diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,12192 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.9982507288629736,
+  "eval_steps": 86,
+  "global_step": 1714,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0011661807580174927,
+      "grad_norm": 0.9131848216056824,
+      "learning_rate": 9.999416262915183e-06,
+      "loss": 2.071,
+      "step": 1
+    },
+    {
+      "epoch": 0.0011661807580174927,
+      "eval_loss": 1.1986331939697266,
+      "eval_runtime": 174.1538,
+      "eval_samples_per_second": 39.643,
+      "eval_steps_per_second": 2.481,
+      "step": 1
+    },
+    {
+      "epoch": 0.0023323615160349854,
+      "grad_norm": 0.8528902530670166,
+      "learning_rate": 9.998831912159796e-06,
+      "loss": 2.0233,
+      "step": 2
+    },
+    {
+      "epoch": 0.003498542274052478,
+      "grad_norm": 0.9434170126914978,
+      "learning_rate": 9.998246946765618e-06,
+      "loss": 1.985,
+      "step": 3
+    },
+    {
+      "epoch": 0.004664723032069971,
+      "grad_norm": 0.8081845045089722,
+      "learning_rate": 9.997661365762395e-06,
+      "loss": 2.0784,
+      "step": 4
+    },
+    {
+      "epoch": 0.0058309037900874635,
+      "grad_norm": 0.656217634677887,
+      "learning_rate": 9.99707516817783e-06,
+      "loss": 2.0497,
+      "step": 5
+    },
+    {
+      "epoch": 0.006997084548104956,
+      "grad_norm": 0.5666890144348145,
+      "learning_rate": 9.996488353037576e-06,
+      "loss": 1.9716,
+      "step": 6
+    },
+    {
+      "epoch": 0.00816326530612245,
+      "grad_norm": 0.5374447703361511,
+      "learning_rate": 9.99590091936523e-06,
+      "loss": 2.0613,
+      "step": 7
+    },
+    {
+      "epoch": 0.009329446064139942,
+      "grad_norm": 0.4860251843929291,
+      "learning_rate": 9.995312866182331e-06,
+      "loss": 2.011,
+      "step": 8
+    },
+    {
+      "epoch": 0.010495626822157435,
+      "grad_norm": 0.5425748229026794,
+      "learning_rate": 9.994724192508354e-06,
+      "loss": 1.9325,
+      "step": 9
+    },
+    {
+      "epoch": 0.011661807580174927,
+      "grad_norm": 0.595158576965332,
+      "learning_rate": 9.994134897360705e-06,
+      "loss": 2.0397,
+      "step": 10
+    },
+    {
+      "epoch": 0.01282798833819242,
+      "grad_norm": 2.011725425720215,
+      "learning_rate": 9.99354497975471e-06,
+      "loss": 2.0171,
+      "step": 11
+    },
+    {
+      "epoch": 0.013994169096209912,
+      "grad_norm": 0.9743179082870483,
+      "learning_rate": 9.992954438703617e-06,
+      "loss": 1.9163,
+      "step": 12
+    },
+    {
+      "epoch": 0.015160349854227406,
+      "grad_norm": 2.6744604110717773,
+      "learning_rate": 9.992363273218586e-06,
+      "loss": 2.0376,
+      "step": 13
+    },
+    {
+      "epoch": 0.0163265306122449,
+      "grad_norm": 1.3725961446762085,
+      "learning_rate": 9.991771482308689e-06,
+      "loss": 1.9759,
+      "step": 14
+    },
+    {
+      "epoch": 0.01749271137026239,
+      "grad_norm": 1.0943834781646729,
+      "learning_rate": 9.99117906498089e-06,
+      "loss": 1.9609,
+      "step": 15
+    },
+    {
+      "epoch": 0.018658892128279883,
+      "grad_norm": 1.5065501928329468,
+      "learning_rate": 9.990586020240058e-06,
+      "loss": 1.9063,
+      "step": 16
+    },
+    {
+      "epoch": 0.019825072886297375,
+      "grad_norm": 0.6291128396987915,
+      "learning_rate": 9.98999234708895e-06,
+      "loss": 1.8891,
+      "step": 17
+    },
+    {
+      "epoch": 0.02099125364431487,
+      "grad_norm": 1.3838165998458862,
+      "learning_rate": 9.989398044528215e-06,
+      "loss": 2.0107,
+      "step": 18
+    },
+    {
+      "epoch": 0.022157434402332362,
+      "grad_norm": 0.4075084924697876,
+      "learning_rate": 9.988803111556368e-06,
+      "loss": 1.988,
+      "step": 19
+    },
+    {
+      "epoch": 0.023323615160349854,
+      "grad_norm": 0.3460110127925873,
+      "learning_rate": 9.988207547169812e-06,
+      "loss": 2.0267,
+      "step": 20
+    },
+    {
+      "epoch": 0.024489795918367346,
+      "grad_norm": 0.4190002381801605,
+      "learning_rate": 9.987611350362811e-06,
+      "loss": 1.9722,
+      "step": 21
+    },
+    {
+      "epoch": 0.02565597667638484,
+      "grad_norm": 0.5147244930267334,
+      "learning_rate": 9.987014520127494e-06,
+      "loss": 1.9987,
+      "step": 22
+    },
+    {
+      "epoch": 0.026822157434402333,
+      "grad_norm": 0.5430766940116882,
+      "learning_rate": 9.986417055453849e-06,
+      "loss": 1.9548,
+      "step": 23
+    },
+    {
+      "epoch": 0.027988338192419825,
+      "grad_norm": 0.48507723212242126,
+      "learning_rate": 9.98581895532971e-06,
+      "loss": 2.0099,
+      "step": 24
+    },
+    {
+      "epoch": 0.029154518950437316,
+      "grad_norm": 0.6483656764030457,
+      "learning_rate": 9.985220218740763e-06,
+      "loss": 1.9746,
+      "step": 25
+    },
+    {
+      "epoch": 0.030320699708454812,
+      "grad_norm": 0.24689602851867676,
+      "learning_rate": 9.984620844670532e-06,
+      "loss": 2.0528,
+      "step": 26
+    },
+    {
+      "epoch": 0.0314868804664723,
+      "grad_norm": 0.2205437421798706,
+      "learning_rate": 9.984020832100374e-06,
+      "loss": 1.9677,
+      "step": 27
+    },
+    {
+      "epoch": 0.0326530612244898,
+      "grad_norm": 0.2769107222557068,
+      "learning_rate": 9.983420180009475e-06,
+      "loss": 1.9209,
+      "step": 28
+    },
+    {
+      "epoch": 0.03381924198250729,
+      "grad_norm": 0.28950074315071106,
+      "learning_rate": 9.982818887374844e-06,
+      "loss": 1.937,
+      "step": 29
+    },
+    {
+      "epoch": 0.03498542274052478,
+      "grad_norm": 0.17652425169944763,
+      "learning_rate": 9.982216953171311e-06,
+      "loss": 1.9548,
+      "step": 30
+    },
+    {
+      "epoch": 0.036151603498542274,
+      "grad_norm": 0.18701910972595215,
+      "learning_rate": 9.981614376371509e-06,
+      "loss": 1.9435,
+      "step": 31
+    },
+    {
+      "epoch": 0.037317784256559766,
+      "grad_norm": 0.24653731286525726,
+      "learning_rate": 9.981011155945882e-06,
+      "loss": 1.9698,
+      "step": 32
+    },
+    {
+      "epoch": 0.03848396501457726,
+      "grad_norm": 0.17641796171665192,
+      "learning_rate": 9.980407290862674e-06,
+      "loss": 1.9009,
+      "step": 33
+    },
+    {
+      "epoch": 0.03965014577259475,
+      "grad_norm": 0.2894304692745209,
+      "learning_rate": 9.979802780087918e-06,
+      "loss": 2.0188,
+      "step": 34
+    },
+    {
+      "epoch": 0.04081632653061224,
+      "grad_norm": 0.16869935393333435,
+      "learning_rate": 9.97919762258544e-06,
+      "loss": 1.8692,
+      "step": 35
+    },
+    {
+      "epoch": 0.04198250728862974,
+      "grad_norm": 0.20803764462471008,
+      "learning_rate": 9.978591817316841e-06,
+      "loss": 1.9433,
+      "step": 36
+    },
+    {
+      "epoch": 0.04314868804664723,
+      "grad_norm": 0.1673920452594757,
+      "learning_rate": 9.977985363241508e-06,
+      "loss": 1.9946,
+      "step": 37
+    },
+    {
+      "epoch": 0.044314868804664724,
+      "grad_norm": 0.1684078872203827,
+      "learning_rate": 9.977378259316587e-06,
+      "loss": 1.9164,
+      "step": 38
+    },
+    {
+      "epoch": 0.045481049562682216,
+      "grad_norm": 0.2573232054710388,
+      "learning_rate": 9.976770504496993e-06,
+      "loss": 1.9789,
+      "step": 39
+    },
+    {
+      "epoch": 0.04664723032069971,
+      "grad_norm": 0.20336005091667175,
+      "learning_rate": 9.9761620977354e-06,
+      "loss": 1.8792,
+      "step": 40
+    },
+    {
+      "epoch": 0.0478134110787172,
+      "grad_norm": 0.16184398531913757,
+      "learning_rate": 9.975553037982233e-06,
+      "loss": 1.9105,
+      "step": 41
+    },
+    {
+      "epoch": 0.04897959183673469,
+      "grad_norm": 0.15008476376533508,
+      "learning_rate": 9.974943324185658e-06,
+      "loss": 1.8857,
+      "step": 42
+    },
+    {
+      "epoch": 0.05014577259475218,
+      "grad_norm": 0.32614490389823914,
+      "learning_rate": 9.97433295529159e-06,
+      "loss": 1.8527,
+      "step": 43
+    },
+    {
+      "epoch": 0.05131195335276968,
+      "grad_norm": 0.13648928701877594,
+      "learning_rate": 9.97372193024367e-06,
+      "loss": 1.9335,
+      "step": 44
+    },
+    {
+      "epoch": 0.052478134110787174,
+      "grad_norm": 0.21451526880264282,
+      "learning_rate": 9.97311024798327e-06,
+      "loss": 1.8619,
+      "step": 45
+    },
+    {
+      "epoch": 0.053644314868804666,
+      "grad_norm": 0.14589788019657135,
+      "learning_rate": 9.972497907449481e-06,
+      "loss": 1.8856,
+      "step": 46
+    },
+    {
+      "epoch": 0.05481049562682216,
+      "grad_norm": 0.14174826443195343,
+      "learning_rate": 9.971884907579111e-06,
+      "loss": 1.8943,
+      "step": 47
+    },
+    {
+      "epoch": 0.05597667638483965,
+      "grad_norm": 0.19878287613391876,
+      "learning_rate": 9.97127124730668e-06,
+      "loss": 1.9532,
+      "step": 48
+    },
+    {
+      "epoch": 0.05714285714285714,
+      "grad_norm": 0.14497295022010803,
+      "learning_rate": 9.970656925564406e-06,
+      "loss": 1.9392,
+      "step": 49
+    },
+    {
+      "epoch": 0.05830903790087463,
+      "grad_norm": 0.13562679290771484,
+      "learning_rate": 9.970041941282207e-06,
+      "loss": 1.9914,
+      "step": 50
+    },
+    {
+      "epoch": 0.05947521865889213,
+      "grad_norm": 0.14476607739925385,
+      "learning_rate": 9.969426293387687e-06,
+      "loss": 1.921,
+      "step": 51
+    },
+    {
+      "epoch": 0.060641399416909623,
+      "grad_norm": 0.13193249702453613,
+      "learning_rate": 9.968809980806143e-06,
+      "loss": 1.8814,
+      "step": 52
+    },
+    {
+      "epoch": 0.061807580174927115,
+      "grad_norm": 0.1686239391565323,
+      "learning_rate": 9.968193002460542e-06,
+      "loss": 1.8901,
+      "step": 53
+    },
+    {
+      "epoch": 0.0629737609329446,
+      "grad_norm": 0.14944055676460266,
+      "learning_rate": 9.967575357271527e-06,
+      "loss": 1.9515,
+      "step": 54
+    },
+    {
+      "epoch": 0.0641399416909621,
+      "grad_norm": 0.13094307482242584,
+      "learning_rate": 9.966957044157405e-06,
+      "loss": 1.8662,
+      "step": 55
+    },
+    {
+      "epoch": 0.0653061224489796,
+      "grad_norm": 0.6770932674407959,
+      "learning_rate": 9.966338062034144e-06,
+      "loss": 1.8838,
+      "step": 56
+    },
+    {
+      "epoch": 0.06647230320699708,
+      "grad_norm": 0.1337747573852539,
+      "learning_rate": 9.965718409815361e-06,
+      "loss": 1.8508,
+      "step": 57
+    },
+    {
+      "epoch": 0.06763848396501458,
+      "grad_norm": 0.12662266194820404,
+      "learning_rate": 9.965098086412324e-06,
+      "loss": 1.8902,
+      "step": 58
+    },
+    {
+      "epoch": 0.06880466472303207,
+      "grad_norm": 0.12304198741912842,
+      "learning_rate": 9.96447709073394e-06,
+      "loss": 1.8934,
+      "step": 59
+    },
+    {
+      "epoch": 0.06997084548104957,
+      "grad_norm": 0.49144798517227173,
+      "learning_rate": 9.963855421686748e-06,
+      "loss": 1.8585,
+      "step": 60
+    },
+    {
+      "epoch": 0.07113702623906705,
+      "grad_norm": 0.15562550723552704,
+      "learning_rate": 9.963233078174915e-06,
+      "loss": 1.9323,
+      "step": 61
+    },
+    {
+      "epoch": 0.07230320699708455,
+      "grad_norm": 0.13582547008991241,
+      "learning_rate": 9.96261005910023e-06,
+      "loss": 1.8194,
+      "step": 62
+    },
+    {
+      "epoch": 0.07346938775510205,
+      "grad_norm": 0.17720484733581543,
+      "learning_rate": 9.961986363362097e-06,
+      "loss": 1.9132,
+      "step": 63
+    },
+    {
+      "epoch": 0.07463556851311953,
+      "grad_norm": 0.21406219899654388,
+      "learning_rate": 9.961361989857523e-06,
+      "loss": 1.9139,
+      "step": 64
+    },
+    {
+      "epoch": 0.07580174927113703,
+      "grad_norm": 0.1265885829925537,
+      "learning_rate": 9.960736937481124e-06,
+      "loss": 1.9459,
+      "step": 65
+    },
+    {
+      "epoch": 0.07696793002915452,
+      "grad_norm": 0.12423531711101532,
+      "learning_rate": 9.960111205125106e-06,
+      "loss": 2.014,
+      "step": 66
+    },
+    {
+      "epoch": 0.07813411078717201,
+      "grad_norm": 0.16045193374156952,
+      "learning_rate": 9.959484791679266e-06,
+      "loss": 1.9674,
+      "step": 67
+    },
+    {
+      "epoch": 0.0793002915451895,
+      "grad_norm": 0.13593930006027222,
+      "learning_rate": 9.958857696030977e-06,
+      "loss": 1.858,
+      "step": 68
+    },
+    {
+      "epoch": 0.080466472303207,
+      "grad_norm": 0.1352033019065857,
+      "learning_rate": 9.9582299170652e-06,
+      "loss": 1.8328,
+      "step": 69
+    },
+    {
+      "epoch": 0.08163265306122448,
+      "grad_norm": 0.11946167796850204,
+      "learning_rate": 9.957601453664447e-06,
+      "loss": 1.9095,
+      "step": 70
+    },
+    {
+      "epoch": 0.08279883381924198,
+      "grad_norm": 0.12354989349842072,
+      "learning_rate": 9.956972304708807e-06,
+      "loss": 1.8284,
+      "step": 71
+    },
+    {
+      "epoch": 0.08396501457725948,
+      "grad_norm": 0.15010464191436768,
+      "learning_rate": 9.956342469075916e-06,
+      "loss": 1.9122,
+      "step": 72
+    },
+    {
+      "epoch": 0.08513119533527697,
+      "grad_norm": 0.1309327632188797,
+      "learning_rate": 9.955711945640964e-06,
+      "loss": 1.8555,
+      "step": 73
+    },
+    {
+      "epoch": 0.08629737609329446,
+      "grad_norm": 0.13184697926044464,
+      "learning_rate": 9.95508073327668e-06,
+      "loss": 1.9427,
+      "step": 74
+    },
+    {
+      "epoch": 0.08746355685131195,
+      "grad_norm": 0.137481227517128,
+      "learning_rate": 9.954448830853326e-06,
+      "loss": 1.9352,
+      "step": 75
+    },
+    {
+      "epoch": 0.08862973760932945,
+      "grad_norm": 0.14024440944194794,
+      "learning_rate": 9.953816237238699e-06,
+      "loss": 1.9089,
+      "step": 76
+    },
+    {
+      "epoch": 0.08979591836734693,
+      "grad_norm": 0.13411511480808258,
+      "learning_rate": 9.953182951298111e-06,
+      "loss": 1.8863,
+      "step": 77
+    },
+    {
+      "epoch": 0.09096209912536443,
+      "grad_norm": 0.12333246320486069,
+      "learning_rate": 9.952548971894392e-06,
+      "loss": 1.8156,
+      "step": 78
+    },
+    {
+      "epoch": 0.09212827988338193,
+      "grad_norm": 0.13011416792869568,
+      "learning_rate": 9.951914297887881e-06,
+      "loss": 1.8687,
+      "step": 79
+    },
+    {
+      "epoch": 0.09329446064139942,
+      "grad_norm": 0.1756213903427124,
+      "learning_rate": 9.95127892813642e-06,
+      "loss": 2.0011,
+      "step": 80
+    },
+    {
+      "epoch": 0.09446064139941691,
+      "grad_norm": 0.15610665082931519,
+      "learning_rate": 9.95064286149534e-06,
+      "loss": 1.9118,
+      "step": 81
+    },
+    {
+      "epoch": 0.0956268221574344,
+      "grad_norm": 0.14427155256271362,
+      "learning_rate": 9.950006096817461e-06,
+      "loss": 1.8729,
+      "step": 82
+    },
+    {
+      "epoch": 0.0967930029154519,
+      "grad_norm": 0.13449522852897644,
+      "learning_rate": 9.94936863295309e-06,
+      "loss": 1.8993,
+      "step": 83
+    },
+    {
+      "epoch": 0.09795918367346938,
+      "grad_norm": 0.13250373303890228,
+      "learning_rate": 9.948730468750002e-06,
+      "loss": 1.8482,
+      "step": 84
+    },
+    {
+      "epoch": 0.09912536443148688,
+      "grad_norm": 0.11890729516744614,
+      "learning_rate": 9.948091603053436e-06,
+      "loss": 1.8796,
+      "step": 85
+    },
+    {
+      "epoch": 0.10029154518950437,
+      "grad_norm": 0.14321458339691162,
+      "learning_rate": 9.947452034706099e-06,
+      "loss": 1.9092,
+      "step": 86
+    },
+    {
+      "epoch": 0.10029154518950437,
+      "eval_loss": 0.9654499888420105,
+      "eval_runtime": 168.1316,
+      "eval_samples_per_second": 41.063,
+      "eval_steps_per_second": 2.569,
+      "step": 86
+    },
+    {
+      "epoch": 0.10145772594752187,
+      "grad_norm": 0.1254715472459793,
+      "learning_rate": 9.946811762548146e-06,
+      "loss": 1.9509,
+      "step": 87
+    },
+    {
+      "epoch": 0.10262390670553936,
+      "grad_norm": 0.11887878179550171,
+      "learning_rate": 9.946170785417178e-06,
+      "loss": 1.9222,
+      "step": 88
+    },
+    {
+      "epoch": 0.10379008746355685,
+      "grad_norm": 0.12081245332956314,
+      "learning_rate": 9.945529102148236e-06,
+      "loss": 1.9551,
+      "step": 89
+    },
+    {
+      "epoch": 0.10495626822157435,
+      "grad_norm": 0.17569242417812347,
+      "learning_rate": 9.944886711573792e-06,
+      "loss": 1.868,
+      "step": 90
+    },
+    {
+      "epoch": 0.10612244897959183,
+      "grad_norm": 0.16713251173496246,
+      "learning_rate": 9.944243612523742e-06,
+      "loss": 1.933,
+      "step": 91
+    },
+    {
+      "epoch": 0.10728862973760933,
+      "grad_norm": 0.12986613810062408,
+      "learning_rate": 9.943599803825405e-06,
+      "loss": 1.8419,
+      "step": 92
+    },
+    {
+      "epoch": 0.10845481049562682,
+      "grad_norm": 0.12703189253807068,
+      "learning_rate": 9.942955284303503e-06,
+      "loss": 1.8522,
+      "step": 93
+    },
+    {
+      "epoch": 0.10962099125364431,
+      "grad_norm": 0.11318150907754898,
+      "learning_rate": 9.942310052780165e-06,
+      "loss": 1.9002,
+      "step": 94
+    },
+    {
+      "epoch": 0.11078717201166181,
+      "grad_norm": 0.13671989738941193,
+      "learning_rate": 9.941664108074917e-06,
+      "loss": 1.8989,
+      "step": 95
+    },
+    {
+      "epoch": 0.1119533527696793,
+      "grad_norm": 0.1254475861787796,
+      "learning_rate": 9.94101744900467e-06,
+      "loss": 1.812,
+      "step": 96
+    },
+    {
+      "epoch": 0.1131195335276968,
+      "grad_norm": 0.12093223631381989,
+      "learning_rate": 9.940370074383723e-06,
+      "loss": 1.9427,
+      "step": 97
+    },
+    {
+      "epoch": 0.11428571428571428,
+      "grad_norm": 0.12198259681463242,
+      "learning_rate": 9.939721983023743e-06,
+      "loss": 1.9005,
+      "step": 98
+    },
+    {
+      "epoch": 0.11545189504373178,
+      "grad_norm": 1.4697738885879517,
+      "learning_rate": 9.939073173733768e-06,
+      "loss": 1.8732,
+      "step": 99
+    },
+    {
+      "epoch": 0.11661807580174927,
+      "grad_norm": 0.1178608238697052,
+      "learning_rate": 9.938423645320198e-06,
+      "loss": 1.8417,
+      "step": 100
+    },
+    {
+      "epoch": 0.11778425655976676,
+      "grad_norm": 0.12745453417301178,
+      "learning_rate": 9.93777339658678e-06,
+      "loss": 1.8802,
+      "step": 101
+    },
+    {
+      "epoch": 0.11895043731778426,
+      "grad_norm": 0.12872348725795746,
+      "learning_rate": 9.937122426334608e-06,
+      "loss": 1.8452,
+      "step": 102
+    },
+    {
+      "epoch": 0.12011661807580175,
+      "grad_norm": 0.12067419290542603,
+      "learning_rate": 9.936470733362117e-06,
+      "loss": 1.9108,
+      "step": 103
+    },
+    {
+      "epoch": 0.12128279883381925,
+      "grad_norm": 0.13692283630371094,
+      "learning_rate": 9.935818316465072e-06,
+      "loss": 1.972,
+      "step": 104
+    },
+    {
+      "epoch": 0.12244897959183673,
+      "grad_norm": 0.17604570090770721,
+      "learning_rate": 9.935165174436556e-06,
+      "loss": 1.9011,
+      "step": 105
+    },
+    {
+      "epoch": 0.12361516034985423,
+      "grad_norm": 0.16985397040843964,
+      "learning_rate": 9.934511306066971e-06,
+      "loss": 1.8849,
+      "step": 106
+    },
+    {
+      "epoch": 0.12478134110787172,
+      "grad_norm": 0.12045660614967346,
+      "learning_rate": 9.933856710144032e-06,
+      "loss": 1.8937,
+      "step": 107
+    },
+    {
+      "epoch": 0.1259475218658892,
+      "grad_norm": 0.13032138347625732,
+      "learning_rate": 9.933201385452747e-06,
+      "loss": 1.9301,
+      "step": 108
+    },
+    {
+      "epoch": 0.1271137026239067,
+      "grad_norm": 0.12779614329338074,
+      "learning_rate": 9.93254533077542e-06,
+      "loss": 1.8464,
+      "step": 109
+    },
+    {
+      "epoch": 0.1282798833819242,
+      "grad_norm": 0.21776288747787476,
+      "learning_rate": 9.93188854489164e-06,
+      "loss": 1.9359,
+      "step": 110
+    },
+    {
+      "epoch": 0.1294460641399417,
+      "grad_norm": 0.26191726326942444,
+      "learning_rate": 9.93123102657828e-06,
+      "loss": 1.9041,
+      "step": 111
+    },
+    {
+      "epoch": 0.1306122448979592,
+      "grad_norm": 0.2020399272441864,
+      "learning_rate": 9.930572774609473e-06,
+      "loss": 1.8888,
+      "step": 112
+    },
+    {
+      "epoch": 0.13177842565597667,
+      "grad_norm": 0.31705787777900696,
+      "learning_rate": 9.929913787756623e-06,
+      "loss": 1.9028,
+      "step": 113
+    },
+    {
+      "epoch": 0.13294460641399417,
+      "grad_norm": 0.12636755406856537,
+      "learning_rate": 9.929254064788382e-06,
+      "loss": 1.9754,
+      "step": 114
+    },
+    {
+      "epoch": 0.13411078717201166,
+      "grad_norm": 0.1725548505783081,
+      "learning_rate": 9.928593604470663e-06,
+      "loss": 1.8643,
+      "step": 115
+    },
+    {
+      "epoch": 0.13527696793002916,
+      "grad_norm": 0.16371938586235046,
+      "learning_rate": 9.927932405566601e-06,
+      "loss": 1.9506,
+      "step": 116
+    },
+    {
+      "epoch": 0.13644314868804663,
+      "grad_norm": 0.17776437103748322,
+      "learning_rate": 9.927270466836576e-06,
+      "loss": 1.8809,
+      "step": 117
+    },
+    {
+      "epoch": 0.13760932944606413,
+      "grad_norm": 0.16379471123218536,
+      "learning_rate": 9.92660778703819e-06,
+      "loss": 1.9044,
+      "step": 118
+    },
+    {
+      "epoch": 0.13877551020408163,
+      "grad_norm": 0.1185624971985817,
+      "learning_rate": 9.925944364926257e-06,
+      "loss": 1.8955,
+      "step": 119
+    },
+    {
+      "epoch": 0.13994169096209913,
+      "grad_norm": 0.19759467244148254,
+      "learning_rate": 9.925280199252803e-06,
+      "loss": 1.8652,
+      "step": 120
+    },
+    {
+      "epoch": 0.14110787172011663,
+      "grad_norm": 0.1337408870458603,
+      "learning_rate": 9.924615288767056e-06,
+      "loss": 1.9349,
+      "step": 121
+    },
+    {
+      "epoch": 0.1422740524781341,
+      "grad_norm": 0.16124458611011505,
+      "learning_rate": 9.923949632215435e-06,
+      "loss": 1.8293,
+      "step": 122
+    },
+    {
+      "epoch": 0.1434402332361516,
+      "grad_norm": 0.15511973202228546,
+      "learning_rate": 9.923283228341547e-06,
+      "loss": 1.9239,
+      "step": 123
+    },
+    {
+      "epoch": 0.1446064139941691,
+      "grad_norm": 0.12374623119831085,
+      "learning_rate": 9.922616075886172e-06,
+      "loss": 1.945,
+      "step": 124
+    },
+    {
+      "epoch": 0.1457725947521866,
+      "grad_norm": 0.1404823213815689,
+      "learning_rate": 9.921948173587263e-06,
+      "loss": 1.9442,
+      "step": 125
+    },
+    {
+      "epoch": 0.1469387755102041,
+      "grad_norm": 0.36530742049217224,
+      "learning_rate": 9.921279520179933e-06,
+      "loss": 1.9364,
+      "step": 126
+    },
+    {
+      "epoch": 0.14810495626822157,
+      "grad_norm": 0.12016047537326813,
+      "learning_rate": 9.92061011439645e-06,
+      "loss": 1.8609,
+      "step": 127
+    },
+    {
+      "epoch": 0.14927113702623906,
+      "grad_norm": 0.1622726321220398,
+      "learning_rate": 9.919939954966225e-06,
+      "loss": 1.9189,
+      "step": 128
+    },
+    {
+      "epoch": 0.15043731778425656,
+      "grad_norm": 0.11802049726247787,
+      "learning_rate": 9.91926904061581e-06,
+      "loss": 1.8771,
+      "step": 129
+    },
+    {
+      "epoch": 0.15160349854227406,
+      "grad_norm": 0.20923790335655212,
+      "learning_rate": 9.91859737006888e-06,
+      "loss": 1.9106,
+      "step": 130
+    },
+    {
+      "epoch": 0.15276967930029153,
+      "grad_norm": 0.15893536806106567,
+      "learning_rate": 9.917924942046238e-06,
+      "loss": 1.8397,
+      "step": 131
+    },
+    {
+      "epoch": 0.15393586005830903,
+      "grad_norm": 0.18280337750911713,
+      "learning_rate": 9.9172517552658e-06,
+      "loss": 1.9626,
+      "step": 132
+    },
+    {
+      "epoch": 0.15510204081632653,
+      "grad_norm": 0.1171150654554367,
+      "learning_rate": 9.916577808442578e-06,
+      "loss": 1.8268,
+      "step": 133
+    },
+    {
+      "epoch": 0.15626822157434403,
+      "grad_norm": 0.13247933983802795,
+      "learning_rate": 9.915903100288692e-06,
+      "loss": 1.8632,
+      "step": 134
+    },
+    {
+      "epoch": 0.15743440233236153,
+      "grad_norm": 0.2886529266834259,
+      "learning_rate": 9.915227629513345e-06,
+      "loss": 1.7797,
+      "step": 135
+    },
+    {
+      "epoch": 0.158600583090379,
+      "grad_norm": 0.11958787590265274,
+      "learning_rate": 9.914551394822822e-06,
+      "loss": 1.8557,
+      "step": 136
+    },
+    {
+      "epoch": 0.1597667638483965,
+      "grad_norm": 0.1151299849152565,
+      "learning_rate": 9.913874394920476e-06,
+      "loss": 1.8642,
+      "step": 137
+    },
+    {
+      "epoch": 0.160932944606414,
+      "grad_norm": 0.21387127041816711,
+      "learning_rate": 9.91319662850673e-06,
+      "loss": 1.8882,
+      "step": 138
+    },
+    {
+      "epoch": 0.1620991253644315,
+      "grad_norm": 0.12334681302309036,
+      "learning_rate": 9.912518094279063e-06,
+      "loss": 1.8464,
+      "step": 139
+    },
+    {
+      "epoch": 0.16326530612244897,
+      "grad_norm": 0.1289202719926834,
+      "learning_rate": 9.911838790931991e-06,
+      "loss": 1.7959,
+      "step": 140
+    },
+    {
+      "epoch": 0.16443148688046647,
+      "grad_norm": 0.18751095235347748,
+      "learning_rate": 9.91115871715708e-06,
+      "loss": 1.9322,
+      "step": 141
+    },
+    {
+      "epoch": 0.16559766763848396,
+      "grad_norm": 0.14587146043777466,
+      "learning_rate": 9.91047787164292e-06,
+      "loss": 1.8752,
+      "step": 142
+    },
+    {
+      "epoch": 0.16676384839650146,
+      "grad_norm": 0.15966948866844177,
+      "learning_rate": 9.909796253075129e-06,
+      "loss": 1.8934,
+      "step": 143
+    },
+    {
+      "epoch": 0.16793002915451896,
+      "grad_norm": 0.2343795746564865,
+      "learning_rate": 9.90911386013633e-06,
+      "loss": 1.848,
+      "step": 144
+    },
+    {
+      "epoch": 0.16909620991253643,
+      "grad_norm": 0.15083608031272888,
+      "learning_rate": 9.908430691506158e-06,
+      "loss": 1.8428,
+      "step": 145
+    },
+    {
+      "epoch": 0.17026239067055393,
+      "grad_norm": 0.12175185233354568,
+      "learning_rate": 9.90774674586124e-06,
+      "loss": 1.9327,
+      "step": 146
+    },
+    {
+      "epoch": 0.17142857142857143,
+      "grad_norm": 0.3265966475009918,
+      "learning_rate": 9.907062021875198e-06,
+      "loss": 1.9093,
+      "step": 147
+    },
+    {
+      "epoch": 0.17259475218658893,
+      "grad_norm": 0.1465921700000763,
+      "learning_rate": 9.906376518218625e-06,
+      "loss": 1.8775,
+      "step": 148
+    },
+    {
+      "epoch": 0.17376093294460643,
+      "grad_norm": 0.1281740367412567,
+      "learning_rate": 9.905690233559086e-06,
+      "loss": 1.8778,
+      "step": 149
+    },
+    {
+      "epoch": 0.1749271137026239,
+      "grad_norm": 0.11814381927251816,
+      "learning_rate": 9.905003166561115e-06,
+      "loss": 1.8744,
+      "step": 150
+    },
+    {
+      "epoch": 0.1760932944606414,
+      "grad_norm": 0.14336846768856049,
+      "learning_rate": 9.904315315886195e-06,
+      "loss": 1.917,
+      "step": 151
+    },
+    {
+      "epoch": 0.1772594752186589,
+      "grad_norm": 0.12930937111377716,
+      "learning_rate": 9.903626680192748e-06,
+      "loss": 2.0256,
+      "step": 152
+    },
+    {
+      "epoch": 0.1784256559766764,
+      "grad_norm": 0.20058688521385193,
+      "learning_rate": 9.902937258136143e-06,
+      "loss": 1.8907,
+      "step": 153
+    },
+    {
+      "epoch": 0.17959183673469387,
+      "grad_norm": 0.11353437602519989,
+      "learning_rate": 9.902247048368668e-06,
+      "loss": 1.8572,
+      "step": 154
+    },
+    {
+      "epoch": 0.18075801749271136,
+      "grad_norm": 0.10796897113323212,
+      "learning_rate": 9.901556049539538e-06,
+      "loss": 1.8753,
+      "step": 155
+    },
+    {
+      "epoch": 0.18192419825072886,
+      "grad_norm": 0.21075373888015747,
+      "learning_rate": 9.900864260294866e-06,
+      "loss": 1.9091,
+      "step": 156
+    },
+    {
+      "epoch": 0.18309037900874636,
+      "grad_norm": 0.1726645678281784,
+      "learning_rate": 9.900171679277675e-06,
+      "loss": 1.8482,
+      "step": 157
+    },
+    {
+      "epoch": 0.18425655976676386,
+      "grad_norm": 0.1095050722360611,
+      "learning_rate": 9.899478305127878e-06,
+      "loss": 1.8419,
+      "step": 158
+    },
+    {
+      "epoch": 0.18542274052478133,
+      "grad_norm": 0.12067454308271408,
+      "learning_rate": 9.898784136482272e-06,
+      "loss": 1.8405,
+      "step": 159
+    },
+    {
+      "epoch": 0.18658892128279883,
+      "grad_norm": 0.2446606457233429,
+      "learning_rate": 9.898089171974523e-06,
+      "loss": 1.8559,
+      "step": 160
+    },
+    {
+      "epoch": 0.18775510204081633,
+      "grad_norm": 0.1668396145105362,
+      "learning_rate": 9.897393410235168e-06,
+      "loss": 1.8883,
+      "step": 161
+    },
+    {
+      "epoch": 0.18892128279883383,
+      "grad_norm": 0.10679066181182861,
+      "learning_rate": 9.896696849891597e-06,
+      "loss": 1.9672,
+      "step": 162
+    },
+    {
+      "epoch": 0.1900874635568513,
+      "grad_norm": 0.10319094359874725,
+      "learning_rate": 9.895999489568047e-06,
+      "loss": 1.9006,
+      "step": 163
+    },
+    {
+      "epoch": 0.1912536443148688,
+      "grad_norm": 0.12684528529644012,
+      "learning_rate": 9.895301327885598e-06,
+      "loss": 1.8878,
+      "step": 164
+    },
+    {
+      "epoch": 0.1924198250728863,
+      "grad_norm": 0.10842933505773544,
+      "learning_rate": 9.894602363462153e-06,
+      "loss": 1.8838,
+      "step": 165
+    },
+    {
+      "epoch": 0.1935860058309038,
+      "grad_norm": 0.13465608656406403,
+      "learning_rate": 9.89390259491244e-06,
+      "loss": 1.9329,
+      "step": 166
+    },
+    {
+      "epoch": 0.1947521865889213,
+      "grad_norm": 0.11536028981208801,
+      "learning_rate": 9.893202020847989e-06,
+      "loss": 1.9027,
+      "step": 167
+    },
+    {
+      "epoch": 0.19591836734693877,
+      "grad_norm": 0.10761712491512299,
+      "learning_rate": 9.892500639877144e-06,
+      "loss": 1.8692,
+      "step": 168
+    },
+    {
+      "epoch": 0.19708454810495626,
+      "grad_norm": 0.09428612142801285,
+      "learning_rate": 9.891798450605033e-06,
+      "loss": 1.8848,
+      "step": 169
+    },
+    {
+      "epoch": 0.19825072886297376,
+      "grad_norm": 0.15564614534378052,
+      "learning_rate": 9.89109545163357e-06,
+      "loss": 1.8733,
+      "step": 170
+    },
+    {
+      "epoch": 0.19941690962099126,
+      "grad_norm": 0.13192029297351837,
+      "learning_rate": 9.89039164156144e-06,
+      "loss": 1.9487,
+      "step": 171
+    },
+    {
+      "epoch": 0.20058309037900873,
+      "grad_norm": 0.1510130763053894,
+      "learning_rate": 9.889687018984096e-06,
+      "loss": 1.7885,
+      "step": 172
+    },
+    {
+      "epoch": 0.20058309037900873,
+      "eval_loss": 0.949571967124939,
+      "eval_runtime": 168.5132,
+      "eval_samples_per_second": 40.97,
+      "eval_steps_per_second": 2.564,
+      "step": 172
+    },
+    {
+      "epoch": 0.20174927113702623,
+      "grad_norm": 0.09452079981565475,
+      "learning_rate": 9.888981582493744e-06,
+      "loss": 1.8608,
+      "step": 173
+    },
+    {
+      "epoch": 0.20291545189504373,
+      "grad_norm": 0.4270099401473999,
+      "learning_rate": 9.888275330679338e-06,
+      "loss": 1.812,
+      "step": 174
+    },
+    {
+      "epoch": 0.20408163265306123,
+      "grad_norm": 0.11568140238523483,
+      "learning_rate": 9.887568262126567e-06,
+      "loss": 1.8958,
+      "step": 175
+    },
+    {
+      "epoch": 0.20524781341107873,
+      "grad_norm": 0.41217589378356934,
+      "learning_rate": 9.886860375417846e-06,
+      "loss": 1.8768,
+      "step": 176
+    },
+    {
+      "epoch": 0.2064139941690962,
+      "grad_norm": 0.09278150647878647,
+      "learning_rate": 9.88615166913231e-06,
+      "loss": 1.9148,
+      "step": 177
+    },
+    {
+      "epoch": 0.2075801749271137,
+      "grad_norm": 0.13748738169670105,
+      "learning_rate": 9.8854421418458e-06,
+      "loss": 1.8546,
+      "step": 178
+    },
+    {
+      "epoch": 0.2087463556851312,
+      "grad_norm": 0.0980186015367508,
+      "learning_rate": 9.884731792130854e-06,
+      "loss": 1.796,
+      "step": 179
+    },
+    {
+      "epoch": 0.2099125364431487,
+      "grad_norm": 0.12085171043872833,
+      "learning_rate": 9.884020618556703e-06,
+      "loss": 1.9873,
+      "step": 180
+    },
+    {
+      "epoch": 0.2110787172011662,
+      "grad_norm": 0.28499236702919006,
+      "learning_rate": 9.883308619689254e-06,
+      "loss": 1.9339,
+      "step": 181
+    },
+    {
+      "epoch": 0.21224489795918366,
+      "grad_norm": 0.1469694972038269,
+      "learning_rate": 9.882595794091086e-06,
+      "loss": 1.9111,
+      "step": 182
+    },
+    {
+      "epoch": 0.21341107871720116,
+      "grad_norm": 0.10729426890611649,
+      "learning_rate": 9.881882140321437e-06,
+      "loss": 1.8494,
+      "step": 183
+    },
+    {
+      "epoch": 0.21457725947521866,
+      "grad_norm": 0.12403575330972672,
+      "learning_rate": 9.881167656936193e-06,
+      "loss": 1.8495,
+      "step": 184
+    },
+    {
+      "epoch": 0.21574344023323616,
+      "grad_norm": 0.11913225799798965,
+      "learning_rate": 9.880452342487884e-06,
+      "loss": 1.9043,
+      "step": 185
+    },
+    {
+      "epoch": 0.21690962099125363,
+      "grad_norm": 0.14571070671081543,
+      "learning_rate": 9.879736195525672e-06,
+      "loss": 1.8098,
+      "step": 186
+    },
+    {
+      "epoch": 0.21807580174927113,
+      "grad_norm": 0.1274808794260025,
+      "learning_rate": 9.87901921459533e-06,
+      "loss": 1.8893,
+      "step": 187
+    },
+    {
+      "epoch": 0.21924198250728863,
+      "grad_norm": 0.20610937476158142,
+      "learning_rate": 9.878301398239255e-06,
+      "loss": 1.9279,
+      "step": 188
+    },
+    {
+      "epoch": 0.22040816326530613,
+      "grad_norm": 0.1038026213645935,
+      "learning_rate": 9.877582744996439e-06,
+      "loss": 1.8286,
+      "step": 189
+    },
+    {
+      "epoch": 0.22157434402332363,
+      "grad_norm": 0.10393866896629333,
+      "learning_rate": 9.876863253402465e-06,
+      "loss": 1.8868,
+      "step": 190
+    },
+    {
+      "epoch": 0.2227405247813411,
+      "grad_norm": 0.09725241363048553,
+      "learning_rate": 9.876142921989495e-06,
+      "loss": 1.926,
+      "step": 191
+    },
+    {
+      "epoch": 0.2239067055393586,
+      "grad_norm": 0.22062404453754425,
+      "learning_rate": 9.875421749286271e-06,
+      "loss": 1.833,
+      "step": 192
+    },
+    {
+      "epoch": 0.2250728862973761,
+      "grad_norm": 0.13754050433635712,
+      "learning_rate": 9.874699733818088e-06,
+      "loss": 1.8325,
+      "step": 193
+    },
+    {
+      "epoch": 0.2262390670553936,
+      "grad_norm": 0.09754852950572968,
+      "learning_rate": 9.873976874106797e-06,
+      "loss": 1.9138,
+      "step": 194
+    },
+    {
+      "epoch": 0.22740524781341107,
+      "grad_norm": 0.10769336670637131,
+      "learning_rate": 9.873253168670784e-06,
+      "loss": 1.8292,
+      "step": 195
+    },
+    {
+      "epoch": 0.22857142857142856,
+      "grad_norm": 0.1064218059182167,
+      "learning_rate": 9.872528616024974e-06,
+      "loss": 1.9038,
+      "step": 196
+    },
+    {
+      "epoch": 0.22973760932944606,
+      "grad_norm": 0.09421267360448837,
+      "learning_rate": 9.87180321468081e-06,
+      "loss": 1.8231,
+      "step": 197
+    },
+    {
+      "epoch": 0.23090379008746356,
+      "grad_norm": 0.10365895926952362,
+      "learning_rate": 9.871076963146245e-06,
+      "loss": 1.8454,
+      "step": 198
+    },
+    {
+      "epoch": 0.23206997084548106,
+      "grad_norm": 0.13903117179870605,
+      "learning_rate": 9.870349859925729e-06,
+      "loss": 1.9344,
+      "step": 199
+    },
+    {
+      "epoch": 0.23323615160349853,
+      "grad_norm": 0.09399236738681793,
+      "learning_rate": 9.86962190352021e-06,
+      "loss": 1.8837,
+      "step": 200
+    },
+    {
+      "epoch": 0.23440233236151603,
+      "grad_norm": 0.17643000185489655,
+      "learning_rate": 9.86889309242711e-06,
+      "loss": 1.8651,
+      "step": 201
+    },
+    {
+      "epoch": 0.23556851311953353,
+      "grad_norm": 0.1239488422870636,
+      "learning_rate": 9.868163425140322e-06,
+      "loss": 1.9106,
+      "step": 202
+    },
+    {
+      "epoch": 0.23673469387755103,
+      "grad_norm": 0.27507391571998596,
+      "learning_rate": 9.8674329001502e-06,
+      "loss": 1.9229,
+      "step": 203
+    },
+    {
+      "epoch": 0.23790087463556853,
+      "grad_norm": 0.11207671463489532,
+      "learning_rate": 9.866701515943544e-06,
+      "loss": 1.9696,
+      "step": 204
+    },
+    {
+      "epoch": 0.239067055393586,
+      "grad_norm": 0.9817244410514832,
+      "learning_rate": 9.865969271003597e-06,
+      "loss": 1.8101,
+      "step": 205
+    },
+    {
+      "epoch": 0.2402332361516035,
+      "grad_norm": 0.10202158987522125,
+      "learning_rate": 9.865236163810024e-06,
+      "loss": 1.9546,
+      "step": 206
+    },
+    {
+      "epoch": 0.241399416909621,
+      "grad_norm": 0.16548047959804535,
+      "learning_rate": 9.86450219283891e-06,
+      "loss": 1.8836,
+      "step": 207
+    },
+    {
+      "epoch": 0.2425655976676385,
+      "grad_norm": 1.556760311126709,
+      "learning_rate": 9.863767356562746e-06,
+      "loss": 1.8732,
+      "step": 208
+    },
+    {
+      "epoch": 0.24373177842565596,
+      "grad_norm": 0.09632741659879684,
+      "learning_rate": 9.863031653450425e-06,
+      "loss": 1.9391,
+      "step": 209
+    },
+    {
+      "epoch": 0.24489795918367346,
+      "grad_norm": 0.09008914977312088,
+      "learning_rate": 9.862295081967214e-06,
+      "loss": 1.8147,
+      "step": 210
+    },
+    {
+      "epoch": 0.24606413994169096,
+      "grad_norm": 0.15885330736637115,
+      "learning_rate": 9.861557640574766e-06,
+      "loss": 1.9179,
+      "step": 211
+    },
+    {
+      "epoch": 0.24723032069970846,
+      "grad_norm": 0.09356313943862915,
+      "learning_rate": 9.860819327731092e-06,
+      "loss": 1.9172,
+      "step": 212
+    },
+    {
+      "epoch": 0.24839650145772596,
+      "grad_norm": 0.12561793625354767,
+      "learning_rate": 9.86008014189056e-06,
+      "loss": 1.8416,
+      "step": 213
+    },
+    {
+      "epoch": 0.24956268221574343,
+      "grad_norm": 0.1658526062965393,
+      "learning_rate": 9.859340081503879e-06,
+      "loss": 1.8412,
+      "step": 214
+    },
+    {
+      "epoch": 0.25072886297376096,
+      "grad_norm": 0.15473313629627228,
+      "learning_rate": 9.858599145018088e-06,
+      "loss": 1.948,
+      "step": 215
+    },
+    {
+      "epoch": 0.2518950437317784,
+      "grad_norm": 0.09785402566194534,
+      "learning_rate": 9.857857330876546e-06,
+      "loss": 1.8449,
+      "step": 216
+    },
+    {
+      "epoch": 0.2530612244897959,
+      "grad_norm": 0.11832020431756973,
+      "learning_rate": 9.857114637518932e-06,
+      "loss": 1.8696,
+      "step": 217
+    },
+    {
+      "epoch": 0.2542274052478134,
+      "grad_norm": 0.11395788937807083,
+      "learning_rate": 9.85637106338121e-06,
+      "loss": 1.9043,
+      "step": 218
+    },
+    {
+      "epoch": 0.2553935860058309,
+      "grad_norm": 0.08991193026304245,
+      "learning_rate": 9.855626606895643e-06,
+      "loss": 1.9178,
+      "step": 219
+    },
+    {
+      "epoch": 0.2565597667638484,
+      "grad_norm": 0.09471415728330612,
+      "learning_rate": 9.854881266490766e-06,
+      "loss": 1.8048,
+      "step": 220
+    },
+    {
+      "epoch": 0.2577259475218659,
+      "grad_norm": 0.09410014003515244,
+      "learning_rate": 9.854135040591381e-06,
+      "loss": 1.8611,
+      "step": 221
+    },
+    {
+      "epoch": 0.2588921282798834,
+      "grad_norm": 0.10619162768125534,
+      "learning_rate": 9.853387927618544e-06,
+      "loss": 1.9756,
+      "step": 222
+    },
+    {
+      "epoch": 0.2600583090379009,
+      "grad_norm": 0.1283230185508728,
+      "learning_rate": 9.85263992598956e-06,
+      "loss": 1.9104,
+      "step": 223
+    },
+    {
+      "epoch": 0.2612244897959184,
+      "grad_norm": 0.09683047980070114,
+      "learning_rate": 9.851891034117959e-06,
+      "loss": 1.9643,
+      "step": 224
+    },
+    {
+      "epoch": 0.26239067055393583,
+      "grad_norm": 0.09329723566770554,
+      "learning_rate": 9.851141250413497e-06,
+      "loss": 1.8649,
+      "step": 225
+    },
+    {
+      "epoch": 0.26355685131195333,
+      "grad_norm": 0.8949700593948364,
+      "learning_rate": 9.85039057328214e-06,
+      "loss": 1.8992,
+      "step": 226
+    },
+    {
+      "epoch": 0.26472303206997083,
+      "grad_norm": 0.09303262829780579,
+      "learning_rate": 9.849639001126052e-06,
+      "loss": 1.8434,
+      "step": 227
+    },
+    {
+      "epoch": 0.26588921282798833,
+      "grad_norm": 0.40761691331863403,
+      "learning_rate": 9.848886532343585e-06,
+      "loss": 1.9339,
+      "step": 228
+    },
+    {
+      "epoch": 0.26705539358600583,
+      "grad_norm": 0.11234267055988312,
+      "learning_rate": 9.848133165329266e-06,
+      "loss": 1.8652,
+      "step": 229
+    },
+    {
+      "epoch": 0.26822157434402333,
+      "grad_norm": 0.1065870076417923,
+      "learning_rate": 9.84737889847379e-06,
+      "loss": 1.8369,
+      "step": 230
+    },
+    {
+      "epoch": 0.2693877551020408,
+      "grad_norm": 0.11718452721834183,
+      "learning_rate": 9.846623730164e-06,
+      "loss": 1.8962,
+      "step": 231
+    },
+    {
+      "epoch": 0.2705539358600583,
+      "grad_norm": 0.09421117603778839,
+      "learning_rate": 9.845867658782887e-06,
+      "loss": 1.9355,
+      "step": 232
+    },
+    {
+      "epoch": 0.2717201166180758,
+      "grad_norm": 0.16175977885723114,
+      "learning_rate": 9.845110682709568e-06,
+      "loss": 1.952,
+      "step": 233
+    },
+    {
+      "epoch": 0.27288629737609327,
+      "grad_norm": 0.15923438966274261,
+      "learning_rate": 9.844352800319277e-06,
+      "loss": 1.8092,
+      "step": 234
+    },
+    {
+      "epoch": 0.27405247813411077,
+      "grad_norm": 0.12237056344747543,
+      "learning_rate": 9.843594009983361e-06,
+      "loss": 1.8877,
+      "step": 235
+    },
+    {
+      "epoch": 0.27521865889212827,
+      "grad_norm": 0.13256801664829254,
+      "learning_rate": 9.84283431006926e-06,
+      "loss": 1.9086,
+      "step": 236
+    },
+    {
+      "epoch": 0.27638483965014576,
+      "grad_norm": 0.09711961448192596,
+      "learning_rate": 9.842073698940495e-06,
+      "loss": 1.8721,
+      "step": 237
+    },
+    {
+      "epoch": 0.27755102040816326,
+      "grad_norm": 0.09064258635044098,
+      "learning_rate": 9.841312174956662e-06,
+      "loss": 1.9445,
+      "step": 238
+    },
+    {
+      "epoch": 0.27871720116618076,
+      "grad_norm": 0.10512492060661316,
+      "learning_rate": 9.840549736473415e-06,
+      "loss": 1.9178,
+      "step": 239
+    },
+    {
+      "epoch": 0.27988338192419826,
+      "grad_norm": 0.09276045858860016,
+      "learning_rate": 9.839786381842456e-06,
+      "loss": 1.8344,
+      "step": 240
+    },
+    {
+      "epoch": 0.28104956268221576,
+      "grad_norm": 0.08836158365011215,
+      "learning_rate": 9.83902210941153e-06,
+      "loss": 1.8755,
+      "step": 241
+    },
+    {
+      "epoch": 0.28221574344023326,
+      "grad_norm": 0.09765245020389557,
+      "learning_rate": 9.838256917524396e-06,
+      "loss": 1.7981,
+      "step": 242
+    },
+    {
+      "epoch": 0.28338192419825076,
+      "grad_norm": 0.10735093802213669,
+      "learning_rate": 9.837490804520832e-06,
+      "loss": 1.9087,
+      "step": 243
+    },
+    {
+      "epoch": 0.2845481049562682,
+      "grad_norm": 0.11153509467840195,
+      "learning_rate": 9.836723768736618e-06,
+      "loss": 1.9154,
+      "step": 244
+    },
+    {
+      "epoch": 0.2857142857142857,
+      "grad_norm": 0.15119706094264984,
+      "learning_rate": 9.835955808503516e-06,
+      "loss": 1.7997,
+      "step": 245
+    },
+    {
+      "epoch": 0.2868804664723032,
+      "grad_norm": 0.08542892336845398,
+      "learning_rate": 9.835186922149271e-06,
+      "loss": 1.8185,
+      "step": 246
+    },
+    {
+      "epoch": 0.2880466472303207,
+      "grad_norm": 0.08899488300085068,
+      "learning_rate": 9.834417107997587e-06,
+      "loss": 1.9115,
+      "step": 247
+    },
+    {
+      "epoch": 0.2892128279883382,
+      "grad_norm": 0.11924991011619568,
+      "learning_rate": 9.833646364368125e-06,
+      "loss": 1.8846,
+      "step": 248
+    },
+    {
+      "epoch": 0.2903790087463557,
+      "grad_norm": 0.2538813650608063,
+      "learning_rate": 9.832874689576482e-06,
+      "loss": 1.9753,
+      "step": 249
+    },
+    {
+      "epoch": 0.2915451895043732,
+      "grad_norm": 0.1084447056055069,
+      "learning_rate": 9.832102081934186e-06,
+      "loss": 1.8326,
+      "step": 250
+    },
+    {
+      "epoch": 0.2927113702623907,
+      "grad_norm": 0.10492334514856339,
+      "learning_rate": 9.831328539748674e-06,
+      "loss": 1.8564,
+      "step": 251
+    },
+    {
+      "epoch": 0.2938775510204082,
+      "grad_norm": 0.08830749243497849,
+      "learning_rate": 9.830554061323293e-06,
+      "loss": 1.9334,
+      "step": 252
+    },
+    {
+      "epoch": 0.29504373177842563,
+      "grad_norm": 0.0963321253657341,
+      "learning_rate": 9.829778644957278e-06,
+      "loss": 1.8599,
+      "step": 253
+    },
+    {
+      "epoch": 0.29620991253644313,
+      "grad_norm": 0.1080334261059761,
+      "learning_rate": 9.82900228894574e-06,
+      "loss": 1.9407,
+      "step": 254
+    },
+    {
+      "epoch": 0.29737609329446063,
+      "grad_norm": 0.08521893620491028,
+      "learning_rate": 9.828224991579658e-06,
+      "loss": 1.8635,
+      "step": 255
+    },
+    {
+      "epoch": 0.29854227405247813,
+      "grad_norm": 0.0925750881433487,
+      "learning_rate": 9.827446751145862e-06,
+      "loss": 1.9474,
+      "step": 256
+    },
+    {
+      "epoch": 0.29970845481049563,
+      "grad_norm": 0.09808698296546936,
+      "learning_rate": 9.826667565927025e-06,
+      "loss": 1.8612,
+      "step": 257
+    },
+    {
+      "epoch": 0.3008746355685131,
+      "grad_norm": 0.8018088936805725,
+      "learning_rate": 9.825887434201648e-06,
+      "loss": 2.0059,
+      "step": 258
+    },
+    {
+      "epoch": 0.3008746355685131,
+      "eval_loss": 0.9415693283081055,
+      "eval_runtime": 168.9757,
+      "eval_samples_per_second": 40.858,
+      "eval_steps_per_second": 2.557,
+      "step": 258
+    },
+    {
+      "epoch": 0.3020408163265306,
+      "grad_norm": 0.088890939950943,
+      "learning_rate": 9.825106354244042e-06,
+      "loss": 1.941,
+      "step": 259
+    },
+    {
+      "epoch": 0.3032069970845481,
+      "grad_norm": 0.10895935446023941,
+      "learning_rate": 9.824324324324325e-06,
+      "loss": 1.9818,
+      "step": 260
+    },
+    {
+      "epoch": 0.3043731778425656,
+      "grad_norm": 0.19524423778057098,
+      "learning_rate": 9.823541342708404e-06,
+      "loss": 1.9107,
+      "step": 261
+    },
+    {
+      "epoch": 0.30553935860058307,
+      "grad_norm": 0.11606555432081223,
+      "learning_rate": 9.822757407657963e-06,
+      "loss": 1.8226,
+      "step": 262
+    },
+    {
+      "epoch": 0.30670553935860057,
+      "grad_norm": 0.088669553399086,
+      "learning_rate": 9.821972517430449e-06,
+      "loss": 1.8858,
+      "step": 263
+    },
+    {
+      "epoch": 0.30787172011661806,
+      "grad_norm": 0.1167503297328949,
+      "learning_rate": 9.821186670279059e-06,
+      "loss": 1.9284,
+      "step": 264
+    },
+    {
+      "epoch": 0.30903790087463556,
+      "grad_norm": 0.0924626961350441,
+      "learning_rate": 9.82039986445273e-06,
+      "loss": 1.792,
+      "step": 265
+    },
+    {
+      "epoch": 0.31020408163265306,
+      "grad_norm": 0.7967892289161682,
+      "learning_rate": 9.819612098196122e-06,
+      "loss": 1.8627,
+      "step": 266
+    },
+    {
+      "epoch": 0.31137026239067056,
+      "grad_norm": 0.08940774202346802,
+      "learning_rate": 9.81882336974961e-06,
+      "loss": 1.8399,
+      "step": 267
+    },
+    {
+      "epoch": 0.31253644314868806,
+      "grad_norm": 0.13215725123882294,
+      "learning_rate": 9.818033677349268e-06,
+      "loss": 1.9075,
+      "step": 268
+    },
+    {
+      "epoch": 0.31370262390670556,
+      "grad_norm": 0.08561773598194122,
+      "learning_rate": 9.81724301922685e-06,
+      "loss": 1.8757,
+      "step": 269
+    },
+    {
+      "epoch": 0.31486880466472306,
+      "grad_norm": 0.09795838594436646,
+      "learning_rate": 9.81645139360979e-06,
+      "loss": 1.8794,
+      "step": 270
+    },
+    {
+      "epoch": 0.3160349854227405,
+      "grad_norm": 0.1317494660615921,
+      "learning_rate": 9.815658798721176e-06,
+      "loss": 1.831,
+      "step": 271
+    },
+    {
+      "epoch": 0.317201166180758,
+      "grad_norm": 0.11524375528097153,
+      "learning_rate": 9.814865232779745e-06,
+      "loss": 1.8672,
+      "step": 272
+    },
+    {
+      "epoch": 0.3183673469387755,
+      "grad_norm": 0.10002207010984421,
+      "learning_rate": 9.814070693999865e-06,
+      "loss": 1.8475,
+      "step": 273
+    },
+    {
+      "epoch": 0.319533527696793,
+      "grad_norm": 0.08794574439525604,
+      "learning_rate": 9.813275180591522e-06,
+      "loss": 1.8453,
+      "step": 274
+    },
+    {
+      "epoch": 0.3206997084548105,
+      "grad_norm": 0.10029341280460358,
+      "learning_rate": 9.812478690760315e-06,
+      "loss": 1.8635,
+      "step": 275
+    },
+    {
+      "epoch": 0.321865889212828,
+      "grad_norm": 0.10724449157714844,
+      "learning_rate": 9.811681222707424e-06,
+      "loss": 1.8943,
+      "step": 276
+    },
+    {
+      "epoch": 0.3230320699708455,
+      "grad_norm": 0.11051566153764725,
+      "learning_rate": 9.810882774629618e-06,
+      "loss": 1.9233,
+      "step": 277
+    },
+    {
+      "epoch": 0.324198250728863,
+      "grad_norm": 0.08395867794752121,
+      "learning_rate": 9.810083344719224e-06,
+      "loss": 1.903,
+      "step": 278
+    },
+    {
+      "epoch": 0.3253644314868805,
+      "grad_norm": 0.09595666825771332,
+      "learning_rate": 9.809282931164126e-06,
+      "loss": 1.9825,
+      "step": 279
+    },
+    {
+      "epoch": 0.32653061224489793,
+      "grad_norm": 0.10128135234117508,
+      "learning_rate": 9.808481532147745e-06,
+      "loss": 1.8628,
+      "step": 280
+    },
+    {
+      "epoch": 0.32769679300291543,
+      "grad_norm": 0.08685976266860962,
+      "learning_rate": 9.807679145849018e-06,
+      "loss": 1.7714,
+      "step": 281
+    },
+    {
+      "epoch": 0.32886297376093293,
+      "grad_norm": 0.08862022310495377,
+      "learning_rate": 9.806875770442406e-06,
+      "loss": 1.8677,
+      "step": 282
+    },
+    {
+      "epoch": 0.33002915451895043,
+      "grad_norm": 0.09113382548093796,
+      "learning_rate": 9.806071404097857e-06,
+      "loss": 1.9141,
+      "step": 283
+    },
+    {
+      "epoch": 0.33119533527696793,
+      "grad_norm": 0.09103240817785263,
+      "learning_rate": 9.805266044980803e-06,
+      "loss": 1.8778,
+      "step": 284
+    },
+    {
+      "epoch": 0.3323615160349854,
+      "grad_norm": 0.11897517740726471,
+      "learning_rate": 9.804459691252145e-06,
+      "loss": 1.8216,
+      "step": 285
+    },
+    {
+      "epoch": 0.3335276967930029,
+      "grad_norm": 0.09001608937978745,
+      "learning_rate": 9.803652341068241e-06,
+      "loss": 1.8742,
+      "step": 286
+    },
+    {
+      "epoch": 0.3346938775510204,
+      "grad_norm": 0.08359828591346741,
+      "learning_rate": 9.80284399258089e-06,
+      "loss": 1.8532,
+      "step": 287
+    },
+    {
+      "epoch": 0.3358600583090379,
+      "grad_norm": 0.08774221688508987,
+      "learning_rate": 9.802034643937313e-06,
+      "loss": 1.7951,
+      "step": 288
+    },
+    {
+      "epoch": 0.33702623906705537,
+      "grad_norm": 0.1321367472410202,
+      "learning_rate": 9.801224293280144e-06,
+      "loss": 1.9615,
+      "step": 289
+    },
+    {
+      "epoch": 0.33819241982507287,
+      "grad_norm": 0.09782902896404266,
+      "learning_rate": 9.80041293874742e-06,
+      "loss": 1.8581,
+      "step": 290
+    },
+    {
+      "epoch": 0.33935860058309036,
+      "grad_norm": 0.08564814925193787,
+      "learning_rate": 9.79960057847256e-06,
+      "loss": 1.9063,
+      "step": 291
+    },
+    {
+      "epoch": 0.34052478134110786,
+      "grad_norm": 0.2194901406764984,
+      "learning_rate": 9.798787210584345e-06,
+      "loss": 1.8307,
+      "step": 292
+    },
+    {
+      "epoch": 0.34169096209912536,
+      "grad_norm": 0.09886486083269119,
+      "learning_rate": 9.797972833206923e-06,
+      "loss": 1.8445,
+      "step": 293
+    },
+    {
+      "epoch": 0.34285714285714286,
+      "grad_norm": 0.11643076688051224,
+      "learning_rate": 9.797157444459777e-06,
+      "loss": 1.866,
+      "step": 294
+    },
+    {
+      "epoch": 0.34402332361516036,
+      "grad_norm": 0.10743511468172073,
+      "learning_rate": 9.796341042457716e-06,
+      "loss": 1.823,
+      "step": 295
+    },
+    {
+      "epoch": 0.34518950437317786,
+      "grad_norm": 0.09972105920314789,
+      "learning_rate": 9.795523625310861e-06,
+      "loss": 1.8057,
+      "step": 296
+    },
+    {
+      "epoch": 0.34635568513119536,
+      "grad_norm": 0.19049879908561707,
+      "learning_rate": 9.79470519112463e-06,
+      "loss": 1.9023,
+      "step": 297
+    },
+    {
+      "epoch": 0.34752186588921286,
+      "grad_norm": 0.08880095183849335,
+      "learning_rate": 9.793885737999724e-06,
+      "loss": 1.9038,
+      "step": 298
+    },
+    {
+      "epoch": 0.3486880466472303,
+      "grad_norm": 0.10895738750696182,
+      "learning_rate": 9.793065264032115e-06,
+      "loss": 1.8915,
+      "step": 299
+    },
+    {
+      "epoch": 0.3498542274052478,
+      "grad_norm": 0.09476173669099808,
+      "learning_rate": 9.79224376731302e-06,
+      "loss": 1.9186,
+      "step": 300
+    },
+    {
+      "epoch": 0.3510204081632653,
+      "grad_norm": 0.10094133019447327,
+      "learning_rate": 9.791421245928904e-06,
+      "loss": 1.9446,
+      "step": 301
+    },
+    {
+      "epoch": 0.3521865889212828,
+      "grad_norm": 0.10069712996482849,
+      "learning_rate": 9.790597697961448e-06,
+      "loss": 1.86,
+      "step": 302
+    },
+    {
+      "epoch": 0.3533527696793003,
+      "grad_norm": 0.11581845581531525,
+      "learning_rate": 9.789773121487547e-06,
+      "loss": 2.0159,
+      "step": 303
+    },
+    {
+      "epoch": 0.3545189504373178,
+      "grad_norm": 0.09166114032268524,
+      "learning_rate": 9.788947514579284e-06,
+      "loss": 1.8884,
+      "step": 304
+    },
+    {
+      "epoch": 0.3556851311953353,
+      "grad_norm": 0.09281602501869202,
+      "learning_rate": 9.788120875303926e-06,
+      "loss": 1.8371,
+      "step": 305
+    },
+    {
+      "epoch": 0.3568513119533528,
+      "grad_norm": 0.1167796403169632,
+      "learning_rate": 9.7872932017239e-06,
+      "loss": 1.8982,
+      "step": 306
+    },
+    {
+      "epoch": 0.3580174927113703,
+      "grad_norm": 0.15306980907917023,
+      "learning_rate": 9.786464491896781e-06,
+      "loss": 1.9376,
+      "step": 307
+    },
+    {
+      "epoch": 0.35918367346938773,
+      "grad_norm": 0.10240183025598526,
+      "learning_rate": 9.78563474387528e-06,
+      "loss": 1.9255,
+      "step": 308
+    },
+    {
+      "epoch": 0.36034985422740523,
+      "grad_norm": 0.10198673605918884,
+      "learning_rate": 9.784803955707223e-06,
+      "loss": 1.8813,
+      "step": 309
+    },
+    {
+      "epoch": 0.36151603498542273,
+      "grad_norm": 0.13608044385910034,
+      "learning_rate": 9.78397212543554e-06,
+      "loss": 1.8732,
+      "step": 310
+    },
+    {
+      "epoch": 0.36268221574344023,
+      "grad_norm": 0.27615493535995483,
+      "learning_rate": 9.783139251098251e-06,
+      "loss": 1.8554,
+      "step": 311
+    },
+    {
+      "epoch": 0.3638483965014577,
+      "grad_norm": 0.10219445824623108,
+      "learning_rate": 9.78230533072844e-06,
+      "loss": 1.9031,
+      "step": 312
+    },
+    {
+      "epoch": 0.3650145772594752,
+      "grad_norm": 0.08255916833877563,
+      "learning_rate": 9.781470362354256e-06,
+      "loss": 1.8441,
+      "step": 313
+    },
+    {
+      "epoch": 0.3661807580174927,
+      "grad_norm": 0.09833742678165436,
+      "learning_rate": 9.780634343998883e-06,
+      "loss": 1.8956,
+      "step": 314
+    },
+    {
+      "epoch": 0.3673469387755102,
+      "grad_norm": 0.14747004210948944,
+      "learning_rate": 9.779797273680532e-06,
+      "loss": 1.8644,
+      "step": 315
+    },
+    {
+      "epoch": 0.3685131195335277,
+      "grad_norm": 0.1037067174911499,
+      "learning_rate": 9.778959149412424e-06,
+      "loss": 1.8639,
+      "step": 316
+    },
+    {
+      "epoch": 0.36967930029154517,
+      "grad_norm": 0.09124871343374252,
+      "learning_rate": 9.778119969202773e-06,
+      "loss": 1.9815,
+      "step": 317
+    },
+    {
+      "epoch": 0.37084548104956266,
+      "grad_norm": 0.11458101123571396,
+      "learning_rate": 9.77727973105477e-06,
+      "loss": 1.9069,
+      "step": 318
+    },
+    {
+      "epoch": 0.37201166180758016,
+      "grad_norm": 0.10865851491689682,
+      "learning_rate": 9.776438432966572e-06,
+      "loss": 1.9049,
+      "step": 319
+    },
+    {
+      "epoch": 0.37317784256559766,
+      "grad_norm": 0.09231556951999664,
+      "learning_rate": 9.775596072931277e-06,
+      "loss": 1.8765,
+      "step": 320
+    },
+    {
+      "epoch": 0.37434402332361516,
+      "grad_norm": 0.0897209569811821,
+      "learning_rate": 9.774752648936917e-06,
+      "loss": 1.9162,
+      "step": 321
+    },
+    {
+      "epoch": 0.37551020408163266,
+      "grad_norm": 0.09530792385339737,
+      "learning_rate": 9.773908158966437e-06,
+      "loss": 1.8334,
+      "step": 322
+    },
+    {
+      "epoch": 0.37667638483965016,
+      "grad_norm": 0.09419985115528107,
+      "learning_rate": 9.773062600997683e-06,
+      "loss": 1.8457,
+      "step": 323
+    },
+    {
+      "epoch": 0.37784256559766766,
+      "grad_norm": 0.1203305646777153,
+      "learning_rate": 9.772215973003374e-06,
+      "loss": 1.8586,
+      "step": 324
+    },
+    {
+      "epoch": 0.37900874635568516,
+      "grad_norm": 0.10866625607013702,
+      "learning_rate": 9.771368272951109e-06,
+      "loss": 1.8052,
+      "step": 325
+    },
+    {
+      "epoch": 0.3801749271137026,
+      "grad_norm": 0.08843393623828888,
+      "learning_rate": 9.770519498803324e-06,
+      "loss": 1.8647,
+      "step": 326
+    },
+    {
+      "epoch": 0.3813411078717201,
+      "grad_norm": 0.10051094740629196,
+      "learning_rate": 9.769669648517294e-06,
+      "loss": 1.9072,
+      "step": 327
+    },
+    {
+      "epoch": 0.3825072886297376,
+      "grad_norm": 0.08930347859859467,
+      "learning_rate": 9.768818720045108e-06,
+      "loss": 1.8767,
+      "step": 328
+    },
+    {
+      "epoch": 0.3836734693877551,
+      "grad_norm": 0.08803662657737732,
+      "learning_rate": 9.767966711333664e-06,
+      "loss": 1.8535,
+      "step": 329
+    },
+    {
+      "epoch": 0.3848396501457726,
+      "grad_norm": 0.11282350867986679,
+      "learning_rate": 9.76711362032463e-06,
+      "loss": 1.9113,
+      "step": 330
+    },
+    {
+      "epoch": 0.3860058309037901,
+      "grad_norm": 0.08783852308988571,
+      "learning_rate": 9.766259444954453e-06,
+      "loss": 1.9796,
+      "step": 331
+    },
+    {
+      "epoch": 0.3871720116618076,
+      "grad_norm": 0.1171242818236351,
+      "learning_rate": 9.765404183154326e-06,
+      "loss": 1.9008,
+      "step": 332
+    },
+    {
+      "epoch": 0.3883381924198251,
+      "grad_norm": 0.11021047830581665,
+      "learning_rate": 9.764547832850175e-06,
+      "loss": 1.8946,
+      "step": 333
+    },
+    {
+      "epoch": 0.3895043731778426,
+      "grad_norm": 0.09998290985822678,
+      "learning_rate": 9.763690391962645e-06,
+      "loss": 1.963,
+      "step": 334
+    },
+    {
+      "epoch": 0.39067055393586003,
+      "grad_norm": 0.15620645880699158,
+      "learning_rate": 9.76283185840708e-06,
+      "loss": 1.8777,
+      "step": 335
+    },
+    {
+      "epoch": 0.39183673469387753,
+      "grad_norm": 0.09890913963317871,
+      "learning_rate": 9.761972230093511e-06,
+      "loss": 1.842,
+      "step": 336
+    },
+    {
+      "epoch": 0.39300291545189503,
+      "grad_norm": 0.12239693850278854,
+      "learning_rate": 9.761111504926632e-06,
+      "loss": 1.8536,
+      "step": 337
+    },
+    {
+      "epoch": 0.39416909620991253,
+      "grad_norm": 0.09670644253492355,
+      "learning_rate": 9.760249680805788e-06,
+      "loss": 1.9457,
+      "step": 338
+    },
+    {
+      "epoch": 0.39533527696793,
+      "grad_norm": 0.09737051278352737,
+      "learning_rate": 9.759386755624957e-06,
+      "loss": 1.9673,
+      "step": 339
+    },
+    {
+      "epoch": 0.3965014577259475,
+      "grad_norm": 0.09580627828836441,
+      "learning_rate": 9.758522727272727e-06,
+      "loss": 1.8868,
+      "step": 340
+    },
+    {
+      "epoch": 0.397667638483965,
+      "grad_norm": 0.12124479562044144,
+      "learning_rate": 9.757657593632294e-06,
+      "loss": 1.9309,
+      "step": 341
+    },
+    {
+      "epoch": 0.3988338192419825,
+      "grad_norm": 0.11707170307636261,
+      "learning_rate": 9.756791352581427e-06,
+      "loss": 1.9016,
+      "step": 342
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.12999343872070312,
+      "learning_rate": 9.755924001992459e-06,
+      "loss": 1.8898,
+      "step": 343
+    },
+    {
+      "epoch": 0.40116618075801747,
+      "grad_norm": 0.09256736189126968,
+      "learning_rate": 9.755055539732272e-06,
+      "loss": 1.834,
+      "step": 344
+    },
+    {
+      "epoch": 0.40116618075801747,
+      "eval_loss": 0.9358909726142883,
+      "eval_runtime": 168.4624,
+      "eval_samples_per_second": 40.982,
+      "eval_steps_per_second": 2.564,
+      "step": 344
+    },
+    {
+      "epoch": 0.40233236151603496,
+      "grad_norm": 0.10559074580669403,
+      "learning_rate": 9.754185963662274e-06,
+      "loss": 1.8349,
+      "step": 345
+    },
+    {
+      "epoch": 0.40349854227405246,
+      "grad_norm": 0.08047611266374588,
+      "learning_rate": 9.753315271638387e-06,
+      "loss": 1.8221,
+      "step": 346
+    },
+    {
+      "epoch": 0.40466472303206996,
+      "grad_norm": 0.09671982377767563,
+      "learning_rate": 9.752443461511023e-06,
+      "loss": 1.8463,
+      "step": 347
+    },
+    {
+      "epoch": 0.40583090379008746,
+      "grad_norm": 0.11827627569437027,
+      "learning_rate": 9.751570531125072e-06,
+      "loss": 1.8393,
+      "step": 348
+    },
+    {
+      "epoch": 0.40699708454810496,
+      "grad_norm": 0.0888730064034462,
+      "learning_rate": 9.75069647831988e-06,
+      "loss": 1.8629,
+      "step": 349
+    },
+    {
+      "epoch": 0.40816326530612246,
+      "grad_norm": 0.13909290730953217,
+      "learning_rate": 9.749821300929236e-06,
+      "loss": 1.8635,
+      "step": 350
+    },
+    {
+      "epoch": 0.40932944606413996,
+      "grad_norm": 0.08804619312286377,
+      "learning_rate": 9.748944996781348e-06,
+      "loss": 1.8617,
+      "step": 351
+    },
+    {
+      "epoch": 0.41049562682215746,
+      "grad_norm": 0.1112079992890358,
+      "learning_rate": 9.748067563698826e-06,
+      "loss": 1.955,
+      "step": 352
+    },
+    {
+      "epoch": 0.41166180758017495,
+      "grad_norm": 0.08929669857025146,
+      "learning_rate": 9.747188999498675e-06,
+      "loss": 1.9581,
+      "step": 353
+    },
+    {
+      "epoch": 0.4128279883381924,
+      "grad_norm": 0.09188219904899597,
+      "learning_rate": 9.746309301992262e-06,
+      "loss": 1.8167,
+      "step": 354
+    },
+    {
+      "epoch": 0.4139941690962099,
+      "grad_norm": 0.09583163261413574,
+      "learning_rate": 9.7454284689853e-06,
+      "loss": 1.8766,
+      "step": 355
+    },
+    {
+      "epoch": 0.4151603498542274,
+      "grad_norm": 0.18377377092838287,
+      "learning_rate": 9.744546498277843e-06,
+      "loss": 1.8673,
+      "step": 356
+    },
+    {
+      "epoch": 0.4163265306122449,
+      "grad_norm": 0.1100630834698677,
+      "learning_rate": 9.743663387664249e-06,
+      "loss": 1.9026,
+      "step": 357
+    },
+    {
+      "epoch": 0.4174927113702624,
+      "grad_norm": 0.09713442623615265,
+      "learning_rate": 9.74277913493318e-06,
+      "loss": 1.8262,
+      "step": 358
+    },
+    {
+      "epoch": 0.4186588921282799,
+      "grad_norm": 0.13300760090351105,
+      "learning_rate": 9.741893737867568e-06,
+      "loss": 1.9011,
+      "step": 359
+    },
+    {
+      "epoch": 0.4198250728862974,
+      "grad_norm": 0.12599875032901764,
+      "learning_rate": 9.741007194244605e-06,
+      "loss": 1.8327,
+      "step": 360
+    },
+    {
+      "epoch": 0.4209912536443149,
+      "grad_norm": 0.35959139466285706,
+      "learning_rate": 9.74011950183572e-06,
+      "loss": 1.864,
+      "step": 361
+    },
+    {
+      "epoch": 0.4221574344023324,
+      "grad_norm": 0.10534382611513138,
+      "learning_rate": 9.73923065840657e-06,
+      "loss": 1.9157,
+      "step": 362
+    },
+    {
+      "epoch": 0.42332361516034983,
+      "grad_norm": 0.09453371167182922,
+      "learning_rate": 9.738340661717005e-06,
+      "loss": 1.9516,
+      "step": 363
+    },
+    {
+      "epoch": 0.42448979591836733,
+      "grad_norm": 0.09826826304197311,
+      "learning_rate": 9.737449509521062e-06,
+      "loss": 1.8652,
+      "step": 364
+    },
+    {
+      "epoch": 0.42565597667638483,
+      "grad_norm": 0.11827974766492844,
+      "learning_rate": 9.736557199566944e-06,
+      "loss": 1.818,
+      "step": 365
+    },
+    {
+      "epoch": 0.4268221574344023,
+      "grad_norm": 0.13238371908664703,
+      "learning_rate": 9.735663729596997e-06,
+      "loss": 1.8837,
+      "step": 366
+    },
+    {
+      "epoch": 0.4279883381924198,
+      "grad_norm": 0.0851769968867302,
+      "learning_rate": 9.734769097347691e-06,
+      "loss": 1.9197,
+      "step": 367
+    },
+    {
+      "epoch": 0.4291545189504373,
+      "grad_norm": 0.09790528565645218,
+      "learning_rate": 9.73387330054961e-06,
+      "loss": 1.9828,
+      "step": 368
+    },
+    {
+      "epoch": 0.4303206997084548,
+      "grad_norm": 0.4292675852775574,
+      "learning_rate": 9.73297633692742e-06,
+      "loss": 1.8726,
+      "step": 369
+    },
+    {
+      "epoch": 0.4314868804664723,
+      "grad_norm": 0.11923214793205261,
+      "learning_rate": 9.732078204199856e-06,
+      "loss": 1.8467,
+      "step": 370
+    },
+    {
+      "epoch": 0.4326530612244898,
+      "grad_norm": 0.10532516986131668,
+      "learning_rate": 9.731178900079706e-06,
+      "loss": 1.8947,
+      "step": 371
+    },
+    {
+      "epoch": 0.43381924198250726,
+      "grad_norm": 0.10830720514059067,
+      "learning_rate": 9.730278422273782e-06,
+      "loss": 1.8763,
+      "step": 372
+    },
+    {
+      "epoch": 0.43498542274052476,
+      "grad_norm": 0.19431138038635254,
+      "learning_rate": 9.729376768482915e-06,
+      "loss": 1.8597,
+      "step": 373
+    },
+    {
+      "epoch": 0.43615160349854226,
+      "grad_norm": 0.09612815082073212,
+      "learning_rate": 9.728473936401918e-06,
+      "loss": 1.8398,
+      "step": 374
+    },
+    {
+      "epoch": 0.43731778425655976,
+      "grad_norm": 0.09484248608350754,
+      "learning_rate": 9.72756992371958e-06,
+      "loss": 1.9256,
+      "step": 375
+    },
+    {
+      "epoch": 0.43848396501457726,
+      "grad_norm": 0.08584627509117126,
+      "learning_rate": 9.72666472811864e-06,
+      "loss": 1.8764,
+      "step": 376
+    },
+    {
+      "epoch": 0.43965014577259476,
+      "grad_norm": 0.1113109439611435,
+      "learning_rate": 9.725758347275771e-06,
+      "loss": 1.8654,
+      "step": 377
+    },
+    {
+      "epoch": 0.44081632653061226,
+      "grad_norm": 0.13657604157924652,
+      "learning_rate": 9.724850778861553e-06,
+      "loss": 1.8398,
+      "step": 378
+    },
+    {
+      "epoch": 0.44198250728862976,
+      "grad_norm": 0.10629869252443314,
+      "learning_rate": 9.723942020540463e-06,
+      "loss": 1.8706,
+      "step": 379
+    },
+    {
+      "epoch": 0.44314868804664725,
+      "grad_norm": 0.0976574495434761,
+      "learning_rate": 9.723032069970846e-06,
+      "loss": 1.8168,
+      "step": 380
+    },
+    {
+      "epoch": 0.4443148688046647,
+      "grad_norm": 0.11477094888687134,
+      "learning_rate": 9.722120924804903e-06,
+      "loss": 1.8465,
+      "step": 381
+    },
+    {
+      "epoch": 0.4454810495626822,
+      "grad_norm": 0.16930529475212097,
+      "learning_rate": 9.72120858268866e-06,
+      "loss": 1.8834,
+      "step": 382
+    },
+    {
+      "epoch": 0.4466472303206997,
+      "grad_norm": 0.10541530698537827,
+      "learning_rate": 9.720295041261959e-06,
+      "loss": 1.8377,
+      "step": 383
+    },
+    {
+      "epoch": 0.4478134110787172,
+      "grad_norm": 0.13728584349155426,
+      "learning_rate": 9.719380298158434e-06,
+      "loss": 1.8699,
+      "step": 384
+    },
+    {
+      "epoch": 0.4489795918367347,
+      "grad_norm": 0.08973895758390427,
+      "learning_rate": 9.718464351005485e-06,
+      "loss": 1.8502,
+      "step": 385
+    },
+    {
+      "epoch": 0.4501457725947522,
+      "grad_norm": 0.11848149448633194,
+      "learning_rate": 9.717547197424266e-06,
+      "loss": 1.9038,
+      "step": 386
+    },
+    {
+      "epoch": 0.4513119533527697,
+      "grad_norm": 0.23062708973884583,
+      "learning_rate": 9.716628835029656e-06,
+      "loss": 1.8654,
+      "step": 387
+    },
+    {
+      "epoch": 0.4524781341107872,
+      "grad_norm": 0.5158998370170593,
+      "learning_rate": 9.715709261430247e-06,
+      "loss": 1.9295,
+      "step": 388
+    },
+    {
+      "epoch": 0.4536443148688047,
+      "grad_norm": 0.19186925888061523,
+      "learning_rate": 9.714788474228317e-06,
+      "loss": 1.8209,
+      "step": 389
+    },
+    {
+      "epoch": 0.45481049562682213,
+      "grad_norm": 0.09066485613584518,
+      "learning_rate": 9.71386647101981e-06,
+      "loss": 1.8895,
+      "step": 390
+    },
+    {
+      "epoch": 0.45597667638483963,
+      "grad_norm": 0.09533058106899261,
+      "learning_rate": 9.712943249394319e-06,
+      "loss": 1.9208,
+      "step": 391
+    },
+    {
+      "epoch": 0.45714285714285713,
+      "grad_norm": 0.23010948300361633,
+      "learning_rate": 9.712018806935057e-06,
+      "loss": 1.9003,
+      "step": 392
+    },
+    {
+      "epoch": 0.4583090379008746,
+      "grad_norm": 0.08680427074432373,
+      "learning_rate": 9.711093141218851e-06,
+      "loss": 1.866,
+      "step": 393
+    },
+    {
+      "epoch": 0.4594752186588921,
+      "grad_norm": 0.10906676948070526,
+      "learning_rate": 9.710166249816097e-06,
+      "loss": 1.8634,
+      "step": 394
+    },
+    {
+      "epoch": 0.4606413994169096,
+      "grad_norm": 0.09538917243480682,
+      "learning_rate": 9.709238130290763e-06,
+      "loss": 1.8307,
+      "step": 395
+    },
+    {
+      "epoch": 0.4618075801749271,
+      "grad_norm": 0.0878988653421402,
+      "learning_rate": 9.708308780200354e-06,
+      "loss": 1.8593,
+      "step": 396
+    },
+    {
+      "epoch": 0.4629737609329446,
+      "grad_norm": 0.09339457750320435,
+      "learning_rate": 9.707378197095896e-06,
+      "loss": 1.8825,
+      "step": 397
+    },
+    {
+      "epoch": 0.4641399416909621,
+      "grad_norm": 0.14582982659339905,
+      "learning_rate": 9.706446378521907e-06,
+      "loss": 1.8049,
+      "step": 398
+    },
+    {
+      "epoch": 0.46530612244897956,
+      "grad_norm": 0.09130845218896866,
+      "learning_rate": 9.705513322016386e-06,
+      "loss": 1.8987,
+      "step": 399
+    },
+    {
+      "epoch": 0.46647230320699706,
+      "grad_norm": 0.16227616369724274,
+      "learning_rate": 9.704579025110782e-06,
+      "loss": 1.8084,
+      "step": 400
+    },
+    {
+      "epoch": 0.46763848396501456,
+      "grad_norm": 0.09542547911405563,
+      "learning_rate": 9.703643485329984e-06,
+      "loss": 1.888,
+      "step": 401
+    },
+    {
+      "epoch": 0.46880466472303206,
+      "grad_norm": 0.10470747947692871,
+      "learning_rate": 9.70270670019228e-06,
+      "loss": 1.8885,
+      "step": 402
+    },
+    {
+      "epoch": 0.46997084548104956,
+      "grad_norm": 0.08773402124643326,
+      "learning_rate": 9.701768667209355e-06,
+      "loss": 1.9137,
+      "step": 403
+    },
+    {
+      "epoch": 0.47113702623906706,
+      "grad_norm": 0.11207730323076248,
+      "learning_rate": 9.700829383886256e-06,
+      "loss": 1.9111,
+      "step": 404
+    },
+    {
+      "epoch": 0.47230320699708456,
+      "grad_norm": 0.11006439477205276,
+      "learning_rate": 9.699888847721379e-06,
+      "loss": 1.8299,
+      "step": 405
+    },
+    {
+      "epoch": 0.47346938775510206,
+      "grad_norm": 0.08384604752063751,
+      "learning_rate": 9.698947056206436e-06,
+      "loss": 1.8199,
+      "step": 406
+    },
+    {
+      "epoch": 0.47463556851311955,
+      "grad_norm": 0.10979919135570526,
+      "learning_rate": 9.698004006826445e-06,
+      "loss": 1.9264,
+      "step": 407
+    },
+    {
+      "epoch": 0.47580174927113705,
+      "grad_norm": 0.09500827640295029,
+      "learning_rate": 9.6970596970597e-06,
+      "loss": 1.8816,
+      "step": 408
+    },
+    {
+      "epoch": 0.4769679300291545,
+      "grad_norm": 0.8866495490074158,
+      "learning_rate": 9.69611412437774e-06,
+      "loss": 1.8834,
+      "step": 409
+    },
+    {
+      "epoch": 0.478134110787172,
+      "grad_norm": 0.11265905201435089,
+      "learning_rate": 9.695167286245353e-06,
+      "loss": 1.909,
+      "step": 410
+    },
+    {
+      "epoch": 0.4793002915451895,
+      "grad_norm": 0.27813753485679626,
+      "learning_rate": 9.694219180120528e-06,
+      "loss": 1.857,
+      "step": 411
+    },
+    {
+      "epoch": 0.480466472303207,
+      "grad_norm": 0.08620649576187134,
+      "learning_rate": 9.69326980345444e-06,
+      "loss": 1.8055,
+      "step": 412
+    },
+    {
+      "epoch": 0.4816326530612245,
+      "grad_norm": 0.1967063695192337,
+      "learning_rate": 9.692319153691427e-06,
+      "loss": 1.8397,
+      "step": 413
+    },
+    {
+      "epoch": 0.482798833819242,
+      "grad_norm": 0.09050402045249939,
+      "learning_rate": 9.691367228268974e-06,
+      "loss": 1.8316,
+      "step": 414
+    },
+    {
+      "epoch": 0.4839650145772595,
+      "grad_norm": 0.0871565192937851,
+      "learning_rate": 9.69041402461768e-06,
+      "loss": 1.9044,
+      "step": 415
+    },
+    {
+      "epoch": 0.485131195335277,
+      "grad_norm": 0.11573964357376099,
+      "learning_rate": 9.689459540161244e-06,
+      "loss": 1.9144,
+      "step": 416
+    },
+    {
+      "epoch": 0.4862973760932945,
+      "grad_norm": 0.2950435280799866,
+      "learning_rate": 9.688503772316427e-06,
+      "loss": 1.7865,
+      "step": 417
+    },
+    {
+      "epoch": 0.48746355685131193,
+      "grad_norm": 0.13304616510868073,
+      "learning_rate": 9.687546718493049e-06,
+      "loss": 1.9359,
+      "step": 418
+    },
+    {
+      "epoch": 0.48862973760932943,
+      "grad_norm": 0.09340868890285492,
+      "learning_rate": 9.68658837609395e-06,
+      "loss": 1.9026,
+      "step": 419
+    },
+    {
+      "epoch": 0.4897959183673469,
+      "grad_norm": 0.10877745598554611,
+      "learning_rate": 9.685628742514973e-06,
+      "loss": 1.8669,
+      "step": 420
+    },
+    {
+      "epoch": 0.4909620991253644,
+      "grad_norm": 0.13621404767036438,
+      "learning_rate": 9.684667815144934e-06,
+      "loss": 1.9072,
+      "step": 421
+    },
+    {
+      "epoch": 0.4921282798833819,
+      "grad_norm": 0.08816006779670715,
+      "learning_rate": 9.683705591365612e-06,
+      "loss": 1.8489,
+      "step": 422
+    },
+    {
+      "epoch": 0.4932944606413994,
+      "grad_norm": 0.10320248454809189,
+      "learning_rate": 9.682742068551714e-06,
+      "loss": 1.8438,
+      "step": 423
+    },
+    {
+      "epoch": 0.4944606413994169,
+      "grad_norm": 0.08581226319074631,
+      "learning_rate": 9.681777244070851e-06,
+      "loss": 1.9125,
+      "step": 424
+    },
+    {
+      "epoch": 0.4956268221574344,
+      "grad_norm": 0.0924190878868103,
+      "learning_rate": 9.680811115283516e-06,
+      "loss": 1.8586,
+      "step": 425
+    },
+    {
+      "epoch": 0.4967930029154519,
+      "grad_norm": 0.09078896045684814,
+      "learning_rate": 9.679843679543064e-06,
+      "loss": 1.8395,
+      "step": 426
+    },
+    {
+      "epoch": 0.49795918367346936,
+      "grad_norm": 0.09030900150537491,
+      "learning_rate": 9.678874934195684e-06,
+      "loss": 1.9119,
+      "step": 427
+    },
+    {
+      "epoch": 0.49912536443148686,
+      "grad_norm": 0.10297784954309464,
+      "learning_rate": 9.677904876580376e-06,
+      "loss": 1.9703,
+      "step": 428
+    },
+    {
+      "epoch": 0.5002915451895044,
+      "grad_norm": 0.12785211205482483,
+      "learning_rate": 9.67693350402892e-06,
+      "loss": 1.8789,
+      "step": 429
+    },
+    {
+      "epoch": 0.5014577259475219,
+      "grad_norm": 0.4867236316204071,
+      "learning_rate": 9.675960813865864e-06,
+      "loss": 1.8745,
+      "step": 430
+    },
+    {
+      "epoch": 0.5014577259475219,
+      "eval_loss": 0.9317024350166321,
+      "eval_runtime": 168.3877,
+      "eval_samples_per_second": 41.001,
+      "eval_steps_per_second": 2.566,
+      "step": 430
+    },
+    {
+      "epoch": 0.5026239067055394,
+      "grad_norm": 0.11284340918064117,
+      "learning_rate": 9.674986803408492e-06,
+      "loss": 1.8539,
+      "step": 431
+    },
+    {
+      "epoch": 0.5037900874635568,
+      "grad_norm": 0.09896525740623474,
+      "learning_rate": 9.6740114699668e-06,
+      "loss": 1.9219,
+      "step": 432
+    },
+    {
+      "epoch": 0.5049562682215744,
+      "grad_norm": 0.09354698657989502,
+      "learning_rate": 9.673034810843466e-06,
+      "loss": 1.8738,
+      "step": 433
+    },
+    {
+      "epoch": 0.5061224489795918,
+      "grad_norm": 0.08065647631883621,
+      "learning_rate": 9.672056823333838e-06,
+      "loss": 1.831,
+      "step": 434
+    },
+    {
+      "epoch": 0.5072886297376094,
+      "grad_norm": 0.09500313550233841,
+      "learning_rate": 9.671077504725898e-06,
+      "loss": 1.8129,
+      "step": 435
+    },
+    {
+      "epoch": 0.5084548104956268,
+      "grad_norm": 0.17649948596954346,
+      "learning_rate": 9.670096852300244e-06,
+      "loss": 1.8149,
+      "step": 436
+    },
+    {
+      "epoch": 0.5096209912536444,
+      "grad_norm": 0.31409019231796265,
+      "learning_rate": 9.669114863330053e-06,
+      "loss": 1.8221,
+      "step": 437
+    },
+    {
+      "epoch": 0.5107871720116618,
+      "grad_norm": 0.08261363208293915,
+      "learning_rate": 9.668131535081073e-06,
+      "loss": 1.8182,
+      "step": 438
+    },
+    {
+      "epoch": 0.5119533527696793,
+      "grad_norm": 0.08719644695520401,
+      "learning_rate": 9.667146864811585e-06,
+      "loss": 1.8735,
+      "step": 439
+    },
+    {
+      "epoch": 0.5131195335276968,
+      "grad_norm": 0.09595973044633865,
+      "learning_rate": 9.666160849772384e-06,
+      "loss": 1.9783,
+      "step": 440
+    },
+    {
+      "epoch": 0.5142857142857142,
+      "grad_norm": 0.6334007978439331,
+      "learning_rate": 9.665173487206742e-06,
+      "loss": 1.8485,
+      "step": 441
+    },
+    {
+      "epoch": 0.5154518950437318,
+      "grad_norm": 0.11343630403280258,
+      "learning_rate": 9.664184774350403e-06,
+      "loss": 1.8379,
+      "step": 442
+    },
+    {
+      "epoch": 0.5166180758017492,
+      "grad_norm": 0.10257852077484131,
+      "learning_rate": 9.663194708431536e-06,
+      "loss": 1.9247,
+      "step": 443
+    },
+    {
+      "epoch": 0.5177842565597668,
+      "grad_norm": 0.19690583646297455,
+      "learning_rate": 9.662203286670725e-06,
+      "loss": 1.8652,
+      "step": 444
+    },
+    {
+      "epoch": 0.5189504373177842,
+      "grad_norm": 0.4637579023838043,
+      "learning_rate": 9.66121050628093e-06,
+      "loss": 1.8223,
+      "step": 445
+    },
+    {
+      "epoch": 0.5201166180758018,
+      "grad_norm": 0.08869434893131256,
+      "learning_rate": 9.660216364467469e-06,
+      "loss": 1.9209,
+      "step": 446
+    },
+    {
+      "epoch": 0.5212827988338192,
+      "grad_norm": 0.08551875501871109,
+      "learning_rate": 9.659220858427994e-06,
+      "loss": 1.8557,
+      "step": 447
+    },
+    {
+      "epoch": 0.5224489795918368,
+      "grad_norm": 0.08415248990058899,
+      "learning_rate": 9.658223985352457e-06,
+      "loss": 1.8189,
+      "step": 448
+    },
+    {
+      "epoch": 0.5236151603498542,
+      "grad_norm": 0.10640854388475418,
+      "learning_rate": 9.657225742423087e-06,
+      "loss": 1.8377,
+      "step": 449
+    },
+    {
+      "epoch": 0.5247813411078717,
+      "grad_norm": 0.09836480766534805,
+      "learning_rate": 9.656226126814363e-06,
+      "loss": 1.8668,
+      "step": 450
+    },
+    {
+      "epoch": 0.5259475218658892,
+      "grad_norm": 0.08484815061092377,
+      "learning_rate": 9.655225135692991e-06,
+      "loss": 1.8054,
+      "step": 451
+    },
+    {
+      "epoch": 0.5271137026239067,
+      "grad_norm": 0.09366266429424286,
+      "learning_rate": 9.654222766217871e-06,
+      "loss": 1.867,
+      "step": 452
+    },
+    {
+      "epoch": 0.5282798833819242,
+      "grad_norm": 0.11442459374666214,
+      "learning_rate": 9.653219015540076e-06,
+      "loss": 1.8367,
+      "step": 453
+    },
+    {
+      "epoch": 0.5294460641399417,
+      "grad_norm": 0.08268333226442337,
+      "learning_rate": 9.65221388080282e-06,
+      "loss": 1.9132,
+      "step": 454
+    },
+    {
+      "epoch": 0.5306122448979592,
+      "grad_norm": 0.08256720751523972,
+      "learning_rate": 9.651207359141435e-06,
+      "loss": 1.9091,
+      "step": 455
+    },
+    {
+      "epoch": 0.5317784256559767,
+      "grad_norm": 0.09420077502727509,
+      "learning_rate": 9.65019944768334e-06,
+      "loss": 1.8812,
+      "step": 456
+    },
+    {
+      "epoch": 0.5329446064139942,
+      "grad_norm": 0.080384761095047,
+      "learning_rate": 9.649190143548017e-06,
+      "loss": 1.8798,
+      "step": 457
+    },
+    {
+      "epoch": 0.5341107871720117,
+      "grad_norm": 0.10223258286714554,
+      "learning_rate": 9.648179443846982e-06,
+      "loss": 1.8443,
+      "step": 458
+    },
+    {
+      "epoch": 0.5352769679300291,
+      "grad_norm": 0.08851838856935501,
+      "learning_rate": 9.647167345683759e-06,
+      "loss": 1.7902,
+      "step": 459
+    },
+    {
+      "epoch": 0.5364431486880467,
+      "grad_norm": 0.09891951829195023,
+      "learning_rate": 9.646153846153848e-06,
+      "loss": 1.9753,
+      "step": 460
+    },
+    {
+      "epoch": 0.5376093294460641,
+      "grad_norm": 0.09817704558372498,
+      "learning_rate": 9.6451389423447e-06,
+      "loss": 1.8701,
+      "step": 461
+    },
+    {
+      "epoch": 0.5387755102040817,
+      "grad_norm": 0.09142075479030609,
+      "learning_rate": 9.644122631335695e-06,
+      "loss": 1.9764,
+      "step": 462
+    },
+    {
+      "epoch": 0.5399416909620991,
+      "grad_norm": 0.1107424646615982,
+      "learning_rate": 9.643104910198105e-06,
+      "loss": 1.764,
+      "step": 463
+    },
+    {
+      "epoch": 0.5411078717201167,
+      "grad_norm": 0.08371910452842712,
+      "learning_rate": 9.642085775995065e-06,
+      "loss": 1.8942,
+      "step": 464
+    },
+    {
+      "epoch": 0.5422740524781341,
+      "grad_norm": 0.40017104148864746,
+      "learning_rate": 9.641065225781552e-06,
+      "loss": 1.9102,
+      "step": 465
+    },
+    {
+      "epoch": 0.5434402332361516,
+      "grad_norm": 0.10847759991884232,
+      "learning_rate": 9.640043256604357e-06,
+      "loss": 1.917,
+      "step": 466
+    },
+    {
+      "epoch": 0.5446064139941691,
+      "grad_norm": 0.192711740732193,
+      "learning_rate": 9.639019865502049e-06,
+      "loss": 1.8491,
+      "step": 467
+    },
+    {
+      "epoch": 0.5457725947521865,
+      "grad_norm": 0.09193003922700882,
+      "learning_rate": 9.637995049504952e-06,
+      "loss": 1.9387,
+      "step": 468
+    },
+    {
+      "epoch": 0.5469387755102041,
+      "grad_norm": 0.16377852857112885,
+      "learning_rate": 9.636968805635112e-06,
+      "loss": 1.8195,
+      "step": 469
+    },
+    {
+      "epoch": 0.5481049562682215,
+      "grad_norm": 0.09493596106767654,
+      "learning_rate": 9.635941130906275e-06,
+      "loss": 1.842,
+      "step": 470
+    },
+    {
+      "epoch": 0.5492711370262391,
+      "grad_norm": 0.1303832232952118,
+      "learning_rate": 9.634912022323853e-06,
+      "loss": 1.8688,
+      "step": 471
+    },
+    {
+      "epoch": 0.5504373177842565,
+      "grad_norm": 0.19474978744983673,
+      "learning_rate": 9.633881476884891e-06,
+      "loss": 1.8956,
+      "step": 472
+    },
+    {
+      "epoch": 0.5516034985422741,
+      "grad_norm": 0.09646501392126083,
+      "learning_rate": 9.632849491578049e-06,
+      "loss": 1.8324,
+      "step": 473
+    },
+    {
+      "epoch": 0.5527696793002915,
+      "grad_norm": 0.4128596782684326,
+      "learning_rate": 9.631816063383564e-06,
+      "loss": 1.8779,
+      "step": 474
+    },
+    {
+      "epoch": 0.5539358600583091,
+      "grad_norm": 0.08190983533859253,
+      "learning_rate": 9.630781189273223e-06,
+      "loss": 1.8944,
+      "step": 475
+    },
+    {
+      "epoch": 0.5551020408163265,
+      "grad_norm": 0.14745251834392548,
+      "learning_rate": 9.629744866210331e-06,
+      "loss": 1.9055,
+      "step": 476
+    },
+    {
+      "epoch": 0.556268221574344,
+      "grad_norm": 0.17714940011501312,
+      "learning_rate": 9.628707091149685e-06,
+      "loss": 1.817,
+      "step": 477
+    },
+    {
+      "epoch": 0.5574344023323615,
+      "grad_norm": 0.10438665747642517,
+      "learning_rate": 9.627667861037545e-06,
+      "loss": 1.8002,
+      "step": 478
+    },
+    {
+      "epoch": 0.558600583090379,
+      "grad_norm": 0.1380535215139389,
+      "learning_rate": 9.6266271728116e-06,
+      "loss": 1.7752,
+      "step": 479
+    },
+    {
+      "epoch": 0.5597667638483965,
+      "grad_norm": 0.11089915037155151,
+      "learning_rate": 9.625585023400937e-06,
+      "loss": 1.8713,
+      "step": 480
+    },
+    {
+      "epoch": 0.560932944606414,
+      "grad_norm": 0.09744633734226227,
+      "learning_rate": 9.624541409726016e-06,
+      "loss": 1.8823,
+      "step": 481
+    },
+    {
+      "epoch": 0.5620991253644315,
+      "grad_norm": 0.09447052329778671,
+      "learning_rate": 9.623496328698641e-06,
+      "loss": 1.9124,
+      "step": 482
+    },
+    {
+      "epoch": 0.563265306122449,
+      "grad_norm": 0.08723996579647064,
+      "learning_rate": 9.62244977722192e-06,
+      "loss": 1.8727,
+      "step": 483
+    },
+    {
+      "epoch": 0.5644314868804665,
+      "grad_norm": 0.17652207612991333,
+      "learning_rate": 9.62140175219024e-06,
+      "loss": 1.9251,
+      "step": 484
+    },
+    {
+      "epoch": 0.565597667638484,
+      "grad_norm": 0.20766834914684296,
+      "learning_rate": 9.620352250489237e-06,
+      "loss": 1.8578,
+      "step": 485
+    },
+    {
+      "epoch": 0.5667638483965015,
+      "grad_norm": 0.07957983762025833,
+      "learning_rate": 9.61930126899577e-06,
+      "loss": 1.7621,
+      "step": 486
+    },
+    {
+      "epoch": 0.567930029154519,
+      "grad_norm": 0.09336891025304794,
+      "learning_rate": 9.61824880457788e-06,
+      "loss": 1.8832,
+      "step": 487
+    },
+    {
+      "epoch": 0.5690962099125364,
+      "grad_norm": 0.1073063537478447,
+      "learning_rate": 9.617194854094761e-06,
+      "loss": 1.8105,
+      "step": 488
+    },
+    {
+      "epoch": 0.570262390670554,
+      "grad_norm": 0.33396124839782715,
+      "learning_rate": 9.616139414396735e-06,
+      "loss": 1.843,
+      "step": 489
+    },
+    {
+      "epoch": 0.5714285714285714,
+      "grad_norm": 0.0911710113286972,
+      "learning_rate": 9.615082482325217e-06,
+      "loss": 1.8782,
+      "step": 490
+    },
+    {
+      "epoch": 0.572594752186589,
+      "grad_norm": 0.09589587152004242,
+      "learning_rate": 9.61402405471268e-06,
+      "loss": 1.8661,
+      "step": 491
+    },
+    {
+      "epoch": 0.5737609329446064,
+      "grad_norm": 0.12544631958007812,
+      "learning_rate": 9.612964128382631e-06,
+      "loss": 1.8945,
+      "step": 492
+    },
+    {
+      "epoch": 0.574927113702624,
+      "grad_norm": 0.1198568046092987,
+      "learning_rate": 9.611902700149571e-06,
+      "loss": 1.9123,
+      "step": 493
+    },
+    {
+      "epoch": 0.5760932944606414,
+      "grad_norm": 0.18756799399852753,
+      "learning_rate": 9.61083976681897e-06,
+      "loss": 1.822,
+      "step": 494
+    },
+    {
+      "epoch": 0.577259475218659,
+      "grad_norm": 0.09770026803016663,
+      "learning_rate": 9.60977532518723e-06,
+      "loss": 1.9039,
+      "step": 495
+    },
+    {
+      "epoch": 0.5784256559766764,
+      "grad_norm": 0.10050517320632935,
+      "learning_rate": 9.608709372041654e-06,
+      "loss": 1.9761,
+      "step": 496
+    },
+    {
+      "epoch": 0.5795918367346938,
+      "grad_norm": 0.0852288231253624,
+      "learning_rate": 9.607641904160417e-06,
+      "loss": 1.8909,
+      "step": 497
+    },
+    {
+      "epoch": 0.5807580174927114,
+      "grad_norm": 0.09355819225311279,
+      "learning_rate": 9.60657291831253e-06,
+      "loss": 1.8471,
+      "step": 498
+    },
+    {
+      "epoch": 0.5819241982507288,
+      "grad_norm": 1.137154459953308,
+      "learning_rate": 9.605502411257808e-06,
+      "loss": 1.8604,
+      "step": 499
+    },
+    {
+      "epoch": 0.5830903790087464,
+      "grad_norm": 0.10329793393611908,
+      "learning_rate": 9.604430379746837e-06,
+      "loss": 1.8765,
+      "step": 500
+    },
+    {
+      "epoch": 0.5842565597667638,
+      "grad_norm": 0.12563008069992065,
+      "learning_rate": 9.603356820520941e-06,
+      "loss": 1.8573,
+      "step": 501
+    },
+    {
+      "epoch": 0.5854227405247814,
+      "grad_norm": 0.13698391616344452,
+      "learning_rate": 9.602281730312155e-06,
+      "loss": 1.8013,
+      "step": 502
+    },
+    {
+      "epoch": 0.5865889212827988,
+      "grad_norm": 0.15192078053951263,
+      "learning_rate": 9.601205105843178e-06,
+      "loss": 1.8418,
+      "step": 503
+    },
+    {
+      "epoch": 0.5877551020408164,
+      "grad_norm": 0.09103371202945709,
+      "learning_rate": 9.600126943827357e-06,
+      "loss": 1.899,
+      "step": 504
+    },
+    {
+      "epoch": 0.5889212827988338,
+      "grad_norm": 0.12354784458875656,
+      "learning_rate": 9.599047240968638e-06,
+      "loss": 1.8468,
+      "step": 505
+    },
+    {
+      "epoch": 0.5900874635568513,
+      "grad_norm": 0.11870899051427841,
+      "learning_rate": 9.597965993961545e-06,
+      "loss": 1.9026,
+      "step": 506
+    },
+    {
+      "epoch": 0.5912536443148688,
+      "grad_norm": 0.13545583188533783,
+      "learning_rate": 9.596883199491136e-06,
+      "loss": 1.9672,
+      "step": 507
+    },
+    {
+      "epoch": 0.5924198250728863,
+      "grad_norm": 0.11981657147407532,
+      "learning_rate": 9.595798854232975e-06,
+      "loss": 1.8977,
+      "step": 508
+    },
+    {
+      "epoch": 0.5935860058309038,
+      "grad_norm": 0.12264847010374069,
+      "learning_rate": 9.594712954853093e-06,
+      "loss": 1.8989,
+      "step": 509
+    },
+    {
+      "epoch": 0.5947521865889213,
+      "grad_norm": 0.08354833722114563,
+      "learning_rate": 9.593625498007969e-06,
+      "loss": 1.8805,
+      "step": 510
+    },
+    {
+      "epoch": 0.5959183673469388,
+      "grad_norm": 0.08214504271745682,
+      "learning_rate": 9.592536480344472e-06,
+      "loss": 1.8239,
+      "step": 511
+    },
+    {
+      "epoch": 0.5970845481049563,
+      "grad_norm": 0.222848042845726,
+      "learning_rate": 9.591445898499842e-06,
+      "loss": 1.8853,
+      "step": 512
+    },
+    {
+      "epoch": 0.5982507288629738,
+      "grad_norm": 0.08465005457401276,
+      "learning_rate": 9.590353749101653e-06,
+      "loss": 1.835,
+      "step": 513
+    },
+    {
+      "epoch": 0.5994169096209913,
+      "grad_norm": 0.14443686604499817,
+      "learning_rate": 9.589260028767782e-06,
+      "loss": 1.8072,
+      "step": 514
+    },
+    {
+      "epoch": 0.6005830903790087,
+      "grad_norm": 0.09332364797592163,
+      "learning_rate": 9.58816473410636e-06,
+      "loss": 1.863,
+      "step": 515
+    },
+    {
+      "epoch": 0.6017492711370263,
+      "grad_norm": 0.16969147324562073,
+      "learning_rate": 9.58706786171575e-06,
+      "loss": 1.9271,
+      "step": 516
+    },
+    {
+      "epoch": 0.6017492711370263,
+      "eval_loss": 0.9282099008560181,
+      "eval_runtime": 168.2666,
+      "eval_samples_per_second": 41.03,
+      "eval_steps_per_second": 2.567,
+      "step": 516
+    },
+    {
+      "epoch": 0.6029154518950437,
+      "grad_norm": 0.08617091923952103,
+      "learning_rate": 9.585969408184513e-06,
+      "loss": 1.8184,
+      "step": 517
+    },
+    {
+      "epoch": 0.6040816326530613,
+      "grad_norm": 0.08569926023483276,
+      "learning_rate": 9.584869370091362e-06,
+      "loss": 1.8141,
+      "step": 518
+    },
+    {
+      "epoch": 0.6052478134110787,
+      "grad_norm": 0.10691308975219727,
+      "learning_rate": 9.583767744005134e-06,
+      "loss": 1.8534,
+      "step": 519
+    },
+    {
+      "epoch": 0.6064139941690962,
+      "grad_norm": 0.10644278675317764,
+      "learning_rate": 9.582664526484752e-06,
+      "loss": 1.8329,
+      "step": 520
+    },
+    {
+      "epoch": 0.6075801749271137,
+      "grad_norm": 0.12132445722818375,
+      "learning_rate": 9.581559714079191e-06,
+      "loss": 1.8374,
+      "step": 521
+    },
+    {
+      "epoch": 0.6087463556851312,
+      "grad_norm": 0.12605313956737518,
+      "learning_rate": 9.58045330332744e-06,
+      "loss": 1.8274,
+      "step": 522
+    },
+    {
+      "epoch": 0.6099125364431487,
+      "grad_norm": 0.1545335054397583,
+      "learning_rate": 9.579345290758467e-06,
+      "loss": 1.8835,
+      "step": 523
+    },
+    {
+      "epoch": 0.6110787172011661,
+      "grad_norm": 0.11216040700674057,
+      "learning_rate": 9.57823567289118e-06,
+      "loss": 1.9019,
+      "step": 524
+    },
+    {
+      "epoch": 0.6122448979591837,
+      "grad_norm": 0.09745314717292786,
+      "learning_rate": 9.577124446234394e-06,
+      "loss": 1.8788,
+      "step": 525
+    },
+    {
+      "epoch": 0.6134110787172011,
+      "grad_norm": 0.09892967343330383,
+      "learning_rate": 9.576011607286798e-06,
+      "loss": 1.8814,
+      "step": 526
+    },
+    {
+      "epoch": 0.6145772594752187,
+      "grad_norm": 0.1541074961423874,
+      "learning_rate": 9.574897152536906e-06,
+      "loss": 1.9136,
+      "step": 527
+    },
+    {
+      "epoch": 0.6157434402332361,
+      "grad_norm": 0.1027301624417305,
+      "learning_rate": 9.57378107846303e-06,
+      "loss": 1.8369,
+      "step": 528
+    },
+    {
+      "epoch": 0.6169096209912537,
+      "grad_norm": 0.20511355996131897,
+      "learning_rate": 9.572663381533241e-06,
+      "loss": 1.904,
+      "step": 529
+    },
+    {
+      "epoch": 0.6180758017492711,
+      "grad_norm": 0.1284116506576538,
+      "learning_rate": 9.571544058205336e-06,
+      "loss": 1.9183,
+      "step": 530
+    },
+    {
+      "epoch": 0.6192419825072887,
+      "grad_norm": 0.08450975269079208,
+      "learning_rate": 9.570423104926788e-06,
+      "loss": 1.881,
+      "step": 531
+    },
+    {
+      "epoch": 0.6204081632653061,
+      "grad_norm": 0.09149329364299774,
+      "learning_rate": 9.569300518134716e-06,
+      "loss": 1.8667,
+      "step": 532
+    },
+    {
+      "epoch": 0.6215743440233236,
+      "grad_norm": 0.16286329925060272,
+      "learning_rate": 9.568176294255855e-06,
+      "loss": 1.8407,
+      "step": 533
+    },
+    {
+      "epoch": 0.6227405247813411,
+      "grad_norm": 0.23538637161254883,
+      "learning_rate": 9.567050429706504e-06,
+      "loss": 1.7848,
+      "step": 534
+    },
+    {
+      "epoch": 0.6239067055393586,
+      "grad_norm": 0.09121550619602203,
+      "learning_rate": 9.565922920892495e-06,
+      "loss": 1.8431,
+      "step": 535
+    },
+    {
+      "epoch": 0.6250728862973761,
+      "grad_norm": 0.09964902698993683,
+      "learning_rate": 9.564793764209159e-06,
+      "loss": 1.8634,
+      "step": 536
+    },
+    {
+      "epoch": 0.6262390670553936,
+      "grad_norm": 0.08488503098487854,
+      "learning_rate": 9.563662956041279e-06,
+      "loss": 1.7448,
+      "step": 537
+    },
+    {
+      "epoch": 0.6274052478134111,
+      "grad_norm": 0.5676212310791016,
+      "learning_rate": 9.562530492763053e-06,
+      "loss": 1.8362,
+      "step": 538
+    },
+    {
+      "epoch": 0.6285714285714286,
+      "grad_norm": 0.1452900469303131,
+      "learning_rate": 9.56139637073806e-06,
+      "loss": 1.8719,
+      "step": 539
+    },
+    {
+      "epoch": 0.6297376093294461,
+      "grad_norm": 0.3458977937698364,
+      "learning_rate": 9.56026058631922e-06,
+      "loss": 1.8989,
+      "step": 540
+    },
+    {
+      "epoch": 0.6309037900874636,
+      "grad_norm": 0.1590295135974884,
+      "learning_rate": 9.559123135848751e-06,
+      "loss": 1.9095,
+      "step": 541
+    },
+    {
+      "epoch": 0.632069970845481,
+      "grad_norm": 0.09185579419136047,
+      "learning_rate": 9.557984015658131e-06,
+      "loss": 1.9091,
+      "step": 542
+    },
+    {
+      "epoch": 0.6332361516034986,
+      "grad_norm": 0.08805494010448456,
+      "learning_rate": 9.556843222068067e-06,
+      "loss": 1.8104,
+      "step": 543
+    },
+    {
+      "epoch": 0.634402332361516,
+      "grad_norm": 0.14134864509105682,
+      "learning_rate": 9.555700751388435e-06,
+      "loss": 1.8781,
+      "step": 544
+    },
+    {
+      "epoch": 0.6355685131195336,
+      "grad_norm": 0.2804355323314667,
+      "learning_rate": 9.554556599918269e-06,
+      "loss": 1.8392,
+      "step": 545
+    },
+    {
+      "epoch": 0.636734693877551,
+      "grad_norm": 0.12909014523029327,
+      "learning_rate": 9.553410763945691e-06,
+      "loss": 1.9602,
+      "step": 546
+    },
+    {
+      "epoch": 0.6379008746355685,
+      "grad_norm": 0.0914696604013443,
+      "learning_rate": 9.552263239747894e-06,
+      "loss": 1.9395,
+      "step": 547
+    },
+    {
+      "epoch": 0.639067055393586,
+      "grad_norm": 0.10263511538505554,
+      "learning_rate": 9.551114023591088e-06,
+      "loss": 1.8783,
+      "step": 548
+    },
+    {
+      "epoch": 0.6402332361516035,
+      "grad_norm": 0.09729939699172974,
+      "learning_rate": 9.549963111730472e-06,
+      "loss": 1.8761,
+      "step": 549
+    },
+    {
+      "epoch": 0.641399416909621,
+      "grad_norm": 0.12668545544147491,
+      "learning_rate": 9.548810500410174e-06,
+      "loss": 1.9171,
+      "step": 550
+    },
+    {
+      "epoch": 0.6425655976676384,
+      "grad_norm": 0.14873801171779633,
+      "learning_rate": 9.54765618586323e-06,
+      "loss": 1.8446,
+      "step": 551
+    },
+    {
+      "epoch": 0.643731778425656,
+      "grad_norm": 0.08422703295946121,
+      "learning_rate": 9.546500164311535e-06,
+      "loss": 1.86,
+      "step": 552
+    },
+    {
+      "epoch": 0.6448979591836734,
+      "grad_norm": 0.08689919114112854,
+      "learning_rate": 9.5453424319658e-06,
+      "loss": 1.7977,
+      "step": 553
+    },
+    {
+      "epoch": 0.646064139941691,
+      "grad_norm": 0.20533175766468048,
+      "learning_rate": 9.544182985025507e-06,
+      "loss": 1.9284,
+      "step": 554
+    },
+    {
+      "epoch": 0.6472303206997084,
+      "grad_norm": 0.09089347720146179,
+      "learning_rate": 9.543021819678881e-06,
+      "loss": 1.9164,
+      "step": 555
+    },
+    {
+      "epoch": 0.648396501457726,
+      "grad_norm": 0.09898319095373154,
+      "learning_rate": 9.541858932102834e-06,
+      "loss": 1.917,
+      "step": 556
+    },
+    {
+      "epoch": 0.6495626822157434,
+      "grad_norm": 0.10247382521629333,
+      "learning_rate": 9.540694318462935e-06,
+      "loss": 1.8471,
+      "step": 557
+    },
+    {
+      "epoch": 0.650728862973761,
+      "grad_norm": 0.0948185995221138,
+      "learning_rate": 9.539527974913353e-06,
+      "loss": 1.9088,
+      "step": 558
+    },
+    {
+      "epoch": 0.6518950437317784,
+      "grad_norm": 0.08687960356473923,
+      "learning_rate": 9.53835989759683e-06,
+      "loss": 1.9397,
+      "step": 559
+    },
+    {
+      "epoch": 0.6530612244897959,
+      "grad_norm": 0.12131970375776291,
+      "learning_rate": 9.537190082644629e-06,
+      "loss": 1.8625,
+      "step": 560
+    },
+    {
+      "epoch": 0.6542274052478134,
+      "grad_norm": 0.11417805403470993,
+      "learning_rate": 9.536018526176497e-06,
+      "loss": 1.8793,
+      "step": 561
+    },
+    {
+      "epoch": 0.6553935860058309,
+      "grad_norm": 0.12107567489147186,
+      "learning_rate": 9.534845224300613e-06,
+      "loss": 1.9037,
+      "step": 562
+    },
+    {
+      "epoch": 0.6565597667638484,
+      "grad_norm": 0.08807747066020966,
+      "learning_rate": 9.53367017311356e-06,
+      "loss": 1.8665,
+      "step": 563
+    },
+    {
+      "epoch": 0.6577259475218659,
+      "grad_norm": 0.12042683362960815,
+      "learning_rate": 9.532493368700265e-06,
+      "loss": 1.8282,
+      "step": 564
+    },
+    {
+      "epoch": 0.6588921282798834,
+      "grad_norm": 0.09194262325763702,
+      "learning_rate": 9.53131480713397e-06,
+      "loss": 1.8688,
+      "step": 565
+    },
+    {
+      "epoch": 0.6600583090379009,
+      "grad_norm": 0.13839340209960938,
+      "learning_rate": 9.530134484476176e-06,
+      "loss": 1.935,
+      "step": 566
+    },
+    {
+      "epoch": 0.6612244897959184,
+      "grad_norm": 0.3190121352672577,
+      "learning_rate": 9.528952396776606e-06,
+      "loss": 1.8083,
+      "step": 567
+    },
+    {
+      "epoch": 0.6623906705539359,
+      "grad_norm": 0.08160486817359924,
+      "learning_rate": 9.527768540073162e-06,
+      "loss": 1.8863,
+      "step": 568
+    },
+    {
+      "epoch": 0.6635568513119533,
+      "grad_norm": 0.12480806559324265,
+      "learning_rate": 9.526582910391881e-06,
+      "loss": 1.8762,
+      "step": 569
+    },
+    {
+      "epoch": 0.6647230320699709,
+      "grad_norm": 0.17753319442272186,
+      "learning_rate": 9.525395503746879e-06,
+      "loss": 1.8534,
+      "step": 570
+    },
+    {
+      "epoch": 0.6658892128279883,
+      "grad_norm": 0.10315371304750443,
+      "learning_rate": 9.524206316140322e-06,
+      "loss": 1.9115,
+      "step": 571
+    },
+    {
+      "epoch": 0.6670553935860059,
+      "grad_norm": 0.14890041947364807,
+      "learning_rate": 9.523015343562375e-06,
+      "loss": 1.9504,
+      "step": 572
+    },
+    {
+      "epoch": 0.6682215743440233,
+      "grad_norm": 0.1643107384443283,
+      "learning_rate": 9.521822581991155e-06,
+      "loss": 1.9732,
+      "step": 573
+    },
+    {
+      "epoch": 0.6693877551020408,
+      "grad_norm": 0.12446587532758713,
+      "learning_rate": 9.520628027392686e-06,
+      "loss": 1.9468,
+      "step": 574
+    },
+    {
+      "epoch": 0.6705539358600583,
+      "grad_norm": 0.14547710120677948,
+      "learning_rate": 9.519431675720853e-06,
+      "loss": 1.8859,
+      "step": 575
+    },
+    {
+      "epoch": 0.6717201166180758,
+      "grad_norm": 0.1336713582277298,
+      "learning_rate": 9.518233522917366e-06,
+      "loss": 1.8651,
+      "step": 576
+    },
+    {
+      "epoch": 0.6728862973760933,
+      "grad_norm": 0.13357402384281158,
+      "learning_rate": 9.517033564911694e-06,
+      "loss": 1.9646,
+      "step": 577
+    },
+    {
+      "epoch": 0.6740524781341107,
+      "grad_norm": 0.08782308548688889,
+      "learning_rate": 9.515831797621042e-06,
+      "loss": 1.8497,
+      "step": 578
+    },
+    {
+      "epoch": 0.6752186588921283,
+      "grad_norm": 0.1420316845178604,
+      "learning_rate": 9.51462821695029e-06,
+      "loss": 1.8719,
+      "step": 579
+    },
+    {
+      "epoch": 0.6763848396501457,
+      "grad_norm": 0.1252625584602356,
+      "learning_rate": 9.513422818791947e-06,
+      "loss": 1.8201,
+      "step": 580
+    },
+    {
+      "epoch": 0.6775510204081633,
+      "grad_norm": 0.08772513270378113,
+      "learning_rate": 9.512215599026111e-06,
+      "loss": 1.8943,
+      "step": 581
+    },
+    {
+      "epoch": 0.6787172011661807,
+      "grad_norm": 0.12914720177650452,
+      "learning_rate": 9.511006553520417e-06,
+      "loss": 1.7758,
+      "step": 582
+    },
+    {
+      "epoch": 0.6798833819241983,
+      "grad_norm": 0.2095118761062622,
+      "learning_rate": 9.509795678129993e-06,
+      "loss": 1.8291,
+      "step": 583
+    },
+    {
+      "epoch": 0.6810495626822157,
+      "grad_norm": 0.11311589926481247,
+      "learning_rate": 9.508582968697409e-06,
+      "loss": 1.8255,
+      "step": 584
+    },
+    {
+      "epoch": 0.6822157434402333,
+      "grad_norm": 0.142658993601799,
+      "learning_rate": 9.507368421052633e-06,
+      "loss": 1.8415,
+      "step": 585
+    },
+    {
+      "epoch": 0.6833819241982507,
+      "grad_norm": 1.5993496179580688,
+      "learning_rate": 9.506152031012979e-06,
+      "loss": 1.9202,
+      "step": 586
+    },
+    {
+      "epoch": 0.6845481049562682,
+      "grad_norm": 0.13388904929161072,
+      "learning_rate": 9.504933794383066e-06,
+      "loss": 1.8299,
+      "step": 587
+    },
+    {
+      "epoch": 0.6857142857142857,
+      "grad_norm": 0.10805674642324448,
+      "learning_rate": 9.50371370695476e-06,
+      "loss": 1.816,
+      "step": 588
+    },
+    {
+      "epoch": 0.6868804664723032,
+      "grad_norm": 0.10440585017204285,
+      "learning_rate": 9.502491764507138e-06,
+      "loss": 1.8559,
+      "step": 589
+    },
+    {
+      "epoch": 0.6880466472303207,
+      "grad_norm": 0.1180737167596817,
+      "learning_rate": 9.501267962806425e-06,
+      "loss": 1.9121,
+      "step": 590
+    },
+    {
+      "epoch": 0.6892128279883382,
+      "grad_norm": 0.08112170547246933,
+      "learning_rate": 9.500042297605956e-06,
+      "loss": 1.8016,
+      "step": 591
+    },
+    {
+      "epoch": 0.6903790087463557,
+      "grad_norm": 0.1432664394378662,
+      "learning_rate": 9.498814764646125e-06,
+      "loss": 1.8358,
+      "step": 592
+    },
+    {
+      "epoch": 0.6915451895043732,
+      "grad_norm": 0.12890785932540894,
+      "learning_rate": 9.497585359654327e-06,
+      "loss": 1.7902,
+      "step": 593
+    },
+    {
+      "epoch": 0.6927113702623907,
+      "grad_norm": 0.0872940942645073,
+      "learning_rate": 9.496354078344922e-06,
+      "loss": 1.8917,
+      "step": 594
+    },
+    {
+      "epoch": 0.6938775510204082,
+      "grad_norm": 0.09819458425045013,
+      "learning_rate": 9.495120916419177e-06,
+      "loss": 1.9117,
+      "step": 595
+    },
+    {
+      "epoch": 0.6950437317784257,
+      "grad_norm": 0.08570613712072372,
+      "learning_rate": 9.49388586956522e-06,
+      "loss": 1.8474,
+      "step": 596
+    },
+    {
+      "epoch": 0.6962099125364432,
+      "grad_norm": 0.09564158320426941,
+      "learning_rate": 9.492648933457976e-06,
+      "loss": 1.8501,
+      "step": 597
+    },
+    {
+      "epoch": 0.6973760932944606,
+      "grad_norm": 0.3149571120738983,
+      "learning_rate": 9.491410103759143e-06,
+      "loss": 1.9041,
+      "step": 598
+    },
+    {
+      "epoch": 0.6985422740524782,
+      "grad_norm": 0.08295410126447678,
+      "learning_rate": 9.490169376117116e-06,
+      "loss": 1.81,
+      "step": 599
+    },
+    {
+      "epoch": 0.6997084548104956,
+      "grad_norm": 0.12500935792922974,
+      "learning_rate": 9.488926746166952e-06,
+      "loss": 1.8759,
+      "step": 600
+    },
+    {
+      "epoch": 0.7008746355685131,
+      "grad_norm": 0.21084064245224,
+      "learning_rate": 9.487682209530306e-06,
+      "loss": 1.8511,
+      "step": 601
+    },
+    {
+      "epoch": 0.7020408163265306,
+      "grad_norm": 0.09151198714971542,
+      "learning_rate": 9.48643576181539e-06,
+      "loss": 1.8656,
+      "step": 602
+    },
+    {
+      "epoch": 0.7020408163265306,
+      "eval_loss": 0.9252361059188843,
+      "eval_runtime": 168.3962,
+      "eval_samples_per_second": 40.999,
+      "eval_steps_per_second": 2.565,
+      "step": 602
+    },
+    {
+      "epoch": 0.7032069970845481,
+      "grad_norm": 0.10267394036054611,
+      "learning_rate": 9.485187398616922e-06,
+      "loss": 1.8167,
+      "step": 603
+    },
+    {
+      "epoch": 0.7043731778425656,
+      "grad_norm": 0.08673790842294693,
+      "learning_rate": 9.483937115516065e-06,
+      "loss": 1.8717,
+      "step": 604
+    },
+    {
+      "epoch": 0.7055393586005831,
+      "grad_norm": 0.32248324155807495,
+      "learning_rate": 9.482684908080378e-06,
+      "loss": 1.9179,
+      "step": 605
+    },
+    {
+      "epoch": 0.7067055393586006,
+      "grad_norm": 0.0887330025434494,
+      "learning_rate": 9.481430771863768e-06,
+      "loss": 1.8035,
+      "step": 606
+    },
+    {
+      "epoch": 0.707871720116618,
+      "grad_norm": 0.08606921136379242,
+      "learning_rate": 9.48017470240644e-06,
+      "loss": 1.8213,
+      "step": 607
+    },
+    {
+      "epoch": 0.7090379008746356,
+      "grad_norm": 0.08339248597621918,
+      "learning_rate": 9.478916695234832e-06,
+      "loss": 1.8126,
+      "step": 608
+    },
+    {
+      "epoch": 0.710204081632653,
+      "grad_norm": 0.1024995669722557,
+      "learning_rate": 9.477656745861567e-06,
+      "loss": 1.8228,
+      "step": 609
+    },
+    {
+      "epoch": 0.7113702623906706,
+      "grad_norm": 0.09097527712583542,
+      "learning_rate": 9.476394849785408e-06,
+      "loss": 1.9285,
+      "step": 610
+    },
+    {
+      "epoch": 0.712536443148688,
+      "grad_norm": 0.20047634840011597,
+      "learning_rate": 9.475131002491195e-06,
+      "loss": 1.9011,
+      "step": 611
+    },
+    {
+      "epoch": 0.7137026239067056,
+      "grad_norm": 0.09709218144416809,
+      "learning_rate": 9.473865199449796e-06,
+      "loss": 1.9978,
+      "step": 612
+    },
+    {
+      "epoch": 0.714868804664723,
+      "grad_norm": 0.10387609153985977,
+      "learning_rate": 9.472597436118044e-06,
+      "loss": 1.845,
+      "step": 613
+    },
+    {
+      "epoch": 0.7160349854227406,
+      "grad_norm": 0.13912031054496765,
+      "learning_rate": 9.471327707938696e-06,
+      "loss": 1.9319,
+      "step": 614
+    },
+    {
+      "epoch": 0.717201166180758,
+      "grad_norm": 0.07962010055780411,
+      "learning_rate": 9.47005601034037e-06,
+      "loss": 1.9118,
+      "step": 615
+    },
+    {
+      "epoch": 0.7183673469387755,
+      "grad_norm": 0.18047647178173065,
+      "learning_rate": 9.468782338737497e-06,
+      "loss": 1.8326,
+      "step": 616
+    },
+    {
+      "epoch": 0.719533527696793,
+      "grad_norm": 0.16449600458145142,
+      "learning_rate": 9.46750668853025e-06,
+      "loss": 1.8778,
+      "step": 617
+    },
+    {
+      "epoch": 0.7206997084548105,
+      "grad_norm": 0.32043328881263733,
+      "learning_rate": 9.46622905510451e-06,
+      "loss": 1.9176,
+      "step": 618
+    },
+    {
+      "epoch": 0.721865889212828,
+      "grad_norm": 0.18332763016223907,
+      "learning_rate": 9.464949433831794e-06,
+      "loss": 1.9424,
+      "step": 619
+    },
+    {
+      "epoch": 0.7230320699708455,
+      "grad_norm": 0.08142529428005219,
+      "learning_rate": 9.463667820069205e-06,
+      "loss": 1.8925,
+      "step": 620
+    },
+    {
+      "epoch": 0.724198250728863,
+      "grad_norm": 0.1315646767616272,
+      "learning_rate": 9.462384209159381e-06,
+      "loss": 1.903,
+      "step": 621
+    },
+    {
+      "epoch": 0.7253644314868805,
+      "grad_norm": 0.15011928975582123,
+      "learning_rate": 9.461098596430428e-06,
+      "loss": 1.916,
+      "step": 622
+    },
+    {
+      "epoch": 0.726530612244898,
+      "grad_norm": 0.08591946959495544,
+      "learning_rate": 9.459810977195875e-06,
+      "loss": 1.9016,
+      "step": 623
+    },
+    {
+      "epoch": 0.7276967930029155,
+      "grad_norm": 0.11847570538520813,
+      "learning_rate": 9.4585213467546e-06,
+      "loss": 1.8089,
+      "step": 624
+    },
+    {
+      "epoch": 0.7288629737609329,
+      "grad_norm": 0.08587644249200821,
+      "learning_rate": 9.457229700390796e-06,
+      "loss": 1.8419,
+      "step": 625
+    },
+    {
+      "epoch": 0.7300291545189505,
+      "grad_norm": 0.09782323241233826,
+      "learning_rate": 9.455936033373893e-06,
+      "loss": 1.9123,
+      "step": 626
+    },
+    {
+      "epoch": 0.7311953352769679,
+      "grad_norm": 0.1493980586528778,
+      "learning_rate": 9.454640340958512e-06,
+      "loss": 1.9243,
+      "step": 627
+    },
+    {
+      "epoch": 0.7323615160349854,
+      "grad_norm": 0.10012844949960709,
+      "learning_rate": 9.453342618384402e-06,
+      "loss": 1.8844,
+      "step": 628
+    },
+    {
+      "epoch": 0.7335276967930029,
+      "grad_norm": 0.3254757821559906,
+      "learning_rate": 9.452042860876383e-06,
+      "loss": 1.8144,
+      "step": 629
+    },
+    {
+      "epoch": 0.7346938775510204,
+      "grad_norm": 0.13703452050685883,
+      "learning_rate": 9.45074106364429e-06,
+      "loss": 1.873,
+      "step": 630
+    },
+    {
+      "epoch": 0.7358600583090379,
+      "grad_norm": 0.18141381442546844,
+      "learning_rate": 9.449437221882909e-06,
+      "loss": 1.9555,
+      "step": 631
+    },
+    {
+      "epoch": 0.7370262390670554,
+      "grad_norm": 0.15716956555843353,
+      "learning_rate": 9.448131330771919e-06,
+      "loss": 1.8405,
+      "step": 632
+    },
+    {
+      "epoch": 0.7381924198250729,
+      "grad_norm": 0.22397468984127045,
+      "learning_rate": 9.446823385475838e-06,
+      "loss": 1.9424,
+      "step": 633
+    },
+    {
+      "epoch": 0.7393586005830903,
+      "grad_norm": 0.08670644462108612,
+      "learning_rate": 9.445513381143957e-06,
+      "loss": 1.846,
+      "step": 634
+    },
+    {
+      "epoch": 0.7405247813411079,
+      "grad_norm": 0.09219978749752045,
+      "learning_rate": 9.444201312910286e-06,
+      "loss": 1.95,
+      "step": 635
+    },
+    {
+      "epoch": 0.7416909620991253,
+      "grad_norm": 0.11161966621875763,
+      "learning_rate": 9.442887175893483e-06,
+      "loss": 1.8607,
+      "step": 636
+    },
+    {
+      "epoch": 0.7428571428571429,
+      "grad_norm": 0.154244065284729,
+      "learning_rate": 9.44157096519681e-06,
+      "loss": 1.9209,
+      "step": 637
+    },
+    {
+      "epoch": 0.7440233236151603,
+      "grad_norm": 0.07722692936658859,
+      "learning_rate": 9.440252675908054e-06,
+      "loss": 1.8373,
+      "step": 638
+    },
+    {
+      "epoch": 0.7451895043731779,
+      "grad_norm": 0.09512025862932205,
+      "learning_rate": 9.438932303099483e-06,
+      "loss": 1.8445,
+      "step": 639
+    },
+    {
+      "epoch": 0.7463556851311953,
+      "grad_norm": 0.15499959886074066,
+      "learning_rate": 9.43760984182777e-06,
+      "loss": 1.9354,
+      "step": 640
+    },
+    {
+      "epoch": 0.7475218658892129,
+      "grad_norm": 0.15092675387859344,
+      "learning_rate": 9.436285287133939e-06,
+      "loss": 1.7867,
+      "step": 641
+    },
+    {
+      "epoch": 0.7486880466472303,
+      "grad_norm": 0.4668864905834198,
+      "learning_rate": 9.434958634043303e-06,
+      "loss": 1.895,
+      "step": 642
+    },
+    {
+      "epoch": 0.7498542274052478,
+      "grad_norm": 0.09367256611585617,
+      "learning_rate": 9.433629877565402e-06,
+      "loss": 1.8791,
+      "step": 643
+    },
+    {
+      "epoch": 0.7510204081632653,
+      "grad_norm": 0.09211242198944092,
+      "learning_rate": 9.432299012693936e-06,
+      "loss": 1.9084,
+      "step": 644
+    },
+    {
+      "epoch": 0.7521865889212828,
+      "grad_norm": 0.07713727653026581,
+      "learning_rate": 9.430966034406706e-06,
+      "loss": 1.8872,
+      "step": 645
+    },
+    {
+      "epoch": 0.7533527696793003,
+      "grad_norm": 0.14356684684753418,
+      "learning_rate": 9.429630937665548e-06,
+      "loss": 1.8771,
+      "step": 646
+    },
+    {
+      "epoch": 0.7545189504373178,
+      "grad_norm": 0.09995918720960617,
+      "learning_rate": 9.428293717416278e-06,
+      "loss": 1.9757,
+      "step": 647
+    },
+    {
+      "epoch": 0.7556851311953353,
+      "grad_norm": 0.11649483442306519,
+      "learning_rate": 9.426954368588612e-06,
+      "loss": 1.8929,
+      "step": 648
+    },
+    {
+      "epoch": 0.7568513119533528,
+      "grad_norm": 0.17411640286445618,
+      "learning_rate": 9.425612886096116e-06,
+      "loss": 1.8871,
+      "step": 649
+    },
+    {
+      "epoch": 0.7580174927113703,
+      "grad_norm": 0.11309409141540527,
+      "learning_rate": 9.424269264836138e-06,
+      "loss": 1.9396,
+      "step": 650
+    },
+    {
+      "epoch": 0.7591836734693878,
+      "grad_norm": 0.08775336295366287,
+      "learning_rate": 9.422923499689746e-06,
+      "loss": 1.9475,
+      "step": 651
+    },
+    {
+      "epoch": 0.7603498542274052,
+      "grad_norm": 0.1125795841217041,
+      "learning_rate": 9.421575585521646e-06,
+      "loss": 1.8973,
+      "step": 652
+    },
+    {
+      "epoch": 0.7615160349854228,
+      "grad_norm": 0.11909424513578415,
+      "learning_rate": 9.420225517180149e-06,
+      "loss": 1.8062,
+      "step": 653
+    },
+    {
+      "epoch": 0.7626822157434402,
+      "grad_norm": 0.07819771021604538,
+      "learning_rate": 9.418873289497067e-06,
+      "loss": 1.8005,
+      "step": 654
+    },
+    {
+      "epoch": 0.7638483965014577,
+      "grad_norm": 0.08252590149641037,
+      "learning_rate": 9.417518897287685e-06,
+      "loss": 1.8929,
+      "step": 655
+    },
+    {
+      "epoch": 0.7650145772594752,
+      "grad_norm": 0.0964493528008461,
+      "learning_rate": 9.41616233535066e-06,
+      "loss": 1.8582,
+      "step": 656
+    },
+    {
+      "epoch": 0.7661807580174927,
+      "grad_norm": 0.18199355900287628,
+      "learning_rate": 9.41480359846798e-06,
+      "loss": 1.8438,
+      "step": 657
+    },
+    {
+      "epoch": 0.7673469387755102,
+      "grad_norm": 0.10357444733381271,
+      "learning_rate": 9.413442681404886e-06,
+      "loss": 1.8697,
+      "step": 658
+    },
+    {
+      "epoch": 0.7685131195335277,
+      "grad_norm": 0.1019044741988182,
+      "learning_rate": 9.412079578909806e-06,
+      "loss": 1.8125,
+      "step": 659
+    },
+    {
+      "epoch": 0.7696793002915452,
+      "grad_norm": 0.15269735455513,
+      "learning_rate": 9.410714285714286e-06,
+      "loss": 1.8913,
+      "step": 660
+    },
+    {
+      "epoch": 0.7708454810495626,
+      "grad_norm": 0.10491170734167099,
+      "learning_rate": 9.409346796532928e-06,
+      "loss": 1.8368,
+      "step": 661
+    },
+    {
+      "epoch": 0.7720116618075802,
+      "grad_norm": 0.08167321234941483,
+      "learning_rate": 9.407977106063317e-06,
+      "loss": 1.9329,
+      "step": 662
+    },
+    {
+      "epoch": 0.7731778425655976,
+      "grad_norm": 0.12786783277988434,
+      "learning_rate": 9.40660520898595e-06,
+      "loss": 2.0004,
+      "step": 663
+    },
+    {
+      "epoch": 0.7743440233236152,
+      "grad_norm": 0.08830664306879044,
+      "learning_rate": 9.405231099964171e-06,
+      "loss": 1.8167,
+      "step": 664
+    },
+    {
+      "epoch": 0.7755102040816326,
+      "grad_norm": 0.08603803813457489,
+      "learning_rate": 9.403854773644106e-06,
+      "loss": 1.8784,
+      "step": 665
+    },
+    {
+      "epoch": 0.7766763848396502,
+      "grad_norm": 0.15332531929016113,
+      "learning_rate": 9.402476224654586e-06,
+      "loss": 1.8376,
+      "step": 666
+    },
+    {
+      "epoch": 0.7778425655976676,
+      "grad_norm": 0.13620127737522125,
+      "learning_rate": 9.401095447607077e-06,
+      "loss": 1.9483,
+      "step": 667
+    },
+    {
+      "epoch": 0.7790087463556852,
+      "grad_norm": 0.1036309003829956,
+      "learning_rate": 9.399712437095615e-06,
+      "loss": 1.8958,
+      "step": 668
+    },
+    {
+      "epoch": 0.7801749271137026,
+      "grad_norm": 0.08685003966093063,
+      "learning_rate": 9.398327187696734e-06,
+      "loss": 1.9102,
+      "step": 669
+    },
+    {
+      "epoch": 0.7813411078717201,
+      "grad_norm": 0.09234420210123062,
+      "learning_rate": 9.396939693969398e-06,
+      "loss": 1.7842,
+      "step": 670
+    },
+    {
+      "epoch": 0.7825072886297376,
+      "grad_norm": 0.08349122107028961,
+      "learning_rate": 9.395549950454915e-06,
+      "loss": 1.9258,
+      "step": 671
+    },
+    {
+      "epoch": 0.7836734693877551,
+      "grad_norm": 0.10310136526823044,
+      "learning_rate": 9.394157951676885e-06,
+      "loss": 1.8247,
+      "step": 672
+    },
+    {
+      "epoch": 0.7848396501457726,
+      "grad_norm": 0.09345996379852295,
+      "learning_rate": 9.392763692141117e-06,
+      "loss": 1.8527,
+      "step": 673
+    },
+    {
+      "epoch": 0.7860058309037901,
+      "grad_norm": 0.5228323340415955,
+      "learning_rate": 9.391367166335562e-06,
+      "loss": 1.846,
+      "step": 674
+    },
+    {
+      "epoch": 0.7871720116618076,
+      "grad_norm": 0.115789994597435,
+      "learning_rate": 9.38996836873023e-06,
+      "loss": 1.8583,
+      "step": 675
+    },
+    {
+      "epoch": 0.7883381924198251,
+      "grad_norm": 0.1000959500670433,
+      "learning_rate": 9.388567293777135e-06,
+      "loss": 1.801,
+      "step": 676
+    },
+    {
+      "epoch": 0.7895043731778426,
+      "grad_norm": 0.0929795652627945,
+      "learning_rate": 9.387163935910201e-06,
+      "loss": 1.8155,
+      "step": 677
+    },
+    {
+      "epoch": 0.79067055393586,
+      "grad_norm": 0.08656638115644455,
+      "learning_rate": 9.385758289545209e-06,
+      "loss": 1.8963,
+      "step": 678
+    },
+    {
+      "epoch": 0.7918367346938775,
+      "grad_norm": 0.1204112246632576,
+      "learning_rate": 9.384350349079699e-06,
+      "loss": 1.9006,
+      "step": 679
+    },
+    {
+      "epoch": 0.793002915451895,
+      "grad_norm": 0.09877511113882065,
+      "learning_rate": 9.382940108892923e-06,
+      "loss": 1.8328,
+      "step": 680
+    },
+    {
+      "epoch": 0.7941690962099125,
+      "grad_norm": 0.09568671882152557,
+      "learning_rate": 9.381527563345745e-06,
+      "loss": 1.789,
+      "step": 681
+    },
+    {
+      "epoch": 0.79533527696793,
+      "grad_norm": 0.08654316514730453,
+      "learning_rate": 9.380112706780586e-06,
+      "loss": 1.8895,
+      "step": 682
+    },
+    {
+      "epoch": 0.7965014577259475,
+      "grad_norm": 0.09575946629047394,
+      "learning_rate": 9.378695533521332e-06,
+      "loss": 1.9844,
+      "step": 683
+    },
+    {
+      "epoch": 0.797667638483965,
+      "grad_norm": 0.09180290997028351,
+      "learning_rate": 9.37727603787327e-06,
+      "loss": 1.8202,
+      "step": 684
+    },
+    {
+      "epoch": 0.7988338192419825,
+      "grad_norm": 0.08106855303049088,
+      "learning_rate": 9.375854214123007e-06,
+      "loss": 1.836,
+      "step": 685
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.07933083176612854,
+      "learning_rate": 9.374430056538392e-06,
+      "loss": 1.8152,
+      "step": 686
+    },
+    {
+      "epoch": 0.8011661807580175,
+      "grad_norm": 0.08843309432268143,
+      "learning_rate": 9.37300355936844e-06,
+      "loss": 1.9088,
+      "step": 687
+    },
+    {
+      "epoch": 0.8023323615160349,
+      "grad_norm": 0.08624003082513809,
+      "learning_rate": 9.371574716843259e-06,
+      "loss": 1.9515,
+      "step": 688
+    },
+    {
+      "epoch": 0.8023323615160349,
+      "eval_loss": 0.9231130480766296,
+      "eval_runtime": 167.9232,
+      "eval_samples_per_second": 41.114,
+      "eval_steps_per_second": 2.573,
+      "step": 688
+    },
+    {
+      "epoch": 0.8034985422740525,
+      "grad_norm": 0.2472323626279831,
+      "learning_rate": 9.370143523173965e-06,
+      "loss": 1.8335,
+      "step": 689
+    },
+    {
+      "epoch": 0.8046647230320699,
+      "grad_norm": 0.19017226994037628,
+      "learning_rate": 9.368709972552608e-06,
+      "loss": 1.8539,
+      "step": 690
+    },
+    {
+      "epoch": 0.8058309037900875,
+      "grad_norm": 0.162383571267128,
+      "learning_rate": 9.367274059152093e-06,
+      "loss": 1.799,
+      "step": 691
+    },
+    {
+      "epoch": 0.8069970845481049,
+      "grad_norm": 0.10180008411407471,
+      "learning_rate": 9.3658357771261e-06,
+      "loss": 1.9108,
+      "step": 692
+    },
+    {
+      "epoch": 0.8081632653061225,
+      "grad_norm": 0.08266910910606384,
+      "learning_rate": 9.364395120609007e-06,
+      "loss": 1.8898,
+      "step": 693
+    },
+    {
+      "epoch": 0.8093294460641399,
+      "grad_norm": 0.09119586646556854,
+      "learning_rate": 9.362952083715809e-06,
+      "loss": 1.8327,
+      "step": 694
+    },
+    {
+      "epoch": 0.8104956268221575,
+      "grad_norm": 0.09465663135051727,
+      "learning_rate": 9.361506660542032e-06,
+      "loss": 1.9319,
+      "step": 695
+    },
+    {
+      "epoch": 0.8116618075801749,
+      "grad_norm": 0.17535577714443207,
+      "learning_rate": 9.360058845163663e-06,
+      "loss": 1.8596,
+      "step": 696
+    },
+    {
+      "epoch": 0.8128279883381925,
+      "grad_norm": 0.11075892299413681,
+      "learning_rate": 9.358608631637065e-06,
+      "loss": 1.9333,
+      "step": 697
+    },
+    {
+      "epoch": 0.8139941690962099,
+      "grad_norm": 0.09013248980045319,
+      "learning_rate": 9.357156013998895e-06,
+      "loss": 1.9447,
+      "step": 698
+    },
+    {
+      "epoch": 0.8151603498542274,
+      "grad_norm": 0.22128431499004364,
+      "learning_rate": 9.355700986266017e-06,
+      "loss": 1.8915,
+      "step": 699
+    },
+    {
+      "epoch": 0.8163265306122449,
+      "grad_norm": 0.17227105796337128,
+      "learning_rate": 9.354243542435426e-06,
+      "loss": 1.9036,
+      "step": 700
+    },
+    {
+      "epoch": 0.8174927113702624,
+      "grad_norm": 0.2741061747074127,
+      "learning_rate": 9.352783676484168e-06,
+      "loss": 1.8678,
+      "step": 701
+    },
+    {
+      "epoch": 0.8186588921282799,
+      "grad_norm": 0.15897901356220245,
+      "learning_rate": 9.35132138236925e-06,
+      "loss": 1.9183,
+      "step": 702
+    },
+    {
+      "epoch": 0.8198250728862974,
+      "grad_norm": 0.16776350140571594,
+      "learning_rate": 9.34985665402756e-06,
+      "loss": 1.8769,
+      "step": 703
+    },
+    {
+      "epoch": 0.8209912536443149,
+      "grad_norm": 0.11619832366704941,
+      "learning_rate": 9.348389485375788e-06,
+      "loss": 1.8843,
+      "step": 704
+    },
+    {
+      "epoch": 0.8221574344023324,
+      "grad_norm": 0.09094776958227158,
+      "learning_rate": 9.34691987031033e-06,
+      "loss": 1.9197,
+      "step": 705
+    },
+    {
+      "epoch": 0.8233236151603499,
+      "grad_norm": 0.1494653970003128,
+      "learning_rate": 9.345447802707214e-06,
+      "loss": 1.9308,
+      "step": 706
+    },
+    {
+      "epoch": 0.8244897959183674,
+      "grad_norm": 0.11436711251735687,
+      "learning_rate": 9.343973276422011e-06,
+      "loss": 1.8203,
+      "step": 707
+    },
+    {
+      "epoch": 0.8256559766763848,
+      "grad_norm": 0.10043874382972717,
+      "learning_rate": 9.342496285289748e-06,
+      "loss": 1.8463,
+      "step": 708
+    },
+    {
+      "epoch": 0.8268221574344023,
+      "grad_norm": 0.15034151077270508,
+      "learning_rate": 9.341016823124827e-06,
+      "loss": 1.8266,
+      "step": 709
+    },
+    {
+      "epoch": 0.8279883381924198,
+      "grad_norm": 0.09148337692022324,
+      "learning_rate": 9.339534883720931e-06,
+      "loss": 1.7939,
+      "step": 710
+    },
+    {
+      "epoch": 0.8291545189504373,
+      "grad_norm": 0.2016422301530838,
+      "learning_rate": 9.338050460850946e-06,
+      "loss": 1.8544,
+      "step": 711
+    },
+    {
+      "epoch": 0.8303206997084548,
+      "grad_norm": 0.08696021139621735,
+      "learning_rate": 9.336563548266866e-06,
+      "loss": 1.8505,
+      "step": 712
+    },
+    {
+      "epoch": 0.8314868804664723,
+      "grad_norm": 0.09068041294813156,
+      "learning_rate": 9.335074139699713e-06,
+      "loss": 1.8874,
+      "step": 713
+    },
+    {
+      "epoch": 0.8326530612244898,
+      "grad_norm": 0.10973275452852249,
+      "learning_rate": 9.333582228859436e-06,
+      "loss": 1.8583,
+      "step": 714
+    },
+    {
+      "epoch": 0.8338192419825073,
+      "grad_norm": 0.22106347978115082,
+      "learning_rate": 9.332087809434845e-06,
+      "loss": 1.9537,
+      "step": 715
+    },
+    {
+      "epoch": 0.8349854227405248,
+      "grad_norm": 0.09458278119564056,
+      "learning_rate": 9.330590875093493e-06,
+      "loss": 1.8031,
+      "step": 716
+    },
+    {
+      "epoch": 0.8361516034985422,
+      "grad_norm": 0.08998506516218185,
+      "learning_rate": 9.329091419481616e-06,
+      "loss": 1.8264,
+      "step": 717
+    },
+    {
+      "epoch": 0.8373177842565598,
+      "grad_norm": 0.14205193519592285,
+      "learning_rate": 9.327589436224013e-06,
+      "loss": 1.9208,
+      "step": 718
+    },
+    {
+      "epoch": 0.8384839650145772,
+      "grad_norm": 0.09513915330171585,
+      "learning_rate": 9.326084918923986e-06,
+      "loss": 1.8483,
+      "step": 719
+    },
+    {
+      "epoch": 0.8396501457725948,
+      "grad_norm": 0.15128940343856812,
+      "learning_rate": 9.324577861163228e-06,
+      "loss": 1.9194,
+      "step": 720
+    },
+    {
+      "epoch": 0.8408163265306122,
+      "grad_norm": 0.11448760330677032,
+      "learning_rate": 9.323068256501739e-06,
+      "loss": 1.8645,
+      "step": 721
+    },
+    {
+      "epoch": 0.8419825072886298,
+      "grad_norm": 0.14805534482002258,
+      "learning_rate": 9.321556098477731e-06,
+      "loss": 1.8437,
+      "step": 722
+    },
+    {
+      "epoch": 0.8431486880466472,
+      "grad_norm": 0.09840725362300873,
+      "learning_rate": 9.320041380607542e-06,
+      "loss": 1.8448,
+      "step": 723
+    },
+    {
+      "epoch": 0.8443148688046648,
+      "grad_norm": 0.15363240242004395,
+      "learning_rate": 9.318524096385542e-06,
+      "loss": 1.8684,
+      "step": 724
+    },
+    {
+      "epoch": 0.8454810495626822,
+      "grad_norm": 0.08850331604480743,
+      "learning_rate": 9.317004239284033e-06,
+      "loss": 1.8741,
+      "step": 725
+    },
+    {
+      "epoch": 0.8466472303206997,
+      "grad_norm": 0.09563957899808884,
+      "learning_rate": 9.31548180275316e-06,
+      "loss": 1.9599,
+      "step": 726
+    },
+    {
+      "epoch": 0.8478134110787172,
+      "grad_norm": 0.08761660754680634,
+      "learning_rate": 9.313956780220817e-06,
+      "loss": 1.7969,
+      "step": 727
+    },
+    {
+      "epoch": 0.8489795918367347,
+      "grad_norm": 0.09987696260213852,
+      "learning_rate": 9.312429165092558e-06,
+      "loss": 1.9103,
+      "step": 728
+    },
+    {
+      "epoch": 0.8501457725947522,
+      "grad_norm": 0.09629472345113754,
+      "learning_rate": 9.31089895075149e-06,
+      "loss": 1.8356,
+      "step": 729
+    },
+    {
+      "epoch": 0.8513119533527697,
+      "grad_norm": 0.13861589133739471,
+      "learning_rate": 9.309366130558185e-06,
+      "loss": 1.8676,
+      "step": 730
+    },
+    {
+      "epoch": 0.8524781341107872,
+      "grad_norm": 0.09675567597150803,
+      "learning_rate": 9.307830697850584e-06,
+      "loss": 1.8549,
+      "step": 731
+    },
+    {
+      "epoch": 0.8536443148688047,
+      "grad_norm": 0.09785202145576477,
+      "learning_rate": 9.306292645943898e-06,
+      "loss": 1.8246,
+      "step": 732
+    },
+    {
+      "epoch": 0.8548104956268222,
+      "grad_norm": 0.08803767710924149,
+      "learning_rate": 9.304751968130514e-06,
+      "loss": 1.8893,
+      "step": 733
+    },
+    {
+      "epoch": 0.8559766763848397,
+      "grad_norm": 0.10022731125354767,
+      "learning_rate": 9.303208657679894e-06,
+      "loss": 1.7756,
+      "step": 734
+    },
+    {
+      "epoch": 0.8571428571428571,
+      "grad_norm": 0.09209822863340378,
+      "learning_rate": 9.30166270783848e-06,
+      "loss": 1.8842,
+      "step": 735
+    },
+    {
+      "epoch": 0.8583090379008746,
+      "grad_norm": 0.17527832090854645,
+      "learning_rate": 9.300114111829594e-06,
+      "loss": 1.8323,
+      "step": 736
+    },
+    {
+      "epoch": 0.8594752186588921,
+      "grad_norm": 0.12856149673461914,
+      "learning_rate": 9.298562862853338e-06,
+      "loss": 1.9057,
+      "step": 737
+    },
+    {
+      "epoch": 0.8606413994169096,
+      "grad_norm": 0.162838414311409,
+      "learning_rate": 9.297008954086493e-06,
+      "loss": 1.8944,
+      "step": 738
+    },
+    {
+      "epoch": 0.8618075801749271,
+      "grad_norm": 0.12630973756313324,
+      "learning_rate": 9.29545237868243e-06,
+      "loss": 1.9289,
+      "step": 739
+    },
+    {
+      "epoch": 0.8629737609329446,
+      "grad_norm": 0.10631614178419113,
+      "learning_rate": 9.293893129770992e-06,
+      "loss": 1.8252,
+      "step": 740
+    },
+    {
+      "epoch": 0.8641399416909621,
+      "grad_norm": 0.12933312356472015,
+      "learning_rate": 9.29233120045841e-06,
+      "loss": 1.8432,
+      "step": 741
+    },
+    {
+      "epoch": 0.8653061224489796,
+      "grad_norm": 0.15817859768867493,
+      "learning_rate": 9.290766583827186e-06,
+      "loss": 1.8735,
+      "step": 742
+    },
+    {
+      "epoch": 0.8664723032069971,
+      "grad_norm": 0.11883293092250824,
+      "learning_rate": 9.289199272936e-06,
+      "loss": 1.806,
+      "step": 743
+    },
+    {
+      "epoch": 0.8676384839650145,
+      "grad_norm": 0.11238062381744385,
+      "learning_rate": 9.28762926081961e-06,
+      "loss": 1.8743,
+      "step": 744
+    },
+    {
+      "epoch": 0.8688046647230321,
+      "grad_norm": 0.08764756470918655,
+      "learning_rate": 9.28605654048874e-06,
+      "loss": 1.8716,
+      "step": 745
+    },
+    {
+      "epoch": 0.8699708454810495,
+      "grad_norm": 0.11982334405183792,
+      "learning_rate": 9.284481104929983e-06,
+      "loss": 1.8533,
+      "step": 746
+    },
+    {
+      "epoch": 0.8711370262390671,
+      "grad_norm": 0.11582168936729431,
+      "learning_rate": 9.282902947105693e-06,
+      "loss": 1.8049,
+      "step": 747
+    },
+    {
+      "epoch": 0.8723032069970845,
+      "grad_norm": 0.09098189324140549,
+      "learning_rate": 9.281322059953884e-06,
+      "loss": 1.8616,
+      "step": 748
+    },
+    {
+      "epoch": 0.8734693877551021,
+      "grad_norm": 0.11811266839504242,
+      "learning_rate": 9.279738436388116e-06,
+      "loss": 1.8809,
+      "step": 749
+    },
+    {
+      "epoch": 0.8746355685131195,
+      "grad_norm": 0.1126595064997673,
+      "learning_rate": 9.278152069297403e-06,
+      "loss": 1.9131,
+      "step": 750
+    },
+    {
+      "epoch": 0.8758017492711371,
+      "grad_norm": 0.09005296975374222,
+      "learning_rate": 9.276562951546094e-06,
+      "loss": 1.8655,
+      "step": 751
+    },
+    {
+      "epoch": 0.8769679300291545,
+      "grad_norm": 0.13342715799808502,
+      "learning_rate": 9.274971075973777e-06,
+      "loss": 1.8901,
+      "step": 752
+    },
+    {
+      "epoch": 0.878134110787172,
+      "grad_norm": 0.09575760364532471,
+      "learning_rate": 9.273376435395158e-06,
+      "loss": 1.7941,
+      "step": 753
+    },
+    {
+      "epoch": 0.8793002915451895,
+      "grad_norm": 0.10113046318292618,
+      "learning_rate": 9.271779022599961e-06,
+      "loss": 1.9195,
+      "step": 754
+    },
+    {
+      "epoch": 0.880466472303207,
+      "grad_norm": 0.08846978843212128,
+      "learning_rate": 9.270178830352827e-06,
+      "loss": 1.8688,
+      "step": 755
+    },
+    {
+      "epoch": 0.8816326530612245,
+      "grad_norm": 0.10775884985923767,
+      "learning_rate": 9.26857585139319e-06,
+      "loss": 1.851,
+      "step": 756
+    },
+    {
+      "epoch": 0.882798833819242,
+      "grad_norm": 0.16205506026744843,
+      "learning_rate": 9.26697007843517e-06,
+      "loss": 1.8828,
+      "step": 757
+    },
+    {
+      "epoch": 0.8839650145772595,
+      "grad_norm": 0.12394057959318161,
+      "learning_rate": 9.265361504167476e-06,
+      "loss": 1.9834,
+      "step": 758
+    },
+    {
+      "epoch": 0.885131195335277,
+      "grad_norm": 0.1062767431139946,
+      "learning_rate": 9.263750121253274e-06,
+      "loss": 1.897,
+      "step": 759
+    },
+    {
+      "epoch": 0.8862973760932945,
+      "grad_norm": 0.09801100939512253,
+      "learning_rate": 9.262135922330098e-06,
+      "loss": 1.8589,
+      "step": 760
+    },
+    {
+      "epoch": 0.887463556851312,
+      "grad_norm": 0.1366087794303894,
+      "learning_rate": 9.260518900009718e-06,
+      "loss": 1.8214,
+      "step": 761
+    },
+    {
+      "epoch": 0.8886297376093294,
+      "grad_norm": 0.3029862642288208,
+      "learning_rate": 9.25889904687804e-06,
+      "loss": 1.8547,
+      "step": 762
+    },
+    {
+      "epoch": 0.889795918367347,
+      "grad_norm": 0.0928243100643158,
+      "learning_rate": 9.257276355494988e-06,
+      "loss": 1.7951,
+      "step": 763
+    },
+    {
+      "epoch": 0.8909620991253644,
+      "grad_norm": 0.11008118838071823,
+      "learning_rate": 9.255650818394389e-06,
+      "loss": 1.8659,
+      "step": 764
+    },
+    {
+      "epoch": 0.892128279883382,
+      "grad_norm": 0.16812412440776825,
+      "learning_rate": 9.254022428083862e-06,
+      "loss": 1.873,
+      "step": 765
+    },
+    {
+      "epoch": 0.8932944606413994,
+      "grad_norm": 0.10276354849338531,
+      "learning_rate": 9.2523911770447e-06,
+      "loss": 1.8475,
+      "step": 766
+    },
+    {
+      "epoch": 0.894460641399417,
+      "grad_norm": 0.08310912549495697,
+      "learning_rate": 9.250757057731757e-06,
+      "loss": 1.8221,
+      "step": 767
+    },
+    {
+      "epoch": 0.8956268221574344,
+      "grad_norm": 0.08800824731588364,
+      "learning_rate": 9.249120062573328e-06,
+      "loss": 1.7546,
+      "step": 768
+    },
+    {
+      "epoch": 0.8967930029154519,
+      "grad_norm": 0.0932387113571167,
+      "learning_rate": 9.247480183971034e-06,
+      "loss": 1.8248,
+      "step": 769
+    },
+    {
+      "epoch": 0.8979591836734694,
+      "grad_norm": 0.1079181358218193,
+      "learning_rate": 9.245837414299707e-06,
+      "loss": 1.8304,
+      "step": 770
+    },
+    {
+      "epoch": 0.8991253644314868,
+      "grad_norm": 0.10514263063669205,
+      "learning_rate": 9.244191745907263e-06,
+      "loss": 1.8529,
+      "step": 771
+    },
+    {
+      "epoch": 0.9002915451895044,
+      "grad_norm": 0.08364567160606384,
+      "learning_rate": 9.2425431711146e-06,
+      "loss": 1.8517,
+      "step": 772
+    },
+    {
+      "epoch": 0.9014577259475218,
+      "grad_norm": 0.10194061696529388,
+      "learning_rate": 9.240891682215459e-06,
+      "loss": 1.8403,
+      "step": 773
+    },
+    {
+      "epoch": 0.9026239067055394,
+      "grad_norm": 0.13915513455867767,
+      "learning_rate": 9.239237271476312e-06,
+      "loss": 1.9176,
+      "step": 774
+    },
+    {
+      "epoch": 0.9026239067055394,
+      "eval_loss": 0.9210033416748047,
+      "eval_runtime": 168.4952,
+      "eval_samples_per_second": 40.974,
+      "eval_steps_per_second": 2.564,
+      "step": 774
+    },
+    {
+      "epoch": 0.9037900874635568,
+      "grad_norm": 0.1305500715970993,
+      "learning_rate": 9.237579931136252e-06,
+      "loss": 1.8842,
+      "step": 775
+    },
+    {
+      "epoch": 0.9049562682215744,
+      "grad_norm": 0.09020356833934784,
+      "learning_rate": 9.235919653406854e-06,
+      "loss": 1.9072,
+      "step": 776
+    },
+    {
+      "epoch": 0.9061224489795918,
+      "grad_norm": 0.11344979703426361,
+      "learning_rate": 9.234256430472061e-06,
+      "loss": 1.9035,
+      "step": 777
+    },
+    {
+      "epoch": 0.9072886297376094,
+      "grad_norm": 0.2470046728849411,
+      "learning_rate": 9.232590254488065e-06,
+      "loss": 1.8946,
+      "step": 778
+    },
+    {
+      "epoch": 0.9084548104956268,
+      "grad_norm": 0.1029474064707756,
+      "learning_rate": 9.230921117583177e-06,
+      "loss": 1.8044,
+      "step": 779
+    },
+    {
+      "epoch": 0.9096209912536443,
+      "grad_norm": 0.07936690747737885,
+      "learning_rate": 9.229249011857708e-06,
+      "loss": 1.873,
+      "step": 780
+    },
+    {
+      "epoch": 0.9107871720116618,
+      "grad_norm": 0.1371641457080841,
+      "learning_rate": 9.22757392938384e-06,
+      "loss": 1.832,
+      "step": 781
+    },
+    {
+      "epoch": 0.9119533527696793,
+      "grad_norm": 0.08980944007635117,
+      "learning_rate": 9.225895862205503e-06,
+      "loss": 1.9369,
+      "step": 782
+    },
+    {
+      "epoch": 0.9131195335276968,
+      "grad_norm": 0.09361738711595535,
+      "learning_rate": 9.224214802338253e-06,
+      "loss": 1.8646,
+      "step": 783
+    },
+    {
+      "epoch": 0.9142857142857143,
+      "grad_norm": 0.08910297602415085,
+      "learning_rate": 9.22253074176914e-06,
+      "loss": 1.752,
+      "step": 784
+    },
+    {
+      "epoch": 0.9154518950437318,
+      "grad_norm": 0.11445487290620804,
+      "learning_rate": 9.220843672456576e-06,
+      "loss": 1.8817,
+      "step": 785
+    },
+    {
+      "epoch": 0.9166180758017493,
+      "grad_norm": 0.09215401858091354,
+      "learning_rate": 9.21915358633022e-06,
+      "loss": 1.8516,
+      "step": 786
+    },
+    {
+      "epoch": 0.9177842565597668,
+      "grad_norm": 0.11525017768144608,
+      "learning_rate": 9.217460475290842e-06,
+      "loss": 1.9253,
+      "step": 787
+    },
+    {
+      "epoch": 0.9189504373177843,
+      "grad_norm": 0.1402161568403244,
+      "learning_rate": 9.215764331210191e-06,
+      "loss": 1.8361,
+      "step": 788
+    },
+    {
+      "epoch": 0.9201166180758017,
+      "grad_norm": 0.08479564636945724,
+      "learning_rate": 9.214065145930871e-06,
+      "loss": 1.8419,
+      "step": 789
+    },
+    {
+      "epoch": 0.9212827988338192,
+      "grad_norm": 0.2989562153816223,
+      "learning_rate": 9.212362911266201e-06,
+      "loss": 1.8583,
+      "step": 790
+    },
+    {
+      "epoch": 0.9224489795918367,
+      "grad_norm": 0.21828261017799377,
+      "learning_rate": 9.2106576190001e-06,
+      "loss": 1.9032,
+      "step": 791
+    },
+    {
+      "epoch": 0.9236151603498542,
+      "grad_norm": 0.12120179831981659,
+      "learning_rate": 9.208949260886937e-06,
+      "loss": 1.9422,
+      "step": 792
+    },
+    {
+      "epoch": 0.9247813411078717,
+      "grad_norm": 0.09822405129671097,
+      "learning_rate": 9.207237828651406e-06,
+      "loss": 1.8808,
+      "step": 793
+    },
+    {
+      "epoch": 0.9259475218658892,
+      "grad_norm": 0.17311765253543854,
+      "learning_rate": 9.205523313988393e-06,
+      "loss": 1.8461,
+      "step": 794
+    },
+    {
+      "epoch": 0.9271137026239067,
+      "grad_norm": 0.08903677761554718,
+      "learning_rate": 9.203805708562844e-06,
+      "loss": 1.8084,
+      "step": 795
+    },
+    {
+      "epoch": 0.9282798833819242,
+      "grad_norm": 0.09159966558218002,
+      "learning_rate": 9.202085004009624e-06,
+      "loss": 1.937,
+      "step": 796
+    },
+    {
+      "epoch": 0.9294460641399417,
+      "grad_norm": 0.08562708646059036,
+      "learning_rate": 9.20036119193338e-06,
+      "loss": 1.914,
+      "step": 797
+    },
+    {
+      "epoch": 0.9306122448979591,
+      "grad_norm": 0.14180518686771393,
+      "learning_rate": 9.198634263908416e-06,
+      "loss": 1.9265,
+      "step": 798
+    },
+    {
+      "epoch": 0.9317784256559767,
+      "grad_norm": 0.08666256070137024,
+      "learning_rate": 9.19690421147854e-06,
+      "loss": 1.753,
+      "step": 799
+    },
+    {
+      "epoch": 0.9329446064139941,
+      "grad_norm": 0.1155882254242897,
+      "learning_rate": 9.195171026156942e-06,
+      "loss": 1.8584,
+      "step": 800
+    },
+    {
+      "epoch": 0.9341107871720117,
+      "grad_norm": 0.21061347424983978,
+      "learning_rate": 9.19343469942604e-06,
+      "loss": 1.9324,
+      "step": 801
+    },
+    {
+      "epoch": 0.9352769679300291,
+      "grad_norm": 0.11247839033603668,
+      "learning_rate": 9.191695222737351e-06,
+      "loss": 1.8117,
+      "step": 802
+    },
+    {
+      "epoch": 0.9364431486880467,
+      "grad_norm": 1.1168044805526733,
+      "learning_rate": 9.189952587511348e-06,
+      "loss": 1.8628,
+      "step": 803
+    },
+    {
+      "epoch": 0.9376093294460641,
+      "grad_norm": 0.08525588363409042,
+      "learning_rate": 9.188206785137318e-06,
+      "loss": 1.7902,
+      "step": 804
+    },
+    {
+      "epoch": 0.9387755102040817,
+      "grad_norm": 0.08768081665039062,
+      "learning_rate": 9.186457806973219e-06,
+      "loss": 1.7508,
+      "step": 805
+    },
+    {
+      "epoch": 0.9399416909620991,
+      "grad_norm": 0.08813807368278503,
+      "learning_rate": 9.18470564434554e-06,
+      "loss": 1.8715,
+      "step": 806
+    },
+    {
+      "epoch": 0.9411078717201167,
+      "grad_norm": 0.11474096775054932,
+      "learning_rate": 9.182950288549154e-06,
+      "loss": 1.9118,
+      "step": 807
+    },
+    {
+      "epoch": 0.9422740524781341,
+      "grad_norm": 0.09381350129842758,
+      "learning_rate": 9.181191730847184e-06,
+      "loss": 1.9112,
+      "step": 808
+    },
+    {
+      "epoch": 0.9434402332361516,
+      "grad_norm": 0.13493011891841888,
+      "learning_rate": 9.179429962470839e-06,
+      "loss": 1.9271,
+      "step": 809
+    },
+    {
+      "epoch": 0.9446064139941691,
+      "grad_norm": 0.17446169257164001,
+      "learning_rate": 9.177664974619289e-06,
+      "loss": 1.8295,
+      "step": 810
+    },
+    {
+      "epoch": 0.9457725947521866,
+      "grad_norm": 0.0914335697889328,
+      "learning_rate": 9.175896758459505e-06,
+      "loss": 1.9726,
+      "step": 811
+    },
+    {
+      "epoch": 0.9469387755102041,
+      "grad_norm": 0.10693757236003876,
+      "learning_rate": 9.17412530512612e-06,
+      "loss": 1.9068,
+      "step": 812
+    },
+    {
+      "epoch": 0.9481049562682216,
+      "grad_norm": 0.09492238610982895,
+      "learning_rate": 9.172350605721267e-06,
+      "loss": 1.9149,
+      "step": 813
+    },
+    {
+      "epoch": 0.9492711370262391,
+      "grad_norm": 0.29552215337753296,
+      "learning_rate": 9.170572651314448e-06,
+      "loss": 1.8509,
+      "step": 814
+    },
+    {
+      "epoch": 0.9504373177842566,
+      "grad_norm": 0.08109095692634583,
+      "learning_rate": 9.168791432942376e-06,
+      "loss": 1.885,
+      "step": 815
+    },
+    {
+      "epoch": 0.9516034985422741,
+      "grad_norm": 0.09659001231193542,
+      "learning_rate": 9.16700694160882e-06,
+      "loss": 1.8139,
+      "step": 816
+    },
+    {
+      "epoch": 0.9527696793002915,
+      "grad_norm": 0.4676586091518402,
+      "learning_rate": 9.165219168284459e-06,
+      "loss": 1.9269,
+      "step": 817
+    },
+    {
+      "epoch": 0.953935860058309,
+      "grad_norm": 0.08867879956960678,
+      "learning_rate": 9.16342810390673e-06,
+      "loss": 1.9196,
+      "step": 818
+    },
+    {
+      "epoch": 0.9551020408163265,
+      "grad_norm": 0.11153396964073181,
+      "learning_rate": 9.161633739379672e-06,
+      "loss": 1.9091,
+      "step": 819
+    },
+    {
+      "epoch": 0.956268221574344,
+      "grad_norm": 0.08935698866844177,
+      "learning_rate": 9.159836065573772e-06,
+      "loss": 1.9463,
+      "step": 820
+    },
+    {
+      "epoch": 0.9574344023323615,
+      "grad_norm": 0.10557762533426285,
+      "learning_rate": 9.158035073325813e-06,
+      "loss": 1.8479,
+      "step": 821
+    },
+    {
+      "epoch": 0.958600583090379,
+      "grad_norm": 0.10766385495662689,
+      "learning_rate": 9.156230753438718e-06,
+      "loss": 1.9359,
+      "step": 822
+    },
+    {
+      "epoch": 0.9597667638483965,
+      "grad_norm": 0.08697371929883957,
+      "learning_rate": 9.154423096681394e-06,
+      "loss": 1.8596,
+      "step": 823
+    },
+    {
+      "epoch": 0.960932944606414,
+      "grad_norm": 0.1766807585954666,
+      "learning_rate": 9.152612093788566e-06,
+      "loss": 1.9232,
+      "step": 824
+    },
+    {
+      "epoch": 0.9620991253644315,
+      "grad_norm": 0.09576472640037537,
+      "learning_rate": 9.150797735460628e-06,
+      "loss": 1.8288,
+      "step": 825
+    },
+    {
+      "epoch": 0.963265306122449,
+      "grad_norm": 0.14343927800655365,
+      "learning_rate": 9.148980012363486e-06,
+      "loss": 1.8505,
+      "step": 826
+    },
+    {
+      "epoch": 0.9644314868804664,
+      "grad_norm": 0.10012530535459518,
+      "learning_rate": 9.147158915128392e-06,
+      "loss": 1.8127,
+      "step": 827
+    },
+    {
+      "epoch": 0.965597667638484,
+      "grad_norm": 0.11603701114654541,
+      "learning_rate": 9.145334434351775e-06,
+      "loss": 1.8145,
+      "step": 828
+    },
+    {
+      "epoch": 0.9667638483965014,
+      "grad_norm": 0.14052139222621918,
+      "learning_rate": 9.143506560595103e-06,
+      "loss": 1.7906,
+      "step": 829
+    },
+    {
+      "epoch": 0.967930029154519,
+      "grad_norm": 0.09027399122714996,
+      "learning_rate": 9.141675284384694e-06,
+      "loss": 1.8421,
+      "step": 830
+    },
+    {
+      "epoch": 0.9690962099125364,
+      "grad_norm": 0.12914690375328064,
+      "learning_rate": 9.139840596211573e-06,
+      "loss": 1.9388,
+      "step": 831
+    },
+    {
+      "epoch": 0.970262390670554,
+      "grad_norm": 0.08926420658826828,
+      "learning_rate": 9.13800248653129e-06,
+      "loss": 1.8656,
+      "step": 832
+    },
+    {
+      "epoch": 0.9714285714285714,
+      "grad_norm": 0.08020973205566406,
+      "learning_rate": 9.136160945763766e-06,
+      "loss": 1.8601,
+      "step": 833
+    },
+    {
+      "epoch": 0.972594752186589,
+      "grad_norm": 0.08306428790092468,
+      "learning_rate": 9.134315964293128e-06,
+      "loss": 1.8361,
+      "step": 834
+    },
+    {
+      "epoch": 0.9737609329446064,
+      "grad_norm": 0.08154436200857162,
+      "learning_rate": 9.132467532467534e-06,
+      "loss": 1.8419,
+      "step": 835
+    },
+    {
+      "epoch": 0.9749271137026239,
+      "grad_norm": 0.09556984156370163,
+      "learning_rate": 9.130615640599002e-06,
+      "loss": 1.9189,
+      "step": 836
+    },
+    {
+      "epoch": 0.9760932944606414,
+      "grad_norm": 0.14374229311943054,
+      "learning_rate": 9.128760278963258e-06,
+      "loss": 1.8935,
+      "step": 837
+    },
+    {
+      "epoch": 0.9772594752186589,
+      "grad_norm": 0.09599879384040833,
+      "learning_rate": 9.126901437799541e-06,
+      "loss": 1.836,
+      "step": 838
+    },
+    {
+      "epoch": 0.9784256559766764,
+      "grad_norm": 0.09107203781604767,
+      "learning_rate": 9.12503910731046e-06,
+      "loss": 1.8651,
+      "step": 839
+    },
+    {
+      "epoch": 0.9795918367346939,
+      "grad_norm": 0.21812140941619873,
+      "learning_rate": 9.123173277661797e-06,
+      "loss": 1.9343,
+      "step": 840
+    },
+    {
+      "epoch": 0.9807580174927114,
+      "grad_norm": 0.09028768539428711,
+      "learning_rate": 9.121303938982343e-06,
+      "loss": 1.8157,
+      "step": 841
+    },
+    {
+      "epoch": 0.9819241982507289,
+      "grad_norm": 0.21123015880584717,
+      "learning_rate": 9.119431081363731e-06,
+      "loss": 1.8909,
+      "step": 842
+    },
+    {
+      "epoch": 0.9830903790087464,
+      "grad_norm": 0.10787145793437958,
+      "learning_rate": 9.117554694860254e-06,
+      "loss": 1.9404,
+      "step": 843
+    },
+    {
+      "epoch": 0.9842565597667638,
+      "grad_norm": 0.4299691319465637,
+      "learning_rate": 9.115674769488686e-06,
+      "loss": 1.8644,
+      "step": 844
+    },
+    {
+      "epoch": 0.9854227405247813,
+      "grad_norm": 0.08584709465503693,
+      "learning_rate": 9.113791295228107e-06,
+      "loss": 1.8833,
+      "step": 845
+    },
+    {
+      "epoch": 0.9865889212827988,
+      "grad_norm": 0.21131350100040436,
+      "learning_rate": 9.111904262019735e-06,
+      "loss": 1.8932,
+      "step": 846
+    },
+    {
+      "epoch": 0.9877551020408163,
+      "grad_norm": 0.1512080281972885,
+      "learning_rate": 9.110013659766734e-06,
+      "loss": 1.8413,
+      "step": 847
+    },
+    {
+      "epoch": 0.9889212827988338,
+      "grad_norm": 0.10453635454177856,
+      "learning_rate": 9.108119478334036e-06,
+      "loss": 1.8959,
+      "step": 848
+    },
+    {
+      "epoch": 0.9900874635568513,
+      "grad_norm": 0.10764706879854202,
+      "learning_rate": 9.106221707548163e-06,
+      "loss": 1.8758,
+      "step": 849
+    },
+    {
+      "epoch": 0.9912536443148688,
+      "grad_norm": 0.10833363980054855,
+      "learning_rate": 9.104320337197048e-06,
+      "loss": 1.7993,
+      "step": 850
+    },
+    {
+      "epoch": 0.9924198250728863,
+      "grad_norm": 0.08290691673755646,
+      "learning_rate": 9.10241535702985e-06,
+      "loss": 1.8149,
+      "step": 851
+    },
+    {
+      "epoch": 0.9935860058309038,
+      "grad_norm": 0.09775282442569733,
+      "learning_rate": 9.100506756756757e-06,
+      "loss": 1.8437,
+      "step": 852
+    },
+    {
+      "epoch": 0.9947521865889213,
+      "grad_norm": 0.08646391332149506,
+      "learning_rate": 9.098594526048822e-06,
+      "loss": 1.8525,
+      "step": 853
+    },
+    {
+      "epoch": 0.9959183673469387,
+      "grad_norm": 0.09908445179462433,
+      "learning_rate": 9.096678654537761e-06,
+      "loss": 1.8512,
+      "step": 854
+    },
+    {
+      "epoch": 0.9970845481049563,
+      "grad_norm": 0.11743435263633728,
+      "learning_rate": 9.094759131815776e-06,
+      "loss": 1.9144,
+      "step": 855
+    },
+    {
+      "epoch": 0.9982507288629737,
+      "grad_norm": 0.09963498264551163,
+      "learning_rate": 9.092835947435355e-06,
+      "loss": 1.8574,
+      "step": 856
+    },
+    {
+      "epoch": 0.9994169096209913,
+      "grad_norm": 0.12829691171646118,
+      "learning_rate": 9.090909090909091e-06,
+      "loss": 1.9464,
+      "step": 857
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.1249983087182045,
+      "learning_rate": 9.088978551709493e-06,
+      "loss": 1.8192,
+      "step": 858
+    },
+    {
+      "epoch": 1.0011661807580174,
+      "grad_norm": 0.0866197869181633,
+      "learning_rate": 9.087044319268786e-06,
+      "loss": 1.8615,
+      "step": 859
+    },
+    {
+      "epoch": 1.0023323615160349,
+      "grad_norm": 0.08765610307455063,
+      "learning_rate": 9.085106382978724e-06,
+      "loss": 1.8255,
+      "step": 860
+    },
+    {
+      "epoch": 1.0023323615160349,
+      "eval_loss": 0.9189937710762024,
+      "eval_runtime": 168.333,
+      "eval_samples_per_second": 41.014,
+      "eval_steps_per_second": 2.566,
+      "step": 860
+    },
+    {
+      "epoch": 1.0034985422740526,
+      "grad_norm": 0.09086284041404724,
+      "learning_rate": 9.083164732190395e-06,
+      "loss": 1.858,
+      "step": 861
+    },
+    {
+      "epoch": 1.00466472303207,
+      "grad_norm": 0.10080280154943466,
+      "learning_rate": 9.081219356214028e-06,
+      "loss": 1.9324,
+      "step": 862
+    },
+    {
+      "epoch": 1.0058309037900874,
+      "grad_norm": 0.14213347434997559,
+      "learning_rate": 9.079270244318787e-06,
+      "loss": 1.8078,
+      "step": 863
+    },
+    {
+      "epoch": 1.0069970845481049,
+      "grad_norm": 0.23918505012989044,
+      "learning_rate": 9.077317385732594e-06,
+      "loss": 1.8261,
+      "step": 864
+    },
+    {
+      "epoch": 1.0081632653061225,
+      "grad_norm": 0.13214808702468872,
+      "learning_rate": 9.075360769641903e-06,
+      "loss": 1.852,
+      "step": 865
+    },
+    {
+      "epoch": 1.00932944606414,
+      "grad_norm": 0.08332618325948715,
+      "learning_rate": 9.073400385191528e-06,
+      "loss": 1.8547,
+      "step": 866
+    },
+    {
+      "epoch": 1.0104956268221574,
+      "grad_norm": 0.0968194380402565,
+      "learning_rate": 9.071436221484416e-06,
+      "loss": 1.8525,
+      "step": 867
+    },
+    {
+      "epoch": 1.0116618075801749,
+      "grad_norm": 0.18036837875843048,
+      "learning_rate": 9.069468267581475e-06,
+      "loss": 1.8237,
+      "step": 868
+    },
+    {
+      "epoch": 1.0128279883381923,
+      "grad_norm": 0.12840190529823303,
+      "learning_rate": 9.067496512501342e-06,
+      "loss": 1.9401,
+      "step": 869
+    },
+    {
+      "epoch": 1.01399416909621,
+      "grad_norm": 0.0885062962770462,
+      "learning_rate": 9.065520945220195e-06,
+      "loss": 1.9239,
+      "step": 870
+    },
+    {
+      "epoch": 1.0151603498542274,
+      "grad_norm": 0.10927964746952057,
+      "learning_rate": 9.06354155467154e-06,
+      "loss": 1.8594,
+      "step": 871
+    },
+    {
+      "epoch": 1.0163265306122449,
+      "grad_norm": 0.10928229987621307,
+      "learning_rate": 9.06155832974602e-06,
+      "loss": 1.8492,
+      "step": 872
+    },
+    {
+      "epoch": 1.0174927113702623,
+      "grad_norm": 0.0840403288602829,
+      "learning_rate": 9.059571259291177e-06,
+      "loss": 1.8909,
+      "step": 873
+    },
+    {
+      "epoch": 1.01865889212828,
+      "grad_norm": 0.1162007600069046,
+      "learning_rate": 9.05758033211128e-06,
+      "loss": 1.8332,
+      "step": 874
+    },
+    {
+      "epoch": 1.0198250728862974,
+      "grad_norm": 0.11552979797124863,
+      "learning_rate": 9.055585536967081e-06,
+      "loss": 1.8592,
+      "step": 875
+    },
+    {
+      "epoch": 1.0209912536443149,
+      "grad_norm": 0.10381388664245605,
+      "learning_rate": 9.053586862575627e-06,
+      "loss": 1.7963,
+      "step": 876
+    },
+    {
+      "epoch": 1.0221574344023323,
+      "grad_norm": 0.10671024769544601,
+      "learning_rate": 9.051584297610035e-06,
+      "loss": 1.9262,
+      "step": 877
+    },
+    {
+      "epoch": 1.0233236151603498,
+      "grad_norm": 0.10842776298522949,
+      "learning_rate": 9.049577830699287e-06,
+      "loss": 1.8576,
+      "step": 878
+    },
+    {
+      "epoch": 1.0244897959183674,
+      "grad_norm": 0.1237003281712532,
+      "learning_rate": 9.047567450427999e-06,
+      "loss": 1.8266,
+      "step": 879
+    },
+    {
+      "epoch": 1.0256559766763849,
+      "grad_norm": 0.0878075659275055,
+      "learning_rate": 9.045553145336226e-06,
+      "loss": 1.9068,
+      "step": 880
+    },
+    {
+      "epoch": 1.0268221574344023,
+      "grad_norm": 0.10865923762321472,
+      "learning_rate": 9.043534903919228e-06,
+      "loss": 1.8889,
+      "step": 881
+    },
+    {
+      "epoch": 1.0279883381924197,
+      "grad_norm": 0.3292934000492096,
+      "learning_rate": 9.041512714627257e-06,
+      "loss": 1.8899,
+      "step": 882
+    },
+    {
+      "epoch": 1.0291545189504374,
+      "grad_norm": 0.11215526610612869,
+      "learning_rate": 9.039486565865331e-06,
+      "loss": 1.8358,
+      "step": 883
+    },
+    {
+      "epoch": 1.0303206997084549,
+      "grad_norm": 0.18438370525836945,
+      "learning_rate": 9.03745644599303e-06,
+      "loss": 1.7884,
+      "step": 884
+    },
+    {
+      "epoch": 1.0314868804664723,
+      "grad_norm": 0.13566279411315918,
+      "learning_rate": 9.035422343324251e-06,
+      "loss": 1.844,
+      "step": 885
+    },
+    {
+      "epoch": 1.0326530612244897,
+      "grad_norm": 0.23604775965213776,
+      "learning_rate": 9.033384246126992e-06,
+      "loss": 1.8865,
+      "step": 886
+    },
+    {
+      "epoch": 1.0338192419825072,
+      "grad_norm": 0.08480603992938995,
+      "learning_rate": 9.03134214262313e-06,
+      "loss": 1.8683,
+      "step": 887
+    },
+    {
+      "epoch": 1.0349854227405249,
+      "grad_norm": 2.1027846336364746,
+      "learning_rate": 9.029296020988195e-06,
+      "loss": 1.8727,
+      "step": 888
+    },
+    {
+      "epoch": 1.0361516034985423,
+      "grad_norm": 0.08974109590053558,
+      "learning_rate": 9.027245869351132e-06,
+      "loss": 1.8463,
+      "step": 889
+    },
+    {
+      "epoch": 1.0373177842565597,
+      "grad_norm": 0.11400934308767319,
+      "learning_rate": 9.025191675794087e-06,
+      "loss": 1.7369,
+      "step": 890
+    },
+    {
+      "epoch": 1.0384839650145772,
+      "grad_norm": 0.13008803129196167,
+      "learning_rate": 9.023133428352156e-06,
+      "loss": 1.8566,
+      "step": 891
+    },
+    {
+      "epoch": 1.0396501457725948,
+      "grad_norm": 0.24522775411605835,
+      "learning_rate": 9.02107111501317e-06,
+      "loss": 1.8446,
+      "step": 892
+    },
+    {
+      "epoch": 1.0408163265306123,
+      "grad_norm": 0.08870100975036621,
+      "learning_rate": 9.019004723717455e-06,
+      "loss": 1.9069,
+      "step": 893
+    },
+    {
+      "epoch": 1.0419825072886297,
+      "grad_norm": 0.09477581083774567,
+      "learning_rate": 9.016934242357598e-06,
+      "loss": 1.8947,
+      "step": 894
+    },
+    {
+      "epoch": 1.0431486880466472,
+      "grad_norm": 0.08861856162548065,
+      "learning_rate": 9.014859658778208e-06,
+      "loss": 1.8765,
+      "step": 895
+    },
+    {
+      "epoch": 1.0443148688046646,
+      "grad_norm": 0.13763193786144257,
+      "learning_rate": 9.012780960775672e-06,
+      "loss": 1.8574,
+      "step": 896
+    },
+    {
+      "epoch": 1.0454810495626823,
+      "grad_norm": 0.08719266206026077,
+      "learning_rate": 9.010698136097938e-06,
+      "loss": 1.8331,
+      "step": 897
+    },
+    {
+      "epoch": 1.0466472303206997,
+      "grad_norm": 0.09142394363880157,
+      "learning_rate": 9.00861117244425e-06,
+      "loss": 1.905,
+      "step": 898
+    },
+    {
+      "epoch": 1.0478134110787172,
+      "grad_norm": 0.09117084741592407,
+      "learning_rate": 9.006520057464915e-06,
+      "loss": 1.8698,
+      "step": 899
+    },
+    {
+      "epoch": 1.0489795918367346,
+      "grad_norm": 0.09127234667539597,
+      "learning_rate": 9.004424778761062e-06,
+      "loss": 1.835,
+      "step": 900
+    },
+    {
+      "epoch": 1.0501457725947523,
+      "grad_norm": 0.09196560084819794,
+      "learning_rate": 9.0023253238844e-06,
+      "loss": 1.8419,
+      "step": 901
+    },
+    {
+      "epoch": 1.0513119533527697,
+      "grad_norm": 0.11504214257001877,
+      "learning_rate": 9.000221680336954e-06,
+      "loss": 1.878,
+      "step": 902
+    },
+    {
+      "epoch": 1.0524781341107872,
+      "grad_norm": 0.08450236171483994,
+      "learning_rate": 8.998113835570843e-06,
+      "loss": 1.8171,
+      "step": 903
+    },
+    {
+      "epoch": 1.0536443148688046,
+      "grad_norm": 0.2504514455795288,
+      "learning_rate": 8.996001776988006e-06,
+      "loss": 1.8202,
+      "step": 904
+    },
+    {
+      "epoch": 1.054810495626822,
+      "grad_norm": 0.11913991719484329,
+      "learning_rate": 8.993885491939967e-06,
+      "loss": 1.7819,
+      "step": 905
+    },
+    {
+      "epoch": 1.0559766763848397,
+      "grad_norm": 0.09809495508670807,
+      "learning_rate": 8.991764967727577e-06,
+      "loss": 1.8898,
+      "step": 906
+    },
+    {
+      "epoch": 1.0571428571428572,
+      "grad_norm": 0.2526063919067383,
+      "learning_rate": 8.989640191600757e-06,
+      "loss": 1.8592,
+      "step": 907
+    },
+    {
+      "epoch": 1.0583090379008746,
+      "grad_norm": 0.08389268815517426,
+      "learning_rate": 8.987511150758253e-06,
+      "loss": 1.9098,
+      "step": 908
+    },
+    {
+      "epoch": 1.059475218658892,
+      "grad_norm": 0.12303835153579712,
+      "learning_rate": 8.985377832347362e-06,
+      "loss": 1.9752,
+      "step": 909
+    },
+    {
+      "epoch": 1.0606413994169097,
+      "grad_norm": 0.09074362367391586,
+      "learning_rate": 8.983240223463687e-06,
+      "loss": 1.8784,
+      "step": 910
+    },
+    {
+      "epoch": 1.0618075801749272,
+      "grad_norm": 0.08091694861650467,
+      "learning_rate": 8.98109831115088e-06,
+      "loss": 1.8595,
+      "step": 911
+    },
+    {
+      "epoch": 1.0629737609329446,
+      "grad_norm": 0.08592483401298523,
+      "learning_rate": 8.978952082400358e-06,
+      "loss": 1.7383,
+      "step": 912
+    },
+    {
+      "epoch": 1.064139941690962,
+      "grad_norm": 0.09130120277404785,
+      "learning_rate": 8.97680152415107e-06,
+      "loss": 1.809,
+      "step": 913
+    },
+    {
+      "epoch": 1.0653061224489795,
+      "grad_norm": 0.08186627924442291,
+      "learning_rate": 8.974646623289207e-06,
+      "loss": 1.7952,
+      "step": 914
+    },
+    {
+      "epoch": 1.0664723032069972,
+      "grad_norm": 0.10355684161186218,
+      "learning_rate": 8.972487366647952e-06,
+      "loss": 1.8991,
+      "step": 915
+    },
+    {
+      "epoch": 1.0676384839650146,
+      "grad_norm": 0.08144935220479965,
+      "learning_rate": 8.970323741007195e-06,
+      "loss": 1.9092,
+      "step": 916
+    },
+    {
+      "epoch": 1.068804664723032,
+      "grad_norm": 0.08058435469865799,
+      "learning_rate": 8.968155733093282e-06,
+      "loss": 1.8046,
+      "step": 917
+    },
+    {
+      "epoch": 1.0699708454810495,
+      "grad_norm": 0.16002431511878967,
+      "learning_rate": 8.965983329578735e-06,
+      "loss": 1.7702,
+      "step": 918
+    },
+    {
+      "epoch": 1.0711370262390671,
+      "grad_norm": 0.34844258427619934,
+      "learning_rate": 8.963806517081973e-06,
+      "loss": 1.8649,
+      "step": 919
+    },
+    {
+      "epoch": 1.0723032069970846,
+      "grad_norm": 0.09179319441318512,
+      "learning_rate": 8.961625282167043e-06,
+      "loss": 1.821,
+      "step": 920
+    },
+    {
+      "epoch": 1.073469387755102,
+      "grad_norm": 0.08796966820955276,
+      "learning_rate": 8.959439611343352e-06,
+      "loss": 1.8927,
+      "step": 921
+    },
+    {
+      "epoch": 1.0746355685131195,
+      "grad_norm": 0.08933998644351959,
+      "learning_rate": 8.95724949106537e-06,
+      "loss": 1.8669,
+      "step": 922
+    },
+    {
+      "epoch": 1.075801749271137,
+      "grad_norm": 0.10361660271883011,
+      "learning_rate": 8.955054907732368e-06,
+      "loss": 1.8871,
+      "step": 923
+    },
+    {
+      "epoch": 1.0769679300291546,
+      "grad_norm": 0.18225805461406708,
+      "learning_rate": 8.952855847688124e-06,
+      "loss": 1.9397,
+      "step": 924
+    },
+    {
+      "epoch": 1.078134110787172,
+      "grad_norm": 0.09028956294059753,
+      "learning_rate": 8.950652297220648e-06,
+      "loss": 1.9464,
+      "step": 925
+    },
+    {
+      "epoch": 1.0793002915451895,
+      "grad_norm": 0.11887269467115402,
+      "learning_rate": 8.948444242561889e-06,
+      "loss": 1.8708,
+      "step": 926
+    },
+    {
+      "epoch": 1.080466472303207,
+      "grad_norm": 0.2907440662384033,
+      "learning_rate": 8.946231669887462e-06,
+      "loss": 1.95,
+      "step": 927
+    },
+    {
+      "epoch": 1.0816326530612246,
+      "grad_norm": 0.1159723624587059,
+      "learning_rate": 8.94401456531634e-06,
+      "loss": 1.9037,
+      "step": 928
+    },
+    {
+      "epoch": 1.082798833819242,
+      "grad_norm": 0.10584396868944168,
+      "learning_rate": 8.941792914910582e-06,
+      "loss": 1.9312,
+      "step": 929
+    },
+    {
+      "epoch": 1.0839650145772595,
+      "grad_norm": 0.1732984483242035,
+      "learning_rate": 8.939566704675029e-06,
+      "loss": 1.8316,
+      "step": 930
+    },
+    {
+      "epoch": 1.085131195335277,
+      "grad_norm": 0.09687864035367966,
+      "learning_rate": 8.937335920557015e-06,
+      "loss": 1.8095,
+      "step": 931
+    },
+    {
+      "epoch": 1.0862973760932944,
+      "grad_norm": 0.09148871898651123,
+      "learning_rate": 8.93510054844607e-06,
+      "loss": 1.8964,
+      "step": 932
+    },
+    {
+      "epoch": 1.087463556851312,
+      "grad_norm": 0.09294304996728897,
+      "learning_rate": 8.932860574173626e-06,
+      "loss": 1.8525,
+      "step": 933
+    },
+    {
+      "epoch": 1.0886297376093295,
+      "grad_norm": 0.12912139296531677,
+      "learning_rate": 8.93061598351271e-06,
+      "loss": 1.8694,
+      "step": 934
+    },
+    {
+      "epoch": 1.089795918367347,
+      "grad_norm": 0.11350937932729721,
+      "learning_rate": 8.928366762177652e-06,
+      "loss": 1.765,
+      "step": 935
+    },
+    {
+      "epoch": 1.0909620991253643,
+      "grad_norm": 0.10226733982563019,
+      "learning_rate": 8.926112895823772e-06,
+      "loss": 1.8898,
+      "step": 936
+    },
+    {
+      "epoch": 1.092128279883382,
+      "grad_norm": 0.10338317602872849,
+      "learning_rate": 8.923854370047089e-06,
+      "loss": 1.8622,
+      "step": 937
+    },
+    {
+      "epoch": 1.0932944606413995,
+      "grad_norm": 0.1497410535812378,
+      "learning_rate": 8.921591170383998e-06,
+      "loss": 1.8306,
+      "step": 938
+    },
+    {
+      "epoch": 1.094460641399417,
+      "grad_norm": 0.08954465389251709,
+      "learning_rate": 8.919323282310968e-06,
+      "loss": 1.8683,
+      "step": 939
+    },
+    {
+      "epoch": 1.0956268221574343,
+      "grad_norm": 0.11858201026916504,
+      "learning_rate": 8.91705069124424e-06,
+      "loss": 1.9044,
+      "step": 940
+    },
+    {
+      "epoch": 1.0967930029154518,
+      "grad_norm": 0.11949458718299866,
+      "learning_rate": 8.9147733825395e-06,
+      "loss": 1.8587,
+      "step": 941
+    },
+    {
+      "epoch": 1.0979591836734695,
+      "grad_norm": 0.08819332718849182,
+      "learning_rate": 8.912491341491573e-06,
+      "loss": 1.804,
+      "step": 942
+    },
+    {
+      "epoch": 1.099125364431487,
+      "grad_norm": 0.10250313580036163,
+      "learning_rate": 8.910204553334104e-06,
+      "loss": 1.8768,
+      "step": 943
+    },
+    {
+      "epoch": 1.1002915451895043,
+      "grad_norm": 0.0847439244389534,
+      "learning_rate": 8.90791300323924e-06,
+      "loss": 1.7775,
+      "step": 944
+    },
+    {
+      "epoch": 1.1014577259475218,
+      "grad_norm": 0.13398543000221252,
+      "learning_rate": 8.905616676317313e-06,
+      "loss": 1.8602,
+      "step": 945
+    },
+    {
+      "epoch": 1.1026239067055394,
+      "grad_norm": 0.09716009348630905,
+      "learning_rate": 8.903315557616508e-06,
+      "loss": 1.8978,
+      "step": 946
+    },
+    {
+      "epoch": 1.1026239067055394,
+      "eval_loss": 0.9173402786254883,
+      "eval_runtime": 168.0531,
+      "eval_samples_per_second": 41.082,
+      "eval_steps_per_second": 2.571,
+      "step": 946
+    },
+    {
+      "epoch": 1.103790087463557,
+      "grad_norm": 0.2865077257156372,
+      "learning_rate": 8.90100963212255e-06,
+      "loss": 1.9193,
+      "step": 947
+    },
+    {
+      "epoch": 1.1049562682215743,
+      "grad_norm": 0.15445224940776825,
+      "learning_rate": 8.898698884758366e-06,
+      "loss": 1.8972,
+      "step": 948
+    },
+    {
+      "epoch": 1.1061224489795918,
+      "grad_norm": 0.09443886578083038,
+      "learning_rate": 8.896383300383765e-06,
+      "loss": 1.8607,
+      "step": 949
+    },
+    {
+      "epoch": 1.1072886297376092,
+      "grad_norm": 0.14032648503780365,
+      "learning_rate": 8.89406286379511e-06,
+      "loss": 1.8443,
+      "step": 950
+    },
+    {
+      "epoch": 1.1084548104956269,
+      "grad_norm": 0.09178216755390167,
+      "learning_rate": 8.891737559724975e-06,
+      "loss": 1.8267,
+      "step": 951
+    },
+    {
+      "epoch": 1.1096209912536443,
+      "grad_norm": 0.08012185990810394,
+      "learning_rate": 8.889407372841812e-06,
+      "loss": 1.7976,
+      "step": 952
+    },
+    {
+      "epoch": 1.1107871720116618,
+      "grad_norm": 0.08699242025613785,
+      "learning_rate": 8.887072287749622e-06,
+      "loss": 1.8455,
+      "step": 953
+    },
+    {
+      "epoch": 1.1119533527696792,
+      "grad_norm": 0.09190432727336884,
+      "learning_rate": 8.88473228898761e-06,
+      "loss": 1.8226,
+      "step": 954
+    },
+    {
+      "epoch": 1.1131195335276969,
+      "grad_norm": 0.08745045214891434,
+      "learning_rate": 8.882387361029841e-06,
+      "loss": 1.9095,
+      "step": 955
+    },
+    {
+      "epoch": 1.1142857142857143,
+      "grad_norm": 0.11255834996700287,
+      "learning_rate": 8.880037488284912e-06,
+      "loss": 1.8488,
+      "step": 956
+    },
+    {
+      "epoch": 1.1154518950437318,
+      "grad_norm": 0.12169434875249863,
+      "learning_rate": 8.87768265509558e-06,
+      "loss": 1.8318,
+      "step": 957
+    },
+    {
+      "epoch": 1.1166180758017492,
+      "grad_norm": 0.24658571183681488,
+      "learning_rate": 8.875322845738438e-06,
+      "loss": 1.8905,
+      "step": 958
+    },
+    {
+      "epoch": 1.1177842565597667,
+      "grad_norm": 0.09736736118793488,
+      "learning_rate": 8.87295804442355e-06,
+      "loss": 1.8449,
+      "step": 959
+    },
+    {
+      "epoch": 1.1189504373177843,
+      "grad_norm": 0.08878316730260849,
+      "learning_rate": 8.870588235294118e-06,
+      "loss": 1.8443,
+      "step": 960
+    },
+    {
+      "epoch": 1.1201166180758018,
+      "grad_norm": 0.12252040207386017,
+      "learning_rate": 8.868213402426098e-06,
+      "loss": 1.8815,
+      "step": 961
+    },
+    {
+      "epoch": 1.1212827988338192,
+      "grad_norm": 0.10524008423089981,
+      "learning_rate": 8.865833529827871e-06,
+      "loss": 1.8863,
+      "step": 962
+    },
+    {
+      "epoch": 1.1224489795918366,
+      "grad_norm": 0.1010618656873703,
+      "learning_rate": 8.863448601439868e-06,
+      "loss": 1.9565,
+      "step": 963
+    },
+    {
+      "epoch": 1.1236151603498543,
+      "grad_norm": 0.10591138899326324,
+      "learning_rate": 8.861058601134216e-06,
+      "loss": 1.8377,
+      "step": 964
+    },
+    {
+      "epoch": 1.1247813411078718,
+      "grad_norm": 0.09028156846761703,
+      "learning_rate": 8.85866351271437e-06,
+      "loss": 1.9247,
+      "step": 965
+    },
+    {
+      "epoch": 1.1259475218658892,
+      "grad_norm": 0.09739995002746582,
+      "learning_rate": 8.856263319914753e-06,
+      "loss": 1.8456,
+      "step": 966
+    },
+    {
+      "epoch": 1.1271137026239066,
+      "grad_norm": 0.1010865792632103,
+      "learning_rate": 8.85385800640038e-06,
+      "loss": 1.8923,
+      "step": 967
+    },
+    {
+      "epoch": 1.128279883381924,
+      "grad_norm": 0.12771500647068024,
+      "learning_rate": 8.851447555766494e-06,
+      "loss": 1.8628,
+      "step": 968
+    },
+    {
+      "epoch": 1.1294460641399418,
+      "grad_norm": 0.12462753057479858,
+      "learning_rate": 8.849031951538188e-06,
+      "loss": 1.8232,
+      "step": 969
+    },
+    {
+      "epoch": 1.1306122448979592,
+      "grad_norm": 0.09688686579465866,
+      "learning_rate": 8.846611177170037e-06,
+      "loss": 1.8441,
+      "step": 970
+    },
+    {
+      "epoch": 1.1317784256559766,
+      "grad_norm": 0.09346160292625427,
+      "learning_rate": 8.844185216045709e-06,
+      "loss": 1.8251,
+      "step": 971
+    },
+    {
+      "epoch": 1.132944606413994,
+      "grad_norm": 0.08626839518547058,
+      "learning_rate": 8.841754051477599e-06,
+      "loss": 1.8731,
+      "step": 972
+    },
+    {
+      "epoch": 1.1341107871720117,
+      "grad_norm": 0.27375560998916626,
+      "learning_rate": 8.839317666706431e-06,
+      "loss": 1.9603,
+      "step": 973
+    },
+    {
+      "epoch": 1.1352769679300292,
+      "grad_norm": 0.0970386490225792,
+      "learning_rate": 8.836876044900886e-06,
+      "loss": 1.8711,
+      "step": 974
+    },
+    {
+      "epoch": 1.1364431486880466,
+      "grad_norm": 0.09839459508657455,
+      "learning_rate": 8.834429169157203e-06,
+      "loss": 1.9599,
+      "step": 975
+    },
+    {
+      "epoch": 1.137609329446064,
+      "grad_norm": 0.10917026549577713,
+      "learning_rate": 8.831977022498803e-06,
+      "loss": 1.8557,
+      "step": 976
+    },
+    {
+      "epoch": 1.1387755102040815,
+      "grad_norm": 0.1299753189086914,
+      "learning_rate": 8.829519587875885e-06,
+      "loss": 1.8165,
+      "step": 977
+    },
+    {
+      "epoch": 1.1399416909620992,
+      "grad_norm": 0.09779125452041626,
+      "learning_rate": 8.827056848165027e-06,
+      "loss": 1.8158,
+      "step": 978
+    },
+    {
+      "epoch": 1.1411078717201166,
+      "grad_norm": 0.09313090145587921,
+      "learning_rate": 8.824588786168808e-06,
+      "loss": 1.8665,
+      "step": 979
+    },
+    {
+      "epoch": 1.142274052478134,
+      "grad_norm": 0.1226193755865097,
+      "learning_rate": 8.822115384615385e-06,
+      "loss": 1.856,
+      "step": 980
+    },
+    {
+      "epoch": 1.1434402332361515,
+      "grad_norm": 0.10279785096645355,
+      "learning_rate": 8.819636626158106e-06,
+      "loss": 1.9217,
+      "step": 981
+    },
+    {
+      "epoch": 1.1446064139941692,
+      "grad_norm": 0.11325406283140182,
+      "learning_rate": 8.81715249337509e-06,
+      "loss": 1.8613,
+      "step": 982
+    },
+    {
+      "epoch": 1.1457725947521866,
+      "grad_norm": 0.09295722842216492,
+      "learning_rate": 8.81466296876884e-06,
+      "loss": 1.818,
+      "step": 983
+    },
+    {
+      "epoch": 1.146938775510204,
+      "grad_norm": 0.10155995190143585,
+      "learning_rate": 8.812168034765814e-06,
+      "loss": 1.8999,
+      "step": 984
+    },
+    {
+      "epoch": 1.1481049562682215,
+      "grad_norm": 0.13603760302066803,
+      "learning_rate": 8.809667673716014e-06,
+      "loss": 1.8653,
+      "step": 985
+    },
+    {
+      "epoch": 1.149271137026239,
+      "grad_norm": 0.16047002375125885,
+      "learning_rate": 8.807161867892573e-06,
+      "loss": 1.9272,
+      "step": 986
+    },
+    {
+      "epoch": 1.1504373177842566,
+      "grad_norm": 0.13859263062477112,
+      "learning_rate": 8.804650599491342e-06,
+      "loss": 1.8657,
+      "step": 987
+    },
+    {
+      "epoch": 1.151603498542274,
+      "grad_norm": 0.10804110765457153,
+      "learning_rate": 8.802133850630456e-06,
+      "loss": 1.7594,
+      "step": 988
+    },
+    {
+      "epoch": 1.1527696793002915,
+      "grad_norm": 0.10662123560905457,
+      "learning_rate": 8.799611603349921e-06,
+      "loss": 1.9349,
+      "step": 989
+    },
+    {
+      "epoch": 1.153935860058309,
+      "grad_norm": 0.09498978406190872,
+      "learning_rate": 8.79708383961118e-06,
+      "loss": 1.9068,
+      "step": 990
+    },
+    {
+      "epoch": 1.1551020408163266,
+      "grad_norm": 0.0938570499420166,
+      "learning_rate": 8.79455054129668e-06,
+      "loss": 1.8357,
+      "step": 991
+    },
+    {
+      "epoch": 1.156268221574344,
+      "grad_norm": 0.09996771067380905,
+      "learning_rate": 8.792011690209449e-06,
+      "loss": 1.8978,
+      "step": 992
+    },
+    {
+      "epoch": 1.1574344023323615,
+      "grad_norm": 0.10613182187080383,
+      "learning_rate": 8.789467268072657e-06,
+      "loss": 1.8875,
+      "step": 993
+    },
+    {
+      "epoch": 1.158600583090379,
+      "grad_norm": 0.11555185914039612,
+      "learning_rate": 8.786917256529168e-06,
+      "loss": 1.8593,
+      "step": 994
+    },
+    {
+      "epoch": 1.1597667638483964,
+      "grad_norm": 0.09469382464885712,
+      "learning_rate": 8.784361637141112e-06,
+      "loss": 1.9177,
+      "step": 995
+    },
+    {
+      "epoch": 1.160932944606414,
+      "grad_norm": 0.08879994601011276,
+      "learning_rate": 8.781800391389433e-06,
+      "loss": 1.8471,
+      "step": 996
+    },
+    {
+      "epoch": 1.1620991253644315,
+      "grad_norm": 0.08643746376037598,
+      "learning_rate": 8.779233500673443e-06,
+      "loss": 1.9231,
+      "step": 997
+    },
+    {
+      "epoch": 1.163265306122449,
+      "grad_norm": 0.09634228050708771,
+      "learning_rate": 8.77666094631037e-06,
+      "loss": 1.7985,
+      "step": 998
+    },
+    {
+      "epoch": 1.1644314868804664,
+      "grad_norm": 0.14457504451274872,
+      "learning_rate": 8.774082709534913e-06,
+      "loss": 1.8705,
+      "step": 999
+    },
+    {
+      "epoch": 1.165597667638484,
+      "grad_norm": 0.0938292145729065,
+      "learning_rate": 8.771498771498772e-06,
+      "loss": 1.812,
+      "step": 1000
+    },
+    {
+      "epoch": 1.1667638483965015,
+      "grad_norm": 0.09335353225469589,
+      "learning_rate": 8.768909113270202e-06,
+      "loss": 1.8898,
+      "step": 1001
+    },
+    {
+      "epoch": 1.167930029154519,
+      "grad_norm": 0.09140753000974655,
+      "learning_rate": 8.766313715833537e-06,
+      "loss": 1.7599,
+      "step": 1002
+    },
+    {
+      "epoch": 1.1690962099125364,
+      "grad_norm": 0.0861637219786644,
+      "learning_rate": 8.763712560088747e-06,
+      "loss": 1.8432,
+      "step": 1003
+    },
+    {
+      "epoch": 1.1702623906705538,
+      "grad_norm": 0.37515968084335327,
+      "learning_rate": 8.761105626850937e-06,
+      "loss": 1.8489,
+      "step": 1004
+    },
+    {
+      "epoch": 1.1714285714285715,
+      "grad_norm": 0.08230669051408768,
+      "learning_rate": 8.758492896849907e-06,
+      "loss": 1.8738,
+      "step": 1005
+    },
+    {
+      "epoch": 1.172594752186589,
+      "grad_norm": 0.14773304760456085,
+      "learning_rate": 8.755874350729658e-06,
+      "loss": 1.8148,
+      "step": 1006
+    },
+    {
+      "epoch": 1.1737609329446064,
+      "grad_norm": 0.11925770342350006,
+      "learning_rate": 8.753249969047915e-06,
+      "loss": 1.8706,
+      "step": 1007
+    },
+    {
+      "epoch": 1.1749271137026238,
+      "grad_norm": 0.19538718461990356,
+      "learning_rate": 8.750619732275657e-06,
+      "loss": 1.878,
+      "step": 1008
+    },
+    {
+      "epoch": 1.1760932944606415,
+      "grad_norm": 0.12930434942245483,
+      "learning_rate": 8.747983620796625e-06,
+      "loss": 1.8791,
+      "step": 1009
+    },
+    {
+      "epoch": 1.177259475218659,
+      "grad_norm": 0.11389373242855072,
+      "learning_rate": 8.745341614906832e-06,
+      "loss": 1.8972,
+      "step": 1010
+    },
+    {
+      "epoch": 1.1784256559766764,
+      "grad_norm": 0.12832175195217133,
+      "learning_rate": 8.742693694814079e-06,
+      "loss": 1.8445,
+      "step": 1011
+    },
+    {
+      "epoch": 1.1795918367346938,
+      "grad_norm": 0.24789541959762573,
+      "learning_rate": 8.740039840637451e-06,
+      "loss": 1.8837,
+      "step": 1012
+    },
+    {
+      "epoch": 1.1807580174927113,
+      "grad_norm": 0.09982195496559143,
+      "learning_rate": 8.73738003240683e-06,
+      "loss": 1.8365,
+      "step": 1013
+    },
+    {
+      "epoch": 1.181924198250729,
+      "grad_norm": 0.08501381427049637,
+      "learning_rate": 8.734714250062391e-06,
+      "loss": 1.7954,
+      "step": 1014
+    },
+    {
+      "epoch": 1.1830903790087464,
+      "grad_norm": 0.16936218738555908,
+      "learning_rate": 8.73204247345409e-06,
+      "loss": 1.8833,
+      "step": 1015
+    },
+    {
+      "epoch": 1.1842565597667638,
+      "grad_norm": 0.10166364163160324,
+      "learning_rate": 8.729364682341172e-06,
+      "loss": 1.8448,
+      "step": 1016
+    },
+    {
+      "epoch": 1.1854227405247812,
+      "grad_norm": 0.12310101836919785,
+      "learning_rate": 8.726680856391637e-06,
+      "loss": 1.852,
+      "step": 1017
+    },
+    {
+      "epoch": 1.186588921282799,
+      "grad_norm": 0.10543253272771835,
+      "learning_rate": 8.723990975181749e-06,
+      "loss": 1.8413,
+      "step": 1018
+    },
+    {
+      "epoch": 1.1877551020408164,
+      "grad_norm": 0.2526230216026306,
+      "learning_rate": 8.721295018195506e-06,
+      "loss": 1.808,
+      "step": 1019
+    },
+    {
+      "epoch": 1.1889212827988338,
+      "grad_norm": 0.11322285979986191,
+      "learning_rate": 8.718592964824122e-06,
+      "loss": 1.8747,
+      "step": 1020
+    },
+    {
+      "epoch": 1.1900874635568512,
+      "grad_norm": 0.1550104320049286,
+      "learning_rate": 8.71588479436549e-06,
+      "loss": 1.8174,
+      "step": 1021
+    },
+    {
+      "epoch": 1.1912536443148687,
+      "grad_norm": 0.21468496322631836,
+      "learning_rate": 8.713170486023673e-06,
+      "loss": 1.8572,
+      "step": 1022
+    },
+    {
+      "epoch": 1.1924198250728864,
+      "grad_norm": 0.22103489935398102,
+      "learning_rate": 8.710450018908358e-06,
+      "loss": 1.8877,
+      "step": 1023
+    },
+    {
+      "epoch": 1.1935860058309038,
+      "grad_norm": 0.08816584199666977,
+      "learning_rate": 8.707723372034329e-06,
+      "loss": 1.8547,
+      "step": 1024
+    },
+    {
+      "epoch": 1.1947521865889212,
+      "grad_norm": 0.12131521105766296,
+      "learning_rate": 8.704990524320909e-06,
+      "loss": 1.8774,
+      "step": 1025
+    },
+    {
+      "epoch": 1.1959183673469387,
+      "grad_norm": 0.11561091989278793,
+      "learning_rate": 8.70225145459145e-06,
+      "loss": 1.917,
+      "step": 1026
+    },
+    {
+      "epoch": 1.1970845481049563,
+      "grad_norm": 0.16639551520347595,
+      "learning_rate": 8.69950614157275e-06,
+      "loss": 1.8332,
+      "step": 1027
+    },
+    {
+      "epoch": 1.1982507288629738,
+      "grad_norm": 0.12102624028921127,
+      "learning_rate": 8.696754563894524e-06,
+      "loss": 1.8796,
+      "step": 1028
+    },
+    {
+      "epoch": 1.1994169096209912,
+      "grad_norm": 0.09900449961423874,
+      "learning_rate": 8.693996700088844e-06,
+      "loss": 1.9125,
+      "step": 1029
+    },
+    {
+      "epoch": 1.2005830903790087,
+      "grad_norm": 0.14569807052612305,
+      "learning_rate": 8.69123252858958e-06,
+      "loss": 1.8894,
+      "step": 1030
+    },
+    {
+      "epoch": 1.2017492711370261,
+      "grad_norm": 0.10679837316274643,
+      "learning_rate": 8.688462027731841e-06,
+      "loss": 1.8702,
+      "step": 1031
+    },
+    {
+      "epoch": 1.2029154518950438,
+      "grad_norm": 0.2888908386230469,
+      "learning_rate": 8.685685175751401e-06,
+      "loss": 1.914,
+      "step": 1032
+    },
+    {
+      "epoch": 1.2029154518950438,
+      "eval_loss": 0.9160641431808472,
+      "eval_runtime": 168.4905,
+      "eval_samples_per_second": 40.976,
+      "eval_steps_per_second": 2.564,
+      "step": 1032
+    },
+    {
+      "epoch": 1.2040816326530612,
+      "grad_norm": 0.10004255920648575,
+      "learning_rate": 8.68290195078414e-06,
+      "loss": 1.8997,
+      "step": 1033
+    },
+    {
+      "epoch": 1.2052478134110787,
+      "grad_norm": 0.08940400183200836,
+      "learning_rate": 8.68011233086546e-06,
+      "loss": 1.8301,
+      "step": 1034
+    },
+    {
+      "epoch": 1.2064139941690961,
+      "grad_norm": 0.3191961348056793,
+      "learning_rate": 8.677316293929712e-06,
+      "loss": 1.8457,
+      "step": 1035
+    },
+    {
+      "epoch": 1.2075801749271138,
+      "grad_norm": 0.10709846764802933,
+      "learning_rate": 8.674513817809621e-06,
+      "loss": 1.8191,
+      "step": 1036
+    },
+    {
+      "epoch": 1.2087463556851312,
+      "grad_norm": 0.2598709464073181,
+      "learning_rate": 8.671704880235685e-06,
+      "loss": 1.9059,
+      "step": 1037
+    },
+    {
+      "epoch": 1.2099125364431487,
+      "grad_norm": 0.10091385245323181,
+      "learning_rate": 8.668889458835599e-06,
+      "loss": 1.8881,
+      "step": 1038
+    },
+    {
+      "epoch": 1.211078717201166,
+      "grad_norm": 0.10616692900657654,
+      "learning_rate": 8.66606753113365e-06,
+      "loss": 1.8343,
+      "step": 1039
+    },
+    {
+      "epoch": 1.2122448979591836,
+      "grad_norm": 0.09889233112335205,
+      "learning_rate": 8.66323907455013e-06,
+      "loss": 1.8037,
+      "step": 1040
+    },
+    {
+      "epoch": 1.2134110787172012,
+      "grad_norm": 0.13797344267368317,
+      "learning_rate": 8.66040406640072e-06,
+      "loss": 1.898,
+      "step": 1041
+    },
+    {
+      "epoch": 1.2145772594752187,
+      "grad_norm": 0.09788721799850464,
+      "learning_rate": 8.657562483895902e-06,
+      "loss": 1.8698,
+      "step": 1042
+    },
+    {
+      "epoch": 1.215743440233236,
+      "grad_norm": 0.08686728030443192,
+      "learning_rate": 8.654714304140334e-06,
+      "loss": 1.8745,
+      "step": 1043
+    },
+    {
+      "epoch": 1.2169096209912535,
+      "grad_norm": 0.09466682374477386,
+      "learning_rate": 8.651859504132233e-06,
+      "loss": 1.9517,
+      "step": 1044
+    },
+    {
+      "epoch": 1.2180758017492712,
+      "grad_norm": 0.08108825981616974,
+      "learning_rate": 8.648998060762768e-06,
+      "loss": 1.9018,
+      "step": 1045
+    },
+    {
+      "epoch": 1.2192419825072887,
+      "grad_norm": 0.12795335054397583,
+      "learning_rate": 8.64612995081543e-06,
+      "loss": 1.8801,
+      "step": 1046
+    },
+    {
+      "epoch": 1.220408163265306,
+      "grad_norm": 0.09538246691226959,
+      "learning_rate": 8.6432551509654e-06,
+      "loss": 1.8612,
+      "step": 1047
+    },
+    {
+      "epoch": 1.2215743440233235,
+      "grad_norm": 0.11823736876249313,
+      "learning_rate": 8.640373637778932e-06,
+      "loss": 1.8409,
+      "step": 1048
+    },
+    {
+      "epoch": 1.222740524781341,
+      "grad_norm": 0.10092376172542572,
+      "learning_rate": 8.637485387712691e-06,
+      "loss": 1.8257,
+      "step": 1049
+    },
+    {
+      "epoch": 1.2239067055393587,
+      "grad_norm": 0.14378590881824493,
+      "learning_rate": 8.634590377113135e-06,
+      "loss": 1.8054,
+      "step": 1050
+    },
+    {
+      "epoch": 1.225072886297376,
+      "grad_norm": 0.11370109021663666,
+      "learning_rate": 8.631688582215856e-06,
+      "loss": 1.8922,
+      "step": 1051
+    },
+    {
+      "epoch": 1.2262390670553935,
+      "grad_norm": 0.18617002665996552,
+      "learning_rate": 8.628779979144942e-06,
+      "loss": 1.7987,
+      "step": 1052
+    },
+    {
+      "epoch": 1.227405247813411,
+      "grad_norm": 0.13146650791168213,
+      "learning_rate": 8.625864543912305e-06,
+      "loss": 1.9013,
+      "step": 1053
+    },
+    {
+      "epoch": 1.2285714285714286,
+      "grad_norm": 0.11215505003929138,
+      "learning_rate": 8.622942252417036e-06,
+      "loss": 1.8811,
+      "step": 1054
+    },
+    {
+      "epoch": 1.229737609329446,
+      "grad_norm": 0.1303459107875824,
+      "learning_rate": 8.620013080444736e-06,
+      "loss": 1.8513,
+      "step": 1055
+    },
+    {
+      "epoch": 1.2309037900874635,
+      "grad_norm": 0.08530925959348679,
+      "learning_rate": 8.617077003666843e-06,
+      "loss": 1.8182,
+      "step": 1056
+    },
+    {
+      "epoch": 1.232069970845481,
+      "grad_norm": 0.12923352420330048,
+      "learning_rate": 8.614133997639964e-06,
+      "loss": 1.9055,
+      "step": 1057
+    },
+    {
+      "epoch": 1.2332361516034984,
+      "grad_norm": 0.0917331799864769,
+      "learning_rate": 8.611184037805197e-06,
+      "loss": 1.855,
+      "step": 1058
+    },
+    {
+      "epoch": 1.234402332361516,
+      "grad_norm": 0.08958417177200317,
+      "learning_rate": 8.608227099487449e-06,
+      "loss": 1.9477,
+      "step": 1059
+    },
+    {
+      "epoch": 1.2355685131195335,
+      "grad_norm": 0.11113128811120987,
+      "learning_rate": 8.605263157894738e-06,
+      "loss": 1.8663,
+      "step": 1060
+    },
+    {
+      "epoch": 1.236734693877551,
+      "grad_norm": 0.08974774926900864,
+      "learning_rate": 8.602292188117509e-06,
+      "loss": 1.8459,
+      "step": 1061
+    },
+    {
+      "epoch": 1.2379008746355686,
+      "grad_norm": 0.09684177488088608,
+      "learning_rate": 8.599314165127935e-06,
+      "loss": 1.8235,
+      "step": 1062
+    },
+    {
+      "epoch": 1.239067055393586,
+      "grad_norm": 0.09551951289176941,
+      "learning_rate": 8.596329063779218e-06,
+      "loss": 1.8208,
+      "step": 1063
+    },
+    {
+      "epoch": 1.2402332361516035,
+      "grad_norm": 0.10326727479696274,
+      "learning_rate": 8.593336858804865e-06,
+      "loss": 1.7836,
+      "step": 1064
+    },
+    {
+      "epoch": 1.241399416909621,
+      "grad_norm": 0.09809739887714386,
+      "learning_rate": 8.590337524818002e-06,
+      "loss": 1.8381,
+      "step": 1065
+    },
+    {
+      "epoch": 1.2425655976676384,
+      "grad_norm": 0.08467419445514679,
+      "learning_rate": 8.587331036310628e-06,
+      "loss": 1.8858,
+      "step": 1066
+    },
+    {
+      "epoch": 1.2437317784256559,
+      "grad_norm": 0.09340142458677292,
+      "learning_rate": 8.584317367652914e-06,
+      "loss": 1.8956,
+      "step": 1067
+    },
+    {
+      "epoch": 1.2448979591836735,
+      "grad_norm": 0.08902006596326828,
+      "learning_rate": 8.581296493092456e-06,
+      "loss": 1.9148,
+      "step": 1068
+    },
+    {
+      "epoch": 1.246064139941691,
+      "grad_norm": 0.08397535979747772,
+      "learning_rate": 8.578268386753558e-06,
+      "loss": 1.8728,
+      "step": 1069
+    },
+    {
+      "epoch": 1.2472303206997084,
+      "grad_norm": 0.097264364361763,
+      "learning_rate": 8.575233022636484e-06,
+      "loss": 1.9507,
+      "step": 1070
+    },
+    {
+      "epoch": 1.248396501457726,
+      "grad_norm": 0.11118023097515106,
+      "learning_rate": 8.572190374616718e-06,
+      "loss": 1.8275,
+      "step": 1071
+    },
+    {
+      "epoch": 1.2495626822157435,
+      "grad_norm": 0.10606839507818222,
+      "learning_rate": 8.569140416444208e-06,
+      "loss": 1.9541,
+      "step": 1072
+    },
+    {
+      "epoch": 1.250728862973761,
+      "grad_norm": 0.09160146117210388,
+      "learning_rate": 8.566083121742618e-06,
+      "loss": 1.8514,
+      "step": 1073
+    },
+    {
+      "epoch": 1.2518950437317784,
+      "grad_norm": 0.08492464572191238,
+      "learning_rate": 8.563018464008562e-06,
+      "loss": 1.7688,
+      "step": 1074
+    },
+    {
+      "epoch": 1.2530612244897958,
+      "grad_norm": 0.09300878643989563,
+      "learning_rate": 8.559946416610851e-06,
+      "loss": 1.8621,
+      "step": 1075
+    },
+    {
+      "epoch": 1.2542274052478133,
+      "grad_norm": 0.14066766202449799,
+      "learning_rate": 8.556866952789701e-06,
+      "loss": 1.914,
+      "step": 1076
+    },
+    {
+      "epoch": 1.255393586005831,
+      "grad_norm": 0.08517010509967804,
+      "learning_rate": 8.55378004565597e-06,
+      "loss": 1.7842,
+      "step": 1077
+    },
+    {
+      "epoch": 1.2565597667638484,
+      "grad_norm": 0.10736294090747833,
+      "learning_rate": 8.550685668190375e-06,
+      "loss": 1.8334,
+      "step": 1078
+    },
+    {
+      "epoch": 1.2577259475218658,
+      "grad_norm": 0.09330728650093079,
+      "learning_rate": 8.547583793242699e-06,
+      "loss": 1.7215,
+      "step": 1079
+    },
+    {
+      "epoch": 1.2588921282798835,
+      "grad_norm": 0.10977832227945328,
+      "learning_rate": 8.544474393530997e-06,
+      "loss": 1.9417,
+      "step": 1080
+    },
+    {
+      "epoch": 1.260058309037901,
+      "grad_norm": 0.09410124272108078,
+      "learning_rate": 8.541357441640803e-06,
+      "loss": 1.8296,
+      "step": 1081
+    },
+    {
+      "epoch": 1.2612244897959184,
+      "grad_norm": 0.09153234958648682,
+      "learning_rate": 8.53823291002432e-06,
+      "loss": 1.8748,
+      "step": 1082
+    },
+    {
+      "epoch": 1.2623906705539358,
+      "grad_norm": 0.19266973435878754,
+      "learning_rate": 8.535100770999596e-06,
+      "loss": 1.8368,
+      "step": 1083
+    },
+    {
+      "epoch": 1.2635568513119533,
+      "grad_norm": 0.09563439339399338,
+      "learning_rate": 8.531960996749729e-06,
+      "loss": 1.793,
+      "step": 1084
+    },
+    {
+      "epoch": 1.2647230320699707,
+      "grad_norm": 0.11086522787809372,
+      "learning_rate": 8.528813559322034e-06,
+      "loss": 1.9878,
+      "step": 1085
+    },
+    {
+      "epoch": 1.2658892128279884,
+      "grad_norm": 0.09368658065795898,
+      "learning_rate": 8.525658430627206e-06,
+      "loss": 1.8434,
+      "step": 1086
+    },
+    {
+      "epoch": 1.2670553935860058,
+      "grad_norm": 0.13267828524112701,
+      "learning_rate": 8.522495582438494e-06,
+      "loss": 1.7913,
+      "step": 1087
+    },
+    {
+      "epoch": 1.2682215743440233,
+      "grad_norm": 0.09418691694736481,
+      "learning_rate": 8.519324986390855e-06,
+      "loss": 1.877,
+      "step": 1088
+    },
+    {
+      "epoch": 1.269387755102041,
+      "grad_norm": 0.1031576469540596,
+      "learning_rate": 8.516146613980108e-06,
+      "loss": 1.8154,
+      "step": 1089
+    },
+    {
+      "epoch": 1.2705539358600584,
+      "grad_norm": 0.08370865136384964,
+      "learning_rate": 8.512960436562074e-06,
+      "loss": 1.8405,
+      "step": 1090
+    },
+    {
+      "epoch": 1.2717201166180758,
+      "grad_norm": 0.18562006950378418,
+      "learning_rate": 8.509766425351728e-06,
+      "loss": 1.8365,
+      "step": 1091
+    },
+    {
+      "epoch": 1.2728862973760933,
+      "grad_norm": 0.12915727496147156,
+      "learning_rate": 8.50656455142232e-06,
+      "loss": 1.8212,
+      "step": 1092
+    },
+    {
+      "epoch": 1.2740524781341107,
+      "grad_norm": 0.09559983015060425,
+      "learning_rate": 8.503354785704506e-06,
+      "loss": 1.7967,
+      "step": 1093
+    },
+    {
+      "epoch": 1.2752186588921282,
+      "grad_norm": 0.11479867994785309,
+      "learning_rate": 8.500137098985468e-06,
+      "loss": 1.8978,
+      "step": 1094
+    },
+    {
+      "epoch": 1.2763848396501458,
+      "grad_norm": 0.08780112862586975,
+      "learning_rate": 8.496911461908032e-06,
+      "loss": 1.9353,
+      "step": 1095
+    },
+    {
+      "epoch": 1.2775510204081633,
+      "grad_norm": 0.1146186888217926,
+      "learning_rate": 8.493677844969763e-06,
+      "loss": 1.9019,
+      "step": 1096
+    },
+    {
+      "epoch": 1.2787172011661807,
+      "grad_norm": 0.08856798708438873,
+      "learning_rate": 8.490436218522086e-06,
+      "loss": 1.9644,
+      "step": 1097
+    },
+    {
+      "epoch": 1.2798833819241984,
+      "grad_norm": 0.08960537612438202,
+      "learning_rate": 8.487186552769358e-06,
+      "loss": 1.8781,
+      "step": 1098
+    },
+    {
+      "epoch": 1.2810495626822158,
+      "grad_norm": 0.08653784543275833,
+      "learning_rate": 8.48392881776797e-06,
+      "loss": 1.7849,
+      "step": 1099
+    },
+    {
+      "epoch": 1.2822157434402333,
+      "grad_norm": 0.08975246548652649,
+      "learning_rate": 8.480662983425414e-06,
+      "loss": 1.8325,
+      "step": 1100
+    },
+    {
+      "epoch": 1.2833819241982507,
+      "grad_norm": 0.08603313565254211,
+      "learning_rate": 8.477389019499377e-06,
+      "loss": 1.8916,
+      "step": 1101
+    },
+    {
+      "epoch": 1.2845481049562681,
+      "grad_norm": 0.10388095676898956,
+      "learning_rate": 8.474106895596787e-06,
+      "loss": 1.9635,
+      "step": 1102
+    },
+    {
+      "epoch": 1.2857142857142856,
+      "grad_norm": 0.0999012216925621,
+      "learning_rate": 8.470816581172882e-06,
+      "loss": 1.8289,
+      "step": 1103
+    },
+    {
+      "epoch": 1.2868804664723033,
+      "grad_norm": 0.4995657503604889,
+      "learning_rate": 8.467518045530261e-06,
+      "loss": 1.8702,
+      "step": 1104
+    },
+    {
+      "epoch": 1.2880466472303207,
+      "grad_norm": 0.1177302747964859,
+      "learning_rate": 8.46421125781793e-06,
+      "loss": 1.8741,
+      "step": 1105
+    },
+    {
+      "epoch": 1.2892128279883381,
+      "grad_norm": 0.20680975914001465,
+      "learning_rate": 8.460896187030338e-06,
+      "loss": 1.8181,
+      "step": 1106
+    },
+    {
+      "epoch": 1.2903790087463558,
+      "grad_norm": 0.09337365627288818,
+      "learning_rate": 8.457572802006409e-06,
+      "loss": 1.8748,
+      "step": 1107
+    },
+    {
+      "epoch": 1.2915451895043732,
+      "grad_norm": 0.18588197231292725,
+      "learning_rate": 8.454241071428572e-06,
+      "loss": 1.8336,
+      "step": 1108
+    },
+    {
+      "epoch": 1.2927113702623907,
+      "grad_norm": 0.15842506289482117,
+      "learning_rate": 8.450900963821763e-06,
+      "loss": 1.9224,
+      "step": 1109
+    },
+    {
+      "epoch": 1.2938775510204081,
+      "grad_norm": 0.2347947508096695,
+      "learning_rate": 8.447552447552449e-06,
+      "loss": 1.7823,
+      "step": 1110
+    },
+    {
+      "epoch": 1.2950437317784256,
+      "grad_norm": 0.10756088048219681,
+      "learning_rate": 8.444195490827616e-06,
+      "loss": 1.8964,
+      "step": 1111
+    },
+    {
+      "epoch": 1.296209912536443,
+      "grad_norm": 0.0999930128455162,
+      "learning_rate": 8.440830061693776e-06,
+      "loss": 1.8308,
+      "step": 1112
+    },
+    {
+      "epoch": 1.2973760932944607,
+      "grad_norm": 0.14524096250534058,
+      "learning_rate": 8.43745612803594e-06,
+      "loss": 1.8321,
+      "step": 1113
+    },
+    {
+      "epoch": 1.2985422740524781,
+      "grad_norm": 0.15969781577587128,
+      "learning_rate": 8.43407365757661e-06,
+      "loss": 1.8322,
+      "step": 1114
+    },
+    {
+      "epoch": 1.2997084548104956,
+      "grad_norm": 0.12388742715120316,
+      "learning_rate": 8.430682617874737e-06,
+      "loss": 1.8004,
+      "step": 1115
+    },
+    {
+      "epoch": 1.3008746355685132,
+      "grad_norm": 0.13578063249588013,
+      "learning_rate": 8.42728297632469e-06,
+      "loss": 1.8686,
+      "step": 1116
+    },
+    {
+      "epoch": 1.3020408163265307,
+      "grad_norm": 0.1148504987359047,
+      "learning_rate": 8.423874700155213e-06,
+      "loss": 1.926,
+      "step": 1117
+    },
+    {
+      "epoch": 1.3032069970845481,
+      "grad_norm": 0.12558364868164062,
+      "learning_rate": 8.42045775642837e-06,
+      "loss": 1.9896,
+      "step": 1118
+    },
+    {
+      "epoch": 1.3032069970845481,
+      "eval_loss": 0.9145452380180359,
+      "eval_runtime": 168.8622,
+      "eval_samples_per_second": 40.885,
+      "eval_steps_per_second": 2.558,
+      "step": 1118
+    },
+    {
+      "epoch": 1.3043731778425656,
+      "grad_norm": 0.09700833261013031,
+      "learning_rate": 8.41703211203848e-06,
+      "loss": 1.8542,
+      "step": 1119
+    },
+    {
+      "epoch": 1.305539358600583,
+      "grad_norm": 0.13098494708538055,
+      "learning_rate": 8.413597733711049e-06,
+      "loss": 1.7873,
+      "step": 1120
+    },
+    {
+      "epoch": 1.3067055393586005,
+      "grad_norm": 0.08546532690525055,
+      "learning_rate": 8.410154588001702e-06,
+      "loss": 1.8343,
+      "step": 1121
+    },
+    {
+      "epoch": 1.3078717201166181,
+      "grad_norm": 0.09071295708417892,
+      "learning_rate": 8.406702641295088e-06,
+      "loss": 1.8121,
+      "step": 1122
+    },
+    {
+      "epoch": 1.3090379008746356,
+      "grad_norm": 0.1094484031200409,
+      "learning_rate": 8.403241859803782e-06,
+      "loss": 1.8394,
+      "step": 1123
+    },
+    {
+      "epoch": 1.310204081632653,
+      "grad_norm": 0.09670306742191315,
+      "learning_rate": 8.399772209567199e-06,
+      "loss": 1.8173,
+      "step": 1124
+    },
+    {
+      "epoch": 1.3113702623906707,
+      "grad_norm": 0.09205988794565201,
+      "learning_rate": 8.396293656450463e-06,
+      "loss": 1.8367,
+      "step": 1125
+    },
+    {
+      "epoch": 1.3125364431486881,
+      "grad_norm": 0.11549568176269531,
+      "learning_rate": 8.392806166143307e-06,
+      "loss": 1.964,
+      "step": 1126
+    },
+    {
+      "epoch": 1.3137026239067056,
+      "grad_norm": 0.11309079080820084,
+      "learning_rate": 8.389309704158924e-06,
+      "loss": 1.9342,
+      "step": 1127
+    },
+    {
+      "epoch": 1.314868804664723,
+      "grad_norm": 0.09797142446041107,
+      "learning_rate": 8.385804235832857e-06,
+      "loss": 1.7716,
+      "step": 1128
+    },
+    {
+      "epoch": 1.3160349854227404,
+      "grad_norm": 0.11019336432218552,
+      "learning_rate": 8.382289726321822e-06,
+      "loss": 1.8852,
+      "step": 1129
+    },
+    {
+      "epoch": 1.3172011661807579,
+      "grad_norm": 0.10901153832674026,
+      "learning_rate": 8.378766140602582e-06,
+      "loss": 1.9073,
+      "step": 1130
+    },
+    {
+      "epoch": 1.3183673469387756,
+      "grad_norm": 0.09650638699531555,
+      "learning_rate": 8.375233443470767e-06,
+      "loss": 1.8875,
+      "step": 1131
+    },
+    {
+      "epoch": 1.319533527696793,
+      "grad_norm": 0.08225961029529572,
+      "learning_rate": 8.371691599539702e-06,
+      "loss": 1.8396,
+      "step": 1132
+    },
+    {
+      "epoch": 1.3206997084548104,
+      "grad_norm": 0.11363028734922409,
+      "learning_rate": 8.368140573239234e-06,
+      "loss": 1.7975,
+      "step": 1133
+    },
+    {
+      "epoch": 1.321865889212828,
+      "grad_norm": 0.09057224541902542,
+      "learning_rate": 8.364580328814536e-06,
+      "loss": 1.8362,
+      "step": 1134
+    },
+    {
+      "epoch": 1.3230320699708455,
+      "grad_norm": 0.10524657368659973,
+      "learning_rate": 8.36101083032491e-06,
+      "loss": 1.8486,
+      "step": 1135
+    },
+    {
+      "epoch": 1.324198250728863,
+      "grad_norm": 0.18719112873077393,
+      "learning_rate": 8.357432041642569e-06,
+      "loss": 1.8873,
+      "step": 1136
+    },
+    {
+      "epoch": 1.3253644314868804,
+      "grad_norm": 0.09170349687337875,
+      "learning_rate": 8.353843926451428e-06,
+      "loss": 1.8808,
+      "step": 1137
+    },
+    {
+      "epoch": 1.3265306122448979,
+      "grad_norm": 0.10101398825645447,
+      "learning_rate": 8.350246448245869e-06,
+      "loss": 1.8793,
+      "step": 1138
+    },
+    {
+      "epoch": 1.3276967930029153,
+      "grad_norm": 0.1014290377497673,
+      "learning_rate": 8.34663957032951e-06,
+      "loss": 1.8858,
+      "step": 1139
+    },
+    {
+      "epoch": 1.328862973760933,
+      "grad_norm": 0.08973164111375809,
+      "learning_rate": 8.343023255813953e-06,
+      "loss": 1.8454,
+      "step": 1140
+    },
+    {
+      "epoch": 1.3300291545189504,
+      "grad_norm": 0.08541316539049149,
+      "learning_rate": 8.339397467617524e-06,
+      "loss": 1.864,
+      "step": 1141
+    },
+    {
+      "epoch": 1.3311953352769679,
+      "grad_norm": 0.09410756081342697,
+      "learning_rate": 8.335762168464006e-06,
+      "loss": 1.8863,
+      "step": 1142
+    },
+    {
+      "epoch": 1.3323615160349855,
+      "grad_norm": 0.08956369757652283,
+      "learning_rate": 8.332117320881365e-06,
+      "loss": 1.8902,
+      "step": 1143
+    },
+    {
+      "epoch": 1.333527696793003,
+      "grad_norm": 0.09305954724550247,
+      "learning_rate": 8.328462887200467e-06,
+      "loss": 1.9742,
+      "step": 1144
+    },
+    {
+      "epoch": 1.3346938775510204,
+      "grad_norm": 0.105361707508564,
+      "learning_rate": 8.324798829553768e-06,
+      "loss": 1.8789,
+      "step": 1145
+    },
+    {
+      "epoch": 1.3358600583090379,
+      "grad_norm": 0.096257284283638,
+      "learning_rate": 8.321125109874011e-06,
+      "loss": 1.8436,
+      "step": 1146
+    },
+    {
+      "epoch": 1.3370262390670553,
+      "grad_norm": 0.10395840555429459,
+      "learning_rate": 8.317441689892916e-06,
+      "loss": 1.8944,
+      "step": 1147
+    },
+    {
+      "epoch": 1.3381924198250728,
+      "grad_norm": 0.10116198658943176,
+      "learning_rate": 8.313748531139837e-06,
+      "loss": 1.8703,
+      "step": 1148
+    },
+    {
+      "epoch": 1.3393586005830904,
+      "grad_norm": 0.09636560082435608,
+      "learning_rate": 8.310045594940433e-06,
+      "loss": 1.838,
+      "step": 1149
+    },
+    {
+      "epoch": 1.3405247813411079,
+      "grad_norm": 0.1379690170288086,
+      "learning_rate": 8.306332842415316e-06,
+      "loss": 1.8711,
+      "step": 1150
+    },
+    {
+      "epoch": 1.3416909620991253,
+      "grad_norm": 0.11059990525245667,
+      "learning_rate": 8.302610234478691e-06,
+      "loss": 1.8046,
+      "step": 1151
+    },
+    {
+      "epoch": 1.342857142857143,
+      "grad_norm": 0.08817456662654877,
+      "learning_rate": 8.298877731836976e-06,
+      "loss": 1.8618,
+      "step": 1152
+    },
+    {
+      "epoch": 1.3440233236151604,
+      "grad_norm": 0.08623939007520676,
+      "learning_rate": 8.295135294987432e-06,
+      "loss": 1.806,
+      "step": 1153
+    },
+    {
+      "epoch": 1.3451895043731779,
+      "grad_norm": 0.38700178265571594,
+      "learning_rate": 8.291382884216761e-06,
+      "loss": 1.9743,
+      "step": 1154
+    },
+    {
+      "epoch": 1.3463556851311953,
+      "grad_norm": 0.08418189734220505,
+      "learning_rate": 8.287620459599704e-06,
+      "loss": 1.8091,
+      "step": 1155
+    },
+    {
+      "epoch": 1.3475218658892127,
+      "grad_norm": 0.09361667186021805,
+      "learning_rate": 8.283847980997625e-06,
+      "loss": 1.8669,
+      "step": 1156
+    },
+    {
+      "epoch": 1.3486880466472302,
+      "grad_norm": 0.20830000936985016,
+      "learning_rate": 8.280065408057084e-06,
+      "loss": 1.8144,
+      "step": 1157
+    },
+    {
+      "epoch": 1.3498542274052479,
+      "grad_norm": 0.09171152859926224,
+      "learning_rate": 8.276272700208396e-06,
+      "loss": 1.9008,
+      "step": 1158
+    },
+    {
+      "epoch": 1.3510204081632653,
+      "grad_norm": 0.08294904232025146,
+      "learning_rate": 8.272469816664183e-06,
+      "loss": 1.8607,
+      "step": 1159
+    },
+    {
+      "epoch": 1.3521865889212827,
+      "grad_norm": 0.09728697687387466,
+      "learning_rate": 8.26865671641791e-06,
+      "loss": 1.7776,
+      "step": 1160
+    },
+    {
+      "epoch": 1.3533527696793004,
+      "grad_norm": 0.10881102085113525,
+      "learning_rate": 8.264833358242417e-06,
+      "loss": 1.8799,
+      "step": 1161
+    },
+    {
+      "epoch": 1.3545189504373178,
+      "grad_norm": 0.09119808673858643,
+      "learning_rate": 8.260999700688415e-06,
+      "loss": 1.8182,
+      "step": 1162
+    },
+    {
+      "epoch": 1.3556851311953353,
+      "grad_norm": 0.12684091925621033,
+      "learning_rate": 8.257155702083021e-06,
+      "loss": 1.8295,
+      "step": 1163
+    },
+    {
+      "epoch": 1.3568513119533527,
+      "grad_norm": 0.08546722680330276,
+      "learning_rate": 8.253301320528211e-06,
+      "loss": 1.8881,
+      "step": 1164
+    },
+    {
+      "epoch": 1.3580174927113702,
+      "grad_norm": 0.08860180526971817,
+      "learning_rate": 8.249436513899325e-06,
+      "loss": 1.7868,
+      "step": 1165
+    },
+    {
+      "epoch": 1.3591836734693876,
+      "grad_norm": 0.10230910778045654,
+      "learning_rate": 8.245561239843514e-06,
+      "loss": 1.7497,
+      "step": 1166
+    },
+    {
+      "epoch": 1.3603498542274053,
+      "grad_norm": 0.09236747026443481,
+      "learning_rate": 8.241675455778213e-06,
+      "loss": 1.9319,
+      "step": 1167
+    },
+    {
+      "epoch": 1.3615160349854227,
+      "grad_norm": 0.0949830412864685,
+      "learning_rate": 8.23777911888956e-06,
+      "loss": 1.8544,
+      "step": 1168
+    },
+    {
+      "epoch": 1.3626822157434402,
+      "grad_norm": 0.5018076300621033,
+      "learning_rate": 8.233872186130838e-06,
+      "loss": 1.8216,
+      "step": 1169
+    },
+    {
+      "epoch": 1.3638483965014578,
+      "grad_norm": 0.11666015535593033,
+      "learning_rate": 8.229954614220878e-06,
+      "loss": 1.8865,
+      "step": 1170
+    },
+    {
+      "epoch": 1.3650145772594753,
+      "grad_norm": 0.0846833810210228,
+      "learning_rate": 8.22602635964248e-06,
+      "loss": 1.914,
+      "step": 1171
+    },
+    {
+      "epoch": 1.3661807580174927,
+      "grad_norm": 0.13993310928344727,
+      "learning_rate": 8.222087378640776e-06,
+      "loss": 1.7886,
+      "step": 1172
+    },
+    {
+      "epoch": 1.3673469387755102,
+      "grad_norm": 0.08744478225708008,
+      "learning_rate": 8.218137627221632e-06,
+      "loss": 1.7672,
+      "step": 1173
+    },
+    {
+      "epoch": 1.3685131195335276,
+      "grad_norm": 0.08427275717258453,
+      "learning_rate": 8.214177061149986e-06,
+      "loss": 1.9099,
+      "step": 1174
+    },
+    {
+      "epoch": 1.369679300291545,
+      "grad_norm": 0.1967243105173111,
+      "learning_rate": 8.21020563594821e-06,
+      "loss": 1.7729,
+      "step": 1175
+    },
+    {
+      "epoch": 1.3708454810495627,
+      "grad_norm": 0.09016139805316925,
+      "learning_rate": 8.206223306894447e-06,
+      "loss": 1.823,
+      "step": 1176
+    },
+    {
+      "epoch": 1.3720116618075802,
+      "grad_norm": 0.09787224978208542,
+      "learning_rate": 8.202230029020926e-06,
+      "loss": 1.7971,
+      "step": 1177
+    },
+    {
+      "epoch": 1.3731778425655976,
+      "grad_norm": 0.14146792888641357,
+      "learning_rate": 8.198225757112266e-06,
+      "loss": 1.8508,
+      "step": 1178
+    },
+    {
+      "epoch": 1.3743440233236153,
+      "grad_norm": 0.11487364768981934,
+      "learning_rate": 8.194210445703783e-06,
+      "loss": 1.8516,
+      "step": 1179
+    },
+    {
+      "epoch": 1.3755102040816327,
+      "grad_norm": 0.16582484543323517,
+      "learning_rate": 8.190184049079755e-06,
+      "loss": 1.8798,
+      "step": 1180
+    },
+    {
+      "epoch": 1.3766763848396502,
+      "grad_norm": 0.10427810996770859,
+      "learning_rate": 8.186146521271694e-06,
+      "loss": 1.8556,
+      "step": 1181
+    },
+    {
+      "epoch": 1.3778425655976676,
+      "grad_norm": 0.10396955907344818,
+      "learning_rate": 8.182097816056596e-06,
+      "loss": 1.8798,
+      "step": 1182
+    },
+    {
+      "epoch": 1.379008746355685,
+      "grad_norm": 0.09509523957967758,
+      "learning_rate": 8.178037886955182e-06,
+      "loss": 1.8198,
+      "step": 1183
+    },
+    {
+      "epoch": 1.3801749271137025,
+      "grad_norm": 0.09663504362106323,
+      "learning_rate": 8.173966687230104e-06,
+      "loss": 1.8595,
+      "step": 1184
+    },
+    {
+      "epoch": 1.3813411078717202,
+      "grad_norm": 0.10184940695762634,
+      "learning_rate": 8.16988416988417e-06,
+      "loss": 1.8793,
+      "step": 1185
+    },
+    {
+      "epoch": 1.3825072886297376,
+      "grad_norm": 0.0961347445845604,
+      "learning_rate": 8.165790287658521e-06,
+      "loss": 1.8676,
+      "step": 1186
+    },
+    {
+      "epoch": 1.383673469387755,
+      "grad_norm": 0.10227195173501968,
+      "learning_rate": 8.16168499303082e-06,
+      "loss": 1.8214,
+      "step": 1187
+    },
+    {
+      "epoch": 1.3848396501457727,
+      "grad_norm": 0.10094721615314484,
+      "learning_rate": 8.157568238213402e-06,
+      "loss": 1.8997,
+      "step": 1188
+    },
+    {
+      "epoch": 1.3860058309037901,
+      "grad_norm": 0.11897652596235275,
+      "learning_rate": 8.153439975151421e-06,
+      "loss": 1.9266,
+      "step": 1189
+    },
+    {
+      "epoch": 1.3871720116618076,
+      "grad_norm": 0.0913107842206955,
+      "learning_rate": 8.149300155520995e-06,
+      "loss": 1.9323,
+      "step": 1190
+    },
+    {
+      "epoch": 1.388338192419825,
+      "grad_norm": 0.12803906202316284,
+      "learning_rate": 8.145148730727301e-06,
+      "loss": 1.9876,
+      "step": 1191
+    },
+    {
+      "epoch": 1.3895043731778425,
+      "grad_norm": 0.09005334973335266,
+      "learning_rate": 8.140985651902685e-06,
+      "loss": 1.9289,
+      "step": 1192
+    },
+    {
+      "epoch": 1.39067055393586,
+      "grad_norm": 0.09677088260650635,
+      "learning_rate": 8.136810869904733e-06,
+      "loss": 1.7966,
+      "step": 1193
+    },
+    {
+      "epoch": 1.3918367346938776,
+      "grad_norm": 0.21331501007080078,
+      "learning_rate": 8.132624335314358e-06,
+      "loss": 1.9196,
+      "step": 1194
+    },
+    {
+      "epoch": 1.393002915451895,
+      "grad_norm": 0.12028966844081879,
+      "learning_rate": 8.128425998433829e-06,
+      "loss": 1.8975,
+      "step": 1195
+    },
+    {
+      "epoch": 1.3941690962099125,
+      "grad_norm": 0.09530478715896606,
+      "learning_rate": 8.12421580928482e-06,
+      "loss": 1.945,
+      "step": 1196
+    },
+    {
+      "epoch": 1.3953352769679301,
+      "grad_norm": 0.10388198494911194,
+      "learning_rate": 8.119993717606409e-06,
+      "loss": 1.7709,
+      "step": 1197
+    },
+    {
+      "epoch": 1.3965014577259476,
+      "grad_norm": 0.08599631488323212,
+      "learning_rate": 8.115759672853099e-06,
+      "loss": 1.8355,
+      "step": 1198
+    },
+    {
+      "epoch": 1.397667638483965,
+      "grad_norm": 0.09817085415124893,
+      "learning_rate": 8.111513624192786e-06,
+      "loss": 1.8494,
+      "step": 1199
+    },
+    {
+      "epoch": 1.3988338192419825,
+      "grad_norm": 0.15047875046730042,
+      "learning_rate": 8.107255520504732e-06,
+      "loss": 1.7907,
+      "step": 1200
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.09506538510322571,
+      "learning_rate": 8.102985310377508e-06,
+      "loss": 1.8241,
+      "step": 1201
+    },
+    {
+      "epoch": 1.4011661807580174,
+      "grad_norm": 0.09735575318336487,
+      "learning_rate": 8.09870294210693e-06,
+      "loss": 1.8779,
+      "step": 1202
+    },
+    {
+      "epoch": 1.402332361516035,
+      "grad_norm": 0.11007897555828094,
+      "learning_rate": 8.094408363693966e-06,
+      "loss": 1.9118,
+      "step": 1203
+    },
+    {
+      "epoch": 1.4034985422740525,
+      "grad_norm": 0.10273466259241104,
+      "learning_rate": 8.09010152284264e-06,
+      "loss": 1.8732,
+      "step": 1204
+    },
+    {
+      "epoch": 1.4034985422740525,
+      "eval_loss": 0.9134606122970581,
+      "eval_runtime": 168.8095,
+      "eval_samples_per_second": 40.898,
+      "eval_steps_per_second": 2.559,
+      "step": 1204
+    },
+    {
+      "epoch": 1.40466472303207,
+      "grad_norm": 0.10155772417783737,
+      "learning_rate": 8.085782366957902e-06,
+      "loss": 1.7712,
+      "step": 1205
+    },
+    {
+      "epoch": 1.4058309037900876,
+      "grad_norm": 0.08998365700244904,
+      "learning_rate": 8.081450843143494e-06,
+      "loss": 1.901,
+      "step": 1206
+    },
+    {
+      "epoch": 1.406997084548105,
+      "grad_norm": 0.105516217648983,
+      "learning_rate": 8.077106898199777e-06,
+      "loss": 1.9235,
+      "step": 1207
+    },
+    {
+      "epoch": 1.4081632653061225,
+      "grad_norm": 0.10353604704141617,
+      "learning_rate": 8.07275047862157e-06,
+      "loss": 1.8113,
+      "step": 1208
+    },
+    {
+      "epoch": 1.40932944606414,
+      "grad_norm": 0.09290672838687897,
+      "learning_rate": 8.068381530595941e-06,
+      "loss": 1.8835,
+      "step": 1209
+    },
+    {
+      "epoch": 1.4104956268221573,
+      "grad_norm": 0.10607472062110901,
+      "learning_rate": 8.064e-06,
+      "loss": 1.8173,
+      "step": 1210
+    },
+    {
+      "epoch": 1.411661807580175,
+      "grad_norm": 0.10617832839488983,
+      "learning_rate": 8.059605832398654e-06,
+      "loss": 1.8249,
+      "step": 1211
+    },
+    {
+      "epoch": 1.4128279883381925,
+      "grad_norm": 0.0866347998380661,
+      "learning_rate": 8.055198973042363e-06,
+      "loss": 1.9098,
+      "step": 1212
+    },
+    {
+      "epoch": 1.41399416909621,
+      "grad_norm": 0.09541983157396317,
+      "learning_rate": 8.050779366864857e-06,
+      "loss": 1.813,
+      "step": 1213
+    },
+    {
+      "epoch": 1.4151603498542273,
+      "grad_norm": 0.134172260761261,
+      "learning_rate": 8.04634695848085e-06,
+      "loss": 1.8206,
+      "step": 1214
+    },
+    {
+      "epoch": 1.416326530612245,
+      "grad_norm": 0.22920455038547516,
+      "learning_rate": 8.041901692183723e-06,
+      "loss": 1.8976,
+      "step": 1215
+    },
+    {
+      "epoch": 1.4174927113702624,
+      "grad_norm": 0.12872587144374847,
+      "learning_rate": 8.03744351194319e-06,
+      "loss": 1.7609,
+      "step": 1216
+    },
+    {
+      "epoch": 1.41865889212828,
+      "grad_norm": 0.1302987039089203,
+      "learning_rate": 8.032972361402942e-06,
+      "loss": 1.8417,
+      "step": 1217
+    },
+    {
+      "epoch": 1.4198250728862973,
+      "grad_norm": 0.11671137809753418,
+      "learning_rate": 8.02848818387828e-06,
+      "loss": 1.8381,
+      "step": 1218
+    },
+    {
+      "epoch": 1.4209912536443148,
+      "grad_norm": 0.08939551562070847,
+      "learning_rate": 8.023990922353705e-06,
+      "loss": 1.9038,
+      "step": 1219
+    },
+    {
+      "epoch": 1.4221574344023324,
+      "grad_norm": 0.28153640031814575,
+      "learning_rate": 8.019480519480521e-06,
+      "loss": 1.8674,
+      "step": 1220
+    },
+    {
+      "epoch": 1.4233236151603499,
+      "grad_norm": 0.12999235093593597,
+      "learning_rate": 8.014956917574376e-06,
+      "loss": 1.8832,
+      "step": 1221
+    },
+    {
+      "epoch": 1.4244897959183673,
+      "grad_norm": 0.09469537436962128,
+      "learning_rate": 8.01042005861283e-06,
+      "loss": 1.8952,
+      "step": 1222
+    },
+    {
+      "epoch": 1.4256559766763848,
+      "grad_norm": 0.09911829233169556,
+      "learning_rate": 8.005869884232838e-06,
+      "loss": 1.8845,
+      "step": 1223
+    },
+    {
+      "epoch": 1.4268221574344024,
+      "grad_norm": 0.08617237955331802,
+      "learning_rate": 8.001306335728282e-06,
+      "loss": 1.7534,
+      "step": 1224
+    },
+    {
+      "epoch": 1.4279883381924199,
+      "grad_norm": 0.09149475395679474,
+      "learning_rate": 7.996729354047424e-06,
+      "loss": 1.7428,
+      "step": 1225
+    },
+    {
+      "epoch": 1.4291545189504373,
+      "grad_norm": 0.08364789187908173,
+      "learning_rate": 7.99213887979037e-06,
+      "loss": 1.885,
+      "step": 1226
+    },
+    {
+      "epoch": 1.4303206997084548,
+      "grad_norm": 0.1523764580488205,
+      "learning_rate": 7.987534853206495e-06,
+      "loss": 1.8503,
+      "step": 1227
+    },
+    {
+      "epoch": 1.4314868804664722,
+      "grad_norm": 0.11572586745023727,
+      "learning_rate": 7.982917214191853e-06,
+      "loss": 1.8064,
+      "step": 1228
+    },
+    {
+      "epoch": 1.4326530612244899,
+      "grad_norm": 0.11969052255153656,
+      "learning_rate": 7.978285902286562e-06,
+      "loss": 1.8839,
+      "step": 1229
+    },
+    {
+      "epoch": 1.4338192419825073,
+      "grad_norm": 0.09760216623544693,
+      "learning_rate": 7.973640856672159e-06,
+      "loss": 1.8021,
+      "step": 1230
+    },
+    {
+      "epoch": 1.4349854227405248,
+      "grad_norm": 0.09661209583282471,
+      "learning_rate": 7.96898201616895e-06,
+      "loss": 1.8833,
+      "step": 1231
+    },
+    {
+      "epoch": 1.4361516034985422,
+      "grad_norm": 0.08976302295923233,
+      "learning_rate": 7.964309319233312e-06,
+      "loss": 1.8478,
+      "step": 1232
+    },
+    {
+      "epoch": 1.4373177842565599,
+      "grad_norm": 0.14566168189048767,
+      "learning_rate": 7.959622703954989e-06,
+      "loss": 1.8545,
+      "step": 1233
+    },
+    {
+      "epoch": 1.4384839650145773,
+      "grad_norm": 0.0864613950252533,
+      "learning_rate": 7.95492210805436e-06,
+      "loss": 1.8473,
+      "step": 1234
+    },
+    {
+      "epoch": 1.4396501457725948,
+      "grad_norm": 0.09550000727176666,
+      "learning_rate": 7.95020746887967e-06,
+      "loss": 1.8183,
+      "step": 1235
+    },
+    {
+      "epoch": 1.4408163265306122,
+      "grad_norm": 0.11403138190507889,
+      "learning_rate": 7.945478723404257e-06,
+      "loss": 1.8279,
+      "step": 1236
+    },
+    {
+      "epoch": 1.4419825072886296,
+      "grad_norm": 0.0981639176607132,
+      "learning_rate": 7.94073580822374e-06,
+      "loss": 1.8207,
+      "step": 1237
+    },
+    {
+      "epoch": 1.4431486880466473,
+      "grad_norm": 0.08320952951908112,
+      "learning_rate": 7.935978659553183e-06,
+      "loss": 1.8695,
+      "step": 1238
+    },
+    {
+      "epoch": 1.4443148688046648,
+      "grad_norm": 0.09434887766838074,
+      "learning_rate": 7.931207213224245e-06,
+      "loss": 1.8748,
+      "step": 1239
+    },
+    {
+      "epoch": 1.4454810495626822,
+      "grad_norm": 0.09806468337774277,
+      "learning_rate": 7.926421404682276e-06,
+      "loss": 1.8831,
+      "step": 1240
+    },
+    {
+      "epoch": 1.4466472303206996,
+      "grad_norm": 0.10040673613548279,
+      "learning_rate": 7.92162116898342e-06,
+      "loss": 1.8029,
+      "step": 1241
+    },
+    {
+      "epoch": 1.4478134110787173,
+      "grad_norm": 0.2568235397338867,
+      "learning_rate": 7.91680644079168e-06,
+      "loss": 1.815,
+      "step": 1242
+    },
+    {
+      "epoch": 1.4489795918367347,
+      "grad_norm": 0.10759842395782471,
+      "learning_rate": 7.911977154375944e-06,
+      "loss": 1.8868,
+      "step": 1243
+    },
+    {
+      "epoch": 1.4501457725947522,
+      "grad_norm": 0.11564269661903381,
+      "learning_rate": 7.907133243606999e-06,
+      "loss": 1.9226,
+      "step": 1244
+    },
+    {
+      "epoch": 1.4513119533527696,
+      "grad_norm": 0.09828026592731476,
+      "learning_rate": 7.902274641954507e-06,
+      "loss": 1.879,
+      "step": 1245
+    },
+    {
+      "epoch": 1.452478134110787,
+      "grad_norm": 0.10777269303798676,
+      "learning_rate": 7.89740128248397e-06,
+      "loss": 1.9128,
+      "step": 1246
+    },
+    {
+      "epoch": 1.4536443148688047,
+      "grad_norm": 0.08794382214546204,
+      "learning_rate": 7.892513097853644e-06,
+      "loss": 1.8634,
+      "step": 1247
+    },
+    {
+      "epoch": 1.4548104956268222,
+      "grad_norm": 0.09294755756855011,
+      "learning_rate": 7.887610020311443e-06,
+      "loss": 1.7943,
+      "step": 1248
+    },
+    {
+      "epoch": 1.4559766763848396,
+      "grad_norm": 0.12424357980489731,
+      "learning_rate": 7.882691981691813e-06,
+      "loss": 1.809,
+      "step": 1249
+    },
+    {
+      "epoch": 1.457142857142857,
+      "grad_norm": 0.11715559661388397,
+      "learning_rate": 7.877758913412564e-06,
+      "loss": 1.8457,
+      "step": 1250
+    },
+    {
+      "epoch": 1.4583090379008747,
+      "grad_norm": 0.09252289682626724,
+      "learning_rate": 7.872810746471689e-06,
+      "loss": 1.8334,
+      "step": 1251
+    },
+    {
+      "epoch": 1.4594752186588922,
+      "grad_norm": 0.09401300549507141,
+      "learning_rate": 7.867847411444141e-06,
+      "loss": 1.8895,
+      "step": 1252
+    },
+    {
+      "epoch": 1.4606413994169096,
+      "grad_norm": 0.09103366732597351,
+      "learning_rate": 7.862868838478595e-06,
+      "loss": 1.8965,
+      "step": 1253
+    },
+    {
+      "epoch": 1.461807580174927,
+      "grad_norm": 0.10641312599182129,
+      "learning_rate": 7.857874957294159e-06,
+      "loss": 1.8707,
+      "step": 1254
+    },
+    {
+      "epoch": 1.4629737609329445,
+      "grad_norm": 0.10444042086601257,
+      "learning_rate": 7.852865697177074e-06,
+      "loss": 1.8855,
+      "step": 1255
+    },
+    {
+      "epoch": 1.4641399416909622,
+      "grad_norm": 0.09730406850576401,
+      "learning_rate": 7.847840986977382e-06,
+      "loss": 1.9491,
+      "step": 1256
+    },
+    {
+      "epoch": 1.4653061224489796,
+      "grad_norm": 0.09691834449768066,
+      "learning_rate": 7.842800755105545e-06,
+      "loss": 1.809,
+      "step": 1257
+    },
+    {
+      "epoch": 1.466472303206997,
+      "grad_norm": 0.1531400829553604,
+      "learning_rate": 7.837744929529048e-06,
+      "loss": 1.9034,
+      "step": 1258
+    },
+    {
+      "epoch": 1.4676384839650145,
+      "grad_norm": 0.21025118231773376,
+      "learning_rate": 7.832673437768979e-06,
+      "loss": 1.9233,
+      "step": 1259
+    },
+    {
+      "epoch": 1.4688046647230322,
+      "grad_norm": 0.08831258863210678,
+      "learning_rate": 7.827586206896552e-06,
+      "loss": 1.8254,
+      "step": 1260
+    },
+    {
+      "epoch": 1.4699708454810496,
+      "grad_norm": 0.1076580360531807,
+      "learning_rate": 7.822483163529614e-06,
+      "loss": 1.8411,
+      "step": 1261
+    },
+    {
+      "epoch": 1.471137026239067,
+      "grad_norm": 0.14009325206279755,
+      "learning_rate": 7.817364233829126e-06,
+      "loss": 1.9235,
+      "step": 1262
+    },
+    {
+      "epoch": 1.4723032069970845,
+      "grad_norm": 0.09746917337179184,
+      "learning_rate": 7.812229343495584e-06,
+      "loss": 1.7835,
+      "step": 1263
+    },
+    {
+      "epoch": 1.473469387755102,
+      "grad_norm": 0.08734530955553055,
+      "learning_rate": 7.807078417765442e-06,
+      "loss": 1.8085,
+      "step": 1264
+    },
+    {
+      "epoch": 1.4746355685131196,
+      "grad_norm": 0.1072184294462204,
+      "learning_rate": 7.801911381407471e-06,
+      "loss": 1.7967,
+      "step": 1265
+    },
+    {
+      "epoch": 1.475801749271137,
+      "grad_norm": 0.0964658185839653,
+      "learning_rate": 7.796728158719108e-06,
+      "loss": 1.8587,
+      "step": 1266
+    },
+    {
+      "epoch": 1.4769679300291545,
+      "grad_norm": 0.09876925498247147,
+      "learning_rate": 7.791528673522747e-06,
+      "loss": 1.9008,
+      "step": 1267
+    },
+    {
+      "epoch": 1.478134110787172,
+      "grad_norm": 0.10326961427927017,
+      "learning_rate": 7.786312849162012e-06,
+      "loss": 1.8328,
+      "step": 1268
+    },
+    {
+      "epoch": 1.4793002915451896,
+      "grad_norm": 0.2745940089225769,
+      "learning_rate": 7.781080608497991e-06,
+      "loss": 1.7534,
+      "step": 1269
+    },
+    {
+      "epoch": 1.480466472303207,
+      "grad_norm": 0.11900675296783447,
+      "learning_rate": 7.775831873905429e-06,
+      "loss": 1.8097,
+      "step": 1270
+    },
+    {
+      "epoch": 1.4816326530612245,
+      "grad_norm": 0.11225635558366776,
+      "learning_rate": 7.7705665672689e-06,
+      "loss": 1.9646,
+      "step": 1271
+    },
+    {
+      "epoch": 1.482798833819242,
+      "grad_norm": 0.09173408895730972,
+      "learning_rate": 7.765284609978918e-06,
+      "loss": 1.7492,
+      "step": 1272
+    },
+    {
+      "epoch": 1.4839650145772594,
+      "grad_norm": 0.09222909063100815,
+      "learning_rate": 7.759985922928032e-06,
+      "loss": 1.9005,
+      "step": 1273
+    },
+    {
+      "epoch": 1.485131195335277,
+      "grad_norm": 0.11145643144845963,
+      "learning_rate": 7.754670426506874e-06,
+      "loss": 1.895,
+      "step": 1274
+    },
+    {
+      "epoch": 1.4862973760932945,
+      "grad_norm": 0.11737634986639023,
+      "learning_rate": 7.749338040600176e-06,
+      "loss": 1.9241,
+      "step": 1275
+    },
+    {
+      "epoch": 1.487463556851312,
+      "grad_norm": 0.10823045670986176,
+      "learning_rate": 7.743988684582744e-06,
+      "loss": 1.9294,
+      "step": 1276
+    },
+    {
+      "epoch": 1.4886297376093294,
+      "grad_norm": 0.09695983678102493,
+      "learning_rate": 7.73862227731539e-06,
+      "loss": 1.9147,
+      "step": 1277
+    },
+    {
+      "epoch": 1.489795918367347,
+      "grad_norm": 0.0868280827999115,
+      "learning_rate": 7.73323873714083e-06,
+      "loss": 1.8046,
+      "step": 1278
+    },
+    {
+      "epoch": 1.4909620991253645,
+      "grad_norm": 0.12231990694999695,
+      "learning_rate": 7.727837981879554e-06,
+      "loss": 1.846,
+      "step": 1279
+    },
+    {
+      "epoch": 1.492128279883382,
+      "grad_norm": 0.09998560696840286,
+      "learning_rate": 7.722419928825623e-06,
+      "loss": 1.8185,
+      "step": 1280
+    },
+    {
+      "epoch": 1.4932944606413994,
+      "grad_norm": 0.10422658920288086,
+      "learning_rate": 7.71698449474247e-06,
+      "loss": 1.8682,
+      "step": 1281
+    },
+    {
+      "epoch": 1.4944606413994168,
+      "grad_norm": 0.2020479291677475,
+      "learning_rate": 7.711531595858622e-06,
+      "loss": 1.9028,
+      "step": 1282
+    },
+    {
+      "epoch": 1.4956268221574345,
+      "grad_norm": 0.11782852560281754,
+      "learning_rate": 7.706061147863401e-06,
+      "loss": 1.8325,
+      "step": 1283
+    },
+    {
+      "epoch": 1.496793002915452,
+      "grad_norm": 0.09054877609014511,
+      "learning_rate": 7.700573065902578e-06,
+      "loss": 1.7925,
+      "step": 1284
+    },
+    {
+      "epoch": 1.4979591836734694,
+      "grad_norm": 0.10027078539133072,
+      "learning_rate": 7.69506726457399e-06,
+      "loss": 1.8281,
+      "step": 1285
+    },
+    {
+      "epoch": 1.4991253644314868,
+      "grad_norm": 0.16419924795627594,
+      "learning_rate": 7.689543657923105e-06,
+      "loss": 1.8089,
+      "step": 1286
+    },
+    {
+      "epoch": 1.5002915451895045,
+      "grad_norm": 0.42977407574653625,
+      "learning_rate": 7.684002159438547e-06,
+      "loss": 1.8491,
+      "step": 1287
+    },
+    {
+      "epoch": 1.501457725947522,
+      "grad_norm": 0.0883215069770813,
+      "learning_rate": 7.678442682047584e-06,
+      "loss": 1.8612,
+      "step": 1288
+    },
+    {
+      "epoch": 1.5026239067055394,
+      "grad_norm": 0.09914589673280716,
+      "learning_rate": 7.672865138111573e-06,
+      "loss": 1.7655,
+      "step": 1289
+    },
+    {
+      "epoch": 1.5037900874635568,
+      "grad_norm": 0.12100101262331009,
+      "learning_rate": 7.66726943942134e-06,
+      "loss": 1.8572,
+      "step": 1290
+    },
+    {
+      "epoch": 1.5037900874635568,
+      "eval_loss": 0.9123232960700989,
+      "eval_runtime": 168.3929,
+      "eval_samples_per_second": 40.999,
+      "eval_steps_per_second": 2.565,
+      "step": 1290
+    },
+    {
+      "epoch": 1.5049562682215742,
+      "grad_norm": 0.143805131316185,
+      "learning_rate": 7.661655497192538e-06,
+      "loss": 1.8935,
+      "step": 1291
+    },
+    {
+      "epoch": 1.5061224489795917,
+      "grad_norm": 0.09935273230075836,
+      "learning_rate": 7.656023222060958e-06,
+      "loss": 1.8658,
+      "step": 1292
+    },
+    {
+      "epoch": 1.5072886297376094,
+      "grad_norm": 0.1100090742111206,
+      "learning_rate": 7.650372524077777e-06,
+      "loss": 1.8457,
+      "step": 1293
+    },
+    {
+      "epoch": 1.5084548104956268,
+      "grad_norm": 0.09527628868818283,
+      "learning_rate": 7.64470331270477e-06,
+      "loss": 1.7856,
+      "step": 1294
+    },
+    {
+      "epoch": 1.5096209912536445,
+      "grad_norm": 0.14890426397323608,
+      "learning_rate": 7.63901549680948e-06,
+      "loss": 1.8649,
+      "step": 1295
+    },
+    {
+      "epoch": 1.510787172011662,
+      "grad_norm": 0.11212334036827087,
+      "learning_rate": 7.633308984660337e-06,
+      "loss": 1.9234,
+      "step": 1296
+    },
+    {
+      "epoch": 1.5119533527696793,
+      "grad_norm": 0.0986352413892746,
+      "learning_rate": 7.627583683921712e-06,
+      "loss": 1.8106,
+      "step": 1297
+    },
+    {
+      "epoch": 1.5131195335276968,
+      "grad_norm": 0.10396338254213333,
+      "learning_rate": 7.621839501648956e-06,
+      "loss": 1.8541,
+      "step": 1298
+    },
+    {
+      "epoch": 1.5142857142857142,
+      "grad_norm": 0.13405925035476685,
+      "learning_rate": 7.616076344283355e-06,
+      "loss": 1.8803,
+      "step": 1299
+    },
+    {
+      "epoch": 1.5154518950437317,
+      "grad_norm": 0.1276298314332962,
+      "learning_rate": 7.61029411764706e-06,
+      "loss": 1.8556,
+      "step": 1300
+    },
+    {
+      "epoch": 1.5166180758017491,
+      "grad_norm": 0.08804850280284882,
+      "learning_rate": 7.604492726937949e-06,
+      "loss": 1.8926,
+      "step": 1301
+    },
+    {
+      "epoch": 1.5177842565597668,
+      "grad_norm": 0.16054490208625793,
+      "learning_rate": 7.598672076724456e-06,
+      "loss": 1.8649,
+      "step": 1302
+    },
+    {
+      "epoch": 1.5189504373177842,
+      "grad_norm": 0.10433606058359146,
+      "learning_rate": 7.59283207094033e-06,
+      "loss": 1.8793,
+      "step": 1303
+    },
+    {
+      "epoch": 1.520116618075802,
+      "grad_norm": 0.10368677973747253,
+      "learning_rate": 7.586972612879349e-06,
+      "loss": 1.9074,
+      "step": 1304
+    },
+    {
+      "epoch": 1.5212827988338193,
+      "grad_norm": 0.09686250239610672,
+      "learning_rate": 7.581093605189991e-06,
+      "loss": 1.8871,
+      "step": 1305
+    },
+    {
+      "epoch": 1.5224489795918368,
+      "grad_norm": 0.08583763241767883,
+      "learning_rate": 7.575194949870035e-06,
+      "loss": 1.7461,
+      "step": 1306
+    },
+    {
+      "epoch": 1.5236151603498542,
+      "grad_norm": 0.09494598954916,
+      "learning_rate": 7.569276548261112e-06,
+      "loss": 1.8421,
+      "step": 1307
+    },
+    {
+      "epoch": 1.5247813411078717,
+      "grad_norm": 0.0952291265130043,
+      "learning_rate": 7.563338301043218e-06,
+      "loss": 1.8123,
+      "step": 1308
+    },
+    {
+      "epoch": 1.5259475218658891,
+      "grad_norm": 0.09135711938142776,
+      "learning_rate": 7.557380108229149e-06,
+      "loss": 1.869,
+      "step": 1309
+    },
+    {
+      "epoch": 1.5271137026239066,
+      "grad_norm": 0.14102980494499207,
+      "learning_rate": 7.551401869158879e-06,
+      "loss": 1.8277,
+      "step": 1310
+    },
+    {
+      "epoch": 1.5282798833819242,
+      "grad_norm": 0.09577342122793198,
+      "learning_rate": 7.545403482493915e-06,
+      "loss": 1.859,
+      "step": 1311
+    },
+    {
+      "epoch": 1.5294460641399417,
+      "grad_norm": 0.09546927362680435,
+      "learning_rate": 7.539384846211553e-06,
+      "loss": 1.895,
+      "step": 1312
+    },
+    {
+      "epoch": 1.5306122448979593,
+      "grad_norm": 0.1553334891796112,
+      "learning_rate": 7.533345857599098e-06,
+      "loss": 1.9195,
+      "step": 1313
+    },
+    {
+      "epoch": 1.5317784256559768,
+      "grad_norm": 0.16446226835250854,
+      "learning_rate": 7.527286413248024e-06,
+      "loss": 1.8562,
+      "step": 1314
+    },
+    {
+      "epoch": 1.5329446064139942,
+      "grad_norm": 0.0915827676653862,
+      "learning_rate": 7.521206409048068e-06,
+      "loss": 1.7919,
+      "step": 1315
+    },
+    {
+      "epoch": 1.5341107871720117,
+      "grad_norm": 0.12070725113153458,
+      "learning_rate": 7.51510574018127e-06,
+      "loss": 2.0143,
+      "step": 1316
+    },
+    {
+      "epoch": 1.535276967930029,
+      "grad_norm": 0.10119042545557022,
+      "learning_rate": 7.5089843011159446e-06,
+      "loss": 1.8471,
+      "step": 1317
+    },
+    {
+      "epoch": 1.5364431486880465,
+      "grad_norm": 0.09589023888111115,
+      "learning_rate": 7.502841985600606e-06,
+      "loss": 1.9293,
+      "step": 1318
+    },
+    {
+      "epoch": 1.537609329446064,
+      "grad_norm": 0.1059497743844986,
+      "learning_rate": 7.496678686657811e-06,
+      "loss": 1.7942,
+      "step": 1319
+    },
+    {
+      "epoch": 1.5387755102040817,
+      "grad_norm": 0.12368198484182358,
+      "learning_rate": 7.490494296577947e-06,
+      "loss": 1.9004,
+      "step": 1320
+    },
+    {
+      "epoch": 1.539941690962099,
+      "grad_norm": 0.08914022892713547,
+      "learning_rate": 7.4842887069129685e-06,
+      "loss": 1.8809,
+      "step": 1321
+    },
+    {
+      "epoch": 1.5411078717201168,
+      "grad_norm": 0.27421316504478455,
+      "learning_rate": 7.478061808470051e-06,
+      "loss": 1.8678,
+      "step": 1322
+    },
+    {
+      "epoch": 1.5422740524781342,
+      "grad_norm": 0.09426335245370865,
+      "learning_rate": 7.4718134913051785e-06,
+      "loss": 1.798,
+      "step": 1323
+    },
+    {
+      "epoch": 1.5434402332361516,
+      "grad_norm": 0.09947004169225693,
+      "learning_rate": 7.465543644716692e-06,
+      "loss": 1.8018,
+      "step": 1324
+    },
+    {
+      "epoch": 1.544606413994169,
+      "grad_norm": 0.09941276907920837,
+      "learning_rate": 7.459252157238734e-06,
+      "loss": 1.8775,
+      "step": 1325
+    },
+    {
+      "epoch": 1.5457725947521865,
+      "grad_norm": 0.15721967816352844,
+      "learning_rate": 7.452938916634653e-06,
+      "loss": 1.8494,
+      "step": 1326
+    },
+    {
+      "epoch": 1.546938775510204,
+      "grad_norm": 0.09494185447692871,
+      "learning_rate": 7.446603809890321e-06,
+      "loss": 1.9228,
+      "step": 1327
+    },
+    {
+      "epoch": 1.5481049562682214,
+      "grad_norm": 0.20144601166248322,
+      "learning_rate": 7.440246723207402e-06,
+      "loss": 1.8498,
+      "step": 1328
+    },
+    {
+      "epoch": 1.549271137026239,
+      "grad_norm": 0.10753075778484344,
+      "learning_rate": 7.433867541996524e-06,
+      "loss": 1.9262,
+      "step": 1329
+    },
+    {
+      "epoch": 1.5504373177842565,
+      "grad_norm": 0.13006146252155304,
+      "learning_rate": 7.4274661508704075e-06,
+      "loss": 1.7475,
+      "step": 1330
+    },
+    {
+      "epoch": 1.5516034985422742,
+      "grad_norm": 0.08969902247190475,
+      "learning_rate": 7.421042433636892e-06,
+      "loss": 1.8281,
+      "step": 1331
+    },
+    {
+      "epoch": 1.5527696793002916,
+      "grad_norm": 0.08640894293785095,
+      "learning_rate": 7.414596273291926e-06,
+      "loss": 1.8369,
+      "step": 1332
+    },
+    {
+      "epoch": 1.553935860058309,
+      "grad_norm": 0.11855178326368332,
+      "learning_rate": 7.408127552012445e-06,
+      "loss": 1.8138,
+      "step": 1333
+    },
+    {
+      "epoch": 1.5551020408163265,
+      "grad_norm": 0.11021913588047028,
+      "learning_rate": 7.401636151149202e-06,
+      "loss": 1.8817,
+      "step": 1334
+    },
+    {
+      "epoch": 1.556268221574344,
+      "grad_norm": 0.0965341255068779,
+      "learning_rate": 7.395121951219513e-06,
+      "loss": 1.8887,
+      "step": 1335
+    },
+    {
+      "epoch": 1.5574344023323614,
+      "grad_norm": 0.23131626844406128,
+      "learning_rate": 7.388584831899923e-06,
+      "loss": 1.9435,
+      "step": 1336
+    },
+    {
+      "epoch": 1.5586005830903789,
+      "grad_norm": 0.1027970016002655,
+      "learning_rate": 7.3820246720187986e-06,
+      "loss": 1.7942,
+      "step": 1337
+    },
+    {
+      "epoch": 1.5597667638483965,
+      "grad_norm": 0.12681303918361664,
+      "learning_rate": 7.375441349548843e-06,
+      "loss": 1.8547,
+      "step": 1338
+    },
+    {
+      "epoch": 1.560932944606414,
+      "grad_norm": 0.09002608805894852,
+      "learning_rate": 7.36883474159953e-06,
+      "loss": 1.867,
+      "step": 1339
+    },
+    {
+      "epoch": 1.5620991253644316,
+      "grad_norm": 0.11985815316438675,
+      "learning_rate": 7.36220472440945e-06,
+      "loss": 1.8515,
+      "step": 1340
+    },
+    {
+      "epoch": 1.563265306122449,
+      "grad_norm": 0.10269878059625626,
+      "learning_rate": 7.355551173338592e-06,
+      "loss": 1.8364,
+      "step": 1341
+    },
+    {
+      "epoch": 1.5644314868804665,
+      "grad_norm": 0.0989355593919754,
+      "learning_rate": 7.348873962860531e-06,
+      "loss": 1.8404,
+      "step": 1342
+    },
+    {
+      "epoch": 1.565597667638484,
+      "grad_norm": 0.09452962875366211,
+      "learning_rate": 7.342172966554523e-06,
+      "loss": 1.8363,
+      "step": 1343
+    },
+    {
+      "epoch": 1.5667638483965014,
+      "grad_norm": 0.08915907144546509,
+      "learning_rate": 7.335448057097542e-06,
+      "loss": 1.8672,
+      "step": 1344
+    },
+    {
+      "epoch": 1.5679300291545188,
+      "grad_norm": 0.11357169598340988,
+      "learning_rate": 7.328699106256208e-06,
+      "loss": 1.8912,
+      "step": 1345
+    },
+    {
+      "epoch": 1.5690962099125363,
+      "grad_norm": 0.08834497630596161,
+      "learning_rate": 7.321925984878632e-06,
+      "loss": 1.7572,
+      "step": 1346
+    },
+    {
+      "epoch": 1.570262390670554,
+      "grad_norm": 0.2512083351612091,
+      "learning_rate": 7.315128562886187e-06,
+      "loss": 1.9365,
+      "step": 1347
+    },
+    {
+      "epoch": 1.5714285714285714,
+      "grad_norm": 0.09856240451335907,
+      "learning_rate": 7.308306709265176e-06,
+      "loss": 1.837,
+      "step": 1348
+    },
+    {
+      "epoch": 1.572594752186589,
+      "grad_norm": 0.12346161156892776,
+      "learning_rate": 7.301460292058412e-06,
+      "loss": 1.8055,
+      "step": 1349
+    },
+    {
+      "epoch": 1.5737609329446065,
+      "grad_norm": 0.09703285247087479,
+      "learning_rate": 7.2945891783567134e-06,
+      "loss": 1.9265,
+      "step": 1350
+    },
+    {
+      "epoch": 1.574927113702624,
+      "grad_norm": 0.24820847809314728,
+      "learning_rate": 7.287693234290304e-06,
+      "loss": 1.8712,
+      "step": 1351
+    },
+    {
+      "epoch": 1.5760932944606414,
+      "grad_norm": 0.11651188135147095,
+      "learning_rate": 7.280772325020114e-06,
+      "loss": 1.8892,
+      "step": 1352
+    },
+    {
+      "epoch": 1.5772594752186588,
+      "grad_norm": 0.09457270801067352,
+      "learning_rate": 7.273826314728995e-06,
+      "loss": 1.7943,
+      "step": 1353
+    },
+    {
+      "epoch": 1.5784256559766763,
+      "grad_norm": 0.10134178400039673,
+      "learning_rate": 7.266855066612838e-06,
+      "loss": 1.8234,
+      "step": 1354
+    },
+    {
+      "epoch": 1.5795918367346937,
+      "grad_norm": 0.23578709363937378,
+      "learning_rate": 7.259858442871588e-06,
+      "loss": 1.8948,
+      "step": 1355
+    },
+    {
+      "epoch": 1.5807580174927114,
+      "grad_norm": 0.11537405848503113,
+      "learning_rate": 7.252836304700162e-06,
+      "loss": 1.7732,
+      "step": 1356
+    },
+    {
+      "epoch": 1.5819241982507288,
+      "grad_norm": 0.12945684790611267,
+      "learning_rate": 7.245788512279277e-06,
+      "loss": 1.938,
+      "step": 1357
+    },
+    {
+      "epoch": 1.5830903790087465,
+      "grad_norm": 0.1158704087138176,
+      "learning_rate": 7.238714924766165e-06,
+      "loss": 1.882,
+      "step": 1358
+    },
+    {
+      "epoch": 1.584256559766764,
+      "grad_norm": 0.09275586903095245,
+      "learning_rate": 7.23161540028519e-06,
+      "loss": 1.9217,
+      "step": 1359
+    },
+    {
+      "epoch": 1.5854227405247814,
+      "grad_norm": 0.0872541144490242,
+      "learning_rate": 7.224489795918367e-06,
+      "loss": 1.7625,
+      "step": 1360
+    },
+    {
+      "epoch": 1.5865889212827988,
+      "grad_norm": 0.09915114939212799,
+      "learning_rate": 7.217337967695768e-06,
+      "loss": 1.8698,
+      "step": 1361
+    },
+    {
+      "epoch": 1.5877551020408163,
+      "grad_norm": 0.08939757943153381,
+      "learning_rate": 7.210159770585826e-06,
+      "loss": 1.8551,
+      "step": 1362
+    },
+    {
+      "epoch": 1.5889212827988337,
+      "grad_norm": 0.09835248440504074,
+      "learning_rate": 7.2029550584855326e-06,
+      "loss": 1.9515,
+      "step": 1363
+    },
+    {
+      "epoch": 1.5900874635568512,
+      "grad_norm": 0.09676311910152435,
+      "learning_rate": 7.195723684210526e-06,
+      "loss": 1.8083,
+      "step": 1364
+    },
+    {
+      "epoch": 1.5912536443148688,
+      "grad_norm": 0.1017245426774025,
+      "learning_rate": 7.188465499485068e-06,
+      "loss": 1.8818,
+      "step": 1365
+    },
+    {
+      "epoch": 1.5924198250728863,
+      "grad_norm": 0.10274836421012878,
+      "learning_rate": 7.181180354931903e-06,
+      "loss": 1.8657,
+      "step": 1366
+    },
+    {
+      "epoch": 1.593586005830904,
+      "grad_norm": 0.13532811403274536,
+      "learning_rate": 7.173868100062022e-06,
+      "loss": 1.8292,
+      "step": 1367
+    },
+    {
+      "epoch": 1.5947521865889214,
+      "grad_norm": 0.12773634493350983,
+      "learning_rate": 7.166528583264293e-06,
+      "loss": 1.837,
+      "step": 1368
+    },
+    {
+      "epoch": 1.5959183673469388,
+      "grad_norm": 0.1027255430817604,
+      "learning_rate": 7.159161651794979e-06,
+      "loss": 1.8384,
+      "step": 1369
+    },
+    {
+      "epoch": 1.5970845481049563,
+      "grad_norm": 0.10105174034833908,
+      "learning_rate": 7.1517671517671525e-06,
+      "loss": 1.9072,
+      "step": 1370
+    },
+    {
+      "epoch": 1.5982507288629737,
+      "grad_norm": 0.11251997947692871,
+      "learning_rate": 7.144344928139971e-06,
+      "loss": 1.9145,
+      "step": 1371
+    },
+    {
+      "epoch": 1.5994169096209911,
+      "grad_norm": 0.11653496325016022,
+      "learning_rate": 7.136894824707848e-06,
+      "loss": 1.857,
+      "step": 1372
+    },
+    {
+      "epoch": 1.6005830903790086,
+      "grad_norm": 0.11509859561920166,
+      "learning_rate": 7.129416684089484e-06,
+      "loss": 1.8933,
+      "step": 1373
+    },
+    {
+      "epoch": 1.6017492711370263,
+      "grad_norm": 0.1637764722108841,
+      "learning_rate": 7.1219103477168e-06,
+      "loss": 1.8278,
+      "step": 1374
+    },
+    {
+      "epoch": 1.6029154518950437,
+      "grad_norm": 0.09219600260257721,
+      "learning_rate": 7.114375655823716e-06,
+      "loss": 1.9093,
+      "step": 1375
+    },
+    {
+      "epoch": 1.6040816326530614,
+      "grad_norm": 0.13009685277938843,
+      "learning_rate": 7.1068124474348195e-06,
+      "loss": 1.8146,
+      "step": 1376
+    },
+    {
+      "epoch": 1.6040816326530614,
+      "eval_loss": 0.9113958477973938,
+      "eval_runtime": 169.5003,
+      "eval_samples_per_second": 40.731,
+      "eval_steps_per_second": 2.549,
+      "step": 1376
+    },
+    {
+      "epoch": 1.6052478134110788,
+      "grad_norm": 0.08670420199632645,
+      "learning_rate": 7.099220560353908e-06,
+      "loss": 1.8024,
+      "step": 1377
+    },
+    {
+      "epoch": 1.6064139941690962,
+      "grad_norm": 0.11450019478797913,
+      "learning_rate": 7.091599831152386e-06,
+      "loss": 1.9402,
+      "step": 1378
+    },
+    {
+      "epoch": 1.6075801749271137,
+      "grad_norm": 0.10502946376800537,
+      "learning_rate": 7.08395009515754e-06,
+      "loss": 1.7888,
+      "step": 1379
+    },
+    {
+      "epoch": 1.6087463556851311,
+      "grad_norm": 0.09427541494369507,
+      "learning_rate": 7.076271186440677e-06,
+      "loss": 1.8692,
+      "step": 1380
+    },
+    {
+      "epoch": 1.6099125364431486,
+      "grad_norm": 0.1922471523284912,
+      "learning_rate": 7.068562937805138e-06,
+      "loss": 1.8449,
+      "step": 1381
+    },
+    {
+      "epoch": 1.611078717201166,
+      "grad_norm": 0.14434024691581726,
+      "learning_rate": 7.06082518077414e-06,
+      "loss": 1.8261,
+      "step": 1382
+    },
+    {
+      "epoch": 1.6122448979591837,
+      "grad_norm": 0.11958174407482147,
+      "learning_rate": 7.053057745578521e-06,
+      "loss": 1.7907,
+      "step": 1383
+    },
+    {
+      "epoch": 1.6134110787172011,
+      "grad_norm": 0.14280973374843597,
+      "learning_rate": 7.045260461144322e-06,
+      "loss": 1.8197,
+      "step": 1384
+    },
+    {
+      "epoch": 1.6145772594752188,
+      "grad_norm": 0.08871062099933624,
+      "learning_rate": 7.0374331550802156e-06,
+      "loss": 1.9331,
+      "step": 1385
+    },
+    {
+      "epoch": 1.6157434402332362,
+      "grad_norm": 0.12191244959831238,
+      "learning_rate": 7.02957565366481e-06,
+      "loss": 1.8482,
+      "step": 1386
+    },
+    {
+      "epoch": 1.6169096209912537,
+      "grad_norm": 0.08608485758304596,
+      "learning_rate": 7.021687781833798e-06,
+      "loss": 1.9124,
+      "step": 1387
+    },
+    {
+      "epoch": 1.6180758017492711,
+      "grad_norm": 0.09362581372261047,
+      "learning_rate": 7.013769363166955e-06,
+      "loss": 1.9187,
+      "step": 1388
+    },
+    {
+      "epoch": 1.6192419825072886,
+      "grad_norm": 0.10901051014661789,
+      "learning_rate": 7.005820219874974e-06,
+      "loss": 1.8052,
+      "step": 1389
+    },
+    {
+      "epoch": 1.620408163265306,
+      "grad_norm": 0.08932410925626755,
+      "learning_rate": 6.997840172786177e-06,
+      "loss": 1.9061,
+      "step": 1390
+    },
+    {
+      "epoch": 1.6215743440233235,
+      "grad_norm": 0.12499213963747025,
+      "learning_rate": 6.989829041333045e-06,
+      "loss": 1.8772,
+      "step": 1391
+    },
+    {
+      "epoch": 1.6227405247813411,
+      "grad_norm": 0.09977991133928299,
+      "learning_rate": 6.981786643538595e-06,
+      "loss": 1.9252,
+      "step": 1392
+    },
+    {
+      "epoch": 1.6239067055393586,
+      "grad_norm": 0.10817747563123703,
+      "learning_rate": 6.973712796002606e-06,
+      "loss": 1.8255,
+      "step": 1393
+    },
+    {
+      "epoch": 1.6250728862973762,
+      "grad_norm": 0.13333562016487122,
+      "learning_rate": 6.96560731388768e-06,
+      "loss": 1.8299,
+      "step": 1394
+    },
+    {
+      "epoch": 1.6262390670553937,
+      "grad_norm": 0.11981835961341858,
+      "learning_rate": 6.957470010905126e-06,
+      "loss": 1.853,
+      "step": 1395
+    },
+    {
+      "epoch": 1.6274052478134111,
+      "grad_norm": 0.16619721055030823,
+      "learning_rate": 6.949300699300698e-06,
+      "loss": 1.9089,
+      "step": 1396
+    },
+    {
+      "epoch": 1.6285714285714286,
+      "grad_norm": 0.09967787563800812,
+      "learning_rate": 6.941099189840158e-06,
+      "loss": 1.9353,
+      "step": 1397
+    },
+    {
+      "epoch": 1.629737609329446,
+      "grad_norm": 0.08756931126117706,
+      "learning_rate": 6.932865291794647e-06,
+      "loss": 1.8595,
+      "step": 1398
+    },
+    {
+      "epoch": 1.6309037900874634,
+      "grad_norm": 0.10072306543588638,
+      "learning_rate": 6.924598812925918e-06,
+      "loss": 1.9183,
+      "step": 1399
+    },
+    {
+      "epoch": 1.6320699708454809,
+      "grad_norm": 0.14198710024356842,
+      "learning_rate": 6.916299559471365e-06,
+      "loss": 1.8682,
+      "step": 1400
+    },
+    {
+      "epoch": 1.6332361516034986,
+      "grad_norm": 0.09007628262042999,
+      "learning_rate": 6.907967336128891e-06,
+      "loss": 1.8971,
+      "step": 1401
+    },
+    {
+      "epoch": 1.634402332361516,
+      "grad_norm": 0.14171400666236877,
+      "learning_rate": 6.899601946041574e-06,
+      "loss": 1.8304,
+      "step": 1402
+    },
+    {
+      "epoch": 1.6355685131195337,
+      "grad_norm": 0.1021130383014679,
+      "learning_rate": 6.891203190782184e-06,
+      "loss": 1.886,
+      "step": 1403
+    },
+    {
+      "epoch": 1.636734693877551,
+      "grad_norm": 0.08812300860881805,
+      "learning_rate": 6.882770870337477e-06,
+      "loss": 1.8556,
+      "step": 1404
+    },
+    {
+      "epoch": 1.6379008746355685,
+      "grad_norm": 0.14782240986824036,
+      "learning_rate": 6.8743047830923245e-06,
+      "loss": 1.7599,
+      "step": 1405
+    },
+    {
+      "epoch": 1.639067055393586,
+      "grad_norm": 0.10316556692123413,
+      "learning_rate": 6.865804725813642e-06,
+      "loss": 1.8055,
+      "step": 1406
+    },
+    {
+      "epoch": 1.6402332361516034,
+      "grad_norm": 0.0902082547545433,
+      "learning_rate": 6.8572704936341295e-06,
+      "loss": 1.8668,
+      "step": 1407
+    },
+    {
+      "epoch": 1.6413994169096209,
+      "grad_norm": 0.10623693466186523,
+      "learning_rate": 6.848701880035811e-06,
+      "loss": 1.9239,
+      "step": 1408
+    },
+    {
+      "epoch": 1.6425655976676383,
+      "grad_norm": 0.10182394087314606,
+      "learning_rate": 6.840098676833371e-06,
+      "loss": 1.8749,
+      "step": 1409
+    },
+    {
+      "epoch": 1.643731778425656,
+      "grad_norm": 0.1506323218345642,
+      "learning_rate": 6.831460674157303e-06,
+      "loss": 1.8323,
+      "step": 1410
+    },
+    {
+      "epoch": 1.6448979591836734,
+      "grad_norm": 0.10350453853607178,
+      "learning_rate": 6.822787660436839e-06,
+      "loss": 1.8632,
+      "step": 1411
+    },
+    {
+      "epoch": 1.646064139941691,
+      "grad_norm": 0.12032853811979294,
+      "learning_rate": 6.814079422382672e-06,
+      "loss": 1.8356,
+      "step": 1412
+    },
+    {
+      "epoch": 1.6472303206997085,
+      "grad_norm": 0.10496199131011963,
+      "learning_rate": 6.805335744969478e-06,
+      "loss": 1.907,
+      "step": 1413
+    },
+    {
+      "epoch": 1.648396501457726,
+      "grad_norm": 0.09652126580476761,
+      "learning_rate": 6.7965564114182156e-06,
+      "loss": 1.8991,
+      "step": 1414
+    },
+    {
+      "epoch": 1.6495626822157434,
+      "grad_norm": 0.09990932047367096,
+      "learning_rate": 6.787741203178207e-06,
+      "loss": 1.8821,
+      "step": 1415
+    },
+    {
+      "epoch": 1.6507288629737609,
+      "grad_norm": 0.09016650915145874,
+      "learning_rate": 6.778889899909009e-06,
+      "loss": 1.9455,
+      "step": 1416
+    },
+    {
+      "epoch": 1.6518950437317783,
+      "grad_norm": 0.1096726655960083,
+      "learning_rate": 6.770002279462048e-06,
+      "loss": 1.8274,
+      "step": 1417
+    },
+    {
+      "epoch": 1.6530612244897958,
+      "grad_norm": 0.0972224548459053,
+      "learning_rate": 6.761078117862039e-06,
+      "loss": 1.8159,
+      "step": 1418
+    },
+    {
+      "epoch": 1.6542274052478134,
+      "grad_norm": 0.09369879215955734,
+      "learning_rate": 6.752117189288168e-06,
+      "loss": 1.8344,
+      "step": 1419
+    },
+    {
+      "epoch": 1.6553935860058309,
+      "grad_norm": 0.10772797465324402,
+      "learning_rate": 6.743119266055046e-06,
+      "loss": 1.7893,
+      "step": 1420
+    },
+    {
+      "epoch": 1.6565597667638485,
+      "grad_norm": 0.08820302039384842,
+      "learning_rate": 6.734084118593428e-06,
+      "loss": 1.9267,
+      "step": 1421
+    },
+    {
+      "epoch": 1.657725947521866,
+      "grad_norm": 0.08872099965810776,
+      "learning_rate": 6.725011515430677e-06,
+      "loss": 1.8581,
+      "step": 1422
+    },
+    {
+      "epoch": 1.6588921282798834,
+      "grad_norm": 0.1653842031955719,
+      "learning_rate": 6.715901223171013e-06,
+      "loss": 1.8305,
+      "step": 1423
+    },
+    {
+      "epoch": 1.6600583090379009,
+      "grad_norm": 0.08434578031301498,
+      "learning_rate": 6.706753006475486e-06,
+      "loss": 1.9022,
+      "step": 1424
+    },
+    {
+      "epoch": 1.6612244897959183,
+      "grad_norm": 0.0915495827794075,
+      "learning_rate": 6.697566628041715e-06,
+      "loss": 1.8516,
+      "step": 1425
+    },
+    {
+      "epoch": 1.6623906705539357,
+      "grad_norm": 0.12227003276348114,
+      "learning_rate": 6.688341848583372e-06,
+      "loss": 1.9502,
+      "step": 1426
+    },
+    {
+      "epoch": 1.6635568513119532,
+      "grad_norm": 0.09166563302278519,
+      "learning_rate": 6.679078426809402e-06,
+      "loss": 1.8557,
+      "step": 1427
+    },
+    {
+      "epoch": 1.6647230320699709,
+      "grad_norm": 0.0955251008272171,
+      "learning_rate": 6.6697761194029865e-06,
+      "loss": 1.9215,
+      "step": 1428
+    },
+    {
+      "epoch": 1.6658892128279883,
+      "grad_norm": 0.13822583854198456,
+      "learning_rate": 6.660434681000234e-06,
+      "loss": 1.8721,
+      "step": 1429
+    },
+    {
+      "epoch": 1.667055393586006,
+      "grad_norm": 0.11127714812755585,
+      "learning_rate": 6.651053864168618e-06,
+      "loss": 1.9336,
+      "step": 1430
+    },
+    {
+      "epoch": 1.6682215743440234,
+      "grad_norm": 0.0910099670290947,
+      "learning_rate": 6.641633419385121e-06,
+      "loss": 1.8222,
+      "step": 1431
+    },
+    {
+      "epoch": 1.6693877551020408,
+      "grad_norm": 0.09762828797101974,
+      "learning_rate": 6.632173095014112e-06,
+      "loss": 1.8681,
+      "step": 1432
+    },
+    {
+      "epoch": 1.6705539358600583,
+      "grad_norm": 0.12968000769615173,
+      "learning_rate": 6.62267263728494e-06,
+      "loss": 1.8713,
+      "step": 1433
+    },
+    {
+      "epoch": 1.6717201166180757,
+      "grad_norm": 0.17734146118164062,
+      "learning_rate": 6.6131317902692495e-06,
+      "loss": 1.8914,
+      "step": 1434
+    },
+    {
+      "epoch": 1.6728862973760932,
+      "grad_norm": 0.09326016902923584,
+      "learning_rate": 6.603550295857989e-06,
+      "loss": 1.824,
+      "step": 1435
+    },
+    {
+      "epoch": 1.6740524781341106,
+      "grad_norm": 0.119589664041996,
+      "learning_rate": 6.59392789373814e-06,
+      "loss": 1.73,
+      "step": 1436
+    },
+    {
+      "epoch": 1.6752186588921283,
+      "grad_norm": 0.11904770880937576,
+      "learning_rate": 6.584264321369147e-06,
+      "loss": 1.8307,
+      "step": 1437
+    },
+    {
+      "epoch": 1.6763848396501457,
+      "grad_norm": 0.0998338982462883,
+      "learning_rate": 6.574559313959028e-06,
+      "loss": 1.8826,
+      "step": 1438
+    },
+    {
+      "epoch": 1.6775510204081634,
+      "grad_norm": 0.09207289665937424,
+      "learning_rate": 6.5648126044402e-06,
+      "loss": 1.8065,
+      "step": 1439
+    },
+    {
+      "epoch": 1.6787172011661808,
+      "grad_norm": 0.10164755582809448,
+      "learning_rate": 6.5550239234449765e-06,
+      "loss": 1.8363,
+      "step": 1440
+    },
+    {
+      "epoch": 1.6798833819241983,
+      "grad_norm": 0.1341422200202942,
+      "learning_rate": 6.545192999280748e-06,
+      "loss": 1.8733,
+      "step": 1441
+    },
+    {
+      "epoch": 1.6810495626822157,
+      "grad_norm": 0.15216612815856934,
+      "learning_rate": 6.5353195579048536e-06,
+      "loss": 1.8699,
+      "step": 1442
+    },
+    {
+      "epoch": 1.6822157434402332,
+      "grad_norm": 0.2008359730243683,
+      "learning_rate": 6.525403322899109e-06,
+      "loss": 1.8852,
+      "step": 1443
+    },
+    {
+      "epoch": 1.6833819241982506,
+      "grad_norm": 0.3109564185142517,
+      "learning_rate": 6.515444015444016e-06,
+      "loss": 1.9048,
+      "step": 1444
+    },
+    {
+      "epoch": 1.684548104956268,
+      "grad_norm": 0.09863084554672241,
+      "learning_rate": 6.505441354292624e-06,
+      "loss": 1.7898,
+      "step": 1445
+    },
+    {
+      "epoch": 1.6857142857142857,
+      "grad_norm": 0.288147509098053,
+      "learning_rate": 6.495395055744063e-06,
+      "loss": 1.8296,
+      "step": 1446
+    },
+    {
+      "epoch": 1.6868804664723032,
+      "grad_norm": 0.0914098396897316,
+      "learning_rate": 6.485304833616712e-06,
+      "loss": 1.8583,
+      "step": 1447
+    },
+    {
+      "epoch": 1.6880466472303208,
+      "grad_norm": 0.1086428165435791,
+      "learning_rate": 6.4751703992210326e-06,
+      "loss": 1.7747,
+      "step": 1448
+    },
+    {
+      "epoch": 1.6892128279883383,
+      "grad_norm": 0.38638582825660706,
+      "learning_rate": 6.464991461332032e-06,
+      "loss": 1.9139,
+      "step": 1449
+    },
+    {
+      "epoch": 1.6903790087463557,
+      "grad_norm": 0.1397067755460739,
+      "learning_rate": 6.4547677261613696e-06,
+      "loss": 1.9103,
+      "step": 1450
+    },
+    {
+      "epoch": 1.6915451895043732,
+      "grad_norm": 0.2751591205596924,
+      "learning_rate": 6.444498897329087e-06,
+      "loss": 1.8437,
+      "step": 1451
+    },
+    {
+      "epoch": 1.6927113702623906,
+      "grad_norm": 0.12566755712032318,
+      "learning_rate": 6.434184675834971e-06,
+      "loss": 1.9328,
+      "step": 1452
+    },
+    {
+      "epoch": 1.693877551020408,
+      "grad_norm": 0.1099894642829895,
+      "learning_rate": 6.4238247600295355e-06,
+      "loss": 1.8462,
+      "step": 1453
+    },
+    {
+      "epoch": 1.6950437317784257,
+      "grad_norm": 0.1555374413728714,
+      "learning_rate": 6.413418845584608e-06,
+      "loss": 1.8609,
+      "step": 1454
+    },
+    {
+      "epoch": 1.6962099125364432,
+      "grad_norm": 0.11200569570064545,
+      "learning_rate": 6.402966625463536e-06,
+      "loss": 1.7974,
+      "step": 1455
+    },
+    {
+      "epoch": 1.6973760932944606,
+      "grad_norm": 0.10128124058246613,
+      "learning_rate": 6.392467789890981e-06,
+      "loss": 1.8191,
+      "step": 1456
+    },
+    {
+      "epoch": 1.6985422740524783,
+      "grad_norm": 0.10733206570148468,
+      "learning_rate": 6.381922026322325e-06,
+      "loss": 1.8118,
+      "step": 1457
+    },
+    {
+      "epoch": 1.6997084548104957,
+      "grad_norm": 0.1016615480184555,
+      "learning_rate": 6.371329019412643e-06,
+      "loss": 1.8838,
+      "step": 1458
+    },
+    {
+      "epoch": 1.7008746355685131,
+      "grad_norm": 0.11140910536050797,
+      "learning_rate": 6.360688450985284e-06,
+      "loss": 1.8402,
+      "step": 1459
+    },
+    {
+      "epoch": 1.7020408163265306,
+      "grad_norm": 0.1253715455532074,
+      "learning_rate": 6.35e-06,
+      "loss": 1.836,
+      "step": 1460
+    },
+    {
+      "epoch": 1.703206997084548,
+      "grad_norm": 0.7384465932846069,
+      "learning_rate": 6.339263342520672e-06,
+      "loss": 1.8555,
+      "step": 1461
+    },
+    {
+      "epoch": 1.7043731778425655,
+      "grad_norm": 0.12195040285587311,
+      "learning_rate": 6.328478151682572e-06,
+      "loss": 1.8372,
+      "step": 1462
+    },
+    {
+      "epoch": 1.7043731778425655,
+      "eval_loss": 0.9103754162788391,
+      "eval_runtime": 167.6488,
+      "eval_samples_per_second": 41.181,
+      "eval_steps_per_second": 2.577,
+      "step": 1462
+    },
+    {
+      "epoch": 1.7055393586005831,
+      "grad_norm": 0.09881063550710678,
+      "learning_rate": 6.3176440976592005e-06,
+      "loss": 1.8492,
+      "step": 1463
+    },
+    {
+      "epoch": 1.7067055393586006,
+      "grad_norm": 0.10664098709821701,
+      "learning_rate": 6.306760847628658e-06,
+      "loss": 1.7584,
+      "step": 1464
+    },
+    {
+      "epoch": 1.707871720116618,
+      "grad_norm": 0.16759252548217773,
+      "learning_rate": 6.295828065739571e-06,
+      "loss": 1.8725,
+      "step": 1465
+    },
+    {
+      "epoch": 1.7090379008746357,
+      "grad_norm": 0.09974395483732224,
+      "learning_rate": 6.2848454130765335e-06,
+      "loss": 1.7934,
+      "step": 1466
+    },
+    {
+      "epoch": 1.7102040816326531,
+      "grad_norm": 0.0997401773929596,
+      "learning_rate": 6.273812547625096e-06,
+      "loss": 1.8203,
+      "step": 1467
+    },
+    {
+      "epoch": 1.7113702623906706,
+      "grad_norm": 0.11043285578489304,
+      "learning_rate": 6.262729124236253e-06,
+      "loss": 1.9336,
+      "step": 1468
+    },
+    {
+      "epoch": 1.712536443148688,
+      "grad_norm": 0.11582586914300919,
+      "learning_rate": 6.2515947945904565e-06,
+      "loss": 1.8114,
+      "step": 1469
+    },
+    {
+      "epoch": 1.7137026239067055,
+      "grad_norm": 0.09255058318376541,
+      "learning_rate": 6.240409207161126e-06,
+      "loss": 1.9433,
+      "step": 1470
+    },
+    {
+      "epoch": 1.714868804664723,
+      "grad_norm": 0.11855728924274445,
+      "learning_rate": 6.229172007177648e-06,
+      "loss": 1.8146,
+      "step": 1471
+    },
+    {
+      "epoch": 1.7160349854227406,
+      "grad_norm": 0.4064196050167084,
+      "learning_rate": 6.217882836587873e-06,
+      "loss": 1.8429,
+      "step": 1472
+    },
+    {
+      "epoch": 1.717201166180758,
+      "grad_norm": 0.17871473729610443,
+      "learning_rate": 6.206541334020087e-06,
+      "loss": 1.8803,
+      "step": 1473
+    },
+    {
+      "epoch": 1.7183673469387755,
+      "grad_norm": 0.10903530567884445,
+      "learning_rate": 6.19514713474445e-06,
+      "loss": 1.767,
+      "step": 1474
+    },
+    {
+      "epoch": 1.7195335276967931,
+      "grad_norm": 0.09271171689033508,
+      "learning_rate": 6.183699870633894e-06,
+      "loss": 1.8082,
+      "step": 1475
+    },
+    {
+      "epoch": 1.7206997084548106,
+      "grad_norm": 0.10191729664802551,
+      "learning_rate": 6.172199170124482e-06,
+      "loss": 1.8726,
+      "step": 1476
+    },
+    {
+      "epoch": 1.721865889212828,
+      "grad_norm": 0.2409660667181015,
+      "learning_rate": 6.160644658175201e-06,
+      "loss": 1.8542,
+      "step": 1477
+    },
+    {
+      "epoch": 1.7230320699708455,
+      "grad_norm": 0.10931321233510971,
+      "learning_rate": 6.149035956227202e-06,
+      "loss": 1.8309,
+      "step": 1478
+    },
+    {
+      "epoch": 1.724198250728863,
+      "grad_norm": 0.1354037970304489,
+      "learning_rate": 6.1373726821624445e-06,
+      "loss": 1.8757,
+      "step": 1479
+    },
+    {
+      "epoch": 1.7253644314868803,
+      "grad_norm": 0.18350885808467865,
+      "learning_rate": 6.125654450261781e-06,
+      "loss": 1.8353,
+      "step": 1480
+    },
+    {
+      "epoch": 1.726530612244898,
+      "grad_norm": 0.23060019314289093,
+      "learning_rate": 6.113880871162424e-06,
+      "loss": 1.8208,
+      "step": 1481
+    },
+    {
+      "epoch": 1.7276967930029155,
+      "grad_norm": 0.1051626056432724,
+      "learning_rate": 6.102051551814835e-06,
+      "loss": 1.861,
+      "step": 1482
+    },
+    {
+      "epoch": 1.728862973760933,
+      "grad_norm": 0.09245733916759491,
+      "learning_rate": 6.0901660954389664e-06,
+      "loss": 1.9221,
+      "step": 1483
+    },
+    {
+      "epoch": 1.7300291545189506,
+      "grad_norm": 0.1456136852502823,
+      "learning_rate": 6.078224101479916e-06,
+      "loss": 1.7845,
+      "step": 1484
+    },
+    {
+      "epoch": 1.731195335276968,
+      "grad_norm": 0.3250628709793091,
+      "learning_rate": 6.0662251655629145e-06,
+      "loss": 1.8909,
+      "step": 1485
+    },
+    {
+      "epoch": 1.7323615160349854,
+      "grad_norm": 0.1504368782043457,
+      "learning_rate": 6.05416887944769e-06,
+      "loss": 1.868,
+      "step": 1486
+    },
+    {
+      "epoch": 1.733527696793003,
+      "grad_norm": 0.08908155560493469,
+      "learning_rate": 6.042054830982167e-06,
+      "loss": 1.7572,
+      "step": 1487
+    },
+    {
+      "epoch": 1.7346938775510203,
+      "grad_norm": 0.3843560814857483,
+      "learning_rate": 6.029882604055497e-06,
+      "loss": 1.8538,
+      "step": 1488
+    },
+    {
+      "epoch": 1.7358600583090378,
+      "grad_norm": 0.09327776730060577,
+      "learning_rate": 6.017651778550414e-06,
+      "loss": 1.8818,
+      "step": 1489
+    },
+    {
+      "epoch": 1.7370262390670554,
+      "grad_norm": 0.09769665449857712,
+      "learning_rate": 6.005361930294906e-06,
+      "loss": 1.9033,
+      "step": 1490
+    },
+    {
+      "epoch": 1.7381924198250729,
+      "grad_norm": 0.09195089340209961,
+      "learning_rate": 5.993012631013169e-06,
+      "loss": 1.8275,
+      "step": 1491
+    },
+    {
+      "epoch": 1.7393586005830903,
+      "grad_norm": 0.09683268517255783,
+      "learning_rate": 5.980603448275862e-06,
+      "loss": 1.892,
+      "step": 1492
+    },
+    {
+      "epoch": 1.740524781341108,
+      "grad_norm": 0.1517692655324936,
+      "learning_rate": 5.968133945449636e-06,
+      "loss": 1.8165,
+      "step": 1493
+    },
+    {
+      "epoch": 1.7416909620991254,
+      "grad_norm": 0.12244709581136703,
+      "learning_rate": 5.955603681645912e-06,
+      "loss": 1.8963,
+      "step": 1494
+    },
+    {
+      "epoch": 1.7428571428571429,
+      "grad_norm": 0.1856079250574112,
+      "learning_rate": 5.943012211668928e-06,
+      "loss": 1.9328,
+      "step": 1495
+    },
+    {
+      "epoch": 1.7440233236151603,
+      "grad_norm": 0.1421637386083603,
+      "learning_rate": 5.930359085963004e-06,
+      "loss": 1.9182,
+      "step": 1496
+    },
+    {
+      "epoch": 1.7451895043731778,
+      "grad_norm": 0.1451270431280136,
+      "learning_rate": 5.9176438505590396e-06,
+      "loss": 1.8566,
+      "step": 1497
+    },
+    {
+      "epoch": 1.7463556851311952,
+      "grad_norm": 0.08269717544317245,
+      "learning_rate": 5.90486604702023e-06,
+      "loss": 1.8783,
+      "step": 1498
+    },
+    {
+      "epoch": 1.7475218658892129,
+      "grad_norm": 0.09546936303377151,
+      "learning_rate": 5.892025212386955e-06,
+      "loss": 1.8732,
+      "step": 1499
+    },
+    {
+      "epoch": 1.7486880466472303,
+      "grad_norm": 0.17464224994182587,
+      "learning_rate": 5.8791208791208794e-06,
+      "loss": 1.8661,
+      "step": 1500
+    },
+    {
+      "epoch": 1.7498542274052478,
+      "grad_norm": 0.10765030980110168,
+      "learning_rate": 5.8661525750481965e-06,
+      "loss": 1.8196,
+      "step": 1501
+    },
+    {
+      "epoch": 1.7510204081632654,
+      "grad_norm": 0.08960511535406113,
+      "learning_rate": 5.853119823302044e-06,
+      "loss": 1.9183,
+      "step": 1502
+    },
+    {
+      "epoch": 1.7521865889212829,
+      "grad_norm": 0.111103355884552,
+      "learning_rate": 5.8400221422640465e-06,
+      "loss": 1.9089,
+      "step": 1503
+    },
+    {
+      "epoch": 1.7533527696793003,
+      "grad_norm": 0.08859553188085556,
+      "learning_rate": 5.826859045504995e-06,
+      "loss": 1.822,
+      "step": 1504
+    },
+    {
+      "epoch": 1.7545189504373178,
+      "grad_norm": 0.12368645519018173,
+      "learning_rate": 5.813630041724619e-06,
+      "loss": 1.8505,
+      "step": 1505
+    },
+    {
+      "epoch": 1.7556851311953352,
+      "grad_norm": 0.26847147941589355,
+      "learning_rate": 5.800334634690463e-06,
+      "loss": 1.7969,
+      "step": 1506
+    },
+    {
+      "epoch": 1.7568513119533526,
+      "grad_norm": 0.1013394221663475,
+      "learning_rate": 5.786972323175846e-06,
+      "loss": 1.8541,
+      "step": 1507
+    },
+    {
+      "epoch": 1.7580174927113703,
+      "grad_norm": 0.0893905982375145,
+      "learning_rate": 5.7735426008968624e-06,
+      "loss": 1.9087,
+      "step": 1508
+    },
+    {
+      "epoch": 1.7591836734693878,
+      "grad_norm": 0.12159135192632675,
+      "learning_rate": 5.76004495644844e-06,
+      "loss": 1.8612,
+      "step": 1509
+    },
+    {
+      "epoch": 1.7603498542274052,
+      "grad_norm": 0.09463372081518173,
+      "learning_rate": 5.7464788732394374e-06,
+      "loss": 1.8422,
+      "step": 1510
+    },
+    {
+      "epoch": 1.7615160349854229,
+      "grad_norm": 0.09465835243463516,
+      "learning_rate": 5.732843829426715e-06,
+      "loss": 1.833,
+      "step": 1511
+    },
+    {
+      "epoch": 1.7626822157434403,
+      "grad_norm": 0.08529729396104813,
+      "learning_rate": 5.719139297848244e-06,
+      "loss": 1.8544,
+      "step": 1512
+    },
+    {
+      "epoch": 1.7638483965014577,
+      "grad_norm": 0.14807681739330292,
+      "learning_rate": 5.705364745955152e-06,
+      "loss": 1.8098,
+      "step": 1513
+    },
+    {
+      "epoch": 1.7650145772594752,
+      "grad_norm": 0.09294959157705307,
+      "learning_rate": 5.6915196357427435e-06,
+      "loss": 1.8886,
+      "step": 1514
+    },
+    {
+      "epoch": 1.7661807580174926,
+      "grad_norm": 0.08552692085504532,
+      "learning_rate": 5.677603423680457e-06,
+      "loss": 1.8259,
+      "step": 1515
+    },
+    {
+      "epoch": 1.76734693877551,
+      "grad_norm": 0.43894219398498535,
+      "learning_rate": 5.663615560640733e-06,
+      "loss": 1.8639,
+      "step": 1516
+    },
+    {
+      "epoch": 1.7685131195335277,
+      "grad_norm": 0.11018567532300949,
+      "learning_rate": 5.649555491826785e-06,
+      "loss": 1.8445,
+      "step": 1517
+    },
+    {
+      "epoch": 1.7696793002915452,
+      "grad_norm": 0.10608147829771042,
+      "learning_rate": 5.635422656699253e-06,
+      "loss": 1.8695,
+      "step": 1518
+    },
+    {
+      "epoch": 1.7708454810495626,
+      "grad_norm": 0.09701573103666306,
+      "learning_rate": 5.621216488901701e-06,
+      "loss": 1.8821,
+      "step": 1519
+    },
+    {
+      "epoch": 1.7720116618075803,
+      "grad_norm": 0.11868813633918762,
+      "learning_rate": 5.606936416184971e-06,
+      "loss": 1.7602,
+      "step": 1520
+    },
+    {
+      "epoch": 1.7731778425655977,
+      "grad_norm": 0.14249104261398315,
+      "learning_rate": 5.59258186033034e-06,
+      "loss": 1.9038,
+      "step": 1521
+    },
+    {
+      "epoch": 1.7743440233236152,
+      "grad_norm": 0.14230656623840332,
+      "learning_rate": 5.578152237071471e-06,
+      "loss": 1.8726,
+      "step": 1522
+    },
+    {
+      "epoch": 1.7755102040816326,
+      "grad_norm": 0.10407228022813797,
+      "learning_rate": 5.5636469560151475e-06,
+      "loss": 1.7816,
+      "step": 1523
+    },
+    {
+      "epoch": 1.77667638483965,
+      "grad_norm": 0.09664175659418106,
+      "learning_rate": 5.549065420560749e-06,
+      "loss": 1.848,
+      "step": 1524
+    },
+    {
+      "epoch": 1.7778425655976675,
+      "grad_norm": 0.11642491817474365,
+      "learning_rate": 5.534407027818448e-06,
+      "loss": 1.9201,
+      "step": 1525
+    },
+    {
+      "epoch": 1.7790087463556852,
+      "grad_norm": 0.1527661383152008,
+      "learning_rate": 5.5196711685261314e-06,
+      "loss": 1.9209,
+      "step": 1526
+    },
+    {
+      "epoch": 1.7801749271137026,
+      "grad_norm": 0.13195447623729706,
+      "learning_rate": 5.504857226964969e-06,
+      "loss": 1.8038,
+      "step": 1527
+    },
+    {
+      "epoch": 1.78134110787172,
+      "grad_norm": 0.0971655324101448,
+      "learning_rate": 5.489964580873672e-06,
+      "loss": 1.8398,
+      "step": 1528
+    },
+    {
+      "epoch": 1.7825072886297377,
+      "grad_norm": 0.11325421184301376,
+      "learning_rate": 5.47499260136135e-06,
+      "loss": 1.8474,
+      "step": 1529
+    },
+    {
+      "epoch": 1.7836734693877552,
+      "grad_norm": 0.15892252326011658,
+      "learning_rate": 5.459940652818992e-06,
+      "loss": 1.8772,
+      "step": 1530
+    },
+    {
+      "epoch": 1.7848396501457726,
+      "grad_norm": 0.1889907568693161,
+      "learning_rate": 5.4448080928295155e-06,
+      "loss": 1.8883,
+      "step": 1531
+    },
+    {
+      "epoch": 1.78600583090379,
+      "grad_norm": 0.1678425371646881,
+      "learning_rate": 5.429594272076373e-06,
+      "loss": 1.8593,
+      "step": 1532
+    },
+    {
+      "epoch": 1.7871720116618075,
+      "grad_norm": 0.14115029573440552,
+      "learning_rate": 5.4142985342506734e-06,
+      "loss": 1.9436,
+      "step": 1533
+    },
+    {
+      "epoch": 1.788338192419825,
+      "grad_norm": 0.12288099527359009,
+      "learning_rate": 5.398920215956809e-06,
+      "loss": 1.8485,
+      "step": 1534
+    },
+    {
+      "epoch": 1.7895043731778426,
+      "grad_norm": 0.09795211255550385,
+      "learning_rate": 5.38345864661654e-06,
+      "loss": 1.8342,
+      "step": 1535
+    },
+    {
+      "epoch": 1.79067055393586,
+      "grad_norm": 0.17868438363075256,
+      "learning_rate": 5.367913148371532e-06,
+      "loss": 1.8864,
+      "step": 1536
+    },
+    {
+      "epoch": 1.7918367346938775,
+      "grad_norm": 0.2338380366563797,
+      "learning_rate": 5.352283035984276e-06,
+      "loss": 1.8529,
+      "step": 1537
+    },
+    {
+      "epoch": 1.7930029154518952,
+      "grad_norm": 0.09090147912502289,
+      "learning_rate": 5.336567616737417e-06,
+      "loss": 1.8763,
+      "step": 1538
+    },
+    {
+      "epoch": 1.7941690962099126,
+      "grad_norm": 0.2063981145620346,
+      "learning_rate": 5.320766190331408e-06,
+      "loss": 1.8385,
+      "step": 1539
+    },
+    {
+      "epoch": 1.79533527696793,
+      "grad_norm": 0.19731548428535461,
+      "learning_rate": 5.304878048780487e-06,
+      "loss": 1.7543,
+      "step": 1540
+    },
+    {
+      "epoch": 1.7965014577259475,
+      "grad_norm": 0.09930050373077393,
+      "learning_rate": 5.2889024763069406e-06,
+      "loss": 1.8934,
+      "step": 1541
+    },
+    {
+      "epoch": 1.797667638483965,
+      "grad_norm": 0.3272360861301422,
+      "learning_rate": 5.2728387492335986e-06,
+      "loss": 1.7964,
+      "step": 1542
+    },
+    {
+      "epoch": 1.7988338192419824,
+      "grad_norm": 0.14644372463226318,
+      "learning_rate": 5.256686135874578e-06,
+      "loss": 1.9061,
+      "step": 1543
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.09581572562456131,
+      "learning_rate": 5.240443896424168e-06,
+      "loss": 1.8824,
+      "step": 1544
+    },
+    {
+      "epoch": 1.8011661807580175,
+      "grad_norm": 0.08885791897773743,
+      "learning_rate": 5.224111282843894e-06,
+      "loss": 1.7989,
+      "step": 1545
+    },
+    {
+      "epoch": 1.802332361516035,
+      "grad_norm": 0.23225639760494232,
+      "learning_rate": 5.207687538747676e-06,
+      "loss": 1.9295,
+      "step": 1546
+    },
+    {
+      "epoch": 1.8034985422740526,
+      "grad_norm": 0.10454729944467545,
+      "learning_rate": 5.191171899285049e-06,
+      "loss": 1.8367,
+      "step": 1547
+    },
+    {
+      "epoch": 1.80466472303207,
+      "grad_norm": 0.12204279005527496,
+      "learning_rate": 5.174563591022445e-06,
+      "loss": 1.8332,
+      "step": 1548
+    },
+    {
+      "epoch": 1.80466472303207,
+      "eval_loss": 0.9097900390625,
+      "eval_runtime": 167.0333,
+      "eval_samples_per_second": 41.333,
+      "eval_steps_per_second": 2.586,
+      "step": 1548
+    },
+    {
+      "epoch": 1.8058309037900875,
+      "grad_norm": 0.10996700078248978,
+      "learning_rate": 5.157861831822445e-06,
+      "loss": 1.931,
+      "step": 1549
+    },
+    {
+      "epoch": 1.806997084548105,
+      "grad_norm": 0.1849961131811142,
+      "learning_rate": 5.141065830721003e-06,
+      "loss": 1.7672,
+      "step": 1550
+    },
+    {
+      "epoch": 1.8081632653061224,
+      "grad_norm": 0.09252898395061493,
+      "learning_rate": 5.124174787802578e-06,
+      "loss": 1.8166,
+      "step": 1551
+    },
+    {
+      "epoch": 1.8093294460641398,
+      "grad_norm": 0.14785408973693848,
+      "learning_rate": 5.10718789407314e-06,
+      "loss": 1.8737,
+      "step": 1552
+    },
+    {
+      "epoch": 1.8104956268221575,
+      "grad_norm": 0.08367530256509781,
+      "learning_rate": 5.090104331331015e-06,
+      "loss": 1.8392,
+      "step": 1553
+    },
+    {
+      "epoch": 1.811661807580175,
+      "grad_norm": 0.11494553089141846,
+      "learning_rate": 5.072923272035512e-06,
+      "loss": 1.8919,
+      "step": 1554
+    },
+    {
+      "epoch": 1.8128279883381926,
+      "grad_norm": 0.1934298872947693,
+      "learning_rate": 5.055643879173291e-06,
+      "loss": 1.7618,
+      "step": 1555
+    },
+    {
+      "epoch": 1.81399416909621,
+      "grad_norm": 0.10867110639810562,
+      "learning_rate": 5.0382653061224495e-06,
+      "loss": 1.8777,
+      "step": 1556
+    },
+    {
+      "epoch": 1.8151603498542275,
+      "grad_norm": 0.10382754355669022,
+      "learning_rate": 5.0207866965142315e-06,
+      "loss": 1.8079,
+      "step": 1557
+    },
+    {
+      "epoch": 1.816326530612245,
+      "grad_norm": 0.3039425015449524,
+      "learning_rate": 5.003207184092367e-06,
+      "loss": 1.8261,
+      "step": 1558
+    },
+    {
+      "epoch": 1.8174927113702624,
+      "grad_norm": 0.10122989118099213,
+      "learning_rate": 4.985525892569959e-06,
+      "loss": 1.8527,
+      "step": 1559
+    },
+    {
+      "epoch": 1.8186588921282798,
+      "grad_norm": 0.08252281695604324,
+      "learning_rate": 4.967741935483871e-06,
+      "loss": 1.8056,
+      "step": 1560
+    },
+    {
+      "epoch": 1.8198250728862972,
+      "grad_norm": 0.11565924435853958,
+      "learning_rate": 4.9498544160465864e-06,
+      "loss": 1.8514,
+      "step": 1561
+    },
+    {
+      "epoch": 1.820991253644315,
+      "grad_norm": 0.15656670928001404,
+      "learning_rate": 4.931862426995458e-06,
+      "loss": 1.8729,
+      "step": 1562
+    },
+    {
+      "epoch": 1.8221574344023324,
+      "grad_norm": 0.13529087603092194,
+      "learning_rate": 4.913765050439311e-06,
+      "loss": 1.9334,
+      "step": 1563
+    },
+    {
+      "epoch": 1.82332361516035,
+      "grad_norm": 0.12258682399988174,
+      "learning_rate": 4.895561357702351e-06,
+      "loss": 1.8579,
+      "step": 1564
+    },
+    {
+      "epoch": 1.8244897959183675,
+      "grad_norm": 0.15220880508422852,
+      "learning_rate": 4.877250409165303e-06,
+      "loss": 1.8735,
+      "step": 1565
+    },
+    {
+      "epoch": 1.825655976676385,
+      "grad_norm": 0.10631298273801804,
+      "learning_rate": 4.858831254103743e-06,
+      "loss": 1.8371,
+      "step": 1566
+    },
+    {
+      "epoch": 1.8268221574344023,
+      "grad_norm": 0.24128419160842896,
+      "learning_rate": 4.840302930523543e-06,
+      "loss": 1.8664,
+      "step": 1567
+    },
+    {
+      "epoch": 1.8279883381924198,
+      "grad_norm": 0.09952948987483978,
+      "learning_rate": 4.821664464993395e-06,
+      "loss": 1.8475,
+      "step": 1568
+    },
+    {
+      "epoch": 1.8291545189504372,
+      "grad_norm": 0.09560183435678482,
+      "learning_rate": 4.80291487247433e-06,
+      "loss": 1.9031,
+      "step": 1569
+    },
+    {
+      "epoch": 1.8303206997084547,
+      "grad_norm": 0.08972621709108353,
+      "learning_rate": 4.78405315614618e-06,
+      "loss": 1.8419,
+      "step": 1570
+    },
+    {
+      "epoch": 1.8314868804664723,
+      "grad_norm": 0.12510262429714203,
+      "learning_rate": 4.765078307230923e-06,
+      "loss": 1.8319,
+      "step": 1571
+    },
+    {
+      "epoch": 1.8326530612244898,
+      "grad_norm": 0.10316437482833862,
+      "learning_rate": 4.745989304812835e-06,
+      "loss": 1.883,
+      "step": 1572
+    },
+    {
+      "epoch": 1.8338192419825075,
+      "grad_norm": 0.4328541159629822,
+      "learning_rate": 4.72678511565538e-06,
+      "loss": 1.8278,
+      "step": 1573
+    },
+    {
+      "epoch": 1.834985422740525,
+      "grad_norm": 0.2721095383167267,
+      "learning_rate": 4.707464694014795e-06,
+      "loss": 1.7546,
+      "step": 1574
+    },
+    {
+      "epoch": 1.8361516034985423,
+      "grad_norm": 0.16511045396327972,
+      "learning_rate": 4.688026981450253e-06,
+      "loss": 1.877,
+      "step": 1575
+    },
+    {
+      "epoch": 1.8373177842565598,
+      "grad_norm": 0.12194500863552094,
+      "learning_rate": 4.668470906630582e-06,
+      "loss": 1.9055,
+      "step": 1576
+    },
+    {
+      "epoch": 1.8384839650145772,
+      "grad_norm": 0.17013604938983917,
+      "learning_rate": 4.6487953851374285e-06,
+      "loss": 1.9098,
+      "step": 1577
+    },
+    {
+      "epoch": 1.8396501457725947,
+      "grad_norm": 0.10560650378465652,
+      "learning_rate": 4.628999319264806e-06,
+      "loss": 1.9009,
+      "step": 1578
+    },
+    {
+      "epoch": 1.8408163265306121,
+      "grad_norm": 0.11002885550260544,
+      "learning_rate": 4.609081597814954e-06,
+      "loss": 1.8348,
+      "step": 1579
+    },
+    {
+      "epoch": 1.8419825072886298,
+      "grad_norm": 0.18719035387039185,
+      "learning_rate": 4.589041095890411e-06,
+      "loss": 1.8763,
+      "step": 1580
+    },
+    {
+      "epoch": 1.8431486880466472,
+      "grad_norm": 0.1343841850757599,
+      "learning_rate": 4.56887667468224e-06,
+      "loss": 1.8767,
+      "step": 1581
+    },
+    {
+      "epoch": 1.8443148688046649,
+      "grad_norm": 0.09223797172307968,
+      "learning_rate": 4.548587181254308e-06,
+      "loss": 1.8141,
+      "step": 1582
+    },
+    {
+      "epoch": 1.8454810495626823,
+      "grad_norm": 0.22785449028015137,
+      "learning_rate": 4.528171448323539e-06,
+      "loss": 1.857,
+      "step": 1583
+    },
+    {
+      "epoch": 1.8466472303206998,
+      "grad_norm": 0.10066588222980499,
+      "learning_rate": 4.507628294036062e-06,
+      "loss": 1.8049,
+      "step": 1584
+    },
+    {
+      "epoch": 1.8478134110787172,
+      "grad_norm": 0.14481768012046814,
+      "learning_rate": 4.486956521739131e-06,
+      "loss": 1.877,
+      "step": 1585
+    },
+    {
+      "epoch": 1.8489795918367347,
+      "grad_norm": 0.3311879336833954,
+      "learning_rate": 4.466154919748779e-06,
+      "loss": 1.816,
+      "step": 1586
+    },
+    {
+      "epoch": 1.850145772594752,
+      "grad_norm": 0.09457577019929886,
+      "learning_rate": 4.4452222611130564e-06,
+      "loss": 1.882,
+      "step": 1587
+    },
+    {
+      "epoch": 1.8513119533527695,
+      "grad_norm": 0.29340437054634094,
+      "learning_rate": 4.424157303370787e-06,
+      "loss": 1.7993,
+      "step": 1588
+    },
+    {
+      "epoch": 1.8524781341107872,
+      "grad_norm": 0.09214404225349426,
+      "learning_rate": 4.402958788305742e-06,
+      "loss": 1.8524,
+      "step": 1589
+    },
+    {
+      "epoch": 1.8536443148688047,
+      "grad_norm": 0.10155724734067917,
+      "learning_rate": 4.381625441696113e-06,
+      "loss": 1.8311,
+      "step": 1590
+    },
+    {
+      "epoch": 1.8548104956268223,
+      "grad_norm": 0.1542295664548874,
+      "learning_rate": 4.360155973059199e-06,
+      "loss": 1.8785,
+      "step": 1591
+    },
+    {
+      "epoch": 1.8559766763848398,
+      "grad_norm": 0.10311226546764374,
+      "learning_rate": 4.3385490753911804e-06,
+      "loss": 1.9154,
+      "step": 1592
+    },
+    {
+      "epoch": 1.8571428571428572,
+      "grad_norm": 0.09403152018785477,
+      "learning_rate": 4.316803424901891e-06,
+      "loss": 1.7946,
+      "step": 1593
+    },
+    {
+      "epoch": 1.8583090379008746,
+      "grad_norm": 0.08991537243127823,
+      "learning_rate": 4.294917680744453e-06,
+      "loss": 1.7892,
+      "step": 1594
+    },
+    {
+      "epoch": 1.859475218658892,
+      "grad_norm": 0.12465853244066238,
+      "learning_rate": 4.2728904847396774e-06,
+      "loss": 1.8818,
+      "step": 1595
+    },
+    {
+      "epoch": 1.8606413994169095,
+      "grad_norm": 0.18682657182216644,
+      "learning_rate": 4.2507204610951e-06,
+      "loss": 1.8763,
+      "step": 1596
+    },
+    {
+      "epoch": 1.861807580174927,
+      "grad_norm": 0.20600618422031403,
+      "learning_rate": 4.22840621611854e-06,
+      "loss": 1.8392,
+      "step": 1597
+    },
+    {
+      "epoch": 1.8629737609329446,
+      "grad_norm": 0.08709615468978882,
+      "learning_rate": 4.205946337926034e-06,
+      "loss": 1.8436,
+      "step": 1598
+    },
+    {
+      "epoch": 1.864139941690962,
+      "grad_norm": 0.1042516753077507,
+      "learning_rate": 4.183339396144053e-06,
+      "loss": 1.9096,
+      "step": 1599
+    },
+    {
+      "epoch": 1.8653061224489798,
+      "grad_norm": 0.21963785588741302,
+      "learning_rate": 4.16058394160584e-06,
+      "loss": 1.8188,
+      "step": 1600
+    },
+    {
+      "epoch": 1.8664723032069972,
+      "grad_norm": 0.09787870943546295,
+      "learning_rate": 4.137678506041743e-06,
+      "loss": 1.8825,
+      "step": 1601
+    },
+    {
+      "epoch": 1.8676384839650146,
+      "grad_norm": 0.15103669464588165,
+      "learning_rate": 4.114621601763409e-06,
+      "loss": 1.8644,
+      "step": 1602
+    },
+    {
+      "epoch": 1.868804664723032,
+      "grad_norm": 0.12370051443576813,
+      "learning_rate": 4.091411721341689e-06,
+      "loss": 1.8243,
+      "step": 1603
+    },
+    {
+      "epoch": 1.8699708454810495,
+      "grad_norm": 0.10909929871559143,
+      "learning_rate": 4.068047337278106e-06,
+      "loss": 1.8995,
+      "step": 1604
+    },
+    {
+      "epoch": 1.871137026239067,
+      "grad_norm": 0.09355130046606064,
+      "learning_rate": 4.044526901669759e-06,
+      "loss": 1.9373,
+      "step": 1605
+    },
+    {
+      "epoch": 1.8723032069970844,
+      "grad_norm": 0.10362165421247482,
+      "learning_rate": 4.020848845867461e-06,
+      "loss": 1.8182,
+      "step": 1606
+    },
+    {
+      "epoch": 1.873469387755102,
+      "grad_norm": 0.11886598914861679,
+      "learning_rate": 3.997011580127008e-06,
+      "loss": 1.8325,
+      "step": 1607
+    },
+    {
+      "epoch": 1.8746355685131195,
+      "grad_norm": 0.08759265393018723,
+      "learning_rate": 3.9730134932533734e-06,
+      "loss": 1.8042,
+      "step": 1608
+    },
+    {
+      "epoch": 1.8758017492711372,
+      "grad_norm": 0.11225347220897675,
+      "learning_rate": 3.948852952237684e-06,
+      "loss": 1.931,
+      "step": 1609
+    },
+    {
+      "epoch": 1.8769679300291546,
+      "grad_norm": 0.14762015640735626,
+      "learning_rate": 3.924528301886793e-06,
+      "loss": 1.8636,
+      "step": 1610
+    },
+    {
+      "epoch": 1.878134110787172,
+      "grad_norm": 0.15451020002365112,
+      "learning_rate": 3.900037864445286e-06,
+      "loss": 1.8852,
+      "step": 1611
+    },
+    {
+      "epoch": 1.8793002915451895,
+      "grad_norm": 0.13228684663772583,
+      "learning_rate": 3.875379939209727e-06,
+      "loss": 1.8321,
+      "step": 1612
+    },
+    {
+      "epoch": 1.880466472303207,
+      "grad_norm": 0.12841254472732544,
+      "learning_rate": 3.85055280213496e-06,
+      "loss": 1.8253,
+      "step": 1613
+    },
+    {
+      "epoch": 1.8816326530612244,
+      "grad_norm": 0.11803828924894333,
+      "learning_rate": 3.825554705432287e-06,
+      "loss": 1.8684,
+      "step": 1614
+    },
+    {
+      "epoch": 1.8827988338192418,
+      "grad_norm": 0.271273672580719,
+      "learning_rate": 3.8003838771593086e-06,
+      "loss": 1.8757,
+      "step": 1615
+    },
+    {
+      "epoch": 1.8839650145772595,
+      "grad_norm": 0.09479683637619019,
+      "learning_rate": 3.775038520801233e-06,
+      "loss": 1.8785,
+      "step": 1616
+    },
+    {
+      "epoch": 1.885131195335277,
+      "grad_norm": 0.0999586209654808,
+      "learning_rate": 3.7495168148434484e-06,
+      "loss": 1.7938,
+      "step": 1617
+    },
+    {
+      "epoch": 1.8862973760932946,
+      "grad_norm": 0.48679253458976746,
+      "learning_rate": 3.723816912335144e-06,
+      "loss": 1.9077,
+      "step": 1618
+    },
+    {
+      "epoch": 1.887463556851312,
+      "grad_norm": 0.15930908918380737,
+      "learning_rate": 3.6979369404437525e-06,
+      "loss": 1.8205,
+      "step": 1619
+    },
+    {
+      "epoch": 1.8886297376093295,
+      "grad_norm": 0.2391028255224228,
+      "learning_rate": 3.6718750000000003e-06,
+      "loss": 1.8757,
+      "step": 1620
+    },
+    {
+      "epoch": 1.889795918367347,
+      "grad_norm": 0.15171706676483154,
+      "learning_rate": 3.6456291650333203e-06,
+      "loss": 1.8107,
+      "step": 1621
+    },
+    {
+      "epoch": 1.8909620991253644,
+      "grad_norm": 0.0930434912443161,
+      "learning_rate": 3.6191974822974043e-06,
+      "loss": 1.8517,
+      "step": 1622
+    },
+    {
+      "epoch": 1.8921282798833818,
+      "grad_norm": 0.10419802367687225,
+      "learning_rate": 3.59257797078563e-06,
+      "loss": 1.8041,
+      "step": 1623
+    },
+    {
+      "epoch": 1.8932944606413993,
+      "grad_norm": 0.09357330203056335,
+      "learning_rate": 3.5657686212361327e-06,
+      "loss": 1.9499,
+      "step": 1624
+    },
+    {
+      "epoch": 1.894460641399417,
+      "grad_norm": 0.28163260221481323,
+      "learning_rate": 3.5387673956262427e-06,
+      "loss": 1.7839,
+      "step": 1625
+    },
+    {
+      "epoch": 1.8956268221574344,
+      "grad_norm": 0.10682091116905212,
+      "learning_rate": 3.5115722266560256e-06,
+      "loss": 1.8626,
+      "step": 1626
+    },
+    {
+      "epoch": 1.896793002915452,
+      "grad_norm": 0.19390399754047394,
+      "learning_rate": 3.484181017220665e-06,
+      "loss": 1.8502,
+      "step": 1627
+    },
+    {
+      "epoch": 1.8979591836734695,
+      "grad_norm": 0.14990460872650146,
+      "learning_rate": 3.456591639871383e-06,
+      "loss": 1.8867,
+      "step": 1628
+    },
+    {
+      "epoch": 1.899125364431487,
+      "grad_norm": 0.08680065721273422,
+      "learning_rate": 3.4288019362646223e-06,
+      "loss": 1.8566,
+      "step": 1629
+    },
+    {
+      "epoch": 1.9002915451895044,
+      "grad_norm": 0.23905427753925323,
+      "learning_rate": 3.4008097165991905e-06,
+      "loss": 1.8616,
+      "step": 1630
+    },
+    {
+      "epoch": 1.9014577259475218,
+      "grad_norm": 0.10342249274253845,
+      "learning_rate": 3.37261275904104e-06,
+      "loss": 1.8958,
+      "step": 1631
+    },
+    {
+      "epoch": 1.9026239067055393,
+      "grad_norm": 0.08906349539756775,
+      "learning_rate": 3.3442088091353997e-06,
+      "loss": 1.8441,
+      "step": 1632
+    },
+    {
+      "epoch": 1.9037900874635567,
+      "grad_norm": 0.29118141531944275,
+      "learning_rate": 3.315595579205895e-06,
+      "loss": 1.8015,
+      "step": 1633
+    },
+    {
+      "epoch": 1.9049562682215744,
+      "grad_norm": 0.2689427137374878,
+      "learning_rate": 3.286770747740345e-06,
+      "loss": 1.8848,
+      "step": 1634
+    },
+    {
+      "epoch": 1.9049562682215744,
+      "eval_loss": 0.9093718528747559,
+      "eval_runtime": 168.9891,
+      "eval_samples_per_second": 40.855,
+      "eval_steps_per_second": 2.556,
+      "step": 1634
+    },
+    {
+      "epoch": 1.9061224489795918,
+      "grad_norm": 0.23355530202388763,
+      "learning_rate": 3.2577319587628864e-06,
+      "loss": 1.8601,
+      "step": 1635
+    },
+    {
+      "epoch": 1.9072886297376095,
+      "grad_norm": 0.16404344141483307,
+      "learning_rate": 3.228476821192053e-06,
+      "loss": 1.8488,
+      "step": 1636
+    },
+    {
+      "epoch": 1.908454810495627,
+      "grad_norm": 0.09277043491601944,
+      "learning_rate": 3.1990029081844615e-06,
+      "loss": 1.9001,
+      "step": 1637
+    },
+    {
+      "epoch": 1.9096209912536444,
+      "grad_norm": 0.1049395352602005,
+      "learning_rate": 3.1693077564637205e-06,
+      "loss": 1.8365,
+      "step": 1638
+    },
+    {
+      "epoch": 1.9107871720116618,
+      "grad_norm": 0.2586573660373688,
+      "learning_rate": 3.1393888656341563e-06,
+      "loss": 1.8968,
+      "step": 1639
+    },
+    {
+      "epoch": 1.9119533527696793,
+      "grad_norm": 0.10180753469467163,
+      "learning_rate": 3.109243697478992e-06,
+      "loss": 1.9178,
+      "step": 1640
+    },
+    {
+      "epoch": 1.9131195335276967,
+      "grad_norm": 0.14484284818172455,
+      "learning_rate": 3.078869675242514e-06,
+      "loss": 1.8565,
+      "step": 1641
+    },
+    {
+      "epoch": 1.9142857142857141,
+      "grad_norm": 0.12210977077484131,
+      "learning_rate": 3.048264182895851e-06,
+      "loss": 1.861,
+      "step": 1642
+    },
+    {
+      "epoch": 1.9154518950437318,
+      "grad_norm": 0.1487254500389099,
+      "learning_rate": 3.0174245643858903e-06,
+      "loss": 1.9547,
+      "step": 1643
+    },
+    {
+      "epoch": 1.9166180758017493,
+      "grad_norm": 0.09462174028158188,
+      "learning_rate": 2.9863481228668946e-06,
+      "loss": 1.8315,
+      "step": 1644
+    },
+    {
+      "epoch": 1.917784256559767,
+      "grad_norm": 0.14547698199748993,
+      "learning_rate": 2.9550321199143473e-06,
+      "loss": 1.8965,
+      "step": 1645
+    },
+    {
+      "epoch": 1.9189504373177844,
+      "grad_norm": 0.12536869943141937,
+      "learning_rate": 2.923473774720551e-06,
+      "loss": 1.8847,
+      "step": 1646
+    },
+    {
+      "epoch": 1.9201166180758018,
+      "grad_norm": 0.25210854411125183,
+      "learning_rate": 2.8916702632714715e-06,
+      "loss": 1.8711,
+      "step": 1647
+    },
+    {
+      "epoch": 1.9212827988338192,
+      "grad_norm": 0.09865819662809372,
+      "learning_rate": 2.859618717504333e-06,
+      "loss": 1.8874,
+      "step": 1648
+    },
+    {
+      "epoch": 1.9224489795918367,
+      "grad_norm": 0.15940511226654053,
+      "learning_rate": 2.827316224445412e-06,
+      "loss": 1.8603,
+      "step": 1649
+    },
+    {
+      "epoch": 1.9236151603498541,
+      "grad_norm": 0.08476684987545013,
+      "learning_rate": 2.794759825327511e-06,
+      "loss": 1.8575,
+      "step": 1650
+    },
+    {
+      "epoch": 1.9247813411078716,
+      "grad_norm": 0.09040199965238571,
+      "learning_rate": 2.761946514686541e-06,
+      "loss": 1.8286,
+      "step": 1651
+    },
+    {
+      "epoch": 1.9259475218658892,
+      "grad_norm": 0.16521970927715302,
+      "learning_rate": 2.7288732394366195e-06,
+      "loss": 1.9238,
+      "step": 1652
+    },
+    {
+      "epoch": 1.9271137026239067,
+      "grad_norm": 0.11156275123357773,
+      "learning_rate": 2.6955368979231107e-06,
+      "loss": 1.8889,
+      "step": 1653
+    },
+    {
+      "epoch": 1.9282798833819244,
+      "grad_norm": 0.2201853096485138,
+      "learning_rate": 2.6619343389529723e-06,
+      "loss": 1.8416,
+      "step": 1654
+    },
+    {
+      "epoch": 1.9294460641399418,
+      "grad_norm": 0.10887926816940308,
+      "learning_rate": 2.628062360801782e-06,
+      "loss": 1.8247,
+      "step": 1655
+    },
+    {
+      "epoch": 1.9306122448979592,
+      "grad_norm": 0.09681367129087448,
+      "learning_rate": 2.59391771019678e-06,
+      "loss": 1.9111,
+      "step": 1656
+    },
+    {
+      "epoch": 1.9317784256559767,
+      "grad_norm": 0.2435135692358017,
+      "learning_rate": 2.559497081275258e-06,
+      "loss": 1.9515,
+      "step": 1657
+    },
+    {
+      "epoch": 1.9329446064139941,
+      "grad_norm": 0.3333396315574646,
+      "learning_rate": 2.524797114517583e-06,
+      "loss": 1.8547,
+      "step": 1658
+    },
+    {
+      "epoch": 1.9341107871720116,
+      "grad_norm": 0.20642785727977753,
+      "learning_rate": 2.4898143956541425e-06,
+      "loss": 1.8544,
+      "step": 1659
+    },
+    {
+      "epoch": 1.935276967930029,
+      "grad_norm": 0.0856885090470314,
+      "learning_rate": 2.454545454545455e-06,
+      "loss": 1.8447,
+      "step": 1660
+    },
+    {
+      "epoch": 1.9364431486880467,
+      "grad_norm": 0.10381123423576355,
+      "learning_rate": 2.4189867640346876e-06,
+      "loss": 1.8857,
+      "step": 1661
+    },
+    {
+      "epoch": 1.9376093294460641,
+      "grad_norm": 0.10432425141334534,
+      "learning_rate": 2.383134738771769e-06,
+      "loss": 1.923,
+      "step": 1662
+    },
+    {
+      "epoch": 1.9387755102040818,
+      "grad_norm": 0.12674321234226227,
+      "learning_rate": 2.3469857340082834e-06,
+      "loss": 1.952,
+      "step": 1663
+    },
+    {
+      "epoch": 1.9399416909620992,
+      "grad_norm": 0.09296005219221115,
+      "learning_rate": 2.3105360443622925e-06,
+      "loss": 1.8114,
+      "step": 1664
+    },
+    {
+      "epoch": 1.9411078717201167,
+      "grad_norm": 0.14433319866657257,
+      "learning_rate": 2.273781902552204e-06,
+      "loss": 1.8445,
+      "step": 1665
+    },
+    {
+      "epoch": 1.9422740524781341,
+      "grad_norm": 0.17025868594646454,
+      "learning_rate": 2.2367194780987884e-06,
+      "loss": 1.8903,
+      "step": 1666
+    },
+    {
+      "epoch": 1.9434402332361516,
+      "grad_norm": 0.11757093667984009,
+      "learning_rate": 2.1993448759943846e-06,
+      "loss": 1.8638,
+      "step": 1667
+    },
+    {
+      "epoch": 1.944606413994169,
+      "grad_norm": 0.2709057927131653,
+      "learning_rate": 2.161654135338346e-06,
+      "loss": 1.8767,
+      "step": 1668
+    },
+    {
+      "epoch": 1.9457725947521864,
+      "grad_norm": 0.14542120695114136,
+      "learning_rate": 2.1236432279377066e-06,
+      "loss": 1.8737,
+      "step": 1669
+    },
+    {
+      "epoch": 1.9469387755102041,
+      "grad_norm": 0.10316893458366394,
+      "learning_rate": 2.085308056872038e-06,
+      "loss": 1.8231,
+      "step": 1670
+    },
+    {
+      "epoch": 1.9481049562682216,
+      "grad_norm": 0.1628752052783966,
+      "learning_rate": 2.0466444550214185e-06,
+      "loss": 1.9143,
+      "step": 1671
+    },
+    {
+      "epoch": 1.9492711370262392,
+      "grad_norm": 0.12916240096092224,
+      "learning_rate": 2.0076481835564054e-06,
+      "loss": 1.8494,
+      "step": 1672
+    },
+    {
+      "epoch": 1.9504373177842567,
+      "grad_norm": 0.09298492968082428,
+      "learning_rate": 1.9683149303888623e-06,
+      "loss": 1.8515,
+      "step": 1673
+    },
+    {
+      "epoch": 1.951603498542274,
+      "grad_norm": 0.2267657071352005,
+      "learning_rate": 1.9286403085824492e-06,
+      "loss": 1.9007,
+      "step": 1674
+    },
+    {
+      "epoch": 1.9527696793002915,
+      "grad_norm": 0.14401975274085999,
+      "learning_rate": 1.8886198547215497e-06,
+      "loss": 1.7951,
+      "step": 1675
+    },
+    {
+      "epoch": 1.953935860058309,
+      "grad_norm": 0.11148157715797424,
+      "learning_rate": 1.848249027237354e-06,
+      "loss": 1.8219,
+      "step": 1676
+    },
+    {
+      "epoch": 1.9551020408163264,
+      "grad_norm": 0.10503887385129929,
+      "learning_rate": 1.8075232046897901e-06,
+      "loss": 1.8909,
+      "step": 1677
+    },
+    {
+      "epoch": 1.9562682215743439,
+      "grad_norm": 0.2999081015586853,
+      "learning_rate": 1.7664376840039258e-06,
+      "loss": 1.8159,
+      "step": 1678
+    },
+    {
+      "epoch": 1.9574344023323615,
+      "grad_norm": 0.1257830560207367,
+      "learning_rate": 1.7249876786594382e-06,
+      "loss": 1.8499,
+      "step": 1679
+    },
+    {
+      "epoch": 1.958600583090379,
+      "grad_norm": 0.0965367779135704,
+      "learning_rate": 1.683168316831683e-06,
+      "loss": 1.8903,
+      "step": 1680
+    },
+    {
+      "epoch": 1.9597667638483967,
+      "grad_norm": 0.13980472087860107,
+      "learning_rate": 1.6409746394828444e-06,
+      "loss": 1.9279,
+      "step": 1681
+    },
+    {
+      "epoch": 1.960932944606414,
+      "grad_norm": 0.13928864896297455,
+      "learning_rate": 1.5984015984015983e-06,
+      "loss": 1.826,
+      "step": 1682
+    },
+    {
+      "epoch": 1.9620991253644315,
+      "grad_norm": 0.1054939329624176,
+      "learning_rate": 1.555444054189664e-06,
+      "loss": 1.8509,
+      "step": 1683
+    },
+    {
+      "epoch": 1.963265306122449,
+      "grad_norm": 0.08890262991189957,
+      "learning_rate": 1.5120967741935483e-06,
+      "loss": 1.7839,
+      "step": 1684
+    },
+    {
+      "epoch": 1.9644314868804664,
+      "grad_norm": 0.14130505919456482,
+      "learning_rate": 1.468354430379747e-06,
+      "loss": 1.8677,
+      "step": 1685
+    },
+    {
+      "epoch": 1.9655976676384839,
+      "grad_norm": 0.10629648715257645,
+      "learning_rate": 1.4242115971515769e-06,
+      "loss": 1.8466,
+      "step": 1686
+    },
+    {
+      "epoch": 1.9667638483965013,
+      "grad_norm": 0.08525505661964417,
+      "learning_rate": 1.3796627491057743e-06,
+      "loss": 1.8398,
+      "step": 1687
+    },
+    {
+      "epoch": 1.967930029154519,
+      "grad_norm": 0.09646148234605789,
+      "learning_rate": 1.3347022587268994e-06,
+      "loss": 1.9219,
+      "step": 1688
+    },
+    {
+      "epoch": 1.9690962099125364,
+      "grad_norm": 0.21308110654354095,
+      "learning_rate": 1.289324394017535e-06,
+      "loss": 1.9197,
+      "step": 1689
+    },
+    {
+      "epoch": 1.970262390670554,
+      "grad_norm": 0.10031432658433914,
+      "learning_rate": 1.2435233160621762e-06,
+      "loss": 1.9009,
+      "step": 1690
+    },
+    {
+      "epoch": 1.9714285714285715,
+      "grad_norm": 1.052268385887146,
+      "learning_rate": 1.1972930765226446e-06,
+      "loss": 1.9162,
+      "step": 1691
+    },
+    {
+      "epoch": 1.972594752186589,
+      "grad_norm": 0.1006222739815712,
+      "learning_rate": 1.1506276150627615e-06,
+      "loss": 1.8743,
+      "step": 1692
+    },
+    {
+      "epoch": 1.9737609329446064,
+      "grad_norm": 0.12004754692316055,
+      "learning_rate": 1.1035207566999474e-06,
+      "loss": 1.8388,
+      "step": 1693
+    },
+    {
+      "epoch": 1.9749271137026239,
+      "grad_norm": 0.13161960244178772,
+      "learning_rate": 1.0559662090813095e-06,
+      "loss": 1.8321,
+      "step": 1694
+    },
+    {
+      "epoch": 1.9760932944606413,
+      "grad_norm": 0.16834133863449097,
+      "learning_rate": 1.0079575596816976e-06,
+      "loss": 1.9027,
+      "step": 1695
+    },
+    {
+      "epoch": 1.9772594752186587,
+      "grad_norm": 0.09627038985490799,
+      "learning_rate": 9.594882729211088e-07,
+      "loss": 1.8538,
+      "step": 1696
+    },
+    {
+      "epoch": 1.9784256559766764,
+      "grad_norm": 0.3764210641384125,
+      "learning_rate": 9.105516871987146e-07,
+      "loss": 1.8522,
+      "step": 1697
+    },
+    {
+      "epoch": 1.9795918367346939,
+      "grad_norm": 0.08656106889247894,
+      "learning_rate": 8.611410118406889e-07,
+      "loss": 1.9142,
+      "step": 1698
+    },
+    {
+      "epoch": 1.9807580174927115,
+      "grad_norm": 0.1837455928325653,
+      "learning_rate": 8.112493239588966e-07,
+      "loss": 1.8474,
+      "step": 1699
+    },
+    {
+      "epoch": 1.981924198250729,
+      "grad_norm": 0.10193949192762375,
+      "learning_rate": 7.608695652173912e-07,
+      "loss": 1.8552,
+      "step": 1700
+    },
+    {
+      "epoch": 1.9830903790087464,
+      "grad_norm": 0.08827048540115356,
+      "learning_rate": 7.0999453850355e-07,
+      "loss": 1.8568,
+      "step": 1701
+    },
+    {
+      "epoch": 1.9842565597667638,
+      "grad_norm": 0.10714339464902878,
+      "learning_rate": 6.586169045005489e-07,
+      "loss": 1.8299,
+      "step": 1702
+    },
+    {
+      "epoch": 1.9854227405247813,
+      "grad_norm": 0.12334258109331131,
+      "learning_rate": 6.067291781577496e-07,
+      "loss": 1.856,
+      "step": 1703
+    },
+    {
+      "epoch": 1.9865889212827987,
+      "grad_norm": 0.15866361558437347,
+      "learning_rate": 5.543237250554324e-07,
+      "loss": 1.8316,
+      "step": 1704
+    },
+    {
+      "epoch": 1.9877551020408162,
+      "grad_norm": 0.10053005814552307,
+      "learning_rate": 5.013927576601672e-07,
+      "loss": 1.8829,
+      "step": 1705
+    },
+    {
+      "epoch": 1.9889212827988338,
+      "grad_norm": 0.19745850563049316,
+      "learning_rate": 4.479283314669653e-07,
+      "loss": 1.8097,
+      "step": 1706
+    },
+    {
+      "epoch": 1.9900874635568513,
+      "grad_norm": 0.4615603983402252,
+      "learning_rate": 3.939223410241981e-07,
+      "loss": 1.9048,
+      "step": 1707
+    },
+    {
+      "epoch": 1.991253644314869,
+      "grad_norm": 0.1014687791466713,
+      "learning_rate": 3.3936651583710414e-07,
+      "loss": 1.8432,
+      "step": 1708
+    },
+    {
+      "epoch": 1.9924198250728864,
+      "grad_norm": 0.09493707865476608,
+      "learning_rate": 2.8425241614553725e-07,
+      "loss": 1.8027,
+      "step": 1709
+    },
+    {
+      "epoch": 1.9935860058309038,
+      "grad_norm": 0.18891943991184235,
+      "learning_rate": 2.2857142857142855e-07,
+      "loss": 1.7992,
+      "step": 1710
+    },
+    {
+      "epoch": 1.9947521865889213,
+      "grad_norm": 0.09644436091184616,
+      "learning_rate": 1.7231476163124643e-07,
+      "loss": 1.9157,
+      "step": 1711
+    },
+    {
+      "epoch": 1.9959183673469387,
+      "grad_norm": 0.12725742161273956,
+      "learning_rate": 1.1547344110854502e-07,
+      "loss": 1.8256,
+      "step": 1712
+    },
+    {
+      "epoch": 1.9970845481049562,
+      "grad_norm": 0.10499086230993271,
+      "learning_rate": 5.803830528148578e-08,
+      "loss": 1.8941,
+      "step": 1713
+    },
+    {
+      "epoch": 1.9982507288629736,
+      "grad_norm": 0.15809349715709686,
+      "learning_rate": 0,
+      "loss": 1.8502,
+      "step": 1714
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 1714,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 2,
+  "save_steps": 86,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 6.3015412893036315e+19,
+  "train_batch_size": 4,
+  "trial_name": null,
+  "trial_params": null
+}