diff --git "a/checkpoint-534240/trainer_state.json" "b/checkpoint-534240/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-534240/trainer_state.json"
@@ -0,0 +1,6339 @@
+{
+  "best_metric": 4.133681774139404,
+  "best_model_checkpoint": "/mmfs1/gscratch/stf/abhinavp/corpus-filtering/outputs/binding-case/lstm/0/checkpoints/checkpoint-534240",
+  "epoch": 1.0250006060157382,
+  "eval_steps": 10,
+  "global_step": 534240,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.999998362119627e-05,
+      "loss": 10.8206,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.999161405248948e-05,
+      "loss": 7.5557,
+      "step": 512
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.998322810497896e-05,
+      "loss": 7.0612,
+      "step": 1024
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.997484215746844e-05,
+      "loss": 6.9955,
+      "step": 1536
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.996645620995792e-05,
+      "loss": 6.949,
+      "step": 2048
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.99580702624474e-05,
+      "loss": 6.9206,
+      "step": 2560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.994968431493688e-05,
+      "loss": 6.7551,
+      "step": 3072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.994129836742636e-05,
+      "loss": 6.6436,
+      "step": 3584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.993291241991584e-05,
+      "loss": 6.5449,
+      "step": 4096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.992452647240532e-05,
+      "loss": 6.4714,
+      "step": 4608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.991615690369854e-05,
+      "loss": 6.3994,
+      "step": 5120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.990777095618801e-05,
+      "loss": 6.3319,
+      "step": 5632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.989938500867749e-05,
+      "loss": 6.2619,
+      "step": 6144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.989099906116697e-05,
+      "loss": 6.1949,
+      "step": 6656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.988261311365645e-05,
+      "loss": 6.1395,
+      "step": 7168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.987422716614593e-05,
+      "loss": 6.0836,
+      "step": 7680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.986584121863541e-05,
+      "loss": 6.0391,
+      "step": 8192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.985745527112489e-05,
+      "loss": 5.9938,
+      "step": 8704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.984906932361437e-05,
+      "loss": 5.9539,
+      "step": 9216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.984068337610385e-05,
+      "loss": 5.9146,
+      "step": 9728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.983229742859333e-05,
+      "loss": 5.8783,
+      "step": 10240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.982391148108281e-05,
+      "loss": 5.8402,
+      "step": 10752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.981552553357229e-05,
+      "loss": 5.8076,
+      "step": 11264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.980713958606178e-05,
+      "loss": 5.7774,
+      "step": 11776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.979875363855125e-05,
+      "loss": 5.7584,
+      "step": 12288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.979036769104073e-05,
+      "loss": 5.7174,
+      "step": 12800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9781998122333946e-05,
+      "loss": 5.6961,
+      "step": 13312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9773612174823426e-05,
+      "loss": 5.6694,
+      "step": 13824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.97652262273129e-05,
+      "loss": 5.6503,
+      "step": 14336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.975684027980238e-05,
+      "loss": 5.623,
+      "step": 14848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9748470711095595e-05,
+      "loss": 5.5988,
+      "step": 15360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.974008476358507e-05,
+      "loss": 5.5924,
+      "step": 15872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.973169881607455e-05,
+      "loss": 5.566,
+      "step": 16384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9723329247367764e-05,
+      "loss": 5.5574,
+      "step": 16896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9714943299857244e-05,
+      "loss": 5.5424,
+      "step": 17408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9706557352346724e-05,
+      "loss": 5.5176,
+      "step": 17920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.969818778363994e-05,
+      "loss": 5.508,
+      "step": 18432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.968980183612942e-05,
+      "loss": 5.4698,
+      "step": 18944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.96814158886189e-05,
+      "loss": 5.4702,
+      "step": 19456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.967302994110837e-05,
+      "loss": 5.4358,
+      "step": 19968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.966464399359785e-05,
+      "loss": 5.4358,
+      "step": 20480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.965625804608733e-05,
+      "loss": 5.418,
+      "step": 20992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.964787209857681e-05,
+      "loss": 5.4185,
+      "step": 21504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.963948615106629e-05,
+      "loss": 5.3915,
+      "step": 22016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.96311165823595e-05,
+      "loss": 5.3852,
+      "step": 22528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.962273063484898e-05,
+      "loss": 5.3764,
+      "step": 23040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.961434468733847e-05,
+      "loss": 5.3599,
+      "step": 23552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.960595873982795e-05,
+      "loss": 5.3653,
+      "step": 24064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.959757279231743e-05,
+      "loss": 5.3357,
+      "step": 24576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.958920322361064e-05,
+      "loss": 5.3203,
+      "step": 25088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.958081727610012e-05,
+      "loss": 5.3299,
+      "step": 25600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.957244770739333e-05,
+      "loss": 5.3239,
+      "step": 26112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.956407813868654e-05,
+      "loss": 5.2986,
+      "step": 26624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9555692191176016e-05,
+      "loss": 5.2878,
+      "step": 27136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9547306243665496e-05,
+      "loss": 5.2856,
+      "step": 27648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9538920296154976e-05,
+      "loss": 5.271,
+      "step": 28160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9530534348644456e-05,
+      "loss": 5.2853,
+      "step": 28672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9522148401133936e-05,
+      "loss": 5.2463,
+      "step": 29184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.951376245362342e-05,
+      "loss": 5.2454,
+      "step": 29696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.95053765061129e-05,
+      "loss": 5.2377,
+      "step": 30208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.949699055860238e-05,
+      "loss": 5.2269,
+      "step": 30720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.948860461109186e-05,
+      "loss": 5.2277,
+      "step": 31232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.948021866358134e-05,
+      "loss": 5.2061,
+      "step": 31744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.947183271607082e-05,
+      "loss": 5.2001,
+      "step": 32256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.94634467685603e-05,
+      "loss": 5.192,
+      "step": 32768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.945507719985351e-05,
+      "loss": 5.2067,
+      "step": 33280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.944669125234299e-05,
+      "loss": 5.1779,
+      "step": 33792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.943830530483247e-05,
+      "loss": 5.1785,
+      "step": 34304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.942991935732195e-05,
+      "loss": 5.1512,
+      "step": 34816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.942154978861516e-05,
+      "loss": 5.1598,
+      "step": 35328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.941316384110464e-05,
+      "loss": 5.167,
+      "step": 35840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.940477789359412e-05,
+      "loss": 5.1598,
+      "step": 36352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.93963919460836e-05,
+      "loss": 5.1372,
+      "step": 36864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.938800599857309e-05,
+      "loss": 5.1505,
+      "step": 37376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.937962005106256e-05,
+      "loss": 5.1458,
+      "step": 37888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.937123410355204e-05,
+      "loss": 5.1229,
+      "step": 38400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9362864534845256e-05,
+      "loss": 5.12,
+      "step": 38912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.935447858733473e-05,
+      "loss": 5.1027,
+      "step": 39424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.934609263982421e-05,
+      "loss": 5.089,
+      "step": 39936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.933770669231369e-05,
+      "loss": 5.0878,
+      "step": 40448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9329337123606905e-05,
+      "loss": 5.0921,
+      "step": 40960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.932095117609638e-05,
+      "loss": 5.0903,
+      "step": 41472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.931256522858586e-05,
+      "loss": 5.096,
+      "step": 41984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.930417928107534e-05,
+      "loss": 5.0813,
+      "step": 42496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9295809712368554e-05,
+      "loss": 5.0617,
+      "step": 43008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9287423764858034e-05,
+      "loss": 5.0622,
+      "step": 43520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9279037817347514e-05,
+      "loss": 5.0622,
+      "step": 44032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9270651869836994e-05,
+      "loss": 5.0509,
+      "step": 44544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9262265922326474e-05,
+      "loss": 5.0423,
+      "step": 45056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.925389635361968e-05,
+      "loss": 5.0321,
+      "step": 45568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.92455267849129e-05,
+      "loss": 5.0325,
+      "step": 46080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.923714083740238e-05,
+      "loss": 5.0333,
+      "step": 46592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.922875488989185e-05,
+      "loss": 5.0206,
+      "step": 47104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.922036894238133e-05,
+      "loss": 5.0233,
+      "step": 47616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.921198299487081e-05,
+      "loss": 5.0122,
+      "step": 48128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.920361342616403e-05,
+      "loss": 5.0066,
+      "step": 48640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.919522747865351e-05,
+      "loss": 4.9943,
+      "step": 49152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.918684153114299e-05,
+      "loss": 4.9911,
+      "step": 49664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.917845558363247e-05,
+      "loss": 4.9846,
+      "step": 50176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.917006963612195e-05,
+      "loss": 4.9879,
+      "step": 50688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9161700067415156e-05,
+      "loss": 4.9896,
+      "step": 51200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9153314119904636e-05,
+      "loss": 4.967,
+      "step": 51712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.914494455119785e-05,
+      "loss": 4.9728,
+      "step": 52224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9136558603687325e-05,
+      "loss": 4.9613,
+      "step": 52736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9128172656176805e-05,
+      "loss": 4.9586,
+      "step": 53248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9119786708666285e-05,
+      "loss": 4.9504,
+      "step": 53760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9111400761155765e-05,
+      "loss": 4.9449,
+      "step": 54272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9103014813645245e-05,
+      "loss": 4.9452,
+      "step": 54784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.909462886613473e-05,
+      "loss": 4.9367,
+      "step": 55296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.908624291862421e-05,
+      "loss": 4.934,
+      "step": 55808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.907785697111369e-05,
+      "loss": 4.9274,
+      "step": 56320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.906947102360317e-05,
+      "loss": 4.9177,
+      "step": 56832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.906108507609265e-05,
+      "loss": 4.9292,
+      "step": 57344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.905269912858213e-05,
+      "loss": 4.9178,
+      "step": 57856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.904432955987534e-05,
+      "loss": 4.9181,
+      "step": 58368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.903594361236482e-05,
+      "loss": 4.9134,
+      "step": 58880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.90275576648543e-05,
+      "loss": 4.9144,
+      "step": 59392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.901917171734378e-05,
+      "loss": 4.9042,
+      "step": 59904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.901078576983326e-05,
+      "loss": 4.8909,
+      "step": 60416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.900241620112647e-05,
+      "loss": 4.8986,
+      "step": 60928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.899403025361595e-05,
+      "loss": 4.9034,
+      "step": 61440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.898564430610543e-05,
+      "loss": 4.8889,
+      "step": 61952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.897725835859491e-05,
+      "loss": 4.8826,
+      "step": 62464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8968872411084396e-05,
+      "loss": 4.8807,
+      "step": 62976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.896048646357387e-05,
+      "loss": 4.8775,
+      "step": 63488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8952116894867085e-05,
+      "loss": 4.8676,
+      "step": 64000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8943747326160294e-05,
+      "loss": 4.8714,
+      "step": 64512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8935361378649774e-05,
+      "loss": 4.8604,
+      "step": 65024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8926975431139254e-05,
+      "loss": 4.8629,
+      "step": 65536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8918589483628734e-05,
+      "loss": 4.8533,
+      "step": 66048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8910203536118214e-05,
+      "loss": 4.8705,
+      "step": 66560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.890181758860769e-05,
+      "loss": 4.8545,
+      "step": 67072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.889343164109717e-05,
+      "loss": 4.8449,
+      "step": 67584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.888506207239038e-05,
+      "loss": 4.8412,
+      "step": 68096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.887667612487986e-05,
+      "loss": 4.8422,
+      "step": 68608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.886830655617308e-05,
+      "loss": 4.8429,
+      "step": 69120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.885992060866256e-05,
+      "loss": 4.843,
+      "step": 69632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.885153466115204e-05,
+      "loss": 4.8318,
+      "step": 70144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.884314871364151e-05,
+      "loss": 4.8329,
+      "step": 70656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.883476276613099e-05,
+      "loss": 4.828,
+      "step": 71168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.882637681862047e-05,
+      "loss": 4.8262,
+      "step": 71680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.881800724991369e-05,
+      "loss": 4.8149,
+      "step": 72192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.880962130240316e-05,
+      "loss": 4.825,
+      "step": 72704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.880123535489264e-05,
+      "loss": 4.813,
+      "step": 73216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.879284940738212e-05,
+      "loss": 4.7998,
+      "step": 73728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.87844634598716e-05,
+      "loss": 4.8142,
+      "step": 74240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.877607751236109e-05,
+      "loss": 4.8032,
+      "step": 74752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.876769156485057e-05,
+      "loss": 4.8036,
+      "step": 75264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.875930561734005e-05,
+      "loss": 4.8102,
+      "step": 75776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.875091966982953e-05,
+      "loss": 4.7895,
+      "step": 76288
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.757783889770508,
+      "eval_runtime": 371.7114,
+      "eval_samples_per_second": 1026.579,
+      "eval_steps_per_second": 32.081,
+      "step": 76320
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.874253372231901e-05,
+      "loss": 4.7832,
+      "step": 76800
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.873414777480849e-05,
+      "loss": 4.7856,
+      "step": 77312
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.872576182729797e-05,
+      "loss": 4.7881,
+      "step": 77824
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.871737587978745e-05,
+      "loss": 4.7823,
+      "step": 78336
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.870898993227693e-05,
+      "loss": 4.7878,
+      "step": 78848
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8700620363570137e-05,
+      "loss": 4.7744,
+      "step": 79360
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8692234416059616e-05,
+      "loss": 4.7772,
+      "step": 79872
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8683848468549096e-05,
+      "loss": 4.7662,
+      "step": 80384
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8675462521038576e-05,
+      "loss": 4.773,
+      "step": 80896
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8667076573528056e-05,
+      "loss": 4.7613,
+      "step": 81408
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.865870700482127e-05,
+      "loss": 4.7605,
+      "step": 81920
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.865032105731075e-05,
+      "loss": 4.7599,
+      "step": 82432
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8641935109800225e-05,
+      "loss": 4.7528,
+      "step": 82944
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8633549162289705e-05,
+      "loss": 4.7481,
+      "step": 83456
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8625163214779185e-05,
+      "loss": 4.7445,
+      "step": 83968
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.86167936460724e-05,
+      "loss": 4.7333,
+      "step": 84480
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8608407698561874e-05,
+      "loss": 4.7424,
+      "step": 84992
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8600021751051354e-05,
+      "loss": 4.7358,
+      "step": 85504
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.859165218234457e-05,
+      "loss": 4.7304,
+      "step": 86016
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.858326623483405e-05,
+      "loss": 4.7491,
+      "step": 86528
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.857488028732352e-05,
+      "loss": 4.7294,
+      "step": 87040
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.856649433981301e-05,
+      "loss": 4.7343,
+      "step": 87552
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.855810839230249e-05,
+      "loss": 4.726,
+      "step": 88064
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.85497388235957e-05,
+      "loss": 4.7448,
+      "step": 88576
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.854135287608518e-05,
+      "loss": 4.7139,
+      "step": 89088
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.853296692857466e-05,
+      "loss": 4.7151,
+      "step": 89600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.852458098106414e-05,
+      "loss": 4.718,
+      "step": 90112
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.851621141235735e-05,
+      "loss": 4.7197,
+      "step": 90624
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.850782546484683e-05,
+      "loss": 4.7058,
+      "step": 91136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.849943951733631e-05,
+      "loss": 4.7054,
+      "step": 91648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8491069948629524e-05,
+      "loss": 4.7126,
+      "step": 92160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8482684001119e-05,
+      "loss": 4.711,
+      "step": 92672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.847429805360848e-05,
+      "loss": 4.7127,
+      "step": 93184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8465912106097964e-05,
+      "loss": 4.706,
+      "step": 93696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.845754253739117e-05,
+      "loss": 4.6987,
+      "step": 94208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.844915658988065e-05,
+      "loss": 4.7059,
+      "step": 94720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.844077064237013e-05,
+      "loss": 4.6771,
+      "step": 95232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.843238469485961e-05,
+      "loss": 4.6969,
+      "step": 95744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.842399874734909e-05,
+      "loss": 4.673,
+      "step": 96256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.841561279983857e-05,
+      "loss": 4.6867,
+      "step": 96768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.840722685232805e-05,
+      "loss": 4.6774,
+      "step": 97280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.839884090481753e-05,
+      "loss": 4.6886,
+      "step": 97792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.839045495730701e-05,
+      "loss": 4.6774,
+      "step": 98304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.838206900979649e-05,
+      "loss": 4.6739,
+      "step": 98816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.837368306228597e-05,
+      "loss": 4.6764,
+      "step": 99328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.836529711477545e-05,
+      "loss": 4.6693,
+      "step": 99840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.835691116726493e-05,
+      "loss": 4.6837,
+      "step": 100352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.834852521975441e-05,
+      "loss": 4.6573,
+      "step": 100864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.834017202985136e-05,
+      "loss": 4.6511,
+      "step": 101376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.833178608234084e-05,
+      "loss": 4.6721,
+      "step": 101888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.832340013483032e-05,
+      "loss": 4.6714,
+      "step": 102400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.83150141873198e-05,
+      "loss": 4.6566,
+      "step": 102912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.830662823980928e-05,
+      "loss": 4.6494,
+      "step": 103424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.829824229229876e-05,
+      "loss": 4.6569,
+      "step": 103936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.828985634478824e-05,
+      "loss": 4.6435,
+      "step": 104448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.828147039727771e-05,
+      "loss": 4.6669,
+      "step": 104960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8273100828570926e-05,
+      "loss": 4.6411,
+      "step": 105472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8264714881060406e-05,
+      "loss": 4.649,
+      "step": 105984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8256328933549886e-05,
+      "loss": 4.6413,
+      "step": 106496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8247942986039366e-05,
+      "loss": 4.6343,
+      "step": 107008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8239557038528846e-05,
+      "loss": 4.6464,
+      "step": 107520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.823118746982206e-05,
+      "loss": 4.6314,
+      "step": 108032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8222801522311535e-05,
+      "loss": 4.6277,
+      "step": 108544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8214415574801015e-05,
+      "loss": 4.6298,
+      "step": 109056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8206029627290495e-05,
+      "loss": 4.6436,
+      "step": 109568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8197643679779975e-05,
+      "loss": 4.6279,
+      "step": 110080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8189274111073184e-05,
+      "loss": 4.6242,
+      "step": 110592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8180888163562664e-05,
+      "loss": 4.6169,
+      "step": 111104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8172502216052144e-05,
+      "loss": 4.6195,
+      "step": 111616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8164116268541624e-05,
+      "loss": 4.6373,
+      "step": 112128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8155730321031104e-05,
+      "loss": 4.6343,
+      "step": 112640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8147344373520584e-05,
+      "loss": 4.6162,
+      "step": 113152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.81389748048138e-05,
+      "loss": 4.6285,
+      "step": 113664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.813058885730328e-05,
+      "loss": 4.6335,
+      "step": 114176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.812220290979276e-05,
+      "loss": 4.6189,
+      "step": 114688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.811381696228224e-05,
+      "loss": 4.6197,
+      "step": 115200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.810543101477172e-05,
+      "loss": 4.6086,
+      "step": 115712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.80970450672612e-05,
+      "loss": 4.5985,
+      "step": 116224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.808865911975068e-05,
+      "loss": 4.6001,
+      "step": 116736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.808027317224016e-05,
+      "loss": 4.6058,
+      "step": 117248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8071936361140827e-05,
+      "loss": 4.6087,
+      "step": 117760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8063550413630307e-05,
+      "loss": 4.6257,
+      "step": 118272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8055164466119786e-05,
+      "loss": 4.6104,
+      "step": 118784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.804677851860927e-05,
+      "loss": 4.593,
+      "step": 119296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.803839257109875e-05,
+      "loss": 4.6001,
+      "step": 119808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.803000662358823e-05,
+      "loss": 4.604,
+      "step": 120320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.802162067607771e-05,
+      "loss": 4.6004,
+      "step": 120832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.801323472856719e-05,
+      "loss": 4.5924,
+      "step": 121344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.80048651598604e-05,
+      "loss": 4.5902,
+      "step": 121856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.799647921234988e-05,
+      "loss": 4.5892,
+      "step": 122368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.798809326483936e-05,
+      "loss": 4.5952,
+      "step": 122880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.797970731732884e-05,
+      "loss": 4.5855,
+      "step": 123392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.797132136981832e-05,
+      "loss": 4.5906,
+      "step": 123904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.79629354223078e-05,
+      "loss": 4.5851,
+      "step": 124416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.795454947479728e-05,
+      "loss": 4.5851,
+      "step": 124928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.794616352728676e-05,
+      "loss": 4.574,
+      "step": 125440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.793777757977624e-05,
+      "loss": 4.5809,
+      "step": 125952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.792940801106946e-05,
+      "loss": 4.574,
+      "step": 126464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.792102206355894e-05,
+      "loss": 4.5791,
+      "step": 126976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.791263611604842e-05,
+      "loss": 4.5818,
+      "step": 127488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.79042501685379e-05,
+      "loss": 4.5675,
+      "step": 128000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.789586422102737e-05,
+      "loss": 4.5711,
+      "step": 128512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.788747827351685e-05,
+      "loss": 4.5672,
+      "step": 129024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.787909232600633e-05,
+      "loss": 4.5684,
+      "step": 129536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7870722757299546e-05,
+      "loss": 4.5594,
+      "step": 130048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.786233680978902e-05,
+      "loss": 4.5605,
+      "step": 130560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.78539508622785e-05,
+      "loss": 4.5649,
+      "step": 131072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.784556491476798e-05,
+      "loss": 4.5554,
+      "step": 131584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.783717896725746e-05,
+      "loss": 4.5589,
+      "step": 132096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.782879301974694e-05,
+      "loss": 4.5524,
+      "step": 132608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7820423451040155e-05,
+      "loss": 4.5454,
+      "step": 133120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.781205388233337e-05,
+      "loss": 4.5666,
+      "step": 133632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7803667934822844e-05,
+      "loss": 4.5497,
+      "step": 134144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7795281987312324e-05,
+      "loss": 4.5564,
+      "step": 134656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7786896039801804e-05,
+      "loss": 4.5522,
+      "step": 135168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7778510092291284e-05,
+      "loss": 4.5631,
+      "step": 135680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7770124144780764e-05,
+      "loss": 4.552,
+      "step": 136192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7761738197270244e-05,
+      "loss": 4.5388,
+      "step": 136704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.775336862856345e-05,
+      "loss": 4.5504,
+      "step": 137216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.774498268105293e-05,
+      "loss": 4.554,
+      "step": 137728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.773659673354241e-05,
+      "loss": 4.5412,
+      "step": 138240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.772821078603189e-05,
+      "loss": 4.5443,
+      "step": 138752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.771982483852137e-05,
+      "loss": 4.5375,
+      "step": 139264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.771143889101086e-05,
+      "loss": 4.5448,
+      "step": 139776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.770305294350034e-05,
+      "loss": 4.5342,
+      "step": 140288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.769468337479355e-05,
+      "loss": 4.5396,
+      "step": 140800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.768629742728303e-05,
+      "loss": 4.5322,
+      "step": 141312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.767792785857624e-05,
+      "loss": 4.5377,
+      "step": 141824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.766954191106572e-05,
+      "loss": 4.5267,
+      "step": 142336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.766117234235893e-05,
+      "loss": 4.5507,
+      "step": 142848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.765278639484841e-05,
+      "loss": 4.5377,
+      "step": 143360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.764440044733789e-05,
+      "loss": 4.523,
+      "step": 143872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.763601449982737e-05,
+      "loss": 4.528,
+      "step": 144384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.762762855231685e-05,
+      "loss": 4.5286,
+      "step": 144896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.761924260480633e-05,
+      "loss": 4.5321,
+      "step": 145408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7610856657295813e-05,
+      "loss": 4.5397,
+      "step": 145920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.760248708858902e-05,
+      "loss": 4.5222,
+      "step": 146432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.75941011410785e-05,
+      "loss": 4.5271,
+      "step": 146944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.758571519356798e-05,
+      "loss": 4.5235,
+      "step": 147456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.757732924605746e-05,
+      "loss": 4.5279,
+      "step": 147968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.756895967735067e-05,
+      "loss": 4.5181,
+      "step": 148480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.756057372984015e-05,
+      "loss": 4.5269,
+      "step": 148992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.755218778232963e-05,
+      "loss": 4.5181,
+      "step": 149504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.754380183481911e-05,
+      "loss": 4.5111,
+      "step": 150016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.753541588730859e-05,
+      "loss": 4.5251,
+      "step": 150528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.752702993979807e-05,
+      "loss": 4.5163,
+      "step": 151040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.751864399228755e-05,
+      "loss": 4.516,
+      "step": 151552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.751025804477703e-05,
+      "loss": 4.5169,
+      "step": 152064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.750187209726651e-05,
+      "loss": 4.5087,
+      "step": 152576
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.479610919952393,
+      "eval_runtime": 353.5697,
+      "eval_samples_per_second": 1079.253,
+      "eval_steps_per_second": 33.727,
+      "step": 152640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.749348614975599e-05,
+      "loss": 4.5003,
+      "step": 153088
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.748510020224547e-05,
+      "loss": 4.501,
+      "step": 153600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.747671425473495e-05,
+      "loss": 4.5149,
+      "step": 154112
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.746832830722443e-05,
+      "loss": 4.5023,
+      "step": 154624
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.745994235971391e-05,
+      "loss": 4.5163,
+      "step": 155136
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.745155641220339e-05,
+      "loss": 4.5021,
+      "step": 155648
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.744317046469287e-05,
+      "loss": 4.5034,
+      "step": 156160
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.743478451718235e-05,
+      "loss": 4.4973,
+      "step": 156672
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.742639856967183e-05,
+      "loss": 4.5084,
+      "step": 157184
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.741801262216131e-05,
+      "loss": 4.496,
+      "step": 157696
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.740962667465079e-05,
+      "loss": 4.4957,
+      "step": 158208
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7401240727140264e-05,
+      "loss": 4.4944,
+      "step": 158720
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.739287115843348e-05,
+      "loss": 4.4925,
+      "step": 159232
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7384485210922966e-05,
+      "loss": 4.4876,
+      "step": 159744
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7376099263412446e-05,
+      "loss": 4.4897,
+      "step": 160256
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.736771331590192e-05,
+      "loss": 4.4799,
+      "step": 160768
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.73593273683914e-05,
+      "loss": 4.4865,
+      "step": 161280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.735094142088088e-05,
+      "loss": 4.4794,
+      "step": 161792
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.734255547337036e-05,
+      "loss": 4.4777,
+      "step": 162304
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.733416952585984e-05,
+      "loss": 4.503,
+      "step": 162816
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.732578357834932e-05,
+      "loss": 4.4823,
+      "step": 163328
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.731741400964253e-05,
+      "loss": 4.4801,
+      "step": 163840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.730902806213201e-05,
+      "loss": 4.4812,
+      "step": 164352
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.730064211462149e-05,
+      "loss": 4.4957,
+      "step": 164864
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.729225616711097e-05,
+      "loss": 4.4715,
+      "step": 165376
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7283886598404184e-05,
+      "loss": 4.4751,
+      "step": 165888
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7275500650893664e-05,
+      "loss": 4.4761,
+      "step": 166400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7267114703383144e-05,
+      "loss": 4.482,
+      "step": 166912
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.725874513467635e-05,
+      "loss": 4.4641,
+      "step": 167424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.725035918716583e-05,
+      "loss": 4.4695,
+      "step": 167936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.724197323965531e-05,
+      "loss": 4.4739,
+      "step": 168448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.723358729214479e-05,
+      "loss": 4.4769,
+      "step": 168960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.722520134463427e-05,
+      "loss": 4.4851,
+      "step": 169472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.721681539712375e-05,
+      "loss": 4.4734,
+      "step": 169984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.720842944961323e-05,
+      "loss": 4.4679,
+      "step": 170496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.720004350210271e-05,
+      "loss": 4.4751,
+      "step": 171008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.719167393339592e-05,
+      "loss": 4.4551,
+      "step": 171520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.71832879858854e-05,
+      "loss": 4.4695,
+      "step": 172032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.717490203837488e-05,
+      "loss": 4.4494,
+      "step": 172544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.716651609086437e-05,
+      "loss": 4.459,
+      "step": 173056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.715813014335385e-05,
+      "loss": 4.46,
+      "step": 173568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.714974419584333e-05,
+      "loss": 4.468,
+      "step": 174080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.714135824833281e-05,
+      "loss": 4.4584,
+      "step": 174592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.713297230082228e-05,
+      "loss": 4.4556,
+      "step": 175104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.712458635331176e-05,
+      "loss": 4.4567,
+      "step": 175616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.711620040580124e-05,
+      "loss": 4.4539,
+      "step": 176128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.710781445829072e-05,
+      "loss": 4.4658,
+      "step": 176640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.709944488958393e-05,
+      "loss": 4.4456,
+      "step": 177152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.709105894207341e-05,
+      "loss": 4.4346,
+      "step": 177664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.708267299456289e-05,
+      "loss": 4.4615,
+      "step": 178176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.707428704705237e-05,
+      "loss": 4.4558,
+      "step": 178688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.706590109954185e-05,
+      "loss": 4.4463,
+      "step": 179200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.705751515203134e-05,
+      "loss": 4.4437,
+      "step": 179712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.704912920452082e-05,
+      "loss": 4.4473,
+      "step": 180224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.70407432570103e-05,
+      "loss": 4.4372,
+      "step": 180736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7032373688303506e-05,
+      "loss": 4.4561,
+      "step": 181248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7024004119596715e-05,
+      "loss": 4.4367,
+      "step": 181760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7015618172086195e-05,
+      "loss": 4.4445,
+      "step": 182272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7007232224575675e-05,
+      "loss": 4.4358,
+      "step": 182784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6998846277065155e-05,
+      "loss": 4.4381,
+      "step": 183296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6990460329554635e-05,
+      "loss": 4.439,
+      "step": 183808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6982090760847844e-05,
+      "loss": 4.4362,
+      "step": 184320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6973704813337324e-05,
+      "loss": 4.4277,
+      "step": 184832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.696533524463054e-05,
+      "loss": 4.4303,
+      "step": 185344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.695694929712002e-05,
+      "loss": 4.443,
+      "step": 185856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.69485633496095e-05,
+      "loss": 4.4328,
+      "step": 186368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.694017740209898e-05,
+      "loss": 4.4282,
+      "step": 186880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.693180783339219e-05,
+      "loss": 4.4228,
+      "step": 187392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.692342188588167e-05,
+      "loss": 4.4255,
+      "step": 187904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.691503593837115e-05,
+      "loss": 4.4473,
+      "step": 188416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.690664999086063e-05,
+      "loss": 4.4407,
+      "step": 188928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.689826404335011e-05,
+      "loss": 4.4272,
+      "step": 189440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.688987809583959e-05,
+      "loss": 4.4373,
+      "step": 189952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.688149214832907e-05,
+      "loss": 4.4424,
+      "step": 190464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.687310620081855e-05,
+      "loss": 4.4242,
+      "step": 190976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.686472025330803e-05,
+      "loss": 4.4311,
+      "step": 191488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.685633430579751e-05,
+      "loss": 4.4257,
+      "step": 192000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.684794835828699e-05,
+      "loss": 4.4143,
+      "step": 192512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.683956241077647e-05,
+      "loss": 4.4195,
+      "step": 193024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6831192842069684e-05,
+      "loss": 4.4171,
+      "step": 193536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6822806894559164e-05,
+      "loss": 4.4254,
+      "step": 194048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6814420947048644e-05,
+      "loss": 4.4366,
+      "step": 194560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.680603499953812e-05,
+      "loss": 4.4322,
+      "step": 195072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.67976490520276e-05,
+      "loss": 4.4069,
+      "step": 195584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.678926310451708e-05,
+      "loss": 4.4236,
+      "step": 196096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.678087715700656e-05,
+      "loss": 4.4212,
+      "step": 196608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.677249120949604e-05,
+      "loss": 4.424,
+      "step": 197120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.676410526198552e-05,
+      "loss": 4.4151,
+      "step": 197632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6755735693278726e-05,
+      "loss": 4.4098,
+      "step": 198144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6747349745768206e-05,
+      "loss": 4.4168,
+      "step": 198656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.673896379825769e-05,
+      "loss": 4.4151,
+      "step": 199168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.673057785074717e-05,
+      "loss": 4.4096,
+      "step": 199680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.672220828204038e-05,
+      "loss": 4.4178,
+      "step": 200192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.671382233452986e-05,
+      "loss": 4.4087,
+      "step": 200704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.670543638701934e-05,
+      "loss": 4.4147,
+      "step": 201216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.669705043950882e-05,
+      "loss": 4.4056,
+      "step": 201728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.66886644919983e-05,
+      "loss": 4.4052,
+      "step": 202240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.668029492329151e-05,
+      "loss": 4.4038,
+      "step": 202752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.667190897578099e-05,
+      "loss": 4.4121,
+      "step": 203264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.666352302827047e-05,
+      "loss": 4.4116,
+      "step": 203776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.665513708075995e-05,
+      "loss": 4.3984,
+      "step": 204288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.664676751205316e-05,
+      "loss": 4.4023,
+      "step": 204800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.663838156454265e-05,
+      "loss": 4.3995,
+      "step": 205312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6630011995835856e-05,
+      "loss": 4.4033,
+      "step": 205824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6621626048325336e-05,
+      "loss": 4.3902,
+      "step": 206336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6613240100814816e-05,
+      "loss": 4.3923,
+      "step": 206848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6604854153304296e-05,
+      "loss": 4.4028,
+      "step": 207360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6596468205793776e-05,
+      "loss": 4.3948,
+      "step": 207872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6588082258283256e-05,
+      "loss": 4.3919,
+      "step": 208384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6579696310772736e-05,
+      "loss": 4.391,
+      "step": 208896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6571310363262216e-05,
+      "loss": 4.3854,
+      "step": 209408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6562940794555425e-05,
+      "loss": 4.401,
+      "step": 209920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6554554847044905e-05,
+      "loss": 4.3958,
+      "step": 210432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6546168899534385e-05,
+      "loss": 4.39,
+      "step": 210944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6537782952023865e-05,
+      "loss": 4.3926,
+      "step": 211456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6529397004513344e-05,
+      "loss": 4.4058,
+      "step": 211968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.652102743580656e-05,
+      "loss": 4.3943,
+      "step": 212480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.651264148829604e-05,
+      "loss": 4.3826,
+      "step": 212992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.650425554078552e-05,
+      "loss": 4.393,
+      "step": 213504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6495869593275e-05,
+      "loss": 4.3963,
+      "step": 214016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.648748364576447e-05,
+      "loss": 4.3865,
+      "step": 214528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.647911407705769e-05,
+      "loss": 4.3882,
+      "step": 215040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.64707445083509e-05,
+      "loss": 4.3818,
+      "step": 215552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.646237493964411e-05,
+      "loss": 4.3915,
+      "step": 216064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.645398899213359e-05,
+      "loss": 4.3784,
+      "step": 216576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.644560304462307e-05,
+      "loss": 4.392,
+      "step": 217088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6437217097112554e-05,
+      "loss": 4.3729,
+      "step": 217600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6428831149602034e-05,
+      "loss": 4.3905,
+      "step": 218112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6420445202091514e-05,
+      "loss": 4.3727,
+      "step": 218624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6412059254580994e-05,
+      "loss": 4.3984,
+      "step": 219136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6403673307070474e-05,
+      "loss": 4.3889,
+      "step": 219648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.639528735955995e-05,
+      "loss": 4.3782,
+      "step": 220160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.638690141204943e-05,
+      "loss": 4.3738,
+      "step": 220672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.637851546453891e-05,
+      "loss": 4.3816,
+      "step": 221184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.637012951702839e-05,
+      "loss": 4.3811,
+      "step": 221696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.636174356951787e-05,
+      "loss": 4.3906,
+      "step": 222208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.635335762200735e-05,
+      "loss": 4.3731,
+      "step": 222720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.634497167449683e-05,
+      "loss": 4.3827,
+      "step": 223232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.633658572698631e-05,
+      "loss": 4.3782,
+      "step": 223744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.632821615827952e-05,
+      "loss": 4.3847,
+      "step": 224256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.631984658957273e-05,
+      "loss": 4.3697,
+      "step": 224768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.631146064206221e-05,
+      "loss": 4.3815,
+      "step": 225280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.630307469455169e-05,
+      "loss": 4.3731,
+      "step": 225792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.629468874704117e-05,
+      "loss": 4.3685,
+      "step": 226304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.628630279953065e-05,
+      "loss": 4.381,
+      "step": 226816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.627793323082386e-05,
+      "loss": 4.3779,
+      "step": 227328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.626954728331334e-05,
+      "loss": 4.3674,
+      "step": 227840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.626116133580282e-05,
+      "loss": 4.3782,
+      "step": 228352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.62527753882923e-05,
+      "loss": 4.3677,
+      "step": 228864
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.3468146324157715,
+      "eval_runtime": 293.3117,
+      "eval_samples_per_second": 1300.975,
+      "eval_steps_per_second": 40.656,
+      "step": 228960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.624440581958551e-05,
+      "loss": 4.3634,
+      "step": 229376
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.623601987207499e-05,
+      "loss": 4.3587,
+      "step": 229888
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6227633924564476e-05,
+      "loss": 4.3707,
+      "step": 230400
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6219247977053956e-05,
+      "loss": 4.3667,
+      "step": 230912
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6210862029543436e-05,
+      "loss": 4.3802,
+      "step": 231424
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6202476082032916e-05,
+      "loss": 4.3583,
+      "step": 231936
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6194090134522396e-05,
+      "loss": 4.3661,
+      "step": 232448
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6185704187011876e-05,
+      "loss": 4.3575,
+      "step": 232960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6177318239501356e-05,
+      "loss": 4.3766,
+      "step": 233472
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6168948670794565e-05,
+      "loss": 4.356,
+      "step": 233984
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6160562723284045e-05,
+      "loss": 4.3575,
+      "step": 234496
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6152176775773525e-05,
+      "loss": 4.3629,
+      "step": 235008
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6143807207066734e-05,
+      "loss": 4.3541,
+      "step": 235520
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6135421259556214e-05,
+      "loss": 4.3556,
+      "step": 236032
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6127035312045694e-05,
+      "loss": 4.3547,
+      "step": 236544
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6118649364535174e-05,
+      "loss": 4.347,
+      "step": 237056
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6110263417024654e-05,
+      "loss": 4.3475,
+      "step": 237568
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.610189384831787e-05,
+      "loss": 4.349,
+      "step": 238080
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.609350790080735e-05,
+      "loss": 4.3446,
+      "step": 238592
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.608512195329683e-05,
+      "loss": 4.37,
+      "step": 239104
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.607673600578631e-05,
+      "loss": 4.3488,
+      "step": 239616
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.606835005827578e-05,
+      "loss": 4.3502,
+      "step": 240128
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.605996411076526e-05,
+      "loss": 4.3525,
+      "step": 240640
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.605157816325474e-05,
+      "loss": 4.362,
+      "step": 241152
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.604319221574422e-05,
+      "loss": 4.3444,
+      "step": 241664
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.603482264703743e-05,
+      "loss": 4.3473,
+      "step": 242176
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.602643669952691e-05,
+      "loss": 4.3452,
+      "step": 242688
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.601805075201639e-05,
+      "loss": 4.352,
+      "step": 243200
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.600966480450588e-05,
+      "loss": 4.3392,
+      "step": 243712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.600127885699536e-05,
+      "loss": 4.3445,
+      "step": 244224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.599289290948484e-05,
+      "loss": 4.3472,
+      "step": 244736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.598450696197432e-05,
+      "loss": 4.3507,
+      "step": 245248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.597613739326753e-05,
+      "loss": 4.3571,
+      "step": 245760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.596775144575701e-05,
+      "loss": 4.3491,
+      "step": 246272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.595936549824649e-05,
+      "loss": 4.3411,
+      "step": 246784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.595097955073597e-05,
+      "loss": 4.3477,
+      "step": 247296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5942609982029176e-05,
+      "loss": 4.3342,
+      "step": 247808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5934224034518656e-05,
+      "loss": 4.3411,
+      "step": 248320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5925838087008136e-05,
+      "loss": 4.3286,
+      "step": 248832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5917452139497616e-05,
+      "loss": 4.3325,
+      "step": 249344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5909066191987096e-05,
+      "loss": 4.3336,
+      "step": 249856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.590069662328031e-05,
+      "loss": 4.3484,
+      "step": 250368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.589231067576979e-05,
+      "loss": 4.3395,
+      "step": 250880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.588392472825927e-05,
+      "loss": 4.3316,
+      "step": 251392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.587553878074875e-05,
+      "loss": 4.3351,
+      "step": 251904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.586716921204196e-05,
+      "loss": 4.3327,
+      "step": 252416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.585878326453144e-05,
+      "loss": 4.345,
+      "step": 252928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.585041369582465e-05,
+      "loss": 4.3255,
+      "step": 253440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.584202774831413e-05,
+      "loss": 4.3153,
+      "step": 253952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.583364180080361e-05,
+      "loss": 4.342,
+      "step": 254464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.582525585329309e-05,
+      "loss": 4.3368,
+      "step": 254976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.581686990578257e-05,
+      "loss": 4.3221,
+      "step": 255488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.580848395827205e-05,
+      "loss": 4.329,
+      "step": 256000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.580009801076153e-05,
+      "loss": 4.3301,
+      "step": 256512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5791712063251017e-05,
+      "loss": 4.318,
+      "step": 257024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5783342494544226e-05,
+      "loss": 4.336,
+      "step": 257536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5774972925837435e-05,
+      "loss": 4.3172,
+      "step": 258048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5766586978326915e-05,
+      "loss": 4.3311,
+      "step": 258560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5758201030816395e-05,
+      "loss": 4.3166,
+      "step": 259072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5749815083305875e-05,
+      "loss": 4.3287,
+      "step": 259584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5741429135795355e-05,
+      "loss": 4.3175,
+      "step": 260096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5733043188284835e-05,
+      "loss": 4.3233,
+      "step": 260608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5724657240774315e-05,
+      "loss": 4.3083,
+      "step": 261120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5716271293263794e-05,
+      "loss": 4.3169,
+      "step": 261632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.570788534575327e-05,
+      "loss": 4.3341,
+      "step": 262144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5699499398242754e-05,
+      "loss": 4.3153,
+      "step": 262656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5691113450732234e-05,
+      "loss": 4.3154,
+      "step": 263168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5682727503221714e-05,
+      "loss": 4.3074,
+      "step": 263680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5674357934514923e-05,
+      "loss": 4.3137,
+      "step": 264192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.56659719870044e-05,
+      "loss": 4.3327,
+      "step": 264704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.565760241829762e-05,
+      "loss": 4.3257,
+      "step": 265216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.564921647078709e-05,
+      "loss": 4.3198,
+      "step": 265728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.564084690208031e-05,
+      "loss": 4.3266,
+      "step": 266240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.563246095456979e-05,
+      "loss": 4.3278,
+      "step": 266752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.562407500705927e-05,
+      "loss": 4.3141,
+      "step": 267264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.561568905954874e-05,
+      "loss": 4.3176,
+      "step": 267776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.560730311203822e-05,
+      "loss": 4.3167,
+      "step": 268288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.559891716452771e-05,
+      "loss": 4.3065,
+      "step": 268800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.559053121701719e-05,
+      "loss": 4.3074,
+      "step": 269312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.558214526950667e-05,
+      "loss": 4.3081,
+      "step": 269824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.557377570079988e-05,
+      "loss": 4.3154,
+      "step": 270336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.556538975328936e-05,
+      "loss": 4.3272,
+      "step": 270848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5557020184582566e-05,
+      "loss": 4.321,
+      "step": 271360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5548634237072046e-05,
+      "loss": 4.2962,
+      "step": 271872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5540248289561526e-05,
+      "loss": 4.3218,
+      "step": 272384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5531862342051006e-05,
+      "loss": 4.3133,
+      "step": 272896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5523476394540486e-05,
+      "loss": 4.3147,
+      "step": 273408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5515090447029966e-05,
+      "loss": 4.3084,
+      "step": 273920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5506704499519446e-05,
+      "loss": 4.3019,
+      "step": 274432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5498318552008926e-05,
+      "loss": 4.3116,
+      "step": 274944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5489932604498406e-05,
+      "loss": 4.3061,
+      "step": 275456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.548156303579162e-05,
+      "loss": 4.3071,
+      "step": 275968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.54731770882811e-05,
+      "loss": 4.3092,
+      "step": 276480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.546479114077058e-05,
+      "loss": 4.304,
+      "step": 276992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.545640519326006e-05,
+      "loss": 4.3139,
+      "step": 277504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.544801924574954e-05,
+      "loss": 4.3004,
+      "step": 278016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.543964967704275e-05,
+      "loss": 4.3009,
+      "step": 278528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.543126372953223e-05,
+      "loss": 4.2977,
+      "step": 279040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.542287778202171e-05,
+      "loss": 4.3121,
+      "step": 279552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.541449183451119e-05,
+      "loss": 4.3053,
+      "step": 280064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.540610588700067e-05,
+      "loss": 4.2955,
+      "step": 280576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.539771993949015e-05,
+      "loss": 4.2973,
+      "step": 281088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.538933399197963e-05,
+      "loss": 4.2992,
+      "step": 281600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.538094804446911e-05,
+      "loss": 4.305,
+      "step": 282112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5372578475762326e-05,
+      "loss": 4.2913,
+      "step": 282624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5364192528251806e-05,
+      "loss": 4.2878,
+      "step": 283136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.535580658074128e-05,
+      "loss": 4.3026,
+      "step": 283648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.534742063323076e-05,
+      "loss": 4.2919,
+      "step": 284160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5339051064523975e-05,
+      "loss": 4.291,
+      "step": 284672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5330681495817184e-05,
+      "loss": 4.2891,
+      "step": 285184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5322295548306664e-05,
+      "loss": 4.2885,
+      "step": 285696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5313909600796144e-05,
+      "loss": 4.301,
+      "step": 286208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5305523653285624e-05,
+      "loss": 4.2883,
+      "step": 286720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5297137705775104e-05,
+      "loss": 4.2915,
+      "step": 287232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.528875175826458e-05,
+      "loss": 4.2951,
+      "step": 287744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5280365810754064e-05,
+      "loss": 4.3052,
+      "step": 288256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5271979863243544e-05,
+      "loss": 4.2988,
+      "step": 288768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.526361029453675e-05,
+      "loss": 4.2839,
+      "step": 289280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.525522434702623e-05,
+      "loss": 4.2928,
+      "step": 289792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.524683839951571e-05,
+      "loss": 4.2927,
+      "step": 290304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.523845245200519e-05,
+      "loss": 4.2924,
+      "step": 290816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.52300828832984e-05,
+      "loss": 4.2931,
+      "step": 291328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.522169693578788e-05,
+      "loss": 4.284,
+      "step": 291840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.521331098827736e-05,
+      "loss": 4.2926,
+      "step": 292352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.520492504076684e-05,
+      "loss": 4.2834,
+      "step": 292864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.519653909325632e-05,
+      "loss": 4.2927,
+      "step": 293376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.518816952454953e-05,
+      "loss": 4.2781,
+      "step": 293888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.517978357703902e-05,
+      "loss": 4.2946,
+      "step": 294400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.517141400833223e-05,
+      "loss": 4.2761,
+      "step": 294912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.516302806082171e-05,
+      "loss": 4.3077,
+      "step": 295424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5154642113311187e-05,
+      "loss": 4.2868,
+      "step": 295936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5146256165800667e-05,
+      "loss": 4.2879,
+      "step": 296448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5137870218290147e-05,
+      "loss": 4.2768,
+      "step": 296960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5129484270779626e-05,
+      "loss": 4.2898,
+      "step": 297472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5121098323269106e-05,
+      "loss": 4.2863,
+      "step": 297984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5112712375758586e-05,
+      "loss": 4.2953,
+      "step": 298496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5104326428248066e-05,
+      "loss": 4.2777,
+      "step": 299008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5095940480737546e-05,
+      "loss": 4.2892,
+      "step": 299520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5087554533227026e-05,
+      "loss": 4.284,
+      "step": 300032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5079168585716506e-05,
+      "loss": 4.2953,
+      "step": 300544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5070799017009715e-05,
+      "loss": 4.2746,
+      "step": 301056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5062413069499195e-05,
+      "loss": 4.2879,
+      "step": 301568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.505402712198868e-05,
+      "loss": 4.2802,
+      "step": 302080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.504564117447816e-05,
+      "loss": 4.2789,
+      "step": 302592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.503727160577137e-05,
+      "loss": 4.2872,
+      "step": 303104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.502888565826085e-05,
+      "loss": 4.2845,
+      "step": 303616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.502049971075033e-05,
+      "loss": 4.2765,
+      "step": 304128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.501211376323981e-05,
+      "loss": 4.2836,
+      "step": 304640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.500376057333675e-05,
+      "loss": 4.2798,
+      "step": 305152
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.26357364654541,
+      "eval_runtime": 294.8232,
+      "eval_samples_per_second": 1294.305,
+      "eval_steps_per_second": 40.448,
+      "step": 305280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.499537462582623e-05,
+      "loss": 4.2723,
+      "step": 305664
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.498698867831571e-05,
+      "loss": 4.2659,
+      "step": 306176
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.497860273080519e-05,
+      "loss": 4.2821,
+      "step": 306688
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.497021678329467e-05,
+      "loss": 4.2769,
+      "step": 307200
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.496183083578415e-05,
+      "loss": 4.2912,
+      "step": 307712
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4953444888273636e-05,
+      "loss": 4.2733,
+      "step": 308224
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4945058940763116e-05,
+      "loss": 4.2721,
+      "step": 308736
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.493667299325259e-05,
+      "loss": 4.2703,
+      "step": 309248
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.492828704574207e-05,
+      "loss": 4.2842,
+      "step": 309760
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.491990109823155e-05,
+      "loss": 4.2715,
+      "step": 310272
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.491151515072103e-05,
+      "loss": 4.2671,
+      "step": 310784
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.490312920321051e-05,
+      "loss": 4.2705,
+      "step": 311296
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.489475963450372e-05,
+      "loss": 4.2652,
+      "step": 311808
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.48863736869932e-05,
+      "loss": 4.2689,
+      "step": 312320
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.487798773948268e-05,
+      "loss": 4.2722,
+      "step": 312832
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.486960179197216e-05,
+      "loss": 4.2543,
+      "step": 313344
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.486121584446164e-05,
+      "loss": 4.2619,
+      "step": 313856
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.485282989695112e-05,
+      "loss": 4.2612,
+      "step": 314368
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4844443949440604e-05,
+      "loss": 4.2566,
+      "step": 314880
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4836058001930084e-05,
+      "loss": 4.2805,
+      "step": 315392
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.482768843322329e-05,
+      "loss": 4.2676,
+      "step": 315904
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.481930248571277e-05,
+      "loss": 4.2627,
+      "step": 316416
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.481091653820225e-05,
+      "loss": 4.2651,
+      "step": 316928
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.480253059069173e-05,
+      "loss": 4.2777,
+      "step": 317440
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.479414464318121e-05,
+      "loss": 4.2582,
+      "step": 317952
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.478577507447442e-05,
+      "loss": 4.2639,
+      "step": 318464
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.477740550576763e-05,
+      "loss": 4.2596,
+      "step": 318976
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.476901955825711e-05,
+      "loss": 4.2633,
+      "step": 319488
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.476063361074659e-05,
+      "loss": 4.2554,
+      "step": 320000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.475224766323607e-05,
+      "loss": 4.2623,
+      "step": 320512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.474386171572556e-05,
+      "loss": 4.2633,
+      "step": 321024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.473547576821504e-05,
+      "loss": 4.2649,
+      "step": 321536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.472710619950825e-05,
+      "loss": 4.2721,
+      "step": 322048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.471872025199773e-05,
+      "loss": 4.265,
+      "step": 322560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.471033430448721e-05,
+      "loss": 4.2559,
+      "step": 323072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.470194835697669e-05,
+      "loss": 4.262,
+      "step": 323584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.469356240946617e-05,
+      "loss": 4.2568,
+      "step": 324096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.468517646195565e-05,
+      "loss": 4.2556,
+      "step": 324608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.467679051444512e-05,
+      "loss": 4.2451,
+      "step": 325120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.46684045669346e-05,
+      "loss": 4.2508,
+      "step": 325632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4660034998227816e-05,
+      "loss": 4.2539,
+      "step": 326144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4651649050717296e-05,
+      "loss": 4.2638,
+      "step": 326656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4643263103206776e-05,
+      "loss": 4.2552,
+      "step": 327168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4634877155696256e-05,
+      "loss": 4.2526,
+      "step": 327680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4626491208185736e-05,
+      "loss": 4.2534,
+      "step": 328192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4618105260675215e-05,
+      "loss": 4.2516,
+      "step": 328704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4609719313164695e-05,
+      "loss": 4.2631,
+      "step": 329216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4601333365654175e-05,
+      "loss": 4.246,
+      "step": 329728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.459298017575112e-05,
+      "loss": 4.2328,
+      "step": 330240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4584594228240594e-05,
+      "loss": 4.2606,
+      "step": 330752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4576208280730074e-05,
+      "loss": 4.2578,
+      "step": 331264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4567822333219554e-05,
+      "loss": 4.2409,
+      "step": 331776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4559436385709033e-05,
+      "loss": 4.2498,
+      "step": 332288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4551050438198513e-05,
+      "loss": 4.2534,
+      "step": 332800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.454266449068799e-05,
+      "loss": 4.2359,
+      "step": 333312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.453429492198121e-05,
+      "loss": 4.2567,
+      "step": 333824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.452590897447069e-05,
+      "loss": 4.2403,
+      "step": 334336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.451752302696017e-05,
+      "loss": 4.2493,
+      "step": 334848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.450913707944965e-05,
+      "loss": 4.2376,
+      "step": 335360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.450075113193913e-05,
+      "loss": 4.2475,
+      "step": 335872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.449236518442861e-05,
+      "loss": 4.2397,
+      "step": 336384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.448397923691809e-05,
+      "loss": 4.2462,
+      "step": 336896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.447559328940757e-05,
+      "loss": 4.2298,
+      "step": 337408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.446722372070078e-05,
+      "loss": 4.242,
+      "step": 337920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.445883777319026e-05,
+      "loss": 4.2526,
+      "step": 338432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.445045182567974e-05,
+      "loss": 4.2427,
+      "step": 338944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.444206587816922e-05,
+      "loss": 4.2368,
+      "step": 339456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.443369630946243e-05,
+      "loss": 4.2302,
+      "step": 339968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4425310361951914e-05,
+      "loss": 4.2373,
+      "step": 340480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4416924414441394e-05,
+      "loss": 4.2522,
+      "step": 340992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4408538466930874e-05,
+      "loss": 4.2516,
+      "step": 341504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.440016889822408e-05,
+      "loss": 4.2449,
+      "step": 342016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.439179932951729e-05,
+      "loss": 4.2504,
+      "step": 342528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.438341338200677e-05,
+      "loss": 4.2527,
+      "step": 343040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.437502743449625e-05,
+      "loss": 4.2377,
+      "step": 343552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.436664148698573e-05,
+      "loss": 4.2405,
+      "step": 344064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.435827191827894e-05,
+      "loss": 4.2452,
+      "step": 344576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.434988597076842e-05,
+      "loss": 4.2307,
+      "step": 345088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.43415000232579e-05,
+      "loss": 4.2326,
+      "step": 345600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4333130454551117e-05,
+      "loss": 4.237,
+      "step": 346112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4324744507040597e-05,
+      "loss": 4.236,
+      "step": 346624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4316358559530076e-05,
+      "loss": 4.2536,
+      "step": 347136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4307972612019556e-05,
+      "loss": 4.2476,
+      "step": 347648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4299586664509036e-05,
+      "loss": 4.2225,
+      "step": 348160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4291200716998516e-05,
+      "loss": 4.243,
+      "step": 348672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4282814769487996e-05,
+      "loss": 4.2398,
+      "step": 349184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4274428821977476e-05,
+      "loss": 4.2407,
+      "step": 349696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4266059253270685e-05,
+      "loss": 4.236,
+      "step": 350208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4257689684563894e-05,
+      "loss": 4.2277,
+      "step": 350720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4249303737053374e-05,
+      "loss": 4.2366,
+      "step": 351232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4240917789542854e-05,
+      "loss": 4.2325,
+      "step": 351744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4232531842032334e-05,
+      "loss": 4.2323,
+      "step": 352256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.422414589452182e-05,
+      "loss": 4.2314,
+      "step": 352768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.42157599470113e-05,
+      "loss": 4.2345,
+      "step": 353280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.420737399950078e-05,
+      "loss": 4.2364,
+      "step": 353792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.419898805199026e-05,
+      "loss": 4.2345,
+      "step": 354304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4190602104479734e-05,
+      "loss": 4.223,
+      "step": 354816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4182216156969214e-05,
+      "loss": 4.229,
+      "step": 355328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.417384658826243e-05,
+      "loss": 4.234,
+      "step": 355840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.41654606407519e-05,
+      "loss": 4.2387,
+      "step": 356352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.415707469324138e-05,
+      "loss": 4.2234,
+      "step": 356864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.414868874573086e-05,
+      "loss": 4.2252,
+      "step": 357376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.414030279822034e-05,
+      "loss": 4.2272,
+      "step": 357888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.413193322951356e-05,
+      "loss": 4.2297,
+      "step": 358400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.412354728200304e-05,
+      "loss": 4.2217,
+      "step": 358912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4115177713296255e-05,
+      "loss": 4.2177,
+      "step": 359424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4106791765785735e-05,
+      "loss": 4.2265,
+      "step": 359936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.409840581827521e-05,
+      "loss": 4.2226,
+      "step": 360448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.409001987076469e-05,
+      "loss": 4.2196,
+      "step": 360960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.408163392325417e-05,
+      "loss": 4.2195,
+      "step": 361472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.407324797574365e-05,
+      "loss": 4.2177,
+      "step": 361984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.406486202823313e-05,
+      "loss": 4.2304,
+      "step": 362496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.405647608072261e-05,
+      "loss": 4.2196,
+      "step": 363008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.404809013321209e-05,
+      "loss": 4.2188,
+      "step": 363520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.403970418570157e-05,
+      "loss": 4.2265,
+      "step": 364032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.403131823819105e-05,
+      "loss": 4.2386,
+      "step": 364544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4022948669484257e-05,
+      "loss": 4.2308,
+      "step": 365056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.401456272197374e-05,
+      "loss": 4.2146,
+      "step": 365568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.400617677446322e-05,
+      "loss": 4.2201,
+      "step": 366080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.39977908269527e-05,
+      "loss": 4.2244,
+      "step": 366592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.398942125824591e-05,
+      "loss": 4.22,
+      "step": 367104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.398106806834285e-05,
+      "loss": 4.2258,
+      "step": 367616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.397268212083233e-05,
+      "loss": 4.2133,
+      "step": 368128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.396429617332181e-05,
+      "loss": 4.2226,
+      "step": 368640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.395591022581129e-05,
+      "loss": 4.2208,
+      "step": 369152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.394752427830077e-05,
+      "loss": 4.2214,
+      "step": 369664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.393913833079025e-05,
+      "loss": 4.2074,
+      "step": 370176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.393075238327973e-05,
+      "loss": 4.2224,
+      "step": 370688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.392236643576921e-05,
+      "loss": 4.2153,
+      "step": 371200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.39139804882587e-05,
+      "loss": 4.2315,
+      "step": 371712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.390559454074818e-05,
+      "loss": 4.2181,
+      "step": 372224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.389720859323766e-05,
+      "loss": 4.2245,
+      "step": 372736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3888839024530866e-05,
+      "loss": 4.2093,
+      "step": 373248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3880453077020346e-05,
+      "loss": 4.2246,
+      "step": 373760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3872067129509826e-05,
+      "loss": 4.222,
+      "step": 374272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3863681181999306e-05,
+      "loss": 4.2265,
+      "step": 374784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3855295234488786e-05,
+      "loss": 4.209,
+      "step": 375296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3846909286978266e-05,
+      "loss": 4.2282,
+      "step": 375808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3838539718271475e-05,
+      "loss": 4.2136,
+      "step": 376320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3830153770760955e-05,
+      "loss": 4.2285,
+      "step": 376832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3821767823250435e-05,
+      "loss": 4.2099,
+      "step": 377344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3813381875739915e-05,
+      "loss": 4.2209,
+      "step": 377856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3804995928229395e-05,
+      "loss": 4.2126,
+      "step": 378368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3796609980718875e-05,
+      "loss": 4.2102,
+      "step": 378880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3788224033208355e-05,
+      "loss": 4.22,
+      "step": 379392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3779854464501564e-05,
+      "loss": 4.22,
+      "step": 379904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3771468516991044e-05,
+      "loss": 4.2113,
+      "step": 380416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3763082569480524e-05,
+      "loss": 4.2147,
+      "step": 380928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3754696621970004e-05,
+      "loss": 4.2197,
+      "step": 381440
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.205178260803223,
+      "eval_runtime": 295.0791,
+      "eval_samples_per_second": 1293.182,
+      "eval_steps_per_second": 40.413,
+      "step": 381600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3746310674459483e-05,
+      "loss": 4.2085,
+      "step": 381952
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3737924726948963e-05,
+      "loss": 4.2003,
+      "step": 382464
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3729538779438443e-05,
+      "loss": 4.2162,
+      "step": 382976
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.372115283192792e-05,
+      "loss": 4.2118,
+      "step": 383488
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.37127668844174e-05,
+      "loss": 4.2211,
+      "step": 384000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.370438093690688e-05,
+      "loss": 4.2094,
+      "step": 384512
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.369599498939636e-05,
+      "loss": 4.2058,
+      "step": 385024
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.368760904188584e-05,
+      "loss": 4.2067,
+      "step": 385536
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.367922309437533e-05,
+      "loss": 4.2189,
+      "step": 386048
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.367083714686481e-05,
+      "loss": 4.2061,
+      "step": 386560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.366245119935428e-05,
+      "loss": 4.2062,
+      "step": 387072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.365406525184376e-05,
+      "loss": 4.2031,
+      "step": 387584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.364569568313698e-05,
+      "loss": 4.2044,
+      "step": 388096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.363730973562645e-05,
+      "loss": 4.2049,
+      "step": 388608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.362892378811593e-05,
+      "loss": 4.2084,
+      "step": 389120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.362053784060541e-05,
+      "loss": 4.1908,
+      "step": 389632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.361215189309489e-05,
+      "loss": 4.1966,
+      "step": 390144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.360376594558437e-05,
+      "loss": 4.2003,
+      "step": 390656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.359537999807385e-05,
+      "loss": 4.1926,
+      "step": 391168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.358699405056333e-05,
+      "loss": 4.217,
+      "step": 391680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.357860810305281e-05,
+      "loss": 4.2081,
+      "step": 392192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.35702221555423e-05,
+      "loss": 4.2024,
+      "step": 392704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.356183620803178e-05,
+      "loss": 4.1983,
+      "step": 393216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.355346663932499e-05,
+      "loss": 4.2128,
+      "step": 393728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.354508069181447e-05,
+      "loss": 4.198,
+      "step": 394240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.353669474430395e-05,
+      "loss": 4.2038,
+      "step": 394752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.352830879679343e-05,
+      "loss": 4.1939,
+      "step": 395264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.351992284928291e-05,
+      "loss": 4.2014,
+      "step": 395776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.351153690177239e-05,
+      "loss": 4.1967,
+      "step": 396288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.350315095426187e-05,
+      "loss": 4.1984,
+      "step": 396800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.349476500675134e-05,
+      "loss": 4.2008,
+      "step": 397312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.348637905924082e-05,
+      "loss": 4.2041,
+      "step": 397824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3478009490534036e-05,
+      "loss": 4.2088,
+      "step": 398336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3469623543023516e-05,
+      "loss": 4.2073,
+      "step": 398848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3461237595512996e-05,
+      "loss": 4.1928,
+      "step": 399360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.345286802680621e-05,
+      "loss": 4.2032,
+      "step": 399872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.344449845809942e-05,
+      "loss": 4.1965,
+      "step": 400384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.34361125105889e-05,
+      "loss": 4.1875,
+      "step": 400896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.342772656307838e-05,
+      "loss": 4.1902,
+      "step": 401408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.341934061556786e-05,
+      "loss": 4.1899,
+      "step": 401920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.341095466805734e-05,
+      "loss": 4.1922,
+      "step": 402432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3402568720546814e-05,
+      "loss": 4.2088,
+      "step": 402944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3394182773036294e-05,
+      "loss": 4.1904,
+      "step": 403456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3385796825525774e-05,
+      "loss": 4.1946,
+      "step": 403968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.337742725681899e-05,
+      "loss": 4.1986,
+      "step": 404480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.336904130930847e-05,
+      "loss": 4.1925,
+      "step": 404992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.336065536179795e-05,
+      "loss": 4.1961,
+      "step": 405504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.335226941428743e-05,
+      "loss": 4.1881,
+      "step": 406016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.334388346677691e-05,
+      "loss": 4.1729,
+      "step": 406528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.333551389807012e-05,
+      "loss": 4.2042,
+      "step": 407040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.33271279505596e-05,
+      "loss": 4.1949,
+      "step": 407552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.331874200304908e-05,
+      "loss": 4.1873,
+      "step": 408064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.331035605553856e-05,
+      "loss": 4.1886,
+      "step": 408576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.330197010802804e-05,
+      "loss": 4.1908,
+      "step": 409088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.329358416051752e-05,
+      "loss": 4.1766,
+      "step": 409600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.328521459181073e-05,
+      "loss": 4.1942,
+      "step": 410112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.327682864430021e-05,
+      "loss": 4.1854,
+      "step": 410624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.326844269678969e-05,
+      "loss": 4.1907,
+      "step": 411136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.326005674927917e-05,
+      "loss": 4.1832,
+      "step": 411648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3251670801768654e-05,
+      "loss": 4.1864,
+      "step": 412160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3243284854258134e-05,
+      "loss": 4.1807,
+      "step": 412672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3234898906747614e-05,
+      "loss": 4.1903,
+      "step": 413184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3226512959237094e-05,
+      "loss": 4.174,
+      "step": 413696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.32181433905303e-05,
+      "loss": 4.1827,
+      "step": 414208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.320975744301978e-05,
+      "loss": 4.1938,
+      "step": 414720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.320137149550926e-05,
+      "loss": 4.1857,
+      "step": 415232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.319298554799874e-05,
+      "loss": 4.1776,
+      "step": 415744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.318461597929195e-05,
+      "loss": 4.1779,
+      "step": 416256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.317623003178143e-05,
+      "loss": 4.1788,
+      "step": 416768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.316784408427091e-05,
+      "loss": 4.1932,
+      "step": 417280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.315945813676039e-05,
+      "loss": 4.1945,
+      "step": 417792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.315108856805361e-05,
+      "loss": 4.1863,
+      "step": 418304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.314271899934682e-05,
+      "loss": 4.195,
+      "step": 418816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.31343330518363e-05,
+      "loss": 4.1955,
+      "step": 419328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3125963483129506e-05,
+      "loss": 4.1889,
+      "step": 419840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3117593914422715e-05,
+      "loss": 4.1815,
+      "step": 420352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3109207966912195e-05,
+      "loss": 4.1842,
+      "step": 420864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3100822019401675e-05,
+      "loss": 4.1817,
+      "step": 421376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3092436071891155e-05,
+      "loss": 4.1731,
+      "step": 421888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3084050124380635e-05,
+      "loss": 4.1849,
+      "step": 422400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3075664176870115e-05,
+      "loss": 4.1781,
+      "step": 422912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3067278229359595e-05,
+      "loss": 4.1951,
+      "step": 423424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3058892281849075e-05,
+      "loss": 4.1948,
+      "step": 423936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.305050633433856e-05,
+      "loss": 4.1655,
+      "step": 424448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.304212038682804e-05,
+      "loss": 4.1903,
+      "step": 424960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.303373443931752e-05,
+      "loss": 4.1798,
+      "step": 425472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3025348491807e-05,
+      "loss": 4.1851,
+      "step": 425984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3016962544296475e-05,
+      "loss": 4.1797,
+      "step": 426496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3008576596785955e-05,
+      "loss": 4.1742,
+      "step": 427008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.300020702807917e-05,
+      "loss": 4.1828,
+      "step": 427520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.299182108056865e-05,
+      "loss": 4.1776,
+      "step": 428032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2983435133058124e-05,
+      "loss": 4.181,
+      "step": 428544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2975049185547604e-05,
+      "loss": 4.1752,
+      "step": 429056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2966663238037084e-05,
+      "loss": 4.1824,
+      "step": 429568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2958277290526564e-05,
+      "loss": 4.1796,
+      "step": 430080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2949891343016043e-05,
+      "loss": 4.1807,
+      "step": 430592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.294150539550553e-05,
+      "loss": 4.1698,
+      "step": 431104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.293313582679874e-05,
+      "loss": 4.1761,
+      "step": 431616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.292476625809195e-05,
+      "loss": 4.1769,
+      "step": 432128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.291638031058143e-05,
+      "loss": 4.1871,
+      "step": 432640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.290799436307091e-05,
+      "loss": 4.1709,
+      "step": 433152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.289960841556039e-05,
+      "loss": 4.1733,
+      "step": 433664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.289122246804987e-05,
+      "loss": 4.1678,
+      "step": 434176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.288283652053935e-05,
+      "loss": 4.1801,
+      "step": 434688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.287446695183256e-05,
+      "loss": 4.1644,
+      "step": 435200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2866097383125766e-05,
+      "loss": 4.1691,
+      "step": 435712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.285771143561525e-05,
+      "loss": 4.1713,
+      "step": 436224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.284932548810473e-05,
+      "loss": 4.1718,
+      "step": 436736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.284093954059421e-05,
+      "loss": 4.165,
+      "step": 437248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.283255359308369e-05,
+      "loss": 4.1715,
+      "step": 437760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.282416764557317e-05,
+      "loss": 4.1653,
+      "step": 438272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.281578169806265e-05,
+      "loss": 4.1707,
+      "step": 438784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.280741212935586e-05,
+      "loss": 4.174,
+      "step": 439296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.279902618184534e-05,
+      "loss": 4.1623,
+      "step": 439808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.279064023433482e-05,
+      "loss": 4.1756,
+      "step": 440320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.27822542868243e-05,
+      "loss": 4.1872,
+      "step": 440832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.277386833931378e-05,
+      "loss": 4.181,
+      "step": 441344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.276548239180326e-05,
+      "loss": 4.1624,
+      "step": 441856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.275709644429274e-05,
+      "loss": 4.1667,
+      "step": 442368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.274871049678222e-05,
+      "loss": 4.1718,
+      "step": 442880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.27403245492717e-05,
+      "loss": 4.1664,
+      "step": 443392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.273193860176118e-05,
+      "loss": 4.1761,
+      "step": 443904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.272355265425066e-05,
+      "loss": 4.1616,
+      "step": 444416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.271518308554388e-05,
+      "loss": 4.1706,
+      "step": 444928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.270679713803336e-05,
+      "loss": 4.1672,
+      "step": 445440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.269841119052284e-05,
+      "loss": 4.1746,
+      "step": 445952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.269002524301231e-05,
+      "loss": 4.1604,
+      "step": 446464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2681655674305526e-05,
+      "loss": 4.1668,
+      "step": 446976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2673269726795006e-05,
+      "loss": 4.1704,
+      "step": 447488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2664883779284486e-05,
+      "loss": 4.1782,
+      "step": 448000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.265649783177396e-05,
+      "loss": 4.1702,
+      "step": 448512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.264811188426344e-05,
+      "loss": 4.1714,
+      "step": 449024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.263972593675292e-05,
+      "loss": 4.1604,
+      "step": 449536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.26313399892424e-05,
+      "loss": 4.1747,
+      "step": 450048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2622954041731886e-05,
+      "loss": 4.1699,
+      "step": 450560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2614584473025095e-05,
+      "loss": 4.1768,
+      "step": 451072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2606198525514575e-05,
+      "loss": 4.1618,
+      "step": 451584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2597828956807784e-05,
+      "loss": 4.1748,
+      "step": 452096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2589443009297264e-05,
+      "loss": 4.1633,
+      "step": 452608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2581057061786744e-05,
+      "loss": 4.1825,
+      "step": 453120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2572671114276224e-05,
+      "loss": 4.1578,
+      "step": 453632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2564285166765704e-05,
+      "loss": 4.174,
+      "step": 454144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2555899219255184e-05,
+      "loss": 4.1635,
+      "step": 454656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2547513271744664e-05,
+      "loss": 4.1607,
+      "step": 455168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2539127324234144e-05,
+      "loss": 4.1697,
+      "step": 455680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.253075775552735e-05,
+      "loss": 4.1666,
+      "step": 456192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.252237180801684e-05,
+      "loss": 4.1703,
+      "step": 456704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.251398586050632e-05,
+      "loss": 4.1597,
+      "step": 457216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.25055999129958e-05,
+      "loss": 4.1719,
+      "step": 457728
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.1641435623168945,
+      "eval_runtime": 300.7161,
+      "eval_samples_per_second": 1268.941,
+      "eval_steps_per_second": 39.655,
+      "step": 457920
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.249721396548528e-05,
+      "loss": 4.1608,
+      "step": 458240
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.248882801797476e-05,
+      "loss": 4.1511,
+      "step": 458752
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.248044207046424e-05,
+      "loss": 4.1627,
+      "step": 459264
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.247205612295372e-05,
+      "loss": 4.1641,
+      "step": 459776
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.246368655424693e-05,
+      "loss": 4.1721,
+      "step": 460288
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.245530060673641e-05,
+      "loss": 4.1654,
+      "step": 460800
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.244691465922589e-05,
+      "loss": 4.1514,
+      "step": 461312
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.243852871171537e-05,
+      "loss": 4.1624,
+      "step": 461824
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.243015914300858e-05,
+      "loss": 4.1693,
+      "step": 462336
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.2421789574301793e-05,
+      "loss": 4.1595,
+      "step": 462848
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.241340362679127e-05,
+      "loss": 4.1568,
+      "step": 463360
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.240501767928075e-05,
+      "loss": 4.1551,
+      "step": 463872
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.239663173177023e-05,
+      "loss": 4.1591,
+      "step": 464384
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.238824578425971e-05,
+      "loss": 4.1545,
+      "step": 464896
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.237985983674919e-05,
+      "loss": 4.1617,
+      "step": 465408
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.23714902680424e-05,
+      "loss": 4.1453,
+      "step": 465920
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.236312069933561e-05,
+      "loss": 4.148,
+      "step": 466432
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.235473475182509e-05,
+      "loss": 4.1522,
+      "step": 466944
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.234634880431457e-05,
+      "loss": 4.1475,
+      "step": 467456
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.233796285680405e-05,
+      "loss": 4.1681,
+      "step": 467968
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.232957690929353e-05,
+      "loss": 4.1631,
+      "step": 468480
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.232119096178301e-05,
+      "loss": 4.1553,
+      "step": 468992
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.231280501427249e-05,
+      "loss": 4.1492,
+      "step": 469504
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.230441906676197e-05,
+      "loss": 4.1644,
+      "step": 470016
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.229603311925145e-05,
+      "loss": 4.1513,
+      "step": 470528
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.228764717174093e-05,
+      "loss": 4.1561,
+      "step": 471040
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.227927760303415e-05,
+      "loss": 4.1492,
+      "step": 471552
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.227089165552362e-05,
+      "loss": 4.1569,
+      "step": 472064
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.22625057080131e-05,
+      "loss": 4.1521,
+      "step": 472576
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.225411976050258e-05,
+      "loss": 4.1495,
+      "step": 473088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.224573381299206e-05,
+      "loss": 4.153,
+      "step": 473600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.223734786548154e-05,
+      "loss": 4.1591,
+      "step": 474112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.222896191797102e-05,
+      "loss": 4.1578,
+      "step": 474624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.222059234926423e-05,
+      "loss": 4.1606,
+      "step": 475136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.221220640175371e-05,
+      "loss": 4.1464,
+      "step": 475648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2203820454243196e-05,
+      "loss": 4.1616,
+      "step": 476160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2195434506732676e-05,
+      "loss": 4.1504,
+      "step": 476672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2187048559222156e-05,
+      "loss": 4.141,
+      "step": 477184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2178662611711635e-05,
+      "loss": 4.1433,
+      "step": 477696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2170276664201115e-05,
+      "loss": 4.1427,
+      "step": 478208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2161890716690595e-05,
+      "loss": 4.1468,
+      "step": 478720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2153504769180075e-05,
+      "loss": 4.159,
+      "step": 479232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2145135200473284e-05,
+      "loss": 4.1446,
+      "step": 479744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2136765631766494e-05,
+      "loss": 4.1492,
+      "step": 480256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2128379684255973e-05,
+      "loss": 4.1542,
+      "step": 480768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2119993736745453e-05,
+      "loss": 4.1485,
+      "step": 481280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.211160778923493e-05,
+      "loss": 4.1535,
+      "step": 481792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.210322184172441e-05,
+      "loss": 4.141,
+      "step": 482304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.209483589421389e-05,
+      "loss": 4.1286,
+      "step": 482816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.208644994670338e-05,
+      "loss": 4.1584,
+      "step": 483328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.207806399919286e-05,
+      "loss": 4.153,
+      "step": 483840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.206969443048607e-05,
+      "loss": 4.1381,
+      "step": 484352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.206130848297555e-05,
+      "loss": 4.1493,
+      "step": 484864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.205293891426876e-05,
+      "loss": 4.1414,
+      "step": 485376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.204455296675824e-05,
+      "loss": 4.1332,
+      "step": 485888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.203616701924772e-05,
+      "loss": 4.1454,
+      "step": 486400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.20277810717372e-05,
+      "loss": 4.1435,
+      "step": 486912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.201939512422668e-05,
+      "loss": 4.147,
+      "step": 487424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.201102555551989e-05,
+      "loss": 4.137,
+      "step": 487936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.200263960800937e-05,
+      "loss": 4.1438,
+      "step": 488448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.199425366049885e-05,
+      "loss": 4.1353,
+      "step": 488960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.1985867712988334e-05,
+      "loss": 4.1485,
+      "step": 489472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.197748176547781e-05,
+      "loss": 4.1309,
+      "step": 489984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.196909581796729e-05,
+      "loss": 4.1375,
+      "step": 490496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.196070987045677e-05,
+      "loss": 4.1496,
+      "step": 491008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.195234030174998e-05,
+      "loss": 4.1421,
+      "step": 491520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.1943954354239456e-05,
+      "loss": 4.1382,
+      "step": 492032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.1935568406728936e-05,
+      "loss": 4.1321,
+      "step": 492544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.1927182459218416e-05,
+      "loss": 4.1312,
+      "step": 493056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.1918796511707896e-05,
+      "loss": 4.1438,
+      "step": 493568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.1910410564197376e-05,
+      "loss": 4.1544,
+      "step": 494080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.1902040995490585e-05,
+      "loss": 4.1419,
+      "step": 494592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.189365504798007e-05,
+      "loss": 4.1524,
+      "step": 495104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.188526910046955e-05,
+      "loss": 4.1493,
+      "step": 495616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.187688315295903e-05,
+      "loss": 4.1505,
+      "step": 496128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.186851358425224e-05,
+      "loss": 4.1376,
+      "step": 496640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.186012763674172e-05,
+      "loss": 4.1396,
+      "step": 497152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.18517416892312e-05,
+      "loss": 4.1392,
+      "step": 497664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.184335574172068e-05,
+      "loss": 4.1271,
+      "step": 498176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.183496979421016e-05,
+      "loss": 4.1417,
+      "step": 498688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.182658384669964e-05,
+      "loss": 4.1379,
+      "step": 499200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.181819789918912e-05,
+      "loss": 4.1503,
+      "step": 499712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.180982833048233e-05,
+      "loss": 4.1516,
+      "step": 500224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.180144238297181e-05,
+      "loss": 4.1233,
+      "step": 500736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.179305643546129e-05,
+      "loss": 4.1517,
+      "step": 501248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.178467048795077e-05,
+      "loss": 4.1377,
+      "step": 501760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.177628454044025e-05,
+      "loss": 4.1423,
+      "step": 502272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.1767898592929736e-05,
+      "loss": 4.1365,
+      "step": 502784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.1759512645419216e-05,
+      "loss": 4.1334,
+      "step": 503296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1751126697908696e-05,
+      "loss": 4.1401,
+      "step": 503808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1742757129201905e-05,
+      "loss": 4.1348,
+      "step": 504320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1734371181691385e-05,
+      "loss": 4.1364,
+      "step": 504832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1725985234180865e-05,
+      "loss": 4.136,
+      "step": 505344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1717615665474074e-05,
+      "loss": 4.1389,
+      "step": 505856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1709229717963554e-05,
+      "loss": 4.1436,
+      "step": 506368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1700843770453034e-05,
+      "loss": 4.1306,
+      "step": 506880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1692457822942514e-05,
+      "loss": 4.1316,
+      "step": 507392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.168407187543199e-05,
+      "loss": 4.1333,
+      "step": 507904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.167571868552894e-05,
+      "loss": 4.134,
+      "step": 508416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.166733273801842e-05,
+      "loss": 4.1465,
+      "step": 508928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.16589467905079e-05,
+      "loss": 4.1309,
+      "step": 509440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.165056084299738e-05,
+      "loss": 4.1306,
+      "step": 509952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.164217489548686e-05,
+      "loss": 4.1298,
+      "step": 510464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.163378894797634e-05,
+      "loss": 4.1408,
+      "step": 510976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.162540300046581e-05,
+      "loss": 4.1185,
+      "step": 511488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.161701705295529e-05,
+      "loss": 4.129,
+      "step": 512000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.160863110544477e-05,
+      "loss": 4.1305,
+      "step": 512512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.160024515793425e-05,
+      "loss": 4.1312,
+      "step": 513024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.159187558922746e-05,
+      "loss": 4.1249,
+      "step": 513536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.158348964171694e-05,
+      "loss": 4.1295,
+      "step": 514048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.157510369420643e-05,
+      "loss": 4.1238,
+      "step": 514560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.156671774669591e-05,
+      "loss": 4.1312,
+      "step": 515072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1558348177989116e-05,
+      "loss": 4.1328,
+      "step": 515584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1549962230478596e-05,
+      "loss": 4.1246,
+      "step": 516096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1541576282968076e-05,
+      "loss": 4.1309,
+      "step": 516608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1533190335457556e-05,
+      "loss": 4.1453,
+      "step": 517120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1524804387947036e-05,
+      "loss": 4.1439,
+      "step": 517632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1516418440436516e-05,
+      "loss": 4.1243,
+      "step": 518144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1508032492925996e-05,
+      "loss": 4.125,
+      "step": 518656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1499646545415476e-05,
+      "loss": 4.1327,
+      "step": 519168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1491276976708685e-05,
+      "loss": 4.1249,
+      "step": 519680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1482891029198165e-05,
+      "loss": 4.1384,
+      "step": 520192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1474505081687645e-05,
+      "loss": 4.1158,
+      "step": 520704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1466119134177125e-05,
+      "loss": 4.1361,
+      "step": 521216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.145774956547034e-05,
+      "loss": 4.1235,
+      "step": 521728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.144936361795982e-05,
+      "loss": 4.1367,
+      "step": 522240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.14409776704493e-05,
+      "loss": 4.1227,
+      "step": 522752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.143260810174251e-05,
+      "loss": 4.1292,
+      "step": 523264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.142422215423199e-05,
+      "loss": 4.1266,
+      "step": 523776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.14158525855252e-05,
+      "loss": 4.1404,
+      "step": 524288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.140746663801468e-05,
+      "loss": 4.1276,
+      "step": 524800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.139908069050416e-05,
+      "loss": 4.1324,
+      "step": 525312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.139069474299364e-05,
+      "loss": 4.1211,
+      "step": 525824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.138230879548312e-05,
+      "loss": 4.1358,
+      "step": 526336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.13739228479726e-05,
+      "loss": 4.1306,
+      "step": 526848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.136553690046208e-05,
+      "loss": 4.133,
+      "step": 527360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1357150952951565e-05,
+      "loss": 4.1259,
+      "step": 527872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1348765005441045e-05,
+      "loss": 4.135,
+      "step": 528384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1340379057930525e-05,
+      "loss": 4.1262,
+      "step": 528896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1331993110420005e-05,
+      "loss": 4.1403,
+      "step": 529408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.132360716290948e-05,
+      "loss": 4.119,
+      "step": 529920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1315237594202694e-05,
+      "loss": 4.1343,
+      "step": 530432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1306851646692174e-05,
+      "loss": 4.126,
+      "step": 530944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.129846569918165e-05,
+      "loss": 4.1253,
+      "step": 531456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.129007975167113e-05,
+      "loss": 4.1243,
+      "step": 531968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.128172656176807e-05,
+      "loss": 4.1273,
+      "step": 532480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.127334061425755e-05,
+      "loss": 4.1296,
+      "step": 532992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.126495466674703e-05,
+      "loss": 4.1242,
+      "step": 533504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.125656871923652e-05,
+      "loss": 4.1346,
+      "step": 534016
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.133681774139404,
+      "eval_runtime": 299.1733,
+      "eval_samples_per_second": 1275.485,
+      "eval_steps_per_second": 39.86,
+      "step": 534240
+    }
+  ],
+  "logging_steps": 512,
+  "max_steps": 3052726,
+  "num_train_epochs": 9223372036854775807,
+  "save_steps": 10,
+  "total_flos": 2.191907311679832e+17,
+  "trial_name": null,
+  "trial_params": null
+}