.922686576843262,
+      "learning_rate": 1.545354314924326e-05,
+      "loss": 2.3468,
+      "step": 2269
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 10.599013328552246,
+      "learning_rate": 1.545153853863887e-05,
+      "loss": 1.6547,
+      "step": 2270
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 9.613250732421875,
+      "learning_rate": 1.5449533928034483e-05,
+      "loss": 2.3148,
+      "step": 2271
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 20.166654586791992,
+      "learning_rate": 1.544752931743009e-05,
+      "loss": 2.3727,
+      "step": 2272
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 14.838998794555664,
+      "learning_rate": 1.54455247068257e-05,
+      "loss": 1.8119,
+      "step": 2273
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 15.088861465454102,
+      "learning_rate": 1.544352009622131e-05,
+      "loss": 1.2957,
+      "step": 2274
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 16.902069091796875,
+      "learning_rate": 1.544151548561692e-05,
+      "loss": 2.637,
+      "step": 2275
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 39.02754592895508,
+      "learning_rate": 1.543951087501253e-05,
+      "loss": 2.1298,
+      "step": 2276
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 13.487964630126953,
+      "learning_rate": 1.543750626440814e-05,
+      "loss": 2.4923,
+      "step": 2277
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 28.00465965270996,
+      "learning_rate": 1.543550165380375e-05,
+      "loss": 2.9849,
+      "step": 2278
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 12.890748977661133,
+      "learning_rate": 1.543349704319936e-05,
+      "loss": 2.6115,
+      "step": 2279
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 14.795134544372559,
+      "learning_rate": 1.543149243259497e-05,
+      "loss": 1.9501,
+      "step": 2280
+    },
+    {
+      "epoch": 0.69,
+      "eval_loss": 0.27775686979293823,
+      "eval_runtime": 43.6573,
+      "eval_samples_per_second": 33.878,
+      "eval_steps_per_second": 33.878,
+      "step": 2280
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 17.335163116455078,
+      "learning_rate": 1.5429487821990577e-05,
+      "loss": 2.4545,
+      "step": 2281
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 19.485933303833008,
+      "learning_rate": 1.542748321138619e-05,
+      "loss": 1.9536,
+      "step": 2282
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 17.225101470947266,
+      "learning_rate": 1.54254786007818e-05,
+      "loss": 1.7893,
+      "step": 2283
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 12.618383407592773,
+      "learning_rate": 1.5423473990177407e-05,
+      "loss": 2.0978,
+      "step": 2284
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 21.85410499572754,
+      "learning_rate": 1.542146937957302e-05,
+      "loss": 2.1725,
+      "step": 2285
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 13.22624397277832,
+      "learning_rate": 1.5419464768968628e-05,
+      "loss": 1.9183,
+      "step": 2286
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 14.312644004821777,
+      "learning_rate": 1.5417460158364238e-05,
+      "loss": 1.764,
+      "step": 2287
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 17.018287658691406,
+      "learning_rate": 1.541545554775985e-05,
+      "loss": 1.6649,
+      "step": 2288
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 19.239450454711914,
+      "learning_rate": 1.5413450937155458e-05,
+      "loss": 2.6163,
+      "step": 2289
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 8.873007774353027,
+      "learning_rate": 1.5411446326551068e-05,
+      "loss": 2.4475,
+      "step": 2290
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 34.43484878540039,
+      "learning_rate": 1.5409441715946678e-05,
+      "loss": 3.171,
+      "step": 2291
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 13.385025024414062,
+      "learning_rate": 1.5407437105342288e-05,
+      "loss": 2.4151,
+      "step": 2292
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 23.839799880981445,
+      "learning_rate": 1.5405432494737898e-05,
+      "loss": 3.4928,
+      "step": 2293
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 13.791393280029297,
+      "learning_rate": 1.5403427884133508e-05,
+      "loss": 1.5945,
+      "step": 2294
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 15.341072082519531,
+      "learning_rate": 1.5401423273529118e-05,
+      "loss": 1.6761,
+      "step": 2295
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 16.246986389160156,
+      "learning_rate": 1.539941866292473e-05,
+      "loss": 1.9111,
+      "step": 2296
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 21.339174270629883,
+      "learning_rate": 1.539741405232034e-05,
+      "loss": 2.3112,
+      "step": 2297
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 19.16057777404785,
+      "learning_rate": 1.539540944171595e-05,
+      "loss": 1.8671,
+      "step": 2298
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 10.452445983886719,
+      "learning_rate": 1.539340483111156e-05,
+      "loss": 1.1408,
+      "step": 2299
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 12.354015350341797,
+      "learning_rate": 1.5391400220507165e-05,
+      "loss": 2.4143,
+      "step": 2300
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 12.557357788085938,
+      "learning_rate": 1.538939560990278e-05,
+      "loss": 1.8995,
+      "step": 2301
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 16.86603546142578,
+      "learning_rate": 1.538739099929839e-05,
+      "loss": 2.0175,
+      "step": 2302
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 22.319530487060547,
+      "learning_rate": 1.5385386388693996e-05,
+      "loss": 3.6267,
+      "step": 2303
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 21.675506591796875,
+      "learning_rate": 1.538338177808961e-05,
+      "loss": 2.8276,
+      "step": 2304
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 25.780988693237305,
+      "learning_rate": 1.5381377167485216e-05,
+      "loss": 2.1939,
+      "step": 2305
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 21.725067138671875,
+      "learning_rate": 1.5379372556880826e-05,
+      "loss": 2.4273,
+      "step": 2306
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 13.079160690307617,
+      "learning_rate": 1.537736794627644e-05,
+      "loss": 2.748,
+      "step": 2307
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 70.98606872558594,
+      "learning_rate": 1.5375363335672046e-05,
+      "loss": 3.8798,
+      "step": 2308
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 30.048757553100586,
+      "learning_rate": 1.5373358725067656e-05,
+      "loss": 1.797,
+      "step": 2309
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 19.938982009887695,
+      "learning_rate": 1.5371354114463266e-05,
+      "loss": 2.9236,
+      "step": 2310
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 19.096269607543945,
+      "learning_rate": 1.5369349503858876e-05,
+      "loss": 3.0036,
+      "step": 2311
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 16.5783634185791,
+      "learning_rate": 1.5367344893254486e-05,
+      "loss": 2.3479,
+      "step": 2312
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 11.655875205993652,
+      "learning_rate": 1.5365340282650096e-05,
+      "loss": 2.0129,
+      "step": 2313
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 12.814486503601074,
+      "learning_rate": 1.5363335672045706e-05,
+      "loss": 2.283,
+      "step": 2314
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 12.23183822631836,
+      "learning_rate": 1.5361331061441317e-05,
+      "loss": 2.0969,
+      "step": 2315
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 23.845361709594727,
+      "learning_rate": 1.5359326450836927e-05,
+      "loss": 2.6363,
+      "step": 2316
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 22.525117874145508,
+      "learning_rate": 1.5357321840232537e-05,
+      "loss": 2.7907,
+      "step": 2317
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 35.314666748046875,
+      "learning_rate": 1.5355317229628147e-05,
+      "loss": 2.434,
+      "step": 2318
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 10.943061828613281,
+      "learning_rate": 1.5353312619023757e-05,
+      "loss": 1.8863,
+      "step": 2319
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 16.291006088256836,
+      "learning_rate": 1.5351308008419367e-05,
+      "loss": 1.8541,
+      "step": 2320
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 12.258851051330566,
+      "learning_rate": 1.5349303397814977e-05,
+      "loss": 2.3217,
+      "step": 2321
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 12.439240455627441,
+      "learning_rate": 1.5347298787210584e-05,
+      "loss": 2.1214,
+      "step": 2322
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 24.879018783569336,
+      "learning_rate": 1.5345294176606197e-05,
+      "loss": 2.4778,
+      "step": 2323
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 14.103094100952148,
+      "learning_rate": 1.5343289566001804e-05,
+      "loss": 3.2186,
+      "step": 2324
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 16.0571346282959,
+      "learning_rate": 1.5341284955397414e-05,
+      "loss": 1.7382,
+      "step": 2325
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 14.569293975830078,
+      "learning_rate": 1.5339280344793027e-05,
+      "loss": 2.6728,
+      "step": 2326
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 19.245534896850586,
+      "learning_rate": 1.5337275734188634e-05,
+      "loss": 2.1571,
+      "step": 2327
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 35.83250427246094,
+      "learning_rate": 1.5335271123584244e-05,
+      "loss": 3.574,
+      "step": 2328
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 22.268566131591797,
+      "learning_rate": 1.5333266512979854e-05,
+      "loss": 2.8414,
+      "step": 2329
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 13.335097312927246,
+      "learning_rate": 1.5331261902375464e-05,
+      "loss": 2.2446,
+      "step": 2330
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 21.128978729248047,
+      "learning_rate": 1.5329257291771075e-05,
+      "loss": 2.1243,
+      "step": 2331
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 15.031562805175781,
+      "learning_rate": 1.5327252681166685e-05,
+      "loss": 2.1856,
+      "step": 2332
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 24.040231704711914,
+      "learning_rate": 1.5325248070562295e-05,
+      "loss": 2.7496,
+      "step": 2333
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 21.210121154785156,
+      "learning_rate": 1.5323243459957905e-05,
+      "loss": 2.1727,
+      "step": 2334
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 14.999757766723633,
+      "learning_rate": 1.5321238849353515e-05,
+      "loss": 1.8486,
+      "step": 2335
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 21.83721160888672,
+      "learning_rate": 1.5319234238749125e-05,
+      "loss": 2.4315,
+      "step": 2336
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 10.136129379272461,
+      "learning_rate": 1.5317229628144735e-05,
+      "loss": 1.8972,
+      "step": 2337
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 14.037278175354004,
+      "learning_rate": 1.5315225017540345e-05,
+      "loss": 1.9107,
+      "step": 2338
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 12.793769836425781,
+      "learning_rate": 1.5313220406935955e-05,
+      "loss": 1.8938,
+      "step": 2339
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 46.66535568237305,
+      "learning_rate": 1.5311215796331565e-05,
+      "loss": 3.1435,
+      "step": 2340
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 24.93342399597168,
+      "learning_rate": 1.5309211185727172e-05,
+      "loss": 2.3244,
+      "step": 2341
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 17.550573348999023,
+      "learning_rate": 1.5307206575122785e-05,
+      "loss": 1.4479,
+      "step": 2342
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 23.929763793945312,
+      "learning_rate": 1.5305201964518392e-05,
+      "loss": 2.6339,
+      "step": 2343
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 18.37746238708496,
+      "learning_rate": 1.5303197353914002e-05,
+      "loss": 1.916,
+      "step": 2344
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 32.94269943237305,
+      "learning_rate": 1.5301192743309616e-05,
+      "loss": 2.7435,
+      "step": 2345
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 14.292465209960938,
+      "learning_rate": 1.5299188132705222e-05,
+      "loss": 1.8105,
+      "step": 2346
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 12.711530685424805,
+      "learning_rate": 1.5297183522100832e-05,
+      "loss": 1.9106,
+      "step": 2347
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 16.853271484375,
+      "learning_rate": 1.5295178911496443e-05,
+      "loss": 1.9825,
+      "step": 2348
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 17.740068435668945,
+      "learning_rate": 1.5293174300892053e-05,
+      "loss": 1.5591,
+      "step": 2349
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 19.151620864868164,
+      "learning_rate": 1.5291169690287663e-05,
+      "loss": 3.9344,
+      "step": 2350
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 14.547835350036621,
+      "learning_rate": 1.5289165079683273e-05,
+      "loss": 2.0179,
+      "step": 2351
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 14.701192855834961,
+      "learning_rate": 1.5287160469078883e-05,
+      "loss": 2.1844,
+      "step": 2352
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 20.547576904296875,
+      "learning_rate": 1.5285155858474493e-05,
+      "loss": 2.5008,
+      "step": 2353
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 13.004501342773438,
+      "learning_rate": 1.5283151247870103e-05,
+      "loss": 1.8492,
+      "step": 2354
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 18.574153900146484,
+      "learning_rate": 1.528114663726571e-05,
+      "loss": 1.7889,
+      "step": 2355
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 21.489116668701172,
+      "learning_rate": 1.5279142026661323e-05,
+      "loss": 2.2442,
+      "step": 2356
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 11.905816078186035,
+      "learning_rate": 1.5277137416056933e-05,
+      "loss": 1.6658,
+      "step": 2357
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 16.373111724853516,
+      "learning_rate": 1.527513280545254e-05,
+      "loss": 1.7949,
+      "step": 2358
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 14.20676326751709,
+      "learning_rate": 1.5273128194848153e-05,
+      "loss": 2.0334,
+      "step": 2359
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 14.752305030822754,
+      "learning_rate": 1.527112358424376e-05,
+      "loss": 2.1005,
+      "step": 2360
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 13.963676452636719,
+      "learning_rate": 1.526911897363937e-05,
+      "loss": 2.9856,
+      "step": 2361
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 11.170400619506836,
+      "learning_rate": 1.5267114363034984e-05,
+      "loss": 1.8121,
+      "step": 2362
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 17.13726043701172,
+      "learning_rate": 1.526510975243059e-05,
+      "loss": 2.1476,
+      "step": 2363
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 79.9980697631836,
+      "learning_rate": 1.52631051418262e-05,
+      "loss": 3.3246,
+      "step": 2364
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 21.73980712890625,
+      "learning_rate": 1.526110053122181e-05,
+      "loss": 2.9358,
+      "step": 2365
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 13.559164047241211,
+      "learning_rate": 1.525909592061742e-05,
+      "loss": 2.0114,
+      "step": 2366
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 14.406414031982422,
+      "learning_rate": 1.525709131001303e-05,
+      "loss": 1.1554,
+      "step": 2367
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 11.396732330322266,
+      "learning_rate": 1.5255086699408641e-05,
+      "loss": 2.1299,
+      "step": 2368
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 22.816299438476562,
+      "learning_rate": 1.5253082088804253e-05,
+      "loss": 2.8568,
+      "step": 2369
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 19.31545066833496,
+      "learning_rate": 1.5251077478199861e-05,
+      "loss": 2.9191,
+      "step": 2370
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 20.314498901367188,
+      "learning_rate": 1.5249072867595471e-05,
+      "loss": 1.6643,
+      "step": 2371
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 21.570932388305664,
+      "learning_rate": 1.524706825699108e-05,
+      "loss": 2.4345,
+      "step": 2372
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 12.214997291564941,
+      "learning_rate": 1.5245063646386691e-05,
+      "loss": 1.8357,
+      "step": 2373
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 15.286229133605957,
+      "learning_rate": 1.5243059035782301e-05,
+      "loss": 1.6836,
+      "step": 2374
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 18.07011604309082,
+      "learning_rate": 1.524105442517791e-05,
+      "loss": 2.6134,
+      "step": 2375
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 13.150678634643555,
+      "learning_rate": 1.5239049814573522e-05,
+      "loss": 2.336,
+      "step": 2376
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 14.786459922790527,
+      "learning_rate": 1.523704520396913e-05,
+      "loss": 1.8392,
+      "step": 2377
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 12.723509788513184,
+      "learning_rate": 1.523504059336474e-05,
+      "loss": 2.4568,
+      "step": 2378
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 17.796947479248047,
+      "learning_rate": 1.5233035982760348e-05,
+      "loss": 2.8035,
+      "step": 2379
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 7.9330854415893555,
+      "learning_rate": 1.523103137215596e-05,
+      "loss": 1.0556,
+      "step": 2380
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 20.23641586303711,
+      "learning_rate": 1.522902676155157e-05,
+      "loss": 2.4805,
+      "step": 2381
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 14.428129196166992,
+      "learning_rate": 1.5227022150947179e-05,
+      "loss": 2.3162,
+      "step": 2382
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 19.485916137695312,
+      "learning_rate": 1.522501754034279e-05,
+      "loss": 2.8538,
+      "step": 2383
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 23.651782989501953,
+      "learning_rate": 1.5223012929738399e-05,
+      "loss": 2.2296,
+      "step": 2384
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 20.420745849609375,
+      "learning_rate": 1.5221008319134009e-05,
+      "loss": 2.8712,
+      "step": 2385
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 13.436408042907715,
+      "learning_rate": 1.5219003708529619e-05,
+      "loss": 2.4493,
+      "step": 2386
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 9.011188507080078,
+      "learning_rate": 1.5216999097925229e-05,
+      "loss": 1.8621,
+      "step": 2387
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 24.090200424194336,
+      "learning_rate": 1.521499448732084e-05,
+      "loss": 2.7823,
+      "step": 2388
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 22.469762802124023,
+      "learning_rate": 1.521298987671645e-05,
+      "loss": 2.1563,
+      "step": 2389
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 20.96700668334961,
+      "learning_rate": 1.521098526611206e-05,
+      "loss": 2.3123,
+      "step": 2390
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 13.538233757019043,
+      "learning_rate": 1.5208980655507668e-05,
+      "loss": 1.8336,
+      "step": 2391
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 19.34774398803711,
+      "learning_rate": 1.520697604490328e-05,
+      "loss": 2.2524,
+      "step": 2392
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 23.267269134521484,
+      "learning_rate": 1.520497143429889e-05,
+      "loss": 3.195,
+      "step": 2393
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 43.13511657714844,
+      "learning_rate": 1.5202966823694498e-05,
+      "loss": 2.9556,
+      "step": 2394
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 12.41313362121582,
+      "learning_rate": 1.520096221309011e-05,
+      "loss": 2.4461,
+      "step": 2395
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 16.363807678222656,
+      "learning_rate": 1.5198957602485718e-05,
+      "loss": 2.2686,
+      "step": 2396
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 18.332998275756836,
+      "learning_rate": 1.5196952991881328e-05,
+      "loss": 2.3713,
+      "step": 2397
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 15.461286544799805,
+      "learning_rate": 1.5194948381276937e-05,
+      "loss": 2.0574,
+      "step": 2398
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 32.771728515625,
+      "learning_rate": 1.5192943770672548e-05,
+      "loss": 2.7732,
+      "step": 2399
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 15.005850791931152,
+      "learning_rate": 1.5190939160068158e-05,
+      "loss": 2.0059,
+      "step": 2400
+    },
+    {
+      "epoch": 0.72,
+      "eval_loss": 0.2799599766731262,
+      "eval_runtime": 43.2863,
+      "eval_samples_per_second": 34.168,
+      "eval_steps_per_second": 34.168,
+      "step": 2400
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 18.537982940673828,
+      "learning_rate": 1.5188934549463767e-05,
+      "loss": 2.3335,
+      "step": 2401
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 28.75159454345703,
+      "learning_rate": 1.5186929938859379e-05,
+      "loss": 2.5293,
+      "step": 2402
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 19.689199447631836,
+      "learning_rate": 1.5184925328254987e-05,
+      "loss": 3.4202,
+      "step": 2403
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 16.85619354248047,
+      "learning_rate": 1.5182920717650597e-05,
+      "loss": 2.1249,
+      "step": 2404
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 15.209796905517578,
+      "learning_rate": 1.5180916107046209e-05,
+      "loss": 2.1942,
+      "step": 2405
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 19.145883560180664,
+      "learning_rate": 1.5178911496441817e-05,
+      "loss": 2.3506,
+      "step": 2406
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 13.587536811828613,
+      "learning_rate": 1.5176906885837427e-05,
+      "loss": 2.483,
+      "step": 2407
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 14.813488960266113,
+      "learning_rate": 1.5174902275233036e-05,
+      "loss": 1.6431,
+      "step": 2408
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 7.390352249145508,
+      "learning_rate": 1.5172897664628648e-05,
+      "loss": 1.329,
+      "step": 2409
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 124.1446304321289,
+      "learning_rate": 1.5170893054024256e-05,
+      "loss": 2.8408,
+      "step": 2410
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 31.081998825073242,
+      "learning_rate": 1.5168888443419868e-05,
+      "loss": 2.8232,
+      "step": 2411
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 18.641647338867188,
+      "learning_rate": 1.5166883832815478e-05,
+      "loss": 1.8497,
+      "step": 2412
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 17.504941940307617,
+      "learning_rate": 1.5164879222211086e-05,
+      "loss": 2.4579,
+      "step": 2413
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 11.096992492675781,
+      "learning_rate": 1.5162874611606698e-05,
+      "loss": 2.1797,
+      "step": 2414
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 14.132731437683105,
+      "learning_rate": 1.5160870001002306e-05,
+      "loss": 1.5483,
+      "step": 2415
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 49.675132751464844,
+      "learning_rate": 1.5158865390397916e-05,
+      "loss": 2.4514,
+      "step": 2416
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 12.76675796508789,
+      "learning_rate": 1.5156860779793528e-05,
+      "loss": 1.2279,
+      "step": 2417
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 10.700181007385254,
+      "learning_rate": 1.5154856169189137e-05,
+      "loss": 2.0745,
+      "step": 2418
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 17.42853546142578,
+      "learning_rate": 1.5152851558584747e-05,
+      "loss": 2.7441,
+      "step": 2419
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 8.670104026794434,
+      "learning_rate": 1.5150846947980355e-05,
+      "loss": 1.8953,
+      "step": 2420
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 24.383087158203125,
+      "learning_rate": 1.5148842337375967e-05,
+      "loss": 2.6507,
+      "step": 2421
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 12.565937042236328,
+      "learning_rate": 1.5146837726771575e-05,
+      "loss": 1.5887,
+      "step": 2422
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 22.554595947265625,
+      "learning_rate": 1.5144833116167185e-05,
+      "loss": 2.4942,
+      "step": 2423
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 8.884187698364258,
+      "learning_rate": 1.5142828505562797e-05,
+      "loss": 1.2549,
+      "step": 2424
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 15.001145362854004,
+      "learning_rate": 1.5140823894958406e-05,
+      "loss": 2.1248,
+      "step": 2425
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 13.459707260131836,
+      "learning_rate": 1.5138819284354016e-05,
+      "loss": 2.0756,
+      "step": 2426
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 15.579310417175293,
+      "learning_rate": 1.5136814673749624e-05,
+      "loss": 2.8265,
+      "step": 2427
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 12.242110252380371,
+      "learning_rate": 1.5134810063145236e-05,
+      "loss": 2.2286,
+      "step": 2428
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 30.915904998779297,
+      "learning_rate": 1.5132805452540844e-05,
+      "loss": 2.2951,
+      "step": 2429
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 15.82615852355957,
+      "learning_rate": 1.5130800841936454e-05,
+      "loss": 2.0723,
+      "step": 2430
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 15.196529388427734,
+      "learning_rate": 1.5128796231332066e-05,
+      "loss": 2.2142,
+      "step": 2431
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 43.0781135559082,
+      "learning_rate": 1.5126791620727674e-05,
+      "loss": 3.249,
+      "step": 2432
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 10.74176025390625,
+      "learning_rate": 1.5124787010123284e-05,
+      "loss": 1.763,
+      "step": 2433
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 14.15688705444336,
+      "learning_rate": 1.5122782399518895e-05,
+      "loss": 1.8488,
+      "step": 2434
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 16.205501556396484,
+      "learning_rate": 1.5120777788914505e-05,
+      "loss": 2.8312,
+      "step": 2435
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 23.575292587280273,
+      "learning_rate": 1.5118773178310115e-05,
+      "loss": 1.8974,
+      "step": 2436
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 12.507033348083496,
+      "learning_rate": 1.5116768567705725e-05,
+      "loss": 2.0892,
+      "step": 2437
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 11.459385871887207,
+      "learning_rate": 1.5114763957101335e-05,
+      "loss": 2.232,
+      "step": 2438
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 13.064870834350586,
+      "learning_rate": 1.5112759346496943e-05,
+      "loss": 2.9855,
+      "step": 2439
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 21.083112716674805,
+      "learning_rate": 1.5110754735892555e-05,
+      "loss": 2.7381,
+      "step": 2440
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 16.428953170776367,
+      "learning_rate": 1.5108750125288163e-05,
+      "loss": 2.8427,
+      "step": 2441
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 13.690245628356934,
+      "learning_rate": 1.5106745514683774e-05,
+      "loss": 1.9179,
+      "step": 2442
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 13.37304401397705,
+      "learning_rate": 1.5104740904079385e-05,
+      "loss": 1.8193,
+      "step": 2443
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 10.359079360961914,
+      "learning_rate": 1.5102736293474994e-05,
+      "loss": 2.5558,
+      "step": 2444
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 12.4136381149292,
+      "learning_rate": 1.5100731682870604e-05,
+      "loss": 2.191,
+      "step": 2445
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 9.771052360534668,
+      "learning_rate": 1.5098727072266212e-05,
+      "loss": 1.5917,
+      "step": 2446
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 12.872698783874512,
+      "learning_rate": 1.5096722461661824e-05,
+      "loss": 1.6524,
+      "step": 2447
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 28.35697364807129,
+      "learning_rate": 1.5094717851057434e-05,
+      "loss": 2.569,
+      "step": 2448
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 18.459850311279297,
+      "learning_rate": 1.5092713240453042e-05,
+      "loss": 2.1667,
+      "step": 2449
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 12.4850492477417,
+      "learning_rate": 1.5090708629848654e-05,
+      "loss": 2.2356,
+      "step": 2450
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 21.161081314086914,
+      "learning_rate": 1.5088704019244263e-05,
+      "loss": 2.1366,
+      "step": 2451
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 22.59266471862793,
+      "learning_rate": 1.5086699408639873e-05,
+      "loss": 2.4504,
+      "step": 2452
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 11.43664836883545,
+      "learning_rate": 1.5084694798035481e-05,
+      "loss": 2.434,
+      "step": 2453
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 12.218484878540039,
+      "learning_rate": 1.5082690187431093e-05,
+      "loss": 1.2119,
+      "step": 2454
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 10.82869815826416,
+      "learning_rate": 1.5080685576826703e-05,
+      "loss": 1.8022,
+      "step": 2455
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 17.006765365600586,
+      "learning_rate": 1.5078680966222311e-05,
+      "loss": 2.1377,
+      "step": 2456
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 21.112869262695312,
+      "learning_rate": 1.5076676355617923e-05,
+      "loss": 2.799,
+      "step": 2457
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 24.335466384887695,
+      "learning_rate": 1.5074671745013532e-05,
+      "loss": 2.4533,
+      "step": 2458
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 20.010364532470703,
+      "learning_rate": 1.5072667134409142e-05,
+      "loss": 1.9907,
+      "step": 2459
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 34.753047943115234,
+      "learning_rate": 1.5070662523804753e-05,
+      "loss": 2.0948,
+      "step": 2460
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 19.572534561157227,
+      "learning_rate": 1.5068657913200362e-05,
+      "loss": 2.1058,
+      "step": 2461
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 13.721003532409668,
+      "learning_rate": 1.5066653302595972e-05,
+      "loss": 1.4905,
+      "step": 2462
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 14.475379943847656,
+      "learning_rate": 1.5064648691991582e-05,
+      "loss": 1.7658,
+      "step": 2463
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 24.027629852294922,
+      "learning_rate": 1.5062644081387192e-05,
+      "loss": 2.6237,
+      "step": 2464
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 15.826460838317871,
+      "learning_rate": 1.50606394707828e-05,
+      "loss": 2.6549,
+      "step": 2465
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 43.245018005371094,
+      "learning_rate": 1.5058634860178412e-05,
+      "loss": 2.9669,
+      "step": 2466
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 13.400920867919922,
+      "learning_rate": 1.5056630249574022e-05,
+      "loss": 1.9453,
+      "step": 2467
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 15.512686729431152,
+      "learning_rate": 1.505462563896963e-05,
+      "loss": 1.9563,
+      "step": 2468
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 18.510339736938477,
+      "learning_rate": 1.5052621028365242e-05,
+      "loss": 1.5118,
+      "step": 2469
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 13.163358688354492,
+      "learning_rate": 1.505061641776085e-05,
+      "loss": 1.8231,
+      "step": 2470
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 8.8781156539917,
+      "learning_rate": 1.5048611807156461e-05,
+      "loss": 1.1656,
+      "step": 2471
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 25.674057006835938,
+      "learning_rate": 1.504660719655207e-05,
+      "loss": 2.7279,
+      "step": 2472
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 16.811283111572266,
+      "learning_rate": 1.5044602585947681e-05,
+      "loss": 2.0451,
+      "step": 2473
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 15.99625301361084,
+      "learning_rate": 1.5042597975343291e-05,
+      "loss": 1.7554,
+      "step": 2474
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 16.740371704101562,
+      "learning_rate": 1.50405933647389e-05,
+      "loss": 2.9046,
+      "step": 2475
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 11.545669555664062,
+      "learning_rate": 1.5038588754134511e-05,
+      "loss": 1.1914,
+      "step": 2476
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 12.485255241394043,
+      "learning_rate": 1.503658414353012e-05,
+      "loss": 2.0212,
+      "step": 2477
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 28.34869956970215,
+      "learning_rate": 1.503457953292573e-05,
+      "loss": 1.4173,
+      "step": 2478
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 14.046238899230957,
+      "learning_rate": 1.5032574922321342e-05,
+      "loss": 2.3209,
+      "step": 2479
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 40.98884201049805,
+      "learning_rate": 1.503057031171695e-05,
+      "loss": 2.7137,
+      "step": 2480
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 16.54846954345703,
+      "learning_rate": 1.502856570111256e-05,
+      "loss": 1.7074,
+      "step": 2481
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 34.91267776489258,
+      "learning_rate": 1.502656109050817e-05,
+      "loss": 3.7392,
+      "step": 2482
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 11.860535621643066,
+      "learning_rate": 1.502455647990378e-05,
+      "loss": 2.6046,
+      "step": 2483
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 14.997228622436523,
+      "learning_rate": 1.5022551869299389e-05,
+      "loss": 2.761,
+      "step": 2484
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 15.930320739746094,
+      "learning_rate": 1.5020547258695e-05,
+      "loss": 2.0798,
+      "step": 2485
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 28.564027786254883,
+      "learning_rate": 1.501854264809061e-05,
+      "loss": 2.5476,
+      "step": 2486
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 18.10663604736328,
+      "learning_rate": 1.5016538037486219e-05,
+      "loss": 2.6392,
+      "step": 2487
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 22.97527313232422,
+      "learning_rate": 1.501453342688183e-05,
+      "loss": 2.0554,
+      "step": 2488
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 18.91280174255371,
+      "learning_rate": 1.5012528816277439e-05,
+      "loss": 2.5951,
+      "step": 2489
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 22.503339767456055,
+      "learning_rate": 1.5010524205673049e-05,
+      "loss": 1.7317,
+      "step": 2490
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 20.57047462463379,
+      "learning_rate": 1.5008519595068661e-05,
+      "loss": 2.3906,
+      "step": 2491
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 17.44088363647461,
+      "learning_rate": 1.500651498446427e-05,
+      "loss": 2.3833,
+      "step": 2492
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 11.56372356414795,
+      "learning_rate": 1.500451037385988e-05,
+      "loss": 2.2393,
+      "step": 2493
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 36.66616439819336,
+      "learning_rate": 1.5002505763255488e-05,
+      "loss": 2.0996,
+      "step": 2494
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 59.76615905761719,
+      "learning_rate": 1.50005011526511e-05,
+      "loss": 1.3206,
+      "step": 2495
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 10.487852096557617,
+      "learning_rate": 1.4998496542046708e-05,
+      "loss": 1.2596,
+      "step": 2496
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 17.123096466064453,
+      "learning_rate": 1.4996491931442318e-05,
+      "loss": 2.1562,
+      "step": 2497
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 13.120617866516113,
+      "learning_rate": 1.499448732083793e-05,
+      "loss": 1.2493,
+      "step": 2498
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 11.029601097106934,
+      "learning_rate": 1.4992482710233538e-05,
+      "loss": 0.918,
+      "step": 2499
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 38.185813903808594,
+      "learning_rate": 1.4990478099629148e-05,
+      "loss": 3.1123,
+      "step": 2500
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 21.2143611907959,
+      "learning_rate": 1.4988473489024757e-05,
+      "loss": 1.9136,
+      "step": 2501
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 18.287214279174805,
+      "learning_rate": 1.4986468878420368e-05,
+      "loss": 2.3058,
+      "step": 2502
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 20.503644943237305,
+      "learning_rate": 1.4984464267815979e-05,
+      "loss": 3.3248,
+      "step": 2503
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 18.117403030395508,
+      "learning_rate": 1.4982459657211587e-05,
+      "loss": 2.5097,
+      "step": 2504
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 16.31001853942871,
+      "learning_rate": 1.4980455046607199e-05,
+      "loss": 1.7266,
+      "step": 2505
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 12.371569633483887,
+      "learning_rate": 1.4978450436002807e-05,
+      "loss": 1.9252,
+      "step": 2506
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 11.64188289642334,
+      "learning_rate": 1.4976445825398417e-05,
+      "loss": 2.0164,
+      "step": 2507
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 17.393896102905273,
+      "learning_rate": 1.4974441214794027e-05,
+      "loss": 2.4099,
+      "step": 2508
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 16.229440689086914,
+      "learning_rate": 1.4972436604189637e-05,
+      "loss": 2.8192,
+      "step": 2509
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 19.683637619018555,
+      "learning_rate": 1.4970431993585247e-05,
+      "loss": 2.0972,
+      "step": 2510
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 23.35310173034668,
+      "learning_rate": 1.4968427382980858e-05,
+      "loss": 2.0364,
+      "step": 2511
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 10.506977081298828,
+      "learning_rate": 1.4966422772376468e-05,
+      "loss": 1.575,
+      "step": 2512
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 24.63909339904785,
+      "learning_rate": 1.4964418161772076e-05,
+      "loss": 1.9545,
+      "step": 2513
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 18.600322723388672,
+      "learning_rate": 1.4962413551167688e-05,
+      "loss": 2.0238,
+      "step": 2514
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 25.803688049316406,
+      "learning_rate": 1.4960408940563296e-05,
+      "loss": 1.7973,
+      "step": 2515
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 13.325533866882324,
+      "learning_rate": 1.4958404329958906e-05,
+      "loss": 1.5489,
+      "step": 2516
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 17.538631439208984,
+      "learning_rate": 1.4956399719354518e-05,
+      "loss": 1.6318,
+      "step": 2517
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 20.671701431274414,
+      "learning_rate": 1.4954395108750126e-05,
+      "loss": 2.2905,
+      "step": 2518
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 18.96885871887207,
+      "learning_rate": 1.4952390498145736e-05,
+      "loss": 1.2961,
+      "step": 2519
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 17.124177932739258,
+      "learning_rate": 1.4950385887541345e-05,
+      "loss": 2.1604,
+      "step": 2520
+    },
+    {
+      "epoch": 0.76,
+      "eval_loss": 0.269967257976532,
+      "eval_runtime": 43.2806,
+      "eval_samples_per_second": 34.172,
+      "eval_steps_per_second": 34.172,
+      "step": 2520
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 22.06162452697754,
+      "learning_rate": 1.4948381276936957e-05,
+      "loss": 2.1107,
+      "step": 2521
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 15.108476638793945,
+      "learning_rate": 1.4946376666332567e-05,
+      "loss": 1.7432,
+      "step": 2522
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 13.295374870300293,
+      "learning_rate": 1.4944372055728175e-05,
+      "loss": 2.03,
+      "step": 2523
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 19.436033248901367,
+      "learning_rate": 1.4942367445123787e-05,
+      "loss": 1.6658,
+      "step": 2524
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 36.37664031982422,
+      "learning_rate": 1.4940362834519395e-05,
+      "loss": 3.2737,
+      "step": 2525
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 11.738432884216309,
+      "learning_rate": 1.4938358223915005e-05,
+      "loss": 1.6208,
+      "step": 2526
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 17.715194702148438,
+      "learning_rate": 1.4936353613310614e-05,
+      "loss": 2.3697,
+      "step": 2527
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 21.08336639404297,
+      "learning_rate": 1.4934349002706226e-05,
+      "loss": 2.4212,
+      "step": 2528
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 19.754703521728516,
+      "learning_rate": 1.4932344392101836e-05,
+      "loss": 1.71,
+      "step": 2529
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 21.412639617919922,
+      "learning_rate": 1.4930339781497446e-05,
+      "loss": 2.807,
+      "step": 2530
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 18.40741729736328,
+      "learning_rate": 1.4928335170893056e-05,
+      "loss": 2.2224,
+      "step": 2531
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 33.87398147583008,
+      "learning_rate": 1.4926330560288664e-05,
+      "loss": 2.8005,
+      "step": 2532
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 14.195319175720215,
+      "learning_rate": 1.4924325949684276e-05,
+      "loss": 2.4552,
+      "step": 2533
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 14.160918235778809,
+      "learning_rate": 1.4922321339079886e-05,
+      "loss": 2.4736,
+      "step": 2534
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 31.502965927124023,
+      "learning_rate": 1.4920316728475494e-05,
+      "loss": 2.5237,
+      "step": 2535
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 17.290319442749023,
+      "learning_rate": 1.4918312117871106e-05,
+      "loss": 2.2399,
+      "step": 2536
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 13.980632781982422,
+      "learning_rate": 1.4916307507266715e-05,
+      "loss": 2.3494,
+      "step": 2537
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 14.719244956970215,
+      "learning_rate": 1.4914302896662325e-05,
+      "loss": 2.5073,
+      "step": 2538
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 19.83281707763672,
+      "learning_rate": 1.4912298286057933e-05,
+      "loss": 2.4892,
+      "step": 2539
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 13.933186531066895,
+      "learning_rate": 1.4910293675453545e-05,
+      "loss": 2.6791,
+      "step": 2540
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 20.118640899658203,
+      "learning_rate": 1.4908289064849155e-05,
+      "loss": 1.996,
+      "step": 2541
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 20.20134162902832,
+      "learning_rate": 1.4906284454244763e-05,
+      "loss": 2.2007,
+      "step": 2542
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 14.23250961303711,
+      "learning_rate": 1.4904279843640375e-05,
+      "loss": 2.2191,
+      "step": 2543
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 14.728852272033691,
+      "learning_rate": 1.4902275233035984e-05,
+      "loss": 1.9154,
+      "step": 2544
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 38.36711120605469,
+      "learning_rate": 1.4900270622431594e-05,
+      "loss": 2.367,
+      "step": 2545
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 17.614601135253906,
+      "learning_rate": 1.4898266011827205e-05,
+      "loss": 1.1838,
+      "step": 2546
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 23.657424926757812,
+      "learning_rate": 1.4896261401222814e-05,
+      "loss": 2.6376,
+      "step": 2547
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 10.91047191619873,
+      "learning_rate": 1.4894256790618424e-05,
+      "loss": 2.0975,
+      "step": 2548
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 19.83192253112793,
+      "learning_rate": 1.4892252180014032e-05,
+      "loss": 2.7231,
+      "step": 2549
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 6.836325645446777,
+      "learning_rate": 1.4890247569409644e-05,
+      "loss": 1.3988,
+      "step": 2550
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 16.332334518432617,
+      "learning_rate": 1.4888242958805252e-05,
+      "loss": 1.9877,
+      "step": 2551
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 23.387727737426758,
+      "learning_rate": 1.4886238348200862e-05,
+      "loss": 2.8849,
+      "step": 2552
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 33.43765640258789,
+      "learning_rate": 1.4884233737596474e-05,
+      "loss": 2.3284,
+      "step": 2553
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 10.308141708374023,
+      "learning_rate": 1.4882229126992083e-05,
+      "loss": 2.6766,
+      "step": 2554
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 20.615131378173828,
+      "learning_rate": 1.4880224516387693e-05,
+      "loss": 2.4127,
+      "step": 2555
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 18.643369674682617,
+      "learning_rate": 1.4878219905783303e-05,
+      "loss": 1.9807,
+      "step": 2556
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 29.241533279418945,
+      "learning_rate": 1.4876215295178913e-05,
+      "loss": 2.6441,
+      "step": 2557
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 19.864036560058594,
+      "learning_rate": 1.4874210684574521e-05,
+      "loss": 2.237,
+      "step": 2558
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 9.182751655578613,
+      "learning_rate": 1.4872206073970133e-05,
+      "loss": 1.6721,
+      "step": 2559
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 17.696592330932617,
+      "learning_rate": 1.4870201463365743e-05,
+      "loss": 1.8645,
+      "step": 2560
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 10.09614086151123,
+      "learning_rate": 1.4868196852761352e-05,
+      "loss": 1.4887,
+      "step": 2561
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 10.682119369506836,
+      "learning_rate": 1.4866192242156963e-05,
+      "loss": 1.296,
+      "step": 2562
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 18.929485321044922,
+      "learning_rate": 1.4864187631552572e-05,
+      "loss": 2.7717,
+      "step": 2563
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 21.349645614624023,
+      "learning_rate": 1.4862183020948182e-05,
+      "loss": 2.4609,
+      "step": 2564
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 7.982182502746582,
+      "learning_rate": 1.4860178410343794e-05,
+      "loss": 1.9265,
+      "step": 2565
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 15.188501358032227,
+      "learning_rate": 1.4858173799739402e-05,
+      "loss": 3.0483,
+      "step": 2566
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 37.26816177368164,
+      "learning_rate": 1.4856169189135012e-05,
+      "loss": 2.585,
+      "step": 2567
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 13.02489948272705,
+      "learning_rate": 1.485416457853062e-05,
+      "loss": 1.8137,
+      "step": 2568
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 15.654376029968262,
+      "learning_rate": 1.4852159967926232e-05,
+      "loss": 2.2203,
+      "step": 2569
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 9.267987251281738,
+      "learning_rate": 1.485015535732184e-05,
+      "loss": 1.362,
+      "step": 2570
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 31.388477325439453,
+      "learning_rate": 1.484815074671745e-05,
+      "loss": 2.3867,
+      "step": 2571
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 19.304162979125977,
+      "learning_rate": 1.4846146136113062e-05,
+      "loss": 2.6042,
+      "step": 2572
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 21.53330421447754,
+      "learning_rate": 1.4844141525508671e-05,
+      "loss": 2.2464,
+      "step": 2573
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 14.074399948120117,
+      "learning_rate": 1.4842136914904281e-05,
+      "loss": 1.6423,
+      "step": 2574
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 30.000886917114258,
+      "learning_rate": 1.484013230429989e-05,
+      "loss": 1.9405,
+      "step": 2575
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 12.38129711151123,
+      "learning_rate": 1.4838127693695501e-05,
+      "loss": 1.6343,
+      "step": 2576
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 15.83830451965332,
+      "learning_rate": 1.4836123083091111e-05,
+      "loss": 1.4498,
+      "step": 2577
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 20.503822326660156,
+      "learning_rate": 1.483411847248672e-05,
+      "loss": 2.8636,
+      "step": 2578
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 15.516641616821289,
+      "learning_rate": 1.4832113861882331e-05,
+      "loss": 2.0966,
+      "step": 2579
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 18.49781036376953,
+      "learning_rate": 1.483010925127794e-05,
+      "loss": 3.8588,
+      "step": 2580
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 11.766064643859863,
+      "learning_rate": 1.4828104640673552e-05,
+      "loss": 3.0766,
+      "step": 2581
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 28.008140563964844,
+      "learning_rate": 1.482610003006916e-05,
+      "loss": 2.8604,
+      "step": 2582
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 11.355888366699219,
+      "learning_rate": 1.482409541946477e-05,
+      "loss": 1.8569,
+      "step": 2583
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 24.921497344970703,
+      "learning_rate": 1.4822090808860382e-05,
+      "loss": 2.3311,
+      "step": 2584
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 49.57196807861328,
+      "learning_rate": 1.482008619825599e-05,
+      "loss": 3.8008,
+      "step": 2585
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 18.97112274169922,
+      "learning_rate": 1.48180815876516e-05,
+      "loss": 2.2731,
+      "step": 2586
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 19.699682235717773,
+      "learning_rate": 1.4816076977047209e-05,
+      "loss": 3.2435,
+      "step": 2587
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 15.111373901367188,
+      "learning_rate": 1.481407236644282e-05,
+      "loss": 2.6261,
+      "step": 2588
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 39.739288330078125,
+      "learning_rate": 1.481206775583843e-05,
+      "loss": 3.0957,
+      "step": 2589
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 7.469949245452881,
+      "learning_rate": 1.4810063145234039e-05,
+      "loss": 1.5037,
+      "step": 2590
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 48.33216857910156,
+      "learning_rate": 1.480805853462965e-05,
+      "loss": 1.5074,
+      "step": 2591
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 14.457548141479492,
+      "learning_rate": 1.4806053924025259e-05,
+      "loss": 2.0293,
+      "step": 2592
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 9.994394302368164,
+      "learning_rate": 1.480404931342087e-05,
+      "loss": 1.6135,
+      "step": 2593
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 16.707799911499023,
+      "learning_rate": 1.4802044702816478e-05,
+      "loss": 2.0803,
+      "step": 2594
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 11.45969009399414,
+      "learning_rate": 1.480004009221209e-05,
+      "loss": 2.0687,
+      "step": 2595
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 17.346237182617188,
+      "learning_rate": 1.47980354816077e-05,
+      "loss": 1.968,
+      "step": 2596
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 16.296354293823242,
+      "learning_rate": 1.4796030871003308e-05,
+      "loss": 2.7071,
+      "step": 2597
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 24.699426651000977,
+      "learning_rate": 1.479402626039892e-05,
+      "loss": 2.575,
+      "step": 2598
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 15.403938293457031,
+      "learning_rate": 1.4792021649794528e-05,
+      "loss": 2.3392,
+      "step": 2599
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 13.177908897399902,
+      "learning_rate": 1.4790017039190138e-05,
+      "loss": 1.4653,
+      "step": 2600
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 34.016021728515625,
+      "learning_rate": 1.4788012428585748e-05,
+      "loss": 2.3153,
+      "step": 2601
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 15.840597152709961,
+      "learning_rate": 1.4786007817981358e-05,
+      "loss": 2.5034,
+      "step": 2602
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 31.41590118408203,
+      "learning_rate": 1.4784003207376968e-05,
+      "loss": 2.0677,
+      "step": 2603
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 12.184178352355957,
+      "learning_rate": 1.4781998596772578e-05,
+      "loss": 2.0374,
+      "step": 2604
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 16.264497756958008,
+      "learning_rate": 1.4779993986168188e-05,
+      "loss": 2.4106,
+      "step": 2605
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 16.872928619384766,
+      "learning_rate": 1.4777989375563797e-05,
+      "loss": 2.3471,
+      "step": 2606
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 47.74468994140625,
+      "learning_rate": 1.4775984764959409e-05,
+      "loss": 3.2204,
+      "step": 2607
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 20.217355728149414,
+      "learning_rate": 1.4773980154355019e-05,
+      "loss": 1.9179,
+      "step": 2608
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 13.135704040527344,
+      "learning_rate": 1.4771975543750627e-05,
+      "loss": 2.2236,
+      "step": 2609
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 10.10952377319336,
+      "learning_rate": 1.4769970933146239e-05,
+      "loss": 1.2541,
+      "step": 2610
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 14.991072654724121,
+      "learning_rate": 1.4767966322541847e-05,
+      "loss": 1.6626,
+      "step": 2611
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 18.55371856689453,
+      "learning_rate": 1.4765961711937457e-05,
+      "loss": 2.59,
+      "step": 2612
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 153.79713439941406,
+      "learning_rate": 1.4763957101333066e-05,
+      "loss": 1.6441,
+      "step": 2613
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 20.791709899902344,
+      "learning_rate": 1.4761952490728678e-05,
+      "loss": 2.2488,
+      "step": 2614
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 26.96784019470215,
+      "learning_rate": 1.4759947880124288e-05,
+      "loss": 2.4947,
+      "step": 2615
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 31.95100975036621,
+      "learning_rate": 1.4757943269519896e-05,
+      "loss": 2.5303,
+      "step": 2616
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 16.064725875854492,
+      "learning_rate": 1.4755938658915508e-05,
+      "loss": 2.2598,
+      "step": 2617
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 19.35765266418457,
+      "learning_rate": 1.4753934048311116e-05,
+      "loss": 2.0552,
+      "step": 2618
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 11.326220512390137,
+      "learning_rate": 1.4751929437706726e-05,
+      "loss": 2.1031,
+      "step": 2619
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 8.919992446899414,
+      "learning_rate": 1.4749924827102338e-05,
+      "loss": 1.8168,
+      "step": 2620
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 14.81724739074707,
+      "learning_rate": 1.4747920216497946e-05,
+      "loss": 1.9064,
+      "step": 2621
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 31.04239845275879,
+      "learning_rate": 1.4745915605893557e-05,
+      "loss": 2.7644,
+      "step": 2622
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 16.15270233154297,
+      "learning_rate": 1.4743910995289165e-05,
+      "loss": 2.0384,
+      "step": 2623
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 20.225746154785156,
+      "learning_rate": 1.4741906384684777e-05,
+      "loss": 2.7575,
+      "step": 2624
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 29.362613677978516,
+      "learning_rate": 1.4739901774080385e-05,
+      "loss": 2.0288,
+      "step": 2625
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 18.233638763427734,
+      "learning_rate": 1.4737897163475995e-05,
+      "loss": 2.4319,
+      "step": 2626
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 15.207660675048828,
+      "learning_rate": 1.4735892552871607e-05,
+      "loss": 2.1672,
+      "step": 2627
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 21.931516647338867,
+      "learning_rate": 1.4733887942267215e-05,
+      "loss": 2.7001,
+      "step": 2628
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 11.918092727661133,
+      "learning_rate": 1.4731883331662825e-05,
+      "loss": 2.1923,
+      "step": 2629
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 24.173839569091797,
+      "learning_rate": 1.4729878721058436e-05,
+      "loss": 2.7606,
+      "step": 2630
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 18.89200782775879,
+      "learning_rate": 1.4727874110454046e-05,
+      "loss": 2.3708,
+      "step": 2631
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 18.636999130249023,
+      "learning_rate": 1.4725869499849654e-05,
+      "loss": 2.3505,
+      "step": 2632
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 24.666654586791992,
+      "learning_rate": 1.4723864889245266e-05,
+      "loss": 2.4304,
+      "step": 2633
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 10.553362846374512,
+      "learning_rate": 1.4721860278640876e-05,
+      "loss": 1.8944,
+      "step": 2634
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 21.21739387512207,
+      "learning_rate": 1.4719855668036484e-05,
+      "loss": 2.6751,
+      "step": 2635
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 38.01958465576172,
+      "learning_rate": 1.4717851057432096e-05,
+      "loss": 1.8805,
+      "step": 2636
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 16.428316116333008,
+      "learning_rate": 1.4715846446827704e-05,
+      "loss": 2.9402,
+      "step": 2637
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 22.54930305480957,
+      "learning_rate": 1.4713841836223315e-05,
+      "loss": 1.9364,
+      "step": 2638
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 14.97050952911377,
+      "learning_rate": 1.4711837225618926e-05,
+      "loss": 1.7904,
+      "step": 2639
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 16.823516845703125,
+      "learning_rate": 1.4709832615014535e-05,
+      "loss": 2.2004,
+      "step": 2640
+    },
+    {
+      "epoch": 0.79,
+      "eval_loss": 0.28386011719703674,
+      "eval_runtime": 44.8703,
+      "eval_samples_per_second": 32.962,
+      "eval_steps_per_second": 32.962,
+      "step": 2640
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 14.531737327575684,
+      "learning_rate": 1.4707828004410145e-05,
+      "loss": 2.1963,
+      "step": 2641
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 24.638132095336914,
+      "learning_rate": 1.4705823393805753e-05,
+      "loss": 2.2464,
+      "step": 2642
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 17.104135513305664,
+      "learning_rate": 1.4703818783201365e-05,
+      "loss": 2.0654,
+      "step": 2643
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 7.9517621994018555,
+      "learning_rate": 1.4701814172596973e-05,
+      "loss": 0.8721,
+      "step": 2644
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 23.127620697021484,
+      "learning_rate": 1.4699809561992583e-05,
+      "loss": 2.0121,
+      "step": 2645
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 18.799203872680664,
+      "learning_rate": 1.4697804951388195e-05,
+      "loss": 2.0454,
+      "step": 2646
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 22.311729431152344,
+      "learning_rate": 1.4695800340783804e-05,
+      "loss": 2.1801,
+      "step": 2647
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 17.162185668945312,
+      "learning_rate": 1.4693795730179414e-05,
+      "loss": 1.814,
+      "step": 2648
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 9.871721267700195,
+      "learning_rate": 1.4691791119575024e-05,
+      "loss": 1.6796,
+      "step": 2649
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 18.128135681152344,
+      "learning_rate": 1.4689786508970634e-05,
+      "loss": 2.3865,
+      "step": 2650
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 15.663517951965332,
+      "learning_rate": 1.4687781898366244e-05,
+      "loss": 1.9406,
+      "step": 2651
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 37.457672119140625,
+      "learning_rate": 1.4685777287761854e-05,
+      "loss": 2.7683,
+      "step": 2652
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 19.439468383789062,
+      "learning_rate": 1.4683772677157464e-05,
+      "loss": 2.354,
+      "step": 2653
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 29.902244567871094,
+      "learning_rate": 1.4681768066553072e-05,
+      "loss": 1.993,
+      "step": 2654
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 11.236685752868652,
+      "learning_rate": 1.4679763455948684e-05,
+      "loss": 1.8682,
+      "step": 2655
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 11.110152244567871,
+      "learning_rate": 1.4677758845344293e-05,
+      "loss": 1.378,
+      "step": 2656
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 15.20004653930664,
+      "learning_rate": 1.4675754234739903e-05,
+      "loss": 2.4361,
+      "step": 2657
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 44.016544342041016,
+      "learning_rate": 1.4673749624135514e-05,
+      "loss": 2.6364,
+      "step": 2658
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 18.134666442871094,
+      "learning_rate": 1.4671745013531123e-05,
+      "loss": 2.1308,
+      "step": 2659
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 83.64218139648438,
+      "learning_rate": 1.4669740402926733e-05,
+      "loss": 3.793,
+      "step": 2660
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 18.634946823120117,
+      "learning_rate": 1.4667735792322341e-05,
+      "loss": 2.0182,
+      "step": 2661
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 46.72085189819336,
+      "learning_rate": 1.4665731181717953e-05,
+      "loss": 2.1095,
+      "step": 2662
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 18.656938552856445,
+      "learning_rate": 1.4663726571113563e-05,
+      "loss": 2.4388,
+      "step": 2663
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 16.998310089111328,
+      "learning_rate": 1.4661721960509172e-05,
+      "loss": 1.9669,
+      "step": 2664
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 25.790451049804688,
+      "learning_rate": 1.4659717349904783e-05,
+      "loss": 2.6492,
+      "step": 2665
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 11.008011817932129,
+      "learning_rate": 1.4657712739300392e-05,
+      "loss": 1.926,
+      "step": 2666
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 8.969502449035645,
+      "learning_rate": 1.4655708128696002e-05,
+      "loss": 0.8433,
+      "step": 2667
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 24.146800994873047,
+      "learning_rate": 1.465370351809161e-05,
+      "loss": 1.7039,
+      "step": 2668
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 32.524452209472656,
+      "learning_rate": 1.4651698907487222e-05,
+      "loss": 2.2074,
+      "step": 2669
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 17.53997802734375,
+      "learning_rate": 1.4649694296882832e-05,
+      "loss": 2.1067,
+      "step": 2670
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 70.99577331542969,
+      "learning_rate": 1.464768968627844e-05,
+      "loss": 2.5416,
+      "step": 2671
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 15.582018852233887,
+      "learning_rate": 1.4645685075674052e-05,
+      "loss": 2.4462,
+      "step": 2672
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 26.792186737060547,
+      "learning_rate": 1.464368046506966e-05,
+      "loss": 1.7403,
+      "step": 2673
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 13.64623737335205,
+      "learning_rate": 1.464167585446527e-05,
+      "loss": 2.7516,
+      "step": 2674
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 17.885374069213867,
+      "learning_rate": 1.463967124386088e-05,
+      "loss": 2.9638,
+      "step": 2675
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 11.002893447875977,
+      "learning_rate": 1.4637666633256491e-05,
+      "loss": 1.5192,
+      "step": 2676
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 18.869712829589844,
+      "learning_rate": 1.4635662022652101e-05,
+      "loss": 2.5806,
+      "step": 2677
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 13.818095207214355,
+      "learning_rate": 1.4633657412047711e-05,
+      "loss": 2.1488,
+      "step": 2678
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 28.199312210083008,
+      "learning_rate": 1.4631652801443321e-05,
+      "loss": 1.8242,
+      "step": 2679
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 19.68671417236328,
+      "learning_rate": 1.462964819083893e-05,
+      "loss": 1.5373,
+      "step": 2680
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 36.305686950683594,
+      "learning_rate": 1.4627643580234541e-05,
+      "loss": 3.0527,
+      "step": 2681
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 22.186141967773438,
+      "learning_rate": 1.4625638969630151e-05,
+      "loss": 3.4701,
+      "step": 2682
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 17.44498634338379,
+      "learning_rate": 1.462363435902576e-05,
+      "loss": 2.0777,
+      "step": 2683
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 25.54764747619629,
+      "learning_rate": 1.4621629748421372e-05,
+      "loss": 1.9166,
+      "step": 2684
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 15.869314193725586,
+      "learning_rate": 1.461962513781698e-05,
+      "loss": 2.0511,
+      "step": 2685
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 16.541519165039062,
+      "learning_rate": 1.461762052721259e-05,
+      "loss": 2.2221,
+      "step": 2686
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 12.096132278442383,
+      "learning_rate": 1.4615615916608198e-05,
+      "loss": 2.2477,
+      "step": 2687
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 32.8545036315918,
+      "learning_rate": 1.461361130600381e-05,
+      "loss": 2.4819,
+      "step": 2688
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 17.078535079956055,
+      "learning_rate": 1.461160669539942e-05,
+      "loss": 2.5493,
+      "step": 2689
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 20.856630325317383,
+      "learning_rate": 1.4609602084795029e-05,
+      "loss": 2.9245,
+      "step": 2690
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 30.489768981933594,
+      "learning_rate": 1.460759747419064e-05,
+      "loss": 2.2884,
+      "step": 2691
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 13.105022430419922,
+      "learning_rate": 1.4605592863586249e-05,
+      "loss": 2.3658,
+      "step": 2692
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 14.78238296508789,
+      "learning_rate": 1.4603588252981859e-05,
+      "loss": 2.5098,
+      "step": 2693
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 17.489654541015625,
+      "learning_rate": 1.460158364237747e-05,
+      "loss": 1.8979,
+      "step": 2694
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 15.65456771850586,
+      "learning_rate": 1.4599579031773079e-05,
+      "loss": 2.2142,
+      "step": 2695
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 24.52695655822754,
+      "learning_rate": 1.459757442116869e-05,
+      "loss": 2.7726,
+      "step": 2696
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 33.155067443847656,
+      "learning_rate": 1.4595569810564298e-05,
+      "loss": 2.2196,
+      "step": 2697
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 9.755550384521484,
+      "learning_rate": 1.459356519995991e-05,
+      "loss": 1.7204,
+      "step": 2698
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 19.83466911315918,
+      "learning_rate": 1.4591560589355518e-05,
+      "loss": 2.5923,
+      "step": 2699
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 21.58110809326172,
+      "learning_rate": 1.458955597875113e-05,
+      "loss": 1.7443,
+      "step": 2700
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 20.373117446899414,
+      "learning_rate": 1.458755136814674e-05,
+      "loss": 2.949,
+      "step": 2701
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 12.08033275604248,
+      "learning_rate": 1.4585546757542348e-05,
+      "loss": 1.7417,
+      "step": 2702
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 16.984411239624023,
+      "learning_rate": 1.458354214693796e-05,
+      "loss": 1.973,
+      "step": 2703
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 24.945363998413086,
+      "learning_rate": 1.4581537536333568e-05,
+      "loss": 1.7648,
+      "step": 2704
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 18.574451446533203,
+      "learning_rate": 1.4579532925729178e-05,
+      "loss": 1.326,
+      "step": 2705
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 32.63951873779297,
+      "learning_rate": 1.457752831512479e-05,
+      "loss": 2.0216,
+      "step": 2706
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 20.653427124023438,
+      "learning_rate": 1.4575523704520398e-05,
+      "loss": 2.0913,
+      "step": 2707
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 41.5728645324707,
+      "learning_rate": 1.4573519093916009e-05,
+      "loss": 3.4579,
+      "step": 2708
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 12.26309585571289,
+      "learning_rate": 1.4571514483311617e-05,
+      "loss": 1.4278,
+      "step": 2709
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 17.044031143188477,
+      "learning_rate": 1.4569509872707229e-05,
+      "loss": 1.7787,
+      "step": 2710
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 22.376020431518555,
+      "learning_rate": 1.4567505262102837e-05,
+      "loss": 2.2045,
+      "step": 2711
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 15.766566276550293,
+      "learning_rate": 1.4565500651498447e-05,
+      "loss": 2.087,
+      "step": 2712
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 10.13654899597168,
+      "learning_rate": 1.4563496040894059e-05,
+      "loss": 2.7784,
+      "step": 2713
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 12.07814884185791,
+      "learning_rate": 1.4561491430289667e-05,
+      "loss": 2.2884,
+      "step": 2714
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 15.864092826843262,
+      "learning_rate": 1.4559486819685277e-05,
+      "loss": 2.1431,
+      "step": 2715
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 18.540136337280273,
+      "learning_rate": 1.4557482209080886e-05,
+      "loss": 2.4553,
+      "step": 2716
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 9.907877922058105,
+      "learning_rate": 1.4555477598476498e-05,
+      "loss": 1.6814,
+      "step": 2717
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 9.542145729064941,
+      "learning_rate": 1.4553472987872106e-05,
+      "loss": 1.1338,
+      "step": 2718
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 18.332304000854492,
+      "learning_rate": 1.4551468377267716e-05,
+      "loss": 1.7549,
+      "step": 2719
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 9.094773292541504,
+      "learning_rate": 1.4549463766663328e-05,
+      "loss": 1.1251,
+      "step": 2720
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 15.948875427246094,
+      "learning_rate": 1.4547459156058936e-05,
+      "loss": 1.5578,
+      "step": 2721
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 13.107511520385742,
+      "learning_rate": 1.4545454545454546e-05,
+      "loss": 1.8671,
+      "step": 2722
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 17.999923706054688,
+      "learning_rate": 1.4543449934850156e-05,
+      "loss": 2.1096,
+      "step": 2723
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 22.923246383666992,
+      "learning_rate": 1.4541445324245767e-05,
+      "loss": 2.1625,
+      "step": 2724
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 12.2493314743042,
+      "learning_rate": 1.4539440713641377e-05,
+      "loss": 2.9069,
+      "step": 2725
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 20.940412521362305,
+      "learning_rate": 1.4537436103036987e-05,
+      "loss": 2.6205,
+      "step": 2726
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 28.031394958496094,
+      "learning_rate": 1.4535431492432597e-05,
+      "loss": 2.167,
+      "step": 2727
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 12.496899604797363,
+      "learning_rate": 1.4533426881828205e-05,
+      "loss": 2.8417,
+      "step": 2728
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 15.856078147888184,
+      "learning_rate": 1.4531422271223817e-05,
+      "loss": 2.5802,
+      "step": 2729
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 19.049468994140625,
+      "learning_rate": 1.4529417660619425e-05,
+      "loss": 2.2789,
+      "step": 2730
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 16.88075828552246,
+      "learning_rate": 1.4527413050015035e-05,
+      "loss": 1.6496,
+      "step": 2731
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 23.054603576660156,
+      "learning_rate": 1.4525408439410647e-05,
+      "loss": 1.8223,
+      "step": 2732
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 27.310646057128906,
+      "learning_rate": 1.4523403828806256e-05,
+      "loss": 2.5239,
+      "step": 2733
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 20.423734664916992,
+      "learning_rate": 1.4521399218201866e-05,
+      "loss": 2.2311,
+      "step": 2734
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 13.929975509643555,
+      "learning_rate": 1.4519394607597474e-05,
+      "loss": 2.5546,
+      "step": 2735
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 12.425309181213379,
+      "learning_rate": 1.4517389996993086e-05,
+      "loss": 1.4081,
+      "step": 2736
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 11.25351333618164,
+      "learning_rate": 1.4515385386388696e-05,
+      "loss": 1.7953,
+      "step": 2737
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 19.53854751586914,
+      "learning_rate": 1.4513380775784304e-05,
+      "loss": 1.207,
+      "step": 2738
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 22.41812515258789,
+      "learning_rate": 1.4511376165179916e-05,
+      "loss": 2.4787,
+      "step": 2739
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 21.234926223754883,
+      "learning_rate": 1.4509371554575524e-05,
+      "loss": 2.2675,
+      "step": 2740
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 19.57403564453125,
+      "learning_rate": 1.4507366943971135e-05,
+      "loss": 2.3916,
+      "step": 2741
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 25.845134735107422,
+      "learning_rate": 1.4505362333366743e-05,
+      "loss": 2.4502,
+      "step": 2742
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 42.33112716674805,
+      "learning_rate": 1.4503357722762355e-05,
+      "loss": 2.0345,
+      "step": 2743
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 15.049954414367676,
+      "learning_rate": 1.4501353112157965e-05,
+      "loss": 2.8193,
+      "step": 2744
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 25.87766456604004,
+      "learning_rate": 1.4499348501553573e-05,
+      "loss": 2.2518,
+      "step": 2745
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 21.152681350708008,
+      "learning_rate": 1.4497343890949185e-05,
+      "loss": 2.3211,
+      "step": 2746
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 36.79154586791992,
+      "learning_rate": 1.4495339280344793e-05,
+      "loss": 2.1298,
+      "step": 2747
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 9.398030281066895,
+      "learning_rate": 1.4493334669740403e-05,
+      "loss": 1.8969,
+      "step": 2748
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 20.946956634521484,
+      "learning_rate": 1.4491330059136015e-05,
+      "loss": 2.3464,
+      "step": 2749
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 9.130115509033203,
+      "learning_rate": 1.4489325448531624e-05,
+      "loss": 1.7633,
+      "step": 2750
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 14.199234962463379,
+      "learning_rate": 1.4487320837927234e-05,
+      "loss": 1.8352,
+      "step": 2751
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 8.78415298461914,
+      "learning_rate": 1.4485316227322844e-05,
+      "loss": 1.6194,
+      "step": 2752
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 32.84090805053711,
+      "learning_rate": 1.4483311616718454e-05,
+      "loss": 1.7946,
+      "step": 2753
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 18.42530632019043,
+      "learning_rate": 1.4481307006114062e-05,
+      "loss": 2.4003,
+      "step": 2754
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 10.56641960144043,
+      "learning_rate": 1.4479302395509674e-05,
+      "loss": 1.698,
+      "step": 2755
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 13.460083961486816,
+      "learning_rate": 1.4477297784905284e-05,
+      "loss": 2.7324,
+      "step": 2756
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 20.335878372192383,
+      "learning_rate": 1.4475293174300893e-05,
+      "loss": 1.5947,
+      "step": 2757
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 44.001834869384766,
+      "learning_rate": 1.4473288563696504e-05,
+      "loss": 1.8942,
+      "step": 2758
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 13.544523239135742,
+      "learning_rate": 1.4471283953092113e-05,
+      "loss": 2.016,
+      "step": 2759
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 30.979524612426758,
+      "learning_rate": 1.4469279342487723e-05,
+      "loss": 3.1479,
+      "step": 2760
+    },
+    {
+      "epoch": 0.83,
+      "eval_loss": 0.2657427191734314,
+      "eval_runtime": 43.5175,
+      "eval_samples_per_second": 33.986,
+      "eval_steps_per_second": 33.986,
+      "step": 2760
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 23.16549301147461,
+      "learning_rate": 1.4467274731883331e-05,
+      "loss": 2.2599,
+      "step": 2761
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 18.973628997802734,
+      "learning_rate": 1.4465270121278943e-05,
+      "loss": 2.1258,
+      "step": 2762
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 26.273101806640625,
+      "learning_rate": 1.4463265510674553e-05,
+      "loss": 3.4268,
+      "step": 2763
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 14.230957984924316,
+      "learning_rate": 1.4461260900070161e-05,
+      "loss": 1.4143,
+      "step": 2764
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 14.468798637390137,
+      "learning_rate": 1.4459256289465773e-05,
+      "loss": 2.3477,
+      "step": 2765
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 16.977825164794922,
+      "learning_rate": 1.4457251678861382e-05,
+      "loss": 2.3843,
+      "step": 2766
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 14.977238655090332,
+      "learning_rate": 1.4455247068256992e-05,
+      "loss": 2.0197,
+      "step": 2767
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 14.926851272583008,
+      "learning_rate": 1.4453242457652603e-05,
+      "loss": 2.9558,
+      "step": 2768
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 23.74448013305664,
+      "learning_rate": 1.4451237847048212e-05,
+      "loss": 2.1472,
+      "step": 2769
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 41.35685729980469,
+      "learning_rate": 1.4449233236443822e-05,
+      "loss": 2.7331,
+      "step": 2770
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 13.470633506774902,
+      "learning_rate": 1.4447228625839432e-05,
+      "loss": 1.9928,
+      "step": 2771
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 12.274412155151367,
+      "learning_rate": 1.4445224015235042e-05,
+      "loss": 1.7747,
+      "step": 2772
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 12.649324417114258,
+      "learning_rate": 1.444321940463065e-05,
+      "loss": 1.599,
+      "step": 2773
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 13.914393424987793,
+      "learning_rate": 1.4441214794026262e-05,
+      "loss": 1.7285,
+      "step": 2774
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 9.780823707580566,
+      "learning_rate": 1.4439210183421872e-05,
+      "loss": 1.3613,
+      "step": 2775
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 13.444343566894531,
+      "learning_rate": 1.443720557281748e-05,
+      "loss": 1.4692,
+      "step": 2776
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 11.597440719604492,
+      "learning_rate": 1.4435200962213093e-05,
+      "loss": 2.5432,
+      "step": 2777
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 66.30996704101562,
+      "learning_rate": 1.4433196351608701e-05,
+      "loss": 1.9169,
+      "step": 2778
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 42.77153778076172,
+      "learning_rate": 1.4431191741004311e-05,
+      "loss": 2.1056,
+      "step": 2779
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 17.99850082397461,
+      "learning_rate": 1.4429187130399923e-05,
+      "loss": 1.6885,
+      "step": 2780
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 11.443147659301758,
+      "learning_rate": 1.4427182519795531e-05,
+      "loss": 2.2285,
+      "step": 2781
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 18.145082473754883,
+      "learning_rate": 1.4425177909191141e-05,
+      "loss": 2.14,
+      "step": 2782
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 12.900897026062012,
+      "learning_rate": 1.442317329858675e-05,
+      "loss": 1.8419,
+      "step": 2783
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 12.71401596069336,
+      "learning_rate": 1.4421168687982361e-05,
+      "loss": 1.6758,
+      "step": 2784
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 10.644526481628418,
+      "learning_rate": 1.441916407737797e-05,
+      "loss": 1.5368,
+      "step": 2785
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 11.878511428833008,
+      "learning_rate": 1.441715946677358e-05,
+      "loss": 2.8906,
+      "step": 2786
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 23.1770076751709,
+      "learning_rate": 1.4415154856169192e-05,
+      "loss": 2.9468,
+      "step": 2787
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 13.855196952819824,
+      "learning_rate": 1.44131502455648e-05,
+      "loss": 2.0071,
+      "step": 2788
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 10.715313911437988,
+      "learning_rate": 1.441114563496041e-05,
+      "loss": 1.3851,
+      "step": 2789
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 11.7968168258667,
+      "learning_rate": 1.4409141024356019e-05,
+      "loss": 1.7609,
+      "step": 2790
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 15.877503395080566,
+      "learning_rate": 1.440713641375163e-05,
+      "loss": 3.124,
+      "step": 2791
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 19.398963928222656,
+      "learning_rate": 1.440513180314724e-05,
+      "loss": 2.2776,
+      "step": 2792
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 13.756909370422363,
+      "learning_rate": 1.4403127192542849e-05,
+      "loss": 2.429,
+      "step": 2793
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 8.20438289642334,
+      "learning_rate": 1.440112258193846e-05,
+      "loss": 1.3021,
+      "step": 2794
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 35.246524810791016,
+      "learning_rate": 1.4399117971334069e-05,
+      "loss": 3.4113,
+      "step": 2795
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 15.750861167907715,
+      "learning_rate": 1.4397113360729679e-05,
+      "loss": 2.1455,
+      "step": 2796
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 12.513425827026367,
+      "learning_rate": 1.4395108750125289e-05,
+      "loss": 2.1875,
+      "step": 2797
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 20.23858070373535,
+      "learning_rate": 1.43931041395209e-05,
+      "loss": 2.686,
+      "step": 2798
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 14.214842796325684,
+      "learning_rate": 1.439109952891651e-05,
+      "loss": 1.5095,
+      "step": 2799
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 13.132567405700684,
+      "learning_rate": 1.438909491831212e-05,
+      "loss": 1.4686,
+      "step": 2800
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 12.941914558410645,
+      "learning_rate": 1.438709030770773e-05,
+      "loss": 1.6349,
+      "step": 2801
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 23.00050926208496,
+      "learning_rate": 1.4385085697103338e-05,
+      "loss": 2.3867,
+      "step": 2802
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 27.743297576904297,
+      "learning_rate": 1.438308108649895e-05,
+      "loss": 2.8904,
+      "step": 2803
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 28.265438079833984,
+      "learning_rate": 1.4381076475894558e-05,
+      "loss": 2.0912,
+      "step": 2804
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 42.51979064941406,
+      "learning_rate": 1.4379071865290168e-05,
+      "loss": 2.9312,
+      "step": 2805
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 13.225244522094727,
+      "learning_rate": 1.437706725468578e-05,
+      "loss": 2.3977,
+      "step": 2806
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 21.915138244628906,
+      "learning_rate": 1.4375062644081388e-05,
+      "loss": 2.3075,
+      "step": 2807
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 27.076114654541016,
+      "learning_rate": 1.4373058033476998e-05,
+      "loss": 2.9827,
+      "step": 2808
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 29.377344131469727,
+      "learning_rate": 1.4371053422872607e-05,
+      "loss": 1.7643,
+      "step": 2809
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 16.541988372802734,
+      "learning_rate": 1.4369048812268219e-05,
+      "loss": 2.0169,
+      "step": 2810
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 11.847968101501465,
+      "learning_rate": 1.4367044201663829e-05,
+      "loss": 1.9688,
+      "step": 2811
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 19.333833694458008,
+      "learning_rate": 1.4365039591059437e-05,
+      "loss": 1.6196,
+      "step": 2812
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 11.35824966430664,
+      "learning_rate": 1.4363034980455049e-05,
+      "loss": 2.2096,
+      "step": 2813
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 21.909666061401367,
+      "learning_rate": 1.4361030369850657e-05,
+      "loss": 2.0165,
+      "step": 2814
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 12.537689208984375,
+      "learning_rate": 1.4359025759246267e-05,
+      "loss": 1.4654,
+      "step": 2815
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 16.89116096496582,
+      "learning_rate": 1.4357021148641876e-05,
+      "loss": 2.2603,
+      "step": 2816
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 29.52833366394043,
+      "learning_rate": 1.4355016538037487e-05,
+      "loss": 2.1104,
+      "step": 2817
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 31.932817459106445,
+      "learning_rate": 1.4353011927433097e-05,
+      "loss": 3.2788,
+      "step": 2818
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 13.009839057922363,
+      "learning_rate": 1.4351007316828708e-05,
+      "loss": 1.8396,
+      "step": 2819
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 20.032115936279297,
+      "learning_rate": 1.4349002706224318e-05,
+      "loss": 2.5377,
+      "step": 2820
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 30.521154403686523,
+      "learning_rate": 1.4346998095619926e-05,
+      "loss": 2.9662,
+      "step": 2821
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 25.22045135498047,
+      "learning_rate": 1.4344993485015538e-05,
+      "loss": 2.4709,
+      "step": 2822
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 36.31182861328125,
+      "learning_rate": 1.4342988874411148e-05,
+      "loss": 2.1305,
+      "step": 2823
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 10.13243293762207,
+      "learning_rate": 1.4340984263806756e-05,
+      "loss": 1.2709,
+      "step": 2824
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 13.83365535736084,
+      "learning_rate": 1.4338979653202368e-05,
+      "loss": 2.4132,
+      "step": 2825
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 13.130005836486816,
+      "learning_rate": 1.4336975042597976e-05,
+      "loss": 1.4074,
+      "step": 2826
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 20.482786178588867,
+      "learning_rate": 1.4334970431993587e-05,
+      "loss": 2.3476,
+      "step": 2827
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 9.617236137390137,
+      "learning_rate": 1.4332965821389195e-05,
+      "loss": 1.595,
+      "step": 2828
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 11.414641380310059,
+      "learning_rate": 1.4330961210784807e-05,
+      "loss": 1.5817,
+      "step": 2829
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 16.799848556518555,
+      "learning_rate": 1.4328956600180417e-05,
+      "loss": 2.1956,
+      "step": 2830
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 10.216323852539062,
+      "learning_rate": 1.4326951989576025e-05,
+      "loss": 1.4594,
+      "step": 2831
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 28.910261154174805,
+      "learning_rate": 1.4324947378971637e-05,
+      "loss": 2.556,
+      "step": 2832
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 26.716176986694336,
+      "learning_rate": 1.4322942768367245e-05,
+      "loss": 2.3864,
+      "step": 2833
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 16.635345458984375,
+      "learning_rate": 1.4320938157762855e-05,
+      "loss": 2.3808,
+      "step": 2834
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 21.195531845092773,
+      "learning_rate": 1.4318933547158467e-05,
+      "loss": 1.7784,
+      "step": 2835
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 16.338502883911133,
+      "learning_rate": 1.4316928936554076e-05,
+      "loss": 2.1817,
+      "step": 2836
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 10.698309898376465,
+      "learning_rate": 1.4314924325949686e-05,
+      "loss": 1.8357,
+      "step": 2837
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 22.48438262939453,
+      "learning_rate": 1.4312919715345294e-05,
+      "loss": 1.9495,
+      "step": 2838
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 72.1753921508789,
+      "learning_rate": 1.4310915104740906e-05,
+      "loss": 1.8956,
+      "step": 2839
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 14.804839134216309,
+      "learning_rate": 1.4308910494136514e-05,
+      "loss": 2.0498,
+      "step": 2840
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 46.27671432495117,
+      "learning_rate": 1.4306905883532124e-05,
+      "loss": 2.3295,
+      "step": 2841
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 20.91038703918457,
+      "learning_rate": 1.4304901272927736e-05,
+      "loss": 1.7985,
+      "step": 2842
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 15.907576560974121,
+      "learning_rate": 1.4302896662323345e-05,
+      "loss": 1.945,
+      "step": 2843
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 11.249744415283203,
+      "learning_rate": 1.4300892051718955e-05,
+      "loss": 2.0269,
+      "step": 2844
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 31.293607711791992,
+      "learning_rate": 1.4298887441114565e-05,
+      "loss": 2.0211,
+      "step": 2845
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 15.472600936889648,
+      "learning_rate": 1.4296882830510175e-05,
+      "loss": 2.2907,
+      "step": 2846
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 15.497085571289062,
+      "learning_rate": 1.4294878219905783e-05,
+      "loss": 2.4152,
+      "step": 2847
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 16.131122589111328,
+      "learning_rate": 1.4292873609301395e-05,
+      "loss": 1.7781,
+      "step": 2848
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 20.631099700927734,
+      "learning_rate": 1.4290868998697005e-05,
+      "loss": 2.6782,
+      "step": 2849
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 28.31133460998535,
+      "learning_rate": 1.4288864388092613e-05,
+      "loss": 2.7064,
+      "step": 2850
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 22.287992477416992,
+      "learning_rate": 1.4286859777488225e-05,
+      "loss": 1.5085,
+      "step": 2851
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 14.099394798278809,
+      "learning_rate": 1.4284855166883834e-05,
+      "loss": 2.1791,
+      "step": 2852
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 23.74030876159668,
+      "learning_rate": 1.4282850556279444e-05,
+      "loss": 1.5725,
+      "step": 2853
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 18.97103500366211,
+      "learning_rate": 1.4280845945675055e-05,
+      "loss": 2.8069,
+      "step": 2854
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 15.047412872314453,
+      "learning_rate": 1.4278841335070664e-05,
+      "loss": 2.3392,
+      "step": 2855
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 11.085140228271484,
+      "learning_rate": 1.4276836724466274e-05,
+      "loss": 1.8397,
+      "step": 2856
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 25.895097732543945,
+      "learning_rate": 1.4274832113861882e-05,
+      "loss": 2.0397,
+      "step": 2857
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 15.016706466674805,
+      "learning_rate": 1.4272827503257494e-05,
+      "loss": 2.7215,
+      "step": 2858
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 30.092390060424805,
+      "learning_rate": 1.4270822892653102e-05,
+      "loss": 2.8657,
+      "step": 2859
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 16.425628662109375,
+      "learning_rate": 1.4268818282048713e-05,
+      "loss": 1.7927,
+      "step": 2860
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 39.504478454589844,
+      "learning_rate": 1.4266813671444324e-05,
+      "loss": 2.4321,
+      "step": 2861
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 15.988898277282715,
+      "learning_rate": 1.4264809060839933e-05,
+      "loss": 3.2789,
+      "step": 2862
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 25.57854461669922,
+      "learning_rate": 1.4262804450235543e-05,
+      "loss": 2.2735,
+      "step": 2863
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 24.288047790527344,
+      "learning_rate": 1.4260799839631151e-05,
+      "loss": 2.4726,
+      "step": 2864
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 9.727755546569824,
+      "learning_rate": 1.4258795229026763e-05,
+      "loss": 1.2214,
+      "step": 2865
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 11.615317344665527,
+      "learning_rate": 1.4256790618422373e-05,
+      "loss": 1.5758,
+      "step": 2866
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 18.63646697998047,
+      "learning_rate": 1.4254786007817981e-05,
+      "loss": 3.0492,
+      "step": 2867
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 18.21438217163086,
+      "learning_rate": 1.4252781397213593e-05,
+      "loss": 2.2775,
+      "step": 2868
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 11.930097579956055,
+      "learning_rate": 1.4250776786609202e-05,
+      "loss": 1.8362,
+      "step": 2869
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 18.075613021850586,
+      "learning_rate": 1.4248772176004812e-05,
+      "loss": 3.2771,
+      "step": 2870
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 17.12186050415039,
+      "learning_rate": 1.4246767565400422e-05,
+      "loss": 1.9388,
+      "step": 2871
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 20.949424743652344,
+      "learning_rate": 1.4244762954796032e-05,
+      "loss": 2.5683,
+      "step": 2872
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 36.75067901611328,
+      "learning_rate": 1.4242758344191644e-05,
+      "loss": 1.8602,
+      "step": 2873
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 12.314196586608887,
+      "learning_rate": 1.4240753733587252e-05,
+      "loss": 1.7533,
+      "step": 2874
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 9.40181827545166,
+      "learning_rate": 1.4238749122982862e-05,
+      "loss": 1.0538,
+      "step": 2875
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 18.72924041748047,
+      "learning_rate": 1.423674451237847e-05,
+      "loss": 1.741,
+      "step": 2876
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 86.12344360351562,
+      "learning_rate": 1.4234739901774082e-05,
+      "loss": 1.8543,
+      "step": 2877
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 11.64871883392334,
+      "learning_rate": 1.4232735291169692e-05,
+      "loss": 1.8377,
+      "step": 2878
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 20.858604431152344,
+      "learning_rate": 1.42307306805653e-05,
+      "loss": 2.407,
+      "step": 2879
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 16.45133399963379,
+      "learning_rate": 1.4228726069960913e-05,
+      "loss": 2.1642,
+      "step": 2880
+    },
+    {
+      "epoch": 0.87,
+      "eval_loss": 0.27140170335769653,
+      "eval_runtime": 43.4249,
+      "eval_samples_per_second": 34.059,
+      "eval_steps_per_second": 34.059,
+      "step": 2880
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 16.983951568603516,
+      "learning_rate": 1.4226721459356521e-05,
+      "loss": 2.1911,
+      "step": 2881
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 12.091941833496094,
+      "learning_rate": 1.4224716848752131e-05,
+      "loss": 1.6567,
+      "step": 2882
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 22.655092239379883,
+      "learning_rate": 1.422271223814774e-05,
+      "loss": 1.7178,
+      "step": 2883
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 13.33409595489502,
+      "learning_rate": 1.4220707627543351e-05,
+      "loss": 1.42,
+      "step": 2884
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 41.79966735839844,
+      "learning_rate": 1.4218703016938961e-05,
+      "loss": 2.5635,
+      "step": 2885
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 9.73848819732666,
+      "learning_rate": 1.421669840633457e-05,
+      "loss": 1.5657,
+      "step": 2886
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 13.572857856750488,
+      "learning_rate": 1.4214693795730181e-05,
+      "loss": 1.8109,
+      "step": 2887
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 19.471956253051758,
+      "learning_rate": 1.421268918512579e-05,
+      "loss": 1.7335,
+      "step": 2888
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 16.479246139526367,
+      "learning_rate": 1.42106845745214e-05,
+      "loss": 2.1827,
+      "step": 2889
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 42.60647964477539,
+      "learning_rate": 1.420867996391701e-05,
+      "loss": 4.2229,
+      "step": 2890
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 29.324525833129883,
+      "learning_rate": 1.420667535331262e-05,
+      "loss": 2.2616,
+      "step": 2891
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 17.365251541137695,
+      "learning_rate": 1.420467074270823e-05,
+      "loss": 1.9783,
+      "step": 2892
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 14.282939910888672,
+      "learning_rate": 1.420266613210384e-05,
+      "loss": 1.9463,
+      "step": 2893
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 18.807802200317383,
+      "learning_rate": 1.420066152149945e-05,
+      "loss": 1.9243,
+      "step": 2894
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 12.65291976928711,
+      "learning_rate": 1.4198656910895059e-05,
+      "loss": 1.5824,
+      "step": 2895
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 24.225215911865234,
+      "learning_rate": 1.419665230029067e-05,
+      "loss": 1.4894,
+      "step": 2896
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 19.319774627685547,
+      "learning_rate": 1.419464768968628e-05,
+      "loss": 2.8796,
+      "step": 2897
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 14.177779197692871,
+      "learning_rate": 1.4192643079081889e-05,
+      "loss": 1.7554,
+      "step": 2898
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 17.083988189697266,
+      "learning_rate": 1.41906384684775e-05,
+      "loss": 2.0681,
+      "step": 2899
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 40.30973434448242,
+      "learning_rate": 1.418863385787311e-05,
+      "loss": 2.4183,
+      "step": 2900
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 9.512898445129395,
+      "learning_rate": 1.418662924726872e-05,
+      "loss": 1.5707,
+      "step": 2901
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 52.70304870605469,
+      "learning_rate": 1.4184624636664328e-05,
+      "loss": 2.9803,
+      "step": 2902
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 16.734039306640625,
+      "learning_rate": 1.418262002605994e-05,
+      "loss": 2.037,
+      "step": 2903
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 19.740087509155273,
+      "learning_rate": 1.418061541545555e-05,
+      "loss": 1.8485,
+      "step": 2904
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 14.386117935180664,
+      "learning_rate": 1.4178610804851158e-05,
+      "loss": 1.0815,
+      "step": 2905
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 16.817588806152344,
+      "learning_rate": 1.417660619424677e-05,
+      "loss": 1.6195,
+      "step": 2906
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 27.766712188720703,
+      "learning_rate": 1.4174601583642378e-05,
+      "loss": 1.9603,
+      "step": 2907
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 33.24577713012695,
+      "learning_rate": 1.4172596973037988e-05,
+      "loss": 1.6909,
+      "step": 2908
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 11.028932571411133,
+      "learning_rate": 1.41705923624336e-05,
+      "loss": 1.3208,
+      "step": 2909
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 12.61074447631836,
+      "learning_rate": 1.4168587751829208e-05,
+      "loss": 1.9588,
+      "step": 2910
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 11.678028106689453,
+      "learning_rate": 1.4166583141224818e-05,
+      "loss": 0.8482,
+      "step": 2911
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 12.001538276672363,
+      "learning_rate": 1.4164578530620427e-05,
+      "loss": 1.9327,
+      "step": 2912
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 16.76682472229004,
+      "learning_rate": 1.4162573920016039e-05,
+      "loss": 2.0711,
+      "step": 2913
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 27.799612045288086,
+      "learning_rate": 1.4160569309411647e-05,
+      "loss": 2.1891,
+      "step": 2914
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 24.618627548217773,
+      "learning_rate": 1.4158564698807257e-05,
+      "loss": 1.8592,
+      "step": 2915
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 24.141916275024414,
+      "learning_rate": 1.4156560088202869e-05,
+      "loss": 2.2455,
+      "step": 2916
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 10.506229400634766,
+      "learning_rate": 1.4154555477598477e-05,
+      "loss": 1.8823,
+      "step": 2917
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 14.979269981384277,
+      "learning_rate": 1.4152550866994087e-05,
+      "loss": 2.0007,
+      "step": 2918
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 34.168128967285156,
+      "learning_rate": 1.4150546256389697e-05,
+      "loss": 2.2599,
+      "step": 2919
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 17.805347442626953,
+      "learning_rate": 1.4148541645785307e-05,
+      "loss": 2.8172,
+      "step": 2920
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 18.875598907470703,
+      "learning_rate": 1.4146537035180918e-05,
+      "loss": 1.7194,
+      "step": 2921
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 44.858428955078125,
+      "learning_rate": 1.4144532424576528e-05,
+      "loss": 1.6975,
+      "step": 2922
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 10.500758171081543,
+      "learning_rate": 1.4142527813972138e-05,
+      "loss": 1.8214,
+      "step": 2923
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 19.17152214050293,
+      "learning_rate": 1.4140523203367746e-05,
+      "loss": 1.9509,
+      "step": 2924
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 19.903955459594727,
+      "learning_rate": 1.4138518592763358e-05,
+      "loss": 2.9421,
+      "step": 2925
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 17.388938903808594,
+      "learning_rate": 1.4136513982158966e-05,
+      "loss": 1.8245,
+      "step": 2926
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 15.076804161071777,
+      "learning_rate": 1.4134509371554576e-05,
+      "loss": 1.5359,
+      "step": 2927
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 11.603935241699219,
+      "learning_rate": 1.4132504760950188e-05,
+      "loss": 1.5563,
+      "step": 2928
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 25.9620418548584,
+      "learning_rate": 1.4130500150345797e-05,
+      "loss": 3.2783,
+      "step": 2929
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 16.795143127441406,
+      "learning_rate": 1.4128495539741407e-05,
+      "loss": 1.5015,
+      "step": 2930
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 11.010591506958008,
+      "learning_rate": 1.4126490929137015e-05,
+      "loss": 1.8603,
+      "step": 2931
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 12.306628227233887,
+      "learning_rate": 1.4124486318532627e-05,
+      "loss": 1.4446,
+      "step": 2932
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 14.734245300292969,
+      "learning_rate": 1.4122481707928235e-05,
+      "loss": 2.5827,
+      "step": 2933
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 23.01751708984375,
+      "learning_rate": 1.4120477097323845e-05,
+      "loss": 2.9274,
+      "step": 2934
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 20.132675170898438,
+      "learning_rate": 1.4118472486719457e-05,
+      "loss": 2.0589,
+      "step": 2935
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 17.275482177734375,
+      "learning_rate": 1.4116467876115065e-05,
+      "loss": 2.1773,
+      "step": 2936
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 40.08796310424805,
+      "learning_rate": 1.4114463265510675e-05,
+      "loss": 2.3652,
+      "step": 2937
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 22.47896957397461,
+      "learning_rate": 1.4112458654906286e-05,
+      "loss": 2.069,
+      "step": 2938
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 21.368188858032227,
+      "learning_rate": 1.4110454044301896e-05,
+      "loss": 2.5254,
+      "step": 2939
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 8.838847160339355,
+      "learning_rate": 1.4108449433697506e-05,
+      "loss": 1.4187,
+      "step": 2940
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 19.025794982910156,
+      "learning_rate": 1.4106444823093116e-05,
+      "loss": 1.7403,
+      "step": 2941
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 26.310054779052734,
+      "learning_rate": 1.4104440212488726e-05,
+      "loss": 2.0741,
+      "step": 2942
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 14.412237167358398,
+      "learning_rate": 1.4102435601884334e-05,
+      "loss": 1.6891,
+      "step": 2943
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 15.92914867401123,
+      "learning_rate": 1.4100430991279946e-05,
+      "loss": 2.542,
+      "step": 2944
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 26.77935028076172,
+      "learning_rate": 1.4098426380675554e-05,
+      "loss": 3.2923,
+      "step": 2945
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 61.19276428222656,
+      "learning_rate": 1.4096421770071165e-05,
+      "loss": 2.9397,
+      "step": 2946
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 15.064374923706055,
+      "learning_rate": 1.4094417159466776e-05,
+      "loss": 1.757,
+      "step": 2947
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 14.561102867126465,
+      "learning_rate": 1.4092412548862385e-05,
+      "loss": 1.7057,
+      "step": 2948
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 17.088661193847656,
+      "learning_rate": 1.4090407938257995e-05,
+      "loss": 2.2723,
+      "step": 2949
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 7.7882256507873535,
+      "learning_rate": 1.4088403327653603e-05,
+      "loss": 1.8397,
+      "step": 2950
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 11.358787536621094,
+      "learning_rate": 1.4086398717049215e-05,
+      "loss": 2.2978,
+      "step": 2951
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 28.74677276611328,
+      "learning_rate": 1.4084394106444825e-05,
+      "loss": 2.7206,
+      "step": 2952
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 16.747051239013672,
+      "learning_rate": 1.4082389495840433e-05,
+      "loss": 1.9734,
+      "step": 2953
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 12.262884140014648,
+      "learning_rate": 1.4080384885236045e-05,
+      "loss": 2.0553,
+      "step": 2954
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 20.407411575317383,
+      "learning_rate": 1.4078380274631654e-05,
+      "loss": 1.3774,
+      "step": 2955
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 25.87983512878418,
+      "learning_rate": 1.4076375664027264e-05,
+      "loss": 2.7685,
+      "step": 2956
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 14.667790412902832,
+      "learning_rate": 1.4074371053422872e-05,
+      "loss": 2.1229,
+      "step": 2957
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 10.724139213562012,
+      "learning_rate": 1.4072366442818484e-05,
+      "loss": 1.6101,
+      "step": 2958
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 25.648866653442383,
+      "learning_rate": 1.4070361832214094e-05,
+      "loss": 2.4417,
+      "step": 2959
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 24.464935302734375,
+      "learning_rate": 1.4068357221609702e-05,
+      "loss": 1.071,
+      "step": 2960
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 22.294677734375,
+      "learning_rate": 1.4066352611005314e-05,
+      "loss": 2.0512,
+      "step": 2961
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 20.227296829223633,
+      "learning_rate": 1.4064348000400923e-05,
+      "loss": 1.4506,
+      "step": 2962
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 35.15935134887695,
+      "learning_rate": 1.4062343389796533e-05,
+      "loss": 2.224,
+      "step": 2963
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 15.452341079711914,
+      "learning_rate": 1.4060338779192144e-05,
+      "loss": 3.0408,
+      "step": 2964
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 13.116132736206055,
+      "learning_rate": 1.4058334168587753e-05,
+      "loss": 2.6705,
+      "step": 2965
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 22.434358596801758,
+      "learning_rate": 1.4056329557983363e-05,
+      "loss": 2.6559,
+      "step": 2966
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 27.88434600830078,
+      "learning_rate": 1.4054324947378973e-05,
+      "loss": 2.2654,
+      "step": 2967
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 29.59733772277832,
+      "learning_rate": 1.4052320336774583e-05,
+      "loss": 1.7697,
+      "step": 2968
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 27.069503784179688,
+      "learning_rate": 1.4050315726170191e-05,
+      "loss": 2.1905,
+      "step": 2969
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 29.36873435974121,
+      "learning_rate": 1.4048311115565803e-05,
+      "loss": 2.3816,
+      "step": 2970
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 10.980319023132324,
+      "learning_rate": 1.4046306504961413e-05,
+      "loss": 1.6791,
+      "step": 2971
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 13.023073196411133,
+      "learning_rate": 1.4044301894357022e-05,
+      "loss": 1.9126,
+      "step": 2972
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 38.787940979003906,
+      "learning_rate": 1.4042297283752633e-05,
+      "loss": 2.6036,
+      "step": 2973
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 29.560956954956055,
+      "learning_rate": 1.4040292673148242e-05,
+      "loss": 2.8441,
+      "step": 2974
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 8.834044456481934,
+      "learning_rate": 1.4038288062543852e-05,
+      "loss": 1.4787,
+      "step": 2975
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 11.226600646972656,
+      "learning_rate": 1.403628345193946e-05,
+      "loss": 1.5387,
+      "step": 2976
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 16.720735549926758,
+      "learning_rate": 1.4034278841335072e-05,
+      "loss": 2.2926,
+      "step": 2977
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 20.5881290435791,
+      "learning_rate": 1.4032274230730682e-05,
+      "loss": 2.5574,
+      "step": 2978
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 31.438135147094727,
+      "learning_rate": 1.403026962012629e-05,
+      "loss": 1.8076,
+      "step": 2979
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 12.068266868591309,
+      "learning_rate": 1.4028265009521902e-05,
+      "loss": 1.2816,
+      "step": 2980
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 25.184410095214844,
+      "learning_rate": 1.402626039891751e-05,
+      "loss": 2.4355,
+      "step": 2981
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 11.62043285369873,
+      "learning_rate": 1.402425578831312e-05,
+      "loss": 1.7278,
+      "step": 2982
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 10.73161506652832,
+      "learning_rate": 1.4022251177708733e-05,
+      "loss": 1.5009,
+      "step": 2983
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 18.191015243530273,
+      "learning_rate": 1.4020246567104341e-05,
+      "loss": 1.7434,
+      "step": 2984
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 35.12432861328125,
+      "learning_rate": 1.4018241956499951e-05,
+      "loss": 1.6909,
+      "step": 2985
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 11.526291847229004,
+      "learning_rate": 1.401623734589556e-05,
+      "loss": 1.4617,
+      "step": 2986
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 61.24599075317383,
+      "learning_rate": 1.4014232735291171e-05,
+      "loss": 1.9207,
+      "step": 2987
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 21.56028175354004,
+      "learning_rate": 1.401222812468678e-05,
+      "loss": 2.4295,
+      "step": 2988
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 15.971736907958984,
+      "learning_rate": 1.4010223514082391e-05,
+      "loss": 1.9247,
+      "step": 2989
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 18.692066192626953,
+      "learning_rate": 1.4008218903478001e-05,
+      "loss": 2.1351,
+      "step": 2990
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 19.446582794189453,
+      "learning_rate": 1.400621429287361e-05,
+      "loss": 2.2271,
+      "step": 2991
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 14.894234657287598,
+      "learning_rate": 1.4004209682269222e-05,
+      "loss": 2.0351,
+      "step": 2992
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 10.384711265563965,
+      "learning_rate": 1.400220507166483e-05,
+      "loss": 1.3777,
+      "step": 2993
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 15.35462760925293,
+      "learning_rate": 1.400020046106044e-05,
+      "loss": 1.4528,
+      "step": 2994
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 9.53518009185791,
+      "learning_rate": 1.3998195850456052e-05,
+      "loss": 2.0396,
+      "step": 2995
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 12.702194213867188,
+      "learning_rate": 1.399619123985166e-05,
+      "loss": 3.5201,
+      "step": 2996
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 16.689231872558594,
+      "learning_rate": 1.399418662924727e-05,
+      "loss": 2.5204,
+      "step": 2997
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 16.812484741210938,
+      "learning_rate": 1.3992182018642879e-05,
+      "loss": 1.823,
+      "step": 2998
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 27.859556198120117,
+      "learning_rate": 1.399017740803849e-05,
+      "loss": 2.3145,
+      "step": 2999
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 40.26885223388672,
+      "learning_rate": 1.3988172797434099e-05,
+      "loss": 1.981,
+      "step": 3000
+    },
+    {
+      "epoch": 0.9,
+      "eval_loss": 0.290047824382782,
+      "eval_runtime": 43.5045,
+      "eval_samples_per_second": 33.996,
+      "eval_steps_per_second": 33.996,
+      "step": 3000
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 16.576269149780273,
+      "learning_rate": 1.3986168186829709e-05,
+      "loss": 1.7576,
+      "step": 3001
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 11.136350631713867,
+      "learning_rate": 1.398416357622532e-05,
+      "loss": 1.1301,
+      "step": 3002
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 7.8271684646606445,
+      "learning_rate": 1.398215896562093e-05,
+      "loss": 1.6797,
+      "step": 3003
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 18.726409912109375,
+      "learning_rate": 1.398015435501654e-05,
+      "loss": 2.3383,
+      "step": 3004
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 13.888818740844727,
+      "learning_rate": 1.3978149744412148e-05,
+      "loss": 1.7,
+      "step": 3005
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 17.36934471130371,
+      "learning_rate": 1.397614513380776e-05,
+      "loss": 1.9836,
+      "step": 3006
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 34.03765869140625,
+      "learning_rate": 1.3974140523203368e-05,
+      "loss": 2.0497,
+      "step": 3007
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 24.520832061767578,
+      "learning_rate": 1.3972135912598978e-05,
+      "loss": 3.1885,
+      "step": 3008
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 13.99000072479248,
+      "learning_rate": 1.397013130199459e-05,
+      "loss": 2.3549,
+      "step": 3009
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 14.597472190856934,
+      "learning_rate": 1.3968126691390198e-05,
+      "loss": 2.2443,
+      "step": 3010
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 10.168909072875977,
+      "learning_rate": 1.3966122080785808e-05,
+      "loss": 2.1224,
+      "step": 3011
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 21.07402229309082,
+      "learning_rate": 1.3964117470181418e-05,
+      "loss": 2.2544,
+      "step": 3012
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 35.29275894165039,
+      "learning_rate": 1.3962112859577028e-05,
+      "loss": 4.147,
+      "step": 3013
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 19.924930572509766,
+      "learning_rate": 1.3960108248972638e-05,
+      "loss": 2.3426,
+      "step": 3014
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 18.175373077392578,
+      "learning_rate": 1.3958103638368249e-05,
+      "loss": 1.6238,
+      "step": 3015
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 16.749662399291992,
+      "learning_rate": 1.3956099027763859e-05,
+      "loss": 1.9444,
+      "step": 3016
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 32.66911697387695,
+      "learning_rate": 1.3954094417159467e-05,
+      "loss": 2.5848,
+      "step": 3017
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 10.309561729431152,
+      "learning_rate": 1.3952089806555079e-05,
+      "loss": 1.2661,
+      "step": 3018
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 32.622840881347656,
+      "learning_rate": 1.3950085195950687e-05,
+      "loss": 3.1314,
+      "step": 3019
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 51.04716491699219,
+      "learning_rate": 1.3948080585346297e-05,
+      "loss": 2.5912,
+      "step": 3020
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 9.933348655700684,
+      "learning_rate": 1.3946075974741909e-05,
+      "loss": 1.6906,
+      "step": 3021
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 11.241128921508789,
+      "learning_rate": 1.3944071364137517e-05,
+      "loss": 2.0141,
+      "step": 3022
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 20.985105514526367,
+      "learning_rate": 1.3942066753533127e-05,
+      "loss": 2.3962,
+      "step": 3023
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 14.399240493774414,
+      "learning_rate": 1.3940062142928736e-05,
+      "loss": 2.4056,
+      "step": 3024
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 23.836753845214844,
+      "learning_rate": 1.3938057532324348e-05,
+      "loss": 1.492,
+      "step": 3025
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 18.37982940673828,
+      "learning_rate": 1.3936052921719958e-05,
+      "loss": 1.8183,
+      "step": 3026
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 13.537169456481934,
+      "learning_rate": 1.3934048311115566e-05,
+      "loss": 2.1759,
+      "step": 3027
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 23.467910766601562,
+      "learning_rate": 1.3932043700511178e-05,
+      "loss": 2.6208,
+      "step": 3028
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 17.59671401977539,
+      "learning_rate": 1.3930039089906786e-05,
+      "loss": 2.3887,
+      "step": 3029
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 12.384592056274414,
+      "learning_rate": 1.3928034479302396e-05,
+      "loss": 2.0166,
+      "step": 3030
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 17.471967697143555,
+      "learning_rate": 1.3926029868698005e-05,
+      "loss": 1.875,
+      "step": 3031
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 12.046953201293945,
+      "learning_rate": 1.3924025258093617e-05,
+      "loss": 1.3814,
+      "step": 3032
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 15.179215431213379,
+      "learning_rate": 1.3922020647489227e-05,
+      "loss": 1.6944,
+      "step": 3033
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 10.748503684997559,
+      "learning_rate": 1.3920016036884835e-05,
+      "loss": 2.1539,
+      "step": 3034
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 20.441787719726562,
+      "learning_rate": 1.3918011426280447e-05,
+      "loss": 2.9749,
+      "step": 3035
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 23.096572875976562,
+      "learning_rate": 1.3916006815676055e-05,
+      "loss": 1.9606,
+      "step": 3036
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 27.173370361328125,
+      "learning_rate": 1.3914002205071665e-05,
+      "loss": 2.5916,
+      "step": 3037
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 11.456189155578613,
+      "learning_rate": 1.3911997594467277e-05,
+      "loss": 1.7153,
+      "step": 3038
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 36.13789367675781,
+      "learning_rate": 1.3909992983862885e-05,
+      "loss": 2.2955,
+      "step": 3039
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 25.46799659729004,
+      "learning_rate": 1.3907988373258496e-05,
+      "loss": 2.2377,
+      "step": 3040
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 18.01449203491211,
+      "learning_rate": 1.3905983762654106e-05,
+      "loss": 2.5831,
+      "step": 3041
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 25.396291732788086,
+      "learning_rate": 1.3903979152049716e-05,
+      "loss": 2.2237,
+      "step": 3042
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 15.773884773254395,
+      "learning_rate": 1.3901974541445324e-05,
+      "loss": 2.6421,
+      "step": 3043
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 16.961776733398438,
+      "learning_rate": 1.3899969930840936e-05,
+      "loss": 1.8037,
+      "step": 3044
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 46.810585021972656,
+      "learning_rate": 1.3897965320236546e-05,
+      "loss": 1.4078,
+      "step": 3045
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 19.385047912597656,
+      "learning_rate": 1.3895960709632154e-05,
+      "loss": 1.9391,
+      "step": 3046
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 14.272333145141602,
+      "learning_rate": 1.3893956099027766e-05,
+      "loss": 1.7094,
+      "step": 3047
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 12.611112594604492,
+      "learning_rate": 1.3891951488423375e-05,
+      "loss": 2.0959,
+      "step": 3048
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 17.307199478149414,
+      "learning_rate": 1.3889946877818985e-05,
+      "loss": 2.1561,
+      "step": 3049
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 20.245845794677734,
+      "learning_rate": 1.3887942267214593e-05,
+      "loss": 2.8128,
+      "step": 3050
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 17.531375885009766,
+      "learning_rate": 1.3885937656610205e-05,
+      "loss": 1.6429,
+      "step": 3051
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 12.451351165771484,
+      "learning_rate": 1.3883933046005815e-05,
+      "loss": 2.0644,
+      "step": 3052
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 16.824329376220703,
+      "learning_rate": 1.3881928435401423e-05,
+      "loss": 2.1751,
+      "step": 3053
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 19.493457794189453,
+      "learning_rate": 1.3879923824797035e-05,
+      "loss": 1.8751,
+      "step": 3054
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 16.869348526000977,
+      "learning_rate": 1.3877919214192643e-05,
+      "loss": 1.9817,
+      "step": 3055
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 29.737041473388672,
+      "learning_rate": 1.3875914603588254e-05,
+      "loss": 2.4454,
+      "step": 3056
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 22.09736442565918,
+      "learning_rate": 1.3873909992983865e-05,
+      "loss": 1.7898,
+      "step": 3057
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 17.416990280151367,
+      "learning_rate": 1.3871905382379474e-05,
+      "loss": 2.9486,
+      "step": 3058
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 13.20331859588623,
+      "learning_rate": 1.3869900771775084e-05,
+      "loss": 2.2757,
+      "step": 3059
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 7.922008514404297,
+      "learning_rate": 1.3867896161170694e-05,
+      "loss": 2.0063,
+      "step": 3060
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 14.735453605651855,
+      "learning_rate": 1.3865891550566304e-05,
+      "loss": 2.4153,
+      "step": 3061
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 13.794514656066895,
+      "learning_rate": 1.3863886939961912e-05,
+      "loss": 1.8875,
+      "step": 3062
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 11.291130065917969,
+      "learning_rate": 1.3861882329357524e-05,
+      "loss": 1.3319,
+      "step": 3063
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 19.629484176635742,
+      "learning_rate": 1.3859877718753134e-05,
+      "loss": 1.4707,
+      "step": 3064
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 15.42160415649414,
+      "learning_rate": 1.3857873108148743e-05,
+      "loss": 1.5667,
+      "step": 3065
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 14.26404094696045,
+      "learning_rate": 1.3855868497544354e-05,
+      "loss": 2.0599,
+      "step": 3066
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 8.3535795211792,
+      "learning_rate": 1.3853863886939963e-05,
+      "loss": 1.3769,
+      "step": 3067
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 13.800082206726074,
+      "learning_rate": 1.3851859276335573e-05,
+      "loss": 1.5491,
+      "step": 3068
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 16.2960262298584,
+      "learning_rate": 1.3849854665731185e-05,
+      "loss": 1.8461,
+      "step": 3069
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 12.411531448364258,
+      "learning_rate": 1.3847850055126793e-05,
+      "loss": 2.1916,
+      "step": 3070
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 10.797046661376953,
+      "learning_rate": 1.3845845444522403e-05,
+      "loss": 1.2897,
+      "step": 3071
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 17.1668643951416,
+      "learning_rate": 1.3843840833918011e-05,
+      "loss": 2.0636,
+      "step": 3072
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 12.581513404846191,
+      "learning_rate": 1.3841836223313623e-05,
+      "loss": 2.0915,
+      "step": 3073
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 18.048904418945312,
+      "learning_rate": 1.3839831612709232e-05,
+      "loss": 1.8037,
+      "step": 3074
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 25.745256423950195,
+      "learning_rate": 1.3837827002104842e-05,
+      "loss": 2.4617,
+      "step": 3075
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 23.290611267089844,
+      "learning_rate": 1.3835822391500453e-05,
+      "loss": 2.3306,
+      "step": 3076
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 18.817665100097656,
+      "learning_rate": 1.3833817780896062e-05,
+      "loss": 2.0089,
+      "step": 3077
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 16.671045303344727,
+      "learning_rate": 1.3831813170291672e-05,
+      "loss": 2.0306,
+      "step": 3078
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 17.590496063232422,
+      "learning_rate": 1.382980855968728e-05,
+      "loss": 2.5062,
+      "step": 3079
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 15.332666397094727,
+      "learning_rate": 1.3827803949082892e-05,
+      "loss": 1.5195,
+      "step": 3080
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 28.62594985961914,
+      "learning_rate": 1.3825799338478502e-05,
+      "loss": 2.3762,
+      "step": 3081
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 15.981874465942383,
+      "learning_rate": 1.382379472787411e-05,
+      "loss": 2.0642,
+      "step": 3082
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 16.672204971313477,
+      "learning_rate": 1.3821790117269722e-05,
+      "loss": 2.0906,
+      "step": 3083
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 15.016446113586426,
+      "learning_rate": 1.381978550666533e-05,
+      "loss": 1.7059,
+      "step": 3084
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 9.689031600952148,
+      "learning_rate": 1.3817780896060941e-05,
+      "loss": 1.625,
+      "step": 3085
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 8.496919631958008,
+      "learning_rate": 1.3815776285456551e-05,
+      "loss": 0.9668,
+      "step": 3086
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 20.545669555664062,
+      "learning_rate": 1.3813771674852161e-05,
+      "loss": 1.827,
+      "step": 3087
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 14.81933307647705,
+      "learning_rate": 1.3811767064247771e-05,
+      "loss": 2.3411,
+      "step": 3088
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 11.812262535095215,
+      "learning_rate": 1.3809762453643381e-05,
+      "loss": 3.2352,
+      "step": 3089
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 50.63083267211914,
+      "learning_rate": 1.3807757843038991e-05,
+      "loss": 1.491,
+      "step": 3090
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 17.194198608398438,
+      "learning_rate": 1.38057532324346e-05,
+      "loss": 1.45,
+      "step": 3091
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 25.95490074157715,
+      "learning_rate": 1.3803748621830211e-05,
+      "loss": 3.7476,
+      "step": 3092
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 12.458473205566406,
+      "learning_rate": 1.380174401122582e-05,
+      "loss": 2.0199,
+      "step": 3093
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 17.711647033691406,
+      "learning_rate": 1.379973940062143e-05,
+      "loss": 2.1909,
+      "step": 3094
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 21.15656852722168,
+      "learning_rate": 1.3797734790017042e-05,
+      "loss": 2.6897,
+      "step": 3095
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 14.197000503540039,
+      "learning_rate": 1.379573017941265e-05,
+      "loss": 2.5904,
+      "step": 3096
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 24.11342430114746,
+      "learning_rate": 1.379372556880826e-05,
+      "loss": 2.4864,
+      "step": 3097
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 18.00290298461914,
+      "learning_rate": 1.3791720958203869e-05,
+      "loss": 2.6676,
+      "step": 3098
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 10.474432945251465,
+      "learning_rate": 1.378971634759948e-05,
+      "loss": 1.9275,
+      "step": 3099
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 15.15256118774414,
+      "learning_rate": 1.378771173699509e-05,
+      "loss": 2.0671,
+      "step": 3100
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 16.475624084472656,
+      "learning_rate": 1.3785707126390699e-05,
+      "loss": 2.206,
+      "step": 3101
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 32.260719299316406,
+      "learning_rate": 1.378370251578631e-05,
+      "loss": 1.7881,
+      "step": 3102
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 15.899197578430176,
+      "learning_rate": 1.3781697905181919e-05,
+      "loss": 1.865,
+      "step": 3103
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 16.820289611816406,
+      "learning_rate": 1.3779693294577529e-05,
+      "loss": 1.9125,
+      "step": 3104
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 24.38895606994629,
+      "learning_rate": 1.3777688683973137e-05,
+      "loss": 1.5985,
+      "step": 3105
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 34.085914611816406,
+      "learning_rate": 1.377568407336875e-05,
+      "loss": 2.2492,
+      "step": 3106
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 14.358930587768555,
+      "learning_rate": 1.377367946276436e-05,
+      "loss": 1.9382,
+      "step": 3107
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 25.895401000976562,
+      "learning_rate": 1.377167485215997e-05,
+      "loss": 1.1291,
+      "step": 3108
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 17.190685272216797,
+      "learning_rate": 1.376967024155558e-05,
+      "loss": 1.4162,
+      "step": 3109
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 13.299299240112305,
+      "learning_rate": 1.3767665630951188e-05,
+      "loss": 1.6711,
+      "step": 3110
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 14.426549911499023,
+      "learning_rate": 1.37656610203468e-05,
+      "loss": 1.9699,
+      "step": 3111
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 25.35927963256836,
+      "learning_rate": 1.376365640974241e-05,
+      "loss": 2.6309,
+      "step": 3112
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 9.705031394958496,
+      "learning_rate": 1.3761651799138018e-05,
+      "loss": 2.4938,
+      "step": 3113
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 12.233379364013672,
+      "learning_rate": 1.375964718853363e-05,
+      "loss": 1.8221,
+      "step": 3114
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 15.394095420837402,
+      "learning_rate": 1.3757642577929238e-05,
+      "loss": 1.7775,
+      "step": 3115
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 13.07430362701416,
+      "learning_rate": 1.3755637967324848e-05,
+      "loss": 1.0702,
+      "step": 3116
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 12.21866226196289,
+      "learning_rate": 1.3753633356720457e-05,
+      "loss": 1.6868,
+      "step": 3117
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 10.357650756835938,
+      "learning_rate": 1.3751628746116069e-05,
+      "loss": 2.2551,
+      "step": 3118
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 18.206022262573242,
+      "learning_rate": 1.3749624135511679e-05,
+      "loss": 1.8297,
+      "step": 3119
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 15.580822944641113,
+      "learning_rate": 1.3747619524907287e-05,
+      "loss": 1.9925,
+      "step": 3120
+    },
+    {
+      "epoch": 0.94,
+      "eval_loss": 0.2486942857503891,
+      "eval_runtime": 43.5686,
+      "eval_samples_per_second": 33.946,
+      "eval_steps_per_second": 33.946,
+      "step": 3120
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 10.770172119140625,
+      "learning_rate": 1.3745614914302899e-05,
+      "loss": 2.3822,
+      "step": 3121
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 25.76287078857422,
+      "learning_rate": 1.3743610303698507e-05,
+      "loss": 2.8554,
+      "step": 3122
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 11.431665420532227,
+      "learning_rate": 1.3741605693094117e-05,
+      "loss": 2.0408,
+      "step": 3123
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 12.141244888305664,
+      "learning_rate": 1.3739601082489729e-05,
+      "loss": 1.5182,
+      "step": 3124
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 12.981203079223633,
+      "learning_rate": 1.3737596471885337e-05,
+      "loss": 1.3628,
+      "step": 3125
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 12.477705955505371,
+      "learning_rate": 1.3735591861280948e-05,
+      "loss": 1.9388,
+      "step": 3126
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 21.606971740722656,
+      "learning_rate": 1.3733587250676556e-05,
+      "loss": 1.7253,
+      "step": 3127
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 15.517189025878906,
+      "learning_rate": 1.3731582640072168e-05,
+      "loss": 2.0146,
+      "step": 3128
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 30.338869094848633,
+      "learning_rate": 1.3729578029467776e-05,
+      "loss": 1.8816,
+      "step": 3129
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 34.77288818359375,
+      "learning_rate": 1.3727573418863386e-05,
+      "loss": 3.4102,
+      "step": 3130
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 29.494319915771484,
+      "learning_rate": 1.3725568808258998e-05,
+      "loss": 1.3876,
+      "step": 3131
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 15.238880157470703,
+      "learning_rate": 1.3723564197654606e-05,
+      "loss": 2.0381,
+      "step": 3132
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 13.510786056518555,
+      "learning_rate": 1.3721559587050216e-05,
+      "loss": 2.5507,
+      "step": 3133
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 24.444250106811523,
+      "learning_rate": 1.3719554976445827e-05,
+      "loss": 1.7889,
+      "step": 3134
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 12.126347541809082,
+      "learning_rate": 1.3717550365841437e-05,
+      "loss": 1.2326,
+      "step": 3135
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 10.488152503967285,
+      "learning_rate": 1.3715545755237045e-05,
+      "loss": 1.2664,
+      "step": 3136
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 30.025480270385742,
+      "learning_rate": 1.3713541144632657e-05,
+      "loss": 1.5859,
+      "step": 3137
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 13.164788246154785,
+      "learning_rate": 1.3711536534028267e-05,
+      "loss": 2.1662,
+      "step": 3138
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 18.30267333984375,
+      "learning_rate": 1.3709531923423875e-05,
+      "loss": 2.1693,
+      "step": 3139
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 16.636390686035156,
+      "learning_rate": 1.3707527312819487e-05,
+      "loss": 2.3369,
+      "step": 3140
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 13.153757095336914,
+      "learning_rate": 1.3705522702215095e-05,
+      "loss": 2.3091,
+      "step": 3141
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 30.390827178955078,
+      "learning_rate": 1.3703518091610706e-05,
+      "loss": 1.786,
+      "step": 3142
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 41.70283889770508,
+      "learning_rate": 1.3701513481006317e-05,
+      "loss": 3.5995,
+      "step": 3143
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 32.86235809326172,
+      "learning_rate": 1.3699508870401926e-05,
+      "loss": 2.4132,
+      "step": 3144
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 14.464203834533691,
+      "learning_rate": 1.3697504259797536e-05,
+      "loss": 2.0061,
+      "step": 3145
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 11.750015258789062,
+      "learning_rate": 1.3695499649193144e-05,
+      "loss": 2.1382,
+      "step": 3146
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 13.602632522583008,
+      "learning_rate": 1.3693495038588756e-05,
+      "loss": 2.4135,
+      "step": 3147
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 15.727476119995117,
+      "learning_rate": 1.3691490427984364e-05,
+      "loss": 2.9052,
+      "step": 3148
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 17.43680763244629,
+      "learning_rate": 1.3689485817379974e-05,
+      "loss": 1.3355,
+      "step": 3149
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 14.329429626464844,
+      "learning_rate": 1.3687481206775586e-05,
+      "loss": 3.1391,
+      "step": 3150
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 25.729965209960938,
+      "learning_rate": 1.3685476596171195e-05,
+      "loss": 1.6923,
+      "step": 3151
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 95.43783569335938,
+      "learning_rate": 1.3683471985566805e-05,
+      "loss": 2.0273,
+      "step": 3152
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 17.75636100769043,
+      "learning_rate": 1.3681467374962413e-05,
+      "loss": 1.5341,
+      "step": 3153
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 19.18214988708496,
+      "learning_rate": 1.3679462764358025e-05,
+      "loss": 1.8777,
+      "step": 3154
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 15.892659187316895,
+      "learning_rate": 1.3677458153753635e-05,
+      "loss": 1.9435,
+      "step": 3155
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 20.043859481811523,
+      "learning_rate": 1.3675453543149243e-05,
+      "loss": 2.1202,
+      "step": 3156
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 13.109265327453613,
+      "learning_rate": 1.3673448932544855e-05,
+      "loss": 2.429,
+      "step": 3157
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 15.493414878845215,
+      "learning_rate": 1.3671444321940463e-05,
+      "loss": 1.768,
+      "step": 3158
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 36.695735931396484,
+      "learning_rate": 1.3669439711336074e-05,
+      "loss": 2.0781,
+      "step": 3159
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 14.086923599243164,
+      "learning_rate": 1.3667435100731684e-05,
+      "loss": 0.8197,
+      "step": 3160
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 13.209424018859863,
+      "learning_rate": 1.3665430490127294e-05,
+      "loss": 1.9733,
+      "step": 3161
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 17.077037811279297,
+      "learning_rate": 1.3663425879522905e-05,
+      "loss": 2.5923,
+      "step": 3162
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 49.46474075317383,
+      "learning_rate": 1.3661421268918514e-05,
+      "loss": 2.0692,
+      "step": 3163
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 14.314860343933105,
+      "learning_rate": 1.3659416658314124e-05,
+      "loss": 2.8229,
+      "step": 3164
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 8.489893913269043,
+      "learning_rate": 1.3657412047709732e-05,
+      "loss": 1.3168,
+      "step": 3165
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 32.04702377319336,
+      "learning_rate": 1.3655407437105344e-05,
+      "loss": 1.6409,
+      "step": 3166
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 16.088096618652344,
+      "learning_rate": 1.3653402826500954e-05,
+      "loss": 1.1645,
+      "step": 3167
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 12.829047203063965,
+      "learning_rate": 1.3651398215896563e-05,
+      "loss": 1.3659,
+      "step": 3168
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 19.811403274536133,
+      "learning_rate": 1.3649393605292174e-05,
+      "loss": 2.6096,
+      "step": 3169
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 15.218836784362793,
+      "learning_rate": 1.3647388994687783e-05,
+      "loss": 1.4327,
+      "step": 3170
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 11.305002212524414,
+      "learning_rate": 1.3645384384083393e-05,
+      "loss": 1.1141,
+      "step": 3171
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 12.872888565063477,
+      "learning_rate": 1.3643379773479001e-05,
+      "loss": 1.6382,
+      "step": 3172
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 18.41777801513672,
+      "learning_rate": 1.3641375162874613e-05,
+      "loss": 2.1711,
+      "step": 3173
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 14.97694206237793,
+      "learning_rate": 1.3639370552270223e-05,
+      "loss": 2.1335,
+      "step": 3174
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 17.031770706176758,
+      "learning_rate": 1.3637365941665832e-05,
+      "loss": 2.1181,
+      "step": 3175
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 15.743291854858398,
+      "learning_rate": 1.3635361331061443e-05,
+      "loss": 2.4674,
+      "step": 3176
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 9.414546966552734,
+      "learning_rate": 1.3633356720457052e-05,
+      "loss": 2.1942,
+      "step": 3177
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 10.637990951538086,
+      "learning_rate": 1.3631352109852662e-05,
+      "loss": 1.5859,
+      "step": 3178
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 13.900732040405273,
+      "learning_rate": 1.3629347499248272e-05,
+      "loss": 1.6751,
+      "step": 3179
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 20.666290283203125,
+      "learning_rate": 1.3627342888643882e-05,
+      "loss": 2.6414,
+      "step": 3180
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 35.015647888183594,
+      "learning_rate": 1.3625338278039492e-05,
+      "loss": 2.7838,
+      "step": 3181
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 15.191834449768066,
+      "learning_rate": 1.3623333667435102e-05,
+      "loss": 1.831,
+      "step": 3182
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 34.37846755981445,
+      "learning_rate": 1.3621329056830712e-05,
+      "loss": 2.7531,
+      "step": 3183
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 12.335997581481934,
+      "learning_rate": 1.361932444622632e-05,
+      "loss": 1.0529,
+      "step": 3184
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 25.120954513549805,
+      "learning_rate": 1.3617319835621932e-05,
+      "loss": 2.2635,
+      "step": 3185
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 11.803996086120605,
+      "learning_rate": 1.3615315225017542e-05,
+      "loss": 1.8602,
+      "step": 3186
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 14.749687194824219,
+      "learning_rate": 1.361331061441315e-05,
+      "loss": 1.0415,
+      "step": 3187
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 15.90971851348877,
+      "learning_rate": 1.3611306003808763e-05,
+      "loss": 1.9429,
+      "step": 3188
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 42.85434341430664,
+      "learning_rate": 1.3609301393204371e-05,
+      "loss": 2.2651,
+      "step": 3189
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 22.296979904174805,
+      "learning_rate": 1.3607296782599981e-05,
+      "loss": 2.2514,
+      "step": 3190
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 16.979684829711914,
+      "learning_rate": 1.360529217199559e-05,
+      "loss": 1.9144,
+      "step": 3191
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 51.25554656982422,
+      "learning_rate": 1.3603287561391201e-05,
+      "loss": 1.4266,
+      "step": 3192
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 32.474002838134766,
+      "learning_rate": 1.3601282950786811e-05,
+      "loss": 2.3617,
+      "step": 3193
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 12.489779472351074,
+      "learning_rate": 1.359927834018242e-05,
+      "loss": 1.4284,
+      "step": 3194
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 14.795775413513184,
+      "learning_rate": 1.3597273729578032e-05,
+      "loss": 0.9304,
+      "step": 3195
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 35.45854568481445,
+      "learning_rate": 1.359526911897364e-05,
+      "loss": 3.7488,
+      "step": 3196
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 6.524545669555664,
+      "learning_rate": 1.359326450836925e-05,
+      "loss": 1.08,
+      "step": 3197
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 13.13879680633545,
+      "learning_rate": 1.3591259897764862e-05,
+      "loss": 2.1342,
+      "step": 3198
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 8.252395629882812,
+      "learning_rate": 1.358925528716047e-05,
+      "loss": 1.6381,
+      "step": 3199
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 12.134532928466797,
+      "learning_rate": 1.358725067655608e-05,
+      "loss": 1.7298,
+      "step": 3200
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 17.264564514160156,
+      "learning_rate": 1.3585246065951689e-05,
+      "loss": 1.6213,
+      "step": 3201
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 17.299379348754883,
+      "learning_rate": 1.35832414553473e-05,
+      "loss": 1.9645,
+      "step": 3202
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 9.386181831359863,
+      "learning_rate": 1.3581236844742909e-05,
+      "loss": 1.0434,
+      "step": 3203
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 10.785475730895996,
+      "learning_rate": 1.3579232234138519e-05,
+      "loss": 1.387,
+      "step": 3204
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 18.141578674316406,
+      "learning_rate": 1.357722762353413e-05,
+      "loss": 1.816,
+      "step": 3205
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 10.100055694580078,
+      "learning_rate": 1.3575223012929739e-05,
+      "loss": 2.0753,
+      "step": 3206
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 13.844127655029297,
+      "learning_rate": 1.3573218402325349e-05,
+      "loss": 1.4975,
+      "step": 3207
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 13.211896896362305,
+      "learning_rate": 1.357121379172096e-05,
+      "loss": 2.011,
+      "step": 3208
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 16.434125900268555,
+      "learning_rate": 1.356920918111657e-05,
+      "loss": 1.3007,
+      "step": 3209
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 25.20293426513672,
+      "learning_rate": 1.356720457051218e-05,
+      "loss": 1.4808,
+      "step": 3210
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 17.422204971313477,
+      "learning_rate": 1.356519995990779e-05,
+      "loss": 2.0575,
+      "step": 3211
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 22.5146484375,
+      "learning_rate": 1.35631953493034e-05,
+      "loss": 2.0943,
+      "step": 3212
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 46.014678955078125,
+      "learning_rate": 1.3561190738699008e-05,
+      "loss": 2.7204,
+      "step": 3213
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 16.03790283203125,
+      "learning_rate": 1.355918612809462e-05,
+      "loss": 2.5307,
+      "step": 3214
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 17.003280639648438,
+      "learning_rate": 1.3557181517490228e-05,
+      "loss": 2.766,
+      "step": 3215
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 21.383115768432617,
+      "learning_rate": 1.3555176906885838e-05,
+      "loss": 1.8888,
+      "step": 3216
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 29.028945922851562,
+      "learning_rate": 1.355317229628145e-05,
+      "loss": 1.2596,
+      "step": 3217
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 14.80164623260498,
+      "learning_rate": 1.3551167685677058e-05,
+      "loss": 2.1869,
+      "step": 3218
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 35.633094787597656,
+      "learning_rate": 1.3549163075072668e-05,
+      "loss": 3.2196,
+      "step": 3219
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 17.463098526000977,
+      "learning_rate": 1.3547158464468277e-05,
+      "loss": 2.077,
+      "step": 3220
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 15.604183197021484,
+      "learning_rate": 1.3545153853863889e-05,
+      "loss": 1.6204,
+      "step": 3221
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 28.1143798828125,
+      "learning_rate": 1.3543149243259497e-05,
+      "loss": 2.9363,
+      "step": 3222
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 18.119604110717773,
+      "learning_rate": 1.3541144632655107e-05,
+      "loss": 1.3793,
+      "step": 3223
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 16.877845764160156,
+      "learning_rate": 1.3539140022050719e-05,
+      "loss": 2.0204,
+      "step": 3224
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 20.121807098388672,
+      "learning_rate": 1.3537135411446327e-05,
+      "loss": 2.4456,
+      "step": 3225
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 15.562324523925781,
+      "learning_rate": 1.3535130800841937e-05,
+      "loss": 1.8663,
+      "step": 3226
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 16.265233993530273,
+      "learning_rate": 1.3533126190237547e-05,
+      "loss": 1.7686,
+      "step": 3227
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 24.303163528442383,
+      "learning_rate": 1.3531121579633158e-05,
+      "loss": 2.5792,
+      "step": 3228
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 19.46893310546875,
+      "learning_rate": 1.3529116969028768e-05,
+      "loss": 1.8132,
+      "step": 3229
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 39.980438232421875,
+      "learning_rate": 1.3527112358424378e-05,
+      "loss": 2.3918,
+      "step": 3230
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 18.492206573486328,
+      "learning_rate": 1.3525107747819988e-05,
+      "loss": 1.6302,
+      "step": 3231
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 11.828949928283691,
+      "learning_rate": 1.3523103137215596e-05,
+      "loss": 1.7631,
+      "step": 3232
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 13.994562149047852,
+      "learning_rate": 1.3521098526611208e-05,
+      "loss": 1.9207,
+      "step": 3233
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 22.90715789794922,
+      "learning_rate": 1.3519093916006816e-05,
+      "loss": 3.6763,
+      "step": 3234
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 45.03823471069336,
+      "learning_rate": 1.3517089305402426e-05,
+      "loss": 2.1425,
+      "step": 3235
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 12.193424224853516,
+      "learning_rate": 1.3515084694798038e-05,
+      "loss": 1.4887,
+      "step": 3236
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 21.530868530273438,
+      "learning_rate": 1.3513080084193647e-05,
+      "loss": 2.9298,
+      "step": 3237
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 30.308269500732422,
+      "learning_rate": 1.3511075473589257e-05,
+      "loss": 2.4928,
+      "step": 3238
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 11.13912582397461,
+      "learning_rate": 1.3509070862984865e-05,
+      "loss": 1.7033,
+      "step": 3239
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 11.577181816101074,
+      "learning_rate": 1.3507066252380477e-05,
+      "loss": 1.4173,
+      "step": 3240
+    },
+    {
+      "epoch": 0.97,
+      "eval_loss": 0.22535113990306854,
+      "eval_runtime": 43.4738,
+      "eval_samples_per_second": 34.021,
+      "eval_steps_per_second": 34.021,
+      "step": 3240
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 27.40611457824707,
+      "learning_rate": 1.3505061641776087e-05,
+      "loss": 1.9135,
+      "step": 3241
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 19.149837493896484,
+      "learning_rate": 1.3503057031171695e-05,
+      "loss": 1.7921,
+      "step": 3242
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 15.974568367004395,
+      "learning_rate": 1.3501052420567307e-05,
+      "loss": 2.5266,
+      "step": 3243
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 19.187986373901367,
+      "learning_rate": 1.3499047809962915e-05,
+      "loss": 1.9754,
+      "step": 3244
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 10.218606948852539,
+      "learning_rate": 1.3497043199358526e-05,
+      "loss": 0.9435,
+      "step": 3245
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 16.653217315673828,
+      "learning_rate": 1.3495038588754134e-05,
+      "loss": 1.8804,
+      "step": 3246
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 19.304424285888672,
+      "learning_rate": 1.3493033978149746e-05,
+      "loss": 2.038,
+      "step": 3247
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 17.04471206665039,
+      "learning_rate": 1.3491029367545356e-05,
+      "loss": 1.9324,
+      "step": 3248
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 16.119550704956055,
+      "learning_rate": 1.3489024756940964e-05,
+      "loss": 2.1206,
+      "step": 3249
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 16.321861267089844,
+      "learning_rate": 1.3487020146336576e-05,
+      "loss": 2.3402,
+      "step": 3250
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 30.2368221282959,
+      "learning_rate": 1.3485015535732184e-05,
+      "loss": 1.683,
+      "step": 3251
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 18.461002349853516,
+      "learning_rate": 1.3483010925127794e-05,
+      "loss": 2.1962,
+      "step": 3252
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 20.721721649169922,
+      "learning_rate": 1.3481006314523406e-05,
+      "loss": 2.1013,
+      "step": 3253
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 31.358562469482422,
+      "learning_rate": 1.3479001703919015e-05,
+      "loss": 2.7242,
+      "step": 3254
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 20.12774085998535,
+      "learning_rate": 1.3476997093314625e-05,
+      "loss": 1.5558,
+      "step": 3255
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 28.31780242919922,
+      "learning_rate": 1.3474992482710235e-05,
+      "loss": 2.1186,
+      "step": 3256
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 11.361303329467773,
+      "learning_rate": 1.3472987872105845e-05,
+      "loss": 1.4945,
+      "step": 3257
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 112.40892028808594,
+      "learning_rate": 1.3470983261501453e-05,
+      "loss": 1.87,
+      "step": 3258
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 48.792320251464844,
+      "learning_rate": 1.3468978650897065e-05,
+      "loss": 2.829,
+      "step": 3259
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 15.994609832763672,
+      "learning_rate": 1.3466974040292675e-05,
+      "loss": 2.33,
+      "step": 3260
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 14.680255889892578,
+      "learning_rate": 1.3464969429688284e-05,
+      "loss": 1.5227,
+      "step": 3261
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 20.413484573364258,
+      "learning_rate": 1.3462964819083895e-05,
+      "loss": 1.3934,
+      "step": 3262
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 10.785961151123047,
+      "learning_rate": 1.3460960208479504e-05,
+      "loss": 1.6991,
+      "step": 3263
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 23.43621063232422,
+      "learning_rate": 1.3458955597875114e-05,
+      "loss": 2.1189,
+      "step": 3264
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 18.55811882019043,
+      "learning_rate": 1.3456950987270722e-05,
+      "loss": 1.6837,
+      "step": 3265
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 25.324520111083984,
+      "learning_rate": 1.3454946376666334e-05,
+      "loss": 2.5444,
+      "step": 3266
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 29.4528751373291,
+      "learning_rate": 1.3452941766061944e-05,
+      "loss": 1.7111,
+      "step": 3267
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 10.598390579223633,
+      "learning_rate": 1.3450937155457552e-05,
+      "loss": 1.1832,
+      "step": 3268
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 16.23030662536621,
+      "learning_rate": 1.3448932544853164e-05,
+      "loss": 1.845,
+      "step": 3269
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 16.992780685424805,
+      "learning_rate": 1.3446927934248773e-05,
+      "loss": 1.5488,
+      "step": 3270
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 16.57977294921875,
+      "learning_rate": 1.3444923323644383e-05,
+      "loss": 2.4538,
+      "step": 3271
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 17.394916534423828,
+      "learning_rate": 1.3442918713039994e-05,
+      "loss": 2.4952,
+      "step": 3272
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 10.907772064208984,
+      "learning_rate": 1.3440914102435603e-05,
+      "loss": 2.2693,
+      "step": 3273
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 15.381023406982422,
+      "learning_rate": 1.3438909491831213e-05,
+      "loss": 1.9541,
+      "step": 3274
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 13.303698539733887,
+      "learning_rate": 1.3436904881226821e-05,
+      "loss": 1.7353,
+      "step": 3275
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 22.471403121948242,
+      "learning_rate": 1.3434900270622433e-05,
+      "loss": 1.8707,
+      "step": 3276
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 12.483773231506348,
+      "learning_rate": 1.3432895660018041e-05,
+      "loss": 1.862,
+      "step": 3277
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 24.261032104492188,
+      "learning_rate": 1.3430891049413652e-05,
+      "loss": 3.0355,
+      "step": 3278
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 27.064979553222656,
+      "learning_rate": 1.3428886438809263e-05,
+      "loss": 2.6927,
+      "step": 3279
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 23.447118759155273,
+      "learning_rate": 1.3426881828204872e-05,
+      "loss": 2.2596,
+      "step": 3280
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 19.78642463684082,
+      "learning_rate": 1.3424877217600484e-05,
+      "loss": 2.1217,
+      "step": 3281
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 11.758055686950684,
+      "learning_rate": 1.3422872606996092e-05,
+      "loss": 1.0749,
+      "step": 3282
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 15.738198280334473,
+      "learning_rate": 1.3420867996391702e-05,
+      "loss": 2.25,
+      "step": 3283
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 21.461877822875977,
+      "learning_rate": 1.3418863385787314e-05,
+      "loss": 2.1378,
+      "step": 3284
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 39.54304504394531,
+      "learning_rate": 1.3416858775182922e-05,
+      "loss": 3.1118,
+      "step": 3285
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 14.904977798461914,
+      "learning_rate": 1.3414854164578532e-05,
+      "loss": 1.2001,
+      "step": 3286
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 19.5516414642334,
+      "learning_rate": 1.341284955397414e-05,
+      "loss": 2.2941,
+      "step": 3287
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 18.030540466308594,
+      "learning_rate": 1.3410844943369752e-05,
+      "loss": 2.1409,
+      "step": 3288
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 17.5920352935791,
+      "learning_rate": 1.340884033276536e-05,
+      "loss": 3.0416,
+      "step": 3289
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 18.530750274658203,
+      "learning_rate": 1.3406835722160971e-05,
+      "loss": 1.1762,
+      "step": 3290
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 15.765362739562988,
+      "learning_rate": 1.3404831111556583e-05,
+      "loss": 2.0459,
+      "step": 3291
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 27.75279998779297,
+      "learning_rate": 1.3402826500952191e-05,
+      "loss": 2.1783,
+      "step": 3292
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 16.64308738708496,
+      "learning_rate": 1.3400821890347801e-05,
+      "loss": 2.4156,
+      "step": 3293
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 19.890602111816406,
+      "learning_rate": 1.339881727974341e-05,
+      "loss": 1.0157,
+      "step": 3294
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 15.657976150512695,
+      "learning_rate": 1.3396812669139021e-05,
+      "loss": 1.3915,
+      "step": 3295
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 14.942137718200684,
+      "learning_rate": 1.3394808058534631e-05,
+      "loss": 2.0334,
+      "step": 3296
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 10.705955505371094,
+      "learning_rate": 1.339280344793024e-05,
+      "loss": 1.7803,
+      "step": 3297
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 11.542771339416504,
+      "learning_rate": 1.3390798837325852e-05,
+      "loss": 1.2149,
+      "step": 3298
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 16.316530227661133,
+      "learning_rate": 1.338879422672146e-05,
+      "loss": 2.4369,
+      "step": 3299
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 23.6247615814209,
+      "learning_rate": 1.338678961611707e-05,
+      "loss": 2.1207,
+      "step": 3300
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 17.557518005371094,
+      "learning_rate": 1.338478500551268e-05,
+      "loss": 1.5633,
+      "step": 3301
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 13.458683967590332,
+      "learning_rate": 1.338278039490829e-05,
+      "loss": 2.0549,
+      "step": 3302
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 10.493450164794922,
+      "learning_rate": 1.33807757843039e-05,
+      "loss": 2.0356,
+      "step": 3303
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 16.16809844970703,
+      "learning_rate": 1.337877117369951e-05,
+      "loss": 1.918,
+      "step": 3304
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 15.41076374053955,
+      "learning_rate": 1.337676656309512e-05,
+      "loss": 2.8471,
+      "step": 3305
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 18.323535919189453,
+      "learning_rate": 1.3374761952490729e-05,
+      "loss": 1.6805,
+      "step": 3306
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 10.888114929199219,
+      "learning_rate": 1.337275734188634e-05,
+      "loss": 1.4534,
+      "step": 3307
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 23.824207305908203,
+      "learning_rate": 1.3370752731281949e-05,
+      "loss": 2.1954,
+      "step": 3308
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 12.056889533996582,
+      "learning_rate": 1.3368748120677559e-05,
+      "loss": 1.9769,
+      "step": 3309
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 20.176801681518555,
+      "learning_rate": 1.3366743510073171e-05,
+      "loss": 2.3137,
+      "step": 3310
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 12.740705490112305,
+      "learning_rate": 1.336473889946878e-05,
+      "loss": 2.1326,
+      "step": 3311
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 13.5363130569458,
+      "learning_rate": 1.336273428886439e-05,
+      "loss": 1.5278,
+      "step": 3312
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 11.068243026733398,
+      "learning_rate": 1.3360729678259998e-05,
+      "loss": 1.186,
+      "step": 3313
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 16.938138961791992,
+      "learning_rate": 1.335872506765561e-05,
+      "loss": 1.0064,
+      "step": 3314
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 13.62263298034668,
+      "learning_rate": 1.335672045705122e-05,
+      "loss": 1.909,
+      "step": 3315
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 11.283188819885254,
+      "learning_rate": 1.3354715846446828e-05,
+      "loss": 1.6886,
+      "step": 3316
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 14.703868865966797,
+      "learning_rate": 1.335271123584244e-05,
+      "loss": 1.9791,
+      "step": 3317
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 13.806490898132324,
+      "learning_rate": 1.3350706625238048e-05,
+      "loss": 1.516,
+      "step": 3318
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 14.787043571472168,
+      "learning_rate": 1.3348702014633658e-05,
+      "loss": 1.8667,
+      "step": 3319
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 17.8718318939209,
+      "learning_rate": 1.3346697404029267e-05,
+      "loss": 2.0014,
+      "step": 3320
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 23.68500518798828,
+      "learning_rate": 1.3344692793424878e-05,
+      "loss": 1.7045,
+      "step": 3321
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 10.535503387451172,
+      "learning_rate": 1.3342688182820488e-05,
+      "loss": 1.183,
+      "step": 3322
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 25.851089477539062,
+      "learning_rate": 1.3340683572216097e-05,
+      "loss": 2.7104,
+      "step": 3323
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 10.395030975341797,
+      "learning_rate": 1.3338678961611709e-05,
+      "loss": 1.7227,
+      "step": 3324
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 14.099095344543457,
+      "learning_rate": 1.3336674351007317e-05,
+      "loss": 1.7941,
+      "step": 3325
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 15.447168350219727,
+      "learning_rate": 1.3334669740402927e-05,
+      "loss": 2.0854,
+      "step": 3326
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 20.244766235351562,
+      "learning_rate": 1.3332665129798539e-05,
+      "loss": 3.0187,
+      "step": 3327
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 56.30267333984375,
+      "learning_rate": 1.3330660519194147e-05,
+      "loss": 2.1849,
+      "step": 3328
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 28.362627029418945,
+      "learning_rate": 1.3328655908589757e-05,
+      "loss": 1.7998,
+      "step": 3329
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 12.778199195861816,
+      "learning_rate": 1.3326651297985367e-05,
+      "loss": 1.5023,
+      "step": 3330
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 9.352411270141602,
+      "learning_rate": 1.3324646687380978e-05,
+      "loss": 1.3491,
+      "step": 3331
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 32.7597770690918,
+      "learning_rate": 1.3322642076776586e-05,
+      "loss": 2.2412,
+      "step": 3332
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 14.371981620788574,
+      "learning_rate": 1.3320637466172198e-05,
+      "loss": 2.2198,
+      "step": 3333
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 15.051918029785156,
+      "learning_rate": 1.3318632855567808e-05,
+      "loss": 1.5761,
+      "step": 3334
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 18.553064346313477,
+      "learning_rate": 1.3316628244963416e-05,
+      "loss": 2.2955,
+      "step": 3335
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 13.871806144714355,
+      "learning_rate": 1.3314623634359028e-05,
+      "loss": 1.7798,
+      "step": 3336
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 24.853288650512695,
+      "learning_rate": 1.3312619023754636e-05,
+      "loss": 2.5074,
+      "step": 3337
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 7.72559928894043,
+      "learning_rate": 1.3310614413150246e-05,
+      "loss": 2.0241,
+      "step": 3338
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 67.96630096435547,
+      "learning_rate": 1.3308609802545858e-05,
+      "loss": 2.9062,
+      "step": 3339
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 15.404458045959473,
+      "learning_rate": 1.3306605191941467e-05,
+      "loss": 1.2167,
+      "step": 3340
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 20.57508087158203,
+      "learning_rate": 1.3304600581337077e-05,
+      "loss": 2.3393,
+      "step": 3341
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 13.644542694091797,
+      "learning_rate": 1.3302595970732685e-05,
+      "loss": 2.7504,
+      "step": 3342
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 27.4990291595459,
+      "learning_rate": 1.3300591360128297e-05,
+      "loss": 2.4179,
+      "step": 3343
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 18.17327308654785,
+      "learning_rate": 1.3298586749523905e-05,
+      "loss": 1.8711,
+      "step": 3344
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 11.454172134399414,
+      "learning_rate": 1.3296582138919515e-05,
+      "loss": 2.2247,
+      "step": 3345
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 14.537096977233887,
+      "learning_rate": 1.3294577528315127e-05,
+      "loss": 1.3509,
+      "step": 3346
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 17.769039154052734,
+      "learning_rate": 1.3292572917710736e-05,
+      "loss": 1.6675,
+      "step": 3347
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 20.219247817993164,
+      "learning_rate": 1.3290568307106346e-05,
+      "loss": 2.1718,
+      "step": 3348
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 12.704767227172852,
+      "learning_rate": 1.3288563696501956e-05,
+      "loss": 1.2209,
+      "step": 3349
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 15.621160507202148,
+      "learning_rate": 1.3286559085897566e-05,
+      "loss": 1.5298,
+      "step": 3350
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 17.41335678100586,
+      "learning_rate": 1.3284554475293174e-05,
+      "loss": 2.7119,
+      "step": 3351
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 27.469499588012695,
+      "learning_rate": 1.3282549864688786e-05,
+      "loss": 2.821,
+      "step": 3352
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 47.88648986816406,
+      "learning_rate": 1.3280545254084396e-05,
+      "loss": 2.4446,
+      "step": 3353
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 9.525798797607422,
+      "learning_rate": 1.3278540643480004e-05,
+      "loss": 0.9091,
+      "step": 3354
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 22.738584518432617,
+      "learning_rate": 1.3276536032875616e-05,
+      "loss": 2.3079,
+      "step": 3355
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 11.072502136230469,
+      "learning_rate": 1.3274531422271225e-05,
+      "loss": 2.7678,
+      "step": 3356
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 12.42871379852295,
+      "learning_rate": 1.3272526811666835e-05,
+      "loss": 1.1835,
+      "step": 3357
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 10.291549682617188,
+      "learning_rate": 1.3270522201062446e-05,
+      "loss": 1.3435,
+      "step": 3358
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 18.563772201538086,
+      "learning_rate": 1.3268517590458055e-05,
+      "loss": 1.2068,
+      "step": 3359
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 16.957143783569336,
+      "learning_rate": 1.3266512979853665e-05,
+      "loss": 2.197,
+      "step": 3360
+    },
+    {
+      "epoch": 1.01,
+      "eval_loss": 0.22327160835266113,
+      "eval_runtime": 43.5349,
+      "eval_samples_per_second": 33.973,
+      "eval_steps_per_second": 33.973,
+      "step": 3360
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 12.542174339294434,
+      "learning_rate": 1.3264508369249273e-05,
+      "loss": 1.7032,
+      "step": 3361
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 12.516151428222656,
+      "learning_rate": 1.3262503758644885e-05,
+      "loss": 1.4823,
+      "step": 3362
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 13.449585914611816,
+      "learning_rate": 1.3260499148040493e-05,
+      "loss": 1.9671,
+      "step": 3363
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 20.43147850036621,
+      "learning_rate": 1.3258494537436104e-05,
+      "loss": 1.5756,
+      "step": 3364
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 9.373948097229004,
+      "learning_rate": 1.3256489926831715e-05,
+      "loss": 2.1166,
+      "step": 3365
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 28.55181312561035,
+      "learning_rate": 1.3254485316227324e-05,
+      "loss": 2.7417,
+      "step": 3366
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 23.936809539794922,
+      "learning_rate": 1.3252480705622934e-05,
+      "loss": 1.5005,
+      "step": 3367
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 14.939528465270996,
+      "learning_rate": 1.3250476095018542e-05,
+      "loss": 2.3812,
+      "step": 3368
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 23.387401580810547,
+      "learning_rate": 1.3248471484414154e-05,
+      "loss": 2.5629,
+      "step": 3369
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 17.187917709350586,
+      "learning_rate": 1.3246466873809764e-05,
+      "loss": 1.8393,
+      "step": 3370
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 15.523499488830566,
+      "learning_rate": 1.3244462263205372e-05,
+      "loss": 2.066,
+      "step": 3371
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 8.489646911621094,
+      "learning_rate": 1.3242457652600984e-05,
+      "loss": 1.5775,
+      "step": 3372
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 12.514199256896973,
+      "learning_rate": 1.3240453041996593e-05,
+      "loss": 1.3532,
+      "step": 3373
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 39.9799919128418,
+      "learning_rate": 1.3238448431392203e-05,
+      "loss": 1.7734,
+      "step": 3374
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 12.430253028869629,
+      "learning_rate": 1.3236443820787813e-05,
+      "loss": 2.0699,
+      "step": 3375
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 32.21778106689453,
+      "learning_rate": 1.3234439210183423e-05,
+      "loss": 2.4161,
+      "step": 3376
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 10.491954803466797,
+      "learning_rate": 1.3232434599579033e-05,
+      "loss": 2.0041,
+      "step": 3377
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 24.133899688720703,
+      "learning_rate": 1.3230429988974643e-05,
+      "loss": 2.884,
+      "step": 3378
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 13.717731475830078,
+      "learning_rate": 1.3228425378370253e-05,
+      "loss": 2.4387,
+      "step": 3379
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 20.04202651977539,
+      "learning_rate": 1.3226420767765862e-05,
+      "loss": 1.764,
+      "step": 3380
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 16.394018173217773,
+      "learning_rate": 1.3224416157161473e-05,
+      "loss": 1.6656,
+      "step": 3381
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 6.424349308013916,
+      "learning_rate": 1.3222411546557083e-05,
+      "loss": 1.2659,
+      "step": 3382
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 12.825157165527344,
+      "learning_rate": 1.3220406935952692e-05,
+      "loss": 1.5172,
+      "step": 3383
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 32.225006103515625,
+      "learning_rate": 1.3218402325348304e-05,
+      "loss": 2.4318,
+      "step": 3384
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 10.880355834960938,
+      "learning_rate": 1.3216397714743912e-05,
+      "loss": 1.5311,
+      "step": 3385
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 23.808944702148438,
+      "learning_rate": 1.3214393104139522e-05,
+      "loss": 2.809,
+      "step": 3386
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 18.621950149536133,
+      "learning_rate": 1.321238849353513e-05,
+      "loss": 1.5185,
+      "step": 3387
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 13.235946655273438,
+      "learning_rate": 1.3210383882930742e-05,
+      "loss": 1.5264,
+      "step": 3388
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 12.991135597229004,
+      "learning_rate": 1.3208379272326352e-05,
+      "loss": 1.8659,
+      "step": 3389
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 38.41598129272461,
+      "learning_rate": 1.320637466172196e-05,
+      "loss": 2.2308,
+      "step": 3390
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 16.32402801513672,
+      "learning_rate": 1.3204370051117572e-05,
+      "loss": 1.7613,
+      "step": 3391
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 18.9952335357666,
+      "learning_rate": 1.320236544051318e-05,
+      "loss": 1.6083,
+      "step": 3392
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 20.658859252929688,
+      "learning_rate": 1.3200360829908791e-05,
+      "loss": 1.7974,
+      "step": 3393
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 15.612691879272461,
+      "learning_rate": 1.31983562193044e-05,
+      "loss": 2.5692,
+      "step": 3394
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 49.652915954589844,
+      "learning_rate": 1.3196351608700011e-05,
+      "loss": 2.6861,
+      "step": 3395
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 10.916786193847656,
+      "learning_rate": 1.3194346998095621e-05,
+      "loss": 1.4085,
+      "step": 3396
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 21.12449073791504,
+      "learning_rate": 1.3192342387491231e-05,
+      "loss": 2.6789,
+      "step": 3397
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 18.018606185913086,
+      "learning_rate": 1.3190337776886841e-05,
+      "loss": 1.6874,
+      "step": 3398
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 22.249736785888672,
+      "learning_rate": 1.318833316628245e-05,
+      "loss": 2.0898,
+      "step": 3399
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 14.536703109741211,
+      "learning_rate": 1.3186328555678062e-05,
+      "loss": 2.2679,
+      "step": 3400
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 16.38414764404297,
+      "learning_rate": 1.3184323945073672e-05,
+      "loss": 1.3831,
+      "step": 3401
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 16.283510208129883,
+      "learning_rate": 1.318231933446928e-05,
+      "loss": 1.622,
+      "step": 3402
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 12.426440238952637,
+      "learning_rate": 1.3180314723864892e-05,
+      "loss": 1.574,
+      "step": 3403
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 10.712419509887695,
+      "learning_rate": 1.31783101132605e-05,
+      "loss": 1.8414,
+      "step": 3404
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 12.0217924118042,
+      "learning_rate": 1.317630550265611e-05,
+      "loss": 1.817,
+      "step": 3405
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 10.142142295837402,
+      "learning_rate": 1.3174300892051719e-05,
+      "loss": 1.7129,
+      "step": 3406
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 17.107044219970703,
+      "learning_rate": 1.317229628144733e-05,
+      "loss": 2.2242,
+      "step": 3407
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 30.633485794067383,
+      "learning_rate": 1.317029167084294e-05,
+      "loss": 2.3054,
+      "step": 3408
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 10.093416213989258,
+      "learning_rate": 1.3168287060238549e-05,
+      "loss": 1.6337,
+      "step": 3409
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 16.132856369018555,
+      "learning_rate": 1.316628244963416e-05,
+      "loss": 1.8572,
+      "step": 3410
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 26.83838653564453,
+      "learning_rate": 1.3164277839029769e-05,
+      "loss": 1.6212,
+      "step": 3411
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 18.7099609375,
+      "learning_rate": 1.3162273228425379e-05,
+      "loss": 2.4694,
+      "step": 3412
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 15.113785743713379,
+      "learning_rate": 1.3160268617820991e-05,
+      "loss": 1.8307,
+      "step": 3413
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 13.659159660339355,
+      "learning_rate": 1.31582640072166e-05,
+      "loss": 2.0593,
+      "step": 3414
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 25.388036727905273,
+      "learning_rate": 1.315625939661221e-05,
+      "loss": 2.2341,
+      "step": 3415
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 7.588253021240234,
+      "learning_rate": 1.3154254786007818e-05,
+      "loss": 1.8049,
+      "step": 3416
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 18.153966903686523,
+      "learning_rate": 1.315225017540343e-05,
+      "loss": 2.1183,
+      "step": 3417
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 12.177376747131348,
+      "learning_rate": 1.3150245564799038e-05,
+      "loss": 1.611,
+      "step": 3418
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 15.74305534362793,
+      "learning_rate": 1.3148240954194648e-05,
+      "loss": 2.5822,
+      "step": 3419
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 13.683473587036133,
+      "learning_rate": 1.314623634359026e-05,
+      "loss": 1.4373,
+      "step": 3420
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 17.658405303955078,
+      "learning_rate": 1.3144231732985868e-05,
+      "loss": 1.9642,
+      "step": 3421
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 10.713934898376465,
+      "learning_rate": 1.3142227122381478e-05,
+      "loss": 1.7652,
+      "step": 3422
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 10.191012382507324,
+      "learning_rate": 1.3140222511777088e-05,
+      "loss": 1.7921,
+      "step": 3423
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 9.874408721923828,
+      "learning_rate": 1.3138217901172698e-05,
+      "loss": 0.9375,
+      "step": 3424
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 18.47734260559082,
+      "learning_rate": 1.3136213290568307e-05,
+      "loss": 2.3192,
+      "step": 3425
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 16.030399322509766,
+      "learning_rate": 1.3134208679963919e-05,
+      "loss": 1.957,
+      "step": 3426
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 8.997057914733887,
+      "learning_rate": 1.3132204069359529e-05,
+      "loss": 1.5734,
+      "step": 3427
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 174.50979614257812,
+      "learning_rate": 1.3130199458755137e-05,
+      "loss": 1.8404,
+      "step": 3428
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 15.005576133728027,
+      "learning_rate": 1.3128194848150749e-05,
+      "loss": 1.4347,
+      "step": 3429
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 13.528023719787598,
+      "learning_rate": 1.3126190237546357e-05,
+      "loss": 2.0983,
+      "step": 3430
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 61.08424377441406,
+      "learning_rate": 1.3124185626941967e-05,
+      "loss": 1.3849,
+      "step": 3431
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 19.432321548461914,
+      "learning_rate": 1.3122181016337579e-05,
+      "loss": 2.5755,
+      "step": 3432
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 10.8148775100708,
+      "learning_rate": 1.3120176405733188e-05,
+      "loss": 2.1681,
+      "step": 3433
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 19.197267532348633,
+      "learning_rate": 1.3118171795128798e-05,
+      "loss": 1.7698,
+      "step": 3434
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 16.004776000976562,
+      "learning_rate": 1.3116167184524406e-05,
+      "loss": 2.1169,
+      "step": 3435
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 36.55406188964844,
+      "learning_rate": 1.3114162573920018e-05,
+      "loss": 2.0371,
+      "step": 3436
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 20.354867935180664,
+      "learning_rate": 1.3112157963315626e-05,
+      "loss": 1.5596,
+      "step": 3437
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 13.395331382751465,
+      "learning_rate": 1.3110153352711236e-05,
+      "loss": 1.3172,
+      "step": 3438
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 17.100326538085938,
+      "learning_rate": 1.3108148742106848e-05,
+      "loss": 1.9774,
+      "step": 3439
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 13.835416793823242,
+      "learning_rate": 1.3106144131502456e-05,
+      "loss": 1.8643,
+      "step": 3440
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 15.755050659179688,
+      "learning_rate": 1.3104139520898067e-05,
+      "loss": 3.0466,
+      "step": 3441
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 19.994091033935547,
+      "learning_rate": 1.3102134910293675e-05,
+      "loss": 1.9489,
+      "step": 3442
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 10.072383880615234,
+      "learning_rate": 1.3100130299689287e-05,
+      "loss": 2.0764,
+      "step": 3443
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 10.777514457702637,
+      "learning_rate": 1.3098125689084897e-05,
+      "loss": 1.2584,
+      "step": 3444
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 24.992101669311523,
+      "learning_rate": 1.3096121078480505e-05,
+      "loss": 2.3589,
+      "step": 3445
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 13.744710922241211,
+      "learning_rate": 1.3094116467876117e-05,
+      "loss": 1.6164,
+      "step": 3446
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 9.5302095413208,
+      "learning_rate": 1.3092111857271725e-05,
+      "loss": 1.4418,
+      "step": 3447
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 16.148855209350586,
+      "learning_rate": 1.3090107246667335e-05,
+      "loss": 1.4431,
+      "step": 3448
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 31.971755981445312,
+      "learning_rate": 1.3088102636062945e-05,
+      "loss": 1.506,
+      "step": 3449
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 18.346967697143555,
+      "learning_rate": 1.3086098025458556e-05,
+      "loss": 1.6131,
+      "step": 3450
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 29.8282527923584,
+      "learning_rate": 1.3084093414854167e-05,
+      "loss": 2.3438,
+      "step": 3451
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 19.195152282714844,
+      "learning_rate": 1.3082088804249776e-05,
+      "loss": 1.9727,
+      "step": 3452
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 17.298931121826172,
+      "learning_rate": 1.3080084193645386e-05,
+      "loss": 1.5604,
+      "step": 3453
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 10.650527954101562,
+      "learning_rate": 1.3078079583040994e-05,
+      "loss": 1.0794,
+      "step": 3454
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 16.008625030517578,
+      "learning_rate": 1.3076074972436606e-05,
+      "loss": 1.7114,
+      "step": 3455
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 15.615730285644531,
+      "learning_rate": 1.3074070361832216e-05,
+      "loss": 2.0135,
+      "step": 3456
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 17.91425132751465,
+      "learning_rate": 1.3072065751227824e-05,
+      "loss": 1.8132,
+      "step": 3457
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 16.860435485839844,
+      "learning_rate": 1.3070061140623436e-05,
+      "loss": 1.9834,
+      "step": 3458
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 12.117476463317871,
+      "learning_rate": 1.3068056530019045e-05,
+      "loss": 1.62,
+      "step": 3459
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 18.71920394897461,
+      "learning_rate": 1.3066051919414655e-05,
+      "loss": 2.2109,
+      "step": 3460
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 18.306224822998047,
+      "learning_rate": 1.3064047308810263e-05,
+      "loss": 2.2506,
+      "step": 3461
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 27.221975326538086,
+      "learning_rate": 1.3062042698205875e-05,
+      "loss": 2.4873,
+      "step": 3462
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 31.973318099975586,
+      "learning_rate": 1.3060038087601485e-05,
+      "loss": 2.6493,
+      "step": 3463
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 10.039262771606445,
+      "learning_rate": 1.3058033476997093e-05,
+      "loss": 1.6217,
+      "step": 3464
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 17.775131225585938,
+      "learning_rate": 1.3056028866392705e-05,
+      "loss": 1.6111,
+      "step": 3465
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 41.91878890991211,
+      "learning_rate": 1.3054024255788314e-05,
+      "loss": 2.4916,
+      "step": 3466
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 19.899276733398438,
+      "learning_rate": 1.3052019645183924e-05,
+      "loss": 2.2337,
+      "step": 3467
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 12.127092361450195,
+      "learning_rate": 1.3050015034579534e-05,
+      "loss": 1.9265,
+      "step": 3468
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 28.322265625,
+      "learning_rate": 1.3048010423975144e-05,
+      "loss": 1.8237,
+      "step": 3469
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 11.918156623840332,
+      "learning_rate": 1.3046005813370754e-05,
+      "loss": 2.1019,
+      "step": 3470
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 10.887943267822266,
+      "learning_rate": 1.3044001202766364e-05,
+      "loss": 2.118,
+      "step": 3471
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 14.735898971557617,
+      "learning_rate": 1.3041996592161974e-05,
+      "loss": 1.5392,
+      "step": 3472
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 22.728878021240234,
+      "learning_rate": 1.3039991981557582e-05,
+      "loss": 1.9459,
+      "step": 3473
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 9.619507789611816,
+      "learning_rate": 1.3037987370953194e-05,
+      "loss": 1.1765,
+      "step": 3474
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 13.704413414001465,
+      "learning_rate": 1.3035982760348804e-05,
+      "loss": 1.5395,
+      "step": 3475
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 13.936915397644043,
+      "learning_rate": 1.3033978149744413e-05,
+      "loss": 1.2055,
+      "step": 3476
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 23.567424774169922,
+      "learning_rate": 1.3031973539140024e-05,
+      "loss": 1.7821,
+      "step": 3477
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 17.09320640563965,
+      "learning_rate": 1.3029968928535633e-05,
+      "loss": 2.4582,
+      "step": 3478
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 12.296806335449219,
+      "learning_rate": 1.3027964317931243e-05,
+      "loss": 1.4382,
+      "step": 3479
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 12.535650253295898,
+      "learning_rate": 1.3025959707326851e-05,
+      "loss": 1.2141,
+      "step": 3480
+    },
+    {
+      "epoch": 1.05,
+      "eval_loss": 0.22333772480487823,
+      "eval_runtime": 43.6637,
+      "eval_samples_per_second": 33.873,
+      "eval_steps_per_second": 33.873,
+      "step": 3480
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 8.423393249511719,
+      "learning_rate": 1.3023955096722463e-05,
+      "loss": 1.2467,
+      "step": 3481
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 13.054819107055664,
+      "learning_rate": 1.3021950486118073e-05,
+      "loss": 1.2692,
+      "step": 3482
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 67.5218734741211,
+      "learning_rate": 1.3019945875513682e-05,
+      "loss": 2.0009,
+      "step": 3483
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 51.58919143676758,
+      "learning_rate": 1.3017941264909293e-05,
+      "loss": 3.2618,
+      "step": 3484
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 24.077619552612305,
+      "learning_rate": 1.3015936654304902e-05,
+      "loss": 2.4753,
+      "step": 3485
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 25.09623146057129,
+      "learning_rate": 1.3013932043700512e-05,
+      "loss": 1.9661,
+      "step": 3486
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 33.02434539794922,
+      "learning_rate": 1.3011927433096124e-05,
+      "loss": 2.3949,
+      "step": 3487
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 16.488039016723633,
+      "learning_rate": 1.3009922822491732e-05,
+      "loss": 2.4064,
+      "step": 3488
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 24.7977237701416,
+      "learning_rate": 1.3007918211887342e-05,
+      "loss": 1.2102,
+      "step": 3489
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 16.36675262451172,
+      "learning_rate": 1.300591360128295e-05,
+      "loss": 1.4976,
+      "step": 3490
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 16.940837860107422,
+      "learning_rate": 1.3003908990678562e-05,
+      "loss": 1.9573,
+      "step": 3491
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 17.5327091217041,
+      "learning_rate": 1.300190438007417e-05,
+      "loss": 1.9867,
+      "step": 3492
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 14.204058647155762,
+      "learning_rate": 1.299989976946978e-05,
+      "loss": 1.8624,
+      "step": 3493
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 41.586021423339844,
+      "learning_rate": 1.2997895158865392e-05,
+      "loss": 3.3605,
+      "step": 3494
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 18.94367218017578,
+      "learning_rate": 1.2995890548261001e-05,
+      "loss": 1.3739,
+      "step": 3495
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 23.99241828918457,
+      "learning_rate": 1.2993885937656611e-05,
+      "loss": 2.0589,
+      "step": 3496
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 11.752195358276367,
+      "learning_rate": 1.2991881327052221e-05,
+      "loss": 1.3307,
+      "step": 3497
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 16.153244018554688,
+      "learning_rate": 1.2989876716447831e-05,
+      "loss": 1.6504,
+      "step": 3498
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 15.57225513458252,
+      "learning_rate": 1.2987872105843441e-05,
+      "loss": 2.247,
+      "step": 3499
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 14.485071182250977,
+      "learning_rate": 1.2985867495239051e-05,
+      "loss": 1.7557,
+      "step": 3500
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 8.054557800292969,
+      "learning_rate": 1.2983862884634661e-05,
+      "loss": 1.3367,
+      "step": 3501
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 19.42696189880371,
+      "learning_rate": 1.298185827403027e-05,
+      "loss": 1.7359,
+      "step": 3502
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 21.340444564819336,
+      "learning_rate": 1.2979853663425882e-05,
+      "loss": 2.8674,
+      "step": 3503
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 16.97898292541504,
+      "learning_rate": 1.297784905282149e-05,
+      "loss": 2.4501,
+      "step": 3504
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 17.556812286376953,
+      "learning_rate": 1.29758444422171e-05,
+      "loss": 1.421,
+      "step": 3505
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 12.702077865600586,
+      "learning_rate": 1.2973839831612712e-05,
+      "loss": 2.1779,
+      "step": 3506
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 17.422500610351562,
+      "learning_rate": 1.297183522100832e-05,
+      "loss": 2.1388,
+      "step": 3507
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 27.59862518310547,
+      "learning_rate": 1.296983061040393e-05,
+      "loss": 3.23,
+      "step": 3508
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 11.814319610595703,
+      "learning_rate": 1.2967825999799539e-05,
+      "loss": 1.5021,
+      "step": 3509
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 10.947737693786621,
+      "learning_rate": 1.296582138919515e-05,
+      "loss": 1.298,
+      "step": 3510
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 14.834168434143066,
+      "learning_rate": 1.2963816778590759e-05,
+      "loss": 1.6569,
+      "step": 3511
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 27.23177719116211,
+      "learning_rate": 1.2961812167986369e-05,
+      "loss": 1.6755,
+      "step": 3512
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 16.08486557006836,
+      "learning_rate": 1.295980755738198e-05,
+      "loss": 1.8572,
+      "step": 3513
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 23.723173141479492,
+      "learning_rate": 1.2957802946777589e-05,
+      "loss": 2.7123,
+      "step": 3514
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 15.125565528869629,
+      "learning_rate": 1.29557983361732e-05,
+      "loss": 2.5594,
+      "step": 3515
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 25.76398468017578,
+      "learning_rate": 1.295379372556881e-05,
+      "loss": 1.7984,
+      "step": 3516
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 18.840059280395508,
+      "learning_rate": 1.295178911496442e-05,
+      "loss": 2.232,
+      "step": 3517
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 35.28987503051758,
+      "learning_rate": 1.294978450436003e-05,
+      "loss": 2.267,
+      "step": 3518
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 7.588108062744141,
+      "learning_rate": 1.294777989375564e-05,
+      "loss": 1.2528,
+      "step": 3519
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 13.926599502563477,
+      "learning_rate": 1.294577528315125e-05,
+      "loss": 1.7358,
+      "step": 3520
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 16.25490951538086,
+      "learning_rate": 1.2943770672546858e-05,
+      "loss": 2.2304,
+      "step": 3521
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 30.820024490356445,
+      "learning_rate": 1.294176606194247e-05,
+      "loss": 2.0343,
+      "step": 3522
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 15.902755737304688,
+      "learning_rate": 1.2939761451338078e-05,
+      "loss": 1.5815,
+      "step": 3523
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 22.500484466552734,
+      "learning_rate": 1.2937756840733688e-05,
+      "loss": 2.0932,
+      "step": 3524
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 11.80160903930664,
+      "learning_rate": 1.29357522301293e-05,
+      "loss": 1.8902,
+      "step": 3525
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 24.187225341796875,
+      "learning_rate": 1.2933747619524908e-05,
+      "loss": 1.9731,
+      "step": 3526
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 13.519028663635254,
+      "learning_rate": 1.2931743008920519e-05,
+      "loss": 1.9801,
+      "step": 3527
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 54.62118911743164,
+      "learning_rate": 1.2929738398316127e-05,
+      "loss": 3.4219,
+      "step": 3528
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 12.734724044799805,
+      "learning_rate": 1.2927733787711739e-05,
+      "loss": 2.3175,
+      "step": 3529
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 15.261761665344238,
+      "learning_rate": 1.2925729177107349e-05,
+      "loss": 2.6037,
+      "step": 3530
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 19.809280395507812,
+      "learning_rate": 1.2923724566502957e-05,
+      "loss": 1.9901,
+      "step": 3531
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 18.573335647583008,
+      "learning_rate": 1.2921719955898569e-05,
+      "loss": 2.4825,
+      "step": 3532
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 17.13604736328125,
+      "learning_rate": 1.2919715345294177e-05,
+      "loss": 1.3552,
+      "step": 3533
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 14.92817497253418,
+      "learning_rate": 1.2917710734689787e-05,
+      "loss": 2.1547,
+      "step": 3534
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 26.906400680541992,
+      "learning_rate": 1.2915706124085396e-05,
+      "loss": 1.741,
+      "step": 3535
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 17.950481414794922,
+      "learning_rate": 1.2913701513481008e-05,
+      "loss": 1.7533,
+      "step": 3536
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 18.598140716552734,
+      "learning_rate": 1.2911696902876618e-05,
+      "loss": 1.9817,
+      "step": 3537
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 20.901409149169922,
+      "learning_rate": 1.2909692292272226e-05,
+      "loss": 1.8785,
+      "step": 3538
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 13.569160461425781,
+      "learning_rate": 1.2907687681667838e-05,
+      "loss": 2.0731,
+      "step": 3539
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 16.909326553344727,
+      "learning_rate": 1.2905683071063446e-05,
+      "loss": 1.912,
+      "step": 3540
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 15.05936050415039,
+      "learning_rate": 1.2903678460459056e-05,
+      "loss": 3.6215,
+      "step": 3541
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 12.6626615524292,
+      "learning_rate": 1.2901673849854668e-05,
+      "loss": 1.6078,
+      "step": 3542
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 12.250699043273926,
+      "learning_rate": 1.2899669239250276e-05,
+      "loss": 1.9456,
+      "step": 3543
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 28.121313095092773,
+      "learning_rate": 1.2897664628645887e-05,
+      "loss": 2.6878,
+      "step": 3544
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 21.93025779724121,
+      "learning_rate": 1.2895660018041497e-05,
+      "loss": 2.364,
+      "step": 3545
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 11.087515830993652,
+      "learning_rate": 1.2893655407437107e-05,
+      "loss": 2.0794,
+      "step": 3546
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 38.5289421081543,
+      "learning_rate": 1.2891650796832715e-05,
+      "loss": 2.4577,
+      "step": 3547
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 29.22125816345215,
+      "learning_rate": 1.2889646186228327e-05,
+      "loss": 2.4881,
+      "step": 3548
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 25.916982650756836,
+      "learning_rate": 1.2887641575623937e-05,
+      "loss": 1.4721,
+      "step": 3549
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 20.641904830932617,
+      "learning_rate": 1.2885636965019545e-05,
+      "loss": 1.6015,
+      "step": 3550
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 14.579924583435059,
+      "learning_rate": 1.2883632354415157e-05,
+      "loss": 1.638,
+      "step": 3551
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 36.379661560058594,
+      "learning_rate": 1.2881627743810766e-05,
+      "loss": 2.2669,
+      "step": 3552
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 16.51966094970703,
+      "learning_rate": 1.2879623133206376e-05,
+      "loss": 1.7893,
+      "step": 3553
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 23.127784729003906,
+      "learning_rate": 1.2877618522601984e-05,
+      "loss": 2.266,
+      "step": 3554
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 17.028398513793945,
+      "learning_rate": 1.2875613911997596e-05,
+      "loss": 1.6085,
+      "step": 3555
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 9.327510833740234,
+      "learning_rate": 1.2873609301393206e-05,
+      "loss": 1.155,
+      "step": 3556
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 9.496240615844727,
+      "learning_rate": 1.2871604690788814e-05,
+      "loss": 1.3495,
+      "step": 3557
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 10.282184600830078,
+      "learning_rate": 1.2869600080184426e-05,
+      "loss": 1.4117,
+      "step": 3558
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 15.490508079528809,
+      "learning_rate": 1.2867595469580034e-05,
+      "loss": 1.5863,
+      "step": 3559
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 17.8187198638916,
+      "learning_rate": 1.2865590858975645e-05,
+      "loss": 1.2819,
+      "step": 3560
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 17.21719741821289,
+      "learning_rate": 1.2863586248371256e-05,
+      "loss": 2.0228,
+      "step": 3561
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 23.23194694519043,
+      "learning_rate": 1.2861581637766865e-05,
+      "loss": 1.858,
+      "step": 3562
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 25.972341537475586,
+      "learning_rate": 1.2859577027162475e-05,
+      "loss": 1.3268,
+      "step": 3563
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 27.290193557739258,
+      "learning_rate": 1.2857572416558083e-05,
+      "loss": 2.174,
+      "step": 3564
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 14.838099479675293,
+      "learning_rate": 1.2855567805953695e-05,
+      "loss": 1.9506,
+      "step": 3565
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 8.05460262298584,
+      "learning_rate": 1.2853563195349303e-05,
+      "loss": 1.2423,
+      "step": 3566
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 10.019596099853516,
+      "learning_rate": 1.2851558584744913e-05,
+      "loss": 1.405,
+      "step": 3567
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 16.532649993896484,
+      "learning_rate": 1.2849553974140525e-05,
+      "loss": 1.5799,
+      "step": 3568
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 14.281306266784668,
+      "learning_rate": 1.2847549363536134e-05,
+      "loss": 1.1582,
+      "step": 3569
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 9.972299575805664,
+      "learning_rate": 1.2845544752931745e-05,
+      "loss": 1.2031,
+      "step": 3570
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 14.451814651489258,
+      "learning_rate": 1.2843540142327354e-05,
+      "loss": 1.5313,
+      "step": 3571
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 15.139348030090332,
+      "learning_rate": 1.2841535531722964e-05,
+      "loss": 1.9977,
+      "step": 3572
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 25.56646156311035,
+      "learning_rate": 1.2839530921118576e-05,
+      "loss": 2.2934,
+      "step": 3573
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 11.112027168273926,
+      "learning_rate": 1.2837526310514184e-05,
+      "loss": 1.5678,
+      "step": 3574
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 14.32687759399414,
+      "learning_rate": 1.2835521699909794e-05,
+      "loss": 1.1919,
+      "step": 3575
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 11.444255828857422,
+      "learning_rate": 1.2833517089305402e-05,
+      "loss": 2.2952,
+      "step": 3576
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 34.31683349609375,
+      "learning_rate": 1.2831512478701014e-05,
+      "loss": 1.8054,
+      "step": 3577
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 16.4061336517334,
+      "learning_rate": 1.2829507868096623e-05,
+      "loss": 1.1196,
+      "step": 3578
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 12.305190086364746,
+      "learning_rate": 1.2827503257492233e-05,
+      "loss": 2.2021,
+      "step": 3579
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 16.25429916381836,
+      "learning_rate": 1.2825498646887845e-05,
+      "loss": 1.629,
+      "step": 3580
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 25.61264419555664,
+      "learning_rate": 1.2823494036283453e-05,
+      "loss": 2.7876,
+      "step": 3581
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 20.26491928100586,
+      "learning_rate": 1.2821489425679063e-05,
+      "loss": 1.1916,
+      "step": 3582
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 10.56248950958252,
+      "learning_rate": 1.2819484815074671e-05,
+      "loss": 1.302,
+      "step": 3583
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 15.701934814453125,
+      "learning_rate": 1.2817480204470283e-05,
+      "loss": 1.5957,
+      "step": 3584
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 17.989017486572266,
+      "learning_rate": 1.2815475593865893e-05,
+      "loss": 1.2969,
+      "step": 3585
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 22.148818969726562,
+      "learning_rate": 1.2813470983261502e-05,
+      "loss": 2.5261,
+      "step": 3586
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 26.945255279541016,
+      "learning_rate": 1.2811466372657113e-05,
+      "loss": 1.8008,
+      "step": 3587
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 13.003181457519531,
+      "learning_rate": 1.2809461762052722e-05,
+      "loss": 1.4448,
+      "step": 3588
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 13.033709526062012,
+      "learning_rate": 1.2807457151448332e-05,
+      "loss": 2.1353,
+      "step": 3589
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 16.16973876953125,
+      "learning_rate": 1.2805452540843942e-05,
+      "loss": 1.847,
+      "step": 3590
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 14.109622955322266,
+      "learning_rate": 1.2803447930239552e-05,
+      "loss": 2.0083,
+      "step": 3591
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 13.897767066955566,
+      "learning_rate": 1.2801443319635162e-05,
+      "loss": 1.2688,
+      "step": 3592
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 18.122596740722656,
+      "learning_rate": 1.2799438709030772e-05,
+      "loss": 1.9293,
+      "step": 3593
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 41.87525939941406,
+      "learning_rate": 1.2797434098426382e-05,
+      "loss": 3.1812,
+      "step": 3594
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 22.537267684936523,
+      "learning_rate": 1.279542948782199e-05,
+      "loss": 1.9805,
+      "step": 3595
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 16.393918991088867,
+      "learning_rate": 1.2793424877217602e-05,
+      "loss": 1.6751,
+      "step": 3596
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 15.888949394226074,
+      "learning_rate": 1.2791420266613211e-05,
+      "loss": 2.2047,
+      "step": 3597
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 13.120323181152344,
+      "learning_rate": 1.2789415656008821e-05,
+      "loss": 1.527,
+      "step": 3598
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 14.314374923706055,
+      "learning_rate": 1.2787411045404433e-05,
+      "loss": 2.3968,
+      "step": 3599
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 20.696285247802734,
+      "learning_rate": 1.2785406434800041e-05,
+      "loss": 1.8987,
+      "step": 3600
+    },
+    {
+      "epoch": 1.08,
+      "eval_loss": 0.23791223764419556,
+      "eval_runtime": 43.6058,
+      "eval_samples_per_second": 33.917,
+      "eval_steps_per_second": 33.917,
+      "step": 3600
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 24.901500701904297,
+      "learning_rate": 1.2783401824195651e-05,
+      "loss": 1.9859,
+      "step": 3601
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 15.03045654296875,
+      "learning_rate": 1.278139721359126e-05,
+      "loss": 1.6967,
+      "step": 3602
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 16.57175064086914,
+      "learning_rate": 1.2779392602986871e-05,
+      "loss": 1.8149,
+      "step": 3603
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 15.594796180725098,
+      "learning_rate": 1.2777387992382481e-05,
+      "loss": 1.4157,
+      "step": 3604
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 14.430771827697754,
+      "learning_rate": 1.277538338177809e-05,
+      "loss": 2.3501,
+      "step": 3605
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 38.617862701416016,
+      "learning_rate": 1.2773378771173702e-05,
+      "loss": 1.98,
+      "step": 3606
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 19.550251007080078,
+      "learning_rate": 1.277137416056931e-05,
+      "loss": 1.2882,
+      "step": 3607
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 19.823959350585938,
+      "learning_rate": 1.276936954996492e-05,
+      "loss": 2.7407,
+      "step": 3608
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 46.88768005371094,
+      "learning_rate": 1.2767364939360528e-05,
+      "loss": 1.9532,
+      "step": 3609
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 15.034628868103027,
+      "learning_rate": 1.276536032875614e-05,
+      "loss": 1.2125,
+      "step": 3610
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 24.965927124023438,
+      "learning_rate": 1.276335571815175e-05,
+      "loss": 2.4233,
+      "step": 3611
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 26.463823318481445,
+      "learning_rate": 1.2761351107547359e-05,
+      "loss": 2.3175,
+      "step": 3612
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 15.110565185546875,
+      "learning_rate": 1.275934649694297e-05,
+      "loss": 2.4443,
+      "step": 3613
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 18.711685180664062,
+      "learning_rate": 1.2757341886338579e-05,
+      "loss": 1.5244,
+      "step": 3614
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 21.489322662353516,
+      "learning_rate": 1.2755337275734189e-05,
+      "loss": 1.9058,
+      "step": 3615
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 15.04526424407959,
+      "learning_rate": 1.27533326651298e-05,
+      "loss": 1.4516,
+      "step": 3616
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 12.51637077331543,
+      "learning_rate": 1.2751328054525409e-05,
+      "loss": 1.7847,
+      "step": 3617
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 29.387882232666016,
+      "learning_rate": 1.274932344392102e-05,
+      "loss": 2.5068,
+      "step": 3618
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 18.071352005004883,
+      "learning_rate": 1.274731883331663e-05,
+      "loss": 2.6933,
+      "step": 3619
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 20.221973419189453,
+      "learning_rate": 1.274531422271224e-05,
+      "loss": 1.3898,
+      "step": 3620
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 8.028789520263672,
+      "learning_rate": 1.2743309612107848e-05,
+      "loss": 1.2193,
+      "step": 3621
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 14.530386924743652,
+      "learning_rate": 1.274130500150346e-05,
+      "loss": 2.0765,
+      "step": 3622
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 16.220090866088867,
+      "learning_rate": 1.273930039089907e-05,
+      "loss": 1.7681,
+      "step": 3623
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 21.331134796142578,
+      "learning_rate": 1.2737295780294678e-05,
+      "loss": 2.7775,
+      "step": 3624
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 11.270216941833496,
+      "learning_rate": 1.273529116969029e-05,
+      "loss": 2.3103,
+      "step": 3625
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 16.944839477539062,
+      "learning_rate": 1.2733286559085898e-05,
+      "loss": 1.7782,
+      "step": 3626
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 8.570146560668945,
+      "learning_rate": 1.2731281948481508e-05,
+      "loss": 1.6687,
+      "step": 3627
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 53.58586502075195,
+      "learning_rate": 1.272927733787712e-05,
+      "loss": 3.6285,
+      "step": 3628
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 24.34233856201172,
+      "learning_rate": 1.2727272727272728e-05,
+      "loss": 2.7752,
+      "step": 3629
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 14.564997673034668,
+      "learning_rate": 1.2725268116668339e-05,
+      "loss": 2.5383,
+      "step": 3630
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 6.960912227630615,
+      "learning_rate": 1.2723263506063947e-05,
+      "loss": 0.6939,
+      "step": 3631
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 17.01753044128418,
+      "learning_rate": 1.2721258895459559e-05,
+      "loss": 1.6664,
+      "step": 3632
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 11.095540046691895,
+      "learning_rate": 1.2719254284855167e-05,
+      "loss": 1.6678,
+      "step": 3633
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 11.197050094604492,
+      "learning_rate": 1.2717249674250777e-05,
+      "loss": 1.9901,
+      "step": 3634
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 11.85187816619873,
+      "learning_rate": 1.2715245063646389e-05,
+      "loss": 0.7637,
+      "step": 3635
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 46.356109619140625,
+      "learning_rate": 1.2713240453041997e-05,
+      "loss": 1.686,
+      "step": 3636
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 72.00715637207031,
+      "learning_rate": 1.2711235842437607e-05,
+      "loss": 2.1146,
+      "step": 3637
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 23.22759437561035,
+      "learning_rate": 1.2709231231833218e-05,
+      "loss": 1.1876,
+      "step": 3638
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 16.499618530273438,
+      "learning_rate": 1.2707226621228828e-05,
+      "loss": 1.4633,
+      "step": 3639
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 35.80891418457031,
+      "learning_rate": 1.2705222010624436e-05,
+      "loss": 1.7426,
+      "step": 3640
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 18.7283992767334,
+      "learning_rate": 1.2703217400020048e-05,
+      "loss": 1.8939,
+      "step": 3641
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 14.170825958251953,
+      "learning_rate": 1.2701212789415658e-05,
+      "loss": 1.992,
+      "step": 3642
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 11.59808349609375,
+      "learning_rate": 1.2699208178811266e-05,
+      "loss": 1.9479,
+      "step": 3643
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 27.232282638549805,
+      "learning_rate": 1.2697203568206878e-05,
+      "loss": 1.9665,
+      "step": 3644
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 13.292805671691895,
+      "learning_rate": 1.2695198957602486e-05,
+      "loss": 1.4353,
+      "step": 3645
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 8.74724292755127,
+      "learning_rate": 1.2693194346998097e-05,
+      "loss": 1.3016,
+      "step": 3646
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 19.11276626586914,
+      "learning_rate": 1.2691189736393708e-05,
+      "loss": 2.3302,
+      "step": 3647
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 46.45100021362305,
+      "learning_rate": 1.2689185125789317e-05,
+      "loss": 2.1202,
+      "step": 3648
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 15.36929702758789,
+      "learning_rate": 1.2687180515184927e-05,
+      "loss": 1.7896,
+      "step": 3649
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 20.07041358947754,
+      "learning_rate": 1.2685175904580535e-05,
+      "loss": 1.826,
+      "step": 3650
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 20.192039489746094,
+      "learning_rate": 1.2683171293976147e-05,
+      "loss": 1.8944,
+      "step": 3651
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 15.358658790588379,
+      "learning_rate": 1.2681166683371755e-05,
+      "loss": 2.1073,
+      "step": 3652
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 11.35305404663086,
+      "learning_rate": 1.2679162072767365e-05,
+      "loss": 1.167,
+      "step": 3653
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 14.477778434753418,
+      "learning_rate": 1.2677157462162977e-05,
+      "loss": 2.1406,
+      "step": 3654
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 25.864349365234375,
+      "learning_rate": 1.2675152851558586e-05,
+      "loss": 1.9479,
+      "step": 3655
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 14.707035064697266,
+      "learning_rate": 1.2673148240954196e-05,
+      "loss": 2.2159,
+      "step": 3656
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 9.004791259765625,
+      "learning_rate": 1.2671143630349804e-05,
+      "loss": 0.9894,
+      "step": 3657
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 16.240209579467773,
+      "learning_rate": 1.2669139019745416e-05,
+      "loss": 1.6444,
+      "step": 3658
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 15.315890312194824,
+      "learning_rate": 1.2667134409141026e-05,
+      "loss": 2.046,
+      "step": 3659
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 14.05276870727539,
+      "learning_rate": 1.2665129798536634e-05,
+      "loss": 2.0308,
+      "step": 3660
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 12.463837623596191,
+      "learning_rate": 1.2663125187932246e-05,
+      "loss": 1.3266,
+      "step": 3661
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 10.216002464294434,
+      "learning_rate": 1.2661120577327854e-05,
+      "loss": 1.4467,
+      "step": 3662
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 52.47303771972656,
+      "learning_rate": 1.2659115966723465e-05,
+      "loss": 3.601,
+      "step": 3663
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 19.366987228393555,
+      "learning_rate": 1.2657111356119075e-05,
+      "loss": 1.9766,
+      "step": 3664
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 11.65963077545166,
+      "learning_rate": 1.2655106745514685e-05,
+      "loss": 1.6192,
+      "step": 3665
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 14.95705509185791,
+      "learning_rate": 1.2653102134910295e-05,
+      "loss": 1.8748,
+      "step": 3666
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 21.516704559326172,
+      "learning_rate": 1.2651097524305905e-05,
+      "loss": 2.2989,
+      "step": 3667
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 12.6599760055542,
+      "learning_rate": 1.2649092913701515e-05,
+      "loss": 1.9577,
+      "step": 3668
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 20.421907424926758,
+      "learning_rate": 1.2647088303097123e-05,
+      "loss": 1.8006,
+      "step": 3669
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 21.49991226196289,
+      "learning_rate": 1.2645083692492735e-05,
+      "loss": 1.5208,
+      "step": 3670
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 16.71851348876953,
+      "learning_rate": 1.2643079081888345e-05,
+      "loss": 1.3844,
+      "step": 3671
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 12.691519737243652,
+      "learning_rate": 1.2641074471283954e-05,
+      "loss": 1.6805,
+      "step": 3672
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 13.60329532623291,
+      "learning_rate": 1.2639069860679565e-05,
+      "loss": 1.3261,
+      "step": 3673
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 12.163605690002441,
+      "learning_rate": 1.2637065250075174e-05,
+      "loss": 1.2984,
+      "step": 3674
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 11.735462188720703,
+      "learning_rate": 1.2635060639470784e-05,
+      "loss": 1.3691,
+      "step": 3675
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 10.237204551696777,
+      "learning_rate": 1.2633056028866392e-05,
+      "loss": 1.4796,
+      "step": 3676
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 16.992961883544922,
+      "learning_rate": 1.2631051418262004e-05,
+      "loss": 1.5819,
+      "step": 3677
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 25.052305221557617,
+      "learning_rate": 1.2629046807657614e-05,
+      "loss": 2.6452,
+      "step": 3678
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 34.612735748291016,
+      "learning_rate": 1.2627042197053223e-05,
+      "loss": 1.888,
+      "step": 3679
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 12.312444686889648,
+      "learning_rate": 1.2625037586448834e-05,
+      "loss": 1.6545,
+      "step": 3680
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 17.487689971923828,
+      "learning_rate": 1.2623032975844443e-05,
+      "loss": 2.225,
+      "step": 3681
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 12.337897300720215,
+      "learning_rate": 1.2621028365240053e-05,
+      "loss": 1.285,
+      "step": 3682
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 12.9970121383667,
+      "learning_rate": 1.2619023754635661e-05,
+      "loss": 1.5476,
+      "step": 3683
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 36.430641174316406,
+      "learning_rate": 1.2617019144031273e-05,
+      "loss": 2.409,
+      "step": 3684
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 11.384819030761719,
+      "learning_rate": 1.2615014533426883e-05,
+      "loss": 1.2839,
+      "step": 3685
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 14.997862815856934,
+      "learning_rate": 1.2613009922822491e-05,
+      "loss": 2.1861,
+      "step": 3686
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 9.791463851928711,
+      "learning_rate": 1.2611005312218103e-05,
+      "loss": 1.6862,
+      "step": 3687
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 20.38786506652832,
+      "learning_rate": 1.2609000701613712e-05,
+      "loss": 1.8915,
+      "step": 3688
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 56.49119567871094,
+      "learning_rate": 1.2606996091009323e-05,
+      "loss": 1.4701,
+      "step": 3689
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 9.169239044189453,
+      "learning_rate": 1.2604991480404933e-05,
+      "loss": 1.0245,
+      "step": 3690
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 18.84854507446289,
+      "learning_rate": 1.2602986869800542e-05,
+      "loss": 1.6371,
+      "step": 3691
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 15.238136291503906,
+      "learning_rate": 1.2600982259196154e-05,
+      "loss": 1.4109,
+      "step": 3692
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 24.6544189453125,
+      "learning_rate": 1.2598977648591762e-05,
+      "loss": 1.8289,
+      "step": 3693
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 15.802862167358398,
+      "learning_rate": 1.2596973037987372e-05,
+      "loss": 2.074,
+      "step": 3694
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 21.196699142456055,
+      "learning_rate": 1.259496842738298e-05,
+      "loss": 2.4331,
+      "step": 3695
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 14.442291259765625,
+      "learning_rate": 1.2592963816778592e-05,
+      "loss": 1.1208,
+      "step": 3696
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 16.67438316345215,
+      "learning_rate": 1.2590959206174202e-05,
+      "loss": 1.9479,
+      "step": 3697
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 30.792251586914062,
+      "learning_rate": 1.258895459556981e-05,
+      "loss": 1.0711,
+      "step": 3698
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 21.835054397583008,
+      "learning_rate": 1.2586949984965423e-05,
+      "loss": 1.4139,
+      "step": 3699
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 23.054439544677734,
+      "learning_rate": 1.2584945374361031e-05,
+      "loss": 2.3756,
+      "step": 3700
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 15.562328338623047,
+      "learning_rate": 1.2582940763756641e-05,
+      "loss": 1.6298,
+      "step": 3701
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 19.600475311279297,
+      "learning_rate": 1.2580936153152253e-05,
+      "loss": 2.0049,
+      "step": 3702
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 15.934944152832031,
+      "learning_rate": 1.2578931542547861e-05,
+      "loss": 2.3699,
+      "step": 3703
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 13.748173713684082,
+      "learning_rate": 1.2576926931943471e-05,
+      "loss": 2.802,
+      "step": 3704
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 10.191062927246094,
+      "learning_rate": 1.257492232133908e-05,
+      "loss": 0.9145,
+      "step": 3705
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 14.744169235229492,
+      "learning_rate": 1.2572917710734691e-05,
+      "loss": 1.7588,
+      "step": 3706
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 17.905502319335938,
+      "learning_rate": 1.25709131001303e-05,
+      "loss": 1.4409,
+      "step": 3707
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 22.17319679260254,
+      "learning_rate": 1.256890848952591e-05,
+      "loss": 1.6653,
+      "step": 3708
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 14.994551658630371,
+      "learning_rate": 1.2566903878921522e-05,
+      "loss": 2.6141,
+      "step": 3709
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 27.965063095092773,
+      "learning_rate": 1.256489926831713e-05,
+      "loss": 1.4174,
+      "step": 3710
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 12.438477516174316,
+      "learning_rate": 1.256289465771274e-05,
+      "loss": 2.034,
+      "step": 3711
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 18.60542106628418,
+      "learning_rate": 1.256089004710835e-05,
+      "loss": 2.1583,
+      "step": 3712
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 11.205132484436035,
+      "learning_rate": 1.255888543650396e-05,
+      "loss": 2.2042,
+      "step": 3713
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 21.031993865966797,
+      "learning_rate": 1.255688082589957e-05,
+      "loss": 1.3351,
+      "step": 3714
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 22.70185661315918,
+      "learning_rate": 1.255487621529518e-05,
+      "loss": 2.8246,
+      "step": 3715
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 19.754451751708984,
+      "learning_rate": 1.255287160469079e-05,
+      "loss": 1.9262,
+      "step": 3716
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 14.8783597946167,
+      "learning_rate": 1.2550866994086399e-05,
+      "loss": 1.958,
+      "step": 3717
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 13.51906681060791,
+      "learning_rate": 1.254886238348201e-05,
+      "loss": 2.0377,
+      "step": 3718
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 13.766884803771973,
+      "learning_rate": 1.2546857772877619e-05,
+      "loss": 1.8277,
+      "step": 3719
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 11.940971374511719,
+      "learning_rate": 1.254485316227323e-05,
+      "loss": 2.0253,
+      "step": 3720
+    },
+    {
+      "epoch": 1.12,
+      "eval_loss": 0.280333936214447,
+      "eval_runtime": 43.5973,
+      "eval_samples_per_second": 33.924,
+      "eval_steps_per_second": 33.924,
+      "step": 3720
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 13.200506210327148,
+      "learning_rate": 1.2542848551668841e-05,
+      "loss": 1.8569,
+      "step": 3721
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 11.850561141967773,
+      "learning_rate": 1.254084394106445e-05,
+      "loss": 1.2431,
+      "step": 3722
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 14.307988166809082,
+      "learning_rate": 1.253883933046006e-05,
+      "loss": 1.2535,
+      "step": 3723
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 33.0361328125,
+      "learning_rate": 1.2536834719855668e-05,
+      "loss": 2.3736,
+      "step": 3724
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 14.496783256530762,
+      "learning_rate": 1.253483010925128e-05,
+      "loss": 1.5165,
+      "step": 3725
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 20.50676155090332,
+      "learning_rate": 1.2532825498646888e-05,
+      "loss": 1.7852,
+      "step": 3726
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 16.414701461791992,
+      "learning_rate": 1.2530820888042498e-05,
+      "loss": 1.5958,
+      "step": 3727
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 9.845596313476562,
+      "learning_rate": 1.252881627743811e-05,
+      "loss": 1.4953,
+      "step": 3728
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 34.321861267089844,
+      "learning_rate": 1.2526811666833718e-05,
+      "loss": 1.9955,
+      "step": 3729
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 11.001937866210938,
+      "learning_rate": 1.2524807056229328e-05,
+      "loss": 1.7248,
+      "step": 3730
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 14.16842269897461,
+      "learning_rate": 1.2522802445624937e-05,
+      "loss": 0.9204,
+      "step": 3731
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 44.32683181762695,
+      "learning_rate": 1.2520797835020549e-05,
+      "loss": 1.5575,
+      "step": 3732
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 21.35757064819336,
+      "learning_rate": 1.2518793224416159e-05,
+      "loss": 2.7646,
+      "step": 3733
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 10.033564567565918,
+      "learning_rate": 1.2516788613811767e-05,
+      "loss": 1.054,
+      "step": 3734
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 11.14671802520752,
+      "learning_rate": 1.2514784003207379e-05,
+      "loss": 1.4729,
+      "step": 3735
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 57.583152770996094,
+      "learning_rate": 1.2512779392602987e-05,
+      "loss": 4.8937,
+      "step": 3736
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 10.513443946838379,
+      "learning_rate": 1.2510774781998597e-05,
+      "loss": 1.1766,
+      "step": 3737
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 20.811086654663086,
+      "learning_rate": 1.2508770171394207e-05,
+      "loss": 1.86,
+      "step": 3738
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 15.33969497680664,
+      "learning_rate": 1.2506765560789817e-05,
+      "loss": 1.9234,
+      "step": 3739
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 16.466957092285156,
+      "learning_rate": 1.250476095018543e-05,
+      "loss": 1.0944,
+      "step": 3740
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 11.51232624053955,
+      "learning_rate": 1.2502756339581038e-05,
+      "loss": 1.9455,
+      "step": 3741
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 16.893362045288086,
+      "learning_rate": 1.2500751728976648e-05,
+      "loss": 2.0934,
+      "step": 3742
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 10.207427024841309,
+      "learning_rate": 1.2498747118372256e-05,
+      "loss": 1.5444,
+      "step": 3743
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 14.303349494934082,
+      "learning_rate": 1.2496742507767868e-05,
+      "loss": 2.0237,
+      "step": 3744
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 15.821151733398438,
+      "learning_rate": 1.2494737897163478e-05,
+      "loss": 1.2218,
+      "step": 3745
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 19.704866409301758,
+      "learning_rate": 1.2492733286559086e-05,
+      "loss": 1.8249,
+      "step": 3746
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 15.844225883483887,
+      "learning_rate": 1.2490728675954698e-05,
+      "loss": 1.4068,
+      "step": 3747
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 26.448633193969727,
+      "learning_rate": 1.2488724065350306e-05,
+      "loss": 2.5801,
+      "step": 3748
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 31.50472640991211,
+      "learning_rate": 1.2486719454745917e-05,
+      "loss": 2.1224,
+      "step": 3749
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 11.211418151855469,
+      "learning_rate": 1.2484714844141525e-05,
+      "loss": 1.3734,
+      "step": 3750
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 16.562360763549805,
+      "learning_rate": 1.2482710233537137e-05,
+      "loss": 2.8084,
+      "step": 3751
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 14.236701965332031,
+      "learning_rate": 1.2480705622932747e-05,
+      "loss": 1.3424,
+      "step": 3752
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 17.27580451965332,
+      "learning_rate": 1.2478701012328355e-05,
+      "loss": 2.5758,
+      "step": 3753
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 12.679486274719238,
+      "learning_rate": 1.2476696401723967e-05,
+      "loss": 1.6972,
+      "step": 3754
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 17.295305252075195,
+      "learning_rate": 1.2474691791119575e-05,
+      "loss": 1.4085,
+      "step": 3755
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 14.256155014038086,
+      "learning_rate": 1.2472687180515185e-05,
+      "loss": 1.5598,
+      "step": 3756
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 14.317242622375488,
+      "learning_rate": 1.2470682569910797e-05,
+      "loss": 1.98,
+      "step": 3757
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 14.427403450012207,
+      "learning_rate": 1.2468677959306406e-05,
+      "loss": 1.4377,
+      "step": 3758
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 19.83202362060547,
+      "learning_rate": 1.2466673348702016e-05,
+      "loss": 1.7858,
+      "step": 3759
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 10.979972839355469,
+      "learning_rate": 1.2464668738097626e-05,
+      "loss": 0.8914,
+      "step": 3760
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 25.953853607177734,
+      "learning_rate": 1.2462664127493236e-05,
+      "loss": 1.768,
+      "step": 3761
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 17.417144775390625,
+      "learning_rate": 1.2460659516888844e-05,
+      "loss": 1.9067,
+      "step": 3762
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 18.727876663208008,
+      "learning_rate": 1.2458654906284456e-05,
+      "loss": 1.8154,
+      "step": 3763
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 28.132505416870117,
+      "learning_rate": 1.2456650295680066e-05,
+      "loss": 2.6384,
+      "step": 3764
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 11.409829139709473,
+      "learning_rate": 1.2454645685075675e-05,
+      "loss": 1.3703,
+      "step": 3765
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 12.39699935913086,
+      "learning_rate": 1.2452641074471286e-05,
+      "loss": 1.56,
+      "step": 3766
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 12.608708381652832,
+      "learning_rate": 1.2450636463866895e-05,
+      "loss": 1.0913,
+      "step": 3767
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 9.895720481872559,
+      "learning_rate": 1.2448631853262505e-05,
+      "loss": 1.6166,
+      "step": 3768
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 15.616667747497559,
+      "learning_rate": 1.2446627242658113e-05,
+      "loss": 2.4802,
+      "step": 3769
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 28.047149658203125,
+      "learning_rate": 1.2444622632053725e-05,
+      "loss": 2.0519,
+      "step": 3770
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 15.117725372314453,
+      "learning_rate": 1.2442618021449335e-05,
+      "loss": 2.0924,
+      "step": 3771
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 11.850608825683594,
+      "learning_rate": 1.2440613410844943e-05,
+      "loss": 1.4544,
+      "step": 3772
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 13.68972396850586,
+      "learning_rate": 1.2438608800240555e-05,
+      "loss": 2.0332,
+      "step": 3773
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 12.880284309387207,
+      "learning_rate": 1.2436604189636164e-05,
+      "loss": 2.0553,
+      "step": 3774
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 14.321048736572266,
+      "learning_rate": 1.2434599579031774e-05,
+      "loss": 1.2826,
+      "step": 3775
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 22.247779846191406,
+      "learning_rate": 1.2432594968427385e-05,
+      "loss": 1.7716,
+      "step": 3776
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 10.209996223449707,
+      "learning_rate": 1.2430590357822994e-05,
+      "loss": 1.5767,
+      "step": 3777
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 12.013636589050293,
+      "learning_rate": 1.2428585747218604e-05,
+      "loss": 1.114,
+      "step": 3778
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 51.70674514770508,
+      "learning_rate": 1.2426581136614212e-05,
+      "loss": 2.013,
+      "step": 3779
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 18.054279327392578,
+      "learning_rate": 1.2424576526009824e-05,
+      "loss": 1.4477,
+      "step": 3780
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 19.923124313354492,
+      "learning_rate": 1.2422571915405432e-05,
+      "loss": 1.3768,
+      "step": 3781
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 23.104625701904297,
+      "learning_rate": 1.2420567304801043e-05,
+      "loss": 2.1138,
+      "step": 3782
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 8.69080638885498,
+      "learning_rate": 1.2418562694196654e-05,
+      "loss": 1.1123,
+      "step": 3783
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 15.912267684936523,
+      "learning_rate": 1.2416558083592263e-05,
+      "loss": 2.4516,
+      "step": 3784
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 12.12271785736084,
+      "learning_rate": 1.2414553472987873e-05,
+      "loss": 1.4647,
+      "step": 3785
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 9.104388236999512,
+      "learning_rate": 1.2412548862383483e-05,
+      "loss": 1.8779,
+      "step": 3786
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 17.152158737182617,
+      "learning_rate": 1.2410544251779093e-05,
+      "loss": 1.4711,
+      "step": 3787
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 9.907875061035156,
+      "learning_rate": 1.2408539641174703e-05,
+      "loss": 1.1329,
+      "step": 3788
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 20.072593688964844,
+      "learning_rate": 1.2406535030570313e-05,
+      "loss": 1.9876,
+      "step": 3789
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 14.026973724365234,
+      "learning_rate": 1.2404530419965923e-05,
+      "loss": 0.7741,
+      "step": 3790
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 15.969365119934082,
+      "learning_rate": 1.2402525809361532e-05,
+      "loss": 1.6111,
+      "step": 3791
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 15.787557601928711,
+      "learning_rate": 1.2400521198757143e-05,
+      "loss": 1.5455,
+      "step": 3792
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 14.39417839050293,
+      "learning_rate": 1.2398516588152752e-05,
+      "loss": 1.0369,
+      "step": 3793
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 13.072067260742188,
+      "learning_rate": 1.2396511977548362e-05,
+      "loss": 2.3197,
+      "step": 3794
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 14.899435997009277,
+      "learning_rate": 1.2394507366943974e-05,
+      "loss": 2.7447,
+      "step": 3795
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 15.155303001403809,
+      "learning_rate": 1.2392502756339582e-05,
+      "loss": 1.4796,
+      "step": 3796
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 28.191987991333008,
+      "learning_rate": 1.2390498145735192e-05,
+      "loss": 1.7832,
+      "step": 3797
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 14.950464248657227,
+      "learning_rate": 1.23884935351308e-05,
+      "loss": 2.1336,
+      "step": 3798
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 21.477602005004883,
+      "learning_rate": 1.2386488924526412e-05,
+      "loss": 1.7525,
+      "step": 3799
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 25.496963500976562,
+      "learning_rate": 1.238448431392202e-05,
+      "loss": 1.6689,
+      "step": 3800
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 12.373186111450195,
+      "learning_rate": 1.238247970331763e-05,
+      "loss": 2.1899,
+      "step": 3801
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 20.684795379638672,
+      "learning_rate": 1.2380475092713243e-05,
+      "loss": 2.6932,
+      "step": 3802
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 9.77327823638916,
+      "learning_rate": 1.2378470482108851e-05,
+      "loss": 1.5063,
+      "step": 3803
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 28.525667190551758,
+      "learning_rate": 1.2376465871504461e-05,
+      "loss": 1.532,
+      "step": 3804
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 14.671927452087402,
+      "learning_rate": 1.2374461260900071e-05,
+      "loss": 2.0253,
+      "step": 3805
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 30.834918975830078,
+      "learning_rate": 1.2372456650295681e-05,
+      "loss": 2.39,
+      "step": 3806
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 16.612564086914062,
+      "learning_rate": 1.2370452039691291e-05,
+      "loss": 1.8819,
+      "step": 3807
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 13.142746925354004,
+      "learning_rate": 1.2368447429086901e-05,
+      "loss": 1.6061,
+      "step": 3808
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 18.517765045166016,
+      "learning_rate": 1.2366442818482511e-05,
+      "loss": 1.6218,
+      "step": 3809
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 19.177627563476562,
+      "learning_rate": 1.236443820787812e-05,
+      "loss": 2.3172,
+      "step": 3810
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 21.176328659057617,
+      "learning_rate": 1.2362433597273732e-05,
+      "loss": 1.8578,
+      "step": 3811
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 21.381567001342773,
+      "learning_rate": 1.236042898666934e-05,
+      "loss": 1.5194,
+      "step": 3812
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 11.9020414352417,
+      "learning_rate": 1.235842437606495e-05,
+      "loss": 1.6908,
+      "step": 3813
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 39.47920608520508,
+      "learning_rate": 1.2356419765460562e-05,
+      "loss": 1.8726,
+      "step": 3814
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 10.872967720031738,
+      "learning_rate": 1.235441515485617e-05,
+      "loss": 1.5569,
+      "step": 3815
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 14.20240592956543,
+      "learning_rate": 1.235241054425178e-05,
+      "loss": 1.9525,
+      "step": 3816
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 16.04691505432129,
+      "learning_rate": 1.2350405933647389e-05,
+      "loss": 1.4398,
+      "step": 3817
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 19.002132415771484,
+      "learning_rate": 1.2348401323043e-05,
+      "loss": 1.9253,
+      "step": 3818
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 9.962849617004395,
+      "learning_rate": 1.234639671243861e-05,
+      "loss": 2.0496,
+      "step": 3819
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 10.417901039123535,
+      "learning_rate": 1.2344392101834219e-05,
+      "loss": 1.9296,
+      "step": 3820
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 16.502777099609375,
+      "learning_rate": 1.234238749122983e-05,
+      "loss": 1.6108,
+      "step": 3821
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 18.25497817993164,
+      "learning_rate": 1.234038288062544e-05,
+      "loss": 1.7453,
+      "step": 3822
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 20.428895950317383,
+      "learning_rate": 1.233837827002105e-05,
+      "loss": 2.3424,
+      "step": 3823
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 13.153669357299805,
+      "learning_rate": 1.2336373659416658e-05,
+      "loss": 1.3317,
+      "step": 3824
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 13.739114761352539,
+      "learning_rate": 1.233436904881227e-05,
+      "loss": 1.4802,
+      "step": 3825
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 17.079265594482422,
+      "learning_rate": 1.233236443820788e-05,
+      "loss": 1.6825,
+      "step": 3826
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 15.86989974975586,
+      "learning_rate": 1.2330359827603488e-05,
+      "loss": 1.7154,
+      "step": 3827
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 14.82314682006836,
+      "learning_rate": 1.23283552169991e-05,
+      "loss": 1.3117,
+      "step": 3828
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 21.10222625732422,
+      "learning_rate": 1.2326350606394708e-05,
+      "loss": 1.4309,
+      "step": 3829
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 15.547961235046387,
+      "learning_rate": 1.2324345995790318e-05,
+      "loss": 1.3606,
+      "step": 3830
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 13.123003005981445,
+      "learning_rate": 1.232234138518593e-05,
+      "loss": 2.1664,
+      "step": 3831
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 14.233585357666016,
+      "learning_rate": 1.2320336774581538e-05,
+      "loss": 1.9827,
+      "step": 3832
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 8.499350547790527,
+      "learning_rate": 1.2318332163977148e-05,
+      "loss": 0.8647,
+      "step": 3833
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 24.76041603088379,
+      "learning_rate": 1.2316327553372758e-05,
+      "loss": 2.5669,
+      "step": 3834
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 11.308053970336914,
+      "learning_rate": 1.2314322942768369e-05,
+      "loss": 1.5768,
+      "step": 3835
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 25.03925132751465,
+      "learning_rate": 1.2312318332163977e-05,
+      "loss": 1.7572,
+      "step": 3836
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 19.462387084960938,
+      "learning_rate": 1.2310313721559589e-05,
+      "loss": 2.0888,
+      "step": 3837
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 12.561948776245117,
+      "learning_rate": 1.2308309110955199e-05,
+      "loss": 1.1124,
+      "step": 3838
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 34.78009796142578,
+      "learning_rate": 1.2306304500350807e-05,
+      "loss": 1.5277,
+      "step": 3839
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 26.024934768676758,
+      "learning_rate": 1.2304299889746419e-05,
+      "loss": 2.0164,
+      "step": 3840
+    },
+    {
+      "epoch": 1.15,
+      "eval_loss": 0.21982242166996002,
+      "eval_runtime": 43.5261,
+      "eval_samples_per_second": 33.98,
+      "eval_steps_per_second": 33.98,
+      "step": 3840
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 6.806481838226318,
+      "learning_rate": 1.2302295279142027e-05,
+      "loss": 0.7476,
+      "step": 3841
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 18.72743797302246,
+      "learning_rate": 1.2300290668537637e-05,
+      "loss": 1.4762,
+      "step": 3842
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 55.82676315307617,
+      "learning_rate": 1.2298286057933246e-05,
+      "loss": 1.809,
+      "step": 3843
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 13.187909126281738,
+      "learning_rate": 1.2296281447328858e-05,
+      "loss": 1.9105,
+      "step": 3844
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 11.694255828857422,
+      "learning_rate": 1.2294276836724468e-05,
+      "loss": 1.6571,
+      "step": 3845
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 20.324060440063477,
+      "learning_rate": 1.2292272226120076e-05,
+      "loss": 1.3941,
+      "step": 3846
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 13.865283012390137,
+      "learning_rate": 1.2290267615515688e-05,
+      "loss": 1.9631,
+      "step": 3847
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 18.18462562561035,
+      "learning_rate": 1.2288263004911296e-05,
+      "loss": 2.1097,
+      "step": 3848
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 12.211462020874023,
+      "learning_rate": 1.2286258394306906e-05,
+      "loss": 1.8528,
+      "step": 3849
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 12.799270629882812,
+      "learning_rate": 1.2284253783702518e-05,
+      "loss": 1.9839,
+      "step": 3850
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 36.99559020996094,
+      "learning_rate": 1.2282249173098127e-05,
+      "loss": 1.7886,
+      "step": 3851
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 15.477313995361328,
+      "learning_rate": 1.2280244562493737e-05,
+      "loss": 1.6013,
+      "step": 3852
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 12.97021198272705,
+      "learning_rate": 1.2278239951889345e-05,
+      "loss": 2.1934,
+      "step": 3853
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 9.071000099182129,
+      "learning_rate": 1.2276235341284957e-05,
+      "loss": 1.2502,
+      "step": 3854
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 13.090424537658691,
+      "learning_rate": 1.2274230730680565e-05,
+      "loss": 1.4692,
+      "step": 3855
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 15.393147468566895,
+      "learning_rate": 1.2272226120076175e-05,
+      "loss": 1.1286,
+      "step": 3856
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 38.615726470947266,
+      "learning_rate": 1.2270221509471787e-05,
+      "loss": 2.5387,
+      "step": 3857
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 28.15321922302246,
+      "learning_rate": 1.2268216898867395e-05,
+      "loss": 2.187,
+      "step": 3858
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 12.39357852935791,
+      "learning_rate": 1.2266212288263007e-05,
+      "loss": 2.2882,
+      "step": 3859
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 30.955167770385742,
+      "learning_rate": 1.2264207677658616e-05,
+      "loss": 1.8665,
+      "step": 3860
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 16.133106231689453,
+      "learning_rate": 1.2262203067054226e-05,
+      "loss": 1.8293,
+      "step": 3861
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 7.856118202209473,
+      "learning_rate": 1.2260198456449837e-05,
+      "loss": 0.9778,
+      "step": 3862
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 12.470169067382812,
+      "learning_rate": 1.2258193845845446e-05,
+      "loss": 1.4214,
+      "step": 3863
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 19.56519317626953,
+      "learning_rate": 1.2256189235241056e-05,
+      "loss": 2.2586,
+      "step": 3864
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 23.616268157958984,
+      "learning_rate": 1.2254184624636664e-05,
+      "loss": 2.7421,
+      "step": 3865
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 15.930326461791992,
+      "learning_rate": 1.2252180014032276e-05,
+      "loss": 2.1892,
+      "step": 3866
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 24.9573974609375,
+      "learning_rate": 1.2250175403427884e-05,
+      "loss": 2.7173,
+      "step": 3867
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 8.64990234375,
+      "learning_rate": 1.2248170792823495e-05,
+      "loss": 1.8587,
+      "step": 3868
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 29.846378326416016,
+      "learning_rate": 1.2246166182219106e-05,
+      "loss": 1.8677,
+      "step": 3869
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 11.002440452575684,
+      "learning_rate": 1.2244161571614715e-05,
+      "loss": 1.6396,
+      "step": 3870
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 16.57766342163086,
+      "learning_rate": 1.2242156961010325e-05,
+      "loss": 1.381,
+      "step": 3871
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 16.003095626831055,
+      "learning_rate": 1.2240152350405933e-05,
+      "loss": 1.7541,
+      "step": 3872
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 16.007123947143555,
+      "learning_rate": 1.2238147739801545e-05,
+      "loss": 1.9818,
+      "step": 3873
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 16.965221405029297,
+      "learning_rate": 1.2236143129197155e-05,
+      "loss": 2.3105,
+      "step": 3874
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 27.36821174621582,
+      "learning_rate": 1.2234138518592763e-05,
+      "loss": 2.4866,
+      "step": 3875
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 13.598553657531738,
+      "learning_rate": 1.2232133907988375e-05,
+      "loss": 1.654,
+      "step": 3876
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 15.706884384155273,
+      "learning_rate": 1.2230129297383984e-05,
+      "loss": 2.687,
+      "step": 3877
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 16.47588348388672,
+      "learning_rate": 1.2228124686779594e-05,
+      "loss": 1.6308,
+      "step": 3878
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 12.876527786254883,
+      "learning_rate": 1.2226120076175204e-05,
+      "loss": 1.4511,
+      "step": 3879
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 20.67323112487793,
+      "learning_rate": 1.2224115465570814e-05,
+      "loss": 1.4801,
+      "step": 3880
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 29.967750549316406,
+      "learning_rate": 1.2222110854966424e-05,
+      "loss": 2.2102,
+      "step": 3881
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 12.949405670166016,
+      "learning_rate": 1.2220106244362034e-05,
+      "loss": 1.8167,
+      "step": 3882
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 26.90522575378418,
+      "learning_rate": 1.2218101633757644e-05,
+      "loss": 1.7836,
+      "step": 3883
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 14.43256950378418,
+      "learning_rate": 1.2216097023153253e-05,
+      "loss": 1.5144,
+      "step": 3884
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 10.095649719238281,
+      "learning_rate": 1.2214092412548864e-05,
+      "loss": 0.8749,
+      "step": 3885
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 7.504431247711182,
+      "learning_rate": 1.2212087801944473e-05,
+      "loss": 1.5074,
+      "step": 3886
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 19.273103713989258,
+      "learning_rate": 1.2210083191340083e-05,
+      "loss": 2.6758,
+      "step": 3887
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 24.052509307861328,
+      "learning_rate": 1.2208078580735695e-05,
+      "loss": 1.6407,
+      "step": 3888
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 19.1132869720459,
+      "learning_rate": 1.2206073970131303e-05,
+      "loss": 1.7163,
+      "step": 3889
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 51.50567626953125,
+      "learning_rate": 1.2204069359526913e-05,
+      "loss": 2.9385,
+      "step": 3890
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 18.24009132385254,
+      "learning_rate": 1.2202064748922521e-05,
+      "loss": 1.7915,
+      "step": 3891
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 14.289188385009766,
+      "learning_rate": 1.2200060138318133e-05,
+      "loss": 1.8599,
+      "step": 3892
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 35.022804260253906,
+      "learning_rate": 1.2198055527713743e-05,
+      "loss": 2.9202,
+      "step": 3893
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 25.463573455810547,
+      "learning_rate": 1.2196050917109352e-05,
+      "loss": 1.622,
+      "step": 3894
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 10.089998245239258,
+      "learning_rate": 1.2194046306504963e-05,
+      "loss": 1.1327,
+      "step": 3895
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 23.384336471557617,
+      "learning_rate": 1.2192041695900572e-05,
+      "loss": 1.2661,
+      "step": 3896
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 12.663224220275879,
+      "learning_rate": 1.2190037085296182e-05,
+      "loss": 2.6986,
+      "step": 3897
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 16.6176700592041,
+      "learning_rate": 1.218803247469179e-05,
+      "loss": 1.894,
+      "step": 3898
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 12.314738273620605,
+      "learning_rate": 1.2186027864087402e-05,
+      "loss": 1.2068,
+      "step": 3899
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 23.991886138916016,
+      "learning_rate": 1.2184023253483012e-05,
+      "loss": 1.6332,
+      "step": 3900
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 23.119836807250977,
+      "learning_rate": 1.218201864287862e-05,
+      "loss": 2.1423,
+      "step": 3901
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 29.719497680664062,
+      "learning_rate": 1.2180014032274232e-05,
+      "loss": 1.4554,
+      "step": 3902
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 13.610468864440918,
+      "learning_rate": 1.217800942166984e-05,
+      "loss": 1.6478,
+      "step": 3903
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 12.19708251953125,
+      "learning_rate": 1.217600481106545e-05,
+      "loss": 1.4697,
+      "step": 3904
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 14.37850570678711,
+      "learning_rate": 1.2174000200461063e-05,
+      "loss": 1.2426,
+      "step": 3905
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 16.169696807861328,
+      "learning_rate": 1.2171995589856671e-05,
+      "loss": 1.7116,
+      "step": 3906
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 14.644316673278809,
+      "learning_rate": 1.2169990979252281e-05,
+      "loss": 1.5071,
+      "step": 3907
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 10.753741264343262,
+      "learning_rate": 1.2167986368647891e-05,
+      "loss": 1.8289,
+      "step": 3908
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 17.671937942504883,
+      "learning_rate": 1.2165981758043501e-05,
+      "loss": 1.3769,
+      "step": 3909
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 10.696758270263672,
+      "learning_rate": 1.216397714743911e-05,
+      "loss": 1.7544,
+      "step": 3910
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 10.894145011901855,
+      "learning_rate": 1.2161972536834721e-05,
+      "loss": 1.3638,
+      "step": 3911
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 14.590895652770996,
+      "learning_rate": 1.2159967926230332e-05,
+      "loss": 1.4625,
+      "step": 3912
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 20.419675827026367,
+      "learning_rate": 1.215796331562594e-05,
+      "loss": 2.223,
+      "step": 3913
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 29.404043197631836,
+      "learning_rate": 1.2155958705021552e-05,
+      "loss": 1.6396,
+      "step": 3914
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 24.126564025878906,
+      "learning_rate": 1.215395409441716e-05,
+      "loss": 1.6138,
+      "step": 3915
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 13.458683967590332,
+      "learning_rate": 1.215194948381277e-05,
+      "loss": 2.0191,
+      "step": 3916
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 9.993988037109375,
+      "learning_rate": 1.2149944873208382e-05,
+      "loss": 1.3779,
+      "step": 3917
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 13.048202514648438,
+      "learning_rate": 1.214794026260399e-05,
+      "loss": 1.8113,
+      "step": 3918
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 11.234829902648926,
+      "learning_rate": 1.21459356519996e-05,
+      "loss": 1.4256,
+      "step": 3919
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 25.56989860534668,
+      "learning_rate": 1.2143931041395209e-05,
+      "loss": 2.0883,
+      "step": 3920
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 12.80074405670166,
+      "learning_rate": 1.214192643079082e-05,
+      "loss": 1.7903,
+      "step": 3921
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 26.22979736328125,
+      "learning_rate": 1.2139921820186429e-05,
+      "loss": 2.3038,
+      "step": 3922
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 15.948308944702148,
+      "learning_rate": 1.2137917209582039e-05,
+      "loss": 1.9847,
+      "step": 3923
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 20.231281280517578,
+      "learning_rate": 1.213591259897765e-05,
+      "loss": 2.6807,
+      "step": 3924
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 26.88908576965332,
+      "learning_rate": 1.213390798837326e-05,
+      "loss": 2.2021,
+      "step": 3925
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 9.647951126098633,
+      "learning_rate": 1.213190337776887e-05,
+      "loss": 1.7718,
+      "step": 3926
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 25.476966857910156,
+      "learning_rate": 1.212989876716448e-05,
+      "loss": 1.7512,
+      "step": 3927
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 17.493688583374023,
+      "learning_rate": 1.212789415656009e-05,
+      "loss": 2.0739,
+      "step": 3928
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 18.608076095581055,
+      "learning_rate": 1.2125889545955698e-05,
+      "loss": 1.8991,
+      "step": 3929
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 26.32684326171875,
+      "learning_rate": 1.212388493535131e-05,
+      "loss": 1.3218,
+      "step": 3930
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 14.290658950805664,
+      "learning_rate": 1.212188032474692e-05,
+      "loss": 1.6064,
+      "step": 3931
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 8.176145553588867,
+      "learning_rate": 1.2119875714142528e-05,
+      "loss": 1.101,
+      "step": 3932
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 77.52426147460938,
+      "learning_rate": 1.211787110353814e-05,
+      "loss": 1.8667,
+      "step": 3933
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 13.218364715576172,
+      "learning_rate": 1.2115866492933748e-05,
+      "loss": 2.0845,
+      "step": 3934
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 13.180416107177734,
+      "learning_rate": 1.2113861882329358e-05,
+      "loss": 1.7175,
+      "step": 3935
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 27.077486038208008,
+      "learning_rate": 1.211185727172497e-05,
+      "loss": 1.5133,
+      "step": 3936
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 14.540903091430664,
+      "learning_rate": 1.2109852661120579e-05,
+      "loss": 2.1343,
+      "step": 3937
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 18.087007522583008,
+      "learning_rate": 1.2107848050516189e-05,
+      "loss": 1.6534,
+      "step": 3938
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 18.4873046875,
+      "learning_rate": 1.2105843439911797e-05,
+      "loss": 1.8601,
+      "step": 3939
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 11.12307071685791,
+      "learning_rate": 1.2103838829307409e-05,
+      "loss": 1.4969,
+      "step": 3940
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 22.688817977905273,
+      "learning_rate": 1.2101834218703017e-05,
+      "loss": 2.0138,
+      "step": 3941
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 32.38698196411133,
+      "learning_rate": 1.2099829608098627e-05,
+      "loss": 3.5612,
+      "step": 3942
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 40.410884857177734,
+      "learning_rate": 1.2097824997494239e-05,
+      "loss": 2.6222,
+      "step": 3943
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 11.373454093933105,
+      "learning_rate": 1.2095820386889847e-05,
+      "loss": 1.8325,
+      "step": 3944
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 57.750003814697266,
+      "learning_rate": 1.2093815776285458e-05,
+      "loss": 2.7973,
+      "step": 3945
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 18.441944122314453,
+      "learning_rate": 1.2091811165681066e-05,
+      "loss": 2.9912,
+      "step": 3946
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 32.177513122558594,
+      "learning_rate": 1.2089806555076678e-05,
+      "loss": 1.9056,
+      "step": 3947
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 22.477294921875,
+      "learning_rate": 1.2087801944472288e-05,
+      "loss": 2.2548,
+      "step": 3948
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 20.485315322875977,
+      "learning_rate": 1.2085797333867896e-05,
+      "loss": 1.4721,
+      "step": 3949
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 14.799079895019531,
+      "learning_rate": 1.2083792723263508e-05,
+      "loss": 1.8698,
+      "step": 3950
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 13.152766227722168,
+      "learning_rate": 1.2081788112659116e-05,
+      "loss": 2.6729,
+      "step": 3951
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 10.631035804748535,
+      "learning_rate": 1.2079783502054726e-05,
+      "loss": 1.7254,
+      "step": 3952
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 15.786643981933594,
+      "learning_rate": 1.2077778891450336e-05,
+      "loss": 1.1798,
+      "step": 3953
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 12.110862731933594,
+      "learning_rate": 1.2075774280845947e-05,
+      "loss": 1.4401,
+      "step": 3954
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 15.332069396972656,
+      "learning_rate": 1.2073769670241557e-05,
+      "loss": 2.4265,
+      "step": 3955
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 26.413911819458008,
+      "learning_rate": 1.2071765059637167e-05,
+      "loss": 1.7556,
+      "step": 3956
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 10.954570770263672,
+      "learning_rate": 1.2069760449032777e-05,
+      "loss": 0.8213,
+      "step": 3957
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 20.662534713745117,
+      "learning_rate": 1.2067755838428385e-05,
+      "loss": 1.6746,
+      "step": 3958
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 9.891891479492188,
+      "learning_rate": 1.2065751227823997e-05,
+      "loss": 1.3557,
+      "step": 3959
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 13.939960479736328,
+      "learning_rate": 1.2063746617219607e-05,
+      "loss": 1.6665,
+      "step": 3960
+    },
+    {
+      "epoch": 1.19,
+      "eval_loss": 0.22422289848327637,
+      "eval_runtime": 43.495,
+      "eval_samples_per_second": 34.004,
+      "eval_steps_per_second": 34.004,
+      "step": 3960
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 50.82015609741211,
+      "learning_rate": 1.2061742006615215e-05,
+      "loss": 2.8102,
+      "step": 3961
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 10.213366508483887,
+      "learning_rate": 1.2059737396010827e-05,
+      "loss": 1.7045,
+      "step": 3962
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 9.090950965881348,
+      "learning_rate": 1.2057732785406436e-05,
+      "loss": 1.0682,
+      "step": 3963
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 14.20166015625,
+      "learning_rate": 1.2055728174802046e-05,
+      "loss": 1.6642,
+      "step": 3964
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 14.564234733581543,
+      "learning_rate": 1.2053723564197654e-05,
+      "loss": 1.6248,
+      "step": 3965
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 16.231090545654297,
+      "learning_rate": 1.2051718953593266e-05,
+      "loss": 1.2251,
+      "step": 3966
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 15.789569854736328,
+      "learning_rate": 1.2049714342988876e-05,
+      "loss": 2.4118,
+      "step": 3967
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 18.06235122680664,
+      "learning_rate": 1.2047709732384484e-05,
+      "loss": 1.4758,
+      "step": 3968
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 18.72103500366211,
+      "learning_rate": 1.2045705121780096e-05,
+      "loss": 1.416,
+      "step": 3969
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 67.29039001464844,
+      "learning_rate": 1.2043700511175705e-05,
+      "loss": 3.1432,
+      "step": 3970
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 9.234892845153809,
+      "learning_rate": 1.2041695900571315e-05,
+      "loss": 1.0072,
+      "step": 3971
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 16.655811309814453,
+      "learning_rate": 1.2039691289966923e-05,
+      "loss": 1.4894,
+      "step": 3972
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 24.546369552612305,
+      "learning_rate": 1.2037686679362535e-05,
+      "loss": 1.6944,
+      "step": 3973
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 22.591285705566406,
+      "learning_rate": 1.2035682068758145e-05,
+      "loss": 1.4835,
+      "step": 3974
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 19.460481643676758,
+      "learning_rate": 1.2033677458153753e-05,
+      "loss": 2.3772,
+      "step": 3975
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 10.695039749145508,
+      "learning_rate": 1.2031672847549365e-05,
+      "loss": 1.3966,
+      "step": 3976
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 16.96231460571289,
+      "learning_rate": 1.2029668236944973e-05,
+      "loss": 1.5844,
+      "step": 3977
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 13.368218421936035,
+      "learning_rate": 1.2027663626340585e-05,
+      "loss": 2.2137,
+      "step": 3978
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 16.881895065307617,
+      "learning_rate": 1.2025659015736195e-05,
+      "loss": 1.809,
+      "step": 3979
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 16.86216926574707,
+      "learning_rate": 1.2023654405131804e-05,
+      "loss": 1.7264,
+      "step": 3980
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 19.68351173400879,
+      "learning_rate": 1.2021649794527415e-05,
+      "loss": 1.4559,
+      "step": 3981
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 27.27216339111328,
+      "learning_rate": 1.2019645183923024e-05,
+      "loss": 1.6525,
+      "step": 3982
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 48.12042999267578,
+      "learning_rate": 1.2017640573318634e-05,
+      "loss": 2.7933,
+      "step": 3983
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 17.5889892578125,
+      "learning_rate": 1.2015635962714242e-05,
+      "loss": 1.2232,
+      "step": 3984
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 15.102045059204102,
+      "learning_rate": 1.2013631352109854e-05,
+      "loss": 2.1175,
+      "step": 3985
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 15.093314170837402,
+      "learning_rate": 1.2011626741505464e-05,
+      "loss": 2.0942,
+      "step": 3986
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 14.149928092956543,
+      "learning_rate": 1.2009622130901073e-05,
+      "loss": 1.4488,
+      "step": 3987
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 23.17877197265625,
+      "learning_rate": 1.2007617520296684e-05,
+      "loss": 2.1352,
+      "step": 3988
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 11.108948707580566,
+      "learning_rate": 1.2005612909692293e-05,
+      "loss": 1.4723,
+      "step": 3989
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 41.934715270996094,
+      "learning_rate": 1.2003608299087903e-05,
+      "loss": 1.3572,
+      "step": 3990
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 30.752838134765625,
+      "learning_rate": 1.2001603688483515e-05,
+      "loss": 1.7268,
+      "step": 3991
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 19.63275718688965,
+      "learning_rate": 1.1999599077879123e-05,
+      "loss": 1.8781,
+      "step": 3992
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 16.867294311523438,
+      "learning_rate": 1.1997594467274733e-05,
+      "loss": 1.2249,
+      "step": 3993
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 12.371650695800781,
+      "learning_rate": 1.1995589856670341e-05,
+      "loss": 2.7085,
+      "step": 3994
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 14.14538860321045,
+      "learning_rate": 1.1993585246065953e-05,
+      "loss": 1.7304,
+      "step": 3995
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 28.63239860534668,
+      "learning_rate": 1.1991580635461562e-05,
+      "loss": 1.6735,
+      "step": 3996
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 10.542112350463867,
+      "learning_rate": 1.1989576024857172e-05,
+      "loss": 2.628,
+      "step": 3997
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 15.562253952026367,
+      "learning_rate": 1.1987571414252784e-05,
+      "loss": 1.8914,
+      "step": 3998
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 14.955403327941895,
+      "learning_rate": 1.1985566803648392e-05,
+      "loss": 1.4507,
+      "step": 3999
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 8.927510261535645,
+      "learning_rate": 1.1983562193044002e-05,
+      "loss": 1.8247,
+      "step": 4000
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 40.698150634765625,
+      "learning_rate": 1.1981557582439612e-05,
+      "loss": 2.7692,
+      "step": 4001
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 8.478541374206543,
+      "learning_rate": 1.1979552971835222e-05,
+      "loss": 1.0191,
+      "step": 4002
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 29.216922760009766,
+      "learning_rate": 1.1977548361230832e-05,
+      "loss": 1.7092,
+      "step": 4003
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 11.635241508483887,
+      "learning_rate": 1.1975543750626442e-05,
+      "loss": 1.7223,
+      "step": 4004
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 23.52641487121582,
+      "learning_rate": 1.1973539140022052e-05,
+      "loss": 1.6989,
+      "step": 4005
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 18.005016326904297,
+      "learning_rate": 1.197153452941766e-05,
+      "loss": 2.5197,
+      "step": 4006
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 59.00607681274414,
+      "learning_rate": 1.1969529918813273e-05,
+      "loss": 3.0074,
+      "step": 4007
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 23.461963653564453,
+      "learning_rate": 1.1967525308208881e-05,
+      "loss": 2.5943,
+      "step": 4008
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 30.815887451171875,
+      "learning_rate": 1.1965520697604491e-05,
+      "loss": 1.9424,
+      "step": 4009
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 29.489526748657227,
+      "learning_rate": 1.1963516087000103e-05,
+      "loss": 2.0126,
+      "step": 4010
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 7.134322166442871,
+      "learning_rate": 1.1961511476395711e-05,
+      "loss": 1.2801,
+      "step": 4011
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 12.04995059967041,
+      "learning_rate": 1.1959506865791321e-05,
+      "loss": 1.0723,
+      "step": 4012
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 13.849966049194336,
+      "learning_rate": 1.195750225518693e-05,
+      "loss": 1.0122,
+      "step": 4013
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 7.143357753753662,
+      "learning_rate": 1.1955497644582541e-05,
+      "loss": 1.961,
+      "step": 4014
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 16.170339584350586,
+      "learning_rate": 1.195349303397815e-05,
+      "loss": 1.9324,
+      "step": 4015
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 22.960071563720703,
+      "learning_rate": 1.195148842337376e-05,
+      "loss": 2.5535,
+      "step": 4016
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 15.098960876464844,
+      "learning_rate": 1.1949483812769372e-05,
+      "loss": 1.142,
+      "step": 4017
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 19.63709831237793,
+      "learning_rate": 1.194747920216498e-05,
+      "loss": 2.072,
+      "step": 4018
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 14.952527046203613,
+      "learning_rate": 1.194547459156059e-05,
+      "loss": 1.2427,
+      "step": 4019
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 13.70450210571289,
+      "learning_rate": 1.1943469980956199e-05,
+      "loss": 1.8315,
+      "step": 4020
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 13.074018478393555,
+      "learning_rate": 1.194146537035181e-05,
+      "loss": 0.9992,
+      "step": 4021
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 12.277508735656738,
+      "learning_rate": 1.193946075974742e-05,
+      "loss": 1.8067,
+      "step": 4022
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 27.505800247192383,
+      "learning_rate": 1.1937456149143029e-05,
+      "loss": 2.1141,
+      "step": 4023
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 18.097551345825195,
+      "learning_rate": 1.193545153853864e-05,
+      "loss": 1.631,
+      "step": 4024
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 15.072179794311523,
+      "learning_rate": 1.1933446927934249e-05,
+      "loss": 1.5262,
+      "step": 4025
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 13.110054016113281,
+      "learning_rate": 1.1931442317329859e-05,
+      "loss": 1.4835,
+      "step": 4026
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 11.2367525100708,
+      "learning_rate": 1.192943770672547e-05,
+      "loss": 1.5453,
+      "step": 4027
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 19.51766014099121,
+      "learning_rate": 1.192743309612108e-05,
+      "loss": 1.4438,
+      "step": 4028
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 58.76639938354492,
+      "learning_rate": 1.1925428485516691e-05,
+      "loss": 2.2727,
+      "step": 4029
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 16.14882469177246,
+      "learning_rate": 1.19234238749123e-05,
+      "loss": 1.4155,
+      "step": 4030
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 9.466064453125,
+      "learning_rate": 1.192141926430791e-05,
+      "loss": 1.3346,
+      "step": 4031
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 16.83454704284668,
+      "learning_rate": 1.1919414653703518e-05,
+      "loss": 1.3689,
+      "step": 4032
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 11.721967697143555,
+      "learning_rate": 1.191741004309913e-05,
+      "loss": 1.5468,
+      "step": 4033
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 9.986775398254395,
+      "learning_rate": 1.191540543249474e-05,
+      "loss": 2.5393,
+      "step": 4034
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 11.999430656433105,
+      "learning_rate": 1.1913400821890348e-05,
+      "loss": 1.7203,
+      "step": 4035
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 15.086469650268555,
+      "learning_rate": 1.191139621128596e-05,
+      "loss": 1.8978,
+      "step": 4036
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 20.604961395263672,
+      "learning_rate": 1.1909391600681568e-05,
+      "loss": 1.0557,
+      "step": 4037
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 11.977767944335938,
+      "learning_rate": 1.1907386990077178e-05,
+      "loss": 1.9318,
+      "step": 4038
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 17.97518539428711,
+      "learning_rate": 1.1905382379472787e-05,
+      "loss": 0.948,
+      "step": 4039
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 9.713574409484863,
+      "learning_rate": 1.1903377768868399e-05,
+      "loss": 1.0446,
+      "step": 4040
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 18.080841064453125,
+      "learning_rate": 1.1901373158264009e-05,
+      "loss": 1.703,
+      "step": 4041
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 35.3074951171875,
+      "learning_rate": 1.1899368547659617e-05,
+      "loss": 2.8692,
+      "step": 4042
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 12.872424125671387,
+      "learning_rate": 1.1897363937055229e-05,
+      "loss": 1.2558,
+      "step": 4043
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 20.24639129638672,
+      "learning_rate": 1.1895359326450837e-05,
+      "loss": 2.013,
+      "step": 4044
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 19.4423885345459,
+      "learning_rate": 1.1893354715846447e-05,
+      "loss": 1.2018,
+      "step": 4045
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 28.46637725830078,
+      "learning_rate": 1.1891350105242059e-05,
+      "loss": 2.3941,
+      "step": 4046
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 15.011366844177246,
+      "learning_rate": 1.1889345494637667e-05,
+      "loss": 1.5718,
+      "step": 4047
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 30.27291488647461,
+      "learning_rate": 1.1887340884033278e-05,
+      "loss": 2.3006,
+      "step": 4048
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 18.10991859436035,
+      "learning_rate": 1.1885336273428888e-05,
+      "loss": 1.7672,
+      "step": 4049
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 34.54472732543945,
+      "learning_rate": 1.1883331662824498e-05,
+      "loss": 2.484,
+      "step": 4050
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 12.643610954284668,
+      "learning_rate": 1.1881327052220106e-05,
+      "loss": 2.0557,
+      "step": 4051
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 12.419498443603516,
+      "learning_rate": 1.1879322441615718e-05,
+      "loss": 1.8953,
+      "step": 4052
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 9.909022331237793,
+      "learning_rate": 1.1877317831011328e-05,
+      "loss": 0.9697,
+      "step": 4053
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 213.60519409179688,
+      "learning_rate": 1.1875313220406936e-05,
+      "loss": 2.7859,
+      "step": 4054
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 43.98239517211914,
+      "learning_rate": 1.1873308609802548e-05,
+      "loss": 2.0812,
+      "step": 4055
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 22.185237884521484,
+      "learning_rate": 1.1871303999198157e-05,
+      "loss": 1.778,
+      "step": 4056
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 20.283220291137695,
+      "learning_rate": 1.1869299388593767e-05,
+      "loss": 1.6555,
+      "step": 4057
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 52.456600189208984,
+      "learning_rate": 1.1867294777989375e-05,
+      "loss": 2.7659,
+      "step": 4058
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 27.247201919555664,
+      "learning_rate": 1.1865290167384987e-05,
+      "loss": 1.9769,
+      "step": 4059
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 19.406150817871094,
+      "learning_rate": 1.1863285556780597e-05,
+      "loss": 2.0779,
+      "step": 4060
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 7.287651062011719,
+      "learning_rate": 1.1861280946176205e-05,
+      "loss": 1.3887,
+      "step": 4061
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 12.292468070983887,
+      "learning_rate": 1.1859276335571817e-05,
+      "loss": 1.7488,
+      "step": 4062
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 9.17116928100586,
+      "learning_rate": 1.1857271724967425e-05,
+      "loss": 1.2331,
+      "step": 4063
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 12.708617210388184,
+      "learning_rate": 1.1855267114363036e-05,
+      "loss": 1.5956,
+      "step": 4064
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 19.217905044555664,
+      "learning_rate": 1.1853262503758647e-05,
+      "loss": 3.3333,
+      "step": 4065
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 9.486623764038086,
+      "learning_rate": 1.1851257893154256e-05,
+      "loss": 1.1654,
+      "step": 4066
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 15.519346237182617,
+      "learning_rate": 1.1849253282549866e-05,
+      "loss": 1.3905,
+      "step": 4067
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 27.3670597076416,
+      "learning_rate": 1.1847248671945474e-05,
+      "loss": 1.8644,
+      "step": 4068
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 12.652219772338867,
+      "learning_rate": 1.1845244061341086e-05,
+      "loss": 1.7432,
+      "step": 4069
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 7.705514430999756,
+      "learning_rate": 1.1843239450736694e-05,
+      "loss": 0.8268,
+      "step": 4070
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 13.351577758789062,
+      "learning_rate": 1.1841234840132304e-05,
+      "loss": 2.1508,
+      "step": 4071
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 44.65596008300781,
+      "learning_rate": 1.1839230229527916e-05,
+      "loss": 2.0205,
+      "step": 4072
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 24.321109771728516,
+      "learning_rate": 1.1837225618923525e-05,
+      "loss": 1.7858,
+      "step": 4073
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 48.96630096435547,
+      "learning_rate": 1.1835221008319135e-05,
+      "loss": 1.9235,
+      "step": 4074
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 16.92742156982422,
+      "learning_rate": 1.1833216397714745e-05,
+      "loss": 1.7837,
+      "step": 4075
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 42.10576629638672,
+      "learning_rate": 1.1831211787110355e-05,
+      "loss": 3.2812,
+      "step": 4076
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 13.819901466369629,
+      "learning_rate": 1.1829207176505965e-05,
+      "loss": 1.4294,
+      "step": 4077
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 11.949419021606445,
+      "learning_rate": 1.1827202565901575e-05,
+      "loss": 1.3175,
+      "step": 4078
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 23.273836135864258,
+      "learning_rate": 1.1825197955297185e-05,
+      "loss": 2.5785,
+      "step": 4079
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 15.00057601928711,
+      "learning_rate": 1.1823193344692793e-05,
+      "loss": 1.4673,
+      "step": 4080
+    },
+    {
+      "epoch": 1.23,
+      "eval_loss": 0.22429148852825165,
+      "eval_runtime": 43.6464,
+      "eval_samples_per_second": 33.886,
+      "eval_steps_per_second": 33.886,
+      "step": 4080
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 16.4818172454834,
+      "learning_rate": 1.1821188734088405e-05,
+      "loss": 1.8456,
+      "step": 4081
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 9.311663627624512,
+      "learning_rate": 1.1819184123484014e-05,
+      "loss": 1.5563,
+      "step": 4082
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 19.122093200683594,
+      "learning_rate": 1.1817179512879624e-05,
+      "loss": 2.2633,
+      "step": 4083
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 33.5509033203125,
+      "learning_rate": 1.1815174902275236e-05,
+      "loss": 1.9949,
+      "step": 4084
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 14.471306800842285,
+      "learning_rate": 1.1813170291670844e-05,
+      "loss": 1.8224,
+      "step": 4085
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 11.821321487426758,
+      "learning_rate": 1.1811165681066454e-05,
+      "loss": 1.8433,
+      "step": 4086
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 26.970781326293945,
+      "learning_rate": 1.1809161070462062e-05,
+      "loss": 2.336,
+      "step": 4087
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 18.373268127441406,
+      "learning_rate": 1.1807156459857674e-05,
+      "loss": 1.4686,
+      "step": 4088
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 17.756778717041016,
+      "learning_rate": 1.1805151849253284e-05,
+      "loss": 1.4351,
+      "step": 4089
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 21.345794677734375,
+      "learning_rate": 1.1803147238648893e-05,
+      "loss": 1.7058,
+      "step": 4090
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 99.40449523925781,
+      "learning_rate": 1.1801142628044504e-05,
+      "loss": 2.2699,
+      "step": 4091
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 29.920364379882812,
+      "learning_rate": 1.1799138017440113e-05,
+      "loss": 1.7795,
+      "step": 4092
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 16.23194122314453,
+      "learning_rate": 1.1797133406835723e-05,
+      "loss": 1.8951,
+      "step": 4093
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 16.518890380859375,
+      "learning_rate": 1.1795128796231331e-05,
+      "loss": 2.0618,
+      "step": 4094
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 15.834227561950684,
+      "learning_rate": 1.1793124185626943e-05,
+      "loss": 1.3648,
+      "step": 4095
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 21.73334503173828,
+      "learning_rate": 1.1791119575022553e-05,
+      "loss": 1.6835,
+      "step": 4096
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 16.20547103881836,
+      "learning_rate": 1.1789114964418163e-05,
+      "loss": 2.8748,
+      "step": 4097
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 13.940154075622559,
+      "learning_rate": 1.1787110353813773e-05,
+      "loss": 1.7865,
+      "step": 4098
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 10.902242660522461,
+      "learning_rate": 1.1785105743209382e-05,
+      "loss": 1.4294,
+      "step": 4099
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 20.06895637512207,
+      "learning_rate": 1.1783101132604993e-05,
+      "loss": 2.0791,
+      "step": 4100
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 21.698528289794922,
+      "learning_rate": 1.1781096522000602e-05,
+      "loss": 1.8301,
+      "step": 4101
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 50.30092239379883,
+      "learning_rate": 1.1779091911396212e-05,
+      "loss": 2.0178,
+      "step": 4102
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 29.254417419433594,
+      "learning_rate": 1.1777087300791824e-05,
+      "loss": 1.7604,
+      "step": 4103
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 13.801785469055176,
+      "learning_rate": 1.1775082690187432e-05,
+      "loss": 1.9442,
+      "step": 4104
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 14.859130859375,
+      "learning_rate": 1.1773078079583042e-05,
+      "loss": 2.0129,
+      "step": 4105
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 14.485882759094238,
+      "learning_rate": 1.177107346897865e-05,
+      "loss": 1.7607,
+      "step": 4106
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 16.90434455871582,
+      "learning_rate": 1.1769068858374262e-05,
+      "loss": 1.5387,
+      "step": 4107
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 13.56885051727295,
+      "learning_rate": 1.1767064247769872e-05,
+      "loss": 1.2367,
+      "step": 4108
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 14.099568367004395,
+      "learning_rate": 1.176505963716548e-05,
+      "loss": 0.9839,
+      "step": 4109
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 22.063432693481445,
+      "learning_rate": 1.1763055026561093e-05,
+      "loss": 1.3803,
+      "step": 4110
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 16.2562255859375,
+      "learning_rate": 1.1761050415956701e-05,
+      "loss": 1.5013,
+      "step": 4111
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 26.141794204711914,
+      "learning_rate": 1.1759045805352311e-05,
+      "loss": 2.49,
+      "step": 4112
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 16.62359619140625,
+      "learning_rate": 1.175704119474792e-05,
+      "loss": 0.9543,
+      "step": 4113
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 21.430021286010742,
+      "learning_rate": 1.1755036584143531e-05,
+      "loss": 1.2435,
+      "step": 4114
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 15.07040786743164,
+      "learning_rate": 1.1753031973539141e-05,
+      "loss": 0.9931,
+      "step": 4115
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 9.915048599243164,
+      "learning_rate": 1.175102736293475e-05,
+      "loss": 1.498,
+      "step": 4116
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 16.46501922607422,
+      "learning_rate": 1.1749022752330362e-05,
+      "loss": 2.0349,
+      "step": 4117
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 16.73762321472168,
+      "learning_rate": 1.174701814172597e-05,
+      "loss": 1.9466,
+      "step": 4118
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 19.883028030395508,
+      "learning_rate": 1.174501353112158e-05,
+      "loss": 1.8937,
+      "step": 4119
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 47.894859313964844,
+      "learning_rate": 1.1743008920517192e-05,
+      "loss": 1.9303,
+      "step": 4120
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 14.77435302734375,
+      "learning_rate": 1.17410043099128e-05,
+      "loss": 1.8005,
+      "step": 4121
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 9.72576904296875,
+      "learning_rate": 1.173899969930841e-05,
+      "loss": 1.5267,
+      "step": 4122
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 27.231468200683594,
+      "learning_rate": 1.173699508870402e-05,
+      "loss": 1.5438,
+      "step": 4123
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 14.572410583496094,
+      "learning_rate": 1.173499047809963e-05,
+      "loss": 1.7702,
+      "step": 4124
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 14.047436714172363,
+      "learning_rate": 1.1732985867495239e-05,
+      "loss": 1.8277,
+      "step": 4125
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 20.151954650878906,
+      "learning_rate": 1.173098125689085e-05,
+      "loss": 1.7252,
+      "step": 4126
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 13.929177284240723,
+      "learning_rate": 1.172897664628646e-05,
+      "loss": 2.2859,
+      "step": 4127
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 16.374547958374023,
+      "learning_rate": 1.1726972035682069e-05,
+      "loss": 1.4863,
+      "step": 4128
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 20.026716232299805,
+      "learning_rate": 1.172496742507768e-05,
+      "loss": 1.4962,
+      "step": 4129
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 14.383737564086914,
+      "learning_rate": 1.172296281447329e-05,
+      "loss": 1.8605,
+      "step": 4130
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 8.625479698181152,
+      "learning_rate": 1.17209582038689e-05,
+      "loss": 1.3414,
+      "step": 4131
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 9.765339851379395,
+      "learning_rate": 1.1718953593264511e-05,
+      "loss": 2.2172,
+      "step": 4132
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 17.315092086791992,
+      "learning_rate": 1.171694898266012e-05,
+      "loss": 1.8639,
+      "step": 4133
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 28.305118560791016,
+      "learning_rate": 1.171494437205573e-05,
+      "loss": 1.2452,
+      "step": 4134
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 23.19142723083496,
+      "learning_rate": 1.1712939761451338e-05,
+      "loss": 1.5799,
+      "step": 4135
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 38.168418884277344,
+      "learning_rate": 1.171093515084695e-05,
+      "loss": 2.4729,
+      "step": 4136
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 12.799455642700195,
+      "learning_rate": 1.1708930540242558e-05,
+      "loss": 0.9976,
+      "step": 4137
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 11.151683807373047,
+      "learning_rate": 1.1706925929638168e-05,
+      "loss": 2.1855,
+      "step": 4138
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 55.58821487426758,
+      "learning_rate": 1.170492131903378e-05,
+      "loss": 1.4985,
+      "step": 4139
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 12.804335594177246,
+      "learning_rate": 1.1702916708429388e-05,
+      "loss": 2.2843,
+      "step": 4140
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 18.690269470214844,
+      "learning_rate": 1.1700912097824998e-05,
+      "loss": 1.8465,
+      "step": 4141
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 13.506994247436523,
+      "learning_rate": 1.1698907487220607e-05,
+      "loss": 1.2933,
+      "step": 4142
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 12.685111999511719,
+      "learning_rate": 1.1696902876616219e-05,
+      "loss": 1.1366,
+      "step": 4143
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 51.746055603027344,
+      "learning_rate": 1.1694898266011827e-05,
+      "loss": 2.0853,
+      "step": 4144
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 16.30849838256836,
+      "learning_rate": 1.1692893655407437e-05,
+      "loss": 1.8595,
+      "step": 4145
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 24.40019989013672,
+      "learning_rate": 1.1690889044803049e-05,
+      "loss": 2.4656,
+      "step": 4146
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 30.043943405151367,
+      "learning_rate": 1.1688884434198657e-05,
+      "loss": 1.5323,
+      "step": 4147
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 16.770814895629883,
+      "learning_rate": 1.1686879823594269e-05,
+      "loss": 2.9425,
+      "step": 4148
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 11.892877578735352,
+      "learning_rate": 1.1684875212989877e-05,
+      "loss": 2.216,
+      "step": 4149
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 20.787343978881836,
+      "learning_rate": 1.1682870602385488e-05,
+      "loss": 2.4695,
+      "step": 4150
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 23.898902893066406,
+      "learning_rate": 1.16808659917811e-05,
+      "loss": 1.5012,
+      "step": 4151
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 24.14310073852539,
+      "learning_rate": 1.1678861381176708e-05,
+      "loss": 1.8126,
+      "step": 4152
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 16.704151153564453,
+      "learning_rate": 1.1676856770572318e-05,
+      "loss": 2.1569,
+      "step": 4153
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 11.908331871032715,
+      "learning_rate": 1.1674852159967926e-05,
+      "loss": 1.5541,
+      "step": 4154
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 14.777883529663086,
+      "learning_rate": 1.1672847549363538e-05,
+      "loss": 1.7314,
+      "step": 4155
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 13.375772476196289,
+      "learning_rate": 1.1670842938759146e-05,
+      "loss": 1.1609,
+      "step": 4156
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 9.32558536529541,
+      "learning_rate": 1.1668838328154756e-05,
+      "loss": 1.6317,
+      "step": 4157
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 14.819250106811523,
+      "learning_rate": 1.1666833717550368e-05,
+      "loss": 1.5169,
+      "step": 4158
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 7.862631797790527,
+      "learning_rate": 1.1664829106945977e-05,
+      "loss": 1.187,
+      "step": 4159
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 26.258054733276367,
+      "learning_rate": 1.1662824496341587e-05,
+      "loss": 2.8375,
+      "step": 4160
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 16.67561912536621,
+      "learning_rate": 1.1660819885737195e-05,
+      "loss": 1.5601,
+      "step": 4161
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 22.342544555664062,
+      "learning_rate": 1.1658815275132807e-05,
+      "loss": 1.5789,
+      "step": 4162
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 13.550854682922363,
+      "learning_rate": 1.1656810664528417e-05,
+      "loss": 2.4655,
+      "step": 4163
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 33.92632293701172,
+      "learning_rate": 1.1654806053924025e-05,
+      "loss": 1.5478,
+      "step": 4164
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 10.819427490234375,
+      "learning_rate": 1.1652801443319637e-05,
+      "loss": 1.5479,
+      "step": 4165
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 16.504961013793945,
+      "learning_rate": 1.1650796832715245e-05,
+      "loss": 1.4122,
+      "step": 4166
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 22.91649055480957,
+      "learning_rate": 1.1648792222110856e-05,
+      "loss": 1.9459,
+      "step": 4167
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 9.941856384277344,
+      "learning_rate": 1.1646787611506466e-05,
+      "loss": 1.7419,
+      "step": 4168
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 12.914164543151855,
+      "learning_rate": 1.1644783000902076e-05,
+      "loss": 1.2351,
+      "step": 4169
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 10.681536674499512,
+      "learning_rate": 1.1642778390297686e-05,
+      "loss": 1.7278,
+      "step": 4170
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 17.06873321533203,
+      "learning_rate": 1.1640773779693296e-05,
+      "loss": 2.0348,
+      "step": 4171
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 20.173357009887695,
+      "learning_rate": 1.1638769169088906e-05,
+      "loss": 1.6212,
+      "step": 4172
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 6.546261310577393,
+      "learning_rate": 1.1636764558484514e-05,
+      "loss": 0.4214,
+      "step": 4173
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 27.385225296020508,
+      "learning_rate": 1.1634759947880126e-05,
+      "loss": 1.6822,
+      "step": 4174
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 37.10486602783203,
+      "learning_rate": 1.1632755337275736e-05,
+      "loss": 1.5469,
+      "step": 4175
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 13.431403160095215,
+      "learning_rate": 1.1630750726671345e-05,
+      "loss": 2.7734,
+      "step": 4176
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 17.27809715270996,
+      "learning_rate": 1.1628746116066956e-05,
+      "loss": 1.8377,
+      "step": 4177
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 6.829921245574951,
+      "learning_rate": 1.1626741505462565e-05,
+      "loss": 1.0531,
+      "step": 4178
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 15.239294052124023,
+      "learning_rate": 1.1624736894858175e-05,
+      "loss": 1.8001,
+      "step": 4179
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 12.837756156921387,
+      "learning_rate": 1.1622732284253783e-05,
+      "loss": 1.1665,
+      "step": 4180
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 16.143875122070312,
+      "learning_rate": 1.1620727673649395e-05,
+      "loss": 0.9885,
+      "step": 4181
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 12.66282844543457,
+      "learning_rate": 1.1618723063045005e-05,
+      "loss": 1.219,
+      "step": 4182
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 27.858238220214844,
+      "learning_rate": 1.1616718452440614e-05,
+      "loss": 2.2014,
+      "step": 4183
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 21.537065505981445,
+      "learning_rate": 1.1614713841836225e-05,
+      "loss": 2.1228,
+      "step": 4184
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 43.390296936035156,
+      "learning_rate": 1.1612709231231834e-05,
+      "loss": 2.343,
+      "step": 4185
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 24.188064575195312,
+      "learning_rate": 1.1610704620627444e-05,
+      "loss": 1.1875,
+      "step": 4186
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 13.248528480529785,
+      "learning_rate": 1.1608700010023052e-05,
+      "loss": 1.3365,
+      "step": 4187
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 26.479053497314453,
+      "learning_rate": 1.1606695399418664e-05,
+      "loss": 2.2343,
+      "step": 4188
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 20.167842864990234,
+      "learning_rate": 1.1604690788814274e-05,
+      "loss": 1.9832,
+      "step": 4189
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 27.049171447753906,
+      "learning_rate": 1.1602686178209882e-05,
+      "loss": 1.6917,
+      "step": 4190
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 7.952903747558594,
+      "learning_rate": 1.1600681567605494e-05,
+      "loss": 1.1251,
+      "step": 4191
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 9.99161148071289,
+      "learning_rate": 1.1598676957001103e-05,
+      "loss": 1.3309,
+      "step": 4192
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 23.682477951049805,
+      "learning_rate": 1.1596672346396713e-05,
+      "loss": 1.7625,
+      "step": 4193
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 14.787646293640137,
+      "learning_rate": 1.1594667735792324e-05,
+      "loss": 0.8436,
+      "step": 4194
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 16.3641414642334,
+      "learning_rate": 1.1592663125187933e-05,
+      "loss": 1.5916,
+      "step": 4195
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 15.075904846191406,
+      "learning_rate": 1.1590658514583543e-05,
+      "loss": 1.5184,
+      "step": 4196
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 13.050864219665527,
+      "learning_rate": 1.1588653903979153e-05,
+      "loss": 1.7963,
+      "step": 4197
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 10.601302146911621,
+      "learning_rate": 1.1586649293374763e-05,
+      "loss": 1.7489,
+      "step": 4198
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 21.2014217376709,
+      "learning_rate": 1.1584644682770371e-05,
+      "loss": 1.6054,
+      "step": 4199
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 17.15383529663086,
+      "learning_rate": 1.1582640072165983e-05,
+      "loss": 1.2979,
+      "step": 4200
+    },
+    {
+      "epoch": 1.26,
+      "eval_loss": 0.2202432006597519,
+      "eval_runtime": 43.4456,
+      "eval_samples_per_second": 34.043,
+      "eval_steps_per_second": 34.043,
+      "step": 4200
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 14.757326126098633,
+      "learning_rate": 1.1580635461561593e-05,
+      "loss": 1.03,
+      "step": 4201
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 12.233736038208008,
+      "learning_rate": 1.1578630850957202e-05,
+      "loss": 1.4463,
+      "step": 4202
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 17.687862396240234,
+      "learning_rate": 1.1576626240352814e-05,
+      "loss": 1.9516,
+      "step": 4203
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 18.68252944946289,
+      "learning_rate": 1.1574621629748422e-05,
+      "loss": 2.6357,
+      "step": 4204
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 24.503774642944336,
+      "learning_rate": 1.1572617019144032e-05,
+      "loss": 1.4327,
+      "step": 4205
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 14.029136657714844,
+      "learning_rate": 1.1570612408539644e-05,
+      "loss": 2.0086,
+      "step": 4206
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 13.784452438354492,
+      "learning_rate": 1.1568607797935252e-05,
+      "loss": 2.5528,
+      "step": 4207
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 29.253156661987305,
+      "learning_rate": 1.1566603187330862e-05,
+      "loss": 0.5661,
+      "step": 4208
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 26.327823638916016,
+      "learning_rate": 1.156459857672647e-05,
+      "loss": 1.8774,
+      "step": 4209
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 50.058536529541016,
+      "learning_rate": 1.1562593966122082e-05,
+      "loss": 2.2888,
+      "step": 4210
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 8.850481033325195,
+      "learning_rate": 1.156058935551769e-05,
+      "loss": 1.476,
+      "step": 4211
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 12.578634262084961,
+      "learning_rate": 1.1558584744913301e-05,
+      "loss": 1.5757,
+      "step": 4212
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 42.25316619873047,
+      "learning_rate": 1.1556580134308913e-05,
+      "loss": 1.574,
+      "step": 4213
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 15.227287292480469,
+      "learning_rate": 1.1554575523704521e-05,
+      "loss": 2.4708,
+      "step": 4214
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 10.333723068237305,
+      "learning_rate": 1.1552570913100131e-05,
+      "loss": 1.6319,
+      "step": 4215
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 9.808292388916016,
+      "learning_rate": 1.1550566302495741e-05,
+      "loss": 2.1226,
+      "step": 4216
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 39.865562438964844,
+      "learning_rate": 1.1548561691891351e-05,
+      "loss": 2.9798,
+      "step": 4217
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 47.3433723449707,
+      "learning_rate": 1.154655708128696e-05,
+      "loss": 1.7633,
+      "step": 4218
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 18.767004013061523,
+      "learning_rate": 1.1544552470682571e-05,
+      "loss": 1.0271,
+      "step": 4219
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 5.930094242095947,
+      "learning_rate": 1.1542547860078182e-05,
+      "loss": 0.7385,
+      "step": 4220
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 17.10114097595215,
+      "learning_rate": 1.154054324947379e-05,
+      "loss": 1.4075,
+      "step": 4221
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 22.982498168945312,
+      "learning_rate": 1.1538538638869402e-05,
+      "loss": 2.3147,
+      "step": 4222
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 29.882640838623047,
+      "learning_rate": 1.153653402826501e-05,
+      "loss": 1.3968,
+      "step": 4223
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 17.916343688964844,
+      "learning_rate": 1.153452941766062e-05,
+      "loss": 1.5234,
+      "step": 4224
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 12.272050857543945,
+      "learning_rate": 1.1532524807056232e-05,
+      "loss": 0.895,
+      "step": 4225
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 14.076766967773438,
+      "learning_rate": 1.153052019645184e-05,
+      "loss": 1.1994,
+      "step": 4226
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 16.016265869140625,
+      "learning_rate": 1.152851558584745e-05,
+      "loss": 1.0423,
+      "step": 4227
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 10.317485809326172,
+      "learning_rate": 1.1526510975243059e-05,
+      "loss": 1.3147,
+      "step": 4228
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 12.768622398376465,
+      "learning_rate": 1.152450636463867e-05,
+      "loss": 2.2527,
+      "step": 4229
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 15.519412994384766,
+      "learning_rate": 1.1522501754034279e-05,
+      "loss": 1.4902,
+      "step": 4230
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 18.289718627929688,
+      "learning_rate": 1.1520497143429889e-05,
+      "loss": 2.5218,
+      "step": 4231
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 17.28578758239746,
+      "learning_rate": 1.1518492532825501e-05,
+      "loss": 0.8004,
+      "step": 4232
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 23.51228141784668,
+      "learning_rate": 1.151648792222111e-05,
+      "loss": 1.8972,
+      "step": 4233
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 12.242810249328613,
+      "learning_rate": 1.151448331161672e-05,
+      "loss": 1.0982,
+      "step": 4234
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 16.621652603149414,
+      "learning_rate": 1.1512478701012328e-05,
+      "loss": 1.8477,
+      "step": 4235
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 9.91673755645752,
+      "learning_rate": 1.151047409040794e-05,
+      "loss": 1.4897,
+      "step": 4236
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 17.222322463989258,
+      "learning_rate": 1.150846947980355e-05,
+      "loss": 1.2186,
+      "step": 4237
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 11.030045509338379,
+      "learning_rate": 1.1506464869199158e-05,
+      "loss": 1.0489,
+      "step": 4238
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 13.282185554504395,
+      "learning_rate": 1.150446025859477e-05,
+      "loss": 2.1427,
+      "step": 4239
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 18.355653762817383,
+      "learning_rate": 1.1502455647990378e-05,
+      "loss": 1.5696,
+      "step": 4240
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 15.735917091369629,
+      "learning_rate": 1.1500451037385988e-05,
+      "loss": 1.789,
+      "step": 4241
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 28.514938354492188,
+      "learning_rate": 1.1498446426781598e-05,
+      "loss": 2.2112,
+      "step": 4242
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 16.573684692382812,
+      "learning_rate": 1.1496441816177208e-05,
+      "loss": 1.6606,
+      "step": 4243
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 18.012393951416016,
+      "learning_rate": 1.1494437205572819e-05,
+      "loss": 1.3518,
+      "step": 4244
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 21.622119903564453,
+      "learning_rate": 1.1492432594968429e-05,
+      "loss": 1.9794,
+      "step": 4245
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 11.884101867675781,
+      "learning_rate": 1.1490427984364039e-05,
+      "loss": 1.4142,
+      "step": 4246
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 19.552974700927734,
+      "learning_rate": 1.1488423373759647e-05,
+      "loss": 1.0571,
+      "step": 4247
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 17.979076385498047,
+      "learning_rate": 1.1486418763155259e-05,
+      "loss": 1.6985,
+      "step": 4248
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 19.238636016845703,
+      "learning_rate": 1.1484414152550869e-05,
+      "loss": 1.396,
+      "step": 4249
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 46.26264953613281,
+      "learning_rate": 1.1482409541946477e-05,
+      "loss": 2.4083,
+      "step": 4250
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 13.219773292541504,
+      "learning_rate": 1.1480404931342089e-05,
+      "loss": 0.8623,
+      "step": 4251
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 9.744564056396484,
+      "learning_rate": 1.1478400320737697e-05,
+      "loss": 2.33,
+      "step": 4252
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 17.062484741210938,
+      "learning_rate": 1.1476395710133308e-05,
+      "loss": 1.4111,
+      "step": 4253
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 11.308719635009766,
+      "learning_rate": 1.1474391099528916e-05,
+      "loss": 1.0592,
+      "step": 4254
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 12.195525169372559,
+      "learning_rate": 1.1472386488924528e-05,
+      "loss": 1.8724,
+      "step": 4255
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 13.376428604125977,
+      "learning_rate": 1.1470381878320138e-05,
+      "loss": 1.5216,
+      "step": 4256
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 7.64074182510376,
+      "learning_rate": 1.1468377267715746e-05,
+      "loss": 0.7377,
+      "step": 4257
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 19.55640983581543,
+      "learning_rate": 1.1466372657111358e-05,
+      "loss": 2.3189,
+      "step": 4258
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 202.1298370361328,
+      "learning_rate": 1.1464368046506966e-05,
+      "loss": 1.7217,
+      "step": 4259
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 44.232112884521484,
+      "learning_rate": 1.1462363435902576e-05,
+      "loss": 1.1155,
+      "step": 4260
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 20.715179443359375,
+      "learning_rate": 1.1460358825298185e-05,
+      "loss": 1.5698,
+      "step": 4261
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 16.672870635986328,
+      "learning_rate": 1.1458354214693797e-05,
+      "loss": 2.0041,
+      "step": 4262
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 15.357919692993164,
+      "learning_rate": 1.1456349604089407e-05,
+      "loss": 1.9595,
+      "step": 4263
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 25.542970657348633,
+      "learning_rate": 1.1454344993485015e-05,
+      "loss": 3.4864,
+      "step": 4264
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 20.45473861694336,
+      "learning_rate": 1.1452340382880627e-05,
+      "loss": 1.8289,
+      "step": 4265
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 18.233806610107422,
+      "learning_rate": 1.1450335772276235e-05,
+      "loss": 1.9064,
+      "step": 4266
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 8.359683990478516,
+      "learning_rate": 1.1448331161671847e-05,
+      "loss": 1.201,
+      "step": 4267
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 9.61418628692627,
+      "learning_rate": 1.1446326551067457e-05,
+      "loss": 2.0508,
+      "step": 4268
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 12.493593215942383,
+      "learning_rate": 1.1444321940463066e-05,
+      "loss": 1.8421,
+      "step": 4269
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 19.759090423583984,
+      "learning_rate": 1.1442317329858677e-05,
+      "loss": 2.7833,
+      "step": 4270
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 12.588993072509766,
+      "learning_rate": 1.1440312719254286e-05,
+      "loss": 1.1135,
+      "step": 4271
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 22.17607879638672,
+      "learning_rate": 1.1438308108649896e-05,
+      "loss": 2.3915,
+      "step": 4272
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 11.478544235229492,
+      "learning_rate": 1.1436303498045504e-05,
+      "loss": 0.9938,
+      "step": 4273
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 10.766488075256348,
+      "learning_rate": 1.1434298887441116e-05,
+      "loss": 1.0582,
+      "step": 4274
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 20.56522560119629,
+      "learning_rate": 1.1432294276836726e-05,
+      "loss": 1.6918,
+      "step": 4275
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 9.447360038757324,
+      "learning_rate": 1.1430289666232334e-05,
+      "loss": 1.839,
+      "step": 4276
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 9.256118774414062,
+      "learning_rate": 1.1428285055627946e-05,
+      "loss": 1.307,
+      "step": 4277
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 13.01025676727295,
+      "learning_rate": 1.1426280445023555e-05,
+      "loss": 1.6272,
+      "step": 4278
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 11.885642051696777,
+      "learning_rate": 1.1424275834419165e-05,
+      "loss": 0.5883,
+      "step": 4279
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 13.292377471923828,
+      "learning_rate": 1.1422271223814776e-05,
+      "loss": 1.4237,
+      "step": 4280
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 7.443901538848877,
+      "learning_rate": 1.1420266613210385e-05,
+      "loss": 0.9989,
+      "step": 4281
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 83.39949035644531,
+      "learning_rate": 1.1418262002605995e-05,
+      "loss": 1.8078,
+      "step": 4282
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 12.701087951660156,
+      "learning_rate": 1.1416257392001603e-05,
+      "loss": 1.5752,
+      "step": 4283
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 58.16669464111328,
+      "learning_rate": 1.1414252781397215e-05,
+      "loss": 2.8532,
+      "step": 4284
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 21.291078567504883,
+      "learning_rate": 1.1412248170792823e-05,
+      "loss": 1.2318,
+      "step": 4285
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 18.41106605529785,
+      "learning_rate": 1.1410243560188434e-05,
+      "loss": 2.264,
+      "step": 4286
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 20.537443161010742,
+      "learning_rate": 1.1408238949584045e-05,
+      "loss": 2.6269,
+      "step": 4287
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 15.127706527709961,
+      "learning_rate": 1.1406234338979654e-05,
+      "loss": 1.9341,
+      "step": 4288
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 20.711387634277344,
+      "learning_rate": 1.1404229728375264e-05,
+      "loss": 1.7335,
+      "step": 4289
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 50.79159164428711,
+      "learning_rate": 1.1402225117770874e-05,
+      "loss": 2.1342,
+      "step": 4290
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 29.805246353149414,
+      "learning_rate": 1.1400220507166484e-05,
+      "loss": 2.0907,
+      "step": 4291
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 12.824877738952637,
+      "learning_rate": 1.1398215896562094e-05,
+      "loss": 1.1383,
+      "step": 4292
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 7.294422149658203,
+      "learning_rate": 1.1396211285957704e-05,
+      "loss": 1.3885,
+      "step": 4293
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 31.934511184692383,
+      "learning_rate": 1.1394206675353314e-05,
+      "loss": 2.1207,
+      "step": 4294
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 18.717044830322266,
+      "learning_rate": 1.1392202064748923e-05,
+      "loss": 1.4102,
+      "step": 4295
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 14.38543701171875,
+      "learning_rate": 1.1390197454144534e-05,
+      "loss": 1.9233,
+      "step": 4296
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 20.979101181030273,
+      "learning_rate": 1.1388192843540143e-05,
+      "loss": 1.509,
+      "step": 4297
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 16.605932235717773,
+      "learning_rate": 1.1386188232935753e-05,
+      "loss": 1.4354,
+      "step": 4298
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 9.505638122558594,
+      "learning_rate": 1.1384183622331365e-05,
+      "loss": 1.1145,
+      "step": 4299
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 15.347785949707031,
+      "learning_rate": 1.1382179011726973e-05,
+      "loss": 1.6882,
+      "step": 4300
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 20.905508041381836,
+      "learning_rate": 1.1380174401122583e-05,
+      "loss": 1.5622,
+      "step": 4301
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 41.201194763183594,
+      "learning_rate": 1.1378169790518192e-05,
+      "loss": 3.4766,
+      "step": 4302
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 15.540892601013184,
+      "learning_rate": 1.1376165179913803e-05,
+      "loss": 1.4535,
+      "step": 4303
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 24.336341857910156,
+      "learning_rate": 1.1374160569309412e-05,
+      "loss": 1.7003,
+      "step": 4304
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 28.202116012573242,
+      "learning_rate": 1.1372155958705022e-05,
+      "loss": 1.2576,
+      "step": 4305
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 12.418293952941895,
+      "learning_rate": 1.1370151348100634e-05,
+      "loss": 1.8355,
+      "step": 4306
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 35.69652557373047,
+      "learning_rate": 1.1368146737496242e-05,
+      "loss": 2.1658,
+      "step": 4307
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 17.411029815673828,
+      "learning_rate": 1.1366142126891852e-05,
+      "loss": 2.4983,
+      "step": 4308
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 10.031928062438965,
+      "learning_rate": 1.136413751628746e-05,
+      "loss": 1.7212,
+      "step": 4309
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 11.622937202453613,
+      "learning_rate": 1.1362132905683072e-05,
+      "loss": 1.4201,
+      "step": 4310
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 19.42033576965332,
+      "learning_rate": 1.1360128295078682e-05,
+      "loss": 1.8876,
+      "step": 4311
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 24.995628356933594,
+      "learning_rate": 1.135812368447429e-05,
+      "loss": 1.682,
+      "step": 4312
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 11.356565475463867,
+      "learning_rate": 1.1356119073869902e-05,
+      "loss": 1.3858,
+      "step": 4313
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 18.608964920043945,
+      "learning_rate": 1.1354114463265511e-05,
+      "loss": 1.2763,
+      "step": 4314
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 16.34697723388672,
+      "learning_rate": 1.1352109852661121e-05,
+      "loss": 1.2931,
+      "step": 4315
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 17.903200149536133,
+      "learning_rate": 1.1350105242056731e-05,
+      "loss": 1.8466,
+      "step": 4316
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 7.10983419418335,
+      "learning_rate": 1.1348100631452341e-05,
+      "loss": 0.967,
+      "step": 4317
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 13.738561630249023,
+      "learning_rate": 1.1346096020847951e-05,
+      "loss": 1.4251,
+      "step": 4318
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 14.061918258666992,
+      "learning_rate": 1.1344091410243561e-05,
+      "loss": 1.8931,
+      "step": 4319
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 28.247364044189453,
+      "learning_rate": 1.1342086799639171e-05,
+      "loss": 2.5285,
+      "step": 4320
+    },
+    {
+      "epoch": 1.3,
+      "eval_loss": 0.21285276114940643,
+      "eval_runtime": 43.3572,
+      "eval_samples_per_second": 34.112,
+      "eval_steps_per_second": 34.112,
+      "step": 4320
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 11.178879737854004,
+      "learning_rate": 1.134008218903478e-05,
+      "loss": 1.5172,
+      "step": 4321
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 16.7074031829834,
+      "learning_rate": 1.1338077578430392e-05,
+      "loss": 1.3458,
+      "step": 4322
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 21.143476486206055,
+      "learning_rate": 1.1336072967826002e-05,
+      "loss": 2.4362,
+      "step": 4323
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 22.83983039855957,
+      "learning_rate": 1.133406835722161e-05,
+      "loss": 2.1729,
+      "step": 4324
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 16.511369705200195,
+      "learning_rate": 1.1332063746617222e-05,
+      "loss": 1.3165,
+      "step": 4325
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 16.411027908325195,
+      "learning_rate": 1.133005913601283e-05,
+      "loss": 2.1342,
+      "step": 4326
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 16.996347427368164,
+      "learning_rate": 1.132805452540844e-05,
+      "loss": 1.9439,
+      "step": 4327
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 27.14801788330078,
+      "learning_rate": 1.1326049914804049e-05,
+      "loss": 2.3698,
+      "step": 4328
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 10.485852241516113,
+      "learning_rate": 1.132404530419966e-05,
+      "loss": 1.1316,
+      "step": 4329
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 48.32770538330078,
+      "learning_rate": 1.132204069359527e-05,
+      "loss": 1.8271,
+      "step": 4330
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 68.1049575805664,
+      "learning_rate": 1.1320036082990879e-05,
+      "loss": 2.5344,
+      "step": 4331
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 22.41739273071289,
+      "learning_rate": 1.131803147238649e-05,
+      "loss": 2.1376,
+      "step": 4332
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 61.55850601196289,
+      "learning_rate": 1.1316026861782099e-05,
+      "loss": 3.21,
+      "step": 4333
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 13.802007675170898,
+      "learning_rate": 1.1314022251177709e-05,
+      "loss": 1.7577,
+      "step": 4334
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 16.4052734375,
+      "learning_rate": 1.1312017640573321e-05,
+      "loss": 1.2041,
+      "step": 4335
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 26.579936981201172,
+      "learning_rate": 1.131001302996893e-05,
+      "loss": 1.6785,
+      "step": 4336
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 24.084991455078125,
+      "learning_rate": 1.130800841936454e-05,
+      "loss": 1.5441,
+      "step": 4337
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 21.47726058959961,
+      "learning_rate": 1.130600380876015e-05,
+      "loss": 1.9604,
+      "step": 4338
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 38.121421813964844,
+      "learning_rate": 1.130399919815576e-05,
+      "loss": 2.2498,
+      "step": 4339
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 23.063236236572266,
+      "learning_rate": 1.1301994587551368e-05,
+      "loss": 1.8673,
+      "step": 4340
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 11.506025314331055,
+      "learning_rate": 1.129998997694698e-05,
+      "loss": 1.3873,
+      "step": 4341
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 11.516486167907715,
+      "learning_rate": 1.129798536634259e-05,
+      "loss": 1.2646,
+      "step": 4342
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 15.611605644226074,
+      "learning_rate": 1.1295980755738198e-05,
+      "loss": 1.2307,
+      "step": 4343
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 17.53089141845703,
+      "learning_rate": 1.129397614513381e-05,
+      "loss": 1.2414,
+      "step": 4344
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 14.397602081298828,
+      "learning_rate": 1.1291971534529418e-05,
+      "loss": 1.3744,
+      "step": 4345
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 23.23992156982422,
+      "learning_rate": 1.1289966923925028e-05,
+      "loss": 1.8092,
+      "step": 4346
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 15.50413990020752,
+      "learning_rate": 1.1287962313320637e-05,
+      "loss": 1.819,
+      "step": 4347
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 12.4719877243042,
+      "learning_rate": 1.1285957702716249e-05,
+      "loss": 0.959,
+      "step": 4348
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 8.75976276397705,
+      "learning_rate": 1.1283953092111859e-05,
+      "loss": 1.0429,
+      "step": 4349
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 34.667945861816406,
+      "learning_rate": 1.1281948481507467e-05,
+      "loss": 2.2953,
+      "step": 4350
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 15.772024154663086,
+      "learning_rate": 1.1279943870903079e-05,
+      "loss": 1.67,
+      "step": 4351
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 19.514604568481445,
+      "learning_rate": 1.1277939260298687e-05,
+      "loss": 2.0143,
+      "step": 4352
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 22.16571807861328,
+      "learning_rate": 1.1275934649694297e-05,
+      "loss": 1.2593,
+      "step": 4353
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 17.157798767089844,
+      "learning_rate": 1.1273930039089909e-05,
+      "loss": 1.6187,
+      "step": 4354
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 33.16340637207031,
+      "learning_rate": 1.1271925428485518e-05,
+      "loss": 2.0045,
+      "step": 4355
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 14.194746971130371,
+      "learning_rate": 1.1269920817881128e-05,
+      "loss": 1.671,
+      "step": 4356
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 15.4343900680542,
+      "learning_rate": 1.1267916207276736e-05,
+      "loss": 1.2274,
+      "step": 4357
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 21.623807907104492,
+      "learning_rate": 1.1265911596672348e-05,
+      "loss": 2.5802,
+      "step": 4358
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 12.904182434082031,
+      "learning_rate": 1.1263906986067956e-05,
+      "loss": 1.4189,
+      "step": 4359
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 16.601791381835938,
+      "learning_rate": 1.1261902375463566e-05,
+      "loss": 2.787,
+      "step": 4360
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 12.915867805480957,
+      "learning_rate": 1.1259897764859178e-05,
+      "loss": 1.5663,
+      "step": 4361
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 17.556509017944336,
+      "learning_rate": 1.1257893154254786e-05,
+      "loss": 1.8852,
+      "step": 4362
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 23.893291473388672,
+      "learning_rate": 1.1255888543650397e-05,
+      "loss": 1.4539,
+      "step": 4363
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 11.98607063293457,
+      "learning_rate": 1.1253883933046007e-05,
+      "loss": 1.7452,
+      "step": 4364
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 13.887165069580078,
+      "learning_rate": 1.1251879322441617e-05,
+      "loss": 1.6717,
+      "step": 4365
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 7.588750839233398,
+      "learning_rate": 1.1249874711837227e-05,
+      "loss": 0.8518,
+      "step": 4366
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 27.59679412841797,
+      "learning_rate": 1.1247870101232837e-05,
+      "loss": 2.483,
+      "step": 4367
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 19.16243553161621,
+      "learning_rate": 1.1245865490628447e-05,
+      "loss": 1.4322,
+      "step": 4368
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 16.66512107849121,
+      "learning_rate": 1.1243860880024055e-05,
+      "loss": 1.9586,
+      "step": 4369
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 17.893173217773438,
+      "learning_rate": 1.1241856269419667e-05,
+      "loss": 2.8739,
+      "step": 4370
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 16.915111541748047,
+      "learning_rate": 1.1239851658815275e-05,
+      "loss": 1.3351,
+      "step": 4371
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 20.749107360839844,
+      "learning_rate": 1.1237847048210886e-05,
+      "loss": 2.3157,
+      "step": 4372
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 33.895973205566406,
+      "learning_rate": 1.1235842437606497e-05,
+      "loss": 3.2345,
+      "step": 4373
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 16.17300033569336,
+      "learning_rate": 1.1233837827002106e-05,
+      "loss": 1.0639,
+      "step": 4374
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 11.130797386169434,
+      "learning_rate": 1.1231833216397716e-05,
+      "loss": 2.0059,
+      "step": 4375
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 12.532829284667969,
+      "learning_rate": 1.1229828605793324e-05,
+      "loss": 1.1101,
+      "step": 4376
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 10.261481285095215,
+      "learning_rate": 1.1227823995188936e-05,
+      "loss": 1.7796,
+      "step": 4377
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 18.463781356811523,
+      "learning_rate": 1.1225819384584546e-05,
+      "loss": 1.9479,
+      "step": 4378
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 12.4391450881958,
+      "learning_rate": 1.1223814773980154e-05,
+      "loss": 2.0404,
+      "step": 4379
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 15.823043823242188,
+      "learning_rate": 1.1221810163375766e-05,
+      "loss": 1.4798,
+      "step": 4380
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 19.075504302978516,
+      "learning_rate": 1.1219805552771375e-05,
+      "loss": 1.712,
+      "step": 4381
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 13.849617958068848,
+      "learning_rate": 1.1217800942166985e-05,
+      "loss": 1.8711,
+      "step": 4382
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 17.942249298095703,
+      "learning_rate": 1.1215796331562593e-05,
+      "loss": 2.6901,
+      "step": 4383
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 9.185188293457031,
+      "learning_rate": 1.1213791720958205e-05,
+      "loss": 1.0339,
+      "step": 4384
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 12.996100425720215,
+      "learning_rate": 1.1211787110353815e-05,
+      "loss": 1.7259,
+      "step": 4385
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 31.83272933959961,
+      "learning_rate": 1.1209782499749425e-05,
+      "loss": 2.1627,
+      "step": 4386
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 11.160262107849121,
+      "learning_rate": 1.1207777889145035e-05,
+      "loss": 1.7325,
+      "step": 4387
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 19.457317352294922,
+      "learning_rate": 1.1205773278540644e-05,
+      "loss": 1.6398,
+      "step": 4388
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 15.514799118041992,
+      "learning_rate": 1.1203768667936255e-05,
+      "loss": 1.5179,
+      "step": 4389
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 55.15019607543945,
+      "learning_rate": 1.1201764057331864e-05,
+      "loss": 4.2946,
+      "step": 4390
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 17.135194778442383,
+      "learning_rate": 1.1199759446727474e-05,
+      "loss": 2.0419,
+      "step": 4391
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 11.672941207885742,
+      "learning_rate": 1.1197754836123086e-05,
+      "loss": 1.4943,
+      "step": 4392
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 27.473955154418945,
+      "learning_rate": 1.1195750225518694e-05,
+      "loss": 2.4105,
+      "step": 4393
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 12.319040298461914,
+      "learning_rate": 1.1193745614914304e-05,
+      "loss": 1.1325,
+      "step": 4394
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 19.93411636352539,
+      "learning_rate": 1.1191741004309912e-05,
+      "loss": 2.4709,
+      "step": 4395
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 25.42889976501465,
+      "learning_rate": 1.1189736393705524e-05,
+      "loss": 2.7937,
+      "step": 4396
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 31.769067764282227,
+      "learning_rate": 1.1187731783101134e-05,
+      "loss": 2.7676,
+      "step": 4397
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 21.329421997070312,
+      "learning_rate": 1.1185727172496743e-05,
+      "loss": 2.0521,
+      "step": 4398
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 17.158628463745117,
+      "learning_rate": 1.1183722561892354e-05,
+      "loss": 1.5289,
+      "step": 4399
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 13.490479469299316,
+      "learning_rate": 1.1181717951287963e-05,
+      "loss": 1.4183,
+      "step": 4400
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 8.754962921142578,
+      "learning_rate": 1.1179713340683573e-05,
+      "loss": 1.2418,
+      "step": 4401
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 11.73357105255127,
+      "learning_rate": 1.1177708730079181e-05,
+      "loss": 1.3139,
+      "step": 4402
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 13.230740547180176,
+      "learning_rate": 1.1175704119474793e-05,
+      "loss": 1.6772,
+      "step": 4403
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 17.465307235717773,
+      "learning_rate": 1.1173699508870403e-05,
+      "loss": 1.6562,
+      "step": 4404
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 32.38056945800781,
+      "learning_rate": 1.1171694898266012e-05,
+      "loss": 2.3572,
+      "step": 4405
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 14.467717170715332,
+      "learning_rate": 1.1169690287661623e-05,
+      "loss": 1.2874,
+      "step": 4406
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 17.778472900390625,
+      "learning_rate": 1.1167685677057232e-05,
+      "loss": 1.9954,
+      "step": 4407
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 8.146272659301758,
+      "learning_rate": 1.1165681066452842e-05,
+      "loss": 0.6304,
+      "step": 4408
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 18.66156005859375,
+      "learning_rate": 1.1163676455848454e-05,
+      "loss": 2.0366,
+      "step": 4409
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 13.149731636047363,
+      "learning_rate": 1.1161671845244062e-05,
+      "loss": 1.5874,
+      "step": 4410
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 15.526867866516113,
+      "learning_rate": 1.1159667234639672e-05,
+      "loss": 1.4642,
+      "step": 4411
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 27.83363151550293,
+      "learning_rate": 1.1157662624035282e-05,
+      "loss": 2.0596,
+      "step": 4412
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 82.84793090820312,
+      "learning_rate": 1.1155658013430892e-05,
+      "loss": 1.8469,
+      "step": 4413
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 16.59908676147461,
+      "learning_rate": 1.11536534028265e-05,
+      "loss": 1.9362,
+      "step": 4414
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 8.759167671203613,
+      "learning_rate": 1.1151648792222112e-05,
+      "loss": 2.0726,
+      "step": 4415
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 32.26197814941406,
+      "learning_rate": 1.1149644181617723e-05,
+      "loss": 2.0811,
+      "step": 4416
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 8.401446342468262,
+      "learning_rate": 1.1147639571013331e-05,
+      "loss": 1.3336,
+      "step": 4417
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 10.307430267333984,
+      "learning_rate": 1.1145634960408943e-05,
+      "loss": 1.1188,
+      "step": 4418
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 16.372766494750977,
+      "learning_rate": 1.1143630349804551e-05,
+      "loss": 1.8886,
+      "step": 4419
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 39.85802459716797,
+      "learning_rate": 1.1141625739200161e-05,
+      "loss": 2.3996,
+      "step": 4420
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 14.044843673706055,
+      "learning_rate": 1.1139621128595773e-05,
+      "loss": 1.4335,
+      "step": 4421
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 23.040878295898438,
+      "learning_rate": 1.1137616517991381e-05,
+      "loss": 1.8451,
+      "step": 4422
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 17.075387954711914,
+      "learning_rate": 1.1135611907386991e-05,
+      "loss": 2.2248,
+      "step": 4423
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 110.57589721679688,
+      "learning_rate": 1.11336072967826e-05,
+      "loss": 1.5452,
+      "step": 4424
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 15.424068450927734,
+      "learning_rate": 1.1131602686178212e-05,
+      "loss": 1.5217,
+      "step": 4425
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 23.823379516601562,
+      "learning_rate": 1.112959807557382e-05,
+      "loss": 1.2601,
+      "step": 4426
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 10.928523063659668,
+      "learning_rate": 1.112759346496943e-05,
+      "loss": 1.4933,
+      "step": 4427
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 13.319860458374023,
+      "learning_rate": 1.1125588854365042e-05,
+      "loss": 1.4411,
+      "step": 4428
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 11.283712387084961,
+      "learning_rate": 1.112358424376065e-05,
+      "loss": 2.1932,
+      "step": 4429
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 7.913817882537842,
+      "learning_rate": 1.112157963315626e-05,
+      "loss": 1.3824,
+      "step": 4430
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 10.032218933105469,
+      "learning_rate": 1.1119575022551869e-05,
+      "loss": 1.2246,
+      "step": 4431
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 25.357025146484375,
+      "learning_rate": 1.111757041194748e-05,
+      "loss": 1.7075,
+      "step": 4432
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 16.777484893798828,
+      "learning_rate": 1.1115565801343089e-05,
+      "loss": 1.6018,
+      "step": 4433
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 27.090877532958984,
+      "learning_rate": 1.1113561190738699e-05,
+      "loss": 2.4302,
+      "step": 4434
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 24.513057708740234,
+      "learning_rate": 1.111155658013431e-05,
+      "loss": 2.2252,
+      "step": 4435
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 20.253862380981445,
+      "learning_rate": 1.1109551969529919e-05,
+      "loss": 1.4782,
+      "step": 4436
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 10.198271751403809,
+      "learning_rate": 1.1107547358925531e-05,
+      "loss": 2.0537,
+      "step": 4437
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 16.26251792907715,
+      "learning_rate": 1.110554274832114e-05,
+      "loss": 1.6344,
+      "step": 4438
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 17.53883934020996,
+      "learning_rate": 1.110353813771675e-05,
+      "loss": 1.4399,
+      "step": 4439
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 9.351251602172852,
+      "learning_rate": 1.1101533527112361e-05,
+      "loss": 1.2029,
+      "step": 4440
+    },
+    {
+      "epoch": 1.33,
+      "eval_loss": 0.22466596961021423,
+      "eval_runtime": 43.3151,
+      "eval_samples_per_second": 34.145,
+      "eval_steps_per_second": 34.145,
+      "step": 4440
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 18.78108787536621,
+      "learning_rate": 1.109952891650797e-05,
+      "loss": 1.8896,
+      "step": 4441
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 14.223852157592773,
+      "learning_rate": 1.109752430590358e-05,
+      "loss": 1.8473,
+      "step": 4442
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 29.34437370300293,
+      "learning_rate": 1.1095519695299188e-05,
+      "loss": 1.8663,
+      "step": 4443
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 11.203985214233398,
+      "learning_rate": 1.10935150846948e-05,
+      "loss": 1.4461,
+      "step": 4444
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 13.125014305114746,
+      "learning_rate": 1.1091510474090408e-05,
+      "loss": 1.3763,
+      "step": 4445
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 17.99578094482422,
+      "learning_rate": 1.1089505863486018e-05,
+      "loss": 2.1525,
+      "step": 4446
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 36.31547927856445,
+      "learning_rate": 1.108750125288163e-05,
+      "loss": 2.2111,
+      "step": 4447
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 13.750075340270996,
+      "learning_rate": 1.1085496642277238e-05,
+      "loss": 1.3206,
+      "step": 4448
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 15.520805358886719,
+      "learning_rate": 1.1083492031672849e-05,
+      "loss": 3.3127,
+      "step": 4449
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 16.05813217163086,
+      "learning_rate": 1.1081487421068457e-05,
+      "loss": 1.5862,
+      "step": 4450
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 18.96892738342285,
+      "learning_rate": 1.1079482810464069e-05,
+      "loss": 1.4905,
+      "step": 4451
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 11.585920333862305,
+      "learning_rate": 1.1077478199859679e-05,
+      "loss": 2.2658,
+      "step": 4452
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 21.121341705322266,
+      "learning_rate": 1.1075473589255287e-05,
+      "loss": 0.8119,
+      "step": 4453
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 39.30922317504883,
+      "learning_rate": 1.1073468978650899e-05,
+      "loss": 2.4467,
+      "step": 4454
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 15.98094654083252,
+      "learning_rate": 1.1071464368046507e-05,
+      "loss": 2.3305,
+      "step": 4455
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 9.627312660217285,
+      "learning_rate": 1.1069459757442117e-05,
+      "loss": 1.2344,
+      "step": 4456
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 14.245975494384766,
+      "learning_rate": 1.1067455146837727e-05,
+      "loss": 0.9343,
+      "step": 4457
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 29.139923095703125,
+      "learning_rate": 1.1065450536233338e-05,
+      "loss": 2.0948,
+      "step": 4458
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 23.93746566772461,
+      "learning_rate": 1.1063445925628948e-05,
+      "loss": 1.8753,
+      "step": 4459
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 8.223407745361328,
+      "learning_rate": 1.1061441315024558e-05,
+      "loss": 1.0049,
+      "step": 4460
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 28.680850982666016,
+      "learning_rate": 1.1059436704420168e-05,
+      "loss": 1.8643,
+      "step": 4461
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 27.831514358520508,
+      "learning_rate": 1.1057432093815776e-05,
+      "loss": 1.8748,
+      "step": 4462
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 17.656972885131836,
+      "learning_rate": 1.1055427483211388e-05,
+      "loss": 1.8081,
+      "step": 4463
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 12.41053581237793,
+      "learning_rate": 1.1053422872606998e-05,
+      "loss": 1.6312,
+      "step": 4464
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 14.747664451599121,
+      "learning_rate": 1.1051418262002606e-05,
+      "loss": 1.8963,
+      "step": 4465
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 20.91558074951172,
+      "learning_rate": 1.1049413651398218e-05,
+      "loss": 1.0479,
+      "step": 4466
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 18.06478500366211,
+      "learning_rate": 1.1047409040793827e-05,
+      "loss": 1.513,
+      "step": 4467
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 25.323034286499023,
+      "learning_rate": 1.1045404430189437e-05,
+      "loss": 2.1898,
+      "step": 4468
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 10.666196823120117,
+      "learning_rate": 1.1043399819585045e-05,
+      "loss": 1.6848,
+      "step": 4469
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 13.26004695892334,
+      "learning_rate": 1.1041395208980657e-05,
+      "loss": 1.3236,
+      "step": 4470
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 11.206985473632812,
+      "learning_rate": 1.1039390598376267e-05,
+      "loss": 1.2008,
+      "step": 4471
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 22.27272605895996,
+      "learning_rate": 1.1037385987771875e-05,
+      "loss": 1.5483,
+      "step": 4472
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 13.044824600219727,
+      "learning_rate": 1.1035381377167487e-05,
+      "loss": 1.4811,
+      "step": 4473
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 15.387162208557129,
+      "learning_rate": 1.1033376766563096e-05,
+      "loss": 1.7151,
+      "step": 4474
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 20.35818862915039,
+      "learning_rate": 1.1031372155958706e-05,
+      "loss": 1.6009,
+      "step": 4475
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 12.326359748840332,
+      "learning_rate": 1.1029367545354314e-05,
+      "loss": 1.7638,
+      "step": 4476
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 29.4976863861084,
+      "learning_rate": 1.1027362934749926e-05,
+      "loss": 1.8569,
+      "step": 4477
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 11.03281307220459,
+      "learning_rate": 1.1025358324145536e-05,
+      "loss": 1.0993,
+      "step": 4478
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 20.460525512695312,
+      "learning_rate": 1.1023353713541144e-05,
+      "loss": 1.4282,
+      "step": 4479
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 11.63111400604248,
+      "learning_rate": 1.1021349102936756e-05,
+      "loss": 2.2856,
+      "step": 4480
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 17.048723220825195,
+      "learning_rate": 1.1019344492332364e-05,
+      "loss": 2.037,
+      "step": 4481
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 7.7821502685546875,
+      "learning_rate": 1.1017339881727975e-05,
+      "loss": 1.2083,
+      "step": 4482
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 16.613143920898438,
+      "learning_rate": 1.1015335271123586e-05,
+      "loss": 1.3636,
+      "step": 4483
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 15.978687286376953,
+      "learning_rate": 1.1013330660519195e-05,
+      "loss": 1.5447,
+      "step": 4484
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 12.686722755432129,
+      "learning_rate": 1.1011326049914805e-05,
+      "loss": 1.308,
+      "step": 4485
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 15.303074836730957,
+      "learning_rate": 1.1009321439310415e-05,
+      "loss": 2.4245,
+      "step": 4486
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 12.841423988342285,
+      "learning_rate": 1.1007316828706025e-05,
+      "loss": 2.0611,
+      "step": 4487
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 14.324323654174805,
+      "learning_rate": 1.1005312218101633e-05,
+      "loss": 1.9151,
+      "step": 4488
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 10.316316604614258,
+      "learning_rate": 1.1003307607497245e-05,
+      "loss": 1.8989,
+      "step": 4489
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 11.872175216674805,
+      "learning_rate": 1.1001302996892855e-05,
+      "loss": 1.4971,
+      "step": 4490
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 13.584701538085938,
+      "learning_rate": 1.0999298386288464e-05,
+      "loss": 1.6093,
+      "step": 4491
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 23.555721282958984,
+      "learning_rate": 1.0997293775684075e-05,
+      "loss": 1.4708,
+      "step": 4492
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 13.753824234008789,
+      "learning_rate": 1.0995289165079684e-05,
+      "loss": 1.4468,
+      "step": 4493
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 23.030887603759766,
+      "learning_rate": 1.0993284554475294e-05,
+      "loss": 1.9327,
+      "step": 4494
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 43.81611251831055,
+      "learning_rate": 1.0991279943870906e-05,
+      "loss": 2.3426,
+      "step": 4495
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 22.91530418395996,
+      "learning_rate": 1.0989275333266514e-05,
+      "loss": 2.1606,
+      "step": 4496
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 30.3094425201416,
+      "learning_rate": 1.0987270722662124e-05,
+      "loss": 2.3948,
+      "step": 4497
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 90.9547348022461,
+      "learning_rate": 1.0985266112057732e-05,
+      "loss": 2.5067,
+      "step": 4498
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 11.258960723876953,
+      "learning_rate": 1.0983261501453344e-05,
+      "loss": 2.0156,
+      "step": 4499
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 66.39472961425781,
+      "learning_rate": 1.0981256890848953e-05,
+      "loss": 2.6879,
+      "step": 4500
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 20.91155242919922,
+      "learning_rate": 1.0979252280244563e-05,
+      "loss": 1.7934,
+      "step": 4501
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 18.402379989624023,
+      "learning_rate": 1.0977247669640175e-05,
+      "loss": 1.3445,
+      "step": 4502
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 15.267261505126953,
+      "learning_rate": 1.0975243059035783e-05,
+      "loss": 1.642,
+      "step": 4503
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 50.35695266723633,
+      "learning_rate": 1.0973238448431393e-05,
+      "loss": 2.0173,
+      "step": 4504
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 10.630228042602539,
+      "learning_rate": 1.0971233837827003e-05,
+      "loss": 1.9372,
+      "step": 4505
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 32.916053771972656,
+      "learning_rate": 1.0969229227222613e-05,
+      "loss": 2.2776,
+      "step": 4506
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 22.993738174438477,
+      "learning_rate": 1.0967224616618223e-05,
+      "loss": 2.2662,
+      "step": 4507
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 13.480685234069824,
+      "learning_rate": 1.0965220006013833e-05,
+      "loss": 1.4548,
+      "step": 4508
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 22.65932273864746,
+      "learning_rate": 1.0963215395409443e-05,
+      "loss": 2.1417,
+      "step": 4509
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 9.409554481506348,
+      "learning_rate": 1.0961210784805052e-05,
+      "loss": 1.1043,
+      "step": 4510
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 51.40016555786133,
+      "learning_rate": 1.0959206174200664e-05,
+      "loss": 2.5073,
+      "step": 4511
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 32.164634704589844,
+      "learning_rate": 1.0957201563596272e-05,
+      "loss": 1.7867,
+      "step": 4512
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 29.420494079589844,
+      "learning_rate": 1.0955196952991882e-05,
+      "loss": 1.7466,
+      "step": 4513
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 10.119473457336426,
+      "learning_rate": 1.0953192342387494e-05,
+      "loss": 1.3604,
+      "step": 4514
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 10.170442581176758,
+      "learning_rate": 1.0951187731783102e-05,
+      "loss": 1.2242,
+      "step": 4515
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 11.852310180664062,
+      "learning_rate": 1.0949183121178712e-05,
+      "loss": 1.2755,
+      "step": 4516
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 14.793402671813965,
+      "learning_rate": 1.094717851057432e-05,
+      "loss": 1.752,
+      "step": 4517
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 42.632144927978516,
+      "learning_rate": 1.0945173899969932e-05,
+      "loss": 1.7613,
+      "step": 4518
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 19.947559356689453,
+      "learning_rate": 1.0943169289365541e-05,
+      "loss": 1.9094,
+      "step": 4519
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 15.529866218566895,
+      "learning_rate": 1.0941164678761151e-05,
+      "loss": 1.3777,
+      "step": 4520
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 21.871538162231445,
+      "learning_rate": 1.0939160068156763e-05,
+      "loss": 1.91,
+      "step": 4521
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 18.05958366394043,
+      "learning_rate": 1.0937155457552371e-05,
+      "loss": 1.6152,
+      "step": 4522
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 13.304652214050293,
+      "learning_rate": 1.0935150846947981e-05,
+      "loss": 1.4861,
+      "step": 4523
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 72.59398651123047,
+      "learning_rate": 1.093314623634359e-05,
+      "loss": 2.0853,
+      "step": 4524
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 11.113508224487305,
+      "learning_rate": 1.0931141625739201e-05,
+      "loss": 1.5427,
+      "step": 4525
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 11.754325866699219,
+      "learning_rate": 1.0929137015134811e-05,
+      "loss": 1.8286,
+      "step": 4526
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 16.077978134155273,
+      "learning_rate": 1.092713240453042e-05,
+      "loss": 1.3027,
+      "step": 4527
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 14.799029350280762,
+      "learning_rate": 1.0925127793926032e-05,
+      "loss": 1.222,
+      "step": 4528
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 12.316436767578125,
+      "learning_rate": 1.092312318332164e-05,
+      "loss": 1.5382,
+      "step": 4529
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 11.659561157226562,
+      "learning_rate": 1.092111857271725e-05,
+      "loss": 1.0118,
+      "step": 4530
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 10.741880416870117,
+      "learning_rate": 1.091911396211286e-05,
+      "loss": 1.1201,
+      "step": 4531
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 17.27454376220703,
+      "learning_rate": 1.091710935150847e-05,
+      "loss": 1.8521,
+      "step": 4532
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 19.821794509887695,
+      "learning_rate": 1.091510474090408e-05,
+      "loss": 1.1864,
+      "step": 4533
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 42.41746520996094,
+      "learning_rate": 1.091310013029969e-05,
+      "loss": 1.7798,
+      "step": 4534
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 17.059722900390625,
+      "learning_rate": 1.09110955196953e-05,
+      "loss": 2.3114,
+      "step": 4535
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 38.7147331237793,
+      "learning_rate": 1.0909090909090909e-05,
+      "loss": 2.4257,
+      "step": 4536
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 17.778711318969727,
+      "learning_rate": 1.090708629848652e-05,
+      "loss": 1.598,
+      "step": 4537
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 55.522010803222656,
+      "learning_rate": 1.090508168788213e-05,
+      "loss": 1.7511,
+      "step": 4538
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 31.59695053100586,
+      "learning_rate": 1.090307707727774e-05,
+      "loss": 1.6838,
+      "step": 4539
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 14.828858375549316,
+      "learning_rate": 1.0901072466673351e-05,
+      "loss": 1.3515,
+      "step": 4540
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 18.156906127929688,
+      "learning_rate": 1.089906785606896e-05,
+      "loss": 2.1229,
+      "step": 4541
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 14.6245698928833,
+      "learning_rate": 1.089706324546457e-05,
+      "loss": 1.0952,
+      "step": 4542
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 14.46635913848877,
+      "learning_rate": 1.0895058634860178e-05,
+      "loss": 1.6335,
+      "step": 4543
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 11.85185432434082,
+      "learning_rate": 1.089305402425579e-05,
+      "loss": 1.6408,
+      "step": 4544
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 9.316162109375,
+      "learning_rate": 1.08910494136514e-05,
+      "loss": 1.0231,
+      "step": 4545
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 16.85337257385254,
+      "learning_rate": 1.0889044803047008e-05,
+      "loss": 1.486,
+      "step": 4546
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 13.948755264282227,
+      "learning_rate": 1.088704019244262e-05,
+      "loss": 2.3643,
+      "step": 4547
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 22.807424545288086,
+      "learning_rate": 1.0885035581838228e-05,
+      "loss": 1.6381,
+      "step": 4548
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 17.668928146362305,
+      "learning_rate": 1.0883030971233838e-05,
+      "loss": 1.8671,
+      "step": 4549
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 61.22587966918945,
+      "learning_rate": 1.088102636062945e-05,
+      "loss": 2.4726,
+      "step": 4550
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 15.125883102416992,
+      "learning_rate": 1.0879021750025058e-05,
+      "loss": 1.6064,
+      "step": 4551
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 12.281584739685059,
+      "learning_rate": 1.0877017139420669e-05,
+      "loss": 1.9036,
+      "step": 4552
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 20.28942108154297,
+      "learning_rate": 1.0875012528816277e-05,
+      "loss": 2.022,
+      "step": 4553
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 12.765894889831543,
+      "learning_rate": 1.0873007918211889e-05,
+      "loss": 1.2068,
+      "step": 4554
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 14.457947731018066,
+      "learning_rate": 1.0871003307607497e-05,
+      "loss": 1.2275,
+      "step": 4555
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 26.644542694091797,
+      "learning_rate": 1.0868998697003109e-05,
+      "loss": 3.3596,
+      "step": 4556
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 17.826154708862305,
+      "learning_rate": 1.0866994086398719e-05,
+      "loss": 2.4254,
+      "step": 4557
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 26.188709259033203,
+      "learning_rate": 1.0864989475794327e-05,
+      "loss": 1.8766,
+      "step": 4558
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 15.313033103942871,
+      "learning_rate": 1.0862984865189939e-05,
+      "loss": 1.5112,
+      "step": 4559
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 15.900823593139648,
+      "learning_rate": 1.0860980254585548e-05,
+      "loss": 2.0939,
+      "step": 4560
+    },
+    {
+      "epoch": 1.37,
+      "eval_loss": 0.21247684955596924,
+      "eval_runtime": 43.2712,
+      "eval_samples_per_second": 34.18,
+      "eval_steps_per_second": 34.18,
+      "step": 4560
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 9.430264472961426,
+      "learning_rate": 1.0858975643981158e-05,
+      "loss": 1.7693,
+      "step": 4561
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 9.332898139953613,
+      "learning_rate": 1.0856971033376766e-05,
+      "loss": 1.048,
+      "step": 4562
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 15.586241722106934,
+      "learning_rate": 1.0854966422772378e-05,
+      "loss": 1.3119,
+      "step": 4563
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 23.395587921142578,
+      "learning_rate": 1.0852961812167988e-05,
+      "loss": 2.1297,
+      "step": 4564
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 11.959819793701172,
+      "learning_rate": 1.0850957201563596e-05,
+      "loss": 1.2409,
+      "step": 4565
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 9.007245063781738,
+      "learning_rate": 1.0848952590959208e-05,
+      "loss": 1.714,
+      "step": 4566
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 56.485538482666016,
+      "learning_rate": 1.0846947980354816e-05,
+      "loss": 2.3396,
+      "step": 4567
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 12.010300636291504,
+      "learning_rate": 1.0844943369750427e-05,
+      "loss": 2.0262,
+      "step": 4568
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 37.940059661865234,
+      "learning_rate": 1.0842938759146038e-05,
+      "loss": 1.7123,
+      "step": 4569
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 13.00949764251709,
+      "learning_rate": 1.0840934148541647e-05,
+      "loss": 1.3661,
+      "step": 4570
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 15.618173599243164,
+      "learning_rate": 1.0838929537937257e-05,
+      "loss": 1.9199,
+      "step": 4571
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 22.606422424316406,
+      "learning_rate": 1.0836924927332865e-05,
+      "loss": 1.8386,
+      "step": 4572
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 13.844141960144043,
+      "learning_rate": 1.0834920316728477e-05,
+      "loss": 1.3548,
+      "step": 4573
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 11.279787063598633,
+      "learning_rate": 1.0832915706124085e-05,
+      "loss": 1.5721,
+      "step": 4574
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 13.838190078735352,
+      "learning_rate": 1.0830911095519695e-05,
+      "loss": 1.6001,
+      "step": 4575
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 18.21175765991211,
+      "learning_rate": 1.0828906484915307e-05,
+      "loss": 1.9498,
+      "step": 4576
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 11.330121994018555,
+      "learning_rate": 1.0826901874310916e-05,
+      "loss": 1.3935,
+      "step": 4577
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 18.719507217407227,
+      "learning_rate": 1.0824897263706526e-05,
+      "loss": 1.4156,
+      "step": 4578
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 49.91222381591797,
+      "learning_rate": 1.0822892653102136e-05,
+      "loss": 2.8363,
+      "step": 4579
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 32.154563903808594,
+      "learning_rate": 1.0820888042497746e-05,
+      "loss": 2.5501,
+      "step": 4580
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 8.45114803314209,
+      "learning_rate": 1.0818883431893356e-05,
+      "loss": 1.6866,
+      "step": 4581
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 44.169677734375,
+      "learning_rate": 1.0816878821288966e-05,
+      "loss": 2.4302,
+      "step": 4582
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 25.283023834228516,
+      "learning_rate": 1.0814874210684576e-05,
+      "loss": 2.2415,
+      "step": 4583
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 16.97034454345703,
+      "learning_rate": 1.0812869600080184e-05,
+      "loss": 1.8726,
+      "step": 4584
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 9.19655990600586,
+      "learning_rate": 1.0810864989475796e-05,
+      "loss": 1.1808,
+      "step": 4585
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 18.277463912963867,
+      "learning_rate": 1.0808860378871405e-05,
+      "loss": 1.8823,
+      "step": 4586
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 21.819673538208008,
+      "learning_rate": 1.0806855768267015e-05,
+      "loss": 2.0347,
+      "step": 4587
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 9.92550277709961,
+      "learning_rate": 1.0804851157662627e-05,
+      "loss": 1.032,
+      "step": 4588
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 22.52004623413086,
+      "learning_rate": 1.0802846547058235e-05,
+      "loss": 2.1637,
+      "step": 4589
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 18.34803581237793,
+      "learning_rate": 1.0800841936453845e-05,
+      "loss": 1.9793,
+      "step": 4590
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 24.83062171936035,
+      "learning_rate": 1.0798837325849453e-05,
+      "loss": 2.2503,
+      "step": 4591
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 31.985370635986328,
+      "learning_rate": 1.0796832715245065e-05,
+      "loss": 1.9019,
+      "step": 4592
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 11.606969833374023,
+      "learning_rate": 1.0794828104640674e-05,
+      "loss": 1.5377,
+      "step": 4593
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 14.753523826599121,
+      "learning_rate": 1.0792823494036284e-05,
+      "loss": 1.6545,
+      "step": 4594
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 19.58027458190918,
+      "learning_rate": 1.0790818883431895e-05,
+      "loss": 1.7343,
+      "step": 4595
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 17.902088165283203,
+      "learning_rate": 1.0788814272827504e-05,
+      "loss": 1.763,
+      "step": 4596
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 10.415716171264648,
+      "learning_rate": 1.0786809662223114e-05,
+      "loss": 1.6063,
+      "step": 4597
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 13.281739234924316,
+      "learning_rate": 1.0784805051618722e-05,
+      "loss": 1.8627,
+      "step": 4598
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 10.292632102966309,
+      "learning_rate": 1.0782800441014334e-05,
+      "loss": 1.7306,
+      "step": 4599
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 12.653688430786133,
+      "learning_rate": 1.0780795830409944e-05,
+      "loss": 1.3753,
+      "step": 4600
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 26.442541122436523,
+      "learning_rate": 1.0778791219805553e-05,
+      "loss": 2.2257,
+      "step": 4601
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 18.668859481811523,
+      "learning_rate": 1.0776786609201164e-05,
+      "loss": 1.8788,
+      "step": 4602
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 19.046268463134766,
+      "learning_rate": 1.0774781998596773e-05,
+      "loss": 2.1165,
+      "step": 4603
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 11.989215850830078,
+      "learning_rate": 1.0772777387992383e-05,
+      "loss": 1.2069,
+      "step": 4604
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 16.762723922729492,
+      "learning_rate": 1.0770772777387993e-05,
+      "loss": 1.3351,
+      "step": 4605
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 21.549854278564453,
+      "learning_rate": 1.0768768166783603e-05,
+      "loss": 1.8815,
+      "step": 4606
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 46.833518981933594,
+      "learning_rate": 1.0766763556179213e-05,
+      "loss": 2.6215,
+      "step": 4607
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 13.283019065856934,
+      "learning_rate": 1.0764758945574823e-05,
+      "loss": 1.9159,
+      "step": 4608
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 41.622859954833984,
+      "learning_rate": 1.0762754334970433e-05,
+      "loss": 1.4153,
+      "step": 4609
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 9.129386901855469,
+      "learning_rate": 1.0760749724366042e-05,
+      "loss": 1.0581,
+      "step": 4610
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 63.40047836303711,
+      "learning_rate": 1.0758745113761653e-05,
+      "loss": 1.7917,
+      "step": 4611
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 119.53295135498047,
+      "learning_rate": 1.0756740503157263e-05,
+      "loss": 1.7074,
+      "step": 4612
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 15.246512413024902,
+      "learning_rate": 1.0754735892552872e-05,
+      "loss": 1.3584,
+      "step": 4613
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 12.469449043273926,
+      "learning_rate": 1.0752731281948484e-05,
+      "loss": 1.8839,
+      "step": 4614
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 38.270023345947266,
+      "learning_rate": 1.0750726671344092e-05,
+      "loss": 1.4275,
+      "step": 4615
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 16.700550079345703,
+      "learning_rate": 1.0748722060739702e-05,
+      "loss": 1.771,
+      "step": 4616
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 23.500558853149414,
+      "learning_rate": 1.074671745013531e-05,
+      "loss": 2.2971,
+      "step": 4617
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 17.58616065979004,
+      "learning_rate": 1.0744712839530922e-05,
+      "loss": 1.05,
+      "step": 4618
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 12.4642972946167,
+      "learning_rate": 1.0742708228926532e-05,
+      "loss": 1.1186,
+      "step": 4619
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 12.4574556350708,
+      "learning_rate": 1.074070361832214e-05,
+      "loss": 1.4973,
+      "step": 4620
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 10.377580642700195,
+      "learning_rate": 1.0738699007717753e-05,
+      "loss": 1.8076,
+      "step": 4621
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 19.34781265258789,
+      "learning_rate": 1.0736694397113361e-05,
+      "loss": 2.2722,
+      "step": 4622
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 16.955293655395508,
+      "learning_rate": 1.0734689786508971e-05,
+      "loss": 1.7219,
+      "step": 4623
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 18.286012649536133,
+      "learning_rate": 1.0732685175904583e-05,
+      "loss": 2.9251,
+      "step": 4624
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 19.848609924316406,
+      "learning_rate": 1.0730680565300191e-05,
+      "loss": 2.2197,
+      "step": 4625
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 10.013169288635254,
+      "learning_rate": 1.0728675954695801e-05,
+      "loss": 0.7125,
+      "step": 4626
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 20.401023864746094,
+      "learning_rate": 1.0726671344091411e-05,
+      "loss": 1.5097,
+      "step": 4627
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 20.424652099609375,
+      "learning_rate": 1.0724666733487021e-05,
+      "loss": 1.6132,
+      "step": 4628
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 17.268680572509766,
+      "learning_rate": 1.072266212288263e-05,
+      "loss": 1.589,
+      "step": 4629
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 10.245696067810059,
+      "learning_rate": 1.0720657512278242e-05,
+      "loss": 2.4915,
+      "step": 4630
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 21.644811630249023,
+      "learning_rate": 1.0718652901673852e-05,
+      "loss": 2.4812,
+      "step": 4631
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 19.70071792602539,
+      "learning_rate": 1.071664829106946e-05,
+      "loss": 1.8817,
+      "step": 4632
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 22.212900161743164,
+      "learning_rate": 1.0714643680465072e-05,
+      "loss": 1.9495,
+      "step": 4633
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 11.387062072753906,
+      "learning_rate": 1.071263906986068e-05,
+      "loss": 1.3702,
+      "step": 4634
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 10.107064247131348,
+      "learning_rate": 1.071063445925629e-05,
+      "loss": 1.0777,
+      "step": 4635
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 34.227264404296875,
+      "learning_rate": 1.0708629848651899e-05,
+      "loss": 1.9404,
+      "step": 4636
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 12.135993003845215,
+      "learning_rate": 1.070662523804751e-05,
+      "loss": 1.2095,
+      "step": 4637
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 15.475272178649902,
+      "learning_rate": 1.070462062744312e-05,
+      "loss": 1.4376,
+      "step": 4638
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 29.245624542236328,
+      "learning_rate": 1.0702616016838729e-05,
+      "loss": 2.0634,
+      "step": 4639
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 33.319969177246094,
+      "learning_rate": 1.070061140623434e-05,
+      "loss": 1.772,
+      "step": 4640
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 15.358686447143555,
+      "learning_rate": 1.0698606795629949e-05,
+      "loss": 1.5593,
+      "step": 4641
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 13.498563766479492,
+      "learning_rate": 1.069660218502556e-05,
+      "loss": 1.8836,
+      "step": 4642
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 11.653413772583008,
+      "learning_rate": 1.0694597574421171e-05,
+      "loss": 1.3999,
+      "step": 4643
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 7.912506103515625,
+      "learning_rate": 1.069259296381678e-05,
+      "loss": 1.0855,
+      "step": 4644
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 18.94805908203125,
+      "learning_rate": 1.069058835321239e-05,
+      "loss": 2.4709,
+      "step": 4645
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 14.374813079833984,
+      "learning_rate": 1.0688583742607998e-05,
+      "loss": 1.4454,
+      "step": 4646
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 77.0107650756836,
+      "learning_rate": 1.068657913200361e-05,
+      "loss": 2.156,
+      "step": 4647
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 14.53071403503418,
+      "learning_rate": 1.0684574521399218e-05,
+      "loss": 1.3087,
+      "step": 4648
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 18.24622917175293,
+      "learning_rate": 1.0682569910794828e-05,
+      "loss": 2.3464,
+      "step": 4649
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 7.772927761077881,
+      "learning_rate": 1.068056530019044e-05,
+      "loss": 1.3107,
+      "step": 4650
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 4.152658462524414,
+      "learning_rate": 1.0678560689586048e-05,
+      "loss": 0.467,
+      "step": 4651
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 14.69251537322998,
+      "learning_rate": 1.0676556078981658e-05,
+      "loss": 1.5746,
+      "step": 4652
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 15.464540481567383,
+      "learning_rate": 1.0674551468377268e-05,
+      "loss": 1.5305,
+      "step": 4653
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 11.303486824035645,
+      "learning_rate": 1.0672546857772879e-05,
+      "loss": 1.8223,
+      "step": 4654
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 23.572040557861328,
+      "learning_rate": 1.0670542247168489e-05,
+      "loss": 1.9877,
+      "step": 4655
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 22.017297744750977,
+      "learning_rate": 1.0668537636564099e-05,
+      "loss": 1.2985,
+      "step": 4656
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 13.40971851348877,
+      "learning_rate": 1.0666533025959709e-05,
+      "loss": 1.3887,
+      "step": 4657
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 17.30711555480957,
+      "learning_rate": 1.0664528415355317e-05,
+      "loss": 1.6859,
+      "step": 4658
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 8.817755699157715,
+      "learning_rate": 1.0662523804750929e-05,
+      "loss": 1.308,
+      "step": 4659
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 29.949804306030273,
+      "learning_rate": 1.0660519194146537e-05,
+      "loss": 1.2951,
+      "step": 4660
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 83.96745300292969,
+      "learning_rate": 1.0658514583542147e-05,
+      "loss": 2.6652,
+      "step": 4661
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 9.81541919708252,
+      "learning_rate": 1.065650997293776e-05,
+      "loss": 0.9763,
+      "step": 4662
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 24.451271057128906,
+      "learning_rate": 1.0654505362333368e-05,
+      "loss": 2.2991,
+      "step": 4663
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 21.38994789123535,
+      "learning_rate": 1.0652500751728978e-05,
+      "loss": 1.4465,
+      "step": 4664
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 9.186866760253906,
+      "learning_rate": 1.0650496141124586e-05,
+      "loss": 0.5958,
+      "step": 4665
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 18.036312103271484,
+      "learning_rate": 1.0648491530520198e-05,
+      "loss": 1.7996,
+      "step": 4666
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 13.11405086517334,
+      "learning_rate": 1.0646486919915808e-05,
+      "loss": 0.8623,
+      "step": 4667
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 15.945577621459961,
+      "learning_rate": 1.0644482309311416e-05,
+      "loss": 1.5742,
+      "step": 4668
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 10.415094375610352,
+      "learning_rate": 1.0642477698707028e-05,
+      "loss": 1.0011,
+      "step": 4669
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 13.777579307556152,
+      "learning_rate": 1.0640473088102636e-05,
+      "loss": 1.6847,
+      "step": 4670
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 29.36578369140625,
+      "learning_rate": 1.0638468477498247e-05,
+      "loss": 1.2423,
+      "step": 4671
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 26.05026626586914,
+      "learning_rate": 1.0636463866893855e-05,
+      "loss": 1.976,
+      "step": 4672
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 26.071765899658203,
+      "learning_rate": 1.0634459256289467e-05,
+      "loss": 2.2839,
+      "step": 4673
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 16.8984317779541,
+      "learning_rate": 1.0632454645685077e-05,
+      "loss": 1.3912,
+      "step": 4674
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 65.57967376708984,
+      "learning_rate": 1.0630450035080687e-05,
+      "loss": 2.2416,
+      "step": 4675
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 12.922877311706543,
+      "learning_rate": 1.0628445424476297e-05,
+      "loss": 1.5951,
+      "step": 4676
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 15.637014389038086,
+      "learning_rate": 1.0626440813871905e-05,
+      "loss": 1.9517,
+      "step": 4677
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 12.687530517578125,
+      "learning_rate": 1.0624436203267517e-05,
+      "loss": 1.3298,
+      "step": 4678
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 47.006107330322266,
+      "learning_rate": 1.0622431592663126e-05,
+      "loss": 2.6819,
+      "step": 4679
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 12.535234451293945,
+      "learning_rate": 1.0620426982058736e-05,
+      "loss": 1.4554,
+      "step": 4680
+    },
+    {
+      "epoch": 1.41,
+      "eval_loss": 0.2101414054632187,
+      "eval_runtime": 43.0905,
+      "eval_samples_per_second": 34.323,
+      "eval_steps_per_second": 34.323,
+      "step": 4680
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 9.197644233703613,
+      "learning_rate": 1.0618422371454347e-05,
+      "loss": 1.1242,
+      "step": 4681
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 11.781989097595215,
+      "learning_rate": 1.0616417760849956e-05,
+      "loss": 1.9253,
+      "step": 4682
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 41.59600830078125,
+      "learning_rate": 1.0614413150245566e-05,
+      "loss": 2.3303,
+      "step": 4683
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 13.190827369689941,
+      "learning_rate": 1.0612408539641174e-05,
+      "loss": 0.8268,
+      "step": 4684
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 57.965003967285156,
+      "learning_rate": 1.0610403929036786e-05,
+      "loss": 2.375,
+      "step": 4685
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 21.654773712158203,
+      "learning_rate": 1.0608399318432396e-05,
+      "loss": 2.0238,
+      "step": 4686
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 21.622922897338867,
+      "learning_rate": 1.0606394707828005e-05,
+      "loss": 1.8048,
+      "step": 4687
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 21.0383358001709,
+      "learning_rate": 1.0604390097223616e-05,
+      "loss": 2.2016,
+      "step": 4688
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 25.455896377563477,
+      "learning_rate": 1.0602385486619225e-05,
+      "loss": 1.9829,
+      "step": 4689
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 27.83086395263672,
+      "learning_rate": 1.0600380876014835e-05,
+      "loss": 1.637,
+      "step": 4690
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 14.252532005310059,
+      "learning_rate": 1.0598376265410443e-05,
+      "loss": 0.7723,
+      "step": 4691
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 16.23126220703125,
+      "learning_rate": 1.0596371654806055e-05,
+      "loss": 1.5121,
+      "step": 4692
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 26.036483764648438,
+      "learning_rate": 1.0594367044201665e-05,
+      "loss": 2.4891,
+      "step": 4693
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 11.481901168823242,
+      "learning_rate": 1.0592362433597273e-05,
+      "loss": 1.8866,
+      "step": 4694
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 15.858734130859375,
+      "learning_rate": 1.0590357822992885e-05,
+      "loss": 1.6499,
+      "step": 4695
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 42.61335754394531,
+      "learning_rate": 1.0588353212388494e-05,
+      "loss": 2.7187,
+      "step": 4696
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 18.403488159179688,
+      "learning_rate": 1.0586348601784104e-05,
+      "loss": 1.5052,
+      "step": 4697
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 12.110722541809082,
+      "learning_rate": 1.0584343991179715e-05,
+      "loss": 1.499,
+      "step": 4698
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 22.188568115234375,
+      "learning_rate": 1.0582339380575324e-05,
+      "loss": 2.2641,
+      "step": 4699
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 22.298145294189453,
+      "learning_rate": 1.0580334769970934e-05,
+      "loss": 1.9698,
+      "step": 4700
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 16.0262508392334,
+      "learning_rate": 1.0578330159366544e-05,
+      "loss": 1.4256,
+      "step": 4701
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 7.636526584625244,
+      "learning_rate": 1.0576325548762154e-05,
+      "loss": 0.9309,
+      "step": 4702
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 16.564451217651367,
+      "learning_rate": 1.0574320938157762e-05,
+      "loss": 0.9893,
+      "step": 4703
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 12.611454963684082,
+      "learning_rate": 1.0572316327553374e-05,
+      "loss": 1.1581,
+      "step": 4704
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 19.108123779296875,
+      "learning_rate": 1.0570311716948984e-05,
+      "loss": 2.3804,
+      "step": 4705
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 28.391921997070312,
+      "learning_rate": 1.0568307106344593e-05,
+      "loss": 1.5563,
+      "step": 4706
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 14.404617309570312,
+      "learning_rate": 1.0566302495740205e-05,
+      "loss": 1.6261,
+      "step": 4707
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 16.0683650970459,
+      "learning_rate": 1.0564297885135813e-05,
+      "loss": 1.5423,
+      "step": 4708
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 15.672560691833496,
+      "learning_rate": 1.0562293274531423e-05,
+      "loss": 1.2338,
+      "step": 4709
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 63.09310531616211,
+      "learning_rate": 1.0560288663927035e-05,
+      "loss": 3.1458,
+      "step": 4710
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 31.956239700317383,
+      "learning_rate": 1.0558284053322643e-05,
+      "loss": 1.1543,
+      "step": 4711
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 18.227100372314453,
+      "learning_rate": 1.0556279442718253e-05,
+      "loss": 1.5385,
+      "step": 4712
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 17.809656143188477,
+      "learning_rate": 1.0554274832113862e-05,
+      "loss": 1.6656,
+      "step": 4713
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 14.160889625549316,
+      "learning_rate": 1.0552270221509473e-05,
+      "loss": 1.4603,
+      "step": 4714
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 19.26714324951172,
+      "learning_rate": 1.0550265610905082e-05,
+      "loss": 1.4625,
+      "step": 4715
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 14.98919677734375,
+      "learning_rate": 1.0548261000300692e-05,
+      "loss": 1.7723,
+      "step": 4716
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 15.679314613342285,
+      "learning_rate": 1.0546256389696304e-05,
+      "loss": 2.7869,
+      "step": 4717
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 21.745820999145508,
+      "learning_rate": 1.0544251779091912e-05,
+      "loss": 1.7639,
+      "step": 4718
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 36.71981430053711,
+      "learning_rate": 1.0542247168487522e-05,
+      "loss": 1.5607,
+      "step": 4719
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 37.882877349853516,
+      "learning_rate": 1.054024255788313e-05,
+      "loss": 2.3179,
+      "step": 4720
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 13.184647560119629,
+      "learning_rate": 1.0538237947278742e-05,
+      "loss": 2.0276,
+      "step": 4721
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 7.884564399719238,
+      "learning_rate": 1.053623333667435e-05,
+      "loss": 2.1323,
+      "step": 4722
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 13.331478118896484,
+      "learning_rate": 1.053422872606996e-05,
+      "loss": 1.2308,
+      "step": 4723
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 10.356979370117188,
+      "learning_rate": 1.0532224115465573e-05,
+      "loss": 1.2406,
+      "step": 4724
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 27.083091735839844,
+      "learning_rate": 1.0530219504861181e-05,
+      "loss": 1.3356,
+      "step": 4725
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 14.349416732788086,
+      "learning_rate": 1.0528214894256791e-05,
+      "loss": 1.337,
+      "step": 4726
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 13.042084693908691,
+      "learning_rate": 1.0526210283652401e-05,
+      "loss": 1.3127,
+      "step": 4727
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 39.22003936767578,
+      "learning_rate": 1.0524205673048011e-05,
+      "loss": 1.7421,
+      "step": 4728
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 7.622201919555664,
+      "learning_rate": 1.0522201062443623e-05,
+      "loss": 0.9831,
+      "step": 4729
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 31.560455322265625,
+      "learning_rate": 1.0520196451839231e-05,
+      "loss": 2.0777,
+      "step": 4730
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 13.57126235961914,
+      "learning_rate": 1.0518191841234841e-05,
+      "loss": 1.4528,
+      "step": 4731
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 25.839723587036133,
+      "learning_rate": 1.051618723063045e-05,
+      "loss": 1.3498,
+      "step": 4732
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 8.843565940856934,
+      "learning_rate": 1.0514182620026062e-05,
+      "loss": 2.0511,
+      "step": 4733
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 13.93571949005127,
+      "learning_rate": 1.051217800942167e-05,
+      "loss": 1.4334,
+      "step": 4734
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 32.184539794921875,
+      "learning_rate": 1.051017339881728e-05,
+      "loss": 2.1138,
+      "step": 4735
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 12.731632232666016,
+      "learning_rate": 1.0508168788212892e-05,
+      "loss": 1.506,
+      "step": 4736
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 9.572550773620605,
+      "learning_rate": 1.05061641776085e-05,
+      "loss": 2.4816,
+      "step": 4737
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 13.963497161865234,
+      "learning_rate": 1.050415956700411e-05,
+      "loss": 1.4354,
+      "step": 4738
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 19.97682762145996,
+      "learning_rate": 1.0502154956399719e-05,
+      "loss": 1.7492,
+      "step": 4739
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 11.23585033416748,
+      "learning_rate": 1.050015034579533e-05,
+      "loss": 1.396,
+      "step": 4740
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 21.9536075592041,
+      "learning_rate": 1.049814573519094e-05,
+      "loss": 1.1373,
+      "step": 4741
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 15.670050621032715,
+      "learning_rate": 1.0496141124586549e-05,
+      "loss": 1.9185,
+      "step": 4742
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 18.927589416503906,
+      "learning_rate": 1.049413651398216e-05,
+      "loss": 1.7097,
+      "step": 4743
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 14.530732154846191,
+      "learning_rate": 1.049213190337777e-05,
+      "loss": 1.1204,
+      "step": 4744
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 29.094755172729492,
+      "learning_rate": 1.049012729277338e-05,
+      "loss": 1.787,
+      "step": 4745
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 25.101224899291992,
+      "learning_rate": 1.048812268216899e-05,
+      "loss": 2.8473,
+      "step": 4746
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 17.750200271606445,
+      "learning_rate": 1.04861180715646e-05,
+      "loss": 1.8474,
+      "step": 4747
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 22.220870971679688,
+      "learning_rate": 1.048411346096021e-05,
+      "loss": 1.8063,
+      "step": 4748
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 7.666170120239258,
+      "learning_rate": 1.048210885035582e-05,
+      "loss": 1.4058,
+      "step": 4749
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 28.113630294799805,
+      "learning_rate": 1.048010423975143e-05,
+      "loss": 2.1896,
+      "step": 4750
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 13.658616065979004,
+      "learning_rate": 1.0478099629147038e-05,
+      "loss": 1.7455,
+      "step": 4751
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 11.242774963378906,
+      "learning_rate": 1.047609501854265e-05,
+      "loss": 1.9807,
+      "step": 4752
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 11.421576499938965,
+      "learning_rate": 1.047409040793826e-05,
+      "loss": 1.5406,
+      "step": 4753
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 24.668371200561523,
+      "learning_rate": 1.0472085797333868e-05,
+      "loss": 2.2752,
+      "step": 4754
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 24.07050132751465,
+      "learning_rate": 1.047008118672948e-05,
+      "loss": 1.4799,
+      "step": 4755
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 38.27885055541992,
+      "learning_rate": 1.0468076576125088e-05,
+      "loss": 1.5369,
+      "step": 4756
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 20.224210739135742,
+      "learning_rate": 1.0466071965520699e-05,
+      "loss": 1.9254,
+      "step": 4757
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 10.304624557495117,
+      "learning_rate": 1.0464067354916307e-05,
+      "loss": 1.5751,
+      "step": 4758
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 23.387136459350586,
+      "learning_rate": 1.0462062744311919e-05,
+      "loss": 1.2816,
+      "step": 4759
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 8.415315628051758,
+      "learning_rate": 1.0460058133707529e-05,
+      "loss": 1.2024,
+      "step": 4760
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 15.556396484375,
+      "learning_rate": 1.0458053523103137e-05,
+      "loss": 1.7196,
+      "step": 4761
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 70.80048370361328,
+      "learning_rate": 1.0456048912498749e-05,
+      "loss": 1.6606,
+      "step": 4762
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 8.032504081726074,
+      "learning_rate": 1.0454044301894357e-05,
+      "loss": 1.4636,
+      "step": 4763
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 13.399280548095703,
+      "learning_rate": 1.0452039691289967e-05,
+      "loss": 2.0113,
+      "step": 4764
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 10.88759708404541,
+      "learning_rate": 1.0450035080685576e-05,
+      "loss": 1.7287,
+      "step": 4765
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 46.82052230834961,
+      "learning_rate": 1.0448030470081188e-05,
+      "loss": 2.7536,
+      "step": 4766
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 19.954078674316406,
+      "learning_rate": 1.0446025859476798e-05,
+      "loss": 1.6187,
+      "step": 4767
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 46.83697509765625,
+      "learning_rate": 1.0444021248872406e-05,
+      "loss": 2.2134,
+      "step": 4768
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 29.12272834777832,
+      "learning_rate": 1.0442016638268018e-05,
+      "loss": 2.3703,
+      "step": 4769
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 37.321956634521484,
+      "learning_rate": 1.0440012027663626e-05,
+      "loss": 1.2239,
+      "step": 4770
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 18.08650779724121,
+      "learning_rate": 1.0438007417059236e-05,
+      "loss": 1.4429,
+      "step": 4771
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 15.265766143798828,
+      "learning_rate": 1.0436002806454848e-05,
+      "loss": 1.2304,
+      "step": 4772
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 16.230012893676758,
+      "learning_rate": 1.0433998195850457e-05,
+      "loss": 1.2576,
+      "step": 4773
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 27.102861404418945,
+      "learning_rate": 1.0431993585246067e-05,
+      "loss": 2.3902,
+      "step": 4774
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 9.05881404876709,
+      "learning_rate": 1.0429988974641677e-05,
+      "loss": 0.8901,
+      "step": 4775
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 24.15192985534668,
+      "learning_rate": 1.0427984364037287e-05,
+      "loss": 1.4729,
+      "step": 4776
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 11.738104820251465,
+      "learning_rate": 1.0425979753432895e-05,
+      "loss": 1.5979,
+      "step": 4777
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 17.82452392578125,
+      "learning_rate": 1.0423975142828507e-05,
+      "loss": 1.2219,
+      "step": 4778
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 10.79785442352295,
+      "learning_rate": 1.0421970532224117e-05,
+      "loss": 1.1237,
+      "step": 4779
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 66.31954193115234,
+      "learning_rate": 1.0419965921619725e-05,
+      "loss": 1.1348,
+      "step": 4780
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 14.166802406311035,
+      "learning_rate": 1.0417961311015337e-05,
+      "loss": 1.5118,
+      "step": 4781
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 19.842845916748047,
+      "learning_rate": 1.0415956700410946e-05,
+      "loss": 1.22,
+      "step": 4782
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 19.89260482788086,
+      "learning_rate": 1.0413952089806556e-05,
+      "loss": 1.4466,
+      "step": 4783
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 8.727401733398438,
+      "learning_rate": 1.0411947479202167e-05,
+      "loss": 1.2331,
+      "step": 4784
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 16.119182586669922,
+      "learning_rate": 1.0409942868597776e-05,
+      "loss": 2.1112,
+      "step": 4785
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 8.426682472229004,
+      "learning_rate": 1.0407938257993386e-05,
+      "loss": 0.7954,
+      "step": 4786
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 11.057311058044434,
+      "learning_rate": 1.0405933647388994e-05,
+      "loss": 1.4438,
+      "step": 4787
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 11.847160339355469,
+      "learning_rate": 1.0403929036784606e-05,
+      "loss": 1.6443,
+      "step": 4788
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 52.34111022949219,
+      "learning_rate": 1.0401924426180214e-05,
+      "loss": 1.8,
+      "step": 4789
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 12.100299835205078,
+      "learning_rate": 1.0399919815575825e-05,
+      "loss": 0.9117,
+      "step": 4790
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 25.9334774017334,
+      "learning_rate": 1.0397915204971436e-05,
+      "loss": 1.6836,
+      "step": 4791
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 16.446218490600586,
+      "learning_rate": 1.0395910594367045e-05,
+      "loss": 1.581,
+      "step": 4792
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 16.541593551635742,
+      "learning_rate": 1.0393905983762655e-05,
+      "loss": 1.6899,
+      "step": 4793
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 14.012060165405273,
+      "learning_rate": 1.0391901373158265e-05,
+      "loss": 1.3736,
+      "step": 4794
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 18.858366012573242,
+      "learning_rate": 1.0389896762553875e-05,
+      "loss": 1.2005,
+      "step": 4795
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 11.462150573730469,
+      "learning_rate": 1.0387892151949485e-05,
+      "loss": 1.0013,
+      "step": 4796
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 13.834492683410645,
+      "learning_rate": 1.0385887541345095e-05,
+      "loss": 0.8818,
+      "step": 4797
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 44.42348861694336,
+      "learning_rate": 1.0383882930740705e-05,
+      "loss": 2.2629,
+      "step": 4798
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 21.56435775756836,
+      "learning_rate": 1.0381878320136314e-05,
+      "loss": 2.646,
+      "step": 4799
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 12.599276542663574,
+      "learning_rate": 1.0379873709531925e-05,
+      "loss": 1.2467,
+      "step": 4800
+    },
+    {
+      "epoch": 1.44,
+      "eval_loss": 0.20445984601974487,
+      "eval_runtime": 43.0646,
+      "eval_samples_per_second": 34.344,
+      "eval_steps_per_second": 34.344,
+      "step": 4800
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 21.611980438232422,
+      "learning_rate": 1.0377869098927534e-05,
+      "loss": 1.8799,
+      "step": 4801
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 26.111543655395508,
+      "learning_rate": 1.0375864488323144e-05,
+      "loss": 1.8642,
+      "step": 4802
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 19.41090965270996,
+      "learning_rate": 1.0373859877718756e-05,
+      "loss": 1.8098,
+      "step": 4803
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 23.62296485900879,
+      "learning_rate": 1.0371855267114364e-05,
+      "loss": 2.0232,
+      "step": 4804
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 25.114355087280273,
+      "learning_rate": 1.0369850656509974e-05,
+      "loss": 1.348,
+      "step": 4805
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 15.635001182556152,
+      "learning_rate": 1.0367846045905583e-05,
+      "loss": 1.4803,
+      "step": 4806
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 12.778668403625488,
+      "learning_rate": 1.0365841435301194e-05,
+      "loss": 1.837,
+      "step": 4807
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 19.424400329589844,
+      "learning_rate": 1.0363836824696803e-05,
+      "loss": 1.6641,
+      "step": 4808
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 9.217851638793945,
+      "learning_rate": 1.0361832214092413e-05,
+      "loss": 0.925,
+      "step": 4809
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 26.674348831176758,
+      "learning_rate": 1.0359827603488025e-05,
+      "loss": 1.5977,
+      "step": 4810
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 13.42351245880127,
+      "learning_rate": 1.0357822992883633e-05,
+      "loss": 1.3927,
+      "step": 4811
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 36.85655975341797,
+      "learning_rate": 1.0355818382279243e-05,
+      "loss": 1.8373,
+      "step": 4812
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 49.87083435058594,
+      "learning_rate": 1.0353813771674851e-05,
+      "loss": 1.6954,
+      "step": 4813
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 30.802719116210938,
+      "learning_rate": 1.0351809161070463e-05,
+      "loss": 2.0961,
+      "step": 4814
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 22.26266860961914,
+      "learning_rate": 1.0349804550466073e-05,
+      "loss": 1.3555,
+      "step": 4815
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 18.78736114501953,
+      "learning_rate": 1.0347799939861682e-05,
+      "loss": 1.4008,
+      "step": 4816
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 10.189375877380371,
+      "learning_rate": 1.0345795329257293e-05,
+      "loss": 0.6236,
+      "step": 4817
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 14.57485580444336,
+      "learning_rate": 1.0343790718652902e-05,
+      "loss": 1.3009,
+      "step": 4818
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 16.516605377197266,
+      "learning_rate": 1.0341786108048512e-05,
+      "loss": 1.3581,
+      "step": 4819
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 12.826788902282715,
+      "learning_rate": 1.0339781497444122e-05,
+      "loss": 1.9941,
+      "step": 4820
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 43.45224380493164,
+      "learning_rate": 1.0337776886839732e-05,
+      "loss": 1.8419,
+      "step": 4821
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 33.118247985839844,
+      "learning_rate": 1.0335772276235342e-05,
+      "loss": 1.7156,
+      "step": 4822
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 29.060731887817383,
+      "learning_rate": 1.0333767665630952e-05,
+      "loss": 1.5526,
+      "step": 4823
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 30.88375473022461,
+      "learning_rate": 1.0331763055026562e-05,
+      "loss": 1.9801,
+      "step": 4824
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 26.168588638305664,
+      "learning_rate": 1.032975844442217e-05,
+      "loss": 1.63,
+      "step": 4825
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 20.537208557128906,
+      "learning_rate": 1.0327753833817783e-05,
+      "loss": 1.582,
+      "step": 4826
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 15.51192855834961,
+      "learning_rate": 1.0325749223213393e-05,
+      "loss": 2.2813,
+      "step": 4827
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 18.50126838684082,
+      "learning_rate": 1.0323744612609001e-05,
+      "loss": 1.5379,
+      "step": 4828
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 19.359989166259766,
+      "learning_rate": 1.0321740002004613e-05,
+      "loss": 1.4196,
+      "step": 4829
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 71.47821807861328,
+      "learning_rate": 1.0319735391400221e-05,
+      "loss": 2.627,
+      "step": 4830
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 8.43783950805664,
+      "learning_rate": 1.0317730780795831e-05,
+      "loss": 1.1434,
+      "step": 4831
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 49.501930236816406,
+      "learning_rate": 1.031572617019144e-05,
+      "loss": 2.0673,
+      "step": 4832
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 11.1619234085083,
+      "learning_rate": 1.0313721559587051e-05,
+      "loss": 1.8033,
+      "step": 4833
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 7.6200408935546875,
+      "learning_rate": 1.0311716948982662e-05,
+      "loss": 0.8928,
+      "step": 4834
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 19.349905014038086,
+      "learning_rate": 1.030971233837827e-05,
+      "loss": 1.6486,
+      "step": 4835
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 8.616739273071289,
+      "learning_rate": 1.0307707727773882e-05,
+      "loss": 1.2562,
+      "step": 4836
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 30.071773529052734,
+      "learning_rate": 1.030570311716949e-05,
+      "loss": 1.4047,
+      "step": 4837
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 12.02312183380127,
+      "learning_rate": 1.03036985065651e-05,
+      "loss": 1.5015,
+      "step": 4838
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 13.285801887512207,
+      "learning_rate": 1.0301693895960712e-05,
+      "loss": 1.4643,
+      "step": 4839
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 20.811077117919922,
+      "learning_rate": 1.029968928535632e-05,
+      "loss": 1.8669,
+      "step": 4840
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 21.803550720214844,
+      "learning_rate": 1.029768467475193e-05,
+      "loss": 2.0773,
+      "step": 4841
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 35.751792907714844,
+      "learning_rate": 1.0295680064147539e-05,
+      "loss": 2.3893,
+      "step": 4842
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 14.943572044372559,
+      "learning_rate": 1.029367545354315e-05,
+      "loss": 1.9117,
+      "step": 4843
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 19.128864288330078,
+      "learning_rate": 1.0291670842938759e-05,
+      "loss": 1.9138,
+      "step": 4844
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 17.63079833984375,
+      "learning_rate": 1.028966623233437e-05,
+      "loss": 1.2719,
+      "step": 4845
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 40.6061897277832,
+      "learning_rate": 1.028766162172998e-05,
+      "loss": 1.8636,
+      "step": 4846
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 11.293282508850098,
+      "learning_rate": 1.028565701112559e-05,
+      "loss": 1.1617,
+      "step": 4847
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 7.563941955566406,
+      "learning_rate": 1.0283652400521201e-05,
+      "loss": 1.266,
+      "step": 4848
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 37.318321228027344,
+      "learning_rate": 1.028164778991681e-05,
+      "loss": 4.1312,
+      "step": 4849
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 15.308874130249023,
+      "learning_rate": 1.027964317931242e-05,
+      "loss": 1.563,
+      "step": 4850
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 12.4258451461792,
+      "learning_rate": 1.0277638568708028e-05,
+      "loss": 1.7674,
+      "step": 4851
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 12.700484275817871,
+      "learning_rate": 1.027563395810364e-05,
+      "loss": 1.2176,
+      "step": 4852
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 26.99094581604004,
+      "learning_rate": 1.027362934749925e-05,
+      "loss": 1.5912,
+      "step": 4853
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 52.29587936401367,
+      "learning_rate": 1.0271624736894858e-05,
+      "loss": 2.6716,
+      "step": 4854
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 28.816272735595703,
+      "learning_rate": 1.026962012629047e-05,
+      "loss": 2.7605,
+      "step": 4855
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 28.791126251220703,
+      "learning_rate": 1.0267615515686078e-05,
+      "loss": 1.8248,
+      "step": 4856
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 17.190807342529297,
+      "learning_rate": 1.0265610905081688e-05,
+      "loss": 1.6728,
+      "step": 4857
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 17.33432960510254,
+      "learning_rate": 1.02636062944773e-05,
+      "loss": 1.427,
+      "step": 4858
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 18.81448745727539,
+      "learning_rate": 1.0261601683872909e-05,
+      "loss": 1.4079,
+      "step": 4859
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 26.18571662902832,
+      "learning_rate": 1.0259597073268519e-05,
+      "loss": 1.3066,
+      "step": 4860
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 22.120250701904297,
+      "learning_rate": 1.0257592462664127e-05,
+      "loss": 1.8271,
+      "step": 4861
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 19.113744735717773,
+      "learning_rate": 1.0255587852059739e-05,
+      "loss": 0.9674,
+      "step": 4862
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 15.18375301361084,
+      "learning_rate": 1.0253583241455347e-05,
+      "loss": 1.3229,
+      "step": 4863
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 14.799633979797363,
+      "learning_rate": 1.0251578630850957e-05,
+      "loss": 1.1113,
+      "step": 4864
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 13.80958366394043,
+      "learning_rate": 1.0249574020246569e-05,
+      "loss": 1.6729,
+      "step": 4865
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 38.37821578979492,
+      "learning_rate": 1.0247569409642177e-05,
+      "loss": 1.5533,
+      "step": 4866
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 28.763893127441406,
+      "learning_rate": 1.0245564799037788e-05,
+      "loss": 1.699,
+      "step": 4867
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 20.876445770263672,
+      "learning_rate": 1.0243560188433398e-05,
+      "loss": 1.3333,
+      "step": 4868
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 52.730857849121094,
+      "learning_rate": 1.0241555577829008e-05,
+      "loss": 1.4987,
+      "step": 4869
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 13.893447875976562,
+      "learning_rate": 1.0239550967224618e-05,
+      "loss": 1.6037,
+      "step": 4870
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 28.038808822631836,
+      "learning_rate": 1.0237546356620228e-05,
+      "loss": 3.0782,
+      "step": 4871
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 26.272857666015625,
+      "learning_rate": 1.0235541746015838e-05,
+      "loss": 1.7294,
+      "step": 4872
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 87.54938507080078,
+      "learning_rate": 1.0233537135411446e-05,
+      "loss": 2.8123,
+      "step": 4873
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 25.847455978393555,
+      "learning_rate": 1.0231532524807058e-05,
+      "loss": 1.3259,
+      "step": 4874
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 10.328527450561523,
+      "learning_rate": 1.0229527914202666e-05,
+      "loss": 0.9254,
+      "step": 4875
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 19.659297943115234,
+      "learning_rate": 1.0227523303598277e-05,
+      "loss": 1.1272,
+      "step": 4876
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 25.877544403076172,
+      "learning_rate": 1.0225518692993888e-05,
+      "loss": 1.8762,
+      "step": 4877
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 22.409603118896484,
+      "learning_rate": 1.0223514082389497e-05,
+      "loss": 2.1701,
+      "step": 4878
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 48.738155364990234,
+      "learning_rate": 1.0221509471785107e-05,
+      "loss": 2.1954,
+      "step": 4879
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 13.183887481689453,
+      "learning_rate": 1.0219504861180715e-05,
+      "loss": 1.5571,
+      "step": 4880
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 18.5463809967041,
+      "learning_rate": 1.0217500250576327e-05,
+      "loss": 1.5617,
+      "step": 4881
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 22.341135025024414,
+      "learning_rate": 1.0215495639971937e-05,
+      "loss": 1.0747,
+      "step": 4882
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 10.601929664611816,
+      "learning_rate": 1.0213491029367545e-05,
+      "loss": 1.539,
+      "step": 4883
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 12.355496406555176,
+      "learning_rate": 1.0211486418763157e-05,
+      "loss": 1.3754,
+      "step": 4884
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 11.556707382202148,
+      "learning_rate": 1.0209481808158766e-05,
+      "loss": 1.5293,
+      "step": 4885
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 19.798812866210938,
+      "learning_rate": 1.0207477197554376e-05,
+      "loss": 0.9688,
+      "step": 4886
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 13.712343215942383,
+      "learning_rate": 1.0205472586949984e-05,
+      "loss": 1.5041,
+      "step": 4887
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 10.897645950317383,
+      "learning_rate": 1.0203467976345596e-05,
+      "loss": 1.0905,
+      "step": 4888
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 28.161396026611328,
+      "learning_rate": 1.0201463365741206e-05,
+      "loss": 1.7225,
+      "step": 4889
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 34.16568374633789,
+      "learning_rate": 1.0199458755136814e-05,
+      "loss": 2.0862,
+      "step": 4890
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 19.398710250854492,
+      "learning_rate": 1.0197454144532426e-05,
+      "loss": 1.5471,
+      "step": 4891
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 24.547252655029297,
+      "learning_rate": 1.0195449533928035e-05,
+      "loss": 3.0927,
+      "step": 4892
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 11.176749229431152,
+      "learning_rate": 1.0193444923323645e-05,
+      "loss": 1.1639,
+      "step": 4893
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 50.76742172241211,
+      "learning_rate": 1.0191440312719255e-05,
+      "loss": 2.6678,
+      "step": 4894
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 9.156387329101562,
+      "learning_rate": 1.0189435702114865e-05,
+      "loss": 0.9565,
+      "step": 4895
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 23.076967239379883,
+      "learning_rate": 1.0187431091510475e-05,
+      "loss": 1.6497,
+      "step": 4896
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 16.60446548461914,
+      "learning_rate": 1.0185426480906085e-05,
+      "loss": 1.5919,
+      "step": 4897
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 12.027603149414062,
+      "learning_rate": 1.0183421870301695e-05,
+      "loss": 2.2724,
+      "step": 4898
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 18.373804092407227,
+      "learning_rate": 1.0181417259697303e-05,
+      "loss": 0.8864,
+      "step": 4899
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 33.86744689941406,
+      "learning_rate": 1.0179412649092915e-05,
+      "loss": 2.4096,
+      "step": 4900
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 19.00617218017578,
+      "learning_rate": 1.0177408038488525e-05,
+      "loss": 1.4766,
+      "step": 4901
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 10.875265121459961,
+      "learning_rate": 1.0175403427884134e-05,
+      "loss": 1.6248,
+      "step": 4902
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 18.423274993896484,
+      "learning_rate": 1.0173398817279745e-05,
+      "loss": 2.1238,
+      "step": 4903
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 14.276087760925293,
+      "learning_rate": 1.0171394206675354e-05,
+      "loss": 1.471,
+      "step": 4904
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 8.41922664642334,
+      "learning_rate": 1.0169389596070964e-05,
+      "loss": 0.8771,
+      "step": 4905
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 17.14394760131836,
+      "learning_rate": 1.0167384985466572e-05,
+      "loss": 1.0387,
+      "step": 4906
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 20.45330047607422,
+      "learning_rate": 1.0165380374862184e-05,
+      "loss": 1.9674,
+      "step": 4907
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 6.636265754699707,
+      "learning_rate": 1.0163375764257794e-05,
+      "loss": 0.3816,
+      "step": 4908
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 15.284008979797363,
+      "learning_rate": 1.0161371153653403e-05,
+      "loss": 1.6652,
+      "step": 4909
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 16.42159080505371,
+      "learning_rate": 1.0159366543049014e-05,
+      "loss": 2.3129,
+      "step": 4910
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 19.10989761352539,
+      "learning_rate": 1.0157361932444623e-05,
+      "loss": 1.8581,
+      "step": 4911
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 30.188232421875,
+      "learning_rate": 1.0155357321840233e-05,
+      "loss": 1.5959,
+      "step": 4912
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 12.307696342468262,
+      "learning_rate": 1.0153352711235845e-05,
+      "loss": 1.6547,
+      "step": 4913
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 12.962136268615723,
+      "learning_rate": 1.0151348100631453e-05,
+      "loss": 1.8137,
+      "step": 4914
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 10.3368558883667,
+      "learning_rate": 1.0149343490027063e-05,
+      "loss": 0.9236,
+      "step": 4915
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 20.91614532470703,
+      "learning_rate": 1.0147338879422673e-05,
+      "loss": 1.7155,
+      "step": 4916
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 38.6691780090332,
+      "learning_rate": 1.0145334268818283e-05,
+      "loss": 2.1508,
+      "step": 4917
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 9.685796737670898,
+      "learning_rate": 1.0143329658213892e-05,
+      "loss": 1.7707,
+      "step": 4918
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 14.471162796020508,
+      "learning_rate": 1.0141325047609503e-05,
+      "loss": 1.5946,
+      "step": 4919
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 18.22927474975586,
+      "learning_rate": 1.0139320437005114e-05,
+      "loss": 1.2375,
+      "step": 4920
+    },
+    {
+      "epoch": 1.48,
+      "eval_loss": 0.2065747231245041,
+      "eval_runtime": 43.2855,
+      "eval_samples_per_second": 34.169,
+      "eval_steps_per_second": 34.169,
+      "step": 4920
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 51.773658752441406,
+      "learning_rate": 1.0137315826400722e-05,
+      "loss": 0.986,
+      "step": 4921
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 105.13095092773438,
+      "learning_rate": 1.0135311215796334e-05,
+      "loss": 2.8983,
+      "step": 4922
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 14.783707618713379,
+      "learning_rate": 1.0133306605191942e-05,
+      "loss": 1.9219,
+      "step": 4923
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 8.431492805480957,
+      "learning_rate": 1.0131301994587552e-05,
+      "loss": 0.9575,
+      "step": 4924
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 14.085335731506348,
+      "learning_rate": 1.0129297383983164e-05,
+      "loss": 1.2251,
+      "step": 4925
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 17.102575302124023,
+      "learning_rate": 1.0127292773378772e-05,
+      "loss": 1.4003,
+      "step": 4926
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 13.873455047607422,
+      "learning_rate": 1.0125288162774382e-05,
+      "loss": 1.3603,
+      "step": 4927
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 8.495047569274902,
+      "learning_rate": 1.012328355216999e-05,
+      "loss": 1.8705,
+      "step": 4928
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 12.480233192443848,
+      "learning_rate": 1.0121278941565603e-05,
+      "loss": 1.6846,
+      "step": 4929
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 9.334129333496094,
+      "learning_rate": 1.0119274330961211e-05,
+      "loss": 1.924,
+      "step": 4930
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 12.200242042541504,
+      "learning_rate": 1.0117269720356821e-05,
+      "loss": 1.4727,
+      "step": 4931
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 10.241484642028809,
+      "learning_rate": 1.0115265109752433e-05,
+      "loss": 1.2527,
+      "step": 4932
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 18.393444061279297,
+      "learning_rate": 1.0113260499148041e-05,
+      "loss": 1.9137,
+      "step": 4933
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 30.996809005737305,
+      "learning_rate": 1.0111255888543651e-05,
+      "loss": 1.9626,
+      "step": 4934
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 22.243946075439453,
+      "learning_rate": 1.010925127793926e-05,
+      "loss": 1.4565,
+      "step": 4935
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 10.407564163208008,
+      "learning_rate": 1.0107246667334871e-05,
+      "loss": 1.0015,
+      "step": 4936
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 17.139793395996094,
+      "learning_rate": 1.010524205673048e-05,
+      "loss": 1.3928,
+      "step": 4937
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 34.3438606262207,
+      "learning_rate": 1.010323744612609e-05,
+      "loss": 1.5863,
+      "step": 4938
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 14.947382926940918,
+      "learning_rate": 1.0101232835521702e-05,
+      "loss": 1.3544,
+      "step": 4939
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 11.249651908874512,
+      "learning_rate": 1.009922822491731e-05,
+      "loss": 1.3051,
+      "step": 4940
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 16.109973907470703,
+      "learning_rate": 1.009722361431292e-05,
+      "loss": 2.0421,
+      "step": 4941
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 14.892985343933105,
+      "learning_rate": 1.009521900370853e-05,
+      "loss": 1.1327,
+      "step": 4942
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 27.952659606933594,
+      "learning_rate": 1.009321439310414e-05,
+      "loss": 2.3613,
+      "step": 4943
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 27.172393798828125,
+      "learning_rate": 1.009120978249975e-05,
+      "loss": 2.1473,
+      "step": 4944
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 9.170857429504395,
+      "learning_rate": 1.008920517189536e-05,
+      "loss": 1.2239,
+      "step": 4945
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 11.424710273742676,
+      "learning_rate": 1.008720056129097e-05,
+      "loss": 0.8955,
+      "step": 4946
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 13.28514289855957,
+      "learning_rate": 1.0085195950686579e-05,
+      "loss": 1.6111,
+      "step": 4947
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 32.51649475097656,
+      "learning_rate": 1.008319134008219e-05,
+      "loss": 1.9871,
+      "step": 4948
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 15.511556625366211,
+      "learning_rate": 1.00811867294778e-05,
+      "loss": 1.37,
+      "step": 4949
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 16.7358341217041,
+      "learning_rate": 1.007918211887341e-05,
+      "loss": 1.6897,
+      "step": 4950
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 13.070718765258789,
+      "learning_rate": 1.0077177508269021e-05,
+      "loss": 0.666,
+      "step": 4951
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 22.743227005004883,
+      "learning_rate": 1.007517289766463e-05,
+      "loss": 1.8756,
+      "step": 4952
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 20.257579803466797,
+      "learning_rate": 1.007316828706024e-05,
+      "loss": 1.9218,
+      "step": 4953
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 32.02715301513672,
+      "learning_rate": 1.0071163676455848e-05,
+      "loss": 1.2685,
+      "step": 4954
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 15.203670501708984,
+      "learning_rate": 1.006915906585146e-05,
+      "loss": 1.5685,
+      "step": 4955
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 7.677938461303711,
+      "learning_rate": 1.006715445524707e-05,
+      "loss": 1.0445,
+      "step": 4956
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 24.447038650512695,
+      "learning_rate": 1.0065149844642678e-05,
+      "loss": 1.7323,
+      "step": 4957
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 12.981391906738281,
+      "learning_rate": 1.006314523403829e-05,
+      "loss": 1.9296,
+      "step": 4958
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 38.580570220947266,
+      "learning_rate": 1.0061140623433898e-05,
+      "loss": 1.909,
+      "step": 4959
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 11.870725631713867,
+      "learning_rate": 1.0059136012829508e-05,
+      "loss": 1.3958,
+      "step": 4960
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 7.389049530029297,
+      "learning_rate": 1.0057131402225117e-05,
+      "loss": 0.8741,
+      "step": 4961
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 12.032258033752441,
+      "learning_rate": 1.0055126791620729e-05,
+      "loss": 1.4206,
+      "step": 4962
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 22.42498207092285,
+      "learning_rate": 1.0053122181016339e-05,
+      "loss": 2.5246,
+      "step": 4963
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 12.639444351196289,
+      "learning_rate": 1.0051117570411949e-05,
+      "loss": 1.8031,
+      "step": 4964
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 33.977386474609375,
+      "learning_rate": 1.0049112959807559e-05,
+      "loss": 1.595,
+      "step": 4965
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 13.614667892456055,
+      "learning_rate": 1.0047108349203167e-05,
+      "loss": 0.975,
+      "step": 4966
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 18.450183868408203,
+      "learning_rate": 1.0045103738598779e-05,
+      "loss": 2.5177,
+      "step": 4967
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 15.293846130371094,
+      "learning_rate": 1.0043099127994389e-05,
+      "loss": 1.472,
+      "step": 4968
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 16.424936294555664,
+      "learning_rate": 1.0041094517389997e-05,
+      "loss": 1.477,
+      "step": 4969
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 13.09062385559082,
+      "learning_rate": 1.003908990678561e-05,
+      "loss": 1.414,
+      "step": 4970
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 21.331588745117188,
+      "learning_rate": 1.0037085296181218e-05,
+      "loss": 2.6522,
+      "step": 4971
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 44.23732376098633,
+      "learning_rate": 1.0035080685576828e-05,
+      "loss": 1.6369,
+      "step": 4972
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 13.58181095123291,
+      "learning_rate": 1.0033076074972436e-05,
+      "loss": 1.6649,
+      "step": 4973
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 25.254730224609375,
+      "learning_rate": 1.0031071464368048e-05,
+      "loss": 2.0687,
+      "step": 4974
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 10.236112594604492,
+      "learning_rate": 1.0029066853763658e-05,
+      "loss": 1.3942,
+      "step": 4975
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 13.936944961547852,
+      "learning_rate": 1.0027062243159266e-05,
+      "loss": 2.3225,
+      "step": 4976
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 11.376901626586914,
+      "learning_rate": 1.0025057632554878e-05,
+      "loss": 1.3294,
+      "step": 4977
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 15.195333480834961,
+      "learning_rate": 1.0023053021950487e-05,
+      "loss": 2.3572,
+      "step": 4978
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 35.94542694091797,
+      "learning_rate": 1.0021048411346097e-05,
+      "loss": 3.3368,
+      "step": 4979
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 21.131452560424805,
+      "learning_rate": 1.0019043800741705e-05,
+      "loss": 1.618,
+      "step": 4980
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 22.635162353515625,
+      "learning_rate": 1.0017039190137317e-05,
+      "loss": 2.0817,
+      "step": 4981
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 22.40682601928711,
+      "learning_rate": 1.0015034579532927e-05,
+      "loss": 1.2649,
+      "step": 4982
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 7.878605365753174,
+      "learning_rate": 1.0013029968928535e-05,
+      "loss": 1.3573,
+      "step": 4983
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 19.825239181518555,
+      "learning_rate": 1.0011025358324147e-05,
+      "loss": 1.9824,
+      "step": 4984
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 14.39383602142334,
+      "learning_rate": 1.0009020747719755e-05,
+      "loss": 1.8926,
+      "step": 4985
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 21.175561904907227,
+      "learning_rate": 1.0007016137115366e-05,
+      "loss": 1.114,
+      "step": 4986
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 17.892271041870117,
+      "learning_rate": 1.0005011526510977e-05,
+      "loss": 1.2682,
+      "step": 4987
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 14.41032600402832,
+      "learning_rate": 1.0003006915906586e-05,
+      "loss": 1.3076,
+      "step": 4988
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 10.802579879760742,
+      "learning_rate": 1.0001002305302196e-05,
+      "loss": 1.1787,
+      "step": 4989
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 16.22612762451172,
+      "learning_rate": 9.998997694697806e-06,
+      "loss": 1.8036,
+      "step": 4990
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 9.207633018493652,
+      "learning_rate": 9.996993084093416e-06,
+      "loss": 1.2887,
+      "step": 4991
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 109.87674713134766,
+      "learning_rate": 9.994988473489026e-06,
+      "loss": 2.7811,
+      "step": 4992
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 10.483904838562012,
+      "learning_rate": 9.992983862884636e-06,
+      "loss": 2.3548,
+      "step": 4993
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 8.791914939880371,
+      "learning_rate": 9.990979252280245e-06,
+      "loss": 1.0692,
+      "step": 4994
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 12.697139739990234,
+      "learning_rate": 9.988974641675855e-06,
+      "loss": 1.2917,
+      "step": 4995
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 15.190261840820312,
+      "learning_rate": 9.986970031071466e-06,
+      "loss": 1.7266,
+      "step": 4996
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 13.911056518554688,
+      "learning_rate": 9.984965420467075e-06,
+      "loss": 1.2805,
+      "step": 4997
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 10.715677261352539,
+      "learning_rate": 9.982960809862685e-06,
+      "loss": 2.3035,
+      "step": 4998
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 23.649269104003906,
+      "learning_rate": 9.980956199258295e-06,
+      "loss": 1.6413,
+      "step": 4999
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 9.821858406066895,
+      "learning_rate": 9.978951588653905e-06,
+      "loss": 1.1268,
+      "step": 5000
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 11.555136680603027,
+      "learning_rate": 9.976946978049515e-06,
+      "loss": 1.2356,
+      "step": 5001
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 15.840431213378906,
+      "learning_rate": 9.974942367445125e-06,
+      "loss": 1.7111,
+      "step": 5002
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 13.110334396362305,
+      "learning_rate": 9.972937756840735e-06,
+      "loss": 1.3668,
+      "step": 5003
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 15.277511596679688,
+      "learning_rate": 9.970933146236345e-06,
+      "loss": 1.524,
+      "step": 5004
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 22.332592010498047,
+      "learning_rate": 9.968928535631954e-06,
+      "loss": 1.8393,
+      "step": 5005
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 10.41042709350586,
+      "learning_rate": 9.966923925027564e-06,
+      "loss": 0.9121,
+      "step": 5006
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 17.307802200317383,
+      "learning_rate": 9.964919314423174e-06,
+      "loss": 1.2376,
+      "step": 5007
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 28.895647048950195,
+      "learning_rate": 9.962914703818784e-06,
+      "loss": 2.9159,
+      "step": 5008
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 54.32232666015625,
+      "learning_rate": 9.960910093214394e-06,
+      "loss": 2.1614,
+      "step": 5009
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 13.556760787963867,
+      "learning_rate": 9.958905482610004e-06,
+      "loss": 0.914,
+      "step": 5010
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 16.66999053955078,
+      "learning_rate": 9.956900872005614e-06,
+      "loss": 1.303,
+      "step": 5011
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 23.729116439819336,
+      "learning_rate": 9.954896261401223e-06,
+      "loss": 1.1319,
+      "step": 5012
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 14.170336723327637,
+      "learning_rate": 9.952891650796833e-06,
+      "loss": 1.2095,
+      "step": 5013
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 8.873191833496094,
+      "learning_rate": 9.950887040192443e-06,
+      "loss": 1.0473,
+      "step": 5014
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 19.656044006347656,
+      "learning_rate": 9.948882429588053e-06,
+      "loss": 1.9538,
+      "step": 5015
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 7.630486011505127,
+      "learning_rate": 9.946877818983663e-06,
+      "loss": 1.2515,
+      "step": 5016
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 15.626066207885742,
+      "learning_rate": 9.944873208379273e-06,
+      "loss": 1.7024,
+      "step": 5017
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 11.820615768432617,
+      "learning_rate": 9.942868597774883e-06,
+      "loss": 1.5467,
+      "step": 5018
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 26.351261138916016,
+      "learning_rate": 9.940863987170493e-06,
+      "loss": 1.0137,
+      "step": 5019
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 15.984896659851074,
+      "learning_rate": 9.938859376566102e-06,
+      "loss": 1.2989,
+      "step": 5020
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 23.20572853088379,
+      "learning_rate": 9.936854765961713e-06,
+      "loss": 1.8128,
+      "step": 5021
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 12.120548248291016,
+      "learning_rate": 9.934850155357323e-06,
+      "loss": 1.295,
+      "step": 5022
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 14.345947265625,
+      "learning_rate": 9.932845544752932e-06,
+      "loss": 1.0609,
+      "step": 5023
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 10.384604454040527,
+      "learning_rate": 9.930840934148542e-06,
+      "loss": 0.8389,
+      "step": 5024
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 50.97624588012695,
+      "learning_rate": 9.928836323544152e-06,
+      "loss": 2.2094,
+      "step": 5025
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 15.930320739746094,
+      "learning_rate": 9.926831712939762e-06,
+      "loss": 1.0719,
+      "step": 5026
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 15.929656982421875,
+      "learning_rate": 9.924827102335372e-06,
+      "loss": 1.4516,
+      "step": 5027
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 20.207595825195312,
+      "learning_rate": 9.922822491730982e-06,
+      "loss": 1.122,
+      "step": 5028
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 10.85169506072998,
+      "learning_rate": 9.920817881126592e-06,
+      "loss": 1.3796,
+      "step": 5029
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 40.899234771728516,
+      "learning_rate": 9.918813270522202e-06,
+      "loss": 2.79,
+      "step": 5030
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 77.47624206542969,
+      "learning_rate": 9.91680865991781e-06,
+      "loss": 2.6326,
+      "step": 5031
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 11.343374252319336,
+      "learning_rate": 9.914804049313421e-06,
+      "loss": 1.2866,
+      "step": 5032
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 35.656246185302734,
+      "learning_rate": 9.912799438709033e-06,
+      "loss": 2.6495,
+      "step": 5033
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 17.106098175048828,
+      "learning_rate": 9.910794828104641e-06,
+      "loss": 1.4173,
+      "step": 5034
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 25.03550148010254,
+      "learning_rate": 9.908790217500251e-06,
+      "loss": 1.5695,
+      "step": 5035
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 14.62669563293457,
+      "learning_rate": 9.906785606895861e-06,
+      "loss": 1.6762,
+      "step": 5036
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 98.68997192382812,
+      "learning_rate": 9.904780996291471e-06,
+      "loss": 3.2678,
+      "step": 5037
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 10.68436050415039,
+      "learning_rate": 9.902776385687081e-06,
+      "loss": 1.6819,
+      "step": 5038
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 17.10354232788086,
+      "learning_rate": 9.900771775082692e-06,
+      "loss": 1.5759,
+      "step": 5039
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 24.183143615722656,
+      "learning_rate": 9.898767164478302e-06,
+      "loss": 1.51,
+      "step": 5040
+    },
+    {
+      "epoch": 1.52,
+      "eval_loss": 0.1983981430530548,
+      "eval_runtime": 43.7634,
+      "eval_samples_per_second": 33.795,
+      "eval_steps_per_second": 33.795,
+      "step": 5040
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 11.218318939208984,
+      "learning_rate": 9.896762553873912e-06,
+      "loss": 2.6052,
+      "step": 5041
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 28.35023307800293,
+      "learning_rate": 9.89475794326952e-06,
+      "loss": 1.5494,
+      "step": 5042
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 39.44877624511719,
+      "learning_rate": 9.89275333266513e-06,
+      "loss": 2.2204,
+      "step": 5043
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 16.99179458618164,
+      "learning_rate": 9.89074872206074e-06,
+      "loss": 1.9028,
+      "step": 5044
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 22.564708709716797,
+      "learning_rate": 9.88874411145635e-06,
+      "loss": 2.0602,
+      "step": 5045
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 10.778221130371094,
+      "learning_rate": 9.88673950085196e-06,
+      "loss": 1.071,
+      "step": 5046
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 25.14896583557129,
+      "learning_rate": 9.88473489024757e-06,
+      "loss": 1.4653,
+      "step": 5047
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 24.445661544799805,
+      "learning_rate": 9.88273027964318e-06,
+      "loss": 1.2668,
+      "step": 5048
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 19.382781982421875,
+      "learning_rate": 9.88072566903879e-06,
+      "loss": 1.6534,
+      "step": 5049
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 48.97084426879883,
+      "learning_rate": 9.878721058434399e-06,
+      "loss": 1.6285,
+      "step": 5050
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 23.401174545288086,
+      "learning_rate": 9.87671644783001e-06,
+      "loss": 2.24,
+      "step": 5051
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 7.404333591461182,
+      "learning_rate": 9.874711837225621e-06,
+      "loss": 0.8004,
+      "step": 5052
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 14.808720588684082,
+      "learning_rate": 9.87270722662123e-06,
+      "loss": 1.193,
+      "step": 5053
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 11.823633193969727,
+      "learning_rate": 9.87070261601684e-06,
+      "loss": 1.1512,
+      "step": 5054
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 20.415996551513672,
+      "learning_rate": 9.86869800541245e-06,
+      "loss": 1.7406,
+      "step": 5055
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 584.0230712890625,
+      "learning_rate": 9.86669339480806e-06,
+      "loss": 1.3769,
+      "step": 5056
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 14.067730903625488,
+      "learning_rate": 9.864688784203668e-06,
+      "loss": 1.1319,
+      "step": 5057
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 27.009614944458008,
+      "learning_rate": 9.86268417359928e-06,
+      "loss": 1.6301,
+      "step": 5058
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 12.77624225616455,
+      "learning_rate": 9.86067956299489e-06,
+      "loss": 1.4832,
+      "step": 5059
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 14.48194408416748,
+      "learning_rate": 9.858674952390498e-06,
+      "loss": 1.3063,
+      "step": 5060
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 23.552724838256836,
+      "learning_rate": 9.856670341786108e-06,
+      "loss": 1.726,
+      "step": 5061
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 35.44766616821289,
+      "learning_rate": 9.854665731181718e-06,
+      "loss": 1.7173,
+      "step": 5062
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 17.609161376953125,
+      "learning_rate": 9.852661120577328e-06,
+      "loss": 1.2193,
+      "step": 5063
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 13.460494041442871,
+      "learning_rate": 9.850656509972939e-06,
+      "loss": 1.2271,
+      "step": 5064
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 19.18136215209961,
+      "learning_rate": 9.848651899368549e-06,
+      "loss": 1.0003,
+      "step": 5065
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 33.11497497558594,
+      "learning_rate": 9.846647288764159e-06,
+      "loss": 1.9836,
+      "step": 5066
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 17.282258987426758,
+      "learning_rate": 9.844642678159769e-06,
+      "loss": 1.7075,
+      "step": 5067
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 12.306963920593262,
+      "learning_rate": 9.842638067555377e-06,
+      "loss": 0.9739,
+      "step": 5068
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 18.901031494140625,
+      "learning_rate": 9.840633456950987e-06,
+      "loss": 1.4667,
+      "step": 5069
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 22.379085540771484,
+      "learning_rate": 9.838628846346599e-06,
+      "loss": 1.1675,
+      "step": 5070
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 14.647932052612305,
+      "learning_rate": 9.836624235742207e-06,
+      "loss": 1.809,
+      "step": 5071
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 8.400578498840332,
+      "learning_rate": 9.834619625137818e-06,
+      "loss": 1.8459,
+      "step": 5072
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 14.352058410644531,
+      "learning_rate": 9.832615014533428e-06,
+      "loss": 1.6325,
+      "step": 5073
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 38.508724212646484,
+      "learning_rate": 9.830610403929038e-06,
+      "loss": 1.4109,
+      "step": 5074
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 24.62671661376953,
+      "learning_rate": 9.828605793324648e-06,
+      "loss": 2.3143,
+      "step": 5075
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 32.014671325683594,
+      "learning_rate": 9.826601182720258e-06,
+      "loss": 2.2224,
+      "step": 5076
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 16.45624351501465,
+      "learning_rate": 9.824596572115868e-06,
+      "loss": 1.3855,
+      "step": 5077
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 24.989152908325195,
+      "learning_rate": 9.822591961511478e-06,
+      "loss": 1.8901,
+      "step": 5078
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 16.87116241455078,
+      "learning_rate": 9.820587350907086e-06,
+      "loss": 2.1003,
+      "step": 5079
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 19.88626480102539,
+      "learning_rate": 9.818582740302697e-06,
+      "loss": 0.9223,
+      "step": 5080
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 29.52408218383789,
+      "learning_rate": 9.816578129698307e-06,
+      "loss": 2.1227,
+      "step": 5081
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 6.786428451538086,
+      "learning_rate": 9.814573519093917e-06,
+      "loss": 1.1044,
+      "step": 5082
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 11.187318801879883,
+      "learning_rate": 9.812568908489527e-06,
+      "loss": 1.0142,
+      "step": 5083
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 9.082967758178711,
+      "learning_rate": 9.810564297885137e-06,
+      "loss": 0.8634,
+      "step": 5084
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 38.15378189086914,
+      "learning_rate": 9.808559687280747e-06,
+      "loss": 1.4613,
+      "step": 5085
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 11.708313941955566,
+      "learning_rate": 9.806555076676357e-06,
+      "loss": 2.1042,
+      "step": 5086
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 15.158965110778809,
+      "learning_rate": 9.804550466071965e-06,
+      "loss": 1.2576,
+      "step": 5087
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 15.897801399230957,
+      "learning_rate": 9.802545855467577e-06,
+      "loss": 1.5028,
+      "step": 5088
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 58.225669860839844,
+      "learning_rate": 9.800541244863187e-06,
+      "loss": 1.0376,
+      "step": 5089
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 26.382015228271484,
+      "learning_rate": 9.798536634258796e-06,
+      "loss": 1.2807,
+      "step": 5090
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 11.506767272949219,
+      "learning_rate": 9.796532023654406e-06,
+      "loss": 1.4462,
+      "step": 5091
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 36.81036376953125,
+      "learning_rate": 9.794527413050016e-06,
+      "loss": 2.7945,
+      "step": 5092
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 13.376030921936035,
+      "learning_rate": 9.792522802445626e-06,
+      "loss": 1.562,
+      "step": 5093
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 17.786754608154297,
+      "learning_rate": 9.790518191841234e-06,
+      "loss": 2.5748,
+      "step": 5094
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 22.214862823486328,
+      "learning_rate": 9.788513581236846e-06,
+      "loss": 1.8133,
+      "step": 5095
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 7.965408802032471,
+      "learning_rate": 9.786508970632456e-06,
+      "loss": 1.2572,
+      "step": 5096
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 16.189109802246094,
+      "learning_rate": 9.784504360028065e-06,
+      "loss": 1.1491,
+      "step": 5097
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 21.910438537597656,
+      "learning_rate": 9.782499749423675e-06,
+      "loss": 1.8276,
+      "step": 5098
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 29.36257553100586,
+      "learning_rate": 9.780495138819285e-06,
+      "loss": 1.69,
+      "step": 5099
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 15.409547805786133,
+      "learning_rate": 9.778490528214895e-06,
+      "loss": 1.83,
+      "step": 5100
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 12.414823532104492,
+      "learning_rate": 9.776485917610505e-06,
+      "loss": 1.3998,
+      "step": 5101
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 25.707353591918945,
+      "learning_rate": 9.774481307006115e-06,
+      "loss": 2.12,
+      "step": 5102
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 43.882991790771484,
+      "learning_rate": 9.772476696401725e-06,
+      "loss": 1.7967,
+      "step": 5103
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 12.999314308166504,
+      "learning_rate": 9.770472085797335e-06,
+      "loss": 1.1704,
+      "step": 5104
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 30.975318908691406,
+      "learning_rate": 9.768467475192944e-06,
+      "loss": 2.4516,
+      "step": 5105
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 10.727395057678223,
+      "learning_rate": 9.766462864588554e-06,
+      "loss": 0.5141,
+      "step": 5106
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 16.37921142578125,
+      "learning_rate": 9.764458253984165e-06,
+      "loss": 1.9043,
+      "step": 5107
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 11.777350425720215,
+      "learning_rate": 9.762453643379774e-06,
+      "loss": 1.3563,
+      "step": 5108
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 10.583935737609863,
+      "learning_rate": 9.760449032775384e-06,
+      "loss": 1.3274,
+      "step": 5109
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 24.881576538085938,
+      "learning_rate": 9.758444422170994e-06,
+      "loss": 1.0478,
+      "step": 5110
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 24.082473754882812,
+      "learning_rate": 9.756439811566604e-06,
+      "loss": 0.8794,
+      "step": 5111
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 15.496257781982422,
+      "learning_rate": 9.754435200962214e-06,
+      "loss": 0.8897,
+      "step": 5112
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 27.82330894470215,
+      "learning_rate": 9.752430590357824e-06,
+      "loss": 2.5503,
+      "step": 5113
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 14.794507026672363,
+      "learning_rate": 9.750425979753434e-06,
+      "loss": 2.2607,
+      "step": 5114
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 25.15207862854004,
+      "learning_rate": 9.748421369149044e-06,
+      "loss": 1.2937,
+      "step": 5115
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 28.609935760498047,
+      "learning_rate": 9.746416758544653e-06,
+      "loss": 1.8648,
+      "step": 5116
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 19.665058135986328,
+      "learning_rate": 9.744412147940263e-06,
+      "loss": 1.3485,
+      "step": 5117
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 16.362789154052734,
+      "learning_rate": 9.742407537335873e-06,
+      "loss": 1.6568,
+      "step": 5118
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 43.66654586791992,
+      "learning_rate": 9.740402926731483e-06,
+      "loss": 1.5276,
+      "step": 5119
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 11.123966217041016,
+      "learning_rate": 9.738398316127093e-06,
+      "loss": 1.4542,
+      "step": 5120
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 10.916337966918945,
+      "learning_rate": 9.736393705522703e-06,
+      "loss": 1.7105,
+      "step": 5121
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 38.372467041015625,
+      "learning_rate": 9.734389094918313e-06,
+      "loss": 2.5102,
+      "step": 5122
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 12.8078031539917,
+      "learning_rate": 9.732384484313923e-06,
+      "loss": 1.573,
+      "step": 5123
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 22.485212326049805,
+      "learning_rate": 9.730379873709532e-06,
+      "loss": 1.986,
+      "step": 5124
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 16.489240646362305,
+      "learning_rate": 9.728375263105144e-06,
+      "loss": 1.5365,
+      "step": 5125
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 48.157371520996094,
+      "learning_rate": 9.726370652500754e-06,
+      "loss": 2.6245,
+      "step": 5126
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 19.086997985839844,
+      "learning_rate": 9.724366041896362e-06,
+      "loss": 2.8709,
+      "step": 5127
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 16.59282112121582,
+      "learning_rate": 9.722361431291972e-06,
+      "loss": 1.6568,
+      "step": 5128
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 77.64139556884766,
+      "learning_rate": 9.720356820687582e-06,
+      "loss": 2.4053,
+      "step": 5129
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 32.31578063964844,
+      "learning_rate": 9.718352210083192e-06,
+      "loss": 2.4112,
+      "step": 5130
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 16.772424697875977,
+      "learning_rate": 9.716347599478802e-06,
+      "loss": 1.8791,
+      "step": 5131
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 18.102991104125977,
+      "learning_rate": 9.714342988874412e-06,
+      "loss": 1.3216,
+      "step": 5132
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 11.130638122558594,
+      "learning_rate": 9.712338378270023e-06,
+      "loss": 1.4781,
+      "step": 5133
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 23.45066261291504,
+      "learning_rate": 9.710333767665631e-06,
+      "loss": 2.065,
+      "step": 5134
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 24.560604095458984,
+      "learning_rate": 9.708329157061241e-06,
+      "loss": 1.383,
+      "step": 5135
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 12.12047290802002,
+      "learning_rate": 9.706324546456851e-06,
+      "loss": 1.5988,
+      "step": 5136
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 10.139289855957031,
+      "learning_rate": 9.704319935852461e-06,
+      "loss": 0.8568,
+      "step": 5137
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 10.667194366455078,
+      "learning_rate": 9.702315325248071e-06,
+      "loss": 1.2382,
+      "step": 5138
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 14.842784881591797,
+      "learning_rate": 9.700310714643681e-06,
+      "loss": 1.474,
+      "step": 5139
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 28.56570816040039,
+      "learning_rate": 9.698306104039291e-06,
+      "loss": 1.9052,
+      "step": 5140
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 14.900588035583496,
+      "learning_rate": 9.696301493434901e-06,
+      "loss": 1.2868,
+      "step": 5141
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 11.330522537231445,
+      "learning_rate": 9.69429688283051e-06,
+      "loss": 1.1221,
+      "step": 5142
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 10.978798866271973,
+      "learning_rate": 9.69229227222612e-06,
+      "loss": 1.2924,
+      "step": 5143
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 10.9464111328125,
+      "learning_rate": 9.690287661621732e-06,
+      "loss": 0.9553,
+      "step": 5144
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 19.434829711914062,
+      "learning_rate": 9.68828305101734e-06,
+      "loss": 1.6625,
+      "step": 5145
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 13.196100234985352,
+      "learning_rate": 9.68627844041295e-06,
+      "loss": 1.5143,
+      "step": 5146
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 13.529244422912598,
+      "learning_rate": 9.68427382980856e-06,
+      "loss": 1.6467,
+      "step": 5147
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 37.84720993041992,
+      "learning_rate": 9.68226921920417e-06,
+      "loss": 3.0862,
+      "step": 5148
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 11.348771095275879,
+      "learning_rate": 9.68026460859978e-06,
+      "loss": 1.3557,
+      "step": 5149
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 11.032979965209961,
+      "learning_rate": 9.67825999799539e-06,
+      "loss": 0.9113,
+      "step": 5150
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 15.761959075927734,
+      "learning_rate": 9.676255387391e-06,
+      "loss": 1.7351,
+      "step": 5151
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 53.275428771972656,
+      "learning_rate": 9.67425077678661e-06,
+      "loss": 2.9133,
+      "step": 5152
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 18.882171630859375,
+      "learning_rate": 9.672246166182219e-06,
+      "loss": 2.1493,
+      "step": 5153
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 13.768794059753418,
+      "learning_rate": 9.67024155557783e-06,
+      "loss": 1.1085,
+      "step": 5154
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 15.78554916381836,
+      "learning_rate": 9.66823694497344e-06,
+      "loss": 1.7214,
+      "step": 5155
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 18.663970947265625,
+      "learning_rate": 9.66623233436905e-06,
+      "loss": 1.5341,
+      "step": 5156
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 91.2535629272461,
+      "learning_rate": 9.66422772376466e-06,
+      "loss": 2.4091,
+      "step": 5157
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 30.0625057220459,
+      "learning_rate": 9.66222311316027e-06,
+      "loss": 1.2433,
+      "step": 5158
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 10.793164253234863,
+      "learning_rate": 9.66021850255588e-06,
+      "loss": 1.3247,
+      "step": 5159
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 14.865962028503418,
+      "learning_rate": 9.65821389195149e-06,
+      "loss": 0.7533,
+      "step": 5160
+    },
+    {
+      "epoch": 1.55,
+      "eval_loss": 0.20735611021518707,
+      "eval_runtime": 43.7679,
+      "eval_samples_per_second": 33.792,
+      "eval_steps_per_second": 33.792,
+      "step": 5160
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 18.602514266967773,
+      "learning_rate": 9.656209281347098e-06,
+      "loss": 1.5636,
+      "step": 5161
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 14.882292747497559,
+      "learning_rate": 9.65420467074271e-06,
+      "loss": 1.1038,
+      "step": 5162
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 56.73546600341797,
+      "learning_rate": 9.65220006013832e-06,
+      "loss": 1.9049,
+      "step": 5163
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 13.342903137207031,
+      "learning_rate": 9.650195449533928e-06,
+      "loss": 1.6425,
+      "step": 5164
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 13.89185905456543,
+      "learning_rate": 9.648190838929538e-06,
+      "loss": 1.0887,
+      "step": 5165
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 72.23283386230469,
+      "learning_rate": 9.646186228325149e-06,
+      "loss": 2.1655,
+      "step": 5166
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 17.974084854125977,
+      "learning_rate": 9.644181617720759e-06,
+      "loss": 2.2678,
+      "step": 5167
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 69.75627136230469,
+      "learning_rate": 9.642177007116369e-06,
+      "loss": 1.7682,
+      "step": 5168
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 11.242698669433594,
+      "learning_rate": 9.640172396511979e-06,
+      "loss": 1.3168,
+      "step": 5169
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 66.39893341064453,
+      "learning_rate": 9.638167785907589e-06,
+      "loss": 2.6081,
+      "step": 5170
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 9.396846771240234,
+      "learning_rate": 9.636163175303199e-06,
+      "loss": 0.8918,
+      "step": 5171
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 30.329811096191406,
+      "learning_rate": 9.634158564698807e-06,
+      "loss": 2.5263,
+      "step": 5172
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 8.505033493041992,
+      "learning_rate": 9.632153954094417e-06,
+      "loss": 0.9187,
+      "step": 5173
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 15.874688148498535,
+      "learning_rate": 9.63014934349003e-06,
+      "loss": 1.2034,
+      "step": 5174
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 12.436075210571289,
+      "learning_rate": 9.628144732885638e-06,
+      "loss": 1.2702,
+      "step": 5175
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 16.405839920043945,
+      "learning_rate": 9.626140122281248e-06,
+      "loss": 1.2595,
+      "step": 5176
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 6.489515781402588,
+      "learning_rate": 9.624135511676858e-06,
+      "loss": 0.699,
+      "step": 5177
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 15.752748489379883,
+      "learning_rate": 9.622130901072468e-06,
+      "loss": 1.8403,
+      "step": 5178
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 21.755474090576172,
+      "learning_rate": 9.620126290468076e-06,
+      "loss": 1.9319,
+      "step": 5179
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 14.760236740112305,
+      "learning_rate": 9.618121679863686e-06,
+      "loss": 1.3553,
+      "step": 5180
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 20.644493103027344,
+      "learning_rate": 9.616117069259298e-06,
+      "loss": 1.2051,
+      "step": 5181
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 18.111047744750977,
+      "learning_rate": 9.614112458654906e-06,
+      "loss": 1.1888,
+      "step": 5182
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 9.958390235900879,
+      "learning_rate": 9.612107848050517e-06,
+      "loss": 1.2644,
+      "step": 5183
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 16.58241081237793,
+      "learning_rate": 9.610103237446127e-06,
+      "loss": 0.8966,
+      "step": 5184
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 27.23797607421875,
+      "learning_rate": 9.608098626841737e-06,
+      "loss": 2.5713,
+      "step": 5185
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 28.486438751220703,
+      "learning_rate": 9.606094016237347e-06,
+      "loss": 2.6203,
+      "step": 5186
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 15.266054153442383,
+      "learning_rate": 9.604089405632957e-06,
+      "loss": 1.8284,
+      "step": 5187
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 27.08075714111328,
+      "learning_rate": 9.602084795028567e-06,
+      "loss": 1.749,
+      "step": 5188
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 15.239933013916016,
+      "learning_rate": 9.600080184424177e-06,
+      "loss": 1.4726,
+      "step": 5189
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 29.341596603393555,
+      "learning_rate": 9.598075573819785e-06,
+      "loss": 1.697,
+      "step": 5190
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 17.9268856048584,
+      "learning_rate": 9.596070963215396e-06,
+      "loss": 1.7262,
+      "step": 5191
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 18.277809143066406,
+      "learning_rate": 9.594066352611006e-06,
+      "loss": 2.4779,
+      "step": 5192
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 24.198942184448242,
+      "learning_rate": 9.592061742006616e-06,
+      "loss": 1.2228,
+      "step": 5193
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 10.138649940490723,
+      "learning_rate": 9.590057131402226e-06,
+      "loss": 2.3378,
+      "step": 5194
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 15.209259986877441,
+      "learning_rate": 9.588052520797836e-06,
+      "loss": 1.8108,
+      "step": 5195
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 22.023107528686523,
+      "learning_rate": 9.586047910193446e-06,
+      "loss": 1.5428,
+      "step": 5196
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 18.7374210357666,
+      "learning_rate": 9.584043299589056e-06,
+      "loss": 1.6044,
+      "step": 5197
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 21.584501266479492,
+      "learning_rate": 9.582038688984664e-06,
+      "loss": 2.0156,
+      "step": 5198
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 17.4039363861084,
+      "learning_rate": 9.580034078380276e-06,
+      "loss": 1.3307,
+      "step": 5199
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 10.785429000854492,
+      "learning_rate": 9.578029467775886e-06,
+      "loss": 1.2009,
+      "step": 5200
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 13.932561874389648,
+      "learning_rate": 9.576024857171495e-06,
+      "loss": 1.6818,
+      "step": 5201
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 13.384397506713867,
+      "learning_rate": 9.574020246567105e-06,
+      "loss": 1.7232,
+      "step": 5202
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 32.59089279174805,
+      "learning_rate": 9.572015635962715e-06,
+      "loss": 2.1975,
+      "step": 5203
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 19.552274703979492,
+      "learning_rate": 9.570011025358325e-06,
+      "loss": 2.2128,
+      "step": 5204
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 15.523249626159668,
+      "learning_rate": 9.568006414753935e-06,
+      "loss": 1.1143,
+      "step": 5205
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 35.84797286987305,
+      "learning_rate": 9.566001804149545e-06,
+      "loss": 2.5374,
+      "step": 5206
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 14.734626770019531,
+      "learning_rate": 9.563997193545155e-06,
+      "loss": 1.2287,
+      "step": 5207
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 13.902031898498535,
+      "learning_rate": 9.561992582940765e-06,
+      "loss": 1.4145,
+      "step": 5208
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 9.233687400817871,
+      "learning_rate": 9.559987972336374e-06,
+      "loss": 1.2157,
+      "step": 5209
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 47.99122619628906,
+      "learning_rate": 9.557983361731984e-06,
+      "loss": 1.6352,
+      "step": 5210
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 18.855663299560547,
+      "learning_rate": 9.555978751127596e-06,
+      "loss": 2.1963,
+      "step": 5211
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 25.103059768676758,
+      "learning_rate": 9.553974140523204e-06,
+      "loss": 2.7753,
+      "step": 5212
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 19.241830825805664,
+      "learning_rate": 9.551969529918814e-06,
+      "loss": 1.5551,
+      "step": 5213
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 13.835005760192871,
+      "learning_rate": 9.549964919314424e-06,
+      "loss": 1.4738,
+      "step": 5214
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 8.090837478637695,
+      "learning_rate": 9.547960308710034e-06,
+      "loss": 1.7791,
+      "step": 5215
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 22.887216567993164,
+      "learning_rate": 9.545955698105643e-06,
+      "loss": 1.4865,
+      "step": 5216
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 18.71986961364746,
+      "learning_rate": 9.543951087501254e-06,
+      "loss": 1.7043,
+      "step": 5217
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 9.452502250671387,
+      "learning_rate": 9.541946476896864e-06,
+      "loss": 1.9759,
+      "step": 5218
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 18.228004455566406,
+      "learning_rate": 9.539941866292473e-06,
+      "loss": 1.4368,
+      "step": 5219
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 27.472970962524414,
+      "learning_rate": 9.537937255688083e-06,
+      "loss": 1.5878,
+      "step": 5220
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 34.94017028808594,
+      "learning_rate": 9.535932645083693e-06,
+      "loss": 1.7284,
+      "step": 5221
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 19.611160278320312,
+      "learning_rate": 9.533928034479303e-06,
+      "loss": 1.9294,
+      "step": 5222
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 181.42205810546875,
+      "learning_rate": 9.531923423874913e-06,
+      "loss": 1.5287,
+      "step": 5223
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 27.391780853271484,
+      "learning_rate": 9.529918813270523e-06,
+      "loss": 1.0979,
+      "step": 5224
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 19.48676872253418,
+      "learning_rate": 9.527914202666133e-06,
+      "loss": 1.9428,
+      "step": 5225
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 60.94241714477539,
+      "learning_rate": 9.525909592061743e-06,
+      "loss": 2.5457,
+      "step": 5226
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 39.271629333496094,
+      "learning_rate": 9.523904981457352e-06,
+      "loss": 1.834,
+      "step": 5227
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 7.752985000610352,
+      "learning_rate": 9.521900370852962e-06,
+      "loss": 1.3035,
+      "step": 5228
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 11.235089302062988,
+      "learning_rate": 9.519895760248572e-06,
+      "loss": 0.659,
+      "step": 5229
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 22.830162048339844,
+      "learning_rate": 9.517891149644182e-06,
+      "loss": 2.1836,
+      "step": 5230
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 40.63211441040039,
+      "learning_rate": 9.515886539039792e-06,
+      "loss": 2.0237,
+      "step": 5231
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 15.20790958404541,
+      "learning_rate": 9.513881928435402e-06,
+      "loss": 0.9827,
+      "step": 5232
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 24.22981071472168,
+      "learning_rate": 9.511877317831012e-06,
+      "loss": 1.6325,
+      "step": 5233
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 69.4743423461914,
+      "learning_rate": 9.509872707226622e-06,
+      "loss": 2.8366,
+      "step": 5234
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 16.03273582458496,
+      "learning_rate": 9.50786809662223e-06,
+      "loss": 1.8027,
+      "step": 5235
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 27.365671157836914,
+      "learning_rate": 9.505863486017843e-06,
+      "loss": 1.7576,
+      "step": 5236
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 24.969091415405273,
+      "learning_rate": 9.503858875413453e-06,
+      "loss": 2.0772,
+      "step": 5237
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 17.248416900634766,
+      "learning_rate": 9.501854264809061e-06,
+      "loss": 1.3726,
+      "step": 5238
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 15.464237213134766,
+      "learning_rate": 9.499849654204671e-06,
+      "loss": 1.4722,
+      "step": 5239
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 24.053791046142578,
+      "learning_rate": 9.497845043600281e-06,
+      "loss": 1.4818,
+      "step": 5240
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 21.699386596679688,
+      "learning_rate": 9.495840432995891e-06,
+      "loss": 2.1324,
+      "step": 5241
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 32.680335998535156,
+      "learning_rate": 9.493835822391501e-06,
+      "loss": 2.5121,
+      "step": 5242
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 9.323308944702148,
+      "learning_rate": 9.491831211787111e-06,
+      "loss": 1.7795,
+      "step": 5243
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 9.845460891723633,
+      "learning_rate": 9.489826601182722e-06,
+      "loss": 1.5512,
+      "step": 5244
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 12.808528900146484,
+      "learning_rate": 9.487821990578332e-06,
+      "loss": 1.6268,
+      "step": 5245
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 30.653648376464844,
+      "learning_rate": 9.48581737997394e-06,
+      "loss": 2.2517,
+      "step": 5246
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 37.43178939819336,
+      "learning_rate": 9.48381276936955e-06,
+      "loss": 1.6323,
+      "step": 5247
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 23.361366271972656,
+      "learning_rate": 9.481808158765162e-06,
+      "loss": 1.6171,
+      "step": 5248
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 43.43690872192383,
+      "learning_rate": 9.47980354816077e-06,
+      "loss": 1.8716,
+      "step": 5249
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 16.062522888183594,
+      "learning_rate": 9.47779893755638e-06,
+      "loss": 2.9676,
+      "step": 5250
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 34.8533935546875,
+      "learning_rate": 9.47579432695199e-06,
+      "loss": 1.6051,
+      "step": 5251
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 20.79990577697754,
+      "learning_rate": 9.4737897163476e-06,
+      "loss": 2.5814,
+      "step": 5252
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 94.41461181640625,
+      "learning_rate": 9.47178510574321e-06,
+      "loss": 3.8194,
+      "step": 5253
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 9.540099143981934,
+      "learning_rate": 9.46978049513882e-06,
+      "loss": 2.1172,
+      "step": 5254
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 18.652847290039062,
+      "learning_rate": 9.46777588453443e-06,
+      "loss": 1.3757,
+      "step": 5255
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 6.206528663635254,
+      "learning_rate": 9.465771273930041e-06,
+      "loss": 0.92,
+      "step": 5256
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 18.400524139404297,
+      "learning_rate": 9.46376666332565e-06,
+      "loss": 1.8978,
+      "step": 5257
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 22.74957275390625,
+      "learning_rate": 9.46176205272126e-06,
+      "loss": 1.8973,
+      "step": 5258
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 15.965211868286133,
+      "learning_rate": 9.45975744211687e-06,
+      "loss": 1.8043,
+      "step": 5259
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 19.99260139465332,
+      "learning_rate": 9.45775283151248e-06,
+      "loss": 1.6596,
+      "step": 5260
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 9.259965896606445,
+      "learning_rate": 9.45574822090809e-06,
+      "loss": 1.4668,
+      "step": 5261
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 49.869110107421875,
+      "learning_rate": 9.4537436103037e-06,
+      "loss": 3.4404,
+      "step": 5262
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 26.978092193603516,
+      "learning_rate": 9.45173899969931e-06,
+      "loss": 2.3785,
+      "step": 5263
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 37.760353088378906,
+      "learning_rate": 9.449734389094918e-06,
+      "loss": 1.1802,
+      "step": 5264
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 24.364702224731445,
+      "learning_rate": 9.447729778490528e-06,
+      "loss": 2.288,
+      "step": 5265
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 17.613224029541016,
+      "learning_rate": 9.445725167886138e-06,
+      "loss": 1.8934,
+      "step": 5266
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 25.8133544921875,
+      "learning_rate": 9.443720557281748e-06,
+      "loss": 1.6238,
+      "step": 5267
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 32.17334747314453,
+      "learning_rate": 9.441715946677358e-06,
+      "loss": 1.8157,
+      "step": 5268
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 16.442485809326172,
+      "learning_rate": 9.439711336072969e-06,
+      "loss": 1.1689,
+      "step": 5269
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 11.316102027893066,
+      "learning_rate": 9.437706725468579e-06,
+      "loss": 2.063,
+      "step": 5270
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 11.197502136230469,
+      "learning_rate": 9.435702114864189e-06,
+      "loss": 1.4149,
+      "step": 5271
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 15.829837799072266,
+      "learning_rate": 9.433697504259797e-06,
+      "loss": 1.6842,
+      "step": 5272
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 44.153804779052734,
+      "learning_rate": 9.431692893655409e-06,
+      "loss": 1.8206,
+      "step": 5273
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 8.531728744506836,
+      "learning_rate": 9.429688283051019e-06,
+      "loss": 0.9972,
+      "step": 5274
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 16.26192283630371,
+      "learning_rate": 9.427683672446627e-06,
+      "loss": 1.8538,
+      "step": 5275
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 10.455742835998535,
+      "learning_rate": 9.425679061842237e-06,
+      "loss": 1.0675,
+      "step": 5276
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 19.000938415527344,
+      "learning_rate": 9.423674451237848e-06,
+      "loss": 1.9213,
+      "step": 5277
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 17.04540252685547,
+      "learning_rate": 9.421669840633458e-06,
+      "loss": 1.4436,
+      "step": 5278
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 15.419578552246094,
+      "learning_rate": 9.419665230029068e-06,
+      "loss": 1.9087,
+      "step": 5279
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 18.85577392578125,
+      "learning_rate": 9.417660619424678e-06,
+      "loss": 2.2146,
+      "step": 5280
+    },
+    {
+      "epoch": 1.59,
+      "eval_loss": 0.1917702704668045,
+      "eval_runtime": 43.7028,
+      "eval_samples_per_second": 33.842,
+      "eval_steps_per_second": 33.842,
+      "step": 5280
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 24.73578453063965,
+      "learning_rate": 9.415656008820288e-06,
+      "loss": 2.0709,
+      "step": 5281
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 10.795376777648926,
+      "learning_rate": 9.413651398215898e-06,
+      "loss": 0.9204,
+      "step": 5282
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 14.635207176208496,
+      "learning_rate": 9.411646787611506e-06,
+      "loss": 1.7222,
+      "step": 5283
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 12.989263534545898,
+      "learning_rate": 9.409642177007116e-06,
+      "loss": 1.3394,
+      "step": 5284
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 12.602009773254395,
+      "learning_rate": 9.407637566402728e-06,
+      "loss": 1.7059,
+      "step": 5285
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 20.175142288208008,
+      "learning_rate": 9.405632955798337e-06,
+      "loss": 1.7654,
+      "step": 5286
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 18.197296142578125,
+      "learning_rate": 9.403628345193947e-06,
+      "loss": 1.7068,
+      "step": 5287
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 34.1553955078125,
+      "learning_rate": 9.401623734589557e-06,
+      "loss": 1.3868,
+      "step": 5288
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 27.550334930419922,
+      "learning_rate": 9.399619123985167e-06,
+      "loss": 1.2817,
+      "step": 5289
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 19.8583984375,
+      "learning_rate": 9.397614513380777e-06,
+      "loss": 1.9351,
+      "step": 5290
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 6.953277587890625,
+      "learning_rate": 9.395609902776387e-06,
+      "loss": 1.1993,
+      "step": 5291
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 15.303847312927246,
+      "learning_rate": 9.393605292171997e-06,
+      "loss": 1.6499,
+      "step": 5292
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 28.220230102539062,
+      "learning_rate": 9.391600681567607e-06,
+      "loss": 1.7453,
+      "step": 5293
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 40.280029296875,
+      "learning_rate": 9.389596070963216e-06,
+      "loss": 1.9758,
+      "step": 5294
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 15.754435539245605,
+      "learning_rate": 9.387591460358826e-06,
+      "loss": 1.1859,
+      "step": 5295
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 64.05276489257812,
+      "learning_rate": 9.385586849754436e-06,
+      "loss": 2.8985,
+      "step": 5296
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 12.505768775939941,
+      "learning_rate": 9.383582239150046e-06,
+      "loss": 1.7193,
+      "step": 5297
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 22.20297622680664,
+      "learning_rate": 9.381577628545656e-06,
+      "loss": 1.8241,
+      "step": 5298
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 17.095487594604492,
+      "learning_rate": 9.379573017941266e-06,
+      "loss": 1.7652,
+      "step": 5299
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 30.220666885375977,
+      "learning_rate": 9.377568407336876e-06,
+      "loss": 2.0137,
+      "step": 5300
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 16.760820388793945,
+      "learning_rate": 9.375563796732484e-06,
+      "loss": 1.4244,
+      "step": 5301
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 74.30367279052734,
+      "learning_rate": 9.373559186128095e-06,
+      "loss": 1.818,
+      "step": 5302
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 8.43071460723877,
+      "learning_rate": 9.371554575523705e-06,
+      "loss": 0.9715,
+      "step": 5303
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 19.98003387451172,
+      "learning_rate": 9.369549964919315e-06,
+      "loss": 1.5193,
+      "step": 5304
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 15.74586009979248,
+      "learning_rate": 9.367545354314925e-06,
+      "loss": 1.7498,
+      "step": 5305
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 16.798616409301758,
+      "learning_rate": 9.365540743710535e-06,
+      "loss": 0.9751,
+      "step": 5306
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 15.125486373901367,
+      "learning_rate": 9.363536133106145e-06,
+      "loss": 2.1381,
+      "step": 5307
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 26.87381362915039,
+      "learning_rate": 9.361531522501755e-06,
+      "loss": 1.6783,
+      "step": 5308
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 23.354263305664062,
+      "learning_rate": 9.359526911897363e-06,
+      "loss": 2.2363,
+      "step": 5309
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 11.7379789352417,
+      "learning_rate": 9.357522301292975e-06,
+      "loss": 1.6394,
+      "step": 5310
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 16.918331146240234,
+      "learning_rate": 9.355517690688585e-06,
+      "loss": 2.0194,
+      "step": 5311
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 10.654472351074219,
+      "learning_rate": 9.353513080084194e-06,
+      "loss": 2.1258,
+      "step": 5312
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 23.184656143188477,
+      "learning_rate": 9.351508469479804e-06,
+      "loss": 2.9148,
+      "step": 5313
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 10.799365997314453,
+      "learning_rate": 9.349503858875414e-06,
+      "loss": 0.941,
+      "step": 5314
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 12.17577838897705,
+      "learning_rate": 9.347499248271024e-06,
+      "loss": 0.8023,
+      "step": 5315
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 25.213882446289062,
+      "learning_rate": 9.345494637666634e-06,
+      "loss": 1.6508,
+      "step": 5316
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 11.12057113647461,
+      "learning_rate": 9.343490027062244e-06,
+      "loss": 1.589,
+      "step": 5317
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 27.033601760864258,
+      "learning_rate": 9.341485416457854e-06,
+      "loss": 1.8537,
+      "step": 5318
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 16.07554054260254,
+      "learning_rate": 9.339480805853464e-06,
+      "loss": 1.4078,
+      "step": 5319
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 16.774127960205078,
+      "learning_rate": 9.337476195249073e-06,
+      "loss": 1.2248,
+      "step": 5320
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 12.099883079528809,
+      "learning_rate": 9.335471584644683e-06,
+      "loss": 0.5273,
+      "step": 5321
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 14.563944816589355,
+      "learning_rate": 9.333466974040295e-06,
+      "loss": 1.2972,
+      "step": 5322
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 6.988048553466797,
+      "learning_rate": 9.331462363435903e-06,
+      "loss": 1.2299,
+      "step": 5323
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 15.345603942871094,
+      "learning_rate": 9.329457752831513e-06,
+      "loss": 1.8004,
+      "step": 5324
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 8.920988082885742,
+      "learning_rate": 9.327453142227123e-06,
+      "loss": 1.799,
+      "step": 5325
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 13.415757179260254,
+      "learning_rate": 9.325448531622733e-06,
+      "loss": 1.3893,
+      "step": 5326
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 14.817873001098633,
+      "learning_rate": 9.323443921018343e-06,
+      "loss": 2.0269,
+      "step": 5327
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 17.05929183959961,
+      "learning_rate": 9.321439310413953e-06,
+      "loss": 1.9584,
+      "step": 5328
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 10.261931419372559,
+      "learning_rate": 9.319434699809563e-06,
+      "loss": 1.5321,
+      "step": 5329
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 18.25226402282715,
+      "learning_rate": 9.317430089205174e-06,
+      "loss": 1.5914,
+      "step": 5330
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 13.496363639831543,
+      "learning_rate": 9.315425478600782e-06,
+      "loss": 1.6435,
+      "step": 5331
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 26.062244415283203,
+      "learning_rate": 9.313420867996392e-06,
+      "loss": 1.892,
+      "step": 5332
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 28.14862632751465,
+      "learning_rate": 9.311416257392002e-06,
+      "loss": 0.9718,
+      "step": 5333
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 9.369063377380371,
+      "learning_rate": 9.309411646787612e-06,
+      "loss": 1.6377,
+      "step": 5334
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 15.546137809753418,
+      "learning_rate": 9.307407036183222e-06,
+      "loss": 1.9211,
+      "step": 5335
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 32.82224655151367,
+      "learning_rate": 9.305402425578832e-06,
+      "loss": 1.2737,
+      "step": 5336
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 58.94491958618164,
+      "learning_rate": 9.303397814974442e-06,
+      "loss": 2.3802,
+      "step": 5337
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 17.832128524780273,
+      "learning_rate": 9.30139320437005e-06,
+      "loss": 2.0701,
+      "step": 5338
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 25.756505966186523,
+      "learning_rate": 9.299388593765661e-06,
+      "loss": 1.6935,
+      "step": 5339
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 12.220959663391113,
+      "learning_rate": 9.297383983161273e-06,
+      "loss": 2.3572,
+      "step": 5340
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 28.716758728027344,
+      "learning_rate": 9.295379372556883e-06,
+      "loss": 1.8092,
+      "step": 5341
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 17.98668670654297,
+      "learning_rate": 9.293374761952491e-06,
+      "loss": 1.61,
+      "step": 5342
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 15.28249454498291,
+      "learning_rate": 9.291370151348101e-06,
+      "loss": 2.0471,
+      "step": 5343
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 10.611080169677734,
+      "learning_rate": 9.289365540743711e-06,
+      "loss": 0.9532,
+      "step": 5344
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 18.82200813293457,
+      "learning_rate": 9.287360930139321e-06,
+      "loss": 1.315,
+      "step": 5345
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 19.420570373535156,
+      "learning_rate": 9.28535631953493e-06,
+      "loss": 1.0041,
+      "step": 5346
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 16.25090980529785,
+      "learning_rate": 9.283351708930542e-06,
+      "loss": 2.2406,
+      "step": 5347
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 36.150394439697266,
+      "learning_rate": 9.281347098326152e-06,
+      "loss": 2.0748,
+      "step": 5348
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 69.37711334228516,
+      "learning_rate": 9.27934248772176e-06,
+      "loss": 2.1891,
+      "step": 5349
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 9.345280647277832,
+      "learning_rate": 9.27733787711737e-06,
+      "loss": 1.2469,
+      "step": 5350
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 51.229942321777344,
+      "learning_rate": 9.27533326651298e-06,
+      "loss": 2.5097,
+      "step": 5351
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 33.678749084472656,
+      "learning_rate": 9.27332865590859e-06,
+      "loss": 2.5323,
+      "step": 5352
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 20.67754364013672,
+      "learning_rate": 9.2713240453042e-06,
+      "loss": 3.1563,
+      "step": 5353
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 19.53748321533203,
+      "learning_rate": 9.26931943469981e-06,
+      "loss": 1.1978,
+      "step": 5354
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 16.690845489501953,
+      "learning_rate": 9.26731482409542e-06,
+      "loss": 1.1068,
+      "step": 5355
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 19.17826271057129,
+      "learning_rate": 9.26531021349103e-06,
+      "loss": 1.8296,
+      "step": 5356
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 15.507776260375977,
+      "learning_rate": 9.263305602886639e-06,
+      "loss": 1.6507,
+      "step": 5357
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 129.35165405273438,
+      "learning_rate": 9.261300992282249e-06,
+      "loss": 1.632,
+      "step": 5358
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 9.657642364501953,
+      "learning_rate": 9.259296381677861e-06,
+      "loss": 1.315,
+      "step": 5359
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 18.087684631347656,
+      "learning_rate": 9.25729177107347e-06,
+      "loss": 1.2448,
+      "step": 5360
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 16.07610511779785,
+      "learning_rate": 9.25528716046908e-06,
+      "loss": 1.2188,
+      "step": 5361
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 14.000444412231445,
+      "learning_rate": 9.25328254986469e-06,
+      "loss": 1.9099,
+      "step": 5362
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 22.69326400756836,
+      "learning_rate": 9.2512779392603e-06,
+      "loss": 2.2418,
+      "step": 5363
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 41.782066345214844,
+      "learning_rate": 9.24927332865591e-06,
+      "loss": 1.8902,
+      "step": 5364
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 9.715392112731934,
+      "learning_rate": 9.24726871805152e-06,
+      "loss": 1.1494,
+      "step": 5365
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 13.382871627807617,
+      "learning_rate": 9.24526410744713e-06,
+      "loss": 1.6138,
+      "step": 5366
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 33.94945526123047,
+      "learning_rate": 9.24325949684274e-06,
+      "loss": 1.562,
+      "step": 5367
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 24.688888549804688,
+      "learning_rate": 9.241254886238348e-06,
+      "loss": 1.3487,
+      "step": 5368
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 19.392419815063477,
+      "learning_rate": 9.239250275633958e-06,
+      "loss": 1.7019,
+      "step": 5369
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 9.527002334594727,
+      "learning_rate": 9.237245665029568e-06,
+      "loss": 1.6555,
+      "step": 5370
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 11.718170166015625,
+      "learning_rate": 9.235241054425179e-06,
+      "loss": 1.143,
+      "step": 5371
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 14.71898078918457,
+      "learning_rate": 9.233236443820789e-06,
+      "loss": 1.3483,
+      "step": 5372
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 12.641324996948242,
+      "learning_rate": 9.231231833216399e-06,
+      "loss": 1.3701,
+      "step": 5373
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 26.643348693847656,
+      "learning_rate": 9.229227222612009e-06,
+      "loss": 1.8719,
+      "step": 5374
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 11.661134719848633,
+      "learning_rate": 9.227222612007619e-06,
+      "loss": 1.5889,
+      "step": 5375
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 14.931244850158691,
+      "learning_rate": 9.225218001403227e-06,
+      "loss": 1.7465,
+      "step": 5376
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 11.898720741271973,
+      "learning_rate": 9.223213390798839e-06,
+      "loss": 1.6241,
+      "step": 5377
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 16.589155197143555,
+      "learning_rate": 9.221208780194449e-06,
+      "loss": 1.4509,
+      "step": 5378
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 23.514259338378906,
+      "learning_rate": 9.219204169590058e-06,
+      "loss": 2.5493,
+      "step": 5379
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 17.52058982849121,
+      "learning_rate": 9.217199558985668e-06,
+      "loss": 1.2928,
+      "step": 5380
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 28.47098159790039,
+      "learning_rate": 9.215194948381278e-06,
+      "loss": 1.2744,
+      "step": 5381
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 15.93428897857666,
+      "learning_rate": 9.213190337776888e-06,
+      "loss": 1.5216,
+      "step": 5382
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 73.04627227783203,
+      "learning_rate": 9.211185727172498e-06,
+      "loss": 0.9938,
+      "step": 5383
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 26.94357681274414,
+      "learning_rate": 9.209181116568108e-06,
+      "loss": 1.0955,
+      "step": 5384
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 15.998095512390137,
+      "learning_rate": 9.207176505963718e-06,
+      "loss": 1.2675,
+      "step": 5385
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 6.676560401916504,
+      "learning_rate": 9.205171895359326e-06,
+      "loss": 0.8448,
+      "step": 5386
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 17.63335609436035,
+      "learning_rate": 9.203167284754936e-06,
+      "loss": 1.771,
+      "step": 5387
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 15.837037086486816,
+      "learning_rate": 9.201162674150547e-06,
+      "loss": 1.1493,
+      "step": 5388
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 15.462651252746582,
+      "learning_rate": 9.199158063546157e-06,
+      "loss": 1.4158,
+      "step": 5389
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 14.927474975585938,
+      "learning_rate": 9.197153452941767e-06,
+      "loss": 2.7956,
+      "step": 5390
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 25.812740325927734,
+      "learning_rate": 9.195148842337377e-06,
+      "loss": 2.2199,
+      "step": 5391
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 17.734699249267578,
+      "learning_rate": 9.193144231732987e-06,
+      "loss": 2.1131,
+      "step": 5392
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 15.593416213989258,
+      "learning_rate": 9.191139621128597e-06,
+      "loss": 1.3466,
+      "step": 5393
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 14.880626678466797,
+      "learning_rate": 9.189135010524205e-06,
+      "loss": 1.4476,
+      "step": 5394
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 17.891508102416992,
+      "learning_rate": 9.187130399919815e-06,
+      "loss": 1.5905,
+      "step": 5395
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 20.343963623046875,
+      "learning_rate": 9.185125789315427e-06,
+      "loss": 2.0189,
+      "step": 5396
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 19.60910415649414,
+      "learning_rate": 9.183121178711036e-06,
+      "loss": 2.392,
+      "step": 5397
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 19.910167694091797,
+      "learning_rate": 9.181116568106646e-06,
+      "loss": 1.713,
+      "step": 5398
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 13.35189151763916,
+      "learning_rate": 9.179111957502256e-06,
+      "loss": 1.2514,
+      "step": 5399
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 27.091657638549805,
+      "learning_rate": 9.177107346897866e-06,
+      "loss": 1.4003,
+      "step": 5400
+    },
+    {
+      "epoch": 1.62,
+      "eval_loss": 0.18608930706977844,
+      "eval_runtime": 43.4962,
+      "eval_samples_per_second": 34.003,
+      "eval_steps_per_second": 34.003,
+      "step": 5400
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 38.99711608886719,
+      "learning_rate": 9.175102736293476e-06,
+      "loss": 2.452,
+      "step": 5401
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 17.14046859741211,
+      "learning_rate": 9.173098125689086e-06,
+      "loss": 1.0333,
+      "step": 5402
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 14.545239448547363,
+      "learning_rate": 9.171093515084696e-06,
+      "loss": 1.1729,
+      "step": 5403
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 21.479055404663086,
+      "learning_rate": 9.169088904480306e-06,
+      "loss": 2.0552,
+      "step": 5404
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 52.615394592285156,
+      "learning_rate": 9.167084293875915e-06,
+      "loss": 1.6672,
+      "step": 5405
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 12.193754196166992,
+      "learning_rate": 9.165079683271525e-06,
+      "loss": 1.3845,
+      "step": 5406
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 10.990750312805176,
+      "learning_rate": 9.163075072667135e-06,
+      "loss": 1.3259,
+      "step": 5407
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 14.663102149963379,
+      "learning_rate": 9.161070462062745e-06,
+      "loss": 1.432,
+      "step": 5408
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 8.971065521240234,
+      "learning_rate": 9.159065851458355e-06,
+      "loss": 0.9139,
+      "step": 5409
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 12.251411437988281,
+      "learning_rate": 9.157061240853965e-06,
+      "loss": 1.2786,
+      "step": 5410
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 10.386260032653809,
+      "learning_rate": 9.155056630249575e-06,
+      "loss": 1.8492,
+      "step": 5411
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 25.967205047607422,
+      "learning_rate": 9.153052019645185e-06,
+      "loss": 1.8648,
+      "step": 5412
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 15.101587295532227,
+      "learning_rate": 9.151047409040794e-06,
+      "loss": 1.3361,
+      "step": 5413
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 21.25626564025879,
+      "learning_rate": 9.149042798436405e-06,
+      "loss": 1.624,
+      "step": 5414
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 12.912954330444336,
+      "learning_rate": 9.147038187832015e-06,
+      "loss": 1.2722,
+      "step": 5415
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 19.924049377441406,
+      "learning_rate": 9.145033577227624e-06,
+      "loss": 1.8115,
+      "step": 5416
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 10.341268539428711,
+      "learning_rate": 9.143028966623234e-06,
+      "loss": 1.1745,
+      "step": 5417
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 19.03986167907715,
+      "learning_rate": 9.141024356018844e-06,
+      "loss": 1.8332,
+      "step": 5418
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 10.237374305725098,
+      "learning_rate": 9.139019745414454e-06,
+      "loss": 1.1372,
+      "step": 5419
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 12.311965942382812,
+      "learning_rate": 9.137015134810064e-06,
+      "loss": 1.3524,
+      "step": 5420
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 17.118541717529297,
+      "learning_rate": 9.135010524205674e-06,
+      "loss": 1.4631,
+      "step": 5421
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 13.466850280761719,
+      "learning_rate": 9.133005913601284e-06,
+      "loss": 2.0476,
+      "step": 5422
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 22.777727127075195,
+      "learning_rate": 9.131001302996893e-06,
+      "loss": 1.6859,
+      "step": 5423
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 24.398303985595703,
+      "learning_rate": 9.128996692392503e-06,
+      "loss": 2.0839,
+      "step": 5424
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 22.156652450561523,
+      "learning_rate": 9.126992081788113e-06,
+      "loss": 1.3133,
+      "step": 5425
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 10.810490608215332,
+      "learning_rate": 9.124987471183725e-06,
+      "loss": 0.9388,
+      "step": 5426
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 17.22280502319336,
+      "learning_rate": 9.122982860579333e-06,
+      "loss": 2.081,
+      "step": 5427
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 11.593786239624023,
+      "learning_rate": 9.120978249974943e-06,
+      "loss": 1.6318,
+      "step": 5428
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 111.95779418945312,
+      "learning_rate": 9.118973639370553e-06,
+      "loss": 2.2157,
+      "step": 5429
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 10.570384979248047,
+      "learning_rate": 9.116969028766163e-06,
+      "loss": 1.8122,
+      "step": 5430
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 47.375701904296875,
+      "learning_rate": 9.114964418161772e-06,
+      "loss": 2.1275,
+      "step": 5431
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 15.159667015075684,
+      "learning_rate": 9.112959807557382e-06,
+      "loss": 2.2646,
+      "step": 5432
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 10.478726387023926,
+      "learning_rate": 9.110955196952994e-06,
+      "loss": 1.4377,
+      "step": 5433
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 27.897777557373047,
+      "learning_rate": 9.108950586348602e-06,
+      "loss": 2.1688,
+      "step": 5434
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 26.705507278442383,
+      "learning_rate": 9.106945975744212e-06,
+      "loss": 1.7822,
+      "step": 5435
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 23.994443893432617,
+      "learning_rate": 9.104941365139822e-06,
+      "loss": 2.0833,
+      "step": 5436
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 66.67716217041016,
+      "learning_rate": 9.102936754535432e-06,
+      "loss": 1.3532,
+      "step": 5437
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 30.51035499572754,
+      "learning_rate": 9.100932143931042e-06,
+      "loss": 2.0045,
+      "step": 5438
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 10.716803550720215,
+      "learning_rate": 9.098927533326652e-06,
+      "loss": 1.6747,
+      "step": 5439
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 13.023651123046875,
+      "learning_rate": 9.096922922722262e-06,
+      "loss": 0.7083,
+      "step": 5440
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 25.49052619934082,
+      "learning_rate": 9.094918312117873e-06,
+      "loss": 1.0234,
+      "step": 5441
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 33.57606887817383,
+      "learning_rate": 9.092913701513481e-06,
+      "loss": 2.2643,
+      "step": 5442
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 11.915518760681152,
+      "learning_rate": 9.090909090909091e-06,
+      "loss": 1.0967,
+      "step": 5443
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 15.158368110656738,
+      "learning_rate": 9.088904480304701e-06,
+      "loss": 1.534,
+      "step": 5444
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 11.216906547546387,
+      "learning_rate": 9.086899869700311e-06,
+      "loss": 1.4261,
+      "step": 5445
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 26.098888397216797,
+      "learning_rate": 9.084895259095921e-06,
+      "loss": 2.0426,
+      "step": 5446
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 13.751211166381836,
+      "learning_rate": 9.082890648491531e-06,
+      "loss": 1.5604,
+      "step": 5447
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 12.62765884399414,
+      "learning_rate": 9.080886037887141e-06,
+      "loss": 1.7308,
+      "step": 5448
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 20.71200942993164,
+      "learning_rate": 9.078881427282752e-06,
+      "loss": 0.9927,
+      "step": 5449
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 23.00042152404785,
+      "learning_rate": 9.07687681667836e-06,
+      "loss": 2.4304,
+      "step": 5450
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 28.70808982849121,
+      "learning_rate": 9.074872206073972e-06,
+      "loss": 1.5417,
+      "step": 5451
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 9.605061531066895,
+      "learning_rate": 9.072867595469582e-06,
+      "loss": 1.3111,
+      "step": 5452
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 7.0347490310668945,
+      "learning_rate": 9.07086298486519e-06,
+      "loss": 0.7902,
+      "step": 5453
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 7.641110420227051,
+      "learning_rate": 9.0688583742608e-06,
+      "loss": 0.5109,
+      "step": 5454
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 39.103248596191406,
+      "learning_rate": 9.06685376365641e-06,
+      "loss": 1.863,
+      "step": 5455
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 10.603463172912598,
+      "learning_rate": 9.06484915305202e-06,
+      "loss": 1.5076,
+      "step": 5456
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 53.90522766113281,
+      "learning_rate": 9.06284454244763e-06,
+      "loss": 3.2397,
+      "step": 5457
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 55.33717727661133,
+      "learning_rate": 9.06083993184324e-06,
+      "loss": 2.0653,
+      "step": 5458
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 18.05723762512207,
+      "learning_rate": 9.05883532123885e-06,
+      "loss": 1.8223,
+      "step": 5459
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 45.47984313964844,
+      "learning_rate": 9.05683071063446e-06,
+      "loss": 1.3857,
+      "step": 5460
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 64.19039916992188,
+      "learning_rate": 9.05482610003007e-06,
+      "loss": 1.9055,
+      "step": 5461
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 16.403478622436523,
+      "learning_rate": 9.05282148942568e-06,
+      "loss": 1.2061,
+      "step": 5462
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 15.951025009155273,
+      "learning_rate": 9.050816878821291e-06,
+      "loss": 2.1111,
+      "step": 5463
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 32.44859313964844,
+      "learning_rate": 9.0488122682169e-06,
+      "loss": 1.3111,
+      "step": 5464
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 22.92570686340332,
+      "learning_rate": 9.04680765761251e-06,
+      "loss": 1.4222,
+      "step": 5465
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 11.310336112976074,
+      "learning_rate": 9.04480304700812e-06,
+      "loss": 1.0347,
+      "step": 5466
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 48.00858688354492,
+      "learning_rate": 9.04279843640373e-06,
+      "loss": 1.5265,
+      "step": 5467
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 22.504119873046875,
+      "learning_rate": 9.040793825799338e-06,
+      "loss": 1.2273,
+      "step": 5468
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 23.860111236572266,
+      "learning_rate": 9.03878921519495e-06,
+      "loss": 1.4192,
+      "step": 5469
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 11.303675651550293,
+      "learning_rate": 9.03678460459056e-06,
+      "loss": 1.0257,
+      "step": 5470
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 12.571292877197266,
+      "learning_rate": 9.034779993986168e-06,
+      "loss": 1.4753,
+      "step": 5471
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 45.509620666503906,
+      "learning_rate": 9.032775383381778e-06,
+      "loss": 2.5814,
+      "step": 5472
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 11.296418190002441,
+      "learning_rate": 9.030770772777388e-06,
+      "loss": 0.8345,
+      "step": 5473
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 19.737348556518555,
+      "learning_rate": 9.028766162172999e-06,
+      "loss": 2.0533,
+      "step": 5474
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 32.28429412841797,
+      "learning_rate": 9.026761551568609e-06,
+      "loss": 1.0527,
+      "step": 5475
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 5.6402106285095215,
+      "learning_rate": 9.024756940964219e-06,
+      "loss": 0.7558,
+      "step": 5476
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 11.547632217407227,
+      "learning_rate": 9.022752330359829e-06,
+      "loss": 1.089,
+      "step": 5477
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 7.933753967285156,
+      "learning_rate": 9.020747719755439e-06,
+      "loss": 1.1976,
+      "step": 5478
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 25.738367080688477,
+      "learning_rate": 9.018743109151047e-06,
+      "loss": 2.3856,
+      "step": 5479
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 14.972805976867676,
+      "learning_rate": 9.016738498546657e-06,
+      "loss": 2.7914,
+      "step": 5480
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 23.992773056030273,
+      "learning_rate": 9.014733887942267e-06,
+      "loss": 2.568,
+      "step": 5481
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 12.318902015686035,
+      "learning_rate": 9.012729277337878e-06,
+      "loss": 1.2272,
+      "step": 5482
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 13.769071578979492,
+      "learning_rate": 9.010724666733488e-06,
+      "loss": 1.4336,
+      "step": 5483
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 29.345212936401367,
+      "learning_rate": 9.008720056129098e-06,
+      "loss": 1.8268,
+      "step": 5484
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 10.752758026123047,
+      "learning_rate": 9.006715445524708e-06,
+      "loss": 0.75,
+      "step": 5485
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 11.773067474365234,
+      "learning_rate": 9.004710834920318e-06,
+      "loss": 1.5268,
+      "step": 5486
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 20.230098724365234,
+      "learning_rate": 9.002706224315926e-06,
+      "loss": 1.2181,
+      "step": 5487
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 19.085262298583984,
+      "learning_rate": 9.000701613711538e-06,
+      "loss": 2.0583,
+      "step": 5488
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 15.87958812713623,
+      "learning_rate": 8.998697003107148e-06,
+      "loss": 1.2279,
+      "step": 5489
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 10.969903945922852,
+      "learning_rate": 8.996692392502757e-06,
+      "loss": 1.4018,
+      "step": 5490
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 16.4309024810791,
+      "learning_rate": 8.994687781898367e-06,
+      "loss": 1.7592,
+      "step": 5491
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 17.328758239746094,
+      "learning_rate": 8.992683171293977e-06,
+      "loss": 1.3829,
+      "step": 5492
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 23.75958824157715,
+      "learning_rate": 8.990678560689587e-06,
+      "loss": 1.7726,
+      "step": 5493
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 7.305182933807373,
+      "learning_rate": 8.988673950085197e-06,
+      "loss": 0.9196,
+      "step": 5494
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 23.182376861572266,
+      "learning_rate": 8.986669339480807e-06,
+      "loss": 2.1978,
+      "step": 5495
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 16.009490966796875,
+      "learning_rate": 8.984664728876417e-06,
+      "loss": 1.3805,
+      "step": 5496
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 25.598642349243164,
+      "learning_rate": 8.982660118272027e-06,
+      "loss": 1.7893,
+      "step": 5497
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 14.513004302978516,
+      "learning_rate": 8.980655507667636e-06,
+      "loss": 1.0957,
+      "step": 5498
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 33.75980758666992,
+      "learning_rate": 8.978650897063246e-06,
+      "loss": 1.9849,
+      "step": 5499
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 17.729082107543945,
+      "learning_rate": 8.976646286458857e-06,
+      "loss": 1.2108,
+      "step": 5500
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 15.00300121307373,
+      "learning_rate": 8.974641675854466e-06,
+      "loss": 1.9234,
+      "step": 5501
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 37.787193298339844,
+      "learning_rate": 8.972637065250076e-06,
+      "loss": 1.5762,
+      "step": 5502
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 18.156917572021484,
+      "learning_rate": 8.970632454645686e-06,
+      "loss": 1.2841,
+      "step": 5503
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 17.250186920166016,
+      "learning_rate": 8.968627844041296e-06,
+      "loss": 1.4742,
+      "step": 5504
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 27.10199737548828,
+      "learning_rate": 8.966623233436904e-06,
+      "loss": 1.9449,
+      "step": 5505
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 17.242645263671875,
+      "learning_rate": 8.964618622832516e-06,
+      "loss": 1.3392,
+      "step": 5506
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 12.527267456054688,
+      "learning_rate": 8.962614012228126e-06,
+      "loss": 1.559,
+      "step": 5507
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 12.160088539123535,
+      "learning_rate": 8.960609401623735e-06,
+      "loss": 1.2504,
+      "step": 5508
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 9.42102336883545,
+      "learning_rate": 8.958604791019345e-06,
+      "loss": 0.9488,
+      "step": 5509
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 16.88702964782715,
+      "learning_rate": 8.956600180414955e-06,
+      "loss": 1.5164,
+      "step": 5510
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 8.981550216674805,
+      "learning_rate": 8.954595569810565e-06,
+      "loss": 1.2683,
+      "step": 5511
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 14.901409149169922,
+      "learning_rate": 8.952590959206175e-06,
+      "loss": 1.3492,
+      "step": 5512
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 38.2581787109375,
+      "learning_rate": 8.950586348601785e-06,
+      "loss": 1.4274,
+      "step": 5513
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 11.266697883605957,
+      "learning_rate": 8.948581737997395e-06,
+      "loss": 1.2673,
+      "step": 5514
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 12.022287368774414,
+      "learning_rate": 8.946577127393005e-06,
+      "loss": 1.7178,
+      "step": 5515
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 18.94673728942871,
+      "learning_rate": 8.944572516788614e-06,
+      "loss": 1.3636,
+      "step": 5516
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 30.44729232788086,
+      "learning_rate": 8.942567906184224e-06,
+      "loss": 1.623,
+      "step": 5517
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 13.971253395080566,
+      "learning_rate": 8.940563295579834e-06,
+      "loss": 1.3185,
+      "step": 5518
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 10.269094467163086,
+      "learning_rate": 8.938558684975444e-06,
+      "loss": 1.2367,
+      "step": 5519
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 19.084022521972656,
+      "learning_rate": 8.936554074371054e-06,
+      "loss": 0.9847,
+      "step": 5520
+    },
+    {
+      "epoch": 1.66,
+      "eval_loss": 0.1914350539445877,
+      "eval_runtime": 43.4865,
+      "eval_samples_per_second": 34.011,
+      "eval_steps_per_second": 34.011,
+      "step": 5520
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 11.31253719329834,
+      "learning_rate": 8.934549463766664e-06,
+      "loss": 1.363,
+      "step": 5521
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 22.407804489135742,
+      "learning_rate": 8.932544853162274e-06,
+      "loss": 1.953,
+      "step": 5522
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 36.859561920166016,
+      "learning_rate": 8.930540242557884e-06,
+      "loss": 2.4004,
+      "step": 5523
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 28.108509063720703,
+      "learning_rate": 8.928535631953493e-06,
+      "loss": 1.4061,
+      "step": 5524
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 20.431781768798828,
+      "learning_rate": 8.926531021349104e-06,
+      "loss": 1.4469,
+      "step": 5525
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 20.850435256958008,
+      "learning_rate": 8.924526410744714e-06,
+      "loss": 1.789,
+      "step": 5526
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 11.911688804626465,
+      "learning_rate": 8.922521800140323e-06,
+      "loss": 2.0367,
+      "step": 5527
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 10.329282760620117,
+      "learning_rate": 8.920517189535933e-06,
+      "loss": 1.4678,
+      "step": 5528
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 12.712068557739258,
+      "learning_rate": 8.918512578931543e-06,
+      "loss": 1.6763,
+      "step": 5529
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 58.54856872558594,
+      "learning_rate": 8.916507968327153e-06,
+      "loss": 2.0221,
+      "step": 5530
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 13.266801834106445,
+      "learning_rate": 8.914503357722763e-06,
+      "loss": 1.7743,
+      "step": 5531
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 27.546993255615234,
+      "learning_rate": 8.912498747118373e-06,
+      "loss": 2.6575,
+      "step": 5532
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 11.523126602172852,
+      "learning_rate": 8.910494136513983e-06,
+      "loss": 1.1144,
+      "step": 5533
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 11.785201072692871,
+      "learning_rate": 8.908489525909593e-06,
+      "loss": 1.1003,
+      "step": 5534
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 14.82313346862793,
+      "learning_rate": 8.906484915305202e-06,
+      "loss": 1.8222,
+      "step": 5535
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 14.647685050964355,
+      "learning_rate": 8.904480304700812e-06,
+      "loss": 0.8845,
+      "step": 5536
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 8.299193382263184,
+      "learning_rate": 8.902475694096424e-06,
+      "loss": 1.3426,
+      "step": 5537
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 31.66522979736328,
+      "learning_rate": 8.900471083492032e-06,
+      "loss": 1.5596,
+      "step": 5538
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 10.633186340332031,
+      "learning_rate": 8.898466472887642e-06,
+      "loss": 0.8703,
+      "step": 5539
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 12.473995208740234,
+      "learning_rate": 8.896461862283252e-06,
+      "loss": 1.1009,
+      "step": 5540
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 16.58210563659668,
+      "learning_rate": 8.894457251678862e-06,
+      "loss": 0.7829,
+      "step": 5541
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 9.069561004638672,
+      "learning_rate": 8.89245264107447e-06,
+      "loss": 0.5217,
+      "step": 5542
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 14.045775413513184,
+      "learning_rate": 8.890448030470083e-06,
+      "loss": 1.2927,
+      "step": 5543
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 10.430898666381836,
+      "learning_rate": 8.888443419865693e-06,
+      "loss": 1.0182,
+      "step": 5544
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 28.892423629760742,
+      "learning_rate": 8.886438809261303e-06,
+      "loss": 2.1135,
+      "step": 5545
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 33.67451095581055,
+      "learning_rate": 8.884434198656911e-06,
+      "loss": 2.2505,
+      "step": 5546
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 22.55853271484375,
+      "learning_rate": 8.882429588052521e-06,
+      "loss": 1.275,
+      "step": 5547
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 11.746838569641113,
+      "learning_rate": 8.880424977448131e-06,
+      "loss": 1.0598,
+      "step": 5548
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 18.632312774658203,
+      "learning_rate": 8.878420366843741e-06,
+      "loss": 2.3075,
+      "step": 5549
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 23.912031173706055,
+      "learning_rate": 8.876415756239351e-06,
+      "loss": 1.5192,
+      "step": 5550
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 19.950305938720703,
+      "learning_rate": 8.874411145634962e-06,
+      "loss": 1.5378,
+      "step": 5551
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 15.393779754638672,
+      "learning_rate": 8.872406535030572e-06,
+      "loss": 1.4361,
+      "step": 5552
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 11.668610572814941,
+      "learning_rate": 8.87040192442618e-06,
+      "loss": 1.1552,
+      "step": 5553
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 11.19984245300293,
+      "learning_rate": 8.86839731382179e-06,
+      "loss": 1.3071,
+      "step": 5554
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 12.745835304260254,
+      "learning_rate": 8.8663927032174e-06,
+      "loss": 0.7712,
+      "step": 5555
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 32.49069595336914,
+      "learning_rate": 8.86438809261301e-06,
+      "loss": 1.9268,
+      "step": 5556
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 8.485462188720703,
+      "learning_rate": 8.86238348200862e-06,
+      "loss": 1.0959,
+      "step": 5557
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 14.367705345153809,
+      "learning_rate": 8.86037887140423e-06,
+      "loss": 1.0967,
+      "step": 5558
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 22.95987892150879,
+      "learning_rate": 8.85837426079984e-06,
+      "loss": 1.6232,
+      "step": 5559
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 15.968416213989258,
+      "learning_rate": 8.85636965019545e-06,
+      "loss": 1.2239,
+      "step": 5560
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 10.333566665649414,
+      "learning_rate": 8.854365039591059e-06,
+      "loss": 1.248,
+      "step": 5561
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 45.59640884399414,
+      "learning_rate": 8.85236042898667e-06,
+      "loss": 1.6009,
+      "step": 5562
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 28.932910919189453,
+      "learning_rate": 8.85035581838228e-06,
+      "loss": 1.88,
+      "step": 5563
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 12.517439842224121,
+      "learning_rate": 8.84835120777789e-06,
+      "loss": 1.8985,
+      "step": 5564
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 17.193233489990234,
+      "learning_rate": 8.8463465971735e-06,
+      "loss": 1.1834,
+      "step": 5565
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 26.31612205505371,
+      "learning_rate": 8.84434198656911e-06,
+      "loss": 2.4826,
+      "step": 5566
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 16.692577362060547,
+      "learning_rate": 8.84233737596472e-06,
+      "loss": 1.3551,
+      "step": 5567
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 19.141315460205078,
+      "learning_rate": 8.84033276536033e-06,
+      "loss": 1.5564,
+      "step": 5568
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 12.95341682434082,
+      "learning_rate": 8.83832815475594e-06,
+      "loss": 1.3491,
+      "step": 5569
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 11.91671085357666,
+      "learning_rate": 8.83632354415155e-06,
+      "loss": 1.4844,
+      "step": 5570
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 17.263015747070312,
+      "learning_rate": 8.83431893354716e-06,
+      "loss": 1.5285,
+      "step": 5571
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 10.586352348327637,
+      "learning_rate": 8.832314322942768e-06,
+      "loss": 1.4477,
+      "step": 5572
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 17.991832733154297,
+      "learning_rate": 8.830309712338378e-06,
+      "loss": 1.3879,
+      "step": 5573
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 19.859966278076172,
+      "learning_rate": 8.82830510173399e-06,
+      "loss": 1.8104,
+      "step": 5574
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 436.9852294921875,
+      "learning_rate": 8.826300491129598e-06,
+      "loss": 2.6175,
+      "step": 5575
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 12.303122520446777,
+      "learning_rate": 8.824295880525209e-06,
+      "loss": 0.9817,
+      "step": 5576
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 13.081160545349121,
+      "learning_rate": 8.822291269920819e-06,
+      "loss": 2.0876,
+      "step": 5577
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 9.616783142089844,
+      "learning_rate": 8.820286659316429e-06,
+      "loss": 1.7439,
+      "step": 5578
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 15.427207946777344,
+      "learning_rate": 8.818282048712039e-06,
+      "loss": 1.205,
+      "step": 5579
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 17.407861709594727,
+      "learning_rate": 8.816277438107649e-06,
+      "loss": 1.9737,
+      "step": 5580
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 30.608501434326172,
+      "learning_rate": 8.814272827503259e-06,
+      "loss": 1.4772,
+      "step": 5581
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 17.050554275512695,
+      "learning_rate": 8.812268216898869e-06,
+      "loss": 1.2003,
+      "step": 5582
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 21.977340698242188,
+      "learning_rate": 8.810263606294477e-06,
+      "loss": 2.4612,
+      "step": 5583
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 48.15430450439453,
+      "learning_rate": 8.808258995690088e-06,
+      "loss": 1.7486,
+      "step": 5584
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 15.752915382385254,
+      "learning_rate": 8.806254385085698e-06,
+      "loss": 1.6484,
+      "step": 5585
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 37.53594207763672,
+      "learning_rate": 8.804249774481308e-06,
+      "loss": 2.1715,
+      "step": 5586
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 10.964682579040527,
+      "learning_rate": 8.802245163876918e-06,
+      "loss": 1.4082,
+      "step": 5587
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 23.61287498474121,
+      "learning_rate": 8.800240553272528e-06,
+      "loss": 0.7415,
+      "step": 5588
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 15.340435981750488,
+      "learning_rate": 8.798235942668138e-06,
+      "loss": 1.5422,
+      "step": 5589
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 12.990081787109375,
+      "learning_rate": 8.796231332063746e-06,
+      "loss": 1.4462,
+      "step": 5590
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 10.152392387390137,
+      "learning_rate": 8.794226721459356e-06,
+      "loss": 0.9929,
+      "step": 5591
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 13.210673332214355,
+      "learning_rate": 8.792222110854968e-06,
+      "loss": 1.7244,
+      "step": 5592
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 13.106926918029785,
+      "learning_rate": 8.790217500250577e-06,
+      "loss": 1.1276,
+      "step": 5593
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 11.7565336227417,
+      "learning_rate": 8.788212889646187e-06,
+      "loss": 1.1467,
+      "step": 5594
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 32.24980163574219,
+      "learning_rate": 8.786208279041797e-06,
+      "loss": 1.8956,
+      "step": 5595
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 15.186212539672852,
+      "learning_rate": 8.784203668437407e-06,
+      "loss": 1.8275,
+      "step": 5596
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 26.510732650756836,
+      "learning_rate": 8.782199057833017e-06,
+      "loss": 1.2493,
+      "step": 5597
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 9.903314590454102,
+      "learning_rate": 8.780194447228625e-06,
+      "loss": 0.8917,
+      "step": 5598
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 9.444249153137207,
+      "learning_rate": 8.778189836624237e-06,
+      "loss": 1.073,
+      "step": 5599
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 17.22249412536621,
+      "learning_rate": 8.776185226019847e-06,
+      "loss": 1.2148,
+      "step": 5600
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 9.89676284790039,
+      "learning_rate": 8.774180615415456e-06,
+      "loss": 1.2275,
+      "step": 5601
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 12.96465015411377,
+      "learning_rate": 8.772176004811066e-06,
+      "loss": 1.0976,
+      "step": 5602
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 15.916191101074219,
+      "learning_rate": 8.770171394206676e-06,
+      "loss": 1.7115,
+      "step": 5603
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 18.037708282470703,
+      "learning_rate": 8.768166783602286e-06,
+      "loss": 1.7032,
+      "step": 5604
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 15.334985733032227,
+      "learning_rate": 8.766162172997896e-06,
+      "loss": 1.3752,
+      "step": 5605
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 9.522956848144531,
+      "learning_rate": 8.764157562393506e-06,
+      "loss": 0.623,
+      "step": 5606
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 21.579017639160156,
+      "learning_rate": 8.762152951789116e-06,
+      "loss": 1.5285,
+      "step": 5607
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 10.721216201782227,
+      "learning_rate": 8.760148341184726e-06,
+      "loss": 0.893,
+      "step": 5608
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 29.23563575744629,
+      "learning_rate": 8.758143730580335e-06,
+      "loss": 2.0828,
+      "step": 5609
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 5.085595607757568,
+      "learning_rate": 8.756139119975945e-06,
+      "loss": 0.6403,
+      "step": 5610
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 16.123058319091797,
+      "learning_rate": 8.754134509371556e-06,
+      "loss": 1.7075,
+      "step": 5611
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 54.834938049316406,
+      "learning_rate": 8.752129898767165e-06,
+      "loss": 1.6164,
+      "step": 5612
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 15.420845031738281,
+      "learning_rate": 8.750125288162775e-06,
+      "loss": 1.3926,
+      "step": 5613
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 14.784860610961914,
+      "learning_rate": 8.748120677558385e-06,
+      "loss": 1.175,
+      "step": 5614
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 32.64680862426758,
+      "learning_rate": 8.746116066953995e-06,
+      "loss": 1.5805,
+      "step": 5615
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 28.060972213745117,
+      "learning_rate": 8.744111456349605e-06,
+      "loss": 1.803,
+      "step": 5616
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 12.223137855529785,
+      "learning_rate": 8.742106845745215e-06,
+      "loss": 1.149,
+      "step": 5617
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 14.413814544677734,
+      "learning_rate": 8.740102235140825e-06,
+      "loss": 1.5615,
+      "step": 5618
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 24.743038177490234,
+      "learning_rate": 8.738097624536435e-06,
+      "loss": 1.5725,
+      "step": 5619
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 37.43949508666992,
+      "learning_rate": 8.736093013932044e-06,
+      "loss": 2.383,
+      "step": 5620
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 19.173534393310547,
+      "learning_rate": 8.734088403327654e-06,
+      "loss": 2.1207,
+      "step": 5621
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 17.49724006652832,
+      "learning_rate": 8.732083792723264e-06,
+      "loss": 1.7264,
+      "step": 5622
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 9.76134967803955,
+      "learning_rate": 8.730079182118874e-06,
+      "loss": 1.5944,
+      "step": 5623
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 11.717310905456543,
+      "learning_rate": 8.728074571514484e-06,
+      "loss": 1.756,
+      "step": 5624
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 21.17192840576172,
+      "learning_rate": 8.726069960910094e-06,
+      "loss": 2.2556,
+      "step": 5625
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 16.35768699645996,
+      "learning_rate": 8.724065350305704e-06,
+      "loss": 1.2782,
+      "step": 5626
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 12.22645378112793,
+      "learning_rate": 8.722060739701313e-06,
+      "loss": 1.3418,
+      "step": 5627
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 19.546850204467773,
+      "learning_rate": 8.720056129096923e-06,
+      "loss": 1.2375,
+      "step": 5628
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 14.505806922912598,
+      "learning_rate": 8.718051518492535e-06,
+      "loss": 1.5247,
+      "step": 5629
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 11.707478523254395,
+      "learning_rate": 8.716046907888145e-06,
+      "loss": 1.2686,
+      "step": 5630
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 12.525168418884277,
+      "learning_rate": 8.714042297283753e-06,
+      "loss": 2.1298,
+      "step": 5631
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 13.70942497253418,
+      "learning_rate": 8.712037686679363e-06,
+      "loss": 0.8167,
+      "step": 5632
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 18.232576370239258,
+      "learning_rate": 8.710033076074973e-06,
+      "loss": 1.0261,
+      "step": 5633
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 19.63440704345703,
+      "learning_rate": 8.708028465470583e-06,
+      "loss": 2.1018,
+      "step": 5634
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 17.513425827026367,
+      "learning_rate": 8.706023854866193e-06,
+      "loss": 1.2959,
+      "step": 5635
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 15.64637565612793,
+      "learning_rate": 8.704019244261803e-06,
+      "loss": 1.8533,
+      "step": 5636
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 7.151846885681152,
+      "learning_rate": 8.702014633657414e-06,
+      "loss": 0.8085,
+      "step": 5637
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 9.188348770141602,
+      "learning_rate": 8.700010023053022e-06,
+      "loss": 1.1322,
+      "step": 5638
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 9.800728797912598,
+      "learning_rate": 8.698005412448632e-06,
+      "loss": 0.9624,
+      "step": 5639
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 14.96066951751709,
+      "learning_rate": 8.696000801844242e-06,
+      "loss": 1.3465,
+      "step": 5640
+    },
+    {
+      "epoch": 1.7,
+      "eval_loss": 0.1950412392616272,
+      "eval_runtime": 43.6602,
+      "eval_samples_per_second": 33.875,
+      "eval_steps_per_second": 33.875,
+      "step": 5640
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 21.22214698791504,
+      "learning_rate": 8.693996191239852e-06,
+      "loss": 1.9302,
+      "step": 5641
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 11.030116081237793,
+      "learning_rate": 8.691991580635462e-06,
+      "loss": 0.9403,
+      "step": 5642
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 10.79010009765625,
+      "learning_rate": 8.689986970031072e-06,
+      "loss": 1.2434,
+      "step": 5643
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 17.979541778564453,
+      "learning_rate": 8.687982359426682e-06,
+      "loss": 1.4883,
+      "step": 5644
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 11.965887069702148,
+      "learning_rate": 8.685977748822292e-06,
+      "loss": 1.381,
+      "step": 5645
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 16.491119384765625,
+      "learning_rate": 8.683973138217901e-06,
+      "loss": 1.1538,
+      "step": 5646
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 12.865142822265625,
+      "learning_rate": 8.681968527613511e-06,
+      "loss": 1.2,
+      "step": 5647
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 22.270700454711914,
+      "learning_rate": 8.679963917009123e-06,
+      "loss": 1.3727,
+      "step": 5648
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 31.27298927307129,
+      "learning_rate": 8.677959306404731e-06,
+      "loss": 2.6597,
+      "step": 5649
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 10.323209762573242,
+      "learning_rate": 8.675954695800341e-06,
+      "loss": 0.9057,
+      "step": 5650
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 14.9830961227417,
+      "learning_rate": 8.673950085195951e-06,
+      "loss": 2.7499,
+      "step": 5651
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 17.780193328857422,
+      "learning_rate": 8.671945474591561e-06,
+      "loss": 1.4829,
+      "step": 5652
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 14.081412315368652,
+      "learning_rate": 8.669940863987171e-06,
+      "loss": 1.159,
+      "step": 5653
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 16.137582778930664,
+      "learning_rate": 8.667936253382782e-06,
+      "loss": 1.0004,
+      "step": 5654
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 21.85881805419922,
+      "learning_rate": 8.665931642778392e-06,
+      "loss": 1.4793,
+      "step": 5655
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 21.70241355895996,
+      "learning_rate": 8.663927032174002e-06,
+      "loss": 1.6126,
+      "step": 5656
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 16.00739288330078,
+      "learning_rate": 8.66192242156961e-06,
+      "loss": 1.7453,
+      "step": 5657
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 41.264095306396484,
+      "learning_rate": 8.65991781096522e-06,
+      "loss": 1.8027,
+      "step": 5658
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 59.908199310302734,
+      "learning_rate": 8.65791320036083e-06,
+      "loss": 1.4352,
+      "step": 5659
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 26.252933502197266,
+      "learning_rate": 8.65590858975644e-06,
+      "loss": 1.1851,
+      "step": 5660
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 14.983129501342773,
+      "learning_rate": 8.65390397915205e-06,
+      "loss": 1.1969,
+      "step": 5661
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 7.037642478942871,
+      "learning_rate": 8.65189936854766e-06,
+      "loss": 0.6684,
+      "step": 5662
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 12.585198402404785,
+      "learning_rate": 8.64989475794327e-06,
+      "loss": 1.4334,
+      "step": 5663
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 21.677915573120117,
+      "learning_rate": 8.64789014733888e-06,
+      "loss": 2.5165,
+      "step": 5664
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 23.204702377319336,
+      "learning_rate": 8.645885536734489e-06,
+      "loss": 1.5472,
+      "step": 5665
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 10.986197471618652,
+      "learning_rate": 8.643880926130101e-06,
+      "loss": 1.6953,
+      "step": 5666
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 11.328323364257812,
+      "learning_rate": 8.641876315525711e-06,
+      "loss": 0.5602,
+      "step": 5667
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 9.922806739807129,
+      "learning_rate": 8.63987170492132e-06,
+      "loss": 1.3713,
+      "step": 5668
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 8.560861587524414,
+      "learning_rate": 8.63786709431693e-06,
+      "loss": 1.0511,
+      "step": 5669
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 38.06193923950195,
+      "learning_rate": 8.63586248371254e-06,
+      "loss": 1.5991,
+      "step": 5670
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 17.680063247680664,
+      "learning_rate": 8.63385787310815e-06,
+      "loss": 2.7477,
+      "step": 5671
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 29.49445915222168,
+      "learning_rate": 8.63185326250376e-06,
+      "loss": 1.7613,
+      "step": 5672
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 32.89970397949219,
+      "learning_rate": 8.62984865189937e-06,
+      "loss": 1.6423,
+      "step": 5673
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 40.36264419555664,
+      "learning_rate": 8.62784404129498e-06,
+      "loss": 1.2661,
+      "step": 5674
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 15.551639556884766,
+      "learning_rate": 8.625839430690588e-06,
+      "loss": 1.6711,
+      "step": 5675
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 30.854095458984375,
+      "learning_rate": 8.623834820086198e-06,
+      "loss": 1.0708,
+      "step": 5676
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 30.149341583251953,
+      "learning_rate": 8.621830209481808e-06,
+      "loss": 1.7262,
+      "step": 5677
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 17.061527252197266,
+      "learning_rate": 8.619825598877418e-06,
+      "loss": 1.4248,
+      "step": 5678
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 78.73602294921875,
+      "learning_rate": 8.617820988273029e-06,
+      "loss": 2.1561,
+      "step": 5679
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 51.467491149902344,
+      "learning_rate": 8.615816377668639e-06,
+      "loss": 1.7134,
+      "step": 5680
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 17.935325622558594,
+      "learning_rate": 8.613811767064249e-06,
+      "loss": 1.576,
+      "step": 5681
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 14.966944694519043,
+      "learning_rate": 8.611807156459859e-06,
+      "loss": 2.0311,
+      "step": 5682
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 14.564214706420898,
+      "learning_rate": 8.609802545855467e-06,
+      "loss": 1.7562,
+      "step": 5683
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 11.899775505065918,
+      "learning_rate": 8.607797935251077e-06,
+      "loss": 1.2177,
+      "step": 5684
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 14.998151779174805,
+      "learning_rate": 8.605793324646689e-06,
+      "loss": 1.6503,
+      "step": 5685
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 18.94357681274414,
+      "learning_rate": 8.603788714042297e-06,
+      "loss": 1.411,
+      "step": 5686
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 14.1841459274292,
+      "learning_rate": 8.601784103437908e-06,
+      "loss": 1.5919,
+      "step": 5687
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 32.609100341796875,
+      "learning_rate": 8.599779492833518e-06,
+      "loss": 1.7305,
+      "step": 5688
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 35.52297592163086,
+      "learning_rate": 8.597774882229128e-06,
+      "loss": 2.0784,
+      "step": 5689
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 13.570123672485352,
+      "learning_rate": 8.595770271624738e-06,
+      "loss": 1.0387,
+      "step": 5690
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 12.182998657226562,
+      "learning_rate": 8.593765661020348e-06,
+      "loss": 1.042,
+      "step": 5691
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 56.89901351928711,
+      "learning_rate": 8.591761050415958e-06,
+      "loss": 2.2758,
+      "step": 5692
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 16.903789520263672,
+      "learning_rate": 8.589756439811568e-06,
+      "loss": 1.0774,
+      "step": 5693
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 36.77326583862305,
+      "learning_rate": 8.587751829207176e-06,
+      "loss": 2.6291,
+      "step": 5694
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 24.694618225097656,
+      "learning_rate": 8.585747218602787e-06,
+      "loss": 1.4135,
+      "step": 5695
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 44.71474075317383,
+      "learning_rate": 8.583742607998397e-06,
+      "loss": 1.5254,
+      "step": 5696
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 27.899707794189453,
+      "learning_rate": 8.581737997394007e-06,
+      "loss": 1.8512,
+      "step": 5697
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 8.52084732055664,
+      "learning_rate": 8.579733386789617e-06,
+      "loss": 1.3119,
+      "step": 5698
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 12.418741226196289,
+      "learning_rate": 8.577728776185227e-06,
+      "loss": 1.7178,
+      "step": 5699
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 11.793305397033691,
+      "learning_rate": 8.575724165580837e-06,
+      "loss": 1.0568,
+      "step": 5700
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 9.70165729522705,
+      "learning_rate": 8.573719554976447e-06,
+      "loss": 1.6653,
+      "step": 5701
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 15.392131805419922,
+      "learning_rate": 8.571714944372055e-06,
+      "loss": 2.2157,
+      "step": 5702
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 35.93635177612305,
+      "learning_rate": 8.569710333767667e-06,
+      "loss": 1.2638,
+      "step": 5703
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 12.030016899108887,
+      "learning_rate": 8.567705723163277e-06,
+      "loss": 1.3019,
+      "step": 5704
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 39.288814544677734,
+      "learning_rate": 8.565701112558886e-06,
+      "loss": 2.3544,
+      "step": 5705
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 23.307493209838867,
+      "learning_rate": 8.563696501954496e-06,
+      "loss": 1.5503,
+      "step": 5706
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 26.393465042114258,
+      "learning_rate": 8.561691891350106e-06,
+      "loss": 2.5251,
+      "step": 5707
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 10.045159339904785,
+      "learning_rate": 8.559687280745716e-06,
+      "loss": 1.1675,
+      "step": 5708
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 35.52949142456055,
+      "learning_rate": 8.557682670141326e-06,
+      "loss": 2.1613,
+      "step": 5709
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 19.822158813476562,
+      "learning_rate": 8.555678059536936e-06,
+      "loss": 1.5757,
+      "step": 5710
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 41.30690383911133,
+      "learning_rate": 8.553673448932546e-06,
+      "loss": 2.7388,
+      "step": 5711
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 20.42682456970215,
+      "learning_rate": 8.551668838328155e-06,
+      "loss": 2.3481,
+      "step": 5712
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 9.312350273132324,
+      "learning_rate": 8.549664227723765e-06,
+      "loss": 0.9268,
+      "step": 5713
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 11.393967628479004,
+      "learning_rate": 8.547659617119375e-06,
+      "loss": 1.6853,
+      "step": 5714
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 22.255510330200195,
+      "learning_rate": 8.545655006514987e-06,
+      "loss": 1.0106,
+      "step": 5715
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 12.787086486816406,
+      "learning_rate": 8.543650395910595e-06,
+      "loss": 1.6355,
+      "step": 5716
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 39.999839782714844,
+      "learning_rate": 8.541645785306205e-06,
+      "loss": 1.8123,
+      "step": 5717
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 7.7507853507995605,
+      "learning_rate": 8.539641174701815e-06,
+      "loss": 1.4238,
+      "step": 5718
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 15.439056396484375,
+      "learning_rate": 8.537636564097425e-06,
+      "loss": 1.7531,
+      "step": 5719
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 17.616294860839844,
+      "learning_rate": 8.535631953493034e-06,
+      "loss": 1.8694,
+      "step": 5720
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 10.567305564880371,
+      "learning_rate": 8.533627342888644e-06,
+      "loss": 1.0895,
+      "step": 5721
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 14.490511894226074,
+      "learning_rate": 8.531622732284255e-06,
+      "loss": 1.4669,
+      "step": 5722
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 9.218794822692871,
+      "learning_rate": 8.529618121679864e-06,
+      "loss": 1.0658,
+      "step": 5723
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 10.850655555725098,
+      "learning_rate": 8.527613511075474e-06,
+      "loss": 1.0692,
+      "step": 5724
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 29.35923194885254,
+      "learning_rate": 8.525608900471084e-06,
+      "loss": 1.4617,
+      "step": 5725
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 19.219905853271484,
+      "learning_rate": 8.523604289866694e-06,
+      "loss": 2.7337,
+      "step": 5726
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 17.148277282714844,
+      "learning_rate": 8.521599679262304e-06,
+      "loss": 1.4809,
+      "step": 5727
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 55.859092712402344,
+      "learning_rate": 8.519595068657914e-06,
+      "loss": 1.3051,
+      "step": 5728
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 12.13316822052002,
+      "learning_rate": 8.517590458053524e-06,
+      "loss": 1.3944,
+      "step": 5729
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 26.604719161987305,
+      "learning_rate": 8.515585847449134e-06,
+      "loss": 1.6258,
+      "step": 5730
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 13.429365158081055,
+      "learning_rate": 8.513581236844743e-06,
+      "loss": 1.0267,
+      "step": 5731
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 13.195096969604492,
+      "learning_rate": 8.511576626240353e-06,
+      "loss": 1.1112,
+      "step": 5732
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 19.855567932128906,
+      "learning_rate": 8.509572015635963e-06,
+      "loss": 1.1505,
+      "step": 5733
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 13.332557678222656,
+      "learning_rate": 8.507567405031573e-06,
+      "loss": 1.5326,
+      "step": 5734
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 11.699455261230469,
+      "learning_rate": 8.505562794427183e-06,
+      "loss": 1.3027,
+      "step": 5735
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 55.0339469909668,
+      "learning_rate": 8.503558183822793e-06,
+      "loss": 1.8905,
+      "step": 5736
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 14.223106384277344,
+      "learning_rate": 8.501553573218403e-06,
+      "loss": 1.1121,
+      "step": 5737
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 19.943954467773438,
+      "learning_rate": 8.499548962614013e-06,
+      "loss": 1.9677,
+      "step": 5738
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 30.25005340576172,
+      "learning_rate": 8.497544352009622e-06,
+      "loss": 1.7023,
+      "step": 5739
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 19.0448055267334,
+      "learning_rate": 8.495539741405234e-06,
+      "loss": 1.3845,
+      "step": 5740
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 16.908838272094727,
+      "learning_rate": 8.493535130800844e-06,
+      "loss": 1.7335,
+      "step": 5741
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 24.786378860473633,
+      "learning_rate": 8.491530520196452e-06,
+      "loss": 1.7752,
+      "step": 5742
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 14.871647834777832,
+      "learning_rate": 8.489525909592062e-06,
+      "loss": 1.7707,
+      "step": 5743
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 9.237364768981934,
+      "learning_rate": 8.487521298987672e-06,
+      "loss": 1.2148,
+      "step": 5744
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 20.291860580444336,
+      "learning_rate": 8.485516688383282e-06,
+      "loss": 1.3439,
+      "step": 5745
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 23.027454376220703,
+      "learning_rate": 8.483512077778892e-06,
+      "loss": 1.6966,
+      "step": 5746
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 14.931220054626465,
+      "learning_rate": 8.481507467174502e-06,
+      "loss": 1.8485,
+      "step": 5747
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 8.827281951904297,
+      "learning_rate": 8.479502856570113e-06,
+      "loss": 0.8949,
+      "step": 5748
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 13.313647270202637,
+      "learning_rate": 8.477498245965723e-06,
+      "loss": 1.3746,
+      "step": 5749
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 40.448280334472656,
+      "learning_rate": 8.475493635361331e-06,
+      "loss": 2.6908,
+      "step": 5750
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 15.871477127075195,
+      "learning_rate": 8.473489024756941e-06,
+      "loss": 1.6084,
+      "step": 5751
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 38.03894805908203,
+      "learning_rate": 8.471484414152553e-06,
+      "loss": 1.693,
+      "step": 5752
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 53.909751892089844,
+      "learning_rate": 8.469479803548161e-06,
+      "loss": 1.3998,
+      "step": 5753
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 23.2731876373291,
+      "learning_rate": 8.467475192943771e-06,
+      "loss": 1.1304,
+      "step": 5754
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 16.913698196411133,
+      "learning_rate": 8.465470582339381e-06,
+      "loss": 1.2357,
+      "step": 5755
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 25.4580135345459,
+      "learning_rate": 8.463465971734992e-06,
+      "loss": 1.669,
+      "step": 5756
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 17.684297561645508,
+      "learning_rate": 8.4614613611306e-06,
+      "loss": 1.8284,
+      "step": 5757
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 11.82136058807373,
+      "learning_rate": 8.459456750526212e-06,
+      "loss": 1.5129,
+      "step": 5758
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 19.27686882019043,
+      "learning_rate": 8.457452139921822e-06,
+      "loss": 2.0577,
+      "step": 5759
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 19.104829788208008,
+      "learning_rate": 8.45544752931743e-06,
+      "loss": 1.8845,
+      "step": 5760
+    },
+    {
+      "epoch": 1.73,
+      "eval_loss": 0.19915775954723358,
+      "eval_runtime": 43.571,
+      "eval_samples_per_second": 33.945,
+      "eval_steps_per_second": 33.945,
+      "step": 5760
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 30.6223201751709,
+      "learning_rate": 8.45344291871304e-06,
+      "loss": 1.4588,
+      "step": 5761
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 20.711313247680664,
+      "learning_rate": 8.45143830810865e-06,
+      "loss": 1.9552,
+      "step": 5762
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 9.937006950378418,
+      "learning_rate": 8.44943369750426e-06,
+      "loss": 1.4769,
+      "step": 5763
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 11.456831932067871,
+      "learning_rate": 8.44742908689987e-06,
+      "loss": 1.486,
+      "step": 5764
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 11.295050621032715,
+      "learning_rate": 8.44542447629548e-06,
+      "loss": 1.1309,
+      "step": 5765
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 10.123924255371094,
+      "learning_rate": 8.44341986569109e-06,
+      "loss": 0.8497,
+      "step": 5766
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 23.44426727294922,
+      "learning_rate": 8.4414152550867e-06,
+      "loss": 1.9237,
+      "step": 5767
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 23.85992431640625,
+      "learning_rate": 8.439410644482309e-06,
+      "loss": 2.8773,
+      "step": 5768
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 12.136680603027344,
+      "learning_rate": 8.43740603387792e-06,
+      "loss": 1.9736,
+      "step": 5769
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 13.099090576171875,
+      "learning_rate": 8.43540142327353e-06,
+      "loss": 1.9405,
+      "step": 5770
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 18.678722381591797,
+      "learning_rate": 8.43339681266914e-06,
+      "loss": 2.2382,
+      "step": 5771
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 14.240285873413086,
+      "learning_rate": 8.43139220206475e-06,
+      "loss": 1.3103,
+      "step": 5772
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 21.930389404296875,
+      "learning_rate": 8.42938759146036e-06,
+      "loss": 1.4554,
+      "step": 5773
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 20.750886917114258,
+      "learning_rate": 8.42738298085597e-06,
+      "loss": 1.1572,
+      "step": 5774
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 13.55035400390625,
+      "learning_rate": 8.42537837025158e-06,
+      "loss": 1.0804,
+      "step": 5775
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 12.210824966430664,
+      "learning_rate": 8.423373759647188e-06,
+      "loss": 1.526,
+      "step": 5776
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 9.945907592773438,
+      "learning_rate": 8.4213691490428e-06,
+      "loss": 1.1015,
+      "step": 5777
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 16.482894897460938,
+      "learning_rate": 8.41936453843841e-06,
+      "loss": 1.9265,
+      "step": 5778
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 16.412376403808594,
+      "learning_rate": 8.417359927834018e-06,
+      "loss": 1.9719,
+      "step": 5779
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 49.32164764404297,
+      "learning_rate": 8.415355317229628e-06,
+      "loss": 2.7723,
+      "step": 5780
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 27.84273910522461,
+      "learning_rate": 8.413350706625239e-06,
+      "loss": 1.1557,
+      "step": 5781
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 25.465152740478516,
+      "learning_rate": 8.411346096020849e-06,
+      "loss": 1.241,
+      "step": 5782
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 21.347864151000977,
+      "learning_rate": 8.409341485416459e-06,
+      "loss": 2.8817,
+      "step": 5783
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 17.83620834350586,
+      "learning_rate": 8.407336874812069e-06,
+      "loss": 1.4303,
+      "step": 5784
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 24.558000564575195,
+      "learning_rate": 8.405332264207679e-06,
+      "loss": 0.9856,
+      "step": 5785
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 16.54327392578125,
+      "learning_rate": 8.403327653603289e-06,
+      "loss": 0.702,
+      "step": 5786
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 21.651714324951172,
+      "learning_rate": 8.401323042998897e-06,
+      "loss": 1.6273,
+      "step": 5787
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 13.498676300048828,
+      "learning_rate": 8.399318432394507e-06,
+      "loss": 0.949,
+      "step": 5788
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 10.88339900970459,
+      "learning_rate": 8.39731382179012e-06,
+      "loss": 1.7288,
+      "step": 5789
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 10.3096284866333,
+      "learning_rate": 8.395309211185728e-06,
+      "loss": 1.1514,
+      "step": 5790
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 17.805374145507812,
+      "learning_rate": 8.393304600581338e-06,
+      "loss": 1.5093,
+      "step": 5791
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 16.813570022583008,
+      "learning_rate": 8.391299989976948e-06,
+      "loss": 0.8186,
+      "step": 5792
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 20.913787841796875,
+      "learning_rate": 8.389295379372558e-06,
+      "loss": 1.7035,
+      "step": 5793
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 52.88319778442383,
+      "learning_rate": 8.387290768768166e-06,
+      "loss": 1.7603,
+      "step": 5794
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 15.752192497253418,
+      "learning_rate": 8.385286158163778e-06,
+      "loss": 1.2862,
+      "step": 5795
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 8.408364295959473,
+      "learning_rate": 8.383281547559388e-06,
+      "loss": 0.9196,
+      "step": 5796
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 18.868669509887695,
+      "learning_rate": 8.381276936954997e-06,
+      "loss": 1.5121,
+      "step": 5797
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 20.32513999938965,
+      "learning_rate": 8.379272326350607e-06,
+      "loss": 2.3994,
+      "step": 5798
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 19.855003356933594,
+      "learning_rate": 8.377267715746217e-06,
+      "loss": 1.8385,
+      "step": 5799
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 15.905160903930664,
+      "learning_rate": 8.375263105141827e-06,
+      "loss": 1.1867,
+      "step": 5800
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 8.156031608581543,
+      "learning_rate": 8.373258494537437e-06,
+      "loss": 0.9248,
+      "step": 5801
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 20.510618209838867,
+      "learning_rate": 8.371253883933047e-06,
+      "loss": 1.6939,
+      "step": 5802
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 39.78818130493164,
+      "learning_rate": 8.369249273328657e-06,
+      "loss": 1.3656,
+      "step": 5803
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 33.16683578491211,
+      "learning_rate": 8.367244662724267e-06,
+      "loss": 1.5406,
+      "step": 5804
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 21.53314971923828,
+      "learning_rate": 8.365240052119875e-06,
+      "loss": 1.0033,
+      "step": 5805
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 24.56745719909668,
+      "learning_rate": 8.363235441515486e-06,
+      "loss": 1.7757,
+      "step": 5806
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 15.24342155456543,
+      "learning_rate": 8.361230830911096e-06,
+      "loss": 1.292,
+      "step": 5807
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 27.435335159301758,
+      "learning_rate": 8.359226220306706e-06,
+      "loss": 1.9828,
+      "step": 5808
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 20.47377586364746,
+      "learning_rate": 8.357221609702316e-06,
+      "loss": 1.9325,
+      "step": 5809
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 6.70631742477417,
+      "learning_rate": 8.355216999097926e-06,
+      "loss": 1.1601,
+      "step": 5810
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 13.876242637634277,
+      "learning_rate": 8.353212388493536e-06,
+      "loss": 1.3469,
+      "step": 5811
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 22.673198699951172,
+      "learning_rate": 8.351207777889146e-06,
+      "loss": 2.9192,
+      "step": 5812
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 9.959436416625977,
+      "learning_rate": 8.349203167284754e-06,
+      "loss": 1.4124,
+      "step": 5813
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 30.46356773376465,
+      "learning_rate": 8.347198556680366e-06,
+      "loss": 2.1221,
+      "step": 5814
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 12.394341468811035,
+      "learning_rate": 8.345193946075976e-06,
+      "loss": 1.2795,
+      "step": 5815
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 71.86808013916016,
+      "learning_rate": 8.343189335471585e-06,
+      "loss": 3.5364,
+      "step": 5816
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 13.528796195983887,
+      "learning_rate": 8.341184724867195e-06,
+      "loss": 1.1612,
+      "step": 5817
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 38.51426696777344,
+      "learning_rate": 8.339180114262805e-06,
+      "loss": 1.6762,
+      "step": 5818
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 15.68415355682373,
+      "learning_rate": 8.337175503658415e-06,
+      "loss": 2.1607,
+      "step": 5819
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 11.136553764343262,
+      "learning_rate": 8.335170893054025e-06,
+      "loss": 0.7781,
+      "step": 5820
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 15.836930274963379,
+      "learning_rate": 8.333166282449635e-06,
+      "loss": 1.9072,
+      "step": 5821
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 27.627634048461914,
+      "learning_rate": 8.331161671845245e-06,
+      "loss": 1.4755,
+      "step": 5822
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 11.358156204223633,
+      "learning_rate": 8.329157061240855e-06,
+      "loss": 1.3935,
+      "step": 5823
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 11.681863784790039,
+      "learning_rate": 8.327152450636464e-06,
+      "loss": 1.2521,
+      "step": 5824
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 37.17042541503906,
+      "learning_rate": 8.325147840032074e-06,
+      "loss": 1.7879,
+      "step": 5825
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 17.618371963500977,
+      "learning_rate": 8.323143229427686e-06,
+      "loss": 1.283,
+      "step": 5826
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 13.247238159179688,
+      "learning_rate": 8.321138618823294e-06,
+      "loss": 0.9504,
+      "step": 5827
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 22.750028610229492,
+      "learning_rate": 8.319134008218904e-06,
+      "loss": 1.7133,
+      "step": 5828
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 23.75035285949707,
+      "learning_rate": 8.317129397614514e-06,
+      "loss": 1.4273,
+      "step": 5829
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 17.772930145263672,
+      "learning_rate": 8.315124787010124e-06,
+      "loss": 1.4838,
+      "step": 5830
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 14.174525260925293,
+      "learning_rate": 8.313120176405733e-06,
+      "loss": 1.5128,
+      "step": 5831
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 59.5333366394043,
+      "learning_rate": 8.311115565801344e-06,
+      "loss": 1.0576,
+      "step": 5832
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 20.53312873840332,
+      "learning_rate": 8.309110955196954e-06,
+      "loss": 1.0399,
+      "step": 5833
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 31.05029296875,
+      "learning_rate": 8.307106344592565e-06,
+      "loss": 1.5115,
+      "step": 5834
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 12.972631454467773,
+      "learning_rate": 8.305101733988173e-06,
+      "loss": 1.2375,
+      "step": 5835
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 13.963759422302246,
+      "learning_rate": 8.303097123383783e-06,
+      "loss": 1.5626,
+      "step": 5836
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 11.903511047363281,
+      "learning_rate": 8.301092512779393e-06,
+      "loss": 1.4446,
+      "step": 5837
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 18.711076736450195,
+      "learning_rate": 8.299087902175003e-06,
+      "loss": 1.3406,
+      "step": 5838
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 84.2789535522461,
+      "learning_rate": 8.297083291570613e-06,
+      "loss": 2.1487,
+      "step": 5839
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 22.546630859375,
+      "learning_rate": 8.295078680966223e-06,
+      "loss": 1.1753,
+      "step": 5840
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 12.994787216186523,
+      "learning_rate": 8.293074070361833e-06,
+      "loss": 1.6728,
+      "step": 5841
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 8.531490325927734,
+      "learning_rate": 8.291069459757442e-06,
+      "loss": 0.8257,
+      "step": 5842
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 21.086061477661133,
+      "learning_rate": 8.289064849153052e-06,
+      "loss": 1.1679,
+      "step": 5843
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 23.762374877929688,
+      "learning_rate": 8.287060238548664e-06,
+      "loss": 3.049,
+      "step": 5844
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 18.857986450195312,
+      "learning_rate": 8.285055627944272e-06,
+      "loss": 1.5555,
+      "step": 5845
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 11.890861511230469,
+      "learning_rate": 8.283051017339882e-06,
+      "loss": 1.1098,
+      "step": 5846
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 8.958333969116211,
+      "learning_rate": 8.281046406735492e-06,
+      "loss": 0.8418,
+      "step": 5847
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 15.934253692626953,
+      "learning_rate": 8.279041796131102e-06,
+      "loss": 1.3008,
+      "step": 5848
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 21.336219787597656,
+      "learning_rate": 8.277037185526712e-06,
+      "loss": 1.1645,
+      "step": 5849
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 12.621298789978027,
+      "learning_rate": 8.27503257492232e-06,
+      "loss": 2.0196,
+      "step": 5850
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 13.761375427246094,
+      "learning_rate": 8.273027964317933e-06,
+      "loss": 2.1065,
+      "step": 5851
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 35.57234191894531,
+      "learning_rate": 8.271023353713543e-06,
+      "loss": 1.1288,
+      "step": 5852
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 10.00034236907959,
+      "learning_rate": 8.269018743109151e-06,
+      "loss": 1.3954,
+      "step": 5853
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 37.05139923095703,
+      "learning_rate": 8.267014132504761e-06,
+      "loss": 1.4784,
+      "step": 5854
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 22.268341064453125,
+      "learning_rate": 8.265009521900371e-06,
+      "loss": 1.3877,
+      "step": 5855
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 13.753389358520508,
+      "learning_rate": 8.263004911295981e-06,
+      "loss": 1.5012,
+      "step": 5856
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 18.250154495239258,
+      "learning_rate": 8.261000300691591e-06,
+      "loss": 1.7476,
+      "step": 5857
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 14.758729934692383,
+      "learning_rate": 8.258995690087201e-06,
+      "loss": 1.5117,
+      "step": 5858
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 20.81943130493164,
+      "learning_rate": 8.256991079482812e-06,
+      "loss": 1.363,
+      "step": 5859
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 26.64165496826172,
+      "learning_rate": 8.254986468878422e-06,
+      "loss": 1.7073,
+      "step": 5860
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 14.987895965576172,
+      "learning_rate": 8.25298185827403e-06,
+      "loss": 0.7878,
+      "step": 5861
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 17.91753387451172,
+      "learning_rate": 8.25097724766964e-06,
+      "loss": 1.4102,
+      "step": 5862
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 6.570464134216309,
+      "learning_rate": 8.248972637065252e-06,
+      "loss": 0.7017,
+      "step": 5863
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 21.44278335571289,
+      "learning_rate": 8.24696802646086e-06,
+      "loss": 1.4487,
+      "step": 5864
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 15.875109672546387,
+      "learning_rate": 8.24496341585647e-06,
+      "loss": 1.178,
+      "step": 5865
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 36.79011154174805,
+      "learning_rate": 8.24295880525208e-06,
+      "loss": 1.9915,
+      "step": 5866
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 21.93402862548828,
+      "learning_rate": 8.24095419464769e-06,
+      "loss": 1.4122,
+      "step": 5867
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 23.80208969116211,
+      "learning_rate": 8.2389495840433e-06,
+      "loss": 2.1488,
+      "step": 5868
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 10.25802993774414,
+      "learning_rate": 8.23694497343891e-06,
+      "loss": 1.7158,
+      "step": 5869
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 38.02273178100586,
+      "learning_rate": 8.23494036283452e-06,
+      "loss": 1.6496,
+      "step": 5870
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 27.261823654174805,
+      "learning_rate": 8.232935752230131e-06,
+      "loss": 2.1812,
+      "step": 5871
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 10.888236999511719,
+      "learning_rate": 8.23093114162574e-06,
+      "loss": 1.3974,
+      "step": 5872
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 33.661293029785156,
+      "learning_rate": 8.22892653102135e-06,
+      "loss": 2.1975,
+      "step": 5873
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 38.529293060302734,
+      "learning_rate": 8.22692192041696e-06,
+      "loss": 1.8204,
+      "step": 5874
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 28.78758430480957,
+      "learning_rate": 8.22491730981257e-06,
+      "loss": 2.9932,
+      "step": 5875
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 10.32004165649414,
+      "learning_rate": 8.22291269920818e-06,
+      "loss": 2.0646,
+      "step": 5876
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 18.924707412719727,
+      "learning_rate": 8.22090808860379e-06,
+      "loss": 2.3191,
+      "step": 5877
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 14.681024551391602,
+      "learning_rate": 8.2189034779994e-06,
+      "loss": 1.1271,
+      "step": 5878
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 13.027987480163574,
+      "learning_rate": 8.216898867395008e-06,
+      "loss": 1.6397,
+      "step": 5879
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 37.9376335144043,
+      "learning_rate": 8.214894256790618e-06,
+      "loss": 1.7215,
+      "step": 5880
+    },
+    {
+      "epoch": 1.77,
+      "eval_loss": 0.19819287955760956,
+      "eval_runtime": 43.4174,
+      "eval_samples_per_second": 34.065,
+      "eval_steps_per_second": 34.065,
+      "step": 5880
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 37.3797492980957,
+      "learning_rate": 8.21288964618623e-06,
+      "loss": 2.3832,
+      "step": 5881
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 20.478801727294922,
+      "learning_rate": 8.210885035581838e-06,
+      "loss": 1.8481,
+      "step": 5882
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 22.586971282958984,
+      "learning_rate": 8.208880424977449e-06,
+      "loss": 1.8847,
+      "step": 5883
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 25.06772232055664,
+      "learning_rate": 8.206875814373059e-06,
+      "loss": 2.011,
+      "step": 5884
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 13.59852123260498,
+      "learning_rate": 8.204871203768669e-06,
+      "loss": 1.5316,
+      "step": 5885
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 15.501117706298828,
+      "learning_rate": 8.202866593164279e-06,
+      "loss": 1.5476,
+      "step": 5886
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 15.47221851348877,
+      "learning_rate": 8.200861982559889e-06,
+      "loss": 1.5741,
+      "step": 5887
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 25.984909057617188,
+      "learning_rate": 8.198857371955499e-06,
+      "loss": 1.8725,
+      "step": 5888
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 23.49461555480957,
+      "learning_rate": 8.196852761351109e-06,
+      "loss": 2.3657,
+      "step": 5889
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 12.506348609924316,
+      "learning_rate": 8.194848150746717e-06,
+      "loss": 1.4631,
+      "step": 5890
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 22.092987060546875,
+      "learning_rate": 8.192843540142327e-06,
+      "loss": 1.3425,
+      "step": 5891
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 24.98388671875,
+      "learning_rate": 8.190838929537938e-06,
+      "loss": 2.064,
+      "step": 5892
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 15.064332962036133,
+      "learning_rate": 8.188834318933548e-06,
+      "loss": 1.257,
+      "step": 5893
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 54.17677307128906,
+      "learning_rate": 8.186829708329158e-06,
+      "loss": 2.009,
+      "step": 5894
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 18.08331298828125,
+      "learning_rate": 8.184825097724768e-06,
+      "loss": 1.2969,
+      "step": 5895
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 32.95471954345703,
+      "learning_rate": 8.182820487120378e-06,
+      "loss": 1.2109,
+      "step": 5896
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 11.22049331665039,
+      "learning_rate": 8.180815876515988e-06,
+      "loss": 1.2192,
+      "step": 5897
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 14.255977630615234,
+      "learning_rate": 8.178811265911596e-06,
+      "loss": 1.887,
+      "step": 5898
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 12.499604225158691,
+      "learning_rate": 8.176806655307206e-06,
+      "loss": 1.3196,
+      "step": 5899
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 15.631196975708008,
+      "learning_rate": 8.174802044702818e-06,
+      "loss": 1.3063,
+      "step": 5900
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 13.534256935119629,
+      "learning_rate": 8.172797434098427e-06,
+      "loss": 1.6275,
+      "step": 5901
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 10.233455657958984,
+      "learning_rate": 8.170792823494037e-06,
+      "loss": 1.1873,
+      "step": 5902
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 11.301042556762695,
+      "learning_rate": 8.168788212889647e-06,
+      "loss": 1.2987,
+      "step": 5903
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 46.812076568603516,
+      "learning_rate": 8.166783602285257e-06,
+      "loss": 1.7821,
+      "step": 5904
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 34.08024597167969,
+      "learning_rate": 8.164778991680867e-06,
+      "loss": 1.3355,
+      "step": 5905
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 12.29112720489502,
+      "learning_rate": 8.162774381076477e-06,
+      "loss": 1.3053,
+      "step": 5906
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 16.215862274169922,
+      "learning_rate": 8.160769770472087e-06,
+      "loss": 1.6378,
+      "step": 5907
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 33.6343994140625,
+      "learning_rate": 8.158765159867697e-06,
+      "loss": 1.9444,
+      "step": 5908
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 27.786399841308594,
+      "learning_rate": 8.156760549263306e-06,
+      "loss": 1.1768,
+      "step": 5909
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 22.714481353759766,
+      "learning_rate": 8.154755938658916e-06,
+      "loss": 1.5109,
+      "step": 5910
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 14.203557968139648,
+      "learning_rate": 8.152751328054526e-06,
+      "loss": 1.2664,
+      "step": 5911
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 25.13650894165039,
+      "learning_rate": 8.150746717450136e-06,
+      "loss": 1.6833,
+      "step": 5912
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 8.750547409057617,
+      "learning_rate": 8.148742106845746e-06,
+      "loss": 0.7758,
+      "step": 5913
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 13.47951889038086,
+      "learning_rate": 8.146737496241356e-06,
+      "loss": 0.9706,
+      "step": 5914
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 25.456212997436523,
+      "learning_rate": 8.144732885636966e-06,
+      "loss": 2.0665,
+      "step": 5915
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 26.591684341430664,
+      "learning_rate": 8.142728275032575e-06,
+      "loss": 2.694,
+      "step": 5916
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 141.93113708496094,
+      "learning_rate": 8.140723664428185e-06,
+      "loss": 2.0094,
+      "step": 5917
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 14.334453582763672,
+      "learning_rate": 8.138719053823796e-06,
+      "loss": 1.156,
+      "step": 5918
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 31.279136657714844,
+      "learning_rate": 8.136714443219406e-06,
+      "loss": 1.2964,
+      "step": 5919
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 13.835858345031738,
+      "learning_rate": 8.134709832615015e-06,
+      "loss": 0.992,
+      "step": 5920
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 20.324710845947266,
+      "learning_rate": 8.132705222010625e-06,
+      "loss": 2.0425,
+      "step": 5921
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 9.181856155395508,
+      "learning_rate": 8.130700611406235e-06,
+      "loss": 1.2267,
+      "step": 5922
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 12.382397651672363,
+      "learning_rate": 8.128696000801845e-06,
+      "loss": 1.4096,
+      "step": 5923
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 19.25230598449707,
+      "learning_rate": 8.126691390197455e-06,
+      "loss": 1.86,
+      "step": 5924
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 9.224576950073242,
+      "learning_rate": 8.124686779593065e-06,
+      "loss": 1.291,
+      "step": 5925
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 26.03525161743164,
+      "learning_rate": 8.122682168988675e-06,
+      "loss": 2.0772,
+      "step": 5926
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 12.974356651306152,
+      "learning_rate": 8.120677558384284e-06,
+      "loss": 1.524,
+      "step": 5927
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 9.397273063659668,
+      "learning_rate": 8.118672947779894e-06,
+      "loss": 1.2373,
+      "step": 5928
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 12.543724060058594,
+      "learning_rate": 8.116668337175504e-06,
+      "loss": 1.4601,
+      "step": 5929
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 10.495887756347656,
+      "learning_rate": 8.114663726571114e-06,
+      "loss": 1.3694,
+      "step": 5930
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 12.478328704833984,
+      "learning_rate": 8.112659115966724e-06,
+      "loss": 1.8051,
+      "step": 5931
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 18.770479202270508,
+      "learning_rate": 8.110654505362334e-06,
+      "loss": 1.3905,
+      "step": 5932
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 14.52056884765625,
+      "learning_rate": 8.108649894757944e-06,
+      "loss": 1.4971,
+      "step": 5933
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 25.12417221069336,
+      "learning_rate": 8.106645284153554e-06,
+      "loss": 1.9117,
+      "step": 5934
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 40.989715576171875,
+      "learning_rate": 8.104640673549163e-06,
+      "loss": 2.4517,
+      "step": 5935
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 7.442935943603516,
+      "learning_rate": 8.102636062944773e-06,
+      "loss": 1.0286,
+      "step": 5936
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 13.013799667358398,
+      "learning_rate": 8.100631452340385e-06,
+      "loss": 1.1536,
+      "step": 5937
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 10.014057159423828,
+      "learning_rate": 8.098626841735993e-06,
+      "loss": 1.193,
+      "step": 5938
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 24.215864181518555,
+      "learning_rate": 8.096622231131603e-06,
+      "loss": 1.3133,
+      "step": 5939
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 14.128569602966309,
+      "learning_rate": 8.094617620527213e-06,
+      "loss": 1.0683,
+      "step": 5940
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 18.90635108947754,
+      "learning_rate": 8.092613009922823e-06,
+      "loss": 1.3165,
+      "step": 5941
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 7.926889896392822,
+      "learning_rate": 8.090608399318433e-06,
+      "loss": 1.3542,
+      "step": 5942
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 15.462095260620117,
+      "learning_rate": 8.088603788714043e-06,
+      "loss": 1.7081,
+      "step": 5943
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 13.095417976379395,
+      "learning_rate": 8.086599178109653e-06,
+      "loss": 1.3562,
+      "step": 5944
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 16.457101821899414,
+      "learning_rate": 8.084594567505264e-06,
+      "loss": 1.1891,
+      "step": 5945
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 9.37852668762207,
+      "learning_rate": 8.082589956900872e-06,
+      "loss": 1.3149,
+      "step": 5946
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 10.058578491210938,
+      "learning_rate": 8.080585346296482e-06,
+      "loss": 1.2129,
+      "step": 5947
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 15.420865058898926,
+      "learning_rate": 8.078580735692092e-06,
+      "loss": 2.2392,
+      "step": 5948
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 17.6695556640625,
+      "learning_rate": 8.076576125087702e-06,
+      "loss": 1.1624,
+      "step": 5949
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 12.30250072479248,
+      "learning_rate": 8.074571514483312e-06,
+      "loss": 1.4646,
+      "step": 5950
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 26.107072830200195,
+      "learning_rate": 8.072566903878922e-06,
+      "loss": 1.4681,
+      "step": 5951
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 16.788949966430664,
+      "learning_rate": 8.070562293274532e-06,
+      "loss": 1.3324,
+      "step": 5952
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 43.306427001953125,
+      "learning_rate": 8.068557682670143e-06,
+      "loss": 1.6477,
+      "step": 5953
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 14.007320404052734,
+      "learning_rate": 8.066553072065751e-06,
+      "loss": 0.9753,
+      "step": 5954
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 9.967177391052246,
+      "learning_rate": 8.064548461461363e-06,
+      "loss": 0.8905,
+      "step": 5955
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 63.55514144897461,
+      "learning_rate": 8.062543850856973e-06,
+      "loss": 2.9485,
+      "step": 5956
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 18.12586212158203,
+      "learning_rate": 8.060539240252581e-06,
+      "loss": 1.9415,
+      "step": 5957
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 9.148773193359375,
+      "learning_rate": 8.058534629648191e-06,
+      "loss": 1.476,
+      "step": 5958
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 10.708128929138184,
+      "learning_rate": 8.056530019043801e-06,
+      "loss": 1.3661,
+      "step": 5959
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 20.42496681213379,
+      "learning_rate": 8.054525408439411e-06,
+      "loss": 1.4746,
+      "step": 5960
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 15.104199409484863,
+      "learning_rate": 8.052520797835022e-06,
+      "loss": 1.8083,
+      "step": 5961
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 20.05451011657715,
+      "learning_rate": 8.050516187230632e-06,
+      "loss": 1.6234,
+      "step": 5962
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 23.32365608215332,
+      "learning_rate": 8.048511576626242e-06,
+      "loss": 1.5622,
+      "step": 5963
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 17.466800689697266,
+      "learning_rate": 8.04650696602185e-06,
+      "loss": 0.6051,
+      "step": 5964
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 26.45103645324707,
+      "learning_rate": 8.04450235541746e-06,
+      "loss": 1.5636,
+      "step": 5965
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 25.677671432495117,
+      "learning_rate": 8.04249774481307e-06,
+      "loss": 1.1426,
+      "step": 5966
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 17.478527069091797,
+      "learning_rate": 8.04049313420868e-06,
+      "loss": 0.8924,
+      "step": 5967
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 15.59758472442627,
+      "learning_rate": 8.03848852360429e-06,
+      "loss": 1.4224,
+      "step": 5968
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 20.508508682250977,
+      "learning_rate": 8.0364839129999e-06,
+      "loss": 2.0134,
+      "step": 5969
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 14.481962203979492,
+      "learning_rate": 8.03447930239551e-06,
+      "loss": 1.176,
+      "step": 5970
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 17.673538208007812,
+      "learning_rate": 8.03247469179112e-06,
+      "loss": 1.5274,
+      "step": 5971
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 16.05086898803711,
+      "learning_rate": 8.030470081186729e-06,
+      "loss": 1.2769,
+      "step": 5972
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 26.023082733154297,
+      "learning_rate": 8.028465470582339e-06,
+      "loss": 0.8421,
+      "step": 5973
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 19.35530662536621,
+      "learning_rate": 8.026460859977951e-06,
+      "loss": 2.123,
+      "step": 5974
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 45.65658950805664,
+      "learning_rate": 8.02445624937356e-06,
+      "loss": 2.4255,
+      "step": 5975
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 50.98988342285156,
+      "learning_rate": 8.02245163876917e-06,
+      "loss": 2.1818,
+      "step": 5976
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 25.256816864013672,
+      "learning_rate": 8.02044702816478e-06,
+      "loss": 2.182,
+      "step": 5977
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 18.960912704467773,
+      "learning_rate": 8.01844241756039e-06,
+      "loss": 1.4124,
+      "step": 5978
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 20.172710418701172,
+      "learning_rate": 8.016437806956e-06,
+      "loss": 0.64,
+      "step": 5979
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 16.025949478149414,
+      "learning_rate": 8.01443319635161e-06,
+      "loss": 1.1168,
+      "step": 5980
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 19.96694564819336,
+      "learning_rate": 8.01242858574722e-06,
+      "loss": 0.9764,
+      "step": 5981
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 28.706392288208008,
+      "learning_rate": 8.01042397514283e-06,
+      "loss": 1.6412,
+      "step": 5982
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 9.621232032775879,
+      "learning_rate": 8.008419364538438e-06,
+      "loss": 0.9458,
+      "step": 5983
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 50.64955520629883,
+      "learning_rate": 8.006414753934048e-06,
+      "loss": 1.558,
+      "step": 5984
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 54.46828842163086,
+      "learning_rate": 8.004410143329658e-06,
+      "loss": 2.4981,
+      "step": 5985
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 41.00409698486328,
+      "learning_rate": 8.002405532725269e-06,
+      "loss": 1.5329,
+      "step": 5986
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 14.016818046569824,
+      "learning_rate": 8.000400922120879e-06,
+      "loss": 1.4779,
+      "step": 5987
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 46.308258056640625,
+      "learning_rate": 7.998396311516489e-06,
+      "loss": 1.1857,
+      "step": 5988
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 10.816703796386719,
+      "learning_rate": 7.996391700912099e-06,
+      "loss": 1.2238,
+      "step": 5989
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 14.56143856048584,
+      "learning_rate": 7.994387090307709e-06,
+      "loss": 1.5197,
+      "step": 5990
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 13.148183822631836,
+      "learning_rate": 7.992382479703317e-06,
+      "loss": 1.0046,
+      "step": 5991
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 54.134552001953125,
+      "learning_rate": 7.990377869098929e-06,
+      "loss": 1.8494,
+      "step": 5992
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 12.82621955871582,
+      "learning_rate": 7.988373258494539e-06,
+      "loss": 1.4167,
+      "step": 5993
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 14.938492774963379,
+      "learning_rate": 7.986368647890148e-06,
+      "loss": 0.6812,
+      "step": 5994
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 39.79289627075195,
+      "learning_rate": 7.984364037285758e-06,
+      "loss": 4.1951,
+      "step": 5995
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 12.317481994628906,
+      "learning_rate": 7.982359426681368e-06,
+      "loss": 1.0172,
+      "step": 5996
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 28.569238662719727,
+      "learning_rate": 7.980354816076978e-06,
+      "loss": 2.539,
+      "step": 5997
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 28.877849578857422,
+      "learning_rate": 7.978350205472588e-06,
+      "loss": 1.4473,
+      "step": 5998
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 42.884979248046875,
+      "learning_rate": 7.976345594868198e-06,
+      "loss": 0.6649,
+      "step": 5999
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 8.961115837097168,
+      "learning_rate": 7.974340984263808e-06,
+      "loss": 1.1251,
+      "step": 6000
+    },
+    {
+      "epoch": 1.8,
+      "eval_loss": 0.19368119537830353,
+      "eval_runtime": 43.5079,
+      "eval_samples_per_second": 33.994,
+      "eval_steps_per_second": 33.994,
+      "step": 6000
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 72.2662124633789,
+      "learning_rate": 7.972336373659416e-06,
+      "loss": 2.015,
+      "step": 6001
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 16.20361328125,
+      "learning_rate": 7.970331763055027e-06,
+      "loss": 1.2574,
+      "step": 6002
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 20.717201232910156,
+      "learning_rate": 7.968327152450637e-06,
+      "loss": 1.5776,
+      "step": 6003
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 23.648340225219727,
+      "learning_rate": 7.966322541846248e-06,
+      "loss": 1.8967,
+      "step": 6004
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 28.407695770263672,
+      "learning_rate": 7.964317931241857e-06,
+      "loss": 2.1085,
+      "step": 6005
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 13.047331809997559,
+      "learning_rate": 7.962313320637467e-06,
+      "loss": 1.9436,
+      "step": 6006
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 12.131548881530762,
+      "learning_rate": 7.960308710033077e-06,
+      "loss": 1.3458,
+      "step": 6007
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 15.92016315460205,
+      "learning_rate": 7.958304099428687e-06,
+      "loss": 0.8177,
+      "step": 6008
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 9.055414199829102,
+      "learning_rate": 7.956299488824295e-06,
+      "loss": 1.3297,
+      "step": 6009
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 13.94730281829834,
+      "learning_rate": 7.954294878219907e-06,
+      "loss": 1.4445,
+      "step": 6010
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 17.391616821289062,
+      "learning_rate": 7.952290267615517e-06,
+      "loss": 0.9038,
+      "step": 6011
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 198.21731567382812,
+      "learning_rate": 7.950285657011126e-06,
+      "loss": 2.3957,
+      "step": 6012
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 12.243670463562012,
+      "learning_rate": 7.948281046406736e-06,
+      "loss": 1.4155,
+      "step": 6013
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 13.951066017150879,
+      "learning_rate": 7.946276435802346e-06,
+      "loss": 1.4097,
+      "step": 6014
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 24.333097457885742,
+      "learning_rate": 7.944271825197956e-06,
+      "loss": 2.4542,
+      "step": 6015
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 17.491724014282227,
+      "learning_rate": 7.942267214593566e-06,
+      "loss": 1.649,
+      "step": 6016
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 11.039308547973633,
+      "learning_rate": 7.940262603989176e-06,
+      "loss": 1.1686,
+      "step": 6017
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 18.201019287109375,
+      "learning_rate": 7.938257993384786e-06,
+      "loss": 1.5321,
+      "step": 6018
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 29.47193717956543,
+      "learning_rate": 7.936253382780396e-06,
+      "loss": 1.5978,
+      "step": 6019
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 24.67177391052246,
+      "learning_rate": 7.934248772176005e-06,
+      "loss": 2.0138,
+      "step": 6020
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 25.735088348388672,
+      "learning_rate": 7.932244161571615e-06,
+      "loss": 3.0204,
+      "step": 6021
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 63.19715118408203,
+      "learning_rate": 7.930239550967225e-06,
+      "loss": 1.5556,
+      "step": 6022
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 12.563130378723145,
+      "learning_rate": 7.928234940362835e-06,
+      "loss": 1.3196,
+      "step": 6023
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 21.176742553710938,
+      "learning_rate": 7.926230329758445e-06,
+      "loss": 2.416,
+      "step": 6024
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 14.44190502166748,
+      "learning_rate": 7.924225719154055e-06,
+      "loss": 1.14,
+      "step": 6025
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 10.227465629577637,
+      "learning_rate": 7.922221108549665e-06,
+      "loss": 0.9925,
+      "step": 6026
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 16.822582244873047,
+      "learning_rate": 7.920216497945275e-06,
+      "loss": 1.413,
+      "step": 6027
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 23.78660774230957,
+      "learning_rate": 7.918211887340884e-06,
+      "loss": 2.2229,
+      "step": 6028
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 22.946805953979492,
+      "learning_rate": 7.916207276736495e-06,
+      "loss": 1.9366,
+      "step": 6029
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 11.842130661010742,
+      "learning_rate": 7.914202666132105e-06,
+      "loss": 2.1797,
+      "step": 6030
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 11.936902046203613,
+      "learning_rate": 7.912198055527714e-06,
+      "loss": 0.9288,
+      "step": 6031
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 12.97545337677002,
+      "learning_rate": 7.910193444923324e-06,
+      "loss": 0.8792,
+      "step": 6032
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 22.72028160095215,
+      "learning_rate": 7.908188834318934e-06,
+      "loss": 1.8991,
+      "step": 6033
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 68.23728942871094,
+      "learning_rate": 7.906184223714544e-06,
+      "loss": 2.0705,
+      "step": 6034
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 17.14826202392578,
+      "learning_rate": 7.904179613110154e-06,
+      "loss": 1.9903,
+      "step": 6035
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 33.424381256103516,
+      "learning_rate": 7.902175002505764e-06,
+      "loss": 1.5851,
+      "step": 6036
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 21.694814682006836,
+      "learning_rate": 7.900170391901374e-06,
+      "loss": 2.1811,
+      "step": 6037
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 11.936399459838867,
+      "learning_rate": 7.898165781296984e-06,
+      "loss": 1.4652,
+      "step": 6038
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 10.102071762084961,
+      "learning_rate": 7.896161170692593e-06,
+      "loss": 1.9421,
+      "step": 6039
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 32.03966522216797,
+      "learning_rate": 7.894156560088203e-06,
+      "loss": 1.4121,
+      "step": 6040
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 31.259422302246094,
+      "learning_rate": 7.892151949483815e-06,
+      "loss": 1.255,
+      "step": 6041
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 11.89094352722168,
+      "learning_rate": 7.890147338879423e-06,
+      "loss": 1.1638,
+      "step": 6042
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 16.84495735168457,
+      "learning_rate": 7.888142728275033e-06,
+      "loss": 1.4586,
+      "step": 6043
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 9.196840286254883,
+      "learning_rate": 7.886138117670643e-06,
+      "loss": 2.2115,
+      "step": 6044
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 20.001394271850586,
+      "learning_rate": 7.884133507066253e-06,
+      "loss": 1.0957,
+      "step": 6045
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 30.092775344848633,
+      "learning_rate": 7.882128896461862e-06,
+      "loss": 1.6151,
+      "step": 6046
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 10.02807331085205,
+      "learning_rate": 7.880124285857474e-06,
+      "loss": 0.872,
+      "step": 6047
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 21.552358627319336,
+      "learning_rate": 7.878119675253084e-06,
+      "loss": 1.4557,
+      "step": 6048
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 21.00843620300293,
+      "learning_rate": 7.876115064648692e-06,
+      "loss": 1.379,
+      "step": 6049
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 16.021032333374023,
+      "learning_rate": 7.874110454044302e-06,
+      "loss": 1.8451,
+      "step": 6050
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 14.853699684143066,
+      "learning_rate": 7.872105843439912e-06,
+      "loss": 1.125,
+      "step": 6051
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 20.046295166015625,
+      "learning_rate": 7.870101232835522e-06,
+      "loss": 1.9366,
+      "step": 6052
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 17.23594856262207,
+      "learning_rate": 7.868096622231132e-06,
+      "loss": 1.6949,
+      "step": 6053
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 16.25902557373047,
+      "learning_rate": 7.866092011626742e-06,
+      "loss": 1.9581,
+      "step": 6054
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 25.339801788330078,
+      "learning_rate": 7.864087401022353e-06,
+      "loss": 1.4696,
+      "step": 6055
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 45.478515625,
+      "learning_rate": 7.862082790417963e-06,
+      "loss": 2.0867,
+      "step": 6056
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 19.254344940185547,
+      "learning_rate": 7.860078179813571e-06,
+      "loss": 1.6101,
+      "step": 6057
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 23.987213134765625,
+      "learning_rate": 7.858073569209181e-06,
+      "loss": 1.1053,
+      "step": 6058
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 33.420997619628906,
+      "learning_rate": 7.856068958604791e-06,
+      "loss": 1.3957,
+      "step": 6059
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 53.56259536743164,
+      "learning_rate": 7.854064348000401e-06,
+      "loss": 2.8107,
+      "step": 6060
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 14.887017250061035,
+      "learning_rate": 7.852059737396011e-06,
+      "loss": 2.4886,
+      "step": 6061
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 15.285018920898438,
+      "learning_rate": 7.850055126791621e-06,
+      "loss": 1.1363,
+      "step": 6062
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 16.956491470336914,
+      "learning_rate": 7.848050516187231e-06,
+      "loss": 1.4113,
+      "step": 6063
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 22.227479934692383,
+      "learning_rate": 7.846045905582842e-06,
+      "loss": 1.2002,
+      "step": 6064
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 44.9412727355957,
+      "learning_rate": 7.84404129497845e-06,
+      "loss": 1.583,
+      "step": 6065
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 37.56648254394531,
+      "learning_rate": 7.842036684374062e-06,
+      "loss": 2.022,
+      "step": 6066
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 15.846146583557129,
+      "learning_rate": 7.840032073769672e-06,
+      "loss": 0.9523,
+      "step": 6067
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 18.317182540893555,
+      "learning_rate": 7.83802746316528e-06,
+      "loss": 1.5857,
+      "step": 6068
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 14.531113624572754,
+      "learning_rate": 7.83602285256089e-06,
+      "loss": 1.2722,
+      "step": 6069
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 9.223918914794922,
+      "learning_rate": 7.8340182419565e-06,
+      "loss": 1.6124,
+      "step": 6070
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 12.137374877929688,
+      "learning_rate": 7.83201363135211e-06,
+      "loss": 1.0094,
+      "step": 6071
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 8.65196418762207,
+      "learning_rate": 7.83000902074772e-06,
+      "loss": 0.9413,
+      "step": 6072
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 16.220958709716797,
+      "learning_rate": 7.82800441014333e-06,
+      "loss": 2.1056,
+      "step": 6073
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 19.515766143798828,
+      "learning_rate": 7.82599979953894e-06,
+      "loss": 1.763,
+      "step": 6074
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 11.028536796569824,
+      "learning_rate": 7.82399518893455e-06,
+      "loss": 1.577,
+      "step": 6075
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 14.080721855163574,
+      "learning_rate": 7.82199057833016e-06,
+      "loss": 1.2928,
+      "step": 6076
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 22.41744041442871,
+      "learning_rate": 7.81998596772577e-06,
+      "loss": 1.6717,
+      "step": 6077
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 29.117725372314453,
+      "learning_rate": 7.817981357121381e-06,
+      "loss": 2.0723,
+      "step": 6078
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 10.062914848327637,
+      "learning_rate": 7.81597674651699e-06,
+      "loss": 0.9723,
+      "step": 6079
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 19.473957061767578,
+      "learning_rate": 7.8139721359126e-06,
+      "loss": 1.6513,
+      "step": 6080
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 13.241780281066895,
+      "learning_rate": 7.81196752530821e-06,
+      "loss": 1.4396,
+      "step": 6081
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 50.16336441040039,
+      "learning_rate": 7.80996291470382e-06,
+      "loss": 1.4343,
+      "step": 6082
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 8.716567039489746,
+      "learning_rate": 7.807958304099428e-06,
+      "loss": 0.7952,
+      "step": 6083
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 17.933069229125977,
+      "learning_rate": 7.80595369349504e-06,
+      "loss": 2.706,
+      "step": 6084
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 30.77072525024414,
+      "learning_rate": 7.80394908289065e-06,
+      "loss": 1.9985,
+      "step": 6085
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 13.682772636413574,
+      "learning_rate": 7.801944472286258e-06,
+      "loss": 0.9191,
+      "step": 6086
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 17.457841873168945,
+      "learning_rate": 7.799939861681868e-06,
+      "loss": 1.518,
+      "step": 6087
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 41.92538833618164,
+      "learning_rate": 7.797935251077479e-06,
+      "loss": 2.1136,
+      "step": 6088
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 22.720664978027344,
+      "learning_rate": 7.795930640473089e-06,
+      "loss": 1.8965,
+      "step": 6089
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 13.168417930603027,
+      "learning_rate": 7.793926029868699e-06,
+      "loss": 1.3237,
+      "step": 6090
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 22.606231689453125,
+      "learning_rate": 7.791921419264309e-06,
+      "loss": 1.6433,
+      "step": 6091
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 22.04924964904785,
+      "learning_rate": 7.789916808659919e-06,
+      "loss": 1.9051,
+      "step": 6092
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 12.943755149841309,
+      "learning_rate": 7.787912198055529e-06,
+      "loss": 1.3566,
+      "step": 6093
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 8.388080596923828,
+      "learning_rate": 7.785907587451137e-06,
+      "loss": 1.0533,
+      "step": 6094
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 39.58984375,
+      "learning_rate": 7.783902976846747e-06,
+      "loss": 2.1753,
+      "step": 6095
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 17.842021942138672,
+      "learning_rate": 7.781898366242357e-06,
+      "loss": 0.9146,
+      "step": 6096
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 12.273473739624023,
+      "learning_rate": 7.779893755637968e-06,
+      "loss": 1.0507,
+      "step": 6097
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 34.619178771972656,
+      "learning_rate": 7.777889145033578e-06,
+      "loss": 2.2011,
+      "step": 6098
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 87.12923431396484,
+      "learning_rate": 7.775884534429188e-06,
+      "loss": 1.4831,
+      "step": 6099
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 9.995388984680176,
+      "learning_rate": 7.773879923824798e-06,
+      "loss": 1.2102,
+      "step": 6100
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 13.057961463928223,
+      "learning_rate": 7.771875313220408e-06,
+      "loss": 1.3872,
+      "step": 6101
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 8.191656112670898,
+      "learning_rate": 7.769870702616016e-06,
+      "loss": 1.3387,
+      "step": 6102
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 25.005573272705078,
+      "learning_rate": 7.767866092011628e-06,
+      "loss": 1.393,
+      "step": 6103
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 11.448406219482422,
+      "learning_rate": 7.765861481407238e-06,
+      "loss": 1.2107,
+      "step": 6104
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 10.444814682006836,
+      "learning_rate": 7.763856870802847e-06,
+      "loss": 1.1085,
+      "step": 6105
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 22.36345863342285,
+      "learning_rate": 7.761852260198457e-06,
+      "loss": 1.4649,
+      "step": 6106
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 12.638916015625,
+      "learning_rate": 7.759847649594067e-06,
+      "loss": 1.0798,
+      "step": 6107
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 30.909666061401367,
+      "learning_rate": 7.757843038989677e-06,
+      "loss": 1.3542,
+      "step": 6108
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 21.984508514404297,
+      "learning_rate": 7.755838428385287e-06,
+      "loss": 1.5787,
+      "step": 6109
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 28.869102478027344,
+      "learning_rate": 7.753833817780897e-06,
+      "loss": 1.8047,
+      "step": 6110
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 8.794825553894043,
+      "learning_rate": 7.751829207176507e-06,
+      "loss": 0.4721,
+      "step": 6111
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 20.192941665649414,
+      "learning_rate": 7.749824596572117e-06,
+      "loss": 1.8928,
+      "step": 6112
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 8.28946304321289,
+      "learning_rate": 7.747819985967726e-06,
+      "loss": 0.8887,
+      "step": 6113
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 7.44697380065918,
+      "learning_rate": 7.745815375363336e-06,
+      "loss": 1.0449,
+      "step": 6114
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 47.45841598510742,
+      "learning_rate": 7.743810764758947e-06,
+      "loss": 1.8076,
+      "step": 6115
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 22.896045684814453,
+      "learning_rate": 7.741806154154556e-06,
+      "loss": 1.3109,
+      "step": 6116
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 51.84198760986328,
+      "learning_rate": 7.739801543550166e-06,
+      "loss": 1.9647,
+      "step": 6117
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 16.364362716674805,
+      "learning_rate": 7.737796932945776e-06,
+      "loss": 1.6399,
+      "step": 6118
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 14.368910789489746,
+      "learning_rate": 7.735792322341386e-06,
+      "loss": 1.0631,
+      "step": 6119
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 11.560124397277832,
+      "learning_rate": 7.733787711736994e-06,
+      "loss": 1.6401,
+      "step": 6120
+    },
+    {
+      "epoch": 1.84,
+      "eval_loss": 0.1900317370891571,
+      "eval_runtime": 43.601,
+      "eval_samples_per_second": 33.921,
+      "eval_steps_per_second": 33.921,
+      "step": 6120
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 11.354382514953613,
+      "learning_rate": 7.731783101132606e-06,
+      "loss": 1.3231,
+      "step": 6121
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 8.842697143554688,
+      "learning_rate": 7.729778490528216e-06,
+      "loss": 0.5822,
+      "step": 6122
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 44.98017501831055,
+      "learning_rate": 7.727773879923826e-06,
+      "loss": 1.6967,
+      "step": 6123
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 29.9891414642334,
+      "learning_rate": 7.725769269319435e-06,
+      "loss": 1.5777,
+      "step": 6124
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 25.591136932373047,
+      "learning_rate": 7.723764658715045e-06,
+      "loss": 0.9665,
+      "step": 6125
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 13.819255828857422,
+      "learning_rate": 7.721760048110655e-06,
+      "loss": 2.2739,
+      "step": 6126
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 28.475740432739258,
+      "learning_rate": 7.719755437506265e-06,
+      "loss": 2.247,
+      "step": 6127
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 12.238512992858887,
+      "learning_rate": 7.717750826901875e-06,
+      "loss": 0.8953,
+      "step": 6128
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 35.80490493774414,
+      "learning_rate": 7.715746216297485e-06,
+      "loss": 3.4695,
+      "step": 6129
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 7.781877040863037,
+      "learning_rate": 7.713741605693095e-06,
+      "loss": 1.8642,
+      "step": 6130
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 13.301283836364746,
+      "learning_rate": 7.711736995088704e-06,
+      "loss": 1.3696,
+      "step": 6131
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 10.512940406799316,
+      "learning_rate": 7.709732384484314e-06,
+      "loss": 0.6859,
+      "step": 6132
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 25.800260543823242,
+      "learning_rate": 7.707727773879926e-06,
+      "loss": 1.9089,
+      "step": 6133
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 15.625998497009277,
+      "learning_rate": 7.705723163275534e-06,
+      "loss": 1.8234,
+      "step": 6134
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 42.901615142822266,
+      "learning_rate": 7.703718552671144e-06,
+      "loss": 2.0355,
+      "step": 6135
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 15.231833457946777,
+      "learning_rate": 7.701713942066754e-06,
+      "loss": 1.1297,
+      "step": 6136
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 21.943153381347656,
+      "learning_rate": 7.699709331462364e-06,
+      "loss": 1.2517,
+      "step": 6137
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 22.251258850097656,
+      "learning_rate": 7.697704720857974e-06,
+      "loss": 1.5025,
+      "step": 6138
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 15.252081871032715,
+      "learning_rate": 7.695700110253583e-06,
+      "loss": 1.0183,
+      "step": 6139
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 26.634531021118164,
+      "learning_rate": 7.693695499649194e-06,
+      "loss": 1.5842,
+      "step": 6140
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 29.84535789489746,
+      "learning_rate": 7.691690889044805e-06,
+      "loss": 1.862,
+      "step": 6141
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 24.699621200561523,
+      "learning_rate": 7.689686278440413e-06,
+      "loss": 1.8952,
+      "step": 6142
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 26.514081954956055,
+      "learning_rate": 7.687681667836023e-06,
+      "loss": 1.948,
+      "step": 6143
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 11.536233901977539,
+      "learning_rate": 7.685677057231633e-06,
+      "loss": 2.2692,
+      "step": 6144
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 30.69923973083496,
+      "learning_rate": 7.683672446627243e-06,
+      "loss": 1.2475,
+      "step": 6145
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 19.856365203857422,
+      "learning_rate": 7.681667836022853e-06,
+      "loss": 1.361,
+      "step": 6146
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 10.169954299926758,
+      "learning_rate": 7.679663225418463e-06,
+      "loss": 1.3548,
+      "step": 6147
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 36.704524993896484,
+      "learning_rate": 7.677658614814073e-06,
+      "loss": 1.8048,
+      "step": 6148
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 14.586971282958984,
+      "learning_rate": 7.675654004209683e-06,
+      "loss": 1.6924,
+      "step": 6149
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 31.322877883911133,
+      "learning_rate": 7.673649393605292e-06,
+      "loss": 2.1292,
+      "step": 6150
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 13.903470993041992,
+      "learning_rate": 7.671644783000902e-06,
+      "loss": 1.5199,
+      "step": 6151
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 206.5603485107422,
+      "learning_rate": 7.669640172396514e-06,
+      "loss": 2.1998,
+      "step": 6152
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 42.4091682434082,
+      "learning_rate": 7.667635561792122e-06,
+      "loss": 1.8714,
+      "step": 6153
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 10.952982902526855,
+      "learning_rate": 7.665630951187732e-06,
+      "loss": 1.7223,
+      "step": 6154
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 12.478805541992188,
+      "learning_rate": 7.663626340583342e-06,
+      "loss": 0.9292,
+      "step": 6155
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 45.16120910644531,
+      "learning_rate": 7.661621729978952e-06,
+      "loss": 1.1461,
+      "step": 6156
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 11.00290298461914,
+      "learning_rate": 7.659617119374562e-06,
+      "loss": 1.3505,
+      "step": 6157
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 12.987632751464844,
+      "learning_rate": 7.657612508770173e-06,
+      "loss": 1.1632,
+      "step": 6158
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 23.716419219970703,
+      "learning_rate": 7.655607898165783e-06,
+      "loss": 1.9068,
+      "step": 6159
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 30.917993545532227,
+      "learning_rate": 7.653603287561393e-06,
+      "loss": 1.9522,
+      "step": 6160
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 9.036660194396973,
+      "learning_rate": 7.651598676957001e-06,
+      "loss": 1.1854,
+      "step": 6161
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 53.36625289916992,
+      "learning_rate": 7.649594066352611e-06,
+      "loss": 1.6606,
+      "step": 6162
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 23.467453002929688,
+      "learning_rate": 7.647589455748221e-06,
+      "loss": 1.7637,
+      "step": 6163
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 16.767295837402344,
+      "learning_rate": 7.645584845143831e-06,
+      "loss": 1.3892,
+      "step": 6164
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 21.95707893371582,
+      "learning_rate": 7.643580234539441e-06,
+      "loss": 2.415,
+      "step": 6165
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 14.439309120178223,
+      "learning_rate": 7.641575623935052e-06,
+      "loss": 1.575,
+      "step": 6166
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 8.983049392700195,
+      "learning_rate": 7.639571013330662e-06,
+      "loss": 1.2169,
+      "step": 6167
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 15.929500579833984,
+      "learning_rate": 7.63756640272627e-06,
+      "loss": 1.3968,
+      "step": 6168
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 16.895946502685547,
+      "learning_rate": 7.63556179212188e-06,
+      "loss": 1.773,
+      "step": 6169
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 11.67673110961914,
+      "learning_rate": 7.633557181517492e-06,
+      "loss": 0.8359,
+      "step": 6170
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 12.90839958190918,
+      "learning_rate": 7.6315525709131e-06,
+      "loss": 1.2435,
+      "step": 6171
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 50.556400299072266,
+      "learning_rate": 7.62954796030871e-06,
+      "loss": 3.8311,
+      "step": 6172
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 9.137083053588867,
+      "learning_rate": 7.6275433497043204e-06,
+      "loss": 0.9147,
+      "step": 6173
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 39.34935760498047,
+      "learning_rate": 7.6255387390999305e-06,
+      "loss": 2.6975,
+      "step": 6174
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 14.056702613830566,
+      "learning_rate": 7.62353412849554e-06,
+      "loss": 1.3314,
+      "step": 6175
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 14.793180465698242,
+      "learning_rate": 7.621529517891151e-06,
+      "loss": 1.8193,
+      "step": 6176
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 6.99417781829834,
+      "learning_rate": 7.619524907286761e-06,
+      "loss": 0.9592,
+      "step": 6177
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 8.927657127380371,
+      "learning_rate": 7.61752029668237e-06,
+      "loss": 0.7251,
+      "step": 6178
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 22.89257049560547,
+      "learning_rate": 7.61551568607798e-06,
+      "loss": 2.0944,
+      "step": 6179
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 33.927913665771484,
+      "learning_rate": 7.613511075473589e-06,
+      "loss": 2.6198,
+      "step": 6180
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 18.200057983398438,
+      "learning_rate": 7.611506464869199e-06,
+      "loss": 1.3679,
+      "step": 6181
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 17.662506103515625,
+      "learning_rate": 7.6095018542648095e-06,
+      "loss": 2.0992,
+      "step": 6182
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 11.235939979553223,
+      "learning_rate": 7.60749724366042e-06,
+      "loss": 1.4116,
+      "step": 6183
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 19.397539138793945,
+      "learning_rate": 7.60549263305603e-06,
+      "loss": 1.3618,
+      "step": 6184
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 10.240893363952637,
+      "learning_rate": 7.60348802245164e-06,
+      "loss": 1.0369,
+      "step": 6185
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 14.981084823608398,
+      "learning_rate": 7.601483411847249e-06,
+      "loss": 0.8672,
+      "step": 6186
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 12.943142890930176,
+      "learning_rate": 7.599478801242859e-06,
+      "loss": 1.19,
+      "step": 6187
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 15.80017375946045,
+      "learning_rate": 7.597474190638468e-06,
+      "loss": 1.2749,
+      "step": 6188
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 15.71901798248291,
+      "learning_rate": 7.595469580034079e-06,
+      "loss": 1.0749,
+      "step": 6189
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 13.598134994506836,
+      "learning_rate": 7.593464969429689e-06,
+      "loss": 1.2334,
+      "step": 6190
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 13.935446739196777,
+      "learning_rate": 7.5914603588252986e-06,
+      "loss": 1.2212,
+      "step": 6191
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 28.57442855834961,
+      "learning_rate": 7.589455748220909e-06,
+      "loss": 2.0672,
+      "step": 6192
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 35.023197174072266,
+      "learning_rate": 7.587451137616518e-06,
+      "loss": 1.8612,
+      "step": 6193
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 11.907183647155762,
+      "learning_rate": 7.585446527012128e-06,
+      "loss": 1.6789,
+      "step": 6194
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 18.430017471313477,
+      "learning_rate": 7.583441916407739e-06,
+      "loss": 1.3619,
+      "step": 6195
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 29.030160903930664,
+      "learning_rate": 7.581437305803349e-06,
+      "loss": 1.116,
+      "step": 6196
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 18.958419799804688,
+      "learning_rate": 7.579432695198958e-06,
+      "loss": 1.6912,
+      "step": 6197
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 19.51006317138672,
+      "learning_rate": 7.577428084594568e-06,
+      "loss": 1.6718,
+      "step": 6198
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 13.113887786865234,
+      "learning_rate": 7.5754234739901775e-06,
+      "loss": 1.2158,
+      "step": 6199
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 13.66335678100586,
+      "learning_rate": 7.573418863385788e-06,
+      "loss": 0.8829,
+      "step": 6200
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 12.062017440795898,
+      "learning_rate": 7.5714142527813986e-06,
+      "loss": 1.7451,
+      "step": 6201
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 12.591232299804688,
+      "learning_rate": 7.569409642177008e-06,
+      "loss": 1.4194,
+      "step": 6202
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 9.007003784179688,
+      "learning_rate": 7.567405031572618e-06,
+      "loss": 0.6593,
+      "step": 6203
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 15.050545692443848,
+      "learning_rate": 7.565400420968227e-06,
+      "loss": 1.2698,
+      "step": 6204
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 53.996299743652344,
+      "learning_rate": 7.563395810363837e-06,
+      "loss": 2.5506,
+      "step": 6205
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 16.08456039428711,
+      "learning_rate": 7.561391199759447e-06,
+      "loss": 0.7814,
+      "step": 6206
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 19.180896759033203,
+      "learning_rate": 7.559386589155057e-06,
+      "loss": 2.2722,
+      "step": 6207
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 12.53034782409668,
+      "learning_rate": 7.5573819785506675e-06,
+      "loss": 1.113,
+      "step": 6208
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 12.597771644592285,
+      "learning_rate": 7.5553773679462775e-06,
+      "loss": 1.7598,
+      "step": 6209
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 8.483002662658691,
+      "learning_rate": 7.553372757341887e-06,
+      "loss": 0.8353,
+      "step": 6210
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 9.132856369018555,
+      "learning_rate": 7.551368146737497e-06,
+      "loss": 1.0596,
+      "step": 6211
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 12.06258773803711,
+      "learning_rate": 7.549363536133106e-06,
+      "loss": 1.3816,
+      "step": 6212
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 12.495207786560059,
+      "learning_rate": 7.547358925528717e-06,
+      "loss": 2.0172,
+      "step": 6213
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 15.568368911743164,
+      "learning_rate": 7.545354314924327e-06,
+      "loss": 2.3213,
+      "step": 6214
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 57.913516998291016,
+      "learning_rate": 7.543349704319936e-06,
+      "loss": 1.5673,
+      "step": 6215
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 11.138136863708496,
+      "learning_rate": 7.5413450937155464e-06,
+      "loss": 1.1155,
+      "step": 6216
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 36.35742950439453,
+      "learning_rate": 7.539340483111156e-06,
+      "loss": 0.877,
+      "step": 6217
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 11.55097484588623,
+      "learning_rate": 7.537335872506766e-06,
+      "loss": 0.539,
+      "step": 6218
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 18.280834197998047,
+      "learning_rate": 7.535331261902377e-06,
+      "loss": 1.5513,
+      "step": 6219
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 30.530216217041016,
+      "learning_rate": 7.533326651297986e-06,
+      "loss": 1.9069,
+      "step": 6220
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 28.866329193115234,
+      "learning_rate": 7.531322040693596e-06,
+      "loss": 1.4386,
+      "step": 6221
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 22.6090145111084,
+      "learning_rate": 7.529317430089206e-06,
+      "loss": 2.3103,
+      "step": 6222
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 9.455753326416016,
+      "learning_rate": 7.527312819484815e-06,
+      "loss": 1.0053,
+      "step": 6223
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 47.22819519042969,
+      "learning_rate": 7.525308208880425e-06,
+      "loss": 2.0746,
+      "step": 6224
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 16.05484962463379,
+      "learning_rate": 7.523303598276035e-06,
+      "loss": 1.4145,
+      "step": 6225
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 23.540828704833984,
+      "learning_rate": 7.521298987671646e-06,
+      "loss": 1.1681,
+      "step": 6226
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 14.86224365234375,
+      "learning_rate": 7.519294377067256e-06,
+      "loss": 1.6844,
+      "step": 6227
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 76.27838134765625,
+      "learning_rate": 7.517289766462865e-06,
+      "loss": 1.9908,
+      "step": 6228
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 18.715179443359375,
+      "learning_rate": 7.515285155858475e-06,
+      "loss": 1.2129,
+      "step": 6229
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 16.809297561645508,
+      "learning_rate": 7.513280545254085e-06,
+      "loss": 1.1011,
+      "step": 6230
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 8.646817207336426,
+      "learning_rate": 7.511275934649694e-06,
+      "loss": 1.5495,
+      "step": 6231
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 35.9256591796875,
+      "learning_rate": 7.509271324045305e-06,
+      "loss": 1.7407,
+      "step": 6232
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 8.104480743408203,
+      "learning_rate": 7.507266713440915e-06,
+      "loss": 0.5633,
+      "step": 6233
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 10.297518730163574,
+      "learning_rate": 7.5052621028365246e-06,
+      "loss": 1.2832,
+      "step": 6234
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 15.735064506530762,
+      "learning_rate": 7.503257492232135e-06,
+      "loss": 1.4141,
+      "step": 6235
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 12.001105308532715,
+      "learning_rate": 7.501252881627744e-06,
+      "loss": 1.4383,
+      "step": 6236
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 15.454205513000488,
+      "learning_rate": 7.499248271023354e-06,
+      "loss": 1.0,
+      "step": 6237
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 19.108205795288086,
+      "learning_rate": 7.497243660418965e-06,
+      "loss": 1.6183,
+      "step": 6238
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 38.10220718383789,
+      "learning_rate": 7.495239049814574e-06,
+      "loss": 2.5108,
+      "step": 6239
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 10.768503189086914,
+      "learning_rate": 7.493234439210184e-06,
+      "loss": 1.4355,
+      "step": 6240
+    },
+    {
+      "epoch": 1.88,
+      "eval_loss": 0.19893887639045715,
+      "eval_runtime": 43.6319,
+      "eval_samples_per_second": 33.897,
+      "eval_steps_per_second": 33.897,
+      "step": 6240
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 7.937902450561523,
+      "learning_rate": 7.4912298286057935e-06,
+      "loss": 0.9729,
+      "step": 6241
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 10.797032356262207,
+      "learning_rate": 7.4892252180014035e-06,
+      "loss": 0.9792,
+      "step": 6242
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 8.380309104919434,
+      "learning_rate": 7.487220607397014e-06,
+      "loss": 0.7776,
+      "step": 6243
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 10.550333023071289,
+      "learning_rate": 7.485215996792624e-06,
+      "loss": 1.1951,
+      "step": 6244
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 16.15055274963379,
+      "learning_rate": 7.483211386188234e-06,
+      "loss": 0.9995,
+      "step": 6245
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 18.002609252929688,
+      "learning_rate": 7.481206775583844e-06,
+      "loss": 1.6159,
+      "step": 6246
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 19.96839714050293,
+      "learning_rate": 7.479202164979453e-06,
+      "loss": 1.9925,
+      "step": 6247
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 22.932233810424805,
+      "learning_rate": 7.477197554375063e-06,
+      "loss": 1.5709,
+      "step": 6248
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 10.494919776916504,
+      "learning_rate": 7.4751929437706724e-06,
+      "loss": 1.1168,
+      "step": 6249
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 37.70262145996094,
+      "learning_rate": 7.473188333166283e-06,
+      "loss": 1.6439,
+      "step": 6250
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 14.26046085357666,
+      "learning_rate": 7.4711837225618935e-06,
+      "loss": 1.4012,
+      "step": 6251
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 19.355712890625,
+      "learning_rate": 7.469179111957503e-06,
+      "loss": 1.3238,
+      "step": 6252
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 43.30799865722656,
+      "learning_rate": 7.467174501353113e-06,
+      "loss": 2.3852,
+      "step": 6253
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 17.540157318115234,
+      "learning_rate": 7.465169890748723e-06,
+      "loss": 0.9667,
+      "step": 6254
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 14.730655670166016,
+      "learning_rate": 7.463165280144332e-06,
+      "loss": 1.454,
+      "step": 6255
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 11.64047908782959,
+      "learning_rate": 7.461160669539943e-06,
+      "loss": 1.209,
+      "step": 6256
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 8.219613075256348,
+      "learning_rate": 7.459156058935553e-06,
+      "loss": 1.3621,
+      "step": 6257
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 15.764402389526367,
+      "learning_rate": 7.457151448331162e-06,
+      "loss": 1.7374,
+      "step": 6258
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 17.738845825195312,
+      "learning_rate": 7.4551468377267724e-06,
+      "loss": 1.5129,
+      "step": 6259
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 18.939716339111328,
+      "learning_rate": 7.453142227122382e-06,
+      "loss": 1.6715,
+      "step": 6260
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 25.509105682373047,
+      "learning_rate": 7.451137616517992e-06,
+      "loss": 1.6497,
+      "step": 6261
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 30.341272354125977,
+      "learning_rate": 7.449133005913603e-06,
+      "loss": 2.2224,
+      "step": 6262
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 29.086957931518555,
+      "learning_rate": 7.447128395309212e-06,
+      "loss": 2.018,
+      "step": 6263
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 13.672849655151367,
+      "learning_rate": 7.445123784704822e-06,
+      "loss": 1.4355,
+      "step": 6264
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 27.003616333007812,
+      "learning_rate": 7.443119174100431e-06,
+      "loss": 1.4442,
+      "step": 6265
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 19.808364868164062,
+      "learning_rate": 7.441114563496041e-06,
+      "loss": 1.2619,
+      "step": 6266
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 17.011503219604492,
+      "learning_rate": 7.439109952891651e-06,
+      "loss": 1.2087,
+      "step": 6267
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 16.227067947387695,
+      "learning_rate": 7.437105342287261e-06,
+      "loss": 2.4985,
+      "step": 6268
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 15.978217124938965,
+      "learning_rate": 7.435100731682872e-06,
+      "loss": 1.3265,
+      "step": 6269
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 16.302669525146484,
+      "learning_rate": 7.433096121078482e-06,
+      "loss": 0.8265,
+      "step": 6270
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 20.792295455932617,
+      "learning_rate": 7.431091510474091e-06,
+      "loss": 1.4503,
+      "step": 6271
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 9.446883201599121,
+      "learning_rate": 7.429086899869701e-06,
+      "loss": 1.8997,
+      "step": 6272
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 24.59245491027832,
+      "learning_rate": 7.42708228926531e-06,
+      "loss": 1.1568,
+      "step": 6273
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 13.666716575622559,
+      "learning_rate": 7.42507767866092e-06,
+      "loss": 0.9083,
+      "step": 6274
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 17.118921279907227,
+      "learning_rate": 7.423073068056531e-06,
+      "loss": 1.5059,
+      "step": 6275
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 56.16904067993164,
+      "learning_rate": 7.4210684574521405e-06,
+      "loss": 3.6439,
+      "step": 6276
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 13.676959037780762,
+      "learning_rate": 7.4190638468477506e-06,
+      "loss": 1.6671,
+      "step": 6277
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 10.493898391723633,
+      "learning_rate": 7.41705923624336e-06,
+      "loss": 1.4543,
+      "step": 6278
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 22.156248092651367,
+      "learning_rate": 7.41505462563897e-06,
+      "loss": 1.5765,
+      "step": 6279
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 15.999430656433105,
+      "learning_rate": 7.41305001503458e-06,
+      "loss": 1.7229,
+      "step": 6280
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 10.846489906311035,
+      "learning_rate": 7.411045404430191e-06,
+      "loss": 1.9086,
+      "step": 6281
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 17.6227970123291,
+      "learning_rate": 7.4090407938258e-06,
+      "loss": 1.2638,
+      "step": 6282
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 16.64601707458496,
+      "learning_rate": 7.40703618322141e-06,
+      "loss": 1.2801,
+      "step": 6283
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 18.028026580810547,
+      "learning_rate": 7.4050315726170195e-06,
+      "loss": 1.4945,
+      "step": 6284
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 35.858795166015625,
+      "learning_rate": 7.4030269620126295e-06,
+      "loss": 1.969,
+      "step": 6285
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 9.260512351989746,
+      "learning_rate": 7.401022351408239e-06,
+      "loss": 1.0171,
+      "step": 6286
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 28.248775482177734,
+      "learning_rate": 7.39901774080385e-06,
+      "loss": 1.2457,
+      "step": 6287
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 23.339683532714844,
+      "learning_rate": 7.39701313019946e-06,
+      "loss": 1.8959,
+      "step": 6288
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 16.68250274658203,
+      "learning_rate": 7.395008519595069e-06,
+      "loss": 1.1061,
+      "step": 6289
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 22.26504898071289,
+      "learning_rate": 7.393003908990679e-06,
+      "loss": 1.7896,
+      "step": 6290
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 16.31560707092285,
+      "learning_rate": 7.390999298386289e-06,
+      "loss": 0.9241,
+      "step": 6291
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 132.97479248046875,
+      "learning_rate": 7.3889946877818984e-06,
+      "loss": 2.0392,
+      "step": 6292
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 26.72028350830078,
+      "learning_rate": 7.386990077177509e-06,
+      "loss": 1.2421,
+      "step": 6293
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 7.662388324737549,
+      "learning_rate": 7.3849854665731195e-06,
+      "loss": 1.324,
+      "step": 6294
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 17.365476608276367,
+      "learning_rate": 7.382980855968729e-06,
+      "loss": 1.5321,
+      "step": 6295
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 12.46912956237793,
+      "learning_rate": 7.380976245364339e-06,
+      "loss": 1.4461,
+      "step": 6296
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 11.624770164489746,
+      "learning_rate": 7.378971634759948e-06,
+      "loss": 1.7608,
+      "step": 6297
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 18.62898826599121,
+      "learning_rate": 7.376967024155558e-06,
+      "loss": 1.0102,
+      "step": 6298
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 7.544929504394531,
+      "learning_rate": 7.374962413551169e-06,
+      "loss": 0.8058,
+      "step": 6299
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 8.430646896362305,
+      "learning_rate": 7.372957802946778e-06,
+      "loss": 1.2943,
+      "step": 6300
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 16.084999084472656,
+      "learning_rate": 7.370953192342388e-06,
+      "loss": 1.6322,
+      "step": 6301
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 26.333581924438477,
+      "learning_rate": 7.368948581737998e-06,
+      "loss": 1.3346,
+      "step": 6302
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 12.999780654907227,
+      "learning_rate": 7.366943971133608e-06,
+      "loss": 1.312,
+      "step": 6303
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 20.878276824951172,
+      "learning_rate": 7.364939360529218e-06,
+      "loss": 1.235,
+      "step": 6304
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 39.3919677734375,
+      "learning_rate": 7.362934749924827e-06,
+      "loss": 1.3573,
+      "step": 6305
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 42.52452087402344,
+      "learning_rate": 7.360930139320438e-06,
+      "loss": 2.1336,
+      "step": 6306
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 10.164920806884766,
+      "learning_rate": 7.358925528716048e-06,
+      "loss": 1.1969,
+      "step": 6307
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 14.353740692138672,
+      "learning_rate": 7.356920918111657e-06,
+      "loss": 1.6327,
+      "step": 6308
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 11.760319709777832,
+      "learning_rate": 7.354916307507267e-06,
+      "loss": 1.8274,
+      "step": 6309
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 65.9858169555664,
+      "learning_rate": 7.3529116969028766e-06,
+      "loss": 1.0124,
+      "step": 6310
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 13.734026908874512,
+      "learning_rate": 7.350907086298487e-06,
+      "loss": 1.2509,
+      "step": 6311
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 21.561870574951172,
+      "learning_rate": 7.348902475694098e-06,
+      "loss": 1.3681,
+      "step": 6312
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 51.21876525878906,
+      "learning_rate": 7.346897865089707e-06,
+      "loss": 2.3598,
+      "step": 6313
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 35.55440139770508,
+      "learning_rate": 7.344893254485317e-06,
+      "loss": 1.6387,
+      "step": 6314
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 9.630316734313965,
+      "learning_rate": 7.342888643880927e-06,
+      "loss": 1.0267,
+      "step": 6315
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 17.24791717529297,
+      "learning_rate": 7.340884033276536e-06,
+      "loss": 2.4284,
+      "step": 6316
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 12.127941131591797,
+      "learning_rate": 7.338879422672146e-06,
+      "loss": 1.3257,
+      "step": 6317
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 11.26683235168457,
+      "learning_rate": 7.336874812067757e-06,
+      "loss": 0.946,
+      "step": 6318
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 7.805086612701416,
+      "learning_rate": 7.3348702014633665e-06,
+      "loss": 0.6828,
+      "step": 6319
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 124.16524505615234,
+      "learning_rate": 7.3328655908589766e-06,
+      "loss": 2.6128,
+      "step": 6320
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 26.658418655395508,
+      "learning_rate": 7.330860980254586e-06,
+      "loss": 1.359,
+      "step": 6321
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 63.66525650024414,
+      "learning_rate": 7.328856369650196e-06,
+      "loss": 2.1401,
+      "step": 6322
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 19.522897720336914,
+      "learning_rate": 7.326851759045805e-06,
+      "loss": 1.6017,
+      "step": 6323
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 18.54653549194336,
+      "learning_rate": 7.324847148441416e-06,
+      "loss": 1.902,
+      "step": 6324
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 40.539337158203125,
+      "learning_rate": 7.322842537837026e-06,
+      "loss": 1.6362,
+      "step": 6325
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 24.200485229492188,
+      "learning_rate": 7.320837927232635e-06,
+      "loss": 2.5706,
+      "step": 6326
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 38.600364685058594,
+      "learning_rate": 7.3188333166282455e-06,
+      "loss": 2.5773,
+      "step": 6327
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 13.316827774047852,
+      "learning_rate": 7.3168287060238555e-06,
+      "loss": 1.6111,
+      "step": 6328
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 20.68040657043457,
+      "learning_rate": 7.314824095419465e-06,
+      "loss": 1.8822,
+      "step": 6329
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 20.918405532836914,
+      "learning_rate": 7.312819484815076e-06,
+      "loss": 0.873,
+      "step": 6330
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 19.907466888427734,
+      "learning_rate": 7.310814874210686e-06,
+      "loss": 1.2545,
+      "step": 6331
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 16.205663681030273,
+      "learning_rate": 7.308810263606295e-06,
+      "loss": 1.2967,
+      "step": 6332
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 13.830738067626953,
+      "learning_rate": 7.306805653001905e-06,
+      "loss": 0.9852,
+      "step": 6333
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 12.994755744934082,
+      "learning_rate": 7.304801042397514e-06,
+      "loss": 1.0101,
+      "step": 6334
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 10.922076225280762,
+      "learning_rate": 7.3027964317931244e-06,
+      "loss": 1.1132,
+      "step": 6335
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 14.2009859085083,
+      "learning_rate": 7.300791821188735e-06,
+      "loss": 1.1271,
+      "step": 6336
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 13.78223705291748,
+      "learning_rate": 7.298787210584345e-06,
+      "loss": 1.4629,
+      "step": 6337
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 47.76295852661133,
+      "learning_rate": 7.296782599979955e-06,
+      "loss": 1.6983,
+      "step": 6338
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 26.619237899780273,
+      "learning_rate": 7.294777989375565e-06,
+      "loss": 2.0114,
+      "step": 6339
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 10.558842658996582,
+      "learning_rate": 7.292773378771174e-06,
+      "loss": 0.7375,
+      "step": 6340
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 11.236355781555176,
+      "learning_rate": 7.290768768166784e-06,
+      "loss": 1.406,
+      "step": 6341
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 10.12108325958252,
+      "learning_rate": 7.288764157562395e-06,
+      "loss": 1.4366,
+      "step": 6342
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 23.887067794799805,
+      "learning_rate": 7.286759546958004e-06,
+      "loss": 1.2171,
+      "step": 6343
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 11.277857780456543,
+      "learning_rate": 7.284754936353614e-06,
+      "loss": 1.1178,
+      "step": 6344
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 8.617938041687012,
+      "learning_rate": 7.282750325749224e-06,
+      "loss": 0.9352,
+      "step": 6345
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 7.352804660797119,
+      "learning_rate": 7.280745715144834e-06,
+      "loss": 1.1608,
+      "step": 6346
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 19.099529266357422,
+      "learning_rate": 7.278741104540443e-06,
+      "loss": 1.3358,
+      "step": 6347
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 7.905605792999268,
+      "learning_rate": 7.276736493936053e-06,
+      "loss": 1.8386,
+      "step": 6348
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 23.328462600708008,
+      "learning_rate": 7.274731883331664e-06,
+      "loss": 1.5516,
+      "step": 6349
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 55.610599517822266,
+      "learning_rate": 7.272727272727273e-06,
+      "loss": 1.2527,
+      "step": 6350
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 14.633689880371094,
+      "learning_rate": 7.270722662122883e-06,
+      "loss": 1.4997,
+      "step": 6351
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 15.966309547424316,
+      "learning_rate": 7.268718051518493e-06,
+      "loss": 1.4803,
+      "step": 6352
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 15.084650039672852,
+      "learning_rate": 7.2667134409141026e-06,
+      "loss": 0.8466,
+      "step": 6353
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 24.249235153198242,
+      "learning_rate": 7.264708830309713e-06,
+      "loss": 1.2289,
+      "step": 6354
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 30.721338272094727,
+      "learning_rate": 7.262704219705324e-06,
+      "loss": 1.5891,
+      "step": 6355
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 16.246719360351562,
+      "learning_rate": 7.260699609100933e-06,
+      "loss": 1.4035,
+      "step": 6356
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 21.17243194580078,
+      "learning_rate": 7.258694998496543e-06,
+      "loss": 1.8486,
+      "step": 6357
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 16.293087005615234,
+      "learning_rate": 7.256690387892152e-06,
+      "loss": 0.9408,
+      "step": 6358
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 96.68749237060547,
+      "learning_rate": 7.254685777287762e-06,
+      "loss": 2.8577,
+      "step": 6359
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 17.761423110961914,
+      "learning_rate": 7.2526811666833715e-06,
+      "loss": 2.8659,
+      "step": 6360
+    },
+    {
+      "epoch": 1.91,
+      "eval_loss": 0.17964056134223938,
+      "eval_runtime": 43.926,
+      "eval_samples_per_second": 33.67,
+      "eval_steps_per_second": 33.67,
+      "step": 6360
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 23.259288787841797,
+      "learning_rate": 7.250676556078982e-06,
+      "loss": 1.1937,
+      "step": 6361
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 28.89937400817871,
+      "learning_rate": 7.2486719454745925e-06,
+      "loss": 1.9101,
+      "step": 6362
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 13.6145601272583,
+      "learning_rate": 7.246667334870202e-06,
+      "loss": 1.4405,
+      "step": 6363
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 28.463932037353516,
+      "learning_rate": 7.244662724265812e-06,
+      "loss": 1.7418,
+      "step": 6364
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 29.748435974121094,
+      "learning_rate": 7.242658113661422e-06,
+      "loss": 1.7554,
+      "step": 6365
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 10.47834300994873,
+      "learning_rate": 7.240653503057031e-06,
+      "loss": 0.9418,
+      "step": 6366
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 66.0386962890625,
+      "learning_rate": 7.238648892452642e-06,
+      "loss": 1.4975,
+      "step": 6367
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 17.965469360351562,
+      "learning_rate": 7.236644281848252e-06,
+      "loss": 1.2229,
+      "step": 6368
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 12.599654197692871,
+      "learning_rate": 7.234639671243861e-06,
+      "loss": 1.606,
+      "step": 6369
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 17.390832901000977,
+      "learning_rate": 7.2326350606394715e-06,
+      "loss": 1.387,
+      "step": 6370
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 15.174606323242188,
+      "learning_rate": 7.230630450035081e-06,
+      "loss": 1.3947,
+      "step": 6371
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 16.567323684692383,
+      "learning_rate": 7.228625839430691e-06,
+      "loss": 1.5866,
+      "step": 6372
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 33.17914962768555,
+      "learning_rate": 7.226621228826302e-06,
+      "loss": 1.8752,
+      "step": 6373
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 18.26946258544922,
+      "learning_rate": 7.224616618221911e-06,
+      "loss": 1.1493,
+      "step": 6374
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 26.841148376464844,
+      "learning_rate": 7.222612007617521e-06,
+      "loss": 1.86,
+      "step": 6375
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 15.99163818359375,
+      "learning_rate": 7.220607397013131e-06,
+      "loss": 1.7793,
+      "step": 6376
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 85.62297821044922,
+      "learning_rate": 7.21860278640874e-06,
+      "loss": 2.2498,
+      "step": 6377
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 14.444759368896484,
+      "learning_rate": 7.2165981758043504e-06,
+      "loss": 1.0914,
+      "step": 6378
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 14.665372848510742,
+      "learning_rate": 7.214593565199961e-06,
+      "loss": 2.3208,
+      "step": 6379
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 6.8967604637146,
+      "learning_rate": 7.212588954595571e-06,
+      "loss": 1.0709,
+      "step": 6380
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 27.484949111938477,
+      "learning_rate": 7.210584343991181e-06,
+      "loss": 1.6652,
+      "step": 6381
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 36.965274810791016,
+      "learning_rate": 7.20857973338679e-06,
+      "loss": 2.1163,
+      "step": 6382
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 16.792247772216797,
+      "learning_rate": 7.2065751227824e-06,
+      "loss": 1.4545,
+      "step": 6383
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 18.809303283691406,
+      "learning_rate": 7.204570512178009e-06,
+      "loss": 1.9265,
+      "step": 6384
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 23.61911964416504,
+      "learning_rate": 7.20256590157362e-06,
+      "loss": 1.4629,
+      "step": 6385
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 15.787665367126465,
+      "learning_rate": 7.20056129096923e-06,
+      "loss": 1.5397,
+      "step": 6386
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 25.38530158996582,
+      "learning_rate": 7.1985566803648395e-06,
+      "loss": 1.9278,
+      "step": 6387
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 52.69976806640625,
+      "learning_rate": 7.19655206976045e-06,
+      "loss": 1.0921,
+      "step": 6388
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 12.26889419555664,
+      "learning_rate": 7.19454745915606e-06,
+      "loss": 0.9164,
+      "step": 6389
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 11.600250244140625,
+      "learning_rate": 7.192542848551669e-06,
+      "loss": 0.8845,
+      "step": 6390
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 21.00747299194336,
+      "learning_rate": 7.190538237947279e-06,
+      "loss": 1.6517,
+      "step": 6391
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 34.60345458984375,
+      "learning_rate": 7.18853362734289e-06,
+      "loss": 1.5433,
+      "step": 6392
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 11.045711517333984,
+      "learning_rate": 7.186529016738499e-06,
+      "loss": 1.2821,
+      "step": 6393
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 15.181732177734375,
+      "learning_rate": 7.184524406134109e-06,
+      "loss": 1.382,
+      "step": 6394
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 15.159256935119629,
+      "learning_rate": 7.1825197955297185e-06,
+      "loss": 1.1485,
+      "step": 6395
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 18.364330291748047,
+      "learning_rate": 7.1805151849253286e-06,
+      "loss": 1.5439,
+      "step": 6396
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 25.671911239624023,
+      "learning_rate": 7.178510574320938e-06,
+      "loss": 1.1944,
+      "step": 6397
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 20.186847686767578,
+      "learning_rate": 7.176505963716549e-06,
+      "loss": 1.7509,
+      "step": 6398
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 13.25575065612793,
+      "learning_rate": 7.174501353112159e-06,
+      "loss": 0.9348,
+      "step": 6399
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 16.35478401184082,
+      "learning_rate": 7.172496742507769e-06,
+      "loss": 1.4031,
+      "step": 6400
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 13.498659133911133,
+      "learning_rate": 7.170492131903378e-06,
+      "loss": 1.4828,
+      "step": 6401
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 88.43494415283203,
+      "learning_rate": 7.168487521298988e-06,
+      "loss": 1.9878,
+      "step": 6402
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 11.417867660522461,
+      "learning_rate": 7.1664829106945975e-06,
+      "loss": 1.5075,
+      "step": 6403
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 12.581781387329102,
+      "learning_rate": 7.164478300090208e-06,
+      "loss": 1.1652,
+      "step": 6404
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 19.474443435668945,
+      "learning_rate": 7.1624736894858185e-06,
+      "loss": 1.7053,
+      "step": 6405
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 21.434553146362305,
+      "learning_rate": 7.160469078881428e-06,
+      "loss": 2.3571,
+      "step": 6406
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 17.52594566345215,
+      "learning_rate": 7.158464468277038e-06,
+      "loss": 2.2389,
+      "step": 6407
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 26.64896583557129,
+      "learning_rate": 7.156459857672647e-06,
+      "loss": 1.8134,
+      "step": 6408
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 13.084494590759277,
+      "learning_rate": 7.154455247068257e-06,
+      "loss": 1.584,
+      "step": 6409
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 12.483121871948242,
+      "learning_rate": 7.152450636463868e-06,
+      "loss": 1.8767,
+      "step": 6410
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 8.219230651855469,
+      "learning_rate": 7.150446025859477e-06,
+      "loss": 1.177,
+      "step": 6411
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 31.770601272583008,
+      "learning_rate": 7.148441415255087e-06,
+      "loss": 1.9082,
+      "step": 6412
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 18.06419563293457,
+      "learning_rate": 7.1464368046506975e-06,
+      "loss": 1.2848,
+      "step": 6413
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 6.39808464050293,
+      "learning_rate": 7.144432194046307e-06,
+      "loss": 0.4952,
+      "step": 6414
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 16.47151756286621,
+      "learning_rate": 7.142427583441917e-06,
+      "loss": 1.4156,
+      "step": 6415
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 35.63877487182617,
+      "learning_rate": 7.140422972837528e-06,
+      "loss": 1.6205,
+      "step": 6416
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 16.693328857421875,
+      "learning_rate": 7.138418362233137e-06,
+      "loss": 1.1112,
+      "step": 6417
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 34.685909271240234,
+      "learning_rate": 7.136413751628747e-06,
+      "loss": 1.704,
+      "step": 6418
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 17.206161499023438,
+      "learning_rate": 7.134409141024356e-06,
+      "loss": 1.1891,
+      "step": 6419
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 19.83222770690918,
+      "learning_rate": 7.132404530419966e-06,
+      "loss": 2.2121,
+      "step": 6420
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 26.431730270385742,
+      "learning_rate": 7.130399919815576e-06,
+      "loss": 1.7221,
+      "step": 6421
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 9.490403175354004,
+      "learning_rate": 7.1283953092111865e-06,
+      "loss": 1.1032,
+      "step": 6422
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 52.717281341552734,
+      "learning_rate": 7.126390698606797e-06,
+      "loss": 1.5446,
+      "step": 6423
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 14.396468162536621,
+      "learning_rate": 7.124386088002406e-06,
+      "loss": 0.762,
+      "step": 6424
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 34.430179595947266,
+      "learning_rate": 7.122381477398016e-06,
+      "loss": 2.3349,
+      "step": 6425
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 7.51977014541626,
+      "learning_rate": 7.120376866793626e-06,
+      "loss": 0.1446,
+      "step": 6426
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 11.881677627563477,
+      "learning_rate": 7.118372256189235e-06,
+      "loss": 1.2101,
+      "step": 6427
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 17.82876968383789,
+      "learning_rate": 7.116367645584846e-06,
+      "loss": 1.2283,
+      "step": 6428
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 16.561769485473633,
+      "learning_rate": 7.114363034980456e-06,
+      "loss": 2.0483,
+      "step": 6429
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 35.203155517578125,
+      "learning_rate": 7.1123584243760655e-06,
+      "loss": 1.7524,
+      "step": 6430
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 24.895706176757812,
+      "learning_rate": 7.110353813771676e-06,
+      "loss": 1.6033,
+      "step": 6431
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 23.870237350463867,
+      "learning_rate": 7.108349203167285e-06,
+      "loss": 1.4187,
+      "step": 6432
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 24.825237274169922,
+      "learning_rate": 7.106344592562895e-06,
+      "loss": 1.594,
+      "step": 6433
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 8.913065910339355,
+      "learning_rate": 7.104339981958505e-06,
+      "loss": 1.0598,
+      "step": 6434
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 41.398555755615234,
+      "learning_rate": 7.102335371354115e-06,
+      "loss": 2.3157,
+      "step": 6435
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 18.028615951538086,
+      "learning_rate": 7.100330760749725e-06,
+      "loss": 2.0904,
+      "step": 6436
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 11.3342866897583,
+      "learning_rate": 7.098326150145335e-06,
+      "loss": 1.7827,
+      "step": 6437
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 20.189563751220703,
+      "learning_rate": 7.0963215395409445e-06,
+      "loss": 1.6553,
+      "step": 6438
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 12.127889633178711,
+      "learning_rate": 7.094316928936555e-06,
+      "loss": 1.4503,
+      "step": 6439
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 37.8315315246582,
+      "learning_rate": 7.092312318332164e-06,
+      "loss": 1.7977,
+      "step": 6440
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 13.1519775390625,
+      "learning_rate": 7.090307707727775e-06,
+      "loss": 1.6393,
+      "step": 6441
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 35.46371841430664,
+      "learning_rate": 7.088303097123385e-06,
+      "loss": 2.3399,
+      "step": 6442
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 12.697428703308105,
+      "learning_rate": 7.086298486518994e-06,
+      "loss": 2.5277,
+      "step": 6443
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 26.87334632873535,
+      "learning_rate": 7.084293875914604e-06,
+      "loss": 2.0531,
+      "step": 6444
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 14.13779354095459,
+      "learning_rate": 7.082289265310213e-06,
+      "loss": 1.8136,
+      "step": 6445
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 12.27436637878418,
+      "learning_rate": 7.0802846547058235e-06,
+      "loss": 1.2326,
+      "step": 6446
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 19.918701171875,
+      "learning_rate": 7.078280044101434e-06,
+      "loss": 1.4257,
+      "step": 6447
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 13.029349327087402,
+      "learning_rate": 7.076275433497044e-06,
+      "loss": 1.7537,
+      "step": 6448
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 9.116921424865723,
+      "learning_rate": 7.074270822892654e-06,
+      "loss": 0.7983,
+      "step": 6449
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 30.727689743041992,
+      "learning_rate": 7.072266212288264e-06,
+      "loss": 1.2988,
+      "step": 6450
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 15.236931800842285,
+      "learning_rate": 7.070261601683873e-06,
+      "loss": 1.4535,
+      "step": 6451
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 38.2266845703125,
+      "learning_rate": 7.068256991079483e-06,
+      "loss": 0.8669,
+      "step": 6452
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 42.05986785888672,
+      "learning_rate": 7.066252380475094e-06,
+      "loss": 1.7596,
+      "step": 6453
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 28.586145401000977,
+      "learning_rate": 7.064247769870703e-06,
+      "loss": 1.823,
+      "step": 6454
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 6.259302616119385,
+      "learning_rate": 7.062243159266313e-06,
+      "loss": 0.3312,
+      "step": 6455
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 8.961036682128906,
+      "learning_rate": 7.060238548661923e-06,
+      "loss": 0.999,
+      "step": 6456
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 19.617019653320312,
+      "learning_rate": 7.058233938057533e-06,
+      "loss": 1.9528,
+      "step": 6457
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 30.003917694091797,
+      "learning_rate": 7.056229327453143e-06,
+      "loss": 1.5707,
+      "step": 6458
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 16.13441276550293,
+      "learning_rate": 7.054224716848753e-06,
+      "loss": 2.1332,
+      "step": 6459
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 11.956790924072266,
+      "learning_rate": 7.052220106244363e-06,
+      "loss": 1.4419,
+      "step": 6460
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 17.357301712036133,
+      "learning_rate": 7.050215495639973e-06,
+      "loss": 1.1238,
+      "step": 6461
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 16.655643463134766,
+      "learning_rate": 7.048210885035582e-06,
+      "loss": 2.1496,
+      "step": 6462
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 17.952760696411133,
+      "learning_rate": 7.046206274431192e-06,
+      "loss": 1.5901,
+      "step": 6463
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 20.259384155273438,
+      "learning_rate": 7.044201663826802e-06,
+      "loss": 1.2221,
+      "step": 6464
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 25.712038040161133,
+      "learning_rate": 7.0421970532224125e-06,
+      "loss": 2.141,
+      "step": 6465
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 17.299470901489258,
+      "learning_rate": 7.040192442618023e-06,
+      "loss": 0.9474,
+      "step": 6466
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 18.08493423461914,
+      "learning_rate": 7.038187832013632e-06,
+      "loss": 1.711,
+      "step": 6467
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 54.114418029785156,
+      "learning_rate": 7.036183221409242e-06,
+      "loss": 0.6773,
+      "step": 6468
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 13.001557350158691,
+      "learning_rate": 7.034178610804851e-06,
+      "loss": 1.1428,
+      "step": 6469
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 8.948216438293457,
+      "learning_rate": 7.032174000200461e-06,
+      "loss": 1.2129,
+      "step": 6470
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 96.37139892578125,
+      "learning_rate": 7.030169389596072e-06,
+      "loss": 1.3139,
+      "step": 6471
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 14.782844543457031,
+      "learning_rate": 7.0281647789916814e-06,
+      "loss": 1.4799,
+      "step": 6472
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 14.976436614990234,
+      "learning_rate": 7.0261601683872915e-06,
+      "loss": 1.4801,
+      "step": 6473
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 26.140769958496094,
+      "learning_rate": 7.024155557782902e-06,
+      "loss": 1.9828,
+      "step": 6474
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 14.462868690490723,
+      "learning_rate": 7.022150947178511e-06,
+      "loss": 1.3346,
+      "step": 6475
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 12.547101974487305,
+      "learning_rate": 7.020146336574121e-06,
+      "loss": 1.8225,
+      "step": 6476
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 13.108290672302246,
+      "learning_rate": 7.01814172596973e-06,
+      "loss": 1.7796,
+      "step": 6477
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 20.673439025878906,
+      "learning_rate": 7.016137115365341e-06,
+      "loss": 1.7685,
+      "step": 6478
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 9.68248462677002,
+      "learning_rate": 7.014132504760951e-06,
+      "loss": 1.1545,
+      "step": 6479
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 11.824162483215332,
+      "learning_rate": 7.01212789415656e-06,
+      "loss": 1.1257,
+      "step": 6480
+    },
+    {
+      "epoch": 1.95,
+      "eval_loss": 0.17668664455413818,
+      "eval_runtime": 43.6307,
+      "eval_samples_per_second": 33.898,
+      "eval_steps_per_second": 33.898,
+      "step": 6480
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 19.132976531982422,
+      "learning_rate": 7.0101232835521705e-06,
+      "loss": 1.4466,
+      "step": 6481
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 13.578184127807617,
+      "learning_rate": 7.00811867294778e-06,
+      "loss": 1.8283,
+      "step": 6482
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 14.636981010437012,
+      "learning_rate": 7.00611406234339e-06,
+      "loss": 1.5467,
+      "step": 6483
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 14.252835273742676,
+      "learning_rate": 7.004109451739001e-06,
+      "loss": 1.0799,
+      "step": 6484
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 7.7530837059021,
+      "learning_rate": 7.002104841134611e-06,
+      "loss": 1.7159,
+      "step": 6485
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 19.219745635986328,
+      "learning_rate": 7.00010023053022e-06,
+      "loss": 1.3202,
+      "step": 6486
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 10.907150268554688,
+      "learning_rate": 6.99809561992583e-06,
+      "loss": 1.3228,
+      "step": 6487
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 7.514638423919678,
+      "learning_rate": 6.996091009321439e-06,
+      "loss": 1.1397,
+      "step": 6488
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 21.773658752441406,
+      "learning_rate": 6.9940863987170495e-06,
+      "loss": 0.9278,
+      "step": 6489
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 10.506806373596191,
+      "learning_rate": 6.99208178811266e-06,
+      "loss": 1.709,
+      "step": 6490
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 15.891654014587402,
+      "learning_rate": 6.99007717750827e-06,
+      "loss": 1.4618,
+      "step": 6491
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 31.406612396240234,
+      "learning_rate": 6.98807256690388e-06,
+      "loss": 1.977,
+      "step": 6492
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 20.738195419311523,
+      "learning_rate": 6.986067956299489e-06,
+      "loss": 1.4744,
+      "step": 6493
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 17.87335968017578,
+      "learning_rate": 6.984063345695099e-06,
+      "loss": 1.8068,
+      "step": 6494
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 13.646010398864746,
+      "learning_rate": 6.982058735090709e-06,
+      "loss": 0.9828,
+      "step": 6495
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 12.331052780151367,
+      "learning_rate": 6.980054124486319e-06,
+      "loss": 1.6284,
+      "step": 6496
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 52.41801452636719,
+      "learning_rate": 6.978049513881929e-06,
+      "loss": 0.973,
+      "step": 6497
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 27.965364456176758,
+      "learning_rate": 6.976044903277539e-06,
+      "loss": 1.0861,
+      "step": 6498
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 25.636484146118164,
+      "learning_rate": 6.974040292673149e-06,
+      "loss": 1.8763,
+      "step": 6499
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 15.626718521118164,
+      "learning_rate": 6.972035682068759e-06,
+      "loss": 1.2224,
+      "step": 6500
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 12.968523979187012,
+      "learning_rate": 6.970031071464368e-06,
+      "loss": 1.4936,
+      "step": 6501
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 38.640384674072266,
+      "learning_rate": 6.968026460859979e-06,
+      "loss": 1.847,
+      "step": 6502
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 35.005760192871094,
+      "learning_rate": 6.966021850255589e-06,
+      "loss": 1.8344,
+      "step": 6503
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 59.90297317504883,
+      "learning_rate": 6.964017239651198e-06,
+      "loss": 1.5804,
+      "step": 6504
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 8.664277076721191,
+      "learning_rate": 6.962012629046808e-06,
+      "loss": 1.1554,
+      "step": 6505
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 15.182535171508789,
+      "learning_rate": 6.9600080184424175e-06,
+      "loss": 1.107,
+      "step": 6506
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 25.827924728393555,
+      "learning_rate": 6.958003407838028e-06,
+      "loss": 1.1151,
+      "step": 6507
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 9.57219409942627,
+      "learning_rate": 6.9559987972336385e-06,
+      "loss": 1.3565,
+      "step": 6508
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 113.87895202636719,
+      "learning_rate": 6.953994186629248e-06,
+      "loss": 1.4153,
+      "step": 6509
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 17.828977584838867,
+      "learning_rate": 6.951989576024858e-06,
+      "loss": 1.8861,
+      "step": 6510
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 27.323060989379883,
+      "learning_rate": 6.949984965420468e-06,
+      "loss": 1.4276,
+      "step": 6511
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 11.81863021850586,
+      "learning_rate": 6.947980354816077e-06,
+      "loss": 0.785,
+      "step": 6512
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 10.938015937805176,
+      "learning_rate": 6.945975744211687e-06,
+      "loss": 1.096,
+      "step": 6513
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 11.981101036071777,
+      "learning_rate": 6.9439711336072965e-06,
+      "loss": 1.3268,
+      "step": 6514
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 19.58521842956543,
+      "learning_rate": 6.9419665230029074e-06,
+      "loss": 1.8049,
+      "step": 6515
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 24.513118743896484,
+      "learning_rate": 6.9399619123985175e-06,
+      "loss": 1.3997,
+      "step": 6516
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 13.00726318359375,
+      "learning_rate": 6.937957301794127e-06,
+      "loss": 2.1062,
+      "step": 6517
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 16.61869239807129,
+      "learning_rate": 6.935952691189737e-06,
+      "loss": 1.4637,
+      "step": 6518
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 18.2260684967041,
+      "learning_rate": 6.933948080585347e-06,
+      "loss": 1.3907,
+      "step": 6519
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 27.4097957611084,
+      "learning_rate": 6.931943469980956e-06,
+      "loss": 2.0212,
+      "step": 6520
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 13.984885215759277,
+      "learning_rate": 6.929938859376567e-06,
+      "loss": 0.8537,
+      "step": 6521
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 29.465072631835938,
+      "learning_rate": 6.927934248772177e-06,
+      "loss": 2.082,
+      "step": 6522
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 13.194611549377441,
+      "learning_rate": 6.925929638167786e-06,
+      "loss": 1.1381,
+      "step": 6523
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 7.926638603210449,
+      "learning_rate": 6.9239250275633965e-06,
+      "loss": 0.9932,
+      "step": 6524
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 16.14949607849121,
+      "learning_rate": 6.921920416959006e-06,
+      "loss": 1.4698,
+      "step": 6525
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 33.72904968261719,
+      "learning_rate": 6.919915806354616e-06,
+      "loss": 3.2966,
+      "step": 6526
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 8.642800331115723,
+      "learning_rate": 6.917911195750227e-06,
+      "loss": 1.1102,
+      "step": 6527
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 12.295087814331055,
+      "learning_rate": 6.915906585145836e-06,
+      "loss": 0.9972,
+      "step": 6528
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 24.457801818847656,
+      "learning_rate": 6.913901974541446e-06,
+      "loss": 1.4332,
+      "step": 6529
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 14.026710510253906,
+      "learning_rate": 6.911897363937055e-06,
+      "loss": 1.8641,
+      "step": 6530
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 12.17983341217041,
+      "learning_rate": 6.909892753332665e-06,
+      "loss": 1.4611,
+      "step": 6531
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 12.283072471618652,
+      "learning_rate": 6.9078881427282755e-06,
+      "loss": 1.3947,
+      "step": 6532
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 17.96526527404785,
+      "learning_rate": 6.9058835321238856e-06,
+      "loss": 1.7153,
+      "step": 6533
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 24.099628448486328,
+      "learning_rate": 6.903878921519496e-06,
+      "loss": 1.2495,
+      "step": 6534
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 16.107059478759766,
+      "learning_rate": 6.901874310915106e-06,
+      "loss": 1.4728,
+      "step": 6535
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 15.214371681213379,
+      "learning_rate": 6.899869700310715e-06,
+      "loss": 1.1217,
+      "step": 6536
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 27.67184829711914,
+      "learning_rate": 6.897865089706325e-06,
+      "loss": 1.4113,
+      "step": 6537
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 20.205726623535156,
+      "learning_rate": 6.895860479101934e-06,
+      "loss": 0.8228,
+      "step": 6538
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 12.976736068725586,
+      "learning_rate": 6.893855868497545e-06,
+      "loss": 1.0787,
+      "step": 6539
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 141.27496337890625,
+      "learning_rate": 6.891851257893155e-06,
+      "loss": 2.659,
+      "step": 6540
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 25.662107467651367,
+      "learning_rate": 6.8898466472887645e-06,
+      "loss": 2.3812,
+      "step": 6541
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 17.610265731811523,
+      "learning_rate": 6.887842036684375e-06,
+      "loss": 1.1345,
+      "step": 6542
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 13.565105438232422,
+      "learning_rate": 6.885837426079985e-06,
+      "loss": 1.7477,
+      "step": 6543
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 72.72836303710938,
+      "learning_rate": 6.883832815475594e-06,
+      "loss": 1.0895,
+      "step": 6544
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 11.967425346374512,
+      "learning_rate": 6.881828204871205e-06,
+      "loss": 1.2826,
+      "step": 6545
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 27.99849510192871,
+      "learning_rate": 6.879823594266815e-06,
+      "loss": 1.4163,
+      "step": 6546
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 17.99080467224121,
+      "learning_rate": 6.877818983662424e-06,
+      "loss": 0.7908,
+      "step": 6547
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 8.94594955444336,
+      "learning_rate": 6.875814373058034e-06,
+      "loss": 1.5791,
+      "step": 6548
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 18.120492935180664,
+      "learning_rate": 6.8738097624536435e-06,
+      "loss": 1.5981,
+      "step": 6549
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 11.887974739074707,
+      "learning_rate": 6.871805151849254e-06,
+      "loss": 2.1877,
+      "step": 6550
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 14.82646369934082,
+      "learning_rate": 6.8698005412448645e-06,
+      "loss": 1.2886,
+      "step": 6551
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 15.733491897583008,
+      "learning_rate": 6.867795930640474e-06,
+      "loss": 0.836,
+      "step": 6552
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 20.574148178100586,
+      "learning_rate": 6.865791320036084e-06,
+      "loss": 0.9531,
+      "step": 6553
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 28.415647506713867,
+      "learning_rate": 6.863786709431693e-06,
+      "loss": 2.1882,
+      "step": 6554
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 17.647674560546875,
+      "learning_rate": 6.861782098827303e-06,
+      "loss": 2.3014,
+      "step": 6555
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 49.27579879760742,
+      "learning_rate": 6.859777488222913e-06,
+      "loss": 1.5323,
+      "step": 6556
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 19.76232147216797,
+      "learning_rate": 6.8577728776185225e-06,
+      "loss": 1.6397,
+      "step": 6557
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 11.850191116333008,
+      "learning_rate": 6.8557682670141334e-06,
+      "loss": 1.027,
+      "step": 6558
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 26.760108947753906,
+      "learning_rate": 6.8537636564097435e-06,
+      "loss": 1.7687,
+      "step": 6559
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 11.1150484085083,
+      "learning_rate": 6.851759045805353e-06,
+      "loss": 1.418,
+      "step": 6560
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 23.722135543823242,
+      "learning_rate": 6.849754435200963e-06,
+      "loss": 1.0756,
+      "step": 6561
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 17.482393264770508,
+      "learning_rate": 6.847749824596572e-06,
+      "loss": 1.2066,
+      "step": 6562
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 17.04688835144043,
+      "learning_rate": 6.845745213992182e-06,
+      "loss": 1.4078,
+      "step": 6563
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 15.257699012756348,
+      "learning_rate": 6.843740603387793e-06,
+      "loss": 1.8723,
+      "step": 6564
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 13.324009895324707,
+      "learning_rate": 6.841735992783402e-06,
+      "loss": 1.3293,
+      "step": 6565
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 18.284709930419922,
+      "learning_rate": 6.839731382179012e-06,
+      "loss": 1.8599,
+      "step": 6566
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 13.653274536132812,
+      "learning_rate": 6.837726771574622e-06,
+      "loss": 1.7576,
+      "step": 6567
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 38.0052490234375,
+      "learning_rate": 6.835722160970232e-06,
+      "loss": 1.656,
+      "step": 6568
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 15.570902824401855,
+      "learning_rate": 6.833717550365842e-06,
+      "loss": 1.3572,
+      "step": 6569
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 10.287487030029297,
+      "learning_rate": 6.831712939761453e-06,
+      "loss": 1.0965,
+      "step": 6570
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 21.403841018676758,
+      "learning_rate": 6.829708329157062e-06,
+      "loss": 1.657,
+      "step": 6571
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 11.690693855285645,
+      "learning_rate": 6.827703718552672e-06,
+      "loss": 1.9188,
+      "step": 6572
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 15.413796424865723,
+      "learning_rate": 6.825699107948281e-06,
+      "loss": 1.1102,
+      "step": 6573
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 14.91450023651123,
+      "learning_rate": 6.823694497343891e-06,
+      "loss": 1.0896,
+      "step": 6574
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 12.081074714660645,
+      "learning_rate": 6.821689886739501e-06,
+      "loss": 1.7378,
+      "step": 6575
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 16.04603385925293,
+      "learning_rate": 6.8196852761351116e-06,
+      "loss": 1.4417,
+      "step": 6576
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 22.271512985229492,
+      "learning_rate": 6.817680665530722e-06,
+      "loss": 1.7555,
+      "step": 6577
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 13.81209945678711,
+      "learning_rate": 6.815676054926331e-06,
+      "loss": 1.1311,
+      "step": 6578
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 16.152433395385742,
+      "learning_rate": 6.813671444321941e-06,
+      "loss": 1.1026,
+      "step": 6579
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 12.994856834411621,
+      "learning_rate": 6.811666833717551e-06,
+      "loss": 1.1081,
+      "step": 6580
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 12.3396577835083,
+      "learning_rate": 6.80966222311316e-06,
+      "loss": 1.3073,
+      "step": 6581
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 8.903776168823242,
+      "learning_rate": 6.807657612508771e-06,
+      "loss": 0.9238,
+      "step": 6582
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 15.200712203979492,
+      "learning_rate": 6.805653001904381e-06,
+      "loss": 1.3466,
+      "step": 6583
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 19.97858428955078,
+      "learning_rate": 6.8036483912999905e-06,
+      "loss": 1.7572,
+      "step": 6584
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 24.35478401184082,
+      "learning_rate": 6.801643780695601e-06,
+      "loss": 1.929,
+      "step": 6585
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 30.171939849853516,
+      "learning_rate": 6.79963917009121e-06,
+      "loss": 1.5492,
+      "step": 6586
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 12.768831253051758,
+      "learning_rate": 6.79763455948682e-06,
+      "loss": 1.0463,
+      "step": 6587
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 21.945470809936523,
+      "learning_rate": 6.795629948882431e-06,
+      "loss": 2.2842,
+      "step": 6588
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 22.516357421875,
+      "learning_rate": 6.79362533827804e-06,
+      "loss": 2.0211,
+      "step": 6589
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 8.880302429199219,
+      "learning_rate": 6.79162072767365e-06,
+      "loss": 1.2581,
+      "step": 6590
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 9.477561950683594,
+      "learning_rate": 6.7896161170692594e-06,
+      "loss": 0.7168,
+      "step": 6591
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 21.51420783996582,
+      "learning_rate": 6.7876115064648695e-06,
+      "loss": 1.3912,
+      "step": 6592
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 50.1191291809082,
+      "learning_rate": 6.78560689586048e-06,
+      "loss": 3.0844,
+      "step": 6593
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 19.533275604248047,
+      "learning_rate": 6.78360228525609e-06,
+      "loss": 2.1365,
+      "step": 6594
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 11.014450073242188,
+      "learning_rate": 6.7815976746517e-06,
+      "loss": 0.7238,
+      "step": 6595
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 35.398197174072266,
+      "learning_rate": 6.77959306404731e-06,
+      "loss": 2.9231,
+      "step": 6596
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 59.179141998291016,
+      "learning_rate": 6.777588453442919e-06,
+      "loss": 1.4903,
+      "step": 6597
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 26.32767677307129,
+      "learning_rate": 6.775583842838529e-06,
+      "loss": 1.0562,
+      "step": 6598
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 14.91978931427002,
+      "learning_rate": 6.773579232234138e-06,
+      "loss": 1.8945,
+      "step": 6599
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 43.29436111450195,
+      "learning_rate": 6.7715746216297485e-06,
+      "loss": 2.4119,
+      "step": 6600
+    },
+    {
+      "epoch": 1.98,
+      "eval_loss": 0.17498844861984253,
+      "eval_runtime": 43.9202,
+      "eval_samples_per_second": 33.675,
+      "eval_steps_per_second": 33.675,
+      "step": 6600
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 20.57763671875,
+      "learning_rate": 6.7695700110253594e-06,
+      "loss": 1.2593,
+      "step": 6601
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 11.902490615844727,
+      "learning_rate": 6.767565400420969e-06,
+      "loss": 0.8341,
+      "step": 6602
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 11.146263122558594,
+      "learning_rate": 6.765560789816579e-06,
+      "loss": 1.3651,
+      "step": 6603
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 11.864297866821289,
+      "learning_rate": 6.763556179212189e-06,
+      "loss": 0.9559,
+      "step": 6604
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 12.255757331848145,
+      "learning_rate": 6.761551568607798e-06,
+      "loss": 1.1177,
+      "step": 6605
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 6.7448506355285645,
+      "learning_rate": 6.759546958003408e-06,
+      "loss": 0.6869,
+      "step": 6606
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 15.666464805603027,
+      "learning_rate": 6.757542347399019e-06,
+      "loss": 1.7579,
+      "step": 6607
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 16.473482131958008,
+      "learning_rate": 6.755537736794628e-06,
+      "loss": 0.9031,
+      "step": 6608
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 12.140952110290527,
+      "learning_rate": 6.753533126190238e-06,
+      "loss": 0.7175,
+      "step": 6609
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 9.83596134185791,
+      "learning_rate": 6.751528515585848e-06,
+      "loss": 0.8216,
+      "step": 6610
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 13.17597770690918,
+      "learning_rate": 6.749523904981458e-06,
+      "loss": 1.694,
+      "step": 6611
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 17.648849487304688,
+      "learning_rate": 6.747519294377067e-06,
+      "loss": 2.1423,
+      "step": 6612
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 13.655058860778809,
+      "learning_rate": 6.745514683772678e-06,
+      "loss": 0.9178,
+      "step": 6613
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 18.347841262817383,
+      "learning_rate": 6.743510073168288e-06,
+      "loss": 1.0027,
+      "step": 6614
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 26.16404914855957,
+      "learning_rate": 6.741505462563897e-06,
+      "loss": 2.206,
+      "step": 6615
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 7.279219150543213,
+      "learning_rate": 6.739500851959507e-06,
+      "loss": 0.7718,
+      "step": 6616
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 12.55292797088623,
+      "learning_rate": 6.737496241355117e-06,
+      "loss": 1.8409,
+      "step": 6617
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 42.25106430053711,
+      "learning_rate": 6.735491630750727e-06,
+      "loss": 2.3152,
+      "step": 6618
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 14.544368743896484,
+      "learning_rate": 6.7334870201463376e-06,
+      "loss": 1.2898,
+      "step": 6619
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 21.00879669189453,
+      "learning_rate": 6.731482409541948e-06,
+      "loss": 1.4371,
+      "step": 6620
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 15.112432479858398,
+      "learning_rate": 6.729477798937557e-06,
+      "loss": 2.2847,
+      "step": 6621
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 16.390920639038086,
+      "learning_rate": 6.727473188333167e-06,
+      "loss": 1.1366,
+      "step": 6622
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 17.105491638183594,
+      "learning_rate": 6.725468577728776e-06,
+      "loss": 1.7725,
+      "step": 6623
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 17.488632202148438,
+      "learning_rate": 6.723463967124386e-06,
+      "loss": 0.5385,
+      "step": 6624
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 29.78978157043457,
+      "learning_rate": 6.721459356519997e-06,
+      "loss": 2.3833,
+      "step": 6625
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 15.036394119262695,
+      "learning_rate": 6.7194547459156065e-06,
+      "loss": 1.0212,
+      "step": 6626
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 17.2657527923584,
+      "learning_rate": 6.7174501353112165e-06,
+      "loss": 1.4752,
+      "step": 6627
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 27.82944679260254,
+      "learning_rate": 6.715445524706826e-06,
+      "loss": 1.741,
+      "step": 6628
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 14.460271835327148,
+      "learning_rate": 6.713440914102436e-06,
+      "loss": 1.6737,
+      "step": 6629
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 30.571147918701172,
+      "learning_rate": 6.711436303498046e-06,
+      "loss": 2.2318,
+      "step": 6630
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 18.02724838256836,
+      "learning_rate": 6.709431692893657e-06,
+      "loss": 1.0354,
+      "step": 6631
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 12.847180366516113,
+      "learning_rate": 6.707427082289266e-06,
+      "loss": 1.5948,
+      "step": 6632
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 13.379613876342773,
+      "learning_rate": 6.705422471684876e-06,
+      "loss": 1.2021,
+      "step": 6633
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 20.53904914855957,
+      "learning_rate": 6.7034178610804854e-06,
+      "loss": 1.9106,
+      "step": 6634
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 23.03459930419922,
+      "learning_rate": 6.7014132504760955e-06,
+      "loss": 1.785,
+      "step": 6635
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 19.824228286743164,
+      "learning_rate": 6.699408639871705e-06,
+      "loss": 0.8569,
+      "step": 6636
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 59.54680633544922,
+      "learning_rate": 6.697404029267316e-06,
+      "loss": 2.1475,
+      "step": 6637
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 10.229207992553711,
+      "learning_rate": 6.695399418662926e-06,
+      "loss": 1.1441,
+      "step": 6638
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 28.467979431152344,
+      "learning_rate": 6.693394808058535e-06,
+      "loss": 1.7708,
+      "step": 6639
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 10.531692504882812,
+      "learning_rate": 6.691390197454145e-06,
+      "loss": 0.5444,
+      "step": 6640
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 13.020025253295898,
+      "learning_rate": 6.689385586849755e-06,
+      "loss": 0.8387,
+      "step": 6641
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 12.514997482299805,
+      "learning_rate": 6.687380976245364e-06,
+      "loss": 1.3939,
+      "step": 6642
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 27.652118682861328,
+      "learning_rate": 6.6853763656409745e-06,
+      "loss": 3.0772,
+      "step": 6643
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 24.15280532836914,
+      "learning_rate": 6.6833717550365854e-06,
+      "loss": 1.5129,
+      "step": 6644
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 13.905855178833008,
+      "learning_rate": 6.681367144432195e-06,
+      "loss": 0.9144,
+      "step": 6645
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 16.91253662109375,
+      "learning_rate": 6.679362533827805e-06,
+      "loss": 1.5355,
+      "step": 6646
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 11.888856887817383,
+      "learning_rate": 6.677357923223414e-06,
+      "loss": 1.1719,
+      "step": 6647
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 14.441067695617676,
+      "learning_rate": 6.675353312619024e-06,
+      "loss": 1.3663,
+      "step": 6648
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 13.287384033203125,
+      "learning_rate": 6.673348702014633e-06,
+      "loss": 0.5206,
+      "step": 6649
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 12.121244430541992,
+      "learning_rate": 6.671344091410244e-06,
+      "loss": 1.0182,
+      "step": 6650
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 44.00876998901367,
+      "learning_rate": 6.669339480805854e-06,
+      "loss": 0.9455,
+      "step": 6651
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 15.983967781066895,
+      "learning_rate": 6.6673348702014636e-06,
+      "loss": 0.8615,
+      "step": 6652
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 13.022303581237793,
+      "learning_rate": 6.665330259597074e-06,
+      "loss": 1.3803,
+      "step": 6653
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 57.43462371826172,
+      "learning_rate": 6.663325648992684e-06,
+      "loss": 1.9661,
+      "step": 6654
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 13.801424026489258,
+      "learning_rate": 6.661321038388293e-06,
+      "loss": 1.6653,
+      "step": 6655
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 14.12394905090332,
+      "learning_rate": 6.659316427783904e-06,
+      "loss": 1.0337,
+      "step": 6656
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 16.065610885620117,
+      "learning_rate": 6.657311817179514e-06,
+      "loss": 1.6578,
+      "step": 6657
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 23.195810317993164,
+      "learning_rate": 6.655307206575123e-06,
+      "loss": 2.0227,
+      "step": 6658
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 39.60324478149414,
+      "learning_rate": 6.653302595970733e-06,
+      "loss": 2.9616,
+      "step": 6659
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 12.803409576416016,
+      "learning_rate": 6.6512979853663425e-06,
+      "loss": 1.3529,
+      "step": 6660
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 27.853031158447266,
+      "learning_rate": 6.649293374761953e-06,
+      "loss": 1.2641,
+      "step": 6661
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 12.550612449645996,
+      "learning_rate": 6.6472887641575636e-06,
+      "loss": 1.032,
+      "step": 6662
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 14.057964324951172,
+      "learning_rate": 6.645284153553173e-06,
+      "loss": 1.4831,
+      "step": 6663
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 13.735244750976562,
+      "learning_rate": 6.643279542948783e-06,
+      "loss": 1.842,
+      "step": 6664
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 10.447699546813965,
+      "learning_rate": 6.641274932344393e-06,
+      "loss": 1.1719,
+      "step": 6665
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 24.760276794433594,
+      "learning_rate": 6.639270321740002e-06,
+      "loss": 1.0642,
+      "step": 6666
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 17.945220947265625,
+      "learning_rate": 6.637265711135612e-06,
+      "loss": 2.1282,
+      "step": 6667
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 40.95531463623047,
+      "learning_rate": 6.635261100531223e-06,
+      "loss": 1.8811,
+      "step": 6668
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 13.15553092956543,
+      "learning_rate": 6.6332564899268325e-06,
+      "loss": 1.134,
+      "step": 6669
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 20.673992156982422,
+      "learning_rate": 6.6312518793224425e-06,
+      "loss": 1.2125,
+      "step": 6670
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 17.39179801940918,
+      "learning_rate": 6.629247268718052e-06,
+      "loss": 1.7538,
+      "step": 6671
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 8.030019760131836,
+      "learning_rate": 6.627242658113662e-06,
+      "loss": 1.5265,
+      "step": 6672
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 23.085996627807617,
+      "learning_rate": 6.625238047509271e-06,
+      "loss": 1.5642,
+      "step": 6673
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 36.12226104736328,
+      "learning_rate": 6.623233436904882e-06,
+      "loss": 2.836,
+      "step": 6674
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 16.673858642578125,
+      "learning_rate": 6.621228826300492e-06,
+      "loss": 1.0128,
+      "step": 6675
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 18.04882049560547,
+      "learning_rate": 6.619224215696101e-06,
+      "loss": 1.2667,
+      "step": 6676
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 14.236947059631348,
+      "learning_rate": 6.6172196050917114e-06,
+      "loss": 1.0438,
+      "step": 6677
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 64.36849212646484,
+      "learning_rate": 6.6152149944873215e-06,
+      "loss": 2.7021,
+      "step": 6678
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 23.676084518432617,
+      "learning_rate": 6.613210383882931e-06,
+      "loss": 1.0589,
+      "step": 6679
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 19.20448875427246,
+      "learning_rate": 6.611205773278542e-06,
+      "loss": 1.8695,
+      "step": 6680
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 10.511244773864746,
+      "learning_rate": 6.609201162674152e-06,
+      "loss": 1.1752,
+      "step": 6681
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 40.56865692138672,
+      "learning_rate": 6.607196552069761e-06,
+      "loss": 1.6004,
+      "step": 6682
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 17.472505569458008,
+      "learning_rate": 6.605191941465371e-06,
+      "loss": 1.5864,
+      "step": 6683
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 10.834254264831543,
+      "learning_rate": 6.60318733086098e-06,
+      "loss": 1.2419,
+      "step": 6684
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 16.969730377197266,
+      "learning_rate": 6.60118272025659e-06,
+      "loss": 1.2284,
+      "step": 6685
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 7.831116199493408,
+      "learning_rate": 6.5991781096522e-06,
+      "loss": 1.4888,
+      "step": 6686
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 24.744110107421875,
+      "learning_rate": 6.597173499047811e-06,
+      "loss": 1.6571,
+      "step": 6687
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 21.574138641357422,
+      "learning_rate": 6.595168888443421e-06,
+      "loss": 2.0631,
+      "step": 6688
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 11.040058135986328,
+      "learning_rate": 6.593164277839031e-06,
+      "loss": 1.1278,
+      "step": 6689
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 46.7620735168457,
+      "learning_rate": 6.59115966723464e-06,
+      "loss": 1.4132,
+      "step": 6690
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 10.559418678283691,
+      "learning_rate": 6.58915505663025e-06,
+      "loss": 0.9851,
+      "step": 6691
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 8.289081573486328,
+      "learning_rate": 6.587150446025859e-06,
+      "loss": 1.1853,
+      "step": 6692
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 8.112937927246094,
+      "learning_rate": 6.58514583542147e-06,
+      "loss": 1.0523,
+      "step": 6693
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 19.774045944213867,
+      "learning_rate": 6.58314122481708e-06,
+      "loss": 1.6196,
+      "step": 6694
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 52.585914611816406,
+      "learning_rate": 6.5811366142126896e-06,
+      "loss": 1.0791,
+      "step": 6695
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 10.37691879272461,
+      "learning_rate": 6.5791320036083e-06,
+      "loss": 0.525,
+      "step": 6696
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 13.222021102905273,
+      "learning_rate": 6.577127393003909e-06,
+      "loss": 1.0604,
+      "step": 6697
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 12.805461883544922,
+      "learning_rate": 6.575122782399519e-06,
+      "loss": 1.4706,
+      "step": 6698
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 43.80728530883789,
+      "learning_rate": 6.57311817179513e-06,
+      "loss": 2.084,
+      "step": 6699
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 15.121885299682617,
+      "learning_rate": 6.571113561190739e-06,
+      "loss": 0.8853,
+      "step": 6700
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 32.659427642822266,
+      "learning_rate": 6.569108950586349e-06,
+      "loss": 1.8597,
+      "step": 6701
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 26.797901153564453,
+      "learning_rate": 6.567104339981959e-06,
+      "loss": 2.3284,
+      "step": 6702
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 3.8142130374908447,
+      "learning_rate": 6.5650997293775685e-06,
+      "loss": 0.2054,
+      "step": 6703
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 15.68226432800293,
+      "learning_rate": 6.563095118773179e-06,
+      "loss": 2.487,
+      "step": 6704
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 42.038726806640625,
+      "learning_rate": 6.5610905081687896e-06,
+      "loss": 2.1623,
+      "step": 6705
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 19.80816650390625,
+      "learning_rate": 6.559085897564399e-06,
+      "loss": 1.0708,
+      "step": 6706
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 9.321516990661621,
+      "learning_rate": 6.557081286960009e-06,
+      "loss": 1.605,
+      "step": 6707
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 159.25819396972656,
+      "learning_rate": 6.555076676355618e-06,
+      "loss": 2.8398,
+      "step": 6708
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 17.203962326049805,
+      "learning_rate": 6.553072065751228e-06,
+      "loss": 1.4595,
+      "step": 6709
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 17.149423599243164,
+      "learning_rate": 6.5510674551468374e-06,
+      "loss": 1.7006,
+      "step": 6710
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 12.44357967376709,
+      "learning_rate": 6.549062844542448e-06,
+      "loss": 0.8718,
+      "step": 6711
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 9.552597999572754,
+      "learning_rate": 6.5470582339380585e-06,
+      "loss": 1.4398,
+      "step": 6712
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 12.471324920654297,
+      "learning_rate": 6.545053623333668e-06,
+      "loss": 2.0504,
+      "step": 6713
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 13.623446464538574,
+      "learning_rate": 6.543049012729278e-06,
+      "loss": 1.7889,
+      "step": 6714
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 8.370439529418945,
+      "learning_rate": 6.541044402124888e-06,
+      "loss": 1.1749,
+      "step": 6715
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 16.073829650878906,
+      "learning_rate": 6.539039791520497e-06,
+      "loss": 1.3742,
+      "step": 6716
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 9.64084529876709,
+      "learning_rate": 6.537035180916108e-06,
+      "loss": 1.142,
+      "step": 6717
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 25.97657585144043,
+      "learning_rate": 6.535030570311718e-06,
+      "loss": 2.3333,
+      "step": 6718
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 118.77561950683594,
+      "learning_rate": 6.533025959707327e-06,
+      "loss": 1.5912,
+      "step": 6719
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 14.903034210205078,
+      "learning_rate": 6.5310213491029374e-06,
+      "loss": 1.897,
+      "step": 6720
+    },
+    {
+      "epoch": 2.02,
+      "eval_loss": 0.16902218759059906,
+      "eval_runtime": 43.4322,
+      "eval_samples_per_second": 34.053,
+      "eval_steps_per_second": 34.053,
+      "step": 6720
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 39.25941848754883,
+      "learning_rate": 6.529016738498547e-06,
+      "loss": 1.8256,
+      "step": 6721
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 9.439811706542969,
+      "learning_rate": 6.527012127894157e-06,
+      "loss": 0.9625,
+      "step": 6722
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 12.62756061553955,
+      "learning_rate": 6.525007517289767e-06,
+      "loss": 0.8624,
+      "step": 6723
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 13.67402458190918,
+      "learning_rate": 6.523002906685377e-06,
+      "loss": 0.785,
+      "step": 6724
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 29.54031753540039,
+      "learning_rate": 6.520998296080987e-06,
+      "loss": 2.0936,
+      "step": 6725
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 13.225194931030273,
+      "learning_rate": 6.518993685476597e-06,
+      "loss": 1.2107,
+      "step": 6726
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 8.916555404663086,
+      "learning_rate": 6.516989074872206e-06,
+      "loss": 1.0452,
+      "step": 6727
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 24.967422485351562,
+      "learning_rate": 6.514984464267816e-06,
+      "loss": 2.1105,
+      "step": 6728
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 14.154083251953125,
+      "learning_rate": 6.512979853663426e-06,
+      "loss": 0.7428,
+      "step": 6729
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 17.682830810546875,
+      "learning_rate": 6.510975243059037e-06,
+      "loss": 1.3675,
+      "step": 6730
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 12.129364013671875,
+      "learning_rate": 6.508970632454647e-06,
+      "loss": 1.7799,
+      "step": 6731
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 37.760650634765625,
+      "learning_rate": 6.506966021850256e-06,
+      "loss": 1.1513,
+      "step": 6732
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 87.66400909423828,
+      "learning_rate": 6.504961411245866e-06,
+      "loss": 1.9592,
+      "step": 6733
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 12.75828742980957,
+      "learning_rate": 6.502956800641475e-06,
+      "loss": 1.1079,
+      "step": 6734
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 12.436376571655273,
+      "learning_rate": 6.500952190037085e-06,
+      "loss": 0.8629,
+      "step": 6735
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 8.984572410583496,
+      "learning_rate": 6.498947579432696e-06,
+      "loss": 0.9553,
+      "step": 6736
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 22.55635643005371,
+      "learning_rate": 6.4969429688283055e-06,
+      "loss": 1.6062,
+      "step": 6737
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 13.196118354797363,
+      "learning_rate": 6.4949383582239156e-06,
+      "loss": 0.5885,
+      "step": 6738
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 46.98003387451172,
+      "learning_rate": 6.492933747619526e-06,
+      "loss": 1.3093,
+      "step": 6739
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 23.900150299072266,
+      "learning_rate": 6.490929137015135e-06,
+      "loss": 1.4142,
+      "step": 6740
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 17.741270065307617,
+      "learning_rate": 6.488924526410745e-06,
+      "loss": 1.4268,
+      "step": 6741
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 26.587739944458008,
+      "learning_rate": 6.486919915806356e-06,
+      "loss": 2.1686,
+      "step": 6742
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 35.67372131347656,
+      "learning_rate": 6.484915305201965e-06,
+      "loss": 1.6213,
+      "step": 6743
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 26.185956954956055,
+      "learning_rate": 6.482910694597575e-06,
+      "loss": 1.1309,
+      "step": 6744
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 17.040918350219727,
+      "learning_rate": 6.4809060839931845e-06,
+      "loss": 1.0778,
+      "step": 6745
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 23.627857208251953,
+      "learning_rate": 6.4789014733887946e-06,
+      "loss": 1.1968,
+      "step": 6746
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 29.324899673461914,
+      "learning_rate": 6.476896862784405e-06,
+      "loss": 1.9221,
+      "step": 6747
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 24.806861877441406,
+      "learning_rate": 6.474892252180015e-06,
+      "loss": 1.0605,
+      "step": 6748
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 32.63029098510742,
+      "learning_rate": 6.472887641575625e-06,
+      "loss": 1.4815,
+      "step": 6749
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 17.42801284790039,
+      "learning_rate": 6.470883030971235e-06,
+      "loss": 1.6286,
+      "step": 6750
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 40.37263488769531,
+      "learning_rate": 6.468878420366844e-06,
+      "loss": 1.6338,
+      "step": 6751
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 17.74640464782715,
+      "learning_rate": 6.466873809762454e-06,
+      "loss": 1.2096,
+      "step": 6752
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 26.730159759521484,
+      "learning_rate": 6.4648691991580634e-06,
+      "loss": 2.1374,
+      "step": 6753
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 13.699724197387695,
+      "learning_rate": 6.462864588553674e-06,
+      "loss": 1.7276,
+      "step": 6754
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 26.75263023376465,
+      "learning_rate": 6.4608599779492845e-06,
+      "loss": 1.0525,
+      "step": 6755
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 24.192272186279297,
+      "learning_rate": 6.458855367344894e-06,
+      "loss": 1.4798,
+      "step": 6756
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 62.38142013549805,
+      "learning_rate": 6.456850756740504e-06,
+      "loss": 1.8068,
+      "step": 6757
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 29.921567916870117,
+      "learning_rate": 6.454846146136113e-06,
+      "loss": 1.619,
+      "step": 6758
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 6.582921981811523,
+      "learning_rate": 6.452841535531723e-06,
+      "loss": 0.648,
+      "step": 6759
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 13.459547996520996,
+      "learning_rate": 6.450836924927334e-06,
+      "loss": 1.4701,
+      "step": 6760
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 19.61090660095215,
+      "learning_rate": 6.448832314322943e-06,
+      "loss": 1.2505,
+      "step": 6761
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 32.77788543701172,
+      "learning_rate": 6.446827703718553e-06,
+      "loss": 1.9251,
+      "step": 6762
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 7.951191425323486,
+      "learning_rate": 6.4448230931141634e-06,
+      "loss": 0.7957,
+      "step": 6763
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 13.6834135055542,
+      "learning_rate": 6.442818482509773e-06,
+      "loss": 1.488,
+      "step": 6764
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 26.919858932495117,
+      "learning_rate": 6.440813871905383e-06,
+      "loss": 1.6017,
+      "step": 6765
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 18.014230728149414,
+      "learning_rate": 6.438809261300992e-06,
+      "loss": 0.8233,
+      "step": 6766
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 18.699932098388672,
+      "learning_rate": 6.436804650696603e-06,
+      "loss": 1.1873,
+      "step": 6767
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 24.57878303527832,
+      "learning_rate": 6.434800040092213e-06,
+      "loss": 1.7365,
+      "step": 6768
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 30.225631713867188,
+      "learning_rate": 6.432795429487822e-06,
+      "loss": 2.7789,
+      "step": 6769
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 22.14498519897461,
+      "learning_rate": 6.430790818883432e-06,
+      "loss": 1.6425,
+      "step": 6770
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 12.216958045959473,
+      "learning_rate": 6.428786208279042e-06,
+      "loss": 1.7486,
+      "step": 6771
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 26.501882553100586,
+      "learning_rate": 6.426781597674652e-06,
+      "loss": 1.3613,
+      "step": 6772
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 30.44036293029785,
+      "learning_rate": 6.424776987070263e-06,
+      "loss": 1.2315,
+      "step": 6773
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 19.72119140625,
+      "learning_rate": 6.422772376465873e-06,
+      "loss": 1.8577,
+      "step": 6774
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 11.062511444091797,
+      "learning_rate": 6.420767765861482e-06,
+      "loss": 1.4272,
+      "step": 6775
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 16.308460235595703,
+      "learning_rate": 6.418763155257092e-06,
+      "loss": 1.3599,
+      "step": 6776
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 35.35619354248047,
+      "learning_rate": 6.416758544652701e-06,
+      "loss": 1.477,
+      "step": 6777
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 13.431120872497559,
+      "learning_rate": 6.414753934048311e-06,
+      "loss": 0.9098,
+      "step": 6778
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 22.69354820251465,
+      "learning_rate": 6.412749323443922e-06,
+      "loss": 1.3123,
+      "step": 6779
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 41.717098236083984,
+      "learning_rate": 6.4107447128395315e-06,
+      "loss": 2.688,
+      "step": 6780
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 9.498014450073242,
+      "learning_rate": 6.4087401022351416e-06,
+      "loss": 0.8678,
+      "step": 6781
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 10.49351692199707,
+      "learning_rate": 6.406735491630751e-06,
+      "loss": 1.8695,
+      "step": 6782
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 31.190580368041992,
+      "learning_rate": 6.404730881026361e-06,
+      "loss": 1.78,
+      "step": 6783
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 15.785634994506836,
+      "learning_rate": 6.402726270421971e-06,
+      "loss": 1.0891,
+      "step": 6784
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 14.310815811157227,
+      "learning_rate": 6.400721659817581e-06,
+      "loss": 0.9196,
+      "step": 6785
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 38.41286087036133,
+      "learning_rate": 6.398717049213191e-06,
+      "loss": 2.4826,
+      "step": 6786
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 10.610367774963379,
+      "learning_rate": 6.396712438608801e-06,
+      "loss": 1.0001,
+      "step": 6787
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 9.231528282165527,
+      "learning_rate": 6.3947078280044105e-06,
+      "loss": 1.0759,
+      "step": 6788
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 7.4241862297058105,
+      "learning_rate": 6.3927032174000206e-06,
+      "loss": 1.9806,
+      "step": 6789
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 14.655599594116211,
+      "learning_rate": 6.39069860679563e-06,
+      "loss": 1.3411,
+      "step": 6790
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 10.586833000183105,
+      "learning_rate": 6.388693996191241e-06,
+      "loss": 0.9932,
+      "step": 6791
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 19.10291290283203,
+      "learning_rate": 6.386689385586851e-06,
+      "loss": 1.5976,
+      "step": 6792
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 15.874801635742188,
+      "learning_rate": 6.38468477498246e-06,
+      "loss": 0.897,
+      "step": 6793
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 13.638111114501953,
+      "learning_rate": 6.38268016437807e-06,
+      "loss": 0.7186,
+      "step": 6794
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 9.389366149902344,
+      "learning_rate": 6.380675553773679e-06,
+      "loss": 0.9544,
+      "step": 6795
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 26.145200729370117,
+      "learning_rate": 6.3786709431692894e-06,
+      "loss": 1.3817,
+      "step": 6796
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 12.559453010559082,
+      "learning_rate": 6.3766663325649e-06,
+      "loss": 1.5977,
+      "step": 6797
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 9.843866348266602,
+      "learning_rate": 6.37466172196051e-06,
+      "loss": 0.9535,
+      "step": 6798
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 15.275118827819824,
+      "learning_rate": 6.37265711135612e-06,
+      "loss": 1.3151,
+      "step": 6799
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 18.97870635986328,
+      "learning_rate": 6.37065250075173e-06,
+      "loss": 1.6633,
+      "step": 6800
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 11.981043815612793,
+      "learning_rate": 6.368647890147339e-06,
+      "loss": 1.0551,
+      "step": 6801
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 25.228471755981445,
+      "learning_rate": 6.366643279542949e-06,
+      "loss": 1.4122,
+      "step": 6802
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 46.05463790893555,
+      "learning_rate": 6.36463866893856e-06,
+      "loss": 2.192,
+      "step": 6803
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 28.93854522705078,
+      "learning_rate": 6.362634058334169e-06,
+      "loss": 1.235,
+      "step": 6804
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 22.576128005981445,
+      "learning_rate": 6.360629447729779e-06,
+      "loss": 1.61,
+      "step": 6805
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 8.201756477355957,
+      "learning_rate": 6.358624837125389e-06,
+      "loss": 1.3716,
+      "step": 6806
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 9.097603797912598,
+      "learning_rate": 6.356620226520999e-06,
+      "loss": 0.69,
+      "step": 6807
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 26.398914337158203,
+      "learning_rate": 6.354615615916609e-06,
+      "loss": 2.6075,
+      "step": 6808
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 15.047880172729492,
+      "learning_rate": 6.352611005312218e-06,
+      "loss": 1.2919,
+      "step": 6809
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 42.7448616027832,
+      "learning_rate": 6.350606394707829e-06,
+      "loss": 1.1902,
+      "step": 6810
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 29.7187557220459,
+      "learning_rate": 6.348601784103439e-06,
+      "loss": 1.9144,
+      "step": 6811
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 13.130889892578125,
+      "learning_rate": 6.346597173499048e-06,
+      "loss": 0.9165,
+      "step": 6812
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 14.863086700439453,
+      "learning_rate": 6.344592562894658e-06,
+      "loss": 2.3026,
+      "step": 6813
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 31.419349670410156,
+      "learning_rate": 6.342587952290268e-06,
+      "loss": 1.929,
+      "step": 6814
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 16.03536605834961,
+      "learning_rate": 6.340583341685878e-06,
+      "loss": 1.0579,
+      "step": 6815
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 9.93125057220459,
+      "learning_rate": 6.338578731081489e-06,
+      "loss": 1.0248,
+      "step": 6816
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 20.598365783691406,
+      "learning_rate": 6.336574120477098e-06,
+      "loss": 1.0575,
+      "step": 6817
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 24.11957359313965,
+      "learning_rate": 6.334569509872708e-06,
+      "loss": 2.0264,
+      "step": 6818
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 12.065218925476074,
+      "learning_rate": 6.332564899268317e-06,
+      "loss": 0.6793,
+      "step": 6819
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 43.54728698730469,
+      "learning_rate": 6.330560288663927e-06,
+      "loss": 2.746,
+      "step": 6820
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 12.573387145996094,
+      "learning_rate": 6.328555678059537e-06,
+      "loss": 1.0912,
+      "step": 6821
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 53.12962341308594,
+      "learning_rate": 6.326551067455147e-06,
+      "loss": 2.5651,
+      "step": 6822
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 27.892974853515625,
+      "learning_rate": 6.3245464568507575e-06,
+      "loss": 2.1102,
+      "step": 6823
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 11.939574241638184,
+      "learning_rate": 6.3225418462463676e-06,
+      "loss": 1.0358,
+      "step": 6824
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 15.528552055358887,
+      "learning_rate": 6.320537235641977e-06,
+      "loss": 1.43,
+      "step": 6825
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 19.028608322143555,
+      "learning_rate": 6.318532625037587e-06,
+      "loss": 1.3059,
+      "step": 6826
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 29.945940017700195,
+      "learning_rate": 6.316528014433196e-06,
+      "loss": 1.4528,
+      "step": 6827
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 8.454259872436523,
+      "learning_rate": 6.314523403828807e-06,
+      "loss": 0.9366,
+      "step": 6828
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 19.378828048706055,
+      "learning_rate": 6.312518793224417e-06,
+      "loss": 0.8679,
+      "step": 6829
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 20.16804313659668,
+      "learning_rate": 6.310514182620026e-06,
+      "loss": 1.6893,
+      "step": 6830
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 9.753872871398926,
+      "learning_rate": 6.3085095720156365e-06,
+      "loss": 1.1253,
+      "step": 6831
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 14.549400329589844,
+      "learning_rate": 6.306504961411246e-06,
+      "loss": 1.3079,
+      "step": 6832
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 15.794536590576172,
+      "learning_rate": 6.304500350806856e-06,
+      "loss": 1.394,
+      "step": 6833
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 32.083927154541016,
+      "learning_rate": 6.302495740202467e-06,
+      "loss": 1.716,
+      "step": 6834
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 12.402176856994629,
+      "learning_rate": 6.300491129598077e-06,
+      "loss": 1.0654,
+      "step": 6835
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 16.35983657836914,
+      "learning_rate": 6.298486518993686e-06,
+      "loss": 1.1802,
+      "step": 6836
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 27.28036117553711,
+      "learning_rate": 6.296481908389296e-06,
+      "loss": 1.8658,
+      "step": 6837
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 10.746267318725586,
+      "learning_rate": 6.294477297784905e-06,
+      "loss": 1.2586,
+      "step": 6838
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 27.72111701965332,
+      "learning_rate": 6.2924726871805155e-06,
+      "loss": 2.0948,
+      "step": 6839
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 13.532320976257324,
+      "learning_rate": 6.290468076576126e-06,
+      "loss": 1.2246,
+      "step": 6840
+    },
+    {
+      "epoch": 2.06,
+      "eval_loss": 0.18061862885951996,
+      "eval_runtime": 43.8871,
+      "eval_samples_per_second": 33.7,
+      "eval_steps_per_second": 33.7,
+      "step": 6840
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 23.459665298461914,
+      "learning_rate": 6.288463465971736e-06,
+      "loss": 1.4494,
+      "step": 6841
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 13.611770629882812,
+      "learning_rate": 6.286458855367346e-06,
+      "loss": 1.1819,
+      "step": 6842
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 25.924118041992188,
+      "learning_rate": 6.284454244762955e-06,
+      "loss": 1.0744,
+      "step": 6843
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 19.138996124267578,
+      "learning_rate": 6.282449634158565e-06,
+      "loss": 1.2561,
+      "step": 6844
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 46.55351638793945,
+      "learning_rate": 6.280445023554175e-06,
+      "loss": 1.4002,
+      "step": 6845
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 16.169979095458984,
+      "learning_rate": 6.278440412949785e-06,
+      "loss": 1.2913,
+      "step": 6846
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 37.923133850097656,
+      "learning_rate": 6.276435802345395e-06,
+      "loss": 1.5886,
+      "step": 6847
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 57.48485565185547,
+      "learning_rate": 6.274431191741005e-06,
+      "loss": 2.0952,
+      "step": 6848
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 13.789143562316895,
+      "learning_rate": 6.272426581136615e-06,
+      "loss": 1.712,
+      "step": 6849
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 21.859037399291992,
+      "learning_rate": 6.270421970532225e-06,
+      "loss": 2.0063,
+      "step": 6850
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 13.258837699890137,
+      "learning_rate": 6.268417359927834e-06,
+      "loss": 1.2955,
+      "step": 6851
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 8.517374038696289,
+      "learning_rate": 6.266412749323444e-06,
+      "loss": 1.0369,
+      "step": 6852
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 10.359166145324707,
+      "learning_rate": 6.264408138719055e-06,
+      "loss": 1.0215,
+      "step": 6853
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 28.84781837463379,
+      "learning_rate": 6.262403528114664e-06,
+      "loss": 1.9747,
+      "step": 6854
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 11.506865501403809,
+      "learning_rate": 6.260398917510274e-06,
+      "loss": 1.1076,
+      "step": 6855
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 15.466108322143555,
+      "learning_rate": 6.2583943069058835e-06,
+      "loss": 1.2779,
+      "step": 6856
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 31.154294967651367,
+      "learning_rate": 6.256389696301494e-06,
+      "loss": 1.9285,
+      "step": 6857
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 15.335009574890137,
+      "learning_rate": 6.254385085697104e-06,
+      "loss": 1.6327,
+      "step": 6858
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 14.675117492675781,
+      "learning_rate": 6.252380475092715e-06,
+      "loss": 1.644,
+      "step": 6859
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 133.6639862060547,
+      "learning_rate": 6.250375864488324e-06,
+      "loss": 1.1432,
+      "step": 6860
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 39.901222229003906,
+      "learning_rate": 6.248371253883934e-06,
+      "loss": 2.097,
+      "step": 6861
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 13.691509246826172,
+      "learning_rate": 6.246366643279543e-06,
+      "loss": 1.6336,
+      "step": 6862
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 14.485090255737305,
+      "learning_rate": 6.244362032675153e-06,
+      "loss": 1.3259,
+      "step": 6863
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 72.5770492553711,
+      "learning_rate": 6.2423574220707625e-06,
+      "loss": 1.6907,
+      "step": 6864
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 10.634366989135742,
+      "learning_rate": 6.240352811466373e-06,
+      "loss": 0.5027,
+      "step": 6865
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 33.018367767333984,
+      "learning_rate": 6.2383482008619835e-06,
+      "loss": 1.7409,
+      "step": 6866
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 9.623235702514648,
+      "learning_rate": 6.236343590257593e-06,
+      "loss": 1.715,
+      "step": 6867
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 47.73927307128906,
+      "learning_rate": 6.234338979653203e-06,
+      "loss": 1.4383,
+      "step": 6868
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 9.162986755371094,
+      "learning_rate": 6.232334369048813e-06,
+      "loss": 1.1415,
+      "step": 6869
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 41.81623077392578,
+      "learning_rate": 6.230329758444422e-06,
+      "loss": 1.6832,
+      "step": 6870
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 21.047195434570312,
+      "learning_rate": 6.228325147840033e-06,
+      "loss": 1.202,
+      "step": 6871
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 19.954036712646484,
+      "learning_rate": 6.226320537235643e-06,
+      "loss": 1.4934,
+      "step": 6872
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 24.60007095336914,
+      "learning_rate": 6.224315926631252e-06,
+      "loss": 0.9536,
+      "step": 6873
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 23.833621978759766,
+      "learning_rate": 6.2223113160268625e-06,
+      "loss": 2.561,
+      "step": 6874
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 51.6407585144043,
+      "learning_rate": 6.220306705422472e-06,
+      "loss": 1.8729,
+      "step": 6875
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 9.72765827178955,
+      "learning_rate": 6.218302094818082e-06,
+      "loss": 1.6889,
+      "step": 6876
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 86.259765625,
+      "learning_rate": 6.216297484213693e-06,
+      "loss": 1.6229,
+      "step": 6877
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 12.195944786071777,
+      "learning_rate": 6.214292873609302e-06,
+      "loss": 1.7568,
+      "step": 6878
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 41.19993591308594,
+      "learning_rate": 6.212288263004912e-06,
+      "loss": 1.2259,
+      "step": 6879
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 10.510581970214844,
+      "learning_rate": 6.210283652400521e-06,
+      "loss": 1.1598,
+      "step": 6880
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 14.523313522338867,
+      "learning_rate": 6.208279041796131e-06,
+      "loss": 1.3958,
+      "step": 6881
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 11.1996431350708,
+      "learning_rate": 6.2062744311917415e-06,
+      "loss": 1.6634,
+      "step": 6882
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 17.171239852905273,
+      "learning_rate": 6.2042698205873515e-06,
+      "loss": 1.8742,
+      "step": 6883
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 83.33916473388672,
+      "learning_rate": 6.202265209982962e-06,
+      "loss": 2.0191,
+      "step": 6884
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 13.033934593200684,
+      "learning_rate": 6.200260599378572e-06,
+      "loss": 0.9672,
+      "step": 6885
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 25.13782501220703,
+      "learning_rate": 6.198255988774181e-06,
+      "loss": 1.5467,
+      "step": 6886
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 129.21408081054688,
+      "learning_rate": 6.196251378169791e-06,
+      "loss": 1.8225,
+      "step": 6887
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 15.809123992919922,
+      "learning_rate": 6.1942467675654e-06,
+      "loss": 1.1815,
+      "step": 6888
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 10.88241195678711,
+      "learning_rate": 6.19224215696101e-06,
+      "loss": 0.5932,
+      "step": 6889
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 17.281024932861328,
+      "learning_rate": 6.190237546356621e-06,
+      "loss": 0.6785,
+      "step": 6890
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 78.56420135498047,
+      "learning_rate": 6.1882329357522305e-06,
+      "loss": 2.5001,
+      "step": 6891
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 7.025608539581299,
+      "learning_rate": 6.186228325147841e-06,
+      "loss": 0.9645,
+      "step": 6892
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 17.67293930053711,
+      "learning_rate": 6.184223714543451e-06,
+      "loss": 1.5476,
+      "step": 6893
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 54.84214782714844,
+      "learning_rate": 6.18221910393906e-06,
+      "loss": 2.5763,
+      "step": 6894
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 16.025243759155273,
+      "learning_rate": 6.18021449333467e-06,
+      "loss": 1.0716,
+      "step": 6895
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 19.868698120117188,
+      "learning_rate": 6.178209882730281e-06,
+      "loss": 1.6816,
+      "step": 6896
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 23.872350692749023,
+      "learning_rate": 6.17620527212589e-06,
+      "loss": 1.1166,
+      "step": 6897
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 17.220922470092773,
+      "learning_rate": 6.1742006615215e-06,
+      "loss": 1.601,
+      "step": 6898
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 37.89704895019531,
+      "learning_rate": 6.1721960509171095e-06,
+      "loss": 1.2833,
+      "step": 6899
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 10.233612060546875,
+      "learning_rate": 6.17019144031272e-06,
+      "loss": 1.137,
+      "step": 6900
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 14.788456916809082,
+      "learning_rate": 6.168186829708329e-06,
+      "loss": 0.7968,
+      "step": 6901
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 47.135398864746094,
+      "learning_rate": 6.16618221910394e-06,
+      "loss": 1.8272,
+      "step": 6902
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 48.95859146118164,
+      "learning_rate": 6.16417760849955e-06,
+      "loss": 1.8223,
+      "step": 6903
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 9.25454044342041,
+      "learning_rate": 6.162172997895159e-06,
+      "loss": 0.9862,
+      "step": 6904
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 34.636192321777344,
+      "learning_rate": 6.160168387290769e-06,
+      "loss": 1.5629,
+      "step": 6905
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 51.81803512573242,
+      "learning_rate": 6.158163776686379e-06,
+      "loss": 1.5237,
+      "step": 6906
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 39.07495880126953,
+      "learning_rate": 6.1561591660819885e-06,
+      "loss": 1.1899,
+      "step": 6907
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 22.247758865356445,
+      "learning_rate": 6.154154555477599e-06,
+      "loss": 0.9568,
+      "step": 6908
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 15.606700897216797,
+      "learning_rate": 6.1521499448732095e-06,
+      "loss": 1.5251,
+      "step": 6909
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 17.546934127807617,
+      "learning_rate": 6.150145334268819e-06,
+      "loss": 1.1845,
+      "step": 6910
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 7.191950798034668,
+      "learning_rate": 6.148140723664429e-06,
+      "loss": 0.7725,
+      "step": 6911
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 10.01009464263916,
+      "learning_rate": 6.146136113060038e-06,
+      "loss": 1.0729,
+      "step": 6912
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 9.743487358093262,
+      "learning_rate": 6.144131502455648e-06,
+      "loss": 1.2371,
+      "step": 6913
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 21.309419631958008,
+      "learning_rate": 6.142126891851259e-06,
+      "loss": 2.0328,
+      "step": 6914
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 14.136860847473145,
+      "learning_rate": 6.140122281246868e-06,
+      "loss": 1.4086,
+      "step": 6915
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 12.133606910705566,
+      "learning_rate": 6.138117670642478e-06,
+      "loss": 1.3393,
+      "step": 6916
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 8.155012130737305,
+      "learning_rate": 6.136113060038088e-06,
+      "loss": 0.8114,
+      "step": 6917
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 14.2897367477417,
+      "learning_rate": 6.134108449433698e-06,
+      "loss": 1.0815,
+      "step": 6918
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 37.01668930053711,
+      "learning_rate": 6.132103838829308e-06,
+      "loss": 2.4772,
+      "step": 6919
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 9.246944427490234,
+      "learning_rate": 6.130099228224919e-06,
+      "loss": 1.0533,
+      "step": 6920
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 14.934267044067383,
+      "learning_rate": 6.128094617620528e-06,
+      "loss": 0.8619,
+      "step": 6921
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 25.984840393066406,
+      "learning_rate": 6.126090007016138e-06,
+      "loss": 1.3991,
+      "step": 6922
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 23.823701858520508,
+      "learning_rate": 6.124085396411747e-06,
+      "loss": 2.2563,
+      "step": 6923
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 25.616357803344727,
+      "learning_rate": 6.122080785807357e-06,
+      "loss": 1.559,
+      "step": 6924
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 25.483346939086914,
+      "learning_rate": 6.120076175202967e-06,
+      "loss": 1.8079,
+      "step": 6925
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 26.09974479675293,
+      "learning_rate": 6.1180715645985775e-06,
+      "loss": 1.8173,
+      "step": 6926
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 33.826454162597656,
+      "learning_rate": 6.116066953994188e-06,
+      "loss": 1.2576,
+      "step": 6927
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 22.255189895629883,
+      "learning_rate": 6.114062343389797e-06,
+      "loss": 1.5849,
+      "step": 6928
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 23.203243255615234,
+      "learning_rate": 6.112057732785407e-06,
+      "loss": 2.0366,
+      "step": 6929
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 48.67564010620117,
+      "learning_rate": 6.110053122181017e-06,
+      "loss": 1.4752,
+      "step": 6930
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 13.649922370910645,
+      "learning_rate": 6.108048511576626e-06,
+      "loss": 1.0061,
+      "step": 6931
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 15.346248626708984,
+      "learning_rate": 6.106043900972236e-06,
+      "loss": 0.975,
+      "step": 6932
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 43.33583450317383,
+      "learning_rate": 6.104039290367847e-06,
+      "loss": 1.8835,
+      "step": 6933
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 11.801863670349121,
+      "learning_rate": 6.1020346797634565e-06,
+      "loss": 0.9421,
+      "step": 6934
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 13.316075325012207,
+      "learning_rate": 6.100030069159067e-06,
+      "loss": 0.9621,
+      "step": 6935
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 7.643853187561035,
+      "learning_rate": 6.098025458554676e-06,
+      "loss": 0.7109,
+      "step": 6936
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 6.656645774841309,
+      "learning_rate": 6.096020847950286e-06,
+      "loss": 0.9677,
+      "step": 6937
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 8.41319751739502,
+      "learning_rate": 6.094016237345895e-06,
+      "loss": 1.4541,
+      "step": 6938
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 17.172225952148438,
+      "learning_rate": 6.092011626741506e-06,
+      "loss": 1.2853,
+      "step": 6939
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 13.03164005279541,
+      "learning_rate": 6.090007016137116e-06,
+      "loss": 1.2991,
+      "step": 6940
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 26.83146095275879,
+      "learning_rate": 6.088002405532725e-06,
+      "loss": 1.2144,
+      "step": 6941
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 15.566588401794434,
+      "learning_rate": 6.0859977949283355e-06,
+      "loss": 1.1742,
+      "step": 6942
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 11.731130599975586,
+      "learning_rate": 6.083993184323946e-06,
+      "loss": 1.5685,
+      "step": 6943
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 23.45868492126465,
+      "learning_rate": 6.081988573719555e-06,
+      "loss": 1.1652,
+      "step": 6944
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 30.693378448486328,
+      "learning_rate": 6.079983963115166e-06,
+      "loss": 1.1192,
+      "step": 6945
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 76.05176544189453,
+      "learning_rate": 6.077979352510776e-06,
+      "loss": 1.03,
+      "step": 6946
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 42.937931060791016,
+      "learning_rate": 6.075974741906385e-06,
+      "loss": 2.9801,
+      "step": 6947
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 21.51824188232422,
+      "learning_rate": 6.073970131301995e-06,
+      "loss": 1.3744,
+      "step": 6948
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 13.035652160644531,
+      "learning_rate": 6.071965520697604e-06,
+      "loss": 1.2832,
+      "step": 6949
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 7.819190502166748,
+      "learning_rate": 6.0699609100932145e-06,
+      "loss": 1.3225,
+      "step": 6950
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 10.307256698608398,
+      "learning_rate": 6.067956299488825e-06,
+      "loss": 1.1998,
+      "step": 6951
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 43.879268646240234,
+      "learning_rate": 6.065951688884435e-06,
+      "loss": 1.2899,
+      "step": 6952
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 14.263348579406738,
+      "learning_rate": 6.063947078280045e-06,
+      "loss": 0.8165,
+      "step": 6953
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 19.4408016204834,
+      "learning_rate": 6.061942467675655e-06,
+      "loss": 1.2222,
+      "step": 6954
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 21.61421012878418,
+      "learning_rate": 6.059937857071264e-06,
+      "loss": 1.6328,
+      "step": 6955
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 59.171634674072266,
+      "learning_rate": 6.057933246466874e-06,
+      "loss": 2.2851,
+      "step": 6956
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 17.417770385742188,
+      "learning_rate": 6.055928635862485e-06,
+      "loss": 1.5558,
+      "step": 6957
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 21.351945877075195,
+      "learning_rate": 6.053924025258094e-06,
+      "loss": 1.9198,
+      "step": 6958
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 21.162813186645508,
+      "learning_rate": 6.051919414653704e-06,
+      "loss": 0.8733,
+      "step": 6959
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 28.082653045654297,
+      "learning_rate": 6.049914804049314e-06,
+      "loss": 1.3162,
+      "step": 6960
+    },
+    {
+      "epoch": 2.09,
+      "eval_loss": 0.1762227714061737,
+      "eval_runtime": 43.6018,
+      "eval_samples_per_second": 33.921,
+      "eval_steps_per_second": 33.921,
+      "step": 6960
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 17.02800941467285,
+      "learning_rate": 6.047910193444924e-06,
+      "loss": 1.4449,
+      "step": 6961
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 48.51543045043945,
+      "learning_rate": 6.045905582840533e-06,
+      "loss": 1.0376,
+      "step": 6962
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 30.203258514404297,
+      "learning_rate": 6.043900972236144e-06,
+      "loss": 1.8892,
+      "step": 6963
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 22.062679290771484,
+      "learning_rate": 6.041896361631754e-06,
+      "loss": 2.034,
+      "step": 6964
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 29.935195922851562,
+      "learning_rate": 6.039891751027363e-06,
+      "loss": 2.1552,
+      "step": 6965
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 21.121536254882812,
+      "learning_rate": 6.037887140422973e-06,
+      "loss": 1.848,
+      "step": 6966
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 13.469535827636719,
+      "learning_rate": 6.035882529818583e-06,
+      "loss": 0.8718,
+      "step": 6967
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 9.78769302368164,
+      "learning_rate": 6.033877919214193e-06,
+      "loss": 0.5627,
+      "step": 6968
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 11.262333869934082,
+      "learning_rate": 6.0318733086098035e-06,
+      "loss": 1.4755,
+      "step": 6969
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 21.476552963256836,
+      "learning_rate": 6.029868698005414e-06,
+      "loss": 0.9482,
+      "step": 6970
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 8.83089542388916,
+      "learning_rate": 6.027864087401023e-06,
+      "loss": 0.9499,
+      "step": 6971
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 16.572052001953125,
+      "learning_rate": 6.025859476796633e-06,
+      "loss": 1.4945,
+      "step": 6972
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 94.96258544921875,
+      "learning_rate": 6.023854866192242e-06,
+      "loss": 1.655,
+      "step": 6973
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 11.988658905029297,
+      "learning_rate": 6.021850255587852e-06,
+      "loss": 1.4834,
+      "step": 6974
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 15.83250904083252,
+      "learning_rate": 6.0198456449834615e-06,
+      "loss": 1.7121,
+      "step": 6975
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 14.637907981872559,
+      "learning_rate": 6.0178410343790724e-06,
+      "loss": 1.849,
+      "step": 6976
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 42.469970703125,
+      "learning_rate": 6.0158364237746825e-06,
+      "loss": 1.3548,
+      "step": 6977
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 14.263885498046875,
+      "learning_rate": 6.013831813170293e-06,
+      "loss": 1.5441,
+      "step": 6978
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 91.35665130615234,
+      "learning_rate": 6.011827202565902e-06,
+      "loss": 2.4573,
+      "step": 6979
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 26.59601402282715,
+      "learning_rate": 6.009822591961512e-06,
+      "loss": 1.0937,
+      "step": 6980
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 21.7692928314209,
+      "learning_rate": 6.007817981357121e-06,
+      "loss": 1.4581,
+      "step": 6981
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 55.85102081298828,
+      "learning_rate": 6.005813370752732e-06,
+      "loss": 1.5007,
+      "step": 6982
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 17.059696197509766,
+      "learning_rate": 6.003808760148342e-06,
+      "loss": 1.2327,
+      "step": 6983
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 20.334264755249023,
+      "learning_rate": 6.001804149543951e-06,
+      "loss": 1.1726,
+      "step": 6984
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 14.4188871383667,
+      "learning_rate": 5.9997995389395615e-06,
+      "loss": 1.4264,
+      "step": 6985
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 10.726988792419434,
+      "learning_rate": 5.997794928335171e-06,
+      "loss": 0.912,
+      "step": 6986
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 22.09978675842285,
+      "learning_rate": 5.995790317730781e-06,
+      "loss": 1.9006,
+      "step": 6987
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 25.827951431274414,
+      "learning_rate": 5.993785707126392e-06,
+      "loss": 2.7776,
+      "step": 6988
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 43.5711784362793,
+      "learning_rate": 5.991781096522001e-06,
+      "loss": 1.5516,
+      "step": 6989
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 20.165645599365234,
+      "learning_rate": 5.989776485917611e-06,
+      "loss": 1.1189,
+      "step": 6990
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 23.63228988647461,
+      "learning_rate": 5.987771875313221e-06,
+      "loss": 1.4044,
+      "step": 6991
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 18.94355583190918,
+      "learning_rate": 5.98576726470883e-06,
+      "loss": 1.3975,
+      "step": 6992
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 43.25507736206055,
+      "learning_rate": 5.9837626541044405e-06,
+      "loss": 2.5693,
+      "step": 6993
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 24.190704345703125,
+      "learning_rate": 5.981758043500051e-06,
+      "loss": 1.0735,
+      "step": 6994
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 16.01865577697754,
+      "learning_rate": 5.979753432895661e-06,
+      "loss": 1.2435,
+      "step": 6995
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 16.649044036865234,
+      "learning_rate": 5.977748822291271e-06,
+      "loss": 1.3125,
+      "step": 6996
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 17.755905151367188,
+      "learning_rate": 5.97574421168688e-06,
+      "loss": 1.6912,
+      "step": 6997
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 18.950183868408203,
+      "learning_rate": 5.97373960108249e-06,
+      "loss": 1.9294,
+      "step": 6998
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 9.899605751037598,
+      "learning_rate": 5.971734990478099e-06,
+      "loss": 1.6834,
+      "step": 6999
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 16.555978775024414,
+      "learning_rate": 5.96973037987371e-06,
+      "loss": 1.773,
+      "step": 7000
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 9.918785095214844,
+      "learning_rate": 5.96772576926932e-06,
+      "loss": 0.8201,
+      "step": 7001
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 7.417164325714111,
+      "learning_rate": 5.9657211586649295e-06,
+      "loss": 0.8608,
+      "step": 7002
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 31.227968215942383,
+      "learning_rate": 5.96371654806054e-06,
+      "loss": 1.3737,
+      "step": 7003
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 16.146133422851562,
+      "learning_rate": 5.96171193745615e-06,
+      "loss": 1.6191,
+      "step": 7004
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 11.509698867797852,
+      "learning_rate": 5.959707326851759e-06,
+      "loss": 1.118,
+      "step": 7005
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 11.928789138793945,
+      "learning_rate": 5.95770271624737e-06,
+      "loss": 1.3745,
+      "step": 7006
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 12.056342124938965,
+      "learning_rate": 5.95569810564298e-06,
+      "loss": 2.4071,
+      "step": 7007
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 15.146832466125488,
+      "learning_rate": 5.953693495038589e-06,
+      "loss": 0.6288,
+      "step": 7008
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 15.896178245544434,
+      "learning_rate": 5.951688884434199e-06,
+      "loss": 1.3848,
+      "step": 7009
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 20.91379737854004,
+      "learning_rate": 5.9496842738298085e-06,
+      "loss": 1.9552,
+      "step": 7010
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 15.576189041137695,
+      "learning_rate": 5.947679663225419e-06,
+      "loss": 1.7155,
+      "step": 7011
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 12.269247055053711,
+      "learning_rate": 5.9456750526210295e-06,
+      "loss": 1.1991,
+      "step": 7012
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 11.45450210571289,
+      "learning_rate": 5.943670442016639e-06,
+      "loss": 0.6404,
+      "step": 7013
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 14.839483261108398,
+      "learning_rate": 5.941665831412249e-06,
+      "loss": 1.1984,
+      "step": 7014
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 11.097734451293945,
+      "learning_rate": 5.939661220807859e-06,
+      "loss": 1.0033,
+      "step": 7015
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 23.61093521118164,
+      "learning_rate": 5.937656610203468e-06,
+      "loss": 2.0833,
+      "step": 7016
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 16.99240493774414,
+      "learning_rate": 5.935651999599078e-06,
+      "loss": 1.4891,
+      "step": 7017
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 11.620327949523926,
+      "learning_rate": 5.9336473889946875e-06,
+      "loss": 0.9176,
+      "step": 7018
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 17.769229888916016,
+      "learning_rate": 5.9316427783902984e-06,
+      "loss": 2.0986,
+      "step": 7019
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 9.187108039855957,
+      "learning_rate": 5.9296381677859085e-06,
+      "loss": 1.0068,
+      "step": 7020
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 47.74100112915039,
+      "learning_rate": 5.927633557181518e-06,
+      "loss": 1.3823,
+      "step": 7021
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 20.690021514892578,
+      "learning_rate": 5.925628946577128e-06,
+      "loss": 1.5737,
+      "step": 7022
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 19.90296745300293,
+      "learning_rate": 5.923624335972737e-06,
+      "loss": 1.6142,
+      "step": 7023
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 19.526081085205078,
+      "learning_rate": 5.921619725368347e-06,
+      "loss": 0.9043,
+      "step": 7024
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 47.775001525878906,
+      "learning_rate": 5.919615114763958e-06,
+      "loss": 3.4549,
+      "step": 7025
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 12.722251892089844,
+      "learning_rate": 5.917610504159567e-06,
+      "loss": 1.7378,
+      "step": 7026
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 10.309690475463867,
+      "learning_rate": 5.915605893555177e-06,
+      "loss": 0.9647,
+      "step": 7027
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 10.273581504821777,
+      "learning_rate": 5.9136012829507875e-06,
+      "loss": 1.6892,
+      "step": 7028
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 16.655803680419922,
+      "learning_rate": 5.911596672346397e-06,
+      "loss": 1.6853,
+      "step": 7029
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 23.953218460083008,
+      "learning_rate": 5.909592061742007e-06,
+      "loss": 0.5064,
+      "step": 7030
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 21.488235473632812,
+      "learning_rate": 5.907587451137618e-06,
+      "loss": 1.7174,
+      "step": 7031
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 54.67894744873047,
+      "learning_rate": 5.905582840533227e-06,
+      "loss": 1.7804,
+      "step": 7032
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 10.905550956726074,
+      "learning_rate": 5.903578229928837e-06,
+      "loss": 0.7826,
+      "step": 7033
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 6.627830982208252,
+      "learning_rate": 5.901573619324446e-06,
+      "loss": 0.9134,
+      "step": 7034
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 12.987330436706543,
+      "learning_rate": 5.899569008720056e-06,
+      "loss": 1.4485,
+      "step": 7035
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 14.574166297912598,
+      "learning_rate": 5.897564398115666e-06,
+      "loss": 1.1571,
+      "step": 7036
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 12.813142776489258,
+      "learning_rate": 5.8955597875112766e-06,
+      "loss": 2.2213,
+      "step": 7037
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 96.96940612792969,
+      "learning_rate": 5.893555176906887e-06,
+      "loss": 2.6402,
+      "step": 7038
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 22.448095321655273,
+      "learning_rate": 5.891550566302497e-06,
+      "loss": 1.2935,
+      "step": 7039
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 8.147926330566406,
+      "learning_rate": 5.889545955698106e-06,
+      "loss": 0.8031,
+      "step": 7040
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 16.05350685119629,
+      "learning_rate": 5.887541345093716e-06,
+      "loss": 1.6077,
+      "step": 7041
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 6.752038955688477,
+      "learning_rate": 5.885536734489325e-06,
+      "loss": 0.6485,
+      "step": 7042
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 13.922531127929688,
+      "learning_rate": 5.883532123884936e-06,
+      "loss": 1.3103,
+      "step": 7043
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 26.871532440185547,
+      "learning_rate": 5.881527513280546e-06,
+      "loss": 1.4686,
+      "step": 7044
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 10.8401460647583,
+      "learning_rate": 5.8795229026761555e-06,
+      "loss": 1.5021,
+      "step": 7045
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 22.18703842163086,
+      "learning_rate": 5.877518292071766e-06,
+      "loss": 1.2032,
+      "step": 7046
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 14.264338493347168,
+      "learning_rate": 5.875513681467375e-06,
+      "loss": 1.3754,
+      "step": 7047
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 14.007697105407715,
+      "learning_rate": 5.873509070862985e-06,
+      "loss": 1.549,
+      "step": 7048
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 45.7745361328125,
+      "learning_rate": 5.871504460258596e-06,
+      "loss": 1.5497,
+      "step": 7049
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 13.82027530670166,
+      "learning_rate": 5.869499849654205e-06,
+      "loss": 0.9325,
+      "step": 7050
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 28.13387107849121,
+      "learning_rate": 5.867495239049815e-06,
+      "loss": 1.5999,
+      "step": 7051
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 6.187619209289551,
+      "learning_rate": 5.865490628445425e-06,
+      "loss": 0.843,
+      "step": 7052
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 16.584333419799805,
+      "learning_rate": 5.8634860178410345e-06,
+      "loss": 1.3149,
+      "step": 7053
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 36.140869140625,
+      "learning_rate": 5.861481407236645e-06,
+      "loss": 1.3137,
+      "step": 7054
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 25.52848243713379,
+      "learning_rate": 5.8594767966322555e-06,
+      "loss": 2.0762,
+      "step": 7055
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 36.224937438964844,
+      "learning_rate": 5.857472186027865e-06,
+      "loss": 1.4368,
+      "step": 7056
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 37.73203659057617,
+      "learning_rate": 5.855467575423475e-06,
+      "loss": 2.0182,
+      "step": 7057
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 10.022035598754883,
+      "learning_rate": 5.853462964819084e-06,
+      "loss": 0.9173,
+      "step": 7058
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 31.495861053466797,
+      "learning_rate": 5.851458354214694e-06,
+      "loss": 1.8068,
+      "step": 7059
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 13.925558090209961,
+      "learning_rate": 5.849453743610303e-06,
+      "loss": 0.8992,
+      "step": 7060
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 45.48153305053711,
+      "learning_rate": 5.8474491330059135e-06,
+      "loss": 2.8715,
+      "step": 7061
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 19.373300552368164,
+      "learning_rate": 5.8454445224015244e-06,
+      "loss": 1.1757,
+      "step": 7062
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 20.65883445739746,
+      "learning_rate": 5.8434399117971345e-06,
+      "loss": 2.2168,
+      "step": 7063
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 18.880720138549805,
+      "learning_rate": 5.841435301192744e-06,
+      "loss": 1.2185,
+      "step": 7064
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 14.757368087768555,
+      "learning_rate": 5.839430690588354e-06,
+      "loss": 1.3524,
+      "step": 7065
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 14.642279624938965,
+      "learning_rate": 5.837426079983963e-06,
+      "loss": 2.0154,
+      "step": 7066
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 7.192394733428955,
+      "learning_rate": 5.835421469379573e-06,
+      "loss": 0.3944,
+      "step": 7067
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 33.72096252441406,
+      "learning_rate": 5.833416858775184e-06,
+      "loss": 1.6128,
+      "step": 7068
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 18.03380584716797,
+      "learning_rate": 5.831412248170793e-06,
+      "loss": 1.4521,
+      "step": 7069
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 10.494623184204102,
+      "learning_rate": 5.829407637566403e-06,
+      "loss": 1.3512,
+      "step": 7070
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 12.263495445251465,
+      "learning_rate": 5.827403026962013e-06,
+      "loss": 1.0091,
+      "step": 7071
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 10.31989860534668,
+      "learning_rate": 5.825398416357623e-06,
+      "loss": 1.4169,
+      "step": 7072
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 24.252532958984375,
+      "learning_rate": 5.823393805753233e-06,
+      "loss": 1.5791,
+      "step": 7073
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 9.647172927856445,
+      "learning_rate": 5.821389195148843e-06,
+      "loss": 0.7097,
+      "step": 7074
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 39.26737976074219,
+      "learning_rate": 5.819384584544453e-06,
+      "loss": 1.2926,
+      "step": 7075
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 18.434120178222656,
+      "learning_rate": 5.817379973940063e-06,
+      "loss": 1.684,
+      "step": 7076
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 23.794525146484375,
+      "learning_rate": 5.815375363335672e-06,
+      "loss": 1.4986,
+      "step": 7077
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 23.152494430541992,
+      "learning_rate": 5.813370752731282e-06,
+      "loss": 1.7962,
+      "step": 7078
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 10.018393516540527,
+      "learning_rate": 5.811366142126892e-06,
+      "loss": 0.6564,
+      "step": 7079
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 14.856311798095703,
+      "learning_rate": 5.8093615315225026e-06,
+      "loss": 1.3591,
+      "step": 7080
+    },
+    {
+      "epoch": 2.13,
+      "eval_loss": 0.1809202879667282,
+      "eval_runtime": 43.5269,
+      "eval_samples_per_second": 33.979,
+      "eval_steps_per_second": 33.979,
+      "step": 7080
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 14.052577018737793,
+      "learning_rate": 5.807356920918113e-06,
+      "loss": 0.7851,
+      "step": 7081
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 49.35251998901367,
+      "learning_rate": 5.805352310313722e-06,
+      "loss": 2.7815,
+      "step": 7082
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 48.668617248535156,
+      "learning_rate": 5.803347699709332e-06,
+      "loss": 1.4805,
+      "step": 7083
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 9.275986671447754,
+      "learning_rate": 5.801343089104941e-06,
+      "loss": 0.7976,
+      "step": 7084
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 50.608848571777344,
+      "learning_rate": 5.799338478500551e-06,
+      "loss": 1.9386,
+      "step": 7085
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 18.53827476501465,
+      "learning_rate": 5.797333867896162e-06,
+      "loss": 1.2232,
+      "step": 7086
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 20.711294174194336,
+      "learning_rate": 5.7953292572917715e-06,
+      "loss": 2.057,
+      "step": 7087
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 12.552864074707031,
+      "learning_rate": 5.7933246466873815e-06,
+      "loss": 0.9687,
+      "step": 7088
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 20.43316078186035,
+      "learning_rate": 5.791320036082992e-06,
+      "loss": 1.6212,
+      "step": 7089
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 8.86861801147461,
+      "learning_rate": 5.789315425478601e-06,
+      "loss": 1.2358,
+      "step": 7090
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 17.865428924560547,
+      "learning_rate": 5.787310814874211e-06,
+      "loss": 1.8776,
+      "step": 7091
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 10.841514587402344,
+      "learning_rate": 5.785306204269822e-06,
+      "loss": 1.0038,
+      "step": 7092
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 10.26215934753418,
+      "learning_rate": 5.783301593665431e-06,
+      "loss": 0.7371,
+      "step": 7093
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 19.391332626342773,
+      "learning_rate": 5.781296983061041e-06,
+      "loss": 1.335,
+      "step": 7094
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 84.89678955078125,
+      "learning_rate": 5.7792923724566504e-06,
+      "loss": 3.5264,
+      "step": 7095
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 18.809005737304688,
+      "learning_rate": 5.7772877618522605e-06,
+      "loss": 1.2989,
+      "step": 7096
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 17.723434448242188,
+      "learning_rate": 5.775283151247871e-06,
+      "loss": 0.9432,
+      "step": 7097
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 11.459822654724121,
+      "learning_rate": 5.77327854064348e-06,
+      "loss": 0.796,
+      "step": 7098
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 10.62411880493164,
+      "learning_rate": 5.771273930039091e-06,
+      "loss": 2.0095,
+      "step": 7099
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 17.644634246826172,
+      "learning_rate": 5.769269319434701e-06,
+      "loss": 1.3346,
+      "step": 7100
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 16.67945671081543,
+      "learning_rate": 5.76726470883031e-06,
+      "loss": 1.6617,
+      "step": 7101
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 12.336973190307617,
+      "learning_rate": 5.76526009822592e-06,
+      "loss": 1.2088,
+      "step": 7102
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 43.285831451416016,
+      "learning_rate": 5.763255487621529e-06,
+      "loss": 1.8144,
+      "step": 7103
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 95.72914123535156,
+      "learning_rate": 5.7612508770171395e-06,
+      "loss": 1.4038,
+      "step": 7104
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 37.060298919677734,
+      "learning_rate": 5.7592462664127504e-06,
+      "loss": 1.6265,
+      "step": 7105
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 45.99290466308594,
+      "learning_rate": 5.75724165580836e-06,
+      "loss": 1.7691,
+      "step": 7106
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 15.260518074035645,
+      "learning_rate": 5.75523704520397e-06,
+      "loss": 1.7976,
+      "step": 7107
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 18.973817825317383,
+      "learning_rate": 5.753232434599579e-06,
+      "loss": 1.355,
+      "step": 7108
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 16.729829788208008,
+      "learning_rate": 5.751227823995189e-06,
+      "loss": 1.2276,
+      "step": 7109
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 15.583281517028809,
+      "learning_rate": 5.749223213390799e-06,
+      "loss": 1.3231,
+      "step": 7110
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 9.61965274810791,
+      "learning_rate": 5.747218602786409e-06,
+      "loss": 0.6578,
+      "step": 7111
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 14.255253791809082,
+      "learning_rate": 5.745213992182019e-06,
+      "loss": 1.0521,
+      "step": 7112
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 28.25699234008789,
+      "learning_rate": 5.743209381577629e-06,
+      "loss": 1.7292,
+      "step": 7113
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 24.005878448486328,
+      "learning_rate": 5.741204770973239e-06,
+      "loss": 1.3648,
+      "step": 7114
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 55.37303924560547,
+      "learning_rate": 5.739200160368849e-06,
+      "loss": 1.104,
+      "step": 7115
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 18.499359130859375,
+      "learning_rate": 5.737195549764458e-06,
+      "loss": 1.4972,
+      "step": 7116
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 14.160520553588867,
+      "learning_rate": 5.735190939160069e-06,
+      "loss": 1.2106,
+      "step": 7117
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 12.180475234985352,
+      "learning_rate": 5.733186328555679e-06,
+      "loss": 1.4281,
+      "step": 7118
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 10.075895309448242,
+      "learning_rate": 5.731181717951288e-06,
+      "loss": 0.811,
+      "step": 7119
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 62.43848419189453,
+      "learning_rate": 5.729177107346898e-06,
+      "loss": 2.8725,
+      "step": 7120
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 23.31486701965332,
+      "learning_rate": 5.7271724967425076e-06,
+      "loss": 2.8676,
+      "step": 7121
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 13.410110473632812,
+      "learning_rate": 5.725167886138118e-06,
+      "loss": 1.0656,
+      "step": 7122
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 25.776451110839844,
+      "learning_rate": 5.7231632755337286e-06,
+      "loss": 1.9623,
+      "step": 7123
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 19.853778839111328,
+      "learning_rate": 5.721158664929339e-06,
+      "loss": 1.2182,
+      "step": 7124
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 30.225332260131836,
+      "learning_rate": 5.719154054324948e-06,
+      "loss": 1.4418,
+      "step": 7125
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 9.551996231079102,
+      "learning_rate": 5.717149443720558e-06,
+      "loss": 1.1343,
+      "step": 7126
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 33.873294830322266,
+      "learning_rate": 5.715144833116167e-06,
+      "loss": 1.9743,
+      "step": 7127
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 11.439074516296387,
+      "learning_rate": 5.713140222511777e-06,
+      "loss": 1.1761,
+      "step": 7128
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 11.673768043518066,
+      "learning_rate": 5.711135611907388e-06,
+      "loss": 1.4744,
+      "step": 7129
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 22.201738357543945,
+      "learning_rate": 5.7091310013029975e-06,
+      "loss": 2.8592,
+      "step": 7130
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 11.81099796295166,
+      "learning_rate": 5.7071263906986075e-06,
+      "loss": 1.2218,
+      "step": 7131
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 26.6683406829834,
+      "learning_rate": 5.705121780094217e-06,
+      "loss": 1.2941,
+      "step": 7132
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 8.380386352539062,
+      "learning_rate": 5.703117169489827e-06,
+      "loss": 0.6369,
+      "step": 7133
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 11.767404556274414,
+      "learning_rate": 5.701112558885437e-06,
+      "loss": 1.0829,
+      "step": 7134
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 25.861289978027344,
+      "learning_rate": 5.699107948281047e-06,
+      "loss": 1.9822,
+      "step": 7135
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 27.93609619140625,
+      "learning_rate": 5.697103337676657e-06,
+      "loss": 1.3616,
+      "step": 7136
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 87.64994812011719,
+      "learning_rate": 5.695098727072267e-06,
+      "loss": 2.3206,
+      "step": 7137
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 11.817543029785156,
+      "learning_rate": 5.6930941164678764e-06,
+      "loss": 1.2847,
+      "step": 7138
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 21.869726181030273,
+      "learning_rate": 5.6910895058634865e-06,
+      "loss": 1.2728,
+      "step": 7139
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 12.761932373046875,
+      "learning_rate": 5.689084895259096e-06,
+      "loss": 1.5109,
+      "step": 7140
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 10.641894340515137,
+      "learning_rate": 5.687080284654706e-06,
+      "loss": 0.7452,
+      "step": 7141
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 17.009889602661133,
+      "learning_rate": 5.685075674050317e-06,
+      "loss": 1.8117,
+      "step": 7142
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 24.577238082885742,
+      "learning_rate": 5.683071063445926e-06,
+      "loss": 1.1146,
+      "step": 7143
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 35.608741760253906,
+      "learning_rate": 5.681066452841536e-06,
+      "loss": 2.1795,
+      "step": 7144
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 11.501810073852539,
+      "learning_rate": 5.679061842237145e-06,
+      "loss": 0.9263,
+      "step": 7145
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 17.368040084838867,
+      "learning_rate": 5.6770572316327554e-06,
+      "loss": 1.4102,
+      "step": 7146
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 16.95510482788086,
+      "learning_rate": 5.6750526210283655e-06,
+      "loss": 1.601,
+      "step": 7147
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 8.293194770812988,
+      "learning_rate": 5.673048010423976e-06,
+      "loss": 0.8742,
+      "step": 7148
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 6.959547519683838,
+      "learning_rate": 5.671043399819586e-06,
+      "loss": 0.4787,
+      "step": 7149
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 15.539986610412598,
+      "learning_rate": 5.669038789215196e-06,
+      "loss": 0.8455,
+      "step": 7150
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 11.356565475463867,
+      "learning_rate": 5.667034178610805e-06,
+      "loss": 1.1054,
+      "step": 7151
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 25.823801040649414,
+      "learning_rate": 5.665029568006415e-06,
+      "loss": 1.7274,
+      "step": 7152
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 43.120635986328125,
+      "learning_rate": 5.663024957402024e-06,
+      "loss": 1.6869,
+      "step": 7153
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 8.684465408325195,
+      "learning_rate": 5.661020346797635e-06,
+      "loss": 1.2637,
+      "step": 7154
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 23.894262313842773,
+      "learning_rate": 5.659015736193245e-06,
+      "loss": 1.4771,
+      "step": 7155
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 26.122549057006836,
+      "learning_rate": 5.6570111255888546e-06,
+      "loss": 1.3112,
+      "step": 7156
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 20.096174240112305,
+      "learning_rate": 5.655006514984465e-06,
+      "loss": 1.0569,
+      "step": 7157
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 36.696006774902344,
+      "learning_rate": 5.653001904380075e-06,
+      "loss": 1.7493,
+      "step": 7158
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 28.245203018188477,
+      "learning_rate": 5.650997293775684e-06,
+      "loss": 1.5218,
+      "step": 7159
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 27.659273147583008,
+      "learning_rate": 5.648992683171295e-06,
+      "loss": 2.6867,
+      "step": 7160
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 47.04728317260742,
+      "learning_rate": 5.646988072566905e-06,
+      "loss": 1.1954,
+      "step": 7161
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 17.212331771850586,
+      "learning_rate": 5.644983461962514e-06,
+      "loss": 2.1451,
+      "step": 7162
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 12.564343452453613,
+      "learning_rate": 5.642978851358124e-06,
+      "loss": 1.1914,
+      "step": 7163
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 38.6114501953125,
+      "learning_rate": 5.6409742407537336e-06,
+      "loss": 0.9675,
+      "step": 7164
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 49.109031677246094,
+      "learning_rate": 5.638969630149344e-06,
+      "loss": 1.4241,
+      "step": 7165
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 17.366100311279297,
+      "learning_rate": 5.6369650195449546e-06,
+      "loss": 1.9756,
+      "step": 7166
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 79.63975524902344,
+      "learning_rate": 5.634960408940564e-06,
+      "loss": 1.9999,
+      "step": 7167
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 28.03964614868164,
+      "learning_rate": 5.632955798336174e-06,
+      "loss": 1.4169,
+      "step": 7168
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 23.203266143798828,
+      "learning_rate": 5.630951187731783e-06,
+      "loss": 1.3344,
+      "step": 7169
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 12.224784851074219,
+      "learning_rate": 5.628946577127393e-06,
+      "loss": 1.4286,
+      "step": 7170
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 9.047876358032227,
+      "learning_rate": 5.626941966523003e-06,
+      "loss": 1.1898,
+      "step": 7171
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 22.1667423248291,
+      "learning_rate": 5.624937355918613e-06,
+      "loss": 1.945,
+      "step": 7172
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 13.911121368408203,
+      "learning_rate": 5.6229327453142235e-06,
+      "loss": 1.3742,
+      "step": 7173
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 9.241662979125977,
+      "learning_rate": 5.6209281347098335e-06,
+      "loss": 1.0261,
+      "step": 7174
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 13.640268325805664,
+      "learning_rate": 5.618923524105443e-06,
+      "loss": 1.1987,
+      "step": 7175
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 21.423112869262695,
+      "learning_rate": 5.616918913501053e-06,
+      "loss": 1.722,
+      "step": 7176
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 33.87396240234375,
+      "learning_rate": 5.614914302896662e-06,
+      "loss": 0.9787,
+      "step": 7177
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 29.940101623535156,
+      "learning_rate": 5.612909692292273e-06,
+      "loss": 1.792,
+      "step": 7178
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 22.763090133666992,
+      "learning_rate": 5.610905081687883e-06,
+      "loss": 1.3948,
+      "step": 7179
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 33.94498062133789,
+      "learning_rate": 5.608900471083492e-06,
+      "loss": 2.2237,
+      "step": 7180
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 15.50265884399414,
+      "learning_rate": 5.6068958604791024e-06,
+      "loss": 1.5329,
+      "step": 7181
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 20.567745208740234,
+      "learning_rate": 5.6048912498747125e-06,
+      "loss": 1.7991,
+      "step": 7182
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 38.924346923828125,
+      "learning_rate": 5.602886639270322e-06,
+      "loss": 1.8058,
+      "step": 7183
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 23.95490264892578,
+      "learning_rate": 5.600882028665932e-06,
+      "loss": 0.9036,
+      "step": 7184
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 8.675360679626465,
+      "learning_rate": 5.598877418061543e-06,
+      "loss": 1.1445,
+      "step": 7185
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 52.8515510559082,
+      "learning_rate": 5.596872807457152e-06,
+      "loss": 0.9069,
+      "step": 7186
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 14.209603309631348,
+      "learning_rate": 5.594868196852762e-06,
+      "loss": 1.1668,
+      "step": 7187
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 19.132238388061523,
+      "learning_rate": 5.592863586248371e-06,
+      "loss": 1.6934,
+      "step": 7188
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 16.72909927368164,
+      "learning_rate": 5.5908589756439814e-06,
+      "loss": 1.127,
+      "step": 7189
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 11.741925239562988,
+      "learning_rate": 5.588854365039591e-06,
+      "loss": 1.9914,
+      "step": 7190
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 34.14512634277344,
+      "learning_rate": 5.586849754435202e-06,
+      "loss": 2.0267,
+      "step": 7191
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 15.906356811523438,
+      "learning_rate": 5.584845143830812e-06,
+      "loss": 1.2969,
+      "step": 7192
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 14.884824752807617,
+      "learning_rate": 5.582840533226421e-06,
+      "loss": 1.6604,
+      "step": 7193
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 14.613624572753906,
+      "learning_rate": 5.580835922622031e-06,
+      "loss": 1.3751,
+      "step": 7194
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 14.830942153930664,
+      "learning_rate": 5.578831312017641e-06,
+      "loss": 1.5889,
+      "step": 7195
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 17.539030075073242,
+      "learning_rate": 5.57682670141325e-06,
+      "loss": 1.0911,
+      "step": 7196
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 28.30144500732422,
+      "learning_rate": 5.574822090808861e-06,
+      "loss": 1.2038,
+      "step": 7197
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 11.865046501159668,
+      "learning_rate": 5.572817480204471e-06,
+      "loss": 1.2387,
+      "step": 7198
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 29.797569274902344,
+      "learning_rate": 5.5708128696000806e-06,
+      "loss": 2.5909,
+      "step": 7199
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 14.200777053833008,
+      "learning_rate": 5.568808258995691e-06,
+      "loss": 1.5443,
+      "step": 7200
+    },
+    {
+      "epoch": 2.16,
+      "eval_loss": 0.1803688108921051,
+      "eval_runtime": 43.5403,
+      "eval_samples_per_second": 33.969,
+      "eval_steps_per_second": 33.969,
+      "step": 7200
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 8.243300437927246,
+      "learning_rate": 5.5668036483913e-06,
+      "loss": 1.2393,
+      "step": 7201
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 29.610637664794922,
+      "learning_rate": 5.56479903778691e-06,
+      "loss": 0.9743,
+      "step": 7202
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 30.562345504760742,
+      "learning_rate": 5.562794427182521e-06,
+      "loss": 1.5775,
+      "step": 7203
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 13.141053199768066,
+      "learning_rate": 5.56078981657813e-06,
+      "loss": 1.026,
+      "step": 7204
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 15.844094276428223,
+      "learning_rate": 5.55878520597374e-06,
+      "loss": 1.8327,
+      "step": 7205
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 7.878782272338867,
+      "learning_rate": 5.5567805953693495e-06,
+      "loss": 1.0848,
+      "step": 7206
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 12.858576774597168,
+      "learning_rate": 5.5547759847649596e-06,
+      "loss": 1.5242,
+      "step": 7207
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 11.919968605041504,
+      "learning_rate": 5.55277137416057e-06,
+      "loss": 1.6723,
+      "step": 7208
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 11.811442375183105,
+      "learning_rate": 5.5507667635561806e-06,
+      "loss": 1.1354,
+      "step": 7209
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 32.790809631347656,
+      "learning_rate": 5.54876215295179e-06,
+      "loss": 1.136,
+      "step": 7210
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 29.143482208251953,
+      "learning_rate": 5.5467575423474e-06,
+      "loss": 1.3966,
+      "step": 7211
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 24.455890655517578,
+      "learning_rate": 5.544752931743009e-06,
+      "loss": 2.0205,
+      "step": 7212
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 23.328998565673828,
+      "learning_rate": 5.542748321138619e-06,
+      "loss": 1.6471,
+      "step": 7213
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 26.35028076171875,
+      "learning_rate": 5.5407437105342285e-06,
+      "loss": 1.5245,
+      "step": 7214
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 23.13437843322754,
+      "learning_rate": 5.538739099929839e-06,
+      "loss": 1.7795,
+      "step": 7215
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 16.176660537719727,
+      "learning_rate": 5.5367344893254495e-06,
+      "loss": 1.5189,
+      "step": 7216
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 16.43119239807129,
+      "learning_rate": 5.534729878721059e-06,
+      "loss": 1.0649,
+      "step": 7217
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 16.297130584716797,
+      "learning_rate": 5.532725268116669e-06,
+      "loss": 1.0049,
+      "step": 7218
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 11.09915542602539,
+      "learning_rate": 5.530720657512279e-06,
+      "loss": 1.3301,
+      "step": 7219
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 10.667206764221191,
+      "learning_rate": 5.528716046907888e-06,
+      "loss": 1.0879,
+      "step": 7220
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 22.059289932250977,
+      "learning_rate": 5.526711436303499e-06,
+      "loss": 1.7633,
+      "step": 7221
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 45.51456069946289,
+      "learning_rate": 5.524706825699109e-06,
+      "loss": 1.451,
+      "step": 7222
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 13.462637901306152,
+      "learning_rate": 5.522702215094718e-06,
+      "loss": 1.4225,
+      "step": 7223
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 11.555397033691406,
+      "learning_rate": 5.5206976044903284e-06,
+      "loss": 1.2796,
+      "step": 7224
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 25.498672485351562,
+      "learning_rate": 5.518692993885938e-06,
+      "loss": 1.1889,
+      "step": 7225
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 18.293188095092773,
+      "learning_rate": 5.516688383281548e-06,
+      "loss": 2.2397,
+      "step": 7226
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 17.86231231689453,
+      "learning_rate": 5.514683772677157e-06,
+      "loss": 1.5173,
+      "step": 7227
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 18.436603546142578,
+      "learning_rate": 5.512679162072768e-06,
+      "loss": 1.2884,
+      "step": 7228
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 10.068883895874023,
+      "learning_rate": 5.510674551468378e-06,
+      "loss": 1.314,
+      "step": 7229
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 9.923480987548828,
+      "learning_rate": 5.508669940863987e-06,
+      "loss": 0.8318,
+      "step": 7230
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 30.2728214263916,
+      "learning_rate": 5.506665330259597e-06,
+      "loss": 1.7006,
+      "step": 7231
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 24.514514923095703,
+      "learning_rate": 5.5046607196552074e-06,
+      "loss": 1.2372,
+      "step": 7232
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 8.947182655334473,
+      "learning_rate": 5.502656109050817e-06,
+      "loss": 0.6531,
+      "step": 7233
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 32.466487884521484,
+      "learning_rate": 5.500651498446428e-06,
+      "loss": 0.9527,
+      "step": 7234
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 19.44408416748047,
+      "learning_rate": 5.498646887842038e-06,
+      "loss": 1.8595,
+      "step": 7235
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 15.075602531433105,
+      "learning_rate": 5.496642277237647e-06,
+      "loss": 1.2715,
+      "step": 7236
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 40.67919921875,
+      "learning_rate": 5.494637666633257e-06,
+      "loss": 1.1642,
+      "step": 7237
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 11.373068809509277,
+      "learning_rate": 5.492633056028866e-06,
+      "loss": 0.8346,
+      "step": 7238
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 19.530393600463867,
+      "learning_rate": 5.490628445424476e-06,
+      "loss": 1.1305,
+      "step": 7239
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 93.64016723632812,
+      "learning_rate": 5.488623834820087e-06,
+      "loss": 1.6901,
+      "step": 7240
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 26.258949279785156,
+      "learning_rate": 5.4866192242156965e-06,
+      "loss": 2.7283,
+      "step": 7241
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 23.18985366821289,
+      "learning_rate": 5.4846146136113066e-06,
+      "loss": 1.3425,
+      "step": 7242
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 18.908994674682617,
+      "learning_rate": 5.482610003006917e-06,
+      "loss": 1.075,
+      "step": 7243
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 16.2775936126709,
+      "learning_rate": 5.480605392402526e-06,
+      "loss": 1.6765,
+      "step": 7244
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 22.71320915222168,
+      "learning_rate": 5.478600781798136e-06,
+      "loss": 0.7902,
+      "step": 7245
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 27.827980041503906,
+      "learning_rate": 5.476596171193747e-06,
+      "loss": 1.4328,
+      "step": 7246
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 11.23440170288086,
+      "learning_rate": 5.474591560589356e-06,
+      "loss": 1.3235,
+      "step": 7247
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 32.618648529052734,
+      "learning_rate": 5.472586949984966e-06,
+      "loss": 1.3066,
+      "step": 7248
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 41.39394760131836,
+      "learning_rate": 5.4705823393805755e-06,
+      "loss": 1.2173,
+      "step": 7249
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 17.75341796875,
+      "learning_rate": 5.4685777287761856e-06,
+      "loss": 0.9759,
+      "step": 7250
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 24.715322494506836,
+      "learning_rate": 5.466573118171795e-06,
+      "loss": 2.2409,
+      "step": 7251
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 16.27878761291504,
+      "learning_rate": 5.464568507567406e-06,
+      "loss": 1.0532,
+      "step": 7252
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 31.52488899230957,
+      "learning_rate": 5.462563896963016e-06,
+      "loss": 2.1667,
+      "step": 7253
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 24.314443588256836,
+      "learning_rate": 5.460559286358625e-06,
+      "loss": 1.6379,
+      "step": 7254
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 9.153544425964355,
+      "learning_rate": 5.458554675754235e-06,
+      "loss": 0.9792,
+      "step": 7255
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 15.48164176940918,
+      "learning_rate": 5.456550065149845e-06,
+      "loss": 1.3478,
+      "step": 7256
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 13.745574951171875,
+      "learning_rate": 5.4545454545454545e-06,
+      "loss": 1.0092,
+      "step": 7257
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 42.99429702758789,
+      "learning_rate": 5.452540843941065e-06,
+      "loss": 2.0053,
+      "step": 7258
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 15.831052780151367,
+      "learning_rate": 5.4505362333366755e-06,
+      "loss": 1.2616,
+      "step": 7259
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 18.406320571899414,
+      "learning_rate": 5.448531622732285e-06,
+      "loss": 1.7931,
+      "step": 7260
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 23.25520133972168,
+      "learning_rate": 5.446527012127895e-06,
+      "loss": 1.7849,
+      "step": 7261
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 11.698332786560059,
+      "learning_rate": 5.444522401523504e-06,
+      "loss": 1.0173,
+      "step": 7262
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 18.50562286376953,
+      "learning_rate": 5.442517790919114e-06,
+      "loss": 1.0716,
+      "step": 7263
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 18.679840087890625,
+      "learning_rate": 5.440513180314725e-06,
+      "loss": 1.331,
+      "step": 7264
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 11.41555118560791,
+      "learning_rate": 5.438508569710334e-06,
+      "loss": 1.2174,
+      "step": 7265
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 12.062613487243652,
+      "learning_rate": 5.436503959105944e-06,
+      "loss": 1.7347,
+      "step": 7266
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 12.057971000671387,
+      "learning_rate": 5.4344993485015544e-06,
+      "loss": 1.2392,
+      "step": 7267
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 14.793939590454102,
+      "learning_rate": 5.432494737897164e-06,
+      "loss": 1.7844,
+      "step": 7268
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 20.210912704467773,
+      "learning_rate": 5.430490127292774e-06,
+      "loss": 2.0136,
+      "step": 7269
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 6.897406578063965,
+      "learning_rate": 5.428485516688383e-06,
+      "loss": 1.0915,
+      "step": 7270
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 12.593406677246094,
+      "learning_rate": 5.426480906083994e-06,
+      "loss": 0.7125,
+      "step": 7271
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 12.755428314208984,
+      "learning_rate": 5.424476295479604e-06,
+      "loss": 1.9477,
+      "step": 7272
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 7.096426486968994,
+      "learning_rate": 5.422471684875213e-06,
+      "loss": 0.718,
+      "step": 7273
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 8.153369903564453,
+      "learning_rate": 5.420467074270823e-06,
+      "loss": 0.7129,
+      "step": 7274
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 14.67286491394043,
+      "learning_rate": 5.418462463666433e-06,
+      "loss": 1.6797,
+      "step": 7275
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 18.533056259155273,
+      "learning_rate": 5.416457853062043e-06,
+      "loss": 1.1951,
+      "step": 7276
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 8.047659873962402,
+      "learning_rate": 5.414453242457654e-06,
+      "loss": 0.6167,
+      "step": 7277
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 11.328049659729004,
+      "learning_rate": 5.412448631853263e-06,
+      "loss": 0.9758,
+      "step": 7278
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 23.584749221801758,
+      "learning_rate": 5.410444021248873e-06,
+      "loss": 1.7252,
+      "step": 7279
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 22.34720230102539,
+      "learning_rate": 5.408439410644483e-06,
+      "loss": 2.0207,
+      "step": 7280
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 19.402429580688477,
+      "learning_rate": 5.406434800040092e-06,
+      "loss": 1.2057,
+      "step": 7281
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 16.989200592041016,
+      "learning_rate": 5.404430189435702e-06,
+      "loss": 1.9234,
+      "step": 7282
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 26.748077392578125,
+      "learning_rate": 5.402425578831313e-06,
+      "loss": 1.8995,
+      "step": 7283
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 97.96756744384766,
+      "learning_rate": 5.4004209682269225e-06,
+      "loss": 2.0797,
+      "step": 7284
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 20.710670471191406,
+      "learning_rate": 5.398416357622533e-06,
+      "loss": 1.0238,
+      "step": 7285
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 13.581433296203613,
+      "learning_rate": 5.396411747018142e-06,
+      "loss": 1.6866,
+      "step": 7286
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 12.058156967163086,
+      "learning_rate": 5.394407136413752e-06,
+      "loss": 1.0288,
+      "step": 7287
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 21.06386375427246,
+      "learning_rate": 5.392402525809361e-06,
+      "loss": 1.271,
+      "step": 7288
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 51.95856857299805,
+      "learning_rate": 5.390397915204972e-06,
+      "loss": 1.5959,
+      "step": 7289
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 39.71308898925781,
+      "learning_rate": 5.388393304600582e-06,
+      "loss": 1.7519,
+      "step": 7290
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 22.637428283691406,
+      "learning_rate": 5.386388693996191e-06,
+      "loss": 2.0241,
+      "step": 7291
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 13.722464561462402,
+      "learning_rate": 5.3843840833918015e-06,
+      "loss": 1.0318,
+      "step": 7292
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 35.6817512512207,
+      "learning_rate": 5.3823794727874116e-06,
+      "loss": 1.7191,
+      "step": 7293
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 13.845057487487793,
+      "learning_rate": 5.380374862183021e-06,
+      "loss": 1.3774,
+      "step": 7294
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 12.214203834533691,
+      "learning_rate": 5.378370251578632e-06,
+      "loss": 1.5962,
+      "step": 7295
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 66.19721984863281,
+      "learning_rate": 5.376365640974242e-06,
+      "loss": 1.2639,
+      "step": 7296
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 14.872382164001465,
+      "learning_rate": 5.374361030369851e-06,
+      "loss": 1.6655,
+      "step": 7297
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 19.328413009643555,
+      "learning_rate": 5.372356419765461e-06,
+      "loss": 1.3506,
+      "step": 7298
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 12.651726722717285,
+      "learning_rate": 5.37035180916107e-06,
+      "loss": 1.2916,
+      "step": 7299
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 13.568867683410645,
+      "learning_rate": 5.3683471985566805e-06,
+      "loss": 1.116,
+      "step": 7300
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 17.879594802856445,
+      "learning_rate": 5.366342587952291e-06,
+      "loss": 1.9263,
+      "step": 7301
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 16.369518280029297,
+      "learning_rate": 5.364337977347901e-06,
+      "loss": 0.7172,
+      "step": 7302
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 37.3539924621582,
+      "learning_rate": 5.362333366743511e-06,
+      "loss": 2.3236,
+      "step": 7303
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 28.66567611694336,
+      "learning_rate": 5.360328756139121e-06,
+      "loss": 1.4089,
+      "step": 7304
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 15.510285377502441,
+      "learning_rate": 5.35832414553473e-06,
+      "loss": 1.4704,
+      "step": 7305
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 18.096101760864258,
+      "learning_rate": 5.35631953493034e-06,
+      "loss": 1.029,
+      "step": 7306
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 34.01200866699219,
+      "learning_rate": 5.354314924325949e-06,
+      "loss": 1.732,
+      "step": 7307
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 12.730477333068848,
+      "learning_rate": 5.35231031372156e-06,
+      "loss": 2.1386,
+      "step": 7308
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 17.955995559692383,
+      "learning_rate": 5.35030570311717e-06,
+      "loss": 0.6812,
+      "step": 7309
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 9.278726577758789,
+      "learning_rate": 5.34830109251278e-06,
+      "loss": 0.5906,
+      "step": 7310
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 7.9487199783325195,
+      "learning_rate": 5.34629648190839e-06,
+      "loss": 0.6598,
+      "step": 7311
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 14.72791576385498,
+      "learning_rate": 5.344291871303999e-06,
+      "loss": 0.7821,
+      "step": 7312
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 8.070356369018555,
+      "learning_rate": 5.342287260699609e-06,
+      "loss": 0.7893,
+      "step": 7313
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 14.87585163116455,
+      "learning_rate": 5.34028265009522e-06,
+      "loss": 1.5793,
+      "step": 7314
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 119.86707305908203,
+      "learning_rate": 5.338278039490829e-06,
+      "loss": 2.2575,
+      "step": 7315
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 22.031450271606445,
+      "learning_rate": 5.336273428886439e-06,
+      "loss": 0.9648,
+      "step": 7316
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 20.54271697998047,
+      "learning_rate": 5.334268818282049e-06,
+      "loss": 1.7682,
+      "step": 7317
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 18.402219772338867,
+      "learning_rate": 5.332264207677659e-06,
+      "loss": 1.3206,
+      "step": 7318
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 21.85274887084961,
+      "learning_rate": 5.330259597073269e-06,
+      "loss": 2.4046,
+      "step": 7319
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 12.441575050354004,
+      "learning_rate": 5.32825498646888e-06,
+      "loss": 1.8915,
+      "step": 7320
+    },
+    {
+      "epoch": 2.2,
+      "eval_loss": 0.18911099433898926,
+      "eval_runtime": 43.726,
+      "eval_samples_per_second": 33.824,
+      "eval_steps_per_second": 33.824,
+      "step": 7320
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 18.565292358398438,
+      "learning_rate": 5.326250375864489e-06,
+      "loss": 1.4864,
+      "step": 7321
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 21.362342834472656,
+      "learning_rate": 5.324245765260099e-06,
+      "loss": 1.0767,
+      "step": 7322
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 15.389693260192871,
+      "learning_rate": 5.322241154655708e-06,
+      "loss": 0.9276,
+      "step": 7323
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 83.4190673828125,
+      "learning_rate": 5.320236544051318e-06,
+      "loss": 2.4627,
+      "step": 7324
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 17.144039154052734,
+      "learning_rate": 5.3182319334469275e-06,
+      "loss": 1.6204,
+      "step": 7325
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 12.209066390991211,
+      "learning_rate": 5.316227322842538e-06,
+      "loss": 0.9967,
+      "step": 7326
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 20.075319290161133,
+      "learning_rate": 5.3142227122381485e-06,
+      "loss": 2.2126,
+      "step": 7327
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 22.937063217163086,
+      "learning_rate": 5.312218101633759e-06,
+      "loss": 1.583,
+      "step": 7328
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 9.0242338180542,
+      "learning_rate": 5.310213491029368e-06,
+      "loss": 0.6662,
+      "step": 7329
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 8.918848037719727,
+      "learning_rate": 5.308208880424978e-06,
+      "loss": 0.8654,
+      "step": 7330
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 32.961360931396484,
+      "learning_rate": 5.306204269820587e-06,
+      "loss": 1.8518,
+      "step": 7331
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 16.80774688720703,
+      "learning_rate": 5.304199659216198e-06,
+      "loss": 1.4526,
+      "step": 7332
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 86.86822509765625,
+      "learning_rate": 5.302195048611808e-06,
+      "loss": 1.354,
+      "step": 7333
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 18.777795791625977,
+      "learning_rate": 5.300190438007417e-06,
+      "loss": 1.2628,
+      "step": 7334
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 47.588802337646484,
+      "learning_rate": 5.2981858274030275e-06,
+      "loss": 3.0136,
+      "step": 7335
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 41.61444091796875,
+      "learning_rate": 5.296181216798637e-06,
+      "loss": 2.3059,
+      "step": 7336
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 16.625478744506836,
+      "learning_rate": 5.294176606194247e-06,
+      "loss": 1.0123,
+      "step": 7337
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 50.9505729675293,
+      "learning_rate": 5.292171995589858e-06,
+      "loss": 1.1573,
+      "step": 7338
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 3.5535824298858643,
+      "learning_rate": 5.290167384985467e-06,
+      "loss": 0.5366,
+      "step": 7339
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 13.707948684692383,
+      "learning_rate": 5.288162774381077e-06,
+      "loss": 1.2388,
+      "step": 7340
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 117.4101333618164,
+      "learning_rate": 5.286158163776687e-06,
+      "loss": 2.3683,
+      "step": 7341
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 13.597126960754395,
+      "learning_rate": 5.284153553172296e-06,
+      "loss": 1.0199,
+      "step": 7342
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 8.304295539855957,
+      "learning_rate": 5.2821489425679065e-06,
+      "loss": 0.7169,
+      "step": 7343
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 14.256132125854492,
+      "learning_rate": 5.280144331963517e-06,
+      "loss": 1.2964,
+      "step": 7344
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 86.91548919677734,
+      "learning_rate": 5.278139721359127e-06,
+      "loss": 3.997,
+      "step": 7345
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 16.045759201049805,
+      "learning_rate": 5.276135110754737e-06,
+      "loss": 1.1771,
+      "step": 7346
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 39.05559158325195,
+      "learning_rate": 5.274130500150346e-06,
+      "loss": 2.0468,
+      "step": 7347
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 13.659476280212402,
+      "learning_rate": 5.272125889545956e-06,
+      "loss": 0.9715,
+      "step": 7348
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 12.945648193359375,
+      "learning_rate": 5.270121278941565e-06,
+      "loss": 1.0217,
+      "step": 7349
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 11.518617630004883,
+      "learning_rate": 5.268116668337175e-06,
+      "loss": 1.2031,
+      "step": 7350
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 43.542816162109375,
+      "learning_rate": 5.266112057732786e-06,
+      "loss": 1.8437,
+      "step": 7351
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 9.149428367614746,
+      "learning_rate": 5.2641074471283955e-06,
+      "loss": 1.1645,
+      "step": 7352
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 17.6467227935791,
+      "learning_rate": 5.262102836524006e-06,
+      "loss": 1.3694,
+      "step": 7353
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 30.845457077026367,
+      "learning_rate": 5.260098225919616e-06,
+      "loss": 1.3288,
+      "step": 7354
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 17.407968521118164,
+      "learning_rate": 5.258093615315225e-06,
+      "loss": 1.3238,
+      "step": 7355
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 22.961395263671875,
+      "learning_rate": 5.256089004710835e-06,
+      "loss": 1.1421,
+      "step": 7356
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 21.641342163085938,
+      "learning_rate": 5.254084394106446e-06,
+      "loss": 1.2616,
+      "step": 7357
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 16.428573608398438,
+      "learning_rate": 5.252079783502055e-06,
+      "loss": 0.9203,
+      "step": 7358
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 17.370075225830078,
+      "learning_rate": 5.250075172897665e-06,
+      "loss": 1.0207,
+      "step": 7359
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 17.009870529174805,
+      "learning_rate": 5.2480705622932745e-06,
+      "loss": 1.3327,
+      "step": 7360
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 16.25709342956543,
+      "learning_rate": 5.246065951688885e-06,
+      "loss": 1.274,
+      "step": 7361
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 11.702857971191406,
+      "learning_rate": 5.244061341084495e-06,
+      "loss": 1.3384,
+      "step": 7362
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 39.857627868652344,
+      "learning_rate": 5.242056730480105e-06,
+      "loss": 1.6856,
+      "step": 7363
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 34.70151138305664,
+      "learning_rate": 5.240052119875715e-06,
+      "loss": 1.7488,
+      "step": 7364
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 8.398064613342285,
+      "learning_rate": 5.238047509271325e-06,
+      "loss": 1.1744,
+      "step": 7365
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 16.08351707458496,
+      "learning_rate": 5.236042898666934e-06,
+      "loss": 1.2976,
+      "step": 7366
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 41.76036834716797,
+      "learning_rate": 5.234038288062544e-06,
+      "loss": 2.1369,
+      "step": 7367
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 14.112876892089844,
+      "learning_rate": 5.2320336774581535e-06,
+      "loss": 0.5372,
+      "step": 7368
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 25.881088256835938,
+      "learning_rate": 5.230029066853764e-06,
+      "loss": 1.3301,
+      "step": 7369
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 28.427711486816406,
+      "learning_rate": 5.2280244562493745e-06,
+      "loss": 1.6783,
+      "step": 7370
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 10.222917556762695,
+      "learning_rate": 5.226019845644984e-06,
+      "loss": 0.9323,
+      "step": 7371
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 8.820756912231445,
+      "learning_rate": 5.224015235040594e-06,
+      "loss": 0.9893,
+      "step": 7372
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 17.932085037231445,
+      "learning_rate": 5.222010624436203e-06,
+      "loss": 1.7065,
+      "step": 7373
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 11.802325248718262,
+      "learning_rate": 5.220006013831813e-06,
+      "loss": 1.3393,
+      "step": 7374
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 23.23512077331543,
+      "learning_rate": 5.218001403227424e-06,
+      "loss": 1.1122,
+      "step": 7375
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 15.658203125,
+      "learning_rate": 5.215996792623033e-06,
+      "loss": 1.3305,
+      "step": 7376
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 39.932865142822266,
+      "learning_rate": 5.213992182018643e-06,
+      "loss": 1.7015,
+      "step": 7377
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 33.233642578125,
+      "learning_rate": 5.2119875714142535e-06,
+      "loss": 1.5578,
+      "step": 7378
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 12.168890953063965,
+      "learning_rate": 5.209982960809863e-06,
+      "loss": 0.9835,
+      "step": 7379
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 9.322324752807617,
+      "learning_rate": 5.207978350205473e-06,
+      "loss": 0.8729,
+      "step": 7380
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 32.88364791870117,
+      "learning_rate": 5.205973739601084e-06,
+      "loss": 1.5136,
+      "step": 7381
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 24.65951156616211,
+      "learning_rate": 5.203969128996693e-06,
+      "loss": 1.2119,
+      "step": 7382
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 9.384584426879883,
+      "learning_rate": 5.201964518392303e-06,
+      "loss": 0.8526,
+      "step": 7383
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 16.68515396118164,
+      "learning_rate": 5.199959907787912e-06,
+      "loss": 1.6815,
+      "step": 7384
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 12.431203842163086,
+      "learning_rate": 5.197955297183522e-06,
+      "loss": 0.8398,
+      "step": 7385
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 10.749804496765137,
+      "learning_rate": 5.1959506865791325e-06,
+      "loss": 1.6542,
+      "step": 7386
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 19.671411514282227,
+      "learning_rate": 5.1939460759747425e-06,
+      "loss": 1.1745,
+      "step": 7387
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 15.82059383392334,
+      "learning_rate": 5.191941465370353e-06,
+      "loss": 1.5952,
+      "step": 7388
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 11.592869758605957,
+      "learning_rate": 5.189936854765963e-06,
+      "loss": 1.375,
+      "step": 7389
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 12.64307975769043,
+      "learning_rate": 5.187932244161572e-06,
+      "loss": 1.5403,
+      "step": 7390
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 21.732891082763672,
+      "learning_rate": 5.185927633557182e-06,
+      "loss": 0.9123,
+      "step": 7391
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 11.803874969482422,
+      "learning_rate": 5.183923022952791e-06,
+      "loss": 1.1493,
+      "step": 7392
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 13.172371864318848,
+      "learning_rate": 5.181918412348401e-06,
+      "loss": 1.5287,
+      "step": 7393
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 11.935776710510254,
+      "learning_rate": 5.179913801744012e-06,
+      "loss": 1.0215,
+      "step": 7394
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 41.79684829711914,
+      "learning_rate": 5.1779091911396215e-06,
+      "loss": 2.0082,
+      "step": 7395
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 20.29804229736328,
+      "learning_rate": 5.175904580535232e-06,
+      "loss": 1.9343,
+      "step": 7396
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 13.398682594299316,
+      "learning_rate": 5.173899969930841e-06,
+      "loss": 1.1503,
+      "step": 7397
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 24.813034057617188,
+      "learning_rate": 5.171895359326451e-06,
+      "loss": 1.0149,
+      "step": 7398
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 23.43954849243164,
+      "learning_rate": 5.169890748722061e-06,
+      "loss": 1.2993,
+      "step": 7399
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 14.955592155456543,
+      "learning_rate": 5.167886138117671e-06,
+      "loss": 1.3491,
+      "step": 7400
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 15.617820739746094,
+      "learning_rate": 5.165881527513281e-06,
+      "loss": 1.1608,
+      "step": 7401
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 18.086448669433594,
+      "learning_rate": 5.163876916908891e-06,
+      "loss": 1.4179,
+      "step": 7402
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 7.104720115661621,
+      "learning_rate": 5.1618723063045005e-06,
+      "loss": 0.8867,
+      "step": 7403
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 15.77996826171875,
+      "learning_rate": 5.159867695700111e-06,
+      "loss": 1.276,
+      "step": 7404
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 28.867000579833984,
+      "learning_rate": 5.15786308509572e-06,
+      "loss": 1.6279,
+      "step": 7405
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 26.46651268005371,
+      "learning_rate": 5.155858474491331e-06,
+      "loss": 1.2349,
+      "step": 7406
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 15.890294075012207,
+      "learning_rate": 5.153853863886941e-06,
+      "loss": 2.3934,
+      "step": 7407
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 37.84567642211914,
+      "learning_rate": 5.15184925328255e-06,
+      "loss": 1.5725,
+      "step": 7408
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 13.57693862915039,
+      "learning_rate": 5.14984464267816e-06,
+      "loss": 1.4944,
+      "step": 7409
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 16.91181755065918,
+      "learning_rate": 5.147840032073769e-06,
+      "loss": 1.7839,
+      "step": 7410
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 36.64265441894531,
+      "learning_rate": 5.1458354214693795e-06,
+      "loss": 2.3358,
+      "step": 7411
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 5.110405921936035,
+      "learning_rate": 5.14383081086499e-06,
+      "loss": 0.6339,
+      "step": 7412
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 13.546368598937988,
+      "learning_rate": 5.1418262002606005e-06,
+      "loss": 2.2864,
+      "step": 7413
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 13.818781852722168,
+      "learning_rate": 5.13982158965621e-06,
+      "loss": 1.8962,
+      "step": 7414
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 8.545878410339355,
+      "learning_rate": 5.13781697905182e-06,
+      "loss": 1.9361,
+      "step": 7415
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 22.833742141723633,
+      "learning_rate": 5.135812368447429e-06,
+      "loss": 1.869,
+      "step": 7416
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 48.628639221191406,
+      "learning_rate": 5.133807757843039e-06,
+      "loss": 1.6338,
+      "step": 7417
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 18.07792091369629,
+      "learning_rate": 5.13180314723865e-06,
+      "loss": 1.0774,
+      "step": 7418
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 12.404952049255371,
+      "learning_rate": 5.129798536634259e-06,
+      "loss": 1.8888,
+      "step": 7419
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 16.634307861328125,
+      "learning_rate": 5.127793926029869e-06,
+      "loss": 0.9383,
+      "step": 7420
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 13.299745559692383,
+      "learning_rate": 5.125789315425479e-06,
+      "loss": 1.4448,
+      "step": 7421
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 20.849573135375977,
+      "learning_rate": 5.123784704821089e-06,
+      "loss": 1.8726,
+      "step": 7422
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 69.56287384033203,
+      "learning_rate": 5.121780094216699e-06,
+      "loss": 2.2076,
+      "step": 7423
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 13.498839378356934,
+      "learning_rate": 5.119775483612309e-06,
+      "loss": 0.9413,
+      "step": 7424
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 18.645475387573242,
+      "learning_rate": 5.117770873007919e-06,
+      "loss": 1.2546,
+      "step": 7425
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 24.639263153076172,
+      "learning_rate": 5.115766262403529e-06,
+      "loss": 1.5101,
+      "step": 7426
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 9.19277572631836,
+      "learning_rate": 5.113761651799138e-06,
+      "loss": 0.789,
+      "step": 7427
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 13.205931663513184,
+      "learning_rate": 5.111757041194748e-06,
+      "loss": 1.2832,
+      "step": 7428
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 26.456480026245117,
+      "learning_rate": 5.109752430590358e-06,
+      "loss": 2.1971,
+      "step": 7429
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 16.407329559326172,
+      "learning_rate": 5.1077478199859685e-06,
+      "loss": 1.7726,
+      "step": 7430
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 24.08245277404785,
+      "learning_rate": 5.105743209381579e-06,
+      "loss": 1.5051,
+      "step": 7431
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 42.648616790771484,
+      "learning_rate": 5.103738598777188e-06,
+      "loss": 1.9396,
+      "step": 7432
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 21.788301467895508,
+      "learning_rate": 5.101733988172798e-06,
+      "loss": 1.0098,
+      "step": 7433
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 19.175148010253906,
+      "learning_rate": 5.099729377568407e-06,
+      "loss": 1.4408,
+      "step": 7434
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 33.719520568847656,
+      "learning_rate": 5.097724766964017e-06,
+      "loss": 1.6809,
+      "step": 7435
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 22.389249801635742,
+      "learning_rate": 5.095720156359627e-06,
+      "loss": 1.5647,
+      "step": 7436
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 21.30531120300293,
+      "learning_rate": 5.0937155457552374e-06,
+      "loss": 1.5027,
+      "step": 7437
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 26.76338768005371,
+      "learning_rate": 5.0917109351508475e-06,
+      "loss": 1.3216,
+      "step": 7438
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 13.98495101928711,
+      "learning_rate": 5.089706324546458e-06,
+      "loss": 1.1699,
+      "step": 7439
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 6.971924304962158,
+      "learning_rate": 5.087701713942067e-06,
+      "loss": 0.8873,
+      "step": 7440
+    },
+    {
+      "epoch": 2.24,
+      "eval_loss": 0.1698659360408783,
+      "eval_runtime": 43.7492,
+      "eval_samples_per_second": 33.806,
+      "eval_steps_per_second": 33.806,
+      "step": 7440
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 15.718365669250488,
+      "learning_rate": 5.085697103337677e-06,
+      "loss": 1.535,
+      "step": 7441
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 23.17116928100586,
+      "learning_rate": 5.083692492733286e-06,
+      "loss": 1.6606,
+      "step": 7442
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 17.73175621032715,
+      "learning_rate": 5.081687882128897e-06,
+      "loss": 1.6878,
+      "step": 7443
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 19.423385620117188,
+      "learning_rate": 5.079683271524507e-06,
+      "loss": 1.3193,
+      "step": 7444
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 11.797534942626953,
+      "learning_rate": 5.077678660920116e-06,
+      "loss": 0.881,
+      "step": 7445
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 15.763280868530273,
+      "learning_rate": 5.0756740503157265e-06,
+      "loss": 1.4587,
+      "step": 7446
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 8.128504753112793,
+      "learning_rate": 5.073669439711337e-06,
+      "loss": 1.0937,
+      "step": 7447
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 31.446063995361328,
+      "learning_rate": 5.071664829106946e-06,
+      "loss": 1.4584,
+      "step": 7448
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 13.80343246459961,
+      "learning_rate": 5.069660218502557e-06,
+      "loss": 0.7605,
+      "step": 7449
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 10.658504486083984,
+      "learning_rate": 5.067655607898167e-06,
+      "loss": 0.8793,
+      "step": 7450
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 12.806272506713867,
+      "learning_rate": 5.065650997293776e-06,
+      "loss": 0.9486,
+      "step": 7451
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 9.60067081451416,
+      "learning_rate": 5.063646386689386e-06,
+      "loss": 1.0777,
+      "step": 7452
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 10.402573585510254,
+      "learning_rate": 5.061641776084995e-06,
+      "loss": 1.0899,
+      "step": 7453
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 24.5078182220459,
+      "learning_rate": 5.0596371654806055e-06,
+      "loss": 1.6828,
+      "step": 7454
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 20.464935302734375,
+      "learning_rate": 5.057632554876216e-06,
+      "loss": 1.5368,
+      "step": 7455
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 13.39626693725586,
+      "learning_rate": 5.055627944271826e-06,
+      "loss": 0.9746,
+      "step": 7456
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 18.0714168548584,
+      "learning_rate": 5.053623333667436e-06,
+      "loss": 1.0704,
+      "step": 7457
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 33.03631591796875,
+      "learning_rate": 5.051618723063045e-06,
+      "loss": 1.4787,
+      "step": 7458
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 101.09317779541016,
+      "learning_rate": 5.049614112458655e-06,
+      "loss": 1.9596,
+      "step": 7459
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 27.06547737121582,
+      "learning_rate": 5.047609501854265e-06,
+      "loss": 0.824,
+      "step": 7460
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 26.54560089111328,
+      "learning_rate": 5.045604891249875e-06,
+      "loss": 1.0234,
+      "step": 7461
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 29.58421516418457,
+      "learning_rate": 5.043600280645485e-06,
+      "loss": 1.8553,
+      "step": 7462
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 11.534109115600586,
+      "learning_rate": 5.041595670041095e-06,
+      "loss": 0.9154,
+      "step": 7463
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 23.17076301574707,
+      "learning_rate": 5.039591059436705e-06,
+      "loss": 2.0129,
+      "step": 7464
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 19.68193244934082,
+      "learning_rate": 5.037586448832315e-06,
+      "loss": 1.0759,
+      "step": 7465
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 21.87065887451172,
+      "learning_rate": 5.035581838227924e-06,
+      "loss": 0.9128,
+      "step": 7466
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 21.481231689453125,
+      "learning_rate": 5.033577227623535e-06,
+      "loss": 1.7024,
+      "step": 7467
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 11.491585731506348,
+      "learning_rate": 5.031572617019145e-06,
+      "loss": 1.0131,
+      "step": 7468
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 14.809111595153809,
+      "learning_rate": 5.029568006414754e-06,
+      "loss": 1.3544,
+      "step": 7469
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 31.751813888549805,
+      "learning_rate": 5.027563395810364e-06,
+      "loss": 1.477,
+      "step": 7470
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 24.474321365356445,
+      "learning_rate": 5.025558785205974e-06,
+      "loss": 1.5531,
+      "step": 7471
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 182.55990600585938,
+      "learning_rate": 5.023554174601584e-06,
+      "loss": 2.9409,
+      "step": 7472
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 17.08757972717285,
+      "learning_rate": 5.0215495639971945e-06,
+      "loss": 1.093,
+      "step": 7473
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 18.484947204589844,
+      "learning_rate": 5.019544953392805e-06,
+      "loss": 1.7707,
+      "step": 7474
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 14.49228572845459,
+      "learning_rate": 5.017540342788414e-06,
+      "loss": 1.4408,
+      "step": 7475
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 10.136235237121582,
+      "learning_rate": 5.015535732184024e-06,
+      "loss": 1.0075,
+      "step": 7476
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 63.96528244018555,
+      "learning_rate": 5.013531121579633e-06,
+      "loss": 2.0736,
+      "step": 7477
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 10.56208610534668,
+      "learning_rate": 5.011526510975243e-06,
+      "loss": 1.0639,
+      "step": 7478
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 11.87157154083252,
+      "learning_rate": 5.0095219003708525e-06,
+      "loss": 1.0202,
+      "step": 7479
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 11.656047821044922,
+      "learning_rate": 5.0075172897664634e-06,
+      "loss": 1.3296,
+      "step": 7480
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 20.969064712524414,
+      "learning_rate": 5.0055126791620735e-06,
+      "loss": 2.0833,
+      "step": 7481
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 12.001461029052734,
+      "learning_rate": 5.003508068557683e-06,
+      "loss": 1.13,
+      "step": 7482
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 17.097755432128906,
+      "learning_rate": 5.001503457953293e-06,
+      "loss": 1.3623,
+      "step": 7483
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 10.47364330291748,
+      "learning_rate": 4.999498847348903e-06,
+      "loss": 0.7631,
+      "step": 7484
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 11.359280586242676,
+      "learning_rate": 4.997494236744513e-06,
+      "loss": 0.6724,
+      "step": 7485
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 10.474048614501953,
+      "learning_rate": 4.995489626140122e-06,
+      "loss": 0.5766,
+      "step": 7486
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 14.837994575500488,
+      "learning_rate": 4.993485015535733e-06,
+      "loss": 1.1248,
+      "step": 7487
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 18.054821014404297,
+      "learning_rate": 4.991480404931342e-06,
+      "loss": 1.0936,
+      "step": 7488
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 9.805904388427734,
+      "learning_rate": 4.9894757943269525e-06,
+      "loss": 0.8442,
+      "step": 7489
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 44.7296257019043,
+      "learning_rate": 4.987471183722563e-06,
+      "loss": 2.3775,
+      "step": 7490
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 13.354528427124023,
+      "learning_rate": 4.985466573118173e-06,
+      "loss": 1.3387,
+      "step": 7491
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 24.598257064819336,
+      "learning_rate": 4.983461962513782e-06,
+      "loss": 2.0694,
+      "step": 7492
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 28.982959747314453,
+      "learning_rate": 4.981457351909392e-06,
+      "loss": 1.3952,
+      "step": 7493
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 16.613788604736328,
+      "learning_rate": 4.979452741305002e-06,
+      "loss": 1.1931,
+      "step": 7494
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 11.446743965148926,
+      "learning_rate": 4.977448130700611e-06,
+      "loss": 0.8703,
+      "step": 7495
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 14.696474075317383,
+      "learning_rate": 4.975443520096221e-06,
+      "loss": 1.3185,
+      "step": 7496
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 33.95481491088867,
+      "learning_rate": 4.9734389094918315e-06,
+      "loss": 1.3152,
+      "step": 7497
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 9.814773559570312,
+      "learning_rate": 4.9714342988874416e-06,
+      "loss": 1.4784,
+      "step": 7498
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 31.25359344482422,
+      "learning_rate": 4.969429688283051e-06,
+      "loss": 1.9966,
+      "step": 7499
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 28.215229034423828,
+      "learning_rate": 4.967425077678662e-06,
+      "loss": 2.6453,
+      "step": 7500
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 16.87133026123047,
+      "learning_rate": 4.965420467074271e-06,
+      "loss": 1.0434,
+      "step": 7501
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 16.18372344970703,
+      "learning_rate": 4.963415856469881e-06,
+      "loss": 1.9418,
+      "step": 7502
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 10.651266098022461,
+      "learning_rate": 4.961411245865491e-06,
+      "loss": 1.5416,
+      "step": 7503
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 37.711185455322266,
+      "learning_rate": 4.959406635261101e-06,
+      "loss": 1.3423,
+      "step": 7504
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 21.620492935180664,
+      "learning_rate": 4.9574020246567105e-06,
+      "loss": 1.435,
+      "step": 7505
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 52.260536193847656,
+      "learning_rate": 4.9553974140523205e-06,
+      "loss": 1.8419,
+      "step": 7506
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 21.462879180908203,
+      "learning_rate": 4.953392803447931e-06,
+      "loss": 0.8966,
+      "step": 7507
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 9.330449104309082,
+      "learning_rate": 4.951388192843541e-06,
+      "loss": 0.8809,
+      "step": 7508
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 31.02956771850586,
+      "learning_rate": 4.949383582239151e-06,
+      "loss": 1.8236,
+      "step": 7509
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 19.69917869567871,
+      "learning_rate": 4.94737897163476e-06,
+      "loss": 1.4695,
+      "step": 7510
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 13.709826469421387,
+      "learning_rate": 4.94537436103037e-06,
+      "loss": 0.9547,
+      "step": 7511
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 14.561116218566895,
+      "learning_rate": 4.94336975042598e-06,
+      "loss": 2.0196,
+      "step": 7512
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 12.211623191833496,
+      "learning_rate": 4.94136513982159e-06,
+      "loss": 1.0027,
+      "step": 7513
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 94.37039947509766,
+      "learning_rate": 4.9393605292171995e-06,
+      "loss": 1.9356,
+      "step": 7514
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 14.454056739807129,
+      "learning_rate": 4.9373559186128105e-06,
+      "loss": 1.0738,
+      "step": 7515
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 88.53865051269531,
+      "learning_rate": 4.93535130800842e-06,
+      "loss": 1.7283,
+      "step": 7516
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 225.10995483398438,
+      "learning_rate": 4.93334669740403e-06,
+      "loss": 1.8414,
+      "step": 7517
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 16.295265197753906,
+      "learning_rate": 4.93134208679964e-06,
+      "loss": 1.7194,
+      "step": 7518
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 102.90641021728516,
+      "learning_rate": 4.929337476195249e-06,
+      "loss": 1.1804,
+      "step": 7519
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 22.84174156188965,
+      "learning_rate": 4.927332865590859e-06,
+      "loss": 1.0631,
+      "step": 7520
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 29.34662628173828,
+      "learning_rate": 4.925328254986469e-06,
+      "loss": 1.2535,
+      "step": 7521
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 11.276557922363281,
+      "learning_rate": 4.923323644382079e-06,
+      "loss": 1.1659,
+      "step": 7522
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 14.418801307678223,
+      "learning_rate": 4.921319033777689e-06,
+      "loss": 1.9511,
+      "step": 7523
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 20.165433883666992,
+      "learning_rate": 4.9193144231732995e-06,
+      "loss": 1.258,
+      "step": 7524
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 15.483535766601562,
+      "learning_rate": 4.917309812568909e-06,
+      "loss": 1.0741,
+      "step": 7525
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 23.793718338012695,
+      "learning_rate": 4.915305201964519e-06,
+      "loss": 1.4441,
+      "step": 7526
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 23.1822452545166,
+      "learning_rate": 4.913300591360129e-06,
+      "loss": 1.3305,
+      "step": 7527
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 18.903297424316406,
+      "learning_rate": 4.911295980755739e-06,
+      "loss": 1.1903,
+      "step": 7528
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 18.78779411315918,
+      "learning_rate": 4.909291370151348e-06,
+      "loss": 1.7469,
+      "step": 7529
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 15.670270919799805,
+      "learning_rate": 4.907286759546958e-06,
+      "loss": 1.1513,
+      "step": 7530
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 22.992828369140625,
+      "learning_rate": 4.905282148942568e-06,
+      "loss": 0.9966,
+      "step": 7531
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 16.692344665527344,
+      "learning_rate": 4.9032775383381785e-06,
+      "loss": 1.3511,
+      "step": 7532
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 17.386524200439453,
+      "learning_rate": 4.901272927733789e-06,
+      "loss": 1.4587,
+      "step": 7533
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 18.931062698364258,
+      "learning_rate": 4.899268317129398e-06,
+      "loss": 2.273,
+      "step": 7534
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 86.52537536621094,
+      "learning_rate": 4.897263706525008e-06,
+      "loss": 3.4529,
+      "step": 7535
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 9.628124237060547,
+      "learning_rate": 4.895259095920617e-06,
+      "loss": 0.8626,
+      "step": 7536
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 10.610788345336914,
+      "learning_rate": 4.893254485316228e-06,
+      "loss": 1.2612,
+      "step": 7537
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 15.068879127502441,
+      "learning_rate": 4.891249874711837e-06,
+      "loss": 1.2702,
+      "step": 7538
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 25.841157913208008,
+      "learning_rate": 4.889245264107447e-06,
+      "loss": 1.5596,
+      "step": 7539
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 13.19726276397705,
+      "learning_rate": 4.8872406535030575e-06,
+      "loss": 1.1938,
+      "step": 7540
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 30.624975204467773,
+      "learning_rate": 4.8852360428986676e-06,
+      "loss": 2.6475,
+      "step": 7541
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 17.774322509765625,
+      "learning_rate": 4.883231432294277e-06,
+      "loss": 0.8152,
+      "step": 7542
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 13.231738090515137,
+      "learning_rate": 4.881226821689887e-06,
+      "loss": 1.0322,
+      "step": 7543
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 18.530324935913086,
+      "learning_rate": 4.879222211085497e-06,
+      "loss": 1.6454,
+      "step": 7544
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 32.340023040771484,
+      "learning_rate": 4.877217600481107e-06,
+      "loss": 1.1846,
+      "step": 7545
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 13.608917236328125,
+      "learning_rate": 4.875212989876717e-06,
+      "loss": 1.307,
+      "step": 7546
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 33.83692932128906,
+      "learning_rate": 4.873208379272326e-06,
+      "loss": 1.4187,
+      "step": 7547
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 41.42520523071289,
+      "learning_rate": 4.8712037686679365e-06,
+      "loss": 1.5752,
+      "step": 7548
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 7.894377708435059,
+      "learning_rate": 4.8691991580635466e-06,
+      "loss": 1.1328,
+      "step": 7549
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 22.25946807861328,
+      "learning_rate": 4.867194547459157e-06,
+      "loss": 2.0451,
+      "step": 7550
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 49.385562896728516,
+      "learning_rate": 4.865189936854766e-06,
+      "loss": 2.2722,
+      "step": 7551
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 13.534929275512695,
+      "learning_rate": 4.863185326250377e-06,
+      "loss": 0.7596,
+      "step": 7552
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 20.882999420166016,
+      "learning_rate": 4.861180715645986e-06,
+      "loss": 1.9043,
+      "step": 7553
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 9.814292907714844,
+      "learning_rate": 4.859176105041596e-06,
+      "loss": 1.0012,
+      "step": 7554
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 23.16642951965332,
+      "learning_rate": 4.857171494437206e-06,
+      "loss": 1.4161,
+      "step": 7555
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 41.616703033447266,
+      "learning_rate": 4.8551668838328154e-06,
+      "loss": 2.3166,
+      "step": 7556
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 14.239105224609375,
+      "learning_rate": 4.8531622732284255e-06,
+      "loss": 1.1065,
+      "step": 7557
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 8.06943130493164,
+      "learning_rate": 4.851157662624036e-06,
+      "loss": 0.6586,
+      "step": 7558
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 16.735288619995117,
+      "learning_rate": 4.849153052019646e-06,
+      "loss": 0.4483,
+      "step": 7559
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 13.807044982910156,
+      "learning_rate": 4.847148441415255e-06,
+      "loss": 1.3318,
+      "step": 7560
+    },
+    {
+      "epoch": 2.27,
+      "eval_loss": 0.1706262081861496,
+      "eval_runtime": 43.7986,
+      "eval_samples_per_second": 33.768,
+      "eval_steps_per_second": 33.768,
+      "step": 7560
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 18.019256591796875,
+      "learning_rate": 4.845143830810866e-06,
+      "loss": 1.3919,
+      "step": 7561
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 16.097368240356445,
+      "learning_rate": 4.843139220206475e-06,
+      "loss": 1.5764,
+      "step": 7562
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 13.278281211853027,
+      "learning_rate": 4.841134609602085e-06,
+      "loss": 1.0933,
+      "step": 7563
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 10.068906784057617,
+      "learning_rate": 4.839129998997695e-06,
+      "loss": 0.8741,
+      "step": 7564
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 16.215852737426758,
+      "learning_rate": 4.837125388393305e-06,
+      "loss": 2.1497,
+      "step": 7565
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 11.731410026550293,
+      "learning_rate": 4.835120777788915e-06,
+      "loss": 1.7863,
+      "step": 7566
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 10.969109535217285,
+      "learning_rate": 4.833116167184525e-06,
+      "loss": 0.8401,
+      "step": 7567
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 10.927252769470215,
+      "learning_rate": 4.831111556580135e-06,
+      "loss": 1.4045,
+      "step": 7568
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 41.4652214050293,
+      "learning_rate": 4.829106945975745e-06,
+      "loss": 3.292,
+      "step": 7569
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 8.78602123260498,
+      "learning_rate": 4.827102335371355e-06,
+      "loss": 1.6186,
+      "step": 7570
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 30.659353256225586,
+      "learning_rate": 4.825097724766964e-06,
+      "loss": 1.3395,
+      "step": 7571
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 9.962449073791504,
+      "learning_rate": 4.823093114162574e-06,
+      "loss": 1.2939,
+      "step": 7572
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 10.435197830200195,
+      "learning_rate": 4.821088503558184e-06,
+      "loss": 1.0703,
+      "step": 7573
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 20.504138946533203,
+      "learning_rate": 4.819083892953794e-06,
+      "loss": 1.5566,
+      "step": 7574
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 15.632285118103027,
+      "learning_rate": 4.817079282349404e-06,
+      "loss": 1.2294,
+      "step": 7575
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 15.574999809265137,
+      "learning_rate": 4.815074671745015e-06,
+      "loss": 1.4429,
+      "step": 7576
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 10.924356460571289,
+      "learning_rate": 4.813070061140624e-06,
+      "loss": 1.0845,
+      "step": 7577
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 24.5418758392334,
+      "learning_rate": 4.811065450536234e-06,
+      "loss": 1.175,
+      "step": 7578
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 29.236740112304688,
+      "learning_rate": 4.809060839931843e-06,
+      "loss": 0.9208,
+      "step": 7579
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 11.928509712219238,
+      "learning_rate": 4.807056229327453e-06,
+      "loss": 0.7309,
+      "step": 7580
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 31.78061294555664,
+      "learning_rate": 4.805051618723063e-06,
+      "loss": 1.4187,
+      "step": 7581
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 18.418073654174805,
+      "learning_rate": 4.803047008118673e-06,
+      "loss": 2.4135,
+      "step": 7582
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 27.633852005004883,
+      "learning_rate": 4.8010423975142835e-06,
+      "loss": 1.7401,
+      "step": 7583
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 48.656333923339844,
+      "learning_rate": 4.799037786909893e-06,
+      "loss": 1.6712,
+      "step": 7584
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 15.722077369689941,
+      "learning_rate": 4.797033176305503e-06,
+      "loss": 1.2471,
+      "step": 7585
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 8.848787307739258,
+      "learning_rate": 4.795028565701113e-06,
+      "loss": 0.993,
+      "step": 7586
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 15.616233825683594,
+      "learning_rate": 4.793023955096723e-06,
+      "loss": 1.3871,
+      "step": 7587
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 24.031694412231445,
+      "learning_rate": 4.791019344492332e-06,
+      "loss": 1.2137,
+      "step": 7588
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 11.082610130310059,
+      "learning_rate": 4.789014733887943e-06,
+      "loss": 1.1946,
+      "step": 7589
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 18.813764572143555,
+      "learning_rate": 4.787010123283552e-06,
+      "loss": 1.1478,
+      "step": 7590
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 20.931020736694336,
+      "learning_rate": 4.7850055126791625e-06,
+      "loss": 1.3695,
+      "step": 7591
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 15.224300384521484,
+      "learning_rate": 4.7830009020747726e-06,
+      "loss": 1.0471,
+      "step": 7592
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 13.662450790405273,
+      "learning_rate": 4.780996291470383e-06,
+      "loss": 1.0492,
+      "step": 7593
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 11.05825424194336,
+      "learning_rate": 4.778991680865992e-06,
+      "loss": 1.4997,
+      "step": 7594
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 14.814659118652344,
+      "learning_rate": 4.776987070261602e-06,
+      "loss": 1.5903,
+      "step": 7595
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 20.72893524169922,
+      "learning_rate": 4.774982459657212e-06,
+      "loss": 0.9354,
+      "step": 7596
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 16.968223571777344,
+      "learning_rate": 4.772977849052821e-06,
+      "loss": 1.7869,
+      "step": 7597
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 12.797143936157227,
+      "learning_rate": 4.770973238448432e-06,
+      "loss": 1.4657,
+      "step": 7598
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 278.7440185546875,
+      "learning_rate": 4.7689686278440414e-06,
+      "loss": 1.1522,
+      "step": 7599
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 11.309410095214844,
+      "learning_rate": 4.7669640172396515e-06,
+      "loss": 1.6979,
+      "step": 7600
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 17.630992889404297,
+      "learning_rate": 4.764959406635262e-06,
+      "loss": 1.2774,
+      "step": 7601
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 64.68533325195312,
+      "learning_rate": 4.762954796030872e-06,
+      "loss": 1.6083,
+      "step": 7602
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 12.072052001953125,
+      "learning_rate": 4.760950185426481e-06,
+      "loss": 0.8717,
+      "step": 7603
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 13.834127426147461,
+      "learning_rate": 4.758945574822091e-06,
+      "loss": 1.385,
+      "step": 7604
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 18.952781677246094,
+      "learning_rate": 4.756940964217701e-06,
+      "loss": 1.2853,
+      "step": 7605
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 18.030406951904297,
+      "learning_rate": 4.754936353613311e-06,
+      "loss": 1.6859,
+      "step": 7606
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 35.28142166137695,
+      "learning_rate": 4.752931743008921e-06,
+      "loss": 2.1156,
+      "step": 7607
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 28.663076400756836,
+      "learning_rate": 4.7509271324045305e-06,
+      "loss": 2.1559,
+      "step": 7608
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 11.263956069946289,
+      "learning_rate": 4.748922521800141e-06,
+      "loss": 0.9984,
+      "step": 7609
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 9.154500007629395,
+      "learning_rate": 4.746917911195751e-06,
+      "loss": 1.3324,
+      "step": 7610
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 18.66733169555664,
+      "learning_rate": 4.744913300591361e-06,
+      "loss": 0.6866,
+      "step": 7611
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 22.89703369140625,
+      "learning_rate": 4.74290868998697e-06,
+      "loss": 2.2213,
+      "step": 7612
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 9.362771987915039,
+      "learning_rate": 4.740904079382581e-06,
+      "loss": 1.069,
+      "step": 7613
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 100.19771575927734,
+      "learning_rate": 4.73889946877819e-06,
+      "loss": 1.1066,
+      "step": 7614
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 12.872753143310547,
+      "learning_rate": 4.7368948581738e-06,
+      "loss": 1.3306,
+      "step": 7615
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 13.48827075958252,
+      "learning_rate": 4.73489024756941e-06,
+      "loss": 0.8473,
+      "step": 7616
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 8.565572738647461,
+      "learning_rate": 4.7328856369650204e-06,
+      "loss": 0.7963,
+      "step": 7617
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 16.844417572021484,
+      "learning_rate": 4.73088102636063e-06,
+      "loss": 0.9599,
+      "step": 7618
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 16.240650177001953,
+      "learning_rate": 4.72887641575624e-06,
+      "loss": 1.0792,
+      "step": 7619
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 13.393976211547852,
+      "learning_rate": 4.72687180515185e-06,
+      "loss": 1.1918,
+      "step": 7620
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 27.26254653930664,
+      "learning_rate": 4.724867194547459e-06,
+      "loss": 1.5554,
+      "step": 7621
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 19.264162063598633,
+      "learning_rate": 4.722862583943069e-06,
+      "loss": 1.0877,
+      "step": 7622
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 14.326976776123047,
+      "learning_rate": 4.720857973338679e-06,
+      "loss": 1.4076,
+      "step": 7623
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 15.193740844726562,
+      "learning_rate": 4.718853362734289e-06,
+      "loss": 1.6355,
+      "step": 7624
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 9.33708381652832,
+      "learning_rate": 4.7168487521298986e-06,
+      "loss": 1.4984,
+      "step": 7625
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 84.12486267089844,
+      "learning_rate": 4.7148441415255095e-06,
+      "loss": 1.9932,
+      "step": 7626
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 21.830821990966797,
+      "learning_rate": 4.712839530921119e-06,
+      "loss": 1.4148,
+      "step": 7627
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 11.36988353729248,
+      "learning_rate": 4.710834920316729e-06,
+      "loss": 1.3706,
+      "step": 7628
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 4.701085567474365,
+      "learning_rate": 4.708830309712339e-06,
+      "loss": 0.4887,
+      "step": 7629
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 40.99164962768555,
+      "learning_rate": 4.706825699107949e-06,
+      "loss": 3.0745,
+      "step": 7630
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 46.76643753051758,
+      "learning_rate": 4.704821088503558e-06,
+      "loss": 1.5348,
+      "step": 7631
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 24.31828498840332,
+      "learning_rate": 4.702816477899168e-06,
+      "loss": 1.1564,
+      "step": 7632
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 32.74114227294922,
+      "learning_rate": 4.700811867294778e-06,
+      "loss": 0.9132,
+      "step": 7633
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 36.050968170166016,
+      "learning_rate": 4.6988072566903885e-06,
+      "loss": 1.0512,
+      "step": 7634
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 11.422075271606445,
+      "learning_rate": 4.6968026460859986e-06,
+      "loss": 1.2059,
+      "step": 7635
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 35.184654235839844,
+      "learning_rate": 4.694798035481608e-06,
+      "loss": 1.4813,
+      "step": 7636
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 45.07157516479492,
+      "learning_rate": 4.692793424877218e-06,
+      "loss": 1.2941,
+      "step": 7637
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 37.6901969909668,
+      "learning_rate": 4.690788814272828e-06,
+      "loss": 3.1126,
+      "step": 7638
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 21.26262664794922,
+      "learning_rate": 4.688784203668438e-06,
+      "loss": 1.3052,
+      "step": 7639
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 26.654037475585938,
+      "learning_rate": 4.686779593064047e-06,
+      "loss": 1.6718,
+      "step": 7640
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 19.267364501953125,
+      "learning_rate": 4.684774982459657e-06,
+      "loss": 1.1093,
+      "step": 7641
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 11.072561264038086,
+      "learning_rate": 4.6827703718552674e-06,
+      "loss": 1.1142,
+      "step": 7642
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 17.212007522583008,
+      "learning_rate": 4.6807657612508775e-06,
+      "loss": 1.1688,
+      "step": 7643
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 14.088850021362305,
+      "learning_rate": 4.678761150646488e-06,
+      "loss": 1.0786,
+      "step": 7644
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 8.411274909973145,
+      "learning_rate": 4.676756540042097e-06,
+      "loss": 1.6672,
+      "step": 7645
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 10.209150314331055,
+      "learning_rate": 4.674751929437707e-06,
+      "loss": 0.6725,
+      "step": 7646
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 8.190345764160156,
+      "learning_rate": 4.672747318833317e-06,
+      "loss": 0.7018,
+      "step": 7647
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 17.66701316833496,
+      "learning_rate": 4.670742708228927e-06,
+      "loss": 1.9426,
+      "step": 7648
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 29.10671043395996,
+      "learning_rate": 4.668738097624536e-06,
+      "loss": 1.8796,
+      "step": 7649
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 10.14902114868164,
+      "learning_rate": 4.666733487020147e-06,
+      "loss": 0.9493,
+      "step": 7650
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 21.747249603271484,
+      "learning_rate": 4.6647288764157565e-06,
+      "loss": 0.8706,
+      "step": 7651
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 52.81174087524414,
+      "learning_rate": 4.662724265811367e-06,
+      "loss": 2.5637,
+      "step": 7652
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 21.7281494140625,
+      "learning_rate": 4.660719655206977e-06,
+      "loss": 1.5396,
+      "step": 7653
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 9.083329200744629,
+      "learning_rate": 4.658715044602587e-06,
+      "loss": 0.701,
+      "step": 7654
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 8.001774787902832,
+      "learning_rate": 4.656710433998196e-06,
+      "loss": 0.9427,
+      "step": 7655
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 8.815278053283691,
+      "learning_rate": 4.654705823393806e-06,
+      "loss": 1.6543,
+      "step": 7656
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 37.267616271972656,
+      "learning_rate": 4.652701212789416e-06,
+      "loss": 0.9176,
+      "step": 7657
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 32.767486572265625,
+      "learning_rate": 4.650696602185025e-06,
+      "loss": 1.6066,
+      "step": 7658
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 15.666519165039062,
+      "learning_rate": 4.648691991580636e-06,
+      "loss": 1.3438,
+      "step": 7659
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 31.18556785583496,
+      "learning_rate": 4.646687380976246e-06,
+      "loss": 2.5968,
+      "step": 7660
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 21.61691665649414,
+      "learning_rate": 4.644682770371856e-06,
+      "loss": 1.7081,
+      "step": 7661
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 6.294021129608154,
+      "learning_rate": 4.642678159767465e-06,
+      "loss": 0.5644,
+      "step": 7662
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 17.326904296875,
+      "learning_rate": 4.640673549163076e-06,
+      "loss": 1.2007,
+      "step": 7663
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 7.6418986320495605,
+      "learning_rate": 4.638668938558685e-06,
+      "loss": 1.5581,
+      "step": 7664
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 36.48931121826172,
+      "learning_rate": 4.636664327954295e-06,
+      "loss": 1.8396,
+      "step": 7665
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 12.859515190124512,
+      "learning_rate": 4.634659717349905e-06,
+      "loss": 1.6651,
+      "step": 7666
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 19.518774032592773,
+      "learning_rate": 4.632655106745515e-06,
+      "loss": 1.0046,
+      "step": 7667
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 17.942481994628906,
+      "learning_rate": 4.6306504961411246e-06,
+      "loss": 1.7626,
+      "step": 7668
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 17.79322624206543,
+      "learning_rate": 4.628645885536735e-06,
+      "loss": 0.8779,
+      "step": 7669
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 17.742807388305664,
+      "learning_rate": 4.626641274932345e-06,
+      "loss": 0.9714,
+      "step": 7670
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 14.361837387084961,
+      "learning_rate": 4.624636664327955e-06,
+      "loss": 1.1075,
+      "step": 7671
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 19.260719299316406,
+      "learning_rate": 4.622632053723565e-06,
+      "loss": 2.0738,
+      "step": 7672
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 19.078481674194336,
+      "learning_rate": 4.620627443119174e-06,
+      "loss": 1.3898,
+      "step": 7673
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 11.892518043518066,
+      "learning_rate": 4.618622832514784e-06,
+      "loss": 1.4389,
+      "step": 7674
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 37.109336853027344,
+      "learning_rate": 4.616618221910394e-06,
+      "loss": 1.5463,
+      "step": 7675
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 35.61430358886719,
+      "learning_rate": 4.614613611306004e-06,
+      "loss": 1.2612,
+      "step": 7676
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 10.030492782592773,
+      "learning_rate": 4.612609000701614e-06,
+      "loss": 1.0428,
+      "step": 7677
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 32.46242141723633,
+      "learning_rate": 4.6106043900972246e-06,
+      "loss": 1.2366,
+      "step": 7678
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 63.76512908935547,
+      "learning_rate": 4.608599779492834e-06,
+      "loss": 1.8638,
+      "step": 7679
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 21.762617111206055,
+      "learning_rate": 4.606595168888444e-06,
+      "loss": 1.7875,
+      "step": 7680
+    },
+    {
+      "epoch": 2.31,
+      "eval_loss": 0.16951656341552734,
+      "eval_runtime": 43.7895,
+      "eval_samples_per_second": 33.775,
+      "eval_steps_per_second": 33.775,
+      "step": 7680
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 10.09822940826416,
+      "learning_rate": 4.604590558284054e-06,
+      "loss": 1.1142,
+      "step": 7681
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 10.086823463439941,
+      "learning_rate": 4.602585947679663e-06,
+      "loss": 0.8318,
+      "step": 7682
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 11.713811874389648,
+      "learning_rate": 4.600581337075273e-06,
+      "loss": 0.9897,
+      "step": 7683
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 20.750566482543945,
+      "learning_rate": 4.598576726470883e-06,
+      "loss": 1.5894,
+      "step": 7684
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 17.660737991333008,
+      "learning_rate": 4.5965721158664935e-06,
+      "loss": 1.4457,
+      "step": 7685
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 11.275466918945312,
+      "learning_rate": 4.594567505262103e-06,
+      "loss": 1.0097,
+      "step": 7686
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 9.963224411010742,
+      "learning_rate": 4.592562894657714e-06,
+      "loss": 1.7383,
+      "step": 7687
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 5.169731140136719,
+      "learning_rate": 4.590558284053323e-06,
+      "loss": 0.5528,
+      "step": 7688
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 10.396340370178223,
+      "learning_rate": 4.588553673448933e-06,
+      "loss": 0.967,
+      "step": 7689
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 15.52597427368164,
+      "learning_rate": 4.586549062844543e-06,
+      "loss": 0.9266,
+      "step": 7690
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 12.798748016357422,
+      "learning_rate": 4.584544452240153e-06,
+      "loss": 1.288,
+      "step": 7691
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 21.744783401489258,
+      "learning_rate": 4.582539841635762e-06,
+      "loss": 1.3389,
+      "step": 7692
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 7.416219711303711,
+      "learning_rate": 4.5805352310313724e-06,
+      "loss": 0.9147,
+      "step": 7693
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 27.035898208618164,
+      "learning_rate": 4.5785306204269825e-06,
+      "loss": 2.5926,
+      "step": 7694
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 18.482648849487305,
+      "learning_rate": 4.576526009822593e-06,
+      "loss": 2.276,
+      "step": 7695
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 52.933494567871094,
+      "learning_rate": 4.574521399218203e-06,
+      "loss": 2.2703,
+      "step": 7696
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 11.743091583251953,
+      "learning_rate": 4.572516788613812e-06,
+      "loss": 0.9347,
+      "step": 7697
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 16.696788787841797,
+      "learning_rate": 4.570512178009422e-06,
+      "loss": 1.3391,
+      "step": 7698
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 23.81439971923828,
+      "learning_rate": 4.568507567405032e-06,
+      "loss": 2.2063,
+      "step": 7699
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 11.924534797668457,
+      "learning_rate": 4.566502956800642e-06,
+      "loss": 1.0567,
+      "step": 7700
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 15.285096168518066,
+      "learning_rate": 4.564498346196251e-06,
+      "loss": 1.6988,
+      "step": 7701
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 14.276802062988281,
+      "learning_rate": 4.562493735591862e-06,
+      "loss": 0.7809,
+      "step": 7702
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 19.88568687438965,
+      "learning_rate": 4.560489124987472e-06,
+      "loss": 1.8969,
+      "step": 7703
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 22.174179077148438,
+      "learning_rate": 4.558484514383082e-06,
+      "loss": 1.4157,
+      "step": 7704
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 26.93759536743164,
+      "learning_rate": 4.556479903778691e-06,
+      "loss": 0.7634,
+      "step": 7705
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 14.25698184967041,
+      "learning_rate": 4.554475293174301e-06,
+      "loss": 0.9706,
+      "step": 7706
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 24.950761795043945,
+      "learning_rate": 4.552470682569911e-06,
+      "loss": 1.938,
+      "step": 7707
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 7.129110336303711,
+      "learning_rate": 4.550466071965521e-06,
+      "loss": 0.538,
+      "step": 7708
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 10.459171295166016,
+      "learning_rate": 4.548461461361131e-06,
+      "loss": 1.3554,
+      "step": 7709
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 26.577316284179688,
+      "learning_rate": 4.5464568507567405e-06,
+      "loss": 2.1107,
+      "step": 7710
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 87.70563507080078,
+      "learning_rate": 4.5444522401523506e-06,
+      "loss": 1.3952,
+      "step": 7711
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 10.34211254119873,
+      "learning_rate": 4.542447629547961e-06,
+      "loss": 0.7507,
+      "step": 7712
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 33.580265045166016,
+      "learning_rate": 4.540443018943571e-06,
+      "loss": 1.8236,
+      "step": 7713
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 13.009629249572754,
+      "learning_rate": 4.53843840833918e-06,
+      "loss": 1.2249,
+      "step": 7714
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 8.66175365447998,
+      "learning_rate": 4.536433797734791e-06,
+      "loss": 0.9531,
+      "step": 7715
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 17.59461784362793,
+      "learning_rate": 4.5344291871304e-06,
+      "loss": 1.6012,
+      "step": 7716
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 26.786701202392578,
+      "learning_rate": 4.53242457652601e-06,
+      "loss": 1.3363,
+      "step": 7717
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 10.355632781982422,
+      "learning_rate": 4.53041996592162e-06,
+      "loss": 0.8374,
+      "step": 7718
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 13.772407531738281,
+      "learning_rate": 4.52841535531723e-06,
+      "loss": 1.9609,
+      "step": 7719
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 10.235556602478027,
+      "learning_rate": 4.52641074471284e-06,
+      "loss": 1.4712,
+      "step": 7720
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 37.70382308959961,
+      "learning_rate": 4.52440613410845e-06,
+      "loss": 1.8212,
+      "step": 7721
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 16.85622215270996,
+      "learning_rate": 4.52240152350406e-06,
+      "loss": 1.3294,
+      "step": 7722
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 13.376880645751953,
+      "learning_rate": 4.520396912899669e-06,
+      "loss": 0.8711,
+      "step": 7723
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 24.79469871520996,
+      "learning_rate": 4.51839230229528e-06,
+      "loss": 1.125,
+      "step": 7724
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 21.333232879638672,
+      "learning_rate": 4.516387691690889e-06,
+      "loss": 1.7883,
+      "step": 7725
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 11.4044771194458,
+      "learning_rate": 4.514383081086499e-06,
+      "loss": 1.1805,
+      "step": 7726
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 51.108253479003906,
+      "learning_rate": 4.512378470482109e-06,
+      "loss": 1.9263,
+      "step": 7727
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 21.30112075805664,
+      "learning_rate": 4.5103738598777195e-06,
+      "loss": 1.3295,
+      "step": 7728
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 26.95627784729004,
+      "learning_rate": 4.508369249273329e-06,
+      "loss": 2.7558,
+      "step": 7729
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 13.784543991088867,
+      "learning_rate": 4.506364638668939e-06,
+      "loss": 2.2716,
+      "step": 7730
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 55.046142578125,
+      "learning_rate": 4.504360028064549e-06,
+      "loss": 1.819,
+      "step": 7731
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 10.930252075195312,
+      "learning_rate": 4.502355417460159e-06,
+      "loss": 0.8165,
+      "step": 7732
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 23.520112991333008,
+      "learning_rate": 4.500350806855769e-06,
+      "loss": 0.9347,
+      "step": 7733
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 8.282185554504395,
+      "learning_rate": 4.498346196251378e-06,
+      "loss": 1.1021,
+      "step": 7734
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 15.41140365600586,
+      "learning_rate": 4.496341585646988e-06,
+      "loss": 1.5547,
+      "step": 7735
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 26.10579490661621,
+      "learning_rate": 4.4943369750425984e-06,
+      "loss": 1.77,
+      "step": 7736
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 11.937204360961914,
+      "learning_rate": 4.4923323644382085e-06,
+      "loss": 1.0362,
+      "step": 7737
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 21.467954635620117,
+      "learning_rate": 4.490327753833818e-06,
+      "loss": 1.802,
+      "step": 7738
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 17.1977596282959,
+      "learning_rate": 4.488323143229429e-06,
+      "loss": 1.2613,
+      "step": 7739
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 11.602217674255371,
+      "learning_rate": 4.486318532625038e-06,
+      "loss": 1.5562,
+      "step": 7740
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 50.02082061767578,
+      "learning_rate": 4.484313922020648e-06,
+      "loss": 1.2008,
+      "step": 7741
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 16.229761123657227,
+      "learning_rate": 4.482309311416258e-06,
+      "loss": 1.5572,
+      "step": 7742
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 16.347227096557617,
+      "learning_rate": 4.480304700811867e-06,
+      "loss": 1.4817,
+      "step": 7743
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 9.144948959350586,
+      "learning_rate": 4.478300090207477e-06,
+      "loss": 0.6118,
+      "step": 7744
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 45.46603775024414,
+      "learning_rate": 4.4762954796030875e-06,
+      "loss": 1.9398,
+      "step": 7745
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 10.924186706542969,
+      "learning_rate": 4.474290868998698e-06,
+      "loss": 1.4317,
+      "step": 7746
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 39.43347930908203,
+      "learning_rate": 4.472286258394307e-06,
+      "loss": 2.4369,
+      "step": 7747
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 124.04064178466797,
+      "learning_rate": 4.470281647789917e-06,
+      "loss": 3.3272,
+      "step": 7748
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 6.773767471313477,
+      "learning_rate": 4.468277037185527e-06,
+      "loss": 0.5183,
+      "step": 7749
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 21.174972534179688,
+      "learning_rate": 4.466272426581137e-06,
+      "loss": 2.1152,
+      "step": 7750
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 15.640390396118164,
+      "learning_rate": 4.464267815976746e-06,
+      "loss": 1.0071,
+      "step": 7751
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 11.41567611694336,
+      "learning_rate": 4.462263205372357e-06,
+      "loss": 0.9003,
+      "step": 7752
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 13.2136869430542,
+      "learning_rate": 4.4602585947679665e-06,
+      "loss": 1.4521,
+      "step": 7753
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 24.674755096435547,
+      "learning_rate": 4.4582539841635766e-06,
+      "loss": 1.1701,
+      "step": 7754
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 27.077539443969727,
+      "learning_rate": 4.456249373559187e-06,
+      "loss": 1.3348,
+      "step": 7755
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 15.865095138549805,
+      "learning_rate": 4.454244762954797e-06,
+      "loss": 0.9423,
+      "step": 7756
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 10.586139678955078,
+      "learning_rate": 4.452240152350406e-06,
+      "loss": 0.9348,
+      "step": 7757
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 14.161200523376465,
+      "learning_rate": 4.450235541746016e-06,
+      "loss": 1.5252,
+      "step": 7758
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 23.082622528076172,
+      "learning_rate": 4.448230931141626e-06,
+      "loss": 1.3711,
+      "step": 7759
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 19.711462020874023,
+      "learning_rate": 4.446226320537235e-06,
+      "loss": 2.655,
+      "step": 7760
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 24.49384880065918,
+      "learning_rate": 4.444221709932846e-06,
+      "loss": 1.0757,
+      "step": 7761
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 9.763320922851562,
+      "learning_rate": 4.4422170993284555e-06,
+      "loss": 0.9065,
+      "step": 7762
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 8.923637390136719,
+      "learning_rate": 4.440212488724066e-06,
+      "loss": 1.1599,
+      "step": 7763
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 193.84384155273438,
+      "learning_rate": 4.438207878119676e-06,
+      "loss": 1.5881,
+      "step": 7764
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 4.945940971374512,
+      "learning_rate": 4.436203267515286e-06,
+      "loss": 0.698,
+      "step": 7765
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 8.121231079101562,
+      "learning_rate": 4.434198656910895e-06,
+      "loss": 0.9452,
+      "step": 7766
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 28.268239974975586,
+      "learning_rate": 4.432194046306505e-06,
+      "loss": 1.5929,
+      "step": 7767
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 70.76728057861328,
+      "learning_rate": 4.430189435702115e-06,
+      "loss": 1.6534,
+      "step": 7768
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 23.936372756958008,
+      "learning_rate": 4.428184825097725e-06,
+      "loss": 1.2137,
+      "step": 7769
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 9.887136459350586,
+      "learning_rate": 4.426180214493335e-06,
+      "loss": 0.8383,
+      "step": 7770
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 14.792794227600098,
+      "learning_rate": 4.424175603888945e-06,
+      "loss": 0.9405,
+      "step": 7771
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 17.71567726135254,
+      "learning_rate": 4.422170993284555e-06,
+      "loss": 1.063,
+      "step": 7772
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 17.9732608795166,
+      "learning_rate": 4.420166382680165e-06,
+      "loss": 1.4261,
+      "step": 7773
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 25.94655418395996,
+      "learning_rate": 4.418161772075775e-06,
+      "loss": 2.0324,
+      "step": 7774
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 39.89909744262695,
+      "learning_rate": 4.416157161471384e-06,
+      "loss": 1.7023,
+      "step": 7775
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 23.909780502319336,
+      "learning_rate": 4.414152550866995e-06,
+      "loss": 0.9315,
+      "step": 7776
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 13.213065147399902,
+      "learning_rate": 4.412147940262604e-06,
+      "loss": 0.9687,
+      "step": 7777
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 45.419342041015625,
+      "learning_rate": 4.410143329658214e-06,
+      "loss": 1.2235,
+      "step": 7778
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 18.985164642333984,
+      "learning_rate": 4.4081387190538244e-06,
+      "loss": 1.544,
+      "step": 7779
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 9.602522850036621,
+      "learning_rate": 4.4061341084494345e-06,
+      "loss": 0.7917,
+      "step": 7780
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 38.75558090209961,
+      "learning_rate": 4.404129497845044e-06,
+      "loss": 1.5868,
+      "step": 7781
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 51.216766357421875,
+      "learning_rate": 4.402124887240654e-06,
+      "loss": 1.2511,
+      "step": 7782
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 9.934971809387207,
+      "learning_rate": 4.400120276636264e-06,
+      "loss": 1.18,
+      "step": 7783
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 10.260887145996094,
+      "learning_rate": 4.398115666031873e-06,
+      "loss": 0.5996,
+      "step": 7784
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 26.906917572021484,
+      "learning_rate": 4.396111055427484e-06,
+      "loss": 1.635,
+      "step": 7785
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 11.61650562286377,
+      "learning_rate": 4.394106444823093e-06,
+      "loss": 0.8312,
+      "step": 7786
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 40.0263786315918,
+      "learning_rate": 4.392101834218703e-06,
+      "loss": 0.9066,
+      "step": 7787
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 31.856773376464844,
+      "learning_rate": 4.390097223614313e-06,
+      "loss": 1.9564,
+      "step": 7788
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 33.76646423339844,
+      "learning_rate": 4.388092613009924e-06,
+      "loss": 1.9249,
+      "step": 7789
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 24.660625457763672,
+      "learning_rate": 4.386088002405533e-06,
+      "loss": 2.0123,
+      "step": 7790
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 16.023160934448242,
+      "learning_rate": 4.384083391801143e-06,
+      "loss": 2.5144,
+      "step": 7791
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 18.865354537963867,
+      "learning_rate": 4.382078781196753e-06,
+      "loss": 1.4033,
+      "step": 7792
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 17.00137710571289,
+      "learning_rate": 4.380074170592363e-06,
+      "loss": 1.7571,
+      "step": 7793
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 22.194595336914062,
+      "learning_rate": 4.378069559987972e-06,
+      "loss": 1.5826,
+      "step": 7794
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 24.60336685180664,
+      "learning_rate": 4.376064949383582e-06,
+      "loss": 1.7541,
+      "step": 7795
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 14.990224838256836,
+      "learning_rate": 4.3740603387791925e-06,
+      "loss": 1.1141,
+      "step": 7796
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 17.370750427246094,
+      "learning_rate": 4.3720557281748026e-06,
+      "loss": 0.947,
+      "step": 7797
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 32.88360595703125,
+      "learning_rate": 4.370051117570413e-06,
+      "loss": 1.035,
+      "step": 7798
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 13.272078514099121,
+      "learning_rate": 4.368046506966022e-06,
+      "loss": 0.8005,
+      "step": 7799
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 28.69300651550293,
+      "learning_rate": 4.366041896361632e-06,
+      "loss": 1.6131,
+      "step": 7800
+    },
+    {
+      "epoch": 2.35,
+      "eval_loss": 0.17263449728488922,
+      "eval_runtime": 43.5025,
+      "eval_samples_per_second": 33.998,
+      "eval_steps_per_second": 33.998,
+      "step": 7800
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 21.5859317779541,
+      "learning_rate": 4.364037285757242e-06,
+      "loss": 1.961,
+      "step": 7801
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 45.89167785644531,
+      "learning_rate": 4.362032675152852e-06,
+      "loss": 0.887,
+      "step": 7802
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 9.14841079711914,
+      "learning_rate": 4.360028064548461e-06,
+      "loss": 0.82,
+      "step": 7803
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 13.917359352111816,
+      "learning_rate": 4.358023453944072e-06,
+      "loss": 0.6227,
+      "step": 7804
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 10.889512062072754,
+      "learning_rate": 4.3560188433396815e-06,
+      "loss": 0.885,
+      "step": 7805
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 11.779256820678711,
+      "learning_rate": 4.354014232735292e-06,
+      "loss": 1.6273,
+      "step": 7806
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 9.816779136657715,
+      "learning_rate": 4.352009622130902e-06,
+      "loss": 1.2789,
+      "step": 7807
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 10.623241424560547,
+      "learning_rate": 4.350005011526511e-06,
+      "loss": 0.888,
+      "step": 7808
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 15.834614753723145,
+      "learning_rate": 4.348000400922121e-06,
+      "loss": 0.8899,
+      "step": 7809
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 8.324065208435059,
+      "learning_rate": 4.345995790317731e-06,
+      "loss": 0.7989,
+      "step": 7810
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 17.0432186126709,
+      "learning_rate": 4.343991179713341e-06,
+      "loss": 2.3516,
+      "step": 7811
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 20.02086639404297,
+      "learning_rate": 4.3419865691089504e-06,
+      "loss": 0.8566,
+      "step": 7812
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 7.930890083312988,
+      "learning_rate": 4.339981958504561e-06,
+      "loss": 1.2351,
+      "step": 7813
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 30.14038848876953,
+      "learning_rate": 4.337977347900171e-06,
+      "loss": 1.7222,
+      "step": 7814
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 12.645387649536133,
+      "learning_rate": 4.335972737295781e-06,
+      "loss": 1.3162,
+      "step": 7815
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 17.415475845336914,
+      "learning_rate": 4.333968126691391e-06,
+      "loss": 1.8934,
+      "step": 7816
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 13.79565143585205,
+      "learning_rate": 4.331963516087001e-06,
+      "loss": 1.4519,
+      "step": 7817
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 14.080390930175781,
+      "learning_rate": 4.32995890548261e-06,
+      "loss": 0.899,
+      "step": 7818
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 71.52656555175781,
+      "learning_rate": 4.32795429487822e-06,
+      "loss": 1.7353,
+      "step": 7819
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 20.801849365234375,
+      "learning_rate": 4.32594968427383e-06,
+      "loss": 1.2644,
+      "step": 7820
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 40.78676223754883,
+      "learning_rate": 4.32394507366944e-06,
+      "loss": 2.8361,
+      "step": 7821
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 14.759114265441895,
+      "learning_rate": 4.3219404630650504e-06,
+      "loss": 1.0354,
+      "step": 7822
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 24.163408279418945,
+      "learning_rate": 4.31993585246066e-06,
+      "loss": 1.1916,
+      "step": 7823
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 8.51475715637207,
+      "learning_rate": 4.31793124185627e-06,
+      "loss": 1.5931,
+      "step": 7824
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 16.245559692382812,
+      "learning_rate": 4.31592663125188e-06,
+      "loss": 1.3994,
+      "step": 7825
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 78.04993438720703,
+      "learning_rate": 4.31392202064749e-06,
+      "loss": 2.4156,
+      "step": 7826
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 11.318324089050293,
+      "learning_rate": 4.311917410043099e-06,
+      "loss": 1.7852,
+      "step": 7827
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 9.104801177978516,
+      "learning_rate": 4.309912799438709e-06,
+      "loss": 0.7264,
+      "step": 7828
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 12.194317817687988,
+      "learning_rate": 4.307908188834319e-06,
+      "loss": 1.4084,
+      "step": 7829
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 39.487327575683594,
+      "learning_rate": 4.305903578229929e-06,
+      "loss": 1.4993,
+      "step": 7830
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 24.5257511138916,
+      "learning_rate": 4.303898967625539e-06,
+      "loss": 1.7682,
+      "step": 7831
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 22.789091110229492,
+      "learning_rate": 4.301894357021149e-06,
+      "loss": 2.2842,
+      "step": 7832
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 14.122344017028809,
+      "learning_rate": 4.299889746416759e-06,
+      "loss": 0.7481,
+      "step": 7833
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 9.010689735412598,
+      "learning_rate": 4.297885135812369e-06,
+      "loss": 1.0806,
+      "step": 7834
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 20.649883270263672,
+      "learning_rate": 4.295880525207979e-06,
+      "loss": 1.4924,
+      "step": 7835
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 24.006229400634766,
+      "learning_rate": 4.293875914603588e-06,
+      "loss": 1.4348,
+      "step": 7836
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 30.37126350402832,
+      "learning_rate": 4.291871303999198e-06,
+      "loss": 1.5881,
+      "step": 7837
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 10.542387962341309,
+      "learning_rate": 4.289866693394808e-06,
+      "loss": 0.7792,
+      "step": 7838
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 19.010889053344727,
+      "learning_rate": 4.2878620827904185e-06,
+      "loss": 1.4441,
+      "step": 7839
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 13.315712928771973,
+      "learning_rate": 4.285857472186028e-06,
+      "loss": 1.0043,
+      "step": 7840
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 11.758505821228027,
+      "learning_rate": 4.283852861581639e-06,
+      "loss": 1.304,
+      "step": 7841
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 14.62195873260498,
+      "learning_rate": 4.281848250977248e-06,
+      "loss": 1.1863,
+      "step": 7842
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 17.724573135375977,
+      "learning_rate": 4.279843640372858e-06,
+      "loss": 1.0244,
+      "step": 7843
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 49.88128662109375,
+      "learning_rate": 4.277839029768468e-06,
+      "loss": 1.6886,
+      "step": 7844
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 28.493061065673828,
+      "learning_rate": 4.275834419164077e-06,
+      "loss": 1.7018,
+      "step": 7845
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 17.226518630981445,
+      "learning_rate": 4.273829808559687e-06,
+      "loss": 1.3155,
+      "step": 7846
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 14.533112525939941,
+      "learning_rate": 4.2718251979552975e-06,
+      "loss": 1.6489,
+      "step": 7847
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 154.60177612304688,
+      "learning_rate": 4.2698205873509075e-06,
+      "loss": 3.9771,
+      "step": 7848
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 24.33751678466797,
+      "learning_rate": 4.267815976746517e-06,
+      "loss": 1.5103,
+      "step": 7849
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 7.450550556182861,
+      "learning_rate": 4.265811366142128e-06,
+      "loss": 0.8299,
+      "step": 7850
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 11.310479164123535,
+      "learning_rate": 4.263806755537737e-06,
+      "loss": 1.0471,
+      "step": 7851
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 22.815073013305664,
+      "learning_rate": 4.261802144933347e-06,
+      "loss": 2.6051,
+      "step": 7852
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 49.23891067504883,
+      "learning_rate": 4.259797534328957e-06,
+      "loss": 1.1229,
+      "step": 7853
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 12.846585273742676,
+      "learning_rate": 4.257792923724567e-06,
+      "loss": 1.0635,
+      "step": 7854
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 7.844564437866211,
+      "learning_rate": 4.2557883131201764e-06,
+      "loss": 0.8797,
+      "step": 7855
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 12.018916130065918,
+      "learning_rate": 4.2537837025157865e-06,
+      "loss": 1.2833,
+      "step": 7856
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 15.914363861083984,
+      "learning_rate": 4.251779091911397e-06,
+      "loss": 1.2172,
+      "step": 7857
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 20.0510311126709,
+      "learning_rate": 4.249774481307007e-06,
+      "loss": 0.7942,
+      "step": 7858
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 45.73860549926758,
+      "learning_rate": 4.247769870702617e-06,
+      "loss": 2.2663,
+      "step": 7859
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 18.577871322631836,
+      "learning_rate": 4.245765260098226e-06,
+      "loss": 1.8624,
+      "step": 7860
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 14.684171676635742,
+      "learning_rate": 4.243760649493836e-06,
+      "loss": 1.2921,
+      "step": 7861
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 20.421449661254883,
+      "learning_rate": 4.241756038889446e-06,
+      "loss": 1.2322,
+      "step": 7862
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 11.350706100463867,
+      "learning_rate": 4.239751428285056e-06,
+      "loss": 1.3288,
+      "step": 7863
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 21.545028686523438,
+      "learning_rate": 4.2377468176806655e-06,
+      "loss": 1.4258,
+      "step": 7864
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 18.22039222717285,
+      "learning_rate": 4.2357422070762764e-06,
+      "loss": 1.4584,
+      "step": 7865
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 28.657711029052734,
+      "learning_rate": 4.233737596471886e-06,
+      "loss": 2.444,
+      "step": 7866
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 7.091336727142334,
+      "learning_rate": 4.231732985867496e-06,
+      "loss": 1.0055,
+      "step": 7867
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 10.07259750366211,
+      "learning_rate": 4.229728375263106e-06,
+      "loss": 0.9513,
+      "step": 7868
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 32.32286834716797,
+      "learning_rate": 4.227723764658715e-06,
+      "loss": 1.2243,
+      "step": 7869
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 14.63747787475586,
+      "learning_rate": 4.225719154054325e-06,
+      "loss": 1.3113,
+      "step": 7870
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 13.216215133666992,
+      "learning_rate": 4.223714543449935e-06,
+      "loss": 1.3608,
+      "step": 7871
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 22.928085327148438,
+      "learning_rate": 4.221709932845545e-06,
+      "loss": 1.5632,
+      "step": 7872
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 14.180635452270508,
+      "learning_rate": 4.2197053222411546e-06,
+      "loss": 1.0618,
+      "step": 7873
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 13.465377807617188,
+      "learning_rate": 4.217700711636765e-06,
+      "loss": 2.2106,
+      "step": 7874
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 21.83644676208496,
+      "learning_rate": 4.215696101032375e-06,
+      "loss": 1.1141,
+      "step": 7875
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 9.465641021728516,
+      "learning_rate": 4.213691490427985e-06,
+      "loss": 0.9889,
+      "step": 7876
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 38.847965240478516,
+      "learning_rate": 4.211686879823594e-06,
+      "loss": 2.6969,
+      "step": 7877
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 12.433218002319336,
+      "learning_rate": 4.209682269219205e-06,
+      "loss": 0.5288,
+      "step": 7878
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 13.307668685913086,
+      "learning_rate": 4.207677658614814e-06,
+      "loss": 0.9523,
+      "step": 7879
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 57.23488235473633,
+      "learning_rate": 4.205673048010424e-06,
+      "loss": 3.0293,
+      "step": 7880
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 16.167064666748047,
+      "learning_rate": 4.203668437406034e-06,
+      "loss": 1.9897,
+      "step": 7881
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 29.6549072265625,
+      "learning_rate": 4.2016638268016445e-06,
+      "loss": 1.3998,
+      "step": 7882
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 10.331962585449219,
+      "learning_rate": 4.199659216197254e-06,
+      "loss": 0.8686,
+      "step": 7883
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 16.08567237854004,
+      "learning_rate": 4.197654605592864e-06,
+      "loss": 1.3522,
+      "step": 7884
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 21.387054443359375,
+      "learning_rate": 4.195649994988474e-06,
+      "loss": 1.0533,
+      "step": 7885
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 18.664920806884766,
+      "learning_rate": 4.193645384384083e-06,
+      "loss": 1.7003,
+      "step": 7886
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 12.59611988067627,
+      "learning_rate": 4.191640773779694e-06,
+      "loss": 1.0916,
+      "step": 7887
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 11.082344055175781,
+      "learning_rate": 4.189636163175303e-06,
+      "loss": 1.3448,
+      "step": 7888
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 15.192607879638672,
+      "learning_rate": 4.187631552570913e-06,
+      "loss": 1.1133,
+      "step": 7889
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 12.27822494506836,
+      "learning_rate": 4.1856269419665235e-06,
+      "loss": 1.4396,
+      "step": 7890
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 18.4125919342041,
+      "learning_rate": 4.1836223313621335e-06,
+      "loss": 0.9025,
+      "step": 7891
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 20.866134643554688,
+      "learning_rate": 4.181617720757743e-06,
+      "loss": 1.1648,
+      "step": 7892
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 16.63862419128418,
+      "learning_rate": 4.179613110153353e-06,
+      "loss": 1.4421,
+      "step": 7893
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 9.658064842224121,
+      "learning_rate": 4.177608499548963e-06,
+      "loss": 0.6854,
+      "step": 7894
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 12.731464385986328,
+      "learning_rate": 4.175603888944573e-06,
+      "loss": 1.374,
+      "step": 7895
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 8.585317611694336,
+      "learning_rate": 4.173599278340183e-06,
+      "loss": 0.5775,
+      "step": 7896
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 24.50328826904297,
+      "learning_rate": 4.171594667735792e-06,
+      "loss": 1.5935,
+      "step": 7897
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 16.08628273010254,
+      "learning_rate": 4.1695900571314024e-06,
+      "loss": 1.0627,
+      "step": 7898
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 15.948634147644043,
+      "learning_rate": 4.1675854465270125e-06,
+      "loss": 1.4136,
+      "step": 7899
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 31.818986892700195,
+      "learning_rate": 4.165580835922623e-06,
+      "loss": 1.8128,
+      "step": 7900
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 37.07505798339844,
+      "learning_rate": 4.163576225318232e-06,
+      "loss": 1.9246,
+      "step": 7901
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 15.36788558959961,
+      "learning_rate": 4.161571614713843e-06,
+      "loss": 1.9555,
+      "step": 7902
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 17.946441650390625,
+      "learning_rate": 4.159567004109452e-06,
+      "loss": 1.1702,
+      "step": 7903
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 12.651650428771973,
+      "learning_rate": 4.157562393505062e-06,
+      "loss": 1.0666,
+      "step": 7904
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 31.045000076293945,
+      "learning_rate": 4.155557782900672e-06,
+      "loss": 1.553,
+      "step": 7905
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 11.69579029083252,
+      "learning_rate": 4.153553172296282e-06,
+      "loss": 1.5206,
+      "step": 7906
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 37.5781135559082,
+      "learning_rate": 4.1515485616918915e-06,
+      "loss": 2.1329,
+      "step": 7907
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 21.697532653808594,
+      "learning_rate": 4.149543951087502e-06,
+      "loss": 1.4265,
+      "step": 7908
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 9.745403289794922,
+      "learning_rate": 4.147539340483112e-06,
+      "loss": 1.1323,
+      "step": 7909
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 15.652579307556152,
+      "learning_rate": 4.145534729878721e-06,
+      "loss": 1.3761,
+      "step": 7910
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 12.394492149353027,
+      "learning_rate": 4.143530119274332e-06,
+      "loss": 0.9479,
+      "step": 7911
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 19.919248580932617,
+      "learning_rate": 4.141525508669941e-06,
+      "loss": 1.2497,
+      "step": 7912
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 13.425742149353027,
+      "learning_rate": 4.139520898065551e-06,
+      "loss": 1.0198,
+      "step": 7913
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 17.13833999633789,
+      "learning_rate": 4.13751628746116e-06,
+      "loss": 2.0541,
+      "step": 7914
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 13.618558883666992,
+      "learning_rate": 4.135511676856771e-06,
+      "loss": 1.3842,
+      "step": 7915
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 5.930107593536377,
+      "learning_rate": 4.1335070662523806e-06,
+      "loss": 0.517,
+      "step": 7916
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 4.825802803039551,
+      "learning_rate": 4.131502455647991e-06,
+      "loss": 0.3964,
+      "step": 7917
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 25.973148345947266,
+      "learning_rate": 4.129497845043601e-06,
+      "loss": 0.9199,
+      "step": 7918
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 9.384675979614258,
+      "learning_rate": 4.127493234439211e-06,
+      "loss": 1.1662,
+      "step": 7919
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 59.6272087097168,
+      "learning_rate": 4.12548862383482e-06,
+      "loss": 2.4387,
+      "step": 7920
+    },
+    {
+      "epoch": 2.38,
+      "eval_loss": 0.17472867667675018,
+      "eval_runtime": 43.5367,
+      "eval_samples_per_second": 33.971,
+      "eval_steps_per_second": 33.971,
+      "step": 7920
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 17.422163009643555,
+      "learning_rate": 4.12348401323043e-06,
+      "loss": 0.9547,
+      "step": 7921
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 19.566699981689453,
+      "learning_rate": 4.12147940262604e-06,
+      "loss": 1.364,
+      "step": 7922
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 13.921435356140137,
+      "learning_rate": 4.11947479202165e-06,
+      "loss": 1.5498,
+      "step": 7923
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 46.268829345703125,
+      "learning_rate": 4.11747018141726e-06,
+      "loss": 1.7792,
+      "step": 7924
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 37.72319793701172,
+      "learning_rate": 4.11546557081287e-06,
+      "loss": 1.582,
+      "step": 7925
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 13.244497299194336,
+      "learning_rate": 4.11346096020848e-06,
+      "loss": 1.7341,
+      "step": 7926
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 10.264554977416992,
+      "learning_rate": 4.11145634960409e-06,
+      "loss": 0.902,
+      "step": 7927
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 16.762826919555664,
+      "learning_rate": 4.1094517389997e-06,
+      "loss": 1.2989,
+      "step": 7928
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 27.864768981933594,
+      "learning_rate": 4.107447128395309e-06,
+      "loss": 1.0548,
+      "step": 7929
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 9.768196105957031,
+      "learning_rate": 4.105442517790919e-06,
+      "loss": 0.7929,
+      "step": 7930
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 8.759325981140137,
+      "learning_rate": 4.103437907186529e-06,
+      "loss": 0.6668,
+      "step": 7931
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 29.228063583374023,
+      "learning_rate": 4.101433296582139e-06,
+      "loss": 1.5968,
+      "step": 7932
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 15.066993713378906,
+      "learning_rate": 4.0994286859777495e-06,
+      "loss": 1.457,
+      "step": 7933
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 12.962151527404785,
+      "learning_rate": 4.097424075373359e-06,
+      "loss": 1.568,
+      "step": 7934
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 9.651278495788574,
+      "learning_rate": 4.095419464768969e-06,
+      "loss": 0.9769,
+      "step": 7935
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 27.313953399658203,
+      "learning_rate": 4.093414854164579e-06,
+      "loss": 1.1726,
+      "step": 7936
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 30.882871627807617,
+      "learning_rate": 4.091410243560189e-06,
+      "loss": 2.3507,
+      "step": 7937
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 16.88389015197754,
+      "learning_rate": 4.089405632955798e-06,
+      "loss": 0.9228,
+      "step": 7938
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 54.354469299316406,
+      "learning_rate": 4.087401022351409e-06,
+      "loss": 2.5421,
+      "step": 7939
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 18.750852584838867,
+      "learning_rate": 4.085396411747018e-06,
+      "loss": 1.2235,
+      "step": 7940
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 7.334155559539795,
+      "learning_rate": 4.0833918011426284e-06,
+      "loss": 0.647,
+      "step": 7941
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 14.79666519165039,
+      "learning_rate": 4.0813871905382385e-06,
+      "loss": 0.8448,
+      "step": 7942
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 35.3593864440918,
+      "learning_rate": 4.079382579933849e-06,
+      "loss": 1.4009,
+      "step": 7943
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 25.9478702545166,
+      "learning_rate": 4.077377969329458e-06,
+      "loss": 1.7573,
+      "step": 7944
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 13.793614387512207,
+      "learning_rate": 4.075373358725068e-06,
+      "loss": 1.5591,
+      "step": 7945
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 15.494868278503418,
+      "learning_rate": 4.073368748120678e-06,
+      "loss": 1.6407,
+      "step": 7946
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 26.147010803222656,
+      "learning_rate": 4.071364137516287e-06,
+      "loss": 1.6419,
+      "step": 7947
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 19.630136489868164,
+      "learning_rate": 4.069359526911898e-06,
+      "loss": 1.5473,
+      "step": 7948
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 75.69630432128906,
+      "learning_rate": 4.0673549163075074e-06,
+      "loss": 2.2357,
+      "step": 7949
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 82.69189453125,
+      "learning_rate": 4.0653503057031175e-06,
+      "loss": 1.675,
+      "step": 7950
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 12.869571685791016,
+      "learning_rate": 4.063345695098728e-06,
+      "loss": 1.1058,
+      "step": 7951
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 16.09258270263672,
+      "learning_rate": 4.061341084494338e-06,
+      "loss": 1.3042,
+      "step": 7952
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 15.368246078491211,
+      "learning_rate": 4.059336473889947e-06,
+      "loss": 0.7744,
+      "step": 7953
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 56.09071350097656,
+      "learning_rate": 4.057331863285557e-06,
+      "loss": 1.4905,
+      "step": 7954
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 11.899616241455078,
+      "learning_rate": 4.055327252681167e-06,
+      "loss": 1.1252,
+      "step": 7955
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 17.915721893310547,
+      "learning_rate": 4.053322642076777e-06,
+      "loss": 1.8459,
+      "step": 7956
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 12.431897163391113,
+      "learning_rate": 4.051318031472386e-06,
+      "loss": 1.0242,
+      "step": 7957
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 15.987489700317383,
+      "learning_rate": 4.0493134208679965e-06,
+      "loss": 2.4057,
+      "step": 7958
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 14.070905685424805,
+      "learning_rate": 4.0473088102636066e-06,
+      "loss": 1.2005,
+      "step": 7959
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 26.776086807250977,
+      "learning_rate": 4.045304199659217e-06,
+      "loss": 1.1287,
+      "step": 7960
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 21.03402328491211,
+      "learning_rate": 4.043299589054827e-06,
+      "loss": 1.6414,
+      "step": 7961
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 16.39931297302246,
+      "learning_rate": 4.041294978450436e-06,
+      "loss": 1.7569,
+      "step": 7962
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 22.343650817871094,
+      "learning_rate": 4.039290367846046e-06,
+      "loss": 1.0703,
+      "step": 7963
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 21.395599365234375,
+      "learning_rate": 4.037285757241656e-06,
+      "loss": 2.1687,
+      "step": 7964
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 53.524776458740234,
+      "learning_rate": 4.035281146637266e-06,
+      "loss": 1.6589,
+      "step": 7965
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 12.918240547180176,
+      "learning_rate": 4.0332765360328755e-06,
+      "loss": 0.878,
+      "step": 7966
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 10.968676567077637,
+      "learning_rate": 4.031271925428486e-06,
+      "loss": 2.1511,
+      "step": 7967
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 21.93174934387207,
+      "learning_rate": 4.029267314824096e-06,
+      "loss": 1.5569,
+      "step": 7968
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 18.74276351928711,
+      "learning_rate": 4.027262704219706e-06,
+      "loss": 0.7945,
+      "step": 7969
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 9.503168106079102,
+      "learning_rate": 4.025258093615316e-06,
+      "loss": 0.8866,
+      "step": 7970
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 39.47118377685547,
+      "learning_rate": 4.023253483010925e-06,
+      "loss": 1.4191,
+      "step": 7971
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 11.129521369934082,
+      "learning_rate": 4.021248872406535e-06,
+      "loss": 1.1143,
+      "step": 7972
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 13.493324279785156,
+      "learning_rate": 4.019244261802145e-06,
+      "loss": 1.5284,
+      "step": 7973
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 12.583949089050293,
+      "learning_rate": 4.017239651197755e-06,
+      "loss": 1.6311,
+      "step": 7974
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 7.225354194641113,
+      "learning_rate": 4.0152350405933645e-06,
+      "loss": 0.5856,
+      "step": 7975
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 15.300594329833984,
+      "learning_rate": 4.0132304299889755e-06,
+      "loss": 1.0028,
+      "step": 7976
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 16.86861228942871,
+      "learning_rate": 4.011225819384585e-06,
+      "loss": 1.7745,
+      "step": 7977
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 20.883081436157227,
+      "learning_rate": 4.009221208780195e-06,
+      "loss": 1.6959,
+      "step": 7978
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 120.636962890625,
+      "learning_rate": 4.007216598175805e-06,
+      "loss": 2.2987,
+      "step": 7979
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 76.55606842041016,
+      "learning_rate": 4.005211987571415e-06,
+      "loss": 1.4258,
+      "step": 7980
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 15.42659854888916,
+      "learning_rate": 4.003207376967024e-06,
+      "loss": 1.5873,
+      "step": 7981
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 9.378588676452637,
+      "learning_rate": 4.001202766362634e-06,
+      "loss": 0.9196,
+      "step": 7982
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 12.483980178833008,
+      "learning_rate": 3.999198155758244e-06,
+      "loss": 0.9071,
+      "step": 7983
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 7.117259502410889,
+      "learning_rate": 3.9971935451538544e-06,
+      "loss": 0.6292,
+      "step": 7984
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 14.306236267089844,
+      "learning_rate": 3.9951889345494645e-06,
+      "loss": 1.0767,
+      "step": 7985
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 46.9137077331543,
+      "learning_rate": 3.993184323945074e-06,
+      "loss": 2.4379,
+      "step": 7986
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 17.20838737487793,
+      "learning_rate": 3.991179713340684e-06,
+      "loss": 1.2174,
+      "step": 7987
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 9.353306770324707,
+      "learning_rate": 3.989175102736294e-06,
+      "loss": 1.0121,
+      "step": 7988
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 17.213781356811523,
+      "learning_rate": 3.987170492131904e-06,
+      "loss": 1.4443,
+      "step": 7989
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 6.010331630706787,
+      "learning_rate": 3.985165881527513e-06,
+      "loss": 0.8314,
+      "step": 7990
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 12.470282554626465,
+      "learning_rate": 3.983161270923124e-06,
+      "loss": 1.9135,
+      "step": 7991
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 25.282251358032227,
+      "learning_rate": 3.9811566603187334e-06,
+      "loss": 1.9866,
+      "step": 7992
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 12.670585632324219,
+      "learning_rate": 3.9791520497143435e-06,
+      "loss": 0.8544,
+      "step": 7993
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 21.99867057800293,
+      "learning_rate": 3.977147439109954e-06,
+      "loss": 0.906,
+      "step": 7994
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 26.105510711669922,
+      "learning_rate": 3.975142828505563e-06,
+      "loss": 1.1245,
+      "step": 7995
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 27.73430824279785,
+      "learning_rate": 3.973138217901173e-06,
+      "loss": 1.5627,
+      "step": 7996
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 31.076269149780273,
+      "learning_rate": 3.971133607296783e-06,
+      "loss": 1.5272,
+      "step": 7997
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 20.252588272094727,
+      "learning_rate": 3.969128996692393e-06,
+      "loss": 1.3878,
+      "step": 7998
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 15.601667404174805,
+      "learning_rate": 3.967124386088002e-06,
+      "loss": 1.6348,
+      "step": 7999
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 15.273148536682129,
+      "learning_rate": 3.965119775483612e-06,
+      "loss": 2.6779,
+      "step": 8000
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 14.738374710083008,
+      "learning_rate": 3.9631151648792225e-06,
+      "loss": 1.3222,
+      "step": 8001
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 12.968079566955566,
+      "learning_rate": 3.9611105542748326e-06,
+      "loss": 0.8718,
+      "step": 8002
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 15.838987350463867,
+      "learning_rate": 3.959105943670442e-06,
+      "loss": 1.1175,
+      "step": 8003
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 11.581687927246094,
+      "learning_rate": 3.957101333066053e-06,
+      "loss": 2.2458,
+      "step": 8004
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 13.473747253417969,
+      "learning_rate": 3.955096722461662e-06,
+      "loss": 1.0507,
+      "step": 8005
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 22.461040496826172,
+      "learning_rate": 3.953092111857272e-06,
+      "loss": 1.8377,
+      "step": 8006
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 9.827531814575195,
+      "learning_rate": 3.951087501252882e-06,
+      "loss": 1.3932,
+      "step": 8007
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 17.85820198059082,
+      "learning_rate": 3.949082890648492e-06,
+      "loss": 1.0887,
+      "step": 8008
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 10.893664360046387,
+      "learning_rate": 3.9470782800441015e-06,
+      "loss": 0.6649,
+      "step": 8009
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 27.924177169799805,
+      "learning_rate": 3.9450736694397116e-06,
+      "loss": 1.7343,
+      "step": 8010
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 7.996644973754883,
+      "learning_rate": 3.943069058835322e-06,
+      "loss": 1.1344,
+      "step": 8011
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 29.98558807373047,
+      "learning_rate": 3.941064448230931e-06,
+      "loss": 1.4769,
+      "step": 8012
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 27.507963180541992,
+      "learning_rate": 3.939059837626542e-06,
+      "loss": 0.83,
+      "step": 8013
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 7.844414710998535,
+      "learning_rate": 3.937055227022151e-06,
+      "loss": 0.9977,
+      "step": 8014
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 51.909061431884766,
+      "learning_rate": 3.935050616417761e-06,
+      "loss": 2.3273,
+      "step": 8015
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 10.7676420211792,
+      "learning_rate": 3.933046005813371e-06,
+      "loss": 0.986,
+      "step": 8016
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 14.182157516479492,
+      "learning_rate": 3.931041395208981e-06,
+      "loss": 1.3547,
+      "step": 8017
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 15.017866134643555,
+      "learning_rate": 3.9290367846045905e-06,
+      "loss": 0.7605,
+      "step": 8018
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 22.285579681396484,
+      "learning_rate": 3.927032174000201e-06,
+      "loss": 1.5245,
+      "step": 8019
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 69.15910339355469,
+      "learning_rate": 3.925027563395811e-06,
+      "loss": 2.0588,
+      "step": 8020
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 57.589107513427734,
+      "learning_rate": 3.923022952791421e-06,
+      "loss": 2.0236,
+      "step": 8021
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 38.282440185546875,
+      "learning_rate": 3.921018342187031e-06,
+      "loss": 1.923,
+      "step": 8022
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 16.257158279418945,
+      "learning_rate": 3.91901373158264e-06,
+      "loss": 1.2573,
+      "step": 8023
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 13.993424415588379,
+      "learning_rate": 3.91700912097825e-06,
+      "loss": 1.1249,
+      "step": 8024
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 18.90182876586914,
+      "learning_rate": 3.91500451037386e-06,
+      "loss": 1.4962,
+      "step": 8025
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 19.353958129882812,
+      "learning_rate": 3.91299989976947e-06,
+      "loss": 1.0571,
+      "step": 8026
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 31.27533531188965,
+      "learning_rate": 3.91099528916508e-06,
+      "loss": 1.6633,
+      "step": 8027
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 15.394329071044922,
+      "learning_rate": 3.9089906785606905e-06,
+      "loss": 1.4122,
+      "step": 8028
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 77.46128845214844,
+      "learning_rate": 3.9069860679563e-06,
+      "loss": 1.7462,
+      "step": 8029
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 33.42092514038086,
+      "learning_rate": 3.90498145735191e-06,
+      "loss": 1.3732,
+      "step": 8030
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 9.593501091003418,
+      "learning_rate": 3.90297684674752e-06,
+      "loss": 0.9003,
+      "step": 8031
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 16.690101623535156,
+      "learning_rate": 3.900972236143129e-06,
+      "loss": 1.6157,
+      "step": 8032
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 14.943684577941895,
+      "learning_rate": 3.898967625538739e-06,
+      "loss": 1.5163,
+      "step": 8033
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 59.907310485839844,
+      "learning_rate": 3.896963014934349e-06,
+      "loss": 1.6031,
+      "step": 8034
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 32.62445068359375,
+      "learning_rate": 3.8949584043299594e-06,
+      "loss": 1.2945,
+      "step": 8035
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 92.56492614746094,
+      "learning_rate": 3.892953793725569e-06,
+      "loss": 1.5538,
+      "step": 8036
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 11.2645902633667,
+      "learning_rate": 3.890949183121179e-06,
+      "loss": 1.8437,
+      "step": 8037
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 13.851442337036133,
+      "learning_rate": 3.888944572516789e-06,
+      "loss": 0.9473,
+      "step": 8038
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 17.675752639770508,
+      "learning_rate": 3.886939961912399e-06,
+      "loss": 1.4754,
+      "step": 8039
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 14.534573554992676,
+      "learning_rate": 3.884935351308008e-06,
+      "loss": 1.2699,
+      "step": 8040
+    },
+    {
+      "epoch": 2.42,
+      "eval_loss": 0.16647757589817047,
+      "eval_runtime": 43.4766,
+      "eval_samples_per_second": 34.018,
+      "eval_steps_per_second": 34.018,
+      "step": 8040
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 10.053756713867188,
+      "learning_rate": 3.882930740703619e-06,
+      "loss": 1.1651,
+      "step": 8041
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 15.086660385131836,
+      "learning_rate": 3.880926130099228e-06,
+      "loss": 1.2307,
+      "step": 8042
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 45.48273468017578,
+      "learning_rate": 3.878921519494838e-06,
+      "loss": 1.7186,
+      "step": 8043
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 34.40294647216797,
+      "learning_rate": 3.8769169088904485e-06,
+      "loss": 2.7345,
+      "step": 8044
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 14.807568550109863,
+      "learning_rate": 3.8749122982860586e-06,
+      "loss": 0.8994,
+      "step": 8045
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 10.855109214782715,
+      "learning_rate": 3.872907687681668e-06,
+      "loss": 1.2174,
+      "step": 8046
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 36.307777404785156,
+      "learning_rate": 3.870903077077278e-06,
+      "loss": 2.1917,
+      "step": 8047
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 22.572141647338867,
+      "learning_rate": 3.868898466472888e-06,
+      "loss": 0.8395,
+      "step": 8048
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 17.936315536499023,
+      "learning_rate": 3.866893855868497e-06,
+      "loss": 2.4869,
+      "step": 8049
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 12.054944038391113,
+      "learning_rate": 3.864889245264108e-06,
+      "loss": 1.4992,
+      "step": 8050
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 10.359464645385742,
+      "learning_rate": 3.862884634659717e-06,
+      "loss": 1.1123,
+      "step": 8051
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 21.69719886779785,
+      "learning_rate": 3.8608800240553275e-06,
+      "loss": 1.9015,
+      "step": 8052
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 9.15548038482666,
+      "learning_rate": 3.8588754134509376e-06,
+      "loss": 1.1113,
+      "step": 8053
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 7.168182373046875,
+      "learning_rate": 3.856870802846548e-06,
+      "loss": 0.7675,
+      "step": 8054
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 23.34624481201172,
+      "learning_rate": 3.854866192242157e-06,
+      "loss": 1.1852,
+      "step": 8055
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 8.906826972961426,
+      "learning_rate": 3.852861581637767e-06,
+      "loss": 0.7492,
+      "step": 8056
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 17.47923469543457,
+      "learning_rate": 3.850856971033377e-06,
+      "loss": 1.1435,
+      "step": 8057
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 8.960250854492188,
+      "learning_rate": 3.848852360428987e-06,
+      "loss": 0.9197,
+      "step": 8058
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 17.509536743164062,
+      "learning_rate": 3.846847749824597e-06,
+      "loss": 1.5889,
+      "step": 8059
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 14.029973030090332,
+      "learning_rate": 3.8448431392202065e-06,
+      "loss": 1.5197,
+      "step": 8060
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 13.85709285736084,
+      "learning_rate": 3.8428385286158165e-06,
+      "loss": 1.1813,
+      "step": 8061
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 11.46907901763916,
+      "learning_rate": 3.840833918011427e-06,
+      "loss": 1.1435,
+      "step": 8062
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 11.111536026000977,
+      "learning_rate": 3.838829307407037e-06,
+      "loss": 1.6378,
+      "step": 8063
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 13.92691421508789,
+      "learning_rate": 3.836824696802646e-06,
+      "loss": 1.0666,
+      "step": 8064
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 11.448002815246582,
+      "learning_rate": 3.834820086198257e-06,
+      "loss": 1.2318,
+      "step": 8065
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 13.583879470825195,
+      "learning_rate": 3.832815475593866e-06,
+      "loss": 0.7262,
+      "step": 8066
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 17.306129455566406,
+      "learning_rate": 3.830810864989476e-06,
+      "loss": 1.5805,
+      "step": 8067
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 17.71708106994629,
+      "learning_rate": 3.828806254385086e-06,
+      "loss": 1.5627,
+      "step": 8068
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 13.716277122497559,
+      "learning_rate": 3.826801643780696e-06,
+      "loss": 1.223,
+      "step": 8069
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 9.647749900817871,
+      "learning_rate": 3.824797033176306e-06,
+      "loss": 2.1291,
+      "step": 8070
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 15.30841064453125,
+      "learning_rate": 3.822792422571916e-06,
+      "loss": 1.1331,
+      "step": 8071
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 18.855783462524414,
+      "learning_rate": 3.820787811967526e-06,
+      "loss": 1.7297,
+      "step": 8072
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 35.978248596191406,
+      "learning_rate": 3.818783201363135e-06,
+      "loss": 0.7591,
+      "step": 8073
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 19.79545021057129,
+      "learning_rate": 3.816778590758746e-06,
+      "loss": 1.878,
+      "step": 8074
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 33.37739562988281,
+      "learning_rate": 3.814773980154355e-06,
+      "loss": 1.5872,
+      "step": 8075
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 19.29977035522461,
+      "learning_rate": 3.8127693695499653e-06,
+      "loss": 0.9943,
+      "step": 8076
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 22.01044273376465,
+      "learning_rate": 3.8107647589455753e-06,
+      "loss": 1.9346,
+      "step": 8077
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 24.812864303588867,
+      "learning_rate": 3.808760148341185e-06,
+      "loss": 1.4286,
+      "step": 8078
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 16.758087158203125,
+      "learning_rate": 3.8067555377367947e-06,
+      "loss": 0.6716,
+      "step": 8079
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 10.366033554077148,
+      "learning_rate": 3.8047509271324047e-06,
+      "loss": 1.3467,
+      "step": 8080
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 20.55314064025879,
+      "learning_rate": 3.802746316528015e-06,
+      "loss": 1.4071,
+      "step": 8081
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 31.28022003173828,
+      "learning_rate": 3.8007417059236245e-06,
+      "loss": 1.7199,
+      "step": 8082
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 6.907041072845459,
+      "learning_rate": 3.798737095319234e-06,
+      "loss": 0.3853,
+      "step": 8083
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 24.64528465270996,
+      "learning_rate": 3.7967324847148447e-06,
+      "loss": 1.4183,
+      "step": 8084
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 10.540509223937988,
+      "learning_rate": 3.7947278741104543e-06,
+      "loss": 1.0129,
+      "step": 8085
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 13.344951629638672,
+      "learning_rate": 3.792723263506064e-06,
+      "loss": 0.989,
+      "step": 8086
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 8.0833158493042,
+      "learning_rate": 3.7907186529016745e-06,
+      "loss": 0.8597,
+      "step": 8087
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 19.457054138183594,
+      "learning_rate": 3.788714042297284e-06,
+      "loss": 1.3676,
+      "step": 8088
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 16.89392852783203,
+      "learning_rate": 3.786709431692894e-06,
+      "loss": 1.1124,
+      "step": 8089
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 9.098607063293457,
+      "learning_rate": 3.784704821088504e-06,
+      "loss": 0.5794,
+      "step": 8090
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 41.21145248413086,
+      "learning_rate": 3.7827002104841136e-06,
+      "loss": 1.9724,
+      "step": 8091
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 5.302923679351807,
+      "learning_rate": 3.7806955998797236e-06,
+      "loss": 0.579,
+      "step": 8092
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 14.981736183166504,
+      "learning_rate": 3.7786909892753337e-06,
+      "loss": 2.1375,
+      "step": 8093
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 20.563865661621094,
+      "learning_rate": 3.7766863786709434e-06,
+      "loss": 1.137,
+      "step": 8094
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 10.908076286315918,
+      "learning_rate": 3.774681768066553e-06,
+      "loss": 1.2199,
+      "step": 8095
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 79.2155532836914,
+      "learning_rate": 3.7726771574621636e-06,
+      "loss": 2.0151,
+      "step": 8096
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 187.99632263183594,
+      "learning_rate": 3.7706725468577732e-06,
+      "loss": 2.0555,
+      "step": 8097
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 38.596275329589844,
+      "learning_rate": 3.768667936253383e-06,
+      "loss": 2.0002,
+      "step": 8098
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 8.95095157623291,
+      "learning_rate": 3.766663325648993e-06,
+      "loss": 1.0331,
+      "step": 8099
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 17.133882522583008,
+      "learning_rate": 3.764658715044603e-06,
+      "loss": 1.5345,
+      "step": 8100
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 8.687376976013184,
+      "learning_rate": 3.7626541044402127e-06,
+      "loss": 0.7787,
+      "step": 8101
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 69.27308654785156,
+      "learning_rate": 3.760649493835823e-06,
+      "loss": 1.252,
+      "step": 8102
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 23.108619689941406,
+      "learning_rate": 3.7586448832314325e-06,
+      "loss": 1.3771,
+      "step": 8103
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 33.36818313598633,
+      "learning_rate": 3.7566402726270425e-06,
+      "loss": 1.6731,
+      "step": 8104
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 13.890486717224121,
+      "learning_rate": 3.7546356620226526e-06,
+      "loss": 1.39,
+      "step": 8105
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 9.70341682434082,
+      "learning_rate": 3.7526310514182623e-06,
+      "loss": 0.7873,
+      "step": 8106
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 15.94449234008789,
+      "learning_rate": 3.750626440813872e-06,
+      "loss": 0.9446,
+      "step": 8107
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 19.849061965942383,
+      "learning_rate": 3.7486218302094825e-06,
+      "loss": 1.6883,
+      "step": 8108
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 19.96930503845215,
+      "learning_rate": 3.746617219605092e-06,
+      "loss": 1.2279,
+      "step": 8109
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 12.898889541625977,
+      "learning_rate": 3.7446126090007018e-06,
+      "loss": 1.6165,
+      "step": 8110
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 12.426019668579102,
+      "learning_rate": 3.742607998396312e-06,
+      "loss": 1.4589,
+      "step": 8111
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 20.18886375427246,
+      "learning_rate": 3.740603387791922e-06,
+      "loss": 1.7677,
+      "step": 8112
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 19.88523292541504,
+      "learning_rate": 3.7385987771875316e-06,
+      "loss": 2.5334,
+      "step": 8113
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 39.80482482910156,
+      "learning_rate": 3.7365941665831417e-06,
+      "loss": 1.2999,
+      "step": 8114
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 20.459688186645508,
+      "learning_rate": 3.7345895559787513e-06,
+      "loss": 1.1679,
+      "step": 8115
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 6.700068950653076,
+      "learning_rate": 3.7325849453743614e-06,
+      "loss": 0.7833,
+      "step": 8116
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 10.873579978942871,
+      "learning_rate": 3.7305803347699715e-06,
+      "loss": 1.1187,
+      "step": 8117
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 11.668147087097168,
+      "learning_rate": 3.728575724165581e-06,
+      "loss": 1.436,
+      "step": 8118
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 24.042890548706055,
+      "learning_rate": 3.726571113561191e-06,
+      "loss": 1.6814,
+      "step": 8119
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 7.250650882720947,
+      "learning_rate": 3.7245665029568013e-06,
+      "loss": 1.062,
+      "step": 8120
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 8.547741889953613,
+      "learning_rate": 3.722561892352411e-06,
+      "loss": 1.0638,
+      "step": 8121
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 9.865437507629395,
+      "learning_rate": 3.7205572817480207e-06,
+      "loss": 0.8298,
+      "step": 8122
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 11.00238037109375,
+      "learning_rate": 3.7185526711436303e-06,
+      "loss": 0.7298,
+      "step": 8123
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 10.733601570129395,
+      "learning_rate": 3.716548060539241e-06,
+      "loss": 0.9721,
+      "step": 8124
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 32.854774475097656,
+      "learning_rate": 3.7145434499348505e-06,
+      "loss": 1.4885,
+      "step": 8125
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 11.409563064575195,
+      "learning_rate": 3.71253883933046e-06,
+      "loss": 0.8726,
+      "step": 8126
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 37.438106536865234,
+      "learning_rate": 3.7105342287260702e-06,
+      "loss": 1.3614,
+      "step": 8127
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 28.72161865234375,
+      "learning_rate": 3.70852961812168e-06,
+      "loss": 1.0996,
+      "step": 8128
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 10.785248756408691,
+      "learning_rate": 3.70652500751729e-06,
+      "loss": 0.7049,
+      "step": 8129
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 9.802728652954102,
+      "learning_rate": 3.7045203969129e-06,
+      "loss": 1.0367,
+      "step": 8130
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 14.082221984863281,
+      "learning_rate": 3.7025157863085097e-06,
+      "loss": 1.0125,
+      "step": 8131
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 63.74949645996094,
+      "learning_rate": 3.7005111757041194e-06,
+      "loss": 2.3996,
+      "step": 8132
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 9.598031044006348,
+      "learning_rate": 3.69850656509973e-06,
+      "loss": 1.0888,
+      "step": 8133
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 15.858400344848633,
+      "learning_rate": 3.6965019544953396e-06,
+      "loss": 1.055,
+      "step": 8134
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 10.83476734161377,
+      "learning_rate": 3.6944973438909492e-06,
+      "loss": 1.9326,
+      "step": 8135
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 10.07654857635498,
+      "learning_rate": 3.6924927332865597e-06,
+      "loss": 1.5472,
+      "step": 8136
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 16.537139892578125,
+      "learning_rate": 3.6904881226821694e-06,
+      "loss": 1.2108,
+      "step": 8137
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 15.194998741149902,
+      "learning_rate": 3.688483512077779e-06,
+      "loss": 1.0623,
+      "step": 8138
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 43.36076736450195,
+      "learning_rate": 3.686478901473389e-06,
+      "loss": 1.2945,
+      "step": 8139
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 9.480472564697266,
+      "learning_rate": 3.684474290868999e-06,
+      "loss": 0.5474,
+      "step": 8140
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 36.88868713378906,
+      "learning_rate": 3.682469680264609e-06,
+      "loss": 1.0721,
+      "step": 8141
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 42.60389709472656,
+      "learning_rate": 3.680465069660219e-06,
+      "loss": 1.4549,
+      "step": 8142
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 22.67746353149414,
+      "learning_rate": 3.6784604590558286e-06,
+      "loss": 2.1135,
+      "step": 8143
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 8.67175006866455,
+      "learning_rate": 3.6764558484514383e-06,
+      "loss": 1.0506,
+      "step": 8144
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 18.72547149658203,
+      "learning_rate": 3.674451237847049e-06,
+      "loss": 0.891,
+      "step": 8145
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 32.578861236572266,
+      "learning_rate": 3.6724466272426585e-06,
+      "loss": 1.6625,
+      "step": 8146
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 9.158719062805176,
+      "learning_rate": 3.670442016638268e-06,
+      "loss": 0.8982,
+      "step": 8147
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 15.746097564697266,
+      "learning_rate": 3.6684374060338786e-06,
+      "loss": 1.349,
+      "step": 8148
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 12.427379608154297,
+      "learning_rate": 3.6664327954294883e-06,
+      "loss": 1.283,
+      "step": 8149
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 30.601055145263672,
+      "learning_rate": 3.664428184825098e-06,
+      "loss": 0.9753,
+      "step": 8150
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 15.714460372924805,
+      "learning_rate": 3.662423574220708e-06,
+      "loss": 1.1643,
+      "step": 8151
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 27.35734748840332,
+      "learning_rate": 3.6604189636163177e-06,
+      "loss": 1.7191,
+      "step": 8152
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 16.40424919128418,
+      "learning_rate": 3.6584143530119278e-06,
+      "loss": 1.4443,
+      "step": 8153
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 9.086562156677246,
+      "learning_rate": 3.656409742407538e-06,
+      "loss": 1.0836,
+      "step": 8154
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 16.159156799316406,
+      "learning_rate": 3.6544051318031475e-06,
+      "loss": 1.0811,
+      "step": 8155
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 32.5626220703125,
+      "learning_rate": 3.652400521198757e-06,
+      "loss": 1.5753,
+      "step": 8156
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 10.6674165725708,
+      "learning_rate": 3.6503959105943677e-06,
+      "loss": 0.9974,
+      "step": 8157
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 11.11431884765625,
+      "learning_rate": 3.6483912999899773e-06,
+      "loss": 1.3574,
+      "step": 8158
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 24.535797119140625,
+      "learning_rate": 3.646386689385587e-06,
+      "loss": 1.4364,
+      "step": 8159
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 49.98130416870117,
+      "learning_rate": 3.6443820787811975e-06,
+      "loss": 1.0015,
+      "step": 8160
+    },
+    {
+      "epoch": 2.45,
+      "eval_loss": 0.16650927066802979,
+      "eval_runtime": 44.1127,
+      "eval_samples_per_second": 33.528,
+      "eval_steps_per_second": 33.528,
+      "step": 8160
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 9.75783634185791,
+      "learning_rate": 3.642377468176807e-06,
+      "loss": 1.1418,
+      "step": 8161
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 15.587701797485352,
+      "learning_rate": 3.640372857572417e-06,
+      "loss": 0.7225,
+      "step": 8162
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 17.10492706298828,
+      "learning_rate": 3.6383682469680265e-06,
+      "loss": 1.1976,
+      "step": 8163
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 23.952566146850586,
+      "learning_rate": 3.6363636363636366e-06,
+      "loss": 1.2036,
+      "step": 8164
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 17.270137786865234,
+      "learning_rate": 3.6343590257592467e-06,
+      "loss": 1.6207,
+      "step": 8165
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 16.140460968017578,
+      "learning_rate": 3.6323544151548563e-06,
+      "loss": 0.9803,
+      "step": 8166
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 42.045997619628906,
+      "learning_rate": 3.6303498045504664e-06,
+      "loss": 2.4721,
+      "step": 8167
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 13.348779678344727,
+      "learning_rate": 3.628345193946076e-06,
+      "loss": 1.1113,
+      "step": 8168
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 36.904014587402344,
+      "learning_rate": 3.6263405833416857e-06,
+      "loss": 2.5081,
+      "step": 8169
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 23.20426368713379,
+      "learning_rate": 3.6243359727372962e-06,
+      "loss": 1.0172,
+      "step": 8170
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 16.895856857299805,
+      "learning_rate": 3.622331362132906e-06,
+      "loss": 1.0231,
+      "step": 8171
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 37.31813049316406,
+      "learning_rate": 3.6203267515285156e-06,
+      "loss": 1.3518,
+      "step": 8172
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 19.724748611450195,
+      "learning_rate": 3.618322140924126e-06,
+      "loss": 1.2338,
+      "step": 8173
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 19.960813522338867,
+      "learning_rate": 3.6163175303197357e-06,
+      "loss": 1.4858,
+      "step": 8174
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 11.546113967895508,
+      "learning_rate": 3.6143129197153454e-06,
+      "loss": 1.2553,
+      "step": 8175
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 18.865678787231445,
+      "learning_rate": 3.6123083091109555e-06,
+      "loss": 1.2506,
+      "step": 8176
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 22.531909942626953,
+      "learning_rate": 3.6103036985065656e-06,
+      "loss": 1.6922,
+      "step": 8177
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 20.21176528930664,
+      "learning_rate": 3.6082990879021752e-06,
+      "loss": 0.8544,
+      "step": 8178
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 15.008464813232422,
+      "learning_rate": 3.6062944772977853e-06,
+      "loss": 1.3876,
+      "step": 8179
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 24.066457748413086,
+      "learning_rate": 3.604289866693395e-06,
+      "loss": 1.4522,
+      "step": 8180
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 7.842005729675293,
+      "learning_rate": 3.6022852560890046e-06,
+      "loss": 0.5896,
+      "step": 8181
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 13.021004676818848,
+      "learning_rate": 3.600280645484615e-06,
+      "loss": 0.7843,
+      "step": 8182
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 21.92922592163086,
+      "learning_rate": 3.598276034880225e-06,
+      "loss": 2.2506,
+      "step": 8183
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 11.565603256225586,
+      "learning_rate": 3.5962714242758345e-06,
+      "loss": 1.0158,
+      "step": 8184
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 18.74056053161621,
+      "learning_rate": 3.594266813671445e-06,
+      "loss": 0.9444,
+      "step": 8185
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 14.562174797058105,
+      "learning_rate": 3.5922622030670546e-06,
+      "loss": 0.7885,
+      "step": 8186
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 37.87458038330078,
+      "learning_rate": 3.5902575924626643e-06,
+      "loss": 1.1931,
+      "step": 8187
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 14.50774097442627,
+      "learning_rate": 3.5882529818582744e-06,
+      "loss": 1.1908,
+      "step": 8188
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 37.05324172973633,
+      "learning_rate": 3.5862483712538845e-06,
+      "loss": 2.294,
+      "step": 8189
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 7.687615394592285,
+      "learning_rate": 3.584243760649494e-06,
+      "loss": 0.7946,
+      "step": 8190
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 10.446063995361328,
+      "learning_rate": 3.582239150045104e-06,
+      "loss": 1.2933,
+      "step": 8191
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 25.05036735534668,
+      "learning_rate": 3.580234539440714e-06,
+      "loss": 1.8224,
+      "step": 8192
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 12.898480415344238,
+      "learning_rate": 3.5782299288363235e-06,
+      "loss": 1.3031,
+      "step": 8193
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 23.968446731567383,
+      "learning_rate": 3.576225318231934e-06,
+      "loss": 2.3017,
+      "step": 8194
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 16.672903060913086,
+      "learning_rate": 3.5742207076275437e-06,
+      "loss": 0.9025,
+      "step": 8195
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 10.29025936126709,
+      "learning_rate": 3.5722160970231534e-06,
+      "loss": 1.0269,
+      "step": 8196
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 11.360820770263672,
+      "learning_rate": 3.570211486418764e-06,
+      "loss": 1.7671,
+      "step": 8197
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 28.387712478637695,
+      "learning_rate": 3.5682068758143735e-06,
+      "loss": 0.9867,
+      "step": 8198
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 37.20726776123047,
+      "learning_rate": 3.566202265209983e-06,
+      "loss": 2.4591,
+      "step": 8199
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 8.600688934326172,
+      "learning_rate": 3.5641976546055933e-06,
+      "loss": 0.8608,
+      "step": 8200
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 42.92110824584961,
+      "learning_rate": 3.562193044001203e-06,
+      "loss": 1.5033,
+      "step": 8201
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 9.241715431213379,
+      "learning_rate": 3.560188433396813e-06,
+      "loss": 0.8824,
+      "step": 8202
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 18.05348777770996,
+      "learning_rate": 3.558183822792423e-06,
+      "loss": 0.8519,
+      "step": 8203
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 13.041126251220703,
+      "learning_rate": 3.5561792121880328e-06,
+      "loss": 0.9648,
+      "step": 8204
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 15.101016998291016,
+      "learning_rate": 3.5541746015836424e-06,
+      "loss": 1.7081,
+      "step": 8205
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 21.521345138549805,
+      "learning_rate": 3.5521699909792525e-06,
+      "loss": 2.7192,
+      "step": 8206
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 27.694257736206055,
+      "learning_rate": 3.5501653803748626e-06,
+      "loss": 1.5167,
+      "step": 8207
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 10.351465225219727,
+      "learning_rate": 3.5481607697704722e-06,
+      "loss": 1.3463,
+      "step": 8208
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 29.087156295776367,
+      "learning_rate": 3.546156159166082e-06,
+      "loss": 1.397,
+      "step": 8209
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 14.024823188781738,
+      "learning_rate": 3.5441515485616924e-06,
+      "loss": 0.8097,
+      "step": 8210
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 11.973261833190918,
+      "learning_rate": 3.542146937957302e-06,
+      "loss": 0.7829,
+      "step": 8211
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 8.828713417053223,
+      "learning_rate": 3.5401423273529117e-06,
+      "loss": 0.7569,
+      "step": 8212
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 32.27656173706055,
+      "learning_rate": 3.538137716748522e-06,
+      "loss": 0.9883,
+      "step": 8213
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 54.36296463012695,
+      "learning_rate": 3.536133106144132e-06,
+      "loss": 2.0888,
+      "step": 8214
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 30.008255004882812,
+      "learning_rate": 3.5341284955397416e-06,
+      "loss": 1.1675,
+      "step": 8215
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 20.43587875366211,
+      "learning_rate": 3.5321238849353517e-06,
+      "loss": 1.0271,
+      "step": 8216
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 9.914734840393066,
+      "learning_rate": 3.5301192743309613e-06,
+      "loss": 1.4436,
+      "step": 8217
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 10.097274780273438,
+      "learning_rate": 3.5281146637265714e-06,
+      "loss": 1.2491,
+      "step": 8218
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 13.336429595947266,
+      "learning_rate": 3.5261100531221815e-06,
+      "loss": 1.1802,
+      "step": 8219
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 14.391281127929688,
+      "learning_rate": 3.524105442517791e-06,
+      "loss": 1.0317,
+      "step": 8220
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 21.83466339111328,
+      "learning_rate": 3.522100831913401e-06,
+      "loss": 2.0406,
+      "step": 8221
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 13.69311809539795,
+      "learning_rate": 3.5200962213090113e-06,
+      "loss": 1.0287,
+      "step": 8222
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 14.4343900680542,
+      "learning_rate": 3.518091610704621e-06,
+      "loss": 1.1169,
+      "step": 8223
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 19.60431671142578,
+      "learning_rate": 3.5160870001002306e-06,
+      "loss": 1.2294,
+      "step": 8224
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 8.042855262756348,
+      "learning_rate": 3.5140823894958407e-06,
+      "loss": 0.8052,
+      "step": 8225
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 13.775824546813965,
+      "learning_rate": 3.512077778891451e-06,
+      "loss": 1.0722,
+      "step": 8226
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 15.00515079498291,
+      "learning_rate": 3.5100731682870605e-06,
+      "loss": 1.0372,
+      "step": 8227
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 52.58286666870117,
+      "learning_rate": 3.5080685576826705e-06,
+      "loss": 4.7955,
+      "step": 8228
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 24.154563903808594,
+      "learning_rate": 3.50606394707828e-06,
+      "loss": 1.4256,
+      "step": 8229
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 13.502425193786621,
+      "learning_rate": 3.50405933647389e-06,
+      "loss": 1.381,
+      "step": 8230
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 9.207113265991211,
+      "learning_rate": 3.5020547258695004e-06,
+      "loss": 0.4176,
+      "step": 8231
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 21.608287811279297,
+      "learning_rate": 3.50005011526511e-06,
+      "loss": 0.7372,
+      "step": 8232
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 13.105839729309082,
+      "learning_rate": 3.4980455046607197e-06,
+      "loss": 0.719,
+      "step": 8233
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 17.119613647460938,
+      "learning_rate": 3.49604089405633e-06,
+      "loss": 1.2089,
+      "step": 8234
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 129.32228088378906,
+      "learning_rate": 3.49403628345194e-06,
+      "loss": 2.5109,
+      "step": 8235
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 11.509007453918457,
+      "learning_rate": 3.4920316728475495e-06,
+      "loss": 1.1199,
+      "step": 8236
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 17.33429718017578,
+      "learning_rate": 3.4900270622431596e-06,
+      "loss": 1.3015,
+      "step": 8237
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 28.055233001708984,
+      "learning_rate": 3.4880224516387697e-06,
+      "loss": 2.121,
+      "step": 8238
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 22.95208740234375,
+      "learning_rate": 3.4860178410343794e-06,
+      "loss": 1.7331,
+      "step": 8239
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 12.567434310913086,
+      "learning_rate": 3.4840132304299894e-06,
+      "loss": 1.5196,
+      "step": 8240
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 40.08758544921875,
+      "learning_rate": 3.482008619825599e-06,
+      "loss": 1.1594,
+      "step": 8241
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 17.78892707824707,
+      "learning_rate": 3.4800040092212088e-06,
+      "loss": 1.0566,
+      "step": 8242
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 28.924680709838867,
+      "learning_rate": 3.4779993986168193e-06,
+      "loss": 1.722,
+      "step": 8243
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 13.244800567626953,
+      "learning_rate": 3.475994788012429e-06,
+      "loss": 1.5066,
+      "step": 8244
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 46.74547576904297,
+      "learning_rate": 3.4739901774080386e-06,
+      "loss": 1.802,
+      "step": 8245
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 19.008142471313477,
+      "learning_rate": 3.4719855668036482e-06,
+      "loss": 1.4982,
+      "step": 8246
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 16.53547477722168,
+      "learning_rate": 3.4699809561992588e-06,
+      "loss": 1.4899,
+      "step": 8247
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 18.363264083862305,
+      "learning_rate": 3.4679763455948684e-06,
+      "loss": 1.022,
+      "step": 8248
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 7.702117443084717,
+      "learning_rate": 3.465971734990478e-06,
+      "loss": 1.3944,
+      "step": 8249
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 8.90109634399414,
+      "learning_rate": 3.4639671243860886e-06,
+      "loss": 0.8569,
+      "step": 8250
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 34.664485931396484,
+      "learning_rate": 3.4619625137816982e-06,
+      "loss": 1.817,
+      "step": 8251
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 12.355849266052246,
+      "learning_rate": 3.459957903177308e-06,
+      "loss": 0.9937,
+      "step": 8252
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 12.493422508239746,
+      "learning_rate": 3.457953292572918e-06,
+      "loss": 1.1074,
+      "step": 8253
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 20.661190032958984,
+      "learning_rate": 3.4559486819685277e-06,
+      "loss": 2.2912,
+      "step": 8254
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 19.867286682128906,
+      "learning_rate": 3.4539440713641377e-06,
+      "loss": 1.6991,
+      "step": 8255
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 13.034092903137207,
+      "learning_rate": 3.451939460759748e-06,
+      "loss": 0.8973,
+      "step": 8256
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 14.20175552368164,
+      "learning_rate": 3.4499348501553575e-06,
+      "loss": 1.1049,
+      "step": 8257
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 12.550244331359863,
+      "learning_rate": 3.447930239550967e-06,
+      "loss": 1.1492,
+      "step": 8258
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 13.553984642028809,
+      "learning_rate": 3.4459256289465777e-06,
+      "loss": 1.1702,
+      "step": 8259
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 17.68621826171875,
+      "learning_rate": 3.4439210183421873e-06,
+      "loss": 1.4533,
+      "step": 8260
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 19.83187484741211,
+      "learning_rate": 3.441916407737797e-06,
+      "loss": 1.4019,
+      "step": 8261
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 11.274413108825684,
+      "learning_rate": 3.4399117971334075e-06,
+      "loss": 1.7506,
+      "step": 8262
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 27.54680061340332,
+      "learning_rate": 3.437907186529017e-06,
+      "loss": 1.5858,
+      "step": 8263
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 9.336912155151367,
+      "learning_rate": 3.435902575924627e-06,
+      "loss": 0.7806,
+      "step": 8264
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 53.21189498901367,
+      "learning_rate": 3.433897965320237e-06,
+      "loss": 1.7986,
+      "step": 8265
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 9.917101860046387,
+      "learning_rate": 3.4318933547158465e-06,
+      "loss": 0.8816,
+      "step": 8266
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 8.793371200561523,
+      "learning_rate": 3.4298887441114566e-06,
+      "loss": 0.8211,
+      "step": 8267
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 6.4236626625061035,
+      "learning_rate": 3.4278841335070667e-06,
+      "loss": 0.7913,
+      "step": 8268
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 22.39311408996582,
+      "learning_rate": 3.4258795229026764e-06,
+      "loss": 1.1071,
+      "step": 8269
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 15.763972282409668,
+      "learning_rate": 3.423874912298286e-06,
+      "loss": 0.7142,
+      "step": 8270
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 16.8847713470459,
+      "learning_rate": 3.4218703016938965e-06,
+      "loss": 1.2469,
+      "step": 8271
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 9.058008193969727,
+      "learning_rate": 3.419865691089506e-06,
+      "loss": 1.3107,
+      "step": 8272
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 19.072185516357422,
+      "learning_rate": 3.417861080485116e-06,
+      "loss": 1.2581,
+      "step": 8273
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 30.158096313476562,
+      "learning_rate": 3.4158564698807264e-06,
+      "loss": 1.2928,
+      "step": 8274
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 11.686331748962402,
+      "learning_rate": 3.413851859276336e-06,
+      "loss": 1.0867,
+      "step": 8275
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 13.757401466369629,
+      "learning_rate": 3.4118472486719457e-06,
+      "loss": 1.5826,
+      "step": 8276
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 18.740379333496094,
+      "learning_rate": 3.4098426380675558e-06,
+      "loss": 1.2056,
+      "step": 8277
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 32.20988082885742,
+      "learning_rate": 3.4078380274631654e-06,
+      "loss": 1.3858,
+      "step": 8278
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 13.55139446258545,
+      "learning_rate": 3.4058334168587755e-06,
+      "loss": 1.0321,
+      "step": 8279
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 10.795275688171387,
+      "learning_rate": 3.4038288062543856e-06,
+      "loss": 1.2745,
+      "step": 8280
+    },
+    {
+      "epoch": 2.49,
+      "eval_loss": 0.17804668843746185,
+      "eval_runtime": 43.714,
+      "eval_samples_per_second": 33.834,
+      "eval_steps_per_second": 33.834,
+      "step": 8280
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 11.243696212768555,
+      "learning_rate": 3.4018241956499953e-06,
+      "loss": 0.8101,
+      "step": 8281
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 10.823609352111816,
+      "learning_rate": 3.399819585045605e-06,
+      "loss": 2.0621,
+      "step": 8282
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 52.87884521484375,
+      "learning_rate": 3.3978149744412154e-06,
+      "loss": 2.0929,
+      "step": 8283
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 11.563464164733887,
+      "learning_rate": 3.395810363836825e-06,
+      "loss": 0.8609,
+      "step": 8284
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 39.08991241455078,
+      "learning_rate": 3.3938057532324348e-06,
+      "loss": 1.3453,
+      "step": 8285
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 72.77201843261719,
+      "learning_rate": 3.391801142628045e-06,
+      "loss": 1.5782,
+      "step": 8286
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 44.493431091308594,
+      "learning_rate": 3.389796532023655e-06,
+      "loss": 1.312,
+      "step": 8287
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 49.612998962402344,
+      "learning_rate": 3.3877919214192646e-06,
+      "loss": 2.2307,
+      "step": 8288
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 30.34926986694336,
+      "learning_rate": 3.3857873108148743e-06,
+      "loss": 1.1781,
+      "step": 8289
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 9.271419525146484,
+      "learning_rate": 3.3837827002104843e-06,
+      "loss": 0.9995,
+      "step": 8290
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 16.43708610534668,
+      "learning_rate": 3.3817780896060944e-06,
+      "loss": 1.193,
+      "step": 8291
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 15.72551441192627,
+      "learning_rate": 3.379773479001704e-06,
+      "loss": 1.3098,
+      "step": 8292
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 7.720631122589111,
+      "learning_rate": 3.377768868397314e-06,
+      "loss": 0.8755,
+      "step": 8293
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 27.38102149963379,
+      "learning_rate": 3.375764257792924e-06,
+      "loss": 1.4355,
+      "step": 8294
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 12.947731018066406,
+      "learning_rate": 3.3737596471885335e-06,
+      "loss": 1.0127,
+      "step": 8295
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 18.314491271972656,
+      "learning_rate": 3.371755036584144e-06,
+      "loss": 1.1366,
+      "step": 8296
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 20.400217056274414,
+      "learning_rate": 3.3697504259797537e-06,
+      "loss": 1.2425,
+      "step": 8297
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 21.475561141967773,
+      "learning_rate": 3.3677458153753633e-06,
+      "loss": 0.9925,
+      "step": 8298
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 21.971511840820312,
+      "learning_rate": 3.365741204770974e-06,
+      "loss": 1.3472,
+      "step": 8299
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 11.167675018310547,
+      "learning_rate": 3.3637365941665835e-06,
+      "loss": 1.0643,
+      "step": 8300
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 55.813114166259766,
+      "learning_rate": 3.361731983562193e-06,
+      "loss": 2.2609,
+      "step": 8301
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 14.630870819091797,
+      "learning_rate": 3.3597273729578032e-06,
+      "loss": 0.6388,
+      "step": 8302
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 21.16858673095703,
+      "learning_rate": 3.357722762353413e-06,
+      "loss": 1.4621,
+      "step": 8303
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 25.93169593811035,
+      "learning_rate": 3.355718151749023e-06,
+      "loss": 1.6922,
+      "step": 8304
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 17.35527229309082,
+      "learning_rate": 3.353713541144633e-06,
+      "loss": 1.1722,
+      "step": 8305
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 33.93075942993164,
+      "learning_rate": 3.3517089305402427e-06,
+      "loss": 1.9865,
+      "step": 8306
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 39.736629486083984,
+      "learning_rate": 3.3497043199358524e-06,
+      "loss": 2.1434,
+      "step": 8307
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 16.60782814025879,
+      "learning_rate": 3.347699709331463e-06,
+      "loss": 1.4811,
+      "step": 8308
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 23.1959285736084,
+      "learning_rate": 3.3456950987270725e-06,
+      "loss": 1.8844,
+      "step": 8309
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 15.47692584991455,
+      "learning_rate": 3.343690488122682e-06,
+      "loss": 1.1461,
+      "step": 8310
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 10.405765533447266,
+      "learning_rate": 3.3416858775182927e-06,
+      "loss": 1.0609,
+      "step": 8311
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 7.443643569946289,
+      "learning_rate": 3.3396812669139024e-06,
+      "loss": 0.5891,
+      "step": 8312
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 40.757137298583984,
+      "learning_rate": 3.337676656309512e-06,
+      "loss": 1.1556,
+      "step": 8313
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 23.71717071533203,
+      "learning_rate": 3.335672045705122e-06,
+      "loss": 1.3438,
+      "step": 8314
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 9.829071998596191,
+      "learning_rate": 3.3336674351007318e-06,
+      "loss": 1.1778,
+      "step": 8315
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 36.342872619628906,
+      "learning_rate": 3.331662824496342e-06,
+      "loss": 1.5935,
+      "step": 8316
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 34.38883590698242,
+      "learning_rate": 3.329658213891952e-06,
+      "loss": 1.1699,
+      "step": 8317
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 13.522592544555664,
+      "learning_rate": 3.3276536032875616e-06,
+      "loss": 1.3218,
+      "step": 8318
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 18.872207641601562,
+      "learning_rate": 3.3256489926831713e-06,
+      "loss": 1.0661,
+      "step": 8319
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 48.556331634521484,
+      "learning_rate": 3.3236443820787818e-06,
+      "loss": 2.3702,
+      "step": 8320
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 26.391420364379883,
+      "learning_rate": 3.3216397714743914e-06,
+      "loss": 1.0047,
+      "step": 8321
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 10.691487312316895,
+      "learning_rate": 3.319635160870001e-06,
+      "loss": 0.6147,
+      "step": 8322
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 11.963746070861816,
+      "learning_rate": 3.3176305502656116e-06,
+      "loss": 1.4303,
+      "step": 8323
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 14.355230331420898,
+      "learning_rate": 3.3156259396612213e-06,
+      "loss": 1.6003,
+      "step": 8324
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 14.952115058898926,
+      "learning_rate": 3.313621329056831e-06,
+      "loss": 1.2936,
+      "step": 8325
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 11.36012077331543,
+      "learning_rate": 3.311616718452441e-06,
+      "loss": 1.3233,
+      "step": 8326
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 73.88568115234375,
+      "learning_rate": 3.3096121078480507e-06,
+      "loss": 2.0371,
+      "step": 8327
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 7.392289638519287,
+      "learning_rate": 3.3076074972436608e-06,
+      "loss": 0.817,
+      "step": 8328
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 26.323123931884766,
+      "learning_rate": 3.305602886639271e-06,
+      "loss": 1.2377,
+      "step": 8329
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 6.049129486083984,
+      "learning_rate": 3.3035982760348805e-06,
+      "loss": 0.725,
+      "step": 8330
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 18.688451766967773,
+      "learning_rate": 3.30159366543049e-06,
+      "loss": 1.0841,
+      "step": 8331
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 19.779504776000977,
+      "learning_rate": 3.2995890548261e-06,
+      "loss": 1.6004,
+      "step": 8332
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 37.673744201660156,
+      "learning_rate": 3.2975844442217103e-06,
+      "loss": 1.7093,
+      "step": 8333
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 15.679644584655762,
+      "learning_rate": 3.29557983361732e-06,
+      "loss": 1.4718,
+      "step": 8334
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 17.381481170654297,
+      "learning_rate": 3.2935752230129297e-06,
+      "loss": 2.5925,
+      "step": 8335
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 9.106160163879395,
+      "learning_rate": 3.29157061240854e-06,
+      "loss": 0.8652,
+      "step": 8336
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 25.941999435424805,
+      "learning_rate": 3.28956600180415e-06,
+      "loss": 0.8842,
+      "step": 8337
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 9.563976287841797,
+      "learning_rate": 3.2875613911997595e-06,
+      "loss": 1.6659,
+      "step": 8338
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 46.76715087890625,
+      "learning_rate": 3.2855567805953696e-06,
+      "loss": 1.8769,
+      "step": 8339
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 20.058713912963867,
+      "learning_rate": 3.2835521699909797e-06,
+      "loss": 1.3938,
+      "step": 8340
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 41.1780891418457,
+      "learning_rate": 3.2815475593865893e-06,
+      "loss": 1.9891,
+      "step": 8341
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 14.344354629516602,
+      "learning_rate": 3.2795429487821994e-06,
+      "loss": 1.4836,
+      "step": 8342
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 12.271919250488281,
+      "learning_rate": 3.277538338177809e-06,
+      "loss": 1.1799,
+      "step": 8343
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 10.539395332336426,
+      "learning_rate": 3.2755337275734187e-06,
+      "loss": 1.0998,
+      "step": 8344
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 18.392601013183594,
+      "learning_rate": 3.2735291169690292e-06,
+      "loss": 1.2702,
+      "step": 8345
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 9.823795318603516,
+      "learning_rate": 3.271524506364639e-06,
+      "loss": 1.4675,
+      "step": 8346
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 43.700782775878906,
+      "learning_rate": 3.2695198957602486e-06,
+      "loss": 2.5785,
+      "step": 8347
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 18.977764129638672,
+      "learning_rate": 3.267515285155859e-06,
+      "loss": 2.1341,
+      "step": 8348
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 11.805031776428223,
+      "learning_rate": 3.2655106745514687e-06,
+      "loss": 1.0303,
+      "step": 8349
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 35.22747802734375,
+      "learning_rate": 3.2635060639470784e-06,
+      "loss": 1.0836,
+      "step": 8350
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 48.45168685913086,
+      "learning_rate": 3.2615014533426885e-06,
+      "loss": 1.3489,
+      "step": 8351
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 22.869895935058594,
+      "learning_rate": 3.2594968427382986e-06,
+      "loss": 2.6811,
+      "step": 8352
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 18.353343963623047,
+      "learning_rate": 3.257492232133908e-06,
+      "loss": 0.976,
+      "step": 8353
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 29.93402862548828,
+      "learning_rate": 3.2554876215295183e-06,
+      "loss": 1.1663,
+      "step": 8354
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 13.383641242980957,
+      "learning_rate": 3.253483010925128e-06,
+      "loss": 0.7903,
+      "step": 8355
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 6.8472747802734375,
+      "learning_rate": 3.2514784003207376e-06,
+      "loss": 1.1989,
+      "step": 8356
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 32.73374938964844,
+      "learning_rate": 3.249473789716348e-06,
+      "loss": 0.9284,
+      "step": 8357
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 11.246085166931152,
+      "learning_rate": 3.2474691791119578e-06,
+      "loss": 1.5323,
+      "step": 8358
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 8.259552955627441,
+      "learning_rate": 3.2454645685075674e-06,
+      "loss": 0.8713,
+      "step": 8359
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 28.066495895385742,
+      "learning_rate": 3.243459957903178e-06,
+      "loss": 1.1289,
+      "step": 8360
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 16.770734786987305,
+      "learning_rate": 3.2414553472987876e-06,
+      "loss": 1.1323,
+      "step": 8361
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 10.900749206542969,
+      "learning_rate": 3.2394507366943973e-06,
+      "loss": 0.7672,
+      "step": 8362
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 21.28752326965332,
+      "learning_rate": 3.2374461260900074e-06,
+      "loss": 1.4755,
+      "step": 8363
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 15.661283493041992,
+      "learning_rate": 3.2354415154856174e-06,
+      "loss": 1.3512,
+      "step": 8364
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 16.886972427368164,
+      "learning_rate": 3.233436904881227e-06,
+      "loss": 1.4028,
+      "step": 8365
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 17.953754425048828,
+      "learning_rate": 3.231432294276837e-06,
+      "loss": 1.2938,
+      "step": 8366
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 9.600143432617188,
+      "learning_rate": 3.229427683672447e-06,
+      "loss": 0.7052,
+      "step": 8367
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 19.288686752319336,
+      "learning_rate": 3.2274230730680565e-06,
+      "loss": 1.588,
+      "step": 8368
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 12.002143859863281,
+      "learning_rate": 3.225418462463667e-06,
+      "loss": 0.5881,
+      "step": 8369
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 18.510684967041016,
+      "learning_rate": 3.2234138518592767e-06,
+      "loss": 1.543,
+      "step": 8370
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 31.459373474121094,
+      "learning_rate": 3.2214092412548863e-06,
+      "loss": 1.0949,
+      "step": 8371
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 7.364177703857422,
+      "learning_rate": 3.219404630650496e-06,
+      "loss": 0.7282,
+      "step": 8372
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 10.49648666381836,
+      "learning_rate": 3.2174000200461065e-06,
+      "loss": 0.6402,
+      "step": 8373
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 25.187376022338867,
+      "learning_rate": 3.215395409441716e-06,
+      "loss": 1.4291,
+      "step": 8374
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 26.098617553710938,
+      "learning_rate": 3.213390798837326e-06,
+      "loss": 1.861,
+      "step": 8375
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 46.278839111328125,
+      "learning_rate": 3.2113861882329363e-06,
+      "loss": 1.8303,
+      "step": 8376
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 12.867042541503906,
+      "learning_rate": 3.209381577628546e-06,
+      "loss": 1.0696,
+      "step": 8377
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 20.583019256591797,
+      "learning_rate": 3.2073769670241557e-06,
+      "loss": 0.992,
+      "step": 8378
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 56.57286071777344,
+      "learning_rate": 3.2053723564197657e-06,
+      "loss": 1.8542,
+      "step": 8379
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 13.195068359375,
+      "learning_rate": 3.2033677458153754e-06,
+      "loss": 1.3004,
+      "step": 8380
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 16.263944625854492,
+      "learning_rate": 3.2013631352109855e-06,
+      "loss": 1.7439,
+      "step": 8381
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 11.452669143676758,
+      "learning_rate": 3.1993585246065956e-06,
+      "loss": 0.7889,
+      "step": 8382
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 17.5844669342041,
+      "learning_rate": 3.1973539140022052e-06,
+      "loss": 1.2866,
+      "step": 8383
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 12.583413124084473,
+      "learning_rate": 3.195349303397815e-06,
+      "loss": 0.9804,
+      "step": 8384
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 13.724156379699707,
+      "learning_rate": 3.1933446927934254e-06,
+      "loss": 0.7936,
+      "step": 8385
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 11.498400688171387,
+      "learning_rate": 3.191340082189035e-06,
+      "loss": 1.1136,
+      "step": 8386
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 41.235206604003906,
+      "learning_rate": 3.1893354715846447e-06,
+      "loss": 1.1562,
+      "step": 8387
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 26.88475799560547,
+      "learning_rate": 3.187330860980255e-06,
+      "loss": 2.1271,
+      "step": 8388
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 16.73887825012207,
+      "learning_rate": 3.185326250375865e-06,
+      "loss": 1.5196,
+      "step": 8389
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 20.1945858001709,
+      "learning_rate": 3.1833216397714746e-06,
+      "loss": 0.9805,
+      "step": 8390
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 34.976627349853516,
+      "learning_rate": 3.1813170291670846e-06,
+      "loss": 1.1922,
+      "step": 8391
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 36.179351806640625,
+      "learning_rate": 3.1793124185626943e-06,
+      "loss": 4.2282,
+      "step": 8392
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 9.375896453857422,
+      "learning_rate": 3.1773078079583044e-06,
+      "loss": 0.9379,
+      "step": 8393
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 132.8348846435547,
+      "learning_rate": 3.1753031973539145e-06,
+      "loss": 3.0371,
+      "step": 8394
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 30.87514877319336,
+      "learning_rate": 3.173298586749524e-06,
+      "loss": 2.5639,
+      "step": 8395
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 12.229470252990723,
+      "learning_rate": 3.171293976145134e-06,
+      "loss": 0.9845,
+      "step": 8396
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 26.513559341430664,
+      "learning_rate": 3.1692893655407443e-06,
+      "loss": 1.1088,
+      "step": 8397
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 7.9161553382873535,
+      "learning_rate": 3.167284754936354e-06,
+      "loss": 0.8013,
+      "step": 8398
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 12.535621643066406,
+      "learning_rate": 3.1652801443319636e-06,
+      "loss": 1.0497,
+      "step": 8399
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 9.409601211547852,
+      "learning_rate": 3.1632755337275737e-06,
+      "loss": 1.132,
+      "step": 8400
+    },
+    {
+      "epoch": 2.53,
+      "eval_loss": 0.17680850625038147,
+      "eval_runtime": 44.0695,
+      "eval_samples_per_second": 33.561,
+      "eval_steps_per_second": 33.561,
+      "step": 8400
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 38.185997009277344,
+      "learning_rate": 3.1612709231231838e-06,
+      "loss": 1.8298,
+      "step": 8401
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 103.94436645507812,
+      "learning_rate": 3.1592663125187934e-06,
+      "loss": 0.8306,
+      "step": 8402
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 105.07266998291016,
+      "learning_rate": 3.1572617019144035e-06,
+      "loss": 1.827,
+      "step": 8403
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 19.526813507080078,
+      "learning_rate": 3.155257091310013e-06,
+      "loss": 1.4366,
+      "step": 8404
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 10.135507583618164,
+      "learning_rate": 3.153252480705623e-06,
+      "loss": 0.6992,
+      "step": 8405
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 35.69138717651367,
+      "learning_rate": 3.1512478701012334e-06,
+      "loss": 1.5329,
+      "step": 8406
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 74.143798828125,
+      "learning_rate": 3.149243259496843e-06,
+      "loss": 0.9668,
+      "step": 8407
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 15.071893692016602,
+      "learning_rate": 3.1472386488924527e-06,
+      "loss": 1.1024,
+      "step": 8408
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 11.287890434265137,
+      "learning_rate": 3.145234038288063e-06,
+      "loss": 1.2887,
+      "step": 8409
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 63.98174285888672,
+      "learning_rate": 3.143229427683673e-06,
+      "loss": 1.8663,
+      "step": 8410
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 15.564671516418457,
+      "learning_rate": 3.1412248170792825e-06,
+      "loss": 1.1972,
+      "step": 8411
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 26.284500122070312,
+      "learning_rate": 3.1392202064748926e-06,
+      "loss": 1.1945,
+      "step": 8412
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 81.35211181640625,
+      "learning_rate": 3.1372155958705027e-06,
+      "loss": 1.7933,
+      "step": 8413
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 20.28238296508789,
+      "learning_rate": 3.1352109852661123e-06,
+      "loss": 1.6545,
+      "step": 8414
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 11.033011436462402,
+      "learning_rate": 3.133206374661722e-06,
+      "loss": 0.8211,
+      "step": 8415
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 17.557846069335938,
+      "learning_rate": 3.131201764057332e-06,
+      "loss": 1.2009,
+      "step": 8416
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 17.877410888671875,
+      "learning_rate": 3.1291971534529417e-06,
+      "loss": 1.1617,
+      "step": 8417
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 17.008520126342773,
+      "learning_rate": 3.127192542848552e-06,
+      "loss": 1.2213,
+      "step": 8418
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 13.909770965576172,
+      "learning_rate": 3.125187932244162e-06,
+      "loss": 0.9544,
+      "step": 8419
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 12.314075469970703,
+      "learning_rate": 3.1231833216397716e-06,
+      "loss": 1.1267,
+      "step": 8420
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 8.14324951171875,
+      "learning_rate": 3.1211787110353812e-06,
+      "loss": 1.052,
+      "step": 8421
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 9.060734748840332,
+      "learning_rate": 3.1191741004309917e-06,
+      "loss": 0.7107,
+      "step": 8422
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 12.2525634765625,
+      "learning_rate": 3.1171694898266014e-06,
+      "loss": 0.9515,
+      "step": 8423
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 18.487070083618164,
+      "learning_rate": 3.115164879222211e-06,
+      "loss": 0.9164,
+      "step": 8424
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 21.11574363708496,
+      "learning_rate": 3.1131602686178216e-06,
+      "loss": 1.1415,
+      "step": 8425
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 10.710432052612305,
+      "learning_rate": 3.1111556580134312e-06,
+      "loss": 0.6223,
+      "step": 8426
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 12.434354782104492,
+      "learning_rate": 3.109151047409041e-06,
+      "loss": 0.7547,
+      "step": 8427
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 5.420670032501221,
+      "learning_rate": 3.107146436804651e-06,
+      "loss": 0.913,
+      "step": 8428
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 23.501205444335938,
+      "learning_rate": 3.1051418262002606e-06,
+      "loss": 1.7807,
+      "step": 8429
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 17.01691436767578,
+      "learning_rate": 3.1031372155958707e-06,
+      "loss": 1.2807,
+      "step": 8430
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 15.571081161499023,
+      "learning_rate": 3.101132604991481e-06,
+      "loss": 1.397,
+      "step": 8431
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 32.079612731933594,
+      "learning_rate": 3.0991279943870905e-06,
+      "loss": 1.4421,
+      "step": 8432
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 13.537861824035645,
+      "learning_rate": 3.0971233837827e-06,
+      "loss": 1.4324,
+      "step": 8433
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 72.99842834472656,
+      "learning_rate": 3.0951187731783106e-06,
+      "loss": 2.2397,
+      "step": 8434
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 32.24034118652344,
+      "learning_rate": 3.0931141625739203e-06,
+      "loss": 1.8425,
+      "step": 8435
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 14.41550350189209,
+      "learning_rate": 3.09110955196953e-06,
+      "loss": 1.0483,
+      "step": 8436
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 47.05146789550781,
+      "learning_rate": 3.0891049413651405e-06,
+      "loss": 1.3004,
+      "step": 8437
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 25.591800689697266,
+      "learning_rate": 3.08710033076075e-06,
+      "loss": 1.1771,
+      "step": 8438
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 9.942156791687012,
+      "learning_rate": 3.08509572015636e-06,
+      "loss": 0.5432,
+      "step": 8439
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 22.738941192626953,
+      "learning_rate": 3.08309110955197e-06,
+      "loss": 2.3535,
+      "step": 8440
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 14.746498107910156,
+      "learning_rate": 3.0810864989475795e-06,
+      "loss": 1.2366,
+      "step": 8441
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 98.85680389404297,
+      "learning_rate": 3.0790818883431896e-06,
+      "loss": 2.4007,
+      "step": 8442
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 23.468538284301758,
+      "learning_rate": 3.0770772777387997e-06,
+      "loss": 1.0986,
+      "step": 8443
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 8.700865745544434,
+      "learning_rate": 3.0750726671344094e-06,
+      "loss": 1.1621,
+      "step": 8444
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 22.677888870239258,
+      "learning_rate": 3.073068056530019e-06,
+      "loss": 1.2522,
+      "step": 8445
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 21.384296417236328,
+      "learning_rate": 3.0710634459256295e-06,
+      "loss": 1.6951,
+      "step": 8446
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 30.282161712646484,
+      "learning_rate": 3.069058835321239e-06,
+      "loss": 1.8486,
+      "step": 8447
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 11.835555076599121,
+      "learning_rate": 3.067054224716849e-06,
+      "loss": 1.0966,
+      "step": 8448
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 16.40022850036621,
+      "learning_rate": 3.0650496141124594e-06,
+      "loss": 1.55,
+      "step": 8449
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 34.12657928466797,
+      "learning_rate": 3.063045003508069e-06,
+      "loss": 1.6477,
+      "step": 8450
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 43.224571228027344,
+      "learning_rate": 3.0610403929036787e-06,
+      "loss": 1.4323,
+      "step": 8451
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 21.87476921081543,
+      "learning_rate": 3.0590357822992888e-06,
+      "loss": 1.117,
+      "step": 8452
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 12.986775398254395,
+      "learning_rate": 3.0570311716948984e-06,
+      "loss": 0.865,
+      "step": 8453
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 12.08977222442627,
+      "learning_rate": 3.0550265610905085e-06,
+      "loss": 0.7838,
+      "step": 8454
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 50.07052230834961,
+      "learning_rate": 3.053021950486118e-06,
+      "loss": 1.4847,
+      "step": 8455
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 7.173743724822998,
+      "learning_rate": 3.0510173398817283e-06,
+      "loss": 0.6167,
+      "step": 8456
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 38.1131591796875,
+      "learning_rate": 3.049012729277338e-06,
+      "loss": 2.517,
+      "step": 8457
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 9.692411422729492,
+      "learning_rate": 3.0470081186729476e-06,
+      "loss": 1.0573,
+      "step": 8458
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 41.6327018737793,
+      "learning_rate": 3.045003508068558e-06,
+      "loss": 1.4349,
+      "step": 8459
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 29.187610626220703,
+      "learning_rate": 3.0429988974641677e-06,
+      "loss": 1.5777,
+      "step": 8460
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 29.35192108154297,
+      "learning_rate": 3.0409942868597774e-06,
+      "loss": 1.8016,
+      "step": 8461
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 20.445228576660156,
+      "learning_rate": 3.038989676255388e-06,
+      "loss": 0.7974,
+      "step": 8462
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 18.133628845214844,
+      "learning_rate": 3.0369850656509976e-06,
+      "loss": 1.1352,
+      "step": 8463
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 16.889158248901367,
+      "learning_rate": 3.0349804550466072e-06,
+      "loss": 1.4672,
+      "step": 8464
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 20.985919952392578,
+      "learning_rate": 3.0329758444422173e-06,
+      "loss": 1.192,
+      "step": 8465
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 9.734439849853516,
+      "learning_rate": 3.0309712338378274e-06,
+      "loss": 0.7422,
+      "step": 8466
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 23.414813995361328,
+      "learning_rate": 3.028966623233437e-06,
+      "loss": 1.8114,
+      "step": 8467
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 12.681184768676758,
+      "learning_rate": 3.026962012629047e-06,
+      "loss": 1.1328,
+      "step": 8468
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 23.120763778686523,
+      "learning_rate": 3.024957402024657e-06,
+      "loss": 1.1875,
+      "step": 8469
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 19.008502960205078,
+      "learning_rate": 3.0229527914202665e-06,
+      "loss": 1.2166,
+      "step": 8470
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 22.20121192932129,
+      "learning_rate": 3.020948180815877e-06,
+      "loss": 1.4729,
+      "step": 8471
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 16.53184700012207,
+      "learning_rate": 3.0189435702114866e-06,
+      "loss": 1.8034,
+      "step": 8472
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 21.981178283691406,
+      "learning_rate": 3.0169389596070963e-06,
+      "loss": 2.2006,
+      "step": 8473
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 36.195106506347656,
+      "learning_rate": 3.014934349002707e-06,
+      "loss": 2.1031,
+      "step": 8474
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 33.1042366027832,
+      "learning_rate": 3.0129297383983165e-06,
+      "loss": 1.5522,
+      "step": 8475
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 14.721949577331543,
+      "learning_rate": 3.010925127793926e-06,
+      "loss": 1.5655,
+      "step": 8476
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 14.444605827331543,
+      "learning_rate": 3.0089205171895362e-06,
+      "loss": 1.2094,
+      "step": 8477
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 14.284456253051758,
+      "learning_rate": 3.0069159065851463e-06,
+      "loss": 0.986,
+      "step": 8478
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 12.816495895385742,
+      "learning_rate": 3.004911295980756e-06,
+      "loss": 1.0418,
+      "step": 8479
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 67.11431884765625,
+      "learning_rate": 3.002906685376366e-06,
+      "loss": 1.5002,
+      "step": 8480
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 24.34355926513672,
+      "learning_rate": 3.0009020747719757e-06,
+      "loss": 1.329,
+      "step": 8481
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 11.59719467163086,
+      "learning_rate": 2.9988974641675854e-06,
+      "loss": 1.4363,
+      "step": 8482
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 32.02231216430664,
+      "learning_rate": 2.996892853563196e-06,
+      "loss": 1.2263,
+      "step": 8483
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 9.977734565734863,
+      "learning_rate": 2.9948882429588055e-06,
+      "loss": 0.6075,
+      "step": 8484
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 48.880313873291016,
+      "learning_rate": 2.992883632354415e-06,
+      "loss": 1.5793,
+      "step": 8485
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 7.916026592254639,
+      "learning_rate": 2.9908790217500257e-06,
+      "loss": 0.7116,
+      "step": 8486
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 29.04728889465332,
+      "learning_rate": 2.9888744111456354e-06,
+      "loss": 1.2325,
+      "step": 8487
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 13.980238914489746,
+      "learning_rate": 2.986869800541245e-06,
+      "loss": 1.496,
+      "step": 8488
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 15.269109725952148,
+      "learning_rate": 2.984865189936855e-06,
+      "loss": 1.4313,
+      "step": 8489
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 20.857440948486328,
+      "learning_rate": 2.9828605793324648e-06,
+      "loss": 1.5363,
+      "step": 8490
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 14.627253532409668,
+      "learning_rate": 2.980855968728075e-06,
+      "loss": 1.3487,
+      "step": 8491
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 7.229565620422363,
+      "learning_rate": 2.978851358123685e-06,
+      "loss": 0.8008,
+      "step": 8492
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 29.895063400268555,
+      "learning_rate": 2.9768467475192946e-06,
+      "loss": 1.7014,
+      "step": 8493
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 26.894834518432617,
+      "learning_rate": 2.9748421369149043e-06,
+      "loss": 1.8372,
+      "step": 8494
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 50.57379150390625,
+      "learning_rate": 2.9728375263105148e-06,
+      "loss": 1.3365,
+      "step": 8495
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 12.997958183288574,
+      "learning_rate": 2.9708329157061244e-06,
+      "loss": 1.9539,
+      "step": 8496
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 11.23823356628418,
+      "learning_rate": 2.968828305101734e-06,
+      "loss": 1.2944,
+      "step": 8497
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 21.072771072387695,
+      "learning_rate": 2.9668236944973438e-06,
+      "loss": 1.4985,
+      "step": 8498
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 61.82595443725586,
+      "learning_rate": 2.9648190838929543e-06,
+      "loss": 2.2333,
+      "step": 8499
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 26.898548126220703,
+      "learning_rate": 2.962814473288564e-06,
+      "loss": 1.2997,
+      "step": 8500
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 14.761646270751953,
+      "learning_rate": 2.9608098626841736e-06,
+      "loss": 1.4087,
+      "step": 8501
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 26.489471435546875,
+      "learning_rate": 2.9588052520797837e-06,
+      "loss": 1.6874,
+      "step": 8502
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 25.354101181030273,
+      "learning_rate": 2.9568006414753938e-06,
+      "loss": 0.7809,
+      "step": 8503
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 8.844145774841309,
+      "learning_rate": 2.9547960308710034e-06,
+      "loss": 0.9255,
+      "step": 8504
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 18.32469940185547,
+      "learning_rate": 2.9527914202666135e-06,
+      "loss": 1.0907,
+      "step": 8505
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 14.326888084411621,
+      "learning_rate": 2.950786809662223e-06,
+      "loss": 1.1902,
+      "step": 8506
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 19.863540649414062,
+      "learning_rate": 2.948782199057833e-06,
+      "loss": 1.5195,
+      "step": 8507
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 9.05129623413086,
+      "learning_rate": 2.9467775884534433e-06,
+      "loss": 0.7649,
+      "step": 8508
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 11.893647193908691,
+      "learning_rate": 2.944772977849053e-06,
+      "loss": 1.4486,
+      "step": 8509
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 18.236352920532227,
+      "learning_rate": 2.9427683672446626e-06,
+      "loss": 1.0206,
+      "step": 8510
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 32.056034088134766,
+      "learning_rate": 2.940763756640273e-06,
+      "loss": 2.3092,
+      "step": 8511
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 72.33465576171875,
+      "learning_rate": 2.938759146035883e-06,
+      "loss": 2.1425,
+      "step": 8512
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 56.66347122192383,
+      "learning_rate": 2.9367545354314925e-06,
+      "loss": 1.1673,
+      "step": 8513
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 14.660852432250977,
+      "learning_rate": 2.9347499248271026e-06,
+      "loss": 1.1077,
+      "step": 8514
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 45.072975158691406,
+      "learning_rate": 2.9327453142227126e-06,
+      "loss": 1.7196,
+      "step": 8515
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 15.052632331848145,
+      "learning_rate": 2.9307407036183223e-06,
+      "loss": 1.245,
+      "step": 8516
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 17.311857223510742,
+      "learning_rate": 2.9287360930139324e-06,
+      "loss": 1.241,
+      "step": 8517
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 71.18388366699219,
+      "learning_rate": 2.926731482409542e-06,
+      "loss": 1.4936,
+      "step": 8518
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 14.947158813476562,
+      "learning_rate": 2.9247268718051517e-06,
+      "loss": 1.164,
+      "step": 8519
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 39.9619140625,
+      "learning_rate": 2.9227222612007622e-06,
+      "loss": 2.3687,
+      "step": 8520
+    },
+    {
+      "epoch": 2.56,
+      "eval_loss": 0.17769810557365417,
+      "eval_runtime": 44.5571,
+      "eval_samples_per_second": 33.193,
+      "eval_steps_per_second": 33.193,
+      "step": 8520
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 17.896486282348633,
+      "learning_rate": 2.920717650596372e-06,
+      "loss": 1.4077,
+      "step": 8521
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 32.224422454833984,
+      "learning_rate": 2.9187130399919815e-06,
+      "loss": 1.8564,
+      "step": 8522
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 20.0201358795166,
+      "learning_rate": 2.916708429387592e-06,
+      "loss": 0.6378,
+      "step": 8523
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 8.714916229248047,
+      "learning_rate": 2.9147038187832017e-06,
+      "loss": 0.8691,
+      "step": 8524
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 13.311516761779785,
+      "learning_rate": 2.9126992081788114e-06,
+      "loss": 1.0312,
+      "step": 8525
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 23.491960525512695,
+      "learning_rate": 2.9106945975744215e-06,
+      "loss": 1.2836,
+      "step": 8526
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 9.72035026550293,
+      "learning_rate": 2.9086899869700315e-06,
+      "loss": 1.1779,
+      "step": 8527
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 14.758877754211426,
+      "learning_rate": 2.906685376365641e-06,
+      "loss": 1.3364,
+      "step": 8528
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 9.038101196289062,
+      "learning_rate": 2.9046807657612513e-06,
+      "loss": 0.7359,
+      "step": 8529
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 9.687514305114746,
+      "learning_rate": 2.902676155156861e-06,
+      "loss": 0.6275,
+      "step": 8530
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 21.076562881469727,
+      "learning_rate": 2.9006715445524706e-06,
+      "loss": 1.3599,
+      "step": 8531
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 8.555130004882812,
+      "learning_rate": 2.898666933948081e-06,
+      "loss": 0.9138,
+      "step": 8532
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 30.38118553161621,
+      "learning_rate": 2.8966623233436908e-06,
+      "loss": 1.4925,
+      "step": 8533
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 34.965641021728516,
+      "learning_rate": 2.8946577127393004e-06,
+      "loss": 1.3331,
+      "step": 8534
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 10.746747016906738,
+      "learning_rate": 2.892653102134911e-06,
+      "loss": 1.2565,
+      "step": 8535
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 62.186431884765625,
+      "learning_rate": 2.8906484915305206e-06,
+      "loss": 2.5688,
+      "step": 8536
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 9.051580429077148,
+      "learning_rate": 2.8886438809261303e-06,
+      "loss": 0.7643,
+      "step": 8537
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 22.37142562866211,
+      "learning_rate": 2.88663927032174e-06,
+      "loss": 1.8968,
+      "step": 8538
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 15.524712562561035,
+      "learning_rate": 2.8846346597173504e-06,
+      "loss": 0.9424,
+      "step": 8539
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 14.384515762329102,
+      "learning_rate": 2.88263004911296e-06,
+      "loss": 1.0429,
+      "step": 8540
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 55.4445686340332,
+      "learning_rate": 2.8806254385085698e-06,
+      "loss": 2.7088,
+      "step": 8541
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 11.720520973205566,
+      "learning_rate": 2.87862082790418e-06,
+      "loss": 0.7067,
+      "step": 8542
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 18.14213752746582,
+      "learning_rate": 2.8766162172997895e-06,
+      "loss": 1.0006,
+      "step": 8543
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 9.360933303833008,
+      "learning_rate": 2.8746116066953996e-06,
+      "loss": 1.1814,
+      "step": 8544
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 54.820091247558594,
+      "learning_rate": 2.8726069960910097e-06,
+      "loss": 1.0313,
+      "step": 8545
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 17.7171688079834,
+      "learning_rate": 2.8706023854866193e-06,
+      "loss": 0.5955,
+      "step": 8546
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 26.265403747558594,
+      "learning_rate": 2.868597774882229e-06,
+      "loss": 1.2873,
+      "step": 8547
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 51.058250427246094,
+      "learning_rate": 2.8665931642778395e-06,
+      "loss": 1.3523,
+      "step": 8548
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 14.056093215942383,
+      "learning_rate": 2.864588553673449e-06,
+      "loss": 0.9431,
+      "step": 8549
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 55.49951171875,
+      "learning_rate": 2.862583943069059e-06,
+      "loss": 1.6338,
+      "step": 8550
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 12.471390724182129,
+      "learning_rate": 2.8605793324646693e-06,
+      "loss": 1.1413,
+      "step": 8551
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 19.35071563720703,
+      "learning_rate": 2.858574721860279e-06,
+      "loss": 0.9613,
+      "step": 8552
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 16.208166122436523,
+      "learning_rate": 2.8565701112558886e-06,
+      "loss": 1.5786,
+      "step": 8553
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 24.832500457763672,
+      "learning_rate": 2.8545655006514987e-06,
+      "loss": 1.6098,
+      "step": 8554
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 51.225799560546875,
+      "learning_rate": 2.8525608900471084e-06,
+      "loss": 1.4513,
+      "step": 8555
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 13.274884223937988,
+      "learning_rate": 2.8505562794427185e-06,
+      "loss": 0.8535,
+      "step": 8556
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 31.125516891479492,
+      "learning_rate": 2.8485516688383286e-06,
+      "loss": 1.388,
+      "step": 8557
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 17.432456970214844,
+      "learning_rate": 2.8465470582339382e-06,
+      "loss": 1.5913,
+      "step": 8558
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 24.500856399536133,
+      "learning_rate": 2.844542447629548e-06,
+      "loss": 2.2924,
+      "step": 8559
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 14.529585838317871,
+      "learning_rate": 2.8425378370251584e-06,
+      "loss": 0.5999,
+      "step": 8560
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 12.131428718566895,
+      "learning_rate": 2.840533226420768e-06,
+      "loss": 1.439,
+      "step": 8561
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 8.231968879699707,
+      "learning_rate": 2.8385286158163777e-06,
+      "loss": 0.9683,
+      "step": 8562
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 7.67575740814209,
+      "learning_rate": 2.836524005211988e-06,
+      "loss": 0.6245,
+      "step": 8563
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 12.068075180053711,
+      "learning_rate": 2.834519394607598e-06,
+      "loss": 0.7527,
+      "step": 8564
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 14.110102653503418,
+      "learning_rate": 2.8325147840032075e-06,
+      "loss": 1.7017,
+      "step": 8565
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 38.348731994628906,
+      "learning_rate": 2.8305101733988176e-06,
+      "loss": 2.0444,
+      "step": 8566
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 9.310153007507324,
+      "learning_rate": 2.8285055627944273e-06,
+      "loss": 1.0862,
+      "step": 8567
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 18.77411460876465,
+      "learning_rate": 2.8265009521900374e-06,
+      "loss": 1.0017,
+      "step": 8568
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 9.130143165588379,
+      "learning_rate": 2.8244963415856475e-06,
+      "loss": 1.1588,
+      "step": 8569
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 37.241661071777344,
+      "learning_rate": 2.822491730981257e-06,
+      "loss": 1.1469,
+      "step": 8570
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 19.131694793701172,
+      "learning_rate": 2.8204871203768668e-06,
+      "loss": 1.8664,
+      "step": 8571
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 16.91837501525879,
+      "learning_rate": 2.8184825097724773e-06,
+      "loss": 1.2044,
+      "step": 8572
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 12.339491844177246,
+      "learning_rate": 2.816477899168087e-06,
+      "loss": 0.8794,
+      "step": 8573
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 41.453243255615234,
+      "learning_rate": 2.8144732885636966e-06,
+      "loss": 1.4408,
+      "step": 8574
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 36.32587814331055,
+      "learning_rate": 2.8124686779593067e-06,
+      "loss": 0.8383,
+      "step": 8575
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 19.15566635131836,
+      "learning_rate": 2.8104640673549168e-06,
+      "loss": 1.0883,
+      "step": 8576
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 18.432010650634766,
+      "learning_rate": 2.8084594567505264e-06,
+      "loss": 1.314,
+      "step": 8577
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 51.9474983215332,
+      "learning_rate": 2.8064548461461365e-06,
+      "loss": 1.7958,
+      "step": 8578
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 21.26700210571289,
+      "learning_rate": 2.804450235541746e-06,
+      "loss": 1.878,
+      "step": 8579
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 23.973962783813477,
+      "learning_rate": 2.8024456249373563e-06,
+      "loss": 1.2876,
+      "step": 8580
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 11.906970024108887,
+      "learning_rate": 2.800441014332966e-06,
+      "loss": 0.5872,
+      "step": 8581
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 14.538961410522461,
+      "learning_rate": 2.798436403728576e-06,
+      "loss": 0.9155,
+      "step": 8582
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 19.3804988861084,
+      "learning_rate": 2.7964317931241857e-06,
+      "loss": 1.2437,
+      "step": 8583
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 21.384546279907227,
+      "learning_rate": 2.7944271825197953e-06,
+      "loss": 0.8738,
+      "step": 8584
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 5.975277423858643,
+      "learning_rate": 2.792422571915406e-06,
+      "loss": 0.4514,
+      "step": 8585
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 13.506431579589844,
+      "learning_rate": 2.7904179613110155e-06,
+      "loss": 1.3215,
+      "step": 8586
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 16.181753158569336,
+      "learning_rate": 2.788413350706625e-06,
+      "loss": 1.9854,
+      "step": 8587
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 23.374975204467773,
+      "learning_rate": 2.7864087401022357e-06,
+      "loss": 1.9213,
+      "step": 8588
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 29.04400634765625,
+      "learning_rate": 2.7844041294978453e-06,
+      "loss": 1.8084,
+      "step": 8589
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 9.371678352355957,
+      "learning_rate": 2.782399518893455e-06,
+      "loss": 1.0408,
+      "step": 8590
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 11.659550666809082,
+      "learning_rate": 2.780394908289065e-06,
+      "loss": 1.3009,
+      "step": 8591
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 27.285043716430664,
+      "learning_rate": 2.7783902976846747e-06,
+      "loss": 1.9198,
+      "step": 8592
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 30.421232223510742,
+      "learning_rate": 2.776385687080285e-06,
+      "loss": 1.3893,
+      "step": 8593
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 14.901866912841797,
+      "learning_rate": 2.774381076475895e-06,
+      "loss": 1.0787,
+      "step": 8594
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 10.58189868927002,
+      "learning_rate": 2.7723764658715046e-06,
+      "loss": 1.3617,
+      "step": 8595
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 83.2371826171875,
+      "learning_rate": 2.7703718552671142e-06,
+      "loss": 1.7557,
+      "step": 8596
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 23.80765724182129,
+      "learning_rate": 2.7683672446627247e-06,
+      "loss": 2.3199,
+      "step": 8597
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 15.422316551208496,
+      "learning_rate": 2.7663626340583344e-06,
+      "loss": 1.1275,
+      "step": 8598
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 20.54278564453125,
+      "learning_rate": 2.764358023453944e-06,
+      "loss": 1.7601,
+      "step": 8599
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 14.661285400390625,
+      "learning_rate": 2.7623534128495546e-06,
+      "loss": 1.0741,
+      "step": 8600
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 33.70438766479492,
+      "learning_rate": 2.7603488022451642e-06,
+      "loss": 1.0387,
+      "step": 8601
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 15.016576766967773,
+      "learning_rate": 2.758344191640774e-06,
+      "loss": 1.7023,
+      "step": 8602
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 7.491636753082275,
+      "learning_rate": 2.756339581036384e-06,
+      "loss": 1.374,
+      "step": 8603
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 13.837325096130371,
+      "learning_rate": 2.7543349704319936e-06,
+      "loss": 0.8613,
+      "step": 8604
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 17.60556983947754,
+      "learning_rate": 2.7523303598276037e-06,
+      "loss": 1.1599,
+      "step": 8605
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 13.817652702331543,
+      "learning_rate": 2.750325749223214e-06,
+      "loss": 1.1847,
+      "step": 8606
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 39.440921783447266,
+      "learning_rate": 2.7483211386188235e-06,
+      "loss": 1.5714,
+      "step": 8607
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 13.094484329223633,
+      "learning_rate": 2.746316528014433e-06,
+      "loss": 1.4254,
+      "step": 8608
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 12.363558769226074,
+      "learning_rate": 2.7443119174100436e-06,
+      "loss": 1.4503,
+      "step": 8609
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 81.3821792602539,
+      "learning_rate": 2.7423073068056533e-06,
+      "loss": 1.6868,
+      "step": 8610
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 28.759248733520508,
+      "learning_rate": 2.740302696201263e-06,
+      "loss": 1.2274,
+      "step": 8611
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 22.243206024169922,
+      "learning_rate": 2.7382980855968735e-06,
+      "loss": 1.5054,
+      "step": 8612
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 16.99969482421875,
+      "learning_rate": 2.736293474992483e-06,
+      "loss": 1.3755,
+      "step": 8613
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 8.332833290100098,
+      "learning_rate": 2.7342888643880928e-06,
+      "loss": 1.0886,
+      "step": 8614
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 17.875459671020508,
+      "learning_rate": 2.732284253783703e-06,
+      "loss": 1.2963,
+      "step": 8615
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 17.195663452148438,
+      "learning_rate": 2.7302796431793125e-06,
+      "loss": 0.9124,
+      "step": 8616
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 45.631954193115234,
+      "learning_rate": 2.7282750325749226e-06,
+      "loss": 2.5502,
+      "step": 8617
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 72.2505874633789,
+      "learning_rate": 2.7262704219705327e-06,
+      "loss": 1.9153,
+      "step": 8618
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 19.540000915527344,
+      "learning_rate": 2.7242658113661424e-06,
+      "loss": 1.3127,
+      "step": 8619
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 16.321380615234375,
+      "learning_rate": 2.722261200761752e-06,
+      "loss": 1.1336,
+      "step": 8620
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 35.40462875366211,
+      "learning_rate": 2.7202565901573625e-06,
+      "loss": 1.4144,
+      "step": 8621
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 18.681198120117188,
+      "learning_rate": 2.718251979552972e-06,
+      "loss": 0.8571,
+      "step": 8622
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 37.2798957824707,
+      "learning_rate": 2.716247368948582e-06,
+      "loss": 1.1746,
+      "step": 8623
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 17.78194808959961,
+      "learning_rate": 2.7142427583441915e-06,
+      "loss": 1.0923,
+      "step": 8624
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 18.684160232543945,
+      "learning_rate": 2.712238147739802e-06,
+      "loss": 1.3139,
+      "step": 8625
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 25.179645538330078,
+      "learning_rate": 2.7102335371354117e-06,
+      "loss": 1.5291,
+      "step": 8626
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 9.155866622924805,
+      "learning_rate": 2.7082289265310213e-06,
+      "loss": 0.6078,
+      "step": 8627
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 25.500167846679688,
+      "learning_rate": 2.7062243159266314e-06,
+      "loss": 1.8845,
+      "step": 8628
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 14.735610008239746,
+      "learning_rate": 2.7042197053222415e-06,
+      "loss": 1.3812,
+      "step": 8629
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 13.087080001831055,
+      "learning_rate": 2.702215094717851e-06,
+      "loss": 2.3241,
+      "step": 8630
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 15.85035228729248,
+      "learning_rate": 2.7002104841134612e-06,
+      "loss": 2.1271,
+      "step": 8631
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 16.836761474609375,
+      "learning_rate": 2.698205873509071e-06,
+      "loss": 0.9277,
+      "step": 8632
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 17.871294021606445,
+      "learning_rate": 2.6962012629046806e-06,
+      "loss": 1.3209,
+      "step": 8633
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 10.736555099487305,
+      "learning_rate": 2.694196652300291e-06,
+      "loss": 0.9247,
+      "step": 8634
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 4.654946804046631,
+      "learning_rate": 2.6921920416959007e-06,
+      "loss": 0.6192,
+      "step": 8635
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 62.396331787109375,
+      "learning_rate": 2.6901874310915104e-06,
+      "loss": 2.6783,
+      "step": 8636
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 37.43707275390625,
+      "learning_rate": 2.688182820487121e-06,
+      "loss": 0.6253,
+      "step": 8637
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 10.426443099975586,
+      "learning_rate": 2.6861782098827306e-06,
+      "loss": 1.1717,
+      "step": 8638
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 77.32369995117188,
+      "learning_rate": 2.6841735992783402e-06,
+      "loss": 2.797,
+      "step": 8639
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 18.485462188720703,
+      "learning_rate": 2.6821689886739503e-06,
+      "loss": 0.9191,
+      "step": 8640
+    },
+    {
+      "epoch": 2.6,
+      "eval_loss": 0.17308995127677917,
+      "eval_runtime": 43.9129,
+      "eval_samples_per_second": 33.68,
+      "eval_steps_per_second": 33.68,
+      "step": 8640
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 10.55259895324707,
+      "learning_rate": 2.6801643780695604e-06,
+      "loss": 1.4176,
+      "step": 8641
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 14.04641056060791,
+      "learning_rate": 2.67815976746517e-06,
+      "loss": 1.3492,
+      "step": 8642
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 11.673070907592773,
+      "learning_rate": 2.67615515686078e-06,
+      "loss": 1.3232,
+      "step": 8643
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 22.69013023376465,
+      "learning_rate": 2.67415054625639e-06,
+      "loss": 1.535,
+      "step": 8644
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 31.60176658630371,
+      "learning_rate": 2.6721459356519995e-06,
+      "loss": 2.397,
+      "step": 8645
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 72.22872161865234,
+      "learning_rate": 2.67014132504761e-06,
+      "loss": 3.1373,
+      "step": 8646
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 12.671222686767578,
+      "learning_rate": 2.6681367144432196e-06,
+      "loss": 0.9173,
+      "step": 8647
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 24.99388885498047,
+      "learning_rate": 2.6661321038388293e-06,
+      "loss": 1.3955,
+      "step": 8648
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 28.980268478393555,
+      "learning_rate": 2.66412749323444e-06,
+      "loss": 1.3139,
+      "step": 8649
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 14.720972061157227,
+      "learning_rate": 2.6621228826300495e-06,
+      "loss": 0.8373,
+      "step": 8650
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 63.49948501586914,
+      "learning_rate": 2.660118272025659e-06,
+      "loss": 1.9721,
+      "step": 8651
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 11.087202072143555,
+      "learning_rate": 2.658113661421269e-06,
+      "loss": 0.9022,
+      "step": 8652
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 21.9641170501709,
+      "learning_rate": 2.6561090508168793e-06,
+      "loss": 1.9667,
+      "step": 8653
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 16.303483963012695,
+      "learning_rate": 2.654104440212489e-06,
+      "loss": 1.1034,
+      "step": 8654
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 15.40654468536377,
+      "learning_rate": 2.652099829608099e-06,
+      "loss": 0.9308,
+      "step": 8655
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 37.456077575683594,
+      "learning_rate": 2.6500952190037087e-06,
+      "loss": 1.5839,
+      "step": 8656
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 189.6568145751953,
+      "learning_rate": 2.6480906083993184e-06,
+      "loss": 1.1821,
+      "step": 8657
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 14.535365104675293,
+      "learning_rate": 2.646085997794929e-06,
+      "loss": 1.221,
+      "step": 8658
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 15.61431884765625,
+      "learning_rate": 2.6440813871905385e-06,
+      "loss": 0.8718,
+      "step": 8659
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 16.955713272094727,
+      "learning_rate": 2.642076776586148e-06,
+      "loss": 1.0111,
+      "step": 8660
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 17.068771362304688,
+      "learning_rate": 2.6400721659817587e-06,
+      "loss": 1.5695,
+      "step": 8661
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 10.242925643920898,
+      "learning_rate": 2.6380675553773684e-06,
+      "loss": 1.4754,
+      "step": 8662
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 30.505434036254883,
+      "learning_rate": 2.636062944772978e-06,
+      "loss": 1.1623,
+      "step": 8663
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 44.106178283691406,
+      "learning_rate": 2.6340583341685877e-06,
+      "loss": 2.1524,
+      "step": 8664
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 15.738425254821777,
+      "learning_rate": 2.6320537235641978e-06,
+      "loss": 1.3509,
+      "step": 8665
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 37.473567962646484,
+      "learning_rate": 2.630049112959808e-06,
+      "loss": 1.1996,
+      "step": 8666
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 20.072607040405273,
+      "learning_rate": 2.6280445023554175e-06,
+      "loss": 1.1615,
+      "step": 8667
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 14.636398315429688,
+      "learning_rate": 2.6260398917510276e-06,
+      "loss": 0.863,
+      "step": 8668
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 9.129220008850098,
+      "learning_rate": 2.6240352811466373e-06,
+      "loss": 1.157,
+      "step": 8669
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 19.694377899169922,
+      "learning_rate": 2.6220306705422473e-06,
+      "loss": 0.8906,
+      "step": 8670
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 13.88318157196045,
+      "learning_rate": 2.6200260599378574e-06,
+      "loss": 1.343,
+      "step": 8671
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 11.803475379943848,
+      "learning_rate": 2.618021449333467e-06,
+      "loss": 1.4005,
+      "step": 8672
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 18.531572341918945,
+      "learning_rate": 2.6160168387290767e-06,
+      "loss": 1.3321,
+      "step": 8673
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 6.691076278686523,
+      "learning_rate": 2.6140122281246872e-06,
+      "loss": 0.7632,
+      "step": 8674
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 8.980220794677734,
+      "learning_rate": 2.612007617520297e-06,
+      "loss": 0.7735,
+      "step": 8675
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 8.376172065734863,
+      "learning_rate": 2.6100030069159066e-06,
+      "loss": 1.0655,
+      "step": 8676
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 10.30313491821289,
+      "learning_rate": 2.6079983963115167e-06,
+      "loss": 1.6599,
+      "step": 8677
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 23.8471622467041,
+      "learning_rate": 2.6059937857071267e-06,
+      "loss": 0.579,
+      "step": 8678
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 25.849637985229492,
+      "learning_rate": 2.6039891751027364e-06,
+      "loss": 1.8009,
+      "step": 8679
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 26.335845947265625,
+      "learning_rate": 2.6019845644983465e-06,
+      "loss": 1.6794,
+      "step": 8680
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 21.52005386352539,
+      "learning_rate": 2.599979953893956e-06,
+      "loss": 1.3504,
+      "step": 8681
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 24.20670509338379,
+      "learning_rate": 2.5979753432895662e-06,
+      "loss": 0.9255,
+      "step": 8682
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 105.50646209716797,
+      "learning_rate": 2.5959707326851763e-06,
+      "loss": 2.7274,
+      "step": 8683
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 28.476221084594727,
+      "learning_rate": 2.593966122080786e-06,
+      "loss": 2.7293,
+      "step": 8684
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 23.278079986572266,
+      "learning_rate": 2.5919615114763956e-06,
+      "loss": 2.1648,
+      "step": 8685
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 25.680274963378906,
+      "learning_rate": 2.589956900872006e-06,
+      "loss": 1.5944,
+      "step": 8686
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 16.657413482666016,
+      "learning_rate": 2.587952290267616e-06,
+      "loss": 0.8337,
+      "step": 8687
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 24.16915512084961,
+      "learning_rate": 2.5859476796632255e-06,
+      "loss": 1.168,
+      "step": 8688
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 13.357244491577148,
+      "learning_rate": 2.5839430690588355e-06,
+      "loss": 0.9006,
+      "step": 8689
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 17.03959083557129,
+      "learning_rate": 2.5819384584544456e-06,
+      "loss": 1.3754,
+      "step": 8690
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 14.713468551635742,
+      "learning_rate": 2.5799338478500553e-06,
+      "loss": 1.3571,
+      "step": 8691
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 48.19596481323242,
+      "learning_rate": 2.5779292372456654e-06,
+      "loss": 1.8712,
+      "step": 8692
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 42.44739532470703,
+      "learning_rate": 2.575924626641275e-06,
+      "loss": 1.2746,
+      "step": 8693
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 35.6419792175293,
+      "learning_rate": 2.5739200160368847e-06,
+      "loss": 1.5702,
+      "step": 8694
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 19.48603630065918,
+      "learning_rate": 2.571915405432495e-06,
+      "loss": 1.9149,
+      "step": 8695
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 15.384188652038574,
+      "learning_rate": 2.569910794828105e-06,
+      "loss": 1.4403,
+      "step": 8696
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 18.034732818603516,
+      "learning_rate": 2.5679061842237145e-06,
+      "loss": 1.2765,
+      "step": 8697
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 12.634345054626465,
+      "learning_rate": 2.565901573619325e-06,
+      "loss": 1.0044,
+      "step": 8698
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 12.71899127960205,
+      "learning_rate": 2.5638969630149347e-06,
+      "loss": 1.3239,
+      "step": 8699
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 597.7841186523438,
+      "learning_rate": 2.5618923524105444e-06,
+      "loss": 1.6701,
+      "step": 8700
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 16.543411254882812,
+      "learning_rate": 2.5598877418061544e-06,
+      "loss": 1.4786,
+      "step": 8701
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 48.39625930786133,
+      "learning_rate": 2.5578831312017645e-06,
+      "loss": 1.538,
+      "step": 8702
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 10.015215873718262,
+      "learning_rate": 2.555878520597374e-06,
+      "loss": 1.4655,
+      "step": 8703
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 18.111173629760742,
+      "learning_rate": 2.5538739099929843e-06,
+      "loss": 1.539,
+      "step": 8704
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 34.77145767211914,
+      "learning_rate": 2.551869299388594e-06,
+      "loss": 1.5545,
+      "step": 8705
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 17.934062957763672,
+      "learning_rate": 2.5498646887842036e-06,
+      "loss": 1.3262,
+      "step": 8706
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 14.223177909851074,
+      "learning_rate": 2.5478600781798137e-06,
+      "loss": 2.0243,
+      "step": 8707
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 21.776586532592773,
+      "learning_rate": 2.5458554675754238e-06,
+      "loss": 1.3405,
+      "step": 8708
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 10.819275856018066,
+      "learning_rate": 2.5438508569710334e-06,
+      "loss": 1.488,
+      "step": 8709
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 12.452916145324707,
+      "learning_rate": 2.541846246366643e-06,
+      "loss": 1.3601,
+      "step": 8710
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 13.645332336425781,
+      "learning_rate": 2.5398416357622536e-06,
+      "loss": 1.2305,
+      "step": 8711
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 157.49203491210938,
+      "learning_rate": 2.5378370251578633e-06,
+      "loss": 1.5407,
+      "step": 8712
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 15.182050704956055,
+      "learning_rate": 2.535832414553473e-06,
+      "loss": 0.874,
+      "step": 8713
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 16.068603515625,
+      "learning_rate": 2.5338278039490834e-06,
+      "loss": 1.3392,
+      "step": 8714
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 11.687101364135742,
+      "learning_rate": 2.531823193344693e-06,
+      "loss": 0.9513,
+      "step": 8715
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 16.209623336791992,
+      "learning_rate": 2.5298185827403027e-06,
+      "loss": 0.8897,
+      "step": 8716
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 42.321136474609375,
+      "learning_rate": 2.527813972135913e-06,
+      "loss": 2.2684,
+      "step": 8717
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 19.06346321105957,
+      "learning_rate": 2.5258093615315225e-06,
+      "loss": 1.3514,
+      "step": 8718
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 10.905227661132812,
+      "learning_rate": 2.5238047509271326e-06,
+      "loss": 0.6565,
+      "step": 8719
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 17.082691192626953,
+      "learning_rate": 2.5218001403227427e-06,
+      "loss": 2.4891,
+      "step": 8720
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 22.291873931884766,
+      "learning_rate": 2.5197955297183523e-06,
+      "loss": 2.0741,
+      "step": 8721
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 9.077491760253906,
+      "learning_rate": 2.517790919113962e-06,
+      "loss": 1.115,
+      "step": 8722
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 10.916313171386719,
+      "learning_rate": 2.5157863085095725e-06,
+      "loss": 1.4171,
+      "step": 8723
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 54.628211975097656,
+      "learning_rate": 2.513781697905182e-06,
+      "loss": 0.9545,
+      "step": 8724
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 11.69243049621582,
+      "learning_rate": 2.511777087300792e-06,
+      "loss": 0.7757,
+      "step": 8725
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 23.144771575927734,
+      "learning_rate": 2.5097724766964023e-06,
+      "loss": 1.4352,
+      "step": 8726
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 23.521812438964844,
+      "learning_rate": 2.507767866092012e-06,
+      "loss": 1.276,
+      "step": 8727
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 29.313697814941406,
+      "learning_rate": 2.5057632554876216e-06,
+      "loss": 1.4241,
+      "step": 8728
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 5.5264716148376465,
+      "learning_rate": 2.5037586448832317e-06,
+      "loss": 0.6715,
+      "step": 8729
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 107.75334167480469,
+      "learning_rate": 2.5017540342788414e-06,
+      "loss": 2.3175,
+      "step": 8730
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 16.5670108795166,
+      "learning_rate": 2.4997494236744515e-06,
+      "loss": 1.7357,
+      "step": 8731
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 25.61155128479004,
+      "learning_rate": 2.497744813070061e-06,
+      "loss": 1.3919,
+      "step": 8732
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 12.228818893432617,
+      "learning_rate": 2.495740202465671e-06,
+      "loss": 2.4026,
+      "step": 8733
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 16.512880325317383,
+      "learning_rate": 2.4937355918612813e-06,
+      "loss": 1.9608,
+      "step": 8734
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 34.81612014770508,
+      "learning_rate": 2.491730981256891e-06,
+      "loss": 1.0885,
+      "step": 8735
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 31.149940490722656,
+      "learning_rate": 2.489726370652501e-06,
+      "loss": 1.0149,
+      "step": 8736
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 13.102104187011719,
+      "learning_rate": 2.4877217600481107e-06,
+      "loss": 1.5993,
+      "step": 8737
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 18.59217071533203,
+      "learning_rate": 2.4857171494437208e-06,
+      "loss": 1.3812,
+      "step": 8738
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 15.023920059204102,
+      "learning_rate": 2.483712538839331e-06,
+      "loss": 1.0019,
+      "step": 8739
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 15.331181526184082,
+      "learning_rate": 2.4817079282349405e-06,
+      "loss": 1.5826,
+      "step": 8740
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 32.79924774169922,
+      "learning_rate": 2.4797033176305506e-06,
+      "loss": 2.171,
+      "step": 8741
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 11.717575073242188,
+      "learning_rate": 2.4776987070261603e-06,
+      "loss": 0.7835,
+      "step": 8742
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 19.555763244628906,
+      "learning_rate": 2.4756940964217704e-06,
+      "loss": 2.7394,
+      "step": 8743
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 37.40543746948242,
+      "learning_rate": 2.47368948581738e-06,
+      "loss": 2.1193,
+      "step": 8744
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 18.472028732299805,
+      "learning_rate": 2.47168487521299e-06,
+      "loss": 1.0426,
+      "step": 8745
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 15.350752830505371,
+      "learning_rate": 2.4696802646085998e-06,
+      "loss": 1.0137,
+      "step": 8746
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 12.595955848693848,
+      "learning_rate": 2.46767565400421e-06,
+      "loss": 1.6324,
+      "step": 8747
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 29.87061309814453,
+      "learning_rate": 2.46567104339982e-06,
+      "loss": 1.5554,
+      "step": 8748
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 17.269176483154297,
+      "learning_rate": 2.4636664327954296e-06,
+      "loss": 1.6175,
+      "step": 8749
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 16.87700653076172,
+      "learning_rate": 2.4616618221910397e-06,
+      "loss": 1.0128,
+      "step": 8750
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 10.00948429107666,
+      "learning_rate": 2.4596572115866498e-06,
+      "loss": 0.9691,
+      "step": 8751
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 21.35502815246582,
+      "learning_rate": 2.4576526009822594e-06,
+      "loss": 1.3124,
+      "step": 8752
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 11.689464569091797,
+      "learning_rate": 2.4556479903778695e-06,
+      "loss": 1.3952,
+      "step": 8753
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 141.92120361328125,
+      "learning_rate": 2.453643379773479e-06,
+      "loss": 2.629,
+      "step": 8754
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 26.330507278442383,
+      "learning_rate": 2.4516387691690893e-06,
+      "loss": 3.4614,
+      "step": 8755
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 10.556838035583496,
+      "learning_rate": 2.449634158564699e-06,
+      "loss": 1.7251,
+      "step": 8756
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 45.57450485229492,
+      "learning_rate": 2.4476295479603086e-06,
+      "loss": 1.7645,
+      "step": 8757
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 11.817948341369629,
+      "learning_rate": 2.4456249373559187e-06,
+      "loss": 1.0033,
+      "step": 8758
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 7.083280563354492,
+      "learning_rate": 2.4436203267515287e-06,
+      "loss": 0.7292,
+      "step": 8759
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 18.485204696655273,
+      "learning_rate": 2.4416157161471384e-06,
+      "loss": 1.5034,
+      "step": 8760
+    },
+    {
+      "epoch": 2.63,
+      "eval_loss": 0.163481205701828,
+      "eval_runtime": 43.3055,
+      "eval_samples_per_second": 34.153,
+      "eval_steps_per_second": 34.153,
+      "step": 8760
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 21.252248764038086,
+      "learning_rate": 2.4396111055427485e-06,
+      "loss": 1.1208,
+      "step": 8761
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 16.638259887695312,
+      "learning_rate": 2.4376064949383586e-06,
+      "loss": 1.8795,
+      "step": 8762
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 20.396753311157227,
+      "learning_rate": 2.4356018843339682e-06,
+      "loss": 1.5712,
+      "step": 8763
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 7.8299455642700195,
+      "learning_rate": 2.4335972737295783e-06,
+      "loss": 0.7808,
+      "step": 8764
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 11.673842430114746,
+      "learning_rate": 2.4315926631251884e-06,
+      "loss": 0.8884,
+      "step": 8765
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 40.24734115600586,
+      "learning_rate": 2.429588052520798e-06,
+      "loss": 2.0446,
+      "step": 8766
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 29.379608154296875,
+      "learning_rate": 2.4275834419164077e-06,
+      "loss": 1.6934,
+      "step": 8767
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 66.81684112548828,
+      "learning_rate": 2.425578831312018e-06,
+      "loss": 1.426,
+      "step": 8768
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 81.16704559326172,
+      "learning_rate": 2.4235742207076275e-06,
+      "loss": 1.3078,
+      "step": 8769
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 13.74372386932373,
+      "learning_rate": 2.4215696101032376e-06,
+      "loss": 1.543,
+      "step": 8770
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 9.779702186584473,
+      "learning_rate": 2.4195649994988476e-06,
+      "loss": 1.0268,
+      "step": 8771
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 14.754984855651855,
+      "learning_rate": 2.4175603888944573e-06,
+      "loss": 1.7506,
+      "step": 8772
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 29.609939575195312,
+      "learning_rate": 2.4155557782900674e-06,
+      "loss": 2.8884,
+      "step": 8773
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 100.57302856445312,
+      "learning_rate": 2.4135511676856775e-06,
+      "loss": 1.1484,
+      "step": 8774
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 29.211971282958984,
+      "learning_rate": 2.411546557081287e-06,
+      "loss": 2.1784,
+      "step": 8775
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 31.128265380859375,
+      "learning_rate": 2.409541946476897e-06,
+      "loss": 1.6176,
+      "step": 8776
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 16.750701904296875,
+      "learning_rate": 2.4075373358725073e-06,
+      "loss": 0.8013,
+      "step": 8777
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 65.17308044433594,
+      "learning_rate": 2.405532725268117e-06,
+      "loss": 1.9283,
+      "step": 8778
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 18.673601150512695,
+      "learning_rate": 2.4035281146637266e-06,
+      "loss": 1.3132,
+      "step": 8779
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 12.052127838134766,
+      "learning_rate": 2.4015235040593367e-06,
+      "loss": 0.8893,
+      "step": 8780
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 33.23358917236328,
+      "learning_rate": 2.3995188934549464e-06,
+      "loss": 1.5921,
+      "step": 8781
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 32.95631408691406,
+      "learning_rate": 2.3975142828505564e-06,
+      "loss": 1.5091,
+      "step": 8782
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 21.939970016479492,
+      "learning_rate": 2.395509672246166e-06,
+      "loss": 1.1536,
+      "step": 8783
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 19.46248435974121,
+      "learning_rate": 2.393505061641776e-06,
+      "loss": 1.7254,
+      "step": 8784
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 20.779407501220703,
+      "learning_rate": 2.3915004510373863e-06,
+      "loss": 0.6734,
+      "step": 8785
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 18.915145874023438,
+      "learning_rate": 2.389495840432996e-06,
+      "loss": 1.4362,
+      "step": 8786
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 15.002533912658691,
+      "learning_rate": 2.387491229828606e-06,
+      "loss": 1.3399,
+      "step": 8787
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 205.16641235351562,
+      "learning_rate": 2.385486619224216e-06,
+      "loss": 1.7637,
+      "step": 8788
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 16.24716567993164,
+      "learning_rate": 2.3834820086198258e-06,
+      "loss": 2.3097,
+      "step": 8789
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 27.199005126953125,
+      "learning_rate": 2.381477398015436e-06,
+      "loss": 2.0923,
+      "step": 8790
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 7.843896865844727,
+      "learning_rate": 2.3794727874110455e-06,
+      "loss": 0.556,
+      "step": 8791
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 15.323102951049805,
+      "learning_rate": 2.3774681768066556e-06,
+      "loss": 0.9552,
+      "step": 8792
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 14.061708450317383,
+      "learning_rate": 2.3754635662022653e-06,
+      "loss": 1.4372,
+      "step": 8793
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 12.933516502380371,
+      "learning_rate": 2.3734589555978753e-06,
+      "loss": 1.2338,
+      "step": 8794
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 7.49314546585083,
+      "learning_rate": 2.371454344993485e-06,
+      "loss": 0.731,
+      "step": 8795
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 15.913431167602539,
+      "learning_rate": 2.369449734389095e-06,
+      "loss": 1.2261,
+      "step": 8796
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 12.59531307220459,
+      "learning_rate": 2.367445123784705e-06,
+      "loss": 1.0611,
+      "step": 8797
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 16.93478012084961,
+      "learning_rate": 2.365440513180315e-06,
+      "loss": 1.6995,
+      "step": 8798
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 11.499807357788086,
+      "learning_rate": 2.363435902575925e-06,
+      "loss": 1.3401,
+      "step": 8799
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 33.20011901855469,
+      "learning_rate": 2.3614312919715346e-06,
+      "loss": 2.086,
+      "step": 8800
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 10.58979606628418,
+      "learning_rate": 2.3594266813671447e-06,
+      "loss": 1.1518,
+      "step": 8801
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 15.411652565002441,
+      "learning_rate": 2.3574220707627547e-06,
+      "loss": 0.9793,
+      "step": 8802
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 68.01510620117188,
+      "learning_rate": 2.3554174601583644e-06,
+      "loss": 2.0531,
+      "step": 8803
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 56.63874053955078,
+      "learning_rate": 2.3534128495539745e-06,
+      "loss": 1.1724,
+      "step": 8804
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 60.3148193359375,
+      "learning_rate": 2.351408238949584e-06,
+      "loss": 2.7668,
+      "step": 8805
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 38.45302963256836,
+      "learning_rate": 2.3494036283451942e-06,
+      "loss": 1.0094,
+      "step": 8806
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 10.030008316040039,
+      "learning_rate": 2.347399017740804e-06,
+      "loss": 1.0792,
+      "step": 8807
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 10.08592700958252,
+      "learning_rate": 2.345394407136414e-06,
+      "loss": 0.7964,
+      "step": 8808
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 125.09164428710938,
+      "learning_rate": 2.3433897965320236e-06,
+      "loss": 3.0581,
+      "step": 8809
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 19.070087432861328,
+      "learning_rate": 2.3413851859276337e-06,
+      "loss": 0.8685,
+      "step": 8810
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 10.671518325805664,
+      "learning_rate": 2.339380575323244e-06,
+      "loss": 1.1835,
+      "step": 8811
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 27.613374710083008,
+      "learning_rate": 2.3373759647188535e-06,
+      "loss": 1.4339,
+      "step": 8812
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 18.462194442749023,
+      "learning_rate": 2.3353713541144636e-06,
+      "loss": 1.0176,
+      "step": 8813
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 13.578800201416016,
+      "learning_rate": 2.3333667435100736e-06,
+      "loss": 2.1367,
+      "step": 8814
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 43.0266227722168,
+      "learning_rate": 2.3313621329056833e-06,
+      "loss": 2.1957,
+      "step": 8815
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 40.213008880615234,
+      "learning_rate": 2.3293575223012934e-06,
+      "loss": 1.9988,
+      "step": 8816
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 14.245367050170898,
+      "learning_rate": 2.327352911696903e-06,
+      "loss": 1.891,
+      "step": 8817
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 13.524105072021484,
+      "learning_rate": 2.3253483010925127e-06,
+      "loss": 1.4238,
+      "step": 8818
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 22.680530548095703,
+      "learning_rate": 2.323343690488123e-06,
+      "loss": 1.2083,
+      "step": 8819
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 13.493782043457031,
+      "learning_rate": 2.3213390798837325e-06,
+      "loss": 1.1022,
+      "step": 8820
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 10.675830841064453,
+      "learning_rate": 2.3193344692793425e-06,
+      "loss": 0.5942,
+      "step": 8821
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 12.953835487365723,
+      "learning_rate": 2.3173298586749526e-06,
+      "loss": 1.445,
+      "step": 8822
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 10.076544761657715,
+      "learning_rate": 2.3153252480705623e-06,
+      "loss": 1.5124,
+      "step": 8823
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 21.92123031616211,
+      "learning_rate": 2.3133206374661724e-06,
+      "loss": 1.8663,
+      "step": 8824
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 25.028993606567383,
+      "learning_rate": 2.3113160268617824e-06,
+      "loss": 1.0345,
+      "step": 8825
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 8.76278305053711,
+      "learning_rate": 2.309311416257392e-06,
+      "loss": 0.9227,
+      "step": 8826
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 7.540186405181885,
+      "learning_rate": 2.307306805653002e-06,
+      "loss": 0.9597,
+      "step": 8827
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 21.144914627075195,
+      "learning_rate": 2.3053021950486123e-06,
+      "loss": 1.3642,
+      "step": 8828
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 17.2004451751709,
+      "learning_rate": 2.303297584444222e-06,
+      "loss": 0.6705,
+      "step": 8829
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 8.921226501464844,
+      "learning_rate": 2.3012929738398316e-06,
+      "loss": 1.096,
+      "step": 8830
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 13.363105773925781,
+      "learning_rate": 2.2992883632354417e-06,
+      "loss": 0.9989,
+      "step": 8831
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 11.100980758666992,
+      "learning_rate": 2.2972837526310513e-06,
+      "loss": 1.4,
+      "step": 8832
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 18.512340545654297,
+      "learning_rate": 2.2952791420266614e-06,
+      "loss": 1.3327,
+      "step": 8833
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 52.814693450927734,
+      "learning_rate": 2.2932745314222715e-06,
+      "loss": 1.4743,
+      "step": 8834
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 37.01776123046875,
+      "learning_rate": 2.291269920817881e-06,
+      "loss": 1.1938,
+      "step": 8835
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 14.084249496459961,
+      "learning_rate": 2.2892653102134913e-06,
+      "loss": 1.4542,
+      "step": 8836
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 34.67374038696289,
+      "learning_rate": 2.2872606996091013e-06,
+      "loss": 1.9464,
+      "step": 8837
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 16.59427833557129,
+      "learning_rate": 2.285256089004711e-06,
+      "loss": 1.5703,
+      "step": 8838
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 20.186216354370117,
+      "learning_rate": 2.283251478400321e-06,
+      "loss": 1.527,
+      "step": 8839
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 34.25483703613281,
+      "learning_rate": 2.281246867795931e-06,
+      "loss": 2.5718,
+      "step": 8840
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 28.123106002807617,
+      "learning_rate": 2.279242257191541e-06,
+      "loss": 1.5148,
+      "step": 8841
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 55.747276306152344,
+      "learning_rate": 2.2772376465871505e-06,
+      "loss": 1.7577,
+      "step": 8842
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 21.762004852294922,
+      "learning_rate": 2.2752330359827606e-06,
+      "loss": 1.3397,
+      "step": 8843
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 38.36250305175781,
+      "learning_rate": 2.2732284253783702e-06,
+      "loss": 1.2055,
+      "step": 8844
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 14.818611145019531,
+      "learning_rate": 2.2712238147739803e-06,
+      "loss": 1.3097,
+      "step": 8845
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 7.749112129211426,
+      "learning_rate": 2.26921920416959e-06,
+      "loss": 0.4492,
+      "step": 8846
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 10.357666969299316,
+      "learning_rate": 2.2672145935652e-06,
+      "loss": 0.9897,
+      "step": 8847
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 15.227676391601562,
+      "learning_rate": 2.26520998296081e-06,
+      "loss": 1.3606,
+      "step": 8848
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 16.850852966308594,
+      "learning_rate": 2.26320537235642e-06,
+      "loss": 1.4527,
+      "step": 8849
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 16.4047794342041,
+      "learning_rate": 2.26120076175203e-06,
+      "loss": 1.6789,
+      "step": 8850
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 7.696395397186279,
+      "learning_rate": 2.25919615114764e-06,
+      "loss": 0.9858,
+      "step": 8851
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 21.263986587524414,
+      "learning_rate": 2.2571915405432496e-06,
+      "loss": 1.7351,
+      "step": 8852
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 7.717570781707764,
+      "learning_rate": 2.2551869299388597e-06,
+      "loss": 1.0206,
+      "step": 8853
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 14.79688549041748,
+      "learning_rate": 2.2531823193344694e-06,
+      "loss": 1.7117,
+      "step": 8854
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 8.93305492401123,
+      "learning_rate": 2.2511777087300795e-06,
+      "loss": 1.839,
+      "step": 8855
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 13.082503318786621,
+      "learning_rate": 2.249173098125689e-06,
+      "loss": 1.252,
+      "step": 8856
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 11.024530410766602,
+      "learning_rate": 2.2471684875212992e-06,
+      "loss": 0.5396,
+      "step": 8857
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 30.2106876373291,
+      "learning_rate": 2.245163876916909e-06,
+      "loss": 1.8259,
+      "step": 8858
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 12.592206954956055,
+      "learning_rate": 2.243159266312519e-06,
+      "loss": 2.2187,
+      "step": 8859
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 29.125699996948242,
+      "learning_rate": 2.241154655708129e-06,
+      "loss": 1.1455,
+      "step": 8860
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 16.28235626220703,
+      "learning_rate": 2.2391500451037387e-06,
+      "loss": 1.9047,
+      "step": 8861
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 30.546987533569336,
+      "learning_rate": 2.237145434499349e-06,
+      "loss": 1.7441,
+      "step": 8862
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 12.745742797851562,
+      "learning_rate": 2.2351408238949585e-06,
+      "loss": 0.5604,
+      "step": 8863
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 10.053805351257324,
+      "learning_rate": 2.2331362132905685e-06,
+      "loss": 1.467,
+      "step": 8864
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 22.58064842224121,
+      "learning_rate": 2.2311316026861786e-06,
+      "loss": 1.3182,
+      "step": 8865
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 10.636580467224121,
+      "learning_rate": 2.2291269920817883e-06,
+      "loss": 1.3195,
+      "step": 8866
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 18.260143280029297,
+      "learning_rate": 2.2271223814773984e-06,
+      "loss": 1.3604,
+      "step": 8867
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 8.831524848937988,
+      "learning_rate": 2.225117770873008e-06,
+      "loss": 1.1861,
+      "step": 8868
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 13.567652702331543,
+      "learning_rate": 2.2231131602686177e-06,
+      "loss": 1.0606,
+      "step": 8869
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 31.04169273376465,
+      "learning_rate": 2.2211085496642278e-06,
+      "loss": 1.1305,
+      "step": 8870
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 17.62767219543457,
+      "learning_rate": 2.219103939059838e-06,
+      "loss": 1.8986,
+      "step": 8871
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 31.392688751220703,
+      "learning_rate": 2.2170993284554475e-06,
+      "loss": 2.3315,
+      "step": 8872
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 18.428953170776367,
+      "learning_rate": 2.2150947178510576e-06,
+      "loss": 1.559,
+      "step": 8873
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 24.9808292388916,
+      "learning_rate": 2.2130901072466677e-06,
+      "loss": 1.6667,
+      "step": 8874
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 23.458446502685547,
+      "learning_rate": 2.2110854966422773e-06,
+      "loss": 0.8407,
+      "step": 8875
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 10.7035493850708,
+      "learning_rate": 2.2090808860378874e-06,
+      "loss": 1.3342,
+      "step": 8876
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 21.982637405395508,
+      "learning_rate": 2.2070762754334975e-06,
+      "loss": 1.8623,
+      "step": 8877
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 10.788835525512695,
+      "learning_rate": 2.205071664829107e-06,
+      "loss": 1.14,
+      "step": 8878
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 12.52689266204834,
+      "learning_rate": 2.2030670542247173e-06,
+      "loss": 0.8905,
+      "step": 8879
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 12.546585083007812,
+      "learning_rate": 2.201062443620327e-06,
+      "loss": 1.4123,
+      "step": 8880
+    },
+    {
+      "epoch": 2.67,
+      "eval_loss": 0.1634318232536316,
+      "eval_runtime": 43.2728,
+      "eval_samples_per_second": 34.179,
+      "eval_steps_per_second": 34.179,
+      "step": 8880
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 23.065811157226562,
+      "learning_rate": 2.1990578330159366e-06,
+      "loss": 1.6073,
+      "step": 8881
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 8.505629539489746,
+      "learning_rate": 2.1970532224115467e-06,
+      "loss": 0.6363,
+      "step": 8882
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 19.605792999267578,
+      "learning_rate": 2.1950486118071563e-06,
+      "loss": 1.1445,
+      "step": 8883
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 62.97642517089844,
+      "learning_rate": 2.1930440012027664e-06,
+      "loss": 3.0308,
+      "step": 8884
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 34.175453186035156,
+      "learning_rate": 2.1910393905983765e-06,
+      "loss": 1.2063,
+      "step": 8885
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 41.114051818847656,
+      "learning_rate": 2.189034779993986e-06,
+      "loss": 1.2233,
+      "step": 8886
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 22.9063663482666,
+      "learning_rate": 2.1870301693895962e-06,
+      "loss": 1.1179,
+      "step": 8887
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 12.28271770477295,
+      "learning_rate": 2.1850255587852063e-06,
+      "loss": 1.1284,
+      "step": 8888
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 18.61870765686035,
+      "learning_rate": 2.183020948180816e-06,
+      "loss": 1.1018,
+      "step": 8889
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 14.14230728149414,
+      "learning_rate": 2.181016337576426e-06,
+      "loss": 2.1612,
+      "step": 8890
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 17.895893096923828,
+      "learning_rate": 2.179011726972036e-06,
+      "loss": 2.1376,
+      "step": 8891
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 10.05941390991211,
+      "learning_rate": 2.177007116367646e-06,
+      "loss": 1.9443,
+      "step": 8892
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 17.179000854492188,
+      "learning_rate": 2.1750025057632555e-06,
+      "loss": 1.3958,
+      "step": 8893
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 18.431400299072266,
+      "learning_rate": 2.1729978951588656e-06,
+      "loss": 1.2758,
+      "step": 8894
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 18.084781646728516,
+      "learning_rate": 2.1709932845544752e-06,
+      "loss": 1.7674,
+      "step": 8895
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 15.693217277526855,
+      "learning_rate": 2.1689886739500853e-06,
+      "loss": 1.0816,
+      "step": 8896
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 18.634521484375,
+      "learning_rate": 2.1669840633456954e-06,
+      "loss": 1.1063,
+      "step": 8897
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 11.23396110534668,
+      "learning_rate": 2.164979452741305e-06,
+      "loss": 0.6603,
+      "step": 8898
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 8.881089210510254,
+      "learning_rate": 2.162974842136915e-06,
+      "loss": 0.7917,
+      "step": 8899
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 6.563168525695801,
+      "learning_rate": 2.1609702315325252e-06,
+      "loss": 0.7384,
+      "step": 8900
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 45.68901824951172,
+      "learning_rate": 2.158965620928135e-06,
+      "loss": 2.6647,
+      "step": 8901
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 10.925772666931152,
+      "learning_rate": 2.156961010323745e-06,
+      "loss": 0.5719,
+      "step": 8902
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 52.76468276977539,
+      "learning_rate": 2.1549563997193546e-06,
+      "loss": 1.5088,
+      "step": 8903
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 17.248638153076172,
+      "learning_rate": 2.1529517891149647e-06,
+      "loss": 1.2967,
+      "step": 8904
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 51.29323196411133,
+      "learning_rate": 2.1509471785105744e-06,
+      "loss": 1.6994,
+      "step": 8905
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 42.033302307128906,
+      "learning_rate": 2.1489425679061845e-06,
+      "loss": 1.3593,
+      "step": 8906
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 9.53936767578125,
+      "learning_rate": 2.146937957301794e-06,
+      "loss": 0.7497,
+      "step": 8907
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 29.297077178955078,
+      "learning_rate": 2.144933346697404e-06,
+      "loss": 1.042,
+      "step": 8908
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 37.91065216064453,
+      "learning_rate": 2.142928736093014e-06,
+      "loss": 2.1151,
+      "step": 8909
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 24.55820655822754,
+      "learning_rate": 2.140924125488624e-06,
+      "loss": 1.9639,
+      "step": 8910
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 30.16676902770996,
+      "learning_rate": 2.138919514884234e-06,
+      "loss": 1.3401,
+      "step": 8911
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 47.085235595703125,
+      "learning_rate": 2.1369149042798437e-06,
+      "loss": 2.4905,
+      "step": 8912
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 19.413427352905273,
+      "learning_rate": 2.1349102936754538e-06,
+      "loss": 1.4043,
+      "step": 8913
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 31.471092224121094,
+      "learning_rate": 2.132905683071064e-06,
+      "loss": 1.9347,
+      "step": 8914
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 26.705747604370117,
+      "learning_rate": 2.1309010724666735e-06,
+      "loss": 2.8265,
+      "step": 8915
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 43.82736587524414,
+      "learning_rate": 2.1288964618622836e-06,
+      "loss": 1.3732,
+      "step": 8916
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 12.801732063293457,
+      "learning_rate": 2.1268918512578933e-06,
+      "loss": 0.7379,
+      "step": 8917
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 14.50540828704834,
+      "learning_rate": 2.1248872406535033e-06,
+      "loss": 1.2645,
+      "step": 8918
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 55.6925163269043,
+      "learning_rate": 2.122882630049113e-06,
+      "loss": 1.05,
+      "step": 8919
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 8.958430290222168,
+      "learning_rate": 2.120878019444723e-06,
+      "loss": 0.9973,
+      "step": 8920
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 12.112589836120605,
+      "learning_rate": 2.1188734088403328e-06,
+      "loss": 0.5669,
+      "step": 8921
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 9.585906982421875,
+      "learning_rate": 2.116868798235943e-06,
+      "loss": 1.455,
+      "step": 8922
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 10.292905807495117,
+      "learning_rate": 2.114864187631553e-06,
+      "loss": 0.822,
+      "step": 8923
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 12.411725997924805,
+      "learning_rate": 2.1128595770271626e-06,
+      "loss": 0.5915,
+      "step": 8924
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 95.50955200195312,
+      "learning_rate": 2.1108549664227727e-06,
+      "loss": 1.2842,
+      "step": 8925
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 15.629396438598633,
+      "learning_rate": 2.1088503558183823e-06,
+      "loss": 1.6067,
+      "step": 8926
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 15.7233304977417,
+      "learning_rate": 2.1068457452139924e-06,
+      "loss": 0.9173,
+      "step": 8927
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 30.911314010620117,
+      "learning_rate": 2.1048411346096025e-06,
+      "loss": 2.3595,
+      "step": 8928
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 12.196179389953613,
+      "learning_rate": 2.102836524005212e-06,
+      "loss": 1.0742,
+      "step": 8929
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 13.497405052185059,
+      "learning_rate": 2.1008319134008222e-06,
+      "loss": 1.1664,
+      "step": 8930
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 14.204902648925781,
+      "learning_rate": 2.098827302796432e-06,
+      "loss": 1.0476,
+      "step": 8931
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 21.779407501220703,
+      "learning_rate": 2.0968226921920416e-06,
+      "loss": 1.0798,
+      "step": 8932
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 30.279451370239258,
+      "learning_rate": 2.0948180815876516e-06,
+      "loss": 1.2667,
+      "step": 8933
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 13.741867065429688,
+      "learning_rate": 2.0928134709832617e-06,
+      "loss": 0.93,
+      "step": 8934
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 9.646475791931152,
+      "learning_rate": 2.0908088603788714e-06,
+      "loss": 1.304,
+      "step": 8935
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 13.321015357971191,
+      "learning_rate": 2.0888042497744815e-06,
+      "loss": 1.1108,
+      "step": 8936
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 15.75296688079834,
+      "learning_rate": 2.0867996391700916e-06,
+      "loss": 0.9929,
+      "step": 8937
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 24.70446014404297,
+      "learning_rate": 2.0847950285657012e-06,
+      "loss": 1.2706,
+      "step": 8938
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 13.372381210327148,
+      "learning_rate": 2.0827904179613113e-06,
+      "loss": 0.6046,
+      "step": 8939
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 40.783939361572266,
+      "learning_rate": 2.0807858073569214e-06,
+      "loss": 1.5435,
+      "step": 8940
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 8.127727508544922,
+      "learning_rate": 2.078781196752531e-06,
+      "loss": 0.8914,
+      "step": 8941
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 16.346284866333008,
+      "learning_rate": 2.076776586148141e-06,
+      "loss": 1.139,
+      "step": 8942
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 83.08602142333984,
+      "learning_rate": 2.074771975543751e-06,
+      "loss": 2.625,
+      "step": 8943
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 13.884016036987305,
+      "learning_rate": 2.0727673649393605e-06,
+      "loss": 1.0258,
+      "step": 8944
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 11.885502815246582,
+      "learning_rate": 2.0707627543349705e-06,
+      "loss": 0.8579,
+      "step": 8945
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 9.849443435668945,
+      "learning_rate": 2.06875814373058e-06,
+      "loss": 0.9329,
+      "step": 8946
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 9.109235763549805,
+      "learning_rate": 2.0667535331261903e-06,
+      "loss": 0.9577,
+      "step": 8947
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 10.755121231079102,
+      "learning_rate": 2.0647489225218004e-06,
+      "loss": 0.6667,
+      "step": 8948
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 10.410956382751465,
+      "learning_rate": 2.06274431191741e-06,
+      "loss": 0.4745,
+      "step": 8949
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 10.473151206970215,
+      "learning_rate": 2.06073970131302e-06,
+      "loss": 1.2898,
+      "step": 8950
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 14.992378234863281,
+      "learning_rate": 2.05873509070863e-06,
+      "loss": 1.5699,
+      "step": 8951
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 9.013774871826172,
+      "learning_rate": 2.05673048010424e-06,
+      "loss": 1.1274,
+      "step": 8952
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 18.34583854675293,
+      "learning_rate": 2.05472586949985e-06,
+      "loss": 1.1243,
+      "step": 8953
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 9.34281063079834,
+      "learning_rate": 2.0527212588954596e-06,
+      "loss": 0.977,
+      "step": 8954
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 19.33155632019043,
+      "learning_rate": 2.0507166482910697e-06,
+      "loss": 0.7991,
+      "step": 8955
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 62.87168502807617,
+      "learning_rate": 2.0487120376866794e-06,
+      "loss": 1.8296,
+      "step": 8956
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 14.451416015625,
+      "learning_rate": 2.0467074270822894e-06,
+      "loss": 0.9752,
+      "step": 8957
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 13.052489280700684,
+      "learning_rate": 2.044702816477899e-06,
+      "loss": 1.0093,
+      "step": 8958
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 10.156307220458984,
+      "learning_rate": 2.042698205873509e-06,
+      "loss": 1.142,
+      "step": 8959
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 32.259788513183594,
+      "learning_rate": 2.0406935952691193e-06,
+      "loss": 2.6409,
+      "step": 8960
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 35.24033737182617,
+      "learning_rate": 2.038688984664729e-06,
+      "loss": 1.7325,
+      "step": 8961
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 15.073776245117188,
+      "learning_rate": 2.036684374060339e-06,
+      "loss": 0.7844,
+      "step": 8962
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 26.254253387451172,
+      "learning_rate": 2.034679763455949e-06,
+      "loss": 0.9144,
+      "step": 8963
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 10.655972480773926,
+      "learning_rate": 2.0326751528515588e-06,
+      "loss": 1.5722,
+      "step": 8964
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 15.8784761428833,
+      "learning_rate": 2.030670542247169e-06,
+      "loss": 1.5385,
+      "step": 8965
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 75.25648498535156,
+      "learning_rate": 2.0286659316427785e-06,
+      "loss": 1.2864,
+      "step": 8966
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 12.602245330810547,
+      "learning_rate": 2.0266613210383886e-06,
+      "loss": 1.6121,
+      "step": 8967
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 26.045513153076172,
+      "learning_rate": 2.0246567104339982e-06,
+      "loss": 1.1818,
+      "step": 8968
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 15.361552238464355,
+      "learning_rate": 2.0226520998296083e-06,
+      "loss": 0.8867,
+      "step": 8969
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 52.4283561706543,
+      "learning_rate": 2.020647489225218e-06,
+      "loss": 2.3197,
+      "step": 8970
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 10.277100563049316,
+      "learning_rate": 2.018642878620828e-06,
+      "loss": 1.2036,
+      "step": 8971
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 56.391624450683594,
+      "learning_rate": 2.0166382680164377e-06,
+      "loss": 1.6374,
+      "step": 8972
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 16.56217384338379,
+      "learning_rate": 2.014633657412048e-06,
+      "loss": 1.2358,
+      "step": 8973
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 30.0115966796875,
+      "learning_rate": 2.012629046807658e-06,
+      "loss": 1.2294,
+      "step": 8974
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 10.70960807800293,
+      "learning_rate": 2.0106244362032676e-06,
+      "loss": 1.2672,
+      "step": 8975
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 12.158578872680664,
+      "learning_rate": 2.0086198255988776e-06,
+      "loss": 0.794,
+      "step": 8976
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 6.594705104827881,
+      "learning_rate": 2.0066152149944877e-06,
+      "loss": 0.964,
+      "step": 8977
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 18.69150161743164,
+      "learning_rate": 2.0046106043900974e-06,
+      "loss": 1.1616,
+      "step": 8978
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 19.576065063476562,
+      "learning_rate": 2.0026059937857075e-06,
+      "loss": 0.9053,
+      "step": 8979
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 30.62327003479004,
+      "learning_rate": 2.000601383181317e-06,
+      "loss": 1.0141,
+      "step": 8980
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 16.1602840423584,
+      "learning_rate": 1.9985967725769272e-06,
+      "loss": 1.0071,
+      "step": 8981
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 16.035024642944336,
+      "learning_rate": 1.996592161972537e-06,
+      "loss": 1.9886,
+      "step": 8982
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 32.672157287597656,
+      "learning_rate": 1.994587551368147e-06,
+      "loss": 1.9142,
+      "step": 8983
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 34.7219352722168,
+      "learning_rate": 1.9925829407637566e-06,
+      "loss": 1.6279,
+      "step": 8984
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 38.268131256103516,
+      "learning_rate": 1.9905783301593667e-06,
+      "loss": 1.3426,
+      "step": 8985
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 10.52354621887207,
+      "learning_rate": 1.988573719554977e-06,
+      "loss": 1.3539,
+      "step": 8986
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 19.19072914123535,
+      "learning_rate": 1.9865691089505865e-06,
+      "loss": 1.7263,
+      "step": 8987
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 35.17476272583008,
+      "learning_rate": 1.9845644983461965e-06,
+      "loss": 1.2377,
+      "step": 8988
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 23.662670135498047,
+      "learning_rate": 1.982559887741806e-06,
+      "loss": 0.9037,
+      "step": 8989
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 25.2396297454834,
+      "learning_rate": 1.9805552771374163e-06,
+      "loss": 1.8102,
+      "step": 8990
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 8.776323318481445,
+      "learning_rate": 1.9785506665330264e-06,
+      "loss": 0.7533,
+      "step": 8991
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 16.68621253967285,
+      "learning_rate": 1.976546055928636e-06,
+      "loss": 1.3865,
+      "step": 8992
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 17.21820640563965,
+      "learning_rate": 1.974541445324246e-06,
+      "loss": 1.0538,
+      "step": 8993
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 13.057286262512207,
+      "learning_rate": 1.9725368347198558e-06,
+      "loss": 1.1173,
+      "step": 8994
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 16.948637008666992,
+      "learning_rate": 1.9705322241154654e-06,
+      "loss": 0.5743,
+      "step": 8995
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 11.57693099975586,
+      "learning_rate": 1.9685276135110755e-06,
+      "loss": 1.2016,
+      "step": 8996
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 25.845434188842773,
+      "learning_rate": 1.9665230029066856e-06,
+      "loss": 1.6141,
+      "step": 8997
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 11.109925270080566,
+      "learning_rate": 1.9645183923022953e-06,
+      "loss": 1.2974,
+      "step": 8998
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 17.424144744873047,
+      "learning_rate": 1.9625137816979054e-06,
+      "loss": 1.0136,
+      "step": 8999
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 13.590910911560059,
+      "learning_rate": 1.9605091710935154e-06,
+      "loss": 2.0274,
+      "step": 9000
+    },
+    {
+      "epoch": 2.71,
+      "eval_loss": 0.16217102110385895,
+      "eval_runtime": 43.386,
+      "eval_samples_per_second": 34.089,
+      "eval_steps_per_second": 34.089,
+      "step": 9000
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 16.488697052001953,
+      "learning_rate": 1.958504560489125e-06,
+      "loss": 1.3625,
+      "step": 9001
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 9.223793983459473,
+      "learning_rate": 1.956499949884735e-06,
+      "loss": 1.2811,
+      "step": 9002
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 13.224336624145508,
+      "learning_rate": 1.9544953392803453e-06,
+      "loss": 1.5226,
+      "step": 9003
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 39.96490478515625,
+      "learning_rate": 1.952490728675955e-06,
+      "loss": 0.9551,
+      "step": 9004
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 8.030670166015625,
+      "learning_rate": 1.9504861180715646e-06,
+      "loss": 0.5237,
+      "step": 9005
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 45.17795944213867,
+      "learning_rate": 1.9484815074671747e-06,
+      "loss": 1.1533,
+      "step": 9006
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 13.118334770202637,
+      "learning_rate": 1.9464768968627843e-06,
+      "loss": 1.3035,
+      "step": 9007
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 13.178396224975586,
+      "learning_rate": 1.9444722862583944e-06,
+      "loss": 1.2787,
+      "step": 9008
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 12.746395111083984,
+      "learning_rate": 1.942467675654004e-06,
+      "loss": 1.3914,
+      "step": 9009
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 25.33348274230957,
+      "learning_rate": 1.940463065049614e-06,
+      "loss": 1.2344,
+      "step": 9010
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 18.61649513244629,
+      "learning_rate": 1.9384584544452242e-06,
+      "loss": 0.9396,
+      "step": 9011
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 98.35498809814453,
+      "learning_rate": 1.936453843840834e-06,
+      "loss": 2.0006,
+      "step": 9012
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 12.018935203552246,
+      "learning_rate": 1.934449233236444e-06,
+      "loss": 1.1039,
+      "step": 9013
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 18.714567184448242,
+      "learning_rate": 1.932444622632054e-06,
+      "loss": 1.2751,
+      "step": 9014
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 41.53782653808594,
+      "learning_rate": 1.9304400120276637e-06,
+      "loss": 1.7534,
+      "step": 9015
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 37.50891876220703,
+      "learning_rate": 1.928435401423274e-06,
+      "loss": 1.1007,
+      "step": 9016
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 14.983277320861816,
+      "learning_rate": 1.9264307908188835e-06,
+      "loss": 1.6123,
+      "step": 9017
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 13.673733711242676,
+      "learning_rate": 1.9244261802144936e-06,
+      "loss": 1.0661,
+      "step": 9018
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 9.97967529296875,
+      "learning_rate": 1.9224215696101032e-06,
+      "loss": 1.5572,
+      "step": 9019
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 12.354124069213867,
+      "learning_rate": 1.9204169590057133e-06,
+      "loss": 1.2915,
+      "step": 9020
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 11.94011116027832,
+      "learning_rate": 1.918412348401323e-06,
+      "loss": 1.3119,
+      "step": 9021
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 11.489762306213379,
+      "learning_rate": 1.916407737796933e-06,
+      "loss": 1.1241,
+      "step": 9022
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 26.69053840637207,
+      "learning_rate": 1.914403127192543e-06,
+      "loss": 1.3275,
+      "step": 9023
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 15.625397682189941,
+      "learning_rate": 1.912398516588153e-06,
+      "loss": 1.2162,
+      "step": 9024
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 23.008207321166992,
+      "learning_rate": 1.910393905983763e-06,
+      "loss": 1.4855,
+      "step": 9025
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 122.37059783935547,
+      "learning_rate": 1.908389295379373e-06,
+      "loss": 1.5003,
+      "step": 9026
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 47.32773208618164,
+      "learning_rate": 1.9063846847749826e-06,
+      "loss": 2.3885,
+      "step": 9027
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 10.500248908996582,
+      "learning_rate": 1.9043800741705925e-06,
+      "loss": 1.3469,
+      "step": 9028
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 12.691232681274414,
+      "learning_rate": 1.9023754635662024e-06,
+      "loss": 1.5982,
+      "step": 9029
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 15.978590965270996,
+      "learning_rate": 1.9003708529618122e-06,
+      "loss": 0.9491,
+      "step": 9030
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 22.821592330932617,
+      "learning_rate": 1.8983662423574223e-06,
+      "loss": 1.3128,
+      "step": 9031
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 33.413455963134766,
+      "learning_rate": 1.896361631753032e-06,
+      "loss": 2.5185,
+      "step": 9032
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 14.448841094970703,
+      "learning_rate": 1.894357021148642e-06,
+      "loss": 1.7291,
+      "step": 9033
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 17.098773956298828,
+      "learning_rate": 1.892352410544252e-06,
+      "loss": 0.7803,
+      "step": 9034
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 22.17733383178711,
+      "learning_rate": 1.8903477999398618e-06,
+      "loss": 2.054,
+      "step": 9035
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 18.0053768157959,
+      "learning_rate": 1.8883431893354717e-06,
+      "loss": 0.9891,
+      "step": 9036
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 7.076298713684082,
+      "learning_rate": 1.8863385787310818e-06,
+      "loss": 0.6043,
+      "step": 9037
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 10.31680965423584,
+      "learning_rate": 1.8843339681266914e-06,
+      "loss": 0.6896,
+      "step": 9038
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 34.10591125488281,
+      "learning_rate": 1.8823293575223015e-06,
+      "loss": 1.6691,
+      "step": 9039
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 19.84099006652832,
+      "learning_rate": 1.8803247469179114e-06,
+      "loss": 1.7946,
+      "step": 9040
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 16.761245727539062,
+      "learning_rate": 1.8783201363135213e-06,
+      "loss": 1.7972,
+      "step": 9041
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 19.77780532836914,
+      "learning_rate": 1.8763155257091311e-06,
+      "loss": 1.0711,
+      "step": 9042
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 12.905445098876953,
+      "learning_rate": 1.8743109151047412e-06,
+      "loss": 0.79,
+      "step": 9043
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 34.32182312011719,
+      "learning_rate": 1.8723063045003509e-06,
+      "loss": 1.5714,
+      "step": 9044
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 15.384803771972656,
+      "learning_rate": 1.870301693895961e-06,
+      "loss": 0.9761,
+      "step": 9045
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 10.162015914916992,
+      "learning_rate": 1.8682970832915708e-06,
+      "loss": 1.0334,
+      "step": 9046
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 21.561569213867188,
+      "learning_rate": 1.8662924726871807e-06,
+      "loss": 0.7688,
+      "step": 9047
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 8.59749698638916,
+      "learning_rate": 1.8642878620827906e-06,
+      "loss": 0.533,
+      "step": 9048
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 17.311405181884766,
+      "learning_rate": 1.8622832514784007e-06,
+      "loss": 0.9795,
+      "step": 9049
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 67.3037109375,
+      "learning_rate": 1.8602786408740103e-06,
+      "loss": 2.6467,
+      "step": 9050
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 12.96740436553955,
+      "learning_rate": 1.8582740302696204e-06,
+      "loss": 1.1722,
+      "step": 9051
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 16.440349578857422,
+      "learning_rate": 1.85626941966523e-06,
+      "loss": 0.6379,
+      "step": 9052
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 35.48670196533203,
+      "learning_rate": 1.85426480906084e-06,
+      "loss": 1.4359,
+      "step": 9053
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 55.38397216796875,
+      "learning_rate": 1.85226019845645e-06,
+      "loss": 2.2976,
+      "step": 9054
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 20.49527931213379,
+      "learning_rate": 1.8502555878520597e-06,
+      "loss": 1.1748,
+      "step": 9055
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 11.808248519897461,
+      "learning_rate": 1.8482509772476698e-06,
+      "loss": 1.6608,
+      "step": 9056
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 25.69436264038086,
+      "learning_rate": 1.8462463666432799e-06,
+      "loss": 1.0124,
+      "step": 9057
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 27.099803924560547,
+      "learning_rate": 1.8442417560388895e-06,
+      "loss": 1.4334,
+      "step": 9058
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 16.551921844482422,
+      "learning_rate": 1.8422371454344994e-06,
+      "loss": 1.4909,
+      "step": 9059
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 14.500848770141602,
+      "learning_rate": 1.8402325348301095e-06,
+      "loss": 1.1442,
+      "step": 9060
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 14.154791831970215,
+      "learning_rate": 1.8382279242257191e-06,
+      "loss": 1.8214,
+      "step": 9061
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 28.4224796295166,
+      "learning_rate": 1.8362233136213292e-06,
+      "loss": 1.2986,
+      "step": 9062
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 14.508337020874023,
+      "learning_rate": 1.8342187030169393e-06,
+      "loss": 1.1846,
+      "step": 9063
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 17.251667022705078,
+      "learning_rate": 1.832214092412549e-06,
+      "loss": 1.2499,
+      "step": 9064
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 37.404449462890625,
+      "learning_rate": 1.8302094818081588e-06,
+      "loss": 0.9916,
+      "step": 9065
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 13.81652545928955,
+      "learning_rate": 1.828204871203769e-06,
+      "loss": 1.2755,
+      "step": 9066
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 21.725704193115234,
+      "learning_rate": 1.8262002605993786e-06,
+      "loss": 1.1399,
+      "step": 9067
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 10.298942565917969,
+      "learning_rate": 1.8241956499949887e-06,
+      "loss": 0.7941,
+      "step": 9068
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 25.29086685180664,
+      "learning_rate": 1.8221910393905988e-06,
+      "loss": 1.3804,
+      "step": 9069
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 18.70489501953125,
+      "learning_rate": 1.8201864287862084e-06,
+      "loss": 1.1317,
+      "step": 9070
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 11.406014442443848,
+      "learning_rate": 1.8181818181818183e-06,
+      "loss": 0.4713,
+      "step": 9071
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 22.935609817504883,
+      "learning_rate": 1.8161772075774282e-06,
+      "loss": 1.6272,
+      "step": 9072
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 33.32746124267578,
+      "learning_rate": 1.814172596973038e-06,
+      "loss": 1.3237,
+      "step": 9073
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 8.96728801727295,
+      "learning_rate": 1.8121679863686481e-06,
+      "loss": 0.556,
+      "step": 9074
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 23.04669761657715,
+      "learning_rate": 1.8101633757642578e-06,
+      "loss": 1.5039,
+      "step": 9075
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 10.999143600463867,
+      "learning_rate": 1.8081587651598679e-06,
+      "loss": 1.7855,
+      "step": 9076
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 21.91204261779785,
+      "learning_rate": 1.8061541545554777e-06,
+      "loss": 1.4705,
+      "step": 9077
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 31.69646644592285,
+      "learning_rate": 1.8041495439510876e-06,
+      "loss": 1.5004,
+      "step": 9078
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 8.526796340942383,
+      "learning_rate": 1.8021449333466975e-06,
+      "loss": 0.8363,
+      "step": 9079
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 91.82820129394531,
+      "learning_rate": 1.8001403227423076e-06,
+      "loss": 2.5966,
+      "step": 9080
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 62.776432037353516,
+      "learning_rate": 1.7981357121379172e-06,
+      "loss": 2.0551,
+      "step": 9081
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 7.502096652984619,
+      "learning_rate": 1.7961311015335273e-06,
+      "loss": 0.9674,
+      "step": 9082
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 13.519929885864258,
+      "learning_rate": 1.7941264909291372e-06,
+      "loss": 0.9669,
+      "step": 9083
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 18.463733673095703,
+      "learning_rate": 1.792121880324747e-06,
+      "loss": 1.557,
+      "step": 9084
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 18.185571670532227,
+      "learning_rate": 1.790117269720357e-06,
+      "loss": 1.9059,
+      "step": 9085
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 11.766922950744629,
+      "learning_rate": 1.788112659115967e-06,
+      "loss": 1.3468,
+      "step": 9086
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 22.52054786682129,
+      "learning_rate": 1.7861080485115767e-06,
+      "loss": 0.9551,
+      "step": 9087
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 8.950528144836426,
+      "learning_rate": 1.7841034379071868e-06,
+      "loss": 0.6039,
+      "step": 9088
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 19.103931427001953,
+      "learning_rate": 1.7820988273027966e-06,
+      "loss": 1.0202,
+      "step": 9089
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 14.955410957336426,
+      "learning_rate": 1.7800942166984065e-06,
+      "loss": 1.4588,
+      "step": 9090
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 49.647361755371094,
+      "learning_rate": 1.7780896060940164e-06,
+      "loss": 3.6037,
+      "step": 9091
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 111.04601287841797,
+      "learning_rate": 1.7760849954896263e-06,
+      "loss": 1.5603,
+      "step": 9092
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 49.5386962890625,
+      "learning_rate": 1.7740803848852361e-06,
+      "loss": 2.5773,
+      "step": 9093
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 12.374490737915039,
+      "learning_rate": 1.7720757742808462e-06,
+      "loss": 1.1015,
+      "step": 9094
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 9.451532363891602,
+      "learning_rate": 1.7700711636764559e-06,
+      "loss": 0.4628,
+      "step": 9095
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 13.187180519104004,
+      "learning_rate": 1.768066553072066e-06,
+      "loss": 0.8442,
+      "step": 9096
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 12.320247650146484,
+      "learning_rate": 1.7660619424676758e-06,
+      "loss": 0.88,
+      "step": 9097
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 8.388055801391602,
+      "learning_rate": 1.7640573318632857e-06,
+      "loss": 0.7241,
+      "step": 9098
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 11.43036937713623,
+      "learning_rate": 1.7620527212588956e-06,
+      "loss": 0.8333,
+      "step": 9099
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 8.495409965515137,
+      "learning_rate": 1.7600481106545057e-06,
+      "loss": 1.1459,
+      "step": 9100
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 11.421164512634277,
+      "learning_rate": 1.7580435000501153e-06,
+      "loss": 1.0815,
+      "step": 9101
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 17.927734375,
+      "learning_rate": 1.7560388894457254e-06,
+      "loss": 1.1274,
+      "step": 9102
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 22.090795516967773,
+      "learning_rate": 1.7540342788413353e-06,
+      "loss": 0.8337,
+      "step": 9103
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 11.036092758178711,
+      "learning_rate": 1.752029668236945e-06,
+      "loss": 1.3825,
+      "step": 9104
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 10.002640724182129,
+      "learning_rate": 1.750025057632555e-06,
+      "loss": 1.174,
+      "step": 9105
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 13.688326835632324,
+      "learning_rate": 1.748020447028165e-06,
+      "loss": 1.0106,
+      "step": 9106
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 18.911972045898438,
+      "learning_rate": 1.7460158364237748e-06,
+      "loss": 1.3007,
+      "step": 9107
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 28.944271087646484,
+      "learning_rate": 1.7440112258193848e-06,
+      "loss": 1.5335,
+      "step": 9108
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 17.66625213623047,
+      "learning_rate": 1.7420066152149947e-06,
+      "loss": 0.7189,
+      "step": 9109
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 28.595500946044922,
+      "learning_rate": 1.7400020046106044e-06,
+      "loss": 1.4251,
+      "step": 9110
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 11.355489730834961,
+      "learning_rate": 1.7379973940062145e-06,
+      "loss": 0.8031,
+      "step": 9111
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 20.97258949279785,
+      "learning_rate": 1.7359927834018241e-06,
+      "loss": 0.8911,
+      "step": 9112
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 21.422672271728516,
+      "learning_rate": 1.7339881727974342e-06,
+      "loss": 1.5909,
+      "step": 9113
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 16.778446197509766,
+      "learning_rate": 1.7319835621930443e-06,
+      "loss": 0.5871,
+      "step": 9114
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 14.981059074401855,
+      "learning_rate": 1.729978951588654e-06,
+      "loss": 0.938,
+      "step": 9115
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 17.02187156677246,
+      "learning_rate": 1.7279743409842638e-06,
+      "loss": 1.4252,
+      "step": 9116
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 20.162086486816406,
+      "learning_rate": 1.725969730379874e-06,
+      "loss": 0.9187,
+      "step": 9117
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 22.232948303222656,
+      "learning_rate": 1.7239651197754836e-06,
+      "loss": 1.1912,
+      "step": 9118
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 21.19387435913086,
+      "learning_rate": 1.7219605091710937e-06,
+      "loss": 0.7006,
+      "step": 9119
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 11.553994178771973,
+      "learning_rate": 1.7199558985667037e-06,
+      "loss": 0.7409,
+      "step": 9120
+    },
+    {
+      "epoch": 2.74,
+      "eval_loss": 0.16245336830615997,
+      "eval_runtime": 43.6231,
+      "eval_samples_per_second": 33.904,
+      "eval_steps_per_second": 33.904,
+      "step": 9120
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 9978,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 120,
+  "total_flos": 435510745758720.0,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}